|
1. |
Methods for the synthesis of α,β-unsaturated trifluoromethyl ketones and their use in organic synthesis |
|
Russian Chemical Reviews,
Volume 68,
Issue 6,
1999,
Page 437-458
Valentin G. Nenaidenko,
Preview
|
|
摘要:
Russian Chemical Reviews 68 (6) 437 ± 458 (1999) Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis V G Nenaidenko, A V Sanin, E S Balenkova Contents I. Introduction II. Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones III. The use of a,b-unsaturated trifluoromethyl ketones in organic synthesis IV. Conclusion Abstract. Published data on the methods of synthesis and proper- ties of a,b-unsaturated trifluoromethyl-containing ketones are surveyed and described systematically. Primary attention is devoted to the use of these compounds in organic synthesis as useful building blocks for the preparation of various compounds bearing a trifluoromethyl group. The bibliography includes 133 references.I. Introduction Organofluorine chemistry has been vigorously developing during the last two decades. A large number of studies have been devoted to modification of natural products by introduction of a fluori- nated or perfluorinated substituent.1±3 These fluorinated deriva- tives often exhibit biological activities; some heterocyclic compounds containing a trifluoromethyl group have already found wide use as medicines (for example, triftazine, trifluorothy- midine).4 a,b-Unsaturated trifluoromethyl ketones (trifluoro- methyl a,b-enones) are fairly convenient building blocks for pre- paring heterocycles with a trifluoromethyl group. Most of the known approaches to the synthesis of these heterocycles suffer from serious drawbacks, associated either with the fact that the initial compounds are rather difficult to obtain or with the fact that they are fairly toxic and inconvenient to work with.1±3 Thus, development of methods for the synthesis and study of properties of trifluoromethyl a,b-enones present substantial interest and have been in the centre of attention of many researchers in recent years. In this review, we survey methods for the synthesis of a,b- enones with a trifluoromethyl (perfluoroalkyl) substituent at the carbonyl group, because these enones are used in organic synthesis much more often than those containing a CF3 group at other positions.5, 6 II.Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones Although the first a,b-unsaturated trifluoromethyl ketones were synthesised almost 40 years ago, the vigorous development of the methods for their synthesis has started only in the last decade.As a V G Nenaidenko, A V Sanin, E S Balenkova Department of Chemistry, Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 939 31 81. Tel. (7-095) 939 22 76 (Nenaidenko). E-mail: nen@acylium.chem.msu.ru Received 23 June, 1998 Uspekhi Khimii 68 (6) 483 ± 505 (1999); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.384:547.321:547.7 437 437 445 456 rule, they cannot be prepared by the same procedures as non- fluorinated a,b-enones. Five possible approaches to the construction of the carbon skeleton of trifluoromethyl a,b-enones can formally be proposed.Four approaches are based on the formation of C(1) ± C(2), C(2) ± C(3), C(3)=C(4) and C(4) ±R3 bonds and one is based on transformations resulting in the appearance of a carbonyl group. R2 O 3 2 4 R3 F3C1 R1 Analysis of published data shows that the following principal methods for the synthesis of trifluoromethyl a,b-enones exist. �Trifluoroacetylation (for example, of alkenes, their organo- metallic derivatives or compounds generating alkenes in situ) to give a C(2) ± C(3) bond. � Condensation of 1,1,1-trifluoroacetone and its derivatives with aldehydes and the formation of the C(3) ± C(4) double bond in an elimination reaction. � Addition to alkynyl ketones (a,b-ynones) or to enones containing a b-substituent capable of being replaced (most often, an alkoxy group), and replacement of one carbonyl group in b- diketones containing a perfluoroalkyl substituent [formation of the C(4) ±R3 bond].In this case, examples described in the literature include formation of both C±C and C± heteroatom bonds. � Addition of perfluoroalkyl organolithium compounds to esters of acrylic acid to give the C(1) ± C(2) bond. This method is scarcely documented and has never been used for the preparation of trifluoromethyl a,b-enones. Apparently, this is due to the fact that synthesis of perfluorinated organolithium compounds requires the use of perfluoroalkyl iodides, which are toxic and difficult to work with (especially, trifluoromethyl iodide).3 � Oxidation of allyl alcohols, which are synthesised using perfluorinated or vinylic organometallic compounds, yielding a C=O group.This approach also has not found wide use. Below, we consider all the above methods for the synthesis of a,b-unsaturated trifluoromethyl ketones in the same order. 1. Trifluoroacetylation of alkenes a. Trifluoroacetylation of electron-rich alkenes by trifluoroacetic anhydride Trifluoromethyl a,b-enones are formed upon trifluoroacetylation of activated alkenes by trifluoroacetic anhydride (or other derivatives of trifluoroacetic acid) at room temperature or on438 cooling. The reaction is often carried out in the presence of pyridine, which increases the yield. R1X R1X (CF3CO)2O Y Y COCF3 (54% ± 100%) Y=H, Alk, Ar, SAr; X=O, S, Te, NSO2R2, NCOR3; R1, R2, R3=Alk, Ar.Activated alkenes can be represented by vinyl ethers 7 (including cyclic ones such as 1),8, 9 vinyl sulfides,10 ketene dithioacetals,10 vinyl tellurides,11 vinylamides,7 cyclic enam- ines 12, 13 and activated dienes such as 1,1-bis(alkylthio)alka-1,3- dienes 2 or N-acetyl-N-isopropyl-1-aminobuta-1,3-diene 3.14 R R (CF3CO)2O, Py COCF3 O O (CH2)n (77% ± 95%) (CH2)n 1 R=H, Me; n=1, 2. SR1 SR1 (CF3CO)2O, Py COCF3 R1S R1S R2 R2 (43% ± 100%) 2 R1=Alk; R2=H, Et. Pri Pri COCF3 N (CF3CO)2O, Py N COMe (96%) COMe 3 Trifluoroacetylation of vinyl ethers, vinylamides and 1,1- bis(alkylthio)alka-1,3-dienes 2 occurs stereospecifically yielding thermodynamically more favourable E-isomers of trifluoro- methyl a,b-enones.Trifluoroacetylation of vinyl tellurides gives rise to Z-isomers, because they are stabilised by an O± Te coordinate bond (X-ray diffraction data 11). Vinyl sulfides are trifluoroacetylated non-stereoselectively. According to the 1H NMR spectra and data on the reaction kinetics,15, 16 this process follows an addition ± elimination mechanism. Trifluoroacetyla- tion of vinyl ethyl ether gives polymers as side products.16 In the absence of pyridine, the highest yield (59%) was attained when the reaction was carried out in chloroform. Trifluoroacetylation of cyclic enamines, for example, 1-morpholinocyclopent-1-ene (or 1-morpholinocyclohex-1-ene), affords a complex mixture of products.In the case of less reactive 1-morpholinocyclohept-1-ene, doubly trifluoroacetylated prod- uct 4 was isolated.12 CF3CO (CF3CO)2O N O N O 15 ± 25 8C, 1.5 h CF3CO 4 (57%) Enamines 5, derived from indene, dihydronaphthalene and benzocycloheptene, react ambiguously with trifluoroacetic anhy- dride. Depending on the structure of the initial enamine, either trifluoromethyl enamino ketones 6 or products of their cyclisa- tion, oxazine derivatives 7, can be isolated. X X N N (CF3CO)2O COCF3 20 8C (CH2)n (CH2)n 6 (83%) 5 V G Nenaidenko, A V Sanin, E S Balenkova X N O CF3 (CH2)n 7 (30% ± 53%) X=CH2, (CH2)2, OCH2; n=1±3. The reactions of vinyl ethers with a threefold excess of trifluoroacetic anhydride in the presence of pyridine result in the formation of 3-(alkoxymethylidene)-1,1,1,5,5,5-hexafluoropen- tane-1,3-diones in high yields.17RO COCF3 RO (CF3CO)2O, Py COCF3 R=Et, Bui.Orthoacetates, acetals 18 and trithioorthoacetates 19 react with excess trifluoroacetic anhydride with elimination of alkyl or aryl trifluoroacetate (thioacetate) to give trifluoromethyl enones, i.e., they are precursors of the corresponding activated alkenes. It should be noted that only trifluoroacetic and trichloroacetic anhydrides enter into this reaction, whereas acetic anhydride does not. OEt OEt (CF3CO)2O, Py OEt Me 20 8C, 24 h CF3CO OEt OEtR1 (CF3CO)2O, Py OR2 Me 20 8C, 3 ± 48 h CF3CO OR2 OR2 (94% ± 100%) R1=Me, Ph; R2=Me, Et.SAr SAr (CF3CO)2O Me SAr 20 8C, 20 h CF3CO SAr SAr (58% ± 100%) b. Trifluoroacetylation of tertiary amines When trialkylamines are made to react with trifluoroacetic anhydride or trifluoroacetyl halides, b-dialkylamino-substituted trifluoromethyl enones are formed. The reaction of two equiv- alents of triethylamine with trifluoroacetyl chloride at 730 8C gives rise to 4-diethylamino-1,1,1-trifluorobut-3-en-2-one (in 18% yield based on the initial trifluoroacetyl chloride).20 The reaction mechanism proposed in the study cited 20 includes oxidation of triethylamine by one equivalent of trifluoroacetyl chloride to give diethyl(vinyl)amine. Trifluoroacetyl chloride is reduced simultaneously to fluoral. The subsequent interaction of diethyl(vinyl)-amine with a second equivalent of trifluoroacetyl chloride results in the formation of the corresponding enaminone. *H7 CF3COCl+NEt3 [CF3COCl/NEt3] + 7 CHMe] [CF3CHCl O Et2 N CF3CHO+HCl+Et2NCH CH2 NEt3 Et2N CF3COCl+Et2NCH CH2 + COCF3 7HNEt3Cl7 It has been found 21 that the reaction of triethylamine or ethyldiisopropylamine with an equimolar amount of trifluoro- acetic anhydride at 0 8C gives rise to doubly trifluoroacetylatedMethods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis products 8a,b (yield *25%), in conformity with the reaction stoichiometry 4R2NEt+4 (CF3CO)2O COCF3 + + CF3CH(OCOCF3)2+3R2EtNHCF3CO¡ R2N 2 COCF3 8a,b R=Et (a), Pri (b).Cyclic amines react in a similar way. For example, N-methylpiperidine is converted into the corresponding enami- none 9 in 96% yield (with allowance for the degree of conversion of the starting amine). (CF3CO)2O Me N Me N COCF3 9 Some alkaloids containing tertiary nitrogen atoms enter into similar reactions with trifluoroacetic anhydride. They are oxidised and the resulting enamine fragment is then trifluoroacetylated.22 c. Trifluoroacetylation of non-activated alkenes The electrophilicity of trifluoroacetic anhydride and other deriv- atives of trifluoroacetic acid is insufficient for trifluoroacetylation of non-activated alkenes. The attempts to increase the reactivity of these reagents by complex formation with Lewis acids prove successful only for acylation of aromatic compounds.Unsatu- rated substrates undergo cationic polymerisation under these conditions. The use of acylium salts for the acylation of unsaturated compounds would permit conducting the reaction at low temperatures and thus diminish the probability of polymer- isation and other side processes. Unfortunately, acylium salts with a perfluoroalkyl radical are unstable and decompose with decar- bonylation during their attempted synthesis.23, 24 RFCOF+AgSbF6 [RFCO+SbF ¡6] RFF+SbF5+CO A new method for direct electrophilic trifluoroacetylation of non-activated alkenes has been proposed.25 The method is based on the use of trifluoroacetic anhydride in the presence of a complex of dimethyl sulfide with boron trifluoride.This reagent is more electrophilic than trifluoroacetic anhydride, which reacts only with alkenes activated by a heteroatom at the double bond. However, the use of the complex BF3 . SMe2 has a drawback; acylation of an equivalent of an alkene results in the liberation of protonated dimethyl sulfide, which then reacts with at least one more equivalent of the alkene to give sulfonium salt 10. Con- sequently, the yield of vinyl trifluoromethyl ketones does not exceed 50%. This reaction proceeds with alkenes able to form benzylic, allylic or tertiary cations (of the type 11); however, in the last-mentioned case, mixtures of a,b- and b,g-enones are formed.26, 27 R1 Me + R1 + Me2SH SMe2 3 R2 R2 CF3CO2BF¡ 10 a R1 7H+ CHCOCF3 (19% ± 49%) R2 R1 + + 7Me2SH R2 3 R1 CH2COCF3 CH2COCF3 CF3CO2BF¡ 11 SMe2 + R2 SMe2 CF3CO2BF¡3 (a) (CF3CO)2O, BF3 .SMe2,760 to730 8C. 439 Cyclopropyl-substituted alkenes enter into a similar reac- tion.28, 29 The reaction of 2-cyclopropylpropene and 1,1-dicyclo- propylethene with trifluoroacetic anhydride in the presence of BF3 . SMe2 follows an unusual route and results in cyclopropane ring opening to give sulfonium salts 12.29 The salts 12, which are b,g-unsaturated ketones, rearrange spontaneously into a,b- enones 13 (2 months, *20 8C) in a quantitative yield. In the presence of a base (KF, DMF), the sulfonium salts 12 undergo cyclisation in which the dimethylsulfonium group is intramolecu- larly substituted to give a,b-enones with a cyclopropyl substituent 14.CF3COCH2 (CF3CO)2O, BF3 . SMe2 + SMe2 760 to740 8C R R H CF3COCH2 CF3CO2BF¡3 + R (CH2)2SMe2 12 (92% ± 95%) 20 8C, 2 months KF, DMF 70 ± 80 8C, 6 h R CF3CO CF3CO2BF ¡3 + COCF3 (CH2)3SMe2 13 (100%) R14 (31% ± 34%) R=Me, cyclo-C3H5. In the presence of the complex BF3 . SMe2 , trifluoroacetic anhydride trifluoroacylates alkynes with a phenyl substituent at the triple bond. The reactions afford sulfonium salts 15 as a result of conjugate addition of the CF3CO group and dimethyl sulfide to the alkyne molecule.30 The demethylation of these salts upon the reaction with dimethyl sulfide results in the formation of trifluoro- methyl a,b-enones 16.Oxidation of the enones 16 by hydrogen peroxide makes it possible to synthesise unsaturated compounds with two electron-withdrawing substituents at the double bond, namely, sulfonyl and CF3CO groups. Ph AlkCF3CO2BF ¡ SMe2 3 PhC CAlk + (CF3CO)2O, BF3 . SMe2 760 to740 8C COCF3 Me2S15 (90% ± 95%) Ph Ph Alk Alk H2O2 MeO2S COCF3 (88% ± 93%) MeS COCF3 16 (87% ± 96%) The reactions of pentafluoronitroacetone 17 with vinyl ethyl ether,31 ketene diethyl acetal 32 or isobutylene 33 give rise to products similar to those formed upon trifluoroacetylation of alkenes with trifluoroacetic anhydride. The reaction with acti- vated alkenes follows a cycloaddition pathway. Thus the com- pound 17 and vinyl ethyl ether react at 0 ± 5 8C to give oxetane 18.Opening of the oxetane ring and subsequent thermolysis of the resulting alcohols in the presence of catalytic amounts of K2CO3 affords 4-ethoxy-1,1,1-trifluorobut-3-en-2-one 19 (yield 96%). Ketene diethyl acetal enters into a similar reaction with the compound 17 (at 730 8C), 4,4-diethoxy-1,1,1-trifluorobut-3-en- 2-one being formed in 90% yield. NO2F2C OEt NO2F2C O F3C O+ 17 F3C 18 (90%) OEt440 OEt NO2F2C OEt K2CO3, 100 ± 140 8C F3C 7CHF2NO2 19 (64%) OH CF3CO The reaction of pentafluoronitroacetone 17 with isobutylene yields an alcohol, which eliminates difluoronitromethane to give 4-methyl-1,1,1-trifluoropent-3-en-2-one. NO2F2C Me 20 8C, 12 h K2CO3, 100 8C 17+ F3C 7CHF2NO2 Me OH Me (96%) Me Me CF3CO(48%) d.Trifluoroacetylation of vinylic and allylic organometallic compounds Trifluoroacetylated derivatives of non-activated alkenes can be prepared by acylation of vinylic organometallic compounds. The first trifluoromethyl-containing enone, 1,1,1-trifluoro-4-phenyl- but-3-en-2-one 20, was synthesised by the reaction of styrylmag- nesium bromide with trifluoroacetic acid in 1959; however, the yield of the product was relatively low.34 The subsequent studies showed that organolithium derivatives are better suited for the synthesis of trifluoromethyl enones; in this case, higher yields of the target products can be attained.35 However, only one example of this type of synthesis is known to date; apparently, this is due to the fact that the corresponding vinyllithium derivatives are difficult to obtain.a or b Br COCF3 Ph Ph 20 (a) Mg, Et2O; CF3COOH (yield 11%); (yield 63%). (b) Li, Et2O; CF3CON Trifluoromethyl a,b-enones can also be prepared from allylic organometallic compounds. For example, the reaction of allyl- magnesium bromide with sodium trifluoroacetate gives a mixture of the corresponding a,b- and b,g-enones (ratio *1 : 1) in an overall yield of 37%.36 CF3CO2Na CHCH2MgBr CH2 0± 15 8C, 5 ± 8 h CHCH2COCF3+MeCH CHCOCF3 CH2 When lithium diisopropylamide is made to react with N-ethylidene-tert-butylamine, a heteroallylic anion is formed; subsequent treatment of the anion with ethyl trifluoroacetate gives rise to 4-(tert-butylamino)-1,1,1-trifluorobut-3-en-2-one.37 7NCMe3 LiNPri2 MeCH NCMe3 H2C NCMe3 CF3CO2Et 7 F3C H2C 775 8C, 30 min NCMe3 O H (86%) Reactions of other vinylic or allylic organometallic com- pounds with trifluoroacetic acid or its derivatives have not been studied.Thus, trifluoroacetylation of alkenes is a convenient method for the synthesis of trifluoromethyl enones; however, it is applicable only to alkenes containing either an activating group (e.g., OR, SR, TeR, NR2) or, at least, a phenyl, vinyl or V G Nenaidenko, A V Sanin, E S Balenkova cyclopropyl substituent or two alkyl substituents (trifluoroacety- lation in the presence of Me2S . BF3). 2. Condensation of 1,1,1-trifluoroacetone and its derivatives; construction of a double bond upon an elimination reaction The condensation of 1,1,1-trifluoroacetone with aromatic or a,b- unsaturated aldehydes catalysed by the piperidine ± acetic acid system in THF makes it possible to prepare trifluoromethyl- containing conjugated enones, dienones and polyenones (reti- noids).38, 39 The reaction is stereospecific and gives E-isomers of enones.A drawback of this method is self-condensation of 1,1,1- trifluoroacetone during the reaction; therefore, it should be taken in more than 10-fold excess. NH, AcOH R RCHO +MeCOCF3 COCF3 20 8C, *4 h (28% ± 85%) a,b-Unsaturated ketones are often prepared by condensation of b-dicarbonyl compounds with aldehydes and ketones (Knoe- venagel condensation). However, b-dicarbonyl compounds con- taining a perfluoroalkyl substituent enter into this reaction with difficulty.Thus the reaction of aromatic aldehydes with 1,1,1- trifluoroacetylacetone yields a mixture of products resulting from condensation at the methyl and methylene groups in a relatively low yield.40 This fact was explained by assuming that the addition of amine, used as the catalyst, to the initial 1,1,1-trifluoroacetyl- acetone is the predominant reaction route. NH ArCHO+CF3COCH2COMe COMe Ar Ar Ar CF3+ COCF3 + O COCF3 H O(0% ± 14%) (0% ± 24%) (1% ± 6%) The reactions of b-diketones having one or two perfluoroalkyl groups with polyfluorinated aldehydes have also been studied.41 In the case of b-diketones with one perfluoroalkyl group, the carbonyl group adjacent to the perfluoroalkyl substituent, being more active, participates in the formation of the cyclic intermedi- ate.As a consequence, the elimination of a carboxylate anion gives rise to b-unsaturated ketones with a non-fluorinated substituent at the carbonyl group. K2CO3 or Et3N R1COCH2COR2+R3CHO 20 8C, 1 ± 2 h 7OCHR3 O CHR3 CHCOR2 R1C R1CCHCOR2 7R1CO¡2 O O7 R3 COR2 (20% ± 70%) R1=CF3, H(CF2)n; R2=Me, Ph, CF3, H(CF2)n; R3=H(CF2)n. When esters of trifluoroacetic acid are made to condense with aldehydes in the presence of traditional catalysts such as piper- idine ± AcOH, the target products are formed in fairly low yields.42 Recently, it has been proposed to use functionalised silica gel [silica gel treated with (3-aminopropyl)triethoxysilane] as the catalyst with toluene-p-sulfonic acid being added subsequently to the reaction mixture.A mixture of Z- and E-isomers of alkoxycarbonyl-substituted trifluoromethyl a,b-enones was obtained in this way in a good yield.42Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis COCF3 COCF3 R1 R1CHO+ 1. SiO2 2. p-TsOH PhH, D CO2R2 CO2R2 (43% ± 70%) R1=Alk, Ar; R2=Et, Pri. The Knoevenagel condensation of pefluoroalkyl-containing b-dicarbonyl compounds is often accompanied by side reactions (e.g., self-condensation, in the case of 1,1,1-trifluoroacetone), the yields of the target products being relatively low. This fact has stimulated the search for new methods for their synthesis.For example, the reaction of b-iminophosphonate anions with alde- hydes (the Horner ± Emmons reaction), which yields perfluoro- alkylated enones as the final products after acid hydrolysis, has been used for this purpose. Two approaches to the synthesis of b-iminophosphonates have been proposed. Perfluorocarboxylic acid chlorides can be converted via a four-stage procedure into compounds 21, which are then treated with a base (n-butyllithium, lithium diisopropylamide or sodium hydride) and with an alde- hyde.43 Ketones do not enter into this reaction. O P(OEt)3 P(OEt)3 RFCF2COCl 0720 8C, 2 h RFCF2CP(O)(OEt)2 OP(O)(OEt)2 RF 1. BuLi, CuI, THF 2. NH4Cl (sat.) 778 8C, 15 min F P(O)(OEt)2 H RF NMe MeNH2 1.B 2. RCHO P(O)(OEt)2 20 8C, 2 h F RF 21 (100%) P(O)(OEt)2 (60% ± 79%) NMe H3O+ RFCO R RF R (74% ± 95%) RF=CF3, C2F5, n-C6F13; R=Alk, Ph; B=BunLi, LiNPri2, NaH. b-Iminophosphonate anions can also be obtained from diethyl alkylphosphonates and trifluoroacetoimidoyl chloride. The synthesis is carried out as a one-pot procedure; first, diethyl alkylphosphonates are treated with two equivalents of lithium diisopropylamide and then with trifluoroacetoimidoyl chloride; after that, the required aldehyde is added to the resulting b-iminophosphonate anion.44 2 1. LiNPr i F3C R1 CF3 2. NPh Cl R1CH2P(O)(OEt)2 1. R2CHO 2. H3O+ 770 to 20 8C (EtO)2P NPh 770 8C O Li R1 R2 COCF3 (48% ± 72%) R1=H, Me; R2=Alk, Ar.A method for the synthesis of substituted trifluoromethyl dienyl ketones with a particular configuration at the double bond has been reported.45 Thus sulfoxides 22 were prepared from 1-trifluoromethyl-2-phenylthioacetylene via a four-stage proce- dure (Claisen rearrangement is the key stage of the process). Subsequent elimination of sulfinic acid from the sulfoxides 22 gave trifluoromethyl dienyl ketones 23 and 24 in high yields. 441 R2 R3 OH SPh F3C R1 [F3CC CSPh] NaH Br R2 F3C SPh CCl4, D R3 R3 O SPh O CF3 R1 R2 R1 (70% ± 98%) R3 m-ClC6H4CO3H COCF3 R1 CH2Cl2 SPh R2 (97% ± 100%) R3 CH2Cl2, D COCF3 R1 SOPh R2 22 COCF3 R3=H R1 R2 23 (83% ± 87%) R3 R1=H COCF3 24 (82% ± 87%) R2 R1, R2, R3=H, Alk, Ph.3. Addition to alkynyl ketones Alkynyl trifluoromethyl ketones are easily available compounds, readily formed upon acylation of lithium or magnesium acetylides (Iotsitch reagents) with trifluoroacetic acid or its derivatives.46 ± 50 The best yields were attained in the reaction of lithium acetylides with ethyl trifluoroacetate in the presence of boron trifluoride etherate.50 Et2O. BF3 RC CLi+CF3CO2Et 778 8C RC CCOCF3 (64% ± 83%) R=Alk, Ph. A method for the synthesis of trifluoromethyl a,b-enones is based on 1,4-addition of dialkylcuprates to acetylenic ketones. This reaction is highly regioselective but it is not stereoselective and gives products in moderate yields.50, 51 OH R1 R22 CuLi COCF3+ R1C C CF3 R1C CCOCF3 R2 R2 (0% ± 25%) (20% ± 64%) R1=Ph, Bun; R2=Me, Bun, But.The use of cyanocuprates results in higher yields and in a nearly 100% regioselectivity of the reaction. However, in some cases (for example, in the case of 1,1,1-trifluoro-4-phenylbut-3- yn-2-one), cyanohydrins are produced.51CF3 OH Ph NC LiMeCuCN OH C +PhC PhC CCOCF3 CF3 Me CN The reactions of alkynyl trifluoromethyl ketones with aro- matic amines afford b-amino-substituted trifluoromethyl enones in good yields.52442 ArHN COCF3 ArNH2 RC CCOCF3 MeOH, 20 8C R (77% ± 99%) R=Alk, Ar. Alkynyl trifluoromethyl ketones can act as dienophiles in the Diels ± Alder reaction. 1,1,1-Trifluoro-4-phenylbut-3-yn-2-one reacts with cyclopentadiene to give bicyclic trifluoromethyl enone 25, a norbornadiene derivative, which undergoes a rever- sible rearrangement to give compound 26 of the quadricyclane series.53 hn COCF3 PhC CCOCF3+ D, CF3CO2H Ph 25 COCF3 26 Ph 4.Addition to enones containing a replaceable group in the b-position A method for the synthesis of trifluoromethyl enones based on the reactions of various zinc dialkyl- and diaryl-cuprates with b-trifluoroacetylvinyl tellurides has been proposed.54 R22 CuX COCF3 R2 R1Te COCF3 778 to730 8C (56% ± 90%) R1=Bun, Bui, Ph; R2=Ar, Alk; X=(CN)(ZnCl)2, ZnCl. However, in the case of zinc dialkylcuprates, the reaction is accompanied by side formation of the double addition products, Alk2CHCH2COCF3, in 11%± 19% yields.The necessity of using organotellurium compounds is also an obvious drawback of this method. A new procedure for the synthesis of trifluoromethyl enones from b-trifluoroacetylvinyl sulfones 27 has been reported quite recently. The compounds 27 are formed as stable diols upon oxidation of trifluoroacetylated vinyl sulfides.55 Thus b-trifluoroacetylvinyl sulfone 27 (R=Ph) reacts with electron-rich aromatic compounds (furans, indoles and pyrroles) to liberate sulfinic acid and to give aryl- and hetaryl-substituted trifluoromethyl a,b-enones.55 The reaction is stereospecific yield- ing E-isomers of the enones. H2O2, CF3COOH RS COCF3 730 8C, 3 h HetH, CH2Cl2 Het RO2S COCF3 C(OH)2CF3 20 8C (64% ± 86%) 27 (94% ± 98%) R=Me, Ph. b-Alkoxy-substituted trifluoromethyl enones (most often, 4-ethoxy-1,1,1-trifluorobut-3-en-2-one) are used as the starting compounds in a large number of syntheses.These compounds can be readily prepared by trifluoroacetylation of vinyl ethers.56 Several studies consider the reactions of the enone 19 with diverse nucleophiles�electron-rich aromatic compounds, organometal- lic and organoboron compounds and amines. This reaction in the presence of zinc chloride can be carried out only for reactive aromatic compounds such as indoles, pyrroles 57, 58 and N,N- dimethylaniline.57 COCF3 COCF3 X EtO XH, ZnCl2, CH2Cl2 20722 8C, 3 ± 14 h 19 (20% ± 84%) X= R1 , NR2, 4-Me2NC6H4; R1=H, Me, Ph; NH R2=H, Me. V G Nenaidenko, A V Sanin, E S Balenkova Less reactive compounds such as anisole do not enter into this reaction, and the use of other Lewis acids (BF3 or TiCl4) results in resinification.57 The reactions of the enone 19 with other hetero- cyclic compounds such as furan, 2-methylfuran and thiophene under the same conditions do not lead to the desired result either.58 The reaction of 3-ethoxymethylidene-1,1,1,5,5,5-hexafluoro- pentane-1,3-dione with indoles and pyrroles occurs faster and gives products in higher yields.58 COCF3 COCF3 RH, ZnCl2, CH2Cl2 20 ± 22 8C, 3 ± 14 h EtO R COCF3 COCF3 (84% ± 88%) NMe.Me , R= NH The enone 19 reacts with quaternary salts derived from nitrogenous heterocycles to give dienones�d-trifluoromethylbu- tadienylmerocyanines 28 or 29.59 Z Z Me COCF3 19+ +NMe NMe 28 (75% ± 78%) X7 Z=S, Me2C, CH=CH.Me Et3N, EtOH 19+ 60 8C, 1 h MeN COCF3 N + Me X7 29 (70%) 4-Butoxy-1,1,1-trifluorobut-3-en-2-one enters into a similar reaction. The enone 19 reacts with the phenylmagnesium bromide to give a mixture of b-trifluoroacetylstyrene 20 and allylic alcohol 30, resulting from its reduction, in an overall yield of 40%± 60%, the proportion of the alcohol 30 increasing with an increase in the reaction time.57 The reaction of phenylmagnesium bromide with 4-diethylamino-1,1,1-trifluorobut-3-en-2-one, prepared from the enone 19, occurs more unambiguously;57 it gives only b-trifluoro- acetylstyrene 20 in 53% yield. PhMgBr, Et2O COCF3 EtO 710 to 20 8C 19 CH(OH)CF3 COCF3 + Ph Ph 30 2021 : 1 Reaction time /h The 20 : 30 ratio Et2NH PhMgBr, Et2O 20 1 : 7 COCF3 20 19 Et2N 0 to 20 8C, 20 h Recently, a new convenient and fairly versatile method for the synthesis of unsaturated trifluoromethyl ketones has been devel- oped; this method is based on the reactions of various organo- lithium compounds with 4-dimethylamino-1,1,1-trifluorobut-3- en-2-one 31 and 3-(N,N-dimethylaminomethylidene)-1,1,1,5,5,5- hexafluoropentane-2,4-dione 8c.The compound 8c can be easily prepared by acylation of the enaminone 31 with trifluoroacetic anhydride in the presence of pyridine.60 COCF3 (CF3CO)2O, Py, CH2Cl2 COCF3 Me2N 710 to 20 8C, 1 h COCF3 Me2N 31 8c (95%) The reactions of the trifluoromethyl enaminones 31 and 8c with organolithium compounds afford trifluoromethyl enones; inMethods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis the case of the enaminone 31, the reactions are stereospecific and give only E-isomers of the enones.60 COCF3 1.RLi, THF,770 to730 8C 31 (or 8c) 2. H3O+, 0 8C R1 R2 (30% ± 97%) R1=Ar, Het; R2=H, COCF3. The reaction of enaminodione 8c with 1-methyl-1H-2-indo- lyllithium 32 gives rise to the corresponding enedione 33 but does not stop at this stage; the subsequent cyclisation yields cyclo- penta[b]indole derivative 34.60 CF3CO 1. 8c COCF3 Li 2. H3O+, 0 8C NMe 32 NMe 33 CF3 HO COCF3 NMe 34 (63%) The same study 60 describes the reaction of the enaminone 31 with a mixture of lithiated ferrocenes, resulting in the formation of the corresponding mono- 35 and bis-enones 36.In the case of the enaminodione 8c, only the product of its interaction with mono- lithioferrocene, enedione 37, was isolated. Li Li 1. THF, 770 to760 8C 2. H3O+, 0 8C BuLi, THF Fe Fe Fe + 20 8C, 2 h Li COCF3 COCF3 31 Fe Fe + COCF3 36 (8%) 35 (38%) COCF3 COCF3 8c Fe 37 (35%) In order to synthesise conjugated trifluoromethyl enones containing an acetylenic fragment (enynones), which are promis- ing from the synthetic point of view, a method based on the reaction of lithium acetylides with the enone 31 has been developed.61 The reaction is stereospecific giving E-isomers of substituted trifluoromethyl enynones.1. 31 RC C RC CLi 2. H3O+, 0 8C COCF3 (26% ± 57%) R=Alk, Ar, Me3Si. The reactions of various trifluoroacetylated vinyl ethers with organoboron compounds, which occur with high stereoselectivity (92% ± 98%) to give the E-isomers of the corresponding dienones, have been studied.62, 63 Thus the reaction of alkenyldialkoxybor- anes with substituted vinylic ethers in the presence of boron trifluoride etherate affords the corresponding trifluoromethyl dienones.62 Similarly, 6-butyl-1,1,1-trifluorohex-3-en-5-yn-2-one is formed in the reaction of alkynyldialkoxyborane 38 with the enone 19.63 443 R4 R1 Et2O. BF3, CH2Cl2 B(OR3)2+EtO Alk COCF3 0 to 20 8C, 1 ± 120 h R2 R1 COCF3 Alk R2 R4 (67% ± 90%) R1=H, Me, Ph, Br; R2=H, Br; R3=Et, Pri; R4=H, Me, Ph.Et2O.BF3, CH2Cl2 BuC COCF3 20 8C,*140 h CB(OPri)2 + EtO 38 19 BuC C COCF3 (71%) b-(Thio)alkoxy-substituted trifluoromethyl enones react with ammonia and primary and secondary amines (including aromatic ones) at room temperature to give b-amino-substituted enones (enaminones) in high (sometimes quantitative) yields.64, 65 R1X R3R4N COCF3 COCF3 R3R4NH 0 ±20 8C R2 R2 (56% ± 100%) X=O, S; R1=Alk; R2=H, Me, Ph; R3=H, Alk; R4=H, Alk, Ar. Diamines enter into a similar reaction. Thus symmetrical enaminones have been synthesised by the reaction of ethylenedi- amine and o-phenylenediamine with two molecules of b-ethoxy trifluoromethyl enone 19.64 The route of transamination in the series of b-amino-substituted enones does not depend on the nature of the solvent, being determined by the ratio of the initial reactants and the basicities of both the entering and leaving amines.The highest yields are attained when the amine entering into the reaction is more basic than the leaving one (for example, an amino group can easily be replaced by a dimethylamino or methylamino group but not vice versa). The reactions occur at room temperature and do not require a large excess of the amine used.65, 66 R4R5N R1R2N R4R5NH COCF3 COCF3 7R1R2NH R3 R3 (75% ± 100%) R1=H, Alk; R2, R3=H, Alk, Ph; R4=H, Alk; R5=Alk. 5. Preparation of trifluoromethyl ketones from b-diketones Trifluoromethyl enaminones can be synthesised by the reaction of amines with b-diketones.When fluorinated b-diketones are made to react with ammonia,67 ammonium acetate, ammonium hydro- gen carbonate,68 alkylamines or dialkylamines, stable salts 39 are initially formed. Refluxing of the salts 39 obtained from b-diketones having one trifluoromethyl (or perfluoroalkyl) sub- stituent in benzene or toluene with azeotropic distillation of water affords b-amino-substituted enones in 12%± 76% yields.67 + R2R3NH2R1 RF b a 7 O O R1 RF R1 RF 39 (100%) NR2R3 O O O R1 RF c O H O RF=H(CF2)n (n=2, 4); R1=Alk, Ar; (a) R2R3NH (R2, R3=H, Alk), 20 8C; (b) PhH or PhMe, D; (c) R2R3NH (R2=Ph, R3=H), MeOH, 20 8C.444 In the case of low-basicity aromatic amines (e.g., aniline), salts are not formed and the correding enamines are obtained by prolonged keeping of the initial compounds in polar solvents such as methanol at room temperature (yields 70% ± 90%).67 As a rule (or always if R1=Alk), this reaction gives enami- nones in which the amino group is removed from the perfluoro- alkyl substituent.Apparently, the electron-withdrawing influence of this group stabilises the gem-amino-alcoholic fragment and hampers elimination of water in the case where the amine has added at the carbonyl group adjacent to the perfluoroalkyl substituent. b-Diketones having two perfluoroalkyl substituents react with amines, including aromatic ones, to give stable salts, which cannot be converted into the corresponding amino-substituted enones.69 However, these enones can be obtained by the reaction of ammonia or aliphatic amines with fluorine-containing b-chloro- vinyl ketones 40.70, 71 The compounds 40 are formed in the reaction of polyfluorinated b-diketones with SOCl2 in the pres- ence of DMF as a catalyst or with the Vilsmeier reagents [DMF± POCl3 orDMF± (COCl)2].71 ± 74 This approach has also been used to synthesise hexafluoromonothioacetylacetone 41, which exists in the enol form.72 RF RF RF RF SOCl2, DMF O 90 8C, 3 h O Cl H O 40 RF RF NH3 NH2 O (42% ± 65%) CF3 F3C NaSH SH O 41 (37%) RF=H(CF2)n.Similarly, treatment of hydroxytetrahydrofuranone 42 � the hemiketal form of a diketone�with ammonia is accompanied by opening of the furan ring and gives rise to the corresponding hydroxy enamino ketone 43.75 HO O O CF3 Me CF3CO2Et LiH OH OH O H2N COCF3 NH3 CF3 O OH OH 43 (45%) 42 The structure and the stereochemistry of b-amino-substituted trifluoromethyl enones were studied by NMR and IR spectro- scopy.64, 76 The configuration of their molecules was found to depend on the nature of the amino substituent. Trifluoromethyl enaminones with NH2, AlkNH or ArNH groups in non-polar solvents or neat exist entirely as Z-isomers, which is apparently due to the presence of an intramolecular hydrogen bond.In polar solvents such as acetonitrile, an equilibrium between the Z- and E- isomers is established. RNH RNH O CF3 CD3CN CF3 CCl4 E-s-E CF3 E-s-Z O RNH O Z-s-Z (20% ± 25%) b-Dialkylamino-substituted enones always exist as E-isomers.Their 1HNMR spectra exhibit two pairs of signals corresponding V G Nenaidenko, A V Sanin, E S Balenkova to the diethylamino group, which seems to be due to the hindrance of its rotation around the C(4) ±N bond caused by conjugation.64 + O O7 Et2N Et2N CF3 CF3 An attempt has been made 77 to prepare unsubstituted perfluoroalkyl enone � trifluoromethyl vinyl ketone � from ethyl trifluoroacetoacetate. The researchers found that this enone (like other unsubstituted perfluoroalkyl vinyl ketones) dimerises spontaneously at temperatures above 730 8C to give dihydro- pyran derivative 44. OH TsCl 1. NaBH4 2. LiAlH4 CF3COCH2CO2Et CF3CHCH2CH2OH (83%) OH OH KCl Na2Cr2O7, H2SO4 CF3CHCH2CH2Cl (61%) CF3CHCH2CH2OTs (62%) PhNEt2 [CF3COCH CH2] CF3COCH2CH2Cl (36%) CF3 O F3C COCF3 O F3C O 44 When two moles of methylmagnesium bromide are made to react with one mole of trifluoroacetylacetone, methylmagnesium bromide adds to the trifluoroacetylacetone anion formed initially (mostly at the carbonyl group further removed from the CF3 substituent). The resulting tertiary alcohol 45 is readily dehy- drated (to a large extent, during the hydrolysis) to give 1,1,1- trifluoro-4,4-dimethylbut-3-en-2-one.78 OH O O O H 1.MeMgBr 2. H3O+ Me F3C 7H2O Me F3C 45 Me Me CF3CO Me (55%) 6. Addition of perfluoroalkyl organolithium compounds to acrylic esters Perfluoroalkyl organolithium compounds, generated by the reaction of perfluoroalkyl iodides with methyllithium (trans- metallation reaction), react with methyl acrylate or methyl methacrylate to give perfluoroalkyl enones, which exist in the reaction medium as anions of the corresponding hemiacetals 46.Pefluoroalkyl a,b-enones having no substituents at the b-position are unstable and readily dimerise to give dihydropyran deriva- tives. Therefore, the hemiacetals 46 were introduced without isolation in reactions with isothiouronium salts or amidines, giving 4-perfluoroalkyltetrahydropyrimidines 47 in good yields.79 R1 CO2Me n-CnF2n+1I, MeLi, LiBr, Et2O 778 8C, 1 h NH2 CF3 HO + OLi R1 NH2 N MeO R1 CnF2n+1 R2 X7 MeOH, 20 8C R2 N 46 H 47 (61% ± 98%) R1=H, Me; R2=AlkS, Me, Ph; X=I, Cl; n=2, 4, 6, 8.Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis This method has not been used to prepare trifluoromethyl enones, probably, because trifluoromethyl iodide is a toxic gaseous compound, inconvenient to work with.3 7.Palladium-catalysed reactions Imino-derivatives of alkenyl (48) and alkynyl (49) trifluoromethyl ketones can be synthesised by condensation of trifluoroacetimi- doyl iodides 50 with various alkenes and alkynes in the presence of a palladium catalyst; no elimination of fluorine, typical of this type of reaction, occurs.80 The reaction is carried out under mild conditions, alkynes reacting much faster and at lower temper- atures than alkenes.R1 R1 F3C I NAr 48 (48% ± 95%) Pd NAr F3C F3C HC CR2 C CR2 50 ArN 49 (47% ± 92%) R1=CO2Me, Ph, CN, SO2Ph, C6H13; R2=Ph, CO2Me, CO2Et, C6H13, CH2OCOPh. 8. Oxidation of allyl alcohols Perfluoroalkyl-containing allyl alcohols can be synthesised in two ways, namely, by the addition of vinyl-containing organometallic reagents to fluoral 81 or to another perfluorinated aldehyde or by the addition of perfluoroalkyl organometallic compounds to a,b- unsaturated aldehydes.82, 83 In both cases, the corresponding halide and the metal are directly introduced in the reaction, because perfluoroalkyl organometallic compounds are unstable and vinyl organometallic compounds are usually formed in low yields. To increase the yield of perfluoroalkyl-containing allylic alcohols, ultrasonic treatment (yield 40% ± 70%) 82, 83 and catal- ysis by palladium or nickel complexes (yield 19%± 62%) 83 have been employed.a CHBr CF3CHO+Mg+RCH OH b RCH CHCHRF RFX +Zn+RCH CHCHO c RFI+Zn+PhCH CHCHO R=H, Me, Ph, Me3Si; RF=CF3, C2F5, C3F7; X=Br, I; (a) ultrasound, THF, 20 8C, 3 h; (b) ultrasound, DMF, 100 8C, 1 h; (c) (Ph3P)2PdCl2 or (Ph3P)2NiCl2, DMF, 20 8C, 3 ± 4 h. In addition, trifluoromethyl-containing allylic alcohols have been prepared by the reduction of alkynyl ketones with lithium tetrahydroaluminate.84 The reaction is stereospecific and gives the alcohols as E-isomers. OH LiAlH4 (2.1 equiv.) RC CCOCF3 THF, D, 8 h R CF3 (71% ± 97%) R=CnH2n+1 (n=4 ± 16), Ph(CH2)2CH(OSiMe2But).Trifluoromethyl-containing allylic alcohols can be oxidised into the corresponding enones on treatment with the Dess ± Martin reagent 84 ± 87 or Swern reagent 88 (DMSO± oxalyl chlor- ide ± triethylamine). Manganese dioxide in CH2Cl2 was also used for this purpose; however, in this case, the yield of the product was somewhat lower.89 (OAc)3 IO But But OH O COCF3 CH2Cl2, 20 8C, 3 h Ph Ph CF3 (85%) 445 OH COCF3 Ph 1. DMSO ± (COCl)2, CH2Cl2,760 8C 2. Et3N, 20 8C Ph CF3 (84%) OH COCF3 MnO2, CH2Cl2 CF3 20 8C (65%) Thus, despite the high yields of the target enones, this method is limited due to the poor accessibility of the initial allylic alcohols and has not found wide use. III. The use of a,b-unsaturated trifluoromethyl ketones in organic synthesis The first example of using a trifluoromethyl enone (b-trifluoro- acetylstyrene) for the synthesis of pyrazolines dates back to 1959;34 however, vigorous studies along this line started in the last decade.Trifluoromethyl a,b-enones are usually employed to prepare heterocyclic compounds. Attempts to use them for other purposes, for example, as protective reagents in peptide synthesis, have also been made (see below). All trifluoromethyl enones can be divided into two main types: (a) enones containing a group capable of being substituted (e.g., OR, SR, NR2) in the b-position and (b) enones without such group. b-Alkoxy- and b-amino- substituted trifluoromethyl enones as well as alkynyl trifluoro- methyl ketones (a,b-ynones) havthe widest application.Below we consider the use of these compounds in organic synthesis, except for their use to prepare trifluoromethyl-contain- ing enones from acetylenic ketones and enones of type (a), which is described in Section II.3. 1. Reactions of b-alkoxy-substituted enones and acetylenic ketones Enones containing a b-alkoxy group (trifluoroacetylated vinyl ethers) and alkynyl trifluoromethyl ketones are readily available reagents (see Sections II.1 and II.3). Their reactions with various nucleophiles are widely used to synthesise five-, six- and seven- membered heterocycles and other compounds containing a CF3 group. Hydrazine and methylhydrazine react with b-alkoxy-substi- tuted trifluoromethyl enones to give trifluoromethyl-containing pyrazoles 50.61, 90 The reaction of trifluoromethyl enones with phenylhydrazine affords pyrazolines 51.On treatment with sulfuric acid, the compounds 51 are not dehydrated.90 Appa- rently, the conjugation of the nitrogen lone electron pair with the benzene ring hampers aromatisation. In addition, the electron- withdrawing influence of the CF3 group stabilises the gem-amino- alcoholic fragment. R1 R2 R4=H,Me N F3C NR4 R1 OR2 R4NHNH2, EtOH, D 50 (59% ± 98%) R1 R3 R2 CF3CO R4=Ph F3C N HO NPh 51 (70% ± 82%) R1, R3=H, Me; R2=Me, Et. The reaction of alkynyl trifluoromethyl ketones with hydra- zine at room temperature gives rise to a mixture of pyrazoles 52 and pyrazolines 53.Subsequent refluxing of the reaction mixture in benzene with azeotropic distillation of water affords only pyrazoles 52 in good yields.91446 H2NNH2 RC CCOCF3 20 8C N NH N NH PhH, D OH 52 (76% ± 92%) 7H2O +R R CF3 CF3 52 53 tert-Butylhydrazones of aldehydes react with b-ethoxy enone 19 giving rise to 4-trifluoroacetylpyrazoles 54. A possible mecha- nism of this reaction includes replacement of the ethoxy group by the hydrazone. The subsequent cyclisation affords pyrazolines 55, which are oxidised to pyrazoles by atmospheric oxygen.92 R R COCF3 COCF3 [O] 19 N HN ButHN7N=CHR AcOH, MeCN, 20 8C, 48 ± 140 h NBut 55 NBut 54 (34% ± 66%) R= Et, Ar. It is noteworthy that the reactions of aldehyde methylhydra- zones do not give the corresponding pyrazoles.In all probability, the bulky tert-butyl group favours cyclisation resulting in the formation of intermediate 55. When the pyrazoles 54 are heated with 90% sulfuric acid, the tert-butyl group is removed, which gives N-unsubstituted 4-trifluoroacetylpyrazoles in high yields. The reactions of b-alkoxy-substituted enones with hydroxyl- amine follow different pathways, depending on the structure of the initial enone. Thus acyclic enones and enones containing no oxygen atom in the ring are converted into isoxazolines 56,9, 64, 93 which are dehydrated on treatment with P2O5 (followed by distillation of the reaction mixture) 64 or concentrated H2SO493 to give the corresponding isoxazoles 57 or 58.R2 R3 R3 COCF3 HO H2NOH. HCl, H2O, Py 35 ± 50 8C, 8 ± 16 h N R2 R1O F3C O 56 (68% ± 97%) P2O5 N R2=R3=H F3C O 57 (68%) F3C H2SO4, 35 8C, 5 h O R2±R3=(CH2)4 N 58 (90%) R1=Me, Et; R2, R3=H, Me; R2±R3=(CH2)4. Cyclic enones prepared by trifluoroacetylation of 2,3-dihyd- rofuran and 3,4-dihydro-2H-pyran are converted in a similar way at 0 ± 20 8C into isoxazolines 59, which result from opening of the furan or pyran ring. However, when the reaction is carried out at higher temperatures, it gives rise to tetrahydrofuran and tetra- hydropyran derivatives 60, formed apparently upon dehydration of aldehyde oximes 61, resulting from recyclisation of the starting enones.9 (CH2)n+1OH b OH N CF3CO CF3 O (CH2)n a 59 NC O HON (CH2)n (CH2)n c HO HO 7H2O F3C O F3C O 60 61 n=1, 2; (a) H2NOH.HCl, H2O, Py; (b) 0±20 8C, 5 ± 170 h; (c) 65±85 8C, 30 ± 170 h. V G Nenaidenko, A V Sanin, E S Balenkova Isoxazolines and isoxazoles were also obtained in good yields in the reaction of alkynyl ketones with hydroxylamine, the regiodirectivity of the reaction being dependent on the reaction conditions.91 Thus ketone 62 reacts with hydroxylamine in an alkaline medium to give isoxazoline 63; on refluxing in benzene with azeotropic distillation of water, this product undergoes dehydration to give the corresponding isoxazole 64. H2NOH, MeONa, MeOH 2 h, D n-H17C8C CCOCF3 62 O N O N PhH, D OH 7H2O CF3 n-H17C8 n-H17C8 CF3 63 64 (64%) This reaction performed in an acid medium gives initially oxime 65, which cyclises on refluxing in benzene giving rise to isoxazole 66, isomeric to the isoxazole 64. NOH PhH, D 62 H2NOH, AcOH, 10% HCl 20 8C, 15 h n-H17C8C CCCF3 65 O N CF3 n-H17C866 (80%) This reaction pathway may be due to the fact that in an acid medium, dehydration is possible, which shifts the equilibrium towards the oxime.In an alkaline medium, the formation of isoxazoline, which exists as the anion under these conditions, appears to be thermodynamically more favourable. The reactions of some b-alkoxy-substituted trifluoromethyl enones with isocyanides have been studied.94 b-Ethoxyvinyl trifluoromethyl ketone 19 reacts with cycohexyl isocyanide to give 2,5-dihydrofuran derivative 67, whereas b,b-diethoxyvinyl trifluoromethyl ketone isomerises under the same conditions to ester 68.OEt C=NC6H11-cyclo EtO COCF3 20 8C, 3 days F3C NC6H11-cyclo 19 O 67 (54%) OEt CO2Et F3C C=NC6H11-cyclo F3C 20 8C, 2 months EtO 68 (70%) O O Et When b-alkoxy-substituted enones are made to react with formamide in the presence ofNH4Cl, with compounds of the urea series 95 or with isothiouronium salts,96 trifluoromethylpyrimi- dines are formed. In the last-named case, the reaction is carried out in the presence of an acid or a base. N N CF3 HCONH2, NH4Cl 160 8C, 2 h (23%) EtO COCF3 HX 19 N X C(NH2 )2 N CF3 a or b, or c (60% ± 75%) (a) HCl, EtOH, 20 8C, 48 h (X=O, S); (b) PhH, D, 6 h (X=NH); (c) AcONa, AcOH, 100 8C, 2 h (X=p-NSO2C6H4NHCOMe).Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis R1 OR3 NH2 Py or HCl, MeOH, H2O + + MeS SO2¡¦ 4 D, 18 ¡À 79 h NH2 CF3CO R2 2 CF3 R1 N SMe R2 N (34% ¡À 94%) R1, R2=H, Me; R3=Me, Et; R1¡ÀR2=(CH2)2, (CH2)3.The reactions of b-alkoxy-substituted enones with enamino nitriles 69 result in the formation of pyridine derivatives contain- ing a CF3 group. Acyclic adducts 70 are formed as intermediates. For R1=H, they can be isolated in good yields on cooling. At room temperature, mixtures of the adducts 70 and cyanopyridines 71 are formed; refluxing of initial compounds in acetonitrile leads to pyridines 71 in high yields.97 CN X CHCl3 H2N 0¡À 5 8C, 24 h COCF3 NC COCF3 R1 70 (68% ¡À 76%) + MeCN, D, 1 h R1 OR2 X R2O 69 X N F3C MeCN D, 2 h CN R1 71 (68% ¡À 95%) Y (Y=O, NMe, NCOOEt).X= N , N Unlike reactions of enamino nitriles with alkoxyenones, which give 2-trifluoromethylpyridines, the reaction of 1,1,1-trifluoro-4- methoxypent-3-en-2-one with cyanothioacetamide results in the regioselective formation of 4-trifluoromethylpyridinethione 72, the structure of which was proved by X-ray diffraction analysis.98 The alkylation of the pyridinethiones 72 obtained by N-aryl- chloro(iodo)acetamides followed by cyclisation of the resulting compounds 73 in an alkaline medium affords 3-amino-4-trifluoro- methylthieno[2,3-b]pyridinecarboxamides 74.CF3 Me CN XCH2CONHAr NCCH2CSNH2 KOH MeO COCF3 S Me CF3 NH 72 (89%)CF3 NH2 CN KOH CONHAr DMF Me N S SCH2CONHAr Me N74 (68% ¡À 85%) 73 (73% ¡À 91%) X=Cl, I; Ar=Ph, 4-MeC6H4, 3-BuOC6H4. The reaction of the butoxy-substituted trifluoromethyl enone, prepared by trifluoroacetylation of vinyl isobutyl ether, with 1,2- diamines (o-phenylenediamine or 1,2-ethylenediamine) gives rise to 1,5-diazepine 75 or 76, while the reactions with o-aminophenols or o-aminothiophenol yield 1,5-oxazepines or 1,5-thiazepines 77, respectively. The bis(trifluoroacetyl) derivative was also intro- duced into this reaction. Good yields were attained by using microwave (MW) radiation, whereas conducting the reaction in boiling xylene resulted in a complex mixture of products (Scheme 1).99, 100 447 Scheme 1 COCF3 MW xylene, 8 ¡À 25 min X BuiO R1 NH2 HN R1(R2) R2 NH2 X R2(R1) N 75 (73% ¡À 93%) CF3 HN H2N NH2 X N CF3 76 (73% ¡À 77%) R3 NH2 HN R3 YH X CF3 YHO 77 (71% ¡À 89%) X=H, COCF3; R1=H, Me; R2=H, Me, Cl, NO2, COPh; R3=H, Cl; Y=O, S.3-Aminopyrazolo[3,4-b]pyridine derivatives 78 enter into a similar reaction as nucleophiles. On exposure to microwave radiation, trifluoromethyl-substituted derivatives of pyrido- [20,30:3,4]pyrazolo[1,5-a]pyrimidine 79 are formed in 62% ¡À78% yields (the yield in the thermal reaction is only 20%).101 R1 NH2 COCF3 MW or D + N X BuiO N R2 N78 R1 N X N CF3 N N R2 79 X=H, COCF3; R1=Me, CF3; R2=Me, Ar.4,4-Diethoxy-1,1,1-trifluorobut-3-en-2-one reacts with o-phe- nylenediamines, o-aminophenols and o-aminothiophenol to give five-membered heterocycles D benzoimidazoles, benzooxazoles and benzothiazoles D in high yields, the reaction occurring both on refluxing in toluene and on exposure to microwave radia- tion.100 R1 NH2 EtO COCF3 MW or + PhMe, D, 8 ¡À 15 min EtO R2 YH R1 N CH2COCF3 R2 Y (86% ¡À 96%) Y=O, S: R1=H, Cl; R2=H; Y=NH: R1, R2=H, Me. The reaction of b-ethoxy(phenoxy)-substituted enones with vinyl ethers affords 3,4-dihydro-2H-pyrans in good yields; thus, these enones are efficient heterodienes in the hetero-Diels ¡À Alder reaction.102448 OR2 O F3C OR2 CF3CO OR1 (72% ± 100%) 80 8C, 30 h OR1 O O F3C O R1=Et OEt (68%) R1=Et, Ph; R2=Alk, Ph.This reaction proceeds especially easily for b-alkoxy-substi- tuted bis(trifluoromethyl) enones. The presence of the second strong electron-withdrawing group, COCF3, increases the reac- tivity of trifluoromethyl enones as heterodienes. The reaction occurs at room temperature to give 5-trifluoroacetyl-3,4-dihydro- 2H-pyrans in high yields.17 OR1 OEt COCF3 COCF3 PhO EtS CF3 PhO CF3 O (100%) O (77%) SEt PhO OPh R1O COCF3 COCF3 (CH2)n OR2 O OEt OR1 COCF3 COCF3 (H2C)n R2O O CF3 CF3 O (86% ± 100%) O (85% ± 98%) R1=Et, Bui; R2=Et, Ph; n=1, 2. Aryl vinyl ethers react with trifluoroacetic anhydride to give the corresponding bis(trifluoroacetyl) derivatives.It was found 17 that these compounds are unstable and cannot be isolated in a pure state; however, they can be introduced without isolation in the reaction with a second equivalent of aryl vinyl ether. Aryloxy- substituted 5-trifluoroacetyl-3,4-dihydro-2H-pyrans 80 were obtained in this way in 50%± 68% yields. In some cases, these products are formed directly in trifluoroacetylation of aryl vinyl ethers.17 ArO ArO COCF3 (CF3CO)2O, Py ArO 40 8C, 24 h 40 8C, 5 h COCF3 OAr COCF3 ArO CF3 O 80 Ar=Ph, 4-MeC6H4, 4-BrC6H4. Ring opening in 5-trifluoroacetyl-3,4-dihydro-2H-pyrans 81 induced by secondary amines in acetonitrile gives rise to 1-amino- 4,4-bis(trifluoroacetyl)buta-1,3-dienes 82. The mechanism pro- V G Nenaidenko, A V Sanin, E S Balenkova posed for this reaction includes elimination of isobutanol, electro- cyclic ring opening and nucleophilic substitution of the amine for the ethoxy group.It is of interest that no reaction of this type occurs in the series of non-fluorinated analogues of pyrans 81.103 OBui COCF3 COCF3 7BuiOH CF3 EtO O EtO CF3 O 81 COCF3 R1R2NH EtO COCF3 COCF3 R1R2N COCF3 82 (77% ± 100%) R1=Alk, Ph; R2=Alk. One trifluoroacetyl group in bis(trifluoroacetyl)-substituted dienes 82 can be eliminated in an acid medium; this gives 1-amino- 4-trifluoroacetylbuta-1,3-dienes, which can be acylated by tri- fluoroacetic anhydride in the presence of pyridine to give the initial compounds 82.103 R1R2N 6MHCl, THF, 30 8C, 2 h 82 COCF3 (CF3CO)2O, Py, 20 8C, 4 h (59% ± 78%) 3,4-Dihydro-2H-pyran 83, prepared by hetero-Diels ± Alder reaction of b-ethoxy trifluoromethyl enone with vinyl ethyl ether, enters into a similar reaction.The dihydropyran ring opening induced by thiols in an acid medium is stereospecific and gives (E)- 1-alkylthio-4-trifluoroacetylbuta-1,3-dienes 84 in good yields.104 OEt RSH, TsOH, CH2Cl2 or PhH 40 ± 80 8C, 0.5 ± 5 h EtO CF3 O83 OEt RS CF3COOH, CHCl3 40 8C, 15 h COCF3 84 (70% ± 81%) O RS CF3 The reactions of b-ethoxy enone 19 with semicarbazide and thiosemicarbazide result in the formation of 2-pyrazoline-1- carboxamide 85 and 5-(1-thiosemicarbazido)-2-pyrazoline-1-thi- ocarboxamide 86, respectively.105 CF3 X=O OH N N a CONH2 85 (73%) COCF3 EtO CF3 H2NC(S)NHNH 19 OH X=S N N86 (22%) NH2CS (a) H2NCXNHNH2, EtONa, EtOH; NH4Cl, H2O.Apart from the synthesis of heterocycles, b-alkoxy-substituted enones have been used to prepare acyclic compounds. Thus the reaction of the lithium salt of (R)-methyl 4-tolyl sulfoxide 87 with the enone 19 (or some other perfluoroalkyl-containing enone) gives a mixture of two diastereoisomers of b-hydroxy sulfoxides 88a,b in a ratio of *1 : 1. These diastereoisomers were separated by chromatography; their configurations were determined by449 Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis 2. Reactions of b-amino-substituted trifluoromethyl enones X-ray diffraction analysis.The subsequent reduction of the sulfoxide group permits the preparation of two enantiomers of chiral allylic alcohols 89a,b with known absolute configurations in a pure state (overall yield 40% ± 94%).106 Me CF3 a OEt p-MeC6H4 S Me S + O 87 O OH 88a (26% ± 36%) CF3 OEt NaI, (CF3CO)2O, Me2O p-MeC6H4 S + 720 8C, 20 min O OH 88b (35% ± 58%) In addition to b-alkoxy-substituted trifluoromethyl enones, their b-dialkylamino-substituted analogues, readily prepared by treat- ment of b-alkoxy enones with dialkylamines, are also used for the synthesis of various trifluoromethyl-containing heterocyclic com- pounds. Since enaminones are less reactive in nucleophilic substitution than b-alkoxy enones, they are activated by virtue of POCl3 or trifluoromethanesulfonic anhydride.Thus the reac- tions of enaminones with POCl3 afford iminium salts 91a,b, vinylogues of the Vilsmeier complexes. Despite the fact that these salts are less reactive than the complex DMF. POCl3, they are still able to aminoformylate (*20 8C, 1 h), e.g., N,N-dime- thylaniline or quaternary salts of nitrogenous heterocycles 92 and 93. These reactions yield hemicyanines 94 ± 96, respectively, which are valuable intermediates in the synthesis of cyanine dyes.59 CF3 CF3 OEt OEt 77 CF3 OP(O)Cl2 p-MeC6H4S + p-MeC6H4S POCl3 R2N COCF3 + 20 8C Cl R2N OH S-89b OH R-89a 91a,b R=Me (a), Et (b). + (a) LiNPr i2,778 8C; EtOCH=CHCOCF3 (19). NEt2 91b Me2N 4 Me2N NaClO4 ClO¡ CF3 94 (65%) Trifluoroacetylated vinyl ethers (including cyclic ones) readily undergo haloform cleavage induced byKOHin boiling benzene in the presence of a small quantity of water to give a,b-unsaturated acids.8 KOH, H2O COCF3 CO2H Z Z CF3 RO RO 91b PhH, D, 1±8 h (60% ± 80%) NEt3, NaClO4 Me NEt2 X7 R=Et, Bun, Bui, Ph.N + Me N + Me ClO¡492 95 (68%) CF3CO HO2C Me KOH, H2O (CH2)n (CH2)n F3C NEt2 PhH, D, 5±7 h 91b O NEt3, NaClO4 + O (81% ± 100%) MeN n=1, 2. N + Me X7 96 (64%) 93 ClO¡4 When two equivalents of the quaternary salt 92 react with one equivalent of triethylamine and with the iminium salt 91, dicarbo- cyanines 97 are formed.59 Z Z Z CF3 a 2 Me X7 NMe N + Me N+Me ClO ¡492 97 (42% ± 60%) The replacement of the alkoxy group in b-alkoxy enones on treatment with amines occurs with high yields and the resulting enaminones are readily hydrolysed. Therefore, these enones can be used as protective reagents in peptide synthesis. Thus the b-ethoxy enone 19 reacts with amino acids in the presence of an equivalent of a base at room temperature to give the correspond- ing enaminones.The amino acids protected in this way were used to synthesise dipeptides 90. The protective group can be easily removed on treatment with 3 M HCl in dioxane at room temper- ature.107 Z=S, CMe2, CH=CH; (a) 91, NEt3, NaClO4, 20 8C, 30 min. R1 CO2H 1. NaOH, H2O, 22 8C, 1 ± 3 h COCF3 + EtO 2. 6 M HCl (pH 3) NH2 19HO2C CF3 R2(CO2R3)CHNH2 .HCl, DCC, Et3N, CH2Cl2 0 8C, 1 h R1 NH O (70% ± 89%) O The electrophilicity of iminium salts proves to be insufficient for them to react with aromatic compounds less reactive than N,N-dimethylaniline. Thus quenching with a solution of KHCO3 of the reaction mixture obtained upon the reaction between the complex of 4-dimethylamino-1,1,1-trifluorobut-3-en-2-one with POCl3 and 1,3-dimethoxybenzene gave a,b-unsaturated aldehyde 98 in 7%± 12% yield,108 more than 80% of the initial 1,3- dimethoxybenzene being recovered unchanged. HN 3 M HCl, dioxane CF3 R3O2C OMe 20 8C, 10 h 7 CF3 OPOCl2 R2 R1 KHCO3, H2O + + Cl NH O (80% ± 97%) O R2N R2 MeO R1 OMe CO2R3 NH CF3 HCl .H2N 90 (*90%) MeO CHO 98 (E:Z=7:1) R1=Me, Pri, (CH2)2CO2Me, (CH2)2CO2But, PhCH2; R2=H, Pri, PhCH2; R3=Et, But, Me; DCC is dicyclohexylcarbodiimide. Due to the low electrophilicity of the iminium salts obtained by the reaction of enaminones with POCl3 it has been proposed to450 use more reactive salt 99, formed from enaminone and trifluoro- methanesulfonic anhydride and containing a better leaving group (OSO2CF3).The reactions of this compound with electron- donating aromatic or heteroaromatic compounds under mild conditions give rise to a,b-unsaturated trifluoromethyl-contain- ing aldehydes.108 The reaction is stereoselective giving mainly E-isomers of the aldehydes; it occurs for relatively reactive compounds such as indoles, pyrroles, furans and 1,3-dimethoxy- benzene, whereas thiophene and anisole do not enter into this reaction, even when it is carried out under more rigorous conditions. (CF3SO2)2O COCF3 Me2N 75 to 0 8C CF3 ArH or HetH + OSO2CF3 Me2N 99 OSO2CF¡¦3 Ar(Het) Ar(Het) KHCO3, H2O +NMe2 F3C F3C OSO2CF¡¦3 CHO E:Z56 : 1 b-Arylamino-substituted trifluoromethyl enones cyclise on treatment with acids giving rise to 2-trifluoromethyl- (100) and 4-trifluoromethylquinolines (101).37, 52, 109, 110 As catalysts, POCl3, ZnCl2 and polyphosphoric acid (PPA) were used.Quino- lines 101 are the products of `normal' cyclisation, while the mechanism of formation of 2-trifluoromethylquinolines 100 needs to be further investigated. The ratio of the reaction products depends on the acidic catalyst used and the structure of the initial enone. Cyclisation of enones with R2=H gives only 2-trifluoromethylquinolines 100,37, 52 and when R2=Alk or Ph, 4-trifluoromethylquinolines 101 are formed predominantly.52 The presence of electron-donating substituents in the meta-positions of the aromatic amine substantially facilitates cyclisation and increases the total yield.NH COCF3 POCl3, 100 8C, 6 h or PPA, 165 8C, 3 h R2 R1 R2 CF3 and/or R2 N CF3 R1 R1 N 101 100 R1=Me, OMe, Hal; R2=H, Alk, Ph; PPA is polyphosphoric acid. R1 O Ph Ph CF3 CO2Et CO2Et N R1 N R1 B7 7 O R2 O R2 CF3 CF3 R1 O CF3 CO2Et S Ph S H CO2Et B7 Ph O CF3 CF3 V G Nenaidenko, A V Sanin, E S Balenkova The reactions of aniline derivatives with the iminium salt 99 afford 2-trifluoromethylquinolines and no 4-trifluoromethyl- derivatives.111 NH2 +99 R N CF3 R R=Me, OMe, SMe, OCF3, Cl, OH.Enaminone 102 cyclises on refluxing in a high-boiling solvent to give substituted 3-fluoromethylpyrrole 103.112 F3C COCF3 1,3,5-Me3C6H3 HNBz D, 8 h Me Bz Me NH102 103 (92%) A detailed study of the mechanism of cyclisation of such enamino ketones in the presence of a base demonstrated that the transformations of trifluoromethyl-containing ketones include not only 5-exo-trigonal cyclisation to give 3-trifluoromethylpyr- roles 104 but also previously unknown 3-exo-trigonal cyclisation, giving 2-trifluoromethylpyrroles 105. Similarly, in the case of sulfide derivatives of enones, 3-trifluoromethylthiophenes were isolated in addition to 2-trifluoromethylthiophenes (Scheme 2).113 Meanwhile, enaminones 106 cyclise in the presence of tri- fluoroacetic acid to yield 3-trifluoroacetylpyrroles.112, 114 COCF3 COCF3 CF3CO2H, H2O HN (MeO)2HC 20 ¡À 50 8C, 4 h R 106 R NH (87% ¡À 100%) R=H, Me, Ph.Reactions of 3-(diethylaminomethylidene)-1,1,1,5,5,5-hexa- fluoropenta-2,5-dione 8a (enaminodione, whose electrophilicity is enhanced by the presence of the second trifluoroacetyl group) with various nucleophiles have been studied.115 When the en- aminodione 8a reacts with aromatic amines in the presence of catalytic amounts of FeCl3, the diethylamino group is substituted to give N-aryl-substituted enaminodiones, which cyclise on treat- ment with PPA or TiCl4, the yields of the reaction products being substantially higher in the case of TiCl4.Scheme 2 R1 F3C R2 N R2 EtO2C N 7 Ph CO2Et Ph 104 (15% ¡À 64%) Ph CO2Et R2 R1 CO2Et R2 R1 N 7 N CO2Et F3 �º 7O N Ph R2 CF3 Ph 105 (26% ¡À 40%)Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis COCF3 Ar(Het)NH2 FeCl3 Et2N COCF3 8a COCF3 TiCl4, CH2Cl2, 20 8C Ar(Het)NH COCF3 (90% ± 100%) Cyclisation Ar(Het)NH product CF3 COCF3 NH N NH CF3CO F3C NH HO COCF3 N CF3 COCF3 F3C NH N N NH N N CF3 HN N COCF3 The reactions of the enaminodione 8a with bifunctional nucleophiles occur similarly to the reactions of b-alkoxy-substi- tuted enones.Thus the reaction of the compound 8a with hydrazine gives rise to trifluoromethyl-substituted pyrazole 107, while its reaction with phenylhydrazine yields a mixture of regioisomers 108 and 109 in 1 : 3 ratio.115 CF3CO H2NNH2 F3C MeCN 8a 20 8C, 4 h CF3CO PhNHNH2 F3C 108 (75%) The reactions of the enaminodione 8a with a-aminocarbonyl compounds (or their acetals)�2,2-dimethoxyethylamine or ethyl N-benzylglycinate�result in the corresponding pyrroles 110 and 111.115 In the reaction with 2,2-dimethoxyethylamine, for aroma- tisation to occur one molecule of CF3CO2H is eliminated; this is the pathway to 3-trifluoroacetylpyrrole 110.114 8a H2NCH2CH(OMe)2, CF3CO2H, H2O 20 8C, 4 h COCF3 PPA, D or COCF3 cyclisation products N Yield (%) BnNHCH2CO2Et, MeCN, 70 8C, 4 h PPA TiCl4 8a F3C 96 10 EtO2C NBn 0 15 The enaminodione 8a has also been introduced in reactions 0 90 with compounds of the urea series.The reaction of 8a with guanidine affords pyrimidine 112 in a good yield.115 The reaction with O-methylisourea affords two products � 1-methoxypyrimi- dine 113 and 1-diethylaminopyrimidine 114 � because diethyl- amine formed in the reaction reacts with methoxypyrimidine 113. Optimisation of the reaction conditions (MeCN, 65 8C, 4 h) makes it possible to prepare the target compound 113 in 65% yield. NH Me2NCNH2 100 7 8a MeCN, 20 8C, 4 h Me2N N 112 (85%) CF3 NH 60 25 N MeOCNH2 8a MeO N 113 (65%) 0 25 Thus, the enaminodione 8a is a readily available reagent, widely used for the synthesis of various heterocyclic compounds containing both trifluoromethyl and trifluoroacetyl groups.3. Reactions of trifluoromethyl enones containing no replaceable group N107 (96%) NH CF3CO + N NPh F3C N NPh 109 (25%) Trifluoromethyl enones containing no b-substituents able to be replaced are less available compounds than b-alkoxy- or b-amino- substituted enones; therefore, their reactions have not been studied so extensively. However, in the early 1990s, new methods for the synthesis of these compounds were developed (e.g., trifluoroacetylation of alkenes in the presence of the complex of dimethyl sulfide with boron trifluoride 25, 26).This stimulated vigorous studies of their synthetic potential. Trifluoromethyl enones containing no replaceable substitu- ents readily react with various bifunctional nucleophiles. Thus the reactions of phenyl-, cyclobutyl- or adamantyl-substituted enones with hydrazine result in the formation of pyrazolidines 115, which contain a gem-amino-alcoholic fragment, stabilised by the elec- tron-withdrawing influence of the CF3 group. Dehydration of the pyrazolidines 115, which occurs either during the reaction, or on prolonged storage, or on refluxing in benzene in the presence of a catalytic amount of toluene-p-sulfonic acid (depending on the nature of substituents R1 and R2) gives the corresponding pyrazo- lines.116 451 C(COCF3)2 (MeO)2CHNH COCF3 H2O 7CF3CO2H HN 110 (83%) CF3CO COCF3 Et2OC NBn COCF3 111 (96%) CF3 COCF3 N CF3 COCF3 COCF3 N + Et2N N114452 R1 COCF3 R1 CF3 H2NNH2 OH R2 7H2O EtOH, 20 8C R2 (80% ± 95%) CF3 R1 R2 HN N (90% ± 95%) R1=H, R2=Ph; R1±R2=(CH2)3, .b-Trifluoroacetylstyrene 20 reacts with substituted aryl- or alkyl-hydrazines. Thus the reaction with p-bromophenyl- and phenyl-hydrazine affords pyrazolines 116. Oxidation of the compound 116 with X=H by lead tetraacetate affords the corresponding pyrazole in a moderate yield.31 The reaction with p-nitrophenylhydrazine does not give pyrazoline; in this case, b-trifluoroacetylstyrene p-nitrophenylhydrazone 117 is formed in a low yield.p-XC6H4NHNH2 COCF3 Ph 20 Ph CF3 Ph CF3 X=H X=H, Br p-XC6H4N N CHCl3, Pb(OAc)4 N 116 (59% ± 67%) PhN(47%) Ph N NHC6H4NO2-p X=NO2 CF3 117 (14%) When b-trifluoroacetylstyrene 20 is made to react with methylhydrazine, a mixture of isomeric pyrazolines 118a,b (in *1 : 3 ratio) is formed, which is apparently due to the fact that the nucleophilicities of the nitrogen atoms in methyl hydrazine are close. The reaction with 1,2-dimethylhydrazine gives pyrazolidine 119, apparently due to the fact that elimination of water yielding a C=C bond is substantially hampered.116 Ph Ph CF3 CF3+ 2MeNHNH2 .H2SO4 AcONa, EtOH D, 5 h N NMe 118b (18%) MeN N 118a (59%) 20 Ph CF3 OH MeNHNHMe.2 HCl AcONa, THF D, 7 h MeN NMe 119 (77%) The enone 20 reacts with semicarbazide or thiosemicarbazide in an acid medium to afford semicarbazone 120a or thiosemicar- bazone 120b. These compounds are stable both in acid and alkaline media and do not tend to cyclise.105 Ph NNHC(X)NH2 20 H2NC(X)NHNH2 . HCl EtOH, D, 6 h 120a,b CF3 X=O(120a, 91%), S (120b, 93%). Conversely, the reaction with semicarbazide carried out in the presence of an equimolar amount of sodium ethoxide gives a cyclic compound, 5-trifluoromethylpyrazolidine-1-carboxamide 121.105 R1 OH R1 R2 CF3 1. H2NC(O)NHNH2, EtONa, EtOH 2. NH4Cl, H2O HN N R2 COCF3 CONH2 121 (46% ± 78%) R1=Ar, R2=H; R1±R2=(CH2)3, . V G Nenaidenko, A V Sanin, E S Balenkova The reaction occurs regio- and stereoselectively, yielding predominantly one diastereoisomer.The bulky substituents R1 and R2 occupy pseudo-equatorial positions, i.e. thermodynami- cally the most favourable isomers are produced.105 When trifluoromethyl-containing enones react with thiosemi- carbazide under the same conditions, either 3-trifluoromethyl-2- pyrazolinethiocarboxamides 122 or 5-trifluoromethyl-2-pyrazoli- nethiocarboxamides 123 are obtained, depending on the structure of the starting enone.105 CF3 R1 N N R1 NH2CS 122 (25% ± 54%) 1. H2NC(S)NHNH2, EtONa, EtOH 2. NH4Cl, H2O R2 COCF3 R1 OH R2 CF3 HN N 122: R1=Ar, R2=H; CSNH2 123 (61% ± 77%) 123: R1=Ph, R2=Me; R1±R2=(CH2)3, . Unlike pyrazolidines 115, pyrazolidine-1-carboxamides and -thiocarboxamides cannot be converted into the corresponding pyrazolines.105 The reaction of b-trifluoroacetylstyrene 20 with hydroxyl- amine in an acid medium gives rise to oxime 124, which does not tend to cyclise in either acid or alkaline medium.117 CF3 COCF3 Ph H2NOH.HCl, EtOH D, 6 h Ph 20 NOH 124 (94%) Meanwhile, the reactions of trifluoromethyl enones with hydroxylamine in the presence of an equimolar amount of sodium ethoxide give isoxazolidines 125 in good yields.117 R1 R1 OH 1. H2NOH, EtONa, EtOH 2. NH4Cl, H2O R2 CF3 R2 COCF3 HN O 125 (72% ± 93%) R1, R2=Ar, Het, Alk. The attempts to carry out dehydration of isoxazolidines 125 to isoxazolines were unsuccessful. Refluxing of these compounds in toluene with azeotropic distillation of water in the presence of toluene-p-sulfonic acid gives rise to the corresponding oxime, the initial enone being formed as a side product.At the same time, b-alkoxy enones or alkynyl ketones are converted into isoxazo- lines, which are dehydrated in an acid medium to give isox- azoles.63, 93 Thus, the possibility of aromatisation has the crucial influence on the reaction route. The isoxazolidines 125 are fairly interesting from the stereo- chemical viewpoint. They exist in solutions at room temperature as equilibrium mixtures (*1 : 1) of diastereoisomers. The energy barrier to nitrogen inversion in the compound 125 is almost a half that in the isoxazolidines described previously by Kostyanovsky et al.;118, 119 however, it is still high enough (51 ± 55 kJ mol71) 117 to retard nitrogen inversion.Therefore, the signals in the 1H and 13C NMR spectra recorded at room temperature are broadened, and when the temperature decreases, the conformation becomes `frozen' and signals for four diastereoisomers are observed (due to the chirality of the nitrogen atom). The reactions of trifluoromethyl enones with compounds of the urea series afford tetrahydropyrimidines 126 in high yields.120 The compounds 126 are easily dehydrated on refluxing in toluene in the presence of catalytic amounts of toluene-p-sulfonic acidMethods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis [unlike cyclohexane derivatives containing a C(OH)CF3 frag- ment]121 giving rise to dihydropyrimidines 127.120 R1 (H2N)2CX, EtOH D, 24 ± 120 h R2 COCF3 OH R1 CF3 R1 PhMe, TsOH R2 NH HN CF3 NH R2HN D, 6 h X 127 (87% ± 95%) X 126 (77% ± 92%) R1=H, R2=Ph; R1±R2=(CH2)3; X=O, S.In the case of b-trifluoroacetylstyrene 20, the reaction occurs stereoselectively, yielding predominantly (595%) one of the two possible diastereoisomers of 126. The configuration of the major diastereoisomer was established by X-ray diffraction analysis. This was found to be the thermodynamically more favourable trans-isomer (regarding the arrangement of the Ph and OH groups), in which the bulkiest substituents, Ph and CF3, occupy equatorial positions.Ph 4 5 HN CF3 3 1 NH H OH XIt should be noted that such stereochemistry has also been observed for cyclohexane derivatives with the same substituents at the 3- and 5-positions;121 this may be due to the relatively large effective size of the trifluoromethyl group { and to the high electron density on this group. The reaction of a sterically hindered trifluoromethyl enone having an adamantane fragment with thiourea gives dihydropyr- imidine 128.120 CF3NH COCF3 (NH2)2CS, EtOH D, 48 h S HN128 (81%) The reactions of phenyl- and adamantyl-substituted trifluoro- methyl enones with guanidine occur in a similar way giving rise to the corresponding aminopyrimidines 129 and 130. CF3 Ph (H2N)2C=NH. 12H2CO3 OH Ph N HN COCF3 EtOH, D, 9 h 20 NH2 129 (77%) CF3N (H2N)2C=NH.12H2CO3 COCF3 NH2 EtOH, D, 9 h NH 130 (81%) b-Trifluoroacetylstyrene 20 reacts with aminoguanidine to give compound 131, resulting from addition of two enone molecules to an aminoguanidine molecule and containing two heterocyclic moieties (tetrahydropyrimidine and pyrazoline).120 { The energy required for substituents to pass from an equatorial to an axial position (conformational energy) was reported 122 to increase in the sequence Me<Pri<CF3<But. 453 In this case, water is eliminated only from the five-membered ring, which is consistent with the general rule according to which trifluoro-substituted pyrazolidines are dehydrated more readily than tetrahydropyrimidines.116, 120 Ph Ph NH NH H2NCNHNH2 .H2CO3 20 N EtOH, D, 9 h N CF3 N F3C OH 131 (61%) The reactions of trifluoromethyl enones with thiourea and thioacetamide in an acid medium afford dihydrothiazines 132.123 Both reactions are regiospecific and give one isomer, formed upon the addition of sulfur at the double bond and nitrogen at the carbonyl group.This reaction route was interpreted 123 in terms of the principle of hard and soft acids and bases. OH R1 R1 COCF3 a or b CF3 R2 N S R2 R3 132 (65% ± 78%) R1=H, R2=Ph; R1±R2=(CH2)3, ; (a) (H2N)2CS, HCl, EtOH, D (R3=NH2); (b) MeC(S)NH2, EtOH, D (R3=Me). The reactions of some trifluoromethyl enones with thiols have been studied. Thus 1,1,1-trifluorodec-3-en-2-one reacts with benzenethiol to give b-oxo sulfide 133.86, 124 Bu4NF, THF n-C8H17 +PhSH COCF3 20 8C, 4 h PhS COCF3 133 (82%) n-C8H17 As opposed to this, the reaction of the enone 20 with 4-methylthiophenol gives rise to two products, b-oxo sulfide 134 and pyran 135.The compound 134 results from Michael addition, while the pyran 135 is formed upon interaction of two molecules of the ketone with one molecule of the thiol and one water molecule.124 4-MeC6H4SH, Et3N 20 EtOH, 20 8C7 Ph Ph H+ COCF3 COCF3 4-MeC6H4S 4-MeC6H4S134 Ph OH OH O COCF3 F3C CF3 H2O 4-MeC6H4S 4-MeC6H4S COCF3 Ph 135 Ph Ph The reaction is stereospecific, the compound 135 being formed as one diastereoisomer of the 16 possible isomers (the molecule has five asymmetric centres); the bulkiest phenyl, trifluoromethyl, and (4-methylphenylthio)phenylmethyl substituents occupy equa- torial positions.The relative configuration of the carbon atom located outside the ring and bearing the phenyl and 4-methylphe- nylthio substituents is also fixed. Thus, effective asymmetric 1,2- induction takes place.124 The reactions of trifluoromethyl enones with ammonium hydrogen sufide have been studied.125 The reaction pathway depends on the structure of the initial enone. For instance, the454 enone 20 reacts stereospecifically yielding tetrahydropyran 136 as one diastereoisomer of the eight possible isomers.125 H Ph S Ph Ph S CF3CO NH4SH, EtOH Ph 20 F3C 20 8C, 30 min COCF3 H H F3C HO OH 136 (92%) The reaction of ammonium hydrogen sulfide with cyclobutyl- substituted enone affords a mixture of cis- and trans-diastereo- isomers of tetrahydrothiopyran 137a,b (total yield 90%) in 1 : 1 ratio.125 S S a COCF3 + COCF3 COCF3 F3C F3C OH 137b OH 137a (a) NH4SH, EtOH, 20 8C, 3 min.The tetrahydrothiopyrans 136 and 137a are oxidised with retention of configuration, giving the corresponding sulfones 138 and 139 in almost quantitative yields. O O R1 R1 R1 R1 S S H2O2, AcOH R2 R2 R2 R2 D, 5 h COCF3 COCF3 F3C F3C OH 136, 137a OH 138, 139 R1=Ph, R2=H(136, 138); R1±R2=(CH2)3 (137a, 139). The reaction of ammonium hydrogen sulfide with adamanty- lidenemethyl ketone follows an unusual route and affords com- pound 140, containing a four-membered thietane ring, as the only product.On treatment with hydrogen peroxide in acetone or acetic acid, this product is oxidised to give 1,3-sultine 141 as a mixture of diastereoisomers.125 CF3 S COCF3 NH4SH, EtOH OH a or b 20 8C, 30 min 140 (86%) O O S OH CF3 141 (a) H2O2, Me2CO, D, 24 h (92%); (b) H2O2, AcOH, D, 1 min (61%). 1,1,1-Trifluoro-4-phenylpent-3-en-2-one with three substitu- ents at the double bond reacts with ammonium hydrogen sulfide to yield b-sulfanyl ketone 142.125 Ph Ph NH4SH, HCl, EtOH COCF3 Me 20 8C, 2 h Me COCF3 SH 142 (70%) Aryl-substituted trifluoromethyl enones were found to react with 2-sulfanylbenzaldehyde at room temperature in ethanol in the presence of a basic catalyst (triethylamine).This gives a mixture of two compounds, thiochromanes 143 and 2H-thiochro- menes 144, resulting from dehydration of 143. When the reaction mixture is refluxed for 1 h, the compounds 144 are formed in good yields. The intermediate thiochromane 143 (Ar=Ph) can be isolated in 65% yield only when the starting enone is b-trifluoro- acetylstyrene 20.124 V G Nenaidenko, A V Sanin, E S Balenkova Ar S COCF3 2-HSC6H4CHO D, 1 h Et3N, EtOH COCF3 Ar 143 OH Ar S COCF3 144 (64% ± 86%) The reactions of trifluoromethyl enones with malonodinitrile and cyanoacetamide in protic (ethanol, propan-2-ol) and aprotic (benzene) solvents in the presence of various basic catalysts (triethylamine, pyrrolidine, calcined potassium fluoride) have been studied.126 The reaction route depends on the reaction conditions and on the structure of the initial ketone.126 On refluxing in benzene with malonodinitrile in the presence of pyrrolidine as a catalyst, b-aryl-substituted enones are converted into pyrans 145.126 R CF3 R CH2(CN)2, HN COCF3 Ph O PhH, D, 4 h NC Ph 145 NH2 R=H, Me.Under the same conditions, cyclobutylidenemethyl trifluoro- methyl ketone is converted into cyclobutylidenemalonodinitrile, while adamantylidenemethyl trifluoromethyl ketone reacts in the presence of a stronger base (calcined potassium fluoride) to give adamantylidenemalonodinitrile. Apparently, in the case of ketones with an exocyclic double bond, the initial Michael addition of malononitrile is not followed by cyclisation to give pyran derivatives; instead, elimination of trifluoroacetone enolate occurs.The researchers cited suggest 126 that otherwise, cyclisa- tion products would contain a spiro-fused fragment and, as a consequence, the ring would be more strained. CH2(CN)2, HN COCF3 PhH, D, 4 h CN NC CN 7 7 CN 7CH2COCF3 COCF3 (83%) CN COCF3 CH2(CN)2, KF CN PriOH, D, 15 h (47%) The reaction of malonodinitrile with two equivalents of b-trifluoroacetylstyrene 20 in the presence of potassium fluoride in isopropanol affords substituted cyclohexanol 146, the product of addition of two molecules of the ketone to malonodinitrile and the subsequent intramolecular aldol condensation. The com- pound 146 is formed in a good yield as a single diastereoisomer of the eight possible isomers; the bulky substituents (Ph, CF3, COCF3) occupy energetically more favourable equatorial posi- tions.126 CN NC Ph Ph CH2(CN)2, KF 2 Ph COCF3 PriOH, D, 10 h 20 COCF3 CF3 HO146 (77%)Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis The reactions of trifluoromethyl enones with malonodinitrile in the presence of ammonium acetate occur ambiguously.Only in the case of the enone 20, was pyridine derivative 147, resulting from oxidation of the corresponding dihydropyridine 148 with atmospheric oxygen, isolated in a low yield. Ph CF3 Ph CF3 [O] CH2(CN)2, NH4OAc N 20 NH NC EtOH, D, 72 h NC NH2 147 (28%) NH2 148 1,1-Dicyanobuta-1,3-dienes 149 126 were synthesised by Knoe- venagel condensation of trifluoromethyl enones with malonodini- trile in the presence of the complex TiCl4 .2Py in dichloromethane. R1 CH2(CN)2, TiCl4 .2Py CH2Cl2, 20 8C, 72 h R2 COCF3 Ph CF3 R1 CN R2 1. H2SO4, 20 8C, 6 h 2. KHCO3, H2O R1=Ph, R2=Me CN F3C CN 149 (37% ± 61%) NH2 150 (66%) R1=Ph; R2=H, Me; R1±R2=(CH2)3, . An attempt to carry out cyclisation of the compounds 149 in an acid medium was undertaken. When 1,1-dicyano-4-phenyl-2- trifluoromethylpenta-1,3-diene was treated with concentrated sulfuric acid, substituted benzonitrile 150 was obtained; in other cases, such cyclisation did not occur. It was found that on refluxing of benzylidene- and cyclo- butylidene-methyl trifluoromethyl ketones with cyanoacetamide in isopropanol in the presence of calcined KF, a stereospecific reaction occurs, giving piperidines 151a,b in high yields.126 R1 KF +NCCH2CONH2 PriOH, D, 6 h R2 COCF3 CF3 R1 OH R2 NH NC 151a,b O R1=Ph, R2=H(a); R1±R2=(CH2)3 (b).Dehydration of the piperidines 151a and 151b gives dihydro- pyridines 152a,b (a mixture of diastereoisomers in *2 : 1 ratio) and spiro compound 153, respectively. CF3 Ph Ph CF3 a 151a NH NH + NC 7H2O NC O 152b O 152a CF3 a 151b NH 7H2O NC O 153 (87%) (a) PhMe, TsOH, D, 48 h. Imino-derivatives of unsaturated trifluoromethyl-containing ketones 154 cyclise in the presence of palladium supported on carbon to give, depending on the reaction conditions, either 2-trifluoromethylquinoline derivative 155 or 5-trifluoromethyl- pyrrolidone derivatives 156.80 455 N CF3 Pd/C 200 8C CO2Me F3C CO2Me 155 (100%) O NC6H4Me 154 Pd/C, H2 NC6H4Me ButOK CF3 156 (91%) The reactions of trifluoromethyl enones with o-phenylenedi- amine afford 2,3-dihydro-1,5-benzodiazepines 157.127 The com- pounds 157 undergo spontaneous dehydration, apparently, due to the fact that this enables the energetically favourable conjugation of the resulting C=N bond with the benzene ring.R1 HN R1 COCF3 R2 o-(NH2)2C6H4 EtOH, D, 6±9 h R2 N F3C157 (75% ± 86%) R1=H, R2=Ph; R1±R2=(CH2)3, . g,d-Unsaturated trifluoromethyl-containing ketones 160 and 161 have been prepared by the reactions of b-trifluoromethylstyr- ene 20 with alk-1-enyl- (158) 128 and alk-1-ynyl-diisopropoxybor- anes (159) 62 in the presence of boron trifluoride etherate.R2 Et2O. BF3, CH2Cl2 Ph B(OPri)2 COCF3 + R1 20 ± 40 8C, 12 ± 144 h 20 158 R3 Ph R2 COCF3 R1 R3 160 (82% ± 99%) R1=Bun, Me; R2=H, Br; R3=H, Me. RC CCHPh Et2O.BF3, CH2Cl2 20+RC CB(OPri)2 40 8C, 2±12 h CH2COCF3 161 (94% ± 98%) 159 R=Bun, Ph. a-Alkoxycarbonyl-substituted trifluoromethyl a,b-enones, prepared by condensation of ethyl trifluoroacetoacetate with aldehydes, react with enamino esters 162 to afford fairly stable hydroxypyridines 163, whereas their non-fluorinated analogues are converted directly into 1,4-dihydropyridines.The hydroxy- pyridines 163 were dehydrated in the presence of a number of catalysts; the best yields were attained when the complex POCl3 . Py adsorbed on silica gel was used.129 CO2R2 Me R1 COCF3 (ClCH2)2 + D, 2±3 h H2N CO2R2 162 R1 R1 CO2R3 CO2R3 R2O2C R2O2C POCl3 . Py HO D, 3±6 h Me Me F3C F3C HN NH163 164 (76% ± 91%) The reaction of the enone 20 with b-aminocrotononitrile results in the formation of 6-trifluoromethylpyridine 165 in a low456 yield; oxidation of this product leads to the corresponding aromatic derivative 166.130 CN COCF3 + Ph Me H2N 20 Ph Ph CN CN Me F3C Me N F3C NH 166 165 The reactions of 1,1,1-trifluoro-4-methylpent-3-en-2-one with isocyanides occur at room temperature without catalysts to give stable 1,4-cycloaddition products, namely, substituted dihydro- furans 167.131 Me MeMe C NR 20 8C, 14 days Me CF3CO NR F3C O 167 (90% ± 92%) R=But, cyclo-C6H11. Treatment of perfluoroalkyl a,b-enones with reagents that are widely used for asymmetric reduction of non-fluorinated ketones (e.g., Binal-H, Darvon-red) 132 results in racemic alcohols.133 The microbial reduction of perfluoroalkyl enones by baker's yeast was studied; however, the corresponding saturated ketones rather than alcohols were formed as the major reaction products. The alcohols were formed with high enantioselectivities (ee 77%± 85%) but in very low yields.baker's yeast R R R + RF RFCO RFCO 35 ± 37 8C, 240 h (42% ± 77%) OH (3% ± 9%) RF=CF3, C2F5, n-C4F9; R=Me, Ph.The reaction of g-hydroxy enone 168, synthesised by oxida- tion of the corresponding protected allyl alcohol by the Dess ± Martin reagent, with benzenethiol affords a tetrahydrofuran derivative � hemiacetal 169. On refluxing in benzene in the presence of sulfuric acid with azeotropic distillation of water, the compound 169 eliminates water and benzenethiol being thus converted into the corresponding furan 170.84 CF3CO (CH2)2Ph PhSH, KOH, MeCN 20 8C, 2 h 168 OH F3C O O (CH2)2Ph (CH2)2Ph F3C p-TsOH HO PhH, D, 1 h 170 (46%) 169 SPh IV. Conclusion The data presented in this review can be summarised by stating that development of facile methods for the synthesis of trifluoro- methyl a,b-enones remains a topical task.Trifluoromethyl enones containing no RO, RS or R2N groups are not very easy to synthesise; nevertheless, they are highly reactive compounds possessing great synthetic potential. A remarkable feature of trifluoromethyl enones is their high reactivity towards nucleophiles combined with regio- and stereo- selectivity of the corresponding reactions. Similar reactions of their non-fluorinated analogues with nucleophiles occur non- selectively giving products in low yields or do not occur at all. Another significant feature that distinguishes trifluoromethyl a,b- enones from non-fluorinated enones is their ability to form V G Nenaidenko, A V Sanin, E S Balenkova compounds incorporating a stable hemiacetal or hemiaminal fragment, CF3C(OH)O or CF3C(OH)NH.Thus, a,b-unsaturated trifluoromethyl-containing ketones are fairly convenient `building blocks' for the synthesis of diverse CF3-derivatives; further research into their properties and the ways of using them in organic synthesis appears quite promising. The review was financially supported by the Russian Founda- tion for Basic Research (Project No. 97-03-38959a). References 1. N Isikava (Ed.) Soedineniya Ftora. Sintez i Primenenie (Fluorine Compounds. Synthesis and Application) (Translated into Russian; Moscow: Mir, 1990) 2. L M Yagupol'skii Aromaticheskie i Geterotsiklicheskie Soedineniya s Ftorsoderzhashchimi Zamestitelyami (Aromatic and Heterocyclic Compounds with Fluorine-Containing Substituents) (Kiev: Naukova Dumka, 1988) 3.M A McClinton, D A McClinton Tetrahedron 48 6555 (1992) 4. J T Welch Tetrahedron 43 3123 (1987) 5. J-P Be'gue', D Bonnet-Delpon Tetrahedron 47 3207 (1991) 6. K I Pashkevich, V I Filyakova, V G Ratner, O G Khomutov Bashkir. Khim. Zh. 3 93 (1996) 7. M Hojo, R Masuda, Y Kokuryo, H Shioda, S Matsuo Chem. Lett. 499 (1976) 8. M Hojo, R Masuda, S Sakaguchi, M Takagawa Synthesis 1016 (1986) 9. A Colla,M A P Martins, G Clar, S Krimmer, P Fischer Synthesis 483 (1991) 10. M Hojo, R Masuda, Y Kamitori Tetrahedron Lett. 1009 (1976) 11. X-S Mo, Y-Z Huang, Y-R Zhao J. Chem. Soc., Chem. Commun. 2769 (1994) 12. W Verboom, D N Reinhoudt,V S Harkema, G J van Hummel J. Org. Chem. 47 3339 (1982) 13. H Trabelsi, A Cambon Tetrahedron Lett.36 3145 (1995) 14. M Hojo, R Masuda, E Okada Tetrahedron Lett. 27 353 (1986) 15. M Hojo, R Masuda, Y Kamitori, E Okada J. Org. Chem. 56 1975 (1991) 16. T Moriguchi, T Endo, T Takata J. Org. Chem. 60 3523 (1995) 17. M Hojo, R Masuda, E Okada Synthesis 347 (1990) 18. M Hojo, R Masuda, E Okada Synthesis 1013 (1986) 19. M Hojo, R Masuda J. Org. Chem. 40 963 (1975) 20. A M Platoshkin, Yu A Cheburkov, I L Knunyants Izv. Akad. Nauk SSSR, Ser. Khim. 112 (1969) a 21. S L Schreiber Tetrahedron Lett. 21 1027 (1980) 22. J Le'vy,M Soufyane, C Mirand,M DoÈ e'de Maindreville, D Royer Tetrahedron Lett., 32 5081 (1991) 23. G A Olah, A Germain, H C Lin J. Am. Chem. Soc. 97 5481 (1975) 24. G A Olah, L Heiliger, G K Surya Prakash J.Am. Chem. Soc. 111 8020 (1989) 25. V G Nenajdenko, E S Balenkova Zh. Org. Khim. 28 600 (1992) b 26. V G Nenajdenko, I D Gridnev, E S Balenkova Tetrahedron 50 11 023 (1994) a 27. V G Nenajdenko, A V Sanin, E S Balenkova Zh. Org. Khim. 30 531 (1994) a 28. V G Nenajdenko, E S Balenkova Zh. Org. Khim. 29 687 (1993) a 29. V G Nenajdenko, I F Leshcheva, E S Balenkova Tetrahedron 50 775 (1994) 30. V G Nenajdenko, E S Balenkova Tetrahedron 50 12407 (1994) 31. N P Gambaryan, L A Simonyan, P V Petrovskii Izv. Akad. Nauk SSSR, Ser. Khim. 918 (1967) a 32. E A Avetisyan, L A Simonyan, N P Gambaryan Izv. Akad. Nauk SSSR, Ser. Khim., 2742 (1972) a 33. N P Gambaryan, E M Rokhlina, Yu V Zeifman Izv. Akad. Nauk SSSR, Ser. Khim. 1466 (1965) a 34.O Neunhoeffer, G Alsdorf, H Ulrich Chem. Ber. 92 252 (1959) 35. Eur. P. 298 478; Chem. Abstr. 111 39 005 (1989) 36. J D Park, R E Noble, J R Lacher J. Org. Chem. 23 1396 (1958) 37. H Keller, M Schlosser Tetrahedron 52 4637 (1996) 38. D Mead, R Loh, A E Asato, R S H Liu Tetrahedron Lett. 26 2873 (1985)Methods for the synthesis of a,b-unsaturated trifluoromethyl ketones and their use in organic synthesis 39. D Mead, A E Asato, M Denny, R S H Liu, Y Hanzawa, T Taguchi, A Yamada, N Kobayashi, A Hosoda, Y Kobayashi Tetrahedron Lett. 28 259 (1987) 40. A Gazit, Z Rappoport J. Chem. Soc., Perkin Trans. 1 2863 (1984) 41. K I Pashkevich, R R Latypov, V I Filyakova Izv. Akad. Nauk SSSR, Ser. Khim. 2576 (1986) a 42. I Katsuyama, K Funabiki, M Matsui, H Muramatsu, K Shibata Chem.Lett. 179 (1996) 43. T Ishihara, T Maekawa, T Ando Tetrahedron Lett. 24 4229 (1983) 44. W S Huang, C Y Yuan J. Chem. Soc., Perkin Trans. 1 741 (1995) 45. B Jiang Chem. Commun. 861 (1996) 46. P Margaretha, C SchroÈ der, S Wolff,W C Agosta J. Org. Chem. 48 1925 (1983) 47. Y Hanzawa, K-i Kawagoe, N Kobayashi, T Oshima, Y Kobayashi Tetrahedron Lett. 26 2877 (1985) 48. A Yu Zenova, V V Platonov,M V Proskurnina, N S Zefirov Vestn. Mosk. Univ., Ser. 2, Khim. 115 (1997) c 49. S I Shergina, I E Sokolov, A S Zanina Izv. Akad. Nauk SSSR, Ser. Khim. 158 (1992) a 50. R J Linderman,M S Lonikar J. Org. Chem. 53 6013 (1988) 51. R J Linderman,M S Lonikar Tetrahedron Lett. 28 5271 (1987) 52. R J Linderman, K S Kirollos Tetrahedron Lett.31 2689 (1990) 53. M V Proskurnina, A G Kaz'min, A Yu Zenova, S A Lermontov, A A Borisenko, N S Zefirov Zh. Org. Khim. 32 146 (1996) b 54. X-S Mo, Y-Z Huang Synlett 180 (1995) 55. V G Nenajdenko, A L Krasovsky,M V Lebedev, E S Balenkova Synlett 1349 (1997) 56. I I Gerus, M G Gorbunova, V P Kukhar J. Fluorine Chem. 195 (1994) 57. M G Gorbunova, I I Gerus, V P Kukhar J. Fluorine Chem. 25 (1993) 58. A V Sanin, V G Nenajdenko, E S Balenkova Zh. Org. Khi (1999) (in the press) b 59. S V Pazenok, I I Gerus, E A Chaika, L M Yagupol'skii Zh. Org. Khim. 25 379 (1989) b 60. A V Sanin, V G Nenajdenko, K I Smolko, D I Denisenko, E S Balenkova Synthesis 842 (1998) 61. A V Sanin, V G Nenajdenko, D I Denisenko, K I Smolko, E S Balenkova Zh.Org. Khim. (1999) (in the press) b 62. S Hara, N Kato, E Takada, A Suzuki Synlett 961 (1994) 63. E Takada, S Hara, A Suzuki Heteroatom. Chem. 483 (1992) 64. I I Gerus, M G Gorbunova, S I Vdovenko, Yu L Yagupol'skii, V P Kukhar' Zh. Org. Khim. 26 1877 (1990) 65. M Hojo, R Masuda, E Okada Tetrahedron Lett. 30 6173 (1989) 66. K I Pashkevich, V I Filyakova Izv. Akad. Nauk SSSR, Ser. Khim. 620 (1986) a 67. V I Filyakova, V G Ratner, N S Karpenko, K I Pashkevich Izv. Akad. Nauk, Ser. Khim. 2278 (1996) a 68. N S Karpenko, V I Filyakova, K I Pashkevich Zh. Org. Khim. 33 1807 (1997) b 69. K I Pashkevich, V I Filyakova Izv. Akad. Nauk SSSR, Ser. Khim. 623 (1984) a 70. K I Pashkevich, A Ya Aizikovich Izv. Akad. Nauk SSSR, Ser.Khim. 1908 (1982) a 71. K I Pashkevich, M B Bobrov, A Ya Aizikovich,M N Rudaya Izv. Akad. Nauk SSSR, Ser. Khim. 2125 (1986) a 72. E Bayer, H P MuÈ ller, R Sievers Anal. Chem. 43 2012 (1971) 73. R Bartnik, A Bensadat, D Cal, Z Cebulska, A Laurent, E Laurent, C Rizzon Tetrahedron Lett. 37 8751 (1996) 74. G Alvernhe, A Bensadat, A Ghobsi, A Laurent, E Laurent J. Fluorine Chem. 81 169 (1997) 75. V Ya Sosnovskikh,M Yu Mel'nikov, V I Kutsenko Izv. Akad. Nauk, Ser. Khim. 1553 (1997) a 76. S I Vdovenko, I I Gerus, M G Gorbunova J. Chem. Soc., Perkin Trans. 2 559 (1993) 77. M Tordeux, C Wakselman J. Fluorine Chem. 20 301 (1982) 78. A L Henne, P E Hinkamp J. Am. Chem. Soc. 76 5147 (1954) 79. H Uno,Y Matsushima, T Tasaka,H Suzuki Bull.Chem. Soc. Jpn. 63 293 (1990) 80. K Uneyama, H Watanabe Tetrahedron Lett. 32 1459 (1991) 81. N Ishikava, M G Koh, T Kitazume, S K Choi J. Fluorine Chem. 24 419 (1984) 82. T Kitazume, N Ishikava J. Am. Chem. Soc. 107 5186 (1985) 83. N J O'Reilly, M Maruta, N Ishikava Chem. Lett. 517 (1984) 457 84. R J Linderman, E A Jamois, S D Tennyson J. Org. Chem. 59 957 (1994) 85. R J Linderman, D M Graves Tetrahedron Lett. 28 4259 (1987) 86. R J Linderman, D M Graves J. Org. Chem. 54 661 (1989) 87. D B Dess, J C Martin J. Org. Chem. 48 4155 (1983) 88. V G Ratner, K I Pashkevich Izv. Akad. Nauk, Ser. Khim. 680 (1996) a 89. Y Hanzawa, A Yamada, Y Kobayashi Tetrahedron Lett. 26 2881 (1985) 90. M E F Braibante, G Clar,M A P Martins J. Heterocycl. Chem. 30 1159 (1993) 91. R J Linderman, K S Kirollos Tetrahedron Lett. 30 2049 (1989) 92. Y Kamitori,M Hojo, R Masuda J. Heterocycl. Chem. 30 389 (1993) 93. M A P Martins, A F C Flores, R Freitag, N Zanatta J. Heterocycl. Chem. 32 731 (1995) 94. L A Simonyan, E A Avetisyan, Z V Safonova, N P Gambaryan Izv. Akad. Nauk SSSR, Ser. Khim. 2061 (1977) a 95. I I Gerus, S I Vdovenko,M G Gorbunova, V P Kukhar' Khim. Geterotsikl. Soedin. 502 (1991) d 96. C C Madruga, E Clerici, M A P Martins, N Zanatta J. Heterocycl. Chem. 32 735 (1995) 97. M T Cocco, C Congiu, V Onnis J. Heterocycl. Chem. 32 543 (1995) 98. K G Nikishin, V P Kislyi, V N Nesterov, A M Shestopalov, Yu T Struchkov, V V Semenov Izv. Akad. Nauk, Ser. Khim. 482 (1998) a 99. A C S Reddy, P S Rao, R V Venkataratnam Tetrahedron Lett. 37 2845 (1996) 100. A C S Reddy, P S Rao, R V Venkataratnam Tetrahedron 53 5847 (1997) 101. A C S Reddy, B Narsaiah, R V Venkataratnam J. Fluorine Chem. 86 127 (1997) 102. M Hojo, R Masuda, E Okada Synthesis 215 (1989) 103. M Hojo, R Masuda, E Okada Synthesis 425 (1990) 104. M Hojo, R Masuda, E Okada Chem. Lett. 113 (1990) 105. A V Sanin, V G Nenajdenko, V S Kuz'min, E S Balenkova Khim. Geterotsikl. Soedin. 634 (1998) d 106. P Bravo, L Bruche, A Farina, I I Gerus, M T Kolytcheva, V P Kukhar, S V Meille, F Viani J. Chem. Soc., Perkin Trans. 1 1667 (1995) 107. M G Gorbunova, I I Gerus, S V Galushko, V P Kukhar Synthesis 207 (1991) 108. I L Baraznenok, V G Nenajdenko, E S Balenkova Tetrahedron 54 119 (1998) 109. I I Gerus, A V Kropachev,M G Gorbunova, A Ya Il'chenko, V P Kukhar' Ukr. Khim. Zh. 59 408 (1993) 110. M Schlosser, H Keller, S Sumida, J Yang Tetrahedron Lett. 38 8523 (1997) 111. I L Baraznenok, V G Nenajdenko, E S Balenkova Eur. J. Org. Chem. 937 (1999) 112. E Okada, R Masuda, M Hojo, R Inoue Synthesis 533 (1992) 113. R Bartnik, A Bensadat, D Cal, R Faure, N Khatimi, A Laurent, E Laurent, C Rizzon Bull. Soc. Chim. Fr. 134 725 (1997) 114. E Okada, R Masuda, M Hojo, R Yoshida Heterocycles 34 1435 (1992) 115. M Soufyane, C Mirand, J Le'vy Tetrahedron Lett. 34 7737 (1993) 116. V G Nenajdenko, A V Sanin, E S Balenkova Zh. Org. Khim. 31 786 (1995) b 117. V G Nenajdenko, A V Sanin, O L Tok, E S Balenkova Khim. Geterotsikl. Soedin. 395 (1999) d 118. R G Kostyanovskii, V F Rudchenko Dokl. Akad. Nauk SSSR 263 897 (1982) e 119. R G Kostyanovsky, V F Rudchenko, O A D'yachenko, I I Chervin, A B Zolotoi, L O Atovmyan Tetrahedron 35 213 (1979) 120. V G Nenajdenko, A V Sanin, V S Kuz'min, E S Balenkova Zh. Org. Khim. 32 1579 (1996) b 121. Ya V Burgart, A S Fokin, I T Bazyl', V I Saloutin Izv. Akad Nauk, Ser. Khim. 992 (1997) a 122. M Schlosser, D Michel Tetrahedron 52 99 (1996) 123. V G Nenajdenko, A V Sanin,M V Lebedev, E S Balenkova Zh. Org. Khim. 31 783 (1995) b 124. V G Nenajdenko, A V Sanin, A V Churakov, J A K Howard, E S Balenkova Khim. Geterotsikl. Soedin. (1999) (in the press) dV G Nenaidenko, A V Sanin, E S Balenkova 458 125. A V Sanin, V G Nenajdenko, V S Kuz'min, E S Balenkova J. Org. Chem. 61 1986 (1996) 126. A V Sanin, V G Nenajdenko, A L Krasovskii, A V Churakov, J A K Howard, E S Balenkova Zh. Org. Khim. 33 236 (1997) b 127. V G Nenajdenko, A V Sanin, E S Balenkova Khim. Geterotsikl. Soedin. 1429 (1994) d 128. E Takada, S Hara, A Suzuki Tetrahedron Lett. 34 7067 (1993) 129. I Katsuyama, K Funabiki, M Matsui, H Muramatsu, K Shibata Tetrahedron Lett. 37 4177 (1996) 130. I Katsuyama, S Ogava, Y Yamaguchi, K Funabiki,M Matsui, H Muramatsu, K Shibata Synthesis 1321 (1997) 131. E A Avetisyan, N P Gambaryan Izv. Akad. Nauk SSSR, Ser. Khim. 2559 (1973) a 132. H C Brown, W S Park, B T Cho, P V Ramachandran J. Org. Chem. 52 5406 (1987) 133. T Kitazume, N Ishikawa Chem. Lett. 587 (1984) a�Russ. Chem. Bull. (Engl. Transl.) b�Russ. J. Org. Chem. (Engl. Transl.) c�Moscow Univ. Bull. (Engl. Transl.) d�Chem. Heterocycl. Compd. (Engl. Transl.) e�Dokl. Chem. Technol., Dokl. Chem. (Engl
ISSN:0036-021X
出版商:RSC
年代:1999
数据来源: RSC
|
2. |
Bipyrroles, furyl- and thienylpyrroles |
|
Russian Chemical Reviews,
Volume 68,
Issue 6,
1999,
Page 459-482
S. E. Korostova (deceased),
Preview
|
|
摘要:
Russian Chemical Reviews 68 (6) 459 ± 482 (1999) Bipyrroles, furyl- and thienylpyrroles S E Korostova (deceased), A I Mikhaleva, B A Trofimov Contents I. Introduction II. Methods of synthesis III. Reactivity IV. Prospects for the use V. Conclusion Abstract. Data on the methods of preparation of bipyrroles and furyl-, thienyl- and selenienylpyrroles are summarised and described systematically. The reactivity and the prospects of application of these compounds are discussed. Particular attention is paid to a new convenient one-step method for the synthesis of hetaryl-substituted pyrroles starting from alkyl hetaryl ketoximes and acetylene. The bibliography includes 191 references. I. Introduction Derivatives of five-membered aromatic heterocycles (pyrrole, furan, thiophene and selenophene) are, as a rule, endowed with high biological activity and other useful properties and find ever- increasing uses in medicine, agriculture and technology.1± 11 Of five-membered aromatic heterocycles, pyrrole possesses the highest reactivity and most diverse chemical properties.It is not accidental that it is pyrrole that Nature has `chosen' for constructing the `vital pigments' chlorophylls and hemoglobin, as well as other vitally important supramolecular structures (various porphyrins, chlorins, prodigiosins, vitamin B12, etc.). For this reason, amongst all five-membered aromatic heterocycles the molecules containing the pyrrole nucleus attract the greatest attention of researchers and are studied in most detail.The chemistry of bipyrroles as well as of pyrroles with furan, thiophene and selenophene substituents is a very recent field in the chemistry of heterocyclic compounds. Until recently, no suffi- ciently efficient and general methods of their preparation existed; therefore, these compounds remained attractive but hard to access and little studied subjects. The development of this field began in the 1970s, when new methods for pyrrole synthesis appeared or the existing procedures were modified to allow for linking five- membered aromatic heterocycles to the pyrrole ring.Avery simple and versatile method based on the reaction of ketoximes with acetylene (the Trofimov reaction) has been developed, which allows the synthesis not only of hetarylpyrroles non-substituted at the nitrogen atom but also their 1-vinyl derivatives from accessible alkyl hetaryl ketoximes.This also stimulated studies of S E Korostova, A I Mikhaleva, B A Trofimov Irkutsk Institute of Chemistry, Siberian Branch of the Russian Academy of Sciences, ul. Favorskogo 1, 664033 Irkutsk, Russian Federation. Fax (7-395) 239 60 46. Tel. (7-395) 246 64 09 (A I Mikhaleva), (7-395) 246 14 11. E-mail: but@acet.irkutsk.su (B A Trofimov) Received 15 April 1998 Uspekhi Khimii 68 (6) 506 ± 531 (1999); translated by V D Gorokhov #1999 Russian Academy of Sciences and Turpion Ltd UDC 547.741 459 459 471 479 479 reactivity, structure, biological activity and of other useful properties of hetarylpyrroles.Some aspects of the chemistry of pyrroles with furyl, thienyl and selenienyl substituents have been presented in mono- graphs,4, 12 reviews,5, 6, 13 ± 21 the US physical and technical encyc- lopedia (chapter `Acetylene'),22 and Ph.D. and D.Sc. theses.23 ± 27 Nevertheless, generalising papers that summarise the progress in the chemistry of hetaryl-substituted pyrroles are missing. We have attempted to fill this gap by generalising and systematising the currently available literature on the methods of synthesis, reac- tivity, biological and other properties of pyrroles linked to five- membered aromatic heterocycles. II. Methods of synthesis Conventional methods for the synthesis of pyrroles�the Knorr, Hantzch, Piloty, Yuriev (except for the Paal ± Knorr synthesis) reactions � have not found wide use for the preparation of pyrroles bound to other five-membered heterocycles.This goal can be reached most efficiently using the method based on the reaction of alkyl hetaryl ketoximes with acetylene (the Trofimov reaction), which is carried out in one step using accessible initial compounds. New methods have been developed for the preparation of hetarylpyrroles based on activated acetylenes, e.g., by conden- sation of hetarylacetylenes with trimethylsilyl cyanide, as well as those with the use of isocyanides, other cyano compounds, azides, etc. 1. Synthesis from ketoximes and acetylene (the Trofimov reaction) In the early 1970s, heterocyclisation of ketoximes with acetylene in a superbasic system (strong base ± polar aprotic solvent) was discovered, which yielded in one step 1H- and 1-vinylpyrroles (the Trofimov reaction).4, 5, 14 ± 20, 23, 24, 26 R2 R1 R2 B, DMSO +HC CH R1 N 70 ± 140 8C RN3 OH R1=Alk, Ar, Het; R2=H, Alk, Ar; R17R2=(CH2)n; n=3 ± 5; B=MOH, MOR, MON= ; Mis alkali metal; R3=H, CH=CH2 .Later it was established that this reaction is of a general character and embraces virtually all ketoximes that have at least460 one CH3 or CH2 group in the a-position relative to the oxime function and contain no groups unstable in the presence of strong bases. This finding has led to successful syntheses of pyrroles and 1-vinylpyrroles with alkyl, alkenyl, cyclopropyl, aryl and hetaryl substituents in the ring; condensed pyrrole systems of the type of cycloheptano[b]pyrrole, dihydrobenzo[g]indole and 5-azaindoles including stable free radicals with the NO.group at position 5, etc.13 ± 24, 26, 28 ± 33 Synthesis of hetarylpyrroles with furyl and thienyl substitu- ents by this method was first described by Trofimov et al.34 ± 36 2-(2-Furyl)-3-methyl-1-vinyl- and 3-ethyl-2-(2-furyl)-1-vinylpyr- roles were synthesised from the corresponding ketoximes and acetylene in yields up to 85%.35 An analogous approach was used to prepare 2-(2-thienyl)- and 2-(2-thienyl)-1-vinylpyrroles in 60% and 50% yields, respectively.36 The reactions were carried out in an autoclave under an initial pressure of acetylene from 6 to 16 atm.R1 R1 KOH, DMSO +HC CH X X NOH NR2 R1=H, Me, Et; R2=H, CH=CH2; X = O, S. Subsequently, this method has successfully been used for the preparation of 3-alkyl-2-(2-furyl)pyrroles 1a ± c and their 1-vinyl derivatives 2a ± f.37 R R O O N HN 1a ± c 2a ± f CH CH2 Yield (%) Compound Yield (%) Compound R R 2c 2d 2e 2f 85 77 81 76 Et Prn Pri Bun 1a 1b 1c 2a 2b 36 23 39 80 92 HMe Et HMe Under conditions close to those established 4 earlier for aliphatic ketoximes [100 8C, ketoxime :KOH=5 : 3 (w/w), excess of acetylene, pressure 10 ± 16 atm.], the reaction of 2-furyl methyl ketoxime with acetylene produces a mixture consisting of 1H- (1a) and 1-vinyl-2-(2-furyl)pyrroles (2a) in the ratio of ca 1 : 4 and an overall yield of 71% (16% and 55%, respectively).37 Under analogous conditions 2-(2-furyl)-3-methyl-1-vinyl- pyrrole (2b, 57% yield) and the corresponding pyrrole 1b (8% yield) are formed from ethyl 2-furyl ketoxime.At a higher temperature (130 8C) and an equimolar ketoxime :KOH ratio this reaction proceeds selectively with the formation of vinyl- pyrrole alone (2b, 92% yield). The yield of furylpyrroles 1a and 2a upon the reaction of 2-furyl methyl ketoxime and acetylene depends on the reaction temperature (Table 1). At temperatures below 100 8C, the rate of ketoxime cyclisation into pyrrole 1a and vinylation of the latter to vinylpyrrole 2a is too low (the ketoxime conversion does not exceed 40%). At 130 ± 150 8C, one observes resinification that decreases the yield of 1-vinylpyrrole 2a.In addition, at a high temperature the process of deoximation of the initial oxime to 2-acetylfuran becomes noticeable. The optimal conditions for the synthesis of 3-alkyl-2-(2-furyl)-1-vinylpyrroles are as follows: 110 ± 130 8C, 3 h, excess of acetylene (initial pressure 10 ± 14 atm) and equimolar ratio ketoxime : KOH). It should be pointed out that lithium hydroxide, which catalyses selectively the stage of heterocyclisation in an analogous reaction with methyl phenyl and benzyl phenyl ketoximes,38, 39 is S E Korostova, A I Mikhaleva, B A Trofimov Table 1. Dependence he yield of 2-(2-furyl)- (1a) and 2-(2-furyl)-1- vinylpyrroles (2a) on the reaction temperature (molar ratio ketox- ime :KOH=1 : 1, C2H2 pressure 10 ± 14 atm, 3 h).37 Composition of mixture (%) a Tem- Ketoxime conversion (%) /8C Yield perature of mixed products (%) com- pound 2a com- pound 1a ket- oxime 60 20 traces " 40 80 *100 *100 *100 *100 14 56 71 85 *100 *100 26 24 39 15 undetected traces " 70 90 110 120 130 150 " 85 82 70 65 46 21 a According to 1H NMR data.not selective in this case: even at 100 8C a mixture of 1H- (1a) and 1-vinylpyrroles (2a) is formed in the ratio 15 : 1, while the total yield of pyrroles is lower than with KOH (*50% instead of 72%, GLC). The corresponding pyrrole is also readily formed from methyl (5-methyl-2-furyl) ketoxime and acetylene.40 Me KOH, DMSO +HC CH Me O NOH Me O NR R=H (53%), CH=CH2 (62%).Under atmospheric pressure (100 8C, 5 h, equimolar ratio ketoxime :KOH) the yield of 2-(5-methyl-2-furyl)pyrrole is 53% and the conversion of ketoxime is 80%. Under these conditions 2-(5-methyl-2-furyl)-1-vinylpyrrole is formed in a rather low yield (15%). Under acetylene pressure of 12 atm (90 8C, 3 h, molar ratio ketoxime :KOH=1 : 2), the same reaction produces vinyl- pyrrole in 62% yield, though being accompanied by resinification. According to GLC data, no more than *3% of the correspond- ing 1H-pyrrole remains unconsumed under these conditions.40 It is known that even at 50 ± 60 8C and atmospheric pressure alkyl phenyl ketoximes react readily with acetylene to form the corresponding pyrroles in yields up to 83%.41 The reaction of ethyl 2-furyl ketoxime with acetylene under a pressure of 10 atm and 100 8C produces 2-(2-furyl)-3-methyl- and 2-(2-furyl)-3- methyl-1-vinyl-pyrroles in yields of 8% and 57%, respectively.37 Quite unexpectedly, in the condensation of the oxime 3 with acetylene (10 atm, 100 8C) the conversion of oxime 3 was no more than 65%, while the yield of pyrrole 4a was as low as 30%.The reaction is accompanied by considerable resinification. HC CH Ph KOH, DMSO O 3 NOH + Ph O O 5 O Ph RN 4a,b R = H (a), CH=CH2 (b). A decrease in the reaction temperature to 75 ± 80 8C did not lead to higher yield of compound 4a, because even at thisBipyrroles, furyl- and thienylpyrroles temperature ketoxime 3 seems to undergo deoximation to the ketone 5.KOH, DMSO H2O 3 Ph O HNOK H2O 5 Ph Ph 7NH3 O O NH NH2 Vinylpyrrole 4b was detected in the reaction mixture only by GLC.40 Low yield of pyrroles 4a,b was explained by the decrease in the reactivity of ketoxime 3 due to the presence of bulky phenyl and furyl groups preventing the approach of oximate anion to the triple bond. An intermediate formed from ketoxime 3 isomerises easily into a substituted vinylfuran 6, which cannot undergo the [3,3]-sigmatropic shift and, consequently, the pyrrole formation. Ph Ph O O NH NH 6 O O These two side reactions seem to account for resinification and formation of by-products, which prevent isolation of pure pyrroles 4a,b.Additional difficulties appear to be associated with oligomerisation of acetylene, which occurs in the NH3 (amine) ±KOH±DMSO systems.42 Condensation of alkyl 2-thienyl ketoximes with acetylene catalysed by superbases (MOH±DMSO) produced 1H- (7a ± f) (50% ± 60%) and 3-alkyl-2-(2-thienyl)-1-vinylpyrroles (8a ± f) (in 60%± 70% yields).43, 44 The reaction proceeds smoothly at 100 ± 130 8C and may be carried out under both elevated (10 ± 16 atm) and atmospheric pressures of acetylene. The effect of reaction conditions (temperature, acetylene concentration, addition of water, the nature of cation in the alkali metal hydroxide) on the ratio of pyrroles to their vinyl derivatives formed in the same synthesis was studied in the example of methyl 2-thienyl ketoxime.R R HC CH HC CH KOH, DMSO MOH, DMSO S S NOH NH 7a ± f R N S 8a ± f CH CH2 R = H (a), Me (b), Et (c), Prn (d), Bun (e), n-C6H13 (f ); M=Li, K. Synthesis of 2-(2-thienyl)-1-vinylpyrrole (8a) is best carried out in an autoclave, at 120 8C for 3 h under elevated acetylene pressure, 17 mass.%± 30 mass.% of KOH with respect to the ketoxime.26, 37 The increase in the reaction time to 5 h and more as well as in the temperature to 140 ± 150 8C lead to more pro- nounced resinification and a lower product yield. 2-(2- Thienyl)pyrrole (7a) free of its vinyl derivative is produced in >60% yield with the use of a superbasic pair LiOH ±DMSO (100 8C, 3 h, 10 ± 16 atm) which, as in the case of condensation of methyl phenyl ketoxime with acetylene,38, 39 catalyses the stage of pyrrole ring formation and is virtually inactive in the stage of vinylation. The amount of lithium hydroxide added is *18 mass.% with respect to the ketoxime.The same pyrrole was obtained in 47% yield under atmospheric pressure (130 ± 140 8C, 7 h, 30% LiOH). The addition of water to DMSO (10% ± 15%) and the use of KOH (equimolar ratio ofKOH to the 461 oxime) inhibit the vinylation process and make possible the preparation of 2-thienylpyrrole (7a) in 53% yield. Heating of propyl 2-thienyl ketoxime with acetylene (100 ± 120 8C, 0.8 mol KOH, 11% H2O, 7 h) leads to the formation of 3-ethyl-2-(2- thienyl)pyrrole (7c) also in 53% yield.The length of the carbon chain of the alkyl radical in the original alkyl 2-thienyl ketoxime does not exert any noticeable effect on the composition and yield of the final products. The starting ketoximes are a mixture of E- and Z-isomers. It was established that only the anti-isomer participates in the reaction with acetylene, while the syn-isomer is recovered unal- tered after the reaction. It could be supposed that the oximes with a large content of anti-isomer would give thienylpyrroles in higher yields. This is not the case because of easy syn ± anti-isomerisation. This was demonstrated for methyl thienyl ketoximes, where pure syn- and anti-isomers isolated by preparative TLC were used in the reaction with acetylene.26 Me Me S S N N Z E HO OH Like in the case of alkyl methyl 45 and alkyl benzyl ketox- imes,46 the reaction of (E)-methyl 2-(5-methyl-2-thienyl)ethyl ketoxime with acetylene involves the methylene group in the anti-position relative to the hydroxy group.40 At 95 8C and atmospheric pressure, the ratio of structurally isomeric pyrroles 9a and 10a is 33 : 1 [yields 46% and 1.4%, respectively (GLC)], whereas vinylpyrrole 10b was detected in the reaction products only in trace amounts.Despite a rather high total yield of pyrroles 9 and 10 (61%), their isolation from the reaction mixture proved to be problematic due to the presence of a number of unidentified by-products. More drastic conditions (120 8C) favour the increased proportion of the reaction products formed with the participation of the methyl group and also the vinylation of pyrroles 9a and 10a; the ratios of pyrroles 9a : 10a and 9b : 10b become equal to 24 : 1 and 5 : 1, respectively.40 HC CH Me Me KOH, DMSO S NOH Me + Me S S RN 10a,b Me 9a,b RN R = H (a), CH=CH2 (b).2-(2-Thienyl)- and 2-(3-thienyl)pyrroles 11 were synthesised recently using a modification of the Trofimov reaction consisting in the addition of methyl 2-thienyl and methyl 3-thienyl ketoximes 12 to methyl propiolate and dimethyl acetylenedicarboxylate.47 Me Me R RC CCO2Me NOH NO CO2Me S S 12 13 (89%) CO2Me R NH S 11 R=H, CO2Me. The reaction is conducted in two stages: intermediate O-vinyloximes 13 are transformed into thienylpyrroles 11 on boiling in xylene.462 Selenienylpyrroles have remained virtually unstudied until now since no efficient methods for their synthesis had been developed.Mikhaleva et al.48 have shown the possibility of preparing the corresponding pyrroles from methyl 2-selenienyl ketoxime (14) by its reaction with acetylene at 95 ± 97 8C: 2-(2- selenienyl)pyrrole (15a) and -2-(2-selenienyl)-1-vinylpyrrole (15b) were produced in yields of 10% and 2%, respectively. Me KOH, DMSO +HC CH 95 ± 97 8C, 5 h Se Se OH N 14 N 15a,b R R = H (a), CH=CH2 (b). Ketoxime 14 was used in this reaction as a mixture of E- and Z-isomers (1 : 1). Under these conditions, the Z-isomer reacted predominantly (or exclusively), which correlates with the results of the analogous reaction with methyl 2-thienyl ketoxime 43 and 2-benzoimidazolyl methyl ketoxime.49 The low yield of pyrrole 15a is explained by its instability in the KOH±DMSO system.48 Further temperature increase to 110 ± 120 8C results only in a lower yield and resinification. Apparently, the C7Se bond is less stable and more sensitive to the attack by a strong base compared with the C7O and C7S bonds (as mentioned above, under the same conditions the total yield of furyl and thienyl analogues reaches 80%± 90%).a. 1,2-Dichloroethane as an acetylene equivalent The reaction of alkyl hetaryl ketoximes 16 with 1,2-dichloro- ethane (DCE), which is a synthetic equivalent of acetylene under the conditions of the Trofimov reaction, is a version of the synthesis of furyl- and thienylpyrroles.50 The ratio of the reaction products [hetarylpyrroles 1 and 7, 1-vinylhetarylpyrroles 2 and 8, intermediate O-2-chloroethyl (17) and O-vinyl (18) hetaryl ketoximes] and by-products, viz.1,2- bis(hetarylalkylideneiminoxy)ethanes (19), depends on many factors: reaction temperature and time, the nature of the hydrox- ide cation, its concentration and the presence of water in the catalytic system (Scheme 1). An advantage of this version consists in the substitution of acetylene by a more accessible, safe and handling-convenient DCE. The synthesis can be performed in a trivial apparatus under atmospheric pressure and in the temperature range 100 ± 135 8C.50 The highest yields of furyl- and thienylpyrroles were attained at 120 8C (6 ± 9 h, molar ratio ketox- ime :NaOH:DCE=1 : 2 : 7).When KOH was used as a base, better results were obtained upon addition of 5%± 10% of water (120 8C, 7 h, molar ratio ketoxime :KOH:DCE=1 : 2 : 10); however, in this version the yields of 2-(2-furyl)- and 2-(2- thienyl)-1-vinylpyrroles did not exceed 25%. In the presence of lithium hydroxide, the yields of pyrroles were lower (18%). In an NaOH±DMSO+10% H2O system, the reaction stops at the stage of intermediates 18, viz., O-methyl 2-thienyl O-vinylketox- CH2R RCH2 16 X NO(CH2)2ON CH2R Cl(CH2)2Cl 19 MOH, DMSO X NOH CH2R 16 X NO(CH2)2Cl 17 R X 20 X=O, S; R=H, Me, Et, Bun;M=Li, Na, K. S E Korostova, A I Mikhaleva, B A Trofimov imes (R=H, yield 43%) or 2-furyl methyl O-vinylketoximes (R=H, yield 6%).50 The detection of O-vinylketoximes 18 and methyl 2-thienyl O-(2-chloroethyl)ketoximes 17 (X= S, R=H, yield 24%) among the main reaction products has made it possible to suggest 50 that in the case of DCE, two alternative pathways of the formation of pyrroles and 1-vinylpyrroles are realised.According to the first pathway,4 in a strongly basic medium DCE serves as the source of acetylene, and the reaction proceeds as usual (see Section II.1). The second pathway consists in the nucleophilic substitution of one chlorine atom in DCE by the oximate anion with the formation of O-(2-chloroethyl)ketoximes 17 with subsequent elimination of HCl to affordO-vinylketoximes (18). Intermediates 18 undergo cyclisation into the corresponding pyrroles, which react with DCE to form, apparently, 1-(2- chloroethyl)pyrroles 20.These are further transformed into N-vinylpyrroles upon elimination of HCl.50 The possibility of formation of 1-(2-chloroethyl)pyrroles 20 under these conditions is in accord with the data reported in Ref. 51. b. Vinyl chloride as an acetylene equivalent A simple and industrially acceptable method of synthesis of furyl- and thienylpyrroles is based on the reaction of the corresponding ketoximes with vinyl chloride in a KOH±DMSO suspension.26 The condensation is carried out at 120 8C with a sixfold excess of KOH relative to the ketoxime. This results in the formation of a mixture of pyrroles and their 1-vinyl derivatives.In this modifica- tion of the Trofimov reaction, vinyl chloride serves as the acetylene equivalent. R KOH, DMSO 1b, 7d+2b, 8d + Cl 120 8C, 3 h X NOH Yield (%) R X Compound 1b 2b 7d 8d 40 36 46 26 Me Me Prn Prn OOSS c. Intermediates and the reaction mechanism Systematic studies on the synthesis of pyrroles from hetaryl ketoximes and acetylene in the superbasic KOH±DMSO medium have allowed not only identification of the reaction intermediates, such as O-vinylketoximes or 3H-pyrroles, but also the development of methods for their preparation.43, 50, 52 ± 55 For instance, 2-furyl methyl (18a) and methyl 2-thienyl (18b) O-vinylketoximes were obtained in yields of 38% and 32%, respectively, by the reaction of 2-furyl methyl (16a) and methyl 2-thienyl ketoximes (16b) with acetylene in the KOH±DMSO system (initial pressure 12 atm, 3 h, 50 ± 60 8C).52 ± 54 Scheme 1 X R CH2R Cl(CH2)2Cl 7HCl X X NO NH 18 1, 7 R MOH, DMSO 7HCl N X N 2, 8 CH CH2 (CH2)2ClBipyrroles, furyl- and thienylpyrroles Me Me a or b X X NO NOH 18a,b 16a,b X = O (a), S (b); (a) HC:CH, MOH, DMSO, H2O; (b)Cl(CH2)2Cl, MOH, DMSO, H2O; M=K, Li.Upon addition of 4% of water to the KOH±DMSO system, the yield of O-vinylketoxime 18b increased to 44%.54 Under similar conditions, the yield of the corresponding furyl derivative 18a did not exceed 13%.54 Under atmospheric pressure of acetylene, the synthesis of O-vinylketoximes is less efficient.1,2-Dichloroethane is also suitable for the preparation of ketox- imes 18a,b. The optimal conditions for the synthesis of compound 18b are the use of LiOH ±DMSO system and addition of 10% of water (100 C, 4 h, yield 43%).50 In the NaOH±DMSO system containing 10% of water, O-vinylketoxime 18b is not formed, whereas the yield of O-vinylketoxime 18a is as low as 6%.50 Ketoximes 18a,b are transformed upon thermolysis into the corresponding pyrroles as do alkyl aryl O-vinylketoximes,56 methyl (1-methyl-2-pyrrolyl) O-vinylketoxime 57 orO-vinylketox- imes formed upon heating of O-(2-iodoethyl)ketoximes with ButOK in ButOH.58 Thus, heating of 2-furyl methyl and methyl 2-thienyl O-vinylketoximes (100 8C) in the KOH±DMSO system led to 2-(2-furyl)- and 2-(2-thienyl)pyrroles in yields of 60% and 50%, respectively.56 2-(2-Thienyl)- and 2-(3-thienyl)-4-methoxy- carbonyl- and -4,5-di(methoxycarbonyl)pyrroles 13 were obtained by heating of the corresponding O-vinylketoximes 12 in xylene 47 (see Section II.1).The reaction of isopropyl 2-thienyl ketoxime (21) with acetylene afforded 3,3-dimethyl-2-(2-thienyl)-3H-pyrrole (23), the yield of which depended on the reaction conditions and in the best case (addition of Al2O3 to the catalytic system) reached 43% with respect to the consumed ketoxime.55 R Pri R Pri HC CH N NOH 21 O 22a Me Me Me R Me R HN NH O O 22b Me Me Me Me 7H2O R OH R N N 23 . R= S The formation of the pyrrole 23 from the oxime 21 indicates that the reaction pathway includes the stage of prototropic isomerisation of O-vinylketoxime 22a into O,N-divinylhydroxy- amine 22b.This scheme of the formation of pyrroles from oximes is also supported by the identification 59, 60 of other reaction intermedi- ates (hydroxypyrrolines) and their transformation into the corre- sponding pyrroles upon condensation of alkyl aryl ketoximes with acetylene.41 The fact that pyrroles cannot be synthesised from O-alkyl ketoximes under the conditions of the Trofimov reac- tion,24 as well as the predominant formation of b-phenylpyrroles in the case of phenylacetylene 61 support the proposed reaction scheme. 2. The Paal ± Knorr synthesis The Paal ± Knorr synthesis is widely used for the preparation of pyrrolyl-, furyl- and thienylpyrroles. As a rule, it is carried out under mild conditions and provides for high yields of the target products in the final stage.However, the preparation of the initial 1,4-diketones and/or amines with furyl and thienyl substituents remains a challenge: their synthesis includes several stages which do not always give high yields, which restricts substantially the applicability of this method. N,N0,N00-Tris(tert-butoxycarbonyl)-5,500-dibromo-2,20 : 50,200- terpyrrole (26) was obtained from 2-formylpyrrole 62 in several steps (Scheme 2). In the intermediate disulfone 24, two sulfonyl groups were substituted by tert-butoxycarbonyl protective groups. Simultaneously, the third pyrrole ring was N-protected; the totally protected terpyrrole 25 was then selectively brominated with N-bromosuccinimide (NBS) to pyrrole 26. H a C N O HN SO2Ph N O O SO2Ph HN N 24 SO2Ph CO2But N NCO2But Br NCO2But (a) NaH, PhSO2Cl, DMF, 20 8C; (b) SO2 , EtOH, D; (c) AcONH4 , Ac2O, EtCO2H, 140 8C; (d ) NaOH, MeOH, 20 8C; (e) (ButO2C)2 , ButOK, THF, D; ( f ) NBS, THF, 770 8C.In a Japanese patent 63 it was proposed to prepare 1-(2-thi- enyl)pyrroles 29 (yields up to 80%) by the reaction of 2-amino- thiophenes 27 with 1,4-diketones 28 on boiling for 20 min in benzene in the presence of p-toluenesulfonic acid. R5 R2 R1C6H4CO R4 R3 + H2N S 27 R1C6H4CO R4 R5 S N 29 Et R4 R3 R2 R1 Me Me Me Me Me Me Me Me Et Me Me Me HHMe 2-Cl 2-Cl HHH H b C O c NSO2Ph d, e NSO2Ph N 25 CO2But CO2But N 26TsOH Et O O 28 R2R3 R5 Me HHHMe 463 Scheme 2 f Br NCO2But464 Furylpyrroles (31a,b and 32a) were prepared from 1-(2- furyl)pentane-1,4-dione (30a) or 1-(2-furyl)-2,4-diphenylbutane- 1,4-dione (30b) and ammonia (NH4Cl, 180 8C, 20 h, yields of 40% and 50%, respectively) or 4-chlorophenylamine (AcOH, EtOH, 7 h, yield 85%) under atmospheric or elevated pressure.64 R1 R2 O O O 30a,b R1 NH3 O 31a,b R1 4-ClC6H4NH2 O 32a R1=H,R2=Me (a); R1=R2=Ph (b).If ammonia is replaced by ammonium carbonate and the reaction is carried out in ethanol in the presence of acetic acid, a twofold increase in the yield of pyrrole 31a is achieved.64 It should be noted that the reaction of diketone 30a with ammonia is accompanied by the formation of pyrrole 33 (yield 8%), while 2,4,5-tri(2-furyl)imidazole and 2,3,5,6-tetra(2-furyl)pyrazine (5% yield of each) are formed from diketone 30b.64 1-(5-Chloro-2-thienyl)pentane-1,4-dione (36) obtained from 5-chlorothiophene (34) and 5-methyl-5-hydroxytetrahydrofuran- 2-one (35) reacts with ammonium acetate in methanol at 25 8C for 3 h to give 2-(5-chloro-2-thienyl)-5-methylpyrrole 37 in 65% yield.65 Cl + O OH O 35 S 34 Cl S O O 36 Cl S 37 Pyrroles 39a ± d containing two thienyl substituents in the ring were obtained in high yields from the corresponding diketones 38a ± d and methylamine in boiling propanol containing methyl- amine hydrochloride.66, 67 R2 R1 Ph+MeNH2 O O 38a ± d R2 Compound 39 R1 abcd 2-thienyl 2-thienyl 3-thienyl 3-thienyl 2-thienyl 3-thienyl 2-thienyl 3-thienyl Synthesis of 2,5-bis(2-pyrrolyl)thiophenes 42 (yield 65%) was brought about 62, 68 by cyclisation of symmetrical 1,4-diketones 40a,b with Lawesson's reagent (41) upon boiling in toluene.+ Me MeC R2 NH NH O 33 (8%) R2 NC6H4Cl-4 Me AlCl3 NH4OAc Me MeOH Me NH R2 Ph R1 NMe 39a ± d Yield (%) 85 97 90 99 S E Korostova, A I Mikhaleva, B A Trofimov (CH2=CH)2SO2 CHO NaOAc, EtOH NR 41 O O NR NR40a,b RN NR S 42a,b S S C6H4OMe-4 P P .R=MeSO2 (a), PhSO2 (b); 41= S 4-MeOC6H4 S Diketones 40a,b are obtained from pyrrole-2-carboxalde- hydes in ethanol in the presence of NaOAc, divinyl sulfone and a thiazolium salt as the catalyst (yields 70% and 74%, respectively). 2,5-Bis[3,5-di(ethoxycarbonyl)-4-methyl-2-pyrrolyl]thiophene (44) was prepared in 76% yield from diketone 43 [synthesised by the reaction of 3,5-bis(ethoxycarbonyl)-4-methyl-2-formylpyrrole with divinyl sulfone in dioxane at 80 8C in the presence of triethylamine] and Lawesson's reagent 41.68 R R Me Me 41 R D, PhH R NH NH O O 43 R Me R Me R R NHHN S 44 R=EtO2C. The reaction of 2-methylsulfinyl-1-phenyl-5-(2-thienyl)pen- tane-1,5-dione (45) with ammonium acetate in acetic acid (D, 10 min) leading to 2-(2-thienyl)pyrrole (46) with a yield of 19% may be regarded as a variant of the Paal ± Knorr synthesis.69 O NH4OAc, AcOH (CH2)2 Ph 7H2O S 45 OSMe O Ph HO SMePh S S O 7MeSH, 7H2O 46 NH O NH2 An elegant and efficient synthesis of 1-(2-thienyl)pyrroles 51 ± 53 is based on the condensation of 2-aminothiophenes 47, 2-amino-4,5-dihydrocyclopenta[b]thiophene 48 or 2-amino- 4,5,6,7-tetrahydrobenzo[b]thiophenes 49 with 2,5-dimethoxy- tetrahydrofuran (50) in acetic acid.70, 71 R3 R2 R3 R2 + NH2 OMe MeO N R1 R1 O S S 50 51a ± e 47 Yield (%) R3 Compound 51 R2 R1 abcde 86 96 97 72 70 HCO2H CO2H CO2Et CO2Et Me Et Me HMe Me Me Me HMeBipyrroles, furyl- and thienylpyrroles R R 50 S N S NH2 52 48 R=CO2H (96%), CO2Et (71%). R R 50 N S S NH2 53 49 R=H (70%), CO2H (97%), CN (86%), CO2Et (59.5%). The oxidation products of aminoacylfuran 54 are supposed to contain, along with the major compounds 55a ± c, insignificant amounts of furylpyrroles 56a,b.72 COCH2NH(CH2)4CHCO2H O 54 NH2 R + COR N O O O 55a ± c (CH2)4CHCO2H 56a,b NH2 R=OH (a), NH(CH2)4CH(NH2)CO2H (b), CONH(CH2)4CH(NH2)CO2H (c).The reaction of cyclopropyl ketones 57 with amines leads to the formation of hetarylpyrroles 58 in high yields.73, 74 The presence of an electron-acceptor substituent X (OMe, OEt, Cl) in the cyclopropane ring facilitates its isomerisation leading to the formation of a carbonyl group equivalent. R2 O R3NH2 R1 X R2 R1 57 58 NR3 R1= 2-furyl, 2-thienyl; R2=H; R3=H, Alk, Ar; X=OMe, OEt, Cl.3. Syntheses involving activated acetylenes Oxazolium oxide 60 formed from N-methyl-N-thenoylamino acid 59 and acetic anhydride is transformed into 2-(2-thienyl)pyrrole 61 upon boiling in benzene with dimethyl acetylenedicarboxylate (yield 89%).75 7O O RC CR Ac2O Me N CO2H + Ph S S 59 Ph O NMe 60 R R Ph S NMe 61 R=CO2Me. A general method for the synthesis of thienylpyrroles 64 has been developed, which is based on 1,3-dipolar cycloaddition of 2-(2-thienyl)aziridines 62 to activated acetylenes.76, 77 The reac- tion proceeds on heating of the reagents in benzene with the formation of intermediate thienylpyrrolines 63, which are readily dehydrogenated under the action of tetrachloro-1,4-benzoqui- none to give pyrroles 64, as a rule, in moderate yields.465 COAr MeO2CC CR S N 62 C6H11-cyclo R R MeO2C MeO2C COAr COAr 7H2 N S N S C6H11-cyclo 64 63 C6H11-cyclo Ar R R Ar Yield of 64 (%) Yield of 64 (%) H 36 53 65 32 2-naphthyl CO2Me 4-MeC6H4 CO2Me 4-MeOC6H4 CO2Me 3-NO2C6H4 32 26 39 25 H Ph H 4-MeC6H4 H 4-MeOC6H4 H 3-NO2C6H4 Heating of 1,4-di(2-thienyl)butadiyne (65) with aniline at 140 ± 160 8C in the presence of cuprous chloride leads to the formation of 2,5-di(2-thienyl)-1-phenylpyrrole (66) in 68% yield.78 PhNH2 C CC C S S S S 65 66 Ph N This reaction has a general character.Asymmetrically sub- stituted butadiynes 67 react with amines, which allows prepara- tion of various 2-(2-thienyl)pyrroles 68.78 Ethanol, dioxane and DMF are used as solvents. R2NH2 C CC R1C R1 S S 67 68 NR2 R1=H, Me, Et, Ph; R2=H, Me, Et, Bun, Ph. A new method has been developed for the synthesis of 2-(2- pyrrolyl)-, 2-(2-furyl)- and 2-(2-thienyl)pyrroles 69a ± c from the corresponding amines 70, acetylenes 71 and carbon monoxide with the use of zirconium cyclopentadienyl complexes.79 Me Cl CH2NSiMe3 CH2NSiMe3 X X Cp2Zr (71) 778 8C, THF 70 Li MeZrCp2 HC CR NSiMe3 0±20 8C, THF X ZrCp2 .THF R R CO, THF Cp2Zr X X N HN 69a ± c 72 SiMe3 X=N, R=Prn (69a, 68%); X=O, R=Ph (69b, 49%); X=S, R=Prn (69c, 41%).The reaction with CO leading to hetarylpyrroles may be realised both with and without isolation of metallocycles 72. This reaction can be carried out in the presence of NH4Cl, under CO pressure of*100 atm, at*20 8C, for 24 h (method A) or in the absence of NH4Cl, at 80 8C, CO pressure of*6 atm and for 36 h (method B). The former method is more efficient, though the role of NH4Cl is still unclear.466 The following mechanism of this reaction has been pro- posed:79 R R Cp2Zr O CO 72 O Cp2Zr N Het Het Me3Si 73 74 R NSiMe3 R Cp2Zr Cp2Zr NH4Cl, H2O + O7 Me3SiO N N Het Het Me3Si 76 75 R H 69, H Het N H+77 R H+HO NH4Cl, H2O 74 69. 77 Het HN78 Insertion of CO into the C7Zr bond gives metallocycle 73.Migration of the nitrogen atom from zirconium to the acyl carbon (formally, reductive elimination) leads to the zirconium complex 74. The intermediate 74 can undergo ring opening with the formation of a zwitter-ion 75, which after migration of the trimethylsilyl group from nitrogen to zirconium gives compound 76. Its hydrolysis yields intermediate 77, which undergoes proto- tropic rearrangement into pyrrole 69. An alternative pathway is possible: hydrolysis of the complex 74 to the alcohol 78, which is dehydrated to give the intermediate 77 with subsequent formation of pyrrole 69.79 Synthesis of pyrroles 80 through carbocupration of N,N- bis(trimethylsilyl) derivatives of alkyl 4-aminobut-2-ynoates with subsequent treatment with acyl chlorides has opened a route to new polyheterocyclic compounds, including furyl- and thienyl- pyrroles and tetrapyrrole macrocycles.Thus, the reaction of lithium methyl(hexynyl)cuprate with methyl (79a) or ethyl 4-aminobut-2-ynoates (79b) with subsequent treatment with 2-furoyl or 2-thienoyl chlorides leads to 2-furyl- (80) and 2-thienyl-pyrroles (81a,b). Me MeO2C b R=Me O HN 80 a Me RO2C RO2CC CCH2N(SiMe3)2 79a,b c S 81a,b HN R=Me (81a, 30%), Et (81b, 58%); (a) (BunC:C)MeCuLi, Et2O,740 8C; (b) COCl . COCl ; (c) S O 4. Synthesis involving isocyanides Isocyanides are rather common components in the synthesis of hetarylpyrroles.81 ± 87 3-(2-Furyl)- (84a,b) and 3-(2-thienyl)pyr- roles 84c were synthesised by addition of 1-tosylalk-1-enyl isocyanides 82 to chalcone analogues 83 in the presence of S E Korostova, A I Mikhaleva, B A Trofimov ButOK.81 The initial isocyanides 82 can be easily obtained by the reaction of tosylmethyl isocyanide with carbonyl compounds. Ts ButOK, THF (H2C)n + 20 8C COR X N C 83 82 X COR (H2C)n NH 84a ± cn Yield (%) R X Compound 84 abc 93 89 85 2-furyl Ph 2-thienyl 132 OOS It is presumed 81 that the formation of pyrroles 84 occurs by the following mechanism: R2 O R1 7 Ts Ts 7 H+ ButOK 83 82 (H2C)n N C 85 C N 86 (CH2)n O R1 R2 7 Ts H+ 84a ± c. H N (CH2)n The Michael addition of the allylic anion 85, formed under the action of a base, to the vinyl ketone 83 and subsequent cyclisation of the enolate anion 86 and elimination of the tosylate anion yield the final reaction product.Pyrroles 89a ± c were obtained from 2-cyano-3-(2-furyl)- and -(2-thienyl)acrylic acids 87a ± c and tosylmethyl isocyanide (88) in a superbasic system KOH± dimethoxyethane.82 R R CN Ts CO2H 1. 710 8C, 1 h 2. 20 8C, 2 h X + CN X N C 88 87a ± c 89a ± c NH Yield of 89 (%) R X Compounds 87, 89 abc 92 97 89 HHMe OSS The addition of methyl isocyanoacetate (90) to thiophene-2- carbaldehyde in the presence of 1,8-diazabicyclo[5.4.0]undec-7- ene (DBU) in THF at 40 ± 50 8C (15 h) leads to 3-(2-thienyl)- pyrrole (91) in 70% yield.85, 87 CO2Me O DBU +MeO2C S N C H 90 S MeO2C91 NH Sakai et al.85 have obtained pyrrole 91 by a different scheme including intermediates 92 and 93.Bipyrroles, furyl- and thienylpyrroles O +90 S HS H Phosphorylated 3-(2-furyl)pyrrole 95 was synthesised in 54% yield by the reaction of isocyanide 94 with 2-(2-nitroprop-1- enyl)furan in the presence of lithium diisopropylamide (LDA) (THF, 778 8C, 3 h; *20 8C, 72 h; reflux, 2 h) with subsequent treatment with ammonium chloride.86 C N P(O)(OEt)2+ 94 O (EtO)2(O)P 95 5.Syntheses from cyanides A simple and efficient method for the synthesis of 2-(2- furyl)pyrroles 98, 99 is based on the reaction of tetracyanoethane with azomethine 96 or azine 97.88 ± 92 CN NPh O 96 NC CN CNNH2 O H NPh 100 (92%) 60 ± 70 8C 7HCN CN NC NH2 O NPh 98 Pyrrole 98 is synthesised either by heating (90 ± 100 8C) of tetracyanoethane and azomethine 96 in aqueous dioxane (yield 60%± 65%) 88 or with the intermediate isolation of pyrroline 100, which is transformed into pyrrole 98 in 98% yield upon heating for 5 ± 25 min at 60 ± 70 8C in DMF.89 Similarly, upon gentle heating in an aqueous-ethanolic or ethanolic medium pyrroline 101 loses one molecule of hydro- cyanic acid and is transformed into pyrrole 99.91 Condensation of bis(2-thienyl)acetylene 102 with an excess of trimethylsilyl cyanide catalysed by palladium or nickel chloride in pyridine yields 3,4-bis(2-thienyl)pyrrole (103).93 NaH, THF NHCHO POCl3 Et3N S H CO2Me 92 NC 90, DBU 91 CO2Me 93 Me LDA 7HNO2 NO2 OMe NH NC CN NC N O 2 97 NC CN CNNH2 N O H N 101 (86%) O 7HCN D CN NC NH2 N O N 99 O 467 S S Cat C C +Me3SiCN S S NC N(SiMe3)2 102 NH 103 (75%) The reaction involves such a profound rearrangement of trimethylsilyl cyanide that it is hardly possible to propose any reasonable mechanism of this reaction.6. Organometallic synthesis Kaufmann and Lexy 94 have synthesised bipyrrole 104 by a-lithiation of N-methylpyrrole with subsequent oxidative cou- pling of the intermediate obtained. Bipyrrole 105 was used for the preparation of terpyrrole and then oligopyrrole. BunLi NiCl2 Li TMEDA NMe NMe Li Li Me N Me N NMe NMe Li Me N Me N 104 Me N 1) BunLi, TMEDA 2) NiCl2 H H NMe NMe NMe n 105 TMEDA is tetramethylethylenediamine; n=6.3,3-Bipyrrole 106 was obtained 95 by a Wurtz-type reaction from 3-bromo-1-(triisopropylsilyl)pyrrole (107) upon treatment of the latter with tert-butyllithium in the presence of CuI. Br a, b N N N SiPri SiPri SiPri 3 3 3 107 106 (76%) (a) ButLi, CuI, n-C5H12 ± THF,778 8C, 10 min; (b) 20 8C, 2 h. Copper bronze Groenendaal et al.96 have described dimerisation of 5-bromo- 1-tert-butoxycarbonyl-2-phenylpyrrole according to Ullmann. CO2But N Ph DMF, 100 8C, 2 h Ph Br Ph N N CO2But CO2But (77%) Under the same conditions,96 2,5-dibromo-1-tert-butoxycar- bonylpyrrole afforded a dimer and other oligomers containing up to 24 pyrrole rings, which were isolated as individual products by preparative HPLC.Copper bronze + DMF, 100 8C Br Br NR NR NR NRn R=CO2But; n=3 ± 24. Gjùs and Gronowitz 97 have developed a method for the synthesis of thienylpyrroles 108 and 109 by reaction of 2-iodo-468 Thienylpyrrole 108 was obtained by the same authors 97 in three steps: thiophene 110 with 1-methyl-2-pyrrolylcopper (111) prepared by metallation of 1-methylpyrrole with butyllithium in a mixture of ether and tetramethylethylenediamine followed by treatment with cuprous bromide. I S S 110 Cu2Br2 BuLi Li Cu 111 NMe NMe NMe + + S S S NMe NMe 113 108 114 NMe + + S S S The reaction of 2-acetylthiophene with dimethylformamide and phosphorus oxychloride followed by treatment of the reaction mixture with HClO4 yields perchlorate 118.The reaction of salt 118 with dimethylamine gives perchlorate 119; its heterocyclisa- tion leads to thienylpyrrole 108, although in a low yield.97 NMe S 115 S 109 + I 112 NMe The use of the Stille reaction allowed one to synthesise oligoheterocyclic compounds containing pyrrole and thiophene units, e.g., N-tert-butoxycarbonyl-di-2-(5-tert-butoxycarbonyl-5- phenyl-2-pyrrolyl)thienyl]pyrrole (123).98 Organotin derivative 120 is prepared by stannylation of a substituted pyrrole. Cross- coupling of compound 120 with bromobenzene according to Stille in the presence of palladium[tetrakis(triphenylphosphine)] fol- lowed by stannylation of the reaction product yields compound 121.Pyrrole 122 is obtained by distannylation of tert-butyl 2,5- bis(2-thienyl)pyrrole-N-carboxylate with subsequent bromina- tion of the reaction product with N-bromosuccinimide. Cross- coupling of compounds 121 and 122 according to the Stille method leads to the oligoheterocycle 123. NH The course of the reaction and the ratio of the products formed are substantially influenced by its duration: when the reaction was conducted for 1 h, a mixture was obtained that contained, along with thienylpyrrole 108, the initial iodothio- phene (110), iodopyrrole 112, bipyrrole 113 and bithiophene 114; when the mixture was allowed to react for 4 h, a decrease in the concentration of bithiophene 114 and an increase in the content of thienylpyrrole 108 and bipyrrole 113 were observed. Trace amounts of 1-methyl-2-[5-(2,20-bithienyl)]pyrrole 109 and terthio- phene 115 were also formed.According to GLC data, the ratio of products 108 : 109 : 113 : 114 : 115 was 53 : 14 : 24 : 7 :<1. The major product, viz., thienylpyrrole 108, was isolated in 40% yield, while the yield of compound 109 was 15%.97 The formation of the latter product is explained by the following scheme: 110+111 Cu + I S NMe 111 110 109 I S S S 110 111 109 Cu S Me N The reaction of the copper intermediate 111 with 3-iodo- thiophene leads to a more complex mixture of products.97 I S 111 121+122 114+108+ + + S S S S and other.+ +109+115 + NMe Me N S S 116 117 S (a) (ButO2C)2, ButOK, THF, D; (b) tetramethylpiperidine, BunLi, THF, 770 8C; (c) Me3SnCl, THF, Pd(PPh3)4,770 8C; (d ) PhBr, Pd(PPh3)4 , PhMe, H2O, Na2CO3 ; (e) NBS, THF; ( f ) Pd(PPh3)4 , PhMe, Na2CO3 , H2O, D. Cross-coupling according to Stille was also used to prepare the following compounds:98 According to GLC data, bicyclic and tricyclic compounds constitute 7% and 30% of this mixture, respectively. Column chromatography with subsequent distillation and recrystallisation gave 1-methyl-2-(3-thienyl)pyrrole (116) (yield 72%), thienyl- pyrrole 108 (yield 6%) and tricyclic compound 117 (yield 9%). S E Korostova, A I Mikhaleva, B A Trofimov Me2NH 1.DMF, POCl3 2. HClO4 + C CHCH COMe NMe2ClO¡4S 118 (31%) Cl NaH + C CH CH NMe2ClO¡4S S Me N 108 119 (60%) NMe2 a d b, c SnMe3 N N CO2But CO2But (66%) 120 (80%) b, c Ph Ph SnMe3 N N CO2But CO2But (100%) 121 (97%) S S S a b, c, e N NH (70%) CO2But S S Br Br N 122 (38%) CO2But f S S Ph Ph N N N CO2But CO2But CO2But 123 (14%)Bipyrroles, furyl- and thienylpyrroles S But But RN RN Ph Ph R N S But S S Ph NR RN Ph S R=ButO2C. Cross-coupling of N-methylpyrrolylmagnesium(zinc) halides with 3-halogenothiophenes in the presence of a palladium catalyst leads to 1-methyl-2-(3-thienyl)pyrrole (116) in high yield.99, 100 Y + MX S Me N M=Mg, Zn; X=Cl, Br, I; Y=Cl, Br; [Pd]=PdCl2, Ph2P(CH2)4PPh2 (2.5 mol %).7. Synthesis from azides Thermolysis of azidobutadienes 125a ± d containing a hetaryl substituent yields the corresponding 2-hetarylpyrroles 126a ± d. The starting compounds for the preparation of azides 125a ± d were b-2(3)furyl- and b-2(3)-thienylacroleins 124a ± d and methyl azidoacetate. Alkaline hydrolysis of compounds 126a ± d and subsequent decarboxylation of acids 127a ± d in quinoline (with barium-promoted copper chromite as a catalyst) lead to the corresponding 2-(2-furyl)- and 2-(2-thienyl)pyrroles (128a,b) or 2-(3-furyl)- and 2-(3-thienyl)pyrroles 128c,d.101 CO2Me + X N3 CHO 124a ± d N3 X 125a ± d X NH X 469 Yield (%) X S Substituent position Com- pound Ph 128 127 126 RN 68 46 43 57 79 75 79 89 90 90 92 87 abcd OSOS 2233 S But S 8.Synthesis from unsaturated aminocarbonyl compounds R N Ph Rapoport et al.102-106 have synthesised 2,2-bipyrrole (130) and 2,20:50,200-terpyrrole (131) by the reaction of 2-pyrrolidone with phosphorus oxychloride and subsequent dehydrogenation of the intermediate product 129. RN R S POCl3 NH + Ph O OPOCl2 Cl7 NR NH NH N H Pd/C + R R NH NH N Cl7 H Cl2OPO 129 HN [Pd] R THF, D, 1 h NMe HN 130, 131 S 116 R = H (130), 2-pyrrolyl (131). 2-(2-Furyl)- and 2-(2-thienyl)pyrroles have been synthesised in 86% yields by the reaction of N-allyl carboxamides 132 with phosgene followed by the elimination of HCl from the chloro- imines 133.107 COCl2 ButOK HN N 7HCl X X 133 Cl 132 O C 7 N X X N+ 135 134 710 8C X HN X=O, S.D, xylene CO2Me The construction of the pyrrole ring is presumed 107 to result from the intramolecular 1,5-dipolar cyclisation of the intermedi- ate nitrile ylides 134 to 3H-pyrroles 135, which are further isomerised to 1H-pyrroles. HO7 CO2Me 126a ± d CO2 1-(2-Thienyl)pyrroles 139 were obtained in good yields in the reaction of sodium derivatives of cyanoamides 136 with vinyl- triphenylphosphonium bromide (137) followed by the intramo- lecular cyclisation of the intermediate 138 under the action of a base with elimination of HCN and Ph3PO.108 X CO2H 127a ± d NH NHR2 7Na+ 128a ± d CN C + THF, D, 24 h, N2 R1N + H2C CHPPh3Br7 137 O C 136 PhS E Korostova, A I Mikhaleva, B A Trofimov 470 7 + d R2 CO2Me PPh3 COPh Ph R2 NC S 147 OAc NR1 138 NR1 139 OAc OAc R1=2-thienyl: R2=Ph, 4-ClC6H4 .e CO2Me CO2Me S S RN NR 9. Miscellaneous methods NH 149 148 (a) NaH, THF, BunLi, 0 8C, 3 h; (b) 1. p-TsCl, Py, 20 8C, 12 h; 2. 80 8C, 0.5 h; (c) MeCOCl, Py, 0 8C, 5 h; (d ) RN=NR (R=Cl3CCH2OCO), D, PhH; (e) Zn, AcOH, MeOH, 55 8C, 3 h. Boberg et al.109 have proposed to obtain 3-(2-furyl)- (140a) and 3-(2-thienyl)pyrroles (140b ± e) from nitronic acids 141 formed upon reaction of nitrovinylfurans or -thiophenes with alkyl acetoacetates. Heterocyclisation of compound 141 takes place at 60 8C in the presence of reducing agents, such as a mixture of sodium sulfide with ammonium chloride, aqueous ammonia with ammonium sulfide or Zn in HCl.This method does not have any preparative value because of low yields of pyrroles. Me CO2R2 O Me X +2 X CO2R2 R1 NO2 O NO2H R1 141 Condensation of 2-formylthiophene 144 with ethyl acetoace- tate leads to hydroxy ketone 145 (yield of a crude product 93%) dehydration of which yields E-olefin 146.114 Its O-acetylation affords a mixture of E- and Z-isomers of diene 147 from which the Z-isomer was isolated by crystallisation from ether (yield 31%). The Diels ± Alder reaction of the Z-diene with trichloroethyl azo- N,N 0-dicarboxylate results in a tetrahydropyridazine derivative 148 as an intermediate; its treatment with Zn powder in acetic acid yields thienylpyrrole 149.114 CO2R2 1-Vinyltetrahydroisoquinoline (150) (obtained from 3,4-di- X Me R1 140a ± e hydroisoquinoline and vinylmagnesium bromide) reacts with 2-formylfuran (151) or 2-formylthiophene (144) to form polycy- clic systems (152a,b) with 2-furyl or 2-thienyl substituents.115 NH Yield (%) R2 R1 X Compound 140 O N C NH + X X H 150 152a,b 144, 151 X X = O (a, 40%), S (b, 53%).Et Et Et But Et Et Me Et Me Ph 117 14 10 8 abcde OSSSS Pyrrolo[1,2-a]quinoxalines 154 containing a 2-(2-furyl)- pyrrole fragment have been synthesised by the cyclisation of 2-(2-furfurylideneacetyl)quinoxalines 153 under the action of HCl or HClO4.116 Synthesis of 1-(2-cyano-3-thienyl)pyrrole (142) and 1-(3- cyano-2-thienyl)pyrrole (143) by the reaction of 2,5-dimethoxy- tetrahydrofuran (50) with 3-amino-2-cyano- and 2-amino-3- cyanothiophenes, respectively, has been described.110 ± 113 R N NH2 R N N N + O OMe MeO CN N O S O CN S 154 153 O 142 CN R=H, Me.S 50 N NH2 S 143 CN Corrole 156 is prepared by the oxidative cyclisation of a linear tetrapyrrole 155 in an alkaline medium.117 Me Me Me Me Et Et Me Me Kresze et al.114 have carried out a multistep synthesis of 4-acetoxy-5-(methoxycarbonyl)-2-(2-thienyl)pyrrole (149) in which the key compound is a tetrahydropyridazine derivative 148. N N HN 155 HN H OMe Me a Me Me + S O O O Et Me 144 NH HN b CO2Me HN N S 145 O OH Et Me c 156 (68% ± 84%) Me Me CO2Me S 146 (46%) O As expected, boiling of meso-thiaphlorin 157 in o-dichloro- benzene resulted in elimination of sulfur with the formation ofBipyrroles, furyl- and thienylpyrroles corrole 158 in 35%± 40% yield.118 In the presence of triphenyl- phosphine, the yield of corrole 158 reaches 60%.X Me Me S X X N NH N N Me Me 157 Et Et X=CO2Et. In the case of meso-dithia macrocycle 159, the extrusion of sulfur occurs less readily (213 8C, 10 h, triphenylphosphine) and leads to two isomers of meso-thiacorrole 160a,b in a total yield of 42% (ratio 1 : 1).118 Me Me S X X N NH HN N Me Me S 159 Et Et Me Me X N NH N HN Me S Et Et 160a X=CO2Et.III. Reactivity Until recently, chemical properties of pyrroles bound to five- membered aromatic heterocycles have been very little studied because of the absence of reliable general methods for their synthesis. There were only fragmentary data on some reactions that included mostly interactions of their functional substituents (with the exception of formylation of the pyrrole fragment). The development of methods for the synthesis of hetaryl-substituted pyrroles in the last two decades has led to a considerable expansion of the array of preparatively accessible representatives of this group of compounds and to the appearance of real conditions for systematic studies of their reactivity and other properties.1. Electrophilic substitution a. Protonation It is known 1 ± 3, 119 ± 125 that the most characteristic reaction of the pyrrole ring is electrophilic substitution, which is also typical of substituted furans and thiophenes.121 ± 125 The simplest reaction of this type is reversible protonation (degenerate electrophilic sub- stitution of hydrogen). Over the last decade, the 1H NMR method was used for the systematic studies of protonation of the series of 2-(2-furyl)- and 2-(2-thienyl)pyrroles and their 1-vinyl derivatives. Me Et Me NH HN 7S N NH Et Me 158 Me X 7S Et Et Me Me X N NH + HN N X X Me S Me Me 160b 471 Numerous studies 21, 25, 27, 126 ± 133 have shown that at 780 to 750 8C all pyrroles, regardless of the nature of the second heterocycle and the acid (except for the `magic' acid HSO3F± SbF5), are protonated at the C(5) atom of the pyrrole ring; in the case of vinyl derivatives (R3=CH=CH2), the double bond is retained.R2 R2 H HA + R1 R1 780 to750 8C X X H RN3 NR3 X=O, S; R1=H, Me; R2=H, Me, Pri; R3=H, CH=CH2 , Et; A=Cl7, Br7, CF3COO7, FSO¡3 . The reaction of furylpyrroles with hydrogen halides (HCl, HBr) at 730 8C yields a mixture of pyrrolium 161 and furanium 162 cations.21, 25, 128 An analogous mixture is also formed on gradual elevation of temperature from780 to730 8C.21, 25, 128 R1 R1 H H + + O O H H 162 R2 N 161 NR2 Protonation of vinyl derivatives (730 8C) is accompanied by the addition of HX to the 1-vinyl group.21, 25, 27, 126 ± 131 On warming from 780 to 730 8C, the cation in which both a-positions in the furan ring are occupied adds HX only to the double bond with the formation of the pyrrolium ion 163.21, 25, 27, 128 R H + Me N O H 163 X Me R=H, Me; X=Cl, Br.As the temperature is increased to 0 8C, the reaction of furylpyrroles with HBr leads to 4-bromo-4,5-dihydro-2-(2-pyrrol- yl)furanium bromides 164 as the result of addition of HBr to the protonated furan ring.21, 25, 27, 128 R1 Br +O 164 RN2 R1=H:R2=H, CHBrMe, Et; R1=Me: R2=H, CHBrMe. At 20 8C, dihydrofuranium cations 164 are completely trans- formed into 4-bromo-4,5-dihydro-2-(2-furyl)pyrrolium bromide 165 with the rearomatisation of the furan moiety of the mole- cule.21, 25, 27, 128 Br + R1 O 165 RN2 R1=H, Me; R2=CHBrMe.In contrast to the furan ring, the thiophene ring is not involved in the reaction with HBr. In this case, the formation of only 4-bromo-4,5-dihydro-2-(2-thienyl)pyrrolium bromide (166) is observed.21, 25, 130, 131472 Br +N S 166 CHBrMe Unlike ordinary acids, the superacidic system HSO3F ± SbF5 reacts with 2-(2-furyl)-1-vinylpyrrole at 770 8C to give an equilibrium mixture of dications: 1-(1-ethanio)-2-(2-furyl)-5H- pyrrolium (167) and 1-(1-ethanio)-2-(2-furyl)-3H-pyrrolium (168) in the ratio 3 : 1.21, 25 H H H + +N O N O H + + 168 167 CHMe CHMe In the reaction with the superacidic system HSO3F± SbF5±SO2, 2-(2-thienyl)-1-vinylpyrrole forms two cations: 1-(1- ethanio)-2-(2-thienyl)-5H-pyrrolium (169) and 2-[1-(1-ethanio)-2- pyrrolyl)-5H-thiophenium (170) in the ratio 2 : 1.Heating of the reaction mixture up to 10 8C changes the ratio 169 : 170 to 1 : 2.21, 25, 130, 131 H H + + S N N S H H + + 170 169 CHMe CHMe b. Formylation Formylation of 2-(2-furyl)-, 2-(2-thienyl)-, 2-(3-furyl)- and 2-(3- thienyl)pyrroles and their a-methoxycarbonyl derivatives under conditions of the Vilsmeier reaction has been studied.101, 134 Thienylpyrroles are formylated only at the a-position of the pyrrole ring with the formation of 2-thienyl- (171) (yield 75%) and 3-thienyl-2-formylpyrrole (172). DMF, POCl3 CHO ClCH2CH2Cl S S HN HN 171, 172 Formylation of furylpyrroles results in a mixture of mono- (173, 175) and diformyl derivatives (174, 176), the former being predominant.101, 134 DMF, POCl3 O 1a NH CHO +HOC CHO NH NH O 174 (14%) O 173 (62%) DMF, POCl3 HN 128 O CHO CHO + NH NH CHO 176 175 O O In 2-(3-furyl)pyrrole (128), both a-positions of the furan ring are free.However, the formyl group enters the position 2 of the furan ring, which is in accord with the standard orientation rule. S E Korostova, A I Mikhaleva, B A Trofimov and (126b) -2-(3- 5-Methoxycarbonyl-2-(2-thienyl)- thienyl)pyrrole (126d) give a mixture of 3-formyl-substituted pyrroles (177, 178) and 2-(5-formyl-2-thienyl)- (179) and 2-(2- formyl-3-thienyl)-derivatives 180, respectively.134 DMF, POCl3 CO2Me S 126b HN HOC + HOC CO2Me CO2Me S S 179 177 HN HN DMF, POCl3 CO2Me NH S 126d HOC CO2Me CO2Me + NH NH CHO 180 178 S S Because of the orienting effect of the a-methoxycarbonyl group, formylation in pyrroles 126b,d occurs mostly in the b-position of the pyrrole ring (95%); the formylation product in the a-position of the thiophene ring is formed in the yield not exceeding 5%.Unlike its thienyl analogues, 2-(2-furyl)- (126a) and 2-(3- furyl)-5-methoxycarbonypyrrole (126c) form in this case virtu- ally only the products of formylation at positions 5 and 2 of the furan ring, viz., pyrroles 181, 182.134 DMF, POCl3 HOC CO2Me CO2Me O O 126a HN 181 HN CO2Me CO2Me DMF, POCl3 NH HN CHO 182 126c O OHeating of 2,5-di(2-pyrrolyl)thiophene 183 with benzoyl chloride in DMF leads to the formation of 2,2-di(5-formyl-2- pyrrolyl)thiophene 184 in 47% yield.62 1.PhCOCl, DMF 2. Na2CO3 , EtOH NH NH S 183 CHO HOC NH NH S 184 c. Acylation Acylation of bipyrroles, furyl- and thienylpyrroles devoid of functional substituents with carboxylic acid anhydrides and halides has not been studied, except for trifluoroacetylation. Acylation of 1-(2-aminomethyl-3-thienyl)pyrrole 185 occurs exclusively at the amino group with the formation of 1-(2- aroylamidomethyl-3-thienyl)pyrroles 186a ± c.111Bipyrroles, furyl- and thienylpyrroles N N RCOCl CH2NHCOR CH2NH2 S S 185 186a ± c R=Ph (a, 82%), 4-ClC6H4 (b, 89%), 4-NO2C6H4 (c, 74%).Starting from an isomeric aminomethylthienylpyrrole 187, thienylpyrrolodiazepines 188 were obtained.111, 113 S S N N COCl2 RNH2 PhH CH2NHCOCl 187 CH2NH2 S N S POCl3 N N NHR CH2NHCNHR 188 O R=H, Me, Et, Ph, MeC6H4 , 2-thienyl, NHPh. Tricyclic compound 191 was obtained from 1-(3-cyano-2- thienyl)pyrrole (143) by a series of consecutive reactions, includ- ing the rearrangement of the azide 189 into an intermediate 190.112 CN CO2H 1. H2NNH2 2. HNO2 H+, D N N S S 143 NCO CON3 NH S O N N N 7N2 S S 191 190 189 Similarly, isomeric tricyclic compound 193 was synthesised from thienylpyrrole 192.112 S S CO2Me NH O N N 193 192 Trifluoroacetyl derivatives of various pyrrole systems have been attracting attention of researchers since the 1960 ± 1970s.135 ± 145 Trifluoroacetylation was often used on many occasions for comparing reactivities of various pyrrole, furan and thiophene derivatives in electrophilic substitution reactions.121, 124, 135 ± 137 The relative trifluoroacetylation rates of pyrrole, furan and thiophene with trifluoroacetic anhydride are 5.36104, 1.46102 and 1, respectively.121, 124 This ratio of rates illustrates the well known fact that pyrrole, owing to the higher, compared to other five-membered aromatic heterocycles, ability of its nitrogen atom to delocalise the positive charge in cationic s-complexes, is appreciably more reactive than furan and the more so thiophene in reactions of electrophilic substitution.122 It was thus believed 146 that 2-(2-furyl)- (1a) and 2-(2- thienyl)pyrroles (7a), as well as their 1-vinyl (2a,b, 8a ± c), 1-ethyl (194a) and 1-triethylsilylmethyl derivatives (194b), would be acylated with trifluoroacetic anhydride only at position 5 of the pyrrole ring.473 R2 R2 (CF3CO)2, Py COCF3 X X RN1 RN1 195a ± i X R2 R1 Yield of 195 (%) Reaction product Initial compound 195a 195b 195c 195d 195e 195f 195g 195h 195i 1a 7a 2a 2b 8a 8b 8c 194a 194b 41 75 18 34 78 76 63 54 75 HHHMe HMe Et HH HHCH2=CH CH2=CH CH2=CH CH2=CH CH2=CH Et CH2SiEt3 OSOOSSSSS It was shown 146, 147 that it is compounds 195 that are formed upon trifluoroacetylation of furyl- and thienylpyrroles. In the case of 2-(2-furyl)pyrrole 1a, the reaction mixture contains, in addition to the initial pyrrole and the final 2-(2-furyl)-5-trifluoroacetyl- pyrrole 195a, an admixture of a third compound, presumably the product of trifluoroacetylation in the a-position of the furan ring or the b-position of the pyrrole ring.The latter is possible for activated pyrroles.122 In contrast to 2-phenylpyrrole,145 under comparable conditions the trifluoroacetylation of 2-(2-furyl)- (1a) and 2-(2-thienyl)pyrroles (7a) is not complete: the degree of conversion is 50%± 60%, and the yields of the corresponding 5-trifluoroacetylpyrroles 195a,b are 40% ± 50% with respect to the initial pyrrole. The use of a twofold excess of anhydride relative to pyrrole and the increase in the reaction temperature to 50 8C do not change its direction, though the yield of compound 195b increases to 75%.2-(2-Thienyl)-1-vinylpyrrole 8a is acylated with both the equimolar amount and the twofold excess of trifluoroacetic anhydride to give virtually only one compound 195e, whereas the major route of trifluoroacetylation of 2-(2-furyl)-1-vinylpyr- role 2a is the attack at the a-position of the furan ring with the formation of compound 196. With the twofold excess of anhy- dride, the content of isomer 195c increases to 40% along with the increase in the overall yield (62.5%).146 R (CF3CO)2O, Py 2a,b 195c,d+ F3CCO N O 196, 197 (28%) R=H(196, 28%);Me (197, 27%).Substitution of a-hydrogen in the pyrrole ring by the trifluor- oacetyl group is facilitated with the increase in its basicity, as is the case with 2-(2-furyl)- and 2-(2-thienyl)-1-vinylpyrroles with alkyl substituents at position 3. In contrast to compound 2a, the attack of 2-(2-furyl)-3-methyl-1-vinylpyrrole (2b) by the trifluoroacetyl cation is directed mostly not at the furan but at the pyrrole ring. Structurally isomeric trifluoroacetylpyrroles 195d and 197 are formed in 34% and 27% yields, respectively.148 3-Alkyl-2-(2- thienyl)-1-vinylpyrroles 8b,c form only a-trifluoroacetylpyrroles 195f,g. Substitution of the vinyl group by the ethyl group at the nitrogen atom (compound 194a) does not change the direction of trifluoroacetylation. It also occurs in the pyrrole ring of com- pound 194b in which there is a rather bulky triethylsilylmethyl substituent.148 In order to interpret the results obtained, Trofimov et al.146 considered the following factors; (a) an appreciably larger nucleophilicity of the furan ring compared to that of thiophene; (b) a decrease in the electron density in the pyrrole ring due to competitive conjugation of the vinyl group with the unshared electron pair on nitrogen and also owing to its negative inductive effect; (c) possible steric hindrance due to the vinyl group for the474 attack at the a-position of the pyrrole ring.The reactivity of the thiophene ring is so low compared to the furan ring, and the more so, the pyrrole ring, that the factors (a) and (b) seem to be insufficiently strong for changing the direction of trifluoro- acetylation of 2-(2-thienyl)-1-vinylpyrrole.The absence of double acetylation products points to a very strong influence of the electron-acceptor trifluoroacetyl substituent that is transferred to the thiophene ring. The mechanism of such a transfer cannot be reduced solely to the inductive effect but should include formation of a common conjugation system.146 CF3 + COCF3 N X N X O7 CF3 + N X O7 It is known that trifluoroacetylation of 1-vinylamides,149 ethyl iminoacetoacetate 150 and 9-alkenylcarbazoles 151 can result in electrophilic substitution of hydrogen in the 1-vinyl group.This reaction route was not observed for 2-(2-furyl)- and 2-(2-thienyl)- 1-vinylpyrroles. d. Aminomethylation The reaction 1-(4,5-dimethyl-2-thienyl)pyrrole (51a) with the aminomethylation system Me2NH± HCl ±CH2O in acetic acid results in a 2-dimethylaminomethyl derivative of this pyrrole (198).109 Me Me Me2NH, HCl, CH2O, AcOH N N Me Me S S 51a CH2NMe2 198 (*2%) 4-Aryl-5,6-dihydro-4H-pyrrolo[1,2-a]thieno[2,3-f]-1,4-diaze- pines 202a ± h were synthesised by the reaction of 1-(2-benzyl- ideneaminomethyl-3-thienyl)pyrroles 200, obtained from 1-(2- aminomethyl-3-thienyl)pyrrole (199) and the corresponding aro- matic aldehydes, with hydrogen chloride in dry ethanol (D, 45 min) with subsequent treatment of the hydrochlorides formed (201) with dilute NaOH.110 N N HCl ArCHO EtOH CH2N CHAr CH2NH2 S S 200 199 N N Ar NaOH + NH S S 202a ± h Ar NH2 Cl7 201 Ar Yield (%) Compound 202 abcdefg 48 62 56 53 65 63 52 Ph 2-ClC6H4 3-ClC6H4 4-ClC6H4 3-MeOC6H4 4-MeC6H4 3,4-(MeO)2C6H3 O h 62 O S E Korostova, A I Mikhaleva, B A Trofimov 2.Nucleophilic addition to the triple bond Vinylation of NH-heterocycles with acetylene is usually carried out under a pressure of 20 ± 40 atm at 160 ± 200 8C. This process has limited laboratory uses because of its explosion hazard and the need for special equipment, though several such technologies have been realised on the industrial scale.22 The use of superbasic catalytic systems of the KOH±DMSO type allows vinylation of alkyl- and arylpyrroles at 100 ± 200 8C and atmospheric pressure.152 Under these conditions, 2-(2-furyl)- (1a) and 2-(2-thienyl)pyrroles (7a) are successfully vinylated.153 The results of vinylation of compounds 1a, 7a and 2-phenyl- pyrrole are presented in Table 2.KOH, DMSO +HC CH N X X 1a, 7a NH 2a, 8a X = O (1a, 2a), S (7a, 8a). KOH, DMSO +HC CH Ph Ph N HN 203 Table 2. Effect of vinylation conditions on the yield of 2-(2-furyl)- (2a), 2-(2-thienyl)- (8a) and 2-phenyl-1-vinyl-pyrroles (203).153 Yield of the vinylation product a, b Tempera- Time ture /8C /h 8a 203 2a (2 : 98) 98 7 (15 : 85) (3 : 97) 98.5 7 7 7 7(5 : 95) (*0.1 : 99.9) 7(12 : 78) (5 : 95) (2 : 98) (*0.1 : 99.9) (7 : 93) 87 86 7 44.5 55.5 634 110 110 110 110 110 120 120 a The yield was calculated after distillation.b The ratio of the initial compound to the reaction product (according to GLC) is indicated in parentheses. Acetylene was passed through the reaction mixture under atmospheric pressure. Vinylation was carried out at a fivefold molar excess of KOH. Virtually quantitative yields (96% ± 98.5%) of 1-vinylpyrroles were attained at 110 8C within 4.5 ± 6 h (see Table 2). With increase in the temperature to 120 8C the duration of vinylation is reduced, although the yield of 1-vinylpyrroles drops by 9%± 16% because of side processes. When vinylation was conducted in an autoclave (initial acetylene pressure 10 ± 14 atm, maximal pressure 20 ± 30 atm, DMSO, threefold molar excess of KOH, 120 8C), the yield of vinylpyrroles did not exceed 80% due to resinification.Thus, vinylation of aryl- and hetarylpyrroles under atmospheric pres- sure is not only safe but also more efficient. The vinylation rate of pyrroles increases in the sequence 2-phenylpyrrole<2-(2-thienyl)pyrrole<2-(2-furyl)pyrrole (see Table 2). Their NH acidity increases in the same order. (Jack- son 3 did not give the value of 2-(2-furyl)pyrrole acidity, but it can be estimated from the cited dependence of pKa on induction constants of substituents in the pyrrole ring). The observed order of reactivity of pyrroles fully confirms the previously established general regularity 153 ± 157 that the rate of base-catalysed addition of structurally similar azoles to acetylene, including 2-furyl- and 2-thienylpyrroles, increases with their acidity.However, it was reported 153 that the yield of 1-vinyl derivatives of tetrahydro-g- carbolines increases with the introduction into their molecule of a methyl group, which decreases the NH-acidity. Thus, the infor-Bipyrroles, furyl- and thienylpyrroles mation available at that time did not give any clear idea about the influence of NH-acidity (nucleophilicity) of azoles on the rate of their addition to acetylene. In order to assess the influence of substituents on the vinyl- ation rate of pyrroles, the kinetic curves of vinylation of the series of NH-pyrroles for which precise values of pKa of relative acidity are known have been compared quantitatively 156 with the correlation equations relating these parameters to the substituent constants.4, 157, 158 Vinylation was carried out in a KOH±DMSO system at 100 8C under atmospheric pressure.The course of the process was monitored using GLC. It was found 156 that the vinylation rates of 2-benzyl-3-phenyl-, 3-hexyl-2-phenyl-, 2,3- diphenyl-, 2-(4-ethylphenyl)- and 2-(4-methoxyphenyl)pyrroles increase with the decrease in their acidity (increase in pKa values 159). However, 2-(2-furyl)- and 2-(2-thienyl)pyrroles are vinylated more readily than expected from consideration only of the values of their acidity, i.e., possessing a stronger acidity than that of 2-phenylpyrroles, they have higher vinylation rates.The observed deviation may be explained by the fact that the nucleophilicity of anions is determined not only by their basicity (pKa values of conjugated acids) but also by their polarisability and ionisation potential. Therefore, furyl- and thienylpyrroles, which are more easily polarised and possess lower ionisation potentials (owing to the heteroatoms conjugated with the pyrrole moiety of the molecule), should form anions of increased nucleophilicity with concomitantly decreased basicity. Despite the fact that the acidity of pyrroles increases and the basicity of the corresponding anions decreases in the order: 2-phenylpyr- role<2-(2-thienyl)pyrrole<2-(2-furyl)pyrrole, the reaction rate increases in the same order since the ionisation potentials 160, 161 decrease in the following order (eV): 2-phenylpyrrole (7.61)>2-(2-thienyl)pyrrole (7.41)>2-(2-furyl)pyrrole (7.29).Thus, the 2-(2-furyl)- and 2-(2-thienyl)pyrrole anions, while being less basic than the 2-phenylpyrrole anion, are superior to it in their nucleophilicity. Yet another peculiarity of the vinylation reaction of pyrroles has been revealed:156 all kinetic curves are clearly S-shaped, which is characteristic of autocatalytic reactions. In principle, autocatal- ysis should be characteristic of vinylation reactions in aprotic and especially superbasic media. Indeed, with the consumption of protogenic substrate (the NH group of pyrrole in this case) one observes an increase in both nucleophilicity of the remaining anions (due to lower solvation) and the degree of ionisation of the remaining substrate molecules (owing to increased medium basicity).At the same time, the chemical activity of acetylene molecules is enhanced owing to complexation with partially dehydrated molecules of the base serving as the catalyst (KOH). The presence of a one-electron route of nucleophilic addition to the triple bond can also contribute to the manifestation of this effect.156 This reaction can occur by a chain mechanism and be inhibited by admixtures that quench radical and non-radical intermediates. Such a mechanism is highly probable for super- basic systems in which highly reactive, dehydrated anions are especially prone to electron transfer.This is also consistent with the reported 156 approximate correlation between the reaction rate and the ionisation potentials of the initial pyrroles. The presumed existence of a one-electron route in the vinylation reaction of pyrroles has been confirmed in later studies.162 ± 164 3. Cyclisation Cyclisation of pyrroles linked to aromatic five-membered hetero- cycles leads to difficultly accessible polycyclic systems. Porphyrin-like macrocycles 207 ± 210 have been obtained by the reaction of 2,5-bis(5-formyl-4-n-propyl-2-pyrrolyl)furan 204 or dipyrroledicarbaldehyde 205 with 2,5-bis(5-formyl-4-n-propyl- 2-pyrrolyl)thiophene 206.165 Prn HOC O NH HN 204 Prn + HOC S NH 206 Prn S NH N O Prn 207 Prn S NH + N S Prn 208 Prn Prn CHO HOC NH HN 205 Prn S NH N Prn 209 Prn N NH + HN N Prn 210 Oxidative photocyclisation of isomeric 2,3-dithienyl-5-phe- nylpyrroles 39a ± d resulted in a series of isomeric dithienoindoles 211a ± d in 28%± 47% yields.66 S hn 7H2 Ph Me N S 211a (40%) 39a 475 Prn + CHOPrnCHO NH Prn N + HN PrnPrn N HN Prn 206 Prn N + HN Prn Prn Prn S S Ph NMe476 S hn 7H2 S Ph NMe 39b S hn 7H2 Ph NMe S 39c S S hn 7H2 Ph NMe 39d Photochemical electrocyclisation of 3-(2-furyl)- and 3-(2- thienyl)pyrroles 84b,c with subsequent dehydrogenation with 2,3-dichloro-5,6-dicyanoquinone (DDQ) leads to indole deriva- tives 212a,b.81 Irradiation was carried out with a high-pressure mercury lamp at room temperature for 12 h in cyclohexane.X COR 1. hn 2. DDQ NMe (CH2)n 84b,c n X Compound 212 ab 32 OS Linear tetrapyrrole intermediates (213) form complexes (214) with transition metal salts (M=Ni, Co; X=Cl, Br, ClO4) upon heating in o-dichlorobenzene (100 8C, 14 h). These complexes are oxidised by atmospheric oxygen to corrole-like macrocycles 215 containing the bipyrrole fragment. Subsequent reduction of the latter under drastic conditions (100 8C, 100 atm) leads to corrins. Their structural element forms the basis of numerous vitally important compounds, including vitamin B12.166, 167 R1 R1 R2 R2 HN NH X7 HN NH + R2 R2 213 R1 R1 R1 R1 R2 R2 N+ + N N X7 M N N R2 R2 215 R1 R1 R1, R2 =Me, Et; M=Ni, Co; X=Cl, Br, ClO4 .S E Korostova, A I Mikhaleva, B A Trofimov S 4. Hydrolysis, decarboxylation, oxidation and other reactions of substituents Ph S Me N 211b (47%) Hetaryl-substituted pyrroles often contain alkoxycarbonyl groups. These can be eliminated by hydrolysis and decarboxy- lation (200 8C) (see Section II.7, synthesis of compounds 1a, 7a, 128, 129).101 3-(2-Thienyl)pyrrole (217) was obtained in this way from the diester 91.85 S HO2C MeO2C Ph S S S HO7 200 8C Me N CO2H CO2Me S NH NH 91 HN 211c (28%) 217 (67%) S Ph Alkaline hydrolysis of the ester 218 leading to thienylpyrrole 219 has been described.76 MeO2C S NMe 211d (33%) NaOH, DMSO 80 8C COC6H4OMe-4 N S 218 C6H11-cyclo COC6H4OMe-4 N S X COR 219 C6H11-cyclo NMe (H2C)n 212a,b When di(pyrrolyl)thiophene 44 with four ethoxycarbonyl groups in the a- and b-positions of the pyrrole rings was boiled in methanolic NaOH, only the a-positions were selectively liberated with the formation of di(pyrrolyl)thiophene 220.68 Me R R Me Yield of 212 (%) R NaOH, MeOH, D R R S 73 78 Ph 2-thienyl NH HN 44 Me R R Me S NH HN 220 R=CO2Et.R1 R1 Esters of 4-(2-thienyl)pyrrole-3-carboxylic acids 222 were synthesised by decarboxylation of 3-(2-thienyl)-2,4-pyrroledi- carboxylic acid derivatives 221.168 R2 R2 N CO2R2 CO2R2 N + S S X7 M 200 8C 7CO2 N N R1 R1 HO2C221 222 HN NH R2 R2 R1=H, Alk; R2=Alk.R1 R1 214 R1 R1 R2 R2 Solvolysis of thienylpyrrole 149 results in its deacetylation and formation of 4-hydroxy-5-methoxycarbonyl-2-(2-thienyl)pyrrole (223).114 Its methylation with dimethyl sulfate gives pyrrole 224.114 N X7 OH OAc M N N a b R2 CO2Me R2 CO2Me S S 223 HN 149 NH 216 R1 R1Bipyrroles, furyl- and thienylpyrroles OMe CO2Me NH S 224 (95%) (a) 5N NaOMe, MeOH, 3 h; (b) Me2SO4 , NaH, THF. Pyrrole-2-carbaldehydes 225, 226 form prodigiosin analogues 227, 228 upon reaction with the pyrrole 224 (yields up to 95%).114 C5H11-n OMe C5H11-n 224 Me Me OHC N S CH 227 HN NH 225 Me Me MeO MeO2C MeO2C 224 OHC CO2Me CO2Me N S CH 228 NH 226 HN The N-sulfonyl protective groups in dipyrrolylthiophenes 42a,b are removed by boiling with methanolic NaOH.68 NaOH, MeOH S S N N HN HN 42a,b SO2R SO2R R=Me (a), Ph (b).Treatment of 2-(2-thienyl)-5-trifluoroacetyl-1-vinylpyrrole 195e with boiling ethanolic NaOH gave 2-(2-thienyl)pyrrole- carboxylic acid (yield 88%), which is decarboxylated to vinyl- pyrrole 8a upon heating (150 ± 175 8C).169 1. NaOH, EtOH, H2O 2. HCl COCF3 N S 195e 150 ± 175 8C CO2H N S N S 8a a-Formyl derivative of 2-(5-chloro-2-thienyl)pyrrole 229 was synthesised in 45% yield by the oxidation of pyrrole 37 withMnO2 in dioxane.65 MnO2 dioxane, D CHO Cl Me Cl S S NH 37 N 229 H The aldehyde 229 reacts with 2,4-pyrrolophane 230 in ethanol in the presence of concentrated HCl (8 8C, 15 min; 0 8C, 24 h) to give pyrrolophane 231.65 S Cl HCl, EtOH HN 229+ N HN230 CHO 231 5.Reactions of the N-vinyl group Systematic studies of the reactivity of 2-(2-furyl)- and 2-(thienyl)- 1-vinylpyrroles became possible after the development of a 477 convenient and efficient method for their synthesis.4 Before that, there were virtually no data on the N-vinylpyrroles linked to five- membered aromatic heterocycles. Studies of reactions involving the N-vinyl groups of furyl- and thienylpyrroles open wide possibilities for the synthesis of new hetaryl-substituted pyrroles.a. Hydrogenation Hydrogenation of substituted pyrroles is a rather well-studied reaction,1, 2 which allows one to synthesise pyrrolines and pyrro- lidines in high yields. Platinum, ruthenium and nickel catalysts are most often used for catalytic hydrogenation of pyrroles.1, 2, 170 Depending on the nature of the substituent and its position in the ring, it is possible to bring about selective hydrogenation of either the pyrrole ring or the side chain.1, 2 Electron-acceptor substitu- ents in the a- and b-positions of the ring impede its reduction, whereas substituents at the nitrogen atom facilitate this process.1, 2 Hydrogenation of 2-(2-furyl)-1-vinyl- and 2-(2-thienyl)-1- vinylpyrroles in the presence of catalysts such as Raney nickel, platinum black, palladium chloride was first studied by Koros- tova et al.148 It was shown that hydrogenation on Raney nickel (80 8C, 60 atm) results in the reduction of the vinyl group in 2-(2- furyl)-1-vinylpyrrole (2a) to the ethyl group which occurs as easily and selectively as in 2-alkyl(aryl)-1-vinylpyrroles,171 giving 1- ethyl-2-(2-furyl)pyrrole 194c in 94% yield.148 H2 X N X NEt 194a,c 2a, 8a X = O (2a, 194c); S (8a, 194a).In the hydrogenation over Raney nickel, a decrease in the hydrogen pressure to 10 ± 12 atm (all other conditions remaining the same) leads to a sharp deceleration of the reaction and a decrease in the yield of ethylpyrrole 194c to 65%. In the hydro- genation over palladium chloride (15 mass.% with respect to vinylpyrrole, 80 8C, 40 atm) the yield of ethylpyrrole 194c amounts to 75%.In contrast with the data reported by Lunn,172 who succeeded in reducing pyrrole to pyrrolidine in the presence of a nickel ± aluminium alloy, the pyrrole and furan rings of compound 2a remain intact under the described conditions.148 The presence of the thiophene ring complicates hydrogenation of 2-(2-thienyl)-1-vinylpyrrole (8a) over nickel catalyst (50 8C, 48 atm) due to poisoning of the catalyst. Only the use of a very large amount of the catalyst (50 mass.% with respect to vinyl- pyrrole) has allowed preparation of ethylpyrrole 194a in 46% yield; in this case, hydrogenolysis resulted in the formation of low- boiling products. When PdCl2 was used at room temperature, the yield of ethylpyrrole 194a was as low as 7%.In the presence of palladium black (80 8C, 40 atm), 2-(2-thienyl)-1-vinylpyrrole is not hydrogenated. b. Addition of alcohols Electrophilic addition of alcohols to 2-(2-furyl)- and -2-(2- thienyl)-1-vinylpyrroles was studied in the example of reaction with propargyl alcohol in the presence of perfluorobutyric acid (PFBA).148, 173, 174 This reaction is conducted at 96 8C. Higher temperature and longer reaction time enhance resinification. The yield of adducts 232a ± c depends on the catalyst's concentration and the ratio of reagents: the optimum conditions are 1.5-fold excess of propargyl alcohol and 1.5% ± 2% of PFBA relative to the substrate. R C3F7CO2H +HC CCH2OH N X2a, 8a, 8d478 R N X Me 232a ± c OCH2C R=H:X=O(232a, 55%), S (232b, 71%); R=Me, X=S (232c, 56%).c. Addition of thiols Free-radical thiylation of 2-(2-furyl)- and 2-(2-thienyl)-1-vinyl- pyrroles initiated by azobis(isobutyronitrile) (AIBN) gives anti- Markovnikoff formation of solely b-adducts 233.148 R1 AIBN +R2SH N X2a, 8a, d Substrate R1 X 2a 8a 8d 8d HHPrn Prn OSSS At 70 ± 80 8C, the reaction is completed within 18 ± 25 h (the yield of adducts 233a ± d is 80%± 95%). Under comparable conditions without initiator, the yield drops to 15%± 20%.148 This reaction has been successfully used for purification of 2-(2- thienyl)pyrroles from their 1-vinyl derivatives.43 d. Hydrosilylation Kopylova et al.175 performed hydrosilylation of 2-(2-furyl)- and 2-(2-thienyl)-1-vinylpyrroles in the presence of different catalysts; they also used various hydrosilylation agents.The highest yields of silyl derivatives were reached upon hydrosilylation with triethylsilane at an equimolar ratio of reagents in the presence of 0.05 M solution of H2PtCl6 . 6H2O in THF (140 8C, 5 h). The reaction proceeded regiospecifically with the formation of b-adducts 234a ± e. (The yields were calculated relative to the consumed vinylpyrrole. 175) R1 +HSiEt3 R2 N X R1 R2 N X SiEt3 234a ± e Compound 234 R1 X abcde HHMe Et H OSSSS The Wilkinson complex [(Ph3P)3RhCl] also catalyses this reaction.Longer reaction time and the use of polar solvents (THF, dioxane) favour an increase in yields of the adducts. No positive results have been reported for the reactions involving trichloro-, triethoxy- and alkylchlorosilanes. Only upon reaction of methyldichlorosilane with 2-(2-thienyl)-1-vinylpyrrole, the CH R1 N X233a ± d Product R2 233a 233b 233c 233d Et Et Et Prn H2PtCl6 THF Yield (%) R2 40 70 65 44 10 HHHHCOCF3 S E Korostova, A I Mikhaleva, B A Trofimov formation of the corresponding adduct in *7% yield was observed. In the absence of catalysts, the reaction does not take place under UV-initiation. The yield of adducts depends not only on the nature of the silane but also on the substituents in the pyrrole ring.Thus the yield of the pyrrole 234b was 70%; upon introduction of an electron-acceptor substituent (R2=COCF3) into the pyrrole ring, the yield of the corresponding pyrrole 234e drops to 10%.175 e. Cyclisation 7,8-Dicyano-5-methoxy-3-(2-thienyl)-5,6-dihydroindolizine (235) was synthesised from 2-(2-thienyl)-1-vinylpyrrole (8a), tetra- cyanoethylene and methanol.176 CN NC MeOH + N S CN NC 8a SR2 Yield (%) N S CN CN 7HCN N S 84 78 95 80 235 MeO CN CN CN NC f. 2-(2-Furyl)-1-vinylpyrroles as protected pyrroles Protection of the NH group of the pyrrole fragment is often the key stage in the synthesis of complex pyrrole derivatives. In this case, not only the acidity of the NH group is temporarily blocked, but the reactivity of different positions of the pyrrole ring is substantially changed due to steric and electronic effects of the protective groups.The role of protective group is played by, e.g., trialkylsilyl, tosyl, pyridylethyl, phenylsulfonyl or N-b-alkyl- thioethyl groups. The N-vinyl group may also be rather success- fully used for the protection of the NH group.51, 95, 177 ± 180 In this case, deprotection can be effected by acid hydrolysis in the presence of hydroxylamine hydrochloride 181 or by mercuration with mercuric acetate followed by treatment with sodium borohy- dride.51 2-(2-Furyl)- and 2-(5-methyl-2-furyl)-1-vinylpyrroles have thus been transformed into the corresponding 1H-pyrroles in 93% and 83% yields, respectively (relative to the consumed vinylpyrrole).182 Hg(OAc)2 NaBH4 R R N O N O HgOAc 236 AcO R O NH R=H, Me.Mercuration is carried out at room temperature until disap- pearance of the initial 1-vinylpyrrole with subsequent in situ reduction of the intermediate mercuric acetate 236 with sodium borohydride. Despite complete disappearance of the starting vinylpyrrole in the mercuration stage, it appears again in small amount after the reduction with sodium borohydride. It is presumed that the intermediate 236 is preceded by a complex of vinylpyrrole with the mercury cation, the reduction of which leads to regeneration of 1-vinylpyrrole.182 g. Polymerisation Polymerisation of furyl- and thienylpyrroles and their vinyl derivatives has been little studied so far.Electrochemical polymer-Bipyrroles, furyl- and thienylpyrroles isation of 2-(2-thienyl)pyrrole yielded poly(thienylpyrrole), appa- rently with alternating thienyl and pyrrole rings.183 NH S S n NH Under the conditions of radical initiation, i.e., under the action of azobis(isobutyronitrile), 2-(2-thienyl)-1-vinylpyrrole forms homopolymers 237 with a molecular mass of 3700 Da.184 CH CH2 n AIBN S N N S 237 The formation of polymers with involvement of only the N-vinyl group was proved by spectral data (IR, 1H NMR) and turbidimetric titration data in different solvent ± precipitant systems. It was suggested 184 that poly(1-vinylpyrroles) 237 form complexes upon contact with atmospheric oxygen; this suggestion is supported by intense EPR signals in polymers following their prolonged exposure in air.Free-radical copolymerisation of 2-(2-thienyl)-1-vinylpyrrole 8a with vinylpyrrolidone 238 yielded a copolymer 239 enriched in the vinylpyrrolidone units.185, 186 The molecular mass and the yield of the copolymer depend directly on the concentration of vinylpyrrolidone. AIBN + O N N S 8a CH CH2 CH CH2 238 CH CH CH2 CH2 m n N N S O 239 IV. Prospects for the use The chemistry of pyrroles linked to five-membered aromatic heterocycles is still in its infancy. Nonetheless, the currently available data on their properties allow one to consider them and their derivatives as promising compounds for medicine, agriculture and some industrial fields.This concerns primarily the 1-vinyl derivatives of furyl- and thienylpyrroles. Reactions of electrophilic substitution in the pyrrole ring and addition reactions to the vinyl group occur highly selectively making it possible to use these compounds as building blocks for the synthesis of various derivatives. The potentialities of furyl- and thienylpyrroles are still far from being fully disclosed. Poly(furylpyrroles) and poly(thienylpyrroles) with regularly alternating rings can differ appreciably in their electrophysical properties 184 from polypyr- roles � organic metals, which are already applied in novel technologies.187 ± 190 Indolizines with thienyl substituents have been recommended as efficient photosensitisers.176 The antiviral activity of a series of 2-(2-furyl)- and 2-(2- thienyl)-1-vinylpyrroles has been investigated.2-(2-Thienyl)-1- vinylpyrrole was found to be moderately active against the herpes virus, whereas 3-ethyl-2-(2-thienyl)-1-vinylpyrrole proved to be active against the pox virus.26 It should be pointed out that furylvinyl- and thienylvinylpyrroles are as a rule moderately or weakly toxic compounds 23, 24, 26 (Table 3). 479 Table 3. Toxicity of 1-vinyl-2-(2-hetaryl)pyrroles, some of their precursors and derivatives. Compound LD50 /mg kg71 (see a) 760 1300 270 260 630 30 830 1370 620 1500 3900 1870 270 3-Ethyl-2-(2-furyl)-1-vinylpyrrole 2-(2-Furyl)-3-isopropyl-1-vinylpyrrole 3-Butyl-2-(2-furyl)-1-vinylpyrrole 2-(2-Thienyl)pyrrole 2-(2-Thienyl)-1-vinylpyrrole 3-Methyl-2-(2-thienyl)-1-vinylpyrrole 3-Ethyl-2-(2-thienyl)-1-vinylpyrrole 3-Propyl-2-(2-thienyl)-1-vinylpyrrole 3-Butyl-2-(2-thienyl)-1-vinylpyrrole 3-Hexyl-2-(2-thienyl)-1-vinylpyrrole 2-(2-Thienyl)-5-trifluoroacetylpyrrole 2-(2-Thienyl)-5-trifluoroacetyl-1-vinylpyrrole 2-(2-Thienyl)-1-vinyl-5-pyrrolecarboxylic acid a Approximate values.The activity of a series of 2-(2-furyl)- and 2-(2-thienyl)pyrroles towards 11 strains of different bacterial and fungal species has been assayed. 2-(2-Thienyl)-1-vinylpyrrole proved to be the most active against staphyllococci and candidas as well as against the anthracoid bacilli.26 Investigation into the repellent and insecticidal activity of furylvinyl- and thienylvinylpyrroles has revealed that 2-(2-furyl)- 1-vinylpyrrole is a highly active and selective repellent for forest pests (black coniferous longhorn beetle) and is non-toxic for warm-blooded animals and the ichthiofauna of rivers; 3-ethyl-2- (2-thienyl)-1-vinylpyrrole was also found to manifest weak repellent activity.4, 24, 26 The same vinylpyrroles proved to be moderately toxic for the larvae of the coniferous longhorn beetle.4, 26 The assay of insecticidal activity of a series of 2-(2-furyl)- and 2-(2-thienyl)pyrroles has revealed moderate activity of 3-propyl- 2-(2-thienyl)-1-vinylpyrrole against sugar-beet root aphid and high activity of 2-(2-thienyl)-, 3-ethyl- and 3-propyl-1-vinylpyr- roles against larvae of the domestic fly.26 3-Propyl-2-(2-thienyl)-1- vinylpyrrole possesses high herbicidal activity, while 2-(2-thienyl)- 1-vinylpyrrole and its copolymer with N-vinylpyrrolidone exhibit the growth-regulating activity.26, 186, 191 A number of 4-cyano-3-furyl- and 3-thienylpyrroles were recommended as starting compounds for the synthesis of medic- inal substances and pesticides,82 while esters of 4-(2-thienyl)-3- pyrrolecarboxylic acids seem to be appropriate drugs for the therapy of cardiac insufficiency.168 V. Conclusion The data discussed in this review demonstrate appreciable progress in the field of the synthesis of pyrroles linked to other heterocycles, viz., bipyrroles, furyl-, thienyl- and selenienyl- pyrroles and their derivatives. Broad synthetic potential of hetarylpyrroles and especially of their N-vinyl derivatives in various reactions, such as electrophilic substitution, nucleophilic addition to the triple bond, cyclisation, hydrogenation, thiylation, hydrosilylation, polymerisation, as well as useful properties inherent in certain compounds of this series open up broad prospects for their use in fine organic synthesis, medicine, agriculture and some technological fields.The preparation of this review was supported by the Russian Foundation for Basic Research (Project No. 96-03-33263a).480 References 1. A Gossauer Die Chemie der Pyrrole (Berlin: Springer, 1974) 2. R A Jones, G P Bean The Chemistry of Pyrroles (New York: Academic Press, 1977) 3.A H Jackson, in Comprehensive Organic Chemistry Vol. 4 (Eds D Barton,W D Ollis) (New York: Pergamon Press, 1979) 4. B A Trofimov, A I Mikhaleva N-Vinylpirroly (N-Vinilpyrroles) (Novosibirsk: Nauka, 1984) 5. B A Trofimov Usp. Khim. 58 1703 (1989) [Russ. Chem. Rev. 58 967 (1989)] 6. B A Trofimov, A I Mikhaleva, L V Morozova Usp. Khim. 54 1034 (1985) [Russ. Chem. Rev. 54 609 (1985)] 7. E Ya Lukevits (Ed.) Uspekhi Khimii Furana (The Progress in the Chemistry of Pyrrole) (Riga: Zinatne, 1978) 8. P Bauerle, T Fischer, B Bidlingmeier, A Stabel, J P Rabe Angew. Chem. 107 335 (1995) 9. L I Belen'kii (Ed.) Khimiya Organicheskikh Soedinenii Sery (The Chemistry of Organic Compounds of Sulfur) (Moscow: Khimiya, 1988) 10. P I Buchin, A E Lipkin Proizvodnye Tiofena i Bitiofena kak Perspektivnye Antiseptiki Novoi Gruppy (Thiophene and Bithiophene Derivatives as Perspective Antiseptics of New Group) (Saratov: Saratov State University, 1974) 11.V P Litvinov, V D Dyachenko Usp. Khim. 66 1025 (1997) [Russ. Chem. Rev. 66 923 (1997)] 12. B A Trofimov, N I Golovanova, N I Shergina, A I Mikhaleva, S E Korostova, A N Vasil'ev Spektry Pogloshcheniya Proizvodnykh 14. B A Trofimov, A I Mikhaleva Khim. Geterotsikl. Soedin. 1299 Pirrola i 1-Vinilpirrola v Infrakrasnoi, Ul'trafioletovoi i Vidimoi Oblastyakh. Atlas Spektrov Aromaticheskikh i Geterotsiklicheskikh Soedinenii (Absorption Spectra of Pyrrole and 1-Vinylpyrrole Deriv- atives in Infrared, Ultraviolet and Visible Regions. Atlas of Spectra of Aromatic and Heterocyclic Compounds) No.21 (Ed. V A Koptyug) (Novosibirsk: Novosibirsk Institute of Organic Chemistry, Siberian Branch of Academy of Sciences of the USSR, 1981) 13. B A Trofimov, L N Sobenina, A I Mikhaleva Organicheskaya Khimiya, T. 7 (Itogi Nauki i Tekhniki) [Organic Chemistry, Vol. 7 (Advances in Science and Engineering Series)] (Moscow: Izd. VINITI, 1987) (1980) a 15. L N Sobenina, A I Mikhaleva, B A Trofimov Usp. Khim. 58 275 (1989) [Russ. Chem. Rev. 58 163 (1989)] 28. M A Yurovskaya, V V Druzhinina, M A Tyurekhodzhaeva, 16. B A Trofimov Adv. Heterocycl. Chem. 51 177 (1990) 17. G P Bean, in The Chemistry of Heterocyclic Compounds Vol. 48 (Ed. R A Jones) (New York: Wiley, 1990) Part 1, p. 105 18.B A Trofimov, in The Chemistry of Heterocyclic Compounds Vol. 48 (Ed. R A Jones) (New York: Wiley, 1992) Part 2, p. 131 19. B A Trofimov, A I Mikhaleva Heterocycles 37 1193 (1994) 20. B A Trofimov Phosphorus Sulfur Silicon Relat. Elem. 95 ± 96 145 (1994) 21. M V Sigalov, B A Trofimov Zh. Org. Khim. 31 801 (1995) b 22. R J Tedeschi, in Encyclopedia of Physical Science and Technology Vol. 1 (New York: Academic Press, 1992) p. 27 23. A I Mikhaleva, Doctoral Thesis in Chemical Sciences, Irkutsk Institute of Organic Chemistry, Siberian Branch of Russian Academy of Sciences, Irkutsk, 1984 24. S E Korostova, Doctoral Thesis in Chemical Sciences, Irkutsk Institute of Organic Chemistry, Siberian Branch of Russian Academy of Sciences, Irkutsk, 1993 25.M V Sigalov, Doctoral Thesis in Chemical Sciences, Irkutsk Institute of Organic Chemistry, Siberian Branch of Russian Academy of Sciences, Irkutsk, 1994 26. R N Nesterenko, Candidate Thesis in Chemical Sciences, Irkutsk Institute of Organic Chemistry, Siberian Branch of Russian Academy of Sciences, Irkutsk, 1985 27. E Yu Shmidt, Candidate Thesis in Chemical Sciences, Irkutsk Institute of Organic Chemistry, Sibean Branch of Russian Academy of Sciences, Irkutsk, 1988 Yu G Bundel' Khim. Geterotsikl. Soedin. 69 (1984) a 29. M A Yurovskaya, A Z Afanas'ev, Yu G Bundel' Khim. Geterotsikl. Soedin. 1077 (1984) a S E Korostova, A I Mikhaleva, B A Trofimov 30. T N Borisova, A V Varlamov, N D Sergeeva, A T Soldatenkov, O V Astakhov, N S Prostakov Khim.Geterotsikl. Soedin. 973 (1987) a 31. A V Varlamov, T N Borisova, A E Aliev, I A Stazharova, A A Sinitsyna, E A Sakhnova Khim. Geterotsikl. Soedin. 681 (1993) a 32. A V Varlamov, T N Borisova, L G Voskresenskii Khim. Geterotsikl. Soedin. 136 (1995) a 33. S E Korostova, A I Mikhaleva, E E Sirotkina Izv. Akad. Nauk SSSR, Ser. Khim. 1448 (1989) c 34. B A Trofimov, A I Mikhaleva, S E Korostova, L N Balabanova, in The Vth International Congress on Heterocyclic Chemistry (Abstracts 35. B A Trofimov, A I Mikhaleva, R I Polovnikova, R N Nesterenko, of Reports), Ljubljana, 1975 A 1, p. 63 F G Trofimova Zh. Org. Khim. 14 2628 (1978) b 36. B A Trofimov, A I Mikhaleva, R N Nesterenko, A N Vasil'ev, A S Nakhmanovich, M G Voronkov Khim.Geterotsikl. Soedin. 1136 (1977) a 37. B A Trofimov, A I Mikhaleva, R I Polovnikova, S E Korostova, R N Nesterenko, N I Golovanova, V K Voronov Khim. Geterotsikl. Soedin. 1058 (1981) a 38. B A Trofimov, S E Korostova, L N Balabanova, A I Mikhaleva Zh. Org. Khim. 14 1733 (1978) b 39. USSR P. 601 282; Byull. Izobret. (13) 79 (1978) 40. S E Korostova, S G Shevchenko, B A Trofimov Zh. Org. Khim. 30 905 (1994) b 41. S E Korostova, A I Mikhaleva, L N Sobenina, S G Shevchenko, V V Shcherbakov Khim. Geterotsikl. Soedin. 1501 (1985) a 42. B A Trofimov, A I Mikhaleva, S E Korostova, T I Vakul'skaya, I S Poguda,M G Voronkov Zh. Prikl. Khim. 41 117 (1978) d 43. S E Korostova, A I Mikhaleva, R N Nesterenko, N V Maznaya, V K Voronov, N M Borodina Zh.Org. Khim. 21 406 (1985) b 44. B A Trofimov, A I Mikhaleva, in Sintez i Reaktsionnaya Sposobnost' Organicheskikh Soedinenii Sery (Tez. Dokl. XVII Vsesoyuz. Konf.), Tbilisi, 1989 [Synthesis and Reactivity of Organic Compounds of Sulfur (Abstracts of Reports of the XVIIth All-Union Conference), Tbilisi, 1989] p. 24 45. B A Trofimov, A I Mikhaleva, A N Vasil'ev,M V Sigalov Khim. Geterotsikl. Soedin. 54 (1978) a 46. B A Trofimov, C E Korostova, L N Sobenina, A I Mikhaleva, V M Bzhezovskii,M V Sigalov, A A Atavin Khim. Geterotsikl. Soedin. 193 (1982) a 47. G A Pinna,M A Pirisi, G Paglietti J. Chem. Res. (S) 210 (1993) 48. A I Mikhaleva, R N Nesterenko, A M Vasil'tsov, G A Kalabin, E N Deryagina, N A Korchevin, N I Golovanova Khim.Geterotsikl. Soedin. 708 (1992) a 49. B A Trofimov, A I Mikhaleva, A N Vasil'ev, V A Lopyrev, I A Titova, V M Bzhezovskii,M V Sigalov Khim. Geterotsikl. Soedin. 1422 (1981) a 50. C E Korostova, R N Nesterenko, A I Mikhaleva, R I Polovnikova, N I Golovanova Khim. Geterotsikl. Soedin. 901 (1989) a 51. S Gonzalez, R Greenhouse, R Tallabs, J M Muchowski Can. J. Chem. 61 1697 (1983) 52. USSR P. 1 095 593; Byull. Izobret. (6) 196 (1994) 53. O A Tarasova, S E Korostova, A I Mikhaleva, B A Trofimov, in Reaktiv 94 (Tez. Dokl. VII Soveshchaniya po Khimicheskim Reaktivam), Ufa, 1994 [Agent 94 (Abstracts of Reports of the VIIth Meeting on Chemical Agents), Ufa, 1994] p. 53 54. O A Tarasova, S E Korostova, A I Mikhaleva, L N Sobenina, R N Nesterenko, S G Shevchenko, B A Trofimov Zh.Org. Khim. 30 810 (1994) b 55. S E Korostova, S G Shevchenko, M V Sigalov Khim. Geterotsikl. Soedin. 1371 (1991) a 56. B A Trofimov, S E Korostova, A I Mikhaleva, L N Sobenina, A N Vasil'ev, R N Nesterenko Khim. Geterotsikl. Soedin. 273 (1983) a 57. S E Korostova, S G Shevchenko, M V Sigalov Khim. Geterotsikl. Soedin. 187 (1991) a 58. D Dhanak, C B Reese, S Romana, G Zappia J. Chem. Soc., Chem. Commun. 903 (1986) 59. S E Korostova, A I Mikhaleva, L N Sobenina, in Tez. Dokl. Vsesoyuz. Simpoziuma po Organicheskomu Sintezu, Moskva, 1984 (Abstracts of Reports of the All-Union Symposium on Organic Synthesis, Moscow, 1984) p. 110Bipyrroles, furyl- and thienylpyrroles 60. B A Trofimov, S E Korostova, A I Mikhaleva, L N Sobenina, V V Shcherbakov, M V Sigalov Khim.Geterotsikl. Soedin. 276 (1983) a 61. S E Korostova, A I Mikhaleva, B A Trofimov, S G Shevchenko, M V Sigalov Khim. Geterotsikl. Soedin. 485 (1992) a 62. B A Merril, E Le Goff J. Org Chem. 55 2904 (1990) 63. Jpn. P. 77 100 496; Chem. Abstr. 88 50 934 (1978) 64. Ning Fugiang, Hua Wenting Beitszin Dasyue Syuebao 28 185 (1992); Ref. Zh. Khim. 19 Zh 189 (1992) 65. H Berner, G Schulz, H Reinshagen Monatsh. Chem. 108 285 (1977) 66. D M Perrine, J Kagan, D-B Huang, K Zeng, B-K Teo J. Org. Chem. 52 2213 (1987) 67. H Stetter, J Krasselt J. Heterocycl. Chem. 14 573 (1977) 68. B A Merril, E Le Goff, in The 197th American Chemical Society National Meeting (Abstracts of Reports), Dallas, TX, 1989 p.732 69. P Messinger, C Kunick Synthesis 213 (1986) 70. USSR P. 251 581; Byull. Izobret. (28) 20 (1969) 71. V I Shvedov, I A Kharizomenova, A N Grinev Khim.-Farm. Zh. 3 (12) 12 (1969) e 72. F Ledl, F Schleicher Angew. Chem. 102 597 (1990) 73. Yu N Romashin, Al' Mokhana Nadim, in Tez. Dokl. II Konf. Molodykh Uchenykh-Khimikov, Donetsk, 1990 (Abstracts of Reports of the Second Conference of Young Chemist Scientists, Donetsk, 1990) p. 167 74. V L Sorokin, Yu N Romashin, Al' Mokhana Nadim, Azzuz Alib, O G Kullinkovich, in Khimiya Azotsoderzhashchikh Geterotsikli- cheskikh Soedinenii (Tez. Dokl. V Vsesoyuz. Konf.),Chernogolovka, 1991 [The Chemistry of Nitrogen-Containing Heterocyclic Compounds (Abstracts of Reports of the Fifth All-Union Conference), Chernogolovka, 1991] Part 1, p.267 75. R Huisgen, H Gotthardt, H O Bayer, F Schaefer Chem. Ber. 103 2611 (1970) 76. J W Lown, K Matsumoto Can. J. Chem. 48 2215 (1970) 77. J W Lown, K Matsumoto Can. J. Chem. 48 3399 (1970) 78. K E Schulte, J Reisch, H Walker Chem. Ber. 98 98 (1965) 79. C L Buchwald, M W Wannamaker, B T Watson J. Am. Chem. Soc. 111 776 (1989) 80. F H Carre, R J P Corrin, G Bolin, J J E Morean, C Vernhet Organometallics 12 2478 (1993) 81. J Moskal, A M van Leusen J. Org. Chem. 51 4131 (1986) 82. Jpn. P. 61 200 984; Ref. Zh. Khim. 23 O 82P (1987) 83. Jpn. P. 76 128 963; Chem. Abstr. 87 5797 (1977) 84. Jpn. P. 5 885 861; Chem. Abstr. 99 12 229 (1983) 85. K Sakai,M Suzuki,K Numami,N Yoneda, Y Onoda, Y Iwasawa Chem.Pharm. Bull. 28 2384 (1980) 86. C Yuan,W Huang Synthesis 473 (1993) 87. K Matsumoto,M Suzuki, Y Ozaki,M Miyoshi Agrich. Biol. Chem. 40 2271 (1976) 88. O E Nasakin, V V Alekseev, V Yu Promonenkov Khim. Geterotsikl. Soedin. 402 (1981) a 89. O E Nasakin, V V Alekseev, P B Terent'ev, A Kh Bulai, M Yu Zablotskaya Khim. Geterotsikl. Soedin. 1062 (1983) a 90. O E Nasakin, V V Alekseev, A Kh Bulai, P B Terent'ev, in Khimiya Geterotsiklicheskikh Soedinenii (Tez. Dokl. III Konf.) Rostov-na-Donu, 1983 [The Chemistry of Heterocyclic Compounds (Abstracts of Reports of the Third Conference), Rostov on Don, 1983] p. 44 91. O E Nasakin, V V Alekseev, P B Terent'ev, A Kh Bulai, M Yu Zablotskaya Khim. Geterotsikl. Soedin. 1067 (1983) a 92.USSR P. 979 339; Byull. Izobret. (45) 99 (1982) 93. N Chatorri, T Hanafusa Tetrahedron Lett. 27 4201 (1986) 94. T Kaufmann, H Lexy Chem. Ber. 114 3674 (1981) 95. B L Bray, P H Mathies, R Naef, D R Solas, T T Tidwell, D R Artis, J M Muchowski J. Org. Chem. 55 6317 (1990) 96. L Groenendaal, H W I Peerlings, J L J van Dongen, E E Havinga, J A J M Vekemans, E W Meijer Am. Chem. Soc., Div. Polym. Chem., Polym. Prepr. 35 194 (1994) 97. N Gjùs, S Gronowitz Acta Chim. Scand. 25 2596 (1971) 98. L Groenendaal, H W I Peerlings, E E Havinga, J A J M Vekemans, E W Meijer Synth. Met. 69 467 (1995) 99. A Minato, K Tamao, T Hayashi, K Suzuki, M Kumada Tetrahedron Lett. 22 5319 (1981) 100. Jpn. P. 5 896 063; Chem. Abstr. 99 139 759 (1983) 481 101.J-P Boukou-Poba,M Farnier, R Guilard Tetrahedron Lett. 1717 (1979) 102. H Rapoport, N Castagnoli J. Am. Chem. Soc. 84 2178 (1962) 103. H Rapoport, K G Holden J. Am. Chem. Soc. 84 635 (1962) 104. H Rapoport, N Castagnoli, K G Holden J. Org. Chem. 29 883 (1964) 105. H Rapoport, J Bordner J. Org. Chem. 29 2727 (1964) 106. J Bordner, H Rapoport J. Org. Chem. 30 3824 (1965) 107. N Engel,W Steglich Angew. Chem. 90 719 (1978) 108. J V Cooney,W E McEwen J. Org. Chem. 46 2570 (1981) 109. F Boberg, K H Garburg, K J Gorlich, E Piepereit, M Ruhr Liebigs Ann. Chem. 239 (1985) 110. S Vega,M S Gil J. Heterocycl. Chem. 28 945 (1991) 111. S Rault,M Cugnon de Sevricourt, M Robba C. R. Hebd. Seances Acad. Sci., Ser. C 284 533 (1977) 112. S Rault,M Cugnon de Sevricourt, M Robba Heterocycles 14 651 (1980) 113.S Rault,M Cugnon de Sevricourt, H El Khashef,M Robba C. R. Hebd. Seances Acad. Sci., Ser. C 290 169 (1980) 114. G Kresze,M Morper, A Bijev Tetrahedron Lett. 2259 (1977) 115. R Grigg, H Q N Gunaratne Tetrahedron 46 1599 (1990) 116. Matoba Katsutude, Itoh Kenicci, Kondo Kakuya, Yamozaki Takao, Nagata Masanori Chem. Pharm. Bull. 29 2442 (1981) 117. D Dolphin, A W Johnson, J Leng, P van den Brock J. Chem. Soc., C 880 (1966) 118. M J Broadhurst, R Grigg, A W Johnson J. Chem. Soc., Perkin Trans. 1 1124 (1972) 119. E Baltazzi, L I Krimen Chem. Rev. 63 511 (1963) 120. J M Patterson Synthesis 281 (1976) 121. Dzh Marino Khim. Geterotsikl. Soedin. 579 (1973) a 122. L I Belen'kii Khim.Geterotsikl. Soedin. 1587 (1980) a 123. I A Abronin, G M Zhidomirov Khim. Geterotsikl. Soedin. 3 (1977) a 124. J A Joule, G F Smith Heterocyclic Chemistry (London: Van Nostrand Reinhold, 1972) 125. V I Ivanskii Khimiya Geterotsiklicheskikh Soedinenii (The Chemistry of Heterocyclic Compounds) (Moscow: Vysshaya Shkola, 1978) 126. M V Sigalov, E Yu Shmidt, B A Trofimov, in Tez. Dokl. IV Vsesoyuz. Konf. po Khimii Azotsoderzhashchikh Soedinenii, Novosibirsk, 1987 (Abstracts of Reports of the Fourth All-Union Conference on the Chemistry of Nitrogen-Containing Compounds, Novosibirsk, 1987) p. 70 127. M V Sigalov, E Yu Shmidt, B A Trofimov Izv. Akad. Nauk SSSR, Ser. Khim. 2136 (1987) c 128. M V Sigalov, E Yu Shmidt, A B Trofimov, B A Trofimov Khim.Geterotsikl. Soedin. 1343 (1989) a 129. M V Sigalov, E Yu Schmidt, A B Trofimov, B A Trofimov J. Org. Chem. 57 3934 (1992) 130. M V Sigalov, A B Trofimov, E Yu Schmidt, B A Trofimov J. Phys. Org. Chem. 6 471 (1993) 131. M V Sigalov, A B Trofimov, E Yu Shmidt, B A Trofimov Khim. Geterotsikl. Soedin. 825 (1993) a 132. M V Sigalov, E Yu Shmidt, A I Mikhaleva, S E Korostova, I M Lazarev, B A Trofimov Khim. Geterotsikl. Soedin. 48 (1993) a 133. M V Sigalov, E Yu Schmidt, B A Trofimov, in The 10th IUPAC Conference on Physical Organic Chemistry (Abstracts of Reports), Haifa, 1990 p. 247 134. J-P Boukou-Poba,M Farnier, R Guilard Can. J. Chem. 59 2962 (1981) 135. S Clementi, G Marino Tetrahedron 25 4599 (1969) 136. S Clementi, G Marino J.Chem. Soc., Perkin Trans 2 71 (1972) 137. A N Kost, V A Budylin Zh. Vses. Khim. O-va im. D I Mendeleeva 22 315 (1977) f 138. K Imuro, T Hanafusa Bull. Chem. Soc. Jpn. 49 1363 (1976) 139. A N Kost, V A Budylin, N N Romanova, A I Matveeva Khim. Geterotsikl. Soedin. 1233 (1981) a 140. D J Chadwick, G D Meakins, C A Rhodes J. Chem. Res. (S) 42 (1980) 141. D J Chadwick, S T Hodgson J. Chem. Soc., Perkin Trans 1 93 (1983) 142. M W Majchrzak, G Simchen Tetrahedron 42 1299 (1986) 143. N K Genkina, V N Eraksina, N N Suvorov Zh. Org. Khim. 16 2154 (1980) b482 144. B A Trofimov, A I Mikhaleva, S E Korostova, L N Sobenina, A N Vasil'ev, L V Balashenko Zh. Org. Khim. 15 2042 (1979) b 145. S E Korostova, A I Mikhaleva, L N Sobenina, S G Shevchenko, I M Karataeva, B A Trofimov Khim.Geterotsikl. Soedin. 48 (1980) a 146. B A Trofimov, S E Korostova, A I Mikhaleva, R N Nesterenko, M V Sigalov, V K Voronov, R I Polovnikova Zh. Org. Khim. 18 894 (1982) b 147. S E Korostova, A I Mikhaleva, R N Nesterenko, R I Polovnikova, V K Voronov, B A Trofimov, in Tez. Dokl. IV Vsesoyuz. Konf. po Khimii Ftororganicheskikh Soedinenii, Tashkent, 1982 (Abstracts of Reports of Fourth All-Union Conference on the Chemistry of Organofluorine Compounds, Tashkent,1982) p. 218 148. S E Korostova, R N Nesterenko, A I Mikhaleva, S G Shevchenko, G A Kalabin, R I Polovnikova Khim. Geterotsikl. Soedin. 337 (1991) a 149. M Hojo,R Masudo, Ho Shioda, S Matsuno Chem. Lett. 499 (1976) 150. S I Yakimovich, L A Kayukova Zh.Org. Khim. 14 2496 (1978) b 151. N V Moskalev, V D Filimonov, E E Sirotkina Khim. Geterotsikl. Soedin. 1066 (1988) a 152. A I Mikhaleva, B A Trofimov, S E Korostova, A N Vasil'ev, L N Balabanova Izv. Sib. Otd. Akad. Nauk SSSR, Ser. Khim. (4) 105 (1979) 153. B A Trofimov, R N Nesterenko, A I Mikhaleva, A B Shapiro, I A Aliev, I V Yakovleva, G A Kalabin Khim. Geterotsikl. Soedin. 481 (1986) a 154. G G Skvortsova, E S Domnina, N P Glazkova, L P Makhno, in Trudy III Vsesoyuz. Konf. po Khimii Atsetilena (Proceedings of the Third All-Union Conference on the Chemistry of Acetylene) (Moscow: Nauka, 1972) p. 115 155. B V Trzhtsinskaya, N D Abramova, L F Teterina, L V Andriyankova, A V Afonin Izv. Akad. Nauk SSSR, Ser. Khim.416 (1989) c 156. B A Trofimov, S E Korostova, S G Shevchenko, E A Polubentsev, A I Mikhaleva Zh. Org. Khim. 26 1110 (1990) b 157. G I Koldobskii, V A Ostrovskii Khim. Geterotsikl. Soedin. 579 (1988) a 158. E S Petrov Usp. Khim. 52 1974 (1983) [Russ. Chem. Rev. 52 1144 (1983)] 159. B A Trofimov, A I Shatenshtein, E S Petrov,M I Terekhova, N I Golovanova, A I Mikhaleva, S E Korostova, A N Vasil'ev Khim. Geterotsikl. Soedin. 632 (1980) a 160. A F Ermikov, V K Turchaninov, K B Petrushenko, A I Vokin, S E Korostova, L A Ostroukhova, Yu L Frolov Zh. Obshch. Khim. 58 450 (1988) g 161. V K Turchaninov, A F Ermikov, S E Korostova, V A Shagun Zh. Obshch. Khim. 59 791 (1988) g 162. B A Trofimov, A B Shapiro, R N Nesterenko, A I Mikhaleva, G A Kalabin, N I Golovanova, I V Yakovleva, S E Korostova Khim. Geterotsikl. Soedin. 350 (1988) a 163. B A Trofimov, T I Vakul'skaya, S E Korostova, S G Shevchenko, A I Mikhaleva Izv. Akad. Nauk SSSR, Ser. Khim. 142 (1990) c 164. T I Vakul'skaya, B A Trofimov, A I Mikhaleva, S E Korostova, S G Shevchenko, L N Sobenina Khim. Geterotsikl. Soedin. 1056 (1992) a 165. D C Miller,M R Johnson, J A Ibers J. Org. Chem., 59 2877 (1994) 166. A H Jackson, K M Smith, in The Total Synthesis of Natural Products Vol 1 (Ed. J ApSimon) (New York: Wiley-Interscience, 1973) p. 144 167. R Grigg, A W Johnson J. Chem. Soc. 3315 (1964) 168. US P. 4 560 700; Chem. Abstr. 105 114 896 (1986) 169. L N Sobenina, M P Sergeeva, A I Mikhaleva, M V Sigalov, S E Korostova, N I Golovanova, V N Salaurov, E V Bakhareva, N A Vasil'eva Khim. Geterotsikl. Soedin. 612 (1990) a 170. H-P Keiser, J M Muchowski J. Org. Chem. 49 4203 (1984) 171. B A Trofimov, S E Korostova, A I Mikhaleva, L N Balabanova, A N Vasil'ev Khim. Geterotsikl. Soedin. 215 (1977) a 172. G Lunn J. Org. Chem. 52 1043 (1984) 173. B A Trofimov, S E Korostova, A I Mikhaleva, in Tez. Dokl. Konf. po Khimii i Tekhnologii Atsetalei i Ikh Geteroanalogov, Ufa, 1981 (Abstracts of Reports of the Conference on the Chemistry and Technology of Acetals and Their Heteroanalogs, Ufa, 1981) p. 60 S E Korostova, A I Mikhaleva, B A Trofimov 174. S E Korostova, A I Mikhaleva, S G Shevchenko, V V Shcherbakov, R N Nesterenko,M V Sigalov, B A Trofimov Zh. Org. Khim. 22 2489 (1986) b 175. L I Kopylova, S E Korostova, L N Sobenina, R N Nesterenko, M V Sigalov, A I Mikhaleva, B A Trofimov, M G Voronkov Zh. Obshch. Khim. 51 1778 (1981) g 176. USSR P. 1 199 757; Byull. Izobret. (47) 111 (1985) 177. K-P Stefan,W Schuhmann, H Parlar, F Kocte Chem. Ber. 122 169 (1989) 178. N A Bumagin, A F Sokolova, I P Beletskaya, G Vol'ts Zh. Org. Khim. 29 162 (1993) b 179. G Simchen, M W Majchrzak Tetrahedron Lett. 26 5035 (1985) 180. J M Muchowski, D R Solos Tetrahedron Lett. 24 3455 (1983) 181. B A Trofimov, S E Korostova, A I Mikhaleva, L N Sobenina, A N Vasil'ev Khim. Geterotsikl. Soedin. 1631(1982) a 182. B A Trofimov, S E Korostova, S G Shevchenko, A I Mikhaleva, N L Matel' Zh. Org. Khim. 32 897 (1996) b 183. S Nalton Synth. Met. 18 237 (1987) 184. L V Morozova, A I Mikhaleva, S E Korostova, I L Filimonova, in Elektronika Organicheskikh Materialov (Electronics of Organic Materials) (Moscow: Nauka, 1985) p. 305 185. L V Morozova, A I Mikhaleva, M V Markova, in Vodorastvori- mye Polimery i Ikh Primenenie (Tez. Dokl. IV Vsesoyuz. Konf.), Irkutsk, 1991 [Water-Soluble Polymers and Their Applications (Abstracts of Reports of the Fourth All-Union Conference), Irkutsk, 1991] p. 79 186. L V Morozova, A I Mikhaleva, M V Markova, B A Trofimov, in Biologicheski Aktivnye Polimery i Polimernye Reagenty dlya Rastenievodstva (Tez. Dokl. Vsesoyuz. Soveshch.), Nal'chik, 1988 [Bioactive Polymers and Polymer Agents for Plant Growing (Abstracts of Reports of the All-Union Meeting), Nal'chik, 1988] p. 33 187. T Skotheim, I Lundstrom, J Prejra J. Electrochem. Soc. 128 1625 (1981) 188. S Komaba, K Fujihana, T Osaka J. Electrochem. Soc. 145 1126 (1998) 189. R A Simon, A J Ricco,M S Wrighton J Am. Chem. Soc. 104 2032 (1982) 190. R Noufi, A J Frank, A J Nozik J. Am. Chem. Soc. 103 1849 (1981) 191. L V Morozova, A I Mikhaleva,MV Markova, O Yu Molchanov, T V Putan, in Khimiya i Primenenie Pestitsidov (The Chemistry and Application of Pesticides) (Moscow: All-Union Research Institute of Chemical Facilities of Plant Defence, 1990) p. 7 a�Chem. Heterocycl. Compd. (Engl. Transl.) b�Russ. J. Org. Chem. (Engl. Transl.) c�Russ. Chem. Bull. (Engl. Transl.) d�Russ. J. Appl. Chem. (Engl. Transl.) e�Pharm. Chem. J. (Engl. Transl.) f�Mendeleev Chem. J. (Engl. Transl.) g�Russ. J. Gen. Chem. (
ISSN:0036-021X
出版商:RSC
年代:1999
数据来源: RSC
|
3. |
C-Alkenylation of pyrimidine nucleosides and their analogues |
|
Russian Chemical Reviews,
Volume 68,
Issue 6,
1999,
Page 483-504
Anatolii F. Nasonov,
Preview
|
|
摘要:
Russian Chemical Reviews 68 (6) 483 ± 504 (1999) C-Alkenylation of pyrimidine nucleosides and their analogues A F Nasonov, G A Korshunova Contents I. Introduction II. Alkenylation of organoelement nucleoside derivatives and their analogues III. Alkenylation of C-halogeno derivatives and triflates of nucleosides and their analogues IV. Alkenylation of non-substituted nucleosides and their analogues V. Conclusion Abstract. Methods for C-alkenylation of heterocycle moieties of pyrimidine nucleosides and their analogues are generalised. The methods employing both organoelement nucleoside derivatives (organomercury, -tin, -boron and -lithium) and organoelement alkene derivatives (organomagnesium, -boron, -aluminium, -tin and -silicon) as well as palladium-catalysed oxidative coupling and photochemical reactions are considered. The reaction con- ditions (the role of solvents, catalysts and co-catalysts, temper- ature and time modes) are discussed.The advantages and the drawbacks of different approaches for introduction of alkenyl groups are discussed. Examples of practical applications of alkenylated nucleoside derivatives and their analogues are given. The bibliography includes 92 references. I. Introduction Derivatives of pyrimidine nucleosides with substituents at posi- tions 5 and 6 have been the permanent focus of attention of investigators. The number of papers devoted to the synthesis and application of these compounds is rapidly increasing. This interest is largely associated with antiviral and antitumour properties of these nucleosides and the possibility of using them as probes for studying the interactions between biomolecules 1, 2 or as fluores- cent probes for DNA sequencing.Very often these properties are exhibited by nucleosides containing a C7C bond with various substituents at positions 5 and 6 (alkyl, alkenyl, alkynyl, aryl, hydroxyalkyl, acyl, etc.). In this context, systematisation of methods used for the synthesis of such compounds is especially important. A review 3 on C- alkylation of nucleosides which is devoted to problems of syn- thesis of the majority of derivatives with the above substituents was published in 1990. A detailed study 4 describing the synthesis of alkynyl derivatives of nucleosides and their analogues has appeared more recently.A F Nasonov Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 939 31 81. Tel. (7-095) 939 30 20. E-mail: anat@peplab.genebee.msu.su G A Korshunova A N Belozersky Research Institute of Physicochemical Biology, Leninskie Gory, 119899 Moscow, Russian Federation. Fax (7-095) 939 31 81. Tel. (7-095) 939 54 12. E-mail: korshunova@peplab.genebee.msu.su Received 16 June 1998 Uspekhi Khimii 68 (6) 532 ± 554 (1999); translated by R L Birnova #1999 Russian Academy of Sciences and Turpion Ltd UDC 577.113.3 :: 547.854 : 547.3 483 484 493 501 502 Alkenyl-substituted nucleosides seem to play the most crucial role among other nucleosides under consideration; these are extensively used in practically all the fields mentioned above.Thus 5-(E)-bromovinyluridine and its carbocyclic analogues, 5-(E)-(propen-1-yl)uridine and 5-(E)-(3,3,3-trifluoropropen-1- yl)uridine, are known as efficient antiviral agents. Many alkenyl derivatives, particularly those containing a porphyrin fragment,5 manifest an antitumour activity. The extensively investigated reactions between certain alkenes and nucleosides are used to prepare probes to study biomolecules. These reactions allow both direct incorporation of a label into a biomolecule and its addition through an alkenyl linker. As an illustration, we can refer to biotin-containing probes,6 photoactivated labels 7, 8 and fluores- cent probes 9 used in DNA sequencing.Evidently, alkenyl linkers can also be used for other purposes. An interesting example is the addition of carborane-containing fragments 10 used in anticancer boron neutron capture therapy (BNCT). It can be mentioned that the synthesis of alkenyl derivatives of pyrimidine nucleosides and their analogues has a number of special features and often presents a more complicated task than the alkenylation of classical aromatic compounds, e.g., benzene. Such complications can be associated with both the preparation of the starting activated nucleosides and the conducting of the alkenylation reaction. They result from the presence of several reactive functional groups in the nucleoside molecule and the complex behaviour of the quasi-aromatic system of pyrimidine nucleic bases.For example, alkenylation of mercury derivatives of nucleosides with non-activated alkenes in methanol yields pre- dominantly a-methoxyalkyl derivatives, which are not typical of the analogous reactions for the benzene series. One can say that generalised information on these reactions has not been adequately addressed in the literature, especially when one takes into consideration that all the presently known methods of C-alkenylation of pyrimidine nucleosides are extremely diverse. The reviews 11, 12 describing the synthesis of alkenyl-substituted nucleosides point to the exclusive use of 5-mercurionucleoside derivatives. It is of note that the first of them 11 was published as long ago as in 1982, while the second one 12 is rather a summary of results obtained by the authors in the preceding period.Even since the publication of the review by KrecÏ merova',3 several novel alkenyl nucleoside analogues and methods of their synthesis have appeared, which necessitates generalisation and critical evaluation of the vast body of exper- imental data concerning the methods of synthesis of pyrimidine nucleosides with C-alkenylated heterocycles. All the currently known methods used for the synthesis of alkenylated pyrimidine nucleosides can be conventionally divided484 into two groups. The most efficient (Group 1) methods consist in the direct formation of the C7C bond of a nucleoside with an alkenyl substituent.In contrast, Group 2 methods in which the pyrimidine ring is formed in the first step with subsequent glycosylation, have serious drawbacks. These include the neces- sity for preliminary preparation of reagents for cyclisation and the formation of a mixture of a- and b-anomers in the glycosylation step. The present review deals with the description of Group 1 methods, viz., direct alkenylation of the pyrimidine ring of nucleosides and their analogues at position 5 or 6. From our standpoint, the methods of direct alkenylation should best be classified according to the structural features of nucleosides and alkenylating agents involved in the reaction. The choice of compounds to be introduced in the reaction is deter- mined, among other factors, by the ease of their synthesis. In many respects, this factor is crucial for a synthesis-oriented chemist. Thus 5-chloromercury derivatives of nucleosides formed under mild conditions are widely employed in the coupling with alkenes; examples of `reversed' coupling of nucleosides by this reaction are unknown.The readers who are interested in the mechanisms of the reactions discussed below can be referred to the monograph.13 According to the structure of reagents (alkenylating reagents) and substrates (nucleosides), the methods used for alkenylation of pyrimidine nucleosides can be classified as follows: 1. Alkenylation of nucleosides containing a C7E bond (E is the electropositive element, viz., Hg, B, Sn and Li). Coupling of organomercury nucleoside derivatives with alkenes and cross- coupling of organoboron and organotin nucleoside derivatives with alkenyl halides are catalysed by palladium compounds.Nucleoside organolithium derivatives and their analogues as nucleophiles substitute the fluorine atom in perfluoro- or fluoro- chloroalkenes. 2. Alkenylation of nucleosides with a C7Hal or a C7OTf bond. The methods referred to this group can be additionally divided into four subgroups: a. Reactions of halogeno derivatives of nucleosides with alkenes in which carbon atoms at the double bond are not linked to heteroelements. b. Reactions of halogeno derivatives of nucleosides with vinyl acetate. c. Reactions of halogeno derivatives or triflates of nucleosides with organoelement derivatives of alkenes [Group III (B, Al) or Group IV elements (Si, Sn, Zr)].d. Photochemical introduction of alkenyl groups into nucleo- sides. The reactions related to the first three groups are catalysed by palladium compounds. 3. Coupling of non-substituted nucleosides with alkenes. These include addition of nucleosides to activated alkenes, cross- coupling with triflate derivatives of alkenes and addition to alkenyl organomagnesium compounds. The first two reactions are catalysed by palladium complexes. II. Alkenylation of organoelement nucleoside derivatives and their analogues 1. Coupling of organomercury derivatives of pyrimidine nucleosides and their analogues with alkenes Coupling of 5-mercury derivatives of pyrimidine nucleosides with alkenes was the first reaction 14 used in the synthesis of alkenylated nucleosides by direct formation of the C7C bond between the substituent being introduced and the heterocycle. It is also the first reaction which has a versatile preparative significance for C-alkylation of nucleosides in general, since the alkenyl group being introduced can be further (or directly, in the course of the reaction) subjected to easy modification.Reduction of vinyl groups with hydrogen over a platinum catalyst and decarboxyla- tion of carboxyvinyl groups are typical examples of such mod- ifications. The regio- and stereoselectivity of addition of certain substituents is an important feature of the alkenylation reaction. A F Nasonov, G A Korshunova As a matter of fact, this coupling represents the Heck reac- tion,15, 16 although its application to nucleosides has a number of specific features (see below).Its mechanism13, 15, 16 can be pre- sented as a sequence of transmetallation reactions (Hg is replaced by Pd), insertion of an alkene and reductive elimination, and for the sake of simplicity can be represented as follows: PdX2 ArHgX ArPdX H2C=CHY 7HgX2 ArCH=CHY; ArCH2CHY 7PdHX PdX PdHX Pd(0)+HX. This reaction occurs easily for all types of nucleic acid components (nucleic bases, nucleosides and nucleotides). 5-Chloromercurio-, 5-acetoxymercurio- and 5-trifluoroacetoxy- mercurio-pyrimidine nucleosides are normally used.17, 18 Depend- ing on the solubility of the starting compounds and reaction products, the reaction is carried out in methanol, dimethylforma- mide, acetonitrile or aqueous ± organic media.Li2PdCl4, Na2PdCl4, K2PdCl4 and Li2Pd(OAc)2Cl2 are most commonly used as catalysts (Table 1). In the alkenylation of 5-mercurionucleoside derivatives, Pd(II) is irreversibly reduced to Pd(0); therefore, the catalyst has to be used in at least equimolar amounts. To avoid this, an agent which oxidises Pd(0) to Pd(II) (as a rule, CuCl2) is added to the reaction mixture. This allows the use of palladium in catalytic quantities (*0.2 mol. %), but the yield of the target product is slightly lower (Table 1, entries 12, 27 and 30). The mechanism of such coupling has been described in the original study by Heck 15 as well as in the review 11 devoted to the behaviour of nucleosides and their analogues in this reaction.As can be seen from Table 1 (entries 1 and 5), significant differences from the classical Heck reaction are observed when the reaction is carried out in methanol: with ethylene and propylene devoid of M-effects, methoxyalkyl nucleo- side derivatives are formed predominantly. Moreover, the reac- tion with propylene gives all types of alkenyl isomers , viz., cis-, trans- and isopropenyl isomers. These regularities are observed with all simple vinyl compounds of the type H2C=CHR (R=Me, OMe, OAc, Hal). In principle, the formation of a-methoxyalkyl derivatives in the reaction of nucleoside 5-mer- cury derivatives with ethylene and certain alkenes that react regioselectively (Table 1, entries 2 ± 4, 8 and 9) is not important, since methoxy derivatives, as alkenes, are reduced with hydrogen over 10% Pd/C to the corresponding saturated derivatives in overall yields of about 60%± 80%.The double bond C(5)=C(6) in the pyrimidine ring is unaffected by this process. Much better results are achieved in reactions with activated alkenes, partic- ularly with styrene derivatives, esters, amides and nitriles of acrylic, metacrylic and crotonic acids. Here, the reaction usually proceeds regio- and stereoselectively (the b-carbon atom of the double bond is linked to the heterocycle, exclusively E-isomers are formed). The product yields vary from 10% to 80%, being, on the average,*40%± 60%.A typical synthetic procedure consists in the following. An appropriate mercurionucleoside (mercury derivatives of nucleo- sides are insoluble in any known solvent, although a recent report 38 describes the synthesis of water-soluble mercury deriva- tives) is suspended in an excess of an alkene (1.2 ± 11.0 moles per mole of nucleoside) or the substrate is stirred under the pressure of the corresponding gaseous alkene in the presence of 1 equiv. of a palladium catalyst or of 0.1 ± 0.2 equiv. of a Pd catalyst and 2 ± 4 equiv. of CuCl2 at room temperature or with heating to 50 ± 60 8C. After termination of the reaction, Pd is separated by filtration and mercury salts are precipitated by passing hydrogen sulfide through the reaction mixture or by reduction with sodium borohydride.The latter is used, in particular, for subsequent hydrogenation over Pd/C. In some cases, after removal ofC-Alkenylation of pyrimidine nucleosides and their analogues Table 1. Alkenylation of 5-mercuriopyrimidine nucleosides. Entry Substrate a O ClHg 1 H2C=CH2 1) 0.1 M Na2PdCl4, NHO NRib (1) 1 2 1 3 O ClHg NH 4 H2C=CH2 1) 0.1 M Li2PdCl4, 5719 NHO O NdRib (4) 3)H2 , 10% Pd/C, (5) dRib 4 5 1 6 1 7 1 8 4 9 485 Ref. Yield (%) Product Conditions of synthesis Reagent MeO O O 19 NH NH 6 (without MeOH, 9 h; + isolation)+ 2) H2S 3 N O N O9 Rib (2) Rib O H2C=CH2 NHO NRib 1) 0.1 M Li2PdCl4, 8619 MeOH, 3 h; 2) NaBH4, ice; 3) H2, 10% Pd/C, (3) MeOH, 16 h 3 19 48 H2C=CH2 1) 0.1 M Li2PdCl4, CuCl2, MeOH, 3 h; 2) H2S; 3) H2, 10% Pd/C, MeOH, 18 h ON MeOH, 12 h; 2) NaBH4, 0 8C; MeOH±H2O, 24 h O O MeCH=CH2 NH + + 1) 0.1 M Li2PdCl4, 11b19 NHO O N N MeOH, 2 h; 2) H2S dRib (6) (7) dRib O MeO O (6 : 9 : 10= 2.3 : 1.4 : 1.0; the yield of products 7 and 8 is low) NH NH + + O N O N (8) dRib (9) dRib MeO O NHO N (10) dRib O PhCH=CH2 NHO N 1) 0.1 M Li2PdCl4, 3 Ph 919 MeOH, Ar, 3 h; 2) NaBH4 ± EtOH Rib (11) O O H2C=CHCO2Me NH MeO O N 1) 0.1 M Li2PdCl4, 4819 MeOH, Ar, 10 h; 2) H2S Rib (12) O O n-H11C5 19 55+12 NH NH + H2C=CHCH2 ± ±CH(OH)Me OH O O N N Rib Rib 1) 0.1 M Li2PdCl4, MeOH, Ar, 2 h; 2) NaBH4±PriOH; 3) H2, MeOH, 14 h H2C=CHCH2 ± ±CH(OH)CH2OBn NH OH O NdRib 1) 0.1 OH M Li2PdCl4, 5 O 719 MeOH, 3 h; 2) NaBH4 ± MeOH; 3) H2, 10% Pd/C, MeOH, 20 h486 Table 1 (continued). Entry Substrate a 4 10 4 11 4 12 4 13 O 14 (see c) NHO NdRib (dU) 15 dU (see c) 16 dU (see d) 17 dU (see c) O AcOHg 18 NHO NdRib (14) 14 19 O NH 20 O N Na2O3POH2C O 1) 0.14 M Li2PdCl4, MeOH±H2O, 50 8C, 0.5 h; (see c) 2)H2 S (pU) HO OH O NH 21 O N Na2O3POH2C O (see c) OH (pdU) pdU 22 Reagent PhCH=CH2 m-O2NC6H4CH=CH2 m-O2NC6H4CH=CH2 PhCH=CH2 PhCH=CH2 m-O2NC6H4CH=CH2 m-O2NC6H4CH=CH2 p-O2NC6H4CH=CH2 m-O2NC6H4CH=CH2 m-H2NC6H4CH=CH2 m-O2NC6H4CH=CH2 m-O2NC6H4CH=CH2 p-O2NC6H4CH=CH2 Product Conditions of synthesis R 1) 0.1 M Li2PdCl4, MeOH, 12 h; 2) H2S (13) 13, R=H the same but t=18 h 13, R = NO2 13, R = NO2 the same but with the addition of CuCl2 O Ph NH NdRib 1) 0.1 M Li2PdCl4, 5818 MeOH, 18 h; 2) NaBH4 ± MeOH; 3) H2, 10% Pd/C, MeOH, 20 h 13, R=H 1) 0.1 M Li2PdCl4, MeOH, 12 h; 2) H2S the same 13, R = NO2 13, R = NO2 1) 0.1 M Li2PdCl4, MeOH,730 8C, 18 h; 2) H2S O2N the same but at 60 8C 1118 13, R = NO2 1) 0.1 M Li2PdCl4, MeOH, THF, 6 h; 2) H2S 13, R = NH2 the same but under reflux, 9 h NO2 H2O3POH2C (15) HO 15, R = OH 15, R=H the same O2NH2O3POH2C 1) 0.1 M Li2PdCl4, 5918 MeOH±Et2O±H2O, 60 8C, 6 h; 2) H2S A F Nasonov, G A Korshunova Yield (%) Ref.O NHO NdRib 18 57 18 50 18 47 O 18 45 18 62 18 52 O NHO NdRib 18 80 18 30 O NHO N OR 18 82 18 66 O NHO N O OHC-Alkenylation of pyrimidine nucleosides and their analogues Table 1 (continued). Entry Substrate a O AcOHg NH 23 O N Na2O3POH2C O (16) OH 16 24 4 25 4 26 4 27 4 28 4 29 4 30 31 dU (see c) H2C=CHCO2Me O NH 32 O N HOH2C O (see c) Reagent PhCH=CH2 m-N3C6H4CH=CH2 F3CCH=CH2 F3CCH=CH2 H2C=CHCO2Me H2C=CHCN H2C=C(Me)CO2Me MeCH=CHCO2Me H2C=CHCO2Me 487 Yield (%) Ref. Product Conditions of synthesis R O NHO N H2O3POH2C O 1) 0.1 M Li2PdCl4, MeOH±THF±H2O, 50 8C, 5 h; 2) H2S OH (17) 18 50 17, R=H 18 25 17, R = N3 1) 0.274 M Li2PdCl4, MeOH±H2O, 50 8C, 2.5 h 2) H2S O F3C NH 20 17+36 + O N 1) Li2PdCl4, MeOH, 20 8C, 3.5 h; 2) H2S (18) dRib MeO O F3C NHO NdRib (19) 21 26+59 18+19 1) 0.1 M Li2PdCl4, MeOH, 20 8C, 6.5 h; 2) H2S O MeO2C NHO N 1) 0.1 M Li2PdCl4, 6022 CuCl2, MeOH, 20 8C, 3 h; 2) H2S (20) dRib O NC NHO N 1) 0.1 M Li2PdCl4, 1622 MeOH, 20 8C, 12 h; 2) H2S dRibO MeO2C NHO N 1) 0.1 M Li2PdCl4, 6322 MeOH, 20 8C, 16 h; 2) H2S dRib O MeO2C NHO NdRib 1) 0.1 M Li2PdCl4, 2122 CuCl2, MeOH, 20 8C, 96 h; 2) H2S O MeO2C 23 76 NHO N ButMe2SiOH2C O OSiMe2But 1) 0.1 M Li2PdCl4, Et3N, MeOH, 20 8C, 12 ± 16 h; 2) H2S 3) ButMe2SiCl, imidazole, DMF, 20 8C, 18 h O MeO2C NH 23 the same 65 O N ButMe2SiOH2C O488 Table 1 (continued).Entry Substrate a O ClHg NH 33 O N HOH2C OHO OH (21) 21 34 21 35 O ClHg NH 36 O N HOH2C O 4 37 4 38 NH2 ClHg 39 N O NdRib (24) NH2 ClHg 40 N O N (26) Rib 26 41 24 42 Reagent PhCH=CH2 p-O2NC6H4CH=CH2 m-O2NC6H4CH=CH2 NN CF3 p-CF3COC6H4CH=CH2 H2C=CHCO2Et H2C=CHCO2Et H2C=CHCO2Me H2C=CHCN H2C=C(Me)CO2Me A F Nasonov, G A Korshunova Yield (%) Ref. Product Conditions of synthesis R1 R2 O NHO N HOH2C 1) 0.1 M Li2PdCl4, MeOH, 24 h; 2) H2S OHO OH (22) 22, R1=R2=H 24 49 24 the same 19 22, R1=H,R2=NO2 24 " 22 22, R1=NO2, R2=H N N F3C O NH 0.1 M Li2PdCl4, 108 MeOH, 20 8C, 5 h O N HOH2C O O F3CC O NHO N ButMe2SiOH2C 1) 0.1 M Li2PdCl4, 2125 MeOH, 20 8C, 2 h; 2) ButMe2SiCl, imidazole, DMF, 20 8C, 4 h O OSiMe2But O EtO2C ± 26 Li2PdCl4 NHO N (23) dRib NH2 EtO2C 26 the same ± N O N (25) dRib NH2 MeO2C N O N 1) 0.1 M Li2PdCl4, 4227 MeOH, 20 8C, 8 h; 2) H2S (27) Rib NH2 NC N O N 1) 0.1 M Li2PdCl4, 3227 DMF, 20 8C, 6 days; 2) NaBH4±H2O RibNH2 MeO2C N + 27 the same O N 12 b (28 : 29= 1 : 2) (28) dRib NH2 MeO2C N O NdRib (29)C-Alkenylation of pyrimidine nucleosides and their analogues Table 1 (continued).Entry Substrate a 26 43 NH2 ClHg N 44 O N HOH2C OF (30) OH 30 45 O 46 AcOHg NMe O (31) NMe 31 47 31 48 O ClHg NH 49 O N HOH2COH O 50 NHO N HOH2C (see c) N2, 16 ± 36 h OH OH Reagent H2C=CHCONHR (R=(CH2)5CO2Me) H2C=CH2 H2C=CHCO2Et O O AcOOAc CH2OAc O OAc AcO H2C=CHCO2Et ClCH2CH=CH2 489 Yield (%) Ref.Product Conditions of synthesis O NH2 RNHC N O N 1) 0.1 M Li2PdCl4, 3428 MeOH, 24 h; 2) H2S Rib NH2 N O N HOH2C O 1) 0.095 M Li2PdCl4, 5729 DMF, 20 8C, 7 h; 2) H2S, MeOH, 0 8C F OH NH2 EtO2C N O N HOH2C O 1) 0.105 M Li2PdCl4, 2329 MeOH, 20 8C, 7 h; 2) H2S F OH O O 30, 31 66+24 NMe + 1) Pd(OAc)2, LiCl, MeCN, 25 8C, 12 h; 2) H2S Me N O O O NMe Me N O OAc O 31 20+32 the same + HO R R AcO OAc O NMe R= Me N O OAc CH2OAc O HO 31 20+73 " + R OAc AcO AcOCH2 R O NMe R= Me N OO EtO2C NH N O HOH2C 1) 0.1 M Li2PdCl4, 5732 MeOH, N2, 20 h; 2) H2S OH O NHO N HOH2C 1) 0.1 M Li2PdCl4, 1933 MeOH, 3 h; 2) H2S; 3) EtOH, RhCl(PPh3)3,490 Table 1 (continued).Entry Substrate aO NH 51 O N HOH2C 1) 0.1 M Li2PdCl4, ±33 MeOH, 3 h; (see e) 2)H2 S; OH O ClHg O 52 N ¡O P O H2C O O7 3 (32) HO R 32, R = H 32, R = OH 53 32, R = H 54 O XHg 55 NHO NdRib O XHg 56 NHO NdRibO ClHg NH 57 O N HOH2C O OH a Abbreviations used: Rib, D-b-ribofuranosyl; dRib, 20-deoxy-b-D-ribofuranosyl; dU, 20-deoxyuridine; pU, uridine 50-phosphate; pdU, 20-deoxyuridine 50-phosphate.b Overall yield. c The organomercury derivative was synthesised in situ by treatment with Hg(OAc)2 in H2O (50 ± 60 8C). d The same under the action of Hg(OCOCF3)2 in THF. e The same under the action of Hg(OAc)2, then NaCl. palladium the mixture is concentrated and immediately subjected to column chromatography on silica gel. The use of DMF as a solvent permits one to obtain nucleo- sides, e.g., 5-vinylcytidine and its analogues with a modified carbohydrate residue (yields *60%) that cannot be obtained when the reaction is performed in methanol (Table 1, entry 44). Unfortunately, under these conditions the reaction of uridine with ethylene gives exclusively a polymeric product.29 The outcome of the alkenylation reaction depends on the nature of the alkene used.Acrolein does not apparently react with 5-mercury derivatives of nucleosides. Under the action of acryl- onitrile, the corresponding C-alkenylated nucleosides are formed in low yields (Table 1, entries 28 and 41); the adduct with cytidine is reduced upon standard treatment of the reaction mixture with Reagent ClCH2CH=CH2 NH H3N áCH2CH=CH2 O H3N áCH2CH=CH2 H2C=CHCO2R0 O (CH2)5NH HN NH C O R0= (CH2)4 S H2C=CHCONHR HOCH2OO(CH2)6 OH R= HO NHAc H2C=CHCONHR HOCH2 O R= OHHOO(CH2)6 HO H2C=CHPh Product Conditions of synthesis O NH N HOH2C OH 3) EtOH, RhCl(PPh3)3, N2, 16 ± 36 h ¡O O H2C H2N OPO7 0.125 M K2PdCl4, 0.1 M AcONa ±H2O (pH 5), 20 8C, 18 ± 24 h 3 HO R (33) 33, R=H 33, R=OH the same O 0.02 M Li2PdCl4, 5 R0O2C 56 MeOH±H2O, 20 8C, 17 h ¡O P O H2C O7 3 O Pd catalysis RNHCO the same RNHC O Ph N HOH2C 1) 0.1 M Li2PdCl4, 1437 MeOH, 24 h; 2) H2S O OH sodium borohydride to give the corresponding saturated deriva- tive (32%), while hydrogen sulfide does not destroy the palladium complex formed in the reaction.High yields of alkenylation products are obtained in the case of various acrylic acid esters and amides (Table 1, entries 7, 27, 28, 31, 32, 40 and 43). Styrene derivatives containing an electron-withdrawing group in the para- position react less readily than their meta-substituted analogues (Table 1, entries 34 and 35). Aminostyrenes react less readily than nitrostyrenes (Table 1, entries 18 and 19).It is not always clear which particular factors, i.e., electronic effects or solubility, influence the yields of reaction products when para- and meta- isomers of styrene derivatives are used as alkenylating reagents. Thus the yields of the reaction products of a 5-mercury derivative of 20-deoxyuridine 50-phosphate (pdU) with p- and m-nitrostyr- A F Nasonov, G A Korshunova Yield (%) Ref. O O NHO N O 33 ± 35 20 33, 34 ± O NHO N O OH O 36 ± NHO NdRib O 36 ± NHO NdRib NHOC-Alkenylation of pyrimidine nucleosides and their analogues enes in H2O±MeOH± Et2O are comparable (59% and 66%, respectively) (Table 1, entries 21 and 22). Reactions of nucleotides with alkenes in aqueous-organic media give rise to several valuable compounds.Thus the reaction of 5-acetoxymercurio-20-deoxyuridine 50-phosphate with 3-azi- dostyrene gave a photoactivated uridine derivative which inhibits thymidylate synthase (Table 1, entry 24). Coupling of 5-chloro- mercuriouridine 50-triphosphate with allylamine (Table 1, entries 52 and 53) results in the formation of a linker useful for labelling studies. Organomercury derivatives of nucleic bases can also react with alkenes. For example, two 5-substituted 1,3-dimethyluracils are formed from 5-acetoxymercurio-1,3-dimethyluracil and 3,4- dihydro-2H-pyran (Table 1, entry 46). An attempt has been made to use this reaction in the synthesis of C-nucleosides;31 however, in this case cyclic derivatives are predominantly formed. In addition to the above-mentioned compounds, allyl chlor- ides were introduced into the reaction with 5-mercurionucleosides and their derivatives (Table 1, entries 50 and 51).33, 39 The allylation of mercury derivatives of aromatic compounds in the presence of a Pd catalyst was described for the first time by Heck.40 Strictly speaking, this reaction is irrelevant to the material described in this Section, but 5-allylated nucleosides obtained by this method are easily isomerised into the corresponding 5-(alk-1- enyl)pyrimidine nucleosides to give exclusively trans-isomers in the presence of rhodium complexes.It should be remembered that direct coupling of non-activated alkenes with mercury derivatives of nucleosides normally gives complex mixtures of products.This method (i.e., coupling with alkenes) was used to obtain a series of substituted nucleosides that were further used as antiviral agents. An example is 5-(E)-bromovinyluridine and its carbocyclic analogues, 5-(E)-(prop-1-enyl)uridine 19 and 5-(E)-(3,3,3-trifluo- roprop-1-enyl)uridine.21 Of definite interest are the attempts to use nucleoside deriva- tives containing a porphyrin fragment as antitumour agents.33, 39 A covalent binding of vinylporphyrins at position 5 of uridine has been carried out recently.5 Thus porphyrin A (Scheme 1) was O ClHg NH Me N O + AcOH2C O Me AcO OAc MeO2C O NH N O Me Me NH N HN N Me MeO2C B Scheme 1 Me 1) LiPdCl3, DMF±MeCN, N2, 30 8C, 18 h; N N 2) HCl ±H2O Zn N N Me CO2Me A O NH N O Me Me + NH N HN N Me Me Me CO2Me CO2Me CO2Me C introduced into the reaction with an excess (3.3 equiv.) of acetylated 5-chloromercuriouridine in the presence of 1 equiv. (with respect to uridine) of a Pd catalyst.This reaction gave a mixture of trans- (B, 16.5%) and gem-isomers (C, 19.6%) irrespective of the fact that the vinyl group of porphyrin was conjugated with an aromatic system. However, in the presence of a styryl group in the porphyrin fragment, this coupling occurs regio- and stereoselectively. It is noteworthy that this reaction is accompanied by partial substitu- tion of Pd for Zn in the porphyrin fragment of the reaction product (Scheme 2).5 Me Me O ClHg NH Et N N N O Zn + AcOH2C O N N Me AcO OAc MeO2C O NH O N Me Me Et Et N N M N N Me Me MeO2C CO2Me D +E M=Zn (D, 29%), Pd (E, 44%). Yet another rapidly developing field of the application of alkenylated nucleosides is their use for probing biological mole- cules in structural and functional studies.To this end, labels are introduced into nucleosides that are directly linked to the pyrimidine ring (Table 1, entries 36 and 37) or separated from it by a linker (Table 1, entry 54). Aminoallyl, aminoalkylacrylamide and acrylate groups at position 5 are the linkers of choice. 2. Cross-coupling of organotin pyrimidine derivatives of nucleosides with alkenyl halides One of approaches to the synthesis of 6-alkenylated pyrimidine derivatives of nucleosides is based on the reaction of 6-tributyl- stannyl derivatives of nucleosides 34 with alkenyl halides (the Stille reaction).41 An alternative approach consists in the reaction of 6-iodo-substituted pyrimidine nucleosides and their analogues with organoelement derivatives of alkenes (see below).Organotin derivatives of pyrimidine nucleosides are prepared from the corresponding lithium derivatives (6-iodopyrimidine nucleosides and their analogues are synthesised in a similar way) according to the following scheme: 491 Scheme 2 Et LiPdCl3 Me CO2Me492 O O NH NH Bun3 Sn O N O N MomOH2C MomOH2C O O 1) LDA, THF, 778 8C; 2) Bun3 SnCl, THF O O O O 34 Mom=MeOCH2.It should be noted that despite the data 42 on the synthesis of the 5-trimethylstannyl derivative of a pyrimidine nucleoside and its arylation, we are not aware of its application for the incorpo- ration of alkenyl groups. The mechanism of this conversion,13 similar to the mechanisms of other cross-coupling reactions of organometallic compounds with alkyl halides in the presence of transition metal complexes, represents a sequence of reactions of oxidative addition of a transition metal with a low oxidation number [most often, Pd(0) or Ni(0); complexes of low-valent metals can be formed in situ] to aryl or vinyl halides with subsequent transmetallation (substitution of R0 for the halogen atom) and reductive elimination.RX R0M0 RM(II)X M(0) RM(II)R0 R7R0+M(0). ¡M0X Alkenylation of 6-tributylstannylnucleosides has made it possible to obtain the corresponding alkenyl derivatives the yields of which vary from moderate to high (Table 2).43 Depend- ing on the reagent's structure, one can synthesise cis-, trans- and gem-isomers. As can be seen from the reaction scheme, transition metal complexes are required exclusively in catalytic amounts in Table 3. Alkenylation of organolithium derivatives of nucleosides and their analogues.46 t /h Solvent Reagent Substrate a Entry OMe 1 C Br l2 C=CF2 Et2O 1.5 N OMe N (37) 37 1.0 2 Et2O CF3CF=CF2 O I 1.0 3 n-C6H14 ClFC=CF2 NH (see c) O N (38) dRib 1.0 4 38 (see c) CF3 CF=CF2 n-C6H14 1.0 5 38 (see c) CF3 CF=CF2 n-C6H14 a The corresponding organolithium derivatives are prepared by treatment of the substrate with BunLi.b Overall yield. c Prior to coupling, the hydroxy groups of the substrate were protected by the action of (Me3Si)2NH±Me3SiCl ± Py. Table 2. Alkenylation of the organotin nucleoside derivative 34.43 Entry Reagent 1 H2C=CHBr 2 H2C=CBrCO2Et 75 3 1 Br CO2Me 0 81 4 6 Br OH HO 0 70 Note: The reaction was carried out in DMF at 80 8C in the presence of 0.1 equiv. of Pd(PPh3)4 as a catalyst and with 0.2 equiv. of CuI; MomOH2C A=Product F Cl Cl F F3C FF F Cl F F CF3 (F2C)n A F Nasonov, G A Korshunova t /min ProductON R A R=H 100 R=CO2Et MeO2C(35) MeO2C + (36) O .a Overall yield. O O OMe N OMe N OMe OMe F F N N + CF3 OMe N N O O F Cl NH NH + F O O N N dRib dRib O O F F3C NH NH + F O O N NdRib dRib F O NHO NdRib n=3 n=4 Yield (%) NHO 80 56 O a NH N O + (35 : 36= 3 : 1) A O NHO NA O NHO NA Yield (%) 55 42 b OMe 6 b 24 b 79C-Alkenylation of pyrimidine nucleosides and their analogues this case. It should be noted that in some syntheses the presence of copper(I) iodide permits one to increase the product yields more than twofold. Presumably, the role of CuI is connected with the formation of an organocopper intermediate in the transmetalla- tion of Pd for Sn. 3. Cross-coupling of organoboron derivatives of nucleoside analogues with alkenyl halides Only one example of such a coupling is known (Scheme 3).Transmetallation of organoboron compounds was extensively studied by Suzuki 44 who was the first to establish that this reaction requires the presence of a base. Organoboron com- pounds are prepared from the corresponding organolithium derivatives. In the above example, the target product is formed in 72% yield (Scheme 3).45 Scheme 3 OBut OBut (HO)2B N N Br + OBut N OBut N Reagents and conditions: Pd(PPh3)4, 1 M NaHCO3, MeO(CH2)2OMe, N2, 4 h. It should be remembered that this type of product cannot be obtained by coupling of organomercury derivatives with alkenes. 4. Reactions of organolithium derivatives of pyrimidine nucleosides and their analogues with fluoroalkenes Alkenyl derivatives of pyrimidine bases and nucleosides can also be obtained by nucleophilic substitution of polyfluoroalkenes with the corresponding organolithium derivatives.The reaction of 5-bromo-2,4-dimethoxypyrimidine (37) and 5-iodo-30,50-di-O- trimethylsilyl-20-deoxyuridine with butyllithium was carried out in an atmosphere of nitrogen at 770 8C. After addition of an excess (5 equiv.) of the fluorinated alkene, the reaction mixture was kept at a temperature of solid carbon dioxide for 4 days (Table 3). The product yields were small, as a rule, but higher with nucleic bases (42% ± 55%) than with nucleosides (6% ± 24%). In addition, the reaction is not stereoselective (Z- and E-isomers are formed).Perfluorocycloalkenes could be added at position 5 of deoxyuridine, but the yields of the products were as low as a few per cent.46 III. Alkenylation of C-halogeno derivatives and triflates of nucleosides and their analogues 1. The Heck-type coupling of iodonucleosides with alkenes Yet another example of the Heck-type coupling 47 is the reaction of iodonucleosides and their analogues with alkenes in the presence of palladium compounds, which has found wide use in nucleoside chemistry. Its application in organic synthesis has been described in several reviews.48, 49 This reaction was discovered independently by Heck 47 and a group of Japanese investigators.50 Its mechanism consists in oxidative addition of vinyl or aryl halides to Pd(0) with subsequent insertion of an alkene at the aryl(vinyl) ± palladium bond and elimination of PdHX with regeneration of Pd(0).RX H2C=CHR0 Pd(0) RPdIIX RCH2CHR0 7PdHX PdX RCH=CHR0 +Pd(0) + HX The reaction of alkenes with both nucleosides and nucleic bases has been described. Depending on the position of the iodine in the pyrimidine ring, one can obtain 5- and 6-substituted derivatives. Most often, activated alkenes of the acrylic ester type are introduced into the reaction; arylvinyl ketones and 493 terminal aliphatic alkenes are also used. Of special interest is the synthesis of 5- and 6-vinyl derivatives of pyrimidine nucleosides from the corresponding iodonucleosides and 1-arylpropynols in the presence of CuI as the co-catalyst and the synthesis of 5-vinyl derivatives of nucleosides and nucleic bases from the correspond- ing iodo derivatives and vinyl acetate.The coupling is carried out in dioxane, DMF, acetonitrile or aqueous-organic mixtures. Pd(OAc)2 in the presence of Et3N and PPh3 or Pd(PPh3)2Cl2/Et3N are used as catalysts. In the latter case, small amounts of CuI are added. In contrast with the reaction of mercurionucleosides with alkenes, this type of the Heck reaction requires catalytic (but not equimolar) amounts of palladium (most often, as small as 0.05 ± 0.15 equiv.). Depending on the reaction conditions and the nature of nucleosides and reagents used, the yields of the products vary from 8% to 90%. In general, the yields are higher than those in analogous reactions of mercurionucleosides with alkenes.The literature data concerning the synthesis of alkenylated nucleosides by the Heck reaction are summarised in Table 4. These examples demonstrate that pretreatment of the catalyst significantly increases the yield of the product (Table 4, entries 1 and 2). When contemplating the influence of reaction conditions, one cannot but note the recently established increase in the product yield in the coupling of 5-iodo-2,4-dimethoxypyrimidine with methyl acrylate in water (Table 4, entry 29) rather than in organic solvents (the heterogenous Heck reaction 64). It is inter- esting to note that under these conditions the coupling occurs at room temperature, while the classical Heck reaction requires heating to 70 ± 90 8C, which slightly restricts its application.An increase in temperature to 50 8C in the heterogenous reaction results in deiodination of the original pyrimidine and decreases the yield from 90% to 79%. It should be noted that deiodination of nucleosides and their analogues is also the main side reaction in other types of the Heck coupling reaction.65 Presumably, milder conditions of the heterogenous reaction favour the inhibition of side reactions and increase the yields of the target products. High yields of the target compounds in this particular reaction are achieved with alkyl acrylates and arylvinyl ketones as alkenylating reagents. In contrast, the use of terminal alkenes with alkyl substituents results in low yields (Table 4, entries 5 ± 7).Low yields are also characteristic of coupling of vinyl acetate with uridine (the yield of 5-vinyluridine is 37%) and cytidine (8%), although vinyl acetate reacts readily with derivatives of nucleic bases under different conditions in*60% yields (Table 4, entries 27 and 28). Alkynylation of the corresponding 6-iodo-substituted uracils with 1-arylpropynols occurs through intermediate forma- tion of allenic alcohols, which are further isomerised into aroylalkenyl derivatives of nucleic bases (Table 4, entries 20 ± 26).60 Obviously, a mixture of cis- and trans-isomers is formed in this case, although this aspect has not been specially discussed.60 The first alkynylation of aryl iodides was described simulta- neously by Heck 66 and Cassar.67 The Sonogashira reaction 68 (addition of CuI) in the Robins and Barr 69 modification (Et3N as the base) is used in the synthesis of the corresponding nucleosides.A detailed description of this reaction is given in the review.4 Of note is a recently published paper 70 devoted to the synthesis of a porphyrin derivative of deoxyuridine (reaction of acrylamidoporphyrin with 5-iodo-20-deoxyuridine, 5-IdU) (Scheme 4) used to investigate the reactions of porphyrins with the DNA double helix. Coupling of a 5-iodo-substituted nucleoside or its analogue with methyl or ethyl acrylates followed by decarboxy- lation ± bromination of the resulting product with N-bromo- succinimide is used in the synthesis of antiviral analogues of nucleosides containing the E-bromovinyl group at position 5 of the pyrimidine ring more often (due to higher yields) than other alkenylation reactions.494 Table 4.Heck alkenylation of iodo derivatives of nucleosides and their analogues. Substrate Entry O 1 H I 2C=CHCO2Me NHO N (39) Rib 39 2 39 3 38 4 38 5 38 6 38 7 NH2 I N 8 H2C=CHCO2Et Pd(OAc)2, PPh3, 3729 N O N HOH2C OF OH O I NH 9 H2C=CHCO2Me Pd(OAc)2, PPh3, 80a54 O N HOH2COH O I NH 10 O N HOH2COH O I NH 11 O N HOH2C OH (racemate) O I NH 12 O N HOH2C (41) CH2OH 41 (racemate) Reagent H2C=CHCO2Me H2C=CHCO2Et H2C=CHCO2Me PrnCH=CH2 BunCH=CH2 ButCH=CH2 H2C=CHCO2Me H2C=CHCO2Me H2C=CHCO2Me A F Nasonov, G A Korshunova Yield (%) Ref.Conditions of synthesis Product 12 19 53 Pd(OAc)2, PPh3, Et3N, D, 12 h 12 51 86 a Pd(OAc)2, PPh3, Et3N± dioxane (1 : 8), 70 8C, 1 h O EtO2C 52 79 NH Pd(PPh3)2Cl2, Et3N, MeCN, 80 8C, 24 h O NRib 20 53 70 a Pd(OAc)2, PPh3, Et3N± dioxane, D, 40 min O R NH Pd(OAc)2, Et3N, 100 8C, 12 h O N (40) dRib 22 25 40, R = Prn 22 the same but t=2 h 40, R = Bun 15 22 the same but t=48 h 40, R = But16 NH2 EtO2C N O HOH2C O Et3N, DMF, Ar, 85 8C, 24 h F OH O MeO2C NHO N HOH2C Et3N, dioxane, 85 8C, 3 h OH O MeO2C NH 55 the same 64 O N HOH2COH O MeO2C NH a O N HOH2C Pd(OAc)2, PPh3, 6656 Et3N, dioxane, Ar, 70 8C, 2.5 h OH (racemate) O MeO2C NH 56 the same but t=4 h 75a O N HOH2C (42) CH2OH 42 (racemate)C-Alkenylation of pyrimidine nucleosides and their analogues Table 4 (continued).Substrate Entry 41 (homochiral) 13 O I NH 14 O N HOH2C O CH2OH O I NH 15 O N HOH2C O O I NH 16 O N TolOH2C O OTol (43) 43 17 43 18 43 19 O 20 NHO I NH(45) 45 21 45 22 45 23 O 24 NMe O I Me N (47) 47 25 47 26 O I NMe 27 O Me N (49) OMe I 28 N OMe N (50) Reagent H2C=CHCO2Me H2C=CHCO2Me H2C=CHCO2H PhC(O)CH=CH2 p-MeC6H4C(O)CH=CH2 p-MeOC6H4C(O)CH=CH2 m-MeC6H4C(O)CH=CH2 PhCH(OH)C:CH p-MeC6H4CH(OH)C:CH p-MeOC6H4CH(OH)C:CH " o-MeOC6H4CH(OH)C:CH " PhCH(OH)C:CH p-MeC6H4CH(OH)C:CH p-MeOC6H4CH(OH)C:CH " H2C=CHOAc H2C=CHOAc Conditions of synthesis Pd(OAc)2, PPh3, Et3N, dioxane, Ar, 70 8C, 4 h Pd(OAc)2, PPh3, 5657 Et3N, DMF, Ar, 75 8C, 10 h Pd(OAc)2, PPh3, 4958 Et3N, dioxane, 1 N NaOH, 95 8C, 3 h Pd(PPh3)2Cl2, CuI, Et3N, DMF, N2, 80 8C, 24 h the same ""Pd(PPh3)2Cl2, CuI, Et3N, DMF, N2, 55 8C, 6 h the same ""Pd(PPh3)2(OAc)2, 6061 Et3N, 100 8C, 5 h the same Product 42 (homochiral)O MeO2C NHO N HOH2C O CH2OH O HO2C NHO N HOH2C OO O R2 NHO N R1 TolOH2C O OTol (44) 44, R1=R2=H 44, R1=Me, R2=H 44, R1=OMe, R2=H 44, R1=H, R2=Me O NH O R2 CH=CH NH R1 (46) 46, R1=R2=H 46, R1=Me, R2=H 46, R1=OMe, R2=H 46, R1=H, R2=OMeO NMe O O CH=CH NMe (48) R48, R=H 48, R=Me 48, R=OMe O NMe O NMe OMe N OMe N 495 Yield (%) Ref.56 69 aa 59 62 59 78 59 62 59 62 O 60 39 60 60 60 57 60 35 60 76 60 80 60 58 61 58496 Table 4 (continued). Substrate Entry 50 29 49 30 38 31 NH2 I 32 N O N (52) dRib a The catalyst was prepared beforehand by heating Pd(OAc)2, PPh3 and Et3Nin dioxane orDMFup to the appearance of deep red colouring (10 ± 20min) at 70 ± 75 8C. O I NH N O N + HOH2C O OH5-IdU The addition of carborane-containing fragments 10 to 5-alkenyl nucleosides (Scheme 5) is an illustrative example of the application of alkenyl fragments as linkers; this reaction is used in cancer boron neutron capture therapy (for a review, see Ref.71). O O HO NH N HOH2C O OH H (CH2)nO B10H10 Reagent H2C=CHCO2Me O H2C=CHOAc H2C=CHOAc N NH N HN N N (CH2)nOH O H + B10H10 O O NH N HOH2C O OH Conditions of synthesis Pd(OAc)2, PPh3, 9062 K2CO3, Bu4NHSO4, H2O, 20 8C, 90 h Pd(PPh3)2(OAc)2, 6430 Et3N, 100 8C, 5 h Pd(OAc)2, PPh3, 3763 Et3N, DMF, 70 8C, 5.5 h Pd(OAc)2, PPh3, 863 Et3N, DMF, 90 8C, 2 h NH C O Scheme 5 K2CO3, DMF O A F Nasonov, G A Korshunova Yield (%) Ref. Product OMe MeO2C N OMe N O O NMe Me N O O a NHO N (51) dRib NH2 a N O NdRib N Scheme 4 NH N NH N Pd(PPh3)2Cl2, Et3N, MeCN HN N C O O NHO N N HOH2C O OH 2.Cross-coupling of nucleoside halogeno derivatives and triflates with organoelement derivatives of alkenes Recently, coupling of 5-halogeno derivatives and 5-triflates of nucleosides with organoelement derivatives of alkenes has acquired growing popularity for the synthesis of the correspond- ing 5-alkenyl derivatives of nucleosides.72 ± 84 As mentioned above, in this case alkene derivatives with Group III (boron and aluminium) and Group IV elements (silicon, tin, zirconium) of the Periodic system are used as the starting compounds. The literature data on alkenylation of 5-halogeno derivatives and 5-triflates of nucleosides are summarised in Table 5. The mechanisms of the corresponding reactions are considered in Section II.2 exempli- fied in the cross-coupling of organotin nucleoside derivatives with alkenyl halides.An essential advantage of this method is the possibility of performing the reaction with non-activated alkenes to obtain the corresponding 5-alkenyl-substituted nucleosides (particularly, 5-vinyl-substituted ones) (Table 5, entries 3, 13, 18, 26, 29 ± 31, 33 and 35) in good yields. It should be remembered that of all 5-mercurionucleosides only the 5-mercuriocytidine derivative reacts with ethylene in DMF under conditions of the Heck reaction (Table 1, entry 44); 5-vinyluridine undergoes polymerisation under these condi-C-Alkenylation of pyrimidine nucleosides and their analogues Table 5.Alkenylation of halogeno derivatives and triflates of nucleosides and their analogues with organoelement derivatives of alkenes. Entry 12345678910 11 12 13 14 15 16 17 18 19 20 21 22 23 Reagent Substrate 38 Ph H O Br Me2Al NHO NdRib O I R NHO N H TMSOCH2 O (54) OTMS 54 54 54 54 54 54 54 54 54 R 43 H 43 43 43 43 O TfO NH R O N AcOCH2O H AcO OAc (56) 56 56 56 56 56 EtO2CHB(OH)2 H H C6H13-n HZrCp2 R=H R=Bun R=C8H17-n R=Ph R=p-MeOC6H4 R=HC:C(CH2)2 R=MeCHCl(CH2)2 R=MeCHCN(CH2)2 R=EtO R=H2C=CMe HSnBun3 R=H R=SiMe3 R=Ph R=CO2Et R=CH2OTHP HSnBun3 R=H R=SiMe3 R=Ph R=PriMe2CSiMe2OCH2 5 h R=CO2Et SnBun3 Conditions of synthesis Product Pd(m-Ph2PC6H4SO3M)3 13, R=H (M=Na, K), H2O± MeOH, N2, 80 8C, 3 h 1) HMDS± (NH4)2SO4; 27 73 +dU 2) PdCl2, PPh3, THF, Ar, 24 h; 3) NH4Cl ±H2O± (53) MeOH, 2.5 ± 24 h 1) Pd(PhCN)2Cl2, 20 8C, 15 h; 2) H2O±MeOH the same """"""""Pd(PPh3)2Cl2, MeCN, N2 20 8C, 6 h 60 8C, 16 h 50 8C, 16 h 50 8C, 20 h 60 8C, 16 h Pd(PPh3)4, LiCl, dioxane, N2, D 4 h 4 h 8 h 1 h Pd(PPh3)4, LiCl, dioxane, N2, D 15 h O n-H13C6 NHO NdRib O R NHO N (51) dRib 51, R=H 51, R = Bun 51, R = C8H17-n 51, R = Ph 51, R = p-MeOC6H4 51, R = HC:C(CH2)2 51, R=MeCHCl(CH2)2 51, R=MeCHCN(CH2)2 51, R=EtO 51, R = H2C=CMe O R NHO N TolOCH2 O OTol (55) 55, R=H 55, R=SiMe3 55, R = Ph 55, R = CO2Et 55, R = CH2OTHP O R NHO N AcOCH2O AcO OAc (57) 57, R=H 57, R=SiMe3 57, R=Ph 57, R = PriMe2CSiMe2OCH2 57, R = CO2Et O NH EtO2C O N AcOH2C O AcO OAc 497 Yield (%) Ref.72 47 a 74, 75 74 74 74 74 74 74 74 74 74 83 95 96 85 39 80 89 76 49 30 76 76 76 76 76 86 82 81 57 82 77, 78 77, 78 77, 78 77, 78 77, 78 87 73 75 92 92 77, 78 46498 Table 5 (continued). Substrate Entry 56 24 O I NH 25 O N HOH2C O N O F O O I 26 NHO NH (58) 58 27 58 28 O I NH 29 N O TolOH2C O TolO OTol (60) 60 30 O I NH ButSiMe2 31 O N OH2C O HO 38 32 O I NTol 33 O N TolOH2C O TolO O NH I N ButMe2SiOH2C 34 O O O Reagent SnMe3 H Me3Si SnBun H 3 H RH SnBun3 R=H R=Ph H HSnBun Me 3 H2C=CHSnBun3 H2C=CHSnBun3 H2C=CHSnBun3 F Me3Si SnBun H 3 (H2C=CH)4Sn O H2C=CHSnBun3 Conditions of synthesis Product the same but t=5 h Pd(PPh3)2Cl2, 5079 THF, Ar, 50 8C, 16 h Pd2(dba)3, TFurP, NMP, 20 8C, 16 h the same Pd2(dba)3, TFurP, NMP, 16 h at 20 8C and 5 h at 50 8C Pd2(dba)3, TFurP, NMP, 20 8C, 72 h Pd2(dba)3, TFurP, THF, 50 8C, 40 h Pd2(dba)3, TFurP, NMP, 20 8C, 72 h Pd(PPh3)4, DMF, Ar, 100 8C, 4 h 1) Pd(PPh3)4, HMPA, N2, 60 8C, 18 h; 2) MeONa±MeOH Pd(PPh3)4, DMF, Ar, 90 8C, 1.5 h O NH Me O N AcOH2C O AcO OAcO Me3Si NHO N HOH2C O N O F OO R NHO NH (59) 59, R=H 59, R = Ph O NH Me O NH O NHO N TolOH2C O TolO OTol (61) 61 ON ButMe2SiOH2C O HO Me3Si F O NHO NdRib 51 ON ButMe2SiOH2C O O O A F Nasonov, G A Korshunova Yield (%) Ref.77, 78 86 80 89 80 92 80 70 80 76 80 98 NH 80 91 O 81 28 82 80 NHO 83 95C-Alkenylation of pyrimidine nucleosides and their analogues Table 5 (continued). Substrate Entry 49 35 49 36 49 37 49 38 49 39 O I 40 NBn O NH (63) 63 41 58 42 O I NBz 43 O N BzOH2C O OBz (66) 66 44 38 45 38 46 Note. The following abbreviations are used: TFurP, tris(2-furyl)phosphine; NMP, N-methylpyrrolidone; HMPA, hexamethylphosphorotriamide (hexametapol); THP, tetrahydropyranyl; dba, dibenzylideneacetone. a Overall yield.Reagent H2C=CHSiMe2Cl H ClMe2Si Ph H H FMe2Si Ph H H FMe2Si C6H13-n HPrn Prn H SiFMe2H F2MeSi Ph H H F2MeSi H C6H13-n H F2MeSi H C6H13-n H F2MeSi Ph H H F2MeSi H C6H13-n H FMe2Si Ph H H FMe2Si H C6H13-n Conditions of synthesis Product O R NMe [(Z3-C3H5)PdCl]2, Bu4NF, THF, Ar, 60 8C, 14 h Me N O (62) 62, R=H 62, R = Ph the same 62, R = Ph "" 62, R = C6H13-n Prn " 6 O 384 Prn NMe O NMe O " Ph NBn O NH O n-H13C6 NBn + O " 5384 (64 : 65= =2:1) NH(64) O NBn n-H13C6 O NH (65) the reaction does not occur " O R NBz " O N BzOH2C O OBz (67) 67, R = Ph " 67, R = C6H13-n O NH R (R=Ph) 13+ O N " 7984 (13 : 68= 5 : 1) dRib (68) " 53+68 (R=C6H13-n) 499 Yield (%) Ref.84 82 84 70 84 69 84 70 84 58 a 84 ± 84 64 84 60 a 84 74 a (53 : 68= 2 : 1)500 tions.29 In the reaction of vinyl acetate with 5-iodouridine, the yield of the target product is as low as 37% (Table 4, entry 31), while the yield of vinylcytidine is even lower (8%) (Table 4, entry 32).63 As can be seen from the results presented in Table 5, there are only a few examples of nucleoside alkenylation with organoboron and organoaluminium compounds and the yields of reaction products are low. Thus styrylboronic acid reacts with 5-iodo- deoxyuridine in the presence of Pd(m-Ph2PC6H4SO3M)3 (M=K, Na) even in aqueous-organic media, the yield of the target product is 47% (Table 5, entry 1).72 The possibility of carrying out the reaction in the presence of water is important, particularly for modification of oligonucleotides.In the coupling of organoalu- minium alkene derivatives with nucleosides, the product yields are significantly lower (Table 5, entry 2).73 However, the use of organozirconium alkene derivatives and Pd(PhCN)2Cl2 as a catalyst makes it possible to obtain a broad spectrum of 5-alkenylnucleoside derivatives, including vinyl (Table 5, entry 3), alkylvinyl (Table 5, entries 4 and 5), containing a second multiple bond in the chain introduced (Table 5, entries 8 and 12), styryl (Table 5, entries 6 and 7) and other derivatives.It is of note that the reaction (e.g., with 5-iodo-20-deoxy-30,50-di-O-trimethyl- silyluridine in THF) occurs even at room temperature.74, 75 The zirconium-containing alkenyl reagent is synthesised from the corresponding alkyne and the Schwartz reagent. H R Cl(H)ZrCp2, C6H6 RC:CH H ZrCp2 Organotin compounds are most often used for C-alkenylation of nucleosides. Alkenyltributyltin is normally employed for this purpose; alkenyltrimethyltin (Table 5, entry 24) and tetraalkenyl- tin (Table 5, entry 33) are used less often. The amounts of the alkenylating reagent vary from 1 to 5 equiv. 5-Iodo-substituted nucleosides and nucleoside 5-triflates (as a rule, with protected Table 6.Photochemical alkenylation of nucleosides. Reagent Substrate Entry 39 1 H2C=CHCO2Me 38 2 H2C=CHCO2Me 39 3 H2C=CHCN 38 4 H2C=CHCN NH2 5 H I MeO2CCH=CH 2C=CHCO2Me N O N (71) Rib 52 6 H2C=CHCO2Me 71 7 H2C=CHCN 52 8 H2C=CHCN Note: The reaction was carried out in three steps: (1) HMDS, Py, Ar, 18 h; (2) alkene, MeCN, hn=254 nm, 24 h; (3) H2O. A F Nasonov, G A Korshunova hydroxy groups) serve as substrates. The reactions are conducted in acetonitrile, dioxane, THF,N-methylpyrrolidone, hexametapol and DMF. Pd(PPh3)2Cl2, Pd(PPh3)4, Pd(PPh3)4 in the presence of LiCl and Pd2(dba)3 in the presence of TFurP serve as catalysts. The yields of the target compounds obtained by this method vary from 28% to 98%. Depending on the reagent's structure, one can obtain trans-, cis- (Table 5, entry 28) and gem-products (Table 5, entries 23 and 24).The reaction of alkenylfluoro- and alkenylchlorosilanes with 5-iodouracil and 5-iodo-20-deoxyuridine derivatives, which gives target products in high yields, has been studied recently.84 To obvious merits of organosilicon reagents one can relate their low toxicity, moisture resistance and low cost. This reaction was carried out in the presence of 0.025 equiv. of [(Z3-C3H5)PdCl]2 and 2 equiv. of Bun4 NF. This method allows the modification of the pyrimidine ring with alkenyl radicals of various structure, e.g., vinyl (Table 5, entry 35), styryl (Table 5, entries 36, 37, 40, 43 and 45), alkylvinyl (Table 5, entries 38, 41, 44 and 46), etc.As a rule, protected substrates are used; non-protected halogeno derivatives of nucleosides react with organosilicon compounds less readily. Thus 5-iodouracil does not react with alkenylfluoro- and alkenyl- chlorosilanes and 5-iododeoxyuridine forms a mixture of trans- and gem-isomers. 3. Photochemical alkenylation of nucleosides The advantages of the photochemical reaction of activated alkenes with 5-iodopyrimidine nucleosides are technologically simple and versatile procedures used for the synthesis and isolation. However, the yields of the target products are small, varying from low (12%) to moderate (41%) (the literature data on photochemical coupling of nucleosides with alkenes are presented in Table 6). In addition, alkenes are linked non-stereoselectively to give a mixture of cis- and trans-isomers. Therefore, it is reasonable to use this reaction exclusively for the synthesis of the corresponding alkyl derivatives by hydrogenation of the alkenyl- Yield (%) Ref.Product (a mixture of E- and Z-isomers) O MeO2CCH=CH NHO NR 85 85 38 41 (69) 69, R=Rib 69, R=dRibO NCCH=CH NHO NR (70) 85 85 70, R=Rib 70, R=dRib 12 14 NH2 N O NR (72) 86 86 72, R=Rib 72, R=dRib 22 24 NH2 NCCH=CH N O (73) NR 86 86 73, R=Rib 73, R=dRib 18 20C-Alkenylation of pyrimidine nucleosides and their analogues ated product over Pd/C. Alkenyl fragments are added to the trimethylsilylated 5-iodonucleoside upon irradiation of the reac- tion mixture with light of the wavelength of 254 nm in deoxy- genated and dry acetonitrile for 24 h.IV. Alkenylation of non-substituted nucleosides and their analogues Pyrimidine nucleosides and their analogues that are not substi- tuted at positions 5 and 6 of the heterocycle can also enter into the alkenylation reaction. Three versions of this reaction are known, viz., oxidative coupling of alkenes (predominantly, activated ones), cross-coupling with vinyl triflates (both reactions are carried out in the presence of Pd compounds) and nucleophilic addition of alkenylmagnesium derivatives at position 6 of the pyrimidine ring. 1. Oxidative coupling of activated alkenes Oxidative coupling of alkenes (the Fujiwara reaction 87) is carried out in the presence of 1 equiv.of Pd(OAc)2 on boiling in acetonitrile or at room temperature (Table 7). The mechanism of this reaction seems to consist in the formation of an arylpalladium Table 7. Alkenylation of non-substituted nucleosides with activated alkenes. Substrate Entry1 H O 2C=CHCO2Me 24 NMe O NMe (74) 74 2 74 3 74 45 dU 6 dUUrd 7 O NH 8 H2C=CHCO2Me 53 a O N HOH2C O O O 74 9 t /h Reagent 16 H2C=CHCN 16 H2C=CHCOMe 13 H2C=CHPh 32 H2 C=CHCO2Me 62 H2 C=CHPh 72 H2C=CHCO2Me O 24 MeO hydride intermediate followed by insertion of the alkene at the Ar7Pd bond and reductive elimination of PdHX to give Pd(0).88 Pd H2C=CHY ArPdH ArH ArCH=CHY +Pd(0)+HX Reactions of 1,3-dimethyluracil with activated alkenes (methyl acrylate, acrylamide, methyl vinyl ketone and styrene) give uracil derivatives substituted at position 5 in good yields; all these reactions result in trans-isomers.In the case of non-activated alkenes, gem-isomers are formed as admixtures. It is of note that acrolein diethyl acetal also enters into the reaction, which permits one to introduce an alkenyl substituent with a terminal aldehyde group into the pyrimidine base (Table 7, entry 13). It should be remembered that this reaction cannot be realised with mercurio- nucleosides. The products of coupling nucleosides with acrolein diethyl acetal could not be obtained thus far. In general, nucleo- sides add alkenes less readily in this reaction (Table 7, entries 6, 7 and 14).Only in the case of 20,30-O-isopropylidene-protected nucleosides could the methyl acrylate group be introduced at Product O MeO2C NMe Me N O O NC NMe Me N O O O Me NMe Me N O O Ph NMe Me N O 20 13, R=H 12 O MeO2C MeO2C NHO N H2C HOH2C + O O O O MeO NMe + O Me N O O O + NMe MeO Me N O 501 ArCH2CHY 7PdHX PdX Yield (%) Ref. 89 92 89 73 89 74 89 86 89 56 89 35 89 47 O NH O 89 74+23 O N O O O 90 36+10502 Table 7 (continued). Substrate Entry 74 10 74 11 74 12 74 13 O NH 14 O N HOH2C O OHOH Note: The reaction was carried out in the presence of 1.2 equiv. of Pd(OAc)2 in boiling acetonitrile.Urd is uridine (1-b-D-ribofuranosyluracil). a The reaction was carried out at room temperature. position 5 of the pyrimidine ring (Table 7, entry 8); the target product was obtained in good yield (74%). 2. Oxidative coupling with vinyl triflates Vinyl triflates react with non-substituted nucleosides in the presence of a catalyst [(Pd(PPh3)4 (*0.05 equiv.)] and an excess of LiCl in THF to give the corresponding alkenyl derivatives in moderate yields (Table 8). This reaction, as the coupling with organoelement alkene derivatives, permits one to introduce non- activated alkenyl substituents into position 5 of the pyrimidine Table 8. Alkenylation of non-substituted nucleosides with vinyl triflates.91 Reagent Entry 1 H2C=C(Me)OTf 2 Me2 C=CHOTf 3 OTf Note: Reaction conditions: LiCl, Pd(PPh3)4, THF, N2, 48 h; Urd, uridine (1-b-D-ribofuranosyluracil) as a substrate.Reagent O MeOO OH2C=CHCO2Bui H2C=CHCH(OEt)2 H2C=CHCO2Et Product O NHO NRibO NHO NRib O NHO NRib t /h 24 24 24 24 a 24 Yield (%) 54 43 45.7 A F Nasonov, G A Korshunova Yield (%) Ref. Product O O 90 25 NMe MeO Me N O O O NMe O 90 23 Me N O O 90 ± BuiO2C NMe Me N O O 90 80 OHC NMe Me N O O EtO2C NH 90 40 O N HOH2C O OHOH ring. Depending on the structure of the vinyl triflate reagent, one can obtain various regio- and stereoisomers. 3. Addition of alkenylmagnesium derivatives Vinylmagnesium bromide and chloride are added at position 6 of the pyrimidine ring of the protected nucleosides; this reaction gives 6-alkenyldihydropyrimidine derivatives of nucleosides in good yields.Thus reaction of protected uridine 75 with vinyl- magnesium chloride 83, 92 and reaction of the 5-bromo derivative 76 with vinylmagnesium bromide (Ar, 1 h) give their 6-vinyl- dihydro derivatives 77 and 78 (yields 61% and 96%, respectively). OTPS OTPS R R NH N O N O N ButMe2SiOH2C ButMe2SiOH2C O O H2C=CHMgX THF, 0 8C ButMe2SiO ButMe2SiO OSiMe2But OSiMe2But 77, 78 75, 76 R = H (75, 77), Br (76, 78). The attempts 83 to oxidise reaction products to pyrimidine derivatives of nucleosides were unsuccessful. V. Conclusion Numerous methods for C-alkenylation of nucleosides and their analogues are currently known, which differ in complexity, reaction conditions and the possibility of obtaining substituted nucleosides with a predetermined structure.In our opinion, the most promising approaches involve reactions of organoelement alkene derivatives, in particular, organotin (in non-aqueousC-Alkenylation of pyrimidine nucleosides and their analogues solvents) and organoboron derivatives (in H2O or aqueous- organic media). Organosilicon derivatives of alkenes seem to be promising reagents that might replace toxic organotin compounds in future, although their cross-coupling is insufficiently developed yet. Mercury derivatives of nucleosides used to introduce alkenyl substituents will hardly lose their significance by virtue of their availability, which can be attributed to the ease of their prepara- tion and the presence of a large body of experimental material that helps selecting the conditions for a specific synthesis.However, in more distant future this method will most likely be replaced by analogous procedures, e.g., the Heck-type coupling of iodonu- cleosides with alkenes, which affords higher yields of the target products and precludes the use of toxic substrates. The methods of oxidative coupling of nucleosides with activated alkenes and alkenyl triflates also attract attention owing to their experimental simplicity. At the same time, literature data concerning these methods are scarce, which makes difficult their application in the synthesis of novel compounds and long-range prognoses. And finally, the use of photochemical alkenylation of nucleosides for preparative purposes will, most probably, hardly have any practical significance. References 1.S A Fleming Tetrahedron 51 12479 (1995) 2. G Ya Sheflyan, E A Kubareva, E S Gromova Usp. Khim. 65 765 (1996) [Russ. Chem. Rev. 65 709 (1996)] 3. M KrecÏ merova' Chem. Listy 84 1282 (1990) 4. V A Korshun, E V Manasova, Yu A Berlin Bioorg. Khim. 23 324 (1997) a 5. X Jiang, R K Pandey, K M Smith J. Chem. Soc., Perkin Trans. 1 1607 (1996) 6. P S Nelson, C Bahl, I Gibbons Nucleosides Nucleotides 5 233 (1986) 7. B Bartholomew, G A Kassavetis, B R Braun, E P Geiduschek EMBO J. 9 2197 (1990) 8. T Yamaguchi,M Saneyoshi Nucleosides Nucleotides 15 607 (1996) 9.J A Brumbaugh, L R Middendorf, D L Grone, J L Ruth Proc. Natl. Acad. Sci. USA 85 5610 (1988) 10. F-G Rong, A H Soloway Nucleosides Nucleotides 13 2021 (1994) 11. D E Bergstrom Nucleosides Nucleotides 1 1 (1982) 12. D E Bergstrom, X Lin, G Wang, D Rotstein, P Beal, K Norrix, J Ruth Synlett 179 (1992) 13. J P Collman, L S Hegedas, J Norton, R G Finke Principles and Applications of Organotransition Metal Chemistry (Mill Valley, CA: University Science Books, 1987) 14. D E Bergstrom, J L Ruth J. Am. Chem. Soc. 98 1587 (1976) 15. R F Heck J. Am. Chem. Soc. 90 5518 (1968) 16. R F Heck Palladium Reagents in Organic Syntheses (New York: Academic Press, 1985) 17. D E Bergstrom, J L Ruth J.Carbohydr. Nucleosides Nucleotides 4 257 (1977) 18. C F Bigge, P Kalaritis, J R Deck,M P Mertes J. Am. Chem. Soc. 102 2033 (1980) 19. D E Bergstrom, M K Ogawa J. Am. Chem. Soc. 100 8106 (1978) 20. Y Wataya, A Matsuda, D V Santi, D E Bergstrom, J L Ruth J. Med. Chem. 22 339 (1979) 21. D E Bergstrom, J L Ruth, P A Reddy, E De Clercq J. Med. Chem. 27 279 (1984) 22. J Goodchild, R A Porter, R H Raper, I S Sim, R M Upton, J Viney, H J Wadsworh J. Med. Chem. 26 1252 (1983) 23. Y-M Cho, F Johnson Tetrahedron Lett. 35 1149 (1994) 24. S Izuta,M Saneyoshi Chem. Pharm. Bull. 35 4829 (1987) 25. T Yamaguchi, K Suyama, K Narita, S Kohgo, A Tomikawa, M Saneyoshi Nucl. Acids Res. 25 2352 (1997) 26. A S Jones, G Verhelst, R T Walker Tetrahedron Lett.4415 (1979) 27. D E Bergstrom, H Inoe, P A Reddy J. Org. Chem. 47 2174 (1982) 28. G W Ashley, P A Bartlett J. Biol. Chem. 259 13615 (1984) 29. M E Perlman, K A Watanabe, R F Schinazi, J J Fox J. Med. Chem. 28 741 (1985) 30. I Arai, G D Daves Jr J. Org. Chem. 43 4110 (1978) 31. I Arai, G D Daves Jr J. Am. Chem. Soc. 100 287 (1978) 503 32. R C Cookson, P J Dudfield, R F Newton, P Ravenscroft, D I C Scopes Eur. J. Med. Chem. 20 375 (1985) 33. J Goodchild, H J Wadsworh, I S Sim Nucleosides Nucleotides 5 571 (1986) 34. P R Langer, A A Waldrop, D C Ward Proc. Natl. Acad. Sci. USA 78 6633 (1981) 35. M Shimkus, J Levy, T Herman Proc. Natl. Acad. Sci. USA 82 2593 (1985) 36. S Pochet, S R Sarfati, J Igolen Nucleosides Nucleotides 8 1073 (1989) 37.T Yamaguchi,M Saneyoshi Nucleosides Nucleotides 11 373 (1992) 38. US Appl. 95-382 892; Chem. Abstr. 125 248 329 (1996) 39. D E Bergstrom, J L Ruth, P Warwick J. Org. Chem. 46 1432 (1981) 40. R F Heck J. Am. Chem. Soc. 90 5531 (1968) 41. J K Stille Angew. Chem., Int. Ed. Engl. 25 502 (1986) 42. P Wigerinck, L Kerremans, P Claes, R Snoeck, P Maudhal, E De Clercq, P Herdewijn J. Med. Chem. 36 538 (1993) 43. G Palmisano, M Santagostino Tetrahedron 49 2533 (1993) 44. N Miyaura, A Suzuki Chem. Lett. 879 (1981) 45. D Peters, A-B HoÈ rnfeldt, S Gronowitz J. Heterocycl. Chem. 27 2165 (1990) 46. P L Coe,M R Harnden, A S Jones, S A Noble, R T Walker J. Med. Chem. 25 1329 (1982) 47. R F Heck, J P Nolley Jr J. Org. Chem. 37 2320 (1972) 48.A de Meijere, F E Meyer Angew. Chem., Int. Ed. Engl. 33 2379 (1994) 49. W Carbi, I Candiani Acc. Chem. Res. 28 2 (1995) 50. T Mizoroki, K Mori, A Ozaki Bull. Soc. Chem. Jpn. 44 581 (1971) 51. E De Clercq, C Desgranges, P Herdewijn, I S Sim, A S Jones, M J McLean, R T Walker J. Med. Chem. 29 213 (1986) 52. R Kumar, L Xu, E E Knaus, L I Wiebe, D R Tovell, D L Tyrrell, T-M Allen J. Med. Chem. 33 717 (1990) 53. M Ashwell, A S Jones, A Kumar, J R Sayers, R T Walker, T Sakuma, E De Clercq Tetrahedron 43 4601 (1987) 54. P Herdewijn, E De Clercq, J Balzarini, H Vanderhaeghe J. Med. Chem. 28 550 (1985) 55. L J J Hronowski, W A Szarek J. Chem. Soc., Chem. Commun. 1547 (1990) 56. W A Slusarchyk, G S Bisacchi, A K Field, D R Hockstein, G A Jacobs, B McGeever-Rubin, J A Tino, A V Tuomari, G A Yamanaka,M G Young, R Zahler J. Med. Chem. 35 1799 (1992) 57. J A Tino, J M Clark, A K Field, G A Jacobs, K A Lis, T L Michalik, B McGeever-Rubin,W A Slusarchyk, S H Spergel, J E Sundeen, A V Tuomari, E R Weaver,M G Young, R Zahler J. Med. Chem. 36 1221 (1993) 58. T-S Lin,M S Chen, Y-S Gao, I Ghazzouli,W H Prusoff J. Med. Chem. 30 440 (1987) 59. N G Kundu, J S Mahatny, C P Spears, G Andrei, R Snoeck, J Balzarini, E De Clercq Bioorg. Med. Chem. Lett. 5 1627 (1995) 60. N G Kundu, P Das J. Chem. Soc., Chem. Commun. 99 (1995) 61. I Arai, G D Daves Jr J. Heterocycl. Chem. 15 351 (1978) 62. I Basnak, S Takatori, R T Walker Tetrahedron Lett. 38 4869 (1997) 63. S G Rahim,M J H Duggan, R T Walker, A S Jones, R L Dyer, J Balzarini, E De Clercq Nucl. Acids Res. 10 5285 (1982) 64. A Lubineau, J Auge , Y Queneau Synthesis 741 (1994) 65. R F Whale, P L Coe, R T Walker Nucleosides Nucleotides 11 1425 (1992) 66. H A Dieck, R F Heck J. Organomet. Chem. 93 259 (1975) 67. L Cassar J. Organomet. Chem. 93 253 (1975) 68. K Sonogashira, Y Tohda,N Hagihara Tetrahedron Lett. 4467 (1975) 69. M J Robins, P J Barr Tetrahedron Lett. 22 421 (1981) 70. H Li, L Czuchajowski,W R Trumble J. Heterocycl. Chem. 34 999 (1997) 71. N M Goudgaon, G F El-Kattan, R F Schinazi Nucleosides Nucleotides 13 849 (1994) 72. A L Casalnuovo, J C Calabrese J. Am. Chem. Soc. 112 4324 (1990) 73. K Hirota, Y Kitade, Y Kanbe, Y Isobe, Y Maki Synthesis 213 (1993) 74. P Vincent, J-P Beacourt, L Pichat Tetrahedron Lett. 23 63 (1982) 75. P Vincent, J-P Beacourt, L Pichat, J Balzarini, E De Clercq Nucleosides Nucleotides 4 447 (1985) 76. G T Crisp Synth. Commun. 19 2117 (1989) 77. G T Crisp, B L Flynn Tetrahedron Lett. 31 1347 (1990) 78. B L Flynn, V Macolino, G T Crisp Nucleosides Nucleotides 10 763 (1991)A F Nasonov, G A Korshunova 504 79. K W Morin, L I Wiebe, E E Knaus Carbohydr. Res. 49 109 (1993) 80. V Farina, S I Hauck Synlett 157 (1991) 81. D P Matthews, R S Gross, J R McCarthy Tetrahedron Lett. 35 1027 (1994) 82. P Herdewijn, L Kerremans, P Wigerinck, F Vandendriessche, A Van Aerschot Tetrahedron Lett. 32 4397 (1991) 83. S Manfredini, P G Baraldi, R Bazzanini, M Marangoni, D Simoni, J Balzarini, E De Clercq J. Med. Chem. 38 199 (1995) 84. H Matsuhashi, Y Hatanaka,M Kuroboshi, T Hiyama Heterocycles 42 375 (1996) 85. M E Hassan Recl. Trav. Chim. Pays-Bas 105 30 (1986) 86. M E Hassan Bull. Soc. Chim. Belg. 100 627 (1991) 87. I Moritani, Y Fujiwara Synthesis 524 (1973) 88. A E Shilov, G B Shul'pin Aktivatsiya i Kataliticheskie Reaktsii Uglevodorodov (Activation and Catalytic Reactions of Hydro- carbons) (Moscow: Nauka, 1995) p. 277 89. K Hirota, Y Isobe, Y Kitade, Y Maki Synthesis 495 (1987) 90. G A Tolstikov, A G Mustafin, R R Gataullin, L V Spirikhin, V S Sultanova, I B Abdrakhmanov Izv. Akad. Nauk, Ser. Khim. 1616 (1992) b 91. M E Hassan Can. J. Chem. 198 (1990) 92. N Bischofberger Tetrahedron Lett. 30 1621 (1989) a�Russ. J. Bioorg. Chem. (Engl. Transl.) b�Russ. Chem. Bull. (Engl. Tran
ISSN:0036-021X
出版商:RSC
年代:1999
数据来源: RSC
|
4. |
Molecular structure descriptors in the computer-aided design of biologically active compounds |
|
Russian Chemical Reviews,
Volume 68,
Issue 6,
1999,
Page 505-524
Oleg A. Raevsky,
Preview
|
|
摘要:
Russian Chemical Reviews 68 (6) 505 ± 524 (1999) Molecular structure descriptors in the computer-aided design of biologically active compounds O A Raevsky Contents I. Introduction II. Classification of descriptors III. Element-level descriptors IV. Structural formula descriptors V. Descriptors of the electronic structure of molecules VI. Molecular shape descriptors VII. Intermolecular interaction descriptors VIII. Indicator descriptors IX. Modern computer software for the calculation of descriptors X. The necessary properties of descriptors XI. Conclusion XII. Appendix Abstract. The current state of description of molecular structure in computer-aided molecular design of biologically active com- pounds by means of descriptors is analysed. The information contents of descriptors increases in the following sequence: element-level descriptors ± structural formulae descriptors ± elec- tronic structure descriptors ± molecular shape descriptors ± inter- molecular interaction descriptors.Each subsequent class of descriptors normally covers information contained in the pre- vious-level ones. It is emphasised that it is practically impossible to describe all the features of a molecular structure in terms of any single class of descriptors. It is recommended to optimise the number of descriptors used by means of appropriate statistical procedures and characteristics of structure ± property models based on these descriptors. The bibliography includes 371 references. I. Introduction The empirical regularities in the variation of properties of compounds as functions of their structure established by the middle of the XIX century underlay the concept of the struc- ture ± property relationship for chemical compounds. The Peri- odic Law discovered by D I Mendeleev can be considered as the scientific base of the structure ± property concept, and the pre- diction of physical properties and existence of new elements, unknown at that time, is the first stable predictive model based on this relationship.1 Systematic studies on the relationship between the structure and the biological activity of organic compounds started at the end of the XIX century 2 ±4 and by now, they have become widespread.The recognised founder of the research on the O.A.Raevsky Institute of Physiologically Active Compounds, Russian Academy of Sciences, 142432 Chernogolovka, Moscow Region, Russian Federation. Fax (7-095) 913 21 13.Tel. (7-096) 524 50 62. E-mail: raevsky@ipac.ac.ru Received 16 November 1998 Uspekhi Khimii 67 (6) 555 ± 576 (1998); translated by Z P Bobkova #1999 Russian Academy of Sciences and Turpion Ltd UDC 54.02+519.852.6 505 506 506 506 509 509 511 515 515 517 518 518 quantitative structure ± activity relationship (QSAR) carried out nowadays is C Hansch, who reported in 1964, together with T Fujita, a relation that has found the most extensive use in the studies on the structure ± biological activity correlations.2 How- ever, whilst acknowledging the merits of this scientist, one should also remember his numerous predecessors.In particular, it was noted 3 that as early as 1944, N V Lazarev proposed 4 that the partition coefficients of chemical compounds in oil ± water sys- tems be used to establish the quantitative structure ± activity relationships. The intense progress in computational hardware and its adoption for research practice, including that in the field of science in question, promoted the formation of a border line of research � computer-aided design of biologically active com- pounds. Many concepts of this rapidly progressing field are being modified and improved. Therefore, it is necessary to analyse at intervals the current state of studies dealing with this subject.The previous review on this topic 5 was published in this journal in 1988. With allowance for the structure ± activity relationship con- cept, three main stages can be distinguished in the computer-aided design of biologically active compounds. set; (1) Formation of a trial set of compounds having a specified property (activity) or a set of properties; (2) description of the molecular structure of compounds in this (3) elucidation of the structure ± biological activity relation- ship followed by creation of a stable prognostic model suitable for predicting the properties of new compounds. This review analyses the modern approaches to the quantita- tive description of molecular structures of chemical compounds within the framework of computer-aided design of biologically active compounds in terms of independent variables character- ising structural features of molecules or their parts (substituents, fragments). These variables are called descriptors.The progress in computational hardware has a crucial influence on the development of this line of research. It should be noted that computer-aided molecular design faces the problem of processing the data on molecular structures in a unified form with their information content being retained as fully as possible after the data have been entered in the computer.506 Complex natural processes are studied by developing their models. Advantages of these models over the original are that they are easier to investigate, can be better manipulated and monitored and that the investigations are cheaper.6 The models can be classified into iconic (ensuring similarity regarding the form), analog (ensuring similarity regarding the form and functions) and abstract (symbolic, conceptual) models, which provide the corre- lation with the original by virtue of symbols. The modern views on the molecular structure are certainly not exhaustive and, in the future, they will be improved with the aid of various models. Here, it is pertinent to quoteM Dewar, who said that we don't know and, perhaps, shall never know what molecules actually are.Our understanding of molecules is based on models, which reproduce their properties sufficiently well to be useful.6 Description of models, which are used in chemical research, can be found in the literature.6±8 In this review, molecular structure descriptors are considered briefly with allowance for the above classification of models.II. Classification of descriptors Various types of classification of descriptors are possible. Three types of classification are considered in a paper by Charton.9 According to the first type, descriptors are classified into simple (describing one effect of interatomic interactions) and composite ones (describing two or more effects). Composite descriptors, in turn, can be divided into unicomposite (describing effects of the same type) and multicomposite descriptors (reflecting the sum of effects of various types). The second classification is based on the way in which descriptors are estimated, i.e., experimental and theoretical ones.The third classification takes into account the sort of effect described by a given descriptor. Three sorts of effects are distinguished � electronic (electric), steric and intermolecular effects. A specific case of using this type of classification is provided by the classical Hansch equation,2, 10 ± 14 which repre- sents biological activity as a function of Hammett constants (electronic effects), Taft constants (steric effects) and hydropho- bicity constants (transport effects arising due to intermolecular interactions). For visualisation of the description of chemical structures, Testa and Kier 15 used geometrical figures corresponding to the level of structure being considered: element level ? two-dimen- sional structure?three-dimensional structure?bulk properties ? stereodynamic structure ? stereoelectronic structure ? interactions with the environment.The information content of each subsequent level embraces the information content of the preceding one. Apparently, a similar approach in a somewhat different form could be used in this review to discuss descriptors. Figure 1 shows a chart for several levels of descriptors in which each subsequent level covers the information content of the preceding one. In this representation, it is asd that structural formula descriptors carry the whole body of information con- tained in the element-level descriptors; electronic structure descriptors carry the information contained in the structural formula descriptors and descriptors of intermolecular interac- tions cover the whole body of information provided by all of the preceding levels.In some cases, descriptors do not bear the information of the previous level (e.g., some descriptors of intermolecular interactions such as polarisability and hydropho- bicity are often calculated without allowance for the three-dimen- sional structure), but in general, this approach to the classification of descriptors is visual and reflects the degree of their complexity and, thus, it is convenient for the analysis. In this review, each class of descriptors is considered. It should be noted that a descriptor of any level can characterise both a molecule as a whole and its part (fragment, screen, functional group, substituent).O A Raevsky Intermolecular interaction descriptors Molecular shape descriptors Electronic structure descriptors Structural formula descriptors Element-level descriptors Figure 1. Information contents of molecular structure decriptors. III. Element-level descriptors The gross formula bears the information about sorts of atoms that make up the molecule and the number of atoms of each sort. The only property that can be predicted precisely based on the gross formula is molecular weight (MW), which can be regarded as an element-level descriptor. The number of atoms of a particular sort is sometimes taken as another descriptor of this level.16 An example of using such descriptors can be found in the Appendix [see Eqn (A.1)].A legitimate construction of the structure ± property correla- tion should be carried out for related compounds. Therefore, the structure variation within a series of compounds being considered reduces to variation of substituents. Hence, the properties of substituents can be regarded as structural descriptors. In partic- ular, the atomic weights of substituents can also be considered to be element-level descriptors. Generally, descriptors of this level carry too little information on the molecular structure and, hence, they cannot be used alone for elucidation of the real structure ± property relationships. IV. Structural formula descriptors The structural formula of a molecule is the most widespread way of describing a chemical compound.As a model of molecular structure, this formula contains elements of iconic, analog and abstract models 6 and provides the basis for the construction of many types of descriptors, most of all, topological descriptors (Fig. 2). 1. Topological indices Topological descriptors (indices) are calculated based on the description of a structural formula of a compound by a molecular graph G, which is a two-dimensional representation of the molecule (the vertices correspond to atoms and the edges match the chemical bonds).17 Normally, the skeleton atoms and the bonds between them are considered, hydrogen atoms being erased. The matrix form of the graph is used to estimate the topological indices.The adjacency matrix A(G) and the distance matrix D(G) are used most often. The adjacency matrix elements aij are equal to either unity or zero depending on whether or not vertex i of graphG is incident to vertex j. The number of units in an ith line or jth column of the matrix is equal to the vertex degree.Molecular structure descriptors in the computer-aided design of biologically active compounds Distance matrix Adjacency matrix Topological indices Figure 2. Elements of molecular structure description based on the structural formula. Each element in the distance matrixD(G) represents the number of edges connecting vertex i to vertex j via the shortest route and is denoted by dij.The adjacency and distance matrices for pentane, 2-methylpentane, cyclopentane and furan are shown in Scheme 1. Note that the adjacency and distance matrices for the two last- mentioned compounds are identical because the molecular graphs are identical. Compound Molecular graph CH3(CH2)3CH3 CH3 CH3CH(CH2)2CH3 Cyclopentane OThe most commonly used topological indices calculated from the adjacency matrix are the Platt index F(G) (the sum of degrees of each edge in graph G); the Gordon ± Scantlebury index Y(G) (the number of paths of length 2); the full adjacency index A0(G) (the sum of all the non-zero elements in the adjacency matrix); the Randic connectivity index w(G) (it will be discussed below); and the Zagreb group indices M1(G) and M2(G). The indices that should be mentioned as the descriptors calculated from the distance matrix are the Wiener index W(G), the Hosoya index Z(G), the polarity number P(G), the distance sum index S(G) and the index of distances between the vertices VDI(G).The formulae for the calculation of these widely used descriptors and other topological indices estimated based on the adjacency and distance matrices as well as examples of using these descriptors in the Structural formula Other topological matrices Information topological indices Scheme 1 Distance matrix Adjacency matrix 1 2 3 4 5 1 0 1 0 0 0 2 1 0 1 0 0 3 0 1 0 1 0 4 0 0 1 0 1 5 0 0 0 1 0 1 2 3 4 5 1 0 1 2 3 4 2 1 0 1 2 3 3 2 1 0 1 2 4 3 2 1 0 1 5 4 3 2 1 0 1 2 3 4 5 6 1 2 3 4 5 6 1 0 1 0 0 0 0 1 0 1 2 3 4 2 2 1 0 1 0 0 1 2 1 0 1 2 3 1 3 0 1 0 1 0 0 3 2 1 0 1 2 2 4 0 0 1 0 1 0 4 3 2 1 0 1 3 5 0 0 0 1 0 1 5 4 3 2 1 0 4 6 0 1 0 0 1 0 6 2 1 2 3 4 0 1 2 3 4 5 1 0 1 2 2 1 2 1 0 1 2 2 3 2 1 0 1 2 4 2 2 1 0 1 5 1 2 2 1 0 1 2 3 4 5 1 0 1 0 0 1 2 1 0 1 0 0 3 0 1 0 1 0 4 0 0 1 0 1 5 1 0 0 1 0 1 2 3 4 5 1 0 1 2 2 1 2 1 0 1 2 2 3 2 1 0 1 2 4 2 2 1 0 1 5 1 2 2 1 0 1 2 3 4 5 1 0 1 0 0 1 2 1 0 1 0 0 3 0 1 0 1 0 4 0 0 1 0 1 5 1 0 0 1 0 507 Linear nomenclatures Structural formula fragments Bond matrices structure ± property correlations can be found in original publica- tions.18 ± 20 The above matrices and topological indices reflect only topological features of molecules without allowance for the types of atoms and the real distances between them.The imperfection of this description of the structures of compounds becomes clear if one recalls that the adjacency and distance matrices (and, hence, all the topological indices mentioned above) for cyclopentane and furan coincide. In addition to the adjacency and distance matrices described above, the extended distance matrix E,21 connectivity matrix C22 and Wiener matrix W23 are used. Many researchers use various descriptors estimated based on connectivity matrix C; therefore, we shall briefly consider them here. In 1975, Randic proposed that the connectivity index w(G) be determined in the following way:24 (1) where ni and nj are the degrees of vertices i and j in graph G, w(G)=PÖvivjÜ¡1=2, respectively.Summation is carried out over all edges of graph G. Kier and Hall 25 have employed the Randic scheme as the basis for the development of a general method for the description of structures of organic compounds and proposed the concept of molecular connectivity for this purpose.25 The first equation for a topological index was as follows: 1 (2) where d corresponds to the number of neighbouring atoms (with X=P(di dj)71/2, hydrogen atoms being omitted). This equation is identical to Eqn (1). However, the researchers made this approach more versatile by extending it not only to bonds but also to atoms (0X=P(d)71/2) and to fragments with several bonds.For example, when a molecular graph is cut into two-bond frag- ments, the index 2X=P(didj dk)71/2 is estimated, when it is cut into three-bond fragments, the index 3X=P(di dj dk dl)71/2 is determined and so on. In addition, calculation of these indices for molecules containing heteroatoms was started. The hydrogen atoms attached to heteroatoms were ignored (like those attached to carbon). (3) Later, Kier and Hall 26 proposed a general equation for the estimation of d for any atom dv=Zv7h, where Zv is the number of valence electrons with allowance for the electrons of lone electron pairs and h is the number of hydrogen atoms attached to the given atom (which are present in the structural formula but lacking in the molecular graph).Thus the dv value for a carbon atom of benzene is equal to three (the atom has four valence electrons and is linked to one hydrogen atom); the dv value for the oxygen atom in alcohols is equal to five (the atom508 has six valence electrons and is linked to one hydrogen atom) and that for the oxygen atom in ethers is six (the same number of valence electrons but no hydrogens at the heteroatom). In the subsequent studies, Kier and Hall 27 ± 30 proposed using electronic state indices (S) to describe quantitatively the mutual influence of atoms. The S value for any atom is estimated by the formula (4) S=I+DI, ij where I=(dv+1)/d and DI=(Ii7Ij)/r2 (r is the number of bonds between the two atoms considered). The indices of the electronic topological state for some substituted amines are presented below (Scheme 2).Scheme 2 11.12 0.62 2.18 0.62 0.85 1.29 F NH2 NH2 1.28 1.19 0.82 70.23 5.05 5.21 11.62 F 0.16 0.54 0.46 70.54 70.16 70.73 2.64 70.79 NH2 NH2 NH2 1.78 1.41 0.49 1.51 4.67 4.91 4.99 F F F 11.76 11.76 11.74 The Kier and Hall indices have been used rather widely in the search for quantitative structure ± property relationships [see, for example, the Appendix, Eqns (A.2) and (A.3)]. These indices became very popular because they can be easily calculated and provide clear representation of the structural formulae of com- pounds. Several recent publications should be noted.31 ± 39 However, description of the electronic structure by the number of valence electrons and description of interatomic distances by the number of bonds between the atoms are too simplified to ensure elucidation of the true molecular structure.This accounts for the criticism levelled at the attempt of using these indices for establishing the real structure ± biological activity correlations.40 A commonly used index calculated from the distance matrix is the Wiener indexW(G), which is the first topological index used in chemistry. It is defined as the half-sum of the elements of the distance matrix D. The Wiener index and related topological indices are often used in the structure ± property studies to characterise the compactness of a molecule.41 ± 43 However, it should be borne in mind that in matrix D, the distance between atoms is expressed as the number of bonds separating these atoms.In order to bring the structure description in terms of the distance matrix closer to the real situation, it has been proposed to use averaged (typical) distances between atoms in molecules placed in a three-dimensional grid 44 ± 47 or distances found from experi- mental data or by quantum-chemical calculations 48, 49 as the off- diagonal matrix elements. Molecular graphs serve as the source of diverse topological indices. Thus Kier 50 ± 54 proposed various kappa-indices for coding the shape or configuration of molecules. Description of some other new indices can be found in several publications.55 ± 64 Particular correlations of these topological indices with various physical properties and biological activity have been consid- ered.65 ± 72 Information topological indices have also been used for structure description. The estimation of these indices is based on the separation of the set of graph vertices into equivalence classes taking account of the closest environment of the vertices.18, 73 Within the framework of this approach, a number of topological indices of the theoretical-information type have been devised, namely, 1IC (information content), 1SIC (structural information content), 1BIC (binding information content) and 1CIC (comple- mentary information content), which are calculated from the following equations: O A Raevsky i (5) log2 nni, 1IC=¡Xnn 1SIC= (6) log 1IC2n , 1BIC= (7) 1IC2q , (8) log 1CIC=log2 n71IC, where ni is the number of vertices in the ith equivalency class, n is the number of graph vertices for the molecule and q is the number of graph edges.Examples of using the information topological indices for describing structures of organic compounds in eluci- dation of structure ± property correlations including structure ± biological activity correlations have been reported.74 ± 80 2. Encoding of structural formulae by linear nomenclatures and bond matrices The necessity of automating the processing of information on the structures of chemical compounds, the amount of which has sharply increased in recent years, initiated the development of some new methods for representation of the structural formulae of chemical compounds, namely, various linear nomenclatures and bond matrices, which can be regarded as abstract models of molecules.6 According to the linear nomenclature, a structural formula is encoded by a string of characters, which, in turn, encode particular fragments of the structure. Each linear nomenclature is characterised by a definitely specified sequence of characters.The character set of a linear notation usually includes standard characters. The initial stage of development of linear nomencla- tures is associated with the name of Wiswesser. As an example, we present the Wiswesser codes for the structures of some com- pounds: (3O2), CH37CH27O7CH27CH27CH3 (Z2VQ), H2N7CH27CH27C(O)OH HOCH2CH(NH2)CH2CH2CH(CH2NH2)CH2OH (Z1Y1Q2YZ1Q).The Wiswesser linear nomenclature has been described in detail.16 In Russia, a similar approach is practised by Avidon,82, 83 who uses the SSFC language (substructure superposition fragmental code). At present, the most widespread linear nomenclature used to represent structural formulae of organic compounds is contained in the SMILES system,84 ± 86 which is used, in particular, to reveal the structural similarity of compounds.87 A linear nomenclature suitable for the description of not only the structural formula but also the stereoisomerism also deserves attention.88 In this nomen- clature, a compound is represented by a canonical line code consisting of two parts; the first part describes the structural formula and the second part characterises the stereochemical features of the molecule.When a structural formula is described using a connectivity matrix, each atom, each bond and each type of atoms are coded directly. Connectivity matrices are similar to adjacency matrices and distance matrices. The main diagonal of a connectivity matrix normally includes codes for the atoms in the structural formula, while off-diagonal elements are the codes for the types of bonds. These matrices were used rather widely in the 1970s and 1980s; they have been described in a monograph.16Aconnectivity matrix of descriptor centres (CMDC) has been proposed;89 in this matrix, the codes of atoms are replaced by codes of functional groups (descriptor centres) and the off-diagonal elements correspond to the numbers of bonds between the descriptors instead of the bond codes.In order to perfect the language of structure description in terms of CMDC, later it has been proposed to use computable distances between the descriptor centres.90Molecular structure descriptors in the computer-aided design of biologically active compounds 3. Structural fragments as descriptors The topological indices and various matrices described above provide certain information on a chemical compound as a whole. However, the crucial role in the interaction of a compound with other chemical and biological objects is sometimes played by separate fragments of molecules; therefore, in the elucidation of the structure ¡À property and structure ¡À biological activity corre- lations, the structures of compounds of a given series are often described fragment-by-fragment.A structural fragment can comprise one atom, a substituent, a functional group or a particular combination of atoms (with indication of the hybrid- isation types).91 Each structural fragment can be used as an independent descriptor. Further detailing can be attained by specifying the local characteristics of atoms forming the fragment (see, e.g., classification of the descriptor centres in terms of their donor ¡À acceptor properties).92, 93 Numerous approaches to the computer-aided construction of new substances based on struc- tural fragments of compounds having a common property have been developed.Various substructure analysis descriptors,94 ¡À 97 topological pharmacophores in the LOGANA programme pack- age procedure,98 ¡À 100 substructure descriptors,101, 102 descriptors of the structural language of the DARC system,103, 104 chemical functionality descriptors 105 and de novo design descriptors 106 can be cited as examples. V. Descriptors of the electronic structure of molecules Quantum-chemical calculations for complex molecules make use of atomic and molecular quantum-chemical descriptors. The methods for calculation, the potential and limitations of these descriptors have been described in detail in a review.107 In this Section, we consider briefly only quantum-chemical descriptors related to intramolecular electronic properties.Atomic charges are quantum-chemical descriptors used most often to elucidate the structure ¡À property correlations. If chemical interactions are divided into electrostatic and covalent compo- nents, atomic charges characterise the electrostatic component. The full charge on an atom can be regarded as a non-directional descriptor, whereas the s- and p-electron densities characterise the possible orientation of chemical interactions; therefore, they can be regarded as directional descriptors.108 Examples of successful use of these descriptors can be found in the litera- ture.109 ¡À 111 When the charges on atoms are known, a new set of descriptors of this type can be used in each particular study. The energies of the highest occupied (EHOMO) and lowest unoccupied (ELUMO) molecular orbitals are descriptors that are frequently used in calculations of electronic structures of mole- cules.112 ¡À 114 It was found that in some cases, the EHOMO values are directly correlated with the ionisation potential (I) and characterise the susceptibility of the molecule to electrophilic attack.The ELUMO values correlate with the electron affinity (A) and characterise the susceptibility of the molecule to nucleophilic attack. Note that I and A are also used as independent electronic structure descriptors.115, 116 The EHOMO7ELUMO difference is believed to reflect the molecule's stability.117 The notions of softness and hardness of molecules are associated with exactly these electronic structure descriptors.118 The molecular hardness descriptor (Z) (9) Z=I ¡¦ A 2 has also found use in the structureDproperty correlations (see, for example, Ref.119).{ { Hardness as a measure of energy stabilisation has been defined by some researchers as the half-sum of the EHOMO and ELUMO values 120 or as the difference between EHOMO and ELUMO taken with a minus.121, 122 509 Dipole moment and related descriptors. The polarity of a molecule is described by its dipole moment (m). The magnitude of this descriptor can be determined experimentally or calculated using a quantum-chemical method. This descriptor is used rather widely in elucidation of structure ¡À property correla- tions.12, 112, 123D125 However, the vector summation of the dipole moments of several polar groups in a molecule can result in a small value.Therefore, in addition to m, the greatest difference between the charges of two atoms, the topological electronic index, the local dipole index, the quadrupole moment tensor, etc. are sometimes used as polarity descriptors.107 Examples of using electronic structure descriptors in struc- ture ¡À biological activity relationships are given in the Appendix [Eqns (A.4) and (A.5)]. VI. Molecular shape descriptors It is now obvious that features of the three-dimensional structure of molecules need to be taken into account in elucidating the structure ¡À property relationsips. Suffice it to say that almost all of the modern automated systems for the search for new compounds contain, as obligatory elements, various procedures for compar- ison of the spatial structures of molecules.126, 127 The description of the three-dimensional structure of a molecule by a single quantitative descriptor (or a set of descrip- tors) is a fairly complicated task.Let us consider the simplest situation when the compounds under consideration have the same framework and different substituents at the same position. In 1976, Verloop and coworkers proposed a set consisting of five steric constants (STERIMOL parameters) for 243 different substituents; the constants were calculated based on standard bond lengths and bond angles.128 These simple parameters have found fairly wide use in structure ¡À property studies.129 ¡À 131 A similar approach based on estimation of shape parameters of substituents has been proposed in another study.132 Amoore 133 was the first to analyse the projections of molec- ular structures onto the three orthogonal planes in a study dealing with the relationship between the structure and odour of com- pounds.An approach describing the shape of a molecule in terms of six descriptors S1 ¡À S6 has been proposed.134 The main problem arising in the comparison of 3D structure parameters of various molecules is associated with conforma- tional mobility. The appearance of an additional internal rotation axis results in an exponentially increased number of possible conformations; therefore, certain restrictions are required as well as special procedures for selecting real conformations and comparing their shapes.Several methods for the estimation of molecular shape descriptors currently exist. These methods are briefly described below. The distance geometry approach has been developed since the late 1970s, being constantly modified by Crippen et al.135 ¡À 141 The consecutive steps of the procedure used to compare the 3D molecular shapes can be outlined as follows. 1. Generation of 3D structures for the compounds considered; the initial structure is based on the bond lengths and angles found in an X-ray diffraction experiment. 2. Ascribing of definite physico-chemical properties to atoms in the molecules. 3. Elucidation of low-energy regions in the conformational space for each of the compounds. 4.Determination of the geometrically possible superpositions of molecules onto the reference structure. First, all combinations of three atoms in the reference structure and in the molecule being analysed are considered. After the best correspondence of dis- tances between some three-atom groups in the pair of compounds being compared has been found, superposition of these three pairs of atoms is fixed and the process of spatial superposition is continued for other atoms.510 5. Analysis of a great number of `good' superpositions of compounds on the reference structure makes it possible to compare the molecular shapes of the compounds; as this is done, the concept of the molecular shape descriptors is visualised and the shape of the binding site in a potential receptor is evaluated.The binding site approach.142 ¡¾ 144 This approach represents the binding sites of a potential receptor as non-overlapping areas covering the whole space rather than as points (as in the previous method). With such representation of a spatial structure, each atom should be located in one (and only one) area and the binding site model should include the list of areas (descriptors) in which each atom is located. Fig. 3 a shows the binding site model calculated for compounds exhibiting competitive properties in relation to the protein isolated from mouse liver,145 and Fig. 3 b presents the optimum binding model for benzopyrene. a b 12.7A E 3.9 AE Figure 3.Binding site model for compounds exhibiting competitive properties towards a protein (a) and optimum binding model for benzopyrene (b).145 Molecular shape analysis. This approach has been developed during the last 15 years in the studies of Hopfinger et al.146 ¡¾ 155 The procedure for the quantitative comparison of the spatial structures of the molecules of various compounds includes six consecutive stages. 1. Conformational analysis with allowance for the internal rotation about all the single bonds and calculation of the energy for each 10 8 increment. This is done using methods of molecular mechanics with a force field comprising contributions of disper- sion, steric and electrostatic interactions and, whenever possible, hydrogen bonds.2. Selection of individual conformations. The real set of possible conformations for each compound is determined by the chosen difference between the conformation considered and the global energy minimum. 3. The choice of the reference compound and the active conformation by pairwise comparison of the molecular shapes in the series of compounds under consideration. 4. Superposition of each molecule in the series of compounds under consideration onto the molecule of the reference com- pound. 5. Quantitative comparison of molecular shapes. The follow- ing descriptors characterising the molecular shape similarity have been proposed: the common overlap steric volume (V0) determined from the following equation (10) V0=Vu\Vv, where Vu and Vv are the space occupancy by the molecules of the compound considered and the reference compound, respectively; the relative measure of the degree of similarity defined as the function O A Raevsky (11) f0 a Vu \ Vv.Vu These descriptors do not contain information regarding the intramolecular stability of each conformation of the pair of compounds. Therefore, one more descriptor, the shape common- ality index (Ic), has been proposed: (12) uODEu a DEv a eUa1=2 a DEu Ic a SOu;v;wU ¢§ aDE a aDEvODEu a DEv a eUa1=2 , DEv where S(u,v,w) is wV0 or wf0, w is the shape similarity parameter, DE is the difference between the conformation energies, DEv = 1 kcal mol71, DEu is taken to be 1, 3 and 6 kcal mol71 in three separate calculations and e is a correction factor. The hypothetical active site lattice.156 ¡¾ 160 This method is based on the assumption that a molecule can be described by descriptors characterising the space that it occupies and the nature of atoms that occupy this space. Three-dimensional space is defined as a set of regularly arranged points in a three-dimensional lattice.Trans- lation of the molecule to the set of these points occurs with retention of those points of the lattice that fall within the van der Waals radius of each atom in the molecule. The resulting molecular lattice consisting of points carries information on the type and properties of atoms in the molecule. The type of atoms is defined simply: the value +1 is attributed to atoms rich in electrons, the value 71 is attributed to atoms poor in electrons and electrically neutral atoms are denoted by 0.Estimation of molecular positions.161 ¡¾ 163 In this approach, the comparison of spatial shapes of molecules also starts with the search for low-energy conformations and selection of energeti- cally stable shapes for each compound in the series under consideration. Molecular position is defined as a specific con- formation associated with the activity exhibited by this com- pound. The molecular position is described by several descriptors. The weighted holistic invariant (WHIM) molecular descrip- tors are constructed in such a way as to cover the information on the three-dimensional structure as fully as possible including the size, shape and symmetry of molecules and the distribution of atoms in them.164 ¡¾ 169 The calculation algorithm includes the analysis of the main components of the centred molecular coordinates using the weighted covariant matrix derived from various weighted schemes for atoms.Different covariance matrices and principal axes are obtained depending on the type of the weighted scheme used. For example, when the atomic weights are used as the weighted scheme, the directions of the three principal axes coincide with the principal inertia axes. For each weighted scheme, a set of statistical indices of atoms projected onto each principal component tm (m=1, 2, 3) is calculated. The invariance of the calculated parameters with respect to translation is attained by the procedure of centring the atomic coordinates, while the invariance with respect to rotation is ensured by using the principal components.The directional WHIM descriptors are statistical indices calculated based on each individual principal component (1, 2, 3). The non-directional WHIM descriptors are found directly from directional descriptors. The latter type of descriptors characterise the overall shape of the molecule, any information related to the principal axes being lost in this case. Eleven directional descriptors (ll, l2, l3, Wl, W2, g1, g2, g3, Z1, Z2, Z3) characterising the molecular properties within each weighted scheme have been proposed.164 ¡¾ 169 Thus, the total number of directional WHIM descriptors is 66.In the case of planar structures, only eight directional (ll, l2, Wl, W2, g1, g2, Z1, Z2) andMolecular structure descriptors in the computer-aided design of biologically active compounds five non-directional descriptors (T, A, V, K, D) should be taken into account for each weighted scheme. Analysis of the shape of electron cloud.170�¢174 This method is based on the assumption that the shape of the electron cloud carries the whole body of information on the properties of compounds. This approach was used primarily to analyse molec- ular surfaces. The surface of a molecule is cut into areas according to the local curvature properties. Convex D2, saddle-like D1 and concave D0 local areas are distinguished.The ratios of these de- scriptors determine groups of molecular shapes, which include two- one- and zero-dimensional truncated surfaces obtained from the initial surface by removing all areas of a given type (D2, D1 or D0). The quantitative measure for the similarity of shapes of 3D molecules can be found by comparing the numerical codes of shapes. This procedure is performed automatically using a com- puter without subjective visual comparison of the shapes of objects. Examples of using molecular shape descriptors in structure ¡¾ biological activity correlations are given in the Appendix [Eqns (A.9) and (A.12)]. VII. Intermolecular interaction descriptors Quantitative description of the intermolecular interactions in the elucidation of structure ¡¾ biological activity correlations can be accomplished using the free energy difference (DG).175, 176 DG=Gf7Gi, (13) where Gf and Gi are the free energies of the final and initial states, respectively.A more detailed quantitative description of intermolecular interactions is provided by using the enthalpy (DH) and entropy (DS) differences, which are related to DG as follows: (14) DG=DH7TDS, where T is the reaction temperature. By measuring experimentally the equilibrium constant (K), one can determine DG (15) DG=DG87RTlnK, where DG8 is the free energy change for a reaction under standard conditions. The DG, K, DHand DS values for a reaction or interaction can be regarded as descriptors of intermolecular interactions.For example, dissociation constants of acids and bases (pKa) are used fairly frequently in the modelling of structure ¡¾ activity correla- tions.177 ¡¾ 180 A large group of intermolecular interaction descriptors estimated from experimental data on reactivity comprises various constants describing the electronic influence of substituents on the reactivity of chemical compounds. This line of studies was initiated by Hammett who defined the parameter s (Hammett constant) based on the relation 181 (16) sX=logKX7logKH, where KX is the corresponding constant for meta- (sm) or para- substituted (sp) benzoic acid, KH is the ionisation constant for benzoic acid in water at 25 8C. By now, a great number of constants have been proposed to describe the electronic effects of substituents. For example, the inductive (si) and resonance (sr, s+, s7) constants, the inductive Taft constant (s*) and the Swain and Lupton constants 182 ¡¾ 184 are widely used as descriptors ointra- and inter-molecular interactions. The reactivity descriptors, calculated by quantum-chemical methods, constitute an important group of intermolecular inter- action descriptors. First of all, mention should be made of indices of electrophilic (Se) and nucleophilic (Sn) superdelocalisabilities on a reaction centre (r), which are estimated from the equations , (17) SeOrU a 2 k OCskU2 0:5OEHOMO a ELUMOU ¢§ Ek sOrU Xvac X 511 (18) siU2 E SnOrU a 2 i i ¢§ 0:5OEHOMO a ELUMOU , OC sOrU Xocc X where the numerators contain coefficients for the atomic orbitals and denominators contain the energies of the corresponding molecular orbitals.Some other descriptors such as self-atom polarisability p(r),185 reflecting the influence of the variation of electronegativity of atomic orbitals on the charge of the same atom si C2sk (19) pOrU a 4 i k sOrU Xocc Xvac XC2 k ¢§ Ei , E and atom ¡¾ atom polarisability p(AB),107 reflecting the effect of a perturbation at one atom (A) on the charge of another atom (B) CAp CApa CBri CBa(20) , pOABU a 4 r p a i Ei ¢§ Ea XXXX can be used as well to describe intermolecular interactions. These descriptors have also found application in the descrip- tion of structure.107, 186 ¡¾ 188 Molecular refraction and polarisability.When a molecule is placed in an external electrostatic field, the charge distribution in it changes. The resultant dipole moment is determined from the relation (21) m=aE, where a is the molecular polarisability and E is the electric field intensity. The polarisability can be estimated experimentally based on the expression relating it to the molecular refraction (MR) (22) MR a 4pNAa , 3 (NA is the Avogadro number) and the relation of the molecular refraction to the refractive index (n) (23) MR a MVn2 ¢§ 1 n2 a 2 , where MV is the molecular volume. The MR and a values are estimated either based on the additivity of the experimentally determined contributions of separate atoms and fragments 189 or by quantum-chemical calculations.191 ¡¾ 194 Lipophilicity. The group of intermolecular interaction descrip- tors used most commonly to analyse the structure ¡¾ activity correlations is related to the notion of lipophilicity, which characterises the transport properties of compounds in biological objects.In the vast majority of cases lipophilicity is described quantitatively in terms of the partition coefficients (P) of com- pounds in the model octanol ¡¾ water system. The log P value for the neutral form of the compound is used as the descriptor for this system. The corresponding descriptors for the cationic, anionic and zwitter-ion forms are designated by log P+, log P7 and log P+/7.195 The partition of ionised compounds between octa- nol and water depends on pKa of the compound itself and pH of the medium.The distribution of substances in biological media is characterised by the index logD with indication of the pH, instead of log P. In the vast majority of publications, logD refers to a physiological medium with pH 7.4. The relationship between the descriptors logD, log P, logP+, log P7 and pKa is given by the following expressions:196 for the equilibrium XH X7+H+, (24) logD=logPXH7log (1+107pKa+pH); for the equilibrium X+H+ XH+, (25) logD=logPX7log (1+107pKa7pH);512 XH+H+ for the equilibria X7+H+ XHa2 , (26) logD=log (A0+A1+A2), where PX¢§ A (27) 0 a 1 a 10pKa2¢§pH a 10pKa1apKa2¢§2pH , 1 a 1 a 10pH¢§pKa2 a 10pKa1¢§pH , PXH PXHa 2 a 1 a 10pH¢§pKa1 a 10 2 2pH¢§pKa2¢§pKa1 .A (28) A (29) The experimental methods used to estimate the lipophilicity parameters have been described in detail in reviews.197 ¡¾ 199 Lipophilicity can be calculated using several schemes, for example, based on the additivity of the contributions of lipophilic constants of atoms 200 or fragments 202, 202 or using conformation- ally dependent approaches,203, 204 molecular descriptors 205 and various solvatochromic parameters.206 The octanol ¡¾ water parti- tion of compounds is mainly governed by two factors �¢ the molecular volume and the capacity of molecules to form hydrogen bonds.207, 208 The contribution of the molecular volume to the lipophilicity has been described quantitatively 209, 210 using several descriptors such as molecular weight MW, molecular volume MV, total molecular surface area SA, the integral of the van der Waals interaction spectra VWI, molecular refraction MRand molecular polarisability a.To describe quantitatively the contribution of hydrogen bonds, free energy factors for hydrogen bond donors and acceptors Cd and Ca were employed, which are considered in greater detail below. Since the descriptors corresponding to the contribution of the molecular volume are correlated with one another, only one of them can be included in the regression equation. The best result was obtained by using the polarisability and the sum of the factors for the hydrogen bond acceptor capacity [see the Appendix, Eqn (A.14)].This approach to the quantitative description of the lipophilicity makes it possible to estimate the contributions of polarisability and hydrogen bonding acting in the opposite directions and thus to modify the structure of compounds in order to optimise their properties. Table 1 presents the values for lipophilicity of some chemicals determined experimentally and calculated by Eqn (A.14) and the contribu- tions to the lipophilicity of the two parameters mentioned above. Analysis of numerous computer programmes meant for the calculation of log P (see 211 ¡¾ 213) demonstrated that they provide satisfactory accuracy in the calculation of the lipophilicities of Table 1. Experimental and calculated values for lipophilicity and polar- isability and acceptor capacity contributions in the formation of hydrogen bonds.210 logP Compound a a PCd PCa calc- exper- ulation iment 8.00 72.57 1.99 70.59 70.53 5.59 71.79 1.39 70.41 70.12 6.43 72.82 1.60 71.23 71.30 4.16 5.88 13.19 73.98 3.28 70.71 70.37 3.00 11.32 72.04 2.82 0.77 0.59 1.50 10.44 71.02 2.60 1.57 1.77 10.90 21.44 77.39 5.34 72.06 71.52 24.10 60.97 716.34 15.18 71.17 70.84 5.78 10.32 73.92 2.57 71.36 70.85 3.14 0.04 0.64 2.47 Chloracetamide 3.80 Acetaldehyde oxime 2.64 Ethanolamine Nicotinamide Hydroquinoline Dimethyl sulfite Acyclovir Sephopyrazone Fluorouracil Hydroflumethiazide 9.31 25.16 76.31 6.26 70.06 Paracetamol 5.49 16.01 73.72 3.99 0.26 Warfarin 7.41 32.83 75.02 8.17 O A Raevsky simple neutral organic compounds but in the case of complex compounds, especially medicines, the accuracy is fairly low.This may be due to the fact that most of the calculations are based on the additivity of the logarithms of the contributions of atoms and fragments rather than on an accurate additive scheme for the contributions of the octanol (oct) ¡¾ water partition coefficients of compounds. In addition, only a few experimentally determined lipophilicities of compounds containing several cationic and/or anionic groups have been published. Meanwhile it is these data that could provide the basis for precise calculations of the lipophilicities of complex ionised compounds. The application of Eqn (A.14) to the calculation of log Poct values for various neutral organic compounds has resulted in the following expression (30) logPoct(exp.)=1.00 (0.02) logPoct(calc.), N=2781, R=0.971, S=0.32 (see {).Comparison of the lipophilicities calculated from Eqn (A.14) with those determined experimentally for compounds that had been tested with 14 appropriate commercial programmes gave the following results: for 90 simple organic compounds, R = 0.993; for 48 medicines, R = 0.935 and for all the 138 chemicals tested, R = 0.971. These statistical criteria are commensurable with those attained in the best of the programmes tested. This out- come, together with the possibility of calculating the full pH- dependent profile for ionised compounds, attests to the value of the above-mentioned physico-chemical approach to the estima- tion of lipophilicity.In rent years, in addition to the above `classical' descriptors, another descriptor called molecular lipophilicity potential (MLP),214 ¡¾ 218 which quantitatively describes lipophilicity in three-dimensional space, has been widely used. The lipophilicity potential at a given point in space is a result of intermolecular interactions governed by the lipophilicity of all fragments of the molecule. Two components are needed to calculate the lipophi- licity potential, namely, the fragment lipophilicities and a function of distance describing the variation of lipophilicity in space 214 f MLP (31) i fctOdikU, k a ia1 XN where k is the label for a given point in space, i is the label of the fragment, N is the total number of fragments in the molecule, fi is the lipophilicity constant for fragment i, fct is the distance function and dik is the distance between fragment i and point k in space.A large number of examples in which the lipophilicity parameters have been used successfully as intermolecular inter- action descriptors in the elucidation of structure ¡¾ property correlations have been reported in a book 184 and in papers.214 ¡¾ 224 Examples of using log Poct in structure ¡¾ property (activity) regression equations are given in the Appendix [Eqns (A.14) ¡¾ (A.16)]. It should be noted that the majority of procedures used to calculate both descriptors of intermolecular interactions and energies of reactions are based on the assumption of additivity of contributions of various types of intermolecular interactions 175 (32) DG=DGsolvent+DGconf+DGint+DGmotion, where DGsolvent is the solvent contribution, DGconf is the change in the free energy caused by conformational changes in the reacting molecules, DGint is the change in the free energy due to specific interactions and DGmotion is the change in the free energy of the molecule caused by changes in the translational, rotational and vibrational constituents. { From here on, N, S and R are statistical criteria of equations: N is the number of observations, S is the standard deviation and R is the correlation coefficient; the Fischer criterion F is also used.Molecular structure descriptors in the computer-aided design of biologically active compounds Specific interactions are calculated on the assumption that the contributions of various types of interaction obey the additivity principle.At present it is believed that the greatest contributions to the specific interactions are made by steric (st) and electrostatic (el) effects and hydrogen bonds (hb). Within the framework of this assumption, the free energy of specific interactions can be represented as follows:225 (33) DGint=DGst+DGel+DGhb . Below we consider the parametrisation of these three types of specific interactions, analyse the existing descriptors and describe computer programmes used to calculate the energies of these interactions. 1. Parametrisation of steric effects The first steric parameter Es was proposed by Taft.226 When studying the reactivity of esters, he obtained the following relation: E (34) s a log kR , kMe where kR and kMe are the rate constants for hydrolysis of the compounds R1COOR2 and MeCOOR3, respectively.It is assumed that the inductive and electronic effects have no influ- ence on the system under consideration. Tables containing refined Es values are presented in the study cited. Examples of successful use of this steric parameter can be found in the literature.15, 184 In the simplest case, steric effects can be characterised by van der Waals radii. Empirical steric parameters based on these radii have been developed in greater detail by Charton.182, 227, 228 The one-parameter method, which is the simplest among those proposed by this researcher, uses steric parameter n, which depends on the van der Waals radii and expresses the steric effect of the first atom in the substituent.In the case of atomic groups having different substituents at the atoms through which the hindered internal rotation axis passes, it is proposed to describe steric effect either by sets of effective steric parameters for certain reactions or by branching equations, which are empirical in their character. Note that the empirical approach developed by Charton involves many simplifications and requires introduction of addi- tional empirical constants and coefficients when considering new sets of compounds. The Lennard ¡¾ Jones potential takes into account the steric constituent of the total energy of an interaction E1j, which is represented as a combination of the attraction and repulsion forces 225 E (35) 1j a A d 12 ¢§ dB6 , where d is the distance between a pair of non-bonded atoms.When d is small, the term A/d 12 accounts for the predominance of the repulsion forces and, as a consequence, positive E1j values. When d is large, the term B/d6 is greater. The minimum E1j value corresponds to the optimum distance between a pair of non- bonded atoms. The Lennard ¡¾ Jones potential can be used as a self-contained descriptor of intermolecular interactions and also as a measure of the contribution of steric interactions to the total interaction energy. In particular, it has been shown that for a series of substituted esters of acetic acid, a steric descriptor based on the Lennard ¡¾ Jones potentials is well correlated with the Taft param- eter Es .2. Description of electrostatic interactions Electrostatic interactions are described using the Coulomb's law for point charges located on atoms and using standard molecular- mechanics approaches.175 When comparing the capacity of molecules for electrostatic interactions V(Ri), molecular electrostatic potential (MEP) is used 513 (36) almost always; this parameter can be calculated at each point R of the space surrounding the molecule 230 O VORiU a ¢§ jRi ¢§ RIj , ZI dORU i ¢§ Rj dr a jR X where d(R) is the electron density, r is the distance, ZI are the charges of the nuclei and RI are the coordinates of the nuclei in the molecule. The results of the calculations of MEP can be repre- sented graphically as maps.Comparison of the MEP maps for a number of compounds having the same type of activity (examples of these studies can be found in Refs 231 ¡¾ 235) may prove useful for judging the character of specific electrostatic interactions of these compounds with a biological target exhibiting the given type of activity. However, it should be borne in mind that visual comparison is subjective. In recent years, formal approaches based on comparison of quantum-chemical parameters related to electrostatic interactions have been vigorously developing. First of all, molecular quantum similarity measures (MQSM) should be mentioned.236 ¡¾ 239 According to this approach, quantum-chemical similarity of two molecules is identified by considering the electron density func- tional ZAB(O) (37) O ZAB(O)= rA(r1)O(r1,r2)rB(r2) dr1dr2, where rA(r1) and rB(r2) are the electron densities of molecules A and B, and O is a positively defined operator.Various operators can be chosen. It is convenient to present the operator in the form (38) O(r1,r2)=d(r17r2), which transforms the integral in Eqn (37) into an integral of the product of density distributions in the given pair of molecules. (39) ZAB= rA(r)rB(r) dr. O If molecules A and B being compared are identical, we obtain a self-similarity measure (40) ZAA= r2A(r) dr.O The self-similarity measure is used to construct similarity indices, which can be regarded as normalised values of MQSM. The Carbo index was the first similarity index to be proposed 240 C (41) AB a OZAAZBBU1=2 . ZAB This value ranges from zero (for comparison of dissimilar molecules) to unity (when the molecules are identical). Examples of using the Carbo index for elucidation of similarity of electronic structures have been reported.241 ¡¾ 244 3. Quantitative description of the ability of compounds to form hydrogen bonds Hydrogen bonding plays a crucial role in the formation of many molecular and ionic complexes in important chemical and bio- logical processes including enzymic catalysis. The most glaring example demonstrating the significant role of hydrogen bonding in biological systems can be found in a monograph 245 in the description of DNA replication.This process is accompanied by a new binding of purine and pyrimidine bases via the formation of two (binding of adenine with thymine) or three (binding of guanidine with cytosine) hydrogen bonds. Even a rather rough calculation of the free energy difference for this complexation, carried out by Pauling (it was assumed that one hydrogen bond provides an energy benefit of 20 kJ mol71), demonstrated that this system of hydrogen bonds ensures a ratio of the numbers of `correct' and `false' recognitions equal to ten million to one!514 Despite this important role of hydrogen binding in the formation of the properties of substances, until recently this intermolecular interaction has been taken into account mainly at the level of an indicator variable.12, 13, 246 ± 248 Therefore, in this review it is pertinent to discuss the quantitative approaches to the description of hydrogen bonding in detail.According to Eqn (14), full thermodynamic description of a hydrogen bond requires a pair of parameters, namely, enthalpy and entropy or enthalpy and free energy. Four combinations of the changes in the enthalpy (DH) and the entropy (DS) are possible.249 ± 252 Note that in the majority of studies devoted to hydrogen bonding in complexes, the change in the entropy is either totally neglected or considered to be proportional to the enthalpy change. The roughness of this approach becomes clear from the examples given in a review.251 The development of the empirical scales of hydrogen bonding based on the experimental values of thermodynamic parameters started in the 1960s.Let us consider two approaches, the additive- multiplicative and multiplicative ones. In the additive-multiplica- tive method, the enthalpy change accompanying the formation of a hydrogen bond is described by six parameters 253 ± 255 (42) 7DHhb=eAeB+cAcB+tAtB, where e is a parameter reflecting the electrostatic contribution, c is a parameter related to the covalent interaction, t is a parameter describing the charge transfer and indices A and B correspond to Lewis acids (hydrogen bond donors) and bases (hydrogen bond acceptors), respectively.Most of studies dealing with the struc- ture ± biological activity correlations present no exact data on the molecular structure of one partner of the interaction (e.g., receptor, biological target). This fact hampers the use of Eqn (42). The method describing the thermodynamic parameters of hydrogen bonds based on the multiplicative approach is more widely distributed. The enthalpy of the hydrogen bond is described by the equation 256 (43) DHij=DH11Pi Ej, where DH11 is the enthalpy change upon the formation of a hydrogen bond between phenol and diethyl ether (standard hydrogen bond partners) in tetrachloromethane,} Pi is a dimen- sionless factor for an ith hydrogen bond donor and Pj is a factor for a jth hydrogen bond acceptor.The change in DHij can also be described using the donor (Ed) and acceptor (Ea) factors for the enthalpy of a hydrogen bond, which have opposite signs 92, 257 (44) DHij=DH11EdEa. The Ed and Ea values for various neutral and ionised compounds have been estimated.258 ± 270 Similar equations have been proposed to describe the change in the free energy (DG) upon the formation of a hydrogen bond 271, 272 (45) DG=aCdCa+a0, where a and a0 are coefficients and Cd and Ca are the donor and acceptor factors for the free energy of the hydrogen bond; and to describe the complex formation constant for hydrogen bond- ing 273 ± 276 (46) logK=bab+b 0, where b and b0 are coefficients and a and b are the donor and the acceptor factors for the hydrogen bond formation constants.Taft et al.277 have estimated a large number of b values for phosphoryl compounds based on the data of several publications.261 ± 263 With } Most of the experiments on the thermodynamics of hydrogen bonding have been carried out in tetrachloromethane. This solvent is non-polar and therefore convenient for determining the values of the thermodynamic factors described below. allowance for the relationship between the change in the free energy and the constants [see Eqn (15)], it is obvious that the difference between Eqns (45) and (46) reduces to the factor 2.302RT. In the case where the same array of experimental data is considered and the same method is used to estimate the hydrogen bond parameters, there should be a strict correlation between Cd and a and between Ca and b.No general correlation has been found between Ed and Cd or between Ea and Ca; however, these correlations do exist for separate groups of compounds.272 Development and successful use of the empirical approaches described above are possible only when the set of experimental data on the thermodynamics of hydrogen bonds is being extended. A detailed description of an extensive set of data of this type has been reported.210 The Ca and Cd values were calculated by selecting 414 hydrogen bond donors and 1298 hydrogen bond acceptors for which experimental thermodynamic parameters for interaction with several partners were available.251 The matrix generated consisted of 414 columns and 1298 rows.This permitted rigorous calculations of the Ca and Cd values. The factors for the free energy of hydrogen bonding for some compounds are presented below Compound Cd 74.78 74.75 74.28 74.09 73.82 73.49 73.31 72.98 72.50 72.06 71.82 71.62 71.42 71.04 70.81 70.50 70.24 70.12 CF3COOH CCl3COOH CHCl2COOH 3,5-(NO2)2C6H3OH C6F5COOH 3-NO2C6H4OH (CF3)2CHOH CF3C(O)NHC(O)CF3 C6H5OH CH(CN)2Br 4-NO2C6H4NH2 CH3C(O)SH C2H5OH C6H5NH2 CHCl3 2-CH3C6H5NH2 C2H5SH C6H5:CH It should be emphasised that the limits of variation of the factors for the free energy of hydrogen bonding for functional groups overlap (Table 2). This indicates that substituents at a donor or acceptor atom play an exceptionally important role in the formation of the capacity of these atoms for hydrogen bonding.Comparison of the experimental values for enthalpy and free energy with the values calculated using Eqns (44) and (45) proves Table 2. Minimum and maximum free energy factors for hydrogen bonds for various functional groups. Cmin d Cmax d Hydrogen bond donors 70.12 70.24 70.50 70.76 71.06 72.50 72.06 71.92 73.32 74.24 73.50 74.78 R1R2R3C7H RS7H R1R2N7H PhO7H AlkO7H RC(O)O7H O A Raevsky Compound Ca0.55 0.60 1.07 1.76 1.92 2.01 2.45 2.57 2.96 3.13 3.65 4.00 4.25 CCl3CN C4H9Cl C6H5OCH3 CH3OH CH3C(O)CH3 (CH3)2SO2 C5H5N (C2H5)3N (CH3)2NC(O)N(CH3)2 (CH3)2SO (C6H5)3PO [(CH3)2N]3PO (C4H9)2NO Cmax a Cmin a Hydrogen bond acceptors 2.01 1.51 3.65 2.77 4.02 3.25 0.78 0.57 0.58 0.55 2.40 1.16 R7O7R R7S7R R1R2R3N RC:N R1R2R3P=O R1R2S=OMolecular structure descriptors in the computer-aided design of biologically active compounds the adequacy of the multiplicative approach to the quantitative (47) (48) description of the thermodynamics of hydrogen bonding 251 DH(calc.)=70.49 (0.29)+0.99 (0.08)DH(exp.), N=2787, R=0.970, S=2.40, F=44350, DG(calc.)=70.07 (0.12)+1.04 (0.05)DG(exp.), N=3301, R=0.991, S=1.12, F=175000.In the case where charge transfer makes a significant contri- bution to the hydrogen bond in a particular complex, the values calculated in this way may deviate from experimental results.278 Using the knownCd values for 414 hydrogen bond donors and Ca values for 1298 hydrogen bond acceptors, one can calculate DGhb for 537 372 hydrogen-bonded complexes.When a similar approach was used to calculate the constants of formation of hydrogen bonds, the a values for 150 hydrogen bond donors and the b values for 500 hydrogen bond acceptors were determined; this permitted the calculation of Khb for 75 000 complexes.279 a a d d A whole series of descriptors have been proposed for the multiplicative description of the thermodynamic parameters of a hydrogen bond, including the donor and acceptor factors for the hydrogen bond free energy and enthalpy for the strongest donor and acceptor sites in a molecule (Cmax, Cmax, E max, Emax), the Secondly, the programme is based on correlations of the sum of the free energy and enthalpy factors over all the donor and calculated values of descriptors with experimental data.acceptors sites in a molecule (PCa, PCd, PEa ,PEd) and the Thirdly, the programme prefers descriptors of substituents to four last-mentioned descriptors divided by the molecular weight molecular descriptors, because the former reflect more accurately (PCa/MW, PCd/MW, PEa/MW, PEd/MW).280, 281 These the variation of properties upon structure modification and descriptors have been used to model the structure ¡À property because their values can be calculated more precisely. relationships in complex processes such as complex forma- 2.The calculations and the use of theoretical descriptors tion,91, 282 ¡À 286 solubility in water,210, lipophilicity,210 inhibition of cholinesterase,287, 288 inhibition of dopamine back capture,289 determination of the boiling points of hydrocarbons and acids,290 insecticide activities of nitrogen and phosphorus organic com- pounds,291 ionophoric activity of macrocycles,292 fungicidal activity of phenols and b-adrenergic activity of alicyclic acids and phenols,293 toxicity against Golden orfe fish 294 and perme- ability of biological membranes.210, 295 Descriptors based on a and b factors have been described in a review.279 Similar indices for hydrogen bonding have been reported in one more study.296 VIII. Indicator descriptors Indicator descriptors (sometimes called de novo constants) have found fairly wide use in modelling the structure ¡À property (activity) relationships.It is difficult to classify these descriptors into any particular class because they can encode in an implicit form various structural features of compounds (properties of hydrogen bond donors and acceptors, the presence or absence of hydrogen bonds within the molecules, ortho-effects, cis ¡À trans isomerism, stereoisomerism, various fragments, etc.). The expediency of using indicator variables in some cases has been demonstrated in a review.297 Thus for two sets of papain ligands, similar two-parameter relations were obtained [see the Appendix, Eqns (A.7) and (A.8)], having fairly close coefficients at MR and s but substantially different constant terms.By combin- ing these equations and by using the indicator variable I (I=1 for mesylamides and I=0 for benzamides), the researchers derived a common relation applicable to both series [see the Appendix, Eqn (A.24)]. Other examples of using indicator descriptors can also be found in the literature.298, 299 Indicator descriptors are especially useful at initial stages of investigations. These descriptors can be used for tentative combi- nations of compounds to give various subsets of compounds with a particular property, which is needed to form a training set meant for the modelling of the structure ¡À property relationship based on a more extensive structure description.In analyses of the structure ¡À biological activity relationships, various physico-chemical molecular properties or experimental 515 parameters obtained by various physical methods are often used as structure descriptors.15 The use of physicochemical properties as descriptors has been discussed in the preceding Sections. A more detailed discussion of these descriptors can be found in the literature.300 IX. Modern computer software for the calculation of descriptors At present the market of computer software meant for chemistry and, in particular, for molecular design is filled by diversified programme products. Therefore, to be competitive, every new programme must have obvious novelty elements or obvious advantages over other programmes.1. MOLSURF is a generator of chemical descriptors for QSAR.193 This programme has three distinctive features. Firstly, it is assumed that a researcher prefers describing molecules using chemical terms such as hydrogen bonds, nucleo- philes, electrophiles, bases, acids, hydrophobicity, polarisability, p-interaction, etc. rather than using quantum-chemical terms (HOMO, LUMO, electrostatic potential, force fields, etc.). There- fore, the MOLSURF programme calculates the descriptors related the above chemical notions. within the framework of the concept of linear correlations of free energy have been described in a number of publications.301 ¡À 306 In this fairly simple approach, a property is described by regression equations containing descriptors related to steric interactions, polarisability and hydrogen bonding as independent variables.The steric interactions are estimated quantitatively using the van der Waals volumes Vmc. A descriptor called polarisability index (pI) is defined as polarisability divided by volume. To describe hydrogen bonds, the researchers distinguish two constituents, the acidic and basic ones. Each constituent is divided in turn into covalent and electrostatic components. The covalent contribution is assumed to be determined by the energy of the highest occupied molecular orbital (HOMO). The electrostatic contribution to the basicity is characterised by the formally most negative atomic charge (q7) in the molecule (it is assumed that the most negative atom should interact most efficiently with the proton in a neighbouring molecule). The covalent component in the acidic constituent (ea) is characterised by the energy of the lowest unoccupied molecular orbital (the difference between the LUMO energy of the compound and the HOMO energy of water divided by hundred).A smaller ea value points to a higher tendency to form hydrogen bonds (the eb component) with water. The contribution of the electrostatic component to the acidity is determined by the most positive charge of a hydrogen atom in the molecule (q+). According to this approach a property (P) can be described by the following six-parameter equation log P �� log P0 �¢ AVmc 100 �¢ BpI �¢ Ceb �¢ Dq¡¦ �¢ Eea �¢ Fq�¢. (49) where A, B, C, D, E and F are coefficients.This method has been used to describe the properties and biological activity of chemical compounds and gave formally rather good results. Nevertheless, this approach seems to be based on a simplified physical model. In addition, when initialis- ing sets are relatively small, multiparameter regression equations of this type are not stable and may lead to faulty results when used to predict the properties of compounds. 3. The CoMFA programme (Comparative Molecular Field Analysis) is currently the most widespread computer programme516 used to calculate the energies of intermolecular interactions in terms of the corresponding descriptors establishing the structure ¡¾ property relationships.307 The parameters calculated initially include the energies of steric and electrostatic interactions between the compound considered and a probe atom being placed into various points of a regular three-dimensional grid.The size of this grid should be sufficiently large to accommodate the largest molecule of the initialising set. The size of the cell in this grid is usually taken to be 2.0 A and, in some cases, 1 A (the smaller the grid dimensions, the larger the number of descriptors and the time it takes to calculate them). An sp3 hybridised carbon atom with the charge +1.0 was used initially as the probe atom; later, various atoms and groups have been used (see, e.g. Ref. 308). It is clear that these calculations give an enormous number of descriptors, exceeding the number of compounds in the initialising set by a large factor.Most of the descriptors thus obtained are correlated with one another. Under these conditions, a stable model can be designed only by using a special procedure. The PLS (Partial Least Squares) 309 method permits selection of orthogonal components. During the design of stable models, the number of these components can be optimised by calculating the correlation coefficients for predictive models, formed by the cross-validation procedure. During the last decade, the CoMFA programntly improved. For example, several studies have been devoted to the methodological aspects of the programme develop- ment 310, 311 (description of non-linear CoMFA, the use of various semiempirical methods for optimisation of the calculations of the geometry and the properties of compounds).The programme has been used to study many types of biological activity of various compounds, for example, the relationship between the structure of 2-hetarylquinoline-4-amines and their biological activity against HIV,312 the structures and steric and electronic effects of phos- phorus ligands 313 and the structure ¡¾ inhibitory activity correla- tions for various compounds,314 steroid inhibitors,315 melatonin receptor ligands 316 and calcium blocking agents.317 However, the CoMFA programme is not free from draw- backs. First of all, it should be noted that until recently it had been unable to calculate the energies of hydrogen bonds, which play a crucial role in many biological processes.In addition, based on the parameters that act as independent variables in the final regression equation, it is hardly possible to construct any useful physical model for a structure ¡¾ property relationship because these varia- bles are linear combinations of many thousands of descriptors. Yet another drawback of this programme is that it cannot be run on a personal computer. 4. The GRID programme,225 having some common features with CoMFA, also uses atoms and functional groups of various natures as the probe atoms to estimate the intermolecular interactions with the macromolecules under study, which are put in a three-dimensional grid. It is significant that this programme, unlike CoMFA, takes into account not only steric and electro- static interactions but also hydrogen bonds (Ehb).Initially this has been done using a direction-dependent 6 ¡¾ 4 function with a broad energy minimum 225, 318 E (50) hb a d 6 ¢§ dD4 cosmQ. C whereQis the DHP angle (D is the donor atom of the protein,His a hydrogen atom and P is the probe electron-withdrawing atom); m is usually taken to be 4. If the interaction involves two identical atoms, the interatomic separation dmin is determined by tabulated values for C and D. If different atoms interact, the geometrical mean of their individual D values and the arithmetic mean of their separations (dmin) are used. Later, the 6 ¡¾ 4 function has been replaced by a 8 ¡¾ 6 function.319 ¡¾ 320 O A Raevsky E (51) r a Cr 8 ¢§ rD6 , where r is the distance between the probe atom and some atom in the compound considered.The hydrogen bond energy is repre- sented by the following expression: (52) Ehb=Er EtEp , where Et and Ep are angle functions taking account of the mutual arrangement in space of the hydrogen atom and the lone electron pair of the hydrogen bond acceptor. The GRID programme has been used successfully to design models of various biological receptors.321 ¡¾ 327 5. As an alternative to the CoMFA and GRID programmes, approaches reflecting a three-dimensional structure and the capacity for intermolecular interactions as a spectrum of intera- tomic distances or interatomic interactions can be consid- ered.251, 270, 328 ¡¾ 340 In these cases, a molecular structure is represented as a radial scattering curve, which can be constructed for a molecule of any complexity provided that its Cartesian coordinates are known.A similar approach is employed in gas phase electron diffraction studies of molecular structures. The radial distribution function [M(s)] in the spherical approximation for atoms i and j and in the absence of intramolecular vibrations is a spectrum of interatomic distances, the intensity of the spectrum being determined by the natures of the atoms spaced at a given distance and by the dynamics of the variation of this distance during intramolecular vibrations , (53) Cij sin srij srij MOsU a j>i ia1 XN XN where Cij is a coefficient characterising the scattering capacity of this pair of atoms, s is an angle parameter: s=(4p/l)sin(Q/2) (l is the electron wavelength and Q is the scattering angle) and rij is the interatomic distance in the molecule.It can be said that the radial distribution function obtained in an electron diffraction experiment is the pattern `seen' by an electron interacting with the molecule in a scattering event. This analogy has been extended to the case where a receptor interacts with molecules of a substance administered into an organism.328 It is known that a receptor interacts with parts of the molecule that are complementary to the active sites of the receptor rather than with the whole molecule. The governing factors are the distances between the active sites and the susceptibility to intermolecular interactions of the atoms located at these distances from one another in the reagent molecules.A sort of scattering of the reagent molecules on the active sites of the receptor occurs, i.e. the receptor `sees' the molecule as a radial distribution function. In this approach, the radial distribution function can be modified in such a way as to take into account quantitatively the ability of atoms of the reagent to undergo various intermolecular interac- tions. This approach was embodied in some studies 251, 340 and in the MOLTRA (Molecular Transform Analysis) programme package, which can be used to calculate the spectra for the steric interactions of atoms; interactions of positively charged atoms with one another, negatively charged atoms with one another and positively charged atoms with negatively charged atoms; inter- actions of hydrogen bond donors with one another, hydrogen bond acceptors with one another and hydrogen bond donors with hydrogen bond acceptors.Each of these spectra is a superposition of bands corresponding to pairs of atoms in the molecule. The positions of the band maxima are matched by the equilibrium distances between these atoms and the areas below the band contours are proportional to the products of the given local properties for a given pair of atoms. Figure 4 shows the spectra for the interactions of hydrogen bond acceptors in some macrocyclic polyethers. Each point in the spectra presented can be considered as a descriptor of intermo- lecular interactions describing the interaction of hydrogen bondMolecular structure descriptors in the computer-aided design of biologically active compounds Figure 4.General view of the spectrum of interactions of hydrogen bond acceptors in some macrocyclic polyethers. acceptors located at a given distance. For distances ranging from 1.1 to 20.0 A and with a step of 0.1 A, each spectral pattern provides 190 descriptors. Seven spectral patterns permit the description of the possible steric and Coulomb interactions and hydrogen bonds in terms of 1340 descriptors. X. The necessary properties of descriptors A great number of descriptors have been reported by now.It is clear that the extension of the views on molecular structure stimulates the design of new structure models and new descrip- tors reflecting these views. The attempts of developing new descriptors which, having been introduced in computer pro- grammes, would preserve and utilise the information on the molecular structure better than the earlier descriptors are also justified. Nevertheless, the introduction of new descriptors is expedient only in two cases�either when none of the existing descriptors or their combination provides the elucidation of stable structure ± property correlations for the training set used or when the new descriptor ensures a substantial improvement of the statistical criteria of the model.9 New descriptors should be readily calculable, convenient to use and versatile; they should ensure comprehensive description of a given structural feature and be interpretable.341 Unfortunately, the above requirements are by far not always taken into account. The efficiency of descriptors of various classes used in establishing the structure ± biological activity correlations becomes clear upon comparison of the statistical criter of the models designed using these descriptors. For example, Raevsky et al.290, 293 compared the models for classification of agonists and antagonists among substituted phenylethylamines based on a number of topological and physico-chemical descriptors using linear discriminant analysis, a discrete regression model 290 and significant cluster analysis.342 In all of these types of classification, the best results were attained by using hydrogen bond descriptors.Other examples can also be found in the literature.14 ± 16, 337, 343, 344 In this review, the descriptors were introduced in the order of increasing amount of information on the molecular structure contained in them. However, this does not mean that the use of a more informative descriptor guarantees better results in elucida- tion of the structure ± property correlations. Indeed, there are many examples of successful employment of topological descrip- tors for the calculation of the melting and boiling points, evaporation enthalpies and other `simple' physical properties within narrow series of related compounds, which differ, for example, in the length of the chain in the alkyl substituent.In these cases, the statistical criteria of regression equations contain- ing more informative descriptors as independent variables prove sometimes to be inferior to those for equations based on simpler descriptors, because the magnitudes of more informative descrip- tors are estimated with lesser accuracy and because they carry `excessive' information for the set considered. The foregoing leads to the following conclusions: it is expedient to use combinations of known descriptors with different information contents; it is useful to develop new descriptors and perfect the procedures to attain more precise calculations of all descriptors; well-founded selection of descriptors for a particular set of compounds is necessary.} Note that comprehensive description of a molecular structure cannot be attained in terms of any single descriptor. One descriptor can predominantly take into account one feature of 517518 the electronic or steric structure or one type of intermolecular interactions. For this reason, description of a molecular structure using several descriptors and construction of structure ¡À property models based on several independent variables appears expedient.In the construction of structure ¡À biological activity models, the modern views on the processes occurring in an organism with participation of chemical compounds play an important role in the selection of appropriate descriptors.For example, it is currently believed that the complex mechanism of action of a substance on an organism consists of three stages D pharmaceu- tic, pharmacokinetic and pharmacodynamic stages.345 The phar- maceutic stage is associated with the administration of a substance into an organism and includes disintegration of the preparation. The pharmacokinetic stage is determined by the transport of the substance from the administration site to the particular biological target and includes the processes of absorption, distribution, metabolism and excretion. At this stage, the reactivity of the initial compounds and their capacity for complex formation are important. The pharmacodynamic stage is characterised by the interaction of the initial substance or the products with biological targets responsible for the particular type of biological activity. At this stage, the three-dimensional structure of the interacting partners is the most important factor.The exceptional popularity of the Hansch ¡À Fujita three- parameter equation is due most of all to the fact that it uses descriptors of the electronic structure (Hammett constants), steric interactions (Taft steric constants), and transport properties (octanol ¡À water partition coefficients). At present, similar mod- els and the corresponding equations are also used successfully in the molecular design of biologically active substances.346 ¡À 349 For example, for a series of bicyclic compounds, an equation describ- ing their ability to bind the muscarinic receptor based on three descriptors has been derived 210 [see the Appendix, Eqn (A.23)]. One descriptor characterises the transporting properties of the compounds, the second descriptor is related to the solubility of the compounds in water and the ability to form hydrogen bonds with the donor sites of the receptor and the third one points to the important role of the distance between two particular hydrogen bond acceptors in the formation of the substrate ¡À receptor complex.The development of models requires the observance of formal criteria. From the statistical viewpoint, the number of compounds considered should exceed the number of descriptors by a large factor. The information contents of the descriptors must be sufficiently large and, simultaneously, they should not correlate with one another.This can be attained by several procedures including experimental design, 350 the SIMCA/PLS method 351, 352 and the principal component and factor analyses.353 XI. Conclusion At present, molecular structure descriptors constitute an integral part of the majority of studies dealing with structure ¡À property (including biological activity) correlations; they are widely used to optimise characteristics of compounds in the molecular design. The Appendix to this review presents examples of successful application of various types of descriptors for quantitative description of diverse properties and types of biological action.It should be emphasised that the molecular structure descrip- tors used should finally provide the design of structure ¡À property models, which satisfy certain requirements. These models should be consistent with the laws of Nature, be relatively simple and have a clear physical meaning; in addition, they must be stable and hence possess predictive capacity and provide the possibility of identifying the relationships that were not specified in an explicit } An expert can rely not only on the generally recognised views on the mechanism of formation of the property being considered but also on the knowledge not yet formalised and even on intuition. O A Raevsky form when the model was developed. All these points broach one more important section of the computer-aided molecular design, namely, modelling of the structure ¡À property relationships, and require separate discussion.XII. Appendix Examples of successful application of molecular structure descrip- tors in structure ¡À property (biological activity) correlatons are presented below. (A.1) 1. Element-level descriptors Solubility of gases in water, log Lw (see 354). log Lw=42.37(1.11)HDCA(2)+0.65(0.02)[2n(N)+ +m(O)]70.16(0.02)DE+0.12(0.01)PCWTE+ +0.82(0.01)Nring +2.65(0.22), where HDCA(2) is a hydrogen bond descriptor, [2n(N)+m(O)] is a descriptor characterising the number of nitrogen and oxygen atoms in the molecules of compounds studied, DE=ELUMO7EHOMO, PCWTE is an electronic topological index, Nring is the number of aromatic rings; N=406, R=0.971, S=0.52, F=1300.2. Topological descriptors Lipophilicity of organic compounds (based on topological and information topological indices), logP ( see 355) (A.2) log P=0.19P1071.46 IC0+1.09 CIC270.77 CIC37 71.36 6Xb +5.340XV73.411XV+0.554XV7 70.413XVC +1.10Xw70.17W75.60, N=219, R=0.955, S=0.35, F=194. b-Blocking activity of medicines, pLD50 (see 356). (A.3) pLD50=1.791w71.811wV+0.362w+2.374wVp 7 72.764wp+1.56, where w are Kier and Hall connectivity indices; N=17, R=0.905, S=0.208, F=10.1. 3. Descriptors of the electronic structure of molecules Binding of an estrogen receptor, pEC50 (see 357). pEC50=3.38(0.12)D2+3.22(0.20) Gap744.3(1.8) (A.4) where D is the Balaban topological index, Gap is the difference; N=33, R=0.99, S=0.30, ELUMO7EHOMO F=595.0.Toxicity against Daphnia magna, EC50 (see 358). EC50 �� ¡¦3:63Vmc 100 ¡¦ 45:8p�¢3:71q¡¦ ¡¦ 2:0q�¢ �¢ 11:4, (A.5) where Vmc is the volume, p is the polarisability index, q7 is the electrostatic basicity, q+ is the electrostatic acidity; N=38, R=0.977, S=0.37, F=176. Fungicide activity of phosphonates, pEC50 (see 359). pEC50=0.296(0.124)Pp70.047(0.012)MR(R2)+ +0.543(0.331)PF 0+3.556(0.215), (A.6) where p is the hydrophobicity constant, MR(R2) is the molar refraction of the second substituent, F 0 is the Swain¡ÀLupton electronic parameter; N=57, R=0.851, S=0.290, F=46.32. Binding of papain ligands by phenyl N-mesylglycinates (see 297). (A.7) log 1 ��0.529(0.230)MR+0.379(0.200) s+ KmMolecular structure descriptors in the computer-aided design of biologically active compounds +1.877( 0.130), where s is an electronic constant; N=13, R=0.935, S=0.105, F=34.51.Binding of papain ligands by phenyl N-benzoylglycinates (see 297). (A.8) log 1 a0.771(0.670)MR+0.728(0.370) s+ Km +3.623(0.340), N=7, R=0.971, S=0.148, F=32.85. (A.9) 4. Molecular shape descriptors Inhibition of ligand binding, pIC50 (see 360). pIC50=10.25(0.97)V +1.43(0.56), where V is the shape index; N=26, R=0.912, S=0.462, F=111.1. Sweetness of aniline derivatives, logRS (see 361). (A.10) logRS=0.52 L71.37W1+3.71, where L and W1 are STERIMOL parameters; N=20, R=0.90, S=0.32. (A.11) Psychomimetic activity of phenyl-2-aminopropane derivatives, logMU (see 362).logMU=1.99(0.13)71.78(0.17)MTD+ +0.93(0.10) S370.77(0.12) q4 , where MTD is the minimum topological difference, S3 is the electron topological index for the atom with number 3 in the benzene ring, q4 is the charge on atom 4 of the benzene ring; N=49, R2=0.885, S=0.241, F=84.7. Inhibition of hepatic receptors, pEC50 (see 169). (A.12) pEC50=0.447Tm+0.242 Ts70.158 Vu+2.361, where Tm, Ts and Vu are WHIM descriptors; N=71, R=0.914, S=0.64. Inhibition of photosystem II, pIC50 (see 363). (A.13) pIC50=1.081 logP70.57s(Y,Al)+0.29L(Y,Al)7 70.95T(Y,Al)72.21T (Y,Ar)70.32W(X)+ +1.02 Iu+8.16, where Y and X are substituents, Al is alkyl, Ar is aryl, L,Wand T are shape descriptors, Iu is an indicator variable (1 for urea derivatives and 0 for all other compounds); N=69, R=0.921.(A.14) 5. Intermolecular interaction descriptors Lipophilicity of various compounds, logP (see 209). logP=70.01(0.08)+0.249(0.005) a7 70.68(0.02)PCa , where a is the molecular polarisability, PCa is the sum of the factors for the hydrogen bond acceptor capacity; N=234, R=0.961, S=0.34, F=1383. Effect of protein binding, log HSAI (see 364). (A.15) log HSAI=0.59(0.03) logP+0.27(0.08), N=57, R=0.92, S=0.28, F=283. Toxicity against Daphnia, log(1/IC50) (see 365). (A.16) 1 log IC a 0:809 log P a 1:64, 50 519 N=23, R2=0.84, S=0.30, F=118. Bactericidal activity, pMIC (see366). (A.17) pMIC=70.35 pKa+0.68, N=8, R=0.973, S=0.063.Ligand ¡¾ receptor binding, DG (see 367). DG=0.06(0.01)EL70.05(0.02)DEsolv+7.74(0.57), (A.18) where EL is the change of the intramolecular energy of the ligand upon binding, DEsolv is the change of the solvation upon binding; N=10, R=0.92, F=20.61. Inhibitory activity of aryl-substituted piperazine and piperylene, log(1/EC50) (see 368). 1 logEC Ca¢§ (A.19) a 0:62O0:04U logP ¢§ 0:39O0:05U 50 X 72.16(0.18), N=17, R2=0.90. (A.20) Toxicity of carboxylic acids against Xenopus embryos, pLC50 (see 358). pLC50=0.696 logP70.358 pKa70.541, N=44, R=0.947. Membrane permeability (human red cell basal), logBP (see 369). (A.21) N=10, R=0.983, S=0.433. logBP=70.70(0.64)+1.08(0.16)PCd , Membrane permeability (cells of the alga Chara ceratophylla), log Per (see 369). (A.22) N=27, R=0.903, S=0.49.log Per=0.83(0.57)+0.59(0.12)PCd , Binding of the muscarinic receptor by bicyclic ligands, logK (see 210). (A.23) +0.27(0.08) HBA5.3A74.58(0.48), logK=1.28(0.15) logP70.09(0.06)PCa+ where HBA5.3A is a descriptor characterising the interaction between two hydrogen bond acceptors spaced 5.3 A E apart; N=27, R=0.918, S=0.38, F=41.1 6. Indicator descriptors Binding of papain ligands by phenyl N-mesylglycinate and N- benzoylglycinate, log(1/Km) (see 297). (The combination of Eqns (A.7) and (A.8) with the aid of an indicator variable I (I=1 for mesylamides and I=0 for benzamides)]. (A.24) log 1 a 0:57O0:26UMR a 0:56O0:19Us¢§ Km ¢§1:92O0:15UI a 3:74O0:17U, N=20, R=0.990, S=0.148, F=272.04.Inhibition of HIV-RT, pIC50 (see 370). (A.25) pIC50=1.47 I+0.41 m+2.69 logMW74.36, (A.26) where I is the indicator variable, MW is the molecular weight, N=14, R=0.94, S=0.42, F=27.4. Cytotoxic activity, pIC50 (see 371). pIC50=0.73(0.46) I2 0+1.14(0.38) I5 0+ +1.12(0.43) Ixy+3.15,520 where I2 0, I5 0, and Ixy are indicator variables; N=22, R=0.915, S=0.32, F=30.78. References 1. C Hansch Drug Metabolism Rev., 15 1279 (1984-85) 2. C Hansch, T Fujita J. Am. Chem. Soc. 86 1616 (1964) 3. R Lipnick, V A Filov Trends Pharm. Sci. 13 56 (1992) 4. N V Lazarev Neelektrolity (Nonelectrolytes) (Leningrad: VMMA, 1944) 5. O A Raevsky, A M Sapegin Usp. Khim. 57 1565 (1988) [Russ.Chem. Rev. 57 897 (1988)] 6. S Nikolic, N Trinajstic Croat. Chem. Acta 70 777 (1997) 7. O A Raevsky Ros. Khim. Zh. 39 109 (1995) a 8. J Tomasi J. Mol. Struct. (Theochem) 179 273 (1988) 9. M Charton Newsl. Int. QSAR Soc. 1 3 (1991) 10. C Hansch Acc. Chem. Res. 26 147 (1993) 11. C Hansch, D Hoekman, H Gao Chem. Rev. 96 1045 (1996) 12. E J Lien SAR. Side Effects and Drug Design (New York: Marcel Dekker, 1987) 13. E J Lien Chin. Pharm. J. 42 93 (1990) 14. P C Jurs, S L Dixon, L Egolf, in Chemometric Methods in Molecular Design Vol. 2 (Ed. H Waterbeemd) (Dusseldorf: VCH, 1995) p. 15 15. B Testa, L B Kier Med. Res. Rev. 11 35(1991) 16. A J Stuper,WE Brugger, P Jurs Computer Assisted Studies of Chemical Structure and Biological Function (New York: Wiley, 1979) 17.N S Zefirov, S I Kuchanov (Ed.) Primenenie Teorii Grafov v Khimii (Application Graph Theory in Chemistry) (Novosibirsk: Nauka, 1988) 18. M I Stankevich, I V Stankevich, N S Zefirov Usp. Khim. 57 337 (1988) [Russ. Chem. Rev. 57 191 (1988)] 19. Yu G Papulov, V R Rozenfel'd, T G Kemenova Molekulyarnye Grafy (Molecular Graphs) (Tver': Tver' State University, 1990) 20. R B King (Ed.) Chemical Applications of Topology and Graph Theory (Amsterdam: Elsevier, 1983) 21. S S Tratch, M I Stankevitch,M S Zefirov J. Comput. Chem. 11 899 (1990) 22. L B Kier, L H Hall Molecular Connectivity in Structure ± Activity Analysis (New York: Wiley, 1986) 23. M Randic, X Guo, T Oxley, H Krishnaprijan J. Chem. Inf. Comput. Sci. 33 709 (1993) 24.M Randic J. Am. Chem. Soc. 97 6609 (1975) 25. L B Kier, L H Hall J. Pharm. Sci. 64 1971 (1975) 26. L B Kier, L H Hall J. Pharm. Sci. 65 1806 (1976) 27. L B Kier, L H Hall Pharm. Res. 8 801 (1990) 28. L H Hall, B Mohney, L B Kier Quantum Struct. ± Act. Relat. 10 43 (1991) 29. L H Hall, L B Kier J. Chem. Inf. Comput. Sci. 35 1039 (1995) 30. L H Hall, L B Kier, B B Brown J. Chem. Inf. Comput. Sci. 35 1074 (1995) 31. A K Saxena Quantum Struct. ± Act. Relat. 14 31 (1995) 32. A K Saxena Quantum Struct. ± Act. Relat. 14 142 (1995) 33. D L Dowdy, T E McKone Environ. Toxicol. Chem. 16 2448 (1997) 34. L H Hall, C T Story J. Chem. Inf. Comput. Sci. 36 1004 (1996) 35. M T D Cronin Toxicol. Vitro 10 103 (1996) 36. P-Z Lang, X-F Ma, G-H Lu, Y Wang, Y Bian Chemosphere 32 1547 (1996) 37.M Soskis, D Plavsic, N Trinajstic J. Chem. Inf. Comput. Sci. 36 829 (1996) 38. R Garcia-Domenech, F J Garcia-March, R M Soler, J Galvez, G M Anton-Fos, J V de Julian-Ortiz Quantum Struct. ± Act. Relat. 15 201 (1996) 39. R R A Abon-Shaaban, H A Al-Khamees, H S Abou-Auda, A P Simonelli Pharm. Res. 13 129 (1996) 40. H Kubinyi Quantum Struct. ± Act. Relat. 14 149 (1995) 41. S Nikolic, N Trinajstic, Z Mihalic Croat. Chem. Acta 68 105 (1995) 42. I Lukovits J. Chem. Inf. Comput. Sci. 34 1079 (1994) 43. A A Dobrynin, A A Kochetova J. Chem. Inf. Comput. Sci. 34 1082 (1994) 44. M Randic Stud. Phys. Theor. Chim. 54 101 (1988) O A Raevsky 45. M Randic Int. J. Quantum Chem. Quantum Biol.Symp. 15 201 (1988) 46. M Randic, A F Kleiner, L M DeAlba J. Chem. Inf. Comput. Sci. 34 277 (1994) 47. M Randic,M Razinger J. Chem. Inf. Comput. Sci. 35 140 (1995) 48. O G Mekenyan, D Peitchev, D Bonchev, N Trinajstic, I Bangov Drug. Design 36 176 () 49. M V Diudea, D Horvath, A Graovac J. Chem. Inf. Comput. Sci. 35 129 (1995) 50. L B Kier Quantum Struct. ± Act. Relat. 4 109 (1985) 51. L B Kier Quantum Struct. ± Act. Relat. 5 1 (1986) 52. L B Kier Quantum Struct. ± Act. Relat. 5 7 (1986) 53. L B Kier Quantum Struct. ± Act. Relat. 6 8(1987) 54. L B Kier Quantum Struct. ± Act. Relat. 6 117 (1987) 55. Y-Y Yao, Lu Xu, Y-Q Yang, X-S Yuan J. Chem. Inf. Comput. Sci. 33 590 (1993) 56. G Rucker, C Rucker J. Chem. Inf. Comput. Sci. 33 683 (1993) 57.H P Schultz, T P Schultz J. Chem. Inf. Comput. Sci. 33 240 (1993) 58. H P Schultz, E B Schultz, T P Schultz J. Chem. Inf. Comput. Sci. 33 863 (1993) 59. H P Schultz, E B Schultz, T P Schultz J. Chem. Inf. Comput. Sci. 34 1154 (1994) 60. M V Diudea, M Topan, A Graovac J. Chem. Inf. Comput. Sci. 34 1072 (1994) 61. T Pisanski, A Zithik, A Graovac, A Baumgartner J. Chem. Inf. Comput. Sci. 34 1090 (1994) 62. J Galvez, R Garcia,M T Salabert, R M Soler J. Chem. Inf. Comput. Sci. 34 520 (1994) 63. E Estrada J. Chem. Inf. Comput. Sci. 35 701 (1995) 64. E Estrada J. Chem. Inf. Comput. Sci. 35 708 (1995) 65. B Llorente, N Rivero, R Carrasco, R M Martinez Quantum Struct. ± Act. Relat. 13 419 (1994) 66. I Lukovits Quantum Struct. ± Act.Relat. 9 227 (1990) 67. G G Gash, J J Breen J. Chem. Inf. Comput. Sci. 33 275 (1993) 68. A R Katritzky, E V Gordeeva J. Chem. Inf. Comput. Sci. 33 835 (1993) 69. J Galvez, R Garcia-Domenech, J V de Julian-Ortiz, R M Soler J. Chem. Inf. Comput. Sci. 35 272 (1995) 70. B Lucic, S Nikolic, N Trinajstic, D Juretic, A Juric Croat. Chem. Acta 68 435 (1995) 71. J Niederfellner, D Lenoir, G Matuschek, F Rehfeldt, H Utschick, R Bruggemann Quantum Struct. ± Act. Relat. 16 38 (1997) 72. J V de Julian-Ortiz, R Garcia-Domenech, J Galvez Alvarez, R Soler Roca, F J Garcia-March, G M Anton-Fos J. Chromatogr., A 719 37 (1996) 73. R B King (Ed.) Chemical Applications of Topology and Graph Theory (Amsterdam: Elsevier, 1983) 74. G Bazylak J. Planar.Chromatogr. 7 202 (1994) 75. S C Basak, S Bertelsen,G D Grunwald J. Chem. Inf. Comput. Sci. 34 27 (1994) 76. W J Boecklen, G J Niemi SAR QSAR Environ. Res. 2 79 (1994) 77. A T Balaban, S C Basak, T Colburn, G D Grunwald J. Chem. Inf. Comput. Sci. 34 1118 (1994) 78. S C Basak, G D Grunwald J. Chem. Inf. Comput. Sci. 35 366 (1995) 79. T Wieland Arzheim.-Forsch. Drug. Res. 46 273 (1996) 80. A T Balaban, T S Balaban J. Chim. Phys. Phys.-Chim. Biol. 89 1735 (1992) 81. W J Wisswesser The Wisswesser Line-Formula Chemical Notation (New York: McGraw-Hill, 1968) 82. V V Avidon Khim.-Farm. Zh. 8 22 (1974) b 83. V V Avidon, V S Aralovich NTI, Ser. 2, Inform. Prots. Sist. (5) 26 (1975) 84. D Weininger J. Chem. Inf. Comput. Sci. 28 31 (1988) 85.D Weininger J. Chem. Inf. Comput. Sci. 29 97 (1989) 86. D Weininger J. Chem. Inf. Comput. Sci. 30 237 (1990) 87. A Drefahl,M Reinhard J. Chem. Inf. Comput. Sci. 33 886 (1993) 88. K K Agarwal, H L Gelernter J. Chem. Inf. Comput. Sci. 34 463 (1994) 89. V V Avidon, V E Golender, A B Rozenblit, in Metody Predstav- leniya i Obrabotki Strukturnoi Informatsii dlya Analiza Svyazi Struktura ± Aktivnost' (The Methods of Presentation and Processing of Structural Information for Analysis of Structure ± ActivityMolecular structure descriptors in the computer-aided design of biologically active compounds Relationship) (Riga: Institute of Organic Compounds, Academy of Science, Latv. SSR, 1981) p. 8 90. A B Rozenblit, V E Golender Logiko-Kombinatornye Metody v Konstruirovanii Lekarstv (Logical Combinatorial Methods in Drug Design) (Riga: Zinatne, 1983) 91.O A Raevsky, A M Sapegin, V V Chistyakov, V P Solov'ev, N S Zefirov Koord. Khim. 16 1175 (1990) c 92. O A Raevsky, V V Avidon, V P Novikov Khim.-Farm. Zh. 16 968 (1982) b 93. V P Avidon, O A Raevsky, V S Arolovich, R I Zhdanov, in Bioactive Spin Labels (Ed. R I Zhdanov) (Berlin: Springer, 1992) p. 549 94. G M Barenboim, A G Malenkov Biologicheski Aktivnye Veshchestva (Bioactive Compounds) (Moscow: Nauka, 1986) p. 90 95. I Paeva, E Golovinsky Quantum Struct. ± Act. Relat. 9 16 (1990) 96. A Ormerod, P Willett, D Bawden Quantum Struct. ± Act. Relat. 8 115 (1989) 97. A Ormerod, P Willett, D Bawden Quantum Struct. ± Act. Relat. 9 302 (1990) 98.W J Streich, R Franke Quantum Struct. ± Act. Relat. 4 13 (1985) 99. R Franke, W J Streich Quantum Struct. ± Act. Relat. 4 51 (1985) 100. R Franke, W J Streich Quantum Struct. ± Act. Relat. 4 63 (1985) 101. V K Gombar, E P Jaeger, P C Jurs Quantum Struct. ± Act. Relat. 7 225 (1988) 102. R A Lewis, A C Good, S D Pickett, in Computer-Assisted Lead Finding and Optimization. Current Tools for Medicinal Chemistry. (Eds H van de Waterbeemd, B Testa, G Folkers) (Weinheim: VCH, 1997) p. 137 103. J E Dubois, M Loukianoff, C Mercier J. Chim. Phys. 89 1493 (1992) 104. C Mercier, G Trouiller, J-E Dubois Quantum Struct. ± Act. Relat. 7 88 (1990) 105. E J Martin, J M Blaney,M A Siani, D C Spellmeyer, A K Wong, W H Moos J. Med. Chem. 38 1431 (1995) 106.H-J Bohm, in Computer-Assisted Lead Finding and Optimization. Current Tools for Medicinal Chemistry. (Eds H van de Waterbeemd, B Testa, G Folkers) (Weinheim: VCH, 1997) p. 125 107. M Karelson, V S Lobanov, A R Katritzsky Chem. Rev. 96 1027 (1996) 108. O Kikuchi Quantum Struct. ± Act. Relat. 6 179 (1987) 109. H Bauknecht, A Zell, H Bayer, P Levi, M Wagener, J Sadowski, J Gasteiger J. Chem. Inf. Comput. Sci. 36 1205 (1996) 110. S Karabunarliev, O G Mekenyan,W Karcher, C L Russom, S P Bradbury Quantum Struct. ± Act. Relat. 15 302 (1996) 111. M A Ordocica, M L Velazquez, J G Ordocica, J L Escobar, P A Lehmann Quantum Struct. ± Act. Relat. 12 246 (1993) 112. F T Hatch,M E Colvin, E T Seidl Environ. Mol. Mutagen. 27 314 (1996) 113.A Inoue, T Kawai, M Iimura, H Sugimoto, H Kawakami J. Med. Chem. 39 4460 (1996) 114. M T D Cronin, T W Schultz Chemosphere 32 1453 (1996) 115. M Cocchi, M C Menziani, P G De Benrdrtti, G Cruciani Chemometr. Intell. Lab. Syst. 14 209 (1992) 116. K Tuppurainen J. Mol. Struct. (Theochem) 306 49 (1994) 117. B Balasubramanian SAR QSAR Environ. Res. 2 59 (1994) 118. R G Parr, R Pearson J. Am. Chem. Soc. 105 7512 (1983) 119. Z Medven, H Gusten, A Subljic J. Chemometr. 10 135 (1996) 120. C Graham, R Gealy, O T Macina, M H Karol, H S Rosenkranz Quantum Struct. ± Act. Relat. 15 224 (1996) 121. W H J Vaes, C Hamwijk, E U Ramos, H J Mverhaar, J L M Hermens Anal. Chem. 68 4458 (1996) 122. D F W Lewis, B G Lake Toxicology 125 31 (1998) 123.R C Young, C Graham,M L Roantree, in QSAR in Drug Design and Toxicology (Eds D Hadzy, B Jerman-Blazic) (Amsterdam: Elsevier, 1987) p. 91 124. T I Oprea, L Rurunczi, E E Moret, in Trends in QSAR and Molecular Modelling 92 (Ed. C G Wermuth) (Leiden: ESCOM, 1993) p. 398 125. S Gibson, R McGuire, D C Rees J. Med. Chem. 39 4065 (1996) 126. D Chapman J. Comput.-Aid. Mol. Design 10 501 (1996) 127. J H van Drie J. Chem. Inf. Comput. Sci. 37 38 (1997) 128. A Verloop,W Hoogenstraaten, J Tipker, in Drug Design (Ed. J Ariens) (New York: Academic Press, 1976) p. 165 521 129. J Jurgens, H Schwedletzky, P Heising, J K Seydel, B Wiedemann, U Holzgrabe Arch. Pharm. Med. Chem. 329 179 (1996) 130. A Tanaka, H Fujiwara J. Med. Chem. 39 5017 (1996) 131.W Draber Z. Naturforsch., C Biosci. 51 1 (1996) 132. M M Davila-Jimenez, M P Elizalde-Gonzales, S Garcia-Diaz J. Liq. Chrom. Rel. Technol. 21 1629 (1998) 133. J E Amoore Ann. N. Y. Acad. Sci. 116 457 (1964) 134. R H Rohrbaugh, P C Jurs Anal. Chim. Acta 199 99 (1987) 135. G M Crippen J. Med. Chem. 22 988 (1979) 136. G M Crippen J. Med. Chem. 23 599 (1980) 137. G M Crippen J. Med. Chem. 24 198 (1981) 138. A K Ghose, G M Crippen J. Med. Chem. 25 892 (1982) 139. A K Ghose, G M Crippen J. Med. Chem. 27 901 (1984) 140. A K Ghose, G M Crippen J. Med. Chem. 28 333 (1985) 141. A K Ghose, G M Crippen, G R Revankar, P A McKernan, D F Smee, R K Robins J. Med. Chem. 32 746 (1989) 142. G M Crippen J. Comput. Chem. 8 943 (1987) 143. G M Crippen J. Comput.Chem. 10 673 (1989) 144. M P Bradley, G M Crippen J. Med. Chem. 36 4171 (1993) 145. L G Bouli, G M Crippen, H A Barton, H Kwon,M A Marletta J. Med. Chem. 33 771 (1990) 146. A J Hopfinger J. Am. Chem. Soc. 102 7196 (1980) 147. M Mabilia, R A Pearlstein, A J Hopfinger Eur. J. Med. Chem. 20 1133 (1985) 148. M G Koehler, K Rowberg-Schaefer, A J Hopfinger Arch. Biohem. Biophys. 266 152 (1988) 149. B J Burke, A J Hopfinger J. Med. Chem. 33 274 (1990) 150. A J Hopfinger, B J Burke,W J Dunn III J. Med. Chem. 37 3768 (1994) 151. B J Burke, W J Dunn III, A J Hopfinger J. Med. Chem. 37 3775 (1994) 152. P I Nagy, J S Tokarski, A J Hopfinger J. Chem. Inf. Comput. Sci. 34 1190 (1994) 153. J S Tokarski, A J Hopfinger J. Med. Chem. 37 3639 (1994) 154.K-B Rhyu,H C Patel, A J Hopfinger J. Chem. Inf. Comput. Sci. 35 771 (1995) 155. J W Dunn, A J Hopfinger, C Catana, C Duraiswami J. Med. Chem. 39 4825 (1996) 156. A M Doweyko J. Med. Chem. 31 1396 (1988) 157. A M Doweyko J. Math. Chem. 7 273 (1991) 158. A M Doweyko, W B Mattes Biochemistry 31 9388 (1992) 159. A M Doweyko J. Med. Chem. 37 1769 (1994) 160. J J Kaminsky, A M Doweyko J. Med. Chem. 40 427 (1997) 161. A N Jain, T G Dietterich, R H Lathrop, D Chapman, R E Critchlow, B E Bauer, T A Webster, T Lozanj-Perez J. Comput.-Aid. Mol. Design 8 635 (1994) 162. A N Jain, K Koile, D Chapman J. Med. Chem. 37 2315 (1994) 163. A N Jain, N L Harris, J Y Park J. Med. Chem. 38 1295 (1995) 164. R Todeschini, M Lasagni, E Marengo J. Chemometr.8 263 (1994) 165. R Todeschini, P Gramatica, R Provenzani Chemometr. Intell. Lab. Syst. 27 221 (1995) 166. R Todeschini, M Vigni, R Provenzani, A Finizio, P Gramatica Chemosphere 32 1527 (1996) 167. R Todeschini, C Bettiol, G Giurin, P Gramatica, P Miana, E Argese Chemosphere 33 71 (1996) 168. R Todeschini, P Gramatica Quantum Struct. ± Act. Relat. 16 113 (1997) 169. R Todeschini, P Gramatica Quantum Struct. ± Act. Relat. 16 120 (1997) 170. P G Mezey J. Chem. Inf. Comput. Sci. 32 650 (1992) 171. P G Mezey Shape in Chemistry. Introduction to Molecular Shape and Topology (New York: VCH, 1992) 172. P D Walker, P G Mezey J. Am. Chem. Soc. 115 12 423 (1993) 173. P D Walker, P G Mezey J. Am. Chem. Soc. 116 12 022 (1994) 174. P G Mezey, Z Zimpel, P Warburton, P D Walker, D G Irvine, S L Dixon, B Greenberg J.Chem. Inf. Comput. Sci. 36 602 (1996) 175. A Murcko,M A Murcko J. Med. Chem. 38 4953 (1995) 176. M Charton, in Lipophilicity in Drug Action and Toxicology (Eds V Pliska, B Testa, H Waterbeemd) (Weinheim: VCH, 1996) p. 387 177. C Dauwe, B Sellergren J. Chromatogr., A 753 191 (1996)522 178. P G Tratnyek, in Perspectives in Environmental Chemistry (Ed. D L Macalady) (New York: Oxford University Press, 1998) p. 167 179. L Quadri, A Cerri, P Ferrari, E Folpini,M Mabilia, P Melloni J. Med. Chem. 39 3385 (1996) 180. Z Nikolovska-Coleska, L Suturkova, K Dorevski, A Krbavcic, T Solmajer Quantum Struct. ± Act. Relat. 17 7 (1998) 181. L P Hammet J. Am. Chem. Soc.59 66 (1937) 182. M Charton Prog. Phys. Org. Chem. 13 119 (1981) 183. C Hanch, A Leo, R W Taft Chem. Rev. 91 165 (1991) 184. C Hansch, A Leo, D Hoekman Exploring QSAR. Hydrophobic, Electronic and Steric Constants (Washington, DC: American Chemical Society, 1995) 185. G Schuurman Quantum Struct. ± Act. Relat. 9 326 (1990) 186. O G Mekenyan, G D Veith SAR QSAR Environ. Res. 2 129 (1994) 187. S Karabunarliev, O G Mekenyan,W Karcher, C L Russom, S P Bradbury Quantum Struct. ± Act. Relat. 15 311 (1996) 188. B W Clare, C T Supuran J. Pharm. Sci. 83 768 (1994) 189. K J Miller J. Am. Chem. Soc. 112 8533 (1990) 190. R M Smith, P S Pearlman, in QSAR and Molecular Modelling. Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giraldo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p.222 191. J Jaerv, K Sak,M Eller, P Ek, A Engstroem, I Engstroem Bioorg. Chem. 24 159 (1996) 192. J D Hirst J. Med. Chem. 39 3526 (1996) 193. P Sjoberg, in Computer-Assisted Lead Finding and Optimization. Current Tools for Medicinal Chemistry (Eds H van de Waterbeemd, B Testa, G.Folker) (Weinheim: VCH, 1997) p. 83 194. S -S So,M Karplus J. Med. Chem. 39 5246 (1996) 195. V Pliska, B Testa, H Waterbeemd, in Lipophilicity in Drug Action and Toxicology (Eds V Pliska, B Testa, H Waterbeemd) (Weinheim: VCH, 1996) p. 1 196. A Avdeef, in Lipophilicity in Drug Action and Toxicology (Eds V Pliska, B Testa, H Waterbeemd) (Weinheim: VCH, 1996) p. 109 197. J C Dearden, G M Bresnen Quantum Struct.± Act. Relat. 7 133 (1988) 198. L-G Danielsson, Y-H Zhang Trends Anal. Chem. 15 188 (1996) 199. M L C Montanari, C A Montanari, D P Veloso, A E Beezer, J C Mitchell, P L O Volpe, Quantum Struct. ± Act. Relat. 17 102 (1998) 200. V N Viswanadhan, A K Ghose, G R Revankar, R K Robins J. Chem. Inf. Comput. Sci. 29 163 (1989) 201. R F Rekker, R Mannhold Calculation of Drug Lipophilicity (Weinheim: VCH, 1992) 202. G Klopman, J-K Wang,M J Dimayuga J. Chem. Inf. Comput. Sci. 34 752 (1994) 203. E Kellogg, D Abraham J. Comput.-Aid. Mol. Design 5 545 (1991) 204. H Waterbeemd, H Karajiannis, H Kansy, D Obrecht, K Muller, C Lehmann, in QSAR and Molecular Modelling. Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giraldo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p.78 205. N Bodor,M J Huang J. Pharm. Sci. 81 272 (1992) 206. M H Abraham, H S Chadra, in Lipophilicity in Drug Action and Toxicology (Eds V Pliska, B Testa, H Waterbeemd) (Weinheim: VCH, 1996) p. 311 207. H Waterbeemd, B Testa Adv. Drug. Res. 16 87 (1987) 208. N El Tayar, B Testa, in Trends in QSAR and Molecular Modelling 92 (Ed. C G Wermuth) (Leiden: ESCOM, 1993) p. 101 209. O A Raevsky, K-J Shaper, J K Seydel Quantum Struct. ± Act. Relat. 14 433 (1995) 210. O A Raevsky, in Computer-Aided Lead Finding and Optimization. Current Tools for Medicinal Chemistry (Eds H Waterbeemd, B Testa, G Folkers) (Wienheim: VCH, 1997) p. 367 211. H Waterbeemd, R Mannhold, in Lipophilicity in Drug Action and Toxicology (Eds V Pliska, B Testa, H Waterbeemd) (Weinheim: VCH, 1996) p.401 212. R Mannhold, K Dross Quantum Struct. ± Act. Relat. 15 403 (1996) 213. H Waterbeemd, R Mannhold Quantum Struct. ± Act. Relat. 15 410 (1996) 214. P A Carrupt, P Gaillard, F Billois, P Weber, B Testa, C Meyer, S Perez, in Lipophilicity in Drug Action and Toxicology O A Raevsky (Eds V Pliska, B Testa, H Waterbeemd) (Weinheim: VCH, 1996) p. 195 215. C Altomare, A Carotti, V Carta, S Kneubihler, P A Carrupt, B Testa, in QSAR and Molecular Modelling. Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giraldo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p. 463 216. M Laguerre,M Saux,A Carpy, in QSAR and Molecular Modelling.Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giraldo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p. 588 217. I Rozas,M Martin J. Chem. Inf. Comput. Sci. 36 872 (1996) 218. S Immel, F W Lichtenhaler Leibigs. Ann. Chem. 39 (1996) 219. T Fujita Quantum Struct. ± Act. Relat., 16 107 (1997) 220. B Testa, P- A Carrupt, P Gaillard, F Billois, P Weber Pharm. Res. 13 335 (1996) 221. C Hansch, D Hoekman, A Leo, L Zhang, P Li Toxicol. Lett. 79 45 (1995) 222. M T D Cronin, J C Dearden Quantum Struct. ± Act. Relat. 14 1 (1995) 223. M T D Cronin, J C Dearden Quantum Struct. ± Act. Relat. 14 117 (1995) 224. M T D Cronin, J C Dearden Quantum Struct. ± Act. Relat. 14 329 (1995) 225. P J Goodford J.Med. Chem. 28 849 (1985) 226. R W Taft, in Steric Effects in Organic Chemistry (Ed. MS Newman) (New York: Wiley, 1956) p. 556 227. M Charton Environ. Health Perspect. 61 229 (1985) 228. M Charton, in Similarity Models in Organic Chemistry, Bio- chemistry and Related Fields (Eds R I Zalewski, TMKrygowski, J Shorter) (Amsterdam: Elsevier, 1991) p. 629 229. K H Kim, Y C Martin, in QSAR, Rational Approaches to the Design of Bioactive Compounds (Eds C Silipo, A Vittoria) (Amsterdam: Elsevier, 1991) p. 151 230. F Sanz,M Martin, F Lapena, F Manaut Quantum Struct. ± Act. Relat. 5 54 (1986) 231. I Alkorta, H O Villar J. Med. Chem. 37 210 (1994) 232. D B Kireev, O A Raevsky, V I Fetisov, in Trends in QSAR and Molecular Modelling 92 (Ed. C G Wermuth) (Leiden: ESCOM, 1993) p.331 233. L Bonati, E Fraschini, M Lasagni, E P Madoni, D Pitea J. Mol. Struct. (Theochem) 340 83 (1995) 234. R Bureau, J C Lancelot, H Prunier, S Rault Quantum Struct. ± Act. Relat. 15 373 (1996) 235. A K Bhattacharjee, J M Karle J. Med. Chem. 39 4622 (1996) 236. R Carbo, B Calabuig Int. J. Quantum Chem. 42 1681 (1992) 237. R Carbo, B Calabuig J. Chem. Inf. Comput. Sci. 32 600 (1992) 238. E Besalu, L Amat, X Fradera, R Carbo, in QSAR and Molecular Modelling. Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giraldo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p. 396 239. X Fradera, L Amat, E Besalu, R Carbo-Dorca Quantum Struct. ± Act. Relat., 16 25 (1997) 240.R Carbo, L Leyde,M Arnau Int. J. Quantum Chem. 17 1185 (1980) 241. A C Good, S-S So,W G Richard J. Med. Chem. 36 433 (1993) 242. R Carbo, B Calabuig J. Mol. Struct. (Teochem) 254 517 (1992) 243. R Carbo, E Besalu, B Calabuig, L Vera Adv. Quant. Chem. 25 253 (1994) 244. D J Wild, P Willett J. Chem. Inf. Comput. Sci. 36 159 (1996) 245. L Pauling, P Pauling Chemistry (San Francisco:WH Freeman and Co., 1975) 246. P Magee Sci. Total Environ. 109/110 155 (1991) 247. N Froloff, A Faurion, P M Leod Chem. Sens. 21 425 (1996) 248. E Forgas, T Cserhati J. Liq. Chromatogr., Relat. Technol. 19 1849 (1996) 249. O A Raevsky Teor. Eksp. Khim. 22 450 (1986) d 250. O A Raevsky Usp. Khim. 59 375 (1990) [Russ. Chem. Rev. 59 219 (1990)] 251.O A Raevsky J. Phys. Org. Chem. 10 405 (1997) 252. O A Raevsky, V Yu Grigor'ev, V P Solov'ev, I V Martynov Dokl. Akad. Nauk SSSR 299 1433 (1988) e 253. R S Drago, B B Wayland J. Am. Chem. Soc. 87 3571 (1965) 254. R S Drago Structure and Bonding (Heidelberg: Springer, 1973) 255. M K Kroeger, R S Drago J. Am. Chem. Soc. 103 3250 (1981)Molecular structure descriptors in the computer-aided design of biologically active compounds 256. A V Iogansen Teor. Eksp. Khim. 7 302 (1971) d 257. O A Raevsky, V P Novikov Khim.-Farm. Zh. 16 583 (1982) b 258. I V Martynov, O A Raevsky Dokl. Akad. Nauk SSSR 265 664 (1982) 259. I V Martynov,O A Raevsky Vestn. Akad. Nauk SSSR (7) 93 (1983) 260. I V Martynov, O A Raevsky Zh. Vses. Khim. O-va im D I Mendeleeva 28 (1983) a 261.O A Raevsky,M M Gilyazov, Ya A Levin Zh. Obshch. Khim. 53 563 (1983) f 262. O A Raevsky,M M Gilyazov, Ya A Levin Zh. Obshch. Khim. 53 1720 (1983) f 263. O A Raevsky,M M Gilyazov, Ya A Levin Zh. Obshch. Khim. 53 1724 (1983) f 264. O A Raevsky, V Yu Grigor'ev, V P Solov'ev Khim.-Farm. Zh. 18 578 (1984) b 265. T G Valeeva, Ya A Levin, O A Raevsky Zh. Obshch. Khim. 55 1590 (1985) f 266. O A Raevsky, V Yu Grigor'ev, V P Solov'ev, A I Ivanov, V B Sokolov, I V Martynov Zh. Obshch. Khim. 57 2073 (1987) f 267. O A Raevsky, V Yu Grigor'ev, V P Solov'ev, I V Martynov Dokl. Akad. Nauk SSSR 298 1166 (1988) e 268. O A Raevsky, A F Solotnov, V P Solov'ev Zh. Obshch. Khim. 57 1240 (1987) f 269. A M Sapegin, O A Raevsky, V V Chistyakov, I V Martynov Khim.-Farm. Zh.21 578 (1987) b 270. O A Raevsky, in QSAR in Drug Design and Toxicology (Eds D Hadzy, B Jerman-Blazic) (Amsterdam: Elsevier, 1987) p. 31 271. O A Raevsky, V Yu Grigor'ev, V P Solov'ev Khim.-Farm. Zh. 23 1294 (1989) b 272. O A Raevsky, V Ju Grigor'ev, D B Kireev, N S Zefirov Quantum Struct. ± Act. Relat. 11 49 (1992) 273. R W Taft, D Gurka, L Loris, P R Schleyer, J W Rakshys J. Am. Chem. Soc. 91 4801 (1969) 274. J-L M Abboud, L Bellon Ann. Chim. 5 63 (1970) 275. M H Abraham, P E Grellier, D V Prior, R W Taft, J J Moris, P J Taylor,C Lourence,M Bertholot,R M Doherty,M J Kamlet, J-L M Abboud, K Sraidi, G Guiheneuf J. Am. Chem. Soc. 110 8534 (1988) 276. M H Abraham Chem. Soc. Rev. 22 73 (1993) 277.R W Taft,W J Shuely, R H Doherty,M J Kamlet J. Org. Chem. 53 1737 (1988) 278. P-C Maria, J-F Gal, J Franceschi, E Fargin J. Am. Chem. Soc. 109 483 (1987) 279. M H Abraham, in Quantitative Treatments of Solute/Solvent Interactions. Theoretical and Computational Chemistry Vol. 1 (Eds P Polizer, J S Murray) (Amsterdam: Elsevier, 1994) p. 83 280. O A Raevsky, V Ju Grigor'ev, V P Solov'ev, in QSAR Rational Approaches to the Design of Bioactive Compounds (Eds C Silipo, A Vittoria) (Amsterdam: Elsevier, 1991) p. 135 281. O A Raevsky, A M Sapegegin Khim.-Farm. Zh. 24 43 (1990) b 282. O A Raevsky, A M Sapegin, V V Chistyakov, A N Razdol'skii, N S Zefirov Dokl. Akad. Nauk SSSR 309 623 (1989) e 283. H-J Schneider, V Rmdiger, O Raevsky J. Org.Chem. 58 3648 (1993) 284. H-J Schneider, T Blatter, A Eliseev, V Rmdiger, O A Raevsky Pure Appl. Chem. 65. 2329 (1993) 285. O A Raevsky, I G Shnaider Zh. Org. Khim. 31 1793 (1995) g 286. V P Solov'ev, N N Strakhova, O A Raevsky, V Rudiger, H-J Schneider J. Org. Chem. 61 5221 (1996) 287. O A Raevsky, V V Chistyakov, R S Agabekyan, A M Sapegin, N S Zefirov Bioorg. Khim. 16 1509 (1990) h 288. I I Kitova, O A Raevsky, V G Blinova, N S Zefirov, in QSAR and Molecular Modelling. Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giralo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p.160 289. S O Bachurin, N V Lukoyanov, L N Petrova, L S Solyakov, S E Tkachenko, O A Raevsky Dokl. Akad. Nauk 346 549 (1996) e 290.O Raevsky, A Sapegin, N Zefirov Quantum Struct. ± Act. Relat. 13 412 (1994) 291. A N Razdol'skii, A M Sapegin, O A Raevsky, E V Shaveleva, G I Smirnova, N G Rozhkova Khim.-Farm. Zh. 26 (9 ± 10) 92 (1992) 523 292. U Z Mirkhodzhaev, N V Lukoyanov, I I Kitova, Kh F Abdulaev, O A Raevsky, A K Tashmukhamedova Uzb. Biol. Zh. (1) 8 (1991) 293. O Raevsky, V Ju Grigor'ev, D B Kireev, N S Zefirov J. Chim. Phys. Phys.-Chim. Biol. 89 1747(1992) 294. O A Raevsky, V Ju Grigor'ev, E Mednikova, in Trends in QSAR and Molecular Modelling 92 (Ed. C G Wermuth) (Leiden: ESCOM, 1993) p. 116 295. H Waterbeemd, G Gamenisch, G Folkers, O A Raevsky Quant. Struct.-Act. Relat. 15 480 (1996) 296. J Li, Y Zhang, A J Dallas, P W Carr J. Chromatogr. 550 101 (1991) 297.H Kubinyi, in Burger's Medicinal Chemistry and Drug Discovery (Ed. ME Wolff) (New York: Willey, 1995) p. 497 298. J J Hostynek, P S Magee Quantum Struct. ± Act. Relat. 16 473 (1997) 299. J Halova, O Strouf, P Z A Sochozova, N Uchida, T Yuzyri, K Sakakibara,M Hirota Quantum Struct. ± Act. Relat. 17 37 (1998) 300. W H Streng Drug Design Today 2 415 (1997) 301. L Y Wilson, G R Famini J. Med. Chem. 34 1668 (1991) 302. G R Famini, V P Loumbev, E K Frykman, L Y Wilson Quantum Struct. ± Act. Relat. 17 558 (1998) 303. G R Famini,W P Ashman, A P Miciewicz, L Y Wilson Quantum Struct. ± Act. Relat. 11 162 (1992) 304. G R Famini, L Y Wilson, C A Penski J. Phys. Org. Chem. 5 395 (1992) 305. A H Lowrey, C J Cramer, J J Urban, G R Famimi Comput.Chem. 19 209 (1995) 306. N A Chester, G R Famini, M V Haley, C W Kurnas, P A Sterling, L Y Wilson ACS Symp. Ser. 643 191 (1996) 307. R D Cramer III, D E Patterson, J D Bunce J. Am. Chem. Soc. 110 5959 (1988) 308. R J Vaz Quantum Struct. ± Act. Relat. 16 303 (1997) 309. W Lindberg, J A Persson, S Wold Anal. Chem. 55 643 (1983) 310. K Hasegava, T Kimura, K Funatsu Quantum Struct. ± Act. Relat. 16 219 (1997) 311. C Navajas, A Poso, K Tuppurainen, J Gynther Quantum Struct. ± Act. Relat. 15 189 (1996) 312. D B Kireev, J R Chretien,O A Raevsky Eur. J. Med. Chem. 30 395 (1995) 313. W E Steinmetz Quantum Struct. ± Act. Relat. 15 1 (1996) 314. M Akamatsu, Y Ozoe, T Ueno, T Fujita, K Mochida, T Nakamura, F Matsumura Pest. Sci.49 319 (1997) 315. T I Oprea, A E Garcia J. Comput.-Aid. Mol. Design 10 186 (1996) 316. S Sicsic, I Serraz, J Andrieux, B Bremont, M Mathe-Allanmat, A Poncet, S Shen, M Langlois J. Med. Chem. 40 739 (1997) 317. F Corelli, F Manetti, A Tafi, G Campiani, V Nacci,M Botta J. Med. Chem. 40 125 (1997) 318. D N A Boobbyer, P J Goodford, P M McWhinne, R C Wade J. Med. Chem. 32 1083 (1989) 319. R C Wade, K J Clark, P J Goodford J. Med. Chem. 36 140 (1993) 320. R C Wade, P J Goodford J. Med. Chem. 36 148 (1993) 321. G Cruciani, P J Goodford J. Mol. Graph. 12 116 (1994) 322. A M Davis, N P Gensmantel, E Johansson, D P Marriot J. Med. Chem. 37 963 (1994) 323. E E Polymeropoulos, N Hofgen Quantum Struct. ± Act. Relat. 16 231 (1997) 324. T Langer Quantum Struct. ± Act. Relat. 15 469 (1996) 325. M Pastor, G Cruciani J. Med. Chem. 38 4637 (1995) 326. M Itzstein, J C Dyason, S W Oliver, H F White, W-Y Wu, G B Kok,M S Pegg J. Med. Chem. 39 388 (1996) 327. S Clementi, G Cruciani, P Fifi, D Riganelli, R Valigi Quantum Struct. ± Act. Relat. 15 108 (1996) 328. V P Novikov, O A Raevsky Khim.-Farm. Zh. 16 574 (1982) b 329. I Csorvassy, L Tozser, in QSAR. Rational Approaches to the Design of Bioactive Compounds (Eds C Silipo, A Vittoria) (Amsterdam: Elsevier, 1991) p. 193 330. J W King, R Kassel, B B King Int. J. Quantum Chem. Quantum Biol. Symp. 17 27 (1990) 331. J W King, R J Kassel Int. J. Quantum Chem. Quantum Biol. Symp. 18 289 (1991) 332. J W King, R J Kassel Int. J. Quantum Chem. Quantum Biol. Symp. 19 179 (1992)O A Raevsky 524 a�Mendeleev Chem. J. (Engl.Transl.) b�Pharm. Chem. J. (Engl. Transl.) c�Russ. J. Coord. Chem. (Engl. Transl.) d�Theor. Exp. Chem. (Engl. Transl.) e�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) f�Russ. J. Gen. Chem. (Engl. Transl.) g�Russ. J. Org. Chem. (Engl. Transl.) h�Russ. J. Bioorg. Chem. (Engl. Transl.) 333. J W King Int. J. Quantum Chem. Quantum Biol. Symp. 20 139 (1993) 334. G R Famini, R J Kassel, J W King, L Y Wilson Quantum Struct. ± Act. Relat. 10 344 (1991) 335. J W King Int. J. Quantum Chem. Quantum Biol. Symp. 21 209 (1994) 336. B I Kuchkaev, B A Knyazev, V K Kurochkin, V A Petrunin Khim.-Farm. Zh. 26 (6) 60 (1992) b 337. B I Kuchkaev, B A Knyazev, V K Kurochkin, V A Petrunin Khim.-Farm. Zh. 31 (8) 30 (1997) 338. J Schuur, P Selzer, J Gasteiger J. Chem. Inf. Comput. Sci. 36 334 (1996) 339. J Gasteiger, J Sadowski, J Schuur, P Selser, L Steinhauer, V Steinhauer J. Chem. Inf. Comput. Sci., 36 1030 (1996) 340. O Raevsky, L Dolmatova, V Grigor'ev, I Lisyansky, S Bondarev, in QSAR and Molecular Modelling. Concepts, Computational Tools and Biological Applications (Eds F Sanz, J Giraldo, F Manaut) (Barcelona: J R Prous Science Publishers, 1995) p. 241 341. R Franke Theoretical Drug Design Method (Berlin: Springer, 1984) 342. J McFarland, D Gan Drug Inform. J. 24 705 (1990) 343. J McFarland, D Gan Quantum Struct. ± Act. Relat. 13 11 (1994) 344. H Matter J. Med. Chem. 40 1219 (1997) 345. E J Ariens, A M Simonis Top. Curr. Chem. 52 1 (1974) 346. K-t Chung, L Kirkovsky, A Kirkovsky,W P Purcell Mutat. Res. 387 1 (1997) 347. C A Montanari, M S Tute Quantum Struct. ± Act. Relat. 16 480 (1997) 348. Y Nezu, N Wada, F Yoshida,M Miyazawa, T Shimizu, T Fujita Pestic. Sci. 52 343 (1998) 349. H Kubinyi Drug Design Today 2 457 (1997) 350. V Austel, in Chemometric Methods in Molecular Design Vol. 2 (Ed. H Waterbeemd) (Dusseldorf: VCH, 1995) p. 49 351. W Dunn, S Wold, in Chemometric Methods in Molecular Design Vol. 2 (Ed. H Waterbeemd) (Dusseldorf: VCH, 1995) p. 179 352. S Wold, in Chemometric Methods in Molecular Design Vol. 2 (Ed. H Waterbeemd) (Dusseldorf: VCH, 1995) p. 195 353. R Franke, A Gruska, in Chemometric Methods in Molecular Design Vol. 2 (Ed. H Waterbeemd) (Dusseldorf: VCH, 1995) p. 115 354. A R Katritzky, L Mu,M Karelson J. Chem. Inf. Comput. Sci. 36 1162 (1996) 355. S C Basak, B D Gute, G D Grundwald J. Chem. Inf. Comput. Sci. 36 1054 (1996) 356. R Carcia-Domenich, C de Grigoria Alapont, J V de Julian-Ortiz, J Galvez, L Popa Bioorg. Med. Chem. Lett. 7 567 (1997) 357. S P Bradbury, O G Mekenyan, G T Ankley Environ. Toxicol. Chem. 15 1945 (1996) 358. D A Dawson, T W Schultz, R S Hunter Teratogen Carcinogen Mutagen 16 109 (1996) 359. N K Roy, E S J Nidiry, K Vasu, S Bedi, B Lalljee, B Singh J. Agric. Food Chem. 44 3971(1996) 360. A Cappelli, M Anzini, S Vomero, P G De Benedetti, M C Menziani, G Giorgi, C Manzoni J. Med. Chem. 40 2910 (1997) 361. J Polanski, J Gasteiger,M Wagener, J Sadowski Quantum Struct. ± Act. Relat. 17 27 (1998) 362. M Mracec, S Muresan, M Mracec, Z Simon, G Naray-Szabo Quantum Struct. ± Act. Relat. 16 459 (1997) 363. R Vaz Weed. Sci. 44 718(1996) 364. M Rowley, J J Kulagowski, A P Watt, D Rathbone, G I Stevenson, R W Carling, R Baker, G R Marshall, J A Kemp, A C Foster, S Grimwood, R Hargreaves, C Hurley, K L Saywell, M D Tricklebank, P D Leeson J. Med. Chem. 40 4053 (1997) 365. Y H Zhao,M T D Cronin, J C Dearden Quantum Struct. ± Act. Relat. 16 131 (1997) 366. M J B Mengelers, P E Hougee, L H M Janssen, A S J P A M van Miert J. Vet. Pharmacol. Ther. 20 276 (1997) 367. J S Tokarski, A J Hopfinger J. Chem. Inf. Comput. Sci. 37 792 (1997) 368. P Chiba,M Hitzler, E Richter, M Huber, C Tmej, E Giovagnoni, G Ecker Quantum Struct. ± Act. Relat. 16 361 (1997) 369. O A Raevsky, K-J Schaper Eur. J. Med. Chem. 33 799 (1998) 370. K C S C Liu, S-S Lee, M-T Lin, C-W Chang, C-L Liu, J-Y Hsu, S Ren, E J Lien Med. Chem. Res. 7 168 (1997) 371. R Garg, S P Gupta J. Enzyme Inhi
ISSN:0036-021X
出版商:RSC
年代:1999
数据来源: RSC
|
|