摘要:

Editorial The collection of papers that form this issue of Journal of many discussions. We hope that in addition this collection of papers will reach a wider community and will serve as a Materials Chemistry was presented at the First Materials state-of-the-art overview of the wide diversity of methods that Chemistry Discussion MD1 held at the Institut de Chimie de are now available to synthesise solids both purely inorganic la Matie`re Condense�e Bordeaux France on September 24–26 solids and composite solids that have a molecular or organic 1998. Papers were submitted and refereed in the usual way component. Sometimes these methods are used to synthesise and then issued as preprints to delegates in advance of the known solids by alternative routes which may give advantages meeting. At the meeting authors were given just 5 minutes to in terms of cost purity or versatility in the form of the final summarise the keypoints of their papers (10 minutes for the product.Or they may be used to synthesise entirely new review papers and 50 minutes for the overview paper by materials many of which are thermodynamically metastable C. N. R. Rao) so that most of the time could be devoted to and cannot be prepared by conventional ‘beat and heat’ routes. discussion. It was abundantly clear from the very extensive discussion The format and philosophy of the meeting was based on of some of the papers at the meeting that the synthesis the very well-established and successful Faraday Discussions procedures and the characterisation problems for some solids on topics in physical chemistry which the Royal Society of are still evolving subjects and that the whole area of synthesis Chemistry has organised for over half a century.Nevertheless and characterisation of solids is very much at the forefront of for almost all of the delegates this was a new-style meeting research into materials chemistry. It is hoped to return to and therefore something of an experiment. In contrast to these and other areas of interest and controversy in future normal practice with Faraday Discussions it was decided not meetings of this MD series. to make a written record of all the discussion comments in At the beginning of the meeting Raymond Brec paid a part because this would have delayed final publication of these personal tribute to Jean Rouxel the text of which follows papers by several months.this editorial. Participants at the meeting will have obtained their own personal benefits from listening to and taking part in the Anthony R. West Jean Etourneau J. Mater. Chem. 1999,9 xi Editorial The collection of papers that form this issue of Journal of many discussions. We hope that in addition this collection of papers will reach a wider community and will serve as a Materials Chemistry was presented at the First Materials state-of-the-art overview of the wide diversity of methods that Chemistry Discussion MD1 held at the Institut de Chimie de are now available to synthesise solids both purely inorganic la Matie`re Condense�e Bordeaux France on September 24–26 solids and composite solids that have a molecular or organic 1998. Papers were submitted and refereed in the usual way component.Sometimes these methods are used to synthesise and then issued as preprints to delegates in advance of the known solids by alternative routes which may give advantages meeting. At the meeting authors were given just 5 minutes to in terms of cost purity or versatility in the form of the final summarise the keypoints of their papers (10 minutes for the product. Or they may be used to synthesise entirely new review papers and 50 minutes for the overview paper by materials many of which are thermodynamically metastable C. N. R. Rao) so that most of the time could be devoted to and cannot be prepared by conventional ‘beat and heat’ routes. discussion. It was abundantly clear from the very extensive discussion The format and philosophy of the meeting was based on of some of the papers at the meeting that the synthesis the very well-established and successful Faraday Discussions procedures and the characterisation problems for some solids on topics in physical chemistry which the Royal Society of are still evolving subjects and that the whole area of synthesis Chemistry has organised for over half a century.Nevertheless and characterisation of solids is very much at the forefront of for almost all of the delegates this was a new-style meeting research into materials chemistry. It is hoped to return to and therefore something of an experiment. In contrast to these and other areas of interest and controversy in future normal practice with Faraday Discussions it was decided not meetings of this MD series. to make a written record of all the discussion comments in At the beginning of the meeting Raymond Brec paid a part because this would have delayed final publication of these personal tribute to Jean Rouxel the text of which follows papers by several months. this editorial. Participants at the meeting will have obtained their own personal benefits from listening to and taking part in the Anthony R. West Jean Etourneau J. Mater. Chem. 1999,

ISSN:0959-9428    

DOI:10.1039/a809185d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Solvothermal processes: a route to the stabilization of new materials† Ge�rard Demazeau Institut de Chimie de la Matie`re Condense�e de Bordeaux (I.C.M.C.B.-UPR-CNRS 9048), 87 Avenue du Dr A. Schweitzer, 33608 Pessac Cedex, France and Interface Hautes Pressions (I.H.P.), Ecole Nationale Supe�rieure de Chimie et de Physique de Bordeaux (E.N.S.C.P.B.), Avenue Pey-Berland BP 108, 33402 Talence Cedex, France Received 16th April 1998, Accepted 16th July 1998 Solvothermal processes are a powerful route for preparing K, Rb, Cs, NH4, Ca, Sr, Ba) with layered structures has been materials.Different examples are given, either in water as studied by Oka et al.1 Several hydrated vanadium(IV,V) oxides solvent or in non-aqueous solvents such as alcohols, liquid with layered structures, also found in natural minerals, have NH3 and NH2NH2.This quite new approach for materials recently been reviewed by Evans and Hughes.2,3 The structures synthesis will probably be strongly developed in the near of such AxVyOz·nH2O compounds are built from VyOz layers future provided that the role of solvents in the supercritical of VMO polyhedral frameworks with A cations and water conditions can be better understood from the chemical molecules placed in the interlayer domain. The frameworks of reactivity point of view.the VyOz layers exhibit diVerent structural types according to their compositions: V2O5, V2O6, V3O8, V4O11, etc. The most Solvothermal processes can be defined as chemical reactions developed preparation routes involve solid state reactions at or transformations in a solvent under supercritical conditions high temperatures and consequently lead to anhydrous comor near such a pressure–temperature domain.The specific pounds. In this solvothermal synthesis the precursors were physico-chemical properties of solvents in these conditions VO(OH)2 and V2O5.Vanadium oxyhydroxide VO(OH)2 was can, in particular, markedly improve the diVusion of chemi- obtained by the hydrothermal treatment of VOSO4 in NaOH cal species. solution at 150 °C for 24 h. To prepare AxV6O16·nH2O oxides, These processes have been mainly developed in the following a suspension of VO(OH)2 and/or V2O5 powders was treated scientific areas: (i) the synthesis of new materials, (ii) the under autogenous pressure at 250–280 °C for 24–48 h in development of new processes for preparing functional mate- aqueous solutions of alkali metal or alkaline earth salts rials, and (iii) the shaping of materials (crystal growth or the (0.1–0.2 M).All the resulting compounds AxV6O16·nH2O (A= preparation of fine microcrystallites). K, Rb, Cs, NH4, x=2, n=1.4–1.6) are characterized relative The purpose of this paper is to present the potentialities of to the natural mineral hewettite CaV2O16·9H2O.This structure solvothermal processes for the preparation of original mate- consists of V3O8 layers and interstitial Ca and H2O species. rials. To achieve this objective the main requirements are the optimization of the precursors and the nature of the solvent used. 1.1.2. Preparation of anhydrous oxides. Anhydrous oxides Significant examples are reported in this paper in order to can also be obtained through hydrothermal processes, the introduce the interest and the potentialities of the solvothermal main factors able to orientate the formation of anhydrous reactions for preparing new materials. Furthermore this materials being the nature of the starting materials and solvent, method is clearly highlighted with the recent preparation of the concentration of the precursors, and in particular the the graphitic form of the carbon nitride C3N4.temperature range used during the synthesis. The main advantage of the solvothermal process is to induce the preparation of metastable phases or to reduce considerably the reaction 1.Preparation of new materials through temperature and consequently the sintering eVects. solvothermal reactions Three examples can be given to illustrate the potential utility of the solvothermal method: (i) the hydrothermal synthesis of The chemical composition of the solvent must be adapted to lithium manganese oxides with the spinel structure,4 or the material to be prepared, consequently the presentation of La0.50Ba0.50MnO35 where the concentration eVect of the pre- the diVerent examples discussed in this paper will be correlated cursors appears a key factor, (ii) the formation of perovskite to this requirement.type oxides such as Pb(Mg1/3Nb2/3)O36 illustrating the eVect of temperature on the structural form stabilized, and (iii) the 1.1.Synthesis using water as solvent preparation of silicates such as Y2Si2O7,7 underlining the role 1.1.1. Hydrothermal synthesis of new hydrated oxides. of the kinetics under such conditions. Hydrothermal chemistry has been extensively developed for The preparation of lithium and sodium manganese oxides the synthesis of advanced inorganic materials which are diY- involves the hydrothermal treatment of d-MnO2 with LiOH cult to obtain by high temperature solid state reactions.In aqueous solution. While a spinel structure is observed for the particular, hydrated materials are mainly prepared through lowest LiOH concentrations (ca. 0.4 M), a monoclinic phase processes involving such chemistry, the reactivity being Li2MnO3 is formed for the highest (ca. 1 M).In the spinel increased under high pressure conditions and at moderate structure Mn ions and Mn vacancies are situated on the 16d temperatures. octahedral sites and Li+ ions on 8a tetrahedral ones. The preparation of compounds such as AxV6O16·nH2O (A= The preparation of manganese oxides with the perovskite structure can involve solid state reactions at high temperatures8,9 and sol–gel processes.10 †Basis of the presentation given at Materials Chemistry Discussion The mixed valence manganese oxide La0.50Ba0.50MnO3 No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. which exhibits giant magnetoresistance has been prepared J. Mater. Chem., 1999, 9, 15–18 15through a hydrothermal method from KMnO4, MnSO4, of diamond formation play an important role due to the small energy diVerence between the graphitic and the diamond forms La(NO3)3 and Ba(OH)2 in aqueous solution.The alkalinity and molar ratio of Mn(VII)/Mn(II) in the initial re- of carbon. A critical analysis of solvothermal diamond deposition indicates that the improvement of the diVusion of carbon action mixture, associated with the hydrothermal reaction temperature, govern the crystallization of the resulting species in a solvent under supercritical conditions could play a key role in such a process.This deposition process must be La0.50Ba0.50MnO3 material. The hydrothermal preparation of Pb(Mg1/3Nb2/3)O3 from more deeply studied in order to approach a better understanding of diamond deposition in the liquid phase in its metastable aqueous solutions of lead and magnesium nitrates and an ethanolic solution of NbCl5 leads to the pyrochlore or the P,T domain of stability.perovskite structure depending on the temperature: the pyrochlore structure is observed at temperatures lower than 600 °C, 1.1.4. Solvothermal preparation of carbon films under hydrothermal conditions. Carbon films can be used in a large range the perovskite structure is formed above 600 °C.Through solid state reactions the formation of Y2Si2O7 of applications from microelectronics to materials science.18 Chemical vapour deposition (CVD) is mainly used for prepar- requires high temperatures (900<T<1800 °C).11 Through hydrothermal processes the chemicals used were yttrium acet- ing carbon coatings but such a process is expensive and not well adapted for a large variety of substrates due to the ate and tetraethylorthosilicate. The reaction time is directly correlated to the temperature: 100 days are required at 300 °C temperature range required.Hydrothermal treatment of SiC fibers (100 MPa, whereas 11 days are necessary at 600 °C. Hydrothermal syntheses of oxides and fluorides have also 300<T<6g;C for 25 h) leads to the formation of a carbon film on the fiber surface.Such a phenomenon can be explained been developed recently by Feng et al.12 (Table 1). through the following equation: 1.1.3. Diamond crystal growth under hydrothermal con- SiC+2H2O�C+2H2+‘SiO2’ ditions: a challenge? During the last ten years several works have been focused on the role of supercritical water on Several investigations with diVerent carbides indicate that such a chemical reaction occurs.19 This hydrolysis of carbides at diamond formation.Yamada et al. have underlined the role of water in the Mg2SiO4–graphite system in diamond forma- moderate temperatures could involve a new carbon chemistry near the surface of the involved substrates. tion under high pressure and high temperature conditions (7.7 GPa, 1900 °C).13 Without water, no diamond formation is detected, while in the presence of water a large number of 2.Solvothermal preparation of new materials using diamond micro-crystals with octahedral shape are observed. non-aqueous solvents Many research works have attempted to reproduce, at moderate pressures and temperatures, the crystal growth of Although water or aqueous solutions were largely used as natural diamond (the nucleation requiring more energetic solvents in most solvothermal reactions, during the last few conditions).DiVerent research groups have tried to grow years several non-aqueous solvents have been investigated for diamond from the decomposition of minerals14 or from diVer- the preparation of diVerent materials.Two recent studies ent systems such as Ni–NaOH–C,15 SiC–H2O,16 or more carried out in our Institute will be described, one involving recently in the system Ni–C–H2O.17 In this hydrothermal the synthesis of phyllosilicate-like oxides without OH groups approach to diamond deposition, the pressure and temperature and the second showing that solvothermal synthesis using conditions are relatively low (140 MPa, 800 °C) by comparison nitriding solvents can open a new route for the preparation of with those used for industrial diamond synthesis (P5 GPa, nitrides or to the stabilization of new nitrided materials.T1200 °C). Owing to the fact that diamond is metastable under these P,T conditions, water in supercritical conditions 2.1.Stabilization of a new class of bidimensional oxides: the seems to play a specific role. ‘phyllosiloxides’ The crystal growth of diamond involving hydrothermal In order to improve the thermal stability of bidimensional conditions can be compared to the deposition of diamond as structures derived from phyllosilicates, new oxide compositions thin films using CVD processes. In the CVD route the kinetics have been designed on the basis of anionic (OH-�O2-) and cationic (M2+�M3+) coupled substitutions.Using the mica phlogopite KMg3(Si3Al )O10(OH)2 as a Table 1 List of hydrothermally synthesized complex oxides and fluorides model, a new composition, K(Mg2Al )Si4O12, has been proposed (Fig. 1). Because of the complicated composition, Hydrothermal a sol–gel process was developed to enable the use of a Compound reaction temp./°C Time/days NaxLa2/3-x/3TiO3 240 7 NaxAgyLa2/3-(x+y)/3TiO3 240 7 NaxLiyLa2/3-(x+y)/3TiO3 240 7 La1-xCaxMnO3 240 3 La1-xSrxMnO3 240 6 La1-xBaxMnO3 240 3 NaNdTi2O6 240 3 NaCeTi2O6 240 3 CaMo(W)O4 240 3 SrMo(W)O4 240 3 BaMo(W)O4 240 3 LiBaF3 140 5 KMgF3 120 8 LiYF4 220 3 KYF4 220 3 BaBeF4 220 3 BaY2F8 240 3 LiYF4:Re3+ 240 3 Fig. 1 Schematic description of the structure of a mica phlogopite (a) BaY2F8:Re3+ 240 3 and related ‘phyllosiloxide’ (b). 16 J. Mater. Chem., 1999, 9, 15–18homogeneous precursor. A gel was firstly prepared from the starting material Si(OC2H5)4, Al(OC4H9)3, Mg(OC2H5)2 and KOCH3. By using the conventional solid state treatment of such a gel at diVerent temperatures in the range 700–1200 °C, only the formation of 3D silicates (MgSiO3, Mg2SiO4, AlSiO5) is observed.20 Since such a thermal treatment at normal pressure did not lead to the formation of a layered structure isotypic with the mica phlogopite, a solvothermal treatment using 2-methoxyethanol as solvent (critical parameters Tc and Pc being: 4.7<Pc<4.9 MPa, 277<Tc<297 °C) was used.The optimized experimental conditions leading to a unique phase were: 50<P<100 MPa, 700<T<800 °C, 48<t<72 h.The Fig. 3 The diVerent possible structures claimed for C3N4. XRD pattern of this phase is isotypic with that of the micaphlogopite (Fig. 2). Physico-chemical characterization (TEM, IR andNMRspectroscopy) confirms the bidimensional 2.2.1. Solvothermal process for preparing molybdenum structure of this new oxide, the absence of OH groups and the nitrides as fine particles.In order to obtain molybdenum cationic distribution between Oh and Td sites, in agreement nitrides (Mo2N and/or MoN) free of oxygen and as finely with the structure of the mica phlogopite. TGA and XRD divided particles, the solvothermal reaction has been develstudies show an increase of the thermal stability (900–950 °C oped.By using precursors such as molybdenum oxides, the vs. 500–700 °C for the mica phlogopite parent structure).21 formation of oxynitrides as final product cannot be avoided. Such a 2D phyllosiloxide, K(Mg2Al )Si4O12, has been tested By using a precursor free of oxygen such as MoCl5 and a as an interphase in ceramic–matrix composites.22,23 nitriding solvent such as liquid NH3 or NH2NH2, Mo2N and In such a case, the use of a solvent under supercritical MoN nitrides were obtained as crystallites characterized by conditions seems to orientate the formation of a unique phase BET isotherms, TEM and SEM.24 Such finely divided molybinstead of a mixture of silicates.In addition, the in situ denum nitrides, free of oxygen and well controlled in shape, stabilization of specific chemical species such as shaping agents may open a new processing route to new catalysts.could orientate a bidimensional structure through a synthesis route comparable to that used for preparing zeolites. 2.2.2. Solvothermal preparation of a new nitride: C3N4 with In order to obtain more information concerning the process the graphitic structure.The prediction of the stability of in situ physical characterization such as for example infrared hypothetical solids through ab initio or ‘first principle’ calcuspectroscopy is necessary. lations appears as a new challenge in solid state chemistry.25 The 3D materials containing light 2p elements such as B, 2.2. Preparation of nitrides using solvothermal processes in C, N, O have received much interest from the scientific nitriding solvents (NH3 or NH2NH2) community due to the exceptional properties of the first members of this family, e.g.diamond or c-BN.26–28 The DiVerent applications of solvothermal synthesis of nitrides can compound C3N4 can be considered as an interesting member be developed: (i) to decrease the temperature of synthesis, of this family due to the high value for the calculated bulk (ii) to set up new processes in order to improve specific modulus of the b form (Bcalc.#430 GPa) isostructural with b- properties for the resulting material through shaping eVects Si3N4, which is comparable to that observed for diamond (e.g.monodispersed nano-sized particles), in particular for (Bobs.#442 GPa).29–31 During the past 10 years five diVerent catalytic applications or ceramic precursors, and (ii) to prepare structural types have been predicted through diVerent ab initio new metastable compounds.calculations: one with the bidimensional graphitic structure (g-C3N4) and four with a three-dimensional lattice (b and a structures isotypic with that of Si3N4, one of the blende type with carbon vacancies and the last derived from the high pressure form of Zn2SiO4)32 (Fig. 3). DiVerent attempts to prepare C3N4 either by CVD or PVD techniques or by high pressure routes would have led, in a few cases, to the formation of C3N4 microcrystallites corresponding to the diVerent structural forms embedded in a large amount of an amorphous phase.33,34 We have he same strategy as that developed for the synthesis of diamond c-BN through the flux assisted conversion process.35 The preparation of the graphitic form of C3N4 as a macroscopic sample appears to be an important challenge to gain access to the 3D structural varieties.In order to prepare such a graphitic form two diVerent routes were investigated, both involving solvothermal reactions in nonaqueous solvents.The first consisted of the condensation of Fig. 2 XRD patterns (Cu-Ka) of: (a) KMg2AlSi4O12 ‘phyllosilox- Fig. 4 Condensation of melamine on cyanuric chloride at medium pressure (130 MPa) and medium temperature (250 °C) using triethyl- ide’, (b) 1 M KMg3(Si3Al )O10(OH)2 phlogopite, (c) 2 M KMg3(Si3Al )O10(OH)2 phlogopite. amine as a weak nucleophilic solvent.J. Mater. Chem., 1999, 9, 15–18 17melamine (2.4.6-triamino-1.3.5-triazine) (a) and cyanuric vations. Such eVorts are currently being made in this direction in our Institute. chloride (2.4.6-trichloro-1.3.5-triazine) (b) at moderate pressure (130 MPa) and temperature (250 °C) using triethylamine (Et3N) as a weak nucleophilic solvent in order to trap the References resulting HCl formed during the reaction (Fig. 4). The second 1 Y. Oka, T. Yao and N. Yamamoto, Proceedings of the first inter- was the thermolysis of melamine C3N6H6 at high pressure national Conference on Solvothermal Reactions, ed. Organizing (2.5–3 GPa) in the temperature range 800–850 °C using hydra- Committee of the 1st International Conference on Solvothermal zine NH2NH2 as the nitriding solvent.In both cases the XRD Reactions, Takamatsu, Japan, 1994, p. 197. analysis of the resulting material after the elimination of the 2 H. T. Evans and J. M. Hughes, Am. Mineral., 1990, 75, 508. unreacted and secondary products confirms the formation of 3 H. T. Evans Jr., Can. Mineral., 1989, 27, 181. 4 Q. Feng, H. Kanoh, Y. Miyai and K. Ooi, Chem. Mater., 1995, a graphitic-type phase.The chemical analysis is in agreement 7, 1226. with the composition C3N4. The first route involves the 5 K. Yanagisawa, J. Mater. Sci. Lett., 1993, 12, 1842. production of a large amount of hydrogen compared to the 6 S. Feng, D. Wang, R. Yu and L. Na, Proceedings of the second one. International Symposium on Solvothermal–Hydrothermal The physical characterization (FTIR spectroscopy, TGA Processes, ed.Organizing Committee for Solvothermal Technology Research, 2nd RIST, Takamatsu, Japan, 1997, p. 12. and XPS analysis) of the material formed from the thermolysis 7 P. A. Trusty, I. MacLaren and C. B. Ponton, Proceedings of the of melamine under solvothermal conditions confirms the for- International Symposium on Solvothermal–Hydrothermal mation of g-C3N4.35 It is worthwhile noticing that the high Processes, ed.Organizing Committee for Solvothermal pressure and high temperature decomposition of melamine, Technology Research, 2nd RIST, Takamatsu, Japan, 1997, p. 20. involving the same experimental P,T conditions (2.5–3 GPa; 8 S. Jin, T. H. Tiefel, M. MacLormack, R. A. Fastnacht, L. Ramesh and L. H. Chen, Science, 1994, 264, 413. 800 °C) as those used for the second route (above) but in the 9 C. N. R. Rao, A. K. Cheetham and R. Mahesh, Chem. Mater., absence of the nitriding solvent NH2NH2, leads to the forma- 1996, 8, 2421. tion of only carbon graphite. Such a result underlines the 10 J. Tanaka, K. Takahashi, Y. Yajima and M. Tsukioka, Chem. chemical role of the solvent under supercritical conditions for Lett., 1982, 1847.stabilizing the C3N4 composition. 11 J. Felsche, J. Less Common Met., 1970, 21, 1. 12 S. Feng, G. Li, C. Zhao, G. Wang, D. Wang and Y. Mao, DiVerent hypotheses can explain the stabilization of carbon Proceedings of Second International Conference on Solvothermal nitride in the presence of NH2NH2 under supercritical con- Reactions, ed. Organizing Committee for Solvothermal ditions.(i) Owing to the stability of N2, an average pressure Technology Research, 2nd RIST, Takamatsu, Japan, 1997, p. 118. of 3 GPa is not suYcient for stabilizing C3N4 at the tempera- 13 H. Yamada, M. Akaishi and S. Yamaoka, International ture required for the thermal decomposition of C3N6H6 and Conference on High Pressure Science and Technology (Joint Conference AIRAPT 16-HPCJ-38, Kyoto, Japan, August 25–29 consequently NH2NH2 acts as nitriding solvent.In this case, 1997); Booklet of Abstracts, p. 35. the synthesis is governed by chemical eVects. (ii) The high 14 A. Szymanski, E. Abgarowicz, A. Baron, A. Niedbalska, pressure thermal decomposition of C3N6H6 leads to carbon R. Salacinski and J. Sentek, Diamond Relat. Mater., 1995, 4, 234. nitride formation only if a solvent under supercritical con- 15 M.Komath, K. A. Cherian, S. K. Kulkarni and A. Ray, Diamond ditions can help (through the high value of the diVusion of Relat. Mater., 1995, 4, 20. 16 R. Roy, D. Ravichandran, A. Badzian and E. Breval, Diamond chemical species) the building of the graphitic variety. Relat. Mater., 1996, 5, 973. Consequently, kinetic eVects could play an important role. 17 X. Z. Zhao, R. Roy, K. A. Cherian and A. Badzian, Nature, 1997, However, the narrow P,T domain for preparing C3N4 385, 513. suggests that the kinetics eVects could be predominant. 18 H. Tsai and D. B. Bogy, J. Vac. Sci. Technol. A, 1987, 5, 3287. 19 Y. G. Gogotsi and M. Yoshimura, Nature, 1994, 367, 628. 20 P. Reig, Thesis of BORDEAUX 1 University, No. 1226, 1995. 21 P. Reig, G. Demazeau and R. Naslain, Eur. J. Solid State Inorg. 3. Conclusion Chem., 1995, 32, 439. 22 P. Reig, G. Demazeau and R. Naslain, J. Mater. Sci., 1997, 32, Solvothermal processes open a fruitful route for improving 4189. the synthesis of well known materials such as for example 23 P. Reig, G. Demazeau and R. Naslain, J. Mater. Sci., 1997, 32, diamond, oxides and nitrides at temperatures and pressures 4195. 24 A. Wang, F. Capitain, V. Monnier, S. Matar and G. Demazeau, much lower than those used in classical solid state chemistry J. Mater. Synth. Proc., 1997, 5, 235. methods. Furthermore the chemical reactions carried out in 25 D. M. Teter, MRS Bull., January 1998, p. 22. supercritical fluid media with appropriate precursors and 26 T. H. Hall, Science, 1985, 148, 1331.solvents allow either the preparation of materials which are 27 T. H. Hall and L. A. Compton, Inorg. Chem., 1985, 4, 1213. diYcult or impossible to obtain by the ceramic route (e.g. 28 D. A. Zhogolev, O. P. Bugaets and I. A. Marushko, Zh. Strukt. Khim., 1981, 22, 46 (Inorg. Chem., 1981, 22, 33). phyllosiloxides) or the synthesis of metastable phases (e.g. 29 A. Y. Liu and M. L. Cohen, Science, 1989, 245, 841. C3N4 with graphitic structure). Another feature of the synthesis 30 J. R. Chelikowsky and S. G. Louie, Phys. Rev. B, 1984, 29, 3470. carried out under (near) supercritical conditions concerns the 31 A. Y. Liu and M. L. Cohen, Phys. Rev. B, 1990, 41, 10727. morphology of the materials which can be governed by varying 32 D. M. Teter and R.J. Hemley, Science, 1996, 271, 53. 33 B. L. Korsounskii and V. I. Pepekin, Russ. Chem. Rev., 1997, three main parameters (pressure, temperature and reaction 66, 901. time). However for controlling the chemical reaction in order 34 H. Montigaud, G. Demazeau and B. Tanguy, J. Mater. Sci., in to obtain materials with specific composition, structure, morpress. phology and properties it is absolutely necessary to understand 35 H.Montigaud, B. Tanguy, G. Demazeau, I. Alves, M. Birot and more fully the chemical reactivity of supercritical fluids. This J. Dunogues, Nature, submitted. purpose requires the evolution of the chemical reaction to be followed in situ with the help of diVerent spectroscopic obser- Paper 8/05536J 18 J. Mater. Chem., 1999, 9, 15–18 J O U R N A L O F C H E M I S T R Y Materials Feature Article Solvothermal processes: a route to the stabilization of new materials† Ge�rard Demazeau Institut de Chimie de la Matie`re Condense�e de Bordeaux (I.C.M.C.B.-UPR-CNRS 9048), 87 Avenue du Dr A.Schweitzer, 33608 Pessac Cedex, France and Interface Hautes Pressions (I.H.P.), Ecole Nationale Supe�rieure de Chimie et de Physique de Bordeaux (E.N.S.C.P.B.), Avey-Berland BP 108, 33402 Talence Cedex, France Received 16th April 1998, Accepted 16th July 1998 Solvothermal processes are a powerful route for preparing K, Rb, Cs, NH4, Ca, Sr, Ba) with layered structures has been materials. Different examples are given, either in water as studied by Oka et al.1 Several hydrated vanadium(IV,V) oxides solvent or in non-aqueous solvents such as alcohols, liquid with layered structures, also found in natural minerals, have NH3 and NH2NH2.This quite new approach for materials recently been reviewed by Evans and Hughes.2,3 The structures synthesis will probably be strongly developed in the near of such AxVyOz·nH2O compounds are built from VyOz layers future provided that the role of solvents in the supercritical of VMO polyhedral frameworks with A cations and water conditions can be better understood from the chemical molecules placed in the interlayer domain.The frameworks of reactivity point of view. the VyOz layers exhibit diVerent structural types according to their compositions: V2O5, V2O6, V3O8, V4O11, etc. The most Solvothermal processes can be defined as chemical reactions developed preparation routes involve solid state reactions at or transformations in a solvent under supercritical conditions high temperatures and consequently lead to anhydrous comor near such a pressure–temperature domain.The specific pounds. In this solvothermal synthesis the precursors were physico-chemical properties of solvents in these conditions VO(OH)2 and V2O5.Vanadium oxyhydroxide VO(OH)2 was can, in particular, markedly improve the diVusion of chemi- obtained by the hydrothermal treatment of VOSO4 in NaOH cal species. solution at 150 °C for 24 h. To prepare AxV6O16·nH2O oxides, These processes have been mainly developed in the following a suspension of VO(OH)2 and/or V2O5 powders was treated scientific areas: (i) the synthesis of new materials, (ii) the under autogenous pressure at 250–280 °C for 24–48 h in development of new processes for preparing functional mate- aqueous solutions of alkali metal or alkaline earth salts rials, and (iii) the shaping of materials (crystal growth or the (0.1–0.2 M).All the resulting compounds AxV6O16·nH2O (A= preparation of fine microcrystallites). K, Rb, Cs, NH4, x=2, n=1.4–1.6) are characterized relative The purpose of this paper is to present the potentialities of to the natural mineral hewettite CaV2O16·9H2O. This structure solvothermal processes for the preparation of original mate- consists of V3O8 layers and interstitial Ca and H2O species.rials. To achieve this objective the main requirements are the optimization of the precursors and the nature of the solvent used. 1.1.2. Preparation of anhydrous oxides. Anhydrous oxides Significant examples are reported in this paper in order to can also be obtained through hydrothermal processes, the introduce the interest and the potentialities of the solvothermal main factors able to orientate the formation of anhydrous reactions for preparing new materials. Furthermore this materials being the nature of the starting materials and solvent, method is clearly highlighted with the recent preparation of the concentration of the precursors, and in particular the the graphitic form of the carbon nitride C3N4.temperature range used during the synthesis. The main advantage of the solvothermal process is to induce the preparation of metastable phases or to reduce considerably the reaction 1.Preparation of new materials through temperature and consequently the sintering eVects. solvothermal reactions Three examples can be given to illustrate the potential utility of the solvothermal method: (i) the hydrothermal synthesis of The chemical composition of the solvent must be adapted to lithium manganese oxides with the spinel structure,4 or the material to be prepared, consequently the presentation of La0.50Ba0.50MnO35 where the concentration eVect of the pre- the diVerent examples discussed in this paper will be correlated cursors appears a key factor, (ii) the formation of perovskite to this requirement.type oxides such as Pb(Mg1/3Nb2/3)O36 illustrating the eVect of temperature on the structural form stabilized, and (iii) the 1.1.Synthesis using water as solvent preparation of silicates such as Y2Si2O7,7 underlining the role 1.1.1. Hydrothermal synthesis of new hydrated oxides. of the kinetics under such conditions. Hydrothermal chemistry has been extensively developed for The preparation of lithium and sodium manganese oxides the synthesis of advanced inorganic materials which are diY- involves the hydrothermal treatment of d-MnO2 with LiOH cult to obtain by high temperature solid state reactions.In aqueous solution. While a spinel structure is observed for the particular, hydrated materials are mainly prepared through lowest LiOH concentrations (ca. 0.4 M), a monoclinic phase processes involving such chemistry, the reactivity being Li2MnO3 is formed for the highest (ca. 1 M). In the spinel increased under high pressure conditions and at moderate structure Mn ions and Mn vacancies are situated on the 16d temperatures. octahedral sites and Li+ ions on 8a tetrahedral ones. The preparation of compounds such as AxV6O16·nH2O (A= The preparation of manganese oxides with the perovskite structure can involve solid state reactions at high temperatures8,9 and sol–gel processes.10 †Basis of the presentation given at Materials Chemistry Discussion The mixed valence manganese oxide La0.50Ba0.50MnO3 No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. which exhibits giant magnetoresistance has been prepared J. Mater. Chem., 1999, 9, 15–18 15through a hydrothermal method from KMnO4, MnSO4, of diamond formation play an important role due to the small energy diVerence between the graphitic and the diamond forms La(NO3)3 and Ba(OH)2 in aqueous solution.The alkalinity and molar ratio of Mn(VII)/Mn(II) in the initial re- of carbon. A critical analysis of solvothermal diamond deposition indicates that the improvement of the diVusion of carbon action mixture, associated with the hydrothermal reaction temperature, govern the crystallization of the resulting species in a solvent under supercritical conditions could play a key role in such a process.This deposition process must be La0.50Ba0.50MnO3 material. The hydrothermal preparation of Pb(Mg1/3Nb2/3)O3 from more deeply studied in order to approach a better understanding of diamond deposition in the liquid phase in its metastable aqueous solutions of lead and magnesium nitrates and an ethanolic solution of NbCl5 leads to the pyrochlore or the P,T domain of stability.perovskite structure depending on the temperature: the pyrochlore structure is observed at temperatures lower than 600 °C, 1.1.4. Solvothermal preparation of carbon films under hydrothermal conditions. Carbon films can be used in a large range the perovskite structure is formed above 600 °C.Through solid state reactions the formation of Y2Si2O7 of applications from microelectronics to materials science.18 Chemical vapour deposition (CVD) is mainly used for prepar- requires high temperatures (900<T<1800 °C).11 Through hydrothermal processes the chemicals used were yttrium acet- ing carbon coatings but such a process is expensive and not well adapted for a large variety of substrates due to the ate and tetraethylorthosilicate.The reaction time is directly correlated to the temperature: 100 days are required at 300 °C temperature range required. Hydrothermal treatment of SiC fibers (100 MPa, whereas 11 days are necessary at 600 °C. Hydrothermal syntheses of oxides and fluorides have also 300<T<600 °C for 25 h) leads to the formation of a carbon film on the fiber surface.Such a phenomenon can be explained been developed recently by Feng et al.12 (Table 1). through the following equation: 1.1.3. Diamond crystal growth under hydrothermal con- SiC+2H2O�C+2H2+‘SiO2’ ditions: a challenge? During the last ten years several works have been focused on the role of supercritical water on Several investigations withiVerent carbides indicate that such a chemical reaction occurs.19 This hydrolysis of carbides at diamond formation.Yamada et al. have underlined the role of water in the Mg2SiO4–graphite system in diamond forma- moderate temperatures could involve a new carbon chemistry near the surface of the involved substrates.tion under high pressure and high temperature conditions (7.7 GPa, 1900 °C).13 Without water, no diamond formation is detected, while in the presence of water a large number of 2. Solvothermal preparation of new materials using diamond micro-crystals with octahedral shape are observed. non-aqueous solvents Many research works have attempted to reproduce, at moderate pressures and temperatures, the crystal growth of Although water or aqueous solutions were largely used as natural diamond (the nucleation requiring more energetic solvents in most solvothermal reactions, during the last few conditions).DiVerent research groups have tried to grow years several non-aqueous solvents have been investigated for diamond from the decomposition of minerals14 or from diVer- the preparation of diVerent materials.Two recent studies ent systems such as Ni–NaOH–C,15 SiC–H2O,16 or more carried out in our Institute will be described, one involving recently in the system Ni–C–H2O.17 In this hydrothermal the synthesis of phyllosilicate-like oxides without OH groups approach to diamond deposition, the pressure and temperature and the second showing that solvothermal synthesis using conditions are relatively low (140 MPa, 800 °C) by comparison nitriding solvents can open a new route for the preparation of with those used for industrial diamond synthesis (P5 GPa, nitrides or to the stabilization of new nitrided materials.T1200 °C). Owing to the fact that diamond is metastable under these P,T conditions, water in supercritical conditions 2.1.Stabilization of a new class of bidimensional oxides: the seems to play a specific role. ‘phyllosiloxides’ The crystal growth of diamond involving hydrothermal In order to improve the thermal stability of bidimensional conditions can be compared to the deposition of diamond as structures derived from phyllosilicates, new oxide compositions thin films using CVD processes.In the CVD route the kinetics have been designed on the basis of anionic (OH-�O2-) and cationic (M2+�M3+) coupled substitutions. Using the mica phlogopite KMg3(Si3Al )O10(OH)2 as a Table 1 List of hydrothermally synthesized complex oxides and fluorides model, a new composition, K(Mg2Al )Si4O12, has been proposed (Fig. 1). Because of the complicated composition, Hydrothermal a sol–gel process was developed to enable the use of a Compound reaction temp./°C Time/days NaxLa2/3-x/3TiO3 240 7 NaxAgyLa2/3-(x+y)/3TiO3 240 7 NaxLiyLa2/3-(x+y)/3TiO3 240 7 La1-xCaxMnO3 240 3 La1-xSrxMnO3 240 6 La1-xBaxMnO3 240 3 NaNdTi2O6 240 3 NaCeTi2O6 240 3 CaMo(W)O4 240 3 SrMo(W)O4 240 3 BaMo(W)O4 240 3 LiBaF3 140 5 KMgF3 120 8 LiYF4 220 3 KYF4 220 3 BaBeF4 220 3 BaY2F8 240 3 LiYF4:Re3+ 240 3 Fig. 1 Schematic description of the structure of a mica phlogopite (a) BaY2F8:Re3+ 240 3 and related ‘phyllosiloxide’ (b). 16 J. Mater. Chem., 1999, 9, 15–18homogeneous precursor. A gel was firstly prepared from the starting material Si(OC2H5)4, Al(OC4H9)3, Mg(OC2H5)2 and KOCH3. By using the conventional solid state treatment of such a gel at diVerent temperatures in the range 700–1200 °C, only the formation of 3D silicates (MgSiO3, Mg2SiO4, AlSiO5) is observed.20 Since such a thermal treatment at normal pressure did not lead to the formation of a layered structure isotypic with the mica phlogopite, a solvothermal treatment using 2-methoxyethanol as solvent (critical parameters Tc and Pc being: 4.7<Pc<4.9 MPa, 277<Tc<297 °C) was used.The optimized experimental conditions leading to a unique phase were: 50<P<100 MPa, 700<T<800 °C, 48<t<72 h.The Fig. 3 The diVerent possible structures claimed for C3N4. XRD pattern of this phase is isotypic with that of the micaphlogopite (Fig. 2). Physico-chemical characterization (TEM, IR andNMRspectroscopy) confirms the bidimensional 2.2.1. Solvothermal process for preparing molybdenum structure of this new oxide, the absence of OH groups and the nitrides as fine particles.In order to obtain molybdenum cationic distribution between Oh and Td sites, in agreement nitrides (Mo2N and/or MoN) free of oxygen and as finely with the structure of the mica phlogopite. TGA and XRD divided particles, the solvothermal reaction has been develstudies show an increase of the thermal stability (900–950 °C oped.By using precursors such as molybdenum oxides, the vs. 500–700 °C for the mica phlogopite parent structure).21 formation of oxynitrides as final product cannot be avoided. Such a 2D phyllosiloxide, K(Mg2Al )Si4O12, has been tested By using a precursor free of oxygen such as MoCl5 and a as an interphase in ceramic–matrix composites.22,23 nitriding solvent such as liquid NH3 or NH2NH2, Mo2N and In such a case, the use of a solvent under supercritical MoN nitrides were obtained as crystallites characterized by conditions seems to orientate the formation of a unique phase BET isotherms, TEM and SEM.24 Such finely divided molybinstead of a mixture of silicates.In addition, the in situ denum nitrides, free of oxygen and well controlled in shape, stabilization of specific chemical species such as shaping agents may open a new processing route to new catalysts. could orientate a bidimensional structure through a synthesis route comparable to that used for preparing zeolites. 2.2.2. Solvothermal preparation of a new nitride: C3N4 with In order to obtain more information concerning the process the graphitic structure.The prediction of the stability of in situ physical characterization such as for example infrared hypothetical solids through ab initio or ‘first principle’ calcuspectroscopy is necessary. lations appears as a new challenge in solid state chemistry.25 The 3D materials containing light 2p elements such as B, 2.2. Preparation of nitrides using solvothermal processes in C, N, O have received much interest from the scientific nitriding solvents (NH3 or NH2NH2) community due to the exceptional properties of the first members of this family, e.g.diamond or c-BN.26–28 The DiVerent applications of solvothermal synthesis of nitrides can compound C3N4 can be considered as an interesting member be developed: (i) to decrease the temperature of synthesis, of this family due to the high value for the calculated bulk (ii) to set up new processes in order to improve specific modulus of the b form (Bcalc.#430 GPa) isostructural with b- properties for the resulting material through shaping eVects Si3N4, which is comparable to that observed for diamond (e.g.monodispersed nano-sized particles), in particular for (Bobs.#442 GPa).29–31 During the past 10 years five diVerent catalytic applications or ceramic precursors, and (ii) to prepare structural types have been predicted through diVerent ab initio new metastable compounds.calculations: one with the bidimensional graphitic structure (g-C3N4) and four with a three-dimensional lattice (b and a structures isotypic with that of Si3N4, one of the blende type with carbon vacancies and the last derived from the high pressure form of Zn2SiO4)32 (Fig. 3). DiVerent attempts to prepare C3N4 either by CVD or PVD techniques or by high pressure routes would have led, in a few cases, to the formation of C3N4 microcrystallites corresponding to the diVerent structural forms embedded in a large amount of an amorphous phase.33,34 We have used the same strategy as that developed for the synthesis of diamond c-BN through the flux assisted conversion process.35 The preparation of the graphitic form of C3N4 as a macroscopic sample appears to be an important challenge to gain access to the 3D structural varieties.In order to prepare such a graphitic form two diVerent routes were investigated, both involving solvothermal reactions in nonaqueous solvents.The first coted of the condensation of Fig. 2 XRD patterns (Cu-Ka) of: (a) KMg2AlSi4O12 ‘phyllosilox- Fig. 4 Condensation of melamine on cyanuric chloride at medium pressure (130 MPa) and medium temperature (250 °C) using triethyl- ide’, (b) 1 M KMg3(Si3Al )O10(OH)2 phlogopite, (c) 2 M KMg3(Si3Al )O10(OH)2 phlogopite. amine as a weak nucleophilic solvent.J. Mater. Chem., 1999, 9, 15–18 17melamine (2.4.6-triamino-1.3.5-triazine) (a) and cyanuric vations. Such eVorts are currently being made in this direction in our Institute. chloride (2.4.6-trichloro-1.3.5-triazine) (b) at moderate pressure (130 MPa) and temperature (250 °C) using triethylamine (Et3N) as a weak nucleophilic solvent in order to trap the References resulting HCl formed during the reaction (Fig. 4).The second 1 Y. Oka, T. Yao and N. Yamamoto, Proceedings of the first inter- was the thermolysis of melamine C3N6H6 at high pressure national Conference on Solvothermal Reactions, ed. Organizing (2.5–3 GPa) in the temperature range 800–850 °C using hydra- Committee of the 1st International Conference on Solvothermal zine NH2NH2 as the nitriding solvent.In both cases the XRD Reactions, Takamatsu, Japan, 1994, p. 197. analysis of the resulting material after the elimination of the 2 H. T. Evans and J. M. Hughes, Am. Mineral., 1990, 75, 508. unreacted and secondary products confirms the formation of 3 H. T. Evans Jr., Can. Mineral., 1989, 27, 181. 4 Q. Feng, H. Kanoh, Y. Miyai and K. Ooi, Chem. Mater., 1995, a graphitic-type phase.The chemical analysis is in agreement 7, 1226. with the composition C3N4. The first route involves the 5 K. Yanagisawa, J. Mater. Sci. Lett., 1993, 12, 1842. production of a large amount of hydrogen compared to the 6 S. Feng, D. Wang, R. Yu and L. Na, Proceedings of the second one. International Symposium on Solvothermal–Hydrothermal The physical characterization (FTIR spectroscopy, TGA Processes, ed.Organizing Committee for Solvothermal Technology Research, 2nd RIST, Takamatsu, Japan, 1997, p. 12. and XPS analysis) of the material formed from the thermolysis 7 P. A. Trusty, I. MacLaren and C. B. Ponton, Proceedings of the of melamine under solvothermal conditions confirms the for- International Symposium on Solvothermal–Hydrothermal mation of g-C3N4.35 It is worthwhile noticing that the high Processes, ed.Organizing Committee for Solvothermal pressure and high temperature decomposition of melamine, Technology Research, 2nd RIST, Takamatsu, Japan, 1997, p. 20. involving the same experimental P,T conditions (2.5–3 GPa; 8 S. Jin, T. H. Tiefel, M. MacLormack, R. A. Fastnacht, L. Ramesh and L. H. Chen, Science, 1994, 264, 413. 800 °C) as those used for the second route (above) but in the 9 C. N. R. Rao, A. K. Cheetham and R. Mahesh, Chem. Mater., absence of the nitriding solvent NH2NH2, leads to the forma- 1996, 8, 2421. tion of only carbon graphite. Such a result underlines the 10 J. Tanaka, K. Takahashi, Y. Yajima and M. Tsukioka, Chem. chemical role of the solvent under supercritical conditions for Lett., 1982, 1847.stabilizing the C3N4 composition. 11 J. Felsche, J. Less Common Met., 1970, 21, 1. 12 S. Feng, G. Li, C. Zhao, G. Wang, D. Wang and Y. Mao, DiVerent hypotheses can explain the stabilization of carbon Proceedings of Second International Conference on Solvothermal nitride in the presence of NH2NH2 under supercritical con- Reactions, ed. Organizing Committee for Solvothermal ditions.(i) Owing to the stability of N2, an average pressure Technology Research, 2nd RIST, Takamatsu, Japan, 1997, p. 118. of 3 GPa is not suYcient for stabilizing C3N4 at the tempera- 13 H. Yamada, M. Akaishi and S. Yamaoka, International ture required for the thermal decomposition of C3N6H6 and Conference on High Pressure Science and Technology (Joint Conference AIRAPT 16-HPCJ-38, Kyoto, Japan, August 25–29 consequently NH2NH2 acts as nitriding solvent. In this case, 1997); Booklet of Abstracts, p. 35. the synthesis is governed by chemical eVects. (ii) The high 14 A. Szymanski, E. Abgarowicz, A. Baron, A. Niedbalska, pressure thermal decomposition of C3N6H6 leads to carbon R. Salacinski and J. Sentek, Diamond Relat. Mater., 1995, 4, 234.nitride formation only if a solvent under supercritical con- 15 M. Komath, K. A. Cherian, S. K. Kulkarni and A. Ray, Diamond ditions can help (through the high value of the diVusion of Relat. Mater., 1995, 4, 20. 16 R. Roy, D. Ravichandran, A. Badzian and E. Breval, Diamond chemical species) the building of the graphitic variety. Relat. Mater., 1996, 5, 973. Consequently, kinetic eVects could play an important role. 17 X. Z. Zhao, R. Roy, K. A. Cherian and A. Badzian, Nature, 1997, However, the narrow P,T domain for preparing C3N4 385, 513. suggests that the kinetics eVects could be predominant. 18 H. Tsai and D. B. Bogy, J. Vac. Sci. Technol. A, 1987, 5, 3287. 19 Y. G. Gogotsi and M. Yoshimura, Nature, 1994, 367, 628. 20 P. Reig, Thesis of BORDEAUX 1 University, No. 1226, 1995. 21 P. Reig, G. Demazeau and R. Naslain, Eur. J. Solid State Inorg. 3. Conclusion Chem., 1995, 32, 439. 22 P. Reig, G. Demazeau and R. Naslain, J. Mater. Sci., 1997, 32, Solvothermal processes open a fruitful route for improving 4189. the synthesis of well known materials such as for example 23 P. Reig, G. Demazeau and R. Naslain, J. Mater. Sci., 1997, 32, diamond, oxides and nitrides at temperatures and pressures 4195. 24 A. Wang, F. Capitain, V. Monnier, S. Matar and G. Demazeau, much lower than those used in classical solid state chemistry J. Mater. Synth. Proc., 1997, 5, 235. methods. Furthermore the chemical reactions carried out in 25 D. M. Teter, MRS Bull., January 1998, p. 22. supercritical fluid media with appropriate precursors and 26 T. H. Hall, Science, 1985, 148, 1331. solvents allow either the preparation of materials which are 27 T. H. Hall and L. A. Compton, Inorg. Chem., 1985, 4, 1213. diYcult or impossible to obtain by the ceramic route (e.g. 28 D. A. Zhogolev, O. P. Bugaets and I. A. Marushko, Zh. Strukt. Khim., 1981, 22, 46 (Inorg. Chem., 1981, 22, 33). phyllosiloxides) or the synthesis of metastable phases (e.g. 29 A. Y. Liu and M. L. Cohen, Science, 1989, 245, 841. C3N4 with graphitic structure). Another feature of the synthesis 30 J. R. Chelikowsky and S. G. Louie, Phys. Rev. B, 1984, 29, 3470. carried out under (near) supercritical conditions concerns the 31 A. Y. Liu and M. L. Cohen, Phys. Rev. B, 1990, 41, 10727. morphology of the materials which can be governed by varying 32 D. M. Teter and R. J. Hemley, Science, 1996, 271, 53. 33 B. L. Korsounskii and V. I. Pepekin, Russ. Chem. Rev., 1997, three main parameters (pressure, temperature and reaction 66, 901. time). However for controlling the chemical reaction in order 34 H. Montigaud, G. Demazeau and B. Tanguy, J. Mater. Sci., in to obtain materials with specific composition, structure, morpress. phology and properties it is absolutely necessary to understand 35 H. Montigaud, B. Tanguy, G. Demazeau, I. Alves, M. Birot and more fully the chemical reactivity of supercritical fluids. This J. Dunogues, Nature, submitted. purpose requires the evolution of the chemical reaction to be followed in situ with the help of diVerent spectroscopic obser- Paper 8/05536J 18 J. Mater. Chem., 1999, 9, 15–18

ISSN:0959-9428    

DOI:10.1039/a805536j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Recent progress in cryochemical synthesis of oxide materials Yuri D. Tretyakov and Oleg A. Shlyakhtin Department of Chemistry, Moscow State University, 119899 Moscow, Russia Received 21st April 1998, Accepted 1st July 1998 Recently conventional cryochemical processes have been is freeze-drying of gels and suspensions obtained by (co)deposcomplemented by cryoextraction, cryoprecipitation, ition of components from an initial solution.Low-temperature cryoimpregnation and freeze casting techniques. dehydration is essential in this case to prevent pore coalescence Considerably more attention is being paid to colloidal and to keep the specific porous structure of the residue. This solution processing and the application of sol–gel pro- method is also used to process separately prepared oxide and cedures like ion exchange treatment and hydrolysis of hydroxide products and components.organometallic compounds. New preparation methods for The product of basic cryochemical processing is a fine nanopowders of various oxides and metals have been voluminous powder with submicron-sized crystallites joined developed while further treatment of cryochemical pow- into strong aggregates (Fig. 2). Owing to the high initial ders using powder engineering techniques allows the homogeneity of freeze-dried salt precursors the formation of powder grain size to vary from 0.2 to 5–6 mm. multicomponent oxide phases, especially the first members of Cryochemical methods are also suitable for preparation homologous series, normally takes place during the course of of various porous media, like porous silica or cryogels thermal decomposition, or, sometimes, a small additional heat with surface areas close to those of aerogels, and for the treatment is needed.The various applications of these powders synthesis of HTSC powders for different applications. are based on the earlier formation of poorly synthesized Fundamental and applied aspects of cryochemical compounds, the small size of the crystallites resulted in the processing are discussed. advanced sinterability of these powders to give specific and stable microstructures with a large volume of diVerent pores Traditions suitable for catalyst and support applications.Various chemical routes using low temperatures as a substantial element of the process are considered as cryochemical methods. Unlike cryogenic studies in biology and pharmaceuticals where slow cooling rates are often applied, almost all the cryochemical methods are based on the fast cooling of solutions (freezing rates over 10 K s-1) to prevent separate crystallization for most of the components involved; this is normally achieved by freezing in liquid nitrogen.Cryochemical techniques also involve the elimination of frozen water by physical (freeze-drying) or chemical means. While multicomponent oxide materials cannot be directly obtained from freezedried solution (except by suspension drying), further heat treatment is needed to convert salts into oxides. Special precautions are needed during thermal processing in order to maintain system homogeneity close to that of the initial frozen state. In spite of the large number of cryochemical techniques developed for materials synthesis and processing over the last Fig. 1 Scheme of the main processes of cryochemical synthesis. 30 years, the majority of over 500 works published in this field deal with the processing sequence (Fig. 1, solid arrows) proposed by Landsberg and Campbell in their first study on the freeze-drying synthesis of metal powders1 and further developed by Johnson and Gallagher and their co-workers for oxides.2,3 An initial single- or multicomponent true aqueous solution containing cations in the stoichiometric ratio required for the final material is sprayed into liquid nitrogen (or, less frequently, into CO2-cooled hexane) by means of a pneumatic or ultrasonic nozzle under vigorous stirring. After evaporation of liquid N2 (or filtering oV hexane) frozen microdroplets in metal trays are placed on shelves precooled to T=220–230 K and subjected to freeze-drying at P=(3–10)×10-2 mbar.The partially dehydrated salt produced by freeze-drying is normally heated at various rates to complete oxide formation.4 An important and frequently used variation of this scheme †Basis of the presentation given at Materials Chemistry Discussion No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, Fig. 2 Microstructure of oxide product of freeze-drying synthesis. France. J. Mater. Chem., 1999, 9, 19–24 19A detailed physico-chemical description of the processes amounts of free HNO3 or HCl in the system: involved in cryochemical synthesis, as well as an extended MeCln+mH2O�Me(OH)mCln-m+mHCl overview of applications, is given in the recently published monograph4 and will not be considered in the present paper.These acids form eutectic mixtures with ice, with rather low melting temperatures (190–210 K); freeze-drying of solutions We would like to focus our attention on rarely used or recently discovered cryochemical procedures, as well as on recent containing such mixtures is accompanied by low-temperature sintering (collapse) of the salt framework during drying.applications of novel and traditional cryochemical methods in materials synthesis and processing. Increasing the pH of the initial solution over threshold values results in the formation of often undesirable gels and residues.One way out of this situation concerns the application of strong complexing agents which strongly influence the equilib- Development rium conditions of residue formation. Most of these complexes Progress in freeze-drying synthesis was substantially influenced have no low-temperature eutectics and can be easily subjected by fundamental chemical studies of cryocrystallization and to freeze-drying.11,12 freeze-drying processes.It was demonstrated that transform- Another way is based on the fact that the formation of the ation of homogeneous liquid solutions of most inorganic salts residue is the final stage of hydroxopolymerization process in into freeze-dried solids cannot be considered as a physical hydrolyzed solutions far below threshold pH values: process due to the large number of chemical transformations x[Me(OH)m](n-m)+=[Me(OH)y]x(n-y)x++(m-y)xOH- involved.For instance, many salts of rare earths and transition metals demonstrated outstanding resistance to crystallization The character and rate of hydroxopolymerization is influenced when the ratio of salt to water corresponded to completed by the pH and, mostly, by the ionic strength of the solucoordination shells of the cation and anion (the so-called tion.When the ionic strength is kept low enough, further boundary of complete hydration).4 More generally, the behav- hydroxopolymerization occurs in colloidal particles until their iour of various inorganic salts during cryocrystallization has coagulation at much higher pH values.This transformation been found to depend on the coordination numbers of both may be carried out by anion exchange treatment of acid the cation and the anion and on the stability of hydrated ions solution with OH- to form an anionite resin: in solution. This dependence is especially important for multi- An-OH+NO3-�An-NO3+OH- component solutions when ions of diVerent chemical nature are present simultaneously and demonstrate diVerent crystalliz- OH-+H+�H2O ation behaviour.when the increasing pH is accompanied by reducing the total Unlike biological and pharmaceutical solutions, homogenionic strength of the solution.4 This process is equivalent to eity of frozen solutions of inorganic salts during freeze-drying the continuous conversion of the initial MeXa salt (X=Cl-, is substantially aVected by specific transformation of crystal- NO3-) into hydroxide Me(OH)a via the formation of an lohydrates at low temperatures and pressures.The evolution intermediate colloidal hydroxy salt Me(OH)aXn-a. of these non-equilibrium and internally strained hydrate forms Application of such a solution in freeze-drying synthesis has after completion of freying also diVers from conventional some additional advantages.The first deals with the reduced hydration processes.4,5 These processes occur within microcryshygroscopicity of freeze-drying products which is essential for tallites of freeze-dried salts, so that final changes in the many chlorides and nitrates.Another advantage is the possibil- homogeneity of the system are sometimes not substantial if ity to change the properties of the oxide product systematically. the formation of the liquid phase is avoided. This can be illustrated by the various crystallization tempera- The simple and attractive idea of conservation of the tures (Fig. 3) and thermal stabilities of Fe2O3 particles microstructure of the residue or gel by means of freeze-drying obtained from Fe(NO3)3 solution by anionite treatment at continuously finds new applications.Advancement of synthesis diVerent pH values.13 A similar tendency of increasing crys- techniques results in the appearance of new gel and suspension tallization temperature for greater exchange values is also preparation methods having various advantages over traobserved for aluminium nitrate where it resulted in crystalliz- ditional deposition from aqueous solutions.The stage of hydrolysis of organometallic precursors makes some of them6,7 closer to sol–gel processing while freeze-drying is believed to be crucial for attaining the final result. Thus, the remarkable softening of alumina aggregates can give the freeze-dried products of hydrolysis an advantage even over products of true Al2(SO4)3 solution freeze-drying which demonstrated poorer densification during sintering.7 The other new eYcient methods used to prepare gel precursors for freeze-drying are hydrothermal synthesis8 and electrochemical reduction of aqueous solutions.9 Non-aqueous solutions have not attracted much attention owing to the limited solubility of inorganic salts and modifi- cations needed to the freeze-drying equipment.An interesting example of their successful application in materials synthesis concerns utilization of liquid ammonia as a solvent for nitrate, acetate and perchlorates followed by freeze-drying of multicomponent solutions either at 179 K and normal pressure or at P=150 mbar without additional refrigeration.10 Nevertheless, the observed rate of target YBa2Cu3Ox phase formation was less than for freeze-dried aqueous solutions, probably owing to local melting during thermal decomposition. Fig. 3 DSC curves of heat evolution during crystallization of Fe2O3 Aqueous solutions of nitrates and chlorides of many from amorphous products obtained by freeze-drying of anionite multicharged cations like Zr4+, Ti4 +, Fe3 + are substantially processed Fe(NO3)3 solutions with pH=(1) 2.0; (2) 2.5; (3) 3.0; (4) 4.0.13 hydrolyzed, which means the appearance of significant 20 J.Mater. Chem., 1999, 9, 19–24ation of c-Al2O3 at temperatures as high as 800 °C.14 Recent because liquid phase diVusion of water from the extraction area is often the limiting stage of the complex cryoextraction analysis of anionite treatment eYciency for various multicharged cations demonstrated the possibility to use a ‘charge process.This last feature makes clear the need for intensive stirring of the solvent throughout the process. Comparison of to ionic radius’ criterion for preliminary estimates: all cations successfully used to date in this way (Zr4+, Ti4+, Fe3+, Cr3+, diVerent solvents by the sum of these factors allows them to be placed in the following order of eYciency: methanol& Al3+, Be2 +) have this ratio in the range 4.6<R<5.6 e A-1.16,17 Further extension of this range can be achieved by acetone>ethanol>ethoxyethanol>isopropanol>propanol. Another unusual but much more promising technique of the application of other processing techniques, including cationite exchange treatment which was recently developed for cryochemical processing is cryoprecipitation. Based on lowtemperature (230–250 K) treatment of frozen granules in a doped SnO2 synthesis.The cryoextraction technique has been developed as an precooled solution of the precipitant, this method has advantages over both room-temperature precipitation and cryoex- alternative to freeze-drying in order to reduce the duration of ice removal from frozen droplets—the main drawback of traction processes.Unlike the first technique, slow layer-bylayer dissolution of droplets allows greater supersaturation freeze-drying processes. Cryogranules, obtained by cryocrystallization of the initial solution, are placed into the low-tempera- values to be achieved and prevents the redistribution of components during precipitation.An additional advantage of ture thermostat with an organic solvent precooled to temperatures below the melting point of the cryogranules low-temperature processing is the smaller diVerence of solubility products and precipitation rates for chemically diVerent (230–240 K). Dissolution of ice proceeds under intensive stirring while the salt framework remains almost unchanged; components.In contrast to cryoextraction, the solubility of precursor salts in water–solvent mixtures is necessary and the resulting salt product is separated from the solvent by filtering. The normal duration of the process is 2–4 hours, desirable here to accelerate precipitation.Moreover, the ability of solvents to dissolve inorganic substances is essential in order which is 10–20 times faster than freeze-drying. Systematic studies of these processes (see ref. 4 and refs. to ensure higher concentrations of inorganic precipitants. The selection of precipitants is similar to the conventional therein) demonstrated that the eYciency of application of various organic solvents in cryoextraction is determined by coprecipitation method—hydroxides, carbonates, oxalates of alkali metals and ammonia, oxalic acid—except for some both thermodynamic and kinetic factors.The capacity of the solvent, i.e. the maximum amount of water that can be limitations for aqueous solutions, concerning their freezing temperatures.Nevertheless, aqueous ammonia remains one of absorbed by the solvent at a given temperature, is determined by the position of the boundary between the ‘solution’ and the most popular cryoprecipitating agents for many systems.4 The application of alcohol solutions of precipitants ‘ice+solution’ fields in the low-temperature part of the water(ice)–solvent phase diagram (point C, Fig. 4 and (ammonium oxalate or oxalic acid) provides more complete precipitation and a broader range of processing temperatures. Table 1); the maximum capacity corresponds to the lowest solvent content at point C. Freezing temperatures for most of The mechanism of the process varies from control by the chemical reaction at the boundary of the solid and liquid the widely used solvents are low enough to be used in these processes, so this factor is usually not the limiting one.More phases to diVusion control depending on the salt system, precipitant and solvent. essential are the diVerent diVusion coeYcients (Table 1) One of the interesting phenomena discovered during studies of cryoextraction processes is so-called ‘liquid phase dehydration’. Treatment of CoSO4 6H2O in absolute ethanol resulted in complete dissolution of the salt followed by precipitation of CoSO4 H2O some 15–120 minutes later.This phenomenon, based on the anomalous stability of some complex ions of the [Me(ROH)a(H2O)6-n]2+ series, seems to be rather general; formation of metastable solutions was observed also for sulfates of some other transition metals (Cu, Ni, Mn) and for other lower alcohols.4 Most current applications of the cryoprecipitation technique are not specific for this method, although definite advantages can be demonstrated when components of the material are poorly compatible in solution or are unsuitable for direct freeze-drying.At the same time the absence of aggregation and the low decomposition temperature of the cryoprecipitation product allowed the formation of silver metal nanoparticles16 used to produce Ag–CdO composite electrical contacts with excellent wear resistance. Even in the case of chemically compatible components, the application of cryoprecipitation led to substantial acceleration of the solid state synthesis of Fig. 4 A fragment of the T –x phase diagram of the ethanol–water system.Bi2Sr2CaCu2Ox.17 Table 1 Some properties of organic solvents relevant to their ability to be used in cryoextraction processes Equilibrium composition (C) of DiVusion coeYcient of liquid phase (wt.% of solvent) water molecules in Freezing extractant DH2O/106 cm2 s-1 Solvent temperature/K T=243 K T=233 K (T=248 K) Methanol 175.2 33.0 40.0 4.0±1.2 Acetone 178.4 71.0 87.0 2.1±0.9 Ethanol 161.0 64.8 75.5 0.33±0.11 Isopropanol 187.2 73.0 77.0 0.26±0.11 Propanol 146.0 87.0 91.0 0.21±0.11 J.Mater. Chem., 1999, 9, 19–24 21Relatively small and continuous changes of microstructure during freeze-drying and further thermal treatment of hydroxides brought about a new method of powderless processing, freeze casting,18 developed mostly for shaping porous SiO2. During this process the slurry containing hydrated silica is moulded into shape and then freeze-dried, so that powder forming stage is avoided and the net-like shape of the mould is kept throughout the process.Another non-traditional application of freeze-drying is cryoimpregnation, when soaking an oxide powder with a solution of another component is followed by freeze-drying and thermal treatment of the product. This method is especially useful for components that are poorly compatible in aqueous solution or when a minor component has to be distributed through the surface of the matrix phase.The most obvious application of this technique is supported catalyst preparation Fig. 5 Formation of the porous network during thermal decomposition of salt products of freeze-drying.or introduction of sintering aids. Along with solution-based cryochemical techniques new kinds of other materials processing methods were developed, where application of low temperatures is a necessary and sition conditions allows the evolution process to stop at the stage of framework formation. High chemical homogeneity of substantial feature of the process. One such procedure is the pulsed-plasma channel method used to prepare nanoparticles cryochemical precursors and their porous, gas permeable microstructure substantially (50–200 °C) decrease the thermal of YBa2Cu3Ox.19 Two electrodes, one made from YBa2Cu3Ox ceramic, are immersed in liquid oxygen.Electrical pulses decomposition temperature, leading, in turn, to better stability of the framework structures.Combination of these features between electrodes generate a large number of spherical 20 nm sized YBa2Cu3Ox particles quenched by the surrounding cool- with the ability to form multicomponent oxide compounds in the course of thermal decomposition makes thermal decompo- ing agent. Application of liquid oxygen as cooling medium resulted in the appearance of superconductivity with Tc>77 K sition of cryochemical precursors a useful way to obtain aggregated, but definitely detectable, nanoparticles of BaTiO322 in the as-quenched state.This method provides the unique possibility to prepare nanosized YBa2Cu3Ox particles with and LiFe5O8.23 More intensive thermal treatment of decomposition prod- undamaged crystalline structure.19 A new original method of cryoelectrophoretic deposition ucts leads to the formation of larger but still submicron (100–150 nm) particles of BaFe12O19.24 The low tendency to was recently developed for the preparation of high temperature superconductor (HTSC) coatings.Liquid nitrogen is used in grain growth during heat treatment is inherent to freeze-dried products owing to their specific microstructure—a poorly this case, not only as a cooling agent to keep particles in the superconducting state but also as an electrophoresis medium. connected network of more closely packed aggregates.Modification of this microstructure by thermal or mechanical In contrast to conventional electrophoresis in organic solvents, much higher voltages (up to 11 kV) are applied to ensure means allows the grain size in powders and ceramics to vary from 0.2–0.3 to 5–6 mm.25,26 reasonable process rates.Special precautions, like external cooling of the processing dewar and a small overpressure When the temperature of decomposition is low enough (~270 C), the formation of individual nanosized Fe2O3 par- therein, allowed boiling of liquid N2 and subsequent disturbances to be minimized.The strongly anisotropic character of ticles can be observed.27 Hydrogen reduction of these 3–5 nm aggregated spherical particles at 200 C leads to formation of HTSC particles makes possible their orientation by applying an additional weak magnetic field near the Ag electrode during ultrafine (6–18 nm) particles of metallic iron.28 As mentioned above, nanoparticles of Ag with outstanding sinterability can deposition.This orientation is maintained during subsequent heat treatment of the obtained 30–80 mm thick YBa2Cu3Ox be obtained directly by thermal decomposition of cryoprecipitation product.16 films at 910 °C and resulted in considerable enhancement of their critical current density values.20 Similar particles of SiO2 have been obtained in the course of a specially developed procedure of cryogel synthesis, considered as an alternative to aerogels used as supports for Applications catalyst coatings.The specific area of the product obtained by hydrolysis of Na2SiO3 followed by freeze-drying reached Owing to the growing interest in preparation of nanosized particles of various phases and compounds, some eVorts have 750 m2 g-1 even without special precautions during water elimination.29 Similar treatment of nickel doped alumina been made to apply cryochemical methods for these purposes. The large diVerence in specific molar volumes of a precursor hydrogel results in lower (but still comparable with aerogels) surface area values (ca. 350 m2 g-1) while the economic salt and the oxide product of its thermal decomposition causes transformation of salt crystallites due to large internal strains.eYciency of this method is proved to be higher than supercritical drying of aerogels.30 Unlike hydroxide precursors, where such a transformation is similar to continuous shrinkage of precursor particles, thermal One of the cryochemical techniques which is closest to industrial application is utilized in a study31 where a way to decomposition of many salts is accompanied by decay of the salt crystallites into a porous oxide framework of almost the prepare porous silicon on a standard Si wafer is proposed.Electrochemical etching of the substrate by HF followed by same size and shape as the initial crystallite (Fig. 5). The rate and direction of further evolution of this strained rinsing with water and freeze-drying allows the preparation of thick silicon film with porosity over 90% and good adhesion intermediate depend on various factors, especially on the following thermal treatment, causing rearrangement and sin- to the initial wafer.One of the substantial advantages of the proposed method is its full compatibility with existing Si tering of the framework elements of 10–40 nanometers in size into larger aggregated particles of 50–300 nm in size21 (Fig. 2).technologies. Similarly to pharmacy and biology, freeze-drying in mate- It is these sintered particles which are referred to as oxide products of freeze-drying synthesis in most preceding works. rials synthesis can be applied to fix not only the microstructure of the powders but also unstable or unusual chemical states At the same time, careful selection of the thermal decompo- 22 J.Mater. Chem., 1999, 9, 19–24and compounds useful in materials synthesis. A good example is Fe(OH)3, which could not be otherwise obtained in the solid state to study its crystallographic structure.32 The newly discovered binary hydroxonitrate LnCu6NO11 is formed during the freeze-drying synthesis of LnBa2Cu3Ox-based materials from nitrate precursors.33 Reducing the temperature of phase formation allowed the observation of new metastable, vacancyordered and catalytically active La1-xSrxCoO3 phases.34 Interesting correlations between the microstructure and phase transitions were observed for TiO2 where thermal evolution of the phase composition and microstructure strongly depended on the drying method of the Ti isopropoxide hydrolysis products.For the freeze-dried product, the undesirable transformation of catalytically active anatase to rutile proceeds at a temperature 170 °C higher than for the ovendried hydrolysis product. The reason for such a behaviour seems to be connected with the later achievement of the critical nucleus size of rutile in voluminous, poorly packed products of freeze-drying.35 Cryochemical synthesis was used in the preparation of new materials with colossal magnetoresistance36 and, especially, high-temperature superconductors (HTSCs).A large number (over 100 references4) of cryochemical synthesis applications over the last 10 years concern HTSC materials; some of them, Fig. 6 Major classes of oxide materials obtained by freeze-drying synthesis. where cryochemical methods were first applied to the synthesis of a specific phase or composition, are mentioned in Table 2. Along with the fast synthesis of good quality samples of new phases for studies of HTSC properties, freeze-drying synthesis freeze-drying of polymers and organic compounds and aimed is used for the preparation of HTSC powders,17,25 comat the creation of hybrid organic–inorganic composites.posites,40,46 ceramics,39,41,47,48 thick films and tapes,42–45 including materials with the highest values of HTSC The authors are grateful to Dr. Yu. G. Metlin and Prof. N. parameters.45 N. Oleynikov for their valuable comments and to V. V. Low-temperature techniques, especially freeze-drying, have Ischenko and A.L. Vinokurov for their kind help during been proved to be simple, easily accessible and useful laboramanuscript preparation. The work is supported by the tory tools for solving many problems of materials synthesis. National Program of Russia ‘Actual directions in the physics These methods are especially eVective for the preparation of of condensed matter’, Russian Foundation for Basic Research a large number of materials (Fig. 6) where chemical homogen- (Grant 96-03-33322a) and State Program ‘Universities of eity is essential for the functional properties of materials. New Russia’. approaches to better homogeneity and reproducibility of their properties can be based upon the concept of deterministic chaos.49 According to this theory, the properties of products References are considered as fundamentally irreproducible due to the 1 A.Landsberg and T. T. Campbell, J. Metals, 1965, 856. strongly non-equilibrium character of the synthesis processes 2 D. W. Johnson Jr. and F. J. Schnettler, J. Am. Ceram. Soc., 1970, and special chaos suppression algorithms should be applied 53, 440. during the development of processing techniques.Another 3 P. K. Gallagher and F. Schrey, Thermochim. Acta. 1970, 1, 465. 4 Yu. D. Tretyakov, N. N. Oleynikov and O. A. Shlyakhtin, promising direction for cryochemical methods development is Cryochemical Processing of Advanced Materials, Chapman & Hall, based on the fundamental background created in the field of London, 1997. 5 O. A.Shlyakhtin, A. B. Kulakov, Yu. V. Badun, A. M. Tesker and A. P. Mozhaev, Trans. Mater. Res. Soc. Jpn. A, 1994, 14, 7. 6 H. L. Chang and I. K. Lloyd, J. Am. Ceram. Soc., 1993, 76, 1357. Table 2 First references for freeze-drying synthesis of various super- 7 C.-T. Wang, L.-S. Lin and S. J. Yang, J. Am. Ceram. Soc., 1992, conducting materials 75, 2240. 8 K. Nakajima, S. Shimada and M.Inagaki, J. Ceram. Soc. Jpn., Year of 1995, 103, 304. Composition publication Research group 9 K. Yamashita, K. V. Ramanujachary and M. Greenblatt, Solid State Ionics, 1995, 81, 53. 1987 Johnson et al.37 YBa2Cu3Ox 10 R. W. SchaeVer, J. Macho, R. E. Salomon, G. Myer, J. E. Crow Shabatin et al.38 and P. Pernambuco-Wise, J. Supercond., 1991, 4, 365. RBa2Cu3Ox 1987 Avdeev et al.39 11 N.M. Borisova, Z. V. Golubenko, T. G. Kuz’micheva, L. P. (R=Eu, Ho, Sm) Ol’khovik and V. P. Shabatin, J. Magn. Magn. Mater., 1992, YBa2Cu3O7-x–Ag 1989 Elashkin et al.40 114, 317. YBa2Cu4O8 1991 Horn et al.41 12 L. I. Martynenko, O. A. Shlyakhtin and D. O. Charkin, Inorg. Bi2Sr2CaCu2Ox 1989 Shabatin et al.17 Mater., 1997, 33, 581. (Bi,Pb)2Sr2Ca2Cu3Ox 1990 Dou et al.42 13 F. Yu.Sharikov, Cryochemical synthesis of finely dispersed oxide Pb-free Bi2Sr2Ca2Cu3Ox 1992 M’Hamdi and powders using ion exchange, PhD thesis, Moscow State University, Lacour43 Moscow, 1991. Bulk (Bi, Pb)2Sr2Ca2Cu3Ox 1995 Lelovic et al.45 14 A. A. Vertegel, S. V. Kalinin, N. N. Oleynikov, Yu. G. Metlin and with Jc>105 A cm-2 at Yu. D. Tretyakov, J. Mater. Res., 1998, 13, 901. T=77 K 15 S.M. Kudryavtseva, A. A. Vertegel, S. V. Kalinin, N. N. Bi2Sr2CaCu2Ox–SrZrO3 1997 Pupysheva et al.46 Oleynikov, L. I. Ryabova, L. L. Meshkov, S. N. Nesterenko, Tl2Ba2Ca3Cu4Oy 1989 Zalishchansky M. N. Rumyantseva and A. M. Gaskov, J. Mater. Chem., 1997, et al.47 7, 2269. HgBa2Ca2Cu3O8+x 1995 Lee et al.48 16 Yu. D. Tretyakov, V. P. Shabatin, V. R. Sokolovskii, J. Mater. Chem., 1999, 9, 19–24 23V.G. Merkulov and L. N. Tartakovskaya, USSR Pat. no. 34 J. Kirchnerova and D. B. Hibbert, J. Mater. Sci., 1993, 28, 5800. 35 H. Izutsu, P. K. Nair and F. Mizukami, J. Mater. Chem., 1997, 4186083/31-02 from 26.01.87. 7, 855. 17 V. P. Shabatin, Venkateshvara Reddy Bommareddy, V. I. Pershin, 36 V. Yu. Pomjakushin, A. M. Balagurov, M. V. Lobanov, Yu. D. Tretyakov and P.E. Kazin, USSR Pat. no. 4749380/33 O. G. Dyachenko, E. V. Antipov, A. M. Abakumov, O. I. Lebedev from 17.10.89. and G. van Tendeloo, in 5th International Workshop on High- 18 J. Laurie, C. M. Bagnall, B. Harris, R. W. Jones, R. G. Cooke, Temperature Superconductors and Novel Materials Engineering R. S. Russell-Floyd, T. H. Wang and F. W. Hammett, J. Non- (MSU-HTSC-V ), Moscow State University, Moscow, 1998, F- Cryst.Solids, 1992, 147–148, 320. 22. 19 Yu. V. Bugoslavsky, I. G. Naumenko, G. A. Emelchenko and 37 S. M. Johnson, M. T. Gusman and D. J. RowcliVe, Adv. Ceram. A. D. Caplin, in 5th International Workshop on High-Temperature Mater., 1987, 2 (3B), 337. Superconductors and Novel Materials Engineering (MSU-HTSC- 38 V. P. Shabatin, Ya. V. Fomina, Yu.G. Metlin, Yu. D. Tretyakov, V ), Moscow State University, Moscow, 1998, W-62. V. V. Moshchalkov, A. M. Tesker and A. B. Kulakov, USSR Pat. 20 R. EggenhoVner, A. Tuccio, R. Masini, A. Diaspro, S. Leporatti no. 4244655/31-33 from 12.05.87. and R. Rolandi, Supercond. Sci. Technol., 1997, 10, 142. 39 L. Z. Avdeev, N. B. Brandt, A. V. Volkozub, A. A. Gippius, 21 V. V. Ischenko, O.A. Shlyakhtin and N. N. Oleinikov, Inorg. I. E. Graboy, A. R. Kaul, L. M. Kovba, L. N. Lykova, V. V. Mater., 1997, 33, 931. Moshchalkov, I. G. Muttik, O. V. Snigirev, Yu. D. Tretyakov, 22 J. M. McHale, P. C. McIntyre and N. V. Coppa, J. Mater. Res., R. V. Shpanchenko, V. V. Khanin and Ho Hyu Nyan, Pis’ma v 1996, 11, 1199. Zh. Eksp. Teor. Fiz., 1987, 46(Suppl.), 82. 23 E. Bermejo, C.Lacour and M. Quarton, J. Mater. Sci. Lett., 1995, 40 M. V. Elashkin, A. A. Zhukov, V. P. Shabatin, D. A. Komarkov, 14, 1346. V. V. Moshchalkov, I. G. Muttik, N. A. Samarin and 24 T. G. Kuz’micheva, L. P. Ol’khovik and V. P. Shabatin, IEEE A. Yu. Galkin, Physica C, 1989, 162–164, 1173. Trans. Magn. Mater., 1995, 31, 800. 41 J. Horn, M. Borner, H. C. Semmelhack, B. Lippold, J.Herrmann, 25 V. V. Ischenko, O. A. Shlyakhtin, A. L. Vinokurov, H. Altenburg M. Wurlitzer, M. Krotzsch, U. Boehnke, F. Schlenkrich and and N. N. Oleinikov, Physica C, 1997, 282–287, 855. Ch. Freuzel, Solid State Commun., 1991, 79, 483. 26 V. V. Ischenko, O. A. Shlyakhtin, N. N. Oleinikov, I. S. Sokolov, 42 S. X. Dou, H. K. Liu, M. H. Apperley, K. H. Song and H. Altenburg and Yu.D. Tretyakov, Dokl. Chem. (Engl. Transl.), C. C. Sorrell, Supercond. Sci. Technol., 1990, 3, 138. 1997, 356, 217. 43 E. M. M’Hamdi and C. Lacour, Ann. Chim. Fr., 1992, 17, 421. 27 E. Bermejo, T. Dantas, C. Lacour and M. Quarton, Mater. Res. 44 A. D. Nikulin, A. K. Shikov and N. E. Khlebova, IEEE Trans. Bull., 1995, 30, 645. Magn., 1994, 30, 2368. 28 E. Bermejo, T. Becue, C. Lacour and M.Quarton, Powder 45 M. Lelovic, P. Krishnaraj, N. G. Eror and U. Balachandran, Technol., 1997, 94, 29. Physica C, 1995, 242, 246. 29 G. M. Pajonk, M. Repellin-Lacroix, S. Abouarnadasse, 46 O. V. Pupysheva, O. A. Shlyakhtin, V. V. Lennikov, V. I. Putlyaev J. Chaouki and D. Klvana, J. Non-Cryst. Solids, 1990, 121, 66. and Yu. D. Tretyakov, Inorg. Mater., 1997, 33, 951. 30 D.Klvana, J. Chaouki, L. Perras and G. Belanger, in Progress in 47 M. E. Zalischansky, V. P. Shabatin and N. S. Kopelev, in Progress Catalysis, Elsevier, Amsterdam, 1992, pp. 239–246. in High Temperature Superconductivity, 1989, 22, 183 (Beijing International Conf. on High Temperature Superconductivity, 31 G. Amato, N. Brunetto and A. Parisini, Thin Solid Films, 1997, Beijing, China, Sept. 4–8, 1989). 297, 73. 48 S. R. Lee, O. A. Shlyakhtin, M.-O.Mun, M.-K. Bae and S.-I. Lee, 32 K. Shinoda, E. Matsubara, A. Muramatsu and Y.Waseda, Mater. Supercond. Sci. Technol., 1995, 8, 60. Trans. JIM, 1994, 35, 394. 49 A. J. Markworth, J. Stringer and R. W. Rollins, MRS Bull., 1995, 33 O. A. Shlyakhtin, R. V. Shpanchenko, E. V. Antipov, V. V. 7, 20. Ischenko and A. P. Mozhaev, in First International Conference on Materials Chemistry, University of Aberdeen, Scotland, UK 19–22 June 1993, Aberdeen 1993, p. 119. Paper 8/05081C 24 J. Mater. Chem., 1999, 9, 19–24 J O U R N A L O F C H E M I S T R Y Materials Feature Article Recent progress in cryochemical synthesis of oxide materials Yuri D. Tretyakov and Oleg A. Shlyakhtin Department of Chemistry, Moscow State University, 119899 Moscow, Russia Received 21st April 1998, Accepted 1st July 1998 Recently conventional cryochemical processes have been is freeze-drying of gels and suspensions obtained by (co)deposcomplemented by cryoextraction, cryoprecipitation, ition of components from an initial solution.Low-temperature cryoimpregnation and freeze casting techniques. dehydration is essential in this case to prevent pore coalescence Considerably more attention is being paid to colloidal and to keep the specific porous structure of the residue.This solution processing and the application of sol–gel pro- method is also used to process separately prepared oxide and cedures like ion exchange treatment and hydrolysis of hydroxide products and components. organometallic compounds.New preparation methods for The product of basic cryochemical processing is a fine nanopowders of various oxides and metals have been voluminous powder with submicron-sized crystallites joined developed while further treatment of cryochemical pow- into strong aggregates (Fig. 2). Owing to the high initial ders using powder engineering techniques allows the homogeneity of freeze-dried salt precursors the formation of powder grain size to vary from 0.2 to 5–6 mm.multicomponent oxide phases, especially the first members of Cryochemical methods are also suitable for preparation homologous series, normally takes place during the course of of various porous media, like porous silica or cryogels thermal decomposition, or, sometimes, a small additional heat with surface areas close to those of aerogels, and for the treatment is needed.The various applications of these powders synthesis of HTSC powders for different applications. are based on the earlier formation of poorly synthesized Fundamental and applied aspects of cryochemical compounds, the small size of the crystallites resulted in the processing are discussed. advanced sinterability of these powders to give specific and stable microstructures with a large volume of diVerent pores Traditions suitable for catalyst and support applications.Various chemical routes using low temperatures as a substantial element of the process are considered as cryochemical methods. Unlike cryogenic studies in biology and pharmaceuticals where slow cooling rates are often applied, almost all the cryochemical methods are based on the fast cooling of solutions (freezing rates over 10 K s-1) to prevent separate crystallization for most of the components involved; this is normally achieved by freezing in liquid nitrogen.Cryochemical techniques also involve the elimination of frozen water by physical (freeze-drying) or chemical means. While multicomponent oxide materials cannot be directly obtained from freezedried solution (except by suspension drying), further heat treatment is needed to convert salts into oxides. Special precautions are needed during thermal processing in order to maintain system homogeneity close to that of the initial frozen state.In spite of the large number of cryochemical techniques developed for materials synthesis and processing over the last Fig. 1 Scheme of the main processes of cryochemical synthesis. 30 years, the majority of over 500 works published in this field deal with the processing sequence (Fig. 1, solid arrows) proposed by Landsberg and Campbell in their first study on the freeze-drying synthesis of metal powders1 and further developed by Johnson and Gallagher and their co-workers for oxides.2,3 An initial single- or multicomponent true aqueous solution containing cations in the stoichiometric ratio required for the final material is sprayed into liquid nitrogen (or, less frequently, into CO2-cooled hexane) by means of a pneumatic or ultrasonic nozzle under vigorous stirring.After evaporation of liquid N2 (or filtering oV hexane) frozen microdroplets in metal trays are placed on shelves precooled to T=220–230 K and subjected to freeze-drying at P=(3–10)×10-2 mbar.The partially dehydrated salt produced by freeze-drying is normally heated at various rates to complete oxide formation.4 An important and frequently used variation of this scheme †Basis of the presentation given at Materials Chemistry Discussion No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, Fig. 2 Microstructure of oxide product of freeze-drying synthesis. France. J. Mater. Chem., 1999, 9, 19–24 19A detailed physico-chemical description of the processes amounts of free HNO3 or HCl in the system: involved in cryochemical synthesis, as well as an extended MeCln+mH2O�Me(OH)mCln-m+mHCl overview of applications, is given in the recently published monograph4 and will not be considered in the present paper.These acids form eutectic mixtures with ice, with rather low melting temperatures (190–210 K); freeze-drying of solutions We would like to focus our attention on rarely used or recently discovered cryochemical procedures, as well as on recent containing such mixtures is accompanied by low-temperature sintering (collapse) of the salt framework during drying.applications of novel and traditional cryochemical methods in materials synthesis and processing. Increasing the pH of the initial solution over threshold values results in the formation of often undesirable gels and residues. One way out of this situation concerns the application of strong complexing agents which strongly influence the equilib- Development rium conditions of residue formation.Most of these complexes Progress in freeze-drying synthesis was substantially influenced have no low-temperature eutectics and can be easily subjected by fundamental chemical studies of cryocrystallization and to freeze-drying.11,12 freeze-drying processes. It was demonstrated that transform- Another way is based on the fact that the formation of the ation of homogeneous liquid solutions of most inorganic salts residue is the final stage of hydroxopolymerization process in into freeze-dried solids cannot be considered as a physical hydrolyzed solutions far below threshold pH values: process due to the large number of chemical transformations x[Me(OH)m](n-m)+=[Me(OH)y]x(n-y)x++(m-y)xOH- involved.For instance, many salts of rare earths and transition metals demonstrated outstanding resistance to crystallization The character and rate of hydroxopolymerization is influenced when the ratio of salt to water correspoed to completed by the pH and, mostly, by the ionic strength of the solucoordination shells of the cation and anion (the so-called tion. When the ionic strength is kept low enough, further boundary of complete hydration).4 More generally, the behav- hydroxopolymerization occurs in colloidal particles until their iour of various inorganic salts during cryocrystallization has coagulation at much higher pH values.This transformation been found to depend on the coordination numbers of both may be carried out by anion exchange treatment of acid the cation and the anion and on the stability of hydrated ions solution with OH- to form an anionite resin: in solution.This dependence is especially important for multi- An-OH+NO3-�An-NO3+OH- component solutions when ions of diVerent chemical nature are present simultaneously and demonstrate diVerent crystalliz- OH-+H+�H2O ation behaviour. when the increasing pH is accompanied by reducing the total Unlike biological and pharmaceutical solutions, homogenionic strength of the solution.4 This process is equivalent to eity of frozen solutions of inorganic salts during freeze-drying the continuous conversion of the initial MeXa salt (X=Cl-, is substantially aVected by specific transformation of crystal- NO3-) into hydroxide Me(OH)a via the formation of an lohydrates at low temperatures and pressures.The evolution intermediate colloidal hydroxy salt Me(OH)aXn-a. of these non-equilibrium and internally strained hydrate forms Application of such a solution in freeze-drying synthesis has after completion of freeze-drying also diVers from conventional some additional advantages. The first deals with the reduced hydration processes.4,5 These processes occur within microcryshygroscopicity of freeze-drying products which is essential for tallites of freeze-dried salts, so that final changes in the many chlorides and nitrates.Another advantage is the possibil- homogeneity of the system are sometimes not substantial if ity to change the properties of the oxide product systematically. the formation of the liquid phase is avoided. This can be illustrated by the various crystallization tempera- The simple and attractive idea of conservation of the tures (Fig. 3) and thermal stabilities of Fe2O3 particles microstructure of the residue or gel by means of freeze-drying obtained from Fe(NO3)3 solution by anionite treatment at continuously finds new applications. Advancement of synthesis diVerent pH values.13 A similar tendency of increasing crys- techniques results in the appearance of new gel and suspension tallization temperature for greater exchange values is also preparation methods having various advantages over traobserved for aluminium nitrate where it resulted in crystalliz- ditional deposition from aqueous solutions. The stage of hydrolysis of organometallic precursors makes some of them6,7 closer to sol–gel processing while freeze-drying is believed to be crucial for attaining the final result.Thus, the remarkable softening of alumina aggregates can give the freeze-dried products of hydrolysis an advantage even over products of true Al2(SO4)3 solution freeze-drying which demonstrated poorer densification during sintering.7 The other new eYcient methods used to prepare gel precursors for freeze-drying are hydrothermal synthesis8 and electrochemical reduction of aqueous solutions.9 Non-aqueous solutions have not attracted much attention owing to the limited solubility of inorganic salts and modifi- cations needed to the freeze-drying equipment.An interesting example of their successful application in materials synthesis concerns utilization of liquid ammonia as a solvent for nitrate, acetate and perchlorates followed by freeze-drying of multicomponent solutions either at 179 K and normal pressure or at P=150 mbar without additional refrigeration.10 Nevertheless, the observed rate of target YBa2Cu3Ox phase formation was less than for freeze-dried aqueous solutions, probably owing to local melting during thermal decomposition.Fig. 3 DSC curves of heat evolution during crystallization of Fe2O3 Aqueous solutions of nitrates and chlorides of many from amorphous products obtained by freeze-drying of anionite multicharged cations like Zr4+, Ti4 +, Fe3 + are substantially processed Fe(NO3)3 solutions with pH=(1) 2.0; (2) 2.5; (3) 3.0; (4) 4.0.13 hydrolyzed, which means the appearance of significant 20 J.Mater. Chem., 1999, 9, 19–24ation of c-Al2O3 at temperatures as high as 800 °C.14 Recent because liquid phase diVusion of water from the extraction area is often the limiting stage of the complex cryoextraction analysis of anionite treatment eYciency for various multicharged cations demonstrated the possibility to use a ‘charge process.This last feature makes clear the need for intensive stirring of the solvent throughout the process. Comparison of to ionic radius’ criterion for preliminary estimates: all cations successfully used to date in this way (Zr4+, Ti4+, Fe3+, Cr3+, diVerent solvents by the sum of these factors allows them to be placed in the following order of eYciency: methanol& Al3+, Be2 +) have this ratio in the range 4.6<R<5.6 e A-1.16,17 Further extension of this range can be achieved by acetone>ethanol>ethoxyethanol>isopropanol>propanol. Another unusual but much more promising technique of the application of other processing techniques, including cationite exchange treatment which was recently developed for cryochemical processing is cryoprecipitation. Based on lowtemperature (230–250 K) treatment of frozen granules in a doped SnO2 synthesis.The cryoextraction technique has been developed as an precooled solution of the precipitant, this method has advantages over both room-temperature precipitation and cryoex- alternative to freeze-drying in order to reduce the duration of ice removal from frozen droplets—the main drawback of traction processes.Unlike the first technique, slow layer-bylayer dissolution of droplets allows greater supersaturation freeze-drying processes. Cryogranules, obtained by cryocrystallization of the initial solution, are placed into the low-tempera- values to be achieved and prevents the redistribution of components during precipitation. An additional advantage of ture thermostat with an organic solvent precooled to temperatures below the melting point of the cryogranules low-temperature processing is the smaller diVerence of solubility products and precipitation rates for chemically diVerent (230–240 K).Dissolution of ice proceeds under intensive stirring while the salt framework remains almost unchanged; components. In contrast to cryoextraction, the solubility of precursor salts in water–solvent mixtures is necessary and the resulting salt product is separated from the solvent by filtering.The normal duration of the process is 2–4 hours, desirable here to accelerate precipitation. Moreover, the ability of solvents to dissolve inorganic substances is essential in order which is 10–20 times faster than freeze-drying.Systematic studies of these processes (see ref. 4 and refs. to ensure higher concentrations of inorganic precipitants. The selection of precipitants is similar to the conventional therein) demonstrated that the eYciency of application of various organic solvents in cryoextraction is determined by coprecipitation method—hydroxides, carbonates, oxalates of alkali metals and ammonia, oxalic acid—except for some both thermodynamic and kinetic factors.The capacity of the solvent, i.e. the maximum amount of water that can be limitations for aqueous solutions, concerning their freezing temperatures. Nevertheless, aqueous ammonia remains one of absorbed by the solvent at a given temperature, is determined by the position of the boundary between the ‘solution’ and the most popular cryoprecipitating agents for many systems.4 The application of alcohol solutions of precipitants ‘ice+solution’ fields in the low-temperature part of the water(ice)–solvent phase diagram (point C, Fig. 4 and (ammonium oxalate or oxalic acid) provides more complete precipitatand a broader range of processing temperatures. Table 1); the maximum capacity corresponds to the lowest solvent content at point C.Freezing temperatures for most of The mechanism of the process varies from control by the chemical reaction at the boundary of the solid and liquid the widely used solvents are low enough to be used in these processes, so this factor is usually not the limiting one. More phases to diVusion control depending on the salt system, precipitant and solvent.essential are the diVerent diVusion coeYcients (Table 1) One of the interesting phenomena discovered during studies of cryoextraction processes is so-called ‘liquid phase dehydration’. Treatment of CoSO4 6H2O in absolute ethanol resulted in complete dissolution of the salt followed by precipitation of CoSO4 H2O some 15–120 minutes later. This phenomenon, based on the anomalous stability of some complex ions of the [Me(ROH)a(H2O)6-n]2+ series, seems to be rather general; formation of metastable solutions was observed also for sulfates of some other transition metals (Cu, Ni, Mn) and for other lower alcohols.4 Most current applications of the cryoprecipitation technique are not specific for this method, although definite advantages can be demonstrated when components of the material are poorly compatible in solution or are unsuitable for direct freeze-drying.At the same time the absence of aggregation and the low decomposition temperature of the cryoprecipitation product allowed the formation of silver metal nanoparticles16 used to produce Ag–CdO composite electrical contacts with excellent wear resistance. Even in the case of chemically compatible components, the application of cryoprecipitation led to substantial acceleration of the solid state synthesis of Fig. 4 A fragment of the T –x phase diagram of the ethanol–water system. Bi2Sr2CaCu2Ox.17 Table 1 Some properties of organic solvents relevant to their ability to be used in cryoextraction processes Equilibrium composition (C) of DiVusion coeYcient of liquid phase (wt.% of solvent) water molecules in Freezing extractant DH2O/106 cm2 s-1 Solvent temperature/K T=243 K T=233 K (T=248 K) Methanol 175.2 33.0 40.0 4.0±1.2 Acetone 178.4 71.0 87.0 2.1±0.9 Ethanol 161.0 64.8 75.5 0.33±0.11 Isopropanol 187.2 73.0 77.0 0.26±0.11 Propanol 146.0 87.0 91.0 0.21±0.11 J.Mater. Chem., 1999, 9, 19–24 21Relatively small and continuous changes of microstructure during freeze-drying and further thermal treatment of hydroxides brought about a new method of powderless processing, freeze casting,18 developed mostly for shaping porous SiO2.During this process the slurry containing hydrated silica is moulded into shape and then freeze-dried, so that powder forming stage is avoided and the net-like shape of the mould is kept throughout the process.Another non-traditional application of freeze-drying is cryoimpregnation, when soaking an oxide powder with a solution of another component is followed by freeze-drying and thermal treatment of the product. This method is especially useful for components that are poorly compatible in aqueous solution or when a minor component has to be distributed through the surface of the matrix phase.The most obvious application of this technique is supported catalyst preparation Fig. 5 Formation of the porous network during thermal decomposition of salt products of freeze-drying. or introduction of sintering aids. Along with solution-based cryochemical techniques new kinds of other materials processing methods were developed, where application of low temperatures is a necessary and sition conditions allows the evolution process to stop at the stage of framework formation.High chemical homogeneity of substantial feature of the process. One such procedure is the pulsed-plasma channel method used to prepare nanoparticles cryochemical precursors and their porous, gas permeable microstructure substantially (50–200 °C) decrease the thermal of YBa2Cu3Ox.19 Two electrodes, one made from YBa2Cu3Ox ceramic, are immersed in liquid oxygen.Electrical pulses decomposition temperature, leading, in turn, to better stability of the framework structures. Combination of these features between electrodes generate a large number of spherical 20 nm sized YBa2Cu3Ox particles quenched by the surrounding cool- with the ability to form multicomponent oxide compounds in the course of thermal decomposition makes thermal decompo- ing agent.Application of liquid oxygen as cooling medium resulted in the appearance of superconductivity with Tc>77 K sition of cryochemical precursors a useful way to obtain aggregated, but definitely detectable, nanoparticles of BaTiO322 in the as-quenched state.This method provides the unique possibility to prepare nanosized YBa2Cu3Ox particles with and LiFe5O8.23 More intensive thermal treatment of decomposition prod- undamaged crystalline structure.19 A new original method of cryoelectrophoretic deposition ucts leads to the formation of larger but still submicron (100–150 nm) particles of BaFe12O19.24 The low tendency to was recently developed for the preparation of high temperature superconductor (HTSC) coatings.Liquid nitrogen is used in grain growth during heat treatment is inherent to freeze-dried products owing to their specific microstructure—a poorly this case, not only as a cooling agent to keep particles in the superconducting state but also as an electrophoresis medium. connected network of more closely packed aggregates.Modification of this microstructure by thermal or mechanical In contrast to conventional electrophoresis in organic solvents, much higher voltages (up to 11 kV) are applied to ensure means allows the grain size in powders and ceramics to vary from 0.2–0.3 to 5–6 mm.25,26 reasonable process rates. Special precautions, like external cooling of the processing dewar and a small overpressure When the temperature of decomposition is low enough (~270 C), the formation of individual nanosized Fe2O3 par- therein, allowed boiling of liquid N2 and subsequent disturbances to be minimized. The strongly anisotropic character of ticles can be observed.27 Hydrogen reduction of these 3–5 nm aggregated spherical particles at 200 C leads to formation of HTSC particles makes possible their orientation by applying an additional weak magnetic field near the Ag electrode during ultrafine (6–18 nm) particles of metallic iron.28 As mentioned above, nanoparticles of Ag with outstanding sinterability can deposition.This orientation is maintained during subsequent heat treatment of the obtained 30–80 mm thick YBa2Cu3Ox be obtained directly by thermal decomposition of cryoprecipitation product.16 films at 910 °C and resulted in considerable enhancement of their critical current density values.20 Similar particles of SiO2 have been obtained in the course of a specially developed procedure of cryogel synthesis, considered as an alternative to aerogels used as supports for Applications catalyst coatings.The specific area of the product obtained by hydrolysis of Na2SiO3 followed by freeze-drying reached Owing to the growing interest in preparation of nanosized particles of various phases and compounds, some eVorts have 750 m2 g-1 even without special precautions during water elimination.29 Similar treatment of nickel doped alumina been made to apply cryochemical methods for these purposes.The large diVerence in specific molar volumes of a precursor hydrogel results in lower (but still comparable with aerogels) surface area values (ca. 350 m2 g-1) while the economic salt and the oxide product of its thermal decomposition causes transformation of salt crystallites due to large internal strains. eYciency of this method is proved to be higher than supercritical drying of aerogels.30 Unlike hydroxide precursors, where such a transformation is similar to continuous shrinkage of precursor particles, thermal One of the cryochemical techniques which is closest to industrial application is utilized in a study31 where a way to decomposition of many salts is accompanied by decay of the salt crystallites into a porous oxide framework of almost the prepare porous silicon on a standard Si wafer is proposed.Electrochemical etching of the substrate by HF followed by same size and shape as the initial crystallite (Fig. 5). The rate and direction of further evolution of this strained rinsing with water and freeze-drying allows the preparation of thick silicon film with porosity over 90% and good adhesion intermediate depend on various factors, especially on the following thermal treatment, causing rearrangement and sin- to the initial wafer. One of the substantial advantages of the proposed method is its full compatibility with existing Si tering of the framework elements of 10–40 nanometers in size into larger aggregated particles of 50–300 nm in size21 (Fig. 2). technologies. Similarly to pharmacy and biology, freeze-drying in mate- It is these sintered particles which are referred to as oxide products of freeze-drying synthesis in most preceding works.rials synthesis can be applied to fix not only the microstructure of the powders but also unstable or unusual chemical states At the same time, careful selection of the thermal decompo- 22 J. Mater. Chem., 1999, 9, 19–24and compounds useful in materials synthesis.A good example is Fe(OH)3, which could not be otherwise obtained in the solid state to study its crystallographic structure.32 The newly discovered binary hydroxonitrate LnCu6NO11 is formed during the freeze-drying synthesis of LnBa2Cu3Ox-based materials from nitrate precursors.33 Reducing the temperature of phase formation allowed the observation of new metastable, vacancyordered and catalytically active La1-xSrxCoO3 phases.34 Interesting correlations between the microstructure and phase transitions were observed for TiO2 where thermal evolution of the phase composition and microstructure strongly depended on the drying method of the Ti isopropoxide hydrolysis products.For the freeze-dried product, the undesirable transformation of catalytically active anatase to rutile proceeds at a temperature 170 °C higher than for the ovendried hydrolysis product. The reason for such a behaviour seems to be connected with the later achievement of the critical nucleus size of rutile in voluminous, poorly packed products of freeze-drying.35 Cryochemical synthesis was used in the preparation of new materials with colossal magnetoresistance36 and, especially, high-temperature superconductors (HTSCs).A large number (over 100 references4) of cryochemical synthesis applications over the last 10 years concern HTSC materials; some of them, Fig. 6 Major classes of oxide materials obtained by freeze-drying synthesis. where cryochemical methods were first applied to the synthesis of a specific phase or composition, are mentioned in Table 2.Along with the fast synthesis of good quality samples of new phases for studies of HTSC properties, freeze-drying synthesis freeze-drying of polymers and organic compounds and aimed is used for the preparation of HTSC powders,17,25 comat the creation of hybrid organic–inorganic composites. posites,40,46 ceramics,39,41,47,48 thick films and tapes,42–45 including materials with the highest values of HTSC The authors are grateful to Dr.Yu. G. Metlin and Prof. N. parameters.45 N. Oleynikov for their valuable comments and to V. V. Low-temperature techniques, especially freeze-drying, have Ischenko and A. L. Vinokurov for their kind help during been proved to be simple, easily accessible and useful laboramanuscript preparation. The work is supported by the tory tools for solving many problems of materials synthesis.National Program of Russia ‘Actual directions in the physics These methods are especially eVective for the preparation of of condensed matter’, Russian Foundation for Basic Research a large number of materials (Fig. 6) where chemical homogen- (Grant 96-03-33322a) and State Program ‘Universities of eity is essential for the functional properties of materials.New Russia’. approaches to better homogeneity and reproducibility of their properties can be based upon the concept of deterministic chaos.49 According to this theory, the properties of products References are considered as fundamentally irreproducible due to the 1 A. Landsberg and T. T. Campbell, J. Metals, 1965, 856. strongly non-equilibrium character of the synthesis processes 2 D.W. Johnson Jr. and F. J. Schnettler, J. Am. Ceram. Soc., 1970, and special chaos suppression algorithms should be applied 53, 440. during the development of processing techniques. Another 3 P. K. Gallagher and F. Schrey, Thermochim. Acta. 1970, 1, 465. 4 Yu. D. Tretyakov, N. N. Oleynikov and O. A. Shlyakhtin, promising direction for cryochemical methods development is Cryochemical Processing of Advanced Materials, Chapman & Hall, based on the fundamental background created in the field of London, 1997. 5 O. A. Shlyakhtin, A. B. Kulakov, Yu. V. Badun, A. M. Tesker and A. P. Mozhaev, Trans. Mater. Res. Soc. Jpn. A, 1994, 14, 7. 6 H. L. Chang and I. K. Lloyd, J. Am. Ceram. Soc., 1993, 76, 1357.Table 2 First references for freeze-drying synthesis of various super- 7 C.-T. Wang, L.-S. Lin and S. J. Yang, J. Am. Ceram. Soc., 1992, conducting materials 75, 2240. 8 K. Nakajima, S. Shimada and M. Inagaki, J. Ceram. Soc. Jpn., Year of 1995, 103, 304. Composition publication Research group 9 K. Yamashita, K. V. Ramanujachary and M. Greenblatt, Solid State Ionics, 1995, 81, 53. 1987 Johnson et al.37 YBa2Cu3Ox 10 R. W. SchaeVer, J. Macho, R. E. Salomon, G. Myer, J. E. Crow Shabatin et al.38 and P. Pernambuco-Wise, J. Supercond., 1991, 4, 365. RBa2Cu3Ox 1987 Avdeev et al.39 11 N. M. Borisova, Z. V. Golubenko, T. G. Kuz’micheva, L. P. (R=Eu, Ho, Sm) Ol’khovik and V. P. Shabatin, J. Magn. Magn. Mater., 1992, YBa2Cu3O7-x–Ag 1989 Elashkin et al.40 114, 317.YBa2Cu4O8 1991 Horn et al.41 12 L. I. Martynenko, O. A. Shlyakhtin and D. O. Charkin, Inorg. Bi2Sr2CaCu2Ox 1989 Shabatin et al.17 Mater., 1997, 33, 581. (Bi,Pb)2Sr2Ca2Cu3Ox 1990 Dou et al.42 13 F. Yu. Sharikov, Cryochemical synthesis of finely dispersed oxide Pb-free Bi2Sr2Ca2Cu3Ox 1992 M’Hamdi and powders using ion exchange, PhD thesis, Moscow State University, Lacour43 Moscow, 1991.Bulk (Bi, Pb)2Sr2Ca2Cu3Ox 1995 Lelovic et al.45 14 A. A. Vertegel, S. V. Kalinin, N. N. Oleynikov, Yu. G. Metlin and with Jc>105 A cm-2 at Yu. D. Tretyakov, J. Mater. Res., 1998, 13, 901. T=77 K 15 S. M. Kudryavtseva, A. A. Vertegel, S. V. Kalinin, N. N. Bi2Sr2CaCu2Ox–SrZrO3 1997 Pupysheva et al.46 Oleynikov, L. I. Ryabova, L. L. Meshkov, S. N. Nesterenko, Tl2Ba2Ca3Cu4Oy 1989 Zalishchansky M.N. Rumyantseva and A. M. Gaskov, J. Mater. Chem., 1997, et al.47 7, 2269. HgBa2Ca2Cu3O8+x 1995 Lee et al.48 16 Yu. D. Tretyakov, V. P. Shabatin, V. R. Sokolovskii, J. Mater. Chem., 1999, 9, 19–24 23V. G. Merkulov and L. N. Tartakovskaya, USSR Pat. no. 34 J. Kirchnerova and D. B. Hibbert, J. Mater. Sci., 1993, 28, 5800. 35 H. Izutsu, P. K. Nair and F. Mizukami, J.Mater. Chem., 1997, 4186083/31-02 from 26.01.87. 7, 855. 17 V. P. Shabatin, Venkateshvara Reddy Bommareddy, V. I. Pershin, 36 V. Yu. Pomjakushin, A. M. Balagurov, M. V. Lobanov, Yu. D. Tretyakov and P. E. Kazin, USSR Pat. no. 4749380/33 O. G. Dyachenko, E. V. Antipov, A. M. Abakumov, O. I. Lebedev from 17.10.89. and G. van Tendeloo, in 5th International Workshop on High- 18 J.Laurie, C. M. Bagnall, B. Harris, R. W. Jones, R. G. Cooke, Temperature Superconductors and Novel Materials Engineering R. S. Russell-Floyd, T. H. Wang and F. W. Hammett, J. Non- (MSU-HTSC-V ), Moscow State University, Moscow, 1998, F- Cryst. Solids, 1992, 147–148, 320. 22. 19 Yu. V. Bugoslavsky, I. G. Naumenko, G. A. Emelchenko and 37 S. M. Johnson, M. T. Gusman and D. J.RowcliVe, Adv. Ceram. A. D. Caplin, in 5th International Workshop on High-Temperature Mater., 1987, 2 (3B), 337. Superconductors and Novel Materials Engineering (MSU-HTSC- 38 V. P. Shabatin, Ya. V. Fomina, Yu. G. Metlin, Yu. D. Tretyakov, V ), Moscow State University, Moscow, 1998, W-62. V. V. Moshchalkov, A. M. Tesker and A. B. Kulakov, USSR Pat. 20 R. EggenhoVner, A. Tuccio, R. Masini, A. Diaspro, S. Leporatti no. 4244655/31-33 from 12.05.87. and R. Rolandi, Supercond. Sci. Technol., 1997, 10, 142. 39 L. Z. Avdeev, N. B. Brandt, A. V. Volkozub, A. A. Gippius, 21 V. V. Ischenko, O. A. Shlyakhtin and N. N. Oleinikov, Inorg. I. E. Graboy, A. R. Kaul, L. M. Kovba, L. N. Lykova, V. V. Mater., 1997, 33, 931. Moshchalkov, I. G. Muttik, O. V. Snigirev, Yu. D. Tretyakov, 22 J. M. McHale, P. C. McIntyre and N. V. Coppa, J. Mater. Res., R. V. Shpanchenko, V. V. Khanin and Ho Hyu Nyan, Pis’ma v 1996, 11, 1199. Zh. Eksp. Teor. Fiz., 1987, 46(Suppl.), 82. 23 E. Bermejo, C. Lacour and M. Quarton, J. Mater. Sci. Lett., 1995, 40 M. V. Elashkin, A. A. Zhukov, V. P. Shabatin, D. A. Komarkov, 14, 1346. V. V. Moshchalkov, I. G. Muttik, N. A. Samarin and 24 T. G. Kuz’micheva, L. P. Ol’khovik and V. P. Shabatin, IEEE A. Yu. Galkin, Physica C, 1989, 162–164, 1173. Trans. Magn. Mater., 1995, 31, 800. 41 J. Horn, M. Borner, H. C. Semmelhack, B. Lippold, J. Herrmann, 25 V. V. Ischenko, O. A. Shlyakhtin, A. L. Vinokurov, H. Altenburg M. Wurlitzer, M. Krotzsch, U. Boehnke, F. Schlenkrich and and N. N. Oleinikov, Physica C, 1997, 282–287, 855. Ch. Freuzel, Solid State Commun., 1991, 79, 483. 26 V. V. Ischenko, O. A. Shlyakhtin, N. N. Oleinikov, I. S. Sokolov, 42 S. X. Dou, H. K. Liu, M. H. Apperley, K. H. Song and H. Altenburg and Yu. D. Tretyakov, Dokl. Chem. (Engl. Transl.), C. C. Sorrell, Supercond. Sci. Technol., 1990, 3, 138. 1997, 356, 217. 43 E. M. M’Hamdi and C. Lacour, Ann. Chim. Fr., 1992, 17, 421. 27 E. Bermejo, T. Dantas, C. Lacour and M. Quarton, Mater. Res. 44 A. D. Nikulin, A. K. Shikov and N. E. Khlebova, IEEE Trans. Bull., 1995, 30, 645. Magn., 1994, 30, 2368. 28 E. Bermejo, T. Becue, C. Lacour and M. Quarton, Powder 45 M. Lelovic, P. Krishnaraj, N. G. Eror and U. Balachandran, Technol., 1997, 94, 29. Physica C, 1995, 242, 246. 29 G. M. Pajonk, M. Repellin-Lacroix, S. Abouarnadasse, 46 O. V. Pupysheva, O. A. Shlyakhtin, V. V. Lennikov, V. I. Putlyaev J. Chaouki and D. Klvana, J. Non-Cryst. Solids, 1990, 121, 66. and Yu. D. Tretyakov, Inorg. Mater., 1997, 33, 951. 30 D. Klvana, J. Chaouki, L. Perras and G. Belanger, in Progress in 47 M. E. Zalischansky, V. P. Shabatin and N. S. Kopelev, in Progress Catalysis, Elsevier, Amsterdam, 1992, pp. 239–246. in High Temperature Superconductivity, 1989, 22, 183 (Beijing International Conf. on High Temperature Superconductivity, 31 G. Amato, N. Brunetto and A. Parisini, Thin Solid Films, 1997, Beijing, China, Sept. 4–8, 1989). 297, 73. 48 S. R. Lee, O. A. Shlyakhtin, M.-O.Mun, M.-K. Bae and S.-I. Lee, 32 K. Shinoda, E. Matsubara, A. Muramatsu and Y.Waseda, Mater. Supercond. Sci. Technol., 1995, 8, 60. Trans. JIM, 1994, 35, 394. 49 A. J. Markworth, J. Stringer and R. W. Rollins, MRS Bull., 1995, 33 O. A. Shlyakhtin, R. V. Shpanchenko, E. V. Antipov, V. V. 7, 20. Ischenko and A. P. Mozhaev, in First International Conference on Materials Chemistry, University of Aberdeen, Scotland, UK 19–22 June 1993, Aberdeen 1993, p. 119. Paper 8/05081C 24 J. Mater. Chem., 1999, 9, 19–24

ISSN:0959-9428    

DOI:10.1039/a805081c

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Relevant examples of intercalation–deintercalation processes in solid state chemistry: application to oxides† J-C. Grenier,*a J-M. Bassat,a J-P. Doumerc,a J. Etourneau,a Z. Fang,b L. Fournes,a S. Petit,a M. Poucharda and A. Wattiauxa aInstitut de la Chimie de la Matie`re Condense�e de Bordeaux, C.N.R.S., Av. du Dr. A. Schweitzer, 33 608 Pessac Cedex, France.E-mail: grenier@chimsol.icmcb.u-bordeaux.fr. bInstitute of Physics, the Chinese Academy of Sciences, P.O. Box 603, 100080 Beijing, China Received 4th May 1998, Accepted 6th July 1998 Four ‘chimie douce’ synthetic methods, to which much (a) Electrochemical oxidation attention has been devoted in the last years, are described. In 1990, we discovered that using the electrochemical potential The electrochemical potential is shown to be efficient for as a ‘driving force’, it was possible to intercalate oxygen into intercalating as well as for deintercalating oxygen atoms in oxides.Gaseous NO2 is able to destroy in situ oxide networks, at room temperature, and correspondingly to ammonium ions either in a soft way or violently, leading overoxidize oxides up to unexpected oxygen stoichiometries.2 to the formation of tunneled or lamellar structures or to Basically this reaction can be achieved thanks to the high oxide nanoparticles.Borohydride solutions appear very oxygen activity resulting from an applied overpotential g on attractive for the synthesis of low valent oxides at room an oxide ceramic electrode according to the reaction: temperature.Typical examples of these methods are given 4OH-�O2(g)(+2 H2O+4e- and the most important features of these reactions are discussed. (Eth0=0.303 V vs. Hg/HgO, at pH=14) (1) Using the Nernst equation, one can estimate this oxygen Introduction activity to be equivalent, at room temperature for a few hundred mV of overpotential, to a very high oxygen pressure.Until the 1960s, most solid state compounds were prepared Indeed, such activity can induce the formation of oxide films using high temperature treatments of starting materials that on metal electrodes. This has been recently used for growing were first ground and pressed. Then new preparative methods tailored thick films of oxides such as Al2O3, ZnO, CuO, etc.,3 were investigated at mild temperatures with the aim of synthesbut they are usually passivating.However, by replacing a izing new compounds, generally unstable at high temperatures: metallic electrode by an oxide and provided that a diVusion that was the beginning of the ‘chimie douce’ era. These methods path exists within the oxide matrix, oxygen atoms have been can be classified in diVerent ways according to the following shown to migrate easily.features: all of them involve ‘precursors’ that must be tailored The most striking result was obtained in 1990 for La2CuO4: with respect to the desired final material; the reaction temperathis insulating antiferromagnetic compound was transformed ture ranges between room temperature and nearly 300 °C, a within a few hours into a metallic and superconducting oxide lower temperature leading usually to poorer crystallization; with TC=44 K, the highest transition temperature ever from a structural viewpoint, they can be either topotactic or obtained for this phase.4 reconstructive; and finally, such reactions may or not imply redox processes.Experimental Nowadays, the largest scale technical applications of this ‘soft solid state chemistry’ are undoubtedly concerned with The electrochemical system used for these experiments consists batteries.The reactions are typically intercalation–deintercal- of a power supply (a potentiostat) and an electrochemical cell ation of either cationic (Li+, H+) or neutral species (Li or of three-electrode type, using alkaline solution, 1 M NaOH or H2) at room temperature; the former involve redox processes KOH, as electrolyte.The electrolyte concentration of 1 M and hence mixed valence oxides [e.g. MnO2-x(OH)x, LixNiO2, (pH=14) is considered as the optimal one. Fig. 1 shows a etc.], the latter deal with alloys such as LaNi5, metals (Al ) or scheme of the electrochemical cell, the major components of graphite.1 which are the three electrodes.This paper reports on some methods recently developed for The working electrodes are usually ceramic pellets of the preparing oxides by either chemical or electrochemical oxi- materials to be oxidized: their relative densities are around dation or reduction of an appropriate precursor. The common 80%, which seems to be the optimal density allowing good feature of these methods is their eYciency even at room permeation of the electrolyte into the ceramic.A rotating temperature. Descriptions and relevant examples of the follow- working electrode is preferred to a fixed one as it produces ing four methods are given: (a) electrochemical oxidation better convection between the electrolyte in the ceramic elecresulting in oxygen intercalation into oxide networks; (b) dein- trode and in the bulk solution.It also prevents evolved gas tercalation of oxygen using electrochemical reduction; bubbles (O2), if any, from accumulating on the electrode (c) chemical oxidation using NO2 gas; and (d) chemical surface. A stable and ohmic contact between the ceramic pellet reduction using borohydride solutions. and the metallic support (made of brass) which is connected with the external circuit, is required for the electrochemical process and meaningful data measurements.The assembly is achieved thanks to a gold foil and silver paste, the whole being †Basis of the presentation given at Materials Chemistry Discussion embedded in a resin. No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France.The reference electrode is a mercury–mercury oxide electrode J. Mater. Chem., 1999, 9, 25–33 25polarization begins (C1, E<300 mV) assigned to the charging of the double layer on the interface between the electrolyte and the electrode surface.5 Then a wave-shaped part (O1, E= 300–450 mV) with the half-wave potential about E1/2= 330 mV: this is ascribed to the oxidation of the electrode material according to reaction (2).In this case it corresponds formally to the transformation from Co3+ to Co4+. At higher potentials the current intensity increases again and the evolution of gaseous oxygen takes place in the third region (O2, E>450 mV) according to reaction (1). Then two kinds of experiments can be performed. (i) Potentiostatic oxidation, using a steady potential set in the oxidation plateau, depending on the starting material (e.g. 400 mV for Sr2Co2O5). The oxidation reaction is followed by measuring the time dependence of the rest potential: the open Fig. 1 The electrochemical cell with three electrodes for experiments at 25 °C–90 °C. circuit voltage, EOCV, measured after 15 min, increases with the polarization time.Actually it can be directly correlated to the chemical composition, i.e. to the oxygen stoichiometry (Hg/HgO) that has excellent stability in alkaline solutions. Its (3-y) or to the amount of Co2+/Co3+/Co4+ (Fig. 3). It standard potential (at aOH-=1 and T=298 K) is 0.098 V shows a logarithmic dependence on the oxygen composition, relative to SHE. All the potentials quoted in this work are which may be explained in a similar way as for the intercalation referred to this electrode.of lithium ions into CoO2,MnO2 or other oxides on the basis of The counter electrode is designed to have a large surface area. The materials usually used are good electronic conduc- the Armand equation.6 The end of the polarization reaction tors, such as gold or platinum foil, or a piece of glassy carbon.is reflected in a drastic increase of the rest potential. (ii) Galvanostatic experiments, which are more convenient for Electrochemical aspects controlling the reaction kinetics, have been used by several groups. As for the lithium intercalation into dichalcogenides Various compounds AxMzOy whose structures derive from or oxides, using small current intensities (a few mA), potal perovskite, have been oxidized: they belong essentially to the steps are observed.Such a feature characterizes the formation so-called AnMnO3n-1 and An+1MnO3n+1 series with A=La, of single phases. Examples are given in Fig. 4. Thus the Nd, Sr or Ba and M=Fe, Co, Ni and Cu. The oxidation existence of two phases is definitively evidenced in the reaction can be summarized as follows: La2CuO4+d system (d#0.04 and 0.09); they exhibit supercon- AxMyOz+2d OH-�AxMyOz+d+dH2O+2de- (2) ducting transitions at 32 and 44 K, respectively.Similarly, in the SrCoO3-y system, several steps are observed involving d representing the amount of inserted oxygen. The fact that again the formation of various intermediate phases before the oxygen atoms and not hydroxide OH- species are intercalated, last stage of oxidation that leads to the fully oxidized material was earlier demonstrated by several groups using various SrCoO3.Similar features have also been reported for techniques such as TGA and chemical analysis, coulometric La2NiO4+d and Nd2NiO4+d, which confirms the close analogy titration, 1H NMR, IR and Mo�ssbauer spectroscopies (see with cationic intercalation phenomena.examples below). Table 1 summarizes the most significant results obtained Fig. 2 shows a typical I–E voltammetry curve (anodic part) during the last few years.7–17 measured on a rotating electrode. This example is concerned with the brownmillerite-type Sr2Co2O5 material. Three regions can be seen in this curve: a linear part immediately after the Fig. 3 Dependence of the rest potential (EOCV is recorded after 15 min) Fig. 2 Voltammetry curve (anodic part) of a brownmillerite Sr2Co2O5- type electrode in a 1 M NaOH solution (RDE 2000 rpm, dE/dt= on the oxygen composition of SrCoO3-y. It shows two branches agreeing with the Armand equation EOCV=f (d)3log [d/(1-d)], d 150 mV min-1, T=300 K, under air). Note the oxidation plateau O1 observed before the oxygen evolution.being the amount of inserted species. 26 J. Mater. Chem., 1999, 9, 25–33Fig. 4 Variation of the O.C.V. potential vs. the oxygen composition (d) for La2CuO4+d and (3-y) for SrCoO3-y calculated from the overall charge transfer and reaction (2) for galvanostatic experiments. Structural aspects The main structural types that have been studied are represented in Fig. 5, all of them being related to the perovskite structure. The AnMnO3n-1 series. These compounds can be considered as oxygen deficient perovskites AMO3-y (n2; 0y0.5); their structures obviously exhibit channels perpendicular to the z axis, which can constitute diVusion paths for intercalating oxygen atoms. However, this precondition is not suYcient as some compounds, especially those containing calcium, do not intercalate oxygen.A tentative and simple model can be proposed by calculating the electrostatic potentials along the channels, which was performed for Ca2Fe2O5, Sr2 Fe2O5 and Fig. 5 Representation of the main structural types of the AnMnO3n-1 Sr2LaFe3O8. The electrostatic potentials along the diVusion and An+1MnO3n+1 series showing the channels and interlayer space paths are negative for Ca2Fe2O5 and positive for Sr2Fe2O5 allowing the oxygen intercalation.and Sr2LaFe3O8. Therefore, for a negatively charged oxygen species, its diVusion along the channels in Sr2Fe2O5 and Sr2LaFe3O8 compounds may be favored by the Coulomb energy, but this is not the case for Ca2Fe2O5. Electrochemical experiments confirm that, actually, it is not possible to insert oxygen within the networks of Ca2Fe2O5 and Ca2LaFe3O8.Table 1 Some examples of electrochemical oxidation of perovskite-related oxides Starting material Electrochemical conditions Final material Electrical behavior Ref. AnMnO3n-1-type Sr2Fe2O5 E=400 mV; 60 h SrFeO3 Metallic 7 Sr2FeCoO5 I=250 mA; 240 h Sr2FeCoO6 Metallic 8 Sr2Co2O5 E=500 mV; 180 h SrCoO3 Metallic 9 Sr2LaFe3O8 E=500 mV; 190 h Sr2LaFe3O8.95 Semi-conducting 10 YBa2Cu3O6.50 E=600 mV; 15 h YBa2Cu3O7-e Metallic 11 An+1MnO3n+1-type La2CuO4.01 E=430 mV; 120 h La2CuO4.09 Metal/superconducting This work La2CuO4.01 E=390 mV; 120 h La2CuO4.04 Metal/superconducting This work La2CuO4.01 I=10 mA; 430 h La2CuO4.09 Metal/superconducting This work La2-xNdxCuO4 E=600 mV; 15 h La2-xNdxCuO4.09 Metal/superconducting 12 La2-xSrxCuO4 I=100 mA cm-2 La2-xSrxCuO4+d Metal/superconducting 13 La2-xBaxCuO4 I=50 mA La2-xBaxCuO4+d Metal/superconducting 14 Nd2NiO4 Potential step (5 mV) Nd2NiO4.25 Semi-conducting 15 La2NiO4 E=600 mV; 100 h La2NiO4.25 Semi-conducting 16 La3Ni2O7 I=10 mA; 360 h La3Ni2O7.10 Metallic 17 La4Ni3O10 I=5 mA; 720 h La4Ni3O10.10 Metallic 17 J.Mater. Chem., 1999, 9, 25–33 27Fig. 6 Representation of the AxOy compact hexagonal-type layers showing the missing rows of oxygen constituting the diVusion paths in the perovskite-related compounds. The Ruddlesden–Popper An+1MnO3n+1 series. In these com- Fig. 7 Mo�ssbauer spectra at 300 K for Sr2LaFe3O8 (before electrochemical oxidation) and for Sr2LaFe3O8.95 (after pounds, such channels do not appear so clearly, but their electrochemical oxidation).structure can be considered as lamellar (Fig. 5), the additional oxygen atoms being inserted into the ‘A2O2’ layers. From a structural viewpoint, X-ray diVraction analysis Table 2 Mo�ssbauer parameters of Sr2LaFe3O8+d phases at room temshows the materials to remain well crystallized after the perature electrochemical treatment. This demonstrates the oxygen inter- Octahedral Tetrahedral calation to be a topotactic reaction and easy diVusion paths sites sites to exist within these structures. Indeed, common features can be found in both series of materials considering that all these Sr2LaFe3O8.00 d=0.35 mm s-1 d=0.18 mm s-1 structures are based on a stacking of compact hexagonal AxOy (Fe3+) H=51.4 T H=37.1 T layers between which metal cations M are located.18 Then, Sr2LaFe3O8.95 d=0.14 mm s-1 — (Fe3.63+) D=0.16 mm s-1 — one can easily notice that oxygen rows along the [010]o are missing in these layers.Examples are given in Fig. 6. The stacking of these layers finally gives rise to channels in the Electrochemical oxidation of An+1MnO3n+1 compounds AnMnO3n-1 series and rather to interlayer spaces in the An+1MnO3n+1 series.Unlike the deficient perovskites, the materials belonging to the An+1MnO3n+1 series are only partly oxidized (dmax#0.10 for Electrochemical oxidation of AnMnO3n-1 compounds: the cuprates and 0.25 for nickelates) with respect to the number example of Sr2LaFe3O8 of vacant sites (Table 1), which means that only a few oxygen atoms are inserted in the missing rows (as shown in Fig. 6). The compounds belonging to the AnMnO3n-1 series (n2) mentioned in Table 1 can be fully oxidized into the corresponding stoichiometric perovskites. All starting materials are almost insulators at room temperature and become semiconductors, metals, or even superconductors upon oxidation. An interesting example is given by Sr2LaFe3O8 (n=3 member of the series) which can be oxidized into Sr2LaFe3O8.95 (cubic perovskite).Mo� ssbauer spectroscopy clearly evidences the drastic changes occurring in the material (Fig. 7). The starting material exhibits, at room temperature, a Mo� ssbauer spectrum characterizing a magnetically ordered compound (TN#720 K) and trivalent iron in octahedral (Oh) and tetrahedral (Td) sites; the intensity ratio between these lines (Oh5Td=68532) agrees well with the distribution of iron in both sites with regard to the structure (Fig. 5). Upon electrochemical oxidation, the spectrum undergoes a complete change and no trace of the starting material remains. It exhibits a single peak that can be fitted using a paramagnetic doublet; the Mo�ssbauer parameters reported in Table 2 characterize an averaged valence state (3.63+) evidencing that iron atoms have been oxidized, so that the perovskite compound is formed.Above a phase transition temperature of 195 K, which can be detected by DSC and magnetic measurements as well as by Fig. 8 Temperature dependence of the logarithm of the conductivity studying the transport properties (electrical conductivity or for Sr2LaFe3O8.95.In the inset is represented the temperature dcoeYcient; Fig. 8), a fast electron transfer between dence of the Seebeck coeYcient. The electronic transition is clearly evidenced at 195 K. iron atoms occurs resulting in rather high conductivity. 28 J. Mater. Chem., 1999, 9, 25–33Fig. 9 Electron diVraction patterns for La2CuO4.09, zone axis [3910].The low temperature pattern exhibits extra spots which, reversibly, disappear near room temperature. However, these materials are oxidized to a larger extent than that using high oxygen pressures. The most interesting feature of these compounds is the Fig. 10 Composition dependence of the oxygen excess (d) for the oxygen mobility near room temperature.According to the La2Cu1-xNixO4+d system, measured either by iodometry or by composition d, it may result in a ‘phase separation’ which was coulometry. extensively discussed earlier.19,20 Thus it appears that given compositions of oxygen excess d lead to stable phases for which the additional oxygen atoms are ordered. For instance, for the nickelate La2NiO4+d, we described in detail the formation of a new phase La8Ni4O17 (d=0.25) for which the ordered stage results from an ‘electronic’ coupling between the oxygen ordering into the La2O2 planes and a disproportionation of nickel cations resulting in a charge density wave within the NiO2 planes.21 For La2CuO4+d such an ordering was not clearly observed but two line phases were characterized, d= 0.04 and d=0.09.22 For the latter, the mobility of oxygen is evidenced using electron microscopy.Fig. 9 shows electron diVraction patterns obtained at 100 K and 340 K: they are obviously diVerent. The low temperature one exhibits extra spots characterizing a commensurate modulated structure which can be interpreted in terms of locally ordered oxygen atoms in the La2O2 layers leading to the formation of a giant supercell. At higher temperature, these weak spots disappear, which results from a disordered state of these additional oxygen atoms near room temperature.Upon lowering the temperature these spots reversibly reappear, thus showing the temperature eVect on the oxygen mobility. (b) Electrochemical reduction Fig. 11 Time dependences of the potential for typical compositions of La2MO4 (M=Ni, Cu) compounds have been widely investi- the La2Cu1-xNixO4+d system (galvanostatic experiments I=-10 mA).gated during the last decade in relation to the superconducting properties exhibited by La2-xBaxCuO4+d .23 These compounds, of the potential down to a constant value, which characterizes annealed in air, show some oxygen overstoichiometry and it the end of the reduction of the material.This step occurs after finally appears rather diYcult to obtain truly stoichiometric a polarization time which depends on the composition of the samples due to the stabilizing role of these additional oxygen starting compound. Then the samples were chemically ana- atoms, which has been discussed earlier in terms of structural lyzed; their oxygen stoichiometry was found to be equal to distortion and electronic eVects.24 4.00 within experimental error, which shows that all the excess We have previously mentioned the interest in using the oxygen was electrochemically removed.The overall reaction electrochemical potential for overoxidizing such materials. can be written as follows: Therefore, is it also possible to use such a method for deintercalating these oxygen atoms? Is the electrochemical La2Cu1-xNixO4+d+dH2O+2d e- oxidation a reversible process? �La2Cu1-xNixO4.00+2dOH- (3) Various compositions of the La2Cu1-xNixO4 system were at first prepared by solid state reaction of nitrates, then pellets In addition, on the basis of this reaction, the value of d in the starting materials can be calculated from the overall charge were sintered at 1080 °C for compositions x<0.25 and at 1250 °C for x0.25.Chemical analysis by iodometry of the transfer (Q=Itr), tr being the polarization time considered when the potential reaches a steady negative value. Thus it oxidation state of the cations led to the determination of the excess oxygen diodo which depends on the composition follows that: (Fig. 10). dcoul=ItrM/2Fm Electrochemical experiments were carried out using the galvanostatic mode with a cathodic current of -10 mA. The in which I is the current intensity (in this example -10 mA), M is the molar mass of the starting compound, F is the time dependence of the potential for typical x values is reported in Fig. 11. All of them are similar and exhibit a drastic decrease Faraday constant (96484 C) and m is the sample mass.The J. Mater. Chem., 1999, 9, 25–33 29Fig. 13 Thermal dependence of the reciprocal magnetic susceptibility of as-prepared and electrochemically reduced La2NiO4+d . ation processes can also be achieved using chemical agents: the most well known example is the lithium deintercalation from some oxide (e.g.LiCoO2 or LiNbO2) or sulfide (e.g. Fig. 12 Thermal dependence of the electrical conductivity for asprepared, electrochemically oxidized and electrochemically reduced LiTiS2) networks using strong oxidizing agents such as I2, La2CuO4+d . Br2, MoF6 , etc. in acetonitrile.27–29 One can also mention oxidizing intercalation processes in aqueous solutions containvalues of dcoul reported in Fig. 10 show a very good agreement ing species such as MnO4-, BrO- or ClO- which were used, with those determined by chemical analysis, which confirms for instance, for overoxidizing cuprates such as La2CuO4+d or the exchange of 2d electrons according to electrochemical LaCuO2+d .30,31 reaction (3) and that it is really oxygen atoms that are However, another way is a solid–gas reaction at moderate deintercalated.temperatures (T350 °C). In this temperature range, the In addition this is corroborated by following the changes in kinetics with O2 is generally slow and one can use F2 or Cl2 the properties of the materials such as for instance the electrical instead but decomposition of the materials is often observed. conductivity of La2CuO4+d compounds.The as-prepared We recently investigated an alternative reaction involving material La2CuO4.01 exhibits a characteristic semiconducting gaseous NO2 and more especially its reaction with ammonium behavior with a large hump at about 40 K, indicative of ions NH4+.32 Actually such a process implies the N+IV/N-III filamentary superconductivity (Fig. 12). Galvanostatic oxi- redox couple as in the explosive decomposition of NH4NO3; dation at 10 mA for 18 days leads to a metallic compound in addition we may expect the in situ destruction of NH4+ in La2CuO4.09 which undergoes a superconducting transition at the starting materials according to the following reaction: 44 K.Then electrochemical reduction at steady potential, NO2(g)+NH4++e-�N2(+2H2O( (4) 200 mV, for 1 day, deintercalates all the oxygen atoms as shown by the perfectly semiconducting behavior of the This mole-to-mole reaction yields neutral species, i.e.gaseresistivity characterizing the stoichiometric La2CuO4.00 ous H2O and N2, as by-products that can be easily expelled (Fig. 12). from the material with respect to their size which is close to In the same way, the lanthanum nickelate prepared in air that of NH4+ cations.Therefore one can presume a topotactic shows a large oxygen excess (d#0.14) and the stoichiometric destruction of these large cations leaving large empty sites in compound can be obtained by reduction using H2 or CO/CO2 the final compound. One should also note that this redox (i.e. under controlled oxygen pressure) at about 400 °C.25 reaction requires one electron.Previous works reported this phase to be a 3D antiferromag- Typically, powders of the starting material deposited on a netic oxide with TN=328 K, the magnetic properties depending sintered glass are treated at temperatures ranging between 20 strongly on the oxygen excess (d).25,26 The galvanostatic and 350 °C by a mixture of NO2–N2. Depending on the reduction of La2NiO4.14, I=-200 mA for 3 days, leads to a materials, the reaction can be violent.Various kinds of reacbrown sample. XRD analysis confirms the orthorhombic tions have been carried out; they ce classified as follows. symmetry (Bmab) characterizing the stoichiometric oxide La2NiO4.00. The thermal dependence of the magnetic suscepti- Topotactic reactions bility shown in Fig. 13 reveals again an important change; the The initial structure remains unchanged after the in situ kink observed near 325 K confirms previous determinations destruction of NH4+. The first example concerns the hexagonal of TN. tungsten bronzes (NH4)xWO3 that are prepared by controlled Both examples demonstrate the eYciency of this method reduction of (NH4)10W12O41·5H2O under Ar–H2 (10%) at ca.that can easily control the oxygen stoichiometry of this kind 380 °C. The composition x determined by TGA and the of material as well as the reversibility of the oxygen Kjeldahl method is close to 0.20. Then a treatment of this intercalation. dark blue sample with NO2, for 3 h, near 250–300 °C leads to a well crystallized white powder, this color being characteristic (c) Oxidation by NO2 gas of the oxidation of the cations from W5+ to W6+.The absence of ammonium ions was checked by the Kjeldahl method and Previous examples were concerned with electrochemically monitored reactions in aqueous solutions. Oxidative deintercal- various physical characterizations (IR, mass spectrometry) 30 J. Mater. Chem., 1999, 9, 25–33Non-topotactic reactions As pointed out, the reaction of ammonium compounds may sometimes be violent, resulting in destruction of the parent structure.Two examples can be quoted. NH4VO3 crystallizes with the structure of a-NaVO3 made of (VO4) chains of tetrahedra; this compound was treated by NO2 at various temperatures. At 100 °C, within a few minutes, it is transformed into the well crystallized a-V2O5 whose lamellar structure is completely diVerent from that of the starting material (V5+ cations are located in corner-sharing pyramids).35 In the same way, starting from ammonium cerium nitrate (NH4)2Ce(NO3)6, one can prepare cerium oxide CeO2, at typically 200–300 °C, by NO2 treatment for about 1 hour.The X-ray diVraction pattern reveals a poorly crystallized material (Fig. 16). SEM images show a particle size of about 0.1–0.2 mm. With respect to its industrial applications (cataly- Fig. 14 X-Ray diVractogram of hexagonal WO3. The inset shows a sis, ionic conductor in solid oxide fuel cells), the preparation high resolution electron diVraction image ([001] zone axis). of this oxide with such a submicronic texture is of great interest.36 Obviously the use of NO2 gas as oxidizing agent at moderate temperatures and more especially its exothermic reaction with reveal that the final compound is a hexagonal form of WO3.ammonium compounds opens a large field of investigation for This is confirmed by the X-ray diVraction pattern as well as preparing oxides in metastable forms or with peculiar textures. by the electron microscopy image (Fig. 14) showing the large tunnels of the structure (Fig. 15). The lattice parameters (a= 7.334 A° , c=7.652 A° ) are somewhat diVerent from those pre- (d ) Chemical reduction by borohydride solution viously reported for h-WO3 obtained by a reconstructive The synthesis of oxides containing metal cations with low process (dehydration of WO3·1/3H2O, a=7.298 A° , c= oxidation states usually requires solid state reactions at high 7.798 A° ), the c parameter being unambiguously doubled.33 temperatures under reducing atmosphere (i.e.low oxygen The second example deals with ammonium molybdenum partial pressure pO2) generated by CO/CO2 or H2, NH3 , etc. oxide; the starting compound is obtained from For preparing metastable phases, at about room temperature, Na2MoO4·2H2O by cationic exchange in a solution of NH4Cl, alternative routes should be used.We mentioned above the the resulting material being ammonium and molybdenum electrochemical reduction that allows the deintercalation of deficient with a composition close to (NH4)0.7Mo5.3O18H3.4, oxygen but also the intercalation of cations such as H+ or but containing only Mo+VI.34 XRD analysis shows a single Li+ in materials devoted to applications in electrochromism phase with hexagonal symmetry (a=10.583 A° , c=3.724 A° ).or batteries. Often, the latter reaction is achieved in non- The NO2 treatment (3 h, 150 to 300 °C) again retains the aqueous solvents; however chemical reduction of oxides has parent structure and the final material was identified to be also been considered in aqueous solutions.Various reducing pure hexagonal MoO3 (Fig. 15). One should point out that in agents can be used; one can quote, for instance, Zn or sulfites this case no oxidation of the cation is required, which raises which have been recognized for their eYciency, especially for the question of the provenance of the electron in reaction (4). preparing metals.37 Recently borohydride solutions have been There could be two explanations: either there is an oxidation of oxygen anions according to O2-�DO2+2e-, reaction (4) remaining the basic one, or the reaction with NO2 occurs via a two step process, the first being the destruction of NH3 leading to the formation of a hydrate or hydroxide without oxidation of the material, according to a scheme which only involves an internal redox process: 4NH4++3NO2(g)�4H++7/2N2(+6H2O( (5) Then this hydrate or hydroxide can be thermally decomposed at moderate temperatures.Fig. 16 X-Ray diVractograms of the commercial cerium oxide and Fig. 15 Hexagonal structures of WO3 and MoO3 showing the large empty channels after NO2 treatment. that obtained by the NO2 route. J. Mater. Chem., 1999, 9, 25–33 31reassessed for the preparation of nanoparticles and low valent soluble species, between which ammonium ions are inserted as pillars. It is quite diVerent from that of a-V2O5.oxides at room temperature.38–40 According to the thermodynamical data, the redox potential One should point out that, in more or less similar conditions, Zhu and Manthiram prepared NaxWO3, which was amorph- of the following reaction: ous,44 which emphasizes that experimental conditions are of H2BO3-+5H2O+8e-uBH4-+8OH- vital importance.is E0=-1.24 V vs. SHE.37 With regard to this high value and (2) Preparation of KxWO3: a topotactic reduction in order to prevent water decomposition, sodium or potassium borohydrides should be stabilized in alkaline solution. We described above the preparation of hexagonal WO3: this Various kinds of reactions can be achieved; recently, Zhang compound contains large tunnels in which various chemical and Manthiram38 reported the preparation of M–boron nano- species can be easily inserted.45 The reduction by KBH4 is particles (M=Fe, Co, Ni) and we also obtained amorphous carried out in a solution in which h-WO3 has been mixed with cobalt by this method.With respect to the reduction of oxides, an excess of KCl. The pH is kept close to 7.5. Within a few the nature of the reaction products as well as the texture of minutes the white starting material becomes blue. The powder the materials seem to depend drastically on the concentration is then filtered and dried. X-Ray diVraction confirms that the of the solutions and on the pH.Often the synthesized oxides hexagonal structure is preserved. Electron microprobe analysis are amorphous, which, for applications such as in batteries, is leads to a potassium content close to 0.27 in agreement with advantageous. Thus Tsang and Manthiram prepared MoO2 the value expected from the addition of borohydride (x#0.30). and VO2(B) whose cyclability makes these compounds attract- This example clearly demonstrates that borohydride ive for Li batteries.40,41 solutions can provide an interesting route for intercalating Another interesting application of these borohydride reduc- cations such as Li+, Na+, K+ or NH4+ into oxide networks ing solutions is the preparation of crystallized oxides at room using aqueous solutions instead of other routes such as butyltemperature, which we recently developed.42 Two kinds of lithium in organic solvents.experiments have been carried out: in the first, the starting oxide is dissolved beforehand, whereas in the secondis not. Conclusion (1) Preparation of (NH4)2V3O8 by precipitation The development of ‘chimie douce’ preparative methods in solid state chemistry has undoubtedly broadened the field of The starting material a-V2O5 is dissolved at first in 1 M KOH new materials either in a metastable form or with particular solution.NH4Cl is then added in excess, the pH being close textures, which cannot be obtained by other routes. This is to 10.5. The reduction of V5+ to V4+ is obtained by addition due to the fact that the nature of the reaction products is of the solution of NaBH4 under nitrogen atmosphere.At last, controlled by kinetics rather than by thermodynamics. These the pH is slowly lowered down to 7.5 by adding dilute HCl methods can be classified into two main types: those using the and black ammonium vanadyl vanadate (NH4)2V3O8 precipielectrochemical potential as the driving force and those based tates.After filtration, washing and drying, the material is on chemical redox reactions. characterized by chemical analysis and TGA and its X-ray For transition metal oxides, provided that the crystal diVraction pattern shows a rather well crystallized compound structure accommodates diVusion paths for oxygen ions, it is (Fig. 17). It is identified as the ‘black hydrate of vanadium’ possible to tune the oxygen stoichiometry over a large range synthesized earlier by reduction, at 100 °C, for 24 hours, using and hence to promote interesting properties such as Zn as reducing agent.43 One should point out that our method superconductivity.leads within 1 hour to a similar result. NO2 gas as well as borohydrides are basic reactants in The structure of this compound is built up of ‘V3O8’ sheets chemistry, but, surprisingly, they have not been much used in formed from the condensation and reduction of (VO4)3- solid state chemistry.We described some attractive reactions, which may open new fields of investigations at moderate temperatures. The solid–gas reaction of NO2 with ammonium compounds leads to two types of processes: those for which the in situ destruction of NH4+ is topotactic and those where the precursor structures collapse giving rise to more stable compounds or to amorphous or finely divided materials.Finally, the use of borohydride aqueous solutions for reducing oxides appears simple and eYcient; it generally yields nanotextured materials interesting for various applications such as catalysis or energy storage in batteries.The authors thank the Rho�ne-Poulenc company for financial support and are very grateful to C. Monroux, Y. Muraoka, P. Dordor, A. Demourgues, F. Arrouy, E. Marquestaut, M. D. Carvolho and F. M. A. Costa for their helpful contributions to this work. References 1 Soft chemistry routes to new materials, Mater. Sci. Forum, 1994, 152–153, ed. J. Rouxel, M. Tournoux and R.Brec, Trans Tech Publications, USA. 2 J-C. Grenier, A. Wattiaux, J-P. Doumerc, P. Dordor, L. Fourne`s, J-P. Chaminade and M. Pouchard, J. Solid State Chem., 1992, 96, 20. 3 M. Izaki and T. Omi, J. Electrochem. Soc., 1997, 144, 1949. Fig. 17 X-Ray diVractogram of (NH4)2V3O8 prepared by the 4 A. Wattiaux, J-C. Park, J-C. Grenier and M. Pouchard, C. R. Acad. Sci.Paris, 1990, 310, 1047. borohydride method. 32 J. Mater. Chem., 1999, 9, 25–335 J. O’M. Bockris and T. Otagawa, J. Electrochem. Soc., 1984, 131, in cuprate superconductors, ed. K. A. Mu� ller Benedeck, World Scientific, Singapore, 1994, p. 237. 290. 25 S. Hosoya, T. Omata, K. Nakajima, K. Yamada and Y. Endoh, 6 M. Armand, Intercalation in Materials for Advanced Batteries, ed.Physica C, 1992, 202, 188. D. W. Murphy, Plenum New York, 1980, p. 145. 26 A. Demourgues, P. Dordor, J-P. Doumerc, J-C. Grenier, 7 A. Wattiaux, L. Fournes, A. Demourgues, N. Bernaben, E. Marquestaut, M. Pouchard, A. Villesuzanne and A. Wattiaux, J-C. Grenier and M. Pouchard, Solid State Commun., 1991, 77, J. Solid State Chem., 1996, 124, 199. 489. 27 G. G. Amatucci, J. M. Tarascon and L.C. Klein, J. Electrochem. 8 P. Bezdicka, L. Fournes, A. Wattiaux, J. C. Grenier and Soc., 1996, 143, 1114. M. Pouchard, Solid State Commun., 1994, 91, 501. 28 R. Gupta and A. Manthiram, J. Solid State Chem., 1996, 121, 483. 9 P. Bezdicka, A. Wattiaux, J. C. Grenier, M. Pouchard and 29 A. R. Wizansky, P. E. Rauch and F. J. Disalvo, J. Solid State P. Hagenmuller, Z. Anorg.Allg. Chem., 1993, 619, 7. Chem., 1989, 81, 203. 10 A.Wattiaux, L. Fourne`s, F. Zhou, J-C. Grenier, M. Pouchard and 30 E. Takayama-Muromachi, T. Sasaki and Y. Matsui, Physica C, J. Etourneau, J. Phys. IV Fr., 1997, 7, C1351. 1993, 207, 97. 11 J. C. Park, A. Wattiaux, J. C. Grenier, K. Fro� hlich, P. Dordor, 31 M. Trari, J. To� pfer, J-P. Doumerc, M. Pouchard, A. Ammar and M. Pouchard and P.Hagenmuller, Z. Anorg. Allg. Chem., 1992, P. Hagenmuller, J. Solid State Chem., 1994, 111, 104. 608, 153. 32 S. Petit, J-P. Doumerc, J-C. Grenier, T. Se�guelong and 12 F. Arrouy, A. Wattiaux, E. Marquestaut, C. Cros, G. Demazeau, M. Pouchard, C.R. Acad. Sci. Paris, Ser. IIb, 1995, 321, 37. J-C. Grenier and M. Pouchard, J. Solid State Chem., 1995, 115, 33 B. Ge�rand, G.Novogrocki, J. Guenot and M. Figlarz, J. Solid 140. State Chem., 1979, 29, 429. 13 F. C. Chou, J. H. Cho and D. C. Johnston, Physica C, 1992, 34 Y. Muraoka, J-C. Grenier, S. Petit and M. Pouchard, Eur. J. Solid State Inorg. Chem., submitted. 197, 303. 35 S. Petit, J. P. Doumerc, J. C. Grenier and M. Pouchard, C. R. 14 M. Itoh, Y. J. Shan, S. Sakamoto, Y. Inaguma and T. Nakamura, Acad.Sci., in press. Physica C, 1994, 223, 75. 36 J-P. Doumerc, J-C. Grenier, M. Pouchard and S. Petit, Rho�ne- 15 S. Bhavaraju, J. D. DiCarlo, D. P. Scarfe, I. Yazdi and Poulenc Chimie, Fr. Pat., 95-01874, 1995. A. J. Jacobson, Chem. Mater., 1994, 6, 2172. 37 Handbook of Chemistry and Physics, ed. R. C. Weast, CRC Press, 16 A. Demourgues, A. Wattiaux, J-C. Grenier, M. Pouchard, Cleveland, OH, 56th edn., 1976, D146.J-M. Dance and P. Hagenmuller, J. Solid State Chem., 1993, 105, 38 L. Zhang and A. Manthiram, J. Mater. Chem., 1996, 6, 999. 458. 39 C. Tsang, A. Dananjay, J. Kim and A. Manthiram, Inorg. Chem., 17 M. D. Carvalho, F. M. A. Costa, I. S. Pereira, A. Wattiaux, 1996, 35, 504. J. M. Bassat, J-C. Grenier and M. Pouchard, to be published. 40 A. Manthiram and Y.T. Zhu, J. Electrochem. Soc., 1996, 143, 18 C. Dussaret, F. Grasset and J. Darriet, Eur. J. Solid State Inorg. L143. Chem., 1995, 32, 557. 41 C. Tsang and A. Manthiram, J. Electrochem. Soc., 1997, 144, 520. 19 Phase separation in cuprate superconductors, ed. K. A. Mu� ller and 42 S. Petit, K. David, J-P. Doumerc, J-C. Grenier, T. Se�guelong and G. Benedeck, World Scientific, Singapore, 1992.M. Pouchard, C.R. Acad. Sci. Paris, Ser. IIc, 1998, 1, 517. 20 Phase separation in cuprate superconductors, ed. E. Sigmund and 43 J. Bernard and F. The�obald, C. R. Acad. Sci. Paris, 1963, 4916. K. A. Mu� ller, Springer Verlag, Berlin, 1994. 44 Y. T. Zhu and A. Manthiram, J. Solid State Chem., 1994, 110, 187. 21 A. Demourgues, F. Weill, B. Darriet, A. Wattiaux, J-C.Grenier, 45 S. Petit, B. Jousseaume, L. Fourne`s, J. P. Doumerc, J. C. Grenier, P. Gravereau and M. Pouchard, J. Solid State Chem., 1993, 105, A. Wattiaux, M. Pouchard and P. Hagenmuller, Essays on Interdisciplinary Topics in Natural Sciences, ed. R. B. Scorzelli, 317, 330. I. Souza Azevedo and E. Baggio Saitovitch, CBPF, Rio de 22 C. Monroux, A. Wattiaux and J-C. Grenier, to be published. Janeiro, Brazil, 1997, p. 199. 23 J. G. Bednorz and K. A. Mu� ller, Z. Phys. B, 1986, 64, 189. 24 J. C. Grenier, F. Arrouy, J-P. Locquet, C. Monroux, M. Pouchard, A. Villesuzanne and A. Wattiaux, Phase separation Paper 8/05213A J. Mater. Chem., 1999, 9, 25–33 33 J O U R N A L O F C H E M I S T R Y Materials Feature Article Relevant examples of intercalation–deintercalation processes in solid state chemistry: application to oxides† J-C.Grenier,*a J-M. Bassat,a J-P. Doumerc,a J. Etourneau,a Z. Fang,b L. Fournes,a S. Petit,a M. Poucharda and A. Wattiauxa aInstitut de la Chimie de la Matie`re Condense�e de Bordeaux, C.N.R.S., Av. du Dr. A. Schweitzer, 33 608 Pessac Cedex,himsol.icmcb.u-bordeaux.fr. bInstitute of Physics, the Chinese Academy of Sciences, P.O.Box 603, 100080 Beijing, China Received 4th May 1998, Accepted 6th July 1998 Four ‘chimie douce’ synthetic methods, to which much (a) Electrochemical oxidation attention has been devoted in the last years, are described. In 1990, we discovered that using the electrochemical potential The electrochemical potential is shown to be efficient for as a ‘driving force’, it was possible to intercalate oxygen into intercalating as well as for deintercalating oxygen atoms in oxides.Gaseous NO2 is able to destroy in situ oxide networks, at room temperature, and correspondingly to ammonium ions either in a soft way or violently, leading overoxidize oxides up to unexpected oxygen stoichiometries.2 to the formation of tunneled or lamellar structures or to Basically this reaction can be achieved thanks to the high oxide nanoparticles. Borohydride solutions appear very oxygen activity resulting from an applied overpotential g on attractive for the synthesis of low valent oxides at room an oxide ceramic electrode according to the reaction: temperature.Typical examples of these methods are given 4OH-�O2(g)(+2 H2O+4e- and the most important features of these reactions are discussed.(Eth0=0.303 V vs. Hg/HgO, at pH=14) (1) Using the Nernst equation, one can estimate this oxygen Introduction activity to be equivalent, at room temperature for a few hundred mV of overpotential, to a very high oxygen pressure. Until the 1960s, most solid state compounds were prepared Indeed, such activity can induce the formation of oxide films using high temperature treatments of starting materials that on metal electrodes. This has been recently used for growing were first ground and pressed.Then new preparative methods tailored thick films of oxides such as Al2O3, ZnO, CuO, etc.,3 were investigated at mild temperatures with the aim of synthesbut they are usually passivating.However, by replacing a izing new compounds, generally unstable at high temperatures: metallic electrode by an oxide and provided that a diVusion that was the beginning of the ‘chimie douce’ era. These methods path exists within the oxide matrix, oxygen atoms have been can be classified in diVerent ways according to the following shown to migrate easily. features: all of them involve ‘precursors’ that must be tailored The most striking result was obtained in 1990 for La2CuO4: with respect to the desired final material; the reaction temperathis insulating antiferromagnetic compound was transformed ture ranges between room temperature and nearly 300 °C, a within a few hours into a metallic and superconducting oxide lower temperature leading usually to poorer crystallization; with TC=44 K, the highest transition temperature ever from a structural viewpoint, they can be either topotactic or obtained for this phase.4 reconstructive; and finally, such reactions may or not imply redox processes.Experimental Nowadays, the largest scale technical applications of this ‘soft solid state chemistry’ are undoubtedly concerned with The electrochemical system used for these experiments consists batteries.The reactions are typically intercalation–deintercal- of a power supply (a potentiostat) and an electrochemical cell ation of either cationic (Li+, H+) or neutral species (Li or of three-electrode type, using alkaline solution, 1 M NaOH or H2) at room temperature; the former involve redox processes KOH, as electrolyte. The electrolyte concentration of 1 M and hence mixed valence oxides [e.g.MnO2-x(OH)x, LixNiO2, (pH=14) is considered as the optimal one. Fig. 1 shows a etc.], the latter deal with alloys such as LaNi5, metals (Al ) or scheme of the electrochemical cell, the major components of graphite.1 which are the three electrodes. This paper reports on some methods recently developed for The working electrodes are usually ceramic pellets of the preparing oxides by either chemical or electrochemical oxi- materials to be oxidized: their relative densities are around dation or reduction of an appropriate precursor.The common 80%, which seems to be the optimal density allowing good feature of these methods is their eYciency even at room permeation of the electrolyte into the ceramic. A rotating temperature.Descriptions and relevant examples of the follow- working electrode is preferred to a fixed one as it produces ing four methods are given: (a) electrochemical oxidation better convection between the electrolyte in the ceramic elecresulting in oxygen intercalation into oxide networks; (b) dein- trode and in the bulk solution. It also prevents evolved gas tercalation of oxygen using electrochemical reduction; bubbles (O2), if any, from accumulating on the electrode (c) chemical oxidation using NO2 gas; and (d) chemical surface.A stable and ohmic contact between the ceramic pellet reduction using borohydride solutions. and the metallic support (made of brass) which is connected with the external circuit, is required for the electrochemical process and meaningful data measurements.The assembly is achieved thanks to a gold foil and silver paste, the whole being †Basis of the presentation given at Materials Chemistry Discussion embedded in a resin. No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. The reference electrode is a mercury–mercury oxide electrode J. Mater. Chem., 1999, 9, 25–33 25polarization begins (C1, E<300 mV) assigned to the charging of the double layer on the interface between the electrolyte and the electrode surface.5 Then a wave-shaped part (O1, E= 300–450 mV) with the half-wave potential about E1/2= 330 mV: this is ascribed to the oxidation of the electrode material according to reaction (2).In this case it corresponds formally to the transformation from Co3+ to Co4+. At higher potentials the current intensity increases again and the evolution of gaseous oxygen takes place in the third region (O2, E>450 mV) according to reaction (1).Then two kinds of experiments can be performed. (i) Potentiostatic oxidation, using a steady potential set in the oxidation plateau, depending on the starting material (e.g. 400 mV for Sr2Co2O5). The oxidation reaction is followed by measuring the time dependence of the rest potential: the open Fig. 1 The electrochemical cell with three electrodes for experiments at 25 °C–90 °C. circuit voltage, EOCV, measured after 15 min, increases with the polarization time. Actually it can be directly correlated to the chemical composition, i.e. to the oxygen stoichiometry (Hg/HgO) that has excellent stability in alkaline solutions.Its (3-y) or to the amount of Co2+/Co3+/Co4+ (Fig. 3). It standard potential (at aOH-=1 and T=298 K) is 0.098 V shows a logarithmic dependence on the oxygen composition, relative to SHE. All the potentials quoted in this work are which may be explained in a similar way as for the intercalation referred to this electrode.of lithium ions into CoO2,MnO2 or other oxides on the basis of The counter electrode is designed to have a large surface area. The materials usually used are good electronic conduc- the Armand equation.6 The end of the polarization reaction tors, such as gold or platinum foil, or a piece of glassy carbon. is reflected in a drastic increase of the rest potential. (ii) Galvanostatic experiments, which are more convenient for Electrochemical aspects controlling the reaction kinetics, have been used by several groups.As for the lithium intercalation into dichalcogenides Various compounds AxMzOy whose structures derive from or oxides, using small current intensities (a few mA), potential perovskite, have been oxidized: they belong essentially to the steps are observed.Such a feature characterizes the formation so-called AnMnO3n-1 and An+1MnO3n+1 series with A=La, of single phases. Examples are given in Fig. 4. Thus the Nd, Sr or Ba and M=Fe, Co, Ni and Cu. The oxidation existence of two phases is definitively evidenced in the reaction can be summarized as follows: La2CuO4+d system (d#0.04 and 0.09); they exhibit supercon- AxMy+2d OH-�AxMyOz+d+dH2O+2de- (2) ducting transitions at 32 and 44 K, respectively. Similarly, in the SrCoO3-y system, several steps are observed involving d representing the amount of inserted oxygen.The fact that again the formation of various intermediate phases before the oxygen atoms and not hydroxide OH- species are intercalated, last stage of oxidation that leads to the fully oxidized material was earlier demonstrated by several groups using various SrCoO3.Similar features have also been reported for techniques such as TGA and chemical analysis, coulometric La2NiO4+d and Nd2NiO4+d, which confirms the close analogy titration, 1H NMR, IR and Mo�ssbauer spectroscopies (see with cationic intercalation phenomena. examples below). Table 1 summarizes the most significant results obtained Fig. 2 shows a typical I–E voltammetry curve (anodic part) during the last few years.7–17 measured on a rotating electrode. This example is concerned with the brownmillerite-type Sr2Co2O5 material. Three regions can be seen in this curve: a linear part immediately after the Fig. 3 Dependence of the rest potential (EOCV is recorded after 15 min) Fig. 2 Voltammetry curve (anodic part) of a brownmillerite Sr2Co2O5- type electrode in a 1 M NaOH solution (RDE 2000 rpm, dE/dt= on the oxygen composition of SrCoO3-y. It shows two branches agreeing with the Armand equation EOCV=f (d)3log [d/(1-d)], d 150 mV min-1, T=300 K, under air). Note the oxidation plateau O1 observed before the oxygen evolution. being the amount of inserted species. 26 J.Mater. Chem., 1999, 9, 25–33Fig. 4 Variation of the O.C.V. potential vs. the oxygen composition (d) for La2CuO4+d and (3-y) for SrCoO3-y calculated from the overall charge transfer and reaction (2) for galvanostatic experiments. Structural aspects The main structural types that have been studied are represented in Fig. 5, all of them being related to the perovskite structure.The AnMnO3n-1 series. These compounds can be considered as oxygen deficient perovskites AMO3-y (n2; 0y0.5); their structures obviously exhibit channels perpendicular to the z axis, which can constitute diVusion paths for intercalating oxygen atoms. However, this precondition is not suYcient as some compounds, especially those containing calcium, do not intercalate oxygen.A tentative and simple model can be proposed by calculating the electrostatic potentials along the channels, which was performed for Ca2Fe2O5, Sr2 Fe2O5 and Fig. 5 Representation of the main structural types of the AnMnO3n-1 Sr2LaFe3O8. The electrostatic potentials along the diVusion and An+1MnO3n+1 series showing the channels and interlayer space paths are negative for Ca2Fe2O5 and positive for Sr2Fe2O5 allowing the oxygen intercalation.and Sr2LaFe3O8. Therefore, for a negatively charged oxygen species, its diVusion along the channels in Sr2Fe2O5 and Sr2LaFe3O8 compounds may be favored by the Coulomb energy, but this is not the case for Ca2Fe2O5. Electrochemical experiments confirm that, actually, it is not possible to insert oxygen within the networks of Ca2Fe2O5 and Ca2LaFe3O8.Table 1 Some examples of electrochemical oxidation of perovskite-related oxides Starting material Electrochemical conditions Final material Electrical behavior Ref. AnMnO3n-1-type Sr2Fe2O5 E=400 mV; 60 h SrFeO3 Metallic 7 Sr2FeCoO5 I=250 mA; 240 h Sr2FeCoO6 Metallic 8 Sr2Co2O5 E=500 mV; 180 h SrCoO3 Metallic 9 Sr2LaFe3O8 E=500 mV; 190 h Sr2LaFe3O8.95 Semi-conducting 10 YBa2Cu3O6.50 E=600 mV; 15 h YBa2Cu3O7-e Metallic 11 An+1MnO3n+1-type La2CuO4.01 E=430 mV; 120 h La2CuO4.09 Metal/superconducting This work La2CuO4.01 E=390 mV; 120 h La2CuO4.04 Metal/superconducting This work La2CuO4.01 I=10 mA; 430 h La2CuO4.09 Metal/superconducting This work La2-xNdxCuO4 E=600 mV; 15 h La2-xNdxCuO4.09 Metal/superconducting 12 La2-xSrxCuO4 I=100 mA cm-2 La2-xSrxCuO4+d Metal/superconducting 13 La2-xBaxCuO4 I=50 mA La2-xBaxCuO4+d Metal/superconducting 14 Nd2NiO4 Potential step (5 mV) Nd2NiO4.25 Semi-conducting 15 La2NiO4 E=600 mV; 100 h La2NiO4.25 Semi-conducting 16 La3Ni2O7 I=10 mA; 360 h La3Ni2O7.10 Metallic 17 La4Ni3O10 I=5 mA; 720 h La4Ni3O10.10 Metallic 17 J.Mater. Chem., 1999, 9, 25–33 27Fig. 6 Representation of the AxOy compact hexagonal-type layers showing the missing rows of oxygen constituting the diVusion paths in the perovskite-related compounds. The Ruddlesden–Popper An+1MnO3n+1 series.In these com- Fig. 7 Mo�ssbauer spectra at 300 K for Sr2LaFe3O8 (before electrochemical oxidation) and for Sr2LaFe3O8.95 (after pounds, such channels do not appear so clearly, but their electrochemical oxidation).structure can be considered as lamellar (Fig. 5), the additional oxygen atoms being inserted into the ‘A2O2’ layers. From a structural viewpoint, X-ray diVraction analysis Table 2 Mo�ssbauer parameters of Sr2LaFe3O8+d phases at room temshows the materials to remain well crystallized after the perature electrochemical treatment. This demonstrates the oxygen inter- Octahedral Tetrahedral calation to be a topotactic reaction and easy diVusion paths sites sites to exist within these structures.Indeed, common features can be found in both series of materials considering that all these Sr2LaFe3O8.00 d=0.35 mm s-1 d=0.18 mm s-1 structures are based on a stacking of compact hexagonal AxOy (Fe3+) H=51.4 T H=37.1 T layers between which metal cations M are located.18 Then, Sr2LaFe3O8.95 d=0.14 mm s-1 — (Fe3.63+) D=0.16 mm s-1 — one can easily notice that oxygen rows along the [010]o are missing in these layers.Examples are given in Fig. 6. The stacking of these layers finally gives rise to channels in the Electrochemical oxidation of An+1MnO3n+1 compounds AnMnO3n-1 series and rather to interlayer spaces in the An+1MnO3n+1 series.Unlike the deficient perovskites, the materials belonging to the An+1MnO3n+1 series are only partly oxidized (dmax#0.10 for Electrochemical oxidation of AnMnO3n-1 compounds: the cuprates and 0.25 for nickelates) with respect to the number example of Sr2LaFe3O8 of vacant sites (Table 1), which means that only a few oxygen atoms are inserted in the missing rows (as shown in Fig. 6). The compounds belonging to the AnMnO3n-1 series (n2) mentioned in Table 1 can be fully oxidized into the corresponding stoichiometric perovskites. All starting materials are almost insulators at room temperature and become semiconductors, metals, or even superconductors upon oxidation. An interesting example is given by Sr2LaFe3O8 (n=3 member of the series) which can be oxidized into Sr2LaFe3O8.95 (cubic perovskite).Mo� ssbauer spectroscopy clearly evidences the drastic changes occurring in the material (Fig. 7). The starting material exhibits, at room temperature, a Mo� ssbauer spectrum characterizing a magnetically ordered compound (TN#720 K) and trivalent iron in octahedral (Oh) and tetrahedral (Td) sites; the intensity ratio between these lines (Oh5Td=68532) agrees well with the distribution of iron in both sites with regard to the structure (Fig. 5).Upon electrochemical oxidation, the spectrum undergoes a complete change and no trace of the starting material remains. It exhibits a single peak that can be fitted using a paramagnetic doublet; the Mo�ssbauer parameters reported in Table 2 characterize an averaged valence state (3.63+) evidencing that iron atoms have been oxidized, so that the perovskite compound is formed.Above a phase transition temperature of 195 K, which can be detected by DSC and magnetic measurements as well as by Fig. 8 Temperature dependence of the logarithm of the conductivity studying the transport properties (electrical conductivity or for Sr2LaFe3O8.95. In the inset is represented the temperature depen- Seebeck coeYcient; Fig. 8), a fast electron transfer between dence of the Seebeck coeYcient. The electronic transition is clearly evidenced at 195 K. iron atoms occurs resulting in rather high conductivity. 28 J. Mater. Chem., 1999, 9, 25–33Fig. 9 Electron diVraction patterns for La2CuO4.09, zone axis [3910]. The low temperature pattern exhibits extra spots which, reversibly, disappear near room temperature.However, these materials are oxrger extent than that using high oxygen pressures. The most interesting feature of these compounds is the Fig. 10 Composition dependence of the oxygen excess (d) for the oxygen mobility near room temperature. According to the La2Cu1-xNixO4+d system, measured either by iodometry or by composition d, it may result in a ‘phase separation’ which was coulometry.extensively discussed earlier.19,20 Thus it appears that given compositions of oxygen excess d lead to stable phases for which the additional oxygen atoms are ordered. For instance, for the nickelate La2NiO4+d, we described in detail the formation of a new phase La8Ni4O17 (d=0.25) for which the ordered stage results from an ‘electronic’ coupling between the oxygen ordering into the La2O2 planes and a disproportionation of nickel cations resulting in a charge density wave within the NiO2 planes.21 For La2CuO4+d such an ordering was not clearly observed but two line phases were characterized, d= 0.04 and d=0.09.22 For the latter, the mobility of oxygen is evidenced using electron microscopy.Fig. 9 shows electron diVraction patterns obtained at 100 K and 340 K: they are obviously diVerent. The low temperature one exhibits extra spots characterizing a commensurate modulated structure which can be interpreted in terms of locally ordered oxygen atoms in the La2O2 layers leading to the formation of a giant supercell. At higher temperature, these weak spots disappear, which results from a disordered state of these additional oxygen atoms near room temperature.Upon lowering the temperature these spots reversibly reappear, thus showing the temperature eVect on the oxygen mobility. (b) Electrochemical reduction Fig. 11 Time dependences of the potential for typical compositions of La2MO4 (M=Ni, Cu) compounds have been widely investi- the La2Cu1-xNixO4+d system (galvanostatic experiments I=-10 mA).gated during the last decade in relation to the superconducting properties exhibited by La2-xBaxCuO4+d .23 These compounds, of the potential down to a constant value, which characterizes annealed in air, show some oxygen overstoichiometry and it the end of the reduction of the material. This step occurs after finally appears rather diYcult to obtain truly stoichiometric a polarization time which depends on the composition of the samples due to the stabilizing role of these additional oxygen starting compound.Then the samples were chemically ana- atoms, which has been discussed earlier in terms of structural lyzed; their oxygen stoichiometry was found to be equal to distortion and electronic eVects.24 4.00 within experimental error, which shows that all the excess We have previously mentioned the interest in using the oxygen was electrochemically removed. The overall reaction electrochemical potential for overoxidizing such materials. can be written as follows: Therefore, is it also possible to use such a method for deintercalating these oxygen atoms? Is the electrochemical La2Cu1-xNixO4+d+dH2O+2d e- oxidation a reversible process? �La2Cu1-xNixO4.00+2dOH- (3) Various compositions of the La2Cu1-xNixO4 system were at first prepared by solid state reaction of nitrates, then pellets In addition, on the basis of this reaction, the value of d in the starting materials can be calculated from the overall charge were sintered at 1080 °C for compositions x<0.25 and at 1250 °C for x0.25.Chemical analysis by iodometry of the transfer (Q=Itr), tr being the polarization time considered when the potential reaches a steady negative value. Thus it oxidation state of the cations led to the determination of the excess oxygen diodo which depends on the composition follows that: (Fig. 10). dcoul=ItrM/2Fm Electrochemical experiments were carried out using the galvanostatic mode with a cathodic current of -10 mA.The in which I is the current intensity (in this example -10 mA), M is the molar mass of the starting compound, F is the time dependence of the potential for typical x values is reported in Fig. 11. All of them are similar and exhibit a drastic decrease Faraday constant (96484 C) and m is the sample mass. The J.Mater. Chem., 1999, 9, 25–33 29Fig. 13 Thermal dependence of the reciprocal magnetic susceptibility of as-prepared and electrochemically reduced La2NiO4+d . ation processes can also be achieved using chemical agents: the most well known example is the lithium deintercalation from some oxide (e.g. LiCoO2 or LiNbO2) or sulfide (e.g. Fig. 12 Thermal dependence of the electrical conductivity for asprepared, electrochemically oxidized and electrochemically reduced LiTiS2) networks using strong oxidizing agents such as I2, La2CuO4+d .Br2, MoF6 , etc. in acetonitrile.27–29 One can also mention oxidizing intercalation processes in aqueous solutions containvalues of dcoul reported in Fig. 10 show a very good agreement ing species such as MnO4-, BrO- or ClO- which were used, with those determined by chemical analysis, which confirms for instance, for overoxidizing cuprates such as La2CuO4+d or the exchange of 2d electrons according to electrochemical LaCuO2+d .30,31 reaction (3) and that it is really oxygen atoms that are However, another way is a solid–gas reaction at moderate deintercalated.temperatures (T350 °C). In this temperature range, the In addition this is corroborated by following the changes in kinetics with O2 is generally slow and one can use F2 or Cl2 the properties of the materials such as for instance the electrical instead but decomposition of the materials is often observed.conductivity of La2CuO4+d compounds. The as-prepared We recently investigated an alternative reaction involving material La2CuO4.01 exhibits a characteristic semiconducting gaseous NO2 and more especially its reaction with ammonium behavior with a large hump at about 40 K, indicative of ions NH4+.32 Actually such a process implies the N+IV/N-III filamentary superconductivity (Fig. 12). Galvanostatic oxi- redox couple as in the explosive decomposition of NH4NO3; dation at 10 mA for 18 days leads to a metallic compound in addition we may expect the in situ destruction of NH4+ in La2CuO4.09 which undergoes a superconducting transition at the starting materials according to the following reaction: 44 K.Then electrochemical reduction at steady potential, NO2(g)+NH4++e-�N2(+2H2O( (4) 200 mV, for 1 day, deintercalates all the oxygen atoms as shown by the perfectly semiconducting behavior of the This mole-to-mole reaction yields neutral species, i.e.gaseresistivity characterizing the stoichiometric La2CuO4.00 ous H2O and N2, as by-products that can be easily expelled (Fig. 12). from the material with respect to their size which is close to In the same way, the lanthanum nickelate prepared in air that of NH4+ cations. Therefore one can presume a topotactic shows a large oxygen excess (d#0.14) and the stoichiometric destruction of these large cations leaving large empty sites in compound can be obtained by reduction using H2 or CO/CO2 the final compound.One should also note that this redox (i.e. under controlled oxygen pressure) at about 400 °C.25 reaction requires one electron. Previous works reported this phase to be a 3D antiferromag- Typically, powders of the starting material deposited on a netic oxide with TN=328 K, the magnetic properties depending sintered glass are treated at temperatures ranging between 20 strongly on the oxygen excess (d).25,26 The galvanostatic and 350 °C by a mixture of NO2–N2.Depending on the reduction of La2NiO4.14, I=-200 mA for 3 days, leads to a materials, the reaction can be violent. Various kinds of reacbrown sample.XRD analysis confirms the orthorhombic tions have been carried out; they can be classified as follows. symmetry (Bmab) characterizing the stoichiometric oxide La2NiO4.00. The thermal dependence of the magnetic suscepti- Topotactic reactions bility shown in Fig. 13 reveals again an important change; the The initial structure remains unchanged after the in situ kink observed near 325 K confirms previous determinations destruction of NH4+.The first example concerns the hexagonal of TN. tungsten bronzes (NH4)xWO3 that are prep by controlled Both examples demonstrate the eYciency of this method reduction of (NH4)10W12O41·5H2O under Ar–H2 (10%) at ca. that can easily control the oxygen stoichiometry of this kind 380 °C.The composition x determined by TGA and the of material as well as the reversibility of the oxygen Kjeldahl method is close to 0.20. Then a treatment of this intercalation. dark blue sample with NO2, for 3 h, near 250–300 °C leads to a well crystallized white powder, this color being characteristic (c) Oxidation by NO2 gas of the oxidation of the cations from W5+ to W6+.The absence of ammonium ions was checked by the Kjeldahl method and Previous examples were concerned with electrochemically monitored reactions in aqueous solutions. Oxidative deintercal- various physical characterizations (IR, mass spectrometry) 30 J. Mater. Chem., 1999, 9, 25–33Non-topotactic reactions As pointed out, the reaction of ammonium compounds may sometimes be violent, resulting in destruction of the parent structure.Two examples can be quoted. NH4VO3 crystallizes with the structure of a-NaVO3 made of (VO4) chains of tetrahedra; this compound was treated by NO2 at various temperatures. At 100 °C, within a few minutes, it is transformed into the well crystallized a-V2O5 whose lamellar structure is completely diVerent from that of the starting material (V5+ cations are located in corner-sharing pyramids).35 In the same way, starting from ammonium cerium nitrate (NH4)2Ce(NO3)6, one can prepare cerium oxide CeO2, at typically 200–300 °C, by NO2 treatment for about 1 hour.The X-ray diVraction pattern reveals a poorly crystallized material (Fig. 16). SEM images show a particle size of about 0.1–0.2 mm. With respect to its industrial applications (cataly- Fig. 14 X-Ray diVractogram of hexagonal WO3. The inset shows a sis, ionic conductor in solid oxide fuel cells), the preparation high resolution electron diVraction image ([001] zone axis). of this oxide with such a submicronic texture is of great interest.36 Obviously the use of NO2 gas as oxidizing agent at moderate temperatures and more especially its exothermic reaction with reveal that the final compound is a hexagonal form of WO3.ammonium compounds opens a large field of investigation for This is confirmed by the X-ray diVraction pattern as well as preparing oxides in metastable forms or with peculiar textures. by the electron microscopy image (Fig. 14) showing the large tunnels of the structure (Fig. 15). The lattice parameters (a= 7.334 A° , c=7.652 A° ) are somewhat diVerent from those pre- (d ) Chemical reduction by borohydride solution viously reported for h-WO3 obtained by a reconstructive The synthesis of oxides containing metal cations with low process (dehydration of WO3·1/3H2O, a=7.298 A° , c= oxidation states usually requires solid state reactions at high 7.798 A° ), the c parameter being unambiguously doubled.33 temperatures under reducing atmosphere (i.e.low oxygen The second example deals with ammonium molybdenum partial pressure pO2) generated by CO/CO2 or H2, NH3 , etc. oxide; the starting compound is obtained from For preparing metastable phases, at about room temperature, Na2MoO4·2H2O by cationic exchange in a solution of NH4Cl, alternative routes should be used.We mentioned above the the resulting material being ammonium and molybdenum electrochemical reduction that allows the deintercalation of deficient with a composition close to (NH4)0.7Mo5.3O18H3.4, oxygen but also the intercalation of cations such as H+ or but containing only Mo+VI.34 XRD analysis shows a single Li+ in materials devoted to applications in electrochromism phase with hexagonal symmetry (a=10.583 A° , c=3.724 A° ).or batteries. Often, the latter reaction is achieved in non- The NO2 treatment (3 h, 150 to 300 °C) again retains the aqueous solvents; however chemical reduction of oxides has parent structure and the final material was identified to be also been considered in aqueous solutions. Various reducing pure hexagonal MoO3 (Fig. 15). One should point out that in agents can be used; one can quote, for instance, Zn or sulfites this case no oxidation of the cation is required, which raises which have been recognized for their eYciency, especially for the question of the provenance of the electron in reaction (4). preparing metals.37 Recently borohydride solutions have been There could be two explanations: either there is an oxidation of oxygen anions according to O2-�DO2+2e-, reaction (4) remaining the basic one, or the reaction with NO2 occurs via a two step process, the first being the destruction of NH3 leading to the formation of a hydrate or hydroxide without oxidation of the material, according to a scheme which only involves an internal redox process: 4NH4++3NO2(g)�4H++7/2N2(+6H2O( (5) Then this hydrate or hydroxide can be thermally decomposed at moderate temperatures.Fig. 16 X-Ray diVractograms of the commercial cerium oxide and Fig. 15 Hexagonal structures of WO3 and MoO3 showing the large empty channels after NO2 treatment. that obtained by the NO2 route. J. Mater. Chem., 1999, 9, 25–33 31reassessed for the preparation of nanoparticles and low valent soluble species, between which ammonium ions are inserted as pillars.It is quite diVerent from that of a-V2O5. oxides at room temperature.38–40 According to the thermodynamical data, the redox potential One should point out that, in more or less similar conditions, Zhu and Manthiram prepared NaxWO3, which was amorph- of the following reaction: ous,44 which emphasizes that experimental conditions are of H2BO3-+5H2O+8e-uBH4-+8OH- vital importance.is E0=-1.24 V vs. SHE.37 With regard to this high value and (2) Preparation of KxWO3: a topotactic reduction in order to prevent water decomposition, sodium or potassium borohydrides should be stabilized in alkaline solution. We described above the preparation of hexagonal WO3: this Various kinds of reactions can be achieved; recently, Zhang compound contains large tunnels in which various chemical and Manthiram38 reported the preparation of M–boron nano- species can be easily inserted.45 The reduction by KBH4 is particles (M=Fe, Co, Ni) and we also obtained amorphous carried out in a solution in which h-WO3 has been mixed with cobalt by this method. With respect to the reduction of oxides, an excess of KCl.The pH is kept close to 7.5. Within a few the nature of the reaction products as well as the texture of minutes the white starting material becomes blue. The powder the materials seem to depend drastically on the concentration is then filtered and dried. X-Ray diVraction confirms that the of the solutions and on the pH. Often the synthesized oxides hexagonal structure is preserved.Electron microprobe analysis are amorphous, which, for applications such as in batteries, is leads to a potassium content close to 0.27 in agreement with advantageous. Thus Tsang and Manthiram prepared MoO2 the value expected from the addition of borohydride (x#0.30). and VO2(B) whose cyclability makes these compounds attract- This example clearly demonstrates that borohydride ive for Li batteries.40,41 solutions can provide an interesting route for intercalating Another interesting application of these borohydride reduc- cations such as Li+, Na+, K+ or NH4+ into oxide networks ing solutions is the preparation of crystallized oxides at room using aqueous solutions instead of other routes such as butyltemperature, which we recently developed.42 Two kinds of lithium in organic solvents.experiments have been carried out: in the first, the starting oxide is dissolved beforehand, whereas in the second it is not. Conclusion (1) Preparation of (NH4)2V3O8 by precipitation The development of ‘chimie douce’ preparative methods in solid state chemistry has undoubtedly broadened the field of The starting material a-V2O5 is dissolved at first in 1 M KOH new materials either in a metastable form or with particular solution.NH4Cl is then added in excess, the pH being close textures, which cannot be obtained by other routes. This is to 10.5. Theuction of V5+ to V4+ is obtained by addition due to the fact that the nature of the reaction products is of the solution of NaBH4 under nitrogen atmosphere. At last, controlled by kinetics rather than by thermodynamics.These the pH is slowly lowered down to 7.5 by adding dilute HCl methods can be classified into two main types: those using the and black ammonium vanadyl vanadate (NH4)2V3O8 precipielectrochemical potential as the driving force and those based tates. After filtration, washing and drying, the material is on chemical redox reactions.characterized by chemical analysis and TGA and its X-ray For transition metal oxides, provided that the crystal diVraction pattern shows a rather well crystallized compound structure accommodates diVusion paths for oxygen ions, it is (Fig. 17). It is identified as the ‘black hydrate of vanadium’ possible to tune the oxygen stoichiometry over a large range synthesized earlier by reduction, at 100 °C, for 24 hours, using and hence to promote interesting properties such as Zn as reducing agent.43 One should point out that our method superconductivity.leads within 1 hour to a similar result. NO2 gas as well as borohydrides are basic reactants in The structure of this compound is built up of ‘V3O8’ sheets chemistry, but, surprisingly, they have not been much used in formed from the condensation and reduction of (VO4)3- solid state chemistry. We described some attractive reactions, which may open new fields of investigations at moderate temperatures.The solid–gas reaction of NO2 with ammonium compounds leads to two types of processes: those for which the in situ destruction of NH4+ is topotactic and those where the precursor structures collapse giving rise to more stable compounds or to amorphous or finely divided materials. Finally, the use of borohydride aqueous solutions for reducing oxides appears simple and eYcient; it generally yields nanotextured materials interesting for various applications such as catalysis or energy storage in batteries. The authors thank the Rho�ne-Poulenc company for financial support and are very grateful to C.Monroux, Y. Muraoka, P. Dordor, A. Demourgues, F. Arrouy, E. Marquestaut, M. D. Carvolho and F. M. A. Costa for their helpful contributions to this work. References 1 Soft chemistry routes to new materials, Mater. Sci. Forum, 1994, 152–153, ed. J. Rouxel, M. Tournoux and R. Brec, Trans Tech Publications, USA. 2 J-C. Grenier, A. Wattiaux, J-P.Doumerc, P. Dordor, L. Fourne`s, J-P. Chaminade and M. Pouchard, J. Solid State Chem., 1992, 96, 20. 3 M. Izaki and T. Omi, J. Electrochem. Soc., 1997, 144, 1949. Fig. 17 X-Ray diVractogram of (NH4)2V3O8 prepared by the 4 A. Wattiaux, J-C. Park, J-C. Grenier and M. Pouchard, C. R. Acad. Sci. Paris, 1990, 310, 1047. borohydride method. 32 J. Mater. Chem., 1999, 9, 25–335 J.O’M. Bockris and T. Otagawa, J. Electrochem. Soc., 1984, 131, in cuprate superconductors, ed. K. A. Mu� ller Benedeck, World Scientific, Singapore, 1994, p. 237. 290. 25 S. Hosoya, T. Omata, K. Nakajima, K. Yamada and Y. Endoh, 6 M. Armand, Intercalation in Materials for Advanced Batteries, ed. Physica C, 1992, 202, 188. D. W. Murphy, Plenum New York, 1980, p. 145. 26 A.Demourgues, P. Dordor, J-P. Doumerc, J-C. Grenier, 7 A. Wattiaux, L. Fournes, A. Demourgues, N. Bernaben, E. Marquestaut, M. Pouchard, A. Villesuzanne and A. Wattiaux, J-C. Grenier and M. Pouchard, Solid State Commun., 1991, 77, J. Solid State Chem., 1996, 124, 199. 489. 27 G. G. Amatucci, J. M. Tarascon and L. C. Klein, J. Electrochem. 8 P. Bezdicka, L. Fournes, A. Wattiaux, J. C. Grenier and Soc., 1996, 143, 1114. M. Pouchard, Solid State Commun., 1994, 91, 501. 28 R. Gupta and A. Manthiram, J. Solid State Chem., 1996, 121, 483. 9 P. Bezdicka, A. Wattiaux, J. C. Grenier, M. Pouchard and 29 A. R. Wizansky, P. E. Rauch and F. J. Disalvo, J. Solid State P. Hagenmuller, Z. Anorg. Allg. Chem., 1993, 619, 7. Chem., 1989, 81, 203. 10 A.Wattiaux, L. Fourne`s, F. Zhou, J-C. Grenier, M. Pouchard and 30 E. Takayama-Muromachi, T. Sasaki and Y. Matsui, Physica C, J. Etourneau, J. Phys. IV Fr., 1997, 7, C1351. 1993, 207, 97. 11 J. C. Park, A. Wattiaux, J. C. Grenier, K. Fro� hlich, P. Dordor, 31 M. Trari, J. To� pfer, J-P. Doumerc, M. Pouchard, A. Ammar and M. Pouchard and P. Hagenmuller, Z. Anorg. Allg. Chem., 1992, P. Hagenmuller, J. Solid State Chem., 1994, 111, 104. 608, 153. 32 S. Petit, J-P. Doumerc, J-C. Grenier, T. Se�guelong and 12 F. Arrouy, A. Wattiaux, E. Marquestaut, C. Cros, G. Demazeau, M. Pouchard, C.R. Acad. Sci. Paris, Ser. IIb, 1995, 321, 37. J-C. Grenier and M. Pouchard, J. Solid State Chem., 1995, 115, 33 B. Ge�rand, G. Novogrocki, J. Guenot and M. Figlarz, J. Solid 140. State Chem., 1979, 29, 429. 13 F. C. Chou, J. H. Cho and D. C. Johnston, Physica C, 1992, 34 Y. Muraoka, J-C. Grenier, S. Petit and M. Pouchard, Eur. J. Solid State Inorg. Chem., submitted. 197, 303. 35 S. Petit, J. P. Doumerc, J. C. Grenier and M. Pouchard, C. R. 14 M. Itoh, Y. J. Shan, S. Sakamoto, Y. Inaguma and T. Nakamura, Acad. Sci., in press. Physica C, 1994, 223, 75. 36 J-P. Doumerc, J-C. Grenier, M. Pouchard and S. Petit, Rho�ne- 15 S. Bhavaraju, J. D. DiCarlo, D. P. Scarfe, I. Yazdi and Poulenc Chimie, Fr. Pat., 95-01874, 1995. A. J. Jacobson, Chem. Mater., 1994, 6, 2172. 37 Handbook of Chemistry and Physics, ed. R. C. Weast, CRC Press, 16 A. Demourgues, A. Wattiaux, J-C. Grenier, M. Pouchard, Cleveland, OH, 56th edn., 1976, D146. J-M. Dance and P. Hagenmuller, J. Solid State Chem., 1993, 105, 38 L. Zhang and A. Manthiram, J. Mater. Chem., 1996, 6, 999. 458. 39 C. Tsang, A. Dananjay, J. Kim and A. Manthiram, Inorg. Chem., 17 M. D. Carvalho, F. M. A. Costa, I. S. Pereira, A. Wattiaux, 1996, 35, 504. J. M. Bassat, J-C. Grenier and M. Pouchard, to be published. 40 A. Manthiram and Y. T. Zhu, J. Electrochem. Soc., 1996, 143, 18 C. Dussaret, F. Grasset and J. Darriet, Eur. J. Solid State Inorg. L143. Chem., 1995, 32, 557. 41 C. Tsang and A. Manthiram, J. Electrochem. Soc., 1997, 144, 520. 19 Phase separation in cuprate superconductors, ed. K. A. Mu� ller and 42 S. Petit, K. David, J-P. Doumerc, J-C. Grenier, T. Se�guelong and G. Benedeck, World Scientific, Singapore, 1992. M. Pouchard, C.R. Acad. Sci. Paris, Ser. IIc, 1998, 1, 517. 20 Phase separation in cuprate superconductors, ed. E. Sigmund and 43 J. Bernard and F. The�obald, C. R. Acad. Sci. Paris, 1963, 4916. K. A. Mu� ller, Springer Verlag, Berlin, 1994. 44 Y. T. Zhu and A. Manthiram, J. Solid State Chem., 1994, 110, 187. 21 A. Demourgues, F. Weill, B. Darriet, A. Wattiaux, J-C. Grenier, 45 S. Petit, B. Jousseaume, L. Fourne`s, J. P. Doumerc, J. C. Grenier, P. Gravereau and M. Pouchard, J. Solid State Chem., 1993, 105, A. Wattiaux, M. Pouchard and P. Hagenmuller, Essays on Interdisciplinary Topics in Natural Sciences, ed. R. B. Scorzelli, 317, 330. I. Souza Azevedo and E. Baggio Saitovitch, CBPF, Rio de 22 C. Monroux, A. Wattiaux and J-C. Grenier, to be published. Janeiro, Brazil, 1997, p. 199. 23 J. G. Bednorz and K. A. Mu� ller, Z. Phys. B, 1986, 64, 189. 24 J. C. Grenier, F. Arrouy, J-P. Locquet, C. Monroux, M. Pouchard, A. Villesuzanne and A. Wattiaux, Phase separation Paper 8/05213A J. Mater. Chem.,

ISSN:0959-9428    

DOI:10.1039/a805213a

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Molecular design of hybrid organic–inorganic nanocomposites synthesized via sol–gel chemistry† C. Sanchez, F. Ribot and B. Lebeau Chimie de la Matie`re Condense�e, UMR CNRS 7574, Universite� Pierre et Marie Curie 4, place Jussieu, 75252 Paris, France. E-mail: clems@ccr.jussieu.fr Received 12th May 1998, Accepted 16th July 1998 The design, synthesis and some optical properties of hybrid as organic polymerization and lead to hybrid organic–inororganic –inorganic nanocomposites materials are pre- ganic copolymers.6–8 This article reviews some of the previous sented.The properties that can be expected for such work we have performed on the design and synthesis of hybrid materials depend on the chemical nature of their compo- organic–inorganic nanocomposites in which organics are nents, but they also depend on the synergy of these simply embedded or chemically bonded to the inorganic components.Thus, the interface in these nanocomposites gel network. is of paramount significance and one key point of their synthesis is the control of this interface. These nanocom- 2 Siloxane based hybrid materials posites can be obtained by hydrolysis and condensation reactions of organically functionalized alkoxide precur- Organic groups can be bonded to an inorganic network as sors.Striking examples of hybrids made from modified network modifiers or network formers. Both functions have silicon, tin and transition metal alkoxides are presented. been achieved in the so-called ORMOSILS.3,9 The precursors Some optical properties (photochromic, luminescence, of these compounds are organo-substituted silicic acid esters NLO) of siloxane based hybrids are also discussed.of general formula R¾nSi(OR)4-n, where R¾ can be any organofunctional group. If R¾ is a simple non hydrolyzable group 1 Introduction bonded to silicon through a SiMC bond, it will have a network modifying eVect (SiMCH3). On the other hand, if R¾ can react Sol–gel chemistry is based on the polymerization of molecular with itself (R¾ contains a methacryl group for example) or precursors such as metal alkoxides M(OR)n.1,2 Hydrolysis and additional components, it will act as a network former.3c condensation of these alkoxides lead to the formation of metal Network modifiers and network formers can also introduce oxo-polymers.The mild characteristics oVered by the sol–gel other physical properties (mechanical, hydrophobic, electro- process allow the introduction of organic molecules inside an chemical, optical, etc.). Several examples related to some inorganic network.3 Inorganic and organic components can optical properties of siloxane based hybrids will be now then be mixed at the nanometric scale, in virtually any ratio described.leading to so-called hybrid organic–inorganic nanocomposites. 4 These hybrids are extremely versatile in their composi- 2-A Photochromic properties tion, processing and optical and mechanical properties.5 The nature of the interface between the organic and inorganic Spiropyrans and spirooxazines are two of the fascinating components has been used recently to classify these hybrids families of molecules exhibiting photochromic properties.into two diVerent classes.4h Class I corresponds to all the Upon irradiation, the colorless spiropyran or spirooxazine systems where there are no covalent or iono-covalent bonds undergoes heterolytic CMO ring cleavage, producing colored between the organic and inorganic components.In such mate- forms of merocyanines (Fig. 1). rials, the various components only exchange interactions such The merocyanines may interact with their environment, i.e. as van der Waals forces, hydrogen bondings or electrostatic solvent, matrix, etc. leading to diVerent photochromic forces. In contrast, in class II materials, some of the organic responses.Levy and Avnir10 first demonstrated the important and inorganic components are linked through strong chemical role played by the dye–matrix interactions on the photobonds (covalent or iono-covalent). Numerous hybrid organic– chromic response of spiropyrans. They studied the photochroinorganic materials have been developed in the past few years. mism of spiropyrans trapped in sol–gel matrices synthesized This development yields many interesting new materials, with via polymerization of Si(OCH3)4 or RSi(OEt)3 (R=ethyl, mechanical properties tunable between those of glasses and those of polymers, with improved optical properties (eYciency, stability, new sensors, etc.) and with improved catalytic or membrane based properties.4g This field of materials research mainly arises from chemists’ skills and demonstrates the major role played by chemistry in advanced materials.Siloxane based hybrids4 can be easily synthesized because SiMCsp3 bonds are rather covalent and therefore they are not broken upon hydrolysis. Similar chemistry can be developed from tin alkoxides. This is not aplicable to transition metals for which the more ionic MMC bond is easily cleaved by water; complexing organic ligands must be used.Such groups can be functionalized for any kind of organic reactions such †Basis of the presentation given at Materials Chemistry Discussion NO NO2 SP N O N SO N O– NO2 N N O hn1 hn2, D hn1 hn2, D + Fig. 1 Molecular structures of SO and SP photochromic dyes. No. 1, 24–26 September 1998, University of Bordeaux, France.J. Mater. Chem., 1999, 9, 35–44 35methyl, etc.) precursors, and observed two types of photochromic behavior. When the photochromic dye is trapped within a hydrophilic domain of the matrix (domain containing residual SiMOH groups), the open zwitterionic colored forms are probably stabilized through hydrogen bonding with the acidic silanol groups present at the pore surface.The result of this stabilization is the observation of the colored forms before irradiation. These colored forms can be bleached by irradiation in the visible range. This has been termed ‘reverse photochromism’. On the other hand spiropyran dyes embedded in a more hydrophobic hybrid network made by hydrolysis of RSi(OEt)3 exhibit direct photochromism, i.e.the colorless form is stable without irradiation. Such photochromic behavior has been reported for many spiropyran or spirooxazine doped sol–gel matrices.10–17 Moreover, for hybrid organic–inorganic matrices containing diVerent chemical environments (hydrophilic and hydrophobic domains) a competition between direct and reverse photochromisms can be observed.17 However many fundamental questions still need to be considered.Little is Fig. 2 Cartoon of the D/Zrx matrices. known concerning the role of the photochromic dye–matrix interactions in the kinetics of coloration and thermal fading. As far as photochromic devices are concerned the tuning between strong and fast photochromic coloration (high DA) and very fast thermal fading is needed. Usually spiropyran or spirooxazine doped sol–gel matrices or even spirooxazine doped polymeric matrices exhibit slow thermal fading.10–18 The photochromic behavior of a spiropyran SP (6-nitro- 1¾,3¾,3¾-trimethylspiro{2H-1-benzopyran-2,2¾-indoline}) and a spirooxazine SO (1,3,3-trimethylspiro{indoline-2,3¾(3H)- naph(2,1-b)(1,4)oxazine}) (Fig. 1) embedded within two new hybrid matrices have been recently studied.19 Photochromic properties of the SP or SO doped D/Zr hybrid matrices.The first kind of matrix obtained through hydrolysis and condensation of (CH3)2Si(OC2H5)2 (D) and appropriate amounts of Zr(OPrn)4 (Zr) is labelled D/Zrx, where Zr stands for the zirconium, x for the molar amount of zirconium. Fig. 3 Photocolouration and photobleaching of SP doped D/Zrx These D/Zrx matrices are hybrid nanocomposites made matrices19 [(a) x=10; (b) x=20; (c) x=30].from polydimethylsiloxane species (chains, cycles etc.) crosslinked by zirconium oxopolymers.20–22 The zirconium oxopolymers are hydrophilic domains that still contain some photochromism is partially reversed and be balanced by tuning the D/Zr ratio. hydroxo groups coming from residual ethanol or ZrMOH ligands.22 The size and the spacing between the ZrO2 based The thermal bleaching behavior of the D/Zr20 samples was fitted with a biexponential equation.The SP doped materials domains is about a few nanometers as evidenced by SAXS.22 17O MAS NMR and FTIR experiments show that inside these exhibited a very long bleaching time (about 24 h) while for the SO doped D/Zr20 materials the thermal fading was composites the hydrophobic polydimethylsiloxane species are interfaced with the zirconium oxo domains both through much faster.19 The kinetic data of the SP or SO doped D/Zr20 samples covalent ZrMOMSi bonds and through weak interactions (hydrogen and van der Waals bonds).19 The structure of the are similar to those reported for other modified sol–gel matrices or in organic polymers.17,23 As in organic polymers, the D/Zrx matrices is shown schematically in Fig. 2. D/Zrx matrices doped with SP or SO are lightly colored (pink with bleaching follows a biexponential equation which can be explained by an inhomogeneous distribution of free volumes SP or blue with SO) before irradiation. However the absorbance (A) in the visible region is weak in comparison in the gel.Moreover the presence of diVerent stereoisomers (cis or trans) could also account for this behavior. The diVerent with the total amount of embedded photochromic dyes. The amount of colored form depends on the x content. isomer–matrix interactions could explain the diVerent kinetics observed for SO and SP. Fig. 3 shows the photochromic behavior of SP doped D/Zrx gels for three compositions of zirconium.When the Zr amount The thermal fading is longer for SP doped hybrids than for SO doped ones. This phenomenon can be correlated to the increases, the A variation due to the coloration decreases while that due to the decoloration increases: there are more and fact that SP open forms are known for their tendancy to form zwitterionic species, while non charged quinonic species are more open forms in the gel. The amount of colored form increases proportionally with the x content.It is much higher usually favored for open SO molecules. Zwitterionic species can be strongly stabilized by hydrogen bonding with the for D/Zr30 samples than for D/Zr10 ones. This indicates that before irradiation the SO and SP dyes matrix, thus lowering the decay times of thermal fading.are roughly split into two populations. The colored merocyanine open forms of SO and SP are stabilized by hydrogen Photochromic properties of the SP or SO doped DH/TH hybrid matrices. The second kind of SP or SO doped matrix bonding within the hydrophilic regions made from the zirconium oxopolymers, while the closed SO and SP forms are prepared from the hydrolysis and cocondensation of (CH3)HSi(OC2H5)2 (DH) and HSi(OC2H5)3 (TH) precursors probably located in the environment of the hydrophobic polydimethylsiloxane chains.So for these D/Zrx matrices the is labelled DH70/TH30 (70/30 refers to the molar composi- 36 J. Mater. Chem., 1999, 9, 35–44few minutes under magnetic stirring. A solution of mixed alkoxides, Zr(OPrn)4 and a corresponding rare earth (M= Nd3+, Sm3 +, Dy3 +, etc.) methoxy ethoxide was then added to the previous solution in order to obtain a Zr5Si molar ratio of 159 and M5Si molar ratios of 159; 0.159; 0.0159.The sols were deposited onto glass sheets and allowed to gel and dry at room temperature. Such transparent coatings were deposited several times till a thin layer of about 50–100 mm was obtained.25 FTIR, 1H MAS NMR and 29Si CP MAS NMR experiments have shown that these hybrids can be described as nanocom- Fig. 4 Schematic representation of the environment experienced by posites built from siloxane polymers crosslinked by metal oxo the SO dye within the matrix DH70/TH30. species. The metal oxo domains are made of mixed zirconium–rare earth oxo species.tion). These hybrids can be described as strongly interpen- Absorption–emission properties of these Nd3+ doped hybrid etrated nanocomposites. The environment experienced by the coatings. The room temperature absorption spectrum of Nd3+ SO dye within the DH70/TH30 matrix is shown in Fig. 4. doped hybrid coatings presented in ref. 25 consisted of some The SO or SP DH70/TH30 doped matrices exhibit normal broad transitions.The 4I9/2A4F5/2,2H9/2 transition around photochromism. All the samples are colorless before 800 nm, particularly important for diode pumping systems, irradiation. For the two photochromic dyes, the thermal fading presents a FWHMof around 15 nm. The absorption coeYcient can be fitted with excellent agreement with a monoexponential at this wavelength is around 10 cm-1 for a coating containing equation.This may be related to the quasi-liquid mobility approximately 4×1020 Nd3+ ions cm-3. This high absorption observed by NMR for this matrix. coeYcient value indicates that a high Nd3+ concentration The time dependence of the absorption upon repeated could be introduced into the system without aVecting the irradiation with 365 nm light for SO doped DH70/TH30 synthesis and the transparency of the films.coatings is reported in Fig. 5. The photochromic behavior is For Nd3+ ions excited at 805 nm, the emission spectra of reversible, extremely fast (the rate constant is 0.2 s-1) and the 4F3/2A4I11/2 transition at room temperature is presented corresponds to a very high absorption jump (DA=1.2).The in Fig. 6. This emission is broad with a FWHM around 40 nm photochromic kinetics of these SO doped hybrid materials are and extends into the NIR range from approximately 1.045 to to the best of our knowledge much faster than those reported 1.095 mm. A broad emission has also been observed for the for SO in any other matrix (sol–gel matrices, organic polymers, 4F3/2A4I9/2 transition.25 Broad emissions for Nd3+ are charac- alcohols, etc.).11,14,15,17,23 teristic of wide sites distribution around the neodymium ions The very high reactivity of the DH/TH precursors towards in this hybrid materials.Such inhomogeneous broadening is hydrolysis–condensation reactions (SiMOH groups are immealso observed in glassy hosts.27,28 Experimentally, one can diatly consumed) and the strong hydrophobicity of the notice that the Nd3+ fluorescence intensity is weaker than resulting matrix are both responsible for the direct and very those reported in glassy or crystalline matrices where the fast photochromic behavior observed.quantum eYciency is relatively high.29 2-B Luminescent properties of rare-earth doped hybrids Fluorescence kinetics of Nd3+ in hybrid siloxane based coat- Neodymium doped sol–gel matrices are suitable luminescent ings.Fig. 7 presents the variation of the fluorescence intensity materials. However room temperature sol–gel derived matrices decay profiles as a function of the Nd3+ content. Lifetimes usually contain a large amount of hydroxy groups which are decrease from 160 ms for 0.4×1020 Nd3+ ions cm-3 to responsible for the quenching of the Nd3+ emission.Therefore approximately 200 ns for 4×1021 Nd3+ ions cm-3. the design of sol–gel matrices inside which the hydroxy content It is quite unusual to observe radiative emission of Nd3+ can be minimized24,25 and/or inside which the rare earth is ions in matrices prepared at room temperature by sol–gel protected via complexation26 or encapsulation is desirable.processing. Usually the numerous hydroxy groups present in Recent work has demonstrated fluorescence emission for the classical xerogels obtained at room temperature prevent any Nd3+, Sm3 +, Dy3 +, Er3 + and Tm3+ ions doped in hybrid Nd3+ radiative emission.5d,5e,32 A thermal treatment at high siloxane–oxide matrices.25 temperatures is necessary to allow the Nd3+ emission to be These hybrids were synthesized by the following procedure. observed.In all our samples fluorescence is detected, however Diethoxymethylsilane [DEMS=SiH(OEt)2(CH3)], absolute ethanol and water in a 15151 molar ratio were mixed for a Fig. 6 RT Nd3+ fluorescence spectra.30 Fig. 5 Photochromic response of the SO doped DH70/TH30 matrix.18 J. Mater. Chem., 1999, 9, 35–44 37case, phonons can interact with the atomic transitions of Nd3+ and this coupling, analogous to weak electric dipole coupling, is enhanced when hydroxy groups are close to the neodymium ions.These interactions may explain the low Nd3+ quantum eYciency values and the weak emission observed for high Nd3+ concentration. All these results need of course to be improved, however they open many possibilities in the field of room temperature processed luminescent films.Energy transfer between an organic dye and Nd3+ as inorganic chromophore. Codoped rhodamine 6G–Nd3+ hybrid samples shows that the rhodamine emission spectra exhibit some dips at wavelengths corresponding to the Nd3+ absorption bands.31 This feature indicates that radiative energy transfer occurs between the organic dye and the Nd3+ ions.33–35 A photon emitted by the rhodamine molecule can be trapped by the rare-earth ions in the hybrid siloxane network.Furthermore, when excited by an argon laser at 488 nm, a wavelength where only the organic dye molecules absorb, the codoped (R6G–Nd3+) hybrid coating exhibits a Nd3+ emission around 1.06 mm. No emission was detected around 1.06 mm under the same experimental conditions without the Fig. 7 Variation of the Nd3+ fluorescence decay31 (Nd3+ content in presence of the organic dye inside the hybrid coating. This atom cm-3, (a) 0.4×10,20 (b) 4×1020 (c) 4×1021. behavior reveals that energy transfer mechanisms may be favorable for Nd3+ emission. the Nd3+ fluorescence decay profiles lead to lifetime values New Eu2+ doped hybrid organic–inorganic nanocomposites lower than those usually recorded for Nd3+ in high temperasynthesized at room temperature.The Eu2+ ion is particularly ture processed glassy or crystalline matrices. These hybrid unique because its broad luminescence band 4f65d1A4f7 is materials can be described as a siloxane polymeric network strongly host dependent with emission wavelengths extending made from (CH3SiO3/2) units crosslinked by mixed nanoaggrefrom the UV to the red range of the electromagnetic spec- gates made from zirconium oxo and neodymium oxo species.trum.36 Therefore, the luminescent properties of Eu2+-doped Depending on the Nd3+ concentration, two diVerent optical solids have been intensively studied during the past three features are observed in these hybrid coatings. Both are well decades.These studies have led to the use of these compounds correlated with the structure. (1) For low Nd3+ concentration, as phosphors, notably blue-emitting Eu2+5BaMgAl10O17 in the initial non-exponential part of the decay profiles vary lamp and plasma display panels and UV-emitting linearly on a root-mean-square time scale showing the strong Eu2+5SrB4O7 for medical applications and skin tanning.cross-relaxation phenomena. Even if a few nOMH vibrations Crystalline or glassy Eu2+ doped materials are usually pro- are observed in the infrared spectrum, these hydroxy groups cessed at relatively high temperatures.36–39 Moreover, the are not located close to the Nd3+ ions and at low concentration synthesis and the stabilization of europium in the divalent the main non-radiative de-excitation mechanisms are the state under mild synthetic conditions is not an easy task.For Nd3+–Nd3+ interactions. The fact that these nanocomposites the first time, we reported recently the room temperature are made with segregated metal oxo species (the dispersion is synthesis of new Eu2+ doped hybrid materials together with not statistical all over the sample) leads to Nd3+–Nd3+ their absorption and emission properties.40 These hybrids are interactions even for low rare earth concentrations.The slope obtained though the hydrolysis and condensation of diethoxy- measured at long times give an indication of the Nd3+ methylsilane (MDES), methyltriethoxysilane (TREOS) and fluorescence quantum yield in this hybrid coating.The ratio g zirconium tetrapropoxide precursors in the presence of of the experimental lifetime at low concentration over the europium trichloride. calculated radiative lifetime (g=0.35) indicates that approxi- Dehydrocondensation of organic hydrosilanes with silanols mately 35% of the relaxation occurs radiatively at low Nd3+ is one of the common methods for the synthesis of the siloxane concentration. This is a high value for a room temperature linkage.41 This reaction, which occurs with the evolution of sol–gel processed material and this is in agreement with the hydrogen gas, has been described as follows:41 fact that, at low concentration, Nd3+ ions form clusters at relatively long distances from the remaining hydroxy groups.OSiMH+HOMSiOCA catalystOSiMOMSiO+H2 (2) When the Nd3+ concentration increases, interactions increase as the Nd3+–Nd3+ distance becomes shorter in the nanoaggregates. Simultaneously when the neodymium concen- In this sense, alkoxide precursors containing SiMH groups show the possibility of using the SiMH groups as an in situ tration increases (the zirconium concentration remaining constant) the probability of finding neodymium ions at the surface reducing agent which allows the formation of metal/silica nanocomposites.42 of the metal oxo species increases, rendering the rare earth prone to interactions with some remaining hydroxy groups In our study,40 the in situ formation of hydrogen provided by the cleavage of the SiMH bonds was used to generate, located close to the surface.As a consequence, short lifetimes are observed. For a concentration of 4×1021 Nd3+ ions cm-3 during the first step of hydrolysis and condensation reactions, europium in the divalent state. the quantum yield is less than 0.01%. This behavior characterizes non-radiative de-excitation processes due to strong A typical absorption spectra (Fig. 8) of these is constituted by a broad absorption band in the UV range (200–400 nm) energy transfers between Nd3+ but also non-radiative deexcitation occurring via the filling of the 4F3/2–4I15/2 energy attributed to the 4f75d0A4f65d1 (Eu2+) transition. The emission spectra (Fig. 9) of the corresponding hybrids gap by nOMH vibrations. Usually, according to the energy gap law, multiphonon non-radiative contributions do not exceed recorded under excitation at 355 nm show a broad emission corresponding to the interconfigurational 4f65d1A4f75d0 50% taking into account the energy diVerence around 5500 cm-1 between the 4F3/2 and 4I15/2 levels. In the present transition centered at 430 nm (ca. 23250 cm-1) and a intracon- 38 J. Mater. Chem., 1999, 9, 35–44are only achieved in a non-centrosymmetric environment, we first demonstrated that orientation of organic chromophores can be performed in hybrid sol–gel matrices43–46 by using electrical field induced second harmonic (EFISH) or corona electrical field poling techniques. Organic molecules such as N-(3-triethoxysilylpropyl )-2,4-dinitrophenylamine (TSDP) were chemically bonded to the oxide backbone of gels.The chemical bonding of the dye to the sol–gel matrix allowed dye concentration to be increased without any crystallization eVects.43–46 A first generation of sol–gel matrices was synthesized by copolymerization of silicon alkoxysilanes [TSDP and SiHCH3(OEt)2] and zirconium propoxide precursors.45 The sols were deposited as transparent coatings and exhibited after corona poling an SHG response of 1.6 pm V-1.45 Even if in this first generation of sol–gel matrices relaxation of the organic chromophores occurred over several hours, these results suggested the feasibility of poling techniques into hybrid inorganic sol–gel matrices more ionic than classical polymers.Fig. 8 Absorption spectra of europium doped hybrid xerogels. Consequently a range of opportunities for the synthesis of optical sol–gel devices with eYcient second harmonic properties was discovered. Since then, there has been increasing interest in second order NLO materials synthesized via sol–gel chemistry.47–60 The optimization of the second order NLO response of hybrid sol–gel matrices with grafted chromophores is currently under investigation by several research groups.Several strategies are used to improve the NLO response of the hybrid coatings.43–60 (i) The intrinsic NLO response of the dye can be increase by using chromophores such as N-(4- nitrophenyl )-L-prolinol (NPP) or disperse red one (DR1) derivatives which exhibit higher non linearities than nitroaniline ones. (ii) The chromophore relaxation can be controlled by increasing the matrix rigidity.This point is without doubt the most important in order to be able to make eYcient NLO devices. The modification of the binary composition (siloxane –crosslinker), the nature of theM(OR)4 crosslinking alkoxide [SiR¾x(OR)4-x–M(OR)4: R¾=any NLO chromophore; M=Zr, Si, Ti, etc.), and the processing of these hybrid Fig. 9 Emission spectrum of europium doped hybrid xerogel (lexc= materials in the presence of polymers with well known mechan- 355 nm).ical properties such as methyl methacrylates or polyimides, are the most commonly used strategies to minimize dye figurational 4f–4f Eu3+ emission in the longer wavelengths. relaxation. The strategies we have used to improve the NLO Several bands are obtained corresponding to the 5D0A7F0,1,2,3 response of hybrid materials will be llustrated in the two transitions. A Stokes’ shift value of the Eu2+ luminescence following sections.around 9000 cm-1 is obtained, this shift between the absorption and emission energies of Eu2+ located in an oxygen ligand TSPD/TMOS based hybrids with NLO properties.58,59 The field has been assigned to a combination of crystal field and second generation of hybrids investigated were made via nephelauxetic eVects.36 These hybrid structures contain oxygen hydrolysis and co-condensation of tetramethoxysilane atoms in higher coordination number environments (highly (TMOS) and N-(3-triethoxysilylpropyl )-2,4-dinitrophenylam- coordinated by metal atoms m3-O–Zr/Eu or m4-O–Zr/Eu) ine (TSDP) [Fig. 10(a)] precursors (T and Q are common which produce Eu2+ emission at longer wavelengths.Moreover notations referring to the oxo trifunctional R¾MSiO3 and distortion of the oxygen polyhedra from ideal coordination tetrafunctional SiO4 central units, respectively). FTIR, 17O geometry results in a large Stokes’ shift. First measurements and 29Si NMR experiments indicated the existence of linear indicate a Eu2+5Eu3+ concentration ratio of about 551.The and cyclic siloxane (T–T)a oligomers and silica (Q–Q)b units high Eu2+ content is probably related to the more eYcient reductive medium provided by the initial mixture of the europium trichloride with the MDES and TREOS precursors. Moreover, the intensity of the Eu2+ luminescence did not change when the xerogels were stored in air for several months, showing that Eu2+ ions are eYciently trapped inside the hybrid matrix. 2-C Quadratic NLO properties of siloxane based hybrids Most of the sol–gel optics research devoted to non-linear optic (NLO) materials was initially related to third order processes which are compatible with the isotropy of amorphous sol–gel matrices. Organic molecules inside amorphous sol–gel matrices are in general randomly oriented thus ruling out the emission O2N NO2 NH Si(OEt)3 TSDP O2N N N N CH2CH2OCONH-(CH2)3Si(OEt)3 CH2CH2OCONH-(CH2)3Si(OEt)3 CH3 (a) (b) Fig. 10 Graftable NLO dyes: (a)=TSDP (b)=ICTES-Red17. of second harmonic generation. As second order non-linearities J. Mater. Chem., 1999, 9, 35–44 39linked through stable T–O–Q bridges formed in the early also strongly depend on their thermal history.Chemical crosslinking is not complete at gelation or even after RT air stages of the process. Films of thickness 1–5 mm were easily obtained through spin-coating. In such systems gelation prob- drying as shown by 29Si NMR experiments.59 Upon ageing and curing the chemical reactions continue towards com- ably occurs through the crosslinking of siloxane polymers with Q silica based species.The degrees of condensation of T and pletion. Consequently, given suYcient time and temperature to allow mobility of the species a network forming system Q units measured in the solid state by 29Si MAS NMR spectroscopy are much higher in xerogels than in sols and this continues to crosslink long after gelation. The increase of the density of crosslinks modifies the thermomechanical properties diVerence demonstrates that a large number of condensation and crosslinking reactions still occur upon solvent removal.of the hybrid as illustrated by the changes observed in Tg upon thermal curing.57 The mobility of the NLO chromophores, as observed by high-resolution solid-state 13C NMR spectroscopy, is also Another processing parameter which has great importance is the electrical field used to poled the NLO chromophores.correlated with the glass-transition phenomenon of the matrix observed by DSC.57 This glass transition phenomenon corre- Accelerated field induced curing must occur in these hybrid TSDP/TMOS materials. The high electrical field provided sponds to the glass transition of the polysiloxane network.Tg, the glass-transition temperature, increases with the TMOS during poling must favor crosslinking and interpenetration of both polymeric T and Q networks. This was described by content, while the apparent variation of heat capacity corresponding to Tg decreases. These results, as well as the analysis Haruvy and co-workers,48b who have shown that greatly accelerated curing occurs under ambient conditions on thin of the polarization transfer in MAS/CP/DD 29Si NMR experiments, are consistent with the relatively high degree of inter- films processed from siloxane resins prepared by the sol–gel process when they are exposed to an intense corona-discharge penetration of T and Q units.Therefore, these hybrid TSDP/TMOS coatings can be described as nanocomposites field.The corona cured sol–gel films exhibited a more compact matrix as manifested by the lower mobility of the embedded made of silica rich domains and siloxane rich domains. Many Q and T species are mutually sequestered at the nanometer chromophores and a more hydrophilic surface than thermally cured ones. They suggest that the field induced removal of scale.Their microstructure is schematically pictured in Fig. 11. The white parts correspond to the silica-rich phase inside condensate small molecules and solvents allows better completion of the reactions and more eYcient crosslinking.48 which some T units (black dots) are sequestered. The black spheres correspond to the polysiloxane rich phase which Compared to the tremendous amount of work and time devoted to polymeric NLO materials, NLO materials made participates in the glass-transition phenomenon and contains Q units (white dots). The sizes of the polysiloxane and silica by sol–gel are still in their infancy. For these system, depending on chemical composition, the SHG values range between 2.5 domains depend not only on the chemical composition but also on the drying procedure and consequently on the solvent and 10 pm V-1.47,57 Moreover, the sol–gel materials described in this work have Tg values in the range of 30–70 °C, well and sample thickness.The TSDP/TMOS ratio, proton concentration, hydrolysis below the state of the art obtained with pure organic polymeric materials based on polyimides61 which are highly non linear ratio, sequence of mixing the reagents and ageing time of the sol are the chemical parameters that should directly influence and stable for hundred of hours at temperatures higher than 100 °C.However, the excellent knowledge of such systems a and b values characterizing the lengh of the constituent linear and cyclic siloxane (T–T)a oligomers and silica (Q–Q)b allowed us within a short period of time to design a third generation of hybrids with a highly improved NLO response.units respectively. However, it has been demonstrated that the mechanical properties of hybrid siloxane–oxide materials, and thus the ICTES-Red17/TMOS based hybrids with NLO properties. The third generation of hybrid organic–inorganic nanocom- relaxation behavior of chromophores grafted in these matrices, posites was designed on the basis of the following specifications: the NLO dye must have a high NLO response, it must be anchored by more than one trifunctional link and silica was kept as the crosslinking agent because coatings of better optical quality were usually obtained with binary silica–siloxane materials.56 In order to be able to perform double grafting of an NLO chromophore, the Red 17 [4-(amino-N,N-diethanol )-2-methyl-4¾-nitroazobenzene] with a very eYcient quadratic hyperpolarizability [b(0) (Red 17)=55×10-30 esu] was functionalized with two alkoxysilyl groups by a coupling reaction between the dye and 3-isocyanatopropyltriethoxysilane (ICTES).56,58 The resulting alkoxysilyl functionalized NLO precursor, ICTES-Red 17 [Fig. 10(b)] was hydrolyzed and co-condensed with TMOS in order to obtain the hybrid siloxane–silica nanocomposite.From the resulting sols, coatings with a thickness of a few mm can be deposited. The resulting hybrid materials do not exhibit Tg according to DSC results. Non-resonant second-order non-linearities as high as 150–200 pm V-1,58,62 measured on these hybrid systems, with significant long-term stability (10% of signal lost after 20 days) have been reported.58 The thermal stability at 80 °C has been shown to be excellent, making the ICTES-Red 17/TMOS systems competitive candidates for non-linear optics systems.Chemical characterization (FTIR, 29Si MAS NMR, UV–VIS spectroscopy) and thermal assisted in situ poling studies performed on these coatings revealed the importance of the processing and history of these systems.Three param- Fig. 11 Schematic representation of TSDP-TMOS based hybrids.52 eters are of paramount importance. (i) Aging of the solution 40 J. Mater. Chem., 1999, 9, 35–44has been shown to greatly influence the amplitude of the final Moreover, these clusters exhibit a high versality for the design of hybrids (Fig. 13).{(RSn)12O14(OH)6}2+2X- can non-linear signal. This results from improvement of crosslinking eYciency and from modifications of the distribution be assembled through organic networks by using the covalent interface provided by the SnKC bond or by using the ionic between cyclic and linear siloxane species. (ii) Thermal precuring of the samples at 150 °C was found to markedly improve interface associated with the charge compensating anions X- or even by using both interfaces.In the first case the organic the non-linear response as well as its stability. (iii) Optical poling recently tested in sol–gel derived matrices can also be moiety carried through the SnKCsp3 links should be polymerizable (R=butenyl, propylmethacrylate, propylcrotonate, 4- used to improve the chromophore anisotropy63 These very reproducible results58,62 are very promising, styrylbutyl, etc.).In the second case charge compensating organic dianions must be able to bridge the clusters. This can however as far as NLO devices are concerned they must be completed by measurements of electro-optical eYciency, be performed by using dicarboxylates,65 or a,v-telechelic macromonomers terminated by carboxylic or sulfonic groups.67 waveguiding properties and the evaluation of the optical losses.As an example, the coupling of these clusters by carboxymethyl terminated PEG macromonomers67 is schematically shown 3 Tin oxo species based hybrid materials in Fig. 14. Another strategy could be to use polymerizable anions Tin is a very interesting element because its characteristics make it intermediate between silicon and the transition metals.(methacrylate, 2-acrylamido-2-methylpropane-1-sulfonate, etc.) as monomers for organic polymerizations reactions.66,68a Like the latter, tin exhibits several coordination numbers (generally from 4 to 6) and coordination expansion makes By a simple acid-base reaction, the oxo-hydroxo butyltin macrocation, {(BuSn)12O14(OH)6}2+, was functionalized hydrolysis–condensation reactions of tin alkoxides fast.But, as for silicon, the Sn–Csp3 bond is stable, especially towards with 2-acrylamido-2-methylpropane-1-sulfonate, aVording nanobuilding blocks with two highly polymerizable groups.68a nucleophilic agents such as water. This last characteristic allows one to chemically link organic moieties to the tin oxo For the first time,68a the direct polymerization of such functionalized oxo-hydroxo butyltin nanoclusters has been polymers/oligomers but it also reduces the inorganic functionality of tin and therefore favors the formation of oxo successfully performed, yielding hybrid materials in which the nanosized inorganic component is perfectly defined.Two types clusters.These oxo clusters can be used as nanobuilding blocks in the design of new hybrid materials.64–72 of organic components are found in such materials. The butyl groups covalently bound onto tin atoms, and, more import- The nanobuilding block [(RSn)12(m3-O)14(m-OH)6]2+, the structure of which is shown in Fig. 12 can be obtained through antly, poly(2-acrylamido-2-methylpropane-1-sulfonate) chains which interact through electrostatic interactions with the several chemical pathways: hydrolysis of RSn(OPri)3 or RSnCl3 or by refluxing in toluene butyltin hydroxide oxide oxo-hydroxo butyltin macrocations and aVord the crosslinking. [BuSnO(OH)] in the presence of sulfonic acids (R¾SO3H)64–71 and more recently Jousseamme et al.opened a new route to Such an approach to the construction of tin-based hybrid materials from bifunctional nanobuilding blocks was pre- this cluster through hydrolysis of functionalized trialkynylorganotin precursors.72 viously attempted with pure {(BuSn)12O14(OH)6}- {O2CC(CH3)NCH2}2 but failed as its homopolymerization This compound is made of a tin oxo-hydroxo cluster with a equal numbers of six- and five-coordinate tin atoms.This appeared impossible.66 Addition of a co-monomer [CH3O2CC(CH3)NCH2] allowed the polymerization, but cage-like cluster is surrounded by twelve organic chains (butyl, butenyl, etc.) which prevent further condensation. Depending recent results have indicated that little crosslinking was achieved, the methacrylate charge compensating anions acting on the synthesis conditions the 2+ positive charge can be compensated by a large variety of anions (OH+,Cl-, sulfon- mainly as termination agents.66b These diYculties may be related to the fairly large molecular weight of the precursor ates, carboxylates, etc.).The position of the charge compensating anions in the structure indicates that the 2+ charge is (ca. 2600 g mol-1), but also to the shortness of the methacrylate functional anions which induce high steric hindrance.equally located at both cage poles, where six-coordinate tin atoms form hydroxylated [RSn(OH)]3O trimers. The second reason seems to prevail, as the use of AAMPS, where the polymerizable acrylamido group is more distant This cluster is confirmed both in solution by 119Sn NMR [it is characterized by two chemical shifts located at about -280 from the anionic anchoring head, allows the formation of a hybrid polymer by simple homopolymerization. More work is and -450 ppm (R=butyl or butenyl )] and a set of scalar tin–tin coupling satellites and in the solid state through 119MAS currently under way to obtain some information on the average length of the poly(2-acrylamido-2-methylpropane-1-sulfonate) NMR spectroscopy.69,71 Thus, this cluster can be followed easily throughout the polymerization or crosslinking reactions chains, which are probably short owing to strong steric hindrance.needed to tranform it into a hybrid material. As a consequence {(RSn)12O14(OH)6}2+ is a good nanobuilding block for the synthesis of well defined tin-oxo based hybrid materials that can be used as models. 4 Molecular design of transition metal alkoxides for the synthesis of hybrid organic–inorganic copolymers The chemical tailoring performed with systems containing a SiMC bond cannot be directly extended to pure transition metals because the more ionic MMC bond is broken down upon hydrolysis. Organic modification can however be performed by means of strong complexing ligands.The best are b-diketones and allied derivatives, polyhydroxylated ligands such as polyols, and a- or b-hydroxyacids. These ligands (HL) react readily with transition metal alkoxides M(OR)4 (M= Ce, Ti, Zr, etc.) to yield new precursors M(OR)3-x(L)x.73,74 Upon hydrolysing these new precursors, most of the alkoxy groups are quickly removed while all strong complexing ligands Fig. 12 Molecular structure of [(RSn)12(m3-O)14(m-OH)6 ]2+. cannot be completely removed. Complexing ligands appear to J. Mater. Chem., 1999, 9, 35–44 41Fig. 13 Schematic representation of the strategies that can be used from a hybrid [(RSn)12(m3-O)14(m-OH)6 ]2+2X- nanobuilding block. by partial hydrolysis of the alkoxy groups and radical polymerization of the allyl functions. However polymerization of allyl functions is slow and the degree of polymerization remains low.More reactive methacrylic acid can also be used as a polymerizable chelating ligand. The sol–gel synthesis of zirconium oxide based monoliths synthesized by UV copolymerization of zirconium oxide sols and organic monomers was recently reported.8 However, as carboxylic functions are weak ligands, they are largely removed upon hydrolysis2,75 and thus a large number of the chemical bonds between organic and inorganic networks is lost in the sol state.Therefore a new approach was chosen with diVerent ligands, such as acetoacetoxyethylmethacrylate (AAEM) and methacrylamidosalicylic acid (MASA), which contain both a strong chelating part and a highly reactive methacrylate group.6 Zirconium–oxo-PAAEM copolymers were synthesized from zirconium propoxide modified at the molecular level with AAEM.6 These hybrid organic–inorganic copolymers are made of zirconium oxo-polymers and polymethacrylate chains.The zirconium oxo species, in which zirconium is coordinated to seven oxygens, are chemically bonded to methacrylate chains Fig. 14 Schematical representation of the tin oxo-hydroxo clusters hybrids obtained through crosslinking of [(RSn)12O14(OH)6 ]2+ with through the b-diketo complexing function.The complexation a,v-PEG carboxylates.67 ratio (AAEM/Zr) is the key parameter which controls the structure and the texture of these hybrid materials (schematic structure Fig. 15). Careful adjustment of this parameter leads be quite stable towards hydrolysis because of chelate and steric to the tailoring of the ratio between organic and inorganic hindrance eVects.Thus, they allow organic groups to be components and also to zirconium oxo species with more or anchored to transition metal oxo-polymeric species and allow less open structures. The inorganic/organic ratio increases the synthesis of new hybrid organic–inorganic materials. when the complexation ratio decreases.For a high com- Organically modified TiO2 gels, which give photochromic plexation ratio (0.75) both networks interpenetrate intimately coatings, were synthesized from an allyl acetylacetone modified Ti(OBun)4 alkoxide.74 Double polymerization was performed at the nanometer scale, while for a low ratio (0.25) the size of 42 J.Mater. Chem., 1999, 9, 35–44References 1 C. J. Brinker and G. W. Scherrer, Sol–Gel Science, The Physics and Chemistry of Sol–Gel Processing, Academic Press, San Diego, CA, 1990. 2 (a) J. Livage, M. Henry and C. Sanchez, Prog. Solid State Chem., 1988, 18, 259; (b) C. Sanchez, F. Ribot and S. DoeuV, Inorganic and Organometallic Polymers with Special Properties, ed.R. M. Laine, NATO ASI Series 206, Kluwer, New York, NY, 1992, p. 267. 3 (a) G. L. Wilkes, B. Orler and H. H. Huang, Polym. Prep., 1985, 26, 300; (b) G-S. Sur and J. E. Mark, Eur. Polym. J., 1985, 21, 1051; (c) H. Schmidt, A. Kaiser, H. Patzelt and H. Sholze, J. Phys., 1982, 43, C9-275. 4 (a) B.M. Novak, Adv. Mater., 1993, 5, 422; (b) Proceeding of the First European Workshop on Hybrid Organic–Inorganic Materials, ed.C. Sanchez and F. Ribot, New J. Chem., 1994, 18; Fig. 15 Schematic structures of class II hybrid materials made from (c) U. Schubert, N. Hu�sing and A. Lorenz, Chem.Mater., 1995, 7, zirconium n-propoxide complexed by acetoacetoxyethylmethacrylate 2010; (d) D. A. Loy and K. J. Shea, Chem. Rev., 1995, 95, 1431; (AAEM).6 (e) J. Sol–Gel Sci.Technol., 1995, 5; ( f ) P. Judenstein and C. Sanchez, J. Mater. Chem., 1996, 6, 511; (g) Mate�riaux Hybrides, Se� rie Arago 17, Masson, Paris, 1996; (h) C. Sanchez and F. Ribot, New J. Chem., 1994, 18, 1007; (i) U. Schubert, J. Chem. the inorganic domains increases to the sub-micrometer range. Soc., Dalton Trans., 1996, 3343; ( j) A. Morikawa, Y. Iyoku, M. Kakimoto and Y.Imai, J. Mater. Chem., 1992, 2, 679; Organic and inorganic growths are not independent and such (k) Y. Chujo and T.Saegusa, Ad. Polym. Sci., 1992, 100, 11; systems probably exhibit similar behavior to the so-called (l ) Hybrid Organic–Inorganic Composites, ed. J. E. Mark, interpenetrating polymer networks. C. Y. C. Lee and P. A. Bianconi, American Chemical Society, Washington, DC, 1995; (m) F.Ribot and C. Sanchez, Comments Inorg. Chem., 1998, in press. 5 (a) Better Ceramics Through Chemistry VI, ed. A. Cheetham, 5 Conclusions C. J Brinker, M. MacCartney, C. Sanchez, D. W. Schaefer and G. L. Wilkes, Mater. Res. Soc. Symp. Proc., Materials Research The combination at the nanosize level of inorganic and organic Society, Pittsburgh, PA, 1994, vol. 435; (b) Better Ceramics or even bio-active components in a single material makes Through Chemistry VII: Organic/Inorganic Hybrid Materials, ed.accessible an immense new area of materials science that has B. K. Coltrain, C. Sanchez, D. W. Schaefer and G. L. Wilkes, extraordinary implications for developing novel multi-func- Mater. Res. Soc. Symp. Proc., Materials Research Society, tional materials exhibiting a wide range of properties.This Pittsburgh, PA, 1996, vol. 435; (c) Sol–Gel Optics I, ed. J. D. Mackenzie and D. R. Ulrich, Proc. SPIE, Washington, 1990, fascinating new field of research brings together scientists vol. 1328; (d) Sol–Gel Optics II, ed. J. D. Mackenzie, Proc. SPIE, working in many diVerent domains. Among soft chemistry Washington, 1992, vol. 1758; (e) Sol–Gel Optics III, ed.processes, sol–gel chemistry oVers versatile access to the chemi- J. D. Mackenzie, Proc. SPIE, Washington, 1994, vol. 2288. cal design of new hybrid organic–inorganic materials. Many 6 (a) C. Sanchez and M. In, J. Non-Cryst. Solids 1992, 147–148, 1; new combinations between inorganic and organic or biological (b) M. In, C. Ge�rardin, J. Lambart and C.Sanchez, J. Sol–Gel Sci. Technol., 1995, 5, 101. components will probably appear in the future. Yet, better 7 (a) U. Schubert, E. Arpac, W. Glaubitt, A. Helmerich and understanding and control of the local and semi-local structure C. Chau, Chem. Mater., 1992, 4, 291; (b) C. Barglik-Chory and of these materials is an important issue, especially if tailored U. Schubert, J. Sol–Gel Sci.Technol., 1995, 5, 135. properties are sought. 8 R. Naß and H. Schmidt, in ‘Sol–Gel Optics I’, ed. J. D. Mackenzie To achieve such a control of the material structure, the and D. R. Ulrich, Proc. SPIE,Washington, 1990, vol. 1328, p. 258. 9 H. Schmidt and B. Seiferling, Mater. Res. Soc. Symp. Proc., 1986, assembly of well defined nanobuilding blocks is an interesting 73, 739.approach. The inorganic components are nanometric and 10 (a) D. Levy and D. Avnir, J. Phys. Chem., 1998, 92, 734; monodispersed. Their structures are perfectly defined, which (b) D. Levy, S. Einhorn and D. Avni, J. Non-Cryst. Solids, 1989, probably facilitates the characterization of the final materials. 113, 137. The variety found in the nanobuilding blocks (nature, struc- 11 Proceedings of the Second International Symposium on ture, functionality) and links, associated with diVerent assemb- Photochromism, ed.R. C. Bertelson, Chroma Chemicals Inc, Dayton, Ohio, USA, 1997. ling strategies, allow very diVerent types of architectures and 12 D. Preston, J. C. Pouxviel, T. Novinson, W. C. Kaska, B. Dunn organic–inorganic interfaces to be formed.ver, the and J.I. Zink, J. Phys. Chem., 1990, 94, 4167. step-by-step preparations of the materials usually allows some 13 H. Nakazumi, R. Nagashiro, S. Matsumoto and K. Isagawa, control over their semi-local structure. One can also combine Sol–Gel Optics III, ed. J. D. MacKenzie, Proc. SPIE,Washington, the nanobuilding blocks approach with the use of organic 1994, vol. 2288. 14 L. Hou, B.HoVmann, M. Menning and H. Schmidt, J. Sol–Gel templates that self-assemble and allow, through weak forces, Sci. Technol., 1994, 2, 635. control of the assembly step. With the aim of organizing/ 15 L. Hou, B. HoVmann, H. Schmidt and M. Menning, J. Sol–Gel structuring the nanobuilding blocks prior to their assembly, Sci. Technol., 1997, 8, 923, 927. one can also functionalize them with mesogenic groups and 16 L.Hou, M. Menning and H. Schmidt, Proc Eurogel’92, 1992, give them self-organizing properties. p. 173. 17 J. Biteau, F. Chaput and J. P. Boilot, J. Phys. Chem., 1996, 100, As described by Mann et al,76 the traditional view of 9024. inorganic solids as condensed matter is currently being 18 D. Levy, Chem. Mater., 1997, 9, 2666. reshaped by new insight in materials synthesis.77–79 A new 19 B.Schaudel, C. Guermeur, C. Sanchez, K. Keitaro and J. Delaire, paradigm, organized matter chemistry, is being formulated. In J. Mater. Chem., 1997, 7, 61. this field the sol–gel chemistry of organized matter is a very 20 S. Dire�, F. Babonneau, C. Sanchez and J. Livage, J. Mater. Chem., 1992, 2, 239. promising field of research which is just beginning to be 21 F.Babonneau, Polyhedron, 1994, 13, 1123. explored. Hybrids of both Class I and II4h and more particu- 22 C. Guermeur and C. Sanchez, forthcoming paper. larly those obtained through the ‘nanobuilding block 23 Applied photochromic polymers systems, ed. C. B. McArdle, approach’ will have paramount importance in the exploration Gordon and Breach Science Publishers, New York, 1991.the theme of ‘synthesis with construction’ of hierarchically 24 S. K. Yuh, E. Bescher, F. Babonneau and J. D. Mackenzie, Mater. Res. Soc. Symp. Proc., 1994, 346, 803. organized (structure and function) materials. J. Mater. Chem., 1999, 9, 35–44 4325 N. Koslova, B. Viana and C. Sanchez, J. Mater. Chem., 1993, M. Dumont, New J. Chem., 1996, 20, 13; (b) B. Lebeau, C.Sanchez, S. Brasselet and J. Zyss, Mater. Res. Soc. Symp. Proc., 3, 111. 1996, 435, 395; (c) B. Lebeau, C. Guermeur and C. Sanchez, 26 L. R. Mathews and E. T. Knobbe, Mater. Res. Soc. Symp. Proc., Mater. Res. Soc. Symp. Proc., 1994, 346, 315. 1993, 286, 259. 57 B. Lebeau, J. Maquet, C. Sanchez, E. Toussaere, R. Hierle and 27 I. M. Thomas, S. A. Payne and G. D. Wilke, J.Non-Cryst. Solids, J. Zyss, J. Mater. Chem., 1994, 4, 1855. 1992, 151, 183. 58 B. Lebeau, C. Sanchez, S. Brasselet and J. Zyss, Chem. Mater., 28 I. M. Thomas, S. A. Payne and G. D. Wilke, J. Non-Cryst. Solids, 1997, 9, 1012. 1992, 151, 183. 59 B. Lebeau, J. Maquet, C. Sanchez, F. Beaume and F. Laupre�tre, 29 A. A. Kaminskii, in Laser Crystals, Springer Verlag, Berlin, J. Mater. Chem., 1997, 7, 989.Heidelberg, New York, 2nd edn., 1981, pp. 361–433. 60 C. Sanchez and B.Lebeau, Pure Appl. Opt., 1996, 5, 689. 30 M. Lecomte, B. Viana and C. Sanchez, J. Chim. Phys., 1991, 61 R. F. Shi, M. H. Wu, S. Yamada, Y. M. Cai and A. F. Garito, 88, 39. Appl. Phys. Lett., 1993, 63, 1173. 31 B. Viana, N. Koslova, P. Aschehoug and C.Sanchez, J. Mater. 62 D. Blanc, P. Peyrot, C.Sanchez and C. Gonnet, Opt. Eng. Chem., 1995, 5, 719. Integrated Opt., 1998, 37, 1203. 32 C. Guizard, J.C. Achddou, A. Larbot, L. Cot, G. Le Flem, 63 C. Fiorini, F. Charra, J. M. Nunzi, I. D. W. Samuel and J. Zyss, C. Parent and C. L. Lurin, in ref. 5(c), p. 208. Opt. Lett., 1995, 20, 2469. 33 M. Genet, V. Brandel, M. P. LaHalle and E. Simoni, in ref. 5(c), 64 F. Ribot, F. Banse and C.Sanchez, Mater. Res. Soc. Symp. Proc., p. 194. 1994, 346, 121. 34 W. Nie, B. Dunn, C. Sanchez and P. Griesmar, Mater. Res. Soc. 65 F. Ribot, F. Banse, F. Diter and C. Sanchez, New J. Chem., 1995, Symp. Proc., 1992, 271, 639. 19, 1145. 35 M. Canva, P. Georges, G. Lesaux, A. Brun, F. Chaput and 66 (a) F. Ribot, F. Banse, C. Sanchez, M. Lahcini and J. P. Boilot, J. Non-Cryst.Solids, 1992, 147, 627. B. Jousseaumme, J. Sol–Gel Sci. Technol., 1997, 8, 529; 36 G. J. Dirksen and G. Blasse, J. Solid State Chem., 1991, 92, 591. (b) L. Angiolini, D. Caretti, C. Carlini, R. De Vito, F. T. Niesel, 37 A. Diaz and D. A Keszler, Chem. Mater. 1997, 9, 2071. E. Salatelli, F. Ribot and C. Sanchez, J. Inorg. Organomet. Polym., 38 A. Diaz and D. A Keszler, Mater. Res.Bull., 1996, 31, 147. 1998, 7, 151. 39 H. F. Folkerts and G. Blasse, J. Mater. Chem., 1995, 5, 1547. 67 (a) F. Ribot,C. Eychenne-Baron and C. Sanchez, Mater. Res. Soc. 40 E. Cordoncillo, P. Escribano, B. Viana and C. Sanchez, J. Mater. Symp. Proc., 1996, 435, 43; C. Eychenne-Baron, F. Ribot and Chem., 1998, 8, 1507. C. Sanchez, J. Organomet. Chem., 1998, 567, 137. 41 J. Chrusciel and Z.Lasocki, Pol. J. Chem., 1983, 57, 121. 68 F. Ribot, C. Eychenne-Baron and C. Sanchez, Mater. Res. Soc. 42 R. Campostrini and S. Dire�, Eurogel Proceedings, Advanced Symp. Proc., 1998, 519, in press. Materials and Processes by Sol–Gel Techniques, Colmar, 1992, ed. 69 F. Banse, F. Ribot, P. Tole�dano, J. Maquet and C. Sanchez, Inorg. S. Vilminot, Strasbourg, 1995, p. 307. Chem., 1995, 34, 6371. 43 G. Pucetti, I. Ledoux, J. Zyss, P. Griesmar and C. Sanchez, Polym. 70 H. PuV and H. Reuter, J. Organomet. Chem., 1989, 373, 173. Prepr., 1991, 32, 61. 71 D. Dakternieks, H. Zhu, E. R. T. Tiekink and R. J. Colton, 44 J. Zyss, G. Pucetti, I. Ledoux, P. Griesmar, J. Livage and J. Organomet. Chem., 1994, 476, 33. C. Sanchez, Fr. Pat., 1152, March 1991. 72 P. Jaumier, B.Jousseaume, M. Lahcini, F. Ribot and C. Sanchez, 45 (a) E. Toussaere, J. Zyss, P. Griesmar and C. Sanchez, Non Linear Chem. Commun., 1998, 369. Optics, 1991, 1, 349; (b) C. Sanchez, P. Griesmar, E. Toussaere, 73 (a) R.C.Mehrotra, R. Bohra and D. P. Gaur, Metal b-diketonates G. Puccetti, I. Ledoux and J. Zyss, Non Linear Optics, 1992, 4, 245. and Allied Derivatives, Academic Press, London, 1978; 46 P.Griesmar, C. Sanchez, G. Pucetti, I. Ledoux and J. Zyss, Mol. (b) D. C. Bradley, R. C. Mehrotra and D. P. Gaur, Metal Eng., 1991, 1, 205. Alkoxides, Academic Press, London, 1978; (c) L. G. Hubert- 47 (a) J. Kim, J. L. Plawsky, R. LaPeruta and G. M. Korenowski, Pfalzgraf, New J. Chem., 1987, 11, 663; ( f ) A. Leaustic, Chem. Mater., 1992, 4, 249; (b) J. Kim, J.L. Plawsky, E. Van F. Babonneau and J. Livage, Chem. Mater., 1989, 1, 248; Wagenen and G. M. Korenowski, Chem.Mater., 1993, 5, 1118. ( g) F. Ribot, P. Tole�dano and C. Sanchez, Chem. Mater., 1991, 3, 48 (a) Y. Haruvy and S. E. Weber, Mater. Res. Soc. Symp. Proc., 759; (h) C.Sanchez, M. In, P. Toledano and P. Griesmar, Mater. 1992, 271, 297; (b) Q. Hibben, E. Lu, Y. Haruvy and S.E.Weber, Res. Soc. Symp. Proc., 1992, 271, 669. Chem. Mater., 1994, 6, 761. 74 C. Sanchez, J. Livage, M. Henry and F. Babonneau, J. Non-Cryst. 49 F. Chaput, D. Riehl, Y Levy and J. P. Boilot, Chem.Mater., 1993, Solids, 1988, 100, 650. 5, 589. 75 S. DoeuV, M. Henry, C. Sanchez and J. Livage, J. Non-Cryst. 50 Y. Zhang, P. N. Prasad and R. Burzynski, Chem. Mater., 1992, Solids, 1987, 89, 206. 4, 851. 76 S. Mann, S. L. Burkett, S. A. Davis, C. E. Fowler, 51 Y. Nosaka, N. Tohriiwa, T. Kobayashi and N. Fujii, Chem. N. H. Mendelson, S. D. Sims, D.Walsh and N. T. Whilton, Chem. Mater., 1993, 5, 930. Mater., 1997, 9, 2300. 52 S. Marturunkakul, J. I. Chen, R. J. Jeng, S. Sengupta, J. Kumar 77 S. Mann and G. A. Ozin, Nature, 1996, 382, 313. and S. K. Tripathy, Chem. Mater., 1993, 5, 743; K.Izawa, 78 C. G. Go� ltner and M. Antonietti, Adv. Mater., 1997, 9, 431. N. Okamoto and O. Sugihara, Jpn. J. Appl. Phys., 1993, 32, 807. 79 (a) A. Monnier, F. Schuth, Q. Huo, D. Kumar, D. Margolese, 53 L. Kador, R. Fischer, D. Haarer, R. Kaseman, S. Bru� ck, R. S. Maxwell, G.D. Stucky, M. Krishnamurty, P. PetroV, H. Schmidt and H. Du� rr, Adv. Mater., 1993, 5, 270.A. Firouzi, M. Janicke and B. F. Chmelka, Scie54 (a) J. R. Caldwell, R. W. Cruse, K. J. Drost, V. P. Rao, A. K.- 1299; (b) A. Firouzi, D. Kumar, L. M. Bull, T. Besier, P. Sieger, Q. Hue, S. A. Walker, J. A. Zasadinski, C. Glinka, J. Nicol, D. Y. Jen, K. Y. Wong, Y. M. Cali, R. M. Mininni, J. Kenney, Margolese, G. D. Stucky and B. F. Chmelka, Science, 1995, 267, E.Binkley, L. R. Dalton, Y. Shi and W. H. Steier, Mater. Res. 1133; (c) D. Y. Zhao, P. D. Yang, Q. S. Huo, B. F. Chmelka and Soc. Symp. Proc., 1994, 328, 531; (b) Z. Yang, C. Xu, B. Wu, L. G. D. Stucky, Curr. Opin. Solid State Mater. Sci., 1998, 3, 111; (d) R. Dalton, S. Kalluri, W. H. Steier, Y. Shi and J. H. Bechtel, C. J. Brinker, Curr. Opin. Colloid Interface Sci., 1998, 3, 166; (e) Chem.Mater., 1994, 6, 1899; (c) H. W. Oviatt, K. J. Shea, S. Y. Lu, R. Ganguli, C. A. Drewien, M. T. Anderson, C. J. Brinker, Kalluri, Y. Shi, W. Steier and L. R. Dalton, Chem. Mater., 1995, W. Gong, Y. Guo, H. Soyez, B. Dunn, M. H. Huang and J. I. 7, 493. Zink, Nature, 1997, 389, 364; ( f ) G. H. Mehl and J. W. Goodby, 55 (a) F. Chaput, D. Reihl, Y. Levy and J.-P. Boilot, Chem.Mater., Chem. Ber., 1996, 129, 521. 1993, 5, 589; (b) D. Reihl, F. Chaput, Y. Levy, J.-P. Boilot, F. Kajzar and P. A. Chollet, Chem. Phys. Lett., 1995, 245, 36. 56 (a) B. Lebeau, C. Sanchez, S. Brasselet, J. Zyss, G. Froc and Paper 8/05538F 44 J. Mater. Chem., 1999, 9, 35–44 J O U R N A L O F C H E M I S T R Y Materials Feature Article Molecular design of hybrid organic–inorganic nanocomposites synthesized via sol–gel chemistry† C.Sanchez, F. Ribot and B. Lebeau Chimie de la Matie`re Condense�e, UMR CNRS 7574, Universite� Pierre et Marie Curie 4, place Jussieu, 75252 Paris, France. E-mail: clems@ccr.jussieu.fr Received 12th May 1998, Accepted 16th July 1998 The design, synthesis and some optical properties of hybrid as organic polymerization and lead to hybrid organic–inororganic –inorganic nanocomposites materials are pre- ganic copolymers.6–8 This article reviews some of the previous sented.The properties that can be expected for such work we have performed on the design and synthesis of hybrid materials depend on the chemical nature of their compo- organic–inorganic nanocomposites in which organics are nents, but they also depend on the synergy of these simply embedded or chemically bonded to the inorganic components.Thus, the interface in these nanocomposites gel network. is of paramount significance and one key point of their synthesis is the control of this interface. These nanocom- 2 Siloxane based hybrid materials posites can be obtained by hydrolysis and condensation reactions of organically functionalized alkoxide precur- Organic groups can be bonded to an inorganic network as sors.Striking examples of hybrids made from modified network modifiers or network formers. Both functions have silicon, tin and transition metal alkoxides are presented. been achieved in the so-called ORMOSILS.3,9 The precursors Some optical properties (photochromic, luminescence, of these compounds are organo-substituted silicic acid esters NLO) of siloxane based hybrids are also discussed.of general formula R¾nSi(OR)4-n, where R¾ can be any organofunctional group. If R¾ is a simple non hydrolyzable group 1 Introduction bonded to silicon through a SiMC bond, it will have a network modifying eVect (SiMCH3). On the other hand, if R¾ can react Sol–gel chemistry is based on the polymerization of molecular with itself (R¾ contains a methacryl group for example) or precursors such as metal alkoxides M(OR)n.1,2 Hydrolysis and additional components, it will act as a network former.3c condensation of these alkoxides lead to the formation of metal Network modifiers and network formers can also introduce oxo-polymers.The mild characteristics oVered by the sol–gel other physical properties (mechanical, hydrophobic, electro- process allow the introduction of organic molecules inside an chemical, optical, etc.).Several examples related to some inorganic network.3 Inorganic and organic components can optical properties of siloxane based hybrids will be now then be mixed at the nanometric scale, in virtually any ratio described.leading to so-called hybrid organic–inorganic nanocomposites. 4 These hybrids are extremely versatile in their composi- 2-A Photochromic properties tion, processing and optical and mechanical properties.5 The nature of the interface between the organic and inorganic Spiropyrans and spirooxazines are two of the fascinating components has been used recently to classify these hybrids families of molecules exhibiting photochromic properties.into two diVerent classes.4h Class I corresponds to all the Upon irradiation, the colorless spiropyran or spirooxazine systems where there are no covalent or iono-covalent bonds undergoes heterolytic CMO ring cleavage, producing colored between the organic and inorganic components. In such mate- forms of merocyanines (Fig. 1). rials, the various components only exchange interactions such The merocyanines may interact with their environment, i.e. as van der Waals forces, hydrogen bondings or electrostatic solvent, matrix, etc. leading to diVerent photochromic forces. In contrast, in class II materials, some of the organic responses. Levy and Avnir10 first demonstrated the important and inorganic components are linked through strong chemical role played by the dye–matrix interactions on the photobonds (covalent or iono-covalent).Numerous hybrid organic– chromic response of spiropyrans. They studied the photochroinorganic materials have been developed in the past few years. mism of spiropyrans trapped in sol–gel matrices synthesized This development yields many interesting new materials, with via polymerization of Si(OCH3)4 or RSi(OEt)3 (R=ethyl, mechanical properties tunable between those of glasses and those of polymers, with improved optical properties (eYciency, stability, new sensors, etc.) and with improved catalytic or membrane based properties.4g This field of materials research mainly arises from chemists’ skills and demonstrates the major role played by chemistry in advanced materials.Siloxane based hybrids4 can be easily synthesized because SiMCsp3 bonds are rather covalent and therefore they are not broken upon hydrolysis. Similar chemistry can be developed from tin alkoxides. This is not aplicable to transition metals for which the more ionic MMC bond is easily cleaved by water; complexing organic ligands must be used.Such groups can be functionalized for any kind of organic reactions such †Basis of the presentation given at Materials Chemistry Discussion NO NO2 SP N O N SO N O– NO2 N N O hn1 hn2, D hn1 hn2, D + Fig. 1 Molecular structures of SO and SP photochromic dyes. No. 1, 24–26 September 1998, University of Bordeaux, France. J. Mater. Chem., 1999, 9, 35–44 35methyl, etc.) precursors, and observed two types of photochromic behavior.When the photochromic dye is trapped within a hydrophilic domain of the matrix (domain containing residual SiMOH groups), the open zwitterionic colored forms are probably stabilized through hydrogen bonding with the acidic silanol groups present at the pore surface. The result of this stabilization is the observation of the colored forms before irradiation.These colored forms can be bleached by irradiation in the visible range. This has been termed ‘reverse photochromism’. On the other hand spiropyran dyes embedded in a more hydrophobic hybrid network made by hydrolysis of RSi(OEt)3 exhibit direct photochromism, i.e. the colorless form is stable without irradiation. Such photochromic behavior has been reported for many spiropyran or spirooxazine doped sol–gel matrices.10–17 Moreover, for hybrid organic–inorganic matrices containing diVerent chemical environments (hydrophilic and hydrophobic domains) a competition between direct and reverse photochromisms can be observed.17 However many fundamental questions still need to be considered.Little is Fig. 2 Cartoon of the D/Zrx matrices. known concerning the role of the photochc dye–matrix interactions in the kinetics of coloration and thermal fading.As far as photochromic devices are concerned the tuning between strong and fast photochromic coloration (high DA) and very fast thermal fading is needed. Usually spiropyran or spirooxazine doped sol–gel matrices or even spirooxazine doped polymeric matrices exhibit slow thermal fading.10–18 The photochromic behavior of a spiropyran SP (6-nitro- 1¾,3¾,3¾-trimethylspiro{2H-1-benzopyran-2,2¾-indoline}) and a spirooxazine SO (1,3,3-trimethylspiro{indoline-2,3¾(3H)- naph(2,1-b)(1,4)oxazine}) (Fig. 1) embedded within two new hybrid matrices have been recently studied.19 Photochromic properties of the SP or SO doped D/Zr hybrid matrices. The first kind of matrix obtained through hydrolysis and condensation of (CH3)2Si(OC2H5)2 (D) and appropriate amounts of Zr(OPrn)4 (Zr) is labelled D/Zrx, where Zr stands for the zirconium, x for the molar amount of zirconium.Fig. 3 Photocolouration and photobleaching of SP doped D/Zrx These D/Zrx matrices are hybrid nanocomposites made matrices19 [(a) x=10; (b) x=20; (c) x=30]. from polydimethylsiloxane species (chains, cycles etc.) crosslinked by zirconium oxopolymers.20–22 The zirconium oxopolymers are hydrophilic domains that still contain some photochromism is partially reversed and can be balanced by tuning the D/Zr ratio.hydroxo groups coming from residual ethanol or ZrMOH ligands.22 The size and the spacing between the ZrO2 based The thermal bleaching behavior of the D/Zr20 samples was fitted with a biexponential equation.The SP doped materials domains is about a few nanometers as evidenced by SAXS.22 17O MAS NMR and FTIR experiments show that inside these exhibited a very long bleaching time (about 24 h) while for the SO doped D/Zr20 materials the thermal fading was composites the hydrophobic polydimethylsiloxane species are interfaced with the zirconium oxo domains both through much faster.19 The kinetic data of the SP or SO doped D/Zr20 samples covalent ZrMOMSi bonds and through weak interactions (hydrogen and van der Waals bonds).19 The structure of the are similar to those reported for other modified sol–gel matrices or in organic polymers.17,23 As in organic polymers, the D/Zrx matrices is shown schematically in Fig. 2. D/Zrx matrices doped with SP or SO are lightly colored (pink with bleaching follows a biexponential equation which can be explained by an inhomogeneous distribution of free volumes SP or blue with SO) before irradiation. However the absorbance (A) in the visible region is weak in comparison in the gel. Moreover the presence of diVerent stereoisomers (cis or trans) could also account for this behavior. The diVerent with the total amount of embedded photochromic dyes.The amount of colored form depends on the x content. isomer–matrix interactions could explain the diVerent kinetics observed for SO and SP. Fig. 3 shows the photochromic behavior of SP doped D/Zrx gels for three compositions of zirconium. When the Zr amount The thermal fading is longer for SP doped hybrids than for SO doped ones.This phenomenon can be correlated to the increases, the A variation due to the coloration decreases while that due to the decoloration increases: there are more and fact that SP open forms are known for their tendancy to form zwitterionic species, while non charged quinonic species are more open forms in the gel. The amount of colored form increases proportionally with the x content. It is much higher usually favored for open SO molecules.Zwitterionic species can be strongly stabilized by hydrogen bonding with the for D/Zr30 samples than for D/Zr10 ones. This indicates that before irradiation the SO and SP dyes matrix, thus lowering the decay times of thermal fading. are roughly split into two populations.The colored merocyanine open forms of SO and SP are stabilized by hydrogen Photochromic properties of the SP or SO doped DH/TH hybrid matrices. The second kind of SP or SO doped matrix bonding within the hydrophilic regions made from the zirconium oxopolymers, while the closed SO and SP forms are prepared from the hydrolysis and cocondensation of (CH3)HSi(OC2H5)2 (DH) and HSi(OC2H5)3 (TH) precursors probably located in the environment of the hydrophobic polydimethylsiloxane chains.So for these D/Zrx matrices the is labelled DH70/TH30 (70/30 refers to the molar composi- 36 J. Mater. Chem., 1999, 9, 35–44few minutes under magnetic stirring. A solution of mixed alkoxides, Zr(OPrn)4 and a corresponding rare earth (M= Nd3+, Sm3 +, Dy3 +, etc.) methoxy ethoxide was then added to the previous solution in order to obtain a Zr5Si molar ratio of 159 and M5Si molar ratios of 159; 0.159; 0.0159.The sols were deposited onto glass sheets and allowed to gel and dry at room temperature. Such transparent coatings were deposited several times till a thin layer of about 50–100 mm was obtained.25 FTIR, 1H MAS NMR and 29Si CP MAS NMR experiments have shown that these hybrids can be described as nanocom- Fig. 4 Schematic representation of the environment experienced by posites built from siloxane polymers crosslinked by metal oxo the SO dye within the matrix DH70/TH30. species. The metal oxo domains are made of mixed zirconium–rare earth oxo species. tion). These hybrids can be described as strongly interpen- Absorption–emission properties of these Nd3+ doped hybrid etrated nanocomposites.The environment experienced by the coatings. The room temperature absorption spectrum of Nd3+ SO dye within the DH70/TH30 matrix is shown in Fig. 4. doped hybrid coatings presented in ref. 25 consisted of some The SO or SP DH70/TH30 doped matrices exhibit normal broad transitions. The 4I9/2A4F5/2,2H9/2 transition around photochromism. All the samples are colorless before 800 nm, particularly important for diode pumping systems, irradiation.For the two photochromic dyes, the thermal fading presents a FWHMof around 15 nm. The absorption coeYcient can be fitted with excellent agreement with a monoexponential at this wavelength is around 10 cm-1 for a coating containing equation. This may be related to the quasi-liquid mobility approximately 4×1020 Nd3+ ions cm-3.This high absorption observed by NMR for this matrix. coeYcient value indicates that a high Nd3+ concentration The time dependence of the absorption upon repeated could be introduced into the system without aVecting the irradiation with 365 nm light for SO doped DH70/TH30 synthesis and the transparency of the films.coatings is reported in Fig. 5. The photochromic behavior is For Nd3+ ions excited at 805 nm, the emission spectra of reversible, extremely fast (the rate constant is 0.2 s-1) and the 4F3/2A4I11/2 transition at room temperature is presented corresponds to a very high absorption jump (DA=1.2). The in Fig. 6. This emission is broad with a FWHM around 40 nm photochromic kinetics of these SO doped hybrid materials are and extends into the NIR range from approximately 1.045 to to the best of our knowledge much faster than those reported 1.095 mm.A broad emission has also been observed for the for SO in any other matrix (sol–gel matrices, organic polymers, 4F3/2A4I9/2 transition.25 Broad emissions for Nd3+ are charac- alcohols, etc.).11,14,15,17,23 teristic of wide sites distribution around the neodymium ions The very high reactivity of the DH/TH precursors towards in this hybrid materials.Such inhomogeneous broadening is hydrolysis–condensation reactions (SiMOH groups are immealso observed in glassy hosts.27,28 Experimentally, one can diatly consumed) and the strong hydrophobicity of the notice that the Nd3+ fluorescence intensity is weaker than resulting matrix are both responsible for the direct and very those reported in glassy or crystalline matrices where the fast photochromic behavior observed.quantum eYciency is relatively high.29 2-B Luminescent properties of rare-earth doped hybrids Fluorescence kinetics of Nd3+ in hybrid siloxane based coat- Neodymium doped sol–gel matrices are suitable luminescent ings.Fig. 7 presents the variation of the fluorescence intensity materials. However room temperature sol–gel derived matrices decay profiles as a function of the Nd3+ content. Lifetimes usually contain a large amount of hydroxy groups which are decrease from 160 ms for 0.4×1020 Nd3+ ions cm-3 to responsible for the quenching of the Nd3+ emission. Therefore approximately 200 ns for 4×1021 Nd3+ ions cm-3.the design of sol–gel matrices inside which the hydroxy content It is quite unusual to observe radiative emission of Nd3+ can be minimized24,25 and/or inside which the rare earth is ions in matrices prepared at room temperature by sol–gel protected via complexation26 or encapsulation is desirable. processing. Usually the numerous hydroxy groups present in Recent work has demonstrated fluorescence emission for the classical xerogels obtained at room temperature prevent any Nd3+, Sm3 +, Dy3 +, Er3 + and Tm3+ ions doped in hybrid Nd3+ radiative emission.5d,5e,32 A thermal treatment at high siloxane–oxide matrices.25 temperatures is necessary to allow the Nd3+ emission to be These hybrids were synthesized by the following procedure.observed.In all our samples fluorescence is detected, however Diethoxymethylsilane [DEMS=SiH(OEt)2(CH3)], absolute ethanol and water in a 15151 molar ratio were mixed for a Fig. 6 RT Nd3+ fluorescence spectra.30 Fig. 5 Photochromic response of the SO doped DH70/TH30 matrix.18 J. Mater. Chem., 1999, 9, 35–44 37case, phonons can interact with the atomic transitions of Nd3+ and this coupling, analogous to weak electric dipole coupling, is enhanced when hydroxy groups are close to the neodymium ions.These interactions may explain the low Nd3+ quantum eYciency values and the weak emission observed for high Nd3+ concentration. All these results need of course to be improved, however they open many possibilities in the field of room temperature processed luminescent films.Energy transfer between an organic dye and Nd3+ as inorganic chromophore. Codoped rhodamine 6G–Nd3+ hybrid samples shows that the rhodamine emission spectra exhibit some dips at wavelengths corresponding to the Nd3+ absorption bands.31 This feature indicates that radiative energy transfer occurs between the organic dye and the Nd3+ ions.33–35 A photon emitted by the rhodamine molecule can be trapped by the rare-earth ions in the hybrid siloxane network. Furthermore, when excited by an argon laser at 488 nm, a wavelength where only the organic dye molecules absorb, the codoped (R6G–Nd3+) hybrid coating exhibits a Nd3+ emission around 1.06 mm.No emission was detected around 1.06 mm under the same experimental conditions without the Fig. 7 Variation of the Nd3+ fluorescence decay31 (Nd3+ content in presence of the organic dye inside the hybrid coating. This atom cm-3, (a) 0.4×10,20 (b) 4×1020 (c) 4×1021. behavior reveals that energy transfer mechanisms may be favorable for Nd3+ emission. the Nd3+ fluorescence decay profiles lead to lifetime values New Eu2+ doped hybrid organic–inorganic nanocomposites lower than those usually recorded for Nd3+ in high temperasynthesized at room temperature.The Eu2+ ion is particularly ture processed glassy or crystalline matrices. These hybrid unique because its broad luminescence band 4f65d1A4f7 is materials can be described as a siloxane polymeric network strongly host dependent with emission wavelengths extending made from (CH3SiO3/2) units crosslinked by mixed nanoaggrefrom the UV to the red range of the electromagnetic spec- gates made from zirconium oxo and neodymium oxo species.trum.36 Therefore, the luminescent properties of Eu2+-doped Depending on the Nd3+ concentration, two diVerent optical solids have been intensively studied during the past three features are observed in these hybrid coatings. Both are well decades. These studies have led to the use of these compounds correlated with the structure.(1) For low Nd3+ concentration, as phosphors, notably blue-emitting Eu2+5BaMgAl10O17 in the initial non-exponential part of the decay profiles vary lamp and plasma display panels and UV-emitting linearly on a root-mean-square time scale showing the strong Eu2+5SrB4O7 for medical applications and skin tanning.cross-relaxation phenomena. Even if a few nOMH vibrations Crystalline or glassy Eu2+ doped materials are usually pro- are observed in the infrared spectrum, these hydroxy groups cessed at relatively high temperatures.36–39 Moreover, the are not located close to the Nd3+ ions and at low concentration synthesis and the stabilization of europium in the divalent the main non-radiative de-excitation mechanisms are the state under mild synthetic conditions is not an easy task.For Nd3+–Nd3+ interactions. The fact that these nanocomposites the first time, we reported recently the room temperature are made with segregated metal oxo species (the dispersion is synthesis of new Eu2+ doped hybrid materials together with not statistical all over the sample) leads to Nd3+–Nd3+ their absorption and emission properties.40 These hybrids are interactions even for low rare earth concentrations.The slope obtained though the hydrolysis and condensation of diethoxy- measured at long times give an indication of the Nd3+ methylsilane (MDES), methyltriethoxysilane (TREOS) and fluorescence quantum yield in this hybrid coating. The ratio g zirconium tetrapropoxide precursors in the presence of of the experimental lifetime at low concentration over the europium trichloride.calculated radiative lifetime (g=0.35) indicates that approxi- Dehydrocondensation of organic hydrosilanes with silanols mately 35% of the relaxation occurs radiatively at low Nd3+ is one of the common methods for the synthesis of the siloxane concentration. This is a high value for a room temperature linkage.41 This reaction, which occurs with the evolution of sol–gel processed material and this is in agreement with the hydrogen gas, has been described as follows:41 fact that, at low concentration, Nd3+ ions form clusters at relatively long distances from the remaining hydroxy groups.OSiMH+HOMSiOCA catalystOSiMOMSiO+H2 (2) When the Nd3+ concentration increases, interactions increase as the Nd3+–Nd3+ distance becomes shorter in the nanoaggregates.Simultaneously when the neodymium concen- In this sense, alkoxide precursors containing SiMH groups show the possibility of using the SiMH groups as an in situ tration increases (the zirconium concentration remaining constant) the probability of finding neodymium ions at the surface reducing agent which allows the formation of metal/silica nanocomposites.42 of the metal oxo species increases, rendering the rare earth prone to interactions with some remaining hydroxy groups In our study,40 the in situ formation of hydrogen provided by the cleavage of the SiMH bonds was used to generate, located close to the surface.As a consequence, short lifetimes are observed.For a concentration of 4×1021 Nd3+ ions cm-3 during the first step of hydrolysis and condensation reactions, europium in the divalent state. the quantum yield is less than 0.01%. This behavior characterizes non-radiative de-excitation processes due to strong A typical absorption spectra (Fig. 8) of these is constituted by a broad absorption band in the UV range (200–400 nm) energy transfers between Nd3+ but also non-radiative deexcitation occurring via the filling of the 4F3/2–4I15/2 energy attributed to the 4f75d0A4f65d1 (Eu2+) transition.The emission spectra (Fig. 9) of the corresponding hybrids gap by nOMH vibrations. Usually, according to the energy gap law, multiphonon non-radiative contributions do not exceed recorded under excitation at 355 nm show a broad emission corresponding to the interconfigurational 4f65d1A4f75d0 50% taking into account the energy diVerence around 5500 cm-1 between the 4F3/2 and 4I15/2 levels.In the present transition centered at 430 nm (ca. 23250 cm-1) and a intracon- 38 J. Mater. Chem., 1999, 9, 35–44are only achieved in a non-centrosymmetric environment, we first demonstrated that orientation of organic chromophores can be performed in hybrid sol–gel matrices43–46 by using electrical field induced second harmonic (EFISH) or corona electrical field poling techniques.Organic molecules such as N-(3-triethoxysilylpropyl )-2,4-dinitrophenylamine (TSDP) were chemically bonded to the oxide backbone of gels. The chemical bonding of the dye to the sol–gel matrix allowed dye concentration to be increased without any crystallization eVects.43–46 A first generation of sol–gel matrices was synthesized by copolymerization of silicon alkoxysilanes [TSDP and SiHCH3(OEt)2] and zirconium propoxide precursors.45 The sols were deposited as transparent coatings and exhibited after corona poling an SHG response of 1.6 pm V-1.45 Even if in this first generation of sol–gel matrices relaxation of the organic chromophores occurred over several hours, these results suggested the feasibility of poling techniques into hybrid inorganic sol–gel matrices more ionic than classical polymers.Fig. 8 Absorption spectra of europium doped hybrid xerogels. Consequently a range of opportunities for the synthesis of optical sol–gel devices with eYcient second harmonic properties was discovered. Since then, there has been increasing interest in second order NLO materials synthesized via sol–gel chemistry.47–60 The optimization of the second order NLO response of hybrid sol–gel matrices with grafted chromophores is currently under investigation by several research groups.Several strategies are used to improve the NLO response of the hybrid coatings.43–60 (i) The intrinsic NLO response of the dye can be increase by using chromophores such as N-(4- nitrophenyl )-L-prolinol (NPP) or disperse red one (DR1) derivatives which exhibit higher non linearities than nitroaniline ones.(ii) The chromophore relaxation can be controlled by increasing the matrix rigidity. This point is without doubt the most important in order to be able to make eYcient NLO devices. The modification of the binary composition (siloxane –crosslinker), the nature of theM(OR)4 crosslinking alkoxide [SiR¾x(OR)4-x–M(OR)4: R¾=any NLO chromophore; M=Zr, Si, Ti, etc.), and the processing of these hybrid Fig. 9 Emission spectrum of europium doped hybrid xerogel (lexc= materials in the presence of polymers with well known mechan- 355 nm).ical properties such as methyl methacrylates or polyimides, are the most commonly used strategies to minimize dye figurational 4f–4f Eu3+ emission in the longer wavelengths. relaxation. The strategies we have used to improve the NLO Several bands are obtained corresponding to the 5D0A7F0,1,2,3 response of hybrid materials will be llustrated in the two transitions.A Stokes’ shift value of the Eu2+ luminescence following sections. around 9000 cm-1 is obtained, this shift between the absorption and emission energies of Eu2+ located in an oxygen ligand TSPD/TMOS based hybrids with NLO properties.58,59 The field has been assigned to a combination of crystal field and second generation of hybrids investigated were made via nephelauxetic eVects.36 These hybrid structures contain oxygen hydrolysis and co-condensation of tetramethoxysilane atoms in higher coordination number environments (highly (TMOS) and N-(3-triethoxysilylpropyl )-2,4-dinitrophenylam- coordinated by metal atoms m3-O–Zr/Eu or m4-O–Zr/Eu) ine (TSDP) [Fig. 10(a)] precursors (T and Q are common which produce Eu2+ emission at longer wavelengths. Moreover notations referring to the oxo trifunctional R¾MSiO3 and distortion of the oxygen polyhedra from ideal coordination tetrafunctional SiO4 central units, respectively). FTIR, 17O geometry results in a large Stokes’ shift.First measurements and 29Si NMR experiments indicated the existence of linear indicate a Eu2+5Eu3+ concentration ratio of about 551. The and cyclic siloxane (T–T)a oligomers and silica (Q–Q)b units high Eu2+ content is probably related to the more eYcient reductive medium provided by the initial mixture of the europium trichloride with the MDES and TREOS precursors.Moreover, the intensity of the Eu2+ luminescence did not change when the xerogels were stored in air for several months, showing that Eu2+ ions are eYciently trapped inside the hybrid matrix. 2-C Quadratic NLO properties of siloxane based hybrids Most of the sol–gel optics research devoted to non-linear optic (NLO) materials was initially related to third order processes which are compatible with the isotropy of amorphous sol–gel matrices.Organic molecules inside amorphous sol–gel matrices are in general randomly oriented thus ruling out the emission O2N NO2 NH Si(OEt)3 TSDP O2N N N N CH2CH2OCONH-(CH2)3Si(OEt)3 CH2CH2OCONH-(CH2)3Si(OEt)3 CH3 (a) (b) Fig. 10 Graftable NLO dyes: (a)=TSDP (b)=ICTES-Red17. of second harmonic generation. As second order non-linearities J. Mater. Chem., 1999, 9, 35–44 39linked through stable T–O–Q bridges formed in the early also strongly depend on their thermal history. Chemical crosslinking is not complete at gelation or even after RT air stages of the process.Films of thickness 1–5 mm were easily obtained through spin-coating. In such systems gelation prob- drying as shown by 29Si NMR experiments.59 Upon ageing and curing the chemical reactions continue towards com- ably occurs through the crosslinking of siloxane polymers with Q silica based species. The degrees of condensation of T and pletion.Consequently, given suYcient time and temperature to allow mobility of the species a network forming system Q units measured in the solid state by 29Si MAS NMR spectroscopy are much higher in xerogels than in sols and this continues to crosslink long after gelation. The increase of the density of crosslinks modifies the thermomechanical properties diVerence demonstrates that a large number of condensation and crosslinking reactions still occur upon solvent removal.of the hybrid as illustrated by the changes observed in Tg upon thermal curing.57 The mobility of the NLO chromophores, as observed by high-resolution solid-state 13C NMR spectroscopy, is also Another processing parameter which has great importance is the electrical field used to poled the NLO chromophores.correlated with the glass-transition phenomenon of the matrix observed by DSC.57 This glass transition phenomenon corre- Accelerated field induced curing must occur in these hybrid TSDP/TMOS materials. The high electrical field provided sponds to the glass transition of the polysiloxane network. Tg, the glass-transition temperature, increases with the TMOS during poling must favor crosslinking and interpenetration of both polymeric T and Q networks.This was described by content, while the apparent variation of heat capacity corresponding to Tg decreases. These results, as well as the analysis Haruvy and co-workers,48b who have shown that greatly accelerated curing occurs under ambient conditions on thin of the polarization transfer in MAS/CP/DD 29Si NMR experiments, are consistent with the relatively high degree of inter- films processed from siloxane resins prepared by the sol–gel process when they are exposed to an intense corona-discharge penetration of T and Q units.Therefore, these hybrid TSDP/TMOS coatings can be described as nanocomposites field. The corona cured sol–gel films exhibited a more compact matrix as manifested by the lower mobility of the embedded made of silica rich domains and siloxane rich domains.Many Q and T species are mutually sequestered at the nanometer chromophores and a more hydrophilic surface than thermally cured ones. They suggest that the field induced removal of scale. Their microstructure is schematically pictured in Fig. 11. The white parts correspond to the silica-rich phase inside condensate small molecules and solvents allows better completion of the reactions and more eYcient crosslinking.48 which some T units (black dots) are sequestered.The black spheres correspond to the polysiloxane rich phase which Compared to the tremendous amount of work and time devoted to polymeric NLO materials, NLO materials made participates in the glass-transition phenomenon and contains Q units (white dots).The sizes of the polysiloxane and silica by sol–gel are still in their infancy. For these system, depending on chemical composition, the SHG values range between 2.5 domains depend not only on the chemical composition but also on the drying procedure and consequently on the solvent and 10 pm V-1.47,57 Moreover, the sol–gel materials described in this work have Tg values in the range of 30–70 °C, well and sample thickness.The TSDP/TMOS ratio, proton concentration, hydrolysis below the state of the art obtained with pure organic polymeric materials based on polyimides61 which are highly non linear ratio, sequence of mixing the reagents and ageing time of the sol are the chemical parameters that should directly influence and stable for hundred of hours at temperatures higher than 100 °C.However, the excellent knowledge of such systems a and b values characterizing the lengh of the constituent linear and cyclic siloxane (T–T)a oligomers and silica (Q–Q)b allowed us within a short period of time to design a third generation of hybrids with a highly improved NLO response. units respectively.However, it has been demonstrated that the mechanical properties of hybrid siloxane–oxide materials, and thus the ICTES-Red17/TMOS based hybrids with NLO properties. The third generation of hybrid organic–inorganic nanocom- relaxation behavior of chromophores grafted in these matrices, posites was designed on the basis of the following specifications: the NLO dye must have a high NLO response, it must be anchored by more than one trifunctional link and silica was kept as the crosslinking agent because coatings of better optical quality were usually obtained with binary silica–siloxane materials.56 In order to be able to perform double grafting of an NLO chromophore, the Red 17 [4-(amino-N,N-diethanol )-2-methyl-4¾-nitroazobenzene] with a very eYcient quadratic hyperpolarizability [b(0) (Red 17)=55×10-30 esu] was functionalized with two alkoxysilyl groups by a coupling reaction between the dye and 3-isocyanatopropyltriethoxysilane (ICTES).56,58 The resulting alkoxysilyl functionalized NLO precursor, ICTES-Red 17 [Fig. 10(b)] was hydrolyzed and co-condensed with TMOS in order to obtain the hybrid siloxane–silica nanocomposite.From the resulting sols, coatings with a thickness of a few mm can be deposited. The resulting hybrid materials do not exhibit Tg according to DSC results. Non-resonant second-order non-linearities as high as 150–200 pm V-1,58,62 measured on these hybrid systems, with significant long-term stability (10% of signal lost after 20 days) have been reported.58 The thermal stability at 80 °C has been shown to be excellent, making the ICTES-Red 17/TMOS systems competitive candidates for non-linear optics systems.Chemical characterization (FTIR, 29Si MAS NMR, UV–VIS spectroscopy) and thermal assisted in situ poling studies performed on these coatings revealed the importance of the processing and history of these systems. Three param- Fig. 11 Schematic representation of TSDP-TMOS based hybrids.52 eters are of paramount importance.(i) Aging of the solution 40 J. Mater. Chem., 1999, 9, 35–44has been shown to greatly influence the amplitude of the final Moreover, these clusters exhibit a high versality for the design of hybrids (Fig. 13). {(RSn)12O14(OH)6}2+2X- can non-linear signal. This results from improvement of crosslinking eYciency and from modifications of the distribution be assembled through organic networks by using the covalent interface provided by the SnKC bond or by using the ionic between cyclic and linear siloxane species.(ii) Thermal precuring of the samples at 150 °C was found to markedly improve interface associated with the charge compensating anions X- or even by using both interfaces.In the first case the organic the non-linear response as well as its stability. (iii) Optical poling recently tested in sol–gel derived matrices can also be moiety carried through the SnKCsp3 links should be polymerizable (R=butenyl, propylmethacrylate, propylcrotonate, 4- used to improve the chromophore anisotropy63 These very reproducible results58,62 are very promising, styrylbutyl, etc.).In the second case charge compensating organic dianions must be able to bridge the clusters. This can however as far as NLO devices are concerned they must be completed by measurements of electro-optical eYciency, be performed by using dicarboxylates,65 or a,v-telechelic macromonomers terminated by carboxylic or sulfonic groups.67 waveguiding properties and the evaluation of the optical losses.As an example, the coupling of these clusters by carboxymethyl terminated PEG macromonomers67 is schematically shown 3 Tin oxo species based hybrid materials in Fig. 14. Another strategy could be to use polymerizable anions Tin is a very interesting element because its characteristics make it intermediate between silicon and the transition metals.(methacrylate, 2-acrylamido-2-methylpropane-1-sulfonate, etc.) as monomers for organic polymerizations reactions.66,68a Like the latter, tin exhibits several coordination numbers (generally from 4 to 6) and coordination expansion makes By a simple acid-base reaction, the oxo-hydroxo butyltin macrocation, {(BuSn)12O14(OH)6}2+, was functionalized hydrolysis–condensation reactions of tin alkoxides fast.But, as for silicon, the Sn–Csp3 bond is stable, especially towards with 2-acrylamido-2-methylpropane-1-sulfonate, aVording nanobuilding blocks with two highly polymerizable groups.68a nucleophilic agents such as water. This last characteristic allows one to chemically link organic moieties to the tin oxo For the first time,68a the direct polymerization of such functionalized oxo-hydroxo butyltin nanoclusters has been polymers/oligomers but it also reduces the inorganic functionality of tin and therefore favors the formation of oxo successfully performed, yielding hybrid materials in which the nanosized inorganic component is perfectly defined.Two types clusters. These oxo clusters can be used as nanobuilding blocks in the design of new hybrid materials.64–72 of organic components are found in such materials.The butyl groups covalently bound onto tin atoms, and, more import- The nanobuilding block [(RSn)12(m3-O)14(m-OH)6]2+, the structure of which is shown in Fig. 12 can be obtained through antly, poly(2-acrylamido-2-methylpropane-1-sulfonate) chains which interact through electrostatic interactions with the several chemical pathways: hydrolysis of RSn(OPri)3 or RSnCl3 or by refluxing in toluene butyltin hydroxide oxide oxo-hydroxo butyltin macrocations and aVord the crosslinking. [BuSnO(OH)] in the presence of sulfonic acids (R¾SO3H)64–71 and more recently Jousseamme et al.opened a new route to Such an approach to the construction of tin-based hybrid materials from bifunctional nanobuilding blocks was pre- this cluster through hydrolysis of functionalized trialkynylorganotin precursors.72 viously attempted with pure {(BuSn)12O14(OH)6}- {O2CC(CH3)NCH2}2 but failed as its homopolymerization This compound is made of a tin oxo-hydroxo cluster with a equal numbers of six- and five-coordinate tin atoms.This appeared impossible.66 Addition of a co-monomer [CH3O2CC(CH3)NCH2] allowed the polymerization, but cage-like cluster is surrounded by twelve organic chains (butyl, butenyl, etc.) which prevent further condensation.Depending recent results have indicated that little crosslinking was achieved, the methacrylate charge compensating anions acting on the synthesis conditions the 2+ positive charge can be compensated by a large variety of anions (OH+,Cl-, sulfon- mainly as termination agents.66b These diYculties may be related to the fairly large molecular weight of the precursor ates, carboxylates, etc.).The position of the charge compensating anions in the structure indicates that the 2+ charge is (ca. 2600 g mol-1), but also to the shortness of the methacrylate functional anions which induce high steric hindrance.equally located at both cage poles, where six-coordinate tin atoms form hydroxylated [RSn(OH)]3O trimers. The second reason seems to prevail, as the use of AAMPS, where the polymerizable acrylamido group is more distant This cluster is confirmed both in solution by 119Sn NMR [it is characterized by two chemical shifts located at about -280 from the anionic anchoring head, allows the formation of a hybrid polymer by simple homopolymerization.More work is and -450 ppm (R=butyl or butenyl )] and a set of scalar tin–tin coupling satellites and in the solid state through 119MAS currently under way to obtain some information on the average length of the poly(2-acrylamido-2-methylpropane-1-sulfonate) NMR spectroscopy.69,71 Thus, this cluster can be followed easily throughout the polymerization or crosslinking reactions chains, which are probably short owing to strong steric hindrance. needed to tranform it into a hybrid material.As a consequence {(RSn)12O14(OH)6}2+ is a good nanobuilding block for the synthesis of well defined tin-oxo based hybrid materials that can be used as models. 4 Molecular design of transition metal alkoxides for the synthesis of hybrid organic–inorganic copolymers The chemical tailoring performed with systems containing a SiMC bond cannot be directly extended to pure transition metals because the more ionic MMC bond is broken down upon hydrolysis.Organic modification can however be performed by means of strong complexing ligands. The best are b-diketones and allied derivatives, polyhydroxylated ligands such as polyols, and a- or b-hydroxyacids.These ligands (HL) react readily with transition metal alkoxides M(OR)4 (M= Ce, Ti, Zr, etc.) to yield new precursors M(OR)3-x(L)x.73,74 Upon hydrolysing these new precursors, most of the alkoxy groups are quickly removed while all strong complexing ligands Fig. 12 Molecular structure of [(RSn)12(m3-O)14(m-OH)6 ]2+. cannot be completely removed.Complexing ligands appear to J. Mater. Chem., 1999, 9, 35–44 41Fig. 13 Schematic representation of the strategies that can be used from a hybrid [(RSn)12(m3-O)14(m-OH)6 ]2+2X- nanobuilding block. by partial hydrolysis of the alkoxy groups and radical polymerization of the allyl functions. However polymerization of allyl functions is slow and the degree of polymerization remains low.More reactive methacrylic acid can also be used as a polymerizable chelating ligand. The sol–gel synthesis of zirconium oxide based monoliths synthesized by UV copolymerization of zirconium oxide sols and organic monomers was recently reported.8 However, as carboxylic functions are weak ligands, they are largely removed upon hydrolysis2,75 and thus a large number of the chemical bonds between organic and inorganic networks is lost in the sol state.Therefore a new approach was chosen with diVerent ligands, such as acetoacetoxyethylmethacrylate (AAEM) and methacrylamidosalicylic acid (MASA), which contain both a strong chelating part and a highly reactive methacrylate group.6 Zirconium–oxo-PAAEM copolymers were synthesized from zirconium propoxide modified at the molecular level with AAEM.6 These hybrid organic–inorganic copolymers are made of zirconium oxo-polymers and polymethacrylate chains. The zirconium oxo species, in which zirconium is coordinated to seven oxygens, are chemically bonded to methacrylate chains Fig. 14 Schematical representation of the tin oxo-hydroxo clusters hybrids obtained through crosslinking of [(RSn)12O14(OH)6 ]2+ with through the b-diketo complexing function.The complexation a,v-PEG carboxylates.67 ratio (AAEM/Zr) is the key parameter which controls the structure and the texture of these hybrid materials (schematic structure Fig. 15). Careful adjustment of this parameter leads be quite stable towards hydrolysis because of chelate and steric to the tailoring of the ratio between organic and inorganic hindrance eVects.Thus, they allow organic groups to be components and also to zirconium oxo species with more or anchored to transition metal oxo-polymeric species and allow less open structures. The inorganic/organic ratio increases the synthesis of new hybrid organic–inorganic materials. when the complexation ratio decreases.For a high com- Organically modified TiO2 gels, which give photochromic plexation ratio (0.75) both networks interpenetrate intimately coatings, were synthesized from an allyl acetylacetone modified Ti(OBun)4 alkoxide.74 Double polymerization was performed at the nanometer scale, while for a low ratio (0.25) the size of 42 J. Mater. Chem., 1999, 9, 35–44References 1 C.J. Brinker and G. W. Scherrer, Sol–Gel Science, The Physics and Chemistry of Sol–Gel Processing, Academic Press, San Diego, CA, 1990. 2 (a) J. Livage, M. Henry and C. Sanchez, Prog. Solid State Chem., 1988, 18, 259; (b) C. Sanchez, F. Ribot and S. DoeuV, Inorganic and Organometallic Polymers with Special Properties, ed. R. M. Laine, NATO ASI Series 206, Kluwer, New York, NY, 1992, p. 267. 3 (a) G. L. Wilkes, B. Orler and H. H. Huang, Polym. Prep., 1985, 26, 300; (b) G-S. Sur and J. E. Mark, Eur. Polym. J., 1985, 21, 1051; (c) H. Schmidt, A. Kaiser, H. Patzelt and H. Sholze, J. Phys., 1982, 43, C9-275. 4 (a) B.M. Novak, Adv. Mater., 1993, 5, 422; (b) Proceeding of the First European Workshop on Hybrid Organic–Inorganic Materials, ed. C. Sanchez and F. Ribot, New J.Chem., 1994, 18; Fig. 15 Schematic structures of class II hybrid materials made from (c) U. Schubert, N. Hu�sing and A. Lorenz, Chem.Mater., 1995, 7, zirconium n-propoxide complexed by acetoacetoxyethylmethacrylate 2010; (d) D. A. Loy and K. J. Shea, Chem. Rev., 1995, 95, 1431; (AAEM).6 (e) J. Sol–Gel Sci. Technol., 1995, 5; ( f ) P. Judenstein and C. Sanchez, J. Mater. Chem., 1996, 6, 511; (g) Mate�riaux Hybrides, Se� rie Arago 17, Masson, Paris, 1996; (h) C.Sanchez and F. Ribot, New J. Chem., 1994, 18, 1007; (i) U. Schubert, J. Chem. the inorganic domains increases to the sub-micrometer range. Soc., Dalton Trans., 1996, 3343; ( j) A. Morikawa, Y. Iyoku, M. Kakimoto and Y. Imai, J. Mater. Chem., 1992, 2, 679; Organic and inorganic growths are not independent and such (k) Y.Chujo and T.Saegusa, Ad. Polym. Sci., 1992, 100, 11; systems probably exhibit similar behavior to the so-called (l ) Hybrid Organic–Inorganic Composites, ed. J. E. Mark, interpenetrating polymer networks. C. Y. C. Lee and P. A. Bianconi, American Chemical Society, Washington, DC, 1995; (m) F. Ribot and C. Sanchez, Comments Inorg. Chem., 1998, in press. 5 (a) Better Ceramics Through Chemistry VI, ed.A. Cheetham, 5 Conclusions C. J Brinker, M. MacCartney, C. Sanchez, D. W. Schaefer and G. L. Wilkes, Mater. Res. Soc. Symp. Proc., Materials Research The combination at the nanosize level of inorganic and organic Society, Pittsburgh, PA, 1994, vol. 435; (b) Better Ceramics or even bio-active components in a single material makes Through Chemistry VII: Organic/Inorganic Hybrid Materials, ed.accessible an immense new area of materials science that has B. K. Coltrain, C. Sanchez, D. W. Schaefer and G. L. Wilkes, extraordinary implications for developing novel multi-func- Mater. Res. Soc. Symp. Proc., Materials Research Society, tional materials exhibiting a wide range of properties. This Pittsburgh, PA, 1996, vol. 435; (c) Sol–Gel Optics I, ed. J. D. Mackenzie and D. R. Ulrich, Proc. SPIE, Washington, 1990, fascinating new field of research brings together scientists vol. 1328; (d) Sol–Gel Optics II, ed. J. D. Mackenzie, Proc. SPIE, working in many diVerent domains. Among soft chemistry Washington, 1992, vol. 1758; (e) Sol–Gel Optics III, ed. processes, sol–gel chemistry oVers versatile access to the chemi- J.D. Mackenzie, Proc. SPIE, Washington, 1994, vol. 2288. cal design of new hybrid organic–inorganic materials. Many 6 (a) C. Sanchez and M. In, J. Non-Cryst. Solids 1992, 147–148, 1; new combinations between inorganic and organic or biological (b) M. In, C. Ge�rardin, J. Lambart and C. Sanchez, J. Sol–Gel Sci. Technol., 1995, 5, 101. components will probably appear in the future. Yet, better 7 (a) U.Schubert, E. Arpac, W. Glaubitt, A. Helmerich and understanding and co the local and semi-local structure C. Chau, Chem. Mater., 1992, 4, 291; (b) C. Barglik-Chory and of these materials is an important issue, especially if tailored U. Schubert, J. Sol–Gel Sci. Technol., 1995, 5, 135. properties are sought. 8 R. Naß and H. Schmidt, in ‘Sol–Gel Optics I’, ed. J.D. Mackenzie To achieve such a control of the material structure, the and D. R. Ulrich, Proc. SPIE,Washington, 1990, vol. 1328, p. 258. 9 H. Schmidt and B. Seiferling, Mater. Res. Soc. Symp. Proc., 1986, assembly of well defined nanobuilding blocks is an interesting 73, 739. approach. The inorganic components are nanometric and 10 (a) D. Levy and D. Avnir, J.Phys. Chem., 1998, 92, 734; monodispersed. Their structures are perfectly defined, which (b) D. Levy, S. Einhorn and D. Avni, J. Non-Cryst. Solids, 1989, probably facilitates the characterization of the final materials. 113, 137. The variety found in the nanobuilding blocks (nature, struc- 11 Proceedings of the Second International Symposium on ture, functionality) and links, associated with diVerent assemb- Photochromism, ed.R. C. Bertelson, Chroma Chemicals Inc, Dayton, Ohio, USA, 1997. ling strategies, allow very diVerent types of architectures and 12 D. Preston, J. C. Pouxviel, T. Novinson, W. C. Kaska, B. Dunn organic–inorganic interfaces to be formed.4m Moreover, the and J. I. Zink, J. Phys. Chem., 1990, 94, 4167. step-by-step preparations of the materials usually allows some 13 H.Nakazumi, R. Nagashiro, S. Matsumoto and K. Isagawa, control over their semi-local structure. One can also combine Sol–Gel Optics III, ed. J. D. MacKenzie, Proc. SPIE,Washington, the nanobuilding blocks approach with the use of organic 1994, vol. 2288. 14 L. Hou, B. HoVmann, M. Menning and H. Schmidt, J. Sol–Gel templates that self-assemble and allow, through weak forces, Sci.Technol., 1994, 2, 635. control of the assembly step. With the aim of organizing/ 15 L. Hou, B. HoVmann, H. Schmidt and M. Menning, J. Sol–Gel structuring the nanobuilding blocks prior to their assembly, Sci. Technol., 1997, 8, 923, 927. one can also functionalize them with mesogenic groups and 16 L. Hou, M. Menning and H. Schmidt, Proc Eurogel’92, 1992, give them self-organizing properties.p. 173. 17 J. Biteau, F. Chaput and J. P. Boilot, J. Phys. Chem., 1996, 100, As described by Mann et al,76 the traditional view of 9024. inorganic solids as condensed matter is currently being 18 D. Levy, Chem. Mater., 1997, 9, 2666. reshaped by new insight in materials synthesis.77–79 A new 19 B. Schaudel, C. Guermeur, C. Sanchez, K. Keitaro and J.Delaire, paradigm, organized matter chemistry, is being formulated. In J. Mater. Chem., 1997, 7, 61. this field the sol–gel chemistry of organized matter is a very 20 S. Dire�, F. Babonneau, C. Sanchez and J. Livage, J. Mater. Chem., 1992, 2, 239. promising field of research which is just beginning to be 21 F. Babonneau, Polyhedron, 1994, 13, 1123. explored. Hybrids of both Class I and II4h and more particu- 22 C.Guermeur and C. Sanchez, forthcoming paper. larly those obtained through the ‘nanobuilding block 23 Applied photochromic polymers systems, ed. C. B. McArdle, approach’ will have paramount importance in the exploration Gordon and Breach Science Publishers, New York, 1991. the theme of ‘synthesis with construction’ of hierarchically 24 S.K. Yuh, E. Bescher, F. Babonneau and J. D. Mackenzie, Mater. Res. Soc. Symp. Proc., 1994, 346, 803. organized (structure and function) materials. J. Mater. Chem., 1999, 9, 35–44 4325 N. Koslova, B. Viana and C. Sanchez, J. Mater. Chem., 1993, M. Dumont, New J. Chem., 1996, 20, 13; (b) B. Lebeau, C. Sanchez, S. Brasselet and J. Zyss, Mater. Res. Soc. Symp. Proc., 3, 111. 1996, 435, 395; (c) B.Lebeau, C. Guermeur and C. Sanchez, 26 L. R. Mathews and E. T. Knobbe, Mater. Res. Soc. Symp. Proc., Mater. Res. Soc. Symp. Proc., 1994, 346, 315. 1993, 286, 259. 57 B. Lebeau, J. Maquet, C. Sanchez, E. Toussaere, R. Hierle and 27 I. M. Thomas, S. A. Payne and G. D. Wilke, J. Non-Cryst. Solids, J. Zyss, J. Mater. Chem., 1994, 4, 1855. 1992, 151, 183. 58 B. Lebeau, C.Sanchez, S. Brasselet and J. Zyss, Chem. Mater., 28 I. M. Thomas, S. A. Payne and G. D. Wilke, J. Non-Cryst. Solids, 1997, 9, 1012. 1992, 151, 183. 59 B. Lebeau, J. Maquet, C. Sanchez, F. Beaume and F. Laupre�tre, 29 A. A. Kaminskii, in Laser Crystals, Springer Verlag, Berlin, J. Mater. Chem., 1997, 7, 989. Heidelberg, New York, 2nd edn., 1981, pp. 361–433. 60 C. Sanchez and B.Lebeau, Pure Appl.Opt., 1996, 5, 689. 30 M. Lecomte, B. Viana and C. Sanchez, J. Chim. Phys., 1991, 61 R. F. Shi, M. H. Wu, S. Yamada, Y. M. Cai and A. F. Garito, 88, 39. Appl. Phys. Lett., 1993, 63, 1173. 31 B. Viana, N. Koslova, P. Aschehoug and C.Sanchez, J. Mater. 62 D. Blanc, P. Peyrot, C. Sanchez and C. Gonnet, Opt. Eng. Chem., 1995, 5, 719. Integrated Opt., 1998, 37, 1203. 32 C. Guizard, J.C. Achddou, A. Larbot, L. Cot, G. Le Flem, 63 C. Fiorini, F. Charra, J. M. Nunzi, I. D. W. Samuel and J. Zyss, C. Parent and C. L. Lurin, in ref. 5(c), p. 208. Opt. Lett., 1995, 20, 2469. 33 M. Genet, V. Brandel, M. P. LaHalle and E. Simoni, in ref. 5(c), 64 F. Ribot, F. Banse and C. Sanchez, Mater. Res. Soc. Symp. Proc., p. 194. 1994, 346, 121. 34 W. Nie, B. Dunn, C.Sanchez and P. Griesmar, Mater. Res. Soc. 65 F. Ribot, F. Banse, F. Diter and C. Sanchez, New J. Chem., 1995, Symp. Proc., 1992, 271, 639. 19, 1145. 35 M. Canva, P. Georges, G. Lesaux, A. Brun, F. Chaput and 66 (a) F. Ribot, F. Banse, C. Sanchez, M. Lahcini and J. P. Boilot, J. Non-Cryst. Solids, 1992, 147, 627. B. Jousseaumme, J. Sol–Gel Sci. Technol., 1997, 8, 529; 36 G. J.Dirksen and G. Blasse, J. Solid State Chem., 1991, 92, 591. (b) L. Angiolini, D. Caretti, C. Carlini, R. De Vito, F. T. Niesel, 37 A. Diaz and D. A Keszler, Chem. Mater. 1997, 9, 2071. E. Salatelli, F. Ribot and C. Sanchez, J. Inorg. Organomet. Polym., 38 A. Diaz and D. A Keszler, Mater. Res. Bull., 1996, 31, 147. 1998, 7, 151. 39 H. F. Folkerts and G. Blasse, J. Mater. Chem., 1995, 5, 1547. 67 (a) F. Ribot,C. Eychenne-Baron and C. Sanchez, Mater. Res. Soc. 40 E. Cordoncillo, P. Escribano, B. Viana and C. Sanchez, J. Mater. Symp. Proc., 1996, 435, 43; C. Eychenne-Baron, F. Ribot and Chem., 1998, 8, 1507. C. Sanchez, J. Organomet. Chem., 1998, 567, 137. 41 J. Chrusciel and Z. Lasocki, Pol. J. Chem., 1983, 57, 121. 68 F. Ribot, C. Eychenne-Baron and C. Sanchez, Mater. Res. Soc. 42 R. Campostrini and S. Dire�, Eurogel Proceedings, Advanced Symp. Proc., 1998, 519, in press. Materials and Processes by Sol–Gel Techniques, Colmar, 1992, ed. 69 F. Banse, F. Ribot, P. Tole�dano, J. Maquet and C. Sanchez, Inorg. S. Vilminot, Strasbourg, 1995, p. 307. Chem., 1995, 34, 6371. 43 G. Pucetti, I. Ledoux, J. Zyss, P. Griesmar and C. Sanchez, Polym. 70 H. PuV and H. Reuter, J. Organomet. Chem., 1989, 373, 173. Prepr., 1991, 32, 61. 71 D. Dakternieks, H. Zhu, E. R. T. Tiekink and R. J. Colton, 44 J. Zyss, G. Pucetti, I. Ledoux, P. Griesmar, J. Livage and J. Organomet. Chem., 1994, 476, 33. C. Sanchez, Fr. Pat., 1152, March 1991. 72 P. Jaumier, B. Jousseaume, M. Lahcini, F. Ribot and C. Sanchez, 45 (a) E. Toussaere, J. Zyss, P. Griesmar and C. Sanchez, Non Linear Chem. Commun., 1998, 369. Optics, 1991, 1, 349; (b) C. Sanchez, P. Griesmar, E. Toussaere, 73 (a) R.C.Mehrotra, R. Bohra and D. P. Gaur, Metal b-diketonates G. Puccetti, I. Ledoux and J. Zyss, Non Linear Optics, 1992, 4, 245. and Allied Derivatives, Academic Press, London, 1978; 46 P. Griesmar, C. Sanchez, G. Pucetti, I. Ledoux and J. Zyss, Mol. (b) D. C. Bradley, R. C. Mehrotra and D. P. Gaur, Metal Eng., 1991, 1, 205. Alkoxides, Academic Press, London, 1978; (c) L. G. Hubert- 47 (a) J. Kim, J. L. Plawsky, R. LaPeruta and G. M. Korenowski, Pfalzgraf, New J. Chem., 1987, 11, 663; ( f ) A. Leaustic, Chem. Mater., 1992, 4, 249; (b) J. Kim, J. L. Plawsky, E. Van F. Babonneau and J. Livage, Chem. Mater., 1989, 1, 248; Wagenen and G. M. Korenowski, Chem.Mater., 1993, 5, 1118. ( g) F. Ribot, P. Tole�dano and C. Sanchez, Chem. Mater., 1991, 3, 48 (a) Y. Haruvy ander, Mater. Res. Soc. Symp. Proc., 759; (h) C.Sanchez, M. In, P. Toledano and P. Griesmar, Mater. 1992, 271, 297; (b) Q. Hibben, E. Lu, Y. Haruvy and S.E. Weber, Res. Soc. Symp. Proc., 1992, 271, 669. Chem. Mater., 1994, 6, 761. 74 C. Sanchez, J. Livage, M. Henry and F. Babonneau, J. Non-Cryst. 49 F. Chaput, D. Riehl, Y Levy and J. P. Boilot, Chem.Mater., 1993, Solids, 1988, 100, 650. 5, 589. 75 S. DoeuV, M. Henry, C. Sanchez and J. Livage, J. Non-Cryst. 50 Y. Zhang, P. N. Prasad and R. Burzynski, Chem. Mater., 1992, Solids, 1987, 89, 206. 4, 851. 76 S. Mann, S. L. Burkett, S. A. Davis, C. E. Fowler, 51 Y. Nosaka, N. Tohriiwa, T. Kobayashi and N. Fujii, Chem. N. H. Mendelson, S. D. Sims, D.Walsh and N. T. Whilton, Chem. Mater., 1993, 5, 930. Mater., 1997, 9, 2300. 52 S. Marturunkakul, J. I. Chen, R. J. Jeng, S. Sengupta, J. Kumar 77 S. Mann and G. A. Ozin, Nature, 1996, 382, 313. and S. K. Tripathy, Chem. Mater., 1993, 5, 743; K. Izawa, 78 C. G. Go� ltner and M. Antonietti, Adv. Mater., 1997, 9, 431. N. Okamoto and O. Sugihara, Jpn. J. Appl. Phys., 1993, 32, 807. 79 (a) A. Monnier, F. Schuth, Q. Huo, D. Kumar, D. Margolese, 53 L. Kador, R. Fischer, D. Haarer, R. Kaseman, S. Bru� ck, R. S. Maxwell, G.D. Stucky, M. Krishnamurty, P. PetroV, H. Schmidt and H. Du� rr, Adv. Mater., 1993, 5, 270. A. Firouzi, M. Janicke and B. F. Chmelka, Science, 1993, 261, 54 (a) J. R. Caldwell, R. W. Cruse, K. J. Drost, V. P. Rao, A. K.- 1299; (b) A. Firouzi, D. Kumar, L. M. Bull, T. Besier, P. Sieger, Q. Hue, S. A. Walker, J. A. Zasadinski, C. Glinka, J. Nicol, D. Y. Jen, K. Y. Wong, Y. M. Cali, R. M. Mininni, J. Kenney, Margolese, G. D. Stucky and B. F. Chmelka, Science, 1995, 267, E. Binkley, L. R. Dalton, Y. Shi and W. H. Steier, Mater. Res. 1133; (c) D. Y. Zhao, P. D. Yang, Q. S. Huo, B. F. Chmelka and Soc. Symp. Proc., 1994, 328, 531; (b) Z. Yang, C. Xu, B. Wu, L. G. D. Stucky, Curr. Opin. Solid State Mater. Sci., 1998, 3, 111; (d) R. Dalton, S. Kalluri, W. H. Steier, Y. Shi and J. H. Bechtel, C. J. Brinker, Curr. Opin. Colloid Interface Sci., 1998, 3, 166; (e) Chem. Mater., 1994, 6, 1899; (c) H. W. Oviatt, K. J. Shea, S. Y. Lu, R. Ganguli, C. A. Drewien, M. T. Anderson, C. J. Brinker, Kalluri, Y. Shi, W. Steier and L. R. Dalton, Chem. Mater., 1995, W. Gong, Y. Guo, H. Soyez, B. Dunn, M. H. Huang and J. I. 7, 493. Zink, Nature, 1997, 389, 364; ( f ) G. H. Mehl and J. W. Goodby, 55 (a) F. Chaput, D. Reihl, Y. Levy and J.-P. Boilot, Chem. Mater., Chem. Ber., 1996, 129, 521. 1993, 5, 589; (b) D. Reihl, F. Chaput, Y. Levy, J.-P. Boilot, F. Kajzar and P. A. Chollet, Chem. Phys. Lett., 1995, 245, 36. 56 (a) B. Lebeau, C. Sanchez, S. Brasselet, J. Zyss, G. Froc and Paper 8/05538F 44 J. Mater. Chem., 1999, 9, 35&nd

ISSN:0959-9428    

DOI:10.1039/a805538f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Synthesis of sol–gel encapsulated heme proteins with chemical sensing properties† Esther H. Lan,a Bakul C. Dave,b Jon M. Fukuto,a Bruce Dunn,*a JeVrey I. Zinkc and Joan S. Valentinec aUniversity of California, Los Angeles, Department of Materials Science and Engineering, Los Angeles, CA 90095, USA bSouthern Illinois University, Department of Chemistry, Carbondale, IL, USA cUniversity of California, Los Angeles, Department of Chemistry and Biochemistry, Los Angeles, CA 90095, USA Received 18th May 1998, Accepted 16th July 1998 Heme proteins such as cytochrome-c (cyt-c), hemoglobin microorganisms.In general, sol–gel matrices are thermally and (Hb), and myoglobin (Mb) have been successfully encap- chemically stable, can be obtained in a variety of forms such sulated in sol–gel derived silica matrices, retaining their as bulk monoliths or thin films, and can store very high spectroscopic properties and chemical function.The ther- concentrations of proteins due to the porous network. Proteins mal stability of cyt-c was significantly improved by immob- and enzymes encapsulated in sol–gel derived glasses can interilization in a porous silica network.Results from optical act with target molecules with a high degree of specificity, and absorption, resonance Raman, and thermal denaturation using an appropriate sensing scheme, produce a detectable studies suggest that biomolecules such as cyt-c design self- signal. While the relatively large biomolecules are immobilized specific pores in the silica network according to the size within the silica network, the porous network allows small and shape requirements of the biomolecule.Hb and Mb, ions or molecules to be transported into the interior of the immobilized using the sol–gel process, bound ligands simimatrix. There is, however, a limit in the molecular size of the lar to the proteins in aqueous buffer, and silica-encapsuanalyte that can be detected since it must be able to travel lated manganese myoglobin (MnMb) was a viable detector through the pore network.for nitric oxide (NO). At the time of gelation, the matrix may be considered as a two-phase system consisting of a porous solid and a trapped Introduction aqueous phase. This two-phase material is termed an aged gel.As the gel is allowed to dry, liquid is expelled from the pores The encapsulation of proteins in sol–gel derived silica glasses and the gel shrinks as the pores collapse, forming a dried gel, has been widely studied in recent years. Research in this field termed a xerogel. Xerogels are approximately one-eighth of has established that, upon encapsulation, proteins retain their the original volume, 50% porous (by volume), and dimen- spectroscopic properties and enzymes retain their biological sionally invariant.The pores of the gels are negatively charged activity.1–4 Immobilizing proteins by physically trapping the (pI of silica ca. 2) at the pH of the buVers (pH#7) generally molecules in individual pores of a matrix permits the protein used to encapsulate biomolecules in the sol–gel process. Such molecules to be isolated and stabilized.Conventional immobilpores can show aYnity towards polar and charged dopant ization techniques such as covalent attachment results in molecules, and they may also alter peripheral hydrogen- chemical modifications of the protein. In contrast, sol–gel bonding interaction of biomolecules.As the conformations of immobilization is characterized by physical entrapment withthe biomolecules are maintained by a large sum of relatively out chemical modification. One potential advantage is that the weaker non-covalent interactions, these eVects can perturb the silica matrix ‘cages’ the biomolecule, providing a far more structural dynamics of trapped entities. The entrapment of ‘rugged’ environment for the dopant molecule in contrast to biomolecules inside this silicate ‘cage’, therefore, can produce surface attachment schemes.The reaction chemistry of sol–gel slight alterations in conformational structure and possibly immobilized biomolecular systems has been shown to be biological function. analogous to that in aqueous solution except for the observed In this paper, we will review the literature published to date rates of chemical reactions which are generally slower due to on sol–gel encapsulation of heme proteins and summarize the diVusion limitations in a porous silica matrix.The first studies key findings. Proteins containing heme are characterized by on sol–gel immobilization of biomolecules by Avnir and cothe presence of Fe-protoporphyrin IX, which gives the protein workers involved trapping of the enzyme alkaline phosphatase its characteristic spectroscopic properties. In addition to which produced a bioactive powder.5 We developed synthesis reviewing the literature in this area, we include new studies on conditions that produced optically transparent monoliths with the heme proteins cytochrome c (cyt-c), hemoglobin (Hb), trapped proteins, enabling these materials to be used as myoglobin (Mb), and manganese myoglobin (MnMb).For optical sensors.1 cyt-c, we investigated possible modifications to protein struc- Physical entrapment of proteins in a silica matrix without ture as a result of encapsulation, the extent to which the chemical modification preserves protein structure and funcmatrix can stabilize the protein from thermal denaturation, tionality and protects the protein.Isolating protein molecules and the eVect of pH on the protein in silica as compared to in individual pores of a silica matrix permits the molecules to the protein in solution. For Hb, Mb, and MnMb, we evaluated be stabilized and prevents degradation due to proteases or the feasibility of using sol–gel encapsulated heme proteins as sensing elements, specifically, for optical detection of nitric oxide (NO).We and other research groups have investigated †Basis of the presentation given at Materials Chemistry Discussion the sol–gel encapsulation of heme proteins for several reasons: No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. 1) they represent model systems for understanding sol–gel J. Mater. Chem., 1999, 9, 45–53 45encapsulation eVects as their protein structure and func- water which had its temperature controlled by a thermal programmer. The absorbance at 400 nm was monitored at tionality have been well documented in the literature, 2) their characteristic spectroscopic properties make optical detection 5 °C intervals from 25 to 95 °C.EVects of pH were tested by immersing a gel in phosphate buVer adjusted to appropriate possible, 3) they can bind ligands, resulting in changes in optical absorption that are readily observed, and 4) they serve pH values between 7 and 0. The gel was equilibrated with the buVer for about one day. The same sample was used as feasible detectors for the ligands of interest, making device applications possible.throughout the length of the experiment. Review of sol–gel encapsulation of heme proteins Experimental Silica sol was prepared using tetramethylorthosilicate (TMOS) Table 1 summarizes the research to date on sol–gel encapsulation of heme proteins and the key findings. Our earliest work as the precursor. 15.27 g of TMOS, 3.36 g of deionized water, and 0.22 g of 0.04MHCl were mixed and sonicated to produce on sol–gel encapsulation of proteins and enzymes demonstrated that optical spectroscopy of encapsulated cyt-c and the sol.For cyt-c, the sol was added to a Fe(III) cyt-c solution buVered with 0.1 M pH 4.25 acetate to achieve a final protein Mb was essentially identical to that in solution.1 In subsequent work on cyt-c, we found that immobilizing the protein in a concentration of 100 mM in the aged gels.For Hb, Mb, and MnMb, the pH of the sol was raised to pH#5.5 using 1.0 M silica matrix can stabilize the protein from aggregation due to methanol, and denaturation due to methanol was reversible in NH4OH and then added to an aqueous protein solution. The final protein concentrations in aged gels were 5 mM for native the encapsulated protein.8 The improvements in protein stability in the presence of alcohol were attributed to confinement Hb, 20 mM for native Mb, and 3 mM for MnMb.All proteins were encapsulated with the heme in the Fe(III) or Mn(III) and isolation of the biomolecule in pores of the silica network. In work by Shen et al., native cyt-c and zinc cyt-c were state.Gels were of dimensions x 1.8 cm×1.0 cm×0.4 cm. In the CO binding experiments with native Hb or Mb silica encapsulated in sol–gel silica glasses and studied using optical spectroscopy, circular dichroism, and resonance Raman spec- gels, aged gels with the protein in the Fe(III) or ‘met’ state were immersed in buVer (0.1 M pH 7.0 Tris–HCl), reduced troscopy.9 The results collectively suggested that encapsulation in the sol–gel glass only slightly perturbed the polypeptide with dithionite, and then exposed to gaseous CO, all under anaerobic conditions.The gels were gently stirred in CO- backbone and did not detectably perturb the heme group. In zinc cyt-c quenching experiments using [Fe(CN)6]3-, O2 and saturated buVer for 20 min.The saturation concentration of CO in water (1 atm, 20 °C) is 1 mM.6 In the NO binding p-benzoquinone, rate constants were consistently lower in the silica-encapsulated samples than in solution. The slower rates experiments with native Hb or Mb silica gels, aged gels in the Fe(III ) or ‘met’ state were immersed in buVer and exposed to can be attributed to diVusion processes through the porous network.Moreover, changes in ionic strength and pH aVected gaseous NO under anaerobic conditions. NO binding to Hb or Mb did not require reduction using sodium dithionite as kinetics diVerently in the gel as compared to solution, possibly due to the fact that positively charged cyt-c adsorbed to the NO can act as the reducing agent. The gels were gently stirred in the NO-saturated buVer for about 20 min.The saturation negatively-charged silica walls. The encapsulation of cyt-c–cyt-c peroxidase complexes in concentration of NO in water (1 atm, 20 °C) is 2 mM.6 MnMb was prepared by incorporating Mn protoporphyrin sol–gel derived gels allowed, for the first time, detection of this system’s electron paramagnetic resonance (EPR) signals into apomyoglobin using the method of Yonetani and Asakura.7 The experiments carried out using MnMb included at room temperature.10 Previously, quantitative EPR measurements of cyt-c and its complexes were performed at tempera- the study of reactions in solution as well as in a sol–gel matrix.All experiments were conducted in air. In the experiments tures below -100 °C, because EPR absorption of the heme group broadens at higher temperatures.By sol–gel encapsul- involving NO-synthase to enzymatically generate NO, the following constituents were mixed in the stated order: 1) 25 ml ation, however, a suitable microenvironment was obtained that permitted a measurable EPR spectrum at ambient tem- 0.01MpH 7.4 Tris–HCl buVer, 2) 25 ml 0.2 mg ml-1 calmodulin, 3) 25 ml 40 mM CaCl2, 4) 25 ml 2 mM NADPH, 5) 50 ml peratures because the immobilized protein was ‘frozen’ in the gel.The charge transfer activities between the Fe and porphyrin 1 mM tetrahydrobiopterin, 6) 25 ml 5mM L-arginine, 7) 30 ml 60 mM MnMb solution, 8) 20 ml 57.5 mM sodium dithionite, ring in cyt-c–cyt-c peroxidase complex were weak in the silica gels as compared to that in solution.Redox processes of the and 9) 350 ml 341 mg ml-1 NOS enzyme solution (semi-purified from rat brains). The temperature of this solution was main- encapsulated cyt-c–cyt-c peroxidase complex were more active in the aged gel stage than in the xerogel stage. tained at 37 °C during the experiment. In MnMb solution experiments with various nitrogen oxides, MnMb in 0.1 M In one study to evaluate biocatalysis of sol–gel encapsulated heme proteins, the sol–gel encapsulated proteins (cyt-c, Mb, pH 7.0 phosphate buVer was reacted with sodium dithionite and exposed to gaseous NO, gaseous NO2, solid NaNO2, Hb, horseradish peroxidase) had catalytic eYciencies in the oxidation of dibenzothiophene similar to those in solution.11 solid Na2N2O3, or solid NaNO3 of the stated concentrations.For the sol–gel encapsulated MnMb, aged gels were immersed Because of the advantage of easy separation of the encapsulated proteins from the liquid reaction mixture, it was suggested in buVer, reduced using sodium dithionite, and exposed to a 0.8 mM NO solution. In the S-nitroso-N-acetylpenicillamine that immobilization of active heme proteins in the solid glass media could serve as a more practical biocatalysis. In another (SNAP) experiments, aged gels were placed in buVer, reduced with dithionite, and then SNAP was added to the solution.study to evaluate the cooperativity of O2 binding, the quaternary protein Hb was immobilized in both aerobic and anaerobic Gels were immersed in the SNAP solution for the stated times. Optical absorption spectra were measured using a Shimadzu silica thin films.12 In both cases, O2 binding to the Hb was non-cooperative.This is in contrast to solution, where it has UV-260 or Cary 3E spectrophotometer. Resonance Raman spectroscopy was performed on cyt-c by exciting the samples been well established that O2 binding is cooperative. The conclusion drawn from these studies was that Hb encapsulated at 514.5 nm (Q-band).Excitation at 406.7 and 413.1 nm laser lines (Soret band) were not feasible as the thermal energy led in the sol–gel glasses remained fixed in their original quaternary structures during the oxygenation or deoxygenation process. to cracking of the gel samples. The cyt-c used for resonance Raman studies was in the Fe(III) state and had concentrations It is likely that structural changes were not totally forbidden but the kinetics were considerably slowed by sol–gel encapsul- of 50 mg ml-1 and 25 mg ml-1 for solution and sol–gel samples, respectively.Thermal profile measurements were per- ation. This finding is quite interesting as it implies that sol–gel encapsulation can preferentially dictate the quaternary formed on a spectrophotometer equipped with circulating 46 J.Mater. Chem., 1999, 9, 45–53Table 1 Review of research on sol–gel encapsulation of heme proteins Heme protein(s) Purpose of study Finding(s) Ref. Biological function: ligand (O2, CO, NO) cyt-c, Mb, Hb Oxidation–reduction properties of heme and ligand 1 (1992) binding binding similar to that in solution 13 (1994) cyt-c, Mb, Hb, Biocatalysis: oxidation of dibenzothiophene Catalytic activities similar to that in solution 11 (1994) horseradish peroxidase Hb Oxygen binding: evaluation of cooperativity Oxygen binding to Hb in sol–gel systems non- 12 (1994) cooperative. Quaternary structures of Hb ‘fixed’ upon encapsulation cyt-c, Mb, Hb Ligand binding and sensing Ligand binding similar to that in solution and higher 14 (1995) absorption intensity with increasing ligand concentration Mb Sensing of dissolved oxygen (DO) in water DO concentrations at ppm levels can be measured within 16 (1995) minutes Mb Sensing of DO in water Detection limit for DO can be enhanced using 17 (1996) fluoresence instead of absorbance signals cyt-c and cyt-c Spin state and charge transfer activity Charge transfer activities between cyt-c and cyt-c 10 (1996) peroxidase peroxidase weak in sol–gels compared to that in solution cyt-c Kinetics of photo-induced electron transfer Encapsulation did not detectably perturb heme group but 9 (1997) reactions diVerent reaction kinetics observed in sol-gel vs.solution cyt-c EVect of alcohol (and other synthesis Encapsulation stabilized protein against aggregation due 8 (1997) conditions) on protein stability to methanol and made denaturation reversible cyt-c, Mb Binding to CO2 Formation of MetMb–CO2 adduct possible in sol–gels 15 (1998) whereas MetMb in solution denatures upon CO2 exposure structure of proteins, opening new areas of protein structure–function relationships. Since heme proteins such as Hb and Mb bind ligands O2, CO, and NO, by encapsulating these proteins in silica matrices, it is possible to create solid-state materials which can be used as sensing elements for these gases.The findings have been consistent in that formation of O2, CO, or NO adducts in sol–gel encapsulated Hb and Mb is similar to those adducts in solution, as observed using optical spectroscopy.1,13,14 Ligand binding experiments with the native heme proteins were conducted anaerobically.The formation of these adducts was reversible. Table 2 details the spectroscopic characteristics of Hb and Mb in the unligated [met or Fe(III )] and ligated (CO, NO) forms. Aged silica gels with Hb or Mb in the Fe(III ) state can be stored for at least one year with no significant changes in spectroscopy or ability to bind ligands.From Table 2, it is apparent that upon CO binding, silica-encapsulated HbCO and MbCO exhibited the same spectroscopic features as in solution. Fig. 1 shows the absorption spectra of metHb/HbCO and metMb/MbCO in solution and in silica gels. Upon NO binding, silica-encapsulated MbNO also exhibited the same spectroscopic features as in Fig. 1 (a) Optical spectra of MetHb and HbCO indicate successful solution, although the Soret band for HbNO was slightly blue CO binding in the silica-encapsulated Hb, resulting in the same shifted as compared to that in solution. The slight blue shift spectroscopic changes as those observed in aqueous buVer. (b) Optical in the HbNO Soret band was consistently observed (Table 2). spectra of MetMb and MbCO in aqueous buVer and aged silica gel also show the same spectroscopic changes upon CO binding.Table 2 Optical absorption characteristics of heme proteins Fig. 2 shows the absorption spectra of MetHb/HbNO and Soret/nm b/nm a/nm MetMb/MbNO in solution and in silica gels. In a recent study, State of the encapsulation of MetMb in sol–gel derived gels permitted protein Soln.Sol–gel Soln. Sol–gel Soln. Sol–gel Ref. formation of a CO2 adduct.15 This contrasts dramatically with MetMb 409 409 13 the behavior of MetMb in solution, where it denatures when 410 410 14 exposed to CO2. Therefore, sol–gel encapsulation made poss- MbCO 423 423 541 541 578 578 13 ible a CO2 complex which could not be obtained in aqueous 424 424 542 540 578 578 14 buVer. MbNO 418 418 533 533 574 574 13 Since the heme proteins retain their reactivity in optically 418 418 534 534 574 574 14 transparent glasses, these materials can be used as sensing MetHb 406 406 13 406 406 14 elements.We have used sol–gel encapsulated Mb as a sensing HbCO 420 420 538 538 569 568 13 element for measurement of dissolved oxygen in water using 420 420 540 540 570 566 14 optical spectroscopy.16 The dissolved oxygen concentration HbNO 418 416 535 533 567 565 13 was determined quantitatively by observing the rate of change 416 412 534 530 566 562 14 of the visible absorption spectrum.The overall change in J. Mater. Chem., 1999, 9, 45–53 47is likely to interact with the silica pore walls. In this section, we specifically address the following issues: 1) the influence of the silica matrix on the protein upon encapsulation and subsequent gel drying, 2) thermal denaturation of silicaencapsulated cyt-c, and 3) pH eVects on silica-encapsulated cyt-c.Optical absorption and resonance Raman studies Absorption spectra of Fe(III) cyt-c in solution (0.1 M acetate, pH 4.5) and encapsulated in an aged silica gel (equilibrated with 0.1 M acetate, pH 4.5) are shown in Fig. 3. In the aged state, the pores are filled with liquid and there has been no change in the original volume of the gel. Both samples exhibited the characteristic pattern of cyt-c, indicating that the heme signature is preserved upon encapsulation of the protein in the gel. Although the overall characteristics are similar, slight changes in peak wavelengths are observed. The spectrum of the protein in solution shows the Soret band maximum at 406 nm, while the spectrum of the protein in an aged gel shows a slight blue shift in the Soret band maximum from 406 to 404 nm.The slight blue shift of cyt-c upon encapsulation was Fig. 2 (a) Optical spectra of MetHb and HbNO indicate successful also observed by Blyth et al. wherein the solution had its Soret NO binding in the silica-encapsulated Hb.The spectroscopic changes band maximum at 410 nm, while the sol–gel sample had its are the same as those in aqueous buVer, except for a slight blue shift Soret band maximum at 408 nm.14 In their experiments, the in the Soret peak maximum and decrease in intensity. (b) Optical samples were equilibrated with pH 7.0 phosphate buVer.spectra of MetMb and MbNO in aqueous buVer and aged silica gel also show the same spectroscopic changes upon NO binding. Xerogels were prepared by ambient drying of the aged gel samples. During the transformation to the xerogel state, there are both weight and volume changes. In the xerogel state, absorbance at 431.5 and 436 nm was linear with time and most of the liquid in the pores has evaporated, accompanied directly proportional to the concentration of dissolved oxygen.by a collapse of the pores and a considerable shrinkage in The dissolved oxygen concentrations examined in the study volume. The xerogel is typically #15% of the original volume ranged from 2 to 8 ppm. Moreover, the optical response can and #20% of the original weight. The spectrum of cyt-c in a be established rapidly as the rate of absorbance change was xerogel is shown in Fig. 4. As previously reported,8 the blue established within 2 min. The gels were also reusable since O2 shift in the cyt-c Soret band continues along the aged gel to binding is reversible, and used gels can be regenerated to xerogel transition in the silica matrix, with the xerogel sample metMb by ‘washing’ with aerated buVer to remove the dithionshowing a Soret band maximum at 395 nm.There is a total ite. Mb encapsulated in sol–gel silica proved to be an accurate downshift of 9 nm (from 404 to 395 nm) that can be attributed and reproducible sensing element for measurement of dissolved to drying eVects. These eVects arise due to either pore collapse oxygen at ppm concentrations.In a subsequent study, the or to loss of solvent phase from the aged gel samples that fluorescent dye brilliant sulfaflavine was used to enhance the occurs during drying. detection limit for DO measurements.17 In order to determine the exact cause of the blue shifts, The accumulation of research data on heme proteins clearly absorption spectra were obtained on rehydrated xerogel demonstrate that these proteins retain their characteristic samples.Xerogels were rehydrated with 0.1 M acetate buVer optical signatures and chemical function upon sol–gel encap- (pH 4.5), with no swelling of the gels. Absorption spectra of sulation. By isolating and ‘caging’ the biomolecules in a porous xerogels immersed in solution show a distinctly red shifted silica matrix, interesting eVects and new features have been observed, such as 1) cyt-c was stabilized from aggregation in methanol and methanol denaturation was reversible, 2) electron paramagnetic resonance on cyt-c could be performed at room temperature, 3) a MetMb–CO2 adduct was formed which was not possible in aqueous buVer, and 4) the quaternary structure of a polymeric protein such as Hb could be ‘fixed’.In addition, sol–gel immobilized heme proteins can bind ligands such as O2, CO, and NO. Results with O2 and immobilized Mb demonstrated that O2 concentrations can be quantitatively determined in a matter of minutes. Therefore, these protein-doped glasses can be explored as potential solidstate optical detectors and sensors. Encapsulation of cytochrome c: biomolecules that design self-specific pores Cytochrome c is an electron transport protein in which its heme Fe is reversibly oxidized and reduced between the Fe(II ) and Fe(III) oxidation states.18,19 Its absorption signature has been carefully studied and established in solution.We report here our investigation of cyt-c to determine changes in its structure as a consequence of sol–gel encapsulation and gel Fig. 3 Optical spectra of cyt-c in aqueous buVer and aged gel show a drying. A biomolecule such as cyt-c, by virtue of a relatively slight blue shift in the Soret band maximum for the aged gel (404 nm) as compared to aqueous buVer (406 nm). high molecular weight (#12,400 Da) and high positive charge, 48 J. Mater. Chem., 1999, 9, 45–53leading to the conclusion that the sol–gel encapsulation did not appreciably alter the spin state, oxidation state, or geometry of the heme site.The finding that both the aged and rehydrated xerogels had the same optical absorption and RR characteristics suggests that the microenvironment of cyt-c is not influenced by the stresses caused by pore shrinkage during drying. One mechanism which could account for this behavior is that the biomolecule designs a self-specific pore as the silica network forms from hydrolysis and condensation reactions during the sol–gel process.That is, the silica cage that defines the pore forms around the biomolecule according to the size and shape requirements of the biomolecule. The presence of the biomolecule prevents its surrounding pore from collapsing during gel drying.Thermal denaturation studies Globular proteins such as cyt-c possess non-covalent interactions (hydrogen bonds, van der Waals bonds, etc.) that maintain the native folded state under physiological conditions. Disruption of the non-covalent interactions by thermal energy Fig. 4 Upon drying, the xerogel cyt-c Soret band is further blue shifted leads to denaturation by unfolding. Thermally induced to 395 nm.After rehydration, however, the Soret band is red shifted unfolding of the proteins in solution, in general, is highly back to 404 nm, the same value as that for an aged gel. The blue shift in the Soret band that accompanies drying, therefore, is reversible cooperative with a sharp transition over a small temperature upon rehydration.range. The transition point at which half of the molecules are denatured is termed Tm. The increased intensity of the cyt-c heme Soret band in the unfolded state as compared to the Soret band that once again has its maximum at 404 nm, as native state can be used to monitor the thermal denaturation shown in Fig. 4.8 This value is exactly the same as that process,21,22 and optical monitoring of the heme band at observed for aged gel samples.Thus, there is a reversible 9 nm 400 nm was used in the experiments described here. As shown shift in the Soret band maximum that accompanies drying in Fig. 5, the solution sample of cyt-c (0.1 M sodium acetate and subsequent rehydration. The reversibility of these shifts buVer, pH 4.5) showed a transition to the denatured state with indicates that the blue shift during drying was due to evapora Tm #65 °C, whereas the aged gel sample (equilibrated with ation of the solvent phase and not due to physical constraint 0.1 M sodium acetate buVer, pH 4.5) did not show any of the protein from pore collapse.The reversibility of the transition up to ca. 95°C. For the sol–gel sample, it was not Soret band shift upon rehydration in xerogels suggests that possible to increase the temperature beyond 95 °C as boiling cyt-c experienced similar metal–heme interactions in the xerogof the buVer led to gel cracking.Since both experiments were els as in aged gels even after substantial shrinkage in gel performed under the same conditions, the observed increase volume and pore size.in thermal tolerance of the protein can be ascribed to the Resonance Raman (RR) frequencies for Fe(III) cyt-c in influence of the silica gel matrix. solution, aged gel, and rehydrated xerogel are summarized in The thermal denaturation experiments underscore the Table 3, along with mode assignments.20 It was not possible pronounced influence of the physical eVects of caging.The to obtain spectra on dried xerogel samples due to thermally induced sample cracking under laser irradiation. For the rehydrated xerogel samples, only the strongly enhanced modes could be observed due to extensive background and poor signal to noise ratio. As seen in Table 3, the overall spectral pattern is preserved upon sol–gel encapsulation, although a few peaks were upshifted. The RR frequencies for the aged and rehydrated xerogel samples were essentially the same, indicating that the structure of the heme group in xerogels was not substantially altered as compared to that in aged gels.Results from RR suggest that volume shrinkage during gel drying did not significantly alter the ground state geometry of the heme in cyt-c. RR studies on cyt-c by Shen and Kostic showed all vibrational bands in the solution and sol–gel samples had identical wavenumbers within experimental error.9 The relative intensity of most bands also remained unchanged, Table 3 Resonance Raman frequencies (cm-1) of cytochrome-c Mode Cyt-c in Cyt-c in Cyt-c in assignment20 aqueous buVer aged gel rehydrated xerogel n22 1130 1131 Fig. 5 Thermal denaturation profiles of cyt-c in aqueous buVer and n30 1175 1175 1172 in an aged gel indicate that there is a substantial improvement in n13 1232 1235 thermal stability as a result of sol–gel encapsulation.Tm#65 °C in n21 1315 1317 1318 aqueous buVer, whereas denaturation does not begin until ca. 95°C n4 1366 1370, 1376 1376 in the aged gel. It was not possible to increase the temperature beyond n29 1404 1410 1410 95 °C as boiling of the buVer led to gel cracking.The absorbance n11 1551 1553 values are higher for the aged samples than for the solution sample n19 1584 1590 because of a higher cyt-c concentration in the aged gel. J. Mater. Chem., 1999, 9, 45–53 49substantial stability provided by the matrix supports the silica matrix depending upon pH. Whether the diVerence is due to the matrix hindering conformational transitions or due hypothesis that the protein molecules dictate the required size of its surrounding pore.The presence of a rigid cage structure to the charged silicate groups on the matrix surface (pI#2) interacting with the charged residues on cyt-c cannot be around the protein then restrains the conformational mobilities of the protein thereby thermally stabilizing it against thermal resolved with the current data.In summary, our studies on cyt-c revealed that the overall denaturation. structure of this protein is retained upon sol–gel immobilization, although there may be subtle diVerences with respect to pH Studies protein conformation. Optical absorption spectra showed a Conformations of cyt-c are pH dependent and several confor- blue shift in the spectrum as a result of encapsulation in aged mational transitions accompany changes in pH.23,24 In the gels that became more pronounced in xerogels.Upon rehyacidic region, at least two pH induced transitions have been dration of the xerogel, however, the eVects were reversible to characterized, one centered at pH#3 and the other at pH#1.the aged gel state. The results suggest that the protein dopant It has been concluded that the transitions accompany a change forms a self-specific ‘cage’, so that even as pores collapse, in the spin-state of the central Fe atom. There is a high-spin there is minimal change in the physical constraint imposed by configuration of the Fe(III) at low pH as compared to a low- the matrix.Our thermal studies demonstrated one of the spin configuration existing at neutral pH. The transition benefits of sol–gel encapsulation, that thermal stability of the centered around pH#3 has been characterized as due to protein was significantly improved in immobilized cyt-c. While replacement of axial Met-80 residue with Lys-79, whereas the cyt-c in solution undergoes unfolding with a Tm#65 °C, the transition centered around pH#1 probably involves replace- protein encapsulated in the aged gel did not show any transment of both axial residues with terminal water molecules.25 ition up to ca. 95°C. The presence of the physical cage in the Optical monitoring of the Soret band wavelength at diVerent sol–gel matrix stabilizes the folded state as compared to the pH values provides a good estimate of the pH induced unfolded state.A similar improvement in protein stability to conformational changes in the protein. Overall conformational alcohol has also been observed for cyt-c in silica gels.8 changes occur in the protein as the polypeptide subunits are protonated or deprotonated. At pH>4, the solution spectrum shows its Soret band maximum centered at 406 nm which Sol–gel encapsulated manganese myoglobin: shifts to ca. 395 nm at pH 1. The optical absorption changes detector for nitric oxide indicate a titration of the amino acid groups with the protons. In solution, two transitions centered around pH#0 and pH#3 As discussed previously, native Mb and Hb with Fe in the can be observed, as shown in Fig. 6.Within the limits of heme can bind O2, CO, and NO. Manganese myoglobin experimental error, these values correspond well with the (MnMb), however, is an excellent alternative to native Mb reported literature. On the other hand, similar experiments on for NO detection because the manganese-substituted proteins cyt-c encapsulated in aged gels indicate a more continuous bind NO but not O2.26 Moreover, NO-binding in MnMb can variation in the pH profile, also shown in Fig. 6. A large also be detected optically by using the a peak of MnMbNO change in Soret peak wavelength is observed in the pH range at 580 nm as the optical marker. 3 to 5 with a midpoint at pH#4. At low pH, the sharp Nitric oxide chemistry has been in the forefront of research transition observed in solution is no longer evident in the aged recently because of its importance in physiological activities gel samples.Instead the conformational change accompanying ranging from maintaining vascular tone to antimicrobial pH variation is spread out over a pH range of 3 to 0. defense.27 NO is synthesized in cells via the oxidation of LThe observed diVerences in the proton dependent behavior arginine to L-citrulline by the enzyme NO synthase (NOS), as of the cyt-c in solution and in aged gels show that the presence shown in eqn.(1). of the silica matrix may have altered the acid–base properties of the protein. As changes in pH induce conformational L-arginine CCCCCCCDA O2, NADPH, NOSynthase L-citrulline+NO (1) transitions in cyt-c, it is possible that the constraints imposed by encapsulation within a pore of finite dimension restrict NO is a free radical and the biological lifetime of NO is of those transitions.Another possibility is that the charged the order of seconds. A variety of sensors have been developed residues on cyt-c react diVerently with the pore walls of the to measure NO levels. A porphyrinic-based microsensor has been developed to detect NO electrochemically,28 and optical methods have also been developed based on bacteria denitrifi- cation,29 conversion of HbO2 to metHb,30 chemiluminescence of ozone treated NO,31 luminol chemiluminescence,32 guanylate cyclase activation,33 and other NO-mediated eVects.More recently Aylott et al.34 demonstrated that cyt-c can also be used for optical NO sensing.By using sol–gel thin films with encapsulated cyt-c, a reversible NO sensor was developed that can measure NO at ppm levels within minutes. Barker et al. developed a reversible fiber-optic NO sensor based on cyt-c¾ that can also measure NO at ppm levels with <1 s response time. In both cases, there was no interference from O21 and with cyt-c¾, interference from NO2 can also be eliminated.An ideal optical sensor for NO is one which specifically binds NO with high aYnity, producing a measurable change in the absorption spectrum. The ability to conduct the experiments aerobically is also an important factor, since working in an O2-free environment is physiologically unreasonable. To evaluate the sensitivity of MnMb for NO sensing under Fig. 6 Optical monitoring of the Soret band wavelength as a function simulated physiological conditions, we conducted an experi- of pH shows a diVerence in the proton-dependent behavior of cyt-c in aged silica gels as compared to aqueous buVer.ment in which NO was generated enzymatically from the 50 J. Mater. Chem., 1999, 9, 45–53Table 4 Reaction of nitrogen oxides with manganese myoglobin enzyme nitric oxide synthase (NOS).In this experiment, MnMb was in solution rather than in a silica gel so that NO Nitrogen oxide Experimental conditions NO-Binding transport would not be impeded by the silica matrix. The experiment was conducted in the presence of air. After mixing NO (gas) [NO]=0.8 mM Yes an aqueous solution of the NOS enzyme, L-arginine, the NO2 (gas) [NO2]=3.7 mM Yes NO2- (solid NaNO2) [NO2 -]<10 mM No required cofactors and MnMb, optical absorption was [NO2-]=45 mM Yes recorded periodically between 500 and 600 nm to detect gener- NO- (solid Na2N2O3) [NO-]=5.5 mM Yes ation of NO as evidenced by the a peak of MnMbNO NO3- (solid NaNO3) [NO3 -]=45 mM No at 580 nm.Fig. 7 shows the optical absorption of the NOS–MnMb solution as a function of time.The data clearly demonstrate that enzymatically generated NO can be detected optically by directly monitoring the NO adduct. The level of NO generated in the NOS–MnMb experiment was estimated by measuring the enzymatic activity, i.e. nmole of citrulline generated per mg of enzyme. As shown in eqn. (1), citrulline and NO are formed in equimolar amounts. The measured levels of citrulline generated per mg of NOS after 30, 60, 90, and 120 min were 161, 239, 212, and 256 nmol, respectively.In the experiment reported here ca. 0.12 mg of enzyme was used. Although optical detection of NO was reported previously using HbO2,30 MnMb represents a better choice of protein because one can measure the NO-adduct directly and also because HbO2 will eventually oxidize to metHb without NO in the presence of air.To determine the specificity of MnMb for NO sensing, we carried out control experiments in which MnMb in solution was reacted with other nitrogen oxides, specifically NO-, NO2, NO2-, and NO3-. All experiments were performed with MnMb in aqueous buVer. We found that MnMb binds NO- and NO2 but not NO2- or NO3-. NO- was generated indirectly via the decomposition of Na2N2O3, which releases HNO (NO-) in aqueous solution36 as shown in eqn.(2); Na2N2O3=HNO+NO2- (2) NO2- does not bind MnMb. However, it produces NO at suYciently high concentrations according to eqn. (3). 3HNO2=HNO3+2NO+H2O (3) Table 4 lists the diVerent nitrogen oxide species tested and their ability or inability to bind MnMb. Fig. 8 Successful NO detection was achieved in silica-encapsulated We successfully encapsulated MnMb in optically transparent MnMb, as the same characteristic absorption changes were observed sol–gel derived silica gels and used these protein-doped mate- upon metal reduction and NO binding as compared to MnMb in rials as NO detectors. As mentioned previously, a significant aqueous buVer.The slight reduction in absorption intensity for the aged gel was most likely caused by the decomposition of the reducing advantage of using MnMb to detect NO is that experiments agent (sodium dithionite).The gel was immersed in a dithionite can be performed aerobically. Fig. 8 shows the absorption solution which became cloudy over time in air. spectra of Mn(III)Mb, Mn(II)Mb, and MnMbNO in solution and in an aged silica gel.Optical absorption characteristics of MnMb and its NO adduct are detailed in Table 5. A comparison of the spectra shows essentially the same spectroscopic properties upon metal reduction and NO binding in the aged gel as in solution. In addition to binding dissolved gaseous NO, silica-encapsulated MnMb can also bind NO released from a chemical. SNitroso- N-acetylpenicillamine (SNAP) releases NO in an aqueous medium.37 Fig. 9 shows the absorption spectra of MnMb aged gels reacted with SNAP, and NO binding is evident. There was a predictable concentration eVect as the 1.5 mM SNAP concentration required a longer incubation Table 5 Optical absorption characteristics of manganese myoglobin Soret/nm b/nm a/nm Description Soln. Sol–gel Soln.Sol–gel Soln. Sol–gel Mn(III )Mb 373 371 Fig. 7 Optical monitoring of a solution containing NO synthase and 468 468 MnMb at varying incubation times demonstrates that NO generated Mn(II )Mb 438 438 by NO synthase can be detected. The a peak at 580 nm of MnMbNO Mn(II )MbNO 424 424 538 538 580 579 was used as the optical marker for NO binding. J. Mater. Chem., 1999, 9, 45–53 51Fig. 9 Optical monitoring of MnMb in silica gels reacted with NO released from S-nitroso-N-acetylpenicillamine (SNAP) shows NO binding. As expected, there was a higher level of the NO adduct with a higher SNAP concentration. time (>120 min) for complete NO binding compared with the The authors thank Dr Yumiko Komori for preparation of the NOS enzyme and SNAP, Naiyma Houston and James Hauser 5.3 mM SNAP concentration (60 min).These experiments demonstrate the potential for using for their experimental assistance, and Dr. Daryl Eggers for his critical reading of the manuscript. The support of this research encapsulated heme proteins in sol–gel matrices as sensing elements. A simple yet eVective optically-based technique to by the National Science Foundation (DMR-9408780) is greatly appreciated. detect NO was developed using sol–gel encapsulated MnMb.The Mn-substituted heme makes aerobic experiments possible, which is critical for physiological conditions. The advantages of using heme proteins [cyt-c, cyt-c¾, MnMb] include direct References detection of NO via formation of an NO-adduct and the 1 L. M. Ellerby, C. R. Nishida, F.Nishida, S. A. Yamanaka, ability to detect NO without interference from O2 or air. One B. Dunn, J. S. Valentine and J. I. Zink, Science, 1992, 255, 1113. disadvantage of MnMb is that it is not specific for the NO 2 D. Avnir, S. Braun and M. Ottolenghi, in Supramolecular free radical but will also react with NO- and NO2 to form Architecture in Two and Three Dimensions, ed. T. Bein, American MnMbNO.In sensor applications, the sol–gel encapsulated Chemical Society, New York, 1992. MnMb should be in thin-film form to maximize kinetics of 3 B. C. Dave, B. Dunn, J. S. Valentine and J. I. Zink, Anal. Chem., analyte diVusion through the pores. 1994, 66, 1120A. 4 D. Avnir, S. Braun, O. Lev and M. Ottolenghi, Chem. Mater., 1994, 6, 1605. Conclusions 5 S. Braun, S.Rappoport, R. Zusman, D. Avnir and M. Ottolenghi, Mater. Lett., 1990, 10, 1. Heme proteins can be successfully encapsulated in silica 6 E. Antonini and M. Brunori, Hemoglobin and Myoglobin in their matrices with minimal changes to their spectroscopic proper- Reactions with Ligands, North-Holland Publishing, Amsterdam, ties. Research by various groups has shown that the immobil- 1971.ized heme proteins also retain their chemical functions of 7 T. Yonetani and T. Asakura, J. Biol. Chem., 1969, 244, 4580. 8 B. C. Dave, J. M. Miller, B. Dunn, J. S. Valentine and J. I. Zink, oxidation/reduction, ligand (O2, CO, NO) binding, and biocat- J. Sol–Gel Sci. Technol., 1997, 8, 629. alysis. The sol–gel encapsulated proteins can be used to 9 C. Shen and N. M. Kostic, J.Am. Chem. Soc., 1997, 119, 1304. quantitatively determine concentration of dissolved gases as 10 C. T. Lin, C. M. Catuara, J. E. Erman, K. C. Chen, S. F. Huang, dissolved O2 concentrations could be measured within minutes. W. J. Wang and H. H. Wei, J. Sol–Gel Sci. Technol., 1996, 7, 19. Moreover, immobilization using the sol–gel approach has 11 S. Wu, J. Lin and S. I. Chan, Appl. Biochem.Biotechnol., 1994, enabled new features such as room temperature electron 47, 11. 12 N. Shibayama and S. Saigo, J. Mol. Biol., 1995, 251, 203. paramagnetic resonance, formation of a MetMb–CO2 adduct, 13 E. H. Lan, M. S. Davidson, L. M. Ellerby, B. Dunn, J. S. Valentine and fixation of Hb’s quaternary structure. The conclusions and J. I. Zink, Mater. Res. Soc. Symp. Proc., 1994, 330, 289.from our new studies include the finding that encapsulation 14 D. J. Blyth, J. W. Aylott, D. J. Richardson and D. A. Russell, in a porous silica matrix can significantly improve the thermal Analyst, 1995, 120, 2725. stability of a protein. 15 Q. Ji, C. R. Lloyd, W. R. Ellis and E. M. Eyring, J. Am. Chem. Thermal denaturation studies on cyt-c demonstrated that Soc., 1998, 120, 221. 16 K. E. Chung, E. H. Lan, M. S. Davidson, B. Dunn, J. S. Valentine whereas the protein in aqueous buVer showed a transition to and J. I. Zink, Anal. Chem., 1995, 67, 1505. the denatured state with Tm#65 °C, the sol–gel immobilized 17 M. F. McCurley, G. J. Bayer and S. A. Glazier, Sens. Actuators B, protein did not show any transition up to ca. 95°C. A similar 1996, 36, 491.improvement in protein stability to alcohol has also been 18 R. A. Scott and A. G. Mauk (Editors), Cytochrome c: A achieved using sol–gel encapsulation. Another key finding from Multidisciplinary Approach, University Science Books, Sausalito, our studies was that sol–gel encapsulated MnMb can serve as CA, 1996. 19 G. R. Moore and G. W. Pettigrew, Cytochromes c: Evolutionary, solid-state detectors or sensors for NO.Our experiments estab- Structural, and Physicochemical Aspects, Springer-Verlag, Berlin, lished that sol–gel immobilized MnMb binds NO without 1990. interference from O2, similar to MnMb in solution. Finally, an 20 S. Hu, I. K. Morris, J. P. Singh, K. M. Smith and T. G. Spiro, intriguing hypothesis based on collective results from optical J.Am. Chem. Soc., 1993, 115, 12446. absorption, resonance Raman, and thermal denaturation studies 21 T. Uno, Y. Nishimura and M. Tsuboi, Biochemistry, 1984, 23, on cyt-c is that this protein designs self-specific pores in the 6802. 22 D. S. Cohen and G. J. Pielak, Protein Sci., 1994, 3, 1253. silica network according to its size and shape requirements. 52 J. Mater. Chem., 1999, 9, 45–5323 T.Kitagawa, Y. Ozaki, J. Teraoka, Y. Kyogoku and 32 K. Kikuchi, T. Nagano, H. Hayakawa, Y. Hirata and M. Hirobe, J. Biol. Chem., 1993, 268, 23106. T. Yamanaka, Biochim. Biophys. Acta, 1977, 494, 100. 24 Y. P. Myer, R. B. Srivasatava, S. Kumar and K. Raghvendra, 33 K. Ishii, H. Sheng, T. D. Warner, U. Forstermann and F. Murad, Am. J. Physiol., 1991, 261, H598. J.Protein Chem., 1983, 2, 13. 25 B. Cartling, Biological Applications of Raman Spectroscopy, ed. 34 J. W. Aylott, D. S. Richardson and D. A. Russell, Chem. Mater., 1997, 9, 2261. T. G. Spiro, Wiley, New York, Vol. III, 1988. 26 C. Bull, R. G. Fisher and B. M. HoVman, Biochem. Biophys. Res. 35 S. L. R. Barker, R. Kopelman, T. E. Meyer and M. A. Cusanavich, Anal. Chem., 1998, 70, 971.Commun., 1974, 59, 140. 27 E. Culotta and D. E. Koshland, Science, 1992, 258, 1862. 36 F. T. Bonner and M. N. Hughes, Commun Inorg. Chem., 1988, 7, 215. 28 T. Malinski and Z. Taha, Nature (London), 1992, 358, 676. 29 J. Goretski and T. C. Hollocher, J. Biol. Chem., 1988, 263, 2316. 37 L. J. Ignarro, H. Lippton, J. C. Edwards, W. H. Baricos, A. L. Hyman, P. J. Kadowitz and C. A. Gruetter, J.Pharmacol. 30 M. P. Doyle and J. W. Hoekstra, J. Inorg. Biochem., 1981, 14, 351. 31 O. C. Zafiriou and M. McFarland, Anal. Chem., 1980, 52, 1662. Exp. Ther., 1981, 218, 739. Paper 8/05541F J. Mater. Chem., 1999, 9, 45–53 53 J O U R N A L O F C H E M I S T R Y Materials Feature Article Synthesis of sol–gel encapsulated heme proteins with chemical sensing properties† Esther H.Lan,a Bakul C. Dave,b Jon M. Fukuto,a Bruce Dunn,*a JeVrey I. Zinkc and Joan S. Valentinec aUniversity of California, Los Angeles, Department of Materials Science and Engineering, Los Angeles, CA 90095, USA bSouthern Illinois University, Department of Chemistry, Carbondale, IL, USA cUniversity of California, Los Angeles, Department of Chemistry and Biochemistry, Los Angeles, CA 90095, USA Received 18th May 1998, Accepted 16th July 1998 Heme proteins such as cytochrome-c (cyt-c), hemoglobin microorganisms.In general, sol–gel matrices are thermally and (Hb), and myoglobin (Mb) have been successfully encap- chemically stable, can be obtained in a variety of forms such sulated in sol–gel derived silica matrices, retaining their as bulk monoliths or thin films, and can store very high spectroscopic properties and chemical function.The ther- concentrations of proteins due to the porous network. Proteins mal stability of cyt-c was significantly improved by immob- and enzymes encapsulated in sol–gel derived glasses can interilization in a porous silica network. Results from optical act with target molecules with a high degree of specificity, and absorption, resonance Raman, and thermal denaturation using an appropriate sensing scheme, produce a detectable studies suggest that biomolecules such as cyt-c design self- signal.While the relatively large biomolecules are immobilized specific pores in the silica network according to the size within the silica network, the porous network allows small and shape requirements of the biomolecule.Hb and Mb, ions or molecules to be transported into the interior of the immobilized using the sol–gel process, bound ligands simimatrix. There is, however, a limit in the molecular size of the lar to the proteins in aqueous buffer, and silica-encapsuanalyte that can be detected since it must be able to travel lated manganese myoglobin (MnMb) was a viable detector through the pore network.for nitric oxide (NO). At the time of gelation, the matrix may be considered as a two-phase system consisting of a porous solid and a trapped Introduction aqueous phase. This two-phase material is termed an aged gel. As the gel is allowed to dry, liquid is expelled from the pores The encapsulation of proteins in sol–gel derived silica glasses and the gel shrinks as the pores collapse, forming a dried gel, has been widely studied in recent years.Research in this field termed a xerogel. Xerogels are approximately one-eighth of has established that, upon encapsulation, proteins retain their the original volume, 50% porous (by volume), and dimen- spectroscopic properties and enzymes retain their biological sionally invariant.The pores of the gels are negatively charged activity.1–4 Immobilizing proteins by physically trapping the (pI of silica ca. 2) at the pH of the buVers (pH#7) generally molecules in individual pores of a matrix permits the protein used to encapsulate biomolecules in the sol–gel process. Such molecules to be isolated and stabilized. Conventional immobilpores can show aYnity towards polar and charged dopant ization techniques such as covalent attachment results in molecules, and they may also alter peripheral hydrogen- chemical modifications of the protein.In contrast, sol–gel bonding interaction of biomolecules. As the conformations of immobilization is characterized by physical entrapment withthe biomolecules are maintained by a large sum of relatively out chemical modification.One potential advantage is that the weaker non-covalent interactions, these eVects can perturb the silica matrix ‘cages’ the biomolecule, providing a far more structural dynamics of trapped entities. The entrapment of ‘rugged’ environment for the dopant molecule in contrast to biomolecules inside this silicate ‘cage’, therefore, can produce surface attachment schemes.The reaction chemistry of sol–gel slight alterations in conformational structure and possibly immobilized biomolecular systems has been shown to be biological function. analogous to that in aqueous solution except for the observed In this paper, we will review the literature published to date rates of chemical reactions which are generally slower due to on sol–gel encapsulation of heme proteins and summarize the diVusion limitations in a porous silica matrix.The first studies key findings. Proteins containing heme are characterized by on sol–gel immobilization of biomolecules by Avnir and cothe presence of Fe-protoporphyrin IX, which gives the protein workers involved trapping of the enzyme alkaline phosphatase its characteristic spectroscopic properties.In addition to which produced a bioactive powder.5 We developed synthesis reviewing the literature in this area, we include new studies on conditions that produced optically transparent monoliths with the heme proteins cytochrome c (cyt-c), hemoglobin (Hb), trapped proteins, enabling these materials to be used as myoglobin (Mb), and manganese myoglobin (MnMb).For optical sensors.1 cyt-c, we investigated possible modifications to protein struc- Physical entrapment of proteins in a silica matrix without ture as a result of encapsulation, the extent to which the chemical modification preserves protein structure and funcmatrix can stabilize the protein from thermal denaturation, tionality and protects the protein. Isolating protein molecules and the eVect of pH on the protein in silica as compared to in individual pores of a silica matrix permits the molecules to the protein in solution. For Hb, Mb, and MnMb, we evaluated be stabilized and prevents degradation due to proteases or the feasibility of using sol–gel encapsulated heme proteins as sensing elements, specifically, for optical detection of nitric oxide (NO).We and other research groups have investigated †Basis of the presentation given at Materials Chemistry Discussion the sol–gel encapsulation of heme proteins for several reasons: No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. 1) they represent model systems for understanding sol–gel J. Mater. Chem., 1999, 9, 45–53 45encapsulation eVects as their protein structure and func- water which had its temperature controlled by a thermal programmer. The absorbance at 400 nm was monitored at tionality have been well documented in the literature, 2) their characteristic spectroscopic properties make optical detection 5 °C intervals from 25 to 95 °C.EVects of pH were tested by immersing a gel in phosphate buVer adjusted to appropriate possible, 3) they can bind ligands, resulting in changes in optical absorption that are readily observed, and 4) they serve pH values between 7 and 0.The gel was equilibrated with the buVer for about one day. The same sample was used as feasible detectors for the ligands of interest, making device applications possible. throughout the length of the experiment. Review of sol–gel encapsulation of heme proteins Experimental Silica sol was prepared using tetramethylorthosilicate (TMOS) Table 1 summarizes the research to date on sol–gel encapsulation of heme proteins and the key findings.Our earliest work as the precursor. 15.27 g of TMOS, 3.36 g of deionized water, and 0.22 g of 0.04MHCl were mixed and sonicated to produce on sol–gel encapsulation of proteins and enzymes demonstrated that optical spectroscopy of encapsulated cyt-c and the sol.For cyt-c, the sol was added to a Fe(III) cyt-c solution buVered with 0.1 M pH 4.25 acetate to achieve a final protein Mb was essentially identical to that in solution.1 In subsequent work on cyt-c, we found that immobilizing the protein in a concentration of 100 mM in the aged gels. For Hb, Mb, and MnMb, the pH of the sol was raised to pH#5.5 using 1.0 M silica matrix can stabilize the protein from aggregation due to methanol, and denaturation due to methanol was reversible in NH4OH and then added to an aqueous protein solution.The final protein concentrations in aged gels were 5 mM for native the encapsulated protein.8 The improvements in protein stability in the presence of alcohol were attributed to confinement Hb, 20 mM for native Mb, and 3 mM for MnMb.All proteins were encapsulated with the heme in the Fe(III) or Mn(III) and isolation of the biomolecule in pores of the silica network. In work by Shen et al., native cyt-c and zinc cyt-c were state. Gels were of dimensions x 1.8 cm×1.0 cm×0.4 cm. In the CO binding experiments with native Hb or Mb silica encapsulated in sol–gel silica glasses and studied using optical spectroscopy, circular dichroism, and resonance Raman spec- gels, aged gels with the protein in the Fe(III) or ‘met’ state were immersed in buVer (0.1 M pH 7.0 Tris–HCl), reduced troscopy.9 The results collectively suggested that encapsulation in the sol–gel glass only slightly perturbed the polypeptide with dithionite, and then exposed to gaseous CO, all under anaerobic conditions.The gels were gently stirred in CO- backbone and did not detectably perturb the heme group. In zinc cyt-c quenching experiments using [Fe(CN)6]3-, O2 and saturated buVer for 20 min. The saturation concentration of CO in water (1 atm, 20 °C) is 1 mM.6 In the NO binding p-benzoquinone, rate constants were consistently lower in the silica-encapsulated samples than in solution. The slower rates experiments with native Hb or Mb silica gels, aged gels in the Fe(III ) or ‘met’ state were immersed in buVer and exposed to can be attributed to diVusion processes through the porous network.Moreover, changes in ionic strength and pH aVected gaseous NO under anaerobic conditions. NO binding to Hb or Mb did not require reduction using sodium dithionite as kinetics diVerently in the gel as compared to solution, possibly due to the fact that positively charged cyt-c adsorbed to the NO can act as the reducing agent.The gels were gently stirred in the NO-saturated buVer for about 20 min. The saturation negatively-charged silica walls. The encapsulation of cyt-c–cyt-c peroxidase complexes in concentration of NO in water (1 atm, 20 °C) is 2 mM.6 MnMb was prepared by incorporating Mn protoporphyrin sol–gel derived gels allowed, for the first time, detection of this system’s electron paramagnetic resonance (EPR) signals into apomyoglobin using the method of Yonetani and Asakura.7 The experiments carried out using MnMb included at room temperature.10 Previously, quantitative EPR measurements of cyt-c and its complexes were performed at tempera- the study of reactions in solution as well as in a sol–gel matrix.All experiments were conducted in air. In the experiments tures below -100 °C, because EPR absorption of the heme group broadens at higher temperatures. By sol–gel encapsul- involving NO-synthase to enzymatically generate NO, the following constituents were mixed in the stated order: 1) 25 ml ation, however, a suitable microenvironment was obtained that permitted a measurable EPR spectrum at ambient tem- 0.01MpH 7.4 Tris–HCl buVer, 2) 25 ml 0.2 mg ml-1 calmodulin, 3) 25 ml 40 mM CaCl2, 4) 25 ml 2 mM NADPH, 5) 50 ml peratures because the immobilized protein was ‘frozen’ in the gel.The charge transfer activities between the Fe and porphyrin 1 mM tetrahydrobiopterin, 6) 25 ml 5mM L-arginine, 7) 30 ml 60 mM MnMb solution, 8) 20 ml 57.5 mM sodium dithionite, ring in cyt-c–cyt-c peroxidase complex were weak in the silica gels as compared to that in solution. Redox processes of the and 9) 350 ml 341 mg ml-1 NOS enzyme solution (semi-purified from rat brains).The temperature of this solution was main- encapsulated cyt-c–cyt-c peroxidase complex were more active in the aged gel stage than in the xerogel stage.tained at 37 °C during the experiment. In MnMb solution experiments with various nitrogen oxides, MnMb in 0.1 M In one study to evaluate biocatalysis of sol–gel encapsulated heme proteins, the sol–gel encapsulated proteins (cyt-c, Mb, pH 7.0 phosphate buVer was reacted with sodium dithionite and exposed to gaseous NO, gaseous NO2, solid NaNO2, Hb, horseradish peroxidase) had catalytic eYciencies in the oxidation of dibenzothiophene similar to those in solution.11 solid Na2N2O3, or solid NaNO3 of the stated concentrations. For the sol–gel encapsulated MnMb, aged gels were immersed Because of the advantage of easy separation of the encapsulated proteins from the liquid reaction mixture, it was suggested in buVer, reduced using sodium dithionite, and exposed to a 0.8 mM NO solution.In the S-nitroso-N-acetylpenicillamine that immobilization of active heme proteins in the solid glass media could serve as a more practical biocatalysis. In another (SNAP) experiments, aged gels were placed in buVer, reduced with dithionite, and then SNAP was added to the solution. study to evaluate the cooperativity of O2 binding, the quaternary protein Hb was immobilized in both aerobic and anaerobic Gels were immersed in the SNAP solution for the stated times.Optical absorption spectra were measured using a Shimadzu silica thin films.12 In both cases, O2 binding to the Hb was non-cooperative.This is in contrast to solution, where it has UV-260 or Cary 3E spectrophotometer. Resonance Raman spectroscopy was performed on cyt-c by exciting the samples been well established that O2 binding is cooperative. The conclusion drawn from these studies was that Hb encapsulated at 514.5 nm (Q-band). Excitation at 406.7 and 413.1 nm laser lines (Soret band) were not feasible as the thermal energy led in the sol–gel glasses remained fixed in their original quaternary structures during the oxygenation or deoxygenation process. to cracking of the gel samples.The cyt-c used for resonance Raman studies was in the Fe(III) state and had concentrations It is likely that structural changes were not totally forbidden but the kinetics were considerably slowed by sol–gel encapsul- of 50 mg ml-1 and 25 mg ml-1 for solution and sol–gel samples, respectively.Thermal profile measurements were per- ation. This finding is quite interesting as it implies that sol–gel encapsulation can preferentially dictate the quaternary formed on a spectrophotometer equipped with circulating 46 J. Mater. Chem., 1999, 9, 45–53Table 1 Review of research on sol–gel encapsulation of heme proteins Heme protein(s) Purpose of study Finding(s) Ref. Biological function: ligand (O2, CO, NO) cyt-c, Mb, Hb Oxidation–reduction properties of heme and ligand 1 (1992) binding binding similar to that in solution 13 (1994) cyt-c, Mb, Hb, Biocatalysis: oxidation of dibenzothiophene Catalytic activities similar to that in solution 11 (1994) horseradish peroxidase Hb Oxygen binding: evaluation of cooperativity Oxygen binding to Hb in sol–gel systems non- 12 (1994) cooperative.Quaternary structures of Hb ‘fixed’ upon encapsulation cyt-c, Mb, Hb Ligand binding and sensing Ligand binding similar to that in solution and higher 14 (1995) absorption intensity with increasing ligand concentration Mb Sensing of dissolved oxygen (DO) in water DO concentrations at ppm levels can be measured within 16 (1995) minutes Mb Sensing of DO in water Detection limit for DO can be enhanced using 17 (1996) fluoresence instead of absorbance signals cyt-c and cyt-c Spin state and charge transfer activity Charge transfer activities between cyt-c and cyt-c 10 (1996) peroxidase peroxidase weak in sol–gels compared to that in solution cyt-c Kinetics of photo-induced electron transfer Encapsulation did not detectably perturb heme group but 9 (1997) reactions diVerent reaction kinetics observed in sol-gel vs.solution cyt-c EVect of alcohol (and other synthesis Encapsulation stabilized protein against aggregation due 8 (1997) conditions) on protein stability to methanol and made denaturation reversible cyt-c, Mb Binding to CO2 Formation of MetMb–CO2 adduct possible in sol–gels 15 (1998) whereas MetMb in solution denatures upon CO2 exposure structure of proteins, opening new areas of protein structure–function relationships.Since heme proteins such as Hb and Mb bind ligands O2, CO, and NO, by encapsulating these proteins in silica matrices, it is possible to create solid-state materials which can be used as sensing elements for these gases.The findings have been consistent in that formation of O2, CO, or NO adducts in sol–gel encapsulated Hb and Mb is similar to those adducts in solution, as observed using optical spectroscopy.1,13,14 Ligand binding experiments with the native heme proteins were conducted anaerobically. The formation of these adducts was reversible.Table 2 details the spectroscopic characteristics of Hb and Mb in the unligated [met or Fe(III )] and ligated (CO, NO) forms. Aged silica gels with Hb or Mb in the Fe(III ) state can be stored for at least one year with no significant changes in spectroscopy or ability to bind ligands. From Table 2, it is apparent that upon CO binding, silica-encapsulated HbCO and MbCO exhibited the same spectroscopic features as in solution. Fig. 1 shows the absorption spectra of metHb/HbCO and metMb/MbCO in solution and in silica gels. Upon NO binding, silica-encapsulated MbNO also exhibited the same spectroscopic features as in Fig. 1 (a) Optical spectra of MetHb and HbCO indicate successful solution, although the Soret band for HbNO was slightly blue CO binding in the silica-encapsulated Hb, resulting in the same shifted as compared to that in solution.The slight blue shift spectroscopic changes as those observed in aqueous buVer. (b) Optical in the HbNO Soret band was consistently observed (Table 2). spectra of MetMb and MbCO in aqueous buVer and aged silica gel also show the same spectroscopic changes upon CO binding. Table 2 Optical absorption characteristics of heme proteins Fig. 2 shows the absorption spectra of MetHb/HbNO and Soret/nm b/nm a/nm MetMb/MbNO in solution and in silica gels. In a recent study, State of the encapsulation of MetMb in sol–gel derived gels permitted protein Soln. Sol–gel Soln. Sol–gel Soln. Sol–gel Ref. formation of a CO2 adduct.15 This contrasts dramatically with MetMb 409 409 13 the behavior of MetMb in solution, where it denatures when 410 410 14 exposed to CO2.Therefore, sol–gel encapsulation made poss- MbCO 423 423 541 541 578 578 13 ible a CO2 complex which could not be obtained in aqueous 424 424 542 540 578 578 14 buVer. MbNO 418 418 533 533 574 574 13 Since the heme proteins retain their reactivity in optically 418 418 534 534 574 574 14 transparent glasses, these materials can be used as sensing MetHb 406 406 13 406 406 14 elements. We have used sol–gel encapsulated Mb as a sensing HbCO 420 420 538 538 569 568 13 element for measurement of dissolved oxygen in water using 420 420 540 540 570 566 14 optical spectroscopy.16 The dissolved oxygen concentration HbNO 418 416 535 533 567 565 13 was determined quantitatively by observing the rate of change 416 412 534 530 566 562 14 of the visible absorption spectrum. The overall change in J.Mater. Chem., 1999, 9, 45–53 47is likely to interact with the silica pore walls. In this section, we specifically address the following issues: 1) the influence of the silica matrix on the protein upon encapsulation and subsequent gel drying, 2) thermal denaturation of silicaencapsulated cyt-c, and 3) pH eVects on silica-encapsulated cyt-c.Optical absorption and resonance Raman studies Absorption spectra of Fe(III) cyt-c in solution (0.1 M acetate, pH 4.5) and encapsulated in an aged silica gel (equilibrated with 0.1 M acetate, pH 4.5) are shown in Fig. 3. In the aged state, the pores are filled with liquid and there has been no change in the original volume of the gel.Both samples exhibited the characteristic pattern of cyt-c, indicating that the heme signature is preserved upon encapsulation of the protein in the gel. Although the overall characteristics are similar, slight changes in peak wavelengths are observed. The spectrum of the protein in solution shows the Soret band maximum at 406 nm, while the spectrum of the protein in an aged gel shows a slight blue shift in the Soret band maximum from 406 to 404 nm.The slight blue shift of cyt-c upon encapsulation was Fig. 2 (a) Optical spectra of MetHb and HbNO indicate successful also observed by Blyth et al. wherein the solution had its Soret NO binding in the silica-encapsulated Hb. The spectroscopic changes band maximum at 410 nm, while the sol–gel sample had its are the same as those in aqueous buVer, except for a slight blue shift Soret band maximum at 408 nm.14 In their experiments, the in the Soret peak maximum and decrease in intensity.(b) Optical samples were equilibrated with pH 7.0 phosphate buVer. spectra of MetMb and MbNO in aqueous buVer and aged silica gel also show the same spectroscopic changes upon NO binding.Xerogels were prepared by ambient drying of the aged gel samples. During the transformation to the xerogel state, there are both weight and volume changes. In the xerogel state, absorbance at 431.5 and 436 nm was linear with time and most of the liquid in the pores has evaporated, accompanied directly proportional to the concentration of dissolved oxygen. by a collapse of the pores and a considerable shrinkage in The dissolved oxygen concentrations examined in the study volume.The xerogel is typically #15% of the original volume ranged from 2 to 8 ppm. Moreover, the optical response can and #20% of the original weight. The spectrum of cyt-c in a be established rapidly as the rate of absorbance change was xerogel is shown in Fig. 4. As previously reported,8 the blue established within 2 min.The gels were also reusable since O2 shift in the cyt-c Soret band continues along the aged gel to binding is reversible, and used gels can be regenerated to xerogel transition in the silica matrix, with the xerogel sample metMb by ‘washing’ with aerated buVer to remove the dithionshowing a Soret band maximum at 395 nm. There is a total ite.Mb encapsulated in sol–gel silica proved to be an accurate downshift of 9 nm (from 404 to 395 nm) that can be attributed and reproducible sensing element for measurement of dissolved to drying eVects. These eVects arise due to either pore collapse oxygen at ppm concentrations. In a subsequent study, the or to loss of solvent phase from the aged gel samples that fluorescent dye brilliant sulfaflavine was used to enhance the occurs during drying.detection limit for DO measurements.17 In order to determine the exact cause of the blue shifts, The accumulation of research data on heme proteins clearly absorption spectra were obtained on rehydrated xerogel demonstrate that these proteins retain their characteristic samples. Xerogels were rehydrated with 0.1 M acetate buVer optical signatures and chemical function upon sol–gel encap- (pH 4.5), with no swelling of the gels.Absorption spectra of sulation. By isolating and ‘caging’ the biomolecules in a porous xerogels immersed in solution show a distinctly red shifted silica matrix, interesting eVects and new features have been observed, such as 1) cyt-c was stabilized from aggregation in methanol and methanol denaturation was reversible, 2) electron paramagnetic resonance on cyt-c could be performed at room temperature, 3) a MetMb–CO2 adduct was formed which was not possible in aqueous buVer, and 4) the quaternary structure of a polymeric protein such as Hb could be ‘fixed’.In addition, sol–gel immobilized heme proteins can bind ligands such as O2, CO, and NO. Results with O2 and immobilized Mb demonstrated that O2 concentrations can be quantitatively determined in a matter of minutes.Therefore, these protein-doped glasses can be explored as potential solidstate optical detectors and sensors. Encapsulation of cytochrome c: biomolecules that design self-specific pores Cytochrome c is an electron transport protein in which its heme Fe is reversibly oxidized and reduced between the Fe(II ) and Fe(III) oxidation states.18,19 Its absorption signature has been carefully studied and established in solution. We report here our investigation of cyt-c to determine changes in its structure as a consequence of sol–gel encapsulation and gel Fig. 3 Optical spectra of cyt-c in aqueous buVer and aged gel show a drying. A biomolecule such as cyt-c, by virtue of a relatively slight blue shift in the Soret band maximum for the aged gel (404 nm) as compared to aqueous buVer (406 nm).high molecular weight (#12,400 Da) and high positive charge, 48 J. Mater. Chem., 1999, 9, 45–53leading to the conclusion that the sol–gel encapsulation did not appreciably alter the spin state, oxidation state, or geometry of the heme site. The finding that both the aged and rehydrated xerogels had the same optical absorption and RR characteristics suggests that the microenvironment of cyt-c is not influenced by the stresses caused by pore shrinkage during drying.One mechanism which could account for this behavior is that the biomolecule designs a self-specific pore as the silica network forms from hydrolysis and condensation reactions during the sol–gel process.That is, the silica cage that defines the pore forms around the biomolecule according to the size and shape requirements of the biomolecule. The presence of the biomolecule prevents its surrounding pore from collapsing during gel drying. Thermal denaturation studies Globular proteins such as cyt-c possess non-covalent interactions (hydrogen bonds, van der Waals bonds, etc.) that maintain the native folded state under physiological conditions.Disruption of the non-covalent interactions by thermal energy Fig. 4 Upon drying, the xerogel cyt-c Soret band is further blue shifted leads to denaturation by unfolding. Thermally induced to 395 nm. After rehydration, however, the Soret band is red shifted unfolding of the proteins in solution, in general, is highly back to 404 nm, the same value as that for an aged gel.The blue shift in the Soret band that accompanies drying, therefore, is reversible cooperative with a sharp transition over a small temperature upon rehydration. range. The transition point at which half of the molecules are denatured is termed Tm. The increased intensity of the cyt-c heme Soret band in the unfolded state as compared to the Soret band that once again has its maximum at 404 nm, as native state can be used to monitor the thermal denaturation shown in Fig. 4.8 This value is exactly the same as that process,21,22 and optical monitoring of the heme band at observed for aged gel samples. Thus, there is a reversible 9 nm 400 nm was used in the experiments described here.As shown shift in the Soret band maximum that accompanies drying in Fig. 5, the solution sample of cyt-c (0.1 M sodium acetate and subsequent rehydration. The reversibility of these shifts buVer, pH 4.5) showed a transition to the denatured state with indicates that the blue shift during drying was due to evapora Tm #65 °C, whereas the aged gel sample (equilibrated with ation of the solvent phase and not due to physical constraint 0.1 M sodium acetate buVer, pH 4.5) did not show any of the protein from pore collapse.The reversibility of the transition up to ca. 95°C. For the sol–gel sample, it was not Soret band shift upon rehydration in xerogels suggests that possible to increase the temperature beyond 95 °C as boiling cyt-c experienced similar metal–heme interactions in the xerogof the buVer led to gel cracking.Since both experiments were els as in aged gels even after substantial shrinkage in gel performed under the same conditions, the observed increase volume and pore size. in thermal tolerance of the protein can be ascribed to the Resonance Raman (RR) frequencies for Fe(III) cyt-c in influence of the silica gel matrix.solution, aged gel, and rehydrated xerogel are summarized in The thermal denaturation experiments underscore the Table 3, along with mode assignments.20 It was not possible pronounced influence of the physical eVects of caging. The to obtain spectra on dried xerogel samples due to thermally induced sample cracking under laser irradiation. For the rehydrated xerogel samples, only the strongly enhanced modes could be observed due to extensive background and poor signal to noise ratio.As seen in Table 3, the overall spectral pattern is preserved upon sol–gel encapsulation, although a few peaks were upshifted. The RR frequencies for the aged and rehydrated xerogel samples were essentially the same, indicating that the structure of the heme group in xerogels was not substantially altered as compared to that in aged gels.Results from RR suggest that volume shrinkage during gel drying did not significantly alter the ground state geometry of the heme in cyt-c. RR studies on cyt-c by Shen and Kostic showed all vibrational bands in the solution and sol–gel samples had identical wavenumbers within experimental error.9 The relative intensity of most bands also remained unchanged, Table 3 Resonance Raman frequencies (cm-1) of cytochrome-c Mode Cyt-c in Cyt-c in Cyt-c in assignment20 aqueous buVer aged gel rehydrated xerogel n22 1130 1131 Fig. 5 Thermal denaturation profiles of cyt-c in aqueous buVer and n30 1175 1175 1172 in an aged gel indicate that there is a substantial improvement in n13 1232 1235 thermal stability as a result of sol–gel encapsulation.Tm#65 °C in n21 1315 1317 1318 aqueous buVer, whereas denaturation does not begin until ca. 95°C n4 1366 1370, 1376 1376 in the aged gel. It was not possible to increase the temperature beyond n29 1404 1410 1410 95 °C as boiling of the buVer led to gel cracking. The absorbance n11 1551 1553 values are higher for the aged samples than for the solution sample n19 1584 1590 because of a higher cyt-c concentration in the aged gel.J. Mater. Chem., 1999, 9, 45–53 49substantial stability provided by the matrix supports the silica matrix depending upon pH. Whether the diVerence is due to the matrix hindering conformational transitions or due hypothesis that the protein molecules dictate the required size of its surrounding pore.The presence of a rigid cage structure to the charged silicate groups on the matrix surface (pI#2) interacting with the charged residues on cyt-c cannot be around the protein then restrains the conformational mobilities of the protein thereby thermally stabilizing it against thermal resolved with the current data. In summary, our studies on cyt-c revealed that the overall denaturation.structure of this protein is retained upon sol–gel immobilization, although there may be subtle diVerences with respect to pH Studies protein conformation. Optical absorption spectra showed a Conformations of cyt-c are pH dependent and several confor- blue shift in the spectrum as a result of encapsulation in aged mational transitions accompany changes in pH.23,24 In the gels that became more pronounced in xerogels.Upon rehyacidic region, at least two pH induced transitions have been dration of the xerogel, however, the eVects were reversible to characterized, one centered at pH#3 and the other at pH#1. the aged gel state. The results suggest that the protein dopant It has been concluded that the transitions accompany a change forms a self-specific ‘cage’, so that even as pores collapse, in the spin-state of the central Fe atom. There is a high-spin there is minimal change in the physical constraint imposed by configuration of the Fe(III) at low pH as compared to a low- the matrix. Our thermal studies demonstrated one of the spin configuration existing at neutral pH.The transition benefits of sol–gel encapsulation, that thermal stability of the centered around pH#3 has been characterized as due to protein was significantly improved in immobilized cyt-c.While replacement of axial Met-80 residue with Lys-79, whereas the cyt-c in solution undergoes unfolding with a Tm#65 °C, the transition centered around pH#1 probably involves replace- protein encapsulated in the aged gel did not show any transment of both axial residues with terminal water molecules.25 ition up to ca. 95°C. The presence of the physical cage in the Optical monitoring of the Soret band wavelength at diVerent sol–gel matrix stabilizes the folded state as compared to the pH values provides a good estimate of the pH induced unfolded state. A similar improvement in protein stability to conformational changes in the protein.Overall conformational alcohol has also been observed for cyt-c in silica gels.8 changes occur in the protein as the polypeptide subunits are protonated or deprotonated. At pH>4, the solution spectrum shows its Soret band maximum centered at 406 nm which Sol–gel encapsulated manganese myoglobin: shifts to ca. 395 nm at pH 1. The optical absorption changes detector for nitric oxide indicate a titration of the amino acid groups with the protons.In solution, two transitions centered around pH#0 and pH#3 As discussed previously, native Mb and Hb with Fe in the can be observed, as shown in Fig. 6. Within the limits of heme can bind O2, CO, and NO. Manganese myoglobin experimental error, these values correspond well with the (MnMb), however, is an excellent alternative to native Mb reported literature.On the other hand, similar experiments on for NO detection because the manganese-substituted proteins cyt-c encapsulated in aged gels indicate a more continuous bind NO but not O2.26 Moreover, NO-binding in MnMb can variation in the pH profile, also shown in Fig. 6. A large also be detected optically by using the a peak of MnMbNO change in Soret peak wavelength is observed in the pH range at 580 nm as the optical marker. 3 to 5 with a midpoint at pH#4. At low pH, the sharp Nitric oxide chemistry has been in the forefront of research transition observed in solution is no longer evident in the aged recently because of its importance in physiological activities gel samples. Instead the conformational change accompanying ranging from maintaining vascular tone to antimicrobial pH variation is spread out over a pH range of 3 to 0.defense.27 NO is synthesized in cells via the oxidation of LThe observed diVerences in the proton dependent behavior arginine to L-citrulline by the enzyme NO synthase (NOS), as of the cyt-c in solution and in aged gels show that the presence shown in eqn.(1). of the silica matrix may have altered the acid–base properties of the protein. As changes in pH induce conformational L-arginine CCCCCCCDA O2, NADPH, NOSynthase L-citrulline+NO (1) transitions in cyt-c, it is possible that the constraints imposed by encapsulation within a pore of finite dimension restrict NO is a free radical and the biological lifetime of NO is of those transitions.Another possibility is that the charged the order of seconds. A variety of sensors have been developed residues on cyt-c react diVerently with the pore walls of the to measure NO levels. A porphyrinic-based microsensor has been developed to detect NO electrochemically,28 and optical methods have also been developed based on bacteria denitrifi- cation,29 conversion of HbO2 to metHb,30 chemiluminescence of ozone treated NO,31 luminol chemiluminescence,32 guanylate cyclase activation,33 and other NO-mediated eVects.More recently Aylott et al.34 demonstrated that cyt-c can also be used for optical NO sensing. By using sol–gel thin films with encapsulated cyt-c, a reversible NO sensor was developed that can measure NO at ppm levels within minutes.Barker et al. developed a reversible fiber-optic NO sensor based on cyt-c¾ that can also measure NO at ppm levels with <1 s response time. In both cases, there was no interference from O21 and with cyt-c¾, interference from NO2 can also be eliminated. An ideal optical sensor for NO is one which specifically binds NO with high aYnity, producing a measurable change in the absorption spectrum. The ability to conduct the experiments aerobically is also an important factor, since working in an O2-free environment is physiologically unreasonable. To evaluate the sensitivity of MnMb for NO sensing under Fig. 6 Optical monitoring of the Soret band wavelength as a function simulated physiological conditions, we conducted an experi- of pH shows a diVerence in the proton-dependent behavior of cyt-c in aged silica gels as compared to aqueous buVer.ment in which NO was generated enzymatically from the 50 J. Mater. Chem., 1999, 9, 45–53Table 4 Reaction of nitrogen oxides with manganese myoglobin enzyme nitric oxide synthase (NOS). In this experiment, MnMb was in solution rather than in a silica gel so that NO Nitrogen oxide Experimental conditions NO-Binding transport would not be impeded by the silica matrix.The experiment was conducted in the presence of air. After mixing NO (gas) [NO]=0.8 mM Yes an aqueous solution of the NOS enzyme, L-arginine, the NO2 (gas) [NO2]=3.7 mM Yes NO2- (solid NaNO2) [NO2 -]<10 mM No required cofactors and MnMb, optical absorption was [NO2-]=45 mM Yes recorded periodically between 500 and 600 nm to detect gener- NO- (solid Na2N2O3) [NO-]=5.5 mM Yes ation of NO as evidenced by the a peak of MnMbNO NO3- (solid NaNO3) [NO3 -]=45 mM No at 580 nm.Fig. 7 shows the optical absorption of the NOS–MnMb solution as a function of time. The data clearly demonstrate that enzymatically generated NO can be detected optically by directly monitoring the NO adduct.The level of NO generated in the NOS–MnMb experiment was estimated by measuring the enzymatic activity, i.e. nmole of citrulline generated per mg of enzyme. As shown in eqn. (1), citrulline and NO are formed in equimolar amounts. The measured levels of citrulline generated per mg of NOS after 30, 60, 90, and 120 min were 161, 239, 212, and 256 nmol, respectively. In the experiment reported here ca. 0.12 mg of enzyme was used. Although optical detection of NO was reported previously using HbO2,30 MnMb represents a better choice of protein because one can measure the NO-adduct directly and also because HbO2 will eventually oxidize to metHb without NO in the presence of air. To determine the specificity of MnMb for NO sensing, we carried out control experiments in which MnMb in solution was reacted with other nitrogen oxides, specifically NO-, NO2, NO2-, and NO3-.All experiments were performed with MnMb in aqueous buVer. We found that MnMb binds NO- and NO2 but not NO2- or NO3-. NO- was generated indirectly via the decomposition of Na2N2O3, which releases HNO (NO-) in aqueous solution36 as shown in eqn. (2); Na2N2O3=HNO+NO2- (2) NO2- does not bind MnMb.However, it produces NO at suYciently high concentrations according to eqn. (3). 3HNO2=HNO3+2NO+H2O (3) Table 4 lists the diVerent nitrogen oxide species tested and their ability or inability to bind MnMb. Fig. 8 Successful NO detection was achieved in silica-encapsulated We successfully encapsulated MnMb in optically transparent MnMb, as the same characteristic absorption changes were observed sol–gel derived silica gels and used these protein-doped mate- upon metal reduction and NO binding as compared to MnMb in rials as NO detectors.As mentioned previously, a significant aqueous buVer. The slight reduction in absorption intensity for the aged gel was most likely caused by the decomposition of the reducing advantage of using MnMb to detect NO is that experiments agent (sodium dithionite).The gel was immersed in a dithionite can be performed aerobically. Fig. 8 shows the absorption solution which became cloudy over time in air. spectra of Mn(III)Mb, Mn(II)Mb, and MnMbNO in solution and in an aged silica gel. Optical absorption characteristics of MnMb and its NO adduct are detailed in Table 5. A comparison of the spectra shows essentially the same spectroscopic properties upon metal reduction and NO binding in the aged gel as in solution.In addition to binding dissolved gaseous NO, silica-encapsulated MnMb can also bind NO released from a chemical. SNitroso- N-acetylpenicillamine (SNAP) releases NO in an aqueous medium.37 Fig. 9 shows the absorption spectra of MnMb aged gels reacted with SNAP, and NO binding is evident.There was a predictable concentration eVect as the 1.5 mM SNAP concentration required a longer incubation Table 5 Optical absorption characteristics of manganese myoglobin Soret/nm b/nm a/nm Description Soln. Sol–gel Soln. Sol–gel Soln. Sol–gel Mn(III )Mb 373 371 Fig. 7 Optical monitoring of a solution containing NO synthase and 468 468 MnMb at varying incubation times demonstrates that NO generated Mn(II )Mb 438 438 by NO synthase can be detected.The a peak at 580 nm of MnMbNO Mn(II )MbNO 424 424 538 538 580 579 was used as the optical marker for NO binding. J. Mater. Chem., 1999, 9, 45–53 51Fig. 9 Optical monitoring of MnMb in silica gels reacted with NO released from S-nitroso-N-acetylpenicillamine (SNAP) shows NO binding.As expected, there was a higher level of the NO adduct with a higher SNAP concentration. time (>120 min) for complete NO binding compared with the The authors thank Dr Yumiko Komori for preparation of the NOS enzyme and SNAP, Naiyma Houston and James Hauser 5.3 mM SNAP concentration (60 min). These experiments demonstrate the potential for using for their experimental assistance, and Dr.Daryl Eggers for his critical reading of the manuscript. The support of this research encapsulated heme proteins in sol–gel matrices as sensing elements. A simple yet eVective optically-based technique to by the National Science Foundation (DMR-9408780) is greatly appreciated. detect NO was developed using sol–gel encapsulated MnMb. The Mn-substituted heme makes aerobic experiments possible, which is critical for physiological conditions.The advantages of using heme proteins [cyt-c, cyt-c¾, MnMb] include direct References detection of NO via formation of an NO-adduct and the 1 L. M. Ellerby, C. R. Nishida, F. Nishida, S. A. Yamanaka, ability to detect NO without interference from O2 or air. One B. Dunn, J. S. Valentine and J. I. Zink, Science, 1992, 255, 1113.disadvantage of MnMb is that it is not specific for the NO 2 D. Avnir, S. Braun and M. Ottolenghi, in Supramolecular free radical but will also react with NO- and NO2 to form Architecture in Two and Three Dimensions, ed. T. Bein, American MnMbNO. In sensor applications, the sol–gel encapsulated Chemical Society, New York, 1992. MnMb should be in thin-film form to maximize kinetics of 3 B.C. Dave, B. Dunn, J. S. Valentine and J. I. Zink, Anal. Chem., analyte diVusion through the pores. 1994, 66, 1120A. 4 D. Avnir, S. Braun, O. Lev and M. Ottolenghi, Chem. Mater., 1994, 6, 1605. Conclusions 5 S. Braun, S. Rappoport, R. Zusman, D. Avnir and M. Ottolenghi, Mater. Lett., 1990, 10, 1. Heme proteins can be successfully encapsulated in silica 6 E.Antonini and M. Brunori, Hemoglobin and Myoglobin in their matrices with minimal changes to their spectroscopic proper- Reactions with Ligands, North-Holland Publishing, Amsterdam, ties. Research by various groups has shown that the immobil- 1971. ized heme proteins also retain their chemical functions of 7 T. Yonetani and T. Asakura, J. Biol. Chem., 1969, 244, 4580. 8 B. C. Dave, J. M. Miller, B. Dunn, J. S. Valentine and J. I. Zink, oxidation/reduction, ligand (O2, CO, NO) binding, and biocat- J. Sol–Gel Sci. Technol., 1997, 8, 629. alysis. The sol–gel encapsulated proteins can be used to 9 C. Shen and N. M. Kostic, J. Am. Chem. Soc., 1997, 119, 1304. quantitatively determine concentration of dissolved gases as 10 C. T. Lin, C. M. Catuara, J. E. Erman, K. C. Chen, S. F. Huang, dissolved O2 concentrations could be measured within minutes. W. J. Wang and H. H. Wei, J. Sol–Gel Sci. Technol., 1996, 7, 19. Moreover, immobilization using the sol–gel approach has 11 S. Wu, J. Lin and S. I. Chan, Appl. Biochem. Biotechnol., 1994, enabled new features such as room temperature electron 47, 11. 12 N. Shibayama and S. Saigo, J. Mol. Biol., 1995, 251, 203. paramagnetic resonance, formation of a MetMb–CO2 adduct, 13 E. H. Lan, M. S. Davidson, L. M. Ellerby, B. Dunn, J. S. Valentine and fixation of Hb’s quaternary structure. The conclusions and J. I. Zink, Mater. Res. Soc. Symp. Proc., 1994, 330, 289. from our new studies include the finding that encapsulation 14 D. J. Blyth, J. W. Aylott, D. J. Richardson and D. A. Russell, in a porous silica matrix can significantly improve the thermal Analyst, 1995, 120, 2725. stability of a protein. 15 Q. Ji, C. R. Lloyd, W. R. Ellis and E. M. Eyring, J. Am. Chem. Thermal denaturation studies on cyt-c demonstrated that Soc., 1998, 120, 221. 16 K. E. Chung, E. H. Lan, M. S. Davidson, B. Dunn, J. S. Valentine whereas the protein in aqueous buVer showed a transition to and J. I. Zink, Anal. Chem., 1995, 67, 1505. the denatured state with Tm#65 °C, the sol–gel immobilized 17 M. F. McCurley, G. J. Bayer and S. A. Glazier, Sens. Actuators B, protein did not show any transition up to ca. 95°C. A similar 1996, 36, 491. improvement in protein stability to alcohol has also been 18 R. A. Scott and A. G. Mauk (Editors), Cytochrome c: A achieved using sol–gel encapsulation. Another key finding from Multidisciplinary Approach, University Science Books, Sausalito, our studies was that sol–gel encapsulated MnMb can serve as CA, 1996. 19 G. R. Moore and G. W. Pettigrew, Cytochromes c: Evolutionary, solid-state detectors or sensors for NO. Our experiments estab- Structural, and Physicochemical Aspects, Springer-Verlag, Berlin, lished that sol–gel immobilized MnMb binds NO without 1990. interference from O2, similar to MnMb in solution. Finally, an 20 S. Hu, I. K. Morris, J. P. Singh, K. M. Smith and T. G. Spiro, intriguing hypothesis based on collective results from optical J. Am. Chem. Soc., 1993, 115, 12446. absorption, resonance Raman, and thermal denaturation studies 21 T. Uno, Y. Nishimura and M. Tsuboi, Biochemistry, 1984, 23, on cyt-c is that this protein designs self-specific pores in the 6802. 22 D. S. Cohen and G. J. Pielak, Protein Sci., 1994, 3, 1253. silica network according to its size and shape requirements. 52 J. Mater. Chem., 1999, 9, 45–5323 T. Kitagawa, Y. Ozaki, J. Teraoka, Y. Kyogoku and 32 K. Kikuchi, T. Nagano, H. Hayakawa, Y. Hirata and M. Hirobe, J. Biol. Chem., 1993, 268, 23106. T. Yamanaka, Biochim. Biophys. Acta, 1977, 494, 100. 24 Y. P. Myer, R. B. Srivasatava, S. Kumar and K. Raghvendra, 33 K. Ishii, H. Sheng, T. D. Warner, U. Forstermann and F. Murad, Am. J. Physiol., 1991, 261, H598. J. Protein Chem., 1983, 2, 13. 25 B. Cartling, Biological Applications of Raman Spectroscopy, ed. 34 J. W. Aylott, D. S. Richardson and D. A. Russell, Chem. Mater., 1997, 9, 2261. T. G. Spiro, Wiley, New York, Vol. III, 1988. 26 C. Bull, R. G. Fisher and B. M. HoVman, Biochem. Biophys. Res. 35 S. L. R. Barker, R. Kopelman, T. E. Meyer and M. A. Cusanavich, Anal. Chem., 1998, 70, 971. Commun., 1974, 59, 140. 27 E. Culotta and D. E. Koshland, Science, 1992, 258, 1862. 36 F. T. Bonner and M. N. Hughes, Commun Inorg. Chem., 1988, 7, 215. 28 T. Malinski and Z. Taha, Nature (London), 1992, 358, 676. 29 J. Goretski and T. C. Hollocher, J. Biol. Chem., 1988, 263, 2316. 37 L. J. Ignarro, H. Lippton, J. C. Edwards, W. H. Baricos, A. L. Hyman, P. J. Kadowitz and C. A. Gruetter, J. Pharmacol. 30 M. P. Doyle and J. W. Hoekstra, J. Inorg. Biochem., 1981, 14, 351. 31 O. C. Zafiriou and M. McFarland, Anal. Chem., 1980, 52, 1662. Exp. Ther., 1981, 218, 739. Paper 8/05541F J. Mater. Chem., 1999, 9, 45–53 53

ISSN:0959-9428    

DOI:10.1039/a805541f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Design of nanosized structures in sol–gel derived porous solids. Applications in catalyst and inorganic membrane preparation† Christian G. Guizard,* Anne C. Julbe and Andre� Ayral Laboratoire des Mate�riaux et Proce�de�s Membranaires-CNRS UMR 5635, Ecole Nationale Supe�rieure de Chimie-8, rue Ecole Normale, 34296 Montpellier Cedex 5, France.E-mail: guizard@cit.enscm.fr Received 6th April 1998, Accepted 27th July 1998 Recent advances in sol–gel processing, with the aim of branes mainly focus on liquid separations. In the meantime porous ceramic oxide preparation and applications in the market for gas separation and related applications is catalysis and membrane separation, are reviewed. Recent considered to be potentially very important.Regarding future results from our group in porous solid preparation are developments, the catalytic membrane reactor concept is cerpresented. Three topics of particular interest are reported tainly the most exciting and promising idea in the area of which highlight the potential of sol–gel methods to tailor inorganic membranes. Combining reaction and separation in microporous and mesoporous structures in solids, and the same unit often creates a synergy and many reports also the possibility to combine catalyst and ceramic mem- agree that acceptance of catalytic membrane reactors on a brane properties.The first one deals with the preparation commercial scale will arise within the next ten years. and the characterisation of supported microporous layers Most of the above related applications concern porous (e.g.with pore diameter of less than 2 nm) and the appliceramic oxide materials. Today the demand in catalysis and cations linked to ceramic nanofilters. The second topic separation applications is for selective materials able to per- illustrates improvements expected in sol–gel processing of form reaction and separation of a wide range of salts, organic oxide catalysts and the possibility of forming supported molecules, gases and vapours.The eYciency of these catalytic catalytic layers and membranes. The last one has to do and separation processes strongly depends on the pore size of with the role of surface active agents in the control of the the used materials (catalysts, adsorbents or membranes) in sol-to-gel transition and in the formation of tailor-made relation to the size of the processed molecules and ions.porous structures in oxide materials. Consistently the control of the porous structure (pore size, pore volume and surface area) is of prime importance in the 1.0 Introduction development of more specific materials. The thermal stability of such materials is also a crucial parameter for the applications Mesoporous and microporous solids are increasingly being listed above.Most of the catalysts, adsorbent or inorganic employed in a number of industrially important areas which membranes can be described as ceramic nanophase materials include catalysis, adsorption processes and membrane separain which ultrafine grain sizes and a large fraction of the tion. Recently Davis and Maxwell edited a review on solid interfacial volume are responsible for a variety of novel and catalysts and porous solids1 in which was pointed out the potentially advantageous properties and characteristics com- importance of catalytic technologies for both economic growth pared to those of conventional materials.This new class of and environmental sustainability.Heterogeneous catalysis materials results from the emphasis of some new ceramic appears as a major concern in the development of new catalyst synthesis methods like the sol–gel process. systems for oil refining, chemical synthesis, natural gas conver- The sol–gel process presents inherent advantages for the sion and environmental technology.Adapted catalysts are based on mesoporous and microporous solids with or without preparation of porous ceramic oxides, because the nanostrucadded metals. The important research and development ture of the derived materials can be controlled together with activity on microporous and mesoporous materials for molecu- their porous structure (e.g. formation of mesopores or microlar sieve and adsorbent applications should also be emphasised. pores).Other quoted advantages of sol–gel processing are The discovery of mesoporous molecular sieves is certainly the compositional homogeneity and the ability to prepare shaped most remarkable advance in this field. These materials exhibit materials such as spherical particles, fibres and thin films. structured inorganic frameworks with pores large enough Depending on the method used, colloidal or polymeric, two (2–10 nm) to be used in a wide range of application including main gel structures can be obtained which are described in the shape-selective catalysis, sorption of large organic molecules, literature4 as: (i) physical gels in which steric or electrolytic chromatographic separations, and use as hosts to confine guest eVects in the sol dominate gel formation. The main charactermolecules and atomic arrays.The mesoporous structure of istic of this type of gel is the way in which individual particles gels derived from colloidal sols of oxide and clay particles has contained in the sol can be arranged during the process. These also been described in the literature as a basis for preparing gels are rather concerned with aqueous media; and (ii) polyadsorbents with controlled and uniform pore geometry.2 meric gels in which the relative rates and extents of chemical Another important concern in the field of porous materials reactions lead to cluster polymerization and interpenetration is the notable development of inorganic membranes over the during gel formation.In this case organic media are preferred. past decade. A recent book edited by Burggraaf and Cot deals The scope of this paper is to review innovative developments with fundamentals of inorganic membrane science and technol- in sol–gel processing of ceramic oxide nanophase materials in ogy.3 The current industrial applications of inorganic mem- relation to the preparation of both catalysts and membranes.Three topics of particular interest are reported which highlight the potential of sol–gel methods to tailor microporous and †Basis of the presentation given at Materials Chemistry Discussion mesoporous structures in solids, and also the possibility to No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. combine catalyst and ceramic membrane properties. The first J.Mater. Chem., 1999, 9, 55–65 55one deals with the preparation and the characterisation of supported microporous layers (e.g. with pore diameter of less than 2 nm) and the applications linked to ceramic nanofilters. The second topic illustrates improvements expected in sol–gel processing of oxide catalysts and the possibility of forming supported catalytic layers and membranes.The last one has to do with the role of surface active agents in the control of sol-to-gel transition and in the formation of tailor-made porous structures in oxide materials. 2.0 Microporous materials and membranes formed from nanoparticulate sol–gel systems 2.1 General aspects of microporous membrane preparation and characterisation The sol–gel process is one of the most appropriate methods for the preparation of microporous oxide layers.Two sol–gel routes, polymeric or colloidal, can be used to prepare supported ceramic membranes in which the porous structure is mainly influenced by the diVerent steps involved in the process Fig. 1 Influence of cluster structures in polymeric sols on the porosity even at the very first stage of precursor chemistry. The first of coated layers: (a) packing of interpenetrated low branched clusters, stage in the sol–gel process consists of the preparation of a (b) packing of non-interpenetrated highly branched clusters.sol using molecular precursors, either metal organics (preferentially metal alkoxides) or metal salts. In both cases condensation reactions occur at the sol stageation of clusters or colloids which collide at the final stage to form the gel.In the case of membrane formation, coating of the active layer has to be carried out at the sol stage with rheological behavior adapted to the porous substrate chosen as membrane support. Actually, inorganic membranes are formed following a three-step thermal treatment of freshly cast gel layers.At first, the supported gel layer is dried at low temperature (<100 °C). Then the dried layer is fired to an intermediate temperature (ca. 350 °C) in order for residual organic groups and carbon to be burned out. Finally the consolidation of the membrane is performed through viscous or conventional sintering depending on the amorphous or crystalline structure of the membrane material.During sol–gel processing of inorganic membranes, sols and gels evolve in a diVerent way depending Fig. 2 Packing density of colloidal particles in solids obtained from on the category of the precursors used. This evolution has a physical gels. great influence on the porous structure of the final membrane materials. On the one hand, polymeric gels formed in organic media formed from colloidal particles in aqueous media yield membrane materials with 30–35% porosity (Fig. 2) and with pore from interpenetrated clusters lead to microporous membrane materials5,6 as shown in Fig. 1. The problem with this category sizes in the intermediate range between mesopores and micropores. A specific problem arises in the case of microporous of membrane is that microporous volumes remain very low; this is a major limitation for liquid filtration and particularly ceramic membranes.Indeed, two main conditions exist in order to prepare nanophase ceramics exhibiting a connected for nanofiltration application of these materials. On the other hand, in particulate sol–gel systems, the corresponding sols microporosity with a narrow pore size distribution.The first one is to prevent particle aggregation at the sol stage respon- contain individual particles surrounded either by a steric barrier or an electrical double layer responsible for interparticle sible for the formation of large and polydispersed sintered grains yielding a residual mesoporosity, the second one is to repulsion and sol stability.7 A strong steric eVect or a high electrostatic repulsion barrier between particles at the sol stage preserve individual grains of less than 10 nm all along the process up to the sintered ceramic.This is possible by using a provides a dense packed bed of particles during gel formation close to a perfect arrangement of spheres (green density d= new preparation route which consists of the formation of nanoparticulate stable sols in organic solvents.In this case the 0.74 dth). In this case materials with a rather low porosity (30%) and capable of readily densifying at high temperature particles, which are less than 10 nm in size, consist of a mineral core surrounded by an organic shell which prevents aggre- are generally obtained. For weaker interactions, a partial aggregation of particles can occur in the sol and consequently gation (Fig. 3) and allows the formation of a connected microporous structure with porosity in the 20–30% range. An a high porous volume is obtained in the gel as well as in the final material. These two eVects, which are shown in Fig. 2, example dealing with the preparation of a microporous zirconia layer according to this method is given hereafter.can be exploited for the preparation of ceramic membranes with a controlled microporous or mesoporous structure.7 The characterisation of the porous structure of supported thin films and membranes is a complex problem, since total Thermal treatment is another parameter which can aVect the porous structure of the final membrane material.In the case porosity, pore size and pore size distribution must be analysed. A quite diYcult problem is the characterisation of microporous of materials obtained from particulate sols, a grain growth phenomenon is generally found to be responsible for an materials8–10 for which thermodynamic laws are no longer valid because of very strong interactions between intrusive increase of pore size corresponding to the evolution of residual voids between the sintered particles.The classical physical gels fluids and pore walls. The methods based on gas adsorption 56 J. Mater. Chem., 1999, 9, 55–65ture in supported ceramic films. Such materials oVer attractive properties when used as ceramic nanofilters.15 Ceramic nano- filters are being recognised as of growing importance owing to increasing demands for membrane systems which are able to separate ions and small molecules in harsh working conditions (high temperature, extreme pH, organic solvent media).Regarding the problem of aggregation, a solution has been to generate steric repulsion in organic media through the interaction of organic shells preventing close contact between nanoparticles in the sol, Fig. 3. Taking into account the role of chelating agents as blocking functional groups in metal alkoxide condensation16 a supported microporous zirconia layer has been obtained from a nanoparticulate organic sol. Zirconium isopropoxide was used as the ceramic precursor to synthesise the organic particulate sol.17 In order to avoid the precipitation of non-homogeneous hydroxide particles during the hydrolysis step, the alkoxide reactivity was modified by a strong complexing ligand, acetylacetonatee (acac).Hacac reacts with the alkoxides to form mixed complexes which have diVerent physico-chemical properties and, more importantly, which are more diYcult to hydrolyse than alkoxy groups:18 Zr(OR)4+Hacac�Zr(OR)3(acac)+ROH Then this ligand acts as a functionality blocker when substoichiometric hydrolysis ratios are used.Consequently, a Hacac/M molar ratio greater than l was used to prevent precipitation and led to a stable nanoparticulate sol prepared in air without any precipitation. Here the stability of nanoparticles in the sol results from the interacting organic shells formed with the acac groups as shown in Fig. 4.The supported zirconia layer obtained after sintering at Fig. 3 Interaction and corresponding potential (V) between steric 500 °C exhibited a tetragonal crystalline structure and revealed barriers due to the presence of organic shells around ceramic particles. a very fine texture when observed by TEM. In the case of a Zr/Hacac molar ratio=2, a mean grain size of about 6 nm are the most convenient for the study of microporous materials.was measured. Powder X-ray diVraction (Scherrer’s formula) The conventional apparatus uses volumetric measurements of was used for the determination of an average individual crystal the adsorbed gas quantities.11 In the case of supported thin size. Results are consistent with particle size observed by TEM films, the amount of porous layer which can be introduced in in the supported membrane, Fig. 5. In this membrane packing a standard-size sample cell is often too small to be correctly of the nanoparticles results in a pore diameter of 1.4 nm and characterised. Other types of adsorption equipment were therea porosity of 18%. fore developed, like the surface acoustic wave device12 and a device based on the measurement of the refractive index by ellipsometry.13 However these techniques require specific conditions concerning the nature of the substrate.In the first case, the substrate must be piezoelectric, while in the second case, the layer must be transparent and the refractive index of the dense substrate must be quite diVerent from that of the layer. In other respects, each substrate exhibits specific surface properties which can influence the final characteristics of the deposited coatings.In a recent work we combined adsorption measurements and ellipsometry for the characterisation of supported microporous silica layers.14 Both silica layers and the corresponding bulk silica materials were prepared by the sol–gel polymeric route using tetraethoxysilane as the starting material. This work led to the conclusion that during the drying step the support has a great influence on theorous structure formation of the coated layer. The support prevents shrinkage of the thin layer resulting in a higher porosity compared to the bulk material.Further modification or adaptation of these techniques must be completed in order to better characterise the porous structure of supported membranes, particularly microporous membranes. 2.2 Preparation of a zirconia nanofilter from a nanoparticulate Fig. 4 Cluster formed from condensation of acetylacetone (Hacac) organic sol modified zirconium isopropoxide; acac groups are at the periphery of Starting from the above considerations it is possible, from a the cluster as an organic shell and prevent interpenetration of clusters which behave as individual nanoparticles.nanoparticulate sol, to create a connected microporous struc- J. Mater. Chem., 1999, 9, 55–65 57of ammonium metavanadate.20,21 The synthesis of original metallo-organic molecular precursors has been attempted in order to obtain a homogeneous liquid, adapted to casting and being a precursor of a pure VMgO catalyst, with a controlled V/Mg ratio, a high specific surface area and a high catalytic activity (dispersion of active sites).A functional magnesium alkoxide Mg(OCHCH3CH2OCH3)2 has been synthesised by direct reaction between magnesium and 1-methoxypropan-2- ol22 and reacted with a new original vanadium oxoalkoxide VO(OCHCH3CH2OCH3)3 prepared by reaction of ammonium metavanadate with 1-methoxypropan-2-ol. The reaction leads to the formation of (OR)n-1VOMg(OR)m-1 species.This heterometallic compound can be easily cast on a support, hydrolysed, dried in vacuum at room temperature and finally fired at 600 °C in air to lead to the final VMgO catalyst with a controlled V/Mg ratio. Apart from the fact that this specific method is well adapted to catalyst casting on a support (e.g.for catalytic membrane reactor applications), the derived VMgO powders have been found to exhibit original specific characteristics when compared to conventional VMgO catalysts derived from precipitated Mg(OH)2. These characteristics concern both the textural properties of the catalysts (high specific surface area: 110 m2 g-1 at 14 wt% V, round particles of about 25 nm instead of platelets for conventional catalysts) and also their structural characteristics (delayed crystallisation of Mg3V2O8, high dispersion of amorphous vanadium species on the surface of MgO particles).Fig. 6 compares the morphology of the VMgO powders (14 wt% V) obtained at 600 °C from the Mg(OH)2 salt precursor21 and from the above-mentioned alkoxides.The specific characteristics of the alkoxide derived catalyst, attributed to a high homogeneity of V and Mg species distribution in the molecular Fig. 5 Electron microscopy images of the active zirconia layer of a precursor, induce attractive catalytic properties for the oxyceramic nanofilter prepared at 500 °C from a sol of Hacac modified dehydrogenation of propane. Some preliminary results have zirconium alkoxide precursor.(a) SEM cross-section image of the supported zirconia layer, (b) TEM image of the zirconia nanoparticles constituting the zirconia layer. 3.0 Catalytically active nanophase materials from the sol–gel process The sol–gel process is also of particular interest for designing catalytically active materials with specific properties and/or with improved performances compared to those of classical catalysts. It allows preparation of catalytically active materials which can be directly cast in/on supports at the sol stage.This is a great advantage for catalytic membrane development.19 The classical synthesis methods for conventional catalysts often start from salts or oxide precursors and involve precipitation, impregnation or even solid/solid reactions.These methods are usually not adapted to a homogeneous casting of the catalyst on a support and may lead to limited specific surface areas and to a heterogeneous distribution of active species. The sol–gel process, starting from a homogeneous distribution of the precursors at the molecular level, can in many cases improve these specific criteria. Indeed a large range of methods and precursors can be investigated to obtain the required powders or film-shaped materials either as a pure catalyst or homogeneously doped or dispersed in a matrix.Furthermore, the specificity of the process can lead to original materials which can help in a better understanding of the catalytically active sites in a specific catalyst.Two examples showing the potential of the sol–gel process for catalyst synthesis are described in the following sections. 3.1 VMgO catalysts prepared from original metallo-organic precursors The method classically used to prepare conventional VMgO Fig. 6 FESEM observation of VMgO (14 wt.% V) catalytic powders catalysts for the oxydehydrogenation of propane involves the obtained at 600 °C from: (a) Mg(OH)2 salt precursor (SBET= 90 m2 g-1), (b) novel metal alkoxides (SBET=110 m2 g-1). precipitation of Mg(OH)2 and its impregnation with a solution 58 J.Mater. Chem., 1999, 9, 55–65been reported recently.23 VMgO powders with various V/Mg ratios are currently being investigated and should help in a better understanding of the nature of the catalytic sites active for the oxydehydrogenation of propane.The corresponding optimised mesoporous membranes will be tested as contactors in a catalytic membrane reactor configuration (European Joule program JOE3-CT-95 0022). 3.2 Synthesis of CeO2 based materials for DeNOx applications Most of the three-way catalysts (TWCs) currently used for car exhausts are made with a washcoat (on honeycomb supports) of a mixture of oxides, namely Al2O3 and CeO2, which is impregnated with Pt, Pd or Rh salts.24 The main problems related to classical TWCs lie both in the elevated price of noble metals and in the instability of the catalysts upon use (oxide sintering and metal aggregation).25 The goal was in this case to investigate sol–gel methods and precursors providing the best conditions for the preparation of eYcient CeO2 based catalysts with a high specific surface areas at 1000 °C and from precursors able to be directly cast on supports.A sol–gel preparation method was used, starting from Ce(acac)3·xH2O and Al(OsBu)3 as ceria and alumina precursors, respectively. Hydrolysis and condensation of the precursors in the presence of water was carried out in hexylene glycol at 100 °C for 24 hours with the following weight% composition for the initial sol: precursors 40/hexylene glycol 50/water 10.Gel formation depends on the precursor molar ratio, for Al(OsBu)3/Ce(acac)3·xH2O>0.25 gel formation occurred immediately after water addition. The as-formed gels were heat treated at 1000 °C and the resulting powders analysed by X-ray, FESEM and N2 adsorption–desorption.DiVerent Fig. 8 Morphological aspect of CeO2–Al2O3 based DeNOx catalyst: Ce(acac)3·xH2O/Al(OsBu)3 ratios were investigated. The (a) powders heat-treated at 1000 °C, (b) corresponding honeycomb homogeneous dispersion of CeO2 in the alumina matrix was supported layer. found to prevent fast growth of cerium oxide crystallites with temperature when the CeO2 content in the final material was less than 25 mol%.The evolution of the powder specific surface remained amorphous up to 700 °C, as only the sole CeO2 areas versus CeO2 content and firing temperature is reported diVraction lines were detected. At 1000 °C, some poorly defined in Fig. 7. diVraction lines of c-Al2O3 began to appear. This delayed Specific surface areas higher than 100 m2 g-1 have been crystallisation of alumina was attributed to an eVect of CeO2, obtained at 1000 °C for CeO2–Al2O3 based materials containas the structure of alumina prepared without any additive was ing less than 25 mol% CeO2.When the CeO2 content was too already crystalline at 450 °C. The CeO2 crystallite sizes deter- high (47 mol%) the dispersive eVect of alumina decreased mined from the X-ray diVraction peak widths were smaller and SBET at 1000 °C was lower than 10 m2 g-1.The SEM than 20 nm. observation of the powders with 25 mol% of CeO2 (Fig. 8a) The corresponding sols lead to homogenous layers when showed that the grain sizes were always homogeneous and cast on honeycomb supports, Fig. 8b. High temperature ageing never exceeded 30–40 nm at 1000 °C.Without any alumina in tests are in progress in order to check the structural stability the powder the grain sizes were larger than 200 nm at 1000 °C, of the derived TWC catalysts. The sol–gel route previously and in the case of alumina alone the grain sizes were about described is also currently being investigated for the direct 15 nm at 1000 °C. It should be noted that the alumina matrix insertion of zirconia (forming a solid solution with ceria and enhancing its redox properties) and of Pd catalysts at the sol stage (European Brite EuRam program BRPR-CT 96 0290). 4.0 New methods based on amphiphilic media to tailor porous structures in sol–gel derived solids 4.1 Interest of surface active agents for the control of pore formation Surface active agents represent a specific variety of additives which has been recently investigated in sol–gel processing and not only to limit gel cracking.Indeed, one of the other interests of such additives concerns the control of the hydrolysis step of highly reactive alkoxides, such as titanium alkoxides, by the use of reverse micelle sol–gel systems.26 Another utilisation under investigation in our group is the modification of gel porous structures when surfactant molecules are added after the hydrolysis of an alkoxide.27 As shown in Fig. 9, the Fig. 7 Evolution of the specific surface area in CeO2–Al2O3 powders surfactant molecules introduced in the reaction medium inter- as a function of the CeO2 content and firing temperature. Sol–gel organic route from Ce(acac)3·xH2O. act with growing clusters and their role is in this case the J.Mater. Chem., 1999, 9, 55–65 59sized ‘crystallites’ can be obtained; a firing treatment removes the surfactant aggregates and liberates the ordered porosity. The initial mixture composition controls the structure (lamellar, hexagonal or cubic) of the mesophase. The main material labelled MCM41 exhibits a hexagonal ordered porous texture.The pore sizes are directly controlled by the length of the surfactant alkyl chain and ranges between 1.8 and 10 nm. Monnier et al.32 explained the co-operative process of oxide polymerisation and mesophase formation by the evolution of coulombic type interactions between the surfactant polar head and the growing silicate oligomers. Subsequent studies33–38 showed that the synthesis of a silica network in ordered amphiphilic media could also be obtained by sol–gel routes at Fig. 9 Schematic representation of the role of non-ionic surfactants in room temperature, under atmospheric pressure and with other the limitation of cluster growth and aggregation: (a) gel formation types of surfactant, anionic or non-ionic. In contrast to the with added surfactants, (b) gel formation in the absence of surfactants.M41S type structured mesoporous solids our main objective has been to create, during the sol-to-gel transition, ordered limitation of both interpenetration and condensation between microporous structures in oxide materials using the directing clusters. In addition to the material pore sizes and pore eVect of lyotropic liquid crystal mesophases.33–36 Moreover volume, it is also of particular interest to control the pore these materials have been shaped as continuous supported connectivity which is an important parameter aVecting the thin films in view of membrane applications for liquid and gas material permeability.Nakanishi et al.28 used the spinodal separation.34 We first worked on ordering of porous structures decomposition of sol–gel solutions into two phases, one rich in sol–gel derived silica materials with further investigation of in solvent and the other rich in polyethylene oxide and other metal oxide systems like alumina.Beyond the prepinorganic compounds, to produce dried gels with intercon- aration of the ordered structures an important improvement nected pores in the micrometer range.Unfortunately this in porous membrane characteristics can be expected in terms method cannot be extended to the nanometer range and the of sharp control of pore tortuosity. Tortuosity is one of the templating approach is much more attractive by using remov- important structural parameters which governs membrane able individual templating units or interconnected networks.permeability. As a prospect we relate at the end of this paper At the nanometer scale the lyotropic liquid crystal mesophases some preliminary works on the orientation of ordered domains appear as very attractive systems which can produce removable in silica supported layers with the aim to prepare membranes interconnected networks. Moreover the crystalline structure of with low tortuosity.these templates can direct the growth of the inorganic network and leads to the formation of ordered structure and porosity. 4.2 Control of residual porosity in silica materials derived from The periodicity of the structures induces monodispersed pore polymeric gels sizes and uniform connectivity of both the porous and solid networks.29 Lyotropic liquid crystals are usually lamellar, Based on the same concept as for zirconia nanoparticulate hexagonal and cubic phases (Fig. 10). They are obtained by a organic sols, an interesting way to tailor the microporous self-assembly process of surfactant molecules for intermediate structure in silica materials and membranes is based on compositions in the water–surfactant binary phase diagrams. hindered cluster formation at the sol stage (Fig. 9) using non- The sizes of the constituent units, i.e. the amphiphilic aggre- ionic surfactants. In this approach, non-ionic surfactants, gates, usually range from 2 to 10 nm and are susceptible to alkylaryl polyether alcohols Triton (TX) of diVerent molecular direct the formation of inorganic networks with an ordered weights (X=1–30) were added to tetraethoxysilane sols (molar porosity from the supermicroporous domain (>0.7 nm) up to composition: TEOS/C2H5OH/H2O/HCl=1/4.5/4/0.02). On the mesoporous domain (>2 nm).the basis of sol and gel characterisation (gelation time, 29Si There is currently considerable interest in this new field of NMR, QELS, SAXS, thermoporometry) and material characmaterial science which integrates both a biomimetic approach terisation (FESEM, N2 adsorption, FTIR) the eVect of surfac- (biomineralisation processes) and the concept of nanochemis- tant chain length (X) and TX/TEOS ratio on derived sols, try.In 1992 researchers at Mobil published a synthesis of gels and ceramic materials has been explored.27 In other mesoporous molecular sieves using the templating eVect of respects thermoporometry is a method which is well adapted lyotropic liquid crystal mesophases.30,31 The aluminosilicate to the study of the evolution of mesoporosity in wet gels.We or silicate materials M41S were prepared by hydrothermal have shown,39 using this technique, that the Triton X surfactreatment of solid or molecular precursors of alumina and tant molecules prevent interpenetration and further condensilica in the presence of cationic surfactants of alkyltrimethyl- sation of clusters in the above-mentioned sols.Because surface ammonium halide type. As in the case of zeolites, micrometer active agents are susceptible to interaction with silica oligomers derived from TEOS by van der Waals forces or by interaction with OH groups, the formation of an organic shell made of surfactant molecules between the clusters can be assumed.On the other hand the possibility of micelle formation during drying, by exceeding the surfactant cmc, cannot be set aside but there was no experimental evidence of that. Anyway, in both cases the resulting steric hindrance should limit further condensation of clusters during sol ageing.This is confirmed experimentally by reduced bonding between clusters which explains the formation of more stable sols with longer gelation time and a reversible sol-to-gel transition upon gel shaking. During the heat treatment (450 °C, 2 h) there is formation of Fig. 10 Drawing of three lyotropic liquid crystal structures: (a) lamela homogeneous microporous material due to the elimination lar, (b) bicontinuous cubic (space group Ia3d) viewed along the 100 of the surfactant molecules in contact with the silica gel matrix.direction, (c) hexagonal. J. Charvolin, Prog. Colloid Polym. Sci., 1990, 81, 6 (reproduced with permission). N2 adsorption experiments revealed that N2 (kinetic diam- 60 J. Mater. Chem., 1999, 9, 55–65be of importance for material reactivity: ionic exchange, organic grafting, specific interactions with gas or liquid phases.41 This type of sol prepared in the presence of surfactant is also well adapted for preparing homogeneous dispersions of metals in a silica matrix.An example is reported39 relating the dispersion of a platinum salt (PtCl4) in a TX–TEOS sol. Thermoporometry of the derived gels showed that the eVect of PtCl4 is opposite to the eVect of Triton X and leads to gels with a higher crosslinking degree.The conjunction of the steric hindrance and rigidity eVects of Triton X and PtCl4 respectively produces rigid gels in which the surfactant molecules are Fig. 11 Evolution of specific surface area and microporous volume of trapped. This is an original method to prepare highly microsilica materials versus the molar ratio TX/TEOS. porous SiO2/Pt supported membranes which contain a homogeneous dispersion of catalytically active nanosized Pt eter=3.96 A° ) does not penetrate the porous structure of SiO2 particles.41 materials prepared without a surface active agent.When surfactants were used, type I isotherms were obtained which 4.3 Templates based on a removable interconnected network.are characteristic of microporous materials. The mean Structure directing eVect of lyotropic liquid crystal mesophases hydraulic pore radius RH (estimated by the MP method) can In a previous work, silica gels with ordered microporosity be varied between 3.2 and 7.0 A° . This parameter is slightly were prepared using short length cationic surfactant molecules aVected by the TX/TEOS molar ratio but increases with the which show less ability to form lyotropic mesophases.33,36 The surfactant chain length X.Surfactant addition greatly increases sol compositions (Table 1) were based on a silica precursor, the pore volume and specific surface area of silica materials. tetramethoxysilane, mixed with a series of cationic surfac- As shown in Fig. 11, these two parameters, in the case of X= tants of alkyltrimethylammonium bromide type: 3, reach a maximum (SBET=500 m2 g-1, Micropore volume= CxH2x+1(CH3)3N+Br- (x=8, 10, 12, or 14) and water. In 0.239 cm3 g-1, 35.5% porosity) for TX/TEOS=0.55. Fig. 12 these systems gelation of the silica network and mesophase shows the evolution of the pore volume and specific surface formation occur simultaneously.The X-ray diagrams area as a function of the surfactant chain length X for a ratio (Fig. 13a,c) of two wet gels prepared according to this method TX/TEOS=0.55. The curves seem to reach a plateau for X (for x=8 and x=14) can be assigned to hexagonal and higher than 10. In fact with long chains, larger pore sizes and lamellar phases respectively.The large Bragg spacing d100 can wider distributions were obtained, whereas to obtain small be related to the existence of a liquid crystal structure and its pores with a sharp distribution, a high pore volume and value varies with the length of the chain of the used surfactant specific surface area, X values of 3 to 10 and TX/TEOS ratio molecule. In the investigated surfactant series, Table 2, the wet of 0.55 were preferred.This kind of sol was successfully used to prepare supported microporous silica membranes.27 A Table 1 Composition of the silica sols synthesised in presence of thorough investigation of residual OH groups by FTIR specalkyltrimethylammonium bromides, CxH2x+1(CH3)3N+Br- (with x= troscopy40 showed another attractive feature of this type of 8, 10, 12, or 14) material related to the presence of single OH surface groups at high temperature. These single OH groups (typical sharp Reagent Surfactant Tetramethoxysilane Water IR absorption band at 3744 cm-1) were clearly distinguished Weight % 21 12 67 on FTIR spectra from adjacent OH groups (broad absorption Addition order 1 2 3 bands at 3670 and 3500 cm-1) which completely disappeared upon firing at 600 °C whereas the band at 3744 cm-1 remained.The unusual presence of these groups is specifically due to the presence of surfactant molecules in the sol and to their interaction by hydrogen bonding with the silica based clusters. At the sol and gel stages, these OH groups are ‘protected’ by the surfactant hydrophilic head. This interaction is maintained till the surfactant thermal decomposition, between 200 °C and 300 °C, when single OH groups are liberated. These groups have been shown to be relatively distant from each other because of surfactant steric hindrance and cannot easily condense as in classical sol–gel derived silica.This original surface characteristic of materials derived from TX–TEOS sols should Fig. 13 X-Ray patterns of various wet and thermally treated gels: Fig. 12 Evolution of specific surface area and microporous volume of (a) x=8, wet gel; (b) x=8, gel treated at 450 °C under nitrogen; (c) x=14, wet gel; (d) x=14, gel treated at 450 °C under nitrogen. silica materials versus surfactant chain length. J. Mater. Chem., 1999, 9, 55–65 61Table 2 Textural properties of the thermally treated gels calculated assuming given mesophase structures Gel wp d100/nm dc/nm SM/m2 g-1 SBET/m2 g-1 DS/m2 g-1 D/nm x=8 0.55 2.3 2.1 1074 1260 186 33 x=10 0.55 2.8 2.5 883 1040 157 39 Gel wp d100/nm wp/nm SM/m2 g-1 SBET/m2 g-1 DS/m2 g-1 D/nm x=12 0.58 2.0 1.2 1082 1100 18 366 x=14 0.58 2.2 1.3 1033 1090 57 120 gels exhibit a hexagonal structure for the lower values of x (8 of the surfactant molecule, and have the same order of magnitude as the length of the surfactant molecules, ls, calcu- and 10) whereas for higher values of x (12 and 14) the wet gels are lamellar.After drying and firing at 450 °C under lated using the Tanford relation.45 It has also been noted that the calculated mesophase surface area SM is lower than the nitrogen, a broad diVraction peak is observed in the X-ray patterns (Fig. 13b,d). The decrease of d100 from the wet to the measured value, SBET. This excess of specific surface area, DS, has been assigned to the existence of an interface between the thermally treated gels has been attributed to a shrinkage of the material during the drying step at low temperature and to ordered domains. To a first approximation the ordered domains have been assumed to be spherical with a diameter the condensation of reactive groups of the silica network.Taking into account the sharpness of the diVraction peaks for D calculated from the following equation:36 the wet gels compared to an important broadening of the D=6/[DS(1-wp)rs] diVraction peaks after heat treatment it can be stated that the The skeletal density, rs, of the oxide network measured by size of the ordered domains is reduced during the transition helium pycnometry is equal to that of amorphous silica, i.e.from the wet to the thermally treated gel. Final ordered 2.2 g cm-3. The pore volume fraction in the mesophase struc- domains of about 20 nm in size were determined using the ture wp, is determined from nitrogen adsorption measurements.Scherrer relation.42 The final size of these domains is small The calculated values of D for the hexagonal material have compared to the size of the liquid crystal cells implied in the the same order of magnitude as the sizes determined from the formation of the material. This small size and the probable width of the diVraction peaks using the Scherrer equation. random orientation of the ordered domains could explain the For the lamellar materials, the mean calculated values are limited collapse of the lamellar structures upon heating at larger but in that case D depends strongly on the d100 and 450 °C.wp values. The main remarkable results (Table 2) obtained with these From these results it has been shown that it is possible to materials exhibiting a directed porous structure can be sumdirect, at room temperature, during the sol-to-gel transition, marised as follows.The BET specific surface area, SBET, and an ordered microporous structure in silica gels using lyotropic the total porosity are very large, more than 1000 m2 g-1 and crystal mesophases as reaction media for silicon alkoxides. around 60% respectively.The calculated diameter of the The diVerent experimental values which characterise the cylindrical pores, dc, for the hexagonal gels and the calculated porous texture of the final material can be correlated to width of the slit-like pores, wp, for the lamellar gels are calculated values assuming a polycrystalline material composed consistent with the size of the micropores experimentally of hexagonal or lamellar domains.Remarkably, the micropore determined by the MP43 and Horvath–Kawazoe44 methods. size of the calcined silica gels can be modulated by the length Fig. 14 shows that the average hydraulic diameter, 2rh, and of the alkyl chain of the used surfactant molecules. Starting the average Horvath–Kawazoe diameter, dHK, are almost from the results obtained on ordered porous silica materials, proportional to x, the number of carbons of the alkyl chain the synthesis method was tentatively adapted to the preparation of alumina porous material.In fact aluminium alkoxides are too reactive as precursors of polymeric gels. Usually alumina colloidal gels are obtained from the destabilisation of pseudo-boehmite hydrosols. Kunitake46 showed that a lamellar mesophase produced from a cationic surfactant can be successfully used to direct the aggregation of such colloidal alumina particles.The initial anisotropy of the material is maintained after the departure of the surfactant until the transformation into a-Al2O3 at high temperature. However it can be deduced from the available observations by electron microscopy that the resulting materials are expanded and highly macroporous.We mentioned in a preliminary work47 the possibility of alumina gel synthesis in amphiphilic media using an aluminium salt as the gel precursor. Following this method an aqueous solution of aluminium chloride (0.1 M) containing a quaternary ammonium type cationic surfactant was used to produce an aluminium hydroxide gel.After a pre-neutralisation step with ammonia (molar ratio [NH3]/[Al ]=1), an in situ controlled neutralisation was carried out by introduction of urea (molar ratio [urea]/[Al ]=1) in the starting solution and gentle heating at 80 °C. The thermal decomposition of urea into carbon Fig. 14 Measured and calculated average size of pores versus x, the dioxide and ammonia induced an increase of the pH of the number of carbons in the alkyl chain of the surfactant.(a) 2rh with rh solution and the formation of the alumina gel. This kind of the measured hydraulic radius; (b) 4rh; (c) dHK, measured average controlled neutralisation was previously used for the synthesis Horwath–Kawazoe diameter; (d) ls, calculated length of the surfactant of monodispersed powders48 and gels.49 The main interest of molecule; (e) dc, calculated diameter of the cylindrical pores in the this process is that it can lead to continuous and homogeneous hexagonal structure; (f ) wp, calculated width of the slit-like pores in the lamellar structure.thin layers by deposition of partially neutralised solutions 62 J. Mater. Chem., 1999, 9, 55–65Table 3 Textural properties of the alumina gels after thermal treatment under nitrogen Name Heat treatment/°C SBET/m2 g-1 Pore volume/cm3 g-1 Pore size/nm WS 450 430 0.29 2.7 8C32 450 460 0.33 2.9 14C32 450 485 0.66 5.5 and further ageing of the coatings under controlled field (11.7 T) used recently by Tolbert et al.53,54 to orient the mesophase templated silica networks.atmosphere. The used cationic surfactants were octyl- [C8H17(CH3)3N+Br-] and tetradecyl- [C14H29(CH3)3N+Br-] Consistently our current approach to control the tortuosity in supported porous layers consists of the introduction of trimethylammonium bromides.The added amounts of surfactant were determined from the analysis of the available binary magnetic nanoparticles in the gelling solution. As a matter of fact these particles could have a seeding role in the hetero- diagram: tetradecyltrimethylammonium bromide–water.50 On this diagram the weight percentage of surfactant corresponding geneous nucleation of templating mesophases with an induced eVect on the size of ordered domains.Moreover magnetic to the boundary between the isotropic (lower wt%) and hexagonal mesophase (upper wt%) areas is located at ca. 32 wt%. nanoparticles could be used to orient the mesophases. A hydrosol of maghemite (c-Fe2O3) nanoparticles55 (9.5 nm in In contrast to silica gels, X-ray analysis carried out on a series of wet gels and on the corresponding calcined alumina size, ZPC=7.3, stability pH=2) was used as a ferrofluid source. The stability of silicon alkoxide sols in the presence of gels showed that the ordered porosity formed in the wet gels cannot be maintained upon heat treatment.47 In the case of surfactant molecules and maghemite particles and the possibility to prepare ordered porous structures from these sols have the wet gel 8C32 (x=8), a main diVraction peak at 2.3 nm can be assigned to the presence of a hexagonal mesophase.In been investigated recently.56 Silica gel precursor (TMOS) and surfactant molecules CxH2x+1(CH3)3N+Br- (x=8, 10) were the case of the wet gel 14C32 (x=14), the position d of the main peak (d<3 nm) and the existence of another peak of used as starting materials for gel preparation.A typical weight% composition of gelling solutions is: seeding hydrosol lower intensity located at ca. d/2 are in favour of the presence of a lamellar mesophase.The diVraction patterns of the gels 44, methanol 1, TMOS 8, NH3 (0.1 M) 26, surfactant (x=8) 21. The seeding level in the resulting gels corresponds to 1 after the thermal treatment at 450 °C are consistent with that of a low temperature transition alumina.47 The lyotropic particle per 105 nm3. Thick layers (1 mm>thickness >0.5 mm) were deposited on flat substrates.The texture crystal phase, present in the wet gels, has an influence on the porous texture of the calcined alumina gels (Table 3). The orientation of the thick layers was deduced from the variation of the X-ray diVraction peak at d=2.45 nm associated to the porous texture characteristics of gel 8C32 calcined at 450 and 800 °C are close to those of gel WS (without surfactant) hexagonal mesophase.The peak intensities were measured at two incident angles, a between the incident beam and the calcined at 450 °C. This is a very interesting result owing to the fact that porous alumina materials, obtained by dehy- substrate, b between the projection of the incident beam on the substrate and the shear stress direction of sheared wet gels.dration of well crystallised hydroxides, exhibit a continuous decrease of the pore size and of the surface area as the heat In the absence of a magnetic field during gel formation, seeded or unseeded wet gels exhibit very reproducible variations of treatment temperature is increased. In the case of gel 14C32 the specific surface area at 450 °C is maintained at a high the intensity of the diVraction peak at 2.45 nm versus a and b.In Fig. 15(a) the strong maximum observed for a equal to value but with an almost two-fold increase of pore size and pore volume compared to the reference gel WS. Unfortunately the diVraction angle h is assigned to an alignment of the micellar cylinders and of the associated diVraction planes the peak located at low diVraction angle and associated with an ordered porosity for the 8C32 and 14C32 wet gels is not parallel to the surface of the substrate. The variation versus b, Fig. 15(b), is attributed to an alignment of the micellar observed with the calcined gels. These results have been explained taking into account the low equivalent fraction of cylinders along the shear stress direction. When a magnetic field (0.6 T) is applied during gel layer formation in a direction oxide (Al2O3) in the initial sol composition (ca. 3 wt% for 8C32) and the fact that the produced aluminium hydroxide perpendicular to the surface of the substrate, no more diVraction peaks versus a or b are observed as expected in the case gel is not a strongly cohesive gel as it is for silica. The strengthening of the inorganic network must clearly occur of alignment of the micellar cylinders along the magnetic field direction.Further X-ray experiments on wet gels are under before the departure of the template to favour the direct templating eVect of the ordered amphiphilic mesophase. way in order to confirm these results and to investigate the specific interactions between the seeding particles and the Another aspect of these templated porous textures is related to membrane applications and has to do with the anisotropy lyotropic crystal phases during gel formation.of the ordered domains in the case of supported layers. Interactions at the gel–substrate and air–gel interfaces favour 5.0 Conclusion the preferential orientation of the hexagonal and lamellar structures along these interfaces.Moreover the shear stress The most recent advances in sol–gel processing of porous ceramic oxides have been reviewed and illustrated by recent which is applied during the deposition can induce an alignment of the crystalline structures in the direction of the stress. These results of our group in porous solid preparation. The current state of development of the sol–gel process allows a precise anisotropies are detrimental to the mass transfer performance of the final supported porous materials.In order to improve control of composition, grain structure and pore structure in ceramic oxide materials. These new developments are based their permeability, the size of the ordered domains must be increased and ordered porous structures oriented with the on a better mechanistic control of the reactivity of precursors and derived growing species.The interplay between liquid resulting pores along a direction perpendicular to the layer surface. Liu et al.51 showed that nanoparticle seeding favours crystal organic chemistry and the growth of inorganic frameworks provides a rare opportunity to synthesize a wide range the formation of the templating mesophases in the synthesis medium.In other respects, Fabre et al.52 obtained ferrosmectics of original structures. These systems are potentially tunable for a desired application and oVer an unequalled ability for by doping swollen lamellar phases with ferrofluid nanoparticles. These hybrid mesophases can be oriented under very low the synthesis of mesoporous and microporous materials.Regarding both fundamental research and applications, magnetic field (<1 T) in comparison to the intense magnetic J. Mater. Chem., 1999, 9, 55–65 6313 M. A. Fardad, E. M. Yeatman, E. J. C. Dawnay, M. Green and F. Horowitz, J. Non-Cryst. Solids, 1995, 183, 260. 14 A. Ayral, A. El Mansouri, M. P. Vieira and C. Pilon, J. Mater.Sci. Lett., in press. 15 C. Guizard, C. Mouchet, R. Vacassy, A. Julbe and A. Larbot, J. Sol–Gel Sci. Technol., 1995, 2, 483. 16 C. Sanchez and J. Livage, New J. Chem., 1990, 14, 513. 17 R. Vacassy, C. Guizard, V. Thoraval and L. Cot, J. Membr. Sci., 1997, 132, 109. 18 C. Sanchez, J. Livage, M. Henry and F. Babonneau, J. Non-Cryst. Solids, 1988, 100, 65. 19 A. Julbe, C. Guizard, A.Larbot, L. Cot and A. Giroir-Fendler, J. Membr. Sci., 1993, 77, 137. 20 A. Pantazidis and C. Mirodatos, 11th International Congress on Catalysis, Studies in Surface Sciences and Catalysis, ed. J. W. Hightower, W. N. Delglass, E. Iglesia and A. T. Bell, Elsevier Science B.V., Amsterdam, 1996, 101-B, 1029. 21 A. Pantazidis, A. Auroux, J.-M. Herrmann and C. Mirodatos, Catal. Today, 1996, 32, 81. 22 L. Albaric, N. Hovnanian, A. Julbe, C. Guizard, A. Alvarez- Larena and J. Piniella, Polyhedron, 1997, 16, 587. 23 A. Julbe, L. Albaric, N. Hovnanian, C. Guizard, A. Pantazidis and C. Mirodatos, in Proc. 4th Workshop of the ESF Network: Catalytic Membrane Reactors—Optimisation of Catalytic Membrane Reactor Systems, 1997, ed. R. Bredesen, SINTEF, Oslo, pp. 215–220. 24 P.Mare�cot, L. Pirault, G. Mabilon, M. Prigent and J. Barbier, Appl. Catal. B: Environment, 1994, 5, 57. 25 C. J. Stournaras, A. Tsetsekou, M. Muhammed, C. Guizard, A. Julbe, P. Bekiaroglou, N. Moschoudis, J. P. Joulin, D. Leonidopoulos and M. Debenedetti, in Proc. First European Conference on Clean Cars and First Hellenic Eco Rally, 1997. 26 C. Guizard, A. Larbot, L. Cot, S. Peres and J.Rouvie`re, J. Chim. Phys., 1990, 87, 1901. Fig. 15 X-Ray diVraction analyses on a sheared hexagonal silica layer 27 A. Julbe, C. Balzer, J. M. Barthez, C. Guizard, A. Larbot and (x=8). (a) Evolution of the peak intensity versus the incidence angle, L. Cot, J. Sol–Gel Sci. Technol., 1995, 4, 89. a; (b) evolution of the peak intensity versus b, angle between the 28 K. Nakanishi, Y.Yamasaki, H. Kaji, N. Soga, T. Inoue and direction of the shear stress and that of the incidence angle. N. Nemoto, J. Sol–Gel Sci. Technol., 1994, 2, 227. 29 S. Hansen, Adv. Mater., 1993, 5, 113. 30 C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature, 1992, 359, 710. many exciting developments are expected in the near future 31 J. S. Beck, J. C.Vartuli, W. J. Roth, M. E. Leonowicz, from sol–gel derived nanophase materials, namely for catalytic C. T. Kresge, K. D. Schmitt, C. T. W. Chu, D. H. Olson, E. W. Sheppard, S. B. McCullen, J. B. Higgins and J. L. Schenker, and membrane separation applications. J. Am. Chem. Soc., 1992, 114, 10 834. 32 A. Monnier, F. Schu� th, Q. Huo, D. Kumar, D. Margolese, R. S. Maxwell, G. D. Stucky, M.Krishnamurty, P. PetroV, References A. Firouzi, M. Janicke and B. F. Chmelka, Science, 1993, 261, 1 Solid catalysts and porous solids, ed. M. E. Davis and 1299. I. E. Maxwell, in Curr. Opin. Solid State Mater. Sci., 1996, 1, 54. 33 T. Dabadie, A. Ayral, C. Guizard, L. Cot, J. C. Robert and 2 J. D. F. Ramsay and B. O. Booth, in Fundamentals of Adsorption, O. Poncelet, Mater.Res. Soc. Symp. Proc., 1994, 346, 849. ed. A. B. Mersmann and S. E. Scholl, United Engineering Trustees 34 T. Dabadie, A. Ayral, C. Guizard, L. Cot, J. C. Robert and Inc., 1991. O. Poncelet, in Proceedings of the Third International Conference 3 Fundamentals of Inorganic Membrane Science and Technology, ed. on Inorganic Membranes, July 1994, Worcester, USA, ed. A. J. Burggraaf and L.Cot, Elsevier, Amsterdam, 1996. Y. H. Ma, 3rd ICIM-Worcester Polytechnic Institute, Worcester, 4 Sol–gel Science, ed. C. J. Brinker and G. Scherer, Academic Press, 1994, p. 411. New York, 1990. 35 T. Dabadie, A. Ayral, C. Guizard, L. Cot, C. Lurin, W. Nie and 5 C. J. Brinker, T. L. Ward, R. Sehgal, N. K. Raman, S. L. Hietala, D. Rioult, J. Sol–Gel Sci. Technol., 1995, 4, 107.D. M. Smith, D.-W. Hua and T. J. Headley, J. Membr. Sci., 1993, 36 T. Dabadie, A. Ayral, C. Guizard, L. Cot and P. Lacan, J. Mater. 77, 165. Chem., 1996, 6, 1789. 6 C. Guizard, in Fundamentals of Inorganic Membrane Science and 37 J. E. Martin, M. T. Anderson, J. Odinek and P. Newcomer, Technology, ed. A. J. Burggraaf and L. Cot, Elsevier, Amsterdam, Langmuir, 1997, 13, 4133. 1996, p. 227. 38 M. T. Anderson, J. E. Martin, J. Odinek, P. Newcomer and 7 A. J. Rubin, Mater. Sci. Res., 1984, 17, 45. J. P. Wilcoxon, Microporous Mater., 1997, 10, 13. 8 Characterization of Porous Solids I, Studies in Surface Science and 39 A. Julbe, J. F. Quinson, A. El Mansouri and C. Guizard, in Catalysis Vol. 39, ed. K. K. Unger, J. Rouquerol, K. S. W. Sing Characterisation of Porous Solids IV, ed. B.M. McEnary, and H. Kral, Proc. of the IUPAC Symposium (COPS I), Bad T. J. Mays, J. Rouque�rol, F. Rodriguez-Reinoso, K. S. Sing and Soden, Germany, April 1988, Elsevier, Amsterdam, 1988. K. M. Unger, The Royal Society of Chemistry, Cambridge, 1997, 9 Characterization of Porous Solids II, Studies in Surface Science and pp. 519–526. Catalysis Vol. 62, ed. F. Rodrigez-Reinoso, J.Rouquerol, Piaggio, A. Bottino, G. Capannelli, E. Carosini and A. Julbe, K. S. W. Sing and K. K. Unger, Proc. of the IUPAC Symposium Langmuir, 1995, 11, 3970. (COPS II), Alicante, Spain, May 1990, Elsevier, Amsterdam, 41 C. Balzer, A. Julbe, A. Larbot, C. Guizard, L. Cot, J. Peureux, 1991. A. Giroir-Fendler and J. A. Dalmon, in Proceedings of the Third 10 Characterization of Porous Solids III, Studies in Surface Science International Conference on Inorganic Membranes, July 1994, and Catalysis Vol. 87, ed. J. Rouquerol, F. Rodrigez-Reinoso, Worcester, USA, ed. Y. H. Ma, 3rd ICIM-Worcester Polytechnic K. S. W. Sing and K. K. Unger, Proc. of the IUPAC Symposium Institute, Worcester, 1994, p. 629. (COPS III), Marseille, France, May 1993, Elsevier, Amsterdam, 42 P.Scherrer, Nachr. Go�ttinger Ges. Dtsch., 1918, 2, 98. 1994. 43 R. H. Mikhail, S. Brunauer and E. E. Bodor, J. Colloid Interface 11 S. Lowell and J. E. Shields, in Powder surface area and porosity, Sci., 1968, 26, 45. Chapman and Hall, London, 1984. 44 G. Horvath and K. Kawazoe, J. Chem. Eng. Jpn, 1983, 16, 470. 12 G. C. Frye, A. J. Ricco, S. J. Martin and C. J.Brinker, Mater. 45 C. Tanford, in The Hydrophobic EVect, 2nd edn. Wiley- Interscience, New York, 1980. Res. Soc. Symp. Proc., 1988, 121, 349. 64 J. Mater. Chem., 1999, 9, 55–6546 T. Kunitake, Mol. Cryst. Liq. Cryst., 1994, 240, 1. 52 P. Fabre, C. Casagrande, M. Veyssie, V. Cabuil and R. Massart, Phys. Rev. Lett., 1990, 64, 539. 47 S. Acosta, A. Ayral, C. Guizard and L. Cot, J.Sol–Gel Sci. Technol., 1996, 8, 195. 53 S. H. Tolbert, A. Firouzi, G. D. Stucky and B. F. Chmelka, Science, 1997, 278, 264. 48 J. E. Blendell, H. K. Bowen and R. L. Colbe, Am. Ceram. Soc. Bull., 1984, 63, 797. 54 A. Firouzi, D. J. Schaefer, S. H. Tolbert, G. D. Stucky and B. F. Chmelka, J. Am. Chem. Soc., 1997, 119, 9466. 49 T. E. Wood, A. R. Siedle, J. R. Hill, R. P. Skarjune and C.J. Goodbrake, Mater. Res. Soc. Symp. Proc., 1990, 180, 97. 55 R. Massart, IEEE Trans. Magn., 1991, 17, 1247. 56 M. Klotz, A. Ayral, A. Van der Lee, C. Guizard, C. Menager and 50 T. Warnheim and A. Jonsson, J. Colloid Interface Sci., 1988, 125, 627. V. Cabuil, Mater. Res. Soc. Symp. Proc., to be published. 51 J. Liu, A. Y. Kim, J. W. Virden and B. C. Bunker, Langmuir, 1995, 11, 659.Paper 8/05867I J. Mater. Chem., 1999, 9, 55–65 65 J O U R N A L O F C H E M I S T R Y Materials Feature Article Design of nanosized structures in sol–gel derived porous solids. Applications in catalyst and inorganic membrane preparation† Christian G. Guizard,* Anne C. Julbe and Andre� Ayral Laboratoire des Mate�riaux et Proce�de�s Membranaires-CNRS UMR 5635, Ecole Nationale Supe�rieure de Chimie-8, rue Ecole Normale, 34296 Montpellier Cedex 5, France.E-mail: guizard@cit.enscm.fr Received 6th April 1998, Accepted 27th July 1998 Recent advances in sol–gel processing, with the aim of branes mainly focus on liquid separations. In the meantime porous ceramic oxide preparation and applications in the market for gas separation and related applications is catalysis and membrane separation, are reviewed.Recent considered to be potentially very important. Regarding future results from our group in porous solid preparation are developments, the catalytic membrane reactor concept is cerpresented. Three topics of particular interest are reported tainly the most exciting and promising idea in the area of which highlight the potential of sol–gel methods to tailor inorganic membranes.Combining reaction and separation in microporous and mesoporous structures in solids, and the same unit often creates a synergy and many reports also the possibility to combine catalyst and ceramic mem- agree that acceptance of catalytic membrane reactors on a brane properties. The first one deals with the preparation commercial scale will arise within the next ten years.and the characterisation of supported microporous layers Most of the above related applications concern porous (e.g. with pore diameter of less than 2 nm) and the appliceramic oxide materials. Today the demand in catalysis and cations linked to ceramic nanofilters. The second topic separation applications is for selective materials able to per- illustrates improvements expected in sol–gel processing of form reaction and separation of a wide range of salts, organic oxide catalysts and the possibility of forming supported molecules, gases and vapours.The eYciency of these catalytic catalytic layers and membranes. The last one has to do and separation processes strongly depends on the pore size of with the role of surface active agents in the control of the the used materials (catalysts, adsorbents or membranes) in sol-to-gel transition and in the formation of tailor-made relation to the size of the processed molecules and ions.porous structures in oxide materials. Consistently the control of the porous structure (pore size, pore volume and surface area) is of prime importance in the 1.0 Introduction development of more specific materials.The thermal stability of such materials is also a crucial parameter for the applications Mesoporous and microporous solids are increasingly being listed above. Most of the catalysts, adsorbent or inorganic employed in a number of industrially important areas which membranes can be described as ceramic nanophase materials include catalysis, adsorption processes and membrane separain which ultrafine grain sizes and a large fraction of the tion.Recently Davis and Maxwell edited a review on solid interfacial volume are responsible for a variety of novel and catalysts and porous solids1 in which was pointed out the potentially advantageous properties and characteristics com- importance of catalytic technologies for both economic growth pared to those of conventional materials.This new class of and environmental sustainability. Heterogeneous catalysis materials results from the emphasis of some new ceramic appears as a major concern in the development of new catalyst synthesis methods like the sol–gel process. systems for oil refining, chemical synthesis, natural gas conver- The sol–gel process presents inherent advantages for the sion and environmental technology.Adapted catalysts are based on mesoporous and microporous solids with or without preparation of porous ceramic oxides, because the nanostrucadded metals. The important research and development ture of the derived materials can be controlled together with activity on microporous and mesoporous materials for molecu- their porous structure (e.g.formation of mesopores or microlar sieve and adsorbent applications should also be emphasised. pores). Other quoted advantages of sol–gel processing are The discovery of mesoporous molecular sieves is certainly the compositional homogeneity and the ability to prepare shaped most remarkable advance in this field. These materials exhibit materials such as spherical particles, fibres and thin films. structured inorganic frameworks with pores large enough Depending on the method used, colloidal or polymeric, two (2–10 nm) to be used in a wide range of application including main gel structures can be obtained which are described in the shape-selective catalysis, sorption of large organic molecules, literature4 as: (i) physical gels in which steric or electrolytic chromatographic separations, and use as hosts to confine guest eVects in the sol dominate gel formation.The main charactermolecules and atomic arrays. The mesoporous structure of istic of this type of gel is the way in which individual particles gels derived from colloidal sols of oxide and clay particles has contained in the sol can be arranged during the process.These also been described in the literature as a basis for preparing gels are rather concerned with aqueous media; and (ii) polyadsorbents with controlled and uniform pore geometry.2 meric gels in which the relative rates and extents of chemical Another important concern in the field of porous materials reactions lead to cluster polymerization and interpenetration is the notable development of inorganic membranes over the during gel formation.In this case orgae preferred. past decade. A recent book edited by Burggraaf and Cot deals The scope of this paper is to review innovative developments with fundamentals of inorganic membrane science and technol- in sol–gel processing of ceramic oxide nanophase materials in ogy.3 The current industrial applications of inorganic mem- relation to the preparation of both catalysts and membranes.Three topics of particular interest are reported which highlight the potential of sol–gel methods to tailor microporous and †Basis of the presentation given at Materials Chemistry Discussion mesoporous structures in solids, and also the possibility to No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France.combine catalyst and ceramic membrane properties. The first J. Mater. Chem., 1999, 9, 55–65 55one deals with the preparation and the characterisation of supported microporous layers (e.g. with pore diameter of less than 2 nm) and the applications linked to ceramic nanofilters. The second topic illustrates improvements expected in sol–gel processing of oxide catalysts and the possibility of forming supported catalytic layers and membranes.The last one has to do with the role of surface active agents in the control of sol-to-gel transition and in the formation of tailor-made porous structures in oxide materials. 2.0 Microporous materials and membranes formed from nanoparticulate sol–gel systems 2.1 General aspects of microporous membrane preparation and characterisation The sol–gel process is one of the most appropriate methods for the preparation of microporous oxide layers. Two sol–gel routes, polymeric or colloidal, can be used to prepare supported ceramic membranes in which the porous structure is mainly influenced by the diVerent steps involved in the process Fig. 1 Influence of cluster structures in polymeric sols on the porosity even at the very first stage of precursor chemistry.The first of coated layers: (a) packing of interpenetrated low branched clusters, stage in the sol–gel process consists of the preparation of a (b) packing of non-interpenetrated highly branched clusters. sol using molecular precursors, either metal organics (preferentially metal alkoxides) or metal salts.In both cases condensation reactions occur at the sol stage with formation of clusters or colloids which collide at the final stage to form the gel. In the case of membrane formation, coating of the active layer has to be carried out at the sol stage with rheological behavior adapted to the porous substrate chosen as membrane support. Actually, inorganic membranes are formed following a three-step thermal treatment of freshly cast gel layers.At first, the supported gel layer is dried at low temperature (<100 °C). Then the dried layer is fired to an intermediate temperature (ca. 350 °C) in order for residual organic groups and carbon to be burned out. Finally the consolidation of the membrane is performed through viscous or conventional sintering depending on the amorphous or crystalline structure of the membrane material.During sol–gel processing of inorganic membranes, sols and gels evolve in a diVerent way depending Fig. 2 Packing density of colloidal particles in solids obtained from on the category of the precursors used. This evolution has a physical gels. great influence on the porous structure of the final membrane materials.On the one hand, polymeric gels formed in organic media formed from colloidal particles in aqueous media yield membrane materials with 30–35% porosity (Fig. 2) and with pore from interpenetrated clusters lead to microporous membrane materials5,6 as shown in Fig. 1. The problem with this category sizes in the intermediate range between mesopores and micropores.A specific problem arises in the case of microporous of membrane is that microporous volumes remain very low; this is a major limitation for liquid filtration and particularly ceramic membranes. Indeed, two main conditions exist in order to prepare nanophase ceramics exhibiting a connected for nanofiltration application of these materials. On the other hand, in particulate sol–gel systems, the corresponding sols microporosity with a narrow pore size distribution.The first one is to prevent particle aggregation at the sol stage respon- contain individual particles surrounded either by a steric barrier or an electrical double layer responsible for interparticle sible for the formation of large and polydispersed sintered grains yielding a residual mesoporosity, the second one is to repulsion and sol stability.7 A strong steric eVect or a high electrostatic repulsion barrier between particles at the sol stage preserve individual grains of less than 10 nm all along the process up to the sintered ceramic.This is possible by using a provides a dense packed bed of particles during gel formation close to a perfect arrangement of spheres (green density d= new preparation route which consists of the formation of nanoparticulate stable sols in organic solvents.In this case the 0.74 dth). In this case materials with a rather low porosity (30%) and capable of readily densifying at high temperature particles, which are less than 10 nm in size, consist of a mineral core surrounded by an organic shell which prevents aggre- are generally obtained.For weaker interactions, a partial aggregation of particles can occur in the sol and consequently gation (Fig. 3) and allows the formation of a connected microporous structure with porosity in the 20–30% range. An a high porous volume is obtained in the gel as well as in the final material. These two eVects, which are shown in Fig. 2, example dealing with the preparation of a microporous zirconia layer according to this method is given hereafter. can be exploited for the preparation of ceramic membranes with a controlled microporous or mesoporous structure.7 The characterisation of the porous structure of supported thin films and membranes is a complex problem, since total Thermal treatment is another parameter which can aVect the porous structure of the final membrane material.In the case porosity, pore size and pore size distribution must be analysed. A quite diYcult problem is the characterisation of microporous of materials obtained from particulate sols, a grain growth phenomenon is generally found to be responsible for an materials8–10 for which thermodynamic laws are no longer valid because of very strong interactions between intrusive increase of pore size corresponding to the evolution of residual voids between the sintered particles.The classical physical gels fluids and pore walls. The methods based on gas adsorption 56 J. Mater. Chem., 1999, 9, 55–65ture in supported ceramic films. Such materials oVer attractive properties when used as ceramic nanofilters.15 Ceramic nano- filters are being recognised as of growing importance owing to increasing demands for membrane systems which are able to separate ions and small molecules in harsh working conditions (high temperature, extreme pH, organic solvent media). Regarding the problem of aggregation, a solution has been to generate steric repulsion in organic media through the interaction of organic shells preventing close contact between nanoparticles in the sol, Fig. 3. Taking into account the role of chelating agents as blocking functional groups in metal alkoxide condensation16 a supported microporous zirconia layer has been obtained from a nanoparticulate organic sol. Zirconium isopropoxide was used as the ceramic precursor to synthesise the organic particulate sol.17 In order to avoid the precipitation of non-homogeneous hydroxide particles during the hydrolysis step, the alkoxide reactivity was modified by a strong complexing ligand, acetylacetonatee (acac).Hacac reacts with the alkoxides to form mixed complexes which have diVerent physico-chemical properties and, more importantly, which are more diYcult to hydrolyse than alkoxy groups:18 Zr(OR)4+Hacac�Zr(OR)3(acac)+ROH Then this ligand acts as a functionality blocker when substoichiometric hydrolysis ratios are used.Consequently, a Hacac/M molaratio greater than l was used to prevent precipitation and led to a stable nanoparticulate sol prepared in air without any precipitation. Here the stability of nanoparticles in the sol results from the interacting organic shells formed with the acac groups as shown in Fig. 4. The supported zirconia layer obtained after sintering at Fig. 3 Interaction and corresponding potential (V) between steric 500 °C exhibited a tetragonal crystalline structure and revealed barriers due to the presence of organic shells around ceramic particles. a very fine texture when observed by TEM. In the case of a Zr/Hacac molar ratio=2, a mean grain size of about 6 nm are the most convenient for the study of microporous materials.was measured. Powder X-ray diVraction (Scherrer’s formula) The conventional apparatus uses volumetric measurements of was used for the determination of an average individual crystal the adsorbed gas quantities.11 In the case of supported thin size. Results are consistent with particle size observed by TEM films, the amount of porous layer which can be introduced in in the supported membrane, Fig. 5.In this membrane packing a standard-size sample cell is often too small to be correctly of the nanoparticles results in a pore diameter of 1.4 nm and characterised. Other types of adsorption equipment were therea porosity of 18%. fore developed, like the surface acoustic wave device12 and a device based on the measurement of the refractive index by ellipsometry.13 However these techniques require specific conditions concerning the nature of the substrate.In the first case, the substrate must be piezoelectric, while in the second case, the layer must be transparent and the refractive index of the dense substrate must be quite diVerent from that of the layer.In other respects, each substrate exhibits specific surface properties which can influence the final characteristics of the deposited coatings. In a recent work we combined adsorption measurements and ellipsometry for the characterisation of supported microporous silica layers.14 Both silica layers and the corresponding bulk silica materials were prepared by the sol–gel polymeric route using tetraethoxysilane as the starting material.This work led to the conclusion that during the drying step the support has a great influence on the porous structure formation of the coated layer. The support prevents shrinkage of the thin layer resulting in a higher porosity compared to the bulk material. Further modification or adaptation of these techniques must be completed in order to better characterise the porous structure of supported membranes, particularly microporous membranes. 2.2 Preparation of a zirconia nanofilter from a nanoparticulate Fig. 4 Cluster formed from condensation of acetylacetone (Hacac) organic sol modified zirconium isopropoxide; acac groups are at the periphery of Starting from the above considerations it is possible, from a the cluster as an organic shell and prevent interpenetration of clusters which behave as individual nanoparticles.nanoparticulate sol, to create a connected microporous struc- J. Mater. Chem., 1999, 9, 55–65 57of ammonium metavanadate.20,21 The synthesis of original metallo-organic molecular precursors has been attempted in order to obtain a homogeneous liquid, adapted to casting and being a precursor of a pure VMgO catalyst, with a controlled V/Mg ratio, a high specific surface area and a high catalytic activity (dispersion of active sites).A functional magnesium alkoxide Mg(OCHCH3CH2OCH3)2 has been synthesised by direct reaction between magnesium and 1-methoxypropan-2- ol22 and reacted with a new original vanadium oxoalkoxide VO(OCHCH3CH2OCH3)3 prepared by reaction of ammonium metavanadate with 1-methoxypropan-2-ol. The reaction leads to the formation of (OR)n-1VOMg(OR)m-1 species. This heterometallic compound can be easily cast on a support, hydrolysed, dried in vacuum at room temperature and finally fired at 600 °C in air to lead to the final VMgO catalyst with a controlled V/Mg ratio.Apart from the fact that this specific method is well adapted to catalyst casting on a support (e.g. for catalytic membrane reactor applications), the derived VMgO powders have been found to exhibit original specific characteristics when compared to conventional VMgO catalysts derived from precipitated Mg(OH)2.These characteristics concern both the textural properties of the catalysts (high specific surface area: 110 m2 g-1 at 14 wt% V, round particles of about 25 nm instead of platelets for conventional catalysts) and also their structural characteristics (delayed crystallisation of Mg3V2O8, high dispersion of amorphous vanadium species on the surface of MgO particles).Fig. 6 compares the morphology of the VMgO powders (14 wt% V) obtained at 600 °C from the Mg(OH)2 salt precursor21 and from the above-mentioned alkoxides.The specific characteristics of the alkoxide derived catalyst, attributed to a high homogeneity of V and Mg species distribution in the molecular Fig. 5 Electron microscopy images of the active zirconia layer of a precursor, induce attractive catalytic properties for the oxyceramic nanofilter prepared at 500 °C from a sol of Hacac modified dehydrogenation of propane.Some preliminary results have zirconium alkoxide precursor. (a) SEM cross-section image of the supported zirconia layer, (b) TEM image of the zirconia nanoparticles constituting the zirconia layer. 3.0 Catalytically active nanophase materials from the sol–gel process The sol–gel process is also of particular interest for designing catalytically active materials with specific properties and/or with improved performances compared to those of classical catalysts.It allows preparation of catalytically active materials which can be directly cast in/on supports at the sol stage. This is a great advantage for catalytic membrane development.19 The classical synthesis methods for conventional catalysts often start from salts or oxide precursors and involve precipitation, impregnation or even solid/solid reactions.These methods are usually not adapted to a homogeneous casting of the catalyst on a support and may lead to limited specific surface areas and to a heterogeneous distribution of active species. The sol–gel process, starting from a homogeneous distribution of the precursors at the molecular level, can in many cases improve these specific criteria.Indeed a large range of methods and precursors can be investigated to obtain the required powders or film-shaped materials either as a pure catalyst or homogeneously doped or dispersed in a matrix. Furthermore, the specificity of the process can lead to original materials which can help in a better understanding of the catalytically active sites in a specific catalyst.Two examples showing the potential of the sol–gel process for catalyst synthesis are described in the following sections. 3.1 VMgO catalysts prepared from original metallo-organic precursors The method classically used to prepare conventional VMgO Fig. 6 FESEM observation of VMgO (14 wt.% V) catalytic powders catalysts for the oxydehydrogenation of propane involves the obtained at 600 °C from: (a) Mg(OH)2 salt precursor (SBET= 90 m2 g-1), (b) novel metal alkoxides (SBET=110 m2 g-1).precipitation of Mg(OH)2 and its impregnation with a solution 58 J. Mater. Chem., 1999, 9, 55–65been reported recently.23 VMgO powders with various V/Mg ratios are currently being investigated and should help in a better understanding of the nature of the catalytic sites active for the oxydehydrogenation of propane.The corresponding optimised mesoporous membranes will be tested as contactors in a catalytic membrane reactor configuration (European Joule program JOE3-CT-95 0022). 3.2 Synthesis of CeO2 based materials for DeNOx applications Most of the three-way catalysts (TWCs) currently used for car exhausts are made with a washcoat (on honeycomb supports) of a mixture of oxides, namely Al2O3 and CeO2, which is impregnated with Pt, Pd or Rh salts.24 The main problems related to classical TWCs lie both in the elevated price of noble metals and in the instability of the catalysts upon use (oxide sintering and metal aggregation).25 The goal was in this case to investigate sol–gel methods and precursors providing the best conditions for the preparation of eYcient CeO2 based catalysts with a high specific surface areas at 1000 °C and from precursors able to be directly cast on supports.A sol–gel preparation method was used, starting from Ce(acac)3·xH2O and Al(OsBu)3 as ceria and alumina precursors, respectively. Hydrolysis and condensation of the precursors in the presence of water was carried out in hexylene glycol at 100 °C for 24 hours with the following weight% composition for the initial sol: precursors 40/hexylene glycol 50/water 10.Gel formation depends on the precursor molar ratio, for Al(OsBu)3/Ce(acac)3·xH2O>0.25 gel formation occurred immediately after water addition. The as-formed gels were heat treated at 1000 °C and the resulting powders analysed by X-ray, FESEM and N2 adsorption–desorption.DiVerent Fig. 8 Morphological aspect of CeO2–Al2O3 based DeNOx catalyst: Ce(acac)3·xH2O/Al(OsBu)3 ratios were investigated. The (a) powders heat-treated at 1000 °C, (b) corresponding honeycomb homogeneous dispersion of CeO2 in the alumina matrix was supported layer. found to prevent fast growth of cerium oxide crystallites with temperature when the CeO2 content in the final material was less than 25 mol%. The evolution of the powder specific surface remained amorphous up to 700 °C, as only the sole CeO2 areas versus CeO2 content and firing temperature is reported diVraction lines were detected.At 1000 °C, some poorly defined in Fig. 7. diVraction lines of c-Al2O3 began to appear.This delayed Specific surface areas higher than 100 m2 g-1 have been crystallisation of alumina was attributed to an eVect of CeO2, obtained at 1000 °C for CeO2–Al2O3 based materials containas the structure of alumina prepared without any additive was ing less than 25 mol% CeO2. When the CeO2 content was too already crystalline at 450 °C. The CeO2 crystallite sizes deter- high (47 mol%) the dispersive eVect of alumina decreased mined from the X-ray diVraction peak widths were smaller and SBET at 1000 °C was lower than 10 m2 g-1.The SEM than 20 nm. observation of the powders with 25 mol% of CeO2 (Fig. 8a) The corresponding sols lead to homogenous layers when showed that the grain sizes were always homogeneous and cast on honeycomb supports, Fig. 8b. High temperature ageing never exceeded 30–40 nm at 1000 °C. Without any alumina in tests are in progress in order to check the structural stability the powder the grain sizes were larger than 200 nm at 1000 °C, of the derived TWC catalysts. The sol–gel route previously and in the case of alumina alone the grain sizes were about described is also currently being investigated for the direct 15 nm at 1000 °C.It should be noted that the alumina matrix insertion of zirconia (forming a solid solution with ceria and enhancing its redox properties) and of Pd catalysts at the sol stage (European Brite EuRam program BRPR-CT 96 0290). 4.0 New methods based on amphiphilic media to tailor porous structures in sol–gel derived solids 4.1 Interest of surface active agents for the control of pore formation Surface active agents represent a specific variety of additives which has been recently investigated in sol–gel processing and not only to limit gel cracking.Indeed, one of the other interests of such additives concerns the control of the hydrolysis step of highly reactive alkoxides, such as titanium alkoxides, by the use of reverse micelle sol–gel systems.26 Another utilisation under investigation in our group is the modification of gel porous structures when surfactant molecules are added after the hydrolysis of an alkoxide.27 As shown in Fig. 9, the Fig. 7 Evolution of the specific surface area in CeO2–Al2O3 powders surfactant molecules introduced in the reaction medium inter- as a function of the CeO2 content and firing temperature.Sol–gel organic route from Ce(acac)3·xH2O. act with growing clusters and their role is in this case the J. Mater. Chem., 1999, 9, 55–65 59sized ‘crystallites’ can be obtained; a firing treatment removes the surfactant aggregates and liberates the ordered porosity. The initial mixture composition controls the structure (lamellar, hexagonal or cubic) of the mesophase. The main material labelled MCM41 exhibits a hexagonal ordered porous texture. The pore sizes are directly controlled by the length of the surfactant alkyl chain and ranges between 1.8 and 10 nm.Monnier et al.32 explained the co-operative process of oxide polymerisation and mesophase formation by the evolution of coulombic type interactions between the surfactant polar head and the growing silicate oligomers.Subsequent studies33–38 showed that the synthesis of a silica network in ordered amphiphilic media could also be obtained by sol–gel routes at Fig. 9 Schematic representation of the role of non-ionic surfactants in room temperature, under atmospheric pressure and with other the limitation of cluster growth and aggregation: (a) gel formation types of surfactant, anionic or non-ionic.In contrast to the with added surfactants, (b) gel formation in the absence of surfactants. M41S type structured mesoporous solids our main objective has been to create, during the sol-to-gel transition, ordered limitation of both interpenetration and condensation between microporous structures in oxide materials using the directing clusters.In addition to the material pore sizes and pore eVect of lyotropic liquid crystal mesophases.33–36 Moreover volume, it is also of particular interest to control the pore these materials have been shaped as continuous supported connectivity which is an important parameter aVecting the thin films in view of membrane applications for liquid and gas material permeability. Nakanishi et al.28 used the spinodal separation.34 We first worked on ordering of porous structures decomposition of sol–gel solutions into two phases, one rich in sol–gel derived silica materials with further investigation of in solvent and the other rich in polyethylene oxide and other metal oxide systems like alumina.Beyond the prepinorganic compounds, to produce dried gels with intercon- aration of the ordered structures an important improvement nected pores in the micrometer range.Unfortunately this in porous membrane characteristics can be expected in terms method cannot be extended to the nanometer range and the of sharp control of pore tortuosity. Tortuosity is one of the templating approach is much more attractive by using remov- important structural parameters which governs membrane able individual templating units or interconnected networks.permeability. As a prospect we relate at the end of this paper At the nanometer scale the lyotropic liquid crystal mesophases some preliminary works on the orientation of ordered domains appear as very attractive systems which can produce removable in silica supported layers with the aim to prepare membranes interconnected networks.Moreover the crystalline structure of with low tortuosity. these templates can direct the growth of the inorganic network and leads to the formation of ordered structure and porosity. 4.2 Control of residual porosity in silica materials derived from The periodicity of the structures induces monodispersed pore polymeric gels sizes and uniform connectivity of both the porous and solid networks.29 Lyotropic liquid crystals are usually lamellar, Based on the same concept as for zirconia nanoparticulate hexagonal and cubic phases (Fig. 10). They are obtained by a organic sols, an interesting way to tailor the microporous self-assembly process of surfactant molecules for intermediate structure in silica materials and membranes is based on compositions in the water–surfactant binary phase diagrams.hindered cluster formation at the sol stage (Fig. 9) using non- The sizes of the constituent units, i.e. the amphiphilic aggre- ionic surfactants. In this approach, non-ionic surfactants, gates, usually range from 2 to 10 nm and are susceptible to alkylaryl polyether alcohols Triton (TX) of diVerent molecular direct the formation of inorganic networks with an ordered weights (X=1–30) were added to tetraethoxysilane sols (molar porosity from the supermicroporous domain (>0.7 nm) up to composition: TEOS/C2H5OH/H2O/HCl=1/4.5/4/0.02). On the mesoporous domain (>2 nm).the basis of sol and gel characterisation (gelation time, 29Si There is currently considerable interest in this new field of NMR, QELS, SAXS, thermoporometry) and material characmaterial science which integrates both a biomimetic approach terisation (FESEM, N2 adsorption, FTIR) the eVect of surfac- (biomineralisation processes) and the concept of nanochemis- tant chain length (X) and TX/TEOS ratio on derived sols, try. In 1992 researchers at Mobil published a synthesis of gels and ceramic materials has been explored.27 In other mesoporous molecular sieves using the templating eVect of respects thermoporometry is a method which is well adapted lyotropic liquid crystal mesophases.30,31 The aluminosilicate to the study of the evolution of mesoporosity in wet gels.We or silicate materials M41S were prepared by hydrothermal have shown,39 using this technique, that the Triton X surfactreatment of solid or molecular precursors of alumina and tant molecules prevent interpenetration and further condensilica in the presence of cationic surfactants of alkyltrimethyl- sation of clusters in the above-mentioned sols. Because surface ammonium halide type.As in the case of zeolites, micrometer active agents are susceptible to interaction with silica oligomers derived from TEOS by van der Waals forces or by interaction with OH groups, the formation of an organic shell made of surfactant molecules between the clusters can be assumed. On the other hand the possibility of micelle formation during drying, by exceeding the surfactant cmc, cannot be set aside but there was no experimental evidence of that.Anyway, in both cases the resulting steric hindrance should limit further condensation of clusters during sol ageing.This is confirmed experimentally by reduced bonding between clusters which explains the formation of more stable sols with longer gelation time and a reversible sol-to-gel transition upon gel shaking. During the heat treatment (450 °C, 2 h) there is formation of Fig. 10 Drawing of three lyotropic liquid crystal structures: (a) lamela homogeneous microporous material due to the elimination lar, (b) bicontinuous cubic (space group Ia3d) viewed along the 100 of the surfactant molecules in contact with the silica gel matrix.direction, (c) hexagonal. J. Charvolin, Prog. Colloid Polym. Sci., 1990, 81, 6 (reproduced with permission). N2 adsorption experiments revealed that N2 (kinetic diam- 60 J.Mater. Chem., 1999, 9, 55–65be of importance for material reactivity: ionic exchange, organic grafting, specific interactions with gas or liquid phases.41 This type of sol prepared in the presence of surfactant is also well adapted for preparing homogeneous dispersions of metals in a silica matrix. An example is reported39 relating the dispersion of a platinum salt (PtCl4) in a TX–TEOS sol.Thermoporometry of the derived gels showed that the eVect of PtCl4 is opposite to the eVect of Triton X and leads to gels with a higher crosslinking degree. The conjunction of the steric hindrance and rigidity eVects of Triton X and PtCl4 respectively produces rigid gels in which the surfactant molecules are Fig. 11 Evolution of specific surface area and microporous volume of trapped.This is an original method to prepare highly microsilica materials versus the molar ratio TX/TEOS. porous SiO2/Pt supported membranes which contain a homogeneous dispersion of catalytically active nanosized Pt eter=3.96 A° ) does not penetrate the porous structure of SiO2 particles.41 materials prepared without a surface active agent. When surfactants were used, type I isotherms were obtained which 4.3 Templates based on a removable interconnected network.are characteristic of microporous materials. The mean Structure directing eVect of lyotropic liquid crystal mesophases hydraulic pore radius RH (estimated by the MP method) can In a previous work, silica gels with ordered microporosity be varied between 3.2 and 7.0 A° .This parameter is slightly were prepared using short length cationic surfactant molecules aVected by the TX/TEOS molar ratio but increases with the which show less ability to form lyotropic mesophases.33,36 The surfactant chain length X. Surfactant addition greatly increases sol compositions (Table 1) were based on a silica precursor, the pore volume and specific surface area of silica materials.tetramethoxysilane, mixed with a series of cationic surfac- As shown in Fig. 11, these two parameters, in the case of X= tants of alkyltrimethylammonium bromide type: 3, reach a maximum (SBET=500 m2 g-1, Micropore volume= CxH2x+1(CH3)3N+Br- (x=8, 10, 12, or 14) and water. In 0.239 cm3 g-1, 35.5% porosity) for TX/TEOS=0.55. Fig. 12 these systems gelation of the silica network and mesophase shows the evolution of the pore volume and specific surface formation occur simultaneously.The X-ray diagrams area as a function of the surfactant chain length X for a ratio (Fig. 13a,c) of two wet gels prepared according to this method TX/TEOS=0.55. The curves seem to reach a plateau for X (for x=8 and x=14) can be assigned to hexagonal and higher than 10.In fact with long chains, larger pore sizes and lamellar phases respectively. The large Bragg spacing d100 can wider distributions were obtained, whereas to obtain small be related to the existence of a liquid crystal structure and its pores with a sharp distribution, a high pore volume and value varies with the length of the chain of the used surfactant specific surface area, X values of 3 to 10 and TX/TEOS ratio molecule.In the investigated surfactant series, Table 2, the wet of 0.55 were preferred. This kind of sol was successfully used to prepare supported microporous silica membranes.27 A Table 1 Composition of the silica sols synthesised in presence of thorough investigation of residual OH groups by FTIR specalkyltrimethylammonium bromides, CxH2x+1(CH3)3N+Br- (with x= troscopy40 showed another attractive feature of this type of 8, 10, 12, or 14) material related to the presence of single OH surface groups at high temperature.These single OH groups (typical sharp Reagent Surfactant Tetramethoxysilane Water IR absorption band at 3744 cm-1) were clearly distinguished Weight % 21 12 67 on FTIR spectra from adjacent OH groups (broad absorption Addition order 1 2 3 bands at 3670 and 3500 cm-1) which completely disappeared upon firing at 600 °C whereas the band at 3744 cm-1 remained.The unusual presence of these groups is specifically due to the presence of surfactant molecules in the sol and to their interaction by hydrogen bonding with the silica based clusters. At the sol and gel stages, these OH groups are ‘protected’ by the surfactant hydrophilic head.This interaction is maintained till the surfactant thermal decomposition, between 200 °C and 300 °C, when single OH groups are liberated. These groups have been shown to be relatively distant from each other because of surfactant steric hindrance and cannot easily condense as in classical sol–gel derived silica. This original surface characteristic of materials derived from TX–TEOS sols should Fig. 13 X-Ray patterns of various wet and thermally treated gels: Fig. 12 Evolution of specific surface area and microporous volume of (a) x=8, wet gel; (b) x=8, gel treated at 450 °C under nitrogen; (c) x=14, wet gel; (d) x=14, gel treated at 450 °C under nitrogen. silica materials versus surfactant chain length.J. Mater. Chem., 1999, 9, 55–65 61Table 2 Textural properties of the thermally treated gels calculated assuming given mesophase structures Gel wp d100/nm dc/nm SM/m2 g-1 SBET/m2 g-1 DS/m2 g-1 D/nm x=8 0.55 2.3 2.1 1074 1260 186 33 x=10 0.55 2.8 2.5 883 1040 157 39 Gel wp d100/nm wp/nm SM/m2 g-1 SBET/m2 g-1 DS/m2 g-1 D/nm x=12 0.58 2.0 1.2 1082 1100 18 366 x=14 0.58 2.2 1.3 1033 1090 57 120 gels exhibit a hexagonal structure for the lower values of x (8 of the surfactant molecule, and have the same order of magnitude as the length of the surfactant molecules, ls, calcu- and 10) whereas for higher values of x (12 and 14) the wet gels are lamellar. After drying and firing at 450 °C under lated using the Tanford relation.45 It has also been noted that the calculated mesophase surface area SM is lower than the nitrogen, a broad diVraction peak is observed in the X-ray patterns (Fig. 13b,d).The decrease of d100 from the wet to the measured value, SBET. This excess of specific surface area, DS, has been assigned to the existence of an interface between the thermally treated gels has been attributed to a shrinkage of the material during the drying step at low temperature and to ordered domains. To a first approximation the ordered domains have been assumed to be spherical with a diameter the condensation of reactive groups of the silica network. Taking into account the sharpness of the diVraction peaks for D calculated from the following equation:36 the wet gels compared to an important broadening of the D=6/[DS(1-wp)rs] diVraction peaks after heat treatment it can be stated that the The skeletal density, rs, of the oxide network measured by size of the ordered domains is reduced during the transition helium pycnometry is equal to that of amorphous silica, i.e.from the wet to the thermally treated gel. Final ordered 2.2 g cm-3. The pore volume fraction in the mesophase struc- domains of about 20 nm in size were determined using the ture wp, is determined from nitrogen adsorption measurements.Scherrer relation.42 The final size of these domains is small The calculated values of D for the hexagonal material have compared to the size of the liquid crystal cells implied in the the same order of magnitude as the sizes determined from the formation of the material. This small size and the probable width of the diVraction peaks using the Scherrer equation.random orientation of the ordered domains could explain the For the lamellar materials, the mean calculated values are limited collapse of the lamellar structures upon heating at larger but in that case D depends strongly on the d100 and 450 °C. wp values. The main remarkable results (Table 2) obtained with these From these results it has been shown that it is possible to materials exhibiting a directed porous structure can be sumdirect, at room temperature, during the sol-to-gel transition, marised as follows.The BET specific surface area, SBET, and an ordered microporous structure in silica gels using lyotropic the total porosity are very large, more than 1000 m2 g-1 and crystal mesophases as reaction media for silicon alkoxides.around 60% respectively. The calculated diameter of the The diVerent experimental values which characterise the cylindrical pores, dc, for the hexagonal gels and the calculated porous texture of the final material can be correlated to width of the slit-like pores, wp, for the lamellar gels are calculated values assuming a polycrystalline material composed consistent with the size of the micropores experimentally of hexagonal or lamellar domains.Remarkably, the micropore determined by the MP43 and Horvath–Kawazoe44 methods. size of the calcined silica gels can be modulated by the length Fig. 14 shows that the average hydraulic diameter, 2rh, and of the alkyl chain of the used surfactant molecules.Starting the average Horvath–Kawazoe diameter, dHK, are almost from the results obtained on ordered porous silica materials, proportional to x, the number of carbons of the alkyl chain the synthesis method was tentatively adapted to the preparation of alumina porous material. In fact aluminium alkoxides are too reactive as precursors of polymeric gels. Usually alumina colloidal gels are obtained from the destabilisation of pseudo-boehmite hydrosols.Kunitake46 showed that a lamellar mesophase produced from a cationic surfactant can be successfully used to direct the aggregation of such colloidal alumina particles. The initial anisotropy of the material is maintained after the departure of the surfactant until the transformation into a-Al2O3 at high temperature.However it can be deduced from the available observations by electron microscopy that the resulting materials are expanded and highly macroporous. We mentioned in a preliminary work47 the possibility of alumina gel synthesis in amphiphilic media using an aluminium salt as the gel precursor. Following this method an aqueous solution of aluminium chloride (0.1 M) containing a quaternary ammonium type cationic surfactant was used to produce an aluminium hydroxide gel.After a pre-neutralisation step with ammonia (molar ratio [NH3]/[Al ]=1), an in situ controlled neutralisation was carried out by introduction of urea (molar ratio [urea]/[Al ]=1) in the starting solution and gentle heating at 80 °C. The thermal decomposition of urea into carbon Fig. 14 Measured and calculated average size of pores versus x, the dioxide and ammonia induced an increase of the pH of the number of carbons in the alkyl chain of the surfactant. (a) 2rh with rh solution and the formation of the alumina gel. This kind of the measured hydraulic radius; (b) 4rh; (c) dHK, measured average controlled neutralisation was previously used for the synthesis Horwath–Kawazoe diameter; (d) ls, calculated length of the surfactant of monodispersed powders48 and gels.49 The main interest of molecule; (e) dc, calculated diameter of the cylindrical pores in the this process is that it can lead to continuous and homogeneous hexagonal structure; (f ) wp, calculated width of the slit-like pores in the lamellar structure.thin layers by deposition of partially neutralised solutions 62 J.Mater. Chem., 1999, 9, 55–65Table 3 Textural properties of the alumina gels after thermal treatment under nitrogen Name Heat treatment/°C SBET/m2 g-1 Pore volume/cm3 g-1 Pore size/nm WS 450 430 0.29 2.7 8C32 450 460 0.33 2.9 14C32 450 485 0.66 5.5 and further ageing of the coatings under controlled field (11.7 T) used recently by Tolbert et al.53,54 to orient the mesophase templated silica networks. atmosphere.The used cationic surfactants were octyl- [C8H17(CH3)3N+Br-] and tetradecyl- [C14H29(CH3)3N+Br-] Consistently our current approach to control the tortuosity in supported porous layers consists of the introduction of trimethylammonium bromides. The added amounts of surfactant were determined from the analysis of the available binary magnetic nanoparticles in the gelling solution. As a matter of fact these particles could have a seeding role in the hetero- diagram: tetradecyltrimethylammonium bromide–water.50 On this diagram the weight percentage of surfactant corresponding geneous nucleation of templating mesophases with an induced eVect on the size of ordered domains.Moreover magnetic to the boundary between the isotropic (lower wt%) and hexagonal mesophase (upper wt%) areas is located at ca. 32 wt%. nanoparticles could be used to orient the mesophases. A hydrosol of maghemite (c-Fe2O3) nanoparticles55 (9.5 nm in In contrast to silica gels, X-ray analysis carried out on a series of wet gels and on the corresponding calcined alumina size, ZPC=7.3, stability pH=2) was used as a ferrofluid source.The stability of silicon alkoxide sols in the presence of gels showed that the ordered porosity formed in the wet gels cannot be maintained upon heat treatment.47 In the case of surfactant molecules and maghemite particles and the possibility to prepare ordered porous structures from these sols have the wet gel 8C32 (x=8), a main diVraction peak at 2.3 nm can be assigned to the presence of a hexagonal mesophase.In been investigated recently.56 Silica gel precursor (TMOS) and surfactant molecules CxH2x+1(CH3)3N+Br- (x=8, 10) were the case of the wet gel 14C32 (x=14), the position d of the main peak (d<3 nm) and the existence of another peak of used as starting materials for gel preparation. A typical weight% composition of gelling solutions is: seeding hydrosol lower intensity located at ca.d/2 are in favour of the presence of a lamellar mesophase. The diVraction patterns of the gels 44, methanol 1, TMOS 8, NH3 (0.1 M) 26, surfactant (x=8) 21. The seeding level in the resulting gels corresponds to 1 after the thermal treatment at 450 °C are consistent with that of a low temperature transition alumina.47 The lyotropic particle per 105 nm3.Thick layers (1 mm>thickness >0.5 mm) were deposited on flat substrates. The texture crystal phase, present in the wet gels, has an influence on the porous texture of the calcined alumina gels (Table 3). The orientation of the thick layers was deduced from the variation of the X-ray diVraction peak at d=2.45 nm associated to the porous texture characteristics of gel 8C32 calcined at 450 and 800 °C are close to those of gel WS (without surfactant) hexagonal mesophase.The peak intensities were measured at two incident angles, a between the incident beam and the calcined at 450 °C. This is a very interesting result owing to the fact that porous alumina materials, obtained by dehy- substrate, b between the projection of the incident beam on the substrate and the shear stress direction of sheared wet gels.dration of well crystallised hydroxides, exhibit a continuous decrease of the pore size and of the surface area as the heat In the absence of a magnetic field during gel formation, seeded or unseeded wet gels exhibit very reproducible variations of treatment temperature is increased.In the case of gel 14C32 the specific surface area at 450 °C is maintained at a high the intensity of the diVraction peak at 2.45 nm versus a and b. In Fig. 15(a) the strong maximum observed for a equal to value but with an almost two-fold increase of pore size and pore volume compared to the reference gel WS. Unfortunately the diVraction angle h is assigned to an alignment of the micellar cylinders and of the associated diVraction planes the peak located at low diVraction angle and associated with an ordered porosity for the 8C32 and 14C32 wet gels is not parallel to the surface of the substrate.The variation versus b, Fig. 15(b), is attributed to an alignment of the micellar observed with the calcined gels. These results have been explained taking into account the low equivalent fraction of cylinders along the shear stress direction.When a magnetic field (0.6 T) is applied during gel layer formation in a direction oxide (Al2O3) in the initial sol composition (ca. 3 wt% for 8C32) and the fact that the produced aluminium hydroxide perpendicular to the surface of the substrate, no more diVraction peaks versus a or b are observed as expected in the case gel is not a strongly cohesive gel as it is for silica.The strengthening of the inorganic network must clearly occur of alignment of the micellar cylinders along the magnetic field direction. Further X-ray experiments on wet gels are under before the departure of the template to favour the direct templating eVect of the ordered amphiphilic mesophase. way in order to confirm these results and to investigate the specific interactions between the seeding particles and the Another aspect of these templated porous textures is related to membrane applications and has to do with the anisotropy lyotropic crystal phases during gel formation.of the ordered domains in the case of supported layers. Interactions at the gel–substrate and air–gel interfaces favour 5.0 Conclusion the preferential orientation of the hexagonal and lamellar structures along these interfaces.Moreover the shear stress The most recent advances in sol–gel processing of porous ceramic oxides have been reviewed and illustrated by recent which is applied during the deposition can induce an alignment of the crystalline structures in the direction of the stress.These results of our group in porous solid preparation. The current state of development of the sol–gel process allows a precise anisotropies are detrimental to the mass transfer performance of the final supported porous materials. In order to improve control of composition, grain structure and pore structure in ceramic oxide materials. These new developments are based their permeability, the size of the ordered domains must be increased and ordered porous structures oriented with the on a better mechanistic control of the reactivity of precursors and derived growing species.The interplay between liquid resulting pores along a direction perpendicular to the layer surface. Liu et al.51 showed that nanoparticle seeding favours crystal organic chemistry and the growth of inorganic frameworks provides a rare opportunity to synthesize a wide range the formation of the templating mesophases in the synthesis medium.In other respects, Fabre et al.52 obtained ferrosmectics of original structures. These systems are potentially tunable for a desired application and oVer an unequalled ability for by doping swollen lamellar phases with ferrofluid nanoparticles. These hybrid mesophases can be oriented under very low the synthesis of mesoporous and microporous materials.Regarding both fundamental research and applications, magnetic field (<1 T) in comparison to the intense magnetic J. Mater. Chem., 1999, 9, 55–65 6313 M. A. Fardad, E. M. Yeatman, E. J. C. Dawnay, M. Green and F. Horowitz, J.Non-Cryst. Solids, 1995, 183, 260. 14 A. Ayral, A. El Mansouri, M. P. Vieira and C. Pilon, J. Mater. Sci. Lett., in press. 15 C. Guizard, C. Mouchet, R. Vacassy, A. Julbe and A. Larbot, J. Sol–Gel Sci. Technol., 1995, 2, 483. 16 C. Sanchez and J. Livage, New J. Chem., 1990, 14, 513. 17 R. Vacassy, C. Guizard, V. Thoraval and L. Cot, J. Membr. Sci., 1997, 132, 109. 18 C. Sanchez, J.Livage, M. Henry and F. Babonneau, J. Non-Cryst. Solids, 1988, 100, 65. 19 A. Julbe, C. Guizard, A. Larbot, L. Cot and A. Giroir-Fendler, J. Membr. Sci., 1993, 77, 137. 20 A. Pantazidis and C. Mirodatos, 11th International Congress on Catalysis, Studies in Surface Sciences and Catalysis, ed. J. W. Hightower, W. N. Delglass, E. Iglesia and A. T. Bell, Elsevier Science B.V., Amsterdam, 1996, 101-B, 1029. 21 A. Pantazidis, A. Auroux, J.-M. Herrmann and C. Mirodatos, Catal. Today, 1996, 32, 81. 22 L. Albaric, N. Hovnanian, A. Julbe, C. Guizard, A. Alvarez- Larena and J. Piniella, Polyhedron, 1997, 16, 587. 23 A. Julbe, L. Albaric, N. Hovnanian, C. Guizard, A. Pantazidis and C. Mirodatos, in Proc. 4th Workshop of the ESF Network: Catalytic Membrane Reactors—Optimisation of Catalytic Membrane Reactor Systems, 1997, ed. R.Bredesen, SINTEF, Oslo, pp. 215–220. 24 P. Mare�cot, L. Pirault, G. Mabilon, M. Prigent and J. Barbier, Appl. Catal. B: Environment, 1994, 5, 57. 25 C. J. Stournaras, A. Tsetsekou, M. Muhammed, C. Guizard, A. Julbe, P. Bekiaroglou, N. Moschoudis, J. P. Joulin, D. Leonidopoulos and M. Debenedetti, in Proc. First European Conference on Clean Cars and First Hellenic Eco Rally, 1997. 26 C. Guizard, A. Larbot, L. Cot, S. Peres and J. Rouvie`re, J. Chim. Phys., 1990, 87, 1901. Fig. 15 X-Ray diVraction analyses on a sheared hexagonal silica layer 27 A. Julbe, C. Balzer, J. M. Barthez, C. Guizard, A. Larbot and (x=8). (a) Evolution of the peak intensity versus the incidence angle, L. Cot, J. Sol–Gel Sci. Technol., 1995, 4, 89.a; (b) evolution of the peak intensity versus b, angle between the 28 K. Nakanishi, Y. Yamasaki, H. Kaji, N. Soga, T. Inoue and direction of the shear stress and that of the incidence angle. N. Nemoto, J. Sol–Gel Sci. Technol., 1994, 2, 227. 29 S. Hansen, Adv. Mater., 1993, 5, 113. 30 C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature, 1992, 359, 710. many exciting developments are expected in the near future 31 J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, from sol–gel derived nanophase materials, namely for catalytic C. T. Kresge, K. D. Schmitt, C. T. W. Chu, D. H. Olson, E. W. Sheppard, S. B. McCullen, J. B. Higgins and J. L. Schenker, and membrane separation applications. J. Am. Chem. Soc., 1992, 114, 10 834. 32 A. Monnier, F. Schu� th, Q. Huo, D. Kumar, D. Margolese, R. S. Maxwell, G. D. Stucky, M. Knamurty, P. PetroV, References A. Firouzi, M. Janicke and B. F. Chmelka, Science, 1993, 261, 1 Solid catalysts and porous solids, ed. M. E. Davis and 1299. I. E. Maxwell, in Curr. Opin. Solid State Mater. Sci., 1996, 1, 54. 33 T. Dabadie, A. Ayral, C. Guizard, L. Cot, J. C. Robert and 2 J. D. F. Ramsay and B. O. Booth, in Fundamentals of Adsorption, O. Poncelet, Mater. Res. Soc. Symp. Proc., 1994, 346, 849. ed. A. B. Mersmann and S. E. Scholl, United Engineering Trustees 34 T. Dabadie, A. Ayral, C. Guizard, L. Cot, J. C. Robert and Inc., 1991. O. Poncelet, in Proceedings of the Third International Conference 3 Fundamentals of Inorganic Membrane Science and Technology, ed. on Inorganic Membranes, July 1994, Worcester, USA, ed. A. J. Burggraaf and L. Cot, Elsevier, Amsterdam, 1996. Y. H. Ma, 3rd ICIM-Worcester Polytechnic Institute, Worcester, 4 Sol–gel Science, ed. C. J. Brinker and G. Scherer, Academic Press, 1994, p. 411. New York, 1990. 35 T. Dabadie, A. Ayral, C. Guizard, L. Cot, C. Lurin, W. Nie and 5 C. J. Brinker, T. L. Ward, R. Sehgal, N. K. Raman, S. L. Hietala, D. Rioult, J. Sol–Gel Sci. Technol., 1995, 4, 107. D. M. Smith, D.-W. Hua and T. J. Headley, J. Membr. Sci., 1993, 36 T. Dabadie, A. Ayral, C. Guizard, L. Cot and P. Lacan, J. Mater. 77, 165. Chem., 1996, 6, 1789. 6 C. Guizard, in Fundamentals of Inorganic Membrane Science and 37 J. E. Martin, M. T. Anderson, J. Odinek and P. Newcomer, Technology, ed. A. J. Burggraaf and L. Cot, Elsevier, Amsterdam, Langmuir, 1997, 13, 4133. 1996, p. 227. 38 M. T. Anderson, J. E. Martin, J. Odinek, P. Newcomer and 7 A. J. Rubin, Mater. Sci. Res., 1984, 17, 45. J. P. Wilcoxon, Microporous Mater., 1997, 10, 13. 8 Characterization of Porous Solids I, Studies in Surface Science and 39 A. Julbe, J. F. Quinson, A. El Mansouri and C. Guizard, in Catalysis Vol. 39, ed. K. K. Unger, J. Rouquerol, K. S. W. Sing Characterisation of Porous Solids IV, ed. B. M. McEnary, and H. Kral, Proc. of the IUPAC Symposium (COPS I), Bad T. J. Mays, J. Rouque�rol, F. Rodriguez-Reinoso, K. S. Sing and Soden, Germany, April 1988, Elsevier, Amsterdam, 1988. K. M. Unger, The Royal Society of Chemistry, Cambridge, 1997, 9 Characterization of Porous Solids II, Studies in Surface Science and pp. 519–526. Catalysis Vol. 62, ed. F. Rodrigez-Reinoso, J. Rouquerol, 40 P. Piaggio, A. Bottino, G. Capannelli, E. Carosini and A. Julbe, K. S. W. Sing and K. K. Unger, Proc. of the IUPAC Symposium Langmuir, 1995, 11, 3970. (COPS II), Alicante, Spain, May 1990, Elsevier, Amsterdam, 41 C. Balzer, A. Julbe, A. Larbot, C. Guizard, L. Cot, J. Peureux, 1991. A. Giroir-Fendler and J. A. Dalmon, in Proceedings of the Third 10 Characterization of Porous Solids III, Studies in Surface Science International Conference on Inorganic Membranes, July 1994, and Catalysis Vol. 87, ed. J. Rouquerol, F. Rodrigez-Reinoso, Worcester, USA, ed. Y. H. Ma, 3rd ICIM-Worcester Polytechnic K. S. W. Sing and K. K. Unger, Proc. of the IUPAC Symposium Institute, Worcester, 1994, p. 629. (COPS III), Marseille, France, May 1993, Elsevier, Amsterdam, 42 P. Scherrer, Nachr. Go�ttinger Ges. Dtsch., 1918, 2, 98. 1994. 43 R. H. Mikhail, S. Brunauer and E. E. Bodor, J. Colloid Interface 11 S. Lowell and J. E. Shields, in Powder surface area and porosity, Sci., 1968, 26, 45. Chapman and Hall, London, 1984. 44 G. Horvath and K. Kawazoe, J. Chem. Eng. Jpn, 1983, 16, 470. 12 G. C. Frye, A. J. Ricco, S. J. Martin and C. J. Brinker, Mater. 45 C. Tanford, in The Hydrophobic EVect, 2nd edn. Wiley- Interscience, New York, 1980. Res. Soc. Symp. Proc., 1988, 121, 349. 64 J. Mater. Chem., 1999, 9, 55–6546 T. Kunitake, Mol. Cryst. Liq. Cryst., 1994, 240, 1. 52 P. Fabre, C. Casagrande, M. Veyssie, V. Cabuil and R. Massart, Phys. Rev. Lett., 1990, 64, 539. 47 S. Acosta, A. Ayral, C. Guizard and L. Cot, J. Sol–Gel Sci. Technol., 1996, 8, 195. 53 S. H. Tolbert, A. Firouzi, G. D. Stucky and B. F. Chmelka, Science, 1997, 278, 264. 48 J. E. Blendell, H. K. Bowen and R. L. Colbe, Am. Ceram. Soc. Bull., 1984, 63, 797. 54 A. Firouzi, D. J. Schaefer, S. H. Tolbert, G. D. Stucky and B. F. Chmelka, J. Am. Chem. Soc., 1997, 119, 9466. 49 T. E. Wood, A. R. Siedle, J. R. Hill, R. P. Skarjune and C. J. Goodbrake, Mater. Res. Soc. Symp. Proc., 1990, 180, 97. 55 R. Massart, IEEE Trans. Magn., 1991, 17, 1247. 56 M. Klotz, A. Ayral, A. Van der Lee, C. Guizard, C. Menager and 50 T. Warnheim and A. Jonsson, J. Colloid Interface Sci., 1988, 125, 627. V. Cabuil, Mater. Res. Soc. Symp. Proc., to be published. 51 J. Liu, A. Y. Kim, J. W. Virden and B. C. Bunker, Langmuir, 1995, 11, 659. Paper 8/05867I J. Mater. Chem., 1999, 9, 55–6

ISSN:0959-9428    

DOI:10.1039/a805867i

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Supercritical fluid processing: a new route for materials synthesis† Franc�ois Cansell,* Bernard Chevalier, Alain Demourgues, Jean Etourneau, Christophe Even, Yves Garrabos, Vincent Pessey, Ste�phane Petit, Alain Tressaud and Franc�ois Weill Institut de Chimie de la Matie`re Condense�e de Bordeaux [ICMCB], CNRS—UPR 9048, Universite� Bordeaux I, Cha�teau Brivazac, Av.du Dr. Schweitzer, 33608 Pessac Cedex, France. E-mail: cansell@chimsol.icmcb.u-bordeaux.fr Received 23rd April 1998, Accepted 29th June 1998 Supercritical fluids exhibit a range of unusual properties that can be exploited for new reactions which are qualitatively diVerent from those involving classical solid state chemistry. After giving a brief introduction to these fluids we describe their use in inorganic chemistry and related fields. We then present two examples concerning diVerent areas of solid state chemistry: (i) the formation of novel inorganic nanoparticles; (ii) the preparation of new open-structure oxy(hydroxy)fluorides, thus showing the advantages of this supercritical fluid processing that can be seen as an alternative method to regular solution chemistry or solid–gas reactions.Thus, it is possible to prepare new materials under mild Introduction conditions by using either physical or chemical transformations A fluid is in the supercritical domain when both pressure and (see Fig. 2). The main results concerning these two possibilities temperature are above their critical values, Pc and Tc. In are reported herein. However, first of all, it should be pointed practice, this definition is restricted to fluids close to their out that supercritical fluids are used in order to provide critical points and, hence, with their density close to rc .There contaminant-free compounds. Moreover, this method allows are numerous fluids for which the critical temperature is one to obtain solid particles exhibiting both a narrow range moderate.A selection of the most usual ones is listed in of size distribution (micronic and nanometric) and specific Table 1.1 features (e.g. well defined stoichiometry, structure, etc.). Nowadays supercritical fluids are widely used in many fields.2–4 These fluids are very attractive for materials pro- Physical transformations cessing5 and more particularly for the formation of particles, fibres, thin films (pharmaceuticals, explosives, coatings)6–9 and Concerning the physical transformations, three main techfor drying materials (highly porous gels).10,11 Furthermore, niques are successively presented: the rapid expansion of a these fluid processes are already used in many other appli- supercritical solution (RESS), supercritical antisolvent precipications, such as: separation (petroleum-chemistry separation tation: gas anti-solvent (GAS), and precipitation with comand purification, food industry);12–15 chromatography (analyt- pressed anti-solvent (PCA).ical and preparative separation, determination of physicochemical properties);16,17 chemical reactivity ( low-density polyethylene, waste destruction, polymer recycling);18–22 and Table 1 Critical coordinates of usual pure fluids.Tc , Pc and rc are earth sciences (volcanism, geothermal energy, hydrothermal the critical temperature, pressure and density, respectively1 synthesis).23 Both the capability of some supercritical fluids to replace Fluid Tc/°C Pc/MPa rc/kg m-3 toxic industrial solvents and the ability of tuning solvent characteristics for highly specific separations or reactions lead Carbon dioxide 31.2 7.38 468 Ammonia 132.4 11.29 235 to the current industrial and scientific interest in supercritical Water 374.1 22.1 317 fluids.24 These fluids possess physicochemical properties, such Ethylene 9.5 5.06 220 as density, viscosity and diVusivity, which are intermediate Ethane 32.5 4.91 212 between those of liquids and gases (Table 2). The main interest Propane 96.8 4.26 225 in supercritical fluids as reaction media relies on their continu- n-Pentane 196.6 3.37 232 ously adjustable properties from gas to liquid with small Cyclohexane 279.9 4.03 270 Methanol 240.0 7.95 275 pressure and temperature variations.Fig. 1 shows that it is Ethanol 243.1 6.39 280 possible to move continuously from the liquid to the gaseous Isopropanol 235.6 5.37 274 state without crossing the liquid–gas equilibrium line. Acetone 235.0 4.76 273 Classically, materials preparation in supercritical fluid can be described in three steps: step one: solubilization of solutes in dense fluid (close to the liquid phase) since the solute solubility varies generally as a power law with respect to Table 2 Characteristic magnitudes of thermophysical properties of density;25 step two: chemical or physical transformation of fluids fluids in the supercritical domain, characterized by specific properties (Table 2);26,27 step three: materials recovery in the Liquid Supercritical fluid Gasa low density fluid domain (close to the gas phase) in order to r/kg m-3 1000 100–800 1 obtain dry materials.g/Pa s 10-3 10-4–10-5 10-5 Db/m2 s-1 10-9 10-8 10-5 †Basis of the presentation given at Materials Chemistry Discussion r=density; g=viscosity; D=diVusion coeYcient. aAt ambient con- No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, ditions. bFor small-molecule solute. France. J. Mater. Chem., 1999, 9, 67–75 67remains soluble in the organic solvent.These processes are used for explosives, polymers, food, pharmaceutical compounds, coloring matter and catalysts. For the latter it can be specified that zinc acetate nanoparticles down to about 30 nm are obtained with an average particle size of 50 nm.33 With the GAS process, a batch of the solution (solute dissolved in organic solvent) with dense carbon dioxide for example, is expanded several times in a vessel.Because of the lower solvent strength of the carbon dioxide–expanded solvent mixture, the fluid medium becomes supersaturated, thus forcing the solute to precipitate in microparticles. The PCA process consists of spraying the solution through a nozzle into dense carbon dioxide, for example. These two processes allow the control of the particle size (micronic and submicronic).Unfortunately their use is limited Fig. 1 Schematic representation of a phase diagram illustrating the by the separation of the solid particles from the solvents. density evolution from the liquid to the gas without crossing the T–C line which corresponds to the liquid–gas equilibrium line, C being the critical point and T being the triple point.Pc and Tc are the critical pressure and temperature respectively. Chemical transformations Physico-chemical properties of supercritical fluids as a reactive medium are relatively easy to adjust by operating conditions. Indeed, these tunable characteristics, such as density, viscosity, and diVusivity, influence directly the reaction rate constants. For the reactions involving ionic species, the variations of the relative permittivity and of the ion products influence both the chemical equilibrium and the evolution of the transition state.By moving continuously from the liquid to the gaseous state, the eYciency of the reaction may be significantly enhanced to create a completely new chemical process. As the reactivity of organic compounds in supercritical fluids has already been intensively studied,34,35 we focused on materials preparation.Supercritical fluids are used for matter transport or as chemical media in order to produce submicronic particles, compounds with specific features, polymer–metal composites, aerogels and thin films. Chemical reactions generally involve the thermal decomposition of a metal-containing precursor.The various applications developed in the field of inorganic materials are also reported. Fig. 2 Possible routes for material elaboration using supercritical fluids. RESS=rapid expansion of supercritical solution; GAS=gas Oxide nanoparticles anti-solvent recrystallisation; PCA=precipitation with compressed anti-solvent. Submicronic TiO2 powder has been obtained in a flow reactor by a sol&ndass.Titanium isopropoxide, Ti(O-iC3H7)4, is Rapid expansion of supercritical solution (RESS) process hydrolytically decomposed by water, produced by the catalytic dehydration of the isopropyl alcohol used as the supercritical In this process, the reactant is first solubilized in a supercritical solvent.36 Titanium hydroxides are formed and their decompo- fluid which is then expanded across a nozzle or a capillary at sition into TiO2 kinetically limits the reaction.At temperature a very high velocity. Such a rapid expansion leads to a high and pressure around 300 °C and 10 MPa respectively, particles solute supersaturation and subsequently to the precipitation with a narrow range of size distribution (20–60 nm), some of of small monodispersed particles.28 This process has been which crystallize with the anatase structure, are weakly associ- applied to inorganic,29 organic30 and pharmaceutical31 comated into spherical agglomerates (500–2000 nm).A model of pounds. It allows the preparation of either an intimately mixed the flow reactor, has been made including hydrodynamic and powder of two inorganic compounds (e.g.SiO2 and KI) or an kinetic studies,37 and experimentally validated for optimizing inorganic and organic combination [e.g. KI and poly(vinyl the process. chloride)].32 Metal oxide submicronic particles can also be obtained in The major limitation of the RESS process is the low supercritical water by the hydrolysis of metal salts in aqueous solubility of many compounds in supercritical fluids.For solutions. Numerous oxides have been studied such as Fe2O3 , example in carbon dioxide, the solute solubilities are about Co3O4 , NiO, ZrO2, TiO2 and CeO2.38 In the case of TiO2, 0.01 wt.% or less. The other challenge for large commercial TiCl4 was used as the precursor and leads, at 30 MPa and plant configuration is associated with nozzle design in order 400–450 °C, to the formation of TiO2 (anatase) with a prismatic to avoid particle accumulation and the freezing caused by the shape and a size of 20 nm.39 rapid expansion.Porous substrates (alumina or zirconia) with high surface areas have been impregnated with supercritical solutions of Supercritical anti-solvent precipitation processes: gas antimetal salts (carbonate, acetate, etc.), which can be adsorbed solvent (GAS) and precipitation with compressed anti-solvent or deposited as a film on the surface of the substrate by (PCA) lowering the pressure.40 Supercritical fluid has also been used as a drying medium In these processes a solute is dissolved in an organic solvent.The addition of a supercritical fluid leads to a lower solubility in the sol–gel process for the preparation of nanometric tin oxide particles (3–5 nm).41 of the solute in the liquid phase since the supercritical fluid 68 J.Mater. Chem., 1999, 9, 67–75Metal nanoparticles expansion of the supercritical fluid causes the vaporisation of the solutes. The precursors can be decomposed at or near a Micronic metallic powders (as pure metals, metal alloys, or substrate surface to form a thin film.This leads to the mixtures thereof ) have been prepared from metal salts.42 formation of pure metal (e.g. Al, Cr, Ni, Cu, etc.), or oxide Precursors are solubilized in methanol at a temperature lower (e.g. CuO, SiO2, Cr2 O3, etc.) films. Moreover, the stoichiothan that employed to generate a metallic powder. As an metric ratios of the precursors for multi-element coatings (e.g.example, micronic Cu particles were obtained from a solution YBa2Cu3O7-x) can be precisely controlled.46–48 of Cu(OAc)2 at 275 °C and about 14 MPa. When Pd(OAc)2 After this brief review, we present our own results concerning was added to a methanol solution of Cu(OAc)2, a mixture of a new route for nanoparticle formation and for the synthesis Pd and Cu metals and a face centered cubic Cu–Pd alloy were of new phases in solid state chemistry such as open-structure obtained; both systems were identified by XRD and ESCA oxy(hydroxy)fluorides. measurements.Compounds with specific features Preparation of nanoparticles Semiconducting quaternary antimony sulfides, such as The aim of this work is to develop a new method for the KAg2SbS4 or RbAg2SbS4 , have been prepared in supercritical preparation of nanoparticles.Indeed, metallic nanoparticles ammonia at 160 °C and about 22 MPa during 4 days from the with dimensions of tens of nanometers or less exhibit size- appropriate mole ratio of K2S4, K2CO3 or Rb2CO3 with Ag, dependent electrical, chemical and magnetic properties. The Sb2S3 and S8.The use of supercritical ammonia leads to high chemical transformation of metallic precursors inside a quality crystal growth, owing to its specific transport properties supercritical fluid is a new route for obtaining nanometric (viscosity, solubility, etc.).43 homogeneous powders. The process mainly deals with the dissolution of precursors in a dense fluid, followed by the Polymer–metal composites decomposition of the precursors in supercritical fluid for Polymer–metal composites consist of isolated nanoclusters of preparing either powders or thin films of materials.metal atoms distributed homogeneously throughout polymeric substrates. For example, ametal precursor such as dimethyl(cy- Experimental setup cloocta-1,4-diene)platinum(II ) is dissolved in supercritical carbon dioxide and impregnated into the polymer [poly(4- The apparatus used for chemical synthesis in supercritical fluid methylpent-1-ene)]. Thermolytic reduction of the precursor in media is shown in Fig. 3. The system itself, placed in a the CO2-swollen polymer medium at 140 °C and under 26 MPa temperature controlled hot-air oven, is composed of: (i) two pressure produces platinum aggregates with a maximum diam- high pressure 316 stainless steel cells.The first one (cell A) is eter of approximately 50 nm.44 the mass exchanger. It is a cylinder of 1.8 cm diameter and 5.5 cm length filled with a glass packed bed of 0.2 cm diameter. Aerogels A glass wool plug is placed at the outlet of the cylinder in order to avoid the dragging of the precursor by the flowing Sol–gel reactions have been directly performed in supercritical fluid. For each experimental run the cylinder is charged with carbon dioxide as a reactive medium by using formic acid as 1 g of the precursor.The second cell (cell B) is the reactor. It condensation reagent.45 Supercritical carbon dioxide appears consists of a cylinder of 1.8 cm diameter and 9.5 cm length.to be a better solvent than ethanol for preparing phenylene- An external heating resistor allows the generation of a flash bridged polysilsesquioxane aerogels from 1,4-bis(triethoxysiof temperature. A glass wool plug is placed at the outlet of lyl )benzene. In addition, this process can be successfully used the cylinder for the reasons mentioned above. A polytetra- to generate monolithic aerogels from their precursors in a fluoroethylene film (PTFE, 0.25 mm thickness) rolled up inside single step.the cell is used for collecting the metallic particles. (ii) A high pressure generator, which is a fluid metering pump (CM 3200 Thin films P/F). This pump provides either constant flow or constant pressure. Only one fluid can be used with this system (in our Supercritical fluid transport and chemical deposition can be used for the formation of thin films on a substrate.This film case it is ammonia). The working pressure is monitored by a digital gauge. (iii) Two thermocouples measure the tempera- deposition technique can use nonvolatile precursors, unlike the chemical vapor deposition process. In supercritical fluid ture, one is placed inside the hot-air oven, the other one is inside cell B.transport and chemical deposition technique, precursors are dissolved in the supercritical fluid. Then, this solution flows Acetylacetonates of metals [Cu(AcAc)2 and Fe(AcAc)3] were used as precursors. They were all provided by Sigma- through a restrictor into a deposition chamber where the rapid Fig. 3 Apparatus for chemical synthesis in supercritical fluid media. 1, Fluid tank. 2, Cooling system. 3, High pressure pump: 1–10 ml min-1; 0–40 MPa. 4, Heat exchanger. 5, Mass exchanger (cell A). 6, Reactor (cell B). 7, Temperature regulator. 8, Contact breacker. 9, Micrometric valve. 10, Solvent trap. 11, Oven. J. Mater. Chem., 1999, 9, 67–75 69Aldrich and used without further purification. Their solubility in ammonia was measured and is reported in Table 3.The solid particle formation results from the solute thermal decomposition in the supercritical fluid medium. The fact that the critical temperature of this solvent must be lower than that of the metal chelate decomposition justifies the choice of ammonia as the supercritical fluid. In addition this fluid exhibits a high reducing power.A known amount of precursor is placed in cell A. The system is filled with solvent. Then the hot-air oven is heated at a temperature TA which allows good solute solubility. Cell B is heated at temperature TB which allows chemical reaction. When both temperatures are stabilized, the pressure is increased up to 17 MPa and then maintained constant. In these conditions ammonia is liquid in the mass exchanger (cell A) and is in the supercritical domain in the reactor (cell B).Then the fluid flow rate is stabilized at the working value of 3 ml min-1 and the fluid mixture (solvent+precursor) is transported to the reactor. The solute is thermally decomposed Fig. 4 SEM picture of the mixture of Fe4N and Fe2O3 obtained by in the reactor and the metallic part is precipitated on the the decomposition of Fe(AcAc)3 in supercritical ammonia.(Treactor= PTFE film. The organic part, still soluble in the supercritical 180 °C; P=16 MPa). solvent, is trapped at the outlet of the installation. Results and discussion Chemical syntheses in supercritical fluids are known to allow the obtention of powders with small particle size and narrow distribution range.Indeed, these syntheses are based on homogeneous nucleation and large supersaturation in a supercritical solvent and the size of the collected particles is very sensitive to the supersaturation: the higher the supersaturation, the smaller the particles.49 Moreover, in a supercritical fluid small changes of pressure give rise to large changes in both density and solubility.50 Thereby the optimization of working pressure and temperature in the reactor allows the achievement of high supersaturation and the formation of nanosized particles.This is demonstrated by our results, obtained from the thermal decomposition of Fe(AcAc)3 made in the dense fluid domain and in the supercritical domain. Obviously, supercritical fluids as reactive media allow the formation of smaller particles (with waverage=50 nm, see Fig. 4) than those obtained in dense fluids (with waverage=800 nm, see Fig. 5). The experiments have shown that the final particle size depends on the variations of Fig. 5 SEM picture of the mixture of Fe4N and Fe2O3 obtained by the working conditions (T, P, solute supersaturation, hydrothe decomposition of Fe(AcAc)3 in liquid ammonia.(Treactor=120 °C; dynamics, etc.). The relationship between size and working P=10 MPa.) conditions is under study. A solvolysis reaction occurs during the decomposition of the precursor. Indeed, the particle reactivity is very high and Table 4 Cu and Fe nitride formation in supercritical ammonia thus the use of ammonia may lead to metal nitrides. In the Precursor Working conditions Products case of Cu(AcAc)2 and Fe(AcAc)3 precursors, we have prepared particles of copper nitride Cu3N at 200 °C and 17 MPa Cu(AcAc)2 Tmass exchanger=90 °C Cu3 N with copper as and iron nitride Fe4N at 180 °C and 16 MPa (Table 4).The Treactor=200 °C impurity sample of Cu3N has been characterized by XRD. The pattern P=17 MPa indicates that the Cu3N phase is obtained with copper as Ammonia flow=3 ml min-1 Fe(AcAc)3 Tmass exchanger=90 °C Mixture of Fe4N and impurity (Fig. 6). In the case of the Fe(AcAc)3 decomposition, Treactor=180 °C Fe2 O3 a mixture of both iron nitride Fe4N and iron oxide Fe2O3 was P=16 MPa observed (Table 4). This mixture has been characterized by Ammonia flow=3 ml min-1 electron diVraction because this sample seems to be amorphous to X-ray diVraction (Fig. 7, Table 5). Aggregates of 50 nm are obtained in the case of the Fe4N–Fe2O3 mixture. They consist of very small particles Table 3 Solubility (at 90 °C and 10 MPa) and decomposition tempera- (10 nm) with a narrow range of size distribution as shown ture (at 16 MPa) values for Cu(AcAc)2 and Fe(AcAc)3 in ammonia in Fig. 8. The iron oxide formation could be due to the high (the critical coordinates of ammonia are Tc=132.4 °C and Pc= 11.29 MPa) reactivity of iron with oxygen present in the acetylacetonate groups.New precursors free from oxygen are under study in Solubility/g g-1 a Decomposition temperature/°C order to obtain metallic nanoparticles only. It must be pointed out that our working temperatures for Cu(AcAc)2 5.8×10-3 190 obtaining metal nitrides are clearly lower than those used in Fe(AcAc)3 3.8×10-3 170 classical solid state chemistry routes.Classical ways for making aGram of precursor per gram of fluid. transition-metal nitrides involve high-temperature and high- 70 J. Mater. Chem., 1999, 9, 67–75Table 5 d-Values obtained for mixture of Fe4N and Fe2O3 (see Fig. 7) dmeasured/A° dFe4N/A° a dFe2O3/A° a 6.47 4.76 4.76 4.01 3.79 3.34 3.24 3.15 2.99 2.981 2.804 2.728 2.684 2.55 2.573 2.548 2.458 2.372 2.295 Fig. 6 X-Ray diVraction pattern of Cu3N obtained by the decompo- 2.243 sition of Cu(AcAc)2 in supercritical ammonia (Treactor=200 °C; P= 2.19 2.191 17 MPa.) *indicate the peaks relating to copper. 2.176 2.018 1.997 1.983 1.973 1.897 1.855 1.807 1.75 1.735 1.697 1.66 1.666 1.58 1.53 1.549 1.31 1.342 1.25 1.265 1.200 1.12 1.144 1.095 1.06 1.053 1.01 1.014 0.98 0.949 0.88 aFrom ASTM files. tions, the electrical properties of the catalyst constitute one of the key features, in particular when the materials show mixed conductivity, i.e.fast transport of ions in addition to electronic conductivity at various temperatures. In the case of oxides, for instance, ionic conductivity is generally due to polarizing cationic species such as H+ and Li+, or polarisable anionic O2-/O- entities.The aim of this work was then to elaborate and to study some (O2-, OH-, F-) mixed-anion systems showing good mixed conductivity, such as Sr(OH)X (X=Cl, Br, I )54 phases, which are known as protonic conductors. From a chemical point of view, it was interesting to use Fig. 7 Electron diVraction pattern of the mixture of Fe4N and Fe2O3 transition metal fluorides as starting materials, i.e. compounds obtained by the decomposition of Fe(AcAc)3 in supercritical ammonia which already contain ionic MMF bonds, instead of substitut- (Treactor=180 °C; P=16 MPa). (See Table 5.) ing oxy- or hydroxy-anions by fluorine in oxides or hydroxides exhibiting strongerMMO (OH) bonds as was done in previous work.55,56 From a structural point of view, and in order to pressure reactions between a pure metal and a nitrogen source.51 However, there are other ways to obtain metal produce good catalytic properties, we were rather interested in open structures based on transition metal networks, such nitrides, based on the use of precursors (such as organometallics, acetates, salts etc.).In this case, the working conditions as tunnelled or layered frameworks, to enhance the ionic conductivity and to favor electronic conductivity. are lowered. Indeed, single crystalline c¾-Fe4N can be obtained by chemical transport from iron over [Fe(NH3)6]I2] in A suitable method would then consist of partial hydrolysis in the case of a transition metal fluoride with the desired supercritical ammonia ( p(NH3)=600–800 MPa, 460 °CT 580 °C)].52 By heating CuF2 at 280 °C in NH3, pure Cu3N is oxidation state, such as FeF3, whereas oxidative hydrolysis would occur for compounds with lower oxidation states, such obtained.53 as FeF2.We could also consider ternary fluorides such as NH4FeF3 from which ammonium ions could be topotactically Open-structure oxy(hydroxy)fluorides deintercalated, thus entailing the replacement of F- by OH-/O2- ions in an aqueous medium, in order to obtain an During the past few years, much work has been devoted to the study of heterogeneous catalytic systems.In many reac- oxyfluoride or an oxy(hydroxy)fluoride. J. Mater. Chem., 1999, 9, 67–75 71obtained a higher fluorine content; however it was limited to e<0.3.A method for preparing oxy(hydroxy)fluorides with the hollandite-type structure which would lead to a much higher fluorine content, involves the oxidation of transition metal fluorides with H2O2 solutions. Using this route, we could expect that F- and OH- would occupy the same type of sites within the network, at the corners of the FeX6 octahedra, instead of being present only as inserted species in the tunnels of the framework.In this connection we describe here the oxidative hydrolysis of the fluoroperovskite NH4FeF3 under supercritical conditions. Experimental setup Our experiments were carried out in scCO2; a second high pressure pump was added for the injection of the solution (H2O2–H2O–ethanol ) at temperatures between 50 and 250 °C approximately, and at pressures varying from 10 to 40 MPa.The reactant used was a concentrated aqueous solution of hydrogen peroxide (H2O2, 30M), diluted in an equal volume of ethanol. The stainless steel reactor (5 mm diameter, 50 mm long) (cell A of Fig. 3) was prepared as described above and filled with freshly ground NH4FeF3 powder in a dry argon glove box.It was subsequently mounted on the pressure line in a programmable hot-air oven. The line was then flushed with CO2 before raising the temperature and starting the reaction. During the treatment, great care should be taken in keeping the system in supercritical conditions in order to avoid water separation: water solubility in scCO2 remains low, and the introduction of large quantities of water would lead to a twophase system.Fig. 8 TEM picture of the mixture of Fe4N and Fe2O3 obtained by After the reaction, the reactor was flushed again with scCO2 the decomposition of Fe(AcAc)3 in supercritical ammonia (Treactor= in order to eliminate residual water. 180 °C; P=16 MPa). The compounds obtained were characterized by electron probe microanalysis (EPMA) (O/Fe and F/Fe ratios), TGA–DTA coupled with mass spectrometry (adsorbed H2O, Unfortunately, several drawbacks arise from these methods: OH and F contents), diVuse reflectance-IR, electron first, fluoride materials are generally sensitive to moisture; microscopy (morphology of the particles) and X-ray secondly, moderate temperatures, necessary to obtain open- diVraction.structure materials, generally lead to low diVusion of the reactive species (gaseous or liquid ) through the solid fluoride Results and discussion material, and then induce a gradient in the chemical composition of the final product. The oxidative process using the water–ethanol solution of H2O2 in scCO2 on NH4FeF3 allowed us to prepare oxy(hyd- Therefore, we have chosen to use a supercritical medium which allows very high diVusivity of the species through the roxy)fluorides with the general formula FeOF1-x(OH)x nH2O (x0.2, n0.4), isostructural with b-FeOOH, i.e. exhibiting transition metal fluorides and good elimination of water.Supercritical CO2 (scCO2) with ethanol as a co-solvent the hollandite-type structure (Fig. 9b). The frameworks of the rutile-type and the hollandite-type structures are built of chains appeared to be a good candidate.In order to demonstrate the interest and capability of this of octahedra sharing opposite edges (Fig. 9). In the case of the rutile structure, single chains are connected to one other method for preparing 3d-transition metal oxyfluorides, we focused our attention on iron-based compounds, owing to the by vertices, whereas in the hollandite structure double chains sharing edges form large square tunnels of about 2×0.5 nm well known stability of the two oxidation states +II (3d6) and +III (3d5) present in iron oxides and fluorides.In addition, section and running along the c direction. The small amount of water which is present in the compounds mainly depends several attempts were made over the last few decades to prepare iron oxyfluorides with open frameworks.57 It should on the conditions of the final CO2 treatment, performed after the reaction.be pointed out that only the rutile form of FeOF (Fig. 9a) has been obtained so far, by high temperature reaction.58,59 After optimization of the experimental supercritical conditions, NH4FeF3 was treated at 150 °C for 30 min under a Numerous works have also been reported concerning an iron oxyhydroxide exhibiting an open structure (hollandite scCO2 pressure of 30 MPa.The volume of the injected H2O2 solution (as described above) was 15 ml. The X-ray diVraction type): b-FeOOH.60,61 This compound was obtained by hydrolysis of FeCl3 .62–64 It has been shown that residual Cl- pattern shows a well crystallized single phase with a hollanditetype structure (sample A; Fig. 10a). The tetragonal cell param- anions are always present in the tunnels of the framework, keeping it from collapsing.65 Some attempts have been made eters (a=1.036 nm, c=0.302 nm) are slightly smaller than those reported for b-FeOOH57 (a=1.053 nm, c=0.303 nm). to prepare the corresponding fluorinated phase FeO(OH)1-eFe via the hydrolysis of FeF3.The material contained only a few The lower a-value could be related to the presence of F- anions instead of OH- in the framework. atom% of F- anions which were located in the tunnel sites of the structure.60 Other authors tried to synthesize such com- We have also prepared, using the standard hydrolysis route, two other compounds also exhibiting the same hollandite pounds by the hydrolysis of iron hydroxides56 or nitrates55,66 in a concentrated NaF or NH4F aqueous solution, and structure: a slightly chlorinated hydrated iron oxyhydroxide 72 J.Mater. Chem., 1999, 9, 67–75Fig. 9 a) Rutile-type FeOF and b) open-structure hollandite-type ‘FeOF nH2O’ prepared in supercritical medium (sample A).Fig. 10 Powder XRD patterns of a) hollandite-type FeOF1-x(OH)x nH2O prepared in supercritical medium (sample A) and b) slightly fluorinated b-FeOOH prepared by the classical hydrolysis route (sample C). (sample B), by hydrolysis of FeCl3 as reported earlier;62–64 by DTA–TGA (under flowing nitrogen) coupled with mass spectrometry. The amount of water molecules depends on the and a slightly fluorinated hydrated iron oxyhydroxide (sample C), by hydrolysis of FeF3 .60 This latter type of compound conditions of the final CO2 flow and varies from 0 to 0.4.In the case of sample A obtained as described above, the composi- reveals a much lower crystallinity than those prepared in supercritical conditions, as shown in Fig. 10b. tion is FeOF0.9(OH)0.1 . The FeOF1-x(OH)x nH2O compounds were checked by In order to check the eVectiveness of the oxidizing reaction and the final oxidation state of iron, samples obtained by infrared spectroscopy (diVuse reflectance), as well as hydrated b-FeO(OH,Cl ) (sample B), as shown in Table 6.supercritical treatment were treated afterwards under flowing HF at moderate temperatures (40<T<150 °C) for 2 hours.The spectrum of sample B shows the characteristic bands of slightly chlorinated b-FeOOH, as reported earlier,57,61,67 in This type of fluorination reaction is known to be non-oxidative, and leads to an exchange of O2- and OH- anions by F-. particular those of the deformation mode of the FeKOKFe bonds (n=1070–1030 cm-1), and the asymmetric and sym- The phase obtained after reaction with HF corresponds to rhombohedral FeF3 (ReO3-derived phase), as characterized metric vibrations of FeKOKFe bonds at n=700 and 420 cm-1, respectively.Large quantities of water are also detected (n= by XRD and EPMA. This result confirms that all Fe2+ from NH4FeF3 had been fully oxidized to Fe3+ during the reaction 3400–3300 cm-1, 1620 cm-1). The spectra of the oxy(hydroxy)fluorides prepared in this with H2O2 in supercritical CO2–H2O–ethanol.The atomic ratios O/Fe and F/Fe have been estimated by work diVer from those of chlorinated b-FeOOH: in addition to the FeKOKFe vibrations mentioned above and located at EPMA and the fluorine and water contents have been obtained J. Mater. Chem., 1999, 9, 67–75 73Table 6 Infared band assignments for FeOF1-x(OH)x nH2O and for slightly chlorinated iron oxyhydroxide (sample B) obtained by conventional hydrolysis of FeCl3 FeOF1-x(OH)x nH2O Hydrated b-FeO(OH,Cl) (sample B) Type of n/cm-1 Intensity (a.u.) vibration n/cm-1 Intensity (a.u.) 3500–3400 s n(OH) 3480–3360 vs 3180 s Characteristic of fluorinated compounda — — 1619 mw d(HKOKH) 1615 s 1080 w L Deformation modes of FeKO(H)KFe61 1070 vw 1020 K 1030 w m 956 m Characteristic of fluorinated compounda — — 845 w Interaction of H2O with hollandite tunnels61 845 m 705 ms Asymmetric vibration of FeKOKFe in hollandite 650–700 ms 424 w Symmetric vibration of FeKOKFe in hollandite 420 vw Intensities: vs=very strong; s=strong; m=medium; w=weak; vw=very weak.aThese bands were assigned in previous works57,61 to the signature of the presence of fluorine species in the structure. n=1080–1020, 705, 425 cm-1, two strong bands appear at n= CO2 medium) through the NH4FeF3 network coupled to the eYcient elimination of water allows the synthesis of new types 3180 and 956 cm-1.They were also found with much lower intensity in slightly fluorinated b-FeOOH,57 and seem charac- of compounds which adopt the hollandite-type structure with a framework exhibiting large tunnels; (ii) the phase corre- teristic of fluorinated oxy-compounds.The amount of water present in the tunnels of sample A, which is in any case smaller sponding to the FeOF0.9(OH)0.1 chemical composition shows a high fluorine content compared to fluorinated iron oxyhyd- than for sample B obtained via conventional hydrolysis, depends on the final CO2 treatment.This may be related to roxides prepared by conventional routes. The use of this method is now in progress for preparing the fact that flushing the reactor with the supercritical CO2–ethanol mixture helps to eliminate residual water.61 This other metastable phases with various transition metals, such as lamellar structures containing large interlayer spaces.is evident by the lowering of the intensity of the band at 845 cm-1. We are grateful to C. Chabot, L. Rabardel, E. Sellier and Electron diVraction and Rietveld refinement of powder XJ. J. Videau for helpful discussions and technical support. ray diVraction data are in progress in order to elucidate the structural parameters of these new oxy(hydroxy)fluorides.A Mo�ssbauer spectroscopy study will specify the oxidation state References and the environment of iron. These results will be presented 1 N. B. Vargaftik, Table on the thermophysical properties of liquids in a forthcoming paper. and gases, Hemisphere Publishing Corporation, London, 1975. 2 Innovations in Supercritical Fluids, ed. K. W. Hutchenson and N. R. Foster, ACS Symp. Ser. 608, Washington DC, 1995. Conclusions 3 Supercritical Fluids—Fundamentals for Applications, ed. E. Kiran It has been pointed out that chemical synthesis in supercritical and J. M. H. Levelt Sengers, NATO ASI Series E273, Kluwer, Dordrecht, 1994. media is a new route for preparing either nanometric powders 4 Fluides Supercritiques et Mate�riaux, ed. F. Cansell and J. P. with narrow size distribution or oxy(hydroxy)fluorides exhibit- Petitet, AIPFS Publishing, Nancy, France, 1995. ing open frameworks. 5 X. Y. Zeng, Y. Arai and T. Furuya, Trends Chem. Eng., 1996, As far as the formation of nanoparticles is concerned the 3, 205. following conclusions may be drawn. (i) The use of supercriti- 6 P. G. Debenedetti, Supercritical Fluids—Fundamentals for cal fluids as reactive medium allows control of the particle size Application, ed.E. Kiran and J. M. H. Levelt Sengers, NATO ASI Series E273, Kluwer, Dordrecht, 1994, pp. 719–729. distribution owing to the tunable density and solubility of the 7 H. Ksibi and P. Subra, Adv. Powder Technol., 1996, 7, 21. solute. Then the solute supersaturation can be continuously 8 K. Chhor, J. F. Bocquet and C. Pommier, Mater.Chem. Phys., adjusted in order to control the size of the nanoparticles. The 1995, 40, 63. systematic variation of working conditions (T, P, solute 9 P. Beslin, P. Jestin, Ph. Desmarest, R. Tufeu and F. Cansell, supersaturation, hydrodynamics, etc.) and their eVects on the Supercritical Fluids—Reactions, Material science and particle size are under study. Thus, metallic nanoparticles Chromatographiy, ed. M.Perrut and G. Brunner, AIPFS Publishing Nancy, France, 1994, vol. 3, pp. 321–326. prepared by this route should be very attractive for their size- 10 R. A. Laudise and D. W. Johnson, J. Non-Cryst. Solids, 1986, dependent electrical, chemical, and magnetic properties. (ii) 79, 155. The nature of the final product depends on the nature of the 11 B.Rangarajan and C. T. Lira, J. Supercritical Fluids, 1991, 4, 1. solvent. The metallic atoms can easily react with ammonia, 12 A. Akgerman and G. Madras, Supercritical Fluids—Fundamentals giving nitrides, Cu3N and Fe4N, at temperatures clearly lower for Application, ed. E. Kiran and J. M. H. Levelt Sengers, NATO than those used in conventional solid state chemistry routes. ASI Series E 273, Kluwer, Dordrecht, 1994, pp. 669–695. 13 Supercritical Fluid Extraction: Principles and Practice, ed. M. A. (iii) The nature of the final product depends also on the nature McHugh and V. J. Krukonis, Butterworths, Stoneham, MA, 1988. of the metallic precursors (e.g. the formation of Fe2O3 results 14 F. Cansell, Ph. Botella, Y. Garrabos, J. L. Six, Y. Gnanou and from the high reactivity of iron with oxygen present in R.Tufeu, Polym. J., 1997, 29, 910. acetylacetonate groups). In this case the preparation of pure 15 E. J. Beckman, Nature, 1996, 271, 613. metallic nanoparticles should require the use of metallic precur- 16 S. H. Page, J. F. Morrison and M. L. Lee, Supercritical Fluids— sors free from oxygen. (iv) The supercritical chemistry route Fundamentals for Application, ed.E. Kiran and J. M. H. Levelt Sengers, NATO ASI Series E273, Kluwer, Dordrecht, 1994, is interesting not only for obtaining either intimate homopp. 641–652. geneous mixtures of nanoparticles, (e.g. Fe2O3+Fe4N) or 17 G. M. Schneider, Fluid Phase Equilibr., 1983, 10, 141. metallic alloys (e.g. Cu–Pd), but also for the metallic coating 18 M. PoliakoV, M.W. Georges and S. M. Howdle, Chemistry under of particles used as substrates and therefore insoluble in the extreme or non-classical conditions, ed. R. V. Eldik and C. D. supercritical fluid medium. Hubbard, Wiley, New York, 1996, pp. 189–219. As far as the preparation of metastable phases is concerned: 19 P. Beslin, F. Cansell, Y. Garrabos, G. Demazeau, B. Berdeu and D. Sentagnes, De�chets, 1997, 5, 17.(i) the high diVusivity of O2-/OH- species (in supercritical 74 J. Mater. Chem., 1999, 9, 67–7520 D. T. Chen, A. P. Craig, E. Reichert and J. Hoven, J. Hazard. 40 D. F. McLaughlin and M. C. Skriba, US Pat. 4,916,108, 1990. Mater., 1995, 44, 53. 41 F. Lu, S. Chen and S. Peng, Catal. Today, 1996, 30, 183. 21 H. Lentz and W. Mormann, Makromol. Chem., Macromol.Symp., 42 J. N. Armor and E. J. Carlson, US Pat. 4,615,736, 1986. 1992, 57, 305. 43 G. L. Schimek, W. T. Pennington, P. T. Wood and J. W. Kolis, 22 K. P. Johnston, K. L. Harrison, M. J. Clarke, S. M. Howdle, J. Solid State Chem., 1996, 123, 277. M. P. Heitz, F. V. Bright, C. Carlier and T. W. Randolph, Nature, 44 J. J.Watkins and T. J. McCarthy, Chem. Mater., 1995, 7, 1991. 1996, 271, 624. 45 D. A. Loy, E. M. Russick, S. A. Yamanaka and B. M. Baugher, 23 J. P. Petitet, Fluides Supercritiques et Mate�riaux, ed. F. Cansell Chem. Mater., 1997, 9, 2264. and J. P. Petitet, AIPFS Publishing, Nancy, France, 1995, 46 B. N. Hansen, B. M. Hyberston, R. M. Barkley and R. E. Sievers, pp. 251–300. Chem. Mater., 1992, 4, 749. 24 C. A. Eckert, B. L. Knuston and P. G. Debenedetti, Nature, 1996, 47 O.A. Louchev, V. K. Popov and E. N. Antonov, J. Cryst. Growth, 383, 313. 1995, 155, 276. 25 J. Chrastil, J. Phys. Chem., 1982, 86, 3016. 48 V. K. Popov, V. N. Bvili, E. N. Antonov and D. A. 26 P. E. Savage, S. Gopalan, T. I. Mizian, C. J. Martino and E. E. Lemenovski, Thin Solid Films, 1996, 279, 66. Brock, AIChE J., 1995, 41, 1723. 49 P. G. Debenedetti, AIChE J., 1990, 36, 1289. 27 F. Cansell, S. Rey and P. Beslin, Rev. Inst. Fr. Pe� trole, 1998, 53, 71. 50 S. K. Kumar, J. Supercritical Fluids, 1988, 1, 15. 28 J. W. Tom and P. G. Debenedetti, J. Aerosol Sci., 1991, 22, 555. 51 D. V. Baxter, M. H. Chisholm, G. J. Gama and V. F. DiStasi, 29 D. W. Matson and R. D. Smith, J. Am. Ceram. Soc., 1989, 72, 871. Chem. Mater., 1996, 8, 1222. 30 H. Ksibi, P. Subra and Y. Garrabos, Adv. Powder Technol., 1995, 52 H. Jacobs and J. Bock, J. Less-Common Met., 1987, 134, 215. 6, 25. 53 R. Juza and H. Hahn, Z. Anorg. Allg. Chem., 1939, 241, 172. 31 B. Subramaniam, R. A. Rajewski and K. Snavely, J. Pharm. Sci., 54 S. Peter, F. Altofer, W. Bu� hrer and H. D. Lutz, VIth European 1997, 86, 885. Conference on Solid State Chemistry, Zu� rich, Sept. 1997. 32 D. W. Matson, J. L. Fulton, R. C. Petersen and R. D. Smith, Ind. 55 F. Kamoun, M. Lorenz, G. Kempe and F. Peger, J. Less-Common Eng. Chem. Res., 1987, 26, 2298. Metals, 1991, 170, 1. 33 E. Reverchon, Supercritical fluids—Materials and natural products 56 K. J. Gallagher and M. R. Ottaway, J. Chem. Soc., Dalton Trans., processing, ed. M. Perrut and P. Subra, AIPFS Publishing, Nancy, 1974, 2347.France, 1998, pp. 221–236. 57 P. Keller, N. Jb. Miner. Abh., 1970, 113, 29. 34 M. PoliakoV and M. W. George, Supercritical Fluids: Materials 58 P. Hagenmuller, J. Portier, J. Cadiou and R. de Pape, C. R. Acad. and Natural Products Processing, ed. M. Perrut and P. Subra, Sci. Paris, 1965, 260, 4768. AIPFS Publishing, Nancy, France, 1998, pp. 833–842. 59 M. Vlasse, J. C. Massies and G. Demazeau, J. Solid State Chem., 35 E. Dinjus, R. Fornika and M. Scholtz, Chemistry under extreme or 1973, 8, 109. non-classical conditions, ed. R. V. Eldik and C. D. Hubbard, 60 A. L. Mackay, Mineral. Mag., 1960, 32, 545. Wiley, New York, 1996, pp. 219–272. 61 J. M. Gonzalez-Calbet, M. A. Alario-Franco and M. Gayoso- 36 V. Gourinchas-Courtecuisse, K. Chhor, J.F. Bocquet and C. Andrade, J. Inorg. Nucl. Chem., 1981, 43, 257. Pommier, Ind. Eng. Chem. Res., 1996, 35, 2539. 62 J. M. Combes, A. Manceau, G. Calas and J. Y. Bottero, Geochim. 37 V. Gourinchas-Courtecuisse, J. F. Bocquet, K. Chhor and C. Cosmochim. Acta, 1989, 53, 583. Pommier, J. Supercritical Fluids, 1996, 9, 222. 63 R. So� derquist and S. Jansson, Acta. Chem. Scand., 1966, 20, 1417. 38 Y. Hakuta, H. Terayama, S. Onai, T. Adschiri and K. Arai, 64 G. Biedermann and J. T. Chow, Acta Chem. Scand., 1966, 20, Supercritical Fluids—Chromatography and Novel Applications, ed. 1376. S. Saito and K. Arai, AIPFS Publishing, Nancy, France, 1997, 65 J. E. Post and V. F. Buchwald, Am. Mineral., 1991, 76, 272. pp. 255–258. 66 M. Ohyabu and Y. Ujihira, J. Inorg. Nucl.Chem., 1981, 43, 3125. 39 T. Adschiri, S. Yamane, S. Onai and K. Arai, Supercritical 67 E. Murad, Clay Miner., 1979, 14, 273. Fluids—Reactions, Material science and Chromatography, ed. M. Perrut and G. Brunner, AIPFS Publishing, Nancy, France, 1994, vol. 3, pp. 241–246. Paper 8/04964E J. Mater. Chem., 1999, 9, 67–75 75 J O U R N A L O F C H E M I S T R Y Materials Supercritical fluid processing: a new route for materials synthesis† Franc�ois Cansell,* Bernard Chevalier, Alain Demourgues, Jean Etourneau, Christophe Even, Yves Garrabos, Vincent Pessey, Ste�phane Petit, Alain Tressaud and Franc�ois Weill Institut de Chimie de la Matie`re Condense�e de Bordeaux [ICMCB], CNRS—UPR 9048, Universite� Bordeaux I, Cha�teau Brivazac, Av.du Dr. Schweitzer, 33608 Pessac Cedex, France.E-mail: cansell@chimsol.icmcb.u-bordeaux.fr Received 23rd April 1998, Accepted 29th June 1998 Supercritical fluids exhibit a range of unusual properties that can be exploited for new reactions which are qualitatively diVerent from those involving classical solid state chemistry. After giving a brief introduction to these fluids we describe their use in inorganic chemistry and related fields.We then present two examples concerning diVerent areas of solid state chemistry: (i) the formation of novel inorganic nanoparticles; (ii) the preparation of new open-structure oxy(hydroxy)fluorides, thus showing the advantages of this supercritical fluid processing that can be seen as an alternative method to regular solution chemistry or solid–gas reactions.Thus, it is possible to prepare new materials under mild Introduction conditions by using either physical or chemical transformations A fluid is in the supercritical domain when both pressure and (see Fig. 2). The main results concerning these two possibilities temperature are above their critical values, Pc and Tc. In are reported herein. However, first of all, it should be pointed practice, this definition is restricted to fluids close to their out that supercritical fluids are used in order to provide critical points and, hence, with their density close to rc .There contaminant-free compounds. Moreover, this method allows are numerous fluids for which the critical temperature is one to obtain solid particles exhibiting both a narrow range moderate.A selection of the most usual ones is listed in of size distribution (micronic and nanometric) and specific Table 1.1 features (e.g. well defined stoichiometry, structure, etc.). Nowadays supercritical fluids are widely used in many fields.2–4 These fluids are very attractive for materials pro- Physical transformations cessing5 and more particularly for the formation of particles, fibres, thin films (pharmaceuticals, explosives, coatings)6–9 and Concerning the physical transformations, three main techfor drying materials (highly porous gels).10,11 Furthermore, niques are successively presented: the rapid expansion of a these fluid processes are already used in many other appli- supercritical solution (RESS), supercritical antisolvent precipications, such as: separation (petroleum-chemistry separation tation: gas anti-solvent (GAS), and precipitation with comand purification, food industry);12–15 chromatography (analyt- pressed anti-solvent (PCA).ical and preparative separation, determination of physicochemical properties);16,17 chemical reactivity ( low-density polyethylene, waste destruction, polymer recycling);18–22 and Table 1 Critical coordinates of usual pure fluids.Tc , Pc and rc are earth sciences (volcanism, geothermal energy, hydrothermal the critical temperature, pressure and density, respectively1 synthesis).23 Both the capability of some supercritical fluids to replace Fluid Tc/°C Pc/MPa rc/kg m-3 toxic industrial solvents and the ability of tuning solvent characteristics for highly specific separations or reactions lead Carbon dioxide 31.2 7.38 468 Ammonia 132.4 11.29 235 to the current industrial and scientific interest in supercritical Water 374.1 22.1 317 fluids.24 These fluids possess physicochemical properties, such Ethylene 9.5 5.06 220 as density, viscosity and diVusivity, which are intermediate Ethane 32.5 4.91 212 between those of liquids and gases (Table 2).The main interest Propane 96.8 4.26 225 in supercritical fluids as reaction media relies on their continu- n-Pentane 196.6 3.37 232 ously adjustable properties from gas to liquid with small Cyclohexane 279.9 4.03 270 Methanol 240.0 7.95 275 pressure and temperature variations. Fig. 1 shows that it is Ethanol 243.1 6.39 280 possible to move continuously from the liquid to the gaseous Isopropanol 235.6 5.37 274 state without crossing the liquid–gas equilibrium line.Acetone 235.0 4.76 273 Classically, materials preparation in supercritical fluid can be described in three steps: step one: solubilization of solutes in dense fluid (close to the liquid phase) since the solute solubility varies generally as a power law with respect to Table 2 Characteristic magnitudes of thermophysical properties of density;25 step two: chemical or physical transformation of fluids fluids in the supercritical domain, characterized by specific properties (Table 2)aterials recovery in the Liquid Supercritical fluid Gasa low density fluid domain (close to the gas phase) in order to r/kg m-3 1000 100–800 1 obtain dry materials.g/Pa s 10-3 10-4–10-5 10-5 Db/m2 s-1 10-9 10-8 10-5 †Basis of the presentation given at Materials Chemistry Discussion r=density; g=viscosity; D=diVusion coeYcient.aAt ambient con- No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, ditions. bFor small-molecule solute. France. J. Mater. Chem., 1999, 9, 67–75 67remains soluble in the organic solvent. These processes are used for explosives, polymers, food, pharmaceutical compounds, coloring matter and catalysts.For the latter it can be specified that zinc acetate nanoparticles down to about 30 nm are obtained with an average particle size of 50 nm.33 With the GAS process, a batch of the solution (solute dissolved in organic solvent) with dense carbon dioxide for example, is expanded several times in a vessel.Because of the lower solvent strength of the carbon dioxide–expanded solvent mixture, the fluid medium becomes supersaturated, thus forcing the solute to precipitate in microparticles. The PCA process consists of spraying the solution through a nozzle into dense carbon dioxide, for example. These two processes allow the control of the particle size (micronic and submicronic).Unfortunately their use is limited Fig. 1 Schematic representation of a phase diagram illustrating the by the separation of the solid particles from the solvents. density evolution from the liquid to the gas without crossing the T–C line which corresponds to the liquid–gas equilibrium line, C being the critical point and T being the triple point. Pc and Tc are the critical pressure and temperature respectively.Chemical transformations Physico-chemical properties of supercritical fluids as a reactive medium are relatively easy to adjust by operating conditions. Indeed, these tunable characteristics, such as density, viscosity, and diVusivity, influence directly the reaction rate constants. For the reactions involving ionic species, the variations of the relative permittivity and of the ion products influence both the chemical equilibrium and the evolution of the transition state.By moving continuously from the liquid to the gaseous state, the eYciency of the reaction may be significantly enhanced to create a completely new chemical process. As the reactivity of organic compounds in supercritical fluids has already been intensively studied,34,35 we focused on materials preparation.Supercritical fluids are used for matter transport or as chemical media in order to produce submicronic particles, compounds with specific features, polymer–metal composites, aerogels and thin films. Chemical reactions generally involve the thermal decomposition of a metal-containing precursor. The various applications developed in the field of inorganic materials are also reported.Fig. 2 Possible routes for material elaboration using supercritical fluids. RESS=rapid expansion of supercritical solution; GAS=gas Oxide nanoparticles anti-solvent recrystallisation; PCA=precipitation with compressed anti-solvent. Submicronic TiO2 powder has been obtained in a flow reactor by a sol–gel process.Titanium isopropoxide, Ti(O-iC3H7)4, is Rapid expansion of supercritical solution (RESS) process hydrolytically decomposed by water, produced by the catalytic dehydration of the isopropyl alcohol used as the supercritical In this process, the reactant is first solubilized in a supercritical solvent.36 Titanium hydroxides are formed and their decompo- fluid which is then expanded across a nozzle or a capillary at sition into TiO2 kinetically limits the reaction.At temperature a very high velocity. Such a rapid expansion leads to a high and pressure around 300 °C and 10 MPa respectively, particles solute supersaturation and subsequently to the precipitation with a narrow range of size distribution (20–60 nm), some of of small monodispersed particles.28 This process has been which crystallize with the anatase structure, are weakly associ- applied to inorganic,29 organic30 and pharmaceutical31 comated into spherical agglomerates (500–2000 nm). A model of pounds.It allows the preparation of either an intimately mixed the flow reactor, has been made including hydrodynamic and powder of two inorganic compounds (e.g. SiO2 and KI) or an kinetic studies,37 and experimentally validated for optimizing inorganic and organic combination [e.g.KI and poly(vinyl the process. chloride)].32 Metal oxide submicronic particles can also be obtained in The major limitation of the RESS process is the low supercritical water by the hydrolysis of metal salts in aqueous solubility of many compounds in supercritical fluids. For solutions.Numerous oxides have been studied such as Fe2O3 , example in carbon dioxide, the solute solubilities are about Co3O4 , NiO, ZrO2, TiO2 and CeO2.38 In the case of TiO2, 0.01 wt.% or less. The other challenge for large commercial TiCl4 was used as the precursor and leads, at 30 MPa and plant configuration is associated with nozzle design in order 400–450 °C, to the formation of TiO2 (anatase) with a prismatic to avoid particle accumulation and the freezing caused by the shape and a size of 20 nm.39 rapid expansion.Porous substrates (alumina or zirconia) with high surface areas have been impregnated with supercritical solutions of Supercritical anti-solvent precipitation processes: gas antimetal salts (carbonate, acetate, etc.), which can be adsorbed solvent (GAS) and precipitation with compressed anti-solvent or deposited as a film on the surface of the substrate by (PCA) lowering the pressure.40 Supercritical fluid has also been used as a drying medium In these processes a solute is dissolved in an organic solvent.The addition of a supercritical fluid leads to a lower solubility in the sol–gel process for the preparation of nanometric tin oxide particles (3–5 nm).41 of the solute in the liquid phase since the supercritical fluid 68 J.Mater. Chem., 1999, 9, 67–75Metal nanoparticles expansion of the supercritical fluid causes the vaporisation of the solutes. The precursors can be decomposed at or near a Micronic metallic powders (as pure metals, metal alloys, or substrate surface to form a thin film.This leads to the mixtures thereof ) have been prepared from metal salts.42 formation of pure metal (e.g. Al, Cr, Ni, Cu, etc.), or oxide Precursors are solubilized in methanol at a temperature lower (e.g. CuO, SiO2, Cr2 O3, etc.) films. Moreover, the stoichiothan that employed to generate a metallic powder. As an metric ratios of the precursors for multi-element coatings (e.g. example, micronic Cu particles were obtained from a solution YBa2Cu3O7-x) can be precisely controlled.46–48 of Cu(OAc)2 at 275 °C and about 14 MPa.When Pd(OAc)2 After this brief review, we present our own results concerning was added to a methanol solution of Cu(OAc)2, a mixture of a new route for nanoparticle formation and for the synthesis Pd and Cu metals and a face centered cubic Cu–Pd alloy were of new phases in solid state chemistry such as open-structure obtained; both systems were identified by XRD and ESCA oxy(hydroxy)fluorides. measurements.Compounds with specific features Preparation of nanoparticles Semiconducting quaternary antimony sulfides, such as The aim of this work is to develop a new method for the KAg2SbS4 or RbAg2SbS4 , have been prepared in supercritical preparation of nanoparticles.Indeed, metallic nanoparticles ammonia at 160 °C and about 22 MPa during 4 days from the with dimensions of tens of nanometers or less exhibit size- appropriate mole ratio of K2S4, K2CO3 or Rb2CO3 with Ag, dependent electrical, chemical and magnetic properties. The Sb2S3 and S8. The use of supercritical ammonia leads to high chemical transformation of metallic precursors inside a quality crystal growth, owing to its specific transport properties supercritical fluid is a new route for obtaining nanometric (viscosity, solubility, etc.).43 homogeneous powders.The process mainly deals with the dissolution of precursors in a dense fluid, followed by the Polymer–metal composites decomposition of the precursors in supercritical fluid for Polymer–metal composites consist of isolated nanoclusters of preparing either powders or thin films of materials.metal atoms distributed homogeneously throughout polymeric substrates. For example, ametal precursor such as dimethyl(cy- Experimental setup cloocta-1,4-diene)platinum(II ) is dissolved in supercritical carbon dioxide and impregnated into the polymer [poly(4- The apparatus used for chemical synthesis in supercritical fluid methylpent-1-ene)].Thermolytic reduction of the precursor in media is shown in Fig. 3. The system itself, placed in a the CO2-swollen polymer medium at 140 °C and under 26 MPa temperature controlled hot-air oven, is composed of: (i) two pressure produces platinum aggregates with a maximum diam- high pressure 316 stainless steel cells. The first one (cell A) is eter of approximately 50 nm.44 the mass exchanger. It is a cylinder of 1.8 cm diameter and 5.5 cm length filled with a glass packed bed of 0.2 cm diameter.Aerogels A glass wool plug is placed at the outlet of the cylinder in order to avoid the dragging of the precursor by the flowing Sol–gel reactions have been directly performed in supercritical fluid.For each experimental run the cylinder is charged with carbon dioxide as a reactive medium by using formic acid as 1 g of the precursor. The second cell (cell B) is the reactor. It condensation reagent.45 Supercritical carbon dioxide appears consists of a cylinder of 1.8 cm diameter and 9.5 cm length. to be a better solvent than ethanol for preparing phenylene- An external heating resistor allows the generation of a flash bridged polysilsesquioxane aerogels from 1,4-bis(triethoxysiof temperature.A glass wool plug is placed at the outlet of lyl )benzene. In addition, this process can be successfully used the cylinder for the reasons mentioned above. A polytetra- to generate monolithic aerogels from their precursors in a fluoroethylene film (PTFE, 0.25 mm thickness) rolled up inside single step.the cell is used for collecting the metallic particles. (ii) A high pressure generator, which is a fluid metering pump (CM 3200 Thin films P/F). This pump provides either constant flow or constant pressure. Only one fluid can be used with this system (in our Supercritical fluid transport and chemical deposition can be used for the formation of thin films on a substrate. This film case it is ammonia).The working pressure is monitored by a digital gauge. (iii) Two thermocouples measure the tempera- deposition technique can use nonvolatile precursors, unlike the chemical vapor deposition process. In supercritical fluid ture, one is placed inside the hot-air oven, the other one is inside cell B.transport and chemical deposition technique, precursors are dissolved in the supercritical fluid. Then, this solution flows Acetylacetonates of metals [Cu(AcAc)2 and Fe(AcAc)3] were used as precursors. They were all provided by Sigma- through a restrictor into a deposition chamber where the rapid Fig. 3 Apparatus for chemical synthesis in supercritical fluid media. 1, Fluid tank. 2, Cooling system. 3, High pressure pump: 1–10 ml min-1; 0–40 MPa. 4, Heat exchanger. 5, Mass exchanger (cell A). 6, Reactor (cell B). 7, Temperature regulator. 8, Contact breacker. 9, Micrometric valve. 10, Solvent trap. 11, Oven. J. Mater. Chem., 1999, 9, 67–75 69Aldrich and used without further purification. Their solubility in ammonia was measured and is reported in Table 3.The solid particle formation results from the solute thermal decomposition in the supercritical fluid medium. The fact that the critical temperature of this solvent must be lower than that of the metal chelate decomposition justifies the choice of ammonia as the supercritical fluid. In addition this fluid exhibits a high reducing power. A known amount of precursor is placed in cell A.The system is filled with solvent. Then the hot-air oven is heated at a temperature TA which allows good solute solubility. Cell B is heated at temperature TB which allows chemical reaction. When both temperatures are stabilized, the pressure is increased up to 17 MPa and then maintained constant. In these conditions ammonia is liquid in the mass exchanger (cell A) and is in the supercritical domain in the reactor (cell B).Then the fluid flow rate is stabilized at the working value of 3 ml min-1 and the fluid mixture (solvent+precursor) is transported to the reactor. The solute is thermally decomposed Fig. 4 SEM picture of the mixture of Fe4N and Fe2O3 obtained by in the reactor and the metallic part is precipitated on the the decomposition of Fe(AcAc)3 in supercritical ammonia.(Treactor= PTFE film. The organic part, still soluble in the supercritical 180 °C; P=16 MPa). solvent, is trapped at the outlet of the installation. Results and discussion Chemical syntheses in supercritical fluids are known to allow the obtention of powders with small particle size and narrow distribution range. Indeed, these syntheses are based on homogeneous nucleation and large supersaturation in a supercritical solvent and the size of the collected particles is very sensitive to the supersaturation: the higher the supersaturation, the smaller the particles.49 Moreover, in a supercritical fluid small changes of pressure give rise to large changes in both density and solubility.50 Thereby the optimization of working pressure and temperature in the reactor allows the achievement of high supersaturation and the formation of nanosized particles.This is demonstrated by our results, obtained from the thermal decomposition of Fe(AcAc)3 made in the dense fluid domain and in the supercritical domain. Obviously, supercritical fluids as reactive media allow the formation of smaller particles (with waverage=50 nm, see Fig. 4) than those obtained in dense fluids (with waverage=800 nm, see Fig. 5). The experiments have shown that the final particle size depends on the variations of Fig. 5 SEM picture of the mixture of Fe4N and Fe2O3 obtained by the working conditions (T, P, solute supersaturation, hydrothe decomposition of Fe(AcAc)3 in liquid ammonia. (Treactor=120 °C; dynamics, etc.).The relationship between size and working P=10 MPa.) conditions is under study. A solvolysis reaction occurs during the decomposition of the precursor. Indeed, the particle reactivity is very high and Table 4 Cu and Fe nitride formation in supercritical ammonia thus the use of ammonia may lead to metal nitrides. In the Precursor Working conditions Products case of Cu(AcAc)2 and Fe(AcAc)3 precursors, we have prepared particles of copper nitride Cu3N at 200 °C and 17 MPa Cu(AcAc)2 Tmass exchanger=90 °C Cu3 N with copper as and iron nitride Fe4N at 180 °C and 16 MPa (Table 4).The Treactor=200 °C impurity sample of Cu3N has been characterized by XRD. The pattern P=17 MPa indicates that the Cu3N phase is obtained with copper as Ammonia flow=3 ml min-1 Fe(AcAc)3 Tmass exchanger=90 °C Mixture of Fe4N and impurity (Fig. 6). In the case of the Fe(AcAc)3 decomposition, Treactor=180 °C Fe2 O3 a mixture of both iron nitride Fe4N and iron oxide Fe2O3 was P=16 MPa observed (Table 4). This mixture has been characterized by Ammonia flow=3 ml min-1 electron diVraction because this sample seems to be amorphous to X-ray diVraction (Fig. 7, Table 5).Aggregates of 50 nm are obtained in the case of the Fe4N–Fe2O3 mixture. They consist of very small particles Table 3 Solubility (at 90 °C and 10 MPa) and decomposition tempera- (10 nm) with a narrow range of size distribution as shown ture (at 16 MPa) values for Cu(AcAc)2 and Fe(AcAc)3 in ammonia in Fig. 8. The iron oxide formation could be due to the high (the critical coordinates of ammonia are Tc=132.4 °C and Pc= 11.29 MPa) reactivity of iron with oxygen present in the acetylacetonate groups.New precursors free from oxygen are under study in Solubility/g g-1 a Decomposition temperature/°C order to obtain metallic nanoparticles only. It must be pointed out that our working temperatures for Cu(AcAc)2 5.8×10-3 190 obtaining metal nitrides are clearly lower than those used in Fe(AcAc)3 3.8×10-3 170 classical solid state chemistry routes.Classical ways for making aGram of precursor per gram of fluid. transition-metal nitrides involve high-temperature and high- 70 J. Mater. Chem., 1999, 9, 67–75Table 5 d-Values obtained for mixture of Fe4N and Fe2O3 (see Fig. 7) dmeasured/A° dFe4N/A° a dFe2O3/A° a 6.47 4.76 4.76 4.01 3.79 3.34 3.24 3.15 2.99 2.981 2.804 2.728 2.684 2.55 2.573 2.548 2.458 2.372 2.295 Fig. 6 X-Ray diVraction pattern of Cu3N obtained by the decompo- 2.243 sition of Cu(AcAc)2 in supercritical ammonia (Treactor=200 °C; P= 2.19 2.191 17 MPa.) *indicate the peaks relating to copper. 2.176 2.018 1.997 1.983 1.973 1.897 1.855 1.807 1.75 1.735 1.697 1.66 1.666 1.58 1.53 1.549 1.31 1.342 1.25 1.265 1.200 1.12 1.144 1.095 1.06 1.053 1.01 1.014 0.98 0.949 0.88 aFrom ASTM files.tions, the electrical properties of the catalyst constitute one of the key features, in particular when the materials show mixed conductivity, i.e. fast transport of ions in addition to electronic conductivity at various temperatures. In the case of oxides, for instance, ionic conductivity is generally due to polarizing cationic species such as H+ and Li+, or polarisable anionic O2-/O- entities. The aim of this work was then to elaborate and to study some (O2-, OH-, F-) mixed-anion systems showing good mixed conductivity, such as Sr(OH)X (X=Cl, Br, I )54 phases, which are known as protonic conductors.From a chemical point of view, it was interesting to use Fig. 7 Electron diVraction pattern of the mixture of Fe4N and Fe2O3 transition metal fluorides as starting materials, i.e.compounds obtained by the decomposition of Fe(AcAc)3 in supercritical ammonia which already contain ionic MMF bonds, instead of substitut- (Treactor=180 °C; P=16 MPa). (See Table 5.) ing oxy- or hydroxy-anions by fluorine in oxides or hydroxides exhibiting strongerMMO (OH) bonds as was done in previous work.55,56 From a structural point of view, and in order to pressure reactions between a pure metal and a nitrogen source.51 However, there are other ways to obtain metal produce good catalytic properties, we were rather interested in open structures based on transition metal networks, such nitrides, based on the use of precursors (such as organometallics, acetates, salts etc.).In this case, the working conditions as tunnelled or layered frameworks, to enhance the ionic conductivity and to favor electronic conductivity. are lowered. Indeed, single crystalline c¾-Fe4N can be obtained by chemical transport from iron over [Fe(NH3)6]I2] in A suitable method would then consist of partial hydrolysis in the case of a transition metal fluoride with the desired supercritical ammonia ( p(NH3)=600–800 MPa, 460 °CT 580 °C)].52 By heating CuF2 at 280 °C in NH3, pure Cu3N is oxidation state, such as FeF3, whereas oxidative hydrolysis would occur for compounds with lower oxidation states, such obtained.53 as FeF2. We could also consider ternary fluorides such as NH4FeF3 from which ammonium ions could be topotactically Open-structure oxy(hydroxy)fluorides deintercalated, thus entailing the replacement of F- by OH-/O2- ions in an aqueous medium, in order to obtain an During the past few years, much work has been devoted to the study of heterogeneous catalytic systems.In many reac- oxyfluoride or an oxy(hydroxy)fluoride. J. Mater. Chem., 1999, 9, 67–75 71obtained a higher fluorine content; however it was limited to e<0.3.A method for preparing oxy(hydroxy)fluorides with the hollandite-type structure which would lead to a much higher fluorine content, involves the oxidation of transition metal fluorides with H2O2 solutions. Using this route, we could expect that F- and OH- would occupy the same type of sites within the network, at the corners of the FeX6 octahedra, instead of being present only as inserted species in the tunnels of the framework.In this connection we describe here the oxidative hydrolysis of the fluoroperovskite NH4FeF3 under supercritical conditions. Experimental setup Our experiments were carried out in scCO2; a second high pressure pump was added for the injection of the solution (H2O2–H2O–ethanol ) at temperatures between 50 and 250 °C approximately, and at pressures varying from 10 to 40 MPa.The reactant used was a concentrated aqueous solution of hydrogen peroxide (H2O2, 30M), diluted in an equal volume of ethanol. The stainless steel reactor (5 mm diameter, 50 mm long) (cell A of Fig. 3) was prepared as described above and filled with freshly ground NH4FeF3 powder in a dry argon glove box.It was subsequently mounted on the pressure line in a programmable hot-air oven. The line was then flushed with CO2 before raising the temperature and starting the reaction. During the treatment, great care should be taken in keeping the system in supercritical conditions in order to avoid water separation: water solubility in scCO2 remains low, and the introduction of large quantities of water would lead to a twophase system.Fig. 8 TEM picture of the mixture of Fe4N and Fe2O3 obtained by After the reaction, the reactor was flushed again with scCO2 the decomposition of Fe(AcAc)3 in supercritical ammonia (Treactor= in order to eliminate residual water. 180 °C; P=16 MPa). The compounds obtained were characterized by electron probe microanalysis (EPMA) (O/Fe and F/Fe ratios), TGA–DTA coupled with mass spectrometry (adsorbed H2O, Unfortunately, several drawbacks arise from these methods: OH and F contents), diVuse reflectance-IR, electron first, fluoride materials are generally sensitive to moisture; microscopy (morphology of the particles) and X-ray secondly, moderate temperatures, necessary to obtain open- diVraction.structure materials, generally lead to low diVusion of the reactive species (gaseous or liquid ) through the solid fluoride Results and discussion material, and then induce a gradient in the chemical composition of the final product.The oxidative process using the water–ethanol solution of H2O2 in scCO2 on NH4FeF3 allowed us to prepare oxy(hyd- Therefore, we have chosen to use a supercritical medium which allows very high diVusivity of the species through the roxy)fluorides with the general formula FeOF1-x(OH)x nH2O (x0.2, n0.4), isostructural with b-FeOOH, i.e.exhibiting transition metal fluorides and good elimination of water. Supercritical CO2 (scCO2) with ethanol as a co-solvent the hollandite-type structure (Fig. 9b). The frameworks of the rutile-type and the hollandite-type structures are built of chains appeared to be a good candidate.In order to demonstrate the interest and capability of this of octahedra sharing opposite edges (Fig. 9). In the case of the rutile structure, single chains are connected to one other method for preparing 3d-transition metal oxyfluorides, we focused our attention on iron-based compounds, owing to the by vertices, whereas in the hollandite structure double chains sharing edges form large square tunnels of about 2×0.5 nm well known stability of the two oxidation states +II (3d6) and +III (3d5) present in iron oxides and fluorides.In addition, section and running along the c direction. The small amount of water which is present in the compounds mainly depends several attempts were made over the last few decades to prepare iron oxyfluorides with open frameworks.57 It should on the conditions of the final CO2 treatment, performed after the reaction.be pointed out that only the rutile form of FeOF (Fig. 9a) has been obtained so far, by high temperature reaction.58,59 After optimization of the experimental supercritical conditions, NH4FeF3 was treated at 150 °C for 30 min under a Numerous works have also been reported concerning an iron oxyhydroxide exhibiting an open structure (hollandite scCO2 pressure of 30 MPa.The volume of the injected H2O2 solution (as described above) was 15 ml. The X-ray diVraction type): b-FeOOH.60,61 This compound was obtained by hydrolysis of FeCl3 .62–64 It has been shown that residual Cl- pattern shows a well crystallized single phase with a hollanditetype structure (sample A; Fig. 10a). The tetragonal cell param- anions are always present in the tunnels of the framework, keeping it from collapsing.65 Some attempts have been made eters (a=1.036 nm, c=0.302 nm) are slightly smaller than those reported for b-FeOOH57 (a=1.053 nm, c=0.303 nm). to prepare the corresponding fluorinated phase FeO(OH)1-eFe via the hydrolysis of FeF3.The material contained only a few The lower a-value could be related to the presence of F- anions instead of OH- in the framework. atom% of F- anions which were located in the tunnel sites of the structure.60 Other authors tried to synthesize such com- We have also prepared, using the standard hydrolysis route, two other compounds also exhibiting the same hollandite pounds by the hydrolysis of iron hydroxides56 or nitrates55,66 in a concentrated NaF or NH4F aqueous solution, and structure: a slightly chlorinated hydrated iron oxyhydroxide 72 J.Mater. Chem., 1999, 9, 67–75Fig. 9 a) Rutile-type FeOF and b) open-structure hollandite-type ‘FeOF nH2O’ prepared in supercritical medium (sample A). Fig. 10 Powder XRD patterns of a) hollandite-type FeOF1-x(OH)x nH2O prepared in supercritical medium (sample A) and b) slightly fluorinated b-FeOOH prepared by the classical hydrolysis route (sample C).(sample B), by hydrolysis of FeCl3 as reported earlier;62–64 by DTA–TGA (under flowing nitrogen) coupled with mass spectrometry. The amount of water molecules depends on the and a slightly fluorinated hydrated iron oxyhydroxide (sample C), by hydrolysis of FeF3 .60 This latter type of compound conditions of the final CO2 flow and varies from 0 to 0.4. In the case of sample A obtained as described above, the composi- reveals a much lower crystallinity than those prepared in supercritical conditions, as shown in Fig. 10b. tion is FeOF0.9(OH)0.1 . The FeOF1-x(OH)x nH2O compounds were checked by In order to check the eVectiveness of the oxidizing reaction and the final oxidation state of iron, samples obtained by infrared spectroscopy (diVuse reflectance), as well as hydrated b-FeO(OH,Cl ) (sample B), as shown in Table 6.supercritical treatment were treated afterwards under flowing HF at moderate temperatures (40<T<150 °C) for 2 hours. The spectrum of sample B shows the characteristic bands of slightly chlorinated b-FeOOH, as reported earlier,57,61,67 in This type of fluorination reaction is known to be non-oxidative, and leads to an exchange of O2- and OH- anions by F-.particular those of the deformation mode of the FeKOKFe bonds (n=1070–1030 cm-1), and the asymmetric and sym- The phase obtained after reaction with HF corresponds to rhombohedral FeF3 (ReO3-derived phase), as characterized metric vibrations of FeKOKFe bonds at n=700 and 420 cm-1, respectively.Large quantities of water are also detected (n= by XRD and EPMA. This result confirms that all Fe2+ from NH4FeF3 had been fully oxidized to Fe3+ during the reaction 3400–3300 cm-1, 1620 cm-1). The spectra of the oxy(hydroxy)fluorides prepared in this with H2O2 in supercritical CO2–H2O–ethanol.The atomic ratios O/Fe and F/Fe have been estimated by work diVer from those of chlorinated b-FeOOH: in addition to the FeKOKFe vibrations mentioned above and located at EPMA and the fluorine and water contents have been obtained J. Mater. Chem., 1999, 9, 67–75 73Table 6 Infared band assignments for FeOF1-x(OH)x nH2O and for slightly chlorinated iron oxyhydroxide (sample B) obtained by conventional hydrolysis of FeCl3 FeOF1-x(OH)x nH2O Hydrated b-FeO(OH,Cl) (sample B) Type of n/cm-1 Intensity (a.u.) vibration n/cm-1 Intensity (a.u.) 3500–3400 s n(OH) 3480–3360 vs 3180 s Characteristic of fluorinated compounda — — 1619 mw d(HKOKH) 1615 s 1080 w L Deformation modes of FeKO(H)KFe61 1070 vw 1020 K 1030 w m 956 m Characteristic of fluorinated compounda — — 845 w Interaction of H2O with hollandite tunnels61 845 m 705 ms Asymmetric vibration of FeKOKFe in hollandite 650–700 ms 424 w Symmetric vibration of FeKOKFe in hollandite 420 vw Intensities: vs=very strong; s=strong; m=medium; w=weak; vw=very weak.aThese bands were assigned in previous works57,61 to the signature of the presence of fluorine species in the structure.n=1080–1020, 705, 425 cm-1, two strong bands appear at n= CO2 medium) through the NH4FeF3 network coupled to the eYcient elimination of water allows the synthesis of new types 3180 and 956 cm-1. They were also found with much lower intensity in slightly fluorinated b-FeOOH,57 and seem charac- of compounds which adopt the hollandite-type structure with a framework exhibiting large tunnels; (ii) the phase corre- teristic of fluorinated oxy-compounds.The amount of water present in the tunnels of sample A, which is in any case smaller sponding to the FeOF0.9(OH)0.1 chemical composition shows a high fluorine content compared to fluorinated iron oxyhyd- than for sample B obtained via conventional hydrolysis, depends on the final CO2 treatment.This may be related to roxides prepared by conventional routes. The use of this method is now in progress for preparing the fact that flushing the reactor with the supercritical CO2–ethanol mixture helps to eliminate residual water.61 This other metastable phases with various transition metals, such as lamellar structures containing large interlayer spaces. is evident by the lowering of the intensity of the band at 845 cm-1. We are grateful to C.Chabot, L. Rabardel, E. Sellier and Electron diVraction and Rietveld refinement of powder XJ. J. Videau for helpful discussions and technical support. ray diVraction data are in progress in order to elucidate the structural parameters of these new oxy(hydroxy)fluorides. A Mo�ssbauer spectroscopy study will specify the oxidation state References and the environment of iron.These results will be presented 1 N. B. Vargaftik, Table on the thermophysical properties of liquids in a forthcoming paper. and gases, Hemisphere Publishing Corporation, London, 1975. 2 Innovations in Supercritical Fluids, ed. K. W. Hutchenson and N. R. Foster, ACS Symp. Ser. 608, Washington DC, 1995. Conclusions 3 Supercritical Fluids—Fundamentals for Applications, ed.E. Kiran It has been pointed out that chemical synthesis in supercritical and J. M. H. Levelt Sengers, NATO ASI Series E273, Kluwer, Dordrecht, 1994. media is a new route for preparing either nanometric powders 4 Fluides Supercritiques et Mate�riaux, ed. F. Cansell and J. P. with narrow size distribution or oxy(hydroxy)fluorides exhibit- Petitet, AIPFS Publishing, Nancy, France, 1995.ing open frameworks. 5 X. Y. Zeng, Y. Arai and T. Furuya, Trends Chem. Eng., 1996, As far as the formation of nanoparticles is concerned the 3, 205. following conclusions may be drawn. (i) The use of supercriti- 6 P. G. Debenedetti, Supercritical Fluids—Fundamentals for cal fluids as reactive medium allows control of the particle size Application, ed.E. Kiran and J. M. H. Levelt Sengers, NATO ASI Series E273, Kluwer, Dordrecht, 1994, pp. 719–729. distribution owing to the tunable density and solubility of the 7 H. Ksibi and P. Subra, Adv. Powder Technol., 1996, 7, 21. solute. Then the solute supersaturation can be continuously 8 K. Chhor, J. F. Bocquet and C. Pommier, Mater. Chem. Phys., adjusted in order to control the size of the nanoparticles.The 1995, 40, 63. systematic variation of working conditions (T, P, solute 9 P. Beslin, P. Jestin, Ph. Desmarest, R. Tufeu and F. Cansell, supersaturation, hydrodynamics, etc.) and their eVects on the Supercritical Fluids—Reactions, Material science and particle size are under study. Thus, metallic nanoparticles Chromatographiy, ed.M. Perrut and G. Brunner, AIPFS Publishing Nancy, France, 1994, vol. 3, pp. 321–326. prepared by this route should be very attractive for their size- 10 R. A. Laudise and D. W. Johnson, J. Non-Cryst. Solids, 1986, dependent electrical, chemical, and magnetic properties. (ii) 79, 155. The nature of the final product depends on the nature of the 11 B. Rangarajan and C.T. Lira, J. Supitical Fluids, 1991, 4, 1. solvent. The metallic atoms can easily react with ammonia, 12 A. Akgerman and G. Madras, Supercritical Fluids—Fundamentals giving nitrides, Cu3N and Fe4N, at temperatures clearly lower for Application, ed. E. Kiran and J. M. H. Levelt Sengers, NATO than those used in conventional solid state chemistry routes. ASI Series E 273, Kluwer, Dordrecht, 1994, pp. 669–695. 13 Supercritical Fluid Extraction: Principles and Practice, ed. M. A. (iii) The nature of the final product depends also on the nature McHugh and V. J. Krukonis, Butterworths, Stoneham, MA, 1988. of the metallic precursors (e.g. the formation of Fe2O3 results 14 F. Cansell, Ph. Botella, Y. Garrabos, J. L. Six, Y. Gnanou and from the high reactivity of iron with oxygen present in R.Tufeu, Polym. J., 1997, 29, 910. acetylacetonate groups). In this case the preparation of pure 15 E. J. Beckman, Nature, 1996, 271, 613. metallic nanoparticles should require the use of metallic precur- 16 S. H. Page, J. F. Morrison and M. L. Lee, Supercritical Fluids— sors free from oxygen. (iv) The supercritical chemistry route Fundamentals for Application, ed.E. Kiran and J. M. H. Levelt Sengers, NATO ASI Series E273, Kluwer, Dordrecht, 1994, is interesting not only for obtaining either intimate homopp. 641–652. geneous mixtures of nanoparticles, (e.g. Fe2O3+Fe4N) or 17 G. M. Schneider, Fluid Phase Equilibr., 1983, 10, 141. metallic alloys (e.g. Cu–Pd), but also for the metallic coating 18 M. PoliakoV, M. W. Georges and S. M.Howdle, Chemistry under of particles used as substrates and therefore insoluble in the extreme or non-classical conditions, ed. R. V. Eldik and C. D. supercritical fluid medium. Hubbard, Wiley, New York, 1996, pp. 189–219. As far as the preparation of metastable phases is concerned: 19 P. Beslin, F. Cansell, Y. Garrabos, G. Demazeau, B. Berdeu and D. Sentagnes, De�chets, 1997, 5, 17.(i) the high diVusivity of O2-/OH- species (in supercritical 74 J. Mater. Chem., 1999, 9, 67–7520 D. T. Chen, A. P. Craig, E. Reichert and J. Hoven, J. Hazard. 40 D. F. McLaughlin and M. C. Skriba, US Pat. 4,916,108, 1990. Mater., 1995, 44, 53. 41 F. Lu, S. Chen and S. Peng, Catal. Today, 1996, 30, 183. 21 H. Lentz and W. Mormann, Makromol. Chem., Macromol. Symp., 42 J.N. Armor and E. J. Carlson, US Pat. 4,615,736, 1986. 1992, 57, 305. 43 G. L. Schimek, W. T. Pennington, P. T. Wood and J. W. Kolis, 22 K. P. Johnston, K. L. Harrison, M. J. Clarke, S. M. Howdle, J. Solid State Chem., 1996, 123, 277. M. P. Heitz, F. V. Bright, C. Carlier and T. W. Randolph, Nature, 44 J. J.Watkins and T. J. McCarthy, Chem. Mater., 1995, 7, 1991. 1996, 271, 624. 45 D. A. Loy, E. M. Russick, S. A. Yamanaka and B. M. Baugher, 23 J. P. Petitet, Fluides Supercritiques et Mate�riaux, ed. F. Cansell Chem. Mater., 1997, 9, 2264. and J. P. Petitet, AIPFS Publishing, Nancy, France, 1995, 46 B. N. Hansen, B. M. Hyberston, R. M. Barkley and R. E. Sievers, pp. 251–300. Chem. Mater., 1992, 4, 749. 24 C. A. Eckert, B. L. Knuston and P. G. Debenedetti, Nature, 1996, 47 O. A. Louchev, V. K. Popov and E. N. Antonov, J. Cryst. Growth, 383, 313. 1995, 155, 276. 25 J. Chrastil, J. Phys. Chem., 1982, 86, 3016. 48 V. K. Popov, V. N. Bagratashvili, E. N. Antonov and D. A. 26 P. E. Savage, S. Gopalan, T. I. Mizian, C. J. Martino and E. E. Lemenovski, Thin Solid Films, 1996, 279, 66. Brock, AIChE J., 1995, 41, 1723. 49 P. G. Debenedetti, AIChE J., 1990, 36, 1289. 27 F. Cansell, S. Rey and P. Beslin, Rev. Inst. Fr. Pe� trole, 1998, 53, 71. 50 S. K. Kumar, J. Supercritical Fluids, 1988, 1, 15. 28 J. W. Tom and P. G. Debenedetti, J. Aerosol Sci., 1991, 22, 555. 51 D. V. Baxter, M. H. Chisholm, G. J. Gama and V. F. DiStasi, 29 D. W. Matson and R. D. Smith, J. Am. Ceram. Soc., 1989, 72, 871. Chem. Mater., 1996, 8, 1222. 30 H. Ksibi, P. Subra and Y. Garrabos, Adv. Powder Technol., 1995, 52 H. Jacobs and J. Bock, J. Less-Common Met., 1987, 134, 215. 6, 25. 53 R. Juza and H. Hahn, Z. Anorg. Allg. Chem., 1939, 241, 172. 31 B. Subramaniam, R. A. Rajewski and K. Snavely, J. Pharm. Sci., 54 S. Peter, F. Altofer, W. Bu� hrer and H. D. Lutz, VIth European 1997, 86, 885. Conference on Solid State Chemistry, Zu� rich, Sept. 1997. 32 D. W. Matson, J. L. Fulton, R. C. Petersen and R. D. Smith, Ind. 55 F. Kamoun, M. Lorenz, G. Kempe and F. Peger, J. Less-Common Eng. Chem. Res., 1987, 26, 2298. Metals, 1991, 170, 1. 33 E. Reverchon, Supercritical fluids—Materials and natural products 56 K. J. Gallagher and M. R. Ottaway, J. Chem. Soc., Dalton Trans., processing, ed. M. Perrut and P. Subra, AIPFS Publishing, Nancy, 1974, 2347. France, 1998, pp. 221–236. 57 P. Keller, N. Jb. Miner. Abh., 1970, 113, 29. 34 M. PoliakoV and M. W. George, Supercritical Fluids: Materials 58 P. Hagenmuller, J. Portier, J. Cadiou and R. de Pape, C. R. Acad. and Natural Products Processing, ed. M. Perrut and P. Subra, Sci. Paris, 1965, 260, 4768. AIPFS Publishing, Nancy, France, 1998, pp. 833–842. 59 M. Vlasse, J. C. Massies and G. Demazeau, J. Solid State Chem., 35 E. Dinjus, R. Fornika and M. Scholtz, Chemistry under extreme or 1973, 8, 109. non-classical conditions, ed. R. V. Eldik and C. D. Hubbard, 60 A. L. Mackay, Mineral. Mag., 1960, 32, 545. Wiley, New York, 1996, pp. 219–272. 61 J. M. Gonzalez-Calbet, M. A. Alario-Franco and M. Gayoso- 36 V. Gourinchas-Courtecuisse, K. Chhor, J. F. Bocquet and C. Andrade, J. Inorg. Nucl. Chem., 1981, 43, 257. Pommier, Ind. Eng. Chem. Res., 1996, 35, 2539. 62 J. M. Combes, A. Manceau, G. Calas and J. Y. Bottero, Geochim. 37 V. Gourinchas-Courtecuisse, J. F. Bocquet, K. Chhor and C. Cosmochim. Acta, 1989, 53, 583. Pommier, J. Supercritical Fluids, 1996, 9, 222. 63 R. So� derquist and S. Jansson, Acta. Chem. Scand., 1966, 20, 1417. 38 Y. Hakuta, H. Terayama, S. Onai, T. Adschiri and K. Arai, 64 G. Biedermann and J. T. Chow, Acta Chem. Scand., 1966, 20, Supercritical Fluids—Chromatography and Novel Applications, ed. 1376. S. Saito and K. Arai, AIPFS Publishing, Nancy, France, 1997, 65 J. E. Post and V. F. Buchwald, Am. Mineral., 1991, 76, 272. pp. 255–258. 66 M. Ohyabu and Y. Ujihira, J. Inorg. Nucl. Chem., 1981, 43, 3125. 39 T. Adschiri, S. Yamane, S. Onai and K. Arai, Supercritical 67 E. Murad, Clay Miner., 1979, 14, 273. Fluids—Reactions, Material science and Chromatography, ed. M. Perrut and G. Brunner, AIPFS Publishing, Nancy, France, 1994, vol. 3, pp. 241–246. Paper 8/04964E J. Mater. Chem., 1999, 9, 67

ISSN:0959-9428    

DOI:10.1039/a804964e

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Recent developments in soft, solution processing: one step fabrication of functional double oxide films by hydrothermal–electrochemical methods† Masahiro Yoshimura,*Wojciech Suchanek and Kyoo-Seung Han Center for Materials Design, Materials and Structures Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta, Midori, Yokohama 226, Japan. Received 14th April 1998, Accepted 18th June 1998 Based upon thermodynamic considerations, we propose ‘soft, solution processing’, as a means to fabrication of shaped, sized, and controlled advanced materials from aqueous solutions without excess heat, energy consumption, expensive equipment, and precursor(s) as one of the most economical and environmentally friendly preparation techniques for advanced materials.We have succeeded in the fabrication of double oxide thin films such as BaTiO3, SrTiO3, their solid solutions, and LiNiO2 at fixed temperatures between 60 and 150 °C by hydrothermal–electrochemical methods. Nucleation and growth of the films takes place at the interface between the alkaline (earth) component in the solutions and anodically treated metal plates. The structures and properties of these films are described.A novel method to fabricate BaTiO3/SrTiO3 layered thin films by changing the solution compositions in flow-cell equipment is presented. defined as ‘environmentally friendly processing using (aqueous) Introduction solutions’. In this paper a thermodynamic and environmental Modern society needs advanced ceramic materials with background to SSP is given.In addition, recent important sophisticated functions and high performances. A number of developments in the field of SSP, such as fabrication of double substances with particular compositions, crystal structures and oxide thin films including multilayered materials and specified properties have been investigated for this purpose. applications of a new solution flow system, are presented.[In our considerations ‘substance’ is characterized by (1) a particular chemical composition, (2) a particular physical state Thermodynamic and environmental features of soft, including structure, and (3) particular properties based upon its composition and structure. Since ‘material’ should be solution processing defined as a substance mostly in the solid state which is used Generally speaking, SSP gives similar results to any other or to be used for certain application(s), we must add the processing using fluids (such as vapor, gas, plasma) and/or following features to define ‘material’: (4) a particular beam/vacuum processing.However, the total energy consump- shape, (5) a certain size, (6) location, and (7) orientation.1] tion of all these processing routes should be the lowest in Unfortunately, it is diYcult to give desired shapes, forms and aqueous systems, because a much larger excess of energy is sizes to inorganic materials, owing to their high brittleness.necessary to create melts, vapor, gas or plasma than to form Organic materials such as polymers and plastics or metallic an aqueous solution at the same temperature.1 This idea can materials can be generally deformed when local stresses (above be easily demonstrated using the simple example of BaTiO3.their yield stresses) are applied to them, but inorganic mate- The energy diagram for the formation of BaTiO3 from various rials, particularly ceramics, tend to break due to brittle fracprecursors is shown in Fig. 1. The driving forces (DG) for ture.2–4 Therefore, ceramics are fabricated by rather special representative synthesis reactions of (1), (2), (3), (4), and (5) ‘ceramic processing’ which consist of two steps: (1) synthesis are 38, 727, 3685, 17, and -14 kcal mol-1, respectively at of powders, and (2) shape-forming by firing/sintering of the room temperature.8,9 powders or melting (in the case of glasses).5 Both steps usually require high temperatures and thus consume a lot of energy.BaO (crystal )+TiO2 (crystal )=BaTiO3 (crystal ) (1) Novel processing using a gaseous phase, like CVD, MOCVD, Ba (vapor)+Ti (vapor)+3/2O2 (gas)=BaTiO3 (crystal ) (2) etc., or vacuum systems such as sputtering, MBE etc., require even higher energy than standard high temperature pro- Ba2+ (gas)+Ti4+ (gas)+3O2- (gas)=BaTiO3 (crystal ) (3) cessing.6 All these techniques have resulted in environmental TiO2 (crystal )+Ba2+ (aq)+2OH- (aq)-H2O (aq)= problems because the consumed energies are emitted as exhaust BaTiO3 (crystal ) (4) gas(es) or exhaust heat (entropy) except for the part involved in the product.In particular, vacuum systems seem to be Ba2+ (aq)+Ti(OH)4 (aq)-H2O (aq)-2H+ (aq)= worse because they need continuous pumping to maintain BaTiO3 (crystal ) (5) vacuum and their exhaust gas(es) cannot be cycled due to their diluted large volumes.This means that any processing using gaseous ions requires a Fortunately, we can fabricate shaped, sized, reacted and/or huge activation energy of 727–3685 kcal mol-1 to make solid oriented ceramic materials, without firing/sintering or melting, BaTiO3 and this energy must be discarded into the environwithout vacuum and expensive equipment, in situ, by soft, ment, because the raw materials of Ba and Ti must be solid solution processing (denoted hereafter as SSP).1,7 SSP is oxide(s) or carbonate(s) ore.On the other hand, since the lattice energy of BaO and TiO2 is almost equal to the hydration (solvation) energy of Ba2+ and Ti4+ ions, solution processing †Basis of the presentation given at Materials Chemistry Discussion consumes very little energy (DG, driving force) if only the No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. synthesis activation energy (DG*) can be overcome. Fig. 1 J. Mater. Chem., 1999, 9, 77–82 77ceramic films by solution processing have been limited except for the ‘sol–gel’ or ‘pyrolysis’ methods,10,11 which are regarded as solid rather than solution methods because solvents must be driven out of the system before formation of the crystalline material.In spite of the environmental and energetic advantages of SSP, film formation of multicomponent inorganic materials is quite diYcult from solution systems.In contrast, film formation is rather easy from gas or plasma systems. The gaseous species are always in high energy states (see also Fig. 1), thus the DG and the DG* are suYciently provided for reaction to yield crystalline compounds with desired shape/size via several steps such as diVusion, adsorption, reaction, nucleation, and growth.6 Generally speaking, DG*3(DG)-n, where n=2 for the nucleation process.On the other hand, species in aqueous solutions are hydrated (or chelated by some complexing agents), thus they have only a small DG for the reaction, thus rather high DG* values are necessary for the reaction to occur by overcoming the hydration (chelation) energies of ions. Electro- or electroless-plating of metals is achieved by reducing of metal ion(s) electrochemically or chemically.12 However, in the case of ceramic materials, particularly multicomponent ones, special activation of all the components is needed, anions must be oxidized at the same time as reduction of cations.Since this simultaneous multiactivation is quite diYcult, ceramics could scarcely be fabricated directly from solutions previously.However, we have demonstrated that direct fabrication of ceramic films can be achieved using interfacial Fig. 1 Energy diagram for the formation of BaTiO3 using solid and/or reactions between reactive substrates of metal, alloy, or oxide gas precursors (g, s and c correspond to the gaseous state solution state and the crystalline state, respectively). Compiled from ref. 8 and 9. and species in aqueous solution activated thermally (hydrothermal ) electrochemically, or by a combination (hydrothermal –electrochemical ).7 Other activation methods like photo-, gives the fundamental thermodynamic background for the sono-, electro-, mechano-, complexo-, or even bio-activation environmental problems related to materials processing. are possible to accelerate the interfacial reactions.1,13 A temperature–pressure map for various kinds of materials processing is shown in Fig. 2. Solution processing is located in the p–T range characteristic of conditions of living on earth. All other processing routes are associated with increasing Applications of soft, solution processing for thin film temperature and/or increasing (or decreasing) pressure, fabrication: hydrothermal–electrochemical technique therefore they are environmentally stressed.From the environ- Preparation techniques of thin films can be divided into two mental point of view, solution processing should be the most categories: (1) dry methods, such as vacuum deposition, environmentally friendly. Nevertheless, previous studies on sputtering, ion plating, CVD etc.,6 and (2) wet methods such as dip, spray coating of organic/inorganic precursors, sol–gel,10,11 electrodeposition,14 etc.All these methods usually require high temperatures (above 500 °C) to crystallize asdeposited amorphous films or to decompose the precursor to yield an appropriate compound. Such heat treatments often result in cracking and/or peeling of the deposited layers, formation of undesired compositions by evaporation, dissociation, and/or reaction of the film with the substrate, moreover they are energy consuming.In particular, it is diYcult to form dense films by organometallic and metallorganic processing, because decomposition of the precursors may result in porous ceramic layers. Therefore, low temperature, in situ fabrication of crystalline, thin films is essential to improve their quality, lower costs and make the whole process environmentally friendly.This was accomplished, by introducing the hydrothermal and/or electrochemical methods of film preparation.15 These techniques allow various double oxide films to be prepared on such substrates as metal, alloy or oxide.7 The substrate works as an electrode which reacts chemically and/or electrochemically (by passing electrical current) with the cation components of the surrounding solution. This processing route may involve the following reactions:16 (1) hydrothermal reaction between anodically oxidized film and solution components; (2) electrochemical reaction with solution component(s); (3) electrolytic oxidation of dissolved species and hydrothermal reaction; (4) hydrothermal reaction between anodically dissolved species Fig. 2 Schematic p–T diagram for preparative techniques for advanced ceramic materials. and solution components; (5) electro-discharge reaction 78 J. Mater. Chem., 1999, 9, 77–82between anodic oxidized film with solution component(s); (6) electro-deposition of charged oxide particles. Advantages of the hydrothermal/electrochemical techniques are: (1) one step (direct) formation of shaped/sized/deposited/ oriented ceramics; (2) low fabrication temperatures (minimized consumption of energy); (3) any shape, any size; (4) possibility of using a closed flow system, allowing easy charging, separation, cycling and recycling; (5) relatively high growth rate of the films (up to several mm per hour); (6) versatility.In addition, SSP gives products with much higher homogeneity than solid state processing and with higher density than gas or vacuum processing (faster growth rate). There is a wide variety of combinations of solvent/solute systems. Liquids may be beneficial for charging, transportation, mixing and/or separation of products. They are completely adaptable for cycling/recycling of material processing in closed systems. Moreover, liquids give the possibility for acceleration of diVusion, adsorption, reaction rate and crystallization Fig. 3 Cyclic voltammograms of the LiNiO2 film prepared at 125 °C (nucleation and growth), especially under hydrothermal with a current density of 1.0 mA cm-2 taken in 0.1 M LiClO4 propylene conditions. carbonate (potential is referred to Li/Li+ electrode) with a scan rate For these reasons, various thin films have been successfully of 5.0 mV s-1.prepared by the hydrothermal and/or electrochemical method. Examples include polycrystalline PZT,17,18 ZnO,19 TiO2,20 recycled by filtering it and replenishing the LiOH, which is ferrites,21–23 PbS,24 CdSe,24 BaTiO3,15,25–31 SrTiO3,32–36 environmentally and economically beneficial.BaxSr1-xTiO3,37 CaTiO3,38 BaFeO4 (BaFeO3-x),16 LiNbO3,16 The electrochemical–hydrothermal treatment of polished PbTiO3,39 ABO4 oxides (A=Ca, Ba, Sr; B=W, Mo)40,41 and nickel plates led to the formation of a visually detectable complete series of their solid solutions.42 The microstructures cobalt-blue film. The structural purity and chemical composiand chemical compositions of these materials can be controlled tion of the films were confirmed by X-ray diVraction and Xin a wide range at temperatures in most cases not exceeding ray photoelectron spectroscopy, respectively.54–56 According 200 °C, sometimes even at room temperature.40–42 The fabri- to X-ray diVraction, scanning electron microscopy (SEM) cation of epitaxial thin films,26,43 monomolecular layers,24 and cross-section views, and atomic force microscopy (AFM) even superlattices44,45 have also been reported.surface images, the films show good crystallinity despite the In the following two sections, new developments in the field low fabrication temperature without any post-synthesis of SSP are described. In both sections the experimental pro- annealing.54–56 Further evidence for the formation of LiNiO2 cedure and advantages of the applied techniques over pre- films was obtained from their cyclic voltammograms (Fig. 3) viously used ones for the particular material are also briefly since the peak potentials characterizing oxidation and mentioned. reduction are similar to those for a LiNiO2 powder electrode.55 Films prepared between 125 and 175 °C can also be reversibly oxidized and reduced in lithium perchlorate (LiClO4) propylene carbonate solutions for >5 electrochemical cycles, and Hydrothermal–electrochemical synthesis of LiNiO2 therefore are possible cathodes for lithium rechargeable films microbatteries.All-solid-state lithium rocking chair secondary microbatteries The films prepared under diVerent conditions show diVerent have been considered as the most suitable power source for electrochemical activity, surface morphology, and film thickmicroelectronics. 46–51 However, the synthetic approach in pre- ness.55,56 The apparent eVect of the fabrication temperature vious works required highly sophisticated multistep procedures on the surface morphology, for instance, is demonstrated by such as chemical vapor deposition, sol–gel methods, and comparing the AFM surface images of the nickel substrate sputtering.46–53 Such fabrication routes require high energy and the films prepared at diVerent temperatures (Fig. 4). and materials consumption, expensive precursors, and While purely hydrothermal treatment of nickel substrates complicated instrumentation.leads to the formation of Ni(OH)2 films, by use of the In the course of trying to develop an alternative, economic, electrochemical–hydrothermal approach under supplementary and low-temperature process that can produce the desired galvanostatic charge with the same hydrothermal conditions, LiNiO2 films can only eVectively be prepared in a single lithiated cathode films on chips, we have succeeded in the synthetic step from nickel substrates.54–56 In addition, the electrochemical preparation of active LiNiO2 films by electroelectrochemical oxidation of nickel plates in alkaline solution chemical–hydrothermal treatment on nickel plates.54–56 This at room temperature leads to the formation of NiOOH films.fabrication can be interpreted in terms of the electrochemical Therefore, the electrochemical anodic process is necessary to oxidation of nickel metal electrodes in a concentrated LiOH obtain trivalent nickel and the hydrothermal process is solution at fixed temperatures between 60 and 200 °C.The necessary for the cationic exchange reaction between Li+ electrochemical process, i.e. electrochemical oxidation of nickel and H+.plates, was realized galvanostatically with a fixed current density between 0.1 and 10 mA cm-2 in a three-electrode arrangement using nickel working and auxiliary electrodes and Multilayered thin films in the BaTiO3–SrTiO3 a temperature controlled Ag/AgCl reference electrode.57 The system: integration issues of soft, solution processing reaction time depends on the applied current density and fabrication temperature.Although detailed experimental It has been shown in the previous sections that a variety of procedures have been described previously,54–56 note that the technologically important thin films can be prepared by hydroprepared LiNiO2 films were not subjected to any post-synthesis thermal and/or electrochemical techniques. It seems therefore heat treatment.The used LiOH solution containing dissolved that we should start to think seriously about how to integrate the hydrothermal/electrochemical methods with functional nickel species, nickelite ions HNiO2- in practice, can be J. Mater. Chem., 1999, 9, 77–82 79Fig. 5 General schematic diagram of the flow cell for hydrothermal– electrochemical synthesis. disadvantages such as high vacuum and/or temperatures in the range of 500–700 °C.All the experiments were accomplished in our flow cell for hydrothermal–electrochemical synthesis. A general schematic diagram of the equipment is shown in Fig. 5. Single-phase, single layer and multilayered thin films in the BaTiO3–SrTiO3 system have been fabricated using titanium substrate and Ba and Sr acetates.A detailed description of the fabrication procedure has been published elsewhere.67–69 Results of syntheses carried out in open and closed flow were almost the same. Single phase BaTiO3, SrTiO3 or Fig. 4 Atomic force microscope (AFM) images of (a) the Ni substrate BaxSr1-xTiO3 have been prepared at 120–200 °C, when the polished with 1.0 mm diamond paste, and (b)–(f ) the LiNiO2 films current density did not exceed 1 mA cm-2.EVects of flow rate prepared at diVerent temperatures with a current density of on grain size are shown in Fig. 6. The grain size increases 1.0 mA cm-2: (b) 100 °C, (c) 125 °C, (d) 150 °C, (e) 175 °C and almost linearly with increasing flow rate, reaching a plateau (f ) 200 °C. Image size is 7.5 mm×7.5 mm in each case. at approximately 20 cm3 min-1. EVects of the flow rate in our experiments are in fact similar to the eVects of stirring applied in growth of crystals from solutions or melts.Usually, they lead to an increase in the probability of spontaneous device technology which has been dominated by physical vapor nucleation, reduction of supersaturation inhomogeneities, and deposition6 and chemical vapor deposition.6 In addition to increase of the growth rate.70 The growth rate of crystals PVD and CVD, some processing routes aimed at fabrication increases with increasing solution flow rate until a limiting of integrated circuits use chemical solution processing (etching, rate is reached when the growth rate becomes controlled by sol–gel ).58 In some cases, the (electro)chemical routes are considered as alternatives for the presently used vapor techniques (e.g.copper deposition).59 From this point of view, fabrication of ceramic thin films in the solution flow below 200 °C (this is the maximum temperature for wide industrial applications of the hydrothermal technique) is very important and may find multiple applications. The best suited equipment for this purpose seems to be a solution flow system for hydrothermal–electrochemical synthesis.Such equipment is similar to flow cells used in geological studies60 or in materials engineering (ferrite plating), 21,22 and allows fabrication of multilayered thin films by simply changing the flowing solution and/or adjusting the processing conditions. Moreover the flow can be closed, enabling easy recycling of the solutions.We have selected the BaTiO3–SrTiO3 system to demonstrate the fabrication of multilayered ceramic films under solution flow. Recently, attention has focused on multilayered films in Fig. 6 Average grain size of the BaTiO3, SrTiO3, and BaxSr1-xTiO3 this system,61–66 as they may find applications as DRAM single-phase thin films prepared in the flow cell by the hydrothermal– capacitors with high relative permittivity and reduced leakage electrochemical method at 150 °C (1 h, 1mAcm-2) as a function of current,62–64 tunable multilayer capacitors, waveguide phase the solution flow rate.Error bars denote standard deviations for shifters, filters, etc.66 The previous preparation routes to the average values. Solution flow rates were in the range of BaTiO3–SrTiO3 multilayered structures, such as tape-casting,66 1–50 cm3 min-1 which corresponds to the Reynolds numbers of about 46–2300, within a region of laminar flow.67 dip-coating,65 and magnetron sputtering62–64 had several 80 J.Mater. Chem., 1999, 9, 77–82environmentally friendly. Several advanced ceramic materials have been successfully fabricated using this technique.Well crystallized, electrochemically active LiNiO2 films for lithium rechargeable microbatteries were fabricated in a single step by the electrochemical–hydrothermal treatment of nickel plates in a concentrated LiOH solution at fixed temperatures between 60 and 200 °C without any post-synthesis annealing. The film properties show the obtained LiNiO2 films to be as suitable cathode films.Layered thin films in the BaTiO3–SrTiO3 system have been prepared in a solution flow system by the hydrothermal– electrochemical method in a single step by simply changing the flowing solution which is its main advantage over closed autoclaves. Changing the flow rate allows additional control of film morphology by enhancing the growth rate. This processing route may serve as an inexpensive and environmentally friendly way to fabricate functionally graded thin films and seems to be an important step in the application of the hydrothermal/electrochemical techniques to fabrication of integrated electronic devices.SSP allows fabrication in aqueous solution of shaped/sized/ oriented ceramics in one step, without firing/sintering or melting. This technique requires simple equipment and low temperatures.The morphology and chemical composition of Fig. 7 Atomic force microscopy images of the SrTiO3/BaTiO3/Ti multilayer (a) after step 1, i.e. the BaTiO3 surface, (b) after step 2, the resulting ceramics can be easily controlled by adjusting the i.e. the SrTiO3 surface; and the BaTiO3/SrTiO3/Ti multilayer (c) after processing conditions.SSP provides an inexpensive and step 1, i.e. the SrTiO3 surface, (d) after step 2, i.e. the BaTiO3 surface. environmentally friendly route to advanced ceramic materials, Image size is 1.0×1.0 mm in each case.67 and therefore is one of the most promising technologies for the 21st century. the interfacial kinetic processes.70 Our results suggest that the This research was supported by the ‘Research for the Future’ microstructure of the films at various flow rates is controlled Program No. 96R06901 of the Japanese Society for the by the growth rate, which initially increases, and then remains Promotion of Science (JSPS). The authors are greatly indebted unchanged for flow rates >20 cm3 min-1. to Prof. M. Kakihana, Prof. M. Yashima, Prof.M. Abe (all SrTiO3/BaTiO3/Ti and BaTiO3/SrTiO3/Ti layered thin films Tokyo Institute of Technology, Japan) for stimulating dis- have been fabricated at 150 °C (step 1: deposition of the first cussions and to Dr. P. Krtil, Mr. B. Sakurai, Mr. T.Watanabe, layer+step 2: deposition of the second layer), by changing the Miss. S-W. Song, Mr. S. Tsurimoto, and Mr. N. Kumagai (all flowing solution.XRD patterns of the double layers showed Tokyo Institute of Technology, Japan) for experimental distinct peaks derived from single-phase SrTiO3 and BaTiO3 assistance. layers.67 The presence of BaxSr1-xTiO3 solid solutions has not been detected by XRD, but we cannot preclude dissolution of small quantities of Ba in the SrTiO3 layers or vice versa. References Results of AFM measurements are shown in Fig. 7.They clearly show diVerent microstructures of the surface of the 1 M. Yoshimura, J. Mater. Res., 1998, 13, 796. single layers and the double layers. These data demonstrate 2 D. W. van Krevelen, Properties of Polymers, Elsevier Science Publishers, Amsterdam, 1990. that initially formed BaTiO3 and SrTiO3 layers have been 3 R. A. Higgins, Engineering Metalurgy, Edward Arnold, London, covered by layers of SrTiO3 and BaTiO3, respectively.XPS 1993. data confirmed presence of the Sr- or Ba-rich layer on the 4 W. D. Kingery, H. K. Bowen and D. R. Uhlmann, Introduction to surface and the Ba- or Sr-rich layer, respectively, below.67,71 Ceramics, John Wiley & Sons, New York, 1976. The formation mechanism of the multilayered thin films in 5 R.A. Ring, Fundamentals of Ceramic Powder Processing and the BaTiO3–SrTiO3 system by the hydrothermal–electrochem- Synthesis, Academic Press, San Diego, 1996. 6 M. Ohring, The Materials Science of Thin Films, Academic Press, ical technique will be discussed in detail elsewhere.72 New York, 1992. The processing described may be applied generally for 7 M. Yoshimura and W.Suchanek, Solid State Ionics, 1997, 98, 197. various single and/or multilayered thin films. The possibility 8 JANAF Thermochemical Tables, National Bureau of Standards, of solution recycling during hydrothermal–electrochemical Ametican Chemical Society, American Institute of Physics, synthesis of the thin films in the BaTiO3–SrTiO3 system is 1968–1982. important because it minimizes the environmental impact of 9 M.M.Lencka and R. E. Riman, Chem. Mater., 1993, 5, 61. 10 Sol–Gel Technology for Thin Films, Fibers, Preforms, Electronics, the process.1 The technique using the solution flow under and Specialty Shapes, ed. L. Klein, Noyes Publ., Mill Road, Park hydrothermal–electrochemical conditions is an important step Ridge, New Jersey, 1988. in integration of the solution processing with functional device 11 Sol–Gel Optics. Processing and Applications, ed.L. Klein, Kluwer technology and may find applications as an inexpensive and Academic Publishers, Boston/Dordrecht/London, 1994. environmentally friendly way to fabricate various single and/or 12 Modern Electroplating, ed. F. A. Lowenheim, John Wiley & Sons, multilayered thin films and functionally graded materials.Inc., New York, 1963. 13 R. Roy, J. Solid State Chem., 1994, 111, 11. 14 C. D. Lokhande and S. H. Pawar, Phys. Status Solidi A, 1989, Summary 111, 17. 15 M. Yoshimura, S.-E. Yoo, M. Hayashi and N. Ishizawa, Jpn. A thermodynamic concept for energetic and environmental J. Appl. Phys., 1989, 28, L2007. problems related to materials synthesis has been proposed.It 16 M. Yoshimura, Mater. Res. Soc. Symp. Proc., 1992, 271, 457. concludes that soft, solution processing, i.e. solution processing 17 Y. Ohba, M. Miyauchi, E. Sakai and M. Daimon, Jpn. J. Appl. Phys., 1995, 34, 5216. using (aqueous) solutions, should be the most economical and J. Mater. Chem., 1999, 9, 77–82 8118 A. T. Chien, J. S. Speck and F. F. Lange, J.Mater. Res., 1997, 47 P. Birke, W. F. Chu and W. Weppner, Solid State Ionics, 1997, 93, 1. 12, 1176. 48 K. A. Striebel, C. Z. Deng, S. J. Wen and E. J. Cairns, 19 Q. Chen, Y. Qian, Z. Chen, G. Zhou and Y. Zhang, Mater. Lett., J. Electrochem. Soc., 1996, 143, 1821. 1995, 22, 93. 49 S. J. Lee, J. K. Lee, D. W. Kim, H. K. Baik and S. M. Lee, 20 Q. W. Chen, Y. T. Qian, Z. Y. Chen, Y.B. Jia, G. E. Zhou, J. Electrochem. Soc., 1996, 143, L268. X. G. Li and Y. H. Zhang, Phys. Status Solidi A, 1996, 156, 381. 50 K. H. Hwang, S. H. Lee and S. K. Joo, J. Electrochem. Soc., 1994, 21 M. Abe, Y. Tamaura, Y. Goto, N. Kitamura and M. Gomi, 141, 3296. J. Appl. Phys., 1987, 61, 3211. 51 F. K. Shokoohi, J. M. Tarascon and B. J. Wilkens, Appl. Phys. 22 M. Abe, T. Itoh and Y.Tamaura, Mater. Res. Soc. Symp. Proc., Lett., 1991, 59, 1260. 1991, 232, 107. 52 M. Antaya, K. Cearns, J. S. Preston, J. N. Reimers and 23 J. S. Lee, T. Itoh and M. Abe, J. Korean Phys. Soc., 1995, 28, 375. J. R. Dahn, J. Appl. Phys., 1994, 76, 2799. 24 J. H. Fendler and F. C. Meldrum, Adv. Mater., 1995, 7, 607. 53 M. Antaya, J. R. Dahn, J. S. Preston, E. Rossen and 25 M.Yoshimura, S.-E. Yoo, M. Hayashi and N. Ishizawa, Ceram. J. N. Reimers, J. Electrochem. Soc., 1993, 140, 575. Trans., 1990, 15, 427. 54 M. Yoshimura, K. S. Han and S. Tsurimoto, Solid State Ionics, 26 F. F. Lange, Science, 1996, 273, 903. 1998, 106, 39. 27 E. B. Slamovich and I. A. Aksay, J. Am. Ceram. Soc., 1996, 55 K. S. Han, P. Krtil and M. Yoshimura, J. Mater. Chem., 1998, 79, 239. 8, 2043. 28 M. E. Pilleux and V. M. Fuenzalida, J. Appl. Phys., 1993, 74, 4664. 56 K. S. Han, S. W. Song and M. Yoshimura, Chem. Mater., 1998, 29 T. Vargas, H. Diaz, C. I. Silva and V. M. Fuenzalida, J. Am. 10, 2183. Ceram. Soc., 1997, 80, 213. 57 D. D. Macdonald and D. Owen, J. Electrochem. Soc., 1973, 120, 30 R. R. Bacsa, G. Rutsch and J. P. Dougherty, J. Mater. Res., 1996, 317. 11, 194. 58 H. L. Tuller, Ceram. Trans., 1996, 68, 97. 31 P. Bendale, S. Venigalla, J. R. Ambrose, E. D. Verink Jr. and 59 R. J. Gutmann, T. P. Chow, S. Lakshminarayanan, D. T. Price, J. H. Adair, J. Am. Ceram. Soc., 1993, 76, 2619. J. M. Steigerwald, L. You and S. P. Murarka, Thin Solid Films, 32 S.-E. Yoo, M. Hayashi, N. Ishizawa and M. Yoshimura, J. Am. 1995, 270, 472. Ceram. Soc., 1990, 73, 2561. 60 J. M. Porter, D. C. Pohl and J. D. Rimstid, in Hydrothermal 33 K. Kajiyoshi, Y. Hamaji, K. Tomono, T. Kasanami and Experimental Techniques, ed. G. C. Ulmer and H. L. Barnes, John M. Yoshimura, J. Am. Ceram. Soc., 1996, 79, 613. Wiley & Sons, New York, 1987, p. 240. 34 K. Kajiyoshi, K. Tomono, Y. Hamaji, T. Kasanami and 61 M. Suzuki, J. Ceram. Soc. Jpn., 1995, 103, 1099 (in Japanese).M. Yoshimura, J. Am. Ceram. Soc., 1995, 78, 1521. 62 M.-H. Song, Y.-H. Lee, T.-S. Hahn and M.-H. Oh, J. Appl. Phys., 35 K. Kajiyoshi and M. Yoshimura, Eur. J. Solid State Inorg. Chem., 1996, 79, 3744. 1996, 33, 623. 63 Q. X. Jia, L. H. Chang and W. A. Anderson, Thin Solid Films, 36 M. E. Pilleux, C. R. Grahmann and V. M. Fuenzalida, J. Am. 1995, 259, 264. Ceram.Soc., 1994, 77, 1601. 64 Z. Q. Shi, Q. X. Jia and W. A. Anderson, J. Electron. Mater., 37 K. Kajiyoshi, M. Yoshimura, Y. Hamaji, K. Tomono and 1991, 20, 939. T. Kasanami, J. Mater. Res., 1996, 11, 169. 65 T. Hayashi and T. Tanaka, Jpn. J. Appl. Phys., 1995, 34, 5100. 38 M. Yoshimura, W. Urushihara, M. Yashima and M. Kakihana, 66 V. K. Varadan, V. V. Varadan, F. Selmi, Z. Ounaies and Intermetallics, 1995, 3, 125.K. A. Jose, Proc. SPIE—Int. Soc. Opt. Eng., 1994, 2189, 433. 39 W.-S. Cho and M. Yoshimura, J. Mater. Res., 1997, 12, 833. 67 M. Yoshimura, W. Suchanek, T. Watanabe, B. Sakurai and 40 W. S. Cho and M. Yoshimura, Jpn. J. Appl. Phys., 1997, 36, 1216. M. Abe, J. Mater. Res., 1998, 13, 875. 41 W. S. Cho, M. Yashima, M. Kakihana, A. Kudo, T. Sakata and 68 W.Suchanek, T.Watanabe and M. Yoshimura, Solid State Ionics, M. Yoshimura, J. Am. Ceram. Soc., 1997, 80, 765. 1998, 109, 65. 42 W. S. Cho, M. Yashima, M. Kakihana, A. Kudo, T. Sakata and 69 W. Suchanek and M. Yoshimura, J. Am. Ceram. Soc., 1998, in M. Yoshimura, Appl. Phys. Lett., 1996, 68, 137. press. 70 D. Elwell and H. J. Scheel, Crystal Growth from High-Temperature 43 A.T. Chien, L. Zhao, M. Colic. J. S. Speck and F. F. Lange, Solutions, Academic Press, New York, 1975. J. Mater. Res., 1998, 13, 649. 71 W. Suchanek, T. Watanabe, B. Sakurai and M. Yoshimura, 44 J. A. Switzer, C. J. Hung, B. E. Breyfogle, M. G. Shumsky, Mater. Res. Soc. Symp. Proc., 1998, 517) High-Density Magnetic R. Vanleeuwen and T. D. Golden Science, 1994, 264, 1573. Recording and Integrated Magneto-Optics: Materials and 45 J.A. Switzer, C. J. Hung, L. Y. Huang, F. C. Miller, Y. Zhou, Devices), in press. E. R. Raub, M. G. Shumsky and E. W. Bohannan, J. Mater. Res., 72 W. Suchanek and M. Yoshimura, unpublished data. 1998, 13, 909. 46 F. K. Shokoohi, J. M. Tarascon, B. J. Wilkens, D. Guyomard and C. C. Chang, J. Electrochem. Soc., 1992, 139, 1845. Paper 8/04647F 82 J.Mater. Chem., 1999, 9, 77–82 J O U R N A L O F C H E M I S T R Y Materials Recent developments in soft, solution processing: one step fabrication of functional double oxide films by hydrothermal–electrochemical methods† Masahiro Yoshimura,*Wojciech Suchanek and Kyoo-Seung Han Center for Materials Design, Materials and Structures Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta, Midori, Yokohama 226, Japan.Received 14th April 1998, Accepted 18th June 1998 Based upon thermodynamic considerations, we propose ‘soft, solution processing’, as a means to fabrication of shaped, sized, and controlled advanced materials from aqueous solutions without excess heat, energy consumption, expensive equipment, and precursor(s) as one of the most economical and environmentally friendly preparation techniques for advanced materials.We have succeeded in the fabrication of double oxide thin films such as BaTiO3, SrTiO3, their solid solutions, and LiNiO2 at fixed temperatures between 60 and 150 °C by hydrothermal–electrochemical methods.Nucleation and growth of the films takes place at the interface between the alkaline (earth) component in the solutions and anodically treated metal plates.The structures and properties of these films are described. A novel method to fabricate BaTiO3/SrTiO3 layered thin films by changing the solution compositions in flow-cell equipment is presented. defined as ‘environmentally friendly processing using (aqueous) Introduction solutions’. In this paper a thermodynamic and environmental Modern society needs advanced ceramic materials with background to SSP is given.In addition, recent important sophisticated functions and high performances. A number of developments in the field of SSP, such as fabrication of double substances with particular compositions, crystal structures and oxide thin films including multilayered materials and specified properties have been investigated for this purpose.applications of a new solution flow system, are presented. [In our considerations ‘substance’ is characterized by (1) a particular chemical composition, (2) a particular physical state Thermodynamic and environmental features of soft, including structure, and (3) particular properties based upon its composition and structure. Since ‘material’ should be solution processing defined as a substance mostly in the solid state which is used Generally speaking, SSP gives similar results to any other or to be used for certain application(s), we must add the processing using fluids (such as vapor, gas, plasma) and/or following features to define ‘material’: (4) a particular beam/vacuum processing. However, the total energy consump- shape, (5) a certain size, (6) location, and (7) orientation.1] tion of all these processing routes should be the lowest in Unfortunately, it is diYcult to give desired shapes, forms and aqueous systems, because a much larger excess of energy is sizes to inorganic materials, owing to their high brittleness.necessary to create melts, vapor, gas or plasma than to form Organic materials such as polymers and plastics or metallic an aqueous solution at the same temperature.1 This idea can materials can be generally deformed when local stresses (above be easily demonstrated using the simple example of BaTiO3.their yield stresses) are applied to them, but inorganic mate- The energy diagram for the formation of BaTiO3 from various rials, particularly ceramics, tend to break due to brittle fracprecursors is shown in Fig. 1. The driving forces (DG) for ture.2–4 Therefore, ceramics are fabricated by rather special representative synthesis reactions of (1), (2), (3), (4), and (5) ‘ceramic processing’ which consist of two steps: (1) synthesis are 38, 727, 3685, 17, and -14 kcal mol-1, respectively at of powders, and (2) shape-forming by firing/sintering of the room temperature.8,9 powders or melting (in the case of glasses).5 Both steps usually require high temperatures and thus consume a lot of energy.BaO (crystal )+TiO2 (crystal )=BaTiO3 (crystal ) (1) Novel processing using a gaseous phase, like CVD, MOCVD, Ba (vapor)+Ti (vapor)+3/2O2 (gas)=BaTiO3 (crystal ) (2) etc., or vacuum systems such as sputtering, MBE etc., require even higher energy than standard high temperature pro- Ba2+ (gas)+Ti4+ (gas)+3O2- (gas)=BaTiO3 (crystal ) (3) cessing.6 All these techniques have resulted in environmental TiO2 (crystal )+Ba2+ (aq)+2OH- (aq)-H2O (aq)= problems because the consumed energies are emitted as exhaust BaTiO3 (crystal ) (4) gas(es) or exhaust heat (entropy) except for the part involved in the product.In particular, vacuum systems seem to be Ba2+ (aq)+Ti(OH)4 (aq)-H2O (aq)-2H+ (aq)= worse because they need continuous pumping to maintain BaTiO3 (crystal ) (5) vacuum and their exhaust gas(es) cannot be cycled due to their diluted large volumes. This means that any processing using gaseous ions requires a Fortunately, we can fabricate shaped, sized, reacted and/or huge activation energy of 727–3685 kcal mol-1 to make solid oriented ceramic materials, without firing/sintering or melting, BaTiO3 and this energy must be discarded into the environwithout vacuum and expensive equipment, in situ, by soft, ment, because the raw materials of Ba and Ti must be solid solution processing (denoted hereafter as SSP).1,7 SSP is oxide(s) or carbonate(s) ore.On the other hand, since the lattice energy of BaO and TiO2 is almost equal to the hydration (solvation) energy of Ba2+ and Ti4+ ions, solution processing †Basis of the presentation given at Materials Chemistry Discussion consumes very little energy (DG, driving force) if only the No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. synthesis activation energy (DG*) can be overcome. Fig. 1 J. Mater. Chem., 1999, 9, 77–82 77ceramic films by solution processing have been limited except for the ‘sol–gel’ or ‘pyrolysis’ methods,10,11 which are regarded as solid rather than solution methods because solvents must be driven out of the system before formation of the crystalline material. In spite of the environmental and energetic advantages of SSP, film formation of multicomponent inorganic materials is quite diYcult from solution systems.In contrast, film formation is rather easy from gas or plasma systems. The gaseous species are always in high energy states (see also Fig. 1), thus the DG and the DG* are suYciently provided for reaction to yield crystalline compounds with desired shape/size via several steps such as diVusion, adsorption, reaction, nucleation, and growth.6 Generally speaking, DG*3(DG)-n, where n=2 for the nucleation process.On the other hand, species in aqueous solutions are hydrated (or chelated by some complexing agents), thus they have only a small DG for the reaction, thus rather high DG* values are necessary for the reaction to occur by overcoming the hydration (chelation) energies of ions.Electro- or electroless-plating of metals is achieved by reducing of metal ion(s) electrochemically or chemically.12 However, in the case of ceramic materials, particularly multicomponent ones, special activation of all the components is needed, anions must be oxidized at the same time as reduction of cations. Since this simultaneous multiactivation is quite diYcult, ceramics could scarcely be fabricated directly from solutions previously.However, we have demonstrated that direct fabrication of ceramic films can be achieved using interfacial Fig. 1 Energy diagram for the formation of BaTiO3 using solid and/or reactions between reactive substrates of metal, alloy, or oxide gas precursors (g, s and c correspond to the gaseous state solution state and the crystalline state, respectively). Compiled from ref. 8 and 9. and species in aqueous solution activated thermally (hydrothermal ) electrochemically, or by a combination (hydrothermal –electrochemical ).7 Other activation methods like photo-, gives the fundamental thermodynamic background for the sono-, electro-, mechano-, complexo-, or even bio-activation environmental problems related to materials processing.are possible to accelerate the interfacial reactions.1,13 A temperature–pressure map for various kinds of materials processing is shown in Fig. 2. Solution processing is located in the p–T range characteristic of conditions of living on earth. All other processing routes are associated with increasing Applications of soft, solution processing for thin film temperature and/or increasing (or decreasing) pressure, fabrication: hydrothermal–electrochemical technique therefore they are environmentally stressed.From the environ- Preparation techniques of thin films can be divided into two mental point of view, solution processing should be the most categories: (1) dry methods, such as vacuum deposition, environmentally friendly.Nevertheless, previous studies on sputtering, ion plating, CVD etc.,6 and (2) wet methods such as dip, spray coating of organic/inorganic precursors, sol–gel,10,11 electrodeposition,14 etc. All these methods usually require high temperatures (above 500 °C) to crystallize asdeposited amorphous films or to decompose the precursor to yield an appropriate compound. Such heat treatments often result in cracking and/or peeling of the deposited layers, formation of undesired compositions by evaporation, dissociation, and/or reaction of the film with the substrate, moreover they are energy consuming. In particular, it is diYcult to form dense films by organometallic and metallorganic processing, because decomposition of the precursors may result in porous ceramic layers.Therefore, low temperature, in situ fabrication of crystalline, thin films is essential to improve their quality, lower costs and make the whole process environmentally friendly. This was accomplished, by introducing the hydrothermal and/or electrochemical methods of film preparation.15 These techniques allow various double oxide films to be prepared on such substrates as metal, alloy or oxide.7 The substrate works as an electrode which reacts chemically and/or electrochemically (by passing electrical current) with the cation components of the surrounding solution.This processing route may involve the following reactions:16 (1) hydrothermal reaction between anodically oxidized film and solution components; (2) electrochemical reaction with solution component(s); (3) electrolytic oxidation of dissolved species and hydrothermal reaction; (4) hydrothermal reaction between anodically dissolved species Fig. 2 Schematic p–T diagram for preparative techniques for advanced ceramic materials. and solution components; (5) electro-discharge reaction 78 J.Mater. Chem., 1999, 9, 77–82between anodic oxidized film with solution component(s); (6) electro-deposition of charged oxide particles.Advantages of the hydrothermal/electrochemical techniques are: (1) one step (direct) formation of shaped/sized/deposited/ oriented ceramics; (2) low fabrication temperatures (minimized consumption of energy); (3) any shape, any size; (4) possibility of using a closed flow system, allowing easy charging, separation, cycling and recycling; (5) relatively high growth rate of the films (up to several mm per hour); (6) versatility.In addition, SSP gives products with much higher homogeneity than solid state processing and with higher density than gas or vacuum processing (faster growth rate). There is a wide variety of combinations of solvent/solute systems. Liquids may be beneficial for charging, transportation, mixing and/or separation of products.They are completely adaptable for cycling/recycling of material processing in closed systems. Moreover, liquids give the possibility for acceleration of diVusion, adsorption, reaction rate and crystallization Fig. 3 Cyclic voltammograms of the LiNiO2 film prepared at 125 °C (nucleation and growth), especially under hydrothermal with a current density of 1.0 mA cm-2 taken in 0.1 M LiClO4 propylene conditions.carbonate (potential is referred to Li/Li+ electrode) with a scan rate For these reasons, various thin films have been successfully of 5.0 mV s-1. prepared by the hydrothermal and/or electrochemical method. Examples include polycrystalline PZT,17,18 ZnO,19 TiO2,20 recycled by filtering it and replenishing the LiOH, which is ferrites,21–23 PbS,24 CdSe,24 BaTiO3,15,25–31 SrTiO3,32–36 environmentally and economically beneficial.BaxSr1-xTiO3,37 CaTiO3,38 BaFeO4 (BaFeO3-x),16 LiNbO3,16 The electrochemical–hydrothermal treatment of polished PbTiO3,39 ABO4 oxides (A=Ca, Ba, Sr; B=W, Mo)40,41 and nickel plates led to the formation of a visually detectable complete series of their solid solutions.42 The microstructures cobalt-blue film.The structural purity and chemical composiand chemical compositions of these materials can be controlled tion of the films were confirmed by X-ray diVraction and Xin a wide range at temperatures in most cases not exceeding ray photoelectron spectroscopy, respectively.54–56 According 200 °C, sometimes even at room temperature.40–42 The fabri- to X-ray diVraction, scanning electron microscopy (SEM) cation of epitaxial thin films,26,43 monomolecular layers,24 and cross-section views, and atomic force microscopy (AFM) even superlattices44,45 have also been reported.surface images, the films show good crystallinity despite the In the following two sections, new developments in the field low fabrication temperature without any post-synthesis of SSP are described. In both sections the experimental pro- annealing.54–56 Further evidence for the formation of LiNiO2 cedure and advantages of the applied techniques over pre- films was obtained from their cyclic voltammograms (Fig. 3) viously used ones for the particular material are also briefly since the peak potentials characterizing oxidation and mentioned.reduction are similar to those for a LiNiO2 powder electrode.55 Films prepared between 125 and 175 °C can also be reversibly oxidized and reduced in lithium perchlorate (LiClO4) propylene carbonate solutions for >5 electrochemical cycles, and Hydrothermal–electrochemical synthesis of LiNiO2 therefore are possible cathodes for lithium rechargeable films microbatteries.All-solid-state lithium rocking chair secondary microbatteries The films prepared under diVerent conditions show diVerent have been considered as the most suitable power source for electrochemical activity, surface morphology, and film thickmicroelectronics. 46–51 However, the synthetic approach in pre- ness.55,56 The apparent eVect of the fabrication temperature vious works required highly sophisticated multistep procedures on the surface morphology, for instance, is demonstrated by such as chemical vapor deposition, sol–gel methods, and comparing the AFM surface images of the nickel substrate sputtering.46–53 Such fabrication routes require high energy and the films prepared at diVerent temperatures (Fig. 4). and materials consumption, expensive precursors, and While purely hydrothermal treatment of nickel substrates complicated instrumentation.leads to the formation of Ni(OH)2 films, by use of the In the course of trying to develop an alternative, economic, electrochemical–hydrothermal approach under supplementary and low-temperature process that can produce the desired galvanostatic charge with the same hydrothermal conditions, LiNiO2 films can only eVectively be prepared in a single lithiated cathode films on chips, we have succeeded in the synthetic step from nickel substrates.54–56 In addition, the electrochemical preparation of active LiNiO2 films by electroelectrochemical oxidation of nickel plates in alkaline solution chemical–hydrothermal treatment on nickel plates.54–56 This at room temperature leads to the formation of NiOOH films.fabrication can be interpreted in terms of the electrochemical Therefore, the electrochemical anodic process is necessary to oxidation of nickel metal electrodes in a concentrated LiOH obtain trivalent nickel and the hydrothermal process is solution at fixed temperatures between 60 and 200 °C. The necessary for the cationic exchange reaction between Li+ electrochemical process, i.e.electrochemical oxidation of nickel and H+. plates, was realized galvanostatically with a fixed current density between 0.1 and 10 mA cm-2 in a three-electrode arrangement using nickel working and auxiliary electrodes and Multilayered thin films in the BaTiO3–SrTiO3 a temperature controlled Ag/AgCl reference electrode.57 The system: integration issues of soft, solution processing reaction time depends on the applied current density and fabrication temperature.Although detailed experimental It has been shown in the previous sections that a variety of procedures have been described previously,54–56 note that the technologically important thin films can be prepared by hydroprepared LiNiO2 films were not subjected to any post-synthesis thermal and/or electrochemical techniques.It seems therefore heat treatment. The used LiOH solution containing dissolved that we should start to think seriously about how to integrate the hydrothermal/electrochemical methods with functional nickel species, nickelite ions HNiO2- in practice, can be J. Mater. Chem., 1999, 9, 77–82 79Fig. 5 General schematic diagram of the flow cell for hydrothermal– electrochemical synthesis.disadvantages such as high vacuum and/or temperatures in the range of 500–700 °C. All the experiments were accomplished in our flow cell for hydrothermal–electrochemical synthesis. A general schematic diagram of the equipment is shown in Fig. 5. Single-phase, single layer and multilayered thin films in the BaTiO3–SrTiO3 system have been fabricated using titanium substrate and Ba and Sr acetates.A detailed description of the fabrication procedure has been published elsewhere.67–69 Results of syntheses carried out in open and closed flow were almost the same. Single phase BaTiO3, SrTiO3 or Fig. 4 Atomic force microscope (AFM) images of (a) the Ni substrate BaxSr1-xTiO3 have been prepared at 120–200 °C, when the polished with 1.0 mm diamond paste, and (b)–(f ) the LiNiO2 films current density did not exceed 1 mA cm-2.EVects of flow rate prepared at diVerent temperatures with a current density of on grain size are shown in Fig. 6. The grain size increases 1.0 mA cm-2: (b) 100 °C, (c) 125 °C, (d) 150 °C, (e) 175 °C and almost linearly with increasing flow rate, reaching a plateau (f ) 200 °C.Image size is 7.5 mm×7.5 mm in each case. at approximately 20 cm3 min-1. EVects of the flow rate in our experiments are in fact similar to the eVects of stirring applied in growth of crystals from solutions or melts. Usually, they lead to an increase in the probability of spontaneous device technology which has been dominated by physical vapor nucleation, reduction of supersaturation inhomogeneities, and deposition6 and chemical vapor deposition.6 In addition to increase of the growth rate.70 The growth rate of crystals PVD and CVD, some processing routes aimed at fabrication increases with increasing solution flow rate until a limiting of integrated circuits use chemical solution processing (etching, rate is reached when the growth rate becomes controlled by sol–gel ).58 In some cases, the (electro)chemical routes are considered as alternatives for the presently used vapor techniques (e.g.copper deposition).59 From this point of view, fabrication of ceramic thin films in the solution flow below 200 °C (this is the maximum temperature for wide industrial applications of the hydrothermal technique) is very important and may find multiple applications.The best suited equipment for this purpose seems to be a solution flow system for hydrothermal–electrochemical synthesis. Such equipment is similar to flow cells used in geological studies60 or in materials engineering (ferrite plating), 21,22 and allows fabrication of multilayered thin films by simply changing the flowing solution and/or adjusting the processing conditions.Moreover the flow can be closed, enabling easy recycling of the solutions. We have selected the BaTiO3–SrTiO3 system to demonstrate the fabrication of multilayered ceramic films under solution flow. Recently, attention has focused on multilayered films in Fig. 6 Average grain size of the BaTiO3, SrTiO3, and BaxSr1-xTiO3 this system,61–66 as they may find applications as DRAM single-phase thin films prepared in the flow cell by the hydrothermal– capacitors with high relative permittivity and reduced leakage electrochemical method at 150 °C (1 h, 1mAcm-2) as a function of current,62–64 tunable multilayer capacitors, waveguide phase the solution flow rate.Error bars denote standard deviations for shifters, filters, etc.66 The previous preparation routes to the average values.Solution flow rates were in the range of BaTiO3–SrTiO3 multilayered structures, such as tape-casting,66 1–50 cm3 min-1 which corresponds to the Reynolds numbers of about 46–2300, within a region of laminar flow.67 dip-coating,65 and magnetron sputtering62–64 had several 80 J. Mater. Chem., 1999, 9, 77–82environmentally friendly.Several advanced ceramic materials have been successfully fabricated using this technique. Well crystallized, electrochemically active LiNiO2 films for lithium rechargeable microbatteries were fabricated in a single step by the electrochemical–hydrothermal treatment of nickel plates in a concentrated LiOH solution at fixed temperatures between 60 and 200 °C without any post-synthesis annealing.The film properties show the obtained LiNiO2 films to be as suitable cathode films. Layered thin films in the BaTiO3–SrTiO3 system have been prepared in a solution flow system by the hydrothermal– electrochemical method in a single step by simply changing the flowing solution which is its main advantage over closed autoclaves. Changing the flow rate allows additional control of film morphology by enhancing the growth rate.This processing route may serve as an inexpensive and environmentally friendly way to fabricate functionally graded thin films and seems to be an important step in the application of the hydrothermal/electrochemical techniques to fabrication of integrated electronic devices. SSP allows fabrication in aqueous solution of shaped/sized/ oriented ceramics in one step, without firing/sintering or melting.This technique requires simple equipment and low temperatures. The morphology and chemical composition of Fig. 7 Atomic force microscopy images of the SrTiO3/BaTiO3/Ti multilayer (a) after step 1, i.e. the BaTiO3 surface, (b) after step 2, the resulting ceramics can be easily controlled by adjusting the i.e.the SrTiO3 surface; and the BaTiO3/SrTiO3/Ti multilayer (c) after processing conditions. SSP provides an inexpensive and step 1, i.e. the SrTiO3 surface, (d) after step 2, i.e. the BaTiO3 surface. environmentally friendly route to advanced ceramic materials, Image size is 1.0×1.0 mm in each case.67 and therefore is one of the most promising technologies for the 21st century.the interfacial kinetic processes.70 Our results suggest that the This research was supported by the ‘Research for the Future’ microstructure of the films at various flow rates is controlled Program No. 96R06901 of the Japanese Society for the by the growth rate, which initially increases, and then remains Promotion of Science (JSPS). The authors are greatly indebted unchanged for flow rates >20 cm3 min-1.to Prof. M. Kakihana, Prof. M. Yashima, Prof. M. Abe (all SrTiO3/BaTiO3/Ti and BaTiO3/SrTiO3/Ti layered thin films Tokyo Institute of Technology, Japan) for stimulating dis- have been fabricated at 150 °C (step 1: deposition of the first cussions and to Dr. P. Krtil, Mr. B. Sakurai, Mr. T.Watanabe, layer+step 2: deposition of the second layer), by changing the Miss.S-W. Song, Mr. S. Tsurimoto, and Mr. N. Kumagai (all flowing solution. XRD patterns of the double layers showed Tokyo Institute of Technology, Japan) for experimental distinct peaks derived from single-phase SrTiO3 and BaTiO3 assistance. layers.67 The presence of BaxSr1-xTiO3 solid solutions has not been detected by XRD, but we cannot preclude dissolution of small quantities of Ba in the SrTiO3 layers or vice versa.References Results of AFM measurements are shown in Fig. 7. They clearly show diVerent microstructures of the surface of the 1 M. Yoshimura, J. Mater. Res., 1998, 13, 796. single layers and the double layers. These data demonstrate 2 D. W. van Krevelen, Properties of Polymers, Elsevier Science Publishers, Amsterdam, 1990. that initially formed BaTiO3 and SrTiO3 layers have been 3 R.A. Higgins, Engineering Metalurgy, Edward Arnold, London, covered by layers of SrTiO3 and BaTiO3, respectively. XPS 1993. data confirmed presence of the Sr- or Ba-rich layer on the 4 W. D. Kingery, H. K. Bowen and D. R. Uhlmann, Introduction to surface and the Ba- or Sr-rich layer, respectively, below.67,71 Ceramics, John Wiley & Sons, New York, 1976.The formation mechanism of the multilayered thin films in 5 R. A. Ring, Fundamentals of Ceramic Powder Processing and the BaTiO3–SrTiO3 system by the hydrothermal–electrochem- Synthesis, Academic Press, San Diego, 1996. 6 M. Ohring, The Materials Science of Thin Films, Academic Press, ical technique will be discussed in detail elsewhere.72 New York, 1992. The processing described may be applied generally for 7 M.Yoshimura and W. Suchanek, Solid State Ionics, 1997, 98, 197. various single and/or multilayered thin films. The possibility 8 JANAF Thermochemical Tables, National Bureau of Standards, of solution recycling during hydrothermal–electrochemical Ametican Chemical Society, American Institute of Physics, synthesis of the thin films in the BaTiO3–SrTiO3 system is 1968–1982.important because it minimizes the environmental impact of 9 M.M. Lencka and R. E. Riman, Chem. Mater., 1993, 5, 61. 10 Sol–Gel Technology for Thin Films, Fibers, Preforms, Electronics, the process.1 The technique using the solution flow under and Specialty Shapes, ed. L. Klein, Noyes Publ., Mill Road, Park hydrothermal–electrochemical conditions is an important step Ridge, New Jersey, 1988.in integration of the solution processing with functional device 11 Sol–Gel Optics. Processing and Applications, ed. L. Klein, Kluwer technology and may find applications as an inexpensive and Academic Publishers, Boston/Dordrecht/London, 1994. environmentally friendly way to fabricate various single and/or 12 Modern Electroplating, ed.F. A. Lowenheim, John Wiley & Sons, multilayered thin films and functionally graded materials. Inc., New York, 1963. 13 R. Roy, J. Solid State Chem., 1994, 111, 11. 14 C. D. Lokhande and S. H. Pawar, Phys. Status Solidi A, 1989, Summary 111, 17. 15 M. Yoshimura, S.-E. Yoo, M. Hayashi and N. Ishizawa, Jpn. A thermodynamic concept for energetic and environmental J.Appl. Phys., 1989, 28, L2007. problems related to materials synthesis has been proposed. It 16 M. Yoshimura, Mater. Res. Soc. Symp. Proc., 1992, 271, 457. concludes that soft, solution processing, i.e. solution processing 17 Y. Ohba, M. Miyauchi, E. Sakai and M. Daimon, Jpn. J. Appl. Phys., 1995, 34, 5216. using (aqueous) solutions, should be the most economical and J.Mater. Chem., 1999, 9, 77–82 8118 A. T. Chien, J. S. Speck and F. F. Lange, J. Mater. Res., 1997, 47 P. Birke, W. F. Chu and W. Weppner, Solid State Ionics, 1997, 93, 1. 12, 1176. 48 K. A. Striebel, C. Z. Deng, S. J. Wen and E. J. Cairns, 19 Q. Chen, Y. Qian, Z. Chen, G. Zhou and Y. Zhang, Mater. Lett., J. Electrochem. Soc., 1996, 143, 1821. 1995, 22, 93. 49 S.J. Lee, J. K. Lee, D. W. Kim, H. K. Baik and S. M. Lee, 20 Q. W. Chen, Y. T. Qian, Z. Y. Chen, Y. B. Jia, G. E. Zhou, J. Electrochem. Soc., 1996, 143, L268. X. G. Li and Y. H. Zhang, Phys. Status Solidi A, 1996, 156, 381. 50 K. H. Hwang, S. H. Lee and S. K. Joo, J. Electrochem. Soc., 1994, 21 M. Abe, Y. Tamaura, Y. Goto, N. Kitamura and M. Gomi, 141, 3296. J. Appl. Phys., 1987, 61, 3211. 51 F. K. Shokoohi, J. M. Tarascon and B. J. Wilkens, Appl. Phys. 22 M. Abe, T. Itoh and Y. Tamaura, Mater. Res. Soc. Symp. Proc., Lett., 1991, 59, 1260. 1991, 232, 107. 52 M. Antaya, K. Cearns, J. S. Preston, J. N. Reimers and 23 J. S. Lee, T. Itoh and M. Abe, J. Korean Phys. Soc., 1995, 28, 375. J. R. Dahn, J. Appl. Phys., 1994, 76, 2799. 24 J. H. Fendler and F. C. Meldrum, Adv.Mater., 1995, 7, 607. 53 M. Antaya, J. R. Dahn, J. S. Preston, E. Rossen and 25 M. Yoshimura, S.-E. Yoo, M. Hayashi and N. Ishizawa, Ceram. J. N. Reimers, J. Electrochem. Soc., 1993, 140, 575. Trans., 1990, 15, 427. 54 M. Yoshimura, K. S. Han and S. Tsurimoto, Solid State Ionics, 26 F. F. Lange, Science, 1996, 273, 903. 1998, 106, 39. 27 E. B. Slamovich and I. A. Aksay, J. Am. Ceram. Soc., 1996, 55 K.S. Han, P. Krtil and M. Yoshimura, J. Mater. Chem., 1998, 79, 239. 8, 2043. 28 M. E. Pilleux and V. M. Fuenzalida, J. Appl. Phys., 1993, 74, 4664. 56 K. S. Han, S. W. Song and M. Yoshimura, Chem. Mater., 1998, 29 T. Vargas, H. Diaz, C. I. Silva and V. M. Fuenzalida, J. Am. 10, 2183. Ceram. Soc., 1997, 80, 213. 57 D. D. Macdonald and D. Owen, J. Electrochem. Soc., 1973, 120, 30 R.R. Bacsa, G. Rutsch and J. P. Dougherty, J. Mater. Res., 1996, 317. 11, 194. 58 H. L. Tuller, Ceram. Trans., 1996, 68, 97. 31 P. Bendale, S. Venigalla, J. R. Ambrose, E. D. Verink Jr. and 59 R. J. Gutmann, T. P. Chow, S. Lakshminarayanan, D. T. Price, J. H. Adair, J. Am. Ceram. Soc., 1993, 76, 2619. J. M. Steigerwald, L. You and S. P. Murarka, Thin Solid Films, 32 S.-E. Yoo, M. Hayashi, N. Ishizawa and M. Yoshimura, J. Am. 1995, 270, 472. Ceram. Soc., 1990, 73, 2561. 60 J. M. Porter, D. C. Pohl and J. D. Rimstid, in Hydrothermal 33 K. Kajiyoshi, Y. Hamaji, K. Tomono, T. Kasanami and Experimental Techniques, ed. G. C. Ulmer and H. L. Barnes, John M. Yoshimura, J. Am. Ceram. Soc., 1996, 79, 613. Wiley & Sons, New York, 1987, p. 240. 34 K. Kajiyoshi, K. Tomono, Y. Hamaji, T. Kasanami and 61 M. Suzuki, J. Ceram. Soc. Jpn., 1995, 103, 1099 (in Japanese). M. Yoshimura, J. Am. Ceram. Soc., 1995, 78, 1521. 62 M.-H. Song, Y.-H. Lee, T.-S. Hahn and M.-H. Oh, J. Appl. Phys., 35 K. Kajiyoshi and M. Yoshimura, Eur. J. Solid State Inorg. Chem., 1996, 79, 3744. 1996, 33, 623. 63 Q. X. Jia, L. H. Chang and W. A. Anderson, Thin Solid Films, 36 M. E. Pilleux, C. R. Grahmann and V. M. Fuenzalida, J. Am. 1995, 259, 264. Ceram. Soc., 1994, 77, 1601. 64 Z. Q. Shi, Q. X. Jia and W. A. Anderson, J. Electron. Mater., 37 K. Kajiyoshi, M. Yoshimura, Y. Hamaji, K. Tomono and 1991, 20, 939. T. Kasanami, J. Mater. Res., 1996, 11, 169. 65 T. Hayashi and T. Tanaka, Jpn. J. Appl. Phys., 1995, 34, 5100. 38 M. Yoshimura, W. Urushihara, M. Yashima and M. Kakihana, 66 V. K. Varadan, V. V. Varadan, F. Selmi, Z. Ounaies and Intermetallics, 1995, 3, 125. K. A. Jose, Proc. SPIE—Int. Soc. Opt. Eng., 1994, 2189, 433. 39 W.-S. Cho and M. Yoshimura, J. Mater. Res., 1997, 12, 833. 67 M. Yoshimura, W. Suchanek, T. Watanabe, B. Sakurai and 40 W. S. Cho and M. Yoshimura, Jpn. J. Appl. Phys., 1997, 36, 1216. M. Abe, J. Mater. Res., 1998, 13, 875. 41 W. S. Cho, M. Yashima, M. Kakihana, A. Kudo, T. Sakata and 68 W. Suchanek, T.Watanabe and M. Yoshimura, Solid State Ionics, M. Yoshimura, J. Am. Ceram. Soc., 1997, 80, 765. 1998, 109, 65. 42 W. S. Cho, M. Yashima, M. Kakihana, A. Kudo, T. Sakata and 69 W. Suchanek and M. Yoshimura, J. Am. Ceram. Soc., 1998, in M. Yoshimura, Appl. Phys. Lett., 1996, 68, 137. press. 70 D. Elwell and H. J. Scheel, Crystal Growth from High-Temperature 43 A. T. Chien, L. Zhao, M. Colic. J. S. Speck and F. F. Lange, Solutions, Academic Press, New York, 1975. J. Mater. Res., 1998, 13, 649. 71 W. Suchanek, T. Watanabe, B. Sakurai and M. Yoshimura, 44 J. A. Switzer, C. J. Hung, B. E. Breyfogle, M. G. Shumsky, Mater. Res. Soc. Symp. Proc., 1998, 517) High-Density Magnetic R. Vanleeuwen and T. D. Golden Science, 1994, 264, 1573. Recording and Integrated Magneto-Optics: Materials and 45 J. A. Switzer, C. J. Hung, L. Y. Huang, F. C. Miller, Y. Zhou, Devices), in press. E. R. Raub, M. G. Shumsky and E. W. Bohannan, J. Mater. Res., 72 W. Suchanek and M. Yoshimura, unpublished data. 1998, 13, 909. 46 F. K. Shokoohi, J. M. Tarascon, B. J. Wilkens, D. Guyomard and C. C. Chang, J. Electrochem. Soc., 1992, 139, 1845. Paper 8/04647F 82 J. Mater. Chem., 1999, 9, 77–82

ISSN:0959-9428    

DOI:10.1039/a804647f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Hydrothermal synthesis and characterisation of BaTiO3 fine powders: precursors, polymorphism and properties† Iain J. Clark,a Tomanari Takeuchi,b Noikazu Ohtoric and Derek C. Sinclaira aChemistry Department, University of Aberdeen, Meston Walk, Aberdeen, UK AB24 3UE bOsaka National Research Institute, 1–8-31, Midorigaoka, Ikeda, Osaka, Japan cGraduate School of Science and Technology, Niigatu University, Ikarashi 2-no cho, Niigata, Japan Received 19th June 1998, Accepted 22nd July 1998 The influence of two Ti-precursors, TiO2 (anatase) and H2TiO3 (b-titanic acid), on the purity and particle size of BaTiO3 powders prepared via hydrothermal synthesis is discussed.Amorphous H2TiO3 was found to be an excellent Ti-precursor material and oVers several advantages over crystalline anatase.Phase pure powders which have small particle sizes, ca. 40–80 nm and narrow particle size distributions can be prepared at 180 °C after 24 h using H2TiO3 as a precursor material. Although the initial reaction is very fast, ca. 90% yield after 8–10 h, extended reaction periods at 180 °C are required in order to drive the reaction to completion.Lowering the reaction temperature from 180 to 85 °C does produce powders with even smaller particle sizes, however, very long reaction periods are required, e.g.>72 h, to ensure complete reaction. Raman spectra of as-prepared and heat treated (1000 °C) powders with average particle sizes as small as ca. 20–40 nm indicate asymmetry within the TiO6 octahedra of the BaTiO3 lattice.These results contradict the widely cited ‘critical’ particle size theory for the stabilisation of the cubic polymorph, at least for particle sizes greater than ca. 20–40 nm. As-prepared powders contain many defects, primarily in the form of lattice OH- ions. Preliminary ac impedance spectroscopy data on samples heat treated to remove lattice hydoxyl ions demonstrate these materials to be modest proton conductors.size eVect but there is still no clear consensus as to its origin.‡ Introduction Two commonly cited models involve the generation of higher- BaTiO3 is one of the most widely used and studied ferroelectric than-usual stress in fine grained ceramics5,6 whereas a third materials in the electro-ceramics industry. It has various suggests a core-shell structure,7,8 whereby grains below a polymorphs, all of which are based on the perovskite struc- ‘critical’ size (<0.2 mm) have cubic symmetry, intermediate ture;1 however, the two most studied are the tetragonal, t, and grain sizes (ca. 1 mm) have a ferroelectric-tetragonal core with cubic, c, polymorphs.t-BaTiO3 forms between ca. 0 and 130 °C a cubic surface layer and large grains (5 mm) are essentially whereas c-BaTiO3 is stable above 130 °C.In the t-polymorph, tetragonal. It should be noted that there is not an accepted titanium ions are displaced from the centrosymmetric position ‘critical’ particle or grain size below which the cubic polymorph within TiO6 octahedra and give rise to spontaneous polaris- is stabilised, on the contrary, a wide range of values, ca. ation.Due to this asymmetry within the crystal structure, t- 25–190 nm have been reported.9–16 In addition, a variety of BaTiO3-based materials have high permittivities and are widely reasons have been proposed for the room-temperature stabilisemployed as dielectrics in ceramic capacitors.2 The transition ation of the cubic polymorph, including the presence of lattice from the polar (ferroelectric) t- to non-polar (paraelectric) c- ‘defects’ such as hydroxyl ions17 (associated with powders polymorph normally occurs at ca. 130 °C, the Curie tempera- formed via wet chemical methods), small deviations in Ba/Ti ture, Tc. Above Tc, the Ti ions occupy, on average, the stoichiometry7 and excess surface energy associated with centrosymmetric position within TiO6 octahedra and there is ultrafine particles.18,19 Clearly, much remains to be done in no net polarisation within the solid.order to establish the factors which influence and control the A typical room temperature permittivity value, e25, for crystal symmetry and electrical properties of sub-micron ceramic t-BaTiO3 with an average grain size of ca. 10 mm is BaTiO3 powders and ceramics. 1000–2000; however, this rises sharply to a maximum value, Despite these fundamental problems, BaTiO3 will continue emax, of ca. 10 000 at Tc. In such ‘large’ grained ceramics, a to be used in the manufacture of thermistors,20 multilayer variety of ferroelectric domain structures form to relieve capacitors,21 electro-optic devices22 and DRAM (dynamic internal stress associated with the cubic to tetragonal phase random access memories)23 into the next century.Improved transformation. Although a detailed discussion of domain performance and miniaturisation of BaTiO3-based devices, structures in BaTiO3 powders and ceramics is outwith the either in the form of thin ceramic layers, ca. 20mm, for scope of this paper, it is important to note that ferroelectric multilayer capacitors or as thin films for integrated electronic domain structures and dipole–dipole interactions control the circuits remains a priority.High permittivities and miniaturispermittivity characteristics of BaTiO3 and that these depend ation can be achieved by controlling the ceramic microstructure on particle or grain size. For example, Arlt et al.3 have shown which, in turn, depends on the homogeneity, composition, that e25 for BaTiO3 ceramics can be optimised to a value surface area and particle size of the starting BaTiO3 powder.approaching 5000 by controlling the grain size to be ca. In addition, it is well known that BaTiO3 undergoes exagger- 0.7–1.0 mm. ated grain growth during sintering at elevated temperatures Several models have been proposed to explain this grain and long sintering periods.This can produce ceramics †Basis of the presentation given at Materials Chemistry Discussion ‡During the preparation of this manuscript, Frey et al.4 published a paper which appears to have clarified many of the problems associated No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. with the origin of the so-called grain size eVect in BaTiO3.J. Mater. Chem., 1999, 9, 83–91 83containing very large grain sizes, e.g. >50 mm or duplex and the surrounding BaTiO3. For this model, complete reaction may be diYcult to achieve due to increasing encapsulation microstructures of low density consisting of both large (>10 mm) and small (<1 mm) grains.For these reasons, many of TiO2 via product formation. Homogeneous and heterogeneous dissolution–precipitation synthetic methods and sintering profiles24–28 have been investigated in an attempt to produce sub-micron, deagglomerated models have also been proposed. In the homogeneous model, a low concentration of TiO2 dissolves in the form of soluble fine powders of BaTiO3 that can be sintered into dense, finegrained ceramics.hydroxytitanium complexes which then react with Ba2+ ions in solution to precipitate BaTiO3. In the heterogeneous model, Traditionally, BaTiO3 has been produced via the mixed oxide route. This method involves repeated calcination and BaTiO3 nuclei form on the surfaces of the dissolving TiO2 particles and, as for the in situ transformation mechanism, the regrinding of BaCO3 and TiO2 powders above 1100 °C.The reaction mechanism in air has been proposed to take place in reaction becomes limited by the increasing isolation of the reactants via BaTiO3 product layers on the surface of TiO2 at least three stages29–32 and relies on the diVusion of Ba2+ ions into TiO2. Firstly, BaCO3 reacts with the outer surface particles.Eckert et al. reviewed these models41 in more detail and regions of TiO2 to form a surface layer of BaTiO3 on individual TiO2 grains. Further diVusion of Ba2+ ions into TiO2 necessi- undertook a kinetic study of hydrothermal reactions involving barium hydroxide octahydrate and anatase precursors. Two tates the formation of Ba2TiO4 between unreacted BaCO3 and the previously formed BaTiO3.After prolonged sintering reaction regimes were clearly identified; during the early stages the reaction is controlled by a dissolution–precipitation pro- periods, the intermediate Ba-rich phase Ba2TiO4 reacts with the remaining TiO2 in the core-regions of the TiO2 grains to cess, whereas, for the second regime at longer reaction times the results were inconclusive.The authors suggest two plausible form BaTiO3. Although the nominal starting Ba/Ti ratio is 151, long reaction times are required to form homogeneous but diVerent explanations for their kinetic data, one model is based on heterogeneous dissolution–precipitation followed by powders free from secondary phases. For such a complex and slow reaction mechanism, it is desirable to have high surface in situ transformation, whereas, the second model suggests that dissolution–precipitation is the controlling mechanism, area TiO2 powders of fine particle size and narrow size distribution in order to control the morphology and grain size with nucleation and growth of BaTiO3 controlling the first regime and the dissolution rate of TiO2 controlling the of the resultant BaTiO3 powder.Nevertheless, the mixed oxide route tends to produce coarse, agglomerated powders that second regime. In the present paper we discuss some of our recent work on require high sintering temperatures to form dense ceramics. In order to produce powders that are suitable for sintering the hydrothermal synthesis of BaTiO3 powders and highlight some advantages of using amorphous H2TiO3 as opposed to into micron-grain-sized ceramics for MLC applications there has been much interest in developing wet and novel chemical crystalline anatase as a precursor material.In particular, H2TiO3 promotes faster and more-complete reactions and also routes. As these reactions take place in the liquid, as opposed to the solid state, more intimate mixing of the cations is produces powders with much smaller particle sizes.We discuss the influence of temperature on the rate of reaction and on achieved and consequently, much shorter diVusion pathways are created. Deagglomerated, sub-micron powders of BaTiO3 the particle size, distribution and water content of the BaTiO3 powders. The importance of using Raman spectroscopy to can then be obtained at low temperatures and short reaction periods, for example via hydrothermal synthesis at 240 °C detect the presence of trace amounts of unreacted Ti-containing precursors and to yield information on the polymorphism of for 24 h.33 Other common methods include, oxalate (Clabauch),34 citrate (Pechini),35 catecholate (Milne)36 and sub-micron powders is discussed.Finally, we present preliminary results on powders and ceramics heat treated above sol–gel.37 Although the reaction mechanisms and advantages and disadvantages of many of these routes have been dis- 800 °C that support the idea that hydrothermal BaTiO3 powders contain many defects which influence their physical cussed,38 for the present purposes it is worth discussing these issues for hydrothermal synthesis. properties.BaTiO3 powders have been prepared via hydrothermal processing since the 1940’s and this method is commonly employed Experimental to produce commercial powders. Although it is well established that BaTiO3 is the only thermodynamically stable binary Ba(OH)2·8H2O (98+%, Aldrich) and anatase (99.9%, Aldrich) or b-H2TiO3 (+99%,Mitsuwa Chemicals) were mixed compound produced under all conditions,39 i.e.pH, Ba/Ti ratio, reaction temperature, etc. products can contain unre- with a Ba5Ti ratio of 1.0551 in a 23 ml Teflon-lined pressure vessel (Model 4749, Parr Instruments) together with 10 ml of acted TiO2 and BaCO3. Normally, an excess of Ba2+ in the starting solution is employed (Ba/Ti=1.05–1.1051) in an de-ionised water. The vessel was sealed, shaken and placed in an oven at either 85, 120 or 180 °C for periods ranging from attempt to remove any unreacted TiO2 and therefore drive the reaction to completion.Any excess Ba, in the form of BaCO3 1 to 72 h. After cooling, the contents of the bomb were diluted in 30 ml of 0.1M formic acid in an attempt to dissolve any can then be removed via acid washing of the powders.Although this procedure is eVective in controlling the purity BaCO3 formed by the addition of excess Ba2+ to the starting solution. The mixture was vacuum filtered using a Buchner of the powders, relatively little is known about the influence of acid washing on powder stoichiometry as Ba2+ may be funnel, the residue thoroughly washed with distilled water and dried overnight in air at 120 °C.The percentage yield of leached from the surfaces of the particles. Although controlling the precise stoichiometry of hydrothermal BaTiO3 powders BaTiO3 was estimated from the weight of the dried powders. Phase purity and polymorphism of the BaTiO3 powders can be problematic, the low cost and easy handling of the reagents, and the fast reaction rate at low temperatures ensures were studied by X-ray diVraction (XRD) and Raman spectroscopy.XRD was carried out on a Stoe STADI P automated that deagglomerated powders consisting of small particles of narrow size distribution are readily obtained. powder diVractometer employing monochromatic Cu-Ka1 radiation and a linear position sensitive detector over the 2h Several mechanisms have been proposed for the hydrothermal synthesis of BaTiO3.Hertl40 has suggested an in situ range 20 to 80 °. Cubic and tetragonal polymorphs were distinguished by the splitting of the (200) peak at 2h#45 °. transformation whereby dissolved Ba2+ ions react with undissolved TiO2 to produce a continuous layer of BaTiO3 on the Unit cell parameters were obtained by refinement of XRD data. Raman spectroscopy was carried out on a Jobin Yvon surface of TiO2 particles. The reaction is then controlled either by diVusion of Ba2+ through the product BaTiO3 layer or T64000 using a 100 mW Ar laser with a wavelength of 514.5 nm.A 300 mm slit and integration time of 5 s were used further reaction between unreacted TiO2 within grain interiors 84 J. Mater. Chem., 1999, 9, 83–91giving a resolution of 8.4 cm-1.Spectra were measured over An important aspect of this work is to obtain homogeneous the range 58–1103 cm-1. Impurity phases such as BaCO3 and particles of controlled stoichiometry which can be sintered TiO2 (anatase) were detected from characteristic peaks at 1060 into dense ceramics of uniform microstructure. The Ba5Ti and 150 cm-1, respectively. The presence of a peak in the ratio of the particles is an important parameter but is not spectra of BaTiO3 at 305 cm-1 indicates asymmetry within readily obtained via XRD or Raman spectroscopy.We choose, TiO6 octahedra42 and was used to distinguish between tetra- therefore, to study the phase assemblage and Ba5Ti ratio of gonal and cubic polymorphs. the grains in sintered ceramics using EPMA and from this, Thermogravimetric analysis (TGA) was carried out using a deduce information about the Ba5Ti ratio of the powders.Stanton Redcroft TG-DTA simultaneous thermal analyser This is an indirect method for obtaining information on the (model STA 1000/1500) to determine the water content of as- stoichiometry of the powders, but since neither barium nor prepared powders.Samples were accurately weighed to five titanium are particularly volatile at the sintering temperature, decimal places in a platinum crucible and weight losses we perceive this method to be valid. recorded from 25–1000 °C against an alumina reference. All experiments were carried out in air at a heating rate of 10 °Cmin-1 with an amplifier setting of 50 mV. Results and discussion Powder morphology was determined via transmission electron microscopy (TEM) using a JEOL electron microscope The results are subdivided into three sections.The first section (model 2000EX TEMSCAN) operating at 200 kV. Particle discusses the influence of two Ti-precursors, TiO2 (anatase) size and distribution were obtained by measuring the cross and H2TiO3 (b-titanic acid), on the purity and particle size of diagonals of approximately 130 particles from scanned nega- the BaTiO3 powders.The second section assesses the role of tives of TEM images. Energy dispersive X-ray analysis (EDX) reaction temperature on the rate of reaction, purity, particle was used in an attempt to determine the Ba5Ti ratio of the size and water content of BaTiO3 powders prepared from BaTiO3 particles and also to detect the presence of any Ba(OH)2·8H2O and H2TiO3.The final section discusses preunreacted TiO2. liminary electrical property measurements on powders heat Electron probe micro-analysis (EPMA) was employed to treated above 800 °C. determine the homogeneity of sintered pellets fabricated from the various powders. Pellets were sintered at ca. 1350 °C and I Titanium precursors a Cameca SX51 electron microprobe employed using an accelerating voltage of 20 kV and a beam current of 40 nA. One of the major problems associated with the hydrothermal Ba-La and Ti-Ka lines were measured using benitoite, synthesis of BaTiO3 is the low solubility of TiO2 in the highly BaTiSi3O9, as a standard. Prior to analysis, small pieces of alkaline environments, ca.pH>12, required for synthesis. pellet were mounted in resin blocks, surface polished and Although it is well documented that diVerent Ti-precursors carbon coated. and reaction temperatures influence the rate of reaction there Electrical measurements were carried out using a Solartron have been few comprehensive studies on the influence of these frequency response analyser (model 1250 or 1260) combined parameters on the purity and particle size of hydrothermal with a Solartron dielectric interface (model 1296).A frequency powders. Kutty et al. suggested that amorphous Ti-precursors range of 10-3–106 Hz was employed with an applied voltage react fastest,17 then anatase with rutile giving the slowest of 100 mV. Cold pressed pellets of various powders were reaction.Various workers reported an increase in average sintered between 800 and 1350 °C and organopaste and gold particle size and a change (via XRD) from cubic to tetragonal foil electrodes applied to the surfaces of the pellets at 800 °C, symmetry with increasing reaction temperature.9–16We decided prior to mounting in a conductivity jig. Measurements were to reinvestigate some of these findings by comparing the recorded over the temperature range 25–500 °C.reaction rates, powder purity and particle size of hydrothermal Before discussing the results it is important to appreciate BaTiO3 powders prepared at various temperatures and reac- the need for a wide variety of techniques to determine the tion periods using two Ti-precursors: amorphous H2TiO3 phase purity and crystal structure of fine-grain BaTiO3 pow- (BET using N2, surface area 193 m2 g-1) and crystalline ders.XRD is an excellent technique for obtaining information anatase (BET, surface area 8.9 m2 g-1). on the average or long range crystal structure of a material; In general, for a set reaction temperature and period, the however, peak broadening is significant for diVraction from reaction rate was faster and the average BaTiO3 particle size sub-micron sized particles and XRD cannot reveal the very smaller for reactions using the amorphous H2TiO3.Raman subtle unit cell distortions which occur in BaTiO3 for microdospectra of powders prepared under the same conditions for mains of dimensions of ca. 100 A° . In addition, routine XRD the two precursors are shown in Fig. 1. Spectra of anatase and is relatively insensitive to small quantities of impurity phases, the tetragonal polymorph of BaTiO3 prepared via the tra- especially for secondary phases which are amorphous/poorlyditional mixed oxide route are included for comparison. The crystalline and/or have weak X-ray scattering power. In our Raman spectrum of H2TiO3 is not shown but is similar to experience, XRD is sensitive to small quantities of BaCO3 in that of anatase and is also dominated by a large peak at ca.hydrothermally processed BaTiO3 but is much less sensitive in 150 cm-1. There are two noticeable features in the Raman the detection of unreacted Ti-containing precursors, either spectra of the products which are not readily discernible from crystalline anatase or amorphous H2TiO3.the XRD patterns (not shown) and which provide important Uncertainties over crystal symmetry can largely be resolved information. using alternative techniques such as Raman spectroscopy to First, the powder products from the H2TiO3 reaction appear probe short range order or local symmetry. There are reliable to be phase pure BaTiO3, whereas unreacted TiO2, as shown Raman spectra in the literature42 for various BaTiO3 polyby the presence of the peak at ca. 150 cm-1, is clearly present morphs; however, it must be stated that the diVerences in in the products of the anatase reaction. Small amounts of spectra are rather subtle, especially for the orthorhombic, unreacted anatase or amorphous H2TiO3 are diYcult to detect tetragonal and cubic (at temperatures slightly greater than Tc) via routine XRD but are readily detected (by the peak at ca.polymorphs. In contrast to XRD, Raman spectroscopy is 150 cm-1) in the Raman spectra. This result demonstrates the sensitive to the presence of both BaCO3 and unreacted Tiimportance of using Raman spectroscopy to monitor the phase containing precursors.For these reasons we choose to present purity of BaTiO3 powders prepared via hydrothermal syn- Raman spectra, as opposed to XRD patterns to illustrate the thesis. We should stress that single phase powders can, in fact, phase purity and crystal symmetry of as-prepared BaTiO3 powders. be prepared from crystalline anatase at 180 °C, however the J. Mater. Chem., 1999, 9, 83–91 85Fig. 2 TEM micrographs of BaTiO3 powders prepared via hydrothermal synthesis at 180 °C for 24 h using anatase (a) and H2TiO3 (b), as Ti-precursors. Fig. 1 Raman spectra of BaTiO3 prepared via a mixed oxide route (top) and via hydrothermal synthesis at 180 °C for 24 h using H2TiO3 and anatase as Ti-precursors. The bottom spectrum is of crystalline II Reaction temperature anatase.Raman spectra of powder products from two reaction temperatures, 85 and 180 °C after various reaction periods are shown in Fig. 3(a) and (b), respectively. With increasing time, both times required are far in excess of those required using sets of spectra illustrate a decrease in peak intensity at amorphous H2TiO3, typically in excess of 72 h. 150 cm-1 associated with the titanium precursor as BaTiO3 is Second, the presence of a peak at 305 cm-1 in both spectra formed.In fact, the spectrum after reaction at 180 °C for 24 h indicates asymmetry within the TiO6 octahedra of BaTiO3 and shows no evidence of this peak, suggesting that the reaction demonstrates clearly, on a local scale, that as-prepared powders has gone to completion. A small peak at 150 cm-1 is still do not have cubic symmetry.In contrast, XRD patterns apparent in the corresponding spectrum for powders prepared showed appreciable peak broadening, presumably associated at 85 °C. A higher reaction temperature of ca. 180 °C, is clearly with the small particle size and poor crystallinity of the required to ensure a more complete reaction for short reaction powders, however, there was little evidence of peak splitting periods, e.g. 24 h. The high surface area (193 m2 g-1) and the and the XRD data ‘appeared’ consistent with that for the fact that hydroxylation of many Ti–O–Ti bridging bonds has cubic polymorph, as reported previously.9–16 The higher sensi- already been achieved in H2TiO3, ensure rapid formation of tivity of Raman spectroscopy to probe the local rather than BaTiO3, even at 85 °C.It should be noted that many spectra long range structure clearly demonstrates that sub-micron contained a small peak at 1060 cm-1 (not shown) associated powders prepared via hydrothermal processing are tetragonal with a small amount of BaCO3 which was not completely rather than cubic. removed by the mild acid wash. The particle size of powders prepared from H2TiO3 were, The weight of the dried powders was used to calculate the on average, 3–5 times smaller than those prepared from TiO2 percentage yield of BaTiO3 and is shown in Fig. 4. We stress and, to our knowledge, are amongst the smallest reported for that these values represent upper estimates of the yield, especi- BaTiO3 powders prepared hydrothermally. Representative ally for short reaction periods where powders contain unremicrographs of powders prepared at 180 °C for 24 h from acted precursor and are also heavily hydrated.Nevertheless, H2TiO3 and TiO2 are shown in Fig. 2 and demonstrate the these estimates show that reaction is rapid and yields in excess much smaller particle size, ca. 80 nm compared with ca. of 90% are obtained within 8 h of reaction.From the Raman 250 nm. The faster rate of reaction and smaller particle size spectra in Fig. 3, extended periods, e.g. 24 h at elevated of powders prepared from amorphous H2TiO3 demonstrate temperatures (180 °C) are required in order to drive the two clear advantages of using this precursor instead of crystal- reaction close to completion, even for a reactive precursor line anatase.The remaining sections of the paper concentrate such as H2TiO3. on powders and ceramics prepared using H2TiO3 as the Ti- As-prepared powders are hydrated and lose water both adsorbed and structural on heating to ca. 1000 °C. In general, precursor. 86 J. Mater. Chem., 1999, 9, 83–91Fig. 3 Raman spectra of powder products from two reaction temperatures, 85 (a) and 180 °C (b), after various reaction periods.dramatic decrease in reaction rate after ca. 8–10 h, as shown by the percentage yield of BaTiO3. Nucleation of BaTiO3 crystals presumably takes place via a structural rearrangement of the amorphous H2TiO3 associated with the incorporation of Ba2+ or via direct reaction in solution of dissolved hydroxytitanium complexes with Ba2+ ions. Either way, the reaction involves dehydration, for example, for the rearrangement mechanism (1).H2TiO3+Ba2++2OH-�BaTiO3+H2O (1) Dehydration is slow in superheated fluids and leads to partial retention of H2O and OH- in BaTiO3, especially for powders prepared at low temperatures. The variation in particle size as a function of temperature and time are shown in Fig. 5.All powders have narrow particle size distribution ranges and for each temperature, the mean particle size approximately doubles over a period of ca. 20 h, Fig. 4 Percentage yield (weight%) of powder products from reactions e.g. 23±5 to 43±12 nm, between 2 and 24 h at 85 °C and at 85 and 180 °C and their associated water content (mol%) as a from 53±14 to 80±15 nm at 180 °C. Particle size values after function of reaction period.reaction at 120 °C were intermediate between those at 85 and 180 °C. On comparing the particle size values with the variations in product yield, it appears that after an initial ‘burst’ weight loss commenced at ca. 100 °C and was complete at ca. 600 °C. The powders prepared at 85 °C clearly contained of nucleation, growth of BaTiO3 particles is a rather slow process. substantially more water, especially for short reaction periods where up to 4 weight% loss was recorded.The water content Micrographs of powders obtained at 85 and 180 °C for 2 h are shown in Fig. 6. At 85 °C, they have poor crystallinity and of powders for each reaction temperature remained reasonably constant after ca. 8 h, Fig. 4. This coincides with the ill-defined morphology, Fig. 6(a), whereas at 180 °C they are J. Mater. Chem., 1999, 9, 83–91 87Fig. 5 Particle size histograms for powders prepared at 85 °C after 2 (a) and 24 h (b) and at 180 °C after 2 (c) and 24 h (d). well-defined, regular particles, Fig. 6(b). The powders prepared at 85 °C have a higher water content and presumably, a higher concentration of lattice defects, such as OH- ions.Despite the poor crystallinity of such powders, the presence of the peak at 305 cm-1 in all Raman spectra, Fig. 3(a), demonstrates that these small particles, ca. 20 nm, are tetragonal rather than cubic. This result suggests that the widely-cited ‘critical’ particle/grain size model is incorrect and that distortion of Ti–O bonds within the BaTiO3 lattice is possible, even for particles as small as ca. 20 nm and which contain substantial amounts of lattice defects, such as OH- ions. Although our results clarify some of the existing problems in the literature we are unable, at this stage, to identify the reaction mechanism(s) involved in hydrothermal synthesis of BaTiO3. In the final section we discuss some of the properties of heat treated, hydrothermal BaTiO3 powders.III Heat treated powders Hydroxyl ions play an important role in the synthesis of BaTiO3 via hydrothermal processing, as high pH environments, >12, are required to obtain single phase materials. Infrared spectroscopy17,43,44 and TGA45 have both been employed to demonstrate that as-prepared hydrothermal powders contain weakly-bound water molecules adsorbed onto particle surfaces and more strongly bonded structural water in the form of lattice OH- ions.In general, only total water contents are reported as deconvolution into the two distinct types is diYcult. In order to maintain electro-neutrality, it is generally accepted that barium vacancies (VBa) are created on the surfaces of individual particles46 to compensate for the incorporation of lattice OH- ions on O2- sites (OHOV), according to eqn.(2). 2[OHOV]=VBa (2) On heating above ca. 300 °C, lattice OH- ions are removed as follows, Fig. 6 TEM micrographs of powders prepared at 85 (a) and 180 °C (b), after a reaction period of 2 h. 2OHOV�Oo x+VOVV+H2O (g) (3) 88 J. Mater. Chem., 1999, 9, 83–91resulting in powders which contain significant concentrations of VBa and VOVV.High concentrations of lattice hydroxyl ions in as-prepared hydrothermal powders therefore influence the stoichiometry (Ba/Ti ratio), defect chemistry and grain growth of sintered BaTiO3 ceramics. Elimination of water in our powders was complete by ca. 600 °C, but there were no significant changes in the Raman spectra of the dehydrated powders until samples were heated >1000 °C.These subtle changes in spectra will be discussed elsewhere,47 however, the peak at 305 cm-1 was present in all processed powders, indicating the distorted crystal symmetry of the particles, irrespective of the presence of adsorbed water or lattice hydroxyl ions. This result contradicts the suggestion that lattice hydroxyl ions play a crucial role in controlling the crystal symmetry of hydrothermal BaTiO3 powders17,44,45 and is in agreement with the work of Frey and Payne who came to the same conclusion for BaTiO3 powders prepared via a Fig. 8 Lattice parameters (&) and c/a ratio (#) of BaTiO3 powders sol–gel route.48 prepared at 85 °C for 72 h as a function of processing temperature. A selection of XRD patterns over a limited 2h range for powders prepared at 85 °C and heated up to 1100 °C are shown As mentioned previously, cation and anion defects influence in Fig. 7. At 800 °C, the broad peak at ca. 45.2 ° (in as-made the stoichiometry, grain growth and electrical properties of hydrated powders) sharpens but clear splitting is not apparent BaTiO3 powders. The diVusion coeYcients for oxygen and until heating at ca. 1100 °C. In general, XRD reflections barium vacancies in BaTiO3 become appreciable above 800 become sharper and more intense after heating at higher and 1100 °C, respectively. Thus, on heating hydrothermal temperatures. XRD patterns for powders heated above 1100 °C powders at elevated temperatures, any barium deficiency, were indexed on a tetragonal unit cell, whereas, all others were either as compensating cation defects for lattice hydroxyl ions indexed on a cubic unit cell.Lattice parameters and c/a ratios or as a consequence of acid washing, or excess titanium, in are shown in Fig. 8 and demonstrate that powders heated at the form of unreacted Ti-containing precursor, should result temperatures &100 °C have c/a ratios of ca. 1.01, which is in in the formation of Ti-rich phases.EPMA and SEM on good agreement with those reported in the literature.7 ceramic pellets sintered at 1350 °C from the powders produced at 85 and 180 °C after 24 h, both demonstrated the existence of Ti-rich secondary phases. For powders produced at 85 °C, where unreacted TiO2 was detected via Raman spectroscopy, Fig. 3(a), secondary phases were clearly visible as dark intergranular regions in back scattered electron images (BSE), Fig. 9(a).EPMA results identified the presence of both BaTi2O5 and Ba6Ti17O40 in inter-granular regions but their volume fractions were too small to be detected via XRD. For powders produced at 180 °C, where the products appeared phase-pure by Raman spectroscopy, sintered pellets contained only a very small Fig. 7 XRD diVractograms over the 2h range 44–46 ° of BaTiO3 Fig. 9 Back scattered electron images of ceramic pellets sintered at prepared at 85 °C (top) and after heat treatment at 800, 1000 and 1100 °C. 1350 °C from powders produced at 85 (a) and 180 °C (b). J. Mater. Chem., 1999, 9, 83–91 89volume fraction of Ba6Ti17O40, as shown by the dark, isolated regions in the BSE image, Fig. 9(b).The Ti-rich secondary phases in powders prepared at 85 °C are clearly associated with unreacted Ti-precursor, which reacts with surrounding BaTiO3 particles at elevated temperatures to form Ti-rich binary phases. It is unclear, however, whether the secondary phase produced from powders prepared at 180 °C is associated with unreacted Ti-precursor which has not been detected via Raman spectroscopy (or analytical TEM) or is associated with a compositional Ba/Ti gradient within individual particles of the as-prepared powders.Direct evidence of VBa on the surfaces of individual particles, either from incorporation of lattice hydroxyl ions during hydrothermal synthesis or via acid/aqueous media wet milling of powders has been limited.Abicht et al.49 used HREM and EELS to demonstrate that the outer surfaces of individual BaTiO3 particles which have been wet milled in aqueous media Fig. 11 Log resistivity versus 1000 K/T for pellets of powder prepared show evidence of Ba2+ leaching. They propose that a Ba/Ti at 85 °C and sintered at 1000 °C (filled squares) and 1350 °C (open concentration gradient exists within individual particles which symbols).Calculated activation energies are in eV. consists of a 3–5 nm thick TiOx-rich outer layer followed by a intermediate layer, ca. 10 nm thick, with a molar Ba/Ti ratio omena at the electrodes and supports the idea that conduction increasing from 0 to 1. These core-shell structures aVect the is mainly by means of ions at these temperatures. sintering of BaTiO3 powders due to the reactivity of the Ti- Bulk resistivity values were extracted from the low frequency rich, outer layers.Our observation of small quantities of Ti- intercept of the semi-circular arc with the Z¾ axis on the Z* rich intergranular regions in sintered pellets of hydrothermal plots. Initially, the bulk resistivity increases on heating, as powders produced at 180 °C is consistent with this model but shown by the increase in diameter of the bulk semi-circular in-depth studies using HREM and EELS are required to arc in the Z* plots, Fig. 10, and in the log resistance, Arrheniusestablish the presence or absence of any Ba/Ti compositional type plot shown in Fig. 11. The bulk resistivity rises by several gradient within individual particles of BaTiO3 (produced via orders of magnitude on heating to ca. 150 °C, then remains hydrothermal synthesis). temperature independent before decreasing in accordance with Heat treatment of hydrothermal powders at 1000 °C is the Arrhenius law at temperatures above ca. 350 °C. In this suYcient to remove any adsorbed or structural water and create high temperature region, there was no evidence of the low oxygen vacancies within the lattice, eqn. (3).It is insuYcient, frequency, inclined spike in Z*; instead, the plots consisted of however, to cause substantial migration of barium vacancies or a single, bulk semi-circular arc, indicating that conduction was densification and grain growth via liquid phase sintering involv- predominantly electronic. The resistivity behaviour was reversing Ti-rich outer particle surfaces and/or any excess unreacted ible on thermal cycling.For comparison, bulk and grain Ti-precursor material. Consequently, ceramic pellets formed at boundary resistivities extracted from high and low frequency 1000 °C are poorly sintered, have low density, ca. 65% of the semicircular arcs in Z* plots for a pellet sintered at 1350 °C theoretical density, and consist of small grains with very poor are also shown in Fig. 11. These values are consistent with inter-granular contact. Despite this the pellets exhibit exception- those reported in the literature for dense BaTiO3 ceramics.50 ally low room temperature resistivities, ca. 10–50MV. It should be noted that the room temperature resistivity of Complex impedance plane, Z*, plots at 32 and 44 °C for a these ceramics is in excess of 1 TV and that no low frequency, pellet sintered at 1000 °C are shown in Fig. 10.The plots consist inclined spike was observed in any Z* plots. of a high frequency semicircular arc with an associated capaci- The exceptionally low bulk resistivity value of 20 MV at tance of 32 pF and a low frequency-spike with an associated 25 °C, the presence of a low frequency spike in Z* plots below capacitance in the order of 1 mF.The capacitance value for the ca. 150 °C and the fact that resistivity initially rises with bulk or intra-granular component is consistent with a poorly temperature, all indicate that BaTiO3 prepared via hydrothermal sintered BaTiO3 ceramic and the low frequency, inclined spike synthesis at 85 °C and sintered at 1000 °C is a modest proton is attributable to ionic polarisation and diVusion-limited phen- conductor, especially at temperatures close to room temperature.Presumably water vapour is adsorbed from the ambient on cooling (from the sintering temperature of 1000 °C) and partially reverses the reaction given in eqn. (3), thus, converting doubly ionised oxygen vacancies into lattice hydroxyl ions.Subsequent reheating removes water rather easily and the bulk resistivity increases rapidly; however, temperatures in excess of 350 °C are required before the predominant charge carriers are electronic. More detailed studies of this unusual bulk resistivity behaviour will be reported elsewhere,51 however, such materials are clearly poor dielectrics. To our knowledge, this is the first time that proton conduction has been demonstrated in oxygen-deficient BaTiO3 materials.Many other oxygen-deficient perovskites such as doped, alkaline earth zirconates52 and cerates53 are well-known proton conductors. Several diVerent mechanisms of incorporation of protonic defects in such oxides have been suggested,54 involving interstitial protons and lattice hydroxyl ions, H2O (g)+VOVV=2HiV+OO x (4) Fig. 10 Complex impedance plane plots at 32 and 44 °C for a pellet H2O (g)+OO x+VOVV=2OHOV (5) of powder prepared at 85 °C and sintered at 1000 °C. Selected frequencies (in Hz) are shown for filled symbols. however, the detailed mechanism remains unclear. 90 J. Mater. Chem., 1999, 9, 83–9111 M. Kataoka, K.Suda, N. Ishizawa, F. Marumo, Y. Shimizugawa Conclusions and K. Ohsumi, J. Ceram. Soc. Jpn., 1994, 102, 213. 12 K. Uchino, E. Sadanaga and T. Hirose, J. Am. Ceram. Soc., 1989, Amorphous H2TiO3 is an excellent Ti-precursor to use in the 72, 1555. hydrothermal synthesis of BaTiO3. Phase pure powders with 13 F. J. Gotor, C. Real, M. J. Dianez and J. M. Criado, J. Solid State small particle sizes, ca. 40–80 nm and narrow particle size Chem., 1996, 123, 301. distributions can be prepared at 180 °C after 24 h. Although 14 S. Schlag and H. F. Eicke, Solid State Commun., 1994, 91, 883. initial reaction is very fast, ca. 90% yield after 8–10 h, extended 15 B. D. Begg, E. R. Vance and J. Nowotny, J. Am. Ceram. Soc., 1994, 77, 3186. heating at 180 °C is required to drive reactions to completion. 16 H. I. Hsiang and F. S Yen, J. Am. Ceram. Soc., 1996, 79, 1053. Lowering the reaction temperature produces powders with 17 R. Vivekanandan and T. R. N. Kutty, Powder Technol., 1989, even smaller particle sizes but very long reaction periods are 57, 181. required, >72 h, to ensure complete reaction. In addition, the 18 K. Uchino, N. Lee, T. Toba, N. Usuki, H.Aburatani and Y. Ito, powders are poorly crystalline and have high water content. J. Ceram. Soc. Jpn., 1992, 100, 1091. In order to obtain phase pure BaTiO3 ceramics it is import- 19 F. Yen, C. T. Chang and Y. Chang, J. Am. Ceram. Soc., 1970, 73, 3422. ant to control the Ba/Ti ratio. As the rate limiting step in the 20 J. Nowotny and M. Rekas, Solid State Ionics, 1991, 49, 135.synthesis involves reaction of the Ti-precursor, it is important 21 D. Hennings, M. Klee and R. Waser, Adv. Mater., 1991, 3, 334. to use a technique which can detect very small quantities of 22 M. Mori and T. Kineri, J. Am. Ceram. Soc., 1995, 78, 2391. unreacted amorphous or crystalline Ti-precursor material. We 23 A. I. Kingon, S. K. StreiVer, C. Basceri and S. R. Summerfelt, have clearly demonstrated that Raman spectroscopy, rather MRS Bull., 1996, 6, 46. 24 K. Wa. Gachigi, U. Kumar and J. P. Dougherty, Ferroelectrics, than XRD, is a simple and eVective technique for this purpose. 1993, 143, 229. Raman spectra of powders with an average particle size as 25 F. Chaput and J. P. Boilt, J. Am. Ceram. Soc., 1990, 73, 942. small as ca. 20–40 nm indicate asymmetry within the TiO6 26 M.Demartin, C. Herard, C. Carry and J. Lemaitre, J. Am. Ceram. octahedra of the BaTiO3 lattice. This contradicts the widely Soc., 1997, 80, 1079. cited ‘critical’ particle size theory for the stabilisation of the 27 W. Zhu, C. C. Wang, S. A. Akbar and R. Asiaie, J. Mater. Sci., cubic polymorph, at least for particle sizes in excess of ca. 1997, 32, 4303. 28 W.Zhu, C. C. Wang, S. A. Akbar, R. Asiaie, P. K. Dutta and 20–40 nm. Raman spectra of powders heat treated at ca. M. A. Alim, Jpn. J. Appl. Phys., 1996, 35, 6145. 1000 °C to remove adsorbed and structural water are very 29 A. Beauger, J. C. Mutin and J. C. Niepce, J. Mater. Sci., 1983, similar to that of as-prepared powders. This contradicts the 18, 3041. suggestion that lattice hydroxyl ions stabilise the cubic poly- 30 A.Beauger, J. C. Mutin and J. C. Niepce, J. Mater. Sci., 1983, morph of BaTiO3 prepared via a wet chemical technique, such 18, 3543. 31 J. C. Mutin and J. C. Niepce, J. Mater. Sci. Lett., 1984, 3, 591. as hydrothermal synthesis. Finally, as-prepared powders con- 32 A. Beauger, J. C. Mutin and J. C. Niepce, J. Mater. Sci., 1984, tain many defects, primarily lattice OH- ions.Preliminary 19, 195. conductivity results on pellets of powders which have been 33 R. Asiaie, W. Zhu, S. A. Akbar and P. K. Dutta, Chem. Mater., heated treated to remove lattice hydroxyl ions reveal these 1996, 8, 226. materials to be modest proton conductors at room 34 W. S. Clabaugh, E. M. Swiggard and R. Gilchrist, J. Res. Natl. temperature. Bur.Stand., 1956, 56, 289. 35 M. P. Pechini, US Pat., 3 330 697, 1967. 36 N. J. Ali and S. J. Milne, Trans. J. Br. Ceram. Soc., 1987, 86, 113. 37 M. H. Frey and D. A. Payne, Chem.Mater., 1995, 7, 123. Acknowledgements 38 A. D. Hilton and R. Frost, Key Eng. Mat., 1992, 66/67, 145. 39 M. M. Lencka and R. E. Riman, Chem. Mater., 1993, 5, 61. The authors would like to thank Dr Eric Lachowski for 40 W.Hertl, J. Am. Ceram. Soc., 1988, 71, 879. assistance with electron microscopy, Dr Alison Coats for 41 J. O. Eckert Jr., C. C. Hung-Houston, B. L. Gersten, EPMA, Dr Susan Blake for XRD, Mr James Marr for BET M. M. Lencka and R. E. Riman, J. Am. Ceram. Soc., 1996, 79, analysis and Professor Tony West for useful discussions. The 2929. EPSRC and AIST (Japan) are gratefully acknowledged for 42 C.H. Perry and D. B. Hall, Phys. Rev. Lett., 1965, 15, 700. financial support. 43 G. Busca, V. Buscaglia, M. Leoni and P. Nanni, Chem. Mater., 1994, 6, 955. 44 D. Hennings and S. Schreinemacher, J. Eur. Ceram. Soc., 1992, 9, 4. References 45 S. Wada, T. Suzuki and T. Noma, J. Ceram. Soc. Jpn., 1996, 104, 383. 1 B. JaVe, W. R. Cook and H. JaVe, Piezoelectric Ceramics, 46 R.Waser, J. Am. Ceram. Soc., 1988, 71, 58. Academic Press, London, 1971. 47 I. J. Clark, T. Takeuchi, N. Ohtori and D. C. Sinclair, unpub- 2 D. Hennings, Int. J. High Technology Ceramics, 1987, 3, 91. lished work. 3 G. Arlt, D. Hennings and G. de With, J. Appl. Phys., 1985, 58, 48 M. H. Frey and D. A. Payne, Phys. Rev. B, 1996, 54, 3158. 1619. 49 H. P. Abicht, D.Voltzke, A. Roder, R. Schneider and 4 M. H. Frey, Z. Xu, P. Han and D. A. Payne, Ferroelectrics, 1998, J. Woltersdorf, J. Mater. Chem., 1997, 7, 487. 206, 337. 50 N. Hirose and A. R.West, J. Am. Ceram. Soc., 1996, 79, 1633. 5 W. R. Bussen, L. E. Cross and A. K. Goswami, J. Am. Ceram. 51 I. J. Clark and D. C. Sinclair, unpublished work. Soc., 1966, 49, 33. 52 H. Iwahara, T. Esaka, H.Uchida and N. Maeda, Solid State 6 G. Arlt, Ferroelectrics, 1990, 104, 217. Ionics, 1981, 3/4, 359. 7 A. Morell and J. C. Niepce, J. Mater. Educ., 1991, 13, 173. 53 H. Uchida, H. Yoshikawa and H. Iwahara, Solid State Ionics, 8 T. Takeuchi, K. Ado, T. Asai, H. Kageyama, Y. Saito, 1989, 34, 103. C. Masquelier and O. Nakamura, J. Am. Ceram. Soc., 1994, 77, 54 K. D. Kruer, Chem.Mater., 1996, 8, 610. 1665. 9 W. Kanzig, Phys. Rev., 1955, 98, 549. 10 K. Kinoshita and A. Yamaji, J. Appl. Phys., 1976, 47, 371. Paper 8/05756G J. Mater. Chem., 1999, 9, 83–91 91 J O U R N A L O F C H E M I S T R Y Materials Hydrothermal synthesis and characterisation of BaTiO3 fine powders: precursors, polymorphism and properties† Iain J. Clark,a Tomanari Takeuchi,b Noikazu Ohtoric and Derek C.Sinclaira aChemistry Department, University of Aberdeen, Meston Walk, Aberdeen, UK AB24 3UE bOsaka National Research Institute, 1–8-31, Midorigaoka, Ikeda, Osaka, Japan cGraduate School of Science and Technology, Niigatu University, Ikarashi 2-no cho, Niigata, Japan Received 19th June 1998, Accepted 22nd July 1998 The influence of two Ti-precursors, TiO2 (anatase) and H2TiO3 (b-titanic acid), on the purity and particle size of BaTiO3 powders prepared via hydrothermal synthesis is discussed. Amorphous H2TiO3 was found to be an excellent Ti-precursor material and oVers several advantages over crystalline anatase.Phase pure powders which have small particle sizes, ca. 40–80 nm and narrow particle size distributions can be prepared at 180 °C after 24 h using H2TiO3 as a precursor material.Although the initial reaction is very fast, ca. 90% yield after 8–10 h, extended reaction periods at 180 °C are required in order to drive the reaction to completion. Lowering the reaction temperature from 180 to 85 °C does produce powders with even smaller particle sizes, however, very long reaction periods are required, e.g.>72 h, to ensure complete reaction.Raman spectra of as-prepared and heat treated (1000 °C) powders with average particle sizes as small as ca. 20–40 nm indicate asymmetry within the TiO6 octahedra of the BaTiO3 lattice. These results contradict the widely cited ‘critical’ particle size theory for the stabilisation of the cubic polymorph, at least for particle sizes greater than ca. 20–40 nm. As-prepared powders contain many defects, primarily in the form of lattice OH- ions. Preliminary ac impedance spectroscopy data on samples heat treated to remove lattice hydoxyl ions demonstrate these materials to be modest proton conductors. size eVect but there is still no clear consensus as to its origin.‡ Introduction Two commonly cited models involve the generation of higher- BaTiO3 is one of the most widely used and studied ferroelectric than-usual stress in fine grained ceramics5,6 whereas a third materials in the electro-ceramics industry.It has various suggests a core-shell structure,7,8 whereby grains below a polymorphs, all of which are based on the perovskite struc- ‘critical’ size (<0.2 mm) have cubic symmetry, intermediate ture;1 however, the two most studied are the tetragonal, t, and grain sizes (ca. 1 mm) have a ferroelectric-tetragonal core with cubic, c, polymorphs. t-BaTiO3 forms between ca. 0 and 130 °C a cubic surface layer and large grains (5 mm) are essentially whereas c-BaTiO3 is stable above 130 °C. In the t-polymorph, tetragonal. It should be noted that there is not an accepted titanium ions are displaced from the centrosymmetric position ‘critical’ particle or grain size below which the cubic polymorph within TiO6 octahedra and give rise to spontaneous polaris- is stabilised, on the contrary, a wide range of values, ca.ation. Due to this asymmetry within the crystal structure, t- 25–190 nm have been reported.9–16 In addition, a variety of BaTiO3-based materials have high permittivities and are widely reasons have been proposed for the room-temperature stabilisemployed as dielectrics in ceramic capacitors.2 The transition ation of the cubic polymorph, including the presence of lattice from the polar (ferroelectric) t- to non-polar (paraelectric) c- ‘defects’ such as hydroxyl ions17 (associated with powders polymorph normally occurs at ca. 130 °C, the Curie tempera- formed via wet chemical methods), small deviations in Ba/Ti ture, Tc.Above Tc, the Ti ions occupy, on average, the stoichiometry7 and excess surface energy associated with centrosymmetric position within TiO6 octahedra and there is ultrafine particles.18,19 Clearly, much remains to be done in no net polarisation within the solid. order to establish the factors which influence and control the A typical room temperature permittivity value, e25, for crystal symmetry and electrical properties of sub-micron ceramic t-BaTiO3 with an average grain size of ca. 10 mm is BaTiO3 powders and ceramics. 1000–2000; however, this rises sharply to a maximum value, Despite these fundamental problems, BaTiO3 will continue emax, of ca. 10 000 at Tc. In such ‘large’ grained ceramics, a to be used in the manufacture of thermistors,20 multilayer variety of ferroelectric domain structures form to relieve capacitors,21 electro-optic devices22 and DRAM (dynamic internal stress associated with the cubic to tetragonal phase random access memories)23 into the next century.Improved transformation. Although a detailed discussion of domain performance and miniaturisation of BaTiO3-based devices, structures in BaTiO3 powders and ceramics is outwith the either in the form of thin ceramic layers, ca. 20mm, for scope of this paper, it is important to note that ferroelectric multilayer capacitors or as thin films for integrated electronic domain structures and dipole–dipole interactions control the circuits remains a priority. High permittivities and miniaturispermittivity characteristics of BaTiO3 and that these depend ation can be achieved by controlling the ceramic microstructure on particle or grain size.For example, Arlt et al.3 have shown which, in turn, depends on the homogeneity, composition, that e25 for BaTiO3 ceramics can be optimised to a value surface area and particle size of the starting BaTiO3 powder. approaching 5000 by controlling the grain size to be ca.In addition, it is well known that BaTiO3 undergoes exagger- 0.7–1.0 mm. ated grain growth during sintering at elevated temperatures Several models have been proposed to explain this grain and long sintering periods. This can produce ceramics †Basis of the presentation given at Materials Chemistry Discussion ‡During the preparation of this manuscript, Frey et al.4 published a paper which appears to have clarified many of the problems associated No. 1, 24–26 September 1998, ICMCB, University of Bordeaux, France. with the origin of the so-called grain size eVect in BaTiO3. J. Mater. Chem., 1999, 9, 83–91 83containing very large grain sizes, e.g. >50 mm or duplex and the surrounding BaTiO3. For this model, complete reaction may be diYcult to achieve due to increasing encapsulation microstructures of low density consisting of both large (>10 mm) and small (<1 mm) grains.For these reasons, many of TiO2 via product formation. Homogeneous and heterogeneous dissolution–precipitation synthetic methods and sintering profiles24–28 have been investigated in an attempt to produce sub-micron, deagglomerated models have also been proposed.In the homogeneous model, a low concentration of TiO2 dissolves in the form of soluble fine powders of BaTiO3 that can be sintered into dense, finegrained ceramics. hydroxytitanium complexes which then react with Ba2+ ions in solution to precipitate BaTiO3. In the heterogeneous model, Traditionally, BaTiO3 has been produced via the mixed oxide route.This method involves repeated calcination and BaTiO3 nuclei form on the surfaces of the dissolving TiO2 particles and, as for the in situ transformation mechanism, the regrinding of BaCO3 and TiO2 powders above 1100 °C. The reaction mechanism in air has been proposed to take place in reaction becomes limited by the increasing isolation of the reactants via BaTiO3 product layers on the surface of TiO2 at least three stages29–32 and relies on the diVusion of Ba2+ ions into TiO2.Firstly, BaCO3 reacts with the outer surface particles. Eckert et al. reviewed these models41 in more detail and regions of TiO2 to form a surface layer of BaTiO3 on individual TiO2 grains. Further diVusion of Ba2+ ions into TiO2 necessi- undertook a kinetic study of hydrothermal reactions involving barium hydroxide octahydrate and anatase precursors.Two tates the formation of Ba2TiO4 between unreacted BaCO3 and the previously formed BaTiO3. After prolonged sintering reaction regimes were clearly identified; during the early stages the reaction is controlled by a dissolution–precipitation pro- periods, the intermediate Ba-rich phase Ba2TiO4 reacts with the remaining TiO2 in the core-regions of the TiO2 grains to cess, whereas, for the second regime at longer reaction times the results were inconclusive.The authors suggest two plausible form BaTiO3. Although the nominal starting Ba/Ti ratio is 151, long reaction times are required to form homogeneous but diVerent explanations for their kinetic data, one model is based on heterogeneous dissolution–precipitation followed by powders free from secondary phases.For such a complex and slow reaction mechanism, it is desirable to have high surface in situ transformation, whereas, the second model suggests that dissolution–precipitation is the controlling mechanism, area TiO2 powders of fine particle size and narrow size distribution in order to control the morphology and grain size with nucleation and growth of BaTiO3 controlling the first regime and the dissolution rate of TiO2 controlling the of the resultant BaTiO3 powder.Nevertheless, the mixed oxide route tends to produce coarse, agglomerated powders that second regime. In the present paper we discuss some of our recent work on require high sintering temperatures to form dense ceramics.In order to produce powders that are suitable for sintering the hydrothermal synthesis of BaTiO3 powders and highlight some advantages of using amorphous H2TiO3 as opposed to into micron-grain-sized ceramics for MLC applications there has been much interest in developing wet and novel chemical crystalline anatase as a precursor material. In particular, H2TiO3 promotes faster and more-complete reactions and also routes.As these reactions take place in the liquid, as opposed to the solid state, more intimate mixing of the cations is produces powders with much smaller particle sizes. We discuss the influence of temperature on the rate of reaction and on achieved and consequently, much shorter diVusion pathways are created. Deagglomerated, sub-micron powders of BaTiO3 the particle size, distribution and water content of the BaTiO3 powders. The importance of using Raman spectroscopy to can then be obtained at low temperatures and short reaction periods, for example via hydrothermal synthesis at 240 °C detect the presence of trace amounts of unreacted Ti-containing precursors and to yield information on the polymorphism of for 24 h.33 Other common methods include, oxalate (Clabauch),34 citrate (Pechini),35 catecholate (Milne)36 and sub-micron powders is discussed.Finally, we present preliminary results on powders and ceramics heat treated above sol–gel.37 Although the reaction mechanisms and advantages and disadvantages of many of these routes have been dis- 800 °C that support the idea that hydrothermal BaTiO3 powders contain many defects which influence their physical cussed,38 for the present purposes it is worth discussing these issues for hydrothermal synthesis.properties. BaTiO3 powders have been prepared via hydrothermal processing since the 1940’s and this method is commonly employed Experimental to produce commercial powders. Although it is well established that BaTiO3 is the only thermodynamically stable binary Ba(OH)2·8H2O (98+%, Aldrich) and anatase (99.9%, Aldrich) or b-H2TiO3 (+99%,Mitsuwa Chemicals) were mixed compound produced under all conditions,39 i.e.pH, Ba/Ti ratio, reaction temperature, etc. products can contain unre- with a Ba5Ti ratio of 1.0551 in a 23 ml Teflon-lined pressure vessel (Model 4749, Parr Instruments) together with 10 ml of acted TiO2 and BaCO3.Normally, an excess of Ba2+ in the starting solution is employed (Ba/Ti=1.05–1.1051) in an de-ionised water. The vessel was sealed, shaken and placed in an oven at either 85, 120 or 180 °C for periods ranging from attempt to remove any unreacted TiO2 and therefore drive the reaction to completion. Any excess Ba, in the form of BaCO3 1 to 72 h.After cooling, the contents of the bomb were diluted in 30 ml of 0.1M formic acid in an attempt to dissolve any can then be removed via acid washing of the powders. Although this procedure is eVective in controlling the purity BaCO3 formed by the addition of excess Ba2+ to the starting solution. The mixture was vacuum filtered using a Buchner of the powders, relatively little is known about the influence of acid washing on powder stoichiometry as Ba2+ may be funnel, the residue thoroughly washed with distilled water and dried overnight in air at 120 °C.The percentage yield of leached from the surfaces of the particles. Although controlling the precise stoichiometry of hydrothermal BaTiO3 powders BaTiO3 was estimated from the weight of the dried powders.Phase purity and polymorphism of the BaTiO3 powders can be problematic, the low cost and easy handling of the reagents, and the fast reaction rate at low temperatures ensures were studied by X-ray diVraction (XRD) and Raman spectroscopy. XRD was carried out on a Stoe STADI P automated that deagglomerated powders consisting of small particles of narrow size distribution are readily obtained.powder diVractometer employing monochromatic Cu-Ka1 radiation and a linear position sensitive detector over the 2h Several mechanisms have been proposed for the hydrothermal synthesis of BaTiO3. Hertl40 has suggested an in situ range 20 to 80 °. Cubic and tetragonal polymorphs were distinguished by the splitting of the (200) peak at 2h#45 °. transformation whereby dissolved Ba2+ ions react with undissolved TiO2 to produce a continuous layer of BaTiO3 on the Unit cell parameters were obtained by refinement of XRD data.Raman spectroscopy was carried out on a Jobin Yvon surface of TiO2 particles. The reaction is then controlled either by diVusion of Ba2+ through the product BaTiO3 layer or T64000 using a 100 mW Ar laser with a wavelength of 514.5 nm. A 300 mm slit and integration time of 5 s were used further reaction between unreacted TiO2 within grain interiors 84 J.Mater. Chem., 1999, 9, 83–91giving a resolution of 8.4 cm-1. Spectra were measured over An important aspect of this work is to obtain homogeneous the range 58–1103 cm-1. Impurity phases such as BaCO3 and particles of controlled stoichiometry which can be sintered TiO2 (anatase) were detected from characteristic peaks at 1060 into dense ceramics of uniform microstructure.The Ba5Ti and 150 cm-1, respectively. The presence of a peak in the ratio of the particles is an important parameter but is not spectra of BaTiO3 at 305 cm-1 indicates asymmetry within readily obtained via XRD or Raman spectroscopy. We choose, TiO6 octahedra42 and was used to distinguish between tetra- therefore, to study the phase assemblage and Ba5Ti ratio of gonal and cubic polymorphs.the grains in sintered ceramics using EPMA and from this, Thermogravimetric analysis (TGA) was carried out using a deduce information about the Ba5Ti ratio of the powders. Stanton Redcroft TG-DTA simultaneous thermal analyser This is an indirect method for obtaining information on the (model STA 1000/1500) to determine the water content of as- stoichiometry of the powders, but since neither barium nor prepared powders.Samples were accurately weighed to five titanium are particularly volatile at the sintering temperature, decimal places in a platinum crucible and weight losses we perceive this method to be valid. recorded from 25–1000 °C against an alumina reference. All experiments were carried out in air at a heating rate of 10 °Cmin-1 with an amplifier setting of 50 mV.Results and discussion Powder morphology was determined via transmission electron microscopy (TEM) using a JEOL electron microscope The results are subdivided into three sections. The first section (model 2000EX TEMSCAN) operating at 200 kV.Particle discusses the influence of two Ti-precursors, TiO2 (anatase) size and distribution were obtained by measuring the cross and H2TiO3 (b-titanic acid), on the purity and particle size of diagonals of approximately 130 particles from scanned nega- the BaTiO3 powders. The second section assesses the role of tives of TEM images. Energy dispersive X-ray analysis (EDX) reaction temperature on the rate of reaction, purity, particle was used in an attempt to determine the Ba5Ti ratio of the size and water content of BaTiO3 powders prepared from BaTiO3 particles and also to detect the presence of any Ba(OH)2·8H2O and H2TiO3.The final section discusses preunreacted TiO2. liminary electrical property measurements on powders heat Electron probe micro-analysis (EPMA) was employed to treated above 800 °C. determine the homogeneity of sintered pellets fabricated from the various powders. Pellets were sintered at ca. 1350 °C and I Titanium precursors a Cameca SX51 electron microprobe employed using an accelerating voltage of 20 kV and a beam current of 40 nA. One of the major problems associated with the hydrothermal Ba-La and Ti-Ka lines were measured using benitoite, synthesis of BaTiO3 is the low solubility of TiO2 in the highly BaTiSi3O9, as a standard.Prior to analysis, small pieces of alkaline environments, ca. pH>12, required for synthesis. pellet were mounted in resin blocks, surface polished and Although it is well documented that diVerent Ti-precursors carbon coated. and reaction temperatures influence the rate of reaction there Electrical measurements were carried out using a Solartron have been few comprehensive studies on the influence of these frequency response analyser (model 1250 or 1260) combined parameters on the purity and particle size of hydrothermal with a Solartron dielectric interface (model 1296).A frequency powders. Kutty et al. suggested that amorphous Ti-precursors range of 10-3–106 Hz was employed with an applied voltage react fastest,17 then anatase with rutile giving the slowest of 100 mV.Cold pressed pellets of various powders were reaction. Various workers reported an increase in average sintered between 800 and 1350 °C and organopaste and gold particle size and a change (via XRD) from cubic to tetragonal foil electrodes applied to the surfaces of the pellets at 800 °C, symmetry with increasing reaction temperature.9–16We decided prior to mounting in a conductivity jig.Measurements were to reinvestigate some of these findings by comparing the recorded over the temperature range 25–500 °C. reaction rates, powder purity and particle size of hydrothermal Before discussing the results it is important to appreciate BaTiO3 powders prepared at various temperatures and reac- the need for a wide variety of techniques to determine the tion periods using two Ti-precursors: amorphous H2TiO3 phase purity and crystal structure of fine-grain BaTiO3 pow- (BET using N2, surface area 193 m2 g-1) and crystalline ders.XRD is an excellent technique for obtaining information anatase (BET, surface area 8.9 m2 g-1).on the average or long range crystal structure of a material; In general, for a set reaction temperature and period, the however, peak broadening is significant for diVraction from reaction rate was faster and the average BaTiO3 particle size sub-micron sized particles and XRD cannot reveal the very smaller for reactions using the amorphous H2TiO3. Raman subtle unit cell distortions which occur in BaTiO3 for microdospectra of powders prepared under the same conditions for mains of dimensions of ca. 100 A° . In addition, routine XRD the two precursors are shown in Fig. 1. Spectra of anatase and is relatively insensitive to small quantities of impurity phases, the tetragonal polymorph of BaTiO3 prepared via the tra- especially for secondary phases which are amorphous/poorlyditional mixed oxide route are included for comparison.The crystalline and/or have weak X-ray scattering power. In our Raman spectrum of H2TiO3 is not shown but is similar to experience, XRD is sensitive to small quantities of BaCO3 in that of anatase and is also dominated by a large peak at ca. hydrothermally processed BaTiO3 but is much less sensitive in 150 cm-1.There are two noticeable features in the Raman the detection of unreacted Ti-containing precursors, either spectra of the products which are not readily discernible from crystalline anatase or amorphous H2TiO3. the XRD patterns (not shown) and which provide important Uncertainties over crystal symmetry can largely be resolved information. using alternative techniques such as Raman spectroscopy to First, the powder products from the H2TiO3 reaction appear probe short range order or local symmetry.There are reliable to be phase pure BaTiO3, whereas unreacted TiO2, as shown Raman spectra in the literature42 for various BaTiO3 polyby the presence of the peak at ca. 150 cm-1, is clearly present morphs; however, it must be stated that the diVerences in in the products of the anatase reaction.Small amounts of spectra are rather subtle, especially for the orthorhombic, unreacted anatase or amorphous H2TiO3 are diYcult to detect tetragonal and cubic (at temperatures slightly greater than Tc) via routine XRD but are readily detected (by the peak at ca. polymorphs. In contrast to XRD, Raman spectroscopy is 150 cm-1) in the Raman spectra.This result demonstrates the sensitive to the presence of both BaCO3 and unreacted Tiimportance of using Raman spectroscopy to monitor the phase containing precursors. For these reasons we choose to present purity of BaTiO3 powders prepared via hydrothermal syn- Raman spectra, as opposed to XRD patterns to illustrate the thesis. We should stress that single phase powders can, in fact, phase purity and crystal symmetry of as-prepared BaTiO3 powders.be prepared from crystalline anatase at 180 °C, however the J. Mater. Chem., 1999, 9, 83–91 85Fig. 2 TEM micrographs of BaTiO3 powders prepared via hydrothermal synthesis at 180 °C for 24 h using anatase (a) and H2TiO3 (b), as Ti-precursors. Fig. 1 Raman spectra of BaTiO3 prepared via a mixed oxide route (top) and via hydrothermal synthesis at 180 °C for 24 h using H2TiO3 and anatase as Ti-precursors.The bottom spectrum is of crystalline II Reaction temperature anatase. Raman spectra of powder products from two reaction temperatures, 85 and 180 °C after various reaction periods are shown in Fig. 3(a) and (b), respectively. With increasing time, both times required are far in excess of those required using sets of spectra illustrate a decrease in peak intensity at amorphous H2TiO3, typically in excess of 72 h. 150 cm-1 associated with the titanium precursor as BaTiO3 is Second, the presence of a peak at 305 cm-1 in both spectra formed.In fact, the spectrum after reaction at 180 °C for 24 h indicates asymmetry within the TiO6 octahedra of BaTiO3 and shows no evidence of this peak, suggesting that the reaction demonstrates clearly, on a local scale, that as-prepared powders has gone to completion. A small peak at 150 cm-1 is still do not have cubic symmetry.In contrast, XRD patterns apparent in the corresponding spectrum for powders prepared showed appreciable peak broadening, presumably associated at 85 °C. A higher reaction temperature of ca. 180 °C, is clearly with the small particle size and poor crystallinity of the required to ensure a more complete reaction for short reaction powders, however, there was little evidence of peak splitting periods, e.g. 24 h. The high surface area (193 m2 g-1) and the and the XRD data ‘appeared’ consistent with that for the fact that hydroxylation of many Ti–O–Ti bridging bonds has cubic polymorph, as reported previously.9–16 The higher sensi- already been achieved in H2TiO3, ensure rapid formation of tivity of Raman spectroscopy to probe the local rather than BaTiO3, even at 85 °C.It should be noted that many spectra long range structure clearly demonstrates that sub-micron contained a small peak at 1060 cm-1 (not shown) associated powders prepared via hydrothermal processing are tetragonal with a small amount of BaCO3 which was not completely rather than cubic.removed by the mild acid wash. The particle size of powders prepared from H2TiO3 were, The weight of the dried powders was used to calculate the on average, 3–5 times smaller than those prepared from TiO2 percentage yield of BaTiO3 and is shown in Fig. 4. We stress and, to our knowledge, are amongst the smallest reported for that these values represent upper estimates of the yield, especi- BaTiO3 powders prepared hydrothermally. Representative ally for short reaction periods where powders contain unremicrographs of powders prepared at 180 °C for 24 h from acted precursor and are also heavily hydrated. Nevertheless, H2TiO3 and TiO2 are shown in Fig. 2 and demonstrate the these estimates show that reaction is rapid and yields in excess much smaller particle size, ca. 80 nm compared with ca. of 90% are obtained within 8 h of reaction. From the Raman 250 nm. The faster rate of reaction and smaller particle size spectra in Fig. 3, extended periods, e.g. 24 h at elevated of powders prepared from amorphous H2TiO3 demonstrate temperatures (180 °C) are required in order to drive the two clear advantages of using this precursor instead of crystal- reaction close to completion, even for a reactive precursor line anatase.The remaining sections of the paper concentrate such as H2TiO3. on powders and ceramics prepared using H2TiO3 as the Ti- As-prepared powders are hydrated and lose water both adsorbed and structural on heating to ca. 1000 °C. In general, precursor. 86 J. Mater. Chem., 1999, 9, 83–91Fig. 3 Raman spectra of powder products from two reaction temperatures, 85 (a) and 180 °C (b), after various reaction periods. dramatic decrease in reaction rate after ca. 8–10 h, as shown by the percentage yield of BaTiO3. Nucleation of BaTiO3 crystals presumably takes place via a structural rearrangement of the amorphous H2TiO3 associated with the incorporation of Ba2+ or via direct reaction in solution of dissolved hydroxytitanium complexes with Ba2+ ions.Either way, the reaction involves dehydration, for example, for the rearrangement mechanism (1). H2TiO3+Ba2++2OH-�BaTiO3+H2O (1) Dehydration is slow in superheated fluids and leads to partial retention of H2O and OH- in BaTiO3, especially for powders prepared at low temperatures.The variation in particle size as a function of temperature and time are shown in Fig. 5. All powders have narrow particle size distribution ranges and for each temperature, the mean particle size approximately doubles over a period of ca. 20 h, Fig. 4 Percentage yield (weight%) of powder products from reactns e.g. 23±5 to 43±12 nm, between 2 and 24 h at 85 °C and at 85 and 180 °C and their associated water content (mol%) as a from 53±14 to 80±15 nm at 180 °C. Particle size values after function of reaction period. reaction at 120 °C were intermediate between those at 85 and 180 °C. On comparing the particle size values with the variations in product yield, it appears that after an initial ‘burst’ weight loss commenced at ca. 100 °C and was complete at ca. 600 °C. The powders prepared at 85 °C clearly contained of nucleation, growth of BaTiO3 particles is a rather slow process. substantially more water, especially for short reaction periods where up to 4 weight% loss was recorded. The water content Micrographs of powders obtained at 85 and 180 °C for 2 h are shown in Fig. 6. At 85 °C, they have poor crystallinity and of powders for each reaction temperature remained reasonably constant after ca. 8 h, Fig. 4. This coincides with the ill-defined morphology, Fig. 6(a), whereas at 180 °C they are J. Mater. Chem., 1999, 9, 83–91 87Fig. 5 Particle size histograms for powders prepared at 85 °C after 2 (a) and 24 h (b) and at 180 °C after 2 (c) and 24 h (d).well-defined, regular particles, Fig. 6(b). The powders prepared at 85 °C have a higher water content and presumably, a higher concentration of lattice defects, such as OH- ions. Despite the poor crystallinity of such powders, the presence of the peak at 305 cm-1 in all Raman spectra, Fig. 3(a), demonstrates that these small particles, ca. 20 nm, are tetragonal rather than cubic.This result suggests that the widely-cited ‘critical’ particle/grain size model is incorrect and that distortion of Ti–O bonds within the BaTiO3 lattice is possible, even for particles as small as ca. 20 nm and which contain substantial amounts of lattice defects, such as OH- ions. Although our results clarify some of the existing problems in the literature we are unable, at this stage, to identify the reaction mechanism(s) involved in hydrothermal synthesis of BaTiO3.In the final section we discuss some of the properties of heat treated, hydrothermal BaTiO3 powders. III Heat treated powders Hydroxyl ions play an important role in the synthesis of BaTiO3 via hydrothermal processing, as high pH environments, >12, are required to obtain single phase materials.Infrared spectroscopy17,43,44 and TGA45 have both been employed to demonstrate that as-prepared hydrothermal powders contain weakly-bound water molecules adsorbed onto particle surfaces and more strongly bonded structural water in the form of lattice OH- ions. In general, only total water contents are reported as deconvolution into the two distinct types is diYcult. In order to maintain electro-neutrality, it is generally accepted that barium vacancies (VBa) are created on the surfaces of individual particles46 to compensate for the incorporation of lattice OH- ions on O2- sites (OHOV), according to eqn.(2). 2[OHOV]=VBa (2) On heating above ca. 300 °C, lattice OH- ions are removed as follows, Fig. 6 TEM micrographs of powders prepared at 85 (a) and 180 °C (b), after a reaction period of 2 h. 2OHOV�Oo x+VOVV+H2O (g) (3) 88 J. Mater. Chem., 1999, 9, 83–91resulting in powders which contain significant concentrations of VBa and VOVV. High concentrations of lattice hydroxyl ions in as-prepared hydrothermal powders therefore influence the stoichiometry (Ba/Ti ratio), defect chemistry and grain growth of sintered BaTiO3 ceramics. Elimination of water in our powders was complete by ca. 600 °C, but there were no significant changes in the Raman spectra of the dehydrated powders until samples were heated >1000 °C. These subtle changes in spectra will be discussed elsewhere,47 however, the peak at 305 cm-1 was present in all processed powders, indicating the distorted crystal symmetry of the particles, irrespective of the presence of adsorbed water or lattice hydroxyl ions.This result contradicts the suggestion that lattice hydroxyl ions play a crucial role in controlling the crystal symmetry of hydrothermal BaTiO3 powders17,44,45 and is in agreement with the work of Frey and Payne who came to the same conclusion for BaTiO3 powders prepared via a Fig. 8 Lattice parameters (&) and c/a ratio (#) of BaTiO3 powders sol–gel route.48 prepared at 85 °C for 72 h as a function of processing temperature.A selection of XRD patterns over a limited 2h range for powders prepared at 85 °C and heated up to 1100 °C are shown As mentioned previously, cation and anion defects influence in Fig. 7. At 800 °C, the broad peak at ca. 45.2 ° (in as-made the stoichiometry, grain growth and electrical properties of hydrated powders) sharpens but clear splitting is not apparent BaTiO3 powders.The diVusion coeYcients for oxygen and until heating at ca. 1100 °C. In general, XRD reflections barium vacancies in BaTiO3 become appreciable above 800 become sharper and more intense after heating at higher and 1100 °C, respectively. Thus, on heating hydrothermal temperatures.XRD patterns for powders heated above 1100 °C powders at elevated temperatures, any barium deficiency, were indexed on a tetragonal unit cell, whereas, all others were either as compensating cation defects for lattice hydroxyl ions indexed on a cubic unit cell. Lattice parameters and c/a ratios or as a consequence of acid washing, or excess titanium, in are shown in Fig. 8 and demonstrate that powders heated at the form of unreacted Ti-containing precursor, should result temperatures >1100 °C have c/a ratios of ca. 1.01, which is in in the formation of Ti-rich phases. EPMA and SEM on good agreement with those reported in the literature.7 ceramic pellets sintered at 1350 °C from the powders produced at 85 and 180 °C after 24 h, both demonstrated the existence of Ti-rich secondary phases.For powders produced at 85 °C, where unreacted TiO2 was detected via Raman spectroscopy, Fig. 3(a), secondary phases were clearly visible as dark intergranular regions in back scattered electron images (BSE), Fig. 9(a). EPMA results identified the presence of both BaTi2O5 and Ba6Ti17O40 in inter-granular regions but their volume fractions were too small to be detected via XRD.For powders produced at 180 °C, where the products appeared phase-pure by Raman spectroscopy, sintered pellets contained only a very small Fig. 7 XRD diVractograms over the 2h range 44–46 ° of BaTiO3 Fig. 9 Back scattered electron images of ceramic pellets sintered at prepared at 85 °C (top) and after heat treatment at 800, 1000 and 1100 °C. 1350 °C from powders produced at 85 (a) and 180 °C (b). J. Mater. Chem., 1999, 9, 83–91 89volume fraction of Ba6Ti17O40, as shown by the dark, isolated regions in the BSE image, Fig. 9(b). The Ti-rich secondary phases in powders prepared at 85 °C are clearly associated with unreacted Ti-precursor, which reacts with surrounding BaTiO3 particles at elevated temperatures to form Ti-rich binary phases.It is unclear, however, whether the secondary phase produced from powders prepared at 180 °C is associated with unreacted Ti-precursor which has not been detected via Raman spectroscopy (or analytical TEM) or is associated with a compositional Ba/Ti gradient within individual particles of the as-prepared powders. Direct evidence of VBa on the surfaces of individual particles, either from incorporation of lattice hydroxyl ions during hydrothermal synthesis or via acid/aqueous media wet milling of powders has been limited.Abicht et al.49 used HREM and EELS to demonstrate that the outer surfaces of individual BaTiO3 particles which have been wet milled in aqueous media Fig. 11 Log resistivity versus 1000 K/T for pellets of powder prepared show evidence of Ba2+ leaching.They propose that a Ba/Ti at 85 °C and sintered at 1000 °C (filled squares) and 1350 °C (open concentration gradient exists within individual particles which symbols). Calculated activation energies are in eV. consists of a 3–5 nmhick TiOx-rich outer layer followed by a intermediate layer, ca. 10 nm thick, with a molar Ba/Ti ratio omena at the electrodes and supports the idea that conduction increasing from 0 to 1.These core-shell structures aVect the is mainly by means of ions at these temperatures. sintering of BaTiO3 powders due to the reactivity of the Ti- Bulk resistivity values were extracted from the low frequency rich, outer layers. Our observation of small quantities of Ti- intercept of the semi-circular arc with the Z¾ axis on the Z* rich intergranular regions in sintered pellets of hydrothermal plots.Initially, the bulk resistivity increases on heating, as powders produced at 180 °C is consistent with this model but shown by the increase in diameter of the bulk semi-circular in-depth studies using HREM and EELS are required to arc in the Z* plots, Fig. 10, and in the log resistance, Arrheniusestablish the presence or absence of any Ba/Ti compositional type plot shown in Fig. 11. The bulk resistivity rises by several gradient within individual particles of BaTiO3 (produced via orders of magnitude on heating to ca. 150 °C, then remains hydrothermal synthesis). temperature independent before decreasing in accordance with Heat treatment of hydrothermal powders at 1000 °C is the Arrhenius law at temperatures above ca. 350 °C. In this suYcient to remove any adsorbed or structural water and create high temperature region, there was no evidence of the low oxygen vacancies within the lattice, eqn. (3). It is insuYcient, frequency, inclined spike in Z*; instead, the plots consisted of however, to cause substantial migration of barium vacancies or a single, bulk semi-circular arc, indicating that conduction was densification and grain growth via liquid phase sintering involv- predominantly electronic.The resistivity behaviour was reversing Ti-rich outer particle surfaces and/or any excess unreacted ible on thermal cycling. For comparison, bulk and grain Ti-precursor material. Consequently, ceramic pellets formed at boundary resistivities extracted from high and low frequency 1000 °C are poorly sintered, have low density, ca. 65% of the semicircular arcs in Z* plots for a pellet sintered at 1350 °C theoretical density, and consist of small grains with very poor are also shown in Fig. 11. These values are consistent with inter-granular contact. Despite this the pellets exhibit exception- those reported in the literature for dense BaTiO3 ceramics.50 ally low room temperature resistivities, ca. 10–50MV. It should be noted that the room temperature resistivity of Complex impedance plane, Z*, plots at 32 and 44 °C for a these ceramics is in excess of 1 TV and that no low frequency, pellet sintered at 1000 °C are shown in Fig. 10. The plots consist inclined spike was observed in any Z* plots. of a high frequency semicircular arc with an associated capaci- The exceptionally low bulk resistivity value of 20 MV at tance of 32 pF and a low frequency-spike with an associated 25 °C, the presence of a low frequency spike in Z* plots below capacitance in the order of 1 mF.The capacitance value for the ca. 150 °C and the fact that resistivity initially rises with bulk or intra-granular component is consistent with a poorly temperature, all indicate that BaTiO3 prepared via hydrothermal sintered BaTiO3 ceramic and the low frequency, inclined spike synthesis at 85 °C and sintered at 1000 °C is a modest proton is attributable to ionic polarisation and diVusion-limited phen- conductor, especially at temperatures close to room temperature.Presumably water vapour is adsorbed from the ambient on cooling (from the sintering temperature of 1000 °C) and partially reverses the reaction given in eqn.(3), thus, converting doubly ionised oxygen vacancies into lattice hydroxyl ions. Subsequent reheating removes water rather easily and the bulk resistivity increases rapidly; however, temperatures in excess of 350 °C are required before the predominant charge carriers are electronic. More detailed studies of this unusual bulk resistivity behaviour will be reported elsewhere,51 however, such materials are clearly poor dielectrics.To our knowledge, this is the first time that proton conduction has been demonstrated in oxygen-deficient BaTiO3 materials. Many other oxygen-deficient perovskites such as doped, alkaline earth zirconates52 and cerates53 are well-known proton conductors.Several diVerent mechanisms of incorporation of protonic defects in such oxides have been suggested,54 involving interstitial protons and lattice hydroxyl ions, H2O (g)+VOVV=2HiV+OO x (4) Fig. 10 Complex impedance plane plots at 32 and 44 °C for a pellet H2O (g)+OO x+VOVV=2OHOV (5) of powder prepared at 85 °C and sintered at 1000 °C.Selected frequencies (in Hz) are shown for filled symbols. however, the detailed mechanism remains unclear. 90 J. Mater. Chem., 1999, 9, 83–9111 M. Kataoka, K. Suda, N. Ishizawa, F. Marumo, Y. Shimizugawa Conclusions and K. Ohsumi, J. Ceram. Soc. Jpn., 1994, 102, 213. 12 K. Uchino, E. Sadanaga and T. Hirose, J. Am. Ceram. Soc., 1989, Amorphous H2TiO3 is an excellent Ti-precursor to use in the 72, 1555.hydrothermal synthesis of BaTiO3. Phase pure powders with 13 F. J. Gotor, C. Real, M. J. Dianez and J. M. Criado, J. Solid State small particle sizes, ca. 40–80 nm and narrow particle size Chem., 1996, 123, 301. distributions can be prepared at 180 °C after 24 h. Although 14 S. Schlag and H. F. Eicke, Solid State Commun., 1994, 91, 883. initial reaction is very fast, ca. 90% yield after 8–10 h, extended 15 B. D. Begg, E. R. Vance and J. Nowotny, J. Am. Ceram. Soc., 1994, 77, 3186. heating at 180 °C is required to drive reactions to completion. 16 H. I. Hsiang and F. S Yen, J. Am. Ceram. Soc., 1996, 79, 1053. Lowering the reaction temperature produces powders with 17 R. Vivekanandan and T. R. N. Kutty, Powder Technol., 1989, even smaller particle sizes but very long reaction periods are 57, 181.required, >72 h, to ensure complete reaction. In addition, the 18 K. Uchino, N. Lee, T. Toba, N. Usuki, H. Aburatani and Y. Ito, powders are poorly crystalline and have high water content. J. Ceram. Soc. Jpn., 1992, 100, 1091. In order to obtain phase pure BaTiO3 ceramics it is import- 19 F. Yen, C. T. Chang and Y.Chang, J. Am. Ceram. Soc., 1970, 73, 3422. ant to control the Ba/Ti ratio. As the rate limiting step in the 20 J. Nowotny and M. Rekas, Solid State Ionics, 1991, 49, 135. synthesis involves reaction of the Ti-precursor, it is important 21 D. Hennings, M. Klee and R. Waser, Adv. Mater., 1991, 3, 334. to use a technique which can detect very small quantities of 22 M. Mori and T. Kineri, J.Am. Ceram. Soc., 1995, 78, 2391. unreacted amorphous or crystalline Ti-precursor material. We 23 A. I. Kingon, S. K. StreiVer, C. Basceri and S. R. Summerfelt, have clearly demonstrated that Raman spectroscopy, rather MRS Bull., 1996, 6, 46. 24 K. Wa. Gachigi, U. Kumar and J. P. Dougherty, Ferroelectrics, than XRD, is a simple and eVective technique for this purpose. 1993, 143, 229.Raman spectra of powders with an average particle size as 25 F. Chaput and J. P. Boilt, J. Am. Ceram. Soc., 1990, 73, 942. small as ca. 20–40 nm indicate asymmetry within the TiO6 26 M. Demartin, C. Herard, C. Carry and J. Lemaitre, J. Am. Ceram. octahedra of the BaTiO3 lattice. This contradicts the widely Soc., 1997, 80, 1079. cited ‘critical’ particle size theory for the stabilisation of the 27 W. Zhu, C. C. Wang, S. A. Akbar and R. Asiaie, J. Mater. Sci., cubic polymorph, at least for particle sizes in excess of ca. 1997, 32, 4303. 28 W. Zhu, C. C. Wang, S. A. Akbar, R. Asiaie, P. K. Dutta and 20–40 nm. Raman spectra of powders heat treated at ca. M. A. Alim, Jpn. J. Appl. Phys., 1996, 35, 6145. 1000 °C to remove adsorbed and structural water are very 29 A. Beauger, J. C. Mutin and J. C. Niepce, J. Mater. Sci., 1983, similar to that of as-prepared powders. This contradicts the 18, 3041. suggestion that lattice hydroxyl ions stabilise the cubic poly- 30 A. Beauger, J. C. Mutin and J. C. Niepce, J. Mater. Sci., 1983, morph of BaTiO3 prepared via a wet chemical technique, such 18, 3543. 31 J. C. Mutin and J. C. Niepce, J. Mater. Sci. Lett., 1984, 3, 591. as hydrothermal synthesis. Finally, as-prepared powders con- 32 A. Beauger, J. C. Mutin and J. C. Niepce, J. Mater. Sci., 1984, tain many defects, primarily lattice OH- ions. Preliminary 19, 195. conductivity results on pellets of powders which have been 33 R. Asiaie, W. Zhu, S. A. Akbar and P. K. Dutta, Chem. Mater., heated treated to remove lattice hydroxyl ions reveal these 1996, 8, 226. materials to be modest proton conductors at room 34 W. S. Clabaugh, E. M. Swiggard and R. Gilchrist, J. Res. Natl. temperature. Bur. Stand., 1956, 56, 289. 35 M. P. Pechini, US Pat., 3 330 697, 1967. 36 N. J. Ali and S. J. Milne, Trans. J. Br. Ceram. Soc., 1987, 86, 113. 37 M. H. Frey and D. A. Payne, Chem.Mater., 1995, 7, 123. Acknowledgements 38 A. D. Hilton and R. Frost, Key Eng. Mat., 1992, 66/67, 145. 39 M. M. Lencka and R. E. Riman, Chem. Mater., 1993, 5, 61. The authors would like to thank Dr Eric Lachowski for 40 W. Hertl, J. Am. Ceram. Soc., 1988, 71, 879. assistance with electron microscopy, Dr Alison Coats for 41 J. O. Eckert Jr., C. C. Hung-Houston, B. L. Gersten, EPMA, Dr Susan Blake for XRD, Mr James Marr for BET M. M. Lencka and R. E. Riman, J. Am. Ceram. Soc., 1996, 79, analysis and Professor Tony West for useful discussions. The 2929. EPSRC and AIST (Japan) are gratefully acknowledged for 42 C. H. Perry and D. B. Hall, Phys. Rev. Lett., 1965, 15, 700. financial support. 43 G. Busca, V. Buscaglia, M. Leoni and P. Nanni, Chem. Mater., 1994, 6, 955. 44 D. Hennings and S. Schreinemacher, J. Eur. Ceram. Soc., 1992, 9, 4. References 45 S. Wada, T. Suzuki and T. Noma, J. Ceram. Soc. Jpn., 1996, 104, 383. 1 B. JaVe, W. R. Cook and H. JaVe, Piezoelectric Ceramics, 46 R. Waser, J. Am. Ceram. Soc., 1988, 71, 58. Academic Press, London, 1971. 47 I. J. Clark, T. Takeuchi, N. Ohtori and D. C. Sinclair, unpub- 2 D. Hennings, Int. J. High Technology Ceramics, 1987, 3, 91. lished work. 3 G. Arlt, D. Hennings and G. de With, J. Appl. Phys., 1985, 58, 48 M. H. Frey and D. A. Payne, Phys. Rev. B, 1996, 54, 3158. 1619. 49 H. P. Abicht, D. Voltzke, A. Roder, R. Schneider and 4 M. H. Frey, Z. Xu, P. Han and D. A. Payne, Ferroelectrics, 1998, J. Woltersdorf, J. Mater. Chem., 1997, 7, 487. 206, 337. 50 N. Hirose and A. R.West, J. Am. Ceram. Soc., 1996, 79, 1633. 5 W. R. Bussen, L. E. Cross and A. K. Goswami, J. Am. Ceram. 51 I. J. Clark and D. C. Sinclair, unpublished work. Soc., 1966, 49, 33. 52 H. Iwahara, T. Esaka, H. Uchida and N. Maeda, Solid State 6 G. Arlt, Ferroelectrics, 1990, 104, 217. Ionics, 1981, 3/4, 359. 7 A. Morell and J. C. Niepce, J. Mater. Educ., 1991, 13, 173. 53 H. Uchida, H. Yoshikawa and H. Iwahara, Solid State Ionics, 8 T. Takeuchi, K. Ado, T. Asai, H. Kageyama, Y. Saito, 1989, 34, 103. C. Masquelier and O. Nakamura, J. Am. Ceram. Soc., 1994, 77, 54 K. D. Kruer, Chem. Mater., 1996, 8, 610. 1665. 9 W. Kanzig, Phys. Rev., 1955, 98, 549. 10 K. Kinoshita and A. Yamaji, J. Appl. Phys., 1976, 47, 371. Paper 8/05756G J. Mater. Chem., 1999, 9, 83–91 91

ISSN:0959-9428    

DOI:10.1039/a805756g

出版商:RSC    

年代:1999

数据来源: RSC