摘要:

Journal of Materials Chemistry Scientific Advisory Editor Professor Martin R. Bryce Department of Chemistry University of Durham South Road Durham DH1 3LE, UK Associate Editor Professor Jean Etourneau ICMCB Avenue du Docteur Schweitzer 33600 Pessac France Editorial board Allan E. Underhill (Chairman) Bangor Neil L. Allan Bristol Peter G. Bruce St. Andrews Martin R. Bryce Durham Michael J. Cook Norwich Managing Editor Janet L. Dean Deputy Editor Zoe G. Lewin Assistant Editor Graham F. McCann Editorial Secretary Miss D. J. Halls Jean Etourneau Bordeaux Wendy R. Flavell UMIST Colin Greaves Birmingham Philip Hodge Manchester Stephen M. Kelly Hull International advisory editorial board K. Bechgaard Riso, Denmark J.Y. Becker Beer-Sheva, Israel A. J. Bruce Murray Hill, USA E. Chiellini Pisu, Italy D. Coates Poole, UK P. Day London, UK D. A. Dunmur Shefield, UK B. Dunn Los Angeles, USA W. J. Feast Durham, UK R. H. Friend Cambridge, UK A. Fukuda Tokyo, Japan D. Gatteschi Florence, Italy J. W. Goodby Hull, UK Information for authors The Royal Society of Chemistry welcomes submission of manuscripts intended for publication in two forms, Articles and Materials Chemistry Communications. These should describe original work of high quality dealing with the synthesis, structures, properties and applications of materials, particularly those associated with advanced technology. Full papers contain original scientific work that has not been published previously.However, work that has appeared in print in a short form such as a Materials Chemistry Communication is normally acceptable. Four copies of Articles including a top copy with figures etc. should be sent to the Managing Editor at the Cambridge address. Journal of Materials Chemistry (ISSN 0959-9428) is published monthly by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF, UK. All orders accompanied with payment should be sent directly to The Royal Society of Chemistry, Turpin Distribution Services Ltd., Blackhorse Road, Letchworth, Herts SG6 lHN, UK. NB Turpin Distribution Services Ltd., distributors, is wholly owned by The Royal Society of Chemistry. 1996 Annual subscription rate EEA (incl.UK) E519.00, USA $934.00, Rest of World E532.00. Customers A. B. Holmes Cumbridge, UK H. Inokuchi Okazaki, Japan W. Jeitschko Miinster, Germany 0.Kahn Bordeaux, France J. Livage Paris, France R. McCullough Pittsburgh, USA J. S. Miller Salt Lake City, USA K. Mullen Mainz, Germany L. Niinisto Espoo, Finland M. Nygren Stockholm, Sweden Y. W. Park Seoul, Korea N. Plate Moscow, Russia M. Prato Trieste, Italy Materials Chemistry Communications contain novel scientific work in short form and of such importance that rapid publication is warranted. The total length is normally restricted to two printed A4 pages. However, special consideration will be given to communications with a large amount of essential diagrammatic information.Submission of a Materials Chemistry Communication can be made either to the Managing Editor at the Cambridge address, or via a member of the International Advisory Editorial Board. In the latter case, the top copy of the manuscript including any figures etc., together with the name of the person to whom the Communication is being submitted, should be sent simultaneously to the Managing Editor at the Cambridge address. All authors submitting work for publication are should make payments by cheque in sterling payable on a UK clearing bank or in US dollars payable on a US clearing bank. Air freight and mailing in the USA by Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11003.USA POSTMASTER: send address changes to Journal of Materials Chemistry, Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11003. Periodicals postage paid at Jamaica, NY 11431. All other dispatches outside the UK by Bulk Airmail within Europe, Accelerated Surface Post outside Europe. PRINTED IN THE UK. Editorial Production Coordinator Stephanie Shah Technical Editors Carole J. Nerney Alan J. Holder Jeff R. White Graphics Designer Ms C. Taylor-Reid Klaus Praefcke Berlin Anthony R. West Aberdeen Janet L. Dean (Secretary) C. N. R. Rao Bangalore, India B. Raveau Caen, France T. Rojo Bilbao, Spain J. Rouxel Nantes, France A. Simon Stuttgart, Germany M. A. Subramanian Wilmington, USA S.Takahashi Osaka, Japan J. 0.Thomas Uppsala, Sweden M. Vallet-Regi Madrid, Spain D. E. W. Vaughan Annandale, USA J. D. Wright Canterbury, UK Y. Yamashita Okuzaki, Japan required to sign an exclusive copyright licence. All submissions should be accompanied by a completed form (a blank for photocopying is reproduced at the end of the Information for Authors in Issue l), without which publication cannot proceed. A completed graphical abstract template should also accompany each submission. Full details of the form of manuscripts for Articles and Materials Chemistry Communications, conditions for acceptance etc. are given in Issue 1 of Journal of Materials Chemistry published in January of each year, on the world wide web (htpp://chemistry.rsc.org/rsc/) or may be obtained from the Managing Editor. There is no page charge for papers published in Journal of Materials Chemistry. Fifty reprints are supplied free of charge. Janet L. Dean, Managing Editor Tel.: Cambridge (01223) 420066 E-Mail (INTERNET): DEANJ@RSC.ORG Fax: (01223) 420247 Advertisement sales: Tel. +44 (0)171-287 3091; Fax +44 (0)171-494 1134 0The Royal Society of Chemistry, 1996. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form, or by any means, electronic, mechanical, photographic, recording, or otherwise, without the prior permission of the publishers.

ISSN:0959-9428    

DOI:10.1039/JM99606FX037

出版商:RSC    

年代:1996

数据来源: RSC

ISSN:0959-9428    

DOI:10.1039/JM996060X039

出版商:RSC    

年代:1996

数据来源: RSC

ISSN:0959-9428    

DOI:10.1039/JM99606BX043

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

ISSN 0959-9428 JMACEP( 11) 1741-1858 (1996) Synthesis, structures, properties and applications of materials, particularly those associated with advanced technology cn CONTENTS Articles 1741 Influence of the host-guest interactions on the ferroelectric properties of mononuclear orthopalladated metallomesogen mixtures Neil J. Thompson, Raquel Iglesias, JosC Luis Serrano, Maria Jesus Baena and Pablo Espinet 1745 Synthesis and magnetic behaviour of polyradical: poly (1,3-phenyleneethynylene) with -toPoregulated pendant stable aminoxyl and imine N-oxide-aminoxyl radicals Yozo Miura, Tsuneki Issiki, Yukio Ushitani, Yoshio Teki and Koichi Itoh 1751 Theoretical determination of the molecular and solid-state electronic structures of phthalocyanine and largely extended phthalocyanine macrocycles Enrique Orti, Raul Crespo, M.Carmen Piqueras and Francisco Tomas 1 763 Improved electroluminescence performance of poly (3-alkylthiophenes) having a high head-to-tail (HT) ratio Electroluminescenceof Freeman Chen, Parag G. Mehta, Larry Takiff and Richard D. McCullough head 'R R 1 1767 A water-resistant precursor in a wet process for TiO, thin film formation Mitsunobu Sato, Hiroki Hara, Toshikazu Nishide and Yutaka Sawada 177 1 Synthesis and characterization of 11-VI semiconductor nanoparticulates by the reaction of a metal alkyl polymer adduct with hydrogen sulfide Stephen W. Haggata, Xiaochang Li, David J. Cole-Hamilton and John R.Fryer 178 1 Crystal structure and magnetic properties of Ba,,(MnFeF,, -xClx)3FxCl,-x (x =0.85). Structural relationships with the apatite-type structure Jacques Darriet, Virginie Nazabal and Jean Fompeyrine 1785 The electronic and magnetic structures of stoichiometric SrCoO,: ASW calculations Samir F. Matar, Antoine Villesuzanne and Michael Uhl 1789 Synthesis and characterization of inorganic gels in a lyotropic liquid crystal medium. Part 2.-Synthesis of silica gels in lyotropic crystal phases obtained from cationic surfactants Thierry Dabadie, Andrk Ayral, Christian Guizard, Louis Cot and Pascale Lacan 1795 Electrochemical synthesis and thermal decomposition of zirconia gels containing various metal ions Kenichiro Nakajima, Shiro Shimada and Michio Inagaki coating-firing TiO, (anastase) thin film on glass w 2PysPB-Polymer Adduct zns Part& isotropicfluid sol ordered wet gel orderedporous oxide 11 1799 Synthesis and characterization of a shock- synthesized cubic B-C-N solid solution of composition BC,,N Tamikuni Komatsu, Masayuki Nomura, Youzou Kakudate and Shuzou Fujiwara 1805 Low-temperature stabilisation of tetragonal zirconia by antimony Antonino Gulino, Russell G.Egdell and Ignazio Fragala 181 1 Preparation and characterisation of high refractive index Pb0-Ti0,-Te02 glass systems Raul F. Cuevas, Ana M. de Paula, Oswaldo L. Alves, Norbert0 Aranha, Jos6 A. Sanjurjo, Carlos L. Cesar and Luiz C. Barbosa 1815 Comparison of X-ray patterns and Raman spectra of n =2 and 3 Aurivillius phases by principal component analysis Peter J.Klar, Limei Chen and Thomas Rentschler 1823 A neutron diffraction study of structural distortions in the Ruddlesden-Popper phase Na2La2Ti3OI0 Adrian J. Wright and Colin Greaves 1827 Synthesis and characterization of a novel layered titanium silicate JDF-L1 Hongbin Du, Min Fang, Jiesheng Chen and Wenqin Pang OB OC ON t~l~l~l~l~"l'l 170 175 180 185 190 195 200 205 binding energy/eV .I.--:+TeO4 ~e03 T a PCA a a-+ 1 -I 10 20 30 40 50 2Bldegrees ... 111 1831 The structure of the calcined aluminophosphate ALP04-5 determined by high-resolution X-ray and neutron powder diffraction Asilok J.Mora, Andrew N. Fitch, Michael Cole, Rajan Goyal, Richard H. Jones, Hervk Jobic and Stuart W. Carr 1 837 Space group symmetry and Al-0-P bond angles in Awo4-5 A. Rabdel Ruiz-Salvador, German Sastre, Dewi W. Lewis and C. Richard A. Catlow 1843 Structure of Zn(03PC2H4C02H) -0.5C6HsNH2 and XANES-EXAFS study of the intercalation of amines into Zn(O,PR)=H,O zinc alkylphosphonates Stkphanie Drumel, Pascal Janvier, Martine Bujoli-Doeuff and Bruno Bujoli 1849 Intercalation of 2-, 4-sulfanylpyridine, 2,2-and 4,4'-dithiobispyridine into VOP04and gel-V20s interlayer spaces Teruyuki Yatabe and Gen-etsu Matsubayashi Materials Chemistry Communications 1853 Solvate-switchable powder second harmonic generation in a push-pull quinonoid system 1064 nmM.Ravi, D. Narayana Rao, Shmuel Cohen, Israel Agranat and T. P. Radhakrishnan average oxygen 4 DPDQ Ib532nm drying It CHCl3 DPDQ1064 nm * mcy b 532 nm iv 1857 Synthesis of Al-intercalated montmorillonites using microwave irradiation "v G. Fetter, G. Heredia, A. M. Maubert and kP. Bosch + Ic'-el) i Cumulative Author Index iv Conference Diary Note: Where an asterisk appears against the name of one or more authors, it is included with the authors' approval to indicate that correspondence may be addressed to this person. COPIES OF CITED ARTICLES The Royal Society of Chemistry Library can usually supply copies of cited articles. For further details contact: The Library, Royal Society of Chemistry, Burlington House, Piccadilly, London W1V OBN, UK.Tel: +44 (0)171-437 8656, Fax: +44 (0)171-287 9798, Telecom Gold 84: BUR210, Electronic Mailbox (Internet) LIBRARY@RSC.ORG. If the material is not available from the Society's Library, the staff will be pleased to advise on its availability from other sources. Please note that copies are not available from the RSC at Thomas Graham House, Cambridge. V BRITISH LIQUID CRYSTAL SOCIETY ANNUAL CONFERENCE 1997 24 -26 March 1997 The British Liquid Crystal Society annual meeting will be held in Southampton and will last from noon Monday 24 March until noon Wednesday 26 March 1997. Accommodation will be in the Glen Eyre Hall of Residence complex, some 10 minutes walk away from the campus.SPEAKERS Sturgeon lecturer Professor A Fukuda (Tokyo Institute of Technology, Japan) Thresholdless antiferroelectricity in liquid crystals and its application to displays Other invited speakers: Professor C Rosenblatt (Case Western Reserve University, Cleveland, Ohio, USA) Exotic behaviour in antiferroelectric liquid crystals Professor C Zannoni (University of Bologna, Italy) Computer simulation of liquid crystals: advances and perspectives Professor J Goodby (University of Hull, UK) Supermolecules and supramolecular assemblies cost Members of BLCS E180; Non-members 5195; Students &115 These are all-in costs. If you want to come for part of the time, or stay elsewhere, please contact the organiser. Further details from: Professor T J Sluckin Faculty of Mathematical Studies University of Southampton SOUTHAMPTON SO17 1BJ UK Tel: (+44)( 1703)59-3680; secretary #5 150; fax #5 147 world-wide web: http://www.soton.ac.uk/blcs97/ e-mail: tj s @maths.soton. ac.uk

ISSN:0959-9428    

DOI:10.1039/JM99606FP108

出版商:RSC    

年代:1996

数据来源: RSC

ISSN:0959-9428    

DOI:10.1039/JM99606BP114

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Influence of the host-guest interactions on the ferroelectric properties of mononuclear orthopalladated metallomesogen mixtures Neil J. Thompson," Raquel Iglesias," Josk Luis Serrano,*" Maria Jesus Baenab and Pablo Espinet*b "Departamento de Quimica Organica, Facultad de Ciencias-ICMA, Universidad de Zaragoza-CSIC, 50009 Zaragoza, Spain bDepartamento de Quimica Inorganica, Facultad de Ciencias, Universidad de Valladolid, 47005 Valladolid, Spain A new series of four mononuclear orthopalladated complexes which contain a Schiff base and an aliphatic P-diketone as ligands is described. The miscibility of the modified complexes in a standard organic host material is improved in comparison with complexes studied previously. The increased solubility of the complexes leads to ferroelectric organometallic-organic mixtures with lower viscosities and switching times than purely organometallic mixtures, while the spontaneous polarization (P,) values are maintained or increased.Ferroelectric metallomesogens constitute an increasingly important area in the field of metallomesogen research. A number of papers dealing with this subject have recently appeared in the literature and these involve two main types of compound: compounds in which the chirality arises from the disposition of organic moieties around the metal atom(s), and compounds in which the ligands bear chiral terminal tails. In the first group, chiral dinuclear orthopalladated derivatives' and vanadyl Schiff-base derivatives2 have been reported as mixtures of diasteroisomers exhibiting ferroelectric properties.Other compounds, such as 1,3-disubstituted ferrocenes3 and butadiene-iron tricarbonyl com~lexes,~ appear to be very promising materials which fall into this category. In the second category mentioned above, Schiff-base derivatives of copper2 and palladium' as well as mono-and di-nuclear orthometallated derivatives of palladium and platinum have been We have recently described a series of mononuclear com-plexes with the structure shown in Fig. 1 and studied the relationship between the magnitude of the spontaneous polariz- ation (P,) and the position and number of chiral centres." It was found that the chiral centres, when situated in the diketone ligand (R3 and R, in Fig.l), do not contribute greatly to the P, of the system. Owing to the unusual geometry of the molecular core unit and its relatively large width, the complexes are not sufficiently miscible with standard organic hosts used ORz lab GO ClO ClO 2a ClO c8* ClO c8* ClO ClO c8* c8* ClO Fig. 1 Mononuclear ortho-palladated comple~es.~.'~ R,, R,, R, and R, =(R)-2-methylheptyl (c8*) or n-decyl (Cl,,). 'Compounds reported previously. Mixture data used for comparison in this article. Host for binary mixtures. for binary mixtures to enable P, values to be measured. The ferroelectric properties of these compounds were therefore studied in mixtures using the achiral complex (compound la in Fig. 1) as the host. These mixtures, although useful to compare the effect of structural changes on P,, are very viscous and consequently the response times are slow (3-4 ms).The problems of high viscosity and slow response time were also encountered in the pure ferroelectric compounds with the structure shown in Fig. 1. In an effort to overcome this problem, it was decided to synthesize compounds with the structure shown in Fig. 2. In these compounds the bulky rigid aromatic P-diketone ligand has been replaced with a far less bulky and more flexible aliphatic ligand. The aim of this modification was to reduce significantly the molecular width and to obtain ferroelectric complexes with lower rotational viscosities and faster switching times. Given the relatively low contribution to the P, of chiral centres in the P-diketone, it was hoped that the benefits gained by the reduction in molecular width would offset the loss of potential sites for the inclusion of additional chiral centres.The compounds were synthesized by the reaction of the dinuclear orthopalladated complex with the thallium salt of the P-diketone, as described in previous article^.^,^,^* Experimental Materials Owing to the repetitive nature of the synthesis of the complexes and their structural similarity, the synthesis of one representa- R1OWNOOR* Pd C10H21 C10H21 compound R, R, transition temperaturesrc Fig. 2 Structure and transition temperatures of the modified complexes containing an aliphatic diketone ligand. R, and R, =n-decyloxy (Clo) or (R)-2-methylheptyl (C,*) J. Muter.Chern., 1996, 6(11), 1741-1744 11741 tive example is described. The analytical and spectroscopic data of the remainder of the final compounds are given below. Synthesis and characterization of compound lb. To a stirred solution of di-p-chloro-bis [4-n-decyloxy-N-( 4-decyloxybenzyl- idene)aniline-C2,N] palladium(11)~ (0.5 g, 0.4 mmol) in dry dichloromethane was added a suspension of thallium p-diketonate (0.44 g, 0.8 mmol) in dry dichloromethane. The mixture was stirred at room temperature overnight. The pre- cipitate was filtered off and purified by crystallisation (ethyl methyl ketone-ethanol) to give a yellow solid. Yield 82%; 'H NMR (300 MHz, CDCl,): 6 0.83-0.88 (m, 12H), 1.2-1.5 (m, 56H), 1.6-1.8 (m, 8H), 2.11 (t, J 7.3 Hz, 2H), 2.28 (t, J 7.3 Hz, 2H), 3.94 (t, J 6.5 Hz, 2H), 4.05 (t, J 6.5 Hz, 2H), 5.30 (s, lH), 6.58 (dd,J 8.4, 2.5 Hz, lH), 6.84(d,J 8.9 Hz, 2H), 7.13 (d, J 2.4Hz, lH),7.25 (d,J 8.3 Hz, lH), 7.31 (d, J 8.9 Hz,2H), 7.94 (s, 1H); IR (Nujol): v/cm-' 1574, 1506, 1465, 1250.Anal. Calc. for C56H9,N0,Pd: C, 70.81; H, 9.80; N, 1.48. Found: C, 70.78; H, 9.63; N, 1.49%. Compound 2b. Yield 60%; 'H NMR (300 MHz, CDC1,): 6 0.8-0.9 (m, 12H), 1.2-1.5 (m, 57H), 1.6-1.8 (m, 4H), 2.11 (t, J 7.3 Hz, 2H), 2.28 (t, J 7.3 Hz, 2H), 4.05 (t, J 6.6 Hz, 2H), 4.31 (m, lH), 5.30 (s, lH), 6.58 (dd, J 8.3, 2.4Hz, lH), 6.82 (d, J 9.0 Hz, 2H), 7.14 (d, J 2.5 Hz, lH), 7.25 (d, J 8.3 Hz, lH), 7.31 (d, J 8.9 Hz, 2H), 7.95 (s, 1H); IR (Nujol): v/cm-' 1567, 1506, 1467, 1264.Anal. Calc. for C5,H8gN04Pd: C, 70.35; H, 9.66; N, 1.52. Found: C, 70.08; H, 9.40; N, 1.48%. Compound 3b. Yield 63%; 'H NMR (300 MHz, CDC1,): 6 0.8-0.9 (m, 12H), 1.2-1.5 (m, 53H), 1.6-1.8 (m, 8H), 2.11 (t, J 7.3 Hz, 2H), 1.28 (t, J 7.3 Hz, 2H), 3.94 (t, J 6.6 Hz, 2H), 4.52 (m, lH), 5.30 (m, lH), 6.56 (dd, J 8.3, 2.4 Hz, lH), 6.84 (d, J 8.9 Hz, 2H), 7.12 (d, J 2.4 Hz, lH), 7.25 (d, J 8.2 Hz, lH), 7.31 (d, J 8.9 Hz, 2H), 7.94 (s, 1H); IR (Nujol): v/cm-' 1568, 1509, 1465, 1260. Anal. Calc. for C5,H8,N0,Pd: C, 70.35; H, 9.66; N, 1.52. Found: C, 70.54; H, 9.42; N, 1.54%. Compound 4b. Yield 50%; 'H NMR (300 MHz, CDC1,): 6 0.8-0.9 (m, 12H), 1.2-1.5 (m, 54H), 1.6-1.8 (m, 4H), 2.11 (t, J 7.3 Hz, 2H), 2.27 (t, J 7.3 Hz, 2H), 4.32 (m, lH), 4.52 (m, lH), 5.30(s, lH), 6.56 (dd, J 8.5,2.4 Hz, lH), 6.83 (d,J 9.0 Hz, 2H), 7.13 (d, J 2.3 Hz, lH), 7.25 (d, J 8.3 Hz, lH), 7.30 (8.9, 2H), 7.94 (s, 1H); IR (Nujol): v/cm-' 1573, 1505, 1463, 1250.Anal. Calc. for C,,H,,NO,Pd: C, 69.88; H, 9.53; N, 1.57. Found: C, 70.71; H, 9.51; N, 1.54%. Methods Microanalysis was performed with a Perkin-Elmer 240-B microanalyser. IR spectra were recorded using a Perkin-Elmer 1600 (series FTIR) spectrometer in the 400-4000 cm-' spectral range. 'H NMR spectra were recorded on a Varian Unity 300 spectrometer operating at 300 MHz. The transition temperatures and the natures of the phases were determined by polarising optical microscopy using an Olympus BH 2 microscope with a Mettler FP-82 heating stage and temperature-control unit.Measurements of transition tem- peratures were carried out by differential scanning calorimetry using a Perkin-Elmer DSC-7 calorimeter calibrated with indium (156.6 "C) and tin (232.1 "C) and with scanning rates of 10 "C min-'. Spontaneous polarisation, switching times and rotational viscosities of the complexes and binary mixtures were deter- mined simultaneously by the triangular-wave method using 4 mm polyimide coated cells with indium tin oxide electrodes (Standish LCD). 1742 J. Muter. Chem., 1996, 6(11), 1741-1744 Results and Discussion The transition temperatures of the pure compounds are given in Fig. 2. Unfortunately, only the achiral complex (lb) exhibits a smectic C (S,) phase.Therefore direct comparisons between the physical properties of the pure compounds described here and the pure compounds with the structure shown in Fig. 1 cannot be made. Compounds 2b and 3b each exhibit a mono- tropic smectic A (S,) phase which crystallises ca. 10°C below the isotropic (I)-SA transition. The compound containing two chiral chains (4b) is not liquid crystalline. The absence of a chiral Sc phase forced us to investigate binary mixtures. Initially, binary mixtures were investigated using compound lb as the host material, but crystallisation shortly after the SA-S, transition made meaningful measure- ment of P, values impossible. Binary mixtures (20mol%) of compounds 2b, 3b and 4b were studied using as the host the achiral complex la which exhibits a large range of S, phase.In addition, owing to the reduced molecular width of the modified complexes, compounds 2b, 3b and 4b were miscible with the standard organic host 4-hexyloxyphenyl 4-decyloxy- benzoate'' (C 62.5 Sc 78.2 SA84.5 N 90.5 I) in low percentages (approximately 10 mol% mixtures). The ferroelectric properties of the mixtures were determined as described in previous paper~.~?~J' The ferroelectric properties of all the mixtures studied are shown in Table 1. The mixtures are represented as the com- pounds (2b, 3b and 4b in Table 1) along with a letter to designate the host system; e.g. mixture 2b.e is the mixture of compound 2b with the ester 4-hexyloxyphenyl 4-decyloxy- benzoate and 2b.k is the mixture of compound 2b with the k- shaped complex host shown in Fig.1 (compound la). In order to undertake a comparative discussion, the data corresponding to 20mol% mixtures of compounds 2a, 3a and 4a with the achiral host la are also included in Table 1. These blends are denoted as 2a.k, 3a.k and 4a.k, respectively. If the mixtures of the complexes shown in Fig. 2 with the palladium complex as host are considered (2b.k, 3b.k and 4b.k), a trend can be seen regarding the magnitude of the spontaneous polarization. The temperature dependence of the spontaneous polarization of these mixtures is shown in Fig. 3. Mixture 2b.k has the lowest P, value (5.2 nC ern-,) and the P, values for mixtures 3b.k and 4b.k show a slight increase (5.6 and 6.4 nC cm-2, respectively).This trend is the same, although clearly diminished, as that found in examples of both dinuclear7 and mononuclear"" orthopalladated derivatives described in the literature. If the results for the series nb.k are compared with those obtained for the series na.k (see Table l), a significant augmentation of the increase in P, between members of the series can be seen (3.3, 4.8 and 8.5 nC for mixtures 2a.k, 3a.k and 4a.k, respectively). The difference in P, values is a consequence of the rotational freedom of the aromatic ring in which the chiral chain is substituted. In compounds 2a and 2b the chiral chain is in the anilinic ring, which is relatively free to rotate. However, in compounds 3a and 3b the chiral chain is substituted in the aldehydic ring which, by virtue of the orthopalladated ring system, is unable to rotate with respect to the rest of the molecular core.This restricted rotation leads to more efficient coupling of the molecular dipoles and, consequently, to a higher spontaneous polarization. Compounds 4a and 4b, which contain two chiral centres, give rise to the highest Ps values. As far as the rotational viscosity and response times are concerned, there is no significant improvement when the mixtures na.k and nb.k are compared. This is to be expected given that the viscosity of the system is dependent mainly on the viscosity of the host material which constitutes 80% of the mixture. The ferroelectric properties of the mixtures of complexes in Table 1 Ferroelectric properties of binary mixtures of compounds of series na and nb in K-shaped achiral palladium complex la, k, and 4-n-hexyloxyphenyl 4-n-decyloxybenzoate, e 2b.k Clo C8* 20.1 * 122 * 116 -5.2 3.8 11.4 8.4 27 27 4.3 1.6 0.11 0.03 3b.k C,* C,, 21.3 -127 -112 * -5.6 4.1 12.8 9.4 26 26 3.5 1.7 0.11 0.04 4b.k Cs* Cs* 19.6 117 * 102 6.4 5.2 14.6 12.8 26 24 4.5 2.1 0.14 0.06 2b.e C,, C8* 11.2 76 * 75 * 70 * 3.7 2.3 9.5 5.9 23 23 1.6 0.6 0.03 0.01 3b.e C,* Clo 11.4 -74 -66 * 5.4 4.0 15.8 11.7 20 20 2.0 0.9 0.05 0.02 4b.e Cs* C8* 9.9 69 * 64 * 60 -8.1 5.5 23.7 16.7 20 19 1.6 0.9 0.03 0.01 2a.k C,, C,* 20.1 141 * 136 * 3.3 2.0 6.8 4.1 29 29 3.9 1.8 0.07 0.02 3a.k Cs* C,, 19.9 * 141 * 135 * 4.8 2.6 9.3 5.5 31 28 3.5 1.3 0.09 0.02 4a.k C,* C,* 19.6 .131 * 124 * 8.5 6.4 17.0 12.8 30 30 3.5 1.5 0.17 0.05 "C,, =n-decyl. C,* =(R)-2-methylheptyl. R, and R, refer to Fig. 1 and 2. b% molar mixtures. "Transition temperatures measured on cooling in electrooptical cells. dNormalized spontaneous polarization Po =P,/sinQ. 0 A 0 0 0 A0 0 0 0 AA 8811 ~~I~,~,,,,~,,.,,,,,,l,,,,l~,,,l ~ 80 85 90 95 100 105 110 115 120 TI'C Fig. 3 Temperature dependence of Ps for complex/complex binary mixtures nb.k n =2 (A), 3 (O), 4 (0) the organic host (4-n-hexyloxyphenyl 4-n-decyloxybenzoate) are also shown in Table 1 (2b.e, 3b.e and 4b.e) and the temperature dependence of the spontaneous polarization for these mixtures is shown in Fig.4. Once again, the dependence of the Ps on the position of the chiral centre is evident and the reasons for this have been outlined above. Owing to the far lower viscosity of the organic host in comparison to the organometallic host, the viscosity of these mixtures is lower in spite of the opposing effect due to the lower temperatures of the appearance of the S,* phase. 10 -8-Y6 6-: 2:\ a" 4-A 2-O.8. D Ao. Fig. 4 Temperature dependence of P, in complex/organic mixtures nb.e: n =2 (O), 3 (A), 4 (0) Consequently, the response times decrease with little or no variation in the spontaneous polarization. Indeed, the switch- ing times are of the order of 2-3 times faster when the organic host is used and are <2 ms.Another significant difference between the mixtures in the organic host and the organometallic host concerns the values of the tilt angles. Larger tilt angles are observed in the mixtures with the palladium complex as the host. In contrast, when the organic ester is used as the host, smaller molecular tilts are observed and, as a consequence, these mixtures appear to be more interesting as they have higher normalized Ps values (Po). An interesting point arises regarding the comparison between the three series of mixtures na.k, nb.k and nb.e. Although the chiral complexes (Fig. 2) in the organic host ,,~l(series nb.e) are present in lower percentages (10 mol%) than in the complex-complex mixtures (series na.k and nb.k) (20 mol%), the Ps values of the comparable compounds from series na.k and nb.e do not differ significantly; compare mixtures 2b.e and 2a.k (3.7 and 3.3 nC cm-2), 3b.e and 3a.k (5.4 and 4.8 nC cmF2) and 4b.e and 4a.k (8.1 and 8.5 nC cm-2) (possible reasons for this will be discussed later).On the other hand, the mixtures in series nb.k behave very differently to those in the other two series because, as discussed previously, only small differences are apparent between the three mixtures in this series (5.2, 5.6 and 6.4 nC cm-2, respectively). A possible explanation for this phenomenon arises from a consideration of the molecular structures involved in each mixture. The complexes represented in Fig. 2 can be considered as a single rod-like core unit (the Schiff base), as it is possible that the long alkyl chains align in a parallel arrangement along the molecular long axis.Similar behaviour has been observed in classical organic liquid crystals12 and in other metallomeso- gens.13 In mixtures nb.e, both the host system and the com- plexes under investigation have a calamitic structure which consists of a single rod-like core unit. In contrast, the structure of the palladium complex used in mixtures na.k is significantly different and consists of a k-shaped core which can, as a rough approximation, be considered as two parallel rod-like units; one being the Schiff base and the other the P-diketone ligand. Consequently, series nb.e and na.k can be considered to be mixtures between compounds with similar structures.In series nb.e the two components each have a single rod-like core unit and in series na.k each component has a 'bi-calamitic' core unit with two rod-like moieties. In each case, the similarity in the molecular core structure of each component could lead to better mixing and stronger intermolecular interactions between the constituents, giving rise to the marked trend observed in the effect of position and number of chiral tails on the magnitude of Ps, which is similar to that observed in the pure corn pound^.^^^^^^ In contrast, the structures of the two compo- nents in the series nb.k are markedly different. The molecular J. Muter. Chem., 1996,6(11), 1741-1744 1743 core units of the host and guest systems have significantly different widths and the alkyl chains on the P-diketone units are in different relative positions with respect to the core unit. These differences may lead to diminished molecular compati- bility and less efficient packing.As a consequence, the mixtures in series nb.k show broadly similar P, values and, although the trend of position and number of chiral centres is apparent, it is far less marked. As mentioned previously, another interesting aspect of these mixtures is the similarity in the P, values of the two series na.k (20 mol%) and nb.e (10 mol%) despite the different concentrations of the chiral material in the achiral host. A possible explanation for this observation arises when the structures of the core units of the compounds are considered.The unusual structures of both the guest and the host in mixtures na.k contain two rod-like core units per molecule. The chiral centre(s), however, are only present in the Schiff base ligand and so the P-diketone unit can effectively act as an achiral core unit. This effectively halves the concentration of core units bearing chiral centres. This could be responsible for the fact that 20% mixtures of the k-shaped complexes lead to P, values which are similar to those of the 10% mixtures of the essentially rod-like complexes nb.e. Conclusions In conclusion, switching times in binary ferroelectric mixtures containing mononuclear palladium complexes can be reduced considerably by modification of the molecular structure to improve miscibility with a standard, less viscous organic host.The trend in the dependence of the magnitude of Ps on the position and number of chiral centres is similar in each series of mixtures studied. However, in mixtures in which both the guest and host systems have a similarly shaped molecular core unit, the trend is far more marked, reflecting that seen in pure compounds. In mixtures between rod-like and k-shaped com- plexes the Ps values are not as sensitive to the position and number of chiral centres, a fact which may reflects the lower degree of compatibility between the guest and host systems. The natures of the core structures of the complexes also appear to have an effect on the magnitude of the spontaneous polarization.10% mixtures of essentially calamitic (rod-like) molecules in a calamitic host have Ps values similar to those of 20% mixtures in which the core of both the guest and the host contain two calamitic moieties. This implies that the number of rod-like moieties present influences the magnitude of Ps, rather than it simply being dependent on the molar percentage of chiral compound present. Owing to the unusual nature of the compounds described here, caution must be taken regarding this conclusion until further studies and a wider variety of compounds are investigated. This work was financed by the CICYT, Spain (projects MAT 94-0717-CO2-01, MAT 93-1460-CE). N. J. T. thanks the DGICYT (Spain) for a post-doctoral grant. References 1 P.Espinet, J. Etxebarria, M. Marcos, J. Perez, A. Remon and J. L. Serrano, Angew. Chem., Int. Ed. Engl., 1989,28, 1065. 2 M. Marcos, J. L. Serrano, T. Sierra and M. J. GimCnez, Angew. Chem., Int. Ed. Engl., 1992,32, 1471. 3 R. Deschenaux and J. Santiago, Tetrahedron Lett., 1994,35,2169. 4 L. Ziminski and J. Malthete, J. Chem. SOC., Chem. Commun., 1990, 1495. 5 M. Marcos, J. L. Serrano, T. Sierra and M. J. GimCnez, Chem. Mater., 1993,5, 1332. 6 M. J. Baena, P. Espinet, M. B. Ros, J. L. Serrano and A. Ezcurra, Angew. Chem., Int. Ed. Engl., 1993,32, 1203. 7 M. J. Baena, J. Barbera, P. Espinet, A. Ezcurra, M. B. Ros and J. L. Serrano, J. Am. Chem. SOC., 1994,116, 1899. 8 P. Espinet, J. Etxebarria, M. Marcos, M. A. Perez Jubindo, M. B. Ros and J. L. Serrano, Mater. Res. SOC. Symp. Proc., 1995, 392, 123. 9 M. Ghedini, D. Pucci, N. Scaramuzza, L. Komitov and S. Lagerwall, Adv. Mater., 1995,7, 659. 10 N. J. Thompson, J. L. Serrano, M. J. Baena and P. Espinet, Chem. Eur. J., 1996, 2, 186. 11 J. W. Goodby, J. S. Pate1 and E. J. Chin, Phys. Chem., 1987, 91, 5151. 12 W. Weissflog and D. Demus, Cryst. Res. Tech., 1983, 18,21; 1984, 19, 55; C. T. Imrie and L. Taylor, Liq. Cryst., 1989, 6, 1; P. Berdague, P. Judeinstein, F. Perez and J. P. Bayle, New J. Chem., 1995, 19, 275. 13 E. Campillos, M. Marcos and J. L. Serrano, J. Mater. Chem., 1993, 3, 1049. Paper 6/03691K; Received 28th May, 1996 1744 J. Muter. Chem., 1996, 6(11), 1741-1744

ISSN:0959-9428    

DOI:10.1039/JM9960601741

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

~ Synthesis and magnetic behaviour of polyradical: poly (1,3=phenyleneethynylene)with Ir-toporegulated pendant stable aminoxyl and imine N-oxide-aminoxyl radicals Yozo Miura,*" Tsuneki Issiki," Yukio Ushitani," Yoshio Tekib and Koichi Itohb "Department of Applied Chemistry, Faculty of Engineering, Osaka City University, Sumiyoshi-ku, Osaka 558, Japan bDepartment of Material Science, Faculty of Science, Osaka City University, Sumiyoshi-ku, Osaka 558, Japan The palladium-catalysed polycondensation of N-tert-butyl-N-( 3,5-diethynylphenyl)aminoxylwith 2-( 3,5-diiodophenyl)-4,4,5,5-tetramethyl-3-oxido-2-imidazolin-3-ium-l-yloxylin pyridine-triethylamine at room temperature afforded poly ( 1,3-phenyleneethynylene) with pendant aminoxyl and imine N-oxide-aminoxyl radicals as a light blue powder in 78-83 % yields.The number-average molecular weights (an)of the polyradical determined by GPC were 2670-3030 and the spin concentrations determined by EPR were 1.91-2.02 x 1021 spins 8-l. Both solution (CH,Cl,) and powder EPR measurements of the polyradicals gave a hyperfine-smeared single line spectrum, suggesting that spin-exchange narrowing took place due to the high spin concentrations of the polyradicals. The magnetic susceptibility (x)measurements of the polyradicals were carried out on a superconducting quantum interference device (SQUID) magnetometer in the temperature range of 1.8-300 K. The ~Tvs.T plots showed that the polyradicals were a paramegnetic species yielding a weak antiferromagnetic coupling (0 = -1.5 K) below 30 K.The absence of the through-bond ferromagnetic interaction is accounted for by masking of the interaction by the stronger through-space antiferromagnetic interactions. Of the many models proposed for the design of new, molecule- based magnetic materials, polyradical magnets have proved to be one of the most interesting.' Our approach to polyradical ferromagnets has developed along two paths: one is poly(ethyn- ylbenzenes) with n-toporegulated pendant stable free rad- R ical~~-~ withand the other is poly( 173-phenyleneethynylenes) n-toporegulated pendant stable free One of the most important requirements in the magnetic studies of poly- radicals is the synthesis of structurally well-defined polyradicals with high spin concentrations. We have therefore prepared polyradicals by the polymerization of monomers bearing a stable free radical moiety under mild conditions such that the radical sites are not decomposed.In a previous paper we reported the synthesis and mag- netic characterization of poly( 173-phenyleneethynylene)with pendant imine N-oxide-aminoxyl groups l.7Although the spin concentrations of the polyradicals were very high [1.12-1.33 x lo2' spins g-' (0.82-0.91 spin per repeating unit)] , their magnetic susceptibility measurements showed no significant magnetic interactions among the spins. This was ascribed to a lack of sufficient polarization of the unpaired electron spins in the 173-phenylenediethynylenecoupling units. Based on this result we decided to synthesize poly( 1,3-pheny- leneethynylene) bearing aminoxyl radicals 2 because these have a more extensively delocalized unpaired electron spin.Since aminoxyl radicals are less stable than imine N-oxide-aminoxyl radicals, some difficulties were anticipated in the synthesis of 2. We therefore investigated the palladium-catalysed polycon- densation of aminoxyl monomer 4 with 173-diiodo-5-tert-butylbenzene giving 2 as a preliminary study.' When the polycondensation was carried out in pyridine-triethylamine (1:l), the resulting polyradicals showed a low spin concen- tration of 0.49 spins per repeating unit, suggesting that signifi- cant decomposition of the aminoxyl radical sites took place during the polycondensation.However, when polycondens- ation was carried out in 1:4 triethylamine-pyridine, the resulting polyradical showed a high spin concentration of 0.86 spins per repeating unit. After this preliminary study, we investigated the synthesis of poly(l,3-phenyleneethynylene) with n-toporegulated pendant aminoxyl and imine N-oxide- aminoxyl stable radicals 3 by the palladium-catalysed polycon- 4 5 densation of aminoxyl monomer 4 with imine N-oxide-ami- J. Mater. Chem., 1996, 6(11), 1745-1750 1745 noxyl monomer 5 in triethylamine-pyridine (1: 4). This polyradical has delocalized electron spins at the aminoxyl sites and the unpaired electron spin centres are closer to each other than in 1 and 2. Herein we report the synthesis and magnetic characterization of 3.Results and Discussion Design of polyradicals According to the valence bond theory of topological symmetry of alternant hydrocarbon n-systems,10 the ground spin states (S)are predicted by eqn. (l), where n* is the number of starred atoms and n is the number of unstarred atoms (where adjacent atoms are respectively starred and unstarred, and identically denoted s= (n*-n)/2 (1) atoms are not adjacent to each other). In Fig. 1, all the atoms in the n-system of the corresponding repeating unit 6 are either starred or unstarred according to this rule. As seen from Fig. 1, n* is 10 and n is 8, predicting that the ground state of 6 is a triplet (S= 1). On the basis of this consideration it is expected that the intramolecular (through-bond) interactions between the n-unpaired electron spins in 3 are ferromagnetic.It is also important to ascertain for 4 and 5 that the unpaired electron spin is delocalized to the benzene ring; it is particularly important to evaluate the spin density at the meta positions because the connection of the radical units is made at the meta positions of the phenyl group. To date, extensive EPR studies on N-aryl-N-tert-butylaminoxylshave been undertaken," and for N-tert-butyl-N-phenylaminoxyl712 the hyperfine coupling constants have been determined as follows: = = 0.19, am-H= 0.08 mT. On the other hand, for 2-phenyl-3-oxido-2- imidazolin-3-ium-1-yloxyl8 the spin density distribution on the benzene ring has been determined by an 'H ENDOR/ TRIPLE resonance study,', a 'H NMR study,14 and the polarized neutron diffraction method.15 For example, the 'H ENDOR/TRIPLE resonance method gave the following hyperfine coupling constants: = 0.0518, am-H= -0.0292, up-H=0.0466 mT.When eqn. (2) is applied, the spin densities on the meta positions are estimated to be 0.033 (7) and 0.012 (8), respectively, from the hyperfine coupling constants. Therefore, Q Q B"'/"O. MewMeMe Me 7 8 Fig. 1 Prediction of the ground spin state of 6 by the starred and unstarred valence bond theory 1746 J. Muter. Chem., 1996, 6(11), 1745-1750 it is obvious that there are some spin densities at the meta positions of both monomers. UH = -2.4~~ (2) Synthesisof monomers 4 and 5 N-tert-Butyl-N-( 3,5-diethynylphenyl)aminoxyl4 was obtained by our previously reported method and purified by sublimation prior to use. The radical purity determined by EPR was 87%.2-( 3,5-Diiodophenyl)-4,4,5,5-tetramethyl-3-oxido-2-imida-zolin-3-ium-1-yloxyl 5 was prepared according to Scheme 1. Thus, treatment of p-toluidine with benzyltrimethylammonium dichloroiodate (BTMA ICl,) in CH,Cl, in the presence of CaCO, gave 2,6-diiodo-p-toluidine in 78% yield.16 Deamination of this compound was accomplished by treating the corresponding toluidine sulfate with NaNO, in EtOH- benzene, giving 3,5-diiodotoluene in 50% yield. 33-Diiodotoluene was then treated with N-bromosuccinimide (NBS) in 1,2-dichloroethane in the presence of catalytic amounts of benzoyl peroxide, affording 3,5-diiodobenzyl bro- mide in ca.100 yield. This benzyl bromide was then treated with hexamethylenetetraamine in CHCl, and the subsequent hydrolysis in aqueous HCl-AcOH gave 3,5-diiodobenzal-dehyde in 47% yield. 3,5-Diiodobezaldehyde was then allowed to react with 2,3-bis(hydroxyamino)-2,3-dimethylbutanein MeOH-THF to give 1,3-dihydroxy-2-( 3,5-diiodophenyl)- 4,4,5,5-tetramethyl-2-imidazolinein 55% yield.17 The resulting imidazoline compound was oxidized with PbO, in benzene- ethanol. Crystallization from hexane-benzene gave 5 as dark blue needles in 62% yield. The radical purity determined by EPR was ca. 100%. Polycondensation Based on previous results showing that the palladium-catalysed polycondensation of 4 with 1,3-diiodo-5-tert-butylbenzenein 1: 4 triethylamine-pyridine gave polyradical2 with a high spin concentration of 0.86 spins per repeating unit, the polycondens- ation of 4 with 5 was also carried out in the same triethylamine- pyridine (1: 4) solvent (Scheme 2).The procedure for the polycondensation was as follows: a mixture of 1 equiv. of 4 and 5, 0.05 equiv. of (PPh,),PdCl,, and 0.02 equiv. of CuI in y2 - 'Q1 ii / the IMe Me iii1 I CHO CH2Br Me*MeMe Me 5 Scheme1 Reagents and conditions: i, BTMA ICI,, CaCO,, CH,Cl,-MeOH, reflux; ii, H2S04, NaNO,, EtOH-benzene, reflux; iii, NBS, CH,ClCH,Cl, reflux; iv, (CH,),N4, CHCl,, reflux, then HCl, AcOH-H,O, reflux; v, HONHCMe,CMe,NHOH, MeOH-THF, room temp.; vi, PbO,, benzene-ethanol, room temp.(PPh&PdC12-CUl based on polystyrene, whose molecular size is different from 4+5 pyndine-Et3N -3 that of the polyradical. Polyradicals 3 are soluble in dimethylformamide (DMF), Scheme 2 triethylamine-pyridine (1:4) was stirred at 20 “C under nitro- gen. After 2-6 h, the mixture was poured into a large amount of methanol, and the light green-blue powder deposited was collected by filtration and dried in uacuo. The results are summarized in Table 1. Table 1 shows that the yields of the polyradical are in the range 78-83 YOand the number-average molecular weights (M,) of the polyradicals determined by GPC using polystyrene standards are 2670-3030, which correspond to a degree of polymerization of 5.5-6.9.It is also shown that the increases in the yields and M, are small after 2 h of polycondensation, indicating that the polycondensation is almost finished. We therefore did not attempt further prolonged polycondensation, since a gradual decrease in the radical concentration of the polyradicals obtained was observed with prolonged polycond- ensation (see Table 1). The IR spectra of the polyradical showed the complete disappearance of the absorption due to the stretching vibration of the C=CH bond (3250 cm-’) characteristic of monomer 4 and the appearance of the absorption (weak) due to the stretching vibration of the CEC bond (2200cm-l). It is therefore suggested that the polyradicals are terminated by iodines. To confirm this IR result, iodine content was deter- mined by combustion analysis of the polyradicals. The iodine percentages of 9.2-9.3 suggest that structure 9 (n=5) has iodine atoms at both ends, in good agreement with the results from the GPC analyses (n=5.5-6.9).The observed C, H and N analyses for the polyradicals [C, 66.2; H, 5.6; N, 8.0; I, 9.2% (run 2 of Table 1) and C, 66.0; H, 5.7; N, 8.2; I, 9.3% (run 3 of Table l)] also agree with the theoretical values calculated for 9 (C, 66.02; H, 5.61; N, 8.84; I, 9.43%). This good agreement of a, is somewhat surprising since the GPC calibration is dichloromethane and chloroform, and partially soluble in benzene and toluene to give a green-blue solution. However, they are insoluble in methanol and hexane.The spin concentrations of 3 were determined by EPR using dichloromethane as solvent and using 1,3,5-triphenylverdazyl as the reference radical. As shown by Table 1, the spin concen- trations of 3 are in the range 1.91-2.02 x 1O2l spins g-l, which corresponds to 1.40-1.48 spins per repeating unit. If the structure of polyradical 9 has two heavy iodine atoms, this increases to 1.55-1.64. The spin concentrations of the polyrad- ical were also determined by magnetic susceptibility measure- ments of the polyradical to be 2.07 x 1021 spins g-l, as mentioned below. Although it is obvious that some decompo- sition of the radical moieties occurs during the polycondens- ation, their spin concentrations are satisfactorily high. Although we do not know what kinds of decomposition reactions occur for the radical moieties during the polycondens- ation, the good agreement between the experimental and calculated elemental analyses suggests that serious decompo- sition of the polyradical moieties does not occur.The UV-visible spectra of 3, 4 and 5 were measured using dichloromethane as the solvent, and the results are shown in Fig. 2. The concentration of the polyradical 3 is calculated as the unit mole concentration, and E values are corrected based on the spin concentration determined by EPR spectroscopy. Monomer 4 shows absorption maxima at 500 (E 121), 416 (280) and 302 nm (12000), and monomer 5 at 633 (E 300), 592 (340), 371 nm (14600) and 357 nm (7750). The orange colour characteristic of the aminoxyl group is attributable to A,,, at 500nm, and the blue colour characteristic of the imine N-oxide-aminoxyl group is attributable to A,,, at 633 and 592 nm.On the other hand, the polyradical3 shows absorption maxima at 622 (E 450), 585 (500) and 370 nm (16200). These absorption maxima are found in the spectra of 4 and 5, and no bathochromic shifts in the visible region are observed. Since the polyradical is based on a cross-conjugated poly ( 1,3-phenyl--o”~N‘om Table 1 The results of the palladium-catalysed polycondensation of 4 eneethynylene) structure, the spectrum of the polyradical3 can be interpreted as a superposition of the spectra of 4 and 5.Me+RMe.. The solution EPR spectra of 3, 4 and 5 were measured at with 5” spin conc./spins g -Id run t/h yield (%) Mb*c (number of spins per repeating unit) 1 2 83 2670 (5.5) 2.02 x 10” (1.48, 1.64f) 2‘ 4 83 2900 (6.6) 1.99 x lo2’ (1.46, 1.62f)g 3“ 6 78 3030 (6.9) 1.91x lo2’ (1.40, 1.55f) “4, 80 mg (0.38 mmol); 5, 183 mg (0.38 mmol); (PPh3),PdC1,, 13.3 mg (0.018 mmol); Cul, 1.5 mg (0.0078 mmol); solvent, Et3N (2.3 cm3)-pyridine (9.2 cm3).bDetermined by GPC. “The value in parentheses refers to the degree of polymerization. dDetermined by EPR. ‘Elemental analyses: C, 66.2; H, 5.6; N, 8.0; I, 9.2% (run 2); C, 66.0; H, 5.7; N, 8.2; I, 9.3% (run 3). fThe number of spins per repeating unit calculated as the structure of 9. gThe spin concentration derived from the magnetic susceptibility measurements is 2.07 x 10” spins g-’ (1.52 spins per repeating unit).20°C using benzene or dichloromethane as the solvent. Their spectra are shown in Fig. 3. The spectrum of 4 consists of three sets of a 1:3 :3 :1 quartet [a,= 1.210, and ap-H=0.198 mT, g =2.0061 (in benzene)], and that of 5 consists of a 1:2 :3 :2 : 1 quintet [a, =0.747 mT, g =2.0067 (in benzene)]. In contrast, the polyradical3 in dichloromethane gives a hyperfine-smeared single line with a peak-to-peak width of 0.65 mT (g=2.0063) (in dichloromethane), suggesting the occurrence of spin- -‘6 B E (1EU Ahm Fig. 2 UV-VIS spectra of (a) 3 (run 2 in Table l), (b)4 and (c) 5 in dichloromethane. The inset shows an expansion of the visible region. Absorptions E is corrected for the spin concentration determined by EPR spectroscopy.J. Mater. Chem., 1996,6( ll), 1745-1750 1747 1.0mT-v Fig. 3 Solution EPR spectra of 3, 4 and 5 at 20 "C: (a) 4 in benzene; (b)5 in benzene; (c) 3 (run 2 in Table 1) in dichloromethane exchange narrowing due to the high spin concentration of the pol yradical. Magnetic susceptibility measurements The magnetic susceptibility (x)measurements of polyradical 3 having the spin concentration of 1.99 x 1021 spins 8-l (run 2 of Table 1)was carried out in the temperature range 1.8-300 K with a superconducting quantum interference device (SQUID) magnetometer. The molar susceptibility is calculated as the unit molar susceptibility. The diamagnetic contribution from the sample was estimated from Pascal's constants. When the magnetic interactions among the unpaired electron spins are smaller than the energy of the thermal fluctuation kT, and the orbit angular momentum is neglected, the magnetic susceptibility x is expressed by eqn.(3), where N is Avogadro's number, pBis the Bohr magneton, g is the Lande factor, and 8 is the Weiss temperature; NpB2g2S(S+ is called the Curie constant, C. (3) 1 T-8 -x c (4) According to eqn.(4), l/x was plotted against T in the temperature range 1.8-300 K, and an almost linear curve with a Weiss temperature (0) of -1.5 K was obtained. Fig. 4 shows the ~Tvs.T plots of the polyradical 3. The curve is flat in the high temperature region (300-30 K), showing that the polyrad- ical is paramagnetic in this temperature region.From the XT value of 0.570 emu K (repeating unit)-' in this temperature region the spin concentration of the polyradical can be deter- mined to be 1.52 spins per repeating unit [theoretical 0.752emuK (repeating unit)-'], provided that S is 1/2, and this value corresponds to 2.07 x 1O2l spins 8-l. If the structure of polyradical is 9, the number of spins per repeating unit 0 50 100 150 200 250 300 TIK Fig. 4 Plot of ~Tvs.T for 3 becomes 1.69. Accordingly, the spin concentration determined by the magnetic susceptibility measurements is in good agree- ment with that determined by EPR though the SQUID magnetic measurements give a somewhat higher value. In the temperature region below 30 K, on the other hand, the ;~Tus.T plot shows a downward turn, indicating that unpaired electron spins interact antiferromagnetically. This antiferro- magnetic interaction is most probably due to intramolecular and intermolecular through-space interactions between the unpaired electron spins. In most cases through-space inter- actions are antiferromagnetic. If the expected through-bond ferromagnetic interactions are sufficiently strong, an upward turn of the ~Tvs.T plots showing ferromagnetic interactions will be observed. However, if the ferromagnetic interaction between the spins is very or negligibly weak, they will be masked by through-space antiferromagnetic interactions. This is assumed to be this case in the present example.Recently, Nishide et al. observed ferromagnetic interactions for poly(- phenyleneviny1ene)-based polyphenoxyl and polyaminoxyl.l* In that case the ferromagnetic through-bond interactions seem to be strong. Consequently, only when there are strong through-bond ferromagnetic interactions between the electron spins will polyradicals show ferromagnetic behaviour. Magnetic measurements of polyradicals diluted with diamag- netic materials are in progress. Experimental All mps were measured on a Yanaco micro-melting point apparatus and are uncorrected. IR spectra were run on a JASCO A-202 spectrophotometer. UV-VIS spectra were meas- ured with a Shimadzu UV-2200 spectrophotometer. 'H NMR. spectra were recorded with a JEOL a-400 spectrometer (400 MHz) with Me,% as internal reference; J values are given in Hz.GPC was run on a Tosoh GPC 8000 series using Shodex G5000HHR, GMHHR-L, and GMHHR-L columns calibrated with polystyrene standards, eluting with THF and with moni- toring of the refractive index. EPR spectra were recorded on a JEOL ME-3X spectrometer operated at the X-band. Hyperfine splitting constants (a) and g values were determined by simultaneous measurements with a dilute Fremy's salt in aqueous K2C03 solution (aN= 1.309 mT, g =2.0055). The spin concentrations of monomers and polyradicals were determined by the double integrated EPR spectra of the samples in CH2C12 recorded on a Bruker ESP 300 spectrometer. Calibration curves were drawn with 1,3,5-triphenylverdazyl solutions using the same EPR cell and solvent and the same instrument settings as for the sample measurements.The magnetic susceptibility measurements were carried out on a Quantum Design SQUID MPMS2 system in the tempera- ture range 1.8-300 K. The diamagnetic contribution of the samples was estimated from Pascal's diamagnetic constants. Materials N-tert-Butyl-N-(3,5-diethynylphenyl)aminoxylwas prepared by our previously reported method and purified by subli- 1748 J. Mater. Chem., 1996, 6(ll), 1745-1750 mation.8 This monomer was again purified by sublimation prior to use. Pyridine and triethylamine used in the polycond- ensation was purified by distillation. CuI was of commercial grade. 2,6-Diiodo-4-toluidine This compound was obtained by treating p-toluidine (13.5 g, 0.126 mol) with benzyltrimethylammonium dichloroiodate (BTMA IC12) (117.5 g, 0.338 mol) in the presence of CaCO, (40 g) in CH2C12 (1170 cm3)-MeOH (473 cm3) at reflux tem- perature for 1 day.16 After the usual workup, the product was refluxed in 100cm3 of methanol for 10min to remove polar byproducts, giving almost pure 2,4-diiodo-4-toluidine in 78 YO yield (35.5 g, 0.099 mol).Recrystallization from hexane-ben- zene gave light brown needles, mp 122-123°C (1it.,I6 123.5-125 "C). 3,5-Diiodotoluene To a stirred solution of 2,6-diiodotoluidine (12.58 g, 0.035 mol) in EtOH (600 cm3)-benzene (230 cm3) was added a solution of NaNO, (4.44 g) in water (31 cm3) and conc. sulfuric acid (8.3 cm3).After stirring for 30 min at room temperature, the mixture was refluxed for 1 day and the solvent removed in vacuo. After the residue had been neutralized with NaOH (2 mol drn-,), the organic products were extracted with CH2C12, dried (MgSO,), evaporated and chromatographed on silica gel (Wako gel C200) with benzene-hexane (1:9) to give pure 3,5-diiodotoluene in 50% yield (6.06 g, 17.6 mmol). Recrystallization from MeOH gave colourless needles, mp 40-41 "C (1it.,lg 44.5-45.5 "C). 3,5-Diiodobenzyl bromide A solution of 3,5-diiodotoluene (6.0 g, 17.4 mmol), N-bromo- succinimide (NBS) (6.0 g, 33.7 mmol) and benzoyl peroxide (20 mg) in 1,2-dichloroethane (60 cm3) was gently refluxed for 6 h. The reaction mixture was then washed with 10Y0 NaHS03 and brine, dried (MgSO,), and solvent was evaporated.The resultant crude product (7.36 g, 100%) was used in the follow- ing step without purification. Recrystallization from hexane- benzene gave colourless needles, mp 79-80 "c; dH(CDC1,) 4.31 (2 H, s, CH,Br), 7.69 (2 H, d, J 1.5, ArH) and 7.98 (1H, t, J 1.5, ArH). 3,5-Diiodobenzaldehyde To a solution of hexamethylenetetraamine (3.0 g, 21.4 mmol) in CHCl, (150cm3) was added 3,5-diiodobenzyl bromide (7.36 g, 17.4 mmol), and the resulting mixture was gently refluxed for 15 h. After the solvent had been removed by evaporation, acetic acid (36 cm3), water (36 cm3) and conc. HCl (5.7 cm3) were added, and the mixture was refluxed for 30 min. The resultant reaction mixture was then neutralized with 10% Na2C03, extracted with CH2C12, dried (MgSO,), the solvent evaporated and the residue chromatographed on silica gel with benzene-hexane (1 :1) as the eluent to give 3,5-diiodobenzaldehyde in 47% yield (2.94 g, 8.21 mmol).Recrystallization from hexane gave colourless needles, mp 131-132°C; dH (CDCl,) 8.15 (2 H, d, J 1.5, ArH), 8.30 (1H, t, J 1.5, ArH) and 9.83 (1 H, s, CHO). 1,3-Dihydroxy-2-(3,5-diiodophenyl)-4,4,5,5-tetramethy1-2-imidazoline A solution of 3,5-diiodobenzaldehyde (2.94 g, 8.21 mmol), 2,3- bis(hydroxyamino)-2,3-dimethylbutane (1.80 g, 12.1 mmol)20 and 2,3-bis(hydroxyamino)-2,3-dimethylbutane monosulfate salt (20 mg)20 in MeOH (40 cm3)-THF (20 cm3) was stirred at room temperature for 3 days.The solvent was then evapor- ated, and addition of methanol (30 cm3) and water (15 cm3) gave crystals of the desired imidazoline in 93% yield (3.74 g, 7.66mmol) as a crude product. The imidazoline was used in the following reaction without purification. Recrystallization of the crude product from benzene gave colourless plates in 55% yield (2.21 g, 4.52 mmol), mp 226-227 "C; d,(CDC13) 1.09 (6 H, s, Me), 1.15 (6 H, s, Me), 4.54 (1 H, s, NCHN), 7.90 (2 H, d, J 1.4, ArH) and 7.97 (1H, t, J 1.4, ArH) (Found: C, 32.0; H, 3.8; N, 5.7. Cl,Hl8I2N,O, requires C, 31.99; H, 3.72; N, 5.74%). 2-(3,5-Diiodophenyl)-4,4,5,5-tetramethyl-3-oxido-2-imidazolin-3-ium-1-yloxyl5 The imidazoline (0.60 g, 1.23 mmol) was dissolved in benzene (50 cm3)-EtOH ( 10 cm3) with stirring.To this stirred solution, K2C03 (6.0 g) and PbO, (6.0 g) were added. After stirring for 30min, the resulting dark blue reaction mixture was filtered and the filtrate was evaporated under reduced pressure to give dark blue powder or needles containing a small amount of white by-products. This dark blue solid was then refluxed for 5 min in hexane (20 cm3) and the hexane was discarded by decantation. Recrystallization of the residue from hexane- benzene gave 5 as dark blue needles in 45% yield (0.268 g, 0.552rnmol). The use of the pure imidazoline also gave 5 in 62% yield, mp 205-207 "C; aN (benzene)/mT 0.747 (g 2.0067); &,,,(CH,Cl,)/nm 633 (&/dm3 mol-I cm-I 300sh), 592 (340), 371 (14600) and 357 (7750) (Found: C, 32.4; H, 3.3; N, 5.7.Cl3Hl5I2N2O2 requires C, 32.19; H, 3.12; N, 5.78%). Polycondensation of 4 with 5 In a two-necked flask was placed 4 (80 mg, 0.38 mmol), 5 (183 mg, 0.38 mmol), (PPh,),PdCl, (13.3 mg, 0.018 mmol), CuI (1.5 mg, 0.0078 mmol), Et,N (2.3 cm3) and pyridine (9.2 cm3). After the flask had been charged with nitrogen, the mixture was stirred at room temperature (ca. 20°C) for 2-6 h under nitrogen. The reaction mixture was then poured into a large amount of methanol, and the light blue powder deposited was collected by filtration and dried in vacuo. The results of the polycondensation are summarized in Table 1. This work was supported (in part) by a Grant-in-Aid for Scientific Research on Priority Area 'Molecular Magnetism' (Area No.228/06218226) from the Ministry of Education, Science and Culture, Japan. References 1 J. S. Miller and A. J. Epstein, Proceedings of the Symposium on the 4th International Conference on Molecule-Based Magnets, Mol. Cryst. Liq. Cryst., 1995, 271, 1; 272, 1; 273, 1; 274, 1; H. Iwamura and N. Koga, Acc. Chem. Rex, 1993, 26, 346; D. A. Dougherty, Acc. Chem. Res., 1991,24,88;Y. Miura, Kobunshi, 1994,43,838. 2 Y. Miura, K. Inui, F. Yamaguchi, M. Inoue, Y. Teki, T. Takui and K. Itoh, J. Polym. Sci., Polym. Chem. Ed., 1992,30,959. 3 Y. Miura, M. Matsumoto and Y. Ushitani, Macromolecules, 1993, 26, 2628. 4 Y. Miura, M. Matsumoto, Y. Ushitani, Y. Teki, T. Takui and K. Itoh, Macromolecules, 1993,26, 6673. 5 H. Nishide, T. Kaneko, M. Igarashi, E.Tsuchida, N. Yoshioka and P. M. Lahti, Macromolecules, 1994, 27, 3082 and reference cited therein; A. Fujii, T. Ishida, N. Koga and H. Iwamura, Macromolecules, 1991,24, 1077. 6 Y. Miura and K. Inui, Makromol. Chem., 1992,193,2137. 7 Y. Miura, Y. Ushitani, K. Inui, Y. Teki, T. Takui and K. Itoh, Macromolecules, 1993,26,3698. 8 Y. Miura and Y. Ushitani, Macromolecules, 1993,26,7079. 9 P. Swoboda, R. Saf, K. Hummel, F. Hofer and R. Czaputa, Macromolecules, 1995,28,4255 and references cited therein. 10 A. A. Ovchinnikov, Tlieor. Chim. Acta, 1978,47,297. 11 For example, see: A. R. Forrester, J. M. Hay and R. H. Thomson, Organic Chemistry of Stable Free Radicals, Academic Press, London and New York, 1968, pp. 180-246. 12 H. Lemaire, Y. Marechal, R. Ramasseul and A. Rassat, Bull. SOC. Chim. Fr., 1965, 372. J. Mater. Chem., 1996, 6( ll), 1745-1750 1749 13 14 T. Takui, Y. Miura, K. Inui, M. Inoue, Y. Teki and K. Itoh, Mol. Cryst.Liq.Cryst., 1995,271, 55. M. S. Davis, K. Morokuma and R. W. Kreilick, J. Am. Chem. SOC., 18 H. Nishide, T. Kaneko, T. Nii, K. Katoh, E. Tsuchida and K. Yamaguchi, J. Am. Chem. SOC., 1995, 117, 548; T. Kaneko, S. Toriu, Y. Kuzumaki, H. Nishide and E. Tsuchida, Chem. Lett., 1972,94,5588. 1994, 2135; H. Nishide, T. Kaneko, S. Toriu, Y. Kuzumaki and 15 16 A. Zheludev, V. Barone, M. Bonnet, B. Delley, A. Grand, E. Ressouche, P. Rey, R. Subra and J. Schweizer, J. Am. Chem. SOC.,1994,116,2019. S. Kajigaeshi, T. Kakinami, H. Yamasaki, S. Fujisaki and 19 20 E. Tsuchida, Bull. Chem. SOC.Jpn., 1996,69,499. H. L. Wheeler and L. M. Liddle, Am. Chem. J., 1909,42,441. M. Lamchen and T. W. Mittag, J. Chem. SOC.C, 1966,2300. 17 T. Okamoto, Bull. Chem. SOC.Jpn., 1988,61,600. E. F. Ullman, J. H. Osiecki, D. G. B. Boocock and R. Darcy, J. Am. Paper 6/03880H; Received 4th June, 1996 Chem. SOC.,1972,94,7049. 1750 J. Muter. Chem., 1996, 6(ll),1745-1750

ISSN:0959-9428    

DOI:10.1039/JM9960601745

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Theoretical determination of the molecular and solid-state electronic structures of phthalocyanine and largely extended phthalocyanine macrocycles Enrique Orti,* Raul Crespo, M. Carmen Piqueras and Francisco Tomas Department de Quimica Fisica, Universitat de Valgncia, Doctor Moliner 50, E-46100 Burjassot (Valt?ncia), Spain The molecular and solid-state electronic structures of metal-free phthalocyanine and a series of linearly benzoannulated phthalocyanines have been investigated using the valence effective Hamiltonian (VEH) quantum-chemical method. Geometry optimizations show that, from the molecular structure standpoint, phthalocyanine-based macrocycles are the result of joining four polyacenic units to the CsNs central ring. The electronic structure calculated for the parent phthalocyanine is compared with that of porphyrin, and the consequences of benzoannulation and meso-tetraaza substitution on the optical properties of phthalocyanines are discussed.The VEH results obtained for extended phthalocyanines are in agreement with available photoemission, cyclic voltammetry and optical absorption data and help to rationalize the evolution of the electronic properties. The first ionization energy is predicted to decrease with linear benzoannulation and asymptotically converges to an extrapolated value of -5.7 eV. Strikingly, a non-convergent behaviour is obtained for the HOMO-LUMO energy gap and very low excitation energies are predicted for extended phthalocyanines. Band-structure calculations have been performed for one-dimensional stacks of the molecules investigated.The variation of the bandwidth with the staggering angle and the intermolecular separation provides a coherent picture of the electrical conductivities observed experimentally in crystals and polymers. Very small bandgaps lower than 0.5 eV are predicted for extended phthalocyanines. Since their accidental synthesis in 1907,l phthalocyanines (Pcs) have become one of the most intensely studied macrocycles owing to the unique properties they exhibit.2 Pcs are among the most stable organic materials and show remarkable optical properties. The conjugated macrocycle, which contains 42 rc electrons, leads to very intense absorption bands in the far red end of the visible spectrum (ca.670nm, Q band) and in the near-ultraviolet (ca. 340nm, B or Soret band). The intense colour and high thermal and chemical stability have for decades determined the use of phthalocyanines as dyes and pigments. Phthalocyanines are also of great interest owing to their electrical proper tie^.^ They were recognized as good photoconductors as early as 194S4 and are at present used in electrophotographic systems,’ photovoltaic cells for energy conversion,6 etc. More recently, phthalocyanines have been reported to form highly conducting materials. Cocrystallization of Pcs with oxidising agents like iodine yields ‘molecular metals’ with conductivities of the order of 10-lOOOScm-l at room tem- perat~re.~The packing of the phthalocyanine molecules in columnar stacks leads to an effective overlap between the rc molecular orbitals on adjacent macrocycles, which provides the electronic pathway for the delocalization of the charge carriers generated upon oxidation.The presence of metal atoms is therefore not a requirement to achieve high conduc- tivities in phthalocyanines and a value of 700 S cm-l has been reported for partially oxidized crystals of metal-free phthalocy- anine (H,Pc) at room temperature.’“ Higher room-temperature conductivities of around lo4 S cm-l have been measured recently for high-purity oxidized phthalocyanines H,PcI and NiPcLsb The presence of the central metal atom is, however, a requirement for the formation of cofacially joined metalloph- thalocyanine polymers [MPcL], where the metallomacro- cycles are linked together by bisaxially metal-bonded bridging ligands, L.9 Depending on L, these polymers exhibit compara- tively high semiconducting properties (oRT=0.05-0.3 S cm-‘)lo without external oxidative doping.These electrical properties make phthalocyanines prime candidates for use in electronics and molecular electronics. Indeed, they have already been employed in highly sensitive gas sensors,” rectifiers (molecular diodes)’, and tran~istors.’~ The potential applicability of phthalocyanines is being investigated in many other fields such us optical data storage,14 non-linear optics,” electrochromic devices,16 liquid crystals” and Langmuir-Blodgett (LB) films.” One of most recent applications is their potential use as photosensitizers in photo- dynamic therapy.” Phthalocyanines are more effective than porphyrins as photosensitizers because of the high absorption coefficient of the Q band, often exceeding lo5 1 mol-’ ern-'.The phthalocyanine ring offers sixteen potential sites of substitution through which the modulation of the electronic properties (optical absorptions, ionization and redox poten- tials, etc.) can be effected, enabling the possibility of obtaining phthalocyanine-based materials with optical or electrical properties adjusted to the desired application. In this context, extended phthalocyanines are of great relevance and, in particu- lar, 2,3-naphthalocyanines (2,3-Ncs) are the object of numerous investigations.Compared to phthalocyanines, the Q absorption band in 2,3-naphthalocyanines is shifted to the near-IR and appears at about 800 nm.20 This absorption makes 2,3-Ncs excellent can- didates for high-density optical recording (ODR) media since a long-wavelength absorption is required in the writing and reading processes using semiconductor lasers.21 2,3-Ncs are also very promising as sensitizers for photomedicine since bodily tissues allow deeper light penetration with increasing wavelength in the range 600-1200 nm.” They have also been used in the synthesis of cofacially linked polymers, showing higher conductivities than the analogous phthalocyanine-based materials.23 Non-linear optical properties have been measured for a variety of 2,3-N~s.,~ In a previous ~aper,~’ we studied the electronic structure of 2,3-naphthalocyanine and 1,2-naphthalocyanine as models of linearly and angularly annulated phthalocyanines.The results showed that the linear extension of the conjugated structure affects the electronic properties of the macrocycle more drasti- cally, suggesting that materials with improved electrical and optical properties might be obtained from linearly annulated phthal~cyanines.~~~~~Our goal in this work is to investigate the electronic properties of largely extended linearly annulated phthalocyanines at both the molecular and the solid-state levels. The geometric and electronic structures of metal-free J. Muter. Chern., 1996,6(11), 1751-1761 1751 2,3-naphthalocyanine (2,3-H2Nc), 2,3-anthracyanine (2,3-H2Ac) and 2,3-tetracenocyanine (2,3-H2Tc) are thus calculated using quantum-chemical methods.The molecular structures of 2,3-H,Nc, 2,3-H2Ac and 2,3-H2Tc are depicted in Fig. 1 together with that of H,Pc and that of the angularly annulated 9,lO-phenanthrenocyanine(H,Phc). The electronic structure of H,Pc is first analysed and compared with those calculated for the closely related metal-free porphyrin ( H2P) and tetra- benzoporphyrin (H,Tbp). Our aim is to provide a deep under- standing of the unique electronic properties of the phthalocyanine macrocycle that serves as a reference in the study of extended phthalocyanines. Electronic band-structure calculations are performed for one-dimensional molecular stacks of H,Pc, 2,3-H2Nc, 2,3-H2Ac, 2,3-H2Tc and H2Phc.Our aim is to discuss the electrical properties of phthalocyanine-based crystals and polymers. The first anthracyanines were obtained in 1971 as octapheny- lated derivative^.'^" The synthesis of different octa-tert-b~tylated,~~and unsubstituted28 2,3-Acs was also reported. Despite these early syntheses, little experimental work has been carried out on these enlarged macrocycles. More recently, metal-free tetra(tert-butyl)-2,3-anthracyanineand the corre-sponding cobalt complex have been obtained by Kobayashi et Two different Fe" complexes have been obtained by ~1.~~~1~~ Hanack et aL3' To our knowledge, only the synthesis of a vanadium complex of 2,3-tetracenocyanine has been reported in the literature.28 The photophysics and photochemistry of 2,3-Acs is at present being investigated in an attempt to obtain a new generation of photosensitizers for photodynamic the rap^.^ Computational details The molecular geometries of the metal-free macrocycles shown in Fig.1 were optimized using the MNDO-PM3 (modified neglect of diatomic overlap, parametric method number 3) 't+ X ' \ H2Phc 2,3-H2Ac g & \/ Fig. 1 Molecular structures of metal-free phthalocyanine ( H2Pc), 2,3-naphthalocyanine (2,3-H2Nc), 2,3-anthracyanine (2,3-H2Ac), 9,lO-phenanthrenocyanine (H2Phc) and 2,3-tetracenocyanine (2,3-H2Tc) 1752 J. Muter. Chem., 1996, 6(11), 1751-1761 semi-empirical method,32 as implemented in the MOPAC-6.0 system of programs,33 and assuming D,, symmetry constraints. The PM3 method has recently been applied to the study of largely extended porphyrins and olig~porphyrins.~~ Both the molecular electronic structure and the electronic 108.9 band structure of one-dimensional stacks of the macrocycles were investigated using the valence effective Hamiltonian 118.0 (VEH) pseudopotential method.The VEH approach was 121.3 originally developed to deal with and was later extended to treat stereoregular polymers,36 and has been used widely in the context of conjugated molecules and poly- mer~.~~,~'It constitutes an especially useful tool for dealing with large molecular or crystalline systems, since it is param- \ Nnl I .N 1.344 m 1 .;91 eterized to yield one-electron energies of ab initio double-l: 1.471 (DZ) quality without performing any self-consistent-field (SCF) process or calculating any two-electron integral.Its suitability for describing the electronic structure of phthalocy- 1.392 anine-type macrocycles is supported by previous ~ork.~~,~~-~~ In particular, an excellent correlation is found between the molecular and crystalline electronic structures calculated for H,Pc and experimental photoemission (UPS or XPS) All the calculations presented here were performed using the atomic potentials previously optimized for carbon, nitrogen and hydrogen.46 Compared to standard ab initio Hartree-Fock (HF) calcu- lations, the VEH method provides good estimates for the energies of the lowest-energy electronic transitions.This feature is due to the fact that the VEH parameterization is obtained only using the energies and functions of the occupied MOs and is therefore not contaminated by any information from the too diffuse virtual MOs that HF calculations pro-vide.38b,47,48Unoccupied MOs are thus expected to be as reliable as occupied MOs when calculated with the VEH method. In the case of phthalocyanine, the VEH method systematically underestimates the energy of the first optical transition associated with the HOMO-LUMO energy gap (E,) by 0.62 eV. This shortcoming comes from the parametriz- ation used for the nitrogen atom based on pyrrole and dimethylamine molecules.46 As was done previo~sly,~~*~~~~~ the 120.5 119.2 120.4 118.2 132.6 1.471 VEH E, values for all the phthalocyanines studied in this work are consistently increased by 0.62 eV.Results and Discussion Optimized molecular structures To our knowledge, experimental data on the molecular struc- ture of metal-free phthalocyanine-based macrocycles are only available for H,Pc. As early as 1935,49 Robertson performed an X-ray diffraction study of the lattice parameters of the p-polymorphic form of H,Pc and a year later reported a detailed structure of the carbon-nitrogen skeleton of the molecule.50 More recently, Hoskins et aL5' carried out a neutron diffraction study on the same p-polymorphic form and Yase et aL5, reported the X-ray structure of the a-polymorphic form. All these works agree in attributing a highly delocalized planar geometry to the carbon-nitrogen backbone.Fig. 2(a) summar- izes the PM3-optimized bond lengths and bond angles calcu- lated for H,Pc. The D,, constraints provide a highly delocalized structure with almost equivalent lengths for the carbon-nitro- gen bonds forming the C8N8 central ring, in agreement with experiment. The higher stability of this structure us. frozen resonant structures of C,, symmetry with alternating single and double bonds has been shown recently for metal-free p~rphyrin.~~ The theoretical parameters calculated for H,Pc cannot be compared easily with available experimental because these data do not differentiate between the two kinds of isoindole moieties present in H,Pc owing to the difficulty in locating the inner hydrogen atoms.The PM3 oresults predict almost identical lengths of 1.344 and 1.349 A for adjacent Fig. 2 DZhPM3-optimbzed parameters for (a) H,Pc and (b) 2,3-H2Nc. Bond lengths are inA and bond angles in degrees. N, denotes a pyrrole nitrogen and N, and N, denote pyrrole aza and rneso-bridging aza nitrogens, respectively. C, -N, bonds. These lengths are longer than the X-rax value of 1.33 A reported for ~-H,Pc.~~ The C,-N, (1.411 A) and C,-N, (1.386 A) bond distances are also calculated to be longer thFn the experimental values, which are about 1.36-1.38 A.51,52 The angles defined by the bonds forming the C8N8 central ring are in general wider than the X-ray angles by 1-3".The largest deviation is found for the C,-N,-C, angles, for which the PM3 method obtains a value of 125.9' and the X-ray data provide a value of 122.5' for ~-H,Pc.~, The PM3 value is in accord with that reported for the closely related tetraazaporphyrin (H,Tap) from ab initio DZ calcu-lations ( 125.0').54 The N;-.N, and N;--N, distances definiqg the size of the central ring have values of 3.98 and 4.17 A, respectively. The C-C bonds forming the peripheral benzene rings in H,Pc have typical aromatic bond lengths, indicating that the aromaticity of these rings is pre:erved in H,Pc. As a consequence, the C, -C, bonds (ca. 1.42 A) fusing the benzene rings to the pyrrole moieties are $gnificantly longer than those calculated for H,Tap (ca.1.36 A). The C,-C, bonds have lengths of 1.46-1.47 A. Fig. 2( b) displays the geometric parameters optimized for 2,3-H2Nc. The central CsNs ring is almost identical to that of J. Muter. Chern., 1996,6(11), 1751-1761 1753 H,Pc, the major differences being 0.002 for the bond lengths and 0.4" for the bond angles. As for H,Pc, the peripheral naphthalene units fused to the central tetraazaporphyrin ring preserve their structural aromaticity. Fig. 3 compares the geometry calculated for a naphthopyrrole unit in 2,3-H2Nc with the PM3-optimized geometries of isoindole, pyrrole and naphthalene molecules. There are marked differences between the geometry of the pyrrole units within naphthalocyanine and those in isoindole and pyrrole molecules.For pyrrole, the C,:C, bonds show a high degree of double-bond character (1.390 A), which is maintained in the isoindole molecule (1.405 A). This molecule must in fact be viewed as resulting from the union of a cis-buta-1,3-diene fragment on the top of the pyrrole ring. The situation is different in 2,3-H2Nc since the structure of pyrrole is not preserved. oThe C,-C, bonds mostly have single-bond character (1.463 A) and the naphtha- lene moieties have a geometry almost identical to that of naphthalene [cf. Fig. 3(a) and (d)]. Thus, 2,3-H,Nc can be visualized as the result of joining four naphthalene units to the central C8Ns macrocycle. The PM3-optimized geometries obtained for the more extended 2,3-H2Ac and 2,3-H2Tc are not displayed because they show the same geometric trends discussed above for H,Pc and 2,3-H2Nc; i.e., the C8N8 central ring is nearly identical to that of H,Pc and the polyacenic units preserve their structural identity showing geometries very close to those of anthracene (2,3-H2Ac) and tetracene (2,3-H2Tc).The size of the C8N8 central ring is predicted to increase slightly but continuously with linear extension. The Np...Np and N;..N, distances lengthen by 0.030 and 0.016 A, respectively, in passing from H~Pcto 2,3-H,Tc. Molecular electronic structures Table 1 summarizes the VEH molecular orbital (MO) distri- butions obtained for H,Pc, 2,3-H2Nc, 2,3-H,Ac and 2,3-H,Tc. It collects all the occupied orbitals with energies above -10.0eV and the lowest five unoccupied MOs.All these orbitals are of n-nature, the first occupied o-orbitals lying below -10.2 eV. The main characteristics of the electronic structure of phthalocyanine are first discussed by comparing them to those calculated for porphyrin (H2P) and tetrabenzo- porphyrin (H,Tbp). The effects of the n-system extension of the macrocycle on the electronic structure are then analysed. Metal-free phthalocyanine. The highest occupied molecular orbital (HOMO) of H,Pc corresponds to the 4a, level. This orbital is calculated to have an energy of -6.42 eV, in perfect agreement with the first ionization energy (6.41 eV) obtained from gas-phase UPS data.57 It is separated by an energy gap of ca. 2.5 eV from the next occupied 7b,, orbital heading a group of very close-lying MOs.The appearance of this gap is also in accord with experimental photoemission data since energy differences of 2.3-2.7 eV are found between the first 1.416 1.415 1.369 1.425 (1.417) (Jl.yJ2, (1.369) 1.463 1.405 (1.421) N 1.397 I 1 (1.370)I H (c) (d) Fig. 3 Comparison of PM3-optimized bond lengths (in A)of (a) naph- thopyrrole units in 2,3-H2Nc; (b) isoindole molecule; (c) pyrrole molecule (experimental microwave data from ref. 55 are given in parentheses); (d) naphthalene molecule (X-ray crystallographic data from ref. 56 are given in parentheses) 1754 J. Mater. Chern., 1996, 6(11), 1751-1761 Table 1 VEH one-electron energy levels (sign-reversed energies in eV) obtained for phthalocyanine (H,Pc), 2,3-naphthalocyanine (2,3-H2Nc), 2,3-anthracyanine (2,3-H2Ac) and 2,3-tetracenocyanine (2,3-H2Tc)" H,Pc 2,3-H,Nc 2,3-H,Ac 2,3-H,Tc 3.09 9b,, 4.02 llb3, 4.47 13b3, 4.63 3.34 llbl, 4.05 13bl, 4.66 15bl, 5.04 4.01 lob,, 4.27 12bl, 4.80 14b,, 5.14 5.16 8b2, 5.00 lob,, 5.03 12b3, 5.20 5.23 8b3, 5.02 lob,, 5.04 12b,, 5.22 6.42 6a, 6.05 8a, 5.88 10a, 5.79 9.01 5a, 8.56 7a, 7.90 9a, 7.48 9.39 7b3, 8.65 9b,, 7.98 llb3, 7.54 9.42 7b,, 8.66 9b,, 7.99 llb,, 7.55 9.43 4a, 8.96 6a, 8.23 8a, 7.73 9.43 9b1, 9.01 llb,, 9.00 7a, 8.96 9.64 9.71 9.72 10.04 8bl, 7b1, 6b2, 6b3, 9.21 9.26 9.26 9.25 lob,, 8b2, 9b1, 8b3, 9.13 9.16 9.16 9.16 lob3, lob,, 13bl, 6a, 9.00 9.00 9.01 9.08 5% 9.63 12bl, 9.10 7b3, 7b2, 4% 9.67 9.68 9.88 9b,, 9b,, llblu 9.11 9.12 9.21 "The occupied orbitals above -10.0eV and the lowest five unoccupied orbitals are included.Orbitals are labelled according to the D,, point group starting from the first occupied valence orbital. All orbitals are of n-nature. Fig. 4 Atomic orbital composition of the highest two occupied molecular orbitals of phthalocyanine (H,Pc) and porphyrin (H,P). The HOMOs of isoindole and pyrrole molecules are also shown for the sake of comparison. The sizes of the circles are proportional to the magnitude of the LCAO coefficients. Contributions smaller than 0.10 are not displayed. The VEH energies (inev) of the orbitals are given in parentheses.two ionization energie~.'~.~~-f- The high energy of the HOMO and its energy separation from the remaining occupied orbitals deserve special attention because they are the main reasons for the unique electronic properties that differentiate phthalocy- anines from other tetrapyrrolic systems like porphyrins. As illustrated in Fig.4, the HOMO of H,Pc spreads over the carbon backbone with nodes on N, and N, atoms and negligible contributions from the meso nitrogens N,. The HOMO of porphyrin shows the same atomic orbital (AO) composition and is calculated at -7.40 eV using a PM3-optimized geometry. It is therefore found to be 1.0eV lower in energy than the HOMO of H2Pc, in agreement with the higher first ionization energy reported for H,P (6.9-7.2 eV).59 The HOMOs of H2P (2a,) and H,Pc (4a,) actually originate 7 A detailed comparison of the valence electronic structure calculated for H,Pc with experimental photoemission data is given in ref.39 and 41. The slight differences between the one-electron energies reported in those works and those presented here are due to the previous use of an old version of the VEH parameterization for carbon and hydrogen atoms (see ref. 36b) and of an averaged crystallographic geometry for H,Pc. in the HOMOs of pyrrole (la,) and isoindole (2a,), respect- ively. These orbitals are calculated at -8.24 and -7.06 eV, in very good agreement with the first ionization energies measured in the gas phase for pyrrole (8.22 eV)60 and N-methylisoindole (7.12 eV).61 The destabilization of the HOMO in passing from pyrrole to H,P or from isoindole to H,Pc is due to the changes undergone by the bond lengths of the pyrrole and isoindole moieties, which tend to weaken the bonding interactions and to reinforce the antibonding interactions in H,P and H2Pc.For instance, the C,-C, bonds lengthen from 1.390 to 1.449 A in going from pyrrole to H2P, thus reducing the bonding interattions, while the C,-C, bonds shorten from 1.421 to 1.371 A, thus enhancing the antibonding interactions. Similar changes are found for these bonds in passing from isoindole to H,Pc [cf Fig. 2(a) and 3(b)]. The destabilization of the HOMO of H,Pc relative to the HOMO of H,P is similar to that obtained for the HOMO of isoindole with respect to the HOMO of pyrrole, and results from the antibonding inter- action of the cis-butadiene fragments with the HOMO of the pyrrole moieties.The HOMO -1 shows the same atomic orbital (AO) com- position for H,Pc and H,P and is mostly localized on nitrogens and on the meso bridging atoms. For H2P, it appears at -7.68 eV, very close in energy to the HOMO, while for H,Pc, it is calculated at -9.01 eV, i.e. 2.6 eV below the HOMO. The stabilization of the HOMO-1 of H,Pc results from the substitution of the methine linking units (CH,) of H,P by aza bridges (N,) in H,Pc, and is due to the higher nuclear charge of nitrogen atoms. Two main electronic effects are therefore derived from the structural differences between H2P and H,Pc.The destabiliz- ation of the HOMO of H,P in passing to H,Pc due to benzoannulation, and the removal of the HOMO-HOMO -1 near-degeneracy present in H2P due to meso-tetraaza substi- tution. The relationship existing between the molecular and the electronic structures is confirmed by examination of the electronic structure of tetrabenzoporphyrin. H,Tbp shows a molecular structure intermediate between those of H,Pc and H2P, since it has isoindole moieties like H,Pc but is linked by methine units as in H2P. The 4a, HOMO of H,Tbp is thus calculated at an energy of -6.50 eV, almost identical to the energy of the HOMO of H,Pc (-6.42 eV), while the 7bl, HOMO -1 is located at -7.74 eV as in H2P (-7.68 eV).The differences in the energies calculated for the HOMO and HOMO-1 of H,P, H,Tbp and H,Pc explain the different optical properties observed for these compounds as discussed below. These differences have recently been analysed on the basis of ab initio calculation^,^^ and have been used to rational- ize the different behaviour of porphyrin-based conductors compared to Pc-based ones.62 The lowest two unoccupied molecular orbitals (LUMO and LUMO + 1)of H,Pc correspond to a pair of almost degenerate 6b3, (-5.23 eV)-6b2, (-5.16 eV) n*-orbitals that are respect- ively located along the x and y axes (see Fig. 5). The quasi- Fig. 5 Atomic orbital composition of the lowest two unoccupied molecular orbitals of phthalocyanine. The sizes of the circles are proportional to the magnitude of the LCAO coefficients.Contributions smaller than 0.10 are not displayed. degeneracy of these two orbitals is due to the almost identical geometry predicted for the isoindole moieties. Their relative energy ordering is in fact very sensitive to the geometry used. Previous calculations on H,Pc using an averaged crystallo- graphic geometry placed the 6b2, orbital below the 6b3, orbital.25 The LUMO and LUMO+l of H,Tbp (-4.71 and -4.67 eV) and H2P (-5.04 and -4.91 eV) show the same topology. The vapour absorption spectra reported by Edwards and Go~terman~~for H,Pc show two sharp peaks in the visible region at 1.81 eV (Qx, 686.0 nm) and 1.99 eV (Qy, 622.5 nm), that constitute the so-called Q band, and a broad band centred at 3.65 eV, which extends from 3.3 to 4.1 eV and is called the B or Soret band.The double-peak Q band results from the HOMO+LUMO and HOMO+LUMO + 1 electronic trans- itions, since no other low-energy excitation is expected because the HOMO of H,Pc is well above the remaining occupied MOs and the LUMO and LUMO + 1 are ca. 2 eV below the 7b2,-7b3, n*-orbitals. The broad B band is usually assigned to the HOMO -1+LUMO,LUMO + 1 excitations, but other electronic transitions coming from the close-lying lower energy occupied orbitals to the LUMO and LUMO+ 1 (e.g. 6b,,, 5bl,, 3a,, etc.+6b3,, 6b2,) or from the HOMO to higher energy virtual orbitals (e.g. 4a, +7b,,, 7b3,) can contribute to this band, as was suggested by Stillman and Ny~kong~~ and has recently been shown by Kobayashi et ~1.~~‘and Zerner et al.65b for MPcs.The situation in H,P is completely different since the HOMO and HOMO -1 are nearly degenerate and lie ca. 2.5 eV above the remaining occupied orbitals. The electronic excitations from these two orbitals to the LUMO and LUMO + 1 therefore account for both the Q and the B bands, to which they contribute almost equally as recent ab initio calculations have demonstrated.66 This is the basis of the ‘four orbital model’ developed by Gouterman and co-worker~~~,~*to explain the spectra of porphyrins. The model is, however, not fully transferable to phthalocyanines owing to the participation of other MOs other than the HOMO and HOMO -1 and the LUMO and LUMO + 1 in the electronic transitions giving rise to the B band.The H2Tbp molecule is an intermediate case because the HOMO -1 is well isolated from both the HOMO and the low-lying occupied orbitals and the B band results from the HOMO-l+LUMO, LUMO + 1 transitions. This explains the sharpness of the B band for H,Tbp, which is well resolved in B, and By components.68 Extended phthalocyanines. Table 1 indicates that, as found for H,Pc, the HOMOs of the extended phthalocyanines 2,3- H,Nc, 2,3-H,Ac and 2,3-H,Tc correspond to levels of a, symmetry. The A0 compositions of these orbitals are identical to that shown in Fig.4 for the 4a, HOMO of H,Pc. The energy of the HOMO increases by 0.37 eV in passing from H,Pc to 2,3-H2Nc.This increase agrees with the shift of 0.41 eV to low binding energies observed when comparing the position of the first photoemission band in the UPS spectra of vapour- deposited thin films of ZnPc (6.29 eV) and 2,3-ZnNc (5.88 eV).69 The calculated HOMO destabilization further justifies the lower oxidation potentials measured for 2,3-naph- thalocyanine compounds. The naphthalocyanine macrocycle is easier to oxidize than the phthalocyanine macrocycle by ca. 0.4 V, e.g. 2,3-SiNc(OR),, +0.58 V; SiPc(OR),, + 1.00 V; 2,3- CoNc, +0.77 V; CoPc, + 1.15 V (potentials US. SCE).70,71 Owing to the low oxidation potentials of the 2,3-Nc macro- cycle, the 1,4-diisocyanobenzene bridged polymer [2,3-FeNc(dib)]. is doped by oxygen in the air and presents higher powder conductivities than those measured for the ‘undoped’ [FePc(dib)].p~lymer.~~?~~ The HOMO undergoes an additional destabilization of 0.17 eV when passing from 2,3-H2Nc to 2,3-H2Ac in agreement with the cyclic voltammetry data reported by Hanack et aL3’ J. Muter. Chem., 1996, 6(11), 1751-1761 1755 for a bis( pyridine) complex of (But),-2,3-FeAc. The oxidation of the macrocycle in this compound appears as a shoulder near the solvent limit at about +0.75 V and is compared with that measured for 2,3-FeNc at about +0.9 V (potentials us. SCE). A small destabilization of 0.09 eV is finally obtained in going from 2,3-H2Ac to 2,3-H2Tc. Fig. 6(a) summarizes the evolution of the energies of the HOMO and the LUMO as a function of the number of benzenes annulated following the axis of each pyrrole unit.The tetraazaporphyrin ring is taken as a reference. H,Pc, 2,3- H,Nc, 2,3-H2Ac and 2,3-H2Tc therefore correspond to annu- lation of one, two, three and four benzene rings per pyrrole unit, respectively. The energy of the HOMO increases along this series and shows an asymptotic behaviour converging to an extrapolated value of -5.69 eV. This value was obtained by fitting the energies calculated for the HOMO to a poly- nomial in inverse powers of N + 1 [y=a +b/(N+1)+ . . . +e/(N + 1)4], N being the number of benzene rings linearly fused to the tetraazaporphyrin ring per pyrrole unit. HOMO destabilizations smaller than 0.05 eV are therefore expected for extended macrocycles beyond 2,3-H2Tc. The rapid convergence achieved for a small number of benzene rings results from the fact that the contributions of the outermost rings to the HOMO decrease as the number of these rings increases.As discussed above for H2P and H,Pc and shown in Fig. 4, the destabilization of the HOMO with linear benzoannulation is due to the antibonding interactions with the fused fragments. The HOMO remains, however, mainly localized on the central tetraazaporphyrin ring and these interactions decrease in inten- sity with the number of benzene rings fused. The evolution of the energy of the lowest-lying two unoccu- pied orbitals with the extension of the macrocycle is not as regular as for the HOMO. The energy of the LUMO increases with linear benzoannulation, changing from -5.84 eV for I I I I I I 1 0 1 2 3 4 5 2.2 2 2 1.8 1 03 1.6 5 $1.4 U 1.2 l! I I I I I 0 1 2 3 4 5 N Fig.6 Variations of (a) the energy (E) of the HOMO and the LUMO and (b) the HOMO-LUMO energy gap (AEHOM~LUMO)for linearly extended phthalocyanines as a function of the number of benzene rings fused per pyrrole unit (N). The tetraazaporphyrin ring is taken as a reference (N=O). H,Pc, 2,3-H2Nc, 2,3-H2Ac and 2,3-H2Tc correspond to N= 1, 2, 3 and 4, respectively. The energies of the LUMO are shifted up by 0.62 eV as explained in the text. H,Tap to -5.23 eV for H,Pc and to -5.02 eV for 2,3-H2Nc. This trend, however, changes in passing to more extended systems like 2,3-H2Ac and 2,3-H2Tc, for which the LUMOs appear at -5.04 and -5.22 eV, respectively.The LUMO and LUMO + 1 of 2,3-H2Ac and 2,3-H2Tc exhibit in principle the same topologies as those shown in Fig. 5 for H,Pc, but with important contributions from the anthracene and tetracene units, respectively. The participation of the peripheral poly- acenic units is augmented along the series 2,3-H2Nc, 2,3-H2Ac, 2,3-H2Tc because the LUMOs of these units decrease in energy as the number of benzene rings increases: naphthalene, -3.91 eV; anthracene, -4.58 eV; tetracene, -4.99 eV. The interaction of the LUMO of the polyacenic units with the C,N, central ring along the x (b3,) or y (b,,) axes determines the stabilization of the lowest unoccupied orbitals in 2,3-H2Ac and 2,3-H,Tc and, as a consequence, the inversion in the evolution of the LUMO energy [see Fig.6(a)]. The energy of the LUMO is therefore not converged and lower energies are to be expected for more extended systems. The variations calculated for the energies of the HOMO and the LUMO jointly determine a continuous decrease of the HOMO-LUMO energy gap along the series H,Tap (2.08 eV), H,Pc ( 1.81eV), 2,3-H2Nc ( 1.65 eV), 2,3-H2Ac (1.46 eV), 2,3-H2Tc (1.19 eV). (As discussed in the Computational details section, these energies are obtained after adding up 0.62eV to compare with optical absorption data.) The narrowing of the HOMO-LUMO gap agrees with the bathochromic shift observed experimentally for the first absorption Q band along this series.Kobayashi20b has studied the UV-VIS-NIR absorption spectra of tetra-tert-butylated H,Tap, H,Pc, 2,3-H2Nc and 2,3-H2Ac in pyridine solution. The lowest optical transition appears at 619 nm (2.00 eV), 698 nm (1.78 eV), 784 nm (1.58 eV) and 858 nm (1.45 eV), respectively. Similar values were obtained by Hanack et ~1.~' for bidentate complexes of FePc (658 nm, 1.88 eV), 2,3-FeNc (751 nm, 1.65 eV) and 2,3-FeAc (829 nm, 1.50 eV). The Q band of vanadyl-2,3-tetracenocyanine, the only tetracenocyanine reported as yet, appears at 1055 nm (1.17 eV).,* The VEH results are in good agreement with all these experimental data. The energy of the Q band therefore seems to be far from being converged since a large bathochromic shift of ca.0.30eV is predicted theoretically and is observed experimentally in pass- ing from anthracyanines to tetracenocyanines. Fig. 6(b) represents the evolution of the HOMO-LUMO energy gap with the number of benzenes fused, N. The fitted curve shows a change in concavity with an inflexion point for N =2, indicating that the difference between the HOMO-LU MO gap calculated for two consecutive linearly annulated macrocycles decreases in going from H,Tap (N=0) to 2,3-H2Nc (N=2) but increases in going from 2,3-H2Nc to 2,3-H2Tc (N=4). This behaviour explains the non-asymptotic evolution of the energy of the Q band and suggests that greatly extended phthalocyanines could exhibit strong absorption bands in the near-IR with wavelengths significantly longer than 1000 nm.This possibility is especially attractive for the obtention of new photosensitizers and for the development of new materials for optoelectronics. We now turn to a discussion of the electronic structure associated with the occupied electronic levels lying between -7 and -10 eV (see Table 1). The 7bl, HOMO -1 of H,Pc (-9.01 eV) is not affected by the extension of the macrocycle because it is localized on the nitrogen atoms (see Fig. 4). Molecular orbitals with this A0 composition are found at exactly the same energy for 2,3-H2Nc (9b1,), 2,3-H2Ac (llblu) and 2,3-H2Tc ( 13bl,). All the remaining occupied orbitals above -10 eV are dominated by contributions from the poly- acenic units. For H,Pc, the four nearly degenerate MOs lying at -9.4 eV are located on the benzene moieties and their A0 composition corresponds to that of the le,, orbital of benzene with no contribution on the para carbons depicted in Fig.7(a). 1756 J. Muter. Chem., 1996, 6(11), 1751-1761 Fig. 7 Atomic orbital compositions and energies of the highest occupied molecular orbitals of (a) benzene, (b) naphthalene, (c) anthracene and (d) tetracene. Only orbitals above -10.0 eV are shown. The sizes of the circles are proportional to the magnitude of the LCAO coefficients. The following group of four orbitals between -9.6 and -10.0eV originates in the other le,, orbital of benzene. Similarly to H,Pc, two groups of four MOs appear for 2,3- H,Nc above -10eV. The A0 composition of the MOs forming these two groups indicates that they correspond to the la, HOMO and 2b1, HOMO-1, respectively, of the naphthalene molecule shown in Fig.7(b). The two groups of orbitals are in fact calculated to be centred at energies of about -8.6 and -9.2 eV, very similar to those found for the la, (-8.5 eV) and 2b,, (-9.3 eV) orbitals of naphthalene. For 2,3- H,Ac and 2,3-H2Tc, three groups of four MOs appear above -10 eV with topologies and energies identical to those shown in Fig. 7(c),(d) for the highest three occupied orbitals of anthra- cene and tetracene, respectively. In summary, each MO of the polyacenic units generates a group of four MOs in the macrocy- cle constituting a pair of b,,-b 3g. orbitals and two a, or b,, orbitals (see Table 1).The electronic structure of the macrocycle is thus dominated by the electronic levels of the peripheral polyacenic units and only the HOMO and the bl, orbital at -9.0 eV remain as electronic features intrinsically correspond- ing to the tetraazaporphyrin macrocycle. This has been shown for 2,3-naphthalocyanine, for which the photoelectron spec- unoccupied polyacenic orbitals that have appeared for extended Pcs. For 2,3-H,Ac, low-energy transitions can occur, for instance, between the 8a, HOMO and the llb3,,llb2, unoccupied levels or between the 7a,,6aU occupied orbitals and the 10b,,,10b3, LUMOs. One-dimensional electronic band structures The crystal structures of H,Pc and MPcs are well known since large monocrystals can be easily grown by sublimation at 400-500 "C, allowing for very precise X-ray diffraction mea~urements.~~~'~~~~The crystal growth of various MPcs has been studied recently in solid films prepared by vacuum deposition73 and molecular beam epita~y~~techniques.Phthalocyanines crystallize in columnar stacks where the planar macrocyclic molecules lie parallel to one another. Most of them are obtained in the so-called a-or in the thermo- dynamically more stable P-polymorphic form, where the Pc molecules adopt an 'eclipsed, slipped' stacking mode, as depicted in Fig. 8(a). The stacking axis forms an angle $ with the normal to the molecular plane, which has values of about 25" in the a-polymorph7' and of about 45" in the p-p~lymorph.~~-~~,~~ The 'face-to-face, staggered' molecular stacking displayed in Fig.8(b), where the stacking axis is perpendicular to the molecular plane and adjacent macrocycles in the stack are rotated alternatively by an angle 8, is obtained for doped Pcs like H,PcI (8= 40.0°)8" or NiPcI (8= 39.5")77 and also for one- dimensional polymeric Pcs like [ SiPcO], (8= 39").78 The dis; tance between adjacent moleFules in the stack is 3.23-3.25 A for doped Pcs~~,~~,~~ Phthalocyanineand 3.33 A for [ S~PCO],.~~ molecules adopt the 'face-to-face, eclipsed' orientation shoyn in Fig. 8(c) when the intrFstack interplanar spacing is > 3.50 A, as in [GePcO], (3.53 A)78 or in fluorinated polymers like [AlPcF], (3.66 A),80a [GaPcF], (3.87 and [ FePcF], (3.86 A).81 Kobayashi and U~eda~~ obtained thin [GePcO], films for w$ch the interplanar spacing was established to be only 3.40 A and adjacent rings in the stack were therefore observed to be staggered by an angle of 37".The crystal growth and molecular stacking of 2,3-metallo- naphthalocyanines (2,3-MNcs) have been investigated recently trum exhibits a one-to-one correspondence with the photoemis- sion bands observed for na~hthalene.~'" The appearance of polyacene-like occupied MOs at increas- ing energies above -9.0 eV and unoccupied MOs close to the b2g-b3g LUMOs explains the evolution of the optical properties $::/;IE -7 in the region of the B band. For H,Pc, the B band is centred 0,at 340 nm63 and has been mainly assigned to electronic exci- M-3--.A -8tations from the 7bl, HOMO-1 to the LUMO and LUMO+l.Since the energy of the 7bl, orbital remains constant upon benzoannulation, similar excitation energies should be expected for extended Pcs. This is exactly what is observed for naphthalocyanines and anthracyanines, for which the B band appears centred at about 340-360 nm.20b,29730 New absorption features are, however, detected on the low-energy side of the B band for these compounds. For instance, 2,3- anthracyanines show two weak absorption bands at about 440 and 570nm.20b,29,30 These new bands are due to electronic transitions involving the high-lying occupied and low-lying I I 0 da da da k k k Fig. 8 (a)-(c) Molecular stackings adopted by phthalocyanine-type macrocycles in crystals and thin solid films: (a) 'eclipsed, slipped'; (b) 'face-to-face, staggered'; (c) 'face-to-face, eclipsed'.(d)-( f ) VEH band structures calculated for 2,3-H2Nc along the one-dimensional stacks depicted on the top in (a)-(c). VB and CB denote valence and conduction bands, respectively. Unoccupied bands are shifted up by 0.62 eV to correct for the low excitation energies provided by the VEH method, J. Muter. Chern., 1996, 6(11), 1751-1761 1757 for various MNcs in epitaxial thin films vacuum deposited on alkali-metal halide substrates by means of X-ray and electron diffraction, IR spectroscopy and transmission electron microscopy.82 Divalent 2,3-H2Nc and 2,3-ZnNc are found to present a 'face-to-face, eclipsed' molecula! stacking with interplanar distances of only 3.29 and 3.31 A, respectively.82c As discussed above, this molecular stacking has not been reported for MPc crystals and is only observed f9r [MPcL] polymers when the inter-ring distance is > 3.50 A.Trivalent AlNcCl and GaNcF and tetravalent VONc are all found to present an 'eclipsed, slipped' stacking mode.82" To the best of our knowledge, no structural determination has been reported for more extended phthalocyanines. The molecular packing in columnar stacks determines a direct n-n interaction between adjacent molecules in the stack, but very weak interactions between molecules in adjacent stacks. One-dimensional VEH calculations have therefore been performed along the direction which gives rise to a significant n interaction, i.e.the stacking direction. Calculations have been carried out for the three molecular stackings depicted in Fig. 8(a)-(c). For the 'eclipsed, slipped' stacking, the crystalline data reported by Hoskins et aL5l for the more stable p-polymorphic form of H2Pc (intermolecular distance along the stacking axis, d = 4.73 A; angle of tilting, I,!I= 45.7") are adopted in the calculations. For the 'face-to-face, staggered' and 'face- to-face, edipsed' stackings, the data reported for doped H,PcI (d= 3.25 +,8 = 40.0")8" and for thin films of undoped 2,3-H2Nc (d= 3.29 A, 8= 0°)82care used, respectively. The VEH electronic band structures calculated for 2,3-H2Nc are drawn schematically in Fig.8(d)-(f ) below the respective molecular stackings. Similar band structures are obtained for H,Pc and 2,3-H2Ac. For 2,3-H2Tc, only the band structure corresponding to the 'face-to-face, eclipsed' stacking was calcu- lated due to the large size of the molecular system. Table 2 summarizes the VEH values computed for the valence band- width (W),ionization energy (&) and bandgap (EJ. The results of band-structure calculations on the angularly benzoannulated H,Phc are included for the sake of comparison. As illustrated in Fig. 8(d)-(f), the band structures show the same general aspect irrespective of the molecular stacking. The valence band (VB) lies alone well above the remaining occupied bands. It originates in the overlap of the a,(n) HOMOs of the macrocyclic monomers along the stack.The conduction band (CB) is formed by two nearly degenerate bands running parallel and resulting from the overlap of the b2g-b3g LUMOs. The electronic properties collected in Table 2 are, however, strongly affected by the stacking mode, which governs the overlap between adjacent molecules in the stack. The width of the VB increases in passing from the 'eclipsed, slipped' mode (0.34-0.36 eV) to the 'face-to-face, staggered' (0.82-1.00 eV) and 'face-to-face, eclipsed' (1.16-1.20 eV) modes owing to the Table 2 VEH-calculated width of the highest occupied electronic band (W),ionization energy (Ei) and bandgap energy (E,) of H,Pc, 2,3- H,Nc, 2,3-H2Ac, 2,3-H2Tc and H2Phc in various one-dimensional molecular stacks" 'eclipsed, slipped' 'face-to-face, 'face-to-face, staggered' eclipsed' W E, Egb W Ei E,b W Ei E,b H,Pc 0.34 6.16 1.76 0.92 5.95 1.64 1.20 5.81 1.74 2,3-H,Nc 0.36 5.87 1.51 0.86 5.61 1.43 1.18 5.45 1.58 2,3-H,Ac 0.35 5.70 1.21 0.82 5.45 1.11 1.17 5.28 1.39 2,3-H2Tc 1.16 5.20 1.13 H2Phc 0.35 6.57 1.74 1.00 6.32 1.58 1.19 6.24 1.72 "All values are ineV.bThe numbers quoted correspond to the minimum value of the direct energy gap. They are obtained after a shift of 0.62 eV to higher energies of the unoccupied bands to correct for the low excitation energies provided by the VEH method. 1758 J. Muter. Chern., 1996, 6(11), 1751-1761 more effective overlap achieved in these two stacking modes. The expansion of the macrocycle affects only slightly the width of the valence band because the outer benzene rings contribute to the HOMO to a lesser degree as the system is extended. Similar W values are thus obtained for all the macrocycles included in Table 2.The band structure of 'face-to-face' H,Pc stacks was calcu- lated for differept values of the staggering angle 8 (inter-ring distance=3.25 A), in order to study how the width of the valence band varies with rotation of neighbouring rings. Fig. 9 shows the variation calculated for W between 8=0 and 45". A symmetric behaviour is to be expected between 8=45 and 90" owing to the almost identical geometries of the isoindole moieties in H,Pc. The bandwidth presents a maximum value of 1.32 eV for the fully eclipsed rotamer (8=Oo) and a second maximum of 0.99 eV for the fully staggered rotamer (8=45").Between the two maxima the valence band becomes completely flat for a staggering angle of about 20" (W=0.03 eV). Similar behaviours were obtained by Whangbo and Stewarts3 using the extended-Huckel method and by Pietro et who calculated the transfer integral between adjacent rings for a phthalocyanine dimer using the Wolfsberg-Helmholtz approach. The variation of the bandwidth with 8 can be understood by reference to Fig. 10, where the interacting patterns at k=O between the HOMOs of two adjacent molecules are sketched for the 8=0 and 45" rotamers using only the C8N8 central ring. For the eclipsed rotamer, the interaction between neigh- bouring a,(n) orbitals is antibonding, while for the staggered rotamer it is less effective and bonding. Thus, in going from 8=0 to 45", the bonding interactions decrease and the snti- bonding interactions increase in such a way that they cancel each other at intermediate angles.The resulting non-bonding 1.4 1.2 1 0.8>z30.6 0.4 0.2 01 " I 0 15 30 45 eldegrees Fig.9 Variation of the valence bandwidth (w)with the staggering angle I3 between neighboucing rings in 'face-to-face' stackings of H,Pc (inter-ring distance = 3.25 A) Fig. 10 Interacting patterns at k=O between the a,(n) HOMOs of neighbouring H,Pc molecules arranged in 'face-to-face' stacks with staggering angles of (a) 8=Oo and (b) I3=45". H,Pc macrocycles are modelled by the C8N8 central ring.interaction explains the vanishing bandwidth obtained at about 20”. Phthalocyanines change from an ‘eclipsed, slipped’ stack- ing in the undoped state to a ‘face-to-face, staggered (8=40”)’ stacking when cocrystallized with oxidizing agents like iod- ine.8n,77,79This structural change is accompanied by a broaden- ing of the partially emptied valence band which facilitates the delocalization of the electrons, i.e. the charge carriers, giving rise to high metallic conductivities ( 101-104 S cm-’).8 As Pietro et al. discussed,84 if the doped material stacked with a staggering angle of ca. 20”, conductivity would only occur through an electron-hopping mechanism and the material would exhibit intrinsically activated conductivity instead of intrinsically metallic.Canadell and Alvarez also noted the relevance of the broadening of the VB for doped Pcs.~’ The MO interacting patterns depicted in Fig. 10 also explain the change in the sign of the slope of the VB in passing from the staggered to the eclipsed ‘face-to-face’ stacking [see Fig. 8(e) and (f)]. For the latter, the interaction is antibonding at k=O (in-phase translation) and bonding at k =n/a (out-of-phase translation). The VB therefore decreases in energy towards k= n/a and the highest occupied level corresponds to k=O. The opposite occurs for the staggered conformation, for which the interaction is bonding at k =0 and antibonding at k =n/a and the VB increases in energy towards k=n/a.This is also the case for the ‘eclipsed, slipped’ stacking mode [Fig. 8(d)]. The variation of the width of the valence band with the intermolecular distance was calculated for both the eclipsed (8=0”)and staggered (0=45”) ‘face-to-face’ stacking modes of H,Pc. As expected, the bandwidth decreases with increasing inter-ring distances owing to the weaker interactions that occur between adjacent macrocycles in the stack. For the eclipsed stacking, W is calculated to have values of 1.20,0.68, 0.52, 0.39 and 0.29 eV fy-intermolecular distances of 3.29, 3.50, 3.60, 3.70 and 3.80 A, respectively. Smaller values of 0.92, 0.81, 0.62, 0.47 and 0.36 eV are found fpr intermolecular distances of 3.25, 3.30, 3.40, 3.50 and 3.60A, respectively, for the staggered conformation.These bandwidths are an aid to understanding the room-temperature conductivities (uRT) measured for polycrystalline samples of doped phthalocyanine-based mate- rials. For example, the conductivity decreases along the series NiPcI (cRT >[(S~PCO)I~.~~],,=7.7 S ~m-~)~~~(0.67)86> [(G~PCO)~~.,~],,(0.11).86 All these materials present a ‘face- to-face, staggered’ stacking with 8=39:40” and intermol- (see Table 2). The decrease in the Ei is due to the larger values of the valence bandwidth. Lower ionization energies are obtained for linearly extended 2,3-H2Nc, 2,3-H2Ac and 2,3- H,Tc because of the continuously increasing energies of the HOMOS along this series. Minimum Ei values of only 5.45, 5.28 and 5.20 eV are thus calculated for the ‘face-to-face, eclipsed’ stackings of 2,3-H2Nc, 2,3-H2Ac and 2,3-H2Tc, respectively. The angularly annulated H,Phc system has Ei values ca.0.4eV higher than those in H,Pc, in accord with the lower energy of the HOMO of H,Phc. The values quoted in Table 2 for the energy gap separating the valence and conduction bands correspond to the minimum values of the direct energy gap found at k =n/afor the ‘eclipsed, slipped’ and ‘face-to-face, staggered’ stacking modes and at k =0 for the ‘face-to-face, eclipsed’ stacking mode. The E, calculated for the ‘face-to-face, staggered’ stacking mode is smaller than that calculated for the ‘face-to-face, eclipsed’ stacking mode because the CB is significantly narrower (e.g.0.48eV for 2,3-H2Nc) than the VB (0.86eV) for the former, while both bands have similar widths (1.04 and 1.18 eV, respectively) for the latter. The energy gap decreases for linearly extended systems and remains constant with respect to H,Pc for angularly extended systems such as H,Phc. Finally, it is important to note that the energy gap separating the top of the valence band from the bottom of the conduction band does not correspond to the direct energy gap at k=O or k =n/a, but to the indirect gap separating these two points of the Brillouin zone [see Fig. 8(d)-(f)]. An indirect E, of only 0.54 eV is calculated for the ‘face-to-face, eclipsed’ stacking observed experimentally for vacuum-deposited thin films of ~,~-H,Nc.~,‘The indirect E, decreases to 0.35 eV for 2,3-H,Ac and almost vanishes (0.09 eV) for 2,3-H2Tc.Since the intrinsic conducting properties of the material are largely determined by the energy gap separating the conduction and valence bands, high intrinsic conductivities are to be expected for stacks of linearly extended phthalocyanines. Summary and Conclusion The geometric and electronic structures of metal-free phthalo- cyanine and a series of linearly benzoannulated phthalocyan- ines have been investigated using the semi-empirical PM3 method and the non-empirical valence effective Hamiltonian ecular distances of 3.244, 3.30 and 3.48 A, re~pectively.~~“*~~ (VEH) pseudopotential technique. Geometry optimizations The increase of the intermolecular distance is accompanied by a narrowing of the partially emptied VB from ca.0.92 eV to ca. 0.47eV, thus explaining the decrease in the observed conductivities. Relatively high conductivities are also measured for doped polymeric Pcs such as [(A~PCF)(IF~)~.~-~.~], (OR= = 6 x lo-, S ~m-’),~~ (4.2 x[(G~PcF)(BF~),.~~]~ and [(F~PCF)I~.~],(5 x 10-3),81 yhich show longer inter-ring dis- tances of 3.66, 3.73 and 3.86 A, respectively. These conductivit- ies are explained as resulting from the fact that the eclipsed ring stacking observed for the undoped polymers persists for the doped materials. Thus, the inter-ring overlap loss due to the increase of the inter-ring separation is made up in large part by the more efficient overlap in the eclipsed conformation.Bandwidths between ca. 0.500and 0.25 eV are found for inter- ring distances of 3.60-3.90A in an eclipsed stacking mode, which are in fact slightly smaller than those found for a staggered stacking mode with inter-ring distances of 3.40-3.50 A (0.62-0.47 eV). The change in the sign of the slope of the VB depending on the molecular stacking mode means that the ionization energy onset corresponding to the top of the VB is situated at the edge of the Brillouin zone (k=n/a) for the ‘eclipsed, slipped‘ and ‘face-to-face, staggered’ stacking modes and at the centre of the Brillouin zone (k=0) for the ‘face-to-face, eclipsed’ stacking mode. For H,Pc, the ionization energy onset is calculated to have values of 6.16, 5.95 and 5.81 eV, respectively show that the peripheral polyacenic units (benzene, naphtha- lene, anthracene, etc.) preserve their structural identity in the macrocycles. Phthalocyanine-based compounds are therefore predicted to be the result of joining four polyacenic units to the C8N8 central ring.Linear benzoannulation strongly affects the electronic properties of the macrocycle. On the one hand, it produces a continuous destabilization of the HOMO, thus leading to lower ionization energies and oxidation potentials. On the other hand, it reduces the HOMO-LUMO energy gap, thus shifting the intense Q absorption band to the near-IR region. It is to be stressed that, in contrast to the HOMO, which shows asymptotic behaviour rapidly converging to an extrapo- lated energy of -5.69 eV, the HOMO-LUMO gap exhibits non-convergent behaviour and very low excitation energies are predicted for greatly extended phthalocyanines. Linear benzoannulation thus allows the modulation of the electronic properties and, in particular, the optical properties of the phthalocyanine macrocycle by controlling the number of ben- zene rings fused to each pyrrole axis.The extension can in fact be effected in a symmetric way on the four pyrrole axes as performed in this work, or in an asymmetric way by only extending one or two pyrrole axes as Kobayashi et al.*’ reported recently. The combination of these two strategies makes very fine adjustments of the redox and optical properties of the system possible.J. Mater. Chern., 1996, 6(11), 1751-1761 1759 VEH electronic band-structure calculations have been per- formed for the three one-dimensional molecular stacks more commonly found in crystals and polymers of phthalocyanines: ‘eclipsed, slipped’, ‘face-to-face, staggered’ and ‘face-to-face, eclipsed’. The width of the valence band, i.e. the band which is partially emptied upon oxidation for doped compounds, is 12 13 K. Abe, H. Sato, T. Kimura, Y. Ohkatsu and T. Kusano, Makromol. Chem., 1989,190,2693; T. L. Anderson, G. C. Komplin and W. J. Pietro, J. Phys. Chem., 1993,97, 6577; W. J. Pietro, Adv. Muter., 1994,6,239. R. Madru, G. Guillaud, M. A1 Sadoun, M. Maitrot, C. Clarisse, M. Le Contellec, J.-J. AndrC and J. Simon, Chem. Phys. Lett., 1987, 142, 103; R.Madru, G. Guillaud, M. A1 Sadoun, M. Maitrot, J- not affected by the extension of the macrocycle but depends strongly on the molecular stacking mode. For ‘face-to-face’ stacks, the bandwidth is determined by the twist angle, 8, between adjacent macrocycles and by the intermolecular dis- tance. A coherent picture of the conductivities observed exper- 14 15 J. Andre, J. Simon and R. Even, Chem. Phys. Lett., 1988, 145, 343; C. Clarisse and M-T. Riou, J. Appl. Phys., 1991,69,3324. J. E. Kuder, J. Zmag. Sci., 1988,32, 51; R. Ao, S. Jahn, L. Kummerl, R. Weiner and D. Haarer, Jpn. J. Appl. Phys., 1992,31, 693; R. Ao, L. Kummerl and D. Haarer, Adv. Muter., 1995,7,495. J. Simon, P. Bassoul and S. Norvez, New J. Chem., 1989, 13, 13; imentally for doped phthalocyanines is obtained by analysing M.K. Casstevens, M. Samoc, J. Pfleger and P. N. Prasad, J. Chem. the evolution of the bandwidth with 8 and the inter-ring distance. The linear expansion of the macrocycle produces a decrease of the ionization potential and the bandgap, similar to that found for the molecular systems. Indirect bandgaps of the order of 0.5 eV and even lower are predicted for ‘face-to- 16 Phys., 1990, 92, 2019; A. Grund, A. Kaltbeitzel, A. Mathy, R. Schwarz, C. Bubeck, P. Vermehren and M. Hanack, J. Phys. Chem., 1992, 96, 7450; H. S. Nalwa, T. Saito, A. Kakuta and T. Iwayanagi, J. Phys. Chem., 1993,97,10515. G. Corker, B. Grant and C. Clecak, J. Electrochem. SOC.,1979,126, 1339; G. C. S. Collins and D. J. Schffrin, J. Electrochem. SOC.,1985, face’ stacks of 2,3-H2Nc, 2,3-H2Ac and 2,3-H2Tc, suggesting that high intrinsic conductivities can be obtained for crystals of greatly extended linearly benzoannulated phthalocyanines.In summary, the results provided by VEH calculations agree with photoemission, cyclic voltammetry and optical absorption experimental data and thus allows us: (i) to understand the 17 18 132, 1835; M-T. Riou and C. Clarisse, J. Electroanal. Chem., 1988, 249, 181; M. M. Nicholson, in ref. 2, 1993, vol. 3, ch. 2, p. 71. J. F. van der Pol, E. Neeleman, J. W. Zwikker, R. J. M. Nolte, W. Drenth, J. Aerts, R. Visser and S. J. Picken, Liq. Cryst., 1989,6, 577; J. Simon and P. Bassoul, in ref. 2, 1993, vol. 2, ch. 6, p. 223. G. G. Roberts, M. C. Petty, S. Baker, M. T. Fowler and N.J. Thomas, Thin Solid Films, 1985, 132, 113; M. J. Cook, distinctive electronic properties that characterize the parent phthalocyanine; (ii) to rationalize the general trends observed for extended phthalocyanines, and (iii) to predict interesting properties for greatly extended systems. Valuable predictions are in particular performed concerning the optical and electri- cal properties. 19 A. J. Dunn, M. F. Daniel, R. C. 0.Hart, R. M. Richardson and S. J. Roser, Thin Solid Films, 1988, 159, 395; S. Palacin, P. Lesieur, I. Stefanelli and A. Barraud, Thin Solid Films, 1988, 159, 83. I. Rosenthal and E. Ben-Hur, in ref. 2, 1989, vol. 1, ch. 6, p. 393; B. A. Henderson and T. J. Dougherty, Photochem. Photobiol., 1992, 55, 145; S. B. Brown and T. G. Truscott, Chem.Br., 1993,29,955; D. Phillips, Pure Appl. Chem., 1995,67,117; R. Bonnett, Chem. SOC. Rev., 1995, 24, 19. The research presented in this work has been supported by the CICYT Grants MAT95-1630-E, IN950326 and PB95- 0428C02 and by the Generalitat Valenciana. We thank the CIUV (Centre d’informatica de la Universitat de Valencia) for the use of their computing facilities. Technical assistance from W. Diaz is gratefully acknowledged. E. 0.is indebted to Drs. R. Viruela and P. M. Viruela for their valuable collaboration. 20 21 (a) E. A. Luk‘yanets, Electronic Spectra of Phthalocyanines and Related Compounds, Tcherkassy, 1989; (b)N. Kobayashi, in ref. 2, 1993,vol. 2, ch. 3, p. 11 1. P. J. Kivits, M. R. J. de Bont and J. van der Veen, Appl. Phys. A, 1981, 36, 101; A.P. Piechowski, G. R. Bird, D. L. Morel and E. L. Stogryn, J. Phys. Chem., 1984, 88, 934; T. Tomiyama, I. Watanabe, A. Kuwano, M. Habiro, N. Takane and M. Yamada, Appl. Opt., 1995,34, 8201. 22 P. A. Firey, W. E. Ford, J. R. Sounik, M. E. Kenney and References M. A. J. Rodgers, J. Am. Chem. SOC., 1988, 110, 7626; B. Roeder, D. Naether, T. Lewald, M. Braune, C. Nowak and W. Freyer, 1 2 3 4 5 6 7 A. Braun and J. Tcherniac, Ber. Dtsch. Chem. Ges., 1907,40,2709. For a recent and very comprehensive review, see: Phthalocyanines: Properties and Applications, ed. C. C. Leznoff and A. B. P. Lever, VCH, New York, 1989-1993, vol. 1-3. J. Simon and J-J. AndrC, Molecular Semiconductors. Photoelectrical Properties and Solar Cells, Springer, Berlin, 1985, p. 73.A. T. Vartanyan, Zh. Fiz. Khim., 1948,22, 769. R. 0. Loutfy and J. H. Sharp, J. Chem. Phys., 1979, 71, 1211; C. W. Tang, Appl. Phys. Lett., 1986,48, 183. R. 0.Loutfy, A. M. Hor, C. K. Hsiao, G. Baranyi and P. Kazmaier, Pure Appl. Chem., 1988, 60, 1047; 0. Stenzel, A. Stendal, K. Voigtsberger and C. von Borczyskowski, Sol. Energy Muter. Sol. Cells, 1995,37, 337. For a review, see: B. M. Hoffman and J. A. Ibers, Acc. Chem. Res., 23 24 25 26 27 28 29 Biophys. Chem., 1990,35,303. S. Deger and M. Hanack, Synth. Met., 1986,13, 319. H. S. Nalwa, A. Kakuta and A. Mukoh, Chem. Phys. Lett., 1993, 203,109; J. Phys. Chem., 1993,97,1097. E. Orti, M. C. Piqueras, R. Crespo and J. L. Bredas, Chem. Muter., 1990,2, 110. E. Orti, R. Crespo, M. C. Piqueras and J. L. Bredas, Synth. Met., 1991,42,2647.(a)V. N. Kopranenkov and E. A. Luk’yanets, J. Gen. Chem. USSR, 1971, 41, 2366; (b) V. N. Kopranenkov, G. I. Rumyantzeva and E. A. Luk’yanets, Ani1.-Dyes Znd., 1974,2, 6. W. Freyer and L. Q. Minh, Monatsh. Chem., 1986, 117, 475; W. Freyer, Z. Chem., 1986,26,216. N. Kobayashi, S. Nakajima and T. Osa, Znorg. Chim. Acta, 1993, 210, 131. 8 1983, 16, 15; T. J. Marks, Science, 1985, 227, 881; Angew. Chem., Znt. Ed. Engl., 1990, 16, 857. (a) T. Inabe, T. J. Marks, R. L. Burton, J. W. Lyding, 30 31 M. Hanack, R. Dieing and U. Rohrig, Chem. Lett., 1993, 399. W. Freyer, H. Stiel, K. Teuchnerand and D. Leupold, J. Photochem. Photobiol. A, 1994,80, 161. W. J. McCarthy, C. R. Kannewurf, G. M. Reisner and F. H. Herbstein, Solid State Commun., 1985, 54, 501; (b) J.A. Thompson, K. Murata, D. C. Miller, J. L. Stanton, W. E. Broderick, B. M. Hoffman and J. A. Ibers, Znorg. Chem., 1993,32,3546. 32 33 34 J. J. P. Stewart, J. Comput. Chem., 1989, 10,209; 221. J. J. P. Stewart, MOPAC: A General Molecular Orbital Package (Version 6.0), QCPE, 1990,455, 10. T. X. Lu, J. R. Reimers, M. J. Crossley and N. S. Hush, J. Phys. Chem., 1994,98,11878. 9 (a)M. Hanack, A. Datz, R. Fay, K. Fischer, U. Keppeler, J. Koch, J. Metz, M. Mezger, 0. Schneider and H. J. Schulze, in Handbook of Conducting Polymers, ed. T. A. Skotheim, Marcel Dekker, New York, 1986, vol. 1, ch. 5, p. 133; (b) M. Hanack, S. Deger and A. Lange, Coord. Chem. Rev., 1988, 83, 115; (c) H. Schultz, H. Lehmann, M. Rein and M. Hanack, Struct. Bonding (Berlin), 1991,74,41; (d) M.Hanack and M. Lang, Adv. Muter., 1994,6,819. 35 36 37 G. Nicolas and Ph. Durand, J. Chem. Phys., 1979,70,2020; 1980, 72,453. (a) J. M. Andrk, L. A. Burke, J. Delhalle, G. Nicolas and Ph. Durand, Int. J. Quantum Chem. Symp., 1979, 13, 283; (b) J. L. Bredas, R. R. Chance, R. Silbey, G. Nicolas acd Ph. Durand, J. Chem. Phys., 1981,75,255. R. Viruela-Martin, P. M. Viruela-Martin and E. Orti, J. Chem. 10 M. Hanack, A. Lange and R. Grosshans, Synth. Met., 1991,45,59. Phys., 1992, 95, 4474. 11 T. Jones and B. Bott, Sens. Actuators, 1986,9,27;R. A. Collins and 38 (a) J. L. BrCdas, in Handbook of Conducting Polymers, ed. K. A. Mohamed, J. Phys. D, 1988, 21, 154; T. A. Temofonte and T. A. Skotheim, Marcel Dekker, New York, 1986, vol. 2, ch.25, K. F. Schoch, J. Appl. Phys., 1989, 65, 1350; Y. Sadaoka, p. 859; (b) J. M. Andre, J. Delhalle and J. L. BrCdas, Quantum T. A. Jones and W. Gopel, Sens. Actuators B, 1990, 1, 148; A. W. Chemistry Aided Design of Organic Polymers, World Scientific, Snow and W. R. Barger, in ref. 2, 1989, vol. 1,ch. 5, p. 341. Singapore, 1991, ch. 3, p. 163. 1760 J. Mater. Chern., 1996, 6(11), 1751-1761 39 E. Orti and J. L. Bredas, J. Chem. Phys., 1988,89,1009. Spectrosc., 1963, 11, 108; (b) C. Weiss, H. Kobayashi and 40 E. Orti and J. L. Bredas, Synth. Met., 1989,29, F115. M. Gouterman, J. Mol. Spectrosc., 1965,16,415. 41 E. Orti and J. L. Bredas, J. Am. Chem. SOC., 1992,114,8669. 69 D. Schlettwein and N. R. Armstrong, J. Phys. Chem., 1994, 98, 42 E.Orti and J. L. BrCdas, Chem. Phys. Lett., 1989, 164,247. 11771. 43 E. Orti and J. L. Bredas, Synth. Met., 1989,33,27. 70 M. Hanack, A. Lange, M. Rein, R. Behnisch, G. Renz and 44 E. Orti, J. L. BrCdas and C. Clarke, J. Chem. Phys., 1990,92, 1228. A. Leverenz, Synth. Met., 1989,29, F1. 45 E. Orti, R. Crespo, M. C. Piqueras, P. M. Viruela and F. Tomas, 71 B. L. Wheeler, G. Nagasubramanian, A. J. Bard, L. A. Schechtman, Synth. Met., 1993,57,4513;E. Orti, M. C. Piqueras, R. Crespo and D. Dininny and M. E. Kenney, J. Am. Chem. SOC., 1984,106,7404. F. Tomas, Synth. Met., 1993,57,4519. 72 R. P. Linstead and J. M. Robertson, J. Chem. SOC., 1936, 1736; 46 (a) J. M. Andre, J. L. BrCdas, J. Delhalle, J. G. Fripiat, J. M.Robertson and I. J. Woodward, J. Chem. SOC., 1937,219. D. J. Vanderveken and D. P. Vercauteren, in Modern Techniques in 73 T. Kobayashi and N. Uyeda, J. Cryst. Growth, 1987,84,589. Computational Chemistry, ed. E. Clementi, Escom, Leiden, 1990, 74 A. J. Dann, H. Hoshi and Y. Maruyama, J. Appl. Phys., 1990, 67, ch. 15, p. 745; (b) J. L. Bredas, B. Themans and J. M. Andre, 1371;H. Hoshi and Y. Maruyama, J. Appl. Phys., 1991,69,3046. J. Chem. Phys., 1983,78,6137. 75 C. J. Brown, J. Chem. SOC. A, 1968,2494. 47 J. L. BrCdas, B. ThCmans and J. M. AndrC, J. Chem. Phys., 1983, 76 R. Mason, G. A. Williams and P. E. Fielding, J. Chem. SOC., Dalton 78, 6137. Trans., 1979, 676. 48 R. Viruela-Martin, P. M. Viruela-Martin and E. Orti, J. Chem.77 C. J. Schramm, R. P. Scaringe, D. R. Stojakovic, B. M. Hoffman, 49 Phys., 1992,97, 8470. J. M. Robertson, J. Chem. SOC., 1935, 615. J. A. Ibers and T. J. Marks, J. Am. Chem. Soc., 1980, 102, 6702; T. Inabe, S. Nakamura, W-B. Liang, T. J. Marks, R. L. Burton, 50 J. M. Robertson, J. Chem. SOC., 1936, 1195. C. R. Kannewurf and K. I. Imaeda, J. Am. Chem. SOC., 1985, 107, 51 B. F. Hoskins, S. A. Mason and J. C. B. White, J. Chem. SOC., Chem. 7224; S. M. Palmer, J. L. Stanton, B. M. Hoffman and J. A. Ibers, Commun., 1969,554. Znorg. Chem., 1986,25,2296;M. Almeida, M. G. Kanatzidis, L. M. 52 K. Yase, N. Yasuoka, T. Kobayashi and N. Uyeda, Acta Tonge, T. J. Marks, W. J. McCarthy, H. 0. Marcy and C. R. Crystallogr., Sect. C, 1988,44, 514. Kannewurf, Solid State Commun., 1987,63,457.53 J. Almlof, T. H. Fischer, P. G. Gassman, A. Ghosh and M. Hase!, 78 C. W. Dirk, T. Inabe, K. F. Schoch and T. J. Marks, J. Am. Chem. J. Phys. Chem., 1993, 97, 10964; M. Merchan, E. Orti and B. 0. SOC.,1983,105, 1539. 54 Roos, Chem. Phys. Lett., 1994,221, 136. A. Ghosh, P. G. Gassman and J. Almlof, J. Am. Chem. SOC., 1994, 79 M. Almeida, M. G. Kanatzidis, L. M. Tonge, T. J. Marks, W. J. McCarthy, H. 0.Marcy and C. R. Kannewurf, Synth. Met., 116,1932. 1989,29, F37. 55 L. Nygaard, J. J. Nielsen, J. Kirchheiner, G. Maltesen, J. Rastrup- 80 (a)J. P. Linsky, T. R. Paul, R. S. Nohr and M. E. Kenney, Znorg. Andersen and G. 0.Sorensen, J. Mol. Struct., 1969,3,491. Chem., 1980, 19, 3131; (b)R. S.Nohr and K. J. Wynne, J. Chem. 56 D. W. J. Cruikshank, Acta Crystallogr., 1957, 10,470. SOC., Chem. Commun., 1981, 1210; (c) K. J. Wynne, Znorg. Chem., 57 J. Berkowitz, J. Chem. Phys., 1979,70,2819. 1985,24,1389. 58 H. Hochst, A. Goldmann, S. Hufner and H. Malter, Phys. Status 81 M. Futamata, S. Higuchi and S. Takahashi, Synth. Met., 1989, 59 Solidi B, 1976, 76, 559; F. L. Battye, A. Goldmann and L. Kasper, Phys. Status Solidi B, 1977, 80, 425; E. Tegeler, M. Iwan and E-E. Koch, J. Electron Spectrosc. Relat. Phenom., 1981,22,297. P. Dupuis, R. Roberge and C. Sandorfy, Chem. Phys. Lett., 1980, 82 30, 39. (a) H. Yanagi, M. Ashida, J. Elbe and D. Wohrle, J. Phys. Chem., 1990, 94, 7056; (b)H. Yanagi, T. Kouzeki, M. Ashida, T. Noguchi, A. Manivannan, K. Hashimoto and A. Fujishima, J. Appl. Phys., 75,434. 1992,71, 5146; (c)H. Yanagi, T. Kouzeki and M. Ashida, J. Appl. 60 P. J. Derrick, L. Asbrink, 0. Edqvist, B-0. Jonsson and E. Lindholm, Znt. J. Mass Spectrom. Zon Phys., 1971, 6, 191; Phys., 1993, 73, 3812; (d) A. Manivannan, L. A. Nagahara, K. Hashimoto, A. Fujishima, H. Yanagi, T. Kouzeki and L. Klasinc, A. Sabljic, G. Kluge, J. Rieger and M. Scholz, J. Chem. M. Ashida, Langmuir, 1993,9,771. SOC., Perkin Trans. 2, 1982, 539; L. Nyulaszi, T. Toth, G. Y. Zsombok, G. Csonka, J. Rkffy, J. Nagy and T. Veszpremi, 83 84 M-H. Whangbo and K. R. Stewart, Zsr. J. Chem., 1983,23,133. W. J. Pietro, T. J. Marks and M. A. Ratner, J. Am. Chem. SOC., J. Mol. Struct., 1990,218, 201. 1985,107,5387. 61 62 63 64 65 W. Rettig and J. Wirz, Helv. Chim. Acta, 1976,59, 1054. A. Rosa and E. V. Baerends, Znorg. Chem., 1993,32,5637. L. Edwards and M. Gouterman, J. Mol. Spectrosc., 1970,33,292. M. J. Stillman and T. Nyokong, in ref. 2, 1989, vol. 1, ch. 3, p. 133. (a) N. Kobayashi, H. Lam, W. A. Nevin, P. Janda, C. C. Leznoff, T. Koyama, A. Monden and H. Shirai, J. Am. Chem. SOC., 1994, 116,879;(b)M. G. Cory, H. Hirose and M. C. Zerner, Znorg. Chem., 1995,34,2969. 85 86 87 88 89 E. Canadell and S. Alvarez, Znorg. Chem., 1984,23,573. B. N. Diel, T. Inabe, J. W. Lyding, K. F. Schoch, Jr., C. R. Kannewurf and T. J. Marks, J. Am. Chem. SOC., 1983, 105, 1551. D. Djurado, A. Hamwi, C. Fabre, D. Avignant and J. C. Cousseins, Synth. Met., 1986, 16,227. M. Futamata, Synth. Met., 1991,41-43,2621. N. Kobayashi, R. Kondo, S. Nakajima and T. Osa, J. Am. Chem. 66 M. Merchan, E. Orti and B. 0. Roos, Chem. Phys. Lett., 1994, 226, 27. SOC., 1990, 112, 9640; N. Kobayashi, T. Ashida, T. Osa and H. Konami, Inorg. Chem., 1994,33,1735. 67 68 M. Gouterman, J. Mol. Spectrosc., 1961,6, 138. (a) M. Gouterman, G. H. Wagnikre and L. C. Snyder, J. Mol. Paper 6/03693G; Received 28th May, 1996 J. Muter. Chem., 1996,6(11), 1751-1761 1761

ISSN:0959-9428    

DOI:10.1039/JM9960601751

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

Improved electroluminescence performance of poly (3-alkylthiophenes) having a high head-to-tail (HT) ratio Freeman Chen," Parag G. Mehta,"" Larry Takiff" and Richard D. McCullough*b "Polaroid Corporation, 750 Main Street--SC, Cambridge, MA 021 39, USA bDepartment of Chemistry, Carnegie Mellon University, Pittsburgh, PA 1521 3-3890, USA Light-emitting diodes are fabricated from regioregular and non-regioregular poly( 3-hexylthiophenes). The regioregular polymer has a higher head-to-tail (HT) ratio than non-regioregular polymer. It is shown that the electroluminescent device containing high HT ratio polymer perform significantly better than polymers with a low HT ratio. Photoluminescence data of both polymers are also reported. The past few years have seen intense research activity'" in the area of electroluminescence (EL) of organic polymers since the initial reports by Friend, Homes et al.lb and Heeger et aE.IC Early reports described the use of conjugated polymers such as polyphenylene vinylene (PPV), and its derivatives such as poly [2-methoxy-5-( 2'-ethylhexy1oxy)- lY4-phenylene vinylene] (MEH-PPV), in which aromatic rings are connected through the double bonds.Later, all-aromatic polymers such as poly(p- phenylene),2" poly(alkylfluorenes)2b and poly(3-alkylthio-phene~)~were used as electroluminescent materials. We report our efforts on the study and use of poly(3- alkylthiophenes). The use of poly( 3-alkylthiophenes) in electro- luminescent devices is rather attractive. These polymers are relatively easy to synthesize, possess excellent processability and environmental stability.In addition, derivatized polythi- ophenes are not only fusible and soluble in common organic solvents but also afford the opportunity to tune the emission wavelength. Previous work3 has focused on the fabrication of electroluminescent devices based on poly (3-alkylthiophenes) synthesized by ferric chloride catalysed oxidative coupling of 3-alkylthiophene (Scheme 1). While thinking about improving the EL performance of the LEDs fabricated from polymers such as 1, we felt that there were at least two problems associated with the polymeric materials themselves and these problems are the result of the synthetic method employed (Scheme 1). One problem is the purity of the sample.Despite exhaustive cleaning of the poly- mer, it is very difficult to completely remove ferric salts from the product because these salts strongly bind with the thio- phene moieties. Contamination by ferric salts can not only affect the bandgap of the polymer but can also create quenching sites for the excitons and may thereby reduce the EL efficiency. A second problem4b is related to the microarchitecture of the polymer chains and is a direct result of the way in which the polymer chains grow in the oxidative coupling process. In 3- alkylthiophene, oxidative coupling takes place predominantly at 2- and 5-positions. Ideally, all repeat units will be linked in a head-to-tail (2,5) fashion as shown in Fig.1 for a triad. In practice, HT ratio varies considerably with the synthesis R R FeCI3 R = s4@2 Chloroform * Poly 3-hexylthiophene conditions as well as the structure of the monomer and is usually between 50-75%. The remainder of the links between the repeat units can be head-to-head (2,2), tail-to-tail(5,5) and to a lesser degree, 2,4-linkagesY as shown in Fig. 2. H-T (2,5) linkages will give maximum overlap of 7t orbitals of the adjacent thiophene A 2,4-linkage will lead to defects that destroy the conjugation between two thiophene rings. The H-H connection will force two connecting thiophene rings to be non-planar due to steric factors. Thus, it is clear that in a poly( 3-alkylthiophene) synthesized by the ferric chloride method, there will be many chain segments having different effective conjugation lengths.The result is likely to be a polymer with a larger bandgap (and hence shorter emission wavelength), a larger emission band- width and a smaller conductivity than the polymer having all H-T links or a very high HT ratio. The increase in effective conjugation length with increase in HT ratio4 also results in an increased conductivity of such a p~lymer.~,~ This increase HT-HT-HT Fig. 1 Polymer with 100% head-to-tail ratio (HT) T-T-H-T T-T-H-H R &8 L-1 1 Tail T Head H S S 2-4 linkage Scheme 1 Synthesis of polymer 1 Fig. 2 Possible regiochemical linkages in polythiophenes J. Mater. Chem., 1996, 6( ll), 1763-1766 1763 in conductivity is particularly important since higher conduc-tivity means reduced heat generation due to resistance; one of the primary modes of failure of an electroluminescent device is thermal and/or oxidative degradation.In recent worksa Holdcroft and Xu examined the effect of the head-to-tail ratio (50-80%) in poly( 3-hexylthiophene) on photoluminescence properties. They showed that both intensity and emission wavelength of polythiophenes are markedly affected by the molecular architecture in solution as well as in the solid state. However, Holdcroft and Xu dealt exclusively with the photoluminescence properties of the polymers; they did not discuss electroluminescence. In another recent publi-cation,8b Hadziioannou and co-workers reported that well-defined regioregularity in polythiophenes having primarily head-to-head linkages can be used to tune the emission wave-length in electroluminescent devices.Work by Heegersc has examined semiconductor polymer blends with HT-poly(3-hexylthiophene) prepared by the Rieke method. Holmes and Friendsd have examined HT-poly (3-dodecylthiophene) and report that the photo-and electro-luminescence were similar to that reported for regioirregular, non-HT poly( 3-dode-cylthiophene). Based on the foregoing discussion, it is possible that poly(3-alkylthiophenes) having all H-T linkages or a very high HT ratio may perform better in an electroluminescent device. It is, however, theoretically unclear whether polymer order or dis-order will enhance the efficiency of singlet excition formation. This paper does not wish to definitely address that question, yet data presented here indicates that differences in regioregul-arity in poly(3-alkylthiophenes) is significant and important in LED operation.We present here results that show improved electroluminescence performance in regioregular (greater than 98% head-to-tail coupled) poly( 3-alkylthiophene~)~"~~~.~rela-tive to the non-regioregular polymers. Experimenta1 3-Hexylthiophene was obtained from TCI and was polymerized as received by ferric chloride-mediated oxidative couplingg to give polymer 1. Polymer 2 was synthesized using the McCullough meth~d~,~$~(Scheme 2) for the synthesis of regio-regular HT-poly( 3-substituted)thiophenes.For spectral study, the samples were prepared by casting thin films of polymers from chloroform solution on quartz plates. UV-VIS spectra of the samples were recorded using a Hewlett Packard 8452A Diode Array Spectrophotometer. Photoluminescence spectra were recorded on a Spex Fluorolog 2 using a depolarized beam of light. The Mg/A1 electrode having total thickness of about 400 nm was vacuum deposited on the plates. The area of each light-emitting diode was approximately 0.7 mm2. All electrolumi-nescence measurements were carried out using a Hewlett Packard 4145B semiconductor parameter analyser at room temperature in air under DC forward bias conditions where IT0 was used as an anode and the metal electrode was used as the cathode.Results and Discussion Polymers 1 and 2 were synthesized by the methods shown in Schemes 1 and 2 respectively. Polymer 1 has an HT ratio of about 70% while HT polymer 2 has an HT ratio of about 98Yo as determined1*by proton NMR spectroscopy. Electronic absorption spectra and fluorescence spectra are shown in Figs. 3 and 4. Photophysical properties of polymers 1 and 2 are summarized in Table 1.The wavelength with strongest emission was 450 nm. The absorption and photoluminscence (PL) maxima are given in both wavelength and energy (cm-l). The higher wavelength for the absorption maximum for polymer 2 compared to polymer 1 is explained by the fact that polymer 2 has a higher HT ratio and this, in turn, results in increased coplanarity, leading to a higher degree of effective conjugation.1.01 10.30 0.25p1 0.20$ 0.15 ' 0.10 O.* 0.0' I300 400 500 600 700 800 900 Alnm Fig. 3 UV-VIS (left, solid line) and photoluminescence (right, broken line) spectra of thin film of polymer 1. Excitation wavelength 460 nm, 2.5 nm bandwidth. 1.o ' ~ 1 0. a ... . 10.12 4 0.8 :.* Electroluminescence spectra were recorded using a JY CP2000 spectrograph equipped with a Prism Research photo-diode array. The LED devices were fabricated as follows. Square plates of glass (2 x 2 inches) having rectangular pads of indium-tin oxide (ITO) were prepared. A thin film of 0.10 0.08 0, 0.06$ p,ZT polymer was coated on the IT0 pads by spin-coating the 1% solution of the polymers in toluene at 1500 and 2000rpm.Average thicknesses of the polymer films, as measured by Tencor A200 profilometer, were about 50 nm. .R .R .R LDA-'THF MgBr2,Et20 LI Br BrMg Br 2 Scheme 2 Synthesis of polymer 2 :,...a Fig. 4 UV-VIS (left, solid line) and photoluminescence (right, broken line) spectra of thin film of polymer 2. Excitation wavelength 450 nm, 2.5 nm bandwidth. Table 1 Photophysical properties of polymers 1 and 2 polymer An+b~)l cm-(nm) Lax$PL)/ cm-(nm) stokes shift/ cm-' 1 21,930 (456) 15,380 (650) 6550 2 19,610 (510) 13,950(717) 5660 1764 J. Mater. Chem., 1996, 6(11), 1763-1766 The positions of the peak PL emission wavelengths for 1 and 2 are also consistent with the extent of planarity of the chains in two polymers.Interestingly, the Stokes shift for polymer 1 is 6550 cm-' while for 2 it is 5660 cm-l. Polymer 1, i.e. the polymer having a lower HT ratio, shows higher Stokes shift than polymer 2. This observation is partly attributed to the less rigid nature of the polymer backbone in polymer 1. Consistent with this explanation are the observations that the photoluminescence and absorption spectra of polymer 1 are broader than that of polymer 2. The decrease in the Stokes shift with increase in HT ratio is also explained'" by invoking greater relief of conformational strain in the excited state for polymers such as 1. Electroluminescence (EL) spectra (Fig. 5) showed that emis- sion from the diodes constructed from polymer 2 was red- shifted about 30-35 nm as compared to the emission from the diodes constructed from polymer 1.Comparison of PL and EL spectra show that the PL spectra are red-shifted. In general, the PL and EL spectra are expected to be identical. However, the shift between the PL and EL spectra is not unusual. Similar shifts have been observed in the case of poly(p-phenylene)," poly (3-cy~lohexylthiophene)~~and poly (p-phenylenevinyl-ene).13 While no definitive explanation has been given, several plausible theories have been put forward. Inganas and co- workers suggested12 that different regions of the samples having different ground state and/or excited state conformation are excited in PL and EL experiments. Heeger and co-workers suggested13 that during the photoluminescence process the regions of the sample having low mobility for charge carriers are predominantly excited while in EL experiments reflect more upon the regions of sample having high mobility.We also feel that the blue shift in the EL spectrum may also be related to the heat generation in the electroluminescent devices. It is well known that both photoluminescent and electronic absorption spectra of poly(3-alkylthiophene) show thermo- chromic ~r0perties.l~ It is also interesting to note that the EL 3bO 4d0 5b0 660 760 860 I' I I 1 I I 300 400 500 600 700 800 Alnm Fig. 5 Electroluminescence spectra of polymers 1 and 2 Table 2 Electroluminescent properties of polymers 1and 2 external external efficiency" efficiencyb turn-on A,,,( EL)/ polymer ("/.I ("/.I voltage/V" nm 1 7 x 10-5 1.3 x 10-5 5.0 630 2 1.5 x 10-4 3.85 x 10-4 2.8 662 "At 6 mA current through the device.At 25 mA current through the device. At 200 pA current through the device. spectrum of polymer 2 is noticeably narrower than that of polymer 1. Diode characteristics for these polymers, including the exter- nal efficiencies, are given in Table 2. The external efficiency of the device is defined as the ratio of amount of light emitted by the device to the electrical current passed through the device. As judged from the data in Table 1, the device fabricated from polymer 2 is about an order of magnitude more efficient than the device fabricated from polymer 1 at a low current value of 6 mA.Upon passing a higher level of current (25 mA) through the diodes, there was a notable increase in the external efficiency of the device having polymer 2. When these diodes were operated repeatedly at 25 mA, it was observed that the diodes having polymer 1 degraded much more rapidly than the diodes containing polymer 2. Turn-on voltage, defined in this work as the voltage required to pass 200 pA of current through the device, is also found to be lower in case of polymer 2. Lower turn-on voltage for polymer 2 reflects higher conduc- tivity of this polymer. In conclusion, we have shown that poly( 3-hexylthiophene) synthesized by a method giving a polymer having a higher H/T ratio shows better electroluminescent properties and therefore diodes made from polymer 2 represent a significant improvement over diodes made from materials such as 1.We have also carried out similar experiments of poly( 3-dodecylthi- ophene) synthesized by two different methods. These results will be reported in a forthcoming publication. We thank Jeffrey Bisberg and William Robinson, both of the Polaroid Corporation, for help in electroluminescence measurements during the early stages of the project and for depositing the metal electrodes, respectively. This work was supported by the NSF CHE-9201198 (R.D.M.) and by a grant from Polaroid (R.D.M.). References 1 (a) Mini-symposium on Polymeric Light-Emitting Diodes, Sept. 15-17, 1993, Eindhoven, The Netherlands; (b) J. H.Burroughes, D. D. C. Bradley, A. R. Brown, R. N. Marks, K. Mackay, R. H. Friend, P. L. Burns and A. B. Holmes, Nature, 1990,347,539; (c) D. Braun and A. J. Heeger, Appl. Phys. Lett., 1991,58, 1982. 2 (a) G. Grem, G. Leditzky, B. Ullrich and G. Leising, Ado. Muter., 1992, 4, 36; (b) Y. Ohmori, M. Uchida, K. Muro and IS.Yoshino, Jpn. J. Appl. Phys., 1991,30, L1941. 3 (a) D. Braun, D. Gustafsson, D. McBranch and A. J., Heeger, J. Appl. Phys., 1992, 72, 564; (b) Y. Ohmori, M. Uchida, K. Muro and K. Yoshino, Solid State Commun., 1991,80,605; (c)Y. Ohmori, M. Uchida, K. Muro and K. Yoshino, Jpn. J. Appl. Phys., 1991,30, L1938; (d) Y. Ohmori, C. Morishima, M. Uchida and K. Yoshino, Jpn. J. Appl. Phys., 1992,31, L568.4 (a) R. D. McCullough and R. D. Lowe, J. Chem. SOC., Chem. Commun., 1992, 70; (b) R. D. McCullough, R. D. Lowe, M. Jayaraman and D. L. Anderson, J. Org. Chem., 1993,58,904. 5 T.-A. Chen and R. D. Rieke, J. Am. Chem. SOC.,1992,114,10087. 6 R. D. McCullough, S. Tristram-Nagle, S. P. Williams, R. D. Lowe and M. Jayaraman, J. Am. Chem. SOC.,1993,115,4910. 7 R. D. McCullough and S. P. Williams, J. Am. Chem. SOC.,1993, 115,11608. 8 (a) B. Xil and S. Holdcroft, Macromolecules, 1993, 26, 4457; (b) R. E. Gill, G. C. Malliaras, J. Wildeman and G. Hadziioannou, Adv. Muter., 1994,6, 132. J. Mater. Chem., 1996, 6( ll), 1763-1766 1765 9 M. Leclerc, F. M. Diaz and G. Wegner, Makromol. Chem., 1989, 190,3105. 10 S. Hotta, M. Soga and N. Sonoda, Synth. Met., 1988, 26,267. 11 G. Grem, G. Leditzky, B. Ullrich and G. Leising, Adv. Muter., 1992, 4, 36. 12 M. Berggren, G. Gustafson, 0. Inganas, M. R. Anderson, 0.Wennerstrom and T. Hjerberg, Adv. Muter., 1994,6,488. 13 C. Zhang, D. Braun and A. J. Heeger, J. Appl. Phys., 1993,73,5177. 14 (a) K. Yoshino, Y. Manda, K. Sawada, M. Onoda and R.-I. Sugmoto, Solid State Commun., 1989,69, 143; (b)C. Roux, K. Faid and M. Leclerc, Makromol. Chem., Rapid Commun., 1993,14,461; (c)C. ROUX,J. Y. Bergeron and M. Leclerc, Makromol. Chem., 1993, 194,869;(d)W. R. Salaneck, 0.Inganas, B. Themans, J. 0.Nillson, B. Sjogren, J.-E. Osterholm, J.-L. Bredas and S. Svensson,J. Chem. Phys., 1988,89,4613. Paper 6/01547F; Received 5th March 1996 1766 J. Mater. Chew., 1996,6( ll), 1763-1766

ISSN:0959-9428    

DOI:10.1039/JM9960601763

出版商:RSC    

年代:1996

数据来源: RSC

摘要:

A water-resistant precursor in a wet process for Ti02 thin film formation Mitsunobu Sate,*" Hiroki Hara,' Toshikazu Nishideb and Yutaka Sawadac "Research Institute for Science and Technology, Kogakuin University, Nakano, Hachioji City, Tokyo 192, Japan bResearch Center, Nissan Motor Co., Ltd., Natsushima, Yokosuka City, Kanagawa 237, Japan 'Department of Industrial Chemistry, Faculty of Engineering, Tokyo Institute of Polytechnics, Iiyama, Atsugi City, Kanagawa 243-02, Japan Deposition of anatase TiO, thin films on soda-lime glass has been achieved by firing an adhered precursor titanium@) complex of ethylenediamine-N,N,N',"-tetraacetic acid (H,edta) between 450 and 550 "C in air. The coating solution was prepared by the reaction of a neutral [Ti( H,O)(edta)] complex, obtained from TiCl, and H4edta in an air-oxidation process, with dipropylamine in ethanol.The crystal structures of the oxide films on glass substrates were examined by XRD, and some of their optical properties and electronic structure were investigated by XPS. It was shown that TiOz film formation was attainable employing a water-resistant precursor derived from a stable metal complex. The thermal properties of the initial [Ti(H,O)(edta)] complex and the facile preparation of the precursor ethanol solution are also reported. Titanium(1v) dioxide thin films are important materials for a wide variety of applications based on their optical, electrical and photoelectrochemical properties, etc.'Y2 The deposition of anatase thin films, one of three crystalline forms of titania, has been studied recently by the spray inductively coupled plasma techniq~e,~low-pressure CVD4 and atomic layer dep~sition.~ Anatase film formation by anodic oxidative hydrolysis6 of TiC13 and by an organic self-assembled monolayer method7 from aqueous solution has also been performed.The conventional sol-gel process employing alkoxides as precursors is still of importance in the formation of anatase thin film~.~-'~ In this process, the rigorous exclusion of water from the system is essential for the synthesis and conservation of the precursor alkoxides and their solutions, since the process is based on partial or complete hydrolysis of such metal a1k0xides.l~ From this point of view, a novel wet process, in which films can be formed by facile coating procedures of precursor solutions of a water-resistant precursor, may afford practical advantages.In recent work, we found that an ethanol solution of a monomeric tributylammonium cobalt (m)-edta complex was a useful precursor in a wet process for the formation of Co,O, thin films on glass s~bstrates.'~ We report here the formation of anatase thin films on soda-lime glass by firing an adhered film consisting of a titanium(1v) complex salt. The precursor solution was prepared by reaction of a neutral [Ti( H,O)(edta)] complex with dipropylamine in ethanol. The starting complex and the TiO, thin films formed by firing in air at various temperatures up to 550°C were studied by several spectroscopic methods.The influence of added water to the precursor solution was also examined. From these results, it was elucidated that such ethanol solutions are not only suitable for the formation of precursor films adherable to substrates, but also robust to humid atmospheric conditions. Experimental Materials Ethylenediamine-N,N,N',N'-tetraacetic acid and a hydro-chloric acid solution of titanium(II1) chloride (min. 20 mass%) of the highest commercially available grade were purchased from Kanto Chemicals Co., Inc. Ethanol was purchased from Amakasu Chemical Industries Co., Ltd., and was of GR grade (99.5%) and dried upon molecular sieves 4A (Wako Pure Chemicals Co., Ltd.), before use. Other materials were used without further purification.Preparation of [Ti( H,O) (edta)] To an aqueous solution (250 ml) of 29.26 g (0.1 mol) of H4edta, was added 54.0 g of a 20% HC1 solution of TiC1, (67 mmol) with stirring at 90°C in an open beaker. The violet solution obtained was stirred at 90 "C for 30 min. During the reaction, white crystals gradually formed in the violet solution. The solution was then continuously stirred at 50°C for 3 h. The white crystals which formed were collected, washed with a small amount of water and ethanol, and air-dried. Yield: 23.5 g (81%). Elemental analysis indicated high purity (calc. for [Ti(H20)(CloH12N208)] (C10H14N209Ti,M, =354.1 1): C, 33.92; H, 3.98; N, 7.91. Found: C, 33.81; H, 4.03; N, 7.95%). Preparation of an ethanol solution for coating An ethanol solution containing 12 mass% precursor was prepared by reaction of [Ti( H,O)(edta)] with dipropylamine in ethanol.[Ti(H,O)(edta)] [1.00 g, 2.83 mmol] was sus-pended in 15 ml of dried ethanol and 0.62 g (6.16 mmol) of dipropylamine was added slowly to the suspended solution in a flask, and the mixed solution was heated at reflux for 10 min. The yellow transparent solution obtained was used to adhere the precursor onto soda-lime glass. An ethanol solution of 20 mass% was also prepared by the same procedure. Each log aliquot of the solution was then adjusted to 20 g by the addition of measured amounts of water and ethanol. Thus 5 and 10 mass% water solutions, with precursor contents of lo%, were prepared, and applied onto soda-lime glass after storing at ambient temperature for 10 days.Note, the transparency of these solutions was retained even after several months and neither the colour nor the viscosity of the solutions changed. A conventional flow-coat procedure was employed for the coating of these precursor solutions onto the substrate in order to prepare the samples for firing. A spinsoating method was also practicable. In both cases, transparent yellow films were obtained after drying. The samples were fired in air at various temperatures for 30min during which the films became colourless, and showed strong adherence to the glass. The film thickness after firing was controlled to be ca. 100nm for X-ray photoelectron J. Mater. Chew., 1996,6(ll), 1767-1770 1767 spectroscopic (XPS) and ellipsometric measurements or ca.200-300 nm for X-ray diffraction (XRD) measurements. Measurements IR spectra were measured on a Perkin Elmer FTIR 1600 spectrophotometer for sample tablets diluted with KBr. Thermogravimetry (TG) and differential thermal analyses (DTA) were performed using a Rigaku TAS-200 instrument. The crystal structures of the oxide films were determined by XRD performed with a Rigaku RINT-2500V model (50 kV, 300 mA) using graphite-monochromated Cu-Ka radiation and a parallel beam optic system (incident angle: 0.1"). The XP spectrum was recorded on a Perkin Elmer PHI model 5600 spectrometer, with Mg-Ka radiation (1253.6 eV) operated at 10 kV and 200 W used as the X-ray excitation source.The C 1s binding energy (284.6 eV) of the trace amount of hydro-carbon originally present in the air was used to calibrate the binding energy. Under these conditions the Ag 3d,,, peak (367.9 eV) had a FWHM value of 0.96 eV. The operating pressure was within (2-8) x Pa. Transmittance spectra were measured with a Hitachi U4000 spectrophotometer. The refractive indices were measured with a Mizojiri DHA-OLX ellipsometer employing a He-Ne laser source of 632.8 nm. The film thicknesses were measured using a Dektak 3030 stylus profilometer. Results and Discussion The complex [Ti( H,O)(edta)] was isolated and the pentagonal-bipyramidal structure around the Ti" ion was crystallograph-ically determined by Fackler, Jr.et a1.16 They isolated the complex by a similar procedure to that used by Weighardt et al., by employing H,edta in place of nitrilotriacetic acid (H,nta) or iminodiacetic acid (H,ida) as the polyaminopolycar-boxylic acid ligand.17 In their method, Ti"' was first oxidized to TiIV by the addition of nitric acid to the starting TiCl, solution prior to reaction with H,edta. In contrast, we synthe-sized the same compound in good yield by a milder air-oxidation process of a solution of Ti"' and H,edta. The IR spectrum of the complex (Fig. 1) is almost identical to that reported by Sawyer and McKinnie." TG-DTA measurements were carried out in order to study the decomposition of [Ti(H,O)(edta)]. Several exothermic peaks were observed by DTA, shown along with the TG curve in Fig.2. The small peak at 240°C in DTA is attributable to the elimination of the coordinated water molecule, as verified by the corresponding decrease of mass in the TG curve. The broad exothermic peak and the accompanying decrease of mass between ca. 300 and 500°C may be due to the sequential decomposition of the complex and oxidation of the organic ligand. The TG-DTA \F cn -0 cn3E 100 200 300 500 T/"C Fig. 2 TG-DTA curves of the complex [Ti(H,O)(edta)] results show that complete decomposition of the complex and organic substrates occurs at 515 "C. Since [Ti(H,O)(edta)] alone shows low solubility in organic solvents we examined the reaction of [Ti(H,O)(edta)] with dipropylamine in ethanol, guided by the good solubility of the alkylammonium salt of an anionic cobalt (m)-edta complex which was prepared as the precursor for Co,O, thin films on glass substrates.15 The reaction of two mole equivalents of dipropylamine with colourless [Ti( H,O)(edta)] gave a trans-parent yellow solution indicating that the precursor complex formed in solution is different from the starting aqua complex, although its structure is as yet unknown. Adhered films obtained from the precursor complex were fired between 400 and 550°C and in all cases transparent films were obtained.In order to identify the transparent films, XRD spectra of the thin films after firing were examined (Fig. 3). In each spectrum, the broad band observed in the range 28= 15-35" may be attributed to the film and the substrate.It is seen that firing at 400 "C leads to an amorphous film, since no peak due to a crystalline compound was observed [Fig. 3(a)]. In contrast, characteristic peaks of anatase were found in the spectra of the samples obtained above 450°C, as shown in Fig. 3(b)-(d). The peaks at 25.4, 37.9, 48.2, 54.2, 55.1, 62.9 and 75.2" are assigned to the (101), (004), (200),(105), (211),(204) and (215) reflections of anatase,19 respectively. The strongest, at 28 =25.4", is characteristic of anatase ( 101), as previously rep~rted.'~?~~Since no peaks due to other compounds were observed, it is evident that the fired films are anatase, although that obtained at 400°C is amorphous. Thus, it is clear that the reaction of [Ti(H,O)(edta)] with dipropylamine in ethanol leads to an excellent precursor solution for the preparation of anatase thin films.Use of the aqua complex alone was not feasible owing to its low solubility in a variety of solvents. 20 40 3000 ' 2000 1500 I000 500 28/degrees wavenumber/cm-' Fig.3 YRD patterns of the thin films formed on soda-lime glass at Fig. 1 IR spectrum of the complex [Ti(H,O)(edta)] various temperatures: (a) 400, (b) 450, (c) 520 and (d) 550 "C 1768 J. Muter. Chern., 1996, 6(ll), 1767-1770 The relationship between the refractive indices and the firing temperatures of the thin films is shown in Fig. 4. The absolute values of the refractive indices and their tendency to increase with increasing firing temperature mirror the results for films prepared by a sol-gel method reported by Nishide and Mizukami,13 as shown in Fig.4. The transmittance spectra of the thin films prepared by either method are also very similar. An XPS measurement was carried out on the thin film fired at 500"C, in order to examine the electronic structure of the TiO, thin film on the glass substrate. The binding energies measured for Ti 2p3,, and Ti 2pl,, are 458.7 and 464.2eV, respectively. These are comparable with the values, 458.8 eV for Ti 2p3,, and 464.3 eV for Ti 2p1,,, reported for TiO, in the literature.21 Significantly, the difference, 5.5 eV, between the binding energies for Ti 2p3,, and Ti 2p1,, is almost identical to the reported value of 5.54 eV.Thus, the observations on the electronic structure of the thin film support the formation of TiO, involving the Ti4+ ion. In order to examine the optical properties of the trans- parent thin films obtained by firing the adhered films which were coated from the precursor solutions, ellipsometric measurements were made; results for precursors B and C, containing 5 and 10% water as well as for the water-free precursor A are given in Fig. 4. The refractive indices at each firing temperature for samples B and C differed slightly from that obtained from the water-free precursor (precursor A) mentioned above. These results suggest that the optical properties of the anatase thin films can be controlled by the addition of water to the precursor solution.On the basis of these results, the present procedure is useful for the prep- aration of TiO, thin films on glass substrates. Note that optical properties, which can be affected by the co-existence of organic substrates, i.e. the ligand and the amine, of the films are comparable with those obtained from the conven- tional sol-gel procedure, where a water-free precursor solution was employed. Furthermore, samples containing up to 10% water can be used as precursor solutions, the presence of water only slightly affecting the refractive indices of the thin films formed. A schematic representation of the novel pro- cedure for the preparation of the thin films is shown in Fig. 5. Diethylamine was used instead of dipropylamine, in order to investigate the formation reaction of the adherable precur- sor.We found that the aqua complex also reacts with this amine in ethanol to form a similar yellow solution which can also adhere to a glass substrate. This result suggests that the chemical species involving TiIV ions in both solutions are comparable. From the viewpoint of the coordination chemistry 2*20 r .2.10 Xa m .-C ' .-92.00 CI .c3 p! 1.90 A I I I1.80 I I 350 400 450 500 550 600 firing temperature/% Fig. 4 Relationship between the reflactive index and the firing tempera- ture for the TiO, thin films. 0,water-free precursor (A); 0,5% water content (B); A 10% water content (C). The corresponding relationship in the sol-gel method, indicated by U, is from ref. 13.Ti /HCI Solution1 edLinH20 [Ti(HzO)(edta)] Precursory Solution Coating FiringI Ti02 Thin Film Fig. 5 Schematic representation of the novel wet process of the TiIv-edta complex, it is important to clarify the structure of the ethanol-soluble form obtained in the presence of amine. Investigations of the effects of the N-alkyl groups of the amines on the properties and the isolation of the adhered precursor complexes obtained from the reaction of the neutral [Ti(H,O)(edta)] complex are now in progress, along with the application of the novel procedure to form mixed-metal oxide systems involving Ti'V. Conclusion It has been shown that an ethanol solution obtained by the reaction of neutral [Ti(H,O)(edta)] and 2 mole equivalents of dipropylamine is useful as a novel precursor for TiO, thin films.By firing the adhered precursor in the temperature range 450-550°C, anatase thin films were obtained, while firing at 400°C gave an amorphous film. The optical properties of the thin films are comparable with those prepared by the conven- tional sol-gel process. It was furthermore elucidated that the adherent property of the precursor ethanol solution is unaffec- ted by the addition of water. This, to our knowledge, is the first report on TiO, film formation employing a facile coating method with a water-resisting precursor derived from a stable metal complex. This work was partially supported by a grant-in-aid for scientific research from the Ministry of Education, Science and Culture of Japan (No.03650687) and a special grant from our institute. References C. J. Brinker and G. W. Scherer, Sol-Gel Science, Academic Press, San Diego, 1990. D. Segal, Chemical Synthesis of Advanced Ceramic Materials, Cambridge University Press, Cambridge, 1989. Y. Mizoguchi, M. Kagawa, Y. Shono and T. Hirai, J. Mater. Sci. Lett., 1993, 12, 1854. T. Go, N. Hara, and K. Sugimoto, Nippon Kinzoku Gakkaishi, 1994,58,448. J. Arik, A. Aidla, T. Uustare and V. Sammelselg, J. Cryst. Growth, 1995,148,268. L. Kavan, B. O'Regan, A. Kay and M. Graetzel, J. Electroanal. Chem., 1993,346,291. H. Shin, R. J. Collins, M. R. De Guire, A. H. Heuer and C. N. Sukenik, J. Mater.Rex, 1995,10,692. K. Kato, A. Tsuzuki, Y. Torii, H. Taoda, T. Kato and Y. Butsugan, Nagoya Kogyo Gijutsu Kenkyusho Hokoku, 1994,42,346. T. Hashimoto, T. Yoko and S. Sakka, Bull. Chem. Soc. Jpn., 1994, 67, 653. Y. Hamasaki, S. Ohkubo, K. Murakami, H. Sei and G. Nogami, J. Electrochem. SOC., 1994, 141, 660. J. Mateu. Chern., 1996, 6(11), 1767-1770 1769 11 12 B. E. Yoldas, Appl. Opt., 1980,21,2960. Y. Takahashi and Y. Matsuoka, J. Muter. Sci., 1988,23,2259. 17 K. Weighardt, U. Quilitzsch, J. Weiss and B. Nuber, Inorg. Chem., 1980,19,2514. 13 14 15 16 T. Nishide and F. Mizukami, J. Ceram. SOC.Jpn., 1992,100,1122. Inorganic Materials, ed. D. W. Bruce and D. O’Hare, John Wiley & Sons, West Sussex, 1993, pp. 519-525. M. Sato, H. Hara, H. Kuritani and T. Nishide, Sol. Energy Muter., in press. J. P. Fackler, Jr., F. J. Kristine, A. M. Mazany, T. J. Moyer and R. E. Shepherd, Inorg. Chem., 1985,24,1857. 18 19 20 D. T. Sawyer and J. M. McKinnie, J. Am. Chem. SOC., 1960, 82, 4191. Joint Committee of Powder Diffraction, JCPDS card no. 21 1272. Handbook of X-ray Photoelectron Spectroscopy, ed. J. Chastain, Perkin-Elmer, Minnesota, 1992,p. 72. Paper 6103125K; Received 3rd May, 1996 1770 J. Muter. Chem., 1996,6( ll), 1767-1770

ISSN:0959-9428    

DOI:10.1039/JM9960601767

出版商:RSC    

年代:1996

数据来源: RSC