摘要:

J. MATER. CHEM., 1991, 1(2), 197-199 Paramagnetic Rod-like Liquid Crystals, Bis[5-(4-al koxybenzoyloxy)sal icylaldehyde]copper(ii) Eduardo Campillos," Mercedes Marcos,*" Jose Luis Serrano" and Pablo J. Alonsob a Quimica Organica, lnstituto de Ciencia de Materiales de Aragon, Facultad de Ciencias, Universidad de Zaragoza-C.S. 1. C., 50009-Zaragoza, Spain Espectroscopia de Solidos, lnstituto de Ciencia de Materiales de Aragon, Facultad de Ciencias, Universidad de Zaragoza-C.S. I.C., 50009-Zaragoza, Spain The synthesis and mesogenic behaviour of a homologous series of copper(I1) complexes derived from 544-alkoxybenzoyloxy)salicylaldehydes are reported. The complexes exhibit smectic C mesomorphism for alkoxy groups containing eight or more carbon atoms in the terminal chains.These aldehydes therefore favour the formation of metal complexes with liquid-crystalline properties. Electron paramagnetic resonance (EPR) measurements were performed as a function of temperature. The signal can be associated with square-planar Cu2+ ions having an lk'-f> ground state. Molecular orientation in the mesophase under a magnetic field is discussed. Keywords: Liquid crystal; Paramagnetism; Metallo-mesogen; Bis(salicylaIdehyde)copper(II)complex In recent years metallo-mesogens of copper(i1) have received considerable attention.' The Cu" unit is well known for its ability to co-ordinate four donor atoms, which makes it possible to obtain structures that favour calamitic and discotic mesomorphism. Furthermore, Cu" systems are ideal for mag- netic investigations, providing information about the geometry of the metal complexes, and so the structure of the mesophase and consequently the physical properties of the compounds.Paramagnetic liquid-crystal properties of copper(i1) with p-diketones2, phthal~cyanines~ and Schiff's bases: have already been described. In this paper we report the synthesis and properties of a homologous series of copper(1r) complexes derived from 5-(4-alkoxybenzoyloxy)salicylaldehyde with the general formula A. which exhibit a smectic C mesophase typical of rod-like compounds. EPR measurements were performed with one of the com- plexes (n= 14) in order to study the paramagnetic properties of the complexes in the mesophase and the orientation relative to the magnetic field.Experimental Synthesis The preparation of the ligands and chelates was carried out according to Scheme 1. The ligands were synthesized using well known literature methods' by reaction of the appropriate acid chloride with 2,5-dihydroxybenzaldehyde. The copper(I1) complexes were prepared by the addition of an ethanolic solution containing Cu(OAc), *H20(1 mmol) to a hot solution of the appropriate aldehyde (2 mmol) in ethanol (100 cm3). The solution was refluxed for 1 h. After cooling, the precipitate was filtered off, washed with ethanol and recrystallized from chloroform. Elemental analysis and yields are collected in Table 1. A n =6-10,12,14 I 2 H' 3 4 4 + Cu(AcO), H20 fiii'lc A Scheme 1 Reagents and conditions: (i) SOCl,, DMF, reflux, 3 h; (ii) CH2Cl,, Et,N, room temp., several hours; (iii) EtOH, AcOH, reflux Table 1 Elemental analytical data (calculated values in parentheses) and yields of the complexes n c (O/O) H (Yo) yield (YO) 6 644 64.4) 5.7(5.7) 61 7 64.9(65.1) 6.2( 6.0) 75 8 65.6( 65.8) 6.5(6.3) 50 9 65.9( 66.5) 6.8(6.6) 50 10 66.8(67.1) 6.5(6.8) 62 12 68.2(69.3) 7.q7.3) 60 14 69.1(69.3) 7.9( 7.6) 75 Techniques Microanalysis was performed with a Perkin Elmer 240 B microanalyser.Infrared spectra for all the complexes were obtained using a Perkin Elmer 1600 (series FTIR) spec- trometer using Nujol mulls between polyethylene plates in the 400-4000 cm-spectral range. EPR measurements were carried out with a Varian E-112 spectrometer working in X-band.For measurements above room temperature (r.t.) the variable-temperature accessory (E- 157) from Varian was used. The powdered samples were placed inside a quartz tube and the temperature was moni- tored by a copper-constantan thermocouple attached to the tube. The error in temperature was estimated to be ca. 0.5 "C. The textures of the mesophases were studied with a polariz- ing optical microscope (Nikon) equipped with a Mettler FP82 hot stage and a Mettler FP82 central processor. Measurements of temperatures and enthalpies of transition were made using a Perkin Elmer DSC-2 differential scanning calorimeter with a heating rate of 10 "C min-' [the apparatus was calibrated with indium (1 56.6 "C, 28.44 J g-') and tin (232.1 "C, 60.5 J g-')I.Thermogravimetric analyses were performed on a Perkin Elmer TGS-2 equipped with a system 4 Microprocessor Controller at a heating rate of 10 "C min-' under nitrogen. Results and Discussion Synthesis and Characterization The copper(I1) complexes were prepared according to a general procedure6 by reacting the appropriate carbonyl compound with copper(r1) acetate monohydrate in warm ethanol. The complexes were isolated as green solids in good yield and were soluble in chloroform and dichloromethane, slightly soluble in toluene, and insoluble in ethanol, ether and hexane. The elemental analyses of the complexes were consistent with their proposed structures.The infrared spectra show a stretch band at ca.1620-1624 cm- ', which is assigned to v(CO), and a stretch band between 1717-1718 cm-', assigned to the ester group v(C=O) (with the exception of the complex where n=6, this band appears as a double band at 1727, I739 cm -'. The stability of the complexes was studied by thermogravi- metric analysis; they showed a weight loss at the transition temperature to isotropic liquid. This is consistent with the DSC data where decomposition at clearing temperature was also observed. Mesogenic Behaviour Table 2 summarizes the transition temperatures and enthalpy changes for the complexes synthesized. The compounds are stable in air within the temperature ranges of the measure- ments, although they begin to decompose at the same point that the complexes become isotropic.The mesophase shown by the complexes was clearly ident- ified as smectic C by the schlieren texture which shows only point singularities with four derived brushes. The nematic phase was excluded as a possibility since the point singularity with two branches was not observed. This mesophase is similar to that observed for classical rod-like organic com- pound~.~,*The influence of the terminal chain length of the Table 2 Optical, thermal and thermodynamic data for complexes n transition T/"C AH/kJ mol-' 6 c1-c2 171.6 7.63 6 7 C2-I(dec.) c1-c2 244.9 169.9 5.39 7 C2-C3 226.2 10.97 7 8 C3-I(dec.) c1-c2 265.3 109.4 13.01 8 C2-C3 196.1 10.57 8 c3-s, 254.2 25.30 8 9 S,-I(dec.) c1-c2 263.9 131.8 3.0 I 9 C2-C3 194.5 10.91 9 c3-sc 247.0 24.43 9 10 S,-I(dec.) c1-c2 256.8 70.8 5 1.40 10 C2-C3 192.3 14.30 10 c3-sc 242.1 26.50 10 Sc-I(dec.) 250.7 J.MATER. CHEM., 1991, VOL. 1 ligands on the mesogenic properties of the complexes is also apparent from Table 2. The complexes where n=6,7 do not show mesogenic behaviour, while the complexes where n =8 exhibit a smectic C mesophase. There is a slight decrease in the transition temperatures as the length of the terminal chain increases. The transition enthalpy of the Sc-I was not deter- mined, owing to the decomposition of the complexes on passing to isotropic liquid. All the complexes exhibit crystalline polymorphism, which was detected by DSC and optical microscopy.However, the crystalline phase, which appears prior to the Sc phase and which has been named CJ, is a highly viscous and birefringent phase. During the optical observations with the polarizing microscope the glass slide and cover slip can be made to glide over one another by exerting a little pressure; the sample does not show fractures but rather a considerable deformation. EPR Measurements The EPR spectra of bis[5-(4-tetradecyloxybenzoyloxy)salicyl-aldehyde]copper(~~)were measured at different temperatures between room temperature and 250 "C (decomposition tem- perature). In Fig. l(a), the spectrum of an untreated sample measured at room temperature is shown.This spectrum can be associated with that of a paramagnetic entity with S= 1/2, and whose spin Hamiltonian consists of an axial electronic Zeeman term characterized by gll=2.28 0.02 and g, = 2.08k0.02. These values are typical of Cu2+ in a square-planar environment, having the lx2-y2 > ground It is noteworthy that in the present case we do not observe any resolved hyperfine structure even in the parallel trace. Recently, we have studied '' organometallic homopolymers of Cu" as well as low-molecuiar-weight copper(1r) complexes 270 290 310 330 HImT Fig. I EPR spectra of bis[5-(4-tetradecyloxybenzoyloxy)- salicylaldehyde]copper(rr) taken (a)at room temperature (b)at 240 "C and (c) at 250 "C J. MATER. CHEM., 1991, VOL.1 derived from Schiff's bases.I2 In these last cases the parallel hyperfine structure was resolved and values of ca. 15 mT are found for this hyperfine structure, indicating that in the present case this interaction is lower (at least by a factor of 3). This can not be explained by a simple model, but indicates that the electronic spin is largely delocalized as a consequence of a strong covalent contribution to the Cu-0 bonds. Structural information about copper(r1) O4 chelates or cop-per(I1)N202chelates has been reported13 and their geometries (distances and angles) depend strongly on the different sub- stituents; this together with the lack of X-ray data for our compounds prevent us from making an in-depth study. Further spectroscopic (Raman and X-ray diffraction) will be made in order to understand the differences in the Cu-ligand bonds.No modification of this EPR spectrum is detected at temperatures lower than 220"C, but some changes are ob- served when the EPR spectra are taken in the Sc phase. Fig. l(bj and I(c) correspond to the EPR spectra taken at 240 and 250 T,respectively. In the former [Fig. l(b)] an increase in the perpendicular signal with respect to the parallel is observed. A slight increase in the temperature tends to restore the initial situation [see Fig. l(c)]. These results can be inter- preted in the same way as those of some vanadyl organo- metallic-mesogen derivatives. l4 The benzene rings in the organic skeleton provide an anisotropic diamagnetic contri- bution to the magnetic susceptibility, which forces the mol- ecule to orientate with the d.c.magnetic field along its molecular axis. The paramagnetic contribution to the an-isotropy of the magnetic susceptibility due to the copper is not enough to prevent the preferential orientation. For this reason the parallel contribution to the spectra decreases as a consequence of the orientation induced by the magnetic field. When the sample is heated a few degrees [see Fig. l(c)] the thermal agitation works against this preferential magnetic- field-induced orientation and the parallel signal increases at the expense of the perpendicular one. In the 250 "C spectrum a narrow line at g= 2.00 (320 mTj is observed. The intensity of this line increases as a function of the time if the sample is kept at this temperature.This is probably due to some decomposition products of our compound, which is in agree- ment with thermal data. Conclusions From the results we can conclude that the 5-substituted salicylaldehydes have a molecular structure suitable for pro- ducing rod-like liquid-crystal metal complexes with paramag- netic properties. We are currently trying to introduce modifications in the chemical structure of the ligands in order to obtain more stable systems with lower transition temperatures. Thanks are due to the Ministerio de Educacion y Ciencia for a grant to E.C. References 1 (a)A. M. Giroud-Godquin and J. Billard, Mol. Cryst. Liq. Cryst., 1981, 66, 147; (b)I.V. Ovchinnikov, Y. G. Galyametinov, G. I. Ivanova and L. M. Yagforava, Dokl. Akad. Nauk SSSR, 1984, 276, 126; (c) K. Otha, H. Muroki, A. Takagi, I. Yamamoto and K. Matsuzaki, Mol. Cryst. Liq. Cryst., 1986, 135, 247; (d) M. Ghedini, S. Armentano, R. Bartolino, F. Rustichelli, G. Torquati, N. Kirov and M. Petrov, Mol. Cryst. Liq. Cryst., 1987, 151, 75; (e)A. Roviello, A. Sirigu, P. Iannelli and A. Immirzi, Liq. Cryst., 1988, 3, 115; (f) R. Paschke, H. Zaschke, A. Madicke, J. R. Chipperfield, A. B. Blake, P. G. Nelson and G. W. Gray, Mol. Cryst. Liq. Cryst. Lett., 1988, 6, 81; (g) S. Chandrasekhar, B. R. Ratna, B. K. Sadashiva, V. N. Raja, Mol. Cryst. Liq. Cryst., 1988, 165, 123; (h) T. P. Shaffer and K. A. Sheth, Mol. Cryst. Liq. Cryst., 1989, 172, 27; (i) M.Marcos, P. Romero, J. L. Serrano, C. Bueno, J. A. Cabeza and L. Oro, Mol. Cryst., Liq. Cryst., 1989, 167, 123; (j) B. Muhlberger and W. Haase, Liq. Cryst., 1989,5, 251; (k)M. Marcos, P. Romero, J. L. Serrano, J. Barbera and A. M. Levelut, Liq. Cryst., 1990, 7, 251. 2 S. Chandrasekhar, B. K. Sadashiva and B. S. Srikanta, Mol. Cryst. Liq. Cryst., 1987, 151, 93. 3 Ch. Piechocki, J. Simon, A. Skolious, D. Guillon and P. Weber, J. Am. Chem. SOC., 1982, 104, 5245. 4 M. Marcos, P. Romero and J. L. Serrano, J. Chem. SOC., Chem. Commun., 1989, 1641. 5 P. Keller and L. Liebert, Solid State Phys. Suppl., 1978, 14, 19. 6 R. H. Holm, J. Am. Chem. Soc., 1961,83,4683. 7 D. Demus and L. Richter, Textures of Liquid Crystals, Verlag Chemie, Leipzig, 1978. 8 G. W. Gray and J. W. Goodby, Smectic Liquid Crystal Textures and Structures, Leonard Hill, Glasgow, 1984. 9 R. H. Sands, Phys. Rev., 1955, 99, 1222. 10 D. L. Griscom, J. Non-Crysr. Solids, 1980, 40, 211. 11 M. Marcos, L. Oriol, J. L. Serrano, P. J. Alonso and J. A. Puertolas, Macromolecules, in the press. 12 P. J. Alonso, M. Marcos, P. Romero and J. L. Serrano, unpub- lished results. 13 R. H. Holm and M. J. O'Connor, Prog. Znorg. Chem., 1971, 14, 241. 14 P. J. Alonso, M. L. Sajuan, P. Romero, M. Marcos and J. L. Serrano, J. Phys. Condens. Matter, 1990, 2, 9173. Paper 0/03182H; Received 16th July, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100197

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 201-203 201 Database Analysis of Crystal-structure-determining Interactions involving Sulphur: Implications for the Design of Organic Metals Gautam R. Desiraju" and Veerapaneni Nalini School of Chemistry, University of Hyderabad, P.O. Central University, Hyderabad 500 134, India An analysis using the Cambridge Database of 926 structures in the short-axis range up to 8.0 A, containing divalent sulphur, Y-S-Z, shows that 77 have short axes less than 4.5 A and are distinguished by a strongly directional pattern of sulphur-heteroatom interactions. Keywords: Organic metal; Sulphur; Crystal structure; Intermolecular interaction Prediction of organic crystal structures is essential for the deliberate design of new materials with specified physical and chemical properties.But organic crystal structures are deter- mined by an interplay of forces that are weak yet numerous and of variable directionality. Therefore computation of these forces has been difficult. An attractive alternative method of predicting crystal structures is through statistical inference from existing crystallographic data on related compounds conveniently retrieved from databases such as the Cambridge Structural Database.' This paper is concerned with divalent sulphur compounds that adopt highly overlapped stack structures (p-structure) characterised by a short crystallographic axis of ca.4 A. In certain cases, such one-dimensional structures are associated with anisotropic electrical conductivity. Here, we consider these compounds as part of a larger set of sulphur-containing compounds which can be grouped into distinct structural families. An earlier study of planar aromatic hydrocarbons showed that four basic structure types need to be ~onsidered.~.~ These are the p-structure [short axis (s.a.) <4.2 A; tribenzopyrene], the y-structure (4.6 <s.a./A <5.4; coronene), the herringbone (HB) structure (5.4<s.a./A <8.0; naphthalene), and the sand- wich-herringbone structure (s.a.>8.0 & pyrene). These ranges of values for the short axis were used since there is a paucity of structures in the intermediate regions around 4.4, 5.4 and 8.0 A. It was also observed that the short axis was the key parameter in separating packing types and that it defines the crystal structure.Molecules in the four families above have distinct geometries and co-ordination patterns and hence different profiles of energy stabilisation by their near neigh- bours in the ~rystal.~ These ranges of short-axis values seem to have some fundamental significance and demarcate crystal structures for other chemical classes of compounds also. For instance, a retrieval from the third version of the Cambridge Database' (69691 compounds) of metal-free, error-free, diffractometer- data (R<0.10) compounds containing sulphur, resulted in a total of 3594 structures, which were naturally grouped accord- ing to the short axis. Three ranges were observed: s.a.<4.5 %, (181 com ounds), 4.5 <s.a./A< 5.4 (454 compounds), and 5.4<s.a./1<8.0 %i (2959 compounds).This is a much more heterogeneous group than planar aromatic hydrocarbons and contains compounds in a wide variety of chemical environ- ments. Both planar and non-planar compounds are included; sulphur may be cyclic or acyclic and present in any bonding situation, and any or no other heteroatom may be present. The first group is almost exclusively populated by planar compounds, owing to the geometrical constraint involved in packing a non-planar molecular with s.a.<4.5 A; this group is therefore identical to the p-structure with crystal stabilis- ation arising primarily from stacking. Similarly, the second group (4.5 <s.a./A <5.4) consists of planar molecules with one or more bulky groups, which prevent p-structure adoption.This group therefore represents the y-structure. Here too, translational stacking is important, but lateral neighbours are inclined at steep angles, indicating that they are related by screw/glide operations. The third group (5.4 <s.a./A<8.0) is much larger and more varied, containing both planar and non-planar molecules. There is a minimum in the population at ca.6.8 A and so this group was divided into two ranges: 5.4<s.a./A <6.8 (1416 compounds) and 6.8 <s.a./A <8.0 (1543 compounds). Both ranges define the HB structure with the main stabilisation arising from interactions between neigh- bouring screw/glide related molecules. Crystal structures with s.a.>8.0 A were not considered here since they are numerous, diverse and complex.A simple short-axis criterion does not seem to be sufficient to categorise structures with s.a. >8.0 A, unless a restricted set such as planar molecules is chosen. It is well known that intermolecular interactions with divalent sulphur (Y-S-Z) of a heteroatom RR (Groups 15, 16, 17) are of two types: (1) a nucleophilic approach within the plane defined by YSZ and posterior to either the Y-S or the Z-S bond; (2) an electrophilic approach nearly perpendicular to the YSZ plane.6 Therefore, the structures retrieved above were analysed using the program GSTAT88. We chose, from the 3594 hits above, those structures where divalent sulphur Y-S-Z is present (Y, Z are any non-hydrogen atoms and the S-Y and S-Z bond lengths are in the range 1.5-2.1 A).A total of 926 structures were obtained with 77, 135, 309 and 405 in each of the four groups. For these structures, the distances S...RR (2.8<D/A <3.8) and the inclination angle (0) of S...RRto the plane YSZ were calculated (Fig. 1). Fig. 2 is a graphical representation of these results and shows scatterplots of D versus 8 for the four structure 70-.120' ___) 1.5-2.18 Fig. 1 Geometry of contacts to divalent sulphur to show the param- eters D and 8. Y and Z are any non-hydrogen atoms. RR is a heteroatom (Groups 15, 16, 17). 0 is the inclination of RR above the plane YSZ J. MATER. CHEM., 1991, VOL. 1 2 80 3 08 3 26 3 44 3 62 3 80+I--------I--------I--------I--------I--------I+ s RR 95 0--ACROSS DOWN I 13 IS RR THETA 12 I MEANI I 1 1 311 1 21 I 3 604 45 636 75 0 1111 121 21 -SOEV I 1 1 3 1 217481 0 182 27 968 I 11 1 1 1226371 MINIW I 2 Ill211 3 015 0 438 111 1 1 -MAXILNH55 0-I 2 11121 I 3 788 80 000 I 1 21 II 2 1 I 186 POINTS PLOTTED I 35 0-1 1 2 1 -0 POINTS OMITTED 30 0-I 21 3 I 1I 212 I1 1 I CORREL COEFF II I 1 1 11 22 1 3 I 0 508 15 0-1 11 11 1I 1 11 11 13 1 13 I 11124 I 121 21 11 3111 1 111 I I 11 1 11 21 -5 0-+1---.___-1--------1--------1--------1--------1+ THETA 3 42 3.62 3 82 -------1---------1---------11 S...RR -ACROSS THETAIS...RR DOWN 11 1 1221 I w. 1 1 1 1 I 1 21111 1 I 3 577 36.084 1 11 11143621 -SOEV 1 2 1 11 1 512131 I 0 182 25 485 11 22 1 1 121311212 I MINIMUM.1 1 Ill 12 1 113412 I 2 855 0 461 211 1 21 -MAXIMUM 1 1 1 1 2 1 1 I 3.800 82.178 12-12 112 Ill2 11 1 1 213 1 312 POINTS PLOTTED 121 2 1 11 0 POINTS OMITTED 1 11 114 3 11 2 21 3 1 11 42 131 1 1 221 212 CORREI. COEFF 121 1 1 211 1 11 1 0 282 1 21 2 2 11 121252 321112 1 1 2112 1 122212 3 1 11312 I3 11 I 211 I 2 84 3 04 2.45 2 85 3 26 3.65 --I+ s . RR+I._____._. ____ ----I-------I.------I---. 4 .051-90 0-I 1 I 11 I1 1 21 70 0-1 1 I1 1 1 I 2 1 22 1 1 50 0-1 1 1 1fL 2 30 0-1 1 21 I 4 1 1 1 I 31 1 11 1 3124 11211 5 1 1 23 121 2 12564 I 1 22 23 3243332542 2 1 142 21111622131144 11 62122325372533 1 1 3 114121 431572 214 11121 21 1 31155 321 I I I I -I 3 574 34 578 0 184 24.982 MINIMUM: 2 886 0 388 MAXIW: 3 800 88 253 SDEV I70 0-I I I 50 0-I 1 13411 2232 1 2112 21142246562 1 21231137C8853 1 1411 22326AIZJ 1 214111217A885 1 11233228866 1 131628664 -ACROSS DOWN IS .RR THETA I MEAN I 3.571 36 830 SDEV I 0 187 26 367 I MINIMUM: I 2 802 0 341 -MAXIMUM: f 3 800 88.213 1 11113313 2412211441 1 1152 2222221322432 111 3 322 522 2 442 2 I I816 POINTS PLOTTED -0 POINTS OMITTED I I 30 0- II1073 POINTS PLOTTED -0 POINTS OMITTED 1 11 4 322127311411311 1211 1 1233133234 21312 11 1 25622441232122442Al I I I CORREL COEFF 0 175 I I I I I CDRREL. COEFF 0 163 10 0-1 12 I1 1 15 111 2 113 1 21221452211117344345344 113 3222322 2112333115A81415452163 121 1 2211211232131 11 32412122 I I I 10.0-I I I -~nn-.-- I I +I- THETA Fig.2 Scatterplots of the S...RR distance D uersus inclination 8 for (a)/?,(b)7, and (c), (d)HB structures groups. These plots contain much useful information. Within the specified range of D,the number of contacts are 186,312, 816 and 1073 and the number of contacts per structure is therefore 2.41 (b),2.31 (y), 2.64and 2.65(HB). Though the number of contacts per structure is nearly the same in the four groups, a fact which might have resulted from the liberal limits set for D and 8 in the GSTAT calculations, the contacts in the @-familyare directed in a highly specific manner (Fig. 3).Fig. 2(a)shows that there are many contacts around 8=0" and 8=90" with a clear minimum around 8=45".The S...RR contacts are therefore of two types.The 8=90" contacts are in general longer (>3.4A) and are between stacked molecules. Of the 77 b-structures considered, 15 have s.a.<3.8 A and so there is at least one translational S...Scontact less than 3.88, for each of compounds in Fig. 2(a). Lateral (8=0") contacts could be much shorter and facilitate the arrangement of the planar molecules in two-dimensional arrays which can be stacked to obtain the three-dimensional structure. Therefore, they may be considered as being crystal-structure determining. There is a significant correlation (0.51) between D and 8, with the shorter contacts being the lateral ones. These features are somewhat blurred in the y-structures [Fig. 2(b)].There is only a hint of a minimum at 0=45" and some of the 8=90" contacts are quite short. This follows from the geometry of the y-structure where planes of neighbouring perpendicular If lateral Fig. 3 Schematic representation of a /?-crystal structure for a planar sulphur compound. Filled and empty circles could be sulphur and other heteroatoms. Lateral forces are interstack and have 8~0". Perpendicular forces are intrastack and have 8~90".The lateral forces align molecules into two-dimensional motifs and are hence structure determining molecules are steeply inclined. The fact that the S...RR contacts are not so directional here indicates that these contacts are no longer structure determining but rather that their geometri-cal preferences are accommodated within the framework of isotropic C...C stacking interactions and C...H screw/glide interactions.Accordingly, the correlation between D and 8 is far less pronounced (0.28).Continuing the trend, the patterns of S...RR interactions in the HB structures are completely obscured by the isotropic forces [Fig. 2(c)and 2(d)].There is no significant difference between these two groups of com-pounds and hardly any preference for a particular 8 value. Consequently, there is very weak correlation between D and 8 (0.18,o.iq. Organic conductors and superconductors of the donor-acceptor type have a segregated stack one-dimensional b-crystal structure.' It has been observed in individual cases that these structures are characterised by lateral S...S, S...N and S...RR interaction^.^,^ The present work shows that two types of sulphur atom interaction, lateral and perpendicular, need to be considered.If the p-structure is to be adopted, lateral interactions should be dominant. This may be achieved through molecular planarity and a profusion of S and other heteroatoms relative to the C and H atoms, in other words by adjusting the C:H:S:RR ratio. If this ratio tends towards higher C and H stoichiometries in the molecule, isotropic C...C and C...H forces may predominate resulting in the adoption of the alternative y and HB structures. These struc-tures also have both lateral and perpendicular sulphur atom contacts. Although they are as numerous as in the p-structures, these contacts are distorted according to the isotropic forces, hence they are not as directionally specific as in the p stack structures and not crystal-structure determining.References 1 G. R. Desiraju, Crystal Engineering. The Design of Organic Solids, Elsevier, Amsterdam, 1989. 2 G. R. Desiraju and A. Gavezzotti, J. Chem. SOC.,Chem. Commun., 1989, 621. 3 G. R. Desiraju and A. Gavezzotti, Acta Crystallogr., Sect. B, 1989, 45, 473. J. MATER. CHEM., 1991, VOL. 1 203 4 5 6 A. Gavezzotti and G. R. Desiraju, Acta Crystallogr., Sect. B, 1988,44, 427. F. H. Allen, S. Bellard, M. D. Brice, B. A. Cartwright, A. Doubleday, H. Higgs, T. Hummelink, B. G. Hummelink-Peters, 0. Kennard, W. D. S. Motherwell, J. R. Rodgers and D. G. Watson, Acta Crystallogr., Sect. B, 1979, 35, 2331. R. E. Rosenfield, R. Parthasarathy and J. D. Dunitz, J. Am. Chem. Soc., 1977, 99, 4860. 7 8 9 For example, J. B. Torrance, Acc. Chem. Res., 1979, 12, 79. J. D. Wright, Molecular Crystals, Cambridge University Press, Cambridge, 1987, pp. 37-40. J. M. Williams, M. A. Beno, H. H. Wang, P. C. W. Leung, T. J. Emge, U. Geiser and K. D. Carlson, Acc. Chem. Res., 1985, 18, 261. Paper 0/03184D; Received 16th July, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100201

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 205-212 Intercalation of 2-Aminoethylferrocene into the Layered Host Lattices MOO,, 2H-TaS2 and a-Zr(HPO,),= H,O Kalyan Chatakondu, Carl Formstone, Malcolm L. H. Green,* Dermot O'Hare,* J. Mark Twyman and Philip J. Wiseman Inorganic Chemistry Laboratory, South Parks Road, Oxford OX 1 3QR, UK The amino-functionalised ferrocene derivative FcCH,CH,NH, [Fc = Fe(q-C,H,)(q-C,H,)] has been intercalated under mild conditions into both single-crystal and microcrystalline samples of the layered host lattices MOO,, 2H-TaS2, and a-Zr(HPO,), H,O. The X-ray powder patterns for the new intercalates have been indexed and yield interlayer lattice expansions of 13.09, 5.89 and 14.54 A, respectively. The large interlayer expansions of 13.09 and 14.5 A observed for the intercalates of both MOO, and r-Zr(HP04),*2H,0 suggest the formation of a bilayer of guest molecules within these host lattices. The 57Fe Mossbauer spectrum of MoO,(FcCH,CH,NH,),,,, exhibits a quadrupole doublet with an isomer shift &=0.442 mm s-' and quadrupole splitting A=2.344 mm s-', indicating the presence of unoxidised ferrocene molecules within the oxide layers.Solid-state 13C CP-MAS NMR spectra of Mo0,(FcCH,CH,NH,),~,6 are consistent with the existence of both neutral FcCH,CH,NH, and FcCH,CH,NHl moieties within the host. In addition, the changes in the electronic structure of MOO, that accompany the intercalation of FcCH,CH,NH, have been investigated by solid-state photoelectron spectroscopy. Keywords: Intercalation; X-Ray diffraction; Aminoferrocene; Solid-state nuclear magnetic resonance spec-troscopy, Mossbauer spectroscopy It has now been well established that amines, especially primary aliphatic amines, readily intercalate into a variety of lamellar host lattices,' including many industrially important heterogeneous catalysts, with often beneficial effects to their catalytic activity.2 Unfortunately, many potentially interesting organometallic guest molecules do not intercalate directly into these lattices, or decompose under the conditions of the intercalation reaction.We therefore decided to investigate whether amino-functionalised organometallic guest molecules could act as potential guest molecules in these reactions. In particular, we were interested in determining if this approach could be used to intercalate organometallic guests that could not be intercalated under normal conditions.Previously, it has been shown that dimethylaminomethylferrocene can be intercalated into HU02P04 4H,03 and a-Sn(HPO,), * H20., We have already briefly reported the synthesis of the intercalation compounds formed by reaction of 2-aminoethyl- ferrocene with the various host lattice^.^ Here we present the full details of the synthesis and characterisation of the intercal- ation of 2-aminoethylferrocene into the host lattices MOO,, 2H-TaS, and x-Zr(HPO,), * H20. These lattices were chosen since they do not react with ferrocene to form intercalation compounds under normal conditions.Experimental Method The reactions were carried out in an inert atmosphere of nitrogen by the use of vacuum-line or inert-atmosphere dry box. Solvents were pre-dried over molecular sieves (type 4A) and refluxed with rigorous drying agents under a continuous stream of nitrogen. Tetrahydrofuran and diethyl ether were refluxed over sodium/potassium alloy, and dimethoxyethane (DME) was refluxed over potassium metal. Dichloromethane was dried by refluxing over P205 and nitromethane and acetonitrile by refluxing over CaH,. Solvents were distilled prior to use and were stored over molecular sieves in flame- dried ampoules under nitrogen. Equipment Infrared spectra were recorded on a Mattson Instruments Polaris Fourier Transform spectrometer as mulls in Nujol between KBr plates.Powder X-ray diffraction spectra were recorded on a Philips PW 1710 powder diffractometer, controlled by a MAP 80 microcomputer, using Cu-Ksr radiation. Zero-point correc-tions were made using Si and MOO, reference samples run immediately after the sample run. Lattice parameters were determined by least-squares refinement of the calculated d-spacing. For all the intercalates discussed here, the crystal system of the new lattice was assumed to be that of the host, i.e. orthorhombic (MOO,), monoclinic ($-hexagonal) [Zr(HPO,),] or hexagonal (TaS,). In the absence of a full indexing the long axis and thus the interlayer spacing cannot be deduced from the position of the 001 reflections. What can be deduced, for example if the crystal symmetry is monoclinic, is c sin p.Since sin PI1 the true value of c may be larger than the quoted figure. Similarly, indexing of reflections other than those of the 001 class should be regarded as tentative as the symmetry may be lower than that assumed. Also it is common, at higher angle, for several reflections to occur at very similar Bragg angles. Photoelectron spectra of solid samples were recorded using a VG Escalab 5 spectrometer. Single-crystal samples were pressed onto indium foil and mounted onto platinum stubs. The crystals were cleaved with Sellotape in the preparation chamber (ca. lO-'mbar) in order to expose a clean fresh surface prior to exposure to the photon sources in the main chamber (ca.lO-"mbar). The photon source in the main chamber was a UV discharge gun operating at 0.5 kV. Solid-state ' cross-polarisation magic angle spinning (CP-MAS) spectra were recorded on a Bruker MSL200 instru- ment (at 50.3 and 200.1 MHz for I3C and 'H, respectively). Samples were packed in 17 mm (outer diameter) zirconia rotors and data were acquired with a Bruker double-gas bearing probe. MAS frequencies v, of ca. 3.3 kHz, 13C--lH dipolar decoupling frequencies V1H of ca. 75 kHz, and cross- polarisation contact times tc,=2 ms were used. The 13C chemical shifts were externally referenced to the upfield reson- ance of adamantane at 29.23 ppm with respect to tetramethyl-silane. Solution NMR spectra were recorded using a Bruker AM 300 (300 MHz) spectrometer.'H spectra were referenced using solvent peaks internally; all shifts are inppm relative to tetramethylsilane. Elemental microanalyses were performed by the Analytical Services of the Inorganic Chemistry Laboratory. Syntheses Synthesis of Host Lattices MOO, (99.5%) was supplied by the Sigma Chemical Com- pany. It was sintered by heating in air for 6 h at 600 "C prior to intercalation. Zr(HP04)2 H,O was prepared according to literature methods.6 2H-TaS2 was prepared as f01lows.~ Tantalum powder (0.738 g, 0.004 mol) and a slight stoichiometric excess of sulphur (0.265 g, 0.0082 mol) were loaded into a 10 cm silica ampoule. The ampoule was sealed under vacuum. The sealed ampoule was then placed in a cold furnace and heated gradually to 450 "C (over ca.1 day) and held at this tempera- ture for 3 days, after which it was heated further to 860 "C for 3 days. The temperature of the furnace was then gradually lowered, first to 750 "C for 1 day, then 650 "C for 3 days, and finally to 520 "C for 2 days after which it was shut down and left to reach room temperature for 5-6 h. This procedure gives polycrystalline 2H(a) TaS,. Any gold-coloured material indicates the presence of the 1T phase and is a result of improper annealing (cooling) of the sample. The presence of a fibrous or needle-like material indicates TaS, formation which decomposes at 650 "C. Reheating to 850 "C and follow- ing the same annealing procedure as above will remove this impurity.The material was isolated in an inert atmosphere and placed in an ampoule which was heated to 120 "C under a reduced pressure of 5 x1Op2 mbar for 2 h to remove all volatile impurities. 2-Aminoethylferrocene (1)' and TaS2(NH3)7 were prepared according to literature procedures. Synthesis of Zr(HP0,)2 -MeOH9 Zr(HP04)2 H20 (1.08 g, 0.00 359 mol) was stirred in 200 cm3 of 0.01 mol dm-, NaCl solution. The resulting mixture was titrated potentiometrically with 0.1 mol dm- NaOH solution until a plateau in the rise of the pH (ca. pH 6) was reached; at this point Zr(HP04)NaP04 5H20 is the composition of the white solid. This required 40cm3 of the NaOH solution (0.0039 mol). The white solid, Zr(HPO,)NaPO, 5H20 was filtered under a reduced pressure, washed once with acetone and dried in ULZCUO.Zr(HPO,)NaPO, * 5H20(1 g, 0.0025 mol) was suspended in 40 cm3 of AnalaR MeOH. A 2 cm3 volume of 40% HBF, was added to the suspension which was then stirred for 1 h.The solid was filtered under reduced pressure, washed with 3 x 10 cm3 of MeOH and dried in uucuo [yield = 0.56 g (5O%)]. Synthesis of MoO,(F~CH~CH~NH,),,~~(2) A solution of 2-aminoethylferrocene (600 mg, 0.0026 mol) in dry, degassed acetonitrile (2cm3) was added to 300mg of Moo3 which was pre-weighed into an ampoule containing a magnetic flea. The reactants were heated at 100 "C with stirring, for 7 days under an inert atmosphere of nitrogen. The intercalate was washed thoroughly with acetonitrile and then with dichloromethane before it was dried in uucuo and stored under nitrogen. Elemental analysis found [calc.for Mo03(FcCH2CH2NH2)0.36, Mo03Fe0.36C4.32H5.4N,.361: c, J. MATER. CHEM., 1991, VOL. 1 22.7(22.7); H, 2.36(2.39); N, 2.17(2.16); Fe, 7.4(7.4); Mo, 47.5(47.1)?/0. Synthesis of TaS2(FcCH2CH 2NH 2)0. g(H20)x (~~0.1-0.5)(3) The synthesis of this intercalate cannot be done by directly reacting the host and guest, but requires exchanging the 2-aminoethylferrocene for NH, pre-intercalated into the host. Single crystals of TaS2NH3 (200 mg, 0.0007 mol) were loaded into an ampoule containing a magnetic flea in a dry box. A solution of 2-aminoethylferrocene (400 mg, 0.0017mol) in 2 cm3 of dry, degassed DME solvent was added to the ampoule under nitrogen.The reaction mixture was heated for 6 days at 100 "C. The intercalate was washed thoroughly with DME and dried under vacuum. Elemental analysis found [calc. for TaS2(F~CH2CH2NH2)o.19(H20)o,5, TaS2Feo.i9C2.28H3.85No.1900.5]:c, 9.39(9-35); H, 1.43(1.3); N, 1.1 l(0.9); Fe, 2.5(3.62)%. Synthesis of Zr(HP04)2 (FCCH~CH~NH~)~.~(H~O), (X =0.1-0.5) (4) Zr(HP0,)2 H20 intercalates 2-aminoethylferrocene when its layers are preintercalated with methanol. A solution of 2-aminoethylferrocene (600 mg, 0.0026 mol) in 2 cm3 of AnalaR methanol was added to an ampoule containing Zr(HP04)2 MeOH (300 mg, 0.0009 mol). The reaction occurred at room temperature overnight. The intercalate was washed with methanol and dried in uacuu. Elemental analysis found [calc.for Zr(HP04)2 (FCCH~CH~NH~)~.~(H,O)~,~, ZrFeo.sP2No.508.sC6,5~~o,5]:C, 17.75( 17.7); H, 2.85(2.58), N, 1.60( 1.82)%. Results and Discussion The previously characterised 2-aminoethylferrocene (1)' can be conveniently prepared from ferrocene using the reaction scheme shown in Fig. I. Although this is a multistep synthesis, overall yields of 30% were possible. Typically 2 g of 2-amino- ethylferrocene can be prepared starting from 5 g of ferrocene. Compound 1 is a red viscous oil at room temperature and so no X-ray structural characterisation has been determined. However, we have computed the molecular structure using eCH2CN Fe CH ,CH2N H2 eCH&H2N+H3CI. (v) Fa Fig. 1 Reaction scheme: (i) Me,NCH,CH,NMe,, H,PO,, CH,CO,H, reflux 5 h, 77%; (ii) MeI, diethyl ether, r.t, 65%; (iii) KCN, H,O, 88%; (iv) LiAlH,, diethyl ether, 78%; (v) HCl(g), diethyl ether, 78%; (vi) NaOH(aq), 94% J.MATER. CHEM., 1991, VOL. 1 1.10 A Fig. 2 Calculated molecular structure for 1, showing (a)covalent and (b)van der Waals representations the molecular-modelling package. ChemX' and literature bond lengths and angles for the various functional groups. The molecular parameters and a van der Waals representation for 1 are shown in Fig. 2. Synthesis and Spectroscopic Study of Mo03(FcCH2CH2NH2)0.36 Refluxing of 1 with either single crystals or microcrystalline samples of MOO, in MeCN for 6 days gives complete intercalation, the intercalate was washed repeatedly with dichloromethane to remove any chemisorbed 2-aminoethylfer- rocene, and then dried in uucuo.The white molybdenum trioxide turned black upon intercalation. Elemental micro- analysis data gave a reproducible stoichiometry of Mo03(FcCH2CH2NH2)0.36 (2)-Powder X-Ray Digraction The indexed powder X-ray diffraction pattern for 2 contains no reflections assignable to the host MOO, (Table 1); the observation of a low-angle 020 reflection at 26 =4.35"indicates a significant increase in the interlayer spacing of the new lattice. If we assume that the structure of 2 is orthorhombic and analogous to that of the parent MOO, lattice, then the interlayer separation is given by b/2. This occurs since con- secutive MOO, octahedral sheets are displaced relative to each other along the stacking axis, which is the b axis in the orthorhombic crystal system (Fig.3). Least-squares refinement lattice parameters confirm an increase in the interlayer spacing (Ab)= 13.2 A (Table 1). This increase in interlayer separation in 2 is significantly larger than that observed for other intercalation complexes involving other metallocenes (Table 2). On this basis we propose that the structure consists of a bilayer of Fig. 3 Representation of the MOO, structure, showing the stacking axis Table 1 Powder X-ray diffraction data for MOO, and the intercalate 2 12.86 24 0 2 0 6.88 6.9 1 4.35 100 0 2 20.3 1 19.9 23.46 59 1 1 0 3.789 3.790 8.9 I 31 0 4 9.92 9.93 25.84 52 0 4 0 3.445 3.455 13.5 15 0 6 6.55 6.62 27.47 100 0 2 1 3.244 3.25 1 18.3 7 0 8 4.84 4.97 29.77 5 1 3 0 2.999 2.995 23.49 60 0 4 3.78 3.7 1 33.28 7 1 0 1 2.690 2.692 33.9 23 0 12 2.64 2.55 35.64 22 0 4 1 2.5 17 2.527 48.3 23 0 8 1.88 1.85 39.13 26 0 6 0 2.300 2.303 54.3 25 0 12 1.69 1.71 39.82 30 1 5 0 2.262 2.263 45.96 86 1 6 0 1.973 1.989 46.45 11 0 6 1 1.953 1.953 49.44 12 0 0 2 1.842 1.842 52.96 20 0 8 0 1.727 1.728 54.27 I0 2 2 1 1.689 1.686 55.37 4 1 1 2 1.658 1.657 56.56 12 0 4 2 1.626 1.626 57.83 9 1 7 1 1.593 1.592 58.97 10 0 8 1 1.565 1.564 67.69 21 0 10 0 1.383 1.382 69.75 5 2 0 2 1.347 1.34 1 Indexed on orthorhombic cell: a =3.941( 16) A, b = 13.82(2) A, c =3.68( 1) A.Indexed on orthorhombic cell: b =40.00( 16) A, c =4.00(6) A. J. MATER. CHEM., 1991, VOL. 1 Table 2 Observed interlayer expansions (Ac=dint-dhost)in a variety of organometallic intercalation compounds host organometallic guest stoichiometry AclA ref. ZrSz co(?-c5 5)Z Cr(q-c5 5) 2 bis(q-benzene)chromium bis( q-t o1uene)molybdenum 1,l'-dime t hylcobal tocene 1,l'-di-n-butylcobaltocene 2H-TaSz Co(q-C5H5)2 cr(q-c5H5)Z SnS2 CO(vl-C5H 42 FeOCl Fe(q-C5 512 CO(?-C5H5)2Fe( q-C5 Me,E t)Z VOCl cr(V-c5 5)2 CdPS, Co(vl-C5H512 NipsJ bis( q-benzene)chromium 2-aminoethylferrocenes as shown in Fig. 4. Jacobson and co- workers" have also proposed a similar bilayer for the struc- ture of pyridine intercalated into MOO,.For 2 it is not possible to infer the relative orientation of the guest molecules within the layers since the X-ray diffraction data are of sufficiently poor quality for further structural refinement. However, molecular modelling suggests a close-packed struc- ture (Fig. 4) which is consistent with the lattice expansion and the observed stoichiometry. Mossbauer Spectroscopy Mossbauer spectroscopy is particularly suited to investigate the redox state of iron-containing guest species in intercalation compounds.'2 For compound 2 we were interested to deter- mine whether the driving force for intercalation was electron transfer from the ferrocene and reduction of the MOO, or donor-acceptor complex formation involving the amine sub- stituents.The "Fe Mossbauer spectrum of 2 at room temperature is shown in Fig. 5. The spectrum consists of a quadrupole Fig. 4 Schematic representation of the proposed packing of 1 viewed perpendicular to the a-b plane in MOO, 0.27 5.35 13 0.25 5.61 13 0.16 5.90 14 0.13 5.80 14 0.25 5.34 14 0.13 5.34 14 0.23 5.47 13 0.28 5.52 13 0.29 5.35 15 0.16 5.15 9 0.16 4.94 16 0.16 7.55 9 0.16 4.86 9 0.36 5.32 17 0.34 5.92 18 doublet centred at aFe=0.442(5) mm s-' and a quadrupole splitting A =2.344 mm s-'. The spectrum clearly indicates that the guest molecules are exclusively neutral ferrocene (Fe") and not ferrocenium ions.Typically ferrocenium (Fe"') salts give a single sharp resonance in the Mossbauer spectrum with an isomer shift dFe of ca. 0.55 mm s-'.', Photoelectron Spectroscopy We have investigated the changes that occur to the band structure of MOO, upon intercalation of 1 by solid-state UV- Photoelectron Spectroscopy (UV-PES) of cleaved singled crys- tals of the intercalate 2. The He I photoelectron spectrum of both MOO, and 2 are shown in Fig. 6. The He I PE spectrum of the host clearly shows the onset of photoemission at ca. 4.5 eV relative to the vacuum level and is consistent with the high electrical resistivity of this lattice. Upon intercalation we observe the appearance of new localised band-gap states (a) which were absent in the spectrum of pure MOO,.Additional features labelled (b),(c) and (d) correlate closely with the emission band in the gas-phase PE spectrum of the neutral 2-aminoethylferrocene guest.' The new populated electronic states [band (a)]above the valence band but below the conduction band in 2 suggest that the intercalate would exhibit an electrical conductivity characteristic of a semicon- ductor. Infrared Spectroscopy The infrared spectrum of MOO, exhibits inter alia three distinct absorbances at 992, 881 and 614cm-', assigned to the terminal, Mo-O( l), doubly bridging, Mo-0(2), and triply bridging, Mo-0(3) stretches, respectively.' ' In con- trast, the infrared spectrum of 2 exhibits relatively weak absorbances for the Mo-O(l), Mo-0(2) stretches which are shifted to the higher frequencies of 1019 and 953 cm-', respectively.However, the absorbance assigned to the Mo-0(3) stretch remains intense and is shifted to slightly lower energy at 572 cm-'. These data suggest that upon intercalation of 1 there is a major distortion of the Mo-0 bonds along the a and b axes. It is generally believed that the Mo-0 bonds along the a and b axes are often broken upon co-ordination of donor ligands and that the Mo-0(3) bonds remain essentially unchanged if the bilayer structure of the MOO, octahedra is retained. '3C Solid-state NMR Spectroscopy Room-temperature magnetic-susceptibility measurements indicate that 2 is diamagnetic, and this material was studied by solid-state cross-polarisation magic angle spinning J.MATER. CHEM., 1991, VOL. 1 -3.0 -2.0 -1.0 I 060 I 1.0 I 2.o . 1 3-0 . velocitytmm s-’ Fig. 5 Room-temperature 57Fe Mossbauer spectrum of MoO,(FcCH,CH,NH,),,,, A I I I 0 4 a 12 16 binding energy/ eV B 1-1 I I 0 4 ti 12 16 binding energy/eV Fig.6 The solid-state UV-PES spectra of (A) MOO, and (B) MoO,(FcCH 2CHZNHZ)0.36 (S=secondary electrons) (CP-MAS) NMR spectroscopy. The room-temperature solid- state 13C CP-MAS spectrum of 1 as the hydrochloride salt is shown in Fig. 7. The spectrum is easily assigned by reference to the solution 13C NMR spectrum. The room-temperature 13C CP-MAS spectrum of 2 is also shown in Fig. 7. The spectrum closely resembles the 3C spectrum of the hydrochlo- ride except for the observation of an additional methylene resonance at 20.1 ppm.Preliminary spin-lattice relaxation experiments indicate that this resonance arises from a more rapidly relaxing species, which may be tentatively assigned to the CH2 group of either free neutral 2-aminoethylferrocene or to 2-aminoethylferro- cene complexed to Mo, analogous to the structure proposed for pyridine intercalated into M003.1 Synthesis and Spectroscopic Study of TaS2(FcCH2CH2NH2)o.2(H20)x=0.3-0.5)(x TaS2(FcCH2CH2NH2),.,(x =0.3-0.5) (3) was prepared by refluxing a solution of 2-aminoethylferrocene in DME for several days at 100 “C with TaS,NH,. Pre-intercalation of the host with ammonia followed by exchange of the ammonia for 2-aminoethylferrocene, was necessary to achieve intercal- ation, since 2-aminoethylferrocene does not intercalate into TaS2 directly. Intercalation of tantalum disulphide with 2-aminoethylferrocene changed its colour from a lustrous blue-black to a dull-black appearance.The intercalate was washed repeatedly with DME, dried in V~CUOand character- ised by elemental microanalysis, and powder X-ray diffraction. The elemental microanalysis data were variable and the C and H ratios were always best interpreted by including a small amount of H20 (we presume that the H20is incorpor- ated into the lattice when the TaS2 is pre-intercalated with undried ammonia). Using this synthetic procedure it was possible to intercalate the aminoferrocene into relatively large J.MATER. CHEM., 1991, VOL. 1 A 170 130 90 50 10 6 (PPm) 1 ,* 170 130 90 50 10 6 (PPm) Fig. 7 Solid-state 13C CP-MAS NMR spectra of (A) FcCH,CH,NH: C1- and (B) Mo03(FcCH2CH,NH2),,,,, (* =spin-ning sidebands) crystalline samples of TaS2(NH3) (ca.2 mm x 2 mm x 0.1 mm). Typically, it was found that essentially complete exchange of NH, for the guest had occurred after heating to 100°C for 7 days. X-Ray Diflraction Preliminary X-ray powder diffraction experiments were per- formed on microcrystalline samples of 3; indexing the 001 reflections gave a lattice expansion Ac =5.89 A. However, since we had relatively large crystalline samples of both TaS, and the intercalate 3 we attempted a more detailed structural characterisation of these materials using one-dimensional Fourier techniques on orientated single-crystal samples.Table 3 lists the 001 reflections observed for single-crystal samples of TaS2 and the intercalate 3, aligned with their layers perpendicular to the X-ray beam plane. In this geometry only 001 class reflections are observed, and so Fourier refinements of these data will only give information along the c axis of the cell. Seven 001 reflections were measured in the two-theta (28) range 5-130" using Cu radiation (Table 3). The intensities were corrected for absorption, Lorentz effects and polarisation effects. The structure-factor phases were calculated from the known host structure [space group P63/mm~; 2Ta in 2b; 4s in 4f; z=O.125].An OOz projection was then computed from the phased data and is shown in Fig. 8(a). Despite the small number of observed reflections due to the Bragg cut-off for Cu radiation, the plot corresponds closely to the reported 12.2 3.0 0.0 electron density (arb. units) (b) 24.0 20.0 Ta 16.0 <12.0 8.0 Ta 4.0 0.0 electron density (arb. units) Fig. 8 The one-dimensional electron-density synthesis for (a) TaS, and (b) TaS,(FcCH,CH,NH,)o~ 19, together with a schematic rep- resentation of the packing Table 3 Powder X-ray diffraction data for TaS, and the intercalate 3 14.69 100 0 0 2 6.02 6.06 7.37 13 0 0 2 11.98 11.95 29.53 24 0 0 4 3.022 3.03 1 14.76 100 0 0 4 6.00 5.97 44.86 50 0 0 6 2.0 19 2.02 1 22.27 16 0 0 6 3.988 3.983 61.08 39 0 0 8 1.516 1.515 29.88 13 0 0 8 2.988 2.987 78.83 7 0 0 10 1.213 1.212 37.66 20 0 0 10 2.386 2.390 99.22 1 0 0 12 1.006 1.010 45.64 28 0 0 12 1.986 1.991 125.33 4 0 0 14 0.867 0.866 53.85 25 0 0 14 1.701 1.707 62.36 17 0 0 16 1.488 1.493 71.25 6 0 0 18 1.322 1.328 80.65 3 0 0 20 1.190 1.195 89.80 1 0 0 22 1.091 1.086 a Indexed on hexagonal cell: c= 12.12( 1) A.* Indexed on hexagonal cell: c=23.90(3) A. J. MATER. CHEM., 1991, VOL. 1 21 1 Table 4 Powder X-ray data for a-Zr(HPO,), H,O and the intercalate 4 a-Zr(HPO,),H,O (ref.9) a-Zr(H FeCH2 CH2NH 201" I h 11.77 98 0 19.89 30 1 25.07 I00 1 27.83 9 2 33.95 21 1 34.26 21 -3 37.40 11 -1 41.18 5 2 41.79 6 0 42.84 7 -2 44.40 10 2 -2 -2 -1 0 5 1.30 8 2 a Indexed on monoclinic hexagonal cell: (I=5.3 A, c k 1 dobslA dcalclAQ 291" I h k 0 2 7.5 12 7.539 4.00 100 0 0 1 0 4.460 4.470 8.1 1 67 0 0 0 2 4.440 9.6 1 55 0 0 1 2 3.549 3.554 12.10 61 0 0 0 4 3.525 19.50 50 1 0 0 2 3.203 3.214 20.36 42 1 0 1 4 3.204 2 1.90 36 1 0 1 4 2.638 2.640 23.22 60 1 0 2 0 2.635 24.45 41 1 0 1 2 2.61 5 2.622 25.15 43 1 0 1 6 2.402 2.398 28.56 28 1 0 1 4 2.190 2.189 33.8 1 36 1 1 2 4 2.160 2.160 64.99 60 3 0 2 4 2.109 2.1 11 78.13 57 3 1 2 2 2.039 2.038 2 4 1.779 1.777 cell: a =9.07(7) A, b =5.27(1) A, c = 16.2(1) A, B = 1 1 1S(3)'.$-hexagonal cell: =44.14(20)A. 2)O. 5(H2 O)O.5 1 &bSIA dcalclAb 2 22.07 22.07 4 10.89 11.03 5 9.19 8.82 6 7.30 7.35 1 4.55 4.56 3 4.36 4.38 5 4.06 4.07 6 3.83 3.89 7 3.64 3.71 8 3.54 3.52 10 3.12 3.18 2 2.65 2.63 10 1.43 1.44 3 1.22 1.26 a =5.24 A, c =22.6 A. Indexed on structure of 2H-TaS2 [S,(obs.) =0.1 15 us. Sz(lit.)=0.1 25].7 Ripples in the interlayer region (emanating from the strong Ta peaks) are due to series termination effects. A total of 11 001 reflections were measured in the 28 range 5-90' (Table 3) for intercalate 3 using Cu radiation. Since the tantalum atoms are the dominant X-ray scatterers in the material the initial phases were those computed for the host TaS,.The resultant one-dimensional electron-density syn-thesis is shown in Fig. 8(b). The principal difference between Fig. 8(b) and the electron-density synthesis to the host is the substantial increase in the electron density in the interlayer region, which is attributable to the guest molecule. Unfortu- nately, the shape and size of the electron density assigned to the guests could not be modelled using various orientations of the 2-aminoethylferrocene. The presence of an indetermi- nate number of disordered water molecules is presumed to be a major contributor to the poor agreement between the observed and calculated X-ray intensities. Discrepancy factors (R,)were typically ca.25%. Synthesis and Spectroscopic Study of Zr(HP04)2(FcCH2CH2NH2)o.5(H20)x(x =0.1-0.5) Zr(HP0,)2(FcCH2CH2NH2)o,5(H20)x(x=0.1-0.5) (4) was prepared by stirring a solution of 2-aminoethylferrocene in methanol with a suspension of zirconium hydrogen phosphate which had been pre-intercalated with methanol to give Zr(HPO,),MeOH.' Pre-intercalation with MeOH was necess- ary to expand the host lattice layers before 2-aminoethylferro- cene can intercalate the lattice. The zirconium hydrogen phosphate turned from white to beige on intercalation of the 2-aminoethylferrocene. The intercalate was washed repeatedly with methanol, dried in uucuo and characterised by elemental microanalysis, and powder X-ray diffraction.The elemental analysis gave the stoichiometry range Zr(HPO), (FCCH~CH,NH~)O.,(H~O),(~=0.1-0.5). X-Ray Powder Diffraction The powder X-ray diffraction pattern for IX-Z~(HPO,)~ can be indexed using the alternative $-hexagonal cell" as given in Table 4.In this cell, the observed interlayer spacing in given by c/3=7.53 A. The X-ray diffraction pattern of 4 has been tentatively indexed using the same crystal system as the host. The first observed strong peak is indexed as 002, suggesting that the layers may have shifted resulting in two layers per hexagonal cell rather than three layers of the host. A substan-tial lateral movement of the layers upon intercalation of the guest would not be surprising as the interlayer hydrogen bonding of the host are broken.The increase in interlayer spacing of 14.54A is again consistent with the formation of a bilayer of guest molecules. Conclusion 2-Aminoethylferrocene readily intercalates into the layered host lattices MOO,, TaS2, and ~zr(HP0,)~ -H20 in a man- ner analogous to simple organic amines. The ferrocene moiety does not seem to contribute to the energetics of the intercal- ation reaction and does not transfer additional electrons to the host. The X-ray data are entirely consistent with the formation of a bilayer structure for the intercalates formed with the lattices MOO, and x-Zr(HPO& H20. We would like to thank the SERC for partial support, Dr. T. Gibb (University of Leeds) for the Mossbauer spectroscopy measurements, Dr.M. Thompson (Princeton University) for helpful discussions and A. Rohl for assistance with the molecu- lar modelling. References 1 F. R. Gamble, J. H. Osiecki, M. Cais, R. Pisharody, F. J. DiSalvo and T. H. Geballe, Science, 1971, 174, 493; Intercalation Chemis- try, ed. M. S. Whittingham and A. J. Jacobson, Academic Press, New York, 1982. 2 T. A. Pecararo and R. R. Chianelli, J. Catal., 1981, 67, 430. 3 L. M. Real, R. P. Tormo, M. M. Lara and S. Bruque, Marer. Res. Bull., 1987, 22, 19. 4 E. Rodriquez-Castellon, A. Jiminez-Lopez, M. Martinez-Lara and E. Moreno-Real, J. Inclusion Phenom., 1987, 5,335. 5 M. L. H. Green, J. Qin, D. O'Hare, H. E. Bunting, M. E. Thomp-son, S. R. Marder and K. Chatakondu, Pure Appl.Chem., 1989, 61,8 17; K. Chatakondu, M. L. H. Green, J. Qin, M. E. Thompson and P. J. Wiseman, J. Chem. SOC.,Chem. Commun., 1988, 223. 6 L. Kihlorg, Ark. Kemi., 1963, 21, 357. 7 J. F. Revelli, Inorg. Synth., 1979, 19, 35. 8 D. Lednicer and C. K. Hauder, Org. Synth., 1965, 5, 434; J. M. Osgerby and P. L. Pauson, J. Chem. Soc., 1961,4600. 9 U. Constantino, J. Chem. Soc., Dalton Trans., 1978, 402. 10 E. K. Davies, ChemX User Manual, Chemical Crystallography Laboratory, Oxford, 1975. 212 11 J. W. Johnson, A. J. Jacobson, S. M. Rich and J. F. Brody, J. Am. Chem. Soc., 1981, 103, 5246. 12 T. R. Halbert, D. C. Johnston, L. E. McCandlish, A. H. Thomp- son, J. C. Scanlon and J. A. Dumesic, Physica B, 1980, 99,128; H. Stahl, Znorg. Nucl. Chem. Lett., 1980, 16, 271. 13 V. E. Fluck and F. Hausser, 2. Anorg. Allg. Chem., 1973, 396, 257. 14 K. Chatakondu and J. C. Green, unpublished results, 1990. J. MATER. CHEM., 1991, VOL. 1 15 R. N. Hider and C. J. Wilkins, J. Chem. Soc., Dalton Trans., 1984, 495. 16 Crystallography and the Crystal Chemistry ofLayered Structures, ed. F. Levy, Reidel, Dordrecht, 1976. 17 A. Clearfield and G. D. Smith, Znorg. Chem., 1969, 8,431. Paper 0/03335I; Received 24th July, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100205

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 213-216 X-Ray Photoelectron Spectroscopy of New Soluble Polyaniline Perchlorates: Evidence for the Coexistence of Polarons and Bipolarons Michiko B. Inoue,ta Kenneth W. Nebesny," Quintus Fernandoa and Motomichi lnoueb aDepartment of Chemistrx University of Arizona, Tucson, AZ 85721, USA bCentro de lnvestigacion en Polimeros y Materiales, Universidad de Sonora, Apdo. Postal 130, Hermosillo, Sonora, Mexico An X-ray photoelectron spectroscopic (XPS) study was carried out on the following soluble polyaniline perchlorates and polyaniline base: (A) a perchlorate prepared in a single step by the use of copper(I1) perchlorate as an oxidative coupling agent; (B) a polyemeraldine base obtained by treating perchlorate A; and (C) a perchlorate obtained by the protonation of the polyemeraldine base.The N Is XPS peak of perchlorate C exhibited a more intense envelope on the higher binding energy side than the N Is peak of perchlorate A. The difference spectrum, with a peak maximum at 402.0 eV, can be assigned to the bipolaron nitrogen in perchlorate C, when referred to the previous spectroscopic study of the two perchlorates. The area of the bipolaron peak was 5% of the whole peak of perchlorate C, although the charged nitrogen was 30% of the total nitrogen. Polarons are the major charge carriers in the two conducting polyaniline perchlorates, and perchlorate C contains bipolarons (in 5% of the total nitrogen) that coexist with polarons. Keywords: Polyan ilin e ; X-Ray photoelectron spectroscopy The electrical conductivity of polyaniline is sensitively depen- dent on the degree of protonation of the polymer chains, and the transition between the metallic state and insulating state is reversible.The relation between charge-transport properties and the structures of polymer chains has been the subject of extensive studies of conducting polyanilines. Usually, electro- conducting polyaniline salts are insoluble in common solvents. This insolubility makes it difficult to perform the structural characterization of positively charged polymer chains. Recently, we have reported that the use of copper(I1) perchlor- ate as an oxidative coupling agent provides a highly conduc- tive polyaniline perchlorate (polymer perchlorate A in Table 1) that is soluble in dimethyl sulphoxide (DMS0).2 The corre- sponding polymer base (polymer B), which is obtained by treating the perchlorate with an alkaline solution, is also soluble in common organic solvents such as DMSO and tetrahydrofuran.The protonation of the polymer base with perchloric acid yields a conducting perchlorate (polymer per- chlorate C). The compositions of these polymers are shown in Table 1. Structural information about these new polyanil- ines has been provided by the solution electronic spectra.2 The as-prepared polyaniline perchlorate (perchlorate A) shows no electronic absorption band in the region 700-900 nm,2 where quinone-iminium groups (i.e. bipolarons) exhibit an intense band.3 Accordingly, the positive charge is due to anilinium groups (i.e.polarons) rather than bipolarons, as shown in Fig. l(a). When the perchlorate is treated with an alkaline solution, deprotonation occurs at the anilinium cat- ions (polarons) so that quinone-imine groups are formed in the polyaniline base [Fig. l(b)]. The electronic spectrum of the polymer base shows an intense band at ca.600 nm. When the polyaniline base is treated with perchloric acid, the quinone-imine groups are preferentially protonated [Fig. l(c)]. The resulting perchlorate exhibits an intense elec- tronic absorption at 830 nm, suggesting the presence of bipolarons.' It is noteworthy that the chain structure of perchlorate C is different from that of the as-prepared per- t Permanent address: Universidad de Sonora, Apdo.Postal 130, Hermosillo, Sonora, Mexico. Table 1Electrical and XPS data of polyanilines: electrical conductivity a at 300 K, activation energy for electrical conduction E, above 200 K, binding energy, Eb,and full width at half-maximum height of N Is XPS peak polymer a/S cm - EJeV EJeV FWHM/eV A 3 0.04 399.6 3.O B 10-lo - 398.9 2.8 C 2 0.04 399.6 3.5 A, As-prepared perchlorate, [(-C,H,NH-)(CIO,),,, .0.4H20],; B, base, [fc6H4- NH -c6H4- NH+, --(-C6H4-N=C6H4=N~, .0.4H20], (x ~0.6);C, perchlorate prepared from the base, [1(36H4NH--XC104)o,4.0.6H20],. chlorate (perchlorate A). These polyanilines, therefore, are ideal model compounds for investigating the relationship between charge transport and polymer-chain structure.In the present work, we have carried out an XPS study of the three pol yanilines. Experimental Electrical and XPS data for polyanilines studied in the present work are presented in Table 1. Polymer A was prepared by an oxidative coupling polymerization of aniline by the use of copper(I1) perchlorate hexahydrate as an oxidative coupling agent:2 5.13 g of copper(I1) perchlorate hexahydrate (Aldrich) in 20 cm3 of acetonitrile (Merck, spectrum grade) was added dropwise to 0.6 g of aniline (Merck, distilled before use) in acetonitrile (20 cm3) with stirring under a nitrogen atmos- phere. The product was washed with acetonitrile in a Soxhlet extractor to eliminate oligomer components and possible contaminants.The polymer base (B) was obtained by treating the perchlorate with 4.5 mol dm-3 ammonia for 20 h.2 When the base was treated with 3 mol dmP3 perchloric acid for 3 h, a perchlorate (C) was obtained. The resulting perchlorate powder was washed with 0.1 mol dmP3 perchloric acid and dried in vacuum. The compositions of the compounds were determined by elemental analyses, which were performed by Huffman Laboratories, Golden, CO, USA. The electrical data J. MATER. CHEM., 1991, VOL. 1 t H t Fig. 1 Structures proposed by solution electronic spectra: (a)as-prepared poiyaniline perchlorate (A);(b)polyaniline base (B) obtained by treating the perchlorate with an alkaline solution; and (c) polyaniline perchlorate (C) obtained by protonating the polyaniline base with perchloric acid shown in Table 1 were determined for a compressed pellet of each compound by van der Pauw's four-probe method.2 The X-ray photoelectron spectrum was obtained with a Vacuum Generators ESCALAB MKII spectrometer with an Mg-Ka X-ray source (1253.6eVf).The sample chamber was maintained at 10-lo Torrz during each measurement. There was no indication of sample decomposition due to X-ray irradiation. The binding energies, Eb, were calculated by assuming that the peak maximum of the C Is spectrum in each sample had a binding energy of 284.6 eV. Results and Discussion Fig. 2 shows the XPS peaks of N 1s core electrons observed for the three polyanilines. The binding energies of the peak maxima are shown in Table 1.The C1 2p peak was located at 207.4 eV and the 0 Is peak at 532.1 eV for the two perchlorates. These binding energies have been calibrated by the use of the C 1s peak as an internal standard, because the binding energy corresponding to the peak maximum of the C 1s spectrum observed for polyanilines is substantially independent of the oxidation le~eI~-~although the full width at half-maximum height (FWHM) is influenced by the oxi-dation level (or dopant c~ncentration).~.~The binding energy of the C 1s peak maximum was assumed to be 284.6 eV as reported by Kang et aL5 and by Kumar et aL6 Snauwaert et d4employed the C12p peak as an internal standard assuming that its binding energy was 208.7 eV. There is a systematic difference of 1.3eV between the corresponding binding ener-gies reported by Snauwaert et aL4 and by the other group^.^.^ Obviously, this difference is caused by the selection of the internal reference standards.The binding energies can be discussed in a meaningful manner as long as they are cali-brated by the same method. The elemental analysis of the polymer base (polymer B in Table 1) showed that the composition was t 1 eVx1.602 x 10P9J. $. 1 Torrzl33.322 Pa. I I I 1 y 406 I 404 I 402 I 400 I 398 1-396 Fig. 2 N Is XPS peaks of: (a)the as-prepared polyaniline perchlorate (A); (b) the polymer base (B); and (c) the perchlorate (C) obtained from the base [fC6H4--NH-C6H4--NHj-, -x fC6H4N=C6H4=N-fx .0.4H20], with ~~0.6:~the value of x is approximate, because it cannot be determined accurately from elemental analysis alone.The composition of the polymer base is almost identical with that of the polyemeraldine base (x =0.5). The N 1s XPS peak was observed at 398.9 eV with an FWHM of 2.8 eV for the polymer base. This energy is 0.7 eV lower than the value 399.6 eV observed for the two perchlorates (Table 1 and Fig. 2). This chemical shift is related to the structural con-version caused by the deprotonation or protonation processes. Fig. 3-5 show the N 1s spectra normalized so that the three spectra have identical peak areas. When the peak of the as-prepared perchlorate [spectrum (a) in Fig. 31 is subtracted from that of the polymer base [spectrum (b)in Fig.31, the difference spectrum (b)-(a) is obtained. This difference spec-trum is attributable to a quinone-imine group formed by the deprotonation (Fig. 1). The peak maximum of the difference spectrum was located at 398.1 eV. Kang et a1.' reported that the N 1s peak of the polyemeraldine base could be decom-posed into two main components: one at a binding energy of 399.3 eV and the other at 398.1 eV. The authors assigned the former component to aniline nitrogen and the latter to J. MATER. CHEM., 1991, VOL. 1 404 402 400 398 396 EJeV Fig. 3 N 1s XPS peaks of: (a)the as-prepared polyaniline perchlorate (A) and (b)of the polymer base (B). The peaks are normalized so as to have an identical peak area. The peaks (a)-(b) and (@-(a) show the difference spectra 404 402 400 398 396 EJeV Fig.4 N 1s XPS peaks of (b)the polymer base (B) and (c)of the perchlorate (C)obtained from the base.The peaks are normalized so as to have an identical peak area. The peaks (b)-(c) and (c)-(b) show the difference spectra EJeV Fig. 5 N 1s XPS peaks of (a)the as-prepared perchlorate (A) and (c) of the perchlorate (C) obtained from the polymer base. The peaks are normalized so as to have an identical peak area. The peaks (c)-(a) and (a)-(c) show the difference spectra quinone-imine nitrogen by comparing them with the XPS data obtained for other nitrogen-containing polymers. The binding energy of the latter component agrees with that of the difference spectrum (b)-(a). This supports the assignment of the 398.1 eV peak to quinone-imine nitrogen.The subtrac- tion of spectrum (b)from spectrum (a) provides the difference peak, (a)-@), with a peak maximum of 401.0 eV (Fig. 3). This peak can be assigned to the nitrogen of anilinium ions involved in the as-prepared perchlorate (A). The peak area of the difference spectrum (a)-(6) is ca. 30% of the whole peak area. This value gives the proportion of the positively charged nitrogen atoms in the perchlorate, and it is reasonable when compared with the dopant concentration determined by elemental analysis. When the polymer base (B) is treated with perchloric acid, the quinone-imine groups are preferentially protonated and positively charged nitrogen atoms are formed in the resulting perchlorate (C).The difference spectrum (b)-(c) in Fig. 4 is therefore assignable to the quinone-imine nitrogen in the polymer base. Its peak maximum is located at 398.1 eV, which agrees with the binding energy of the peak maximum of the difference peak (b)-(a). The difference spectrum (c)-(b) is due to the positively charged nitrogen of perchlorate C. Its peak area is ca. 30% of the whole peak area of spectrum (c), suggesting that ca. 30% of the nitrogen atoms are positively charged. This is consistent with the dopant concentration determined by elemental analysis. The N 1s spectra of the two perchlorates exhibit intense envelopes on the higher binding energy side (Fig. 2), and the envelope of the perchlorate C obtained by the protonation of polyemeraldine base is more intense than that of the as-prepared perchlorate (A).The envelope on the higher binding energy side of an XPS peak is caused in part by the final- state effects such as an electron-hole pair excitation in the final state and screened us. unscreened final states in conduc- tors.’ To the first approximation, the final-state effects can be assumed to contribute equally to the two perchlorate spectra. Therefore, the peak asymmetry due to these effects is absent in the difference spectra obtained for the two perchlorates. Hence, the difference spectrum (c)-(a) with a peak maximum of 402.0 eV (Fig. 5) is attributable to an additional positively charged nitrogen species involved in perchlorate C.The solution electronic spectrum of perchlorate C has been reported to exhibit a 830nm band, which is characteristic of protonated quinone-imine groups (bipolarons), whereas per- chlorate A shows no band in the region of 500-900 nm.2 A molecular-orbital calculation has suggested that the polaron band appears at longer wavelengths.* On the basis of these results, it has been concluded that perchlorate C obtained from the polyemeraldine base involves bipolarons, whereas perchlorate A (the as-prepared polymer) has no significant number of bipolarom2 Therefore, the difference spectrum (c)-(a) can be assigned to bipolaron nitrogen. Kumar et aL6 reported that the N 1s XPS peak of an electrochemically prepared polyaniline salt consisted of four components at 398.5, 399.5, 400.8 and 402.2eV.The authors have proposed that the N 1s peak of the bipolarons appears at a higher binding energy than that of the polarons by comparing the charge-delocalization properties of the two charged species; the 402.2 eV peak was assigned to the bipolaron nitrogen and the 400.8eV peak to the polaron nitrogen. This assignment has been confirmed experimentally by the present study. The area of the difference spectrum (c)-(a) is ca. 5% of the whole peak area of spectrum (c). Ca. 5% of the nitrogen atoms are, therefore, bipolaron nitrogens in perchlorate C. As discussed above, the proportion of the positively charged nitrogen amounts to ca. 30%. The difference may be attributed to the polaron nitrogen.This suggests that the polarons and the bipolarons coexist in perchlorate C and the polarons are in the majority. Ginder et aL9 have proposed that a bipolaron formed in a polyemeraldine salt by the protonation of a polyemeraldine base is dissociated to two polarons by an internal redox reaction, and the two resulting polarons migrate in succession so as to reduce the electrostatic repulsion energy. Since this model provides a satisfactory explanation for the paramagnet- ism that is induced by the protonation of polyemeraldine base, the authors have concluded that the major charge- transport carrier of polyemeraldine salts is a polaron rather than a bipolaron. It should be noted, however, that polarons (radical cations) are chemically more active than bipolarons, although the electrostatic repulsion energy of the former is lower than that of the latter. Moreover, interconversion between polarons and bipolarons requires considerable acti- vation energy, because it is accompanied by an sp3-sp2 configuration change.The relative stability between polarons and bipolarons is, therefore, dependent on the environment around the positively charged nitrogen. For example, the steric effects of the polymer chains and the electrostatic interaction with dopant anions may be important factors. These factors, which define the relative stability, are different from one nitrogen site to another in polymeric materials. Therefore, bipolarons and polarons can coexist in some polymer chains.The present XPS study demonstrates that the coexistence is realized in perchlorate C. Kang et al.’ have reported that the N 1s peak of polyemeraldine chloride (prepared by the use of ammonium persulphate as an oxidant) involves component peaks attributable to different kinds of positively charged nitrogen atoms: a stronger component is at ca. 401 eV and a weaker component at ca. 403 eV. The comparison with the results of our present study shows that the 401 eV component is due to polaron nitrogen and the 403 eV component to bipolaron nitrogen. Polarons and bipolarons also coexist in the polyemeraldine chloride, with the former being in the majority. In conclusion, the positively charged nitrogen is due to the J. MATER. CHEM., 1991, VOL.1 polarons in the as-prepared polyaniline perchlorate (A), and in perchlorate C obtained from the polyemeraldine base, a small number of the bipolarons coexist with the polarons. Since the charge-transport properties of the two perchlorates are essentially identical (Table l), the polarons are responsible for the charge transport in the conducting polyanilines. This is consistent with the polaron model proposed by Ginder et aL9 This work was supported in part by the University of Arizona Center for Advanced Studies in Copper Recovery and Utiliz- ation under Defense National Stockpile Center, Grant No. DN-004. The work at the Universidad de Sonora was sup- ported by the Direccion General de Investigacion Cientifica y Superacion Academica, SEP (Grant No. C90-07-0388- 1). References Proc. Int. Con$ Sci. Technol. Synth. Met., June 26-July 2, 1988, Santa Fe, NM, Synth. Met., 1989, 29. M. Inoue, R. E. Navarro and M. B. Inoue, Synth. Met., 1989, 30, 199. J. Tanaka, N. Mashita, K. Mizoguchi and K. Kume, Synth. Met., 1989, 29, El75 P. H. Snauwaert, R. Lazzaroni, J. Riga and J. J. Verbist, Synth. Met., 1986, 16, 245. E. T. Kang, K. G. Neoh, S. H. Khor, K. L. Tan and B. T. G. Tan, J. Chem. SOC., Chem. Commun., 1989, 695; E.T. Kang, K.G. Neoh, T. C. Tang, S. H. Khor and K. L. Tan, Macromolecules, 1990, 23, 2918. S. N. Kumar, F. Gaillard, G. Bouyssoux and A. Sartre, Synth. Met., 1990, 36, 111. For example, G. K. Wertheim, Emission and Scattering Tech-niques, ed. P. Day, D. Reidel, Dordrecht, 1981, p. 61. S. Stafstrom, B. Sjogren and J. L. Bredas, Synth. Met., 1989, 29, E219. J. M. Ginder, A. F. Richter, A. G. MacDiarmid and A. J. Epstein, Solid State Commun., 1987, 63, 97. Paper 0/03441J; Received 30th July, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100213

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 217-222 Alkyloxy-substituted CTTV Derivatives that exhibit Columnar Mesophases Virgil Percec,* Chang G. Cho and Coleen Pugh Department of Macromolecular Science, Case Western Reserve University, Cleveland, OH 44 106, USA The homologous series of 2,3,6,7,10,11,14,15-octaalkyloxytetrabenzo[a,d,g,jlcyclododecatetraenes(CTTV-n) con- taining n=4-15 methylenic units in the alkyloxy substituents has been prepared both by alkylation of octahydroxytetrabenzocyclododecatetraenes (CTTV-OH), and by cyclotetramerization of the corresponding 3,4- (dialky1oxy)benzyl alcohols using excess CF,CO,H. CTTV-4 is crystalline. All other compounds display an enantiotropic columnar mesophase in addition to melting and crystallization transitions. The phase-transition temperatures of the CTTV-n compounds prepared by cyclotetrarnerization are lower than those synthesized by etherification of CTTV-OH. Melting and isotropization temperatures decrease in a continuous manner as n increases, without any odd-even alternation.Keywords: Liquid crystal; Cyclotetraveratrylene; Columnar mesophase Recent reports from several laboratories have demonstrated that hexaalkyloxy-and hexa(alkanoy1oxy)-tribenzocyclo-nonatriene (hexaalkyloxy and hexaalkanoyloxy derivatives of cyclotricatechylene) exhibit columnar mesophases. ’-’ This columnar mesophase has since been named pyramidic’ since the cyclotriveratrylene (CTV) core exhibits a rigid cone- or crown-shaped conformation (Scheme 1 Structure 1).6-8 In contrast, the corresponding octa(alky1oxy)tetrabenzocyclodo-decatetraene tetramers based on the cyclotetraveratrylene (CTTV) core are conformationally flexible as shown in Scheme Structure 2.6-8 Nevertheless, the three octaalkyloxy and three octaalkanoyloxy derivatives of CTTV reported in the literature also form columnar me so phase^.^^' The synthesis of a polymer based on the CTTV mesogen has also been reported recently.’ Our primary research interest in polymers containing mesogenic units that exhibit columnar mesophases is in developing an efficient and short-step synthesis.The general strategy is to construct the mesogenic units responsible for the generation of a columnar mesophase during the polymeriz- CH,O‘ bCH, 1 6CH3 OCH, 2 Scheme 1 Most stable conformations of CTV and CTTV: (1) crown or cone conformation of CTV; (2) sofa or saddle conformation of CTTV ation process.One such mesogen-forming reaction may be the electrophilic cyclotrimerization or cyclotetramerization of 3,4-dialkyloxybenzyl alcohols to form derivatives based on either the CTV trimer or the CTTV tetramer. In order to be successful as both a polymerization reaction and a mesogen- forming reaction, the conversion must be quantitative and the reaction must occur with a high selectivity towards either the trimer or the tetramer. According to the literature,8 CTV is the main product in the electrophilic cyclooligomerization of 3,4-dimethoxybenzyl alcohol, with CTTV obtained only as a side product.We have recently developed a synthetic procedure which leads predominantly to the cyclic tetramer from both 3,4-dimethoxybenzyl alcohol and the 3,4-di-alkyloxybenzyl alcohols. l1 The present paper describes the synthesis and characterization of a complete series of octa- alky1oxy)tetrabenzocyclododecatetraene(CTTV-n) derivatives containing alkyloxy substituents with n =4-15. Most of these compounds were synthesized by both alkylation of the 2,3,6,7,10,11,14,15-octahydroxytetrabenzo[u,dgjlcyclodode-catetraene (CTTV-OH) and direct cyclotetramerization of 3,4-dialkyloxybenzyl alcohols. The mesomorphic behaviour of CTTV-n synthesized by both routes is discussed and compared. Experimental Materials I-Bromobutane (97%), 1-bromopentane (99”/0), I-bromohex- ane (98%), 1-bromooctane (%)YO), 1-bromodecane (98%), 1-bromododecane (98%), 1-bromotridecane (98%), 1-bromot-etradecane (98%) and 1-bromopentadecane (98%) were used as received from Aldrich.1-Bromoheptane (Fluka, 99%), 1-bromononane (Fluka, >%’YO),1-bromoundecane (Fluka, >98%), 3,4-dimethoxybenzyl alcohol (Lancaster Synthesis, 97%), 3,4-dihydroxybenzaldehyde (Lancaster Synthesis, 97%), trifluoroacetic acid (Fisher Scientific, 99.7%), and all other reagents were used as received. CHC1, and CH2C12 used interchangeably in the demethylation of CTTV and in the cyclotetramerization of 3,4-(dialkyloxy)benzyl alcohols were dried by distillation from CaH,. Techniques 200 MHz ‘H-NMR spectra were recorded on a Varian XL-200 spectrometer.Unless noted otherwise, all spectra were recorded in CDC1, with SiMe, as the internal standard. All spectra of 2,3,6,7,10,11,14,15-octahydroxytetrabenzo[a,dgj]-cyclododecatetraene (CTTV-OH) and of octaalkyloxytetra- benzocyclododecatetraene (CTTV-n) derivatives were recorded at 55 "C; all other spectra were recorded at room temperature. Purity was determined by high-pressure liquid chromatog- raphy/gel permeation chromatography (HPLC/GPC) with a Perkin Elmer Series 10 LC instrument equipped with an LC- 100 column oven (40 "C), an LC-600 autosampler, and a Nelson Analytical 900 Series data station. Measurements were made using a UV detector with CHC1, as solvent and a 100 A PL gel column (0.9 cm3 min-').A Perkin Elmer DSC-4 differential scanning calorimeter equipped with a TADS 3600 data station was used to deter- mine the thermal transitions which were read as the maximum or minimum of the endothermic or exothermic peaks. All heating and cooling rates were 20 "C min-'. Tabulated ther- mal transitions were read from reproducible second or later heating scans and first or later cooling scans. Both enthalpy changes and transition temperatures were determined using indium as a calibration standard. A Carl Zeiss optical polarized microscope (magnification 100x) equipped with a Mettler FP 82 hot stage and a Mettler FP 800 central processor was used to observe the thermal transitions and to analyse the anisotropic textures. Synthesis and Isolation of Cyclotetraveratrylene (CTTV)' ' A solution of 3,4-dimethoxybenzyl alcohol (5.0 g, 30 mmol) in CH2CI, (20cm3 was added dropwise to an ice-cooled solution of trifluoroacetic acid (25 cm', 0.32 mol) in CH2Cl, (200 cm3).This resulted in an immediate violet colour. After stirring the reaction mixture for 4 h in an ice-water bath, it was neutralized with aqueous NaOH and the two layers were separated. The solvent was removed on a rotary evaporator, and the residue was washed several times in a fritted glass with H,O and twice with acetone. The resulting white solid was then recrystallized from CHC1,-benzene (80 cm3 : 30 cm3) to yield 2.5 g CTTV (55%) in two fractions; purity 99.1%; m.p. 342 "C (literature5 319-321 "C). 'H-NMR: 6 3.62 (24 H, s, -OCH3), 3.80 (8 H, s, Ar-CH2-Ar), 6.60 (8 aromatic H, s).Synthesis of 2,3,6,7,10,11,14,15-0ctahydroxytetrabenzo-[a,d,gj] cyclododecatetraene (CTTV-OH) CTTV-OH was prepared by the demethylation of CTTV following the procedure of White and Ge~ner.~ Boron tribrom- ide (40cm3 1.0 rnol dmP3 in CH2CI2, 40mmol) was added dropwise to a solution of CTTV (2.81 g, 37.4 mmol OCH3) in dry CHCl, (50 cm3) at 0 "C. The reaction mixture was then refluxed for 1 h. After cooling to room temperature, water was slowly added, and the precipitate was collected. Recrys- tallization from water (50 cm3) containing a little acetone yielded 2.28g (78%) CTTV-OH. 'H-NMR: 6 (CDC1,-DMSO, d6) 3.30 (8 H, s -CH,-), 6.45 (8 aromatic H, s), 8.33 (8 H, S, -OH).Synthesis of Octaalkyloxy Derivatives of CTTV-OH (CTTV-n) CTTV-n derivatives (where n represents the number of carbon atoms in the alkyloxy chain) were prepared by the etherifi- cation of CTTV-OH with 1-bromoalkanes as in the following preparation of CTTV-8. A solution of CTTV-OH (0.40g, 6.6 mmol OH), K2C03 (4.0 g, 29 mmol), and KI (0.1 g, 0.6 mmol) in ethanol (40 cm3) and DMF (16 cm3) was stirred at reflux for 30 h with 1-bromooctane (5.0 g, 26 mmol). The J. MATER. CHEM., 1991, VOL. 1 reaction mixture was then filtered to remove KBr, and the KBr was washed several times with CHCl,. The filtrate was condensed and the crude product was washed several times in a fritted glass filter with methanol. Precipitation with methanol from CHC1, yielded 0.64 g (56%) octa(octy1oxy)- tetrabenzocyclododecatetraene as a white powder; purity 98.3Yo.When necessary, these compounds were purified by column chromatography using silica gel as the stationary phase and CHC1,-hexanes (I :1 or 2: 1) as the eluent. The results of the synthesis of the CTTV-n homologous series by etherification of CTTV-OH are presented in Table 1. With the exception of CTTV-5 (98.2% pure), CTTV-8, and CTTV-14 (98.8% pure), the purity of all products is 299.9%. Their 'H-NMR spectra are identical: 6 0.91 (24 H, t, -CH3), 1.31 (16[n-3) H, m, -[CH2],,-,-), 1.74 (16 H, m, -OCH,CH,-), 3.52 (8 H, s, Ar-CH,-Ar), 3.89 (16 H, t, -OCH,-), 6.60 (8 aromatic H, s). Synthesis of 3,4-(Dialkyloxy)benzaldehydes The 3,4-(dialky1oxy)benzaldehydeswere prepared in 48-85% yield by etherification of 3,4-dihydroxybenzaldehyde with 1-bromoalkanes as in the following example. A solution of 3,4-dihydroxybenzaldehyde (10 g, 0.15 mol OH) and K2C03 (90 g, 0.65 mol) in ethanol (250 cm3) was stirred at reflux for 24 h with I-bromopentane (22 g, 0.15 mol), The solvent was then removed on a rotary evaporator, and Et,O and water were added until two clear layers formed.The two layers were separated and the aqueous phase was extracted further with Et,O. The organic phase was dried over MgSO,, filtered, and the solvent was removed on a rotary evaporator. Recrystalliz- ation from ethanol yielded 10 g (50%) 3,4-(di-penty1oxy)benzaldehyde. 'H-NMR : 6 1.0 (6 H, t,-CH,), 1.5 (8 H, m, -[CH2I2-), 1.9 (4 H, m, -OCH,CH2-), 4.1 (4 H, q, -OCH,-), 7.0 (1 aromatic H meta to -CHO, s), 7.3 (2 aromatic H ortho to-CHO, s), 9.87 (1 H, s, -CHO).Synthesis of 3,4-(Dialkyloxy)benzyl Alcohols 3,4-(Dialky1oxy)benzyl alcohols were prepared in 58-96% yield by reduction of the corresponding 3,4-(dialky1oxy)benz- aldehydes as in the following example. A solution of NaBH, (0.46 g, 12 mmol) in 0.4 mol dm -NaOH (3 cm3) and absolute ethanol (40cm3) was added dropwise to a mixture of 3,4- (dipenty1oxy)benzaldehyde(5.0 g, 18 mmol) in absolute etha- nol. The mixture was stirred at 70 "C for 3 h. The solvent was removed on a rotary evaporator, and the residue was dissolved in a mixture of Et20 and water. Following further extraction of the aqueous phase with Et,O, the organic extracts were dried over MgS04, filtered and the solvent was removed on a rotary evaporator.Recrystallization from Et,O yielded 3.9 g (77%) 3,4-(dipenty1oxy)benzyl alcohol. 'H-NMR : 6 0.97 (6 H, t, -CH,), 1.46 (8 H, m,-[CH,],-), 1.84 (4 H, m, ---OCH,CH,-), 4.00 (5 H, m, -0CH2- and -OH), 4.60 (1 H, s,-CH20H), 6.86 and 6.94 (3 aromatic H, two s). Synthesis of CTTV-n by Cyclotetramerization of 3,4-(Di-alkyloxy)benzyl Alcohols In a typical procedure, trifluoroacetic acid (0.50 cm3, 6.5 mmol) was added to a solution of 3,4-(didecy1oxy)benzyl alcohol (0.20 g, 0.48 mmol) in CH2C12 (9.5 cm3). After stirring at room temperature for 4 h, the reaction (purple-coloured mixture) was terminated with triethylamine, and the resulting yellow organic layer was washed four times with water.The solvent was then removed on a rotary evaporator, and the residue was dissolved in CHC1, and precipitated in methanol J. MATER. CHEM., 1991, VOL. 1 Table 1 Synthesis of octa(alky1oxy)tetrabenzocyclododecatetraene (CTTV-n) by etherification of octahydroxytetrabenzocyclododecatetraene (CTTV-OH) with I-bromoalkanes" OH,' n mmol bromoalkane/mmol per mmol OH 4 5.7 5.1 5 5.7 4.6 6 6.6 4.6 7 8.2 2.8 8 6.6 4.0 9 5.7 3.4 10 5.7 3.9 11 4.9 3.5 12 4.9 4.1 13 5.3 2.0 14 2.5 3.7 15 2.5 3.7 " Reflux 30 h: DMSO &CO3/mmol per mmol OH KI/mmol per mmol OH EtOH/cm3 per mmol OH 4.4 0.10 6.I 4.4 0.10 6.1 4.4 0.092 6.1 3.6 0.037 6.2 4.4 0.092 6.1 4.4 0.10 6.1 4.4 0.10 6.1 4.4 0.12 6.1 4.4 0.12 6.1 1.o 0.025 2.8 4.4 0.24 8.1 4.4 0.24 8.1 DMF/cm3 per mmol OH yield (%) ~ 2.1 50 2.8 55 2.4 50 1.2 36 2.4 56 2.4 42 2.4 48 4.1 52 4.1 44 4.1b 63 6.1 31 6.1 17 to yield 0.14 g (19Y0) octa(decy1oxy)tetrabenzocyclodode-ca te t raene.When necessary, these compounds were also washed with acetone to remove any adhering cyclotrimer. The results of the cyclotetramerizations are presented in Table 2. All cyclo-tetramers prepared by this method elute as a single HPLC/ GPC peak. However, the chromatograms of CTTV-9, CTTV- 10, and CTTV-12 also contain a small, higher molecular weight, shoulder. Considering the high molecular shoulder as a part of the single peak, the purity of these CTTV-n deriva- tives is 299.9'/0.Results and Discussion Octasubstituted alkyloxy derivatives of CTTV-OH (CTTV-n) can be prepared by either of the two routes outlined in Scheme 2. As summarized in Table 1, CTTV-n homologues with n=4-15 were obtained in 17-56% yield by the etherifi- cation of CTTV-OH with 1-bromoalkanes. This results in pure cyclotetramers devoid of any other cycles. These cyclo- tetramers can therefore be used to determine accurately the transition temperatures and corresponding thermodynamic parameters of mesomorphic CTTV-n. However, our ultimate goal is to obtain discotic liquid-crystalline polymers by in situ co-cyclization reactions. Therefore, it was necessary to prepare the CTTV-n homologues by cyclotetramerization reactions, and compare the thermal behaviour of the resulting average cyclotetramers with the pure model compounds.Table 2 summarizes the reaction conditions that were used to obtain tetrabenzocyclododecatetraene with ether substitu- ents by cyclization reactions. CTTV-n derivatives were isolated in 15-25Y0 yield by cyclotetramerization of the corresponding benzyl alcohols using a large excess of trifluoroacetic acid. Although these conditions favour tetramer formation and result in quantitative, or nearly quantitative conversion,' The analogous CTV-n cyclotrimer and higher molecular Table 2 Synthesis of octa(alky1oxy)tetrabenzocyclododecatetraene (CTTV-n) by cyclotetramerization of 3,4-(dialky1oxy)benzyl alcohols n benzyl alcoholimmol CF,CO,H/mol per mol alcohol CH,Cl,/cm3 per mmol alcohol temperature time/h yield (Oh) 5 0.7 1 9.2 13.4 reflux 4 15 6 0.68 9.6 14.0 ambient 4 16 7 0.59 11.0 16.1 reflux 4 18 8 0.55 11.8 17.3 reflux 4 24 9 0.5 1 12.7 18.6 ambient 8 22 10 0.48 13.5 19.8 ambient 4 19 12 0.42 15.5 22.6 reflux 4 25 Scheme 2 Synthesis of CTTV-n derivatives by either etherification of CTTV-OH or by cyclotetramerization of 3,4-(dialky1oxy)benzyl alcohols weight compounds also form. The ratio of cyclotetra-mer :cyclotrimer :other products using this procedure is gener- ally 65 : 10: 25, as measured by integration of the 'H-NMR aromatic resonances at 6.84 (cyclotrimer), 6.60 (cyclotetramer), and 6.49 (higher molecular weight analogues).' ' The waxy CTV-n contaminants were removed by precipitating a CHC1, solution of the crude product in methanol, and when neces- sary, by additional acetone washes.Both procedures result in some loss of CTTV-n. Some of the samples, however, still contain small amounts of a higher molecular weight fraction which was not removed by precipitation. This is demonstrated by both HPLC/GPC and by 'H-NMR analyses. Although only the HPLC/GPC chromatograms of cyclotetramerized CTTV-n derivatives with n =9,10,12 have a higher molecular weight shoulder, all 'H-NMR spectra of this series show the same extraneous resonances.Fig. 1 compares the 'H-NMR spectrum of CTTV-5 ob- tained by cyclotetramerization of 3,4-(dipenty1oxy)benzyl alcohol with that obtained by etherification of CTTV-OH. In both cases, the resonances at 6.84 ppm (s, aromatic protons), 4.78 (d, Hax)and 3.56 (d, Heq) due to the CTV-5 cyclotrimer are absent. However, although the spectrum of CTTV-5 prepared by etherification of CTTV-OH is completely clean, that of CTTV-5 prepared by cyclotetramerization shows at least two additional resonances at 6.41 pprn and at 3.59 ppm due to higher molecular weight analogues.' ' All CTTV-n compounds were characterized both by differential scanning calorimetry and by thermal optical polar- ized microscopy. Representative DSC traces observed on heating and cooling CTTV-4 through CTTV-15 synthesized 0.9/-CH3 3.9/-OCH2-1 6.6/aromatic 1.7 -0CH2CH2-3.5/ArCH2Ar J J.MATER. CHEM., 1991, VOL. 1 by alkylation of CTTV-OH are presented in Fig. 2. With the exception of CTTV-4, which is crystalline, all compounds display an enantiotropic columnar mesophase in addition to melting and crystallization peaks. However, preliminary investigations by polarized optical microscopy indicate that some of the phases labelled as crystalline phases are in fact columnar mesophases. More extensive characterization is necessary to elucidate their identity. The phase-transition temperatures and the corresponding enthalpy changes of this CTTV-n series are summarized in Table 3.All mesophases exhibit textures that are characteristic of columnar mesoph- ases and are similar to those reported by Zimmermann et a[.' for other CTTV-n derivatives. Several representative optical polarized microscopic textures are presented in Plate 1. The heating and cooling DSC traces of the CTTV-n deriva- tives synthesized by cyclotetramerization are presented in Fig. 3, and the corresponding thermal transitions and enthalpy changes are summarized in Table 4. Comparison of the DSC traces from Fig. 2 and 3 show that the phase-transition temperatures of the CTTV-n series synthesized by cyclotetra- merization are lower than those synthesized by alkylation. Nevertheless all CTTV-n homologues synthesized by cyclo- tetramerization exhibit the same type of mesophase.Therefore, although the CTV-n and CTTV-n mesophases are not mis- cible,' the higher molecular weight impurities present in this concentration range in the CTTV-n compounds prepared by cyclotetramerization must be miscible with the mesophase 1 n =6 IKJ n =7k ki n =11 101 93h n = 13 k)i1d27 1 v -4 k D 112 i I-n=14k % lki 0 50 100 150 200 250 ,0 50 100 150 200 2 i0 temperature/' C Fig. 1 'H-NMR (CDCl,, 55 "C) spectra of CTTV-5 prepared by (a) etherification of CTTV-OH, and (b) cyclotetramerization of Fig.2 DSC heating (second) and cooling (first) scans (20°C) of 3,4-(dipentyloxy1)benzylalcohol CTTV-n derivatives prepared by etherification of CTTV-OH J. MATER. CHEM., 1991, VOL. 1 Plate 1 Polarized optical micrographs (100 x) of the textures exhibited by CTTV-n derivatives prepared by etherification of CTTV-OH : (a) CTTV-5, 180.6 "C, columnar mesophase obtained by slow cooling from the isotropic liquid; (h) CTTV-5, 25 "C, crystalline phase obtained by cooling from the columnar mesophase shown in part (a);(c)CTTV-6, 168.1 "C,formation of the columnar mesophase by slow cooling from the isotropic liquid; (d)CTTV-7, 160.7 "C, formation of the columnar mesophase by cooling from the isotropic liquid; (e)CTTV-12, 129.7 "C, formation of the columnar mesophase by cooling from the isotropic liquid; (f)CTTV-14, 117.6 "C,formation of the columnar mesophase by cooling from the isotropic liquid V.Percec et al. (Fucingp. 220) J. MATER.CHEM., 1991, VOL. 1 22 1 Table 3 Thermal transitions and thermodynamic parameters of Table 4 Thermal transitions and thermodynamic parameters of CTTV-n prepared by etherification of CTTV-OH" CTTV-n prepared by cyclotetramerization of 3,4-(dialky1oxy)benzyl alcohols" phase-transition temperatures/ "C and the corresponding n enthalpy changes/kJ mol-' (in parentheses) phase-transition temperatures/ "C and the corresponding n enthalpy changes/kJ mol-' (in parentheses)k 116.6 (3.10)k 222.6 (48.8) i i 212.6 (47.4) k 77.4 (3.88) k 5 k 46.3 (12.6) k 70.6 (9.53) k 170.7 D 184.8 (53.6)bi k 38.3 (13.8) k 69.8 (12.2) k 181.7 (6.98) D 190.4 (29.6) i i 176.7 D 162.3 k 153.0 (54.8)bk 148.7 (10.6) 18.0 (10.6) k i 182.1 (30.6) D 176.4 (5.66) k 49.9 (12.7) k 17.7 (11.8) k 6 k 32.0 (9.19) k 155.0 D 163.5 (60.6)bi k 31.9 (8.17)k 161.6 (27.0) D 172,4 (19.5) i i 156.4 D 151.0 k 142.7 (60.6)bk 21.5 (9.67) k i 163.9 (22.6) D 148.0 (29.8) k 18.0 (8.60) k 7 k 92.5 (39.1) k 125.9 (7.36) k 140.4 D 155.3 k/D 146.0 k/D k 90.1 (47.0) k 123.8 (9.49) k 139.1 (15.1) D 161.6 (19.3) i 150.0 D 155.3 (39.0)bi i 154.3 (19.9) D 131.5 (17.2) k 119.7 (9.91) k 82.0 (48.4) k i 146.0 (12.7) D 139.8 (8.94) D/k 132.0 (19.7) k 120.3 (12.0) k 17.1 (2.84) k 67.9 (19.0) k 103.9 (2.67) k 137.4 (32.1) k 80.5 (37.3) k D 154.4 (19.I) i 8 k 20.3 (22.3) k 70.6 (13.9) k 105.5 (2.26) k 139.3 i 146.6(18.7) D 120.4 (36.4) k 87.2 (5.74) k 37.8 (18.2) D 150.1 (56.4)bi k 4.8 (1.86) k i 141.8 D 136.9 (20.5)bD/k 124.4 (38.5) k 88.3 (6.44) k 39.1 9 k 102.1 (55.4) k 113.6 (33.8) D 147.9 (18.0) i k 9.9 (34.0)bk i 139.7 (17.3) D 107.7 (40.4) k 86.6 (53.7) k 9 k 102.9 (41.6) k 115.0 (25.8) D 135.2 (20.8) i 10 k 92.0 (44.6)k 112.9(39.8) D 140.3(16.4) i i 126.1 (13.0) D 108.3 (27.9) k 82.7 (32.9) k i 131.6 (14.5) D 100.5 (44.7) k 75.6 (45.8) k 10 k 92.3 (33.0) k 115.1 (25.3) D 130.8 (17.3) i 11 k 98.2 k 100.5 (103)bD 136.9 (14.4) i i 120.4 (12.2) D 100.9 (35.0) k 75.9 k 67.4 (27.0)bk i 127.2 (14.0) D 92.8 (42.8) k 82.8 (50.0) k 12 k 13.0 (I 1.4)k 83.6 (56.9) k 98.6 (27.5) D 116.9 (6.07) i 12 k 81.7 (85.2) k 90.7 (20.7) k 96.6 (50.1) D 131.2 (12.8) i i 107.5 (6.91) D 85.7 (31.2) k 65.7 (54.1) k 1.4 (10.4) k i 122.7 (1 1.8) D 86.7 (57.3) k 67.0 (82.0) k 13 k 92.9 (149) D 126.6 (14.3) i " k =crystalline, D =columnar mesophase, i =isotropic; first line of i 118.5 (13.6) D 76.7 (152) k data obtained on heating, second line on cooling.Overlapped with 14 k 97.4 (135) D 120.2 (11.1) i previous transition(s). i 112.4 (9.57) D 77.2 (138) k 15 k 95.7 (161) D 117.7 (11.1) i i 110.3(10.6) D 79.9 (174) k " k =crystalline, D =columnar mesophase, i =isotropic; first line of data obtained on heating, second line on cooling. Overlapped with previous transition. jUIn =5 n =6 177 n =7 0 2 4 6 8 10 12 14 16k n 81 n =8 n =9 124 I k 1 83 ICE IIn =IO In =10 1 101ioi In =12 In 1 I ....._.. 1 0 50 100 150 200 0 50 100 150 200 Fig. 4 The dependence of the phase-transition temperatures versus temperature/-C the number of carbons (n)in the alkyloxy substituent of CTTV-n derivatives prepared by etherification of CTTV-OH : (a)on heating, Fig.3 DSC heating (second) and cooling (first) scans (20 "C) of ( 0)crystal-mesomorphic transition, ( .) mesomorphic-isotropic transition, (A)crystal-isotropicCTTV-n derivatives by cyclotetramerization of 3,4-(dialky1oxy)benzyl(.) on cooling,(b)transition; iso-alcohols tropic-mesomorphic transition, (0) 222 1 c& 30 E 0 -xQ (d5 10 Ql I. I *o!. I. 1 -I -I 4 6 8 10 12 14 16 n Fig. 5 The dependence of the change in enthalpy associated with the mesomorphic-isotropic (heating, W), and isotropic-mesomorphic (cooling, Cl) transitions as a function of the number of carbons (n) in the alkyloxy substituent of CTTV-n derivatives prepared by etherification of CTTV-OH exhibited by pure CTTV-n.Similar experiments performed with CTV-n derivatives have shown that the mesophase of CTV-n prepared by cyclotrimerization is completely destroyed by small concentrations of impurities.' These results demon- strate that while cyclotetramerization of 3,4-dialkyloxybenzyl alcohols is a suitable polymerization reaction for the in situ formation of disc-like mesogens, the corresponding cyclo- trimerization reaction is not. This is consistent with the report of Kranig et al." that although polycondensates based on the flexible CTTV core form columnar mesophases, they were unable to obtain liquid-crystalline mesophases from polymers based on the rigid CTV core.The melting, crystallization, and isotropization transition temperatures of CTTV-n derivatives synthesized by alkylation are plotted in Fig.4 as a function of n. Both melting and isotropization temperatures decrease as the number of carbons J. MATER. CHEM., 1991, VOL. 1 in the alkyloxy chain increases. This decrease is continuous and does not display an odd-even alternation as in rigid rod- like mesogens. As shown in Fig. 5, the enthalpy changes associated with the mesomorphic-isotropic and isotropic- mesomorphic phase transitions also decrease continuously with increasing alkyloxy chain length. The absence of an odd-even alternation of the thermal transition temperature was also observed previously in a series of hexasubstituted CTV-n derivatives.' However, in contrast to the CTV-n homologues, which are strongly supercooled upon forming the mesophase compared to isotr~pization,~ the CTTV-n derivatives undergo very little supercooling upon forming the mesophase (Table 3).Financial support from the US Army Research Office is gratefully acknowledged. References 1 J. Malthete and A. Collet, Nouv. J. Chim, 1985, 9, 151. 2 A. M. Levelut, J. Malthete and A. Collet, J. Phys., 1986, 47, 351. 3 J. Malthete and A. Collet, J. Am. Chem. SOC., 1987, 109, 7544. 4 J. Malthete, A. Collet and A. M. Levelut, Liq. Cryst., 1989, 5, 123. 5 H. Zimmerman, R. Poupko, Z. Luz and J. Billard, Z. Naturforsch., Teil A, 1985, 40, 149. 6 J. D. White and B. D. Gesner, Tetrahedron Lett., 1968, 1591. 7 J. D. White and B. D. Gesner, Tetrahedron, 1974, 30, 2273. 8 A. Collet, in Inclusion Compounds, ed. J. L. Atwood, J. E. D. Davies and D. D. MacNicol, Academic Press, London, 1984, vol. 2, p. 97. 9 H. Zimmermann, R. Poupko, Z. Luz and J. Billard, Liq. Cryst., 1988, 3, 749. 10 W. Kranig, H. W. Spiess and H. Zimmermann, Liq. Cryst., 1990, 7, 123. 11 V. Percec, C. G. Cho and C. Pugh, Macromolecules, submitted. 12 D. Demus and L. Richter, Textures of Liquid Crystals, Verlag Chemie, Weinheim, 1978. 13 G. W. Gray and J. W. Goodby, Smectic Liquid Crystals, Textures and Structures, Leonard Hill, Glasgow, 1984. 14 V. Percec, C. G. Cho and C. Pugh, Macromolecules, submitted. Paper 0/03716H; Received 13th August, 1990.

ISSN:0959-9428    

DOI:10.1039/JM9910100217

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 223-231 Oxonol Dyes: X-Ray Crystallographic and Solid-state 13C Nuclear Magnetic Resonance Studies of some Organic Semiconductors Martin C. Grossel,*" Douglas J. Edwards," Anthony K. Cheetham,b Michael M. Eddy,b Owen Johnsonb and S. Roderick Postlet" a The Bourne Laboratory, Royal Holloway and Bedford New College, University of London, Egham Hill, Egham, Surrey TW200EX, UK Chemical Crystallography Laboratory, University of Oxford, Hooke Building, 9 Parks Road, Oxford OX1 3PD, UK llford Ltd., Mobberley, Knutsford, Cheshire WA 16 7HA, UK X-Ray structural studies, together with solution and solid-state multinuclear nuclear magnetic resonance (NMR) spectroscopy and d.c. electrical conductivity data, are reported for the hydroxypyridone trimethine oxonol dyes (3).The structural work reveals that, as in the related cyanine materials, the dye molecules can adopt two different types of packing arrangement: herringbone and infinite parallel stacks. Cations of the type R,X+ (X = N, P) favour infinite parallel stacks, whereas less symmetrical cations yield herringbone structures; 13C MAS NMR appears to provide a convenient probe for these structural features. All the materials are semi-insulating with the notable exception of one of the phosphonium dyes, which is a moderate semiconductor (€,=0.33 eV). Crystals of the Et,NH+ salt (3a) are monoclinic: a= 19.035(4) A, b=8.476(3) A, c= 17.131(4)A, /?=101.85(2)", space group P2,/n, Z=4, R=7.7 (Rw=7.3)%; those of the Ph,P+ salt (3b)are monoclinic: a=12.634(6) A, b= 11.802(9)&, c=12.875(9)A, p=95.69", space group P2/c, Z=2, R=6.4 (Rw=6.1)%; and those of the Me,NH+CH,CH,O-CO-C(Me)=CH, salt (3c) are monoclinic: a= 31.313(5) A, b= 7.436(3)A, c= 25.495(4) A, p= 100.63(7)",space group C2/c, Z=8, R=7.7 (Rw=6.5)%.Keywords: Oxonol dye; Crystal structure; Solid-state nuclear magnetic resonance spectroscopy; Organic semiconductor Much is known about the solid-state structures of a wide range of the cationic cyanine dyes,'.2 but until recently no studies had been reported of another important dye class, the anionic oxonols, which have the general structure l.3Etter et al.4 have described the structures of two polymorphs of the cyanine-oxonol salt (2, n= 1) containing a barbiturate trimethine oxonol anion.We are studying the properties of a further oxonol dye class, the hydroxypyridone trimethine oxonols (3), which are of importance as photographic underlayer dyes.5 We have found that such dye salts show a variety of electrical behaviour, ranging from semi-insulating to moderately semiconducting, depending on the counterion, and we have consequently been interested in determining the 1 2 t Present address: Coates Coatings International, Station Lane, Witney, Oxon OX8 6XZ, UK solid-state structures of these materials. We now wish to report the results of our work on the structures of three such compounds, 3a-3c, and the properties of several others. It will be shown that, as for the cyanines,'T2 the oxonol structures fall into two major structural classes: (i) infinite parallel stacks, and (ii) herringbone arrays.Me Me a M+=Et3NH+, R=Et, n= 1 b Mf=Ph4P+, R=Et, n=l c M+=Me2NH+(CH2)20*CO*C(Me)=CH2,R=Et, n= 1 d M+=Bu,N+, R=Et, n=l e M+=Ph3PfMe, R=Et, n=l (TTF'+),R =Et, n = 1 h M+ =Et,NH+, R=Et, n=O i M+=Et,NH+, R=Et, n=2 R=Bu, n=l /\Me H Experimental Instrumentation Solution 'H, 13C, and 31P NMR spectra were recorded in C2H6]-DMS0 on a JEOL FX90Q, and Bruker WM-250 and WH-400 spectrometers ('H and '3C spectra were referenced to TMS, and 31Pspectra referenced to external H3PO4). Solid-state high-resolution 13C and 31P CP MAS spectra were obtained on Bruker CXP 200 and MSL 300 instruments. Single-crystal UV-VTS spectra were recorded on a Nanospec Microspectro-photometer at the Home Office Forensic Sci- ence Laboratories, Aldermaston; no attempt was made to orient the crystals relative to crystallographic axes.Synthesis of Materials The hydroxypyridone dyes, 3a-3f, were prepared, as pre-viously de~cribed,~.~ by refluxing 1-ethyl-3-cyano-6-hydroxy-4-methyl-2-pyridone (0.2 mol), 1,1,3,3-tetramethoxypropane (0.1 mol), and the cation salt [e.g. Et3N for 3a, Ph,P+Br-for 3b, 2-(dimethy1amino)ethyl methacrylate for 3c, Bu,N+I -for 3d) (0.1 mol) in ethanol (100 cm3) for ca. 6 h. As it cooled the dye crystallised; it was isolated by filtration and then recrystallised from an appropriate solvent to give the hydroxy- pyridones reported below.Each salt had the following common spectroscopic features characteristic of the dye anion (see Fig. I for assignments): 6, 1.09[6 H, t, H(l) H(20)], 2.44[6 H, s, H(7), H(15)], 3.90[4 H, q, H(2) H(19)], 7.71[2 H, d, J= 13 Hz, H(I0) H(12)], 9.0[1 H, t, J=13 Hz, H(ll)]; 6, 13.1[C(1) C(20)], 18.5[C(7) C(l5)], 33.8[C(2) C(19)], 92.2[C(6) C(14)], 110.4[C(8) C(13)], 117.7[C(5) C( 17)], 120.9[C( 1 I)], 157.4[C( 10) C(12)], 158.1 [C(4) C(16)], 161.4 and 161.9[C(3) C(9) C(18) C(21)]; v,,, 2975-2930, 2200(C~N), 1660(C=O), 161 5(C=C), 1215(C-N)cm-'; i,,,(H20) 589 and 548shnm (E,,, 8.3 x lo4 and 3.53 x lo4 dm3 cm- mol-'). The spectral features assigned to the cation are reported with the-individual dyes as follows. (a) Triethylammonium 5-cyano-3-[3-(5-cyano-1-ethyl-4-met hyl-2,6-dioxo- 1,2,3,6-tetrahydropyridin-4-yIidene)prop-1-enyll-1-ethyl-4-methyl-6-oxo-1,6-dihydropyridin-2-01ate (3a) as black needles (68%), m.p.259-260 "C from methanol. 6, 1.21(9 H, 9, CH3), 3.12(6 H, 9, CH2); 6, 8.5(CH3), 45.9(CH2). (Found: C, 65.62; H, 7.13; N, 14.06%. C27H35N504 requires C, 65.70; H, 7.15; N, 14.19%). (b) Tetraphenylphosphonium 5-cyano-3-[3-( 5-cyano- 1-ethyl-4-methyl-2,6-dioxo-1,2,3,6-tetrahydropyridin-4-ylidene) prop-1-enyl]-1-ethyl-4-methyl-6-0~0- 1,6-dihydropyridin-2- olate (3b) as gold octahedra (%TO), m.p. 254-255 "C from methanol. dH 7.69-8.01(20 H, m, Ph-P+); 6, 117.6(d, J=89 Hz, C-P'), 130.4(d, J= 13 Hz, Ar-C), 134.5(d, J= 10 Hz, Ar-C), 135.3br(p-Ar-C); bp 23.6. (Found: C, 72.63 H, 5.45; N, 7.33%.C45H39N404P requires C, 73.96; H, 5.38; N, 7.67%). (c) 2-Acryloyloxy(dimethyl)ammonium 5-cyano-3-[3-(5-cyano- 1 -ethyl-4-methyl-2,6-dioxo-1,2,3,6-tetrahydropyridin-4-y1idene)prop-1-enyl] -1 -ethyl-4-methyl-6-0~0- 1,6-dihydro- pyridin-2-olate (3c) as blue needles (38%), m.p. 217 "C from ethanol. 6, 1.91br(3 H, s, CH3-C=C), 2.90[6 H, s, (CH,),N+], 3.49(2 H, m, CH2-0), 4.43(2 H, m, CH2-Nf), 5.36(1 H, s, NH), 5.76(1 H, m, =CH), 6.15(1 H, m, =CH); 6, 17.6(CH3-C=), 42.8[(CH3)2N+], 55.3(CH2), 58.6(CH2), 126.4(C=C), 135.1(C=C), 165.8(C=O). (Found: C, 62.97; H, 6.43; N, 12.54%. C29H35N506 requires C, 63.37; H, 642; N, 12.74%). (d) Tetrabutylammonium 5-cyano-3-[3-(5-cyano- 1 -ethyl-4- methyl-2,6-dioxo-1,2,3,6-tetrahydropyridin-4-yIidene)prop-1-J. MATER.CHEM., 1991, VOL. 1 7 27 26 CH2 5./ 3a 2526q27 1 I+ 45 30 42 36 37 3b 60 3c Fig. 1 Numbering scheme used for crystallographic and spectroscopic data enyll-1-ethyl-4-methyl-6-oxo-1,6-dihydropyridin-2-olate (3d) as metallic green needles (50°/0), m.p. 186-187 "C from meth- anol. 6, 1.09(12 H, t, CH3), 1.50(16 H, m, CH2Me and CH2Et), 3.31(8 H, m, CH2N); 6, 19.0(CH3), 23.0(CH2Me and CH2Et), 57.6(CH2N). (Found: C, 70.03; H, 8.74; N, 11.04%. C37H55N504requires C, 70.1 1; H, 8.75; N, 11.05%). (e) Methyltriphenylphosphonium 5-cyano-3-[3-(5-cyano- 1- ethyl-4-methyl-2,6-dioxo-1,2,3,6-tetrahydropyridin-4-y1idene)-prop-1-enyl] -1-ethyl -4- methyl -6-0x0 -1,6-dihydropyridin- 2 -olate (3e) as blue needles (89Oh), m.p.220-221 "C (decomp.) from methanol. &, 3.16(3H, d, CH3-P+), 7.84(15 H, m, ArH); 6c 7.4(d, J3ipi3c=56.1 Hz, CH,-P+), 119.8(d, Jp-c=88.3 Hz, Ar-C,-P+), 129.9(d, J,-,= 12.6 Hz, Ar-C3, C5), 133.l(d, Jp-C= 10.9 Hz, Ar-C2, c6), 134.7(brS, Ar-CC,); BP 23.7; these compare well with those previously reported for this ~ation.~(Found: C, 71.79; H, 5.61; N, 8.37%. C40H37N404Prequires C, 71.84; H, 5.58; N, 8.38%). J. MATER. CHEM., 1991, VOL. 1 (f) N-Methylmorpholinium 5-cyano-3-[3-(5-cyano-l-ethyl-4- methyl -2,6-dioxo-1,2,3,6-tetrahydropyridin -4-y1idene)prop- 1-enyl] -1 -ethyl-4-methyl-6-0~0-1,6-dihydropyridin-2-olate (3f) as purple needles (72%), m.p. 272 "C (decomp.) from methanol. 6, 2.83(3 H, s, CH3-Nf), 3.47(8 H, m, N-CH2-CH2-0), 9.60br(l H, N+-H); 6, 42.5-(CH3-N'), 52.5(CH2-0), 63.4(CH2-N+).(Found: C, 63.29; H, 6.13; N, 14.03%. C26H31N505 requires C, 63.27; H, 6.33; N, 14.19%). (g) The preparation of the tetrathiafulvalene (TTF) salt is reported fully elsewhere.* (h) Triethylammonium 5-cyano-3-(5-cyano- 1 -ethyl-4-methyl- 2,6 -dioxo -1,2,3,6 -tetrahydropyridin -4 -ylidenemethyl) -1- ethy1-4-methyl-6-0~0-1,6-dihydropyridin-2-olate(3h) was pre- pared as (3a) but using triethyl orthoformate in place of the tetramethoxypropane to give purple needles (53%), m.p. 186 "C (decomp.) from ethanol. 6, (C2H6]DMSO) 1.06(6 H, t, CH3CH2Ar), 1.18(9 H, t, CH3CH2N+) 2.31(6 H, s, CH3-Ar), 3.11(6 H, m, CH2N+), 3.83(4 H, q, CH,Ar), 7.96(1 H, S, =CH), 8.86br(l H, NH); 6, (C2H6]DMSO) 8.6(CH3CH2N+), 13.1(CH3CH,Ar), 19.2(CH3Ar), 34.3(CH2Ar), 45.7(CH2N+ ), 92.q CA,-CH 3), 1 1 3.2( CA,-CH= CAr), 117.5(C-N), 149.3(=CH-), 159.1(C-C-N), 161.3 and 161.6(C=O); I.,,, (EtOH) 549 nm (E,,, 8.90 x103 dm3 cm-' mol-').(Found: C, 63.92; H, 6.85; N, 14.70%. C,5H33N504 requires C, 64.22; H, 7.1 1; N, 14.98%). (i) Triethylammonium 5-cyano-3-[5-(5-cyano- 1-ethyl-4- methyl -2,6-dioxo -1,2,3,6- tetrahydropyridin -4-y1idene)penta- 1,3-dienyl] -1 -ethyl -4-methyl -6-0x0 -1,6-dihydropyridin -2 -olate (3i) was prepared as 3a but using glutaconalde- hyde dianil hydrochloride (0.1 mo1)6 in acetic anhydride (200 cm3) as solvent to give metallic green microcrystals (37%), m.p. 139 "c from acetonitrile. 6, ([2H6]DMSO) 1.21(15 H, m, 5 CH3-CH,), 2.39(6H, s, CH3Ar), 3.81(10 H, m, 5 CH,-Me), 7.75(4 H, m, CH=CH-Ar), 8.92(1 H, t, CH-(CH),-Ar); hC (['H6]DMs0) 8.5(N'-CH,-CH3), 12.9(CH3-CH2N), 18.5(CH3-Ar), 33.7(CH2 -N), 45.9(CH2 -Nf), 91 .9(CH3 -CAr), 110.4 (CA,-CH =CH), 117.6(CEN), 1233Ar -CH=CH- CH), 146.1(Ar-CH=CH), 156.9(Ar-CH=CH), 157.9 (C-CEN), 161.2 and 162.1(C=O); A,,, (H20) 680nm (E,,, 8.42 x lo4 dm3 cm- mol-').(Found: C, 66.70; H, 6.98; N, 13.38%. C29H37N504 requires C, 67.03; H, 7.18; N, 13.48Yo). (j) N-Methylmorpholinium 5-cyano-3-[3-(5-cyano-l-ethyl-4-methyl-2,6-dioxo-1,2,3,6-tetrahydropyridin-4-ylidene)prop-1-enyl] -1 -ethyl -4-methyl -6- 0x0- 1,6-dihydropyridin- 2-olate (3j) was prepared as outlined above (190/,) but using 1-butyl- 3-cyano-6-hydroxy-4-methyl-2-pyridoneand was found to crystallise in two habits: as blue needles, m.p.219 "C, and green cubes, m.p. 219 "C (from ethanol). 8, 0.91(6 H, t, CH,-C3H6), 1.30(4 H, m. Me-CH2-C2H,), 1.48(4 H, m, Et-CH2-CH2), 2.44[6 H, S, H(7), H(l5)], 2.83(3 H, S, CH3-N'), 3.36br(8 H, m, O-CH2-CH2-Nf), 3.83[4 H, t, N-CH2-C3H7, i.e. equivalent to H(2), H(19)], 7.77(2 H, d, J= 13 Hz), 9.04(1 H, t, J= 13 Hz), 9.60br(l H, N+-H); dC 13.7(CH3-C3H6), 18.5[C(7), C(l5)], 19.8(Me-CH2-C2H4), 29.6(Et-CHz-CH,), 38.6(N-CH2-C3H7), 42.5 (CH3-N+), 52.6 and 63.4(0-CH,-CH,-N+), 92.2 CC(6), C(14)l, 110.3CC(8), C(13)1, 117.7CC(5), C(17)1,120.9[C(1l)], 157.6[C(lO), C(12)], 158.2[C(4), C(16)], 161.6 and 162.1[C(3), C(9), C(18), C(21); vmaX 2957, 2869, 221O(C-N), 1665(C=O), 1612, 1493, 1355, 1310, 1263, 1202 cm- '; iL,,, (EtOH) 602 and 555sh nm (E,,, 1.68 x105 and 4.07 x lo4 dm3 cm-' mol-I).(Found: C, 65.91; H, 7.20; N, 12.80%. C3,#36N505 requires C, 65.55; H, 7.10; N, 12.74%). X-Ray Studies For the X-ray diffraction studies, either Cu-Ka or Mo-Ka radiation was used, in each case monochromated using a graphite single crystal. Data were collected on a CAD4-F diffractometer (Enraf Nonius Delft) controlled by a PDP8-A minicomputer. All subsequent computational work was per- formed on either an ICL 2980 computer or an ICL 1906A computer using the 'Crystals' suite of programs.' The crystal data for the three structures studied are shown in Table 1. In all cases the orientation matrix and unit-cell parameters were optimised by a least-squares refinement using the angular coordinates of 25 reflections.Lorentz and polaris- ation corrections were applied and the data were merged to give independent structure amplitudes with I >30(I), where I is the final observed intensity and o(I) the standard deviation derived from the counting statistics. No absorption corrections were made. The structures were solved by direct methods using MULTAN." In all structures except 3a, the final cycle of refinement involved positional and anisotropic temperature factors (isotropic for hydrogen), and a three-term Chebyshev series was used for the weighting scheme.' ' A full anisotropic refinement was not feasible for 3a because of the large number of atoms (36) and the relatively small data set (only 992 reflections).Consequently, isotropic temperature factors were used for the 17 atoms that are closest to the centre of the molecule, i.e. the 10 ring carbon atoms, the three bridging carbon atoms, the cyano group carbon atoms, and the methyl carbons attached directly to the aromatic rings. A final difference Fourier calculation gave no significant irregularities. For 3a and 3c the hydrogen atoms were placed geometrically, whereas for 3b hydrogen atoms were found from an electron- density difference Fourier synthesis. We note that, as in previous work on similar materials: the R factors are some- what higher than might normally be expected. Fig.2 shows the packing arrangements adopted by dyes 3a-3c. Fractional atomic coordinates for the non-hydrogen atoms in each of the structures are listed in Table 2, together with selected bond lengths in Table 3, and selected bond angles in Table4; the numbering schemes for the data are shown in Fig.1. Other structural data, including full sets of fractional atomic coordinates with anisotropic temperature factors, full tables of bond lengths and bond angles for the three data sets, together with additional figures, are collected in the Supplementary Data.? Attempts were also made to grow single crystals of 3d-3f, but only in the case of 3d were crystal suitable for data collection obtained. The structure of 3d was refined, but did not converge below R =8.35% (R, = 11.45%; a =8.399(5)A, b= 10.445(5)A, c =2 1.265(7) A, p =83.34", P2/n).Results and Discussion Solid-state Structures The structure of salt 3a shown in Fig. 3(a) reveals an almost planar anion (angle of twist as defined by the angle between the planes of the two hydroxypyridone rings, (=12", see Fig. 4) in a cisoid configuration (i.e. the nitrogen atoms of the hydroxypyridone rings are disposed cis across the trimethine bridge) with the N-ethyl groups oriented approximately per- pendicular but mutually trans with respect to the dye plane. The triethylammonium cation is located close to one of the external pyridone oxygens and appears to be hydrogen bonded to it [Fig. 2(a)]. The dye anions form arrays of tilted columnar ~~ ~~ -f Supplementary data available from the Cambridge Crystallo- graphic Data Centre: see Information for Authors, J.Mater. Chern., 1991, Issue 1. J. MATER. CHEM., 1991, VOL. 1 Table 1 Summary of crystal data, intensity collection and data processing for 3a-3c 3a 3b 3c compound C27H35N5O4 MW 493 19.035(4) blA 8.4763) CIA 17.1 3 l(4) PI" 10 1.85( 2) u/A3 2705.2 Dmlg cm -33 1.3 D,/g cm -1.3 z 4 space group P2,lncrystal shape needle size/mm 2.0 x 0.4 x 0.4 colour black crystallisation solvent EtOH radiation Mo-Ka p/cm -6.537 fvA 0.7 1069 data collection 8-28 T/K 294 28 limits/" 20 no. standards 3 frequency (scans per check) 100 no. reflections 6619 no. data points used in least squares 992 Rw W) 7.7 R (Yo) 7.3 stacks within the crystal lattice.Individual anions are paired with a separation of 4.03 8, within these stacks [Fig. 3(b)], and within the dimers neighbouring anions are slightly slipped (4 =75")and relatively inverted about the oxonol long axis such that the ion pairs are related centrosymmetrically. The anion 'dimers' are further apart (5.07 A) and show greater long-axis slip =58"). Consequently, neighbouring cat- ions within a 'dimer' are located on alternate sides of the stack [Fig. 2(a)], presumably to minimise electrostatic and steric interference, and lie in the channels separating neigh- bouring anion columns. The ethyl substituents of anion neigh- bours lie in a parallel fashion with respect to adjacent molecules within the stack.The stacking pattern observed within the anion columns is similar to that found for the dye cations in a number of cyanine~,'~-'~ including, for example, cyanine (TCNQ -)(TCNQo) salts. The anion columns form sheets within the crystal lattice, and within each sheet the dye anion columns are tilted parallel to each other. However, neighbouring sheets are relatively inverted to afford a herringbone lattice [see Fig. 2(a)].Similar behaviour is observed in some ~yanines,'~,'~but in the thiacarbocyanines is apparently only seen when there is a C(9) substituent which controls the stacking mode (i.e. in the centre of the trimethine bridge). The 'herringbone' stack is also seen in the methacrylate salt [3c; Fig. 2(c)].Once again the anion is almost planar (5 = 14") with nitrogens cisoid, and the cation appears to be hydrogen bonded to an external pyridone oxygen, but here the N-ethyl substituents are disposed mutually cis with respect to the dye plane [Fig.5(a)]. Dimers, in which the component anions are relatively inverted through a centre of symmetry, are again formed. The anions are 3.99 A apart and long-axis slipped by 66" [Fig. 5(b)],and the hydrogen-bonded cations are located at the opposite ends of the dimer. Neighbouring dimers are 3.63 8, apart (the shorter distance here arising since all alkyl substituents are oriented into the dimer [Fig. 5(b)] and the anionic columns have a long-axis slip of ca. 72". Adjacent layers of columns are relatively inverted in a herringbone fashion, but the overlap of columns within neighbouring layers c45H 3904N4P C29H35N@6 730 549 12.634(6) 31.3 13(5) 11.802(0) 7.436(3) 12.875(9) 25.495(4) 95.69( 3) 100.63(7) 1909.1 5834.7 1.3 1.3 1.3 1.3 2 8 P2/c c2/c octahedron needle 1.0 x 0.5 x 0.5 1.O x 0.4 x 0.4 gold dark blue EtOH EtOH Cu-Ka Cu-Ka 96.69 96.69 1.5418 1.5418 8-28 8-28 294 294 75 45 3 3 120 100 5478 4141 2679 1065 6.4 7.7 6.1 6.5 differs from that in 3a to accommodate the increased bulk of the cation.By contrast, in the tetraphenylphosphonium salt (3b) the dye anions are significantly twisted about the trimethine bridge (5=40.4") and form very shallow infinite parallel stacks [4 =30"; Fig. 2(b)] in which the interplanar separation of individual dye molecules is 6.44 A.Such shallow stacks have previously been observed in solid-state structures of J-aggre- gating cyanine dyes and are similar to those proposed for dye aggregates on silver halide surfaces, which are thought to be the most efficient arrangement for photosensitisation of silver halides.'q2*'5,16 However, there is no evidence for oxonol dyes showing sensitising behaviour. Within each dye anion, the ethyl groups are located trans across the dye plane and the nitrogen atoms are located cisoid as usual [Fig. 6(a)]. In each stack, the substituents all lie on the same side of the dye column (and are parallel to each other, i.e. on one side of the dye all are cis) but neighbouring stacks are relatively inverted [Fig.6(a) and 6(c)]. Neighbouring dye anions in a column are considerably long-axis slipped (by almost a complete anion length [Fig. 6(c)]. The tetrabutylammonium salt (3d) appears to adopt a structure in which the dye anion is rather less twisted (5 GZ25"), but once again the arrangement of molecules within the stack is of the infinite parallel type (4=30"), as found in 3b. It should be noted that in all cases (except 3c) the ethyl groups are trans disposed across the dye anion. The first conclusion to be drawn is that the two main types of stack observed in the cyanine dyes are also found for the oxonols. It is interesting to note that in the herringbone structures 3a and 3c the two halves of the oxonol dye molecule are unrelated by symmetry.Conversely, in the stacked struc- tures the two halves of each molecule are equivalent. We also draw attention to the fact that the herringbone structures are associated with less symmetrical cations of the types R3XR'+ and R2XR'H+ (X =N, P), which are hydrogen bonded to one end of the anion. Within such anions (i.e. 3a and 3c) there appears to be a degree of bond alternation in the trimethine bridge of the anion (Table 3). J. MATER. CHEM., 1991, VOL. 1 Table 2 Fractional atomic coordinates for non-hydrogen atoms in Table 2 (continued)3a-3c with estimated standard deviations in parentheses 0.5865(3) 0.1035(12) 0.5929(4) 1.192(2) 1.130(3) 0.765( 1) 0.5556(3) 0.0795( 14) 0.63 1 l(4) 1.232(1) 0.997( 3) 0.806( 1) 0.57 24 3) 0.1652(1 1) 0.5398(4) I .2809(8) 1.103(2) 0.9300(8) 0.6033(3) 0.1860( 13) 0.5057(5) 1.2702(8) 1.1 12(2) 1.001 3(7) 0.5276(3) 0.2068( 12) 0.52 15(5) 1.3303(8) 1.205(2) 1.0462(7) 0.5078(3) 0.2675( 12) 0.4715(5) 1.2102(7) 1.042(2) 1.0259(6) 0.4622( 3) 0.305q13) 0.4668(5) 1.2057(8) 1.057(2) 1.1020( 7) 0.4342( 3) 0.3668( 12) 0.421 l(4) 1.1524(6) 0.954(1) 0.9759(6) 0.3893(3) 0.3931( 12) 0.4243(5) 1.I578(7) 0.944(2) 0.9029(7) 0.3718(3) 0.3556( 13) 0.4735(4) 1.0904(7) 0.876(2) 0.9999(6) 0.361 7(3) 0.4549( 13) 0.3786(5) 1.0278(7) 0.783(1) 0.9618(6) 0.3 160( 3) 0.4902( 13) 0.3775(5) 0.9749( 6) 0.717( 1) 0.9994(6) 0.3574(3) 0.4913(13) 0.3293(5) 0.908 l(6) 0.6 16( I) 0.973 5(6) 0.4342(4) 0.4932( 15) 0.2774(5) 0.86 16(7) 0.561( 1) 1.0204(6) 0.437 l(4) 0.3 188( 17) 0.2479(5) 0.8837(8) 0.59 l(2) 1.0996(7) 0.4490(4) 0.4024( 14) 0.3720(5) 0.7962( 7) 0.469(2) 0.9941(6) 0.7561(3) 0.1676(14) 0.721 7(4) 0.7459(8) 0.407(2) 1.0374(7) 0.8 1 12(3) 0.2 169( 14) 0.6670(5) 0.7734(7) 0.421(1) 0.921 l(6) 0.7792(3) -0.0884(14) 0.6737(4) 0.8004(7) 0.439( 2) 0.7983(7) 0.7900( 3) -0.1 55q15) 0.6218(5) 0.85 18(9) 0.3 lO(2) 0.78 12(7) 0.8624(3) -0.2684(15) 0.6426(5) 0.8875( 7) 0.580(1) 0.8994(6) 0.9089(4) -0.2446(20) 0.6394( 6) 0.503(2) 0.4 19( 3) 0.815(1) 0.9220(4) -0.0988(18) 0.6203(5) 0.550(1) 0.456(2) 0.759(1) 0.9375(4) -0.3968( 18) 0.661 7(7) 0.548(1) 0.132(2) 0.829(1) 0.6476(3) 0.1495( 12) 0.5273(4) 0.472(1) 0.064(2) 0.797(1) 0.6593(3) -0.0689( 12) 0.7044(4) 0.512(2) 0.296( 3) 0.932(2) 0.2797(3) 0.5 155( 13) 0.3768(5) 0.543( 2) 0.197(4) 0.976( 2) 0.4 186(3) 0.4648(1 1) 0.3275(4) I .2222(6) 1.019(1) 0.8800(6) 0.7720( 2) 0.1 124(10) 0.6723(4) 1.3772(7) 1.278(2) 1.0846(7) 0.7105(2) 0.0522(9) 0.5937(3) 0.7066(8) 0.361(2) 1.0752(7) 0.5952(2) 0.2303(1 1) 0.4582(3) 0.8194(5) 0.481(1) 0.8755(5) 0.4865( 2) 0.3886( 12) 0.3653(3) 0.5498(8) 0.297(2) 0.8626(9) 0.3495(2) 0.54 15( 10) 0.2883(3) I .3356(6) I.162(1) 0.9054(6) 0.8355(2) -0.1379(9) 0.6183(3) 1.1099(5) 0.877(1) 0.8539(5) 0.8487(3) -0.3929( 12) 0.6664(4) 0.9230(5) 0.622( 1) 0.85 19(4) 0.7164(4) 0.334( 1) 0.8954(5) Solid-state 13C and 31PNMR Spectra The solid-state '3C NMR spectra of the tetraphenylphos- 0.009 l(6) 0.38 8 8(7) 0.1045(6) phonium (3b)and tetrabutylammonium (3d)salts are compar- 0.1192(5) 0.3 62 6(5) 0.0838(5) 0.1868(4) 0.1726(4) 0.1332(4) able to those found for solutions of these dyes (Fig.7). 0.2000(3) 0.0565(4) 0.1018(3) Assignments of the latter, which are reported in the experimen- 0.2600( 4) 0.0120(5) 0.1793(4) tal section, were deduced from a series of selective solution 0.1558(3) 0.0129(4) 0.0062( 3) {' H-' 3C} decoupling experiments, detailed analysis of fully 0.1659(5) 0.1 116(5) -0.0176(5) proton-coupled solution '3C NMR spectra, and {' 3C-1H}0.1007(3) 0.0866(4) -0.069q3) two-dimensional correlation (HSC'') experiments. Quadru- 0.0974( 3) 0.2070(4) -0.0455(3) 0.0506(3) 0.0430(4) -0.1649(3) polar broadening of those resonances due to carbon atoms o.oooqo) 0.1027(6) -0.250qO) attached to nitrogen is clearly evident in the solid-state 0.4898(3) 0.2984(4) 0.1337(4) spectra.0.5594(4) 0.2093(5) 0.1280(4) The spectra of the two ammonium salts 3a and 3c are,0.5539(5) 0.14 14(5) 0.0377(5) however, more interesting, for in each case there is seen to be 0.4774(5) 0.1615(5) -0.046615) 0.4081(5) 0.2487(6) -0.0396(4) a splitting of those peaks assigned to C(I)/C(20),C(6)/C(14), 0.4 128(4) 0.3 182( 5) 0.0489(4) and C(8)/C(13)(Fig.7). This result is in agreement with the 0.6 1 17(4) 0.4798(4) 0.25 3 6( 4) crystallographic studies which show that the two halves of 0.6901(4) 0.4644(5) 0.1879(4) the oxonol anion are inequivalent in these structures. Similar 0.7750(4) 0.5403(5) 0.1972(5) splitting phenomena are observed in the solid-state I3CNMR 0.7804(5) 0.6269(6) 0.2697(6) spectra of the Ph3P+Me salt 3e and the N-methylmorpholin- 0.7009(6) 0.6408( 6) 0.3345(6) 0.6157(5) 0.5685(5) 0.3259(5) ium salts 3f and 3j, each of which have the lower-symmetry 0.1343(3) 0.2428(3) 0.0575(3) cations.The case of 3e is particularly interesting since there 0.3103(4) 0.0656(5) 0.2400(4) is no possibility of conventional hydrogen bonding. 0.2182(3) 0.2099( 3) 0.2223(3) Solid-state CP MAS 31Pspectra of the Ph4P+ salt 3b and 0.0547(3) 0.2804(3) -0.1062(2) Ph3P+Me salt 3e each show a sharp singlet at 23.9 and OSOOO(0) 0.3872(1) 0.2500(0) 23.0 ppm, respectively, there being no evidence for more than one phosphorus environment in either structure. 0.6909(5) 0.0387(21) 0.4699(6) 0.68 18(4) 0.198q20) 0.4939(5) UV-VIS Spectra0.66 15(3) 0.0872( 14) 0.579 l(5) 0.6297(3) 0.0627( 12) 0.6 107(4) A key feature of the behaviour of dyes in solution lies in their 0.6455(4) 0.0 120( 14) 0.6629(5) ability to form aggregates that can significantly affect their Fig.2 Solid-state packing of dyes 3a-3c. (a) The Et,NHf salt (3a), (b) the Ph,P+ salt (3b), (c) the Me2NH+(CH2)20-CO-C(Me)=CH, salt (3c) optical behaviour.18 For example, the cationic cyanine dyes are believed to form two major types of aggregate stack on silver halide crystal surfaces.’” ’,19 In the H-aggregate, dye molecules on the crystal surface are stacked at a relatively steep angle ($x 60-90”) with little lateral displacement between neighbours. Such aggregation leads to the presence of shorter-wavelength bands in the UV-VIS spectra of these materials.By contrast, some dyes form a J-aggregate in which the stack is very shallow (optimally 4 x18-25” depending on the dye structure) and there is a large lateral displacement between neighbours. Such stacks lead to the presence of additional longer-wavelength bands in the UV-VIS spectra of the materials; these dyes behave as spectral sensitisers for silver halides, a role crucial in the photographic process.’ ’,16 Similar aggregate band absorptions are observed in solution. The anionic oxonol dyes also show aggregation behaviour in solution and in gelatin films, and we have observed more intense shorter-wavelength aggregation bands in single crys- tals of 3a (see Fig.8) consistent with a steep aggregate stack angle. By contrast, powder visible spectra (in Nujol) of the phosphonium salt 3b and the ammonium salt 3d, which form J. MATER. CHEM., 1991, VOL. 1 Table 3 Selected bond lengths measured for 3a-3c with estimated standard deviations in parentheses bond length/A bond 3a 3b“ 3c C(3)-N(1) 1.425( 17) 1.389(5) 1.389( 12) C(3)-0( 1) 1.236( 17) 1.239(5) 1.268(11) C(3)-C(4) 1.408(18) 1.44 l(6) 1.403( 13) C(4)-C(5) 1.432( 17) 1.434(6) 1.443( 13) C(4)-C(6) 1.351( 15) 1.383(6) I .378( 12) C(6)-C(8) 1.437( 15) 1.424(5) 1.427( 13) C(8)-C(9) 1.41 3( 15) 1.456(6) 1.427( 13) C(8)-C( 10) 1.400( 15) 1.412(6) 1.426( 12) C(9)- O(2) 1.248( 15) I .236(5) 1.235( 12) C(9)-N( 1) 1.4 16( 16) 1.4 14(5) 1.421(11) C(lO)-C(11) 1.4 10( 15) 1.393(5) 1.378( 13) C(l1)-C(12) 1.382( 15) 1.435(13) C( 12)- C( 13) 1.432( 14) 1.40 1( 12) C( 13)-C( 14) 1.394(15) 1.438( 13) C(13)- C(2 1) 1.4 15( 14) 1.436( 13) C( 14)- C( 16) 1.370( 15) 1.393(12) C( 16)- C( 17) 1.4 12( 16) 1.452( 13) C( 16)-C( 18) 1.423( 15) 1.428( 13) C(18)-O(4) 1.24 1( 14) 1.256( 11) C( 18)-N(4) 1.385(15) 1.377( 12) C( 2 I )-N(4) 1.434( 15) 1.418(12) C(21)-O(3) 1.240( 14) 1.222( 12) “Centrosymmetric anion, i.e.two halves of the anion are equivalent. Table 4 Selected bond angles measured for 3a-3c with estimated standard deviations in parentheses bond angle/” 3a 3b” 3c C(6)-C(8)-C( 10) 119.6(1.1) 120.3(4) 119q1.0) C(9)-C(8)-C( 10) 120.2(1.1) 120.2(4) 120.9(1.0) C(8)-C( 10)-C( 11) 129.9( 1.2) 128.1(5) 128.1( 1.1) C(10)-C(11)-C(12) 1 18.1 (1.2) 119.2(6) 114.2(1.1 ) C(1 1)-C( 12)- C(13) 128.8( 1.2) 126.3(1.1) C( 12)- C( 13)- C( 14) 120.3( 1.1) 117.8(1.1) C(12)- C( 13)- C(2 1) 119.3(1.1) 122.I( 1.O) ‘Centrosymmetric anion, i.e.two halves of the anion are equivalent. shallow solid-state stacks, show a broad absorption having A,,, ca. 680 nm. However, whilst oxonol dyes strongly absorb to silver halide surfaces, they do not show sensitising properties (the dye LUMO lying below the conduction band of the silver halide), and indeed act as densensitisers, trapping photoelec- trons.” The aggregation of these materials is instead exploited to increase the ability of gelatin coatings to absorb light over a broad area of the visible spectrum.The crystal structures reported here show that oxonol dyes form solid-state stacks reminiscent of those proposed for cyanine aggregates. How- ever only shorter-wavelength aggregation bands are observed for oxonol dyes in solution though solvation may well modify the nature of the solution aggregate structure. Caution is therefore required when extrapolating solid-state aggregate structures to those of species adsorbed on silver halide surfaces and in solution. D.C. Conductivities of Dye Salts The propensity of cyanine dyes to form one-dimensional stacks results in semiconducting behaviour in many simple saltsz1 and almost metallic conductivity in some of their TCNQ complexes.2’ We have therefore investigated the d.c.conductivities for the oxonol dye salts already discussed, J. MATER. CHEM., 1991, VOL. 1 n Fig. 3 Packing arrangement within two pairs of anion dimers in a column of 3a. (a) End view, (b)top view, (c) side view \ Fig. 4 Diagrammatic representation of the stacking angle, $J,and the anion twist angle, 5 together with data on some other salts that were prepared during the course of this work (Table 5). With the exception of the TTF'+ salt (see below), the data refer to measurements on compressed powders using silver paint contacts. In general, the materials are very poorly conducting and d.c. measure- ments for some samples (notably 3d and 3h) were restricted to a limited elevated temperature range (ca.SOOC), being bounded by the melting/decomposition point of the material n u ( 4 Fig. 5 Packing arrangement within two pairs of anion dimers in a column of 3c. (a) End view, (b)top view, (c) side view Y Fig. 6 Packing arrangement within two pairs of dimer units in neighbouring columns of 3b. (a) End view, (b)top view, (c) side view and the detection limits (noise level) of the conductance bridge. In other cases, particularly for the more conducting materials 3e and 3g, a much greater temperature range of measurement (100<AT/°C<180) was possible. All materials, with the notable exception of 3g, showed linear I-V characteristics suggesting good ohmic contacts. E, values were calculated from log CT versus l/Tplots, i.e. assuming the materials to be extrinsic semiconductors. Since sample purity (in an electrical rather than an analytical sense) is a major problem in such J.MATER. CHEM., 1991, VOL. 1 Table 5 D.c. conductivities of dyes 3a-3i dye cation, M+ 3a Et,NH+ 3b Ph,P + 3c Me,NH +(CH,),O.CO.C(Me)= 3d Bu,N+ 3e Ph,P+Me 3f ("1N+ /\Me H 3g TTF+ 3h Et,NH+ 3i Et3NH+ no. of bridge double bonds, n conductivity' a,/S cm- at 373 K E,/eVb 5.75 XIO-" 0.8 1 7.24 x 10-l1 0.96 CH2 9.98 x lo-'' 0.6 1 6.92 xlO-" 1.06 4.82 x lo-* 0.33 8.81 x 10-9 0.75 ca. 1 x10-3' d 3.05 x 10-l1 1.78 6.12 x lo-' 0.88 'Measurements were made on compacted powders using Ag paint contacts unless otherwise specified; bcalc. using aT=a.exp (-E,/k,T); 'pressure (Cu disc) contacts used; dNon-linear I-V behaviour: d.c. conductivity estimated by measurement of resistance at 0.1 K Fig. 7 Typical solution and solid-state 13C NMR spectra of oxonol dyes: (a) solution spectrum of 3a (in ['HIDMSO); (b) solid-state spectrum of 3a showing the typical 'herringbone-structure' splitting of the resonances at 6 13, 92, 110, and 118 ppm; and (c) solid-state spectrum of 3d which adopts an infinite parallel-stacked structure. compounds it is highly unlikely that the semiconductivity is truly intrinsic. We consider that these values should be treated with caution and be used rather as a qualitative guide for screening the electrical properties of the dyes. As is reported el~ewhere,'~a.c. dielectric spectroscopy provides far more useful information about both the contact and the bulk properties of these materials.Nonetheless, some observations on the d.c. data are worth making. Particularly noticeable is the dramatic increase in d.c. conductivity that is observed when the cation is changed from Ph,P+ (3b) to Ph3P+CH3 (3e). The other salts of lower- symmetry cations so far prepared do not show this effect, but it is notable that these latter are rather different since each can directly hydrogen bond with the anion. The TTF salt 3g has been described in detail elsewhere,8 the properties of this material being dominated by the TTF wavelengthhm 400 500 600 700 I \ . Fig.8 UV-VIS spectra of 3a as (a) a single crystal (solid line) and (b)in methanol solution (dashed line) radical cation. Attempts to measure d.c.conductivity using Ag paint contacts were hampered by silver diffusion into the sample. Accordingly, measurements were repeated using cop- per disc pressure contacts, but the d.c. measurements thus obtained showed non-linear I-Vcharacteristics, i.e.non-ohmic behaviour for the voltage range investigated (> 0.1 V). So far it has proved possible to carry out single-crystal d.c. measurements on only one of the dyes, 3a. These reveal conductivity values (along axis =7.0 x 10-lo S cm-', ~intermedaxis=2.0X10-10 Sun-', bsho~axis=5.0~10-11 Scm-') which are very close to those measured for the corresponding powder. Thanks are due to: Dr. T. C. Webb, Ilford Ltd., and SERC for a CASE studentship (to D.J.E.); Dr.G.E. Hawkes and Mr. P. Haycock (U.L.I.R.S. NMR Service, Queen Mary College, University of London), and Mrs. J. Hawkes and Dr. F. Galloway [U.L.I.R.S. NMR Service, King's College (KQC), J. MATER. CHEM., 1991, VOL. 1 University of London] for solution NMR spectra; Dr. C. Groombridge (U.L.I.R.S. NMR Service, Royal Holloway and Bedford New College, University of London), and Dr. N. J. Clayden (Inorganic Chemistry Laboratory, University of Oxford) for solid-state NMR spectra; Dr. M. Isaacs and Dr. D. Laing (Home Office Laboratories, Aldermaston) for solid- state UV-VIS spectra; and Dr. W. E. Long and Dr. G. P. Wood (Ilford Ltd.) for valuable advice. References D. Sturmer, in Special Topics in Heterocyclic Chemistry, ed.A. Weissberger and E. C. Taylor, Interscience, New York, 1977, vol. 30, ch. 8. D. L. Smith, Photogr. Sci. Eng., 1974, 18, 309; K.Nakatsu, H. Yoshioka and S. Nishigaki, Kwansei Gakuin University Annual Studies. 1980, 29,213. D. L. Farmery, D. J. Fry and J. P. Stonham, Br. Pat., 1978, 1 521 083. M. C. Etter, R. B. Kress, J. Bernstein and D. J. Cash, J. Am. Chem. SOC.,1984, 106, 6921. T. H. James, The Theory of the Photographic Process, Macmillan, London, 4th edn., 1977. S. W. Bland, Br. Pat., 1972, 1 278 621. T. A. Albright and W. J. Freeman, J. Am. Chem. SOC., 1975, 97, 2942. M.C. Grossel, D. J. Edwards, D. B. Hibbert and H. El Nil, Electrochimica Acta, 1989, 34, 425; D. J. Edwards, S.R. Postle, M. C. Grossel and T. C. Webb, Br. Pat., 1986, 29352. D. J. Watkin, J. R. Carruthers and P. W. Betteridge, CRYSTALS User Guide, Chemical Crystallography Laboratory, University of Oxford, Oxford, 1985. Figures were prepared using Chem-X, developed and distributed by Chemical Design Ltd., Oxford, UK. 231 10 G. Germain, P. Main and M. M. Woolfson, Acta Crystallogr., Sect. A, 1971, 27,368. 11 J. R. Carruthers and D. J. Watkin, Acta Crystallogr., Sect. A, 1979, 35, 698. 12 V. F. Kaminskii, R. P. Shibaeva and L. 0.Atovmyan, Zh. Strukt. Khim., 1973, 14, 700. 13 (a)D. L. Smith and H. R. Luss, Acta Crystallogr., Sect. B, 1972, 28 2793; (b)T. Kaneda, S. Yoon and J. Tanaka, Acta Crystallogr., Sect B, 1977, 33,2065.14 K. Nakao, K. Yakeno, H. Yoshioka and K. Nakatsu, Acta Crystallogr., Sect. B, 1979, 35,415. 15 Ref. 5, pp. 540-549. See also: (a) P. J. Wheatley, J. Chem. Soc., 1959, 3245 and 4096; (b) D. L. Smith and H. R. Luss, Acta Crystallogr., Sect. B, 1972, 28, 2793; (c) J. Potenza and D. Mastropaolo, Acta Crystallogr., Sect. B, 1974, 30, 2353; (d) T. Kaneda, S. Yoon and J. Tanaka, Acta Crystallogr., Sect. B., 1977, 33,2065. 16 K. Norland, A. Ames and T. Taylor, Photogr. Sci. Eng., 1970, 14, 295; C. Reich, W. D. Pandolfe and G.R. Bird, Photogr. Sci. Eng., 1973, 17,334. 17 A. E. Derome, in Modern NMR Techniques for Chemistry Research, Pergamon, Oxford, 1987, p. 245. 18 J. Griffiths, Colour and Constitution of Organic Molecules, Aca-demic Press, London, 1976; D.G. Duff and G. H. Giles, in Water-A Comprehensive Treatise, ed. F. Franks, Plenum, New York, 1975, vol. 4, ch. 3. 19 R. Steiger and F. Zbinden, J. Zmag. Sci., 1988, 32,64. 20 T. Tani, J. Zmag. Sci., 1987, 31 263; C. R. Berry, J. Photogr. Sci., 1973, 21,202. 21 M. Heider, P. Lochan and J. Neel, C. R. Acad. Sci., Paris, 1968, 267C,797; M.Heider and J. Neel, J. Chem. Phys., 1973, 3, 547. 22 B. H. Klanderman and D. C. Hoesterey, J. Chem. Phys., 1969, 51, 377. 23 M. C. Grossel, D. J. Edwards, R. M. Hill and L. A. Dissado, J. Am. Chem. SOC., submitted; M. C. Grossel, F. A. Evans, S. C. Weston, R. M. Hill and L. A. Dissado, Chemtronics, in the press. Paper 01038 12A; Received 21st August, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100223

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 233-238 Elastic and Coulombic Contributions to Real-space Hole Pairing in Doped La,CuO, Xiaozhong Zhang and C. Richard A. Catlow Davy Faraday Research Laboratory, The Royal Institution, 21 Albemarle Street, London WlX 4BS, UK The energetics of three types of possible polaron pair in doped La,CuO, have been investigated by computer- simulation techniques based on lattice-energy minimisation. The most energetically favourable pairing configur- ations have been obtained. O--O-pairing is calculated to be bound by 0.119 eV. For Cu3+-Cu3+ and Cu3+-O-pairing, interlayer configurations are found to be more favourable than intralayer pairing and the nearest-neighbour pairing is unfavourable. The binding energy of the pairs is found to be related to the pair distance and geometry of the pair.The Coulomb repulsive energy between hole pairs is estimated and the interlayer pairing is found to be more effective at screening Coulomb interactions between holes. The lattice distortion energy caused by polaron pairing is estimated and is taken into account in the calculation. Keywords: Oxide superconductor; Hole pairing; Polaron; Bipolaron 1. Introduction It has become increasingly clear that before a proper under- standing can be acquired of the recently discovered high-T, superconductors'y2 it will be necessary to provide a detailed mechanism for the hole-pairing process if the BCS theory provides the mechanism for the superconductivity of these new materials. For doped superconducting La,CuO,, which contains divalent ions substituting for the La3 + host cations, with charge compensation by hole formation, the question of whether such holes are predominantly in Cu 3d or 0 2p bands remains controversial. Experimental results3 indicate that holes are mainly localised in the d,l-y2 Cu orbital in undoped materials and that after the materials are doped additional holes are created in the oxygen p orbitals.Shiba and Ogata4 performed band-structure calculations on a two- dimensional Cu02 cluster and showed that the binding of two holes on the 0 2p orbital can occur. Balseiro et a/.' presented results to show that in a simple model including Cu 3d and 0 2p orbitals, pairing between holes can occur. Feinberg et aL6 reported band calculation on a one-dimen- sional Cu02 plane which indicated that pairing takes place in the 0 2p band.Sarma and Ramasesha' undertook calcu- lations of the binding energy of the hole pair within the extended Anderson Hamiltonian in two-dimensional Cu02, and reported that the hole pairing takes place primarily within the 0 2p band. A number of authors have suggested that polaron pairs could be responsible for superconductivity of high-tempera- ture superconductors.8-' Recently, Alexandrov' presented a systematic polaron theory of high-temperature superconduc- tors and indicated that polaron theory can give a satisfactory description of the basic properties of high-temperature metal oxide superconductors. Experimental results' have demon- strated that the charge carriers in La,CuO, and YBa2Cu307 are polarons (or bipolarons) and that polarons in YBa2Cu307 are situated in Cu02 planes with a 0.13 eVt excitation energy.It remains unknown how hole species on the sublattices of La,CuO, might pair and which sublattice is favourable to the formation of bipolarons in La2Cu04. Moreover, it is clear that, whatever the coupling mechanism, knowledge of the extent to which the lattice can screen Coulomb interactions is of considerable value. Catlow et al.' have investigated t 1 eVz1.602 x J. various possible pairing mechanisms in La,CuO, by computer simulation techniques' and reported that negative- U pro-cesses are energetically unfavourable compared to Cu 3d pairing and 0 2p pairing.The same computer-simulation techniques as those used in ref. 15 are employed in this study because they model accurately the Coulomb and polarisation energies, which are the major terms in any localised coupling process, and take account of the lattice distortion, which is essential in polaron and bipolaron stabilisation. Our concern here is first with which type of real-space pair (Cu3+-Cu3+ pair, O--O-pair and Cu3+-0- pair) and second which sublattices are favourable to the formation of bipolarons. It is important to stress that our approach, which employs a lattice-energy minimisation methodology, omits many import- ant terms, e.g. the antiferromagnetic coupling energy. (Esti- mates of these are, however, available from other sources as discussed in section 4.)The merit of our approach is that it includes detailed estimates of lattice distortion and Coulomb energies, which are difficult to make from other sources.2. Methodology Lattice and defect-energy minimisation techniques are used in this study. In our calculations, interactions between atoms in the lattice are represented by an effective potential including both Coulomb and short-range terms. Holes are treated as either Cu3+ or 0-species. The energy of a region of the crystal surrounding the hole (or hole pair) state is then minimised with respect to the coordinates of the ions within the region. The explicitly relaxed region contains ca. 200-300 ions. The response of the more distant regions of the crystal is calculated using approximate procedures based on con-tinuum models employing the relative permittivity of the material.This approach, which rests ultimately on the work of Mott and Littleton" has been used very extensively during the last 10 years in modelling the energies of defects and impurities in ionic and semi-ionic solids. Detailed reviews are available in ref. 20 and 21, and successful applications to defects in superconductors are reported in ref. 18 and 21-23. As noted, this method does not include quantum-mechanical terms explicitly. It does, however, give an accurate and reliable estimate of the Coulomb interaction between holes and the way in which this is modified by lattice relaxation and polarisation; indeed, our estimates of these terms are accurate J.MATER. CHEM., 1991, VOL. 1 within 0.01 eV, although the uncertainties in the total energies of the component holes are much higher. The interatomic potentials used for orthorhombic La2Cu0, are as in our previous The holes on the Cu and 0 sublattices are treated as Cu3+ and 0-polaron species, respectively. The short-range potential parameters for the Cu3+-0-interaction were taken to be the same as for the Cu2+ -02-interaction. The covalent interaction between pairs of 0-ions is modelled by a Morse potential using parameters appropriate for the isoelectronic F2 molecule as discussed in ref. 18. 3. Results In doped superconducting La,CuO,, Sr2+ substitutes for the La3+ host cation with compensation by hole formation.The holes may occur on the Cu or 0 sublattices and they are treated as Cu3 + and 0-species, respectively. Three types of possible polaron pair (Cu3+-Cu3+, O--O-and Cu3+-0-pair) may form in doped La2Cu0, and computer- modelling results of the energetics of all three types are presented below. 3.1 Cu3+ -Cu3+ Pairing The pairing of Cu3+ species (at Cu2+ sites) was studied for a variety of separations. A number of configurations have been studied, for which results are presented in Table 1 and Fig. 1, where the binding energy is given with respect to isolated substitutional Cu3+ ions and our sign convention is such that a positive sign indicates an energetically unfavour- able process. It is shown in Table 1 that configuration Cu(3) and con- 0.5 1 0.44 4 2 0.3 \ Lu' i.luo 0.2 \,\, \i1 A 0 2 4 6 8 10 12 dlA Fig.1 Binding energy, Eb, and Coulomb repulsive energy, E,, vs. pair separation for Cu3+ -Cu3 pair. 0,Intralayer interaction; +,+ interlayer interaction; A, Coulomb repulsive energy figuration Cu(4) are the most energetically favourable intra- layer pairs with a repulsion energy of 0.09eV, whereas configuration Cu(7) is the most favourable interlayer pair with a repulsion energy of 0.036 eV. Note that all interlayer Cu3+ pairs are much more energetically favourable than intralayer Cu3+ pairs and interlayer Cu3+ pairs are only slightly un- stable (ca. 0.04 eV). This repulsive energy is smaller than estimates of the attractive magnetic energies of 0.05-0.08 eV (discussed in section 4), which may lead to its stabilisation.It is worth stressing that the favourable pair sites are not the nearest-neighbour sites and that they are separated by 7-9 A. It is interesting to compare the binding energy of polaron pairs with the Coulomb repulsive energy because the repulsive Table 1 Cu3+-Cu3+ pairs: configurations and energies separation binding of pair energy configuration number d/A Eb lev intrala yer CW) 3.8 1 0.414Cu3' acu3. CU3' BCU3+CU(2) 5.41 0.254 CU3+mcu3+ 7.62 0.090 CU3' bEjcu3+ CN3) CU(4) 8.54 0.092 CU3+ CU3+ W5) 10.81 0.120 interlayerBC"3< Coulomb repulsive energy E,/eV 0.135 0.095 0.068 0.060 0.048 CU3' CU(6) 7.10 0.046 0.073 cu (7) 8.9 1 0.036 0.058 CUW 10.74 0.048 0.048 J.MATER. CHEM., 1991, VOL. 1 Coulomb energy between charged polarons evidently opposes bipolaron formation. As experimental data for the relative permittivity of La2Cu04 are not available at present, it is difficult to compare directly the binding energy of polaron pairs with the Coulomb repulsive energy. But we can calculate the static relative permittivity E from our potentials, giving E = 27.9 The Coulomb repulsive energies for various pair configur- ations are shown in Table 1 and Fig. 1. It is seen from Table 1 that the Coulomb repulsive energy for interlayer pairing is less than the Coulomb repulsive energy for intralayer pairing at almost the same separation of the pair.This means that interlayer pairing is particularly effective at screening Cou- lomb interactions between holes. It is seen from Fig. 1 that the binding energy for interlayer pairing is less than the Coulomb repulsive energy by ca. 0.02-0.03 eV. As our calcu- lation technique takes into account only Coulomb repulsive energy and lattice distortion energy we can use the difference between the binding energy and the Coulomb repulsive energy as an estimate of the lattice distortion energy, I/;att, obtained by pairing. The lattice distortion energy, Katt,is thus estimated as 0.02-0.03 eV, which is in agreement with the value of 0.02- 0.08 eV given by de Jongh' and the estimate of 0.03 eV of Egami for the case of a Cu3+ polaron pair.24 3.2 Cu3 -0-Pairing+ The results for the Cu3+-0- polaron pair in various con- figurations are presented in Tables 2 and 3 and Fig.2. It is seen from Table 2 that the most energetically favourable intralayer configuration is OCu(6), which has a repulsive energy of 0.053 eV with a pair separation of 5.79 A. This small repulsive energy is roughly competitive with the attractive magnetic energy. Interesting results are found for the interlayer pairing (see Table 3). The configuration OCu(12) has an extremely small repulsive energy of 0.006 eV at a pair distance of 6.76A and it is possible that this could become negative if more accurate interatomic potentials were available. The second favourable Configuration is OCu(15), of which the repulsive energy is only 0.041 eV.These two configurations have a common feature that the pairs are situated close to the (100) plane. It is again clear in Fig. 2 that the interlayer pairing is in general more energetically favourable than the intralayer pairing and that the favourable pairing sites are not in the nearest-neighbour position. The Coulomb repulsive energy is shown again in Table 2 and Fig. 2. Fig. 2 shows that only for configuration OCu(6) are the binding energies of intralayer pairs less than the Coulomb repulsive energy, whereas the binding energies of interlayer pairs are less than the Coulomb repulsive energy for several separations. It is seen again that interlayer pairing is particularly effective at screening Coulomb interactions between holes.Using the arguments discussed in section 3.1, the lattice-distortion energy Katt due to Cu3 + -0-interlayer pairing is estimated as 0.03-0.09 eV. 3.3 0--0-Pairing The results of O--O-pairing studies are shown in Table 4 and Fig. 3. We find stability for two types of oxygen bipolaron. For configuration O(l), the bipolaron is bound by ca. 0.06 eV, whereas bipolaron configuration O(2) is bound by ca. 0.12 eV. These two bound oxygen pairs are situated at the nearest- neighbour site (d= 2.66 A) and next-nearest-neighbour site (d=3.11 A), respectively. When the distance of the pair is larger we find a slightly unstable 0-pair [configuration Table 2 Cu3+-O-pairs (intralayer): configurations and energies Coulomb separation binding repulsive of pair energy energy configuration number d/A E,IeV E,IeV OCu( 1) 1.91 1.010 0.270 OCu(2) 2.46 0.228 0.210 OCu(3) 4.31 0.237 0.120 OCu(4) 4.64 0.129 0.111 OCu(5) 5.72 0.114 0.090 -3+m0 OCu (6) 5.79 0.053 0.089 ..3+@0-OCu (7) 6.89 0.145 0.075 cu3+ 0-OCu(8) 7.87 0.078 0.066 0-CU3' OCu(9) 7.89 0.137 0.065 CU3+ 0-OCu(l0) 9.53 0.092 0.054 J.MATER. CHEM., 1991, VOL. 1 Table 3 Cu3+-0' pairs (interlayer): configurations and energies Coulomb separation binding repulsive configurationDo-_ _ number OCu( 11) of paird/A 4.92 energy E,/eV 0.061 energy E,IeV 0.105 c"?+ OCu( 12) 6.75 0.006 0.076 CU3f Cu3' mo- OCu(13) 7.23 0.096 0.071 CU3' OCu(14) 7.76 0.079 0.067 cu3+wo- ocu (1 5) 8.64 0.041 0.060 r,,?+mo- OCu(16) 10.17 0.059 0.05 1 1.1 1.01 7 0.94 i !0.8-! 0.7-i i.0.6-i Lu' lu; 0.5- \~'.0.4- 0.3- A *\ \ 0 2 4 6 8 10 12 dlA Fig. 2 Binding energy, Eb,and Coulomb repulsive energy, E,, us. pair separation for Cu3+-O- pair. 0,Intralayer interaction; +, inter-layer interaction; A, Coulomb repulsive energy 0(3)] with a repulsive energy of 0.036 eV at d= 3.17 A and a slightly bound bipolaron [configuration 0(4)] with a binding energy of 0.001 eV at d =3.58 A. When d is larger than 3.81 8, all the configurations are energetically unfavourable. In order to distinguish pairing from phase separation of holes we calculated in addition to the binding energy of two holes (Eb,2), the binding energy of four holes with respect to hole pairs (Eb,4),25If E~,~is negative then phase separation would be expected.In the nearest-neighbour site pair and the next-nearest-neighbour site pair we find, however, a negative Eb,2and a positive &,4, which will promote superconductivity by leading to pair formation without phase separation. Compared to the Coulomb repulsive energy shown in Fig. 3 the non-Coulombic interaction energy is large being ca.0.2-0.3 eV when the oxygen pair is bound. It is found that after lattice relaxation, the distance between 0-species decreases by 0.674 8, in configuration 0(1) and 0.564 8, in configuration O(2). The relative change of the pair distances are 25% for configuration 0(1) and 18% for configuration O(2).This enhanced attraction is due to the covalent interac- tion between two 0-species, which as we have noted is modelled by a Morse potential in our simulations. 4. Discussion We find only slightly unstable interlayer Cu3+ pairs with repulsive energies Eb smaller than 0.05 eV. It should be recalled that our binding energy Eb takes into account the Coulomb repulsive energy Er and the lattice distortion energy rfatt, i*e. However, for the nearest-neighbour bipolaron, the magnetic energy Umagwill provide an additional attractive energy favouring bipolaron formation as argued in detail by de J~ngh.~,"He argued that, in general, the binding energy, Eb, needed to form a bipolaron will consist of several contri- butions and can be written as He estimated that the nearest-neighbour Cu3-t bipolaron can gain an attractive energy, Umag,of ca.0.05-0.10 eV due to antiferromagnetic exchange coupling between Cu3 + ions.' O This would not be sufficient to outweigh the unfavourable repulsive energy for this configuration reported in Table 1.This antiferromagnetic exchange energy is expected to decrease as the separation of the pair increases. If this energy is as large as 0.05 eV at large separations, formation of interlayer Cu 3d bipolarons will be possible. Moreover, pairing might be favoured further in the metallic superconducting state owing to metallic screening effects [normally modelled by a screened Coulomb potential of the form (e2/r)exp (-r/ A)], the magnitude of which, however, is difficult to estimate.Apart from the Cu 3d polaron pairing, the antiferromag- netic exchange interaction can also play an important role in the other two types of polaron pairing. Kuramoto and Watan- abe reported that the combination of a polaron effect and the exchange interaction between 0 2p and Cu 3d electrons could lead to electron pairing.26 Shiba and Ogata suggested that a favourable binding energy could occur with two 0 2p holes in real space and the origin of this binding is mainly magneti~.~ It would be useful if the antiferromagnetic exchange energy on the most favourable sites reported in this paper could be estimated to see how significant this magnetic energy is in the polaron pairing.An important feature of our results is the demonstration J. MATER. CHEM., 1991, VOL. 1 Table 4 O--O-pairs: configurations and energes Coulomb separation binding repulsive of pair energy energy configuration number d/A Ehlev E,lev 2.66 -0.059 0.194 3.1 1 -0.119 0.166 3.17 0.036 0.163 O-@O-3.58 -0.001 0.144 O-m0-0-Bo-3.81 0.228 0.135 3.81 0.127 0.135 4.40 0.246 0.117 4.74 0.298 0.109 4.93 0.269 0.105 0-0-5.37 0.315 0.096 0.4 to the Friedel oscillations, pairing may occur in the absence 0.31 of electron-phonon coupling. It is seen from Tables 1-3 that the binding energy for interlayer pairing is generally less than the Coulomb repulsive energy, whereas the binding energy for intralayer pairing is generally greater than the Coulomb repulsive energy.As the 0.oj +: difference between binding energy and Coulomb repulsive Ice 0.1-+;I energy is the lattice distortion energy, Vatt,it appears that the 1;lattice distortion for interlayer pairing is also favourable to the formation of polaron pairs in the cases of Cu3+-Cu3+-0.2-y. I., . , . , . 1. 1 and Cu3+ -0-pairs. Study of the interlayer interaction is of significance in understanding the superconductivity of high- T, superconduc-tors because most high- T,superconductors contain more than one Cu06 layer and the greater the number of layers the higher is T,. Our calculations, which find that nearest-neigh- bour interlayer pairing is more energetically favourable than the intralayer pairing, would suggest that interlayer pairs can cause an increase in T,.In future studies we aim to investigate the next-nearest-neighbour layer pairing. If this is more favourable than nearest-neighbour layer pairing it may help our understanding of why YBa2Cu307 compounds have higher T, than La,Cu04 compounds and why T1 compounds have higher T, than YBa2Cu307 compounds. Our demonstration that the binding energy of polaron pairs is strongly related not only to the distance of the pair but also to the detailed geometry of the site where the polaron is situated deserves particular emphasis. Suma et also reported that the microscopic structure of the electron pair affects the macroscopic response of the superconductor.It seems necessary to study the pairing mechanism of doped La,CuO, using three-dimensional rather than two-dimen- sional models because the latter omit important physical effects. Fig. 3 Binding energy, Eb(+), and Coulomb repulsive energy, &(A), vs. pair separation for 0--0-pair that the interlayer pairing is much more energetically favour- able than the intralayer pairing in the cases of Cu3+-Cu3+ and Cu3+-0- pairs. The origin of this effect is not clear. Stoneham14 has argued that Jahn-Teller terms may play an important role. These are not explicitly included in our calculations, but since our potential model correctly repro- duces the structure of the CuO, octahedron, this effect may at least be partially represented.Suma et ~21.~~studied the superconductivity in layered materials with intralayer and interlayer couplings and reported that interlayer pairing reduces the coherence length for the normal direction to the layer as compared to that for the intralayer pair. The reduction of coherence length in the [OOl] direction has been reported in the case of YBa,Cu,O, compound but not yet in the case of L~,CUO~.~' reported that Burmistrov and Dub~vskii~~ strong reduction of the Coulomb repulsion favours the super- conducting pairing of electrons in neighbouring layers. Owing Recently Allan and Mackrodt3' studied the dielectric con- tribution to bipolaron formation in tetragonal La2Cu04 and Nd2Cu04.They reported that, in La,Cu04, the interaction energy of the large bipolaron is close to the Coulomb repulsive energy and the smallest interaction energy is larger than 0.2 eV. The interaction energy for the small bipolaron is much less than that for the large bipolaron. They also found that the interplanar pairing is more favourable than intraplanar pairing for the Cu3+ -Cu3+ pair. The smallest repulsive interaction energy for the interplanar pair is 0.01 eV, whereas that for the intraplanar pair is 0.02 eV. An attractive interac- tion of 0.03-0.04 eV for the nearby 0-hole pairs in La2Cu04 is found. Their results are fully compatible with ours and the similarity of the results from two independent studies employing different potentials is encouraging.For Nd2 Cu04, Allen and Mackrodt find much greater difference between intraplanar and interplanar interactions than in La2Cu04. They find no confining interactions and the smallest interac- tion energy is ca. 0.09eV, which is much larger than that in La2Cu04. It seems that the difference in bipolaron results for these two materials may arise from their structural difference (there is a CuO, layer in La2Cu04, whereas there is Cu04 plane in Nd2Cu04). It is worth noting that the detailed structure of the sublattice may play an important role in the formation of the bipolaron. Finally we note that similar calculations were performed on hole interactions in NiO. The results differ from those presented here in that the repulsive interactions were much larger than in the calculation reported above, and no con- figurations were found in which hole pairing might be expected.Our results show therefore that the La2Cu04 struc- ture is especially effective at screening Coulomb interactions between holes. 5. Conclusions In summarising this paper we should repeat the points made earlier concerning the restricted nature of our techniques. Effects requiring an explicit quantum-mechanical treatment, e.g. those due to hole kinetic energy and magnetic coupling energies are not included, but we do include other key terms, notably lattice relaxation and polarisation energies, which are difficult to treat using other methods. We are able to provide estimates of these terms whose magnitude must be important in determining the order of hole coupling energies.Our main conclusions are as follows. Three types of polaron pairing have been studied by computer-modelling techniques. The most energetically favourable Cu 3d polaron pair has an interlayer geometry, is slightly unstable (0.036eV) and has a separation distance of 5.91 A. The most probable Cu3+-0- polaron pair is un- stable by only 0.006 eV with a pair distance of 6.76 A. Three bound 0 2p bipolarons are found. Two of them [configur- ations 0(1) and 0(2)] are appreciably bound by 0.059eV at a distance of 2.66 A and by 0.1 19 eV at a distance of 3.1 1 A, respectively. The third bound 0 2p bipolaron is configuration 0(4),which is only slightly bound by 0.001 eV at a distance of 3.58 A.Interlayer pairing is found to be much more energetically favourable than intralayer pairing in both Cu3 + -Cu3 and+ Cu3+ -0 pairs. We found that the nearest-neighbour pairing is not energeti- cally favourable in the cases of Cu3+-Cu3+ pairing and Cu3+ -0 -pairing, whereas the nearest-neighbour pairing is energetically favourable in the 0--0-pairing. The binding energy of the pair is found to be strongly related to the pair distance and geometry of the pair. It seems necessary to study the pairing mechanism in a three-dimen- sional structure rather than via a simple two-dimensional model. J. MATER. CHEM., 1991, VOL. 1 Cu3+ bipolarons can form if the antiferromagnetic exchange energy is greater than 0.04eV.Cu3+-0- pairing is more favourable than Cu3+ -CuJf pairing and needs only 0.006 eV additional energy to result in binding. 0 2p bipolarons can occur, bound by ca. 0.12 eV. We have estimated the lattice distortion energy obtained by polaron pairing. The Cu3+-Cu3+ pair can gain 0.02- 0.03 eV in lattice distortion energy, whereas the gain by Cu3+-0-pairing is 0.03-0.09 eV. The largest gains occur in the O--O-pairing, which is enhanced by the covalent interaction between these species. Finally we should stress that we are not necessarily arguing for 'bipolaron' models of superconductivity. Rather, we are suggesting that La2Cu04 is particularly effective at screening Coulomb interactions between holes, a feature which is surely relevant to the high-T, behaviour of the material.The authors are grateful to Dr. J. H. Harding, Dr. W. C. Mackrodt and Professor A. M. Stoneham for helpful dis- cussions. Dr. X. Zhang would like to thank SERC for a research grant. References 1 J. G. Bednorz and K.A. Muller, Z. Phys. B,1988, 64, 189. 2 M. K. Wu, J. R. Ashburn, C. J. Torng, P. H. Hor, R. L. Meng,L.Gao, Z. J. Huang, Y.Q. Wang and C. W. Chu, Phys. Rev. Lett., 1987, 58, 908. 3 E. Fujimori, Takayma-Muromachi, Y. Uchida and B. Okai, Phys. Rev. B, 1987, 35, 8814. 4 H. Shiba and M. Ogata, J. Magn. Magn. Muter., 1988, 76/77, 59. 5 C. A. Balseiro, B. Alascio, E. Gagliano and A. Rojo, Ann. Phys. Fr., 1988, 13, 415. 6 D. Feinberg, M. Avignon, M. Boiron and B. K. Chakraverty, Ann.Phys. Fr., 1988, 13, 447. 7 D. D. Sarma and S. Ramasesha, Phys. Rev. B,1988, 39, 12286. 8 L. J. de Jongh, Solid State Commun., 1988, 65, 963. 9 L. J. de Jongh, Physica C, 1988, 152, 171. 10 P. Prelovsek P, T. M. Rice and F.C. Zhang, J. Phys. C, 1987, 20, L229. 11 N. F. Mott, Nature (London), 1987, 327, 185. 12 D. K. Ray, Philos. Mag. Lett., 1987, 55, 251. 13 A.S. Alexandrov, D.A. Samarchenko and S.V. Traven, Zh. Eksp. Teor. Fiz., 1987, 93, 1007. (Engl. Transl. Sov. Phys. JET, 1987, 66, 567.) 14 A. M. Stoneham, AERE Report, 1987, M3639. 15 C. R. A. Catlow, S. M. Tomlinson, M. S. Islam and M. Leslie, J. Phys. C, 1988, 21, L1085. 16 A. S. Alexandrov, Sov. Phys. JET, 1989, 68, 167. 17 Y. H. Kim, C. M. Foster and A. J. Heeger, Phys. Scr., 1989, 127, 19. 18 M. S. Islam, M. Leslie, S. M. Tomlinson and C. R. A. Catlow, J. Phys. C, 1988, 21, L109. 19 N. F. Mott and M. J. Littleton, Trans. Faraday SOC., 1938, 34, 485. 20 Computer Simulation of Solids, Lecture Notes in Physics, ed. C. R. A. Catlow and W. C. Mackrodt, Springer-Verlag, Berlin, 1982, vol. 166, p. 3. 21 C. R. A. Catlow, J. Chem. SOC., Faraday Trans. 2, 1989, 85, 335. 22 R. C. Baetzold, Phys. Rev. B, 1988, 38, 11304. 23 N. L. Allan and W. C. Mackrodt, J. Chem. SOC., Faraday Trans. 2, 1989, 85, 385. 24 T. Egami, Solid State Commun., 1987, 63, 1019. 25 J. A. Riera, Phys. Rev. B, 1989, 40,833. 26 Y. Kuramoto and T. Watanabe, J. Magn. Magn. Muter., 1988, 76/77, 571. 27 Y. Suma, Y. Tanaka and M. Tsukada, Phys. Rev. B., 1989, 39, 9113. 28 C. P. Pool Jr., T. Datta and H. A. Farach, Copper Oxide Supercon- ductors, Wiley, New York, 1988, p. 30. 29 S. N. Burmistrov and L. B. Dubovskii, Phys. Lett. A, 1989, 136, 332. 30 N. L. Allan and W. C. Mackrodt, J. Chem. SOC.,Faraday Trans., 1990,86, 1227. Paper 0/04135A; Received 1 lth September, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100233

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 239-244 Preparation and Crystal Structure of U-Phase Ln3(Si,-,A13+x)0,,+xN,_, (x 0.5, Ln =La, Nd) Per-Olov KaIl,*" J. Grins," P-0. OIsson,at K. Liddell,b P. Korgulb and D. P. Thompsonb a Arrhenius Laboratory, Department of Inorganic Chemistry, University of Stockholm, S-106 91, Stockholm, Sweden Wolfson Laboratory, Materials Division, Department of Mechanical, Materials & Manufacturing Engineering, The University, Ne wcastle upon Tyne NE I 7RU, UK U-phase Ln3(Si3-J413+x)0,2+xN2-x(Ln =La, Nd) occurs as a crystalline phase in rare-earth sialon ceramics formed by devitrification of grain-boundary glasses at 1000-1400 "C. The crystal structure of Nd U-phase has been determined from Cu-Ka X-ray powder diffractometer data and refined by the Rietveld full-profile technique to RF=0.028.The space group is P321 and the cell dimensions are a=7.974(1) A, c=4.873(1) A and V=268.29 A3. The structure is isomorphic with the La,Ga,GeO,, structure, and exhibits corner-shared layers of (Si,AI)(O,N), tetrahedra interconnected by A106 octahedra. The rare-earth cations occupy sites between the tetrahedral layers. Transmission electron microscopy and lattice imaging studies support the X-ray structural findings. The structural relationship of the U-phase to other nitrogen-containing ceramic phases is discussed. Keywords: Salon material; Grain boundary phase; Rietveld analysis; Nitrogen ceramic The crystal structures of phases in metal-silicon-oxynitride and metal-silicon-aluminium-oxynitride (sialon) systems are in many cases related to silicates and aluminosilicates, especially at compositions of high O:N ratio.They can, therefore, be described in terms of arrangements of Si(0,N)4 or (Si,Al)(O,N), tetrahedra occurring either as individual tetrahedra or joined together to form chains, rings, sheets or more complex networks, with metal cations occupying sites between these units to join the structure together and balance valencies.' A useful preparative route for obtaining new phases in sialon-based systems is the devitrification of sialon glasses at 1000-1400 "C. In this way, phases unstable at the high temperatures ( 1650-1800 "C) required for solid-state reactions can be prepared at lower temperatures and their structures characterized. The interest in these compounds is three-fold.First, sialon glasses occur as grain-boundary phases in com- mercial sialon ceramics and can be devitrified as part of the firing cycle to give a fully crystalline product with considerably improved high-temperature properties. Secondly, many sialon phases are glass ceramics, which are a relatively new class of materials.* Nitrogen glass ceramics offer improved refractori- ness and high-temperature mechanical properties compared with oxide glass ceramics, whilst still preserving the ease of fabrication and the densification characteristic of this class of materials. Finally, new crystalline sialons are of interest because of their crystal chemistry. During the last 20 years, attempts have been made to prepare nitrogen analogues of all the mineral silicates and although examples exist of most structure types, there is still considerable scope for preparing new analogues.The sialon U-phase was originally reported in the La-, Ce- and Y-sialon systems as a minor phase produced in P'-glass compositions by heat treatment at low (1000-1300 "C) tem- perature~.~X-Ray powder data were presented for this phase and interpreted on the basis of a body-centred cubic unit cell with a=9.6-9.8 A. At this stage, U-phase had been observed only in amounts of 10-20% in the grain boundaries of P'-sialon ceramics. and discusses the possibility of preparing pure samples of U-phase in the La-, Ce- and Nd-sialon systems. The crystal structure of Nd U-phase has been determined from X-ray powder diffractometer data and the La U-phase from Guinier- Hagg powder diffraction data by the Rietveld technique. The X-ray studies are corroborated by high-resolution electron microscopy (HREM) studies.Experimental Previously, the U-phase has been observed in the P'-YAG plane of the Y-Si-AI-0-N system, and in similar parts of the Ce and Nd sialon system^.^ Subsequent studies of heat treatment in the Y-Si-Al-0-N system close to the area reported previously and closer to the glass-forming region failed to produce more than trace amounts of U-phase. It is therefore believed that U-phase has lower stability in this system than e.g. the B-phase, Y2SiA105N.5 Further studies have therefore concentrated on compositions in the La-, Ce- and Nd-sialon systems, where U-phase displays a much higher stability.A detailed study of phase relationships in the Nd-Si-Al- 0-N system at compositions close to those reported in ref. 3 has been carried out. Compositions were prepared by mixing together weighed amounts of Si3N4 (Starck-Berlin, Grade LClO), A1203 (Alcoa, Grade A17), Nd203 (Rare Earth Prod- ucts, 99.9%) and AIN (Starck-Berlin, Grade A) powders, mixing in isopropyl alcohol, drying and compacting by uniax- ial compression in steel dies to 5000 lbf Compacts of 1 g were then placed in BN-lined graphite crucibles and fired in a graphite element sintering furnace in a nitrogen atmos- phere at temperatures in the range 1600-1800 "C.Subsequent heat treatment was carried out in the temperature range 1000-1400 "C in nitrogen in an Sic-element furnace. After preparation, product phases were identified by X-ray powder diffraction using a Guinier-Hagg focusing camera and Cu- Ka, radiation. Microstructural characterization was carried out by scanning electron microscopy (SEM) using a JSM35 The present paper is a continuation of previous ~ork,~,~ instrument fitted with EDX facilities. Some of the samples t Present address: National Defence Research Establishment, S-172 90 Sundbyberg, Sweden t 1 Ibf i11-~=6.89 x103 Pa were also studied with a JEOL 2000 FX TEM microscope, equipped with a Link loo00 EDS system. Powder diffractometer data for the structure refinements were collected with a Rigaku diffractometer equipped with a J.MATER. CHEM., 1991, VOL. 1 required for the cubic cell. This is most marked in the La- sialon system, and on high-quality Guinier-Hagg photo-graphs the resolution of broadened lines into doublets at higher 8 values is apparent. Conclusive evidence that the rotating anode, Cu-Ka radiation, and operating at 45 kV and 190 mA. A graphite monochromator was placed between the U-phase is not cubic is provided by absent reflections. The 1' +k2+h2presence of the reflection corresponding to =7 in sample and the X-ray detector, thereby reducing the back- ground in the diffraction pattern. The diffraction pattern was scanned in steps of 0.02" (28), and fixed-time counting (4s) was employed.For the HREM studies Nd U-phase samples were crushed under butanol and dried on to a porous carbon film supported by a copper grid. The TEM studies were made with a JEOL 200CX microscope with a point-to-point resolution of ca. 2.5 8, and equipped with a top-entry goniometer stage allowing a tilt of *loo about two axes. Image calculations were made according to the multislice method using a locally modified version of the SHRLI program package.6 Results Preparation Table 1 gives the results of firing and heat treating compo- sitions on the /I'-'NdA1O3' plane close to the region where the U-phase was previously identified. Clearly, no region of pure U-phase exists, but in many cases substantially two- phase compositions (considering crystalline phases only) were observed. The combination U-phase- 15R occurred frequently, and a sample used later for crystal-structure determination was taken from this region.X-Ray Diffraction Unit-cell Results The X-ray powder diffraction patterns of the U-phase have an apparently cubic appearance, especially for the Nd U-phase (see below). Extracting detailed information from the patterns obtained in earlier preparations3 was difficult, as the samples were generally multiphase with the U-phase as a minor constituent. Moreover, the U-phase had formed by devitrifi- cation at low temperatures (1000-1 300 "C) and yielded diffuse X-ray lines with decreasing intensities as 8 increased. In rare- earth sialon systems, the U-phase seems to be more stable than in corresponding Y compositions and crystallizes to form larger grains.In these systems it is easier to identify X-ray reflections that deviate from the perfect sin28 values the pseudocubic pattern was interpreted in the previous work3 in terms of the structure being body-centred. However, in this case there should be absent reflections corresponding to h2+k2+12=28 and 60. Careful examination of the X-ray photographs showed that reflections were present in these positions with intensities greater than could be explained by trace amounts of other phases. Careful indexing of the pattern showed that the correct unit cell was trigonal, with unit-cell dimensions of a=7.974(1) 8, and c=4.873(1) A, for the Nd U-phase, and a=8.072(4) 8, and c=4.895(3) 8, for the La U-phase. The pseudocubic appearance of the Nd U-phase is accounted for by the almost precise ,/(3/8) ratio between c and a.Table 2 gives indexed X-ray diffraction patterns for the Nd and La U-phases. Similar conclusions regarding the hexagonal nature of the unit cell of U-phase were deduced by Fernie et aL7 from a combination of X-ray diffraction and electron diffraction results. Fig. 1 shows a plot of unit-cell dimensions uersus ionic radius of the Ln cation for a range of U-phases. As expected, an approximately linear variation is observed. Fig. 1 also shows the variation of c/a with Ln3+ ionic radius and there is a noticeably uniform increase with increasing atomic num- ber. A c/a value of exactly ,/(3/8) corresponds most closely to the Nd and Sm U-phases and explains why in much of the early neodymium work the true unit cell was not deduced.At either end of the rare-earth series, separation of lines into doublets occurs more readily and identification is easier. However, preparation of the U-phase is easier at the lower Z end of the rare-earth series, presumably because the larger cation size favours structural stability. Composition Cation contents and O:N ratio for the Nd U-phase have been measured by Fernie et who deduced the most plausible composition to be Nd3Si3Al3OI2N2. In the present work a range of neodymium, lanthanum and cerium U-phases were analysed by EDX analysis and showed cation contents close to the 1:l:l atomic ratio but with the amount of aluminium Table 1 Preparative results for composition fired on or close to the P'-NdAlO, plane (s =strong; m =medium; w =weak; Me1 =Nd-melilite, Nd,Si,O,N,) composition (atom ratio) no.Nd Si A1 0 N T/"C t/h U NdAlO, p' 15R P-Al,O, others 1.5 3.0 5.5 9.0 5.0 1700 1.7 - W 1100 16 vs vvw 2.0 3.0 5.0 9.0 5.0 1600 1.0 - vs 1100 70 S vs 2.8 2.1 4.9 10.5 3.5 1710 1.0 - vw 1100 16 S S 1.3 2.5 6.4 8.9 5.1 1710 1.0 - vw 1100 16 vs - 1.3 1.3 1.5 3.8 1.3 6.0 5.1 7.6 2.5 7.6 10.2 6.0 6.5 3.8 8.0 1720 1050 1700 1050 1700 1.0 48 1.0 48 1.2 -ms -mw - vw -S S - 12H(mw) 12H(m) NdSiO,N(w) NdSiO,N(w) 2.8 4.2 2.8 8.4 5.6 1050 1700 1050 22 0.5 22 S -m --- NdSiO,N(s) Mel(w) NdSiO,N(m) Mel(w) J.MATER. CHEM., 1991, VOL. 1 24 1 Table 2 Observed and calculated 28 values for the Guinier-Hagg powder diffraction patterns of the Nd and La U-phase" Nd U-phase La U-phase hkl 2@,,bs/' A(28)l' dobs/A Iobs Icalc 280bs/o dobs/A lobs Icalc 100 12.763 -0.046 6.930 42 47 12.599 -0.052 7.020 34 36 001 18.179 -0.013 4.876 268 253 18.077 -0.035 4.903 174 213 110 22.275 -0.006 3.9879 309 256 21.976 -0.027 4.0414 162 238 200 25.792 0.009 3.45 15 157 162 25.439 -0.02 1 3.4985 101 132 111 28.923 0.010 3.0845 lo00 lo00 28.629 -0.015 3.1155 lo00 1000 20 1 31.755 0.018 2.8 156 679 665 31.417 -0.006 2.8452 593 665 210 34.381 0.050 2.6063 217 224 33.889 -0.007 2.6430 159 204 002 36.887 0.023 2.4348 93 96 36.686 -0.012 2.4470 65 92 300 0.038 79 0.075 84 21 1 39.131 0.019 2.2996 304 177 38.677 -0.016 2.3261 185 150 102 -0.038 40 38.970 0.004 2.3094 35 47 30 1 0.048 58 42.922 -0.028 2.1054 66 49 112 43.494 -0.004 2.079 1 219 158 43.209 0.009 2.092 1 118 120 220 0.086 42 44.868 -0.006 2.01 85 34 31 202 45.550 0.018 1.9899 200 150 45.193 -0.003 2.0047 116 137 310 47.440 0.008 1.9149 298 307 46.816 0.003 1.9389 259 29 8 31 1 0.03 1 350 50.586 -0.006 1.8029 273 274 212 5 1.240 -0.015 1.7814 59 1 232 50.784 -0.029 1.7964 197 197 302 54.837 0.012 1.6728 89 88 54.306 -0.01 1 1.6879 69 75 320 58.2 12 0.024 1.5836 49 37 57.455 0.05 1 1.6026 30 25 32 1 0.01 1 143 60.708 -0.006 1.5243 115 141 312 61.51 1 -0.030 1.5063 420 159 60.883 -0.027 1.5204 135 139 113 -0.098 92 61.240 0.005 1.5123 65 73 a The observed intensities originate from diffractometer data (Nd U-phase) and Guinier-Hagg data (La U-phase), respectively.The intensities are calculated with use of the DBW 3.2 program." A(28)= 2@obs-28ca~c. 8.10 that the composition of the Nd U-phase samples in the present work was ca. Nd3Si2~,A1,~,012~5N,~s. 8.05 Further experimental work was therefore carried out by melting glasses of the compositions specified by x=O, 1/2, 1 8.00 in the general formula Ln,(Si,-.Al, +.)Ol2+xN2--x where Ln =La, Nd.Melting these glasses at temperatures above 7.95 1700 "C resulted in considerable nitrogen loss and, on quench- ing, significant amounts of LnAlO, were observed. Lowering 7.90 the preparation temperature reduced considerably the amount of aluminate formed and gave much better glasses on quench- 7.85 ing. Heat treatment at 1000-1 100 "C gave mostly U-phase \. with trace amounts of LnA10, and some apatite\7.80 4.90 [Ln,(SiO,),N]. Alternatively, the same U-phase compositions 2 were prepared by solid-state reaction at 1250-1350 "C for 4.85 periods of 20h or longer. However, only the Nd U-phase could be prepared in this way, with the largest amounts found 0 4.80 for xzO.5.From these results, it is clear that the IJ-phase has a composition close to that given above. 0.615 Attempts to synthesize U-phase-type compounds contain- ing only oxygen, i.e. Ln,SiA1,014, with Ln =La,Nd,Y by 0.610 solid-state reaction at temperatures 1300- 1650 "C were not successful. The main crystalline products were the aluminate 0.605 phases LaA10, and NdAlO,, and for Y the garnet phase Y3A15012* Densities were measured on a range of Nd U-phases and La Ce Nd Srn Gd Dy Er gave values in the range 4.0-4.6 g ern-,. The calculated den- sity of an Nd U-phase with the gallogermanate structure type 1.05 1.00 0.95 0.90 0.85 (see below) is ca. 5.1 g ern-,. Most samples measured did, rl A however, contain significant amounts of the 15R phase as Fig.1 Unit-cell dimensions of Ln U-phases as a function of Ln ionic impurity, which has a density of 3.4 gcrn-,, and would radius satisfactorily account for the observed difference. always slightly higher than the silicon content. The analyses Structure Refinement yielded compositions for Ln:Si:Al of 3:3:3 and 3:2:4, which, taking account of the O:N ratio given by Fernie et al., would A literature search for compounds with a=8-10 A and c= give U-phase compositions between Ln3Si3A13012N2 and 5-6 A revealed a series of metal gal lo germ an ate^^^^ of general Ln,Si2A1401 ,N. In fact, repeated measurements indicated formula M,(Ga,Ge),014 with M=Ca, Ln and the Ga:Ge ratio 5:1 for trivalent M and 2:4 for divalent M. These are trigonal (space group P321) with a z 8 8, and c z 5 A.The Nd U-phase structure was refined with P321 symmetry and starting coordinates taken from the Ln,Ga,GeO 14struc-ture.’ The stoichiometry of the Nd U-phase was assumed to be Nd3Si,Al3Ol2N2. A1 atoms were placed in the six-fold co- ordinated octahedral origin sites, with the remaining A1 atoms and Si atoms randomly distributed over the four-fold co- ordinated tetrahedral sites, and 0 and N atoms disordered on non-metal sites in the 12:2 ratio predicted by the nominal composition. Owing to the similar scattering factors of A1 and Si, and 0 and N, the X-ray powder data used did not allow a precise determination of the atom ordering scheme. The program used in the refinements was a local version of DBW 3.2.The atomic form factors used in the least-squares refinements were corrected for anomalous dispersion. Step intensities in the range lO”~28190” were included in the refinements, comprising 67 theoretical Bragg peaks and 114 hkl reflections. The refinements were considered to have converged when the parameter shifts were less than 0.1 of the e.s.d. values. Besides peaks assignable to the Nd U-phase, the X-ray diffractometer data contained Bragg peaks from the 15R polytype phase, SiA1402N4, which was included in the refinements as a second phase with atomic coordinates from ref. 11. At this stage, it was noted that the refinements yielded one unrealistically short (A1,Si)-(0,N) distance of ca. 1.2-1.3 A. In the final refinements the z coordinates for these (A1,Si) and (0,N) atoms were therefore linked together to yield a fixed (A1,Si)-(0,N) distance of 1.54 8, (cf: Discussion, later).The final refinements were applied to 24 parameters: (i) nine positional, three temperature, two cell parameters and one scale factor for the Nd U-phase; (ii) two cell parameters, one overall temperature factor and one scale factor for the 15R impurity phase; (iii) three halfwidth (U,vW),one asym- metry and one zero-point parameter. The final atomic coordinates and temperature factors ob- tained for Nd U-phase, Nd3Si3A13012N2, are listed in Table 3. Because of serial correlation as indicated by Durbin- Watson statistics (d=0.35), the e.s.d. values given in Table 3 are multiplied by 3.The calculations yielded the following R factors; R, =0.077, R,, =0.096, R,=0.033 and RF =0.028. The fit between the observed and calculated patterns is shown in Fig. 2. A corresponding refinement of the La U-phase, La3Si3AI3Ol2N2, using Guinier-Hagg film data, yielded similar atomic coordinates and an RF value 0.082. HREM Ca. 20 crystals were investigated by electron diffraction. HREM studies verified the trigonal symmetry and no super- structure reflections were revealed. Table 3 Atomic coordinates and temperature factors for Nd U-phase“ position xja Ylb z/c B/AZ Nd 3(e) 0.41 32(6) 0 0 2.0( 1) A1 l(a) 0 0 0 1.5(5) (Al,Si)(1) 2(d) 113 213 0.534(10) 1.5(5) (Al,Si)(2) 3(f) 0.760(3) 0 1I2 1.5(5) (0” 1) 2(d) 113 213 0.218(10) 1.5(7) (O,N)(2) 6(g) 0.441(9) 0.298(4) 0.325(6) 1.5(7) (O,N)(3) 6(g) 0.222(4) 0.092(4) 0.771(6) 1.5(7) “The space group is P321.The nominal composition is Nd,Si,A1,0,2N2 with occupancy factors for (A1,Si) sites n,, =0.6, nA,=0.4 and for (0,N) sites no=516, nN= 116. The z/c coordinate for (O,N)(l) is linked to the zjc coordinate for (Al, Si)(l), (see text). J. MATER. CHEM., 1991, VOL. 1 Table 4 Selected interatomic distances/A in Nd U-phase“ 1.90( 3) 2.67(5) 2.57(8) 2.86( 6) 1.54 1.81(6) 2.79(6) 2.91(11) 1.66(6) 1.73(3) 2.60(9) 2.57( 8) 2.76(5) 2.29(3) 2.55(3) 2.77(3) 2.63( 2) “Standard deviations in parentheses. The (Al,Si)( 1)-(O,N)( 1) distance is fixed in the corresponding refinement (see text). 3 I Q 30001 1 Fig.2 Final Rietveld plot for the Nd U-phase. The upper portion shows the observed intensity data and the lower portion the difference between observed and calculated intensities. The arrows indicate the two strongest, non-overlapped, Bragg reflections for the 15R impurity phase Comparing experimental images (see Fig. 3) with calculated ones, obtained using the atomic coordinates given above, revealed that it was only possible to relate the calculated pattern to the observed one for a very narrow range of thicknesses and defocus. The three images given in Fig. 4, inserted into an expanded part of Fig. 3, are calculated for three different thicknesses and with a defocus of ca. -648 8,, close to the Scherzer focus, -633 8,.From this figure it is seen that it is possible to interpret the observed contrast in Fig. 3 HREM image of the Nd U-phase on (001) J. MATER. CHEM., 1991, VOL. I Fig. 4 Calculated images of the Nd U-phase, inserted in an enlarged part of the observed image. The images are calculated for a defocus of ca. 648 A and, from top to bottom, for thicknesses 20, 40 and 60 A, respectively terms of the X-ray crystal structure in the thinnest part of the crystal. Discussion The U-phase structure is illustrated in Fig. 5. The structure exhibits layers of (Si,Al)(O,N), tetrahedra running perpendicu- lar to the c axis. The (A1,Si) tetrahedra centred on the three- fold axis share three corner oxygens with three neighbouring (A1,Si) tetrahedra and exhibit one free apex oxygen atom.The tetrahedral layers are interconnected, along the c axis, by A1-(0,N) octahedra, which share their corner oxygens with (A1,Si) tetrahedra. The Nd atoms are co-ordinated by eight (0,N) atoms forming a distorted cubic anti-prism. The presence of octahedra in the U-phase structure con- forms with the principle observed for the sialon polytypoid phases, As the oxygen content in the sialon polytypoid phases increases and the aluminium content remains high, increasing proportions of layers of octahedra are inserted into the otherwise tetrahedral network. The octahedra in the U-phase serve to link the structure together along the c axis, but also provide the sixth side of the irregular hexagon of polyhedral edges which co-ordinate the large cation sites.It is interesting to compare this structural arrangement with other rare-earth nitride and oxynitride structures that do not contain octahedra. In the original work on Ca,Ga2Ge4014,8 analogy was drawn with the compositionally similar structure of gehlinite, Ca2Ga2Ge07, which has the melilite type of struc- ture. The melilite structure has almost identical unit cell dimensions with the trigonal gallogermanate. The distribution of tetrahedra in the melilite structure consists of (Ga,Ge),O, groups joined together by individual tetrahedra sharing all corners (see Fig. 6). In this way, five edges of tetrahedra surround each large cation site, and there is no linkage of tetrahedra joining the structure in the c direction.The melilites are therefore more correctly regarded as layer structures, but the U-phase structure can be regarded as a melilite in which individual octahedra have been inserted to form a sixth component of the five-membered rings, also linking them together along c. The M2Si3(0,N)7 general formula of the nitrogen melilites means that 10 of the 14 non-metal atoms in the unit cell are sharing two silicon atoms, whilst the remaining four are joined to only one silicon atom. In the U-phase, all non-metal atoms are two-co-ordinated by (Si,Al) with the exception of the (0,N) atom in Wyckoff position 2d, (O,N)(l), which is co- ordinated to the (Si,Al)(l) atom on the same three-fold axis and also to three rare-earth atoms.The overall (Si,Al):(O,N) ratio in the U-phase is 6:14, which is less than the 1:2 required to two-co-ordinate every non-metal atom, and therefore two non-metals per cell must have only one (Si,Al) linkage. The refinements of both the Nd as well as the La U-phase yielded, as noted above, one very short (Al,Si)( l)-(O,N)(l) bond length, typically 1.2-1.3 A. This distance is unrealistically short, even allowing for some shortening of the bond length due to a higher bond strength of the Si-0 bond compared with the Ln-0 bond. The considerable deviation from an expected value of 1.55-1.65 A for a Si-0 bond is thus believed to be associated with a general difficulty with X-ray powder data in defining light-atom positions in structures containing heavy atoms.The (Si,Al)(l)-(O,N)( 1) bond length was therefore fixed in the final refinements to 1.54 A, a value given for an average non-bridging Si-0 distance.I2 This resulted in a small increase of the R value (2.6-2.8%), but had no significant influence on the other atomic positions. Since it was not possible to determine the position of the 0(1) atom accurately, the (O,N)(1)-(O,N)(2) and (O,N)(1)-Nd bond distances are associated with similar uncertainties. Fig. 5 An illustration of the U-phase structure viewed down (a) [OOl] and (b)[210]. The Nd atom positions are shown by the filled circles Fig. 6 An illustration of the melilite structure on (001) 244 J. MATER. CHEM., 1991, VOL. 1 The refinement of the La U-phase, using Guinier-Hagg powder diffraction data, yielded similar atomic coordinates to the Nd U-phase, and RF=0.082, thus confirming that the compounds are isostructural.The e.s.d.3 for the La phase atomic coordinates were found to be ca. 1.5 times larger than 3 4 meeting ‘EuroCeramics’, ed. G. de With, R. A. Terpstra and R. Metselaar, Elsevier Applied Science, London, 1989, pp. 1530-1535. C. J. Spacie, K. Liddell and D. P. Thompson, J. Muter. Sci. Lett., 1988, 7, 95. J. Grins, P-0. Kall, K. Liddell, P. Korgul and D. P. Thompson, those obtained for the Nd phase, owing to the somewhat inferior Guinier-Hagg data compared with the diffractometer data. Attempts to synthesize corresponding pure oxygen U-phase compounds, i.e. Ln3SiA15014, were unsuccessful (cf.Results). These findings conform with those of Mill et who found that compounds with the Ln3Ga5GeO14 structure do not form in the Ln203-A1203-Si02 systems. It may therefore be concluded that the U-phase is stabilized by its nitrogen content and, as a consequence, exhibits a narrow homogeneity range. 6 7 8 5 Con$ Proc. New Materials and their Applications, 10-12 April 1990, in the press. D. P. Thompson, in Tailoring Multiphase and Composite Cer- amics, ed. R. E. Tressler, G. L. Messing, C. G. Pantanc and R. E. Newnham, Plenum, New York, 1986, pp. 79-91. M. A. O’Keffe, in Electron Optical Systems for Microscopy, Microanalysis and Microlithography, ed. J. J. Hren, F. A. Lenz, E. Munro and P. B. Server, Proc. 3rd Pfeffercorn Conf., SEM Inc., AMF OHare, IL, 1984. J. A. Fernie, G. Leng-Ward and M. H. Lewis, Muter. Lett. 1989, 9, 24. E. L. Belokoneva and N. V. Belov, Sov. Phys. Dokl., 1981, 26, 931. Professors T. Ekstrom and M. Nygren are thanked for their 9 B. V. Mill, A. V. Butashin, G. G. Khodzhabagyan, E. L. Belo-koneva and N. V. Belov, Sov. Phys. Dokl., 1982, 27,434. great interest and valuable comments on this work. The work has been supported financially by the Swedish Board for Technical Development and by AB Sandvik Hard Materials (Stockholm). 10 11 12 D. B. Wiles and R. A. Young, J. Appl. Crystallogr., 1981, 14, 1491. Y. Bando, M. Mitomo, Y. Kitami and F. Izumi, J. Microsc., 1986, 142, 235. J. Lang, in Progress in Nitrogen Ceramics, ed. F. L. Riley, Martinus Nijhoff, Boston, 1983, p. 23. References Paper 0/04301J; Received 24th September, 1990 1 D. P. Thompson, Muter. Sci. Forum, 1989, 47, 21. 2 D. S. Perera, P. Korgul and D. P. Thompson, Proc. 1st. ECers

ISSN:0959-9428    

DOI:10.1039/JM9910100239

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 245-249 Non-linear Optical Properties of Group 10 Metal Alkynyls and their Polymers Werner J. Blau," Hugh J. Byrne," David J. Cardinb and Andrew P. Daveyb a Department of Pure and Applied Physics and Department of Chemistry, Trinity College, University of Dublin, Dublin 2,Ireland Conjugated diacetylene and phenylacetylene systems incorporating Group 10 transition-metal complexes have been obtained via lithium acetylide intermediates. In the UV-VIS spectra, the positioning of the longest- wavelength absorption shows a strong dependence on the incorporated metal. Polymers of both the trans- nickel and the trans-platinum bisphosphine diacetylide [M(PR3),(C-C-CrCH),] (M =Ni, or Pt; R =alkyl group) were synthesised. UV absorption spectra show a significant shift of the absorption maxima from those of the monomers, towards longer wavelengths, consistent with an increase in the degree of electron delocalisation per monomer unit.The third-order non-linear optical properties of these materials in solution were examined using the method of self-diffraction from laser-induced gratings with 70 ps pulses of wavelength 1.064 pm. In the case of the monomer solutions, intensity dependences of the diffraction efficiencies show the influence of the proximity of a three-photon resonance at low concentrations. At higher concentration the influence of this resonance is not apparent. Among the materials synthesised, a clear trend, down the group, is seen, the hyperpolarisabilities decreasing with increasing atomic number of the metal.Measured hyperpolarisabilities are compared with those of enyne oligomer solutions and are found to be substantially higher. Metal-containing polymeric systems show an intensity dependence that is characteristic of a pure third-order process. Molecular hyperpolarisabilities show trends similar to those of the monomeric species. Both polymers possess a non-linearity higher than that of equivalent polydiacetylene solutions. Keywords: Non-linear optical material; Transition-metal complex; Conjugated polymer In the search for materials that might be useful for all-optical switching, organic conjugated polymers have received much attention.' Their delocalised n-electron backbones exhibit sizeable non-linear optical susceptibilities with ultrafast response and recovery times.2 In addition, their structural and electronic properties can be varied chemically and the fact that they are easily processed renders them potentially useful in integrated optical technologies.Over the past decade, much research has been carried out in this field and the result of intensive synthetic efforts is a range of n-conjugated organic polymeric systems, within which polydiacetylenes have shown particular promise with off-resonant non-linearities as high as 10-lo esu.t3 It is important, however, to consider the minimum non-linearity required for realistic application to optical switching. This has been estimated to be >lop9~su.~This value may have been a~hieved,~ but not exceeded using an electronic non-lin- earity.Furthermore, studies of the dependence of the molecu- lar hyperpolarisability (y) on chain length in enyne oligomers indicate that it does not extrapolate to the polymer which exhibits a non-linearity corresponding to a chain length of only 7-10 repeat units.6 This saturation of the material non-linearity may be linked with strong imaginary compo- nents of the non-linearity, associated with a strong electron- lattice coupling, characteristic of these quasi one-dimensional systems.' These results suggest that greater hyperpolarisabilities may not be achieved simply by increasing the electron density per monomer unit. For example, chemical doping of polymers has been observed to lower the non-linear optical response.' Therefore, consideration must be given to ways of modifying the electronic and lattice configurations.Modification of the electronic configuration of a polymer backbone must be accommodated by the lattice to avoid charge localisation and a break-up of the backbone c~nfiguration.~ Incorporation of transition metals into conjugated organic systems has been reported for other purposes. The presence of polarisable low oxidation state metals having occupied d orbitals in the conjugated backbone should increase the hyperpolarisability. In this study, the effect on the non-linear optical properties of incorporating such metals into diacetyl- enic systems is examined. The metals are incorporated in a square-planar geometry leading to a linear backbone, which should result in increased conjugation.The variation of the hyperpolarisabilities with the metal have been studied and the results compared with those for enyne oligomers and pol ydiacetylenes. Experimental Chemical Synthesis and Characterisation The complexes trans-[M(PhC~C),(PEt~)~],(M =Ni, Pd, or Pt), were first synthesised by Chatt and Shaw9,l1 and Calvin and Coates," and truns-[(HC=C-C=C),M(PEt,),l by Hagihara et a!.' In these reports sodium acetylide intermedi- ates were prepared in liquid ammonia. We have adopted a more convenient procedure, affording these complexes in reasonable to good yields, employing lithium acetylides, read- ily obtainable in ether from the appropriate acetylene and n-butyllithium. All reactions were performed under an N2 atmosphere, solvents were purified by standard methods and distilled and degassed before use.HC-C-C-CH was prepared as published.' truns-[Ni(C~CPh),(PEt,),] was made by a modification of the method of Chatt and Shaw. To a stirred solution of [Ni(PEt3)2C12], (2.69 g, 7.35 mmol) in ether (30 cm3) was added dropwise a solution of PhC-CLi (14.7 mmol) in ether (30 cm3) at 30 "C over a period of 30 min. A change from deep red to orange/brown was observed. Following the additions, the solution was allowed to warm to room tempera- ture. Removal of solvent under vacuum yielded a brown oil which was purified by silica column chromatography (eluent hexane) yielding an orange/yellow solution which, when reduced to low volume, yielded the product as orange needles (2.63g, 71%), m.p.148-151 "C. IR (KBr disc): 2961sh, 2928sh, 2122sh, 1587sh, 1021sh, 595sh and 573mcm-'. 'H NMR 6 (solvent CDC1,; standard SiMe,) 1.38(q), 1.59(t) and 7.22(m). UV imax(CHC1,) 370 nm (E 4.0 x lo5 dm3 mol-' cm-'). The following were similarly prepared. trans-[Pd(C=CP h)2 Pd( PEt 3)2] using cis-[Pd( PEt 3)2C1 2], (2.03 g, 6.22 mmol) and PhC=:CLi, (1.34 g, 12.44 mmol) in ether: yield 1.62 g, 57%; m.p. 163-164 "C. IR(KBr disc): 2965sh, 2931sh, 21 18sh, 1988sh, 1597sh, 1486sh, 1034sh, 766m, 757m, 581m and 543m cm-'. 'H NMR 6 (solvent CDC1,; standard SiMe,), 1.32(q), 1.51(t) and 7.18(m). UV A,,, (CHC1,) 370 nm (E 3.71 x103 dm3 mol-' cm-').trans-[Pt(C=CPh),(PEt,),l using ~is-[Pt(PEt,)~Cl,] (1.27 g, 3.3 mmol) and PhC-CLi (0.7 g, 6.6 mmol) in ether: yield 1.32 g, 74%; m.p. 186-188 "C. IR(KBr disc): 2971sh, 2935sh, 2177sh, 1979sh, 1601sh, 1483sh, 1027sh, 752m, 585m and 552m cm-'. 'H NMR 6 (solvent CDC1,; standard SiMe,), 1.34(q), 1.55(t) and 7.21(m). UV A,,, (CHCl,) 332 nm (E 1.36 x104 dm3mol-' cm-'). trans-[Ni(C C-C =CH),(PEt,),] using cis-[Ni(PEf3),Cl2] (2.29 g, 5.81 mmol) and HC-C-C'CLZ (1.25g, 11.62 mmol) in ether: yield 1.56 g, 68%; m.p. 140-142 "C. IR (KBr disc): 3210sh, 2970sh, 2938sh, 2105sh, 1990sh, 1591sh, 1478sh, 1035sh, 770m, 588m and 570m cm-'. 'H NMR 6 (solvent CDCl,; standard SiMe,), 1.39(q), 1.61(t) and 6.12(s). UV A,,, (CHCl,) 336 nm (E 6.54 xlo3 dm3 mol-' cm-I).trans-[Pd(CGC-C =CH)2( PEt using cis-[Pd(PEt3)2C12], (2.03 g, 6.22 mmol) and HC-C-C'CLi, (1.34 g, 12.44 mmol) in ether: yield 1.97g, 72%; m.p. 98-99 "C. IR (KBr disc): 3225sh, 2975sh, 2934sh, 2122sh, 1991sh, 1597sh, 1481sh, 1040sh, 773m, 599m and 593mcm-'. 'H NMR 6 (solvent CDCl,; standard SiMe,) 1.30(q), I .52(t) and 6.2 I(s). UV A,,, (CHC1,) 290 nm (E 1.32 x lo2 dm3 mol-' cm-'). tr~n~-[Pt(C=C-C=CH)2(PEt,)J using cis-[Pt(PEt,),Cl,] (1.16 g, 2.3 mmol) and HC-C-C-CLi (0.49 g, 4.60 mmol) in ether: yield 0.95 g, 78%; m.p. 121-123 "C. IR (KBr disc): 3221sh, 2971sh, 2931sh, 2138sh, 1994sh, 1602sh, 1476sh, 1047sh, 761m, 576m and 401mcm-'. 'H NMR 6 (solvent CDCl,; standard SiMe,), 1.35(q), 1.55(t) and 6.29(s).UV A,,, (CHCl,) 318 nm (E 1.03 x103 dm3 mol-' cm-'). The polymers [-M( PBU,)~ -C=C-Cr C-)n where M=Ni or Pt were prepared by copper(1) coupling of the alkynyl with the dihalometal phosphine compound in diethyl- amine as described by Hagihara et a!.'4315 Non-linear Optical Studies The experimental method employed was that of Forced Light Scattering from Laser Induced Gratings, a technique that corresponds to a degenerate four-wave mixing process in the forward direction.' The light source is an amplified, passively mode-locked Nd3 +:YAG laser emitting linearly polarised pul- ses of 50 +25 ps duration and of wavelength i.= 1.064 pm at a frequency of 3 Hz. Peak powers of up to 50 MW were readily available. The experimental method is described in detail elsewhere' and the set-up is depicted schematically in Fig.1. It is based on the interference at the sample of two spatially and temporally overlapped beams, producing a spa- tial modulation of the intensity-dependent refractive index of the material. This modulation acts as a diffraction grating from which the pulses may self-diffract. Under thin grating J. MATER. CHEM., 1991, VOL. 1 beam splitter delay line photodiode Fig. 1 Experimental set-up for self-diffraction technique conditions,' satisfied experimentally by keeping the angle between the two beams small (<I"), an expression relating the diffraction efficiency into the first order (q),to the third- order material non-linearity may be derived: where c is the speed of light, E~ is the permittivity of free space, n is the refractive index of the sample, d is the sample thickness and I, is the input pulse intensity. In the experiments reported here, d= 1 mm and n is taken to be the refractive index of the solvent, because of the low fractional volume of solute. Eqn.(1) holds for materials that are transparent at the operating wavelength. It can be seen from eqn. (1) that verification of the presence of a true third-order non-linear process may be performed by monitoring the intensity dependence of the diffraction efficiency. For a true third-order process, 11q=--z; (2)10 where I, is the intensity diffracted into the first order. Such a verification is important as fifth- and seventh-order pro- cesses, originating in two- and three-photon resonant enhance- ment of the material non-linearities, have been observed in organic conjugated materials.' 7*1* 1x(3)1may have both real and imaginary components orig- inating from the solute as well as a contribution from the solvent, x(3)s01v,which is purely real and positive in the organic solvents empl~yed.'~ For the concentration range used in this work the solute fractional volume is negligible. Hence where Re x(3)s01u are the real and imaginary and Im x(3)s01u components of the solute non-linearity.By monitoring the concentration dependence of 1~(~)1, the contribution x(3)s01v may be extracted and the magnitudes of ReX(3)soluand Im x(3)s0,umay be determined. Furthermore, the sign of Re x(3)s01umay be determined from the concentration depen- dence of the real part of Ix(3)1.17 Results and Discussion A series of solutions, of different concentrations, was made up in chloroform for each material.Maximum concentrations used were 2.7, 10.3 and 6.2 g dm-, for the nickel, palladium and platinum diacetylides and 1.4, 2.1 and 6.2 g dm -for the nickel, palladium and platinum phenyl acetylides, respectively. Measurements of the diffraction efficiency as a function of pump intensity and polymer concentration were performed with calibrated photodiodes and a digital storage oscilloscope as described above. At high concentrations the intensity dependence of the diffraction efficiency was found to be characteristic of a third- J.MATER. CHEM., 1991, VOL. 1 order non-linear optical process, as is that of the solvent. At low concentrations, however, all monomers exhibit a high- order intensity dependence, characteristic of a multiphoton resonant enhancement of the diffraction process. As an example, the concentration dependence of the order of the intensity dependence of the diffraction efficiency in trans-[Pt(C=C-C=CH),(PEt,),] solutions is shown in Fig. 2. A maximum intensity dependence of 16,indicative of a seventh- order non-linear optical process, was observed. A similar intensity dependence is seen in polythiophene solutions,’ and may be identified as the result of a three-photon resonant enhancement of the diffraction process. Notably, a strong concentration dependence of this resonant enhancement has also been observed in polythiophene solutions, although it is not as pronounced as is seen in the present case. Such concentration dependence may be associated with concen-tration-dependent interchain coupling which causes a vari- ation in backbone electron correlations, a strong determining factor in the positioning of multiphoton absorption levels.” Whereas this behaviour has been frequently observed for polymers, our results indicate that aggregation also plays a significant role for these monomeric systems, producing effects (i.e.concentration and intensity dependence) analogous to those associated with interchain coupling in polymers. The observation of this multiphoton absorption at low concen- trations is indicative of a reasonable degree of backbone electron correlation, suggesting a significant interaction of the metallic d electrons with the organic .n-conjugated system.This concentration-dependent resonant enhancement of the molecular non-linearity complicates the interpretation. An example is shown in Fig. 3 for the concentration dependence of the diffraction efficiency of the trans-[Ni(C =C-C=CH),(PEt J2]. The diffraction efficiency 0 2 4 6 concentration/g drn-3 Fig. 2 Concentration dependence of the order (I”) of the intensity dependence of the diffraction efficiency for the trans-[Pt(C=C-CECH)~( PEt,),] solutions : *,4.04‘51 3.51 ? 3.0 concentration/g drn -3 Fig. 3 Concentration dependence of the diffraction efficiency of the trans-[Ni(CrC-C~CH),(PEt,),] solutions.Solid line is a fit to eqn. (3) in the region without three-photon resonance peaks sharply initially, under the influence of the three-photon resonance, but then falls markedly below the level of the solvent, and subsequently shows the expected parabolic depen- dence, see eqn. (3). An estimate of the molecular hyperpolaris- ability may be made by fitting this concentration dependence to the regions that show a purely third-order non-linearity. For a realistic comparison of the non-linearity with that of other materials, the third-order molecular hyperpolarisability, given by (4) should be used, where C is the molar concentration, NA is Avogadro’s number and LL is the Lorentz local field factor, which for practical purposes is taken to be that of a spherical or randomly coiled molecule, and is given by’’ n2+2 LL=---3 IyI may similarly be resolved into real and imaginary compo- nents, yR and yI.The estimated values for the respective components of the molecular hyperpolarisabilities for each monomer are given in Table 1. A trend of increasing molecular hyperpolarisability with decreasing atomic number may be clearly seen for both sets of monomers, the phenyl acetylides giving larger values than the diacetylides. Indeed this is consistent with the increase in maximum absorption wave- length observed in the linear optical spectra. Of particular interest is a comparison between the molecular hyperpolarisabilities of the metal-containing systems with organic materials.Fig. 4 shows a plot of lyl versus chain length for a series of enyne oligomers,6 which serve as model building blocks for polydiacetylenes. In this figure, the values measured for the metal diacetylide and phenyl acetylide monomers have also been plotted. Chain lengths for the metal-containing monomer systems are taken from X-ray crystallographic data [for the phenylethynyl monomers from ref. 22(a) and for the butadiynyls from ref. 22(h)]. It can be seen that all of the metal-containing systems possess a non-linearity which is substantially larger than for an enyne oligomer of equivalent length. Indeed, the hyperpolarisability of the nickel diacetylide is larger than that of the enyne hexamer.Apart from the influence at low concentrations of the three-photon absorp- tion, qualitatively these materials behave similarly to the enyne oligomers. The sign of the real component of the non-linearity is negative, as in the case of enyne oligomers longer than the dimer.6 In addition, a strong contribution from imaginary components of the non-linearity is observed establishing the expected strong electron-vibration coupling. As these results were promising we also examined the nickel and platinum polymers. Solutions of different concentrations were made, of maximum concentration 1.0 and 1.4 g dm-,, respectively. The intensity dependence of the diffraction efficiency was monitored for each solution and, as is shown in Fig.5, was found to be characteristic of a true third-order process for both polymers. The concentration dependence of the diffraction efficiency at fixed intensity is shown in Fig. 6. An initial decrease of the diffraction efficiency from that of the solvent is seen (indicative of a negative real component of the polymer non-linearity), followed by an increase. How- ever, as the concentration is increased, the non-linearity falls below the parabolic dependence predicted by eqn. (3). A similar behaviour has been observed in a number of polymer solutions, including polydiacetylene-chloroform solution^,'^ and may be attributed to concentration-dependent interchain interactions. Such interactions appear to decrease the mono- meric hyperpolarisability.Efforts towards a qualitative under- J. MATER. CHEM., 1991, VOL. 1 Table 1 Measured hyperpolarisabilities of metal-containing systems and comparison with poly (4BCMU) NiDA 344 336 -7.87 10-44 1.72 x 10-43 1.89 x 10-43 PdDA 392 290 -3.85 10-44 9.19 x 10-45 3.96 x 10-44 PtDA 480 318 -1.93 x 10-44 7.71 x 10-45 2.08 x 10-44 pNiDA pPtDA NiPA 51 1 647 497 412 364 370 -2.63 x 10-42 -1.48 x 10-42 -2.75 10-43 2.41 x 10-42 1.74 x 10-42 1.46 x 10-43 3.57 x 10-42 2.28 x 10-42 3.11 x 10-43 PdPA 545 3 70 -2.10 x 10-43 3.39 10-44 2.13 x 10-43 PtPA 633 332 -1.12 x 10-43 2.15 10-44 1.14 x 10-43 p4BCMU 480 480 - - 2.00 x 10-42 DA, diacetylide; PA, phenyl acetylide; p, poly; 4BCMU, 4-(butoxycarbonylmethylurethane)diacetylene.--m -43 --4 4 -45 2 4 6 810 Llnm Fig. 4 Comparison of hyperpolarisabilities of metal-containing mono- meric systems to those of enyne oligomers (+, from ref. 6). 0, IyJof metal diacetylides; x ,IyI of metal phenyl acetylides. Solid line depicts an L4 dependence -0 log [intensity (rel. units)] Fig. 5 Intensity dependence of the diffraction efficiency of chloroform (0),nickel polymer (A) and platinum polymer (+) A150A AA0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 concentration/ g dm -Fig. 6 Concentration dependence of Ix3I2. 0,Nickel-substituted poly- mer; A,platinum-substituted polymer standing of such effects are at present underway. An estimate of the 'infinite dilution' value of the hyperpolarisability can be made by fitting eqn.(3) to the low-concentration region of the curves. The values obtained from such a fit are tabulated in Table 1 for both the nickel and the platinum polymers as well as the values obtained from the same procedure applied to polydiacetylene solutions.23 As in the case of the monomers, the non-linear polarisability is greater for the nickel polymer than for the platinum. For both polymers, the magnitude of the molecular hyperpolaris- ability is larger than that of the corresponding polydiacetylene solutions. It should be noted, however, that the difference is not as pronounced, as is seen for the monomeric species. If the length dependence exhibited by the enyne oligomers (L4), also observed for thiophene oligomers' applies for these metal-containing systems, then the polymers may be deemed to have an effective length of only 6-7 repeat units, whereas isopiestic molecular-weight determinations show that the number of repeat units is substantially in excess of this (e.g.for the platinum polymer, ca. 150units). This compares to 7-10 effective repeat units for polydiacetylenes. Qualitatively, the non-linearity of the metal-containing polymers behaves quite differently from that of polydiac- etylene. The real component of the non-linearity is negative, in agreement with an extrapolation from the monomeric systems. In polydiacetylene solutions, the real component is positive. The non-linearity of polydiacetylene solutions is dominated by the influence of a two-photon res~nance~',~~ and the sign of the non-linearity is governed by the positioning of the laser wavelength with respect to this resonance.24 With the addition of this resonant enhancement, however, the susceptibility of the polydiacetylene solutions is still substan- tially less than those of the metal-containing polymers studied here.Conclusions Consideration of the non-linear optical susceptibilities of Group 10 metal alkynyls, measured in the transparency region, gives a clear indication of a strong involvement of the d orbital electrons in the n-conjugated electron system of the backbone. Non-linear susceptibilities of monomeric systems are substantially larger than those measured for short-chain enyne oligomers.Such an increase of the non-linearity with respect to organic analogues is seen also in the metal-contain- ing polymers. This study of Group 10transition metal contain- ing compounds reveals clear dependences of the non-linearity on the metal which will lead to a further understanding of the role of the metal in the polarisability of the polymer electron backbone. J. MATER. CHEM., 1991, VOL. I 249 References 1 2 3 4 5 6 7 W. M. Dennis, W. Blau and D. J. Bradley, Appl. Phys. Lett., 1985, 47, 200. G. M. Carter, J. V. Hryniewicz, M. K. Thakur, Y. J. Chen and S. E. Meyler, Appl. Phys. Lett., 1986, 49, 998. C. Sauteret, J-P. Hermann, R. Frey, F. Pradere, J. Ducuing, R. H. Baughman and R. R. Chance, Phys.Rev. Lett., 1976,36,956. Non-linear Optical Eflects in Organic Polymers, ed. J. Messier, F. Kajzar, P. N. Prasad and D. Ulrich, Kluwer, Nato AS1 series, vol. 162, 1989. G. M. Carter, M. K. Thakur, Y. J. Chen and J. V. Hryniewicz, Appl. Phys. Lett., 1985, 47, 457. H. J. Byrne, W. Blau, R.Giesa and R. C. Schulz, Chem. Phys. Lett., 1990, 167, 484. W. P. Su, J. R. Schrieffer and A. J. Heeger, Phys. Rev. B, 1980, 22, 2099. 14 15 16 17 18 19 20 21 N. Hagihara, S. Takahashi, K. Ohga and K. Sonagashira, J. Organomet. Chem., 1980, 188, 237. N. Hagihara, S. Takahashi and K. Sonogashira, Macromolecules, 1977, 10, 879. H. J. Eichler, P. Gunter and D. W. Pohl, Laser Znduced Gratings, Springer Series in Optical Sciences 50, Springer Verlag, New York, 1986. H. J. Byrne, W.Blau and K. Y. Jen, Synth. Met., 1989, 32, 229. J. M. Nunzi and D. Grec, J. Appl. Phys., 1987, 62, 2198. P. D. Maker, R. W. Terhune and C. M. Savage, Phys. Rev. Lett., 1964, 12, 507. C. Grossman, J. R. Heflin, K. Y. Wong, 0.Zamani-Khamiri and A. F. Garito, in Nonlinear Optical Eflects in Organic Polymers, ed. J. Messier, F. Kajzar, P. Prasad and D. Ulrich, Kluwer, Dord- recht, Nato AS1 series, vol. 162, 1989. Y. R. Shen, The Principles of Nonlinear Optics, Wiley-Intersci-ence, New York, 1984. 8 9 10 11 12 13 P. N. Prasad, in Nonlinear Optical Eflects in Organic Polymers, ed. J. Messier, F. Kajzar, P. Prasad and D. Ulrich, Kluwer, Dord- recht, Nato AS1 series, vol. 162, 1989, pp. 351-363. J. Chatt and B. L. Shaw, J. Chem. SOC.,1960, 1718. G. Calvin and G. E. Coates, J. Chem. SOC., 1960, 2008. J. Chatt and B. L. Shaw, J. Chem. SOC., 1959,4020. N. Hagihara, H. Masai and K. Sonagoshira, J. Organomet. Chem., 1971, 26, 271. S. B. Armitage, E. R. H. Jones and M. C. Whiting, J. Chem. SOC., 1951, 44. 22 23 24 (a) C. J. Cardin, D. J. Cardin, M. F. Lappert and K. W. Muir, J. Chem. SOC., Dalton Trans., 1978, 46; (b) H. Yamasaki and K. Aoki, quoted in N. Hagihara, K. Sonogashira and S. Takahashi, Adv. Polym. Sci., 1981, 41, 151. H. J. Byrne and W. Blau, Synth. Met., 1990, 37, 231. F. Kajzar and J. Messier, in Polydiacetylenes, ed. D. Bloor and R. R. Chance, Nato AS1 Series, vol. 12, Martinus-Nijhoff, Dord- recht, 1985, pp. 325-333. Paper 0/04326E; Received 25th September, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100245

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. MATER. CHEM., 1991, 1(2), 251-254 25 1 Synthesis and Mesomorphism of Stilbazole Complexes of Rhodium(1) and Iridium([) Duncan W. Bruce,*" David A. Dunmur," Miguel A. Esteruelas,b Susan E. Hunt," Ronan Le Lagadec," Peter M. Maitlis," Julian R. Marsden/ Eduardo Solab and John M. Stacey" a Department of Chemistry, The University, Sheffield S3 7HF; UK lnstituto de Ciencias de Materiales de Aragon, Universidad de Zaragoza-CSIC, 50009 Zaragoza, Spain The complexes cis-[MCI(CO),(n-OPhVPy)] (M= Rh, Ir; n-OPhVPy = trans-4-alkyloxy-4'-stiIbazole) are mesomorphic showing nematic and smectic A phases at temperatures below 140 "C. Keywords: Liquid crystal; Metallo-mesogen; Rhodium; Iridium Liquid-crystalline materials containing transition metals have received increasing attention in recent years, owing to the possibilities offered by the presence of a metal in an ordered fluid system.' -3 The liquid-crystal chemistry of the Ni triad of elements has been quite well developed,2 but reports of mesomorphic materials containing Rh or Ir are scarce.We reported3 the Rh complexes cis-[RhCI(CO),(CB)] [CB = alkyl(oxy)cyanobiphenyl], which were thermally unstable, probably owing to the weak Rh-nitrile bond. Giroud-God- quin et al. reported4 the dirhodium tetracarboxylates which were shown to have columnar phases, and later the complexes [MCl(CO),(n-OPhIPy)] [M =Ir,' Rh;6 n-OPhIPy =4'-alkyl-oxy-N-(4-pyridylbenzylideneaniline)] were found to show nematic and smectic A phases, despite the fact that the substituted pyridine ligand was non-mesomorphic.We now report analogous complexes of Rh and Ir where the mesogenic ligand, L, is 4-alkyloxy-4'-stilbazole7 which we have previously used to generate other metal-containing liquid crystals.* Experimental All reactions were carried out under an inert atmosphere and solvents were dried before use. Elemental microanalyses were by the University of Sheffield Microanalytical Service and infrared spectra were measured on a Perkin Elmer 1600 FT instrument. 'H NMR spectra were measured on a Bruker AM250 instrument in CDCl,, while 3'PJ1HJ NMR spectra were recorded on a Bruker WP80-SY instrument at 32.4 MHz in CDC13: proton chemical shifts are relative to tetramethyl- silane, while phosphorous chemical shifts are quoted relative to 85% H3P04.Solution electronic spectra were measured on a Philips 8700 instrument in chloroform, and solid-state measurements were made on a Perkin Elmer model 330 spectrometer. Mesophase behaviour was determined using a CZ Scientific Labpol polarising microscope equipped with a Linkam PR600 controller and TH600 hot stage and using a Perkin Elmer DSC7 differential scanning calorimeter. DSC traces were recorded at 10 K min-' in hermetically sealed A1 pans with typical sample weights in the range 3-5 mg. [RhCl(COD)]2,9 [IrC1(COD)I2 lo and 4-alkyloxystilba-zoles7 were synthesised by literature procedures. All rhodium complexes were prepared in the same way, based on the method described in ref. 6. Yields and microanalytical results are collected in Table 1.Preparation of [RhCl(CO),(n-OPhVPy)]. Dichloromethane (10 cm3) was added to a mixture of 4-octyloxystilbazole Table 1 Microanalytical data for rhodium complexes microanalysis calculated (found) n yield (YO) C H N 3 4 5 6 7 8 9 10 11 12 56 91 67 45 78 77 74 74 50 84 49.9 (49.9) 51.0 (51.6) 52.0 (51.7) 53.0 (52.8) 53.9 (53.8) 54.8 (54.8) 55.6 (55.9) 56.4 (56.2) 57.3 (57.1) 57.9 (57.6) 4.0 (3.9) 4.3 (4.1) 4.5 (4.4) 4.8 (4.7) 5.1 (5.2) 5.4 (5.3) 5.6 (5.7) 5.8 (5.8) 6.1 (6.2) 6.2 (6.1) 3.2 (3.1) 3.1 (3.0) 3.0 (2.9) 2.9 (3.1) 2.9 (2.9) 2.8 (2.8) 2.7 (2.7) 2.6 (2.5) 2.6 (2.4) 2.5 (2.3) (124 mg, 0.4 mmol) and [RhCI(COD)], (100 mg, 0.2 mmol) and after stirring (5 min), carbon monoxide was bubbled through the solution (10 min).The resulting solution was concentrated in uucuo and hexane (15 cm3) was added to give a yellow precipitate which was recovered by filtration and air-dried. All iridium complexes were prepared in the same way5 and a sample preparation is given. Yields and microanalytical results are collected in Table 2. Preparation of [IrCI(CO),(n-OPhVPy)]. Dichloromethane (10 cm3) was added to a mixture of [IrCI(COD)], (100mg, 0.148 mmol) and 4-pentyloxystilbazole (79.7 mg, 0.298 mmol). After stirring (5 min), carbon monoxide was bubbled through the solution (1 0 min). The resulting solution was concentrated in uucuo and hexane (15 cm3) was added to give a burgundy precipitate which was recovered by filtration and air-dried.Reaction of [Rh CI (C0),(8-0P h V Py)] with P(0Me)3. [RhCl(C0),(8-OPhVPy)] (150 mg, 0.30 mmol) and P(OMe), Table 2 Microanalytical data for iridium complexes microanalysis calculated (found) n yield (YO) C H N 4 72 42.5 (42.3) 3.5 (3.4) 2.6 (2.3) 5 91 43.6 (43.5) 3.9 (3.9) 2.6 (2.4) 6 49 44.6 (44.2) 4.1 (4.2) 2.5 (2.5) 7 87 45.8 (45.5) 4.0 (4.5) 2.4 (2.4) 8 83 46.6 (46.1) 4.6 (4.6) 2.4 (2.3) 9 86 47.5 (47.6) 4.8 (4.6) 2.3 (2.3) 10 81 48.3 (48.0) 5.0 (4.9) 2.3 (2.2) 11 74 49.2 (49.0) 5.3 (5.1) 2.2 (2.1) 12 56 50.0 (50.0) 5.4 (5.4) 2.2 (2.2) 252 (35 x dm3, 0.30 mmol) were stirred together in dichloro- methane for 30 min. The resulting solution was concentrated in vucuo and hexane was added to give [RhCl(C0)(8-OPhVPy)(P(OMe),)] as a yellow/beige solid (82 mg, 45%).[Calculated (found): C, 50.1 (49.9); H, 6.1 (6.0); N, 2.3 (2.4); C1, 5.9% (5.9%).] v,,,(CO) 1984cm-' (Nujol); 1989 cm-' (CHC1,); 1993 cm-' (Me,CO); 1990 cm-' (CH,NO,) (signals in acetone and nitromethane are broader than those in Nujol or chloroform). 31P NMR: 6 135.7 [d, 'J(RhP) 238 Hz]. Reaction of cis-[IrC1(C0),(8-OPhVPy)] with P(OMe),. A solution of cis-[IrCl(C0),(8-OPhVPy)](150 mg, 0.25 mmol) in ether (20 cm3) was treated with P(OMe), (31.4 x lop6dm3, 0.25 mmol) and stirred for 1 h. The solution was concentrated in uucuo and pentane (10 cm3) was added, precipitating slightly impure [IrC1(CO)(8-OPhVPy)(P(OMe)3]as a beige/yellow solid which was recovered by filtration, washed with pentane and dried in uucuo.Yield 60mg (40%). [Calculated (found): C, 43.6 (42.4); H, 5.3 (5.2); N, 2.0% (1.8Yo).] v,,,(CO) 1970 cm-' (Nujol); 1978 cm-' (CHCl,). ,'P NMR: 6 97.5 (s); 'H NMR: 6 3.83 [d, 4J(PH) 12.5 Hz, P(OCH,),]. Results and Discussion Synthesis The complexes cis-[MCl(CO),(n-OPhVPy)] [M =Rh, I(n); M =Ir, II(n)] were formed in a one-pot synthesis by reaction of the dimer [MCl(COD)], with two equivalents of 4-alkyl- oxystilbazole in dichloromethane under an atmosphere of co. I M=Rh II M=lr The reactions were easily carried out and proceeded in good yield giving yellow/orange (rhodium) or burgundy (iridium) products. Air was excluded from these reactions and this was particularly important in the synthesis of the Ir complexes where it appeared from infrared evidence that polynuclear carbonyl derivatives were formed if anaerobic conditions were not maintained.Complexes of both metals dissolved in CH2C12 to give yellow solutions which showed the expected two v(C0) bands in the infrared spectrum (at 2086 and 2012 cm- ' for rhodium and at 2074 and 1993 cm- for iridium), consistent with a mononuclear cis-dicarbonyl. The cis geometry was confirmed by 13C NMR of [IrCl(C0),(3- OPhVPy)], which showed two CO resonances at 169.5 and 168.2 ppm, rather than the single resonance which would be expected if the complex were trans. In the solid state, however, three infrared bands were observed for rhodium at 2083,2069 and 2007 cm- 'while for iridium, four bands were found 2075, 2050, 1988 and 1982 cm-'.In the case of rhodium, this extra band is attributed to solid-state splitting effects as the solid complexes were yellow/orange, typical of square-planar com- plexes of mononuclear Rh'. However, the solid iridium com- plexes were burgundy coloured which is not typical of square- planar mononuclear Ir' (usually yellow). In this case, the extra bands and the non-characteristic colour are attributed to a stacking of the Ir complexes in a manner similar to that observed with potassium tetracyanoplatinate.' The solid- state and solution electronic spectra were also different. In chloroform solution, the iridium complexes showed a strong absorption at 364 nm (E 30, 304 dm3 mol-' cm-'), while in J.MATER. CHEM., 1991, VOL. 1 the solid state, an additional band was seen at 500 nm. Such behaviour has previously been observed in the chemistry of Ir',', for example with [IrX(CO),] and with anions of the general formula [IrCl,(CO),]-and is more likely to be observed in Ir chemistry (with respect to Rh chemistry), owing to the greater spatial extension of the 5d orbitals facilitating stronger intermolecular interactions.' As yet, we have been unable to obtain suitable crystals to confirm this by X-ray crystallography. Comparison with [IrCl(CO),(pyridine)] ' reveals that the latter is also yellow in solution [v(CO) 2075 and 1981 cm-'1 but gives lustrous, deep-green crystals in the solid state where only two v(C0) are reported (at 2072 and 1998 cm-').One preparation of [Ir(CO),Cl(9-OPhVPy)] [II(9)] led to a brown/orange material (instead of the more usual burgundy solid) the analysis of which was consistent with the postulated complex, [Ir(CO),C1(9-0PhVPy)]. The mesomorphism of the two modifications was identical, although DSC showed an exotherm at 64 "C on heating, which corresponded to an observed orange/brown+burgundy transition (AH = -4 J g-') before the melting point; this was irreversible on cooling. The solid-state infrared spectra were also different. Whereas the burgundy solids showed strong v(C0) at 2077, 2051, 1988 and 1982 cm-', the orange/brown solid showed two of the bands at 2051 and 1982cm-' as shoulders with significantly reduced intensities. Both forms show exactly the same solution infrared spectrum [v(CO) at 2074 and 1993 cm- '1.Solid-state powder X-ray diffraction patterns were obtained for both modifications and confirmed that the two forms were quite different. Given the much lighter colour of the orange/brown form and the difference in the infrared spectra in the solid state, we suggest that there are few, if any, intermolecular Ir-Ir interactions present. Reaction of [RhCl(CO),(S-OPhVPy)] with P(OMe)3 led to [RhCl(CO)(P(OMe),}(8-OPhVPy)] which, in common with the related complexes [RhC1(CO)(P(OMe),)(n-OPhIPy)],6 was non-mesomorphic. Similarly, reaction of [IrCl(C0),(8-OPhVPy)] with P(OMe), in ether followed by addition of pentane led to a yellow/beige complex in moderate yield.Good microanalyses (error <f0.5%) could not be obtained for this material, although all the spectroscopic data pointed to [IrCl(CO){ P(OMe),)(8-0PhVPy)] with the phos- phite trans to the stilbazole. This complex was non-meso- morphic. Mesomorphism The rhodium complexes I(4)-1(12) and the iridium complexes II(5)-II(12) were all found to be mesomorphic. Rhodium The pale-yellow solids melted at temperatures between 85 and 110 "C into either a nematic phase (n=5, 6) or a smectic A (S,) phase (n=7-12); a monotropic nematic phase was ob- served for the complex with n=4. Once in the mesophase, rapid decomposition set in although this was slowed down when the observations were carried out under an inert atmos- phere. Even so, the clearing points quoted for the complexes are those of partially decomposed materials; enthalpy values for this transition are therefore of little use and hence not quoted. The transition temperatures given in Table 3 show the mesomorphic range to be somewhat greater than that observed for the related complexes based on 4-alkyloxy-N- (4-pyridylben~ylideneanilines).~Fig.1 shows the phase diagram. J. MATER. CHEM., 1991, VOL. 1 Table 3 Transition temperatures and thermal data for [RhCl(CO),(n-OPhVPy)] n transition Ti "C AH/J g-' 3 116 4 128 4 118 5 63 4.8 5 110 49.2 5 121 6 106 52.3 -6 124 7 87 47.0 7 117 a -7 130 8 85 46.9 8 123 a -8 130 9 85 51.1 -9 133 10 82 59.3 -10 137 11 85 63.8 -11 139 12 68 6.9 12 82 64.9 12 143 "Not seen by DSC.160r 140 100 80 , III 60' a I a I ' I * 2 4 6 8 10 12 carbon chain length Fig. 1 Phase diagram for [RhCI(CO),(n-OPhVPy)]: 0,K-I; +, N-I; 0, K-N; 0,K-SA; H, SA-1; 0,SA-N Iridium The iridium complexes were deep burgundy in the solid state and melted at temperatures between 80 and 110 "C to give either N or SA phases; the melting event was accompanied by a colour change to bright yellow which reversed on crystallisation. This is attributed to the complexes probably being stacked in the solid state and the stacking breaking down to give discrete molecular species on either dissolution (see above) or melting. These materials were more thermally stable than the corresponding rhodium complexes and no decomposition was apparent on entering the isotropic melt.The transition temperatures given in Table 4 again show the mesomorphic ranges of these complexes to be somewhat greater than those observed for the related complexes based on 4'-alkyloxy-N-(4-pyridylbenzylideneaniline~).~Fig. 2 shows the phase diagram. General Discussion The mesomorphism of these complexes deserves comment. The 4-alkyloxystilbazoles themselves are mesomorphic, show- ing a narrow range SB phase and a wider range (crystal) Table 4 Transition temperatures and thermal data for [IrCI(CO),(n-OPhVPy)] n transition T/ "C AHIJ g-' 5 108 43.6 5 (93) -0.2 6 106 6 110 33.2 (unresolved) 7 92 43.8 7 113 0.6 7 (87) -14.1 8 92 44.7 8 113 a 8 120 1.1 9 64 -4.0 9 87 55.6 9 121 2.4 10 87 57.2 10 131 2.1 11 88 65.2 11 133 2.6 12 85 57.1 12 137 6.1 " Not seen by DSC.160-140 -t 100"4 t t K 4 6 a 10 12 carbon chain length Fig. 2 Phase diagram for [IrCl(CO),(n-OPhVPy)]: 0, K-N;+, K-SA; 0,N-I; 0,SA-N; H, SA-1; 0,K-I E phase. However, complexation of these species to the d8 fragments [MCl(CO),] (M =Rh, Ir) results in complexes that are strongly mesomorphic, showing nematic and SA phases. Furthermore, while the melting point (regarded as the temperature at which a disordered, fluid mesophase is formed) of the stilbazoles is typically in the range 86-95 "C (359-368 K), complexation to, for example, iridium with a concomitant doubling of the molecular mass gives melting points in the range 82-1 10 "C (355-383 K).This contrasts strongly with their complexes of palladium and platinum (trans-[MCl,(n-OPhVPy)J) where the observed mesophases are at quite elevated temperatures (>200 "C).'" However, comparison with the platinum complexes trans-[PtCl,(n-OPhVPy)(olefin)] l4 is instructive as these latter materials show an SA mesophase at temperatures of ca. 50-70 "C. The materials described in this paper then support our earlier proposal14 that complexes with lower symmetry can have lower melting points. This is also sup- ported by the results obtained for [MC1(CO)z(n-OPhIPy)].5*6 In the literature of organic liquid crystals, it is known that extension of the molecular length without increase of the breadth of a molecule will increase the thermal stability of a mesophase.Furthermore, it is known that certain groups have a greater ability than others to promote the formation of Table 5 ATN1 (x-x')/OC data from this work and from ref. 7 and 16 X' X C-H (R=C8H17) C-H (R=C6H13)P C-NO2 2 I0 2 -8 C-CN 222 2 22 [N-IrCI(CO),] 2 24.1 221.5 [N -RhCI(CO)2] 241.1 235.5 a Hexyloxystilbene melts from crystal to isotropic liquid at 11 1 "C. No monotropic phases are observed on cooling. A B X = CN, NO2 C D Fig. 3 Species for which ATNI is evaluated nematic (or cholesteric) phases. This order:' Ph >NHCOCH3>CN >OCH3>NO2>C1> Br >N(CH3)2>CH3 >F >H is based on the result from simple statistical theories, which conclude that the internal energy stabilising the nematic phase is proportional to the nematic/isotropic transition temperature (TN1)in Kelvin.Thus, the higher the transition temperature, the more stable the mesophase. It is of interest to compare metal-containing moieties with organic groups for their relative effectiveness in promoting mesophase formation. Given the limited range of materials and metals studied, we can only make some observations concerning the groups [MCl(CO),] (M =Rh, Ir) based on the corresponding stilbazole mesogens. In order to do this, it is necessary to evaluate the difference (ATNI) between TNIof a metal stilbazole complex and that of the parent stilbazole and then to compare this with the corresponding difference between TNIof a cyano- or nitro-alkyloxystilbene (as represen- tative examples of terminal groups) and that of an unsubsti- tuted alkyloxystilbene [i.e.compare TNI(A)-TNI(B) with TNI(C)- TNI(D) (see Fig. 3)]. This cannot be done precisely because the core stilbazole and stilbene fragments do not have accessible nematic mesophases. However, they will have a virtual TNIwhich is below their melting point or mesophase clearing point (Kp) and hence the change, ATNI,which is J. MATER. CHEM., 1991, VOL. 1 promote nematic (and other phases) will eventually emerge from studies of metal-containing liquid-crystal systems. We thank Professors Jose-Luis Serrano and Luis Oro (Insti- tuto de Ciencias de Materiales de Aragon) for useful dis- cussions and Dr.David Nicholls (University of Liverpool) for the solid-state electronic spectrum. We acknowledge support from the Royal Society (D.W.B.), SERC (J.R.M., S.E.H.),. the Spanish Ministry of Education (E.S.), the EC Twinning (Grant No. ST2J-O387C), the EC Erasmus Programme (RLeL), Mal- colm P. McDonald for help with the powder diffraction data, Johnson Matthey for generous loans of iridium and rhodium salts and BDH (S.E.H.) and Johnson Matthey (J.R.M.) for CASE awards. References 1 See e.g. H. Abied, D. Guillon, A. Skoulios, P. Weber, A-M. Giroud-Godquin and J-C. Marchon, Liq. Cryst., 1987, 2, 269; C. Piechocki and J. Simon, Noun J.Chim., 1985,9, 159; K. Ohta, H. Ema, H. Muroki, I. Yamamoto and K. Matsuzaki, Mol. Cryst. Liq. Cryst., 1987, 147,61; M. Khan, J. Bhatt, B. M. Fung, K. M. Nicholas and E. Wachtel, Liq. Cryst., 1989,5, 285; Yu. G. Galya- metdinov, I. G. Bikchantaev and I. V. Ovchinnikov, Zh. Obs. Khim., 1988, 58, 1326; M. Ghedini, S. Armentano, R. Bartolino, F. Rustichelli, G. Torquati, N. Kirov and M. Petrov, Mol. Cryst. Liq. Cryst., 1987, 151A,75. 2 H. Adams, N. A. Bailey, D. W. Bruce, R. Dhillon, D. A. Dunmur, S. E. Hunt, E. Lalinde, A.A. Maggs, R. Orr, P. Styring, M. S. Wragg and P. M. Maitlis, Polyhedron, 1988, 7, 1861; U. T. Mueller-Westerhoff, A. Nazzal, R. J. Cox and A-M. Giroud, J. Chem. SOC., Chem. Commun., 1980, 497; M.Ghedini, S.Licoccia, S. Armentano and R. Bartolino, Mol. Cryst. Liq. Cryst., 1984, 108, 269; P. Espinet, E. Lalinde, M. Marcos, J. Perez and J-L. Serrano, Organometallics, 1990, 9, 555; J. Barbera, P. Espinet, E. Lalinde, M. Marcos and J. L. Serrano, Liq. Cryst., 1987, 2,833. 3 D. W. Bruce, E. Lalinde, P. Styring, D. A. Dunmur and P. M. Maitlis, J. Chem. SOC., Chem. Commun., 1985, 581; H. Adams, N. A. Bailey, D. W. Bruce, D. A. Dunmur, E. Lalinde, M. Marcos, C. Ridgway, A. J. Smith, P. Styring and P. M. Maitlis, Liq. Cryst., 1987, 2, 381. 4 A-M. Giroud-Godquin, J-C. Marchon, D. Guillon and A. Skoulios, J. Phys. Chem., 1986, 90,5502. 5 M. A. Esteruelas, E. Sola, L. A. Oro, M. B. Ros and J. L. Serrano, J. Chem. SOC., Chem. Commun., 1989, 55. 6 M. A.Esteruelas, E. Sola, L. A. Oro, M. B. Ros, M. Marcos and J. L. Serrano, J. Organomet. Chem., 1990, 387,103. 7 D. W. Bruce, D. A. Dunmur, E. Lalinde, P. M. Maitlis and P. Styring, Liq. Cryst., 1988, 3,385. 8 D. W. Bruce, D. A. Dunmur, E. Lalinde, P. M. Maitlis and P. Styring, Nature (London), 1986, 323,791; D. W. Bruce, D. A. Dunmur, P. M. Maitlis, P. Styring and (in part) M. A. Esteruelas, L. A. Oro, M. B. Ros, J. L. Serrano and E. Sola, Chem. Muter., 1989, I, 479; C. Bertram, D. W. Bruce, D. A. Dunmur, S. E. Hunt, P. M. Maitlis and M. McCann, J. Chem. SOC., Chem. Commun., 1991, 69. 9 G. Giordano and R. H. Crabtree, Inorg. Synth., 1979, 19, 218. 10 J. J. Herde, Znorg. Synth., 1975, 15, 18. 11 See e.g. A. E. Underhill, in Comprehensive Coordination Chemis- try, ed. G. Wilkinson, R. D. Gillard and J. A. McCleverty, Perga- mon, Oxford, 1987, vol. 6, ch. 60, p. 133. 12 G. J. Leigh and R. L. Richards, in Comprehensive Organometallic Chemistry, ed. G. Wilkinson, E. W. Abel and E. G. A. Stone, Pergamon, Oxford, 1982, vol. 5, ch. 36, p. 541. 13 W. Hieber and V. Frey, Chem. Ber., 1966,99, 2607. assumed to be directly proportional to the nematic stabilis- 14 J. P. Rourke, F. P. Fanizzi, N. J. Salt, D. W. Bruce, D. A. Dunmur ation energy, can be defined as A TNI2TNl(complex) 15 and P. M. Maitlis, J. Chem. SOC., Chem. Comrnun., 1990, 229. -zp(core).K. J. Toyne, in Thermotropic Liquid Crystals, ed. G. W. Gray, These values are collected in Table 5. From Table 5 it can be Wiley, Chichester, 1987. concluded that the group [IrCl(CO),] is at least as good as 16 D. Demus and H. Zaschke, Fliissige Kristalle in Tabellen,-C=N in its ability to stabilise a nematic phase, while Deutscher Verlag fur Grundstoffindustrie, Leipzig, 1984, vol. 11. [RhCl(CO),] is significantly more effective. This suggests that an order for the ability of certain inorganic fragments to Paper 0104532B; Received 9th October, 1990

ISSN:0959-9428    

DOI:10.1039/JM9910100251

出版商:RSC    

年代:1991

数据来源: RSC