摘要:

Linearly extended hybrid tetrathiafulvalene analogues with bridged dithienylethylene p-conjugating spacers Hugues Brisset,a Soazig Le Moustarder,a Philippe Blanchard,a Bertrand Illien,a Ame�de�e Riou,a Jesus Orduna,b Javier Garinb and Jean Roncali*a aInge�nierie Mole�culaire et Mate�riaux Organiques, CNRS UMR 6501, Universite� d’Angers, 2 Bd L avoisier, 49045 Angers, France bL aboratorio de Quimica Organica Universidad de Zaragoza, Zaragoza, Spain New linearly extended p-electron donors based on bridged dithienylethylene (DTE) end-capped with diversely substituted 1,3- dithiol-2-ylidene electron-releasing groups have been synthesized by Wittig–Horner olefination of appropriate aldehydes.Cyclic voltammetry shows that whereas the unbridged analogues are directly oxidized to the dication state through a two-electron transfer, rigidification of the DTE spacer leads to the splitting of the two-electron wave into two successive one-electron steps due to the decrease of the potential corresponding to the formation of the cation radical.This unusual electrochemical behaviour is interpreted with the help of theoretical calculations which suggest that these eVects are related to the enhanced electron delocalization resulting from the rigidification of the DTE spacer.This conclusion is supported by an X-ray diVraction structural analysis which reveals that in addition to a fully planar conformation stabilized by intramolecular S,S interactions, the bridging of the DTE spacer leads to a significant reduction of bond length alternation.Linear p-conjugated systems end-capped with 1,3-dithiol-2- X-ray structural analyses which indicate that the bridging of the DTE spacer produces a decrease in bond length alternation ylidene groups have recently emerged as a growing class of resulting in an enhanced p-electron delocalization. organic compounds with specific electronic properties.1 Interest in these systems was initially motivated by the increased dimensionality and hence improved electrical properties anticipated for the corresponding cation-radical salts, in the general frame of the chemistry of extended tetrathiafulvalene (TTF) analogues.2 However, more recent work has shown that these hybrid p-conjugated systems are also potentially interesting as small-bandgap molecular semiconductors3 or as building blocks for push–pull or push–push molecules for quadratic or cubic non-linear optics.4,5 Since these various applications are closely related to the delocalization of p-electrons, the design of a p-conjugating spacer group showing optimal electron transmission properties associated with good thermal and photochemical stability appears to be a priority.In this context oligoheteroarylenevinylene spacers have recently emerged as a good trade-oV between the eYcient but unstable polyalkenic systems6 and the more stable poly(hetero) aromatic ones which present excessive p-electron confinement.7 We have already reported the synthesis of linearly extended TTF analogues (LETTFs) based on oligoheteroarylenevinylenes end-capped with 1,3-dithiol-2-ylidene units.8 Whereas for S S S S S S R R R R S S S S S S R R R R a R = CO2Me b R = SMe c R = Prn 4a–c 1a–c LETTFs involving oligothiophenes, internal rotational disorder leads to rapid saturation of the eVective conjugation beyond a bithiophenic spacer, i.e.eight conjugated carbons,1e such a saturation has not been observed yet even for the Results and Discussion longest known LETTFs built around a tetrathienylenevinylene spacer (22 conjugated carbons).9 The synthesis of compounds 1a–c is depicted in Scheme 1.Recently, we have shown that the bridging of the thiophene Bridged dithienyethylene 3 has been prepared by McMurry rings with the central ethylene linkage of dithienyethylene coupling of 4,5-dihydro-6H-cyclopenta[b]thiophen-6-one11 as (DTE) produces a 0.40 eV decrease of the bandgap of the already described.10,12 Dilithiation of 3 using n-butyllithium resulting electrogenerated polymer.10 As a further step we followed by reaction with DMF aVorded the dicarbaldehyde report here the synthesis of new LETTFs built by grafting the 2 in 81% yield.The target compounds 1a–c were then obtained 1,3-dithiol-2-ylidene moiety to both ends of a bridged DTE in 45–50% yield by double Wittig–Horner olefination of 2 spacer 1a–c.Comparison of the electrochemical properties of using an appropriately substituted phosphonium salt 5a or these new p-donors with those of their non bridged analogues phosphonate anions derived from the dithiolium salts 6b,c.13 4a–c shows that rigidification of the spacer leads to an increase The cyclic voltammograms (CV) of 1a and 1c in methylene of p-donor ability with a stabilisation of the cation radical chloride are shown in Fig. 1. The CV of 1a exhibits two reversible one-electron oxidation waves with anodic peak state. These results are discussed in the light of theoretical and J. Mater. Chem., 1997, 7(10), 2027–2032 2027Table 1 Electrochemical data for compounds 1 and 4,b 10-4 M in 0.1 M Bu4NPF6/CH2Cl2.Scan rate 100 mVs-1, all potentials in V vs. SCE compound Epa1/V Epa2/V Epa2-Epa1/V 1a 0.49 0.65 0.16 1b 0.33 0.45 0.11 1c 0.22 0.31 0.09 4ab 0.59a 0.66 0.07 4bb 0.43 — 0.00 4cb 0.29 — 0.00 aShoulder. bFrom ref. 8(a). for each compound 1, decoalescence and hence the increase of Epa2-Epa1 results from the negative shift of Epa1 while Epa2 values remain similar to those of the unbridged analogues.This result implies that the bridging of the spacer leads to a S S R R S S H, PF6 – R R H P+Bu3,BF4 – S OHC S CHO S S 5a R = CO2Me 6b R = SMe c R = Prn 1. BunLi, THF 2. DMF 1a–c 5a, Et3N, CH3CN or 6b,c, 1. P(OMe)3, NaI 2.BunLi, THF 2 3 + Scheme 1 stabilization of the cation radical state. Previous work on LETTFs containing polyalkenic or oligoheteroarylenelvinylenes p-conjugating spacers has shown that the lengthening of the spacer group produces a negative shift potentials Epa1 and Epa2 at 0.49 and 0.65 V corresponding to of Epa1 and Epa2 together with a decrease of their diVerthe successive generation of the cation radical and dication.ence.1a,1e,8 Thus, beyond a certain conjugation length which As expected, replacement of the electron-withdrawing depends on the structure of the spacer, direct formation of the CO2Me by the electron-releasing SMe and especially n-propyl dication state through a two-electron transfer occurs.1a,8,9 In groups leads to a negative shift of Epa1 and Epa2 while the this context, the behaviour of compounds 1 appears as rather potential diVerence Epa2-Epa1 decreases from 0.16 to 0.09 V, surprising since, as far as we know, it is the first time that a indicating a reduction of the on-site coulombic repulsion decrease in oxidation potential is associated with an increase between positive charges in the dication [Fig. 1(b) and of Epa2-Epa1 and hence a stabilization of the cation radical Table 1]. The comparison of the CV data of compounds 1 instead of the dication.with those of their analogues built around a single thiophene Previous work has shown that the bridging of DTE leads ring1b reveals a ca. 0.20 V negative shift of the peak potentials to a negative shift of Epa1 and Epa2 from 1.10 and 1.40 to 0.72 thus confirming that the insertion of the bridged DTE spacer and 1.16 V, respectively while Epa2-Epa1 increases from 0.30 leads to a significant improvement of the p-donor ability.to 0.44 V.10 This increase in the HOMO level leads to a However, comparison of the CV data of compounds 1a–c 0.40 eV reduction of the HOMO–LUMO gap (DE) of the with those of 4a–c containing a classical DTE spacer shows molecule and of the bandgap of the resulting polymer. The that the bridging of the spacer produces two noticeable great similarity between these eVects and those observed for changes.Firstly, Epa1 undergoes a 70–100 mV negative shift 1a–c clearly shows that the bridging of the DTE spacer is the which is indicative of enhanced p-donor ability. However, origin of their unusual electrochemical properties. The smaller whereas compounds 4b and 4c are dized into their magnitude of the eVects observed here can be related to the dication state through a single step two-electron transfer,9 fact that the 1,3-dithiole moieties provide the major contririgidification of the DTE spacer leads to the splitting of this bution to the HOMO level in LETTFs.1a,8 two-electron wave into two successive one-electron oxidation In a recent joint X-ray diVraction and theoretical analysis steps. A closer examination of the data in Table 1 shows that of bridged DTEs, we have shown that the reduction of DE induced by the bridging of DTE results from a decrease of bond length alternation.12 As widely acknowledged, this parameter represents the main cause for the existence of a finite bandgap in linearly p-conjugated systems.14 In order to analyse the role of bond length alternation in the electronic properties of compounds 1, the structure of a single crystal of 1c has been investigated by X-ray diVraction.The ORTEP view in Fig. 2 shows that the molecule is non-centrosymmetric and adopts a fully planar geometry except for the n-propyl chains that lie outside the plane of the molecule. While the planarity of the conjugated DTE spacer is ensured by the bridge, both lateral parts of the molecule are rigidified by strong 1,5 intramolecular S,S interactions between the sulfur atom of the thiophene ring and a sulfur of the 1,3-dithiole moiety.The S1,S3 and S4,S5 distances (3.154 and 3.171 A ° respectively) are larger than a covalent SMS bond (2.04 A ° ) but shorter than twice the van der Waals radius of sulfur (3.60 A ° ).Such interactions have already been observed for related compounds.1e,8,15 The crystal structure of 1c involves columns of molecules along the [001] crystallographic direction (Fig. 3). The molecules centrosymmetrically related are stacked in a head-totail way along the c axis. Molecules within a column are equidistant with an average separation of 3.95 A ° .The contact distances between molecules in neighbouring columns are too Fig. 1 CVs of (a) 1a and (b) 1c (10-4 M in 0.1 M Bu4NPF6/CH2Cl2). Scan rate 100 mVs-1. large to take any interaction into account. 2028 J. Mater. Chem., 1997, 7(10), 2027–2032Table 2 Bond distances in the DTE conjugated path (see formulae for bond labelling) S S a b' c' d' e' b c d e bond DTEa 7b 1c e 1.351(8) 1.350(1) 1.36(2) d 1.44(1) 1.409(8) 1.41(2) c 1.40(1) 1.356(8) 1.37(2) b 1.457(7) 1.440(7) 1.44(2) a 1.309(8) 1.335(6) 1.33(2) b¾ 1.447(7) 1.440(7) 1.43(1) c¾ 1.40(1) 1.356(8) 1.34(2) Fig. 2 ORTEP view of 1c d¾ 1.44(1) 1.409(8) 1.40(2) e¾ 1.351(8) 1.350(1) 1.38(1) dr/A ° 0.096 0.078 0.064 aFrom ref. 16. bFrom ref. 12. Comparison of the bond distances in the DTE moiety for DTE,16 4,4¾-dibutyl-6,6¾-bi(4,5-dihydro-6H-cyclopenta[b]thienylidene) 7, a substituted analogue of 3,12 and 1c shows that the presence of the bridge produces a lengthening of the a, e and e¾ double bonds and a compression of the b, d, b¾ and d¾ single bonds (Table 2).An important consequence of these bond length modifi- cations is a significant decrease of bond length alternation expressed as the diVerence between the average length of single and double bonds (dr).Thus, as shown in Table 2, dr decreases from 0.096 A ° for DTE to 0.075 A °for 712 and 0.064 A ° for 1c. The smaller dr value obtained for 1c might be related to the further constraint imposed on the DTE moiety by the 1,5 intramolecular S,S interactions. Although the large uncertainties related to the rather limited quality of the 1c crystal mean that these data should be considered with care, the trend expressed here confirms that the reduction of bond length alternation already observed for 7 persists when this system is used as p-conjugating spacers in LETTFs.These results allow us to propose a possible explanation for the peculiar electrochemical behaviour of compounds 1.Oxidation of 1 and 4 to the cation radical state induces geometrical changes in the conjugated path, the thiophene rings adopt a quinoid structure while the central ethylene linkage acquires the character of an essentially single bond (Scheme 2). In this context, the changes in bond distances induced by the bridging of the DTE moiety and in particular the compression of bonds b, d, b¾, d¾ and the lengthening of bonds a and e contribute to prefigure the final geometry of the cation radical.Consequently, formation of the cation radical becomes easier and thus requires less energy, in agreement with the observed decrease of Epa1. On the other hand, since there is much less diVerence between the geometries of 1·+ and 4·+ than between their respective neutral forms, the bridging of the DTE spacer has much less influence on the energy required by the second oxidation step, which is consistent with for the quasi invariance of Epa2. In order to gain more information on this question, theoretical calculations have been performed for the neutral, cation radical and dication forms of compounds 1b and 4b.The geometry of neutral and dicationic species have been optimized using the PM3/RHF method and that of the cation radical with the PM3/UHF method.Table 3 lists the computed values Fig. 3 Crystal structure of 1c of the ionization potential (Ei), dipole moment (m), heat of J. Mater. Chem., 1997, 7(10), 2027–2032 2029Fig. 4 AO coeYcients in HOMO of (a) 4b and (b) 1b a strong localization of the positive charge in one 1,3-dithiole unit.In contrast, there is practically no change in the m value between 1b and 1b·+ which suggests that the positive charge is delocalized over the entire molecule. This result which agrees well with the reduced dr value of 1b appears consistent with the observed stabilization of the cation radical as indicated by CV data. S S S S S S R R R R S S S S S S R R R R S S S S S S R R R R • –e– +e– –e– +e– + + + Scheme 2 Conclusion To summarize, new conjugated p-donors incorporating bridged Table 3 Computed values of ionization potential (Ei), dipole moment DTE spacers have been synthesized.The combined eVects of (m), heat of formation (DHf) and degree of bond length alternation the bridge in the spacer and of 1,5-intramolecular S,S inter- (dr) for compounds 4b and 1b and their cationic species actions lead to a fully planar rigid structure with reduced bond length alternation.The geometrical modifications induced by compound Ei/eV m/D DHf/kcal mol-1 dr/A ° the rigidification of the DTE spacer result in the splitting of the two-electron oxidation wave observed for the unbridged systems into two successive one-electron steps with a concomi- 1b 7.82 0.1 177.9 0.062 1b·+ 0.2 331.7 -0.030 tant decrease of the first oxidation potential.This rather 1b++ 0.0 -0.098 unusual behaviour is attributed to a stabilization of the cation 4b 7.94 0.1 180.7 0.063 radical state associated with an enhanced delocalization of the 4b·+ 24 339.4 -0.018 positive charge. 4b++ 0.2 -0.113 Experimental Electrochemical experiments were carried out with a PAR 273 formation (DHf). The dr values used refer only to the DTE Potentiostat-Galvanostat in a three-electrode singlesystem.compartment cell equipped with platinum microelectrodes The Ei value for 1b is found 0.12 eV lower than for 4b; this of 7.85×10-3 cm2 area, a platinum wire counter electrode and diVerence exactly matches that found betwen Epa1 values a saturated calomel reference electrode (SCE).Solutions were (Table1), but this agreement might be fortuitous. The geodeaerated by nitrogen bubbling prior to each experiment which metries of 1b and 4b are very symmetrical and present a quasi was run under a nitrogen atmosphere. inversion centre leading to a dipole moment close to zero. The computed geometries do not provide evidence for the eVect of X-Ray structural analyses the bridge and show only a slight diVerence in dr (0.062 and 0.063 A ° for 1b and 4b respectively). The fact that the noticeable Crystal data for 1c.C34H40S6 MW 641.08, monoclinic, P21/c, Z=4, a=16.618(7), b=17.850(11), c=11.005(9) A ° , b= diVerence in dr indicated by X-ray data is not found here may reflect a limitation of the computing method used.The ca. 92.25(5)°, V=3262(6) A ° 3, l=0.71069 A ° . 12 kJ mol-1 decrease of DHf observed for 1b suggests that, as expected, the presence of the bridge enhances the stability of Data collection. Data collection by the zig-zag v scan technique, 2°H25°, tmax=40 s, range h, k, l (h 0,13; k 0,21; the molecule. As shown in Fig. 4, whereas for 4b the highest coeYcients of the HOMO are located on the 1,3-dithiole units, l -19,19), intensity controls without appreciable decay (0.2%) gives 6202 reflections from which 1279 were independent for 1b the distribution is more homogeneous which suggests a better delocalization of p-electrons over the whole molecule.with I>3s(I). The data of the cation radical show that in both cases dr becomes negative, in agreement with the expected inversion of Structure refinement.After Lorentz and polarisation corrections the structure was solved with direct methods (SIR) which bond alternation (see Scheme 2). Oxidation of 4b into 4b·+ leads to a large increase in the dipole moment, consistent with reveal all the non-hydrogen atoms. After isotropic and aniso- 2030 J.Mater. Chem., 1997, 7(10), 2027–2032tropic refinement of all the C and S atoms respectively, the 6.62 (s, 2H), 6.68 (s, 2H). MS m/z 657 (M·+), HRMS calc. for coordinates of H atoms were determined using the HYDRO C26H24S10 655.9085, found 655.9064. UV–VIS (CH2Cl2) program. The whole structure was refined by the full-matrix lmax/nm (log e) 481(4.84), 516 (4.82) least-squares techniques {use of F magnitude; Uij for S atoms, x, y, z and B fixed for H; 191 variables and 1276 observations, weighting w=1/s(F0)2=4(F0)2/ [s(I)2+(0.04 F02)2]} with the 2,2¾-Bis(4,5-dipropyl-1,3-dithiol-2-ylidenemethyl )-6,6¾-bi(4,5- resulting R=0.061, Rw=0.060.dihydro-6H-cyclopenta[b]thienylidene) 1c. This compound was prepared using the same procedure from dithiolium salt 6c Theoretical calculations (0.90 g (2.72 mmol), trimethyl phosphite (0.32 ml, 2.72 mmol), KI, (0.41 g, 2.72 mmol) and dicarbaldehyde 2 (0.20 g, The semi-empirical PM3 method17 has been parametrized to 0.67 mmol).After the usual work-up the product was purified reproduce gas-phase properties, i.e. geometry, dipole moment, by column chromatography (silica gel, CH2Cl2). Yield (47%), ionization potential and heat of formation.The PM3 method red powder, mp 175–180 °C. dH (CDCl3) 1.39–1.82 (m, 24H), was used in the framework of the HYPERCHEM 5.0 pack- 2.21–2.71 (m, 8H), 6.88 (s, 2H), 7.29 (s, 2H). dC (CDCl3) 13.51, age.18 The geometry of 1b, 4b and of their dication has been 13.63, 22.65, 29.66, 30.54, 35.27, 106.18, 118.47, 122.96, 127.55, optimized at the RHF level of theory and the geometry of 128.78, 128,92, 131.86. MS m/z 641 (M·+100), 455 (17%), 320 1b·+ and 4b·+ has been computed at the UHF level.All (21%), 111(34%). HRMS calc. for C34H40S6 640.1454, found optimized geometries are almost planar and have root-mean- 640.1442. UV–VIS (CH2Cl2) lmax/nm (log e) 484(4.83), square gradient values lower than 0.1 kcal mol-1A ° -1 (1 cal= 520(4.83). 4.184 J). Ionization potential values were obtained through Koopmans’ theorem.19 6,6¾-Bi(4,5-dihydro-6H-cyclopenta[b]thienylidene) 3. Com- References pound 3 was synthesized according to a previously described procedure.10,12 1 (a) T. Sugimoto, H. Awaji, I. Sugimoto, Y. Misaki, T. Kawase, S. Yoneda and Z. Yoshida, Chem. Mater., 1989, 1, 535; (b) A. Benahmed-Gasmi, P.Fre` re, B. Garrigues, A. Gorgues, 2,2¾-Diformyl-6,6¾-Bi(4,5-dihydro-6H-cyclopenta[b]thienyli- M. Jubault, R. Carlier and F. Texier, T etrahedron L ett., 1992, 33, dene 2. In a round-bottomed flask equipped with a dropping 6457; (c) T. K. Hansen, M. V. Lakshmikantam, M. P. Cava, funnel and nitrogen inlet was added 3 (0.1 g, 0.41 mmol) in R. E. Niziurski-Mann, F. Jensen and J. Becher, J.Am. Chem. Soc., 20 ml of dry tetrahydrofuran (THF). The mixture was cooled 1992, 114, 5035; (d) J. Roncali, M. GiVard, P. Fre`re, M. Jubault to 0 °C and BuLi (1.6 M in hexanes) (0.54 ml, 0.86 mmol) was and A. Gorgues, J. Chem. Soc., Chem. Commun. 1993, 689; (e) added dropwise. After 15 min stirring at 0 °C, anhydrous J. Roncali, L. Rasmussen, C. Thobie-Gautier, P. Fre` re, H.Brisset, dimethylformamide (DMF) (0.20 ml, 2.34 mmol) was added; M. Salle�, J. Becher, O. Simonsen, T. K. Hansen, A. Benahmed- Gasmi, J. Orduna, J. Garin, M. Jubault and A. Gorgues, Adv. the mixture was allowed to warm to room temp. and stirred Mater., 1994, 6, 841. for 30 min. Water was then added and the mixture extracted 2 (a) M. R. Bryce, J. Mater. Chem., 1995, 5, 1481; (b) M.Adam and with diethyl ether. The organic phase was dried over CaCl2. K. Mu� llen, Adv. Mater., 1994, 6, 439; (c) Y. Misaki, N. Higuchi, Removal of the solvent and column chromatography of the H. Fujiwara, T. Yamabe, T. Mori, H. Mori and S. Tanaka, Angew. residue (silica gel, CH2Cl2) gave 0.10 g (81%) of a red powder, Chem., Int. Ed. Engl., 1995, 34, 1222. mp 198–200 °C, dH (CDCl3) 3.00–3.15 (m, 8H), 7.79 (s, 2H), 3 H.Brisset, C. Thobie-Gautier, M. Jubault, A. Gorgues and J. Roncali, J. Chem. Soc., Chem. Commun., 1994, 1765. 9.86 (s, 2H). n/cm-1 (KBr) 1650 (CNO).MS m/z 300 (M.+ 100). 4 (a) V. P. Rao, K.-Y. Jen, K. Y. Wong and K. J. Drost T etrahedron L ett., 1993, 34, 1747; (b) U. Scho� bert, J. Salbeck and J. Daub, Adv. 2,2¾ - Bis[4,5 - bis(methoxycarbonyl) - 1,3 - dithiol-2 - ylidene - Mater., 1992, 4, 41; (c) K.-Y.Jen, V. P. Rao, K. Y. Wong and methyl]-6,6¾-bi(4,5-dihydro-6H-cyclopenta[b]thienylidene) 1a. K. J. Drost, J. Chem. Soc., Chem. Commun., 1993, 90. To a stirred solution of phosphonium salt 5a (1 g, 2 mmol) 5 (a)M. Sylla, J. Zaremba, R. Chevalier, G. Rivoire, A. Khanous and A. Gorgues, Synth. Met., 1993, 59, 111; (b) T.T. Nguyen, M. Salle�, and dialdehyde 2 (0.1 g, 0.33 mmol) in 20 ml of CH3CN, B. Sahraoui, M. Sylla, J. P. Bourdin, G. Rivoire and J. Zaremba, triethylamine (0.32 ml, 2.33 mmol) was added dropwise. After J.Modern Opt., 1995, 42, 2095. 2 h stirring at room temp., the solvent was removed by 6 (a) I. Cabrera, O. AlthoV, H.-T. Man and H. N. Yoon, Adv.Mater., evaporation and the residue purified by column chromatogra- 1994, 6, 43; (b) M.Blanchard-Desce, C. Runser, A. Fort, phy (silica gel, light petroleum–ethyl acetate 951) to give 67 mg M. Barzoukas, J. M. Lehn, V. Bloy and V. Alain, Chem. Phys., (30%) of a dark green powder with a metallic lustre, mp 1995, 199, 253; (c) H. Gibson and J. Pochan,Macromolecules, 1991, >280 °C, dH (CDCl3) 2.83–3.45, (m, 8H), 3.87, (s, 6H), 3.89 (s, 24, 4834. 7 (a) S. Gilmour, R. A. Montgomery, S. R. Marder, L.-T. Cheng, K.- 6H), 6.59 (s, 2H), 6.72 (s, 2H). MS (EI) m/z 704 (M.+ 100). Y. Jen, Y. Cai, J. W. Perry and L. R. Dalton, Chem. Mater., 1994, n/cm-1 (KBr) 1705 and 1737 (CNO). UV–VIS (CH2Cl2) 6, 1603; (b) V. Hernandez, C. Castiglioni, M. Del Zopo and lmax nm (log e) 469 (3.99), 501 (3.97). G. Zerbi, Phys.Rev. B, 1994, 50, 9815. 8 (a) E. Elandaloussi, P. Fre`re, J. Roncali, P. Richomme, M. Jubault 2,2¾ - Bis[4,5 - bis(methylthio) - 1,3 - dithiol -2 -ylidenemethyl] - and A. Gorgues, Adv.Mater., 1995, 7, 390; (b) A. Benahmed-Gasmi, P. Fre` re, E. Elandaloussi, J. Roncali, J. Orduna, J. Garin, 6,6¾-bi(4,5-dihydro-6H-cyclopenta[b]thienylidene) 1b. In a M. Jubault, A. Riou and A.Gorgues, Chem. Mater., 1996, 8, 2291; round-bottomed flask equipped with a dropping funnel and (d) E. Elandaloussi, P. Fre`re, A. Benahmed-Gasmi, A. Riou, nitrogen inlet were introduced dithiolium salt 6b (0.28 g, A. Gorgues and J. Roncali, J.Mater. Chem., 1996, 6, 1859. 1 mmol), trimethyl phosphite (0.12 ml, 1 mmol) and KI (0.15 g, 9 E. Elandaloussi, P. Fre` re and J. Roncali, T etrahedron L ett. 1996, 1 mmol) in 3 ml of acetonitrile. After 2 h stirring at room 37, 6121. temp., evaporation of the solvent and excess of trimethyl 10 J. Roncali, Thobie-Gautier, E. Elandaloussi and P. Fre`re, J. Chem. phosphite left the phosphonate a. Dry THF (3 ml) and Soc., Chem. Commun., 1994, 2249. 11 D. W. H. MacDowell, T. B. Patrick, B. K. Frame and D. L. Ellison, dicarbaldehyde 2 (0.08 g, 0.25 mmol) were then added and the J. Org. Chem., 1967, 32, 1227. mixture cooled to 0 °C. n-Butyllithium (0.63 ml, 2 mmol) (1.6 M 12 (a) H. Brisset, P. Blanchard, B. Illien, A. Riou and J. Roncali, Chem. in hexanes) was added dropwise, and the mixture stirred for Commun., 1997, 569 (b) P. Blanchard, H. Brisset, B. Illien, A. Riou 2 h at room temp. Upon addition of methanol a red precipitate and J. Roncali, J. Org. Chem., 1997, 62, 2401. was formed which is filtered, washed with methanol–diethyl 13 (a) K. Akiba, K. Ishikawa and N. Inasoto, Bull. Chem. Soc. Jpn., ether and dried. Yield 0.07 g (45%), mp 261–263 °C. dH (CDCl3) 1978, 51, 2674; (b) T. K. Hansen, M. V. Lakshmikantham, M. P. Cava, R. M. Metzger and J. Becher, J. Org. Chem., 1991, 56, 2.46 (s, 6H), 2.48 (s, 6H), 2.90–3.10 (m, 4H), 3.20–3.35 (m, 4H), J. Mater. Chem., 1997, 7(10), 2027–2032 20312720; (c) A. J. Moore, M. R. Bryce, D. T. Ando and 17 (a) J. J. P. Stewart, J. Comput. Chem., 1989, 10, 209; (b) J. J. P. Stewart, J. Comput. Chem., 1989, 10, 221. M. B. Hursthouse, J. Chem. Soc., Chem. Commun., 1991, 320. 14 J. Roncali, Chem. Rev., 1997, 97, 173. 18 HYPERCHEM 5.0, Hypercube Inc.Waterloo, Canada 1996. 19 T. Koopmans, Physica, 1933, 1, 104. 15 T. K. Hansen, M. R. Bryce J. A. K. Howard and D. S. YYt, J. Org. Chem., 1994, 59, 5324. Paper 7/01463E; Received 3rd March, 1997 16 G. Ruban and D. Zobel, Acta Crystallogr., Sect. B, 1975, 31, 2632. 2032 J. Mater. Chem., 1997, 7(10), 2027–2032

ISSN:0959-9428    

DOI:10.1039/a701463e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Langmuir–Blodgett films of a tetrathiafulvalene derivative substituted with an azobenzene group Leonid M. Goldenberg,a,b† Martin R. Bryce,*‡a Stefan Wegener,a,b Michael C. Petty,*b John P. Cresswell,b Igor K. Lednev,c Ronald E. Hester*c and John N. Moorec aDepartment of Chemistry and Centre forMolecular Electronics, University of Durham, Durham, UK DH1 3L E bSchool of Engineering and Centre forMolecular Electronics, University of Durham, Durham, UK DH1 3L E cDepartment of Chemistry, University of York, Heslington, York, UK YO1 5DD The new tetrathiafulvalene derivative 1, functionalised with an azobenzene substituent, has been synthesised.Cyclic voltammetric and spectroelectrochemical studies in solution demonstrate the reversible formation of the radical cation of 1. UV–VIS spectroscopy suggests that there is a weak interaction between the TTF and azobenzene moieties in compound 1, and demonstrates that trans–cis isomerisation occurs upon photolysis of the azobenzene substituent.Semi-conducting LB films of 1 have been assembled without the need for added fatty acid: room temperature conductivity values of the films before and after doping with iodine vapour were srt=10-3–10-5 S cm-1 and 2×10-2–10-3 S cm-1, respectively.No change in the conductivity of the LB films was observed under irradiation. Tetrathiafulvalene (TTF) derivatives have attracted consider- layer at the air–water interface unless at least 50% mole ratio of a fatty acid was added,14 and a bis(EDT-TTF) derivative, able interest in recent years as a number of their crystalline cation-radical salts are molecular metals and superconduc- which did not require the addition of any fatty acid.15 The azobenzene group was incorporated into compound 1 tors.1–5 In order to achieve high conductivity in thin films,6 several amphiphilic analogues of TTF have been prepared and because azo derivatives are known to undergo photochemical cis–trans isomerisation in LB films and this reaction can their Langmuir–Blodgett (LB) films built up.These generally exhibit in-plane direct current (dc) room temperature conduc- be monitored electrochemically.16–19 Moreover, for pyridinium–TCNQ LB films, where the pyridinium moiety tivity values in the range srt=10-3–10-1 S cm-1 after formation of a mixed valence state by doping with iodine vapour,7,8 was substituted with an azobenzene derivative, Matsumoto et al.18 reported a 30% conductivity change upon photochem- although higher conductivities, srt=ca. 1 S cm-1, have been achieved with a few derivatives.9,10 ical isomerisation of the azo group in the LB film. A polypyrrole copolymer film, where polypyrrole was substituted with an In the present work we report on the properties of compound 1 in solution and in LB films.This compound is a novel non- azobenzene group, is reported to change its conductivity by up to 50% under illumination; however, the authors of this amphiphilic derivative of TTF. work concluded that the conductivity change was not triggered by isomerisation of the azo group.20 Experimental Synthesis Compound 1 was synthesised as follows.Equimolar quantities of 4-phenylazophenol (Aldrich) and triethylamine were dis- Based upon experimental and theoretical data for related solved in dichloromethane. After 10 min, a solution of tetrathi- TTF–C(O)R derivatives (R=OBu and NMe2)11 the electron- afulvalenecarbonyl chloride [TTF–C(O)Cl] (prepared from withdrawing ester group attached to the TTF ring in com- tetrathiafulvalene carboxylic acid,21 by a modification of the pound 1 should increase the polar nature of the TTF ring literature procedure)22 in dichloromethane was syringed into system, which presumably serves as the hydrophilic portion of the solution, and the mixture was stirred at room temp.for the molecule, while the aromatic rings replace the ‘traditional’ 12 h. The mixture was then acidified with 2 M aqueous HCl; hydrophobic alkyl chain(s).This present study is timely in the the organic layer was separated, washed with water and dried light of current interest in the formation of LB films of charge- (MgSO4). The solvent was evaporated in vacuo and the residue transfer materials which do not possess long alkyl chains. In was chromatographed on a neutral alumina column (eluent this context, we have recently reported the formation of dichloromethane) to aVord compound 1 as a purple solid in conductive LB films of three derivatives of ethylenedithio-TTF 51% yield, mp 148 °C (Found: C, 52.9; H, 3.0; N, 6.2.(EDT-TTF) bearing aromatic substituents (phenyl, pyridyl C19H12N2O2S4 requires: C, 53.2; H, 2.8; N, 6.5%); m/z (DCI) and pyridinium); for these compounds 25 mol% of fatty acid 429 (M++1); dH (200 MHz, CDCl3) 7.99 (2H, dd), 7.92 (2H, was needed to stabilise monolayer formation.12,13 Other dd), 7.63 (1H, s), 7.52 (3H, m), 7.33 (2H, dd) and 6.37 (2H, s).examples of non-amphiphilic TTF materials which form LB films are (PhCH2S)4TTF, which did not form a stable mono- Characterisation Cyclic voltammetry (CV) was performed using an EG&G † Visiting scientist from the Institute of Chemical Physics in PARC model 273 potentiostat with an Advanced Bryans XY Chernogolovka, Russian Academy of Sciences, Chernogolovka 142432, recorder. Pt mesh served as the counter electrode, a saturated Moscow Region, Russia. ‡ E-mail M.R.Bryce@durham.ac.uk calomel electrode (SCE) served as the reference electrode in J.Mater. Chem., 1997, 7(10), 2033–2037 2033HClO4 solution, and Ag wire as the quasi-reference electrode (Aldrich) in acetonitrile (ca. 10-5 M) (Aldrich, spectrometric grade) were contained in 10 mm cells and studied at 18 °C. A in acetonitrile solution. Potentials for solution CV were corrected to Ag/AgCl with the ferrocene/ferrocenium couple as xenon lamp (1 kW) with a glass bandpass filter centred at 350 nm was used for photolysis of the LB films.the internal reference (+0.35 V vs. Ag/AgCl). CV in solution was performed in 0.2 M Bu4NPF6–acetonitrile on a Pt disk working electrode (1.6 mm diameter, Bioanalytical System Inc.) Results and Discussion employing IR compensation. Bu4NPF6 (Fluka, electrochemical grade), HClO4 (Aldrich, ACS reagent), acetonitrile (Aldrich, Solution studies HPLC) and ultrapure water were used for preparation of the The cyclic voltammetry of compound 1 in acetonitrile solution electrolyte solutions.revealed two, reversible, one-electron waves, which are typical Spectroelectrochemistry was undertaken using a Perkinof TTF esters,11 at E1/2=+0.40 and +0.75 V, vs. Ag/AgCl. Elmer Lamda 19 spectrometer with a Ministat (Thomson These redox potentials are raised slightly relative to TTF Electrochem.Ltd, Newcastle upon Tyne, UK) using a 0.2 M under the same conditions (E1/2=0.34 and 0.71 V) by conju- Bu4NPF6–acetonitrile solution of compound 1. The spectrogation of the TTF system and the electron-withdrawing carelectrochemical cell was based on a 1 cm thick cuvette; Pt wire bonyl group.11 No reduction peak was observed between 0 was used as the counter electrode, while Ag wire (with a and -1.8 V in the cyclic voltammetric measurements where potential approximately equal to that of Ag/AgCl in this reduction of an azo group would be expected;16,17 this is solution) served as a quasi-reference.Thin layer electrodes consistent with a trans-azo group, which usually gives a very were constructed from indium tin oxide (ITO, sheet resistance broad, ill-defined reduction wave, whereas the cis-isomer gives 30 V per square, from Balzers) and glass with a ca. 100 mm a clearly observable peak.16 We have measured the spectroelecthick PTFE spacer. trochemistry of compound 1 in acetonitrile solution (Fig. 1). The Durham LB troughs were housed in a class 10 000 The appearance of a new absorption peak at lmax 430 nm microelectronics clean room and have been described prealong with a shoulder at lmax 580 nm, when the spectrum was viously.23 Compound 1 was spread on the surface of ultrapure obtained at +1.2 V, are consistent with the generation of the water (obtained by reverse osmosis, deionisation and ultraviolet cation radical of compound 1.27 (We note that the potentials sterilisation) from CH2Cl2 solutions (0.1 g l-1).The surface are shifted considerably with respect to the CV data reported pressure versus molecular area isotherm was measured at above, due to uncompensated resistance in the thin-layer 20±2 °C, pH=5.8±0.2 and a compression rate ca. spectroelectrochemical cell.) These spectroscopic changes were 4×10-3 nm2 molecule-1 s-1.The optimal dipping pressure reversible and the spectrum of compound 1 reverted to that was found to be 35 mN m-1. LB films were deposited onto of the neutral species measured initially when the potential glass slides, quartz, conducting ITO glass slides (sheet resistwas returned to 0 V. ance 300 V per square, from Balzers) and Au- and Ag-coated The optical absorption spectrum of compound 1 in acetoglass slides by the conventional vertical dipping technique. nitrile [Fig. 2(a)] is similar but not identical to a linear Unless specified otherwise, a dipping speed of 10 mm min-1 combination of the absorption spectra of TTF28 and transwas employed and the first monolayer was dipped on the azobenzene Fig. 2(d) (inset).This suggests that there is some upstroke when the slide was immersed in the subphase before interaction between the TTF and azobenzene groups, but the compression of the monolayer. To improve the hydrophilic small diVerences in comparison with those between azobenzene properties of ITO, the slides were pretreated with saturated and its disubstituted donor–acceptor derivatives29 indicates Na2Cr2O7–concentrated H2SO4 solution for approximately that this interaction is relatively weak.The diVerences between 10 s and carefully washed with ultrapure water.24 Substrates the spectra of compound 1 shown in Fig. 1 (0 V) and Fig. 2(a) with areas between 20 and 30 cm2 were used for LB film are due to absorption by the ITO thin layer electrode in the transfer. After LB film deposition, the slides were cut carefully former case.with a diamond tipped stylus to form several electrodes with Photolysis of 1 in acetonitrile resulted in changes in the contact areas between 0.1 and 0.5 cm2. Au electrodes with a absorption spectrum [Fig. 2(b,c)] which were similar to those gap of ca. 30 mm and interdigitated Au electrodes with a gap observed on photolysis of trans-azobenzene [Fig. 2(e,f,g)]. of 20 mm were used for electrochemical doping of the LB films. Photolysis of trans-azobenzene is known to result in trans–cis These electrodes were produced as previously described.25 photoisomerisation,30 and the changes in the spectra are Electrochemical doping during LB film transfer was achieved characteristic of this process. These changes were observed to on a KI subphase (0.1 M) with a current of 6 mA, using a persist for at least 1 h in the dark, indicating that cis-1, like similar procedure to that described earlier.25 Chemical doping of the LB films was carried out by exposure to iodine vapour for a given time in a sealed vessel.The dc conductivity data were obtained in air by a standard two-contact method using silver paste contacts.By varying the distance between the electrodes, it was established that the contact resistance was negligible. The conductivity values were calculated using a monolayer thickness of 1.5 nm (estimated from molecular modelling). The capacitance was measured by a Boonton Electronics model 72BD capacitance meter. The conductivity normal to the film surface was measured by using evaporated Au top contact dots (diameter 0.1 cm, slowly evaporated at rate of about 0.5–1.0 nm min-1) for films deposited on Au-coated glass slides. The Pockels electro-optic eVect was measured for monolayers deposited on Ag-coated glass using the technique of surface plasmon resonance at a wavelength of 633 nm.26 Optical absorption spectra of solutions were obtained using a Hitachi U-3000 spectrometer, and of LB films using a Perkin-Elmer Lamda 19 spectrophoto- Fig. 1 Optical absorption spectra of compound 1 in 0.2 M meter. A mercury lamp with a 314 nm interference filter was Bu4NPF6–acetonitrile solution using a thin layer electrode: measurements at (a) 0 and (b) +1.2 V vs. Ag wire used for photolysis in solution. Solutions of 1 or azobenzene 2034 J.Mater. Chem., 1997, 7(10), 2033–2037Fig. 2 Optical absorption spectra of compound 1 and azobenzene (inset) in acetonitrile solution. Spectra of 1 obtained: (a) before photolysis, (b) 3 and (c) 8 min after photolysis at 314 nm. Spectra of trans-azobenzene obtained: (d) before photolysis, (e) 1.5, ( f ) 4 and (g) 10 min after photolysis at 314 nm. Fig. 4 A model of an LB monolayer of compound 1; molecular geometry optimisation obtained using Chem3D for Macintosh cis-azobenzene,30 is relatively stable to thermal cis–trans iso- LB film characterisation merisation.Prolonged photolysis (>10 min) of the azobenzene The in-plane dc conductivities for as-deposited LB films of solution resulted in no further changes in the spectrum, indicatcompound 1 were in the range srt=10-3–10-5 S cm-1.After ing that Fig. 2(g) is the spectrum of the photostationary state doping with iodine vapour, the room temperature conductivity mixture. In contrast, prolonged photolysis of 1 resulted in value of each sample rose by ca. one order of magnitude to further changes in the spectrum for which the earlier isosbestic values of srt=2×10-2–10-3 S cm-1.For LB films with a top points were not maintained, indicating that additional photo- Au contact, ohmic current–voltage characteristics were chemical pathways are available to 1. observed. Normal to the film surface, conductivity values of srt=10-6 and 5×10-5 S cm-1 were measured for a 25-layer Monolayer behaviour of 1 on the air–water interface and LB LB film, before and after iodine doping, respectively.The film transfer capacitance of the same device was found to be 550±27 and The condensed pressure vs. area isotherm for compound 1 is 627±56 pF, for the as-deposited and doped films, respectively, shown in Fig. 3. This was reproducible and stable at the which correspond to permittivity values, er=2.9 and 3.3 (using deposition pressure and was not aVected by the time that the a film thickness of 1.5 nm per monolayer obtained from monolayer remained on the subphase before compression.31 molecular modelling studies).There was no evidence of collapse during compression of the It is well known that azo compounds undergo trans–cismonolayer up to the highest pressure measured (40 mN m-1). isomerisation under illumination, even as thin films, and that The extrapolated limiting area (to zero pressure) is 0.21 nm2 this can lead to a change in conductivity.16–19 However, molecule-1, which is slightly less than that expected for the illumination of a 15 layer LB film of 1 using either visible cross-sectional area of the molecule obtained from molecular light, a wide-range UV source, a sodium lamp or monochromodelling studies with geometry optimisation using Chem3D matic light at 320 nm, did not result in any detectable change for Macintosh.We have observed this previously in isotherms in the value of the lateral conductivity. This result suggests of TTF derivatives,32 and we suspect that this is simply due that for LB films of 1 either: (a) trans–cis-isomerisation does to a very slight solubility of the compound in the subphase.A not proceed due to steric hindrance in a compact LB film schematic representation of a possible close-packing arrangestructure, or (b) the isomerisation proceeds but this structural ment of molecules of 1 in the LB film structure is shown in change does not influence the conductivity of the film. Fig. 4 Fig. 4. LB films of compound 1 were built up by predominantly shows a geometry optimisation of an LB monolayer of 1 which Z-type deposition with a transfer ratio on the upstroke of indicates that the molecules are bent to aVord the experimen- 0.9±0.1.tally observed molecular area of 0.21 nm2. This tightly packed structure could imply that the first explanation is more plausible. We note that the LB films studied by Matsumoto et al.18 were assembled at a lower surface pressure than we used in the present work, and by using the horizontal touching technique.Both these experimental conditions may result in a less dense film structure which would allow isomerisation to proceed. However, the optical absorption spectra of LB films of compound 1 (Fig. 5) suggest that some trans–cis-isomerisation does occur upon photolysis: after 10 min irradiation there was a slight decrease in the intensity of the absorption peaks at 320 and 235 nm, and a slight increase in the absorption in the range 420–550 nm.These data are qualitatively similar to those of the solution spectra shown in Fig. 2, and, therefore, we favour explanation (b) above. We also attempted to increase the conductivity values of LB films of compound 1 by electrochemical oxidation either during or after LB film deposition.33 However, neither of these methods aVected the conductivity of the films, and a low Fig. 3 Pressure vs. area isotherm for compound 1 transfer ratio was observed in the former experiments. This J. Mater. Chem., 1997, 7(10), 2033–2037 2035Conclusions We have synthesised the new TTF derivative 1 and demonstrated that trans–cis isomerisation of the azobenzene substituent occurs upon photolysis.Semi-conducting LB films of 1 have been assembled without the need for added fatty acid: presumably, the TTF group (the polarity of which is increased by conjugation with the carbonyl substituent) is hydrophilic, and the azobenzene unit serves as the hydrophobic portion of the molecule, instead of the traditional alkyl chain(s).No change in the conductivity of the LB films was observed under irradiation. Further studies on LB films of non-amphiphilic TTF systems will be reported in due course. We are grateful to the EPSRC for an Advanced Fellowship to J.N.M. and for a research grant (S.W. and J.P.C.). L.M.G. Fig. 5 Optical absorption spectra for a six-layer LB film of compound thanks the Royal Society, the ERSPC, the Russian Foundation 1 on quartz: (a) before photolysis and (b) after photolysis for 10 min for Fundamental Research (project 97-03-32268a) and the University of Durham for financial support.References 1 J. R. Ferraro and J. M. Williams, Introduction to Synthetic Electrical Conductors, Academic Press, London, 1987. 2 M. R. Bryce, Chem. Soc. Rev., 1991, 20, 355. 3 A. E. Underhill, J.Mater. Chem., 1992, 2, 1. 4 J.M. Williams, J. R. Ferraro, R. J. Thorn, K. D. Carlson, U. Geiser, H. H. Wang, A. M. Kini and M-H. Whangbo, Organic Superconductors (including Fullerenes) Prentice Hall, New Jersey, 1992. 5 J. Mater. Chem., Special Issue on Molecular Conductors, 1995, 5, 1469. 6 For a review of electronically conductive LB films see: M.R. Bryce and M. C. Petty, Nature 1995, 374, 771. 7 J. Richard, M. Vandevyver, A. Barraud, J. P. Morand, R. Lapouyade, P. Delhaes, J. F. Jacquinot and M. Roulliay, J. Chem. Soc., Chem. Commun., 1988, 754. Fig. 6 Cyclic voltammogram for a five-layer LB film of compound 1 8 L. M. Goldenberg, R. Andreu, M. Saviro� n, A. J. Moore, J.Garý�n, deposited on an ITO electrode, 0.2 M HClO4, scan rate 50 mV s-1 M. R. Bryce and M. C. Petty, J.Mater. Chem., 1995, 5, 1593. 9 A. S. Dhindsa, Y. P. Song, J. P. Badyal, M. R. Bryce, Y. M. Lvov, M. C. Petty and J. Yarwood, Chem. Mater., 1992, 4, 724. Table 1 Pockels eVect measurements 10 L. M. Goldenberg, V. Yu. Khodorkovsky, J. Y. Becker, P. J. Lukes, M. R. Bryce, M. C.Petty and J. Yarwood, Chem. Mater., 1994, compound thickness/nm x(2) (-v;v,0)/pm V-1 r/pm V-1 6, 1426. 11 A. S. Batsanov, M. R. Bryce, J. N. Heaton, A. J. Moore, 1 1.2 5.0 1.6 P. J. Skabara, J. A. K. Howard, E. Ortý�, P. M. Viruela and R. Viruela, J.Mater. Chem., 1995, 5, 1689. 12 L. M. Goldenberg, J. Y. Becker, O. Paz-Tal Levi, V. Yu. Khodorkovsky, M. R. Bryce and M. C. Petty, J.Chem. Soc., Chem. Commun., 1995, 475. may be explained by two factors: (a) hindered anion diVusion 13 L. M. Goldenberg, J. Y. Becker, O. Paz-Tal Levi, V. Yu. within the multilayer assembly, which is quite compact as Khodorkovsky, L. M. Shapiro, M. R. Bryce, J. R. Cresswell and judged by the molecular area obtained from the isotherm; (b) M. C. Petty, J. Mater. Chem., 1997, 7, 901.instability of the films upon application of an electrochemical 14 Y. Xiao, Z. Yao and D. Jin, L angmuir, 1994, 10, 1848. potential. The cyclic voltammetric response of LB films of 1 15 R. P. Parg, J. D. Kilburn, M. C. Petty, C. Pearson and T. G. Ryan, J.Mater. Chem., 1995, 5, 1609. was measured and the best response was obtained for a five- 16 Z. F. Liu, B. H. Loo, K. Hashimoto and A.Fujishima, layer film (Fig. 6). However, the electroactivity disappeared J. Electroanal. Chem., 1991, 297, 133. after a few cycles, which is consistent with film desorption. We 17 Z. F. Liu, K. Hashimoto and A. Fujishima, Faraday Discuss., 1992, consider, therefore, that (a) explains the results of attempted 94, 221. post-deposition electrochemical oxidation, and (b) aVects the 18 H.Tochibana, T. Nakamura, M. Matsumoto, H. Komizu, E. electrochemical doping during film deposition. Mauda, H. Niino, A. Yabe and Y. Kawabata, J. Am. Chem. Soc., 1989, 111, 3080. The Pockels electro-optic eVect was measured for mono- 19 For a review of organic switches based on azobenzene derivatives, layers of compound 1. The surface plasmon resonance method see: F. Vo� gtle, Supramolecular Chemistry, Wiley, Chichester, 1991, also allowed an estimate of the film thickness to be made by ch. 7. assuming a value of the permittivity (er=2.5 was used in this 20 S-A. Chen and C. S. Liao, Synth.Met., 1993, 57, 4950. case). The results are given in Table 1. In each case the 21 J. Garý�n, J. Orduna, S. Uriel, A. J. Moore, M. R. Bryce, S.Wegener, nonlinear optical r coeYcient is relatively small.This is likely D. S. Yufit and J. A. K. Howard, Synthesis, 1994, 489. For an eYcient synthesis of TTF see: A. J. Moore and M. R. Bryce, to be due to poor alignment of the molecules of 1, as the Synthesis, 1997, 407. chromophores themselves should have large values of hyperpo- 22 C. A. Panetta, J. Baghdadchi and R. Metzger, Mol. Cryst. L iq. larisability.The thickness obtained for a monolayer of 1 Cryst., 1984, 107, 103. (1.2 nm) is consistent with the value for the length of the 23 C. A. Jones, M. C. Petty, G. G. Roberts, G. Davies, J. Yarwood, molecule obtained from molecular modelling studies (1.5 nm, N. M. RatcliVe and J. W. Barton, T hin Solid Films, 1987, 155, 187. see above) and suggests that the film is one molecule in 24 Y.Fu, J. Ouyang and A. B. P. Lever, J. Phys. Chem., 1993, 97, 13753. thickness. 2036 J. Mater. Chem., 1997, 7(10), 2033–203725 L. M. Goldenberg, C. Pearson, M. R. Bryce and M. C. Petty, 31 It has been noted that the structure of multilayer films of amphiphilic metal(dmit)2 charge-transfer salts (metal=Ni, Pd, Pt) J.Mater. Chem., 1996, 6, 699. 26 G. H. Cross, I. R. Girling, I. R. Peterson and N. A. Cade, depends upon the time that the floating film is left on the subphase surface before compression. S. K. Gupta, D. M. Taylor, Electrooptics L ett., 1986, 22, 1111. 27 J. B. Torrance, B. A. Scott, B. Welber, F. B. Kaufman and P. E. P. Dynarowicz, E. Barlow, C. E. A. Wainwright and A. E. Underhill, L angmuir, 1992, 8, 3057; C. Pearson, A. S. Dhindsa, Seiden, Phys. Rev. B, 1979, 19, 730. 28 R. Dieing, V. Morisson, A. J. Moore, L. M. Goldenberg, M. R. L. M. Goldenberg, R. A. Singh, R. Dieing, A. J. Moore, M. R. Bryce and M. C. Petty, J. Mater. Chem., 1995, 5, 1601. However, Bryce, J-M. Raoul, M. C. Petty, J. Garý�n,M. Saviro�n, I. K. Lednev, R. E. Hester and J. N. Moore, J. Chem. Soc., Perkin T rans. 2, this is not usually observed with TTF derivatives. 32 A. S. Dhindsa, J. P.Badyal, M. R. Bryce, M. C. Petty, A. J. Moore 1996, 1587. 29 J. GriYths, Colour and Constitution of Organic Molecules, and Y. M. Lvov, J. Chem. Soc., Chem. Commun., 1990, 970. 33 B. Tieke, Adv. Mater., 1991, 2, 222. Academic Press, London, 1976. 30 H. Rau, in Photochromism.Molecules and Systems, ed. H. Durr and H. Bouas-Laurent, Elsevier, Amsterdam, 1990, ch. 4, p.165. Paper 7/02797D; Received 24tJ. Mater. Chem., 1997, 7(10), 2033–2037 2037

ISSN:0959-9428    

DOI:10.1039/a702797d

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Chemically modified thin organic films supported on polished silica substrates David S. Boyle and John M. Winfield* Department of Chemistry, University of Glasgow, Glasgow, UK G12 8QQ Organic films, of thickness up to ca. 1000 nm, derived from polycyclic aromatic hydrocarbons containing 3–15 rings, have been grown on highly polished silica glass supports by vacuum sublimation. The quality of the films, as judged by their electronic spectra, is very dependent on achieving a near subnanometre surface finish on the silica.Exposure of the films to MoF6 at room temperature results in irreversible adsorption of the latter, as demonstrated by electronic spectroscopy. The fluorides WF6 and AsF5 show similar behaviour, although in many cases their interactions with the films are less marked.The reactions of organic molecules have been studied tradition- Annealing processes observed at higher substrate temperature can lead to greater alignment of molecules with the sub- ally in solution but there is increasing emphasis on selectivity in syntheses and by the use of an ordered, structurally well strate12–14 and interactions with a polar substrate such as cleaved single-crystal KCl can be important in determining characterized environment it is often possible to obtain more information than would be the case in solution. Our interest the structure of the film, a notable example being films derived from a polymeric m-fluoroaluminium(III ) phthalocyanine.15 in the chemistry involved in the chemical–mechanical polishing of silica glass and silicon,1 led us to investigate the use of these materials as supports for organic films that could then be Experimental modified by interaction with volatile inorganic reagents, silica supports being particularly useful for investigations of reactions Wafers of silica (2×1 cm; Spectrosil B, Multilab Ltd) or silicon between the films and high oxidation state fluorides by elec- (2×1 cm; (100) p-type, MCP Wafer Technology Ltd), chemotronic spectroscopy.mechanically polished (surface roughness RA1 nm over Polycyclic aromatic hydrocarbons are potential p donors 250 mm) using a procedure described elsewhere,1 were degreand an obvious way of modifying their electronic properties is ased and cleaned carefully by ultrasonic agitation for 5 min in via the formation of charge-transfer complexes with volatile 68% nitric acid (GPR, Rho�ne Poulenc), Genklene (ICI) and Lewis-acid halides adsorbed on the surface of the film.isopropyl alcohol (GPR, M & B). High-purity (Gold Label, Interactions between p-donor aromatic hydrocarbons and high Aldrich) organic compounds were employed without further oxidation state metal halides have been studied in solution for purification wherever possible for the preparation of thin films, many years.2 In contrast to n-donor molecules, where the otherwise starting materials were purified by recrystallisation.formation of insoluble adducts with acceptor molecules is well Owing to the hygroscopic nature of many of the materials documented,3 studies involving p-donor aromatic hydro- employed, reactions were conducted in vacuo (10-4 Torr) using carbons indicate that the interactions are very much weaker, a flamed-out Pyrex glass vacuum system.Molybdenum hexafor example in systems exhibiting ‘contact charge transfer’ fluoride, tungsten hexafluoride, phosphorus pentafluoride (all behaviour.4 Fluorochem) and arsenic pentafluoride (Matheson) were puri- In this initial survey study, fourteen polycyclic aromatic fied by repeated trap-to-trap vacuum distillation over activated compounds possessing 3–15 aromatic rings and bis(ethylenedi- NaF at 77 K.They were transferred finally to Pyrex breakseal thio)tetrathiafulvalene (BEDT-TTF) were investigated. vessels containing activated NaF and stored at 77 K until Although intrinsically related, there are important diVerences required.Instrumentation was as follows: electronic spectra between series of structurally comparable compounds, for PE Lambda 9, IR Nicolet 5DXC with a SpectraTech collector example the cata- and peri-condensed systems. In the former, for diVuse reflectance and an MTEC 100 cell for PAS, IR all the carbon atoms participate in a conjugated system in microscopy Nicolet Nic-Plan with ZnSe ATR accessory.contrast to the latter, which often possess individual discrete cycles of p-electron conjugation, with peri bonds having bond Preparation of organic thin films lengths approaching those of a C–C single bond.5 The eVect Since this was a survey study, films were synthesised using a of substituents on thin film formation has been examined in, straightforward thermal evaporation procedure in vacuo based for example, rubrene, which is a non-planar cata-condensed on previous work in this Department.9,12,15 The apparatus hydrocarbon by virtue of the four phenyl groups attached to consisted of an Edwards 306 coating unit, capable of 10-7 Torr the tetracene backbone, decacyclene, a non-planar peri-conwith a liquid-nitrogen cooled cold trap.A copper–constantan densed hydrocarbon and coronene tetracarboxylic acid. thermocouple was employed to measure the substrate tempera- The structures and optical properties of thin films derived ture. Cooling the silica support below ambient temperature from polycyclic hydrocarbons and their derivatives on various was not possible with the equipment used. A measured quantity supports have been widely reported.6–15 Structural information (ca. 10 mg) of material was placed in a cleaned molybdenum has been obtained usually by transmission electron or stainless-steel boat at a measured distance from the sub- microscopy9,10,12,15 or by electronic spectra recorded at low strate. The substrate was heated to 493 K at 10-5–10-6 Torr temperatures.6–8,11 Particularly relevant to the present work and allowed to equilibrate for 30 min.This procedure ensured are studies of tetracene and pentacene films supported on clean and dry substrates were obtained. The substrate was glass6–8,11 or carbon10 from which it has been concluded that allowed to cool to the required temperature, whereupon the amorphous structures are formed when films of these hydroorganic material was evaporated by heating resistively the carbons are formed by condensation of the vapour at low boat for 1 min.The substrate was allowed to cool to ambient temperature. Although long-range order is absent, short-range order, to the extent of a few lattice parameters, is present. temperature before being removed for spectroscopic analysis J.Mater. Chem., 1997, 7(10), 2039–2042 2039at room temperature. Film thicknesses were estimated annealing process being significant only above a temperature of half the mp of the compound being evaporated.12 The ratios, assuming a sticking coeYcient of unity, which has been demonstrated to yield values correct to within an order of substrate temperature5mp for the low-melting pyrene and fluoranthrene were somewhat larger than the optimal (Table 1) magnitude.16 owing to experimental limitations.FTIR spectra of the silica and silicon supported films, Preparation of chemically modified thin organic films although restricted in the information available using silica, A thin film supported on silica was loaded into an evacuable were indistinguishable from the spectra of the parent hydro- Pyrex cylindrical cell fitted with Spectrosil B windows and a carbons, strongly suggesting that no chemical change, for Pyrex/PTFE (J.Young) stopcock. Its electronic spectra was example decomposition or decarboxylation, occurred during then obtained. The cell was attached to a reaction manifold evaporation. Well resolved transmission electronic spectra which incorporated a breakseal vessel containing the fluoride which resembled closely those obtained from solutions, were and the entire system evacuated, pumped out for 24 h and obtained from silica-supported films derived from tetracene repeatedly flamed out thoroughly to remove residual moisture.and perylene over a range of thickness ca. 30–800 nm. All the The breakseal vessel was cracked open, the fluoride transferd bands expected in the 600–200 nm region of the spectra were by vacuum distillation to a second vessel containing activated resolved easily.However there was no evidence for solid-state NaF and vapour allowed to fill the cell and manifold. When splittings which might have been expected from previous the reaction cell containing the thin film was supplied with work,6,7,11 possibly because spectra were recorded only at suYcient fluoride to achieve a stoichiometry fluoride5 room temperature.The situation for ovalene films was similar film>151, the cell was isolated from the manifold and allowed except that thick films, ca. 900 nm, showed some loss of to equilibrate for ca. 30 s. Volatile material was removed and resolution particularly below 300 nm.Of the other cata-conthe cell pumped for 5 min to remove any volatile material densed hydrocarbon films examined, anthracene gave well remaining. The evacuated cell was removed from the line and resolved spectra from films up to ca. 100 nm thickness, thick the electronic spectrum of the chemically modified film pentacene films gave rather poorly resolved spectra at waverecorded at room temperature. lengths 400 nm, chrysene spectra were virtually unaVected by increasing the film thickness while the information obtained from rubrene films was limited by interference eVects.Similar Results practical limitations were apparent in the spectra of peri- Determination of experimental conditions for film growth hydrocarbon films.Reasonably resolved bands from violanthrene, decacyclene and fluoranthrene were obtained only The room-temperature electronic spectra of films deposited on from very thin films, thickness ca. 50 nm. Perylene tetracar- commercially polished silica wafers were characterised by very boxylic dianhydride produced well resolved spectra at all film broad absorption bands from which little information could thicknesses whereas for coronene tetracarboxylic dianhydride, be extracted, since the new bands formed on the subsequent spectra were satisfactory only for thin films.The situation was addition of a high-oxidation state fluoride could hardly be similar for coronene tetracarboxylic acid. discerned. However using wafers polished by the method Despite their limited nature, the electronic spectra enabled described previously,1 satisfactory electronic spectra were the experimental suitability of a supported film for further obtained in all cases with the exceptions of pyrene and to study to be assessed, the operational criterion being a compari- some extent, rubrene where the spectra, irrespective of film son with the appropriate solution electronic spectrum.thickness, were dominated by interference patterns. Films, up to an estimated thickness of ca. 1000 nm were stable in vacuo over long time periods apart from anthracene and fluor- Chemically modified supported hydrocarbon films anthrene films which fractured readily. In all cases films were annealled at the temperature at which the substrate was heated Exposure of films derived from the hydrocarbons listed in during evaporation, the most satisfactory temperatures being Table 1 to either boron trifluoride or phosphorus pentafluoride those approaching or slightly greater than, half the melting at room temperature resulted in no observable changes in point of the hydrocarbon, Table 1.The observations are con- electronic spectra.In contrast, spectral changes were substansistent with those made previously using this procedure, the tial in many cases on exposure to molybdenum or tungsten hexafluoride vapours or to arsenic pentafluoride under similar conditions. The most detailed studies were undertaken for Table 1 Optimum temperature range for thin film growth on polished tetracene, perylene and ovalene silica-supported films since silica glass high quality spectra could be obtained readily from a range of substrate temperature films of diVerent thicknesses, ca. 50–1000 nm. compound TS/K TS/mp anthracene 288–298 0.6 Modified tetracene, perylene and ovalene films. Admission of tetracene 333–353 0.54 MoF6 or AsF5 to tetracene films at room temperature led to pentacene 333–353 0.60 immediate colour changes, orange–yellow to blue–green with chrysene 293–303 0.56 rubrene 293–303 0.49 MoF6 and to emerald-green in the case of AsF5.No colour perylene 313–333 0.60 change was observed using WF6. The electronic spectra of the ovalene 333–353 0.46 three modified films, recorded after removal of any unchanged decacyclene 313–333 0.46 volatile materials by pumping at room temperature, are compyrene 288–298 0.68 pared with the unmodified film spectrum in Table 2.In all fluoranthene 288–298 0.76 cases high-energy bands due to tetracene were observed and PTCDAa 313–333 0.56 CTCDAb 333–353 0.48 in the WF6-treated film there was no evidence for any signifi- CTCAc 333–353 0.48 cant change throughout. In the spectrum of AsF5-treated violanthrone 333–353 0.49 tetracene additional weak bands at low energy, lmax>600 nm BEDT-TTFd 293–303 0.59 were apparent.The striking features of the spectrum obtained from MoF6-treated tetracene films were strong bands, lmax 350 aPTCDA=perylene tetracarboxylic dianhydride. bCTCDA=coronene and 670 nm, and its general appearance is reminiscent of the tetracarboxylic dianhydride. cCTCA=coronene tetracarboxylic acid.dBEDT-TTF=bis(ethylenedithio)tetrathiafulvalene. spectrum of tetracene in dimethyl sulfate after oxidation with 2040 J. Mater. Chem., 1997, 7(10), 2039–2042Table 2 Electronic spectra, lmax/nm (absorption coeYcient/cm-1),a of Table 4 Electronic spectra, lmax/nm (absorption coeYcient/cm-1),a of supported ovalene films before and after exposure to MoF6, WF6 supported tetracene films before and after exposure to MoF6, WF6 or AsF5 or AsF5 lmax/nm(a/cm-1) after exposure to lmax/nm(a/cm-1) after exposure to lmax/nm(a/cm-1) lmax/nm(a/cm-1) before exposure MoF6 WF6 AsF5 before exposure MoF6 WF6 AsF5 225(sh) 235(sh) 206(22.7) 205(sh) 213(45.9) 215(45.9) 255(sh) 260(sh) 285(sh) 285(sh) 225(sh) 225(sh) 230(sh) 225(sh) 275(11.6) 280(10.6) 275(11.7) 275(sh) 310(6.5) 350(49.5) 325(sh) 337(sh) 350(sh) 305(sh) 305(sh) 305(sh) 338(sh) 385(8.3) 385(13.2) 413(sh) 413(6.2) 350(26.8) 387(4.1) 387(4.7) 385(sh) 385(5.7) 438(31.2) 437(11.7) 475(sh) 462(sh) 465(6.9) 465(12.0) 412(sh) 412(5.9) 412(sh) 438(7.4) 438(7.2) 438(3.8) 505(30.3) 495(sh) 500(11.7) 525(sh) 530(10.5) 460(4.4) 472(1.1) 478(sh) 472(11.0) 472(3.4) 550(5.5) 550(10.2) 620(sh) 503(14.0) 503(13.9) 505(2.5) 525(7.3) 520(13.6) 525(2.4) 670(3.9) 750(2.6) 750(9.1) 605(sh) 625(0.8) 670(25.0) 675(1.0) 850(2.9) 962(3.6) 1000(0.9) 1000(9.1) 725(sh) 775(4.1) 765(1.8) 1010(3.1) aSee footnote to Table 2.aDefined by: absorbance=104a×film thickness(cm); sh=shoulder. Table 5 New, low-energy bands, lmax/nm (absorption coef- ficient/cm-1),a of some cata-condensed hydrocarbons after exposure to MoF6, WF6 or AsF5 Table 3 Electronic spectra, lmax/nm (absorption coeYcient/cm-1),a of supported perylene films before and after exposure to MoF6, WF6 new bands lmax/nm(a/cm-1) or AsF5 after exposure to (lmax/nm(a/cm-1) after exposure to hydrocarbon MoF6 WF6 AsF5 lmax/nm(a/cm-1) before exposure MoF6 WF6 AsF5 anthracene 920(0.5)b 460(2.3) 430(2.5) pentacene 785(4.9) —c —c 225(sh) 975(0.8) 260(13.8) 270(sh) 270(sh) chrysene 455(sh) 433(sh), 480(sh) 330(8.9) 330(sh) 533(sh) 500(1.4) 560(0.7) 340(8.4) 360(sh) 358(sh) 365(sh) 680(1.5) 680(1.1) 435(sh) 407(17.7) rubrene 615(10.1) —c —d 470(sh) 495(5.6) 479(13.3) 490(4.8) 505(sh) aSee footnote to Table 2.bLoss of resolution at higher energies. cNo 660(4.4) 623(13.3) 650(4.4) significant change. dNot examined. 1000(3.8) 1230(5.5) aSee footnote to Table 2. Table 6 New, low-energy bands, lmax/nm (absorption coef- ficient/cm-1),a of some peri-hydrocarbon derivatives after exposure to MoF6 SO3. The latter spectrum has been attributed to a mixture of mono- and di-positive tetracene cations.17 hydrocarbon derivative lmax/nm(a/cm-1) Perylene films darkened immediately on exposure to MoF6, perylene tetracarboxylic dianhydride 620(sh) WF6 or AsF5 at room temperature, the colour change being 770(4.4) most marked with MoF6 where the sequence golden– coronene tetracarboxylic dianhydride 500(14.5) yellow�red�crimson�violet�black was observed over 1 h.violanthroneb 725(6.3) The electronic spectra obtained after removal of volatile mate- 920(12.4) rial (Table 3) were all similar and, as in the tetracene–AsF5 decacyclene 770(3.2) case (Table 2) new bands at low energies were observed, lmax aSee footnote to Table 2.bExposure to AsF5 resulted in a new weak ca. 650 and 1000 nm with MoF6 and AsF5 and lmax band at lmax 690(3.7). ca. 623 nm with WF6. The behaviour of supported ovalene films was similar. Blue– green colouration of the orange films was immediately apparent a blue colouration on exposure to MoF6 and the spectra of modified films had two prominent new features, lmax=775 and on their exposure to MoF6 or AsF5 and although no change in colour was apparent when WF6 was admitted, the electronic 895 nm.Fluoranthrene films showed no change in their spectra after MoF6 treatment and those derived from coronene tetra- spectra in all cases showed several new bands at lmax>500 nm (Table 4).carboxylic acid (CTCA) lost spectral resolution. Other modified films. New spectral bands observed after Discussion modification of four cata-condensed hydrocarbon films and four films derived from peri-hydrocarbon derivatives are sum- The most striking finding from this work is the strong adsorption of molybdenum and tungsten hexafluorides and of arsenic marised in Tables 5 and 6.New low-energy bands after MoF6 treatment were observed in all cases, after AsF5 treatment in pentafluoride on many of the supported organic films as demonstrated by the appearance of new bands at low energies three cases but only for anthracene and chrysene was there any evidence for modification by WF6.Films derived from that may generally be described as being due to charge transfer between the organic molecules and the adsorbed fluorides. bis(ethylenedithio)tetrathiafulvalene (BEDT-TTF) developed J. Mater. Chem., 1997, 7(10), 2039–2042 2041Monocyclic p donors such as benzene or hexafluorobenzene, cases and by AsF5 in a substantial number. Adsorption of BF3 and PF5 on the films, if it occurs, does not result in any change interact only weakly with high oxidation state halides and in these latter cases isolable complexes are not formed.2 Apart in electronic properties of the hydrocarbons, while WF6 occupies an intermediate position.A more detailed consideration from the reaction between tetracene and MoF6, where the spectrum of the modified film (Table 2) provides some evidence requires structural information to be obtained on the precise nature of the surface layer but the importance of the oxidizing for formation of tetracene radical cations,17 there is no compelling evidence that electron transfer is complete.The presence ability of the fluoride is clear. of multiple charge-transfer transitions in the spectra of many of the modified films is not unexpected and has been reported We thank Dr.T. Baird and Dr. J. R. Fryer (both University for numerous solid complexes.3 of Glasgow) and Dr. R. Richardson (Glasgow Caledonian There have been many studies of charge-transfer complexes University) for practical assistance and provision of facilities in which attempts have been made to correlate the energy of and are grateful to EPSRC for financial support of this work. the charge-transfer absorption band(s) with intrinsic properties of the donor or acceptor species.18,19 However the results References obtained suggest that in the systems studied here the appearance or otherwise of charge-transfer absorption bands cannot 1 D.S. Boyle and J. M. Winfield, J.Mater.Chem., 1996, 6, 227. be correlated with any single property of a given molecule. 2 P. R. Hammond and R. R. Lake, J. Chem. Soc. A, 1971, 3800; 3806; This may reflect the diversity of polycyclic aromatic com- 3819; P. R. Hammond and W. S. McEwan, J. Chem. Soc. A, 1971, 3812; R. R. McLean, D. W. A. Sharp and J. M. Winfield, J. Chem. pounds investigated; diVerences in chemical behaviour and Soc., Dalton T rans., 1972, 676.spectroscopic characteristics between series of structurally 3 R. Foster, Organic Charge T ransfer Complexes, Academic Press, related polycyclic aromatic hydrocarbons (PAHs) have been London, 1969, and references therein. widely reported.20 Correlations within any given series, for 4 R. S. Mulliken, Recl. T rav. Chim. Pays-Bas., 1956, 75, 845.example cata-condensed PAHs, are expected and often found.21 5 E. Clar, Ber. Dtsch. Chem. Ges., 1936, 69, 607. For example, the linear acene series exhibits most clearly the 6 W. Hofberger, Phys. Status Solidi A, 1975, 30, 271. 7 R. Hesse, W. Hofberger and H. Ba�ssler, Chem. Phys., 1980, 49, 201. eVect of annelation, the three main absorption bands a, para 8 L. Sebastian, G.Weiser and H. Ba�ssler, Chem. Phys., 1981, 61, 125. and b, (characteristic to all PAHs) are progressively red-shifted 9 J. R. Fryer and D. J. Smith, Proc. R. Soc. L ondon Ser. A, 1982, with each additional aromatic ring, representing a dilution of 381, 225. the aromaticity; hence anthracene is colourless, tetracene is 10 R. Elermann, G. M. Parkinson, H. Ba�ssler and J. M. Thomas, golden–yellow and pentacene violet–blue. These compounds J.Phys. Chem., 1983, 87, 544. (D2h symmetry) provide a sample group for which it would be 11 R. Jankowiak, K. D. Rockwitz and H. Ba� ssler, J. Phys. Chem., 1983, 87, 552. expected that for a series of donor–acceptor complexes, the 12 J. R. Fryer, Mol. Cryst. L iq. Cryst., 1983, 96, 275. intermolecular repulsion and relative orientations of donor 13 M.Migita, Y. Taniguchi, H. Ishihara, M. Akagi, T. Ishiba and and acceptor molecules would be similar. However, no simple H. Tamura, J. Appl. Phys., 1985, 58, 1187. conclusions can be drawn from a comparison of the lowest 14 W. Dietsche, Th. Rapp and H. C. Basso, J. Appl. Phys., 1986, energy CT band observed after exposure to MoF6 with the 59, 1431.MO coeYcient xi19 or Ei22 of the donor PAH for the acene 15 J. R. Fryer and M. E. Kennedy, Macromolecules, 1988, 21, 259; J. R. Fryer, MSA Bull., 1994, 24, 521 and references therein. series anthracene, tetracene and pentacene. 16 L. Holland, Vacuum Deposition of T hin Films, Chapman and Hall, In principle CT bands might be correlated with the Lewis- London, 1956. acid strengths or oxidizing abilities of the binary fluorides.The 17 W. Ij. Aalbersberg, G. J. Hoijtink, E. L. Mackor and former appear not to be relevant, at least using gas-phase F- W. P.Weijland, J. Chem. Soc., 1959, 3049. aYnities as the indicator of Lewis acidity. Thermochemical 18 R. S. Mulliken and W. D. Person, Annu. Rev. Phys. Chem., 1962, estimates of gas-phase F- aYnities are in the order 13, 107. 19 M. J. S. Dewar and H. Rogers, J. Am. Chem. Soc., 1962, 84, 395; AsF5>PF5>BF3,23 however gas-phase F- transfer reactions A. R. Lepley and C. C. Thompson, Jr., J. Am. Chem. Soc., 1967, observed by ICR24 and F- ion displacement reactions in 89, 5523. MeCN25 both point to WF6 being a weaker Lewis acid than 20 C. A. Coulson, B. O’Leary and R. B. Mallion, Hu�ckel T heory for BF3.The existence of the [MF7]-, M=Mo and W, anions in Organic Chemists, Academic Press, London, 1978, and references equilibrium with the hexafluorides MF6 in MeCN at room therein; E. Clar, T etrahedron, 1959, 5, 98; 1959, 6, 355; 1960, 9, 202. temperature, suggests that WF6 and MoF6 have comparable 21 Z. H. Zaidi and B. N. Khanna, J. Chem. Phys., 1969, 50, 3291; Z. H. Khan and B.N. Khanna, J. Chem. Phys., 1973, 59, 3015; F- ion aYnities.26 It appears therefore that although Lewis E. Clar and W. Schmidt, T etrahedron, 1975, 31, 2263; Z. H. Khan, acidity may be a necessary criterion, it is not a suYcient Can. J. Spectrosc., 1978, 23, 8; 1984, 29, 63; Z. Naturforsch., T eil A, criterion for modification of PAH films observed here. 1984, 39, 668; 1987, 42, 91; Spectrochim.Acta, Part A., 1988, 44, The superior one-electron oxidizing ability of MoF6 com- 313; 1125. pared with WF6 is well established both in the gasmidt, J. Chem. Phys., 1977, 66, 828. and in solution, anhydrous HF28 or MeCN,29 and studies of 23 T. E. Mallouk, G. L. Rosenthal, G. Mu� ller, R. Brusasco and N. Bartlett, Inorg. Chem., 1984, 23, 3167.the behaviour of these and other fluorides with respect to the 24 P. M. George and J. L. Beauchamp, Chem. Phys., 1979, 36, 345. oxidative intercalation of graphite30–33 make an interesting 25 M. F. Ghorab and J. M. Winfield, J. Fluorine Chem., 1990, 49, 367. comparison with the present work. Bartlett and co-workers 26 M. F. Ghorab and J. M. Winfield, J. Fluorine Chem., 1993, 62, 101. have established that a thermodynamic threshold exists for the 27 N. Bartlett, Angew Chem., Int. Ed. Engl., 1968, 7, 433. intercalation of fluoroanions derived from binary fluorides, 28 A. M. Bond, I. Irvine and T. A. O’Donnell, Inorg. Chem., 1975, 14, MFx.30 The fluorides MoF6 and AsF5 are suYciently oxidizing 2408; 1977, 16, 841. 29 G. M. Anderson, J. Iqbal, D. W. A. Sharp, J. M. Winfield, for spontaneous intercalation to occur at room tempera- J. H. Cameron and A. G. McLeod, J. Fluorine Chem., 1984, 24, 303. ture,30–32 the estimated changes in DG2980 for 30 N. Bartlett and B. W. McQuillan, in Intercalation Compounds, ed. MoF6(g)+e-�[MoF6]-(g) and for 3 2 M. S. Whittingham and A. J. Jacobson, Academic Press, New York AsF5(g)+e-�[AsF6]-(g)+DAsF3(g) being similar. Tungsten and London, 1982, ch. 2, p. 19. hexafluoride and phosphorus pentafluoride are insuYciently 31 D. Vaknin, D. Davidov and H. Selig, J. Fluorine Chem., 1986, strong oxidizing agents to initiate spontaneous intercalation 32, 345. 32 M. Lerner, R. Hagiwara and N. Bartlett, J. Fluorine Chem., 1992, of [WF6]- or [PF6]-,30 although PF5 or BF3 do react with 57, 1. graphite in the presence of HF and with Cl2 as the oxidizing 33 F. Okino and N. Bartlett, J. Chem. Soc., Dalton T rans., 1993, 2081. agent.32 In the present work modification of the supported polycyclic hydrocarbon films by MoF6 has been demonstrated in all Paper 7/01893B; Received 18th March, 1997 2042 J. Mater. Chem., 1997, 7(10), 2039&ndas

ISSN:0959-9428    

DOI:10.1039/a701893b

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Microwave assisted preparation and sintering of Al2O3, ZrO2 and their composites from metalorganics Bhuvaragasamy G. Ravi, Peelamedu D. Ramesh, Navneet Gupta† and Kalya J. Rao* Solid State and Structural Chemistry Unit, Indian Institute of Science, Bangalore 560012, India Controlled pyrolysis of Al(OBus)3, Zr(OPrn)4 and their mixtures in ethyl acetate induced using microwaves of 2.45 GHz frequency has been carried out.Microwave irradiation yields second-stage precursors for the preparation of respective oxides and their composites. It is observed that the microwave irradiation has a directive influence on the morphology of the ultimate oxide products. Al2O3, ZrO2 and the two composites 90% Al2O3–10% ZrO2 and 90% ZrO2–10% Al2O3 are also found to be sintered to very high densities within 35 min of microwave irradiation by the use of b-SiC as a secondary susceptor.Use of microwaves for the synthesis and processing of ceramic preparation of the composites calculated quantities of the precursors were mixed in ethyl acetate and a uniform solution materials has attracted much attention recently. Microwave methods oVer several advantages such as unique synthetic was made.However for the preparation of pure Al2O3 and ZrO2, the precursors were used directly. Precursors were pathways, rapid heating rates, short processing durations, low power requirements, product uniformity, etc.1–9 We have placed in Borosil glass beakers and in the first stage of the experiment precursors were decomposed by microwave reported rapid and clean methods of synthesis of b-SiC10 AlN,11 MoO2,12 etc.using microwaves. We have also achieved irradiation in the oven for brief periods. Within a few minutes of heating (see later) we observed a rapid increase in the very rapid sintering of ZrO2–CeO2 ceramics using microwave irradiation.13 Metalorganics such as metal alkoxides, amides, viscosity of the precursors leading to formation of a foam-like product in the beakers.Further exposure to microwaves did imides, esters, etc. are generally employed as precursors for preparation of high purity ceramic powders. Use of alkoxides not cause any change in the foam-like products. There was no further increase in temperature or further decomposition. The for the preparation of oxides through sol–gel techniques is widely established.14 We have found that direct pyrolysis of oven was switched oV and the foam-like powder was collected from the beaker.metalorganics by conventional heating is unsuitable for powder production because it gives rise to non-uniform powder mor- Powders were characterized by X-ray diVraction studies (Philips X-ray DiVractometer, model PW 1050/70, Cu-Ka phology.Several other technological problems have been reported to be associated with complete combustion using radiation, l=1.5418 A ° ), thermogravimetry and diVerential thermal analysis (CAHN Instruments, California, USA, heat- conventional heating methods.15 However, when microwave irradiation is employed, very rapid and volumetric heating of ing rate 10 K min-1, in air) and particle size distribution (Micrometer Photosize, SKC 2000) (Table 1).metalorganic liquids is achieved. This can be expected to result in the formation of products of good structural and morpho- In the second stage, the powders were heated to 1673 K (only to 1473 K in the case of Al2O3) in air, with a heating logical uniformity. Microwave pyrolysis is known to be superior to conventional heating as it is self-regulating (micro- rate of 10 K min-1 in a conventional furnace (Thermolyne 46100).X-Ray diVraction studies were also performed on wave absorption characteristics change drastically once the susceptor is chemically altered) and therefore products having powders recovered from various temperatures. Transmission electron microscopy (JEOL-200CX TEM) was used to charac- uniform particle size distribution can be expected to form.Further microwave heating can be used with advantage in the terize the completely crystallized composites. The composite powders were ultrasonically dispersed in acetone medium. A preparation of ceramic composites because of the diVerential susceptibilities (and hence the heating rates) of the diVerent drop of the suspension was placed on the holey carbon film, supported on a copper grid and the bright field images precursors used in their preparations to generate unique morphologies.were recorded. In the third stage, the crystallized powders were subjected There is, however, very little known in the literature with regard to the use of microwaves in the preparation of ceramic to sintering again in the microwave oven by a method described earlier13 using b-SiC as a secondary heater.Sintered products powders and composites starting from organic precursors. In this paper we report our investigation on the unique influence were characterized using scanning electron microscopy (Cambridge Instruments, Stereoscan 360). of microwave irradiation of organic precursors during the preparation of ceramic composites of ZrO2 and Al2O3.We have also examined the sintering of Al2O3–ZrO2 composites by microwave irradiation and using a secondary heater. Results and Discussion Pure alkoxides Experimental We first consider the eVect of microwave irradiation of the The experimental set-up used for microwave heating is an metalorganics. Both Al(OBus)3 (ASB) and Zr(OPrn)4 (ZIP) ordinary kitchen microwave oven (Batliboi Eddy, India) appear to produce fumes when irradiated and become viscous operating at a frequency of 2.45 GHz and a maximum output liquids.This results in the formation of foam-like solids desigpower level of 980 W. We have chosen as organic precursors nated ASB-F and ZIP-F respectively. The products were Al(OBus)3 and Zr(OPrn)4 (Fluka AG, H-9470, Buchs).For the scraped from the beakers and the scraped materials were powdery (very fine particles) and are produced by irradiation after only a few minutes. Combined TG-DTA of the powders † Undergradute Research Scholar under linkages programme. J. Mater. Chem., 1997, 7(10), 2043–2048 2043Table 1 Microwave preparation conditions and the compounds formed particle starting material exposure durationa/min product colour size/mm Al(OBus)3 7+7+7+7 Al2O3 white 0.72 Zr(OPrn)4 7+7 ZrO2 yellowish white 0.41 Al(OBus)3+Zr(OPrn)4+MeCo2Et 7+7+7+7 90Al2O3–10ZrO2 white 0.75 +MeCo2Et Al(OBus)3+Zr(OPrn)4+MeCO2Et 7+7 10Al2O3–90ZrO2 yellowish white 0.73 aThe microwave power (560 W) was briefly interrupted every 7 min to examine the nature of the contents in the beaker. Fig. 1 Combined TG-DTA of microwave obtained powders of (a) Fig. 2 X-Ray diVractogram of ASB-F heated to various temperatures ASB-F and (b) ZIP-F (see text for definition of ASB-F and ZIP-F) (Cu-Ka radiation, l=1.5418 A ° ; #, a-Al2O3; *, c-Al2O3) is shown in Fig. 1. The loss of mass for ASB-F is gradual and most of the mass loss (25%) occurs below 500 K, corresponding to the first fairly large endotherm in the DTA.The remaining part of the mass loss occurs well below 873 K which corresponds to the region of a minor exotherm in DTA. The mass of the residual product is 65% of the original and remains constant up to 1473 K. The DTA suggests that there are two smaller exotherms in this region occurring at 1173 and 1273 K, respectively. For ZIP-F the first endotherm is rather shallow and extends to ca. 493 K and the mass loss is ca. 21%. The remaining mass loss occurs in small shallow steps and no further loss occurs beyond 1173 K. These step-like losses are associated with one large initial and two smaller subsequent exotherms. The first endotherm observed during heating of both ASBF and ZIP-F can be associated with the loss of water (H2O).The total mass loss for ASB-F (35%) is almost equal (34.6%) to the mass loss expected if ASB-F is Al(OH)3. The negligible diVerence may be due to residual organics which are burnt out in the region 473–673 K. However, for ZIP-F the exotherm in this region is large and the final yield is only 67%, which is much less than would be expected if ZIP-F corresponded to Zr(OH)4 (77.4%).Thus, microwave-irradiated material ZIP-F contains a significant proportion of undecomposed organics. It was noted in the IR spectrum (not shown) that there was a significant amount of residual ZIP in the ZIP-F. The spectra Fig. 3 X-Ray diVractogram of ZIP-F heated to various temperatures of both ASB-F and ZIP-F exhibited H2O related absorptions (Cu-Ka radiation, l=1.5418 A° ).(a) 1673 K, (b) 843 K, (c) 743 K, at 3300 and 1600 cm-1. (d) 623 K. We have examined XRD of the initial and heated powders of ASB-F and ZIP-F. Both are amorphous up to 623 K (Fig. 2 and 3). For ASB-F the XRD of the sample heated to 1273 K a-Al2O3 since samples cooled from 1473 K gave XRD typical of a-Al2O3 (Fig. 2) only. Generally, under conventional con- ( just above the second exotherm in DTA of Fig. 1) was found to be crystalline and could be indexed to a mixture of a- and ditions, the transition of c- to a-Al2O3 occurs at 1373 K.16 For ZIP-F, however, the major product was found to be cubic (c) c-Al2O3. A cubic unit cell (a=7.90 A° ) and a trigonal unit cell (a=b=4.758 A ° , c=12.991 A ° ) were used to index the c- and ZrO2 at 843 K [with monoclinic (m) ZrO2 as a minor phase].A cubic unit cell (a=5.09 A ° ) was used to index cubic ZrO2 a-Al2O3 structures respectively. The high-temperature small exotherm around 1373 K observed for alumina can be associ- while a monoclinic unit cell (a=5.146 A ° , b=5.213 A ° , c= 5.311 A ° , b=99.20) was used to index monoclinic ZrO2. ated with a transition of the small amount of c-Al2O3 to 2044 J.Mater. Chem., 1997, 7(10), 2043–2048Conventionally, under atmospheric conditions, ZrO2 the two Al2O3–ZrO2 composites examined were 90Al2O3–10ZrO2 (CA) and 90ZrO2–10Al2O3 (CZ) respectively. undergoes the following transformation sequences before it melts,17 We found that the time required to obtain foam-like powders (second-stage precursors) was 28 min for the alumina-rich mixture (CA-F) and 14 min for the zirconia-rich mixture monoclinic CA 1443 K tetragonal CA 2646 K cubic CA 2953 K liquid (CZ-F).The powders CA-F and CZ-F were then directly heated to A metastable cubic phase of ZrO2 was also observed at 673 K 1673 K and kept at that temperature for 2 h. The X-ray before the formation of stable monoclinic ZrO2 at 1273 K in diVraction patterns of the resulting oxide composites are shown crystallization studies of zirconia gels.18 in Fig. 4. It is evident that, upon heating, CA-F gives a The crystallization appears to start around 700 K, just at composite product which consists of tetragonal ZrO2 particles the beginning of the first of the twin peaked exotherms in the (a tetragonal unit cell with a=b=5.12 A ° and c=5.25 A ° was DTA (Fig. 1). In fact, the XRD at 743 K (middle of the twin used to index tetragonal ZrO2) in an a-Al2O3 matrix while exothermic peaks) reveals incomplete crystallization. Samples CZ-F gives an a-Al2O3 dispersion in monoclinic ZrO2. heated to 1673 K revealed that only monoclinic ZrO2 (Fig. 3) Fig. 5 shows transmission electron micrographs of CA and was present. We are therefore led to believe that the exotherm CZ composites.In the TEM of the CA composite, ZrO2 at 1173 K is associated with a cubic to monoclinic transformparticles (dark) are seen to be finely dispersed in a-Al2O3 ation of ZrO2 while the exotherm in the region 723–773 K matrix. The diameters of ZrO2 particles vary from 3 to 35 nm. could be largely due to burning (oxidation) of either strongly On the other hand in CZ composites only large crystallites of attached organic fragments or oxidation of any residual carbon ZrO2 were seen and Al2O3 particles could not be identified formed during the rapid initial combustion of organics in with certainty.The X-ray evidence from Fig. 4 however clearly insuYcient oxygen. suggests that formation of a ZrO2–Al2O3 solid solution does The above observation suggests that the alkoxides ASB and not occur.We do not consider the possibility that CA-F and ZIP do not decompose in a single step to their respective CZ-F are actually solid solutions which decompose to give oxides by microwave irradiation. The extent of residual rise to composites at higher temperatures because there are no organics in the amorphous powder products ASB-F and ZIPreports of such ready formation of solid solutions during F are also diVerent (higher in ZIP-F) at the stage where thermal decomposition of alkoxide mixtures.microwave interaction almost ceases. Both ASB-F and ZIP-F We therefore note that microwave heating of the alkoxides contain some water and their masses suggest that there is influences the nature of the final products. This is due to enough for formation of Al(OH)3 (AlOOH H2O?) or Zr(OH)4 diVerential microwave susceptibilities and to a certain extent [ZrO (OH)2 H2O? or ZrO2 2H2O?]. In the second stage of diVerences in quantities of undecomposed organics. It has been furnace heating ASB-F and ZIP-F are converted to oxides at observed by Yoshimatsu et al.19 that a fine distribution of higher temperatures. The various steps leading to the formation ZrO2 particles can result by decomposition of a zircanoalumin- of oxides are visualized as follows.ium compound under optimum conditions. The procedure of Yoshimatsu et al.19 and the present procedure can not be Al(OBus)3 CA microwaves [Al(OH)3(AlOOH H2O?)+ compared although the product morphologies are similar.The composite preparation is achieved here starting from homogeneous liquid mixtures of alkoxides and therefore a dehomo- little organic residue] CA 623 K amorphous Al2O3 genization step must be involved in the formation of the observed composite. This we attribute tentatively to the diVer- CA 1273 K [a-Al2O3+c-Al2O3] CA 1473 K a-Al2O3 ences in the microwave characteristics.Microwave sintering Zr(OPrn)4 CA microwaves [Zr(OH)4? (or ZrO (OH)2 H2O Another focus of this work has been to show that ceramics which are normally microwave inert at low temperatures such or ZrO2 2H2O ?)+organic residue] CA 623 K as Al2O3, ZrO2 and ZrO2–Al2O3 can still be sintered remarkably rapidly in microwaves by a technique which employs a amorphous ZrO2 CA 843 K [c-ZrO2+m-ZrO2] CA 1173 K m-ZrO2 Thus ASB-F and ZIP-F are actually second-stage precursors for the formation of oxides.Mixture of alkoxides We have also noted in our experiment that the time required to obtain the dry foam-like product ASB-F was 28 min compared to only 14 min for ZIP-F. We attribute this to the significant diVerence in the microwave susceptibilities of two alkoxides themselves since the decomposition products do not couple to microwaves eVectively.However the diVerence may also arise from diVerent thermal stabilities of the alkoxides. The presence of a high level of remnant organics in ZIP (which decomposes faster) suggests that thermal instability may not be the cause. On both counts we expect microwave decomposition of the alkoxide mixture to produce a mixture of fine amorphous powders since the microwave assisted decompo- Fig. 4 X-Ray diVractogram of 90Al2O3–10ZrO2 (CA) and sition rates are diVerent. The tendency of ZIP to be associated 90ZrO2–10Al2O3 (CZ) composites heat-treated at 1673 K for 2 h (Cuwith a greater proportion of remnant organics may also help Ka radiation, l=1.5418 A° ). A=Al2O3; Z=monoclinic ZrO2; ZT= tetragonal ZrO2.diVerentiation of ASB-F and ZIP-F phases. Compositions of J. Mater. Chem., 1997, 7(10), 2043–2048 2045Pt–Pt13%Rh thermocouple. The maximum temperatures of the pellets were usually attained after ca. 15 min of exposure and are given in Table 2. The densities of the pellets were also measured after 35 min of sintering (Table 2). Pure Al2O3 could be sintered to 95% of its theoretical density in under 35 min using microwaves: this, we believe, is extraordinary.Pure ZrO2 pellets sintered well but developed visible cracks upon cooling due to the well known tetragonal to monoclinic transformation. Pellets of Al2O3–ZrO2 composites also sintered well but to a maximum of only ca. 90% theoretical density. Alumina can be sintered to a high density when a small amount of MgO is added.20 Addition of 3% Y2O3 with zirconia resulted in tetragonal zirconia polycrystals (TZP) having very good mechanical properties.21 The eVect of adding MgO to Al2O3 and Y2O3 to ZrO2 on the microwave sintering was examined.Using sintering durations as above with the addition of 3% MgO as a sintering aid, Al2O3 was found to sinter to 97% of its theoretical density.ZrO2 also sintered to 98% of its theoretical density with the addition of 3% Y2O3. SEM micrographs of sintered Al2O3 (without and with 3% MgO), ZrO2 and the two composites (both with 3% Y2O3) are shown in Fig. 6 and the corresponding densities, sintering temperatures and other relevant data are given in Table 2. The flaky morphology of the sintered high density Al2O3 is similar in both cases, with or without the additives, and agrees well with the microstructures reported in the literature.9 The microstructure of ZrO2 consists of multifaceted particles with virtually complete elimination of enclosed porosity accounting for its high density.The SEM images of the sintered composites are quite similar to the morphology of the major components in the composite in their respective pure states.Particles of the two diVerent phases in the composites, however, are not so clearly distinguished. Fig. 5 TEMs of (a) CA composite (dispersion of ZrO2 particles is seen clearly in the Al2O3 matrix), and(b) CZ composite Conclusions secondary heater. In this ‘hitch-hiking’ approach low-temperature microwave inert materials are embedded in other strongly Two observations of this work are significant to ceramic microwave active materials (secondary heater) and subjected science and both are related to the use of microwaves. First to microwave irradiation.The critical requirement is that the microwave irradiation has a direct influence on the structure material of the secondary heater is chemically unreactive with of the initial powders obtained from metalorganic mixtures the material to be sintered.In the present case Al2O3, ZrO2 used as precursors in the preparation of composites. This and Al2O3–ZrO2 composite samples were first pelletized under eliminates the need to employ the rather slow and tedious 196 MN m-2 pressure to obtain 10 mm diameter pellets of sol–gel route to make composites starting from metalorganics.approximately 3 mm thickness using 1% poly (vinyl alcohol) Also, the metalorganics upon microwave irradiation yield as a binder. These pellets were surrounded by b-SiC powder second-stage precursors for the preparation of ceramic comin a silica crucible (ca. 10g of b-SiC is suYcient) and the posites. Secondly, the rate of sintering of composites in microcrucible was placed inside a microwave oven.In the present waves is fascinating in the novel ‘hitch-hiking’ heating process set-up 35 min was required for complete sintering of the pellets. where use is made of a secondary microwave susceptor. It is Details of the microwave sintering procedure with the use possible to achieve very high densities of products in extremely of the secondary heater has been described earlier.13 The short times.temperature of b-SiC used as a secondary heater reaches initially about 1073 K. The temperature of the pellet also rises, as a consequence, to nearly 1000 K. Around this temperature The authors acknowledge the help of Dr. G. N. Subbanna and Mr. Sam Philip for carrying out electron microscopy experi- the ceramic materials themselves couple well to microwaves.EVorts were made to monitor temperatures of the pellet surface ments. The Authors also thank the European Commission for financial assistance. by quickly interrupting the microwaves and inserting a Table 2 Data related to sinteringa of Al2O3, ZrO2 and their composites by microwave irradiation green sintered relative sample density/g cm-3 density/g cm-3 density(%) Al2O3 2.60 3.78 95 Al2O3+3%MgO 2.58 3.85 97 ZrO2 3.80 5.65 96 ZrO2+3%Y2O3 3.81 5.73 98 (90Al2O3–10ZrO2)+3% Y2O3 2.70 3.89 95 (90ZrO2–10Al2O3)+3%Y2O3 3.56 5.4 96 aMicrowave power=980 W, sintering duration=35 min, maximum temperature=1773 K. 2046 J. Mater. Chem., 1997, 7(10), 2043–2048Fig. 6 SEMs of freshly fractured surfaces of microwave sintered (a) Al2O3, (b) Al2O3 (3% MgO), (c) ZrO2 (3% Y2O3), (d) 90Al2O3–10ZrO2 and (e) 90ZrO2–10Al2O3 composites Materials, ed.W. H. Sutton, M. H. Brooks and I. J. Chabinsky, References Materials Research Society, Pittsburgh, PA, 1988, vol. 124, p. 235. 1 D. R. Baghurst and D. M. P. Mingos, J. Chem. Soc., Chem. 6 E. J. A. Pope, Am. Ceram. Soc. Bull., 1991, 70, 1777. Commun., 1988, 829. 7 T. T.Meek, C. E. Holcom and N. Dykes, J. Mater. Sci. L ett., 1987, 2 D. R. Baghurst, A. M. Chippindale and D. M. P. Mingos, Nature 6, 1060. (L ondon), 1988, 332, 311. 8 P. D. Ramesh and K. J. Rao, Bull.Mater. Sci., 1995, 18, 447. 3 N. H. Sutton, Am. Ceram. Soc. Bull., 1989, 68, 376. 9 M. Willert-Porada,MRS Bulletin, 1993, 18 51. 4 C. C. Landry and A. R. Barron, Science, 1993, 260, 1653. 10 P. D. Ramesh, B. Vaidhyanathan, M. Ganguli and K. J. Rao, J.Mater. Res., 1994, 9, 3025. 5 S. Komarneni, E. Breval and R. Roy, in Microwave Processing of J. Mater. Chem., 1997, 7(10), 2043–2048 204711 P. D. Ramesh and K. J. Rao, Adv.Mater., 1993, 7, 177. 18 V. S. Nagarajan and K. J. Rao, J. Mater. Sci., 1989, 24, 2140. 19 H. Yoshimatsu, Y. Miura, A. Osaka and H. Kawasaki, J. Mater. 12 B. Vaidhyanathan, M. Ganguli and K. J. Rao, J. Mater. Chem., 1996, 6, 391. Sci., 1990, 25, 961. 20 R. I. Taylor, J. P. Coad and A. E. Hughes, J. Am. Ceram. Soc., 1976, 13 P. D. Ramesh, P. Sarin, S. Jeevan and K. J. Rao, J. Mater. Synth. Processing, 1996, 4, 163. 59, 374. 21 D. J. Green, R. H. J. Hannink and M. V. Swain, T ransformation 14 C. J. Brinker and G. W. Scherer, Sol-Gel Science, the Physics and Chemistry of Sol-Gel Processing, Academic Press, New York, 1989. toughening of Ceramics, CRC Press Inc., Boca Raton, FL, 1989, p. 97. 15 D. W. Sproson, G. L. Messing and T. J. Gardner, Ceram. Int., 1986, 12, 3. 16 Encyclopedia ofMaterials Science and Engineering, ed. M. B. Bever, Pergaman Press, Oxford, 1986, p. 155. 17 E. C. Subbarao, H. S. Maiti and K. K. Srivastava, Phys. Status Paper 7/01976I; Received 21stMarch, 1997 Solidi A, 1974, 21, 9. 2048 J. Mater. Chem., 1997, 7(10), 2043–2048

ISSN:0959-9428    

DOI:10.1039/a701976i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Synthesis of titanium tetraalkoxides from hydrous titanium dioxide and dialkyl carbonates Eiichi Suzuki, Satoru Kusano, Hiroshi Hatayama, Masaki Okamoto and Yoshio Ono Department of Chemical Engineering, T okyo Institute of T echnology, Ookayama,Meguro-ku, T okyo 152, Japan The reaction of hydrous titanium dioxide (TiO2·nH2O, n=0.15–1.23) and dialkyl carbonates at 453–573 K oVers a convenient synthetic method for titanium tetraalkoxides which are free from chlorine-containing impurities.Thus, hydrous titanium dioxide was almost completely converted into Ti(OEt)4 by its reaction with diethyl carbonate at 493 K for 16 h. The reaction proceeded faster in the presence of a sodium hydroxide catalyst. From the hydrous titanium dioxide and dipropyl carbonate, Ti(OPrn)4 was obtained in a high yield.Metal alkoxides are important chemicals as starting materials heating under reduced pressure at 303 K to dryness and then at 353 K. The value of n did not change by this catalyst- for preparing ceramics by the sol–gel method. One of the methods to prepare metal alkoxides is the reaction of metal loading procedure. A 20 mmol portion of hydrous titanium dioxide (1.96 g for chloride with an alcohol or with sodium alkoxide.1 Metal oxides could be starting materials for chlorine-free n=1) and diethyl carbonate (24.4 cm3, 200 mmol) were introduced into a 120 cm3 autoclave.After the atmosphere was synthesis of metal alkoxides, if they were reactive enough towards appropriate organic compounds. The reactivity of a replaced with nitrogen, the autoclave was heated to the desired temperature at a heating rate of 90 K h-1 and reactions were metal oxide, silicon dioxide, has been demonstrated.Rosenheim et al.2 reported the transformation of silica into hexacoordi- conducted under autogenous pressure with stirring. After cooling, the reaction mixture was transferred into a round-bot- nated dianion complexes using pyrocatechol as a complexing agent under basic conditions.Laine et al.3,4 reported the tomed flask (capacity 50 cm3) in a nitrogen atmosphere, and titanium tetraethoxide was separated by distillation under formation of pentacoordinated silicates from silica, ethylene glycol and a base. Silicon alkoxides can be obtained by the reduced pressure. The product thus isolated was weighed to determine its yield.The product, titanium tetraethoxide, was reaction of silica with alcohols such as ethanol using potassium hydroxide as a catalyst under the conditions where water is identified by 1H NMR and IR absorption spectroscopy.† For the reactions of hydrous titanium dioxide with dipropyl continuously removed by azeotropic distillation.5 Ono and coworkers6–8 reported that silicon alkoxides can be easily carbonate, the same reaction and analytical procedures as those with diethyl carbonate were adopted, except that the synthesized by the reactions of silica gel with gaseous dialkyl carbonates in the presence of an alkali hydroxide [eqn. (1)].amounts of hydrous titanium dioxide and dipropyl carbonate charged into the autoclave were 15 mmol (1.47 g for n=1.0) SiO2+2ROC(NO)OR�Si(OR)4+2CO2 (R=Me, Et) and 23.2 cm3 (150 mmol), respectively.(1) In this work, we report the synthesis of titanium tetraalkox- Results and Discussion ides by the reaction of hydrous titanium dioxide (TiO2·nH2O, n=0.15–1.23) with dialkyl carbonates. The reaction is given Reaction of hydrous titanium oxide with diethyl carbonate by eqn. (2). (a) EVect of reaction temperature.The reaction of hydrous TiO2·nH2O+(2+n)R¾OC(NO)OR¾� titanium dioxide (TiO2·H2O) with diethyl carbonate was carried out at various temperatures for 16 h. Fig. 1 shows the Ti(OR¾)4+(2+n)CO2+2nR¾OH (R¾=Et, Prn) (2) yield of titanium tetraethoxide, Ti(OEt)4, as a function of This reaction oVers a simple method to synthesize chlorine- reaction temperature.At 433 K, Ti(OEt)4 was not obtained. free titanium tetraalkoxides. At 453 K, the yield was 77% and reached ca. 95% at 493 and 533 K. The yield slightly decreased at 573 K. Thus, in the temperature range 493–533 K, practically complete conversion Experimental of hydrous titanium dioxide to Ti(OEt)4 can be attained. Hydrous titanium dioxide was prepared as follows. To a Fig. 2 shows the change in the yield of Ti(OEt)4 with reaction 160 cm3 portion of an aqueous solution of titanium(IV) sulfate, time at 453 and 493 K.At 453 K, the presence of the induction with stirring, was added dropwise aqueous ammonia until the time was clear. The yield was almost negligible over the first pH reached 7.0. After the suspension of the precipitate was 4 h, and then sharply increased with reaction time to reach ca.kept stirring overnight at a constant pH of 7.0, the precipitate 60% at 10 h. The yield only slightly increased to 65% by was separated by filtration, washed with water four or five times, and then dried in an ambient atmosphere at 383 K for † The IR spectrum for the product obtained in this work is in 1–18 h. The hydrous titanium dioxide (TiO2·nH2O) thus agreement with that of the published data (T he Aldrich L ibrary of FT - IR Spectra Edition 1, vol. 1, ed. C. J. Pouchert, 1985), except that obtained had n in the range 0.15–1.23 depending on the drying several minor bands in the published spectrum are missing. Titanium time, as determined by thermal gravimetric analysis. tetraethoxide containing chlorine-containing impurities from a com- When a catalyst, for instance, sodium hydroxide, was used, mercial source, which had been produced from titanium tetrachloride, a 5 mass% portion (based on TiO2) of the catalyst was showed these minor bands together with the bands observed for the supported by an impregnation method: hydrous titanium product obtained in this work.This indicates that the bands which dioxide was immersed for 1 h in an aqueous solution of the are missing for the sample prepared in this work are probably attributed to chlorine-containing titanium species.catalyst (as small an amount of water as possible), followed by J. Mater. Chem., 1997, 7(10), 2049–2051 2049Fig. 3 EVect of n-value in TiO2·nH2O on the titanium tetraethoxide Fig. 1 Reaction-temperature dependence of the titanium tetraethoxide yield.Reaction conditions: TiO2·nH2O 1.65–2.04 g (20 mmol based on yield. Reaction conditions: TiO2·nH2O (n=1.0) 1.96 g (20 mmol based TiO2), diethyl carbonate 24.4 cm3 (200 mmol), reaction time 16 h, and on TiO2), diethyl carbonate 24.4 cm3 (200 mmol), and reaction time reaction temperature 453 K (#), 473 K (6) and 493 K (%). 16 h. Alkali-metal chlorides, NaCl and KCl, exhibited a retarding eVect, the yield being 72 and 71%, respectively. Sodium hydroxide is the most eVective catalyst.Fig. 2 also shows the time courses of the Ti(OEt)4 yield obtained for the reactions using the NaOH catalyst. Here the promoting eVect of NaOH is clearly seen; at 453 K, the induction time was shortened, and higher yields of Ti(OEt)4 were obtained.At 493 K, a very high yield of 96% was attained after only 6 h. (d) EVect of diethyl carbonate/hydrous titanium dioxide ratio. The eVect of the molar ratio of diethyl carbonate to hydrous titanium dioxide (TiO2·nH2O, n=0.94) on the Ti(OEt)4 yield was examined. The reactions were carried out at 503 K for 4 h Fig. 2 Time course of the titanium tetraethoxide yield in the presence and the results are shown in Fig. 4. Without a catalyst, a high and absence of sodium hydroxide catalyst. Reaction conditions: molar ratio of 10 was required to obtain a high yield of TiO2·nH2O (n=0.70) 1.85 g (20 mmol based on TiO2), diethyl carbon- Ti(OEt)4. Upon loading NaOH as a catalyst on hydrous ate 24.4 cm3 (200 mmol), and reaction temperature 453 K (#, $) and titanium dioxide, a molar ratio of 6 was suYcient to obtain a 493 K (%, &).Open and closed symbols represent the yield obtained high yield. for the reactions in the absence and presence of 5 mass% of a sodium Hydrous titanium dioxide is considered to consist of hydroxide catalyst, respectively. flocculates of small anatase crystals9 and the strre of hydrous titanium dioxide is schematically shown in Fig. 5.10 It contains a number of hydroxy groups as well as water further increasing the reaction time to 16 h. The presence of molecules coordinated to titanium ions. an induction time indicates that the reaction proceeds through Under the reaction conditions, the hydroxy groups on the an intermediate rate (or species) before the monomeric surface may react with diethyl carbonate to form ethoxy Ti(OEt)4 is finally formed.At 493 K, no induction period was observed. The yield monotonously increased with reaction time, and reached 95% at 16 h. (b) EVect of dehydration of hydrous titanium dioxide. The eVect of the extent of dehydration of hydrous titanium dioxide (TiO2·nH2O) on the yield of Ti(OEt)4 was examined. The extent of dehydration was varied by changing the time of drying the samples at 383 K, with n varying from 0.15 to 1.23.The reactions were carried out at 453, 473 and 493 K for 16 h. As shown in Fig. 3, at each reaction temperature, higher yields of Ti(OEt)4 were obtained with samples with higher n, i.e. with lower degree of dehydration. (c) EVect of the catalysts. In the reaction of silica gel with dialkyl carbonates, alkali-metal hydroxides and halides are eVective catalysts.6,7 Therefore, these compounds were tested as catalysts for the reaction of hydrous titanium dioxide with Fig. 4 EVect of molar ratio of diethyl carbonate (DEC) to hydrous diethyl carbonate. After the catalysts were supported by 5 titanium dioxide on the titanium tetraethoxide yield. Reaction con- mass%, the reactions were carried out at 493 K for 4 h.When ditions: TiO2·nH2O (n=0.94) 0.484 g (5 mmol based on TiO2), diethyl the catalyst was LiOH, NaOH, KOH and CsOH, the yield carbonate 2.4–6.1 cm3 (20–50 mmol), reaction time 4 h, and reaction of Ti(OEt)4 was 91, 93, 90 and 82%, respectively, while temperature 503 K. Open and closed symbols represent the yield the reaction using no catalyst gave a 79% yield.All the obtained for the reactions in the absence and presence of 5 mass% of a sodium hydroxide catalyst, respectively. alkali-metal hydroxides examined showed a promoting eVect. 2050 J. Mater. Chem., 1997, 7(10), 2049–2051Conclusion Titanium tetraethoxide (or tetrapropoxide) is prepared by the reaction of hydrous titanium dioxide and diethyl (or dipropyl) carbonate.The reaction proceeds almost to completion at 493–533 K after 16 h. Titanium tetraalkoxide is formed by the successive cleavage of TiMOMTi bonds in hydrous titanium dioxide by reaction with dialkyl carbonate molecules with formation of TiMOEt bonds and release of carbon dioxide. Use of a catalyst such as sodium hydroxide remarkably reduces the time required to accomplish the reaction.Reaction of the catalyst with hydrous titanium dioxide presumably occurs, giving rise to the cleavage of the TiMOMTi bond and simultaneously, the formation of a TiMOEt bond. The catalyst may facilitate the formation of an EtO group from a diethyl Fig. 5 Structure of hydrous titanium dioxide10 carbonate molecule, which attacks the TiMOMTi bond to form the TiMOEt bond.groups, ethanol and carbon dioxide [eqn. (3)]. TiMOH+EtOC(NO)OEt�TiMOEt+EtOH+CO2 (3) References The water molecules coordinated to titanium ions readily react 1 D. C. Bradley, R. C. Mehrotra and D. D. Gaur, Metal Alkoxides, with diethyl carbonate to give ethanol and carbon dioxide. Academic Press, London, 1978, p. 23. 2 A. Rosenheim, B. Raibmann and G. Schendel, Z. Anorg. Allg.H2O+EtOC(NO)OEt�2EtOH+CO2 (4) Chem., 1931, 196, 160. By these reactions, the hydroxy groups and the water molecules 3 R. M. Laine, K. Y. Blohowiak, T. R. Robinson, M. L. Hoppe, P. Nardi, J. Kampf and J. Uhm, Nature (L ondon), 1991, 353, 642. in hydrous titanium dioxide are scavenged as ethanol, and the 4 K. Y. Blohowiak, D. R. Treadwell, B. L. Mueller, M. L. Hoppe, reaction system is always kept free from water, which may S.Jouppi, P. Kansal, K. W. Chew, C. L. S. Scotto, F. Babonneau, promote the recondensation of the alkoxide product, Ti(OEt)4. J. Kampf and R. M. Laine, Chem. Mater., 1994, 6, 2177. 5 D. L. Bailey and F. M. O’Connor, US Pat., 2 881 198, 1959. Synthesis of titanium tetrapropoxide 6 E. Suzuki, M. Akiyama and Y. Ono, J. Chem. Soc., Chem. Commun., 1992, 136. Hydrous titanium dioxide (TiO2·nH2O, n=0.70) reacted with 7 Y. Ono, M. Akiyama and E. Suzuki, Chem. Mater., 1993, 5, 442. dipropyl carbonate to aVord titanium tetrapropoxide, 8 M. Akiyama, E. Suzuki and Y. Ono, Inorg. Chim. Acta, 1993, Ti(OPrn)4. The reaction at 493 K for 16 h without catalyst 207, 259. 9 W. F. Sullivan and S. S. Cole, J. Am. Ceram. Soc., 1959, 42, 127. gave an 84% yield of Ti(OPrn)4. When sodium hydroxide was 10 F. Inoue and M. Kaneko, Kogyo Kagaku Zasshi, 1971, 74, 591. used as a catalyst (5 mass%), reaction at 493 K for 6 h gave a yield of 92%. The product was identified by IR and 1H NMR spectra. Paper 6/08327G; Received 10th December, 1996 J. Mater. Chem., 1997, 7(10), 2049–2051

ISSN:0959-9428    

DOI:10.1039/a608327g

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Chemical routes to oxides: alkoxide vs. alkoxide–acetate routes: synthesis, characterization, reactivity and polycondensation of MNb2(OAc)2(OPri)10 (M=Mg, Cd, Pb) species Souad Boulmaa�z,a Rene�e Papiernik,a Liliane G. Hubert-Pfalzgraf *,a Bernard Septea and Jacqueline Vaissermannb aL aboratoire de ChimieMole�culaire, URA-CNRS, Universite� de Nice-Sophia Antipolis, 06108 Nice Ce�dex 2, France bL aboratoire de Chimie des Me� taux de T ransition, URA-CNRS, 75230 Paris Ce�dex, France The molecular constitution of solutions containing niobium alkoxides and divalent metal acetates M(OAc)2 (M=Mg, Ba, Pb, Zn, Cd) has been examined.The heterometallic aggregates MNb2(m-OAc)2(m-OPri)4(OPri)6 (M=Mg 1, Cd 2, Pb 3a) have been isolated and characterized by elemental analysis, FTIR, multinuclear NMR (1H, 13C, 207Pb, 113Cd) and by single-crystal X-ray diVraction for M=Mg and Cd.The magnesium derivative crystallizes in the monoclinic system, space group P21/c with the unitcell parameters a=21.238(5), b=10.127(10), c=24.861(3) A ° , b=107.77(2)° and Z=4. The thermal stability of the various species has been investigated. Condensation is induced thermally for PbNb2(OAc)2(OPri)10. The reactivity between MNb2(m-OAc)2- (m-OPri)4(OPri)6 (M=Pb, Mg) and other metallic species has been evaluated.A terheterometallic compound, PbMgNb2O(OAc)2(OPri)10 has been isolated. Whereas no reaction is observed between Ba(OAc)2 and Nb(OPri)5, the reaction between metal alkoxides aVords BaNb2(OPri)12(PriOH)2. Its reactivity shows that the absence of an assembling acetate ligand induces facile separation between the metals.Results of the hydrolysis experiments of 1–3a are given. The powders have been analyzed by TG, SEM–EDX, light scattering and XRD. The merit of the assembling acetate ligand for avoiding the segregation of the metals is emphasized. Mixed-metal oxides represent an important class of advanced reactions has been achieved.Results on the Nb–Cd system have been reported in a communication.5 materials due to their palette of technological applications.1 Their uses include catalysts and structural ceramics, but they have mostly been acclaimed for their applications as sensors, actuators and smart materials.2 Techniques for the formation Experimental of such materials, including chemical vapor deposition (CVD), All manipulations were routinely performed under nitrogen sol–gel processes, metal–organic decomposition and molecular atmosphere using Schlenk tubes and vacuum line techniques beam epitaxy, to name a few, require metal–organic molecules with dried and distilled solvents.Niobium6 and lead alkoxides,7 which have specific physical and chemical properties, as BaNb2(OPri)12(PriOH)2,8 lead9 and zinc10 trimethylsilylam- precursors.The preparation of inorganic materials from ides were prepared according to the literature. Anhydrous metal–organic precursors generally has the advantages over metal acetates were obtained by refluxing metal acetate ‘traditional routes’ of low temperatures of formation and/or hydrates with acetic anhydride over 15 h. 1H, 13C, 113Cd and crystallization, better compositional uniformity and conformal 207Pb NMR spectra were recorded on solutions (concen- coverage in the case of films.3 Metal alkoxides M(OR)n or trations: 1 M for 113Cd and 0.3 M for 207Pb) on a Bruker AC- oxoalkoxides MO(OR)n are versatile molecular precursors of 200 spectrometer. The chemical shifts are reported with respect metal oxides, one of their most attractive features being their to M(NO3)2 (M=Cd, Pb) in aqueous solutions as external solubility in a large variety of solvents and their ability to references, they are positive to low fields.IR spectra were form heterometallic species, especially by mixing alkoxides of registered with a FT–IRS 45 Bruker spectrometer as Nujol diVerent metals.4 However, to overcome the diYculty in handmulls between KBr plates for the air-sensitive derivatives and ling alkoxides and/or in their availability, more commonly as KBr pellets for the powders resulting from hydrolysis. accessible derivatives such as carboxylates (often acetates), Analytical data were obtained from the Centre de nitrates, halides, hydroxides or b-diketonates have often been Microanalyses du CNRS.IR and NMR data are listed in used in conjunction with metal alkoxides. The molecular Table 1. constitution of such solutions is almost unknown, although a Hydrolyses were achieved at room temperature in THF or better understanding could allow a better control of the isopropyl alcohol (0.1–0.05 M) without additives, the water hydrolysis–polymerization process and thus of the properties being added via these solvents.The powders were characterized of the resulting material. by TG–DTA with a Setaram system (nitrogen, heating rate of We report here the results of the investigations of the 5 °C min-1). Powder X-ray diVraction patterns were collected M(OAc)2–Nb(OR)5 (M=Mg, Ba, Zn, Cd, Pb) systems. Insight using Cu-Ka radiation after calcination at various into the molecular composition of homogeneous solutions has temperatures.been gained by using a variety of spectroscopic techniques, IR and nuclear magnetic resonance (1H and 13C as well as metal NMR, namely 207Pb or 113Cd), and in the most favorable cases Synthesis of MgNb2(OAc)2(OPri)10 1 by single-crystal X-ray diVraction. The heterometallic species MNb2(m-OAc)2(m-OR)4(OR)6 (M=Mg, Cd; R=Pri) have thus Anhydrous Mg(OAc)2 (1 g, 7.02 mmol) was added to a solution of Nb(OPri)5 (5.02 g, 12.94 mmol) in 25 ml of hexane (or been unequivocally characterized.Their reactivity as well as that of a Ba–Nb species having a similar stoichiometry, toluene) at room temperature. After stirring for 1 to 2 h, the excess of Mg(OAc)2 was eliminated by filtration. Crystallization BaNb2(OPri)12(PriOH)2 has been examined.The transformation of the acetatoalkoxides by hydrolysis–polycondensation was achieved at 5 °C (5.88 g, 99%). Anal. Calc. for J. Mater. Chem., 1997, 7(10), 2053–2061 2053Table 1 IR and NMR (1H, 13C, 207Pb) spectroscopic data for mixed-metal acetatoalkoxides NMR (25 °C, d CDCl3)a IR/cm-1 207Pb, 1H 13C{1H} nMOZ 113Cd{1H} compound nCO2 (Z=Ac, Et, Pri) (toluene) CH3(Ac) CH2(Et), CH (Pr) CH3 (Et, Pri) CO(Ac) CH2(Et), CH (Pri) CH3( Et, Pri, Ac) MgNb2(OAc)2(OPri)10 1 1590vs, 1429vs 662m, 616w, 581m, — 1.90 (s,3H) 4.75, 4.67, 4.54 1.33, 1.30, 1.20 177.4 75.4, 75.3 25.8, 25.4, 564m, 529m, 478s, (spt, 25251, 5H) (d, 25251, 30H) 71.8, 70.6 24.6, 24.2 391m CdNb2(OAc)2(OPri)10 2 1574vs, 1423s 579s, 557(sh), 518w, 33.8 1.95 (s,3H) 4.70 (spt, 5H) 1.28 (d, 30H) 177.6 74.1 25.4, 24.7 446s PbNb2(OAc)2(OPri)10 3a 1560vs, 1415vs 659m, 644m, 608(sh), 2390 25 °C: 1.97(s,3H) 25 °C: 4.99, 4.69 25 °C: 1.27 (d,30H) 571vs, 512w, 466m (spt, 352, 5H) 177.8 74.7, 73.9 25.8, 25.3, 432m -50 °C: 1.95(s) -50 °C 4.95, 4.83, 4.65 -50 °C: 1.22 (ov.d) 24.6, 23.8 (spt, 25251) PbNb2(OAc)2(OEt)10 3b 1566vs, 1414m 662s, 555s, 403m 2325 1.75(s,3H) 4.50, 4.30 1.21 (ov.t, 15H) 178.6 67.2, 62.9 23.7, 20.0, (q, 253, 10H) 18.2, 18.1 as=Singlet, d=doublet, q=quartet, spt=septet, m=multiplet, ov.d=overlapping of doublets, ov.t=overlapping of triplets.Coupling constants J=7 Hz (Et), 6 Hz (Pri). 2054 J. Mater. Chem., 1997, 7(10), 2053–2061C34H76MgNb2O14: C, 44.43; H, 8.27; Mg, 2.64; Nb, 20.23.CH3, OAc), 1.35, 1.30, 1.20 (d, J=6.1 Hz, 45452, 60H, CH3); 13C{1H} NMR (CDCl3), d 177.5 (CO, Ac), 75.7, 75.5, 72.9, Found: C, 43.53; H, 8.12; Mg, 2.35; Nb, 20.30%. 71.9, 70.7 (CH), 26.7, 25.5, 25.2, 24.7, 24.3 (CH3, Pri). 207Pb NMR (C7D8), d 4323. Synthesis of CdNb2(OAc)2(OPri)10 2 and of PbNb2(OAc)2(OR)10 (R=Pri 3a, Et 3b) Structure determination of MgNb2(OAc)2(OPri )10·0.5C7H8 Colourless needles were obtained by the same procedure (93%) The selected crystal was mounted on a Enraf-Nonius CAD-4 for 2 and for 3a (99%).Anal. Calc. for C34H76CdNb2O14: automatic diVractometer. The unit-cell parameters and basic C, 39.23; H, 7.45; Cd, 10.76; Nb, 18.92. Found: C, 38.99; H, 7.17; information about data collection at -100 &cture Cd, 10.82; Nb, 18.74.Anal. Calc. for C34H76Nb2O14Pb: C, 37.05; refinement are given in Table 2. Lattice parameters and orien- H, 6.90; Pb, 18.81; Nb, 16.87. Found: C, 36.99; H, 6.87; tation matrices were obtained from least-squares refinement of Pb, 18.79; Nb, 16.85%. the setting angles of 25 well centred reflections in the range Unit-cell parameters for 3a (-100 °C): a=10.308(3), b= 6<2h<35°.The intensities of three standard reflections moni- 14.114(3), c=34.121(3) A ° , b=99.08(2)°. tored every hour showed no decay. Corrections for Lorentz PbNb2(OAc)2(OEt)10 3b was obtained accordingly from and polarization eVects were applied. Nb(OEt)5 and Pb(OAc)2 in toluene in a 65% yield. Anal. Calc. Computations were performed using the PC version of for C24H56Nb2O14Pb: C, 29.96; H, 5.82; Nb, 19.33; Pb, 21.56.CRYSTALS.11 Scattering factors and corrections for anomal- Found: C, 29.36; H, 5.48; Nb, 19.19; Pb, 21.19%. ous dispersion were taken from ref. 12. The structure was The various compounds were soluble in common organic solved using direct methods (SHELXS)13 and standard Fourier solvents including hydrocarbons, 3b is poorly soluble in techniques.One of the isopropyl groups [on O(11)] showed ethanol. relatively large thermal parameters compared with the others. A best solution was found by the introduction of two dis- Synthesis of Pb2Nb4O5(OAc)2(OPri)12 4 ordered carbon atoms [C(23) and C(231)] each with half Anhydrous Pb(OAc)2 (0.67 g, 2.06 mmol) was added to a occupancy. All non-hydrogen atoms, except the two disordered solution of Nb(OPri)5 (1.60 g, 4.12 mmol) in 20 ml of toluene carbon atoms, were refined anisotropically.A diVerence Fourier and the reaction medium was refluxed for 40 h. After evapor- map showed the presence of a toluene molecule located on the ation of the solvent, a yellow oil was obtained. Addition of C2 axis, leading to the given complete formula. The very large isopropyl alcohol induced crystallization of thin needles (1.5 g, values of the thermal parameters for this molecule suggests a 86%) which were highly soluble in organic solvents.Anal. disorder around the C2 axis which could not be solved. Atomic coordinates, thermal parameters, and bond lengths Calc. for C40H90O21Nb4Pb2: C, 28.38, H, 5.36; Nb, 21.95; and angles have been deposited at the Cambridge Pb, 24.47.Found: C, 28.10; H, 5.23; Nb, 22.02; Pb, 24.64%. Crystallographic Data Centre (CCDC). See Information for IR(cm-1): 1570s (nasCO2), 1414m (nsCO2); 1160s, 1122s, 1014s, Authors, J. Mater. Chem., 1997, Issue 1. Any request to the 991s, 960s, 850m, 837m, 826w, 800w (nMMOMM), 663s, 618w; CCDC for this material should quote the full literature citation 598s, 569s, 516s, 467m, 430w (nMMOAc, nMOR). 1H NMR and the reference number 1145/47. (CDCl3, -30 °C): 4.85, 4.69, 4.62 (spt, J=6 Hz, 15454, 12H, CH), 2.02 (6 H, O Ac), 1.29, 1.22, 1.16 (d, J=6 Hz, 72H, Me); 207Pb NMR, d 2482. Results and Discussion Synthesis Synthesis of [PbNb2O(OPri)10]m We have investigated the reactions between anhydrous metal Lead iodide (1.12 g, 2.42 mmol) was added to a suspension of acetates M(OAc)2 (M=Mg, Ba, Cd, Pb) and niobium, namely KNb(OPri)6 (2.36 g, 4.85 mmol) in 25 ml of toluene.After stirring at room temperature for ca. 4 h, refluxing was carried Table 2 Crystallographic data for MgNb2(OAc)2(OPri)10·0.5C7H8 at out for 12 h. Potassium iodide was separated by filtration. -100 °C Cooling of the filtrate at -30 °C aVorded large platelets (1.7 g, 70%), soluble in toluene and isopropyl alcohol.Anal. Calc. for Mw 965.1 C30H70Nb2O11Pb: C, 36.03; H, 7.00; Nb, 18.59; Pb, 20.74. a/A° 21.238(5) Found: C, 35.35; H, 6.77; Nb, 17.90; Pb, 21.05%. IR (cm-1): b/A °10.127(10) c/A ° 24.861(3) 1169m, 1135m, 1122s, 1025m, 998s, 984m, 955s; 852m, 829m, a/° 90 800w, 771w, 721w; 577vs, 461m (nMMOR). 1H NMR (CDCl3), b/° 107.77(2) d 5.00, 4.77, 4.65 (spt, J=6 Hz, 75251, 10H, CH), 1.33, 1.31 (d, c/° 90 J=6 Hz, 60H, Me); 13C{1H} NMR (CDCl3), d 75.2 (CH), V /A ° 3 5110(5) 26.1, 25.1, 22.9 (CH3).Z 4 The same product was obtained by reacting [Pb(OPri)2]2 crystal system monoclinic space group P21/c and Nb(OPri)5 in toluene at room temperature. linear absorption coeYcient m/cm-1 4.9 density r/g cm-3 1.26 Synthesis of PbMgNb2O(OAc)2(OPri)10 diVractometer CAD4 Enraf-Nonius radiation Mo-Ka (0.71069) [Pb(OPri)2]2 (0.87 g, 2.67 mmol) was added to a solution of scan type v–2h MgNb2(OAc)2(OPri)10 (1.76 g, 1.91 mmol) in 32 ml of a mixscan range/° 0.80+0.345 tan h ture of hexane–toluene (1551). After stirring for 20 h at room h limits/° 2–20 temperature, the excess of lead isopropoxide was removed by octants collected (hkl) h 0–20, k 0–9, l -23 to 23 filtration; crystallization of 5 as thin needles occurred at no.of data collected 4940 no. of unique data collected 4761 -30 °C (1.25 g, 57.3%). Anal. Calc. for C34H76O15MgNb2Pb: no. of unique data used for refinement 2891 [(Fo)2>3s(Fo)2] C, 35.76; H, 6.71; Mg, 2.13; Nb, 16.27; Pb, 18.14. Found: Rint 0.034 C, 35.53; H, 6.58; Mg, 2.5; Nb, 16.20; Pb, 18.07%.IR(cm-1): R=SdFo|-|Fcd/S|Fo| 0.0544 1605ns, 1582ns, (nasCO2); 1428s (nsCO2); 1329m, 1260w, 1161ns, Rw=[Sw(|Fo|-|Fc|)2/SwFo2]1/2 0.0656(w=1.0) 1125ns, 1017ns, 998ns, 969ns, 953(sh), 899m, 847m, 836m, extinction parameter 0 828(sh), 667m; 604s, 580s, 560ns, 516m, 500m, 467m, 425m, goodness of fit, s 3.9 no. of variables 496 415m, 314m, 291m (nMMOAc, nMOR). 1H NMR (CDCl3), Drmin,max/e A°-3 -0.30, 0.38 d 4.75, 4.67, 4.54 (spt, J=6.1 Hz, 45452; 10H, CH), 1.9 (s, 6H, J. Mater. Chem., 1997, 7(10), 2053–2061 2055ethoxide and isopropoxide. The choice of these systems was is inert toward Nb(OR)5, even by heating in the presence of the parent alcool. The poor reactivity of barium acetate, which motivated by the attractive properties of niobates and tantalates as electrooptical ceramics [LiNbO3, (Sr,Ba)Nb2O6 is actually a polymer based on tetranuclear units, is generally overcome by adding acetic acid,20 but we observed no dissolu- (SBN), (Pb,La)(Ti,Nb)O3 (BLNT), Pb(Sc,Nb)O3 (PSN)],14 as ceramics for microwave resonators [PbMg1/3Nb2/3O3 (PNM), tion even in refluxing conditions by adding variable amounts of AcOH (up to 17 equivalents per Ba, this amount leading to BaZn1/3Ta2/3O3 (BZT), BaMg1/3Ta2/3O3 (BMT)],15,16 or as dielectric ceramics (CdNb2O6).5 Metal acetates are the most gelification).A BaNb2(OPri)12(PriOH)2 species, obtained either by mixing the isopropoxides or by metathesis reaction common precursors associated with metal alkoxides. Among the various coordination modes of the acetate ligand, the between BaI2 and KNb(OPri)6, provides an alternative ‘singlesource’ precursor for Ba–Nb materials.8 Anhydrous zinc acetate assembling ones (bridging or bridging–chelating) are generally favored, this suggests that carboxylates could be a means of is actually a soluble oxide acetate Zn4O(OAc)6, its reaction with Nb(OPri)5 is thus more diYcult to control and less maintaining the stoichiometry along the various steps through the hydrolysis–polycondensation process.These characteristics, selective than for the other divalent metal acetates as shown by the several nasCO2 absorption bands (1590, 1577, which have been exploited for the formation of gels or fibers,17 are however less attractive for MOCVD purposes.18 1515 cm-1). With the exception of the barium acetate, the anhydrous acetates of divalent metals M(OAc)2 (M=Mg, Cd, Reactions between metal alkoxides and carboxylates have generally been considered to proceed by elimination of an Pb) are more reactive than these based on lanthanides since heating was required for the latter.21 ester as a volatile byproduct, thus giving oxo derivatives.However, such reactions can occur in very mild conditions The importance of the solvent on the formation of mixedmetal acetatoalkoxides is noteworthy.Whereas the reaction (room temperature and non-polar solvents), giving compounds whose formulation results from a simple addition. The reac- between Cd(OAc)2 and Nb(OPri)5 proceeds at room temperature, no reaction is observed with [Nb(OEt)5]2, even in tions between divalent metal acetates of magnesium, cadmium and lead, and niobium isopropoxide, illustrate these features.refluxing toluene. Addition of small amounts of ethanol allows the reaction to proceed, probably as a result of the formation These reactions proceed smoothly and quantitatively in hydrocarbons over 1 or 2 h, with progressive dissolution of the of the Nb(OEt)5(EtOH) monomer; CdNb2(OAc)2(OEt)10 is thus obtained.5 The Pb(OAc)2–Nb2(OEt)10 system is acetate according to eqn. (1).The excess of metal acetates is easily removed by filtration, while the novel compounds are more reactive than the Cd(OAc)2–Nb2(OEt)10 system; PbNb2(OAc)2(OEt)10 3b is obtained in toluene, under con- crystallized out almost quantitatively from the filtrate. ditions similar to these of the isopropoxide analog, 3a.The choice of a polar solvent which might act as a ligand toward M(OAc)2+2Nb(OPri)5 CCCCA hexane, 25 °C M=Mg 1, Cd 2, Pb 3a MNb2(OAc)2(OPri)10 M(OAc)2 can be an unfavorable feature for complexation by a metal alkoxide. By contrast with hydrocarbons, no reaction The nasCO2 stretching frequencies of the carboxylate ligands occurs between Pb(OAc)2 and Nb(OEt)5 (152 stoichiometry) in the resulting complexes MNb2(OAc)2(OPri)10 are generally in THF, although solubilization of the acetate is achieved.shifted to higher frequencies with respect to those of the 207Pb NMR, which is a quite convenient tool [l=1/2, sensihomometallic acetates (Table 1). This shift excludes the forma- tivity=20.6%, wide range (#10 000 ppm of chemical shift)]22 tion of mixed crystals M(OAc)2,2Nb(OPri)5. The diVerence for the analysis of the solutions of lead derivatives, only shows D=nasCO2- nsCO2 suggests a chelating or bridging–chelating the presence of complexed lead acetate (d=2230).One can behavior for the OAc ligands.19 notice that whereas Pb6O4(OEt)4 reacts with Nb(OEt)5 to While the reactions between niobium isopropoxide and form Pb6(m4-O)4(m3-OEt)4Nb4(OEt)20,23 the use of Pb(OAc)2 cadmium, magnesium or lead acetates occur easily, diVerent as the source of lead oxide allows to acceed to precursors behaviors are observed in the case of barium and of zinc.having a diVerent Pb–Nb stoichiometry and thus to expand Scheme 1 summarizes the various routes investigated for access the range of ‘single-source’ precursors available for Nb–Pb to Nb–MII species (M=Mg, Ba, Zn, Cd, Pb).Barium acetate oxide materials. The new species are soluble in common organic solvents, thus allowing their characterization by NMR. All compounds 1, 2, 3a, 3b are fluxional, the exchange rate being dependent upon the size of the central nucleus. The exchanges between the diVerent types of alkoxide ligands are frozen out already at room temperature for 1 while lower temperatures (-50 °C) are necessary for 2, 3a and 3b.The acetate ligands are observed as a unique peak in the 1H NMR spectra whereas the alkoxide signals appear as three resonances in a 25251 ratio. The acetate ligands act as clips between the diVerent metals and the solidstate structure is retained upon dissolution in non-polar as well as polar solvents.This is also confirmed by the NMR of the other nuclei (13C, 207Pb, 113Cd). Molecular structures of Nb2M(m-OAc)2(m-OPri)4(OPri)6 The connectivity between the diVerent metals has been established for the magnesium and cadmium derivatives by a singlecrystal X-ray structure determination. Selected bond lengths and angles are collected in Table 3 for 1 and in Table 4 for 2; the molecular structure of MgNb2(OAc)2(OPri)10 is displayed in Fig. 1. The structure is related to that of CdNb2(OAc)2(OPri)10. These trinuclear species display a bent, open-shell structure (NbMMgMNb 139.45° for 1), with alternating Nb and M atoms, all metal atoms being 2 KNb(OPri)6 BaI2 BaNb2(OPri)12(PriOH)2 toluene–PriOH, heat –2 KI MNb2(OAc)2(OPri)10 2 Nb(OR)5 Ba(OPri)2 toluene–PriOH, heat Ba(OAc)2 toluene–ROH, heat M(OAc)2 toluene, RT Zn4O(OAc)6 non-selective Pb(OAc)2 THF Cd(OAc)2 toluene, heat toluene PbNb2(OAc)2(OEt)10 CdNb2(OAc)2(OEt)10 toluene–EtOH M = Mg 1 Cd 2 Pb 3a 3b R = Et and/or Pri Scheme 1 Various routes to Nb–MII species (M=Mg, Ba, Cd, Pb) six-coordinate. The NbMO bond distances span the range 2056 J.Mater.Chem., 1997, 7(10), 2053–2061Table 3 Selected bond distances (A° ) and angles (degrees) for MgNb2(m-OAc)2(m-OPri)4(OPri)6 0.5C6H5CH3 1 Mg(1)MO(1) 2.042(9) Mg(1)MO(3) 2.044(5) Nb(1)MO(10) 1.872(8) Nb(1)MO(11) 1.879(8) Mg(1)MO(5) 2.096(4) Mg(1)MO(6) 2.093(8) Nb(2)MO(2) 2.177(8) Nb(2)MO(5) 2.012(7) Mg(1)MO(7) 2.081(8) Mg(1)MO(8) 2.080(9) Nb(2)MO(7) 2.027(5) Nb(2)MO(12) 1.878(8) Nb(1)MO(4) 2.175(6) Nb(1)MO(6) 2.025(7) Nb(2)MO(13) 1.882(8) Nb(2)MO(14) 1.877(7) Nb(1)MO(8) 2.012(7) Nb(1)MO(9) 1.878(6) Mg(1)···Nb(1) 3.203(4) Mg(1)···Nb(2) 3.194(3) O(3)MMg(1)MO(1) 90.9(3) O(5)MMg(1)MO(1) 89.0(3) O(10)MNb(1)MO(9) 96.0(3) O(11)MNb(1)MO(4) 84.4(3) O(5)MMg(1)MO(3) 168.3(3) O(6)MMg(1)MO(1) 93.7(3) O(11)MNb(1)MO(6) 165.3(4) O(11)MNb(1)MO(8) 92.5(4) O(6)MMg(1)MO(3) 90.5(3) O(6)MMg(1)MO(5) 101.2(3) O(11)MNb(1)MO(9) 95.9(4) O(11)MNb(1)MO(10) 95.0(4) O(7)MMg(1)MO(1) 89.7(3) O(7)MMg(1)MO(3) 93.6(3) O(5)MNb(2)MO(2) 85.5(3) O(7)MNb(2)MO(2) 83.3(3) O(7)MMg(1)MO(5) 74.7(3) O(7)MMg(1)MO(6) 174.7(3) O(7)MNb(2)MO(5) 77.7(3) O(12)MNb(2)MO(2) 85.8(3) O(8)MMg(1)MO(1) 167.1(4) O(8)MMg(1)MO(3) 90.5(3) O(12)MNb(2)MO(5) 168.5(3) O(12)MNb(2)MO(7) 93.7(3) O(8)MMg(1)MO(5) 92.1(3) O(8)MMg(1)MO(6) 73.5(3) O(13)MNb(2)MO(2) 178.4(2) O(13)MNb(2)MO(5) 93.8(3) O(8)MMg(1)MO(7) 103.0(3) O(6)MNb(1)MO(4) 85.2(3) O(13)MNb(2)MO(7) 95.2(3) O(13)MNb(2)MO(12) 94.8(4) O(8)MNb(1)MO(4) 85.7(3) O(8)MNb(1)MO(6) 76.4(3) O(14)MNb(2)MO(2) 84.7(3) O(14)MNb(2)MO(5) 93.3(3) O(9)MNb(1)MO(4) 179.6(3) O(9)MNb(1)MO(6) 94.5(3) O(14)MNb(2)MO(7) 165.5(3) O(14)MNb(2)MO(12) 93.4(3) O(9)MNb(1)MO(8) 94.1(3) O(10)MNb(1)MO(4) 84.2(3) O(14)MNb(2)MO(13) 96.8(3) O(10)MNb(1)MO(6) 94.2(3) O(10)MNb(1)MO(8) 166.8(3) Nb(1)MO(6)MMg(1) 102.2(3) Nb(2)MO(5)MMg(1) 102.0(3) Nb(1)MO(8)MMg(1) 103.1(3) Nb(2)MO(7)MMg(1) 102.0(3) C(1)MO(2)MNb(2) 130.4(7) C(14)MO(8)MNb(1) 136.5(8) C(26)MO(12)MNb(2) 163.7(13) C(5)MO(5)MMg(1) 120.50 C(3)MO(4)MNb(1) 130.9(5) C(20)MO(10)MNb(1) 163.2(12) C(32)MO(14)MNb(2) 163.6(11) C(8)MO(6)MMg(1) 121.4(7) C(8)MO(6)MNb(1) 134.8(6) C(23)MO(11)MNb(1) 168.8(11) C(29)MO(13)MNb(2) 145.4(11) C(11)MO(7)MMg(1) 121.8(5) C(17)MO(9)MNb(1) 147.2(9) C(5)MO(5)MNb(2) 136.8(4) C(1)MO(1)MMg(1) 128.2(7) C(14)MO(8)MMg(1) 120.4(8) C(23)MO(11)MNb(1) 142.7(10) C(11)MO(7)MNb(2) 135.0(6) C(3)MO(3)MMg(1) 128.1(7) 1.872(8)–2.175(6) A ° for 1 and vary according to work is related to that of the structurally characterized Ta–Zn oxoisopropoxide, Ta4Zn2(m3-O)2(m-O)2I2(m-OPri)6(OPri)8.25 NbMOR(t)<NbMm-OR<NbMOAc.The magnesium– oxygen distances spread over the range 2.042(9)–2.096(4) A ° ; Similar observations, formation of diVerent species if the reaction is performed at room temperature or at reflux were the MgMOAc distances are slightly shorter than the MgMOR ones.These data are in accord with the observations made on reported by us for the Pb(OAc)2–Ti(OPri)4 system although the species obtained diVered more in their solubility prop- CdNb2(m-OAc)2(OPri)10.5 The non-bonding Mg,Nb distances are 3.369(8) A ° (av.) and thus reflect the smaller size of erties.26 In contrast to 3a, the oxoacetatoalkoxide Pb2Ti2O(OAc)2(OPri)8, obtained at room temperature, is not the central metal as compared to the Cd–Nb species.The p character of the terminal NbMOR bonds, a feature commonly modified by further heating, probably because the metals are already assembled by an oxo ligand. observed for early transition metals,24 is evidenced by the short NbMO bond distances [1.88 A ° (av.)] as well as by the large NbMOMC angles, 164.8(12)° (av.) and 144.1(10)° (av.) for the equatorial and for the apical NbMOR linkages, respectively.The various metals are six-coordinate but display a distorted surrounding, the distortion is the most severe for the central metal, magnesium, with OMMgMO angles ranging from 73.5(3) to 168.3(3)°. The distortion of the central atom toward a trigonal prismatic surrounding has also been observed for the related heterometallic species CdNb2(m- OAc)2(m-OPri)4(OPri)6 (Table 4) and BaNb2(m-OPri)4- (OPri)8(PriOH)2.8 Reactivity The reactivity of the mixed-metal acetatoalkoxides has been investigated (Scheme 2).Compounds 1 and 2 are quite stable with respect to further condensation by elimination of ester and no evolution is detected after ca. 20 h in refluxing toluene. The behavior of the lead system is more complex. In the case of 3a, FTIR monitoring has shown the formation of isopropylacetate (nasCO2=1745 cm-1) during heating along with Significant diVerences were found if the reaction between that of a new metallic derivative 4, characterized by nasCO2= Nb(OPri)5 and Pb(OAc)2 (151 stoichiometry) was performed 1586 cm-1 (the absorption bands corresponding to 3a pro- in refluxing toluene instead of at room temperature: the gressively decrease); 4, analyzing as Pb2Nb4O5(OAc)2(OPri)12, formation of another oxoacetatoalkoxide 5 is observed.In is more soluble than 3a probably as a result of a close structure contrast to 4, 5 is poorly soluble, the presence of broad with peripheral organic groups.Its crystallization could be absorptions bands in the range 800–650 cm-1 in its FTIR achieved in the presence of isopropyl alcohol. Unfortunately, spectrum suggests extended NbMOMNb bonds and thus a the thin needles obtained were unsuitable for single-crystal X- polymeric nature.27 Compound 5 also displays absorption ray diVraction studies. Low-temperature 1H NMR data indi- bands characteristic of nasCO2 and nsCO2 vibrations at 1547 cate three types of OR groups in a 45454 integration ratio and 1401 cm-1 respectively.These results may explain the and one type of acetate ligand. A structure of type A is in evolution of the solubility over time reported in the literature by refluxing Pb(OAc)2 and Nb(OEt)5 in the course of the accord with the overall spectroscopic data.The overall frame- J. Mater. Chem., 1997, 7(10), 2053–2061 2057Table 4 Selected bond lengths (A ° ) and angles (degrees) for Nb2Cd(m-OPri)4(m-OAc)2(OPri)6 2 Cd(1)MNb(1) 3.3820(8) Cd(1)MNb(2) 3.3962(8) Nb(1)MO(8) 1.991(4) Nb(1)MO(9) 1.878(4) Cd(1)MO(1) 2.260(4) Cd(1)MO(3) (2) 2.270(4) Nb(1)MO(10) 1.877(4) Nb(1)MO(11) 1.882(4) Cd(1)MO(5) 2.291(4) Cd(1)MO(6) 2.312(4) Nb(2)MO(2) 2.169(4) Nb(2)MO(5) 2.008(4) Cd(1)MO(7) 2.332(4) Cd(1)MO(8) 2.313(4) Nb(2)MO(7) 2.003(4) Nb(2)MO(12) 1.876(4) Nb(1)MO(4) 2.167(4) Nb(1)MO(6) 2.025(4) Nb(2)MO(13) 1.887(4) Nb(2)MO(14) 1.876(4) O(5)MCd(1)MO(1) 88.2(2) O(5)MCd(1)MO(3) 159.7(2) O(7)MCd(1)MO(6) 116.2(1) O(8)MCd(1)MO(6) 68.4(1) O(6)MCd(1)MO(1) 159.1(1) O(6)MCd(1)MO(3) 86.4(2) O(7)MCd(1)MO(5) 68.3(1) O(8)MCd(1)MO(5) 110.4(1) O(6)MCd(1)MO(5) 97.2(1) O(8)MCd(1)MO(1) 90.8(1) O(7)MCd(1)MO(6) 116.2(1) O(8)MCd(1)MO(7) 175.2(1) O(7)MCd(1)MO(3) 92.1(1) O(8)MCd(1)MO(3) 89.6(2) O(6)MNb(1)MO(4) 82.6(2) O(7)MNb(2)MO(5) 80.6(2) O(10)MNb(1)MO(8) 170.0(2) O(14)MNb(2)MO(2) 84.1(2) O(8)MNb(1)MO(4) 87.9(2) O(12)MNb(2)MO(2) 86.0(2) O(11)MNb(1)MO(6) 166.1(2) O(14)MNb(2)MO(7) 164.9(2) O(9)MNb(1)MO(6) 95.2(2) O(12)MNb(2)MO(7) 93.7(2) O(11)MNb(1)MO(9) 97.8(2) O(14)MNb(2)MO(13) 96.3(2) O(10)MNb(1)MO(6) 90.8(2) O(13)MNb(2)MO(2) 179.4(2) Nb(1)MO(6)MCD(1) 102.3(2) Nb(2)MO(5)MCD(1) 104.2(2) O(10)MNb(1)MO(9) 93.2(2) O(13)MNb(2)MO(7) 96.1(2) Nb(1)MO(8)MCD(1) 103.3(2) Nb(2)MO(7)MCD(1) 102.9(2) O(11)MNb(1)MO(4) 84.4(2) O(14)MNb(2)MO(5) 90.2(2) Nb(1)MO(4)MC(3) 138.8(4) Nb(2)MO(2)MC(1) 137.2(4) O(11)MNb(1)MO(8) 94.0(2) O(14)MNb(2)MO(12) 93.8(2) Nb(1)MO(6)MC(8) 127.2(4) Nb(2)MO(5)MC(5) 127.8(4) O(11)MNb(1)MO(10) 93.1(2) O(5)MNb(2)MO(2) 86.1(2) Nb(1)MO(8)MC(14) 137.0(4) Nb(2)MO(7)MC(11) 137.2(4) O(8)MNb(1)MO(6) 80.7(2) O(7)MNb(2)MO(2) 83.4(2) Nb(1)MO(9)MC(17) 137.0(5) Nb(2)MO(12)MC(26) 167.5(5) O(9)MNb(1)MO(4) 177.6(2) O(12)MNb(2)MO(5) 170.7(2) Nb(1)MO(10)MC(20) 150.9(5) Nb(2)MO(13)MC(29) 141.7(5) O(9)MNb(1)MO(8) 92.7(2) O(13)MNb(2)MO(5) 93.5(2) Nb(1)MO(11)MC(23) 155.6(4) Nb(2)MO(14)MC(32) 153.6(5) O(10)MNb(1)MO(4) 85.8(2) O(13)MNb(2)MO(12) 94.4(2) 2058 J.Mater. Chem., 1997, 7(10), 2053–2061The reactivity of BaNb2(OPri)12(PriOH)2 7 was also con- Fig. 1 Molecular structure of MgNb2(m-OAc)2(m-OPri)4(OPri)6 show- sidered (Scheme 3).Attempts to control its hydrolysis by ing the atom numbering scheme (ellipsoids at 20% probability) adding acetic acid in toluene at room temperature oVered poorly soluble species, one of them having being identified as Nb4O4(m-OAc)4(OPri)8 (nsCO2 1591, nMMOR 489, 410 cm-1) by comparison with an authentic sample.30 The reaction between 7 and lead acetate is governed by redistribution phenomena with extrusion of barium as barium acetate and formation of a Nb–Pb isopropoxide derivative, [PbNb2O(OPri)10]m 8, which was identified by comparison with an authentic sample obtained more directly by reacting lead iodide and KNb(OPri)6.BaNb2(OPri)12(PriOH)2 displays two solvating alcohol molecules, these may act as functional ligands since their acidity is enhanced by coordination thus allowing their reactivity toward labile metallic species.4,31 Trimethylsilylamides are attractive candidates for the introduction of another metal, the by-product being the volatile amine HN(SiMe3)2.Zinc trimethylsilylamide however was found to be inert toward BaNb2(OPri)12(PriOH)2 at room temperature in hexane. Heating promotes a reaction but zinc separates out from the MNb2(OAc)2(OPri)10 toluene, heat PbMgNb2O(OAc)2(OPri)10 6 Pb4O(OPri)6 toluene–hexane RT M = Mg ePriOH [Pb(OPri)2] M = Mg, Cd toluene–hexane RT [Ba(OPri)2] hexane RT no modification Pb2Nb4O5(OAc)2(OPri)12 4 1– 2 M = Pb toluene, heat –AcOPri reaction medium as insoluble zinc isopropoxide and the initial Scheme 2 Reactivity of the MNb2(OAc)2(OPri)10 (M=Mg, Cd, Pb) Ba–Nb species is recovered by adding isopropyl alcohol.Lead species trimethylsilylamide Pb[N(SiMe3)2]2 is more reactive. Its reaction with 7 at room temperature is evidenced by the discoloration of the reaction medium as well as by the presence in the 207Pb NMR spectrum of two species having chemical shifts preparation of a PNM ceramic.28 Allowing the formation of diVerent from lead isopropoxide derivatives known so far.The the mixed-metal species to proceed, prior to heating, can be major species, which could be isolated, corresponds to a crucial for the homogeneity of the system. derivative 9 containing three metals, Ba, Nb and Pb, and Terheterometallic alkoxides have been reported recently.29 characterized by a single peak at d 4059 in its 207Pb NMR The reactivity of some of the heterometallic species toward spectrum.Attempts to obtain crystals suitable for X-ray other metal complexes has been estimated. In view of the diVraction studies were unsuccessful as a result of its poor formulation of ternary oxides, MgNb2(OAc)2(OPri)10 1 seemed stability. In the presence of small amounts of isopropyl alcohol, to be a good candidate for such investigations.No reaction lead oxoisopropoxide, Pb4O(OPri)6, is obtained together with was observed between 1 and [Ba(OPri)2]2 or Pb4O(OPri)6 at the initial Ba–Nb species 7. The overall results indicate that room temperature, even after two days. The polymeric lead terheterometallic species based on a combination of metals of isopropoxide was more reactive and its dissolution was interest for materials are quite diYcult to stabilize and an observed at room temperature in toluene, 207Pb NMR spectroscopy was used to get some insight into the molecular composition of the reaction medium.Several signals are detected (d 4323, 3838, 3803 and 2591). The predominant species, characterized by the chemical shift at d 4323, could be isolated by crystallization (57% yield) and contains niobium, magnesium and lead. Analytical data imply the empirical formula PbMgNb2O(OAc)2(OPri)10 6.The proton NMR spectrum at room temperature indicates the presence of three types of methine groups in an integration ratio 45452. Lowtemperature spectra show a broadening but give no additional information. 13C NMR spectra however show five types of methine groups.The overall NMR data (1H, 13C and 207Pb NMR) are in accord with a structure of type B. This structure, in which all metals display their usual coordination numbers, namely six for magnesium and niobium, and five for lead, can be viewed as the association between 1 and a PbO moiety. It BaNb2(OPri)12(PriOH)2 7 [PbNb2O(OPri)10]m + Pri 2O 8 1–m –[Ba(OAc)2] –PriOH Pb(OAc)2 toluene, heat toluene AcOH Nb4O4(OAc)4(OPri)8 + poorly soluble species –Pb4O(OPri)6 ePriOH ePriOH 'PbBaNb...' 9 Zn[N(SiMe3)2]2 hexane [BaNb2(OPri)12]m 1–m [Zn(OPri)2] + + 2HN(SiMe3)2 heat RT Pb[N(SiMe3)2]2 is also supported by the facile dissociation of 6 in the presence Scheme 3 Reactivity of BaNb2(OPri)12(PriOH)2 (ePriOH stands for small amount) of trace amounts of isopropyl alcohol to give 1.J. Mater. Chem., 1997, 7(10), 2053–2061 2059appropriate set of ligands is required as in the case of the smooth hydrolysis of PbNb2(OAc)2(OPri)10, that of [PbNb2O(OiPr)10]m 8 in isopropyl alcohol proceeds with compound 6. immediate precipitation. The X-ray diVraction patterns after thermal treatment of the powder (obtained for h=30) show Hydrolysis–condensation reactions no PbNb2O6 but crystalline Pb3Nb4O13 at 600 °C together Hydrolysis–condensation reactions have been achieved for with PbNb4O11 at 750 °C.These observations suggest that various hydrolysis ratios h [h=[H2O]/[MM¾(OR)n+n¾] in THF segregation between the metals occurs at the early stages of or in the parent alcohol (0.1–0.01 M). Powders obtained for the hydrolysis–polycondensation process.This is also conlarge excess of water (h=30) have been analyzed by FTIR, firmed in the Mg–Nb–OR system. thermogravimetry (TG), diVerential thermal analysis (DTA) MgNb2(OAc)2(OPri)10 1, MgNb2(OEt)12(EtOH)2 10 and and X-ray diVraction (XRD) at various temperatures. MgNb2(OPri)12 are all potential precursors of MgNb2O6. Ligands such as carboxylates or b-diketonates are expected These molecules are fragments of a chain made up of to reduce the hydrolysis rates and to modify morphology three octahedra and are thus related to a multicompoand/ or size of the particles.14 The influence of the ancillary nent oxide which displays a columbite-type structure ligands on the properties of the final material can be estimated [(NbO6) (MO6) (NbO6)]2.Hydrolysis proceeds with the forwhen mixed-metal species having the same MM¾ stoichiometry mation of amorphous powders in all cases.As for the powder but a diVerent set of ligands are available. The Mg–Nb, Pb–Nb derived from the hydrolysis of 3a, the thermal elimination of systems, discussed here, and the Pb–Ti system reported prethe acetate ligand is complete below 500 °C for the powder viously26 oVer such opportunities. derived from acetatoalkoxide 1 as precursor [Fig. 3(a)]. The influence of the ancillary ligand, OR vs. OAc, can be Crystallisation starts around 500 °C and the MgNb2O6 phase illustrated with the Pb–Nb system. As expected, diVerential is obtained pure and well crystallized at 600 °C [Fig. 3(b)]. By hydrolysis is observed for the acetatoalkoxide derivatives and contrast, the crystallization only starts at ca. 700 °C for the the powders resulting from the hydrolyses indicate residual powder resulting from hydrolysis of 10 and quite high temperacarboxylate ligands (nasCO2 at 1572 and 1545 cm-1 for 1 and tures are required for the formation of the multicomponent 3a respectively). The hydrolysis of PbNb2(OAc)2(OPri)10 in oxide [Fig. 4(b)]. Partial retention of carboxylate ligands after isopropyl alcohol (0.1 M) gives a homogeneous clear solution the hydrolysis of 1, as shown by FTIR, assists the build-up of up to h=5. PbNb2(OAc)2(OPri)10 remains actually inert an ordered, crystalline material. The formation of sols of toward the addition of water up to h=2 in isopropyl alcohol nanosize particles over a large range of hydrolyses ratios (up (FTIR and 207Pb NMR).Fig. 2(a) shows representative TG to h=24) in isopropyl alcohol, which makes MgNb2(OPri)12 and DTA curves for the powder derived from the hydrolysis an interesting precursor for thin film applications, is however (large hydrolysis ratios) of 3a. A nearly continuous mass loss an unfavorable feature on account of crystallization and homois observed between 80 and 420 °C.This indicates that the geneity. Light scattering measurements have shown that the residual carboxylate ligands are burnt out at relative low sols are polydisperse [20 (88%) and 223 (12%) nm]. Electron temperatures. The DTA curve shows an exotherm around dispersive X-ray analysis (EDAX) by transmission electron 600 °C suggesting the formation of a crystalline material.This microscopy (TEM) indicate that the small particles are mixedis confirmed by XRD. The powders appear amorphous at metal species (Mg5Nb 152) whereas larger particles result room temperature but crystallize at 600 °C as the pure from metal segregation and more reorganization is thus neces- PbNb2O6 phase [Fig. 2(b)], traces of PbNb4O11 are observed sary.Ceramics, such as for instance CdNb2O6 and MgNb2O6 at 800 °C as a result of some loss of lead oxide. By contrast to Fig. 3 TG profile (a) and XRD patterns at various temperatures (b) Fig. 2 TG profile (a) and XRD patterns at various temperatures (b) for the powder resulting from the hydrolysis of PbNb2(OAc)2(OPri)10 for the powder resulting from the hydrolysis of MgNb2(OAc)2(OPri)10 2060 J.Mater. Chem., 1997, 7(10), 2053–20612 L. G. Hubert-Pfalzgraf, Polyhedron, 1994, 13, 1181; Mater. Res. Soc. Symp. Proc., 1992, 271, 15. 3 C. D. Chandler, C. Roger and M. J. Hampden-Smith, Chem. Rev., 1993, 93, 1205. 4 K. G. Caulton and L. G. Hubert-Pfalzgraf, Chem. Rev., 1990, 90, 969. 5 S. Boulmaa� z, R. Papiernik, L. G. Hubert-Pfalzgraf, J. C. Daran and J. Vaissermann, Chem.Mater., 1991, 3, 779. 6 D. C. Bradley, B. N. Chakravarti and W. Wardlaw, J. Chem. Soc., 1956, 2381. 7 M. C. Massiani, R. Papiernik and L. G. Hubert-Pfalzgraf, Polyhedron, 1991, 10, 1667; S. C. Goel, M. Y. Chiang and W. E. Buhro, Inorg. Chem., 1990, 29, 4640. 8 S. Boulmaa�z, R. Papiernik, L. G. Hubert-Pfalzgraf and J. C. Daran, Eur. J. Solid State Inorg. Chem., 1993, 30, 583; E.P. Turevskaya, N. Ya Turova, A. V. Korolev, A. I. Yanovsky and Yu T. Struchkov, Polyhedron, 1995, 14, 1531. 9 D. H. Harris and M. F. Lappert, J. Chem. Soc., Chem. Commun., 1974, 895. 10 H. Burger, W. Sawodny and H. Wannaghat, J. Organomet. Chem., 1965, 3, 113. 11 D. J. Watkin, J. R. Carruthers and P. W. Betteridge, Crystals User Guide, Chemical Crystallography Laboratory, University of Oxford, Oxford, 1986. 12 D. T. Cromer, International T ables for X-Ray Crystallography, Kynoch Press, Birmingham, 1974, vol. 4. 13 G. M. Sheldrick, SHEL XS in Crystallographic Computing 3, ed. G. M. Sheldrick, C. Kruger and R. Goddard, Oxford University Press, Oxford, 1985, p. 175. 14 L. G. Hubert-Pfalzgraf, Chemical Processing of Ceramics, ed. B. I. L. Lee and E. J. A. Pope, M.Dekker, New York, 1994, ch. 2, p. 23; M. I. Yanovskaya, E. P. Turevskaya, V. G. Kessler, I. E. Obvintseva and N. Ya. Turova, Integrated Ferroelectrics, 1992, 1, 343. Fig. 4 TG profile (a) and XRD patterns at various temperatures (b) 15 F. Chaput, J. P. Boilot, M. Lejeune, R. Papiernik and for the powder resulting from the hydrolysis of MgNb2(OPri)12 L. G. Hubert-Pfalzgraf, J. Am. Ceram. Soc., 1989, 72, 1355. 16 O. Renoult, J. P. Boilot, F. Chaput, R. Papiernik, L. G. Hubert- Pfalzgraf and M. Lejeune, Ceramics T oday—T omorrow’s are already well crystallized at 600 °C, while temperatures Ceramics, Elsevier, London, 1991, p. 1991. >1000 °C are required for conventional routes.32 17 J. Livage, M. Henry and C. Sanchez, Prog. Sol. State Chem., 1988, 18, 259. 18 L. G.Hubert-Pfalzgraf, Appl. Organomet. Chem., 1992, 6, 627. Conclusion 19 G. B. Deacon and R. J. Phillips, Coord. Chem. Rev., 1980, 33, 227. 20 A. Mosset, I. Gautier-Luncau, J. Galy, P. Strehlow and Mixed-metal acetatoalkoxides can be obtained readily by H. Schmidt, J. Non-Cryst. Solids, 1988, 100, 339. ting niobium isopropoxide and anhydrous acetates 21 S. Daniele, L. G. Hubert-Pfalzgraf, J.C. Daran and R. Toscano, M(OAc)2 (M=Mg, Cd, Pb) at room temperature in non-polar Polyhedron, 1993, 12, 2091. solvents. The reactions are selective and proceed to the forma- 22 J. J. Dechter, Prog. Inorg. Chem., 1982, 29; R. K. Harris, tion of the smallest aggregates which allow the metals to J. J. Kennedy and W. McFarlane, NMR and the Periodic Table, ed. R. K. Harris and B.E. Mann, Academic Press, London, 1978. achieve their common coordination numbers regarding the 23 R. Papiernik, L. G. Hubert-Pfalzgraf, J. C. Daran and Y. Jeannin, steric demand of the alkoxide ligand. Niobium tends to be six- J. Chem. Soc., Chem. Commun., 1990, 695. coordinate and this is possible via simple addition products. 24 M. H. Chisholm, Chemtracts: Inorg. Chem., 1992, 4, 273. The experimental conditions, choice of solvent and temperature 25 S. Boulmaaz, L. G. Hubert-Pfalzgraf, S. Halut and J. C. Daran, of the reaction are important in determining the reaction J. Chem. Soc., Chem. Commun., 1994, 601. products. The system based on lead is the most reactive and 26 S. Daniele, R. Papiernik, L. G. Hubert-Pfalzgraf, S. Jagner and M. Hakansson, Inorg. Chem., 1995, 34, 628; L. G. Hubert-Pfalzgraf, condensation can be induced thermally. The acetate groups S. Daniele, R. Papiernik, M. C. Massiani, B. Septe, J. Vaissermann act always as assembling ligands as evidenced by the X-ray and J. C. Daran, J.Mater. Chem., 1997, 7, 753. structural data and by the diVerence in the stretching frequen- 27 L. G. Hubert-Pfalzgraf, M. Postel and J. G. Riess, Comprehensive cies nasCO2-nsCO2<200 cm-1. It assists the building up of Coordination Chemistry, Pergamon Press, London, 1987, ch. 34. an ordered, crystalline material in the course of the hydrolysis– 28 T. Fukui, C. Sakurai and M. Okuyama, J. Non Cryst. Solids, 1991, polycondensation process, thus favoring the formation of 134, 293. 29 M. Veith, S. Mathur and V. Huch, J. Am. Chem. Soc., 1996, 118, single-phase materials at quite low temperatures. 903; M. Veith, S. Mathur and V. Huch, J. Chem. Soc., Dalton T rans., 1996, 2485. The authors are grateful to the CNRS (GRECO) for financial 30 R. Papiernik, L. G. Hubert-Pfalzgraf and J. Vaissermann, to be support and Dr F. Chaput (Ecole Polytechnique, Palaiseau) published. for powder X-ray diVraction experiments. 31 B. A. Vaarstra, J. C. HuVman, W. E. Streib and K. G. Caulton, Inorg. Chem., 1991, 30, 3068; V. G. Kessler, L. G. Hubert-Pfalzgraf, S. Halut and J. C. Daran, J. Chem. Soc., Chem. Commun., 1994, 705. References 32 Gmelin, Handbuch der anorganische Chemie ‘Vanadium, Niobium, T antalum,’ Springer-Verlag, Berlin, 1973. 1 S. L. Swartz and V. E. Wood, Condens. Matter News, 1992, 1, 4; G. H. Haertling, Electroceramics, ed. M. L. Levinson, M. Dekker, Paper 7/01381G; Received 27th February, 1997 New York, 1988. J. Mater. Chem., 1997, 7(10), 2053–2061 2061

ISSN:0959-9428    

DOI:10.1039/a701381g

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Preparation, solution behaviour and electrical properties of octasubstituted phthalocyaninato and 2,3-naphthalocyaninato oxotitanium(IV ) complexes Wing-Fong Law,a K. M. Luib and Dennis K. P. Ng*a† aDepartment of Chemistry, T he Chinese University of Hong Kong, Shatin, N.T ., Hong Kong bMaterials T echnology Research Centre, T he Chinese University of Hong Kong, Shatin, N.T ., Hong Kong The octasubstituted oxo(phthalocyaninato)titanium(IV) complexes TiO[Pc¾(CH2OC5H11)8] (Pc¾=2,3,9,10,16,17,23,24- octasubstituted phthalocyaninate) and TiO[Pc(OC4H9)8] (Pc=1,4,8,11,15,18,22,25-octasubstituted phthalocyaninate) and the first 2,3,-naphthalocyaninato titanium complex TiO[Nc¾(C6H13)8] (Nc¾=2,5,11,14,20,23,29,32-octasubstituted naphthalocyaninate) have been prepared by treating the corresponding dicyano-benzenes or -naphthalene with titanium(IV) ethoxide and urea in n-pentanol.These compounds and their analogues TiO[Pc¾(R)8] (R=C7H15, OC5H11) tend to form molecular aggregates in solutions and the eVects of solvent and concentration on their aggregation behaviour were investigated by UV–VIS and 1H NMR spectroscopy. The electrochemistry and electrical properties of these compounds were also studied.Oxo(phthalocyaninato)titanium(IV) (TiOPc) is a well known, method or treating the corresponding dinitriles with TiCl4 followed by hydrolysis.11 All of these macrocycles could be near-IR-active photoconductive dye used practically as xerographic photoreceptors in copiers and GaAs laser printers.1 purified by column chromatography or simply washing with appropriate solvents.The use of this material in optical disk information recording2 and as a p-type semiconductor in photovoltaic cells3 has also been documented. Owing to its poor solubility in usual organic solvents, the purification of TiOPc usually requires tedious procedures such as train sublimation4 and acid pasting,5 and most of the studies have been concentrated on its physical properties in the solid state such as single crystals6,7 and thin solid films.3,7,8 The solution properties of this compound, however, remain relatively unexplored.9 Substitution on the periphery of phthalocyanine will not only enhance its solubility that may facilitate the fabrication of homogeneous thin films, but will also provide entries to tailor the properties of this material. Surprisingly, substituted TiOPc complexes are still scarce.10,11 Here, we describe the preparation of a series of octasubstituted phthalocyaninato and 2,3-naphthalocyaninato oxotitanium(IV) complexes along with their spectroscopic and electrochemical behaviour in solutions.The electrical properties of thin films of TiO[Pc¾(C7H15)8] and TiO[Nc¾(C6H13)8] are also discussed.Results and Discussion Treatment of substituted dinitriles 1, 5 and 6 with titanium(IV) ethoxide and urea in n-pentanol led to the formation of the corresponding metallophthalocyanines 7 and 10, and 2,3- naphthalocyanine 11, respectively. The method was similar to that reported by Pac and co-workers in the synthesis of unsubstituted TiOPc,12 but the yields were considerably lower (12–25%) for substituted analogues.It is noteworthy that compound 11, to our knowledge, represents the first titanium The UV–VIS spectra of these complexes displayed a typical complex of 2,3-naphthalocyanine reported. Reaction of dinitrile Q band and a Soret band which are attributed to the p–p* 2 under similar conditions aVorded a dark green powder.transitions of the macrocycles. As shown in Table 1, the Q Although its UV–VIS spectrum in CHCl3 showed characteristic B and Q bands at 351 and 700 nm, respectively, other spectroscopic methods [1H NMR, IR and liquid secondary ion (LSI) Table 1 UV–VIS data of the oxotitanium complexes 7–11 in CHCl3 MS] and elemental analyses did not support the formation of TiO[Pc¾(CH2OPh)8]. This unknown species seemed to be compound lmax/nm (log e/dm3 mol-1 cm-1) highly aggregated in solutions which hampered the purification 7 352 (4.90), 631 (4.55), 665 (4.60), 700 (5.28) and characterisation processes. The heptyl (8) and pentyloxy 8 347 (4.80), 639 (4.48), 680 (4.42), 710 (5.27) (9) analogues could however be prepared by either using this 9 348 (4.68), 439 (4.37), 631 (4.28), 669 (4.28), 702 (5.12) 10 340 (4.63), 479 (3.99), 704 (4.46), 791 (5.10) 11 337 (4.85), 451 (4.23), 734 (4.56), 782 (4.58), 826 (5.31) † E-mail: dkpn@cuhk.edu.hk J.Mater. Chem., 1997, 7(10), 2063–2067 2063mined for some other substituted phthalocyanines in nonpolar solvents.13a,d The increased aggregation of these complexes in hexanes can be rationalised by the fact that hexanes, having a lower relative permittivity (er=1.89) than CHCl3 (er= 4.81),15 has a weaker screening eVect to disrupt the Pc–Pc interactions. For compounds 9 and 10, such spectral analysis could not be performed because of their poor solubility in hexanes.The naphthalocyaninato complex 11, as expected, is even more highly aggregated in hexanes solution. No absorption spectrum assignable to the monomeric species could be obtained.However, by gradual addition of CHCl3 into dilute hexanes solution of 11, the initial broad band around 770 nm became sharpened giving a characteristic spectrum of metallonaphthalocyanines (Fig. 2). As revealed by UV–VIS spectroscopy, the aggregation of compounds 7–11 in CHCl3 appears to be insignificant. However, at suYciently high concentration, these complexes Fig. 1 Concentration dependence of UV–VIS spectra of 8 in hexanes still have a tendency to form aggregates as shown by 1H NMR spectroscopy. We recorded the 1H NMR spectra of 7–10 in band absorption of 1,4-substituted phthalocyanine 10 shows a CDCl3 over the concentration range 1×10-3–2×10-2 M. bathochromic shift (lmax=791 nm) in comparison with those Fig. 3 shows the concentration dependence of chemical shift of 2,3-substituted analogues 7–9 (lmax=700–710 nm).The Q of the aromatic ring protons for 7–10. The resonances for 2,3- band absorption of naphthalocyanine 11 is even more red- substituted phthalocyanines 7–9 shift upfield by 0.3–0.4 ppm shifted (lmax=826 nm) owing to the more extended p as the concentration approaches 2×10-2 M. It is likely that conjugation.the cone of aromaticity generated by the ring current of one It is well known that phthalocyanines, even in dilute solution, phthalocyanine macrocycle causes upfield shifts of its aggretend to form molecular aggregates such as dimers, trimers and oligomers. These aggregated species may have very diVerent characteristics from the corresponding monomer and the degree of association is largely dependent on the polarity of the solvent and the concentration of the solution.13 We measured the absorption spectra of compounds 7–11 in CHCl3 over the concentration range of 10-6–10-5 M and found no significant spectral change with concentration.The species exhibited spectra typical of monomeric phthalocyanines. Thus, it is apparent that aggregation is not important for these complexes in CHCl3 in these concentrations.In contrast, the absorption spectra of 7 and 8 in hexanes were concentration dependent. Fig. 1 shows the variation of the UV–VIS spectrum of 8 with concentration ranging from 7.34×10-7 to 2.20×10-5 M. The Q band absorption maximum at 696 nm is unshifted but broadened as the concentration increases.Since further spectral change was not observed at concentrations lower than 7.34×10-7 M, the spectrum observed in this concentration can be attributed to the purely monomeric 8. Fig. 2 UV–VIS spectral change of 11 in hexanes upon addition of Compound 7 behaved similarly and the absorption spectrum CHCl3; (a) hexanes only, gradual addition of CHCl3 from (b) to (d) due to monomeric species was obtained at concentrations below 3.25×10-6 M.We assumed that a one-step equilibrium between phthalocyanine monomer (Pc) and aggregated phthalocyanine (Pcn) exists [eqn. (1)], where K is the aggregation constant and n is the aggregation number. nPc ,b) K Pcn (1) By following the treatment described by Mataga,14 eqn. (2) could be derived in which Ct is the total concentration of phthalocyanine, e and em are the observed molar absorptivity and the corresponding value for pure monomer, respectively.log[Ct(1-e/em)]=log(nK)+nlog[Ct(e/em)] (2) Plots of log[Ct(1-e/em)] vs. log[Ct(e/em)] for 7 and 8 gave straight lines from which the values of n and K were determined to be 2.07 and 6.84×104 (for 7) and 2.34 and 1.75×106 (for 8), respectively.The aggregation numbers for both complexes are close to two suggesting that dimer of these compounds may be the dominant species in hexanes solution. The higher value of K for 8 indicates that this compound has a higher aggregation tendency than compound 7 in hexanes and the Fig. 3 Chemical shift of aromatic protons of 7–10 in CDCl3 as a function of log (concentration); (&) 7, (2) 8, (+) 9, ($) 10 value is comparable with the dimerisation constants deter- 2064 J.Mater. Chem., 1997, 7(10), 2063–2067gated partners.16 However, for the 1,4-substituted phthalocyan- oxidation couples, which resembles the energy gap between the highest occupied and the lowest unoccupied molecular ine 10, the chemical shift of ring protons is essentially independent of concentration.This may be attributed to the orbitals (HOMO and LUMO), for 2,3-substituted TiO[Pc¾(R)8] 7–9 was found to be 1.54–1.57 V. These values fact that the ring protons are farther away from the core of phthalocyanine or a weaker aggregation occurs for this substi- are in the range (1.5–1.7 V) reported for the HOMO–LUMO separation of phthalocyanines.17 The corresponding diVerence tution pattern.The 1H NMR spectrum of naphthalocyanine 11 in CDCl3 showed only broad bands for the hexyl side for 1,4-substituted analogue 10 was however smaller (1.29 V) which is in accord with the bathochromic shift of Q band in chains; the aromatic protons’ signals were not observed. However, by adding a few drops of [2H5]pyridine, a broad its UV–VIS spectrum.According to the electrochemical data, the narrowing of the HOMO–LUMO gap in 10 was due to band at d 9.42 and a relatively sharp singlet at d 7.36 appeared which could be ascribed to H1 and H3, respectively. Thus an increase of the HOMO level. The data for 7–9 were also consistent with the electron donating ability of the substituents pyridine is able to disrupt the interactions among these macrocycles to some extent.which follows the order OC5H11>C7H15>CH2OC5H11. The solubility of the substituted macrocyclic compounds The electrochemistry of compounds 7–11 was examined by cyclic voltammetry in CH2Cl2. The voltammograms for8 7–11 in organic solvents renders these compounds suitable for deposition with technique such as spin-coating. Fig. 5 shows (Fig. 4) and 9 revealed two reversible one-electron reduction couples, one quasi-reversible one-electron oxidation couple, the absorption spectra of spin-coated films of 8 and 11 (thickness: ca. 500 A ° ). In comparison with the solution spectra, together with one irreversible oxidation which may be associated with decomposition. Compound 7 gave a similar voltam- the Q bands are significantly broadened, in particular for compound 11.In addition, the lmax for 11 (ca. 775 nm) shows mogram except that an extra quasi-reversible reduction wave was observed, while for compound 10, the second oxidation a hypsochromic shift while that for 8 (ca. 715 nm) remains relatively unchanged. These observations may also indicate also appeared to be quasi-reversible. The voltammogram for naphthalocyanine 11 also displayed two reversible reductions, that the naphthalocyanine 11 has a higher columnar stacking tendency.18 but the oxidation waves were poorly defined. The reversibility of the reduction waves of 7–11 was judged by the separations The electrical properties of the films of 8 and 11 were also briefly examined. The dark conductivity increased with tem- between the anodic and cathodic potentials (62–84 mV) which were almost invariable with scan rates of 50–200 mV s-1, the perature and followed the Arrhenius equation of temperature dependence of conductivity [s=s1 exp(-E1/kT )+s2 cathodic to anodic peak current ratios (ipc/ipa) which approached unity and the linear plots of peak current vs.exp(-E2/kT )]. The Arrhenius plots for both compounds (Fig. 6) show two linear regions of diVerent activation energies square root of the scan rate. A summary of the electrochemical data for these compounds is given in Table 2. All of these (Table 3) with no indication of sharp transition. Since the samples were not intentionally doped, the low activation redox couples are attributed to the macrocyclic ligand as TiIVNO can be considered as a redox inactive moiety.17 energies (E1 ca. 0.04 and 0.24 eV for 8 and 11, respectively) indicate that extrinsic conduction, which may arise from The potential diVerence between the first reduction and structural defects, dominates in the low-temperature regime. Both complexes exhibited semiconducting properties with room-temperature conductivity in the range 10-8 –10-9 V-1 cm-1 which seem to be higher than the values reported for related phthalocyanines.10c As all the conductivity measurements were carried out in complete darkness, we exclude the possibility of extra charge carrier generations due to photo-absorption.On the other hand, it is well known that Fig. 4 Cyclic voltammogram of 8 in CH2Cl2 containing 0.1 M Fig. 5 UV–VIS spectra of spin-coated films of 8 (—) and 11 (A) with a thickness of ca. 500 A ° [NBu4][PF6] at a scan rate of 50 mV s-1 Table 2 Redox potentials for the oxotitanium complexes 7–11a compound Epa b(ox.2) E1/2 c(ox.1) E1/2(red.1) E1/2(red.2) E1/2(red.3) 7 1.54 0.96 -0.58 -0.93 -1.39 8 1.54 0.86 -0.71 -1.06 — 9 1.40 0.80 -0.77 -1.11 — 10 1.00 0.59 -0.70 -1.06 — 11 — — -0.65 -0.97 — aRecorded with [NBu4][PF6] as electrolyte in CH2Cl2 (0.1 M) at ambient temperature. Scan rate=50 mV s-1.Expressed in volts relative to SCE. bAnodic peak potential. cHalf-wave potential. J. Mater. Chem., 1997, 7(10), 2063–2067 2065trations were in the range 10-4 M. All potentials were referenced to the ferrocenium–ferrocene couple (internal standard) at +0.45 V relative to the saturated calomel electrode (SCE).Thin films of 8 and 11 were prepared by dissolving the compounds in tetrahydrofuran and CHCl3, respectively, then spin-coating the solutions on Corning 7059 glass substrates with a spin-coater (Cost EVective Equipment, Model 100 CB) at a spinning rate of 6000 rev min-1 for 30 s. The films were baked at 70 °C for 30 min prior to subsequent measurements. The thicknesses of the films were determined by a film thickness profiler (Tencor Instrument, a-step 500).For conductivity measurements, silver electrodes which defined a gap width of ca. 2 mm were first evaporated onto the samples, then the resistances of the samples were measured in a dynamic vacuum (ca. 10-6 Torr) with an electrometer (Keithley, Model 617) in the V–I mode. The measuring processes started at 200 °C with a ramping down rate of 0.5 °C min-1 to 30 °C.TiO[Pc¾(CH2OC5H11)8] 7 A mixture of 1,2-dicyano-4,5-bis(pentyloxymethyl)benzene 1 (200 mg, 0.61 mmol) and urea (19 mg, 0.31 mmol) was dis- Fig. 6 Plots of log (conductivity) vs. 103 T -1 for spin-coated films of solved in n-pentanol (0.5 ml) to which titanium(IV) ethoxide 8 (#) and 11 ($) (0.04 ml, 0.18 mmol) was added via a micropipette.The mixture was refluxed under nitrogen for 20 h, then mixed with methanol physical properties of thin film materials depend strongly on (15 ml ). After refluxing for a further 15 min, the mixture was their film structures, which may exhibit vast diVerences as a filtered and the residue was washed with methanol (10 ml ), result of diVerent preparation methods and conditions, and of water (10 ml ), and methanol (5 ml) again.The resulting dark any subsequent thermal treatments.19 A direct comparison of blue solid was chromatographed with ethyl acetate–hexanes these data should thus be made with great care. (151) as eluent giving a greenish blue band which was collected and rotary evaporated. The greenish blue solid was dried in vacuo. Yield 53 mg (25%). 1H NMR (CDCl3, 1.6×10-2 M): d Experimental 9.49 (s, 8 H, Pc¾H), 5.14–5.26 (m, 16 H, Pc¾CH2), 3.84 (t, J General 6.7 Hz, 16 H, OCH2), 1.83–1.95 (m, 16 H, CH2), 1.42–1.65 (m, 32 H, CH2), 1.02 (t, J 7.0 Hz, 24 H, CH3); 13C{1H} NMR n-Pentanol and hexanes were distilled from sodium and anhy- (CDCl3): d 151.6, 140.4, 136.3, 123.4, 71.3 (two overlapping drous calcium chloride, respectively.Chromatographic purifi- signals), 29.7, 28.6, 22.7, 14.2. IR: n=978 cm-1 (TiNO); MS cations were performed on silica gel columns (Merck, 70–230 (LSI): an isotopic envelope peaking at m/z 1376.82 (100%) mesh). Dichloromethane used for electrochemical studies was {calc. for M+ based on 48Ti, 1376.81}; Anal. Calc. for freshly distilled from calcium hydride and the electrolyte [NBu4][PF6] was recrystallised from tetrahydrofuran.All C80H112N8O9Ti: C, 69.74; H, 8.19; N, 8.13. Found: C, 67.17; H, other reagents and solvents were of reagent grade and used 8.16; N, 7.68%. without prior purification. The compounds 1,2-dicyano-4,5- bis(pentyloxymethyl)benzene 1,20 3,6-bis(butyloxy)-1,2-dicyanobenzene 5,21 and 2,3-dicyano-5,8-dihexylnaphthalene 622 TiO[Pc(OC4H9)8] 10 were prepared according to the literature procedures.To a mixture of 3,6-bis(butyloxy)-1,2-dicyanobenzene 5 NMR spectra were recorded on a Bruker WM 250 spec- (200 mg, 0.73 mmol) and urea (22 mg, 0.37 mmol) in n-penta- trometer (250 MHz for 1H and 62.9 MHz for 13C) with nol (1 ml) was added titanium(IV) ethoxide (0.05 ml, Si(CH3)4 as an internal standard (d=0).UV–VIS spectra were 0.23 mmol). The mixture was refluxed under nitrogen for 48 h, measured on a Hitachi U-3300 spectrophotometer. IR spectra then mixed with methanol (20 ml ). After refluxing for a further were recorded on a Perkin Elmer 1600 or a Nicolet Magna 30 min, the mixture was cooled to room temperature then 550 spectrometer as KBr pellets. LSI mass spectra were taken filtered, and the solid was washed with diethyl ether (3×10 ml).on a Bruker APEX 47e Fourier transform ion cyclotron The resulting fine dark brown microcrystals were collected resonance spectrometer with 3-nitrobenzyl alcohol as matrix. and dried in vacuo. Yield 26 mg (12%). 1H NMR (CDCl3, Elemental analyses were performed by the Shanghai Institute 2.2×10-2 M): d 7.70 (s, 8 H, PcH), 4.80–5.05 (m, 16 H, OCH2), of Organic Chemistry, Chinese Academy of Sciences. 2.16–2.31 (m, 16 H, CH2), 1.60–1.78 (m, 16 H, CH2), 1.10 (t, Electrochemical measurements were carried out with a BAS J 7.4 Hz, 24 H, CH3); 13C{1H} NMR (CDCl3): d 151.6, 151.2, CV-50W potentiostat using a conventional three-electrode cell 127.0, 118.4, 71.7, 31.5, 19.4, 14.1. IR: n=966 cm-1 (TiNO); equipped with a glassy carbon disc working electrode (3 mm MS (LSI): an isotopic envelope peaking at m/z 1153.59 (100%) diameter), a platinum wire counter electrode and a silver wire {calc.for MH+ based on 48Ti, 1153.56}; Anal. Calc. for pseudo-reference electrode. Typically, experiments were per- C64H80N8O9Ti: C, 66.65; H, 6.99; N, 9.72. Found: C, 65.12; H, formed in CH2Cl2 containing 0.1 M [NBu4][PF6] at ambient temperature under a dry nitrogen atmosphere.The concen- 6.97; N, 9.48%. Table 3 Electrical properties of thin films of compounds 8 and 11 compound temp. range E1/eV temp. range E2/eV s303/V-1 cm-1 8 303–413 K 0.04 413–473 K 0.47 2.5×10-8 11 303–353 K 0.24 353–473 K 0.41 2.2×10-9 2066 J. Mater. Chem., 1997, 7(10), 2063–20678 T. J. Klofta, J.Danziger, P. Lee, J. Pankow, K. W. Nebesny and TiO[Nc¾(C6H13)8] 11 N. R. Armstrong, J. Phys. Chem., 1987, 91, 5646; H. Yanagi, A mixture of 2,3-dicyano-5,8-dihexylnaphthalene 6 (200 mg, S. Chen, P. A. Lee, K. W. Nebesny, N. R. Armstrong and A. Fujishima, J. Phys. Chem., 1996, 100, 5447. 0.58 mmol) and urea (18 mg, 0.30 mmol) was dissolved in n- 9 T. Harazono and I. Takagishi, Bull.Chem. Soc. Jpn., 1993, 66, 1016; pentanol (0.5 ml). Then titanium(IV) ethoxide (0.04 ml, K. Ogawa, J. Yao, H. Yonehara and C. Pac, J.Mater. Chem., 1996, 0.18 mmol) was introduced via a micropipette. The mixture 6, 143; J. Zhou, Y. Wang, J. Qiu, L. Cai, D. Ren and Z. Di, Chem. was refluxed under nitrogen for 20 h, then methanol (20 ml ) Commun., 1996, 2555. was added. After refluxing for a further 30 min, the mixture 10 (a) T.Kashima, Jpn. Kokai T okkyo Koho, JP 63 149 188, 1988 was cooled to room temperature and decanted. The black (Chem. Abstr., 1989, 110, P105174c); (b) JP 63 149 189, 1988 (Chem. Abstr., 1989, 110, P105175d); (c) P. Haisch, G. Winter, M. Hanack, residue was dissolved in hexanes (20 ml ) and precipitated with L. Lu� er, H.-J. Egelhaaf and D.Oelkrug, Adv. Mater., 1997, 9, 316. methanol (10 ml ). The solid was chromatographed with CHCl3 11 W.-F. Law, R. C. W. Liu, J. Jiang and D. K. P. Ng, Inorg. Chim. as eluent. The dark green band developed was collected and Acta, 1997, 256, 147. rotary evaporated to give a dark green solid which was dried 12 J. Yao, H. Yonehara and C. Pac, Bull. Chem. Soc. Jpn., 1995, 68, in vacuo. Yield 50 mg (24%). 1H NMR {CDCl3+ca. 0.6 M 1001. [2H5]pyridine, 4.0×10-3 M): d 9.42 (br s, 8 H, Nc¾H), 7.36 (s, 13 (a) A. R. Monahan, J. A. Brado and A. F. DeLuca, J. Phys. Chem., 1972, 76, 446; (b) A. W. Snow and N. L. Jarvis, J. Am. Chem. Soc., 8 H, Nc¾H), 3.29 (br s, 16 H, Nc¾CH2), 2.00 (br s, 16 H, CH2), 1984, 106, 4706. (c) Ot E. Sielcken, M. M. van Tilborg, M. F. M. 1.66 (br s, 16 H, CH2), 1.47 (br s, 32 H, CH2), 0.95 (t, J 6.9 Hz, Roks, R.Hendriks, W. Drenth and R. J. M. Nolte, J. Am. Chem. 24 H, CH3). IR: n=970 cm-1 (TiNO); MS (LSI): an isotopic Soc., 1987, 109, 4261; (d) W. J. Schutte, M. Sluyters-Rehbach and envelope peaking at m/z 1449.94 (100%) {calc. for M+ based J. H. Sluyters, J. Phys. Chem., 1993, 97, 6069. on 48Ti, 1449.91}; Anal. Calc.for C96H120N8OTi: C, 79.52; H, 14 N. Mataga, Bull. Chem. Soc. Jpn., 1957, 30, 375; S. Tai and 8.34; N, 7.73. Found: C, 77.74; H, 8.27; N, 7.31%. N. Hayashi, J. Chem. Soc., Perkin T rans. 2, 1991, 1275. 15 J. A. Dean, L ange’s Handbook of Chemistry, 14th edn., McGraw- Hill, New York, 1992. We thank The Chinese University of Hong Kong for support 16 D. S. Terekhov, K. J. M. Nolan, C.R. McArthur and C. C. LeznoV, (Direct Grant 94/95). J. Org. Chem., 1996, 61, 3034; C. C. LeznoV and D. M. Drew, Can. J. Chem., 1996, 74, 307. 17 A. B. P. Lever, E. R. Milaeva and G. Speier, in Phthalocyanines: References Properties and Applications, ed. C. C. LeznoV and A. B. P. Lever, VCH, New York, 1993, vol. 3, pp. 1–69. 1 K.-Y. Law, Chem. Rev., 1993, 93, 449. 18 M. J. Cook, J.Mater.Chem., 1996, 6, 677 and references therein. 2 T. N. Gerasimova and V. V. Shelkovnikov, Russ. Chem. Rev., 1992, 19 S. Hasegawa, M. Arai and Y. Kurata, J. Appl. Phys., 1992, 71, 1462; 61, 55. Y. Wu and A. Stesmans, J. Non-Cryst. Solids, 1987, 90, 151; 3 T. Tsuzuki, N. Hirota, N. Noma and Y. Shirota, T hin Solid Films, K. P. Chik and S. H. Fung, J. Non-Cryst. Solids, 1977, 24, 431; 1996, 273, 177; H. Yonehara and C. Pac, T hin Solid Films, 1996, B. A. Orlowski, W. E. Spicer and A. D. Baer, T hin Solid Films, 278, 108. 1976, 34, 31; A. J. Mountvala and G. Abowitz, Vacuum, 1965, 4 J. Mizuguchi, Cryst. Res. T echnol., 1981, 16, 695; H. J. Wagner, 15, 359. R. O. Loutfy and C. K. Hsiao, J.Mater. Sci., 1982, 17, 2781. 20 M. Hanack, A. Beck and H. Lehmann, Synthesis, 1987, 703; 5 G. A. Page, E. G. Tokoli, R. T. Cosgrove and J. W. Spiewak, US M. Hanack, P. Haisch, H. Lehmann and L. R. Subramanian, Pat. 4 557 868, 1985 (Chem. Abstr., 1986, 104, P131453p); Synthesis, 1993, 387. G. Liebermann, A. M. Hor and A. E. J. Toth, Eur. Pat. Appl., EP 21 M. J. Cook, A. J. Dunn, S. D. Howe, A. J. Thomson and 280 520, 1988 (Chem. Abstr., 1989, 110, P97171g). K. J. Harrison, J. Chem. Soc., Perkin T rans. 1, 1988, 2453. 6 W. Hiller, J. Stra�hle, W. Kobel and M. Hanack, Z. Kristallogr., 22 Y.-O. Yeung, R. C. W. Liu, W.-F. Law, P.-L. La. Jiang and 1982, 159, 173; O. Okada and M. L. Klein, J. Chem. Soc., Faraday D. K. P. Ng, T etrahedron, 1997, 53, 9087. T rans., 1996, 92, 2463. 7 J. Mizuguchi, G. Rihs and H. R. Karfunkel, J. Phys. Chem., 1995, 99, 16 217. Paper 7/02637D; Received 17th April, 1997 J. Mater. Chem., 1997, 7(10), 2063–2067 2067

ISSN:0959-9428    

DOI:10.1039/a702637d

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Non-classical FeII spin-crossover behaviour in polymeric iron(II ) compounds of formula [Fe(NH2trz)3]X2 xH2O (NH2trz=4-amino-1,2,4-triazole; X=derivatives of naphthalene sulfonate) Petra J. van Koningsbruggen, Yann Garcia, Epiphane Codjovi, Rene� Lapouyade, Olivier Kahn,* Leopold Fourne`s and Louis Rabardel L aboratoire des Sciences Mole�culaires, Institut de Chimie de la Matie`re Condense�e de Bordeaux, UPR CNRS no. 9048, Avenue du Docteur Schweitzer, F-33608 Pessac, France A new series of iron(II) spin-crossover materials of formula [Fe(NH2trz)3]X2 xH2O [X=1-naphthalene sulfonate (1ns), 2-naphthalene sulfonate (2ns), 4-hydroxy-1-naphthalene sulfonate (4OH-1ns), 4-amino-1-naphthalene sulfonate (4NH2-1ns) and 6-hydroxy-2-naphthalene sulfonate (6OH-2ns)] are reported. The structure of these compounds consists of linear chains in which the Fe(II) ions are linked by triple N1,N2-1,2,4-triazole bridges.All compounds show non-classical spin-crossover behaviour. Optical and magnetic measurements recorded upon heating show an abrupt low-spin to high-spin transition accompanied by a pronounced thermochromic eVect occuring between 330 and 340 K depending on the anion.Thermogravimetric analyses show that the transition is induced by the removal of the two lattice water molecules, which initially stabilized the low-spin state. The dynamical character of this transition has been monitored by extended 57Fe Mo� ssbauer spectroscopic studies on [Fe(NH2trz)3](2ns)2 xH2O. Upon cooling, the dehydrated modifications show classical spin-crossover behaviour with hysteresis at much lower temperatures, ranging from 229 to 297 K depending on the anion.[Fe(NH2trz)3](2ns)2 represents one of the first iron(II) spin-crossover materials showing a spin transition in the close vicinity of room temperature (290 K) accompanied by hysteresis (14 K). Nowadays there is an increasing interest in new bistable iron(II) ing in the following order: ClO4-, I-, Br-, BF4-, NO3-.25,27 spin-crossover compounds, which show a transition from the EXAFS studies on the low-spin forms have shown that the high-spin state (HS, S=2) to the low-spin state (LS, S=0) on FeMN bond distances tend to decrease with the abovecooling, upon increasing pressure, or by light irradiation.1–10 mentioned order of substitution of the anions.25,26 Recently, the use of such materials as molecular-based memory Furthermore, 57Fe Mo� ssbauer spectroscopy studies27 and devices and displays has been investigated.8,11,12 The require- analyses of the X-ray fluorescence spectra24 confirm that this ments for this type of application can be formulated as follows: decrease in interatomic distances correlates with the covalence the compound must exhibit abrupt transitions with large of the FeMN bond in the same series of anions.Therefore, hysteresis (ca. 50 K), the middle of which should preferably be Lavrenova and coworkers concluded that this increase in situated close to room temperature. To allow the material to transition temperatures appears to be associated with the act as a display, the spin transition should be accompanied by increasing anion–cation interactions, and the thereby arising a change in color (thermochromism). Finally, the compound increasing ‘electrostatic pressure’ which causes a subsequent should be stable under normal operating conditions. 2,7,9,11,13,14 increasing compression of the FeN6 octahedron.25 Moreover, A promising class of materials meeting these criteria are the it appears that a rather low transition temperature (T c, where linear chain iron(II) compounds of general formula 50% high-spin FeII and 50% low-spin FeII are present) is [Fe(NH2trz)3]X2 xH2O (NH2trz=4-amino-1,2,4-triazole; accompanied by a rather small hysteresis loop, whereas an X=NO3-,9,15–17 ClO4-,18 BF4-,11,18 I-,19 Br-,11,18 increase in T c leads to an increase in hysteresis width.28 CH3SO3-,20 tosylate21). Although no X-ray crystallographic Within this family of 4-amino-1,2,4-triazole compounds the data are available on these iron(II ) compounds, it is obvious derivative containing tosylate21 shows a rather peculiar FeII that their structure must be comparable to that of spin-crossover behaviour involving a large apparent hysteresis [Cu(NH2trz)3](ClO4)2 0.5H2O.22 The EXAFS study on of about 80 K, which could be ascribed to the synergy between the related materials [Fe(Htrz)2(trz)](BF4) and the FeII spin-crossover behaviour and a dehydration process.[Fe(Htrz)3](BF4)2 H2O (Htrz=1,2,4-4H-triazole; trz=1,2,4- In fact, this type of synergy appears to be a general feature triazolato) confirms that a similar structure can indeed be particular to FeII linear chain compounds containing this obtained with FeII.23 Very recently, a detailed EXAFS study type of planar aromatic anions.Recently, we have observed finally allowed the acquisition of direct information on the the same phenomenon in [Fe(hyetrz)3]X2 3H2O (hyetrz=4- eVect of FeII spin transition on the spatial and electronic (2¾-hydroxyethyl)-1,2,4-triazole; X=3-nitrophenylsulfonate), structure of [Fe(NH2trz)3]X2 (X=NO3-, BF4-, Br-, ClO4-) which gave rise to an unprecedented apparent hysteresis and the magnetically diluted phases [FexZn1-x width of 270 K.29 The structure of the copper(II) analogue (NH2trz)3](NO3)2.24–26 In these compounds the FeII ions are [Cu(hyetrz)3](ClO4)2 3H2O has been elucidated and indeed linked by triple N1,N2-1,2,4-triazole bridges.This direct linkage consists of linear chains in which the CuII ions are linked by of the iron(II) centers results in a large cooperativity of the triple N1,N2-1,2,4-triazole bridges.30 It is worth noting that the spin-crossover behaviour, and therefore considerable hysteresis large thermal hysteresis (60 K) of the mononuclear compound has been found in these materials.7,9 Variation of the non- [Fe{HB(pz)3}]2 [HB(pz)3=tris(1-pyrazolyl)borate]31 is not coordinated anion in [Fe(NH2trz)3]X2 xH2O leads to comdue to the mechanism mentioned above.In fact, the transition pounds with significantly diVerent spin-crossover characterfrom low-spin to high-spin FeII occurring at about 400 K is istics. The temperature of the transition to the high-spin state increases when the radius of the anion diminishes upon chang- associated with a crystallographic phase transition, whereby J.Mater. Chem., 1997, 7(10), 2069–2075 2069the initially well formed crystals shatter into extremely small Synthesis of [Fe(NH2trz)3]X2 2H2O (X=1ns, 2ns, 4OH-1ns, 4NH2-1ns and 6OH-2ns) fragments.31 To study the role of planar aromatic anions in further detail A methanolic solution (40 ml ) containing 2 mmol (1.2 g) of in the 4-amino-1,2,4-triazole systems we have selected the [Fe(H2O)6](2ns)2 and a small amount of ascorbic acid was sulfonate derivatives of naphthalene. This choice allows a added under stirring to a methanolic solution (15 ml ) containsystematic variation of the anion.The sulfonate group can be ing 6 mmol (0.5 g) of 4-amino-1,2,4-triazole.Instantaneously, in two diVerent positions, and other substituents capable of a white precipitate formed, which was filtered and dried in air. forming hydrogen bonds (such as the hydroxy and the amino During the drying the compound turned pink. Yield: 1.04 g group) may be attached to the naphthalene ring system. Here (69%).Elemental analyses. Calc. for C26H30S2O8FeN12 we report on the spin-crossover behaviour of the {[Fe(NH2trz)3](2ns)2 2H2O}: C, 41.17; H, 3.99; N, 22.16; S, [Fe(NH2trz)3]X2 xH2O derivatives containing the mono- 8.45; Fe, 7.36. Found: C, 39.61; H, 3.88; N, 22.77; S, 8.19; valent anions 1-naphthalene sulfonate (abbreviated as 1ns), 2- Fe, 7.20%. naphthalene sulfonate (2ns), 4-hydroxy-1-naphthalene sulfon- The compounds containing other naphthalene sulfonate ate (4OH-1ns), 4-amino-1-naphthalene sulfonate (4NH2–1ns) anions have been prepared analogously.The elemental analyses and 6-hydroxy-2-naphthalene sulfonate (6OH-2ns) (see Fig. 1). of these compounds suggest the same composition as for the 2ns derivative. However, in all cases the C and N analyses indicate a small deficit of ligand.This feature has been generally Experimental ob this type of polynuclear iron(II ) compounds and may be attributed to the presence of a mixture of linear triply Materials N1,N2-1,2,4-triazole bridged compounds diVering in chain Commercially available solvents were used without further length.10 The terminal FeII ions of such a chain are supposed purification.FeCl2·4H2O, 4-amino-1,2,4-triazole and the to have one or more water ligands; consequently, these ions sodium salts of 1-naphthalene sulfonic acid and 6-hydroxy-2- remain in the high-spin state over the whole temperature naphthalene sulfonic acid were purchased from Aldrich. The range. Furthermore, it may not be excluded that the polymeric sodium salts of 2-naphthalene sulfonic acid, 4-hydroxy-1- nature of these compounds gives rise to particular diYculties naphthalene sulfonic acid and 4-amino-1-naphthalene sulfonic in the determination of the elemental analyses.acid were purchased from Acros. The iron(II) salts of the various naphthalene sulfonates were prepared by mixing aque- Results and Discussion ous solutions of iron(II ) chloride and the corresponding naphthalene sulfonate sodium salt.The compounds [Fe(NH2trz)3]X2 xH2O (X=1ns, 2ns, 4OH- 1ns, 4NH2-1ns, 6OH-2ns) appear as purple–pink powders. This pink color is due to the 1A1g�1T1g d–d transition of low- Measurements spin FeII occurring at 520 nm. The color of the compounds changes to white upon heating to ca. 340 K. This white color Elemental analyses were performed by the Service Central is due to the fact that the spin-allowed d–d transition of lowest d’Analyse (CNRS) in Vernaison, France. 57Fe Mo� ssbauer energy of the compound in high-spin state, 5T2g�5Eg, occurs measurements were performed using a constant acceleration at the limit of the visible and IR regions. Halder-type spectrometer with a room temperature 57Co source (Rh matrix) in a transmission geometry.All isomer Optical measurements shifts reported in this work refer to natural iron at room temperature. The spectra were fitted to the sum of Lorentzians Since these compounds are highly thermochromic, the FeII by a least-squares refinement. Thermogravimetric measure- spin transition has been studied optically using a home-made ments were carried out with a Setaram apparatus in the device.7,21 This technique provides an accurate determination temperature range 300–400 K under ambient atmosphere.of the transition temperatures, however it does not give any Magnetic susceptibilities were measured in the temperature information on the percentage of FeII ions involved in the spin range 77–400 K with a fully automatized Manics DSM-8 transition. Since the possible use of these materials in molecular susceptometer equipped with an Oxford instruments DN170 electronics requires stability of the physical behaviour of the continuous-flow cryostat and a Bruker BE15f electromagnet compounds (i.e.retaining of the hysteresis loop) the spinoperating at ca. 0.8 Tesla. Data were corrected for magnetizcrossover behaviour has been investigated during at least three ation of the sample holder and for diamagnetic contributions, thermal cycles.In all cases the temperature has firstly been which were estimated from the Pascal constants. Optical raised to 400 K followed by additional cooling and heating measurements have been carried out using the device described experiments. The results of these optical measurements are previously.7,21 shown in Table 1.All compounds virtually show the same physical properties. It is worth noting that the spin-crossover behaviour shows a paramount analogy with that of the tosylate derivative.21 For Table 1 Results of the optical measurements for [Fe(NH2trz)3]X2 xH2Oa first cycle second cycle third cycle X Tc( Tc3 Tc( Tc3 Tc( Tc3 1ns 330 229 235 229 235 229 4OH-1ns 332 230 240 230 240 230 4NH2-1ns 340 225 235 225 235 225 SO3 – SO3 – SO3 – OH NH2 SO3 – HO SO3 – Fig. 1 Structures of 1-naphthalene sulfonate (1ns), 4-hydroxy-1- 2ns 340 283 297 283 297 283 6OH-2ns 334 260 270 265 270 265 naphthalene sulfonate (4OH-1ns), 4-amino-1-naphthalene sulfonate (4NH2-1ns), 2-naphthalene sulfonate (2ns), and 6-hydroxy-2-naphthalene sulfonate (6OH-2ns) aTc/K has been taken at the half intensity height. 2070 J. Mater. Chem., 1997, 7(10), 2069–2075instance, [Fe(NH2trz)3](1ns)2 xH2O (see Fig. 2) shows upon Interestingly, under these conditions we observed that the compound gradually transforms to its high-spin state. heating a very abrupt low-spin�high-spin transition occurring at 330 K. Decreasing the temperature reveals a very smooth Apparently, even at temperatures lower than 340 K, the temperature at which the low-spin to high-spin transition is found high-spin�low-spin transition with Tc3=229 K.A second heating of the sample shows a rather smooth low-spin�high- in the heating mode while heating at 1 K min-1, the compound is capable of exhibiting spin-crossover behaviour. This feature spin transition at Tc(=235 K.Subsequent heating and cooling cycles indicate that this small hysteresis (Tc(=235 K, Tc3= may be explained by the mechanism associated with this spin transition in the first heating mode, which will be analyzed in 229 K) is retained. Compounds containing the related anions 2ns, 4OH-1ns, 4NH2-1ns and 6OH-2ns show a similar behav- detail in the discussion.iour, diVering only in the position of the transition temperatures. Thermogravimetry Among these compounds the 2ns derivative shows very Thermogravimetric analysis (see Fig. 4) carried out with the interesting characteristics of the stable spin-crossover behavsame velocity of heating (1 K min-1) as for the optical iour occurring in the second and further heating and cooling measurements has been carried out for the 1ns and 2ns experiments (see Fig. 3). A first heating (1 K min-1) results in derivatives. These measurements reveal a continuous loss of an abrupt low-spin�high-spin transition at 340 K. Upon mass starting at room temperature. This decrease in mass cooling a rather abrupt high-spin�low-spin transition occurs proceeds rapidly in the temperature range 315–350 K, after at Tc3=283 K.Heating the compound once more reveals which it continues in a much smoother fashion. It is important another abrupt low-spin�high-spin transition, now situated to notice that at the transition temperatures T c, 330 and 340 K at 297 K. Additional heating and cooling experiments show for 1ns and 2ns, respectively, the mass% lost corresponds to that this hysteresis of 14 K centered in the close vicinity of the removal of 0.5 and 1.3 molecules of lattice water for 1ns room temperature (290 K) remains stable.and 2ns, respectively. If we assume that the measuring con- The rate of heating has a paramount eVect on the spinditions are really identical in the optical and thermogravimetric crossover behaviour. In fact, when the optical measurements measurements, this would imply that the spin transition already for the 2ns derivative are carried out at a heating rate of 0.1 K starts while there are still water molecules present in the sample min-1, the low-spin to high-spin transition proceeds in an and proceeds in a moderately abrupt way until all water abrupt fashion at 330 K.The increasing of the heating rate molecules have been released.This is in sharp contrast to above 1 K min-1 does not aVect the spin-crossover charac- [Fe(hyetrz)3](3-nitrophenylsulfonate)2 3H2O,29 where under teristics with respect to the measurements carried out at identical measuring conditions the spin transition has been 1 K min-1: in all cases, the transition takes place abruptly found to take place in a very abrupt ‘explosive’ fashion at the at 340 K.In addition, optical measurements of moment all water molecules have been removed from the [Fe(NH2trz)3](2ns)2 xH2O have also been carried out by sample. The measurement cell used in the thermogravimetric keeping the sample at a fixed temperature lower than 340 K. analyses is in direct contact with the air. Upon cooling to In a typical experiment the temperature was fixed at 326 K.room temperature an increase in the mass of the 1ns and 2ns samphich indicates that the compounds are being rehydrated: the 1ns compound then has recovered 1.35 molecules of water per formula unit, while the 2ns derivative has only been rehydrated by 1 molecule of water per FeII ion. On the contrary, [Fe(hyetrz)3](3-nitrophenylsulfonate) 2 3H2O has been found not to rehydrate under these experimental conditions.29 57Fe Mo� ssbauer spectroscopy 57Fe Mo�ssbauer spectra have been recorded for [Fe(NH2trz)3](2ns)2 xH2O in the heating mode in the temperature range 293–400 K.Representative Mo�ssbauer spectra are shown in Fig. 5, whereas detailed values of the Mo�ssbauer parameters resulting from the least-squares fitting procedure are listed in Table 2.The area fractions have been calculated Fig. 2 Optical measurement (intensity vs. temperature; recorded at 1 K min-1) for [Fe(NH2trz)3](1ns)2 xH2O Fig. 4 Thermogravimetric analysis (recorded at 1 K min-1) for Fig. 3 Optical measurement (intensity vs. temperature; recorded at [Fe(NH2trz)3](1ns)2 xH2O (dotted line) and [Fe(NH2trz)3](2ns)2 xH2O (full line) 1 K min-1) for [Fe(NH2trz)3](2ns)2 xH2O J.Mater. Chem., 1997, 7(10), 2069–2075 2071Fig. 6 Temperature dependence of the high-spin area fraction xHS as determined by 57Fe Mo� ssbauer spectroscopy for [Fe(NH2trz)3](2ns)2 xH2O. Data obtained from the first (#), second (2) and third (6) heating experiment are included. NO3-, BF4-, Br-, I- and ClO4-.27 This supports the fact that we are dealing with polynuclear compounds in which the active iron(II) spin-crossover sites are provided by FeII ions in a six nitrogen environment.The spin transition is neither complete at low temperatures nor at higher temperatures. The percentage of high-spin FeII detected at lower temperatures may be attributed to ‘defects’ in the crystal lattice, for instance, caused by terminal FeII ions having water molecules in the coordination sphere.Indeed, it has already been reported for this type of iron(II) compounds that the residual high-spin FeII fraction is generally rather high.10 Fig. 6 shows the temperature dependence of the relative Fig. 5 Selected 57Fe Mo� ssbauer spectra for [Fe(NH2trz)3](2ns)2 high-spin area fraction (AHS) as obtained from accumulation xH2O of the spectra for about two days.The data obtained from three heating experiments have been included. Interestingly, all measured data fall upon the same curve showing the assuming identical Lamb–Mo�ssbauer factors for the high-spin and the low-spin state. At 293 K, where according to the recovery of the sample after each heating and subsequent cooling.From these data it may be deduced that the transition optical data the spin transition has not yet started, the spectrum is characterized by a central doublet with a small quadrupole temperature would be 326 K, a value significantly lower than the 340 K obtained from the optical measurements. On first splitting (DEQ) of 0.187(3) mm s-1 and an isomer shift (d) of 0.432(3) mm s-1.The spectral contribution for this doublet is sight these findings seem to be in contradiction. However, in order to interpret all data recorded during the spin transition 92%. This doublet may be assigned to FeII ions in low-spin state. Furthermore, a doublet with a significantly larger quad- it is important to focus on the mechanism of the present FeII spin crossover.In fact, we are dealing with a non-classical rupole splitting [2.70(4) mm s-1] and d of 1.07(4) mm s-1 (spectral contribution=8%) has been observed, which can be spin-crossover behaviour which is induced by the removal of water molecules (vide infra). Clearly, this loss of water molecules attributed to FeII in high-spin state. At increasing temperatures this FeII low-spin doublet gradually decreases in intensity, is governed by a kinetics that is rather slow at lower temperatures, and increases at higher temperatures. Consequently, the while the doublet with larger quadrupole splitting gains intensity.At 373 K, only 13% of FeII in low-spin state is present temperature at which the equilibrium value of 50% high-spin and 50% low-spin FeII is detected significantly depends on the [DEQ=0.341(3) mm s-1 and d=0.18(3) mm s-1]. The fraction of FeII in high-spin state is 87% and the doublet is characterized measuring conditions.If the increase of temperature is rather fast, e.g. 1 K min-1, in the optical measurements, the transition by DEQ=2.532(3) mm s-1 and d=0.986(3) mm s-1. Similar values for these Mo�ssbauer parameters have also been reported temperature is found at 340 K.By Mo�ssbauer spectroscopic measurements a time-averaged spectrum is collected over a in an extended study on a series of linear iron(II ) chain spincrossover compounds of 4-amino-1,2,4-triazole with the anions much longer period of time. This allows the dehydration of Table 2 Mo�ssbauer parameters (mm s-1) for [Fe(NH2trz)3](2ns)2 xH2Oa T /K d(LS) DEQ(LS) C/2(LS) d(HS) DEQ(HS) C/2(HS) AHS(%) 293 0.432(3) 0.187(3) 0.274(3) 1.07(4) 2.70(4) 0.43(7) 8 316 0.422(4) 0.191(4) 0.261(4) 1.06(3) 2.65(3) 0.39(6) 15 321 0.415(7) 0.178(7) 0.284(7) 1.01(6) 2.67(6) 0.6(1) 18 326 0.413(7) 0.173(7) 0.318(9) 1.017(4) 2.700(4) 0.294(7) 51 331 0.40(1) 0.16(3) 0.34(2) 1.011(3) 2.680(3) 0.284(5) 72 335 0.39(2) 0.12(2) 0.35(3) 1.006(6) 2.664(6) 0.29(1) 80 345 0.26(3) 0.30(3) 0.47(9) 1.002(6) 2.651(6) 0.26(1) 78 354 0.24(4) 0.25(4) 0.44(5) 0.995(2) 2.603(2) 0.278(4) 85 364 0.22(4) 0.30(3) 0.41(8) 0.988(5) 2.561(5) 0.280(9) 87 398 0.18(2) 0.29(2) 0.32(3) 0.971(2) 2.434(2) 0.284(4) 87 a=isomer shift, DEQ=quadrupole splitting, C/2=half-width of the lines, AHS=area fraction of the high-spin doublets. 2072 J. Mater. Chem., 1997, 7(10), 2069–2075the sample to occur at a much lower temperature.Therefore, it might not be correct to directly compare these d values, since the variation observed in these values is extremely small a lower transition temperature of 326 K has been observed. Indeed, these findings are in agreement with the optical and may fall within the uncertainty range. Furthermore, it is certainly not clear whether the present compound would obey measurements performed with a heating rate of 0.1 K min-1, where a transition temperature of 330 K has been found.this relation, since its non-classical spin-crossover mechanism involving the release of lattice water molecules certainly diVers Furthermore, the optical measurements carried out at a fixed temperature below 340 K (326 K) have shown that the from those of the compounds investigated by the Russian group. compound is also capable of exhibiting the spin-crossover behaviour at these lower temperatures.In order to confirm the dynamical character of this non- Magnetic measurements classical spin transition, additional Mo�ssbauer spectroscopic Variable-temperature magnetic susceptibility measurements measurements have been performed.In an experiment, the have been carried out in order to obtain additional infor- temperature has been fixed at 326 K and aMo� ssbauer spectrum mations on the completeness of the non-classical spin transition has been recorded every 30 minutes. As the measurement of [Fe(NH2trz)3](2ns)2 xH2O. At 307 K the xT value [x being proceeds at this fixed temperature, the compound starts to the magnetic susceptibility per iron(II) ion and T the tempera- transform to the high-spin state.Probably, the compound ture] is 0.248 cm3 K mol-1. Upon increasing the temperature, slowly starts to loose its lattice water molecules. The xT attains a value of 3.512 cm3 K mol-1 at 364 K, which Mo�ssbauer data listed in Table 3 show the evolution of the roughly corresponds to the value expected for a quintet state.percentage of FeII in high-spin state as a function of time. The percentages listed correspond to the accumulation of the timeaveraged spectra over the total measuring time indicated. After Concluding remarks 30 min 40.08% of FeII in high-spin state is observed. As the time proceeds, the high-spin FeII fraction gradually increases.Here we have described a new series of polynuclear iron(II) spin-crossover materials based on the family [Fe(NH2- Finally, the spectrum recorded after accumulation of the data over 300 min shows Interestingly, trz)3]X2 xH2O (X=1ns, 2ns, 4OH-1ns, 4NH2-1ns and 6OH-2ns). All these compounds which contain derivatives of the same percentage of high-spin FeII ions (51%) has been found at 326 K, while accumulating the Mo�ssbauer data for naphthalene sulfonate as non-coordinating anions show similar characteristics in their spin-crossover behaviour as the pre- about two days.Apparently, to reach the equilibrium ratio of high-spin and low-spin FeII under these experimental con- viously reported tosylate derivative.21 In particular, the physical properties of the compounds are determined by the synergy ditions and at this temperature a measuring time of about 300 min is required, after which this ratio remains stable.These between the FeII spin-crossover behaviour and a dehydration– rehydration process. For all the compounds, the mechanism measurements clearly indicate the non-classical character of this spin transition.We may then interpret the identical describing the physical properties is as follows: at room temperature the thermodynamically stable state for the physical behaviour observed during the three heating experiments (see Fig. 6) by a complete rehydration taking place upon hydrated compound [Fe(NH2trz)3](naphthalene sulfonate derivative)2 xH2O is the low-spin state.Evidently, this low- cooling of the sample, which resides in a measurement cell open to the air. spin state is stabilized by the hydrated nature of this modifi- cation. Indeed, studies on mononuclear iron(II ) spin-crossover In addition, the significant line broadening of the high-spin absorption observed in the spectrum recorded at 321 K [C/2= compounds have already revealed that the low-spin state may be stabilized by interactions with lattice water molecules.32–35 0.6(1) mm s-1] may also be indicative of the presence of highspin FeII ions in diVerent FeN6 environments.Upon heating, the compound starts to lose its lattice water molecules, and consequently the stabilization of the low-spin For a series of related iron(II) compounds exhibiting classical spin transitions Varnek and Lavrenova have observed a state ceases. For the present series of naphthalene sulfonate compounds, it appears that the destabilization of the low-spin decrease in the isomer shift value for the low-spin state (determined either at 78 K or at 295 K) as a function of state already occurs while the compounds are not yet completely dehydrated.At first sight this may seem to be increasing transition temperatures.27 This linear correlation showed a much steeper dependence at 78 than at 295 K. These in contrast with [Fe(hyetrz)3](3-nitrophenylsulfonate)2·xH2O, where optical and thermogravimetric measurements carried authors have concluded that this corresponds to a higher degree of covalency of the FeMN bonds as the transition out at an identical heating rate (1 K min-1) have shown that no lattice water molecules are present in the sample as the temperature increases.The d value at 293 K of 0.432(3) mm s-1 for the present compound is somewhat higher than the low-spin to high-spin transition occurs in an extremely abrupt way at 370 K.29 However, recently, new experiments on values for the series of compounds reported by the Russian group, which range from 0.427 to 0.416 mm s-1.This would [Fe(hyetrz)3](3-nitrophenylsulfonate)2 xH2O have been carried out: the optical data recorded while fixing the temper- imply that a transition temperature lower than room temperature would be expected for the present compound. However, ature at a value below 370 K show that also at these lower Table 3 Evolution of the Mo�ssbauer parameters (mm s-1) as a function of time at 326 K for [Fe(NH2trz)3](2ns)2 xH2Oa t/min d(LS) DEQ(LS) C/2(LS) d(HS) DEQ(HS) C/2(HS) AHS(%) 30 0.41(3) 0.14(3) 0.28(3) 0.99(3) 2.71(3) 0.26(4) 40.08 60 0.41(1) 0.16(1) 0.28(2) 1.01(2) 2.69(2) 0.26(3) 40.20 90 0.41(1) 0.16(1) 0.31(2) 1.01(1) 2.69(1) 0.28(2) 41.82 120 0.41(1) 0.17(1) 0.28(1) 1.01(1) 2.03(1) 0.27(2) 43.39 150 0.41(1) 0.19(1) 0.28(1) 1.01(1) 2.70(1) 0.28(2) 43.62 180 0.41(1) 0.19(1) 0.28(1) 1.01(1) 2.71(1) 0.27(1) 44.45 210 0.412(9) 0.193(9) 0.29(1) 1.014(9) 2.711(9) 0.29(1) 46.65 240 0.414(9) 0.188(9) 0.29(1) 1.015(9) 2.708(9) 0.30(1) 47.81 270 0.41(1) 0.18(1) 0.29(1) 1.01(1) 2.701(1) 0.30(1) 48.51 300 0.413(7) 0.173(7) 0.318(9) 1.017(4) 2.700(4) 0.294(7) 51.00 ad=Isomer shift, DEQ=quadrupole splitting, C/2=half-width of the lines, AHS=area fraction of the high-spin doublets.J. Mater. Chem., 1997, 7(10), 2069–2075 2073temperatures the material slowly transforms to the high-spin completely rehydrate the dehydrated sample back to its initial state. This process of spin crossover induced by dehydration state.36 Therefore, also in that case the spin transition is associated with the loss of lattice water molecules.There are (low-spin to high-spin) and reversible by hydration (high-spin to low-spin) is entirely reproducible. In fact this process may indications that [Fe(NH2trz)3](tosylate)2 xH2O has similar characteristics for the low-spin to high-spin transition taking be regarded as a self-assisted process related to the solvateinduced change of spin state, which has already been reported place at 361 K. 21 For the series of compounds with derivatives of naphthalene sulfonate as anion, the transition temperatures for [Fe(hyetrz)3](3-nitrophenylsulfonate)2 3H2O.29 The dehydrated modifications show spin-crossover behav- as determined by optical measurements using a velocity of 1 K min-1 are in the range 330–340 K.Obviously, these lower iour at lower temperatures. The genuine spin transition with stable hysteresis is centered at temperatures ranging from 229 transition temperatures reflect the easier destabilization of the low-spin state in compounds containing naphthalene sulfonates to 297 K for the various compounds. The diVerences observed in transition temperatures within the series may be due to as compared to phenyl sulfonates. From the proposed mechanism it follows that this spin- slight structural variations induced by the anions.For [Fe(NH2trz)3](2ns)2 a stable hysteresis of 14 K centered crossover behaviour should be entirely governed by the removal of lattice water molecules. Indeed, various experiments around 290 K, i.e. in close vicinity to room temperature, has been found.Up to now, such features have been observed have shown that the observed transition temperatures depend very much on the measurement conditions, in particular the for a few compounds with 4-amino-1,2,4-triazole. For [Fe(NH2trz)3](CH3SO3)2·H2O, Bronisz et al. reported a hys- heating rate. For all hydrated naphthalene sulfonate compounds, the teresis of 26 K centered around 282 K.20 Surprisingly, these authors also reported that the dehydrated form of this com- transition temperatures for this first low-spin to high-spin transition all occur in a very narrow temperature range.This pound shows spin-crossover behaviour with a small hysteresis (4–8 K) centered around ca. 296 K.20 Furthermore, indicates that the destabilization of the low-spin state or alternatively, the removal of the lattice water molecules occurs [Fe(NH2trz)3](tosylate)2 has been reported to have a hysteresis of 17 K around 290 K.21 Moreover, the mixed-ligand material in a similar way.This is also supported by the results of the thermogravimetric studies showing very similar dehydration [Fe(Htrz)3-3x(NH2trz)3x](ClO4)2 xH2O has been reported to exhibit hysteresis of 17 K centered around 304 K. 14 Therefore, characteristics for the 1ns and 2ns derivatives (see Fig. 4). Therefore, it may be supposed that the anions of the naphtha- [Fe(NH2trz)3](2ns)2 represents one of the very few iron(II) spin-crossover materials showing a spin transition and an lene sulfonate type also have similar capabilities for hydrogen bonding with the lattice water molecules.On the other hand, associated thermochromic eVect in close vicinity of room temperature with hysteresis. compounds containing derivatives of phenyl sulfonate generally show higher transition temperatures.21,29 Clearly, in these latter compounds the loss of water molecules is more diYcult, which may re from a diVerent way of incorporation of the References anions in the crystal lattice allowing stronger hydrogen-bond- 1 P.Gu� tlich, Struct.Bonding (Berlin), 1981, 44, 83. ing interactions with the lattice water molecules. 2 J. Zarembowitch and O. Kahn, New J. Chem., 1991, 15, 181. Furthermore, in view of the striking similarities (identical 3 P. Gu� tlich, in Mo�ssbauer Spectroscopy Applied to Inorganic transition temperature for the first low-spin to high-spin Chemistry, ed.G. J. Long, Modern Inorganic Chemistry Series, transition) between the spin-crossover behaviour of tosylate Plenum Press, New York, 1984, vol. 1. compounds containing the 4-amino-1,2,4-triazole21 and 4- 4 E.Ko� nig, Prog. Inorg. Chem., 1987, 35, 527. 5 P.Gu� tlich and A. Hauser, Coord. Chem. Rev., 1990, 97, 1.alkyl-substituted 1,2,4-triazoles,37 it may be excluded that 6 P. Gu� tlich, A. Hauser and H. Spiering, Angew. Chem., Int. Ed. hydrogen bonding interactions involving the 4-amino substitu- Engl., 1994, 33, 2024. ent and the lattice water molecules are the determining factor 7 O. Kahn and E. Codjovi, Philos. T rans. R. Soc. L ondon A, 1996, in maintaining these lattice water molecules. 354, 359. This leads us to the idea that the lattice water molecules in 8 O. Kahn,MolecularMagnetism, VCH, New York, 1993. such compounds are bound in a very loose fashion, which 9 O. Kahn, E. Codjovi, Y. Garcia, P. J. van Koningsbruggen, R. Lapouyade and L. Sommier, in Molecule-Based Magnetic might be comparable to the way water molecules are incorpor- Materials, ed.M. M. Turnbull, T. Sugimoto and L. K. Thompson, ated in zeolites. Indeed, a zeolite is an aluminosilicate with a Symp. Ser. No. 644, ACS,Washington, DC, 1996, p. 298. structure enclosing cavities occupied by large ions and water 10 J. G. Haasnoot, in Magnetism: A Supramolecular Function, ed. molecules, both having considerable freedom of movement O. Kahn, Kluwer Academic, Dordrecht, Netherlands, 1996, p. 299. which permits ion exchange and reversible dehydration. For 11 O. Kahn, J. Kro�ber and C. Jay, Adv.Mater., 1992, 4, 718. [Fe(NH2trz)3](aryl sulfonate)2 xH2O one can propose that 12 C. Jay, F. Grolie`re, O. Kahn and J. Kro�ber,Mol. Cryst. L iq. Cryst., 1993, 234, 255. the shape and size of the diVerent aryl sulfonates induce the 13 J. Kro� ber, J.-P. Audie`re, R.Claude, E. Codjovi, O. Kahn, pore size of the polymeric compounds which, in turn, lead to J. G. Haasnoot, F. Grolie`re, C. Jay, A. Bousseksou, J. Linare`s, a considerable variation in the rate of water uptake or release. F. Varret and A. Gonthier-Vassal, Chem.Mater., 1994, 6, 1404. It may be proposed that the stabilization of the low-spin state 14 J. Kro� ber, E. Codjovi, O.Kahn, F. Grolie`re and C. Jay, J. Am. of FeII by lattice water molecules results from a cooperative Chem. Soc., 1993, 115, 9810. solvation, where hydrogen bonding of these water molecules 15 L. G. Lavrenova, V. N. Ikorskii, V. A. Varnek, I. M. Oglezneva and S. V. Larionov, J. Struct. Chem., 1993, 34, 960. to the sulfonate group increases the nucleophilicity of the 16 L. G. Lavrenova, V.N. Ikorskii, V. A. Varnek, I. M. Oglezneva water oxygen atom, and hence its interaction with FeII mainand S. V. Larionov, Koord. Khim., 1986, 12, 207. tained in its low-spin state. 17 V. A. Varnek and L. G. Lavrenova, J. Struct. Chem., 1994, 35, 842. Interestingly, the thermogravimetric studies also show that 18 L. G. Lavrenova, V. N. Ikorskii, V. A. Varnek, I. M. Oglezneva the compounds are easily being rehydrated.Therefore, the spin and S. V. Larionov, Koord. Khim., 1990, 16, 654. transition from high-spin to low-spin may also be induced by 19 L. G. Lavrenova, N. G. Yudina, V. N. Ikorskii, V. A. Varnek, I. M. Oglezneva and S. V. Larionov, Polyhedron, 1995, 14, 1333. rehydration. This feature has been tested by several experimen- 20 R. Bronisz, K.Drabent, P. Polomka and M. F. Rudolf, Conference tal techniques. For the 2ns derivative, the results of the optical, Proceedings, ICAME95, 1996, 50, 11. magnetic susceptibility, and 57Fe Mo� ssbauer spectroscopic 21 E. Codjovi, L. Sommier, O. Kahn and C. Jay, New J. Chem., 1996, measurements recorded on a fresh sample have been found to 20, 503. be exactly identical to those obtained on an already measured 22 V. P. Sinditskii, V. I. Sokol, A. E. Fogel’zang, M. D. Dutov, (i.e. dehydrated) sample left in contact with the air for a few V. V. Serushkin, M. A. Porai-Koshits and B. S. Svetlov, Russ. J. Inorg. Chem., 1987, 32, 1149. minutes. Evidently, ambient vapour pressure is suYcient to 2074 J. Mater. Chem., 1997, 7(10), 2069–207523 A. Michalowicz, J. Moscovici, B. Ducourant, D. Cracco and 31 F. Grandjean, G. J. Long, B. B. Hutchinson, L. Ohlhausen, P. Neill and J. D. Holcomb, Inorg. Chem., 1989, 28, 4406. O. Kahn, Chem. Mater., 1995, 7, 1833. 24 N. V. Bausk, S. B. E� renburg, L. N. Mazalov, L. G. Lavrenova and 32 K. H. Sugiyarto, D. C. Graig, A. D. Rae and H. A. Goodwin, Aust. J. Chem., 1994, 47, 869. V. N. Ikorskii, J. Struct. Chem., 1994, 35, 509. 25 N. V. Bausk, S. B. E� renburg, L. G. Lavrenova and L. N. Mazalov, 33 K. H. Sugiyarto and H. A. Goodwin, Aust. J. Chem., 1988, 41, 1645. 34 M. Sorai, J. Ensling, K. M. Hasselbach and P. Gu� tlich, Chem. J. Struct. Chem., 1995, 36, 925. 26 S. B. E� renburg, N. V. Bausk, V. A. Varnek and L. G. Lavrenova, Phys., 1977, 20, 197. 35 T. Buchen, P. Gu� tlich, K. H. Sugiyarto and H. A. Goodwin, Chem. J.Magn.Magn. Mater., 1996, 157/158, 595. 27 V. A. Varnek and L. G. Lavrenova, J. Struct. Chem., 1995, 36, 104. Eur. J., 1996, 2, 1134. 36 Y. Garcia, P. J. van Koningsbruggen, E. Codjovi, R. Lapouyade 28 V. Ksenofontov and P. Gu� tlich, personal communication, 1995. 29 Y. Garcia, P. J. van Koningsbruggen, E. Codjovi, R. Lapouyade, and O. Kahn, in preparation. 37 J. Kolnaar and J. G. Haasnoot, personal communication, 1996. O. Kahn and L. Rabardel, J.Mater. Chem., 1997, 7, 857. 30 Y. Garcia, P. J. van Koningsbruggen, G. Bravic, P. Guionneau, D. Chasseau, G. L. Cascarano, J. Moscovici, K. Lambert, Paper 7/02690K; Received 21st April, 1997 A. Michalowicz and O. Kahn, Inorg. Chem., submitted. J. Mater. Chem.,

ISSN:0959-9428    

DOI:10.1039/a702690k

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Fluorination of the Ruddlesden–Popper type cuprates, Ln2-xA1+xCu2O6-y (Ln=La, Nd; A=Ca, Sr) Peter R. Slater,*a Jason P. Hodges,b M. Grazia. Francesconi,b Colin Greavesb and Marcin Slaskic aSchool of Chemistry,University of St. Andrews, St. Andrews, Fife, UK KY16 9ST bSchool of Chemistry, University of Birmingham, Edgbaston, Birmingham, UK B15 2T T cSchool of Physics and Space Research, University of Birmingham, Edgbaston, Birmingham, UK B15 2T T The low-temperature (200–350 °C) fluorination of the Ruddlesden–Popper type cuprates, Ln2-xA1+xCu2O6-y (Ln=La, Nd; A= Ca, Sr) using F2 gas, CuF2, and NH4F is reported.The incorporation of large levels of fluoride ions is observed for each of these fluorinating agents. The general characteristics of each method are discussed, and it is shown that for this system, fluorination mainly occurs by insertion of fluorine for reaction with F2, substitution of fluorine for oxygen for NH4F, and a mixture of the two processes for CuF2.For A=Sr, it is assumed that fluorine inserts mainly between the two CuO2 layers, since large expansions of the unit cell along the c direction are observed. No evidence for bulk superconductivity has so far been observed after fluorination.Recently we have reported that the reaction of the alkaline- La1.9Sr1.1Cu2O6-y and La1.9Ca1.1Cu2O6-y. The intimately earth cuprates A2CuO3 (A=Ca, Sr) with F2 gas at low ground powders were heated in air at a temperature between temperatures (ca. 200 °C) yields the oxide fluorides 1050 and 1075 °C for 18 h, reground, and then reheated at the A2CuO2F2+d , with superconductivity (Tc ca. 46 K) being same temperature for a further 18 h, followed by furnace observed for A=Sr.1,2 We have subsequently demonstrated cooling. two other simple low-temperature solid-state fluorination routes to these oxide fluorides, involving the facile reaction (at Fluorination by solid/F2 gas reaction ca. 230 °C) of A2CuO3 with NH4F3 or transition-metal difluor- The samples were heated in 10%F2–90%N2 (passed over NaF ides (e.g.CuF2, ZnF2).4 Moreover partial substitution of Ba to remove HF) at a temperature between 200 and 250 °C for Sr has been shown to raise the Tc as high as 66 K, for a time ranging from 15 to 150 min. The temperature significantly the highest Tc for any phase with the confirmed used depended on the system [Nd1.3Sr1.7Cu2O6-y (205 °C), La2CuO4 structure.Since for other cuprates, it has been shown La1.9Sr1.1Cu2O6-y (240 °C), La1.9Ca1.1Cu2O6-y (250 °C)] and that Tc increases with increasing number of Cu layers (generally reaction times, which ranged from 15 min to 5 h, determined up to three layers), it was of extreme interest to look at rethe extent of fluorination.lated materials with multiple Cu layers. Sr2CuO3 may be considered as the first member of the homologous series, Fluorination by NH4F (solid-state reaction) Srn+1CunO2n+1+d. The n=2 member has been prepared by high-pressure synthesis,5 but for ambient pressure synthesis, For fluorination using NH4F a reaction temperature of rare-earth elements, Ln, are required to partially substituted 300–350 °C was required and up to 20 mol of NH4F were used for the alkaline-earth metals, and this represents the widely known Ruddlesden–Popper type double-layer cuprates, Ln2-xA1+xCu2O6-y (Ln=rare-earth, A=Ca, Sr).6–9 As in the case of Sr2CuO3, significant oxygen vacancies are located in these phases, mainly between the double CuO2 layers such that the coordination of Cu is square pyramidal (Fig. 1). These phases are non-superconducting as synthesised, but a number of groups have induced superconductivity (Tc up to 60 K) using a variety of routes including high-oxygen-pressure annealing, control of synthesis conditions, or use of an oxidizing agent such as KClO3.10–12 Our aim was to attempt to induce superconductivity through fluorination. In phases with high Sr content, e.g.Nd1.3Sr1.7Cu2O6-y, significant vacancies are also located in the CuO2 planes, and an ordering of these vacancies occurs leading to a tripled cell along one direction.9 This gives rise to a unique one-dimensional tunnelled copper–oxygen sublattice built from vertex linked CuO5 square pyramids (Fig. 2), where it should be possible to insert fluorine similar to the case of A2CuO3 (A= Ca, Sr, Ba).In this paper we report detailed studies on the fluorination of these double-layer cuprates, showing the extent and characteristics of fluorination by each method. In addition we highlight the diVerences in the characteristics of the three methods. Experimental High-purity Nd2O3, La2O3, SrCO3, CaCO3, and CuO were Fig. 1 Structure of La2ACu2O6 (A=Ca, Sr) showing square-pyramidal Cu (spheres: A, Sr) used to prepare the following samples, Nd1.3Sr1.7Cu2O6-y, J.Mater. Chem., 1997, 7(10), 2077–2083 2077for the sample. The fluorine contents of samples containing significant amounts of this impurity should therefore be treated as only approximate, and this is referred to further in the results section. The TISAB solution was prepared as follows; 57 cm3 glacial acetic acid, 58 g of NaCl and 4 g of trans-1,2-diaminocyclohexane- N,N,N,N-tetraacetic acid were dissolved in 500 cm3 of distilled water, and suYcient 5 M NaOH was added to adjust the pH to ca. 5.3. The buVer solution was then made up to 1000 cm3 with distilled water. Copper oxidation states were determined by iodometric titration. Two complementary titrations were performed for each sample.The first titration involved dissolving ca. 0.05 g of the sample in 50 cm3 of distilled water by the addition of dilute (2 M) HCl. The solution was then boiled to ensure complete conversion of any Cu3+/Cu+ to Cu2+, before excess KI was added. The I2 generated was then titrated under N2 against standard (0.02 M) Na2S2O3 solution. This titration determines the total amount of Cu in the sample.The second titration involved dissolving under N2 the same mass of sample Fig. 2 Structure of Nd1.3Sr1.7Cu2O6-y showing square-pyramidal Cu in excess KI solution by addition of dilute HCl. The liberated (spheres: Nd, Sr) I2 was then titrated as before. From the diVerence in the titration values obtained, the copper oxidation state can be to prepare samples with the highest F content.In order to determined. In samples fluorinated by direct solid-state reacminimize SrF2 impurities at higher F levels, the NH4F was tion with CuF2, CuO is present as an impurity, and so this added in stages, e.g. for 10 mol of NH4F, 5 mol was added will influence the titration. The quantity of CuO present is first and reacted at 300–350 °C; the sample was then reground however known (equal to the amount of CuF2 added) and so with a further 5 mol of NH4F and heated as before. this can be corrected for in the determination of the Cu oxidation state of the sample: good agreement between the Cu Fluorination by CuF2 (solid-state reaction) oxidation state calculations was obtained for samples fluori- Anhydrous CuF2 (0–2 mol) was added to the precursor oxides; nated to a similar extent by the direct solid-state reaction the mixture was ground and heated in the range 245–350 °C method and the autoclave method (where the CuF2 is kept in air for 12 h.separate and so no correction is required). Autoclave method (solid–gas reaction with CuF2) Results The drawback of the CuF2 route is the presence in the samples Fluorination of Ln2-xA1+xCu2O6-y (L=La, Nd; A=Ca, Sr) of CuO impurity deriving from the decomposition of CuF2.In order to avoid this problem, a number of experiments have Powder X-ray diVraction suggested significant fluorine incorbeen performed in enclosed vessels with the starting oxides poration for all fluorination methods, with large peak shifts in and CuF2 separated. In this route, the oxides were weighed most cases.The magnitude of the shifts increased with increased out into a Teflon vessel and the required amount of CuF2 was fluorine content. Fig. 3–5 show the X-ray diVraction patterns then placed in a small nickel pot, which was put into the for the starting materials, and the ‘fully’ fluorinated products Teflon vessel, and the lid fitted.The Teflon vessel was then for each method. placed in an autoclave and heated to 245 °C. Similar results were obtained to the solid-state reaction with CuF2, with the (a) Reaction with NH4F. In order to successfully fluorinate exception that no CuO impurities were observed in this case, the double-layer cuprates by this method, a higher temperature as the CuF2 was kept separate.(300–350 °C) was required compared with similar reactions with Sr2-xAxCuO3 (230 °C). Moreover, since the double-layer Characterization methods materials appear to be less moisture sensitive, Sr/CaF2 impurities were only observed for samples with the highest fluorine The resulting products were characterised by powder X-ray diVraction (Cu-Ka1 radiation, Siemens D-5000 diVractometer).levels. Tables 1 and 2 list cell parameters and fluorine contents for a range of samples prepared by this method. The presence Potential superconducting properties were examined using a DC SQUID magnetometer (Cryogenics Model S100). of Sr/CaF2 impurities in the high fluorine content samples means that the calculated fluorine contents for these samples The fluorine contents were determined using a fluoride ion selective electrode.Prior to measurements being made, the should be viewed as only approximate. For La1.9Sr1.1Cu2O6-y and Nd1.3Sr1.7Cu2O6-y there was no electrode was calibrated using freshly prepared solutions containing known concentrations of NaF. The sample solution change in cell symmetry but the cell parameters increased by up to ca. 5% along a and b, and ca. 1% along c after was then prepared as follows: ca. 0.02–0.05 g of sample was dissolved in 5 cm3 of 0.5 M HCl, to which was added 45 cm3 fluorination. Fluorination of La1.9Sr1.1Cu2O6-y proved more interesting. In this system at low fluorine contents the cell was of distilled water followed by 50 cm3 of a pH ca. 5.3 total ionic strength adjustment buVer (TISAB) solution (preparation tetragonal as for the parent compound, while for high fluorine levels the cell became orthorhombic. At intermediate fluorine described below).The fluorine content of the sample was then determined from the electrode reading of the solution using contents both phases were observed (Fig. 6), with the ratio of the phases with low fluorine content (tetragonal, a=b#3.92, the NaF calibration graph.No noticeable residual fluorinating agent (NH4F or CuF2) was present in any of the samples c#20.14 A° ) and high fluorine content (orthorhombic, a#3.86, b#3.90 A ° , c#21.68 A ° ) varying with overall fluorine content; analysed. The major errors in the analysis resulted from any presence of Sr/CaF2 impurities which tended to be observed at high fluorine levels, only the latter phase was observed.In the low fluorine content phase, the largest changes in cell mainly for samples with high fluorine contents. The presence of this impurity will result in a higher apparent fluorine content parameters were along a and b (similar to La1.9Ca1.1Cu2O6-y 2078 J. Mater. Chem., 1997, 7(10), 2077–2083Fig. 4 Powder X-ray diVraction patterns for (a) La1.9Sr1.1Cu2O6.05, (b) Fig. 3 Powder X-ray diVraction patterns for (a) La1.9Ca1.1Cu2O5.95, La1.9Sr1.1Cu2O6.05F1.5 (F2 gas), (c) La1.9Sr1.1Cu2O4.9F2.3 (NH4F) and (b) La1.9Ca1.1Cu2O5.95F0.4 (F2 gas), (c) La1.9Ca1.1Cu2O5.3F1.3 (NH4F) (d) La1.9Sr1.1Cu2O5.45F2 (CuF2, 245 °C autoclave method) and (d) La1.9Ca1.1Cu2O5.5F0.9 (CuF2, 350 °C) and Nd1.3Sr1.7Cu2O6-y) whereas an expansion along c by 8% was observed for the high fluorine content phase, with much lower expansions along a and b (Table 1).If we consider the change in cell parameters from the low to the high fluorine content phase, then we can see that a and b actually contract slightly, with a large expansion along c. Moreover a change from tetragonal to orthorhombic symmetry is observed which may be related to an ordering of O/F and vacancies.It is possible that the cell may be tripled along b similar to that observed for Nd1.3Sr1.7Cu2O6-y, although there is no conclusive evidence for this. Iodometric titrations indicated a negligible increase in the copper oxidation state following fluorination, with the copper oxidation state for all three systems remaining close to 2.0+.Thus fluorination via NH4F appears to be essentially a nonoxidative process, and so probably involves a substitution reaction in which one oxygen is replaced by two fluorine atoms. The reaction with NH4F appears to proceed via the addition of HF, derived from the decomposition reaction NH4F�NH3+HF. Evidence for this is provided by an attempt to fluorinate La1.9Sr1.1Cu2O6-y via a quantitative reaction with a solution of NH4F in H2O in a hydrothermal bomb.Reaction at 100 °C suggested only a minor change in cell parameters, and subsequent heating of the filtered and washed solid at 350 °C resulted in a large change in cell parameters indicating successful fluorine incorporation. Moreover, the filtrate was alkaline, suggesting that the cuprate had incorporated HF, leaving NH3 in solution.Monitoring the reaction more carefully with a pH meter showed that at 60 °C, a fast reaction (total reaction time ca. 10 min) occurred with the pH changing from acidic to alkaline as the cuprate was added to Fig. 5 Powder X-ray diVraction patterns for (a) Nd1.3Sr1.7Cu2O5.65, the aqueous solution of NH4F. A similar experiment performed (b) Nd1.3Sr1.7Cu2O5.65F1.8 (F2 gas), (c) Nd1.3Sr1.7Cu2O4.65F2 (NH4F) and (d) Nd1.3Sr1.7Cu2O4.8F2.8 (CuF2, 245 °C autoclave method) at room temperature, however, showed a pH change to neutral J.Mater. Chem., 1997, 7(10), 2077–2083 2079Table 1 Cell parameters for La1.9Sr1.1Cu2O6.05 (LSC), of fluorine (Table 3), whereas La1.9Sr1.1Cu2O6-y and La1.9Ca1.1Cu2O5.95 (LCC) and Nd1.3Sr1.7Cu2O5.65 (NSC) fluorinated Nd1.3Sr1.7Cu2O6-y showed larger expansions, consistent with using NH4F high fluorine levels.For these phases the largest increase in cell parameters was along the c direction, suggesting the compound mol. NH4F a/A° b/A° c/A° probable incorporation of F between Cu ions in adjacent LCC — 3.828(1) =a 19.410(3) CuO2 layers. Unlike the case of the fluorination of A2CuO3 LCC 1 3.831(1) =a 19.442(6) (A=Ca, Sr, Ba), where fluorination led immediately to a large LCC 2 3.834(1) =a 19.475(5) structural change to the oxide fluoride Sr2CuO2F2+d, with LCC 10 3.895(2) =a 19.56(1) little non-stoichiometry range (d) for F, the fluorination of the LCC 20 3.899(1) =a 19.57(1) Nd1.3Sr1.7Cu2O6-y (L=Eu, Nd) showed a wide fluorine solid LSC — 3.851(1) =a 20.048(6) solution from the oxide endmember to the highest fluorine LSC 0.5 3.896(3) =a 20.12(2) LSC 1 3.922(1) =a 20.141(9) content, the XRD peaks shifting to lower 2h values as the LSC 2 3.929(1) =a 20.177(7) fluorine content was increased.Indeed, broad peaks, or should- LSC 5 3.937(1) =a 20.199(5) ers on peaks were commonly observed, indicating a mixture LSC 20 3.859(2) 3.903(2) 21.68(1) of compositions with slightly diVerent cell parameters in the NSC — 3.767(3) 11.366(9) 20.10(1) sample, due to the inhomogeneity of fluorination.NSC 0.5 3.798(5) 11.49(1) 20.29(4) As in the case of fluorination using NH4F, a distinct two- NSC 1 3.818(4) 11.54(1) 20.35(3) NSC 2 3.847(1) 11.555(6) 20.38(2) phase region (high and low F content phases) was observed NSC 5 3.851(2) 11.56(1) 20.37(2) for La1.9Sr1.1Cu2O6-y at intermediate F levels, the cell param- NSC 10 3.863(3) 11.62(1) 20.31(2) eters for these two phases being vastly diVerent due to a large NSC 20 3.944(3) 11.92(1) 20.10(3) expansion along c for the high F content phase (a=b#3.90, c#20.32 A ° compared to a=b#3.90, c#22.11 A ° ).In this case, however, no change to orthorhombic symmetry was observed Table 2 F contents, Cu oxidation states and suggested final for the high fluorine content phase.The fluorine contents for compositions for selected samples of La1.9Sr1.1Cu2O6.05 (LSC), a range of samples prepared using F2 gas are given in Table 3. La1.9Ca1.1Cu2O5.95 (LCC) and Nd1.3Sr1.7Cu2O5.65 (NSC) fluorinated using NH4F These results show that for the phases with A=Sr, very high flrine contents are achieved, such as 1.8 fluorine atoms per F Cu suggested formula unit after fluorination of Nd1.3Sr1.7Cu2O5.65. mol.content oxidation composition Iodometric titrations showed that the reaction with F2 gas compound NH4F mol-1 state of sample was highly oxidative, as might be expected, with the copper oxidation state increasing with increasing fluorination up to LCC — — 2.0 La1.9Ca1.1Cu2O5.95 LCC 1 0.5 2.0 La1.9Ca1.1Cu2O5.7F0.5 values close to 3.0+ (Table 3).These results, and the fluorine LCC 2 0.7 2.0 La1.9Ca1.1Cu2O5.6F0.7 contents determined, suggest that the fluorination of the LCC 10 1.3 2.0 La1.9Ca1.1Cu2O5.3F1.3 double-layer cuprates by F2 gas proceeds via a simple insertion LSC — — 2.1 La1.9Sr1.1Cu2O6.05 process, in which the anion site vacancies are gradually filled LSC 1 0.6 2.1 La1.9Sr1.1Cu2O5.75F0.6 by fluorine.The fluorine contents and copper oxidation states LSC 20 2.3 2.1 La1.9Sr1.1Cu2O4.9F2.3 determined, however, indicate that fluorine must also occupy NSC — — 2.0 Nd1.3Sr1.7Cu2O5.65 NSC 5 1.2 2.0 Nd1.3Sr1.7Cu2O5.05F1.2 some interstitial sites for high fluorine contents, similar to that NSC 10 1.8 2.0 Nd1.3Sr1.7Cu2O4.75F1.8 observed in Sr2CuO2F2+d, since, for Nd1.3Sr1.7Cu2O5.65F1.8 NSC 20 2.0 2.0 Nd1.3Sr1.7Cu2O4.65F2.0 and La1.9Sr1.1Cu2O6F1.5 there are seven ‘ideal’ anion sites, leaving the remaining 0.45 or 0.5 fluorine atoms to occupy interstitial sites presumably in the Ln/Sr (L=Nd, La) bilayers. The alternative possibility of partial substitution of fluorine for oxygen was not supported by the copper oxidation state determinations.Slightly higher fluorine contents were also observed on prolonged fluorination, demonstrated by larger peak shifts. The Nd1.3Sr1.7Cu2O5.65 sample heated at 205 °C for >2.5 h showed a shoulder at higher d-spacing, which increased with increasing fluorination time, thus indicating the presence of two phases, Nd1.3Sr1.7Cu2O5.65F1.8 and a phase with larger peak shift and therefore presumably higher F content.Unfortunately, at these high fluorination levels, it is extremely diYcult to avoid decomposition to give SrF2, and Fig. 6 Powder X-ray diVraction patterns for La1.9Sr1.1Cu2O6 fluori- it was not possible to isolate the pure phase with the higher F nated with 7.5 moles of NH4F, showing two fluorinated phases: main content.Fluorine analysis of the mixed sample gave an anomalpeaks from the high fluorine content phase are marked* ously high fluorine content, 3.5, presumably due to the presence of the SrF2 impurity, and so the exact fluorine content of this phase is not known. only, suggesting that NH4F itself was adsorbed by the sample, with no decomposition to HF and NH3 occurring at this temperature.(c) CuF2 method (solid-state and autoclave routes). Both the solid-state and the autoclave methods gave similar results, although the latter had the advantage that because the sample (b) F2 gas. In order to achieve successful fluorination, the samples La1.9Ca1.1Cu2O6-y and La1.9Sr1.1Cu2O6-y required a and CuF2 were kept separate, no CuO impurity was observed.Cell parameters, copper oxidation states and F contents are higher temperature (240–250 °C) than Nd1.3Sr1.7Cu2O6-y (205 °C), which may relate to the slightly diVerent structures. given in Tables 4 and 5. At high temperatures (>300 °C) the XRD patterns of the Moreover, at temperatures >205 °C the Nd containing phase decomposed to give SrF2.The fluorine content was observed products were similar to those from reaction with NH4F, whereas at lower temperatures (ca. 245 °C) the reaction to increase with increasing reaction time at these temperatures. Fluorination of La1.9Ca1.1Cu2O6-y resulted in a small appeared to be similar to reaction with F2 gas. Iodometric titrations supported this view, with the copper oxidation state increase in the cell parameters, indicating a low incorporation 2080 J.Mater. Chem., 1997, 7(10), 2077–2083Table 3 Cell parameters, Cu oxidation states and F contents (x) for La1.9Sr1.1Cu2O6.05Fx (LSC), La1.9Ca1.1Cu2O5.95Fx (LCC) and Nd1.3Sr1.7Cu2O5.65Fx (NSC) prepared using F2 Cu reaction reaction oxidation compound temp./°C time/min a/A ° b/A ° c/A ° state x LCC — — 3.828(1) =a 19.410(3) 2.0 — LCC 250 15 3.834(2) =a 19.475(5) 2.1 0.2 LCC 250 50 3.839(1) =a 19.506(4) 2.15 0.4 LSC — — 3.851(1) =a 20.048(6) 2.1 — LSC 220 30 3.853(1) =a 20.05(1) 2.1 0.1 LSC 220 60 3.876(3) =a 20.18(4) 2.25 0.4 LSC 240 15 3.898(1) =a 20.32(2) 2.3 0.6 LSC 240 60 3.893(1) =a 21.95(1) 2.9 1.5 NSC — — 3.767(3) 11.366(9) 20.10(1) 2.0 — NSC 205 20 3.787(4) 11.399(6) 20.18(1) 2.15 0.2 NSC 205 60 3.842(3) 11.544(7) 20.54(2) 2.3 0.8 NSC 205 120 3.837(1) 11.600(4) 21.04(2) 2.6 1.3 NSC 205 150 3.851(2) 11.679(5) 21.37(2) 2.9 1.8 Table 4 Cell parameters for La1.9Sr1.1Cu2O6.05 (LSC), achieved by this method was, however, lower than that La1.9Ca1.1Cu2O5.95 (LCC) and Nd1.3Sr1.7Cu2O5.65 (NSC) fluorinated obtained using F2 gas (2.5+ compared to 3.0+), suggesting using CuF2 that although highly oxidative, the oxidising power of the reaction with MF2 is not as great as for the reaction with F2 mol.reaction gas. Copper oxidation states and fluorine contents for selected compound CuF2 temp./°C a/A ° b/A ° c/A ° samples are listed in Table 5. Taken together, the copper LCC — — 3.828(1) =a 19.410(3) oxidation states and fluorine contents indicate that at high LCC 0.5 245 3.827(1) =a 19.427(5) temperatures the reaction proceeds via substitution (2FO10), LCC 1 245 3.834(1) =a 19.51(1) whereas at low temperatures the reaction is probably a mixture LCC 2 245 3.843(3) =a 19.58(2) of insertion (as in the case of F2) and substitution.Suggested LCC 1 350 3.860(2) =a 19.50(1) final compositions of the samples after fluorination were LSC — — 3.851(1) =a 20.048(6) determined from the fluorine contents and copper oxidation LSC 0.25 245 3.898(3) =a 20.18(2) LSC 0.5a 245 3.902(2) =a 20.14(2) states and are reported in Table 5.The slight diVerence in the LSC 1b 245 3.887(7) =a 21.99(6) MF2 (low temperature) and F2 reaction routes is demonstrated LSC 1.5 245 3.896(5) =a 22.09(2) by the fact that for Nd1.3Sr1.7Cu2O5.65, the orthorhombic LSC 2 245 3.93(1) =a 22.2(1) splitting at the highest fluorine content, designated by the ratio LSC 1.5 300 3.864(2) 3.929(2) 21.53(2) a : b/3, was significantly larger for samples derived from the LSC 1.5 350 3.856(1) 3.911(1) 21.45(1) MF2 route (ca. 1.025 compared to ca. 1.01). NSC — — 3.767(3) 11.366(9) 20.10(1) In all three methods, the fluorine content could be varied NSC 0.25 245 3.817(1) 11.455(1) 20.33(1) NSC 0.5 245 3.832(3) 11.55(1) 20.50(2) by suitable control, e.g.amount of CuF2/NH4F or temperature NSC 1.25b 245 3.804(1) 11.690(4) 21.70(2) and time for F2 gas. For the CuF2 and F2 methods, the copper NSC 1.5 245 3.798(1) 11.699(2) 21.91(1) oxidation state can also be controlled between 2+ and 3.0+, NSC 1.5 270 3.823(1) 11.746(4) 21.75(2) such that it is quite readily possible to prepare samples with NSC 1.5 300 3.937(1) 11.800(2) 19.96(1) copper oxidation states in the region 2.2–2.3+, which should NSC 1.5 350 3.884(7) 11.81(1) 19.89(2) be optimum for superconductivity.aSample consisted of two phases: the cell parameters for the lower F All samples were examined for possible superconductivity.content (most abundant) phase are given. bSample consisted of two For the sample La1.9Sr1.1Cu2O6-y, a very weak superconphases: the cell parameters for the higher F content (most abundant) ducting signal (ca. 0.1% volume fraction), Tc ca. 20 K, was phase are given. observed after light fluorination using CuF2 (ca. 0.1–0.2 moles). Under such conditions, samples showed broad X-ray peaks increasing at low temperatures, while remaining close to 2.0+ consistent with a mixture of fluorine contents.Further work has failed to increase the superconducting fraction, or increase at higher temperatures. The maximum copper oxidation state Table 5 F contents, Cu oxidation states and suggested final compositions for selected samples of La1.9Sr1.1Cu2O6.05 (LSC), La1.9Ca1.1Cu2O5.95 (LCC) and Nd1.3Sr1.7Cu2O5.65 (NSC) fluorinated using CuF2 mol.reaction F content Cu oxidation suggested composition compound CuF2 temp./°C mol-1 state of sample LCC — — — 2.0 La1.9Ca1.1Cu2O5.95 LCC 2 245 0.5 2.1 La1.9Ca1.1Cu2O5.8F0.5 LCC 1 350 0.9 2.0 La1.9Ca1.1Cu2O5.5F0.9 LSC — — — 2.1 La1.9Sr1.1Cu2O6.05 LSC 0.25 245 0.4 2.25 La1.9Sr1.1Cu2O6.0F0.4 LSC 1.5 245 2.0 2.5 La1.9Sr1.1Cu2O5.45F2.0 LSC 1.5 300 1.8 2.15 La1.9Sr1.1Cu2O5.2F1.8 LSC 1.5 350 1.7 2.1 La1.9Sr1.1Cu2O5.2F1.7 NSC — — — 2.0 Nd1.3Sr1.7Cu2O5.65 NSC 0.5 245 0.7 2.2 Nd1.3Sr1.7Cu2O5.5F0.7 NSC 1.5 245 2.8 2.55 Nd1.3Sr1.7Cu2O4.8F2.8 NSC 1.5 270 2.5 2.35 Nd1.3Sr1.7Cu2O4.75F2.5 NSC 1.5 300 2.2 2.1 Nd1.3Sr1.7Cu2O4.65F2.2 NSC 1.5 350 1.8 2.1 Nd1.3Sr1.7Cu2O4.85F1.8 J.Mater. Chem., 1997, 7(10), 2077–2083 2081Tc.No evidence for superconductivity was found for any atom by two fluorine atoms in La1.9Sr1.1Cu2O6), with octahedral copper similar to the situation in Sr2CuO2F2+d . This is other sample. supported by fluorine analysis (Table 2) although the presence of SrF2 impurities at the higher fluorine levels does cast some Discussion doubts over the absolute reliability of the values.Such a compound, with all the anion sites filled, might be expected to The results clearly demonstrate that the Ruddlesden–Popper type cuprates, Ln2-xA1+xCu2O6-y (L=Nd, La, Eu; A=Sr, be very stable, and this perhaps explains the two-phase region. The reason why fluorination of Nd1.3Sr1.7Cu2O6-y with NH4F Ca) can incorporate significant fluorine levels.This system is interesting because it demonstrates some diVerent character- does not form a similar phase may be related in some way to the slightly diVerent structure of the parent oxide or to the istics of each of the three fluorination routes leading to large diVerences in the final fluorinated products. In particular, fact that the starting oxygen content is lower ( y#0.35) so that an anion content (O+F) of 7.0 is not achieved after fluorination reaction with F2 gas appears to proceed via fluorine insertion, while reaction with NH4F appears to involve fluorine substi- of this phase.One problem with this postulation is that the fully fluorinated La1.9Sr1.1Cu2O6-y is orthorhombic, whereas tution for oxygen. The reaction with CuF2 at low temperatures (ca. 245 °C) probably involves both insertion and substitution the proposed compound La1.9Sr1.1Cu2O5F2 might be expected to be tetragonal. This indicates some sort of ordering must be mechanisms, while at higher temperatures (>300 °C) the substitution process dominates, and the copper oxidation state present, which may be related to O/F ordering, perhaps in the apical sites (if we assume that fluorine occupies only the apical remains at 2.0+.In addition, it has been found that heating samples prepared by the F2 gas route to temperatures >300 °C sites, then the composition of the apical positions consists of one oxygen plus two fluorine atoms). Alternatively it may be also results in a similar XRD pattern to reaction with NH4F. Thus the high copper oxidation states are only stable at low- related to a small excess of interstitial fluorine, which is suggested by the fluorine analysis.temperatures (ca. 200–250 °C), and so this fact can be used as a post-synthesis means of control of the fluorine content and The origin of the weak superconducting signal in La1.9Sr1.1Cu2O6-y at 20 K for low fluorine levels is unclear. copper oxidation state.If we consider the reactions with F2 gas and MF2 (low This signal has been seen reproducibly in a number of lightly fluorinated samples, but without any significant increase in temperature), the highest fluorine levels were achieved for A= Sr leading to a large expansion of the lattice parameters along signal strength, and is not present in the parent undoped material.It is possible that this signal may be due to the c (up to 2 A ° ) for such samples, suggesting that fluorine is being inserted into the vacant sites between the CuO2 layers, so that presence of a very small amount of the single copper layer phase La2-xSrxCuO4 impurity, which becomes supercon- the copper coordination changes from square pyramidal to octahedral. The magnitudes of the changes in c are also ducting after fluorination. The lack of bulk superconductivity in these systems may be possibly explained by the partial consistent with the expansion that would be needed for insertion of fluorine between the CuO2 layers to give two CuMF occupation of the anion sites between the CuO2 layers, since the location of isolated fluorine or oxygen atoms in these sites bonds of reasonable length (ca. 1.8 A ° ). There is also evidence that some excess fluorine is introduced in these systems, which would be expected to result in hole trapping. Recently superconducting related phases, Sr2Can-1 must be located in interstitial sites, presumably between the Ln/Sr bilayers, similar to the interstitial fluorine sites between CunO2n+xF2-y (n=2, 3), have been synthesised by a highpressure route with high Tcs (n=2, 99 K; n=3, 111 K).15 These the Sr bilayers in Sr2CuO2F2+d .4 A further interesting point is the distinct two-phase character observed for results support the view that it should be possible to induce superconductivity in the systems studied by suitable synthesis La1.9Sr1.1Cu2O6-y on fluorination, suggesting the stability of the high fluorine content phase.In the case of A=Ca, lower control. In this respect, if we assume that the occurrence of superconductivity is quenched by hole trapping by localised fluorine levels are observed, and the expansion along c is small (ca. 0.17 A ° ), suggesting that very little, if any, fluorine is now oxygen or fluorine atoms linking the CuO2 layers, then in principle, complete occupation of these anion sites should located between the CuO2 layers, with the fluorine possibly being located in the interstitial sites instead.eliminate such hole trapping. However, such a situation ultimately results in copper oxidation states which are too high These results can be explained by considering the size of the cation separating the two CuO2 layers. For phases with A= for superconductivity, e.g.for La1.9Sr1.1Cu2O6F the copper oxidation state would be 2.55+. It may therefore be possible Sr, there is a mixture of rare earth, Ln, and Sr separating the CuO2 layers, whereas for phases containing Ca, it is the Ca to use a combination of fluorination methods, involving the non-oxidative fluorination of the sample first with NH4F, that occupies these sites.13 Thus for the former systems the larger size of Sr compared to Ca (ionic radii for 8-coordination, which we believe leads to the replacement of one oxygen by two fluorine atoms and a consequent increase in the occupation Sr2+ 1.26 A ° , Ca 1.12 A ° )14 means that the separation of the layers is larger (e.g.La2SrCu2O6 ca. 3.7 A° , La2CaCu2O6 of the anion sites between the CuO2 layers.Reaction with F2 or MF2 could then be used to fill the remaining sites and ca. 3.3 A ° )13 and so fluorine can be inserted much more readily into the vacant sites between the layers. provide the necessary Cu oxidation. Alternatively, we have suggested that the high fluorine content phase produced by An interesting point is the fact that for the reactions with NH4F and CuF2 (high temperature) the samples the reaction of La1.9Sr1.1Cu2O6 with NH4F consists of replacement of one oxygen atom by two fluorine atoms to give a La1.9Sr1.1Cu2O6-y and Nd1.3Sr1.7Cu2O6-y showed diVerent behaviour, whereas for the other methods they behaved simi- stable system La1.9Sr1.1Cu2O5F2, where all the anion sites are filled, i.e.larily. For the former, a large increase in c was observed along with a distinct two-phase region similar to that observed for La1.9Sr1.1Cu2O6+2 F�La1.9Sr1.1Cu2O5F2+1 O reaction with F2 and MF2 (low temperature). In contrast, for both La1.9Ca1.1Cu2O6-y and Nd1.3Sr1.7Cu2O6-y the major (Cu oxidation state=2.05) expansion was along a and b, with little change along c.The Fluorine could then be inserted into the ierstitial sites of this origin of this diVerence is unclear.Further work, includphase to raise the Cu oxidation state to the optimum level by ing powder neutron diVraction studies and Eu Mo�ssbauer the oxidative fluorination with F2 gas or CuF2, i.e. experiments on Eu1.3Sr1.7Cu2O6-y (isostructural with Nd1.3Sr1.7Cu2O6-y), is planned to help rationalise this result. La1.9Sr1.1Cu2O5F2+0.5 F�La1.9Sr1.1Cu2O5F2.5 We suggest that the high fluorine content phase has a composition La1.9Sr1.1Cu2O5F2 (i.e.the substitution of one oxygen (Cu oxidation state=2.3) 2082 J. Mater. Chem., 1997, 7(10), 2077–2083Such two-stage fluorination processes could be used for a of superconductivity, these results further demonstrate that low-temperature fluorination is a powerful tool for the incor- range of cuprate systems to vary the O/F ratio, while achieving a copper oxidation state (ca. 2.3+) suitable for super- poration of fluorine and control of copper oxidation states in cuprates. Recent preliminary results have shown that these conductivity. Neutron diVraction studies are planned to try to confirm fluorination routes can also be applied to non-cuprate systems.these conclusions, particularly relating to the location of the We thank the EPSRC for financial support. fluorine atoms. In addition Eu Mo�ssbauer studies of the Eu1.3Sr1.7Cu2O5.65 system are planned to examine the variation in the Eu environment with increasing F content. With respect References to this system, and the related Nd-containing system, 1 M. Al-Mamouri, P.P. Edwards, C. Greaves and M. Slaski, Nature Nd1.3Sr1.7Cu2O5.65, one might expect that fluorine would (L ondon), 1994, 369, 382. initially insert along the one-dimensional channels in 2 M. Al-Mamouri, P. P. Edwards, C. Greaves, P. R. Slater and the structure (Fig. 2). Then after these sites have been filled, M. Slaski, J.Mater. Chem., 1995, 5, 913. the fluorine would presumably occupy the sites between the 3 P.R. Slater, P. P. Edwards, C. Greaves, I. Gameson, J. P. Hodges, CuO2 sheets. M. G. Francesconi, M. Al-Mamouri and M. Slaski, Physica C, 1995, 241, 151. Despite the lack of superconductivity, these results taken 4 P. R. Slater, J. P. Hodges, M. G. Francesconi, P. P. Edwards, together with previous studies on A2CuO3 (A=Ca, Sr, Ba), C. Greaves, I.Gameson and M. Slaski, Physica C, 1995, 253, 16. show that low-temperature fluorination is a powerful tool for 5 Z. Hiroi, M. Takano, M. Azuma and Y. Takeda, Nature (L ondon), the incorporation of fluorine in samples of this type. Moreover, 1993, 364, 315. these results on the double-layer cuprates show a clear diVer- 6 N. Nguyen, L. Er-rakho, C. Michel, J. Choisnet and B. Raveau, ence between the nature of the fluorination processes. Namely Mater. Res. Bull., 1980, 15, 891. 7 C. Michel and B. Raveau, Rev. Chim.Miner., 1984, 21, 407. the fluorination with NH4F appears to be essentially a 8 N. Nguyen, C. Michel, F. Studer and B. Raveau, Mater. Chem., non-oxidative substitution reaction, although the fact that 1982, 7, 413. superconducting Sr2CuO2F2+d can be prepared from Sr2CuO3 9 N. Nguyen, J. Choisnet and B. Raveau, Mater. Res. Bull., 1982, suggests that some oxidation, probably aerial oxidation, can 17, 567. occur.3 In contrast, the fluorination with F2 gas appears to 10 R. J. Cava, B. Batlogg, R. B. van Dover, J. J. Krajewski, proceed via an oxidative insertion reaction in this system, while J. V. Waszczak, R. M. Fleming, W. F. Peck Jr., L. W. Rupp Jr., P. Marsh, A. C. W. P. James and L. F. Schneemeyer, Nature reaction with MF2 (low temperature) involves both substi- (L ondon), 1990, 345, 602. tution and insertion. The reaction with Sr2CuO3 to give 11 K. Kinoshita, H. Shibata and T. Yamada, Physica C, 1991, 176, Sr2CuO2F2+d shows that the fluorination by F2 gas can also 433. involve partial substitition in addition to oxidative insertion. 12 R. Mahesh, R. Vijayaraghavan and C. N. R. Rao,Mater. Res. Bull., 1994, 29, 303. 13 R. C. Lobo Ph.D. Thesis, University of Birmingham, 1990. Summary 14 R. D. Shannon, Acta. Crystallogr., Sect. A, 1976, 32, 751. 15 T. Kawashima, Y. Matsui and E. Takayama-Muromachi, Physica The results clearly demonstrate that large amounts of fluorine C, 1996, 256, 313. can be incorporated into the Ruddlesden–Popper type cuprates Ln2-xA1+xCu2O6-y (L=La, Nd; A=Ca, Sr). Despite the lack Paper 7/03735J; Received 29th May, 1997 J. Mater. Chem., 1997, 7(10), 2077–208

ISSN:0959-9428    

DOI:10.1039/a703735j

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

An experimental study of copper self-diVusion in CuO, Y2Cu2O5 and YBa2Cu3O7-x by secondary neutral mass spectrometry Jan A. Rebane,a Nikolay V. Yakovlev,a Dmitry S. Chicherin,a Yuri.D. Tretyakov,a Lidia I. Leonyukb and Valery G. Yakuninc aChemistry Department,Moscow State University, 119899 Moscow, Russia bGeology Department,Moscow State University, 119899 Moscow, Russia cPhysics Department, Moscow State University, 119899Moscow, Russia Copper self-diVusion has been studied in CuO, Y2Cu2O5 and YBa2Cu3O7-x ceramics by the SNMS (secondary neutral mass spectrometry) technique in the temperature range 700–900 °C with the stable isotope 63Cu used as a tracer.The lowest diVusion rate and the highest value of activation energy were found for the Y2Cu2O5 phase. Copper self-diVusion was also studied along the c axis of YBa2Cu3O7-x single crystals.It was shown that the anisotropy of Cu diVusion in YBa2Cu3O7-x is rather high. The diVusion rate in the c direction of single crystals is more than two orders of magnitude lower than that in ceramic samples. Since the discovery of high-temperature superconductivity a surface of polished ceramic substrates.The roughness of the number of experimental studies have been concerned with substrates before the deposition was between 0.1 and 0.2 mm mass transport in superconducting oxides. The most common as a determined by the profilometer Talystep (Rank–Taylor– method applied involved a study of the diVusion of radioactive Hobson). The films were deposited by RF sputtering using tracers and the use of diVerent sectioning methods or SIMS thick-film 63CuO targets.The ratio of copper isotopes (secondary ion mass spectrometry) for the determination of 63Cu/65Cu in the films was 95/5. The ratio of copper isotopes depth profiles. In this work we tried to apply a relatively new 63Cu/65Cu in the substrates was 69/31 (the natural ratio). method of depth profile analysis, plasma-SNMS, that has The ‘film’ copper with an isotope ratio 63Cu/65Cu=95/5 was emerged as a result of the development of SIMS.The former used as a tracer in this work. has two advantages in comparison to SIMS. First, a very low To study the anisotropy of copper diVusion in YBa2Cu3O7-x energy of primary ions (several hundred eV) allows the per- the same films were deposited on the (001) surface of formance of depth profiling with very high depth resolution, YBa2Cu3O7-x single crystals (the size of the crystals was ca.which is especially valuable for the analysis of mass transport 3×3×0.1 mm). For YBa2Cu3O7-x all diVusion annealings in thin films and heterostructures. Secondly, relatively easy were performed in an oxygen flow, and for Y2Cu2O5 and CuO quantification of the data is possible, as SNMS is less matrix- in air.dependent than SIMS. Depth profile analyses of the diVusion couples were performed by the SNMS technique in direct bombardment mode (DBM) using an INA-3 set-up (Leybold AG). In this exper- Experimental imental mode the low pressure Kr (Ar) plasma in the analysis chamber of the spectrometer is supported by electron-cyclotron Polycrystalline Y2Cu2O5 was synthesized by the conventional resonance.The sample is situated inside the plasma and ceramic process starting from Y2O3 and CuO. Y2Cu2O5 and separated from it by a grounded aperture. The primary ions CuO ‘low density’ ceramic substrates were prepared by convenare extracted from the plasma volume by a negative voltage tional sintering at 1050 and 950 °C, respectively.The final applied to the sample surface (600 V in our case). These ions density of these substrates was around 90% of the theoretical are used for the sputtering of the sample. Neutral particles values for Y2Cu2O5 and CuO. The average grain size was ca. generated in the course of the sputtering have to travel 10 mm for the both specimens. approximately 5 cm through the plasma to the entrance of the High-density CuO substrates were prepared by hot pressing ion optics.This distance is long enough for eVective post- (pressure 4 GPa, temperature 700 °C, duration 60 min). The ionization by plasma electrons. If plasma parameters, sample density reached in this case was 97%. It was not possible to holder assembly and the extraction voltage applied to the estimate the grain size from SEM photographs. The density of sample are properly adjusted, very high depth resolution can substrates was determined by picnometric weighing in CBr3H.be achieved (up to 2–5 nm).1 The uncertainty of this procedure was determined as ±1% Some problems related to sample charging arise in the (absolute) by using Ge single crystals as a reference material.course of the analysis of insulating materials (Y2Cu2O5 in our CuO substrates were pre-annealed in an oxygen flow prior case). The depth profiling of 63CuO/Y2Cu2O5 diVusion couples to polishing to avoid possible reduction of copper. was performed with a Ni grid (grid period=50 mm) pressed The single crystals of the YBa2Cu3O7-x phase were grown onto the sample surface by a mask.The profilometric measure- from a melt of composition 3525572% (YO1.5–BaCO3–CuO). ments confirmed the rectangular shape of the small craters The mixture was heated in an alumina crucible to 1000 °C and restricted by conductive stripes of the grid and the uniform cooled to room temperature at the rate of 1–5 °C h-1. Single depth of these craters over the ion bombarded area (typically crystals were separated by breaking apart the crucible. 1–2 mm in diameter). EPMA (electron probe microanalysis) was performed with The following working parameters of SNMS analysis were an accuracy better than 1%, using a CAMECA analyzer and used: RF power=150 W; Kr (Ar) pressure=3.0×10-3 mbar; showed that the crystals had the stoichiometric composition. Helmholtz coil current=4.9 A and accelerating potential for The diVusion couples were prepared by deposition of thin films of CuO enriched with the stable isotope 63Cu onto the the primary ions=600 V.J. Mater. Chem., 1997, 7(10), 2085–2089 2085Discussion The experimental procedures employed in this work are quite diVerent from the traditional and very well developed methods of diVusion experiments such as techniques using radioactive tracers and serial sectioning for depth profiling.2,3 The first problem related to the application of thin film diVusion couples is that the thin film, which actually is the source of the tracer, constitutes approximately a quarter of the whole diVusion profile.Moreover it would be advisable to perform some correction for diVerences in sputtering rates between substrates and films.However, our preliminary studies showed that the sputtering rates for CuO, Y2Cu2O5 and YBa2Cu3O7-x do not diVer by more than 25–30%. So the average sputtering rate was applied for the transformation of sputtering time into depth. Fig. 1 Experimental points and simulated curve for ‘film’ copper The other problem that arises in the case of thin film diVusion profile in a CuO/CuO diVusion couple (‘low density’ sub- diVusion couples is related to the grain growth.The temperastrate) annealed for 20 min at 800 °C ture of CuO deposition was relatively low (400 °C) and insuYcient to allow perfectly crystalline material. Intensive Table 1 Copper diVusion coeYcients in polycrystal CuO samples grain growth during the first few minutes of a diVusion annealing under elevated temperature changes the surface T/°C; t /min D/cm2 s-1 D/cm2 s-1 morphology and the resultant broadening of transitions of diVusion experiment high density (97%) low density (90%) between layers in an experimental SNMS depth profile is 850; 10 (8.4±0.2)×10-14 caused not only by mass transport but also by recrystallization. 800; 10 (2.3±0.7)×10-14 (4.8±0.4)×10-14 It is almost impossible to separate these two processes. 800; 20 (4.0±0.2)×10-14 Taking this eVect into account all diVusion couples were 750; 15 (1.5±0.1)×10-14 pre-annealed at 800 °C for 1–3 min. The depth profiles of 750; 30 (1.7±0.08)×10-14 copper isotopes in these pre-annealed samples were used as 700; 15 (3.6±0.8)×10-15 the starting point for further calculations. 700; 30 (3.6±0.8)×10-15 (2.8±0.2)×10-15 700; 60 (5.8±0.2)×10-15 The depth profiles of ‘film’ copper in the starting samples 650; 30 (1.6±0.1)×10-15 before diVusion annealings were not step-like. To correct this 650; 60 (2.3±0.1)×10-15 the ‘additional time’ was calculated in the following way. In 600; 30 (4.0±0.4)×10-16 the first stage the copper diVusion coeYcient was calculated 600; 60 (4.4±0.4)×10-16 assuming that the initial profile was step-like.Then the calculated diVusion coeYcient was used to estimate the time that would be necessary to allow the development of the real initial The second series of diVusion experiments was performed profile starting from an ideal one. In the next stage the sum with ‘high-density’ substrates (97% of the theoretical value) of ‘additional time’ and the real time of a diVusion annealing prepared by hot-pressing.In this case it was not possible to was applied for the calculation of the copper diVusion simulate the whole diVusion profile by a simple error function. coeYcient. As a rule after four to seven iterations both A combined fitting function with an additional term corre- ‘additional time’ and the diVusion coeYcient stop changing.sponding to grain-boundary diVusion was applied in this case The final diVusion coeYcients were on average two to three [eqn. (2)] times lower than the ones calculated under the assumption of C(x,t)=A{1-erf [(x-x¾)/(4Dt)1/2]}+C exp[-B(x-x¾)6/5] an ideal shape of the initial diVusion profile. (2) In this experimental series the thickness of 63CuO films was ca. 0.3 mm and as a consequence the surface concentration of Results the tracer decreased in the course of the diVusion annealings. An experimental study of copper self-diVusion in CuO Experimental diVusion profiles at 700 °C of ‘film‘ copper are shown in Fig. 2. It is clearly seen that the concentration in the The first series of experiments was performed with ‘low density’ films is uniform but decreases with increasing diVusion time.CuO polycrystal substrates. The average grain size for the This means that diVusion in the film is much quicker than in substrate material was ca. 10mm and the thickness of the the substrate, probably due to higher defect concentration. 63CuO film was ca. 1 mm. The diVusion profiles were simulated This uniform decrease of the tracer concentration in the film using the error function solution of the Fick’s second law means that the Gaussian term for volume diVusion is more equation for the case of contact of two semi-infinite bodies appropriate for the description of volume diVusion than the where C(x,t)=concentration of the tracer, D=copper diVusion error function term.However, application of eqn.(3) does not coeYcient and x¾=the thickness of the CuO film. change the calculated values of the copper diVusion coeYcients by more than a factor of two. C(x,t)=A{1-erf [(x-x¾)/(4Dt)1/2]} (1) C(x,t)=A exp[-(x-x¾)2/4Dt]+C exp[-B(x-x¾)6/5] An experimental profile in this diVusion couple at 800 °C (3) and annealing time of 20 min is shown in Fig. 1. The parameters A, x¾ and D were chosen to minimize the value of x2, the The temperature dependences of copper self-diVusion measure of the goodness of the fit.x2 values were in the range coeYcients for both series of experiments are compared in 0.5–2. The position of the film/substrate boundary was deter- Fig. 3 and Table 1 and the data are fitted to an Arrenius mined by the position of the 50% point of the ‘film’ copper equation [eqn.(4)] diVusion profile. Calculated values of copper self-diVusion D=D0 exp(Ea/RT) (4) coeYcients are given in Table 1 together with the errors determined in the course of fitting. where Ea is the activation energy and D0 is the intercept. 2086 J. Mater. Chem., 1997, 7(10), 2085–2089Fig. 4 The temperature dependence of the copper self-diVusion coeYcient in Y2Cu2O5 (Ea=415±20 kJ mol-1) Fig. 2 Experimental SNMS profiles of ‘film’ copper in a CuO/CuO diVusion couple (‘high density’ substrate). Filled circles, before annealing; open circles, after 15 min annealing at 700 °C; triangles, Table 2 Copper diVusion coeYcients in a polycrystal Y2Cu2O5 sample after 30 min annealing at 700 °C. temp./°C; time of diVusion experiment D/cm2 s-1 850; 10 min (1.4±0.1)×10-13 850; 30 min (7.7±0.6)×10-14 800; 190 min (1.5±0.2)×10-14 750; 21 h (1.0±0.2)×10-15 700; 76 h (1.4±0.2)×10-16 presented in Fig. 4 and Table 2. The activation energy of copper diVusion in this phase is almost twice as large as that in YBa2Cu3O7-x and even more than twice the value for CuO. At the same time the absolute values of diVusion coeYcients are the lowest for the temperature range studied.The structure of Y2Cu2O5 can be considered as a stack of CuMO layers parallel to the ab-plane, separated by layers of YO6 octahedra. Copper atoms have distorted squarepyramidal coordination (CuMO distance around 2 A° ) with a fifth oxygen atom at a distance of 2.8 A ° .7,8 The copper diVusion mechanism in Y2Cu2O5 is probably rather complex.First, there are no data in the literature on any cation or oxygen non-stoichiometry of this phase. As a Fig. 3 The temperature dependence of the copper self-diVusion consequence the concentration of vacancies might be much coeYcient in CuO determined for ‘low’ ($; Ea=162±7 kJ mol-1) and ‘high density’ (#; Ea=150±15 kJ mol-1) substrates lower than that in CuO or YBa2Cu3O7-x.Secondly, the value of intercept (Table 3), which is proportional to the entropy of diVusion, is too high even for an interstitial mechanism. The Both absolute values of diVusion coeYcients and activation value of intercept for interstitial diVusion in metals is in the energies are in good agreement. range 10-3–1 cm2 s-1.9 The literature data shows that CuO is non-stoichiometric, most probably due to cation vacancies4 as in other divalent Experimental study of anisotropy of copper diVusion in oxides of the elements of the first transition series such as NiO, YBa2Cu3O7-x CoO and FeO.Hence the very low value of the intercept (5×10-6–5×10-7 cm2 s-1) for the temperature dependence The last set of experiments was performed with YBa2Cu3O7-x of the copper self-diVusion coeYcient can be considered as a single crystals.consequence of a vacancy diVusion mechanism in CuO. This mechanism of cation diVusion has also been established for Table 3 Activation energies and pre-exponentials of the temperature NiO and CoO.5 However the values of the intercept for NiO dependences of copper self-diVusion coeYcients in CuO, YBa2Cu3O7-x and CoO (5×10-2 and 5×10-3 cm2 s-1, respectively)6 are and Y2Cu2O5 much higher than that found for CuO.Such a diVerence can sample Ea/kJ mol-1 log10 (D0/cm2 s-1) be caused by diVerent types of crystal structures: the rocksalt structure of NiO and CoO diVers from the structure of CuO. CuO (97%) 150±15 -6.1±0.9 CuO (90%) 162±7 -5.5±0.4 Copper self-diVusion in Y2Cu2O5 polycrystals YBa2Cu3O7-x (polycrystal) 240±9 -0.65±0.4 YBa2Cu3O7-x (crystal c axis) 280±30 -1.3±1.5 The same experimental procedures were applied to study of Y2Cu2O5 415±20 6.3±0.9 copper self-diVusion in Y2Cu2O5 and the data obtained are J.Mater. Chem., 1997, 7(10), 2085–2089 2087Fig. 5 The temperature dependence of the copper self-diVusion coeYcient in YBa2Cu3O7-x. Filled circles, literature data for YBa2Cu3O7-x polycrystals (Ea=256±4 kJ mol-1); open circles, our data for ceramic YBa2Cu3O7-x samples (Ea=240±9 kJ mol-1); triangles, copper self-diVusion coeYcients along the c axis of Fig. 6 The structure of YBa2Cu3O7-x YBa2Cu3O7-x single crystals (Ea=280±35 kJ mol-1). The data on copper diVusion in YBa2Cu3O7-x found in the Table 5 The anisotropy of cation diVusion in YBa2Cu3O7-x literature10 and obtained in the course of our studies11 are cation Dpoly/Dc ref.shown in Fig. 5. The good agreement between our data and literature data for polycrystalline samples gives us faith in our Ba 1000 14 experimental procedures. The new data on copper diVusion Cu 200–400 this work along the c axis are also presented in Table 4. Ni 50–60 10,13 The anisotropy of copper self-diVusion in this phase is Co 1000 10 surprisingly large.The ratio Dpoly/Dc is estimated to be 200–400 (Dc=cation diVusion coeYcient along the c axis). This result can be explained in the following way. The layers of square diVusion.14 For Ba, Dpoly/Dc is at least 3000–4000, i.e. one pyramids CuO5, separated by yttrium ions, are presumed to order of magnitude higher. The authors of ref. 14 emphasise be rigid. No additional ordering or disordering at elevated that this value represents a lower limit for the anisotropy of temperatures was observed in these layers. On the other hand, Ba diVusion as the measurements were performed on melt- O(1) atoms are very mobile and statistically distributed textured ceramic samples rather than single crystals. The actual between O(1) and O(5) positions (Fig. 6) under the conditions value of the anisotropy could be even larger as there is no of the diVusion experiments. This is confirmed by very high continuous chain of Ba positions in YBa2Cu3O7-x along the anisotropy of oxygen diVusion in YBa2Cu3O7-x (four to six c axis and diVusing in this direction Ba has to substitute for Y. orders of magnitude).12 The anisotropy of Co diVusion measured by Routbort et al.10 Gupta et al.13 suggested that the most probable mechanism is of the same order of magnitude as that for Ba.He suggested of copper diVusion in YBa2Cu3O7-x is via vacancies. The that the reason for this is that Co preferentially substitutes negatively charged copper vacancies that inevitably exist in only for copper in the position Cu(1), as confirmed by neutron the lattice will attract highly mobile and positively charged diVraction.As a consequence the jump distance along c axis oxygen vacancies in the Cu(1)MO(1) layer. As a consequence is at least two times longer than the ab plane. the energy barrier for a copper atom to jump onto an adjacent vacant Cu(1) position will be significantly lower than that for Conclusions a jump onto a Cu(2) vacancy.The available data on the anisotropy of cation diVusion in It is shown that the combination of stable isotope tracers, thin YBa2Cu3O7-x are summarised in Table 5. film diVusion couples and secondary neutral mass spectrometry In spite of large absolute value, the anisotropy of copper can be successfully used for an experimental study of cation diVusion is lower than the anisotropy measured for Ba diVusion in complex oxide materials.Copper self-diVusion has been studied for CuO, Y2Cu2O5 Table 4 Copper diVusion coeYcients along the c axis of YBa2Cu3O7-x and YBa2Cu3O7-x ceramic samples. The lowest diVusion rate single crystals and highest values of the intercept and activation energy were found for the Y2Cu2O5 phase.temp./°C; t /h Self-diVusion in CuO most probably involves a vacancy of diVusion experiment Da/cm2 s-1 mechanism, because the non-stoichiometry in this compound 925; 4 (2.3±0.3)×10-14 is accommodated by vacancies in the cation sublattice exactly 900; 4 (1.8±0.2)×10-14 as found in NiO, CoO and FeO. This assumption was con- 850; 5 (7.7±0.5)×10-15 firmed by low values of the activation energy and the intercept. 800; 10 (6.5±0.7)×10-16 The anisotropy of copper diVusion in YBa2Cu3O7-x was 750; 17 (2.1±0.3)×10-16 found to be in the range 200–400: approximately one order of 750; 36 (3.5±0.3)×10-16 magnitude smaller than that for Ba and Co.Such a high value for anisotropy of copper self-diVusion may be related to the aTwo exponential functions. 2088 J.Mater. Chem., 1997, 7(10), 2085–20897 J. L. Garcia-Munoz and J. Rodriguez-Carvayal, J. Solid State very high mobility of oxygen vacancies in the Cu(1)MO(1) Chem., 1995, 115, 324. layer of YBa2Cu3O7-x structure. 8 R. D. Adams, Ja. A. Estrada and T. Datta, J. Supercond, 1992, 5, 33–38. This research was partly supported by RFBR grant N 9 W. K.Warburton and D. Turnbull, in DiVusion in solids—recent 96–03–33027. developments, ed. A. S. Nowick and J. J. Burton, Academic, New York, 1975, p.180. 10 J. L.Routbort, S. J.Rothman, Nan Chen and J. N. Mundy, Phys. References Rev. B., 1991, 43, 5489. 11 A. Strelkov, J. Rebane and Y. Metlin, J. Mater. Chem., 1993, 3, 1 K.-H.Muller and H. Oechsner,Mikrochim. Acta, 1983, 10, 51. 735. 2 K. HauVe, Reaktionen in und an Festen StoVen, Springer-Verlag, 12 Yupu Li, J. A. Kilner et al., Phys. Rev. B., 1995, 51, 8498. Berlin, 1955. 13 D. Gupta, R. B.Laibowitz and J. A.Lacey, Phys. Rev. L ett., 1990, 3 I. Kaur and W. Gust, Fundamentals of Grain and Interphase 64, 2675. Boundary DiVusion, Ziegler Press, Stuttgart, 1989. 14 N. Chen, S. J.Rothman and J. L. Routbort, J. Mater. Res., 1992, 4 S. Asbink and A. Waskowska, J. Phys.: Condens. Matter, 1991, 7, 2308. 3, 8173. 5 N. L. Peterson, in DiVusion in solids—recent developments, ed. Paper 7/02798B; Received 24th April, 1997 A. S. Nowick and J. J. Burton, Academic, New York, 1975, p.157. J. Mater. Chem., 1997, 7(10), 2085–2089 2089

ISSN:0959-9428    

DOI:10.1039/a702798b

出版商:RSC    

年代:1997

数据来源: RSC