摘要:

MATERIALS CHEMISTRY COMMUNICATIONS A macrocyclic aromatic thioether ketone: synthesis, structure and anionic ring-opening polymerisation Howard M. Colquhoun,*a David F. Lewis,a Richard A. Fairman,†a Ian Baxterb and David J. Williams*b aDepartment of Chemistry, University of Manchester, Manchester, UK M13 9PL bDepartment of Chemistry, Imperial College, South Kensington, L ondon, UK SW 7 2AY reaction mixture on cooling.This novel polymer was recovered Nucleophilic polycondensation at high concentration between in 86% yield by filtration, and proved to be crystalline in the benzene-1,3-dithiol and 4,4¾-difluorobenzophenone affords not ‘as-made’ state [Tm=195°C by differential scanning caloronly linear polymer 1 but also a significant yield (ca. 8%) of imetry (DSC)].It did not, however, crystallise from the melt, cyclic oligomers, mainly the [2+2] cyclodimer 2. Under high- showing only a glass transition (Tgonset=107°C) on the second dilution conditions 2 becomes by far the major reaction heating scan. The original filtrate slowly (over several days) product. Macrocycle 2 has been isolated, structurally deposited colourless crystals of a pure cyclic oligomer (3% characterised by single-crystal X-ray methods and found to yield) identified as the [2+2] cyclodimer 2 by mass spec- undergo rapid anionic ring-opening polymerisation in the melt trometry (M+ 640) and 13C NMR spectroscopy.‡ Further (330–360°C), affording a linear, high molecular weight oligomeric material (2% yield) was precipitated by diluting poly(arylthioether ketone).the DMAc solution with MeOH, and mass spectrometry indicated that this comprised a mixture of the cyclodimer 2 with the homologous [3+3] and [4+4] cyclic condensation products 3 and 4 respectively. In order to increase the yield of cyclic material, the reaction Since the first reports of the synthesis and anionic ring-opening was next run at much lower concentration (1.5 wt% conden- polymerisation of macrocyclic aromatic ethers,1,2 there has sation products).The cyclic dimer 2 was now isolated in a been a rapid increase in interest in this potential approach to remarkable 72% yield by column chromatography on silica the production and fabrication of high-performance aromatic gel (351 CH2Cl2MCHCl3 as eluent), although yields of the polymers.3 In general, the very high melt viscosities of commer- higher cyclic oligomers 3 and 4 were still relatively modest at cially available aromatic polyethers mean that their molecular ca. 1% and 8%, respectively. weights (and dependent properties such as fracture-toughness) must be deliberately restricted if melt-processing is to be feasible.4 However, no such restriction holds for reactive processing of low-viscosity cyclic oligomers which, once fabricated, can in principle be advanced to extremely high molecular weights.Moreover, the low initial viscosity of this type of material promotes rapid penetration and wetting of reinforcing fibres, so that cyclic oligomers are potentially attractive intermediates in the fabrication of very high fibre-content composite materials.Key features of this approach are that (i) there are no volatile co-products to generate voids in the final product, and (ii) the final polymer is linear rather than cross-linked, leading to much greater ductility and toughness than is possible for thermosetting materials such as the epoxy resins and bis(maleimide)s. Recent results from Wang et al. suggest that the thioether linkage may be a particularly versatile initiation site for the ring-opening polymerisation of aromatic macrocycles.5 To date however only one specific macrocyclic thioether ketone, the all-para cyclic trimer (SC6H4COC6H4)3, has been isolated and characterised in detail.6 Here we report the high yield synthesis, structural characterisation and remarkably facile polymerisation chemistry of a new type of macrocyclic aromatic thioether ketone, obtained by nucleophilic cyclocondensation of Scheme 1 Reagents and conditions: i, K2CO3 , DMAc, 160 °C benzene-1,3-dithiol with 4,4¾-difluorobenzophenone. Macrocycle 2 was initially isolated as a by-product of the polycondensation between benzene-1,3-dithiol and 4,4¾- difluorobenzophenone (Scheme 1).The reaction was run at high concentration (ca. 25 wt% polymer) in N,N-dimethylac- ‡ Selected data for 2: Mp (DSC) 329°C; IR (KBr disc) n(CNO) etamide (DMAc), and polymer 1 precipitated rapidly from the 1650 cm-1; 13C NMR (300 MHz, CD2Cl2–CH3SO3H, 251 v/v): d 127.34, 128.19, 131.88, 132.73, 135.74, 137.22, 141.05, 155.69, 201.13 (Found: C, 70.9; H, 3.9; S, 20.2. Calc. for C38H24O2S4: C, 71.2; H, 3.8; † Present address: Department of Chemistry, Faculty of Natural Sciences, University of the West Indies, St Augustine, Trinidad, W.I.S, 20.0%). J. Mater. Chem., 1997, 7(1), 1–3 1Scheme 2 Fig. 1 Molecular structure of macrocycle 2, showing the solvating molecules of chloroform. Selected bond lengths and angles: S(1)–C(2), 1.762(3); C(12)–S(15), 1.769(3); C(5)–C(8), 1.485(4); S(15)–C(66), Oligomer 2 was structurally characterised (Fig. 1) as its 1.781(3); C(8)–O(8), 1.223(3); C(20)–S(1A), 1.783(3); C(8)–O(9), chloroform solvate by single crystal X-ray methods.§ The 1.489(4)A° ; C(8)–C(9)–C(14), 123.0(3); S(1)–C(2)–C(3), 115.9(2); C(11)–C(12)–S(15), 115.5(2); S(1)–C(2)–C(7), 124.9(2); macrocycle has a rather open geometry with crystallographic C(13)–C(12)–S(15), 124.4(2); C(4)–C(5)–C(8), 117.9(2); inversion symmetry at its centre. The benzene-1,3-dithiolate C(12)–S(15)–C(16), 104.2(1); C(6)–C(5)–C(8), 123.7(3); residues are directed above and below the mean plane of the S(15)–C(16)–C(17), 119.6(2); C(5)–C(8)–O(8), 118.8(3); macrocycle (as defined by the four sulfur atoms) and are S(15)–C(16)–C(21), 119.8(3); C(5)–C(8)–C(9), 121.3(2); inclined at ca. 95° to this plane, leading to an overall chair- C(19)–C(20)–S(1A), 119.9(3); O(8)–C(8)–C(9), 119.9(3); type conformation. Non-bonded repulsions within the diaryl C(21)–C(20)–S(1A), 119.2(3); C(8)–C(9)–C(10), 118.6(3); C(20)–S(1A)–C(2A), 103.7(1)°. thioether unit lead to significant bond-angle distortions at aromatic carbon atoms [C(2) and C(12)] linked to the thioether bridges.The nominally trigonal bond angles C(3)MC(2)MS(1) and oxide anion, which is present at a relatively high level (ca. C(7)MC(2)MS(1) have, in this structure, values of 115.9(2) and 5 mol% relative to the cyclic oligomer), is ultimately displaced 124.9(2)°, respectively. Similar distortions are also found in from the growing polymer chain, and thus behaves as a genuine linear aryl ethers7 and thioethers,8 and in the present structure catalyst for ring-opening polymerisation rather than as a the bridge-bond angles at sulfur and the carbonyl carbon atoms classical, irreversible initiator.are entirely normal at ca. 104 and 121° respectively. There is This ring-opening polymerisation process can readily be however clear evidence for some degree of ring-strain, in that followed by DSC, the characteristically sharp oligomer-melting the sulfur atoms lie ca. 0.16A° out of the plane of the benzene- peak observed at 329 °C in the initial heating scan (Fig. 2) 1,3-dithiolate residue. As shown in Fig. 1, the crystal contains being replaced in subsequent scans by the glass transition two molecules of chloroform per macrocycle. Although the (onset at 110°C) of the polymeric product.Given the timescale solvent molecules exhibit some orientational disorder in the crystal, both orientations result in their C–H bonds being directed almost linearly (C–H–O=160°) towards the carbonyl oxygens of the macrocycle; the H,O separation of 2.14 A° being indicative of a reasonably strong hydrogen bond. On heating to ca. 350 °C in the presence of an anionic initiator such as the potassium salt of 4-hydroxybenzophenone (2 wt%), macrocycle 2 was found to undergo rapid polymerisation in the melt (Scheme 2), to give tough, transparent, fully soluble, high molecular weight polymer 1, identical by 13C NMR spectroscopy with the material produced by solution polycondensation. The ring-opening polymerisation however gives material of very much higher inherent viscosity (see below).This result suggests that the initiating benzoylphen- § Crystal data for 2: C38H24O2S4·2CHCl3 , M=879.55, triclinic, space group P1 � , a=9.118(1), b=10.333(2), c=12.262(1), a=102.40(1), b= 95.663(1), c=114.052(1)°, Z=1, U=1007.6(2) A° 3, Dc=1.449 g cm-3, F(000)=448. The structure was solved by direct methods and the non-H atoms were refined anisotropically by full-matrix least-squares, based on F2.The H atoms were refined isotropically (riding model). R1 [I>2s(I)] was 0.057, and wR2 was 0.147 for 3043 independent observed reflections. Fig. 2 DSC thermograms (10 °C min-1 heating rate) showing (a) the Atomic coordinates, thermal parameters, and bond lengths and angles have been deposited at the Cambridge Crystallographic Data melting endotherm of pure macrocycle 2, (b) the melting endotherm of macrocycle 2 containing 2 wt% of initiator, and (c) the glass Centre (CCDC).See Information for Authors, J. Mater. Chem., 1997, Issue 1. Any request to the CCDC for this material should quote the transition of polymer 1 produced by ring-opening polymerisation during the course of scan (b) full literature citation and the reference number 1145/23. 2 J. Mater. Chem., 1997, 7(1), 1–3of the DSC experiment, in which the sample is simply heated and United Utilities plc. The award of a Royal Society Industry Fellowship to H. M. C. is gratefully acknowledged. to 360 °C at 10°C min-1 and then immediately allowed to cool, high molecular weight polymer must be produced within a few minutes of the macrocycle reaching its melting point.References Polymer 1 is amorphous when produced by ring-opening polymerisation in the melt, but it crystallises rapidly on contact 1 (a) J. A. Cella, J. J. Talley and J. M. Fukuyama, Polym. Prepr. Am. Chem. Soc., Div. Polym. Chem., 1989, 30(2), 581; (b) J. A. Cella, with dipolar aprotic solvents such as N-methylpyrrolidone J.M. Fukuyama and T. L. Guggenheim, Polym. Prepr. Am. Chem. (NMP) or DMAc, in which solvents the crystalline polymer is Soc., Div. Polym. Chem., 1989, 30(2), 142. insoluble. Viscosity and NMR measurements were therefore 2 H. M. Colquhoun, C. C. Dudman, M. Thomas, C. A. O’Mahoney carried out using a mixture of CH2Cl2 and methanesulfonic and D. J. Williams, J. Chem. Soc., Chem.Commun., 1990, 336. acid (251 v/v) as solvent. Inherent viscosities of 2.98 and 1.15 dl 3 (a) M. J. Mullins, E. P. Woo, D. J. Murray and M. T. Bishop, CHEMTECH, 1993 (August), 25; (b) D. Xie and H. W. Gibson, g-1 for the polymers derived from ring-opening and from Polym. Prepr. Am. Chem. Soc., Div. Polym. Chem., 1994, 35(1), 401; polycondensation respectively clearly indicate that a high (c) K.P. Chan, Y-F. Wang, A. S. Hay, X. L. Hrnowski and molecular weight polythioether ketone is very readily produced R. J. Cotter, Macromolecules, 1995, 28, 6705; (d) Y. Ding and by anionic (nucleophilic) ring-opening polymerisation of A. S. Hay, Macromolecules, 1996, 29, 3090 and references cited macrocycle 2, more readily indeed than by a conventional therein. nucleophilic polycondensation reaction.Moreover, the absence 4 High Performance Polymers: T heir Origin and Development, ed. R. B. Seymour and G. S. Kirshenbaum, Elsevier, New York, 1986. of side reactions during ring-opening polymerisation (as indi- 5 Y-F. Wang, K. P. Chan and A. S. Hay, Macromolecules, 1996, 29, cated by the clean 13C NMR spectrum and complete solubility 3717. in CH2Cl2–MeSO3H of the polymer produced) suggests that 6 Y. E. Ovchinnikov, V. I. Nedelkin, S. I. Ovsyannikova and this class of macrocycle may prove to be of particular value in Y. T. Struchkov, Russ. Chem. Bull., 1994, 43, 1384. the development of ring-opening methodology for the synthesis 7 H. M. Colquhoun, C. A. O’Mahoney and D. J. Williams, Polymer, of high-performance aromatic polymers. 1993, 34, 218. 8 K. Hasebe, T. Asahi, A. Ishazawa and K. Izumi, Acta Crystallogr., Sect. C. 1989, 45, 2023. This work was supported by the Engineering and Physical Sciences Research Council of the United Kingdom, the Communication 6/07044B; Received 15th October, 1996 University of the West Indies Staff Development Programme J. Mater. Chem., 1997, 7(1), 1–3

ISSN:0959-9428    

DOI:10.1039/a607044b

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Thiophenotribenzoporphyrazines: novel near-IR absorbing dyes Michael J. Cook* and Ali Jafari-Fini School of Chemical Sciences, University of East Anglia, Norwich, UK NR4 7T J The Ærst examples of thiophenotribenzoporphyrazines have been synthesised. The compounds obtained bear either six or eight alkyl chains and show Q-band absorptions which are red shifted relative to analogous phthalocyanine derivatives.Their formulations as spin coated Ælms show broad band absorption extending into the near IR. The compounds exhibit discotic mesophase behaviour. Phthalocyanines (Pcs) show interesting optoelectronic and photophysical properties, many of which are associated with the intense and relatively narrow-band absorption in the visible region spectrum, the Q-band.1 Manipulation of the energy levels of the p orbitals associated with the Q-band through the incorporation of ring substituents or benzofusion (as in the naphthalocyanines) provides access to far-red±near IR absorbing dyes.2 Such materials have potential in medical applications such as photodynamic therapy3 and in optical data storage devices.4 Here we describe the Ærst examples of the related thiophenotribenzoporphyrazine system and demonstrate signiÆcant red shifting of the Q-band which arises on replacement of one of the benzenoid rings of the Pc core with a thiophene unit.Furthermore, the distortion of the p system in the new series of macrocycles may lead to interesting nonlinear optic (NLO) behaviour and, as shown here, gives rise to a broad band absorption which may render the new materials beneÆcial for applications requiring light harvesting within the far-red±near IR region of the spectrum.5 Linstead and co-workers obtained tetrathiophenoporphyrazine, presumably as an isomeric mixture, from 2,3-dicyanothiophene during their classic studies of phthalocyanines in the 1930's, describing the product as being greener than phthalocyanine itself.6 The thiophenotribenzoporphyrazines described here, containing just one thiophene unit, have been obtained by reacting both 3,4- and 2,3-dicyanothiophene,7 as well as the 2,5- dioctyl derivative of the former, with an excess of a 3,6- dialkylphthalonitrile8 (Scheme 1).The required products were obtained by conventional workup as the metal-free analogues 1 (7%), 2a (12%), 2b (1%) and 2c (11%) after separation (column chromatography over silica) from by-products, the major one being the octaalkyl Pc 3.The metallated analogue 2d was prepared from 2c by reaction with nickel acetate in pentanol solution heated to reØux. Each compound gave satisfactory analytical data, and the 1H NMR spectra of the metal- N N N N N N S N N H H N N N N N N N N M S R R R R R R R R¢ R¢ R R R R R N N N N N N N N H H R R R R R R S NC NC S NC R¢ S Br R¢ S R¢ R¢ CN R¢ Br R¢ R NC i ii NC iii R iii 1 R = C8H17 2a R = C8H17, R¢ = H, M = 2H 2b R = R¢ = C8H17, M = 2H 2c R = C6H13, R¢ = H, M = 2H 2d R = C6H13, R¢ = H, M = Ni 3 R R free compounds gave signals readily interpreted in terms of their Scheme 1 Reagents and conditions: i, Br2, MeCO2H, -5°C, 18 h; ii, structure (Table 1).Compound 2d showed a satisfactory low CuCN, DMF, 6 h; iii, LiOC5H11MC5H11OH, AcOH [followed by resolution FAB-MS but the metal-free derivatives were prone M(OAc)2 for the metallated derivative] to fragmentation. The compounds are soluble in solvents such as THF, toluene and cyclohexane, a property attributable to the presence of the long alkyl chains.The latter also promote develop a second, lower temperature liquid crystal phase characcolumnar mesophase behaviour (Scheme 2); polarised light terised by a needle-like texture; this we tentatively assign to a microscopy shows that each compound gives rise to amesophase columnar mesophase having rectangular symmetry, i.e. Drd. with a fan type structure on cooling from the isotropic liquid The Q-band absorptions in the visible region spectra of (I).This is characteristic of the hexagonal columnar mesophase solutions in cyclohexane at ca. 1×10-6 M are given in Table 1 (Dhd) also exhibited by non-peripherally octaalkyl-substituted and examples are displayed in Fig. 1. The band shapes of 1, 2a and 2b are more complex than for metal-free and metallated phthalocyanines of series 3.9a Compounds 2b and 2d also J.Mater. Chem., 1997, 7(1), 5±7 5Table 1 Characterisation data for thiophenotribenzoporphyrazines molecular formula lmax (e×105) for solutions in cyclohexane Found; C, H, N 1H NMR [lmax (relative peak intensity) compound (required) dH (270 MHz, C6D6) of the spin-coated Ælms] C78H112N8S 1 -1.32 (s, 2H), 0.8±1.0 (m, 18H), 1.2±1.9 (m, 56H), 720 (1.64), 700 (0.85), 679 (0.80), 648 (1.11), 347 78.50 9.24 9.39 1.9±2.08 (t, 2H), 2.08±2.25 (t, 2H), 2.3±2.6 (m, 12H), (0.79) (78.47 9.46 9.39) 4.35 (t, 4H), 4.58 (t, 4H), 4.70 (t, 4H), 7.54 (d, 1H), [784 (0.57), 624 (0.68), 334 (1.0)] 7.77 (m, 4H), 7.94 (s, 2H), 8.40 (d, 1H). 2a C78H112N8S -0.58 (s, 2H), 0.8±0.95 (m, 18H), 1.2±2.0 (m, 60H), 742 (2.40), 705 (0.71), 687 (0.63), 663 (1.15), 333 78.48 9.55 9.26 2.28±2.45 (m, 12H), 4.34 (t, 4H), 4.62±4.72 (m, 8H), (0.78) (78.47 9.46 9.39) 7.79±7.90 (m, 6H), 8.28 (s, 2H).[793 (0.68), 658 (0.63), 334 (1.0)] 2b C94H144N8S -0.17 (s, 2H), 0.87 (t, 24H), 1.2±1.9 (m, 80H), 2.11 768 (1.39), 729 (0.63), 708 (0.68), 677 (0.79), 337 79.59 10.37 7.75 (t, 4H), 2.2±2.4 (m, 12H), 3.82 (t, 4H), 4.27 (t, 4H), (0.75), 307 (0.69), 284 (0.6) (79.61 10.23 7.90) 4.56±4.67 (m, 8H), 7.7±7.84 (m, 6H). [817 (0.55), 657 (0.61), 341 (1.0)] 2c C66N88N8S -1.27 (s, 2H), 0.92 (m, 18H), 1.2±2.0 (m, 36H), 2.1±2.4 742 (1.81); 705 (0.54); 687 (0.48), 662(0.88), 386 76.99 8.64 10.90 (m, 4H), 4.03 (m, 4H), 4.49 (m, 4H), 4.60 (m, 4H), (0.36), 336 (0.56) (77.29 8.65 10.93) 7.64±7.67 (d, 2H), 7.74±7.78 (d, 2H), 7.83 (s, 2H), [795 (0.99), 657 (0.69), 371 (0.91), 328 (1.0)] 8.03 (s, 2H). 2d C66H86N8SNi 0.91 (m, 18H), 1.05±1.85 (m, 36H), 2.15±2.35 (m, 12H), 716 (2.35), 681 (1.89), 614 (0.43), 333 (0.6), 298 73.22 7.89 10.40 3.95 (m, 4H), 4.39 (m, 4H), 4.53 (m, 4H), 7.50±7.54 (1.1) (73.25 8.01 10.35) (d, 2H), 7.60±7.64 (d, 2H), 7.8 (s, 2H), 8.07 (s, 2H). [765 (1.0), 660 (0.77), 335 (0.85)] 1 2a 2b 2c 2d Scheme 2 Phase transitions determined by differential scanning calorimetry (DSC) and optical microscopy.Enthalpy measurements were Fig. 1 The visible region spectra of (a) 1, (b) 2a, (c) 2b and (d) 2d as ca. determined by DSC (DH/J g-1) 1×10-6 M solutions in cyclohexane Pcs due to their lower symmetry.10 The lmax of the Q-band is sensitive to the type of thiophene fusion and to substituents. give 2d, the Q-band is blue shifted and the higher symmetry of the system (D2h) leads to a simpler band structure which is Thus the lowest energy component of the Q-band of 1 (lmax 720 nm) is marginally to the red of the Q-band of a similarly reminiscent of that for metal-free Pcs such as 3; these are also of D2h symmetry.substituted hexaalkylphthalocyanine (Qx 714, Qy 676 nm).8 The corresponding band absorbs further to the red in the Many actual or potential applications of Pc derivatives utilise the compounds in the solid state.We therefore formu- isomeric compound 2a (lmax 742 nm) and further still upon addition of two alkyl chains onto the thiophene ring as in 2b lated all Æve compounds as spin-coated Ælms by administering a drop of a solution of each compound in THF (ca. 2 mg in (lmax 768 nm). This is 40 nm to the red of the corresponding phthalocyanine derivative 3 (R=octyl). On traversing the 0.5 ml) onto a glass slide rotating at 2000 rpm.11 Examples of the absorption spectra of the Ælms obtained once the solvent series 1, 2a and 2b, there is an increase in the spacing between the two main Q-band components. Upon metallation of 2c to had evaporated appear in Fig. 2 (see also Table 1). The spectra 6 J. Mater. Chem., 1997, 7(1), 5±7series of compounds are under further investigation in order to evaluate their efficiency as photoconducting materials. We thank the EPSRC and ICI Wilton Research Centre for a CASE studentship and Professor T. J. Ryan of EPIGEM Ltd for his support. FAB-MS were obtained from the EPSRC service centre at Swansea.References 1 (a)Phthalocyanines–Properties and Applications, ed. C. C. Leznoff and A. B. P. Lever, VCH, New York, 1989; (b) M. J. Cook, in Spectroscopy of NewMaterials, eds. R. J. H. Clark and R. E. Hester, Wiley, Chichester, 1993, p. 87. 2 (a) N. Kobayashi, in Phthalocyanines–Properties and Applications, eds. C. C. Leznoff and A. B. P. Lever, VCH, New York, 1993, vol. 2, p. 101; (b) M. J. Cook, A. J. Dunn, S. D. Howe, A. J. Thomson and K. J. Harrison, J. Chem. Soc., Perkin T rans. 1, 1988, 2453. 3 (a) R. Bonnett, Chem. Soc. Rev., 1995, 19; (b) P. Margaron, R. Langlois, J. E. van Lier and S. Gaspard, J. Photochem. Photobiol. B: Biol., 1992, 14, 187. 4 (a) P. A. Hunt, in Chemistry and T echnology of Printing and Imaging Systems, ed.P. Gregory, Blackie Academic & Professional, Glasgow, 1996, p. 168; (b) R. Ao, L. Kummerl and D. Haarer, Adv.Mater., 1995, 7, 495. 5 P. Gregory, High-T echnology Applications of Organic Colorants, Plenum Press, New York, 1991, p. 45. 6 R. P. Linstead, E. G. Noble and J. M. Wright, J. Chem. Soc., 1937, 911. 7 J. Morel, C. Paulmier and P. Pastour, C. R. Acad. Sci. Paris, Ser.Fig. 2 The visible region spectra of spin coated Ælms of (a) 1, (b) 2b C, 1968, 266, 1300. and (c) 2c 8 I. Chambrier, M. J. Cook, S. J. Cracknell and J. McMurdo, J.Mater. Chem., 1993, 3, 841. 9 (a) A. S. Cherodian, A. N. Davies, R. M. Richardson, M. J. Cook, N. B. McKeown, A. J. Thomson, J. Feijoo, G. Ungar and show unusually broad Q-band absorptions extending well into K.J. Harrison, Mol. Cryst. L iq. Cryst., 1991, 196, 103; (b) J. Simon the near-IR region. The broadest band absorption is observed and P. Bassoul, in ref. 2(a), p. 223; (c) D. W. Bruce, J. Chem. Soc., for the Ælm of the octaoctyl substituted compound 2b, which Dalton T rans., 1993, 2983. includes lmax at 657 and 817 nm; the spin-coated Ælm of the 10 (a) J. A. Elvidge, J. H. Golden and R.P. Linstead, J. Chem. Soc., corresponding phthalocyanine 3 (R=C8H17) shows lmax 635 1957, 2466; (b) N. Kobayashi, R. Kondo, S. Nakajima and T. Osa, J. Am. Chem. Soc., 1990, 112, 9640; (c) T. F. Baumann, J. W. Sibert, and 765 nm.11 M. M. Olmstead, A. G. M. Barrett and B. M. Hoffman, J. Am. In summary, we have obtained the Ærst examples of some Chem. Soc., 1994, 116, 2639; (d) N. Kobayashi, M. Togashi, T. Osa, thiophenotribenzoporphyrazine derivatives. The site of fusion K. Ishii, S. Yamauchi and H. Hino, J. Am. Chem. Soc., 1996, 118, of the thiophene ring, substituents and incorporation of a 1073. metal ion provide a means of tuning the energy of the visible 11 S. M. Critchley, M. R. Willis, M. J. Cook, J. McMurdo and region absorption band and extending it into the far-red. Films Y. Maruyama, J.Mater. Chem., 1992, 2, 157. fabricated by the spin coating technique show exceptional broad band absorption extending into the near IR. These Communication 6/07004C; Received 14th October, 1996 J. Mater. Chem., 1997, 7(1), 5±7 7

ISSN:0959-9428    

DOI:10.1039/a607004c

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Symmetric and non-symmetric liquid crystal dimers with branched terminal alkyl chains: racemic and chiral Andrew E. Blatch, Ian D. Fletcher and Geoffrey R. Luckhurst* Department of Chemistry, University of Southampton, Highfield, Southampton S017 1BJ, UK Two new series of racemic dimeric liquid crystals containing terminal alkyl chains with methyl branches are presented; in one the dimers are symmetric and in the other they are non-symmetric.Comparisons are made with the structurally isomeric series with no methyl branches in their terminal alkyl chains. The stability and clearing temperatures of the liquid crystal phases are markedly reduced for the dimers with methyl branches, particularly for the symmetric dimers. Chiral analogues of both series were also examined. The liquid crystal behaviour of the non-symmetric dimers is of particular interest. Four homologues were studied with the number of methylene units in the spacer ranging from six to nine.An odd–even effect was observed for the chiral properties of these materials. Blue phase I behaviour was exhibited by the heptane and nonane homologues but not by the even membered hexane and octane members of the series.The chiral properties of low molar mass liquid crystals have or even-membered flexible alkyl spacers. Even-membered spacers possess more conformers which allow an essentially paral- continued to fascinate researchers ever since the discovery of the liquid crystalline state in 1888.1 Today a plethora of chiral lel arrangement of the mesogenic cores whereas odd-membered spacers tend to have more conformers which cause the mesog- liquid crystal phases are known to exist.2 Such interest stems not only from the applications of chiral liquid crystal materials ens to adopt an unfavourable bent configuration. Given such a marked dependence on the parity of the spacer we might in, for example, thermochromic devices and the use of the ferroelectric smectic C phase in electro-optics3 but also results expect the form chirality of the chiral phases exhibited by dimers with an odd- or even-membered spacer also to vary. from their considerable fundamental importance as illustrated by the discoveries of the blue phases and the twist grain In order to further our understanding of the structure–property relationships of racemic dimeric liquid crystals with branched boundary (TGB) phase.Examples in the literature of chiral liquid crystal compounds are dominated to a large degree by terminal alkyl chains, and to examine the effect of the flexible alkyl spacer on the chiral properties of their optically pure low molar mass monomeric materials and to a lesser extent by chiral polymer liquid crystals.However, examples of chiral analogues, we have synthesized two new homologous series of dimeric liquid crystals containing the 2-methylbutyl group in liquid crystal dimers, i.e. compounds composed of two mesogenic groups linked together by a flexible spacer, are scarce. the terminal position: first the symmetric a,v-bis{4-[4-(2- methylbutyl)phenyliminomethyl]phenoxy}alkanes and sec- The studies which do exist have either considered the influence of a chiral, flexible alkyl spacer on the phase behaviour of ondly the non-symmetric a-(4¾-cyanobiphenyl-4-yloxy)-v-{4- [4-(2-methylbutyl)phenyliminomethyl]phenoxy}alkanes. As liquid crystal dimers4,5 or have examined the induction of chiral nematic phases in nematic hosts by the addition of non- well as the racemic compounds we have also prepared a selection of their chiral analogues.The symmetric dimers were mesogenic dimers with chiral spacers.6 Barbera� et al.4 have prepared a series of alkyloxy-substituted Schiff’s base dimers chosen as their monomeric counterparts are known and typically have a nematic character for short terminal alkyl chains with a spacer derived from (S)-2-chlorosuccinic acid.By increasing the dipole moment with a chlorine atom at the with increased smectic polymorphism as the terminal alkyl chain is lengthened.9,10 The structurally isomeric, unbranched chiral centre it was hoped that low molar mass compounds with enhanced ferroelectric properties could be obtained. analogues of these dimers, the a,v-bis[4-(4-pentylphenyliminomethyl) phenoxy)alkanes, have also been studied in detail However, although a ferroelectric smectic C phase was formed by all of the homologues, the compounds proved unstable and and exhibit a rich smectic polymorphism.11 The corresponding non-symmetric dimers, the a-(4¾-cyanobiphenyl-4-yloxy)-v-[4- decomposed on heating.One of the few published reports on liquid crystal dimers with branched terminal alkyl chains is by (4-alkylphenyliminomethyl)phenoxy]alkanes are well characterised12,13 and are of particular interest as they form Shiraishi and co-workers.7 They synthesized the dimer a,vbis{ 4-[(S)-2-methylbutylcinnamoate]-aminobenzylidene-4¾- intercalated smectic phases (see the Results and Discussion section for a description of the structure of this phase).To carbonyloxy}icosane† which exhibits a ferroelectric smectic C phase.gauge the effect of branching in the terminal alkyl chain more fully for the non-symmetric dimers, the remaining homologues Liquid crystal dimers serve as useful models for semi-flexible, main chain liquid crystal polymers. Like polymers, it is well- of the structurally isomeric a-(4¾-cyanobiphenyl-4-yloxy)-v-[4- (4-pentylphenyliminomethyl)phenoxy]alkanes were synthe- known that the order within the liquid crystal phases exhibited by dimers depends greatly on the number of methylene units sized as only the homologues with propane to hexane spacers in the spacer.8 Thus, as a result of their greater molecular have been described.12,13 anisotropy, the liquid crystal phases of even-membered dimers For convenience, we shall denote the a,v-bis{4-[4-(2-methylfrequently possess higher clearing temperatures and entropies butyl)phenyliminomethyl]phenoxy}alkanes and the a-(4¾- of transition than those of neighbouring odd-membered dimers.cyanobiphenyl-4-yloxy)-v-{4-[4-(2-methylbutyl)phenylimino- The observed odd–even effect results principally from the methyl]phenoxy}alkanes by the mnemonics 2MB.OnO.2MB difference in the distribution of conformers available to odd- and CBOnO.2MB, respectively.Here, CB denotes the cyanobiphenyl group, 2MB the 2-methylbutyl group, and n is the number of methylene units in the flexible alkyl spacer. The †The structure of the icosane is (CH2 )20[COOC6H4CH=NC6H4CH= CHCOOCH2CH(CH3)C2H5]2. symmetric straight chain dimers are denoted by 5.OnO.5, J.Mater. Chem., 1997, 7(1), 9–17 9where 5 indicates the pentyl chain. The straight chain analogues a function of temperature using a TMS90 Linkam hot stage to heat the sample. The hot stage was placed on a purpose- of CBOnO.2MB, the a-(4¾-cyanobiphenyl-4-yloxy)-v-[4-(4- pentylphenyliminomethyl)phenoxy]alkanes, are denoted by built housing which allowed the heated sample to be positioned in the beam of a Philips PU8730 UV–VIS spectrometer CBOnO.5.The chiral analogues of the 2MB.OnO.2MB and CBOnO.2MB series are referred to as (S)2MB.OnO.(S)2MB interfaced to an Opus PC II computer. The Grandjean planar texture used in the measurement was obtained by shearing the and CBOnO.(S)2MB, respectively, where S is the absolute configuration of the 2-methylbutyl substituent.sample between two glass slides. The helical twisting power was also determined for several chiral materials by dissolving the chiral compound in a room temperature nematic host Experimental (Merck E7, TCN=-10°C and TNI=61 °C) and measuring the pitch of the induced chiral nematic phase using the Cano- The synthetic steps taken to the racemic and chiral 2MB.OnO.2MB and CBOnO.2MB series are shown in wedge method.16 Two solute concentrations, approximately 5 and 11 mass%, were used and the inverse of the pitch was Schemes 1 and 2, respectively, and in this section we give the experimental details for the syntheses of (S)2MB.O6O.(S)2MB then extrapolated to give the helical twisting power.and CBO6O.(S)2MB as examples. The syntheses of (S)-(+)-2- methylbutylbenzene9 and the a-(4&fr34;-cyanobiphenyl-4-yloxy)-v- (S,S)-(+)-a,v-Bis{4-[ 4-(2-methylbutyl )phenyliminomethyl] (4-formylphenoxy)alkanes12,13 where carried out using litera- phenoxy}hexane ture procedures.(S)-(+)-4-(2-Methylbutyl)aniline ([a]23 -5.8, (S)-(+)-4-(2-Methylbutyl)aniline (0.5 g, 3.07 mmol) was added neat) was synthesized using the methods outlined in ref. 14. to a hot solution of a,v-bis(4-formylphenoxy)hexane (0.5 g, All intermediates were structurally characterized by IR (Perkin 1.46 mmol) with a few crystals of toluene-p-sulfonic acid in Elmer 1600 series FTIR spectrometer) and 1H NMR spec- absolute ethanol (30 ml) in a 50 ml conical flask fitted with a troscopy (JEOL FX90Q NMR spectrometer). The purity of calcium chloride guard tube.While cooling to room tempera- the final products was controlled by high field 1H NMR ture, the reaction mixture was stirred for 5 h. A creamy yellow spectroscopy (Bruker AM 360MHz NMR spectrometer) and precipitate formed which was filtered off and recrystallized mass spectroscopy (VG 70-250). Specific rotations were meas- twice from ethyl acetate to give white crystals which were ured with an AA-100 polarimeter (Optical Activity Ltd.). The dried in vacuo at 30°C (0.61 g, 68%).The yields of the other liquid crystal properties were investigated using an Olympus homologues were all in excess of 65%; nmax(film)/cm-1 1624 BH-2 polarizing microscope together with a TMS90 Linkam (CNN); 1H NMR (CDCl3): d 0.8–1.0 (t, 4H), 1.1–1.9 (m, 22H), hot stage. The thermal characteristics of the liquid crystal 2.1–2.8 (2q, 4H), 3.9–4.1 (t, 4H), 6.7 (d, 4H), 6.9 (s, 8H), 7.6 transitions were determined by differential scanning calor- (d, 4H), 8.1 (s, 2H); m/z (EI) 469 (M+-C5H11Ph), 616 (M+); imetry (DSC) using a Perkin Elmer DSC7 differential scanning [a]20+17.5 (c 1.04, CHCl3).calorimeter. The layer spacings of the smectic phases were determined by X-ray diffraction: a Guinier diffraction camera (S)-a-(4¾-Cyanobiphenyl-4-yloxy)-v-{4-[ 4-(2-methylbutyl ) was used with nickel-filtered Cu-Ka radiation (l=0.154 nm) phenyliminomethyl]phenoxy}hexane and a magnetic field (1.1 T) was used to align the samples.The wavelength of selective reflection, which is proportional (S)-(+)-4-(2-Methylbutyl)aniline (0.5 g, 3.1 mmol) was added to the helical pitch of pure chiral nematics, was measured as to a stirred solution of a-(4¾-cyanobiphenyl-4-yloxy)-v-(4-formylphenoxy) hexane (1.3 g, 3.2 mmol) with a few crystals of toluene-p-sulfonic acid in hot absolute ethanol in a 50 ml conical flask fitted with a calcium chloride guard tube.The reaction mixture was stirred for 5 h while cooling. A creamy yellow precipitate formed which was filtered off and recrystallized twice from ethyl acetate to give white crystals which were dried in vacuo at 30°C (1.1 g, 66%).The yields of the other homologues were all in excess of 65%; nmax(film)/cm-1 1602 (CNN), 2248 (CON); 1H NMR (CDCl3): d 0.8–1.2 (t, 2H), 1.2–2.0 (m, 15H), 2.3–2.7 (2q, 2H), 3.9–4.2 (m, 4H), 6.8–7.4 (m, 8H), 7.5–8.0 (m, 8H), 8.4 (s, 1H); m/z (EI) 487 (M+-C4H9), 544 (M+); [a]20+8.5 (c 0.9, CHCl3 ).Results and Discussion Scheme 1 The synthetic route to the 2MB.OnO.2MB series of The racemic dimers compounds The 2MB.OnO.2MB series: the symmetric dimers. The transition temperatures, enthalpies and entropies of transition for the 2MB.OnO.2MB series are listed in Table 1. From these results we see that this series exhibits both smectic A and nematic phases.The smectic A phase was characterised by its optical texture which showed the coexistence of focal-conic fan and homeotropic textures. The nematic phase was recognised by its schlieren texture which had both two and four point singularities and flashed when subjected to mechanical stress. As expected for this series, linking two mesogenic groups together by a flexible spacer led to a reduction in the smectic polymorphism compared to the analogous monomers.Fig. 1(a) shows the effect of varying the number of methylene units n in the flexible spacer on the liquid crystal behaviour for this series. We see that there is a strong odd–even effect for the Scheme 2 The synthetic route to the CBOnO.2MB series of compounds liquid crystal-to-isotropic transition temperature which is 10 J.Mater. Chem., 1997, 7(1), 9–17Table 1 Transition temperatures, enthalpies and entropies of transition for thin samples. The remaining homologues are all nematog- for the racemic 2MB.OnO.2MB series. Parentheses indicate a mono- enic. An X-ray investigation of the smectic A phase of the n= tropic transition 4 homologue gave a small angle reflection which corresponds to a smectic periodicity of 33.7 A° which is relatively close to T /°C DH/kJ mol-1 DS/R the all-trans molecular length of 38 A° as measured from a n C– SA–N N–I C– N–I C– N–I CPK molecular model.This result indicates that the molecules are arranged in monolayers and that, as expected, there is no 3 136 39.7 11.7 significant interpenetration of the molecules between the layers. 4 156 183a 32.2 9.9a 9.1 2.6a The effect of the branched terminal alkyl chains on the 5 124 (91)b 35.3 10.7 liquid crystalline behaviour is clearly highlighted by a compari- 6 154 (138)c (151) 42.6 4.5 12.2 1.3 son of the phase behaviour for the structurally isomeric 7 147 (96)b 37.0 11.4 8 132 133 38.6 4.0 11.7 1.2 5.OnO.5 series with no methyl branches in their terminal 9 108 (94)b 38.5 12.2 pentyl chains [see Fig. 1(b)].11 We see that the 5.OnO.5 series 10 121 120 52.3 4.6 16.2 1.4 has higher clearing temperatures and exhibits a greater smectic 11 110 (95)b 42.0 13.2 polymorphism than their branched analogues. The smaller 12 119 (114) 65.3 20.2 values for the branched compounds may be due to the methyl branch in the alkyl chain, which tends to reduce the structural aSA–I transition.bObserved only with polarising microscopy by supercooling isotropic droplets. cThe phase transition could only be anisotropy and, in addition, the lateral interactions responsible observed by polarising microscopy. for smectic phase stabilisation.17 Furthermore, the effect is known to be greatest for methyl branches close to the mesogenic core.17 Normally the insertion of a methyl branch also leads to a reduction in the melting point of the material and yet strangely for this series the reverse is the case.This result is hard to explain and presumably reflects a greater packing ability in the crystal phase for the branched dimers compared to their unbranched analogues. From a comparison of Figs. 1(a) and (b) we also notice that the magnitude of the odd–even effect for the clearing temperatures is slightly greater for the branched dimers. The contrasting liquid crystalline behaviour for the two series is further highlighted in Fig. 2, which shows the difference in the transition temperatures [T (2MB.OnO.2MB)-T (5.OnO.5)] versus the number of methylene units n in the flexible alkyl spacer.The results show an odd–even effect for the differences in the clearing temperatures which diminishes across the series and also emphasises the considerably higher melting points for the branched dimers. The attenuation of the differences in the clearing temperatures across the series appears to reflect the diminishing effect of the methyl branch on the length-to-breadth ratio, that is the shape anisotropy, as the spacer is lengthened. The entropy changes at the clearing point could only be measured for the n=4,6,8 and 10 homologues of the 2MB.OnO.2MB series and they are all lower than their unbranched analogues11 (see Fig. 3). This is strange but not unknown, as lateral substitution has also been shown to have an effect on the clearing entropy change.For example, the a,v-bis[4-(2,4-dimethylphenyliminomethyl)- phenoxy]alkanes have been found to exhibit a reduced Fig. 1 The influence of the number of methylene groups n in the flexible alkyl spacer on the transition temperatures for (a) the 2MB. OnO.2MB and (b) the 5.OnO.5 series. C–(#), SA–I (+), SA–N ('), N–I ($), SB–SA (2). attenuated on ascending the series.It is also evident that the liquid crystal phases occur predominantly below the melting point; indeed only the n=4 and 8 homologues exhibit enantiotropic phases. The nematic-to-isotropic transition temperatures of homologues with an odd numbered spacer could only Fig. 2 The difference [T (2MB.OnO.2MB)-T (5.OnO.5)] between the be observed with polarizing microscopy by supercooling iso- melting (#) and clearing ($) temperatures for the 2MB.OnO.2MB lated, isotropic droplets. The n=4 and 6 homologues form and 5.OnO.5 series as a function of the number of methylene groups n in the spacer smectic A phases which show strong homeotropic alignment J.Mater. Chem., 1997, 7(1), 9–17 11Fig. 3 The influence of the number of methylene groups n in the flexible alkyl spacer on the clearing entropies of transition for the n= 4,6,8 and 10 members of the 2MB.OnO.2MB series [SA–I (+), N–I ($)] and the 5.OnO.5 series [SA–I ('), N–I (#)] nematic-to-isotropic entropy change, which was attributed to either an increase in the molecular biaxiality or a reduction in the conjugation of the mesogenic core.18 From Fig. 3 we see that the differences between the entropy changes for the unbranched and branched compounds are constant for both the nematic-to-isotropic and smectic A-to-isotropic transitions.It would appear that the difference in the average conformation of both odd- and even-membered dimers in the isotropic and nematic phases is somewhat less pronounced for the branched dimers compared to their unbranched counterparts.The lower clearing entropy changes for the branched dimers might also be a consequence of the increased molecular biaxiality, resulting from a decreased length-to-breadth ratio relative to their unbranched analogues. Fig. 4 The influence of the number of methylene groups n in the flexible alkyl spacer on the transition temperatures for (a) the The CBOnO.2MB series: the non-symmetric dimers.Table 2 CBOnO.2MB and (b) the CBOnO.5 series. C–(#), SA–I (+), SA–N gives the transition temperatures, enthalpies and entropies of ('), N–I ($), Sl–SA (+), SB–SA (2), Sl–N (x). transition for the CBOnO.2MB series. Smectic A and nematic phases are exhibited by the members of the series and were characterised, as before, by polarizing microscopy. All but one for the n=12 homologue.The improved stability of the smectic A phase as the spacer is lengthened, which is not of the homologues are enantiotropic mesogens; the n=3 member exhibits a nematic phase just 12°C below the melting observed for symmetric dimers, is well-known for non-symmetric dimers of this type.12,13 X-Ray diffraction studies of point. Fig. 4(a) shows the effect of changing the number of methylene groups in the spacer on the liquid crystalline aligned smectic A phases of the n=8 and 11 homologues and a powder sample of the n=12 homologue all gave smectic properties of this series.Both the smectic A-to-nematic and nematic-to-isotropic transitions show the anticipated odd–even periodicities equal to approximately half their all-trans molecular lengths as measured from CPK molecular models (see effect across the series. The n=6–12 homologues all form a smectic A phase, the stability of which increases as the series Table 3).These results indicate that the smectic A phases have an intercalated structure (see Fig. 5) and are comparable to is ascended such that it clears directly into the isotropic phase Table 2 Transition temperatures, enthalpies and entropies of transition for the racemic CBOnO.2MB series.Parentheses indicate a monotropic transition T /°C DH/kJ mol-1 DS/R n C– SA–N N–I C– SA–N N–I C– SA–N N–I 3 92 (80) 25.0 (0.3) 8.3 (0.1) 4 148 200 32.0 4.0 9.2 1.0 5 104 121 31.2 0.5 10.0 0.1 6 137 (102) 178 29.3 (0.4) 4.2 8.6 (0.1) 1.1 7 97 (70) 130 48.0 (3.9) 0.6 16.0 (1.4) 0.2 8 124 140 163 32.4 2.2 5.0 9.8 0.6 1.4 9 109 (100) 133 40.0 (3.7) 1.3 13.0 (1.2) 0.4 10 114 149 151 41.4 3.5 2.9 13.0 1.0 0.8 11 84 116 127 28.4 2.8 2.4 9.6 0.9 0.7 12 112 145a 48.4 14.0a 15.0 4.0a aSA–I transition. 12 J. Mater. Chem., 1997, 7(1), 9–17Table 3 Smectic periodicity d of some members of the CBOnO.2MB and CBOnO.5 series measured by X-ray diffraction. The values are compared to their all-trans molecular lengths l as estimated from CPK molecular models Compound l/A° d/A° d/l CBO8O.2MB 39.1 18.9 0.48 CBO11O.2MB 41.4 21.4 0.52 CBO12O.2MB 44.0 21.6 0.49 CBO6O.5a 38.0 19.5 0.51 CBO9O.5 40.2 20.1 0.50 CBO12O.5 45.0 21.9 0.49 CBO12O.5b 45.0 24.8 0.55 aTaken from ref. 12. bSmectic B phase. Fig. 6 The difference [T (CBOnO.2MB)-T (CBOnO.5)] between the melting (#) and clearing ($) temperatures for the CBOnO.2MB and CBOnO.5 series as a function of the number of methylene groups n in the spacer interactions of the CBOnO.5 series compared to their branched analogues.The monotropic smectic phases exhibited by the n=3,5 and 6 homologues could not be identified as they were only observed fleetingly on rapidly cooling the isotropic liquid prior to crystallization. The periodicities of the smectic A phases for the n=6 and 9 homologues of the CBOnO.5 series were determined by X-ray diffraction and these phases were found to be intercalated as was the smectic B phase of the n=12 homologue.Fig. 6 shows the difference between the melting and clearing temperatures for the two series Fig. 5 A schematic representation of the proposed molecular organis- [T (CBOnO.2MB)-T(CBOnO.5)] as a function of the number ation in the intercalated smectic A phase (ref. 13) of methylene units n in the spacer. Apart from the n=3 homologue, we see that the melting points of the CBOnO.5 the measured smectic periodicities of similar compounds pre- series are lower than those of the CBOnO.2MB series; a similar viously reported.12,13 The proposed structure of the intercalated difference was observed for the 2MB.OnO.2MB and 5.OnO.5 smectic phase shown in Fig. 5 is thought to result from a series. Again the higher melting points for the branched dimers specific interaction between the unlike mesogenic cores19 are hard to explain and we can only speculate that perhaps enhanced by an entropy gain resulting from the uniform mixing the methyl branch helps fill space more efficiently in the crystal of such groups.20 In addition, the stability of these phases is phase than for dimers without branches in their terminal alkyl expected to be influenced by the amount of space between the chains. The difference in the clearing temperatures exhibits an mesogenic groups available to accommodate the terminal alkyl odd–even effect which is attenuated across the series and again chain; an intercalated smectic phase is likely if the terminal reflects the diminishing effect of the methyl branch on the alkyl chain can be accommodated in this space.Thus the length-to-breadth ratio as the spacer length grows. The magni- increase in thermal stability of the intercalated smectic A phase tude of the odd–even effect for the clearing temperatures is for the CBOnO.2MB series as the spacer length grows can be again slightly greater for the branched dimers.Fig. 7 shows attributed to the increasing amount of space available to the the transitional entropies as the number of methylene units n 2-methylbutyl chain. However, a smectic A phase is not in the spacer is varied for the CBOnO.2MB and CBOnO.5 observed for the n=3–5 homologues as the space available to series.We see that for both series the smectic A-to-nematic the 2-methylbutyl group is presumably insufficient. We should and the nematic-to-isotropic transitional entropies exhibit an note that another somewhat more exotic model has recently odd–even effect. The nematic-to-isotropic entropy changes are been proposed for the rationalisation of the properties of the all smaller for the CBOnO.2MB series, with some of the odd- smectic A phase of non-symmetric mesogens; in this the membered homologues possessing particularly small entropies molecules are depicted as adopting a hairpin or horseshoe of transition, which suggest incipient biaxial nematic behav- conformation.20 Thus a thorough and definitive understanding iour.21 Interestingly, the odd–even effect for the smectic A-to- of the structure of the intercalated smectic A phase remains nematic transitional entropies of the CBOnO.2MB series has unavailable.an opposite sense to that of the CBOnO.5 series. Indeed, the The effect of the methyl branch on the transition tempera- entropy changes for the smectic A-to-nematic transitions do tures of the CBOnO.2MB series is clearly seen if we compare not agree with those expected from the ratio of the smectic A- them with those of their structural analogues, the CBOnO.5 to-nematic and nematic-to-isotropic transition temperatures.22 series.Fig. 4(b) shows the dependence of the transition tempera- Thus, when the difference between the smectic A-to-nematic tures of the CBOnO.5 compounds on the number of methylene and nematic-to-isotropic transition temperatures is large it units in the spacer. We see that not only do the CBOnO.5 does not always follow that the entropy change for the species have higher clearing temperatures but, in addition to smectic A-to-nematic transition is small, as the heptane homo- smectic A and nematic phases, they also exhibit smectic B logue shows (see Table 4).Such behaviour is hard to rationalise phases, as indicated by the smoothness of the focal-conic fans and might be a consequence of the restricted conformational in their optical textures. As mentioned previously the greater smectic polymorphism is consistent with the stronger lateral freedom of the 2-methylbutyl group. J.Mater. Chem., 1997, 7(1), 9–17 13bright threads appear in the homeotropic regions of the smectic A phase which thicken as the temperature is increased. The resulting web-like structure then coalesces into the Grandjean texture. Similar effects have been observed for the chiral, monomeric analogues.9 The sense of the chiral helix was determined to be dextrorotatory by contact preparations with the laevorotatory compound cholesteryl benzoate (C 150°C N* 178 °C I, Eastman Kodak Co., Rochester, USA) which showed a nematic phase at the junction of the two components.23 This result agrees with the Gray and McDonnell rules,24 which state that for an S absolute configuration the sense of the chirality should be dextrorotatory when the chiral centre is an even number of atoms away from the mesogenic group, and laevorotatory when it is an odd number of atoms away.As these rules were devised principally for monomeric liquid crystals, we have overcome the problem of counting the number of atoms between the chiral centre and the mesogenic group in dimers by simply splitting the dimer into its constituent monomeric parts. Since the chiral nematic phase of 2MB.O6O.2MB is monotropic and that of 2MB.O8O.2MB has a very short range, we have not determined the wavelength of selective reflection for the pure materials. However, we have obtained an indication of the form chirality of these two chiral symmetric dimers by measuring their helical twisting powers bM.25 This is achieved by plotting the inverse of the pitch p of the induced chiral nematic phase as a function of the solute concentration.According to theory,26 p-1 should be linear in the solute concentration which is best expressed as the mole fraction x,27 although claims to have found a better linearity with the weight fraction have been made.28 However, for the low concentrations used in most measurements, both methods of expressing the solute concentration give linear plots and so we shall use the mole fraction [eqn.(1)]. bM=p-1/x (1) The mole fractions for the solutions were calculated using an average molar mass for the nematic solvent E7 of 271 g since this is a mixture of cyanobiphenyls. The helical twisting powers obtained in this way for the hexane and octane dimers are given in Table 5. We see that they are the same (12.5 mm-1) which is not so surprising Fig. 7 The influence of the number of methylene groups n in the because theory suggests29 that the helical twisting power is flexible alkyl spacer on the entropies of transition for (a) the proportional to the scalar product of the ordering shape tensor CBOnO.2MB and (b) the CBOnO.5 series. SA–N ('), N–I ($), SA–I (+), SB–SA (2), Sl–N (x).of the chiral solute and the chiral shape tensor. Since the two dimers have a similar structure and their nematic-to-isotropic transitional entropies are the same (see Table 1), it might be Table 4 Comparison of the ratio of the smectic A-to-nematic to expected that their orientational order in a common solvent nematic-to-isotropic transition temperatures versus the smectic A-tonematic entropy of transition for the CBOnO.2MB series would be similar.What is surprising is that the helical twisting powers of monomers with the same chiral 2-methylbutyl group n TSA–N/TN–I DSSA–N/R in the same nematic solvent is essentially half that for the two dimers.30 This is unexpected because the orientational order 6 0.831 0.1 of the monomer is much less than that of an even-membered 7 0.851 1.4 dimer in the same nematic solvent.31 As a consequence, after 8 0.947 0.6 9 0.915 1.2 allowing for the presence of two chiral centres in the dimers, 10 0.995 1.0 their bM values should be greater than twice that for a 11 0.973 0.9 monomer.We shall return to this point when we consider the helical twisting powers for the chiral non-symmetric dimers in the following section.The chiral dimers The (S)2MB.OnO.(S)2MB series: the symmetric dimers. The The CBOnO.(S)2MB series: the non-symmetric dimers. Table 6 presents the transition temperatures, enthalpies and transition temperatures, enthalpies and entropies for the n=6 and 8 homologues of the (S)2MB.OnO.(S)2MB series are given entropies of transition for the phase behaviour of the n=6 to 9 homologues of the CBOnO.(S)2MB series.The transition in Table 5. We see that, as expected, the data are similar to those for the racemic analogues. On cooling the n=6 and 8 temperatures as well as the clearing entropy changes agree closely with those of the analogous racemic compounds, as homologues, a focal-conic fan texture is obtained which, when sheared, gave the Grandjean texture which is indicative of a expected.On slow cooling of the isotropic liquid for the n=6 and 8 homologues the focal-conic fan texture of the chiral chiral nematic phase. On cooling the chiral nematic phase of the n=6 homologue a change to a focal-conic fan coexistent nematic phase forms which, when sheared, showed the oily streaks characteristic of the Grandjean plane texture.In con- with a homeotropic texture occurred which was characterized as that of a smectic A phase. On slow heating (0.2 °C min-1) trast, the n=7 and 9 homologues show different behaviour. 14 J. Mater. Chem., 1997, 7(1), 9–17Table 5 Transition temperatures, enthalpies and entropies of transition for the chiral (S)2MB.OnO.(S)2MB series. Parentheses indicate a monotropic transition T /°C DH/kJ mol-1 DS/R n C– SA–N* N*–I C– SA–N* N*–I C– SA–N* N*–I bM/mm-1 6 149 (138) (148) 49.3 0.8 4.6 14.1 0.2 1.3 12.5 8 127 131 42.0 4.3 12.6 1.3 12.5 Table 6 Transition temperatures, enthalpies and entropies of transition for the chiral CBOnO.(S)2MB series.Parentheses indicate a monotropic transition T /°C DH/kJ mol-1 DS/R n C– SA–N* N*–BPI BPI–I C– SA–N* N*–I C– SA–N* N*–I bM/mm-1 6 137 (104) 176a 32.0 0.7 4.5 9.4 0.2 1.2 5.6 7 97 (74) 131.5b 131.9b 45.8 4.3 0.7 15.0 1.5 0.2 5.9 8 128 143 165a 36.3 1.9 5.1 11.0 0.6 1.5 6.7 9 110 (102) 132.2b 132.8b 45.2 3.8 1.3 14.0 1.2 0.4 c aN*–I transition. bTransition could only be observed by polarising microscopy.cValue was not measured. On slow cooling (0.2 °C min-1) the isotropic liquid is seen to exhibit a homeotropic blue colour of low birefringence before a platelet texture composed of blue and red plates quickly grows; crinkled lines between the plates are clearly observed and this is indicative of blue phase I (Plate 1).The transition to the blue phase could not be detected by DSC. If the blue phase is submitted to mechanical stress the Grandjean plane texture forms.The thermal range of blue phase I is 0.4 and 0.6 °C for the n=7 and 9 homologues, respectively. At the smectic A-to-chiral nematic transition ba�tonnets form which coalesce into the smectic A focal-conic fan texture. There was no indication from the optical texture that the smectic A phase was of the twist grain boundary (TGB) type.32 Indeed, the intercalated nature of the smectic A phase might well inhibit the layers from twisting into a helix.The sense of the helix in the chiral nematic phase was determined to be dextrorotatory for all four compounds by contact preparations with cholesteryl Fig. 8 The temperature dependence of the wavelength lo of maximum reflection for the chiral non-symmetric dimers CBOnO.2MB with n= benzoate and again this agrees with the Gray and McDonnell 6,7 and 8.The nature of the phases and the transitions between them rules. The helical pitch p can be determined from the central are also shown. wavelength lo of the peak in the UV–VIS spectrum obtained for normal incidence, but to do this it is necessary to know the spacer the results for the wavelengths show that the helical the mean refractive index of the material n since p=lo/n.33 As pitch for the odd dimer is significantly smaller than for the we do not have accurate values for the refractive indices of the two even dimers, even allowing for the pretransitional growth dimers which we have studied we shall discuss instead the in p for these compounds.Analogous results have been found wavelength of the maximum reflection.This has been measured for the non-symmetric dimers formed by linking a cyanobi- for the chiral non-symmetric dimers CBOnO.2MB with both phenyl group to either a cholesteryl34 or a dihydrocholesteryl odd and even spacers (n=6, 7 and 8) and the results are group.35 Such a difference in the pitch for odd and even chiral plotted as a function of temperature in Fig. 8. Since the mean dimers can be understood in terms of their elastic properties.refractive index is not expected to change significantly with It is to be expected that just as the pitch of an induced chiral nematic phase is proportional to the twist elastic constant of the nematic solvent,29 so the pitch of the pure chiral nematic would be proportional to its twist elastic constant. Although the elastic constants have not been determined for the dimers which we have studied, results are available for analogous systems.36 These show that the twist elastic constant for a dimer with an odd spacer is about half as large as that for a neighbouring dimer with an even spacer.This difference is in accord with the higher orientational order found for even dimers in comparison with their odd counterparts.8 We see, therefore, that the smaller pitch found for the odd dimer can be understood in terms of the lower twist elastic constant and orientational order found for the odd dimers relative to the even. This relationship would then suggest that the increase in the helical pitch with the length of the even spacer shown by our results is caused by a corresponding increase in the twist Plate 1 The blue phase I platelet texture of CBO7O.2MB on cooling elastic constant.However, we should note that this trend is from the isotropic phase (0.2°C min-1); T=131.7°C, magnification ×200 reversed for the cyanobiphenyl-cholesteryl dimers, where the J. Mater. Chem., 1997, 7(1), 9–17 15pitch appears to decrease with increasing length of the even minal alkyl chains have been presented.The branched chains spacer.34 This would, then, imply a decrease in the twist have a marked effect on the liquid crystal properties of the elastic constant. 2MB.OnO.2MB series, as they exhibited lower clearing tem- The small value of the pitch found for the chiral non- peratures and a reduced smectic polymorphism with respect symmetric dimer with an odd spacer is consistent with the to their unbranched analogues, the 5.OnO.5 series.Indeed, the observation of a blue phase for the two odd dimers, but not nematic-to-isotropic transition temperatures of some homolfor those with even spacers. Experimentally, the temperature ogues of the 2MB.OnO.2MB series were so far below the range over which the blue phase exists is found to decrease meltingpoint that they could only be measured by supercooling with an increase in the helical pitch.37 The critical value of the small, isolated isotropic droplets.In contrast, for the pitch at which the blue phase vanishes depends on the molecu- CBOnO.2MB series the inclusion of a branched terminal alkyl lar structure; for cholesteryl alkanoates it is about 300 nm, chain led to a less dramatic reduction in the clearing temperawhereas for mixtures of chiral and achiral materials it is tures and smectic polymorphism (as compared to their straight approximately 400 nm.If the mean refractive index is taken to chain analogues the CBOnO.5 series), presumably because be approximately 1.5 then the pitch for the odd dimer is about only one branched terminal alkyl chain is involved.X-Ray 300 nm, whereas for the even dimers p is about 430 nm. The diffraction studies of the n=8, 11 and 12 homologues of the critical pitch for the chiral non-symmetric dimers is between CBOnO.2MB series showed their smectic A phases to be these two values, in reasonable agreement with results found intercalated as expected for non-symmetric dimers of this type.for chiral monomers.37 In concluding our discussion of the Of the chiral compounds studied, those of the CBOnO.(S)2MB helical pitch, we should comment on its temperature depen- series possessed the most interesting properties. Four homol- dence. For the odd dimer (CBO7O.2MB) the pitch increases ogues were synthesized with the number of methylene units in slightly from the chiral nematic-to-isotropic transition, but the flexible alkyl spacer ranging from 6 to 9.An interesting then is essentially independent of temperature. It was not odd–even effect was observed for the chiral properties of these possible to make measurements near the monotropic SA–N* materials. Blue phase I behaviour was only observed for the transition because the sample crystallised.However, the rela- odd non-symmetric dimers and not their even spacer counter- tively high transitional entropy (see Table 6) suggests that the parts. This marked difference is consistent with the smaller pretransitional growth in the pitch would be weak. In contrast, pitch found for the odd dimer relative to the even dimers. This the entropy change at the smectic A-to-chiral nematic trans- can be related to the smaller twist elastic constant of the odd ition is relatively small for the even dimers (see Table 6) and dimers, which is related to their lower orientational order.In so the pretransitional growth in the pitch is expected to be contrast to this strong odd–even effect, the helical twisting strong. This is certainly the case for CBO8O.2MB, where it powers of the dimers seem to depend solely on the nature of was possible to make measurements very close to the transition the 2-methylbutyl chiral centre and not on the details of the to the smectic A phase (see Fig. 8). For the other even dimer, remaining molecular structure. CBO6O.2MB, the pitch is essentially constant over most of the wide chiral nematic range.However, it then starts to grow We are grateful to the former Science and Engineering Research as the monotropic transition to the smectic A phase is Council for the award of a research studentship to A. E. B. approached, but the sample crystallises before the pitch can and a CASE research studentship with GEC to I. D. F. We become very large. Such pretransitional growth in the pitch, would also like to thank Merck Ltd., Poole, UK, for allowing which is observed for the even dimers, is typical of that found A.E. B. and I. D. F. to carry out the Cano-wedge measurements for monomers when the smectic A-to-chiral nematic transition at their laboratories and to Dr M. J. Goulding for assisting is weak. them. For comparison with the results for thchiral symmetric dimers, we have also determined the helical twisting powers of the chiral non-symmetric dimers with hexane, heptane and References octane spacers.The results are given in Table 6 and we see 1 F. Reinitzer, Monatsh. Chem., 1888, 9, 421. that bM is about 6 mm-1 for these dimers, which is essentially 2 See for example, J. W. Goodby, I. Nishiyama, A. J. Slaney, half the value found for the analogous symmetric dimers with C.J. Booth and K. J. Toyne, L iq. Cryst., 1993, 14, 37 and references two 2-methylbutyl groups. It is also approximately the same cited therein. as for monomers with the same chiral centre.30 These results 3 S. Chandrasekhar, L iquid Crystals, 2nd edn., Cambridge also support the view that it is the nature of the chiral centre University Press, 1992, ch. 4 and 5. which seems to control the helical twisting power and not the 4 J. Barbera�, A. Omenat and J. L. Serrano, Mol. Cryst. L iq. Cryst., 1989, 166, 167. anisotropy of the group to which it is attached. Such an 5 J. Barbera�, A. Omenat, J. L. Serrano and T. Sierra, L iq. Cryst., explanation is at variance with theory,29 and an alternative 1989, 5, 1775. interpretation might be that as the anisotropy of the group 6 G.Heppke, D. Lo�tzsch and F. Oestreicher, Z.Naturforsch., T eil A: attached to the chiral centre grows, the increase in the orien- Phys. Sci., 1987, 42, 279. tational ordering tensor is compensated for by a decrease in 7 K. Shiraishi, K. Kato and K. Sugiyama, Chem. L ett., 1990, 971. the chiral ordering tensor. It remains to be seen whether such 8 J. W.Emsley, G. R. Luckhurst and G. N. Shilstone, Mol. Phys., 1984, 53, 1023. an explanation is correct. Although the bM values for the non- 9 D. Dolphin, Z. Muljiani, J. Cheng and R. B. Meyer, J. Chem. Phys., symmetric dimers are essentially constant, there is a slight 1973, 58, 413. increase with the length of the spacer. There is not, however, 10 Y. Y. Hsu and D.Dolphin, Mol. Cryst. L iq. Cryst., 1977, 42, 327. any indication of an odd–even effect for the helical twisting 11 R. W. Date, C. T. Imrie, G. R. Luckhurst and J. M. Seddon, L iq. power. This is in marked contrast with the behaviour of the Cryst., 1992, 12, 203. helical pitch, for which a strong dependence on the parity of 12 J. L. Hogan, C. T. Imrie and G. R. Luckhurst, L iq.Cryst., 1988, the spacer is found. Such an observation emphasises the fact 3, 645. 13 G. S. Attard, R. W. Date, C. T. Imrie, G. R. Luckhurst, that the helical twisting power can be a poor indicator of the S. J. Roskilly,J. M. Seddon and L. Taylor, L iq. Cryst., 1994, 16, 529. form chirality of the chiral nematic phase exhibited by the 14 P. Keller and L. Liebert, Solid State Phys., 1978, Suppl. 14, 19. pure chiral solute. 15 See for example, H. Kelker and R. Hatz, Handbook of L iquid Crystals, Verlag Chemie, 1980, ch. 7 and references cited therein. 16 R. Cano, Bull. Soc. Fr.Mineral. Cristallogr., 1968, 91, 20. Conclusions 17 See for example, G. W. Gray, L iquid Crystals and Plastic Crystals, The structure–property relationships of two series of dimers, eds. G. W. Gray and P. A. Winsor, Ellis Horwood, 1974, vol. 1, ch. 4. one symmetric, the other non-symmetric, with branched ter- 16 J. Mater. Chem., 1997, 7(1), 9–1718 C. T. Imrie, L iq. Cryst., 1989, 6, 391. 30 M. J. Goulding, PhD Thesis, 1995, University of Southampton. 31 P. J. Le Masurier, G. R. Luckhurst, S. J. Roskilly and G. Saielli, 19 J. W. Park, C. S. Bak and M. M. Labes, J. Am. Chem. Soc., 1975, 97, 4398. unpublished data. 32 J. W. Goodby, M. A. Waugh, S. M. Stein, E. Chin, R. Pindak and 20 A. E. Blatch, I. D. Fletcher and G. R. Luckhurst, L iq. Cryst., 1995, 18, 801. J. S. Patel, J. Am. Chem. Soc., 1989, 111, 8119. 33 See for example, N. H. Hartshorne, L iquid Crystals and Plastic 21 M. J. Freiser, Phys. Rev. L ett., 1970, 24, 1041. 22 W. J. McMillan, Phys. Rev. A, 1972, 6, 936. Crystals, eds. G. W. Gray and P. A. Winsor, Ellis Horwood, 1974, vol. 2, ch. 2. 23 See for example, H. Kelker and R. Hatz, Handbook of L iquid Crystals, Verlag Chemie, 1980, ch. 1 and references cited therein. 34 A. T. M. Marcelis, A. Koudijs and E. J. R. Sudho�lter, Rec. T rav. Chim. Pays-Bas, 1994, 113, 524. 24 G. W. Gray and D. G. McDonnell, Electron. L ett., 1975, 11, 556. 25 See for example, G. Solladie and R. Zimmermann, Angew. Chem., 35 A. T. M. Marcelis, A. Koudijs and E. J. R. Sudho�lter, L iq. Cryst., 1995, 18, 843. Int. Ed. Engl., 1984, 23, 328. 26 W. J. A. Goossens,Mol. Cryst. L iq. Cryst., 1971, 12, 237. 36 G. A. DiLisi, E. M. Terentjev, A. C. Griffin and C. Rosenblatt, J. Phys. II, 1993, 3, 597. 27 J. M. Ruxer, G. Solladie and S. Candau, Mol. Cryst. L iq. Cryst. L ett. Sect., 1978, 41, 109. 37 H. Stegemeyer, T. Blumel, K. Hiltrop, H. Onusseit and F. Porsch, L iq. Cryst., 1986, 1, 1. 28 M. Tsukamoto, T. Ohtsuka, K. Morimoto and Y. Murakami, Jpn. J. Appl. Phys., 1975, 14, 1307. 29 A. Ferrarini, G. J. Moro and P. L. Nordio, L iq. Cryst., 1995, 19, 397. Paper 6/02980I; Received 29th April, 1996 J. Mater. Chem., 1997, 7(1), 9&n

ISSN:0959-9428    

DOI:10.1039/a602980i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Effects of isocyanide substituents on the mesogenic properties of halogeno(isocyanide)gold complexes: calamitic and discotic liquid crystals Silverio Coco,a Pablo Espinet,*a Jose� M. Martý�n-Alvareza and Anne-Marie Levelutb aDepartamento de Quý�mica Inorga� nica, Facultad de Ciencias, Universidad de Valladolid, E-47005 Valladolid, Spain bL aboratoire de Physique des Solides, Ba�timent 510, Universite� Paris-Sud, 91405 Orsay Ce�dex, France The liquid crystal behaviour of linear [AuX(CNC6H4OCnH2n+1)] (X=halogen) complexes is tuned by systematically changing the structure of the isocyanide ligand.The changes studied are lateral fluorination at the 2- and 3-positions of the phenyl ring and modification of the rod-like structure of the complexes by addition of two alkoxy chains at the meta positions.None of the free isocyanides used here are liquid crystals, but all the gold complexes prepared display mesomorphic properties, except iodo(2- fluoro-4-butoxyphenyl isocyanide)gold(I). The (2- and (3-fluorophenyl isocyanide)gold(I) complexes show smectic A mesophases, except the 2-fluoro derivative with n=6, which shows a nematic phase, while the (3,4,5-trialkoxyphenyl isocyanide)gold compounds display columnar hexagonal phases at room temperature.In recent years there has been increasing interest in liquid sponding RNO2 compounds in three steps, involving reduction crystals based on molecules containing transition metals (so- to the amine RNH2 by hydrazine–graphite or SnCl2, forcalled metallomesogens).1 Thermotropic liquid crystalline mylation to the formamide RNHCHO, and finally dehydration phases of calamitic molecules have been extensively studied, with bis(trichloromethyl)carbonate (‘triphosgene’) and triethyland many systematic studies on the influence of molecular amine to the isocyanide RNC as presented in Scheme 1.constitution on mesomorphism have been reported on conven- None of the free ligands show liquid crystal behaviour.At tional organic liquid crystals. However, this kind of study of room temperature, the fluorinated isocyanides are isotropic metallomesogens is rare.1 The continued development of liquid crystal-based technologies requires new types of mesomorphic materials. Therefore it is important to understand the influence of molecular structure on thermal behaviour in order to design and prepare new liquid-crystalline materials.Most metallomesogens consist of coordinated compounds derived from a metal of d8–d10 electron configuration, usually with planar or linear geometries. However, it is surprising that only a few examples of gold mesogens have been reported, namely a family of gold(III ) alkoxydithiobenzoate complexes,2 and some gold(I) derivatives containing styrylpyridine (stilbazole), 3 isocyanide4–6 and carbene ligands.7 We have reported previously a family of gold complexes6 [AuX(CNC6H4OCnH2n+1)] (X=halogen; n=2, 4, 6, 8, 10, 12) which show smectic A phases in spite of the fact that the isocyanide ligands are not mesogenic, and contain only one aryl ring.These compounds show high thermal stability and have a very simple structure (linear coordination for gold), which makes them particularly suitable for studying the effect of modifications in the molecule on the liquid crystal behaviour of the material.Thus, we prepared similar mesogenic halogeno- (biphenyl isocyanide)gold(I) complexes and studied the influence of the biphenyl moiety on transition temperatures.5 In a continuation of our previous studies of these coordinatively simple gold(I) isocyanide complexes, aimed at understanding the influence of molecular structure on liquid-crystalline behaviour, we now report the effect of two changes on the p-alkoxyphenyl isocyanide system, namely (i) the introduction of a lateral substituent on the phenyl ring by fluorination at the 2- and 3-positions, and (ii) the modifi- cation of its rod-like structure by addition of two alkoxy chains in the meta positions.The halogeno(3,4,5-trialkoxyphenyl isocyanide)gold(I) complexes prepared show room temperature hexagonal columnar mesophases. Results and Discussion Syntheses. The isocyanides used in this work have not been Scheme 1 Reagents: i, N2H4–graphite or SnCl2; ii, HCO2H; iii, (Cl3CO)2CO, Et3N reported before and were prepared starting from the corre- J.Mater. Chem., 1997, 7(1), 19–23 19liquids, while the trialkoxyphenyl species are solids with melting points in the range 25 to 42°C, except for the hexyloxy derivative which melts at 16°C. They can be stored for long periods in the freezer. The gold(I) compounds were easily prepared as shown in Scheme 2.The reactions of the isocyanides with [AuCl(tht)] (tht=tetrahydrothiophene) in CH2Cl2 afford white complexes [AuCl(CNR)] [R=2-F-4-OCnH2n+1C6H3, 3-F-4-OCnH2n+1- C6H3 , n=6, 8, 10, 12; R=3,4,5-(OCnH2n+1)3C6H2, n=4, 6, 8, 10], and exchange reactions in acetone with the appropriate potassium salt give the corresponding white bromo and iodo derivatives [AuX(CNR)] [X=Br, I; R=2-F-4-OC12H25C6H3; 3-F-4-OC12H25C6H3; R=3,4,5-(OC10H21)3C6H2].Fig. 1 Thermal behaviour of the complexes [AuCl(CNC6- The elemental analyses for the complexes, the yields and H4OCnH2n+1-4)] (no F), [AuCl(CNC6H3F-2-OCnH2n+1-4)] (F-2) and relevant IR data are given in the Experimental section. The [AuCl(CNC6H3F-3-OCnH2n+1-4)] (F-3) IR spectra show one n(CON) absorption in CH2Cl2 solution, in each case at higher wavenumbers (ca. 100 cm-1) than for the free isocyanide. This shift is well documented and is due to substitution of Cl by Br and I in the 3-F derivatives produces two factors; the s-donation of the antibonding carbon lone adecrease in the transition temperatures in the order Cl>Br>I pair to Au, and the p-backbonding from the AuI 5d orbital to (Fig. 2), according to the decrease in polarity of the Au–X the p* ligand orbitals.8 In Nujol mulls the IR spectra are bond, as reported for similar halogenogold isocyanide comsimilar but the 3-fluorinated gold derivatives show a splitting plexes.6 However, the melting point of the monotropic bromo of the n(CON) isocyanide band, possibly due to solid state effects. 2-F derivative is slightly lower than that of the non-meso- The 1H NMR spectra of the complexes containing fluori- morphic iodo 2-F complex.The transition temperatures of the nated isocyanides are all very similar (see Experimental section corresponding non-fluorinated derivatives are intermediate for details). At 300 MHz the hydrogen atoms of the aromatic between the 3-F and 2-F fluorinated complexes. ring give three resonances as expected for an AMNX spin Although some care has to be exercised when comparing system in the range d 7.0–7.35.In addition, the first methylene conventional organic molecules with organometallic systems, group of the alkoxy chain is observed as a virtual triplet at ca. in general lateral fluorination causes a broadening of the d 4.0. The remaining chain hydrogens appear in the range d molecule, reducing intermolecular attractions and leading to 0.8–1.8.The 19F NMR spectra of these compounds show one lower transition temperatures. On the other hand, polarization signal due to the fluorine atom present in the ring. effects can cause increased intermolecular interactions, leading On the other hand, the 1H NMR spectra of the trialkoxy- to higher transition temperatures.9 It is clear that the position phenyl isocyanide complexes are all very similar and show a of the fluorine atom (at either the 2- or 3-positions) does not singlet corresponding to two equivalent aromatic protons at affect significantly the breadth of the molecule.Thus the 2- ca. d 7, plus signals for the aliphatic chains. and 3-fluorinated derivatives will differ mainly in their polarization effects, and it is not surprising that an electronegative Mesogenic behaviour The behaviour of the fluorinated com- substituent(F) ortho to theOR group (3-fluorination) produces plexes is given in the Experimental section and summarized in a larger effect than when it is placed in the meta position (2- Fig. 1 and Fig. 2. fluorination). Moreover the fluorine atom in the 3-position Alomplexes display a smectic A (SA) mesophase would possibly favour the more polar anti conformation of except the 2-fluoro derivative with n=6, which shows a nematic the fluorine and the alkoxy chain, as has been found on 2- (N) mesophase, and the 2-fluoro iodo derivative, which is not fluoro-6-methoxy- and 2-fluoro-3-methoxy-pyridine derivamesomorphic.The SA mesophase presents the typical mielinic tives.10 Accordingly, the transition temperatures should and homeotropic textures reorganizing to the fan-shaped tex- decrease in the order 3-F>2-F, in agreement with the trend ture at temperatures close to the clearing point and the fan observed for the melting and the clearing points. texture in the cooling from the isotropic melt.The nematic The introduction of two new alkoxy chains in the phenyl phase shows the schlieren texture. moiety produces a dramatic change in the structure of The variation in properties observed is quite regular and the molecule, which cannot be considered rod-like any can be summarized as follows. The transition temperatures of longer. All the (trialkoxyphenyl isocyanide)gold compounds chlorogold derivatives decrease in the order 3-F>2-F, while [AuX{CNC6H2-3,4,5-(OCnH2n+1)3}] (X=Cl, n=4, 6, 8, 10; an increase of the length of the alkoxy chain produces little X=Br, I, n=10) display enantiotropic liquid crystal behaviour variation in the melting and clearing points.Moreover, the at room temperature, except the chloro derivative with n=4, which is monotropic (Fig. 3 and Experimental section). The Fig. 2 Thermal behaviour of the complexes [AuX(CNC6H3F-2- Scheme 2 OC12H25-4)] (F-2) and [AuX(CNC6H3F-3-OC12H25-4)] (F-3) 20 J. Mater. Chem., 1997, 7(1), 19–23in ppm relative to internal Me4Si for 1H and to external CFCl3 for 19F; J values are given in Hz. Microscopy studies were carried out using a Leitz microscope fitted with a hot stage and polarizers at a heating rate of ca. 10°C min-1. For differential scanning calorimetry (DSC), a Perkin Elmer DSC7 instrument was used, calibrated with water and indium; the scanning rate was 10°C min-1, the samples were sealed in aluminium capsules in the air and the holder atmosphere was dry nitrogen. X-Ray diffraction experiments were performed with a point focussing beam with Cu-Ka and a flat film.The samples were Fig. 3 Thermal behaviour of the complexes [AuCl{CNC6- held in Lindemann glass tubes situated in the gap of a magnet H2(OCnH2n+1)3-3,4,5}] with a field strength of 0.3–1.7 T. Literature methods were used to prepare [AuCl(tht)]13 optical textures, when viewed with a polarizing microscope on (tht=tetrahydrothiophene) and 3,4,5-trialkoxyphenylamine.14 cooling from the isotropic melt, are characteristic of hexagonal Only example procedures are described as the syntheses were columnar phases11 and display linear birefringent defects, large similar for the rest of the compounds.Yields, IR, analytical areas of uniform extinction and fan domains. X-Ray diffraction and thermal data (transition temperatures in °C and enthalpies for the chlorogold complex with n=10 from the mesophase in kJ mol-1 in parentheses) are given for all the gold complexes.reveals four low-angle rings and a broad halo at wide angle. The d-spacing of the four first diffraction rings scales as 3-Fluoro-4-octyloxynitrobenzene 15(1/3)1/251/25(1/7)1/2, consistent with a hexagonal lattice (a= Anhydrous K2CO3 (0.879 g, 6.36 mmol) was added to a solu- 2.97 nm).The presence of a single broad halo (0.45 nm) at tion of 2-fluoro-4-nitrophenol (0.500 g, 3.18 mmol) and octyl wide angle indicates that there are only weak liquid-like bromide (0.738 g, 3.82 mmol) in 50 ml of dry acetone. The interactions between the mesogens. Assuming an intracolum- resulting suspension was refluxed under N2 for 10 h and then nar distance of 0.4 nm, the data are consistent with a disc allowed to cool to room temperature.The reaction mixture formed by two molecules of the complex in antiparallel dispo- was poured into 200 ml of H2O and the solution was acidified sition, similar to the stacking in an alternating fashion reported with dilute HCO2H. To this was added CH2Cl2 and the organic for diketonate Schiff-base complexes of Ni, Cu and Pd.12 layer was separated, dried over MgSO4, filtered and the solvent The chloro compound with n=4, as above discussed, shows evaporated on a rotary evaporator.The residue was chromato- only a mesophase on cooling from the isotropic melt (exother- graphed (silica gel, CH2Cl2 as eluent) and the CH2Cl2 was mic peak at 13.3 °C) and its crystallization is not observed, evaporated to obtain the product as an oil (0.742 g, 87%); 1H possibly due to supercooling of the mesophase (Fig. 4). The NMR (CDCl3) d 7.98, 8.04 and 7.01 (Ha, Hb and Hc, AMNX compound recrystallizes on heating (exothermic peak at 15°C) spin system, 4Jab 2.7, 3Jbc 8.5, Ar-H), 4.12 (t, J 6.5, CH2O), and then undergoes a transition corresponding to the melting 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR (CDCl3 ) d -131.0 observed in the first heating cycle. The chloro compound with (m, 3JaF 10.7, 5JbF 1.5, 4JcF 8.5).n=6 shows, in addition to the columnar to isotropic transition, a partial crystallization and the subsequent melting to the 2-Fluoro-4-octyloxynitrobenzene mesophase in the heating cycles. For the rest of the (trialkoxyphenyl isocyanide)gold(I) complexes, only the transitions 1H NMR (CDCl3) d 6.71, 8.08 and 6.74 (Ha, Hb and Hc, between the hexagonal columnar mesophase and the isotropic AMNX spin system, 4Jac 2.7, 3Jbc 9.7, Ar-H), 4.02 (t, J 6.5, liquid are observed in the heating and cooling cycles.CH2O), 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR (CDCl3) d -113.2 (m, 3JaF 13.1, 4JbF 9.0, 5JcF 1.3).Experimental 3-Fluoro-4-octyloxyaniline Combustion analyses were performed with a Perkin Elmer 2400 microanalyser. IR spectra were recorded on a Perkin SnCl2 2H2O (3.1 g, 13.74 mmol) was added to a solution of Elmer FT 1720X instrument. 1H and 19F NMR spectra were 3-fluoro-4-octyloxynitrobenzene (0.740 g, 2.75 mmol) in 20 ml recorded at 300 and 282.38 MHz, respectively, on a Bruker of EtOH.The flask was purged with N2 and the suspension AC 300 instrument in CDCl3. Chemical shifts (d) are reported was refluxed for 30 min. The reaction mixture was allowed to cool to room temperature and poured into water. Solid K2CO3 was added until basic pH was achieved. The mixture was extracted with CH2Cl2 and the organic layer was separated, dried over MgSO4 and filtered.The solvent was removed on a rotary evaporator to obtain the product as white crystals (0.461 g, 70%); 1H NMR (CDCl3) d 6.46, 6.36 and 6.79 (Ha, Hb and Hc, AMNX spin system, 4Jab 2.7, 3Jbc 8.9, Ar-H), 3.93 (t, J 6.5, CH2O), 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR (CDCl3 ) d -133.2 (m, 3JaF 12.5, 5JbF 1.4, 4JcF 8.9). 2-Fluoro-4-octyloxyaniline 1H NMR (CDCl3) d 6.61, 6.70 and 6.53 (Ha, Hb and Hc, AMNX spin system, 4Jac 2.7, 3Jbc 8.6, Ar-H), 3.85 (t, J 6.5, CH2O), 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR (CDCl3) d -132.4 (m, 3JaF 12.6, 4JbF 10.1, 5JcF 1.1).N-(3-Fluoro-4-octyloxyphenyl)formamide A flask fitted with a Dean–Stark apparatus was charged with Fig. 4 DSC scans of [AuCl{CNC6H2(OCnH2n+1)3-3,4,5}] (a) first heating, (b) first cooling and (c) second heating a solution of 3-fluoro-4-octyloxyaniline (0.590 g, 2.47 mmol) J.Mater. Chem., 1997, 7(1), 19–23 21in 100 ml of toluene. Formic acid (10 ml, 98%) was added and H, alkyl chain); 19F NMR (CDCl3) d -129.2 (m, 4JcF 8.3); n(CON)/cm-1 (CH2Cl2) 2217; (Nujol) 2233, 2218 (Calc. for the resulting solution was refluxed for 1 h and then cooled to room temperature. The solvent was removed on a rotary C15H20NAuClFO: C, 37.40; H, 4.18; N, 2.91. Found: C, 37.65; H, 3.97; N, 3.08%); DSC/°C K–K¾ 146.9 (2.1), K¾–SA 174.7 evaporator and the residue recrystallized in pentane to give white crystals of a mixture of Z and E isomers of the product (17.3), SA–I 201.5 (8.9).(0.614 g, 93%); Z-isomer: 1H NMR (CDCl3 ) d 7.43, 7.14 and 6.8–7.0 (overlapped with Ha and Hc of the E-isomer) (Ha, Hb Chloro(3-fluoro-4-hexyloxyphenyl isocyanide) gold(i ) and Hc, AMNX spin system, 4Jab 2.5, 3Jbc 8.8, Ar-H), 8.32 (CH, Yield 66%; n(CON)/cm-1 (CH2Cl2) 2218; (Nujol) 2238, 2220 J 1.6), 7.20 (NH, br), 4.03–3.99 (CH2O, overlapped with the (Calc.for C13H16NAuClFO: C, 34.42; H, 3.55; N, 3.09. Found: CH2O of the E-isomer), 1.86–0.88 (m, alkyl chains of the two C, 34.54; H, 3.38; N, 3.57%).DSC: K–K¾ 156.0 (2.6), K¾–SA isomers); 19F NMR (CDCl3) d -132.4 (m, 3JaF 12.5, 5JbF 1.3, 177.7 (16.1), SA–I, 205.5 (8.9). 4JcF 9.1). E-isomer: 1H NMR (CDCl3 ) d 6.8–7.0 (Ha and Hc) (overlapped with Hc of the Z-isomer) and 6.79 (Hb) (AMNX spin Chloro( 3-fluoro-4-decyloxyphenyl isocyanide) gold(i ) system, 4Jab 2.6, 3Jbc 8.8, Ar-H), 8.51 (CH, J 11.4), 7.70 (NH, br), Yield 69%; n(CON)/cm-1 (CH2Cl2) 2217; (Nujol) 2235, 2218 4.03–3.99 (CH2O, overlapped with the CH2O of the Z-isomer), (Calc.for C17H24NAuClFO: C, 40.05; H, 4.75; N, 2.75. Found: 1.86–0.88 (m, alkyl chains of the two isomers); 19F NMR C, 40.27; H, 4.54; N, 2.65%); DSC/°C K–K¾ 152.5 (2.2), K¾–SA (CDCl3) d -131.5 (m, 3JaF 11.5, 5JbF 1.4, 4JcF 8.8, JCH–F 1.4). 169.9 (17.5), SA–I 196.5 (9.0). N-(2-fluoro-4-octyloxyphenyl )formamide Chloro( 3-fluoro-4-dodecyloxyphenyl isocyanide) gold(i ) Z-isomer: 1H NMR (CDCl3) d 8.10 (Hb) and 6.73–6.64 (Ha Yield 58%; n(CON)/cm-1 (CH2Cl2 ) 2218; (Nujol) 2240, 2221 and Hc) (overlapped with Ha and Hc of the E-isomer) (AMNX (Calc. for C19H28NAuClFO: C, 42.43; H, 5.25; N, 2.60. Found: spin system, 3Jbc 8.9, Ar-H), 8.40 (CH, J 1.6), 7.30 (NH, br), C, 42.70; H, 5.07; N, 2.72%); DSC/°C K–K¾ 148.0 (2.2), K¾–SA 3.94–3.89 (CH2O, overlapped with the CH2O of the E-isomer), 167.7 (18.3), SA–I 190.6 (8.7). 1.86–0.88 (m, alkyl chains of the two isomers); 19F NMR (CDCl3) d -128.4 (m, 4JbF 8.9). Chloro( 2-fluoro-4-octyloxyphenyl isocyanide)gold(i ) E-isomer: 1H NMR (CDCl3) d 6.73–6.64 (Ha and Hc) (overlapped with Ha and Hc of the Z-isomer) and 7.11 (Hb) Yield 64%; 1H NMR (CDCl3 ) d 6.76, 7.46 and 6.75 (Ha, Hb (AMNX spin system, 3Jbc 8.8, Ar-H), 8.46 (CH, J 11.5), 7.35 and Hc, AMNX spin system, 4Jac 2.6, 3Jbc 8.5, Ar-H), 3.99 (t, (NH, br), 3.94–3.89 (CH2O, overlapped with the CH2O of J 6.5, CH2O), 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR the Z-isomer), 1.86–0.88 (m, alkyl chains of the two isomers); (CDCl3 ) d -113.9 (m, 3JaF 8.0, 4JbF 8.5, 5JcF 1.7); n(CON)/cm-1 19F NMR (CDCl3) d -125.1 (m, 4JbF 8.8, JCH–F 1.4).(CH2Cl2) 2223; (Nujol) 2237 (Calc. for C15H20NAuClFO: C, 37.40; H, 4.18; N, 2.91. Found: C, 37.58; H, 4.04; N, 2.91%); 3-Fluoro-4-octyloxyphenyl isocyanide DSC/°C K–SA 114.7 (32.4), SA–I 117.8 (4.8). The procedure described by Ugi15 was followed, using triphosgene as dehydrating agent.Toa solution of 3-fluoro-4-octyloxy- Chloro(2-fluoro-4-hexyloxyphenyl isocyanide) gold(i ) formanilide (0.156 g, 0.58 mmol) and triethylamine (0.2 ml, Yield 58%; n(CON)/cm-1 (CH2Cl2) 2222; (Nujol) 2232 (Calc. 1.17 mmol) in 50 ml of CH2Cl2 was added dropwise a solution for C13H16NAuClFO: C, 34.42; H, 3.55; N, 3.09. Found: C, of triphosgene (0.053 g, 0.19 mmol) in 25 ml of CH2Cl2.The 34.53; H, 3.38; N, 2.99%); DSC/°C K–K¾ 62.4 (4.6), K¾–N, N–I mixture was stirred for 1 h and then the solvent was removed 114.1 (24.0). on a rotary evaporator. The resulting residue was chromatographed (silica gel, CH2Cl2–hexane, 351 as eluent) and the Chloro( 2-fluoro-4-decyloxyphenyl isocyanide) gold(i ) solvent was evaporated to obtain the product as a colourless oil (0.112 g, 77%); 1H NMR (CDCl3) d 7.14–7.10 (Ha and Hb) Yield 52%; n(CON)/cm-1 (CH2Cl2) 2223; (Nujol) 2238 (Calc.and 6.91 (Hc) (AMNX spin system, 3Jbc 8.5, Ar-H), 4.03 (t, J for C17H24NAuClFO: C, 40.05; H, 4.75; N, 2.75. Found: C, 6.5, CH2O), 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR 40.18; H, 4.47; N, 2.99%); DSC/°C K–K¾, K¾–SA 115.7 (32.9), (CDCl3) d -131.4 (m, 3JaF 11.2, 5JbF 2.4, 4JcF 8.5).SA–I 124.2 (6.0). 2-Fluoro-4-octyloxyphenyl isocyanide Chloro( 2-fluoro-4-dodecyloxyphenyl isocyanide) gold(i ) 1H NMR (CDCl3 ) d 6.69, 7.30 and 6.66 (Ha, Hb and Hc, Yield 68%; n(CON)/cm-1 (CH2Cl2) 2223; (Nujol) 2237 (Calc. AMNX spin system, 4Jac 2.6, 3Jbc 8.8), 3.95 (t, J 6.5, CH2O), for C19H28NAuClFO: C, 42.43; H, 5.25; N, 2.60. Found: C, 1.86–0.88 (m, 15 H, alkyl chain); 19F NMR (CDCl3) d -116.2 42.52; H, 5.00; N, 2.65%); DSC/°C K–K¾ 95.6 (0.9), K¾–SA (m, 3JaF 11.4, 4JbF 8.1, 5JcF 1.2). 119.3 (38.9), SA–I 126.7 (6.6). 3,4,5-Trioctyloxyphenyl isocyanide Chloro( 3,4,5-trioctyloxyphenyl isocyanide) gold(i ) 1H NMR (CDCl3) d 6.5 (s, 2 H, C6H2), 3.9 (m, 6 H, CH2O), Yield 83%; 1H NMR (CDCl3) d 6.7 (s, 2 H, C6H2), 3.9 (m, 6 1.9–0.8 (m, 45 H, alkyl chains).H, CH2O), 1.9–0.8 (m, 45 H, alkyl chains); n(CON)/cm-1 (CH2Cl2) 2223; (Nujol) 2225 (Calc. for C31H53NAuClO3: C, Chloro(3-fluoro-4-octyloxyphenyl isocyanide) gold(i) 51.70; H, 7.42; N, 1.95. Found: C, 51.01; H, 7.05; N, 1.96%); To a solution of [AuCl(tht)] (0.131 g, 0.41 mmol) in 30 ml of DSC/°C Colh–I 71.1 (3.2). CH2Cl2 was added 3-fluoro-4-octyloxyphenyl isocyanide (0.112 g, 0.45 mmol).After stirring for 10 min the solvent was Chloro( 3,4,5-tributoxyphenyl isocyanide)gold(i ) removed on a rotary evaporator and the residue was washed with diethyl ether. Recrystallization from CH2Cl2–EtOH Yield 83%; n(CON)/cm-1 (CH2Cl2) 2223; (Nujol) 2228 (Calc. for C19H29NAuClO3 : C, 41.35; H, 5.30; N, 2.54. Found: C, afforded the product as white crystals (0.145 g, 74%); 1H NMR (CDCl3) d 7.34–7.26 (Ha and Hb) and 7.00 (Hc) (AMNX spin 41.42; H, 5.10; N, 2.57%); DSC/°C K–I 38.0 (8.8), I–Colh 13.3 (-0.6).system, 3Jbc 8.3, Ar-H), 4.07 (t, J 6.5, CH2O), 1.86–0.88 (m, 15 22 J. Mater. Chem., 1997, 7(1), 19–23Chloro(3,4,5-trihexyloxyphenyl isocyanide)gold(i ) Bromo( 2-fluoro-4-dodecyloxyphenyl isocyanide) gold(i ) Yield 76%; 1H NMR (CDCl3) d 6.78–6.73 (Ha and Hc) and Yield 79%; n(CON)/cm-1 (CH2Cl2) 2223; (Nujol) 2226 (Calc.for C25H41NAuClO3: C, 47.21; H, 6.50; N, 2.20. Found: C, 7.45 (Hb) (AMNX spin system, 3Jbc 8.5), 3.99 (t, J 6.5, CH2O), 1.86–0.88 (m, 23 H, alkyl chain); 19F NMR (CDCl3) d -113.9 46.84; H, 6.18; N, 2.54%); DSC/°C CoIh–K 5.9 (-0.9), K–CoIh 22.0 (0.9), Colh–I 62.2 (2.3). (m, 4JbF 8.5); n(CON)/cm-1 (CH2Cl2) 2219; (Nujol) 2231 (Calc.for C19H28NAuBrFO: C, 39.19; H, 4.85; N, 2.41. Found: C, 39.27; H, 4.73; N, 2.36%); DSC/°C K–K¾ 109.4 (11.5), K¾–I Chloro(3,4,5-tridecyloxyphenyl isocyanide)gold(i ) 116.4 (35.4), I–SA 106.0 (-5.4), SA–K¾ 103.1 (-43.9). Yield 65%; n(CON)/cm-1 (CH2Cl2) 2223; (Nujol) 2226 (Calc. for C37H65NAuClO3: C, 55.25; H, 8.15; N, 1.74.Found: C, Bromo( 3,4,5-tridecyloxyphenyl isocyanide) gold(i) 55.27; H, 7.85; N, 1.79%); DSC/°C Colh–I 80.0 (3.0). Yield 88%; 1H NMR (CDCl3) d 6.7 (s, 2 H, C6H2), 3.9 (m, 6 H, CH2O), 1.9–0.8 (m, 57 H, alkyl chains); n(CON)/cm-1 Iodo( 3-fluoro-4-dodecyloxyphenyl isocyanide) gold(i ) (CH2Cl2) 2220; (Nujol) 2220 (Calc. for C37H65NAuBrO3: C, 52.36; H, 7.72; N, 1.65. Found: C, 53.04; H, 7.62; N, 1.61%); Potassium iodide (0.077 g, 0.46 mmol) was added to chloro(3- DSC/°C Colh–I 56.6 (2.1). fluoro-4-dodecyloxyphenyl isocyanide)gold(I ) (0.050 g, 0.09 mmol) dissolved in acetone (30 ml).A white precipitate We gratefully acknowledge the Comisio�n Interministerial de immediately appeared, and the resulting suspension was stirred Ciencia y Tecnologý�a for financial support (Project MAT93- for 10 min. The solvent was removed under reduced pressure 0329 and MAT96-0708).and the residue was extracted with CH2Cl2 (2×15 ml). The volume was reduced to 10 ml and EtOH (10 ml) was added References to obtain white crystals of the product (0.037 g, 63%); 1H NMR (CDCl3) d 7.35–7.26 (Ha and Hb) and 7.00 (Hc) (AMNX 1 A. M. Giroud-Godquin and P. M. Maitlis, Angew.Chem., Int. Ed. spin system, 3Jbc 8.3), 4.07 (t, J 6.5, CH2O), 1.86–0.88 (m, 23 Engl., 1 30, 375; P. Espinet, M. A. Esteruelas, L. A. Oro, H, alkyl chain); 19F NMR (CDCl3) d -129.2 (m, 4JcF 8.3); J. L. Serrano and E. Sola, Coord. Chem. Rev., 1992, 117, 215; Inorganic Materials, ed. D. W. Bruce and D. O’Hare, Wiley, n(CON)/cm-1 (CH2Cl2) 2212; (Nujol) 2233, 2211 (Calc.for Chichester, 1992, ch. 8; S. A. Hudson and P. M. Maitlis, Chem. C19H28NAuFIO: C, 36.26; H, 4.49; N, 2.23. Found: C, 36.46; Rev., 1993, 93, 861; Metallomesogens, ed. J. L. Serrano, VCH, H, 4.25; N, 2.21%); DSC/°C K–K¾ 56.0 (2.9), K¾–SA 93.0 (26.2), Weinheim, 1996. SA–I 117.5 (4.5). 2 H. Adams, A. C. Albe�niz, N. A. Bailey, D. W. Bruce, A. S. Cherodian, R. Dhillon, D. A. Dunmur, P.Espinet, J. L. Feijoo, E. Lalinde, P. M. Maitlis, R. M. Richardson and G. Ungar, Iodo( 2-fluoro-4-dodecyloxyphenyl isocyanide) gold(i ) J. Mater. Chem., 1991, 1, 843. 3 D. W. Bruce, E. Lalinde, P. Styring and P. M. Maitlis, J. Chem. Yield 87%; 1H NMR (CDCl3 ) d 6.78–6.73 (Ha and Hc) and Soc., Chem. Commun., 1986, 581. 7.45 (Hb) (AMNX spin system, 3Jbc 8.5), 3.99 (t, J 6.5, CH2O), 4 T.Kaharu, R. Ishii and S. Takahashi, J. Chem. Soc., Chem. 1.86–0.88 (m, 23 H, alkyl chain); 19F NMR (CDCl3) d -113.8 Commun., 1994, 1349; T. Kaharu, R. Ishii, T. Adachi, T. Yoshida (m, 4JbF 8.5); n(CON)/cm-1 (CH2Cl2) 2214; (Nujol) 2223 (Calc. and S. Takahashi, J.Mater. Chem., 1995, 5, 687. for C19H28NAuFIO: C, 36.26; H, 4.49; N, 2.23. Found: C, 5 M. Benouazzane, S. Coco, P.Espinet and J. M. Martý�n-Alvarez, 36.21; H, 4.24; N, 2.49%); DSC/°C K–I 120.6 (54.3). J. Mater. Chem., 1995, 5, 441; P. Alejos, S. Coco and P. Espinet, New J. Chem., 1995, 19, 799. 6 S. Coco, P. Espinet, S. Falaga�n and J. M. Martý�n-Alvarez, New Iodo( 3,4,5-tridecyloxyphenyl isocyanide) gold(i) J. Chem., 1995, 19, 959. 7 R. Ishii, T. Kaharu, N. Pirio, S.-W. Zhang and S. Takahashi, Yield 90%; 1H NMR (CDCl3) d 6.7 (s, 2 H, C6H2), 3.9 (m, 6 J.Chem. Soc., Chem. Commun., 1995, 1215. H, CH2O), 1.9–0.8 (m, 57 H, alkyl chains); n(CON)/cm-1 8 G. Jia, R. J. Puddephat, J. J. Vittal and N. C. Payne, (CH2Cl2) 2215; (Nujol) 2209 (Calc. for C37H65NAuIO3: C, Organometallics, 1993, 12, 263; G. Jia, N. C. Payne, J. J. Vittal and 49.61; H, 7.31; N, 1.56. Found: C, 49.55; H, 7.22; N, 1.44%); R. J. Puddephat, Organometallics, 1993, 12, 4771; G. Jia, R. J. Puddephat, J. D. Scott, J. J. Vittal and N. C. Payne, DSC/°C Colh–I 45.1 (4.6). Organometallics, 1993, 12, 3565. 9 D. W. Bruce and S. A. Hudson, J. Mater. Chem., 1994, 4, 479. Bromo(3-fluoro-4-dodecyloxyphenyl isocyanide)gold (i) 10 V. Reiffenrath and M. Bremer, Angew. Chem., Int. Ed. Engl., 1994, 33, 1386. The method followed was the same as above, using KBr 11 A. G. Serrette, T. M. Swager, Angew. Chem., Int. Ed. Engl., 1994, instead of KI and stirring overnight. Yield 71%; 1H NMR 33, 2342. 12 H. Zheng, C. K. Lai and T. M. Swager, Chem.Mater., 1994, 6, 101. (CDCl3) d 7.33–7.27 (Ha and Hb) and 7.00 (Hc) (AMNX spin 13 R. Uso�n, A. Laguna and J. Vicente, J. Organomet. Chem., 1977, system, 3Jbc 8.6), 4.08 (t, J 6.5, CH2O), 1.86–0.88 (m, 23 H, 131, 471. alkyl chain); 19F NMR (CDCl3) d -129.2 (m, 4JcF 8.6); 14 A. Zinsou, M. Veber, H. Strzelecka, C. Jallabert and P. Fourre�, n(CON)/cm-1 (CH2Cl2) 2215; (Nujol) 2236, 2217 (Calc. for New J. Chem., 1993, 17, 309. C19H28NAuBrFO: C, 39.19; H, 4.85; N, 2.41. Found: C, 39.17; 15 I. Ugi and R. Meyr, Chem. Ber., 1960, 93, 239. H, 4.71; N, 2.33%); DSC/°C K–K¾ 110.2 (2.4), K¾–K, K–SA 129.8 (22.7), SA–I 161.5 (7.1). Paper 6/04258I; Received 18th June, 1996 J. Mater. Chem., 1997, 7(1), 19

ISSN:0959-9428    

DOI:10.1039/a604258i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Tetrafluoro and dichloro derivatives of thiophene-fused DCNQI- and TCNQ-type acceptors: a synthetic, electrochemical and crystallographic study Nazario Martý�n,*,a Pilar de Miguel,a Carlos Seoane,*,a Armando Albertb and Fe� lix H. Canob aDepartamento de Quý�mica Orga�nica I, Facultad de Quý�mica, Universidad Complutense, 28040-Madrid, Spain bDepartamento de Cristalografý�a, Instituto de Quý�mica Fý�sica ‘Rocasolano’ (CSIC), Serrano 119, 28006-Madrid, Spain Novel thiophene-fused DCNQI derivatives 8 and 11 bearing four fluorine atoms have been obtained in good yield from the corresponding quinones by reaction with bis(trimethylsilyl)carbodiimide (BTC).The presence of four fluorine atoms leads to good acceptor molecules which form charge transfer complexes in solution with N,N¾-tetramethyl-p-phenylenediamine.The effect of chlorine atoms on the crystal packing in the analoguous thiophene-fused TCNQ derivatives 3a and 3b is also reported. Tetracyano-p-quinodimethane (TCNQ) and dicyano-p-quino- to the less sterically demanding cyanoimino group in comparison with the dicyanomethylene fragment. In a previous paper5b diimine (DCNQI) have been used most widely as acceptors1 for the preparation of novel charge transfer complexes (CTC) we carried out a comparative crystallographic study of compounds of this type and important differences in terms of and charge transfer salts.2 On the other hand, the electron acceptors containing ben- planarity and crystal packing were found due to the highly distorted geometry in 3a.Considering the effect that substitu- zene-fused TCNQ and DCNQI derivatives 1 and 2, prepared from 9,10-anthraquinone,3,4 do not form CTC when they are ents can play on the stacking pattern, we have determined the X-ray structure of compound 3b bearing two chlorine atoms.mixed with donor molecules such as tetrathiafulvalene (TTF). This is due to their poor acceptor ability and, particularly, the highly distorted structure of 1.Although the same result was Results and Discussion found for the acceptors 3 and 4 obtained from 4,9-dihydronaphtho[ 2,3-b]thiophene-4,9-dione, the condensation of a thio- The synthesis of the novel p-extended DCNQI-type derivatives 8 and 11 was carried out from the corresponding halogeno- phene ring instead of a benzene ring resulted in a more planar molecule.5 substituted quinones 7 and 10 by reaction with bis(trimethylsilyl) carbodiimide (BTC) by following Hu�nig’s procedure (Scheme 1).7 The quinones 7 and 10 were prepared from tetrafluorophthalic anhydride 5 by reaction with benzene and aluminium trichloride to yield the corresponding 2-benzoyl- 3,4,5,6-tetrafluorobenzoic acid 6 which, by treatment with concentrated sulfuric acid at 100 °C, led to the 1,2,3,4-tetra- fluoroanthraquinone 7.In a similar way, the reaction of anhydride 5 with thienyllithium in diethyl ether at -78 °C yielded the acid 9 which, after treatment with PCl5 and AlCl3 in dry nitrobenzene gave the novel 5,6,7,8-tetrafluoro-4,9- dihydronaphtho[2,3-b]thiophene-4,9-dione 10 (see Experimental section). An alternative procedure for the preparation of substituted naphtho[2,3-b]thiophenediones by microwave assisted cyclization of substituted 2-thienylcarbonylbenzoic acids, catalysed with clays in dry media, has been recently described.8 Compounds 8 and 11 can exist as the syn and anti isomers.The presence of isomers in other DCNQI analogues has been previously established by NMR spectroscopy4 and molecular mechanics analyses.9 The high resolution 1H NMR spectrum of compound 8 shows a sharp doublet of doublets at d 7.93 and a broad signal at d 8.67 due to the rapid isomerization of the cyanoimino group at room temperature.This fact could be accounted for by the similar steric hindrance of hydrogen and fluorine peri atoms, resulting in a non-favoured configuration. 4,9 Accordingly, compound 11 showed a sharp doublet at d 8.09 and another at d 8.80 suggesting a favoured configuration with the cyano groups towards the thiophene ring, in agreement with the previously reported X-ray data for the Taking into account the strong effect of fluorine or chlorine atoms on the acceptor capacity,6 we have carried out the unsubstituted compound 4a.5b Cyclic voltammetric (CV) measurements of compounds 8 synthesis of novel DCNQI acceptors 8 and 11 derived from 9,10-anthraquinone and 4,9-dihydronaphtho[2,3-b]thiophene- and 11, as well as the starting quinones 7 and 10, were carried out inCH2Cl2 at room temperature using tetrabutylammonium 4,9-dione bearing four fluorine atoms on the benzene ring.The geometry of these DCNQI-type acceptors may present greater perchlorate as the supporting electrolyte.The data obtained are summarized in the Table 1. All compounds showed two planarity than the corresponding TCNQ-type analogues, due J. Mater. Chem., 1997, 7(1), 25–29 25Table 1 Cyclic voltammetry data of acceptors compound E1/21/V E1/22/V DE log K 1a -0.20 (2e-) 2a -0.11 -0.46 0.35 5.93 3aa -0.18 (2e-) 3bb -0.10 (2e-) 4ab -0.11 -0.61 0.50 8.62 4bb -0.04 -0.50 0.46 7.93 7c -0.72 -1.23 0.51 8.79 8c -0.19 (-0.19)b -0.50 (-0.48)b 0.31 5.34 10c -0.57 -1.08 0.51 8.79 11c 0.00 -0.46 0.46 7.93 DCNQIc 0.21 -0.41 0.62 10.68 a CH2Cl2/Bu4N+BF4-/Pt vs.Ag/AgCl/MeCN (refs. 4, 5). Fig. 1 Cyclic voltammetry of compound 8 (scan rate 50 mV s-1) b MeCN/Bu4N+ClO4-/glassy carbon vs. Ag/Ag+ (ref. 6). c CH2Cl2/Bu4N+BF4-/glassy carbon vs.SCE; scan rate 50 mV s-1. found for compounds 8 and 11 clearly confirm these trends, one-electron reduction waves to the corresponding anion- and the thiophene-containing compound 11 exhibits a first radical and dianion (Fig. 1). Replacing the benzene ring with reduction potential which is shifted 190 mV to more positive a thiophene ring leads to a better acceptor due to the lowering values (Table 1).of the steric hindrance. The presence of four fluorine atoms as Increasing benzannulation with regard to the parent DCNQI substituents on the benzene ring significantly decreases the leads to a decrease in the acceptor ability and also to a reduction potential values. The reduction potential values diminishing of the thermodynamic stability of the corresponding radical-anions (lower log K values10 in Table 1).Thus, compound 8 exhibits the least stable radical-anion due to the stronger steric interaction between the cyano group and the peri-hydrogens. Unlike the non-fluorine substituted analogues, compounds 8 and 11 show evidence of complexation when they were mixed, in equimolecular amounts in dichloromethane under inert atmosphere, with the strong donor N,N,N¾,N¾-tetramethyl- p-phenylenediamine (TMPD).Although solid stable CTCs were not isolated, a colour change was observed. The UV–VIS spectra of the reaction solution, under argon atmosphere, showed a typical low-energy charge-transfer band (8 TMPD: lmax=538 nm; 11 TMPD: lmax=545 nm). Attempts to form CT-complexes with acceptor 11, which showed the best acceptor ability in this study, with other donors such as tetrathiafulvalene (TTF) did not afford the corresponding solid CT-complex.A weak broad band was observed at around 450 nm in the electronic spectrum which might indicate the presence of a very low energy CT-band in solution. These findings are also in agreement with those previously reported for the poorer electron-acceptor dithiophene- fused TCNQ {4,8-bis(dicyanomethylene)-4,8-dihydrobenzo-[ 1,2-c:4,5-c¾]dithiophene}11 which, in contrast to the other isomers, did not form a CT-complex with TTF. Finally, attempts to form a metal salt were carried out by reaction of acceptors 8 and 11 with copper iodide in acetonitrile under inert atmosphere.A black solid precipitate was only obtained with 8, which showed the IR stretching cyano band at 2177 cm-1.According to the elemental analysis, a rather unusual 155 acceptor–copper stoichiometry was found. The molecular structures of compounds 3b and the unsubstituted analogue 3a are shown in Fig. 2, together with the atomic numbering scheme. The main geometrical features of the common moiety of both compounds are quite similar, and only bongles involving atoms of the thiophene ring show significant differences (Table 2).This effect could be due to an artifact of the disorder which both molecules possess. Thus, the presence of the two chlorine atoms in compound 3b does not affect significantly the conformation of this molecule. Both compounds present a butterfly shape centred at the two substituted carbon atoms, with an interplanar angle between the S(1)–C(2)–C(3)–C(4)–C(5)– C(12)–C(13) and C(5)–C(6)–C(7)–C(8)–C(9)–C(10)–C(11)– C(12) planes of 147.5° (1) for compound 3b and 149.7° (1) for Scheme 1 Reagents and conditions: i, C6H6, AlCl3 ; ii, H2SO4; compound 3a. This distortion from planarity has been iii, Me3SiN=C=NSiMe3, TiCl4, CH2Cl2; iv, 2-thienyllithium, then NH4Cl, HCl; v, PCl5, C6H5NO2, AlCl3 accounted for by the steric hindrance between the cyano 26 J.Mater. Chem., 1997, 7(1), 25–29Table 2 Geometrical features of compounds 3a bond lengths/A° 3b 3a S(1)–C(2) 1.80(1) 1.68(2) S(1)–C(13) 1.648(3) 1.66(2) C(2)–C(3) 1.37(1) 1.30(6) C(3)–C(4) 1.541(5) 1.46(6) C(4)–C(5) 1.462(4) 1.458(4) C(4)–C(13) 1.393(6) 1.391(5) C(5)–C(6) 1.465(5) 1.471(5) C(5)–C(14) 1.36(1) 1.373(5) C(6)–C(7) 1.469(5) 1.383(5) C(6)–C(11) 1.405(6) 1.415(4) C(7)–C(8) 1.29(1) 1.399(6) C(8)–C(9) 1.31(1) 1.388(5) C(8)–Cl(24) — 1.713(4) C(9)–C(10) 1.265(6) 1.380(5) C(9)–Cl(25) — 1.719(4) C(10)–C(11) 1.484(5) 1.392(5) C(11)–C(12) 1.469(4) 1.468(5) C(12)–C(13) 1.453(4) 1.454(5) C(12)–C(19) 1.363(5) 1.362(5) C(14)–C(15) 1.432(7) 1.425(6) C(14)–C(17) 1.430(5) 1.435(5) C(15)–N(16) 1.142(7) 1.150(6) C(17)–N(18) 1.152(5) 1.141(6) C(19)–C(20) 1.436(5) 1.441(5) C(19)–C(22) 1.440(5) 1.433(6) C(20)–N(21) 1.140(5) 1.133(6) C(22)–N(23) 1.144(6) 1.139(6) bond angles/degrees C(2)–S(1)–C(13) 93.1(3) 94(1) S(1)–C(2)–C(3) 114.1(5) 113(3) C(2)–C(3)–C(4) 105.4(5) 110(4) C(3)–C(4)–C(13) 115.4(4) 112(3) C(3)–C(4)–C(5) 124.9(4) 128(3) Fig. 2 Molecular structure of compounds (a) 3a and (b) 3b, showing C(5)–C(4)–C(13) 119.7(3) 120.3(3) the atomic numbering C(4)–C(5)–C(14) 122.4(3) 122.3(3) C(4)–C(5)–C(6) 114.6(3) 114.4(3) C(6)–C(5)–C(14) 122.8(3) 123.3(3) groups and the peri hydrogens. This butterfly shape of the C(5)–C(6)–C(11) 119.0(3) 118.8(3) C(5)–C(6)–C(7) 125.3(4) 122.2(3) molecules 3a and 3b with a boat type quinonoid group was C(7)–C(6)–C(11) 115.7(3) 119.0(3) also found in 11,11,12,12-tetracyano-9,10-anthraquinodime- C(6)–C(7)–C(8) 121.7(4) 120.8(3) thane (TCAQ).12 Different isomeric benzodithiophene ana- C(7)–C(8)–C(9) 122.6(5) 119.4(4) logues of TCAQ have also been obtained and their electronic C(7)–C(8)–Cl(24) — 117.0(3) and steric properties were found to depend on the mode of C(9)–C(8)–Cl(24) — 123.6(3) fusion of the two thiophene units to the central TCNQ ring.11 C(8)–C(9)–C(10) 122.3(5) 120.7(4) C(8)–C(9)–Cl(25) — 120.5(3) The crystal structures of these thiophene-fused TCNQs showed C(10)–C(9)–Cl(25) — 118.8(3) a more planar butterfly shape, in comparison with the TCAQ C(9)–C(10)–C(11) 122.2(4) 120.0(4) molecule.C(6)–C(11)–C(10) 115.4(3) 119.9(4) On the other hand, crystal packing is severely modified C(10)–C(11)–C(12) 124.9(3) 121.1(3) (Fig. 3).Compound 3b shows aromatic interactions, forming C(6)–C(11)–C(12) 119.7(3) 118.9(3) dimers and Cl,N13 contacts in the crystal packing, and with C(11)–C(12)–C(19) 122.4(3) 122.9(3) C(11)–C(12)–C(13) 114.4(3) 113.9(3) a contact Cl(25),Cl(24) (x-1/2, y, 1/2-z) of 3.461 (2) A° . C(13)–C(12)–C(19) 123.2(3) 122.9(3) However, compound 3a presents the aromatic interactions as C(4)–C(13)–C(12) 120.0(3) 119.1(3) the main intermolecular contacts, forming a stacking pattern S(1)–C(13)–C(12) 128.2(3) 130.9(4) along the (2 0 1) direction with tight packing.Thus it seems S(1)–C(13)–C(4) 111.8(3) 109.9(4) that, here again, the chlorine atoms, when competing with the C(5)–C(14)–C(17) 124.0(4) 124.2(4) rings interactions, play a role in the packing.C(5)–C(14)–C(15) 124.0(4) 123.3(4) C(15)–C(14)–C(17) 111.6(4) 112.4(4) Both molecules 3a and 3b present structural disorder, con- C(14)–C(15)–N(16) 174.1(5) 175.8(4) sisting of the superposition of two ordered structures related C(14)–C(17)–N(18) 175.9(5) 176.2(5) by a pseudo mirror plane through C(2) and the mid-point of C(12)–C(19)–C(22) 123.5(4) 122.3(3) C(8)–C(9).The population factors were 0.53–0.47(1) and C(12)–C(19)–C(20) 124.8(4) 123.2(4) 0.72–0.298(1) for both compounds 3b and 3a, respectively, C(20)–C(19)–C(22) 111.7(4) 114.0(3) thus showing the symmetrizing effect of the chlorine atoms. C(19)–C(20)–N(21) 175.4(5) 175.9(5) C(19)–C(22)–N(23) 177.2(5) 178.5(5) The disorder was constrained in the refinement so as to fit the pseudo mirror plane with the populations equal to 1.a Both compounds present disorder. Distances and angles are given In conclusion, we have carried out the synthesis, electro- for the major structure. chemical and crystallographical study of some novel halogencontaining acceptors. The presence of four fluorine atoms is responsible for the greater acceptor ability of 8 and 11 as shown by the cyclic voltammetry study.The reasonably good first reduction potentials, in addition to the expected planar J. Mater. Chem., 1997, 7(1), 25–29 271.56 g cm-3, F(000)=760, m=49.44 cm-1. Refined cell parameters were obtained from setting angles of 72 reflections. A prismatic brown crystal (0.30×0.25×0.15 mm) was used for the analysis.Data collection. Automatic four circle diffractometer Philips PW 1100 with graphite oriented monochromated Cu-Ka radiation. The intensity data were collected using the v–2h scan mode with 2<h<65°; two standard reflections were measured every 90 min with no intensity variation. A total of 2743 reflections were measured and 2286 were considered as observed [I>3s(I )].The data were corrected for Lorentz and polarization effects. Structure solution and refinement. The structure was solved by direct methods using SIR8814 and DIRDIF.15 Hydrogen atoms were calculated and fixed in the final mixed refinement; isotropic thermal parameters of these atoms were considered as fixed contributors. A convenient weighting scheme was applied16 to obtain no dependence in wD[14]F vs.F0 and sin h/l .The final R (Rw) value was 5.9 (7.4). Atomic scattering factors for the compound were taken from International Tables for X-Ray Crystallography17 and calculations were performed using XRAY80,18 XTAL,19 HSEARC20 and PARST.21 Fig. 3 Crystal packing for (a) compound 3a along the axis perpendicu- Atomic coordinates, thermal parameters, and bond lengths lar to the (2 0 1) direction and (b) compound 3b along the a axis.and angles have been deposited at the Cambridge Aromatic interactions are represented with dashed lines. Crystallographic Data Centre (CCDC). See Information for Authors, J. Mater. Chem., 1996, Issue 1. Any request to the geometry for compounds 8 and 11, are responsible for the CCDC for this material should quote the full literature citation complexation observed in the UV–VIS spectra when they are and the reference number 1145/15.in solution in the presence of N,N,N¾,N¾-tetramethyl-p-phenylenediamine. Although these novel acceptors did not form CT- 2-Benzoyl-3,4,5,6-tetrafluorobenzoic acid 6 complexes with TTF, a copper CT-salt was formed with This compound was obtained from tetrafluorophthalic anhy- compound 8.On the other hand, the presence of halogen dride, benzene and AlCl3 by following the standard Friedel– atoms also plays an important role in the crystal packing, as Crafts procedure.22 Yield 70%; mp 177 °C (from water) (Calc. observed in the intermolecular interactions of the related for C14H6F4O3: C, 56.39; H, 2.03. Found: C, 56.42; H, 2.12%); compounds 3a and 3b.These data indicate that although the dH(300 MHz, [2H6]DMSO) 7.41–7.89 (5H, m, ArH), 14.20 presence of fluorine atoms increases the acceptor ability in (1H, br s, CO2H); nmax(KBr)/cm-1 3000, 1715 (CNO, acid), these aryl-fused DCNQI derivatives, they are not strong 1680 (CNO), 1630, 1600, 1590, 1530, 1470, 1440, 1370, 1320. enough acceptors to form CT-complexes with molecules which form stable radical-cation such as TTF.The presence of 2-(2-Thienylcarbonyl )-3,4,5,6-tetrafluorobenzoic acid 9 halogen atoms on molecules containing two five-membered heterocyclic rings fused to the central TCNQ and DCNQI 2-Thienyllithium (4.6 ml; 1.0 M in tetrahydrofuran) was slowly acceptors should lead to stronger acceptors with more planar added to a mixture of tetrafluorophthalic anhydride (1 g, geometries. 4.54 mmol) in dry diethyl ether (15 ml) at-20 °C. The mixture was maintained under argon and allowed to warm to room temperature, then stirred for 24 h. The cold reaction mixture Experimental containing the lithium complex was decomposed by the General addition of saturated NH4Cl (30 ml) and then acidified with 6 M HCl.The organic layer was separated and the aqueous Melting points were determined using a Gallenkamp melting phase washed with diethyl ether; the organic extracts were point apparatus and are uncorrected. IR spectra were recorded washed with saturated brine and water, and treated with 5% on a Perkin-Elmer 398 spectrometer. UVspectra were recorded sodium hydrogen carbonate (3×50 ml).The aqueous alkaline on a Perkin-Elmer Lambda-3 instrument. 1H NMR spectra layers were collected, cooled and acidified by the dropwise were determined with a Varian XL-300S spectrometer and J addition of 6 M HCl with vigorous stirring. The keto acid values are given in Hz. Elemental analyses were performed on which precipitated was filtered and washed free of the mineral a Perkin-Elmer CHN 2400 apparatus.acid with water. The crude product was crystallized from Cyclovoltammetric measurements were performed on a EG boiling water. Yield 60%; mp 175 °C (Calc. for C12H4F4O3S: & PAR Versastat potentiostat using 250 electrochemical analy- C, 47.38; H, 1.33. Found C, 47.60; H 1.35%); dH(300 MHz, sis software. A Metrohm 6.084.C10 glassy carbon electrode CDCl3) 7.14 (1H, dd, J 3.9 and 4.8, H-4-thiophene), 7.43 (1H, was used as indicator electrode in voltammetric studies.d, J 3.9, H-5-thiophene), 7.80 (1H, dd, J 4.8 and 1.2 H-3- Tetrafluorophthalic anhydride, 2-thienyllithium, titanium thiophene); nmax(KBr)/cm-1 3200, 1750 (CNO, acid), 1640 tetrachloride and bis(trimethylsilyl)carbodiimide (BTC) are (CNO), 1520, 1480, 1420, 1370, 1300.commercially available. X-Ray crystallographic measurements 1,2,3,4-Tetrafluoroanthraquinone 7 Compound 7 was obtained by treatment of 6 (1 g, 3.3 mmol) Crystal data for compound 3b. C18H4N4S1Cl2, MW= 379.222, monoclinic, P21/c, a=9.229 (1), b=19.609 (4), c= with conc. sulfuric acid (5 ml) at 100 °C. The reaction mixture was cooled and poured into ice–water, and the product was 9.457 (1) A° , b=109.29°, V=1615.4 (4) A° , Z=4, DC= 28 J.Mater. Chem., 1997, 7(1), 25–29collected by filtration to yield an orange solid. Yield 85%; mp nitrile was transferred via cannula under argon atmosphere to a boiling solution of CuI (77 mg, 0.4 mmol) in 10 ml of dry 224 °C (Calc. for C14H4F4O2: C, 60.02; H, 1.44. Found: C, 60.31; H, 1.21%); dH(300 MHz, CDCl3) 7.80 (2H, dd, J 5.6 acetonitrile; upon cooling, the resulting precipitate was filtered, washed and dried, giving the solid product in 30% yield (Calc.and 3.3, ArH), 8.31 (2H, dd, J 5.6 and 3.3, ArH); nmax(KBr)/cm-1 3450, 1680 (CNO), 1620, 1510, 1470, 1400, for C16H4F4N4Cu5: C, 29.87; H, 0.63. Found: C, 30.67; H, 1.04%); nmax(KBr)/cm-1 2177 (CN), 1568, 1505, 1389, 1266. 1370. Financial support by the European Commission (Project 5,6,7,8-Tetrafluoro-4,9-dihydronaphtho[2,3-b]thiophene-4,9- CT93-0066) is gratefully acknowledged.dione 10 This compound was obtained according to the reported pro- References cedure.23 To a solution of 6 (0.5 g, 1.64 mmol) and PCl5 (0.512 g, 2.46 mmol) in 20 ml of dry nitrobenzene, AlCl3 (0.33 g, 1 M. R. Bryce and L. C. Murphy, Nature, 1984, 309, 119; S.Hu�nig, 2.46 mmol) was added. The mixture was kept at room tempera- Pure Appl. Chem., 1990, 62, 395; S. Hu�nig and P. Erk, Adv. Mater., ture for 1 h, and then at 140°C for 4 h. The solvent was 1991, 225; S. Hu�nig, J. Mater. Chem., 1995, 5, 1469. 2 See for example: Proceedings of the International Conference on distilled under vacuum and a black oil was obtained.The Science and T echnology of SyntheticMetals, Tu� bingen, 1990, Synth. crude compound was purified by column chromatography Met., 1991, 41; Goteborg, 1992, Synth. Met., 1993, 55. over silica gel using hexane–ethyl acetate (451) as eluent. 3 S. Yamaguchi, H. Tatemitsu, Y. Sakata and S. Misumi, Chem. L ett., Further purification was accomplished by recrystallization 1983, 1229; A.Aumu�ller and S. Hu�nig, L iebigs Ann. Chem., 1984, from ethanol. Yield 55%, mp 190 °C (Calc. for C12H2F4O2S: 618; B. Ong and S. Keoshkerian, J. Org. Chem., 1984, 49, 5002; C, 50.36; H, 0.70. Found: C, 50.14; H, 0.77%); dH(300 MHz, A. M. Kini, D. O. Cowan, F. Gerson and R. Mo�ckel, J. Am. Chem. Soc., 1985, 107, 556. CDCl3) 7.69 (1H, d, J 5.1, thiophene), 7.79 (1H, d, J 5.1, 4 A.Aumu�ller and S. Hu�nig, L iebigs Ann. Chem., 1986, 142. thiophene); nmax(KBr)/cm-1 1680 (CNO), 1630, 1540, 1510, 5 (a)P. Cruz, N. Martý�n, F. Miguel, C. Seoane, A. Albert, F. H. Cano, 1480, 1410, 1380. A. Leverenz and M. Hanack, Synth.Met., 1992, 48, 59; (b) P. Cruz, N. Martý�n, F. Miguel, C. Seoane, A. Albert, F. H. Cano, Condensation reaction of quinones 7 and 10 with BTC A.Gonza�lez and J. M. Pingarro�n, J. Org. Chem., 1992, 57, 6192. 6 N. Martý�n, J. L. Segura, C. Seoane, C. Torý�o, A. Gonza�lez and General procedure. To a solution of the corresponding J. M. Pingarro�n, Synth. Met., 1994, 64, 83. See also refs. 2 and 3. quinone (2 mmol) in dry CH2Cl2 (50 ml) at room temperature 7 A. Aumu�ller and S. Hu�nig, Angew. Chem., Int. Ed. Engl., 1984, 23, and under argon atmosphere, TiCl4 followed by BTC were 447; A.Aumu�ller, P. Erk, G. Klebe, S. Hu�nig, U. Schu�tz and H- added dropwise with a syringe by using a variable stoichio- P.Werner, Angew. Chem., Int. Ed. Engl., 1986, 25, 740. 8 A. Acosta, P. de la Cruz, P. de Miguel, E. Diez-Barra, A. de la Hoz, metric ratio (see below). The reaction was stirred for the F. Langa, A. Loupy, M.Majdoub, N. Martý�n, C. Sa�nchez and required time and monitored by thin layer chromatography C. Seoane, T etrahedron L ett., 1995, 36, 2165. (TLC) until the starting quinone had been consumed, when 9 E. Barranco, N. Martý�n, J. L. Segura, C. Seoane, P. Cruz, F. Langa, CH2Cl2 (200 ml) was added and the reaction mixture was A. Gonza�lez and J. M. Pingarro�n, T etrahedron, 1993, 49, 4881.poured into ice–water (200 g). The reaction was vigorously 10 B. S. Jense and V. D. Parker, J. Am. Chem. Soc., 1975, 97, 5211. stirred until the solution reached room temperature. The 11 K. Kobayashi,C. L. Gajurel, K. Umemoto, Y. Mazaki, Bull. Chem. Soc. Jpn., 1992, 65, 2168; F. Iwasaki, N. Toyota, M. Hirota, organic phase was separated and washed with water N. Yamazaki, M.Yasui and K. Kobayashi, Bull. Chem. Soc. Jpn., (3×50 ml), dried (MgSO4) and concentrated to 10 ml. The 1992, 65, 2173. same volume of hexane (10 ml) was added and the solid 12 U. Shubert, S. Hu�nig and A. Aumu�ller, L iebigs Ann. Chem., 1985, precipitated. It was collected by filtration and washed with 1216. hexane. 13 R. Gautam Desiraju, Crystal Engineering, Elsevier, Amsterdam, 1989. 14 G. Cascarano, C. Giacovazzo, M. G. Burla, G. Polidori, N,N¾-Dicyano-1,2,3,4-tetrafluoroanthraIR88, 1988. Compound 8 was obtained, by stirring for 24 h at room 15 The DIRDIF Program System, P. T. Beurskens, G. Admiraal, temperature and refluxing for 3 h using 7 (2.1 mmol), TiCl4 H. Behm, G. Beurskens, W. P. Bosman, S.Garcý�a-Granda, (8.7 mmol) and BTC (7.4 mmol), in 60% yield; mp 149 °C R. O. Gould and C. Smykalla, Z. Kristallogr., 1990, suppl. 4, 99. (Calc. for C16H4F4N4: C, 58.55; H, 1.23. Found: C, 58.34; H, 16 M. Martý�nez-Ripoll and F. H. Cano, PESOS, A computer pro- 1.37%); dH(300 MHz, CDCl3) 7.93 (2H, dd, J 5.5 and 3.1, gram for the automatic treatment of weighting schemes, Instituto Rocasolano C.S.I.C., Serrano 119, 28006, Madrid, Spain. ArH), 8.67 (2H, br s, ArH); nmax(KBr)/cm-1 2190 (CN), 1580, 17 International T ables for X-Ray Crystallography, Birmingham 1510, 1480, 1410, 1390. Press. Birmingham, 1974, vol. IV. 18 S. R. Hall and J. M. Stewart, 1990 XTAL System, University of N,N¾-Dicyano-5,6,7,8-tetrafluoro-4,9-dihydronaphtho [2,3- Western Australia, Perth, Australia. b]thiophene-4,9-diimine 11. Compound 11 was obtained, by 19 J. M. Stewart, F. A. Kundell and J. C. Baldwin, The X-ray 76 Computer Science Center, Univ. of Maryland, College Park, stirring for 24 h at room temperature using 10 (0.17 mmol), Maryland, EEUU. TiCl4 (1.05 mmol) and BTC (1.19 mmol), in 90% yield; mp 20 J. Fayos and M. Martý�nez-Ripoll, HSEARCH, A computer pro- 194 °C (Calc. for C14H4F4N4S: C, 50.31; H, 0.60. Found: C, gram for the geometric calculations of H atom Coordinates, 50.14; H, 0.77%); dH(300 MHz, CDCl3) 8.09 (1H, d, J 5.2, Instituto Rocasolano, CSIC, Serrano 119, 28006, Madrid, Spain, thiophene), 8.80 (1H, d, J 5.2, thiophene); nmax(KBr)/cm-1 1978. 2190 (CN), 1580, 1510, 1400, 1300, 1270. 21 M. Nardeli, PARST, Comput. Chem., 1983, 7, 95. 22 Friedel–Crafts and Related Reactions, ed. G. A. Olah, New York, 1964. Copper salt of compound 8 23 R. Gonc�alves and E. V. Brown, J. Org. Chem., 1952, 17, 698. A boiling solution of N,N¾-dicyano-1,2,3,4-tetrafluoroanthraquinone diimine 8 (200 mg, 0.6 mmol) in 20 ml of dry aceto- Paper 6/01628F; Received 7thMarch, 1996 J. Mater. Chem., 1997, 7(1), 25&nda

ISSN:0959-9428    

DOI:10.1039/a601628f

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Synthesis of a new extended p-donor with 1,4-oxathiane annulation Jonas Hellberg,*a Karlis Balodis,a Mikael Moge,a Peter Koralla and J-U. von Schu� tzb aOrganic Chemistry, Royal Institute of T echnology, S-100 44 Stockholm, Sweden b3. Physikalisches Institut, Universita�t Stuttgart, D-70550 Stuttgart, Germany 4,5;4¾,5¾-Bis(1,4-oxathiane-2,3-diyldithio)tetrathiafulvalene 7 (TOET) is synthesised in four steps from 1,4-oxathiane.Compound 7 shows two reversible redox couples at 0.56 and 0.90 V vs. SCE. Analysis of NMR spectra combined with energy minimisation calculations indicated that at least three isomers of TOET should be present in solution. TOET forms a charge-transfer complex with TCNQ which shows a conductivity of 3×10-3 S cm-1. Tetrathiafulvalene p-donors for molecular metals and super- diyldithio)-1,3-dithiole-2-thione 5 in 38% yield.Thione 5 was converted to the corresponding carbonyl derivative 6 via conductors have attracted considerable research interest. Most successful cation radical systems are based on ET 1 as donor, treatment with mercuric acetate in acetonitrile. Dimerization to TOET 7 could be achieved by treating either the thione, or with various counter ions.1 Taking a simplified view, the pcore is responsible for hosting the conduction process, whereas the carbonyl, or a 151 mixture of the two, with trimethyl phosphite in benzene at reflux temperatures.Best yields (ca. the peripheral substituents act as guides for the docking of the donor moieties between the anions. Changing these side groups 40%) were achieved with either the carbonyl or the mixture, while the thione alone gave slightly less product.The isolated will therefore lead to new unforeseen packing modes in the solid state. Although the initial high hopes for oxygen substi- TOET showed a good elemental analysis, whereas the EI-MS conditions produced clean retro Diels–Alder reactions and tution2 has not been fulfilled, a strong tendency to form metallic compounds has been noted.3 therefore showed only the mass spectrum of 4.Proton NMR spectra in CDCl3 or [2H6]DMSO could only give an indication of the purity because of the insolubility of TOET. However, in [2H5]pyridine at 353 K, a reasonable spectrum could be recorded which showed a heavily coupled system with the individual protons well separated (Fig. 1). At this temperature the retro Diels–Alder reaction was too fast to allow a 13C Newly aroused interest in the 1,4-dioxane fused-ET (BDDT- NMR spectrum to be recorded, i.e. after 6 h ca. 50% had TTF) 2,4 together with the availability of 2,3-dihydro-1,4- reverted to starting material. Also apparent from the spectrum oxathiine 4 in our group, prompted us to synthesise the was the presence of at least two stereoisomers, since one of analogous 4,5;4¾,5¾-bis(1,4-oxathiane-2,3-diyldithio)tetrathia- the bridgehead methine protons was doubled.fulvalene (ThiOxane-ET or TOET), which we report here. At this point we found it necessary to try to estimate the The synthesis sequence is depicted in Scheme 1. isomeric purity of the TOET. If the cycloaddition step is Commercially available 1,4-oxathiane 3 was treated with sulfu- assumed to give a product with the vicinal hydrogens between ryl chloride in carbon tetrachloride and refluxed for 3 days to the oxathiane and dithiine ring arranged in a cis-fashion, the yield 2,3-dihydro-1,4-oxathiine 4 in reasonable yield after distil- number of theoretically possible stereoisomers is reduced from lation.Compound 4 is less stable than the corresponding dioxa sixteen to eight. Among these eight there are two pairs of derivative, but can be stored in a freezer for several months if identical molecules, which means that the final number of pure; impurities seem to initiate polymerisation. Reaction of 4 stereoisomers are six, namely the two non-chiral diastereomers with 1,3-dithiole-2,4,5-trithione oligomer was achieved in anti-trans A and syn-cis B, and the two pairs of enantiomers, refluxing dioxane to yield the cis-fused 4,5-(1,4-oxathiane-2,3- syn-trans C/D and anti-cis E/F, respectively.Rotation around the central double bond has been shown to occur even in Scheme 1 Reagents and conditions: i, SO2Cl2, CCl4, reflux, 1 week; Fig. 1 500 MHz 1H NMR spectrum of TOET 7 in [2H5 ]pyridine ii, dioxane, 1,3-dithiole-2,4,5-trithione oligomer, reflux, 18 h, 38%; iii, Hg(OAC)2 , MeCN; iv, (MeO)3P, benzene, reflux, 3.5 h at 353 K J.Mater. Chem., 1997, 7(1), 31–34 31weakly protic solutions in related systems, which should make A equilibrate with B, as well as C with E and D with F, respectively. Our nomenclature for the stereoisomers, and representations of these, are given in Fig. 2 and 3.A simulation of the coupling pattern present in the 1H NMR spectrum of the carbonyl compound 6, using coupling constants calculated from the global minimum conformation derived with the MACROMODEL program, was in good agreement with that obtained from the experimental spectrum (Fig. 4). The conformation of the global minimum is shown in Fig. 5. The second lowest conformation turned out to be 16 kJ mol-1 higher in energy (12 kJ mol-1 using AM1 calculations), indicating a highly populated global minimum. Starting with this conformation we energy-minimized the four different diastereomers of 7 (A, B, C/D, E/F) assumed to be present. The four conformations so achieved were thus all in firm energy-minima with no other conformations energetically available.The energy differences between the four stereo- Fig. 4 (a) Experimental and (b) calculated 1H NMR spectra for 6 isomers were negligible, the largest being 0.4 kJ mol-1. This energy difference is probably within the calculation error-limit, and would lead to only a few percent difference in concentration between two isomers in equilibrium.AM1 optimization of the four isomers gave energy differences very close to zero. The four energy-minimized diastereomers are shown in Fig. 6. Noteworthy is the strong tendency for the sulfur atom in the oxathiane ring to point outwards from the convex part of the molecule, whereas the oxygen atom in the same ring is on the concave side towards the planar p-system.The experimental 1H NMR spectrum for TOET 7 could be simulated, anticipating two stereoisomers with identical coupling systems but slightly different shifts. Good agreement was found for all protons using the coupling constants found for 6, with the two isomers present in approximately 253 ratio (see Fig. 7 for the proton around d 3.95). Since there do not seem to be any large differences in energy between the stereoisomers, the isomeric outcome is probably governed by solubility factors rather than thermodynamic Fig. 5 Energy-minimized structure of carbonyl derivative 6 (MACROMODEL, MM3*). Experimental coupling constants: H11–H12=12.1 Hz, H11–H21=2.3 Hz, H11–H22=11.9 Hz, H12–H21= 1.9 Hz, H12–H22=4.3 Hz, H21–H22=14 Hz, H3–H4=1.5 Hz. stabilities. Our work-up procedure relies on the crystallisation of TOET from the cooled reaction mixture, so there is probably enrichment of the less soluble stereoisomers at this stage.As it has been shown that proton-assisted rotation–isomerization around the central double bond is prevented by the addition Fig. 2 Definitions of different stereoisomers of TOET 7. (a) ‘S–S’ cis- of pyridine, we can anticipate that this isomerization reaction trans iomers formed through proton-assisted rotation around central is not present in our spectrum of TOET in pyridine.double bond. (b) ‘H–H’ syn-anti isomers formed during phosphite- So, without any definitive proof, is it then possible to say assisted dimerization of 5+6. (c) Only cis ‘H–H’ isomers are formed in the cycloaddition reaction to form 5.something about the stereoisomeric distribution in this system? Fig. 3 Different possible stereoisomers of TOET 7 32 J. Mater. Chem., 1997, 7(1), 31–34Well, if one believes that it is the less soluble isomer that is crystallising, we get some guidance from the fact that Kini et al.4b were able to isolate a cation radical salt of an anti isomf the analogous dioxaderivative BDDT-TTF.If the same situation is valid in our case, we can assume that the recorded spectrum is showing a mixture of isomers A and E/F. Cyclic voltammetry of TOET in benzonitrile showed two reversible redox couples at E1/2=0.56 and 0.90 V. These values are ca. 50 mV higher than ET 1 measured under the same conditions. Electrocrystallization of the new donor in benzonitrilein the presence of the anionsNO3 or PF6 gave microcrystalline red–brown powders of poor quality.We could also isolate a 151 charge transfer complex with TCNQ as a microcrystalline powder, with a rather low conductivity (3×10-3 S cm-1 measured on a compressed pellet). In conclusion, we have synthesised the new extended donor TOET in a short synthesis, and showed that there are at least two stereoisomers present in the crude reaction product.We are currently investigating alternative conditions for the electrocrystallization experiments in order to facilitate monoisomeric salts, but the difficulty of isolating pure isomers coupled with the inherent ease of isomerisation in these systems will probably prevent any extensive use of TOET as a donor for organic metals.Experimental NMR spectra were recorded on a Bruker AM 400 (400 MHz) or Bruker DMX 500 (500 MHz) spectrometer; J values are given in Hz. Mass spectra were recorded on a Finnigan SSQ 7000. THF was distilled from sodium–benzophenone, all other solvents were either distilled or HPLC grade. Cyclic voltammetry was carried out at 25°C in benzonitrile at 100 mV s-1 Fig. 6 The four different energy-minimized diastereomers of TOET 7 with Bu4NPF6 (0.15 M) as electrolyte and measured vs. a SCE (AM1 calculated) reference electrode. Electrocrystallizations were performed at ambient temperature, with a platinum wire anode of about 1 cm2. Currents were typically 1.0 mA cm-2 and the experiment was run for 1–2 weeks. Molecular mechanics modelling was carried out using the MACROMODEL program (ver. 4.5) using a Silicon Graphics Indigo 2 workstation. The following parameters were used in the conformational search on 7: force field, MM3*; minimization algorithm, truncated Newton-Raphson; derivative convergent criteria, 0.05 kJ A° -1 mol-1; permissible window above lowest energy conformation, 50 kJ mol-1; number of Monte Carlo steps, 1000. Vicinal proton coupling constants (3J) were calculated within the MACROMODEL program (using the method of Haasnoot et al.5).No Boltzmann averaging was performed since the energy difference of 16.2 kJ mol-1 between the global minimum and the second lowest conformer will result in an almost exclusive population of the former. Semiempirical calculations were performed using the AM1 parametrization as implemented in SPARTAN SGI ver. 4.0.3 GL, using global minima derived from MACROMODEL. 2,3-Dihydro-1,4-oxathiine 4 1,4-Oxathiane 3 (143,9 g, 1.38 mol) dissolved in carbon tetrachloride (1 l) was brought to reflux. Sulfuryl chloride (1.05 equiv. 196 g) dissolved in carbon tetrachloride (500 ml) was added dropwise to the refluxing solution over ca. 90 min. The solution was then refluxed for 3 days.The carbon tetrachloride was then evaporated under reduced pressure and the dark residue distilled under water aspirator vacuum. Pure compound 4 was collected at 70°C (lit.,6 54°C, 20 mmHg) in yields up to 73%. Yields varied considerably since polymerisation frequently occurred, especially at temperatures above 100°C. Attempts to inhibit this side reaction previous to distillation by various drying and/or neutralizing procedures seemed to Fig. 7 (a) Experimental and (b) calculated 1H NMR spectra for the proton at d 3.95 in TOET 7 have little or no effect on the yield. 1H NMR (400 MHz, J. Mater. Chem., 1997, 7(1), 31–34 33CDCl3) d 6.57 (d, J 6.5, 1H), 5.04 (d, J 6.5, 1H), 4.30 (2H, m), 4,5;4¾,5¾-Bis(1,4-oxathiane-2,3-diyldithio)tetrathiafulvalene 7 3.00 (2H, m); 13C NMR (126 MHz, CDCl3) d 140.2, 93.3, 65.3, Freshly distilled trimethyl phosphite (6 ml, 51 mmol) was 25.4; m/z 102 (100%, M+), 74, (84, M+-C2H4 ), 45 (67, added to a stirred suspension of thione 5 (0.62 g, 2.08 mmol) M+-C3H5O).and carbonyl 6 (0.57 g, 2.02 mmol) in benzene (35 ml), and the mixture was refluxed under nitrogen during 3.5 h. After 4,5-(1,4-Oxathiane-2,3-diyldithio)-1,3-dithiole-2-thione 5 cooling to room temperature, fine crystals were filtered off and washed with EtOH, acetone and diethyl ether.Attempts to A suspension of 2,3-dihydro-1,4-oxathiine 4 (4.2 g, 40.7 mmol) recrystallize the product led only to reversal of the Diels–Alder and 1,3-dithiole-2,4,5-trithione oligomer7 in 1,4-dioxane reaction. Comparable yields could be achieved when the same (75 ml) was refluxed for 18 h.The hot mixture was filtered and procedure was used with only one of the reactants 5 or 6; 1H the residue was washed with hot toluene. The solvent was NMR (500 MHz, [2H5]Pyridine) Isomer 1: d 5.72 (d, J 1.80, evaporated under reduced pressure and the resulting dark 1H), 4.86 (d, J 1.8, 1H), 4.32 (m, J1 12.10, J2 3.20, J3 3.20, 1H), brown sticky residue was dissolved in toluene and filtered 3.96 (m, J1 12.10, J2 10.50, J3 2.30, 1H), 3.20 (m, J1 14.00, J2 through a short column of silica gel to remove unreacted 10.50, J3 3.20, 1H), 2.38 (m, J1 14.00, J2 3.20, J3 2.30, 1H).oligomer. Evaporation of the toluene gave pure 5 (by NMR) Isomer 2: d 5.71, 4.80, 4.31, 3.97, 3.23, 2.34; m/z 102 (100%), (4.6 g 38%) as yellow–brown crystals.An analytically pure 74, (84) (Calc. for C14H12O2S10: C, 31.55; H, 2.27. Found: C, sample could be obtained by chromatography with a 251 31.67; H, 2.33%). hexane–CH2Cl2 gradient; mp 142–143 °C; 1HNMR (400 MHz, CDCl3) d 5.47 (d, J 1.50, 1H), 4.66 (d, J 1.5, 1H), 4.46 (m, J1 12.10, J2 3.20, J3 3.20, 1H), 4.07 (m, J1 12.10, J2 10.50, J3 2.30, References 1H), 3.31 (m, J1 14.00, J2 10.50, J3 3.20, 1H), 2.41 (m, J1 14.00, J2 3.20, J3 2.30, 1H); 13C NMR (126 MHz, CDCl3 ) d 208.8, 1 See for instance, J.M. Williams, J. R. Ferraro, R. J. Thorn, K. D. Carlson, U. Geiser, H. H. Wang, A. M. Kini and M- 124.6, 117.9, 78.0, 71.6, 44.3, 23.8; m/z 298 (17%, M+), 102 H. Whangbo, Organic Superconductors, Prentice Hall, 1992. (100, M+-C3S5) (Calc.for C7H6OS6: C, 28.16; H, 2.03. Found: 2 T. Suzuki, H. Yamochi, G. Srdanov, K. Hinkelmann and F. Wudl, C, 28.28; H, 2.09%). J. Am. Chem. Soc., 1989, 111, 3108; F. Wudl, H. Yamochi, T. Suzuki, H. Isotalo, C. Fite, H. Kasmai, K. Liou and G. Srdanov, J. Am. Chem. Soc., 1990, 112, 2461. 4,5-(1,4-Oxathiane-2,3-diyldithio)-1,3-dithiol-2-one 6 3 H. Yamochi, S. Horiuchi and G.Saito, Synth. Met., 1993, 55–57, Thione 5 (1.28 g, 4.3 mmol) was suspended in acetonitrile 2096. 4 (a) A. I. Kotov, C. Faulmann, P. Cassoux and E. Yagubskii, J. Org. (130 ml) and mercuric acetate (1.60 g, 5.0 mmol) was added. Chem., 1994, 59, 2626; (b) A. M. Kini, U. Geiser, H-H. Wang, The resulting mixture was refluxed for 3 h and then additional K. R. Lykke, J. M. Williams and C.F. Campana, J. Mater. Chem., mercuric acetate (0.8 g, 2.5 mmol) was added and the mixture 1995, 5, 1647. was stirred for an additional 10 min. The dark precipitate was 5 C. A. G. Haasnoot, F. A. A. M. de Leeuw and C. Altona, filtered off and washed withCH2Cl2 , and the combined filtrates T etrahedron, 1980, 36, 2783. were evaporated to give crude 6. Chromatography on silica 6 Compound 4 was first synthesised by W. E. Parham, I. Gordon and J. D. Swalen, J. Am. Chem. Soc., 1952, 74, 1824. Other approaches gel with a 154 hexane–CH2Cl2 gradient gave pure 6 as yellow are, A. H. Haubein, J. Am. Chem. Soc., 1959, 81, 144; N. de crystals (0.8 g, 66%). An analytically pure specimen could be Wolf, P. W. Henniger and E. Havinga, Recl. T rav. Chim. Pays-Bas, achieved via recrystallization from EtOH; mp 155–157 °C; 1H 1967, 86, 1227; C. Berglund and S-O. Lawesson, Ark. Kemi, 1963, NMR (400 MHz, CDCl3) d 5.47 (d, J 1.70, 1H), 4.63 (d, J 1.7, 20, 225. 1H), 4.47 (m, J1 12.10, J2 3.20, J3 3.20, 1H), 4.09 (m, J1 12.10, 7 O. Ya. Neilands, Ya. Ya. Katsens and Ya. N. Kreitsberga, Zh. Org. J2 10.50, J3 2.30, 1H), 3.32 (m, J1 14.00, J2 10.50, J3 3.20, 1H), Khim., 1989, 25, 658; J. Becher and N. Svenstrup, Synthesis, 1995, 215. 2.41 (m, J1 14.00, J2 3.20, J3, 2.30, 1H); 13C NMR (126 MHz, CDCl3) d 198.4, 114.6, 108.5, 79.2, 71.3, 45.1, 23.9; m/z 282 (10%, M+), 102 (100, M+-C3S5). Paper 6/03837I; Received 3rd June, 1996 34 J. Mater. Chem., 1997, 7(1), 31–34

ISSN:0959-9428    

DOI:10.1039/a603837i

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Perfect layered arrangement of ion-paired chromophores in a crystalline non-linear optical organic salt: 2-amino-3-nitropyridinium chloride Jean-Franc�ois Nicoud,*a Rene� Masse,b Cyril Bourgognea and Cara Evansa aGroupe desMate�riaux Organiques, Institut de Physique et Chimie desMate�riaux de Strasbourg, UM CNRS-UL P 380046, 23 rue du L oess, BP 20CR, 67037 Strasbourg Cedex, France bL aboratoire de Cristallographie, associe� a` l’Universite� Joseph Fourier, CNRS, BP 166, 38042, Grenoble-Cedex 09, France 2-Amino-3-nitropyridinium chloride is reported as a new crystalline non-linear optical organic salt, built around a twodimensional hyperpolarisable cationic chromophore.Its crystal structure (space group P21) reveals a quasi-perfect layered arrangement of the cations and the anions.Within each layer the chromophores are arranged in a herringbone structure and strong hydrogen bonds are present. The first hyperpolarisabilities of ion-paired chromophores calculated for 2-amino-3- nitropyridinium chloride as well as 2-amino-5-nitropyridinium chloride emphasize the favourable contribution of the inorganic anionic sublattice to the enhancement of the molecular bijk and macroscopic x(2) susceptibilities.In the field of materials for non-linear optics (NLO), there is position and the N+ pyridinium site in the ortho position relative to the amino donor group. This leads to a two- still an interest in the engineering of crystalline materials with dimensional (2D) hyperpolarisable chromophore. Two-dimen- satisfactory quadratic NLO properties. Several efficient inorsional charge transfer molecules have been receiving greater ganic, organic and organomineral NLO crystals have been attention as NLO chromophores, as shown by the recent proposed in recent years; some of these have even reached a review paper by Nalwa et al.4 We thought that it might be high level of development.However, several problems concernpossible to capitalize on the two-dimensional character of the ing the overall quality of these materials remain to be solved, hyperpolarisable ionic chromophore by using the isomeric 2- and more research is needed to improve their optical as well amino-3-nitropyridinium cation (2A3NP+) as a non-linear as mechanical properties.1,2 The engineering of crystalline optical chromophore.In this cation the amino donor group is materials for quadratic non-linear optics is guided by two between two different acceptor sites (the nitro group and the main requirements: N+ pyridinium site), and charge transfer is possible in two (i) Ideally, non-linear optical chromophores self-assemble in very different directions, as shown in Fig. 1(b). To the best of a non-centrosymmetric structure in such a way that the our knowledge, no crystalline NLO material built from this contribution of hyperpolarisabilities bijk of individual chromotype of chromophore has been previously reported.Before any phores results in high macroscopic tensor components x(2). investigation, the following behaviour might be expected from This double goal of ‘non-centrosymmetry–efficiency’ is difficult this new cationic chromophore: to obtain.The engineering is often focused on non-centrosym- (i) The anchorage of the 2A3NP+ cation onto an anionic metric materials from a given hyperpolarisable chromophore. host matrix will occur through short hydrogen bonds originat- The chromophores are modified or oriented in host matrices ing from the NH2 and NH+ groups as observed experimentally in order to improve the efficiency of the x(2) tensor.in all the 2-amino-5-nitropyridinium salts.5 (ii) A qualified material must also be designed for a precisely (ii) Intra-cation hydrogen bonds should be present in such targeted application: electrooptical modulation (optimisation structures because the nitro and amino groups are located of rijk), frequency doubling, or optical parametric oscillation ortho to one another, as clearly shown by Panunto et al.in a (optimisation of xijk and phase-matching conditions). All the study of hydrogen bond patterns of nitroaniline derivatives.6 structural modifications which act on the robustness (bond We now report our successful use of this new crystal strengths), thermal conductivity (damage threshold) and refrac- engineering strategy. 2-Amino-3-nitropyridinium chloride tive indices of a non-linear optical crystal can be presently (2A3NPCl), a relatively simple crystalline 2A3NP+ salt pre- directed by engineering based on the modification of the host matrix of chromophores. Generally, purely organic compounds lack mechanical strength for practical uses, although the electric susceptibility x(2) is often quite high relative to that of inorganic materials.In addition, many difficulties are encountered in growing single crystals of sufficient quality. That is why recently a new crystal engineering strategy, combining mineral and organic moieties, has been proposed with the aim of building more cohesive crystalline structures. The goal of this strategy was to mix the advantages of inorganic ionic structures (cohesion, optical and other damage resistance) with those of organic molecules (structural flexibility, high hyperpolarisability).Thus, numerous 2-amino-5-nitropyridinium (2A5NP+) organic salts, in which the 2A5NP+ cation is anchored onto inorganic or organic host matrices, have been Fig. 1 Mesomeric forms showing the two-dimensional character of the reported for their high NLO efficiency.3a,b The 2A5NP+ cation charge transfer in the cationic chromophores (a) 2A5NP+ and (b) 2A3NP+ has two electron-accepting centres, the nitro group in the para J.Mater. Chem., 1997, 7(1), 35–39 35pared from 2-amino-3-nitropyridine, gives an intense second gram.9 Full-matrix least-squares refinements were performed on F: the function minimized was SvFo-Fc with a unitary harmonic signal at 530 nm by the powder test of Kurtz and Perry7 when illuminated by Nd3+–YAG laser light at 1.06 mm weighting scheme.Scattering factors for neutral atoms and f ¾, Df ¾, f , Df were taken from International T ables for X-ray [I2v observed >I2v (POM=3-methyl-4-nitropyridine-1- oxide)8].In order to elucidate the origin of this strong NLO Crystallography.10 All calculations were carried out using the Enraf-Nonius SDP program11 operating on a micro-Vax II effect, the crystal structure of 2A3NPCl has been investigated. Since the experimental values of the molecular hyperpolarisa- computer. The structure was drawn using the MOLVIEW program.12 The main geometrical features of the 2-amino-3- bilities for 2A3NP+ and 2A5NP+ are not yet known, we have also investigated theoretical values for their bijk tensor compo- nitropyridinium cation are described in Table 2, and the hydrogen bonds are described in Table 3.nents. In addition, we calculated bijk for the ion pairs 2A3NP+ Cl- and 2A5NP+ Cl- in order to see how the overall Atomic coordinates, thermal parameters, and bond lengths and angles have been deposited at the Cambridge Cryst- hyperpolarisability is affected by the presence of the counterion.allographic Data Centre (CCDC). See Information for Authors, J. Mater. Chem., 1997, Issue 1. Any request to the CCDC for Experimental this material should quote the full literature citation and the reference number 1145/18.Crystal structure 2-Amino-3-nitropyridine (0.01 mol) is easily dissolved in aque- NLO Chromophores ous acidic solution (20 cm3 containing 0.025 mol HCl) at 20°C. The solution, slowly evaporated at room temperature, The UV–VIS spectrum of a low concentration solution of the organic salt 2A3NPCl in ethanol shows an intense character- yields pale yellow crystalline platelets up to 3×2×5 mm in size.The crystallization of 2-amino-3-nitropyridinium chloride istic charge transfer absorption band with lmax=385 nm. This absorption, due to the charge transfer shown in Fig. 1, is occurs when only 2 cm3 of solvent remains, indicating a high solubility of the salt in HCl solutions. The chemical formula responsible for the yellow colour of the 2A3NPCl material and is similar to that observed for classic 4-nitroaniline deriva- was establie X-ray crystal structure investigation.The P21 space group of 2-amino-3-nitropyridinium chloride tives. In order to understand better the highly efficient second harmonic signal in the new crystalline material and compare was confirmed by the unique limiting condition (0k0 with k= 2n) and the high second harmonic signal in the Kurtz and with the already known 2A5NPCl material, we calculated the first hyperpolarisabilities for several species.The calculation Perry powder test. The cell parameters, space group and crystal structure were determined from single crystal X-ray diffraction data obtained Table 2 Selected interatomic distances and bond angles in the 2A3NP+ with a four circle diffractometer.Crystal data, experimental chromophore conditions, and structural refinement parameters are described bond distances/A° in Table 1. No absorption correction was applied; only Lorentz N(1)–C(1) 1.357(3) and polarization effects were taken into account. The structure N(1)–C(5) 1.338(4) was solved by direct methods using the MULTAN 77 pro- N(1)–H(1) 0.86(4) N(2)–C(1) 1.324(4) N(2)–H(2) 0.79(2) Table 1 Crystal data, intensity measurements and structural refinement parameters for 2-amino-3-nitropyridinium chloride N(2)–H(3) 0.89(4) N(3)–C(2) 1.459(3) O(1)–N(3) 1.233(3) formula C5H6ClN3O2 molecular weight 175.58 O(2)–N(3) 1.209(4) C(1)–C(2) 1.407(4) space group P21 a/A° 6.496(1) C(2)–C(3) 1.360(4) C(3)–C(4) 1.380(4) b/A° 9.040(4) c/A° 7.263(1) C(3)–H(4) 0.96(4) C(4)–C(5) 1.366(5) b (°) 116.6(9) V /A°3 381.3(3) C(4)–H(5) 0.80(3) C(5)–H(6) 1.03(4) Z, Dx/g cm-3 2, 1.529 unit-cell refinement 25 reflect (10<h<12.8°) bond angles/degrees F(000) 180 m/cm-1 2.417 (l Ag-Ka) C(1)–N(1)–C(5) 124.3(3) C(1)–N(1)–H(1) 107(2) crystal size/mm 0.25×0.30×0.40 temperature/K 293 C(5)–N(1)–H(1) 128(2) C(1)–N(2)–H(2) 127(2) apparatus Nonius CAD4 monochromator graphite (220) C(1)–N(2)–H(3) 114(2) H(2)–N(2)–H(3) 119(3) radiation/A° 0.5608 (Ag-Ka) Bragg angle limits (°) 3–30 O(1)–N(3)–O(2) 123.6(3) O(1)–N(3)–C(2) 119.0(3) h,k,l, limits (-11,11; 0,16; 0,12) scan technique v scan O(2)–N(3)–C(2) 117.3(2) N(1)–C(1)–N(2) 116.4(2) background/s 5–22 scan speed/deg s-1 0.025 to 0.111 N(1)–C(1)–C(2) 114.8(2) N(2)–C(1)–C(2) 128.7(2) scan width (°) 1.10 control reflections (0 4 3), (0 -4 3) N(3)–C(2)–C(1) 119.8(2) N(3)–C(2)–C(3) 117.9(2) period between intensity measurements/s 7200 reflections between orientations 400 C(1)–C(2)–C(3) 122.3(2) C(2)–C(3)–C(4) 119.5(3) reflections collected 2518 unique data 1150 C(2)–C(3)–H(4) 116(2) C(4)–C(3)–H(4) 124(2) data used in refinement 754 [I>2s(I)] refined parameters 123 C(3)–C(4)–C(5) 118.7(3) C(3)–C(4)–H(5) 122.3(3) R (Rw) 0.027 (0.026) weighting scheme unitary C(5)–C(4)–H(5) 118.4(3) N(1)–C(5)–C(4) 120.2(3) goodness-of-fit 0.262 largest shift/error 0.30 N(1)–C(5)–H(6) 110.8(3) C(4)–C(5)–H(6) 128.7(3) max residual density/e A° -3 0.18 36 J.Mater. Chem., 1997, 7(1), 35–39Table 3 Hydrogen bonds and angles in the 2A3NPCl crystal structure Table 4 Static bijk components (in 10-30 esu) calculated by the finite- field method using AM1 parameters, based onthe optimizedgeometries and angles around the Cl- anion for neutral molecules and isolated cations and on the actual conformations in the crystals for the ion pairs bond lengths/A° N(1)–H(1) 0.86(4) Cl–H(1) 2.15(4) bxxx bxyy byyy byxx Cl–N(1) 3.016(2) N(2)–H(2) 0.79(2) O(1)–H(2) 2.19(2) O(2)–H(2) 2.27(2) 4.1 2.0 1.4 1,8 N(2)–O(1) 2.658(3) N(2)–O(2) 3.038(4) N(2)–H(3) 0.89(4) H(3)–Cl 2.55(3) Cl–N(2) 3.364(2) C(3)–H(4) 0.96(4) H(4)–Cl 2.70(4) H(4)–O(2) 2.28(4) 13.3 2.3 0 0.5 C(3)–O(2) 2.654(4) C(3)–Cl 3.492(3) C(4)–H(5) 0.80(3) C(5)–H(6) 1.03(4) H(6)–Cl 2.74(4) Cl–C(5) 3.485(3) 1.2 1.7 0.5 1.1 bond angles/degrees Cl–H(1)–N(1) 177(4) O(1)–H(2)–N(2) 119(2) O(2)–H(2)–N(2) 162(2) Cl–H(3)–N(2) 151(3) Cl–H(4)–C(3) 139(3) O(2)–H(4)–C(3) 102(3) 3.4 1.6 0.5 0.7 Cl–H(6)–C(5) 128(3) adjacent angles around the Cl anion/degrees H(1)–Cl–H(3) 47.8 H(3)–Cl–H(4) 101.7 H(4)–Cl–H(6) 91.4 24.4 2.6 3.4 4.7 H(6)–Cl–H(1) 119 was performed for the neutral pyridines (2A3NP and 2A5NP), the protonated pyridines (2A3NP+ and 2A5NP+) and the ion pairs corresponding to the organic salts (2A3NP+Cl- and 2A5NP+Cl-), as they are positioned in the crystals. The geometries of the pyridines and the pyridiniums, with a Cs 24.2 4.4 4.5 7.2 imposed symmetry, were optimized by ab initio DFT methods in DMOL ver. 3.00 software from BIOSYM. The b tensor components were then computed on these optimized structures by using the semi-empirical AM1 parameters and the finite field method available in MOPAC 6.13 The results for the most significant values of b (in 10-30 esu units) are summarized in Table 4.All the chromophores are planar, so the z axis is not another,6 precludes the rotation of the nitro group with respect drawn. The orientation of each calculated species in the xy to the pyridinium ring. The dihedral angle between the planes plane is given in Table 4.of the NO2 group and the heterocycle is 0.3°, indicating a coplanar geometry. Only an approximate figure can be given Discussion for the dihedral angle between the plane of the NH2 group and the pyridinium plane because the H atoms are located Crystal structure less accurately than C, N, O, or Cl atoms by X-ray diffraction; the value is ca. 4.2° for 2A3NPCl. Distortion or twisting of The cations and anions of 2A3NPCl are packed in layers parallel to the crystallographic plane defined by the b and the NH2 and NO2 groups modifies the efficiency of the intramolecular charge transfer and consequently the values of a+2c vectors (Fig. 2). These planes of ions are perpendicular to the direction defined by the a vector and intersect the unit bijk and xijk.In structures built with the 2-amino-5-nitropyridinium cations a twisting of the NO2 group under the influence cell at a/4 and 3a/4. The strongest hydrogen bonds form within the layer; no long H-bonds are detected between adjacent of C–H,O bonds of neighbouring cations has always been observed. For instance, the twisting angle of the nitro group planes separated by 3.25 A° .The screw axes located at (a/2,0,c/2) and (0,0,c/2) act on the crystallographic motif so that the is 3.6° in 2-amino-5-nitropyridinium acetophosphonate,14 7(1)° in 2-amino-5-nitropyridinium monohydrogenphosphite,15 and chromophores arrange in a herringbone structure (Fig. 3). The anion charge is balanced through N–H,Cl and C–H,Cl as large as 16.7° in 3-methyl-4-nitropyridine 1-oxide.16 The second harmonic generation (SHG) signal at 530 nm is strong, hydrogen bonds (Table 3).The unique intercation contact N(2)–H(2),O(2) induces the aggregation of chromophores which implies that phase-matching conditions occur for at least some xijk. Furthermore, because of the layered structure, in zigzag chains. The three-centre hydrogen bonds C(3)–H(4),O(2), Cl and N(2)–H(2),O(1), O(2) involve high electrooptical coefficients rijk are expected in the (-1 0 2) plane.If X, Y and Z are the dielectric axes with X along a, Y two intra-cation links. This situation, observed in nitroaniline derivatives in which nitro and amino groups are ortho to one along b, and Z belonging to the (-1 0 2) plane perpendicular J. Mater. Chem., 1997, 7(1), 35–39 37Fig. 3 Herringbone structure of cations viewed in a plane intersecting the a direction at a/4. Hydrogen bonds, except the N(2)–H(2)··O(2) contact which induces the cation aggregation, are shown. Fig. 2 Perfect layered arrangement of ion pairs 2A3NP+Cl- in planes parallel to the crystallographic plane (-1 0 2) arisabilities of the ion pairs 2A3NP+Cl- and 2A5NP+Cl-, keeping the same conformations and positions of the ions as to a, the indices of refraction are expected to rank as in the crystals.20 Both the ion pairs have higher theoretical nY$nZ&nX.bxxx values (24.4 and 24.2×10-30 esu, respectively) than 2A3NP+ and 2A5NP+, the x axis being the chloride–nitro Non-linear optical properties axis. According to our calculations it appears that, in this type Calculations on the 2A3NP neutral molecule indicate that a of non-linear organic salt, the chloride anion may make a more pronounced two-dimensional character of the hyperpol- significant contribution to the hyperpolarisability of the cat- arisability exists for this species than for 2A5NP. The latter ionic chromophore and hence to the x(2) of the material.In has a high value of bxxx corresponding to the 2-amino–5-nitro our case we should consider the actual chromophore to be the charge transfer along the x axis, as shown in Fig. 1(a). The ion pair 2A3NP+Cl-, for which the Cl-–NO2 axis is nearly calculated value of bxxx corresponds well to the already coincident with the NH2–NO2 axis. reported static b(0) value of (13.9±0.5)×10-30 esu, extrapolated to zero frequency from the electric field induced second Conclusion harmonic generation (EFISHG) measurement at 1.06 mm.17 When the amino donor group is placed between the nitro We have investigated the 2-amino-3-nitropyridinium cation group and the pyridine nitrogen atom, the charge transfer and other two-dimensional charge transfer chromophores as a described in Fig. 1(b) leads to off-diagonal components of b new family of non-linear optical organic salts. 2-amino-3- that can reach nearly 50% of the bxxx value. This is in nitropyridinium chloride (2A3NPCl) crystals are highly accordance with the results recently reported for the diagonal efficient for the second harmonic generation of laser light at and off-diagonal components of hyperpolarisability, bxxx and 1.06 mm. The analysis of the crystal structure reveals a perfect bxyy respectively, of tri- and tetra-substituted benzene with C2V layered arrangement of the chromophores in layers parallel to symmetry.18 For molecules with the p-conjugation extended the crystallographic plane (-1 0 2).As shown by theoretical by three substituents, bxyy is enhanced when one donor is calculations, within the layer each chloride anion forms an ion present with two acceptors located in ortho rather than meta pair with a neighbouring pyridinium cation, leading to an positions.enhanced hyperpolarisability of the whole. The possible mutual As for the protonated molecules, the calculations gave interaction between organic and inorganic sublattices in non- relatively low values for their hyperpolarisabilities. We notice, linear organic salts deserves further attention and could lead however, a value of the off-diagonal component bxyy for to a new crystal engineering strategy for the design of new 2A3NP+ which is higher than bxxx. For crystalline materials efficient organic NLO materials.built with such chromophores, we cannot easily deduce their NLO efficiency from the orientation of the molecules in the Work at Laboratoire de Cristallographie, CNRS, Grenoble lattice.The classical one-dimensional analysis is no longer was sponsored by the Groupement de Recherche ‘Mate�riaux valid, and optimum values of the NLO coefficients as a pour l’Optique Non-line�aire’ (No. 1181), CNRS. C.E., on leave function of the orientation of molecules with respect to the from Department of Chemical Engineering and Materials two-fold axis is now dependent on the ratio bxyy/bxxx, as was Science, University of Minnesota, acknowledges support from discussed by J.Zyss et al. in their work concerning 2-amino- a National Science Foundation Graduate Fellowship. 5-nitropyridinium hydrogen L-tartrate.19 The low values which we calculated appeared inconsistent with the high second References harmonic signal of 2A3NPCl crystalline powder.We then investigated the actual contribution of the anionic sublattice 1 Nonlinear Optical Properties of Organic Molecules and Crystals, to the NLO properties of the material, since the anions and eds. D. S. Chemla and J. Zyss, Academic Press, Orlando, 1987, vol. I and II. the cations lie in the same plane.We calculated the hyperpol- 38 J. Mater. Chem., 1997, 7(1), 35–392 Molecular Nonlinear Optoelectronics: Materials, Physics and 12 MOLVIEW, J.-M. Cense, T etrahedron Comput. Method., 1989, 2, Devices, ed. J. Zyss, Academic Press, San Diego, London, 1994. 65. 3 R. Masse, M. Bagieu-Beucher, J. Pecaut, J-P. Le�vy and J. Zyss, 13 H. A. Kurtz, J. J. P.Stewart and K. M. Dieter, J. Comput. Chem., Nonlinear Optics, 1993, 5, 413. 1990, 11, 82. 4 H. S. Nalwa, T.Watanabe and S. Miyata, Adv.Mater, 1995, 7, 754. 14 J. Pe�caut and R. Masse, J. Mater. Chem., 1994, 4, 1851. 5 R. Masse, Nonlinear Optics, 1995, 9, 113. 15 J. Pe�caut and M. Bagieu-Beucher, Acta Crystallogr., Sect. C, 1993, 6 T. W. Panunto, Z. Urbanczyk-Lipkowska, R. Johnson and M. C. 49, 834. Etter, J. Am. Chem. Soc., 1987, 109, 7786. 16 M. Shiro, M. Yamakawa and T. Kubota, Acta Crystallogr., Sect. B, 7 S. K. Kurtz and T. T. Perry, J. Appl. Phys., 1968, 39, 3798. 1977, 33, 1549. 8 J. Zyss, D. S. Chemla and J-F. Nicoud, J. Chem. Phys., 1981, 17 Z. Kotler, R. Hierle, D. Josse, J. Zyss and R. Masse, J. Opt. Soc. 74, 4800. Am. B, 1992, 9, 534. 9 MULTAN 77, P. Main, L. Lessinger,M. M.Woolfson,G. Germain 18 M. Tomonari, N. Ookubo and T. Takade, Chem. Phys. L ett., 1995, and J.-P. Declercq, A system of computer programs for the auto- 236, 475. matic solution of crystal structures from X-ray diffraction data, 19 J. Zyss, R. Masse, M. Bagieu-Beucher and J-P. Le�vy, Adv. Mater., Universities of York, England and Louvain-La Neuve, Belgium, 1993, 5, 120. 1977. 20 J. Pe�caut, J. P. Le�vy and R. Masse, J.Mater. Chem., 1993, 3, 999. 10 R. Steward, E. R. Davidson and W. T. Simpson, International T ables for X-ray Crystallography, The Kynoch Press, Birmingham, vol. IV, Table 2-2C. Paper 6/05836A; Received 27th August, 1996 11 Structure Determination Package RSX11M, Enraf-Nonius, Delft, The Netherlands, 1979. J. Mater. Chem., 1997, 7(1)

ISSN:0959-9428    

DOI:10.1039/a605836a

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

High-pressure differential thermal analysis study of the phase behaviour in some tert-butyl compounds pivalic acid 2-methylpropane-2-thiol and tert-butylamine Jo�rg Reuter,a Dirk Bu�sing,a Josep Ll. Tamarit*b and AlbertWu� rflingera aInstitute of Physical Chemistry II Ruhr-University D-44780 Bochum Germany bDepartament de Fý�sica i Enginyeria Nuclear Universitat Polite`cnica de Catalunya Diagonal 647; 08028 Barcelona Catalonia Spain Pressure–temperature phase diagrams of the tert-butyl compounds (ButX) pivalic acid (X=CO2H) 2-methylpropane-2-thiol (X=SH) and tert-butylamine (X=NH2) have been determined at temperatures between 100 K and the melting curve and up to 300 MPa with the aid of differential thermal analysis (DTA) under pressure. New normal as well as high-pressure induced phases have been found for 2-methylpropane-2-thiol and tert-butylamine.The results for the studied compounds are compared to those previously obtained for compounds belonging to the same tert-butyl family (i.e. X=Cl Br CH2OH NO2 ,Me and CN) in order to establish similar trends in the phase behaviour as well as in the thermodynamic properties of the phase transitions. For the related tert-butyl compounds with the exception of the X=CN compound the average of the slopes (dT/dp) of the melting and the II to I transition curves have been determined to be (0.64±0.14) and (0.27±0.13) K MPa-1 respectively. tert-Butyl compounds ButX (X=Cl Br NO2 SH NH2 CO2 H (pivalic acid PA) X=SH (2-methylpropane-2-thiol TBT) and X=NH2 (tert-butylamine TBA). The results will be CO2H Me CH2OH CN etc.) are typical representatives of analysed in comparison with the previous findings for similar plastic crystals due to their almost globular shape.1 Such compounds such as X=Cl (tert-butyl chloride TBC),21 X=Br compounds display a rich polymorphism that has stimulated (tert-butyl bromide TBB),25 X=NO2 (2-methyl-2-nitropro- a large number of experimental investigations as well as pane TBN),16,26 and X=CH2OH (neopentyl alcohol NPA),27 theoretical studies in order to rationalize the thermodynamic as well as some results obtained from the literature for X= properties and molecular dynamics of the various solid Me (neopentane NP) and X=CN (tert-butyl cyanide phases.2–12 Guthrie and McCullough13,14 have tried to derive TBCN).28 the entropies of transition from symmetry and steric considerations.They assumed ten distinguishable orientations for tertbutyl compounds in the high-temperature phase corresponding Experimental to DS=Rln10 in approximate accordance with the experimen- Differential thermal analysis tal entropy change of the transition from the ordered lowtemperature phase into the plastic phase.The method of The experimental device used for the high-pressure differential finding the number of orientations used by Guthrie and thermal analysis (DTA) measurements has been described McCullough however was heavily questioned by Clark et al.15 elsewhere.29,30 The measurements were performed in closed Also the volume-dependent part of the entropy change has to indium capsules. be taken into consideration as has been shown in a recent The limit of experimental error for the transition tempera- paper.16 Furthermore the orientational disorder is not neces- tures determined from the DTA curves (usually on heating sarily characterised by a discrete number of distinguishable runs at a rate of 1 K min-1) is less than 0.5 K.The error for orientations (Frenkel model). Results from incoherent quasi- the pressure generated by compressed argon and measured by elastic neutron scattering suggested an (almost) isotropic tum- using Bourdon gauges is expected to be less than 0.5 MPa. bling motion for many molecules in their plastic phases.5,17,18 Also computer simulations discount the Frenkel model in Materials many cases.19 Pivalic acid obtained from Aldrich (99%) was distillated It is interesting to note that there are many substances under reduced pressure and dried with molecular sieves.A which exhibit pressure-induced disordered phases. Examples sample with purity higher than 99.85% (GC) was attained. 2- can also be given for tert-butyl compounds e.g. in tert-butyl Methylpropane-2-thiol and tert-butylamine were purchased chloride (X=Cl).20–22 from Aldrich and Fluka with purities higher than 99 and The pressure–temperature behaviour of plastic crystals can 99.5% respectively. The substances were used without further qualitatively be explained by the Pople–Karasz theory,23 which purification. was later extended by Amzel and Becka.24 The essential parameter of this theory is the ratio of two energy barriers n=Er/Ed Er and Ed being the barriers against reorientation Results and diffusion respectively. The values of such a parameter in Common features a homologous series would be a measure for the variation of the anisotropy of the molecular shape.From this point of view The details of the measurements on particular substances can a comparison of the polymorphism for similar plastic crystals be found in the thesis of Reuter31 for TBN TBT TBA the can elucidate the nature of the phase situation. thesis of Wilmers32 for TBC the thesis of Kreul33 for TBB and Here we report on the pressure dependence of the phase NPA and in the diploma thesis of Bu�sing34 for PA. behaviour of some selected tert-butyl compounds ButX estab- Enthalpy changes at atmospheric pressure were calculated lished with the aid of differential thermal analysis under from the peak areas of the DTA curves whenever they were not available from the literature.pressure; particularly we have studied the compounds X= J. Mater. Chem. 1997 7(1) 41–46 41 The transition temperatures determined at the half height quasielastic neutron scattering,35 (QNS),39 NMR,40–45 X-ray diffraction,46 dielectric methods47 and Raman spectroscopy.48 of the DTA peaks after the onset of the phase transition as At atmospheric pressure PA melts at 308.3 K from the a function of pressure were fitted by polynomials. Volume orientationally disordered phase (I). Below 279.8 K it trans- changes were derived from the enthalpy changes and the slopes forms into an ordered solid form. In phase I the molecules of the transition lines using the Clausius–Clapeyron equation. which are known to form dimers display overall molecular This set of thermodynamic values together with the associated tumbling while in phase II two types of motions are present entropy changes are collected in Table 1.Fig. 1 contains the methyl group reorientations about the methyl axes (C3) and pressure–temperature phase diagrams of the compounds stud- reorientations of the tert-butyl groups about the C–CO2H axis ied in this work (PA TBT and TBA) together with those (C3¾).40,41,45,46 obtained in recent years (TBC TBB NPA and TBN) as well The pressure–temperature phase diagram of PA derived as those obtained from the literature (NP and TBCN). from DTA measurements34 is shown in Fig. 1( f ). The slope of the melting transition line (0.78 K MPa-1) differs significantly Pivalic acid (ButCO2H) from that given in the previous work of Hasebe et al. Pivalic acid is a well-known plastic crystal which has been (0.45 K MPa-1),45 who used NMR under pressure.Nevertheless the value for the slope of the coexistence line of thoroughly studied by means of several techniques such as Table 1 Thermodynamic properties of the tert-butyl compounds compound ButX PA TBT TBA TBC TBB NPA TBN NP TBCN parameter phase transition (X=CO2H (X=SH (X=NH2) (X=Cl) (X=Br) (X=CH2OH) (X=NO2) (X=Me) (X=CN) T /K I�l 308.7 274.4a 205.2 248.4c 256.2c 331.4f 299.2g 257i 292.1k II�I 279.8 199.4a 201.3 219.4c 231.5c 236.2f 260.1g 140i 232.7k III�II 157.0a 197.5 183.1c 208.7c 215.3g 213.0k IV�III 151.6a 148.3 DH/kJ mol-1 I�l 2.1 2.5a 0.9b 2.0c 2.0c 3.9f 2.6g 3.3i 9.3k II�I 8.6 1.0a 5.6 5.7c 1.1c 4.0f 4.7g 2.6i 1.9k III�II 0.7a ~0.45 1.9c 5.7c 4.2g 0.2k IV�III 4.1a ~0.45 DS/J mol-1 K-1 I�l 6.8 9.0a 4.3b 8.0c 7.5c 11.8f 8.7g 12.7i 31.8k II�I 30.7 4.9a 27.8 25.8c 4.6c 16.9f 17.9g 18.4i 7.8k III�II 4.1a ~2.3 10.2c 27.2c 19.6g 1.1k IV�III 26.86 4.6d 3.9e 7.9f 5.0h 9.9j 11.5l II�I 8.8 1.8 5.2 4.4d 1.8e 4.0f 2.5h 5.5j III�II 1.0 0.4 1.8d 5.4e 5.0h IV�III 2.7 (dT/dp)/K MPa-1 I�l 0.78 0.65 0.61 0.57d 0.51e 0.67f 0.57h 0.78i 0.36i II�I 0.35 0.37 0.18 0.17d 0.40e 0.24f 0.14h 0.30i III�II 0.21 0.17 ~0.18d 0.20e 0.26h IV�III 0.10 0.20 aRef.13. bRef. 35. cRef. 36. dRefs. 21 and 22. eRef. 25. fRef. 27. gRef. 37. hRefs. 16 and 26. iRef. 28. jCalculated from ref. 28. kRef. 38. lCalculated from refs. 28 and 38. Fig. 1 Pressure–temperature phase diagrams of selected tert-butyl compounds (a) TBT (X=SH) (b) TBA (X=NH2) (c) TBC (X=Cl) (d) NP (X=Me) (e) TBB (X=Br) (f) PA (X=CO2H) (g) NPA (X=CH2OH) (h) TBN (X=NO2) and (i ) TBCN (X=CN) 42 J.Mater. Chem. 1997 7(1) 41–46 Table 2 Transition temperatures of PA (X=CO2H) as a function of pressure [T/K=a+b×(p/MPa)+c×(p/MPa)2] transition a/K b/K MPa-1 c/104 K MPa-2 II–I 279.7±0.8 0.347±0.015 1.53±0.56 I–l 308.1±0.9 0.784±0.044 8.21±4.15 phases II and I (0.35 K MPa-1) is closer to the previous value (0.31 K MPa-1).45 The coefficients of the fitted polynomials for the transition lines of PA are shown in Table 2. The temperature range of phase I is enlarged with increasing Fig. 3 DTA traces for TBT (X=SH) at 50 MPa (a) without and (b) pressure a result that is known to be the normal behaviour of with detour showing the transitions III–III* and III*–II (see text) plastic crystals. 2-Methylpropane-2-thiol (ButSH) the pressure to values lower than 150 MPa (ii) cooling to temperatures lower than 140 K and (iii) increasing the pressure It has been shown earlier49 that TBT displays three solid–solid again.It should be mentioned that it seems probable that phase transitions at about 152 157 and 199 K in addition to phase V should also be obtainable at pressures above 200 MPa the melting process at 274 K. On decreasing temperature there by increasing the annealing time appropriately. are then four different solid phases denoted as solid I II III Table 3 summarizes the determined coexistence lines for the and IV. The dominant motions are overall molecular tumbling different transitions their pressure ranges as well as the in the liquid phases I II and III and reorientation of the tert- coefficients of the fitted polynomials.According to the obtained butyl group in solid phase IV.50–53 results two triple points appear in the phase diagram (i) III– III*–II at (160±5) MPa and (192±1) K and (ii) III–IV–V at Normal pressure measurements. The DTA traces of TBT at (99±5) MPa and (161±1) K. normal pressure (0.1 MPa) for a heating run clearly show the mentioned sequence of the phases from IV to liquid. On the tert-Butylamine (ButNH2 ) other hand as has been previously reported in the literature the above sequence is not observed on cooling where phase So far the phase behaviour of TBA has been the subject of II transforms directly to phase IV. not more than two studies.35,54 Finke et al.,35 who measured the heat capacities of TBA between 12 and 340 K in an High pressure measurements.The pressure–temperature adiabatic calorimetry study found two solid–solid transitions. phase diagram of TBT [Fig. 1(a)] shows two new high-pressure The first transition at 91.30 K was described by the authors induced phases (III* and V). The temperature ranges of the as a second-order or lambda transition. The second solid– plastic phases II and I increase with increasing pressure as was solid transition being of first order was found at 202.27 K observed in the case of pivalic acid (PA). For values of pressure (DS=29.9 J mol-1 K-1) and the melting transition was higher than 75 MPa a peak is observed in the thermograms detected at a temperature of 206.19 K (DS=4.28 J mol-1 K-1). that indicates the transition from phase III to phase III* In a dielectric study no evidence for the existence of a plastic (Fig.2) while at a lower pressure the transition III–III* is not phase was found,54 probably due to the use of temperature observed which means that phase IV transforms directly into intervals of about 10 K while phase I extends over only 3.9 K. phase III* which transforms into phase II at higher temperature. In order to obtain the phase III* at pressures lower than Normal pressure measurements. The second-order transition 75 MPa the transition line IV–III/III* must be crossed at at 91.3 K was not detected in our measurements due to the pressures higher than 75 MPa and then the pressure has to be limited sensitivity of the equipment. Fig. 4 shows a DTA diminished in phase III until the desired value is reached. Such thermogram of TBA that was measured at normal pressure.a detour process allows the determination of the coexistence The first peak at 197.5 K belongs to a transition that was line of the phases III and III* in the low pressure region. not mentioned in the work of Finke et al.35 We designate the Unfortunately we did not succeed in obtaining the III–III* respective transition as III–II. The second intense peak is due transition at normal pressure due to the narrow temperature to transition II–I and is detected at 201.3 K. The melting takes region of phase III at such low pressures [see Fig. 1(a)]. Fig. 3 place at 205.2 K. Comparing the determined melting tempera- displays two thermograms of measurements at 50 MPa (with ture with the value given by Finke et al. the calorimetric and without detour) which demonstrate the described purity of the sample can be calculated using the van’t Hoff behaviour.equation as being >99.75%. On cooling only two transitions For pressures higher than 200 MPa the IV–V transition is show up in the DTA curves the freezing at about 203.5 K and obtained by means of a detour consisting of (i) diminution of transition I–III near 193 K. High pressure measurements. The phase behaviour of tertbutylamine was examined in the pressure range from 0.1 to 300 MPa. Fig. 1(b) shows the pressure–temperature phase diagram that resulted from these measurements. Two new phases IV and V [Fig. 1(b)] were discovered in the high-pressure measurements. The transition from phase IV to phase III was not detected at low pressures; it appeared for the first time in the DTA thermograms at a pressure of about 190 MPa.Obviously phase IV crystallises (in the timescale of the DTA measurements) only if phase III is cooled down to pressures between 190 and 220 MPa. To obtain the coexistence line at lower pressure it is therefore necessary to pressurize phase III Fig. 2 DTA trace for TBT (X=SH) at 110 MPa to about 200–220 MPa and to cool down to achieve the III–IV J. Mater. Chem. 1997 7(1) 41–46 43 Table 3 Transition temperatures as a function of pressure [T/K=a+b×( p/MPa)+c×(p/MPa)2] and pressure ranges for the transition lines of TBT (X=SH) transition pressure range/MPa a/K b/K MPa-1 c/104 K MPa-2 I–l 0<p<100 272.3±0.3 0.650±0.018 4.44±1.38 II–I 0<p<300 201.5±0.4 0.368±0.006 0.65±0.21 III–II p>160 158.3±7.9 0.213±0.075 0.23±1.75 III*–II p<160 159.2±0.7 0.236±0.023 1.84±1.58 III–III* 0<p<160 153.7±0.4 0.241±0.004 — V–III p>99 140.3±1.6 0.212±0.008 — IV–III p<99 151.3±0.2 0.099±0.003 — IV–V p>99 149.6±2.7 0.149±0.031 2.92±0.81 256 MPa the peak of the II–I transition already exhibits a shoulder on its high temperature side.In the lower thermogram obtained at ca. 274 MPa the two peaks originating from the phase transitions II–V and V–I are clearly separated. The seven coexistence lines that appear in the pressure– temperature phase diagram of TBA [Fig. 1(b)] have been fitted to polynomials the coefficients of which are shown in Table 4. From these polynomials the coordinates of the triple points can be calculated (i) II–III–IV at (224±3) MPa amn;0.5) K (ii) I–II–V at (247±2) MPa and (238.5±0.5) K.The slopes of the phase transition lines dT/dp at normal Fig. 4 DTA thermogram of TBA (X=NH2 ) at normal pressure pressure can be used to calculate the volume changes of the transitions via the Clausius–Clapeyron equation. In addition transition. Afterwards the pressure can be reduced down to to the slopes the enthalpy changes are needed. The work of the desired value at which the measurement of the IV–III Finke et al.35 contains accurate values of the transition enthalp- transition can be performed. The result of this detour is ies but it is rather difficult to attach these values to the depicted in Fig. 5. The upper thermogram [Fig. 5(a)] is the transitions that were detected in our measurements. It is result of a ‘normal’ measurement (i.e. the pressure was kept questionable whether the enthalpy change of the transition the same on cooling as well as on heating) whereas the lower from ‘crystals II’ to ‘crystals I’ as designated by Finke et al.thermogram [Fig. 5(b)] was obtained after performing the corresponds to the enthalpy change of ‘our’ transition II–I or described detour process. Unlike the III–III* transition of to the sum of the enthalpy changes of the transitions III–II TBT the IV–III transition of TBA could be measured—by and II–I. Assuming the latter the two transition enthalpies using the described detour process—even at normal pressure. can be estimated by comparingthe areas of the two correspond- The second high-pressure induced phase V appears in the ing DTA peaks. The transition enthalpies and volumes thermograms at pressures above ca. 250 MPa.Fig. 6 contains obtained in this way are combined in Table 5. If the enthalpy the DTA curves of two measurements performed at different change measured by Finke et al.35 is fully assigned to the II pressures. In the upper thermogram obtained at about to I transition a volume variation of (5.24±0.09) cm3 mol-1 is obtained. Discussion A theory of fusion of molecular crystals that takes into account orientational as well as positional disorder was developed by Pople and Karasz23 on the basis of the Lennard–Jones– Table 4 Transition temperatures of TBA (X=NH2 ) as a function of pressure [T/K=a+b×(p/MPa)+c×(p/MPa)2] transition a/K b/K MPa-1 c/104 K MPa-2 I–l 205.1±0.8 0.608±0.021 0.22±1.10 Fig. 5 Thermograms for TBA (X=NH2) at about 135 MPa (a) without II–I 201.2±0.2 0.175±0.003 1.02±0.12 and (b) with detour III–II 197.8±0.3 0.167±0.006 0.98±0.25 IV–III 148.5±1.2 0.197±0.024 -(7.47±1.04) IV–II 198.7±1.0 0.142±0.037 — II–V 205.7±3.2 0.133±0.012 — V–I 180.2±8.8 0.236±0.032 — Table 5 Enthalpy and volume changes for the transitions of TBA (X=NH2 ) transition DH/J mol-1 DV/cm3 mol-1 I–l 882.0±0.8a 2.6±0.1 II–I 5600±100 4.9±0.2 III–II 450±100 0.4±0.1 Fig.6 Thermograms of TBA (X=NH2) at (a) ca. 256 MPa and (b) ca. 274 MPa aRef. 35. 44 J. Mater. Chem. 1997 7(1) 41–46 Devonshire approach.55 Amzel and Becka24 extended the rather similar slopes the given average value being ca. 0.35 K MPa-1. model of Pople and Karasz by introducing the existence of more than two possible positions of minimum orientational The entropy and volume changes at the melting transitions of the considered compounds can be compared by using a energy in the crystal in addition to the fact that molecules can occupy either a or b sites i.e.one of two interpenetrating reduced temperature defined as the ratio between the solid– solid phase transition temperature (II to I) and the melting lattices. In both theories the authors defined a non-dimensional parameter n (temperature and volume independent) character- temperature (Tt/Tm). In such a way Fig. 8 and 9 display the entropy and volume changes at the melting transitions as a istic of the molecular crystal as the ratio between the energy barriers for the reorientation and diffusion of the molecules. function of the defined reduced temperature. Both indicate that the further the solid–plastic phase transition is away from Both energy barriers can be obtained by several experimental techniques such as NMR or dielectric methods.Nevertheless the plastic–liquid transition the more the plastic phase differs from the liquid phase thermodynamically. This relationship after a careful and detailed search in the literature for the compounds mentioned in this paper the energy barriers were seems to be independent of the shapes and sizes of the molecules. For example the volumes of the Cl atom and of either not found or the scatter in the values reported by different groups (even with similar techniques) was consider- the Me group are very close; that means that the asymmetry factor (defined as the ratio of the distance from the central able. Nevertheless some results of the model can be analysed in terms of different correlations.carbon to the van der Waals envelope of the Me groups and from the same central carbon atom to the X substituent57) for The slopes (dT/dp) for the melting process are quite similar for the related compounds except for TBCN (X=CN) the molecules like TBC (X=Cl) and NP (X=Me) is almost the same (ca. 1) but the temperature range of the orientationally behaviour of which does not obey the general rules that can be derived for the other compounds. disordered phase is very different as is clearly seen from the Tt/Tm values (0.883 and 0.545 respectively). Moreover com- Fig. 7 displays the volume change as a function of the entropy change at the melting and at the II–I processes for pounds which display a similar temperature range for phase I like TBN (X=NO2) and TBC (X=Cl) (the Tt/Tm values of the collected compounds.The slope of the line relating the melting values which runs through the origin corresponds to which are 0.872 and 0.883 respectively) have clearly different behaviour when their molecular shapes are compared; in the the average of the experimental slopes (dT/dp) for all the compounds (except for TBCN) and it can be considered as the former the asymmetry factor is 1.07 whilst for the latter it is 1.00. Secondly the mentioned result seems also to be indepen- ‘normal slope’ for tert-butyl compounds. The average value was determined to be (0.64±0.14) K MPa-1 (the error being dent of the intermolecular interactions present in the orientationally disordered phase. This is clearly seen if one compares calculated in order to cover all the known values for the compounds studied so far).It should be mentioned that this tert-butyl compounds which exhibit intermolecular interactions by means of hydrogen bonds such as PA (X= correlation between DSm and DVm was predicted by the model of Amzel and Becka.24 However the correlation was derived CO2H)45,47,58 and NPA (X=CH2OH).27 The temperature domain of the plastic phase is in these cases very different on the basis of a common value of D i.e. the number of the positions of minimum orientational energy in the crystal; in (Tt/Tm values are 0.908 and 0.714 for PAand NPA respectively). The entropy and volume changes at the II–I transition as a other words the number of distinguishable orientations of the molecule in the lattice. It is obvious that this number varies function of Tt/Tm does not give a reasonable correlation.This is not surprising if one assumes that the thermodynamic from one tert-butyl compound to another due to the fact that the substituent groups in the ButX molecules (X=SH NO2 Me etc.) can generate additional distinguishable orientations for each one of the tetrahedral orientations. Moreover it must be taken into account that the entropy change of the melting (and also for the solid–solid phase transitions) is not only a result of the change in the number of orientations but also of the volume change as has been proved recently from the pV T data of TBN (X=NO2).16 With regard to the line joining the II–I transition values the correlation is relatively more scattered. This fact is a direct consequence of the differences of the disorder in phases II of the related compounds.The obtained average value for the experimental slopes is determined as (0.27±0.13) K MPa-1. In this context it should be mentioned that Schneider56 found as an experimental evidence that very different phase transitions (solid–solid solid–smectic smectic– Fig. 8 Entropy change (DSm) at the melting transition as a function of the reduced temperature Tt/Tm smectic smectic–nematic nematic–isotropic liquid etc.) exhibit Fig. 9 Volume change (DVm) at the melting transition as a function of Fig. 7 Volume change (DV ) as a function of entropy change (DS) for the melting ($) and II–I transition processes (#) the reduced temperature Tt/Tm J. Mater. Chem. 1997 7(1) 41–46 45 16 M. Jenau J. Reuter J. L. Tamarit and A. Wu� rflinger J. Chem. Soc.properties of phase II (with considerably different character- Faraday T rans. 1996 92 1899. istics for the tert-butyl compounds) cannot be related to the 17 J. C. Frost A. J. Leadbetter and R. M. Richardson Philos. T rans. temperature domain stability (given by Tt/Tm values) of the R. Soc. L ondon B 1980 290 567. disordered forms. 18 A. J. Leadbetter R. C. Ward and R. M. Richardson J. Chem. Soc. It was mentioned in the preceding discussion that TBCN Faraday T rans. 2 1985 81 1067. 19 M. Ferrario M. L. Klein R. M. Lynden-Bell and I. R. McDonald (X=CN) behaves differently than the other tert-butyl com- J. Chem. Soc. Faraday T rans. 2 1987 83 2097. pounds in spite of the shape and size similarities of the 20 U. Wenzel and G. M. Schneider Mol. Cryst. L iq. Cryst. L ett. Sect. molecules.For this compound it has been shown that in phase 1982 72 255; U. Wenzel Doctoral T hesis Bochum 1988. II the molecules undergo rapid reorientational motions about 21 M. Riembauer Doctoral T hesis Bochum 1988. their C–CN axes.5 These motions are quite similar to those 22 J. Wilmers M. Briese and A. Wu� rflinger Mol. Cryst. L iq. Cryst. 1984 187 293. found for TBC (X=Cl)59 and TBB (X=Br)18 in phase III. In 23 J. A. Pople and F. E. Karasz J. Phys. Chem. Solids 1961 18 28; the high-temperature phase I of TBCN the molecules possess F. E. Karasz and J. A. Pople J. Phys. Chem. Solids 1961 20 294. a large degree of freedom of motion along the dipole axes but 24 L. M. Amzel and L. N. Becka J. Phys. Chem. Solids 1969 30 521. only small fluctuations (librational motions) of these axes 25 H.G. Kreul M. Hartmann R. Edelmann A. Wu� rflinger and occur (about 10–15°)18,60 as in the case of phase II of TBC S. Urban Ber. Bunsen-Ges. Phys. Chem. 1989 93 612. (X=Cl). However in phase II of TBB (X=Br) the molecules 26 D. Bu�sing M. Jenau J. Reuter A. Wu� rflinger and J. L. Tamarit Z. Naturforsch. T eil A 1995 50 502. appear to have large fluctuations of the dipole axes (about 60° 27 H. G. Kreul R. Waldinger and A.Wu� rflinger Z.Naturforsch. T eil from their mean direction).18 On the basis of dielectric measure- A 1992 47 1127. ments22 the high-pressure phase IV of TBC (X=Cl) [Fig. 1(c)] 28 M.Woznyj F. X. Prielmeier and H. D. Lu�demann Z.Naturforsch. appeared to be more closely related to phase II of TBB (X= T eil A 1984 39 800. Br) than the phase II of TBC (X=Cl) according to the static 29 A.Wu� rflinger Ber. Bunsen-Ges. Phys. Chem. 1975 79 1195. 30 N. Pingel U. Poser and A. Wu� rflinger J. Chem. Soc. Faraday permittivity. This conclusion was strengthened by comparing T rans. 1 1984 80 3221. the entropy change of the II–I transition for TBB (X=Br) 31 J. Reuter Doctoral T hesis Bochum 1996. (4.60 kJ mol-1) and for TBC (X=Cl) (25.8 kJ mol-1). In this 32 J. Wilmers Diplom. T hesis Bochum 1982. context phase I of TBCN (X=CN) should not be considered 33 H. G. Kreul Doctoral T hesis Bochum 1991. as an orientationally disordered phase; strictly speaking this 34 D. Bu�sing Diplom. T hesis Bochum 1995. 35 H. L. Finke J. F. Messerly and S. S. Todd J. Chem. T hermodyn. phase should be reported as a librational phase much like 1972 4 359. phase II of TBB (X=Br) and TBN (X=NO2 ).36 S. Urban Adv. Mol. Relax. Interact. Processes 1981 21 221. 37 S. Urban Z. Tomkowicz J. Mayer and T.Waluga Acta Phys. Pol. A 1975 48 61. The authors would like to express their gratitude to the UPC- 38 E. F. Westrum and A. Ribner J. Phys. Chem. 1967 71 1216. Olivetti joint committee for the financialsupport which allowed 39 S. Urban J. Mayer and A. I. Belushkin Acta Phys. Pol. A 1983 J.Ll.T. to spend some time in the Institute of Physical 64 161. Chemistry II Ruhr Universita�t where this work was carried 40 T. Hasebe N. Nakamura and H. Chihara Bull. Chem. Soc. Jpn. out. DGICYT grant (PB95–0032) is also acknowledged. 1980 53 896. 41 R. L. Jackson and J. H. Strange Mol. Phys. 1971 22 313. 42 D. W. Aksnes L. L. Kimtys V. J. Balevicius and M. Z. Balevicius Acta Chem. Scand.Ser. A 1984 38 163. References 43 L. L. Kimtys,Mol. Cryst. L iq. Cryst. 1979 56 83. 44 D. W. Aksnes V. J. Balevicius and L. L. Kimtys J. Magn. Reson. 1 N. G. Parsonage and L. A. K. Staveley Disorder in crystals 1983 53 171. Clarendon Press Oxford 1978. 45 T. Hasebe G. Soda and H. Chihara Bull. Chem. Soc. Jpn. 1981 2 Y. A. Atanov and M. I. Shakhparonov Russ. J. Phys. Chem. (Engl. 54 2583. T ransl.) 1969 43 958. 46 H. Namba and T. Oda Bull. Chem. Soc. Jpn. 1952 25 225. 3 M. Brissaud-Lancin and D. Wolf J. Phys. Chem. Solids 1984 47 S. Kondo and T. Oda Bull. Chem. Soc. Jpn. 1954 27 567. 45 733. 48 V. J. Balevicius B. Oral and D. Hadzi Spectrochim. Acta Part A 4 H. Chihara and T. Shinoda Bull. Chem. Soc. Jpn. 1964 37 125. 1981 37 639. 5 J. C. Frost A. J. Leadbetter and R. C. Ward,J.Chem. Soc. Faraday 49 S. Kondo Bull. Chem. Soc. Jpn. 1965 38 527. T rans. 2 1982 78 1009. 50 G. W. Smith J. Chem. Phys. 1969 51 3569. 6 T. Hasebe and H. Chihara Bull. Chem. Soc. Jpn. 1986 59 1141. 51 S. Mooibroek and R. E.Wasylishen Can. J. Chem. 1985 63 2926. 7 T. Hasebe and J. H. Strange J. Chem. Soc. Faraday T rans. 2 1985 52 E. Szczesniak Mol. Phys. 1986 58 551. 81 735. 53 D. W. Aksnes and K. Ramstad Acta Chem. Scand. Ser. A 1987 8 J. Mazur and W. Nosel Acta Phys. Pol. A 1977 52 477. 41 1. 9 R. M. Richardson and P. Taylor,Mol. Phys. 1984 52 525. 54 Krishnaji and A. Mansingh J. Chem. Phys. 1965 42 2503; 1966 10 J. L. Tamarit B. Legendre and J. M. Buisine Mol. Cryst. L iq. 44 1590. Cryst. 1994 250 347. 55 S. E. Lennard-Jones and A. F. Devonshire Proc. R. Soc. L ondon 11 S. Urban J.A. Janik J. Lenik J. Mayer T.Waluga and S.Wrobel Ser. A 1939 168 3171. Phys. Status Solidi A 1972 10 271. 56 G. M. Schneider Faraday Discuss. Chem. Soc. 1980 69 132. 12 V. C. Vani T. Michelberger and E. U. Franck Ber. Bunsen-Ges. 57 C. P. Smyth J. Phys. Chem. Solids 1961 18 40. Phys. Chem. 1990 94 766. 58 W. Longueville M. Bee J. P. Amoureux and R. Fouret J. Phys. 13 G. B. Guthrie and J. P. McCollough J. Phys. Chem. Solids 1961 (Paris) 1986 47 291. 18 53. 59 P. A. C. Gane A. J. Leadbetter R. C. Ward R. M. Richardson and J. Pannetier J. Chem. Soc. Faraday T rans. 2 1982 78 995. 14 J. P. McCullough D. W. Scott H. L. Finke W. N. Hubbard 60 Z. M. El Saffar P. Shultz and E. F. Meyer J. Chem. Phys. 1972 M. E. Gross C. Kartz R. E. Pennington J. F. Messerly and 56 1477. G. Waddington J. Am. Chem. Soc. 1953 75 1818. 15 T. Clark M. A. McKervey H. Mackle and J. J. Rooney J. Chem. Soc. Faraday T rans. 1 1974 7 1279. Paper 6/04182F; Received 14th June 1996 46 J. Mater. Chem.

ISSN:0959-9428    

DOI:10.1039/a604182e

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

Organometallic ferrocenyl dendrimers: synthesis, characterization and redox properties Ching-Fong Shu* and Hsiu-Ming Shen Department of Applied Chemistry, National Chiao T ung University, 1001 T a-Hsueh Road, Hsin-Chu, T aiwan, 30035, Republic of China A series of dendritic poly(aryl ether)s containing 3, 6, 12 and 24 ferrocene functionalities located exclusively at the peripheries of the dendritic structures have been synthesized using the stepwise convergent approach. The structures of these dendrimers have been characterized using 1H and 13C NMR spectroscopy.Cyclic and normal pulse voltammetric studies indicate that the ferrocenyl moieties located on the outer surfaces of these dendrimers are non-interacting redox centres, are electrochemically equivalent, and are oxidizable at the same potential.The results of controlled-potential coulometric oxidation show that nearly all the ferrocene residues in the dendrimers are accessible to electron transfers in electrode reactions. Macromolecular materials with skeletons of transition-metal reported procedures. Other commercially available chemicals were reagent grade and were used as purchased.atoms in close proximity are attracting increasing attention because of their potentially interesting electrical, redox and 3-Bromopropylferrocene (Fc-G0-Br)11 optical characteristics. Among the organotransition-metal complexes, ferrocene has been shown to have excellent thermal To a mixture of 3-ferrocenylpropanol (5.43 g, 22.3 mmol) and and photochemical stability, and to undergo a facile and carbon tetrabromide (14.63 g, 44.6 mmol) in 10 ml of THF reversible one-electron oxidation to ferrocenium cation; the was added triphenylphosphine (11.53 g, 44 mmol).The reac- reaction leads to a marked change in electrical and spectro- tion mixture was stirred at room temperature under nitrogen scopic properties and may be effected chemically, electrochemi- for 1 h.The resulting mixture was then poured into water and cally or photochemically. Ferrocenyl-based polymers have been extractedwith diethyl ether. The extract was dried over Na2SO4 used in chemical modification of electrodes, in the construction and evaporated to dryness. The crude product was purified by of amperometric biosensors, and in the area of non-linear column chromatography (SiO2), eluting with 1510 ethyl acet- optical materials.1 ate-hexane, to give 4.67 g (69%) of Fc-G0-Br as an orange oil.Recently, dendritic macromolecules have received consid- 1H NMR (CDCl3 ) d 2.04 (m, 2H, CH2), 2.50 (t, 2H, J=7.8 Hz, erable attention because of their unique hyperbranched poly- CpCH2), 3.42 (t, 2H, J=6.3 Hz, CH2O), 4.07 (m, 4H, C5H4), meric structure and well defined three-dimensional architec- 4.10 (s, 5H, C5H5); 13C NMR (CDCl3) d 27.91 (CH2), 33.63 tures.2 These dendritic macromolecules are characterized by a (CH2Br), 33.91 (CpCH2), 67.36, 6814 (C5H4 ), 68.57 (C5H5 ), central polyfunctional core, from which arise successive layers 88.25 (C5H4).of monomer units with branches occurring at each monomer unit.3 This results in a nearly entanglement-free hyperbranched General procedure for the synthesis of Fc-Gx-OH (x=1, 2, 3) structure that may adopt a spherical shape, the periphery of A mixture of the appropriate Fc-Gx-Br (x=0, 1, 2) (2 equiv.), which consists of a large number of chain ends or surface 3,5-dihydroxybenzyl alcohol (1 equiv.), potassium carbonate, functional groups.4 This unique macromolecular architecture (3.5 equiv.) and 18-crown-6 (0.2 equiv.) in dry acetone was allows precise control of molecule size, as well as of the heated at reflux and stirred vigorously under nitrogen for 48 h.disposition of the desired functionalities. The synthesis of well The mixture was allowed to cool, added to water and extracted defined, highly branched macromolecules possessing functional with CH2Cl2 (3×).The combined extracts were dried over components on their exterior surfaces, within their dendritic MgSO4 and evaporated to dryness. The crude product was branches, and at their interior cores, giving rise to new purified as outlined in the following text. materials with desirable properties has sparked much interest. 4–6 In this paper, we report full details on the synthesis, General procedure for the synthesis of Fc-Gx-Br (x=1, 2, 3)12 characterizations and redox behaviour of a family of dendritic poly(aryl ether)s that were synthesized using a stepwise conver- To a solution of Fc-Gx-OH (x= 1, 2, 3) (1 equiv.) in benzene gent approach,7 and possess ferrocene functional groups was added dropwise PBr3 (0.4 equiv.).The reaction mixture located exclusively at the peripheries of their dendritic was stirred at room temperature under nitrogen for 4 h.For structures. the latter generation (x=3), larger excesses of PBr3 were required to force the reaction to completion. PBr3 was added in 0.4 equiv. amounts at hourly intervals until TLC showed Experimental no starting material. The mixture was then added to water, Materials neutralized with saturated aqueous NaHCO3, and extracted with CH2Cl2 (3×).The combined extracts were dried over Acetone was distilled from CaSO4. Benzene was distilled from MgSO4 and evaporated to dryness. The crude product was CaH2. Tetrahydrofuran (THF) was distilled from Na/K alloy purified as outlined in the following text. and benzophenone. n-Tetrabutylammonium hexafluorophosphate (TBAPF6) was prepared as described previously,8 and Fc-G1-OH was purified by recrystallization three times from ethyl acetate and dried in vacuo at 60°C. 3-Ferrocenylpropanol9 and 3,5- This was prepared from Fc-G0-Br and purified by column chromatography (SiO2), eluting with 255 ethyl acetate–hexane, dihydroxybenzyl alcohol10 were synthesized according to the J. Mater.Chem., 1997, 7(1), 47–52 47to give Fc-G1-OH (79%) as an orange solid. 1HNMR (CDCl3) to give Fc-G3-Br (24%) as an orange solid. 1H NMR (CDCl3) d 1.98 (m, 16H, CH2), 2.51 (t, 16H, J=7.8 Hz, CpCH2), 3.96 d 1.98 (m, 4H, CH2), 2.51 (t, 4H, J=7.8 Hz, CpCH2 ), 3.96 (t, 4H, J=6.3 Hz, CH2O), 4.06 (m, 4H, C5H4), 4.08 (m, 4H, C5H4), (t, 16H, J=6.3 Hz, CH2O), 4.07 (m, 16H, C5H4), 4.09 (m, 16H, C5H4), 4.11 (s, 40H, C5H5), 4.40 (s, 2H, CH2Br), 4.95 (s, 12H, 4.11 (s, 10H, C5H5), 4.62 (d, 2H, J=4.8 Hz, CH2OH), 6.40 (t, 1H, J=2.0 Hz, ArH), 6.52 (d, 2H, J=2.0 Hz, ArH); 13C NMR ArCH2O), 6.42 (t, 4H, J=2.0 Hz, ArH), 6.54 (m, 3H, ArH), 6.57 (d, 8H, J=2.0 Hz, ArH), 6.60 (d, 2H, J=2.0 Hz, ArH), (CDCl3) d 25.89 (CH2), 30.50 (CpCH2), 65.37, 67.27 (CH2O), 67.27, 68.14 (C5H4), 68.55 (C5H5), 88.19 (C5H4), 100.57, 105.12, 6.68 (d, 4H, J=2.0 Hz, ArH); 13C NMR (CDCl3) d 25.89 (CH2), 30.47 (CpCH2 ), 33.59 (CH2Br), 67.27 (C5H4), 67.33, 143.26, 160.41 (ArC). 67.71, (CH2O), 67.74 (C5H4 ), 68.55 (C5H5), 70.10 (CH2O), Fc-G1-Br 88.25 (C5H4 ), 100.86, 101.62, 102.15, 105.75, 106.40, 108.12, 138.97, 159.92, 160.10, 160.36 (ArC).This was prepared from Fc-G1-OH and purified by column chromatography (SiO2), eluting with 155 ethyl acetate–hexane General procedure for the synthesis of dendritic molecules to give Fc-G1-Br (85%) as an orange solid. 1H NMR (CDCl3) d 1.98 (m, 4H, CH2), 2.51 (t, 4H, J=7.8 Hz, CpCH2 ), 3.95 (t, A mixture of Fc-Gx-Br (x=0, 1, 2, 3) (3 equiv.), 1,1,1-tris(4- 4H, J=6.3 Hz, CH2O), 4.06 (m, 4H, C5H4), 4.06 (m, 4H, C5H4), hydroxyphenyl)ethane (1 equiv.), potassium carbonate (4.5 4.11 (s, 10H, C5H5), 4.41 (s, 2H, J=4.8 Hz, CH2Br), 6.39 (t, equiv.), and 18-crown-6 (0.3 equiv.) in acetone was heated at 1H, J=2.0 Hz, ArH), 6.52 (d, 2H, J=2.0 Hz, ArH); 13C NMR reflux and stirred vigorously under nitrogen for 72 h.The (CDCl3) d 25.89 (CH2), 30.44 (CpCH2 ), 33.76 (CH2Br), 67.27 mixture was allowed to cool, added to water and extracted (CH2O), 67.21, 68.09 (C5H4), 68.49 (C5H5), 99.10 (C5H4), with CH2Cl2 (3×).The combined extracts were dried over 101.42, 139.59, 160.27 (ArC). MgSO4 and evaporated to dryness. The crude product was purified as outlined in the following text. Fc-G2-OH Dendrimer 1. This was prepared from Fc-G0-Br and purified This was prepared from Fc-G1-Br and purified by column by column chromatography (SiO2 ), eluting with 154 ethyl chromatography (SiO2), eluting with 153 hexane–chloroform acetate–hexane, to give dendrimer 1 (58%) as an orange solid.to give Fc-G2-OH (85%) as an orange solid. 1HNMR (CDCl3) 1H NMR (CDCl3) d 1.98 (m, 6H, CH2), 2.11 (s, 3H, CH3 ), d 1.98 (m, 8H, CH2), 2.51 (t, 8H, J=7.8 Hz, CpCH2 ), 3.96 (t, 2.51 (t, 6H, J=7.8 Hz, CpCH2), 3.95 (t, 6H, J=6.3 Hz, CH2O), 8H, J=6.3 Hz, CH2O), 4.05 (m, 8H, C5H5), 4.08 (m, 8H, C5H4), 4.05 (m, 6H, C5H4), 4.08 (m, 6H, C5H4), 6.79 (d, 6H, J=8.7 Hz, 4.11 (s, 20H, C5H5), 4.62 (d, 2H, J=4.8 Hz, CH2OH), 4.96 (s, core Ar¾H), 6.99 (d, 6H, J=8.71 Hz, core Ar¾H); 13C NMR 4H, ArCH2O), 6.42 (t, 2H, J=2.0 Hz, ArH), 6.54 (t, 1H, J= (CDCl3 ) d 25.92 (CH2), 30.50 (CpCH2), 30.73 (CH3), 50.54 2 Hz, ArH), 6.57 (d, 4H, J=2.0 Hz, ArH), 6.61 (d, 2H, J= (CCH3), 67.15 (CH2O), 67.17, 68.03, (C5H4), 68.46 (C5H5 ), 2.0 Hz, ArH); 13C NMR (CDCl3) d 25.89 (CH2), 30.50 88.25 (C5H5), 113.60, 129.56, 141.71, 156.92 (Ar¾C).(CpCH2), 65.27 (CH2O), 67.23 (C5H4), 67.41 (CH2O), 68.41 (C5H4), 68.52 (C5H5), 70.03 (CH2O), 88.20 (C5H4), 100.80, Dendrimer 2. This was prepared from Fc-G1-Br and purified 101.27, 105.65, 105.73, 139.03, 143.37, 160.08, 160.36 (ArC).by column chromatography (SiO2), eluting with 151 chloro- Fc-G2-Br form–hexane to give dendrimer 2 (57%) as an orange solid. 1H NMR (CDCl3 ) d 1.98 (m, 12H, CH2), 2.11 (s, 3H, CH3 ), This was prepared from Fc-G2-OH and purified by column 2.51 (t, 12H, J=7.8 Hz, CpCH2), 3.95 (t, 12H, J=6.3 Hz, chromatography (SiO2), eluting with 253 hexane–chloroform CH2O), 4.05 (m, 12H, C5H4), 4.08 (m, 12H, C5H4), 4.95 (s, 6H, to give Fc-G2-Br (44%) as an orange solid. 1H NMR (CDCl3) CH2O), 6.42 (t, 3H, J=2.0 Hz, ArH), 6.58 (d, 6H, J=2.0 Hz, d 1.98 (m, 8H, CH2), 2.51 (t, 8H, J=7.8 Hz, CpCH2 ), 3.96 (t, ArH), 6.87 (d, 6H, J=8.7 Hz, core Ar¾H), 7.00 (d, 6H, J= 8H, J=6.3 Hz, CH2O), 4.07 (m, 8H, C5H4), 4.09 (m, 8H, C5H4), 8.7 Hz, core Ar¾H); 13C NMR (CDCl3) d 25.89 (CH2), 30.47 4.11 (s, 20H, C5H5), 4.42 (s, 2H, CH2Br), 4.95 (s, 4H, ArCH2O), (CpCH2 ), 30.76 (CH3), 50.60 (CCH3 ), 65.37 (CH2O), 67.21 6.43 (t, 2H, J=2.0 Hz, ArH), 6.55 (t, 1H, J=2.0 Hz, ArH), (C5H4 ), 67.30 (CH2O), 68.08 (C5H4), 68.49 (C5H5), 70.01 6.57 (d, 2H, J=2.0 Hz, ArH), 6.64 (d, 2H, J=2.0 Hz, ArH); (CH2O), 88.16 (C5H4 ), 100.77, 105.76, 113.95, 129.59, 139.35, 13C NMR (CDCl3) d 25.89 (CH2), 30.47 (CpCH2), 33.65 142.00, 156.80, 160.36 (Ar and Ar¾C).(CH2Br), 67.27 (C5H4), 67.33 (CH2O), 68.14 (C5H4), 68.58 (C5H5), 70.12 (CH2O), 88.25 (C5H4), 100.89, 102.17, 105.78, Dendrimer 3. This was prepared from Fc-G2-Br and purified 108.09, 138.80, 139.70, 159.98, 160.39 (ArC). by column chromatography (SiO2), eluting with 152 chloroform –hexane to give dendrimer 3 (59%) as an orange solid.Fc-G3-OH 1H NMR (CDCl3 ) d 1.97 (m, 24H, CH2), 2.11 (s, 3H, CH3 ), This was prepared from Fc-G2-Br and purified by column 2.50 (t, 24H, J=8.1 Hz, CpCH2), 3.95 (t, 24H, J=6.3 Hz, chromatography (SiO2 ), eluting with 153 hexane–chloroform, CH2O), 4.05 (m, 24H, C5H4), 4.08 (m, 24H, C5H4), 4.10 (s, to give Fc-G3-OH (63%) as an orange solid. 1HNMR (CDCl3) 60H, C5H5), 4.95 (s, 18H, CH2O), 6.42 (t, 6H, J=2.0 Hz, ArH), d 1.97 (m, 16H, CH2), 2.51 (t, 16H, J=7.8 Hz, CpCH2), 3.95 6.57 (m, 15H, ArH), 6.68 (d, 6H, J=2.0 Hz, ArH), 6.87 (d, 6H, (t, 16H, J=6.3 Hz, CH2O), 4.05 (m, 16H, C5H4), 4.08 (m, 16H, J=8.7 Hz, core Ar¾H), 7.02 (d, 6H, J=8.7 Hz, core Ar¾H); C5H4 ), 4.10 (s, 40H, C5H5 ), 4.61 (d, 2H, J=4.8 Hz, CH2OH), 13C NMR (CDCl3) d 25.84 (CH2), 30.44 (CpCH2), 50.72 4.96 (s, 8H, ArCH2O), 4.97 (s, 4H, ArCH2O), 6.42 (t, 4H, J= (CCH3), 67.18 (C5H4 ), 67.24 (CH2O), 68.06 (C5H4), 68.46 2.0 Hz, ArH), 6.54 (m, 3H, ArH), 6.57 (d, 8H, J=2.0 Hz, ArH), (C5H5 ), 69.83, 70.01 (CH2O), 88.10 (C5H4), 100.81, 101.45, 6.60 (d, 2H, J=2.0 Hz, ArH), 6.68 (d, 4H, J=2.0 Hz, ArH); 105.73, 106.37, 113.89, 129.57, 139.38, 141.98, 156.69, 160.01, 13C NMR (CDCl3) d 25.89 (CH2), 30.47 (CpCH2), 65.23 160.30 (Ar and Ar¾C).(CH2O), 67.27 (C5H4 ), 67.33 (CH2O), 67.74 (C5H4), 68.55 (C5H5), 69.95, 70.10 (CH2O), 88.25 (C5H5), 100.86, 101.21, Dendrimer 4. This was prepared from Fc-G3-Br and purified 101.56, 105.67, 105.79, 106.31, 138.97, 139.18, 143.37, 160.01, by column chromatography (SiO2), eluting with 152 chloro- 160.07, 160.36 (ArC).form–hexane to give dendrimer 3 (41%) as an orange solid. 1H NMR (CDCl3 ) d 1.96 (m, 48H, CH2), 2.09 (s, 3H, CH3 ), Fc-G3-Br 2.49 (m, 48H, CpCH2), 3.93 (t, 48H, CH2O), 4.05 (m, 96H, C5H4), 4.10 (s, 120H, C5H5), 4.94 (s, 36H, CH2O), 6.41 (m, This was prepared from Fc-G3-OH and purified by column chromatography (SiO2), eluting with 253 hexane–chloroform 12H, ArH), 6.57 (m, 33H, ArH), 6.68 (d, 18H, J=2.0 Hz, ArH), 48 J. Mater.Chem., 1997, 7(1), 47–526.87 (d, 6H, J=8.7 Hz, core Ar¾H), 7.01 (d, 6H, J=8.7 Hz, rocene. The bromide Fc-G0-Br was obtained by bromination of 3-ferrocenylpropanol, which was prepared according to the core Ar¾H); 13C NMR (CDCl3) d 25.89 (CH2), 30.47 (CpCH2), 67.21 (C5H4), 67.33 (CH2O), 67.71 (CH2O), 68.08 (C5H4), 68.28 literature procedures.9 Bromination of Fc-G1-OH with PBr3 restored the reactive bromomethyl functionality at the focal (CH2O), 68.49 (C5H5 ), 70.10 (CH2O), 100.89, 105.79, 106.40, 113.92, 138.94, 160.10, 160.36, 160.48 (Ar and Ar¾C).point of the dendritic wedge to give the corresponding benzyl bromide Fc-G1-Br.Reaction of Fc-G1-Br with 3,5-dihydroxybenzyl alcohol resulted in the formation of the next-generation Electrochemical apparatus benzyl alcohol Fc-G2-OH, and subsequent bromination of Fc- Electrochemical measurements were carried out with a BAS G2-OH with PBr3 gave the second generation bromide Fc-G2- 100B/W Electrochemical Workstation. The working electrode Br. Repeated alkylation and bromination led to the third for cyclic voltammetry and normal pulse voltammetry was a generation alcohol Fc-G3-OH and the bromide Fc-G3-Br. In platinum disk electrode (0.5 mm diameter, sealed in soft glass) the convergent approach, the dendritic wedges obtained are that was polished prior to use with 1 mm diamond paste and attached to a polyfunctional core.The polyfunctional core we rinsed thoroughly with water and acetone.For coulometry chose was 1,1,1-tris(4-hydroxyphenyl)ethane. Coupling reac- a large platinum gauze electrode was employed. All poten- tions of the phenolic groups of the core molecule with each tials are referenced to the Ag/Ag+ (0.01 mol dm-3 AgNO3, bromide generation, from Fc-G0-Br to Fc-G3-Br, as shown in 0.1 mol dm-3 TBAPF6–CH3CN) reference electrode.A coiled Scheme 2, were carried out. Thus, a family of dendritic poly- platinum wire was used as a counter electrode and the electro- (arylether)s, dendrimers 1, 2, 3 and 4, containing 3, 6, 12 and chemical cells were of conventional design. 24 ferrocene functionalities, respectively, at their peripheries were obtained. The products were purified by column chromatography, and their purities were evaluated by 1H and Results and Discussion 13C NMR spectroscopy.No signal of the starting materials Synthesis of dendritic macromolecules was detected in the NMR experiments. A series of poly(aryl ether) dendrimers containing ferrocene functional groups located at the peripheries of their dendritic Characterization structures were prepared using the convergent approach developed by Hawker and Fre� chet.7 The synthetic route to the The ferrocene-terminated dendrimers were characterized structurally using 1H and 13C NMR spectroscopy. 1H NMR ferrocene-terminated dendritic wedges is illustrated in Scheme 1. This was made on the base of the Williamson ether spectroscopy is particularly crucial in confirming the structures produced by our synthetic strategy. 1H NMR spectra showed synthesis for the formation of aryl ether from phenol and benzylic bromide (or alkyl bromide). We used 3,5-dihydroxy- the exterior functional group gave five sets of resonances at d 4.10 (C5H5), 4.08, 4.05 (C5H4), 3.95 (CH2O), 2.50 (CpCH2) benzyl alcohol as the monomer unit. The ferrocene groups were introduced to the peripheries of the growing dendrimers and 1.98 (CH2).This result is consistent with the expected highly symmetricaltructures of the ferrocene dendrimers. The by coupling two molecules of 3-bromopropyl ferrocene Fc- G0-Br to the phenolic groups of the monomer unit in the resonances for the aromatic protons of the internal 3,5- dihydroxybenzyl group occurred in the d 6.50–6.70 region. All presence of potassium carbonate and 18-crown-6 in refluxing acetone, thereby obtaining the first generation alcohol Fc-G1- benzylic protons resonate at d 4.91–5.02, except those at the focal point at which CH2OH appears at d 4.62 and CH2Br at OH.We selected c-propylferrocene as the terminal group because bromomethylferrocene, which is an a-functional ferro- d 4.42. Changes in resonance of benzylic protons occurred upon conversion of the dendritic alcohol to the corresponding cene derivative,is hydrolytically unstable and readily converted to ferrocenemethanol, and 2-bromoethylferrocene undergoes bromide, and on attachment of the bromide to the polyfunctional core, were used to monitor product purities.When elimination during the alkylation reaction, producing vinylfer- Scheme 1 Reagents and conditions: (a) 3,5-dihydroxybenzyl alcohol, K2CO3 , 18-crown-6, reflux in acetone; (b) PBr3, C6H6 , room temperature J.Mater. Chem., 1997, 7(1), 47–52 49Scheme 2 dendritic wedges were attached to the core molecule 1,1,1- tris(4-hydroxyphenyl)ethane, two doublets of the aromatic rings of the core moiety were observed at d 6.87 and 7.02, and were not obscured by the aromatic resonances of the dendritic wedges; the methyl group of the core resonated at d 2.11. Integration data for the exterior ferrocene groups, the internal aromatic groups, and the aromatic rings of the core moiety confirmed the structures and the generation numbers of the dendrimers.The 13C NMR spectra provided further support for the structural assignments.Each of the different carbon resonances was found to agree with the observed spectral resonances. Electrochemistry Electrochemistry studies of the ferrocene dendrimers in CH2Cl2 solutions containing 0.1 mol dm-3 TBAPF6 as the supporting electrolyte were performed. Cyclic voltammograms of 1 and 2 show the characteristics of a reversible one-electron oxidation, with production of soluble, stable cations.However, with 3 and 4, a stripping peak appeared on scan reversal. The shape of the reduction peak indicated that the oxidized products were insoluble and accumulated on the electrode’s surface. The precipitation problems led us to use the normal pulse voltammetry (NPV) technique, which is less susceptible to problems of adsorption and precipitation as a means of determining the wave parameters.13 The normal pulse voltammograms of 3 in CH2Cl2 are shown in Fig. 1A. Surprisingly, the slope of E vs. log(id-i )/i plot, where id is the limiting diffusion current, obtained from the normal pulse voltammogram is 42 mV, Fig. 1 Normal pulse voltammograms for 50 mmol dm-3 of dendrimer which is different from the slopes (60 mV) reported for most 3 in 0.1 mol dm-3 TBAPF6–CH2Cl2.A, Potential scanned in the polymers (or dendrimers) containing multiple ferrocenyl moiet- forward direction, initial potential 0 mV vs. Ag/Ag+; B, potential ies.13–15 We speculated that the reason why a value of 42 mV scanned in the reverse direction, initial potential 500 mV vs. Ag/Ag+. Scan rate 20 mV s-1; sample width 17 ms; pulse width 50 ms.was obtained instead of 60 mV may be due to severe adsorption 50 J. Mater. Chem., 1997, 7(1), 47–52of the oxidized product. The effects of absorbed reactants and technique. Normal pulse voltammograms of the dendrimers are displayed in Fig. 3. The half-wave potentials, limiting products on the plateau currents and half-wave potentials in normal pulse voltammograms have been reported by Anson diffusion currents (id), and slopes of E vs.log(id-i)/i plots, obtained from the normal pulse voltammograms are given in et al.16 In the absence of an independent determination, it is difficult to deduce the presence of product adsorption from Table 1. The E1/2 values, as determined by NPV, are similar to those determined by cyclic voltammetry. Interestingly, the normal pulse voltammograms.However, the appearance of a peak current in the voltammogram recorded in the reverse- E1/2 values for the ferrocene redox couples remain essentially constant upon successive generation buildup. This can be scan direction verifies the occurrence of adsorption. Indeed, for dendrimer 3 there is a peak current recorded in the reverse- explained by noting that the ferrocene functional groups are located on the outer surfaces of the dendrimers and their scan voltammogram, as shown in Fig. 1B. In order to reduce perturbations in the shape of normal microenvironments are independent of dendrimer generation. This result is in contrast to the recent observation by Diederich pulse voltammograms induced by adsorption, we used a mixed solvent, 355 CH3CN–CH2Cl2 in the electrochemical studies, et al.17 that for zinc porphyrin dendrimers the redox potential shifts as generations build up.Because the redox group is where the adsorption of the dendrimers and their oxidized products was minimized. Fig. 2 shows the cyclic voltammo- located in the interior of the porphyrin macromolecules, the electrophore environment, as well as the redox potential of the gram of the ferrocene-terminated dendrimers in the mixed solvent containing 0.1 mol dm-3 TBAPF6 as the supporting zinc porphyrin, changes markedly with increasing generational cascading.It was observed that oxidation of the terminated electrolyte. The cyclic voltammogram of each dendrimer exhibits a single reversible oxidation wave corresponding to oxi- ferrocenyl units in the dendrimers occurred at a potential ca. 40 mV more negative than the corresponding process in ferro- dation of the exterior ferrocene groups. The peak current is linearly proportional to the square root of the scan rate, the cene. This was due to the electron-donating effect of the alkyl group on the ferrocenyl moiety, making the chain-end ferro- peak shape is symmetrical, with ipc/ipa#1, and Ep is independent of the scan rate.The peak potential separation DEp is ca. cenyl groups more easily oxidized. For the dendrimers, the values of DEp determined from CV and the values of slopes 59 mV for dendrimers 1 and 2, 52 mV for 3 and 50 mV for 4. The smaller DEp values for dendrimers 3 and 4 may be obtained from NPV are close to the theoretical value of 59 mV for a reversible one-electron transfer reaction, indicating that attributable to minor adsorption of the dendrimers (or oxidation products) on the electrode surface.The half-wave poten- all the ferrocenyl redox centres located at the peripheries of tials (E1/2) and DEp obtained from the cyclic voltammograms are given in Table 1. Note that the sharp anodic or cathodic peak current due to the adsorption of reactant or product is not observed in the mixed solvent system.The electrochemical properties of the dendrimers were also studied using the NPV Fig. 3 Normal pulse voltammograms for the oxidation of A, Fig. 2 Cyclic voltammograms for A, 200 mmol dm-3 of 1; B, 100 mmol dm-3 of 2; C, 50mmol dm-3 of 3; and D, 25 mmol dm-3 200 mmol dm-3 of 1; B, 100 mmol dm-3 of 2; C,50 mmol dm-3 of 3; and D,25 mmol dm-3 of 4 in 0.1 mol dm-3TBAPF6–355 CH3CN–CH2Cl2 .of 4 in 0.1 mol dm-3 TBAPF6–355 CH3CN–CH2Cl2. Scan rate 60 mV s-1. Scan rate 20 mV s-1; sample width 17 ms; pulse width 50 ms. Table 1 Results of cyclic voltammetry, normal pulse voltammetry and controlled-potential electrolysis for the oxidation of ferrocene-terminated dendrimers 1–4a compound E1/2b/mV E1/2c/mV DEp/mV sloped/mV id/10-1 mA Qce/10-1 C npf 1 211 208 59 60 8.16 2.90 (0.03) 3.0 (0.1) 2 211 208 59 58 7.21 2.95 (0.03) 6.1 (0.1) 3 213 212 52 56 5.90 2.74 (0.06) 11.4 (0.2) 4 214 212 50 58 5.14 2.70 (0.05) 22.4 (0.4) aExperimental conditions are given in the captions to Fig. 2 and 3. bBy cyclic voltammetry. cBy normal pulse voltammetry. dSlope of plot of E vs.log(id-i)/i. eThe electrolysis solutions were 5 ml, and the concentration of each solution was as indicated in the captions to Fig. 2 and 3. Values are an average of three independent determinations, and standard deviations are listed in parentheses. fNumber of ferrocenes per molecule oxidized. J. Mater. Chem., 1997, 7(1), 47–52 51and M. S. Sigman, Macromolecules, 1992, 25, 6055; A.Togni and these dendrimers are equivalent, non-interacting and exchange G. Rihs, Organometallics, 1993, 12, 3368. electrons with the electrode in waves characteristic of one- 2 D. A. Tomalia, A. M. Naylor and W. A. Goddard, Angew. Chem., electron transfer processes. Int. Ed. Engl., 1990, 29, 138; H. Mekelburger, W. I. Jaworke and To evaluate the total number of electrons transferred during F.Vo� gtle, Angew. Chem., Int. Ed. Engl., 1992, 31, 1571; oxidation without knowing the relative diffusion coefficients, J. M. J. Fre� chet, Science, 1994, 263, 1710. 3 D. A. Tomalia, H. Baker, J. R. Dewald, M. Hall, G. Kallos, controlled-potential electrolyses were carried out. Coulometric S. Martin, J. Roeck, J. Ryder and P. Smith, Macromolecules, 1986, oxidation of the dendrimers was performed at a large-area 19, 2466; G.R. Newkome, Z. Yao, G. R. Baker, V. K. Gupta, platinum gauze electrode in a CH2Cl2–CH3CN mixture at P. S. Russo and M. J. Saunders, J. Am. Chem. Soc., 1986, 108, 849; Eapp 150 mVmore positive than the respective anodic potential. G. R. Newkome, A. Nayak, R. K. Behara, C. N. Moorefield and The results are given in Table 1.It is shown that values of np, G. R. Baker, J. Org. Chem., 1992, 57, 358; H. Uchida, Y. Kabe, K. Yoshino, A. Kawamata, T. Tsumuraya and S. Masamune, the number of electrons per molecule consumed in the oxi- J. Am. Chem. Soc., 1990, 112, 7077. dation, calculated from the total number of coulombs, come 4 J. L. Fillaut and D. Astruc, J. Chem. Soc., Chem. Commun., 1993, close to matching the number of ferrocenyl groups anticipated 1320; F.Moulines, L. Djakovitch, R. Boese, B. Gloaguen, W. Thiel, in the synthetic strategy bound to the chain-ends of these J. L. Fillaut, M. H. Delville and D. Astruc, Angew. Chem., Int. Ed. dendrimers. Note that the values of np obtained for 3 and 4 Engl., 1993, 32, 1075; Y-H. Liao and J. R. Moss, J. Chem. Soc., Chem.Commun., 1993, 1774; B. Alonso, I. Cuadrado, M. Moran are somewhat smaller than the number of ferrocene groups and J. Losada, J. Chem. Soc., Chem. Commun., 1994, 2575; Y- attached to these dendrimers. This may have been because the H. Liao and J. R. Moss, Organometallics, 1995, 14, 2130;K. Lorenz, rate of electrolysis of dendrimers 3 and 4 slowed drastically R. Muhaupt, H.Frey, U. Rapp and F. J. Mayer-Posner, near the end of coulometric oxidation, and at the time the Macromolecules, 1995, 28, 6657. controlled-potential electrolysis was stopped, the ferrocene 5 T. Nagasaki, M. Ukon, S. Arimori and S. Shinkai, J. Chem. Soc., Chem. Commun., 1992, 608; G. R. Newkome, F. Cardullo, redox groups that were bound to the dendrimers in the solution E. C. Constable, C.N. Moorefield and A. M. W. C. Thompson, were not totally oxidized. J. Chem. Soc., Chem. Commun., 1993, 925; S. Achar and R. J. Puddephatt, Angew. Chem., Int. Ed. Engl., 1994, 33, 847; G. R. Newkome, C. N. Moorefield, J. M. Keith, G. R. Baker and Conclusion G. H. Escamilla, Angew. Chem., Int. Ed. Engl., 1994, 106, 701; S. Serroni, G. Denti, S. Campagna, A. Juris, M. Ciano and Using the stepwise convergent approach we have synthesized V.Balzani, Angew. Chem., Int. Ed. Engl., 1992, 104, 1540. a series of dendritic poly(aryl ether)s that contain 3, 6, 12 and 6 R-H. Jin, T. Aida and S. Inoue, J. Chem. Soc., Chem. Commun., 24 ferrocene functionalities located exclusively at the peripher- 1993, 1260; C. J. Hawker, K. L.Wooley and J. M. J. Fre� chet, J.Am. Chem. Soc., 1993, 115, 4375; K. L. Wooley, C. J. Hawker, ies of their dendritic structures. The structures of these dendri- J. M. J. Frechet, F. Wudl, G. Srdanov, S. Shi, C. Li and M. Kao, mers were characterized using 1H and 13C NMR spectroscopy. J. Am. Chem. Soc., 1993, 115, 9836; P. J. Dandliker, F. Diederich, The results of electrochemical studies show that electron J-P.Gissebrecht, A. Louati and M. Gross, Angew. Chem., Int. Ed. transfers from the dendritic macromolecules yielded voltam- Engl., 1995, 34, 2725. metric waves with shapes matching those of corresponding 7 C. Hawker and J. M. J. Fre� chet, J. Am. Chem. Soc., 1990, 112, 7638; K. L. Wooley, C. J. Hawker and J. M. J. Fre� chet, J. Chem. Soc., molecules with electroactive centres, but with magnitudes Perkin T rans. 1, 1991, 1059; C. J. Hawker and J. M. J. Fre�chet, determined by the total number of redox centres present. On J. Am. Chem. Soc., 1992, 114, 8405. the basis of this observation, we conclude that the ferrocenyl 8 D. T. Sawyer and J. L. Roberts Jr., Experimental Electrochemistry moieties located on the outer surfaces of these dendrimers are for Chemists, Wiley, New York, 1979. non-interacting redox centres, are electrochemically equivalent 9 K. Schlogel,Monatsh. Chem., 1957, 88, 601. and oxidizable at the same potential. This fully reversible 10 C. G. Pitt, H. H. Seltzman, Y. Sayed, C. E. Twine Jr. and D. L. Willians, J. Org. Chem., 1979, 44, 677. multi-electron redox system may be useful for multi-electron 11 J. Hooz and S. S. H. Gilani, Can. J. Chem., 1968, 46, 86. redox catalysis. 12 E. Reimann, Chem. Ber., 1968, 102, 2887. 13 J. B. Flanagan, S. Margel, A. J. Bard and F. C. Anson, J. Am. Chem. Soc., 1978, 100, 4248. Dr. Fred C. Anson (California Institute of Technology) is 14 T. W. Smith, J. E. Kuder and D. Wychich, J. Polym. Sci., 1976, gratefully acknowledged for helpful discussions. We thank the 15, 2433. National Science Council (ROC) (NSC 84-2113-M-009-003) 15 C. D’Silva, S. Afeworki, O. L. Parri, P. K. Baker and for its financial support of this research. A. E. Underhill, J. Mater. Chem., 1992, 2, 225. 16 J. B. Flangan, K. Takahashi and F. C. Anson,J. Electroanal. Chem., 1977, 85, 257. 17 P. J. Dandliker, F. Diederich, M. Gross, C. B. Knobler, A. Louati References and E. M. Sanford, Angew. Chem., Int. Ed. Engl., 1994, 33, 1739. 1 P. D. Hale, J. Inagaki, H. I. Karan, Y. Okamoto and T. A. Stotheim, J. Am. Chem. Soc., 1989, 111, 3482; M. E. Wright Paper 6/04225B; Received 17th June, 1996 52 J. Mater. Chem., 1997, 7(1),

ISSN:0959-9428    

DOI:10.1039/a604225b

出版商:RSC    

年代:1997

数据来源: RSC

摘要:

29Si NMR study on co-hydrolysis processes in Si(OEt)4–RSi(OEt)3–EtOH– water–HCl systems (R=Me, Ph): effect of R groups Yoshiyuki Sugahara,*a Tsuyoshi Inouea and Kazuyuki Kuroda*a,b aDepartment of Applied Chemistry, School of Science and Engineering, Waseda University, Ohkubo-3, Shinjuku-ku, T okyo 169, Japan bKagami Memorial L aboratory for Materials Science and T echnology, Waseda University, Nishi-waseda-2, Shinjuku-ku, T okyo 169, Japan Hydrolysis and initial polycondensation processes in the Si(OEt)4 (TEOS)–RSi(OEt)3–EtOH–water–HCl systems (R=Me, Ph) [TEOS5RSi(OEt)35EtOH5water5HCl=1515245x5y (x=12, 2/3; y=2×10-3, 4×10-3, 8×10-2)] have been investigated by using 29Si NMR spectroscopy.For comparison, the alkoxysilanes [TEOS, MeSi(OEt)3 (MTES), PhSi(OEt)3 (PTES)] were hydrolysed separately in a similar manner.In a water-rich TEOS–MTES system (water5Si=1251), a silanol-terminated codimer [Me(HO)2SiOSi(OH)3] was detected as well as (HO)3SiOSi(OH)3 and Me(HO)2SiOSi(OH)2Me, suggesting that the hydrolysed monomers were condensed rather randomly. In contrast, in a water-rich TEOS–PTES system, hydrolysed monomers derived from TEOS and those from PTES were condensed independently; a silanol-terminated codimer [Ph(HO)2SiOSi(OH)3 ] did not form, and only (HO)3SiOSi(OH)3 and Ph(HO)2SiOSi(OH)2Ph were detected.These observations in the TEOS–PTES system suggest the presence of the association of the phenyl groups. In water-restricted systems (water5Si=153), monomers were only partially hydrolysed, and ethoxy-terminated codimers [R(EtO)2SiOSi(OEt)3 (R=Me, Ph)] formed in both the TEOS–MTES and the TEOS–PTES systems.A new family of sol–gel-derived inorganic–organic composite tified.16,17 In the co-hydrolysis process of the TEOS–methyltriethoxysilane [MeSi(OEt)3; MTES] system, we have identified materials called ORMOCERs (organically modified ceramics) two 29Si NMR signals due to (HO)2MeSiOSi(OH)3, as [or ORMOSILs (organically modified silicates)] have attracted well as those due to (HO)2MeSiOSiMe(OH)2 and increasing attention.1–3 They can be prepared by two strategies: (HO)3SiOSi(OH)3.37 Similar results were independently the co-hydrolysis of tetraalkoxysilane [Si(OR)4 ] and organoalreported for other TEOS–MTES systems.39,40 On the other koxysilane [R¾Si(OR)3 or R¾RSi(OR)2]1–3 or the hydrolysis hand, Kim et al. studied the co-hydrolysis process of TEOS of tetraalkoxysilane in the presence of silanol-terminated poly and a chromophore-tagged trialkoxysilane {N-[3-(triethoxysi- (dimethylsiloxane).4–6 Co-hydrolysis processes are advantalyl) propyl-2,4-dinitrophenylamine; TDP}, and revealed that geous for the preparation of advanced materials, since the cross-linking between the TEOS-derived and TDP-derived trialkoxysilanes where functional groups are covalently species was limited.38 Very recently, Babonneau et al.utilized attached to silicon can be used; typically chromophores and liquid-state 17O NMR instead of 29Si NMR spectroscopy, and ligand groups available for coordination to metals have been successfully identified signals due to bridging oxygen for incorporated into silica matrices by co-hydrolysis processes.7 OSiMe2OSi(OM)3 (D–Q) and (MO)2SiMeOSi(OM)3 (T–Q) Another recently developed type of material derived from units.41 organoalkoxysilanes is the interpenetrating network (IPN), This paper describes the 29Si NMR study on the co- where organic polymers are incorporated into sol–gel glass hydrolysis processes of TEOS and organotriethoxysilanes structures.8,9 [RSi(OEt)3; R=Me or Ph], and aims to clarify the effect of When two kinds of alkoxysilanes are co-hydrolysed, the the R groups on co-condensation between TEOS and properties of the final xerogels should depend on their homo- RSi(OEt)3.We focus on the dimer formation behaviour, geneity. If two alkoxysilanes condense mainly independently and discuss the tendency of condensation based on the after their hydrolysis, the products should possess segregated hydrophobicity of the R groups.domain structures, while homogeneous materials could be obtained if hydrolysed species are condensed randomly. Solidstate 29Si NMR spectroscopy is capable of investigating the structures of xerogels, so that several solid-state NMR studies Experimental have been reported on ORMOCERs.10–20 Because of Sample preparation insufficient resolution, however, the discussion on the homogeneity is very limited.Thus, as far as we know, only when TEOS, MTES and phenyltriethoxysilane [PhSi(OEt)3 ; PTES] the two-dimensional 1H–29Si correlation CP MAS NMR tech- were used as received. No hydrolysed products were detected nique was applied,18,19 could detailed structural information by 29Si NMR spectroscopy.Ethanol was dehydrated by molbe obtained. ecular sieves before use. HCl was added as a 0.1 or 1 mol dm-3 Liquid-state 29Si NMR is also a powerful tool, in particular aqueous solution. for the hydrolysis and initial polycondensation process of Reactions were conducted in TEOS–RSi(OEt)3 (MTES or tetraalkoxysilanes21 and organoalkoxysilanes,22–35 and it has PTES)–EtOH–water (D2O/H2O)–HCl systems.Alkoxysilanes also been applied to co-hydrolysis processes.11,14,16,17,33,35–40 were used separately or as 151 mixtures; thus five different The initial polycondensation process of a tetraethoxysilane systems (TEOS–MTES, TEOS–PTES, TEOS, MTES, PTES) [Si(OEt)4; TEOS]–dimethyldiethoxysilane [Me2Si(OEt)2; were prepared.Two alkoxysilane5water ratios (water5Si= DMDES] system has been studied extensively, and the signals 1251, 153) were applied for every system. When water5Si= 1251, reactions were conducted in systems with due to D–Q [OSiMe2OSi(OM)3] bonds have been iden- J. Mater. Chem., 1997, 7(1), 53–59 53alkoxysilane5EtOH5water5HCl=25245245y (y=2×10-3, Results 4×10-3, 8×10-2).D2O and hydrochloric acid were added The environments of the silicon atoms are represented using dropwise to a mixture of alkoxysilanes and ethanol, and the the number of siloxane bonds and hydroxy groups. For TEOS- resulting homogeneous solution was stirred for 3 min. For the derived units, an environment Si(O0.5)n(OH)m(OEt)4-n-m is systems with water5Si=153, samples were prepared with represented as Qn(mOH).For units derived from MTES and alkoxysilane5EtOH5water5HCl=252452/354×10-3. A mix- PTES, an environment RSi(O0.5)n(OH)m(OEt)3-n-m (R=Me, ture of ethanol, H2O and hydrochloric acid was added to the Ph) is expressed as TMn(mOH) (R=Me), TPn(mOH) (R=Ph), or alkoxysilane, and the resulting solution was heated at reflux Tn(mOH) (for both).for 2 days under a protective nitrogen atmosphere. Reactions in water-rich systems (water5Si=1251) were conducted to hydrolyse most of the ethoxy groups before condensation. Thus, condensation should proceed mainly via a water- 29Si NMR measurements producing reaction: Silicon-29 NMR spectra were obtained using a JEOL NMOSiOH+ HOSiO�OSiOSiO+H2O GSX-400 spectrometer (9.4 T) operated at 79.42 MHz.A sample solution was put in a 10 mm o.d. glass tube and a trace Fig. 1 shows typical 29Si NMR spectra (collected after ca. 3 h of Cr(CH3COCHCOCH3 )3 was added to the sample solution hydrolysis) for the five systems (TEOS–MTES, TEOS–PTES, to reduce the 29Si spin–lattice relaxation time (T1). For samples TEOS, MTES and PTES), and enlarged profiles of the T1 and with water5Si=153, since no D2O was used, a small amount Q1 regions are demonstrated in Fig. 2. The signals in the of CDCl3 was added to the sample solution to obtain lock TEOS and MTES systems [Fig. 2(a) and 2(c)] are reasonably signals. The spectra were essentially similar to those obtained assigned to monomers and oligomers on the basis of previous without the addition of CDCl3.Measurements were performed reports (see Table 1).21,28 The time evolution of spectra for the using inversely gated decoupling. The repetition time was 20 s PTES system (not shown) is very similar to that of the MTES and p/2 puls were used. For qualitative measurements in system,28 although the signal region is shifted considerably water-rich systems (water5Si=1251), only 12 FIDs were upfield from that of the MTES system.Thus, these signals are accumulated, since the distributions of species changed even assigned in the way used for the MTES system.28 Assignments after 3 h. In contrast, the spectra for qualitative measurements for monomeric species derived from PTES are based on in water-restricted systems (water5Si=153) were accumuprevious work.32 These assignments are summarized in Table 1. lations of 128 FIDs, since the distributions of species barely When the 151 mixture of TEOS and MTES is hydrolysed, changed after refluxing for 2 days with the minimum amount the spectrum is not a simple overlay of the spectra of the of water.For quantitative analyses of the water-rich systems, TEOS and MTES systems; two new intense signals (AB1 and 24 (hydrolysis time up to ca. 3 h) and 48 (hydrolysis times ca. AB2) are clearly observed in Fig. 2(b), and are assigned to 4 and ca. 6 h) FIDs were accumulated. The spectra in water- TM1(2OH)Q1(3OH).37 In contrast, the spectrum for the rich systems were characterized by using the time elapsing TEOS–PTES system [Fig. 2(d)] is essentially an overlay of from the end of the addition of hydrochloric acid to the middle those of the TEOS and PTES systems [Fig. 2(c) and 2(e)]; no of the measurement period. Chemical shifts were reported with respect to internal tetramethylsilane. detectable new signals appear in the TP1 and Q1 regions. Even Fig. 1 Silicon-29 NMR spectra of the solution in the system with alkoxysilane: EtOH5D2O/H2O5HCl=252452454×10-3 after hydrolysis for ca. 3 h. (a) MTES only; (b) MTES5TEOS=151 mixture; (c) TEOS only; (d) PTES5TEOS=151 mixture; (e) PTES only. 54 J. Mater. Chem., 1997, 7(1), 53–59Fig. 2 Enlarged profile (T1 and Q1 region) of Fig. 1. (a) MTES only; (b) MTES5TEOS=151 mixture; (c) TEOS only; (d) PTES5TEOS=151 mixture; (e) PTES only. Table 1 Assignments of labelled 29Si NMR signals formonomeric and oligomeric species obtained by the hydrolysis of alkoxysilanes (water5Si=1251) signal structure formula A1 TM0(3OH) MeSi(OH)3 A2 TM0(2OH) MeSi(OH)2(OEt) A3 TM1(2OH)TM2(1OH)TM1(2OH) Me(HO)2SiOMe(HO)SiOSi(OH)2Me A4 TM1(2OH)TM1(2OH) Me(HO)2SiOSi(OH)2Me A5 TM1(2OH)TM1(1OH) Me(HO)2SiOSi(OH)(OEt)Me A6 TM1(2OH)TM1(1OH) Me(HO)2SiOSi(OH)(OEt)Me A7 TM1(2OH)TM2(1OH)TM1(2OH) Me(HO)2SiOMe(HO)SiOSi(OH)2Me B1 Q0(4OH) Si(OH)4 B2 Q0(3OH) Si(OH)3(OEt) B3 Q1(3OH)Q2(2OH)Q1(3OH) (HO)3SiO(HO)2SiOSi(OH)3 B4 Q1(3OH)Q1(3OH) (HO)3SiOSi(OH)3 B5 Q1(3OH)Q1(2OH) (HO)3SiOSi(OH)2(OEt) B6 [Q2(2OH)]3cyc [(HO)2SiO]3 B7 Q1(3OH)Q1(2OH) (HO)3SiOSi(OH)2(OEt) B8 [Q2(2OH)]4cyc [(HO)2SiO]4 B9 Q1(3OH)Q2(2OH)Q1(3OH) (HO)3SiO(HO)2SiOSi(OH)3 C1 TP0(3OH) PhSi(OH)3 C2 TP0(2OH) PhSi(OH)2(OEt) C3 TP0(1OH) PhSi(OH)(OEt)2 C4 TP1(2OH)TP2(1OH)TP1(2OH) Ph(HO)2SiOPh(HO)SiOSi(OH)2Ph C5 TP1(2OH)TP1(2OH) Ph(HO)2SiOSi(OH)2Ph C6 TP1(2OH)TP1(1OH) Ph(HO)2SiOSi(OH)(OEt)Ph C7 TP1(2OH)TP1(1OH) Ph(HO)2SiOSi(OH)(OEt)Ph C8 TP1(2OH)TP2(1OH)TP1(2OH) Ph(HO)2SiOPh(HO)SiOSi(OH)2Ph AB1 TM1(2OH)Q1(3OH) Me(HO)2SiOSi(OH)3 AB2 TM1(2OH)Q1(3OH) Me(HO)2SiOSi(OH)3 when the amount of HCl was increased to 8×10-2 (not [Fig. 3(a)].For the TEOS–PTES system [Fig. 3(b)], the hydrolysis behaviour of TEOS is similar to that of the shown), no evidence for the presence of codimer was obtained. Fig. 3 illustrates the quantitative results for the systems of TEOS–MTES system. In contrast, hydrolysis of PTES is slower than that of MTES, and only ca. 40% of the PTES the 151 mixtures of alkoxysilanes [TEOS5MTES (or PTES)5EtOH5water5HCl=15152452452×10-3; Fig. 3(a) is hydrolysed completely to form TP0(3OH) after 19 min hydrolysis. and (c) for the TEOS–MTES system and Fig. 3(b) and (d) for the TEOS–PTES system]. After hydrolysis for 19 min, the As the reactions proceed, hydrolysed monomers are involved in condensation reactions to form oligomers and polymers. In amounts of oligomers are very small [Fig. 3(c)].At this point, nearly 70% of the MTES is completely hydrolysed to form the time range investigated, T1, T2, Q1 and Q2 environments are mainly detected as condensed ones, and branched environ- TM0(3OH), while less than 50% of the TEOS is present as Q0(4OH) J. Mater. Chem., 1997, 7(1), 53–59 55Fig. 3 Time evolution of signals due to monomeric and condensed species in the system with TEOS5RSi(OEt)35EtOH5D2O/H2O5HCl= 15152452452×10-3.(a) monomeric species for the MTES5TEOS=151 mixture; (b) monomeric species for the PTES5TEOS=151 mixture; (c) condensed species for the MTES5TEOS=151 mixture; (d) condensed species for the PTES5TEOS=151 mixture. ments (T3, Q3 and Q4) are hardly detected. From the results mainly form: presented in Fig. 3(b) and (d), we can calculate the average OSiOR+HOSiO�OSiOSiO+ROH number of siloxane bonds per silicon atom, as shown in Fig. 4. Clearly, the condensation rate of PTES-derived species is much Since only weak signals are observed in the T2 and Q2 regions, lower than those of TEOS- and MTES-derived species. and no signals are detected in the T3 and Q3 regions, monomers We also prepared dimers with a minimum amount of water and dimers are the dominant species in these systems.In the (water5Si=153). Fig. 5 represents the whole 29Si NMR spectra separate alkoxysilane systems, unhydrolysed monomers [a1, of the five systems, and their enlarged profiles for the T1 and TM0(0OH); b1, Q0(0OH); c1, TP0(0OH)] are clearly detected. In Q1 regions are shown in Fig. 6. Since the amount of silanol addition, one intense signal is observed for each system in groups is very restricted, the alcohol-producing condensation the T1 and Q1 regions, and should be ascribed to ethoxy- should be dominant, thus ethoxy-terminated species should terminated dimers [a2, TM1(0OH)TM1(0OH); b2, Q1(0OH)Q1(0OH); c2, TP1(0OH)TP1(0OH)] which can form by condensation between an unhydrolysed monomer and a hydrolysed species possessing one hydroxy group [from T0(0OH) and T0(1OH) or from Q0(0OH) and Q0(1OH)]. In the spectrum of the TEOS–MTES system, in addition to the signals observed in the TEOS and MTES systems, two new intense signals appear, and are assigned to an ethoxy-terminated codimer [TM1(0OH)Q1(0OH)].40 Similarly, two new signals are detected for the TEOS–PTES system (cb1 and cb2), which should also be ascribed to an ethoxy-terminated codimer [TP1(0OH)Q1(0OH)].Table 2 summarizes these assignments. Discussion When the water5Si ratio is 1251, the hydrolysis and condensation behaviour in the TEOS–MTES system is different from that in the TEOS–PTES one. The hydrolysis rate of MTES is Fig. 4 Time evolution of the average number of SiOSi bonds formed much larger than those of TEOS and PTES.Under acidic per Si atom in the system with TEOS5RSi(OEt)35EtOH5 conditions, the hydrolysis of alkoxysilanes is reported to be D2O/H2O5HCl=15152452452×10-3 (R=Me, Ph). $, TM; #, Q(TEOS–MTES); +, TP; ', Q(TEOS–PTES). initiated by the fast protonation of a leaving alkoxy group, 56 J. Mater. Chem., 1997, 7(1), 53–59Table 2 Assignments of labelled 29Si NMR signals for nomomeric and derived species should be ascribed to the bulkiness of the dimeric species obtained by the hydrolysis of alkoxysilanes phenyl groups.(water5Si=153) When only one alkoxysilane is hydrolysed with sufficient water, the distribution of dimers reflects the distribution of signal structure formula monomers.27,28 In the TEOS–MTES system, although MTES a1 TM0(0OH) MeSi(OEt)3 (MTES) is hydrolysed faster than TEOS, fully hydrolysed monomers a2 TM1(0OH)TM1(0OH) Me(EtO)2SiOSi(OEt)2Me [Q0(4OH) and TM0(3OH)] are the dominant monomeric species.As dimers, TM1(2OH)TM1(2OH), Q1(3OH)Q1(3OH) and TM1(2OH)Q1(3OH) b1 Q0(0OH) Si(OEt)4 (TEOS) b2 Q1(0OH)Q1(0OH) (EtO)3SiOSi(OEt)3 are mainly observed, consistent with the monomer distribution; thus monomeric species are condensed rather randomly in c1 TP0(0OH) PhSi(OEt)3 (PTES) this system.c2 TP1(0OH)TP1(0OH) Ph(EtO)2SiOSi(OEt)2Ph In the TEOS–PTES system, the fully hydrolysed monomers ab1 TM1(0OH)Q1(0OH) Me(EtO)2SiOSOEt)3 [Q0(4OH) and TP0(3OH)] are also dominant when dimers start to ab2 TM1(0OH)Q1(0OH) Me(EtO)2SiOSi(OEt)3 form. Although the condensation rate of the PTES-derived cb1 TP1(0OH)Q1(0OH) Ph(EtO)2SiOSi(OEt)3 species is much slower, the estimated Q0(4OH)/TP0(3OH) values cb2 TP1(0OH)Q1(0OH) Ph(EtO)2SiOSi(OEt)3 range from 1.4 (after 64 min) to 2.4 (after 184 min) at the initial stage.Moreover, based on the steric hindrance to the condensation of PTES-derived species,the formationof TP1(2OH)Q1(3OH) and subsequent nucleophilic attack of water molecule leads to should be much easier than that of TP1(2OH)TP1(2OH). Thus, if a five-coordinate transition state.42,43 The substitution of an we assume random condensation in the TEOS–PTES system, alkoxy group with an alkyl group increases the hydrolysis rate the observed distribution of the dimers [the absence of by stabilizing the development of positive charge by providing TP1(2OH)Q1(3OH) and the presence of TP1(2OH)TP1(2OH)] cannot electrons (the so-called polar effect).In addition, if the trans- yet be interpreted. ition state becomes sterically more crowded, the hydrolysis Since phenyl groups are highly hydrophobic, immiscible rate is reduced (the so-called steric effect). Thus, the substisolutions were obtained for a wide range of compositions in tution of the ethoxy groups with methyl groups should result the TEOS–PTES system.With the present composition in increased hydrolysis rates according to the polar effect.42 In (TEOS5PTES5EtOH5water5HCl=15152452454×10-3), a contrast, substitution with the bulky phenyl group has very certain amount of water should be still present with ethanol little effect on the hydrolysis rate, probably because of the (only 7 mol of water is consumed for the complete hydrolysis combination of these two effects.of 1 mol of TEOS and 1 mol of PTES), and additional water The acid-catalysed condensation rates of TEOS- and MTESforms from the condensation of two silanol groups. Compared derived species are comparable, whereas that of PTES-derived with silicic acid [Q0(4OH); Si(OH)4], TP0(3OH) [PhSi(OH)3] is species is much lower (Fig. 4). Acid-catalysed condensation highly hydrophobic because of the presence of the phenyl reactions can involve protonated silanol groups (OSiOH2+); groups. Thus, the hydrophobic effect of the phenyl groups thus, the polar effect can change the condensation rates.42 should lead to their association. Hence, although a homo- Since the steric effect is reported to be important for condengeneous solution was obtained with the present composition, sation reactions of silanol groups,42,43 however, the conden- PTES-derived species appear to be associated and the forma- sation reactions in the present system appear to be controlled sterically.Thus, the observed slow condensation rate of PTES- tion of TP1(2OH)Q1(3OH) is very limited.Fig. 5 Silicon-29 NMR spectra of the solution in the system with alkoxysilane5EtOH5D2O/H2O5HCl=252452/354×10-3 after refluxing for 2 days. (a) MTES only; (b) MTES5TEOS=151 mixture; (c) TEOS only; (d) PTES5TEOS=151 mixture; (e) PTES only. J. Mater. Chem., 1997, 7(1), 53–59 57Fig. 6 Enlarged profile (T1 and Q1 region) of Fig. 5.(a) MTES only; (b) MTES5TEOS=151 mixture; (c) TEOS only; (d) PTES5TEOS=151 mixture; (e) PTES only. This assumption is supported by the results for water- the TEOS–PTES system with the large amount of water (TEOS5PTES5EtOH5water=151524524), the dimer restricted systems. In addition to T1(0OH)T1(0OH) and Q1(0OH)Q1(0OH) codimers, T1(0OH)Q1(0OH) dimers are clearly possessing TP–Q bonds [(MO)2SiPhOSi(OM)3] is not detected, and only those with TP–TP and Q–Q bonds are detected in both the TEOS–MTES and TEOS–PTES systems. In these systems, most of the water is consumed and is finally detected. If the amount of water is very small (TEOS5PTES5EtOH5water=15152452/3), however, the converted into ethanol via condensation.Moreover, the average number of hydroxy groups attached to silicon is very dimer possessing TP–Q bonds is also obtained.These results suggest that, in the water-rich system, the hydrophobic phenyl small, thus all the species should be hydrophobic. These conditions can allow the homogeneous distribution of the groups are associated in solution, while the methyl groups appear to be randomly dispersed. Hence, in the co-hydrolysis monomers in the solutions and random condensation should occur even in the TEOS–PTES system.processes of alkoxysilanes, compositions, in particular the amount of water, should be selected carefully based on the These results indicate that, if a large amount of water is present (water5Si=1251) in the TEOS–PTES system, hydro- hydrophobicity of the organic groups attached to silicon.lysed monomers tend to aggregate as reactions proceeded. Thus, although a homogeneous gel was obtained by ageing This work is partly supported by Waseda University as a special research project. the uncovered water-rich solution of the TEOS–PTES systems, a segregated domain structure may form. Two-dimensional 1H–29Si correlation CP MAS NMR has been applied to both References the TEOS–MTES and the TEOS–PTES systems, and homogeneous structures were suggested based on the clear evidence 1 H.Schmidt, J. Non-Cryst. Solids, 1985, 73, 681. 2 H. K. Schmidt, ACS Symp. Ser., 1988, 360, 333. for the presence of coupling between Q4 silicon and organic 3 H. K. Schmidt, Mater. Res. Soc. Symp. Proc., 1990, 180, 961. protons (those of methyl and phenyl groups).18,19 In the 4 G.L. Wilkes, B. Orler and H. H. Huang, Polym. Prepr., 1985, reported TEOS–PTES system,19 the amount of water is equiv- 26, 300. alent to that of alkoxy groups. Thus, in contrast to the present 5 H. H. Huang, B. Orler and G. L. Wilkes, Polym. Bull., 1985, 14, 557. water-rich TEOS–PTES system, co-condensation can fre- 6 H. H. Hsin, B. Orler and G. L. Wilkes, Macromolecules, 1987, 20, 1322. quently occur and a certain amount of TP–Q bonds formed. 7 U. Schubert, N. Hu�sing and A. Lorenz, Chem. Mater., 1995, 7, 2010. Conclusions 8 B.M. Novak, Adv. Mater., 1993, 5, 422. 9 A. Morikawa, Y. Iyoku, M. Kakimoto and Y. Imai, J. Mater. We have investigated the hydrolysis and initial polycondens- Chem., 1992, 2, 679. ation process of Si(OEt)4–RSi(OEt)3–EtOH–water–HCl sys- 10 N.Nishiyama, K. Horie and T. Asakura, J. Appl. Polym. Sci., 1987, 34, 1619. tems (R=Me, Ph), and have discussed the effect of the 11 F. Babonneau, K. Thorne and J. D. Mackenzie, Chem. Mater., R groups on condensation reactions on the basis of the 1989, 1, 554. dimer formation behaviour. In the TEOS–MTES system, 12 Y. T. Lee, K. Iwamoto, H. Sekimoto and M. Seno, J. Membr.Sci., irrespective of the amount of water, the dimers possessing 1989, 42, 169. TM–Q bonds [(MO)2SiMeOSi(OM)3] are formed 13 R. H. Glaser, G. L. Wilkes and C. E. Bronnimann, J. Non-Cryst. as well as those possessing TM–TM [(MO)2SiMeOSiMe(OM)2] Solids, 1989, 113, 73. 14 M. P. Besland, C. Guizard, N. Hovnanian, A. Larbot, L. Cot, and Q–Q [(MO)3SiOSi(OM)3 ] bonds. On the other hand, in 58 J.Mater. Chem., 1997, 7(1), 53–59J. Sanz, I. Sobrados and M. Gregorkiewitz, J. Am. Chem. Soc., 29 S. Suda, M. Iwaida, K. Yamashita and T. Umegaki, J. Non-Cryst. Solids, 1994, 176, 26. 1991, 113, 1982. 15 W. G. Fahrenholts, D. M. Smith and D-W. Hua, J. Non-Cryst. 30 F. Brunet, P. Lux and J. Virlet, New J. Chem., 1994, 18, 1059. 31 R. C. Chambers, W. E. Jones, Jr., Y.Haruvy, S. E. Webber and Solids, 1992, 144, 45. 16 F. Babonneau, L. Bois and J. Livage,Mater. Res. Soc. Symp. Proc., M. A. Fox, Chem. Mater., 1993, 5, 1486. 32 T. Jermouni, M. Smaihi and N. Hovnanian, J.Mater. Chem., 1995, 1992, 271, 237. 17 F. Babonneau, L. Bois, J. Maquet and J. Livage, Eur. Mater. Res. 5, 1203. 33 G. D. Soraru`, G. D’Andrea, R. Campostrini and F. Babonneau, Monogr., 1992, 5, 319. 18 C. A. Fyfe, Y. Zhang and P. Aroca, J. Am. Chem. Soc., 1992, J. Mater. Chem., 1995, 5, 1363. 34 F. Rousseau, C. Poinsignon, J. Garcia and M. Popall, Chem. 114, 3252. 19 M. P. J. Peeters, W. J. J. Wakelkp and A. P. M. Kentgens, Mater., 1995, 7, 828. 35 L. Delattre and F. Babonneau, Mater. Res. Soc. Symp. Proc., 1992, J. Non-Cryst. Solids, 1995, 189, 77. 20 M. Smaihi, T.Jermoumi and J. Marignan, Chem. Mater., 1995, 346, 365. 36 M. J. Bommel, T. N. M. Bernards and A. H. Boonstra, J. Non- 7, 2293. 21 C. J. Brinker and G. W. Scherer, Sol–Gel Science, Academic Press, Cryst. Solids, 1991, 128, 231. 37 Y. Sugahara, Y. Tanaka, S. Sato, K. Kuroda and C. Kato, Mater. San Diego, 1990, pp. 152–174. Res. Soc. Symp. Proc., 1992, 271, 231. 22 K. A. Smith,Macromolecules, 1987, 20, 2514. 38 J. Kim, J. L. Plawsky, E. V. Wagenen and G. M. Korenowski, 23 F. Devreux, J. P. Boilot, F. Chaput and A. Lecomte, Phys. Rev. A, Chem. Mater., 1993, 5, 1118. 1990, 41, 6901. 39 C. A. Fyfe, P. P. Aroca and Y. Zhang, Bull. Magn. Reson., 1994, 24 R. Nass, E. Arpac, W. Glaubitt and H. Schmidt, J. Non-Cryst. 15, 195. Solids, 1990, 121, 370. 40 S. Prabakar, R. A. Assink, N. K. Raman and C. J. Brinker, Mater. 25 D. Hoebbel, I. Pitsch, H. Jancke and B. Costisella, Z. Anorg. Allg. Res. Soc. Symp. Proc., 1994, 346, 979. Chem., 1990, 588, 199. 41 F. Babonneau, J. Maquet and J. Livage, Chem. Mater., 1995, 7, 26 I. Hasegawa, S. Sakka, Y. Sugahara, K. Kuroda and C. Kato, 1050. Nippon Seramikkusu Kyokai Gakujyutu Ronbunshi (J. Ceram. Soc. 42 Ref. 21, pp. 116–152. Jpn.), 1990, 98, 647. 43 F. D. Osterholtz and E. R. Pohl, J. Adhesion Sci. T echnol., 1992, 27 Y. Sugahara, S. Okada, K. Kuroda and C. Kato, J. Non-Cryst. 6, 127. Solids, 1992, 139, 25. 28 Y. Sugahara, S. Okada, S. Sato, K. Kuroda and C. Kato, J. Non- Cryst. Solids, 1994, 167, 21. Paper 6/03741K; Received 29thMay, 1996 J. Mater. Chem., 1997, 7(1), 53–59 59

ISSN:0959-9428    

DOI:10.1039/a603741k

出版商:RSC    

年代:1997

数据来源: RSC