摘要:

Abstract. Data on the synthesis, conformational behaviour and complexing properties of calix[4]resorcinolarene and calix[n]arene derivatives modified by phosphorus-containing fragments are surveyed. The bibliography includes 123 references. I. Introduction The chemistry of calixarenes�cyclic products of condensation of phenols with aldehydes � has vigorously developed during the last 20 years.Calixarenes are readily accessible compounds (many of them can be easily obtained by a one-stage synthesis); they contain reactive sites permitting modification of the structure. In addition, they are capable of forming inclusion complexes of the `host±guest' type both with charged and non-polar molecules. The first mention of the interaction of phenols with alde- hydes 1 dates back to 1872 (A Bayer); however the structure of the reaction products has remained barely known for almost 70 more years.It was not until the 1940s that tetrameric cyclic structures A and B were ascribed to the products of condensation of p-alkyl- phenols with formaldehyde in an alkaline medium 2 and to the products of acid-catalysed reaction of resorcinol with aldehydes.3 The term `calixarene' was brought into use by Gutsche et al.4 to characterise the shape of these cyclic tetramers, which resembles a bowl (`calix' is the Greek for bowl or cup).Although later it was found 5±7 that these macrocycles can also exist in other confor- mations, this term has been adopted to denote the whole class of [1n]-metacyclophanes (including compounds C). Calixarenes are widely used as molecular platforms for the development of pre-organised three-dimensional receptors.5±8 Upon functionalisation of the phenolic groups, the aromatic rings and the bridging fragments in the initial macrocycles by appro- priate organic or organoelement reagents, the receptor capacity can be increased by a large factor.The number of publications dealing with calixarene derivatives doubles every three years.8 The number of patents also swiftly increases, indicating that these compounds are important not only for theoretical studies but also for practical purposes and for industry.The latest achievements in the studies of organic derivatives of calixarenes were summarised in a special issue of the journal Gazzetta Chimica Italiana (No. 11 in 1997), which was based on the contributions to the 4th International Conference on Calixarenes held in Parma (Italy) fromAugust 31 to September 4, 1997. Organic derivatives of phosphorus are of special interest for functionalisation of calixarenes, because phosphorus compounds play an important role in many biological and catalytic properties. The complex-forming ability of organophosphorus groups, the strictly definite size of the molecular cavity in the initial calixarene and the possibility of introducing fragments with a specified spatial orientation open up the way to the synthesis of a new generation of receptors capable of recognising complex molecular images.However, the size and the vicinity of functional groups in the supramolecular systems impart some specific features to their chemical behaviour.This makes development of an efficient methodology for selective functionalisation of calix[n]arenes a fundamentally important task. Studies on phosphorylation of calixarenes were started in the late 80s ± early 90s. This review is the first attempt to describe systematically experimental data now accumulated along this line.II. O-Phosphorylation of calixarenes by P(III) compounds The pioneering investigations dealing with the reactions of P(III) compounds with calixarenes were published by a group of American scientists,9± 13 who accomplished a series of studies dealing with functionalisation of the hydroxy groups in p-tert- butylcalix[4]arene (1). B HC R OH HO 4 A H2C OH R 4 C H2C n I S Antipin Kazan State University, ul.Kremlevskaya 18, 420008 Kazan, Russian Federation. Fax (7-843) 275 22 53. Tel. (7-843) 231 54 86. E-mail: iantipin@ksu.ru E Kh Kazakova, A I Konovalov A E Arbuzov Institute of Organic and Physical Chemistry of the Kazan Scientific Centre, ul. Arbuzova 8, 420083 Kazan, Russian Federation. Fax (7-843) 275 22 53. Tel. (7-843) 276 73 74 (E Kh Kazakova). Tel. (7-843) 276 82 54 (A I Konovalov) WD Habicher Institute of Organic Chemistry of Dresden Technical University, 13 Mommsenstrasse, D-01062 Dresden, Germany.Fax (49-0351) 463 40 93 Received 2 July 1998 Uspekhi Khimii 67 (11) 995 ± 1012 (1998); translated by Z P Bobkova UDC 547.260118 Phosphorus-containing calixarenes I S Antipin, E Kh Kazakova, WD Habicher, A I Konovalov Contents I. Introduction 905 II.O-Phosphorylation of calixarenes by P(III) compounds 905 III. O-Phosphorylation of calixarenes by P(V) compounds 911 IV. Phosphorylation of the aromatic rings in calixarenes 917 V. Introduction of phosphorus-containing fragments via a chain of atoms (spacer) 918 Russian Chemical Reviews 67 (11) 905 ± 922 (1998) #1998 Russian Academy of Sciences and Turpion LtdIt was shown that the inner cavity of calix[4]arene can stabilise the `hypervalent' hexacoordinated state of the phosphorus atom.The reaction of the compound 1 with tris(dimethylamino)-phos- phine 10 is accompanied by elimination of two dimethylamine molecules and gives spirocyclic zwitter-ionic product 2a (dp=7120 ppm, JP±H=737 Hz). The two apical positions are occupied by an N+HMe2 group and a hydrogen atom, the bond of the latter with phosphorus being directed inside the aromatic cavity; this was confirmed by NMR spectroscopy and X-ray diffraction analysis.12, 13 On treat- ment of the compound 2 with one equivalent of butyllithium, the dimethylammonium group is deprotonated, which results in the formation of salt 3.When the zwitter-ion 2 is thermolysed at 200 8C or treated with strong acids (CF3CO2H or HBF4), dime- thylamine is eliminated and cage phosphite 5 is produced.Spectral studies (1H and 31P NMR) showed that the reaction with HBF4 occurs via the formation of pentacoordinated phosphorane inter- mediate 4 (dp = 722 ppm). Oxidation of phosphite 5 with tert- butyl hydroperoxide affords the corresponding phosphate 6; according to X-ray diffraction analysis, this product occurs in a conformation intermediate between 1,2-alternate and partial cone.The reaction of the phosphite 5 with one equivalent of n-butyl- lithium gives product 7 with a hexacoordinated phosphorus atom. The same group of researchers 14 succeeded in the introduc- tion of more than one phosphorus atom in the calixarene plat- form.Treatment of the compound 1 with di[bis(N,N- dimethylamino)phosphino]methane in dichloromethane at room temperature gives rise to disubstituted product 8. An X-ray diffraction study showed that the compound 8 is the (R,S) isomer in relation to the phosphorus atoms. Binding of the remaining hydroxyl groups upon the reaction with tris(diethyla- mino)phosphine gave the product 2b and diphosphine.Keeping compound 8 in toluene for 14 h at 1008C resulted in the evolution of two moles of diethylamine and the formation of product 9 containing no free hydroxyl groups. Subsequently this product can be converted into bisphosphonate 10 or bis(phosphono- thioate) 11. NMR monitoring of the formation of the bis(phos- phonothioate) 11 showed that reaction consists of two stages.The monothio-derivative formed initially is responsible for two 31P NMR signals, with dp = 84 and 179 ppm (JP±P = 29 Hz). To convert this compound into the bisthio-derivative, an additional heating for 24 h is required. When the compound 8 is made to react with electrophilic reagents such as phenyl azide, methyl iodide and tetrachloro-1,2-benzoquinone, compounds 12 ± 14 are formed in which only one phosphorus atom of the diphosphino- methylene bridge has changed its coordination.The products 12 and 14 are unstable and rapidly decompose. OH But But OH OH OH But But 1 7 + 2a O But But O O O But But P H NHMe2 P(NMe2)3 7NHMe2 7 BunLi O But But O O O But But P H NMe2 3 Li+ O But But OH O O But But P H NHMe2 + 4 H+ D O But But OH O O But But P 5 [O] O 5 O But But HO O O But But P 6 BunLi O But But O O O But But 7 Li+ 5 Bun P (Me2N)2PCH2P(NMe2)2 1 7 + 2b O But But O O O But But P H NHEt2 (Et2N)3P OH But But HO O O But But P P NMe2 Me2N 8 O But But O O O But But P P 9 O But But O O O But But P P E E 10, 11 8 100 8C E=O (10), S (11) 906 I S Antipin, E Kh Kazakova,WD Habicher, A I KonovalovSynthesis of polyfunctional phosphinite derivatives has been reported in several studies.15 ± 17 Both p-tert-butylcalix[4]arene 1 itself and its 1,3-diethers 15 have been subjected to phosphoryla- tion by chloro(diphenyl)phosphine.Treatment of these substrates with lithium diisopropylamide or butyllithium at a temperature below 50 8C followed by reaction with Ph2PCl results in the selective formation of bis(phosphinites) 17.These products exist in a cone conformation, except for the compound 17a in which fast interconversion of the cone and partial cone conformations occurs. When Et3N is used as the base in boiling THF, the compound 15d can be converted only into the monophosphinite 16d having a cone conformation. Under the same conditions, the calixarene 1 is converted into 1,3-bisphosphinite as a mixture of two conformers, the cone and the 1,2-alternate. The structures of the 1 : 1 complexes formed by 16d with Au(I) cations and the 2 : 1 complexes with Pd(II) and Pt(II) cations have been studied by X-ray diffraction analysis.18 Deprotonation of the dimethyl ether 15a by BunLi in THF followed by the reaction with Ph2PCl at room temperature also gives a mixture of conformers, cone 18 (25%) and 1,2-alternate 19 (60%).At lower temperatures (778 8C), the content of the conformer 19 increases to 82%; the two possible 1,2-alternate structures of this compound undergo fast interconversion.15 It has been noted 19 that treatment of the calixarene 15a with two equivalents ofKHfollowed by reaction with Ph2PCl results in the selective formation of diphosphorylated product 20 in a partial cone conformation.The data of dynamic 1H NMR in solution point to the occurrence of a dynamic equilibrium between the cone 18 and partial cone 20 conformations, the former conformer predominat- ing (the ratio 18 : 20=2.7 : 1).16 The number of introduced diphenylphosphinite fragments can be controlled by preliminary methylation of the phenolic hydroxy groups by methyl tosylate in the presence of potassium carbonate.19 The complex-forming ability of the phosphinite ligands 16d and 17 in relation to transition metal cations [Pt(II), Pd(II) and Rh(I)] has been studied.16, 17 The introduction of P(III) atoms into the lower rim of calix[4]arene by preliminary silylation of hydroxy groups has been described by Schmutzler et al.20, 21 Treatment of the calixar- ene 1 with hexamethyldisilazane gives rise to 1,3-disubstituted derivative 21.The reaction of this product with PF2Cl affords mono- and bis-difluorophosphites, which are converted into monofluoro- phosphite 22 upon elimination of Me3SiF or PF3. The macro- cycles 21 and 22 occur in a cone conformation. In order to synthesise completely O-phosphorylated calix[4]- arenes, the compound 1 is first converted into the corresponding tetraanion by treatment with a strong base (BunLi or KH).The highly nucleophilic tetraanion is made to react with phosphorus- containing electrophiles with P ± Cl bonds. Thus the reaction of the lithium derivative 23a with N-substituted 2-chloro-4-oxo- benzo-1,3,2-diazaphosphorinanes or 2-chloro-4,6-dioxobenzo- 1,3,5,2-triazaphosphorinane yields calixarenes 24a ± c.The mac- rocycles 24d,e were synthesised by similar reactions of the sodium derivative 23b with 3-tert-butyl-2-chloro-1,3,2-oxa- azaphospholidine and chloro(diphenyl)phosphine.22 ± 24 The stereochemical outcome of the above reactions is determined by the nature of the organophosphorus group entering the molecule.The compounds 24a, d, e exist exclusively in a cone conforma- tion,23 whereas for compounds 24b, c all of the four possible conformers are formed.21 Under similar conditions, calix[6]arene is converted into hexakis(diphenylphosphinite),24 while calix[8]- 8 14 MeI PhN3 O O Cl4 O O Me2N Me2N P P But But O O HO But But OH Cl4 OH But But HO O O But But P Me2N NMe2 Me 13 + I7 P OH But But HO O O But But P P Me2N NMe2 NPh 12 OH But But Y OH X But But 15a ± h 16d Y But But OPPh2 X OH But But Y But But OPPh2 X OPPh2 But But 17 X=Y=OMe (a), OEt (b), OPrn (c), OCH2COOEt (d), OCH2COOC10H19 (e); X=OEt: Y=OCH2COOEt (f), OCH2COOC10H19 (g), OCH2COO[(7)-menthyl)] (h). OPPh2 But But OMe OPPh2 OMe But But 15a a, b (a) BunLi, THF; (b) Ph2PCl. 18 Ph2PO But But O Me Me O But But OPPh2 + 19 OMe But OPPh2 Ph2PO But But But OMe 20 Phosphorus-containing calixarenes 907arene gives octaphosphorylated derivatives with diphenylphos- phine and 1,3,2-oxaazaphospholane fragments.22 The cyclophane 24c in a cone conformation reacts with dichloro(cycloocta-1,5-diene)platinum(II) giving 1 : 1 trans-com- plex 25, in which two phosphorus atoms are coordinated to the Pt(II) cation.20 The tetrakis(phosphinites) 24e and 24d also form stable complexes.For example, homodimetal complexes of 24e with Fe(CO)3,23 [Pd(COD)Cl2] and Pt[COD]Cl2 (COD is cyclo- octa-1,5-diene) 25 and a polymeric 1 : 1 complex of 24d with copper(I) chloride 23 have been obtained. The tetradentate ligand 24e is capable of forming both dimeric complexes, for example, {(24e)[Rh(CO)Cl]2}, and binuclear heterometal complexes, for example, [(24e){NiCl2}{Mo(CO)4}].25 Phosphorylation of the calix[4]arene 1 26 and the calix[6]arene 26 27 with phosphorus(III) chloride (in some cases, in the presence of Et3N as a base) yields cyclic bisphosphites 27 and 28.The product 27 is hydrolytically unstable; an attempt to recrystallise it from methanol resulted in the isolation of the initial calixarene 1.According to X-ray diffraction data, the lone electron pairs of the phosphorus atoms in the compounds 27 and 28 point in the same direction; this makes them promising bidentate ligands for complexation with many d- and f-metal cations. Thus the compound 28 forms cis-complexes with Pd(II), Pt(II) and Re(I). In the complex [PdCl2L], the metal coordinates the bisphosphite fragments at an angle of 90 8C.Calix[6]arene derivatives containing cyclic (29) and acyclic (30) phosphorus fragments were synthesised by the reaction of calixarene 26 with phosphorous diethylamide.28 These compounds were found 22 to exhibit antioxidant properties. 25 OR But But O O But But OR P P Pt Cl Cl N N N CO OC CO NMe N OC MeN Me Me Me Me O But But O O O But But PCl PCl 1 27 PCl3 But But But But But But OH OH HO HO OH OH 26 PCl3 But But But But But But O O O O O O P P 28 But But But But But But OH O O O HO O P P R R 29 R= (a), (b), (c), N P NCH2C6H4F-4 O Me N P NMe O Me MeN NMe P N O O Me PPh2 (e).(d), O P NBut 22 1 O But But HO OH But But O P F OR But But OR OR OR But But 24a ± e OM But But OM OM OM But But 23a,b M =Li (a), Na (b) OSiMe3 But But HO OSiMe3 OH But But 21 908 I S Antipin, E Kh Kazakova,WD Habicher, A I KonovalovDue to the conformational flexibility of calix[4]resorcinol- arenes 31,29, 30 their phosphorylation can follow two major path- ways: (a) functionalisation of separate hydroxy groups or (b) cyclophosphorylation of two neighbouring hydroxy groups.It is obvious that the reaction route depends both on the nature of the phosphorylating reagent (most of all, the number of reaction sites) and on the conformation of the macrocycle.The most favourable conformation of octol 31 for cyclophosphoryla- tion is the crown conformation (C4u symmetry). In this conforma- tion, the aromatic rings are directed upwards and form a bowl-like cavity, stabilised by a system of intramolecular hydrogen bonds.29, 30 Monochlorophosphines, monochlorophosphites, dichlorophosphines and aminophosphines are usually employed as phosphorylation agents. Phosphorylation using aminophosphines is distinguished by experimental simplicity and mild reaction conditions (room temperature).31 A series of P(III)-containing cavitands 32 have been prepared by the reactions of calixarenes 31b ± e with a number of amino- phosphines (the reactants were taken in 1 : 4 molar ratio).32 ± 37 The influence of the nature of the substituentRin the bridging fragment on the reaction route has been studied.36 It was found that the reaction of tris(diethylamino)phosphine with tetraphenyl- octol 31a mainly follows pathway (a).Even when the ratio of the reactants is 1 : 2, the proportion of the cyclophosphorylation product is only 23%.In the case of tetramethyloctol 31b, product mixtures containing both cyclic and acyclic phosphorus-contain- ing fragments are formed at various ratios of the reactants. For tetrabutyloctol, cyclophosphorylation becomes the prevailing process; when the reactant ratio is 1 : 4, only pathway (b) is realised.It was suggested 36 that the dependence of the phosphor- ylation route on the nature of the substituentRis a consequence of the calix[4]resorcinolarene preorganisation (i.e. the position of the conformation equilibrium in the given solvent). Bulky alkyl radicals stabilise the crown conformation, due to lipophilic interactions with one another, and thus they promote cyclo- phosphorylation.The reactions of aminomethylated calix[4]re- sorcinol-arenes at a 1 : 4 reactant ratio also follow pathway (b). Recently triaminophosphines were used to phosphorylate p-tert-butylcalix[5]arene. This substrate reacts with P(NMe2)3 to afford monocylic aminophosphine, O,O0-(dimethylaminophos- phinediyl)calix[5]arene.38 It is noteworthy that, unlike ordinary aminophosphines, aminophosphine cavitands 32 are stable against alcohols, water and oxygen and add sulfur only under rigorous conditions.Oxidation of these compounds or addition of sulfur gives macro- cyclic phosphoroamidates or phosphorothioamidates 33. Octa- phosphorylated macrocycles 34 react with sulfur in a similar way. But But But But But But OH O O O RPO O P P R R NEt2 30 R= NH.O R=Ph (a), Me (b), Bun (c), C6H13 (d), C7H15 (e), C9H19 (f), C11H23 (g), (CH2)2Ph (h). OH OH OH HO HO HO HO OH R R R R 31a ± h R R X2PO OPX2 a b R O O O O P X 32b ± e O O O O R R R R O O O O XP PX PX XP R R R R HO HO OH OH HO OH OH HO 31b ± e a (a) 4 equiv. (R02N)2PX; R0=Me, Et; X=NMe2, NEt2; E =O, S. S or [O] 32b ± e O O O O R R R R O O O O XP PX PX XP E E E E 33 R=Ph, Me; X=NEt2, OMe; Y =NEt2.R R R R YX(S)PO YX(S)PO OP(S)XY OP(S)XY YX(S)PO OP(S)XY OP(S)XY YX(S)PO R R R R YXPO YXPO OPXY OPXY YXPO OPXY OPXY YXPO 34 31a,b 4 (R02N)2PX S Phosphorus-containing calixarenes 909When calixarenes 31d,e were made to react with methyl and ethyl phosphorous diamides, the phosphates 33 were isolated from the reaction mixture instead of the expected macrocyclic phosphites 32; the latter proved to be extremely unstable and were easily oxidised to phosphates by atmospheric oxygen.The macro- cyclic phosphites 32 were detected only in the reactions of 31d,e with heptyl and octyl phosphorous diamides.39 Phosphorylation of the calixarene 31e with (7)-menthyl tetraethylphosphorous diamide under mild conditions with differ- ent ratios of the reactants affords completely and partially phosphorylated optically active derivatives containing both cyclic phosphite groups and acyclic menthyl phosphorous amide frag- ments (for example, compound 35).When the ratio of the reactants is 1 : 2 and the reaction conditions are more drastic the `ball' structure 36 is formed.40 The reaction of the calixarene 31d with 2-chloro-1,3,2-benzo- dioxaphospholene at any ratio of the reactants gives phosphite 37, occurring in an equilibrium with a slight amount of spirophos- phorane 38.41 Cavitands containing acyclic phosphinite (39) and cyclic phosphonite (40) fragments have been prepared by the reaction of the calixarene 31h with chloro(diphenyl)phosphine and dichlor- o(phenyl)phosphine, respectively, in the presence of pyridine.42 The high complexation capacity of these calixarenes was con- firmed by introducing them into reactions with Pt(II) and Au(I) salts.The structures of the resulting complexes (their composition is 1 : 4 or 1 : 8) depend on the nature of the metal (its coordination number). For the 1 : 4 complex, an unusual arrangement of the AuCl groups was established. Three of these groups are located at the upper rim of the calixresorcinolarene due to coordination to the phosphorus atom, while the fourth AuCl group resides in the macrocycle cavity; as a result, solvent molecules are arranged in the intermolecular space rather than inside the cavity.Only in the presence of amines, does the inclusion endo-complex 41 form. This property of the complex was used for extraction of alkali metal picrates.42 An unusual m3- and m4-bridging position of halide anions has been established by X-ray diffraction analysis for inclusion com- plexes 42 and for the complexes [Py][40 .Ag4(m-Cl)4 . (m4-Cl)] and [Py][40 . Cu4(m-Cl)4 . (m3-Cl)], in which one of the Cl7 anions is located in the cavity formed by the four cationic centres.43, 44 However, the bridging anion is only weakly fixed in the complex and can be easily replaced by other halide anions.The use of phosphorus(III) halides for phosphorylation of calix[4]resorcinolarenes has been described.45, 46 The reaction of the macrocyclic compound 31b with excess PCl3 affords cyclic phosphorochloridites, similar to compounds 32 (X = Cl), rather than octakis(phosphorodichloridites) 34 (X = Y = Cl).45 The structure of the reaction products was confirmed by the data of IR R0=Me, Et; X=OC7H15, OC8H17; Y=OMe, OEt.O O O O R R R R O O O O P P P P X X X X 32 O O O O R R R R O O O O P P P P Y Y Y Y O O O O 33 31d,e 4 (R02N)2PX 4 (R20N)2PY, O2 R=C7H15, R0 is menthyl. 31e R O R R OH OH O HO O R R P O O P O P O R O O O R R O OH 36 O P R R R R R0OPO R0OPO O OPOR0 R0OPO O OPOR0 R0OPO POR0 NEt2 NEt2 NEt2 NEt2 NEt2 NEt2 35 8 (Et2N)2POR0 2 (Et2N)2POR0 R=CH2CH2Ph R R R R Ph2PO Ph2PO OPPh2 OPPh2 Ph2PO OPPh2 OPPh2 Ph2PO 39 40 O O O O R R R R O O O O PhP PPh PPh PhP HO 2 37 O P O O O P O O C6H13 C6H13 O H 2 O 38 O P O O H O P O O H C6H13 C6H13 O 42 Cu Cl Cu Cl Cl Cu Cu Cl Cl PhP PPh PPh O O O O R R R R O O O O PhP H2N CH2 H2C CH2 Me Au Cl Au Cl 41 Au Cl Au Cl 910 I S Antipin, E Kh Kazakova,WD Habicher, A I Konovalovspectroscopy, mass spectrometry and elemental analysis.The halogen atom in these compounds can be readily replaced by a methyl or dimethylamino group upon the reaction with MeMgI or Me3SiNMe2, respectively. The latter reaction gives the cyclic phosphorous amide 32b, which reacts with tetrachloro-o-benzo- quinone or hexafluoroacetone yielding spirophosphorane struc- tures containing four s5l5 phosphorus atoms.The octakis(phosphorodifluoridite) 34 (X = Y = F) was formed in the reaction with PF2Cl 46 only when the calix[4]resor- cinolarene 31g had been first treated with excess BunLi. This product was relatively unstable and decomposed over a period of 24 h at 735 8C to give PF3 and a mixture of various compounds containing P(=O)F(O) and P(=O)F2 fragments. III.O-Phosphorylation of calixarenes by P(V) compounds Phosphorylation of the tetrahydroxy-derivative 1 with diethyl phosphorochloridate was carried out in the presence of sodium hydride,22 potassium carbonate in THF47 or triethylamine in chloroform.48, 49 In all cases, the reaction gave bisphosphate 43a, which exists in a cone conformation.Synthesis of the compound 43a by the Atherton ± Todd reaction (diethyl phosphite ± triethylamine ± tetrachloromethane, 0 8C) has been reported.50 ± 53 The phosphorylation of 25,27- dihydroxy-26,28-dimethoxy-p-tert-butylcalix[4]arene with diethyl phosphorochloridate in the presence of NaH47 gives bisphos- phate in the 1,2-alternate conformation in which the methoxy groups are oppositely directed with respect to the main plane of the macrocycle.A detailed study 49 of the reaction of the compound 1 with diethyl chlorophosphate in the presence of triethylamine in chloroform has shown that under these conditions, in addition to the previously described bisphosphate 43a, occurring in a cone conformation, a 1,2-alternate with anti-orientation of the dieth- oxyphosphoryl groups with respect to the main plane of the macrocycle is also formed (its structure in the crystal was determined in another study 54).The content of this compound in the reaction mixture changes from 9% to 15% following an increase in the excess of triethylamine. Phosphorylation of unsub- stituted calix[4]arene under the same conditions gives the bisphos- phate 43b exclusively in a cone conformation.Alkylation of the phenolic hydroxy groups in the phosphate 43a with ethyl iodide in the presence of NaH (THF :DMF=4 : 1) gives rise to 5,11,17,23- tetra-tert-butyl-25,27-diethoxy-26,28-bis(diethoxyphos-phorylox- y)calix[4]arene in a partial cone conformation.49 Phosphorylation of the calixarene 1 with excess diethyl phosphorochloridate 50 or diethyl phosphite 55 under phase trans- fer catalysis conditions (CCl4±H2O, 50% NaOH, tetrabutylam- monium bromide) yields tetraphosphorylated derivative 44. When a less lipophilic catalyst (tetraethylammonium bro- mide) is used, the reaction affords a mixture of the tetraki- s(phosphate) 44 in a partial cone conformation and bis(phosphate) 45, which contains both cyclic and acyclic phos- phate fragments.56 Complex-forming properties of a number of phosphates based on calix[4]arene have been studied by extrac- tion.57 It was found that the introduction of these compounds in the organic phase markedly increases the extraction of lanthanide picrates; however, no pronounced selectivity was observed. Mild oxidation of the compound 1 gives rise to spirodienone 46, which can be used as the starting compound in the synthesis of mono- and 1,2-di-substituted calixarenes.58, 59 In particular, treat- ment of this compound with lithium diisopropylamide (THF, 778 8C) followed by the reaction of the resulting mono- or di- lithium salt with excess diisopropyl phosphorochloridate and subsequent reduction of the spirodienone fragment result in the formation of the corresponding phosphorylated calixarenes 47.Synthesis of asymmetrically substituted phosphorylated cal- ix[4]arenes 50 has been described by Markovskii et al.60 The treatment of 1,3-bis(diethoxyphosphoryl)calix[4]arenes 43a,b with sodium hydride in boiling benzene followed by interaction with electrophilic reagents, such as benzoyl chloride or methyl bromoacetate, affords asymmetrically substituted calix[4]arenes 50a ± c of the AABH type.The use of diethyl phosphorochlor- idate as an electrophilic reagent gives rise to triphosphorylated calixarenes 50c. R=P(O)(OEt)2, X=But (a), H (b). OR X X HO RO OH X X 43a,b OR But But RO RO OR But But OH But But O O O But But P P EtO EtO O OEt O 44 45 R=P(O)(OEt)2 O OH OH O But But But But 1 46 a b, c, d R=P(O)(OPri)2, R0=H(a); R =R0 =P(O)(OPri)2 (b); (a) Me3PhN+Br¡3 , CH2Cl2, 0 8C; (b) LiNPri 2, THF,778 8C; (c) ClP(O)(OPri)2; (d) EtOH, HBr.OH OR0 OR But But But But HO 47a,b 43a,b ONa R1 R1 OR2 HO OR2 R1 R1 48 Phosphorus-containing calixarenes 911The key stage in the synthesis is the formation of the unstable monoanion of 48 and its O,O-phosphorotropic rearrangement to the 1,2-bis(diethoxyphosphoryl)calix[4]arene monoanion of 49; the rearrangement occurs via phosphorane 51, which was detected by 31P NMR spectroscopy.All these products exist in a cone conformation, while the asymmetrically substituted calix[4]arenes 50a,b have C1 symmetry and are formed as racemic mixtures, which is indicated by doubling (Dd = 0.01 ppm) of the phosphorus signals in the 31P NMR spectra recorded in a chiral solvent. O-Phosphorylation of calix[5]-, -[6]- and -[8]arenes is less studied.Data on selective phosphorylation are available only for calix[6]arene. Selective functionalisation of the hexahydroxy- derivative 26 by diethyl phosphorochloridate and O,O-diethyl phosphorothiochloridate has been performed and the conforma- tional behavior of the products synthesised has been studied.61, 62 In the case of diethyl phosphorochloridate, the reaction was carried out with various amounts of triethylamine in boiling chloroform.In all cases, the obtained product mixtures were analysed by HPLC and separated by column flash chromatog- raphy. This gave mono-, 1,3-bis-, 1,4-bis-, and 1,2,4-tris(phos- phorylated) derivatives 52a ± d.Depending on the reaction conditions, the optimum yields of these compounds vary from 28% to 56%. It is of interest that in another study,48 1,3,5-tris(phosphorylated) product was obtained in 60% yield under the same conditions. The fully substituted derivative 52e was synthesised in 20% yield under conditions of phase transfer catalysis in the CH2Cl2 ± 50% NaOH system in an excess of diethyl phosphorochloridate.The use of anhydrous solvent and solid alkali permitted the yield to be increased to 37%.61 Under the same conditions, pentakis(- diisopropoxyphosphoryloxy)calix[5]arene 26 (in a cone conforma- tion) and octakis(diethoxyphosphoryloxy)calix[8]-arene 63 have been obtained. It was also noted 63 that phosphorylation of the compound 26 with diethyl phosphorochloridate can give not only the product 52e but also the pentaphosphorylated derivative 52m with one cyclic phosphate fragment.The use of sodium hydride as a base in a THF±DMFmixture (10 : 1) affords the compound 52e as a flattened 1,2,3-alternate conformation in 75% yield.64 When the calixarene 26 is made to react with a large excess of less reactive O,O-diethyl phosphorothiochloridate in the CH2Cl2 ± 50% aqueous NaOH system, the 1,4-bis(thiophos- phorylated) product 52f is formed selectively in 60% ±70% yield.Variation of the conditions of phase transfer catalysis makes it possible to synthesise also mono-, 1,3-bis-, and penta- kis-thiophosphorylated derivatives 52g ± i.65 The phosphates 52j ± l were prepared by phosphorylation of partially methylated calix[6]arenes under similar conditions.66 Study of the mono- (52g) 62 and 1,4-bis(thiophosphorylated) (52f) 61 products by X- ray diffraction analysis and 1H NMR spectroscopy (solution in CDCl3) showed that they exist in a pinched cone conformation (C2u symmetry), in which the two opposite methylene bridges are directed towards the centre of the macrocycle.It is notable that 1,4-diphosphorylated compound 52c exists in a 1,2,3-alternate conformation. According to NMR spectroscopy, the macrocycle 52j in CDCl3 solutions exists as an equilibrium of two conformers, cone (C3u symmetry) and 1,2,3-alternate (Cs symmetry), the former conformation being 4 kJ mol71 more stable.66 The researchers believe 66 that the cone conformation is stabilised due to the weak CH± p interactions of the methyl groups, which, according to X-ray diffraction analysis, are directed inside the ring.OR2 R1 R1 NaO HO OR2 R1 R1 49 OR2 R1 R1 R3O HO OR2 R1 R1 50a ± c R1=H or But, R2=P(O)(OEt)2; R3=CH2C(O)OMe (50a), PhCO (50b), P(O)(OEt)2 (50c). O P O O7 OEt EtO O P(OEt)2 O OH 51 R1=OP(O)(OEt)2, R2=R3=R4=R5=R6=OH(a); R1=R3=OP(O)(OEt)2, R2=R4=R5=R6=OH(b); R1=R4=OP(O)(OEt)2, R2=R3=R5=R6=OH(c); R1=R2=R4=OP(O)(OEt)2, R3=R5=R6=OH(d); R1=R2=R3=R4=R5=R6=OP(O)(OEt)2 (e); R1=R4=OP(S)(OEt)2, R2=R3=R5=R6=OH(f); R1=OP(S)(OEt)2, R2=R3=R4=R5=R6=OH(g); R1=R3=OP(S)(OEt)2, R2=R4=R5=R6=OH(h); R1=R2=R3=R4=R5=OP(S)(OEt)2, R6=OH(i); R1=R3=R5=OMe, R2=R4=R6=OP(O)(OEt)2 (j); R1=R3=R5=OMe, R2=R4=R6=OP(S)(OEt)2 (k); R1=R2=OMe, R3=R4=R5=R6=OP(S)(OEt)2 (l).R1 R2 R3 R4 R5 R6 CH2 52a ± l But But But But But But O OR OR O OR OR P EtO O 52m (R=P(O)(OEt)2) OH OH OH OH HO O P EtO O OEt 52a But But But But But But But OH OH OH OH HO O P EtO O OEt But But But But But 912 I S Antipin, E Kh Kazakova,WD Habicher, A I KonovalovStudy of partially phosphorylated and thiophosphorylated p-tert-butylcalix[6]arenes, for example 52a, by dynamicNMRhas shown 62 that these macrocycles participate in, at least, three dynamic processes: (a) interconversion of the macrocyclic ring; (b) redistribution of hydrogen bonds; (c) interconversion of two pinched cone conformations.Monosubstituted derivatives proved to be conformationally more rigid than 1,3- and 1,4-disubstituted derivatives. This is due to the fact that the activation energy of conformation transitions depends on the number of hydrogen bonds, which can be formed by the hydroxy groups that remain free.The contributions of hydrogen bonding to DG= are 45 and 31 kJ mol71 for the mono- and di-substitiuted derivatives, respectively. The calix[4]arenes 53, containing two or four dihydroxyphos- phoryl groups on the `lower' rim of the macrocycle, were prepared in high yields 67 from readily acessible bis- and tetrakis-diethox- yphosphoryl derivatives of calixarenes 43 and 44.When these products are treated with excess bromotrimethyl- silane in a solution in anhydrous chloroform, the corresponding silyl esters are prepared in quantitative yields. On treatment of these compounds with methanol, the labile P ±O± Si bond rup- tures; this gives the acids 53 as colorless crystalline compounds, readily soluble in polar organic solvents (alcohols, DMSO, DMF) and in aqueous solutions of alkali; they are stable in aqueous alcohols and in amine ± alcohol mixtures for several days.The compounds 53a ± c are relatively strong acids [pKa 2.85, 3.10 and 2.90 (aqueous methanol)]. They form 1 : 1 and 1 : 2 exo-complexes with tert-butylamine, a-phenylethylamine and ephedrine; the complexes exist in a solution in deuterated methanol as contact ion pairs.67 In order to develop selective agents for the transfer of lithium cations through liquid membranes, the phosphates 54 were synthesised by the reaction of 25,26,27-trimethoxy-28-hydroxy- p-tert-butylcalix[4]arene with alkyl phosphorodichloridates fol- lowed by hydrolysis.68 Spectral data indicate that the compounds 54a ± c occur in solution as mixtures of two conformers, a cone and a partial cone, the equilibrium being sensitive to the influence of the medium.The fraction of the cone conformation in which all the four dipoles point in the same direction increases with increase in the solvent polarity.It is noteworthy that the complexation of the compounds 54 with Li+ and Na+ cations occurs more readily for the cone conformation of the calixarenes; therefore, no macro- cycles in the partial cone conformation were found in the com- plexes. Pyrolysis in vacuo (230 ± 240 8C) of mono-, bis-, and tetrakis- phosphates 43, 44 and 47, irrespective of the number of phospho- rus-containing groups, gives the same product, pyrophosphate 55, in which all the hydroxy groups are bound intramolecularly.26, 69 This means that the formation of the compound 55 involves stages of fragmentation and intermolecular migration of the phosphate groups.Study of the crystal structure of the pyrophosphate 55 showed that the two terminal oxygen atoms and the oxygen atom of the P ±O± P bridge are directed away from the plane of the molecule.The conformation of the pyrophosphate can be defined as a pinched or flattened cone. Pyrolysis of a mixture of regioisom- ers of phosphorylated p-tert-butylcalix[6]arenes yields isomeric bridged bis(phosphates) 56, which differ in the arrangement of the phosphoryl oxygen atoms (either syn- or anti-). Yet another pathway to phosphorus-containing bridged mac- rocyclic structures is reaction of the compound 1 with phosphorus halides.Thus refluxing of 1 with POCl3 in chloroform in the presence of Et3N gives rise to the mono-bridged system 57. At higher temperatures (xylene, 85 8C), the reaction affords structure 58 with two bridges, which might be a mixture of stereoisomers with different spatial arrangements of the P ± Cl and P=O bonds.23 43, 44 R=H, X=H, Y =(HO)2PO (a); R=But: X=H, Y=(HO)2PO (b); X=Y=(HO)2PO (c).OY R R XO YO OX R R 53a ± c 1. SiMe3Br 2. MeOH R=Me (a), Et (b), Prn (c). 54 OMe But O OMe But But O OR OH P MeO But OMe But But OMe O OMe But But P OH OR O O O O But But But But O P P O O O 55 O O O But But But P O O P But But But O O O 56 OPOCl2 O OH But But But But O P O Cl 57 Phosphorus-containing calixarenes 913Interestingly, pyrolysis of the compound 57 also gives the pyrophosphate 55.Glode et al.70 have studied the reactivity of the compound 57 towards nucleophilic reagents such as water, alcohols and orthoformates. It was found that the halogen atoms can be successfully substituted by hydroxy or alkoxy groups to give the corresponding acids or esters; these reactions do not involve the P ±O±C bonds.70 When the calixarene 1 is treated with three equivalents of PCl5, the macrocyclic compound 59, containing three phosphorus atoms, is formed.71, 72 According to 31P NMR spectroscopy, one of these atoms forms a phosphonium cation upon linking to the three phenolic groups, a second phosphorus atom forms the quasiphosphonium group OPCl4, while the third atom occurs as the PCl¡6 anion (dP = 8.1, 766.0 and 7296 ppm, respectively).The X-ray diffraction analysis of the compound 59 confirmed the presence of tetra-, penta- and hexa-coordinated phosphorus atoms. The macrocycle conformation is intermediate between a partial cone and 1,2-alternate.73 All the three phosphorus atoms in the compound 59 are reactive.On treatment with SbCl5, the PCl¡6 anion is replaced by SbCl¡6 ; reactions with nucleophiles occur as substitution at the P ± Cl bond and, under more rigorous con- ditions, also at the P ±O bond. For example, the reaction of 59 with water gives phosphate 60 in which the P ±O±C bonds remain intact. However, upon prolonged action of acids or alkali, the 1,2,3-linked phosphate is converted into 1,2-phosphate 61 as a result of hydrolysis of one P ±O±C bond.The compounds 60 and 61 contain free hydroxy groups and can be easily converted into the corresponding esters on treatment with diazomethane or orthoformates. Reactions with sulfur-containing nucleophiles give phosphorothioates 62 and 63, the character of binding in the lower rim of calix[4]arene remaining invariable; this points to high stability of phosphates and phosphorothioates bound to the 1,2,3-positions of the calix- arene system.According to 1H and 13C NMR spectroscopy and X-ray diffraction analysis, the compounds 60 ± 63 occur in a partial cone conformation.71, 72, 74, 75 The reaction of the calix[6]arene 26 with four equivalents of PCl5 yields hexakis(chlorophosphate) 64, which, like 59, is capa- ble of exchanging the PCl¡6 anion for the SbCl¡6 .72 O O O But But But But O P P O O Cl Cl 58 R=H, Me, Et.OH But But OP(OR)2 O O But But O P OR O 57 HC(OR)3 HOR 7HCl OH But But OP(OR)2 O O But But O P Cl O X = O (56), S (65). + + O O O But But But P Cl Cl P But But But O O O 64 2 PCl¡6 H2X O O O But But But P X X P But But But O O O 56, 65 EtSH NaOH (HCl) H2O H2S O But O O But But Cl4PO But P Cl + PCl¡6 59 (HO)2PO 60 O But O O But But But P O O But O O But But Cl2(S)PO But P S 63 O But O O But But HO But P S 62 OH But O O But But (HO)2PO But P O HO 61 NaOH (HCl) O O 914 I S Antipin, E Kh Kazakova,WD Habicher, A I KonovalovThe reactions with water and hydrogen sulfide afford the phosphate 56 and phosphothioate 65.The phosphoryl oxygen and sulfur atoms point in the same direction in relation to the main plane of the macrocycle. This makes these compounds promising chelating agents for cations of many d- and f-elements. The reaction of the compound 26 with ethyl phosphorodi- chloridate gives rise to cyclic mono- and bis-phosphates.76 The reaction route depends on the nature of the base used.In the presence of CsF, monophosphate is formed, while the use of calcium hydride results in the formation of the diphosphorylated product.76 The spatial structure of the macrocycles obtained was not determined. The calix[4]resorcinolarenes 31 are readily phosphorylated by traditional reagents, chlorides of acids of tetracoordinated phos- phorus and the (RO)2PHO± CCl4±Et3N ternary system.The reactions of the octahydroxy-derivative 31 with excess diorganyl phosphorochloridates and triethylamine (20 8C, THF) afford octaphosphorylated cyclophanes 66.77 ± 79 The reaction outcome depends on the nature of the radical in the diorganyl phosphorochloridate. Even when excess phosphor- ylating reagent is used, the reactions with diethyl and dipropyl phosphorochloridates give tetraphosphorylated products 67 together with octasubstituted derivatives.In the case where diisopropyl phosphorochloridate is employed, the reaction stops at the stage of formation of the tetraphosphorylated product, due to the steric hindrance during the process and the low solubility of the corresponding tetrasubstituted products, resulting in their removal from the reaction area.79 The tetrakis(phosphates) 67 were obtained by using a large excess of the reagent (31 : phos- phorochloridate : triethylamine = 1 : 5 : 5) and by treating the compound 31 with the (RO)2PHO± CCl4±Et3N ternary system (31 : reagent=1 : 10).It should be noted that in early studies,77, 78 the structure with alternating positions of the phosphoryl groups along the macrocycle rim had been assigned to the tetrakis(phos- phates) formed in these reactions; however, later it was found 79 that a structure of the type 67 in which the phosphoryl groups occupy the opposing positions is more probable.The tetraki- s(phosphates) 67 were further modified. Thus hydrolysis of the phosphate groups in 67, via the intermediate synthesis of the corresponding trimethylsilyl ester, gave tetrakis(dihydroxy- phosphoryloxy)tetrahydroxycalix[4]resorcinolarene 67g.81 Reac- tion of the free hydroxy groups in 67 with 4-chlorosulfonylbenzo- 15-crown-5 in the presence of triethylamine gave the derivative 67h.82 A study of the conformational behaviour of the tetrakis(phos- phates) 67 by dynamic NMR spectroscopy has shown that at room temperature, fast pseudo-rotation of the boat ± boat type occurs, which consists in the mutual transition of the two pairs of opposing benzene rings between the perpendicular and coplanar arrangements with respect to the main macrocycle plane.79 Estimation of the complex-forming properties of the cyclo- phanes synthesized showed that they can form stable complexes both with some metal cations and with neutral aromatic mole- cules.The phosphoryl groups present in the molecule impart pronounced ionophoric properties to it. When the octakis(phos- phate) 66d is used as an ion transferring agent, the rate of transfer of alkali and alkaline earth metal picrates through a liquid membrane increases 1.3 ± 9.0-fold compared to that attained with 15-crown-5; however, the transfer is almost nonselective.79 The compound 67h has demonstrated selectivity towards potas- sium cations.82 The octakis(phosphates) 66c,d, owing to their lipophilic cavity, form crystalline `host ± guest' complexes with aromatic compounds, toluene and m-cresol.79 Later,80 the struc- tures of several complexes of this type in solution and in the solid state were studied by X-ray diffraction analysis and NMR spectroscopy.The cyclophane 67g forms `host ± guest' complexes with benzene, toluene, anisole, xylenes, phenol and 1-phenyl- alanine in aqueous solutions. The complexes result from inclusion of a `guest' molecule into the cavity formed by the preorganised pentyl radicals.81 Tetrakis(phenylenephosphoryloxy)[1.4]metacyclophane 68a was synthesised by the reaction of the octahydroxy-derivatives 31b with four equivalents of o-phenylene phosphorochloridate and triethylamine.78, 79 Since the phosphorus atoms are located closely to the hydroxy groups of the neighbouring aromatic rings, intramolecular nucle- ophilic addition of the hydroxy groups to the P=O bonds occurs to give spirophosphorane structure 69a.The driving force of this process is formation of the energetically more favourable penta- coordinated phosphorus atom in the dioxaphospholane ring.This reaction is reversible; the position of the equilibrium depends on the temperature and the nature of the solvent. On treatment with triethylamine, the 68a>69a equilibrium shifts towards the more acidic form, 68a. A similar equilibrium is observed for the 68b>69b system.In pyridine, which is capable of binding to a pentacoordinated phosphorus atom,83 the equilibrium shifts towards the spirocyclic tautomer 69b. R1=Me: R2=Et (a), Prn (b), Ph (c), p-MeC6H4 (d); R1=n-C5H11, R2=Ph (e). R1 R1 R1 R1 PO PO OP OP PO OP OP PO O O O O O O (OR2)2 (OR2)2 (OR2)2 (OR2)2 (R2O)2 (R2O)2 (R2O)2 (R2O)2 66a ± e O O X=H, R1=Me: R2=Et (a), Prn (b), Pri (c), Bui (d); X=H, R1=n-C5H11: R2=Et (e), Pri (f), H (g); X= ; R1=n-C5H11, R2=Et (h).O O O O O SO2 R1 R1 R1 R1 XO XO OX OX PO OP OP PO O O O O (OR2)2 (OR2)2 (R2O)2 (R2O)2 67a ± h X = H (a), SiMe3 (b). 69a,b O O P O O OX 2 68a,b XO O P O O O 2 Me Me Me Me Phosphorus-containing calixarenes 915When the compound 31f (R=C9H19) was made to react with excess bis(chloromethyl)phosphinic chloride (acetone, K2CO3), instead of the expected bis(chloromethyl)phosphinic acid deriva- tives, octakis(chloromethylphosphonate) was isolated, due to the extremely easy rupture of one of the P ±C bonds in the phosphi- nate.The product exists in solution as dimer 70 formed according to the `head-to-head' pattern due to intermolecular hydrogen bonds.84 The first representatives of monophosphorylated cyclophanes 71 were prepared by refluxing in benzene of aminomethylated octahydroxy-derivatives with diethyl phosphorochloridate.85 The use of organyl phosphorodichloridates or organylphos- phonodichloridates for phosphorylation of the cyclophanes 31 has been reported.84, 86 ± 88 The reaction with alkyl and aryl phosphorodichloridates in acetone in the presence of triethyl- amine results in the formation of diastereoisomeric cavitands 72, which differ in the spatial arrangement of the P=O groups in the four phosphocine rings: i (the P=O group points to the centre of the upper rim) and o (the P=O group points outwards).86, 87 The reaction with ethyl phosphorodichloridate proceeds with a low yield (12%), owing to the competing intermolecular cross-linking of the calixarene fragments, resulting in the formation of poly- mers.The use of aryl phosphorodichloridates permits the yields of the target products to be increased to 50%± 80%. The reaction gives all of the six possible stereoisomers, although three of them, iiio, iioo and iooo, proved to be more preferable (the proportion of each isomer was determined by HPLC).The proportions of the iiii (compound 72) and oooo isomers in the reaction mixture are not higher than 10%; in some cases, they do not form at all. It is quite natural that as the substituent R3 becomes larger, the content of the iiio isomer increases, whereas the content of the iooo isomer correspondingly decreases. It is clear that the receptor capacity of the macrocyclic compounds obtained depends on the spatial arrangement of the phosphoryl groups.The iiii and iiio isomers are the most promising complex-forming agents. Thus the iiio isomer forms a complex with cyclohexylammonium (Ka = 1370 litre mol71) as a result of three-point hydrogen binding, whereas the iioo and iooo isomers do not form complexes of this type. It has been reported 88 that the iiii stereoisomer of compound 73 is formed as the major product in the reaction of the octahydroxy-derivative 31a with phenylphosphonic dichloride.The publication does not present the conditions of the synthesis. The product was identified as having the iiii configuration based on its C4u symmetry (although the oooo isomer has the same symmetry). The strong complexing properties of the reaction product served as additional evidence for its iiii configuration.The compound 73 forms stoichiometric complexes with alkali metal and alkylammonium cations, characterised by association constants of 108± 1010 litre mol71; the composition of the com- plexes is 1 : 1 (K+, Rb+, Cs+, NHá4 , MeNHá3 , But NHá3 ) or 1:2 (Li+, Na+). The reaction of the long-chain calix[4]resorcinolarene 31f with chloromethylphosphonic dichloride in the presence of triethyl- amine gives completely phosphorylated product 74; when potas- sium carbonate is used as the base, partially phosphorylated product 75 is formed.84 The use of POCl3 (in the presence of pyridine as the base) for introduction of phosphate groups has been reported.89 In relation to C-undecylcalix[4]resorcinolarene, it was shown that at various ratios of the reactants (ranging from 1 : 1 to 1 : 39), complex mixtures of cyclic and acyclic phosphates with different numbers of phosphate fragments (from 1 to 6.3) are formed.Complete phosphorylation was not attained even with a large excess of POCl3. These compounds exhibit surfactant properties and form stable complexes with UO2á 2 . 70 2 XO OX R R OX XO OX OX R R OX OX 31f (ClCH2)2P(O)Cl Me2CO, K2CO3 R=C9H19, X = CH2Cl. O P OH ClP(O)(OEt)2 R R R R HO HO OH OH HO OH OH HO CH2NMe2 CH2NMe2 CH2NMe2 Me2NH2C R=Me, Hexn. R R R R HO HO OH O HO OH OH HO CH2NMe2 CH2NMe2 CH2NMe2 Me2NH2C P(OEt)2 O HCl 71 R1=Ph, Alk; R2=H, Br; R3=OEt, OAr. O O O O R1 R1 R1 R1 O O O O P P P P R3 R3 R3 R3 O O O O R2 R2 R2 R2 72 O O O O R R R R O O O O P P P P Ph Ph Ph Ph O O O O 73 (R=Me) 916 I S Antipin, E Kh Kazakova,WD Habicher, A I KonovalovIV. Phosphorylation of the aromatic rings in calixarenes Phosphorylation of the aromatic rings in calixarenes gives rise to a new type of receptors in which, unlike O-phosphorylated deriva- tives, the molecular cavity of the cyclophane can also be involved in the molecular binding of a substrate.The presence of organo- phosphorus compounds in the upper rim of the macrocycle permits the receptor to interact both with polar or ionic groups of a substrate and with non-polar fragments, the latter being included in the molecular cavity of the cavitand. Thus, multi-point `host ± guest' binding is attained, which is known to increase sharply the efficiency and selectivity of complex formation.90, 91 Moreover, P ±C bonds are thermally and hydrolytically more stable than P ±O bonds, which are formed upon phosphorylation of the lower rim of calix[n]arenes. The first representative of this type of cyclophanes, tetrakis(- diphenylphosphino)calixarene 76, was synthesised by successive reactions of the tetramethyl ether of p-tetrabromo-calix[4]arene 76a with butyllithium and chloro(diphe-nyl)phosphine.92 Nevetheless, the Arbuzov reaction catalysed by nickel(II) salts is now the simplest and the most efficient method for direct phosphorylation of the upper rim of calix[4]arenes.93 The reaction of the compound 76a with triisopropyl phosphite or isopropyl diphenylphosphinite gives the corresponding tetra- phosphorylated calix[4]arenes 76c,d in high (75% ± 80%) yields.The same approach was used in another study 94 in order to prepare bis- and ABAB-tetrakis-phosphorylated derivatives. The reactions of O-alkylated dibromides 77a,b with triisopropyl phosphite or alkyl diphenylphosphinite in the presence of NiBr2 gave bisphosphonates 78a,b and bis(phosphine oxide) 79 (Scheme 1).Treatment of the compounds 78 and 79 with octyl chloromethyl ether in the presence of SnCl4 resulted in the introduction of CH2Cl groups into the free p-positions of the aromatic rings. It is of interest that in the case of the methyl ethers 78a and 79, the chloromethyl derivatives are formed in high yields (60% ± 70%); when the dodecyl ether 77b is used, the yield decreases to 26%.It has been reported that the size of the alkyl radical has a considerable influence on the kinetics of phosphor- ylation of calix[4]resorcinolarenes.95, 96 This was explained by assuming formation of different types of aggregates in the solution. The subsequent substitution of the chlorine atom by phosphorus-containing nucleophiles gave rise to ABAB-tetra- phosphorylated cyclophanes 80 ± 82.The first representative of calixarenes containing a low- coordinate phosphorus atom, phosphinylidenemethylcalixarene 83, was synthesised as a mixture of stable Z- and E-stereoisomers capable of forming complexes with transition metals, by the following reaction:97 R R R R HO O O OH HO O OH O P CH2Cl O P O ClH2C 75 R R R R O O O O O O O O P CH2Cl O P O ClH2C P P O ClH2C O CH2Cl 74 31f ClCH2P(O)Cl2 NEt3 PhH orMe2CO K2CO3 CH2Cl2 R=Br (a), PPh2 (b), P(O)(OPri)2 (c), P(O)Ph2 (d).OMe R R OMe MeO OMe R R 76a ± d OH PrO HO PrO CH O OH PrO HO PrO CH PMes 83 H2PMes PhMe, 110 8C Et2O.BF3 81 80 2 OP(OPri)2 OR OR0 CH2 CH2Cl 2 OP(OPri)2 OR OR0 CH2 P(O)Ph2 2 OP(OPri)2 OMe OMe CH2 OP(OPri)2 d c (a) P(OPri)3, NiBr2; (b) C8H17OCH2Cl, SnCl4; (c) AlkOPPh2; (d ) P(OPri)3, R=R0=Me.R=R0=Me (a); R=Me, R0 =C12H25 (b). 2 Br OR OR0 CH2 78a,b 77a,b a 2 OP(OPri)2 OR OR0 CH2 b 77a C8H17OCH2Cl SnCl4 AlkOPPh2 NiBr2 2 OPPh2 OMe OMe CH2 79 2 OPPh2 OMe OMe CH2 CH2Cl 2 OPPh2 OMe OMe CH2 OP(OEt)2 82 NaOP(OEt)2 Scheme 1 Phosphorus-containing calixarenes 917V. Introduction of phosphorus-containing fragments via a chain of atoms (spacer) A vigorously developing route of the functionalisation of calix[- n]arenes aimed at producing more efficient and selective complex- forming agents is introduction of organophosphorus groups in which phosphorus atoms are not attached directly to the lower or upper rim of the macrocycle but are linked to the calixarene platform via a chain of other atoms (spacer). This approach makes it possible to extend substantially the synthetic potential in the construction of new macrocyclic receptors.Deprotonation of 1,3-bis(ethoxycarbonylmethyl) calix[4]- arene derivative 84a with sodium tert-butoxide and subsequent alkylation of the resulting dianion with Ph2P(O)CH2OTs gives rise to bis(phosphine oxide) cyclophane 85a in a cone conforma- tion.98 When potassium tert-butoxide is used as the base in this reaction, the product 85a is formed in a partial cone conformation.The crucial influence of the nature of the cation on the stereo- chemical outcome of the reaction was explained by the difference in the energy of binding of potassium and sodium cations to the intermediate diphenoxide. The harder sodium cation interacts more strongly with phenoxide anions and thus it fixes the cone conformation of the calixarene platform.When the compounds 85 are refluxed with phenylsilane in toluene, the phosphine oxide group is quantitatively reduced to give the corresponding bis(phosphine) macrocycles 86; according to X-ray diffraction and 1H and 13C NMR data, this product also occurs in a cone conformation.98 The compounds 86 can act as both cis- and trans- chelating ligands.cis-Complexes are formed upon the reactions of the calixarene 86a with PtCl2(PhCN)2 and [Rh(COD)Cl]2 (87) in the presence of AgBF4. The reaction of the compound 86a with RuCl3 and CO in boiling 2-ethoxyethanol affords the complex 88 with trans-P,P stereochemistry in a quantitative yield.98 This method was used to prepare a series of metal-containing calixarenes99±104 able to participate in catalytic processes.Based on the bis(phosphine) cyclophanes 86, a number of new Pt(II), Pd(II), Rh(I) and Ru(II) complexes were prepared. The reaction of the compound 86c with the platinum(II) bis(triphenylphosphine) complex in boiling THF is accompanied by displacement of a PPh3 molecule and the chloride anion from the metal coordination sphere and yields the cationic complex 89.The replacement of the ester side groups by amide groups, which interact more strongly with cations, crucially changes the reaction route. In this case, both triphenylphosphine ligands are replaced and neutral hydridochloride complex 90 is formed. Apparently, this outcome is due to the participation of two amide groups in the interaction with Pt(II); this accelerates the formation of intermediate species with pentacoordinated plati- num and facilitates the ligand substitution.101 ± 103 The structure of the complex 90 was studied.Two-dimen- sional (ROESY) NMR spectra and X-ray diffraction data attest convincingly that the hydride ion, responsible for the resonance signal at d = 715.04 ppm (JH±Pt=1150 Hz, JP±H=15 Hz), points to the centre of the calixarene cavity.101, 103 The stereo- chemistry of the Pt ±H bond can be changed by involving the amide groups in complex formation. Thus the reaction of the complex 90 with silver tetrafluoroborate yields hydrido complex 91 in which the Pt ±H bond is directed outwards.102, 103 In this environment, the hydride becomes much more reactive and can participate in various reactions.It is of interest that addition of a strong donor to the complex 91 results in displace- ment of the amide group from the coordination sphere yielding complex 92, in which hydrogen is again directed inwards. This reaction makes it possible to change the reactivity of the hydride ion and creates conditions for the synthesis of bimetal hydrido complexes upon reactions with bidentate donors.103 R=C(O)OEt (a), C(O)NEt2 (b), CH2OMe (c), C(O)NHCH(Me)Ph (d). 84a ± d OH But But OCH2R HO RCH2O But But 1. 2 NaOBut, THF ±DMF 2. 2 Ph2P(O)CH2OTs But But OCH2R OCH2PPh2 RCH2O But But 86a ± d OCH2PPh2 But But OCH2R RCH2O But But 85a ± d O OCH2PPh2 O OCH2PPh2 PhSiH3 R=CH2CO2Et; COD is cycloocta-1,5-diene. O But But RO O OR But But PPh2 Ph2P RhCOD 87 O But But RO O OR But But PPh2 Ph2P Ru Cl Cl CO OC 88 But R=C(O)NEt2 (86b, 90), CH2OMe (86c, 89); (a) (Ph3P)2PtHCl, THF; (b) (Ph3P)2PtHCl, CH2Cl2.Cl O But But OCH2R O RCH2O But Ph2P PPh2 Pt PPh3 H 89 + Cl7 O But But OCH2R O RCH2O But But Ph2P PPh2 Pt H 90 86b,c a b X is a donor. X O But But O O O But But Ph2P PPh2 H 92 O NEt2 NEt2 O + O But But O O O But But Ph2P PPh2 Pt H 91 BF¡4 O Et2N NEt2 O Pt X 918 I S Antipin, E Kh Kazakova,WD Habicher, A I KonovalovSelective introduction of one, two or four CH2P(O)Ph2 sub- stituents into the lower rim of tert-butylcalix[4]arene 1 was attained by using NaH and an appropriate alkylating reagent.105, 106 When iodomethyl(diphenyl)phosphine oxide is employed as the alkylating reagent, the reaction gives, depending on the temperature and the amount of NaH used, either mono- (93) or 1,3-bis-(phosphine oxide) 94.The reaction with more reactive hydroxymethyl(diphenyl)phosphine oxide tosylate allows the syn- thesis of the corresponding 1,2-di- (95) and tetra-substituted (96) derivatives. According to NMR and X-ray diffraction analysis, the cyclophanes 93 ± 96 occur in a cone conformation; they are widely used as complexing reagents and membrane transferring agents for various metal cations.107 ± 109 The reduction of the compound 96 with phenylsilane gives a tetrakis(phosphine) ligand capable of forming polynuclear complexes.105, 110 A new series of polyfunctional phosphine oxide derivatives of calix[4]-, -[6]- and -[8]arenes 97, containing (CH2)2P(O)Ph2 frag- ments, have been modelled and synthesised for the extraction and complexation of actinide and lanthanide ions.111 The synthetic strategy is presented in the scheme.The properties of these compounds as extractants have been studied.112 ± 115 The further development of the methods for functionalisation of calixarenes followed the path of introduction of groups containing both an organophosphorus fragment and another binding site into the upper or lower rim of the macrocycle.Cavitands with the carbamoylmethylphosphine oxide [NHC(O)CH2P(O)Ph2]116 ± 118 and a-aminophosphonate [NHCR2P(O)(OR02)2] 119 groups have now been obtained. It has been shown 116 that the extraction properties of the traditional extractant for actinides, (N,N-diisopropylcarbamoyl- methyl)octylphenylphosphine oxide (CMPO), can be markedly improved by grafting four or five carbamoylphosphine oxide groups to the upper rim of calix[4]- or -[5]arenes 98 and 99.It is worth noting that the extraction capacity of the specially synthesised116 linear analogues of receptors 98 and 99�the dimer and the trimer 100a,b �was equal to that of the initial CMPO. These results demonstrate the synergistic effect of combining carabamoylphosphine oxide fragments with the calixarene plat- form within the same molecule.The compounds 98 and 99 were prepared by aminolysis of p-nitrophenyl diphenylphosphorylace- tate with O-tetra- or O-pentaalkyl ethers of p-aminocalix[4]- or -[5]arenes. The calixarenes 98 occur in a cone conformation, whereas 99 can be a mixture of cone, partial cone and 1,3-alternate, because for mixed ethers with two methoxy groups, inversion of the macrocycle is possible.Several studies have been devoted to the introduction of carbamoylmethylphosphonate fragments into the calix[4]resor- cinolarenes 31 and to the extraction properties of the products towards a number of metal cations.117, 118 Mono-, 1,2-bis-, 1,3-bis, tris- and tetrakis(carbamoylmethyl)phosphonates and -phosphine oxides 101 were synthesised starting from tetrakis(bromo-meth- yl)calix[4]resorcinolarene.The conditions for partial functionali- sation were created at the stage of formation of the aminomethyl derivatives. The amino group was introduced via phthalimide derivatives, by their hydrazinolysis. The transition to the cyclo- phanes 101 was accomplished either by acylation of the amines with chloroacetyl chloride followed by Arbuzov introduction of the phosphorus-containing fragment or by aminolysis of p-nitro- phenyl diphenylphosphorylacetate.(a) 3 equiv. NaH, 2.2 equiv. Ph2P(O)CH2I, THF, 50%; (b) 5 equiv. NaH, 2.5 equiv. Ph2P(O)CH2I, PhMe, 80%; (c) 2.5 equiv. NaH, 2.2 equiv. Ph2P(O)CH2OTs, THF±DMF, 25%; (d ) 6 equiv.NaH, 4.2 equiv. Ph2P(O)CH2OTs, THF±DMF, 70%. OH But But HO O OH But But 93 Ph2(O)P O But But HO O OH But But P(O)Ph2 94 Ph2(O)P 1 a b d c O But But O O O But But P(O)Ph2 Ph2(O)P Ph2(O)P 96 P(O)Ph2 Ph2(O)P 95 O But But HO OH O But But P(O)Ph2 R=But, H; n=4, 6, 8. O(CH2)2OH R n O(CH2)2OTs R n OCH2CO2Et R n 97 O(CH2)2PPh2 R n O(CH2)2P(O)Ph2 R n 2 OR HN O O NH OMe 99a ± c OR HN O n 98a ± h n=4: R = C5H11 (a), CH2CHBu2 (b), C10H21 (c), C12H25 (d), C14H29 (e), C16H33 (f), C18H37 (g); n=5, R = C14H29 (h).CH2P(O)Ph2 CH2P(O)Ph2 CH2P(O)Ph2 R=CH2CHBu2 (a), C10H21 (b), C18H37 (c). m = 0 (a), 1 (b). 100a,b m OC18H37 HN O O NH OC18H37 O HN OC18H37 CH2P(O)Ph2 CH2P(O)Ph2 CH2P(O)Ph2 R=H: L=Ph (a), OEt (b); R =Prn: L=Ph (c), OEt (d). NR O L2(O)P O O C5H11 4 101a ± d Phosphorus-containing calixarenes 919Study of the extraction of Na+, Cs+, Sr2+, UO2á 2 , Fe3+ and Eu3+ with the compounds 101 showed that Na+ and Cs+ are not extracted, while the degree of extraction of Sr2+ reaches 85%.The cavitands containing three or four carbamoylmethylphosphine oxide groups were found to be the most efficient extractants. As the number of these groups decreases, the efficiency of extraction and its selectivity with respect to Eu3+ decrease.Interestingly, 1,2- disubstituted cavitands are more efficient than the 1,3-isomers. Calixarenes 102 and 103, which contain a-amino phosphonate fragments in the lower and upper rims of the macrocycle, have been constructed as receptors and membrane transferring agents for a-amino acids occurring in the zwitter-ionic form.119 These compounds were synthesised by the Kabachnik ± Fields reaction from the corresponding 1,3-bis(aminoalkyl)-derivatives, diethyl phosphite and a carbonyl compound (yields 50%± 65%).The calixarenes 102 and 103 occur in a cone conformation. It should be noted that transfer of highly hydrophilic amino acid zwitter-ions through lipophilic organic membranes is a difficult task due to the very high energy of hydration of these species.Therefore, the receptor should interact efficiently with as many active sites in the substrate as possible. As applied to a-amino acids, this can be interaction with the carboxylate and ammonium groups and with the side chain, which normally does not contain any polar or charged substituents.The presence of a-amino phosphonate fragments, containing both proton-donor (NH bond) and proton-acceptor (P=O bond and the lone electron pairs of nitrogen) groups together with the calixarene molecular cavity in the receptor molecule ensures efficient trans- port of a-amino and a-hydroxy acids.120, 121 The cavitands 102 and 103 exhibited high selectivity of transfer of a-amino acids, the compounds 102 being more efficient towards histidine, and 103, towards d,l-phenylalanine.A water-soluble receptor for the uranyl ion UO2á 2 was synthesised by introducing six phosphonic acid fragments [CH2P(O)(OH)2] into the calix[6]arene upper rim. This was done by the Arbuzov reaction of the corresponding chloromethyl derivatives with triethyl phosphite and subsequent hydroly- sis.122, 123 This review surveys the main routes of functionalisation of calix[4]resorcinolarenes and calix[n]arenes by organophosphorus compounds of various types and clearly demonstrates the great potential and the prospects for the use of these macrocycles for complex formation and molecular recognition (extraction, mem- brane transfer) of both charged and neutral substrates and for the assembly of novel supramolecular systems.References 1. A Bayer Chem. B5 25; 280 (1872) 2. A Zinke, E Ziegler Chem. Ber. 77 264 (1944) 3. J B Niederl, H J Vogel J. Am. Chem. Soc. 62 2512 (1940) 4. C D Gutsche, R J Muthukrishnan J. Org. Chem. 43 4905 (1978) 5. J Vicent, V Bohmer (Eds) Calixarenes. A Versatile Class of Macro- cyclic Compounds (Dordrecht, Netherlands: Kluver Academic, 1991) 6.V Bohmer Angew. Chem., Int. Ed. Engl. 34 713 (1995) 7. A Ikeda, S Shinkai Chem. Rev. 97 1713 (1997) 8. A Casnati Gazz. Chim. Ital. 127 637 (1997) 9. D V Khasnis,M Lattman, C D Gutsche J. Am. Chem. Soc. 112 9422 (1990) 10. D V Khasnis, J M Burton, M Lattman, H Zhang J. Chem. Soc., Chem. Commun. 562 (1991) 11. D V Khasnis, J M Burton, J D McNeil, H Zhang,M Lattman Phosphorus Sulfur Silicon Relat.Elem. 75 253 (1993) 12. D V Khasnis, J M Burton, J D McNeil, H Zhang,M Lattman Phosphorus Sulfur Silicon Relat. Elem. 87 93 (1994) 13. D V Khasnis, J M Burton, J D McNeil, H Zhang, C J Santini, M Lattman Inorg. Chem. 33 2657 (1994) 14. A Shevchenko, H Zhang,M Lattman Inorg. Chem. 34 5405 (1995) 15. D Matt,C Loeber, J Vicens, Z Asfari J.Chem. Soc., Chem. Commun. 604 (1993) 16. C Loeber, D Matt, P Briard, D Grandiean J. Chem. Soc., Dalton Trans. 513 (1996) 17. B R Cameron, F C J M van Veggel, D N Reinhoudt J. Org. Chem. 60 2802 (1995) 18. P Faidherbe, C Wieser, D Matt, A Harriman, A De Cian, J Fischer Eur. J. Inorg. Chem. 451 (1998) 19. J K Moran, D M Roundhill Inorg. Chem. 31 4213 (1992) 20.I Neda, H-J Plinta, R Sonnenburg, A Fischer, P G Jones, R Schmutzler Chem. Ber. 128 267 (1995) 21. I Neda, H-J Plinta, A Fischer, P G Jones, R Schmutzler Phosphorus Sulfur Silicon Relat. Elem. 109 ± 110 113 (1996) 22. C Floriani, D Jacoby, A Chiesi-Villa, C Guastini Angew. Chem., Int. Ed. Engl. 28 1376 (1989) 23. D Jacoby, C Floriani, A Chiesi-Villa, C Rizzoli J. Chem.Soc., Dal- ton. Trans. 813 (1993) 24. D M Roundhill, E Georgiev, A Yordanov J. Incl. Phen. Mol. Recogn. 19 101 (1994) 25. M Stolmar, C Floriani, A Chiesi-Villa, C Rizzoli Inorg. Chem. 36 1694 (1997) 26. O Aleksiuk, F Grynszpan, S E Biali J. Incl. Phen. Mol. Recogn. 19 237 (1994) 27. F J Parlevliet, A Olivier, W G J de Lange, P C J Kamer, H Kooijman, A L Spek, P W N M van Leeuwen J.Chem. Soc., Chem. Commun. 583 (1996) 28. I Bauer,W D Habicher, C Rautenberg, S Al-Malaika Polym. Degrad. Stab. 48 427 (1995) 29. A G C Hogberg J. Org. Chem. 45 4498 (1980) 30. L Abis, E Dalcanale, A DuVosel, S Spera J. Org. Chem. 53 5475 (1988) 31. E E Nifant'ev, M K Grachev Uspekhi Khimii 63 602 (1994) [Russ. Chem. Rev. 63 575 (1994)] 32. E E Nifant'ev, V I Maslennikova, L K Vasyanina, E V Panina Zh.Obshch. Khim. 64 154 (1994) a 33. E E Nifant'ev, V I Maslennikova, L K Vasyanina, E V Panina, A R Bekker, K A Lisenko,M Yu Antipin, Yu Struchkov Mende- leev Commun. 131 (1995) 34. A R Burilov, I L Nikolaeva, R D Galimov, M A Pudovik, V S Reznik Zh. Obshch. Khim. 65 1745 (1995) a 35. V I Maslennikova, E V Panina, A R Bekker, L K Vasyanina, E E Nifant'ev Phosphorus Sulfur Silicon Relat.Elem. 113 219 (1996) 36. V I Maslennikova, E V Shkarina, L K Vasyanina, K A Lysenko, M Yu Antipin, E E Nifant'ev Zh. Obshch. Khim. 68 379 (1998) a 37. A R Burilov, I L Nikolaeva, T B Makeeva, M A Pudovik, V S Reznik, L A Kudryavtseva, A I Konovalov Zh. Obshch. Khim. 67 875 (1997) a HOP(OEt)2 R2CO OH But But O HO O But But (CH2)2NH (EtO)2PO R R 102a,b (CH2)2NH R R OP(OEt)2 OH But But O HO O But But (CH2)2NH2 (CH2)2NH2 R=Me (a), R2=(CH2)5 (b). OPrn CH2NH2 CH2NH2 OPrn OPrn OPrn HOP(OEt)2 R2CO OPrn CH2NH OPrn OPrn OPrn R R 103a,b OP(OEt)2 HNCH2 R R OP(OEt)2 920 I S Antipin, E Kh Kazakova,WD Habicher, A I Konovalov38.M Fan, H Zhang, M Lattman J. Chem. Soc., Chem.Commun. 99 (1998) 39. A I Konovalov, V S Reznik, M A Pudovik, E Kh Kazakova, A R Burilov, I L Nikolaeva, N A Makarova, G R Davletschina, L V Ermolaeva, R D Galimov, A R Mustafina Phosphorus Sulfur Silicon Relat.Elem. 123 277 (1997) 40. E Kh Kazakova, N A Makarova, V V Zotkina, A R Burilov, M A Pudovik, A I Konovalov Mendeleev Commun. 157 (1996) 41. A R Burilov, I L Nikolaeva, R D Galimov, M A Pudovik, V S Reznik Zh. Obshch. Khim. 66 689 (1996) a 42.W Xu, J P Rourke, J J Vittal, R J Puddephatt Inorg. Chem. 34 323 (1995) 43. W Xu, J J Vittal, R J Puddephatt J. Am. Chem. Soc. 115 6456 (1993) 44. W Xu, J P Rourke, J J Vittal, R J Puddephatt J. Chem. Soc., Chem. Commun. 145 (1993) 45. A Vollbrecht, I Neda, H Thonnessen, P G Jones, R K Harris, L A Crowe, R Schmutzler Chem. Ber. 130 1715 (1997) 46. A Vollbrecht, I Neda, R Schmutzler Phosphorus Sulfur Silicon Relat.Elem. 107 173 (1995) 47. Y Ting,W Verboom, L C Groenen, J D Loon , D N Reinhoudt J. Chem. Soc., Chem. Commun. 1432 (1990) 48. L N Markovskii , V I Kal'chenko, N A Parkhomenko Zh. Obshch. Khim. 60 2811 (1990) a 49. V I Kal'chenko, Ya Lipkovskii, Yu A Simonov,M A Vysotskii, K Suvinska, A A Dvorkin, V V Pirozhenko, I F Tsymbal, L N Markovskii Zh.Obshch. Khim. 65 1311 (1995) a 50. Z Goren, S E Biali J. Chem. Soc., Perkin Trans. 1 1484 (1990) 51. F Grynszpan, Z Goren, S E Biali J. Org. Chem. 56 532 (1991) 52. F Grynszpan, N Dinoor, S E Biali Tetrahedron Lett. 32 1909 (1991) 53. F Grynszpan, S E Biali Tetrahedron Lett. 32 5155 (1991) 54. J Lipkovsky,M Visotsky, J Slowikovska, V I Kalchenko, L N Markovsky J. Phys. Org. Chem. 11 63 (1998) 55. J E McMurry, J C Phelan Tetrahedron Lett. 41 5655 (1991) 56. L T Birne, J M Harrowfield, D C R Hockless, B J Peachey, B W Skelton, A H White Aust. J. Chem. 46 1673 (1993) 57. J M Harrowfield, M Mocerino, B J Peachey, B W Skelton, A H White J. Chem. Soc., Dalton. Trans. 1687 (1996) 58. O Aleksiuk, F Grynspan, S E Biali J. Chem. Soc., Chem. Commun. 11 (1993) 59. F Grynspan,O Aleksiuk, S E Biali J.Org. Chem. 59 2070 (1994) 60. L Markovsky,M A Visotsky, V V Pirozhenko, V I Kalchenko, Y A Simonov J. Chem Soc., Chem. Commun. 69 (1996) 61. R G Janssen,W Verboom, S Harkema, G J van Hummel, D N Reinhoudt, A Pochini, R Ungaro, P Prados, J de Mendoza J. Chem. Soc., Chem. Commun. 506 (1993) 62. R G Janssen, J P M van Duynhoven,W Verboom, G J van Hummel, S Harkema, D N Reinhoudt J.Am. Chem. Soc. 118 3666 (1996) 63. J-B DeVais, S Pellet-Rostaing, R Lamartine Tetrahedron Lett. 35 8147 (1994) 64. L Markovsky, V I Kalchenko, M A Visotsky, V V Pirozhenko, Y A Simonov, A A Dvorkin, A V Latsenko, J Lipkovski Supramol. Chem. 8 85 (1997) 65. W Wryblevsky, Z Brzozka, R G Janssen,W Verboom, D N Reinhoudt New J. Chem. 20 419 (1996) 66. J P M van Duynhoven, R G Janssen,W Verboom, S M Franken, A Casnati, A Pochini, R Ungaro, J de Mendoza, P M Nietu, P Prados, D N Reinhoudt J. Am.Chem. Soc. 116 5814 (1994) 67. V I Kal'chenko, M A Vysotskii V V Pirozhenko, A N Shivanyuk, L N Markovskii Zh. Obshch. Khim. 64 1560 (1994) a 68. S Akabori, H Itabashi, H Shimura,MInoue J. Chem. Soc., Chem. Commun. 2137 (1997) 69. F Grynszpan, O Aleksiuk, S E Biali J.Chem. Soc., Chem. Commun. 13 (1993) 70. J Glode, I Keitel, B Costisella, A Kunath, M Schneider Phosphorus Sulfur Silicon Relat. Elem. 117 67 (1996) 71. J Glode, B Costisella,M Ramm,R Bienert Phosphorus Sulfur Silicon Relat. Elem., 84 217 (1993) 72. J Glode, I Keitel Phosphorus Sulfur Silicon Relat. Elem. 104 104 (1995) 73. H Thonnessen, P G Jones, R Schmutzler, J Gloede Acta Cryst., Sect.C 53 1310 (1997) 74. B Costisella, J Glode Phosphorus Sulfur Silicon Relat. Elem. 89 39 (1994) 75. B Costisella, J Glode Phosphorus Sulfur Silicon Relat. Elem. 108 103 (1996) 76. J K Moran, D M Roundhill Phosphorus Sulfur Silicon Relat. Elem. 71 7 (1992) 77. V I Kal'chenko, D M Rudkevich, L N Markovskii Zh. Obshch. Khim. 60 2813 (1990) a 78. L Markovsky, V I Kal chenko, D M Rudkevich, A N Shevchuk Mendeleev Commun. 106 (1992) 79. V I Kal'chenko, D M Rudkevich, A N Shevalyuk, I F Tsymbal, V V Pirozhenko L N Markovskii Zh. Obshch. Khim. 64 731 (1994) a 80. V I Kalchenko, A V Solov'ov, N R Gladin, A N Shivanyuk, L I Atamas, V V Pirozhenko, L N Markovsky, J Lipkovsky Supramol. Chem. 8 269 (1997) 81. V I Kal'chenko,A N Shivanyuk, V V Pirozhenko, L N Markovskii Zh.Obshch. Khim. 64 1562 (1994) a 82. A N Shivanyuk, V I Kal'chenko, V V Pirozhenko, L N Markovskii Zh. Obshch. Khim. 64 1558 (1994) a 83. V I Kal'chenko, L I Atamas', Yu A Serguchev, L N Markovskii Zh. Obshch. Khim. 54 1754 (1984) a 84. E Kh Kazakova, G R Davletshina, A I Konovalov Zh. Obshch. Khim. 66 407 (1996) a 85. A R Burilov, I L Nikolaeva, T B Makeeva, M A Pudovik, V S Reznik, A I Konovalov Zh.Obshch. Khim. 67 870 (1997) a 86. T Lippmann, E Dalcanale,G Mann Tetrahedron Lett. 35 1685 (1994) 87. T Lippmann, H Wilde, E Dalcanale, L Mavilla, G Mann, U Heyer, S Spera J. Org. Chem. 60 235 (1995) 88. P Delangle, J-P Dutasta Tetrahedron Lett. 36 9325 (1995) 89. Y Koide, H Terasaki, H Sato, H Shosenji, K Yamada Bull. Chem. Soc. Jpn. 69 785 (1996) 90. T H Webb, C S Wilcox Chem. Soc. Rev. 383 (1993) 91. C Seel, A Galan, J de Mendoza Top. Curr. Chem. 175 101 (1995) 92. F Hamada,T Fukugaki , K Murai , G W Orr, J L Atwood J. Incl. Phen. Mol. Recogn. 10 57 (1991) 93. V I Kal'chenko, L I Atamas', V V Pirozhenko, L N Markovskii Zh. Obshch. Khim. 62 2623 (1992) a 94. S Ozegowski, B Costisella, J Glode Phosphorus Sulfur Silicon Relat. Elem. 111 209 (1996) 95. I S Ryzhkina, L A Kudryavtseva, E Kh Kazakova, A R Burilov Mendeleev Commun. 88 (1997) 96. I S Ryzhkina, L A Kudryavtseva, A R Burilov, E Kh Kazakova, A I Konovalov Izv. Akad. Nauk, Ser. Khim. 275 (1998) b 97. V I Kal'chenko, L I Atamas, A V Solovyov, O V Klimchuk, L N Markovsky, in Proceedings of the 4th International Conference on Calixarenes (Abstracts of Reports), Parma, 1997 p. 142 98. C Loeber, D Matt, A De Cian, J Fischer J. Organomet. Chem. 475 297 (1994) 99. C Loeber, C Wieser,D Matt, A De Cian, J Fischer, L Toupet Bull. Soc. Chim. Fr. 132 166 (1995) 100. J Garnon, C Loeber, D Matt, P D Harvey Inorg. Chim. Acta 242 137 (1996) 101. C Wieser, D Matt, L Toupet, H Bourgeois, J-P Kintzinger J. Chem. Soc, Dalton Trans. 4041 (1996) 102. C Wieser, D Matt, J Fischer, A Harriman J. Chem. Soc., Dalton Trans. 2391 (1997) 103. C Wieser, C Dieleman, D Matt Coord. Chem. Rev. 93 165 (1997) 104. C Wieser, D Matt Platinum Metals Rev. 42 11 (1998) 105. C Dieleman, C Loeber, D Matt, A Decian, J Fischer J. Chem. Soc., Dalton Trans. 3097 (1995) 106. C Dieleman, D Matt, A Decian, J Fischer J. Orgnomet. Chem. 545 ± 546 461 (1997) 107. M R Yaftian, M Burgard, D Matt, C Wieser, C Dieleman J. Incl. Phenom. Mol. Recogn. 27 127 (1997) 108. M R Yaftian, M Burgard, A El Bachiri, D Matt, C Wieser, C Dieleman J. Incl. Phenom. Mol.Recogn. 29 137 (1997) 109. M R Yaftian, M Burgard, C Dieleman, D Matt Membr. Sci. 144 57 (1998) 110. C Dieleman, D Matt, I Neda, R Schmutzler, H Thonnessen, P J Jones, A Harriman J. Chem. Soc., Dalton Trans. 2115 (1998) 111. M A McKervey, P O Hagan, N Tompson, A Walker, F Arnaud-Neu, O Mauprivez, M-J Schwing-Weill, J F Dozol, H Rouquette, N Simon J. Chem. Soc., Chem. Commun. 2151 (1995) Phosphorus-containing calixarenes 921112. T Grady, S Maskula, D Diamond, D J Marrs,M A McKervey, P O'Hagan Anal. Proc. 32 471 (1995) 113. T McKittrick, D Diamond, D J Marrs, P O'Hagan, M A McKervey Talanta 43 1145 (1996) 114. M-J Schwing-Weill, F Arnaud-Neu Gazz. Chim. Ital. 127 687 (1997) 115. S O'Neil, P Kane, D Diamont Anal. Commun. 35 127 (1998) 116. F Arnaud-Neu, V Bo È hmer, J F Dozol, G GruÈ tner , R A Jacobi, D Kraft, O Mauprivez, H Rouquette, J Schwing-Weill, N Simon, W Vogt J. Chem. Soc., Perkin Trans. 2 1175 (1996) 117. H Boerrigter, W Verboom, D N Reinhoudt J. Org. Chem. 62 7148 (1997) 118. H Boerrigter, W Verboom, D N Reinhoudt Liebigs Ann. / Recueil 2247 (1997) 119. I S Antipin, I I Stoikov, N A Fitseva, E M Pinkhassik, I Stibor, A I Konovalov Tetrahedron Lett. 38 5865 (1997) 120. I S Antipin, I I Stoikov, A R Garifzyanov, A I Konovalov Dokl. Akad. Nauk 347 626 (1996) c 121. I S Antipin, I I Stoikov, A R Garifzyanov, A I Konovalov Zh. Obshch. Khim. 66 402 (1996) a 122. T Arimura, T Nagasaki, S Shinkai, T Matsuda. J. Org. Chem. 54 3766 (1989) 123. T Nagasaki, T Arimura, S Shinkai Bull. Chem. Soc. Jpn. 64 2575 (1991) a�Russ. J. Gen. Chem. (Engl. Transl.) b�Russ. Chem. Bull. (Engl. Transl.) c�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) 922 I S Antipin, E Kh Kazakova,WD Habicher, A I Kon

ISSN:0036-021X    

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Abstract. Data on the synthesis of b,b-disubstituted ketene N,N-acetals as well as their physicochemical properties and reac- tions with electrophilic and nucleophilic reagents are generalised. Special attention is paid to the synthesis of various heterocyclic compounds based on these compounds. The bibliography includes 182 references. I. Introduction Ketene S,S-, S,N- and N,N-acetals { 1 ± 3 draw the chemists' attention due to their high reactivity and the prospects of their broad and versatile synthetic utilisation, in particular, for the synthesis of heterocyclic compounds.Compounds 1 have been considered in comprehensive reviews.1 ±4 These reviews cover the properties and transforma- tions of N,S-acetals 2 to a considerably smaller extent. Ketene N,N-acetals 3 have been discussed there rather incompletely.The results of the use of keteneN,N-acetals for the synthesis of heterocyclic compounds are generalised in the review;5 however, it mostly discusses monosubstituted ketene N,N-acetals 3 (Y = H). On the other hand, a considerable number of studies have been devoted to the synthesis and transformations of disubstituted ketenes 3 in which X and Y are functional, mostly electron- acceptor substituents.The present review considers the methods for the synthesis of ketene N,N-acetals, their physicochemical and chemical properties and prospects of their use in the synthesis of various heterocyclic systems. II. Synthesis of disubstituted ketene N,N-acetals 1. Synthesis based on ketene S,S- andN,S-acetals and ketene dihalides The majority of publications concerning the synthesis of ketene N,N-acetals 3 deal with the reactions of substituted ketene S,S- (1) or N,S-acetals (2) with amines.6 ±39 It is well known that the alkyl(aryl, hetaryl)thio group (SR) is a `good leaving group', and its replacement by an amine-containing fragment is usually not problematic.However, replacement of one alkylthio group in ketene S,S-acetals 1 by the amino group results in a considerable decrease in the reactivity of the other SR-substituent due to the strong electron-donor effect of the amino group.Thus, it is possible to obtain asymmetrical N,N-acetals 3 in several stages from N,S-acetals 2. A typical example of such a sequential synthesis of ketene N,N-acetals is presented in Ref. 29. In some cases, ketene S,S-acetals were made to react with ethylenediamines.8, 10, 12, 13, 17, 23, 24, 28, 30, 33 ± 36, 39 This resulted in 2-methyleneimidazoline derivatives 4a.10 If substituted propane- 1,3-diamines are used instead of ethylenediamine deriva- tives,12, 28, 33 hydrogenated pyrimidine derivatives 4b are formed. Reactions of cyclic ketene S,S-acetal 5 with various diamines, including o-phenylenediamine derivatives (see also Refs 14, 17, 19, 24, 30, and 34) and 1,8-diaminonaphthalene, have been studied.28 Is was found that the rate of formation of cyclic systems 6 ± 8 depends on the nature of the diamine.The reaction of ketene S,S-acetal 5 with butane-1,4-diamine derivatives (n=4) or aro- matic amines occurs much more slowly than that with ethane(- propane)diamines (n=2, 3). 1 2 3 X Y RS RS a b X Y N RS a b X Y N N a b COR1 CN MeS MeS COR1 CN MeS R2HN COR1 CN R3HN R2HN R2NH2 R3NH2 4a,b X =COR0, CO2R0, CN, Ar; Y=COR0, CO2R0, CN; n = 2 (a), 3 (b). X Y MeS MeS (CH2)n NR Y X NR RHN(CH2)nNHR V A Makarov, V G Granik Russian Research Centre `Research Institute of Organic Intermediaries and Dyes', ul. Bol'shaya Sadovaya 1/4, 103787 Moscow, Russian Federation.Fax (7-095) 254 12 00. Tel. (7-095) 254 97 24 (V A Makarov), (7-095) 254 94 65 (V G Granik) Received 13 October 1997 Uspekhi Khimii 67 (11) 1013 ± 1031 (1998); translated by S S Veselyi UDC 547.233; 7 Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals V A Makarov, V G Granik Contents I. Introduction 923 II. Synthesis of disubstituted ketene N,N-acetals 923 III.Properties of ketene N,N-acetals 930 IV. Conclusion 938 { Other names for these compounds, such as ketene aminals, highly polarised or `push-pull' enamines, and enediamines are also widely used in the literature. Russian Chemical Reviews 67 (11) 923 ± 940 (1998) #1998 Russian Academy of Sciences and Turpion LtdThe reaction of S,S-acetals 1 or S,N-acetals 2 with various compounds containing the unsubstituted amino group opens up no less wide prospects for the synthesis of diverse derivatives of ketene N,N-acetals.For example, spiro compounds 9 incorporat- ing two enediamine fragments have been synthesised from ketene S,S-acetals and the tetramine 10.40 The reaction of S,N-acetal 11a with hydrazine afforded 3-amino-2-cyano-3-hydrazinoacrylamide.22 Reactions of substi- tuted acetals 11b with various amidines and guanidines result in the corresponding `push-pull' enamidines 12.20, 21 2-Bis(methylthio)methylenindane-1,3-dione reacts with tri- methylsilylmethylamine on refluxing in methanol to give a mix- ture of mono- and disubstituted products 13 and 14.32 Monoaziridino-R-amino derivatives of ketene N,N-acetals 15 are formed upon successive introduction of, first, monosubsti- tuted amines and then ethyleneimine in the reaction with ketene S,S-acetals 16.The action of the iodide ion on compounds 15 results in ketene acetals of the imidazolidine series 17.18 Direct reaction of ethyleneimine with compounds 16 gives bisaziridino derivatives of ketene N,N-acetals 18.18 Ketene S,S-acetals 5 can react even with such weakly basic compounds as carboxamides.This reaction gives N-acyl deriva- tives of ketene S,N-acetals 19 as the final products. The latter readily give N-acylketene N,N-acetals 20 on treatment with primary amines.29 Reaction of ketene S,S-acetals with anthranylamide 37 or glycinamide 16 results in the corresponding pyrimidinone (21) or imidazolidone (22) derivatives.The reaction of keteneN,S-acetals with pyrrolyl-1-magnesium bromide readily gives ketene N,N-acetals 23 containing the pyrrole ring as one of the amine moieties.31 It should be noted that ketene S,S-acetals in which the sulfur atoms are incorporated in the cyclic systems can also be converted to ketene N,N-acetals. For example, 1,3-dithiolane 24 reacts with N,N0-substituted ethylenediamines to give imidazolidine deriva- tives 25.8 6 7 8 O O MeS MeS O O Me Me 5 (CH2)n NR NR NR NR RHN(CH2)nNHR n=2± 4 o-(RHN)2C6H4 1,8-(RHN)2C10H6 NR NR + 10 X=CN, CO2Me. 9 MeS MeS X X NH NH X X HN HN X X NH2 NH2 H2N H2N 11a R=Me, Ph, Bn. H2N MeS CONH2 CN H2N H2NHN CONH2 CN N2H4 .H2O 11b 12 RHN N CO2Et CN R0 H2N RHN MeS CO2Et CN R0C(=NH)NH2 MeS MeS O O H2NCH2SiMe3 + 13 (51%) 14 (23%) Me3SiCH2NH Me3SiCH2NH Me3SiCH2NH MeS 16 MeS MeS CN X 15 H2NR NH KI Me2CO MeS RHN CN X N RHN CN X 7I7 7 ICH2CH2N RHN CN X 7 ICH2CH2NH X CN RN 17 CN X NH NR X=ArCO, HetCO. 16 18 N N CN X NH H MeS MeS RCONH 5 O O MeS MeS O O Me Me RCONH2 DMSO, 90 8C KOH 19 (50% ± 80%) 20 RCONH MeS RCONH R0HN R0NH2 AcOH, 50 ± 70 8C X, Y=CN, CO2Et. 22 21 NH NH X Y O NH Y X NH O o-NH2C6H4CONH2 H2NCH2CONH2 MeS MeS Y X X=CH2, O.+ THF N MgBr CN SMe N NC X CN N N NC X 23 R=Ar, ArAlk. 24 25 (CH2NHR)2 Me Me O O NR NR NaH, CS2 (BrCH2)2 S S 924 V A Makarov, V G GranikThe reaction of substituted 1,3-dithietanes 26 with o-phenyl- enediamine results in benzodiazepine derivatives 27, which are oxidised in alkaline medium to give ketene N,N-acetals of the benzoimidazole series 28.These compounds have also been obtained by independent synthesis from alkyl 2,2-bismethylthio- methylenecyanoacetate and o-phenylenediamine.19 Chlorination of ketene S,S-acetals in acetic acid readily gives ketene dichlorides, which represent convenient starting com- pounds for subsequent synthesis of ketene N,N-acetals. For example, acetal 29 was obtained by the reaction of the appropriate dichloride 30 with aniline.41 Similar reactions are also known for other ketene dichlorides, viz., 2-dichloromethyleneindane-1,3-dione 42 and 2-dichloro- methylene-4,5-dichlorocyclopent-4-ene-1,3-dione,43 as well as for some ketene dibromides 44 and ketene diiodides.45 2.Synthesis based on ketene O,O- and O,N-acetals and related compounds Ketene O,O-acetals and related compounds are promising start- ing reagents for the synthesis of ketene N,N-acetals.For example, bisalkoxymethylenemalonodinitriles 31 react with sodium cyana- mide under mild conditions with formation of the sodium salt 32, which givesN-cyanoketene N,N-acetal 33 on heating with methyl- amine.46 Dimethoxymethylenemalonodinitrile 31 (R=R0=Me) reacts similarly with ammonia and ethylenediamine.47 The reaction of the 1,3-dioxolane derivative 34 with various diamines results in cyclic ketene N,N-acetals, while the reaction of compound 34 with aromatic diamines requires more drastic conditions than that with aliphatic ones.48 Amino(aryloxy)methylenemalonodinitrile (35a) and ethyl amino(aryloxy)methylenecyanoacetate (35b) react with amines to give the corresponding enediamine derivatives 36a,b.49 3.Synthesis based on thiourea derivatives It is well known that ketene S,S- and S,N-acetals react with amines rather smoothly, since the alkylthio group is a `good leaving group' (see Section 1). Yet another group of reactions based on the replacement of the RS group, viz., the reactions of compounds containing the reactive methylene group with S-alkylisothioureas, also result in the corresponding `push-pull' enediamines.For instance, treatment of N,N0-dibenzyl-S-methylisothiourea with nitroacetonitrile in alkaline medium gave a small yield (*20%) of b-nitro-b-cyanoenediamine 37.50 Cyclic thiourea derivatives react similarly.51 ± 57 For example, the reaction of compound 38 with aroylacetonitriles results in cyclic acetals 39.Isothiouronium salts also react with compounds containing the reactive methylene group in the presence of sodium hydride.58 For example, the reaction of salt 40 with malonodinitrile results in the corresponding symmetric enediamine. Not only S-methyl derivatives of isothiouronium salts, but also compounds with more complex S-substituents can serve as starting compounds in the synthesis of ketene N,N-acetals.59 Treatment of salts 41 containing an electron-acceptor group at the sulfur atom with sodium hydride in tetrahydrofuran gives enediamines 42 in moderate yields.It is believed 59 that this reaction gives intermediate ylides, which then undergo cyclisation to thiiranes 43. Elimination of the sulfur atom from the thiirane ring results in compounds 42. 26 S NC RO2C S CO2R CN 27 (89%) 28 S RO2C H2N NH NH RO2C NC HN HN H2O2/OH7 NC RO2C SH N H2N H o-(NH2)2C6H4 28 MeS CN CO2R MeS o-(NH2)2C6H4 CO2Me MeS MeS CO2Me Cl2 AcOH 29 30 CO2Me Cl Cl CO2Me CO2Me PhHN PhHN CO2Me PhNH2 CN R0O RO CN MeNH2 EtOH NaNHCN EtOH Na+ 7 CN EtO N CN NC 31 32 R=R0 =Et; R, R0 =(CH2)2. 33 CN MeHN NCHN CN 34 CO2Me CN O O (75%) (45%) NH CN CO2Me NH MeO MeO NH NH CO2Me CN MeOH, D, 1 h H2N(CH2)3NH2 H2O, 0.5 h, 20 8C MeO MeO NH2 NH2 X=CN (a), CO2Et (b). 35a,b 36a,b ArO H2N X CN RR0N H2N X CN RR0NH, H2O D, 15min 37 MeS BnHN NBn BnHN BnHN CN NO2 NCCH2NO2 OH7 N NH Me COAr CN 39 N N Me SMe 38 ArCOCH2CN OH7 CH2(CN)2 NaH NMe CN CN NMe + I7 NMe SMe NMe 40 41 S Me2N Me2N R Ph + Br7 S Me2N Me2N RCH(Br)Ph Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 925In the same study,59 a method is proposed for the synthesis of ketene N,N-acetal 44 with the use of diethyl diazomalonate and tetramethylthiourea as the starting compounds.An interesting method for the synthesis of enediamine 45 is described in Ref 60. It was reported 61 that keteneN,N-acetals can also be obtained from thiourea derivatives without preliminary S-alkylation.For example, enediamine 46 is obtained by condensation of com- pound 47 with malonodinitrile. It is evident from the scheme shown above that the formation of the enamine moiety is accompanied by the pyrazole ring closure. 4. Synthesis based on `activated' amides and ureas It is known that `activated amides', i.e., amide acetals, imidates, imidoyl chlorides and related compounds, are rather promising starting reagents for the formation of various enamine systems.Enediamines, including highly polarised ones, are obtained by the condensation of `activated amides' with compounds having a reactive methylene group.62 ± 64 For example, imidoyl chloride 48 reacts readily with such compounds as ethyl cyanoacetate, malo- nodinitrile and dimedone in dimethyl sulfoxide in the presence of sodium hydride to give substituted ketene N,N-acetals.65 Chloroamidines 49 obtained by the action of phosgene on tetraalkylurea are converted to `push-pull' enamines 50a ± c in the reaction with ethyl malonate and ethyl cyanoacetate or malono- dinitrile.66 Derivatives of cyclic ureas, particularly imidates, also react readily with compounds that incorporate a reactive methylene unit.67 The urea acetals 51 can undergo condensation with ethyl cyanoacetate or malonodinitrile to afford ketene N,N-acetals 50a,c.68, 69 Orthoamidoesters 52 also enter into similar reactions to give aminals 53.70, 71 It should be noted that compounds 52 are much more reactive than urea acetals 51.The bisamidinium salt 54 obtained by the reaction of an urea derivative with trifluoromethanesulfonic acid was also used for the synthesis of `push-pull' enediamines 53.72 A modified procedure has been reported,73 in which salt 54 was initially treated with 4-dimethylaminopyridine, and its con- densation with malonodinitrile was carried out only thereafter.R=CO2R0, Bz. + S Me2N Me2N R Ph 7 NaH THF + S Me2N Me2N R Ph Br7 43 42 (30% ± 45%) Me2N R Ph Me2N S R Me2N Me2N Ph S CO2Et Me2N Me2N CO2Et + S Me2N Me2N CO2Et CO2Et 7 44 Me2N CO2Et CO2Et Me2N N2C(CO2Et)2 (Me2N)2CS Cu CO2Et CO2Et Cu 7N2 S NH2 Me2N + SMe N Me2N Me S SMe I7 CH2(CN)2 Et3N NaH, CS2 MeI NC CN N Me2N SMe SMe 45 46 Me N N EtO2C S 47 NHPh H CH2(CN)2 Et3N N NHPh NC NC N Me OH X=CN, CO2Et.ArCON N Cl 48 X ArCONH N CN N ArCONH Me Me O O O O Me Me XCH2CN X=Y=CN (a), CO2Et (b); X=CN, Y=CO2Et (c). 49 50a ± c (60% ± 70%) O R2N R2N Cl R2N R2N + Cl7 R2N R2N X Y COCl2 XCH2Y Et3N O2NCH2CO2R3 (CH2)n N OR1 NR2 (CH2)n NH CO2R3 NO2 NR2 X=CN, CO2Et. 51 50a,c R2N OR0 R2N OR0 R2N CN X R2N NCCH2X 52 53 Me2N X Y Me2N Me2N OR Me2N NMe2 XCH2Y X, Y=CN, CO2Et. 53 + + NMe2 NMe2 O Me2N Me2N 54 2CF3SO73 CF3SO3H XCH2Y O Me2N Me2N 54+ N NMe2 CH2(CN)2 Me2N CN CN Me2N 2CF3SO73 + Me2N Me2N N NMe2 + 926 V A Makarov, V G GranikBis(dialkylamino)malononitriles, like urea acetals, can undergo condensation with compounds containing the reactive methylene unit.For example, the reaction of bis(dimethyl-ami- no)malonodinitrile with alkyl malonate or acetoacetate in the presence of sodium hydride in acetonitrile gives enediamines 55.74 5.Synthesis based on various S-, N- and C-cyano derivatives A number of studies deal with the synthesis of enediamines under consideration by addition of the cyano group to the reactive methylene unit. For example, the reaction of organic thiocyanates with b-diketones in the presence of nickel acetylacetonate gives ketene S,N-acetals, which can react with amines (in particular with morpholine) to afford ketene N,N-acetals in high yields.75 Cyanamide and its derivatives react in a similar way.76 ± 80 For example, the reaction of cyanamide with ethyl acetoacetate in the presence of nickel acetylacetonate results in the aminal 56.Cyanamide reacts similarly with b-dicarbonyl compounds in the presence of SnCl4.81 Another way for the synthesis of ketene N,N-acetals is based on the reaction of tricyanomethane (cyanoform) and related compounds with amines.82, 83 Trichloroacetonitrile proved to be a convenient starting reagent in the synthesis of the target enamines. Its condensation with CH2-acids initially gives (a-trichloromethyl)enamines, in which the CCl3 group is readily replaced by the amino group with the formation of enediamines.84 ± 86 The reaction of trichloro- acetonitrile with benzoylacetonitrile followed by the reaction of the resulting compound 57 with 2-aminopyridine can be presented as an example.86 Similar replacement of the CCl3 and CF3 groups in enamines by the amino groups with the formation vicinal ethylenediamine derivatives has been reported.87, 88 Reactions of trichloromethyldialkylamines 58 with com- pounds containing reactive methylene groups also result in ketene N,N-acetals.89 For example, compounds 58 react with malonodi- nitrile to give chloroenamines 59.The latter react with aniline to afford enediamines 60. Finally, the reaction of tetracyanoethylene with primary amines (e.g., pentylamine) readily yields symmetrical enedi- amines. Secondary amines can also be used in this reaction.90, 91 6.Synthesis based on enamine derivatives a-Chloroenamines 92 and a-chloroenamidines 93 react with amines to give ketene N,N-acetals. For example, the reaction of chlor- oenamidine 61 with a secondary amine results in the product 62.93 A wide range of b,b-disubstituted ketene N,N-acetals have been synthesised by C-acylation of b-monosubstituted a,a-diami- noethylenes.94 ± 99 Cyanoacetic acid in the presence of Ac2O,94 acetic anhydride,95 isocyanates, isothiocyanates,96, 97 and benzoy- lisothiocyanate were used as acylating agents.98, 99 Scheme 1 It has been shown97 that acylation of enamines 63 with isocyanates occurs not only at the b-carbon atom but also at the amino group.b-Alkylation of b-monosubstituted ketene aminals with com- pounds containing reactive halogen has also been reported.100 X=CO2R0, COMe. +XCH2CO2R 55 Me2N CN Me2N CN X Me2N Me2N CO2R NaH, MeCN 7NaCN R2OC NH2 N R1OC O O R1 R2 O SR3 NH2 R1OC R2OC R3SCN Ni(acac)2 O HN H2NCN+ COMe CO2Et CO2Et H2N 56 H2N COMe Ni(acac)2 THF, D (NC)3C7K+ R2NH (NC)3C7H2N+R2 H2N CN CN R2N D BzCH2CN CCl3CN NaOEt, EtOH Cl3C Bz CN H2N 57 N H2N N NH H2N CN Bz 58 RR0NCSCl +(RR0NCS2)2 RR0NCCl3 PCl5 CH2(CN)2 59 60 PhNH2 RR0N CN CN Cl RR0N CN CN PhHN 2Me(CH2)4NH2 Me(CH2)4NH CN CN Me(CH2)4NH NC CN CN NC 62 CN N RR0N CHO Me2N CN N Cl CHO Me2N 61 RR0NH (a) Ac2O, R7R=(CH2)2, X = CO2Et; (b) R0NCY, R0 =Bz, R7R=(CH2)3, X=CSAr, Y=S; (c) R0NCY, R0 =Bz, X=NO2, Y=S; (d) R0NCY, R0 =Ph, X=CN, Y=O, S; (e) NCCH2CO2H, Ac2O, X =NO2.RHN RHN X RHN X COMe RHN RHN X CNHR0 RHN Y RHN X CCH2CN RHN O b or c or d e a 63 H2N NC N X (CH2)n RNCO X=CH2, n=1, 2; X = O, n=2.+ H2N NC N X (CH2)n CONHR RHNCN NC N X (CH2)n H O Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 927A series of ketene aminals have been synthesised by acylation of derivatives of 2-methylimidazoline,101 2-alkylbenzoimida- zoles 102 ± 104 and 2-methyltetrahydropyrimidines.105 Examples of such syntheses based on 2-alkylbenzoimidazoles are shown in the scheme below. 2-Methylbenzoimidazole also undergoes bisformylation according to Vilsmeier in the presence of phosphorus oxychlor- ide.106 The reaction of 1,3-dimethyl-2-methylenebenzoimidazoline (64) with compounds of general formula O=C=Y, where Y= O, S, NPh,107 or with 1,4-naphthoquinone108 results in ketene N,N-acetals 65 or 66a, respectively.Acylation of compound 64 with acyl halides or related halogen-containing compounds in triethylamine gives the enediamines 66b ± d.108 7. Synthesis based on pyrimidine derivatives One of the most peculiar preparative methods for the synthesis of `push-pull' enamines is that starting from pyrimidines with a strong electron-acceptor substituent at position 5 of the pyrimi- dine ring.In 1967, it was found109 that the action of organic or mineral acids on such pyrimidines results in easy opening of their ring. For example, the reaction of substituted pyrimidine 67 with acetic, chloroacetic, trifluoroacetic or hydrochloric acids results in the opening of the pyrimidine ring and formation of the enedi- amine 68.It was established 109 that if X is a tertiary amino group, the yields of enediamines 68 range from 50% to 95%. If X is a secondary amino group, the yields of compounds 68 are much lower. For example, if X=NHMe, the yield of the product is 18%, and if X=NH2, it becomes impossible to obtain the corresponding enamine at all. In the latter case, the halogen at the pyrimidine ring is replaced by the hydroxy group. The following mechanism of the pyrimidine ring opening has been proposed (Scheme 2).110 First, the nitrogen atom of the starting pyrimidine is protonated.The resulting pyrimidinium cation adds a molecule of water at position 2. Then elimination of the substituent from position 6 occurs with simultaneous opening of the pyrimidine ring at N(1)7C(2) bond.Subsequent N-defor- mylation of the resulting nitrile finally results in the enamine. Scheme 2 The effect of the extent of substitution of the amino group in 4-amino-6-chloro-5-nitropyrimidines on the processes that occur on their heating in acetic or trifluoroacetic acids has been studied in detail.111 This confirmed the conclusions 109 that ketene aminals 68 are formed in good yields if a tertiary amino group is present at position 4.The action of trifluoroacetic acid on pyrimidine derivatives containing a primary or secondary amino group at position 4 was monitored by 1H NMR spectroscopy.112 X NHMe NHEt NHPri NHBut NH2 Yield of com- 47 20 21 25 30 pound 68 (%) Yield of com- 53 80 79 75 70 pound 69 (%) The formation of `push-pull' enamines of the type 68 is also observed upon ring opening of 5-formyl-, 5-cyano- and 5-carbox- ypyrimidines.112, 113 However, compounds of this type tend to form pyrimidones 69 to a greater extent than nitropyrimidines.The mechanism of the pyrimidine ring opening shown in Scheme 2 has been confirmed.112 The authors managed to isolate and identify the N-formyl derivative 70.It was also found that prolonged heating of 5-formylpyrimidines in an acidic medium results in compounds 70, then 3,3-diamino-2-formylacrylonitrile derivatives 71 are formed, and finally their C-deformylation occurs to give cyanoenediamines 72.112 R=CO2Et, Ph; Hal=Cl, Br. HalCH2R NH COAr NH (CH2)n NH COAr CH2R NH (CH2)n NR N CH2X NH Bz Bz NH O HN NR COCO2Et COCO2Et NH NR CONHCO2Et X NH N Bz NBz OBz Ph ClCOCO2Et X=H,R=Et OCNCO2Et R=H X=CN, CO2Et BzCl, Et3N X, R=H POCl3 DMF OH7 NH N Me N N NMe2 N+Me2 N+Me2 2Cl7 NH CHO CHO NH (b) RCl, Et3N, R =MeCO (66b), Bz (66c), MeSO2 (66d).(a) RH, R= (66a); O O 64 NMe NMe CH2 O=C=Y PhH a or b 66a ± d NMe NMe R R 65 (Y=O, S, NPh) NMe NMe COYH COYH 67 N N Cl NO2 X 68 H2N NC NO2 X a or b or c, or d X=NMe2, N O; (a) AcOH; (b) ClCH2CO2H; (c) CF3CO2H; (d ) HCl.H+ H2O N N Cl NO2 NR2 H + N N Cl NO2 NR2 HO H H N N Cl NO2 NR2 1 2 3 4 5 6 7HCO2H H2N NC NO2 NR2 OHCHN NC NO2 NR2 + N N Cl NO2 X H2N NC NO2 X HN N O NO2 X CF3CO2H 69 68 67 928 V A Makarov, V G GranikIt was shown114 that treatment of 4-amino-6-chloro-5-for- mylpyrimidines with an alkali also results in ketene N,N-acetals. It is of note that heating of 4-chloro- or 4-methoxy-5-nitro-6- X-pyrimidines in an alkaline medium made it possible to obtain potassium cyanonitroacetamide 73.115 Ring opening of 4,6-dichloro-5-nitropyrimidine can be achieved on heating with secondary amines.However, treatment of chloronitroisothiazole with piperidine or morpholine gives b-cyano-b-nitroenamines 74 rather than enediamines.116 The bisketene aminal 75 was obtained according to the following scheme:117 The reaction of 4-chloro-5-nitro-6-R-pyrimidines with S-alkyl(aryl)isothiouronium salts on heating in an alkaline medium results in substituted pyrimidines 76.Subsequently, the products 76 were converted to `push-pull' enamines 77, which then underwent cyclisation into substituted sym-triazines.118 It has recently been shown119 that 4-arylamino-6-chloro-5- nitropyrimidines are transformed to enediamines 68 on boiling in dilute HCl to give considerable amounts of the corresponding 4-arylamino-5-nitropyrimidin-6-ones 69.A peculiar method for the synthesis of nitroenediamines containing a tetrazole moiety at the b-position is based on the reaction of 4-amino-6-chloro-5-nitropyrimidine derivatives with sodium azide.120 Bromination of methylenedihydropyrimidine 78 gives a mix- ture of cyclic enamines.121 8.Miscellaneous methods for the synthesis of ketene aminals The reaction of triethylammonium (2-chloro-1,3-dioxoindan-2- yl)dichloromethanesulfonate (79) with various amines results in ketene N,N-acetals 80 ± 82.122 In certain studies, carbodiimides have been used for the synthesis of ketene aminals.This reaction was first described in 1899 with the synthesis of enediamine 83 from N,N0-diphenylcar- bodiimide and diethyl malonate as an example.123 Compound 83 can also be obtained by the reaction of N,N0- diphenylthiourea with sodium diethyl malonate.124 Dicyclohexylcarbodiimide and other carbodiimides react with Meldrum's acid to give the corresponding enediamines.125, 126 N N Cl CHO NR2 OHCHN NC CHO NR2 H2N NC CHO NR2 H+ 70 71 72 H2N NC NR2 X=SC(S)NEt2, Y=OMe; X=Y=Cl.N N X NO2 Y KOH 73 OK NO2 NC H2N N S NO2 Cl N X N S NO2 Cl H 7HCl 7S + X HN X=CH2, O. NC H N O2N X 74 N N N N N N Cl NO2 O2N Cl AcOH H2O N N H2N 75 NH2 NC NO2 CN O2N R=Cl, OMe; X=Alk, Ar. + N N R NO2 Cl + NH2 H2N XS HSO74 N N R NO2 NH SX HN 76 OH7 N N NH NO2 NC SX NO2 NH SX HN NC N OHC H 77 N N Cl NO2 NR2 NaN3 MeOH H NO2 NR2 H2N N N N N 7 NO2 NR2 H2N N N N N Na+ N NH CO2Et CN 78 Br2, EtOH CHCl3, 20 8C + NH NH CO2Et CN Br EtO EtO NH NH CO2Et CN Br HO EtO + NH NH CN CO2Et Br EtO HO + 79 O O CCl2SO73 Cl Et3N+H O O HN HN 80 R 81 O O NH(CH2)2OH NH(CH2)2OH O O HN(CH2)2S HN(CH2)2S 82 H2N(CH2)2OH H2N(CH2)2SH NH2 NH2 R PhN C NPh PhHN CO2Et CO2Et PhHN 83 CH2(CO2Et)2 Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 929Vinyl azides have also been used as the starting compounds in the synthesis of ketene N,N-acetals.127, 128 For example, the reaction of vinyl diazide 84 with diaminopropane gave a small yield of cyclic product 85.The reaction of ketene N,S-acetals with sodium azide prob- ably gives vinyl azides 86 as intermediates, which are then converted into heterocyclic ketene aminals 87.129 The reaction of 2-diazo-1,3-diketones with tetraaminoethy- lene derivatives results mainly in enediamino diketones 88.130 The synthesis of four-membered heterocyclic compounds, viz., diazetidine derivatives incorporating the enediamine moiety, has been reported.131 For example, condensation of N-(thioacyl)- carbamic acid derivative 89 with para-substituted anilines yields cyclic products 90.Heating of 1-hydroxy-5-methyl-1,2,3,4,4a,5-hexahydro-1,4a- propanopyrido[1,2-a]benzoimidazole (91) with acetic anhydride in toluene results eventually in ketene aminal of the benzoimida- zole series 92.132 This process is believed 132 to occur according to the following scheme: III.Properties of ketene N,N-acetals 1. Physicochemical parameters The major characteristics of ketene N,N-acetals is the direct polar conjugation of two a-amino groups with b,b-electron-acceptor substituents. Naturally, this conjugation affects the spectral and other physicochemical properties of ketene N,N-acetals. For example, according to X-ray diffraction data, abrupt elongation of theC=Cbond in b,b-dinitroenediamines is observed, while the N7C bonds are much shortened.This can be rationalised as the superposition of the following resonance structures: The lengths of the C=C bonds in b,b-dinitroenediamines lie within 1.451 ± 1.474 A, those of the N7C bonds lie within 1.315 ± 1.316 A and those of the C7NO2 bonds lie within 1.399 ± 1.494 A.45 It is known133 that the barrier to cis ë trans- isomerisation in b,b-diacyl N,N-acetals 93 (`rotation' around the carbon ± carbon double bond) is extremely low and is DG#<8 kcal mol71 (to compare with enamines, see Refs 134 ± 136).If weaker electron-acceptor substituents are present in the molecule (for example, in compounds 94 and 95), the rotation barrier increases to 15 and 21 kcal mol71, respectively.10 The same trend is observed if the electron-donor effect of the a-sub- stituents decreases.For example, for ketene S,S-acetal 96, DG#>25 kcal mol71. It is essential that in the cases where the formation of intramolecular hydrogen bonds in diacyl ketene aminals is possible, cis ë trans-isomerisation is slowed down con- siderably. For example, the 1H NMR spectrum of compound 97 (in CDCl3) contains a double set of signals corresponding to the E- and Z-isomers, whereas one series of signals is observed in a solvent that favours the scission of intramolecular hydrogen bonds (DMSO-d6).8, 77, 78 RNCNR O O O O Me Me O O O O Me Me RHN RHN N3 CO2Me CN N3 NH NH CO2Me CN H2N(CH2)3NH2 85 (14%) 84 86 87 PhHN R CN MeS N N NPh NH R CN PhHN R CN N3 NaN3 DMSO O7 O Ná2 + Me Me N N N N Ph Ph Ph Ph (6.5%) 88 (37%) + O O PhN PhN Me Me O O N N PhN PhN Me Me 89 90 p-RC6H4NH2 EtOH, 7H2S CO2Et CO2Et N NH O C6H4R-p HO2CNH CO2Et CO2Et S 91 N N OH Me H+ Ac2O N N COMe O Me 7H2O N N A Me OH Me MeHN N COMe O O A= . 92 NA CH2 NMe NA NMe COMe COMe Ac2O NA NMe Me O Me O + 7 7 etc. N NO2 NO2 N N+ NO2 NO2 N N+ NO2 N N O7 O 93 94 95 96 Me2N COMe COMe Me2N Me2N C6H4NO2-p CN Me2N MeS C6H4NO2-p CN MeS Me2N Ph CN Me2N O H N N H O O H Ph R R0 97 98 NR COMe COMe NR (CH2)n 930 V A Makarov, V G GranikIn cyclic enediamines 98, the rotation barriers are also very low and are*12 ± 13 kcal mol71 (see Refs 137, 138). It is of note that replacement of the C=O group in the b-substituent by the C=S group increases the DG# to *30 kcal mol71.This effect is related to the stabilisation of the ground state and destabilisation of the transition state.139 IR spectra of `push-pull' enediamines containing the carbonyl group are their important characteristics, the absorption bands of the CO groups are observed in the region below 1650 cm71. This indicates strong conjugation of the carbonyl groups with the electron-donor, substituted a-amino groups.59, 77, 126, 140, 141 13C NMR spectra of this type of compounds are also rather characteristic.28, 78, 142 For example, the signal for the Cb atom in N-benzoylenediamino diketones 97 is in the region d *87 ± 100 ppm, while the signal for the Ca atom is at d *161 ± 163 ppm.These chemical shifts indicate that the electron pairs of the nitrogen atoms participate in the `enamine conjugation'.A detailed study of 1H and 13C NMR spectra carried out for nitrocyanoenediamines 99143 showed that the stereochemistry of these compounds is determined by the possibility of formation of intramolecular hydrogen bonds between the amino group and the nitro group. 13C NMR spectra of compounds 99 are also characterised by an upfield shift of signals for the Cb atom (d *90 ± 103 ppm), while the signals for the Ca atoms are in the region d 157 ± 170 ppm.Specific conformation features of cyclic `push-pull' ethylenes with sterically hindered electron-donor and electron- acceptor groups were studied.144, 145 It was shown that the electron-donor and electron-acceptor fragments in these com- pounds are turned relative to the central carbon ± carbon double bond.The tautomerism in ketene aminals and related compounds was studied.102, 146, 147 It was shown that compound 100 exists predominantly in the tautomeric form 100a. The position of tautomeric equilibrium depends on the solvent. For example, the equilibrium 101a>101b in CHCl3, dioxane, methanol and water is shifted towards the enamine 101a, whereas inDMSOandDMFit is shifted towards the enol 101b.In acetonitrile, a mixture of tautomers 101a,b is formed.146 It was shown79 that diacyl ketene aminals represent bases and are protonated on the oxygen atom. However, if the amino group has a strong electron-acceptor substituent, these compounds are strong organic acids. For example, the ionisation constants (pKa) of N-cyano-b,b-dicyano- ketene aminals 102 in water are *2.6 ± 2.8, i.e., they are even stronger acids than formic acid (pKa 3.75) Finally, the pronounced effect of direct polar conjugation in `push-pull' enamines is also characterised by their high dipole moments.58 For example, the dipole moments of imidazoline or pyrimidine derivatives of ketene N,N-acetals are*8 D, while it is only 6.5 D for such a strongly polarised compound as N-phenyl- sydnone.148 2.Reactions of b,b-disubstituted ketene N,N-acetals with electrophilic and nucleophilic reagents The presence of two a,a-amino groups in ketene aminals increases the nucleophilicity of the Cb-position considerably, and hence increases the possibility of attack on this position by electrophilic reagents.For example, it is known149 that the reaction of ketene aminals 103 with tert-butyl hypochlorite smoothly gives the corresponding chloroamidines 104 in high yields despite the presence of two substituents at the Cb-atom. Nitration of `push-pull' enediamines under various conditions has been studied.150 It was found that the use of a mixture of nitric and sulfuric acids resulted in Cb-nitration and N-nitrosation, while in a mixture of HNO3 and trifluoroacetic anhydride, Cb- and N-nitration occurred. In the latter case, the order of nitration steps on theCb andNatoms has not been determined.Nitration of the ketene aminal 105 under various conditions results in products 106 and 107.150 It is essential that compound 107 is also formed upon nitration of b-monosubstituted `push-pull' enediamine 108 with an HNO3 ± trifluoroacetic anhydride mixture.etc. N+ X Y N 7 N X Y N N+ X7 Y N Alk N NC N H O N Alk Alk O a b 99 100a 100b NEt N O H O EtO2C EtO2C NEt COCO2Et COCO2Et NH 101b HN N NH RO2C CO2R CO2R RO2C O H O RO RO HN NH NH CO2R CO2R RO2C CO2R 101a CO2R RO2C COMe H2N H2N+ Me HO COMe H2N H2N COMe H+ X NH2 NHMe NMe2 pKa 2.83 2.70 2.63 NCHN CN CN X 102 NMe CN n 2 3 m (benzene), D 7.93 8.02 CN NMe (CH2)n R1=Ph, Me, Pri; R2, R3=CN, CO2Et. 103 104 (76% ± 97%) R3 R1HN Me2N R2 ButOCl R1N Me2N R2 Cl R3 107 106 NH NO2 NO2 NH 105 N NO2 N NO2 NO2 NO2 N NO2 N NO NO2 NO2 HNO3 H2SO4 (CF3CO)2O HNO3 Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 931A method for reductive elimination of one NO2 group from the Cb atom in compound 107 upon treatment with potassium iodide has been described.151 The reaction of b,b-disubstituted ketene N,N-acetals with acyl halides results in N-acylation products.88 If dimethyltrichlorome- thylamine is used as the acylating reagent, the reaction occurs at the nitrogen atom to give the derivative 109 containing the chloroformamidine moiety. Alkylation of ketene aminals is an important reaction that characterises enamines and deserves separate discussion.For example, the alkylation of thioamides 110 with methyl iodide occurs mostly on the sulfur atom. In addition to S-methyl derivatives 111, N,S-dimethyl derivatives 112 are formed.152 Oxidation of aminals 111 with mercuric acetate in acidic medium results in enediamines 113.152 It should be noted that the oxidation of the ammonium salts of cyano(ethoxycarbonyl)thiocarbamoylmethanes with mercury(II) oxide readily gives ketene aminals.153 Enediamino ketone 114 undergoes alkylation with triethy- loxonium tetrafluoroborate or dimethyl sulfate on the oxygen atom to give the corresponding complex 115, which, in turn, can react with alkoxide anions to produce alkoxydienediamine 116.Compound 116 is a convenient starting reagent for the synthesis of various dienamines, e.g., dienediamine 117 or dienetriamine 118.70 Yet another reaction of ketene aminals with electrophilic reagents worthy of note is that of enediamino diketones with the Lawesson reagent (LR). For example, enediamino diketone 119 can be converted to the corresponding enediamino dithione 120.8 A series of reactions of ketene aminals with nucleophilic reagents has been described.125 For example, acid hydrolysis of a Meldrum's acid derivative 121 resulting in the b-monosubstituted derivative 122 is worth mentioning. The same authors described degradation of compound 121 upon treatment with ethanolic alkali.Monodeacylation of enediamino diketones has been described.12 For example, the action of the hydroxy or methoxide anion on diketones 123 results in ketones 124.A possible version of the mechanism of this process includes the addition of the nucleophile at the carbonyl group.12 A version involving the attack of the nucleophile on the a-position may not be ruled out, either. 107 HNO3 (CF3CO)2O KI N NO2 NO2 N NO2 K+ 7 NH H NO2 NH 108 H+ NH NO2 NO2 N NO2 (a) BzCl, R1=H, R2, R3=CN, CO2R4; (b) CCl3NMe2, R1=Me, R2=R3=CN.R3 R12 N H2N R2 a b BzHN R2 R12 N R3 109 CN R12 N N CN Cl Me2N MeI, Ag2O DMF RHN NO2 RHN S NHCOR0 110 + RMeN NO2 RHN MeS NCOR0 112 (4% ± 10%) RHN NO2 RHN MeS NCOR0 111 111 H+ Hg(OAc)2 RHN NO2 RHN R0CON S+ AcO7 HgOAc Me 113 RHN NO2 RHN CN HgO H3N+R0 (50% ± 92%) RHN CO2Et CN R0HN RHN CO2Et CN S7 114 Me2N CN COMe Me2N Et3O+BF74 or Me2SO4 Me2N CN Me2N Me MeO X7 + 115 MeONa Me2N CN Me2N CH2 MeO 116 XCH2Y HN O 118 Me2N CN Me2N CH2 N O 117 Me2N CN Me2N Me Y X NBn NBn S S Me Me 120 NBn NBn O O Me Me 119 RL, PhMe 100 8C, 4 h P S P S S S C6H4OMe-p.p-MeOC6H4 LR= cyclo-C6H11NH2+CH2(CO2H)2+Me2CO+CO2 121 NaOH EtOH cyclo-C6H11NH Me Me O O O O cyclo-C6H11NH 121 MeOH, H+ 7CO2 122 cyclo-C6H11NH cyclo-C6H11NH H CO2Pri 123 NH COMe COMe NMe (CH2)n HO7orMeO7 124 NH COMe NMe (CH2)n Alk2N COMe COMe Alk2N Nu7 Alk2N COMe Alk2N O7 Me Nu Alk2N C Alk2N Me O7 H2O or ROH Alk2N COMe H Alk2N 932 V A Makarov, V G GranikRecently,154 an important method for C-deacylation of ene- diamino diketones upon treatment with cobalt acetate with intermediate formation of chelate complexes 125 and 126 has been suggested. Elimination of the b-formyl group in acidic medium has been described 93 for ketene aminals incorporating the formamidine moiety.N-Acylketene N,N-acetals readily undergo N-deacylation on treatment with sodium methoxide in methanol.80 The transformation of b-substituents in `push-pull' enamines under the action of nucleophilic reagents has been reported.47, 48, 92 For example, treatment with hydrobromic acid gives bromoimines,92 and heating with aniline hydrochloride results in salts of ketene aminals.Treatment with alkali converts the cyano group in `push-pull' enamines into the carbamoyl group.48 If a ketene aminal contains two unsubstituted amino groups, it can react with diethyl malonate to give a cyclic diacylketene aminal.47 Diverse transformations of the cyano group in dicyanodiami- noethylene 127 have been discovered in a study of its reaction with chloromethylenemalononitrile.155 In addition, one should note such a reaction involving ketene N,N-acetals as thermolysis of compounds 128 based on Mel- drum's acid.9, 25 The result of this process depends on the reaction temperature and the nature of the N-substituent.It has been found recently 119 that the `push-pull' enamines 129 with at least one nonsubstituted amino group readily react with amidoacetals to give the corresponding amidines 130 and 131 and the reaction rate depends on the extent of substitution of the other amino group (NH2>NHMe>NMe2). Transamination was studied using enamidine 131 as an example.It is known156 that enamidines undergo transamination more readily than the corresponding enamines. However, it was found that compound 131 reacts in a non-standard way due to the additional a-amino-containing fragment, and replacement of enamidine fragments is accompanied by the reaction on the meso-carbon atom of the amidine group.119 Alk2N COMe COMe Alk2N Alk2N COMe Alk2N Nu H COMe Alk2N COMe H Alk2N NuH NH2 COR0 R0OC RHN O Co N RHN R0 R0OC H 125 Co(OAc)2 . 4H2O 126 O Co N RHN R0 H H NH2 H COR0 RHN RR0N CHO CN H2N RR0N H CN H2N RR0N CHO CN N Me2N H+ OH7 H+ MeONa MeOH RCONH H2N CO2Et COR0 H2N H2N CO2Et COR0 H2N CN CN RHN H2N CN RHN NH Br H2N CONH2 CN RHN H2N+ CN PhHN NH2 PhHN Cl7 PhNH2 HCl, 190 8C HBr OH7 (H2O) NH NH CN CN O O H2N CN CN H2N CH2(CO2Et)2 EtONa, EtOH + Me2N CN CN Me2N 127 Me2N Me2N N Cl CN CN H NC (50% ± 80%) H CN CN Cl N RS N BnS N S R1R2NC S N S EtOC S RSH BnSCS2Na R1R2NCS2Na EtOCS2K H H NH Me2N O O O O 128 560 8C R=Allyl O NMe2 CN [3,3]-sigmatropic rearrangement N C C O NMe2 N Me2N C CO NH Me2N C CO *1,3-H *1,3-R Me2N O O O O RHN 128 C O O O O PhN R=Ph 5310 8C 4310 8C CO C C PhN 129 RR0N CN NO2 H2N RR0N CN NO2 N Me2N 130 N CN NO2 N Me2N Me2N 131 R=H, R0=Me or R=R0=Me R=R0=H D (EtO)2CHNMe2 Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 933A very important property of ketene N,N-acetals is their ability to form chelate complexes with metals 115, 157 ± 160 and boron compounds.161 ± 166 This enabled the development of a new direction of heterocyclic synthesis that opens a route to diverse azaheterocyclic compounds (see below).We consider here only some of the unusual properties of chelate complexes, which provide the possibility to perform processes non-character- istic of the starting ligands. It is known167, 168 that enamines incorporating a primary amino group react with amide and lactam acetals primarily on theNH2 group. If boron chelates are used, an opportunity appears to perform reactions on the active methylene group without involving the amino group.For example, it has been shown161 ± 162, 165 that the chelate 132 reacts with dimethylform- amide and dimethylacetamide acetals selectively on the CH2 group to give compounds 133. 3. Synthesis of heterocyclic compounds based on ketene N,N-acetals a. Four-membered heterocycles Benzoylation of ketene aminals results in azetine derivatives.88 The formation of diazetidine derivatives 131 has been discussed above (see Section II.8).b. Five-membered heterocycles The closure of pyrrole rings according to the Thorpe ± Ziegler and/or Dieckmann reaction starting from ketene N,N-acetals of the benzoimidazole series has been described.14 The pyrrole ring could also be obtained by cyclisation of aminals 134 on refluxing in dioxane.100 Synthesis of pyrazole derivatives occupies an important place in the chemistry of `push-pull' enamines.For example, the strategy of chelate synthesis considered above was also found to be very efficient for the synthesis of various substituted pyrazole- containing systems.162, 165 An example of this kind of synthesis is shown in the Scheme 3.Scheme 3 Similar synthesis of indazole derivatives 135 from the chelate 133 and hydrazine hydrate has been described.165 A wide range of substituted pyrazoles have been obtained by transamination of the dialkylamino group in ketene aminals with hydrazine hydrate followed by cyclisation on the cyano group.50, 84, 113, 143, 169, 170 For example, 3,5-diamino-4-pyrazole has been obtained this way.Similarly, the hydrazino group can replace the good leaving group, viz., CHal3.87 For example, this method was used to obtain alkyl 3,5-diaminopyrazole-4-carboxylates. 131 + + N CN NO2 BnHN BnHN BnHN CN NO2 BnHN BnNH2 (28%) (16%) + (12%) N CN NO2 H2N BnHN O O BzHN B HN Ph Ph 132 O O H2N BzHN (MeO)2CRNMe2 Ph2BBu O O BzHN B HN Ph Ph R Me2N 133 R=H, Me.R3 R1R2N H2N CO2Et N R1R2N CO2Et O Bz BzCl X=CN, CO2Et; Z=NH2, OH. N N CN NC NH2 Z N N CN X CH2CN CH2CN NH NH CN X D ClCH2CN 134 NH NH CH2CO2Et COAr N O COAr NH D O O CO2Et BzHN H2N COMe CO2Et BzHN HN O B Me F F F2BOBu THF (MeO)2CRNMe2 THF H2N4 .H2O CO2Et BzHN HN O B F F NMe2 R N N H O H N H O N Ph R OEt H N N O H H N H O N R OEt Ph H 133 O O BzHN B HN Ph Ph R Me2N O BzHN H2N N N R H 135 N2H4 .H2O N N NH2 NO2 H2N H N2H4 .H2O NO2 NH2 RR0N NC NO2 NH2 H2NHN NC N2H4 .H2O CO2R NH2 F3C NC CO2R NH2 H2NHN NC N N H2N CO2R NH2 H 934 V A Makarov, V G GranikCyclisation can involve not only cyano groups but also other functional groups.85, 86 N-Aminoketene aminals obtained in a different way can also be used as starting compounds in the synthesis of fused pyra- zoles.52 An example of such a synthesis based on compound 136 is shown in the scheme: The synthesis of pyrazolylbenzoimidazole 137 is based on the reaction of gem-formyl groups of the starting benzoimidazole with hydrazine hydrate.106 Several studies171, 172 dealt with the synthesis of imidazole derivatives from `push-pull' enamines (see also Ref. 18). For example, the reaction of ethyl diaminomethylenecyanoacetate with phosgene or diethyl oxalate results in imidazolidinediones.171 Imidazoles are formed from salts of alkyl dicyanoacetate or tricyanomethane and amino acids,172 ketene aminals being the intermediate products.Substituted isothiazoles were synthesised by the oxidation of the appropriate enediaminothioamides.26, 98 The synthesis of tetrazoles from ketene aminals has already been discussed above (see Section II.8).129 Finally, the reaction of hydrazinoenamine 138 with carbon disulfide results in the closure of both the pyrazole and pyrimidine rings to give a purine isoster, viz., pyrazolo[3,4-d]pyrimidine 139.22 c.Synthesis of six-membered heterocycles Thiapyran derivatives were obtained according to the following scheme.145 A few studies 166, 173 deal with the synthesis of substituted pyridines based on ketene aminals with the use of the chelate synthesis strategy.Recyclisation of pyrazolylketene aminal 140 at high temper- ature results in pyrazolo[3,4-b]pyridone 141.7 N N CN H2N Ph H N2H4 .H2O COPh CCl3 H2N NC COPh NHNH2 H2N NC N NH S NO N NH S N CHPh N N SMe N CHPh O2NCH2CO2Et NaOMe N NH N NO2 CO2Et CHPh 136 N NH NH2 NO2 CO2Et H+ (EtOH) + (55%) (14%) N NH N NO2 O H N NH N NO2 OEt NH CHO CHO NH NH N N NH 137 N2H4 .H2O + NR NH O O CO2Et CN NR NH O O CN CO2Et CN RHN H2N CO2Et (CO2Et)2 or COCl2 NH CN CO2R NH O R0 NC OK OR NC H2NCHR0CO2H 7 CN NC X K+ MeNHCH2CO2H + NMe CN CONH2 NH O NMe CN CN NH O R2N NO2 CNHR0 R2N S N S R2N R NR0 NO2 [O] 138 H2N CONH2 CN H2NHN CS2, KOH EtOH 139 (69%) N HN N N O S NH2 H H + S CO2Me CO2Me Me Me SMe O NR1 MeS Me Me MeO2C CO2Me S NR2 (CH2)n H+ (MeOH) + S Me Me SMe NR2 NR1 I7 (CH2)n + (CH2)n NR1 NR2 Me Me S7 S MeI NaHCO3 + + S CO2Me CO2Me Me Me SMe OMe S CO2Me CO2Me Me Me O O + S7 Me Me SMe NR2 NR1 (CH2)n CCO2Me CCO2Me OMe R0 NMe2 OMe BX2 H2N COMe COMe RHN O B HN RHN COMe Me X X BuOH HN RHN COMe R0 O O B HN RHN COMe X X R0 Me2N 141 N O O O O 140 H2N N N H Bn Ph2O 220 8C Bn N N N O NH2 H Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 935Methyl bisanilinomethylene cyanoacetate undergoes cyclisa- tion into quinolinone 142.15 Similar cyclisations have been described in Refs 128 and 140.Intramolecular cyclisation of substituted b-cyanoacetyl-b- nitroketene N,N-acetals results in 4-pyridone derivatives 143a,b.94 It is of note that the hydroxypyridine tautomeric form predominates for pyridone 143a (R=H), whereas the pyridone form predominates for pyridone 143b (R=Me).Intramolecular cyclisation of ketene N,N-acetal 144 upon treatment with sodium hydride to give benzopyridone 145 has been reported.18 Finally, the synthesis of various pyridones based on alkoxy- dienediamine vinylogues 146 and 147 should be noted.70 Thermolysis of N-acylaminals 148 results in substituted oxa- zin-6-ones 149.29 Another method for the synthesis of substituted 1,3-oxazin-6- ones has also been reported.174 The majority of studies devoted to the synthesis of hetero- cyclic compounds from ketene aminals deal with the synthesis of pyrimidine derivatives.In this case, chelate synthesis is also an important approach to their synthesis. The use of boron chelates 132 has enabled the synthesis of substituted pyrimidines 150 and 151.161, 163, 164 The condensation of b-diketones with N-cyanoamidines (R3=Me, Ph) or dicyanoguanidine (R3=NH2) in the presence of nickel acetate, which is required for the formation of reactive chelates, occurs through intermediate derivatives of ketene N,N-acetals with their subsequent cyclisation into substituted 2,4-diaminopyrimidines.160, 175 Ph2O, D PhHN MeO2C CN NHPh N O CN NHPh H 142 PhHN X X PhHN N Cl CO2Et NHPh N N O O H H N N NH2 NH2 POCl3, Et3N X=CO2Et CF3SO3H, CH2Cl2, 20 8C X=CN Ph2O, D X=CO2Et RHN NCCH2C NO2 NHR O 1.MeONa, MeOH 2. H+ N OH NO2 NHR RHN 143a,b N O NO2 NHR RHN H N O CN Ph N NaH Cl O CN N NHPh 144 N O CN Ph 145 OH Me2N Me2N MeO NHR X NC Me2N Me2N CN CH2 MeO 146 RNCX N NC Me2N O OMe H N NC Me2N SMe O Ph Me2N CN Me2N O PhHN SMe 150 ± 160 8C X=S, R=Ph 150 8C X=O, R=H 146 MeNCS N S NMe2 OMe Me CN 147 148 Me Me O O O O R0HN RCOHN 108 8C C CO R0HN RCOHN N O NHR0 O R 149 C CO N R HO R0HN R=Me, Ph.BzHN CO2Et COR H2N N O COR NH2 O Ph Ph2O 260 8C 132 O O NHBz B NH Ph Ph 150 N O NH2 N Ph NH H2N C6H4Me-p C5H5N, NH3 110 ± 120 8C 133 R=H (MeO)2CRNMe2 151 O NH2 N N C6H4Me-p BzHN 936 V A Makarov, V G GranikEnediaminodicarbonyl compounds react with isocyanates to give pyrimidines.176, 177 For example, cyclic aminal 152 was converted into pyrimidinone 153 by treatment with phenyl iso- cyanate.Ketene aminals with the N-ethoxycarbonylcarbamoyl sub- stituent 103 (154) or with the N-benzoylthiocarbamoyl substitu- ent 178 (155) at the b-position undergo cyclisation into pyrimidine derivatives on heating or under conditions of basic catalysis.Enediamino diketones 156 react with aryl isocyanates giving intermediate 4-arylamino-5-acetylpyrimidin-2-one derivatives 157, which are converted into pyrido[2,3-d]pyrimidines 158 under the reaction conditions.179 Pyrimidine derivatives 159 have been obtained as products of the reaction of 6-aryl-2,2-dimethyl-1,3-dioxin-4-ones with ketene aminals.180, 181 The ethoxycarbonyldicyanomethane salt 160 reacts with methyl 3-aminothiophene-2-carboxylate and amino acid esters to give ketene N,N-acetals which can undergo cyclisation into pyrimidine and thieno[3,2-d]pyrimidine derivatives.142 Compound 61 reacts with aniline in acetic acid to give the intermediate ketene aminal 161, which then undergoes cyclisation into pyrimidinone imine 162.93 a,a-Diamino-b-carbamoyl-b-cyanoethylene derivatives are transformed on heating with methyl orthoformate to give sub- stituted 4-amino-5-cyanopyrimidin-6-ones.97 Cyclisation of `push-pull' enaminoamidines 163 under various conditions results in moderate yields of 6-amino-2-phenylpyrimi- dine derivatives.20 Alkoxydienediamines are also convenient starting compounds for pyrimidine synthesis.70 For example, diamine 116 reacts with amidines to give substituted pyrimidines 164.+ Ni(OAc)2 130 ± 140 8C R1OC R2OC H2N N CN R3 N H2N R1OC R3 NH2 COR2 N N R3 NH2 COR2 R1 153 N N N N O O Ph Ph O O PhNCO 160 8C 152 HN MeO2C CO2Me NH 154 N N N X O O H H NH CONHCO2Et X NH PhBr D 155 COR2 R1HN H2N CSNHBz 1.MeONa, MeOH 2. H+ N N R1HN COR2 S Ph H N N O Ar0 Me MeOC ArHN 157 156 ArHN MeOC COMe NH2 Ar0NCO Ar0NCO N N O Ar0 Me N ArHN O H 158 N N Me Me Ar= . 7Me2CO COR2 H2N H2N COR1 O O O Me Me Ar O Ar CO 160 NH NH COR1 COR2 O Ar 159 N COR1 COR2 H2N O H ArOC CN CO2R N S H MeO2C NH NH CN CO2R O S DMF, D NH2 S NH2 CO2Me HCl + NC OK OR NC NH CN CO2R EtO2C Na PhMe, D NH2 NH NH CN CO2R O + NC OK OR NC + CO2Et NH3 Cl7 D O O PhNH2 AcOH, 20 8C CN N Cl CHO Me2N 61 N N NPh CN Ph 162 161 CN N NHPh CHO PhHN NC CONHR NH2 NC N O N R HC(OMe)3 N N (50%) CN N ArHN CO2Et H2N Ph 163 N N NHAr CO2Et Cl Ph HCl O O EtOH EtONa p-TsOH (55%) (65%) N HN NHAr CN O Ph N N NHAr CO2Et NH2 Ph Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 937Finally, let us mention an unusual synthesis of 4,6-bis(dime- thylamino)-5-nitropyrimidine 165 from diamidine 131.182 Compound 165 is also formed upon treatment of diamidine 131 with DMF diethyl acetal. The following scheme of this non- trivial transformation was suggested:182 d. Seven-membered heterocycles Simultaneous closure of five- and seven-membered heterocycles in the reaction of aminal 166 with triamine 167 to give compound 168 has been reported in a study13 dealing with the synthesis of diazepine derivatives from ketene aminals.A different approach19 to the synthesis of benzodiazepines has been considered in Section II.1.IV. Conclusion To sum up the data presented in this review, it can be stated that a fruitful direction based on the use of b,b-bifunctionally substi- tuted ketene N,N-acetals in organic synthesis has been created by now. The preparative availability of these compounds and their high reactivity form considerable prospects for their application in the synthesis of diverse organic and especially heterocyclic systems that are of interest both in theoretical and practical aspects.The authors are grateful to Dr. Christoph Hoock (BYK GULDEN, Germany) for his help in using the databases. This review has been written with financial support of the Ministry of Science and Technology of Federal Republic of Germany [the project `Transform' 01 KX9812�Bundesministe- rium fuÈ r Forschung und Technologie (BMFT)].References 1. R K Dieter Tetrahedron 42 3029 (1986) 2. H Junjappa, H Ila, C V Asokan Tetrahedron 46 5423 (1990) 3. Y Tominaga, S Kohra, H Honkawa, A Hosomi Heterocycles 29 1409 (1989) 4. H Takahata, T Yamazaki Heterocycles 27 1953 (1988) 5. A V Komkov, Candidate Thesis in Chemical Sciences, Institute of Organic Chemistry, Russian Academy of Sciences, Moscow, 1995 6.R Bazzen,W Schunak Arch. Pharm. 315 680 (1982) 7. Y S Sanghvi, S B Larson, R K Robins, G R Revankar J. Chem. Soc., Perkin Trans. 1 2943 (1990) 8. J Sandstrom, K Stenvall, N Sen, K Venkatesan J. Chem. Soc., Per- kin Trans 2 1939 (1985) 9. T Mosandi, C O Roppe, R Flammang J. Chem. Soc., Chem. Commun. 1571 (1992) 10. E Ericsson, J Sandstrom, J Wennerbeck Acta Chem.Scand. 24 3102 (1970) 11. R Gompper,W Topfl Chem. Ber. 95 2871 (1962) 12. H-T Wang, X-J Wang, Z-T Huang Chem. Ber. 123 2141 (1990) 13. R Schwesinger, M Missfeldt, K Peters, H G Schnering Angew. Chem. 99 1210 (1987) 14. A K El-Shafei, A M Soliman, A A R Sultan, A M M El-Saghier Gazz. Chim. Ital. 125 115 (1989) 15. Y Tominaga, T Michioka, K Moriyama, A Hosomi J. Heterocycl.Chem. 27 1217 (1990) 16. Y Kuwayama, S Kataoka Yakugaku Zasshi 85 387 (1965); Chem. Abstr. 63 8342 (1965) 17. S Sasho, H Obase, S Ichikawa, T Kitazawa, H Nonaka, R Yoshizaki, A Ishii, K Shuto J. Med. Chem. 36 572 (1993) 18. W-D Rudorf Tetrahedron 36 1791 (1980) 19. K Peseke Tetrahedron 32 483 (1976) 20. C J Shishoo, M B Devani, V S Bhadti, S Ananthan, G V Ullas Tetrahedron Lett. 25 1291 (1984) 21. C J Shishoo, M B Devani, V S Bhadti, S Ananthan, G V Ullas J. Chem. Res. (M) 3882 (1985) 22. V J Ram, N Hague, A Shoeb J. Prakt. Chem. 334 190 (1992) 23. W-D Rudorf, M Augustin J. Prakt. Chem. 319 545 (1977) 24. M Augustin, Ch Groth, H Kristen, K Peseke, Ch Wiechmann J. Prakt. Chem. 321 205 (1979) 25. A B Cheikh, J Chuche, N Manisse, J C Pommelet, K-P Netsch, P Lorencak, C Wentrup J.Org. Chem. 56 970 (1991) 26. US P. 4 075 001; Chem. Abstr. 88 170 135 (1978) 27. BRD P. 2 423 813; Chem. Abstr. 82 170 943 (1975) 28. Z-T Huang, X Shi Synth. Commun. 20 1321 (1990) 29. X Huang, B Chen, G Wu, H Chen Synth. Commun. 21 1213 (1991) 30. Z-T Huang, M-X Wang Synthesis 1273 (1992) 31. K Hartke, S Radau Liebigs Ann. Chem. 2110 (1974) 32. Y Tominaga, S Takada, S Kohra Heterocycles 40 105 (1995) 33.X-J Wang, Z-T Huang Acta Chim. Sinica 47 890 (1989); Chem. Abstr. 112 198 216 (1990) 34. M Augustin, S Bielka Z. Chem. 20 96 (1980) 35. M-X Wang, J-M Lvanjo, Z-T Huang J. Chem. Res. (M) 1001 (1994) 36. E Ericsson J. Mol. Struct. 24 373 (1975) 37. Y Cheng, H-X Wang,W-X Gan, Z-T Huang Synth. Commun. 26 475 (1996) 38. T Mosandl, S Stadtmueller, H W Wang, C Wentrup J.Phys. Chem. 98 1080 (1994) 39. W-D Rudorf, J Koeditz, N Nenze, A Teraskian Phosphorus Sulfur Silicon Relat. Elem. 107 87 (1995) 40. C-Y Yu, L B Wang, Z-T Huang Synth. Commun. 26 2297 (1996) 41. R Gompper, R Kunz Chem. Ber. 99 2900 (1966) 42. W Hanefeld, D Kluck Arch. Pharm. 315 68 (1982) 116 164 + Me2N CN Me2N CH2 MeO NH H2N R N N Me NC Me2N R N N CN NO2 Me2N Me2N 131 N N NMe2 NO2 NMe2 165 BuONa BuOH (EtO)2CHNMe2 + EtO NMe2 EtO7 131 N CN NO2 N Me2N EtO 7Me2N7 N7 CN NO2 N Me2N Me2N EtO H 7 Me2N7 N CN NO2 Me2N EtO N NO2 Me2N EtO NMe2 N7 7EtO7 NMe2 N7 N N NMe2 NO2 NMe2 165 (90%) NMe2 N7 TsOH + 166 N N (CH2)2OTs TsO(CH2)2 NC CN 167 (CH2)2NH2 HN (CH2)2NH2 168 N N N N N N 938 V A Makarov, V G Granik43.L N Chernova, V D Simonov, Yu N Alyamkin, Z M Minnulin Zh.Org. Khim. 15 512 (1979) a 44. N V Nguyen, K Baum Tetrahedron Lett. 33 2949 (1992) 45. K Baum, S S Bigelow, N V Nguyen, T A Archibald, R Gilardi, J L George J. Org. Chem. 57 235 (1992) 46. E Allenstein, R Fuchs Chem. Ber. 101 1232 (1968) 47. W Middleton, V A Engelhardt J. Am. Chem. Soc. 80 2788 (1958) 48. R Neidlein, D Kikely, W Kramer J. Heterocycl.Chem. 26 1335 (1989) 49. D Martin, K-H Schwarz, S Rackow, P Reich, E Grundemann Chem. Ber. 99 2302 (1966) 50. M I Kanishchev, N V Korneeva, S A Shevelev Izv. Akad. Nauk SSSR, Ser. Khim. 2342 (1986) b 51. S Rajappa,M P Nair, B G Advani, R Sreenivasan J. Chem. Soc., Perkin Trans 1 3161 (1981) 52. K Pilgram J. Heterocycl. Chem. 17 1413 (1980) 53. Z-T Huang, L-H Tzai Chem. Ber. 119 2208 (1968) 54.US P. 4 053 619; Chem. Abstr. 88 37 798 (1978) 55. US P. 405 2411; Chem. Abstr. 88 22 902 (1978) 56. L B Wang, C-Y Yu, Z-T Huang Synth. Commun. 25 1353 (1995) 57. M A F Shoraf, E-E H M Eral H A A Hummonda Phosphorus Sulfur Relat. Elem. 92 19 (1994) 58. E Ericsson, T Marnung, J Sandstrom, I Wennerbeck Acta Chem. Scand. 28 1109 (1974) 59. S Mitamura, M Takaku, H Nozaki Bull.Chem. Soc. Jpn. 47 3152 (1974) 60. M Yokoyama, K Arai, T Imamoto J. Chem. Soc., Perkin Trans 1 1059 (1982) 61. R M Mohareb, H Z Shoms, Y M Elkoly Phosphorus Sulfur Relat. Elem. 72 93 (1992) 62. R G Gluschko V G Granik Adv. Heterocycl. Chem. 12 185 (1970) 63. V G Granik Usp. Khim. 53 651 (1984) [Russ. Chem. Rev. 53 383 (1984)] 64. V G Granik Khim. Geterotsikl. Soedin. 762 (1992) c 65.W Ried, H-E Erle Liebigs Ann. Chem. 201 (1982) 66. H Bredereck, K Bredereck Chem. Ber. 94 2278 (1961) 67. BRD P. 2 624 530; Chem. Abstr. 86 189 967 (1977) 68. H Meerwein, W Florian, N Schon, G Stopp Liebigs Ann. Chem. 641 1 (1961) 69. W Kantlehner, H Jaus, L Kienitz, H Bredereck Liebigs Ann. Chem. 2096 (1979) 70. W Kantlehner, I C Ivanov,W W Mergen,H Bredereck Liebigs Ann.Chem. 372 (1980) 71. W Kantlehner, W W Mergen Synthesis 343 (1979) 72. G Maas, B Feith Synth. Commun. 14 1073 (1984) 73. B Feith, H M Weber, G Maas Chem. Ber. 119 3276 (1986) 74. W Kantlehner, U Greiner Liebigs Ann. Chem. 965 (1990) 75. V A Dorokhov, M F Gordeev, E M Shashkova, A V Komkov, V S Bogdanov Izv. Akad. Nauk SSSR, Ser. Khim. 2600 (1991) b 76. V A Dorokhov, M F Gordeev, V S Bogdanov, A L Leister, V V Semenov Izv.Akad. Nauk SSSR, Ser. Khim. 2660 (1990) b 77. V A Dorokhov, M F Gordeev, Z K Dem'yanets, M N Bochkareva, V S Bogdanov Izv. Akad. Nauk SSSR, Ser. Khim. 1806 (1989) b 78. V A Dorokhov, M F Gordeev, Z K Dem'yanets, V S Bogdanov Izv. Akad. Nauk SSSR, Ser. Khim. 1683 (1987) b 79. V A Dorokhov, M F Gordeev, V S Bogdanov Izv. Akad. Nauk SSSR, Ser.Khim. 1431 (1988) b 80. V A Dorokhov, Z K Dem'yanets Izv. Akad. Nauk, Ser. Khim. 419 (1993) b 81. A C Veronese, V Gandolji J. Chem. Res. (M) 1837 (1988) 82. S Trofimenko, E L Little, H F Mower J. Org. Chem. 27 433 (1962) 83. J A Elvidge, P N Judson, A Pezeival, R Shah J. Chem. Soc., Perkin Trans 1 1741 (1983) 84. M Coennen, J Faust, C Ringel, R Mayer J. Prakt. Chem. 27 239 (1965) 85.B B Gavrilenko, S I Miller J. Org. Chem. 40 2720 (1975) 86. M H Elnagdi, S M Fahmy, E A A Hafez, M R H Elmoghayar, S A R Amer J. Heterocycl. Chem. 16 1109 (1979) 87. N D Bodnarchuk, B B Gavrilenko, V V Momot Zh. Org. Khim. 9 36 (1973) a 88. N D Bodnarchuk, A A Yatsishin Zh. Org. Khim. 13 954 (1977) a 89. V P Kukhar', V I Pasternak, G V Pesotskaya Zh. Org. Khim. 9 39 (1973) a 90.F Lautenschlaeger,M Myhre, F Hopton, J Wilson J. Heterocycl. Chem. 8 241 (1971) 91. Y Masaki, T Miuro, M Ochiai Bull. Chem. Soc. Jpn. 69 195 (1996) 92. E Allenstein Chem. Ber. 96 3230 (1963) 93. K Klemm,W PruÈ sse, L Baron, E Daltrozzo Chem. Ber. 114 2001 (1981) 94. H Mertens, R Troschutz Arch. Pharm. 319 947 (1986) 95. T Yamazaki, K Matoba, T Imai, R N Castle J. Heterocycl.Chem. 16 517 (1979) 96. T Sasaki, A Kojima J. Chem. Soc., C 476 (1970) 97. M T Cocco, C Congiu, A Maccioni, V Onnis J. Heterocycl. Chem. 31 329 (1994) 98. S Rajappa, B G Advani,R Sreenivasan Tetrahedron 33 1057 (1977) 99. S Rajappa,M D Nair, R Sreenivasan, B G Advani Tetrahedron 38 1673 (1982) 100. Z-T Huang, Z-R Liu Chem. Ber. 122 95 (1988) 101. G Dannhardt, S Laufer, K Ziereis Arch.Pharm. 321 429 (1988) 102. J D Albiright, R G Shepherd J. Heterocycl. Chem. 10 899 (1973) 103. E-S A H Badawey, S M Rida, F S G Soliman, T Kappe J. Heterocycl. Chem. 26 405 (1989) 104. I V Zhvinchuk,M O Lozinskii, A V Vypiralenko Zh. Org. Khim. 30 909 (1994) a 105. M D Nair, S Rajappa, J A Desai Indian J. Chem. 21B 1 (1982) 106. H A Naik, V Purnaprajna, S Seshadri Indian J.Chem. 15B 338 (1977) 107. J Bourson Bull. Soc. Chim. Fr. 2373 (1973) 108. J Bourson Bull. Soc. Chim. Fr. 525 (1974) 109. J Clark, J Gelling, G Neath J. Chem. Soc., Chem. Commun. 859 (1967) 110. J Clark,M Curphey, J W Southon J. Chem. Soc., Perkin Trans 1 1611 (1974) 111. J Clark, J Gelling, J W Southon,M S Morton J. Chem. Soc., Perkin Trans 1 494 (1970) 112. J Clark, B Parvizi, J W Southon J.Chem. Soc., Perkin Trans 1 125 (1976) 113. J Clark, B Parvizi, R Colman J. Chem. Soc., Perkin Trans 1 1004 (1976) 114. D St C Black, N E Rothinie Aust. J. Chem. 36 2413 (1983) 115. V A Makarov, A L Sedov,M P Nemeryuk, N P Solov'eva, O S Anisimova, T S Safonova Khim. Geterotsikl. Soedin. 1420 (1994) c 116. M Winn J. Org. Chem. 40 955 (1975) 117. V A Makarov, A L Sedov,M P Nemeryuk, N P Solov'eva, O S Anisimova, T S Safonova Khim.Geterotsikl. Soedin. 208 (1995) c 118. M P Nemeryuk, A L Sedov, V A Makarov, N P Solov'eva, T S Safonova Khim. Geterotsikl. Soedin. 999 (1991) c 119. V A Makarov, A L Sedov, O S Anisimova, V G Granik Khim. Geterotsikl. Soedin. 811 (1996) c 120. V A Makarov, A L Sedov,M P Nemeryuk, N P Solov'eva, T S Safonova Khim.Geterotsikl. Soedin. 976 (1994) c 121. I V Oleinik, O A Zagulyaev, O A Grigor'eva, V M Mamaev Khim. Geterotsikl. Soedin. 1110 (1991) c 122. W Hanefeld, B Spangenberg Arch. Pharm. 321 253 (1988) 123. W Traube, A Eyme Ber. Dtsch. Chem. Ges. 32 3176 (1899) 124. V E Tishchenko, N V Koshkin Zh. Org. Khim. 4 1021 (1934) a 125. A Stephen Monatsh. Chem. 97 695 (1966) 126. M Augustin, E Gunther Z.Chem. 30 169 (1990) 127. R W Saalfrank, H Wirth Chem. Ber. 122 969 (1989) 128. S Shi, F Wude J. Org. Chem. 53 5379 (1988) 129. R W Saalfrank, R Harbig, O Struck, E M Peters, K Peters, H G von Schering Z. Naturforsh. B, Chem. Sci. 51 399 (1996) 130. M Regitz, A Liedhegener, D Stadler Liebigs Ann. Chem. 713 101 (1968) 131. T N Ghosh, D Das-Gupta J. Indian Chem. Soc. 41 (1942) 132.E Malamidou-Xenikaki J. Chem. Soc., Perkin Trans 1 2523 (1996) 133. J Sandstrom Top. Stereochem. 14 83 (1983) 134. J Shvo, H Shanan-Atidi J. Am. Chem. Soc. 91 6683 (1969) 135. J Shvo, H Shanan-Atidi J. Am. Chem. Soc. 91 6689 (1969) 136. N P Kostyuchenko, V G Granik, A M Zhidkova, R G Glushkov, Yu N Sheinker Khim. Geterotsikl. Soedin.1212 (1974) c 137. U Sjostrand, J Sandstrom Tetrahedron 34 3305 (1978) 138.G Isaksson, J Sandstrom Tetrahedron Lett. 24 2233 (1967) Synthesis and properties of b,b-bisfunctionalised ketene N,N-acetals 939139. A Z-Q Khan,R Isaksson, J Sandstrom J. Chem. Soc., Perkin Trans 2 491 (1987) 140. R Gompper, R Kunz Chem. Ber. 98 1391 (1965) 141. D A Long,W O George Spectrochim. Acta 20 1799 (1964) 142. R Neidlein, Z Sui Helv.Chim. Acta 74 579 (1991) 143. N P Solov'eva, V A Makarov, V G Granik Khim. Geterotsikl. Soedin. 89 (1997) c 144. J Sandstrom, U Sjostrand Tetrahedron 34 371 (1978) 145. A Z-Q Khan, J Sandstrom J. Org. Chem. 56 1902 (1991) 146. G F Reynolds, A F Saari J. Heterocycl. Chem. 12 295 (1975) 147. G F Reynolds, R Berber, W Boutwell J. Heterocycl. Chem. 9 1009 (1972) 148. A P Katritskii (Ed.) Fizicheskie Metody v Khimii Geterotsikli- cheskikh Soedinenii (Physical Methods in the Chemistry of Heterocyclic Compounds) (Moscow: Khimiya, 1966) 149. A A Yatsishin, N D Bodnarchuk Zh.Org. Khim. 15 1381 (1979) a 150. K Baum, N V Nguyen, R Gilardi, J J Flippen-Anderson, C George J. Org. Chem. 57 3026 (1992) 151. D G Glover,M J Kramer J. Org. Chem. 26 4734 (1961) 152. M J G Trmino, A Linden, H Heimgarther, A M Cabrera Helv. Chim. Acta 76 2817 (1993) 153. K Peseke Z. Chem. 15 146 (1975) 154. V A Dorokhov, M F Gordeev, A V Komkov, V S Bogdanov Izv. Akad. Nauk SSSR, Ser. Khim. 401 (1990) b 155. W Ried, K Schlpke Liebigs Ann. Chem. 389 (1986) 156. V G Granik, V N Dozorova, N B Marchenko, L I Budanova, V A Kuzovkin, R G Glushkov Khim.-Farmatsevt. Zh. 1249 (1987) 157. C Schmidt, K Polborn, W Beck Chem. Ber. 125 61 (1992) 158. V A Dorokhov, M F Gordeev,M N Bochkareva, M G Kurella, L G Vorontsova, S V L'vov, O S Chizhov Izv. Akad. Nauk SSSR, Ser. Khim. 1134 (1989) b 159. V A Dorokhov, M F Gordeev Izv. Akad. Nauk SSSR, Ser. Khim. 1211 (1989) b 160. V A Dorokhov, M F Gordeev, A V Komkov, V S Bogdanov Izv. Akad. Nauk SSSR, Ser. Khim. 159 (1991) b 161. V A Dorokhov, M A Prezent Izv. Akad. Nauk, Ser. Khim. 888 (1994) b 162. V A Dorokhov, M A Prezent, V S Bogdanov Izv. Akad. Nauk, Ser. Khim. 2211 (1994) b 163. V A Dorokhov, M A Prezent Izv. Akad. Nauk, Ser. Khim. 1504 (1993) b 164. V A Dorokhov, M A Prezent, V S Bogdanov Izv. Akad. Nauk, Ser. Khim. 1638 (1994) b 165. V A Dorokhov, M F Gordeev,M A Prezent Izv. Akad. Nauk SSSR, Ser. Khim. 525 (1991) b 166. V A Dorokhov, A V Komkov, A M Sakharov, V S Bogdanov Izv. Akad. Nauk, Ser. Khim. 177 (1996) b 167. V G Granik, N B Marchenko, E O Sochneva, T F Vlasova, A B Grigor'ev, M K Polievktov, R G Glushkov Khim. Geterotsikl. Soedin. 1506 (1976) c 168. V G Granik, Doctoral Thesis in Chemical Sciences, All-Union Research Chemical Pharmacological Institute, Moscow 1978 169. W D Rudorf, M Augustin J. Prakt. Chem. 320 585 (1978) 170. V A Makarov, O S Anisimova, V G Granik Khim. Geterotsikl. Soedin. 329 (1997) c 171. N D Bodnarchuk, A A Yatsishin Zh. Org. Khim. 13 2432 (1977) 172. R Neidlein, Z Sui Chem. Ber. 123 2203 (1990) 173. V A Dorokhov, M F Gordeev Izv. Akad. Nauk SSSR, Ser. Khim. 2874 (1989) b 174. V A Dorokhov, V S Bogdanov, M F Gordeev Izv. Akad. Nauk SSSR, Ser. Khim. 2874 (1989) b 175. V A Dorokhov, M F Gordeev, A V Komkov, V S Bogdanov Izv. AN SSSR. Ser. khim. 145 (1990) 176. V A Dorokhov, M F Gordeev, A V Komkov, V S Bogdanov Izv. Akad. Nauk SSSR, Ser. Khim. 2593 (1991) b 177. H Wamhoff, W Lamers Synthesis 111 (1993) 178. V A Dorokhov, A V Komkov, E M Shashkova, V S Bogdanov, M N Bochkareva Izv. Akad. Nauk, Ser. Khim. 1932 (1993) b 179. M F Gordeev, A V Komkov, V A Dorokhov Khim. Geterotsikl. Soedin. 1286 (1990) c 180. V L Gein, S G Pitirimova, O V Vinokurova, Yu S Andreichikov, A V Komkov, V S Bogdanov, V A Dorokhov Izv. Akad. Nauk, Ser. Khim. 1475 (1994) b 181. V L Gein, S G Pitirimova, Yu S Andreichikov,M F Gordeev, V S Bogdanov, V A Dorokhov Izv. Akad. Nauk, Ser. Khim. 1225 (1992) b 182. V A Makarov, V A Tafeenko, V G Granik Khim. Geterotsikl. Soedin. 343 (1997) c a�Russ. J. Org. Chem. (Engl. Transl.) b�Russ. Chem. Bull. (Engl. Transl.) c�Chem. Heterocycl. Conpd. (Engl. Transl.) d�Pharm. Chem. J. (Engl. Transl.) 940 V A Makarov, V

ISSN:0036-021X    

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Abstract. Characteristic features of the synthesis of isoalkanes and arenes from CO and H2 in the presence of oxide, metallic and zeolite-containing catalysts are considered. Primary attention is paid to cobalt ± zeolite systems. The bibliography includes 96 references. I. Introduction Nowadays, petroleum is the main source of hydrocarbons. How- ever, since petroleum resources are limited, development of methods for production of hydrocarbons from coal, natural gas, oil shales, plant biomass, etc.becomes a topical task. One of these methods is the Fischer±Tropsch synthesis, which includes three stages: (1) production of synthesis gas (a mixture of hydrogen and carbon monoxide) by gasification of solid raw materials (coal, peat, biomass, etc.) or reforming of natural gas; (2) synthesis of hydrocarbons from synthesis gas; (3) upgrading (hydrocracking, hydroisomerisation) of the resulting mixture giving commercial products. The term `Fischer±Tropsch synthesis' is often used in the literature to denote only the second stage of the process, that is, production of a mixture of liquid and solid hydrocarbons from synthesis gas.It is in this sense that we use this term in this review.The synthesis of hydrocarbons from carbon monoxide and hydrogen in the presence of Group VIII transition metals is a widely employed polymerisation reaction on heterogeneous cata- lysts. This process yields mostly alkenes and alkanes the yields and composition of which depend on the nature of the catalyst and the reaction conditions. In some countries, iron and cobalt catalysts are used in industrial syntheses of hydrocarbons from non-petroleum raw materials.At present, catalysts of this type are preferred because they are highly selective in relation to the formation of paraffins the proportion of which in the synthesis products usually exceeds 90%. Cobalt catalysts are traditionally used to produce linear alkanes.These products can be separated to give petrol (C5±C10) and diesel (C11±C18) fractions of liquid hydrocarbons and a wax fraction (synthetic ceresin C19+). Mixtures of liquid hydrocarbons are mostly needed to pro- duce components of engine fuels, detergents, solvents and some medicines and perfumes. At present, production of fuels is the predominant field of application of the Fischer ± Tropsch prod- ucts.The hydrocarbon mixtures synthesised from CO and H2 are not finished liquid fuels. However, the main characteristics of the diesel fraction obtained on Co catalysts conform to those required of diesel fuels (for example, its cetane number is greater than 50). In contrast, the substantial amount of linear alkanes (70% or more) present in the petrol fraction considerably reduces the possibility of using it directly in internal combustion engines.Figure 1 presents the plots for the octane numbers of individ- ual hydrocarbons of various homologous series.1 It can be seen A L Lapidus, A Yu Krylova N D Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninskii prosp. 47, 117913 Moscow, Russian Federation. Fax (7-095) 135 53 28.Tel (7-095) 135 69 93 (A L Lapidus) Received 8 September 1997 Uspekhi Khimii 67 (11) 1032 ± 1043 (1998); translated by Z P Bobkova UDC 665.652.72 Catalytic synthesis of isoalkanes and aromatic hydrocarbons from CO and H2 A L Lapidus, A Yu Krylova Contents I. Introduction 941 II. Oxide catalysts 942 III. Transformations of hydrocarbons on acidic catalysts 942 IV. Zeolite-containing catalysts 943 V.Synthesis of aromatic hydrocarbons and isoalkanes on zeolite-containing catalysts 943 VI. The mechanism of the Fischer ± Tropsch synthesis 946 VII. Distribution of the products of the Fischer ± Tropsch synthesis 946 VIII. Secondary transformations in the Fischer ± Tropsch synthesis 947 IX. Conclusion 949 The number of carbon atoms in the molecule 100 50 0 750 3 5 7 9 Octane number �1 �2 �3 �4 Figure 1.Octane numbers of individual hydrocarbons of various homol- ogous series. (1) saturated hydrocarbons (paraffins and naphthenes); (2) isoparaffins; (3) olefins; (4) aromatic hydrocarbons. Russian Chemical Reviews 67 (12) 941 ± 950 (1998) #1998 Russian Academy of Sciences and Turpion Ltdthat aromatic compounds possess the highest octane numbers and that the octane numbers of isoalkanes are markedly higher than those of the corresponding linear alkanes (especially when the number of carbon atoms in the molecule is large).Thus, isopar- affins, aromatic compounds and cycloalkanes, which are formed in relatively small amounts in the Fischer ± Tropsch synthesis in the presence of traditional Co catalysts, are the desirable compo- nents of high-octane petrols.These compounds can be produced from linear alkanes by isomerisation, cyclisation and cracking; aromatics can also be obtained by dehydrogenation. At present, it is considered expedient to use branched alkanes for increasing the octane numbers of petrols, because these substances give less harmful exhausts into the atmosphere upon transformations in internal combustion engines than do aromatic hydrocarbons, which are also used to improve the quality of petrol.A fairly promising method is direct production of hydro- carbon mixtures with increased contents of isoalkanes from syn- thesis gas. This can be accomplished using oxide and metal- containing catalysts. II. Oxide catalysts Let us consider synthesis of branched hydrocarbons from CO and H2 in the presence of oxide catalysts.1 For example, the process catalysed by TiO2, ZrO2 or CeO2 (375 ± 475 8C, 30 ± 60 MPa) gives isobutane as the major product, its yield being 20 ± 50 g m73 of synthesis gas.{ The synthesis of branched hydrocarbons from CO and H2 in the presence of oxide catalysts includes apparently the following stages: formation of methanol and dimethyl ether, hydrocarbon chain growth, dehydration of higher alcohols and hydrogenation of alkenes to give isoalkanes.2,3 In the presence of mixed oxide catalysts such as TiO2±Al2O3, ThO2 ±ZnO and ZnO ±Al2O3 at 400 ± 450 8C and 2.5 MPa, synthesis gas is converted into hydrocarbons in a yield of 105 g m73 at a degree of CO conversion of 70% ± 80%.Liquid hydrocarbons (C5+) account for about 57 g m73 of the products formed.3 At temperatures above 450 8C, gaseous hydrocarbons are mainly produced (up to 50 ± 70 g m73 of isobutane), whereas below 375 8C alcohols are formed as the major products.In addition to isoparaffins, the reaction products contain unsatu- rated aliphatic hydrocarbons (more than 90% of which are branched), and small amounts of aromatic hydrocarbons and naphthenes.To produce high-quality petrols from CO and H2, catalysts consisting of aluminium, tungsten, zirconium, uranium and lanthanide oxides with ZnO, Cr2O3 and CuO present as additives have been developed. These catalysts are active at >10 MPa and at temperatures above 350 8C.4 A two-stage process for produc- tion of olefins and high-octane petrol from synthesis gas in the presence of iron and oxide catalysts has also been proposed.5 The first stage is carried out over the Fe3O4 (96.8%) ±Al2O3 (2.4%) ± K2O (0.8%) catalyst at 310 8C and at 1 ± 3 MPa and affords a mixture of alkenes in a yield of 171 g m73.In the second stage catalysed by ThO2, the remaining CO and H2 are converted at 400 ± 450 8C and 28 ± 30 MPa into petrol (yield 137 g m73).The process of selective production of branched hydrocarbons from CO and H2 on oxide catalysts has not yet found practical application, because it usually requires relatively drastic condi- tions. In addition, most often, the yield of the main products of this synthesis in the presence of oxide catalysts is rather low. Therefore, hybrid or bifunctional metal-containing catalysts, which are active under milder conditions, are considered to be more promising for the synthesis of branched hydrocarbons.These catalytic systems combine the properties of hydrogenating catalysts (to convert CO and H2 into hydrocarbons) and catalysts for hydrocarbon transformations (in particular, isomerisatiis known6± 9 that hydrocarbons can undergo various trans- formations in the presence of solid acids.Mixed oxides (for example, amorphous aluminosilicate) and zeolites belong to this class of catalysts. Deposition of metal salts onto solid acids or introduction of reducible cations into zeolites by ion exchange gives (after reduction of the metal cations) bifunctional catalysts, which combine the properties of metallic and acidic contacts.Hydrocarbons or intermediates formed from CO and H2 with participation of the metallic component of the catalyst can undergo secondary transformations on the oxide (acid) surface. The majority of these transformations resemble homogeneous catalytic reactions occurring in the presence of strong acids. Therefore, it is usually assumed that the mechanisms of these processes are similar.III. Transformations of hydrocarbons on acidic catalysts In reactions occurring on solid acidic catalysts (as well as in homogeneous acid-catalysed reactions), carbocations play an important role. These species are generated on bifunctional catalysts upon proton transfer from a Brùnsted acid site to an alkene formed on the metallic phase. Carbocations can be classified into carbenium and carbonium ions.Carbenium ions contain a positively charged carbon atom coordinated to three alkyl groups or three hydrogen atoms (R3C+). Carbonium ions are based on a pentacoordinated positively charged carbon atom surrounded by the same substitu- ents, at least one of them being hydrogen (R4C+H). The formation of carbocations on solid acids Z (in particular, on zeolites) includes the following reactions:9 According to reaction (1), a zeolite proton reacts with an alkene to give a carbenium ion.Reaction (2) describes protonation of alkanes resulting in a carbonium ion. The latter is then converted into a smaller carbenium ion upon elimination of an alkane molecule or molecular hydrogen. Both reactions occur on Brùnsted acid sites.It should be noted that process (1) occurs more rapidly and at lower temperatures than process (2). Transformations with hydride ion transfer can also occur on Lewis acid sites. The formation of carbenium ions on bifunctional metal- containing catalysts occurs mostly by reaction (1), i.e. these ions result from the transfer of a proton from a Brùnsted acid site to an alkene formed on the metallic phase.The proton transfer is terminated as a result of fast hydrogenation of unsaturated frag- ments on metal particles. The presence of a metal permits the reaction temperature to be decreased by *100 8C.9 Simultane- ously, isomerisation and cracking of isomerised hydrocarbons occur on acid sites. The rearrangements and cleavage of alkylcar- benium ions are the rate-determining steps.In bifunctional catalysts, a metal ensures the occurrence of hydrogenation ± dehydrogenation reactions, whereas the role of a solid acid (zeolite or a mixed oxide) consists in the formation of (2) R1 R2 +ZH H2+ H R2 R1 + Z7 . Z7 + R1 R1 H H H (1) R1 R2 +ZH H R2 R1 + Z7 , +Z R1 R2 (3) . H R2 R1 + ZH7 { The activity of catalysts in the Fischer ± Tropsch synthesis is usually estimated based on the yield of hydrocarbons produced from 1 m3 of synthesis gas under normal (or standard) conditions (s.t.p.).The max- imum theoretical yield of hydrocarbons amounts to 208.5 g m73 (s.t.p.). 942 A L Lapidus, A Yu Krylovacarbocations on acid sites. These cations undergo isomerisation and destruction and then they are desorbed, mostly as alkenes.The latter are converted into alkanes upon hydrogenation on the metal component of the catalyst. The classical transformation chart can be represented as follows: In accordance with this scheme, carbenium ions result from protonation of alkenes. These ions can rearrange and be desorbed as isomers, or they can be cracked (b-cleavage). Thus, cracking and isomerisation are competing processes.In the general case, isomerisation of alkanes with high molecular masses occurs at lower temperatures than hydrocracking.9 Alkylcarbenium ions can participate in the following trans- formations:9 intramolecular hydride shift, skeletal isomerisation, intermolecular transfer of a hydride ion from a carbenium ion to a substrate molecule to give a new carbenium ion and an alkene (this is normally the slowest step occurring on the strongest acid sites of the catalyst),10 alkylation by alkenes resulting in the formation of longer alkylcarbonium ions, b-cleavage yielding shorter alkylcar- benium ions and alkene fragments and deprotonation (chain termination).All alkylcarbenium ions formed from both neutral molecules and various intermediates can enter into the above reactions.In the presence of bifunctional catalysts, lower alkanes par- ticipate in more complex processes, namely, dehydrogenation and oligomerisation of `internal alkenes' (those with an internal double bond) followed by cracking of the oligomers.11 IV. Zeolite-containing catalysts In recent years, studies have been started on bifunctional catalysts that combine the properties of traditional Fischer ± Tropsch catalysts (redox function) and selective zeolite catalysts (acid ± base function).Both types of catalytic sites are involved in consecutive and parallel transformations.12 Alkenes formed from CO and H2 on the metal-containing component of the catalyst are converted on the zeolite into aromatic and branched aliphatic hydrocarbons, while alkanes are cracked, which increases the petrol fraction.The catalytic sites for acid ± base reactions are hydroxy groups of the zeolites or water molecules dissociated in the cation field.13 The zeolite hydroxy groups can result from the following proc- esses: (a) hydrolysis Na+-zeolite +H2O NaOH +H+-zeolite (since the sodium form of a zeolite is the salt of a weak acid and a strong base, this equilibrium is shifted to the right, and aqueous suspensions of zeolites are normally alkaline); (b) decomposition of the ammonium form of a zeolite (c) heterolytic dissociation of water molecules in the field of a polyvalent cation and the zeolite framework atoms The aluminium atom has a vacant p orbital; occupation of this orbital results in withdrawal of the electron density from oxygen in the [Al ± O(H) ± Si] group and thus in polarisation of the O±H bond.Thus, the OH group in the [SiOH] fragment, which exhibits weakly acidic properties when incorporated in silica gel, behaves as a strong Brùnsted acid site in the presence of aluminium atoms. Bifunctional catalysts are usually prepared on the basis of HZSM-5 zeolite.This choice is due to the following reasons:8 the average size of the zeolite pores is 5.5 A, which facilitates selective formation of molecules containing less than 10 carbon atoms, i.e. hydrocarbons of the petrol fraction; the strong acid sites of HZSM-5 are responsible for oligomerisation, cracking and aro- matisation; this zeolite is stable against coking, because the presence of medium-size pores prevents the formation of coke precursors; the zeolite is stable under hydrothermal conditions and resistant against dealumination, which is possible under the Fischer ± Tropsch synthesis conditions because this reaction yields a substantial amount of water.The high selectivity of zeolites of the HZSM-5 type towards the formation and subsequent transformations of light hydro- carbons is determined by their structure.In the pores containing strong acid sites, organic molecules are adsorbed according to the `tail-to-tail' pattern.14 In the majority of cases, only alkenes and/ or carbonium ions containing five or fewer carbon atoms are able to participate in reactions occurring in zeolite channels with a size of *9A; therefore, the oligomers formed in these reactions contain not more than 10 carbon atoms.In addition, HZSM-5 zeolites `control the movement of molecules': the reactants enter the zeolite structure through sinusoidal channels, whereas the products are removed through linear channels. As this takes place, alkenes, which slowly diffuse through the pores, are transformed into branched alkanes.The absence of back diffusion is a crucial factor determining the use of HZSM-5 zeolites in catalytic trans- formations. The use of HZSM-5 and other high-silica zeolites as supports for the catalysts for the synthesis of hydrocarbons from CO and H2 has been started quite recently.15 ± 17 To increase the content of aromatic hydrocarbons or isoalkanes in the synthesis products, two types of catalysts are usually employed, namely, hybrid systems comprising an oxide catalyst for methanol synthesis and a zeolite (for example, Zn/HZSM-5 and Zn ± Cr/HZSM-6) 16 and also bifunctional systems based on metallic catalysts of the Fischer ± Tropsch synthesis and zeolites. Both types of catalysts have advantages and drawbacks.Systems of the former type permit selective and highly efficient synthesis of mixtures of aromatic hydrocarbons; however, they are insufficiently stable (the period between catalyst regenerations varies from one to several months).Systems of the latter type are characterised by a lower productivity but they are suitable for the synthesis of both aromatic hydrocarbons and branched paraffins. V. Synthesis of aromatic hydrocarbons and isoalkanes on zeolite-containing catalysts Zeolite-containing catalysts based on Group VIII transition metals are usually prepared by co-precipitation or ion exchange followed by activation with hydrogen. The transition metal cations occur initially within the zeolite structure and on its surface and can be reduced almost completely.13 In the general n-Alkene H+ Secondary n-alkyl cation Isoalkene Tertiary isoalkyl cation H+ Rearrangement Cracking products Cracking products Cracking products Cracking products b-Cleavage b-Cleavage Hydrogenolysis Hydrogenolysis 72H +2H n-Alkane Isoalkane ONHá4 [Al Si] OH+ [Al Si]+NH3 ; OM [Al Si]+H2O where Mis metal.OH+ [Al Si]+MOH, Catalytic synthesis of isoalkanes and aromatic hydrocarbons from CO and H2 943case, the degree of reduction depends on the method of cation activation with hydrogen, the nature of these cations and the reactivity of their oxygen environment in the zeolite.In addition, it depends on the degree of dispersion of the initial compound containing the transition element and on the diffusion processes in the zeolite during reduction. The Co ±O bond of an isolated cobalt species in a zeolite framework is ionic.18 The reduction of cations in the coordination sphere of these zeolites occurs at temperatures above 450 8C,19 cation-exchange zeolites behaving as ordinary salt systems.The mechanism of reduction of transition metal cations in zeolites differs from the mechanism of reduction of the corre- sponding metal oxides.20 The latter is irreversible, the hydrogen that has reacted being removed as water.On the contrary, in zeolites, the protons are retained by the framework where they form hydroxy groups. Cations located in different crystallographic positions have different electrostatic environments and, therefore, different reducibilities.21 The reducibility depends also on the silicon to aluminium ratio in the zeolite framework,22 i.e.on the Si ±O bond length, the residual charge on the framework, and the effective charge on the cation itself. The character of localisation of cations and their reducibility depend appreciably on the catalyst pretreatment conditions. In particular, dehydration leads to the destruction of aqua complexes of transition metal cations; the cations migrate to positions with higher degrees of coordination unsaturation, which decreases their reducibility.23 After reduction, transition metals can occur in zeolite-con- taining catalysts as separate atoms, clusters or crystallites with a size of 5100 A located on the external surface of the zeolite.13 It should be noted that decationised zeolites contain large amounts of strong proton-releasing and electron-withdrawing sites.Thus, the use of a zeolite support makes it possible to prepare catalytic systems differing in the degree of metal dispersion and in the strength of its interaction with the surface. The activity of the most widely used Co catalysts for the Fischer ± Tropsch synthesis is determined to a large extent by the nature of the zeolite support.24 Thus the yield of liquid hydro- carbons on the Co catalysts supported on X and Y zeolites is higher than those on similar specimens based on zeolite A and mordenite.The highest yield of liquid products (166 g m73) over this series of specimens, attained on the Co ± MgO/NaX catalyst, was 1.3 times higher than that on the Co ± MgO/kieselguhr catalyst with the same component ratio (the latter catalyst was close in composition to the commercial one).The nature of synthetic zeolites has a substantial influence on the polymerising and isomerising properties of Co catalysts for the Fischer ± Tropsch synthesis.25 An increase in the SiO2:Al2O3 ratio in a zeolite results in an increased content of high-boiling products in the catalysate and in an increased proportion of normal aliphatic hydrocarbons in these products. Catalysts containing A, X, Y or M zeolites in the H-form or with exchangeable Ca2+,Mg2+ and ZrO2+ cations are less active in the synthesis of hydrocarbons from CO and H2 than catalysts containing zeolites in the Na-form.25 The highest yield of liquid hydrocarbons (125 ± 130 g m73) was achieved on catalysts con- taining NaX and NaY zeolites (SiO2:Al2O3 = 2.8 ± 5).26 When more acidic catalysts were used, the proportion of iso hydro- carbons reached 50 mass %.The composition of the products is substantially influenced by the nature of the exchangeable cation.27 Thus introduction of Mg2+ by ion exchange into NaX zeolite used as a support increases the yield of liquid products 1.5-fold. Simultaneously, the proportions of branched hydrocarbons and alkenes in the catalysate decrease.The catalytic properties of metal zeolite systems depend on the methods of their preparation and modification and on the operation conditions.6 For example, the cobalt particles in the Co/HZSM-5 catalyst prepared by impregnation of the zeolite with an aqueous solution of cobalt nitrate 28 are 3 times larger than those in the specimen prepared using dicobaltoctacarbonyl Co2(CO)8.29 The initial system used to produce the 9.5 mass% Co/ HZSM-5 catalyst contains cobalt nitrate as amorphous particles with a size of*20A.30 After calcination in a flow of air, the nitrate is converted into the oxide Co3O4.The reduction in a flow of hydrogen at 350 8C affords cobalt crystallites with a size of *300 A, which remain virtually unchanged during the Fischer ± Tropsch synthesis, and the cobalt oxide CoO (*90 A).The ratio of these components, equal to 4, remains approximately the same after treatment with synthesis gas, i.e. the catalyst is not reduced during the synthesis. The Co/HZSM-5 catalysts prepared by impregnation contain cobalt in two states: exchangeable non-reducible Co2+ ions within the zeolite structure and crystalline Co3O4 on its external sur- face.15 The latter can be reduced with hydrogen to metallic cobalt and CoO.Specimens containing only exchangeable cobalt ions do not exhibit catalytic activity in the synthesis of hydrocarbons from CO and H2. After preliminary reduction with hydrogen of the 30 mass% Co/HZSM-5 catalyst prepared by mixing basic cobalt carbonate with the support, the degree of Co reduction is 52%± 58%.31 The rest of the cobalt occurs as Co2+.Treatment of the catalyst with synthesis gas results in partial oxidation of the metal on the catalyst surface: the degree of Co reduction decreases to 31%± 35% and Co3O4 (14% ± 17%) is formed. As the Si :Al ratio in the zeolite increases, the Co0 surface concentration somewhat decreases, resulting in a decrease in both the total catalyst activity and the yield of liquid products.Catalysts produced by impregnation or mixing and containing a metallic cobalt phase on the surface of HZSM-5 zeolite crystals are fairly active in the Fischer ± Tropsch synthesis (the degree of CO conversion reaches 60%).15 At 280 8C, these catalysts permit the synthesis of aromatic hydrocarbons, the yield and composi- tion of which depend on the composition of the catalyst and the method by which it was prepared.Thus the hydrocarbon mixtures obtained on the 8.5 mass% Co/HZSM-5 catalyst prepared by mechanical mixing of the components contained up to 49 vol. % of aromatic compounds, whereas the content of aromatics in the products of synthesis on 7.5 mass% Co/HZSM-5 catalyst pre- pared by impregnation was not more than 20 vol.%.Anumber of Co/HZSM-5 catalysts containing from 3 mass% to 9 mass% of the metal and prepared by various methods have been studied.32 The following methods were used: decomposition of C5H5Co(CO)2 on the zeolite (specimen I), impregnation of the support by an aqueous solution of cobalt nitrate (specimen II) and mechanical mixing of cobalt oxide with the zeolite (specimen III).The conversion of synthesis gas at 2.1 MPa and 280 8C on any of these speciments increased with an increase in the cobalt loading. When the Co concentration was 9 mass %, the highest yield of liquid hydrocarbons C5+ (74%) was attained on catalyst I. This specimen contained much less ion-exchangeable cobalt than catalyst II. The rest of the cobalt in I occurred as Co3O4 on the external surface of the zeolite. Catalyst III did not contain exchangeable cobalt ions at all.Thus, the metallic component had no influence on the acidic properties of HZSM-5 zeolite, and this catalyst was best suited for the synthesis of aromatic hydro- carbons from CO and H2.The concentration of aromatics in the liquid products obtained on this catalyst was *51 vol. %. The liquid products produced on specimens I and II were enriched in alkenes (their content was 14 vol.%and 22 vol. %, respectively). The Co/HZSM-5 catalysts employed for the synthesis of aromatic hydrocarbons are prepared conventionally by mechan- ical (dry) mixing of Co3O4 and the zeolite to keep the latter in the H-form.Mixtures formed in the presence of these catalysts [Mn+Z7] + (n/2)H2 [nH+Z7, M0] [nH+Z7] +M0 . [nH+Z7, M0] , 944 A L Lapidus, A Yu Krylovacontain up to 70 mass%of aromatic hydrocarbons.15, 25, 33 How- ever, the yield of the products is not more than 10 ± 20 g m73, which is largely due to the low porosity of the catalyst samples. The yield of liquid aliphatic hydrocarbons in the presence of analogous catalysts prepared by co-precipitation (using basic cobalt carbonate) and promoted by magnesia reaches 150 ± 170 g m73.25, 34 It has been found 35 that the nature of the initial cobalt compound used to prepare the Co/HZSM-5 catalyst influences the yield and the composition of the liquid products formed.The syntheses were carried out at a pressure of 0.1 MPa, a temperature of 220 ± 280 8C, volume ratio H2 :CO=1 and a gas hourly space velocity of 100 h71.The use of basic cobalt carbonate makes it possible to increase the catalyst porosity (the specimen is loosened during reductive decomposition of the carbonate) and increases the yield of liquid products (up to 42 g m73) and the selectivity of their formation (up to 44%).However, simultaneously the con- tent of aromatic hydrocarbons decreases markedly (to 42%). The introduction of 0.4 mass% to 1.5 mass% of ThO2 into the 9 mass% Co/HZSM-5 catalyst prepared by mechanical mixing of the components substantially increases the degree of conversion of synthesis gas (up to 74%) and the selectivity of the formation of the petrol fraction of hydrocarbons (to 70%).32 The proportion of aromatic compounds in the liquid products decreases to 26% but, nevertheless, the octane number of the products is 90.It has been shown by IR spectroscopy that when the Co/HZSM-5 catalyst contains ThO2, the relative concentration of the Brùnsted acid sites in the zeolite diminishes; this accounts for the lower yield of aromatic compounds.A 59Co NMR study has shown the presence of a- and b-cobalt phases in the Co/HZSM-5 catalyst.33 The reduced Co ±ThO2/HZSM-5 specimen contained no hexagonal cobalt. In addition, upon the introduction of ThO2, the size of the cobalt particles arising decreased from 40 to 10 nm;32, 36 this facilitated adsorption of synthesis gas and ensured higher product yields.32 However, after treatment of the catalyst with synthesis gas for 48 h at 280 8C, the size of Co crystallites again increased (apparently, due to sintering of metal particles) and the yield of aromatic hydrocarbons decreased.A similar decrease in the content of aromatic hydrocarbons in the liquid products of the synthesis on Co/ HZSM-5 catalysts prepared using CoCO3 has been observed upon introduction of 2%± 3% of Mg, Zr or La oxides into the catalysts.32 The synthesis carried out under relatively mild conditions (0.1 MPa, 250 ± 280 8C) in the presence of the Co ±Ni ± ZrO2/ HZSM-5 catalyst gave aromatic hydrocarbons with a selectivity of 30 mass%± 35 mass %.37 A decrease in the space velocity, an increase in the temperature or an increase in the zeolite proportion in the catalyst resulted in a higher content of aromatic hydro- carbons in the C6+ fraction.A study dealing with the adsorption properties of bifunctional cobalt ± zeolite catalysts prepared by mechanical (dry) mixing of the components has shown that the adsorption isotherm deviates noticeably from the corresponding curve for the initial zeolite.35 This implies destruction of the zeolite during the catalyst prepa- ration, the degree of destruction being dependent on the duration of the mechanical impact on the specimen.The hydrogenating activity of cobalt can be decreased by introducing Cu into the catalyst. The yield of methane on a mechanical mixture of 4% Co/SiO2 and HZSM-5 zeolite is 1.7 times higher than that on the 4% Co/SiO2 specimen itself (2.1 MPa, 280 8C, H2 :CO=1).However, the proportion of methane formed in the presence of the 6% Co ± 0.05% Cu/SiO2 + HZSM-5 specimen under similar conditions is only 1.2 times higher than that on 6% Co ± 0.05% Cu/SiO2. Thus, the addition of copper decreases somewhat the efficiency of methane forma- tion.38 When developing and employing bifunctional catalysts, one should bear in mind that the rate of synthesis of aromatic hydro- carbons on these catalysts at a given temperature and at the optimum activity of HZSM-5 cannot be higher than the rate of formation of alkenes on the metallic component of the catalyst (the Fischer ± Tropsch component).8 The temperature for the synthesis of aromatic compounds is normally 280 8C.At this temperature, zeolite is active towards aromatisation of alkenes, and methane formation on the Fischer ± Tropsch component is not too pronounced.The maximum rate of formation of aromatic hydrocarbons from alkenes at 280 8C is 0.084 g (g of HZSM-5)71 h71.8 At the same time, a catalyst for the Fischer ± Tropsch synthesis (for example, 4% Co/SiO2) can ensure forma- tion of 0.123 g of alkenes per hour under similar conditions.It is the ratio of 0.084 to 0.123 that determines the ratio of the components in the bifunctional 4% Co/SiO2+HZSM-5 catalyst. The main problem associated with the use of bifunctional zeolite catalysts is their deactivation during the interaction with synthesis gas. The deactivation is due to the following factors:8 deposition of inactive carbon on the metallic component, sintering of metal particles and coking of the zeolite component.The first and the second processes result in a gradual decrease in the overall rate of the synthesis gas conversion. The third process diminishes the selectivity of formation of aromatic hydrocarbons due to a decrease in the number of active sites capable of converting alkenes. An attempt has been made to regenerate the Co ± ThO2/ HZSM-5 catalyst by conducting redox cycles, because oxidative treatment favours the removal of the carbon deposits.However, in each regeneration cycle, cobalt particles sintered, and this retarded the reforming of synthesis gas. In order to suppress metal sintering during the oxidative regeneration of bifunctional zeolite-containing catalysts, the use of oxygen enriched in ozone was proposed.40 Synthesis of branched alkanes from CO and H2 can also be carried out in the presence of bifunctional catalysts containing Group VIII noble metals.41 Most often, these catalytic systems exhibit low activities.Conversely, Co/HZSM-5 catalysts are highly active: at an atmospheric pressure and a temperature of 175 ± 200 8C, they permit production of liquid hydrocarbons (C5±C10), containing 60% ±80% of isoparaffins, in 46% ±59% yield with respect to the converted CO.42 In the presence of 10% Co ± 6% MgO/HZVM catalysts (SiO2:Al2O3 = 30 ± 104) prepared by co-precipitation and con- taining high-silica zeolites, analogues of HZSM-5, the degree of conversion of carbon monoxide is 70%± 100% at 175 ± 200 8C.43 On these catalysts, liquid hydrocarbons enriched in isoalkanes can be synthesised selectively in 70%± 90% yield.The C5±C10 fraction prepared on a catalyst supported on a zeolite with a silica modulus (the SiO2:Al2O3 ratio) of 30 ± 68 contains 36%± 46% of isoalkanes. The zeolite silica modulus has an influence on the yield of isoalkanes, the highest yield being attained at SiO2:Al2O3 = 68. An increase in the reaction temperature from 175 to 220 8C permits selective synthesis of C5±C10 hydrocarbons containing more than 50% of isoalkanes.The influence of the synthesis conditions on the formation of isoalkanes on Ru ± Pt catalysts has been studied.41 It was found that the shorter the contact time, the higher the yield of branched hydrocarbons. When theW: F ratio (Wis the mass of the catalyst, and F is the space velocity of synthesis gas through the catalyst) was proportional to the residence time and was equal to 7 (g of the catalyst) h mol71 (2% Ru ± 2% Pt/HY, 240 8C, 1.5 MPa, CO:H2=2 : 3), the iso to normal ratio (iso/n) for C4 and C5 hydrocarbons was an order of magnitude higher than that expected from the thermodynamic viewpoint (this ratio cannot be achieved upon hydroisomerisation of linear alkanes). As the residence time increased, the iso/n ratio reached the theoretical (thermodynamic) value.This means that the ratio of branched to linear alkanes is controlled by kinetic factors, and an increase in the residence time leads to gradual hydroisomerisation of branched alkanes into linear alkanes. It was found 41 that the Fischer ± Tropsch synthesis on Ru/ zeolite catalysts yields alkenes as the primary products.The predominant formation of branched alkanes in the initial reaction Catalytic synthesis of isoalkanes and aromatic hydrocarbons from CO and H2 945stage may be due to transformation of alkenes on the zeolite acid sites followed by hydrogenation on ruthenium. It can be expected that the ratio of the synthesis products would be determined by the relative reaction rates on the acidic and metallic sites.Since raising the temperature usually enhances the hydrogenating activity of a catalyst, the iso/n ratio simultaneously decreases. A decrease in pressure results in lower iso/n ratios, because hydrogenation is facilitated at low partial pressure of CO. The highest iso/n ratio (ranging from 10 to 40 for various C4+ hydrocarbons) on the 2% Ru ± 2% Pt/HY catalyst was attained at a pressure of 1.5 MPa.41 As the pressure increased, this ratio substantially decreased.The W: F ratio remains constant upon a pressure increase, whereas the real residence time (i.e. the residence time of hydrocarbons in a catalyst layer) becomes longer, and this causes a decrease in the iso/n ratio.It was found that this ratio is inversely proportional to the selectivity of methane formation. When the reaction is carried out on the 2% Ru ± 2% Pt/HY catalyst, the iso/n ratio in the C4+ hydrocarbons depends on the nature of the support and increases in the order AS < CaY < DAHY < HY (where AS is a synthetic amorphous aluminosili- cate, and DAHY is dealuminised HY).41 As the Si :Al ratio in the zeolite increases, the degree of COconversion diminishes, whereas the selectivity of CH4 formation grows. The H-forms of zeolites with higher silica moduli give lower yields of isoalkanes. These results are inconsistent with those obtained in other studies,44, 45 according to which Ru/zeolite catalysts with higher Si :Al pro- portions give products characterised by greater iso/n ratios.This inconsistency may be due to the fact that the iso/n ratio is determined by the relationship between the rates of various steps involved in the synthesis of isoalkanes from CO and H2, which depend on the nature of the metal and the type of the zeolite used. To elucidate the general regularities of the synthesis of hydro- carbons from CO and H2, it is necessary to study systematically various catalysts under identical conditions.VI. The mechanism of the Fischer ± Tropsch synthesis The surface of supported Co catalysts contains two types of metal- containing active sites A and B (Scheme 1).46 ± 49 Sites of type A are metallic cobalt crystallites; on these sites, CO undergoes dissocia- tive adsorption giving surface carbon, which is hydrogenated yielding methane.Thus, sites A are responsible for CO methana- tion. On the same type of sites, as a result of CO dissociation, adsorbed oxygen appears, which is able to react with carbon monoxide to give CO2 (or with hydrogen to give water). Type B sites comprise metallic cobalt and an oxide component. They might be located at the Co ± oxide interface.50 The oxide compo- nent can be represented by cobalt oxide resulting from incomplete reduction of the initial cobalt salt, by an oxide support or by a surface mixed oxide formed upon interaction of cobalt, the support and an oxide promoter during high-temperature pretreat- ment of the catalyst (calcination and/or reduction). On the metallic part of site B, reagents are activated giving intermediates of the type COHx or CHx; these species spill over to the oxide part of the site on which they polymerise.The oxide part of site B is normally a Lewis acid site (L-site). It affects the neighbouring metallic cobalt in such a way that some of the electron density shifts from the metal to the support 51 yielding Cod+. Carbon monoxide is adsorbed on site B in the molecular form; apparently it binds simultaneously to both components of the site.It is also possible thatCOis adsorbed on the interface in such a way that its oxygen atom binds to the oxide component of the active site.52, 53 In this case, the Lewis acid site located near the Co ± oxide interface weakens the C±O bond in the carbon monoxide molecule and increases the rate of its hydrogenation.54 The role of metallic cobalt in active site B is not limited to adsorption of CO (which is also adsorbed efficiently by cobalt oxide); cobalt also ensures hydrogen activation and transport to the polymerisation site.In the case where hydrogen can be delivered to the polymerisation site without participation of the metal (for example, by dissociative adsorption of C2H2 on an oxide 55), CHx radicals can polymerise directly on the oxide.It should also be noted that in addition to the two types of metal-containing active sites mentioned above, the surface of Co catalysts for the Fischer ± Tropsch synthesis contains Lewis and Brùnsted acid sites of various strengths, which are located on the oxide surface and are able to participate in secondary trans- formations of the intermediates formed from CO and H2.In addition, the energetic characteristics of sites of a particular type (A or B) can differ owing to the inhomogeneity of the catalyst surface. VII. Distribution of the products of the Fischer ± Tropsch synthesis The process of formation of hydrocarbons from CO and H2 is an example of catalytic hydropolymerisation.The distribution of the products formed in this reaction can be described using the Schulz or Flory equations, which virtually coincide for high chain growth rates.56 In particular, the Schulz equation is written as follows C O Cos+ L-site or L-site C O Cos+ A Co07Co0 (dissociative adsorption of CO) CO H2 Cads+4Hads CH4,ads CH4,ads B Co07oxide (associative adsorption of CO) CO COads(Co0) COads(oxide)+H2 COHx,ads(oxide) COHx,ads(Co0) CHx,ads(Co0) + H2 a b [COHx,ads(oxide)] n Cads+Oads 2Hads CH4 Cads(Co0) COads(oxide) COHx,ads(oxide) COHx,ads(Co0) CHx,ads(Co0) CH4 n CH2 +H2O (polymerisation) Scheme 1 (a) Proposed by Bartholomew in 1989; (b) proposed by Lapidus in 1991. 946 A L Lapidus, A Yu Krylovagn=(ln2a)nan , where gn is the weight fraction of a product containing n carbon atoms, and a is a factor of the hydrocarbon chain growth characterising the ratio of the probabilities of chain growth and chain termination.In some cases, the distribution of the products of the Fischer ± Tropsch synthesis does not conform to the classical Schulz ± Flory distribution; this is due to the fact that there are at least two centres of chain growth on the catalyst surface, polymerisation on which occurs by different mechanisms.57 ± 59 For example, it has been suggested 60 that hydrocarbons, the chain growth of which is characterised by factors a1 and a2, can arise in the following way: The ratio of the a1- to the a2-products is determined by the hydrogen pressure.It has been assumed 61 that on different active sites of the surface, different intermediates can be formed, which are con- verted, in their turn, into different surface compounds.According to these views, all the reactions occurring on the surface including the formation of the primary complex are reversible. VIII. Secondary transformations in the Fischer ± Tropsch synthesis The deviation of the product composition from the Schulz ± Flory distribution can also be explained by assuming that the alkenes formed fromCOandH2 are adsorbed once again on one or several active surface sites.62, 63 In particular, the readsorption of alkenes substantially decreases the overall rate of the surface chain termination by changing the direction of the hydrogen elimination step, which leads to alkenes with a double bond located at the end of the molecule (terminal alkenes).64 Study of the compositions of the products formed on the Fe/ SiO2 catalyst has demonstrated that there are two types of distribution, namely, a `classical' distribution obeying the Schulz ± Flory law and a `non-classical' distribution, which does not obey this law.65 The non-classical distribution appears because the primary reaction products can be rapidly converted into isomeric secondary products on the acid sites of the catalyst.The secondary transformations include double bond shift, cis ± trans isomerisation, cyclisation and aromatisation. There exists a direct correlation between the content of silica gel in a precipitated iron catalyst and the selectivity of this catalyst towards the formation of internal alkenes and branched hydro- carbons.66 The content of alkenes with an internal double bond increases with an increase in the proportion of silica gel in the catalyst.Terminal alkenes are known to isomerise on silica gel,67 alumina 68 and zeolites.69 In the presence of these catalysts, alkenes undergo the following transformations (listed in order of increasing difficulty of transformation): cis ± trans isomerisation, double bond shift, skeletal isomerisation, cyclisation, hydrogen transfer and polymerisation.These reactions occur on Brùnsted and Lewis acid sites. It is generally believed that termination of a growing alkyl chain upon elimination of a b-hydrogen atom yields terminal alkenes, whereas termination occurring by interaction of hydro- gen with the a-carbon atom in the chain affords linear alkanes.70 The formation of small amounts of cis- and trans-alkenes as the primary products occurs apparently in those rare cases where the growing carbon chain comes in contact with the active site of the catalyst through its b-carbon atom, the chain being terminated upon elimination of d-hydrogen. As the content of SiO2 in the support of the Fe catalyst increases, the yield of alkenes with an internal double bond increases, whereas the amount of terminal alkenes simultaneously decreases; therefore, it can be concluded that internal alkenes result from isomerisation of terminal alkenes on silica gel, which occurs according to the following scheme:66 This mechanism implies that isomerisation proceeds on the same sites on which hydrocarbons are synthesised from CO and H2 and that the two hydrogen atoms attached to the third carbon atom can be removed with equal probabilities to give cis- and trans-isomers in a ratio close to unity.This mechanism explains adequately the composition of the products formed on Fe catalysts supported on a material devoid of SiO2. In the presence of silica gel, the probability of isomerisation of terminal alkenes increases, because this support contains Brùnsted acid sites.In this case, the scheme of formation of the cis- and trans-isomers can be represented as follows:71 The interaction of an acid site proton with the double bond of an alkene yields a carbenium ion. Elimination of a hydrogen atom from the adjacent carbon gives rise to a cis- or trans-alkene.This mechanism accounts for the predominance of the cis-isomers in the reaction products, since it is known that on acidic catalysts (which include SiO2) cis-isomers are formed predominantly.72 The secondary transformations of the hydrocarbons pro- duced from CO and H2 have been studied using the precipitated catalysts 14% Co/SiO2 (I) and 15% Co/SiO2+HZSM-5 (II).38 In addition, a two-stage process was carried out using two consec- utively connected reactors, one of which (the first stage) contained the 14% Co/SiO2 catalyst, while the other one (the second stage) contained HZSM-5.The tests were carried out at a pressure of 2.1 MPa, a temperature of 280 8C, and the ratio H2 :CO=1. The degree of conversion of CO on specimen I reduced at 350 8C reached 50%, and the content ofC5+ hydrocarbons in the carbon- containing products of the synthesis was 72%.The C5+ fraction consisted of alkanes (63%), alkenes (34%) and aromatic com- pounds (3%). In the case of the two-stage process with catalyst I in the first stage and the zeolite at the second stage, the yield of liquid hydrocarbons somewhat decreased (to 65%). However, the C2±C4 alkenes formed in the first stage (*8%) were converted almost completely on HZSM-5.As a consequence, the content of aromatics in the liquid products increased to 33%. The proportion of C3±C4 alkanes markedly increased (from 7% to 21%), that of methane remained virtually unchanged (14%) and the content of C5+ hydrocarbons diminished to 65%.When the 14% Co/SiO2 catalyst was mixed with HZSM-5 (specimen II) and the synthesis was carried out in one stage, the methane formation was enhanced (up to 25%), whereas the yield of C5+ hydrocarbons and the content of aromatics in them barely changed with respect to those attained in the two-stage process. This led to the assumption that methane is formed in a secondary reaction (apparently, hydro- genolysis of light alkanes) occurring on metallic cobalt.This assumption is supported by the fact that the yield of C2±C4 (CO)ads (CHxO)ads (O)ads+(C)ads a1-products a2-products. H2C CH CH2R H2C CH2 CH2R H+ 7H H+ 7H C C H3C H H R H3C CH CH2R H+ 7H H+ 7H H2C CH CH2R H3C CH CH2R + C C H3C H H R cis C C H3C H R H trans 7H 7H Catalytic synthesis of isoalkanes and aromatic hydrocarbons from CO and H2 947alkanes on catalyst II is lower than that in the two-stage process (15% and 21%, respectively).This type of side reaction proceeds to a noticeable extent only on catalyst II in which cobalt particles occur in tight contact with HZSM-5 zeolite. Hybrid catalysts RuNaY+SO2¡¦ 4 /ZrO2 consisting of a mechanical mixture of ion-exchange RuNaY zeolite and sulfi- dised zirconium oxide have been studied.73 It was found that under the synthesis conditions (0.1 MPa, 250 8C,H2 :CO=1 : 1), the propene formed initially on the Fischer ¡À Tropsch component (RuNaY) undergoes two subsequent transformations on the sulfidised zirconium oxide, which behaves as a superacid, namely, hydrogenation through hydrogen transfer to give propane and transformations giving C4+ alkanes.The formation of C4+ hydrocarbons from propene follows an oligomerisation ¡À crack- ing mechanism with hydrogen transfer.11 Propene is formed on the Fischer ¡À Tropsch component and is readsorbed on the acid sites of the hybrid catalyst where it oligomerises to give heavier products. The resulting C4 fragments can be converted into C4 and C5 surface groups, which, in turn, react with propene giving rise to C7 and C8 hydrocarbons, respectively.The probability of interaction with propene for the C5 fragments is higher than that for C4, since heavier hydro- carbons are more readily adsorbed and more difficultly des- orbed.74, 75 The formation of branched oligomeric intermediates during the conversion of alkenes on zeolites depends appreciably on the concentration and the strength of acid sites.76 A higher concen- tration and a higher strength of acid sites lead to more branched oligomeric alkene intermediates.Oligomers synthesised on strong acid sites form larger amounts of isoalkanes. The transformations of linear alkenes on acid sites of the catalysts can follow two routes:73 (1) alkenes undergo direct skeletal isomerisation to isoalkenes and subsequent hydrogena- tion via hydrogen transfer; (2) oligomers formed on acid sites decompose. The fact that the amount of isobutane produced is larger than its equilibrium amount 77 indicates that isoalkanes are mainly formed via reaction (2).It should be noted that hydrogen transfer is also catalysed by acid sites.78 ¡À 80 Upon the hydrogen transfer reactions, some of the alkene molecules are saturated with hydrogen at the expense of other alkene molecules and/or oligomers attached to the catalyst surface.Consequently, oligomers giving off hydrogen become more unsaturated. Polymerisation of these highly unsaturated fragments results in the formation of coke precursors.73 The presence of the metallic component in the hybrid catalyst stabilises its activity towards isomerisation of hydrocar- bons.75, 81, 82 This can be explained in the following way.Molec- ular hydrogen is adsorbed on the metallic phase of the catalyst and undergoes homolytic dissociation. The atomic hydrogen thus formed spills over to the support, migrates to a Lewis acid site and then gives off an electron, being thus converted into H+.The interaction of the proton with oxygen located near the L-site affords a Br��nsted site, which is also capable of participating in acid-catalysed reactions. This is accompanied by weakening of the L-site which has accepted the electron. Thus, the catalyst is deactivated to a lesser degree. Another possible explanation is neutralisation of carbocations on the catalytic surface by hydride ions, which result from heterolytic dissociation of hydrogen on the metal.73 The appearance of hydride ions decreases the lifetime of intermediates and prevents their subsequent transformations.The compositions of the products obtained upon cracking of oct-1-ene on DAHY zeolite and upon the Fischer ¡À Tropsch synthesis on the 2% Ru¡À 2% Pt/DAHY bifunctional catalyst were found to be similar.83 However, the products of catalytic cracking contain larger amounts of branched alkanes in the C4 and C5 fractions (especially in the initial stage of the reaction).This is explained by the increase in the lifetime of a carbenium ion due to the low reactivity of hydrogen in the absence of a metal, which results in a higher probability of isomerisation of the carbenium ion before it is desorbed.The presence of a metallic component in the catalyst favours hydrogenation of alkenes to alkanes;84, 85 this promotes the formation of large amounts of linear saturated hydrocarbons and accelerates the spillover of hydrogen to acid sites. Therefore, to equalise the functions of hybrid catalysts, it is necessary that they contain similar numbers of metallic and acid sites.The distribution of the Fischer ¡À Tropsch synthesis products formed on the 2% Ru ¡À 2% Pt/zeolite catalyst shifts towards lighter hydrocarbons in the series NaY < HY < DAHY.83 The number of strong acid sites of the zeolites determined by the method of temperature-programmed desorption of NH3 (Tmax> 350 8C) increases in the same sequence.Among the products of the synthesis catalysed by the DAHY-containing material, mono- and di-branched alkanes predominate. This is due to the fact that relatively strong acid sites are unsuitable for skeletal isomer- isation.86 Under conditions of the Fischer ¡À Tropsch synthesis, the growing linear chain is chemically bound to the surface.This bond ruptures upon elimination of b-hydrogen from the hydro- carbon yielding a terminal alkene or upon a-hydrogenation giving rise to a linear alkane.87 The latter is inert under the Fischer ¡À Tropsch reaction conditions; 88 therefore, it does not participate in secondary transformations. Meanwhile, terminal alkenes can be partly hydrogenated or involved once again in the hydrocarbon chain growth after readsorption.89, 90 Upon the repeated adsorp- tion of alkenes, the possibility of chain termination as a result of hydrogenation decreases, resulting in a higher probability of chain growth and a greater alkane : alkene ratio.62, 91 The probability of secondary reactions involving terminal alkenes substantially increases with an increase in the number of carbon atoms in the molecule.92, 93 Correspondingly, the alkane : alkene ratio 94, 95 and the probability of chain growth 62, 95 also increase.The probability of secondary transformations of alkenes is normally assumed to be proportional to the contact time. Some investigators 63, 93 believe that the contact time increases appreciably as the chain becomes longer, because hydro- carbons with longer chains are better soluble in heavy products of the Fischer ¡À Tropsch synthesis, which fill the catalyst pores.However, the dependence of the solubility on the chain length cannot be determined or calculated; therefore, this parameter is difficult to use as a quantitative characteristic. In addition, this interpretation does not seem to be fully justified because the presence of a liquid phase does not influence the chemical potential and, hence, the secondary transformations.62 The dependence of the chain length on the residence time might also be due to transport limitations (i.e.diffusion). When there are no transport limitations, the components of the gas and liquid phases of the reaction system are in thermodynamic equilibrium.There- fore, in this case, neither the residence time of the products in the reaction area nor the kinetic driving force (volatility or chemical potential) depend on the presence of the liquid phase.62 The secondary transformations of hydrocarbons are not influenced by the liquid phase either. When there are external-diffusion limitations, the number of carbon atoms in the molecules being formed depends on the factor of the hydrocarbon chain growth a and on the concentration of alkenes. In this case, the residence time of hydrocarbons in the pores filled with a liquid increases and, hence, the probability of secondary transformations also increases.An attempt has been made 88 to elucidate the reason for the dependence of the rate of secondary reactions on the hydrocarbon chain length using cobalt foil as a model catalyst.Diffusion C3H6 [C9H19]+ [C3H7]+ [C6H13]+ [C9H19]+ , C4H8 + [C5H11]+ [C8H17]+ C4H8 + [C4H9]+ [C3H7]++C5H10 . 948 A L Lapidus, A Yu Krylovalimitations were modelled using a foil coated preliminarily by wax. When a non-coated foil was used, there were no such limitations. Below we consider the results of this study in more detail.The ratio of the rates of hydrogenation and repeated involve- ment of readsorbed alkenes in the chain is determined by a large number of parameters, in particular, CO and H2O pressures.62 Under the conditions studied (0.1 MPa, 220 8C, H2 :CO = 1, polycrystalline cobalt), virtually all readsorbed alkenes should be hydrogenated. Thus, the distribution of hydrocarbons up to C14 cannot deviate from the Schulz ± Flory distribution, i.e.it cannot depend on secondary transformations. The chemisorbed inter- mediates which undergo hydrogenation differ from thosvolved once again in the chain growth. To be repeatedly involved in the synthesis, an alkene should be chemisorbed on a vacant site on the catalyst surface. Under the conditions of the Fischer ± Tropsch synthesis, vacant sites are scarce, because the surface is occupied almost completely by various fragments.95 In order to occupy a growth site which has become vacant, an alkene should compete with CO and H2.It is unlikely that an alkene molecule that has moved far away from a vacant active site would reach it before it is occupied again. Of all the alkenes present in the reactor, only those located in the close vicinity of a site have a chance to be readsorbed.Thus, the probability of the repeated involvement of alkenes in the chain is intimately connected with their concentration at the interface in the physically adsorbed layer and with their average concentration over the whole reactor (i.e. the contact time). The alkene concentration at the interface is influenced, in addition to the transport limitations, by the preferred physical adsorption and solubility.Chemisorption of alkenes is a non-activated process. On passing from a physically adsorbed state to chemisorption, there is no competition between the reactions. Thus, the rate of readsorption is controlled only by the effective `rate of collisions', which depends on the concentration of the molecules in the physical adsorption region at the interface, i.e.on the total contact time in the physically adsorbed layer and on the alkene concen- tration in wax. The chemisorbed phase occurs in a steady state; however, there is no equilibrium between the physically and chemically adsorbed phases. As a consequence, at low degrees of CO conversion on the Co foil, the alkane/alkene ratio grows exponentially as the chain is lengthened.The coating of the foil with wax results in lower yields of alkanes and a less pronounced dependence of the alkane/alkene ratio on the chain length. These findings can be explained by the fact that the rate of readsorption of alkenes increases exponen- tially with chain length.A model was proposed which assumes competitive participation of alkenes in the readsorption, so that the readsorption rate is proportional to the alkene concentration at the catalyst ± wax interface. Thus, the dependence of the alkane/ alkene ratio on the chain length is determined by the solubility of hydrocarbons and by diffusion limitations; the latter, in turn, are caused by the predominant physical adsorption in the boundary layer.Under normal synthesis conditions (when the degree of CO conversion is high), the transport limitations are more significant. IX. Conclusion The foregoing leads to a number of general conclusions concern- ing the current state of the problem of catalytic synthesis of isoalkanes and aromatic hydrocarbons from CO and H2. The transformation of CO and H2 on oxide catalysts giving mainly isoalkanes (`iso-synthesis') normally requires high pres- sures.96 In this case, the yield of liquid hydrocarbons is low, and the overall catalyst activity is not sufficiently high.A more promising method is the synthesis of isoalkanes as mixtures with linear alkanes on catalysts containing high-silica zeolites together with Group VIII metals.Apparently, either intermediates or alkenes which undergo skeletal isomerisation serve as the source of isoalkanes. Aromatic hydrocarbons can also be formed on these catalysts but this requires higher temperatures. The zeolite component of the catalyst ensures the occurrence of isomerisation and aromatisation processes. Isomerisation occurs on the Brùnsted and Lewis acid sites.Isomerisation of alkenes is more efficient on bifunctional metal-containing cata- lysts. Thus, the transformation of CO and H2 on the Co ± MgO/ HZVM catalyst affords hydrocarbons (yield 70% ± 90%) con- taining 36%± 40% of isoalkanes.40 The `two-sites' model for the Co catalysts for the Fischer ± Tropsch synthesis that we proposed previously makes it possible to divide all the active sites on the catalyst surface into three main types, namely, methanation sites (Co0), hydrocarbon chain growth sites (Cod+±MxOy) and isomerisation and cyclisation sites (acid sites). The deviation of the product composition from the Shulz ± Flory distribution is due to the repeated adsorption of alkenes on one or several active sites. Readsorption of alkenes and their involvement in growing chains results in an appreciably decreased total rate of hydrocarbon chain termination and in a change in the step of hydrogen elimination giving terminal alkenes.Thus, read- sorption of alkenes followed by their skeletal isomerisation and hydrogenation ensure the formation of isoalkanes. Yet another pathway to isoalkanes is decomposition of unsaturated branched oligomers on acid sites and subsequent hydrogenation. Of course, it is as yet impossible to draw ultimate conclusions about the true mechanism of the formation of isoalkanes and aromatic hydrocarbons under the Fischer ± Tropsch synthesis conditions.However, in all probability, a fairly important role in the processes catalysed by cobalt-containing catalytic systems is played by unsaturated intermediates and by their subsequent readsorption with involvement in the growing hydrocarbon chain or their oligomerisation and isomerisation.References 1. M Belloum, Ch Travers, J P Bournonville Rev. Inst. Fr. Pet. 46 89 (1991) 2. H Pichler, K H Ziesecke Brennstoff-Chemie 30 13; 60; 81 (1949) 3. A N Bashkirov, S M Loktev Izv. Akad. Nauk SSSR, Otd. Tekhn.Nauk 147 (1954) 4. BRD P. 903 572; Ref. Zh. Khim 53 350 (1955) 5. Jpn. P. 7 483 702; Chem. Abstr. 82 101 068 (1975) 6. J A Rabo (Ed.) Zeolite Chemistry and Catalysis Vol. 2 (Washington, DC: Monograph. American Chemical Society, 1976) 7. Al A Petrov Kataliticheskaya Izomerizatsiya Uglevodorodov (Catalytic Isomerisation of Hydrocarbons) (Moscow: Izd. Akad. Nauk SSSR, 1960) 215 p. 8. V U S Rao, R J Gormley Catal. Today 6 207 (1990) 9. H Van Bekkum, E M Flanigen, J C Jansen Stud. Surf. Sci. Catal. 58 455 (1991) 10. J A Martens, P A Jacobs, J Weitkamp Appl. Catal. 20 239 (1986) 11. C W R Engelen, J P Wolthuizen, J H C van Hoff Appl. Catal. 19 153 (1985) 12. P A Jacobs Carboniogenic Activity of Zeolites (Amsterdam: Elsevier, 1977) 13. K G Ione Polifunktsional'nyi Kataliz na Tseolitakh (Polyfunctional Catalysis on Zeolites) (Novosibirsk: Nauka, 1982) 269 p. 14. J Valyon, J Mihalyfi, J A Jacobs, in Proceedings of the Workshop on Adsorption (Abstracts of Reports), Berlin, 1979 p. 15 15. J M Stensel, V U S Rao, J R Diehl, K H Rhee, A G Dhere, R J De Angelis J. Catal. 84 109 (1983) 16. C D Chang, W H Lang, A J Silvestri J. Catal. 56 268 (1979) 17. P D Caesar, J A Brennan,W E Carwood, J Cirie J.Catal. 56 274 (1979) 18. J Barry, L A Lay J. Phys. Chem. Sol. 29 1995 (1968) 19. G V Anthoshin, K M Minachev, R N Sevastjanov, in The 2nd International Conference on Molecular Sieves (Abstracts of Reports), Worchester, 1970 p. 776 20. L Rickert Ber. Bunsenges. Phys. Chem. 73 331 (1969) 21. J A Rabo, C L Angell, P H Kasai, V Schomaker Discuss.Faraday Soc. 41 328 (1966) Catalytic synthesis of isoalkanes and aromatic hydrocarbons from CO and H2 94922. Kh M Minachev, G V Antoshin, E S Shpiro Metalloézika (60) 55 (1975) 23. H Bremer, K H Bager, F Vogt Z. Chem. 14 (5) 199 (1974) 24. Kh M Minachev, A L Lapidus, in XII Mendeleevskii S'ezd po Obshchei i Prikladnoi Khimii (Tez. Dokl.) [The XIIth Mendeleev Congress on General and Applied Chemistry (Abstracts of Reports)] (Moscow: Nauka, 1981) p. 194 25. Kh M Minachev, A L Lapidus, A Yu Krylova Khim. Tv. Topl. 6 5 (1988) 26. A L Lapidus, I V Guseva, Kh M Minachev, Ya T Eidus Khim. Tv. Topl. 6 86 (1977) 27. A L Lapidus, V I Mashinskii, Ya I Isakov, Kh M Minachev Izv. Akad. Nauk SSSR, Ser. Khim. 2694 (1978) a 28. R T Obermyer, L M Melay, L S Lo, M Oskoole-Tabrizi, V U S - Rao J.Appl. Phys 53 2683 (1982) 29. R J Gormley, V U S Rao, D J Fouth, in The 7th North American Meeting of Catalytic Society (Abstracts of Reports), Boston, 1981 p. 23 30. A G Dhere, R J Angelis J. Catal. 81 464 (1983) 31. N E Varivonchik, A Yu Krylova, Kh M Minachev, A L Lapidus, Khoang Chong Iem, in Khimicheskie Sintezy na Osnove Odnougle- rodnykh Molekul (Tez.Dokl.) [Chemical Syntheses Based on One-Carbon Molecules (Abstracts of Reports)] (Moscow: Nauka, 1987) p. 3 32. A Shami, V U S Rao, R J Gormley, R T Obermyer, R R Schehl, J M Stencel Ind. Eng. Chem., Prod. Res. Dev. 23 513 (1984) 33. A Yu Krylova, T M Tsertsvadze, E L Berman, A L Lapidus Khim. Tv. Topl. 5 75 (1988) 34. A L Lapidus, A Yu Krylova, N E Varivonchik, V M Kapustin, Khoang Chong Iem Neftekhimiya 25 640 (1985) 35.A L Lapidus, A Yu Krylova, B V Kuklin, A Zukal, I Ratkhouski, I Starek Izv. Akad. Nauk, Ser. Khim. 342 (1995) a 36. A N Murty, A A Williams, R T Obermyer, V U S Rao, R J Gormley Stud. Surf. Sci. Catal. 38 73 (1987) 37. R L Varma, N N Bakshi, J F Mathews, S H Ng Can. J. Chem. Eng. 63 612 (1985) 38. R J Gormley, V U S Rao, R R Anderson, R R Schele, R D H Chi J.Catal. 33 193 (1988) 39. H W Pennline, S S Pollack Ind. Eng. Chem. Prod. Res. Dev. 25 11 (1986) 40. R G Copperthwaite, G J Hutchings, P Johnston, S W Orchard J. Chem. Soc., Faraday Trans. 1 82 1007 (1986) 41. T Tatsumi, Y G Shul, T Sugiura, H Tominaga Appl. Catal. 21 119 (1986) 42. Yu B Yan, K V Kosygina, I N Nikiforova, L P Skorobogotava, B K Nefedov, F K Shmidt Sovremennye Protsessy Pererabotki Uglya i Fiziko-Khimicheskie Metody Issledovaniya (Tez.Dokl. Vsesoyuzn. Konf.), Irkutsk, 1982 [Modern Processes of Carbon Processing and Physicochemical Methods of Investigation (Abstracts of Reports of the All-Union Conference), Irkutsk, 1982] p. 79 43. K V Kosygina, I N Nikiforova, Yu B Yan, F K Shmidt, B K Nefedov, L D Konoval'chikov, L P Skorobogatova Neftekhimiya 24 389 (1984) 44. Y W Chen, H T Wang, J G Goodwin J.Catal. 85 499 (1984) 45. D L King J. Catal. 51 51 (1978) 46. A Lapidus, A Krylova, J Rathovsky, A Zukal,M Janchalkova Appl. Catal. 80 1 (1992) 47. V H Lee, C H Bartholomew J. Catal. 120 256 (1989) 48. B Sen, J S Falkoner J. Catal. 122 68 (1990) 49. P G Gugla, K M Bailey, J S Falkoner J. Catal. 115 24 (1989) 50. R Burch, A R Flambard J. Catal. 86 384 (1982) 51. A L Lapidus Izv. Akad. Nauk SSSR, Ser. Khim. 2681 (1991) a 52. W M H Sachtler, M J Ichikawa J. Phys. Chem. 90 4752 (1986) 53. W M H Sachtler, D F Shriver,W B Ollenberg, A F Lang J. Catal. 106 401 (1987) 54. A B Boffa, C Lin, A T Bell, G A Somorjai Catal. Lett. 27 243 (1987) 55. A L Lapidus, A Yu Krylova, M S Kharson, V M Kogan, L V Sineva Izv. Akad. Nauk, Ser. Khim. 396 (1994) a 56. G Henrici-Olive', S Olive' The Chemistry of the Catalyzed Hydrogena- tion of Carbon Monoxide (Berlin: Springer, 1984) 57. R J Madon,W F Taylor J. Catal. 69 32 (1981) 58. G A Huff, C N Satterfield, J. Catal. 85 370 (1981) 59. L Konig, J Gaube Chem. Ing. Tech. 55 14 (1983) 60. L-M Tau, H Dabbagh, S Rao, B H Davis Catal. Lett. 7 127 (1990) 61. L-M Tau, H Dabbagh, S Rao, J Halasz, B H Davis Catalysis 87 61 (1988) 62. E Iglesia, S C Reyer, R J Madon J. Catal. 129 39 1988) 63. H Pichler, H Schulz, M Elstner Brennstoff-Chemie 48 78 (1988) 64. S Novak, R J Madon, H Shulz J. Catal. 77 141 (1982) 65. N O Egiebor, K R Ungar, B W Wojciechowski Can. J. Chem. Eng. 62 432 (1982) 66. N O Egiebor, W C Cooper Appl. Catal. 17 47 (1985) 67. P W West, G L Haller, R L Burwell J. Catal. 29 486 (1973) 68. E A Naragon Ind. Eng. Chem. 42 2490 (1950) 69. A N Ko, B W Wojciechowski Int. J. Chem. Kinet. 15 1249 (1983) 70. J A Baker, A T Bell J. Catal. 78 165 (1982) 71. W C Whitmore Chem. Eng. News 26 668 (1947) 72. J E German Catalytic Conversion of Hydrocarbons (London: Academic Press, 1969) p. 144 73. X Song, A Sayari Appl. Catal., A 110 121 (1994) 74. R A Finn, O A Larson, H Beuther Ind. Eng. Chem. 52 153 (1960) 75. M Y Wen, I Wender, Y W Thierney Energy Fuels 4 372 (1990) 76. J Datka Stud. Surf. Sci. Catal. 5 121 (1980) 77. R Oukaci, J C S Wu, J G Goodwin J. Catal. 107 471 (1987) 78. P B Venuto, L A Hamilton, P S Landis J. Catal. 5 484 (1966) 79. B E Langner J. Catal. 65 416 (1980) 80. F E Shephard, J J Rooney, C Kemball J. Catal. 1 379 (1962) 81. T Hosoi, T Shimidzu, S Iton, S Baba, H Takasoka, T Imai, N Yokoyama Prepr. Am. Chem. Soc., Div. Pet. Chem. 33 562 (1988) 82. K Ebitani, J Konishi, H Hattori J. Catal. 130 257 (1991) 83. Y-G Shul, Y Arai, T Tatsumi, H Tominaga Bull. Chem. Soc. Jpn. 60 2335 (1987) 84. K Fujimoto, M Adachi, H Tominaga Chem. Lett. 783 (1985) 85. B K Hodnett, B Delmon Catalytic Hydrogenation (Ed. L Cerveny) (Amsterdam: Elsevier, 1986) p. 53 86. J Abbot, B W Wojciechovski Ind. Eng. Chem., Prod. Res. Dev. 24 501 (1985) 87. E F Herrington Chem. Ind. 347 (1985) 88. E W Kuipers, I H Vinkenburg, H Oosterbeek J. Catal. 152 137 (1995) 89. W K Hall, R J Kokes, P H Emmett J. Am. Chem. Soc. 82 1027 (1960) 90. H Schulz, B R Rao,M Elstner Erdoel Kohle, Erdgas, Pertochem. Brennst. Chem. 23 651 (1970) 91. R J Madon, E Iglesia J. Catal. 139 576 (1993) 92. D Van Hove Proc. Int. Meet. Soc. Fr. Chim., Div. Chim. Phys. 147 (1993) 93. L M Tau, H A Dabbadh, B H Davis Energy Fuels 4 94 (1990) 94. H Pichler, H Schulz, F Hojabri Brennstoff-Chemie 44 5 (1963) 95. C A Mims, L E McCandlish J. Phys. Chem. 91 929 (1963) 96. Yu B Yan, B K Nefedov Sintezy na Osnove Oksidov Ugleroda (The Syntheses Based on Carbon Oxides) (Moscow: Khimiya, 1987) 263 p. a�Russ. Chem. Bull. (Engl. Transl.) 950 A L Lapidus, A Yu Krylo

ISSN:0036-021X    

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Abstract. The results of studies on the structures and mechanical properties of polymeric materials subjected to isostatic treatment and extrusion in the solid state, mainly rigid-chain and network polymers, polymer blends, and reinforced compositions, are described systematically and surveyed. The bibliography includes 204 references. I. Introduction The modification of polymers, the manufacture of which has been undertaken on an industrial scale, is one of the principal trends in the creation of materials with a required set of characteristics. Various procedures for the deliberate alteration of the structures and properties, including methods such as chemical and physical modification, are employed.Despite the advances achieved on these lines, the search for new modification methods, among which methods involving treatment by the application of pressure are promising, is being continued.A high hydrostatic pressure induces both reversible and irreversible changes in the physical and mechanical properties of materials, promotes the formation of their new modifications, and makes it possible to achieve the synthesis of products the prepa- ration of which is impossible under normal conditions.1± 7 The first experiments using high pressure were carried out more than 100 years ago but since then the interest in its employment for scientific and technological purposes has not weakened. In order to obtain materials with the required performance properties, various methods based on the employment of high pressures, are used namely pressing, rolling, stamping, drawing, etc.They are distinguished by the scheme underlying the stressed state of the material and the hydrostatic component of the stress tensor.8 Among the technological processes based on the direct application of a high hydrostatic pressure, the greatest develop- ment has been achieved in isostatic pressing and hydrostatic extrusion.7, 9±17 The isostatic pressing method is used basically in the treatment of powders.Its advantages include extremely effective mechanical characteristics of the final products, the possibility of obtaining articles of any shape, however complex, the simplicity of equipment, and the ease of automation. Virtually all the advantages of the plastic deformation of materials, includ- ing the most favourable scheme underlying the stressed state, the reduction to a minimum of frictional forces, and the plasticisation effect determined by the application of pressure, are applied in the hydroextrusion method.Using the advantages of the method listed above, it is possible to alter the structure and the strain ± - strength characteristics of the treated material. Compared with other substances, for example metals, poly- mers are much more sensitive to the effect of pressure, but the first few studies which considered various aspects of the influence of high pressure on polymers were carried out comparatively recently.18 ± 22 Active interest in the study of the properties of polymers under pressure began to be manifested in the 1960s while the fundamental theoretical and experimental results, which clarify the role of high pressures in the establishment of a set of mechanical and performance properties of polymeric materials, were obtained in the 1970s ± 1980s, which may be regarded as the period of the most vigorous development of this field.23 ± 30 The methods for the treatment of polymers by the application of pressure, used for their structural modification, can be divided conventionally into two groups.The first comprises various procedures based on the application of an isostatic treatment; isostatic pressing of polymer powders, treatment by the hydro- static compression of articles in the form of a block, and structure formation under pressure. This group of methods is characterised by a spherical external stress tensor.The second group includes methods based on the joint operation of a high pressure and shear stresses (the total external stresses tensor): deformation using Bridgman anvils, ultrasound, and shock waves; pressing; elastic- deformation grinding; solid-state extrusion; etc. This review deals with structural modification methods, which are some of the most promising as regards practical employ- ment � different versions of the isostatic treatment and solid- state extrusion.They are characterised by a high effectiveness, ease of technological application, and the possibility of treating comparatively large volumes of the polymeric raw material. The main bulk of studies devoted to isostatic treatment or solid-state extrusion have been carried out on flexible-chain amorphous or amorphous-crystalline homopolymers.Their results have been reflected to a large extent in a series of review articles and monographs, for example in the studies described in Refs 25, 26, 30, and 31. In the present review, attention is therefore V A Beloshenko, V N Varyukhin A A Galkin Donetsk Physicotechnical Institute, National Academy of Sciences of Ukraine, ul.R Lyuksemburg 72, 340114 Donetsk, Ukraine. Fax (38-062) 255 02 08. E-mail: bel@hpress.dipt.donetsk.ua (V A Beloshenko) A A Askadskii ANNesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, ul. Vavilova 28, 117813 Moscow, Russian Federation. Fax (7-095) 135 50 85. Tel. (7-095) 135 93 98. E-mail: andrey@ineos.ac.ru Received 3 March 1998 Uspekhi Khimii 67 (11) 1044 ± 1067 (1998); translated by A K Grzybowski UDC 678.01 : 539.89 Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures V A Beloshenko, A A Askadskii, V N Varyukhin Contents I.Introduction 951 II. Isostatic treatment of crystallisable polymers 952 III. Isostatic treatment of glassy polymers 953 IV.Solid-state extrusion of polymers and polymer composites 960 V. Conclusion 970 Russian Chemical Reviews 67 (11) 951 ± 973 (1998) #1998 Russian Academy of Sciences and Turpion Ltdconcentrated on the less well known aspects of the problem, namely on questions concerning the modification of rigid-chain and network polymers and polymer composites. New aspects of the development of the structural modification methods under discussion, in particular the solid-state extrusion of a powdered billet, are examined in greater detail.II. Isostatic treatment of crystallisable polymers 1. Crystallisation under pressure The vast majority of studies on isostatic treatment have been devoted to crystallisation under pressure. The main bulk of such investigations have been carried out on polyethylene (PE).25, 32 ± 39 In the first experiments, it was already observed that pressure not only increases the crystallisation temperature but also induces certain structural changes in the polymer.Thus Matsuoka and Maxwell 32 obtained at high pressures PE with an unusually high density, which they explained by more effective chain packing.Subsequently, Wunderlich 40 showed that crystallisation under a high pressure entails not only more satisfactory chain packing but also leads to the formation of crystals of a new type�crystallites with extended chains (CEC). The pressure corresponding to the onset of the formation of CEC is in the range 300 ± 350 MPa according to the phase diagram.41 There are two main hypotheses concerning the appearance of CEC.34 According to the first, the CEC are formed on nuclei consisting of straightened chains or packs of chains,35 whilst according to the second, folded lamellae are thickened under pressure and are gradually converted into CEC.36, 37 The molecular mass distribution (MMD) has a significant effect on the course of crystallisation. For example, it has been shown 38 that the curves for the distribution of the thickness (length) of crystallites in the direction of molecular chains and of the molecular mass (M) distribution are identical forM<10 000. For M>10 000, mixed crystallites are formed.Fractions with narrow MMD crystallise in two stages,42 while the crystallisation of specimens with a broadMMDusually involves a single stage.In mixtures of fractions, fractionation and segregation take place durcrystallisation, which results in the formation of a eutectic- like mixture of crystallites comprising fully straightened molecules with a lowMand crystallites made up of molecules with a highM having both straightened and folded shapes.39 The principal differences in the morphology of PE specimens, obtained at atmospheric and high pressures, are as follows. In the pressure-modified PE, the lamellar thickening occurs only up to a certain limiting extent, the twisting of the lamellae is not observed, and spherulitic formations with a fanlike structure are present.The latter are characterised by a tangential orientation of the C-axes. The formation of CEC on crystallisation from the melt under pressure has been observed for a series of other polymers 43, 44 and copolymers.45 Together with the formation of CEC, phase trans- formations occur in many polymers during crystallisation under pressure, as a result of which it is possible to obtain materials with a new crystal system.For example, it has been noted 23 that, together with the orthorhombic phase, a triclinic phase is present in PE crystallised at 436 MPa.If the crystallisation of syndiotactic polypropylene (PP) is carried out at pressures amounting to several hundreds of megapascals, it is possible to obtain a new crystalline modification.46 Poly(vinylidene fluoride) (PVDF), crystallised at 507 MPa, consists of a mixture of two phases with different melting points (Tm); the PVDF obtained at atmospheric pressure consists of a single phase.47 The transition to a new modification on crystallisation under high pressure has been noted also for polybut-1-ene.25 The study of the crystallisation kinetics showed that a high pressure as a rule slows down the crystallisation process, i.e. a higher degree of supercooling of the melt is required in order to attain the same rate of crystallisation as at atmospheric pres- sure.48, 49 At the same time a nonmonotonic dependence of the rate of crystallisation of PE on pressure, due to changes in the kinetics of the growth of crystals and the phenomenon of top- omorphy has also been discovered.50 The structural transformations listed above, which occur in polymers as a result of crystallisation under pressure, induce changes in their properties.As a rule, denser packing appears in the polymers and the degree of crystallinity and Tm measured under atmospheric conditions increase.32, 40 Thus the differences between the densities (r) of high-density polyethylene (HDPE) specimens, obtained at normal and elevated pressures, amount to 2%.32 The dynamic Young's modulus of pressure-modified poly- ethylene is twice as high, its strength is greater, but its plasticity is lower.A similar effect has been obtained for polytetrafluoro- ethylene (PTFE) and polypropylene.25 The increase in the density of PE and the formation of CEC on crystallisation under pressure cause a decrease in its compressibility and in the intensity of g-relaxation.35 2.Annealing under pressure The influence of annealing under pressure has also been inves- tigated mainly in relation to PE.25 It was established that, when PE specimens comprising crystals with folded chains (CFC) are annealed, the rearrangement of the structure with CFC to a structure containing CEC is observed in the corresponding temperature ranges.51 This has been confirmed by the appearance of a second (high-temperature) fusion peak and the results of morphological investigations; lamellae with transverse dimen- sions exceeding by more than an order of magnitude the dimen- sions of the lamellae in the unannealed specimens are formed after such annealing. There is a simultaneous increase in r for the annealed material, which reaches values characteristic of the structure with CEC.Slight changes in Tm, r, and morphology are observed during the annealing of specimens with the CEC structure, indicating a further improvement of their structural arrangement. It is assumed 40 that the formation of CEC during the annealing of PE under a high pressure takes place in accord- ance with a mechanism similar to the mechanism of crystallisation under the conditions of hydrostatic compression, i.e. via the isothermal thickening of lamellae.Similar features (an increase in the degree of crystallinity, an improvement in the structure, and an increase in Tm and r) have also been established in the course of the annealing of other amorphous-crystalline isotropic polymer specimens.25, 46 During annealing under pressure, oriented polymers exhibit a more complex behaviour.For example, uniaxially extended HDPE specimens undergo a phase transformation in addition to the increase in the thickness of the crystallites, in Tm, and in r already mentioned.52 A proportion of the material undergoes the tran- sition from the orthorhombic modification to the oriented triclinic modification. At the same time, there is a sharp increase in the degree of crystallinity: the intensity of the amorphous halo for the annealed specimen is five times smaller than for the original specimen.52 In the course of the annealing of oriented specimens of syndiotactic PP, the formation of a new high-pressure crystalline modification was observed.46 The annealing of oriented PVDF specimens also induces a phase transition in the solid state with retention of the primary orientation.46 The phase formed then has r higher by 10% compared with r for the low-pressure phase.Annealing under pressure (by crystallisation under pressure) of isotropic and oriented polymer specimens is reflected in their mechanical properties. In particular, a significant increase in the moduli of elasticity has been noted.53, 54 An improvement in the properties of the polymers can be achieved in a number of instances also by isostatic treatment at room temperature.The imposition of an hydrostatic pressure onto PTFE specimens with an increased porosity leads to the `healing' of the micropores at pressures up to 80 MPa.55 On further increase in pressure, the `healing' process is retarded.Thermal activation analysis showed that the `healing' of the 952 V A Beloshenko, A A Askadskii, V N Varyukhinpores is associated with the viscous flow of the macromolecules. It continues until the free volume is `exhausted'. The contact between the polymer and the pressure-trans- mitting medium is very important in the isostatic treatment. In the absence of appropriate insulation, the negative effect of the medium can abolish the positive role of pressure or can impair the properties of the created material.56 3.Isostatic treatment of powders The component methods for the treatment of polymer powders are extrusion, pressing with subsequent free sintering, pressing with sintering under pressure, and casting under pressure.57 The choice of a particular procedure is determined both by the properties of the polymer and by the nature of the technical problem to be solved.The advantage of the pressing ¡À free sinter- ing scheme consists in the possibility of the effective manufacture of long articles by separating the pressing and sintering opera- tions. However, in order to obtain good-quality articles, it is necessary to achieve the highest possible density and uniformity of its distribution with respect to length and the cross-section of the billet.It has been shown in relation to PTFE of brands F-2M, F-4, F-4MB, F-40, and F-40-P 58 that the isostatic treatment makes it possible to obtain the above properties with retention of the specified form of the article, provided that this operation is carried out after preliminary pressing of the powder.Hydrostatic pres- sures of 200 ¡À 250 MPa are sufficient for the attainment of the state with the minimum porosity (Fig. 1). A mathematical model describing the isostatic pressing of polymer powders and permitting the determination of the values of the principal technological parameter (the pressure) has been proposed.59 The isostatic pressing process is regarded as consist- ing of two stages.The first stage is associated with the relative displacement and rotation of individual particles. The second involves elastic deformations of the particles filling the existing free space. After the removal of the external pressure, the polymeric material is loosened as a consequence of the reversi- bility of the elastic deformations and its density returns to the level attained at the end of the first stage.The relative density of the pressed polymer powder is therefore always less than unity. Comparison of the calculated and experimental r(P) relations for PTFE confirms that the model is consistent with the isostatic treatment under consideration.59 In the practical application of powder technologies, moulding methods, differing from isostatic pressing, for example pressing in a rigid die of complex shape or pressing in special shells are widely used.The technological problems arising under these conditions are associated with the determination of the force necessary to attain the specified density, with the estimation of the pressure exerted on the walls of the die, with the calculation of the distribution of density with respect to the volume of the billet, etc.The parameters of the model,59 found in experiments involv- ing isostatic treatment, may be used to calculate the force parameters of the processes indicated. III. Isostatic treatment of glassy polymers 1. The influence of pressure on the properties of amorphous polymers There have been extremely few studies devoted to the influence of the isostatic treatment on the structure and properties of amor- phous polymers.This is evidently associated with the possibilities for structural rearrangements which are much smaller than for crystalline polymers. Fairly detailed studies on polymeric glasses have been carried out.60 It has been shown for a series of amorphous polymers that cooling of the specimens at 150 MPa from a temperature exceeding the glass transition temperature (Tg) to room temperature makes it possible to increase r com- pared with the original state by *1%.The state with increased density is metastable, but relaxation under atmospheric condi- tions proceeds slowly. With increase in temperature, this process accelerates, but at T<Tg there is no complete transition to the original state.Compaction of the glass is characterised by increased values of the dynamic modulus of elasticity and a lower dielectric constant. Higher values of r for a polymer vitrified under pressure have also been established in another study.61 Here the instability of the material was noted. The nature of the recovery of r when the material is heated depends on the pressure employed: in the original stage of the recovery of r, the rate of the process is higher the higher the pressure P.There have been few technological studies devoted to the application of a high hydrostatic pressure for the structural modification of polymers. Among them, mention should be made primarily of Shturman's studies 62, 63 in which the favour- able effect of annealing under pressure on the mechanical and performance characteristics of a series of polymers and polymer composites was demonstrated. For example, annealing of impact- resistant polystyrene promotes an increase in its strength, impact strength, and antifriction properties.A similar thermomechanical treatment of glass-reinforced polyamide specimens significantly increased their hardness together with an increase in the ultimate bending strength and impact strength.It is believed 62 that the strengthening effect is due to the compaction of the surface layer and the appearance in the latter of compressive stresses which increase the cohesive strength of the polymers. The molecular mass of the polymers also increases under the influence of temper- ature and pressure (the content of low-molecular fractions dimin- ishes), which is correspondingly reflected in their properties. 2. The structures and properties of network polymers obtained under the conditions of hydrostatic compression The chemical reactions as a result of which network polymers are formed are accompanied by the formation of a three-dimensional network structure.Compared with linear glassy polymers, one may expect new effects in the course of their isostatic treatment, in particular irreversible changes in structure and properties induced by the high pressure. The study of the kinetics of the curing of network polymers under the conditions of hydrostatic compression showed that in the vast majority of cases the rate of this process increases with increase in pressure 64 ¡À 69 and the increase is extremely significant.For a series of epoxide compositions, the increase in the reaction rate constant (k) at 250 ¡À 350 MPa reaches several orders of magnitude.64, 67 The volume activation effect (DV=) for these reactions is negative and the increase in k with increase in pressure follows from the familiar relation.70 q ln k qP T�� ¡¦ DV6�� RT .r /g cm73 2.0 1.8 1.6 1.4 250 500 750 P /MPa 1 2 3 Figure 1. Pressure dependence of the density of polytetrafluoroethylenes for different powders: (1) F-4; (2) F-4MB; (3) F-2M. Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 953The nature of the influence of pressure is determined both by the type of the reagents selected and their proportions.For example, in the presence of an excess of an epoxide oligomer (EO) there is a possibility of the solvation of the reaction centres, which prevents their interaction, as a result of which the intensity of the effect of the pressure diminishes.69 The results of studies by different investigators 64 ± 69, 71 involving the determination of k agree well.The shortening of the gel and glass formation times during the curing of EO by amines under the conditions of hydrostatic compression has been observed.71 Under the influence of pressure, the structure-formation process may be retarded as well as being accelerated. This result has been established, for example, in the curing of EO by an amine complex of boron trifluoride 66, 67 and is associated with the influence of pressure on the equilibrium constant (K) for the dissociation of this complex to the amine (A) and BF3, initiating the polymerisation ABF3 A+BF3 .Under the influence of pressure, the equilibrium in this reaction shifts to the left and the concentration of BF3 in the reaction mixture falls, which means that the rate of curing of the EO diminishes.It has been established 65 ± 69 that the differences in the kinetics of the chemical reactions occurring at pressures from 0.1 to 250 MPa are quantitative. This makes it possible to postulate the absence of an appreciable influence of pressure in the selected range (0.1 ± 250 MPa) on the curing mechanism and hence on the molecular level of the structural arrangement of the epoxide polymers (EP).At the same time, the data obtained for poly(- ethylene glycol dimethacrylate)72 indicate that at sufficiently high pressures there is a possibility of changes in the structure of the molecular chains. Analysis of the IR spectra of poly(ethylene glycol dimethacrylate) showed that the nature of the disposition of the methacryloyl groups changes in the pressure range 600 ± 2000 MPa, which is manifested by an increase in the probability of their `tail to tail' addition.The most significant rearrangement of the molecular structure is observed at P4800 MPa. An increase in pressure hardly influences the type of addition of the polymethacrylate chain units. The influence of hydrostatic pressure on the topological and supermolecular structures of network polymers is more effective than its influence on the molecular level of the structural arrange- ment.Fig. 2 presents thermomechanical curves for specimens of a series of EP cured at different pressures.73 The following compo- sitions were employed: a blend of the epoxybisphenol A{ resin ED-20 and polyepichlorohydrin cured with monocyanoethylene- triamine (EP-1); epoxyaniline resin cured with a eutectic mixture of aromatic diamines (EP-2); epoxybisphenol A resin cured with bis(4-amino-3-chlorophenyl)methane (EP-3); the epoxybisphenol A resin ED-20 cured with methyltetrahydrophthalic e (EP-4); a blend of a flexible-chain block oligomer comprising an epoxide resin and acid oligomer and a diglycidyl ester cured with methyltetrahydrophthalic anhydride (EP-5); a blend of the epox- ybisphenol A resin ED-20 and oligomeric diglycidyl esters based on diethylene glycol cured with triethanolamine titanate (EP-6); the epoxybisphenol A resin ED-16 cured with the methyltetrahy- drophthalic anhydride (EP-7).Curing under pressure results in an increase in Tg and in the initial temperature of forced rubber elasticity and a decrease in the deformability of the polymers in the region of rubber elasticity.These facts indicate the intensification of intermolecular interac- tion. Pressure exerts the greatest influence on EP with large units (EP-4) or long chains between cross-links (EP-5). The increase in Tg reaches 30 K and the relative decrease in the quasiequilibrium elastic deformation reaches 30%.The latter is equivalent to the corresponding increase in the equilibrium modulus of elasticity, i.e. the increase in the rigidity of the polymer network. In the case of EP with compact units (EP-1), the changes under consideration are less pronounced. These findings indicate that the change in Tg under the influence of pressure is caused predominantly by the change in the packing of the units.The structure produced is `frozen' at T<Tg and is retained after the removal of the pressure. The appearance of anomalies on the thermomechanical curves for the modified EP-1 and EP-5 specimens in the vicinity of Tg is apparently a consequence of the relaxation of the `frozen' com- pressive stresses.The main parameter of the topological structure of the type of polymers considered is the concentration of the network cross- links (n). Table 1 presents the values of n for certain EP inves- tigated.74 For all the compositions, the curing under the condi- tions of hydrostatic compression promotes an increase in n and the highest values of Dn are attained for EP-4 and EP-5 (Dn was estimated as the ratio of the increment in n, caused by the pressure, to the value of n for the control specimen).The results of the thermomechanical analysis agree with pulse NMR data,68 according to which structure formation under pressure increases the `nuclear' glass transition temperature (Tg 0), corresponding to the temperature of the appearance of two `kinetic phases' with the nuclear magnetic transverse relaxa- tion times T2b and T2c, and diminishes molecular mobility (Fig. 3). At T>Tg 0, the changes in molecular mobility are observed only for the most mobile (T2b) fragments of the polymer network, which indicates the influence of pressure also on the linear chains between the cross-links. Differences between the internal friction spectra for the pressure-modified and control EP specimens have been demon- strated.75 This indicates a lower intensity of the small-scale relaxation processes and a narrower set of relaxation times corresponding to these processes.{ Bisphenol A (`dian' in Russian) is 2,2-bis(4-hydroxyphenyl)propane (Translator) e (%) 12 8 4 0 320 340 360 380 400 T /K 3 1 5 6 4 2 Figure 2. Thermomechanical curves for the epoxide polymers EP-4 (1, 2), EP-5 (3, 4), and EP-1 (5, 6) at different pressures (MPa): (1), (3), and (5) 0.1; (2), (4), and (6) 200.Table 1. Effective concentrations of network cross-links in EP.74 Polymer 10720 n /cm73 Dn (%) Polymer 10720 n /cm73 Dn (%) EP-1 7:3 7:8 7 EP-4 5:7 8:1 42 EP-3 12:1 13:9 15 EP-5 3:6 4:4 22 Note. The values of n above the horizontal line correspond to control specimens and those below the horizontal line were obtained at a pressure of 200 MPa. 954 V A Beloshenko, A A Askadskii, V N VaryukhinThree endothermic maxima appear on the differential scan- ning calorimetric (DSC) curves for the cured EP specimens (Fig. 4). They correspond to regions with different scales of the `unfreezing' of molecular mobility: the low-temperature maxi- mum I is associated with the `unfreezing' of the noncooperative motion of segments at sites with a lower packing density (b-tran- sition); the maximum II at 440 K corresponds to the cooperative glass a-transition; the high-temperature maximum III at 490 K is due to the molecular mobility in the rubbery state.76 Comparison of curves 1 and 2 in Fig. 4 shows that treatment with pressure leads to a decrease in the intensity of the maximum I and its shift towards higher temperatures (I 0 on curve 2), which is caused by the suppression of the small-scale molecular motion at the level of the segments (sections of the chain). The second maximum for the modified polymer is also located at higher temperatures (II 0 on curve 2).The intensity of the molecular motion does not then change.Thus the set of data obtained by thermomechanical analysis, pulse NMR, internal friction spectroscopy, and DSC makes it possible to assume that the curing of the EP under the conditions of high pressures induces appreciable changes at the topological and supermolecular levels of the structural arrangement, man- ifested by an increase in the effective density of the cross-links and intensification of the intermolecular interaction.This conclusion conflicts at first sight with the results of a study 77 in which the structure formation in EP was also inves- tigated under the conditions of high pressures. According to these results, the specimens cured at 500 ± 1000 MPa have lower values of Tg than those cured at atmospheric pressure.Estimation of the degree of cross-linking of the modified EP demonstrated that this process does not go to completion in the specimens indicated. Possibly the negative effect of pressure appeared because the investigators did not apply a correction for the temperature regime in the curing process. The change in Tg with increase in pressure is known to be 0.15 ± 0.3 K MPa71. By tradition, in the epoxide resin curing practice the final curing temperature for the composition is chosen so that it exceeds the expected glass transition temperature.79 In this case, the highest degree of conversion is attained.When the temperature regime correspond- ing to atmospheric pressure is applied in the region P>500 MPa, the specimen is vitrified, for the reason mentioned above, before it attains the necessary degree of conversion.The reaction is inhibited and an undesirable result appears as a consequence. There is also a possibility of another mechanism of the loosening of the polymer network on curing EP under pressure, which is analogous to that of Ozerkovskii et al.72 In considering the pressure dependence of n for PDMA, the authors of this investigation suggested that in the network polymer cross-linked to the limiting extent an equilibrium is established between the cross-links, leading to a decrease in the volume of the system and the loosening cross-links.Calculation showed that at high pres- sures n decreases, because, starting with a certain critical concen- tration, the cross-links prevent the packing of the network skeleton.In contrast to EP, for bis(methacryloyloxyethyleneoxycar- bonyloxyethylene) oxide (MECDE) polymerised at 85 ± 550 MPa, a decrease in the glass transition and enforced rubber elasticity temperatures as well as an increase in deformability in the rubbery state are observed with increase in pressure.79 Accord- ing to the authors,79 this is caused by the changes occurring at the molecular level in the structural arrangement of MECDE.On polymerisation under pressure, there is an increase in the proba- bility of the isotactic addition of units in the main macrochains of MECDE. Lower values of Tg are more characteristic of this modification than of the modification with the atactic type of addition of units. A microstructure, the morphological unit of which is a globule, is at the uppermost level in the `structural hierarchy' of network polymers 78, 80, 81 The globular character of the micro- structure is manifested either on the surface of a section after etching or of the break produced by the application of a load.In the first case, the weakest sections of the polymer network are etched away, i.e. the globular relief reflects the chemical rupture front, the velocity of which is inversely proportional to the molecular packing density.82 In the second case, the rupture takes place at the interglobular boundaries.83 This makes it possible to regard the globule as a local thickening of the three- dimensional network, manifested as a particle on etching or mechanical rupture.82 Comparison of the microstructures of control EP specimens and specimens cured at high pressures shows that the dimensions of the globular formations and the thicknesses of the boundaries separating them diminish under the influence of hydrostatic pressure and the structure becomes more uniform.62, 68, 84, 85 Table 2 presents the results of optical microscope studies of a series of epoxide compositions, illustrating this postulate.When the resolution of the microscope is high, it is seen that each globule consists of smaller formations. The results of electron microscope studies of the EP structure 84 are correlated with optical micro- scope data, which indicates a link between the observed morpho- logical formations. The more finely disperse structure of the EP obtained under a high pressure compared with the structure of the EP cured at atmospheric pressure is due to the fact that the acceleration of the structure formation processes entails the pc 1.0 0.8 0.6 0.4 320 370 420 470 T /K log(T2b /ms); log(T2c /ms) 5 4 3 2 5 6 1 2 3 4 Figure 3.Temperature dependences of the nuclear magnetic relaxation times T2b (1, 3) and T2c (2, 4) and of the proton population pc (5, 6) for EP-1 at different pressures (MPa): (1), (2), and (5) 0.1; (3), (4), and (6) 150. 300 400 500 T /K I 0 II 0 III 0 I II III 1 2 dH dt =1 mcal s71 Figure 4. DSC curves for EP-1 at different pressures (MPa): (1) 0.1; (2) 100. I ± III and I 0 ± III 0 �see text. Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 955formation of a larger number of centres with reacted reactive groups, which create a basis for the subsequent formation and establishment of the heterogeneous morphology of the EP.86 The influence of pressure on the microstructure of MECDE is analogous to its influence on the structure of EP.Electron micro- scope studies 79 have shown that polymerisation under the con- ditions of hydrostatic compression is responsible for the decrease in the size of the globules from 500 A (at 0.1 MPa) to 300 ± 350 A.Increasingly more pronounced aggregates with an elongated shape, consisting of a set of globules arranged one on top of another, are formed with increase in pressure. An interesting result has been obtained in a study 87 devoted to the structure of a phenol ± furfural composition based on iron tris(hydroxymethyl)phenoxide (Fe-TMP).In the original compo- sition, the Fe-TMP forms ionic clusters which coagulate in the course of structure formation in the system with formation of associated species. High pressure promotes the latter process and leads to the appearance of larger metal-containing particles. At the same time pressure influences the physicochemical processes occurring in these particles during the curing of the composition, which alters their phase structure from amorphous to crystalline.The increase in the size of the cluster is accompanied by the appearance of a large amount of magnetite and an increase in the degree of order in its structure. The pressure-induced changes at different levels of the struc- tural arrangement of network polymers are reflected in their mechanical properties.In the first place, one should point to the increase in density, which has been noted in numerous investiga- tions (see, for example, Refs 75, 79, and 85). This can be explained by the fact that pressure prevents the formation of macroscopic defects in the form of pores or cracks and promotes a decrease in free volume as a result of a more effective packing of the polymer.In the region of comparatively low pressures, the r(P) relation is linear, but, as P increases, it approaches asymptotically a certain limiting value.79 Table 3 lists the strain ± strength characteristics of MECDE specimens at different pressures, established in compressive tests.79 Evidently, with increase in pressure, the modulus of elasticity (E) and the ultimate tensile strength (st) increase to certain values.In the range 425 ± 500 MPa, these parameters diminish. The extremum change in mechanical properties may be caused by the competition between the changes at the microlevels (r) and topological levels (n) of the structure and at the level of the macrochains.At a comparatively low pressure, effects associated with the increase in r and n predominate. The mechanism of the addition of methacryloyl groups predominates in the region5400 MPa.72 The results of compressive tests on EP specimens cured at atmospheric and elevated pressures agree 74, 85 with the data of the above study.79 The more effective mechanical properties of the modified EP can be accounted for by the intensification of the intermolecular interaction and a decrease in structural inhomoge- neity on curing under the conditions of hydrostatic compression.The most significant changes in properties are attained when the epoxide compositions deviate greatly from stoichiometry, where- upon the number of unreacted epoxide groups increases signifi- cantly.The intermolecular interactions at the boundary between the globular and interglobular phases are appreciably weakened under these conditions with a consequent sharp decrease in the strength characteristics of the control specimens. The density of the cross-links in the polymer increases under the influence of pressure (the degree of inhomogeneity diminishes) and the con- nectivity between the supermolecular formations is enhanced, which promotes the retention of fairly high strengths.The data obtained in tensile tests on EP are similar. Fracto- graphic investigations75 have shown that modification by pressure does not entail a change in the rupture mechanism. The differences between the fractograms of the specimens compared are merely quantitative; in the modified EP, the structural elements of the fractograms are smaller.A promising method of improving the mechanical properties of polymers is believed to be the cure under pressure of reinforced compositions, especially if one employs fillers with a low surface energy or fillers tending to undergo sedimentation. Fig. 5 presents the distribution of microhardness along sections of cylindrical Table 2.The average sizes of the globules (d) and the content of groups of globules of different sizes at different pressures.62, 68 Polymer P /MPa d /mm Content (%) of groups of globules of different sizes /mm 1 ± 1.4 1.4 ± 2 2 ± 2.8 2.8 ± 4 4 ± 5.6 5.6 ± 8 8 ± 11.2 11.2 ± 16 16 ± 22.4 22.4 ± 32 32 ± 44.8 EP-1 0.1 5.6 7 0.2 0.4 16.4 41.7 30.8 6.8 3.4 0.2 7 7 200 3.5 7 4.4 29.2 35.8 19.9 9.5 1.2 7 7 7 7 EP-3 0.1 9.8 7 7 0.9 6.2 11.3 20.3 24.1 28.6 7.8 0.6 0.2 100 6.7 0.6 5.6 5.8 11.6 18.5 26.9 19.4 9.4 2.0 0.2 7 EP-6 0.1 6.0 0.8 6.0 11.4 16.1 16.1 21.3 20.1 6.8 1.2 0.2 7 200 5.6 0.8 7.2 9.0 19.8 23.2 21.8 11.4 6.2 0.8 7 7 Table 3.The strain ± strength properties ofMECDEat different pressures. P 1073E st P 1073E st /MPa /MPa /MPa /MPa /MPa /MPa 0.1 2.20 230 340 2.60 250 85 2.35 235 425 2.55 245 170 2.35 220 550 2.40 230 255 2.45 230 Hm /MPa 370 340 310 280 250 edge 4 2 0 2 4 d /mm centre edge 1 2 3 4 Figure 5.Distribution of the microhardness Hm in cylindrical specimens of EP-2 (1, 2) and EP-2 with 10 mass% of graphite (3, 4) at different pressures (MPa): (2) and (4) 0.1; (1) and (3) 200. d=Diameter of the cylinder. 956 V A Beloshenko, A A Askadskii, V N VaryukhinEP-2 specimens obtained at different pressures. The level of microhardness Hm for the modified polymer is higher compared with the control specimen, while the scatter of the values is approximately the same. When finely dispersed graphite powder is introduced, the picture changes sharply: a slight increase in inhomogeneity in the distribution of Hm observed for the polymer obtained under the conditions of hydrostatic compres- sion, while in the control specimens the maximum deviation ofHm from the mean reaches 20% for a filler content of 10 mass% ± 30 - mass%.Such a change in Hm in the latter case is caused by the separation of the composition into layers as a consequence of sedimentation processes.The viscosity of the cured composition increases under the influence of a high hydrostatic pressure, with a consequent suppression of the sedimentation processes, the microstructure of the composite becoming more homogeneous. The dielectric properties of EP obtained at atmospheric and elevated pressures have been investigated.75 It was established that the specific electrical volume resistance (rv) of the modified polymers is higher than that of the control specimens and that the highest differences between the rv are observed for specimens in the rubbery state.The EP investigated exhibited the ionic type of conductivity and the increase in rv is therefore accounted for by the decrease in the mobility of the ions induced by the higher values of r and n for the modified polymers.The temperature dependences of rv, the dielectric loss factor (tan d), and the dielectric constant (e0) are correlated.75 The maxima in tan d and e0 caused by the a-relaxation in the modified EP, as well as the corresponding break in the rv(T) relation, are located at higher temperatures compared with the corresponding maxima for the control specimens. Furthermore, the modified polymers have lower absolute values of e0. Studies on the thermal stability of EP showed 88 that appreci- able differences between the characteristic temperatures of the objects compared were not observed for specimens in the form of a block.When the modified specimens are ground, an increase in the original thermal degradation temperature compared with the corresponding temperature of the control specimen is observed.Analysis of the change in the thermal degradation temperature in the course of the annealing of the ground specimens in oxygen and under an inert atmosphere led to the hypothesis that curing under pressure influences the nature of the packing of the molecular chains, as a result of which the most `vulnerable' chemical bonds are shielded from oxygen.Polymers based on interpenetrating polymer networks (IPN) have great potential possibilities for structural modification. Such polymers are obtained by simultaneous or consecutive curing.89 In a series of studies, Lee and Kim 90 ± 93 examined the influence of hydrostatic pressure on the structures and properties of simulta- neously cured IPN in relation to the polyurethane ± polystyrene (PU ± PS) system.The characteristic features of the behaviour of the PU± PS compositions, consisting of blends of linear polymers or of polymers of the type of semi-IPN, were discovered. It was established that, in the course of synthesis under the conditions of hydrostatic compression, r of the final reaction product increases (the increase in r is observed in the range 0.1 ± 250 MPa, while in the region of higher pressures r is independent of pressure).The nature of the temperature dependence of the mechanical loss factor also varies with increase in pressure. The tan d(T) relation for the compositions synthesised at atmospheric pressure has two maxima corresponding to each component of the composition (Fig. 6). In the synthesis under a high pressure, the positions of the temperature maxima are displaced towards one another and ultimately the maxima merge, forming a single peak. This peak is produced as a result of the degeneration of the PU maximum. The glass transition temperature of the dominant phase (PS) decreases with increase in pressure and this is the reason for the low slope of the relation Tg(P) at *1000 MPa. The authors associate the observed effect with the increase in the compatibility of the phases both for the blend of linear polymers and for the semi-IPN and IPN (Fig. 7). The results of electron microscope studies indicate that the morphology of the PU± PS system, synthesised at atmospheric pressure, consists of a fairly homogeneous structure made up of the PU and PS phases.With increase in the synthesis pressure, fully dispersed domains are formed in the PU phase, their size diminishing with increase in the compatibility of the phases. At 1000 MPa, the size of the domains is 30 ± 50 A, which corresponds to the molecular level of the blending. The size of the domains increases with increase in the cross-linking density in the IPN.The results of the study of the morphology of the IPN PU± PS system by dynamic mechanical analysis and electron microscopy agree satisfactorily. In the case of the semi-IPN and the blend of linear polymers obtained at atmospheric pressure, the PU phase forms a matrix, but with increase in pressure a tendency towards the dispersion of this phase is manifested. Such a tendency is most notable for systems in which the PU component has a linear structure.Prolonged annealing of the pressure-modified PU± PS blends at 0.1 MPa and T < Tg promotes the partial restoration of the tan d(T) spectrum characteristic of control specimens synthesised under atmospheric conditions. With increase in the annealing temperature to T>Tg, complete phase separation (characteristic of the control specimens) takes place, which is indicated by the identity of the mechanical loss spectra of the annealed and control specimens.tan d 0.6 0.4 0.2 0 270 310 350 T /K 3 1 2 Figure 6. Temperature variations of the mechanical loss factor for the polyurethane ± polystyrene IPN system containing 50 mass% polyur- ethane at different pressures (MPa): (1) 0.1; (2) 250; (3) 1000.Tg /K 390 370 350 330 0 250 500 750 P /MPa 2 3 1 Figure 7. Dependence of the glass transition temperature Tg on the synthesis pressure for the polyurethane ± polystyrene system. (1) IPN; (2) semi-IPN; (3) polymer blend. Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 957The characteristic features of the structure formation in and the properties of the consecutively cured IPN synthesised under pressure have been investigated 94 ± 96 in relation to epoxide allyl polymers.Diethylene glycol bis(allyl carbonate) (DEGBAC) and epoxybisphenol resins (EO) were used as the allyl monomers. The study of the prepolymerisation (formation of the allyl network) kinetics by rotational viscometry and by measuring rv showed that together with accelerating the process and increasing the average degree of polymerisation, pressure induces or intensifies the gel effect.Under these conditions, even a comparatively slight increase in the original DEGBAC concentration alters the direction of the reaction towards the formation of the network product. The pressure-induced changes in the structure and properties of the prepolymer are inherited on final curing of the composition (curing of the epoxide component) at 0.1 MPa. According to the results of thermomechanical analysis, high glass transition and rubber elasticity temperatures are characteristic of polymers obtained under pressure, but the vitrification temperature range is comparatively narrow. Fig. 8 presents the dependences of Tg for epoxide allyl compositions on the DEGBAC concentration corresponding to different conditions in the synthesis of the prepolymer. The increase in DEGBAC concentration, starting from a certain value, leads to a linear growth of Tg. The change in n induced by an increase in the content of the allyl component and by pressure is correlated with the `behaviour' of Tg (Table 4).The study of the relaxation and thermophysical properties of the polymers obtained at DEGBAC concentrations ensuring the formation of type IPN polymer ± polymer structures demon- strated the following. In the region of Tg, a symmetrical tan d maximum and a jump in heat capacity (DCp) are observed. Higher values of Tg, a narrower vitrification range (DTg), and lower values of DCp have been noted in pressure-modified polymer compositions. The existence of a single glass transition temperature and its linear dependence on the concentration of the allyl component in the compositions investigated indicate a high degree of inter- penetration of the epoxide and allyl polymers.97, 98 The influence of pressure on the change in the compatibility of the networks and the morphology of the polymers is difficult to trace for such systems. Nevertheless, the experimental results permitted a series of conclusions. The characteristic differences between the values of Tg , DTg and DCp for the polymers (Fig. 8 and Table 4), associated with the use of a high pressure in the prepolymerisation state, indicate the presence of a modifying effect of the pressure.Pressure affects primarily the degree of polymerisation of DEG- BAC: for identical original concentrations of the allyl monomer, a higher degree of conversion of the prepolymer synthesised under the conditions of hydrostatic compression is observed and the probability of the formation also of a denser network of chemical bond cross-links therefore increases.In the curing of EO, the differences between the densities of the allyl networks are retained, which is reflected in the values of n and Tg for the polymers compared. The uniformity of the polymer structures may be inferred to some extent from the values of DTg and DCp: DTg and DCp diminish with increase in homogeneity.76 Hence it follows that pressure promotes the homogenisation of the structures of poly- mer compositions.This is indicated also by the results of small- angle X-ray scattering studies: the scattering intensity in the case of specimens obtained using hydrostatic pressure is much lower than in the control specimens (Fig. 9). This effect may be induced by different causes, for example, by the extensive penetration of the EO into the allyl network in the preparation of the prepolymer under a high pressure and also by the formation of a less defective allyl network.Ultimately the factors considered improve the compatibility of the networks. The results of the measurements of the microhardness Hm, which serves as a measure of the local plasticity (Table 4) agree with this conclusion. The lower values of Hm for the pressure- modified polymers may be induced by the plasticisation phenom- enon caused by the increase in the compatibility of the networks.99 ** ** Tg /K 390 370 350 10 11 13 17 21 [DEGBAC] (%) 2 1 Figure 8.Dependence of the glass transition temperature Tg for epoxide- allyl polymers on the DEGBAC concentration. P (MPa): (1) 0.1 (control specimen); (2) 200. Table 4. The structures and properties of epoxide-allyl polymers obtained at different pressures.[DEGBAC] 10720 n Tg DTg DCp Hm (%) /cm73 /K /K /kJ kg71 K71 /MPa 16 6:0 7:8 343 353 50 30 0:24 0:21 150 180 17 7:0 9:2 343 358 60 30 0:29 0:21 220 180 18 7:6 9:2 343 358 60 20 0:38 0:21 250 185 Note. The quantities above the horizontal line represent the characteristics of the control specimens and those below the horizontal line represent the characteristics of the specimens obtained using a pressure of 200 MPa. 1 2 Scattering intensity /pulse s71 220 160 100 0 40 80 120 2 y /min Figure 9. Results of the study of epoxide-allyl polymers (DEGBAC content 18 mass%) by the small-angle X-ray scattering method for prepolymerisation at different pressures (MPa): (1) 0.1; (2) 200. 958 V A Beloshenko, A A Askadskii, V N Varyukhin3.Preparation of oligomeric products by fusion under pressure The modifying effect of pressure may be used not only for the preparation of network polymers and polymer ¡À polymer systems but also in the synthesis of their reactive components. The effect of hydrostatic pressure on the fusion of the epoxybisphenol A resin ED-22F and bisphenol A has been examined 100 and a study has been made of the fractional composition of the reaction product.In a study of the fusion kinetics, use was made of the method involving the titration of the reaction groups: the epoxide (E), alcoholic OH (A), and phenolic OH (P) groups. The fractional composition was studied by the gel-chromatographic method. Figure 10 presents typical chromatograms of the reaction product obtained on fusion of the epoxide resin for 1 h.The assignments of the peaks are presented in Table 5. The influence of the duration of exposure and pressure on the fractional composition of the epoxide resins is reflected in the data in Table 6 and the change in the molecular parameters is reflected in the data in Table 7. Despite a certain scatter of the parameters presented in Table 7, the acceleration of the process by the pressure can be clearly traced.Pressure leads to an increase in the proportion of the high-molecular fractions and in the degree of polycondensa- tion (Table 6). On the other hand, an integral parameter such as polydispersity changes comparatively little. The synthesis (chain growth) under pressure of oligomeric products may be represented by the following simplified scheme:101 ¡À 105 It may be accompanied by a side reactionDchain branching: CH27CH7CH27Ar7O7CH27CH7CH2+HO7Ar7OH O O CH27CH7CH27O[7Ar7O7CH27CH7CH27O]n7Ar O O H2C7CH7CH27O OH OH O7CH27CH7CH27O7Ar CH27CH7CH2 CH27CH7CH2 +CH27CH7CH27O7Ar O OH 1 2 3 4 5 6 I II 28 25 22 19 16 VR /min Figure 10.Chromatograms of the product of the fusion of an epoxide resin with bisphenol A.P (MPa): (I) 0.1; (II) 100; the numerals opposite the curves represent the numhers of the peaks in Table 5. Table 5. The retention times (t), the degrees of polycondensation (m), and the molecular masses of the product of the fusion of an epoxide resin with bisphenol A (Fig. 10). No. of t m M No. of t m M peak /min peak /min 1 26.9 0 340 5 21.5 4 1476 7 1364 2 24.2 1 568 6 20.2 5 1592 3 22.8 2 908 19.4 a 7 1936 796 18.2 a 10 2620 4 22.0 3 1136 17.0 a 14 3532 a These polymeric products are not represented by individual peaks on the chromatogram but are `located' in the tail part of peak No. 6. Table 6. Variation of the fractional composition of epoxide resins in the course of their synthesis. Time /h Content of fraction (%) for different degrees of polycondensation 0 1 2 3 4 ¡À 6 7 ¡À 9 10 ¡À 13 514 1.00 17:9 8:8 19:0 8:0 19:6 13:5 14:9 16:8 23:9 28:0 4:1 15:5 0:6 7:6 ¡¦ 1:8 1.25 15:8 8:0 17:4 6:0 22:0 14:0 6:6 19:0 32:2 27:1 5:2 15:8 0:8 8:0 ¡¦ 2:1 1.50 14:6 7:9 16:5 5:5 18:3 13:7 14:1 20:4 29:7 25:9 5:8 16:0 1:0 8:1 ¡¦ 2:5 1.75 12:9 7:5 15:0 5:4 17:0 13:4 15:1 21:1 32:3 25:5 6:5 16:1 1:2 8:2 ¡¦ 2:8 2.00 10:7 7:4 14:4 5:3 16:2 14:5 14:5 20:8 36:0 24:6 6:8 16:2 1:4 8:2 ¡¦ 3:0 2.50 8:6 7:3 12:3 5:1 15:0 13:3 12:3 22:2 43:3 24:4 7:0 16:3 1:5 8:3 ¡¦ 3:1 Note.Here and in Table 7 the quantities above the horizontal line were obtained for the control specimens and those below it refer to specimens synthesised at 100 MPa. Table 7. Kinetic parameters of the synthesis and the molecular parameters of the epoxide resins obtained.Time 104k b Mn 7 Mw 7 Mz 7 U f n E f w E f Pn f An /h /s71 0.5 1:1 1:4 0:15 0:04 7 7 7 7 7 7 7 7 0.75 1:1 1:5 0:15 0:04 7 7 7 7 7 7 7 7 1.00 1:1 1:6 0:16 0:04 799 1059 1100 1548 1387 1968 1:4 1:4 1:6 1:4 2:2 2:0 1:1 0:8 1:4 2:6 1.25 1:1 1:6 0:16 0:04 863 1100 1215 1600 1547 2020 1:4 1:5 1:6 1:2 2:2 1:8 1:1 0:6 1:6 2:9 1.50 1:1 1:5 0:17 0:04 867 1100 1227 1600 1562 2000 1:4 1:5 1:5 1:2 2:1 1:7 1:0 0:5 1:7 2:9 Note.The average molecular masses (the number-average Mn 7 , the mass- average Mw 7 , and z-average Mz 7 molecular masses) were calculated using chromatographic data, while the average functionalities were obtained from the formulae f E n =Mw 7 cE/1000, f E W=Mw 7 cE/1000, f P n =Mn 7 cP/1000, and f A n =Mn 7 cA/1000, where cE, cP, and cA are the concentrations of the E, P, and A groups in the polymer blend. The selectivity is designated by b.Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 959It is believed 105 that the polar complex tertiary amine (R3N) ¡À phenol (HOPh) interacts with the epoxy-group in the main reaction and that the reaction proceeds in accordance with the scheme The kinetics of this process are described by the second-order equation 7d[E]/dt=k[E][R3N]0 , because the concentration of the reaction complex R3N.. .HOPh is determined exclusively by the concentration of the tertiary amine. Bearing in mind that [R3N]=const in the course of the reaction,102 the process considered may be described by a first- order kinetic equation for the specified R3N concentration.In reality the process is much more complicate, because the complexes Pm, PmE, and PmA (m=1¡À?), where m is the degree of association of the phenol in the complex andA is the amine, are formed reversibly in the system. The complexes PmA with m=1 and 4 are reactive.All these features significantly complicate the kinetic description of the fusion process because H-complexes and (or) contact ion pairs may be regarded as the reactive species in the chain propagation and branching reactions:102 A+*OH H-complex contact ion pair. The low acidity of the alcohol formed in the chain propagation reaction precludes its competition with phenol in the complex formation process.For this reason, the side reaction is hindered in the presence of phenol and may be neglected to a first approx- imation.101 ¡À 103 Table 7 presents the effective reaction rate constants (k) calculated from the consumption of the epoxy-groups (a first- order equation). Evidently, the approximate constancy of k is observed under the conditions investigated. This factor confirms the hypothesis that the main contribution to the rate of the given reaction comes from the amine ¡À phenol complexes and not from the free ions.102 The fact that the formal-kinetic descriptions of the reactions at 0.1 and 100 MPa are the same may be interpreted as evidence in support of the constancy of the mechanisms of these processes in the system investigated. This permits the employment of the known pressure dependence of k (Zhulin 106) for the calculation of the volume activation effect (DV6��0 ) corresponding to atmo- spheric pressure: log kP k0 �� ¡¦ DV6�� 0 F T , where F is a function of pressure.With the aid of this relation and the values of k presented in Table 7 DV6��0 =712.0 cm3 mol71 was found. The value of DV6��0 obtained differs significantly from that established for the analogous polycondensation process:67 DV6��0 =727.0 cm3 mol71 (interaction in the epoxide ¡À amine system with formation of a network polymer).Consequently the influence of pressure on the interaction in the epoxide ¡À phenol system is appreciably weaker than in the epoxide ¡À amine system, which may be associated both with the differences in the transition states and with the solvation of the transition state in the epoxide ¡À phenol system by oligomeric molecules, which may play the role of a `solvent' (the volume activation effect for the viscous flow of such a solvent is positive 106).Using the approach based on the cascade theory of branching processes,107 the selectivity of the process investigated was esti- mated 100 as the ratio of the rate constants for the side and main reactions.It was established that the synthesis of epoxide resins under the conditions of hydrostatic compression improves the selectivity significantly. IV. Solid-state extrusion of polymers and polymer composites 1. Crystallisable polymers Porter and coworkers were some of the first investigators who studied solid-state extrusion.108, 109 They achieved the extrusion of PE through a capillary viscometer at temperatures close to the melting point.The transparent ultraoriented fibre thus obtained was characterised by high moduli of elasticity. In order to obtain a continuous oriented thread with the aid of a capillary rheometer, in a later study 110 PE was crystallised under pressure, after which it was extruded at T<Tm.This resulted in the appearance of a rigid strengthened fibre. The structure of the polyethylene fibres obtained by solid- state extrusion is microfibrillar and is in the main similar to the structure produced in the usual stretching. The main bunch of microfibrils (fibrilla) is ribbon-shaped. The microfibrils are linked sideways by penetrating chains, which has been confirmed by X-ray investigations.111 The amorphous component of the extruded specimens is also oriented, although to a lesser extent than the crystalline component. The maximum orientation of the amorphous phase is attained at extrusion temperatures close to the a-relaxation temperature.112 The properties of the extruded fibres depend both on the extrusion conditions (extrusion temperature; draw ratio l=d2 b/d2 d, where d2 b and d2 d are respectively the diameters of the billet and the calibrating aperture of the die; the geometry of the deforming instrument) and on the original morphology of the polymer.It has been shown for HDPE that there is an optimum ratio between the CEC and CFC in the original material, which makes it possible to obtain the best set of mechanical properties. On the other hand, the maximum tensile modulus of elasticity is attained on extrusion in the vicinity of the a-process temper- atures.25 An increase in external pressure also corresponds to an increase in the modulus of elasticity.113 The tensile strength is determined mainly by the value of l.25 The solid-state extrusion method, used to obtain ultraoriented fibres, may be employed also for the preparation of films, which are likewise characterised by effective mechanical properties.114 The solid-state coextrusion process has been frequently used in recent years for the uniaxial stretching of films.115 ¡À 117 HDPE is usually the material of the shell in which the film is placed.The choice of HDPE is determined primarily by its satisfactory deformability.The characteristic features of the solid-state extrusion of block specimens of isotactic polypropylene have been investi- gated.118, 119 The authors proposed a scheme for the process dynamics, incorporating nonstationary and stationary extrusion. The description of the process includes relations agreeing satisfac- torily with experimental data.Within the framework of the chosen approach,118 taking into account the occurrence of the marked orientational strengthening of polymers, the formation of spiral defects on extrusion with high draw ratios l becomes under- standable:120 For such values of l, the yield point (sy) becomes close to the ultimate tensile strength (st) and any local inhomoge- neity in the flow leads to the rupture of the extrudate.A new series of publications on this question appeared in connection with the development of the hydrostatic extrusion method. When a liquid is used to transmit a high pressure, the most favourable conditions are created for solid-state moulding and it is possible to obtain rod-like and tubular articles with different profiles and a satisfactory quality of the surface and performance properties.121 R7CH7CH2+R3N_HOPh R7CH7CH2_NR3 _ O _H7OPh RCH7CH2NR3 OH + OPh7 960 V A Beloshenko, A A Askadskii, V N VaryukhinThe vast majority of studies on hydroextrusion have been carried out in relation to PE (see, for example, Refs 120 and 122 ± 132).The very significant role of the technological process parameters, particularly extrusion temperature, the rate of defor- mation, and the angle of the drawing die (a) has been noted.It has been established 126, 128 that the most favourable deformation regime is the creep regime, in which the rate of emergence of the extrudate is *161073 m s71. The extrusion temperature range may be fairly broad but not unduly close to the melting point, where there is a possibility of the sticking of the material on the deforming instrument owing to its local fusion.At the same time, there is an optimum extrusion temperature for each draw ratio l, which makes it possible to achieve the maximum effect (improve- ment of mechanical properties) due to orientation. The data for the optimum value of a are contradic- tory,123, 128, 133 but it is at least clear that it is in the range 20 84a460 8.For low conical ity, the nonuniformity of the distribution of deformations on the cross- section of the extrudate is small, but, owing to the increase in the contact area, there is an increase in the friction of the billet against the die, which increases the extrusion pressure. For large angles, there is an increase in the nonuniformity of the distribution of the deformations, which promotes an increase in the extrusion pressure and has an unfavourable effect on the distribution of properties over the cross-section of the extrudate.With increase in l, the variation of the density of the extruded PE passes through a minimum,124, 126 the modulus of elasticity, the strength, and the hardness increase, while the plasticity diminishes.25, 31, 124, 128 It is noteworthy that the increment in the tensile modulus of elasticity induced by hydroextrusion with high draw ratios, may exceed two orders of magnitude, while the stress may exceed one order of magnitude.128 The capacity for strength- ening is determined to a large extent by the molecular mass: in high-molecular PE, the increment in the modulus of elasticity is smaller than in low-molecular PE for the same technical param- eters of the hydroextrusion process.31, 124, 128 Together with the change in mechanical properties, an improvement of dielectric parameters has been noted for extrudates. Thus it has been observed 128 that the extruded HDPE is characterised by a greater resistance to the formation of incomplete electrical breakdown channels compared with the original material.This parameter increases with increasing draw ratio. The study of the structures and properties of PE made it possible to put forward the following mechanism of the structural transformations in the hydroextrusion process.120, 127 For low values of l, the material has a typical fibrillar structure with slightly skewed crystallites and intra- and inter-fibrillar layers.With increase in l, the crystallites become thinner in a transverse direction, are displaced relative to one another in the direction of the extrusion axis, the fraction of the interfibrillar amorphous layers diminishes, and the layers are partly drawn into the fibrils. On further increase in l, the interfibrillar layers vanish and a structure made up of layers perpendicular to the extrusion axis arises.Dense interfibrillar strands of the type ofCECappear in the amorphous sections. The increase in the modulus of elasticity for small and moderate draw ratios is due to the increase in the degree of orientation of the structural elements considered and in the degree of crystallinity, but for high values of l it is caused by the uniform distribution of the mechanical load over all the fibrils and by the growing role of the increasing number of intercrystallite strands.Overall, the proposed mechanism agrees with that discussed by Ciferri and Ward.31 Fairly numerous studies have been carried out on the hydro- extrusion of PP. Features in the main similar to those character- istic of PE have been established for this material.134, 135 In contrast to PE, the limiting draw ratios are significantly smaller for PP (l<6).A high degree of die swell (elastic recovery), characteristic of PP, reaching more than 30% for low values of l, has also been noted.134 The characteristic features of the hydroextrusion process and the properties of the extrudates have been investigated to a lesser extent for other representatives of the class of amorphous- crystalline polymers.The results of studies carried out on PTFE have been published.136, 137 The variation of the strength and elastic properties of fluoro-plastics as a function of the extrusion conditions obeys the same rules as in the case of other polyolefins, but the limiting degree of deformation on hydroextrusion does not exceed 4.The elastic and strength parameters for extension in the direction of the extrusion axis increase by a factor of 1.2 ± 3.5. The breaking stress and tensile strength in the transverse direction also increase compared with the undeformed material. Extruded fluoro-polymers virtually lack cold fluidity.The density, the mechanical properties in extension, and the linear coefficient of thermal expansion along the extrusion axis of hydroextruded poly-e-caproamide have been investigated as a function of ratio.138 The distribution of the densities of the axial and shear deformations has been obtained and comparative analysis of the experimental and calculated data on the extrusion pressure and the distribution of the deformations in the extrudates has been carried out.An important performance characteristic of the oriented polymers is thermal shrinkage or the degree of recovery of the original dimensions (the shape memory effect). For HDPE, it does not exceed 10% and diminishes with increase in l.128 In the case of low-density polyethylene (LDPE), the degree of recovery is much greater: for low values of l, it can reach > 50%.120 Numerous studies of the characteristic features of the thermal shrinkage of oriented specimens of crystallisable polymers have been carried out, but they were virtually all performed on films or fibres. The shrinkage characteristics and the shrinkage mechanism have been examined in detail for block specimens 139 in relation to LDPE deformed by hydroextrusion. It was established that the thermal shrinkage of extruded LDPE specimens is a relatively rapid process in which the degree of shrinkage (c) increases with increase in the annealing temper- ature and with decrease in the extrusion temperature.Comparison of the maximum values of c for oriented PE films and specimens obtained by the hydroextrusion method showed that the latter exhibit a greater shape stability.140 ± 143 The changes in the maximum shrinkage stresses (sm.sh) and linear dimensions of LDPE as a function of l and extrusion temperature are similar (Fig. 11) The shrinkage stresses vanish completely at temper- atures corresponding to the fusion of the crystallites in the oriented LDPE. l Te /K 2 1 4 3 2 350 330 310 1 2 3 sm.sh /MPa Figure 11.Relations between the draw ratio l, the extrusion temperature Te , and the maximum shrinkage stresses sm.sh for LDPE. (1) sm.sh(l), Te=313 K; (2) sm.sh(Te), l=4. Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 961During the orientation of LDPE, layered structures consisting of crystalline lamellae arranged perpendicular to the direction of stretching, arise in the extrudates.Stressed tied chains or clusters also appear as a result of the orientation. Heating of the oriented specimens shortens the extended macromolecules or parts of them: under the influence of entropy factors, the crystalline framework is transformed, which makes possible the shortening of the molecules.140, 142, 144 The higher the temperature, the easier is the slip of individual sections of the macromolecules in the amorphous-crystalline polymer.With increase in extrusion tem- peratures, the number of stressed tied chains therefore diminishes, which decreases the stress and the degree of shrinkage. The decrease is caused primarily by the disappearance of the stressed tied chains in the defective crystallites, the disruption of which requires a low activation energy.At the same time, the displace- ment of the maximum on the temperature dependence of the shrinkage stress towards higher temperatures, which occurs with increase in the extrusion temperature,139 is induced by the appearance of layers consisting of crystallites with increased transverse dimensions, the disruption of which requires a higher activation energy.Yet another factor determining the structural rearrangement on heating of the extruded LDPE is the increase in the thickness of the lamellae during the annealing process. Individual segments are then `drawn' into the body of the lamella, which induces the deorientation of the crystalline blocks and the appearance of internal stresses.145, 146 All the results examined above for the solid-state extrusion of crystallisable polymers were obtained in the extrusion of low- porosity monolithic billets.When highly-porous polymeric billets are employed, a kind of anisotropic structure, which makes it possible to achieve a state with a negative Poisson ratio (m), may be formed in the extrudates.147 ± 149 The microstructure of such materials consists of grains joined by fibrils.These materials are interesting for a number of reasons: firstly, the effect of the negative m may be used directly in clamps and seals; secondly, the presence of this effect influences favourably other mechanical properties, for example, the modulus of shear, the resistance to identation, etc.150, 151 Finally, the complex microstructures responsible for the establishment of a negative m may induce also other unusual phenomena associated with internal rotational degrees of freedom.152, 153 Not only monolithic but also powder billet may be subjected to solid-state extrusion.Strips, which were subsequently placed in HDPE were prepared from superhigh-molecular polyethylene (SHMPE) powder 154, 155 by pressing at elevated temperatures (T<Tm).Next the coextrusion process was carried out. Thus a highly modular high-strength state was achieved for the SHMPE. For example, the tensile modulus of elasticity achieved in one of the above investigations 155 amounted to*60 GPa. An even greater rigidity was attained on further stretching of the extruded materials by uniaxial extension.According to the data of Kanamoto et al.,156 the modulus of elasticity of such films exceeded 100 GPa. Porter 157 obtained from powdered PTFE by the solid-state coextrusion method strips having a modulus of elasticity and an ultimate tensile strength of 3.26104 and 36103 MPa respectively. Oriented films have been obtained similarly 158 from polyacrylonitrile powder.They had effective strain ± strength characteristics. The difficulty of the process in which a polymeric billet is converted into a monolithic material, to be subjected to extrusion may in many instances render solid-state extrusion economically unjustified. This concerns particularly polymers with a high viscosity of the melt or reinforced compositions for which tradi- tional highly productive processing methods (conversion of the billet into a monolithic material) are unsuitable.The solid-state extrusion method proposed in two investigations 159, 160 makes it possible to solve this problem. The procedure includes the follow- ing operations: compaction of the powder billet at room temper- ature, free heating of the billet, and extrusion using a heated high- pressure container and die.The process is fairly pro-ductive, since it can be made quasi-continuous, whereupon each successive billet, expels the mould residue from the preceding one. It has been established empirically that the optimum extrusion temperature range for the extrusion of materials such as SHMPE is (0.94 ± 0.96)Tm,161 while for the polyetherketone based on isophthalic and terephthalic acids (brand PKIT) it is (0.95 ± 0.98)Tm.At lower temperatures, the extrudates are obtained in a nonmonolithic form whilst at higher temperatures the polymer sticks to the deforming instrument, inducing an increase in the extrusion pressure and the formation of an uneven surface on the articles obtained (SHMPE) or the cracking of the extrudates (PKIT).With increase in the degree of extrusion, the moduli of elasticity of the polymers considered increase and the increment in the modulus of elasticity for the flexible-chain polymer (SHMPE) is greater than for the semirigid-chain polymer (PKIT) (Table 8). The extrusion-induced changes in st are small in both cases. At the same time, the plasticity diminishes with increase in the draw ratio.The results of mechanical tests on extruded SHMPE specimens are consistent with the familiar rules established for flexible chain crystallisable polymers,126, 128 namely the increase in the modulus of elasticity is faster than the increase in the strength parameters associated with orientational extrusion. The conditions in the plunger extrusion of a polymeric powder billet differ fundamentally from the conditions in the traditional variants of solid-state extrusion and we shall therefore consider in greater detail the dynamics of the formation of the oriented structure in the extruded SHMPE specimens.Figure 12 presents the dependence of the degree of crystal- linity of SHMPE specimens on the draw ratio, obtained from Table 8. The properties of polymer specimens obtained by the plunger extrusion of a powder billet.Polymer l 1073E /MPa st /MPa er a(%) PKIT brand 1.0 2.0 80 7 1.7 2.4 75 7 2.3 2.5 70 7 2.8 2.6 73 7 SHMPE 1.0 0.6 28 b 9.7 b 3.0 0.7 32 11.7 4.0 0.8 30 11.0 5.0 1.1 30 10.0 6.0 1.2 34 10.6 SHMPE± kaolin 1.1 0.7 31 b 10.0 b 2.0 0.6 28 11.8 3.0 0.7 31 10.0 4.0 0.8 41 10.3 5.0 0.9 40 9.2 6.0 1.3 43 7.1 7.0 1.5 47 7.0 SHMPE±Al 1.0 0.7 26 b 8.1 b 3.0 0.8 37 7.4 5.0 1.2 41 5.3 6.0 1.1 40 5.4 7.0 1.1 38 4.8 9.0 0.9 34 6.1 SHMPE± Al(OH)3 1.0 0.7 14 b 4.5 b 3.0 0.8 19 3.0 4.0 0.9 22 3.5 5.0 0.8 17 3.0 6.0 0.8 16 2.5 7.0 0.4 13 4.9 a er is deformation at the rupture point; b sy and ey are quoted respectively instead of st and er. 962 V A Beloshenko, A A Askadskii, V N VaryukhinX-ray data (Kx) and from density measurements (Kd).162 Evi- dently the absolute values of the X-ray degree of crystallinity are smaller than those found from the measurements of r.The observed difference between the values of Kx and Kd agrees with the available literature data 163, 164 and can have two explanations. According to Egorov et al.,163 X-ray diffraction analysis yields the so called `geometrical' degree of crystallinity, because the macro- molecules arranged parallel to one another but without a three- dimensional order are detected by this method as disordered regions.On the other hand, in dilatometric measurements they can contribute to the crystalline phase. Mandelkern 164 expressed a different view, according to which the difference under discus- sion is caused by the presence of interfacial regions possessing a partial degree of order. In these regions, there is a smooth transition from long-range order (crystalline phase) to disorder (amorphous layers).Thus the quantity DK=Kd ±Kx defines the fraction of either oriented macromolecules in the noncrystalline regions or the fraction of interfacial regions.An increase in l leads to an increase in the degree of crystallinity of SHMPE (Fig. 12), i.e. crystallisation due to orientation take place. The increment in the crystalline phase (aor) can be expressed by the difference between the degrees of crystallinity (DK) corresponding to an arbitrary l and l=1. The quantity Kx yields the true characteristic of the orthorhombic crystalline phase without taking into account the partial order in the noncrystalline regions and, by virtue of this factor, is accurate. It is therefore used to determine aor.The parameter DK may be treated as the decrease in the fraction of oriented macromolecules in the noncrystalline regions 163 or in the volume of the interfacial regions.164 Comparison of the data presented in Fig. 12 supports the second version, since, firstly, a decrease in the fraction of stretched through chains with increase in l is unlikely; secondly, the values of DK for the undeformed SHMPE specimens are close to those established earlier by Mandelkern.164 The E(l) relations for SHMPE156 and HDPE,124 obtained in compressive, bending, and tensile tests, are similar (Fig. 13). Without dwelling on the discussion of the differences between the absolute values of E, we may note that, with increase in l, one observes an increase in E which is steeper than a linear increase.There exist two limiting models describing the distribution of stresses in the oriented amorphous-crystalline monomers.31 In the first, one postulates a uniform distribution of stresses in the crystalline block, where the tied chains are fairly uniformly distributed among the microfibrils formed in the stretching process.In the second, one postulates the complete independence of the tied chains and the crystalline blocks, the internal regions of the crystals alternating with amorphous layers which do not contain tied chains. The K(l) relations of the type presented in Fig. 12a correspond to the first model.31 In this case, the volume fraction of the volume (b) occupied by tied chains is calculated by means of the simplified Takayanaga formula 31 b= E EaxÖ1 ¡ KxÜ , where Eax is the axial modulus of elasticity of the crystalline component. Figure 13b presents the dependence of b on the fraction of the crystallised interfacial regions Da. The moduli of elasticity calcu- lated from the results of compressive tests were adopted as E.A linear relation, passing through the origin of coordinates, is observed between b and Da. This means that the interfacial regions participate actively in the formation of the system of tied chains in the extruded SHMPE. The b(Da) relation shows that *45% of the macromolecules in the interfacial regions become tied chains.The result obtained agrees well with theoretical 165 and experimental 166 estimates. Two components of the crystalline phase may be differenti- ated for the extruded SHMPE (taking into account the results examined above): the original crystalline phase transformed by deformation and the phase arising as a result of orientational crystallisation. The latter consists of crystallised amorphous sections and crystallised interfacial regions.Analysis of data on Kd , Kx 0.8 0.7 0.6 0.5 1 3 5 l 1 2 a DK 0.15 0.10 0.05 0.1 0.2 aor b Figure 12. Dependences of the degrees of crystallinity Kd (1) and Kx (2) on the draw ratio l (a) for SHMPE and of DK on the increment in the crystalline phase aor (b). 3 2 1 0 E /GPa a 2 3 1 1 3 5 7 l 102 b 6 4 2 0 5 10 102 Da b Figure 13.Dependence of the modulus of elasticity E for SHMPE and HDPE on the draw ratio l (a) and of b on Da for SHMPE (b). (1) SHMPE, compression; (2) SHMPE, bending; (3) HDPE, extension. Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 963the thermal shrinkage of extruded SHMPE specimens showed 167 that c increases with increase in aor and Da, the absolute values of c and the corresponding aor and Da being similar.The above factors made it possible to postulate the fusion sequence for the components of the crystalline phase. The least perfect crystalline formations, obtained by the orientational crystallisation of the interfacial regions, melt initially. Their relatively high degree of imperfection can be explained by the presence in the interfacial regions of a large number of irregularities in the chains torn away from the crystallites.The next to melt are amorphous regions crystallised in the orientation process in which the number of irregularities is appreciably smaller. Finally, at higher temper- atures the original crystalline phase melts.Thus the evolution of different structural components of SHMPE during the extrusion of a powder billet are described satisfactorily within the frame- work of the model of the three-component structure of the amorphous-crystalline monomer. The mechanism of this process has been examined in greater detail (resorting to fractal analysis and the cluster model of the structure of amorphous polymers) in another study.168 X-Ray diffraction investigations carried out on extruded PKIT specimens showed that the degree of crystallinity of all the specimens studied is *40% and changes little with increase in l from 1 to 3.The specimens are characterised mainly by reflections which correspond to interplanar spacings of 5.71, 3.87, 3.19, and 4.79 A. This state of PKIT is characteristic of an unstable crystalline modification,169 i.e.the processing conditions are insufficient for a transition to the stable crystalline form having four principal reflections with interplanar spacings of 4.79, 3.91, 3.86, and 3.08 A. Experiments showed that such an equilibrium form is formed on pressing at 623 K. It may be that the retention of the instability of the crystalline modification on extrusion in the temperature range 533 ± 593 K is one of the causes of the in- significant change in the strain ± strength characteristics of PKIT.Porter and coworkers 157, 170 investigated the processes involved in the solid-state extrusion of a high-molecular PTFE powder billet. The extrusion was carried out in the temperature range 373 ± 613 K until the attainment of the maximum values of l.It was established that at a temperature below 373 K or above 613 K (this is higher than the melting point, which amounts to 607 K), the process under consideration cannot be achieved. In the range 373 K<Te<553 K, the maximum value of l is 10 and depends only slightly on Te. On further increase in temperature, the limiting value of l increases rapidly and amounts to 60 for Te=603 ± 613 K.The structure and properties of PTFE are determined by the values of l and Te. In particular, the dimensions of the crystallites in both parallel and perpendicular directions relative to the extrusion axis increase with increase in l. The presence of microfibrils, having a transverse dimension (*45 nm) greatly exceeding that usually observed for oriented monomers (6 ± 20 nm), is characteristic of the highly oriented PTFE.The maximum value of the flexural modulus elasticity, corresponding to the optimum extrusion conditions (Te=613 K, l=40) reaches 20 GPa. The method involving the solid-state extrusion of a powder billet proved to be fairly effective also in the case of thermoplastic polyimides.171 The transition of the polymers to the molten state induces cross-linking of the macromolecules, which hinders their further processing.On the other hand, if the extrusion is carried out below Tm, using billets made up of powders pressed into discs beforehand, then for 563 K4Te4593 K it is possible to obtain high-quality articles with a high degree of crystallinity. 2.Glassy polymers There have been few studies of the properties of flexible-chain amorphous polymers extruded in the solid state. An increase in E and st with increase in l is observed for polymeric materials of this class, as for the amorphous-crystalline polymers, although it is not quite so pronounced. A distinctive feature of the flexible-chain amorphous polymers is that their plastic characteristics increase appreciably as a result of extrusion.For example, the deformation at the rupture point (er) of poly(methyl methacrylate) subjected to tensile tests increased from several per cent for the control speci- mens to tens per cent for the extrudates.172 A severalfold increase in impact strength has been noted for impact-resistant polystyrene.173 The sharp increase in the plasticity of these polymers is attributed to the orientation of the molecular chains along the extrusion axis and the breaking away of side chains.172 ± 174 The mechanical properties of polymers extruded in the solid state and of low-molecular substances, for example, metals, differ.The dependence of the strain ± strength characteristic of oriented polymers on l is determined to an appreciable extent by the scheme underlying the stressed state chosen in the test.Whereas on extension of polymers many investigators have noted an increase in the elastic and strength parameters with increase in the degree of orientation of the specimens, on compression the reverse effect is frequently observed or there are relations with a minimum. This applies both to crystallisable and flexible-chain amorphous polymers (Fig. 14).175 The characteristic features of the solid-state extrusion of rigid- chain amorphous polymers have been scarcely investigated. In this connection, a definite interest attaches to data obtained in a study 176 of the influence of the deformation by hydroextrusion on the structure and properties of such materials in relation to the polyketones PAEKN (the product of the polycondensation of 4,40-difluorobenzophenone and bisphenol A) and PAEKF (the product of the polycondensation of 4,40-difluorobenzophenone and phenolphthalein, a card polymer).Table 9 presents the parameters of the hydroextrusion process: l, Te, the extrusion pressure P, and the properties of the extruded polyketones in compressive and microhardness tests.Evidently the pressure diminishes with increase in the extrusion temperature and increases with increase in the extrusion draw ratio. In the case of extrusion at room temperature, the extrusion pressure reaches such high values already at l>3 that the extruded materials are `shot out' and break up. Hydroextrusion improves the elastic characteristics of poly- ketones.For example, the modulus of elasticity of PAEKN increases with increase in l and diminishes with increase in Te. In the case of PAEKN, the highest value of E is observed at Te=488 K. The dependence of the characteristics st and Hm on * * sy /MPa 90 100 110 25 50 75 20 0 40 60 e (%) 1 2 3 4 Figure 14. Dependence of the yield point sy in compression on the degree of deformation (e) for polyethylene (1), polypropylene (2), poly(vinyl- chloride) (3), and poly(methyl methacrylate) (4). 964 V A Beloshenko, A A Askadskii, V N Varyukhinthe conditions in orientational extrusion is more complex. For PAEKN deformed at room temperature, these characteristics were lower than for the cast polymer (l=1). On the other hand, the extrusion of this material at an elevated temperature improves significantly the properties indicated.The values of st and Hm for PAEKF initially fall and then increase as l increases, but the maximum values of these parameters for the deformed polyketone are smaller (Hm) or of the same order ot magnitude (st) than for the original material. The measurements of the density of PAEKF indicate the compaction of the polymer on deformation.This factor is associated primarily with the decrease in the volume of the pores, which are present in large numbers in the cast PAEKF according to optical microscope data. The pattern in the scattering of X-rays by specimens of the original PAEKF consists of an amorphous halo. Annealing of the specimens for 2 h at 488 or 523 K with subsequent slow cooling does not alter it.Hydroextrusion at room temperature also does not change the pattern. At Te=488 K, the separation of the centre of the amorphous halo into two broad low-intensity reflections is observed, which indicates the onset of the crystal- lisation of the polymer. Cast PAEKF specimens are amorphous. Analysis of the X-ray diffraction data showed that hydroextrusion does not lead to any appreciable changes in the structure of the specimens.Comparison of the results of mechanical tests and X-ray diffraction studies makes it possible to understand the character- istic features of the changes in the strain ± strength properties of the aromatic polyketones studied. The slight increase in E and the nonmonotonic variation of Hm observed with increase in l (Table 9) for specimens subjected to hydroextrusion 175 are induced (as for flexible-chain amorphous polymers) by the ori- entation of polymer chains and the values of these quantities are determined by the degree of orientation and the structure of the polymer.The higher the extrusion temperature, the lower the degree of orientation, other conditions being equal. This is responsible for the low-slope of the E(l) and Hm(l) relations observed for PAEKF as Te increases. The strength of the polymer is to a large extent determined by the presence of different kinds of defects in it.The presence of a large number of micropores in cast PAEKF specimens and their `healing' (elimination of microdefects) during hydroextrusion probably account for the nature of the st(Te) relation. The effectiveness of the `healing' increases with increase in pressure,54 so that st increases with decrease in Te (this is accompanied by an increase in pressure) (Table 9).The st(l) relation corresponds to the variation of the strength properties of flexible-chain amor- phous polymers established by Nakayama and Inoue.175 It was noted above that the creation of oriented order in crystallisable polymers enhances more significantly their elastic and strength properties than in the case of amorphous polymers.Extrusion with l=2.5 and T=488 K entails the transformation of the amorphous modification of PAEKN into a semi-crystalline modification. This tends to improve the mechanical character- istics of the given material. The relaxation properties are well correlated with the results of mechanical tests and structural studies.Hydroextrusion increases the original and equilibrium values of the relaxing modulus (Er) and stress (s). The highest values of Er and s characteristic of PAEKN correspond to the modification with a partially crystal- line structure. With the aid of the ideas developed by Askadskii 177 it has been possible to describe in detail the kinetics of the relaxation processes in the extruded aromatic polyketones. In particular, it has been established that the rate of interaction of the relaxators in the extrudates is higher than in the cast materials.This is quite reasonable if relaxators are understood as different kinds of microheterogeneities which are converted into nonrelax- ing structures faster the higher the level of internal stresses (the latter naturally increases on deformation). Large single deformations occurring without the disruption of brittle materials such as network polymers have been achieved as a result of the plasticisation effect during hydrostatic extrusion after the attainment of the oriented state.68, 75, 178 ± 180 EP �character- istic representatives of the class of network polymers � were selected as the object of study in the above investigations. It was established that the solid-state extrusion of EP induces a rear- rangement of the structure at different levels of the structural arrangement, which is indicated by the results of NMR studies and optical-microscope studies.68 Fig. 15 illustrates the variation of the molecular mobility of the structural fragments of high- crosslinked network EP, characterised by the nuclear magnetic transverse relaxation times T2 for the mobile (T2b) and more rigid (T2c) `kinetic phases' and also the variation of the fraction of relatively immobile protons (the pc populations of the relatively Table 9.The influence of hydroextrusion on the properties of polyketones.l Te /K P /MPa 1073E /MPa st /MPa Hm /MPa Polymer �polyketone PAEKN 1.0 7 7 2.0 80 180 2.2 293 800 2.5 70 170 2.4 293 900 3.0 70 165 2.8 293 1100 3.2 70 162 4.2 293 2000 7 7 7 2.5 488 30 3.6 120 237 Polymer �polyketone PAEKF 1.0 7 7 0.7 56 143 1.8 443 3 0.9 63 135 2.6 443 25 1.1 58 126 1.8 453 1 0.7 55 130 2.6 453 3 0.8 48 122 4.0 453 15 1.0 59 127 6.0 453 27 1.3 64 133 1.8 463 1 0.7 40 135 2.6 463 3 0.7 38 132 6.0 463 17 0.8 50 136 logT2b (ms), logT2c (ms) 3 2 1 a 1 2 3 4 5 7 8 6 pc 1.0 0.6 0.2 370 390 410 430 T /K b 4 3 2 1 Figure 15.Temperature dependences of the nuclear magnetic relaxation times (a) and the proton population (b) for EP-7. (1) ± (4) T2b; (5) ± (8) T2c ; e (%): (1) and (5) original specimen; (2) and (6) 15; (3) and (7) 30; (4) and (8) 40.Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 965immobile `kinetic phase') as a function of the degree of deforma- tion (e) and temperature. The differences between the kinetic properties of the frag- ments of the topological structure are induced by the differences between the numbers of degrees of freedom.181 From these stand- points, the more mobile protons (characterised by the values of T2b and pb) belong to the linear chains between the cross-links and defective fragments of the network�the trifunctional cross-links CL3 and chain branches.The less mobile protons (characterised by T2c and pc=1±pb) belong to the tetrafunctional cross-links CL4. The hierarchy of these fragments in order of decreasing mobility is as follows: linear chains between the cross-links and branches � CL3 � CL4.Mainly the molecular mobility of the more mobile structural elements changes on hydroextrusion (Fig. 15). The decrease in T2b in the case of e=15% compared with the control specimen may be a consequence of the occurrence of orientational processes.Some of the mobile fragments are converted into the less mobile `kinetic phase', which is manifested by an increase in pc at ared with pc for the original material. For e530%, an increase in T2b and a decrease in pc are observed. This is apparently associated with the predominance of disordering processes over orientational processes. However, the significant decrease in the `nuclear' glass transition temperature indicates that the disordering processes are accompanied by the mechanodegradation of the network, in the first place of CL3 and CL4 cross-links.The results of thermomechanical studies agree well with NMR data.178 The microstructure of the undeformed high-crosslinked net- work EP consists of globule-like formations made up of smaller elements.Hydroextrusion leads to characteristic changes in the morphology of the polymer. For e=15%, the distinct boundaries separating the globules in the original material already vanish. A further increase in e promotes the obliteration of the relief, which is observed on the etched surfaces of the specimens, bending, crushing, and twisting of the structural elements, and the for- mation of what resembles an `eddy' structure.14 With increase in e, the modulus of elasticity of the high- crosslinked network EP-7 (n=961020 cm73) increases strongly in tensile tests and diminishes in compressive tests.178 For indi- vidual specimens, deformed to the extent of 45%, the values of E obtained in extension exceed those for the control specimens by a factor greater than 2.The density varies nonmonotonically, reaching a maximum at e&25%. The dependence of the strength parameters on e, like the dependence on r, passes through an extremum (Fig. 16); on extension, sy and st vary in parallel. In compressive tests, sy and st reach minimum and maximum values respectively at e=25%. In both types of tests, er increases and then decreases with increase in e to 25%, while Hm decreases monotonically for EP-7 as e increases (Fig. 17). The nonmonotonic variation of the r(e) relation for hydro- extrusion is associated with the following factors. Deformation under a high hydrostatic pressure promotes the `healing' of the micropores formed in the EP curing stage 25, 80 and the orientation of the polymer chains in the direction of the application of force leads under these conditions to the greater ordering of the structure compared with the original structure. All these features are responsible for the increase in r.At the same time, as e increases, the material becomes loosened with formation of interfaces and microcavities and the intramolecular and intermo- lecular bonds are broken,25, 182, 183 which leads in its turn to the appearance of cracks and a decrease in r.The changes in r noted above are undoubtedly reflected in the level of the mechanical properties of the hydroextruded polymer. This is indicated by the variation of st and er which is identical with that of r in both types of tests. The healing of the microcavities, predetermining to a large extent the strength of the continuous medium, leads to an increase in st and er, while the increase in porosity for e>25% induces an appreciable decrease in these quantities.At the same time, the results of mechanical tests cannot be accounted for solely by the change in r. As already mentioned, processes involving the orientation of the chains of the macro- molecules and mechanodegradation develop in EP on hydro- extrusion.The orientation is probably achieved via the rotation of the elementary network, leading to the arrangement parallel to the direction of extrusion of the individual sections of the chains between the cross-links. For low values of e, orientation processes predominate. They are responsible for the appearance in the polymer of internal microstresses, influencing the nature of the subsequent deformation of the specimen in tensile and compres- sive tests.As a result of the operation of the microstresses, the Bauschinger effect is observed:184 the preliminary plastic defor- mation, induced by the extension, decreases the resistance to plastic deformation on compression. The moulding process dur- ing hydroextrusion can be regarded to a first approximation as deformation by extension (especially for a small angle of the die cone).31 Therefore the decrease in sy, observed in compressive tests for specimens with e=25% compared with the original specimen, may be induced by the identity of the signs of the microstresses in the extrudates caused by the tension in the linear chains between the cross-links, and of the applied stresses. It is noteworthy that the changes established in the mechanical proper- ties of the deformed high-crosslinked polymer network are similar sy , st /MPa 110 90 70 50 20 40 e (%) st /MPa 280 260 240 220 3 2 1 4 Figure 16.Dependence of the yield point sy (1, 3) and the breaking (tensile) stress st (2, 4) on the degree of deformation e for EP-7. (1) and (2) extension; (3) and (4) compression. Hm /MPa 180 140 0 20 40 e (%) 2 3 1 Figure 17.Dependence of the microhardness Hm on the degree of defor- mation for EP-7 (1) and EP-5 (2, 3). Te(K): (1) and (2) 293; (3) 358. 966 V A Beloshenko, A A Askadskii, V N Varyukhinto the changes in the strain ± strength properties of oriented crystallisable and amorphous polymers,175 which indicates the similarity of the mechanisms of their deformation.The study of the relaxation properties of the high-crosslinked network EP carried out on the original specimens and the extrudates has shown that, for e=35% and 45%, there is a sharp increase in molecular mobility induced by the loosening of the material and the partial mechanodegradation of the net- work.178 The relaxation of the stress in the deformed polymer proceeds more vigorously than in the control specimen with increase in e, the stress required for the maintenance of the specified deformation diminishes, the rate of relaxation increases, and the relaxation continues until the attainment of lower stresses.The fitting of the experimental relations with the aid of new relaxation memory functions 177 made it possible to establish that the stress relaxation process in the extruded high-crosslinked network EP is determined with equal probability by the kinetics of the interaction of the relaxators (microheterogeneities) and the kinetics of the diffusion of the nonrelaxators.178 Compared with the original material, the number of microheterogeneities in the extrudates diminishes.In order to increase the probability of breaks in the network, experiments have been carried out 71 on the hydroextrusion of an incompletely cured EP-1 epoxide composition.The curing of the specimens was completed after hydroextrusion. It was established that the elastic and strength properties depend only slightly on the degree of the preliminary deformation of the specimens.If the curing processes are not taken to completion, it is possible to detect at e=50% an increase in E by an order of magnitude compared with the undeformed specimens. Such an increase is induced by the orientation of the EP macrochain along the deformation axis. Probably a state with a high level of elastic and strength characteristics may be achieved for EP under appropriate experimental conditions when the time required for the completion of the curing of the extrudates is significantly shorter than the relaxation time of the oriented structure.A study of the behaviour of the extruded EP having an open network (n=461020 cm73) showed 179 that the nature of the dependences for sy and er on e in these materials is the same as for the high-crosslinked network EP.The difference consists only in the absolute values of sy and er as well as the values of e for which a change in the type of variation of sy and er is observed; in the open network polymer (EP-5), the increase in sy and the fall of er begin for higher values of e. On the other hand, the dependences of E, sy and r on e are qualitatively different. They increase over the entire range of e.The microstructures of the original EP-5 specimens consist of globule-like formations of different dimensions. Defor- mation by hydroextrusion homogenises the microstructure and renders it finely disperse. On the other hand, an eddy structure is not formed up to e=45%. The observed differences between the changes in the micro- structure, r, and the mechanical properties of the EP compared are associated with the different network densities and mobilities of the linear chains between the cross-links.In the more rigid polymer (EP-7), the plastic deformation processes accompanied by the rupture of the network of chemical bonds begin sooner and proceed more vigorously. In the case of EP with an open network (EP-5), they are scarcely manifested (for the range of e inves- tigated) in the behaviour of the integral macroscopic character- istics (r, E, sy, st, er).On the other hand, the use of a local method such as the procedure involving the measurement of microhard- ness made it possible to follow the evolution of network disruption processes even for low values of e.179 Fig. 17 (curve 2) illustrates the variation ofHm in the extrusion of EP-5 in the glassy state.The nonmonotonic variation of Hm with increase in e indicates com- petition between the orientational strengthening and disordering processes caused by the loosening of the network. In the case of EP-7, in which the degradation phenomena develop sooner, such a maximum in Hm is probably displaced towards e<15%, and for this reason one observes on the experimental curve only its right- hand (descending) branch (curve 1).It is logical to suppose that, in creating more favourable conditions for the orientation of the internodal fragments in EP, the maximum under consideration should be observed for larger values of e. The Hm(e) relation, presented in Fig. 17 (curve 3), confirms this hypothesis. It was obtained for the EP-5 polymer deformed in the rubbery (non- equilibrium) state with subsequent `freezing' of this state. Under conditions where the mobility of the linear chains between the cross-links increases significantly compared with the mobility of such chains in glassy polymers and the values of e for which the disruption of the network is possible increase, this is reflected by the displacement of the maximum in the Hm(e) relation towards higher values of e.The behaviour in hydroextrusion of polymers of the IPN type with components differing appreciably in their rigidity is unusual. The results of studies 179, 180 carried out on the polyurethane ± po- lyesteracrylate (PU ± PEA) system suggest that mainly the elastic component (PU) undergoes orientational strengthening.The strengthening effect is achieved because the rigid network (PEA) prevents the relaxation of the stressed PU chains, creating constrains in the form of macromolecular entanglements. At a low PEA concentration, the deformation-induced strengthening, caused by hydroextrusion, is insignificant, because the density of such entanglements is low and the elastic component relaxes readily.With increase in the PEA concentration, the degree of orientational strengthening increases, but mechanodegradation processes begin to be manifested already for a PEA content of 30%. Both elastic and rigid chains are probably disrupted. The latter follows from the fact that, for PU: PEA550 : 50, the possibility of the deformation of the IPN is reduced sharply.The proposed explanation is supported by the following features. When the PU: PEA ratio changes from 90 : 10 to 70 : 30, the system undergoes phase separation.185 This makes it possible to consider separately the evolution of the rigid and elastic networks during orientational stretching. It is known that in the solid-state extrusion of materials incorporating plastic and brittle (stronger) components with the former predominating, the weaker and more plastic component undergoes strengthening 186 Under these conditions the brittle phase may be disrupted with increase in e.The role of such phases in the PU±PEA composition is assumed by the PU and PEA networks respectively. The characteristic features of the thermal shrinkage of net- work polymers have been investigated 179, 180 in relation to EP and the PU±PEA system.It was established that the recovery of the original dimensions of the extruded EP when they are heated in a free state begins near the glass transition temperature (T<Tg) and terminates at T'Tg with attainment of the 100% effect. With increase in e, the temperature of the onset of shrinkage diminishes. In contrast to linear amorphous and crystallisable polymers, network polymers have a narrower shrinkage temperature range and are characterised by the ability to retain the restored form at temperatures significantly exceeding Tg.The latter is known to be due to the presence of a network of chemical cross-links prevent- ing the transition to the viscous flow state. The temperature of the onset of shrinkage is correlated with Tg (Fig. 18). When extruded specimens with fixed ends are heated, the shrinkage stresses originally increase and then fall, this being accompanied by the disruption of the specimens in the vicinity of the clamps (Fig. 19). The maximum shrinkage stresses increase with increase in e and n. The latter fact reflects the sm.sh(Tg) relation (Fig. 18), since n and Tg vary in parallel in the case under consideration.180 It is remarkable that the values of sm.sh are approximately the same as for oriented block specimens of amorphous-crystalline polymers.The characteristic features of the variation of the size of the oriented PU±PEA specimens during heating are on the whole similar to those established for EP: the shrinkage temperature range is comparatively narrow, the recovery of the original dimensions reaches 100%, and To.sh diminishes with increase in e.The temperature range in which the characteristic ssh(T) Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 967relations for the IPN investigated have an extremum (Fig. 20) indicates the dominant role of the elastic component, while the decrease in the half-width of the ssh(T) maximum with increase in e indicates an increase in the contribution of destructive processes (the specimens are disrupted earlier for the material with a defective structure in measurements of ssh) under these condi- tions.The fact that sm.sh varies nonmonotonically (the relation has a maximum) with increase in e is also striking.Such variation of ssh may be caused by the loosening of the network and hence by the partial relaxation of the stresses in the extrudate. The breaks on the right-hand branch of the ssh(T) maximum for the PU±PEA system, preceding the disruption of the specimens, may be associated with the contribution of the rigid component. The `driving forces' of the thermal shrinkage of network and linear amorphous polymers 187, 188 are apparently the same.The difference consists in the fact that in the case of network polymers the twisting process involves not molecular chains but internodal fragments oriented along the deformation axis during extrusion. Evidently, depending on the nature of the molecular mass distribution, the nature of the thermal shrinkage spectrum, especially its width, should also vary in these materials.For IPN with a high degree of segregation of the components, the stress relaxation processes in different networks are separated by a temperature range. This makes it possible to identify them on the basis of the ssh(T) relation (Fig. 20): by varying the composition and ratio of the components, it is possible to control the form of the relation, obtaining different versions of the shrinkage spec- trum.For amorphous polymers (in contrast to crystallisable ones), the version of solid-state extrusion involving the use of a powder billet has not so far come to be widely employed. Nevertheless, individual investigations indicate that this processing procedure is promising for the establishment of an oriented structure. The extrusion process and the properties of extruded DF-10 (an aromatic polyester), consisting of a composition based on a polyarylate and a copolysulfoneformal, have been investigated.189 The results of these studies have shown that the optimum process is achieved if the extrusion temperature is 0.95 ± 0.98 of the temperature corresponding to the transition of the polymer to the viscous flow state and l&6 ± 7.Table 10 presents the strain ± strength characteristics of DF-10 in tests involving com- pression and three-point bending (nicked specimens). Evidently, with increase in l, the modulus of elasticity changes only slightly, whereas sy and st increase appreciably. At the same time the plasticity of the material increase.These features are correlated with those established earlier for hydroextruded amorphous polymers.172 ± 175 3. Polymer blends and reinforced compositions Polymer systems consisting of polymer blends and reinforced compositions have been vigorously investigated in recent years. One of the reasons for the increased interest of investigators in these materials is their great potential possibilities as regards regulation of mechanical properties.However, there have been extremely few studies on the solid-state extrusion of polymer blends and reinforced compositions. The possibility of obtaining oriented block specimens of the LDPE± PP system by the hydro- static extrusion method has been examined.190 Uniaxial solid- state extrusion of such compositions is difficult owing to the weak interfacial interaction of the components caused by their thermo- dynamic incompatibility.In order to eliminate this obstacle, various procedures are resorted to, for example, radiation- induced or photochemical cross-linking.191, 192 Deformation under the conditions of a high hydrostatic pressure made it To.sh /K 330 320 340 360 Tg /K 2 4 6 sm.sh /MPa 1 2 Figure 18.Dependence of the temperature of the onset of shrinkage To.sh (1) and of the maximum shrinkage stresses sm.sh (2) on the glass transition temperature Tg for EP-1. ssh /MPa 2 1 0 350 380 410 T /K 1 2 3 Figure 19. Temperature dependences of the shrinkage stresses (ssh) for EP-5. e(%): (1) 15; (2) 20; (3) 25. ssh /MPa 3 2 1 0 360 380 T /K 3 2 1 Figure 20.Temperature dependence of the shrinkage stresses for the PU±PEA IPN system containing 30 mass% of PEA. e(%) (1) 25; (2) 30; (3) 40; Te=358 K. Table 10. The properties of DF-10 specimens obtained by the plunger extrusion of a powder billet. l Compression Bending E sy , ey (%) st er E st er /GPa /MPa /MPa /MPa /GPa /MPa /MPa 1 1.5 82 6.0 81 37 1.1 108 13 3 1.4 95 6.0 130 44 1.3 109 12 4 1.5 91 6.0 120 40 1.4 118 12 5 1.5 91 6.5 140 40 1.3 115 12 6 1.5 103 8.0 175 58 1.2 115 12 7 1.6 107 8.0 180 50 1.4 129 18 968 V A Beloshenko, A A Askadskii, V N Varyukhinpossible to abandon these procedures and to achieve higher values of l without disruption of the material.190 The change in the mechanical properties of polymer blend specimens on compression and extrusion as a function of l corresponds to the change in these properties established for extruded PE and PP specimens.135, 175 The thermal shrinkage characteristics of the blends have a number of unusual features.An increase in the content of PP increases the shape stability of the extrudates (the residual deformation). At the same time there is an increase in the values of ssh and in the width of the shrinkage temperature range (Fig. 21). For all the PP concentrations, the increase in l entails also an increase in sm.sh. Analysis of the results of the study of Beloshenko et al.190 established that the thermal shrinkage mechanisms in the LDPE¡À PP compositions and in LDPE are similar. At low PP concentrations, the stressed LDPE chains play the main role in the shrinkage process.With increase in the content of PP, the contribution of this phase to the establishment of the thermal shrinkage effect increases, which is manifested by the increase in sm.sh and in the shape stability of the extrudates. The PP phase hinders the relaxation processes in the material: the temperatures at which the internal stresses are fully relaxed are displaced towards the region of higher Tm for the polyethylene phase.However, owing to the occurrence of phase separation, the observed changes are relatively small. The ssh(T) relations for the blends in the region of the descending branch of the maximum differ from the relations established 191 for crosslinked composi- tions: at T<Tm for polypropylene, the right-hand branch of the maximum passes to a plateau.This fact can be explained by the formation during cross-linking of a network of chemical cross- links intensifying the interfacial interaction. In the hydroextrusion of noncrosslinked compositions, one is dealing merely with the appearance of additional physical cross-links, the network of which is unstable during annealing, and also with a decrease in the stresses at the interfaces compared with the stresses in the course of uniaxial extrusion as a result of the favourable deforma- tion scheme.Nevertheless the result obtained is of definite practical interest, since the expansion of the range of ssh increases the functional possibilities of the material. The behaviour of amorphous and crystalline films specimens of the polyether-ether-ketone ¡À polyester-imide (PEEK ¡À PEI) sys- tem subjected to uniaxial extrusion by the coextrusion method has been studied.116 The latter was achieved at temperatures below Tg (amorphous films) or below and above the crystallisation temper- ature (crystallisable films).The influence of extrusion on Tg, crystallisation and the degree of orientation was noted.In the case of amorphous films of the PEEK¡À PEI system, extrusion lowers the glass transition temperature of the polymer blend, although the density of the specimens increases somewhat under these conditions. The degree of orientation of these films increases with increase in the PEEK concentration. In the crystallisable films, the density diminishes monotonically with increase inAand extrusion intensifies the crystallisation process during the subse- quent annealing of the films.The traditional method for the preparation of reinforced polymer compositions is mechanical blending of the components. In the 1970s, a fundamentally new method for the introduction of fillers was proposed D the polymerisation filling method, which makes it possible to obtain polymer compositions directly in the course of the synthesis of the polymer.193, 194 One of the promising aspects of the polymerisation filling method involves the creation of compositions based on SHMPE.Such materials have strain ¡À - strength parameters similar to those of SHMPE, but are harder, have a higher modulus of elasticity, and are more resistant to wear.195 At the same time they are greatly superior to activation- reinforced compositions as regards deformability and impact strength.195 The characteristic features of the solid-state extrusion process and the mechanical properties of extrudates of reinforced compositions based on SHMPE have been investi- gated.161, 167, 196 ¡À 200 The extrusion was carried out with the aid of the method proposed in a patent,159 using polymer powder billets.The extrusion parameters and the properties of the extrudates for a series of compositions investigated, measured in compressive tests, are presented in Table 8. The variation of E, st, and er for the SHMPE¡À kaolin system is analogous to that observed for SHMPE with the sole difference that the introduction of kaolin increases the limiting value of l, the rigidity, and the strength, but reduces the deformability of the material.For other compositions, dependences of E and st on l with an extremum are observed, the values of l corresponding to the maxima in the characteristics diminishing in the sequence kaolin?Al?Al(OH)3. The last composition has the lowest level of strain ¡À strength properties. Comparison of the properties of the extruded compositions with those of control specimens obtained by hot pressing revealed an increase in rigidity and strength during solid-state extrusion for all the compositions investigated.The deformability diminishes at the same time. This feature is manifested even more strikingly in tensile tests.196 Appreciable changes in r during the variation of l have not been observed for the SHMPE¡À kaolin systems.For other compositions, r diminishes with increase in l. The extrusion draw ratio is known 201 not to reflect the real deformation of the macromolecules (lm) in oriented polymers. For reinforced compositions, the discrepancy between l and lm is greater than for unreinforced ones, because the former quantity is calculated in relation to the entire volume of the polymer, whereas only part of it (the polymer matrix) is actually deformed.In order to eliminate this, a corrected molecular draw ratio has been introduced:167 l0m �� lm 1 ¡¦ j , where j is the volume degree of filling. The inequality l0m<l holds for SHMPE and l0m>l is valid for reinforced composi- tions. This relation between the parameters considered reflects the differences in the mechanisms of the deformation of SHMPE and the compositions based on it.For higher values of l, the polymer macromolecules are disrupted at the polymer ¡À filler boundary, which lowers the effective density of the network of macromolec- ular entanglements. This tends to increase l0m and has a positive effect on E. However, owing to the loosening of the interfaces, the fraction of free cavities increases with increase in l, which leads to a decrease in E.For l>5, the second effect predominates and this induces a decrease in E for the reinforced compositions (Table 8). The above conclusion has been confirmed by measurements of r and by the results of a study of the rupture surfaces of the extrudates.198 For low values of l, the surface relief of the break is homogeneous, which indicates a satisfactory adhesion between ssh /MPa 4 2 0 340 390 440 T /K 1 2 3 Figure 21.Temperature dependences of the shrinkage stresses ssh for the LDPE ¡À PP blends. LDPE content (mass%): (1) 90; (2) 80; (3) 40; l=3 Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 969SHMPE and the filler.In the case wherel57, the inhomogeneity of the relief, induced by the loss of adhesion at the polymer ± filler boundary and hence by the rupture of the material at this boundary, can be easily seen. For the majority of the compositions investigated, st varies with l in parallel with E and reaches the maximum values in the region l=4 ± 6 (Table 8).Fig. 22 presents the dependences of the limiting extrusion draw ratio corresponding to the rupture of the specimen in mechanical tests (lr) as a function of the extrusion draw ratio. The values of lr were determined from the formula 202 lr=1+er . Evidently lr increases with increase in l (for oriented polymers, an increase in l diminishes lr as a consequence of the exhaustion of the mobility of the molecular chains).31 On the basis of the above findings, one can also postulate that the rupture of the extruded compositions is related directly to the presence of fillers, namely to the breakdown of the continuity of the specimens owing to the loosening of the polymer ± filler boundaries and the formation of numerous microcracks.On deformation of the specimen in the course of mechanical tests, the microcracks formed during extru- sion aggregate and the specimen is ruptured.The difference between the strain ± strength properties of the SHMPE± kaolin compositions and the analogous properties of other compositions is caused primarily by the shape of the filler particles. The kaolin particles have the shape of plates, while the Al and Al(OH)3 particles are spherical.The presence of particles anisotropic as regards shape promotes the appearance of an intense interfacial interaction, manifested in the partially cohesive mechanism of the destruction of the SHMPE± kaolin composi- tions.203 At the same time, owing to the weak interaction at the polymer ± filler interface, the fully adhesive rupture mechanism operates for the SHMPE± Al(OH)3 system.A mathematical model has been proposed for the solid-state extrusion of polymer composites.204 It is based on the idea that the formation and `healing' of microcavities are possible on deforma- tion under conditions of high pressure. Relations have been obtained which make it possible to calculate the density of the extrudate and to plot it against l.Comparison of the calculated and experimental values of r showed that the model describes adequately the plunger extrusion of a powder billet comprising polymerisation-reinforced compositions. At the same time, it can be used to predict the defectiveness of polymeric materials also in other processing schemes involving high pressures. V. Conclusion The data presented show that a high pressure is frequently used in theoretical and technological research and that its role in the structure modification of polymers is considerable and varied.The methods involving treatment with pressure examined in the review may be used for the effective control of shaping processes and of the structure and properties of polymeric materials with different structures.However, the functional possibilities of the different ways of achieving the modifying effect of pressure are not the same. Thus crystallisation or curing under pressure make it possible to obtain objects in the form of a block, as a rule cylindrical in shape, which limits the range of their practical applications. On the other hand, the substances formed in series- connected IPN (preparation of a prepolymer) or in oligomeric products serve as starting components for more complex com- pounds.The latter are manufactured under traditional conditions, which tends to expand their applications. For example, the IPN can be used as adhesives and also for the preparation of bulky polymeric articles of different shape and for different purposes. Finished articles, which do not require further mechanical treat- ment, can be usefully subjected to annealing under pressure, because favourable changes in the material take place to the greatest extent in its surface layer. The results of the study of solid-state extrusion indicate the considerable potential possibilities of this method in the process- ing of not only flexible-chain polymers but also of rigid-chain network polymers, polymer blends, and reinforced compositions.As a result of the plasticising effect under the influence of high hydrostatic pressure, it is possible to achieve an orientational order in polymers and polymer composites which are brittle and difficult to work. The appearance of such order is accompanied by an improvement in the strain ± strength properties of the materials or by the establishment in the latter of a qualitatively new state (the shape memory effect, a negative Poisson ratio). One may hope that further studies of the application of the isostatic treatment or solid-state extrusion for the structural modification of polymers will help in discovering new aspects of the influence of high pressure which are of interest from both theoretical and practical points of view.References 1. P W Bridgman Rev. Mod. Phys. 18 (1) (1946) 2. P W Bridgman Studies in Large Plastic Flow and Fracture (London: McGraw-Hill, 1952) 3. K Swenson Physics at High Pressure (New York: Academic Press, 1960) 4. M G Gonikberg Khimicheskoe Ravnovesie i Skorost' Reaktsii pri Vysokikh Davleniyakh (Kinetic Equilibrium and Reaction Rate at High Pressure) (Moscow: Khimiya, 1969) 5.H L D Pugh (Ed.) Mechanical Behaviour of Materials under Pressure (Amsterdam: Elsevier, 1970) 6. V I Zaitsev Fizika Prochnosti i Plastichnosti Gidrostaticheski Szhatykh Kristallov (The Physics of the Strength and Plasticity of Hydrostatically Compressed Crystals) (Kiev: Naukova Dumka, 1983) 7. A N Pilyankevich (Ed.) Vliyanie Vysokikh Davlenii na Veshchestvo (The Influence of High Pressures on Matter, Vol. 1) (Kiev: Naukova Dumka, 1987) 8.V L Kolmogorov Napryazheniya, Deformatsii, Razrusheniya (Stresses, Deformations, Degradations) (Moscow: Metallurgiya, 1970) 9. L V Prozorov, A A Kostava, V D Revtov Pressovanie Metallov Zhidkost'yu Vysokogo Davleniya (Pressing of Metals by a Liquid under High Pressure) (Moscow: Mashinostroenie, 1972) 10.A lpashnikov, V V Vyalov Gidropressovanie Metallov (Hydropressing of Metals) (Moscow: Metallurgiya, 1973) 11. A A Galkin, A P Getmanskii Pressovanie Metallov Zhidkost'yu (Pressing of Metals by a Liquid) (Donetsk: Donbass, 1974) 12. B I Beresnev, E V Trushin Protsess Gidroekstruzii (The Hydro- extrusion Process) (Moscow: Nauka, 1976) 13.M V Mal'tsev, E D Doron'kin, K I Ezerskii Gidrostaticheskaya Obrabotka Tugoplavkikh Metallov (Hydrostatic Processing of High- Melting Metals) (Moscow: Metallurgiya, 1978) lr 1.2 1.1 0 3 5 7 9 l �1 �2 Figure 22. Dependence of the limiting draw ratio lr on the draw ratio l for the SHMPE± 70 mass% Al (1) and SHMPE± 45 mass% of bauxite (2) systems. 970 V A Beloshenko, A A Askadskii, V N Varyukhin14.B I Beresnev, K I Ezerskii, E V Trushin Fizicheskie Osnovy i Prakticheskoe Primenenie Gidroekstruzii (Physical Foundations and Practical Use of Hydroextrusion) (Moscow: Nauka, 1981) 15. B I Beresnev (Ed.) Vliyanie Vysokikh Davlenii na Veshchestvo (The Influence of High Pressures on Matter, Vol. 2) (Kiev: Naukova Dumka, 1987) 16. B I Beresnev, K I Ezerskii, E V Trushin, B I Kamenetskii Vysokie Davleniya v Sovremennykh Tekhnologiyakh Obrabotki Materialov (High Pressures in Modern Technologies for the Processing of Materials) (Moscow: Nauka, 1988) 17.P J James (Ed.) Isostatic Pressing Technology (London: Applied Science Publ., 1979) 18. R B Dow Chem. Phys. 7 201 (1939) 19. W Parks, R B Richards Trans. Faraday Soc. 45 203 (1949) 20.C E Weir J. Res. Natl. Bur. Stand. 46 207 (1951) 21. C E Weir J. Res. Natl. Bur. Stand. 50 95 (1953) 22. P W Bridgman J. Appl. Phys. 24 560 (1953) 23. K D Pae, S K Bhateja J. Macromol. Sci., C 13 1 (1975) 24. S K Bhateja, K D Pae J. Macromol. Sci., C 13 77 (1975) 25. S B Ainbinder, E L Tyunina, K I Tsirule Svoistva Polimerov v Razlichnykh Napryazhennykh Sostoyaniyakh (Properties of Polymers in Various Stressed States) (Moscow: Khimiya, 1981) 26.A Ya Gol'dman Ob'emnoe Deformirovanie Plastmass (Volume Deformation of Plastics) (Leningrad: Mashinostroenie, 1984) 27. K Kishore, R Vasanthakumari High Temp.-High Pres. 16 241 (1984) 28. A L Kovarskii Vysokomol. Soedin., Ser. A 28 1347 (1986) a 29. V F Skorodumov, Yu K Godovskii Vysokomol. Soedin., Ser.B 35 214 (1993) a 30. E V Prut, in High-Pressure Chemistry and Physics of Polymers (Ed. A L Kovarskii) (Boca Raton, FL: CRC Press, 1994) p. 341 31. A Ciferri, I MWard (Eds) Ultra-High Modulus Polymers (London: Applied Science Publ., 1979) 32. S Matsuoka, B Maxwell J. Polym. Sci. 32 131 (1958) 33. T Takemura Am. Chem. Soc., Polym. Prepr. 20 270 (1979) 34. D V Rees, D C Bassett J.Polym. Sci., Part A-2 9 385 (1971) 35. P D Calvert, D R Uhlmann J. Polym. Sci., Part A-2 10 1811 (1972) 36. A Peterlin Polymer 6 25 (1965) 37. D V Rees, D C Bassett Nature (London) 219 368 (1986) 38. B Wunderlich, L Melillo Macromol. Chem. 118 250 (1968) 39. J L Kardos, H-M Li,K A Huckshold J. Polym. Sci., Part A-2 9 2061 (1971) 40. B Wunderlich Macromolecular Physics Vols 1, 2 (New York: Aca- demic Press, 1973, 1976) 41.D C Bassett, B Turner Nature (London) 240 146 (1972) 42. S Matsuoka J. Polym. Sci. 42 511 (1960) 43. Y Miyato, C Nakafuku, T Takemura Polym. J. 3 122 (1972) 44. G P Andrianova Fiziko-Khimiya Polioleénov (The Physical Chemistry of Polyolefins) (Moscow: Khimiya, 1974) 45. T Arakava, B Wunderlich J. Polym. Sci., Part A-2 4 53 (1966) 46. R Hasegava, Y Tanabe, M Kobayashi, H Tadokoro, A Sawaoka, N Kawai J.Polym. Sci., Part A-2 8 1073 (1970) 47. W W Doll, J B Lando J. Macromol. Sci., B 4 889 (1970) 48. W W Doll, J B Lando J. Macromol. Sci., B 2 219 (1968) 49. V M Baranovskii, A M Tarara, S I Bondarenko, V P Skvortsov, A I Nepomnyashchii Plast. Massy (3) 52 (1992) 50. A M Tarara, V M Baranovskii, S I Bondarenko, A A Khomik, V A Tsendrovskii, V V Lapinskii Plast.Massy (2) 13 (1995) 51. C L Gruner, B Wunderlich, R C Bopp J. Polym. Sci., Part A-2 7 2099 (1969) 52. Yu F Zubov, V I Selikhova, M B Konstantinopol'skaya, F F Sukhov, N A Slovokhotova, N F Bakeev, A V Kryukov, V A Sokol'skii, G P Belov Vysokomol. Soedin., Ser. A 14 2090 (1972) a 53. D C Bassett, D R Carder Philos. Mag. 28 513 (1973) 54.D C Bassett, D R Carder Philos. Mag. 28 535 (1973) 55. A I Petrov, M V Razuvaeva, A B Sinani, V I Betekhtin Mekhanika Kompozit. Mater. 1121 (1989) 56. A I Shul'gin, I O Kurochkina, V A Aulov Vysokomol. Soedin., Ser. B 33 224 (1991) a 57. Primenenie Polioleénov, Polistirolov, Ftoroplastov i Polivinilatsetat- nykh Plastikov. Katalog [Application of Polyolefins, Polystyrenes, Fluoroplastics and Poly(Vinyl Acetate Plastics)] (Cherkassy: NIITEKhIM, 1981) 58.V A Beloshenko, V G Slobodina, S A Tsygankov, V M Shepel' Fiz. Tekhn. Vys. Davl. 1 (2) 85 (1991) 59. Ya E Beigel'zimer, V A Beloshenko, E V Prut Vysokomol. Soedin., Ser. B 37 2085 (1995) a 60. R E Wetton, H G Moneypenny Br. Polym. J. 7 51 (1975) 61. G J Kogowsky, F E Filisko Macromolecules 19 828 (1986) 62.A A Shturman Plast. Massy (2) 47 (1989) 63. A A Shturman Plast. Massy (11) 51 (1990) 64. O E Ol'khovik, E M Blyakhman Vysokomol. Soedin., Ser. B 18 36 (1976) a 65. V A Beloshenko, V I Zaika,M K Pakter, V M Shepel' Fiz. Tekhn. Vys. Davl. (33) 82 (1990) 66. V A Beloshenko,M K Pacter, B I Beresnev, V I Zaika, T P Zaika High Pres. Res. 6 203 (1994) 67. B I Beresnev, M K Pakter, V A Beloshenko, V I Zaika Dokl Akad.Nauk Ukr. SSR 9B 28 (1989) 68. M K Pakter, V A Beloshenko, B I Beresnev, T P Zaika, L A Abdrakhmanova, N I Bezai Vysokomol. Soedin., Ser. A 32 2039 (1990) a 69. V A Beloshenko,M K Pakter, G I Sviridov Ukr. Khim. Zh.. 61 68 (1995) 70. M G Evans, M Polanyi Trans. Faraday Soc. 31 875 (1935) 71. J K Lee, K D Pae J. Macromol. Sci., B 32 78 (1993) 72.B V Ozerkovskii, V D Plotnikov, V P Roshchupkin Vysokomol. Soedin., Ser. A 25 1816 (1983) a 73. B I Beresnev, V A Beloshenko,M K Pakter, G T Evtushenko, E V Amosova Dokl. Akad. Nauk Ukrainy 3 119 (1991) 74. V A Beloshenko,M K Pacter, V N Varyukhin Acta Polym. 46 328 (1995) 75. VABeloshenko,MKPakter, B I Beresnev, T P Zaika, VGSlobodina, V M Shepel' Mekhanika Kompozit.Mater. 195 (1990) 76.V A Bershtein, V M Egorov Differentsial'naya Skaniruyushchaya Kalorimetriya v Fizikokhimii Polimerov (Differential Scanning Calorimetry in Physical Chemistry of Polymers) (Leningrad: Khimiya, 1990) 77. N Katsuhiko, N Takashi, A Xu, T Kohro Polym. J. 23 1157 (1991) 78. V I Irzhak, B A Rozenberg, N S Enikolopyan Setchatye Polimery. Sintez, Struktura, Svoistva (Network Polymers.Synthesis, Structure, and Properties) (Moscow: Nauka, 1979) 79. A V Grachev,M R Kiselev Yu M Sivergin Vysokomol. Soedin., Ser. A 27 1197 (1985) a 80. Yu S Zaitsev, Yu S Kochergin, M K Pakter, R V Kucher Epoksidnye Oligomery i Kleevye Kompozitsii (Epoxy-Oligomers and Adhesive Compositions) (Kiev: Naukova Dumka, 1990) 81. I Z Chernin, F M Smekhov, Yu V Zherdev Epoksidnye Polimery i Kompozitsii (Epoxy-Polymers and Compositions) (Moscow: Khimiya, 1982) 82. V G Khozin, A A Polyanskii, Yu M Budnik, V A Voskresenskii Vysokomol.Soedin., Ser. A 24 2308 (1982) a 83. V S Kuksenko, M I Karyakina, N V Maiorova, T A Prokof'eva, A I Slutsker Mekh. Polim. 1 157 (1974) 84. V A Beloshenko,M K Pakter, T P Zaika, G V Borisenko Fiz. Tekhn. Vys. Davl. 2 (1) 16 (1992) 85. V A Beloshenko, G V Borisenko, G T Evtushenko, G I Sviridov Fiz.Tekhn. Vys. Davl. 3 (4) 34 (1993) 86. V B Gupta, L T Drzal, W W Adams J. Mater. Sci. 20 3439 (1985) 87. P G Babaevskii, I G Agapov, S V Bukharov, A E Chalykh, V V Matveev Vysokomol. Soedin., Ser. B 35 1697 (1993) a 88. M K Pakter, V A Beloshenko, G V Borisenko Fiz. Tekhn. Vys. Davl. 2 (4) 115 (1992) 89. L Sperling Interpenetrating Polymer Networks and Related Materials (New York: Plenum, 1981) 90.D S Lee, S C Kim Macromolecules 17 268 (1984) 91. D S Lee, S C Kim Macromolecules 17 2193 (1984) 92. D S Lee, S C Kim Macromolecules 17 2222 (1984) 93. D S Lee, S C Kim Macromolecules 18 2173 (1985) 94. V A Beloshenko, L G Nechitailo, G I Sviridov, V F Stroganov, V P Privalko Ukr. Polym. J. 3 49 (1993) 95.L G Nechitailo, V A Beloshenko, V P Privalko, V F Stroganov Ukr. Khim. Zh.. 61 133 (1995) 96. V A Beloshenko, V F Stroganov, Bufistov, V V Korskanov Ukr. Khim. Zh.. 62 138 (1996) 97. Yu S Lipatov, V F Rosovitskii Dokl. Akad. Nauk SSSR 283 910 (1985) b Promising aspects of the structural modification of polymers and polymer composites with the aid of high pressures 97198.D Thomes, L Sperling Polymer Blends (Eds D R Paul, S Newman) (New York: Academic Press, 1978) 99. V F Stroganov, V M Mikhal'chuk, Yu S Zaitsev, L I Maklakov, Yu S Lipatov Dokl. Akad. Nauk SSSR 292 670 (1987) b 100. M K Pakter, V A Beloshenko, A N Daragan, G I Sviridov Zh. Prikl. Khim. 66 2086 (1993) c 101. M A Shcheblanova, S Z Rogovina, L V Vladimirov, A N Zelenetskii,M A Markevich, N S Enikolopyan Dokl. Akad.Nauk SSSR 226 390 (1976) b 102. M A Tovmasyan, A N Zelenetskii, V V Ivanov, G A Grigoryan, O B Salamatina Vysokomol. Soedin., Ser. A 25 862 (1983) a 103. M A Tovmasyan, A N Zelenetskii, V V Ivanov Vysokomol. Soedin., Ser. A 26 1848 (1984) a 104. M F Sorokin, L G Shode Zh. Org. Khim. 2 1463 (1966) d 105. M F Sorokin, L G Shode Zh. Org. Khim. 2 1469 (1966) d 106.V M Zhulin, in Fizicheskaya Khimiya. Sovremennye Problemy. Ezhegodnik (Physical Chemistry. Current Problems. Yearbook) (Ed. YaMKolotyrkin) (Moscow: Khimiya, 1984) p. 144 107. S A Zahir, S Bantle, in Epoxy Resin Chemistry. The 2nd Symposium of the 183rd Meeting of the American Chemical Society (Abstracts of Reports), Las Vegas, 1982 p. 245 108. J H Southern, R S Porter J.Appl. Polym. Sci. 14 2305 (1970) 109. N E Weeks, R S Porter J. Polym. Sci., Polym. Phys. Ed. 12 635 (1974) 110. N Capiati, S Kojima, W Perkins, R S Porter J. Mater. Sci. 12 334 (1977) 111. J Clements, R Jakeways, I M Ward Polymer 19 639 (1978) 112. W T Mead, C R Desper, R S Porter J. Polym. Sci., Polym. Phys. Ed. 17 859 (1979) 113. J A Sauer, K D Pae, S K Bhateja J. Macromol.Sci. 8B 631 (1973) 114. J H Southern, G L Wilkes J. Polym. Sci., Polym. Phys. Ed. 11 555 (1973) 115. Y H Lee, R S Porter J. Macromol. Sci., Phys. 34 295 (1995) 116. H L Chen, R S Porter Macromolecules 28 3918 (1995) 117. R S Porter, L H Wang J. Macromolecul. Sci., Rev. Macromol. Chem. Phys. C 35 63 (1995) 118. A N Kryuchkov, A O Baranov, I Ya Dorfman, N A Erina, E V Prut, N S Enikolopyan Vysokomol.Soedin., Ser. A 26 1993 (1984) a 119. I Ya Dorfman, A N Kryuchkov, E V Prut, N S Enikolopyan Dokl. Akad. Nauk SSSR 278 141 (1984) b 120. A G Korchagin, M A Martynov, S A Tsygankov Vysokomol. Soedin., Ser. B 27 335 (1985) a 121. N S Enikolopyan, B I Beresnev, G D Myasnikov, E V Prut, S A Tsygankov, A V Kryuchkov, N V Shishkova Dokl. Akad. Nauk SSSR 219 368 (1986) b 122.A G Gibson, I M Ward J. Mater. Sci. 14 1838 (1979) 123. R Gupta, P G McCormick J. Mater. Sci. 15 619 (1980) 124. P S Hope, A G Gibson, I M Ward, J. Polym. Sci., Polym. Phys. Ed. 18 1243 (1980) 125. S Abdul Jawad Indian J. Technol. 24 709 (1986) 126. S A Tsygankov, N V Shishkova, B I Beresnev Fiz. Tekhn. Vys. Davl. (17) 77 (1984) 127. A G Korchagin, M A Martynov, S A Tsygankov Vysokomol.Soedin., Ser. A 26 2529 (1984) a 128. S A Tsygankov, N V Shishkova, B I Beresnev Fiz. Tekhn. Vys. Davl. (19) 57 (1985) 129. S A Tsygankov, I Yu Khanukov, N V Shishkova, G D Myasnikov, B I Beresnev Fiz. Tekhn. Vys. Davl. (23) 70 (1986) 130. K Nakayama, H Kanetsuma J. Mater. Sci. 12 1477 (1977) 131. J B Sahary, B Parsons, I M Ward J. Mater. Sci. 20 346 (1985) 132. A A Turetskii, O Yu Zinov'eva,M B Konstantinopol'skaya, V A Aulov, Yu A Zubov, N F Bakeev Vysokomol.Soedin., Ser. A 31 1644 (1989) a 133. N Inoue, T Nakayama Technocrat 10 23 (1977) 134. H N Yoon,K D Pae, J A Sauer J. Polym. Sci., Polym. Phys. Ed. 14 1611 (1976) 135. S A Tsygankov, N V Shishkova, B I Beresnev Fiz. Tekhn. Vys. Davl. (26) 74 (1987) 136. G D Myasnikov, S A Tsygankov Plast. Massy (2) 34 (1985) 137.B I Beresnev, N S Enikolopov, S A Tsygankov, N V Shishkova Dokl. Akad. Nauk Ukr. SSR 4B 47 (1985) 138. E V Slavnov, V M Timofeev Vest. Perm Gos. Tekhnol. Univ., Mekhanika 2 188 (1995) 139. P V Zamotaev, V A Beloshenko Zh. Prikl. Khim. 67 1863 (1994) c 140. F Decandia, R Russo, V Vittoria J. Polym. Sci., Polym. Phys. Ed. 20 1176 (1982) 141. W Chen, K Xing, L Sun Rad. Phys. Chem. 22 593 (1983) 142. V I Dankin,V L Karpov Vysokomol. Soedin., Ser. B 25 130 (1983) a 143. R J Yan, Y Luo, B Jiang J. Appl. Polym. Sci. 47 789 (1993) 144. M Hoff, Z Pelzbauer Polymer 32 3316 (1991) 145. P Zamotaev, T Privalko,O Mitukhin, V Bogdanovich Ukr. Polym. J. 1 221 (1992) 146. J Spells, P Sadler Macromolecules 22 3941 (1989) 147. B D Caddock, K E Evans J. Phys. D., Appl. Phys. 22 1877 (1989) 148. K E Evans, B D Caddock J. Phys. D., Appl. Phys. 22 1883 (1989) 149. K L Anderson, K E Evans Polymer 33 4435 (1992) 150. R Lakes Science 235 1038 (1987) 151. K E Evans Chem. Ind. 20 654 (1990) 152. A C Eringen J. Math. Mech. 15 909 (1966) 153. A C Smith Int. J. Eng. Sci. 6 65 (1968) 154. P D Griswald, A E Zachariades, R S Porter Polym. Eng. Sci. 18 861 (1978) 155. G T Pawlikowski, D S Mitchell J. Polym. Sci., Part B, Polym. Phys. 26 1865 (1988) 156. T Kanamoto, T Ohama, K Tanaka,M Takeda Polymer 28 1517 (1987) 157. R S Porter, in Macroakron 94. The 35th IUPAC International Symposium on Macromolecules (Abstracts of Reports), Akron, OH, 1994 p. 598 158. T Kameda, A Yamane, T Kanamoto,M Ito, R S Porter Vysokomol. Soedin., Ser. A 38 1152 (1996) a 159. USSR P. 1 839 148; Byull. Izobret. (47 ± 48) 101 (1993) 160. Russ. P. 2 083 369; Byull. Izobret. (19) (1997) 161. V A Beloshenko, G V Kozlov, V N Varyukhin, V G Slobodina Acta Polym. 48 181 (1997) 162. G V Kozlov, V A Beloshenko, V G Slobodina, E V Prut Vysokomol. Soedin., Ser. B 38 1056 (1996) a 163. E A Egorov, V V Zhizhenkov, V A Marikhin, L P Myasnikova, A Popov Vysokomol. Soedin., Ser. A 25 693 (1983) a 164. L Mandelkern Polym. J. 17 337 (1985) 165. P J Flory, D Y Yoon, K A Dill Macromolecules 17 862 (1984) 166. D Y Yoon, P J Flory Polymer 18 509 (1977) 167. V A Beloshenko, G V Kozlov, V G Slobodina, E V Prut, V G Grinev Vysokomol. Soedin., Ser. B 37 1089 (1995) a 168. G V Kozlov, V A Beloshenko, V M Varyukhin, V U Novikov Zh. Fiz. Dosl. 1 204 (1997) 169. Ya V Genin, D Ya Tsvankin, S N Salazkin, V A Sergeev Vysokomol. Soedin., Ser. B 31 74 (1989) a 170. H Okuyama, T Kanamoto, R S Porter J. Mater. Sci. 29 6485 (1994) 171. Y D Wang,M Cakmak, F W Harris J. Appl. Polym. Sci. 56 837 (1995) 172. NInoue, T Nakayama, in Proceedings of the 23rd Japan Congress on Material Research, Tokyo, Kyoto, 1980 p. 287 173. B I Beresnev, S A Tsygankov, N V Shishkova, K A Vylegzhanina, O K Belomutskaya, B S Polonskii, E E Manusevich Dokl. Akad. Nauk Ukr. SSR 7B 42 (1987) 174. N Inoue, T Nakayama, T Ariyama J. Macromol. Sci. 19B 543 (1981) 175. T Nakayama, N Inoue Bull. Jpn. Soc. Mech. Eng. 20 688 (1977) 176. A A Askadskii, V A Beloshenko, K A Bychko, V V Shaposhnikova, A V Samoryadov, O V Kovriga, S N Salazkin, V A Sergeev, V G Slobodina, Ya V Genin Vysokomol. Soedin., Ser. A 36 1143 (1994) a 177. A A Askadskii Mekhan. Kompozit. Mater. 403 (1987) 178. A A Askadskii, V A Beloshenko, M K Pakter, K A Bychko, M P Valetskii Vysokomol. Soedin., Ser. A 33 2206 (1991) a 179. V A Beloshenko, V N Varyukhin, T P Zaika, S I Skiba, V I Sheludchenko Fiz. Tekhn. Vys. Davl. 6 (1) 65 (1996) 180. V A Beloshenko, P V Zamotaev, O P Mityukhin, L M Sergeeva, V F Stroganov Zh. Prikl. Khim. 69 1203 (1996) c 181. V M Lantsov, M K Pakter, in Reaktsionnosposobnye Oligomery i Materialy na Ikh Osnove. Khimicheskii Analiz i Fiziko-Khimicheskie Issledovaniya (Metody, Metodiki i Rekomendatsii po Ikh Primene- niyu) [Reactive of Oligomers and Materials Based on Them. Chemical Analysis and Physicochemical Studies (Methods and 972 V A Beloshenko, A A Askadskii, V N VaryukhinRecommendations Concerning Their Use)] (Cherkassy: NIITEKhIM, 1987) p. 75 182. A A Askadskii Deformatsiya Polimerov (Deformation of Polymers) (Moscow: Khimiya, 1973) 183. P G Cheremskoi, V V Slezov, V I Betekhtin Pory v Tverdom Tele (Pores in Solid) (Moscow: Energoatomizdat, 1990) 184. L S Moroz Mekhanika i Fizika Deformatsii i Razrusheniya Materialov (Mechanics and Physics of the Deformation and Frac- ture of Materials) (Moscow: Mashinostroenie, 1984) 185. Yu S Lipatov, L M Sergeeva, L V Karabanova, V F Rosovitskii, S I Skiba, N V Babkina Vysokomol. Soedin., Ser. A 30 649 (1988) a 186. Yu F Chernyi, V Z Spuskanyuk, A A Lyadskaya, A I Opanashchuk Gidropressovanie Instrumental'nykh Stalei (Hydropressing of Tool Steels) (Kiev: Tekhnika, 1987) 187. N I Shishkin,M F Milagin Mekhan. Polim. 3 323 (1966) 188. N I Shishkin Mekhan. Polim. 5 787 (1972) 189. Russ. P. 2 105 670; Byull. Izobret. (16) (1998) 190. V A Beloshenko, P V Zamotaev, O P Mityukhin, V G Slobodina Fiz. Tekhn. Vys. Davl. 5 (3) 43 (1995) 191. P V Zamotaev, I Khodak, S A Sergienko Zh. Prikl. Khim. 66 2071 (1993) 192. V P Gordienko Radiatsionnoe Modiétsirovanie Kompozitsionnykh Materialov na Osnove Polioleénov (Radiation-Induced Modifica- tion of Composite Materials Based on Polyolefins) (Kiev: Naukova Dumka, 1985) 193. F S D'yachkovskii, L A Novokshonova Usp. Khim. 53 200 (1984) [Russ. Chem. Rev. 53 117 (1984)] 194. L A Novokshonova, I N Meshkova Vysokomol. Soedin. 36 629 (1994) a 195. L O Bunina, V A Teleshov, V I Sergeev, S B Ainbinder, K I Tsirule Plast. Massy (8) 13 (1985) 196. V A Beloshenko, V N Varyukhin, G V Kozlov, V G Slobodina, iIn High Pressure Science and Technology (Ed.WA Trzeciakowski) (Singapore: World Sci. Publ., 1996) p. 153 197. G V Kozlov,M A Gazaev, V O Beloshenko, V M Varyukhin, V G Slobodina Ukr. Fiz. Zh. 40 883 (1995) 198. G V Kozlov, V A Beloshenko, T P Zaika, V G Grinev Plast. Massy (6) 29 (1995) 199. G V Kozlov, V A Beloshenko, V G Slobodina Plast. Massy (3) 14 (1996) 200. V A Beloshenko, V G Slobodina, V G Grinev, E V Prut Vysokomol. Soedin., Ser. B 36 1021 (1994) a 201. M P C Watts, A E Zachariades, R S Porter J. Mater. Sci. 15 426 (1980) 202. A S Argon, M I Bessonov Philos. Mag. 35 917 (1977) 203. I L Dubnikova, A I Petrosyan, V A Topolkaraev, Yu M Tovmasyan, I N Meshkova, F S D'yachkovskii Vysokomol. Soedin., Ser. A 30 2345 (1988) a 204. Ya E Beigel'zimer, V A Beloshenko, A P Borzenko, V N Varyukhin Mekhan. Kompozit. Mater. 31 834 (1995) a�Polym. Sci. (Engl. Transl.) b�Dokl. Chem. Technol., Dokl. Chem. (Engl. Transl.) c�Russ. J. Appl. Chem. (Engl. Transl.) d�Russ. J. Org. Chem. (Engl. Transl.) Promising aspects of the structural modification of polymers and polymer composites with the aid of high press

ISSN:0036-021X    

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Abstract. The synthesis and structure ± activity relationships of inhibitors of steroid aromatase which catalyses the last stage of a multistep biotransformation of cholesterol into estrogens, viz., aromatisation of C19-steroids into C18-phenolic steroids, are discussed. Compounds of the androstane series which are struc- turally related to the natural substrate, viz., androst-4-ene-3,17- dione, are the subjects of consideration.The review encompasses problems of synthesis of various substituted androstanes and their aromatase-inhibiting activities and structural requirements for selective specific aromatase inhibitors based on in vitro and in vivo structure ± activity studies of compounds synthesised, their bio- logical properties and the results of clinical trials. Special attention is paid to practical applications of aromatase inhibitors in the treatment of hormone-dependent mammary and ovarian tumours as well as benign prostatic tumours.In writing this report, the author has used all the information currently available in the chemical, biochemical, endocrinological and medicinal literature as well as in patents. The bibliography includes 173 references. I.Introduction The search for selective and specific therapeutic steroid agents in the past few decades has given the main impetus to the develop- ment of synthesis of modified steroids. It is commonly known that many physiological processes occurring in the organism are controlled by a relatively small set of steroid hormones produced upon cholesterol biosynthesis.These compounds represent small hydrophobic molecules which penetrate easily through mem- branes and interact with their specific receptors thus regulating gene transcription. One most important factor that stimulated studies of steroids was the discovery that analogues of some endogenous hormones can be used for controlling the synthesis and modifications of steroids involved in proliferation of malig- nant tumours.In many clinical situations, estrogens produced in normal or excess quantities play a prominent role in the pathogenesis of various diseases. Therefore, suppression of estrogen biosynthesis is one of the ways to achieve progress in the treatment of such diseases as mammary and ovarian tumours, endometriosis, etc.1± 4 The main efforts in the design of therapeutic preparations of the androstane series which are aromatase (estrogen synthase) inhibitors in their pharmacological effect have been directed at the synthesis of drugs against hormone-dependent mammarian tumours.This approach has indisputable advantages over cyto- toxic therapy which often causes serious complications.3, 5 At present, mammary tumours represent the most widespread type of malignant neoplasms and are the second major cause of mortality in women.In the USA alone, casuality of this disease amounts to about 130 000 annually.6 About a half of these mal- ignant tumours require a source of estrogens for their growth and development. The present-day endocrine therapy relies upon the use of either antiestrogens which act on tumour cells directly, via their specific receptors, or high doses of progestins the mechanism of action of which has not been finally established.An alternative strategy consists in the application of aromatase inhibitors (AI) which suppress production of estrogens in the organism. Aromatase represents an enzyme complex made up of two major components: a flavoprotein (NADPH-cytochrome P-450 reductase) which transfers electrons fromNADPHto the terminal enzyme and a specific form of cytochrome P-450 commonly known as aromatase cytochrome P-450.This protein is involved in the binding of the C(19) atom of the steroid substrate and catalyses the multistep reaction resulting in the aromatisation of the ring A of the steroid.7± 9 The conversion of D4-3-ketoandro- gens into phenolic estrogens is the ultimate stage in the multi- enzyme transformation of cholesterol into female sex hormones.Although the details of this mechanism are still questionable, it is believed 10,11 that aromatase performs three successive steps of the androgen oxidation. The first two steps yield 19-hydroxy- and 19-oxo-androgens. However, the site of incorporation of the third hydroxy group is not established.Presumably, hydroxylation occurs at C(2),12 which results in aromatisation of the ring A of androst-4-ene-3,17-dione (AD). The loss of the angular C(19)- methyl group and cis-elimination of the hydrogen atoms from positions 1 and 2 give estrone and formic acid.3, 12 ± 19 1 2 5 3 4 6 9 8 7 10 11 19 12 13 14 15 18 17 16 AD O O O2, NADPH I S LevinaNDZelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninskii prosp. 47, 117913 Moscow, Russian Federation. Fax (7-095) 135 53 28. Tel. (7-095) 938 36 14 Received 3 October 1997 Uspekhi Khimii 67 (11) 1068 ± 1093 (1998); translated by R L Birnova UDC 547.577.17.591.147 Substituted androstanes as aromatase inhibitors I S Levina Contents I. Introduction 975 II.Classification of steroid aromatase inhibitors and methods for assessment of their activities in vitro and in vivo 976 III. Methods for the synthesis of aromatase inhibitors of the androstane series and their activities 976 IV. Structure ± activity relationships in substituted androstanes as aromatase inhibitors 995 V. In vivo activities and applications of major steroid aromatase inhibitors 995 VI.Conclusion 996 Russian Chemical Reviews 67 (11) 975 ± 998 (1998) #1998 Russian Academy of Sciences and Turpion LtdThe enzymic oxidation described above yields an enzyme-bound derivative, which in turn gives the aromatic steroid by virtue of rapid elimination and simultaneous recovery of the unchanged, active enzyme. It is obvious that various structural modifications of the natural substrate (AD) can alter this elimination in such a way that the steroid intermediate will remain covalently bound with the enzyme rather than undergo conversion into the aromatic steroid.Elucidation of the mechanism of androgen aromatisation and demonstration of its possible inhibition has led to successful design and development of a broad range of AI.Since aromatisation is a unique reaction of steroid biosyn- thesis and estrogens are its final products, in vivo inhibition of aromatase does not affect other essential enzymes involved in this process. This is an important advantage of AI over other inhib- itors recommended for therapeutic use (see Refs 4, 20). Studies of the mechanism of aromatisation of the natural substrate (AD) have specified the type of its structural modifica- tion required for the design of AI. It consists in the introduction of substituents into the ring A or at adjacent positions with simulta- neous preservation, in the majority of cases, of polar functional groups at C(3) and C(17).The sites for substitution have been identified as C(1), C(4), C(6), C(7), and C(19).The nature of the substituents varies widely depending on the substitution site. Syntheses of AI often pose serious problems and are as such a considerable progress in both the design of synthetic approaches and their practical implementation. In each particular case, the synthesis of steroid AI employs original approaches which are based on modern methods of synthetic organic chemistry or a combination of several methods. Syntheses of structures with 2,19-bridges or a series of androsta-4,9(11)-diene-3,17-diones carrying a sulfur-containing group at C(10) (Section III.1) provide illustrative examples.Stereochemical aspects of the reactions are very important; modern physicochemical methods are employed for establishing the structure and stereochemistry of the reaction products.From the standpoint of their chemical structure, AI can be classified according to the site of substitution in the AD molecule. It is this principle that is used in the description of AI syntheses in Section III of the present review. However, it should be remem- bered that classification of AI according to the mechanism of their action may be different. In this review, this classification was used to analyse the relationships between the structure of AI and their inhibitory activity. The reviews on steroid AI that cover simultaneously the problems of their synthesis, relationships between their structures and biological properties and pharmacological data are virtually absent.One can mention the book by Zeelen `Medicinal Chem- istry of Steroids',20 which includes the analysis of steroid and nonsteroid AI, as well as the surveys 9, 15, 21 devoted predomi- nantly to the mechanisms of action of AI and their inhibitory activity in vitro.The present review describes the methods for the synthesis of substituted androstanes active as aromatase inhibitors, the rela- tionships between the structure of AI and the extent of their binding with the enzyme as a criterion for directed search for selective AI, as well as biological properties and practical appli- cations of the most potent steroid preparations currently avail- able.II. Classification of steroid aromatase inhibitors and methods for assessment of their activities in vitro and in vivo Steroid AI have come under the scrutiny of science with the publication of the pioneering work by Brodie et al.22 in 1973. In the subsequent two decades, a vast variety of steroid compounds have been synthesised and assayed for AI activity.In vitro inhibition of human placenta aromatase is the most common procedure used in the study of modified androgens.23 This method is used primarily for testing steroids as candidates for AI. The inhibition is evaluated by several parameters.The most important of them are the inhibition constant, Ki, which is determined in experiments where both AI and a steroid substrate (AD) are introduced, and the concentration of AI that results in 50% inhibition (IC50). The analysis of inhibition kinetics, which allows determination of the type of inhibitory activity for the given steroid compound, is also important.The following methods are used for in vivo studies of AI: (1) inhibition of ovarian estrogen secretion; (2) inhibition of periph- eral (extraovarian) aromatisation which is the main source of estrogens in post-menopausal women; (3) evaluation of the antitumour effect of AI in an experimental model of a 7,12- dimethylbenzo[a]anthracene (DMBA)-induced mammary tumour.According to the mechanism of their action, AI are classified into reversible and irreversible ones. Compounds which, like usual androgenic substrates, reversibly bind with the active centre of aromatase to form a complex that does not undergo rapid dissociation can serve as rather efficient AI. Competitive aroma- tase inhibitors that are capable of binding with iron of the cytochrome P-450 heme (e.g., due to the presence of a hetero atom with a specific stereochemical environment in their mole- cule) can sometimes manifest inhibitory activities both in vitro and in vivo.A great number of steroid AI are irreversible. Irreversible inhibition of the enzyme implies the conversion of a steroid substrate into an active complex which can irreversibly bind with the enzyme.These compounds are also referred to as `suicide inhibitors'.9, 24 ± 26 The latter have a number of salient advantages over competitive reversible AI owing to their high specificity and the ability for irreversible inhibition. Thus, irreversible AI pro- duce long-term inactivation of aromatase in vivo. III. Methods for the synthesis of aromatase inhibitors of the androstane series and their activities 1. 19-Substituted androstanes 19-Substituted androstanes are the most numerous and diverse aromatase inhibitors. It is quite natural because many investiga- tors consider the angular 19-methyl group to be the site of the primary enzyme attack leading to aromatisation. Formally, the functionalisation of the 19-methyl group of the steroid molecule is a very difficult task.This class of steroids are synthesised mostly from compounds having a 5(10) double bond or from 19-hydroxy(oxo)androstanes using diverse methods of synthetic organic chemistry for their further transformations, such as epoxidation followed by oxirane ring opening with appropriate reagents; substitution of 19-mesyl(tosyl)oxy- and 19-halogeno-derivatives; fluorination; the Wittig, Corey, and Mitsunobu ethynylations, and other sophisticated approaches.All the known methods for the synthesis of 19-substituted androstanes are discussed below in detail. 10b-Propynyl-steroids (19-ethynylandrostanes) synthesised in the 80's simultaneously and independently in several laborato- ries 14, 27 ± 31 pertain to irreversible suicide inhibitors.Some 19-ethynyl-steroids of the androstane series have been described in patents.32 ± 35 O2, NADPH O HO O O H O2, NADPH estrone HO 7HCOOH O O H HO 976 I S Levina19-Ethynylandrost-4-ene-3,17-dione 1 was obtained from commercially available estr-5(10)-ene-3,17-dione 2. Diketal 3 formed upon protection of the carbonyl groups reacted with N-bromosuccinimide to give the bromohydrin 4 which was easily converted into the 5a,10a-epoxide 5.Epoxide opening with an alkynylcuprate reagent gave the 5a,10b-disubstituted derivative 6 in 60% yield. Deprotection and elimination of the hydroxy group from position 5 of the 3,17-diketone 7 yielded the target andros- tane 1 in 23% yield starting from the oxirane 5.14 Structurally related diastereomeric 19-ethynyl-19-hydroxy derivatives 10a,b were obtained from the bisketal-aldehyde 8.The latter reacted with ethynylmagnesium bromide to give a chromatographically separable mixture of (19S)-9a and (19R)-9b derivatives in a 3 : 2 ratio (>90% yield). The authors 27 ascribed the (19S)-configuration to the main product by analogy with the stereoselectivity of the reaction of the diketal-aldehyde 8 with methyllithium.After removal of the ketal protection, the dike- tones 10a,b were obtained in the individual state and further oxidised into the triketone 11 27 (Scheme 1). The acetate 12 prepared from 19-ethynyl-19-hydroxy-steroids 9a,b served as the starting compound in the synthesis of an analogous allenic AI (14) 29 based on cuprate-initiated reductive elimination of acetic acid.The allene 13 was converted into the final product 14 after deprotection (see Scheme 1). Yoshida et al.29 have also developed an original procedure for the synthesis of 19-ethynyl-AD 1 using 6b-hydroxyestrenedione bisketal 15 as the starting compound. The vinyl ether 16 obtained from it was subjected to the Claisen rearrangement (66% yield). The Wittig reaction of the aldehyde 17 with chloromethylene- phosphorane gave a mixture of cis- and trans-vinyl chlorides 18, which were dehydrochlorinated with lithium diisopropylamide at O 2 3 a O O O O b O d O O O 4 5 c O O Br OH C O O O O OH Me3SiC 6 e (a) HO(CH2)2OH, H+, PhH; (b) N-bromosuccinimide (NBS), DMF; (c) NaOMe, MeOH; (d) [Me3SiC:CCH2]2CuLi, Et2O; (e) H+, Me2CO.O OH C Me3SiC O 7 O c 1 C HC O a 15 O O OH O O b 16 O O O (a) H2C=CHOEt, Hg(OAc)2, 78%; (b) collidine, 165 8C, 4 h, 66%; (c) ClCH=PPh3, 84%; (d ) lithium diisopropylamide (LDA),770 8C, 1 h, 95%.H+ O 1 (79%) O C HC 19 O O C HC c d O O H Cl O O OHC 18 17 c 8 a O O O O O H (a) HC:CMgBr, D, THF, 2 h, 94%; (b) 10% H2SO4, Me2CO; (c) CrO3, H2SO4; (d) Ac2O, Py; (e) MeLi, CuI, Et2O,778 8C, 50%; ( f ) p-TsOH, Me2CO, 70%. 11 (*70%) C O O O HC 9a 10a (79%) O H OH C HC O b 9b 10b (88%) O OH H O C HC b e f d 9a,b C O O OAc HC 12 13 C O O CH2 14 O O C H2C (19R)-9b (19S)-9a + O O H OH C HC O O OH H C HC Scheme 1 Substituted androstanes as aromatase inhibitors 977770 8C to yield the ethynyl derivative 19 in 95% yield. Removal of the ketal protection gave the final product, viz., 19-ethynyl-AD 1, in high yield. The steroid 1 and its derivatives have been extensively assayed as AI.It was found that 19-ethynyl-AD 1 and its 19-hydroxy- (10a,b) and 19-oxo-derivatives (11) are irreversible suicide inhib- itors of aromatase and manifest high specificity 14, 15, 26, 28, 29, 31, 36 (the data on their inhibitory activities are summarised in Table 1). The steroid 1 possessed high activity on oral administration with respect to the inhibition of ovarian aromatase.It is noteworthy that 19-ethynyl-17-propionyloxyandrost-4-ene-3-one manifests only in vivo inhibitory activity.37 The allenic derivative AD 14 also pertains to irreversible AI and has rather high Ki value.29 Several 19-halogeno-substituted androstanes 22 were obtained from the bisketal 21 by substitution of 19-methane- sulfonates, which is well known in the series of D5-3-acet- oxysteroids; 19-mesyloxy diketone 20 being used as the starting compound.38 19-Difluoroandrostane 25 was obtained by selective fluori- nation of 17-benzoyloxy-3-oxoestr-4-en-19-al (23) with an excess of diethylaminosulfur trifluoride in a mixture of dichloromethane and trichlorofluoromethane.Removal of the benzoyl protection and oxidation of the resulting alcohol gave 19-difluoro-AD (25) which manifested moderate inhibitory activity 39, 40 (see Table 1).The synthesis of androstanes with functional groups at position 19 containing sulfur or nitrogen exploited the ability of 19-iodo derivatives 26a,b for nucleophilic substitution without homoallylic rearrangement.41 This approach was used to obtain a series of 19-substituted androstanes 27a,b and 22d ± f.It is appropriate to mention in this connection the synthesis of 19-azido-AD 22e described in a patent.42 Compound 22e was obtained by esterification of 19-hydroxy-AD 28 with trifluoro- acetic anhydride with subsequent treatment of the ester 29 with sodium azide in DMF. 19-Substituted AD 22d ± f competitively inhibit aroma- tase.38, 42 Of special interest is the 19-methylthio derivative 22f which is the first highly efficient AI (Ki=1 nM); its inhibitory effect is determined not only by its interaction with the steroid- binding site of aromatase but also by coordination of the steroid sulfur atom with the heme iron of cytochrome P-450.41 The attention of several research groups was focussed on 10b- and 19-sulfur-containing AD derivatives as aromatase inhibitors.Compounds with the sulfur atom linked directly with C(10) or separated from the C(10) atom by one or several methylene groups have been synthesised and assayed. Thus the introduction of the SH group at position 10 of estr-4-ene-3,17-dione resulted in an efficient suicide AI 32. The opening of the 5a,10a-epoxide ring of estrane 5 with potassium sulfide in ethylene glycol in the presence of 18-crown-6 gave 5a,10b-disubstituted estrane 30 in 70% yield.The latter was converted in two steps into 10b-mercaptoestr-4- ene-3,17-dione (32) via 5a-hydroxy diketone 31. The yield of the final product was 10% with respect to estr-5(10)-ene-3,17-dione 2.14 The corresponding 17b-hydroxy derivative possessing aroma- tase-inhibiting activity was synthesised by Bednarski and Nel- son 14 following an analogous scheme.43 In vitro studies of the biological activity of the thiol 32 revealed that the free thiol group is required for the binding to the enzyme, because the correspond- ing acetylthio derivative was inactive.19, 43 O 20 MsO O 21 O O MsO O O X=Cl (a), Br (b), I (c), CN (d), N3 (e), SMe (f). 22a ± f O X O (a) excess Et2NSF3 (DAST), CH2Cl2 ± FCCl3 (1 : 1), 20 8C, 48 h; (b) OH7; (c) CrO3, H2SO4. 23 OBz O O H 24 OBz O F F a b,c 25 O F F O 26a,b R2 I R1 O a R2 X R1 27a,b b, c, d, or e or f, g, h, i R1=OAc, R2=H(a); R1, R2=O(CH2)2O (b); X=CN, N3, SSO2Me; (a) NaCN (X=CN); NaN3, HMPA (X=N3); NH4SSO2Me (X=SSO2Me); (b) KOH, MeOH; (c) CrO3, H2SO4; (d ) Al(OPri)3, cyclohexanone (CHex) (27a, X=CN, N3); (e) (COOH)2, EtOH (27b, X=CN, N3); ( f ) LAH, THF; (g) MeI; (h) K2CO3; (i) Al(OPri)3, CHex (27a, X=SSO2Me); (j ) (CF3CO)2O, Py; (k) NaN3, DMF. 22d: X=CN, 22e: X=N3, 22f: X=SMe. 22d ± f O X O 22e (45%) j k 28 HO O 29 O CF3OCO O 30 OH O O SH O O b 5 a O O O O O 978 I S LevinaTable 1. Data on in vitro inhibition of aromatase from human placenta by 19-substituted androstanes .R X Y The position of a double bond a Ki /nM IC50 /nM Ref. D1 D6 D9(11) Me (ADD) O O + 7 7 320 24 Me (114) O O + +7 180 171 Me (183) O O 7 + 7 42 24 Me H2 O 7 7 7 37 152, 145 Me O H2 + 7 7 1100 172 Et O O 7 7 7 9 15 (CH2)2Me O O 7 7 7 3000 15 CH=CH2 O O 7 7 7 3 15, 30 CH27C:CH (1) O O 7 7 7 4.3 31 23 28 68 15 31.3 9 CH27CH=CH2 O O 7 7 7 13 31 56 30 C:CMe O O 7 7 7 7 31 CH2C:CMe O O 7 7 7 26 31 CH(OH)C:CH (10) O O 7 7 7 49 31 COC:CH (11) O O 7 7 7 350 31 CH=C=CH2 (14) O O 7 7 7 14 31 Me CH2 O 7 7 7 13 148 CH2OH CH2 O 7 7 7 4.7 148 CHO CH2 O 7 7 7 24 94, 148 CHF2 (25) O O 7 7 7 1000 1300 15, 40 CH2N3 (22e) O O 7 7 7 22.5 31, 41 SH O OH 7 7 7 106 43 CH2SH (35) O O 7 7 7 34 43 CH2SMe (22f) O O 7 7 7 1 42 41, 46 (CH2)2SMe (55) O O 7 7 7 0.5 22 49, 46 (CH2)2SMe (40a) O O 7 7+ 12 46 (CH2)2SEt (40b) O O 7 7 + 870 46 (CH2)3SMe (51) O O 7 7 + 10000 46 (CH2)2SAc (40c) O O 7 7 + 10000 46 (CH2)2SPh (40d) O O 7 7 + 400 46 (CH2)2S(CH2)2OH (40e) O O 7 7 + 1000 46 (CH2)2SCHF2 (40f) O O 7 7 + 1000 46 (CH2)2SSC6H3(NO2)2 (40g) O O 7 7 + 1000 46 (CH2)2SCH2C:CH (40h) O O 7 7 + 1000 46 (CH2)2SCH2Me (40i) O O 7 7 + 570 46 (CH2)2SCH2CH=CH2 (40j) O O 7 7 + 270 46 (CH2)2SCH=CH2 (40l) O O 7 7+ 50 46 (CH2)2SSMe (40m) O O 7 7 + 1000 46 (CH2)2SSPh (40n) O O 7 7 + 1000 46 (CH2)2SC3H5-cyclo (40o) O O 7 7 + 1000 46 (CH2)2SC:CH (40p) O O 7 7 + 330 46 [(19R)-59] O O 7 7 7 7 51 [(19S)-59] O O 7 7 7 75 51 [(19S)-60] O O 7 7 7 1 51 [(19R)-60] O O 7 7 7 75 51 [(20S)-64] O O 7 7 7 22 55 [(19R)-76] O O 7 7 7 3.4 56 [(19S)-76] O O 7 7 7 56 56 [(19R)-78] O b-OH 7 7 7 30 56 [(19S)-78] O b-OH 7 7 7 693 56 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19 X A R 18 Y O S O H2C N H Substituted androstanes as aromatase inhibitors 979The homologous 19-mercapto-AD 35 was obtained by the reaction of the 19-triflate 33 with potassium ethyl xanthate in the presence of 18-crown-6 with subsequent cleavage of the thio derivative 34.43 The inhibition constant of the steroid 35 is 34 nM.43 It represents an irreversible AI; its activity is 16 times as high as that of formestane both in vitro and in vivo (see Section II).However, the thiol 35 has a limited use because of its insufficient pharmaco- logical stability due to the presence of a free thiol group. To overcome this limitation, its ethylthio derivative, viz., 19-ethyl- thio-AD, has been synthesised 44 (for the relevant patent, see Ref. 45). One of recent papers devoted to 19-functionalised andro- stanes presents the results of extensive studies on the synthesis and structure ± aromatase affinity relationships of several series of androstane-4,9(11)-diene-3,17-diones carrying a sulfur-contain- ing group, [(CH2)nSR], at the C(10) atom, where n=0 ± 3 and R stands for various alkyl, acyl and aryl substituents.46 19-Ethoxy- carbonylandrosta-4,9(11)-diene-3,17-dione 37 obtained from the allylic alcohol 36 by the [3.3]-Claisen rearrangement was used as the starting compound in these syntheses (Scheme 2).(a) KSH, HO(CH2)2OH, 18-crown-6, 70%; (b) H+, Me2CO, 78%; (c) NaOH, MeOH, 32%.c 31 OH SH O O 32 SH O O 28 33 a b HO O O CF3O2SO O (a) (CF3SO2)2O, Py; (b) EtOCSSK, 18-crown-6, THF, 20 8C, 20 h; (c) H2N(CH2)2NH2, THF. c 35 HS O O 34 EtOCS S O Table 1 (continued). R X Y The position of a double bond a Ki /nM IC50 /nM Ref. D1 D6 D9(11) (CH2)37C(2)-b (86) O O 7 7 7 24 59 2b,10b-(CH2)2 O O 7 7 7 70 11 2b,10b-CH2CH(OH) [(R)-82a] O O 7 7 7 870 11 2b,10b-CH2CH(OH) [(S)-82b] O O 7 7 7 926 11 CH(OH)CH2Br [(19R)-80] O O 7 7 7 27 11 CH(OH)CH2Br [(19S)-80] O O 7 7 7 1130 11 CH(OH)CH2Cl [(19R) : (19S)=9: 1] O O 7 7 7 63 11 CH(OH)CH2I [(19R) : (19S)=9: 1] O O 7 7 7 11 11 a Symbols +and7denote the presence and absence of the double bond in this position, respectively.Scheme 2 d j g, h 39 O O O O OMs i, f 40f ± j 42 O O O O O 41 SH O O O O e, f 40a ± e O O SR k, f 43a,b R=Me (a, 51%), Et (b, 55%) O O RS SR O O O O OH 38 O O HO 36 a O O O OEt 37 b, c 40: R=Me (a), Et (b), Ac (c), Ph (d), (CH2)2OH (e), CHF2 (f), 2,4- (NO2)2C6H3S (g), CH2C:CH (h), CH2SMe (i), CH2CH=CH2 (j); (a) (EtO)3CMe, EtCO2H, 137 8C, 4 h, 81%; (b) HO(CH2)2OH, CH2Cl2, D, 8 h, 71%; (c) LAH, THF, 750 to 20 8C, 2 h, 85%; (d ) MsCl, Et3N, CH2Cl2, 0 8C, 100%; (e) RSNa, DMF; (f ) HCl; (g) AcSH, PPh3, diethylazodicarboxylate (DEAD), THF, 20 8C, 2.5 h, 60%; (h) NH2NH2, THF, 730 8C, 72 h, 97%; (i) But OK, THF, RHal; ( j ) (COCl)2, DMSO, CH2Cl2, 750 8C; (k) RSH, BF3 .Et2O, 778 to 20 8C. 980 I S LevinaThus the reaction of steroid 36 with triethyl orthoacetate in the presence of propionic acid gave the ethoxycarbonyl derivative 37 in >80% yield.47 The latter was reduced to alcohol 38 via the bisketal.Several methods presented in Scheme 2 have been elaborated for its conversion into various 19-thiomethyl andros- tane derivatives. One of them consists in conversion of the alcohol 38 into the corresponding mesylate 39, subsequent substitution of the mesyloxy group by various thiolates and deprotection to give compounds 40a ± e.The first member of this series, viz., 3,17- diketo analogue of compound 41, could not be obtained using this approach; various attempts at its deprotection gave compound 46, which is the product of intramolecular Michael addition (Scheme 3). The thiol 41 was used as the starting compound in the synthesis of thio derivatives 40f ± j. This has been obtained by the Mitsunobu reaction from the alcohol 38 and thioacetic acid in the presence of PPh3 and diethyl azodicarboxylate followed by hydrazinolysis.Alkylation of the thiol group of compound 41 with alkyl halides or 2,4-dinitrophenylsulfenyl chloride gave compounds 40f ± j in good yields (see Scheme 2). Thioacetals 43a,b were prepared by the reaction of the aldehyde 42 obtained by the Swern oxidation of the alcohol 38 with methane- or ethane- thiol, respectively, in the presence of boron trifluoriole etherate.Compounds 40k ± p were synthesised by the reaction of the disulfide 40g with various organometallic reagents. The methyl thiocarboxylate 45 was obtained from the key compound 37 in three steps which included hydrolysis of the ester function, activation of the resulting acid 44 with ethyl chloroformate and reaction with sodium methanethiolate (Scheme 3).Methylthiovinyl steroids 48 and 50 have also been described.46 The acetoxy derivative 47 was deacetylated into the corresponding 3,17-diol which was introduced into the reaction with diethyl methylthiomethylphosphonate and butyllithium. The thus pre- pared (10E)-methylthiovinyl derivative was further oxidised into steroid 48.An analogous reaction of the homologous aldehyde 49 with (methylthiomethyl)triphenylphosphorane gave a 1 : 1 mixture of chromatographically separable E- and Z-isomers 50 in total yield 76%. Hydrogenation of this mixture in the presence of the Wilkinson catalyst gave compound 51 in 80% yield which contained a 3-methylthiopropyl residue at C(10).And, finally, the opening of the 5a,10a-epoxide ring of estrane 52 with lithium methanethiolate which is analogous to that described above for compound 5,14, 43 and subsequent deketalisa- tion and elimination of the 5a-hydroxy group resulted in D4,9(11)- steroid 53 (17%). The opening of the 5a,10a-epoxide ring of the estrane 5 with monolithium acetylide with subsequent partial hydrogenation of the hydroxy alkyne and hydroboration of the resulting alkene gave the diol 54.Its mesylation and subsequent reaction of the mesylate with sodium methanethiolate followed by acid hydrol- ysis gave the ring C-saturated analogue of the steroid 40a, viz., 19-methylthiomethyl-AD 55. (a) KOH, MeOH (78%); (b) MeSCH2PO(OEt)2, BuLi (76%); (c) Al(OPri)3, CHex (74%). 47 48 O SMe O H AcO O OAc a, b, c 49 O O O + 7 MeSCH2PPh3Cl BuLi 51 (80%) O MeS (Z,E)-50 (76%) O H SMe H2, RhCl(PPh3)3 PhMe (a) MeSLi, THF, 1 h, 20 8C; (b) 6N HCl; (c) 0.1N NaOH (17%).a, b, c 53 O MeS O 52 O O O O O 40k7p 40:R=C:CSiMe3 (k), CH=CH2 (l), SMe (m), SPh (n), cyclo-C3H5 (o), C:CH (p); (a) KOH, MeOH, 18 h, 95%; (b) ClCO2Et, Et3N, 20 8C, 1 h; (c) MeSNa, DMF, 2 h; (d) HCl; (e) 2,4-(NO2)2C6H3SCl, Py, CH2Cl2, 20 8C, 0.5 h; ( f) RM, THF.O O SR e 40g f 41 d 46 S O O O O O O O OH 45 (30%) 44 O O O SMe b, c, d 37 a Scheme 3 e ± g 54 O O OH OH O O a ± d 5 O O O O O Substituted androstanes as aromatase inhibitors 981The results of in vitro and in vivo studies of activities of the majority of compounds synthesised by Lesuisse et al.46 suggest that the steroids with one- or two-carbon chains having a terminal thioether function at position 10 (e.g., 22f, 40a,b, 55) are the most efficient AI. The presence or the absence of the 9(11)-double bond does not practically affect their inhibitory activity.48 Compound 40a was the most reactive among other compounds assayed and was considered 46 to be one of the best AI available to date.On the contrary, the steroids with the sulfur atom attached directly to the steroid skeleton (53) or separated from it by three carbon atoms (51) have practically no affinity for the enzyme.The key property of the efficient AI of this series is the ability of the sulfur atom to react with the heme iron of the porphyrin moiety of the enzyme (compounds 22f, 40a and 55), whereas compound 53 reacts only with the lipophilic steroid-binding centre of the enzyme. This behaviour is consistent with the hypothesis 14, 41 that the incorpo- ration of one methylene group between the sulfur atom and position 10 of the steroid skeleton alters the character of the interaction of the steroid substrate with the enzyme.The fact that AI with a two-carbon chain at C(10) react most efficiently with the heme Fe(III) allowed Lesuisse et al.46 to suppose that the iron atom is separated from the 19-methyl group of the substrate by a distance equivalent to two methylene groups at the oxidation steps of AD.This hypothesis is consistent with the generally accepted mechanims of aromatisation.15, 16 The effect of the substituent at the sulfur atom in the molecule of AD analogues on the inhibition of aromatase has also been studied.Sulfur-containing steroids with methyl, vinyl or ethynyl substituents at the sulfur atom (compounds 40a, l and p, respectively) possessed the highest affinity for the enzyme. Substitution of the ethyl group (40b) for the methyl group (40a) decreases the activity of the steroid substrate 72-fold. Compounds with hydroxyethyl (40e) and cyclo- propyl groups (40o) are completely devoid of activity, while compounds with allyl (40j), phenyl (40d), prop-2-ynyl (40h) and methylthiomethyl (40i) groups possess certain activity.Disulfides 40g,m,n, thioacetals 43a,b and analogues 48 and 50 with a double bond in the side chain at C(10) are also inactive as AI. Compound 40a, which is the most active in this series, decreases the estradiol level and inhibits the growth of endocrine tumours in vivo when used at a dose of 0.4 mg kg71.49 Yet another type of 19-substituted androstanes, viz., 10b- (oxiran-2-yl)- and -(thiiran-2-yl)-steroids, represent competitive reversible AI.The synthesis of diastereomeric 10b-(epoxy- ethyl)estr-4-ene-3,17-diones (19S)-59 and (19R)-59 was carried out in four steps.50 ± 52 Reaction of the tetrahydropyranyl derivative of 19-aldehyde 56 with dimethylsulfonium methylide (or dimethylsulfoxonium methylide) in THF gives a mixture of oxiranes (19R)-57 and (19S)-57 in 26% and 74% yields, respectively.Under other reaction conditions, i.e., in DMSO in the presence of dimethyl- sulfonium methylide at 0 8C, the ratio of diastereomeric products is changed: the (19R)-57 isomer becomes predominant.51 Depro- tection and chromatographic separation of diols (19R)-58 and (19S)-58 with subsequent Oppenhauer oxidation of each of the diastereomeric alcohols gave individual isomers (19R)-59 and (19S)-59, respectively. Steroid thio analogues (19R)-60 and (19S)-60 were obtained from oxiranes 59 by reaction of the latter with excess of triphenylphosphine sulfide and picric acid in benzene.53 The configuration of isomeric 10b-oxiranes 59 (R or S) was established from X-ray analysis.It was shown that the oxirane ring in the stereoisomer (19R)-59 is positioned over the ring A in the direction towards C(1) and C(2) in a manner similar to that of the 19-hydroxy group in 19-hydroxy-AD 28 . On the contrary, the three-membered ring in the molecule of (19S)-59 is positioned over the ring B and directed toward C(6).31, 51, 54 Presumably, it is these stereochemical peculiarities that are responsible for the difference in the binding and inhibitory activities of these com- pounds.Both of them are potent competitive AI with Ki 7 and 75 nM, respectively. Thiiranes (19R)-60 and (19S)-60 manifest an even greater stereoselectivity of binding: Ki for (19R)-60 is 1 ± 2 nM and that for (19S)-60 is 75 nM.51 In another study 54 it was shown that (19R)-isomers of oxiranes and thiiranes are more active (36-fold and 80-fold, respectively) than the corresponding (19S)-diastereomers. Spectroscopic studies with purified aroma- tase showed that the inhibition involved reversible binding of the oxygen (or sulfur) atom of the oxirane (or thiirane) with the heme iron of the enzyme.15, 51, 54 The knowledge of the spatial structure of these compounds made it possible to assume that the modified steroid inhibitor is bound in exactly the same way as does the natural substrate (AD) and that the heme iron is localised in close proximity to the 19-methyl group of the androgen molecule.54 It is of note that with neither 19-hydroxy, nor 19-methoxy derivatives of AD is such coordination possible.15 Thus, coordination with the heme iron of the enzyme, together with the binding selectivity inherent in 10b-(oxiran-2-yl)- and -(thiiran-2-yl)- steroids make these AD analogues highly specific aromatase inhibitors.(a) LiC:CH, ethylenediamine (EDA), 50 8C, 96 h, 81%; (b) H2 ± 10% Pd(OH)2/BaSO4, EtOAc ± Py; (c) BH3 .Me2S, THF, 35 8C, 4 h; (d ) 30% H2O2, NaOH, 0 to 20 8C, 12 h, 67%; (e) MsCl, Et3N,CH2Cl2, 0 8C, 45 min; ( f ) MeSNa, DMF, 20 8C, 1 h; (g) HCl, EtOH, 20 8C, 3 h. 55 (52%) O SMe O THPO OTHP O a THPO O 56 (19R)-57, (19S)-57 b +7 + 7 (a) Me2SCH2 (or Me2SOCH2),THF or DMSO; (b) Py.HSO3C6H4Me, 20 8C, 4 h; (c) Al(OPri)3, CHex; (d) Ph3PS is picric acid, PhH, 80 8C, 16 h.(19R)-58, (19S)-58 HO O OH c + (19S)-59 H O O O (19R)-59 O O O d H d (19S)-60 H O S O (19R)-60 O O H S 982 I S LevinaThe synthesis of homologous 19-(oxiran-2-yl)- (63) and 19-(thiiran-2-yl)-AD (64) continued the series of investigations aimed at establishing the localisation of the aromatase heme iron during its binding with the steroid substrate.55 The methodology of the synthesis of these compounds is analo- gous to those used in the synthesis of 10b-(oxiran-2-yl) and -(thiiran-2-yl) steroids 52 with 3b,17b-dihydroxyandrost-5-ene 19-carbaldehyde 61 as the starting compound.30 Unlike the 10-formyl analogue 56, this reacted with both dimethylsulfonium and dimethylsulfoxonium methylide to give a chromatographi- cally separable mixture of diastereomeric oxiranes (20R)-62 and (20S)-62 of the same composition (3 : 1) in 75% yield.This result is rationalised by higher spatial accessibility of the aldehyde group in the molecule of compound 61. After separation into individual diastereomers, epoxy diols 62 were converted into diketones (20R)- 63 and (20S)-63 by the Collins oxidation with subsequent isomer- isation of resulting D5-3,17-diketones with diazabicyclonene.The configuration of epoxide (20R)-63 was established by X-ray analysis. Reaction of this oxirane with triphenylphosphine sulfide and picric acid gave thiirane (20S)-64 in 76% yield. Analogously, thiirane (20R)-64 was obtained from (20S)-63 in 79% yield. The R- and S-configuration of thiiranes 64 was ascribed on the basis of the established stereochemistry of the oxirane precursors and the known stereochemical regulation of thiirane formation.55 Oxir- anes 63 and thiiranes 64 are competitive AI; the latter possess higher activity than the corresponding oxiranes. (20S)-Isomers of both compounds are more efficient inhibitors than 20R-diaster- eomers.The inhibition constant of the strongest inhibitor, viz., the thiirane (20S)-64, is 22 nM.The oxygen and sulfur atoms in the oxirane (20S)-63 and the thiirane (20S)-64 molecules coordinate with the aromatase heme iron during their interaction with the enzyme similarly to the corresponding (19S)-diastereomers 59 and 60. Here, the weaker inhibitor possesses weaker affinity for the heme iron of cytochrome P-450.The configuration of the 20-centre of the three-carbon chain at C(10) is crucial for the stereoselectiv- ity of inhibition.55 Thus, strong competitive inhibition of aroma- tase and coordination of the heteroatom in the steroid substrate are observed even in those cases where the oxygen or the sulfur atom in the three-membered ring are separated from C(10) by a methylene group (cf.Ref. 46). Having considered theADanalogues containing an oxygen or a sulfur atom at positions 10 or 19, one may conclude that sulfur- containing steroids are the most potent inhibitors and that compounds with a methylene group between the steroid skeleton and the heteroatom manifest higher inhibitory activity. A charac- teristic feature of compounds of this series is the stereoselectivity of their interaction with the enzyme.Synthesis of AD derivatives containing nitrogen at position 19, which bind with iron of the aromatase heme by virtue of coordination similarly to sulfur-containing steroids was described by Robinson in 1993,56 although attempts to synthesise such derivatives were undertaken earlier.57 Thus 10b-aminoestr-4-ene- 3,17-dione 66 was obtained by the Curtius rearrangement of the corresponding derivative of 19-acid 65.However, all the attempts to obtain its homologue, viz., 19-aminomethyl-AD 69, failed, although its diketal derivative 68 was obtained in high yield from 19-hydroxy-AD 28. Compounds 66 and acetamido and trifluoro- acetamido derivatives of diketone 69 displayed low AI activity.57 Oxime 74 obtained from the accessible 3,17-diacetate of the triol 70 in seven steps (Scheme 4) was the key compound in the synthesis of 10b-(aziridin-2-yl)-steroids.56 Oxidation of com- pound 70 with tetrapropylammonium perruthenate in the pres- ence of N-methylmorpholine N-oxide as a co-oxidant gave 19-aldehyde diacetate 71 in high yield; the latter was further converted into diol and then into bis-tert-butyldimethylsilyl ether 72.Methylation of the aldehyde 72 with methyllithium occurred in quantitative yield. Oxidation of the 19-hydroxy-19- methyl derivative and selective desilylation of the thus obtained 10b-acetyl derivative gave 17-silyl ether 73. The latter was refluxed with hydroxylamine hydrochloride in pyridine to give a mixture of key oximes 74a,b in 70% yield. On treatment with lithium aluminium hydride in THF, the oxime 74b was converted into a mixture of (aziridin-2-yl)-steroids (19R)-75 and (19S)-75 isolated in the individual states with 35% and 53% yields, respectively.Each of them was further converted into (19R)- and (19S)-10b- (aziridin-2-yl) diketones 76 (in 39% and 41% yields, respectively) via the corresponding diols. 17b-Hydroxy analogues (19R)- and + O (2 0S)-63 (2 0R)-63 a b, c 61 HO OH O OH (2 0R)-62, (20S)-62 HO O d O O O O O d + + 7 7 (a) Me2SCH2 (or Me2SOCH2), THF, 4 8C (75%); (b) CrO3 .Py, CH2Cl2, 25 8C, 20 min; (c) diazabicyclononene (DBN), CH2Cl2, 25 8C, 18 h; (d) Ph3PS, PhH, D, 18 h. (2 0S)-64 (76%) O O S (2 0R)-64 (79%) O O S 66 O H2N O c, d 11 O O O 67 O O HO O O b 28 O 65 O OH 28 O O HO a O (a) CrO3±H2SO4; (b) Py.CrO3 (93%); (c) NH2OH. HCl (88%); (d) Ni/Ra, EtOH, 2 MNaOH (89%). 68 O O H2N 69 O O H2N Substituted androstanes as aromatase inhibitors 983Scheme 4 (19S)-78 were synthesised from the aziridines (19R)- and (19S)-75 via intermediate D4-3-ketones 77. The stereochemistry of these compounds was established by deamination of diastereomeric aziridines 76 and 78 with sodium nitrite in aqueous acetic acid into the known 10b-vinylestr-4-ene-3,17-diones and the corresponding 17b-alcohols.The configuration of the (aziridin-2-yl)-steroid (19R)-78 was also confirmed by X-ray analysis. Compounds (19R)- and (19S)-76 as well as (19R)- and (19S)- 78 are potent stereoselective AI with Ki=3.4, 56.0, 30.0 and 693 nM, respectively. 10b-(Aziridin-2-yl)-steroid (19R)-76 mani- fested the highest efficiency.Spectroscopic analysis of its inter- action with preparations of microsomal aromatase have shown the coordination of the nitrogen atom of the aziridine ring with the iron atom of the cytochrome P-450 heme. Earlier, such a coordi- nation was established for 19-azido-AD 22e.17 These results suggest that the orientation and accessibility of the nitrogen atom are crucial for efficient coordination.56 It should be noted that N-acetylation of compounds 76 strongly decreases their inhibitory activity.56 The recently described 10b-aziridinoestr-4-ene-3,17 dione was less active in comparison with the 10b-aziridine (19R)-76.58 The synthesis of hydroxylated 2,19-bridged structures related to AD was described in 1991.11 The key compound in this synthesis was 10b-epoxide 79 [a mixture of (19R)- and (19S)- isomers] obtained by reaction of the aldehyde 8 with dimethylsul- fonium methylide.The opening of the epoxide ring and removal of the ketal protection with 48% HBr in acetone gave a mixture of bromohy- drins 80 [the (19R)-80/(19S)-80 ratio was 7 : 1]. Chromatograph- ically separable trimethylsilyl derivatives 81a and 81b were obtained by the action of bis(trimethylsilyl)-acetamide in DMF.Intramolecular alkylation at C(2) under the action of lithium bis(trimethylsilyl)amide in THF and subsequent treatment of the reaction mixture with an acid complete the synthesis of diastereo- meric alcohols 82a,b. The synthesis of diketone 86, which is homologous to these bridge structures, was carried out in five steps.59, 60 70 AcO OAc HO a 71 AcO OAc O b, c d, a, e 72 TBDMSO OTBDMS O h, i j (a) Tetrapropylammonium perruthenate (TPAP), N-methyl- morpholine N-oxide (NMO), CH2Cl2, 20 8C; (b) 10% KOH, MeOH, 90%; (c) ButMe2SiCl ((TBDMS)Cl), DMF; (d ) MeLi, Et2O; (e) TBAF, THF; ( f) NH2OH.HCl, Py, D; (g) LAH, THF, D, 18 h; (h) HF, H2O, EtOH, 20 8C, 2.5 h; (i ) Al(OPri)3, CHex; ( j ) Al(OPri)3, 4-methylpiperi- done; (k) HF, H2O, MeCN, 20 8C, 2 h.O g f 73 HO OTBDMS O R=TBDMS (a), H (b). 74a,b RO OTBDMS NOH (19R)-75 (35%), (19S)-75 (53%) H HO OTBDMS N (19R)-76, (19S)-76 H O N O (19R)-77, (19S)-77 H OTBDMS N O k (19R)-78, (19S)-78 H OH N O 8 O O H a b (19R)-79, (19S)-79 O O O (19R)-80, (19S)-80 O OH Br O c O O O O 82a (42%) 82b (54%) OH O OH O 81a + OSiMe3 Br O 81b OSiMe3 Br O d d +7 (a) Me3SI, NaCH2SOCH3, DMSO, THF, 98%; (b) HBr, H2O, Me2CO; (c) CH3C(=NSiMe3)OSiMe3 (BSA), DMF; (d ) LiN(SiMe3)2, HCl, THF.(a) (EtO)2P(O)CH2CO2Et; (b) Mg, MeOH; (c) LAH (93%); (d ) TsCl, Py, 60%; (e) H+ (60%); ( f ) LiN(SiMe3)2, THF, 778 8C (62%). 85 O O TsO 86 O O e, f 8 a b c, d 83 (89%) O O O O EtO2C 84 (63%) O O EtO2C 984 I S LevinaIn this case, too, the same bisketal 8 was used as the starting compound.Its reaction with triethyl phosphonoacetate gave the unsaturated derivative 83 in high yield, and the double bond was reduced with magnesium in methanol. The ester function in the thus prepared compound 84 was reduced into the alcohol function which was then tosylated. Removal of 3,17-diketal protection in the tosylate 85 and cyclisation of 19-(2-tosyloxyethyl)-AD with lithium bis(trimethylsilyl)amide in THF gave the final product, 2b,19-ethylene-AD 86. The section devoted to bridge structures can be completed by the description of a synthesis of 2b,19-methyleneoxy-AD 89.The methoxyethoxymethyl ether (MEM) 87 prepared from 19-hydroxymethyl-AD 28 was converted into the homoannular dienol 88 silyl ether.The latter underwent Lewis acid-catalysed intramolecular alkoxyalkylation at C(2) involving the acetal carbon atom of the MEM-protective group to give the bridge steroid 89. It is noteworthy that the MEM ether group played the role of both the protective group and the reagent in these conversions. Some sulfur-containing analogues of the steroid 89 have been recently patented.59 2b,19-Bridged steroids 82a,b, 86 and 89 appeared to be sufficiently active AI.11, 59 ± 61 Bromohydrins 80 also manifested a certain AI activity (see Table 1).11 2. 4-Substituted androstanes Of all substituted androstanes, particularly of those substituted at position 4, 4-hydroxyandrost-4-ene-3,17-dione (4-OHA) 91 is the most popular and well-studied.It was first described in 1977 62 as a competitive AI. Later, it was shown to represent a suicide irreversible AI; according to different estimates, its Ki ranges from 1 to 50 nM.19, 55, 62 ± 64 This steroid is the basis of one of a few (if not the only) steroid inhibitors of aromatase that is already used in clinical practice.65 This compound is known under the trade name `formestane' and is used for the treatment of hormone- dependent mammary tumours in women in the post-menopausal period.8, 15, 48, 66 ± 71 All the syntheses of 4-OHA 91 described in the literature involved acid- or base-catalysed opening of 4,5-epoxides 90a,b.The latter are obtained by treatment of AD with hydrogen peroxide in an alkaline medium. Epoxidation of conjugated steroid ketones is a well-known reaction.In the case of D4-3-ketones, it affords a mixture of 4a,5a- and 4b,5b-oxides in which the latter predominates. The conditions for this reaction have not practically changed since the publication by Bible et al.72 in 1957. However, the overall yields described therein were rather low and poorly reproducible. Various mod- ifications of reaction conditions failed to increase the yields considerably,73 ± 75 although a successful synthesis of epoxides 90a and 90b (1 : 5, total yield 90%) is documented.48 The opening of the epoxide ring in compounds 90a,b results in 4-hydroxy-substituted AD 91 in up to 50% yields.14, 48, 74 ± 78 The steroid 91 was also obtained in a satisfactory yield by opening of the isomeric epoxides 90a,b and hydrolysis of the ether 92.13 Rearrangement of the epoxides 90a,b in the presence of boron trifluoride etherate in benzene and chromatographic separation of side products (e.g., A-nor-3-oxo-5a-carbaldehyde) gave the target steroid 91 in 34% yield.14 The ether 92 was synthesised in 25% yield by treatment of AD with o-iodosylbenzoic acid and alkali in methanol.79 Special mention should also be made of a recently developed procedure for the synthesis of 4-OHA 91 (47% yield) which involves hydroxylation of AD with OsO4/H2O2 with subsequent dehydration of the resulting mixture of 4,5-diols in an alkaline medium.70 Introduction of 4-OHA 91 into the clinical practice and the steadily increasing number of publications devoted to the analysis of the topography of the active site of aromatase using various analogues of 4-OHA 91 (see, e.g., Refs 80 and 81) have stimulated the search for alternative efficient methods of its synthesis.A recent publication,82 which describes a new approach to the synthesis of 4-OHA 91 by oxidation of easily accessible 5a-androst-3-en-17-one (93), is an example. The latter was obtained by reduction of AD with zinc in acetic acid.83 Treatment of steroid 93 with performic acid generated in situ gives trans- diaxial diol 95 in 96% yield.Oxidation of androstenone 93 in dichloromethane resulted in 3a,4a-epoxide 94 as the only identi- fied and isolated reaction product. Further oxidation of the diol 95 with dimethyl sulfoxide activated with trifluoroacetic anhy- dride gives hydroxy ketone 96 in quantitative yield.The latter is isomerised into the final steroid 91 on treatment with a base. a b 28 HO O O 87 (76%) O O OMe O O (a) ClCH2OCH2CH2OMe, (Me2CH)2NEt, CH2Cl2, 20 8C, 20 h; (b) LDA,Me3SiCl, THF,720 8C, 0.5 h; (c) TiCl4, CH2Cl2, 720 8C, 0.5 h. OSiMe3 88 (100%) Me3SiO O OMe O 89 (35%) O O O c AD O O a 4a,5a-90a, 4b,5b-90b O O (a) H2O2, OH7; (b) AcOH, H2SO4, or Me2CO, H2O, H2SO4, or HF, Py, or BF3 .Et2O, PhH; (c) MeOH, NaOH (62%); (d ) HCl, dioxane, D (53%); (e) OsO4, H2O2, ButOH, 20 8C; ( f) OIC6H4CO2H, MeOH, KOH (25%). AD e, c f 91 92 O OH 91 d 91 O OMe 92 b c O c, d b a H 93 O HO H OH 95 (96%) 94 O H Substituted androstanes as aromatase inhibitors 985High yields in all the stages of this synthesis and the accessi- bility of the starting steroid 93 (the authors propose to use a mixture of 5a- and 5b-isomers) illustrate the high outlook of this approach.The mechanism of the inhibitory effect of 4-OHA 91 on aromatase has not been completely elucidated. An inactivation mechanism, which involves the formation of a covalent bond between the enzyme and position 4 of the steroid substrate (which, in turn, changes the direction of the final stage of elimination), has been proposed by Covey.24 Protonation of the hydroxy group at position 4 of the substrate and its elimination as a water molecule occurs in the enzyme-bound intermediate A (instead of removal of the enzyme as the nucleophile, which normally takes place).Inactivation of the enzyme, which cannot `kick off' the steroid substrate, follows aromatisation of the steroid ring A with the formation of structure B.A great number of 4-OHA 91 analogues have been synthesised aimed at elucidating the mechanism of its action and obtaining more efficient aromatase inhibitors. Thus its 6a-methyl analogue obtained by hydrolysis of the corresponding 4,5-epoxide in a mixture of acetic and sulfuric acids has been described.84 4-Acetoxy-AD 63, 64, 85 possessed inhibitory proper- ties, whereas 4-methoxy-AD 92 was inactive.74 However, recent in vitro studies have shown that 4-methoxy-AD 92 suppressed the proliferative activity of cancer cells.86 Although 4-OHA 91 is an active AI, it is characterised by low bioavailability and short-term action and manifests in vivo activity only on subcutaneous administration, which is due to the ease of its glucuronidation.67, 87, 88 The quest for the synthesis and anal- ysis of its various analogues thus becomes understandable.Besides, these studies were aimed at establishing the structural demands for the design of novel AI. A prominent role among formestane 91 analogues belongs to 4-alkylthio- and 4-arylthio-AD. Several compounds of this series (98a ± e) were obtained in the 80's 89, 90 by the opening of epoxides 90a,b with the corresponding thiol reagents.All of them appeared to be efficient competitive AI with Ki ranging from 36 to 73 nM.90 It was found that the chain length at C(4) of alkylthio analogues of such steroids should not exceed three carbon atoms in order to preserve a sufficiently high inhibitory activity.90 4-Phenylthio-AD 98d (25 and 50 mg per kg per day) inhibited the growth of mammary tumours in rats.91 The thiol 97 was synthesised in high yield by the opening of epoxides 90a,b with thioacetic acid in dioxane. Its subsequent alkylation with various functionalised alkyl halides gave a series of 4-alkylthio-AD 98f ± k containing hetero atoms in the substituent at C(4).48, 92 4-Thio-substituted 98f ± h were comparable in activity with 4-OHA 91 (see Table 2).The difluoromethyl analogue 98h was found to be a reversible AI, whereas the fluoromethyl and chloromethyl analogues 98f and 98g are suicide AI with Ki=30 nM (cf. Ki for 4-OHA 91 is 27 nM48). In vivo activity of compound 98f is comparable with that of 4-OHA 91 and, like the latter, it is inactive on per os administration.48 It is noteworthy that the presence of an additional 9(11)-double bond in these analogues, like in 19-sulfur-containing steroids 40a ± o, does not influence their affinity and inhibitory activity with respect to aromatase.A similar conclusion was made from the biochemical analysis 48, 88 of some 4-alkyl(aryl)thio-D9(11)-AD 102 ± 106 (their synthesis is shown below).The most active compound of this series is the steroid 104 (R=CH2Cl) (Ki=27 nM).48 (a) H2O2, HCO2H, CH2Cl2, 20 8C, 6 h, (b) H2O2, HCO2H, 20 8C, 1 h; (c) DMSO, (CF3CO)2O,760 8C, 3 h; (d) Et3N,760 8C, 15 min; (e) NaOMe, 20 8C, 1 h. e 91 (80%) OH O 96 (100%) H HO O Nu is the nucleophile, E is the enzyme. B NuE HO A B7 HO NuE H O 91 OH O O 99 O O a 100 (b : a=3 : 2) (99%) O O b c 101 (52%) OH O OR 102 O O R=Ac, CH2Ph 103 (28%) SH O d e, f R=Me R=Me, CH2Cl 104 SR O (a) H2O2, NaOH, CH2Cl2, MeOH, 4 8C, 1 h; (b) H2SO4, AcOH, 4 h, 20 8C; (c) AcSH ± dioxane, MeOH, HCl; (d ) RX, ButOK, THF; (e) HC(OEt)3, EtOH, p-TsOH; ( f ) chloranil; (g) MCPBA; (h) DAST, CHCl3, D. 105 SMe O g, h 106 (43%) SCH2F O 97 (80%) SH O O c a 90a,b (a) AcSH, dioxane, MeOH, HCl; (b) RX, ButOK, THF; (c) RSH; R=Me (a), Et (b), Pr (c), Ph (d), CH2Ph (e), CH2F (f), CH2Cl (g), CHF2 (h), CH2CN (i), CH2OMe (j), CH2SMe (k).b 98a ± k SR O O 986 I S LevinaOther 4-substituted AD have been also obtained and assayed as AI. Thus 4-fluoro-AD 109 has been recently prepared from 4-bromo-AD 107.93 Conversion of bromide 107 into the 4-tri- methylstannylated steroid 108 by hexamethyldistannane cata- lysed by tetrakis(triphenylphosphine)palladium and its subse- quent treatment with cesium fluorosulfate gave 4-fluoro-AD 109 in high yield. 4-Fluoro-AD 109 is comparable with 4-OHA 91 in its inhib- itory activity; however, its practical application is limited due to its low bioavailability.94 The synthesis of 4-chloro-AD is given in a patent.95 Among other 4-substituted AD, special mention should be made of 4-azido-AD 111 and 4-amino-AD 112 prepared from 4-mesyloxy-AD 110 according to patent data.96 Highly active 4-amino-androstenediones 114 and 119 present special interest.One of them, 4-aminoandrosta-1,4,6-triene-3,17- dione 119, appeared to be an efficient AI and was recommended for treatment of mammary tumours. This compound has passed successfully clinical trials under the trade name of `minames- tane'.97 ± 99 4-Aminoandrosta-1,4-diene-3,17-dione (114) was obtained by ammonolysis of the corresponding 4-chloro derivative 113 95 in dioxane (Scheme 5).Scheme 5 The oxirane 116 obtained by selective epoxidation of the trienedione 115 was the key compound in the synthesis of minamestane 119. Its reaction with potassium thiocyanate gave the 6-thiocyanate derivative 117.Heating of the latter in DMF resulted in 4-isothionate 118 and its alkaline hydrolysis yielded the target compound, 4-aminotrienedione 119 (see Scheme 5).100 Its 6- and 7-substituted analogues have been described in the pat- ent.101 Studies of structure ± activity relationships revealed the enhanced rate and selectivity of aromatase inhibition upon introduction of additional 1,2- and 6,7-double bonds into the molecule of 4-amino-AD 112.98 Steroids 114 and 119 are highly efficient and manifest in vivo antitumour activity both on sub- cutaneous and oral administration.9, 81, 97 ± 99 Thus, the majority of 4-substituted AD described above represent highly active irreversible AI with Ki ranging from 30 to 200 nM.100 3. 1-Substituted androstanes 1-Methylandrosta-1,4-diene-3,17-dione (121) is the most valuable and perhaps the only steroid AI of this series. It was synthesised in 1983 102 and patented under the trade name atamestane in more than a dozen countries. Atamestane is an irreversible inhibitor of estrogen biosynthesis in vitro and in vivo.102 ± 105 Androsta-1,4-diene-3,17-dione (ADD) is the starting com- pound in the synthesis of atamestane 121.This synthesis required two principal problems to be solved, viz., introduction of a methyl group into position 1 of the steroid skeleton and regeneration of a b AD O O 107 (61%) O Br (a) Br2, AcOH, ,730 8C; (b) Pd(PPh3)4, (Me3Sn)2, PhMe; (c) FOSO3Cs, CH2Cl2, MeOH, N2, 0 to 20 8C. O c 108 (84%) O SnMe3 109 (88%) F O O 113 Cl O O a 114 (31%) NH2 O O (a) 30% NH4OH, dioxane, 24 h; (b) 2 KSO5 .KHSO4 .K2SO4 ± phos- phate buffer, Me2CO, 5 ± 10 8C, 15 min; (c) KSCN; (d ) DMF, 110 ± 135 8C, 1.5 h; (e) dioxane ±H2O, 35% NaOH, 0.5 h. 117 OH SCN O d e 118 NCS O 116 b O O c 115 O O 119 O NH2 O Table 2. Data on in vitro inhibition of aromatase from human placenta by 4-substituted androstanes. R The presence Ki IC50 Ref.of a D9(11) /nM /nM double bond OH (91) 7 10 ± 37 31 27 48, 123 50 55 67 124 43.7 9 OPh (176) 7 412 80 NH2 (112) 7 37 15 SH (103) + 72 48 SCH2Cl (98g) 7 30 48 SCH2Cl (104) + 27 48 SCH2F (98f) 7 30 48 SCH2CN (98i) 7 169 48 O O R (a) NaN3, H2O, 60 8C, 1 h; (b) NaN3, 100 8C, 1.5 h. 110 OMs O NH2 112 O a b O N3 111 O O Substituted androstanes as aromatase inhibitors 987the 1,2-double bond in the resulting 1a-methyl-D4-3-oxo deriva- tive 120.The first problem can be solved through selective 1,4- addition of organometallic compounds (e.g., methyllithium cup- rates) to ADD.106 However, the high cost of 2,3-dichloro-5,6- dicyano-1,4-benzoquinone (DDQ) employed in 1,2-dehydrogen- ation of D4-4-ketosteroid 120 and the formation of a 4,6-dehydro derivative as a side product are the limiting factors.107, 108 Early patents applied by the `Schering AG' company dealt with the synthesis of atamestane 121 by oxidation of its 17b- hydroxy derivative 126 102 or by microbiological dehydrogenation of the 1-methyl-D1-5a-derivative 127 in up to 65% yields.109, 110 Compounds 126 and 127 were prepared in a multistep procedure (see below) and in unsatisfactory yields.Thus the 4,5-double bond in the 1a-methyl derivative 123 obtained from the dienone 122 106 was reduced and the resulting compound 124 was brominated to give the 2,4-dibromosteroid 125. Elimination of two HBr molecules from the dibromide 125 gave the dienone 126.111, 112 Oxidation of the 17b-hydroxy group in compound 126 gave the target product 121. The characteristic feature of this synthetic scheme is that the 4,5-double bond in the intermediate compounds is first reduced and then recovered together with the formation of the 1,2-double bond.Obvious disadvantages of this scheme consist in the formation of an undesirable side product, viz., 17b-hydroxy-1a-methylandro- stane-4,6-dien-3-one, and the necessity of tedious chromato- graphic purification following dehydrobromination.Nevertheless, this scheme has recently been applied 113 in the synthesis of atamestane 121 from 1a-methyldihydrotestosterone acetate (128) 114 with satisfactory yields at all steps. Dibromination has led to the epimeric dibromides 129a,b isolated in the individual state, and their configuration was established by 1H NMR. However, dehydrobromination of com- pounds 129a,b with lithium carbonate and lithium bromide occurred in only 66% yields.No formation of side products has been reported. The alcohol 126 prepared by deacetylation of the 17b-acetoxy derivative 130 was oxidised into atamestane 121 using a tetrapropylammonium perruthenate ±N-methylmorpho- line N-oxide system in quantitative yield. The overall yield was 60%.113 A shorter and more efficient method for the synthesis of atamestane 121 which does not require chromatographic proce- dures has been patented.115 ADD O O MeLi Cu(I) 120 O DDQ 121 O O 126 O OH 121 125 124 H O H Br Br O 123 O 122 O OH H 127 O O OAc H 128 O a H Br Br O H Br Br O + b 129b (7%) 129a (91%) (a) 2 equiv. Br2, 4N HBr, AcOH, THF, 50 8C, 0.5 h; (b) LiBr, Li2CO3, DMF, 80 8C, 18 h; (c) MeOH, NaOH, D, 1.5 h; (d ) TPAP7NMO, CH2Cl2, 20 8C. 130 (66%) OAc O 126 (94%) OH O 121 (97%) d c (a) MeLi, CuI, Cl(CH2)2Cl, 730 to 0 8C; (b) Ac2O, Cl(CH2)2Cl, 20 8C; (c) NH4Cl, H2O, 20 8C; (d ) 1,3-bromo-5,5-dimethylimid- azolidin-2-one (dibromantin), dioxane, H2O, 20 8C; (e) MgO, DMF, 110 8C, 2 h; ( f ) NaOAc, ICl, Me2CO,710 8C, 99% (2b-I-isomer); (g) (CH2CO)2NI, abs. MeOH, N2, 20 8C, 55% (2a-I-isomer); (h) Li2CO3, DMF, 130 8C, 1.5 h; (i) Me3Al, 1 mol.% ± 5 mol.% CuBr, Me3SiCl, THF± PhMe, 20 8C; ( j ) I2, CuO, glacial AcOH, N2, 60 8C, 24 h. 131 ADD i 120 (90%) O 121 (80%) h f or g j 133 (100%) I O 132 (97%) Br O e 121 (70%) O O a, b, c d ADD O O 131 (58%) AcO 988 I S LevinaThe intermediate 1a-methylenolate formed upon 1,4-addition of dimethyllithium cuprate to ADD in the presence of acetic anhy- dride is converted into the crystalline dienol acetate 131.Bromi- nation of the latter in aqueous dioxane gave a mixture of epimeric 2x-bromo-1a-methylandrost-4-en-17-ones (132) in which the 2b- isomer was predominant. Dehydrobromination of the bromo ketone 132 with magnesium oxide resulted in the target product 121 in 70% yield. Dehydroiodination of the 2-iodo derivative 133 is a modifica- tion of the above-described procedure.116 This compound is formed from dienol acetate 131 and monochloroiodine in acetone in quantitative yield.When N-iodosuccinimide was used, the yield of the iodide 133 dropped to 55%. The 2-iodosteroid 131 was also obtained by iodination of 1a-methyl-D4-3-ketone 120 with ele- mentary iodine in glacial acetic acid.Dehydroiodination pro- ceeded smoothly on boiling of the iodo derivative 131 with lithium carbonate in DMF. Yet another method for obtaining the intermediate 120 for the synthesis of atamestane 121 was the use of trimethylaluminium in the presence of catalytic amounts (1 mol.% ± 5 mol.%) of mono- valent copper salts for the introduction of the 1a-methyl group into ADD.117 Various modifications of this approach, in partic- ular, the use of Lewis acids for acceleration of the reaction and supression of competitive 1,2-addition and the mechanism of these reactions are discussed in detail with steroid 1,4-dien-3- ones and 1-en-3-ones as examples.114 It was shown that selective 1,4-addition of trimethylaluminium to ADD occurs in the pres- ence of equimolar amounts of chlorotrimethylsilane and catalytic amounts of copper(I) bromide, the yield of 1a-methylandrostat-4- ene-3,17-dione 120 being 90%.Finally, a single-step procedure for the synthesis of atames- tane 121 fromADDhas been patented.118 This method consists in successive treatment of ADD with trimethylaluminium in hexane with a stoichiometric amount of sterically hindered di-tert-butyl- phenol in the presence of catalytic amounts of nickel(II) acetyla- cetonate.The yield of steroid 121 was 42%. Syntheses of 15a-alkyl atamestane analogues 107, 119, 120 and 1b,2b-cyclopropa-4-OHA have been patented.121 4. 6-Substituted androstanes 6-Methyleneandrosta-1,4-diene-3,17-dione (136) synthesised in the late 80's as an anticancer drug known under the trade name `exemestane' is the most well-known aromatase inhibitor of the 6-substituted androstane series.It represents an orally active irreversible selective AI.9, 122 ± 124 Exemestane 137 is shown to be 2.5 times more active than formestane 91.124 Exemestane can be synthesised from both AD and ADD (Scheme 6). The key stage in this synthesis is methylenation of a steroid at position 6.When AD is used as the starting compound, the methylene group at position 6 is introduced by direct meth- ylenation of 3-oxo-4,5-dehydrosteroid using the method of Annen et al. 125 The yields in this stage do not exceed 45%; the penultimate stage consists in 1,2-dehydrogenation of 6-methyl- ene-AD 138 by boiling either with DDQ in anhydrous dioxane or with selenium dioxide in tert-butyl alcohol.The yields in the second stage are 40%± 50%.117 Thus, the overall yield of exemes- tane was as low as 20%± 25%. This scheme involves tedious chromatographic purification of reaction products in both stages and the use of an expensive reagent (DDQ). An Scheme 6 improved procedure for the synthesis of the steroid 136 has been patented. Although this requires a larger number of steps, it yields only crystalline intermediates and is suitable for large-scale syntheses.It includes the Mannich aminomethylation of a 3,5- dien-3-ol ester prepared from AD (see Scheme 6). The 6-methyl- ene steroid 138 is obtained in>70% yield. Compound 138 is then brominated with bromine in acetic acid. Dehydrobromination of the tribromide thus formed is carried out in two steps.At first, the 6-methylene group is regenerated by treatment of the tribromide with sodium iodide in acetone. The resulting 2-bromo-6-methyl- ene steroid (not shown in Scheme 6) is further dehydrobromi- nated without purification by boiling with lithium salts inDMFto give exemestane 136 in 47% yield for the last two stages.126 Two approaches for the synthesis of exemestane 136 from 3-oxo-1,4-didehydrosteroids are shown in Scheme 6.Direct g-methylenation, which is applicable for a wide range of sub- strates,125 has failed in the case of ADD. However, the 17b- hydroxy derivative 122, which is an accessible industrial raw material, can participate in the Mannich reaction. The resulting product 137 formed in 31% yield is oxidised to exemestane 136.127 And, finally, ADD was used directly as the starting material for introducing the 6-methylene group via the 1,3-bispyrrolidino derivative 134 which was converted into 6b-hydroxymethyl ste- (a) Me3Al, n-C6H14; (b) 3,5-(But)2C6H3OH, EtOAc, Ni(acac)2, 58 8C, 2 h.a, b ADD O O 121 O ADD O O a 134 N N 136 d c 122 OH O 137 OH O (a) HCHO, PhH, EtOH, 90%; (b) MeOH, HCl, 20 8C, 5 h (80%); (c) HCHO, Me2NH.HCl, AmiOH, 131 8C (31%); (d ) CrO3, H2SO4 (79%); (e) HC(OEt)3, p-TsOH, THF± EtOH, 40 8C, 2 h; ( f ) PhNHMe, HCHO, 40 8C, 2 h; (g) conc. HCl, 1 h (73%); (h) Br2, Et2O, HBr, AcOH, 2 ±3 8C (84%); (i ) NaI, Me2CO, D, 15 min; ( j) Li2CO3, LiCl, DMF, 120 8C, 2.5 h; (k) H2C(OR)2, POCl3, CHCl3, NaOAc, D (40% ± 50%). 136 (40% ± 50%) h DDQ 136 (47%) Br CH2Br Br O i, j AD O O 138 O (e, f, g), or k b 135 CH2OH O 136 O O Substituted androstanes as aromatase inhibitors 989roid 135 by standard procedures.128 Dehydration of the latter gave exemestane 136 in 80% yield.129 A series of exemestane analogues modified at position 6 and its 17b-hydroxy derivatives have been synthesised in order to investigate the structure ± activity relationships.124 The reason for these studies was the experimental evidence about a wide set of exemestane metabolites.The design of its analogues was based on the hypothesis according to which exemestane metabolism results in modification of not only the carbonyl group at C(17) but also the exocyclic double bond.124 Synthesis of exemestane 136 analogues is considered below.Epoxidation of exemestane 136 or its 17b-hydroxy derivative 137 with a peroxy acid gave epimeric mixtures of spirooxiranes 139a,b, respectively. Their subsequent hydrolysis with aqueous perchloric acid gave 6-hydroxymethyl-1,4,6-trienes 141a,b along with epi- meric 6-hydroxy-6-hydroxymethyl derivatives 140a,b.124 The Jones oxidation of 6b-hydroxymethyl-AD 135 gave 6b- carboxy-AD 142, while selective reduction of the diketone 135 with sodium borohydride gave the 17b-hydroxy derivative 143.124 A series of 6-methyl-substituted AD and ADD have been synthesised.Thus 6-methylene-AD 138 was isomerised over Pd/C into a 4,6-didehydro-6-methyl analogue. Its dehydrogenation with DDQ gave 1,4,6-triene 144 in 50% yield, while reduction over Pd/C in the presence of cyclohexene as a hydrogen donor yielded 6a-methyl-AD 145.The latter was further dehydrogen- ated into 6a-methyl-1,4-diene steroid 146. Finally, oxidation of ADD with oxygen in the presence of alkali gave androsta-1,4- diene-3,6,17-trione 147.124 Preparation of the 2-fluoro derivative of exemestane is patented.130 Exemestane 136 is a highly selective irreversible AI. The irreversibility of its inhibitory action is assigned to the presence of a 1,2-double bond in the molecule;123 the corresponding 1,2- hydrogenated analogues represent reversible AI.It is these prop- erties rather than pharmacokinetic ones that account for the prolonged action of exemestane on the organism.123 None of the synthetic exemestane analogues displays in vitro activity which would be equal or greater than that of exemestane itself.Thus the 17b-hydroxy derivative 137 is active in vitro, but its activity is 2.6-fold less than that of compound 136. According to their inhibitory activity, exemestane analogues modified at C(6) and having a 17-keto group can be arranged in the following order: 6-methylene (136)>6-spirooxirane (139a)>6b-hydroxy- methyl (135)>6-hydroxy-6-hydroxymethyl (140a)>6b-carbo- xy derivative (142). 17b-Hydroxy analogues of some of these compounds, e.g., 139b, 140b and 143, manifest 3 ± 8-fold lower activity than the corresponding 17-oxo compounds. Isomerisation or reduction of the double bond (leading to compounds 144 and 146) decreases their inhibitory activity threefold in comparison with exemestane itself.124 A series of 6-alkyl- and 6-aryl-substituted AD has been described and their inhibitory activity has been studied with the view of determining the topography of the hydrophobic binding pocket in the active site of the enzyme.131 ± 134 The key compound in this synthesis was a mixture of isomers of 5,6-epoxide 149 obtained by epoxidation of androst-5-ene-3,17-dione diketal 148. The opening of the oxirane ring with Grignard reagents gave 5a- hydroxy-6b-substituted androstanes 150.Treatment of the latter with HClO4 in THF gave the corresponding 3,17-diketones 151 in high yields. Dehydration of these hydroxy diketones with thionyl chloride in pyridine has led to a series of 6b-substituted AD 152a ± i. Epimerisation of the 6b-centre in 6b-substituted D4-3- ketones 152 into thermodynamically more stable 6a-isomers 145a ± g occurred in 16% ±56% yields.131, 132 R1, R2=O(136, 139a ± 141a); R1=b-OH, R2=H(137, 139b ± 141b); (a) MCPBA, CH2Cl2, 20 8C, 48 h (80%); (b) HClO4, THF, 20 8C, 3 h. 140a (45%), 140b (35%) + HO CH2OH R2 O R1 141a (42%), 141b (38%) CH2OH R2 O R1 136, 137 R2 O R1 139a,b R2 O O R1 b a 135 O CH2OH O 142 (65%) O COOH O 143 (65%) OH CH2OH O a b (a) CrO3, H2SO4; (b) NaBH4, MeOH.a AD O O b 138 O 146 (50%) O (a) CH2(OEt)2, POCl3, CHCl3 (60%); (b) 5%Pd/C, EtOH, D, PhCH2OH (90%); (c) DDQ, dioxane, D, 15 h; (d ) Pd/C, cyclohexene (60%); (e) O2, KOBut, ButOH (15%). ADD 147 O O e 145 144 (50%) O d c O c O 990 I S LevinaBiochemical assessment of both 6a-(145) and 6b-(152) substituted ADhas been carried out (Table 3).These compounds appeared to be competitive aromatase inhibitors with Ki ranging from 1.4 to 63 nM. Introduction of a methyl group into 6a- and 6b-positions of AD increases its affinity for aromatase. Introduction of an additional methylene group into position 6 of the steroid molecule changes its inhibitory properties. Thus 6b-ethyl-AD 152b had the highest Ki value (1.4 nM) in this series of compounds.Further elongation of the chain does not result in any significant alteration in the inhibitory activity. 6a-Benzyl-AD 145g and 6b-vinyl-AD 152h also manifest high activity. Bulky substituents, such as 6-isopropyl (152e and 145e) or polar substituents, like 6b- ethynyl-(152i), decrease the inhibitory activity (Ki=21 ± 63 nM). The activity of 6b-substituted AD decreases in the following order: Et>Pr>CH=CH2>Bu>Me>Pri>Ph>C:CH >CH2Ph, while that of 6a-substituted AD decreases in the order: Et>Me> Prn>CH2Ph>Bu >Ph>Pri.The stereo- chemistry of the C(6) atom affects the inhibitory activity. Thus 6b- alkyl isomers possess greater affinity for the enzyme than the corresponding 6a-isomers with the exception of 6a- and 6b-methyl analogues 145a and 152a.Exchange of the 6b-ethyl substituent by the 6b-vinyl substituent decreases the affinity of compound 152h compared to that of compound 152b. It was shown that there is a hydrophobic binding pocket in the active site of the enzyme in the b-region of position 6 of the steroid substrate. Judging by the data on the crystal structure of cytochrome P-450 and the inhibitory activity of steroid substrates of this series, the following sizes for the pocket have been proposed: length 6,27 A, width 5,25 A, height 7,74 A.131 In 1995, Numazawa and Oshibe 131 synthesised and assayed 6a- and 6b-alkyl-substitutedADwith 5 ± 7 methylene groups in the side chains.TheirKi values varied from 2.8 to 80 nM.The earlier structure ± activity regularities were also confirmed for compounds of this series. Thus 6b-pentyl-AD manifested the highest activity; further elongation of the chain sharply decreased the affinity.133 The analysis of effects of the chain length and configuration of the alkyl substituent at position 6 of ADD analogues 135 on their inhibitory activity is an extension of earlier studies.131, 133 The 6a- and 6b-alkyl-substituted ADD 146a ± g and 153a ± g were obtained by dehydrogenation of their D4-3-keto precursors. Compounds 146a ± g and 153a ± g appeared to be competitive AI with Ki ranging from 4.7 to 54 nM(see Table 3).Among those, only 6a-ethyl-ADD 146b is the suicide AI. 6b-Isomers manifest greater activity than 6a-analogues. However, there was no corre- lation between the length of the alkyl chain and the affinity for the enzyme.The ability of 6b-alkyl isomers to inhibit aromatase decreases in the following order: Et>(CH2)4Me> Pr>(CH2)6Me>(CH2)5Me>Bu>Me, which is consistent with the results obtained for 6b-alkyl derivatives of AD.133 The activity of 6a-isomers of alkyl-substituted ADD 146a ± g R=Me (a), Et (b), Prn (c), Bu (d), Pri (e), Ph (f), CH2Ph (g), CH=CH2 (h), C:CH (i).(a) MCPBA, CH2Cl2, 20 8C, 4 h (a mixture of isomers*1 : 1); (b) RMgBr, THF, D (86% ± 98%); (c) 3MHClO4, THF, 20 8C, 3 h; (d) SOCl2, Py, 0 8C, 3 min; (e)1M HCl, EtOH, D (16% ± 56%). d c 150 O O R HO 151 O HO R O 148 O O O O b a 149 (91%) O O O 152a ± i O R O 145a ± g O R O e Table 3. Data on in vitro inhibition of aromatase from human placenta by 6-substituted androstanes.R The position Ki IC50 Ref. of double bond a /nM /nM D1 D6 a-Me (145a) 7 7 5.6 350 131 b-Me (152a) 7 7 11 720 131 a-Et (145b) 7 7 4.7 260 131 b-Et (152b) 7 7 1.4 140 131 a-Prn (145c) 7 7 6.7 390 131 b-Prn (152c) 7 7 4.6 240 131 a-Bun (145d) 7 7 12 790 131 b-Bun (152d) 7 7 8.8 510 131 a-Pri (145e) 7 7 31 1400 131 b-Pri (152e) 7 7 22 1100 131 a-Ph (145f) 7 7 21 1100 131 b-Ph (152f) 7 7 37 1700 131 a-CH2Ph (145g) 7 7 10 660 131 b-CH2Ph (152g) 7 7 63 4000 131 b-CH=CH2 (152h) 7 7 5.1 320 131 b-C:CH (152i) 7 7 62 1900 131 Pri 7 7 4.9 131 a-Br 7 7 3.4 173 b-Br 7 7 800 173 =O (147) + 7 268 124 =CH2 (136) + 7 26 ± 27 42.5 9, 97, 124 a-Me (146a) + 7 54 135 b-Me (153a) + 7 43 135 a-Et (146b) + 7 19 135 b-Et (153b) + 7 4.7 135 a-Prn (146c) + 7 22 135 b-Prn (153c) + 7 7.0 135 a-Bun (146d) + 7 16 135 b-Bun (153d) + 7 14 135 a-(CH2)4Me (146e) + 7 11 135 b-(CH2)4Me (153e) + 7 5.0 135 a-(CH2)5Me (146f) + 7 25 135 b-(CH2)5Me (153f) + 7 8.0 135 a-(CH2)6Me (146g) + 7 20 135 b-(CH2)6Me (153g) + 7 7.8 135 Me (144) + + 91 124 a Symbols+and7denote the presence and absence of the double bond in this position, respectively.O O R Substituted androstanes as aromatase inhibitors 991decreases in the order: (CH2)4Me>Bu>Et>(CH2)6Me> Pr>(CH2)5Me>Me, which does not match that for 6a-alkyl- AD.133 Based on the experimental data and the results of molecular modelling studies, Numazawa et al.135 assumed that at least 6b-alkyl-ADD 153 can bind with the same hydrophobic pocket in the active centre of aromatase as does the corresponding 6b-alkyl-AD 152. 6-Hydroxyimino-AD 160 and its 17b-hydroxy analogue 161 are competitive AI.136 The key derivative 158 was obtained by standard methods from 3,17-diacetate 154 via 5a,6a-epoxide 155 with its subsequent oxidative conversion into 5a-hydroxy-6- ketone 156 with chromic anhydride. Dehydration of the latter and hydrolysis of 3,17-acetoxy groups in D4-6-ketone 157 gave the dihydroxy derivative 158 in high yield.Treatment of the latter with hydroxylamine hydrochloride resulted in the (E)-oxime 159; its oxidation with chromic anhydride in pyridine gave 6-hydrox- yimino-AD 160 in 70% yield. Oxidation of oxime 159 with manganese dioxide under mild conditions resulted in the testos- terone derivative 161. 5. 7-Substituted and 6,7-disubstituted androstanes The key stage in the synthesis of 7-substituted androstanes is 1,6- conjugate addition of the corresponding reagents to 6-dehydro derivatives of AD and ADD. Thus 7a-thio-substituted ADD 162, one of the first highly active 7a-substituted androstanes, was obtained in 14% yield by addition of p-aminothiophenol to 6-dehydro-ADD 115.137 This is accompanied by the formation of its inactive 2a-isomer 163 (yield *40%).The attempts to improve the regioselectivity and to increase the yield of 7a-substituted compound 162 by changing the reaction conditions (solvent, temperature and reagent ratio) failed.137 Further transformations of 7a-(p-aminophenylthio)-ADD 162 have led to its analogues with bromine or iodine atoms in the para-position of the aryl substituent. 7a-Naphthalene derivatives 164 and 165 are also described.138 All these steroids are irreversible suicide AI (Ki=12 ± 27 nM).138 According to the patent,139 O R O 145a7d, i7l 152a7d, k7m 146a7g, 153a7g O R O DDQ D 156 AcO HO O c d 157 AcO O 158 HO OH O e 159 HO NOH b a 154 AcO OAc 155 AcO O (a) MCPBA, CH2Cl2; (b) CrO3, H2O; (c) SOCl2, Py (90%); (d ) KOH, MeOH (82%); (e) NH2OH.HCl (75%); ( f ) CrO37Py (70%); (g) MnO2 , CHCl3, 20 8C (45%). f g 160 O NOH O O OH 161 NOH 164 O S O 165 O S O R=H(a), NH2 (b), I (c), OMe (d), N3 (e). 166a7e O S C6H4R-p O + 163 (38%) NH2C6H4S O 162 (14%) SC6H4NH2 O 115 Na, dioxane HS NH2 O O R a-Isomer b-Isomer original product original product compound compound Me 145a 146a 152a 153a Et 145b 146b 152b 153b Pr 145c 146c 152c 153c Bu 145d 146d 152d 153d Am 145i 146e 152k 153e (CH2)5Me 145k 146f 152l 153f (CH2)6Me 145l 146g 152m 153g 992 I S Levina7a-alkyl(acyl)thio- and 7a-mercapto-ADD are highly active prep- arations for in vivo treatment of mammary tumours. 7a-Arylthio-substituted AD 166a ± e were obtained by analo- gous addition of substituted thiophenols to 6-dehydro-AD.24 The activity of the steroid 166b was twice as low as that of its ADD analogue 162.The steroid 166e (Ki=1 nM) was the most efficient in this series (Table 4).25, 140 The compounds described above contain an aryl substituent which is linked to the steroid skeleton by the C7S bond. It was found, however, that the presence of a sulfur atom and the stereochemistry of the C(7) atom are not essential for the manifestation of activity, since 7a- and 7b-substituted AD 170a,b, 171a,b and 173a,b with one or several methylene groups between the steroid skeleton and the phenyl substituent appear to be efficient AI.141 These androgens with arylaliphatic substituents were obtained by the same approach as that used for 7a-thio-substi- tuted compounds described above.The silylated testosterone derivative 167 was used as the starting compound. Introduction of a phenylethyl or a phenylpropyl substituent was carried out by addition of the cuprate reagent prepared from the corresponding lithium reagent and 1 equiv. of copper(I) tetrakis (tributylphos- phine) iodide. Subsequent addition of 0.2 equiv. of the steroid 167 and deprotection gave the desired 1,6-addition products as mixtures of a,b-diastereomers 168a and 169a (71%, a :b=2:1) or 168b and 169b (50%, a :b=1 : 1) isolated in the individual state by chromatography.The structure of the 7a-isomer 168a was additionally confirmed by X-ray analysis. Oxidation of 17b- alcohols 168a, 169a, 168b and 169b gave 17-ketones 170a, 171a and 171b, respectively. 7a- And 7b-benzyl derivatives 172 and 173 were obtained in a similar way. 167 OTBDMS O a or b, c, d + R=CH2Ph (a), (CH2)2Ph (b). 168a,b OH R O R=CH2Ph (a), (CH2)2Ph (b). 169a,b OH O R e e R=CH2Ph (a), (CH2)2Ph (b). R=CH2Ph (a), (CH2)2Ph (b). 171a,b O O R 170a,b O O R 167 f, c, d 33% e 172a,b (33%) OH O Ph (a) Ph(CH2)2I, ButLi, Et2O,778 8C; (b) Ph(CH2)3I, ButLi, Et2O, 778 8C; (c) [CuI(Bun3 P)]4, Et2O,740 8C; (d ) 6N HCl, CH2Cl2, 20 8C, (7a : 7b=5 : 1); (e) PCC, CH2Cl2, 20 8C; ( f ) PhCH2Br, Mg, Et2O, 20 8C. 7a-isomer (a), 7b-isomer (b) 173a,b O O Ph Table 4. Data on in vitro inhibition of aromatase from human placenta by 7-substituted and 6,7-disubstituted androstanes. Compound R1 R2 R3 X Y The presence Ki IC50 Ref. of a D1 /nM /nM double bond 166a H H a-SPh O O 7 212 650 80 166b H H a-SC6H4NH2 O O 7 18 25, 141 166c H H a-SC6H4I O O 7 12 25, 15 166d H H a-SC6H4OMe O O 7 31 25 166e H H a-SC6H4N3 O O 7 1 25 173a H H a-CH2Ph O O 7 18.9 141 173b H H b-CH2Ph O O 7 44.5 141 170a H H a-(CH2)2Ph O O 7 13.1 141 170a H H b-(CH2)2Ph O O 7 40.2 141 170b H H a-(CH2)3Ph O O 7 16.5 141 172 H H a-CH2Ph O b-OH 7 39.5 141 168a H H a-(CH2)2Ph O b-OH 7 36.0 141 162 H H a-SC6H4NH2 O O + 9.9 25 179 OPh H a-SPh O O 7 472 1650 80 187 H a-CF2 O O 7 50 145 189 H a-CH2 H2 O 7 5.0 145 190 H a-CH2 H2 b-OH 7 120 145 X Y R3 R1 R2 Substituted androstanes as aromatase inhibitors 9937a-Substituted AD 170a,b and 173a are efficient competitive AI with Ki 13 ± 19 nM.The corresponding 17b-alcohols 168a,b and 172 as well as 7b-substituted AD 171a,b and 173b are less active (Ki=36 ± 44 nM).Thus, introduction of bulky aryl- aliphatic substituents into position 7 of AD favours high inhib- itory activity. It was found that the length of the carbon chain between the steroid and the aryl group does not significantly affect the activity, thus indicating the presence of a bulky pocket in the active site of the enzyme near position 7 of the steroid substrate.According to patent data,142 7a-allyl-AD 174 synthesised from the corresponding 17b-acetoxy-androsta-4,6-dien-3-one is a rather active AI. Disubstituted 4-phenoxy-7a-phenylthio-AD 179 has been synthesised in order to explore the nature of the binding site of aromatase in the vicinity of the C(4), C(6) and C(7) atoms of AD.80 The starting compound in this synthesis, 4-OHA 91, was converted into the 4-phenoxy derivative 176 via the intermediate copper enolate 175 (yield >50%).81 The formation of the necessary 6,7-double bond was achieved by treating the ether 177 with an equimolar amount of DDQ in aqueous acetone.Addition of thiophenol to D4,6-4-phenoxy-AD formed completes the synthesis of the 4-phenoxy-7-phenylthio-substituted com- pound 179.However, its inhibitory activity was much lower than that of the monosubstituted analogues, viz., 4-phenylthio- (98d) and 7a-phenylthio-AD 166a. These results together with molec- ular modelling data for all the three compounds have led Liu et al.80, 81 to infer the existence of a single hydrophobic pocket in the active site of aromatase in the vicinity of the C(4) and C(7) atoms of the steroid substrate.In conclusion, it is appropriate to mention the synthesis of 6,7- disubstituted androstanes. Thus 6,7-epimino-AD 143 represent a practically novel class of derivatives. The key compounds in this synthesis are 7a-azido-6b-hydroxy- (181) and 6b-azido-7a- hydroxy-AD 184 (Scheme 7). Treatment of 4,6-dienedione 180 with sodium azide and chromic anhydride in acetic acid gives the azido alcohol 181, which was converted into 6a,7a-aziridine 182 in 80% yield upon heating with 2 equiv.of triphenylphosphine in toluene. Epoxidation of 6-dehydro-AD 180 with m-chloro- perbenzoic acid in boiling chloroform gave the 6a,7a-oxide 183; its diaxial opening in 6b-azido-7a-hydroxy-AD 184 occurred in high yield and was performed by treatment of an equimolar mixture of sodium azide and concentrated sulfuric acid in DMSO.Reaction of the azide 184 with triphenylphosphine in toluene gives 6b,7b-epimino-AD 185. The b-configuration of its three-membered ring was con- firmed by 1H NMR spectral data and by independent synthesis, (a) H2C=CHCH2SiMe3, TiCl4, CH2Cl2,770 8C; (b) HC/MeOH; (c) PCC. 174 O O OAc O a, b, c c 175 OCu O 91 OH O O a, b d e OPh O OEt OPh EtO 176 177 (a) CuI, Bu2S, MeLi, Et2O,778 to 20 8C; (b)Et2O, THF, Py,778 8C; (c) PhI, D (55%); (d ) CH(OEt)3, p-TsOH, 20 8C, 24 h; (e) DDQ (5%) aqueous Me2CO, 20 8C, 3 min (50%); ( f) PhSH, Na, dioxane, 60 8C, 8 h (40%).OPh O O 178 O OPh O SPh 179 f Scheme 7 O O 185 (68%) O NH O 184 OH N3 O b 183 d O O b 181 OH O N3 182 (84%) NH O c a 180 g f 186 O CF2 S S 187 CF2 e + 188 (40%) 189 (15%) O OH (a)NaN3, CrO3, AcOH, 20 8C(42%); (b) Ph3P, PhMe, D, 3 h; (c) MCPBA, D, 3 h (57%); (d )NaN3, conc.H2SO4,DMSO, 20 8C(84%); (e) ClF2CCO2Na, diglyme D, 5 h; ( f ) (CH2)2SH, p-TsOH, AcOH, 20 8C, 24 h (77%); (g) Na7NH3, THF, 0.5 h (40%). 994 I S Levinaviz., by reduction of 7-mesylate 190a with lithium aluminium hydride. The stereochemical aspect of the formation of steroids 182 and 185 deserves mention: in each case, the corresponding azido alcohol yields only one isomer with the aziridine ring configuration identical with that of the azido group. The compounds obtained appeared to be weak aromatase inhibitors; only 6b-azido-7a-acetate 190b manifested moderate inhibitory activity.143, 144 This section is finished with a description of a series of 6a,7a- cyclopropaandrost-4-en-17-ones.145, 146 Addition of difluoro- carbene generated from sodium difluorochloroacetate to 6-dehy- dro-AD 180 gave 6a,7a-difluorocyclopropane-AD 186. Reduction of its 3-ethylenedithio derivative 187 with lithium in ammonia gave a mixture of 3-deoxy-6a,7a-cyclopropaandrost-4- en-17-one 188 and its 17b-hydroxy derivative 189.These com- pounds proved to be efficient competitive AI,145 the cyclopropane derivative 188 (Ki=5 nM) being the most active one. However, introduction of a 6a,7a-difluorocyclopropane group into the AD molecule (compound 186) did not affect considerably its affinity for the enzyme. This suggests that it is the keto group at position 17 of AD that plays the crucial role in the binding of the steroid with aromatase.A recent study 147 devoted to the synthesis and biochemical assay of compound 191, yet another representative of substituted androstanes obtained by the base-catalysed Michael addition of 2-aminothiophenol to 3,17-diacetoxyandrost-4-en-6-one with subsequent cyclisation, is also worth mentioning. Compound 191 manifested weak inhibitory activity of a competitive type (Ki=42 300 nM).IV. Structure ± activity relationships in substituted androstanes as aromatase inhibitors The data considered in Section III and those listed in Tables 1 ± 4 allow one to make certain conclusions about the structural requirements for steroid substrates to act as aromatase inhibitors. Although the substituted androstanes described above are rather diverse in nature and mechanisms of their inhibitory action have hitherto been studied relatively little, one may note some general regularities concerning the relationships between their structure and inhibitory activity judging mainly by the affinity for the enzyme and inhibitory properties of a series of similarly substi- tuted androstanes.According to Cole and Robinson,15 efficient AI have two domains which play an important role in binding of the steroid substrate with the enzyme, namely, an iron-coordinat- ing domain and a hydrophobic domain.The structural require- ments for the latter are considered to be a crucial factor, although chemical details of its specificity are unknown. The introduction into, or removal from, certain positions of theADmolecule of definite substituents and/or additional double bonds affect the inhibitory activity as indicated by arrows in Fig. 1. The steroid molecule must contain at least one keto group, specifically at position 17 to manifest aromatase-inhibiting activ- ity. All substituted androstanes sharply decrease their activity as AI upon reduction of the 17-keto group to the hydroxy group and especially following its esterification.Evidently, in some cases the carbonyl function at position 3 is not strictly required as a structural element, because, as has been shown in the preceding section, some 3-deoxo analogues retain high inhibitory activ- ity.14, 33, 35, 36, 145 Moreover, replacement of the 3-keto group in AD or in 19-substituted AD by a 3-methylene group (or a halogenomethylene group) affords compounds possessing high competitive inhibitory activity.148, 149 The study of inhibitory activity of androst-4-ene-3,6,17-trione and its 3-hydroxy analogue revealed that both of them can irreversibly bind with aroma- tase.150, 151 An important role in manifestation of biological activity is played by the conformation of the ringA.For example, androst-2- ene-3,17-dione binds with aromatase very weakly, whereas its 4-dehydro isomer is an efficient AI with Ki=37 nM.15,145,152 The presence of a substituent at position 10 of the steroid skeleton of ADis indispensable for activity. As has been shown in Section III, this substituent plays an essential role in the oxidation of the steroid substrate by the enzyme.Thus 19-nor-4-hydroxyandrost- 4-ene-3,17-dione binds with aromatase but is devoid of inhibitory activity.15, 64 Introduction of additional double bonds into the molecule of AD and its substituted analogues enhances their inhibitory activity. It is particularly true for 1,2- and 6,7-double bonds. It is assumed 123 that the presence of a 1,2-double bond in a steroid molecule is responsible for the irreversible type of inhibition and, correspondingly, for the long-lasting effect of AI in vivo.Introduction of various substituents into the rings A and B (positions 1, 4, 6, 7 and 19) as well as alkylation or arylation of 4-amino-, -mercapto- or -hydroxy-androstanes are the most common structural modifications. High affinity for the enzyme has been noted for steroid AI with small substituents at C(1), C(4) and C(19) and with bulky substituents at C(7).Introduction of hydrophobic substituents is a general trend in the design of efficient AI, although the optimal volume and position of func- tional groups are often difficult to predict. 2a-Substitution and involvement of the C(6) and C(7) atoms in the formation of a three-membered heterocycle have a negative effect.14, 19 Thus, studies of structure ± activity relationships of steroid AI and their interactions with the active site of aromatase provides valuable information about the nature of the enzyme and allows one to gain better insight into the mechanism of action of this enzyme.Investigation of this mechanism is far beyond the scope of the present review; it is evident, however, that the synthesis and analysis of activity of various series of steroid AI represent a step on the way to its elucidation.V. In vivo activities and applications of major steroid aromatase inhibitors As can be seen from Tables 1 ± 4, highly efficient AI usually have similar Ki which are close to the lower limits of sensitivity (0.5 ± 1.0 nM). The possibility of their practical application will further be determined by their biological properties in vivo, pharmacological and pharmacokinetic characteristics, the pres- 191 N S O O R=Ms (a), Ac (b). 190a,b O N3 OR O 19 1 2 4 6 7 O 17 O Me, D1 19-Nor 19-substitution 3-deoxy SH OH, OR, SR Alk, D6, O 7a-substitution 17b-OH(OR) Figure 1. The functional groups affecting the inhibitory activity and their position in the AD molecule.The arrows directed towards a specific position denote enhancement of activity, those directed from this position denote decrease in activity. Substituted androstanes as aromatase inhibitors 995ence or absence of side hormonal effects typical of some steroids of the androstane series, and, naturally, by the methods of their synthesis and the possibility of scaling.This section considers pharmacological activities and clinical applications of several major substituted androstanes the proper- ties of which have been studied in sufficiently great detail. Owing to their antitumour activity, some of these steroid compounds, viz., 4-hydroxy-AD 91, 1-methyl-ADD 121 and 6-methylene- ADD136, have found wide use in the clinical practice.The former known under the trade name formestane is already used in the therapeutic practice, while the other two are undergoing clinical trials. Besides, there is ample literature evidence concerning the pharmacological activity of another two AI. The first of them, 19- ethynyl-AD 1 (promestane), manifests high activity in inhibiting ovarian aromatase on oral administration and efficiently reduces mammary and ovarian tumours in both experimental models and the clinic.15, 36, 37 The synthetic `Organon' product, 19-ethyldithio- AD, possesses very weak androgenic activity but very efficiently inhibits estrogen biosynthesis in vivo.43, 44 Antitumour activities of these AI have been surveyed recently.21, 153, 154 4-Hydroxy-AD (91, formestane) was designed in 1984 by the `Ciba-Geigy' company for treatment of post-menopausal mam- mary tumours in women and is now commercially available in Great Britain, Germany, Austria and some other countries.71 Some patents issued by the `Schering AG' company deal with their application in the therapy of benign prostatic hyperplasia.65, 155 Studies of formestane effects on human placenta aromatase revealed its high inhibitory activity.62, 64 Formestane produced a significant antitumour effect in rats with experimental mammary tumours.36, 70, 87 This inhibition is highly selective.63 However, there is evidence that formestane also inhibits 5a-reductase.69, 156 One main disadvantage of formestane reported by some authors 48, 67, 88, 98, 157 is its low bioavailability and short-term effect.Since this drug is highly active on subcutaneous admin- istration,9 it is normally used in the form of injections. The results of clinical trials of this drug in patients with mammary tumours (500 ± 1000 mg weekly, i.m.) testify to its high efficiency.157 ± 160 According to the experimental protocol elaborated for IInd phase of clinical trials, formestane injected once or twice a week causes partial or complete remission of the tumour.66, 68, 158, 159 This drug is generally well tolerated and produces only nonspecific side effects.One should mention in this connection that replacement hydrocortisone therapy was necessary for patients receiving for- mestane for a prolonged period, since this drug inhibits pregne- nolone biosynthesis and, as a consequence, biosynthesis of corticosteroids.67, 161 According to the literature data, atamestane and exemestane have no such shortcomings. Comparative characteristics of in vivo activities of formestane, atamestane and exemestane and analysis of mechanisms of their action in animals have been docu- mented.9, 69 1-Methyl-ADD or atamestane was designed by the `Schering AG' company in 1983 102 and patented in more than 30 countries.The main objective of its clinical trials initiated in 1986 in Europe and Japan was to estimate its efficiency in the treatment of mammary tumours and endometriosis in women and of benign prostatic hyperplasia in men. As mentioned above, atamestane is a selective irreversible inhibitor of estrogen biosynthesis in vitro and in vivo.Studies in rats revealed that its activity exceeds that of formestane almost 1.5-fold;102, 162 according to other data, its effect is weaker than those of formestane and examestane.9 DMBA-induced mammary tumours in rats reduced appreciably after subcutaneous administration of atamestane (30 or 150 mg per kg per day).103 Neither atamestane nor its main metabolites, viz., 5b- and 17b-reduced compounds, manifest any affinity for steroid receptors, which correlates with the endocrine-pharmaco- logical profile of this drug.104 In vivo studies failed to detect estrogen, anti-estrogen, androgen, progestane or antigonado- tropic activities in atamestane.105 As the affinity of atamestane for enzymes responsible for steroid biosynthesis in adrenal glands is negligibly small in comparison with its affinity for aromatase, it does not exert any appreciable influence on these glands even when used in high doses and upon prolonged administration.This finding has been confirmed experimentally.105 As has been mentioned above, atamestane has been recom- mended for treatment of benign prostatic hyperplasia.154, 163 Along with other factors, estrogens play an important role in the etiology of this disease.Apart from the finging that the incidence of this disease correlates with a shift of the androgens/estrogens ratio towards estrogens upon ageing in males, the presence of the estrogen receptor meeting the classical criteria has also been demonstrated by Habenicht et al.164, 165 The role of estrogens in the pathogenesis of this disease revealed in animals has been confirmed by clinical tests.164 The results of successful clinical trials of atamestane administered in an oral dose of 200 mg 3 times daily for 3 months in the endocrine therapy of benign prostatic hyperplasia were published several years ago.166 The use of atamestane in the form of tablets for increasing testosterone content in the organism has been patented.162 Exemestane 136 has also been recommended for clinical use.This preparation was designed in 1988 by the Italian company `Farmitalia Carlo Erba' and tested in many countries of the world as an independent therapeutic agent for treating estrogen-depend- ent tumours, especially post-menopausal mammary tumours in women. The choice of exemestane is based on its pharmacological properties which characterise this drug as a highly active (on oral administration) selective AI.9, 97 Studies in animals revealed long- term inhibition of ovarian aromatase on subcutaneous and oral administration.9, 97, 167 The activity of exemestane exceeds that of formestane twofold on subcutaneous administration and 27-fold on oral administration.123 This drug does not manifest any affinity for the majority of steroid receptors with the exception of its weak binding with the androgenic receptor; the androgenic activity of exemestane was as low as 10% of the testosterone activity.98, 123 The results of endocrinological clinical trials of small doses of exemestane published in 1995 testify to a long-term selective decrease in estrogen levels and good tolerance of this drug.168, 169 The antitumour effect of exemestane has been studied in a model of DMBA-induced rat mammary tumours.167 Subcutaneous or oral administration of exemestane caused partial or complete remission of the tumour.123, 124 Exemestane has a specific action on women: it does not influence the levels of hydrocortisone, aldosterone, 17-hydroxyprogesterone and luteinising and follicle- stimulating hormones even when used in very high doses.123 The composition of tablets containing exemestane 170 and designed for the treatment of mammary tumours and the use of this AI as a component of a combined preparation for prophylaxis and treat- ment of benign prostatic hyperplasia have been patented.155 VI.Conclusion Thus, substituted androstanes are aromatase inhibitors in their pharmacological effect and antitumour agents in the mode of action.They may compete successfully with other drugs and are sometimes indispensable in the treatment of hormone-dependent tumours. Special mention should be made of the relative simplicity of synthesis of these preparations from industrially available steroid sources.Analysis of methods used for the synthesis of AI together with detailed studies of their activity in vitro and in vivo give a strong impetus to the design of new efficient drugs and the development of methods for their synthesis and offers a possibility to choose the most efficient drug in each concrete case. References 1. A M H Brodie Biochem. Pharmacol. 34 3213 (1985) 2.R J Santen Breast Cancer Res. Treat. (Suppl. ) 7 23 (1986) 996 I S Levina3. A M H Brodie, R C Coombes, M Dowsett J. Steroid Biochem. 27 899 (1987) 4. H V Bossche J. Steroid Biochem., Mol. Biol. 43 1003 (1992) 5. P S Lonning,M Dowsett, T J Powles J. Steroid Biochem. 35 355(1990 ) 6. J L Young Breast Cancer (Ed.ARLiss) (New York: Academic Press, 1989) p. l 7. E A Thompson, P K Siiteri J. Biol.Chem. 249 5373 (1974) 8. P L Bellino J. Steroid Biochem. 17 261 (1982) 9. E DiSalle, D Giudici, G Briatico, G Ornati Ann. N.Y. Acad. Sci. 595 357 (1990) 10. P K Siiteri, E A Thompson J. Steroid Biochem. 6 317 (1975) 11. J P Burkhart,N P Peet, C L Wright, J O N Johnston J. Med. Chem. 34 1748 (1991) 12. J Fishman, E F Hahn Steroids 50 339 (1987) 13. J Mann, B Pietrzak J.Chem. Soc., Perkin. Trans. 1 2681 (1983) 14. P J Bednarski, S D Nelson J. Med. Chem. 32 203 (1989) 15. P A Cole, C H Robinson J. Med. Chem. 33 2933 (1990) 16. J N Wright,M Akhtar Steroids 55 142 (1990) 17. J N Wright, G Slatcher,M Akhtar Biochem. J. 273 533 (1991) 18. D F Covey, in Sterol Biosynthesis Inhibitors (Eds D Berg, MPlempel) (Chichester: Ellis Horwood, 1988) p. 535 19. P J Bednarski, S D Nelson J. Steroid Biochem. 32 309 (1989) 20. F J Zeelen Medicinal Chemistry of Steroids (Amsterdam: Elsevier, 1990) p. 218 21. J M O'Reilly, R W Brueggemeier Curr. Med. Chem. 3 11 (1996) 22. W C Schwarzel, W G Krugel, H J Brodie Endocrinology 92 866 (1973) 23. E A Thompson, P K Siiteri J. Biol. Chem. 249 5364 (1974) 24. D F Covey,W F Hood Cancer Res. (Suppl. 8) 42 3327 (1982) 25. R W Brueggemeier, P K Li, C E Snider, M V Darby, N E Katlic Steroids 50 163 (1987) 26. J O Johnston, C L Wright, G W Holbert J. Steroid Biochem., Mol. Biol. 52 17 (1995) 27. D F Covey, V D Parikh,W W Chien Tetrahedron. Lett. 2105 (1979) 28. D F Covey,W F Hood, V D Parikh J. Biol. Chem. 256 1076 (1981) 29. B W Metcalf, C L Wright, J P Burkhart, J O Johnston J. Am.Chem. Soc. 103 3221 (1981) 30. P A Marcotte, C H Robinson Steroids 39 325 (1982) 31. J O'N Johnston Steroids 50 105 (1987) 32. BRD P. 3 644 358; Chem. Abstr. 108 75 715 (1988) 33. BRD P. 3 644 301; Chem. Abstr. 108 75 716 (1988) 34. Eur. P. 227 472; Chem. Abstr. 107 21 7931 (1987) 35. BRD P. 3 706 647; Chem. Abstr. 109 93 426 (1988) 36. E Perel, S P Davis, D F Covey, D W Killinger Steroids 38 397 (1981) 37.S J Zimnitzki,M T Brandt, D F Covey, D Puett Steroids 50 135 (1987) 38. J N Wright, P T van Leersum, S G Chamberlin, M Akhtar J. Chem. Soc., Perkin Trans. 1 1647 (1989) 39. P A Marcotte, C H Robinson Cancer Res. (Suppl. 8) 42 3322 (1982) 40. M G B Drew, J Mann, E Pietrzak J. Chem. Soc., Chem. Commun. 1191 (1985) 41. J N Wright,M Calder,M Akhtar J. Chem.Soc., Chem. Commun. 1733 (1985) 42. Jpn. P. 62 155 295; Chem. Abstr. 107 198 752 (1987) 43. P J Bednarski, D J Porubek, S D Nelson J. Med. Chem. 28 775 (1985) 44. J A A Geelen, G H J Deckers, J T H van der Wardt, H J J Loozen, L J W Tax, H J Kloosterboer J. Steroid Biochem., Mol. Biol. 38 181 (1991) 45. Eur. P. 149 499; Chem. Abstr. 104 51 017 (1986) 46. D Lesuisse, J F Gourvest, Q Benslimane, F Canu, Ch Delaisi, B Doucet, Ch Hartmann, J M Lefrancois, B Tric, D Mansuy, D Philibert, G Teutsch J.Med. Chem. 39 757 (1996) 47. D Lesuisse, F Canu, B Tric Tetrahedron 50 8491 (1994) 48. D Lesuisse, J F Courvest, C Hartmann, B Tric, O Benslimane, D Philibert, J P Vevert J. Med. Chem. 35 1588 (1992) 49. C Delaisi, B Doucet, C Hartmann, B Tric, J F Gournet, D Lesuisse J.Steroid Biochem., Mol. Biol. 41 773 (1992) 50. M J Shih,M H Carrell, H L Carrell, C L Wright, J O N Johnston, G H Robinson J. Chem. Soc., Chem. Commun. 213 (1987) 51. W E Childers, M J Shih, P S Furth, C H Robinson Steroids 50 121 (1987) 52. W E Childers, P S Furth,M J Shih, C H Robinson J. Org. Chem. 53 5927 (1988) 53. W E Childers, C H Robinson J. Chem. Soc., Chem. Commun. 320 (1987) 54. J T Kellis,W E Childers, C H Robinson, L E Vickery J. Biol. Chem. 262 4421 (1987) 55. W E Childers, J V Silverton, J T Kellis, L E Vickery, C H Robinson J. Med. Chem. 34 1344 (1991) 56. V C O Njar, E Safi, J V Silverton, C H Robinson J. Chem. Soc., Perkin. Trans. 1 1161 (1993) 57. J A Lovett, M V Darby, R S Counsell J. Med. Chem. 27 734 (1984) 58. V C O Njar, J Duerkop, R W Hartmann Steroids 61 138 (1996) 59.Eur. P. 488 383; Chem. Abstr. 117 131 436 (1992) 60. US P. 5 491 136 ; Chem. Abstr. 124 289 995 (1996) 61. N P Peet, J P Burkhart, C L Wright, J O'N Johnston J. Med. Chem. 35 3303 (1992) 62. A M H Brodie, W C Schwarzel, A A Shaikh, H J Brodie Endocrinology 100 1684 (1977) 63. A M H Brodie, W M Garrett, J R Hendrickson, C H Tsai-Morris, P A Marcotte, C H Robinson Steroids 38 693 (1981) 64.D F Covey,W F Hood Mol. Pharmacol. 21 173 (1982) 65. Eur. P. 181 287; Chem. Abstr. 105 102 601 (1986) 66. R C Coombes, P E Goss, M Dowsett, G Hutchinson, D Cunningham, M Jarman, A M H Brodie Steroids 50 245 (1987) 67. A M H Brodie, L Y Wing Steroids 50 89 (1987) 68. M J Reed, L C Lai, A M Owen, A Singh, N G Coldham, A Purohit,M W Ghilchik, N A Shaikh,W H T James Cancer Res. 50 193 (1990) 69. A M H Brodie, P K Banks, S E Inkster,M Dowsett, R C Coombes J. Steroid Biochem., Mol. Biol. 37 327 (1990) 70. P G Ciattini, E Morera, G Ortar Synth. Commun. 22 1949 (1992) 71. A M H Brodie J. Steroid Biochem. 49 281 (1994) 72. R H Bible, C Placek, R D Muir J. Org. Chem. 22 607 (1957) 73. H B Henbest, W R Jackson J. Chem. Soc., C 2459 (1967) 74.D A March, H J Brodie,W M Garrett, C H Tsai-Morris, A M H Brodie J. Med. Chem. 28 788 (1985) 75. S Hrycko, P Morand J. Chem. Soc., Perkin. Trans. 1 2899 (1990) 76. R D Burnett, D H Kirk J. Chem. Soc., Perkin. Trans. 1 1830 (1973) 77. B H Jennings, J M Bengtson Steroids 31 49 (1978) 78. D P Jindal, M R Yadav Ind. J. Chem. 30B 515 (1991) 79. M Numazawa,M Ogata J. Chem.Soc., Chem. Commun. 1092 (1986) 80. X P Liu, D M Lambert, Y J Abul-Hajj J. Med. Chem. 38 4135 (1995) 81. Y J Abul-Hajj, X P Liu,M Hedge J. Steroid Biochem., Mol. Biol. 54 111 (1995) 82. E J Tavares da Silva,M L Sa e Melo, A S Campos Neves J. Chem. Soc., Perkin. Trans. 1 1649 (1996) 83. J A R Salvador, M L Sa e Melo, A S Campos Neves Tetrahedron. Lett. 34 357 (1993) 84. Br. P. 2 191 198; Chem. Abstr. 108 204 885 (1988) 85. A M H Brodie, D A Marsh, H J Brodie J. Steroid Biochem. 10 423 (1979) 86. J L Block, N L Block, B L Lokeshwar Cancer Lett. 101 143 (1966) 87. A M H Brodie, W M Garrett, J P Hendrickson, C H Tsai-Morris Cancer Res. (Suppl. 8) 42 3360 (1982) 88. A B Foster,M Jarman, J Mann, I B Parr J. Steroid Biochem. 24 607 (1986) 89. BRD P. 3 213 328; Chem.Abstr. 100 85 993 (1984) 90. Y J Abul-Hajj J. Med. Chem. 29 582 (1986) 91. Y J Abul-Hajj J. Steroid Biochem. 34 439 (1989) 92. Eur. P. 375 559; Chem. Abstr. 114 24 321d (1991) 93. H F Hodson, D J Madge, D H Widdowson J. Chem. Soc., Perkin. Trans. 1 2965 (1995) 94. M G Rowlands, A B Foster, J Mann, B Pietrzak, J Wilkinson, R C Coombes Steroids 49 371 (1987) 95. Eur. P. 291 290; Chem.Abstr. 110 251 323 (1989) 96. BRD P. 3 604 179; Chem. Abstr. 105 191 504 (1986) 97. D Giudici, G Ornati, G Briatico, F Buzzetti, P Lombardi, E DiSalle J. Steroid Biochem. 30 391 (1988) 98. E Di Salle, D Giudici, G Ornati, G Briatico, R, D'Alessio, V Villa, P Lombardi J. Steroid Biochem., Mol. Biol. 37 369 (1990) 99. T Zaccheo, E DiSalle Cancer Chemother. Pharmacol. 31 308 (1993) Substituted androstanes as aromatase inhibitors 997100.Br. P. 2 284 605; Chem. Abstr. 123 228 634 (1995) 101. Jpn. P. 05 178 888; Chem. Abstr. 120 192 083 (1994) 102. BRD P. 3 322 285; Eur. P. 129 500; Chem. Abstr. 102 204 169 (1985) 103. Y Nishino, M R Schneider, H Michna,M F El Etreby J. Steroid Biochem. 34 435 (1989) 104. D Herderson, in Atamestaneìa New Aromatase Inhibitor Sympo- sium (Abstracts Reports of VIIIth International Congress on Hormones Steroids), The Hague, 1990 p. 1 105.M F El Etreby, in Atamestaneìa New Aromatase Inhibitor Symposium (Abstracts Reports of VIIIth International Congress on Hormones Steroids), The Hague, 1990 p. 6 106. BRD P. 2 046 640; Chem. Abstr. 76 154 025 (1972) 107. BRD P. 3 720 234; Chem. Abstr. 111 7673 (1989) 108. A B Turner, H J Ringold J. Chem. Soc. C 1720 (1967) 109. BRD P. 3 512 328; Chem. Abstr. 106 48 639 (1987) 110. WO PCT 86 05 813 (1986) 111. BRD P. 3 338 212; Chem. Abstr. 103 123 792 (1985) 112. BRD P. 3 539 244; Chem. Abstr. 108 75 717 (1988) 113. M Lourdusamy, F Labrie, S M Singh Synth. Commun. 25 3655 (1995) 114. J Westermann, K Nickisch Angew. Chem., Int. Ed. Engl. 32 1368 (1993) 115. BRD P. 3 715 869; Chem. Abstr. 111 7671 (1989) 116. BRD P. 4 015 247; Chem. Abstr. 116 59 740 (1992) 117. Eur. P. 534 582; Chem. Abstr. 119 117 611 (1993) 118. BRD P. 4 227 053; Chem. Abstr. 120 270 960 (1994) 119. BRD P. 3 703 722; Chem. Abstr. 110 95 625 (1989) 120. BRD P. 3 705 990; Chem. Abstr. 110 95 626 (1989) 121. Eur. P. 265 119; Chem. Abstr. 109 93 434 (1988) 122. E DiSalle, G Briatico, D Giudici, G Ornati, T Zaccheo J. Steroid Biochem. 34 431 (1989) 123. E DiSalle, G Ornati, D Giudici,M Lassus, T R J Evans, R C Coombes J. Steroid Biochem., Mol. Biol. 43 137 (1992) 124. F Buzzetti, E DiSalle, A Longo, G Briatico Steroids 58 527 (1993) 125. K Annen, H Hofmeister, H Laurent, R Wiechert Synthesis 34 (1982) 126. Br. P. 8 801 697; Chem. Abstr. 112 36 257 (1990) 127. Eur. P. 307 314; Chem. Abstr. 111 134 643 (1989) 128. GDR P. 258 820; Chem. Abstr. 111 7672 (1989) 129. GDR P. 264 220; Chem. Abstr. 111 78 502 (1989) 130. WO PCT 95 01366; Chem. Abstr. 122 291 310 (1995) 131. M Numazawa,M Oshibe J. Med. Chem. 37 1312 (1994) 132. Jpn. P. 063 321 981; Chem. Abstr. 122 265 773 (1995) 133. M Numazawa,M Oshibe Steroids 60 506 (1995) 134. M Numazawa, T Kamijama, M Tachibana,M Oshibe J. Med. Chem. 39 2245 (1996) 135. M Numazawa,M Oshibe, S Yamaguchi, M Tachibana J. Med. Chem. 39 1033 (1996) 136. H L Holland, S Kumaresan, L Tan, V C O Njar J. Chem. Soc., Perkin. Trans. 1 585 (1992) 137. C E Snider, R W Brueggemeier J. Biol. Chem. 262 8685 (1987) 138. S Ebrahimain, H H Chen, R W Brueggemeier Steroids 58 414 (1993) 139. Jpn. P. 07 215 992; Chem. Abstr. 123 340 545 (1995) 140. M V Darby, J A Lovett, R W Brueggemeier,M P Groziak, R E Counsell J. Med. Chem. 28 803 (1985) 141. J M O'Reilly, N Li, R W Brueggemeier J. Med. Chem. 38 2842 (1995) 142. WO PCT 91 12 206 ; Chem. Abstr. 115 232 607 (1991) 143. V C O Njar, R W Hartmann, C H Robinson J. Chem. Soc., Perkin. Trans. 1 985 (1995) 144. V C O Njar, G Gruen, R W Hartmann J. Enzyme Inhib. 9 195 (1995) 145. M Numazawa, A Mutsumi Biochem. Biophys. Res. Commun., 177 401 (1991) 146. WO PCT 94 10 190; Chem. Abstr. 121 231 153 (1994) 147. H L Herbert, S Kumaresan, G Lakshmaiah Can. J. Chem. 73 2185 (1995) 148. S Miyairi, J Fishman J. Biol. Chem. 261 6772 (1986) 149. Eur. P. 289 451; Chem. Abstr. 110 95 629 (1989) 150. M Numazawa,M Tsuji, A Mutsumi J. Steroid Biochem. 28 337 (1987) 151. M Yoshihama, K Tamura,M Nakakoshi, J Nakamura, N Fujise, G Kawanishi Chem. Pharm. Bull. 38 2834 (1990) 152. M Numazawa, A Mutsumi, K Hoshi, R Koike Biochem. Biophys. Res. Commun. 160 1009 (1989) 153. K Hoeffken Contrib. Oncol. (Hormone-Dependant Tumors) 50 324 (1995); Chem. Abstr. 124 249 307 (1996) 154. U F Habenicht, U W Tunn, T Senge, F H Schroder, H U Schweikert, G Barsch,M F El Etreby J. Steroid Biochem., Mol. Biol. 44 557 (1993) 155. WO PCT 91 00 731; Chem. Abstr. 115 174 681 (1991) 156. A M H Brodie, C Son, D A King, K M Meyer, S E Inkster Cancer Res. 49 6551 (1989) 157. R C Coombes, P Goss, M Dowsett, J C Gazet, A M H Brodie The Lancet 2 1237 (1984) 158. A M H Brodie, L-Y Wing, P Goss, M Dowsett, R C Coombes J. Steroid Biochem. 24 91 (1986) 159. P Goss, T J Powles, M Dowsett Cancer Res. 46 4823 (1986) 160. P E Lonning, D C Johannessen Drug Today 27 117 (1991) 161. P E Lonning, S Kvinnsland Drugs 35 685 (1988) 162. BRD P. 4 435 368 FRG; Chem. Abstr. 124 279 193 (1996) 163. M F El Etreby J. Steroid Biochem., Mol. Biol. 44 565 (1993) 164. U F Habenicht, in Atamestaneìa New Aromatase Inhibitor Symposium (Abstracts Reports of VIIIth International Congress on Hormones Steroids), The Hague, 1990 p. 10 165. D Henderson, U F Habenicht, Y Nishino,M F El Etreby Steroids 50 219 (1987) 166. H V Schweikert, U W Tunn, U F Habenicht J. Steroid Biochem. 44 573 (1993) 167. T Zaccheo, D Giudici, P Lambardi, E Di Salle Cancer Chemother. Pharmacol. 23 47 (1989) 168. T R J Evans, E Di Salle, G Ornati,M Lassus, M Benedetti, E Pianezzola, R C Coombes Cancer Res. 52 5933 (1992) 169. N Zilembo, C Noberasco, E Bajetta, A Mortinetti, L Mariani, S Orefice, R Buzzoni Br. J. Cancer 72 1007 (1995) 170. BRD P. 3 622 841; Chem. Abstr. 106 162 586 (1987) 171. D F Covey,W F Wood Endocrinology 108 1599 (1981) 172. P F Sherwin, P C McMullan, D F Covey J. Med. Chem. 32 651 (1989) 173. Y Osawa, M J Coon Endocrinology 121 1010 (1987) 998 I S Levina

ISSN:0036-021X    

出版商:RSC    

年代:1998

数据来源: RSC