ISSN:0306-0012    

DOI:10.1039/CS99120FX009

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

ISSN 0306-0012 CSRVBR 20(3)271-404 (1991) Chemical Society Reviews Vol20 No 3 1991 Page Averrnectins and Milbemycins Part I1 By H. Geoffrey Davies and Richard H. Green 27 1 Spin Trapping of Inorganic Radicals By Detlef Rehorek 34 1 Recent Progress on Conducting Organic Charge-Transfer Salts By Martin R. Bryce 355 Poly(pyrro1e) as a Support for Electrocatalytic Materials By Dominic Curran, James Grimshaw, and Sarath D. Perera 39I The Royal Society of Chemistry Cambridge Chemical Society Reviews EDITORIAL BOARD Professor H. W. Kroto FRS (University of Sussex) (Chairman) Professor M. J. Blandamer (University of Leicester) Dr. A. R. Butler (University of St. Andrews) Professor B. T. Golding (University of Newcastle upon Tyne) Professor M.Green (University of Bath) Professor J. A. McCleverty (University of Bristol) Professor J. F. Stoddart (University of Birmingham) Consulting Editors Dr. G. G. Baht-Kurti (University of Bristol) Professor S. A. Benner (Swiss Federal Institute of Technology, Zurich) Dr. J. M. Brown (University of Oxford) Dr. J. Burgess (University of Leicester) Dr. N. Cape (Institute of Terrestrial Ecology, Lothian) Professor A. Hamnett (University of Newcastle upon Tyne) Dr. T. M. Herrington (University of Reading) Dr. R. Hillman (University of Bristol) Professor R. Keese (University of Bern) Dr. T. H. Lilley (University of Sheffield) Dr. H. Maskill (University of Newcastle upon Tyne) Professor Dr. A. de Meijere (University of Gottingen) Professor J.N. Miller (Loughborough University of Technology) Dr. D. M. P. Mingos (University of Oxford) Professor S. M. Roberts (University of Exeter) Professor B. H. Robinson (University of East Anglia) Dr. A. J. Stace (University of Sussex) Staff Editors Mr. K. J. Wilkinson (Royal Society of Chemistry, Cambridge) Dr. J. A. Rhodes (Royal Society of Chemistry, Cambridge) 0The Royal Society of Chemistry, 1991 All Rights Reserved No part of this publication may be reproduced or transmitted in any form or by any means -graphic, electronic, including photocopying, recording, taping, or injormation storage and retrieval systems -without written permission from The Royal Society of Chemistry Published by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF Printed in England by Clays Ltd, St Ives plc NOTICE TO READERS OF CHEMICAL SOCIETY REVIEWS Beginning in 1992 this well-established RSC review journal will concentrate upon shorter articles (usually in the range 6-8000 words) and eight of these articles will be published in each quarterly issue.The journal will appear in a new format -A4, double column, with distinctive illustrated covers. Colour will be used where intrinsic to the subject matter. Turget Audience -chemists and students of chemistry together with scientists in allied disciplines. Approach -The journal will have the broad appeal necessary for researchers to benefit from an awareness of advances in areas outside their own specialities.Deliberate efforts will be made to solicit authors and articles from Europe and further afield, in order to present a truly international outlook on the major advances in a wide range of chemical areas. It will be of interest and help to students planning a career in research. In particular, it will be the place to find succinct and authoritative overviews of timely topics in modern chemistry. In line with the above, review articles will not be comprehensive, too detailed, or heavily referenced; they should act as a springboard to further reading. Interdisciplinary awareness should thereby be heightened and the student should be stimulated to take a more professional interest in the topic of the review.We see this new approach and format for Chemical Society Reviews as an exciting development in RSC publishing activities. Although the majority of articles are intended to be specially commissioned, the Society is always prepared to consider offers of articles for publication. In such cases a short synopsis, rather than the completed article, should be submitted to the Senior Editor (Reviews), Books and Reviews Department, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF. Members of the Royal Society of Chemistry may subscribe to Chemical Society Reviews at E25.00 per annum; they should place their orders on the Annual Subscription renewal forms in the usual way. All other orders accompanied with payment should be sent directly to The Royal Society of Chemistry, The Distribution Centre, Blackhorse Road, Letchworth, Herts. SG6 1HN England. 1992 annual subscription rate E.C. E72.00, Overseas E81.00, U.S.A. $158.00. Air freight and mailing in the U.S.A. by Publications Expediting Inc., 200 Meacham Avenue, Elmont, New York 11003. U.S.A. Postmaster: Send address changes to Chemical Society Reviews, Publications Expediting Inc., 200 Meacham Avenue, Elmont, New York 11003. Second class postage is paid at Jamaica, New York 11431. All other despatches outside the U.K. by Bulk Airmail within Europe, Accelerated Surface Post outside Europe.

ISSN:0306-0012    

DOI:10.1039/CS99120FP011

出版商:RSC    

年代:1991

数据来源: RSC

ISSN:0306-0012    

DOI:10.1039/CS99120BP015

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

Chem. SOC.Rev., 1991,20,271-339 Avermectins and Milbemycins Part I1 * By H. Geoffrey Davies and Richard H. Green MEDICINAL CHEMISTRY DEPARTMENT, GLAXO GROUP RESEARCH, G R E EN FOR D, M I D D LE S EX, UB6 OHE 7 Total Synthesis A. Mi1bernycins.-The first reported synthesis of one of this large family of macrolides was that of Smith and colleagues,' 32*1who described the preparation of racemic milbemycin p3. Shortly after this synthesis was published, Williams and co-workers in Indiana '33*1described their synthesis of optically pure milbemycin B3, which was obtained by the condensation of two asymmetric moieties both derived from the same chiral precursor, ( -)-(3s)-citronellol. In addition to the pioneering work of Smith and Williams, three other groups have reported total syntheses of milbemycin p3,one group the synthesis of milbemycin Bt and one further group the synthesis of milbemycin E.In each case optically active material was obtained. Chronologically, the first of the three milbemycin p3 syntheses was that described by Baker and CO-WO~~~~S.~~~~~*~~~-~~~Th e chiral spiroacetal diol (279) 68 was transformed into the aldehyde (280), and this was immediately treated with ethoxycarbonylethylidene triphenylphosphorane to give the a,P-unsaturated ester (281) (Scheme 36). This was shown by both 'H and 13C NMR spectroscopy to be a single component possessing, exclusively, the E-stereochemistry about the double bond. Reduction of the ester followed by mesylation and iodination gave an unstable iodide which was immediately used to alkylate the oxazolidinone (282), resulting in the introduction of the C-12 methyl group of the milbemycin skeleton. Reductive removal of the oxazolidinone ring from (283), followed by tosylation, nucleophilic substitution, and oxidation, afforded the C-11 to C-25 fragment of the milbemycins (284).In the original report of their total synthesis of milbemycin p3,t34 Baker and colleagues had used an adaptation of Williams' work 133,1to introduce the C-14-C-15 double bond. However, an alternative route to this tri-substituted double bond was later established,'36 and it was this route which was utilized in the full paper on the total synthesis in question.' 35This new approach involved treatment of (284) * Avermectins and Milbemycins Part I (Fermentation and Isolation; Structure Determination; Biosynthesis; Metabolism and Assay; Partial Synthesis) appeared in Chemical Soriel) Reuirws, 1991, Vol.20, Issue 2 (June). I" A. B. Smith, S. R. Schow, J. D. Bloom, A. S. Thompson, and K. N. Winzenberg, J. Am. Chrm. SOL.., 1982, 104, 4015; S. R. Schow, J. D. Bloom, A. S. Thompson, K. N. Winzenberg, and A. B. Smith, J. Am. Chem. Soc., 1986,108,2662. L33D. R. Williams, B. A. Barner, K. Nishitani, and J. G. Phillips, J. Am. Chem. Soc., 1982, 104.4708. 134 R. Baker, M. J. O'Mahony, and C. J. Swain, J. Chem. Soc., Chem. Commun., 1985, 1326. 13' R. Baker, M. J. O'Mahony, and C. J. Swain. J. Chem. SOL..,Perkin Trons. 1, 1987, 1623. 136R. Baker, M. J. O'Mahony.and C. J. Swain, Tetrahedron Lett., 1986, 27, 3059. 271 Avernzectins and Milbemycins Part II HO*=-0 ---0OHC+ OH OTBDPS (279) (280) Milbemycin ...-(28) S0,Ph OTBDPS (284) Reagents: i, Dibal-H, THF, -78 "C; ii, MeSOzCI, pyr, DMAP; iii, NaI, THF; iv, LDA, THF, (282), -78 "C, then add (281);v, LiAlH4, EtzO; vi, p-TsC1, pyr; vii, Na, PhSH, MeOH; viii, KHS05, EtOH, HzO; ix, Bu'Li, THF, -78 "C, then (285); x, PhCOCl; XI, Na/Hg,MeOH, THF; xii, ButNF, THF xiii, KH, THF; xiv, EtSH, Na, DMF Scheme 36 with t-butyl-lithium followed by addition of the E-aldehyde (285), readily available from 4-methoxy-3-methylbenzoicacid. Quenching of the reaction mixture with benzoyl chloride afforded a mixture of diastereoisomeric benzoates which were not separated but treated directly with sodium amalgam.This served to generate the required trans-butadienyl system in nearly 80% yield with complete stereocontrol in the formation of the diene system. Protecting group removal followed by macrolactonization and demethoxylation then provided (+)-milbemycin p3 (28). Two different approaches to milbemycin 03 have been described by Kocienski and colleagues. In their earlier 37 the hydroxy spiroacetal (286), prepared by a Lewis acid catalysed intramolecular aldol reaction,' 38 was transformed using routine chemistry into the aldehyde (287) (Scheme 37) in an overall yield of ca 41%. Chain extension of this aldehyde and construction of the I3'S. D. A. Street, C. Yeates, P. Kocienski, and S.F. Campbell, J. Chem. SOC.,Chem. Commun., 1985, 1386. 13' P. Kocienski and S. D. A. Street, J. Chem. SOC.,Chem. Cnrnrnun., 1984, 1381. HO* 5stvsOHC*%o -i-iii Davies and Green ___c ..-I 0 OTBDMS OCH2Ph OTBDMS (289) iwii1 Milbernycin p3 3steps-(281 Reagents: i, PhS02CH(Li)CH3, THF, -78°C followed by Ac20; ii, NaOH, dioxane; ill, Fe(acac)~, (288); iv, Na, NH,; v, Cr03-2pyr, CH2C12; vi, LDA, THF, -78"C, (290); vii, Ac2O; viii, Na/Hg, THF, MeOH, -20 "C;ix, MeS02CI, Et3N, CH2C12, -10 "C Scheme 37 C-144-15 double bond was then achieved using conditions first employed by Julia and Condensation of the aldehyde (287) with the lithium salt of phenylethyl sulphone, followed by acetylation, gave a mixture of four diastereoisomeric P-acetoxy sulphones. These were not separated but simply subjected to basic hydrolysis to give exclusively an (E)-vinylsulphone.Reaction of this with the Grignard reagent (288) in the presence of a catalytic amount of tris(acetonylacetonato)iron(m) then provided the required olefin (289) with '39 M. Julia, M. Launay, J.-P. Stacino, and J.-N. Verpaux, Tetrahedron Lett., 1982, 24, 2465; J.-L. Fabre, M. Julia, and J.-N. Verpaux, ibid., p. 2469. Averniectrnsand Milbemvcms Part II >95% retention of double bond stereochemistry Although this process suffered the disadvantage of a low overall yield (26%), mainly because of the capricious nature of the Grignard reaction, this was offset by the high stereoselectivity observed in the formation of the double bond Further compensation for the low yield was the fact that any unreacted vinyl sulphone could be recovered and recycled Although conversion of this primary alcohol (289) into the correspond- ing aldehyde, earlier prepared in racemic form by Smith and colleagues,' 32 proceeded without event, final conversion of the aldehyde into milbemycin p3 proved problematic Coupling of the phenylsulphone (290) with the aldehyde proceeded smoothly to afford the P-acetoxysulphone (291) as a mixture of diastereoisomers However, reductive elimination of this material to give the required (E,E)-diene did not proceed smoothly, the required compound only being obtained in low yield There were three main reasons for this disappointing result First, the reaction was very sluggish and thus base-catalysed side reactions were a problem In addition the trans-stereoselectivity of the process was low, and simple reductive desulphonation and ester methanolysis predominated, resulting in the diastereomeric alcohols (292) being the major products of the reaction However, in spite of these limitations, the required diene (294) could be isolated in 39% yield as a 5 1 mixture of E 2 isomers Although the alcohols (292) could be dehydrated via the mesylate (293) to give the diene (294) using DBN, stereoselectivity was very poor (E 2 = 3 2) Interestingly, simply by interconversion of the benzoyl and sulphonyl groups, Baker 134 was able to demonstrate a remarkably facile reduction to occur with none of the inherent problems described above Conversion of the diene (294) into milbemycin p3 (28) was then achieved employing Smith's conditions giving the required macrocycle in 60% overall yield Kocienski and his associates 85 have also reported an alternative, more practical, synthesis of the aldehyde (300) and its transformation (as detailed above) into milbemycin 83 The two chiral synthons (107) and (295), readily available from (S)-(-)-malic acid and (I?)-( -)-methyl 3-hydroxy-2-methylpropionate respectively, were transformed, in ten steps, into the epoxide (296) This was then treated with the mixed organocuprate (297) (Scheme 38), resulting in nucleophilic scission of the epoxide to give the dihydropyran derivative (298) This highly sensitive intermediate was not purified but im- mediately treated with camphorsulphonic acid in methanol giving the spiroacetal (299) in an overall yield of 55% from the epoxide (296) Conversion of this material into the aldehyde (300) was then smoothly accomplished as shown in the Scheme The third synthesis of milbemycin p3 to appear in the literature was that of Barrett and his associates (Scheme 39) Their synthesis involved the preparation of three key intermediates and their coupling and subsequent elaboration to give the required macrocycle The syntheses of these key synthons (301), (302), and (303) were first described at a symposium on the Chemistry of Insect Contr01,'~' and they have since been successfully utilized'41 142 in a total synthesis of milbemycin B3 Hydrogenation of the spirodihydropyrone (301) over rhodium on Davies and Green Li (297) li Reagents: i, Camphorsulphonic acid, MeOH, 20 "C; ii, Bu'Me2SiCl, DMF, Et3N; iii, Na, NH3(L);iv, Cr03*2pyridine, CH2C12 Scheme 38 alumina generated the alcohol (304),although the yield was variable.On the small scale, yields of 90% were reported, but on scale up only 58% of the alcohol (304) was observed along with approximately 6% yield of the epimeric alcohol. In addition, extreme care had to be observed in the hydrogenation process, since a trace of acid totally inhibited the formation of the alcohol (304); the sole observed product (85%) was the a$-unsaturated ester (305). Protection of the alcohol (304) as its t-butyldiphenylsilyl ether followed by kinetically controlled epimerization of the ester group afforded the spiroacetal (306) in 94% yield. Homologation of this ester in routine fashion provided the aldehyde (307). Condensation of this aldehyde with the lithium anion of the phenylsulphone (302) I4O S.V. Attwood, A. G. M. Barrett, R. A. Carr, M. A. W. Finch, and G. Richardson, 'The Application of Novel Carbanion Chemistry in Milbemycin-Avermectin Synthesis' in 'Recent Advances in the Chemistry of Insect Control', ed. N. F. Janes, The Royal Society of Chemistry Special Publication No. 53, London, 1985, p. 257. 14' S. V. Attwood, A. G. M. Barrett, R. A. Carr, and G. Richardson, J. Chem. Soc., Chem. Commun., 1986,479. 14' A. G.M. Barrett, R. A. E. Carr, S. V. Attwood, G. Richardson, and N. D. A. Walshe, J, Org. Chem., 1986,51,4840. 27 5 Avermectins and Milbemycins Part II CO2H n I OH (304) Il, OTBDPS OTBDPS OTBDPS IX-XIIII Milbemycin h -XH'---XV (28) OMe (309) Reagents 1, Bu'Ph2SiC1, DMF, imldazole, 11, LDA, THF, -78 "C then AcOH, 111, Dlbal-H, toluene, -78 "C, iv, Ph3P=CH2, THF, 0 "C, v, B2H6, Et20 then NaOH, H202, v1, PCC, CH2C12,vii, (302), BuLi, EtzO, 0 "C then Ac20, vii1, Na/Hg, THF, MeOH, -20 "C, IX, AcOH, H20, THF, x, PCC, CH2C12, xi, (303), NaH, Bu'Li, THF, -50°C then TFA, x11, TBAF, THF, 45 "C, xiii, KH, 18-crown-6, Et20, -3 "C then chromatography, xiv, Ph3P, DEAD, THF, xv, EtSNa, DMF, heat Scheme 39 followed by acetylation gave a complex mixture of diastereoisomers However, reductive eliminatlon afforded the required alkene (308) in high yield, although the geometric selectivity of the process was poor (AI4 E Z = 5 3) Although this selectivity could be increased by performing the elimination at reduced temperature, the yield of the process deteriorated dramatically Since Dauies and Green chromatographic separation of the isomers was impractical at this stage, the mixture was processed until separation could be achieved.Selective deprotection of the t-butyldimethylsilyl ether followed by oxidation gave an intermediate aldehyde, which was condensed with the double anion of the benzoic acid (303). Acidic work-up, followed by removal of the t-butyldiphenylsilyl ether, provided the benzopyrone (309) in an overall yield of 73%.Although this was obtained as a complex mixture of diastereoisomers at C-8 and C-9 and geometric isomers at C-14, base-mediated elimination, using a variation of Williams' chemistry,' 33 followed by chromatography, gave the required trans, trans-diene. The use of 18-crown-6 in this process considerably enhanced the rate of the reaction, thus allowing it to be performed at a lower temperature. Macrolactonization under Mitsunobu conditions 73 followed by Smith's phenol de-0-methylation pro- cedure'32 finally completed the synthesis of milbemycin p3 (28) in 51% yield. Although the spectroscopic data for this material were in excellent agreement with published data on milbemycin p3, one minor discrepancy was the value observed for the optical rotation.There is much conflict in the literature over this data since different groups have observed different values. For example, Williams 133 reported a figure of +26.5" (c. 0.2 in methanol) in his original synthesis, and similar values were obtained by both Baker 134 (+26.1') and Kocienski'37 (+32.8", c. 0.3 in methanol). In a later c~rrigendum,'~~ Baker amended his value to + 105" but, again, did not report the solvent employed. In addition, Kocienski 74 has reported a value of + 105" when the measurement was carried out in chloroform and stressed in his report that it was essential for the solvent to be alcohol-free chloroform. Barrett's rep~rt,'~','~~ on the other hand, claims the value to be + 102" (c.0.2 in methanol) and presents a reasoned argument for the discrepancy. In addition, he also reports that Williams now observes the following values for his synthetic material: + 103' (c. 0.28 in methanol), + 125.5" (c. 0.27 in chloroform), and +115.6" (c. 0.25 in acetone). Since the Sankyo chemists65 did not report the optical rotation for natural milbemycin 03 and the material is now no longer available, no conclusions can safely be drawn as to the correct value, although on the basis of Barrett's argument it would appear that his observed value is correct. In addition to these total syntheses of milbemycin p3, Crimmins and colleagues 144 have reported a formal synthesis which culminated in the attainment of the Williams intermediate (317). The lactone (310), an important intermediate in a number of synthetic approaches to milbemycin ~3,85~132~133~136~137~141~142was transformed in three steps into a mixture of the spiroenone (311) and the spiroacetal (312) (Scheme 40).The unwanted (312) could be readily converted into the required spiroenone by simple treatment with wet AmberIystB resin and thus the spiroenone was obtained in an overall yield of 52% from the lactone (310).Addition of vinyl magnesium bromide to the a,P-unsaturated ketone (3 11) in the presence of a catalytic amount of [Bu3PCu1l4 143 R. Baker, M. J. O'Mahony, and C. J. Swain, J. Chem. SOC.,Chem. Commun., 1986,276. 144 M. T. Crirnrnins, D. M. Bankaitis-Davis, and W. G.Hollis, Jr., J. Org. Chem., 1988,53,652. Auermectins and Milberntcins Part II t 3 steps 0a-0 OTBDPS (314) Reugents 1, wet amberlyst, CH2C12, heat, ii, eMgBr, [Bu3PCuII4, Et20, -55 "C, 111, NaBH4, DME, iv, Bu'Ph2SiC1, DMF, DMAP, imidazole, v, 9-BBN, THF, ultrasound, vi, (COC1)2, DMSO, CH2C12, -78 "C, vii, Ph,P=C(CH3)C02Et, CH2CI2, viii, Dibal-H, THF, IX, Ph3P, CBr4, CH3CN, x, (315), LDA, XI, LiAIH4, Et20, xi], (C0C1)2, DMSO, CH2C12, -78 "C, ~111,PhjP=CH*C02Et, CH2CI2, heat, XIV,Dibal-H, THF, -78 "C Scheme 40 gave a 5 1 mixture of diastereoisomers which were separated to give the olefin (313) in 60% yield, the stereochemistry of the product was verified by 250 MHz H spectroscopy The stereoselectivity of the process is dramatically affected by the size of the substituent at C-25 since earlier work" had shown a selectivity of 25 1 when there was an isopropyl group at this position Reduction of the ketone with borohydride gave a 3 1 mixture of epimers which were separated and the minor, unwanted, epimer recycled by Jones oxidation Protection of the required equatorial alcohol, followed by hydrobora- tion and Swern oxidation, then led to the spirocyclic aldehyde (307), an intermediate common to Williams' embryonic synthesis of optically active milbemycin p3, in excellent yield The C-14-C-15olefin was then smoothly introduced by Wittig coupling with complete stereocontrol of olefin geometry Reduction and bromination then gave the bromide (314) Addition of this bromide to an excess of the lithium enolate of (315), followed immediately by reduction, gave the alcohol (316) in 50% yield as a single diastereoisomer as observed by 400 MHz proton spectroscopy This alcohol was then routinely Davies and Green AOH transformed into the allylic alcohol (317), a late stage synthon in Williams' synthesis 33 of milbemycin p3. A number of model studies by Ley and colleagues at Imperial College have recently culminated in the fourth total synthesis of a milbemycin, namely milbemycin In early 1987, these workers had described a synthesis of the seco-analogue (318) which, disappointingly, was reported to be devoid of biological activity.146 Shortly after this synthesis was published, a series of papers appeared dealing with the syntheses of the two model compounds (319) 147 and (320),148 and it was this work that provided the basis for the total synthesis of milbemycin 01.In the earlier paper,I4' the crucial intermediate (321) was obtained in 13 steps from 4-methylanisole. This synthesis involved initial Birch reduction and Prins alkylation followed by trivial protection, resolution using (1s)-( -)-camphanic acid chloride, and subsequent elaboration to give the required intermediate (321). Following conditions elucidated in their second paper,'48 these workers were able to modify this molecule further such that a feasible synthesis of milbemycin PI was available. Thus, stereoselective hydrobora- tion of (321) (Scheme 41) gave a 4: 1 mixture of primary alcohols from which the predominant isomer was separated.Treatment of this with benzaldehyde under Dean-Stark conditions gave a 1:1 mixture of products from which the acetal (322) could be isolated in approximately 30% overall yield. Formation of the X-sulphonyl carbanion followed by quenching with phenylselenyl chloride gave a diastereoisomeric pair of selenides which, upon oxidation, underwent smooth syn-elimination providing the key intermediate (323). The intermediate (88) (see references 78 and 80) was then routinely converted into the aldehyde (324) and, on addition of this to the dianion of (323), the olefin (325) was produced in high yield. The exclusive E-geometry of this olefin was attributed to the minimization of non-bonded interactions between the bulky phenylsulphonyl and t-butyldiphenylsilyl groups.This material underwent smooth reductive elimination on treatment with sodium amalgam and the resulting E,E-diene was converted as shown in Scheme 42 into the carboxylic acid (326). Removal of the benzoyl '45 N. J. Anthony, A. Armstrong, S V. Ley, and A Madin, Tetrahedron Lett., 1989,30, 3209. '" T. Clarke and S. V. Ley, J. Chem. Soc., Perkin Trans. 1, 1987, 131 14' C. Greck, P Grice, A. B. Jones, and S. V. Ley. Tetrahedron Lett., 1987, 28, 5759. 14' N. J. Anthony, P Grice, and S. V Ley, Terrakedron Lett., 1987, 28, 5763. Avermectins and Milbemycins Part II SOgh 3DPSI ,OTBDPS i. ii -"W.... A OH ISO2Ph OH (322)1 iii, iv ...-CHO + 6COPh SO2Ph OH (324) 1.(323) OH (325) Reagents: i, BH3.DMS, THF then aq. NaOH, H202; ii, PhCHO, PPTS, benzene, reflux; iii, 2.2 eq. Bu'Li, THF, -78 "C then PhSeC1; iv, mcpba, CH2C12, aq. NaHC03; v, 2.2 eq.Bu'Li, THF, -78 "C Scheme 41 groups followed by macrolactonization under Mukaiyama conditions 149 gave the desired 16-membered lactone, which was then oxidized at C-5 and the benzylidene acetal cleaved to afford the ketone (327). Treatment of this material with t-butyldimethylsilyl triflate served not only to protect the C-8 alcohol but also provide the thermodynamic silyl enol ether at (2-5. Obtention of this enol ether was critical to the successful conclusion of the synthesis. Thus treatment with phenylselenyl chloride followed by oxidation gave the intermediate selenoxides which spontaneously syn-eliminated to give a 1:2 mixture of exo and endo olefin products.In order to prevent aromatization, this mixture was immediately reduced with sodium borohydride to give the mixture of alcohols 149 T. Mukaiyama, M. Usui, and K. Saigo, Chem. Lett., 1976,49. Davies and Green i-v MIX .___)(325) OCOPh OH OH OH (328) (329) (330) XUi, XN xii 1 Milbemycin (24) Reagents: i, 6%Na/Hg, Na2HP04, THF, MeOH, -40 "C; ii, PhCOCl, DMAP, pyridine, CH2C12; iii, BuSNF, THF, reflux, 15 min; iv, TPAP, NMO, 4A ground sieves, CH2C12; v, NaC102, 2-methyl-2-butene, KH2P04, Bu'OH, H20; vi, NaOMe, MeOH. vii, 2-chloro- 1-methyl pyridinium iodide, Et3N, CH3CN, reflux, 9 h; viii, TPAP, NMO, 4h ground sieves, CH2C12;ix, TFA, CH2C12; x, 4 eq.TBDMS.OTf, 20 eq. Et3N, CH2C12, r.t., then PhSeC1, CH2CI2, -78 "C; xi, 2-benzenesulphonyl-3-(p-nitrophenyl)oxaziridine,CDC13, r.t., then NaBH4, CeCI3, MeOH, r.t.; xii, TPAP, 4A ground sieves, CH2C12, then NaBH4, CeCI3, MeOH, r.t.; xiii, MeI, Ag20, ultrasound; xiv, HF, pyridine, CH3CN Scheme 42 (328), (329), and (330) in a 1: 1:l ratio. These were readily separated by chromatography and the least polar component (330) recycled, to provide further quantities of the equatorial alcohol (329), by the previously successful oxidation- reduction sequence. Finally, methylation of the C-5 alcohol (329) with methyl iodide under ultrasonication followed by deprotection of the t-butyldimethylsilyl ether gave the desired natural product, (+)-milbemycin (24), in 55% yield from the alcohol (329). A synthesis of 3,4-dihydromilbemycin E has been reported by Thomas and colleagues and the chemistry involved adapted to a synthesis of milbemycin E (Scheme 43).l5O The racemic Robinson annelation product (33 1) l6 was resolved, through its derived C-5 alcohol, by the use of (S)-(+)-acetoxymandelic acid, to give the dihydroxy ester (332), and this was transformed in nine steps into the hydroxy butenolide (333).Employing similar chemistry to that first described by 150E. R. Parmee, P. G. Steel, and E. J. Thomas, J. Chem. SOC.,Chem. Commun., 1989, 1250. 281 Avermectins and Milbemyins Part I/ OH OMe* (333) - \ ''73 mulbtep '..o 1OH -2.PO3 OTEDMS Milbemycin E vi I-(26) OMe OMe (336) (335) Reagenfs: i, 2 eq. LiHMDS, add (333), -78°C to -15"C, 3h; ii, CH2N2, Et20; iii, I2 benzene; iv, TBAF, THF, 10 h; v, DCC, DMAP, CH2C12, 0 "C, 16 h; vi, Dibal-H, toluene, -78"C, 1 h Scheme 43 Baker and colleague^,'^^ the spiroacetal (1 12)82*84 was transformed into the phosphorane (334) and reaction of this with the hydroxy butenolide (333) afforded the required diene (335) along with some of its C-10,C-11 geometrical isomer. These were not separated but esterified with diazomethane, and the double bond was isomerized using a catalytic amount of iodine to provide the required (E)-isomer in about 37% yield. Removal of the silyl protecting groups followed by macrocylic cyclization using DCC then provided the cyclic structure (336).Reduction of this ester as shown in the Scheme then provided milbemycin E (26) in high yield. A notable feature of the synthesis was that the A3 double bond was not found to migrate into conjugation with the carboxylic acid, a problem which has beset a number of syntheses of the avermectins (vide infra). Finally, Smith and Thompson have reported an efficient synthesis of an Is' A. B. Smith and A. S. Thompson, Tetrahedron Lett., 1985,26,4283. Davies and Green CO,Me %Me OMe Reagents: i, TBDMS.OTf, CH2C12, 2,6-lutidine, 0 "C, 15 min; ii, NaBH4, MeOH, 0 "C to r.t., ih; iii, Mo(CO)~,Bu'OOH, benzene, A, 1.5h;iv, H5IO6,Et20, r.t., 40min; v, NaHMDS, (339), THF, HMPA; vi, BulNF, THF, 18 h; vii, KH, KHMDS, THF, r.t., 1 h; viii, EtSNa, DMF, A Scheme 44 avermectin-milbemycin hybrid.' ' The aldehyde (337), readily available in five steps by degradation of avermectin Bla152(vide infra), was protected as its t- butyldimethylsilyl ether and the aldehyde functionality reduced to a primary alcohol (Scheme 44). Sharpless epoxidation of the C-10-C-11 olefin, followed by oxidative cleavage, afforded the aldehyde (338) in 13% overall yield.Horner- Emmons condensation of the aldehyde with the phosphine oxide (339) '32 in the presence of sodium hexamethyldisilylazide, followed by desilylation, macrolactonization, and demethylation, led to the formation of the macrocyclic hybrid (340). The amount of base employed was critical to the successful Horner-Emmons condensation. Two equivalents led only to p-elimination, whereas with three equivalents and excess phosphine oxide, a 77% yield of product was observed.B. Avermectins.-While there has been a wealth of literature on the partial synthesis of the avermectins and milbemycins as well as a number of reports of A. B. Smith and A. S. Thompson, Tetrahedron Lett., 1985,26,4279. Avermectins and MiIbemycins Part II total syntheses of milbemycins, a total synthesis of an avermectin had, until recently, proved elusive. However, in 1986, Hanessian and his co-worker's successfully surmounted this obstacle and reported the first total synthesis of a member of this group of macrocycles 94*95 when they described the total synthesis of avermectin B1, (Scheme 45). These workers had earlier described the synthesis of the spiroacetal (344) from chiral precursor^,'^^^^ and this was utilized in the total synthesis by conversion into its phenylsulphonyl derivative (345) in routine fashion.Condensation of the anion of this with the chiral ketone (341), obtained in nine steps from (S)-malic acid, followed by reductive elimination and protecting group removal, resulted in the formation of the trio1 (346). Chemical differentiation between the primary and secondary alcohol moieties then served to generate the phenylsulphonyl derivative (347) as shown in the Scheme. Condensation of this with the aldehyde (342), reportedly obtained by a controlled oxidative degradation of avermectin B1, '54 (vide infra), followed by reductive elimination and desilylation, provided the trans-diene (348) as the only dienic product.Removal of the ester protection and macrolactonization then gave the cyclic compound (349), which was protected at C-5 as a t-butyldimethylsilyl ether (350). All that now remained to complete the total synthesis was the coupling of the disaccharide unit at C-13 and deconjugation of the C-2 olefin. The first problem was achieved by resort to a procedure developed by Hanessian,lS5 whereby the protected ether (350) was treated with the 2-pyridylthioglycoside derivative of the disaccharide (343) in the presence of silver triflate. This process provided a mixture of glycosides from which the desired product (351) could be isolated after chromatography.The final stage of the synthesis relied upon a pivotal deconjugation of the C-2 olefin. This was successfully achieved by initial protection of the C-7 hydroxyl group, to preclude its participation in the subsequent acid quench, followed by treatment with strong base in the presence of trimethylsilyl chloride. Acid hydrolysis of the so formed keteneacetal then served to deliver its proton from the requisite face of the intermediate, resulting in the formation of the elusive avermectin (5) after desilylation. In spite of this prodigious achievement, later work by Hanessian and colleague^,'^^ in an attempt to optimize the deconjugation process, showed the reaction to be extremely capricious. They noted 'an unpredictable variation in the nature of the products, even with only the slightest change in experimental procedure'.In relation to this work, Fraser-Reid and colleagues also investigated the deconjugation stage 157 but were unable to duplicate the work of Hanessian, who concludes '56 that 'the material produced in the original deconjugation was not the primary product of deconjugation, but possibly the result of a subsequent epimerization of an initially formed 2-epi isomer'. In spite of the mutable nature of the final deconjugation step of the synthesis, this work nevertheless constitutes S. Hanessian, A. Ugolini, and M. Therien, J. Org. Chem., 1983,48,4427. S. Hanessian, A. Ugolini, P. J. Hodges, and D. Dube, Tetrahedron Left., 1986,27,2699.155 S. Hanessian, C. Bacquet, and N. LeHong, Curbohydr. Res., 1980,80, C17. S. Hanessian, D. Dube, and P. J. Hodges, J. Am. Chem. Soc., 1987,109,7063. B. Fraser-Reid, H. Wolleb, R. Faghih, and J. Barchi, Jr., J. Am. Chem. SOC.,1987,109,933. Davies and Green TBDMSO... OTBDPS OTBDMS (342) (343) R -TMS I I I OCHPh R2 OR" (344)R -OH (346) R' -R" -H; R2-OH n-xl"IIK(345) R -PhSOz (347)R' -R3 -TBDMS; R2-SOPh o&*--. H OR (349)R-H (350)R-TBDMS AOJ X1X--XXI Avermain Bla P (5) (351) R -TBDMS Reagents: i, PhSSPh, Ph3P, THF; ii, rn-CPBA, CH2C12, -10 "C; iii, Bu"Li, THF, -78 "C then add (341); iv, Na/Hg, MeOH, THF, KHlPO,; v, BuZNF, THF; vi, Li/NH3; vii, Bu'OCl, Et3N, CH2C12; viii, Bu'Me2SiCl, imidazole, DMAP, DMF; ix, NaOMe, MeOH, CH2C12; x, PhSSPh, Bu?P, THF; xi, rn-CPBA, CHZCl,; xii, Bu"Li, THF, -78 "C then add (342); xiii, S0Cl2, pyr then Na/Hg, MeOH; xiv, BuSNF, THF; xv, aqueous KOH, THF, then Dowex 50 (H ), xvi, DCC, DMAP, CH2C12; xvii, Bu'Me2SiC1, imidazole, DMF; xviii, (343), CH2C12, AgOTf, toluene; xix, Me3SiC1, Et3N, DMAP, CH2C12; xx, LDA, Me3SiC1, THF, -78 "C then AcOH, THF, -78 "C to r.t.; xxi, BuZNF, THF Scheme 45 285 Avermectins and Milbemycins Part 11 ....OPV OHC OPV (352) (353) (355) (358) (357) (356) Reagents 1, (354), MgBrz, CH2C12, ii, NaBH4, CeC13, CH2C12, EtOH, -78"C, iii, TBDMS OTf, CH2C12, lutidine, -78 "C, 30min, iv, NBS, NaHC03, aq THF, v, Bu3SnH, AIBN, toluene, VI, LiBH4, THF, vii, PrCl, CH2C12, Et3N, DMAP, viii, 5% HF in CH3CN, ix, HgO, 12, CC14, hv, 1 5 h, x, LIOH, MeOH, THF, H20 Scheme 46 an immense scientific achievement due to the high degree of stereoselectivity that is realised A fuller explanation of the results of the deconjugation/epimerization experiments of Hanessian and Fraser-Reid is presented in a later Section of this review More recently, Danishefsky and his associates have reported their investiga- tions, which culminated in the total synthesis of another member of the avermectin family, namely avermectin Al, 96 97 15' In order to obtain their goal, they identified two chiral intermediates, (358)and (362), as being the keys to the synthesis Thus the D-glucal derivative (352) was transformed, in eight high yielding steps (30% overall yield), into the chiral aldehyde (353) (Scheme 46) which, on treatment with 1-methoxy-3-(trimethylsilyloxy)-1,3-butadiene (354), afforded a diastereoisomeric pair containing the isomer (355) as the major constituent This mixture was reduced and the product purified by chromatography to give the homogeneous unsaturated alcohol (356) Progression of this, as shown in the Scheme, furnished the alcohol (357), in 49% overall yield, and oxidative cyclization followed by depivaloylation gave the target spiroacetal (358) This material was identical to an authentic sample obtained by osmium tetroxide degradation of A' avermectin Al, aglycon '59 (vide znfra) The second lSs S Danishefsky H G Selnick D M Armistead and F E Wincott, J Am Chem Sol 1987 109 81 19 lSy H G Selnick and S J Danishefsky Tetrahedron Letf , 1987 28 4955 286 Davies and Green fOTBDMS "Me (359) (360) (361) (362) Reagents: i, (E)-trimethylcrotylsilane, BF3-Et20, CH2C12, -78 "C;ii, NaH, THF, MeI; iii, HCl, MeOH; iv, Et3SiH, BF3-Et20, CH2C12; v, Me2C(OAc)C0.Br, CH2C12, 1 h; vi, Amberlite IRA-400, MeOH; vii, LiEt3BH, THF, 0 "C, 6 h; viii, 03,CH2C12, -78 "C,5 min then Zn, AcOH; ix, Ph3P=CH*C02Me, CH2C12, 12h; x, Dibal-H, Et20, O'C, 2h; xi, Bu'Me2SiC1, Et3N, DMAP, CH2C12; xii, PCC, NaOAc, CH2C12 Scheme 47 target intermediate (362) was obtained by an adaptation of some earlier model studies.' Thus the aldehyde (359), readily obtainable from D-ribose, was reacted with (E)-trimethylcrotylsilane followed by methylation, yielding the olefin (360) (Scheme 47).Conversion of this acetonide to the epoxide (361) was successfully achieved by resort to a well established process, and this epoxide was regioselectively reduced with lithium triethylborohydride to provide the potential C-7 hydroxyl group of the final product. Ozonolysis, with reductive work-up, of the terminal olefin, followed by Wittig coupling, reduction and protection, served to generate an alcohol, which was oxidized to afford the second target intermediate (362). The next stage of the synthesis involved the stereoselective introduction of the C-12 methyl and C-13 hydroxyl substituents. Protecting group manipulation of the spiroacetal (358) gave the aldehyde (363), and this was homologated, as shown in Scheme 48, which served to introduce the required E-olefin and afforded the a,P-unsaturated aldehyde (364) in excellent yield.Utilization of the elegant crotonyl borate chemistry of Roush and colleagues 160 provided access to the stereocontrolled introduction of the C-12 and C-13 substituents, resulting in the formation of the vital C-1 1 to C-25 fragment of the avermectins. The product was obtained as a 4:l mixture of diastereoisomers which proved difficult to separate. However, protection of the secondary alcohol to give (366), followed by selective dihydroxylation of the monosubstituted olefin and cleavage of the resulting diol with lead tetraacetate, provided a mixture of aldehydes from which the major diastereoisomer (367) was obtained by chromatography. Simple homologation as shown in the Scheme then gave the a,P-unsaturated aldehyde (368).The total synthesis was then completed by the process depicted in Scheme 49. The lithium enolate of (362) was condensed with the x,P-unsaturated aldehyde (368) and the product dehydrated with methanesulphonyl chloride to afford the (E,E)-diene system, which was selectively deprotected and oxidized to afford the aldehyde (369). Cyclization to the hexahydrobenzofuran system (370) was accomplished as shown and the product oxidized and deprotected to give the hydroxy acid (371) in high yield. I6O W. R. Roush and A. D. Palkowitz, J. Am. Chem. SOC.,1987,109,953. Avermectins and Milbemycins Part II A A OPV OPV (363) OPV (367) (368) Reagenfs 1, Ph3P=C(CHO)CH3, benzene, reflux, 36 h, 11, (365), toluene, -78 "C, I 5 h, 111, Bu'Me2SiOTf, 2,6-lutidine, CH2CI2, 0 "C, 0 5 h, iv, oso4,THF, pyridine, v, P~(OAC)~,CH2CI2,0 "C, 5 min then chromatography, vi, Ph3P=CH-C02Me, CH2C12, r t , 24 h, vii, Dibal-H, THF, -78 "C,2 h, viii, (C0C1)2, CH2C12, DMSO, -78 "C, 1 5 h Scheme 48 Macrolactonization using the Mukaiyama procedure 149 followed by desilylation, then produced the A2 isomer of avermectin Al, aglycon (372) The final stage of the synthesis was achieved by employment of Hanessian's modified deconjugation methodology 56 Thus action of lithiumdiisopropylamide on (372) followed by quenching with aqueous hydrochloric acid gave a 75% yield of C-2 epi-avermectin Al, aglycon (373) along with a 21% yield of recovered starting material These were readily separated by chromatography and the C-2 epi-compound (373) heated with a concentrated solution of imidazole in benzene for 15 hours This resulted in a mixture of 33% recovered starting material (373), 21% of A2 avermectin Al, aglycon (372)' and a 32% yield of the elusive aglycon of avermectin Al, (374)' all of these compounds were readily separated by chromatography To complete the synthesis of the natural product, Danishefsky has developed a new protocol 158 for the formation of the disaccharide portion and its coupling with the aglycon (374) Thus the unsaturated disaccharide (375) was coupled to the aglycon (374) through the use of N-iodosuccinimide, followed Davies and Green iiii ix. x H bMe (372) (370)R'=Pv R=CHO vi-viii 1(371)R' = H R = C02H xi, xii Avermectin Ala (11 (373) X-a-H (374)XI P-H (375) Reagents: i, LiHMDS, (368), THF, -78 "C then MeS02C1, Et3N; ii, HF, CH3CN, -20°C; iii, PCC, NaOAc, CH2C12; iv, Me3AI, PhSLi, THF, O'C, 10min then rn-CPBA, CH2C12, -2O"C, 2 h; v, toluene, heat, 30 min; vi, Bu'OH, 2-methyl-Z-butene, NaC102, NaH2P04; vii, CH2N2, Et20; viii, LiOH, MeOH, H20; ix, 2-chloro-N-methylpyridiniumiodide, Et3N, CH2CI2, CH3CN, heat; x, Bu~NF, THF, r.t., 2h; xi, LDA, THF, -78"C, 15 min then 1N-HCl; xii, imidazole, benzene, reflux, 1.5 h; xiii, N-iodosuccinimide, CH3CN, (375), r.t., 1 h; xiv, Bu!jSnH, AIBN, toluene, reflux, 10min; xv, LiEt3BH, THF, -78 "C, 3 h Scheme 49 by reductive dehalogenation and deacylation, which resulted in the completion of the total synthesis of avermectin Al, (1).Avermectins and Milbemycins Part II 0°C- -*-0 + SEMo...(., /*- ..a- OMPM 0Pv OPv OMPM (376) (377) (378) ~SEMO...e vi-viii SEMCL-.,** ..*' OH OTBDMS OPv OMPM xii-xiv / 'CH2S02P h (3811 (382) Reagents: i, LDA, THF, -78"C, 0.5h; ii, Ac20, Et,N, DMAP, CH2Clz then DBU; iii, MeMgC1, THF, 0 "C, 1 h; iv, PhSCl, Et3N, CHzCl2, -78 "C to 25 "C; v, oxone, MeOH, H20; vi, (NH4)2Ce(N03)6, H20, CH3CN; vii, Bu'Me2SiOTf, 2,6-lutidine, CH2Clz; viii, LiAlH4, THF; ix, Na(Hg), NaZHPO4, MeOH, 25"C, 2h; x, (C0C1)2, DMSO, Et,N, CH2ClZ,-78 "C to 25 "C; xi, Ph3PCH-C02Me, toluene; xii, Dibal-H, toluene, -75 "C; xiii, NCS, Mez& CHzClZ, 0 "C to 25 "C; xiv, PhSOzNa, DMF, 25 "C, 48 h Scheme 50 To date, two other total syntheses of an avermectin have been described; the first of these was reported by White and colleague^.^^ These workers had previously published routes to both the spiroacetal 87 (Scheme 11) and oxahydrindene121 (Scheme 28) moieties of avermectin Bl,, and they have now succeeded in linking these two portions to provide the third total synthesis of an avermectin, namely avermectin B 1, aglycon.The linear segment (377) required for the completion of the synthesis was obtained in routine fashion from ethyl levulinate in ten steps. The lithium anion of this molecule (377) was condensed with the aldehyde (376) [readily obtainable by Swern oxidation of the primary alcohol (124)87] giving the crossed aldol product (378) (Scheme 50) in high yield.P-Elimination of this hydroxy ketone and reaction with methyl Grignard, followed by elaboration as shown in the Scheme, afforded the sulphone (379). Protecting group adjustment followed by reductive removal of the pivalate protection then provided the primary alcohol (380). Reductive removal of the phenyl sulphonyl group was accomplished in reasonable yield and the product Davies and Green oxidized and subjected to Wittig coupling, giving the a$-unsaturated ester (38 1) in high yield. Standard chemistry then served to provide the sulphone (382) as shown in Scheme 50.In order to complete the synthesis of avermectin Bl,, some minor modifications had to be made to the previously described oxahydrindene ''I synthesis in order to introduce the C-7 hydroxyl function (avermectin numbering) of the final product. Thus the previously described silyl ether (230) was converted to a SEM ether and the lactone opened with methoxide ion; accompanying migration of the double bond then occurred, providing the bicyclic ketone (383) (Scheme 51). That the double bond had migrated into conjugation with the ester moiety proved advantageous, in that conversion of the ketone into its silyl ether followed by epoxidation provided the required C-7 hydroxylated oxahydrindene. Routine protection then gave the silyl ether (384).Julia coupling'6' of this ketone (384) with the previously described sulphone (382) (Scheme 50) gave the hydroxy sulphone (385) in 50% yield, but all attempts to effect elimination failed and in each case the major product was the lactone (386). Fortunately these workers were able to take advantage of this unexpected occurrence and thus by sodium amalgam reduction were able to produce the requisite diene (387). Deprotection of the t-butyldimethylsilyl ether followed by macrolactonization under Mukaiyama conditions '49 afforded the lactone (388) of inverse stereochemistry at C-2 to that present in the naturally occurring avermectins. In order to overcome this problem, resort to Hanessians epimerization conditions 56 provided a 34: 50 ratio of C-2 epimers, along with 16% of a A2 isomer which was removed by chromatography.Finally, cleavage of the remaining protecting groups followed by chromatographic purification pro- vided the target avermectin aglycon (389), whose spectroscopic properties were identical to those of an authentic sample derived from hydrolysis of the naturally occurring macrocycle.' 62 Very recently, Ley and his colleagues at Imperial College have published a series of papers 99-1 O' describing their success in achieving a total synthesis of avermectin B1,. The optically active cyclohexanone (390), a precursor in Ley's total synthesis of milbemycin was dehydrated and the resulting unsaturated ketone epoxidized with dimethyldioxirane (Scheme 52).99 This pro- vided a 5: 1 mixture of a and p epoxides which were readily separated by chromatography.Addition of 2-lithio-4-phenylthiobut-1-eneto the major epoxide then provided the sulphide (391) as the only detectable isomer. Confirmation of the absolute stereochemistry of this product was obtained by X-ray crystal- lography of a derivative of the corresponding sulphone. Oxidation of the sulphide (391) followed by epoxide ring opening gave the tetraol (392). Unfortunately, the vigorous conditions required for epoxide opening also removed the protecting group at C-1. Protecting group adjustment followed by inversion of the alcohol at C-6, utilizing an oxidation-reduction sequence, gave a 6: 1 mixture of alcohol diastereoisomers from which the major one (393) was 16' M Julia, Pure Appl Chem, 1985,57,163 H Mrozik, P Eskola, B H Arison, G Albers-Schonberg, and M Fisher, J Org Chem, 1982, 47, 489 291 '* Avermectins and Milbemycins Part II 0% Me0.6 HI I-0.0OSiEt, A 0 A OSEM OSEM (383) (384)R-Et3Si Vlll c--OSEM (386)R = EtjSl (385) R = EtSt x XI -OSEM (387)R = Et$l Avermectin Bla aglycon (389) Reagents 1, TFA, THF, H20, 11, Me3SiCH2CH20CH2Cl, PriNEt, CH2C12, 111, K2C03, MeOH, iv, Et,SiOTf, 2,6-lutidine, CH2C12, v, rn-CPBA, CH2C12, vi, Et,SiOTf, 2,6-lutidine, CH2CI2,vii, Bu"Li, THF, (382), -78 "C, ~111,NaOMe, Na2HP04, MeOH, 0 "C, 0 5 h, ix, Na(Hg), Na2HP04, MeOH, 0 "C, 3 h, x, TBAF, THF, xi, 2-chloro-1-methylpyridiniumiodide, Et3N, CH3CN, reflux, 2 5 h, xii, imidazole, benzene, reflux, 1 5 h,XIII, HF, MeCN Scheme 51 obtained pure by chromatography Conversion of this material into the important intermediate (394) was then achieved by utilization of similar chemistry to that 36employed in the total synthesis of milbemycin Davies and Green 1v.v -OH (392) S0,PhI ,OTBDPS /OTBDPS -... OTBDMS X - OH iq.... OTBDMS (394) (393) Reagents: i, MsCl, Et3N, CH2C12; ii, dimethyldioxirane, Me2C0, CH2C12; iii, PhS(CH2)2C(=CH2)Li, THF, -78 "C;iv, oxone; v, 15% H2S04, THF, 60 "C;vi, Bu'Ph2SiC1,imidazole, DMF; vii, Bu'Me2SiOTf, Et3N, CH2C12; viii, (COCl)*, DMSO then Et,N; ix, NaBH4, MeOH; x, ref. 36 Scheme 52 The next stage of the synthesis involved preparation of the C-11-C-25 spiroacetal portion, and this was achieved by resorting to the use of n-ally1 iron complexes."' The readily available allylic alcohol (395) was converted, in three steps, into the epoxide (396) (Scheme 53); this, on treatment with di-iron nonacarbonyl, gave a mixture of diastereoisomeric n-allyltricarbonyliron lactone complexes.Interestingly, although these products are diastereoisomeric, the key homochiral asymmetric carbon atom is common to both isomers and thus on exhaustive carbonylation only one enantiomer (397) was obtained. Conversion of this lactone into the sulphone (398) was then achieved using previously reported chemistry.' 63 Condensation of the sulphone (398) with the epoxide (399), readily obtained from 2-propynyl-1-01, gave an unstable enol ether which was im-mediately transformed using standard methodology into the spiroacetal diol (400).Protection of the two alcohol moieties followed by oxidative cleavage of the double bond and oxidation of the resulting diol gave the required intermediate (401). With these two important precursors (401) and (394) in hand, the total synthesis of avermectin B 1, was now within reach. Thus sulphone-stabilized anion coupling of (401) and (394) gave the diastereoisomeric hydroxy sulphones (402) (Scheme 54)."' This unstable mixture was immediately subjected to reductive elimination and the resulting diene fully deprotected. The so formed penta-hydroxy derivative was then selectively oxidized at C-1, in two steps, to 163 D.S. Brown, M. Bruno, R. J. Davenport, and S. V. Ley, Tetrahedron, 1989,46,4293. 293 Avermectins and Milbemjvins Part II i.iiBsleps_ yH \ (395) (396) (397) liii + L OTBDMS Reagents: i, Fez(C0)9, THF; ii, CO, 250 atm, benzene, A; iii, ref. 167; iv, BF3-Et20, BuLi, -78"C, THF; v, PhSeC1, MeOH, Et2N, CH2C12 then CSA, MeOH, CHzClz; vi, p-nitrophenyl-N-sulphonyl oxaziridine, CHC13, 50 "C; vii, TBAF, THF, A; viii, Bu'MezSiCl, imidazole, DMF; ix, Os04, NMNO, Bu'OH, THF, HzO; x, NaI04, KH2PO4, HzO, MeOH Scheme 53 afford the carboxylic acid (403). Cyclization under Mukaiyama conditions 149 provided the 16-membered ring macrocycle and this was subsequently oxidized at C-5. Treatment of the resulting ketone with trimethylsilyl triflate provided the silylenol ether at C-5 along with simultaneous protection of the C-7 and C-13 alcohols.Conversion of the enol ether to the intermediate selenide followed by removal of the unwanted silyl protecting groups then provided the ketone (404) as a 1 : 1 mixture of diastereoisomers. Selenoxide formation followed by syn-elimination and reduction then provided the elusive avermectin aglycon (389) in reasonable yield. The C-4-exomethylene isomer was also obtained from the selenoxide elimination but was readily removed by chromatography. An intrigu- ing feature of the synthesis is the lack of protecting groups employed in the final stages; this was achieved by the innovative exploitation of the differential chemical reactivity of the individual hydroxyl groups.With the aglycon in hand, all that now remained to complete the synthesis of avermectin B1, was to introduce the disaccharide moiety at C-13. In order to achieve this final stage, Ley and colleagues have developed a new synthesis of the bis-oleandrose fragment involving n-ally1 tricarbonyliron lactone complexes as synthetic intermediates.lo2 Thus, by utilization of this type of chemistry, the fourth total synthesis of an avermectin was successfully completed. Davies and Green I OTBDMS (4011 PhSOp ,OTBDPS % OTBDMS (394) OH XI, YIII Avermectin Bla (389) Reagenfs: i, 2.2~Bu'Li, THF, -78°C; ii, 6% Na/Hg, THF, MeOH, -30°C; iii, TBAF, THF, A; iv, RuCI~(PP~~)~,C6H6; v, NaOC12, KH2P04, 2-methyl-2-butene, Bu'OH; vi, 2-chloro-1-methylpyridinium iodide, Et3N, MeCN, A; vii, TPAP, CH2C12, 0 "C; viii, TMSOTf, Et3N, CH2C12, 0 "C;ix, PhSeC1, CH2CI2, -78 "C; x, TBAF, THF, 0 "C to r.t.; xi, p-nitrophenyl-N-sulphonyloxaziridine,CHC13, r.t.; xii, NaBH4, CeCI3, MeOH, 0 "C; xiii, ref.166 Scheme 54 295 Avermectins and Milbemycins Part I1 i-iv Milbernycin A OH Reagents: i, 03, CH2C12, -78 "C; ii, Me2$ iii, NaBH4; iv, OH-Scheme 55 In view of the vast amount of excellent work that has been published on the synthesis of a variety of fragments of the avermectins, it is to be expected that a number of other total syntheses will be reported in the near future. It is also to be hoped that the content of these will be as chemically stimulating as the four syntheses already published.8 General Chemistry A. Degradation of Natural Avermectins and Mi1bemycins.-Avermectin and milbemycin fragments derived from degradation of the natural products are necessary both to confirm the structures of synthetic intermediates and to act as relay compounds in the course of a total synthesis. This problem was first addressed by Smith and his co-workers, who degraded milbemycin p1 (24) to provide the spiroacetal fragment (405) in 28% overall yield (Scheme 55).'32 A more difficult problem was to degrade one of these macrolides in such a way as to provide intact C-9 to C-28 and C-1 to C-8 fragments. For avermectin B1, (5), the solution lay in recognizing the C-8-C-9 double bond as part of an allylic alcohol system and thus susceptible to chemoselective functionalization.' 52 Protection of the C-5 and C-4" hydroxyl groups, as t-butyldiphenylsilyl and t- butyldimethylsilyl ethers respectively (or as a bis-acetate), gave the allylic alcohol (406), which was subjected to Sharpless epoxidation (Scheme 56).It was then necessary to protect the C-7 hydroxyl group as a methyl ether (407) before opening the epoxide with aqueous fluoroboric acid. Reduction of the macrolide ester group with lithium aluminium hydride and oxidative cleavage of the 8,9-diol with lead tetra-acetate gave the required fragments (408) and (409) in good overall yield [a protected form of the aldehyde (408) was later used in a synthesis of a milbemycin-avermectin hybrid].Degradation of avermectin B1, (5)into C-1-C-10 and C-11-C-28 segments has been reported by Hanessian and his co-workers (Scheme 57).'54In this approach. the A3 double bond was deliberately moved into conjugation with the ester group to minimize the risk of aromatization of the oxahydrindene group during subsequent manipulations. The macrolide ester was simultaneously hydrolysed during the isomerization to give the seco-ester (410) after esterification with diazomethane; the seco-ester lacking the disaccharide unit was obtained from the aglycon. After silylation of (410) the C-1W2-11 double bond was selectively ozonized in the presence of Sudan 7B as an indicator, and the resulting ozonides were reduced with sodium borohydride to give the fragments (411) and (412).Davies and Green )l.Q 1 OTBDPS OTBDPS OMe R = TBDMSO,.J$...oJ 6 OTBDPS0 *-*o (409) Reagents: i, Bu’Ph2SiC1, imidazole; ii, Bu’Me2SiC1, imidazole; iii, Bu‘OOH, VO(acac)2; iv, CH2N2;v, HBF4, HzO, Et2O; vi, LiAIH4; vii, P~(OAC)~ Scheme 56 The intact disaccharide moiety (413) could be obtained from (41 1) by pyridinium chlorochromate oxidation and base-catalysed elimination. This degradative procedure to obtain the disaccharide unit has been improved by workers at Merck, who eliminated the conjugation/hydrolysis and esterifica- tion steps and ozonized 4”,5-bis-O-t-butyldimethylsilylavermectin B 1 directly. 64 The resulting ozonide was decomposed with methyl sulphide, and DBU was then added to eliminate the disaccharide unit (413).Thus, at the expense of the loss of the aglycon sub-unit, the disaccharide unit could readily be obtained in 57% overall yield with a single day’s work. A method for cleaving avermectin Al, (1) at C-14-C-15/C-13-C-14 and the macrolide ester has been provided by Selnick and Danishefsky (Scheme 58).lS9 The aglycon of avermectin Al, was converted into the conjugated isomer (414) with DBU in benzene, and this was treated with osmium tetroxide to provide the tetra-ol (415) as a single isomer. This was assigned the 14~~,lS~~-cis-hydroxyl structure, as osmylation on the p face of the avermectin would be sterically hindered. When treated with lead tetra-acetate the tetra-ol was cleaved into an ester-dial which was reduced and hydrolysed to give the two fragments (416) and (417).A corresponding osmylation on the non-conjugated avermectin A 1, aglycon gave a mixture of products corresponding to osmylation at C-3-C-4 as well as C- 144-15. ‘64 T. A. Blizzard, G. Marino, H. Mrozik, and M. H. Fisher, J. Org. Chem., 1989,54, 1756. 297 Avermectcns and Milbemjwns Part I1 I I1Avermectin Bla -(5) R'. 'Hc-c;"l-r OH OTBDMS R'H + "0&%H (413) A OTBDMS R' = TBDIJSO,H(15 d. A= =fi.o Reagents 1, KOH, H20, DME, 11, CH2N2,111, Bu'Me2SiCI, imidazole, iv, 03,Sudan Red, v, NaBH4, VI, PCC, vii, KHMDS, THF Scheme 57 Upon treatment of avermectin B1 with DBU in methanol at 55"C, the seco- ester (410), previously obtained by Hanessian, was formed in 25-30% yield 165 An interesting minor by-product (418) was also obtained, presumably formed by a series of retro-aldol reactions When the DBU reaction was repeated using diethylamine as solvent, the sole product, obtained in 90% yield, was the A'-isomer of avermectin B1 Ozonolyis of the milbemycin S541 factor B (now called nemadectin B), followed by reductive work-up with dimethyl sulphide, gave an approximately 165 T A Blrzzard, H Mrozik, and M H Fisher Tetruhedron Let! 1988,29 3163 Davies and Green i,ii iii-v1 OH OMe (416) (417) Reagents: i, 0~0,;ii, NaHS03;iii, Pb(OAc)z,MeOH, C6H6;iv, NaBH4; v, KzC03, MeOH Scheme 58 equal mixture of three components (4 19)-(42 1) corresponding to non-specific cleavage of the macrolide and C-25 side-chain.166 However, when the side-chain was protected by conversion into the chloro-compound (422) (see Section SH), the ozonolysis was more selective and excised the C-11-C-14 fragment to give the dial (423) in 59% yield (Scheme 59).When the chlorination and ozonolysis were performed concurrently with no intermediate purification and the ozonide was reduced with sodium borohydride, the corresponding diol (424) could be obtained in 28% overall yield. The C-25 side-chain could then be resurrected by reduction with tributyltin hydride with only 20% contamination by the isomeric 166 C. E. Mowbray, M. J. V. Ramsay, and S. M. Roberts, J. Chem. SOC.,Perkin Trans. 1, 1990, 1813. Auermectins and Milbemycins Part II OH OMe (419) H (422) (423) X = 0 (424) X = H, OH Reagents: i, 03;ii, NaBH4 Scheme 59 1-isopropylethene analogue.Various ‘semi’-milbemycins (425) were prepared from the diol (424) by cyclization with dicarboxylic acid chlorides, and the dechlorinated material was also converted into lactones by treatment with malonyl and glutaryl chlorides. The macrocyclic ether (426) was prepared in 18% yield by tosylation of the diol (424). The specific cleavage of the outer spiroacetal ring of avermectins and Duvies und Green (425) n = 0to 3 (426) milbemycins is a desirable goal as it offers a means of preparing these compounds with totally novel spiroacetal substituents. This goal has been achieved by workers at Merck and at Beecham by very different methods.The Merck group formed the enolate of 23-0x0 avermectin B2, (silyl protected at C-4" and C-5 and trapped this with trimethylsilyl chloride to produce the TMS-enol ether (427) (Scheme 60).'67The choice of base for the enolization was critical in view of the labile proton at C-2 of the avermectin (see Section 8F), but it was found that lithium bis(trimethylsi1yl)amide regioselectively gave the required enolate in 75% yield. As the silyl enol ether is electron rich, it was preferentially epoxidized with rn-chloroperoxybenzoic acid, only a little reaction being observed at the 14,15- double bond. Dilute acid treatment of the silyloxy-epoxide (428) gave the X-hydroxy-ketone (429), which was readily oxidized with lead tetra-acetate to the aldehydo-acid (430).Transacetalization then gave a 1:1 mixture of the methoxy- aldehydes in high yield, which could be isomerized to the thermodynamically most stable diastereoisomer (431a) by extending the reaction time for the transacetalization. These methoxy-aldehydes were ideal precursors to avermectin analogues, as described in the subsequent paper.'68 Reaction of either isomer of the aldehyde (431) with unstabilized Wittig reagents (432) (Scheme 61) gave the expected cis-alkene (433), which was spiroacetalized with pyridinium p-toluenesulphonate in methanol to give the required avermectin analogue (434) in 3349% overall yield. A range of avermectin analogues were readily prepared in this manner.It should be noted that the use of TMS protection for the primary alcohol is critical; replacement with a TBDMS group gives complex mixtures as the methoxyacetal then undergoes hydrolysis and fragmentation before the primary alcohol is free to trap the incipient carbocation. Whichever diastereoisomer of the Wittig product was used for the spiroacetalization, a single isomer at the spiroacetal centre was isolated. Ultimate confirmation that this was the case was provided by reconstituting avermectin B1, (5) from the methoxy-aldehyde (43 l), the resulting product being identical to the natural material. lh7 T. L. Shih, H. Mrozik, M. A. Holmes, and M. H. Fisher, Tetrahedron Lett., 1990,31,3525. 16' T. L. Shih, H. Mrozik, M. A. Holmes, and M.H. Fisher, Tetrahedron Left., 1990,31,3529. 301 Avermectins and Milbemycins Purt II\& \&Avermectin B2a i. ii ''-0 iii --..o ---H (7) P P (430) (429) 00 (431a) a-OMe;0-CHO (431b)a-CHO; P-OMe Reagenrs: i, Bu'Me2SiCl, Et3N; ii, (COC1)2, DMSO; iii, LiHMDS, TMSCl; iv, mcpba; v, AcOH, MeOH; vi, Pb(OAc),; viii, H+, MeOH Scheme 60 (432) (433) (434) Scheme 61 The Beecham milbemycins VM44864 and VM44866, (39b) and (39d) respectively, possessing a C-22 hydroxyl group, were degraded by a Beckmann degradation of the derived oxime (435) (Scheme 62).'69 By this process, low to G. H. Baker, R. J. Dorgan, D. 0.Morgan, R. M. Banks, S. E. Blanchflower, M. E. Poulton, and P. R. Shelley (Beecham Group plc), European Patent, EP 319 142 (1989).Davies and Green (39b) RP Me (VM 44864) (39d) R=H (VM44866) (438) P=TBDMS 01 THP Reagents: i, Bu‘Me2SiC1 (for 39d); ii, (C0C1)2, DMSO; iii, NH20H.HCI, H20, MeOH; iv, (CF3SOz)20 or p-O2N-C6H4.SO2C1, p-TsOH; v, (437); vi, H2, Lindlar catalyst; vii, H Scheme 62 medium yields of the lactone (436) were obtained which were reacted with lithium acetylides (437).’ 69*170 The resulting alkynes were semi-hydrogenated and cyclized to give novel milbemycins (438), a huge range of which are claimed in these patents. The stereochemistry of the C-24 and C-25 substituents clearly depends upon the alkyne used as starting material; the stereochemistry at the spiroacetal junction is not described in these patents although, from the Merck paper,’68 it could be assumed that this corresponds to that found in the natural material.This approach is conceptually related to the synthesis of the spiroacetal fragment recently reported by Takano, Sekiguchi, and Ogasawara (see Scheme 5).76 Treatment of 4”-oxo-5-O-t-butyldimethylsilylavermectin B 1 with samarium di-iodide at low temperature selectively removed the 3”-methoxy group.’ 71 Borohydride reduction of the product gave access to the 4”-epi-3”-desmethoxy avermectin. Other examples are provided in this patent. A final example of milbemycin interconversion is provided by the hydrolysis of the 13-iso-butanoate group of N787-182-9 (42i). This was achieved with lithium aluminium hydride in ether at -23 0C.23 17” R.J. Dorgan, D. 0. Morgan, R. M. Banks, S. E. Blanchflower, and P. R. Shelley (Beecharn Group pic), European Patent, EP 353 959 (1990). P. J. Sinclair (Merck and Co. Inc.), US Patent, US 4 897 393 (1989). Avermectins and Milbemj cins Part II B. Reduction and Photo-isomerization.-Wilkinson’s catalyst, [tris-(tripheny1phosphine)-rhodium(rj chloride has been frequently used in avermectin and milbemycin chemistry to selectively reduce the A22 double bonds A further example can be found in the reduction of 5-a~etoxy-A~~-S541 Factor A, only the C-22-C-23 double bond is reduced, with the unsaturated double bond being unaffected 172 This method has also been used to prepare [22,23-733H2]dihydroavermectin B1, of high specific activity A patent and a paper concerning the hydrogenation of avermectins and milbemycins over a supported palladium catalyst have been published by Mrozik and his co-workers 74a ’Reduction of 5-0-t-butyldiphenylsilyl avermectin B1 in ethanol over 5% palladium on charcoal at 90 pounds pressure was stopped when one molar equivalent of hydrogen had been absorbed The major component at this stage was the 22,23-dihydroavermectin After the absorption of a further mole of hydrogen the major product became 10,11,22,23-tetrahydroavermectin The patent reported that a similar reduction of 22,23-dihydroavermectin B1 at 90 p s I, when stopped after the uptake of one molar equivalent of hydrogen, gave a mixture which was estimated by HPLC to contain 23% of starting material and 46% of 10,11,22,23-tetrahydroavermectinB1, small amounts of 3,4,10,11,22,23- hexahydroavermectin B 1 were also isolated The paper, however, reports that this reduction, when performed at 20 psi with one equivalent of hydrogen, yields a 2 1 ratio of the 10,11,22,23-tetrahydro and 3,4,10,11,22,23-hexahydroavermectins B 1 The reduction of avermectin B2 over 5% palladium on charcoal in ethanol with 1 5 equivalents of hydrogen at atmospheric pressure gave d mixture containing 50% of 10,ll-dihydroavermectin B2 3,4-Dihydroavermectin B2 (30%) and 3,4,10,1l-tetrahydroavermectinB2 (1004) were also present Reduction of avermectin B1 with half an equivalent of hydrogen at 20 ps i pressure gives, in addition to starting material, a mixture of 10,ll-dihydro (20%) and 22,23-dihydroavermectins (10%) In contrast, when 13-deoxy- 22,23-dihydroavermectin B 1 aglycon was reduced with 5% palladium on charcoal in ethanol at ‘slightly elevated pressure’ and the reduction was stopped at an intermediate stage, the major product was 10,11,22,23-tetrahydroavermectinB1a aglycon (28%) A considerable amount (17%) of a further product was also isolated This was assigned the structure (439), corresponding to reduction of the A22-double bond and a 1,4-reduction of the diene unit between C-8 and C-11 (small amounts of the corresponding B2, products were also isolated) Analogues of this curious reduction product were also produced from 5-0-t-butyldimethylsilyl avermectin B 1 itself, and milbemycins x1 and x3 Simrlar reductions of 22,23-dihydroavermectin B 1 aglycon or its 5-0-t-butyldimethylsilyl derivative are described as only giving the 10,11,22,23-tetrahydroavermectin The reductions of avermectin A2 aglycon or LL-F28249P ”’0 Z Pereira and M V J Ramsay (American Cyanamid Co ) European Patent EP 346 133 (1988) G Toth J Kardos A Fodor and F Sirokman J LahelletiConipdr Radropharni 1986 24 683 1’4 (N) H Mrozik and T L Smith (Merck and Co Inc) Europedn Patent EP 266 131 (1988) (h) T L Shih H Mrozik J Ruis-Sanchez and M H Fisher J Org Ckent 1989 54 1459 (0T A Blizzard H Mrozik F A Preiser and M H Fisher Tetrahedron Lett 1990 31 4965 Davies and Green OH Averrnectin Bla .'I are similarly unexceptional and yield the 10,ll-dihydro derivatives.' 74a When 22,23-dihydroavermectin B1 was reduced at 60 p.s.i.pressure until four equivalents of hydrogen were absorbed, a 2: 1 mixture of 3,4,10,11,22,23-hexahydroavermectin B1 and the fully reduced avermectin was produced. The conclusion that the Merck workers derived from these studies was that, for palladium/carbon catalyst, the relative ease of hydrogenation for each double bond was A'' and > A3 > Ag % Al4. The interest in semi-reduced avermectins and milbemycins is not solely a result of chemical interest. Such systems should have greater photo-stability which would lead to a prolonged shelf life for commercial products, and a greater half- life in applications which involve their exposure to sunlight. Studies of the photo- stability of avermectins and semi-reduced avermectins have been performed by workers at Merck, Sharp, and Dohme who photolysed avermectin B 1, (containing 6% B 1b).The major photo-product (41%) was (8,9-Z)-avermectin B 1 a (440) (Scheme 63), a small amount of the (10,ll-2) isomer (441) (8%) was also H. Mrozik, P. Eskola, G. F. Reynolds, B. H. Arison, G. M. Smith, and M. H. Fisher, J Org. Ciirm.. 1988,53,1820. Avermectins and Mrlbemj cins Purr /I I c OTBDMS OTBDMS R = TBDMSO... R' = Bus or Pr'd. Reagents 1, Li(OMe)3AIH, CuBr, THF Scheme 64 isolated Similar results were obtained for 22,23-dihydroavermectin B 1, aglycon Time-course experiments have shown that after 20 h of ultraviolet irradiation (maximum at 300nm) only 5% of 22,23-dihydroavermectin B1 remains '74a For 10,11,22,23-tetrahydroavermectinB',under similar conditions, 750/, remains after 74a20 h irradiation, and 40 h irradiation still leaves 36% undecomposed Previous work has demonstrated that conjugate reduction of the A3 double bond of 5-0x0 milbemycin D can be performed with triethylsilane in the presence of catalytic amounts of rhodium or ruthenium,' but a more spectacular example of conjugate reduction is provided by the reduction of the lO-ox0-10,11-dihydroavermectin (442) with cuprous bromide/lithium trimethoxy-aluminiumhydride '74a This gave a 50% yield of the 8,9,10,1l-tetrahydro-lO-oxo product (443) (Scheme 64) Further examples of the reduction of the carbonyl group of various 5-0XO milbemycins with sodium borohydride have been reported 76a An alternative method, however, is to use the inherent reductase activity of various milbemycin- producing strains and their mutants 177 While hydride reduction of 5-0x0-milbemycins regenerates the natural configuration, this is not true for the 22-0x0 Beecham compounds which, with lithium tri-s-butylborohydride, give the epz isomer "* Epimerization at C-23 of the Glaxo S-541 compounds was achieved For example (a) 0 Z Pereira M V J Ramsay and S Freeman (American Cyanamid Co) European Patent, EP 307219 (1987) (h) K Sat0 and T Otsu (Sankyo Co Ltd) Japanese Patent JP62170 379 (1987) B A M Rudd and M V J Ramsay (American Cyanamid Co) European Patent EP 333404 ( 1988) G H Baker R J Dorgan, R M Banks, and M E Poulton (Beecham Group pic) European Patent EP 288 205 (1988) Duvies and Green by sodium borohydride reduction of the 23-ketones7 which yielded a 2: 1 mixture of the natural product and its epimer.'79 C.Dehydration, Dehydroxylation, and Halogenation.-The previously described '8o pyrolytic elimination of a 23-thiocarbonate group has been used to prepare A22 analogues of the Glaxo and Cyanamid S-541 and LL-F compounds. 18' Dehydration of the 23-hydroxy group of 5-acetoxy S541 Factor A, or its 5-0XO analogue,' 82 with diethylaminosulphur trifluoride, however, gave exclusively the A23 analogues.'"" A23-Avermectins A2a7 A2b, B2a7 and B2b were also prepared by DAST-induced dehydration of 23-hydroxy avermectins. '83 Burgess' reagent has also been used to prepare A23-5-acetoxy S541 Factor A.' la Reductive removal of the 23-hydroxy group of Cyanamid's LL-F milbemycins in two steps has been de~cribed.'~~",~ Conversion of the 23-hydroxy group into the corresponding bromide with triphenylphosphine dibromide was followed by radical debromination with tributyltin hydride to provide the required derivatives.A similar dehydrobromination was used to prepare 10,ll -dihydro-10-hydroxy avermectins and 10,ll-dihydro-10,13-dihydroxymilbemycins from the cor-responding 1 1-bromo corn pound^.'^^" Dehydrochlorination by this procedure was used to prepare 5-0-t-butyldimethylsilyl-13-deoxy-l0,ll-dihydro- 10-fluoro avermectin B1 aglycon from the 13-chloro- 10-fluoro compound.' 74a 23-Bronio S541 Factor A has also been dehydrobrominated by treatment with zinc in an acetic acid and isopropanol mixture.18' The reaction of the previously described 23-thiocarbonate avermectins and milbemycins with tributyltin hydride provided a ready means of preparing 23- deoxy analogues which has been widely a~plied.~~.'~~ An alternative one-pot method to prepare 23-deoxy-5-acetoxy S541 Factor A was to react the milbemycin with oxalyl chloride and then add the resulting reaction mixture to a refluxing solution of 2-mercaptopyridine-N-oxide and trityl thiol in toluene in the presence of a catalytic quantity of 4,4-dimethylaminopyridine. 87 Avermectin and milbemycin analogues dehydroxylated at C-5 have also been prepared by the reaction of tributyltin hydride with 5-halo or 5-thiocarbonate N.E. Beddall, P. D. Howes, M. V. J. Ramsay, S. M. Roberts, A. M. Z. Slawin, D. R. Sutherland, E. P. Tiley, and D. J. Williams, Tetrahedron Lett., 1988,29, 2595. lB0H. H. Mrozik (Merck and Co. Inc.), US Patent, US 4 550 160 (1979). "' (a) J. B. Ward, H. M. Noble, N. Porter, R. A. Fletton, D. Noble, D. R. Sutherland, and M. V. J. Ramsay (Glaxo Group Ltd.), European Patent, EP 215654 (1986); (b)G. Asato and S. Y. Tamura (American Cyanamid Co.), European Patent, EP 259 688 (1988). G. Asato and Z. Ahmed (American Cyanamid Co.), European Patent, EP 264 576 (1986). H. H. Mrozik and F. S. Waksmunski (Merck and Co. Inc.), European Patent, EP 326 357 (1988). I H4 (a)G. Asato and S. Y. Tamura (American Cyanamid Co.), European Patent, EP 262 384 (1988); (b) ibid., European Patent EP 280928 (1988).M. V. J. Ramsay, S. Freeman, A. H. Shingler, 0.2. Pereira, and S. C. Dolan (American Cyanamid Co.), European Patent, EP 307 221 (1987). (a) B. G. Christensen, M. H. Fisher, and H. H. Mrozik (Merck and Co. Inc.), European Patent, EP 284255 (1987); (b)M. V. J. Ramsay, P. D. Howes, R. Bell, E. P. Tiley, and D. R. Sutherland (Glaxo Group Ltd.), European Patent, EP 307220 (1989); (c) G. Asato and S. Y. Tamura (American Cyanamid Co.), European Patent, EP 280929 (1988). M. V. J. Ramsa.y and S. C. Dolan (American Cyanamid Co.), European Patent, EP 307 226 (1989). Auermectins and Milbemycins Part II analogues Examples may be found in the reaction of tin hydride with 5-chloro- 23-(E)-methoxyimino S541 Factor A,' 88 or with 5-bromo-23-deoxyavermectin B2, 189 No isomerization of the double bond is reported in these reactions The 5-halo compounds used in these dehydroxylations were prepared from the corresponding alcohols by treatment with triphenylphosphine/carbon tetrachloride 18' or from the corresponding mesylates or tosylates '89 The hydroxylation at C-13 of milbemycin derivatives by allylic bromination,' rearrangement of 15-hydroxy-A' 3-analogues (see Section 8H), or the use of avermectin aglycons, offers the possibility of introducing 13-halo substituents on to the macrocyclic skeleton by direct halogenation Both 13a- and 13P-fluoro milbemycins have been prepared by the use of diethylaminosulphur trifluoride, while the chloro analogues were prepared with thionyl chloride, bromo analogues with phosphorus tribromide, and iodo analogues with trimethylsilyl iodide In the search for a selective method of saturating the A'' double bond of avermectin B1a (5), Shih and his co-workers found that treatment with hypobromous acid (from N-bromoacetamide in aqueous acetone) regioselectively gives the unstable 11-bromo-l0,ll-dihydro- 10-hydroxy avermectin B1, (444a) (Scheme 65) as the only bromohydrin 174b When the avermectin is protected at the 5-and/or 4"-positions the reaction is equally successful and similar reactions were observed for 22,23-dihydroavermectin B1 and 5-O-t-butyldimethylsilyl-13-hydroxy milbemycins a1 and a3 1740 Reductive removal of the bromine from (444a) or (444b) with tributyltin hydride gave 10,ll-dihydro-10-hydroxyderiva-tives (445a and b), a starting point for the preparation of further derivatives Oxidation by a Swern procedure gave 10-0x0 derivatives,' 740 while 10-fluoro derivatives were prepared using diethylaminosulphur trifluoride '74b Reaction of 4',5-diprotected 10,ll-dihydro-10-hydroxy avermectin B1 (445b) with triphenylphosphine/hexachloroacetoneat 0 "C in THF gave the corresponding 10-chloro analogue (446a),' 74a but the use of triethylamine/dichlorotriphenyl-phosphine in methylene chloride on the 5-0-TBDMS protected avermectin (44%) gave a mixture of the expected 10-chloro compound (446b) and the allylically rearranged 8-chloro-8,ll -dihydro compound (447) '740 Hydrode-chlorination of this mixture gave a 2 1 mixture of 10,11- and 8,11- dihydro- avermectins '74 Mixed halo derivatives of these reduced milbemycins have been prepared from the avermectins by preparing the 10-halo aglycons and halogenating further 13-Chloro-10-fluoro and 10,13-difluoro analogues have been prepared in this way, and 13-deoxy- 10,ll-dihydro- 10-fluoro avermectin B 1 aglycon was prepared by tin hydride dehydrochlorination of the 13-chloro- 10-fluoro analogue ' 74 M V J Ramsay, D N Evans, D R Sutherland, E P Tiley, J B Ward, N Porter R A Fletton, and D Noble (American Cyanamid Co ), European Patent, EP 327 270 (1988) lS9 W Roben, W Stendel, and P Andrews (Bayer AG) European Patent EP 303 933 (1987) 190 (a)B Frei, A C 0Sullivan, and P Maienfisch (Ciba-Geigy AG), European Patent, EP 180 539 (1985), (6)K Sato, T Yanai, N Kitang, A Nishida, B Frei, and A C O'Sullivan (Sankyo Co Lrd ) European Patent, EP 203 832 (1986) Davies and Green Avermectin Bla (5) I -or 4".5-bis-OTBDMSAvermectin Bla ""?"r} (444a;4", 5-bis-OH) (444b;4", 5-bis-OTBDMS) ?$} (447;4"-OH, 5-OTBDMS) (445a;4", 5-bis-OH) + (445b;4",5-bis-OTBDMS) 4"-OH, 5-OTBDMS)(44%;4"-OH, 5-OTBDMS) (445~; lv (446a;4", 5-bis-OTBDMS) (446b;4"-OH.SOTBDMS) Reagents: i, HOBr; ii, Bu",nH; iii, Bu'Me2SiCl, DMAP, CH2CIZ; iv, Ph3PC12, Et,N, CH2C12; V, Ph3P, CC13COCC13, THF Scheme 65 D. Acylation and Protection.-Methods of selectively protecting and revealing the separate hydroxyl groups in avermectins and milbemycins were carefully detailed by early workers in this field,' and no radically new methods of protection have been reported.Further examples exemplifying extant knowledge are, however, available. Thus the easiest site to protect is the allylic C-5 hydroxyl group, which can be selectively silylated in the presence of all other hydroxyl groups. Examples can be found in the 5-silylation of avermectin B1, mon~saccharide,'~~10-hydroxy-10,11,22,23-tetrahydroaverrnectin B 1 aglycon,' 74a and the Glaxo/Cyanamid milbemycins, which can be selectively silylated at C-5 with t- butyldimethylsilyl chloride or trimethylsilyl chloride under mild conditions. '7b*1 ' Avermec tins and Milhem) em Part II Many different acyl groups have been introduced to the avermectins and milbemycins in a search for more active analogues or to improve bio-dynamic parameters An exhaustive catalogue of these would take many pages, and thus this review will only detail selected examples to illustrate any new chemistry Further examples of selective acetylation at C-5 in the presence of a free C-23 hydroxyl group are available from work on the Glaxo/Cyanamid milbemycins 17' 19' 192 The acetylation of both the 5-and 23-hydroxyl groups of S541 Factor A requires excess acetic anhydride, pyridine as solvent, and 4,4- dimethylamino pyridine as catalyst for extended periods I 7h 23-acetoxy Factor A can then be prepared by mild hydrolysis of the 5,23-diacetate 17' An alternative method is to acetylate the 5-0-TBDMS milbemycin and then remove the silyl group 193 Acylation of milbemycin D with N-(trichloroethoxycarbony1)-glycine and dicyclohexylcarbodiimide, unsurprisingly, is also selective in acylating at C-5 94 DCC has also been used to prepare other 5-0-acetyl substituted milbemycins such as 5-0-(1,2,4-triazol-l-yl)acetyl-l3-methyland 13-vinyl milbemycin r3 19' Milbemycins and avermectins have been acetylated at C-4a with a range of acid chlorides, using pyridine as a catalyst 25 Among the compounds thus prepared from 4a-hydroxylated 5-0-TBDMS milbemycin ~3 were milbemycins x11, x13, and ~14,which had previously been isolated as fermentation products (see Section 2) Aryloxy and alkyloxy-acetate analogues at C-5 of mil bemycins were prepared by simple acylation at O"C, using the acid chloride with 4,4-dimethylaminopyridine and di-isopropylethylamine or pyridine as catalyst 19' 193 195' Thus LL-F28249a gives 5-methoxyacetoxy-LL-F282496: at O"C, whereas repetition of the reaction at 80°C for 20h gives the 5,23-bis- acetylated compound '93 Mono-acyloxy derivatives at C-23 of LL-F compounds were prepared by acylation at 80°C of C-5 protected derivatives with an acyl halide in the presence of di-isopropylethylamine 93 Acetylation with chloroacetyl chloride provides active analogues and also gives a further handle for manipulation Thus, 5-O-(chloromethoxycarbonyl) milbemycin D has been shown to be active as an anthelmintic agent against Asearis suum larvae 5-Iodoacetyl milbemycin D was prepared from the chloro derivative by a Finkelstein reaction, and the iodide was displaced by azide, 1,2,4- triazole, and imidazole to give the corresponding 5-azoacetyloxy milbemycins I 96 A similar displacement with 4-methyl-imidazole gave a separable mixture of 5-0-(4 -and 5'-methylimidazol-l'-yl)-acetyl milbemycins D, but it could not be Iy' J B Ward H M Noble N Porter R A Fletton D Noble D R Sutherland and M V J Ramsay (Glaxo Group Ltd ) British Patent GB 2 176 182 (1986) 19' G Asato (American Cyanamid Co ) European Patent EP 259 668 (1986) 19' G Asdto (Americdn Cyandmid co ) European Patent EP 259686 (1986) IY4A Terada S Naruto R Matsueda H Rei K Susumu and K Kitano (Sankyo KK) Japanese Patent JP 60/184085 (1985) (a)P Maienfisch and A C 0Sullivan (Ciba Geigy AG) British Patent GB 2 187453 (1986) (h) P Maienfisch (Ciba Geigy AG) European Patent EP 282456 (1988) 196 E Sturm and P Maienfisch (Ciba Geigy AG) European Patent EP 184989 (1985) 310 Dames and Green ascertained with certainty which isomer was which.An extensive list of such aza-heterocyclic milbemycin D, x1 and x3 analogues prepared in this way is described in this patent. 5-0-Chloroacetyl milbemycins have also been converted into other analogues such as 5-O-(fluoroacetoxyacetyl)-and 5-0-[(S)-2- hydroxy- propionyloxy]acetyl milbemycins 313 by simple displacement reactions in DMF solution. '97 The C-23-chloroacetate of S541 Factor D has also been prepared and similarly transformed into the iodoacetyl analogue.'7b This was used as a precursor for the 23-azidoacetate and the 23-glycylester.17' This paper also describes the prepara- tion of the 23-trichloroethylcarbonates of S541 Factor D and 5-acetoxy Factor A which were converted into the corresponding urethanes (OCONHOMe). Now that 13-hydroxy milbemycins are available (see Section 8H), their reaction to give 13-acyloxy milbemycins,' 98 substituted alkyl and aryloxyacetyl milbemycins,' 99 and 13-carbonate esters 2oo by previously detailed methods has been described. The well-known reaction of isocyanates with alcohols has been applied to 13-hydroxy milbemycins D, x1 and x3 to provide a range of 13-carbamoyloxy milbemycins for testing.201 5-Phosphate esters of avermectins, and their corresponding acids, have been described previously by workers at Merck,' and similar analogues of the S541 compounds have been prepared by Glaxo workers.202 E.Oxidation and Derivatives of Oxidized Products.-S541 Factor A, its 5-acetate and its TBDMS ether, have been converted into their 4a-hydroxy analogues by the previously described method ' utilizing selenium dioxide and t-butyl hydroperoxide as oxidant at room temperature.203 Acylation of such 4a-hydroxy milbemycins with the appropriate acid chloride was used by workers at Sankyo to prepare synthetic samples of the milbemycins a1 1, ~13,and a14 [(43a), (43g), and (43c) respectively] 25 which had been obtained by fermentation (see Section 2). Some selectivity could be obtained in the acetylation of the new hydroxy group at C-4a in the presence of a free 5-hydroxy group by the use of triphenyl phosphine/diethyl azodicarboxylate/acetic acid.203 Several further examples of milbemycins being selectively oxidized at the allylic C-5 hydroxyl group with manganese dioxide to give the corresponding x,P-This hasunsaturated ketones have been des~ribed.'~~,~~~ an important ")'P Maienfisch and E Sturm (Ciba-Geigy AG), European Patent, EP 217 742 (1985) (a) B Frei, H B Merevala, A C O'Sullivan, K Sato, and T Yanai (Ciba-Geigy AG), British Patent.GB 2 168 345 (1986), (b)B Frei (Ciba-Geigy AG), European Patent, EP 253 767 (1988) 199K Sato, T Yanai, T Kinoto, T Toyama, K Tanaka, A Nishida, B Frei, and A C OSullivan (Sankyo Co Ltd ), European Patent, EP 246 739 (1987) P Maienfisch (Ciba-Geigy AG), European Patent, EP 245 209 (1987) 'O' P Maienfisch (Ciba-Geigy AG), European Patent, EP 239 528 (1987) 'O' M V J Ramsay, E P Tiley, 0 2 Pereira, J B Ward, N Porter, H M Noble, R A Fletton, D Noble.and D R Sutherland (Glaxo Group Ltd ). European Patent, EP 238 259 (1986) '03 J B Ward, N Porter, H M Noble, R A Fletton, D Noble, and D R Sutherland (Glaxo Group Ltd ), European Patent, 237 341 (1987) '04 D R Sutherland, N Porter, B M Bain, M V J Ramsay, H M Noble, E P Tlley, R A Fletton, J B Ward and D Noble (Glaxo Group Ltd ), European Patent, EP 238 258 (1987) 31 1 Avermectins and Milbemycins Part II OH Pri -i-iii OH (344 Reagents: MnOz, Et20; ii, Me2NSF3, (MeOCH2)2; iii, NaBH4 Scheme 66 -HPPri -Pr' -Pri Reagent: i, MeONHz Scheme 67 consequence for the Glaxo S541 Factor A as the resulting ketone is readily crystallized and purified, an important consideration in a series which is generally non-crystalline and hence more difficult to handle on an industrial scale.Many oxime, hydroxylamine, semicarbazide, and hydrazone derivatives of the 5-ketones have been prepared.'99*204*205 As the C-5 alcohol is the easiest position to oxidize, this can also be used as a means of protecting this alcohol in order to carry out a conversion elsewhere in the molecule; reduction with sodium borohydride then regenerates the correct stereochemistry at C-5. An example of this strategy can be seen in the synthesis of A23-LL-F28249a (Scheme 66).'82 Both hydroxyl groups at C-5 and C-23 of the milbemycin analogue S541 Factor A have been oxidized with Jones reagent in a two-phase mixture of ethyl acetate and water using tetra-n-butylammonium sulphate as phase-transfer catalyst.'06 When the resulting 5,23-diketone was treated with one equivalent of methoxylamine hydrochloride, the 23-E-methoxyimine was by far the major product.This accelerated reaction at C-23 was considered to be a consequence of intramolecular hydrogen bonding between the intermediate hemi-aminal hydroxy group and the oxygen atom at C-17 (Scheme 67), the exclusive formation of the E isomer of the oxime was thought to be due to control of the loss of water by 205 (a) Sankyo KK, Japanese Patent, J6 3045 281 (1987); (b) B.Frei and A. C. O'Sullivan (Ciba-Geigy AG), European Patent, EP 279 783 (1987). 206 S. Freeman (American Cyanamid Co.), European Patent, EP 307 222 (1989). Davies and Green -i.ii OMe Reagents: i, PhC002Bu1,CuCI; ii, AcOH, HzO Scheme 68 the adjacent C-24 group.' 79 A range of 5-0x0-23-alkoximes have also been prepared. '7a Avermectins hydroxylated at C-8a have been isolated as metabolites from soil ' and can be prepared in the laboratory by oxidation with t-butyl peroxybenzoate/cuprous chloride 207 (Scheme 68). 8a-0x0 avermectins (448) are now synthetically available by treating the suitably protected natural compounds with an excess of pyridinium dichromate in DMF at room temperature.207 Reduction of the 8a-ox0 or 8a-hydroxy avermectins with sodium borohydride/cerium chloride gives the diol(449) (Scheme 69).Oxidation of 4",5-di-O-t-butyldimethylsilyl-l0-hydroxyavermectin B1 to the corresponding 10-ketone by a Swern procedure has been described.' 74a In contrast to the above described hydroxylations at C-4a, reaction of selenium dioxide with 5-0x0-milbemycins a2, a4,and D, or 5-0~0-LL-F28249a, in formic or acetic acid gave milbemycins which were acetoxylated at C-1 3.208 Hydrolysis of these crude products gave 13-hydroxy milbemycins in yields greater than 40%. This oxidation can equally well be performed on the 5-~ximes.~'~A curious result has been reported for the selenium dioxide/t-butyl hydroperoxide oxidation of 23-desoxy-S541 Factor A 5-acetate; by analogy with 207 (a) H.H. Mrozik and F. S. Waksmunski (Merck and Co. Inc.), US Patent, US 4547491 (1985); (b) M. H. Fisher, M. J. Wyvratt, and H. Mrozik (Merck and Co. Inc.), European Patent, EP 343715 (1 989). (a)K. Sato, T. Yanai, T. Kinoto, and S. Mio (Sankyo Co. Ltd.), European Patent, EP 184308 (1986); (h)Y. Tsukamoto, K. Sato, and T. Yanai (Sankyo KK Ltd.), Japanese Patent, 587-210805 (1987). '09 T. Yanai, K. Sato, and T. Otsu (Sankyo KK Ltd.), Japanese Patent, JP 860425 (1986). Auermectins and Milbemycins Part II (449) nu-oA+H OTBDMS (448) Reagent7 1, NaBH4, CeC13 Scheme 69 previous work this would be expected to produce a 4a-hydroxylated milbemycin, yet the patent claims the product to be (13R)-hydroxy-23-desoxy Factor A 5-acetate, albeit obtained in low yield (ca.12%).'86b The oxidation of avermectin B aglycones (i.e. 13-hydroxy-milbemycin type structures) with chromium trioxide in the presence of 3,5-dimethyl-pyrazole, or under Swern conditions, to give 13-oxo-derivatives, and their conversion into 13- imino and 13-amino derivatives, has been described.2'o*2'' Th ese 13-0x0 derivatives have been converted into the 13-methylene compounds on treatment with Tebbe's reagent in toluene at -40 cC.212 The initial step in the Beecham degradation of the outer spiroacetal ring (see Section 8A) utilizes the oxidation of the hydroxyl group at C-22 by a Swern procedure.' 78 Various oximes were thus prepared of VM44864 and 5-0-TBDMS VM44866.Swern-type oxidation has also been used to prepare 23-0x0 analogues of suitably protected 13-deoxy avermectin aglycones 2' and S541 type milbemycins; pyridinium dichromate and chlorochromate have also been used.' 79,214 The formation of methoxyimines of these 23-0x0 milbemycins was highly stereospecific providing only the E isomer in 80% yield. The same product was obtained from the corresponding 5,23-diketone in 54% yield, only trace amounts of the isomeric 5-methoxyimine being formed. As previously described, hydrogen bonding in the transition state between the 17-oxygen atom and the transient 23-hydroxy group was postulated to explain the preferential functionalization at C-23. The preferential formation of the E imine was thought to be due to the equatorial C-24 methyl group controlling the direction of 210 M V J Ramsay, P D Howes, R Bell, E P Tiley, and D R Sutherland (American Cyanamid Co ), European Patent, EP 307 220 (1987) 211 B 0 Linn and H H Mrozik (Merck and Co Inc ), European Patent, EP 165029 (1985) 'I2 P Maienfisch dnd M Riediker (Ciba-Geigy AG), German Patent, DE 3 631 387 (1986) 'I3 G Asato and D J France (American Cyanamid Co ), European Patent, EP 260537 (1988) 'I4 G Asato and D J France (American Cyanamid Co ), European Patent, EP 259779 (1988) Davies and Green (450) R = Pr' or Bus elimination of water from the aminal intermediate.Many other oximes, imines, and semicarbazones of 23-0x0 milbemycins and avermectins have been prepared as the resulting compounds are frequently very active.2 '4921 Few medicinal chemists, when presented with a double bond, can resist the temptation of preparing an epoxide! There are many reasons for this: preservation of overall molecular shape, change in electronic characteristics, and as a synthetic handle.Avermectin and milbemycin chemists have not been immune to this temptation. Earlier work has shown that epoxidation of milbemycins or avermectins under Sharpless conditions selectively provides 8,9-epoxy derivatives, directed by the 7-hydroxy group;' this was the basis of the degradative procedure utilized by Smith and Thompson (see Section 8A). A further example of this reaction can be found in the epoxidation of 10,11,22,23-tetrahydroavermectinB1 with slightly over one equivalent of t-butyl hydroperoxide to give the 8,9-epoxide (450).'740 When the 7-hydroxy group is blocked by silylation, Sharpless epoxidation gives the 3,4-oxide (451) in moderate yield.'74c As the free 5-hydroxy group was presumed to direct the epoxidation, the stereochemistry of the epoxide group was assigned as p; this was confirmed by NMR analysis of the product obtained by thiophenol opening of the epoxide (452) (Scheme 70). Reaction of the epoxide (453) with anisylethylamine in methanol, however, did not yield products derived by ring opening of the epoxide (Scheme 71). The major product was the conjugated ally1 alcohol (454) while the minor product was the corresponding seco-ester (455).Replacement of the secondary amine with DBU in this reaction resulted in a reversal of the yields of the two products. An interesting acid-catalysed rearrangement of 4",5,7-tris-O-trimethylsilyl avermectin B1 8,9-oxide (456) has been reported by Blizzard and his co-workers (Scheme 72).'65 Upon treatment with p-toluene sulphonic acid in wet *I5 (a)D. R. Sutherland, M. V. J. Ramsay, E. P. Tiley, 0.Z. Pereira, J. B. Ward, N. Porter, H. M Noble, R. A. Fletton, and D. Noble (Glaxo Group Ltd.), British Patent, GB 2 192630 (1987); (h) G. Asato and D J. France (American Cyanamid Co.),European Patent, EP 250 536 (1988). Avermectinsand Milbemycins Part II 2 steps I_f OH OH Scheme 70 tetrahydrofuran this epoxide gave 60-70% of the novel tricycle (457). The presence of the 7-0-trimethylsilyl group was indicated as necessary to the success of the reaction, as a similar reaction on the non-silylated epoxide only gave the expected 8,9-diol.4",5,7-Tris-O-trimethylsilylavermectin B 1 itself only gave desilylated avermectin B 1 under these conditions. Preparation of 14,15-epoxides of the milbemycins by reaction with one equivalent of m-chloroperoxybenzoic acid at ambient temperature is a well established procedure for milbemycins,' and the resulting epoxides have proved of great utility in the preparation of novel avermectins and milbemycins (see Section 8H). The Glaxo S-54 l/Cyanamid LL-F milbemycins, however, contain an additional exocyclic, double bond with approximately the same reactivity as the 14,15-double bond.This is shown by treatment of S541 Factor B (34d) with one equivalent of m-chloroperoxybenzoic acid at 0 "C-r.t., which gave an (approximately) equal amount of 14,15- and 26,27-epoxides C(458a) and (458b) respectively, R = Me, R' = Similar treatment of S541 Factor A gave a mixture of 14,15- (458a) and 26,27-epoxides (458b), and 14,15,26,27-bis-epoxide (458~).~~~,~"Some selectivity for the epoxidation can be obtained by reducing the reaction temperature; epoxidation of 5,23-di-O-TBDMS LL-F28249a, initially at -70 "C and warming by stages to -20 "C, gives a 4: 1 mixture of the 26-epoxide [(458b) R = R' = TBDMS] and bis-epoxide [(458c) R = R' = TBDMS].lg6' Deprotection of the 26-epoxide and re-protection with one equivalent of t-butyldimethylsilyl chloride gave the 5-0-TBDMS epoxide [(458b) R = TBDMS, R' = H] which was oxidized with pyridinium chlorochromate to the 23-0x0 analogue, from which a range of oximes and imines were prepared.21 2'6 J B Ward, N Porter, H M Noble, R A Fletton, and D Noble (Glaxo Group Ltd), European Patent, EP 241 146 (1987) 217 G Asato and S Y Tamura (American Cyanamid Co ), European Patent, EP 280929 (1988) 218 G Asato and S Y Tamura (American Cyanamid Co ), European Patent, EP 293 549 (1987) Davies and Green (455) I -+ Reagenfs:i, DBU, MeOH or p-Et-NH*CH2*C6H4.0Me Scheme 71 Oxidation of S54 1 Factor B with two equivalents of rn-chloroperoxybenzoic acid at room temperature, predictably, gave the bis-epoxide C(458c) R = Me, R’ = The exocyclic 13-epoxide (459) has been prepared by reaction of 13-0x0 milbemycins with dimethylsulphonium meth~lide.’~’~ 13-Formyl milbemycins were obtained from this epoxide by acid-catalysed rearrangement with dilute hydrofluoric acid or camphor-sulphonic 25-( 1-Methylthioethyl) avermectins, prepared by directed biosynthesis (see Section 2), have been converted into 25-ethenyl avermectins (460) by rn-chloroperoxybenzoic acid oxidation to the sulphoxide and thermal elimination at high temperat~re,~” and these have been subjected to Wacker oxidation (Scheme 73).25-Ethenyl avermectin A2 [(460) R = Me, X = CH2, Y = CHOH], when treated with palladium chloride and cupric chloride and stirred vigorously B.J. Banks and M. J. Witty (Pfizer Ltd.), European Patent, EP 335 541 (1989) Avermectins and Milbemycins Part II HAOH (456) (457) Scheme 72 OR’ Pri OR (458) (a) X=O,Y =bond (b) X = bond, Y = 0 (c) X=Y=O in the air, gave a 50% yield of 25-acetyl avermectin A2 [(461) R = Me, X = CH2, Y = CHOH].’’’ A similar reaction of 25-ethenyl avermectin B1 [(460) R = €3, B. J. Banks and M. J. Witty (Pfizer Ltd.), European Patent, EP 340932 (1989). Davies and Green + Reagents: i, H20, DMF, CuC12, PdC12,Oz Scheme 73 Table 1 Avermectin produced Avermectin,fed Al, A2, B1a B2a B2,AG B2,MS '41, 100 A2a 93 B1a 2.7 93.6 B2, 29.7 68.3 B2, AG 9.3 31.6 46.3 5.6 B2, MS 18.6 45.2 32.2 AG = Aglycon.MS = Monosaccharide XY = CHSH), quenched with methanol, gave a mixture of 25-acetyl avermectin B1 [(461) R = H, XY = CH=CH], 25-formylmethyl avermectin B1 [(462) R = H, XY = CH=CH], and their corresponding acetals. When the Wacker procedure was applied to 25-( 1-methylbut-3-enyl) avermectin A2 (463) (also prepared by directed biosynthesis) the ketone (464) was obtained (Scheme 74). The 25-ethenyl avermectin B1 [(460) R = H, XY = CH=CH] was also converted into the 1,2-diol (465) (Scheme 75) and thence the 25-epoxide (466), while Lemieux-Johnson oxidation of 25-ethenyl avermectin B1 gave the 25- formyl avermectin (467) and, by further oxidation with pyridinium dichromate in the presence of methanol, the 25-methoxycarbonyl avermectin (468).220 Microbiological hydroxylation of 22,23-dihydroavermectin B 1, aglycon with Cunninghamella Blakesleeana gave avermectins mono-hydroxylated at C-12a and Avermectins and Milbemycins Part I1 Reagents: i, H20, DMF, CuC12, PdC12,02 Scheme 74 (460) ii-\%OH -(465)liii Reagenrs: i, OsO,; ii, TsCI; iii, NaI04;iv, PDC, DMF, MeOH Scheme 75 C-24, or in the C-24 and C-25 side-chains (Scheme 76).221 Hydroxylations of milbemycin analogues with Cunninghamella Blakesleeana (ATCC 8688a) gave a similar spectrum of products, but additional dihydroxylated products were formed with a hydroxyl group at C-14a and with a P-hydroxy group introduced into the C-13 position to produce a 13-epi-avermectin.Hydroxylations of avermectin B 1, and 22,23-dihydroavermectin B1, with Nocardia autotrophica (ATCC 35203) were more specific, producing compounds mono-hydroxylated only in the C-27 position, e.g.(469).222 Avermectin Blb gave a 26-hydroxy '"R. T. Goegelman, E. S. Inamine, and R. F. White (Merck and Co. Inc.), European Patent, EP 194 125 (1986); US Patent 4666937 (1987). '*'R. T. Goegelman, E. S. Inamine, and R. F. White (Merck and Co. Inc.), European Patent, EP 212867 (1985). 3 20 Davies and Green X = positionsof hydroxylation by C.Blakesleeana (469) R’ = Me*CH(OH);A2 = H (470) R’= Me;R2= OH Scheme 76 avermectin (470) under these conditions. Different specificity was shown by Streptornyces bikiniensis MA 5853 which hydroxylated 22,23-dihydroavermectin B1, aglycon at C-13, C-23, and C-24a to give (471).223 Milbemycins can be hydroxylated at C-14a with S.nigricans NRRL12479, S. racernosurn IF04287,48 14 or 4828 and Rhizopur circinans ATCC 1225224 while Arnycolata autotrophica FERM P-6182 converts the C-12 methyl group of milbemycins into a hydroxymethyl gro~p.~~~,~~ When cultured with Arnycolata autotrophica FERM P-6183, milbemycin a3 gave the three compounds (472a- c)226 while a P450 enzyme isolated from S. carbophilus SAND 62585 gave 13-hydroxy milbemycin a3.227 As such a one-step method of production of 13-223 Merck and Co. Inc., Japanese Patent, JP 62029 589 (1987). 224 K. Nakagawa, Y. Tsukamoto, K. Sato, and A. Torigata (Sankyo KK), Japanese Patent, JO 1 199 591 (1988). 225 I. Yamamoto and Y.Sugino (Sugiyo KK), Japanese Patent, JO 1 243 966 (1988).226 S. Nakagawa, A. Torigata, K. Sat0 and H. Kajino (Sankyo KK) Japanese Patent J 63 264484 (1987). ”’T. Matsuoka and S. Miyakoshi (Sankyo KK), European Patent, EP 281 245 (1988). Auermectins and Milbemycins Part II OH HO (471) (472) (a) R’ = CH~~H,R~ = ~e (b) R’ -CHPH, R2 -CHPH (c) R’ -CO,H, R2 .IMe Table 2 Added sohent 7;Conversion 130-OH % 14,lS-Epoxide nil 37 70 30 Acetone 98 77 23 Acetonit rile 57 60 40 t-Butyl alcohol 71 63 37 DMF 98 64 36 DMSO 97 74 26 Ethyl alcohol 100 59 41 hydroxy milbemycins would be preferred to the multi-step chemical route, workers at Ciba-Geigy have studied this bio-transformation in some depth 228 From initial screening of eleven micro-organisms S violascens ATCC 31 560 was chosen and shown to produce, at 37% conversion, a 7 3 mixture of 13-hydroxy milbemycin 5~4and the 14,15-epoxide of milbemycin x4 from milbemycin ~4 Separate fermentations of S violascens with 14,15-epoxy milbemycin a4 and 15-hydroxy milbemycin a4 (see Section 8H for preparations) produced no 13-hydroxy milbemycins, implying that the epoxide is not an intermediate in the formation of the 13-alcohol and that it may be possible to adjust the conditions of the fermentation to maximize the production of the alcohol Adding 5% vjv organic solvents to the fermentation had a marked effect on the ratio of products (see Table 2), with optimization of the fermentation medium seven days incubation gave 92% of 13-hydroxy milbemycin and 8% of 14,15-epoxide at 91% conversion Lipophilic analogues of milbemycin a4 such as the 5-acetate, 5-0- silyl ether or 5-oxime were not hydroxylated under the original conditions and the 5-ester (473) was only hydroxylated after the ester group had been hydrolysed off Addition of 5% DMSO to the fermentation medium of 5 oximtno milbemycin (x4 resulted in a 35% yield of the 13-hydroxy compound but had no effect on the fermentations of the 5-acetoxy or 5-0-TBDMS analogues ’’’(u) G M R Tombo 0 Ghisalba H P Schar B Frei P Maienfisch and A C 0 Sullivan Agrrc Biol Chm 1989 53 1531 (h) G Ramos 0 Ghisalba H P Schar B Frei P Maienfisch and A C 0Sullivan (Ciba Geigy AG) European Patent EP 277916 (1988) Dauies and Green X ..-.OCOCHzOCO(CHJ2CO2H (473) X = H, OH Or 0 (474) OH OH (475) R' = H; R2 = H; R3 = CH20H (479, (476) R' = OH; R2= H; R3 = CHzOH LacriminA (477) R' = H; R2 = OH; R3 = Me (478) R' = H; R2 = H; R3 = COZH Milbemycins r3 and D were also converted into mixtures of 13-hydroxy and 14,15-epoxides, but in these cases the epoxides were the major product. S541 Factor A or its 5-0x0 analogue, when incubated with S. pfatensis ssp. mafuinus NRRL 3761, gave the corresponding milbemycin 12-carboxylic acid (474). The 12a-hydroxymethyl milbemycin could also be isolated, leading to the supposition that this must be an intermediate in the oxidation process.229 Other Streptomyces strains, S. mashuensis ISP 5221 and S.rimosus NRRL 2455, hydroxylated S-541 Factor A to give a 12-hydroxymethyl analogue; the latter strain also produced 4a, 12a-dihydroxy Factor A while Absidia cyfindrospora NRRL 2796 produced the tertiary alcohol 12-hydroxy S541 Factor A.229When a different fermentation medium was used with this micro-organism two metabolites from S541 Factor A were isolated which were mono-and di-hydroxylated in the C-25 side-chain, (475) and (476).230 S. eurythermus ISP5014 also gave the mono-hydroxylated product (475) in addition to the tertiary alcohol (477), while S. auermitifis ATCC 31272 gave the carboxylic acid (478) in addition to (475) and the 5-methoxylated analogue of (475).230 229 M. J. Dawson, D. Noble, G. C. Lawrence, R. A.Fletton, S. J. Lane, M V. J. Ramsay, 0.Z. Pereira, D. R. Sutherland, and E. P. Tiley (American Cyanamid Co.).European Patent, EP 341 974 (1989). 230 G. C. Lawrence, M. J. Dawson, D. Noble, R. A. Fletton, and S. J. Lane (American Cyanamid Co.), European Patent, EP 345 078 (1989). Avermectins and Milbemycins Part I1-'24-'b1% 0, 0' 0. H H HOTBDMS OTBDMS OTBDMS (480) Reagent 1, AcOH Scheme 77 OMe (482) (a) R' = TBDMS; TMS @) R' = TBDMS; R2= H F. Conjugation and Deconjugation at C-2.-Previous work has shown that avermectins and milbemycins are inherently unstable to strong base.' Sankyo, however, have serendipitously found that the aqueous base-induced decomposi- tion of milbemycin (24) at a pH > 12 gives a compound which is an anti- hypotensive agent.231 This compound, given the name Lacrimin A (479), has recently been synthesi~ed.~ 32 Avermectins react with hydroxide ion in aqueous methanol to give two major products.' Initially, epimerization at C-2 occurs followed by movement of the A3 double bond into conjugation to give a A2 avermectin.This facile conjugation of the hydrindene double bond has recently been a cause of some discussion. In Hanessian's synthesis of avermectin Bl, (5)' the last step involves a deconjugation of the double bond to give, reportedly, the correct isomer of the required product in good yield. This was rationalized as a topside proton delivery to the ketene acetal derivative (480) under carefully controlled conditions (Scheme 77).94995As Fraser-Reid had experienced difficulties with trying to prevent double bonds 231 H Takiguchi, J Ide, H Koike, and M Terao (Sankyo Co Ltd), Japanese Patent, JP 58-69886 (1983) 232 A Takle and P Kocienski, Terrnhedron Leir ,1989,30,1675 Davies and Green moving into conjugation in intermediates such as (481), he determined to study this facile deconjugation.lS7 Thus he prepared trisilylated A2 Ivermectin (482a) and attempted the deconjugation under a variety of conditions. The only products isolated, however, were deconjugated 2-epi Ivermectin and the 7-hydroxy avermectin (482b). These experiments were then repeated with silylated A2 avermectin B1, in case the additional C-22-C-23 double bond in avermectin B 1,provided a subtle conformational change which favoured the deconjugation reaction.Yet again, however, a wide range of conditions only sufficed to produce 2-epi avermectin where deconjugation had taken place. As this appears to be the kinetic product an attempt was made to epimerize at C-2 under protic conditions (NaOH/MeOH/H20). For the trislylated 2-epi avermectin Bl,, this gave none of the required natural product. Significantly, however, unprotected 2-epi avermectin B1, did give 25% of the natural avermectin B1a under these conditions. Fraser- Reid attributes this to the polarity of the reaction medium and the substrate. An additional factor may be that hydrogen bonding between the macrolide and the solvent encourages the formation of the correct isomer.Such a hydrogen bonding network has been shown to be important in the crystal structure.’ Later in the same year, Hanessian, Dube, and Hodges published a further paper’ 56 on the last stages of their avermectin B1, synthesis. Subsequent work had revealed that the deconjugation step was ‘prone to unpredictable variation in the nature of the products, even with the slightest change in reaction conditions, scale of operation, or mode of work-up’ and that ‘the material produced in our original deconjugation was not the primary product of deconjugation, but possibly the result of a subsequent epimerization of an initially formed 2-epi isomer’. They thus determined to produce the 2-epi isomer deliberately and then epimerize at the 2-position in a subsequent step.Addition of 4”,5,7-tris[O- (trimethylsilyll-A2-4(R)-avermectinB 1, to lithium diethylamide and trimethylsilyl chloride followed by rapid quenching with aqueous acid gave a high yield of the 2-epi derivative which was subjected to various conditions in order to epimerize the C-2 position. While a variety of conditions, both protic and non-protic, were successful, the best conditions were found to be with imidazole in benzene under reflux. This gave a 40% yield of the required natural avermectin, 8% of the A2 isomer and 34% of the starting material which could be recycled through the process to enhance the overall yield. A similar experiment on unprotected 2-epi avermectin B1a gave 40% of avermectin B1, along with 34% recovery of the 2-epi starting material.That this process was a true equilibrium was shown by subjecting avermectin B1, to the same process when an identical ratio of products was observed. The final stage of Danishefsky’s synthesis of avermectin Ala aglycon also involved a deconjugation of the A2 double bond.97 Conversion of the synthetic, conjugated, aglycon (372) into the C-2 epi-isomer was followed by epimerization with imidazole in benzene to yield 32% of the correct isomer of avermectin Al, aglycon admixed with 33% of the 2-epi avermectin and 31% of the A’ isomer. The facile conjugation of the A3 double bond and consequent loss of stereochemical integrity at C-2 poses problems at a late stage of synthesis which Avermectins and Milbemy ins Part 11 8 stepsAvermctin Bla --(5) H bMOM (484) X=OH (485) X=SePhbH I OCOCHpPh /K(489) OMOM Me0 Me0 (487) -0 -O -P--R-Ho---P-.Reagent5 I, Se02, Bu‘OOH, 11, H202, pyndme, 111, DCC, DMAP, IV, N-(phenylselenophthalimide), BuYP, CH2CI2, v, PhOCH2COCI, VI, H202, pyndme, VII, CH3S02CI,viii, LiBr, IX,NaBH4, DMF, x, MeOH, NH3 Scheme 78 are best avoided The solution to this problem lies in ‘parking’ the double bond in an exocyclic position at C-4, as has been described by Fraser-Reid and his co- workers in a carefully considered, and well written paper (Scheme 78) 233 Two possible problems in this approach needed to be assessed (I) can seco-acid 233 B Fraser-Reid, J Barchi, and R Faghih, J Org Chm , 1988,53,923 Duvies and Green intermediates still be macro-lactonized, and (ii) can the double bond be moved between the A394 and 434a positions without any problems of C-2 epimerization or conjugation.Answers to these questions were sought by transformations of avermectin B1, (5). Allylic oxidation of (5) at C-4a followed by cleavage of the macrocyclic lactone with lithium aluminium hydride and protection gave the alcohol (484). This was converted into the selenide (485), which was oxidized and allylically rearranged to the model C-3 alcohol (486). Deprotection and oxidation of the primary alcohol at C-1 gave a 3-hydro~y-A~.~"-seco-acid(487) which was successfully lactonized to reform the macrolide ring (488). The exocyclic double bond was then readily isomerized to the A3 position by a sequence of sulphonation and brominolysis.Debromination with sodium borohydride in dimethylformamide and deprotection then gave back avermectin Bl,. Selenylation of unprotected 4a-hydroxy avermectin B1a (483) with the Nicolau reagent specifically gave the required 4a-selenide, which was oxidatively rear- ranged and protected to give the exocyclic olefin (489). This was proved to have the same skeleton as (487) by conversion into a common triol. As this alcohol (489) had been converted back into avermectin B1, with no loss of stereochemical integrity or problems with conjugation, it can be seen that this strategy represents an exceptionally useful method of circumventing the problems which have bedevilled the synthesis of these macrolides. G.Alkylation and Alkyl Derivatives.-Previous work has shown that avermectins and milbemycins can be 0-alkylated by treatment with an alkyl halide in the presence of silver oxide;' later work has shown that other silver salts; the carbonate, salicylate, and perchlorate, can be used in this type of reaction.' 917204 For S541 Factors A and C, treatment with methyl iodide/silver oxide in ether gives methylation almost exclusively at the allylic C-5 oxygen; prolonged reaction times, or a change of solvent to HMPA, are necessary for methylation to occur at the C-23 hydroxyl group."' On occasions, even methylation of the hindered C-7 group has been 0b~erved.l'~ An extreme example of this can be seen in the alkylation of 5-0-TBDMS-22,23-dihydroavermectin B1, aglycon with methyl iodide/silver oxide which gave exclusively the 7-0-methyl ether.234 In contrast, alkylation of this aglycon with 2-methoxyethoxymethyl chloride and N,N-di- isopropylethylamine gave only the 13-0-ether.234 Similarly surprising results have been reported, indicating that the C-7-hydroxyl group of S-541 milbemycin analogues is not always the least favourable position for reaction.'" Reaction of 5-methoxy S-541 Factor D with an excess of methyl isocyanate, catalysed by 4,4- dimethylaminopyridine, gave the 7-N-methylcarbamate as the major product, while 5-acetoxy S541 Factor A (491a), when alkylated with t-butyl bromoacetate and a sterically demanding base (KF/alumina or thallium oxide), gave the 7-substituted milbemycin (491 b) as the only significant product in quite reasonable yield; very little reaction at C-23 was observed in either case.The isocyanate 234 H. Mrozik, B. 0.Linn, P. Eskola, A. Lusi, A. Matzuk, F. A. Preiser, D. A. Ostlind, J. M. Schaeffer, and M. H. Fisher, J. Med. Clirm., 1989,32, 375. Avermectins and Milbemycins Part I1 Hi OCOMe (491) (a) R = H (b) R = CH2CQBu’ result was explained as a consequence of the compact and fairly rigid spiroacetal structure blocking approach of the isocyanate to the C-23 oxygen atom after the base had associated with the hydroxyl group. A similar situation could arise with the alkylation by bromoacetate. Preparation of a range of 13-alkoxymethoxy avermectins from 5-0-silylated avermectin B aglycons by N,N-di-isopropylethylaminecatalysed etherification has been described; 235 13-epi alkoxymethoxy avermectins were also prepared from the corresponding agly~ons.~~~ Methylation and ethylation at the C-23 group of 5-protected S541 analogues was achieved with the appropriate trialkyloxonium tetraflu~roborate.~~~ Analo-gous reactions were performed on the epimeric 23-hydroxy compounds (prepared by a Mitsunobu acylation/hydrolysis inversion sequence) 238 and on 13R- hydroxy-- 23-desoxy S541 (prepared by selenium dioxide oxidation, see Section 8E).239 Reaction of 5-0-TBDMS-13-hydroxy milbemycin a4 with dihydropyran under acid catalysis gave the two diastereoisomers of the 2-tetrahydropyranyl analogue which were active ant helm in tic^.^^' C-Substituted milbepycins have been prepared by addition of alkyl-lithium or Grignard reagents to the 23-ketone group of the S541 analogue (492) (Scheme 79).79*241 These additions are highly stereoselective. For example, methyl magnesium iodide gave a 19: 1 mixture of P:a isomers while trimethylsilylmethyl magnesium iodide gave only the p isomer. The Peterson product was converted into the 23-methylene derivative by acid treatment; the use of methylenetriphenylphosphorane or zinc/di-iodomethane with titanium tetrachloride to prepare this compound was markedly inferior.242 ”* Y Morisawa, A Saito, T Toyama, dnd S Kaneko (Sankyo KK), European Patent, EP 357460 (1988) 236 B 0 Linn and H H Mrozik (Merck and Co Inc ), US Patent, US 4 587 247 (1986) 237 E P Tiley and M V J Ramsay (American Cyanamid Co ), European Patent, EP 307 223 (1989) 238 R Bell, M V J Ramsay, H M Noble, D Noble, N Porter, J B Ward, and R A Fletton (American Cyanamid Co ), European Patent, EP 307 224 (1989) 239 M V J Ramsay, R Bell, P D Howes, E P Tiley, and D R Sutherland (American Cyanamid Co ), European Patent, EP 341 972 (1988) 240 B Frei, A C O’Sullivan, and E Sturm (Ciba-Geigy AG), SWISS Patent, CH 669 382 (1989) 241 M V J Ramsay, R A Fletton, J B Ward, D Noble, and N Porter (Glaxo Group Ltd ), European Patent, EP 241 145 (1986) Davies and Green (492) (493) R = Me or CH2SiMe3 Reagents: i, MeMgI or Me3SiCH2MgI Scheme 79 Reagents: ii, p-TsOH, ROH Scheme 80 (494) Reagents: i, Me3A1, CH2C12 Scheme 81 Similar Grignard additions to 23-0x0 avermectins have been reported although the stereochemistry of the newly created centre was not reported.243 Grignard reactions at C-5 of 5-0XO S541 and at C-5, C-10, C-13, C-4’, and C-4” of 5-0x0 avermectins have also been described.244 A Grignard reaction was also used to append a 5’-alkoxy-tetrahydrofurangroup onto the C-13 position of milbemycins; in this case the stereochemistry was as shown in the diagram with a 3 : 1 ratio of isomers at C-5’ on the tetrahydrofuran ring (Scheme 80).244 Reaction of the allylic 15-acetoxy compound (494) (Scheme 81) (see Section 242 M.V. J. Ramsay, B. M. Bain, J. B. Ward, H. M. Noble, N. Porter, R. A. Fletton, D. Noble, D. R.Sutherland, and P. D. Howes (Glaxo Group Ltd.), European Patent, EP 231 104 (1987). 243 P. Eskola, T. L. Shih, and H. Mrozik (Merck and Co. Inc.), European Patent, EP 351 923 (1990). 244 P. Maienfisch (Cib-Geigy AG), European Patent, EP 284 563 (1988). 329 Aoermectins and Milbemycins Purt I1 OH (496) Reagents I, RMgBr, CuX Scheme 82 (498) (499) Reagent7 I, Bu"L1, CuCN Scheme 83 8H for preparation) with trimethylaluminium gave the 13-methyl substituted milbemycin (495),'95" which could also be obtained from the reaction of trimethylaluminium with 5-0-t-butyldimethylsilyl-13-ethoxycarbonyl milbemycin D.245Other 13-alkyl milbemycins were also obtained from (494) by reaction with trialkylalumini~ms.~~~ The S-541 halides (496), deriving from hypohalous acid addition (see Section 8H), were used to prepare a range of differently substituted C-25 analogues (497) by conjugate addition of alkyl or arylcopper lithiums, or alkyl vanadates (Scheme 82).246 The chloride (498) (for preparation see Section 8H) could similarly be reacted with n-butyl-lithium/cuprous cyanide to give the 14-pentyl milbemycin analogue (499) (Scheme 83).247 25-Methylene avermectin A2, e.g.(500) (see Section 8E for preparation) has 245 K Gubler, Y Tsukamoto, K Sato. and T Tanai (Ciba-Gelgy AG). European Patent, EP 253378 ( 1986) 246 B M Bain, N Porter, P F Lambeth, H M Noble, A C Rosemeyer, R A Fletton, J B Ward, D Noble, D R Sutherland, M V J Ramsay, and E P Tiley (Glaxo Group Ltd ), European Patent, EP 237 339 (1987), Australian Patent AU 307 050 (1987) 247 U Burckhardt (Ciba-Geigy AG), European Patent, EP 165 900 (1985) 330 Davies and Green c Reagents: i, (O-M~-C~H~)~P, Et3N, MeCN, 3-bromopyridine; ii, Pd/C, Et,N, RI, MeCN Scheme 84 been utilized to prepare a range of 25-arylethylene substituted avermectins by palladium-catalysed Heck reaction; a 2-carbomethoxyethylene derivative was also prepared (Scheme 84) as were several B2 analogues.219 This combination of directed biosynthesis and Heck alkylation is a very powerful method of preparing new analogues; a further example can be found in the but-3-enyl analogue (501).Isomerization of the double bond into the chain is simply achieved (502) (Scheme 85), and Heck alkylations on (501) have been reported to give mixtures of 3- and 4-substituted butyl analogues (503) and (504) (Scheme 85).219 For two of these alkylations, only physical data for the 4-substituted butyl analogue were reported, implying this to be the sole product.This is not explicitly stated in the patent. Exocyclic methylene groups, both substituted and unsubstituted, have been added to the 13-position of milbemycins by reaction of the 13-0x0 compound with Tebbe type titanocenediolefin complexes, e.g. (505).212,248 13-Formyl milbemycins, prepared from the epoxide (459) by acid-catalysed rearrangement, were reacted with titanium-aluminium complexes, e.g. (506), to give a separable 2: 1 mixture of p: x 13-vinyl milbemycins in good ~ie1ds.l~~' Treatment of 5-0-t-butyldimethylsilyl Ivermectin with zinc/copper couple and di-iodomethane gave a mixture of 3,4- and 8,9-methano and 3,4-8,9-dimethano Ivermectins in the ratio 8:2:3.249 All of the introduced cyclopropanes were on the a-face of the molecule.Presumably the direction of addition is controlled by 24H P. Maienfisch and K. Oertle (Ciba-Geigy AG), German Patent, DE 3 808 634 (1988) 249 M. J. Wyvratt (Merck and Co. Inc.), US Patent, US 4 581 345 (1985). Avermectins and Milbemycins Part I1 OMe HO. ..A HA OMe + (503) (504) (502) Reagents I, COD(MePh2P)21r+ PF,, Hz,11, Pd(OAc)z, Et,N, RI, MeCN Scheme 85 the 7-hydroxy group.2s0 The major product, however, from the addition of dichlorocarbene to 5-0-t-butyldimethylsilyl milbemycin ~3,is the 14,15-dichloromethylene adduct (507), a small amount of the 3,4-14,15 bis-adduct was also i~olated.~ A similar reaction on milbemycin a1, with dichlorocarbene (again generated from chloroform and aqueous sodium hydroxide) with the phase-transfer agent tetra-butylammonium chloride added, gave only the 14,15- adduct.Similarly, dichlorocarbene generated from n-butyl-lithium and chloroform reacted with 5-0-t-butyldimethylsilyl milbemycin a3 to give similar specificity. Dibromocarbene addition (from magnesium turnings and bromoform) to milbemycin a3 gave only the 14,15-dibromomethylene adduct, which could be 250 Unpublished work reported in ‘Recent Advances in the Chemistry of Insect Control’, ed N F Janes, Royal Society of Chemistry Special Publication No 53, London, 1985, p 70 251 J C Gehret (Ciba-Geigy AG), European Patent, EP 285 561 (1988) Dauies and Green Ci OTBDMS (508)(a) R = H (b) RrTBDMS (509) (a) R = H (b) R=TBDMS (510) (a) R= H (b) R-TBDMS Reagents: i, Et,AI, HN3 Scheme 86 reduced with zinc in acetic acid to give a mixture of monobromo-cyclopropane and the unsubstituted cyclopropane.This patent claims dihalocarbene adducts with all conceivable combinations of halides, but no experimental details are given apart from those detailed above. H. Reactions of the C-1SC-15 Sub-unit.-The allylic grouping between C-13 and C-15 of the avermectins and milbemycins merits special attention, as a range of reactions occur in this region which can be modulated allylically by the double bond, or homo-allylically by the A'' double bond.Reaction of the 14,lSepoxide of milbemycin D (508a) with a 1:l molar mixture of hydrazoic acid and triethylaluminium in toluene solution gave a 61% yield of 14-azido- 15-hydroxy-milbemycin D (509a), no other product being reported.252 The use of a 3:2 (approximately) mixture of triethylamine and hydrazoic acid in diethyl ether solution, however, gave only 10% of the azido- alcohol (509a); the major product (45%) was the allylic alcohol (510a).'96 When this reaction was repeated on 5-0-t-butyldimethylsilyl milbemycin D (508b), using THF as solvent, the major product (47%) was, again, the allylic alcohol (510b) along with a very small amount (2%) of 13-P-azido milbemycin D 252 H.B. Merevala and B. Frei, Helv. Chim. Acta, 1986,69,415. Auermectins and Miibemjcrns Purt /I (51 1) 196 Derivatives of the allylic alcohol (510a) were also obtained from 5-0-methyldiphenylsilyl milbemycin a3, milbemycin a4, and 5-0-t-butyldimethylsilyl-13-deoxy-22,23-dihydroavermectinB 1a Reaction of the 15-hydroxy milbemycins (5lob) with diethylaminosulphur trifluoride gave the allylic rearrangement products, 13-f$-flUOrO milbemycins (512, X = F) in good yield 13-P-Fluoro milbemycins have also been prepared from 13- hydroxy milbemycins by reaction with diethylaminosulphur trifluoride '90h The susceptibility of A' 15-hydroxy milbemycins to allylic rearrangement can be seen by the wide range of 13-eubstituted compounds prepared therefrom 13-Chloro, bromo, and iodo corn pound^,'^^ and 13-alkoxy, th~oalkyl.~~~ thioa~yl,~~~ and 13-acyloxy 198a milbemycins have all been reported, as has the rearrange- ment of 5-0-t-butyldimethy1sily1-15-hydroxyA' milbemycin D to the correspond- ing 13-hydroxy milbemycin with pyridinium dichromate in DMF 253 Swern oxidation of this 13-hydroxy milbemycin and reduction of the 13-0x0 product with sodium borohydride gave 13-hydroxy milbemycin D, dn analogue of an avermectin aglycon '90a There is very little conformational flexibility in the macrolide ring of avermectins and milbemycins, and because of this the atoms 0-13, C-13, C-14, C- 14a, and C-15 lie in almost the same plane 234 Thus the (T bond of a 13p leaving group lies in the plane of the C-14-C-15 7t orbitals, and any reactions of a 13p-substituted compound should show considerable allylic character In contrast the (T orbitals of a 13x leaving group are orthogonal to the plane of the C-14-C-15 n: orbitals, and thus there should be no allylic interactions in reactions observed at this centre Homoallylic interactions with the C-10-C-11 double bond are, however, possible A further facet of the chemistry at C-13 is provided by apparent steric hindrance in the reactions of the natural 13~-substituted aglycon This can be clearly observed in the reaction of 22,23-dihydroavermectin B1, aglycon with t-butyldimethylsilyl chloride to give only the 5-0-silylated product, no silylation at C-13 is seen under normal conditions 234 Displacements at C- 13 of the avermectin aglycon have been studied extensively using the readily available 5-O-t-butyldimethylsilyl-22,23-dihydroavermectinB 1, aglycon (513, R = H) (see above) 234 250 Solvolysis of the derived tosylate in methanol gave the 13-methoxy compound (514, R = MeO) with retention of configuraion, probably by participation of the A'' double bond forming a homoallylic cation (Scheme 87) When the solvolysis was conducted in the presence of HCl the 13a-chloro derivative (514, R = Cl) was obtained, while solvolysis in a mixture of HF/THF/pyridine gave the 13~ fluoride (514, R = F) as a major product along with some of the 13p fluoride (515d) as a minor product (Scheme 87) Other similar solvolyses have been reported 236 In an intermediate step in the previously described correlation of the avermectins and milbemycins, the mono-protected aglycon (513, R = H) was reacted with 2-nitrobenzenesulphonyl chloride to give the inverted 13p-253 B Frei and A C 0 Sullivan (Ciba-Geigy plc) British Patent GB 2 167 751 (1986) 254 A C 0 Sullivan and B Frei (Cibd Geigy AG) European Patent FP 252 879 (1986) 334 Davies and Green (512) R' = Me.TBDMS. TBDPS R~ = ~e,Et, Pi, BUS X = F. CI, Br, I, OR, SR. SCOR, OCOR (514) R P OH, OMe, OEt, OAC,CI, F (515) (a) R = CI (b) R=I (c) R= Br (d) R= F Scheme 87 chloro derivative (515a).'-4 This was obviously formed by displacement of the 2-nitrobenzenesulphonate intermediate with chloride ions in solution, a supposition supported by the observation that addition of an excess of tetrabutylammonium iodide gives the 13P-iodo analogue (515b) in good yield.234 The expected allylic reactivity of this iodide is demonstrated by its reaction with silver acetate in glacial acetic acid, in that the allylically rearranged 15-acetate (516) was obtained in addition to the 13-acetate (517) (Scheme 88).234*236A further illustration of the allylic character of the 13P-iodide (515b) can be found in its reaction with methylamine, followed by acetylation, to give the A13 15-amide (518) (Scheme 88).234The reaction of 13-0-iodo-50x0 milbemycin a3 with substituted phenethyl alcohols to give 13-P-phenylethoxy milbemycins has been described in a Sankyo Avermectins and Milbemycins Part II -.yy (518) Reagents 1, AgOAc, 11, MeNH2, 111, Ac20, pyridine Scheme 88 patent 235 No allylically rearranged products were described but as experimental details in patents are rarely complete this does not necessarily mean they were not formed The retention of stereochemistry observed, however, argues for the intervention of a homoallylic cation Dehydroiodination of (515b) with collidine at 100 OC gave a mixture of the tetra-ene (519) (of undetermined stereochemistry about the C-12-C-13 double bond),234 whereas the use of DBU gave the tetra- ene with a concomitant shift of the A3 double bond into the A’ position 250 A by-product obtained from the collidine reaction was the 130 isomer of the initial aglycon (515 R = OH), this epz aglycon could be obtained in good yield if the reaction was repeated in aqueous collidine 234 Reaction of milbemycin D with hypochlorous acid or sulphuryl chloride gives a ‘surprisingly high yield’ of 15H-15-chloro milbemycin D (520) 255 Similar reactions have been reported for milbemycins a1 and a3, and 13-deoxy-22,23- dihydroavermectin B1, aglycon ”’S-541 Factor A, which possesses an additional double bond in the 25-alkyl substituent, has also been reacted with hypochlorous ’55 (a) U Burckhardt (Ciba-Gelgy Corp), US Patent, US 4584314 (1986) (b) European Patent EP 143 747 (1984) Davies and Green y S541-FactorA i (%a) (521) X = CI or Br Reagents: i, HOCl or HOBr Scheme 89 ..-Two products were isolated, a monochloride substituted in the C-25 side chain (521, X = Cl), and a dichloride (522, X = Cl) analogous to the milbemycin D product but also substituted in the side chain (Scheme 89).Hypobromous acid reacted in a similar way, while phenylselenyl bromide gave only the side-chain substituted compound (521, X = Br). Cyanamid have reported that the use of N-bromoacetamide in aqueous acttone or N-chlorosuccinimide in methanol give the side-chain halogenated products (521, Hal = C1 or Br).’84b*257 Oxidation of milbemycin D with singlet oxygen followed by reduction with triphenylphosphine generates the 15-hydroxy A14,140milbemycin (523) in good yield; a small quantity of the tertiary alcohol (524) was also i~olated.”~ This reaction was also carried out on 5-OXO milbemycin D and various 5-protected milbemycins with similar results.Soivolysis of the mesylate of the allylic alcohol (523) in glacial acetic acid gave the 14a-acetoxy milbemycin analogue (525); 2’8 the allylic aldehyde was also prepared and converted into various ~ximes.~’~ The preparation of 5-OXO milbemycins and avermectins by mercuric acetate induced isomerization of the 3,4-double bond of a 5-methoxy derivative followed by acid hydrolysis of the resulting enol ether has been rep~rted.’,’~~,~~~ This process, however, has been reported to be unsatisfactory on an industrial scale ”13 S. Y. Tamura and G. Asato (American Cyanamid Co.), European Patent, EP 297 205 (1988). ’” (a)J. C. Gehret (Ciba-Geigy AG), European Patent, EP 144 285 (1985); (b) B.Frei, H. B. Merevala, U. Burckhardt, and J. C. Gehret (Ciba-Geigy AG), Swiss Patent, CH 656 129 (1986). 258 J. C. Gehret (Ciba-Geigy AG), European Patent, EP 281 522 (1987). 259 H. Mrozik, P. Eskola, and M. H. Fisher, J. Org. Chem., 1986,51,3058. 337 Avermectins and Milbemycrns Part II for several reasons, not the least of which is the ecological problem induced by using large quantities of mercury salts 208a In the course of circumventing this problem, these workers oxidized 5-0XO milbemycins with selenium dioxide in formic or acetic acids and, after hydrolysis of the total crude products, obtained good yields of 13-hydroxy milbemycins by allylic oxidation 1. Unnatural Sugar Derivatives of Avermectins and Milbemycins.-Treatment of 5-0-t-butyldimethylsilyl avermectins with sulphur trioxide/pyridine complex gives the 4 ’-0-sulphate derivatives 260 and several simple derivatives, such as semicarbdzones and hydrazones, of 4”-0xo avermectins have been described ’74a 261 Reformatsky reactions have been performed on these 4”-OXO avermectins (and also on the 4-monosaccharide) to give, for example, 4’- [(ethoxycarbonyl)methyl] avermectin B1 262 More fundamental changes in the saccharide groups at C-13 are provided by the alkylation of 13-hydroxy milbemycins (avermectin aglycons) The standard Koenigs-Knorr procedure used previously has been largely supplanted by the coupling of 1-fluoro and 1-phenylthio sugars, which gives better yields 263 An example can be found in the coupling of l-S-(2-pyridyl)-sugars to 5-protected 13- hydroxy milbemycin D with silver perchlorate followed by mild acid treatment, compounds such as 13-0-(3,4-di-0-acetyl-2-deoxy-~-rhamnosyl)and 13-0-(4-0- acetyl-L-oleandrosyl) milbemycin D and 13-0-[4’-O-(~-oleandrosyl)-~-oleandrosyl] milbemycin a3 were prepared in this way 264 1-Fluoro sugars were used to prepare C-2”p and C-2”a-fluoro akermectins B1, 265 An avermectin with a third oleandrose group at the C-4” position has been prepared which is equipotent with avermectin Bl 266 The preparation of 13-j3-(x-~-oleandrosyI-a-~-oleandrosyloxy)milbemycin derivatives, and thus the mono-oleandrosyl analogues, by the incubation of 130- hydroxy milbemycins with S uvermitzlis ATCC 31272 or ATCC 31780 has been described 267 This should present a very effective method for many milbemycin derivatives as such bio-organic approaches frequently require little in the way of protection of the substrate 268 5-0-sugar derivatives of milbemycin D have been prepared by coupling acetohalo-sugars or S-pyridylaceto-sugars in the presence of di-isopropylethylamine and a silver salt 269 260 M J Wyvratt (Merck and Co Inc ), US Patent, US 4 622 3 13 (1986) B 0 Linn and H H Mrozik (Merck and Co Inc ), European Patent, EP 343 708 (1989) M J Wyvratt (Merck and Co Inc 1, US Patent, US 4 833 168 (1989) 263 K C Nicolaou R E Dolle, D P Papahatjis, and J L Randall, J Am Chem SOC,1984, 106.4189 264 B Frei and H B Merevala (Ciba-Geigy AG), European Patent, EP 235085 (1987) 265 C Bliard F C Escribano G Lukacs A Oleskar and P Sarda J Chem Soc Chem Cornmun 1987, 768 266 M J Wyvratt el a/,unpublished work reported in Topics in Medicinal Chemistry 4th SCI RSC Medicinal Chemistry Symposium, ed P R Leeming, Royal Society of Chemistry Special Publication No 65, 1988 267 G T Lawrence, M J Dawson, D Noble M V J Ramsay R Bell D R Sutherland, and E P Tiley (American Cyanamid Co ), European Patent, EP 341 973 (1989) H G Davies, R H Green, D R Kelly, and S M Roberts, ‘Biotransformations in Organic Chemistry’ Academic Press, London, 1989 Davies and Green 9 Conclusion The commercial success of Ivermectin has stimulated many other companies to produce their own macrolide antiparasitics.Scientifically, however, the greatest advance must be the stimulus given to synthetic chemistry, methods of structure determination, and fermentation techniques.The complex nature of the avermectins and milbemycins has caused many problems. Perusal of any of the avermectin and milbemycin total syntheses proves this point. Is there any point in further tempting fate by devising other avermectin and milbemycin syntheses? The answer, surely, is a resounding yes. The successful total synthesis of such a macrocycle with multiple stereogenic centres represents a formidable proof of a particular synthetic concept and needs all the power of modern organic chemistry. The present syntheses detailed in this review are outstanding examples of the synthetic art but there may well be shorter, more efficient, methods to the natural avermectins and milbemycins which can be devised in the future.For the medicinal chemist future synthetic work will be aimed at producing analogues with improved pharmacological properties, either by total synthesis or semi-synthetic methods. Such is the potential market for improved endectocides that there is sure to be much more work to be done in the coming years. In the introduction to our first review,’ we noted that parasitic infections of livestock in America have been estimated to cost more than $3 billion annually. If this is the cost borne by one of the richest nations in the world, how much greater must be the cost in both economic and human terms in those third world countries where parasites take such a terrible toll of both animal and human health.If avermectins and milbemycins can reduce this toll in any way then the scientific work put into them will have reaped a rich harvest. Acknowledgements. The authors gratefully acknowledge the assistance of Miss J. Bradshaw of Glaxo’s Information Science and Services Department, Glaxo Library (Greenford) and also the staff of Glaxo’s Intellectual Property Depart- ment. 269 J C Gehret, E. Sturm, and B. Frei (Ciba-Geigy AG), European Patent, EP 185 623 (1985)

ISSN:0306-0012    

DOI:10.1039/CS9912000271

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

Chern. SOC.Rev. 1991,20,341-353 Spin Trapping of Inorganic Radicals By Detlef Rehorek DEPARTMENT OF CHEMISTRY, UNIVERSITY OF LEIPZIG, TALSTRASSE 35, 0-7010 LEIPZIG, GERMANY 1 Introduction About 30 years ago it was demonstrated by both American' and a group of Hungarian and Russian 'researchers that nitroso compounds add organic free radicals to form persistent aminoxyls which are readily detectable by EPR spectroscopy. This work was extended by Dutch3 and Japanese4 workers, the latter showing that aminoxyls are also formed by addition of short-lived free radicals to nitrones. Shortly after this, the importance of this reaction as an analytical tool for detection and identification of short-lived radicals was recognized by Janzen and, independently, by Lagercrantz in Sweden, Perkins ' in the U.K., de Boer * in the Netherlands, Leaver and Ramsay in Australia and Terabe and Konaka'' in Japan.Later Janzen and Blackburn" coined the expression spin trapping for a reaction where a short-lived radical 'R is scavenged by a diamagnetic compound S to form a persistent radical adduct R-S', equation 1. 'R + S R-S' (1) In this reaction the diamagnetic scavenger S is called the spin trap, and the resulting persistent radical R-S' is named the spin adduct. Unlike many short- lived radicals 'R, the spin adduct R-S' is readily detectable by EPR even in liquid solution at higher temperatures. Although more than 100 different compounds have proven to be suitable spin ' A K Hoffman and A T Henderson, J Am Chem SOL.,1961,86,4671 F Tudos, I Kende, T Berezsnich, S Szolodovnikov, and V Vojevodskij, Magi.Ken1 Foli , 1963. 69. 371, F Tudos, I Kende, T Berezhnykh, S P Solodovnikov, and V V Voevodskii, Kinet Kalal. 1965 6,203 A Mackor, Th A J W Wajer, J D W van Voorst, and Th J de Boer, Tetrahedron Lett , 1966, 21 15. Th A J W Wajer, A Mackor, Th J de Boer, and J D W van Voorst, Tetrahedron, 1967,23,4021 M Iwamura and N Inamoto, Bull Chem Soc Jpn ,1967,40,702,703 E G Janzen, Chem Eng Netts, 1965,43, 50, E G Janzen and J L Gerlock, Nature, 1967, 222, 867, E G Janzen and B J Blackburn, J Am Chem Soc, 1968,90,5909'C Lagercrantz, J Phqs Chem, 1971,75, 3466, C Lagercrantz and S Forshult, Nature. 1968,218, 1247, C Lagercrantz and K Torssell, Acta Chem Scund, 1968,22, 1935 'M J Perkins, in 'Essays on Free-radical Chemistry', Special Publication No 24, The Chemical Society, London 1970, p 97, G R Chalfont, M J Perkins, and A Horsfield, J Am Chem Soc , 1968, 90,7141 Th J de Boer, Can J Chem , 1982,60, 1602 and references therein I H Leaver and G C Ramsay, Tefrahedron,1969,25,5669 lo S Terabe and R Konaka, J Am Chem Soc, 1969,91,5655 E G Janzen and B J Blackburn, J Am Chem Sot , 1969,91,4481 34 1 Spin Trapping of Inorganic Radicals 0.(3)R3--N=O + 'R -R3-h--R traps, nitrones and C-nitroso compounds, which both give persistent aminoxyls upon radical addition (equations 2 and 3), are preferred by most researchers. The spin trapping technique has been successfully applied to EPR detection of organic free radicals." Applications in biochemistry and medicine l3 and in various areas of chemistry such as gas phase reactions, l4 electrochemistry, l5 and photochemistry l6 have also been reviewed.An extensive compilation l7 of spin adduct EPR data have been published recently. Although the spin trapping was originally designed for the detection of short-lived organic radicals, many inorganic radicals have also been shown to form persistent spin adducts with both nitroso compounds and nitrones. This is of particular importance since many inorganic radicals are not only short-lived but have also extremely short spin-lattice relaxation times due to their orbitally degenerate ground states. This effect, which operates for radicals such as 'OH, 'SH, 'Br, 'Cl, and others, causes severe line-broadening and makes direct EPR detection in fluid solutions impossible.In order to detect those radicals, cryogenic temperatures, very often as low as 4 K, are required. Moreover, many inorganic radicals absorb light only in the near UV region, z.e. the range where the precursor compounds usually also exhibit high absorbance. Therefore techniques based on the measurement of the light absorption, e.g. flash photolysis, are unsuitable. In some previous publications * particular reference has been made to l2 E G Janzen, Acc Chem Res, 1971, 4, 31, M J Perkins, Adv Phys Org Chem, 1980, 17, I, R Kh Freidlina, I I Kandror, and R G Gasanov, Usp Khrm , 1978,47, 508, V E Zubarev, V N Belevskij, and L T Bugaenko, Vsp Khrrn, 1979,48, 1361, D Rehorek, Z Chem , 1980,20,325, I Rosenthal and P Riesz, Radiat Phvs Chem, 1987,30,381, P Riesz and S Rustgi, Radiat Phjs Chem, 1979,13,21 l3 E G Janzen, in 'Free Radicals in Biology', ed W A Pryor, Vol 4, Academic Press, New York 1980, 115, P J Thornalley, Llfe Chem Rep, 1986,4, 57, G M Rosen, Adv Free Radrcal Biol Med, 1985, 1, 345, E G Janzen, H J Stronks, C M DuBose, J L Poyer and P B McCay, Enuiron Health Perspect , 1985, 64, 151, E Albano, A Tomasi, K H Cheeseman, V Vannini, and M U Dianzani, in 'Free Radicals in Liver Injury', ed G Poli, K H Cheeseman, M U Dianzani, and T F Slater, IRL Press, Oxford, 1985, p 7, P B McCay, T Noguchi, K -L Fong, E K Lal, and J L Poyer, in 'Free Radicals in Biology', ed W A Pryor, Vol 4, Academic Press, New York, 1980, p 155 '4 E G Janzen, Creat Delect Excited State, 1976,4,83 l5 T H Walter, E E Bancroft, G L McIntire, E R Davis, L M Gierasch, H N Blount, H J Stronks, and E G Janzen.Can J Chon, 1982,60, 1621 J R Harbour and M L Hair, Adv Collord Interface Scr, 1986, 24, 103, D Rehorek and H Hennig, Can J Chem , 1982, 60,1565, D Rehorek, S Di Martino, T J Kernp, and H Hennig, J Inf Rec Mat, 1989,17,469 "G R Buettner, Free Radical Biol Med, 1987,3,259 M C R Syrnons, Phrlos Trans R SOL London B, 1985, 311, 451, D Rehorek, Proc 10th Conf Coord Chem, SmolenicelCSSR 4 7th June 1985, p 329, D Rehorek, H Hennig, C M DuBose, T J Kernp, and E G Janzen, Free Rad Res Commun, 1990, 10, 75, R P Mason and C Mottley, in 'Electron Spin Resonance', Vol IOB, ed M C R Symons, A Specialist Periodical Report, The Royal Society of Chemistry, London, 1987, p 185 Rehorek problems arising from spin trapping inorganic radicals.Possible pitfalls in the course of spin trapping have also been pointed out by various worker~.’~-~l Here we will discuss problems specific for spin trapping inorganic and organometallic radicals. 2 Spin Trapping using Nitrones Except for some platinum radicals 22 and metal-centred radicals 23 of Group IVB, persistent spin adducts of metal-centred paramagnetic species with nitrones have not yet been described. On the other hand, many inorganic radicals and atoms that are not trapped by nitroso compounds form persistent spin adducts with nitrones.The structures of some typical nitrones used in spin trapping experiments are given below. 0-BU~-N=CH-CH=L-BU~(T PBN DMPO GBBN ‘S03-Na+ (MObPBN 2-SSPBN DEN As a result of their high solubility in both polar and non-polar solvents, as well as their reasonable photochemical and thermal stability, PBN and DMPO are probably the most popular spin traps. Usually the EPR spectrum of a nitrone spin adduct exhibits hyperfine coupling of the unpaired electron with nuclear spins of the 14N and the P-H which leads to a triplet of doublets. However, if the centre of the trapped radical has a magnetically active nucleus, i.e. a nucleus with a spin I > 0, additional splitting may be seen. The size of the P-H and 14N hyperfine splittings (and, in some cases, other hyperfine splittings) in the spin adduct may serve as a diagnostic tool for the identification of the trapped radical ‘R.EPR hyperfine splittings for inorganic spin adducts to PBN are summarized in Table 1. Very often the influence of the l9 C. Mottley and R. P. Mason, Biol. Magn. Reson., 1989, 8, 489; K. Stolze and R. P. Mason, Biochem. Biophys. Rex Commun., 1987, 143, 941; C. Mottley, B. Kalyanaraman, and R. P. Mason, FEBS Lert., 1981, 130, 12; J. M. Coxon, B. C. Gilbert, and R. 0.C. Norman, J. Chem. SOC., Perkin Trans. 2, 1981, 329. 2o A. R. Forrester and S. P. Hepburn, J. Chem. SOC. (C), 1971, 701. ” H. Chandra and M. C. R. Symons, J. Chem. SOC.,Chem. Commun., 1986,1301. 22 H. C. Clark and C.S. Wong, J. Am. Chem. SOC.,1975,95,7073. 23 P. Riviere, S. Richelme, M. Riviere-Baudet, J. Satge, M. J. S. Cynane, and M. F. Lappert. J. Chem. Rex (M), 1978, 2801; J. Chem. Res. (S), 1978, 218; P. Riviere, S. Richelme, M. Riviere-Baudet, J. Satge, P. I. Riley, M. F. Lappert, J. Dunogues, and R. Calas, J. Chem. Rex (S), 1981, 130; J. Chem. Res. (M),1981, 1663; A. Alberti, R. Leardi, G. F. Pedulli, A. Tundo, and G. Zanardi, Gazz. Chim. lrnl., 1983, 113, 869; H. Chandra, I. M. T. Davidson, and M. C. R. Symons. J. Chem. Soc., Perkin Trans. 2, 1982, 1353. 343 Spin Trapping of Inorganic Radicals Table I EPR hiper-ne splitting constants for rpin adductr of inorganic radiculr to PBN Radical Soltrent aN/mT aH/mT ax/mT Ref 'H benzene 1480 0 741 (2H) 57 'D ccl4 147 0 73 011 (2D) 58 'F benzene 1 22 0 118 456 ("F) 30 'C1 CH3CN 1 270 0 082 o 620 (35~1) 0 512 (j7Cl) 59 'Br benzene 113 -349 (81Br) 3 24 (79Br) 31 'OH water 155 0 272 0 336 ('70) 60 'OOH water 1481 0 27 027 ("0) 60 'N 3 water 1501 0 201 0 201 (14N) 28 'NCO CH3CN 1 509 0 315 0 184 (14N) 15 'NCSSCN CH3CN 1444 0 109 0 368 (14N) 38 'oso4 CH3CN 1390 0 123 61 'SO3 water 1495 0 197 0 034 (2H) 62 TO2 water 1 580 0 452 63 'CN CH3CN 1 504 0 198 0 985 ( 3C) 15 'CONH2 water 1553 0 320 0 050 (14N) 0 050 ('H) 25 'NC(0H) CH3CN 1481 0 215 0215 (14N) 26 'NCHOS03 CH3CNIH20 1485 0 082 0 170 (14N) 25 'OP05 water 1546 0 184 63 "Formed by hydrolysis of CN Formed by addition of OS03 radicals to cyanide solvent on the hyperfine splitting exceeds that of the radical structure24 Therefore solvent effects have to be studied carefully before literature data are used for assigning spin adducts When a bulky methoxy group is introduced into the ortho-position of the phenyl ring of PBN the P-H coupling constants are much more affected by the nature of the trapped radical than in the unsubstituted analogue On the other hand, the steric hindrance in (M0)3PBN strongly reduces the spin trapping reactivity which, together with the poor solubility in polar solvents, renders (M0)3PBN in general less suitable for the detection of inorganic radicals, although this spin trap appears to be extremely useful for some special cases as will be shown below In addition to the 14N coupling and hyperfine splittings due to magnetically active nuclei in P-position, long-range hyperfine splittings may be observed for some radicals Examples are spin adducts of 303,'PO$ ,and 'HP02 radicals to PBN which exhibit additional splittings due to the two ortho-protons of the phenyl ring adjacent to the nitrone function In the case of the 'HP02 spin adduct to PBN, hyperfine splitting by the proton attached to the phosphorus nucleus could also be resolved Long-range hyperfine couplings are also observed in the carbamoyl ('CONH2) spin adduct to PBN In inorganic reactions carbamoyl radicals are formed by 24 E G Jdnzen G A Coulter U M Oehler and J P Bergsmd Carl J Clitni 1982 60 2725 344 Rehorek either stepwise hydrolysis of cyanyl radicals ('CN) or 'OH radical addition to cyanide ions 25 The spin adduct of the intermediate radical 'N=COH- could also be detected 26 It was shown by Janzen et al 27 that deuteration of PBN leads to sharpening of the EPR lines which allows the resolution of splittings caused by magnetically active nuclei in the y-position, eg proton splittings in the spin adduct of 'CH3 and 'CH2CH3, respectively Using 'HI4-PBN as a spin trap, the hyperfine splitting caused by 14N nuclei in y- and &positions, respectively, could be observed in the spin adduct of the ON3 radical confirming the previous assignment 28 of the 'N3 spin adduct Interestingly, well-resolved ESR spectra with long-range hyperfine couplings could be obtained with the spin trap glyoxal-bis(t-butyl nitrone) (GBBN) 29 This spin trap also turned out to be very useful, since spin adducts of bromine atoms are easily detectable during photolysis of carbon bromides 30 The lifetime of the 'Br spin adduct to GBBN is about two orders of magnitude higher than with PBN 31 However, no spin adducts of iodine atoms have been reported so far Although PBN is one of the most popular spin traps, other spin traps may be recommended for special purposes Thus, 5,5-dimethyl-pyrroline-1 -oxide (DMPO)32 is widely used for studies in aqueous systems The use of DMPO as a spin trap for oxygen-centred radicals has been the subject of various excellent review articles 33 and will not be discussed here It is generally assumed that spin adducts are exclusively formed by addition of free radicals to the spin trap However, alternative routes leading to spin adducts of nitrones should also be considered One of these alternative routes, which is quite common in inorganic spin trapping is the hydrolysis of primary spin adducts leading to the hydroxyl spin adduct, although no free hydroxyl radicals are formed in the reaction (see Scheme 1) Thus, chlorine atoms35 and 'OSOJ radicals26 were shown to react according to Scheme 1 Using the spin trap (M0)3PBN36it is possible to distinguish between hydroxyl spin adducts formed by hydrolysis (a), and those formed by spin trapping free hydroxyl radicals (b), since oxygen-centred radicals, unlike other radicals, preferentially 25 D Rehorek and E G Janzen, Z Chem, 1985,25,69 26 D Rehorek, T J Kemp, and E G Janzen, to be submitted for publication "E G Janzen, U M Oehler, D L Haire, and Y Kotake, J Am Chem Soc 1986 108 6858 28 D Rehorek, P Thomas, and H Hennig, Inorg Chim Acta, 1979 32 L1 "D Rehorek and E G Janzen, J Praki Chem ,1985,327,969 3" E G Janzen, D Rehorek, and H J Stronks, J Magn Reson 1984,56, 174 31 D Rehorek and E G Janzen, Z Chem, 1984 24,441 32 E G Janzen and J I -P Liu, J Magn Reson, 1973 9 510 33 E Finkelstein G M Rosen, and E J Rauckman, Arch Biothem Biophi.s 1980, 200 1 G M Rosen and E Finkelstein Ah Free Radrtd Biol Men 1985 1 345, G M Rosen and E J Rauckman Methods in En=imo/ogt, 1984, 105, 198 G R Buettner In 'Superoxide Dismutase Vol I1 ed L W Oberley, CRC Press, Bocd Rdton, Florida, 1982, p 63 34 D Rehorek C M DuBose, and E G Janzen Inorg Chrm Acta, 1984 83 L7 Z C/irni 1984 24 188 "D Rehorek, E G Janzen, and Y Kotake, Cun J Chem ,in press 36 K Sommermeyer and W Seiffert Z Naturforrch B, 1975, 30, 807, C M DuBose Spin trapping with alpha-2 4 6-trimethoxyphenyl N-tert-butyl nitrone, Ph D Thesis, University of Georgia Athens U S A 1985 E M Jdnzen C M DuBose dnd Y Kotake Tetrahedron Lett 1990 31 7395 345 Spin Trapping of Inorganic Radicals Scheme I OMe Scheme 2 react with (M0)3PBN by hydrogen abstraction at the methoxy group (see Scheme 2) The final aminoxyl (4), 2H,3H-(4,6-dimethoxy)benzo[b]furan-3-yl t-butyl aminoxyl, is formed by cyclization (internal spin trapping) of the intermediate (3) In addition, about 10 to 20% of the regular hydroxy spin adduct (9,ie the spin adduct with an x-hydroxy group, are formed The aminoxyl (4) (uN = 1 644mT, aA = 0 150mT, a$ = 0 123mT, in water) can be easily distinguished from the hydroxy spin adduct (5) (uN = 1 621 mT, UH = 0 885 mT) on the basis of the hyperfine coupling constants In addition to the reaction sequence shown in Scheme 2 partial hydrolysis of the spin adduct may also occur Thus, cyanyl spin adducts (and probably also the corresponding hydroxylamine) undergo hydrolysis to form eventually the carbamoyl spin adduct, equation 4 Reaction is more pronounced for 2-SSPBN than for PBN However, in both cases hydrolysis of the spin adduct is rather slow (k = 5 x lop3s I) and does not compete with hydrolysis of the cyanyl radical leading also to the carbamoyl spin adduct Hence, unambiguous discrimination between hydrolysis of cyanyl spin adducts and spin trapping offree carbamoyl radicals should be possible The spin trap 2-SSPBN, which is commercially available, has been introduced 37 Rehorek because it is a very water-soluble spin trap.This spin trap should be used with caution, however, since it readily undergoes nucleophilic addition of a number of inorganic anions, followed by oxidation of the thus formed hydroxylamine anion, equations 5 and 6. Since only mild oxidants are required in order to bring about oxidation of the hydroxylamine anion, aminoxyl radicals are detectable in the reaction mixture even when no short-livedfree radicals are formed in the course of the reaction, e.g.fairly intense EPR signals of hypophosphite radical spin adducts (ON = l.574mT7 aA = 0.370mT, ai = 0.251 mT, ap = 2.01 1mT) were recorded simply upon addition of sodium hypophosphite to an air-saturated aqueous solution of 2-SSPBN. A similar behaviour was reported2' for PBN in some cases. Fortunately, for most reactions PBN is far less reactive towards nucleophilic addition. Thus, contrasting the observation made by Forrester and Hepburn" at lower pH, cyanide ions lead only to traces of cyanyl and carbamoyl spin adducts at pH 9-10. Although additional problems arise from the enhanced photosensitivity of 2-SSPBN which results in the formation of various persistent aminoxyls upon photolysis,26 this spin trap, nevertheless, may be of particular interest for specific applications.Thus, 2-SSPBN appears to be unique with respect to spin trapping of some radicals formed during oxidation of cyanide ions. A triplet of triplets (ak = 1.175 mT, aft = 0.347mT) assigned to the aminoxyl (6) has been detected26 when cyanide ions are oxidized by Cu" ions. 0. Apparently, the SO; group is involved in the formation of the final spin adducts. It is noteworthy, that (6) is also formed when 'N3 and 'NCO radicals react with 2-SSPBN, suggesting the mechanism shown in Scheme 3. Participation of other groups than the nitrone function in the formation of the final spin adduct has also been observed 38 for reaction of 'SCN and 'NCSSCN-radicals with PBN where cyclization involving the phenyl ring occurs, leading to quite complicated EPR spectra.On the other hand, C-alkyl nitrones such as di-t- butyl nitrone (DBN) that do not have phenyl rings adjacent to the nitrone ''E. G. Janzen and R. V. Shetty, Tetrrrhetlron Lett., 1979, 3229. 38 D. Rehorek and E. G. Janzen. Inorg Clzmm. Acfn, 1986, 118, L29. Spin Trapping of Inorganic Radicals "*C(CN), -OK. -XY *NW + 2SSPBN XY = NN. CO Scheme 3 x function have been proven to be excellent spin traps for both 'SCN and 'NCSSCN -radicals. Distinguishing clearly between spin adducts formed by true radical addition and those formed by 'non-radical' processes is an important problem when using the spin trapping technique, and this problem becomes particularly relevant when dealing with inorganic radicals, since their formation requires either strong oxidants or polar solvents or strongly nucleophilic precursors.All these conditions are unfavourable for spin trapping using nitrone spin traps. In addition to the reaction (a) shown in Scheme 1, the presence of strong oxidants may lead to the formation of short-lived nitrone radical-cations as reported by Chandra and Symons 21 which may eventually yield aminoxyls, equation 7. However, oxidation of nitrones requires very strong oxidants '' which are not present under usual chemical conditions. This situation may change dramatically when photolysis is involved. Although it is quite common to avoid direct excitation of spin traps by using appropriate filters, one has to be aware that energy or electron transfer processes may occur.Thus, Baumann et al.39have shown that electronically excited organic dyestuffs may transfer their energy to nitrones. The resulting excited nitrone is easily oxidized to the cation radical which may react with even weak nucleophiles to form the corresponding aminoxyl. For PBN the energy for the lowest excited triplet state was measured4' as 16200cm-'. The rate constant for quenching the excited state of [Ru(bipy)3I2+ by PBN was found4' to be about 3.3 x lo7 dm3 mol-' s-*. The 39 H. Baumann, U. Dertel, H. J. Timpe, V. E. Zubarev, N. V. Fok, and M. J. Mel'nikov, Z. Chem., 1984, 24, 182. 40 A. P. Darmanyan and G. Moger, J. Photochern., 1984,26,269.41 D. Rehorek and H. Knoll, unpublished results. Rehorek resulting excited PBN molecule is easily oxidized to the cation radical which may react with even very weak nucleophiles to form an aminoxyl radical. A somewhat different situation arises when sensitizers are used which absorb in the high energy region, say below 400 nm. Since a molecule in its excited state is a much stronger oxidant (and reductant) than in its electronic ground state, electron transfer between the excited molecule and the spin trap may occur. This has been proven 42 to be the case for 9,10-dicyanoanthracene, which is known 43 to be a powerful oxidant in its excited singlet state capable of oxidizing a large variety of inorganic anions. Lifetime quenching experiments with PBN gave a good Stern-Volmer plot with k, = (7.00 0.14) x lo9 dm3 mol-' s-'.In the presence of inorganic anions, spin adducts of both sensitizer anion and inorganic radicals were observed. The latter are, however, formed by addition of anions to PBN'. 3 Spin Trapping using Nitroso Compounds Next to nitrones, aliphatic and aromatic nitroso compounds are the most commonly used spin traps. Typical examples for nitroso spin traps are 2-methyl- 2-nitrosopropane (MNP),7 2,3,5,6-tetramethylnitrosobenzene(nitrosodurene, ND),44 2,4,6-tri-t-butylnitrosobenzene(BNB),45 and sodium 3,5-dibromo-4-nitrosobenzene sulphonate (SBNS).46 MNP ND BNB SBNS Unlike nitrones, many nitroso compounds are dimers in the solid state. In order to act as spin traps, dissociation into monomers has to occur.This may lead to some difficulties in kinetic spin trapping experiments. However, because of steric hindrance by the bulky t-butyl groups, BNB exists exclusively as a monomer. SBNS46 and its deuterated analogue appear to be particularly useful for the study of aqueous solutions, although only very few applications to inorganic radicals have been reported so far. Attempts to detect cyanyl-free radicals formed by catalytic reaction of hydrogen peroxide in the presence of horseradish per~xidase~~have failed. It was found that persistent spin adducts of 'SO; radicals to SBNS may be formed by decomposition of SBSN.47*48 42 A. Schleitzer, Diplomarbeit, University of Leipzig 1991; D.Rehorek, A. Schleitzer, H. Knoll, and T. J. Kemp, J. Photochem. Photobid. A: Chemistry, submitted for publication. 43 K. A. Abdullah and T. J. Kemp, J. Photochem., 1985,28, 61. 44 S. Terabe, K. Kuruma, and R. Konaka, J. Chem. Soc., Perkin Trans. 2, 1973, 1252. 45 S. Terabe and R. Konaka, J. Chem. Soc., Perkin Trans. 2, 1973,369. 46 H. Kaur, K. H. W. Leung, and M. J. Perkins, J. Chem. Soc., Chem. Cornmun.. 1981, 142. 4' K. Stolze, S. N. J. Moreno, and R. P. Mason, J. Inorg. Biochem., 1989,37,45. 48 K. Stolze and R. P. Mason, Biochem. Biophys. Res. Commun., 1987, 143,941. Spin Trapping of Inorganic Radicals Table 2 EPR parameters.for spin adducts of inorganic radicals with C-nifroso compounds Spin Radical trap Solvent admT aH/mT ax/mT k! Re$ 'CN MNP water 0.964 0.178 (1N) 47 'BH i MNP benzene 1.39 1.26 (3H) 0.55 (1"B) 0.18 (1'OB) 2.0059 64 'S02F MNP benzene 1.24 65 'S02Cl MNP benzene 1.17 65 'S02H MNP MeOH/H20 1.47 0.46 (113C) 2.0055 66 'S02NH2 MNP water 1.39 2.0055 67 'so; MNP water 1.47 2.0054 67 'H MNP water 1.441 1.391 (1H) 68 'D MNP DzO 1.40 0.22 (1D) 69 'PO$ - MNP water 1.34 1.20 (1P) 70 'As02 MNP water 1.410 0.772 (1 As) 71 "3 ND CHzCl2 0.238 0.721 (2N) 49 'NCO ND CHzCl2 0.240 0.732 (2N) 51 'HPO; water 1.608 0.317 1.603 (31P) 0.184 ('H) 0.027 (2'H) 63 *PO$- water 1.587 0.313 2.166 (31P) 0.020 (2'H) 63 "(CN)2 water 1.587 0.291 0.151 (I4N) 63 'C(CN13 'NH2 water CH3CN 1.437 1.614 0.569 0.354 0.123 (I4N) 0.054 (2 'H) 63 63 Formed by hydrolysis of 'CN.Formed by addition of 'OSO; radicals to cyanide. MNP is one of the most reactive of spin traps; it is also soluble in most solvents including water. On the other hand MNP is very photosensitive, both in the UV and in the red region, forming di-t-butyl aminoxyl which may mask the ESR signals of spin adducts. Nitroso compounds add free radicals directly to the N=O bond. Therefore, spin adducts of nitroso compounds exhibit more line-rich ESR spectra, and the identification of the trapped radicals is more straightforward. There are, however, only relatively few inorganic free radicals which lead to persistent aminoxyls upon addition to nitroso compounds (see Table 2).Among those radicals that form persistent spin adducts with C-nitroso compounds are azidyl radicals ("3). This reaction deserves some comment. 'N3 radicals produced by photolysis of either [Co(NH3)5N3I2+ 28 or CBr4/Ni,49350 may add to both nitroso monomers and dimers leading to a triazene-1,3-dioxyl radical which exhibits hyperfine splitting by three 14N nuclei. The same spin 4y D. Rehorek and E. G. Janzen, Z. Ckem., 1984,24,68.'' H. Knoll, D. Rehorek, and D. J. Stufkens, Z. Chem., 1990,30,220. Rehorek ‘NXY + (XY = N, CO) I R-NO0. 0. ?u? R-N=O -R-A-NXY -R-N=N -R-N-N-N-R Scheme 4 adduct has been observed when cyanate is ~xidized.~’ The mechanism shown in Scheme 4 has been proposed for the formation of azidyl and cyanatyl spin adducts.On the other hand, many short-lived paramagnetic fragments of coordination compounds (metallo radicals, organometallic radicals) form fairly persistent aminoxyls upon addition to nitroso compounds. Spin trapping of short-lived metallo radicals formed during photolysis of organometallic compounds was first demonstrated by Hudson et aLS2While the g-values for spin adducts of simple inorganic radicals to nitroso compounds do not vary markedly with the nature of the radical trapped, the g-values of spin adducts of organometallic radicals very often are strongly influenced by the metallo fragment (see Table 3).Therefore, many metal-centred radical adducts to nitroso compounds should be regarded as paramagnetic metal complexes with organic nitroso ligands rather than metallo nitroxides.In some cases adducts between nitroso compounds and metallo fragments are formed spontaneously by ligand repla~ement.~~ When using nitroso compounds as spin traps for organometallic radicals, one should be aware that nitroso compounds readily undergo electron-transfer reactions with electron-rich organometallic compounds and that nitroso anion radicals are therefore common by-products in spin trapping reactions. Spin trapping using nitroso compounds has recently been proven54 for the detection of a Rh-centred metallo radical formed during sonolysis of an organometallic rhodium compound. Surprisingly, spin traps are not decomposed by ultrasonic irradiation to a great extent which renders this a promising technique for studying radical processes initiated by ultrasound.Electron transfer between the organometallic complex and the nitroso compound may also be achieved photochemically as we have shown55 recently for the reaction of triply bonded organometallic Cr and Mo compounds with both aliphatic and aromatic nitroso compounds. In addition, spin adducts of ” D. Rehorek and E. G.Janzen, Can. J. Chem., 1984,62, 1598. 52 A. Hudson, M. F. Lappert, P. W. Lednor, and B. K. Nicholson, J. Chem. Soc., Chem. Commun., 1974. 966; A. Hudson, M. F. Lappert, and B. K. Nicholson, J. Chem. Soc.. Dalron Trans., 1977,551. 53 E. Dinjus, D. Walther, R. Kirmse, and J. Stach, J. Organornet. Chem., 1980, 198, 215.54 D. Rehorek, S. Di Martino, S. Sostero, D. Traverso, and T. J. Kemp, Inorg. Chim. Acta, 1990, 178, 1. 5s D. Rehorek, S. Di Martino, S. Sostero, 0.Traverso, and T. J. Kemp, J. Prakt. Chem., in press. 35 1 Spin Tt upping of Inorgunic Radicals Table 3 EPR htperjne coupling ton,tunt, und g-culuec of .\pin adducts of metul-coltred rudiculs M ith nitroso cornpound Spin Radical 'Co(N)Z, trup MNP Soh ent methanol W/mT 187 uM/mT 1 11 ('9c~) g 2 0054 Ref 72 'MO(CN)< ND CHzCl2 0 38 3 56 (95 "MO) 19823 73 'W(CN)< ND CHzCl2 0 959 74 'M O( CO)3Cp ND CHzClz 145 0 40 (y5 "Mo) 2 005 52 'Re(CO)5 ND CHIC12 1403 3 80 (Is5 ls7Re) 2 0098 75 'Re(C0)4PPh3 ND CH2Cl2 139 3 43 (Is5 l8'Re) 2 0087 75 Mn( CO) ND CH2Cl2 I 59 0 88 (5sMn) 2 006 52 'R h(CO)2 ND CHIC12 1 60 0 42 ( lo3Rh) 2 012 76 'Fe(C0)2CpMe5' ND benzene 1 708 2 0064 77 'Fe(C0)zCp 'Cr(C0)4CI ND ND toluene ccl4 1787 183 1 86 (53Cr) 2 0052 19983 77 78 'UCP3 ND THF 1 306 2 0076 79 'Ni(bipy)Et ND benzene 125 2 0130 80 'ZrCp2H MNP benzene 137 043 (91Zr)' 81 'Al(TPP)I BNB benzene 1 20 o 19 (27~1) 2 0049 82 'Pd(PPh3)N3 ND CH2C12 1 574 0 801 (105Pd)y 2 0096 83 "Cp = cyclopentadienyl anion PPh3 = triphenyl phosphine additional splitting up = 0 71 mT (31P) 'CpMe5 = pentamethylcyclopentadienyl anion bipy = 2 2 bipyridine Et = ethyl 'Additional splitting OH = 0 15 mT TPP = tetraphenylporphyrin dianion Additional splitting nP = 0 432 mT metallo radicals to nitroso compounds may undergo photolytic cleavage of the metal-N bond forming nitroso anion radicals The main disadvantage of using nitroso compounds as spin traps in photoreac- tions is their photochemical instability 56 Symmetric aminoxyls are formed by NO-carbon bond cleavage and subsequent addition of the alkyl (or aryl) radical to the intact nitroso compounds Owing to their high stability, symmetric aminoxyls have very intense EPR signals which may mask weaker signals of other spin adducts Therefore, direct excitation of the nitroso spin trap should be avoided by using band-pass filteres As discussed above for nitrone spin traps, nitroso compounds may also react with electronically excited sensitizers Thus, reductive quenching of zinc porphyrins and oxidative quenching of 9,f 0-dicyanoanthracene by ND has been observed 42 Acknowledgement The author would like to express his sincere thanks to Professors Edward G Janzen (University of Guelph/Canada), Terence J Kemp (University of Warwick/U K ), and Horst Hennig (University of Leipzig) for their continuous support and encouragement Dr Helmut Knoll and Anna Schleitzer are thanked for their assistance in luminescence measurements 56 D Rehorek and E G Janzen J Photothem 1986 35 251 C Chatgilialoglu and K U Ingold J Am Chem Soc 1981 103 4833 Rehorek 57 T Uda, A Kazusaka, R F Howe, and G W Keulks, J Am Chem Soc ,1979,101,2758 T Matsuzaki, T Uda, A Kazusaka, G W Keulks, and R F Howe, J Am Chem Soi , 1980, 102, 751 I 5y E G Janzen, H J Stronks, D E Nutter, E R Davis, H N Blount, J L Poyer, and P B McCay, Ccrri J Chrni , 1980, 58, 1596 6o C Mottley, H D Connor, and R P Mason, Bioihem Bioph,s Res Commun, 1986,141,622 61 D Rehorek and E G Janzen, Z Cliem, 1985.25,451 62 D Rehorek, E G Janzen.and Y Kotake, J Prakt Cliem, in press 63 D Rehorek, unpublished results fi4 M P Crozet and P Tordo, J Am Cheni Soc ,1980,102,5696 65 I I Kandror, R G Gasanov, and R Kh Freidlina, Terraliedron Lett, 1976, 1075 66 D Mulvey and W A Waters, J Chem Soc , Perkin Trany 2, 1974,771 67 C F Chignell, B Kalyandraman, R P Mason, and R H Sik, Phorocliem Pliotohrol, 1980.32, 563 N H Anderson and R 0 C Norman, J Chem Sot ,Farah Trans 1, 1977,73,776 "B Kalyanararnan.E Perez-Reyes, and R P Mason, Tetrahedron Lett. 1979,4809 7" H Karlsson and C Lagercrantz, Acta Chem Stand, 1970,24,3411 'I D Rehorek and E G Janzen, Pol~lieilron,1984, 3,631 72 D Mulvey and W A Waters, J Clieni Soi ,Perkin Trany 2, 1974, 666 73 D Rehorek, J Salvetter, A Hantschrnann, H Hennig, Z Stasicka, and A Chodkowska Inorg Chirn Ac fa. 1979,37. L471 74 T J Kemp, M A Shand, and D Rehorek, J Chem Soc ,Dalton Trans, 1988,285 75 D Rehorek. S Di Martino, and T J Kernp, J Prakt Chem, 1989,331,778 76 S DI Martino, S Sostero, D Traverso, T J Kemp, and D Rehorek, Inorg Chin? At fa, 1989, 165. L13 "S Di Martino, S Sostero, 0 Traverso, D Rehorek, and T J Kernp, Inorg Chin1 4tti1, 1990, 176 107 "R G Gasdnov and R Kh Freidlina, Dokl Akacl Nauk SSSR, 1980,254, 113 7y E Klaehne, C Giannotti. H Marquet-Ellis, G Folcher, and R D Fischer, J Orguriomrt Clieni 1980,201,399 E Dinjus, D Walther, R Kirmse, and J Stach, Z Anorg Allg Cliem, 1981,481, 71 " R G Gasanov, L I Strunkina, and E M Brainina, Dokl Akad Nauk SSSR, 1987,294, 1146 82 S Tero-Kubota, N Hoshima, M Kato, V L Goedken, and T Ito, J Chem Sot , Clieni Conintun. 1985,959 83 H Hennig, R Stich, D Rehorek, P Thomas, and T J Kemp, Inorg Chrm Aim, 1988, 143, 7

ISSN:0306-0012    

DOI:10.1039/CS9912000341

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

Chem. SOC.Rev., 1991, 20, 355-390 Recent Progress on Conducting Organic Charge-Transfer Salts By Martin R. Bryce DEPARTMENT OF CHEMISTRY. UNIVERSITY OF DURHAM, DURHAM, DHI 3LE 1 Introduction The past two decades have witnessed unabated interest in the synthesis and characterization of organic charge-transfer (CT) salts that display unusual solid- state properties. Metallic conductivity and superconductivity are the most glamorous phenomena to have been discovered in these materials, and recently attention has also been directed to the novel magnetic and optical properties which CT salts can display. This is a multi-disciplinary field that challenges the skills of synthetic chemists, crystallographers, applied physicists, theoreticians, and materials scientists.The motivation behind research on these exotic materials is twofold: (i) there is a wealth of fundamental solid-state chemistry and physics to be uncovered and (ii) there are potentially many far-reaching technological applications within the arena of molecular electronic devices. The same is true of conducting polymers (e.g. polyacetylene and polythiophene) and progress in the two areas (CT salts and conducting polymers) has largely gone hand-in-hand. Herein, we are concerned with CT salts that are electronic con- ductors and we will highlight some of the materials that are currently at the fore- front of attention. The vast majority of work has involved studies on single crystals. However, innovative results are emerging with Langmuir-Blodgett film materials.It is, therefore, timely to devote a significant part of this review to this new area. This article is written by a chemist, and the approach to the subject will, there- fore, be different from that of a scientist from another discipline studying the same materials. Several review articles,' books,2 and detailed conference proceedings ((I) J. B. Torrance, Acc. C/7en7. Rex, 1979, 12, 79; (h)D. Jerome and H. J. Schultz, Ah.. fhjx, 1982, 31. 299; (c,) K. Bechgaard and D. Jerome, Sci. Am., 1982, 247, 52; (d)F. Wudl, A(,c..Clirni. Rev., 1984, 17. 227; (e) M. R. Bryce and L. C. Murphy, Nnturr, 1984. 309, 119; (f)T. J. Marks, Scienc~~.1985, 227, 881: (g) J. M. Williams, M. A. Beno, H. H. Wang, P. C. W. Leung, T. J.Emge, U. Geiser. and K. D. Carlson, AN. Chem. Res., 1985, 18, 261; (11) D. 0.Cowan and F. M. Wiygul. Chem. Eng. New.\. 1986. 64(29), 28; (i) P. M. Chaikin and R. L. Greene. fh?.sics Todq., 1986, 39(5), 24; (j) J. M. Williams. H. H. Wang, T. J. Emge, U. Geiser, M. A. Beno, K. D. Carlson, R. J. Thorn, A. J. Schultz. M.-H. Whangbo, frog. Inorg. Chem., 1987. 35, 51; (k) D. 0. Cowan, in 'New Aspects of Organic Chemistry 1'. ed. Z. Yoshida. T. Shiba, and Y. Oshiro, V.C.H. Publishers, New York, 1989, 177. ((I) J. R. Ferraro and J. M. Williams, 'Introduction to Synthetic Electrical Conductors'. Academic Press, London, 1987: (h) T. lshiguro and K. Yamaji, 'Organic Superconductors', Springer-Verlag. Berlin, 1990. (0) Proceedings of ICSM 1988, Santa Fe, published in Sjxrh.Mer., 1988 and 1989, 27 29; (h) Proceedings of ICSM 1990, Tiibingen, published in Syntli. Met., 1991, 41-43; (c) 'The Physics and Chemistry of Organic Superconductors', ed. G. Saito and S. Kagoshima, Springer-Verlag. Berlin. 1990: ((4 'Organic Superconductivity', ed. V. Z. Kresin and W. A. Little, Plenum Press, New York. 1990. Recent Progress on Conducting Organic Charge- Transfer Salts are recommended for comprehensive coverage of the field. 2 Historical Perspective In 1954 Japanese workers reported that an unstable perylene-bromine salt was c~nducting,~and during the 1960s many salts of TCNQ were found to be organic semicond~ctors.~ Thus the foundations were laid for the discovery, in 1973, that the crystalline 1:1 CT complex formed by the donor tetrathiafulvalene (TTF) (1) and the acceptor tetracyano p-quinodimethane TCNQ (14) exhibited metallic behaviour (crt= 500 Scm-', omax= lo4 Scm-' at 59K).' This was a fulfilment of the prediction made in 1911 that organic solids might exhibit electrical conductivity comparable with that of metals,' and the science of organic conductors was now truly inaugurated.Also, in the 1970s it was found that the conductivity of polyacetylene could be increased by ca. 13 orders of magnitude by doping it with various donor or acceptors species;' this initiated a major research effort into the study of conjugated organic polymers. At about the same time, two other important classes of synthetic metals emerged: uzz., polymers of certain main group elements, notably (SN),," and linear-chain chelated transition metal compounds where intra-chain overlap involves the ligand n-system as well as metal-metal interactions. l1 Major landmarks in the development of organic conductors based on CT complexes and ion radical salts are presented below (Table 1) in chronological order.Some of these compounds will be discussed in more detail in this review. Variable temperature conductivity values at ambient pressure for a range of highly conducting materials are shown in Figure 1. Table 1 Year Discovery 1954 Perylene-bromine salt. the first conducting molecular c~mpound,~ CT,,= ca 1Scm-'. 1962 Semi-conducting salts of TCNQ rep~rted.~ 1973 TTF-TCNQ prepared the first organic metal; B,,= 5OOScm-'; TM-[at 53 K ' 1974 Tetraselenafulvalene (TSF)-TCNQ prepared.orl= 7W800Scm and TM-lat LU 40K. The metallic state is stabilized and transport properties are dominated by the donor stack.12 H Akamatsu, H Inokuchi, and Y Matsunaga, Nature, 1954,173, 168 L R Melby, R J Harder, W R Hertler, W Mahler, R E Benson, and W E Mochel, J Am Chem Soc, 1962,843374 TTF was first synthesized independently by two groups (a)F Wudl, G M Smith, and E J Hufnagel, J Chem Soc, Chem Commun, 1970, 1453, (b) S Hunig, G Kiesslich, D Scheutzow, R Zahradnik, and P Carsky, Int J Sulphur Chern, Port C, 1971,109'J Ferraris, D 0 Cowan, V V Walatka, and J H Perlstein, J Am Chem Soc , 1973,95,948 H N McCoy and W C Moore, J Am Chem Soc, 191 1,33, 1273 H Shtrakawa, E J LOUIS,A G MacDiarmtd, C K Chiang, and A J Heeger, J Chem Soc , Chem Commun, 1977,578 lo M M Labes, P Lowe, and L F Nichols, Chem Rev, 1979,79, 1'' (a) A E Underhill and D M Watkins, Chem Soc Rev, 1980, 9, 429, (6) P I Clemenson, Coord Chem Rev, 1990,106,171 356 Bryce Copper K-(BEDT-TTF)~ ''.\ Cu(NCS)2 Doped polyacetylene TTF-TCNQI -~~~ 1 10 lo2 103 Temperature / K Figure 1 Variable temperature conductivity values at ambient pressure for a range oj highlyconducting materials 1975 HMTSF-TCNQ: metallic character retained down to < 1 K; increased dimensionality due to close interstack Se N contacts.13 1978 HMTSF-2S-DMTCNQ: TM-~suppressed under pressure; o = lo5Scm-' at 1 K and 10kbar.l4 1979 TMTTF-tetrahalo-p-benzoquinones: metallic behaviour without TCNQ, or deriva-tive, as the acceptor." 1980 (TMTSF)2 X salts: organic superconductivity first reported at 0.9 K and 12kbar for X = PFZ, and at 1.4Kand ambient pressure for X = C10i.16 1982 (BEDT-TTF)z c104 (1,1,2-trichloroethane)05: metallic conductivity over the temperature range 298-1.4 K in a sulphur-based system.' ' 1983 (BEDT-TTF)z ReOo: the first sulphur-based organic superconductor; T, = 1.4K at 4 kbar.'' 1984 P-(BEDT-TTF)213: T, at 1.4K and ambient pressure.l9 1986 P-(BEDT-TTF)213: T, raised to 8 K under anisotropic pressure." 1986 TTF[Ni(dmit),],: the first superconducting system based on a CT salt of a TI-anion molecule; T, = 1.6 K at 7 kbar.21 Recent Progress on Conducting Organic Charge- Transfer Salts 1987 Cu(2,5-DMDCNQI)2 extremely high conductivity for a radical ion salt, metallic behaviour between 295-1 3K with CJ = 5 x 105Scm at 3 5K 22 1987 (DMET)2 Au(CN)2 superconductivity first observed in a salt of an unsymmetrical donor, T, = 1 1K at 2 5 kbar 23 1987 Me4N[Ni(dmit)2]2 superconductivity in a n-anion molecule having a closed shell cation, T, = 5 OK at 7kbar 24 err,1988 TTeF-TCNQ synthe~ized,~~ = ca 2000Scm 1988 (MDT-TTF)z AuI2 sulphur-based superconductivity observed with an unsym- metrical donor, T, = 3 5 K at ambient pressure 1988 K-CBEDT-TTF), CU(SCN)~ ambient pressure superconductivity at 10 4K 27 1990 (BEDO-TTF)2 I3 the first organic metal of an oxygen-containing donor, 28 CJ,,= IS280Scm 1990 (BEDO-TTF)3 CU~(NCS)~ superconductivity observed in an oxygen-contdining donor, T, = 1 OK at ambient pressure 29 1990 K-(BEDT-TTF)~ CU[N(CN)~] X, X = Br, CI currently the highest T, organic superconductors, T, = 11 6K at ambient pressure for X = Br,30u T, = 12 5K at 0 3kbar for X = C1 30b E M Engler and V V Patel, J Am Chem SOC,1974,%,7376 l3 A N Bloch, D 0 Cowan, K Bechgaard, R E Pyke, and R H Banks, Phys Rev Lett, 1975,34, 1561 l4 C S Jacobson, K Mortensen, J R Anderson, and K Bechgaard, Phys Rev, 1978, B18,905 l5 J B Torrance, J J Mayerle, V Y Lee, and K Bechgaard, J Am Chem SOC,1979,101,4747 I6(a) D Jerome A Mazand, M Ribault, and K Bechgaard, J Phys Lett, 1980, 41, L95 (h) K Bechgaard, C S Jacobsen, K Mortensen, H J Pedersen, and N Thorup, Solid State Commun , 1980, 33, 11 19 G Saito, T Enoki, K Toriumi, and H Inokuchi, SolidState Commun ,1982,42,557 Is S S P Parkin, E M Engler, R R Schumaker, R Lagier, V Y Lee, J C Scott, and R L Greene, Phjs Rev Lett, 1983,50,270 B Yagubskii, I F Schegolev, V N Laukhin, P A Karatsovnik, M V Karatsovnik, A V Zvarykina, and L I Buravov, J E T P Lett (Engl Trans ), 1984,39, 12 ’O J E Schirber, L J Azevedo, J K Kwak, E L Venturini, P C W Leung, M A Beno, H H Wang, and J M Williams, Phbs Rev B, 1986,33,1987 ” L Brossard, M Ribault, L Valade, and P Cassoux, Phjsica B, 1986, 143,378 ”A Aumuller, P Erk G Klebe, S Hunig, J U von Schutz, and H -P Werner, AngeM Chem ,Int Edn Engl, 1986 25,140 ”K Kikuchi, K Murata, Y Honda, T Namiki, K Saito T Ishiguro, K Kobayashi, and I Ikemoto, J PIIJSSOL Jpn ,1987,55,3435 24 A Kobayashi, H Kim, Y Sasaki, H Kobayashi, S Moriyama, Y Nishio, K Kajita, and W Sasaki, Chem Lett, 1987, 1819 25 (a) M D Mays, R D McCullough, D 0 Cowan, T 0 Poehler, W A Bryden, and T J Kistenmacher, Solid State Commun, 1988, 65, 1089, (6) D 0 Cowan, M D Mays, T J Kistenmacher, T 0 Poehler, M A Beno, A M Kim, J M Williams, Y K Kwok, K D Carlson L Xiao, J J Nuova, and M -HWhangbo, Mol Crjst Liq Cryst, 1990,181,43 26 G C Papavassiliou, G A Mousdis, J S Zambounis, A Terzis, A Hountas, B Hilti, C W Mayer and J Pfeiffer, Sjnh Met, 1988, B27, 379 ”H Urayama, H Yamochi, G Saito, K Nozawa, T Sugano, M Kinoshita, S Sato, K Oshima A Kawamoto, and J Tanaka, Chem Lett, 1988,55 ”(a)F Wudl, H Yamochi, T Suzuki, H Isotalo, C Fite, H Kasmai, K Liou, G Srdanov, P Coppens, K Maly and A Frost-Jensen, J Am Chem Soc, 1990, 112, 2461, (b) F Wudl, H Yamochi T Suzuki H Isotalo C Fite K Liou H Kasrnai, and G Srdanov, in ref 3c, p 358 ”M A Beno, H H Wang, A M Kini, K D Carlson, U Geiser, W K Kwok, J E Thompson J M Williams J Ren and M -H Whangbo Inorg Ciiem, 1990,29 1599 ”(a) A M Kim, U Geiser, H H Wang, K D Carlson J M Williams, W K Kwok K G Vandervoot J E Thompson, D L Stupka, D Jung, and M -H Whangbo, Inorg Chem 1990, 29 2555, (h) J M Williams, A M Kim, H H Wang, K D Carlson, U Geiser, L K Montgomery, G J Pyrka D M Watkins J K Kommers S J Boryschuk, A V Crouch, W K Kwok, J E Schirber D L Overmyer D Jung, and M -H Whangbo, Inorg Chem, 1990,29,3274 358 Bryce (12) DMET (13) MDT-TTF R NC' N+NFN R (14) R=H TCNQ (16) Ni(dmit), (17) R= H DCNQI(15) R IMe 2.5-DMTCNQ (18) R = Me 2.5-DMDCNQI 3 Physical Concepts Rigorous treatment of this topic is available A few basic points outlined here are of particular relevance to chemists engaged in the design and study of new materials.Conducting CT salts are composed of highly ordered arrays of donor [e.g.(1)-( 13)] and acceptor [e.g.(14)-(18)] species, one or both of which must be a radical ion which is thermodynamically stable. The materials can be either single-chain conductors, e.g. (TMTSF);' X-salts where the anion is a closed shell species, or two-chain conductors, e.g.TTF+'-TCNQ -charge-transfer complexes in which both components are open shell molecules. The presence of a supermolecular orbital provides a mechanism for metallic delocaliza- tion of electrons, with the width of the conduction band dependent upon interactions between molecular orbitals on neighbouring molecules. However, the extensive interaction of molecular orbitals is not enough in itself to produce metallic (or even semiconducting) properties in a material. The occupancy of these energy bands is critically important: when the energy gap between the valence band (derived from the HOMO of the molecules) and the conduction band (derived from the LUMO of the molecules) is large, the material is an insulator. As the band gap decreases, thermal excitation of electrons from the valence band to the conduction band is possible and the material is an intrinsic semiconductor.Metallic behaviour requires partially filled bands in which it is possible for a large number of electrons to move easily into infinitesimally higher energy states within the band (Figure 2). The highest occupied state is called the Fermi level and it is the electrons in energy states very near to this level that influence the physical properties. In CT complexes these states are derived from the HOMOS of the donor species and the LUMOs of the acceptor species. The temperature dependence of conductivity in organic (and inorganic) metals is dominated by the interaction (scattering) of electrons with vibrations of the atomic lattice (phonons); as the temperature is lowered there are fewer lattice vibrations which, in effect, increases the intermolecular orbital overlap, so the conductivity increases. In contrast to this, the conductivity of a semiconductor 359 Recent Progress on Conducting Organic Charge- Transfer Salts iilFilled levels insulator semiconductor metal Figure 2 Band structure of a solid; this determines the electrical properties.Eg is the energy gap between occupied and empty states decreases as the temperature is lowered, because there is less energy available to excite charge carriers across the band gap. Superconductivity (the passage of an electric current without resistance) as described by the Bardeen, Cooper, and Shrieffer theory for inorganic materials, is caused by the highly coordinated motion of electron pairs (Cooper pairs.)31 This mechanism is clearly distinct from metallic conduction for which individual electrons are the charge carriers.Electron pairing can be driven by phonons below a critical temperature, T,, if a strict set of structural and electronic energy conditions is fulfilled; at higher temperatures the thermal agitation of the electrons dissociates the pairs and superconduction is lost. Currently it is not clear whether or not BCS theory can adequately explain organic superconduction.2b In order to obtain efficient intermolecular interactions in a CT salt, the structural and redox properties of the constituent molecules are clearly of crucial importance.(There are, of course, other important design constraints, and these have been discussed recently. lk) The donor and/or acceptor moieties are typically planar (or nearly planar) molecules that form either segregated stacks or planar sheets. Thus the resultant electrical properties of the salt are highly anisotropic, i.e. associated with a unique direction within the crystal. Many organic metals are therefore termed quasi-one-dimensional metals. The behaviour of such a system was discussed over thirty years ago when Frohlich32 and Peierls33 pointed out that at low temperature a quasi-one-dimensional metal could not sustain long-range order but would be unstable with respect to lattice distortions. (The chemical analogue is the well-known Jahn-Teller distortion.) The degree of 31 J.Bardeen, L. N. Cooper, and J. R. Schrieffer, Phys. Rev., 1957,108, 1175. 32 H. Frohlich, Proc. R. SOC.London., Ser. A, 1954,223,296. 33 R. E. Peierls, ‘Quantum Theory of Solids’,Oxford University Press, London, 1955. Bryce -c--c-+ -v .-c--c. -A -c--4-+ -v -c--c-+--A-+ -v-c-af-T---4--AT (8) (b) (c) (a Figure 3 Several possible conjigurations of a one-dimensional stack of molecules with one electron in the HOMO: (a) metal with uniform lattice constant a: (6) insulator with dimerization resulting from a Peierls transition: (c) SD W insulator with spin (T) periodicitycaused by Coulomb interactions; (d) insulator with periodicity caused by ordering of non-symmetric anions (Redrawn from R.L. Greene and G. B. Street, Science, 1984,226,651) instability depends upon the level of band filling (i.e., in simple terms, the number of radical ionic and neutral molecules in the stack). A half-filled band has no neutral molecules in the stack, and as each molecule has an unpaired spin there is an electronic driving force for spin pairing. When molecules dimerize in this way there is concomitant creation of an energy gap between bonding and antibonding energy levels and so metallic conduction is lost. This is known as the Peierls distortion and the alternating regions within the lattice of higher and lower charge density result in the generation of a charge density wave (CDW). When neutral molecules are present in the stack, the resulting band will be less than half filled. For non-integral fractions of band filling the CDW periodicity will no longer be commensurate with the lattice and the CDW will be free to translate to new positions, so acting as a charge carrier.34 For the complex TTF-TCNQ, the bandwidths of the TTF and TCNQ stacks are quite different, and although the CDW and the lattice periodicities are incommensurate (the charge on each stack is non-integral, uiz.0.59) ordering of the CDWs on different stacks, relative to each other, provides a mechanism for the Peierls instability which leads to an insulating state below ca. 40K.35 In low-dimensional conductors, metal-insulator transitions can also be driven by either spin density waves (SDW) or anion ordering (Figure 3).The former instability was first identified in the salt (TMTSF)* PF6: the electron spins order antiferromagnetically (alternating spin up and spin down) which restricts electron mobility by introducing an energy gap in the electronic band.36 Application of external pressure on crystals of TMTSF salts may suppress this transition, and for some systems, result in a superconducting ground state.lc Why a magnetic (SDW) transition and not a Peierls (CDW) transition should occur in 34 D. Jerome and H. J. Schultz, Springer Ser. Solid State Phys., 1981,23,239. 35 P. Coppens, V. Petricek, D. Levendis, F. K. Larsen, A. Paturle, G. Yan, and A. D. LeGrand, Phy.~ Rev. Lett., 1987.59, 1695. 36 W. M. Walsh, F. Wudl, E.Aharon-Shalom, L. W. Rupp, J. M. Vandenberg, K. Andres, and J. B. Torrance, Phys. Rev. Lett.. 1982,49, 885. 36 1 Recent Progress on Conducting Organic Charge- Transfer Salts these salts is not clear at present The effect of anion ordering on metal-insulator transitions was also first appreciated from studies of a series of (TMTSF)* X salts (X = PF6, AsF6, Re04, C104 etc) ''I At room temperature the anions have random orientations, whereas at low temperature they orient to give three- dimensional superstructures If the periodicity of this superstructure matches that of the electrons at the Fermi level, an energy gap will open 4 New Conducting CT Materials; Design, Synthesis, and Properties A. General Aspects -The global challenge for chemists working on conducting CT salts is the preparation of new molecular systems which meet the stringent requirements, both at the intra- and inter-molecular level, for high conductivity or superconductivity Since our ability to induce molecules to pack within a crystal lattice in a prescribed fashion (eg in dimer pairs or in sheets) is still very limited, the rational design of new conducting salts is essentially restricted to controlling the key properties of the individual molecules which can be deduced a prrorz, eg planarity, ionization potential/electron affinity, extent of conjugation [Some control over intermolecular arrangements in CT salts can be obtained in the form of Langmuir-Blodgett films (Section 6)] When a new donor or acceptor molecule is in hand it is invariably necessary to crystallize several salts and screen their properties (e g conductivity, magnetic susceptibility, and X-ray crystal structure)-only then will the effects of the new structural modification become apparent t The desire to prepare CT systems with increased dimensionality of structural, and hence electrical, properties has been prevalent in the work of many groups This stems from the fact that the superconducting salts of TMTSF (4) 37 and BEDT-TTF (8) 'I were quickly recognized to be two-dimensional materials (not one-dimensional like TTF-TCNQ) with a strong network of inter-stack, as well as intra-stack, chalcogenxhalcogen interactions playing a key role in eliminating the Peierls transition While TMTSF salts do form donor stacks with close Se Se contacts, the structures of (BEDT-TTF);? X salts are dominated by short Inter-stack S S interactions rather than intra-stack interactions, leading to little or no columnar stacking, and, for some BEDT-TTF salts, conductivity values are higher in the 'sheet' rather than in the 'stack' direction (cf Figures 4 and 5) An increase in dimensionality can be encouraged in two ways (1) by placing polarizable heteroatoms (usually S, Se, or Te) at peripheral sites in the constituent donor or acceptor molecules, (11) by judicious (or serendipitous!) choice of the inorganic counterions for subsequent CT salt formation Three major goals of recent years have been (1) to stabilize the metallic state down to the lowest observable temperatures by thwarting the Peierls distortion, (11) to increase the critical temperature for the onset of superconductivity, (111) to t Two general methods are available for salt formation slow cooling of a saturated solution of the donor and acceptor (eg TTF-TCNQ) and electrolysis The latter method is particularly suitable when the anion is an inorganic species eg (TMTSF)2 X or (BEDT TTF)2 X Indeed this is the only method by which most of these salts are obtained 37 F Wudl J Am Cliem Soc 1981 103 7064 Figure 4 X-Ray crystal structure of (TMTSF):' PF; u1hic.h is a superconductor under pressure. Selenium atonis are shaded.The overlap of selenium n-orbitals along the stacks forms a conduction band rriith a M'idth of cu.1eV (Redrawn from N. Thorup, G. Rindorf, H. Soling, and K. Bechgaard, Acta. Crj-st., 1981, B37, 1236) discover new families of metallic and superconducting salts based on new donor or acceptor molecules that are structurally quite different from those already known. B. Metals and Superconductors Based on IT-Donors.-(i) Sulphur-bused Systems. Since the discovery in 1983 of superconductivity in salts of BEDT-TTF (8),'* sulphur-based systems have attracted the most attention. More than half the known ambient pressure superconductors are salts of BEDT-TTF (8).2b The first reported synthesis of the neutral donor (8) was by Cava et from dithiolate salt (19).39 This procedure has recently been irnpr~ved,~' and an alternative route from (24) is equally efficient41 (Scheme 1).The non-planar structure of donor (8) together with the large thermal vibration of the peripheral ethylene bridges impedes good IT-overlap along a face-to-face stacking axis. This fact, combined with prevalent intercolumn S -S networks, leads to increased dimensionality in the structural and transport properties of BEDT-TTF salts. The study of cation radical salts of BEDT-TTF (8) is complicated by the existence of multiple stoicheiometries and structural phases for the same anion. For example, four stoicheiometrically unique phases, termed Z, p, 8, and K phases, are known for the (BEDT-TTF)2 I3 salt! While the a-phase is a metal (TM-, at 135K) the other three phases are ambient pressure superconductors with T, 1.5K (8 K at 0.5kbar), 3.6 K, and 3.6 K, respectively.The intricate relationship between crystal packing and solid state properties of salts of (8) has been extensively studied, notably by American 1J342 and Japanese teams.43 Two 38 M. Mizuno, A. F. Garito, and M. P. Cava, J. Chem. Soc., Chem. Conimun., 1978, 18. 39 G. Steimecke, H.-J. Sieler, R. Kirmse, and E. Hoyer, Pkospltorous and Su&. 1979. 7.49. 40 K. S. Varma, A. Bury, N. J. Harris, and A. E. Underhill, Sjwlhesis, 1987, 837. 41 J. Larsen and C. Lenoir, Synthesis, 1989, 134. 42 A. M. Kini, M. A. Beno, K. D. Carlson, J. R. Ferraro, U. Geiser, A. J. Schultz, H. H. Wang. J. M. Williams, and M.-H. Whangbo, in ref. 3c, p. 334. "H. Urayama, H. Yamochi, G. Saito, S.Sato, A. Kawamoto, A. Tamaka, T. Mori, Y. Maruyama, and H. Inokuchi, Chem. Lelr., 1985, 463 and references therein. Recent Progress on Conducting Organic Charge- Transfer Salts W a3Q (b) Figure 5 X-Ray crystal structure of the superconductor K-(BEDT-TTF)~ CU[N(CN)~]B~ (a) stereoview of the donor layer showing Ka pa arrangement of orthogonal dimers, and intermolecular S S contacts shorter than 3.60 R;(b)polymeric anion layer (From reference 30a; figures kindly supplied by Professor A. M. Kini) common structural features of the (BEDT-TTF)2 X salts that are known to have T, values in excess of 10K [uiz. X = CU(NCS)~,~~ andCU[N(CN)~]B~,~'~ Cu[N(CN)J C130b have been identified.30 First, there is K-type packing of the BEDT-TTF molecules: this motif does not comprise stacks or sheets of the donors, but, instead, interacting dimers which are positioned approximately orthogonal to each other forming a conducting two-dimensional S S network (Figure 5a).Secondly, the anions in these three salts form insulating V-shaped polymeric chains (Figure 5b). The use of these anions followed froin the observation that T, for BEDT-TTF salts increases concomitantly with increased Bryce -2-Reagents: i, Na, DMF; ii, ZnCl2, Et4NBr; iii, PhC(0)CI; iv, NaOEt; v, NH40Ac, BrCH2CH2Br; vi, Hg(OAc),; vii, P(OEt)3; viii, N-chlorosuccinimide; ix, K + -SC(S)OPr'; x, HzS04 Scheme 1 linear anion length.'' Currently the highest T, organic superconductor at ambient pressure, is K-(BEDT-TTF)~ CU[N(CN)~]B~ with T, = 11.6K; 30a the isostructural chloride salt has T, = 12.5Kat 0.3kbar pressure30b (Figure 6).Both these salts were discovered by Kini, Williams, and co-workers at Argonne. The K-type dimer structure is not unique to (BEDT-TTF),; it is present in cation radical salts derived from three other donor systems, uiz. (BMDT-TTF)2 Au(CN)2 (metalli~),~~(M DT-TTF) 2 AuI2 (superconduct ing),2 6.4 and the mixed S,Se system (DMET)zAuBrz (s~perconducting).~~ It is notable that these last three donors are unsymmetrical molecules and it is tempting to suggest that this favours the formation of the K-structure. However, as pointed out by Kini et uI.,~~there are, as yet, no clear guidelines to follow in this respect. P. J.Nigrey, B. Morosin, J. F. Kwak, E. L. Venturini, and R. J.Baughman, Synfh. Met., 1986, 16, I. 45 A. M. Kin], M. A. Beno, D.Son,H. H.Wang, K. D. Carlson, L. C. Porter, U. Welp, B. A. Vogt,J. M. Williams, D.Jung, M. Evain, M.-H. Whangbo, D. L. Overmyer, and J. E. Schirber, Solid SINI~ Commun., 1989,69, 503. 46 K. Kikuchi, Y. Honda, Y. Ishikawa, K. Saito, 1. Ikemoto, K. Murata, H. Anzai, T. Ishiguro, and K. Kobayashi, Solid State Commun., 1988,66,405. 365 Recent Progress on Conducting Organic Charge- Transfer Salts 599 ....,.,..,....,.... .'..I.... I.... K-(ET)~CU[N(CN)~JB~ (D---------_____ 590 -1.---_-_-. '\ N "-"* -n7-z 125 K -s 597 K-( ET)2C~(NCS)2 .'"C Many other all-sulphur organic donors that are variants of TTF (1) or BEDT- TTF (8) have been reported re~ently.~ Only one of these, MDT-TTF (13), has provided new superconductors.Papavassiliou et al. first reported that the gold iodide salt (MDT-TTF)2 AuIz is a superconductor with T, =3.5K at ambient pressure.26 (It should be noted that neither of the parent symmetrical donors, r.e. TTF and BMDT-TTF, yield superconductors.) The discovery of this second family of superconductors based on an unsymmetrical donor [the first was DMET (12) (Section 4Bii)l has highlighted the need for efficient synthetic routes to unsymmetrical system^.^' While a few new approaches have been published, none is yet established as a general Studies on most unsymmetrical TTF and BEDT-TTF analogues are, therefore, still hampered by synthetic problems.A quite different skeletal variation to the donor system, vzz. vinylogous derivatives of TTF (1) and BEDT-TTF (8) have attracted attention for the following reasons: increased separation of the 1,3-dithiole rings should lead to considerably reduced on-site Coulombic repulsion in the dication redox state,49 and hence, a non-correlated type of conductivity may be possible. Yoshida et al. were the first to prepare extended TTFs of this type, e.g. compound (27).50We ''For a review of the synthesis of TTF systems. see A Krief, Tetrahedron, 1986,42, 1209 48 For examples see ((I) A M Kim, S F Tytko, J E Hunt, and J M Williams, Tetrtrhe~lronLett, 1987, 28, 4153, (b) K Lerstrup, I Johannsen, and M Jorgensen, Sinfh Mrr ,1988, 27, B9.(0 K Lerstrup. M Jorgensen, I Johannsen, and K Bechgaard, in ref 3c, p 383, (d) M R Bryce, A J Moore, D Lorcy, A S Dhindsa, and A Robert, J Clrem Soc .Clrenr Coniniun ,1990,470 49 Cf other highly conjugated redox systems discussed previously S Hunig and H Berneth. Top Cirri Chem ,1980, 92, 1 50 (a) Z Yoshida, T Kawase, H Awaji, I Sugimoto, T Sugimoto, and S Yoneda, Tertnhedron Lett, 1983, 24, 3469, (b)T Sugimoto, H Awaji, I Sugimoto, Y Misaki, T Kawase, S Yoneda. and Z Yoshida, Clreni Mater ,1989, 1. 535 Bryce J iv1. (36) (37) O Rengenfs: i, (Me0)2S02,AcOH, HBF,; ii, NaBH4; iii, Ac20, HBF,; iv, P(OMe)3,NaI; v, PPh,; vi, Et3N glyoxal; vii, (37), Bu"Li Scheme 2 have recently reported the synthesis of (28), the first vinylogous derivative of BEDT-TTF (8), along with several other new, extended donors, e.g.molecules (29)-(31)?' Two other groups have independently prepared compound (28).52.53 Our synthesis of (28) is shown in Scheme 2.51a,54 It is notable that this methodology is readily adaptable to the preparation of unsymmetrical systems, e.g. (30) and (31). Compound (30) has furnished the first X-ray crystal structure of a vinylogous donor of the TTF or BEDT-TTF families: molecule (30) is significantly more planar than BEDT-TTF (8) (Figure 7). Cyclic voltammetry has established the solution redox behaviour of the new 51 ([I) A. J. Moore, M. R. Bryce, D. Ando, and M. B. Hursthouse, J. Chem. Soc., Cfiem. Commuri., 1991, 320; (b)A. J. Moore and M. R. Bryce, manuscript in preparation.52 V. Y. Khodorkovskii, L. N. Veselova, and 0. Ya. Neiland, Kfiim. Geterotsikl. Soedin.. 1990, 130; (Chem.Ahs., 1990, 113.22868t). 53 T. K. Hansen, M. V. Lakshmikantham. M. P. Cava, R. M. Metzger, and J. Becher, J. Org. Chem., 1991,56,2720. A. J. Moore and M. R. Bryce, Syntiieris, 1991, 26. Recent Progress on Conducting Organic Charge Transfer Salts Figure 7 Molecular structure of the neutral donors (a) BEDT TTF (8) and (6) molecule (30) as determined by single crystal X ray analysis a view along the best plane formed bj the sulphur atoms ’Vote the marked deviation from planarity of BEDT TTF TTF (1) -BEDT-TTF(8) ---Donor (28) Potential (V) Figure 8 Cyclic voltammetric data for TTF (1) BEDT TTF (8) and vinylogous BEDT TTF derivative (28)(Ptelectrode versus Ag/AgCl, electrolyte EtN; PFs in CHZC12) vinylogues (28)-(3 1) They undergo two, reversible, single-electron oxidations (as do TTF and BEDT-TTF) and two important consequences of ‘stretching’ the conjugated n-system in this way are observed (I) both the first and second oxidation potentials of (28) are substantially lowered in comparison with BEDT- TTF, ze molecule (28) is a stronger donor than BEDT-TTF (8), and E2* (the radical cation-dication redox wave) for (28) and TTF (1) are seen at very similar potentials, (11) the difference between the two redox waves, A@, is also reduced in the vinylogue (Figure 8) 51 These stretched donors appear, therefore, to be very promising candidates for the preparation of new families of organic metals and superconductors that could be structurally and electronically very different from their predecessors Donor (38), which can also be viewed as a stretched TTF system, ha3 been Bryce Me MeMe-4e+s’AsMH* MeHMe (38)reduced form (383 oxidized form f Figure 9 X-Ray crystal structure of the conducting 1 :4complex formed by donor (38) and TCNQ (14), i.e.(38’)2+-(TCNQ)i-studied recently.55 An unusual 1:4 salt is formed with TCNQ, and magnetic susceptibility data establish that the donor is present in the dication redox state (38’). The X-ray structure reveals four TCNQ stacks surrounding each dicatjon molecule (Figure 9). The two 1,3-dithiolium cations in structure (38’) are almost orthogonal to the planar anthracene ring, which is a conformation very different from the butterfly shape adopted by the neutral donor (38).The conductivity of the salt derives from charge delocalization on the uniform TCNQ stacks; at room temperature orl= 60Scm-’ and this value is maintained down to ca. lOOK when the salt becomes insulating. This work establishes that (i) extended conjugation between the two heterocyclic rings reduces on-site Coulombic repulsion such that ”M. R. Bryce, A. J. Moore, M. Hasan, G. J. Ashwell, A. T. Frazer, W. Clegg, M. B. Hursthouse, and A. 1. Karaulov, Angew. Chem. Int. Ed. Engl., 1990, 29, 1450. Recent Progress on Conductin p Organic Charge- Transfer Salts in the solid state the dication redox form is stable and (11) interesting materials can result from the study of TTF analogues which have very distorted (non- planar) structures (11) Selenium-and Tellurium-conturning Sjstems In the late 1970s and early 1980s organoselenium donors provided a bonanza for the physicists the metallic state was stabilized to very low temperatures (eg HMTSF-TCNQ)13 and the first generation of organic superconductors [uiz the (TMTSF)2X salts] were discovered by Bechgaard l6 Much work concerned the specific role played by the inorganic anions with respect to the ground state of these salts, which can be either insulating, metallic, or superconducting The origin of the insulating ground stdte appears to be related to the symmetry of the anion For centrosymmetric anions, eg PF6 and AsF6, the ground state is due to formation of a spin density wave phase (T, <12K) while in the salts with non-centrosym- metric anions, eg Re04 and FS03, the metal-to-insulator phase transition is driven by ordering of the anions (see Section 3) In both cases, when single crystals are subjected to a finite external pressure the insulating phase is suppressed, leading to metallic behaviour and eventually a superconducting ground state Within the TMTSF series, the perchlorate salt remains the only one to superconduct at ambient pressure More recently, selenium-containing systems have been overshadowed by their sulphur bretheren in comparison, the selenium donors are invariably much harder to synthesi~e,~~ are insoluble, and the oxidation potential is raised by sequential selenium incorporation 48d 57 In contrast to BEDT-TTF (8), superconductivity has not been found in salts of the selenium analogues BEDS- TSF (10)58 or BEDS-TTF (1 1),59eken though the P-phase of the triiodide salt (BEDS-TTF2)13 is isostructural 59 with (BEDT-TTF2)13, which has T, = 8 K under pressure 2o Of the many mixed S,Se donors that are now known,60 only the DMET system (12) has so far yielded superconducting salts 61 The synthesis of DMET (12) is shown in Scheme 3 62 The desired product (12) has to be separated from the symmetrical products TMTSF (4)and BEDT-TTF (8), which is the problem inherent in the synthesis of an unsymmetrical donor by a cross-coupling reaction An efficient synthesis of selenone (43) had been reported previously 63 The gold "Review D Cowan dnd A Kini in The Chemistry of Organic Selenium dnd Tellurium Compounds Vol 2 ed S Patai J Wiley 1987 p 463 E M Engler F B Kaufman D C Green C E Klots and R N Compton J Am Ciiem Sac 1975 97 2921 (a) V Y Lee E M Engler R R Schurnaker and S S P Parkin J Ciiem Soc Ciiem Comniuti5" 1983 235 (h)R Kato H Kobayashi and A Kobayashi Clieni Lett 1986 785 '9 H W Wang L K Montgomery H Geiser L C Porter K D Carlson J R Ferrdro J M Williams C S Cdriss R L Rubinstein and J R Whitworth CIieni Mater 1989 1 140 dnd references therein 6" ((1) V Y Lee Sinrli Met 1987 20 161 (h)G C Pdpavassiliou G A Mousdis S Y Yiannopoulos V C Kdkoussis and J S Zambounis Sintii Met 1988 27 B373 "K Kikuchi Y Honda Y Ishikawa K Sdito I lkemoto K MUrdtd Y Anzdi I Ishiguro dnd K Kobayashi Solid Stare Comniuti I988 66 405 "K Kikuchi T Namiki I Ikernoto and K Kobayashi J Cliem Soc Chem Comrnun 1986 1472 63 A Moradpour V Peyrussen I Johannsen and K Bechgaard J Org Chem 1981 48 388 370 Me Me,+ i Me, 4'" __c Na' ii Me'Ie'NMe, Reugents: i, NaHSe, Et3N; ii, 2-chlorobutanone; iii, HzS04, HPF,; iv, NaHSe, AcOH; v, P(OMe)3 Scheme 3 cyanide salt (DMET)z Au(CN)z was the first superconductor in this series, reported in 1987 by Japanese workers.23 This material broke new ground as the donor (12) is an unsymmetrical molecule [it can be viewed as a hybrid of the TMTSF (4)and BEDT-TTF (8) molecules].A range of electronic transport properties, from insulator to superconductor, are displayed by siblings of the (DMET)2 X family, the highest T, being 1.9K at ambient pressure for the AuBr2 salt.61 Some DMET salts have the important K-type structure discussed earlier (Section 4.2a).Overall the behaviour of DMET superconductors is intermediate between that of the TMTSF and BEDT-TTF systems. Within the whole field of organic metals, the preparation of tetratellurafulvalene (TTeF) derivatives has proved to be one of the most daunting synthetic challenges. Work in the mid 1970s established that the substitution of the sulphur atoms of TTF with selenium generally had beneficial effects on the properties of the TCNQ salt: viz. increased room temperature conductivity and a lower M-I transition temperature; also, transport properties become dominated by the donor stack.Further advantages that tellurium substitution could afford (over and above selenium) were identified by Cowan et a/.:64 vi:. (i) the more diffuse p and d orbitals centred on Te should give larger conduction bandwidths (enhanced metallic conductivity) due to increased intrastack interactions, (ii) interchain interactions (i.e. two-and three-dimensional contacts) should be favoured which, in turn, should suppress Peierls distortions, (iii) the greater polarizability of Te should reduce on-site Coulombic repulsion in the dication state (i.e. TTeF2+)[cf compounds (28) and (38) discussed in Section 4.B(i)]. The first substituted TTeF derivatives were reported in 198265 and the long- awaited parent tellurium donor TTeF (5) arrived in 1987.66 The five-step, one- 64 D.0.Cowan, M. Mays. M. Lee. R. McCullough, A. Bailey. K. Lerstrup. F. Wiygul. T. Kistenmacher. T. Poehler. and L.-Y. Chiang, Mol. Cryst. Liq. Crysr.,1985, 125, 191. ((1) K. Lerstrup. D. Talham, A. Bloch, T. Poehler. and D. Cowan, J. Cl7em. Soc., Chm. Comn7uri., 1982,336; (h) F. Wudl and E. Aharon-Shalom. J. Am. Chem. SOL,.,1982.104. 11 54. 66 (0) R. D. McCullough. G. B. Kok. K. A. Lerstrup, and D. 0.Cowan, J. hi. Ciimi. Soc., 1987. 109. 41 15: (h)for an improved synthesis of TTeF see: R. D. McCullough. M. D. Mays, A. R. Bailey, and D. 0.Cowan, SjxrA. Met.. 1988,27. B487. 37 1 Recent Progress on Conducting Organic Charge- Transfer Salts HfnMe3 1,Il 1,11,111 [;+Tej H SnMe, Te (44) (5) Reagents I, Bu"L1, -78 'C, 11, -Ie, -78 "C, 111, BrzC=CBrZ Scheme 4 (45) pot synthesis of (5) developed in Cowan's laboratory is shown in Scheme 4.66b Electrochemical data, obtained by cyclic voltammetry, indicate an ionization potential for TTeF intermediate between that of TTF and TSF, with the difference between the first and second oxidation waves, A,??, following the expected trend TTeF < TSF < TTF.66" The conductivity of the complex TTeF- TCNQ is very high, crrt= 2200 t-300Scm-', and this value increases down to 2K (i.e.there is no Peierls distorti~n).~~ It has, therefore, been established that within the TXF-TCNQ series, electrical conductivity increases as X varies from S to Se to Te, and this increase is primarily caused by conductivity enhancement in the donor stacks.The X-ray crystal structure of TTeF-TCNQ reveals layers of donor molecules with close inter- and intra-stack Te... Te contacts. This mode of packing is quite different from that of TXF-TCNQ (X = S, Se) which possess no interstack chalcogen interactions, and is more like the structure found for most (TMTSF)2 X and (BEDT-TTF)2 X salts. TTeF-TCNQ is, therefore, considered to be a two-dimensional metal.2sb Some new TTeF derivatives, e.g. tetramethylhexamethylene-TTeF (49,have been reported,67 although tetramethyl-TTeF and molecules with mixed S, Te or Se, Te atoms in the fulvalene core are still unknown. However, attachment of tellurium atoms to the periphery of the TTF frame is relatively straightforward to accomplish by reaction of elemental tellurium with lithiated TTF species.68 Several alkyltelluro sidechains have been attached in this way, e.g.to yield molecule (46), which forms a semi-conducting complex with TCNQ.69 (iii) Oxygen-contammg Systems. Analogues of TTF (1) where any of the sulphurs in the fulvalene core have been replaced by oxygen are unknown. The first derivative with oxygen bonded directly to the periphery of TTF was the symmetrical BEDO-TTF molecule (9) reported by Wudl's group in 1989.70The synthesis is presented in Scheme 5.70 Contrary to the popular view that the more 67 A R Bailey. R D McCullough, M D Mays, D 0 Cowan, and K A Lentrup, Sinth ,Met, 1988, 27, B425 68 (a) E Aharon-Shalom, J Y Becker, J Bernstein, S Bittner, and S S Shaik, Tetrrrkedion Lett, 1985, 26, 2783, (h)J Y Becker, J Bernstein, S Bittner, J A R P Sharma, and L Shahal, Tettdietiron Lett, 1988, 29.6177, (0N Iwasavva, F Shinozaki, G Saito, K Oshima, T Mori, and H Inokuchi. Clieni Leli, 1988, 215 69 E Aharon-Shalom, J Y Becker, J Bernstein, S Bittner, and S Shaik, Shnth Met, 1985, 11.213 70 T Susuki, H Yamochi. G Srdanov. K Hinkelmann, and F Wudl, J Air? Chem Soc , 1989,111,3108 Bryce (9) BEDO-llF (51) (50) Reagents: i, 110°C,DMSO; ii, Brz; iii, llO°C, 25Torr; iv, H2Se;v, (MeO)3P Scheme 5 polarizable chalcogens Se and Te offered the best hopes for metallic and superconducting behaviour, [Section 4.B(ii)], Wudl reasoned that oxygen substitu- tion was a promising way forward.70 First, according to BCS theory of superconductivity, the lighter the component atoms within a series of donors, the higher should be the T, value in an isostructural salt.? Secondly, there may be a remote link between 0-containing organic superconductors and the 'high T,' copper oxide superconductors, where oxygen radical cations may be involved in the conduction process.$ Studies on the first organic metal formed by this new donor, (BEDO-TTF)2.s 13, suggest that oxygen substitution can increase the metallic bandwidth: the metallic state is stabilized down to low temperatures and the physical behaviour is reminiscent of a classical 3-dimensional metal.The X-ray structure reveals short interstack S S and S 0 contacts.28 One superconducting salt of BEDO-TTF has been discovered: this is (BEDO-TTF)3 Cu2(NCS)3 with T, = 1.06Kz9 Some unsymmetrically substituted ethylenedioxytetrathiafulvalene donors have been synthesized but their salts have not yet been ~haracterized.'~ 5 Metals and Superconductors Based on Organic n-Acceptors In recent years two families of n-acceptor molecules have provided exciting results: these are the metal(dmit)z and DCNQI systems.We will consider these in turn and then briefly mention a few other promising acceptor systems. A. Metal(dmit)z Acceptors.-These acceptors (52) which are metal complexes of the 1,2-dithiolene ligand (19) have three notable features: (i) the M(dmit)z system is nearly planar; (ii) there are ten peripheral sulphur atoms which can engage in t For a BCS superconductor, T, should vary inversely with the square root of the mass of the ions in the lattice.$ Other workers have commented on some general similarities between organic CT salts and CuO superconductors: e.g. both types of compound are layered, narrow band structures with close proximity of superconducting and antiferromagnetic ground state^.^'.^ (u) A. M. Kini, T. Mori, U. Geiser, S. M. Budz, and J. M. Williams, J. Chem. SOL..,Chem. Conimun.. 1990, 647; (h)T. Mori, H. Inokuchi, A. M. Kini, and J. M. Williams, Ckrm. Lett., 1990, 1279. Recent Progress on Conducting Organic Charge- Transfer Salts Reagent7 I NIC14 NBu4Br Scheme 6 Figure 10 X-Raj cri Ttal structure of TTF[Nl(dmlt 2]2 t.ze,ted in the plane of the TTF ringThin lines indicate S -* * S di3tantes shorter than 3 70d (From reference 72a) intra- and inter-stack interactions, and (111) the redox properties can be tuned by varying the central metal atom (Ni, Pt, Pd, etc) The metal complexes (52) are synthesized from ligand (19) as shown in Scheme 6 39 (cf Scheme 1) Metal(dm~t)~ anions have been studied in combination with alkali metal cations, open shell organic cations (eg TTF)" 72 73 and closed shell organic cations (eg tetraalkylammonium salts) 24 74 Superconductivity has been found in the last two classes of material It was first reported by Cassoux et af in 1984 that TTF[Ni(dmit)2]2 remained metallic at very low temperatures and that the crystal structure was essentially three-dimensional, comprising segregated stacks of donors and acceptors with close lateral interstack S-..S contacts (Figure 10) 75 The electronic structure of the system IS, however, still under discussion and, based on extended Huckel band calculations, Kobayashi er af concluded that the lateral interactions are very weak and the salt should be considered as a one-dimensional metal The origin of the stable metallic state in TTF[Ni(dmit)2]z lies not in the multi- dimensionality of the band structure, but rather in the nature of the multi-Fermi "(ri) M Bousseau L Valade J P Legros P Cassoux M Garbauskas and L V Interrank J Atii Chem soc 1986 108 1908 (h) P Cassoux L Valade J P Legros C Tejel J P Ulmet dnd L Brosbard in ref 3( p 22 "((I) L Brossard H Hurdequint M Ribault L Valade J P Legros and P Cassoux Sirirli Mtr 1988 27 B157 (h) L Brossard M Ribault L Valade and P Cassoux J Plilc.France 1989 50 1521 74 K Kajita Y Nishio S Moriyama R Kato H Kobayashi W Sasaki A Kobayashi H Kim and Y Sasdkl Solid Slcite Committi 1988 65 361 75 M Boussedu L Valade M F Bruniquel P Cassoux M Garbauskas L V Interrmte and K Kdsper NOULJ Cliim 1984 8 653 374 Bryce Figure 11 X-Ray crystal structure of the anion stuck superconductor Me4N +[Ni(drn~t)~]~ -(From reference 24) N,CN R3 R2R4@1: -"yJR' 0 (53) NC,"(54) R = alkyl. halogen, alkoxy Reagents I, TiC14,Me3SiN=C=NSiMe3 Scheme 7 surface.76 The salt shows superconductivity under pressure (T, = 1.6K at 7kbar).2 Replacement of Ni by Pd affords the salt TTF[Pd(dmit)z]z, one phase of which (the a' phase) has the highest T, value (6.5K at 20kbar) within the metal (dmit), series of corn pound^.^^ This material provides the first example of competition at very low temperatures between a charge density wave and superconductivity [cf: it 1s a spin density wave that operates in salts of TMTSF and BEDT-TTF].The partially oxidized TTF stacks, as well as the anion stacks, play an important role in the conduction process in these M(dm~t)~ salts, which are, therefore, two-chain conductors. Salts of Ni(dmit)z anions with closed shell cations have also provided very interesting materials: the mixed valence salt NMe4[Ni(dmit)2]2 is an anion-stack superconductor24 (T, = 5 K at 7 kbar) and the sodium salt Na[Ni(dmit)2] remains metallic down to at least 25mK.77 The X-ray crystal structure of the tetramethylammonium salt (Figure 11) reveals stacks of planar Ni(dmit), anions separated by channels which contain the inorganic cations.This is the reverse situation to that of the TMTSF salts (cJ Figure 4). B. N,N'-Dicyanoquinonediimine (DCNQI) Acceptors This family of acceptors, which has been developed by Hiinig and co-workers, '' A Kobayashi, H Kim, Y Sasaki, R Kato, and H Kobayashi, Solid Stale Comniuri , 1987.62, 57 " A Clark, A E Underhill, R H Friend, M Allen, I Marsden, A Kobayashi, and H Kobayashi. in ref 3c.p 28 Recent Progress on Conducting Organic Charge- Transfer Salts Figure 12 X-Ray crystal structure of Cu(2,5-DMTCNQI)2, nitrogen atoms are shaded (From reference 22) has three attractive features: (1) synthesis-DCNQIs (54) are produced in good yield in a one-pot synthesis from benzoquinones (53) (Scheme 7);78 (ii) structure-since the =NCN group is flexible, and sterically less demanding than the =C(CN)2 group, planarity of the DCNQI system is retained even upon tetrasubstitution 79 so close face-to-face stacking can occur in a wide range of derivatives [cf.tetrasubstituted TCNQ derivatives (14) have a buckled ring],79 (iii) redox properties-the acceptor strength of DCNQI IS similar to that of TCNQ and by appropriate substitution, the electron affinity can be finely tuned.Conducting DCNQI radical anion salts have been characterized with n-donors (e.g. TTF), organic cations (e.g. tetramethylammonium) and metal cations.80 The last class of compounds is the most interesting, especially those containing 2,5- disubstituted DCNQI acceptors, e.g. (1 8). Very high conductivities are achieved with a variety of metal monocations in salts of general formula M+ (2,5- DMDCNQI);' (M = Cu, Ag, T1, Ni, Na, K, Rb).80 The X-ray crystal structure of Cu(2,5-DMDCNQI)2 reveals that each metal atom is tetrahedrally coordinated to four DCNQI ligands by strong Cu 0.-N interactions (d = 1.99A) and the acceptor forms one-dimensional columns (just like TCNQ) with an intrastack distance of 3.2 A (Figure 12).22 Many of the M(2,5-X2DCNQI)2 salts possess this structure" and the distance between the cations (3.78-3.97A) is too large A Aumuller and S Hunig, Angeu Chant,In1 Ed Engl, 1984,23, 441 79 U Schubert, S Hunig, and A Aumuller, L~ebig~Ann Cheni, 1985, 1216 S Hunig and P Erk,Ah Muter, 1991,3,225 and references therein S Hunig, H Meixner, T Metzenthin, U Langohr, J U von Schutz, H-C Wolf, and E Tillmans.Ah Muter, 1990,2,361 Bryce ,CN Br Br NC (55)R'--R4 H or MeP for electron transport to occur along a metal ion chain; conductivity is, therefore, ascribed to a partially filled band formed by the LUMOs of the DCNQI molecules. The conductivity of the salt Cu(2,5-DMDCNQI)Z is ort= 103Scm-' and this value steadily increases as the temperature is lowered to 1.3K without any interruption from a Peierls distortion.22 Indeed at 3.5 K, CJ = 5 x lo5Scm-'.This conductivity is pseudo-three-dimensional with a value as high as 100Scm-' perpendicular to the stacking axis, with the N-..Cu.**N bridges providing a channel for conduction in this direction.82 Based on low temperature X-ray diffraction and X-ray photoelectron spectra, Kobayashi et al. conclude that the copper atoms are present in a mixed valence state, uzz. Cu' 3+, which could arise from admixture of the 3d orbitals on copper with the pn conduction band of the DMDCNQI ligand~.~~ The spin susceptibility is Pauli paramagnetic and larger than that of most organic metals. Electron spin resonance studies indicate that the unpaired electrons residing on the Cu ions order antiferromagnetically below 10K through the organic n-sy~tem.'~ Some other Cu(2,5-XY-DCNQI)2 salts (X,Y = Me, I; MeO, MeO) retain metallic behaviour at very low temperatures, while many in this series (X,Y = Me, C1; Me, Br; C1, C1; C1, Br; Br, Br) undergo a metal-to-insulator transition at low temperatures, due to extensive deformation of the tetrahedral coordination around copper and a CDW instability on the DCNQI C.Other Acceptors.-A few other quinonoid acceptor systems have been characterized recently. Several derivatives of compound (53, which is a hybrid of the TCNQ (14) and DCNQI (17) systems, yield semiconducting copper salts.85 Incorporation of sulphur into the acceptor ring system with a view to increasing dimensionality has led to heterocyclic TCNQ and DCNQI analogues, e.g.compounds (56) 86 and (57),*' respectively, which have provided some complexes with TTF of high conductivity. '* T Mori. K Imaeda, R Kato, A Kobayashi, H Kobayashi, and H Inokuchi, J Phi F Soc Jpn. 1987, 56,3429 83 (a) A Kobayashi, R Kato, H Kobayashi, T Mori, and H Inokuchi, Solid Slate Commun , 1987, 64, 64, (6) R Kato, H Kobayashi, and A Kobayashi, J Am Clzem Sor , 1989,111, 5224 84H-P Werner, J U von Schutz, H C Wolf, R Kremer, M Gehrke, A Aumuller, P Erk, and S Hunig, Solid State Commun, 1988,65,809*' M R Bryce and S R Davies, J Cliem Soc ,Cliern Cornmun , 1989,328 86 (u) K YUI, Y Aso, T Otsubo, and F Ogura, Bull Chem Soc Jpn, 1989. 62, 1539, (h) F Ogura, K Yui, H Ishida, Y Aso, and T Otsubo, in ref 3c, p 403 E Gunther, S Hunig, K Peters, H Rider, H G von Schenng, J -U von Schutz, S Soderhoh, H -P Werner, and H C Wolf, Angel$ Cliern ,Int Ed End, 1990,29,204 Recent Progress on Conducting Organic Charge- Transfer Salts substrate 10) lb) (C) [d) (e) Figure 13 Schematic representation of molecular arrangements of amphiphilic molecules in LB Jilms (a) X-ytructure, (h) Y-structure, (c) 2-structure, (d) mixed structure, (e) alternate layer structure 6 Langmuir-Blodgett (LB) Films of CT Salts General Aspects.-Currently there is burgeoning interest, from both the scientific and practical viewpoint, in using the LB technique to produce highly-conducting, ultra-thin films of CT salts.LB films offer the possibility for preparing organic structures with a greater level of control over the molecular architecture than in the corresponding single crystal materials.Practical interest stems from the electrical and optical properties of CT salts being available in the form of thin films which should be far easier to fabricate into electronic devices than would frail single crystals. It is, therefore, the combination of tailor-made, intermolecular chemistry within the LB film, together with interactions that can occur at the LB film-solid surface interface, that may provide novel advanced technological applications.88 Before reviewing the CT materials that form conducting films, a brief description of the fundamentals of LB films is appropriate as many chemists are not familiar with the technique.An amphiphilic compound (which may be a neutral molecule or a salt) is dissolved in a volatile organic solvent and spread at the air-water interface of an LB trough. Compression of the resulting monomolecular layer orients the molecules at the interface, and in this condensed, oriented state, transfer of the monolayer IS achieved onto a solid substrate (e.g. a glass slide) as it is dipped perpendicularly through the air-water interface. A single monolayer on the substrate will exhibit quasi two-dimensional character and subsequent deposition of film layers (which may consist of different chemical compounds) will assemble a three-dimensional superstructure, with the film thickness being controlled precisely by the number of dipping cycles.The structure of the amphiphile and the nature of the substrate (ie. hydrophobic or hydrophilic surface) will determine whether deposition occurs during both the up- and down-strokes, or only during either the up- or the down-strokes. Alternate-layer films comprising two different amphiphiles (e.g. a donor and an acceptor molecule) can be assembled in successive layers using a specially designed LB trough. The different molecular orientations that can result are shown in Figure 13. This degree of control over the orientation of individual molecules is clearly one of the most attractive features of the LB technique when applied to the formation of layers (which, in effect, are segregated stacks) of n-donor or n-(a)B Tieke, Arh Muter, 1990,2,222,(h)H Fuchs, H Ohst, and W Prass, Adv Muter, 1991.3, 10 Bryce acceptor species.Structural studies show that most LB films possess a crystalline, lamellar structure consisting of domains with diameters of 0.1-10 pm. Water molecules or adventitious ions may be present in the hydrophilic interlayers. Oxidation (doping) of the molecules within the LB film is frequently needed to yield a mixed-valence conducting system. This is most readily achieved chemically, by exposure of the film to a gaseous oxidant, e.g. iodine or bromine, for 1-2 min. Electrochemical oxidation of a film, during, or after, deposition on a conducting support is an alternative method, by which a wider range of dopant anions (C104, PFs, BF4, etc.) can be introduced into the structure.88a Film preparation takes place under ambient conditions where most amphiphilic organic molecules are stable; however, the monolayer is in a metastable state far removed from thermodynamic equilibrium.Consequently, spontaneous reorienta- tion of the molecules can occur on transfer to the substrate, resulting in structural defects in the assembled multilayer structure. A major short-coming of the LB technique is, therefore, the lack of long-term chemical, thermal, and mechanical stability of the films. B. Anion Radical Salts.-The materials used to fabricate conducting LB films have been collated el~ewhere.*~',~~ Initial work on CT salts concerned long-chain pyridinium cation-TCNQ anion radicals salts first reported by Barraud and co- workers in 1985.90 LB films of the 1: 1 charge-transfer complex N-docosyl- pyridinium-TCNQ (58), as deposited, exhibited low lateral conductivity. However, on doping with iodine vapour, higher conductivity films (art = ca. 10-'Scm-') were produced.Detailed investigations by Richard et revealed that the precursor film consisted of sheets of (TCNQ-')2 dimers with their molecular planes almost parallel to the plane of the substrate: the effect of the iodination process was to re-orient the TCNQ molecules so that they stood edge-on with their molecular planes and long axes roughly aligned to the substrate normal. Work on the same system by the group of Nakamura et has also shown that, as deposited, the films were not highly conducting (art = 10-5-10-7 Scm-'); these studies also indicated, however, that the orientation of the TCNQ molecules could be changed by varying the LB deposition conditions. Subsequently, this group reported on the properties of the 1:2 complex (59) of the same charge-transfer system; LB films of this material exhibited a high conductivity (ca.lop2Scm-') even without doping or any further treatment.93 It should be noted that two distinct means of evaluating the lateral conductivity of multilayer films from the measured resistance values exist in the literature. In the first method, used by Barraud and co-worker~,~~ the calculation neglects the 89 T. Nakamura and Y. Kawabata, Techno Japan., 1989,22,8. 90 (a)A.Ruaudel-Teixier, M. Vandevyver, and A. Barraud, Mol. Cryst. Liq. Cryst., 1985, 120, 319; (b)A. Barraud, A. Ruaudel-Teixier, M. Vandevyver, and P. Leisieur, Nouu. J. Chim., 1985,2,365. "J. Richard, M. Vandevyver, P. Lesieur, A. Ruaudel-Teixier, A. Barraud, R. Bozio, and C. Pecile, J. Chem. Phys., 1987,86,2428. 92 T. Nakamura, M. Tanaka, T. Sekiguchi, and Y. Kawabata,J. Am. Chem. Sor., 1986, 108, 1302. 93 M. Matsumoto, T. Nakamura, F. Takei, M. Tanaka, T. Sekiguchi, M. Mizuno, E. Manda, and Kawabata, Synth. Met., 1987, 19, 675. Recent Progress on Conducting Organic Charge- Transfer Salts I 7-6-I 5-iiic 4-.-s d-b 3->r c.-m E0) 2-28 (dQg.1 -Figure 14 Low angle X-ray diffraction data for 42 LB layers of salt (60).[The arrows identify maxima due to a secondphase (<10% of the sample volume) within thefilm] (From reference 94b) insulating (hydrocarbon) portion of the molecules; in contrast, in the work of Nakamura and co-workers 92 the full length of the substituted charge-transfer system is used. The former technique can produce a conductivity figure up to an order of magnitude larger than the latter method. We have studied multilayers of the 1: 1 salt (60), as deposited, and found conductivity values of crrt = loT2Scm-' (calculated using the full molecular length). UV-visible dichroism and polarized IR spectra indicate that in these films of salt (60) the TCNQ molecules and the alkyl chains have their long axes inclined at an angle of ca.30" to the substrate normal. From low angle X-ray diffraction data (Figure 14) the Bragg spacing was found to be 3.2nm; taken together these data are consistent with the interdigitated multilayer structure represented schematically in Figure 15. This structure is in contrast to the orientation of salt (58) discussed ab~ve.~' The nature of the conduction process in films (60) is uncertain (a 1:1 salt should be a Mott insulator I"). It seems likely that the films are unintentionally doped, with an anion such as OH-, during depo~ition.~~' 380 Bryce 3.2nm Figure 15 Schematic diagram showing the proposed average orientation of salt (60) in LB jilmform (From reference 94b) LB films of a 1 :1 salt formed by TCNQ and the dimethyloctadecylsulphonium cation are conducting after iodine doping; from spectroscopic analysis of the films it was concluded that the TCNQ molecules exist as dimers with their long axes parallel to the substrate,95 in a similar fashion to salt (5S).9' C.TTF-TCNQ CT Complexes.-LB films that incorporate both TTF and TCNQ derivatives have been fabricated: a long hydrophobic chain is attached 94 (a) A. S. Dhindsa, M. R. Bryce, J. P. Lloyd, and M. C. Petty, Sjnth. Met., 1987, 22, 185; (b) A. S. Dhindsa, G. H. Davies, M. R. Bryce, J. Yarwood, J. P. Lloyd, M. C. Petty, and Y. M. Lvov, J. Mol. Elect., 1989, 5, 135; (c) R. J. Ward, A. S. Dhindsa, M. R. Bryce, M. C. Petty, and H. S. Munro, Thin Solid Films, 1991, 198, 363. 95 A. Barraud, M. Lequan, R. M.Lequan, P. Lesieur, J. Richard, A. Ruaudel-Teixier, and M. Vandevyver, J. Chem. Soc., Chem. Commun., 1987,797. 38 1 Recent Progress on Conducting Organic Charge- Transfer Salts n rn (58) 22 1 (59) 22 2 (60) 18 1 either to (i) the TCNQ moiety, or (ii) the TTF moiety, or (iii) both components. The first TTF-TCNQ films were reported by Japanese workers who prepared the 1:1 complex (61).9" The conductivity of the films is orl= 1W2Scm-', without doping, which is an order of magnitude higher than a compressed pellet of the complex. No anistropy of the conductivity was observed in the film plane, and the temperature dependence of the conductivity showed typical semiconductivity behaviour with E, = 0.08eV.97 We have studied LB films of a TTF-TCNQ complex in which a long, hydrophobic chain is attached to the TTF moiety, uzz.complex (62) which has a relatively low conductivity value (orl= 10-3Scm-') that is further reduced to ort= lo-" Scm-' upon iodine doping.98 French workers have studied in detail the complex (63) formed by a double- chain tetrathio-TTF donor and TCNQF4. In the deposited form, LB films are insulating, due to complete charge-transfer, but the films become conducting (or, = 5 x 1(V2Scm-') after iodine doping, i.e. on formation of a mixed-valance, ternary complex.99 Spectroscopic data suggest that the TCNQF4 molecules are fully ionized both before and after doping and are, therefore, not involved in the conduction process. Doping converts the donor into a mixed valence species and it is these molecules, not I;, that are directly responsible for the conduc- tivity.By using a specially designed LB trough, we have alternated layers of octadecyl-TCNQ and octadecanoyl-TTF (64) to afford the semiconducting multilayer structure represented schematically in Figure 13e,lo0 for which orI= 5 x lt3Scm-'. This procedure provides a novel way of enforcing segregated stacking on a donor-acceptor complex. Further studies on amphiphilic TTF and TCNQ derivatives that have different structures and redox potentials are needed 96 T Nakamura, F Takei, M Tanaka, M Matsumoto, T Sekiguchi, E Manda, Y Kawabata, and G Saito, Chem Lett, 1986, 323 97 Y Kawabata, T Nakamura, M Matsumoto, M Tanaka, T Sekiguchi, H Komizu, E Manda, and G Saito, Sqnth Met, 1987, 19,663 98 A S Dhmdsa, C Pearson, M R Bryce, and M C Petty, J Phys D Appf Phys, 1989,22, 1586 99 (a) J Richard, M Vandevyver, A Barraud, J P Morand, R Lapouyade, P Delhaes, J F Jacquinot, and M Roulliay, J Chem SOC,Chem Commun, 1989, 754, (b) J P Morand, R Lapouyade, P Delhaes, M Vandevyver, J Richard, and A Barraud, Synth Met, 1988,27, B569 'O0 C Pearson, A S Dhindsa, M R Bryce, and M C Petty, Synth Met, 1989,31,275 Bryce to determine if the conductivity values of alternate layer structures of this type can be increased to the level of single crystals of TTF-TCNQ.D. Cation Radical Salts.-Cation radical salts of amphiphilic x-donors have provided the best quality data for conducting LB films: to date, the temperature dependence of the conductivity is that of a semiconductor not a metal (cmax= 1 Scm-').A multilayer film of the neutral donor is assembled which is then oxidized to form a conducting salt of general formula (Donor+'), (X-),,, where n < 1, which has a partially filled band structure. The presence of only one bulky component, i.e. the donor, favours a more regular film structure than that of the binary or ternary compounds discussed above, where two bulky species must be ordered. Our strategy has been to decorate the TTF ring with one hydrophobic tail, e.g. molecules (64F(7O),l0' for which a chain length of at least sixteen atoms is needed to give a stable monolayer. The acyl, thioester, and ester groups in compounds (64)-(67) appear to play a key role in ordering the structure of the LB films which form by Y-type deposition (Figure 136).These electron- withdrawing groups reduce the donor ability of the TTF system [by ca. 0.18 V for El* relative to TTF], and their presence has the beneficial effect of significantly increasing the polar, hydrophilic nature of the TTF ring. Indeed, the single crystal X-ray structure of acyl-TTF derivative (71) shows that the carbonyl group lies in the same plane as the TTF ring, thereby maximizing its IT-conjugative effect.' O' Hexadecanoyl-TTF (65) was the first derivative to be studied.'02 Upon doping with iodine vapour the LB films initially remain insulating, then after several hours in air, semiconducting behaviour is established, cr1= 10-2Scm-', E, = 0.19eV. This doping process has been monitored by UV-visible, IR, and X-ray photoelectron spectroscopy: immediately upon doping, the TTF system is oxidized to a full charge transfer (insulating) state, then with time a partial release of iodine occurs to afford the stable mixed-valence (conducting) state.In contrast to this, films doped with bromine remain insulating. Absorption spectra lo' M. R. Bryce, G. Cooke, A. S. Dhindsa, D. Lorcy, A. J. Moore, M. C. Petty, M. B. Hursthouse, and A. I. Karaulov, J. Chem. Soc., Chem. Commun., 1990,816. Io2 (a)A. S. Dhindsa, M. R. Bryce, J. P. Lloyd, and M. C. Petty, Thin Solid Films, 1988, 165, L97; (b)A. S. Dhmdsa, R. J. Ward, M. R. Bryce, Y. M. Lvov, H. S. Munro, and M.C. Petty, Synth. Met., 1990. 35, 307; (0A. S. Dhindsa, M. R. Bryce, H. Ancelin, M. C. Petty, and J. Yarwood, Langmuir, 1990, 6, 1680. w P Wavelength (nm) Figure 16 Absorption spectra of LBJilms of compound (65) Curve a 40 layers as deposited Curve b 40 layers 20 h after iodine doping (I e in the conducting state) Curve c 50 layers 48h after bromine doping The inset shoMs the uariation of optical density with thickness for the iodine- dopedfilm (measured 20 h after iodination) (From reference 1026) Bryce I I a= as deposited b= 0.4V C-0*5V d= 0.6V e= 0.7V 0.-c p 0.2 0 v)a a -0 u 0.1 Q0 500 1000 1500 2000 Wavelength (nm) Figure 17 Absorption spectra for a 40 layer LBJilm of octadecanoyl-TTF (64).The different curves correspond to different applied voltages in the electrochemical cell (From reference 103) of LB films of compound (65) are shown in Figure 16. The intermolecular CT band at ca. 2100nm for the iodine doped films is characteristic of an organic conductor. The optical density of the iodine-doped film scales approximately linearly with the number of layers in the LB assembly. Thus the ordered nature of the multilayers is not disrupted upon doping, and we conclude that doping is uniform throughout the forty layers (Figure 16, inset). Low angle X-ray diffraction and polarized IR spectra of films of (65) point to near-vertical alignment of the donor molecules on the substrate surface, with lateral electron mobility via well-ordered, partially oxidized TTF rings.102c More controlled oxidation of films of an analogous acyl-TTF derivative, viz.octadecanoyl-TTF, (64), has been achieved electrochemically in aqueous tetrabutylammonium perchlorate solution. lo3 The doping process was monitored by measuring the optical spectra at different applied voltages (Figure 17). The charge-transfer band at ca. 2000nm starts to develop at 0.4V, is at a maximum at 0.54.6V and then decreases at higher applied voltages, when the full charge-transfer state of the TTF system has been reached. This is supported by the steady increase in '03 M. C. Petty, A. S. Dhindsa, C. Pearson, A. P. Monkman, and M. R. Bryce, Proceedings IEEE/EMBS, ed. P. C. Pederson and B. Onoral, 1990,1693.Recent Progress on Conducting Organic Charge- Transfer Salty *R bde11H~~S(CHd15C02H s-s Me Me (73) R = -CF3 (72) (74) R= -O-CI,Hn intensity of the band at ca 800nm which has been assigned to the full CT state in crystalline TTF salts Similar structural ordering and doping characteristics are observed for LB films of TTF-thioester (66) lo4 with one important difference the lateral (in- plane) room temperature conductivity of films of (66) is two orders of magnitude higher (art = 1 OScm ', E, = 0 09eV) than for acyl- or ester-derivatives (65) lo2 and (67) lo' Iodine doping has been monitored by IR spectroscopy (Figure 18) Before doping, the C=S band is present at 1250cm (Figure 18a), this band shifts immediately upon doping, consistent with oxidation having occurred, I e neutral (66)-radical cation (66) The new sharp band which appears at 1350cm is a vibronic coupling band and a broad CT band is present at CCI 5000cm (Figure 18b) Two hours after doping, when the conductivity value has reached its maximum, the CT band has shifted to lower wavenumber and the vibronic coupling band is much weaker (Figure 18c) A schematic representation of structural ordering within the conducting films of (66), based on low angle X-ray diffraction and polarized spectroscopic data, is shown in Figure 19 The role ot the C=S group of compound (66) in improving the conduction properties of the LB films is not yet understood, it is tempting to speculate that intermolecular S...S interactions may be involved, but we have no evidence for this The conductivity of (66) is presently the highest for d TTF LB film, it is, however, still considerably lower than that of single crystal TTF halide salts,lo6 eg TTF Bro 77, ort= ca 800Scm ', and a current goal is to achieve comparable values for LB films It is clear that LB film quality within the series of monochain-TTFs (65)-(70) is reduced when the C=O or C=S group is absent Compounds (68))(70) form less-ordered films of lower conductivity, ort= 10 after doping lo7 The related compound (72), which carries a terminal carboxylic acid group, prepared by Bechgaard and Lerstrup, is also a relatively poor LB film material lo' LB films of tetraselenafulvalene derivatives have not yet been reported, but they are clearly prime targets Workers at Ciba-Geigy, Basel, have studied tetrathiotetracene (TTT) deriva- tives (73) lo* and (74) lo9 These donors are not classical amphiphiles and LB film Io4 A S Dhindsa J P Badyal M R Bryce M C Petty A J Moore and Y M Lvov J Clirm Soc Ckem Commun 1990 970 A S Dhindsa Y Song M R Bryce M C Petty and J Yarwood to be published lo6 B A Scott S J Placa J B Torrance B D Silverman and B Welber J Am Chem Soc 1977 99 663 1 lo' A S Dhindsa G Cooke K Lerstrup K Bechgaard M R Bryce and M C Petty to be published Io8 A Wegmann B Tieke C W Mayer and B Hilt1 J Clzem Suc Cliem Commun 1989 716 loY(u)B Tieke and A Wegmann Thin Solid Film, 1989 179 109 (h) B Tieke A Wegmann W Fischer B Hilti C W Mayer and J Pfeiffer Thin SolidFilm~1989 179 233 0 0 9 0 0 0 N h e I5 0-8ki mo5C Q, x> 0 0 0U Recent Progress on Conducting Organic Charge- Transfer Salts 1 h=’l As deposited I hs3 hn4 246 20 “I -2 hr after I2 dopng Figure 19 Low angle X-ray dffraction data for LB layers of compound (66) and schematic diagrams showing the proposed average orientation of the molecules (a) Data for the film as deposited (1 e Y-type deposition) (b) data for the conducting film 2h after iodine doping [The arrow in Figure 19a ident@es maxima due to a second phase within the film which disappear upon doping] formation, as Y-type structures, required mixing the donor with methylarachidate (ca 3 1 ratio) The oxidized films of (74) were prepared in a variety of ways (1) by the usual iodine vapour method, (11) by treatment with aqueous KIj solution Bryce (75) Ni(drnit), before or after film formation; (iii) by anodic oxidation with aqueous LiC104 solution either during, or after, film formation.The conductivity and ordering within the LB film varied with the doping method applied."' X-Ray studies revealed that doping resulted in considerable disordering of the layer structure. The deposited films are insulators (or1 = ca. lC7Scm-') and maximum conductivi- ties obtained after doping are or1= 1.3 x 1t4 Scm-' and Scm-' for (73) and (74), respectively. In contrast to the doped, mixed-valence TTF films (65)-(67), which retain high conductivity values on storage in air for several weeks, the doped TTT films (73) lo' and (74),'09" are quite unstable to loss of iodine, which is accompanied by a marked decrease in conductivity.E. Organometallic Systems.-Conducting LB films of metal(dmit)z complexes with long-chain ammonium or pyridinium cations (75)-(77) have also been studied recently, with the aim of increasing dimensionality within the film structure (cf. Section 5A). The first reports came from Japanese workers 'lo who found conductivity, orl= 10-2Scm-', in bromine-doped films of complex (75). However, the complex had to be mixed with as much as 50% icosonoic acid to prevent collapse of the monolayer on the water surface and transfer to the substrate required a horizontal lifting technique.Subsequently, the same group studied Au(dmit)2 complex (76)-fatty acid mixtures which had significantly higher room temperature conductivity (orl= 25Scm-') and this value was retained on cooling down to ca. 200K.'" Both complexes (75) and (76) are reported to have 1 :1 stoicheiometries. Work in our laboratory has shown that complex (77) of 2:l stoicheiometry produces stable films at the air-water interface without the need of added fatty acid, and efficient Y-type LB deposition of (77) proceeds in the normal way.'12 After iodine doping the films are conducting, orl= 8 x 10-2Scm-1, and variable temperature data (3G100K) provide evidence of space charge injection in both the undoped and doped films. The doping process has been probed by a 'lo (a) T.Nakamura, H. Tanaka, M. Matsumoto, H. Tachibana, E. Manda, and Y. Kawabata, Chem. Lelr., 1988, 1667; (b)idem, Synth. Met., 1988,27, B601. 'I1 T. Nakamura, K. Kojima, M. Matsumoto, H. Tachibana, M. Tanaka, E. Manda, and Y. Kawabata, Chem. Lett., 1989, 367. 112 A. S. Dhindsa, J. P. Badyal, C. Pearson, M. R. Bryce, and M. C. Petty, J. Chem. Soc., Chem. Commun., 1991,322. Recent Progress on Conducting Organic Charge- Transfer Salts combination of IR and XPS data it is clear that oxidation occurs in the vicinity of the nickel centre to yield a mixed-valence CT system 7 Future Directions There are now six families of organic superconductor, ze those based on molecules (4),(8), (9), (12), (13), and (16) The study of these compounds has been most fruitful in recent years and is expected to remain so in the immediate future Exciting physics IS provided by the close proximity of superconducting and antiferromagnetic (insulating) ground states, and the control of dimensional- ity in the solid state is a fascinating challenge for synthetic chemists Only where two-dimensional interstack interactions become predominant is superconductiv- ity observed, it is essential that new donor and acceptor systems are synthesized so that the subtle factors which govern the structural, electrical, and magnetic properties of CT salts can be better understood A'-Ray crystallographic studies at liquid helium temperatures, as well as at high pressures, are needed to shed light on the low temperature transitions that are characteristic of many charge- transfer organic conductors The preparation and study of conducting thin films of CT salts is a growing topic where major developments can be expected Good quality Langmuir- Blodgett films are now available, but their conductivity values are still relatively low Nevertheless, novel properties and technological applications may result, especially from the controlled fabrication of superlattice structures The recent reports of superconductivity in polycrystalline compressed pellets of a BEDT- TTF salt '13 and a thin film of BEDT-TTF iodide, formed by vapour deposition,'14 may prove to be important milestones on the long road to device fabrication D Schweitzer E Gogu H Grimm S Kahlich and H J Keller Angen c'hem Int Ed Engl Adi Mnter 1989 28 953 K Kdwabata K Tanaka and M Mizutani Adt .&later 1991 3 157

ISSN:0306-0012    

DOI:10.1039/CS9912000355

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

Chenz. SOC.Rev., 1991,20,00-00 Poly(pyrro1e) as a Support for Electrocatalytic Materials By Dominic Curran, James Grimshaw,* and Sarath D. Perera SCHOOL OF CHEMISTRY, QUEEN’S UNIVERSITY, BELFAST BT9 SAG, NORTHERN IRELAND 1 Introduction Pyrrole undergoes oxidative polymerization from acetonitrile solution at a metallic anode to form a dark coloured skin that adheres firmly to the metal surface. This simple electropolymerization step, when applied to pyrrole monomers with covalently attached redox centres, offers a method for fixing these redox centres to an electrode. The very large range of modified electrodes prepared by this method is discussed here. A main aim of this area of research is the construction of electrodes on which a catalytic centre is present, and which can be activated by placing the electrode at the required redox potential.A number of important oxidation and reduction processes involve the transfer of atoms or groups of atoms and the making or breaking of bonds which involve a pair of electrons. These reactions do not occur readily at a normal electrode surface which promotes the transfer of single electrons only. Hence, there is a need to develop an electrode surface coating containing an electrocatalyst. If we consider an oxidation process, the catalyst is so designed that it reacts with the substrate and is itself reduced to an inactive form. This inactive form is in turn oxidized at the electrode surface by a sequence of single electron transfer steps. The advantage of such a system lies in constraining the catalyst to the electrode surface where it can be regenerated.The rate of the catalysed oxidation or reduction is limited by the diffusion of substrate to the coated electrode surface. Examples will be given where electrocatalytic processes have been achieved on a small scale. 2 Poly(pyrro1e) The first electrochemical polymerization of pyrrole was reported in 1968 and involved the anodic polymerization of pyrrole in aqueous sulphuric acid to give a laminar electrically conducting polymer containing oxygen (oxypyrrole black). Diaz2q3 in 1979 reported the formation of highly stable and continuous films of poly(pyrro1e) by the oxidation of the monomer at a platinum anode in acetonitrile or, better, acetonitrile-water (99: 1) using tetraethylammonium tetrafluoroborate as background electrolyte. Oxidation occurs at about 1.2 V vs.saturated calomel electrode. Monomer concentration is in the range 2-20 mM. ’ A. Dall’Olio, G. Dascola, V. Varacca, and V. Bocchi, Compt. Rend. C,1968,267,433. A. F. Diaz, K. Kanazawa, and G. P. Gardini, J. Chem. Soc., Chem. Commun., 1979, 635. K. K. Kanazawa, A. F. Diaz, R. H. Geiss, W. D. Gill, J. F. Kwak, J. A. Logan, J. F. Robolt, and G. 9. Street, J. Cliem. SOL..,Cliem. Commun., 1979, 854. 39 1 Poly(pyrro1e) as a Support for Electrocatalytic Materials Films have been grown either by maintaining the anode at a suitable potential or by cycling the anode potential over a range. Poly(pyrro1e) is insoluble in all solvents. Films of the polymer have a high electrical conductivity around 10-100 R-' cm-I and are black in colour.The polymer consists of chains of pyrrole units where one in four units has been oxidized to a radical-cation. Positive charges are balanced by anions from the solution (e.g.BF4-).2 Raman and IR reflectance spectroscopy have confirmed the presence of pyrrole rings in the ~tructure.~ Solid state 13C-NMR studies indicate a,al-coupling.4 Evidence for the a,al-coupling was also provided by Diaz who examined the polymerization of substituted pyrrole~.~ 2,5-Disubstituted pyrroles did not polymerize and 2-substituted pyrroles formed only soluble dimers. Further, the destructive oxidation of pyrrole blacks yields pyrrole-2,5-dicarboxylic acid.The initial step in this anodic polymerization of pyrrole involves oxidation to pyrrole radical-cation which then reacts with a second molecule of pyrrole leading to dipyrrole. The oxidation potentials for pyrrole oligomers become less positive with increasing chain length. Thus, this dimer is converted preferentially into the radical-cation which reacts with monomer leading to pyrrole trimer. The continuing preferential reaction is a lengthening of polymer chains already started, and this leads ultimately to the precipitation of polypyrrole with a negligible loss of material as soluble oligomers. Polymer films are formed in this way by a nucleation and growth mechanism.' Scanning electron micrographs reveal that the films are very poorly crystalline and are space filling rather than fibrillar.8 The forming film is itself electrically conducting, so that thickening of the film does not inhibit growth. The apparent stoicheiometry for the reaction is 2.2-2.4 electrons per pyrrole ring.' Two electrons per pyrrole ring are involved in the polymerization process and the remaining charge is used up in partial oxidation of the pyrrole rings.Cyclic voltammetry studies show that the polymer can be cycled between the conducting (oxidized, doped) form which is black in colour and the insulating (reduced, neutral, undoped) form which is pale yellow. Polymerization to form a coherent film on a platinum anode is tolerant of a wide range of substituents attached to the pyrrole nitrogen atom.The reaction fails when the substituent has a basic function attached such as pyridine,'Oq'' G B Street, T C Clarke, M Krounbi, K Kanazawa, V Lee, P Pfluger, J C Scott, and G Weiser, Mol Crjst Lig Cryst, 1982,83, 253 A F Diaz, A Martinez, K K Kanazawa, and M Salmon, J Electroanal Chem , 1981, 130, 181 A F Diaz, J Crowley, J Bargon, G P Gardini, and J B Torrance, J Electroanaf Chem , 1981, 121. 355'A J Downard and D Pletcher, J Electroanal Chem , 1986,206,139 K K Kanazawa, A F Diaz, W D Gill, P M Grant, G B Street, and G P Gardini, S~.ntliMet, 1979/80, 1,329 R J Waltman and J Bargon, Can J Chem, 1986,64,76, A F Diaz, Chem Scr , 1981.17, 145 lo G Bidan, A Deronzier, and J -C Moutet, Nouv J Chim, 1984,8,501 I' S Panaro, P Prosperi, F Bonino, and B Scrosati, Electrochim Acta, 1987, 32, 1007, T Osaka, K Nan, and S Ogano, J Electrochem Soc , 1988, 135, 1071, T Sata and K Saeki, J Chem Sol, Chem Commun , 1989,230 392 Curran, Grimshaw, and Perera bipyridyl,' 2, or dialkylamino.l4 These groups act as nucleophiles towards the pyrrole radical-cation centres and stop interaction of these radical-cations with a pyrrole monomer. Addition of a trace of acid to protonate the basic centres usually allows polymerization to proceed. 3 Electrochemical Characterization of Substituted Poly(pyrro1es) Substituted poly(pyrro1es) considered here show electrochemical behaviour due to two functions. First, the polypyrrole chain itself can be oxidized and reduced: it exists in the oxidized form at potentials greater than 0.6V us.SCE. Secondly, the electroactive group in the side chain shows its own characteristic electrochemistry which may be in the same potential range as that shown by related monomeric compounds in solution. Occasionally groups show unusual electrochemical behaviour when immobilized within a polypyrrole. The most widely used electrochemical techniques for characterizing these polymers are cyclic voltammetry (CV) and chronoamperometry (CA).' Spectroscopic techniques such as multiple reflection infra-red spectroscopy, Raman scattering, and transmission spectroscopy at optically transparent electrodes have also been used to characterize the species involved in the electrochemistry. CV is a technique in which the potential of the working electrode can be scanned in one direction, anodic or cathodic, at a predetermined rate, whilst the current due to oxidation or reduction of the electroactive group is observed.The electrode is then immediately scanned in the opposite direction to observe the peaks due to reduction or oxidation of the intermediates formed during the forward scan. During the process, cell current is recorded as a function of the applied potential. Continuous CV is a convenient means of recording the gradual deterioration of a film under electrical stress. In CA the film potential is stepped up to a region where the electroactive group must change its redox level. Potential is maintained constant and the current required to charge the film is recorded.Integration of this current gives the number of redox sites available in the film. During the initial phase of charging, the film current is proportional to t-t and the constant of proportional-ity is a measure of the diffusion coefficient for charge through the film.16 After this initial phase and when charge has migrated to the outer surface of the film, the current becomes less than that expected from extrapolating the region where i cc t-+. 4 1-and 3-Substituted Poly(pyrro1es) 3-Substituted pyrroles are relatively difficult to prepare, with the consequence l2 M. Salmon, M. E. Carbajal, M. Aguiler, M. Saloma, and J. C. Jaurez, J. Chem. Soc., Chem. Commun., 1983, 1532. '' F. Daire, F. Bedioni, J. Devynck, and C.Bied-Charreton, J. Electroanal. Chem., 1986, 205, 309. l4 J. Grimshaw and S. D. Perera, J. Electroanal. Chem., 1989,265, 335. l5 For a discussion see A. J. Bard and L. R. Faulkner, Electrochemical Metho&. Funrlamentals and Applications, J. Wiley and Sons, New York, 1980. l6 P. Daum, J. R. Leonard, and R. W. Murray, J. Am. Chem. Soc., 1980,102,4649. Polybyrrole) as a Supportfor Electrocatalytic Materials 0 (3).n -0to 6.11 that most of the work on poly(pyrro1es) has concentrated on using 1-substituted pyrroles. A direct comparison of properties is available for the polymers prepared from isomers (1) l7 and (2).18 Both monomers were polymerized in the usual way in acetonitrile and the electrochemistry of the polymers was examined in dimethyl sulphoxide.The films showed two redox processes on CV for the anthraquinone group corresponding to the processes AQ/AQ'-and AQ'-/AQ* -at almost the same potentials found for the monomer. With thicker films of poly(1) however, this ideality disappeared. Side peaks formed on the cyclic voltammogram and the relative size of the dianion response decreased. The decrease was attributed to slow kinetics of dianion formation. Films of poly(2) showed much more nearly ideal behaviour with little evidence for the slow kinetics of dianion formation." The dianion formed in poly(2) was protonated rapidly, presumably by the free pyrrole NH groups. Charge transfer through films of poly(2) is faster than through films of poly( 1). Polymers containing the anthraquinone group are attractive because their electrochemistry in water can be expected to show changes with pH.Unfortunately poly(1) is totally inactive in water, probably because of the hydrophobic nature of the pyrrole backbone. This polymer was electroactive in 70% aqueous acetonitrile, and in this medium its electrochemistry showed changes with PH.~' Poly( 1) catalysed the reduction of dioxygen at an electrode.20 This catalysis is brought about by a rapid redox reaction between the reduced anthraquinone and dioxygen. The anthraquinone is regenerated in the process and then re-reduced at the metal electrode. Electron transfer between a bare platinum electrode and dioxygen is much slower at the same electrode potential.The group of pyrroles (3)21 and the compound (4)22 offer a comparison of "P Audebert, G Bidan, and M Lapkowski, J Elec rroannl Cliem , 1987,219, 165 l8 C P Andrieux and P Audebert, J Electroanal Chem, 1989,261,443 l9 C P Andrieux, P Audebert, and J -M Saveant, Sjnth Met, 1990,359, 155 2o P Audebert and G Bidan, J Electroanal Chem ,1987,238, 183 21 A Hairnerl and A Merz, Angeii Cliem ,Int Ed Engl, 1986,259. 180 22 T Inagaki, M Hunter, X Q Yang, T A Skotheirn, and Y Okaoto, J Cliem Soc , Chem Coniniun , 1988. 126 Curran, Grimshaw, and Perera m properties for closely similar 1-and 3-substituted pyrroles. Under the conditions necessary for oxidative polymerization (E ca. 1.OV us. SCE) the ferrocene residues will also be oxidized to ferricinium.For all the examples, homopolymerization resulted in uneven deposits. Satisfactory films could only be obtained by co-polymerization in the presence of pyrrole or N-methylpyrrole. 5 Anion Exchange in Oxidized Poly(pyrro1e) The ion exchange properties of the oxidized form of poly(pyrro1e) have been used to immobilize anionic species on coated electrode surfaces. Fixation of [Fe(CN)6J3- in this way can be demonstrated by monitoring the induced redox behaviour. The counterion is released into the solution by pulsing the film at -0.4V us. SCE, at which potential the pyrrole units are in the neutral The much larger anionic species, meso-tetrakis(4- sulphonatopheny1)porphyrin-cobalt, is also incorporated into poly(pyrro1e). It behaves as a catalyst for oxygen reduction. Thus, the potential for oxygen reduction at a bare gold electrode is -0.2V us.SCE and this potential is shifted to +0.2V when the gold is coated using poly(pyrro1e) doped with the anion. Complexes of tetrasulphonatophthalo- cyanin with manganese,24 iron,25 and cobalt 26 can be incorporated into poly- (pyrrole) and these materials also catalyse the reduction of oxygen. The binding capacity of poly(pyrro1e) varies with potential and disappears when the film is reduced. Also, bulky ions cannot be incorporated into a preformed poly(pyrro1e) by ion exchange. Rather, they must be the anion present in solution during the polymerization stage. The capability of polypyrroles to bind by an ion exchange mechanism is much improved by linking ionic substituents to the pyrrole monomer.In one approach to the binding of cations, the monomer (5) was polymerized and the resulting polymer used to bind cobalt ions.27 In another approach to the binding of anions, the monomers (6)-(8) have been polymerized and the resulting films used to bind ferricyanide or ruthenate ions.28 Poly(6) incorporates ferricyanide ions when soaked in an acetonitrile solution and these ions are 23 L. L. Miller, B. Zinger and Q.-N. Zhou, J. Am. Chem. Sor., 1987, 109. 2267; J. Tietje-Girault. J. M Anderson, I. MacInnes, M. Schroder, G. Tennant, and H. H. Girault, J. Chem. Sor., Cliern. Commun.. 1987, 1095; G. Lian and S. Dong, J. Eiecrroanal. Chem., 1989, 260, 127. 24 F. Bedioui, C. Bongas, J.Devynk, C. Bied-Charreton, and C. Hinnen, J. Electroonol. CIieni., 1986. 207. 87 25 A. Elzing, A. van der Putten, W. Visscher, and E. Barendrecht, J. Eiec rroannl. Cliem., 1987,233. 113. 26 T. Skotheim, M. Velazquez-Rosenthal, and C. A. Linkous, J. Cliem. SOC..Chem. Comniun., 1985, 612. T. Osaka, K. Naoi, T. Hirabayashi, and S. Nakamura, Bull. Cliem. So(. Jpn., 1986,59,2717.’’P. G. Pickup, J. Electroanol. Cliem., 1987,225,273. 28 S. Cosnier, A. Deronzier, J.-C. Moutet, and J. F. Roland, J. Electround. Cliem.. 1989,271.69. Poly(pyrro1e) as a Support for Electrocatalytic Materials retained on rinsing in fresh solution or on cyclic voltammetry. Poly(7) also incorporates ferricyanide ions but these are slowly rinsed out. Incorporation of ferricyanide ions into poly(8) is very small.In general, poly(6) offers the best possibilities of the three materials for the incorporation of electroactive anions. The anions are retained even when the polypyrrole backbone has been destroyed by oxidation at + 1.2V us. SCE. Poly(6) and poly(7) have been used to adsorb anions from potassium ruthenate. The adsorbed anion was reduced electrochemically to ruthenium dioxide. Such an electrode based on poly(6) prepared on carbon felt has been used to catalyse the oxidation of benzyl alcohol to benzaldehyde at 1.OV us. SCE on a 10 millimolar scale. The ruthenium is oxidized to the R~04~-ion which then oxidizes the alcohol with formation of ruthenium dioxide. The catalyst -is subsequently regenerated electrochemically.28 Platinum microparticles have been incorporated into a polypyrrole film by first carrying out an ion-exchange step with chloroplatinate ion followed by the electroprecipitation of platinum. Poly(pyrro1e) 29 has been loaded in this way as also has p01y(5).~' The microparticulate platinum electrode incorporated into poly(6) grown on carbon felt was very effective in aqueous ethanol for the electrochemical hydrogenation of the alkene bond in enenones, of benzaldehyde to benzyl alcohol, and of nitrobenzene to aniline.30 The bulky cluster [Fe4S4(SPh)4I2-can be bound from acetonitrile onto preformed poly(6).Transferring the electrode to an aqueous solution achieves the interfacing of this hydrophobic cluster with an aqueous ele~trolyte.~' 6 Poly(pyrro1e)with Reducible Substituents An initial objective for research has been to explore the size and type of substituent that can be attached to pyrrole while still retaining good mechanical and adhesive properties for the corresponding polymer.It is also interesting to see if the crosslinking introduced by polymerization of monomers having two pyrrole units per reducible group will influence the electrochemistry of this group within the polymer matrix. All the pyrroles discussed in this section carry N-substit uen ts. In the case of polymers derived from monomers (9) and (10) the only indication of an effect of polymer structure was found with poly(9) where the response of the anthraquinone in dimethyl sulphoxide led to slightly broader peaks than usual on cyclic voltammetry.Peak maxima were at the same potential for the two polymers. These polymers were very stable to cycling through the AQ/AQ' -region. However, when cycles included the AQ'-/AQ2 -couple, then a decrease in peak height with time was observed on continuous cyclic voltammetry. This decrease is presumably due to protonation of the negative ions. These anthraquinone-containingpolymers, coated on platinum, catalysed the reduction of oxygen.' 29 S. Holdcroft and B. C. Funt, J.Electroanat. Chem., 1988,240,89. 'O L. Coche and J.-C. Moutet, J.Am. Chem. Sor., 1987,109,6887.'' C. J. Pickett and J.-C. Moutet, J. Chem. SOC., Chem. Commun., 1989, 188. Curran, Grimshaw,and Perera II (10) R= -CH0 \CH2-ND / o-CH2CH#H@\ I CI NHCH2CH2-ND 0 0 By contrast, the pyrromelitimide group in poly( 11) showed a considerably different redox behaviour compared to that for the monomer.In acetonitrile the monomer shows two reversible one-electron transfer steps centred at -0.8 and -1.5V us. SCE, respectively. The more negative step is not seen in the redox behaviour of the polymer and the first step shows a much broader wave in cyclic voltammetry. A colour change from green to red accompanies the second electron transfer step of the monomer. Since the polymer film becomes red at the negative side of the observed redox peak, here both one-electron transfer steps must be concealed under the one peak. Thus, the second electron transfer step of the pyrromelitimide group in the polymer is moved to substantially less negative potentials compared to its position for the monomer.32 Not all attempts to polymerize a substituted pyrrole are successful.Thus, attempts to polymerize (12) oxidatively did not lead to a clean reaction. Substantial amounts of dark coloured soluble products were formed. The polymer film which did form under these conditions showed a redox response due to the NQ/NQ’-couple. This response rapidly decreased on cyclic voltammetry and had disappeared after three cycles.33 The benzoquinone (13), was successively polymerized by oxidation of the pyrrole rings. The electrochemistry of the quinone unit in this polymer is modified compared to that shown by the monomer.The anodic and cathodic waves on cyclic voltammetry due to the BQ/BQ’- couple are broader in the polymer and the cathodic wave was moved 0.22V more negative. Not all the BQ’-units were discharged during the anodic phase of cyclic voltammetry so that, after repeated scanning, the BQ/BQ*- couple response decreased in height. ’’J. Grimshaw and S. D. Perera, J. Electroanal. Chem., 1990,278,279. 33 J. Grimshaw and S. D. Perera,J. Electroanal. Chem., 1990,281, 125. Poly(pyrro1e) as a Support for Electrocatalytic Materials I w After resting the film to allow complete discharge, the original electrochemical behaviour is restored. It is suggested that the cross-linking within this film limits ion motion and hence the transport of charge through the film.33 Films of poly( 14) show unusual electrochemical behaviour.When the polymer is placed at a potential where the cyanoanthracene group accepts an electron to form the radical-anion, charging of the film occurs rapidly at approximately constant current density, then in a few milliseconds the current fall to This behaviour strongly suggests that in the film the cyanoanthracene units are stacked in zones which allow rapid transport of charge, and also counter-ions, through the structure. Charging ceases either when the electron acceptor units are all engaged, or when no more counter-ions can be accommodated in the cross-linked structure. Pyrroles (15) and (16), with attached pyridinium groups, have been successfully polymerized by controlled potential oxidation to give films which show the reversible redox chemistry of the bipyridinium gro~p.~~,~~This group was reduced at the expected potential in two one-electron steps.A film of poly( 16) on platinum or on glassy carbon has been used to promote the reduction of aliphatic halogen compounds, dibromostilbene to stilbene 36 and hexachloro- acetone to pentachlor~acetone.~~ In these reactions the reduced form of the bipyridinium acts as an electron shuttle, transferring an electron to the organic substrate, then being itself re-reduced by reaction with the electrode and adjacent redox centres. When conditions are such that several layers of electrode coating are involved in this way with electron transfer to the substrate, the reaction overall is faster than at a bare metal where only the surface can be involved in electron transfer.There has been a theoretical discussion of the advantages of this mediated, outer-sphere type of electron tran~fer.~' The neutral species from addition of two electrons per bipyridinium group mediate the reduction of dibromostilbene, whereas both the radical-cation and the neutral species mediate the reduction of hexachloroacetone. 34 J Grimshaw and S D Perera, J Elec troanal Chew . 1989,265,335 35 G Bidan, A Deronzier, and J -C Moutet, J Chem Soc , Chem Coniniun, 1984, 1185 36 L Coche, A Deronzier, and J -C Moutet, J Electroanal Cliem , 1986, 198, 187 37 L Coche and J -C Moutet, J Elerrrocma1 Climi, 1988, 245, 313, L Coche and J -C Moutet, J Elec troanal Chem , 1987,224, 1 1 1 38 W J Albery and A R Hillman, Annu Rep Progr C/iem,Sec, C, 1981.78, 414, and references cited therein Curran, Grimshaw, and Perercr 7 Poly(pyrrole) with Oxidizable Substituents Polymerization of monomers in this class has the complication that the substituent can be oxidized at a less positive potential than that necessary to polymerize the pyrrole ring. These monomers are usually polymerized successfully by repeated cyclic voltammetry through the electroactive range and farther out to positive potentials where the pyrrole will be polymerized. Comment was made previously that the ferrocene-containing polypyrroles derived from (3) and (4)do not give rise to satisfactory Similar results have been found with poly(l7) which can utilize the two pyrrole rings to form a cross-linked film.39 Copolymerization of these compounds with pyrrole or N-methylpyrrole results in much more stable films.All the homopolymer films from (3), (4),and (17) deteriorated rapidly on cyclic voltammetry about the ferrocene response. After about 10 scans the redox behaviour of ferrocene/ferricinium had almost disappeared and could not be restored. In the polymer derived from (18) there are two electroactive groups on the pyrrole side chain. The triphenylamine residue is reversibly oxidized at 0.92V us. SCE and the bipyridinium residue is reversibly reduced at -0.71 V us. SCE. The monomer is polymerized by repeated cyclic voltammetry between -0.2 and 1.05V us.SCE.40 When this film is placed in a clean solution, repeated cyclic voltammetry at the response of the triphenylamine groups results in a gradual decrease in the current response which is replaced by a new redox couple oxidized at 0.78V us. SCE while redox behaviour of the bipyridinium group remains unchanged. This new couple is due to the tetraphenylbenzidine residue (1 9) formed during oxidative coupling of the triphenylamine. Related oxidative coupling reactions have been observed for other triphenylamines in solution. Coupling occurs much more rapidly in the polymer film because of the close proximity of the reacting groups. Dilution of the loading of triphenylamine residues by copolymerization of the monomer with another pyrrole leads to a film that no longer shows the oxidative dimeri~ation.~’ 8 Catalysed Oxidation of Alcohols The oxidation of alcohols to ketones does not proceed in a satisfactory manner ’’J.G. Eaves, R. Mirrazaei, D. Parker, and H. S. Munro, J. Ciiem. Soc., Perkin Trans. 2, 1989,373. 40 A. Deronzier, M. Essakalli, and J.-C. Moutet, J. Electroanal. Chem., 1988. 244, 163. Poly(pyrro1e) as a Support for Electrocatalytic Materials (22)L= N3(CH2)2-N3 at a bare anode. A number of attempts have been made to devise a polypyrrole based electrocatalyst for this process. Mention has already been made of ruthenate based systems28 where the ruthenate ion is held by ion-pairing with positively charged groups on the polymer side chain.A more elaborate positively charged ion, poly[R~~~(20)3]~+, has been used in the same way to bind ruthenate ions.41 In another approach to this problem an electrode has been coated with poly(2 l).42 This material shows a reversible redox couple due to the oxidation shown below. ;N-0 e>&=O +e Addition of an alcohol to the electrolyte solution enhanced the peak in cyclic voltammetry caused by the oxidation N6 -N6 and partially suppressed the corresponding cathodic peak. This establishes that the nitroxonium ions oxidize the alcohol to the ketone and are reduced to NOH. Reoxidation of the hydroxylamine so formed to the nitroxonium ion is responsible for the enhanced peak in cyclic voltammetry. Although oxidation of alcohol to ketone could be demonstrated on an analytical scale, on a preparative scale these films were largely destroyed during the process.More success has been obtained with catalysts derived from ruthenium clusters where the films were obtained by polymerization of monomers (22) and (23).43 Best results were obtained using a platinum cylinder coated with poly(22). This catalysed the oxidation of benLyl alcohol to benzaldehyde, at 1.62V us. Ag+, from dilute solution in acetonitrile. Reduced Ru species generated during this oxidation are re-oxidized electrochemically. This electrode system was stable for several hours. 9 Photoassisted Electron Transfer The demonstration of photoassisted electron transfer between a group attached to the polymer and a species in solution has been the object of research.To this end monomers (24) and (25) were polymerized by repeated cycling at positive potentials. Poly(24) showed reversible redox behaviour due to the phenothiazine 41 L Coche and J -C Moutet, J Electroanal Chem ,1988, 245, 313, S Cosnier, A Deronzier, and J -C Moutet, Inorg Chem, 1988,21,2390 42 A Deronzier, D Limosin, and J -C Moutet, Electrochim Acta, 1987,32, 1643 43 S Cosnier, A Deronzier, and A Llobet, J Electroanal Cliem, 1990,280,213 Curran, Grimshaw,and Perera (24)R= -NZ The film was kept at a potential negative of this redox couple and irradiated with visible light while immersed in a solution of tropylium cation. This resulted in a photocurrent. Photoassisted electron transfer between the polymer-based phenothiazine and tropylium cation leads to the phenothiazine radical-cation.Reduction of this radical-cation back to the level of phenothiazine is responsible for the current. The tropylium cation is converted into a radical which undergoes irreversible dimerization. The polymer prepared from (25) can be expected to undergo photoassisted electron transfer within the polymer because of the two redox units that are present in the side chain.45 Accounts of experiments in this direction are expected. An electrode coated with p0ly[Ru"(28)3]~+ when kept at a potential where the stable species is Ru" generated a photocurrent in visible light when immersed in a solution containing 4-methylbenzenediazonium fluoroborate as the electron acceptor in a~etonitrile.~~ The current results from the reduction of the so formed Ru"' complex back to the Ru" state.10 Polypyrroles with Attached Metal Complexes A. Pyridine Complexes-The first reports of the polymerization of pyridine- metal complexes with pyrrole groups attached to the pyridine moiety concluded that [Ru(byp)2(26)2I2+ and [Ru(byp)2(27)2I2 will form films after repeated + cyclic voltammetry at a platinum ele~trode.'~.' Later, two groups investigated the number of pyrrole units per metal centre that are necessary to allow film formation. They worked with the related ligands (28) 47 and (29),' respectively, and reached the same conclusions on the polymer forming ability of this type of ligand-metal complex.The tris complexes of (28) and (29) with Ru" and the mixed complexes [Ru(byp)(28)2I2+ and [Ru(byp)(29)2I2 + gave stable polymeric films by oxidation on platinum, glassy carbon, or tin oxide. The mixed complexes with only one pyrrole-containing bipyridyl ligand would not form a polymeric film. Thus the film forming ability and the stability of the films obtained from these ruthenium 44 A. Deronzier, M. Essakalli, and J.-C. Moutet, J. Chem. Soc., Chem. Commun., 1987, 773. "A. Deronzier, M. Essakalli, and J.-C. Moutet, J. Elertroanal. Chem., 1988, 244, 163 46 S. Cosnier, A. Deronzier, and J.-C. Moutet, J. Phys. Chem., 1985,89,4895. 4' J. G. Eaves, H. S. Munro, and D. Parker, Inorg. Chem., 1987,26,644. 401 Poll (pirrole) as a Support for Electrocatalytic Materials (29)n = 2 (30) n =3 complexes increases substantially as the number of attached pyrrole groups increases This may be due to cross-linking which makes the resulting polymer less soluble The complexes [Fe(28)3I2+ and [Cu(28)2I2 + have also been polymerized to give stable films on platin~m,~' Filmsas has a related bis complex of CU"~~ containing Ru" and Fe" were unstable towards reductive cycling but stable on oxidative cycling between the metal centred redox reactions Tris (28) complexes of CO"~~ and Ni115' could be oxidatively polymerized on gold but not on platinum or vitreous carbon The cobalt-containing polymer showed a reversible redox couple at -1 09 V us SCE due to Co"/Co' Reduction of ally1 chloride was catalysed by the Co' species formed at potentials more negative than -1 09V49 In contrast to the previous example, the complex [Re(30)(C0)3Cl), which has only one pyrrole ring per metal atom, could be oxidatively polymerized to a stable coating The resulting polymer has been used to effect the catalytic reduction of carbon dioxide to carbon monoxide with good current efficiency However during the experiment about 80% of the electroactivity of the film was destroyed 51 B.Porphyrin Complexes.-The formation of coherent polypyrrole films contain- ing metalloporphyrin units illustrates the adaptability of this polymer backbone unit Murray 52 first reported the polymerization of tetrakis(p-N-pyr-rolypheny1)porphyrin and its cobalt complex (31) on platinum or glassy carbon electrodes The monomer was dissolved in dichloromethane containing 0 1 M tetrabutylammonium perchlorate and polymerization was effected by repeated scanning between 0 and 1 1V us Ag/AgCl The cobalt complex could also be prepared by metallating the porphyrin substituted polypyrrole Other workers 53 reported polymerization of the Ni" porphyrin (32) on platinum by repeated scanning of a solution In dichloromethane This nickel complex showed two 48 G Bidan B D Blohorn J M Kern and J P Sauvage J Cliem Soc Chem Comniun 1988 723 "F Daire F Bedioui J Dveynck dnd C Bied Chdrreton J Elertroonol Clieni 1987 224 95'"F Daire F Bedioui J Devynck dnd C Bied Chdrreton Elecrroclirni Acin 1988 33 567 51 S Gosnier A Deronzier and J C Moutet J Electronnu1 Cltem 1986 207 315 s2 A Bettelheim B A White S A Raybuck dnd R W Murray Inorg Clieni 1987 26 1009 53 A Deronzier and J M Latour J Electrorinol Chem 1987 224 295 Currun GrrmshaM. and Perera (33) R' = a N 3 R2= p tolyl M = Mn& reversible redox couples at 1 24V and -0 84V us SCE respectively The redox couple at positive potential is due to the metal centre, whereas the couple at negative potential is due to the bipyridinium group The manganese porphyrin (33) has also been successfully polymerized on vitreous carbon 54 This shows a stable and reversible redox couple at -0 3V LS SCE due to Mn"'iMn" The Mn" species formed has been shown to activate molecular oxygen Thus, the system consisting of a polymer coated electrode at -0 5 V LS SCE and oxygen effects the oxidation of 2,6-di-t-butylphenol to 2,6-di- t-butylbenzoquinone This same system converts cis-cyclooctene into the epoxide During these reactions the active centre of Mn" is converted into Mn"' and then reactivated by reduction at the electrode 55 C.Cyclam Complexes.-Collin and Sauvage 56 reported the polymerization of 3-[4-(pyrrol-1-yl)-butyll-1,5,8,12-tetraazacyclotetradecane-Ni2 on glassy carbon + electrodes Cyclic voltammetry of the monomeric Ni+yclam complex showed two redox couples due to Ni"'/Ni" and Ni"/Ni' at 0 98 V and -1 41 V u s SCE, respectively Polymer films were grown by repeated cyclic voltammetry at a vitreous carbon electrode They showed the characteristic Ni"'/Ni" redox couple but the Ni"/Ni' couple at negative potentials was not observable 11 Conclusions The oxidative polymerization of pyrroles with a very wide range of N-substituents has been shown to give stable films on electrodes Substituents with redox centres varying in size and character from small organic groups to bulky metal '' F Bedioui A Merino J Devynck C E Mestes and C Bled Charreton J Electront 01 Chem 1988 239 433 55 F Bedioui P Moiav J Devynck L Salmon and C Bled Chdrreton J Mol Cciml 1989 56 267''J C Collin and J P Sauvage J Chtni Soc Chmi Conitnun 1987 1075 Poly(pyrro1e) as a Support for Electrocatalytic Materials complexes are tolerated Redox centres may be active at either positive or negative potentials Attempts to attach catalytic centres to the polypyrrole are continuing, achievements so far have been with catalysts for the oxidation of alcohols and for the activation of molecular oxygen Polypyrrole films with positively charged centres on the N-substituent have been used to immobilize ruthenate ions, which are useful electrocatalysts for the oxidation of alcohols So far only very small scale reactions have been acheived, but the principle of binding catalysts to an electrode in order to achieve an electrocatalytic process has been demonstrated

ISSN:0306-0012    

DOI:10.1039/CS9912000391

出版商:RSC    

年代:1991

数据来源: RSC