摘要:

Russian Chemical Reviews 71 (9) 707 ± 720 (2002) Nucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology for constructing heterocyclic systems O N Chupakhin, D G Beresnev Contents I. Introduction II. Main transformation paths of sH adducts III. Construction of azine-based heterocyclic systems IV. Construction of nitroarene-based heterocyclic systems V. Conclusion N , Abstract. by azines fused of synthesis the for methods new on Data Data on new methods for the synthesis of fused azines by nucleophilic of atom carbon unsubstituted the on attack nucleophilic attack on the unsubstituted carbon atom of azines azines and surveys review The generalised. are nitroarenes and nitroarenes are generalised. The review surveys tandem tandem reactions comprising binucleophiles with azines of reactions of azines with binucleophiles comprising nucleophilic nucleophilic addition addition A substitution ± addition nucleophilic , N±AN, nucleophilic addition ± substitution AN±Sipso and nucleophilic substitutions S N Intramolec- N and ±Sipso and SHN ±SHN .Intramolec- and nucleophilic substitutions SH ular ular SH reactions of azines and nitroarenes and other methods for N reactions of azines and nitroarenes and other methods for the on based heterocycles fused of synthesis the synthesis of fused heterocycles based on SHN nucleophilic substitution are also discussed. The bibliography includes 69 substitution are also discussed. The bibliography includes 69 references. I.Introduction The main procedures for the functionalisation of aromatic com- pounds involve electrophilic (SEAr) and nucleophilic (SNAr) substitution reactions. The SEAr reactions, typical primarily of carbocyclic compounds, proceed predominantly as the electro- philic substitution of hydrogen. Strictly speaking, these trans- formations would be more correctly designated as SHE Ar because less common reactions accompanied by the displacement of non- hydrogen groups (CO2R, NO2, NR2 , etc.) are conventionally symbolised by Sipso E Ar. Transformation SEAr (SHE Ar): E +E+ +H+. X XTransformation Sipso E Ar: E Y +Y+ X +E+ X E is an electrophile. Nucleophilic aromatic substitution reactions involve most often the ipso-substitution (Sipso N Ar) at the CAr atom bound to the O N Chupakhin, D G Beresnev Institute of Organic Synthesis, Urals Branch of the Russian Academy of Sciences, ul.S Kovalevskoi 20, 620219 Ekaterinburg, Russian Federation. Fax (7-343) 274 11 89. Tel. (7-343) 274 11 89. E-mail: chupakhin@ios.uran.ru (ONChupakhin). Tel. (7-343) 249 30 58. E-mail: beresnev@ios.uran.ru (D G Beresnev) Received 24 June 2002 Uspekhi Khimii 71 (9) 803 ± 818 (2002); translated by T N Safonova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n09ABEH000747 707 707 709 718 719 Hal, N, O, P or S atoms.1 Some SNHreactions have long been used in industry (synthesis of nitrophenols, alizarin, 2-aminopyridine). The preparation of p-phenylenediamine from nitrobenzene and benzamide is currently one of the most advanced organic synthe- ses used in industrial processes.2 SNAr (Sipso N Ar) Nu Z +Z7 EWG +Nu7 EWG SHN (SHN Ar) Nu EWG +Nu7 + [H7] EWG Nu is a nucleophile; EWG is an electron-withdrawing group.N Presently, the Sipso and SHN reactions and related processes are used as the main procedures for the modification of p-electron- deficient aromatic systems.1, 3 It should be noted that methods based on nucleophilic attack on the unsubstituted carbon atom of aromatic compounds have some advantages. Thus, the prelimi- nary introduction of nucleofuges (Cl, NO2, SO3H, etc.) into the ring is not needed here and the hydride ion serves formally as the leaving group.In many reactions, water is removed from the reaction mixture as a by-product, which is of importance for the commercial use of these reactions in connection with increasing industrial requirements for environmental safety. It should be noted that an aromatic substrate must be activated to be involved in the SHN processes. In reactions with arenes, this is most often achieved by the introduction of an electron-withdrawing substituent, for example, the nitro group, into the ring. In the reactions with heterocyclic compounds, the presence of the aza group is often quite sufficient. This group is comparable to NO2 in its electron-withdrawing effect and also directs nucleophilic attack at positions 2 and 4.1 The present review surveys the data on the use of nucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes for the construction of new heterocyclic systems.II. Main transformation paths of sH adducts According to the modern notion of nucleophilic aromatic sub- stitution reactions proceeding as `addition ± elimination', the first step of the nucleophilic attack on activated aromatic substrate 1 gives rise to intermediates of two types, viz., sX adducts (Meisen- heimer complexes) 2 or sH adducts (Servis complexes) 3. Even in the presence of a leaving group in the ring, the ipso-substitution is708 always preceded by the rapid and reversible formation of sH adducts as kinetically controlled products.4 EWG EWG Sipso N (slowly) 7 X 7X7 EWG Nu Nu Nu7 2 sX adduct X EWG EWG SHN1 (rapidly) [O] 7 H 7[H7] Nu X X Nu 3 sH adduct X is a leaving group; EWG is N2, NO2, NO, CO2R, etc.The sH adducts 3 differ substantially in stability. Thus, both unstable short-lived species and stable anionic complexes and dihydro compounds are known.1 In the nitroarene series, the stability of sH adducts increases as the number of electron- withdrawing groups in the nucleus is increased. Thus, the exis- tence of sH adducts, viz., di- and trinitrobenzene derivatives, was unambiguously proved, whereas the corresponding mononitro- benzene derivatives were not, to our knowledge, detected. The sH intermediates derived from azines can be stabilised both by the introduction of electron-withdrawing groups, includ- ing the aza group, and through protonation of anionic forms and their transformations into neutral dihydro compounds.Clearly, the sH adducts generated from cationic forms of azines do not bear a charge. On the whole, sH adducts (Servis complexes) are generally more stable than their sX analogues (Meisenheimer complexes). The nature of the leaving group is the main factor responsible for the high stability of sH adducts. Unlike good leaving groups, such as Hal, CN, NO2, SO3 H, etc., which are readily solvated asX7 anions, the hydride ion is not susceptible to solvation and it was not detected as a kinetically independent species in solutions. The occurrence of hydrogen in the form of the proton H+ or the radical H.is thermodynamically much more favourable than its occurrence as the hydride ion H7. Therefore, aromatisation, which proceeds either through oxidative processes or with the involvement of auxiliary leaving groups (autoaroma- tisation) rather than via the direct abstraction of a hydride ion, is the bottleneck of SHN processes. Hence, nucleophilic attack on the unsubstituted carbon atom of arenes (hetarenes) affords (under particular conditions) SHN -reaction products. If none of the mech- anisms of aromatisation takes place, sH adducts can undergo dissociation. This is particularly true for sH adducts generated from nitroarenes. The sH adducts derived from heteroarenes can be involved in alternative reactions resulting in the ring opening or recyclisation (for example, ANRORC transformations).5, 6 Oxidative aromatisation is a variation of the SHN reactions in which elimination occurs as a redox process.1 In this type of reaction, both inorganic (atmospheric oxygen, halogens, hypoha- logenites, sulfur, metal cations, etc.) and organic (quinones, triarylmethyl, tropylium and heterocyclic cations, etc.) reagents can be used as oxidisers.Besides, oxidative aromatisation can be performed electrochemically. In the absence of an oxidising agent, the starting p-electron-deficient arene (heteroarene) can serve as such a reagent. An oxidiser abstracts electrons from the sH adduct and hydrogen is most often eliminated as a proton.1 As an example of the oxidative SHN processes, we refer to the reactions of naphthalene activated by chromium carbonyl with organolithium compounds.The resulting sH adduct undergoes aromatisation under the action of elemental iodine.7 O N Chupakhin, D G Beresnev CN Me Me Me Me CN H I2 LiCMe2CN 7 THF, 0 8C Cr(CO)3 Cr(CO)3 (96%) Autoaromatisation proceeds if the sH adduct contains a nucleofuge, which is initially present in either the substrate or the reagent. In this case, the nucleofuge group X (which is readily solvated) serves as an oxidiser and is eliminated together with the electron pair from the sH adduct as the X7 anion, whereas hydrogen leaves as a proton. There are several types of autoaromatisation of sH adducts. If the leaving group is located in the aromatic ring in the ortho position with respect to the carbon atom subjected to the attack, these reactions are said to be cine-substitution reactions.Thus in the reactions of 6-bromoazolopyrimidines with indole and its derivatives, 8 the hydrogen atom at position 7 of the triazolopyr- imidine system is replaced and the bromine atom leaves the adjacent position 6. HN Br N H N + Br N N N HN NN NH HN HN H Br 7HBr N N N H N N N N N (78%) Numerous SHN reactions in the series of azine N-oxides and their derivatives (deoxygenative substitution of hydrogen) 9 belong to the cine-substitution processes. In this case, the N-oxide, N-alkoxide or N-acyloxy functional group serves as the auxiliary leaving group in the second step involving elimination of the hydrogen atom from the sH adduct.For example, in the reactions of substituted 1,2,4-triazine N-oxides with cyanamide, sH adduct 4 undergoes aromatisation through 1,2-elimination of the water molecule.10 Ph N N N N H2NCN, B + 7H2O R R N N PhH HNCN O7 OH 4 N Ph Ph NH N N R N R HN N NCN CN R=H, Et, Ph; B is a base. If the reaction centre and the nucleofuge group X are separated by one or more atoms, the reactions are said to belong to tele-substitution processes. Here, the substituent X can be either bound to the aromatic nucleus (Scheme 1) or located in the side chain (Scheme 2). In the former case (see Scheme 1), the reactions proceed through the formation of sH adduct 5, which undergoes aromatisation via 1,4-elimination of HBr with abstrac- tion of the tele-arranged bromine atom.Nucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology... Scheme 1 Br Br 7 N N NH2 KNH2 H+ N N H 7HBr 5 N NH2 N Scheme 2 Cl CCl3 CCl2 MeO2CCH2SH H Et3N, MeOH, D MeO S NO2 N O 7 O 7O CCl2 H MeO S NO2 O CHCl2 O Zn, CaCl2 S S H2O, MeOH, D NH MeO NO2 O (52%) O 6 In the latter case (see Scheme 2), aromatisation of the sH adduct is achieved through the abstraction of the chlorine atom from the trichloromethyl substituent.Reduction of SHN -reaction product 6 with zinc led to the thiazine-ring closure.11 The so-called vicarious nucleophilic substitution where a good leaving group X is present in the nucleophile residue belongs to well-studied autoaromatisation reactions of sH adducts.12 ± 15 For example, the attack of sulfenamides on the unsubstituted carbon atom of 3-chloro-1-nitrobenzene (7) afforded adduct 8 whose structure allows the RSH group to undergo b-elimination.15 SR NH H NH Cl Cl H+ Cl+H2NSR ButOK DMF 7RSH 7 7 NO2 NO2 8 NO2 7 NH2 Cl NO2 Vicarious autoaromatisation of sH adducts readily proceeds in the case of azoles,16 azines 17 and nitroarenes 17 and this process provides the basis for the construction of heteroaromatic systems, for example, of the indole 18, 19 and quinoline 20 rings.III. Construction of azine-based heterocyclic systems Azines containing two or more heteroatoms differ substantially from nitroarenes in that they can be subjected to the double nucleophilic attack.One-pot syntheses of fused heterocyclic 709 systems are based on the tendency of azines to form cycloadducts in reactions with bifunctional nucleophilic reagents. 1. Tandem reactions AN±AN There are a large number of examples of the construction of fused heterocyclic systems by the replacement of two adjacent good leaving groups in pyridine, quinoline, pyrimidine, pyrazine, qui- noxaline and other rings under the action of bifunctional nucleo- philes.21 The main limitation of this method is that only azaaromatic compounds containing two nucleofuge groups at positions 1 and 2 can be used in such reactions. The ability of unsubstituted azines to form cycloadducts in reactions with binucleophiles provides the basis for another procedure for the synthesis of fused heterocyclic systems.It should be noted that although heterocyclic systems contain aza groups comparable to the nitro group in its electron-with- drawing properties, additional activation of the substrate is often required for the reaction to proceed. One of the commonly used procedures for the activation of azines in these transformations involves quaternisation of the nitrogen atoms in the ring. Some nucleophiles can react with azine salts. However, activation by acids is not necessarily applicable because it is accompanied by deactivation of the nucleophile.Another method involving alky- lation of the azine ring is free from this drawback. The latter method was used in a large number of syntheses of fused hetero- cycles from N-alkylazinium salts. Thus, the ortho-cyclisation processes occurring in the reactions of 1,3-binucleophiles with quaternary 1,4-diazinium salts have been described.20 One-pot syntheses of imidazopyrazines, imidazopteridines, furoquinoxa- lines, pyrroloquinoxalines, imidazoquinoxalines and thiazoloqui- noxalines have been reported.21 Table 1 gives examples of the construction of fused azine-based heterocyclic systems using tandem reactions of the AN±AN type. All these systems were prepared by cyclisation of triazinium (9), N-alkylquinoxalinium (10), N-alkylpyrazinium or N-alkylpteridinium salts with such 1,3-binucleophiles as b-diketones, thioamides, dithiocarbamates, thiourea derivatives, etc.X HN N X + Y N N +Y Z Z X = H, Alk; X, Y are nucleophilic groups. N N Unlike the Sipso ±Sipso reactions in which thiourea derivatives act as N,S-binucleophiles,30 ± 33 the AN±AN reactions involve these derivatives exclusively as N,N 0-binucleophiles.34 An important characteristic feature of the AN±AN reactions is the reversible formation of cyclic products. Due to the reversi- bility of the double addition and ambident character of the nucleophiles used, these reactions can yield regioisomeric cyclisa- tion products. For example, tetrahydrothiazolo[4,5-b]quinoxaline 11 formed by the reaction of the N-alkylquinoxalinium salt 10 with thioamide 12 under kinetically controlled conditions in aprotic solvents can be subjected to different types of isomer- isation.Upon heating in ethanol, compound 11, containing the aryl group in the side chain (R2=Ph), underwent isomerisation to give thiazoloquinoxaline 13. When R2 = Me, isomerisation pro- ceeded with the involvement of the methyl group to give pyr- rolo[2,3-b]quinoxaline-2-thione 14 as the cyclisation product.21 Due to the tendency of cycloadducts to dissociation, not only regio- but also stereoisomers can be generated. An example is the construction of the tetrahydropyrrolo[2,3-b]quinoxaline ring. The reaction under the kinetically controlled conditions (activation of the nucleophile with a base) gave rise to stereoisomer 15 having the exo configuration, whereas the reaction in the absence of a base afforded both the exo isomer and the thermodynamically more stable endo isomer 16, the rearrangement 15?16 being possible.35710 Table 1.Syntheses of fused heterocycles on the basis of 1,4-diazinium and triazinium salts with the use of AN±AN reactions. Substrate N MeO2C +N MeO2C R1 R1=Me, Et. NN + R1=Me, Et. R1 10 10 (R1=Me) Binucleophilic reagent X R3 R2 NH NH X R3 R2 NH NH OH COR2 Ar N 7 OEt R2 HN 7 R3 R2 O O O 7 COR2 Me NHCOPh H2N COMe HN HN R3 R2 S R2 NH2Me HNCO2Et N Ar S Me NH2 S7 HN R2 S Product R2 H H N MeO2C N X N N MeO2C H R2, R3=H, Me, Ph, NH2, NHPh; X =O, S.R3 R1 R2 NH N X N N R2, R3=H, Me, Ph, NH2, NHPh; X =O, S. R3 R1 Ar H H N N O R2=Me, Ph, 4-MeOC6H4, 2-Tol, 4-Tol, 4-ClC6H4. COR2 N H R1 NH N OEt R2=Ph, 4-O2NC6H4. R2 COR2 NR1 HN R3 R2, R3=Me, OEt. OCOR2 NR1 HN Me R2=Me, OEt, Bui. O NR1 H H N NH NHCOPh N H COMe R1 R3 NH N R2 N R2=Ph, 2-ClC6H4; R3=Ph, 4-ClC6H4, 4-BrC6H4, 2-Py. NR1 HN S R2, R2=Me, Ph. N NR1 Ar N HN N N Me CO2Et Ar=4-O2NC6H4, 2,4-(O2N)2C6H3. R1 HN HN S NMe R2 HN N S, S N R2=Bn, cyclo-C6H11, thiazol-2-yl. Me O N Chupakhin, D G Beresnev Ref. 21 21 21 21 21 21 21 21 HN N R2 21 S NR1 21 21 HN S S 21 N N R2 MeNucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology...Table 1 (continued). Substrate 10 (R1=Me) N N +N R N Et R is morpholinyl. N +N N Me R1 + N NN R2 R1=H, Et; R2=SMe, morpholinyl. 9 Binucleophilic reagent XH H2N X=NH, O. H2N HX X=NH, O. S NH Ph Ph NH N OH R2 R2=Me, Ph, Bn, 4-BrC6H4. NH2 N NHR3 Ph NHR2 R2=H, Ph; R3=H, COPh. O Ph NH Ph NH NH R3 N HNR2 NH2 R2=H, Ph; R3=Me, Ph. S Me NMe O COMe Py N 7 HNNH2 Ph S R Me O O R=Me, OEt, Bui. O O NHAr Me Ar=Ph, 4-Tol, 2-HOC6H4. Product HN HN , N Me N NH O X Me NHN HN Me HNS Ph , N N N Ph Me O N R2 HNR3 N N N Ph HNNMe NHNMe R2 COPh N N Ph HNR2 N S N HNNMe HN N R3 HN Me O HN NMe Py NH N N O N R N COMe Et Ph HN N N S R N N NHEt Me ROC COR N Me + O N HN Me R1 ArN N N O R2 NH Me O 711 Ref.HN N 21 N NMe 21 Ph N HNN 21 Ph S NMe 21 22 21 21 21 23 21 21 24, 25712 Table 1 (continued). Substrate 9 (R1=Et; R2=SMe, morpholinyl) 9 (R1=H; R2=SMe, NH2) N N Ar N Ar=Ph, 4-O2NC6H4, 4-MeOC6H4. X N YZ N X=N, CCN; Y=N, CH, CCN, CSMe, CCF3; Z=N, CCO2Et. Note. In all reactions, the yields of the products were higher than 80%.N +R2C(S)NH2 12 10 NR +1 N R2=Ph RN1 N R2=Ph RN1 NH R2=Me RN1 R1=Me, Et. N HN 7 OR + NMe + 10 Ph The reactions of 1,4-diazinium salts with 1,4-bifunctional nucleophiles are very similar to the annelation of five-membered rings to the pyrazinium ring. The reactions of N-alkyl-1,4-dia- zinium salts with 1,4-binucleophiles afforded a series of tetrahy- dropyrazines and tetrahydroquinoxalines fused with different Binucleophilic reagent NHCOPh H2N COMe OH HO S , Ac2O R H2N R=Ph, Me. OR HO R=H, Me. HN S SH R2 N N R2 11 NR1 NH N NH R2 S S R2 13 RN1 HN S 14 H H N H N NMe OR 16 Ph HN N B H OR Me N H 15 Ph Product Et N HN N NHCOPh R2 HN COMe NH O N R2 OH HNAc N S N N Ph, N Ar Ar NH X Y OR N Z HN O heterocycles, such as pyrazine, oxazine, thiadiazine and triazine rings (see Table 1).For example, cyclisation of quinoxalinium salts with amide oximes and amidrazones afforded partially hydrogenated derivatives of new heterocyclic systems, viz., 1,2,4- oxadiazino- (17) and 1,2,4-triazolo[5,6-b]quinoxalines (18).22 OH N R1 NH2 HNEt2 N+Me N 10 HNEt2 R1=Me, Ph, Bn, 4-BrC6H4; R2=H, Ph; R3=H, COPh. The reactions with amidrazones were carried out with pre- isolated adduct 19 prepared by the reaction of the N-methylqui- noxalinium cation 10 with diethylamine to prevent the formation of tetrazines from amidrazones in a basic medium.22 Like the products obtained in the AN±AN reactions of 1,4- diazinium salts with 1,3-binucleophiles, cycloadducts prepared by the addition of 1,4-binucleophiles to 1,4-diazinium salts readily undergo dissociative destruction and isomerisation in protic solvents in the presence of catalytic amounts of acids.36 Dissociation of the cycloadducts is initiated both by mineral acids 37 and CH-acids, in particular, by acetylacetone and aceto- acetic esters.In the latter case, the reactions afforded new fused heterocyclic systems.38 First, the starting thiadiazinoquinoxaline O N Chupakhin, D G Beresnev Ref. 25 26, 27 Ac N S 28 CH2 N NH Ac 29 HN N O R1 N N Me H 17 N NEt2 19 Me N R2 N H Ph NHR3 N R3 HN N N Ph N NMe R2 18Nucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology...20 dissociated to yield the quinoxalinium cation 10 and thiobenz- hydrazide (21). The latter reacted with a b-dicarbonyl compound to give hydrazone, which underwent intramolecular cyclisation to form thiadiazole 22. The final step involved the tandem reaction AN±AN in which the compound 22 acted as a binucleophile. R Me HN O O HN N Ph S Me N 20 Me N H2N O N + + EtOH 60 ± 70 8C + O 7S Ph Me N R 21 10 Ph N N Ph N HN HN N S + S + Me Me COR Me N Me N 22 10 COR (63% ± 85%) R=OAlk.To prevent dissociation of the cycloadducts, the NH group of the tetrahydropyrazinium ring can be subjected to acylation.21 The acyl derivatives are much more stable than the starting cycloadducts and do not dissociate even in acidic solutions. Another way of stabilising hydrogenated sH adducts is based on abstraction of hydrogen atoms together with the electron pair from the reaction centres in the presence of oxidising agents 1, 12 (oxidative SHN process). In some reactions, sH adducts are pro- duced in quasistationary concentrations. Here the presence of an oxidising agent in the reaction mixture is a necessary condition for an arene to interact with a nucleophile. This type of reaction will be considered below. 1,2,4-Triazines are similar to 1,4-diazines in reactivity.In the reactions with binucleophilic reagents, these compounds are prone to form addition products at the ortho-atoms. For instance, protonated 1,2,4-triazines and their alkyl derivatives 9 readily reacted with acetoacetamides to give tetrahydropyrrolo[1,2,4]tria- zines.24, 25 Cycloadducts 23 derived from protonated 1,2,4-tria- zines are unstable and can be isolated only as salts of trifluoroacetic acid. Dissociation of these compounds can be prevented by nitrosation of the nitrogen atom adjacent to the reaction centre. N-Nitroso derivatives 24 are more stable and could be isolated as free bases.24 R1 O O NHAr Me + N N R2 N9 NO R1 Ar Ar N N NaNO2, HCl N N N N O O R1=H R2 R2 NH NH Me Me O 24 O 23 R1=H, Et; R2=SMe, morpholinyl.Another procedure for the synthesis of pyrrolotriazines involves the reaction of 1,2,4-triazine derivative 25 with ketene aminal 26. These reagents, which are widely used in the inverse- electron demand Diels ± Alder reactions,39 react with triazinium salts as C,N-binucleophiles.40 713 Et +N NHCOPh N H2N EtOH + N N COMe O 26 25 Et N HN N NHCOPh N NH COMe O (76%) The reactions of 1-alkyl-3-phenyl-1,2,4-triazinium salts 27 with nitromethane in the presence of triethylamine are extraordi- nary processes. Depending on the nature of the N-alkyl group, either one or two molecules of the triazinium salt can participate in these reactions giving rise to adduct 28 or cage compound 29, respectively.41 The reactions are accompanied by the addition of the solvent molecule (EtOH).In the latter reaction, nitromethane acts as a 1,1-binucleophile. This is a rare example of one-step syntheses of complex cage heterocycles. Me N N R=Me OEt H Ph NO2 H NH MeNO2, Et3N RN N + 28 Ph EtOH Et N Ph N27 N OEt N Ph H H N R=Et N H H Et N 29 NO2 In addition to N-alkyltriazinium salts, N-acyl derivatives can also react with binucleophiles, activation by acylation often being advantageous over activation by alkylation.28 Thus, a new approach to the synthesis of thiazolo[4,5-e] [1,2,4]triazines 30 and 31 was proposed on the basis of the tandem nucleophilic addition of alkyl- and arylthioamides to 1,2,4-triazine derivatives 32 in acetic anhydride.28 Attempts to prepare thiazolo[4,5-e] [1,2,4]triazines from 1-alkyl-1,2,4-triazinium salts failed.NAc S R2=Ph N Ph N R1 N S HN 30 Ac2O N R2 +H2N R1 NAc S N R2=Me N32 CH2 NAc R1 NH 31 R1=Ar. Aromatic ambident C-nucleophiles can participate in the AN±AN reactions with 1,2,4-triazinium salts.26 For instance, 3-methylthio- and 3-amino-1,2,4-triazines 32 reacted with resor- cinol in the presence of trifluoroacetic acid. The addition of the aromatic ring of resorcinol at the C(5) atom of the triazine followed by nucleophilic attack of the oxygen atom on the C(6) atom afforded benzofurotriazines 33. Intermediate monoaddition product 34 was isolated only with the use of a softer activating agent, viz., boron trifluoride etherate.26 N N + OH HO R N32714 CF3CO2H, CHCl3 R BF3 .Et2O, MeOH R R=SMe, NH2. In the reactions with binucleophilic resorcinol, azolo-anne- lated 1,2,4-triazines 35 are more inert and their reactions stop at the stage of formation of monoaddition products. N X N + NN N HO 35 N X N NN NH O 36 H X=CH, N. This is attributable to the fact that the adduct 36 occurs predom- inantly in the conformation unfavourable for cyclisation, the conformation being stabilised by an intramolecular hydrogen bond.27 HO X N Y Z N 37a ± f R=H, Me. X Y Z Compounds 37, 38, 39 abcdef N CH CCO2Et CCN CCN N N CH N N CSMe N N CCF3 N N N N NH O N OH NH 33 (75% ± 82%) CF3CO2H OH HO N N H HN 34 (23%) CF3CO2H OH HN O X N NN OH NH OH OR X Y N Z O HN38a ± f OR HO H OR X N Y Z HN X OH Y N Z O HN39a ± f O N Chupakhin, D G Beresnev A different situation occurs when azines with the 1,3-arrange- ment of the heteroatoms are used as the substrate.These com- pounds tend to add binucleophiles at the atoms located in the meta positions of the heteroaromatic ring with respect to one another. Actually, the reactions of azolopyrimidines 37a ± f with 1,3- binucleophilic phenols gave rise to cage structures 38a ± f and 39a ± f.29 To summarise, it should be noted that the nature of the interactions of azinium cations with binucleophiles depends on the relative arrangement of the aza groups.The presence of heteroatoms at positions 1 and 4 (pyrazine, quinoxaline or 1,2,4- triazine rings) is most favourable for the formation of fused heterocycles. The reactions of diazines containing heteroatoms at positions 1 and 2 or at positions 1 and 3 are often accompanied by side reactions.21 Examples of the construction of new hetero- cyclic systems via the addition of binucleophiles at the ortho or meta positions are few in number.29, 42 N N It should also be remembered that the AN±AN reactions are often complicated by the dissociation of primary cycloadducts formed under kinetically controlled conditions and their trans- formations into thermodynamically more stable compounds.These processes can be accompanied by changes not only in the regioorientation but also in the nature of the annelated ring. The AN±Sipso and SHN ±Sipso reactions considered in the following section lack this property. 2. Tandem reactions AN±Sipso and SHN ±Sipso N N N N Azaaromatic compounds containing heteroatoms at positions 1 and 4 can add binucleophiles to the unsubstituted carbon atoms at positions 2 and 3.21 Let us consider the reactions of azines bearing a leaving group in one of these positions. The use of these substrates substantially increases the preparative potential of the reactions with binucleophiles due to a combination of the AN, SH and Sipso processes.The transformations of this type can involve both non-activated azines and their quaternary salts. For exam- ple, the reactions of 2-chloroquinoxaline (40) with enolates led to the addition of the carbanion to the unsubstituted C(3) atom. The replacement of the chlorine atom in the resulting adduct followed by oxidation under the reaction conditions gave rise to furoqui- noxaline 41.43 7 H N But N 7 + But O N Cl Cl O N 40 7N H+, [O] But O N N But O N 41 (15%) The reactions of 2-chloroquinoxaline with aromatic nucleo- philic reagents proceed as the tandem sequence SHN ±Sipso N (appa- rently, with the involvement of hydrogen) to form benzofuro- quinoxalines 42.44 N R R 40+ O HO N 42 (60%) The presence of a leaving group possessing the electron- donating properties (for example, the methoxy group) in the azine ring leads to a decrease in the electrophilicity of the ring.The reactions with such non-activated substrates often do not take place. In this case, the azine ring can be activated with the use of aNucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology... N conventional approach, viz., by quaternisation of the nitrogen atoms of the heterocycle. Thus, 5-methoxy-1-methyl-3-phenyl- 1,2,4-triazinium tetrafluoroborate (43) is more reactive than its non-quaternised analogue and its reaction with m-phenylenedi- amine at 20 8C proceeded by the tandem process AN±Sipso to form triazacarbazole 44.45 Me N+N MeOH, 20 8C + OMe Ph NH2 H2N N 43 Me N N NH2 Ph NH NH 44 (68%) The reaction of the triazinium salt 43 with resorcinol pro- ceeded through a cascade of transformations.This reaction was complicated by dequaternisation of the triazine ring to yield 5-methoxy-3-phenyl-1,2,4-triazine (45). The reaction afforded 6-(2,4-dihydroxyphenyl)-5-methoxy-3-phenyl-1,2,4-triazine (46) as the major product generated via the nucleophilic substitution of hydrogen in the triazine ring.45 HO N N N N a OH Ph + N Ph N OMe OMe 45 (6%) 46 (25%) 43 N N b 45+46+ OH Ph N O47 (25%) (a) resorcinol, methanol, D; (b) resorcinol, DMF, 20 8C. The reaction performed in dimethylformamide afforded ben- zofurotriazine 47 as the major product generated through several reactions, viz., dequaternisation, SHN and Sipso N .It is most likely that the reaction involves dequaternisation as the first step followed by the attack of the nucleophile on the C(6) atom. In the final step, the methoxy group at position 5 is replaced and cyclisation occurs. This assumption was confirmed by a special experiment in which non-quaternised 5-methoxy-3-phenyl-1,2,4-triazine was used as the substrate (45). The latter reacted with resorcinol in the presence of boron trifluoride etherate to give benzofurotriazine 47. N OH HO N DMF, 20 8C + BF3 . Et2O OMe Ph N45 OH air NH N Ph H N OH OMe N N OH Ph N O 47 (30%) Analogously, triazinium salts react with N,N-binucleo- philes.45 The reaction of the salt 43 with thiosemicarbazide 715 N proceeded through the AN and Sipso transformations to yield imidazo[4,5-e] [1,2,4]triazine-6-thione (48).S Me +N N MeOH, NEt3, 20 8C NH2 + 1 h H2N HN OMe Ph N 43 NH2 NH2 S N Me N Me N N N N S 5 h NH2 N Ph Ph NH48 (14%) N O 49 (28%) Me The isolation of intermediate 49 has been reported.45 This is additional evidence that the unsubstituted carbon atom of the triazine ring is subjected to the initial nucleophilic attack, whereas the substitution of the nucleofuge (OMe) is a secondary process resulting in cyclisation. O MeOH, Et3N It should be noted that the reaction can also follow the `traditional' (Sipso N ) pathway. The reaction of the salt 43 with semicarbazide afforded imidazotriazine 50 along with triazine 51, which is a product of ipso-substitution of the methoxy group in 5-methoxy-3-phenyl-1,2,4-triazine 45 generated through dequa- ternisation of the substrate.45 Me N N + NH2 20 8C, 1 h NH + H2N OMe N Ph 43 NH2 O NH2 Me N N N N N O + NH 45 + (3%) N Ph N Ph NH NH 51 (30%) 50 (48%) It is noteworthy that the nucleophilic attack in all the reactions under consideration occurs primarily on the unsubstituted carbon atom, although the methoxy group would be expected to be replaced according to the classical theory of nucleophilic aromatic substitution. Moreover, the methoxy group is located at posi- tion 5 of the 1,2,4-triazine ring, which is most reactive 46 with respect to nucleophiles.However, steric effects play the key role in these reactions 1 and sH adducts are initially generated in all processes. In the salt 43, the methyl group at the N(1) atom adjacent to the reaction centre is responsible for substantial stabilisation of the compound 49, which was isolated in the crystalline state. The reactions of the triazine 45 with nucleophiles (in particular, with semicarbazide) also can afford sH adducts upon attack on the C(6) atom. However, this process is reversible due to the absence of the methyl group at the nitrogen atom adjacent to the reaction centre and ipso-substitution of the methoxy group giving rise to compound 51 prevails.N The preparation of 6-azapurine derivatives by the reactions of 5-methoxy-1,2,4-triazines, for instance, of the compound 45, with urea derivatives (Scheme 3) provides another example of the use of the AN±Sipso and SHN ±Sipso N tandem reactions in the synthesis of fused heterocycles. In these reactions, acetic and trifluoroacetic anhydrides were used for activation of the substrate.46 The reactions of N-acyl-1,2,4-triazinium salts generated in situ with urea derivatives include a series of successive steps. Depend- ing on the nature of the reagents and reaction conditions, both open-chain adducts 52 and their cyclisation products 53 ± 55 can be isolated (see Scheme 3).47 The use of acylating reagents makes it possible, first, to activate the triazine ring to nucleophilic attack and, second, to stabilise a sH adduct.Besides, the acyl substituent can serve as an auxiliary group responsible for the occurrence of the SHN reaction. For example, the final step of the reaction of the 1,2,4-triazine 45716 O Me R1 O N N N HNR2 OMe Ph N 52 HN HN R1 R2 , Ac2O O N N OMe Ph N 45 HN HN R1 R2 , (CF3CO)2O O O F3C R1 N N N O 7CF3CHO N Ph N R2 R1, R2=H, Me. with urea derivatives involves elimination of acetic or trifluoro- acetic aldehyde from the sH adduct, i.e., the process as a whole can be considered as a variation of cine-substitution reactions. N Note also that the reaction of compound 45 with dimethylurea at 20 8C stopped at the stage of formation of adduct 53, which is a cyclic Meisenheimer complex.The latter was isolated in the crystalline state. This is a rare example of a reaction proceeding according to the Sipso scheme in which an intermediate was isolated.2 To summarise, combination of the AN and SHN strategies with ipso-substitution is, undoubtedly, a powerful tool for the con- struction of various fused heterocyclic systems. The only limita- tion is that there is a need to synthesise a substrate containing a leaving group in the position required for cyclisation. However, this drawback is compensated by the high stability of the resulting products and the fact that they cannot undergo dissociation, which is often observed in the AN±AN reactions.3. Tandem reactions SHN ±SHNOnly a few examples of the double nucleophilic substitution of hydrogen are known. All these reactions involve oxidative aroma- tisation of sH adducts.48, 49 Thus, oxidation of cycloadducts 56 ± 58, which were prepared by the AN±AN reactions of quinox- alinium salts, by potassium permanganate led to their aromatisa- tion.48 OH N N R2 NH2 + 56 10 RN1 N N ON R2 RN1 59 (20% ± 78%) R1=Alk; R2=Me, Ph, Bn, 4-BrC6H4. O N Chupakhin, D G Beresnev HO Scheme 3 O Me O HN R1 KMnO4 H2N 10 N N Me2CO R1=Me N O NH Me N 57 Ph HN O N N OMe R2 53 (60% ± 64%) N NMe 7MeOH 60 (10%) O Me Bn N HN R1 Ph Ph N NH2 S HN S KMnO4 N N 10 N Me2CO R1=Me O Bn N Me N N N H Ph N 58 R2 54 (48% ± 80%) Ph N S N 7MeCHO N N Bn R1 Me N 61 (70%) N N N O N N Ph R2 55 (22% ± 87%) In this case, the SHN process is a convenient way of stabilising sH adducts.Compounds 59 ± 61, unlike their partially hydro- genated analogues, cannot undergo dissociation at all. To our knowledge, the reactions of pyrimido[4,1-c]pyrid- azines 62 with diamines in the presence of an oxidising agent are the only examples of the tandem reactions SHN ±SHN proceeding in situ without isolation of intermediate sH adducts.49 The reactions involve the addition of the nucleophile at position 4 of the heterocyclic system followed by intramolecular cyclisation to form tricyclic product 63.O O HN (CH2)n NH MeN MeN NH2(CH2)nNH2 N N Ag(C5H5N)2MnO4 N O N O NMe NMe62 63 (25% ± 80%) n=2±4. 4. Intramolecular SHN reactions New heterocycles are often generated by the intramolecular attack of a nucleophilic functional group present in the side chain of the heterocyclic substrate. The intramolecular SHN reactions were used for structural modifications of fluoroquinolone antibiotics.50, 51 For example, the reactions of 1-amino-substituted fluoroquino- lones 64 with acetylacetone afforded tricyclic compound 65.50 O Me Me F CO2Et O O N X HN 64 NH2 N O KMnO4 O Me2CO F CO2Et R2 NH NR1 N X O N 66 Me MeNucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology...O OH F F CO2Et CO2Et H O O [O] N X X N Me Me HN N 67 Me Me 65 (60% ± 65%) X=F, Cl, N NMe. The mechanism of formation of the tricyclic product 65 involves condensation of the N-aminoquinolone 64 with acetyl- acetone followed by the intramolecular nucleophilic attack on the C(2) atom of the resulting diketone 66 and oxidation of sH adduct 67 with atmospheric oxygen. The synthesis of pyrazolo[4,3-e] [1,2,4]triazines 68 using the N reactions of 5-acyl-1,2,4-triazines 69 with arylhydrazines in an acidic medium is another example of the intramolecular SH process.52 Ar N N NH2, H3O+ NH R1 N R2 69 O Ar Ar NH N N N N N N N R2 N R2 R1 R1 N 68 70 R1=Me, Et, Prn; R2=Ph, SMe; Ar=Ph, 4-Tol, 4-ClC6H4, 2,4-Cl2C6H3.This synthesis involved the successive formation of hydr- azones 70 and intramolecular substitution of hydrogen.52 Analo- gous transformations were also observed upon alkylation of quinoxaline-3-carbaldehyde hydrazones 71. These reactions led to the intramolecular nucleophilic substitution of hydrogen at the a position with respect to the quaternary nitrogen atom.53 N N N N MeI HN HN R R NMe + N71 N N N Me N + R R=H, Me, Ph, Bn, 4-O2NC6H4, 4-O2CC6H4. Yet another example of the intramolecular oxidative nucleo- philic substitution of hydrogen is the reaction of 2-aminoquin- oxaline 72 with acetylacetone or cyclohexanone.54 Schiff's bases 73 generated in the course of the reactions underwent spontaneous cyclisation to pyrrolo[2,3-b]quinoxaline derivatives 74.N R Me Me AcOH + O O R N NH2 72 COMe N R [O] Me N N R 73 717 COMe COMe N R R N Me Me N N R R NH NH 74 R=H, F. The intramolecular tele-substitution to form thienoquinoxa- lines 75 was observed upon treatment of compounds 76 derived from chloro- or iodoquinoxaline derivatives 77 with sodium trithiocarbonate.55 N Hal HC CR2 Pd(0) R1 N 77 C CR2 N Br2 N R1 Br R2 N Na2CS3 Br N R1 76 N R2 S N R1 75 (13% ± 80%) R1=H, Cl; R2=H, SiMe3, Bun, But, 4-MeOC6H4, 4-O2NC6H4, 4-ClC6H4, 4-IC6H4, 5-F3CC5H3N. Aromatisation of sH adduct 78 occurred through elimination of the bromine atom from the thiophene ring that formed.Br H N Na2CS3 Br R2 76 S:7 S N R1 S N R2 75 Br S N R1 H 78 The reaction of pyrazine 79 with pyridine provides an interest- ing example of an intramolecular SHN reaction.56 + N N Cl N N CN CN N N 80 79 Tos Tos N N N N 7HTos N N Tos NC CN 81 (66%) This reaction was not completed by the formation of product 80 by the ipso-substitution of the chlorine atom. Under the action of an excess of pyridine, the carbanionic centre was generated in the side chain of the pyrazine ring and intramolecular cyclisation took place. Cycloadduct 81 underwent aromatisation through elimination of the proton together with the tosyl substituent.718 For nitroarenes, commonly used procedures for the construc- tion of heterocycles are based on the nucleophilic substitution of hydrogen followed by cyclisation of the SHN -reaction product involving the ortho-substituent.14 This strategy, as applied to compounds of the heterocyclic series, is exemplified only by the reactions of 6-nitrotriazolo[1,5-a]pyrimidines 82 with acyl deriv- atives 83 of aromatic and heteroaromatic compounds giving rise to products of addition at the C(7) atom (84).57 NO2 N R1 N R1R1 R1 R1=H, Me; R2=Ph, 4-HalC6H4, 2-thienyl, 2-furyl.Reduction of the nitro group in the adducts 84 by tin(II) chloride did not stop at the formation of the corresponding amine 85 and afforded triazolopyrrolopyrimidine 86 as a result of oxidative aromatisation of the amine 85 followed by intramolec- ular cyclisation of product 87.IV. Construction of nitroarene-based heterocyclic systems Most of the procedures for the direct annelation of heterocycles to nitroarenes consist in intramolecular SHN processes followed by cyclisation. In spite of the fact that the ring closure is not necessarily associated with the nucleophilic attack on the unsub- stituted carbon atom, the SHN reactions of nitroarenes with binucleophiles are the key steps in these syntheses. These methods open up approaches to a wide range of biologically active indoles and quinolines, including those of natural origin. Since these methods have been surveyed in the recent review,14 we consider these questions only briefly.Procedures for annelation of five- and six-membered hetero- cycles to nitroarenes are based on oxidative and intramolecular vicarious nucleophilic substitution of hydrogen. The syntheses of nitroxyindoles 88 58 and 8914 are examples. In the former case, NO2 Me O N Et3N + Me R2 N 83 82 R2 O SnCl2 NO2 N N N N 84 H R2 R2 O O [O] NH2 NH2 N N N N R1 N N N NH87 85 R2 NH N N N N 86 NO2 Cl O ButOK, DMF, 720 8C N O N 7HCl Me 88 (69%) O N Chupakhin, D G Beresnev NO2 NO2 Me Me ButOK, DMSO, [O] O O NMe Me N 89 (70%) aromatisation of the sH adduct proceeds through the vicarious mechanism. The latter reaction involves oxidative aromatisation. The intramolecular vicarious substitution of hydrogen in nitrophenyl- (90) and nitrobenzylchloromethylsulfonamides (91) led to annelation of five- and six-membered rings, respectively, to the arene ring.59 ± 62 N N NaOH 215 8C SO2 SO2 DMSO 7SO2 Cl 92 90 NO2 NO2 N N 93 94 (44%) NO2 NO2 O2N NMe NMe NMe KOH + SO2 SO2 DMSO SO2 NO2 NO2 Cl 91 Interestingly, the initially formed cyclisation product 92 can be readily transformed into the quinoline derivative.Thermolysis of the compound 92 led to SO2 extrusion with the formation of azaxylylene 93. The subsequent intramolecular [4+2]-cycloaddi- tion afforded tetrahydroquinoline 94.59 ± 62 The intramolecular SHN reactions can be employed for the preparation of the key compounds used in the syntheses of some natural substances.For instance, the synthesis of the natural antitumour drug makaluvamine C (95) found in marine sponges involved the closure of the six-membered ring on the basis of the oxidative SHN reaction.63, 64 OMe OMe NO2 O2N NO2 O2N CO2Me ButOK, THF, CAN MeONa N HN O O O HN NO2 CO2Me O MeO 5 steps + O O H2N O2N NH Me N 95 CAN=(NH4)2Ce(NO3)6. The reactions of nitroarenes with bifunctional nucleophilic reagents can be accompanied by intramolecular cyclisation with the participation of such substituents in the aromatic substrate, as CN, NO2, etc. For example, the presence of the amino group in substituted m-nitroanilines made it possible to synthesise indole derivatives by its reactions with acetophenones 65 or phenylaceto- nitrile.66 Isocyanonitrobenzenes can also be involved in analogous transformations.Thus, the product of the SHN reaction of com- pound 96 with a-chloromethyl sulfone 97 underwent spontaneous cyclisation to give nitroindole 98.67Nucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes as an efficient methodology... Cl + Cl 97 NC O2N 96Cl SO2Ph NC O2N In some cases, cyclisation of nitroarenes proceeds with the involvement of the nitro group itself. For instance, nitrophenyl- acetonitriles 99 (Z=CN) generated by the SHN reactions can be readily alkylated. The resulting compounds 100 contain the nucleophilic centre in the side chain. In the presence of bases, these compounds undergo cyclisation involving the nitro group to yield N-hydroxyindole derivatives 101.19, 68, 69 NO2 NO2 Z Hal Y Y R NO2 B Z 100 Y Y=Cl, Br, OMe; Z=CN, CO2Bu, Tos, SO2Ph; R=Et, n-C7H15, CH=CH2, CO2Et, CONMe2.This approach can also be used in the annelation of a six- membered ring to an arene ring. For example, benzyl sulfone 102 was subjected to Michael condensation with a,b-unsaturated esters of carboxylic acids. In the presence of bases, the intermedi- ate generated in this process underwent a series of transformations involving the nitro group to form quinoline N-oxide 103.14 OMe SO2Ph+ EtO2C 102 NO2 OMe CO2EtCO2Et SO2Ph NO2 MeO CO2Et +N CO2Et 103 (70%) O7 The reactions in which the heterocycle closure takes place in situ without isolation of intermediates are of particular interest.20 For example, the reaction of nitronaphthalene 104 with sulfone 105 was not completed by the formation of the sH adduct.Under the reaction conditions, this adduct underwent aromatisation through an intramolecular redox process and cyclisation to give quinoline derivative 106. NaOH SO2Ph DMSO SO2Ph Cl O2N NH 98 (70%) Z RCH2Hal K2CO3 99HO R N Z 101 Y K2CO3, 18-crown-6, MeCN, 80 8C CO2Et OMe CO2Et 7H2O CO2Et NO2 719 NO2 DBU, Me3SiCl, DMF + Ph SO2Ph 105 104 Ph NSO2Ph 106 DBU is 1,8-diazabicyclo[5.4.0]undec-7-ene. The syntheses in which annelation of the heterocycle proceeds with the participation of the amino group generated by reduction of the nitro group of nitroarenes can be considered separately. In these syntheses, acetonitrile and acetophenone derivatives, silyl ethers of ketones, CH-active derivatives of carboxylic acids and other polyfunctional derivatives are used as nucleophiles.Both SHN -reaction products and sH adducts can be subjected to reduc- tive cyclisation. For example, the reaction of o-nitroanisole 107 with (4-chlorophenoxy)acetonitrile was accompanied by vicarious nucleophilic substitution SHN to produce cyanomethyl derivative 108. Reduction of the latter led to annelation of the pyrrole ring to the anisole ring to give methoxy-substituted indole 109.68 OMe NO2 ButOK, DMF + NC OC6H4Cl-4 7ClC6H4OH Cl 107 OMe OMe NO2 NH H2, Pd(C) CN Cl 109 (67%) 108 (64%) Examples of reductive cyclisation of sH adducts of nitro- arenes have been reported.Thus, the sH adduct formed in the course of the fluoride-induced addition of trimethylsilyloxy- ethylene 110 to nitroarenes underwent cyclisation into indolin-2- one 111 under the action of tin(II) chloride.69 NO2 OSiMe3 TASF + OMe Me 110 Me O O7 7O + HN Me N Me SnCl2, H+ CO2Me H Me Me 111 (24%) TASF=(Me2N)3S(Me3SiF2). V. Conclusion In the present review, we generalised and analysed the reactions of azines and nitroarenes containing at least one unsubstituted aromatic carbon atom with bifunctional nucleophilic reagents. These transformations were considered in the context of the construction of fused heterocycles.All the above-considered syntheses share a common trait, viz., the initial attack of a nucleophile on the unsubstituted carbon atom of azine or nitro- arene to give sH adducts, which are generated rapidly and720 reversibly regardless of the presence of nucleofugal groups in the substrate. Subsequent transformations of sH adducts depend on the nature of the substrate and nucleophilic reagent. N The formation of cyclisation products is most typical of azines containing heteroatoms at positions 1 and 2, 1 and 3 or 1 and 4. Cycloadducts prepared by theAN±AN reactions can be subjected to dissociation and isomerisation. These products can be stabi- lised by acylation, nitrosation, alkylation or further SH formations.Fused heterocycles prepared in the tandem reactions AN±Sipso and SH isomerisation. N-trans- N ±Sipso N are more stable and are not prone to To construct heterocycles from nitroarenes, it is most often necessary to introduce an additional functional group into the aromatic ring although the nitro group by itself can be of use. To summarise, the data surveyed in this review provide evidence that nucleophilic attack on the unsubstituted carbon atom of azines and nitroarenes is a powerful tool for the synthesis of fused heterocycles. References 1. O N Chupakhin, V N Charushin, H C van der Plas Nucleophilic Aromatic Substitution of Hydrogen (New York; San Diego: Academic Press, 1994) 2.F Terrier Nucleophilic Aromatic Displacement (Weinheim: Verlag Chemie, 1991) 3. V F Shvets Soros. Obrazov. Zh. 49 (1997) 4. K L Servis J. Am. Chem. Soc. 89 1508 (1967) 5. H C van der Plas Adv. Heterocycl. Chem. 74 9 (1999) 6. H C van der Plas Adv. Heterocycl. Chem. 74 87 (1999) 7. E P KuÈ ndig, V Desobry, D P Simmons, E Wenger J. Am. Chem. Soc. 111 1804 (1989) 8. G L Rusinov, P V Plekhanov, A U Ponomareva, O N Chupakhin Mendeleev Commun. 233 (1999) 9. A R Katrizky, G M Lagovsky Chemistry of Heterocyclic N-Oxides (London; New York: Academic Press, 1971) 10. V N Kozhevnikov, A M Prokhorov, D N Kozhevnikov, V L Rusinov, O N Chupakhin Izv. Akad. Nauk, Ser. Khim. 1128 (2000) a 11. T Giannopoulos, J R Ferguson, B J Wakefield, G Varvounis Tetrahedron 56 447 (2000) 12.M Makosza Izv. Akad. Nauk, Ser. Khim. 531 (1996) a 13. M Makosza, J Winiarski Acc. Chem. Res. 20 282 (1987) 14. M Makosza, K Wojciechowski Heterocycles 54 445 (2001) 15. M Makosza, M Bialecki J. Org. Chem. 57 4784 (1992) 16. V N Elokhina, O V Krylova, L I Larina, A S Nakhmanovich, M S Sorokin, K A Volkova, V A Lopyrev Khim. Geterotsikl. Soedin. 551 (2000) b 17. M Makosza Usp. Khim. 58 1298 (1989) [Russ. Chem. Rev. 58 747 (1989)] 18. Z Wrobel,M Makosza Tetrahedron 53 5501 (1997) 19. M Makosza, J Stalewski, K Wojciechowski, W Danikiewicz Tetrahedron 53 193 (1997) 20. Z Wrobel Tetrahedron 54 2607 (1998) 21. V N Charushin, O N Chupakhin,H C van der Plas Adv. Heterocycl. Chem. 43 301 (1988) 22. L M Naumova, V N Charushin, O N Chupakhin Khim.Geterotsikl. Soedin. 1118 (1987) b 23. I V Kazantseva, V G Baklykov, V N Charushin, O N Chupakhin Khim. Geterotsikl. Soedin. 420 (1986) b 24. S G Alekseev, V N Charushin, O N Chupakhin, G G Alexandrov Tetrahedron Lett. 29 1431 (1988) 25. S G Alekseev, V N Charushin, O N Chupakhin, G G Aleksandrov, S V Shorshnev, A I Chernyshev Izv. Akad. Nauk SSSR, Ser. Khim. 1637 (1989) a 26. G L Rusinov,D G Beresnev,O N Chupakhin Zh. Org. Khim. 34 450 (1998) c 27. M A Grishina, D G Beresnev, V A Potemkin, A V Belik Molodezh- naya Nauchnaya Shkola po Organicheskoi Khimii (Tez. Dokl.), Ekaterinburg, 2000 [Youth Research School on Organic Chemistry (Abstracts of Reports), Ekaterinburg, 2000] p. 313 O N Chupakhin, D G Beresnev 28.V N Charushin, N N Mochulskaya, A A Andreiko, M I Kodess, O N Chupakhin Mendeleev Commum. 28 (2002) 29. E V Bartashevich, P V Plekhanov, G L Rusinov, V A Potemkin, A V Belik,O N Chupakhin Izv. Akad. Nauk, Ser. Khim. 1573 (1999) a 30. V K Chadka J. Ind. Chem. Soc. 53 1170 (1976) 31. I M Ismail, R Iacobi, W Sauer Z. Chem. 17 15 (1977) 32. I MIsmail, W Sauer Ind. J. Chem., Sec. B 16 683 (1978) 33. V K Chadka, V K Saxema J. Ind. Chem. Soc. 57 946 (1980) 34. V G Baklykov, V N Charushin, O N Chupakhin, N N Sorokin Khim. Geterotsikl. Soedin. 960 (1985) b 35. O N Chupakhin, V N Charushin, L M Naumova, A I Rezvukhin Dokl. Akad. Nauk SSSR 261 384 (1981) d 36. V G Baklykov, V N Charushin, O N Chupakhin, V N Drozd Khim. Geterotsikl.Soedin. 557 (1987) b 37. V N Charushin, O N Chupakhin Usp. Khim. 53 1648 (1984) [Russ. Chem. Rev. 53 956 (1984)] 38. V G Baklykov, V N Charushin,O N Chupakhin,G G Aleksandrov Khim. Geterotsikl. Soedin. 697 (1990) b 39. D L Boger, S N Weinreb Hetero Diels ± Alder Methodology in Organic Synthesis (New York: Academic Press, 1987) 40. S G Alekseev, V N Charushin, O N Chupakhin,M F Gordeev, V A Dorokhov Izv. Akad. Nauk SSSR, Ser. Khim. 494 (1989) a 41. B V Rudakov, Candidate Thesis in Chemical Sciences, Urals State Technology University, Ekaterinburg, 1994 42. W P K Girke Chem. Ber. 112 1 (1979) 43. D R Carver, J S Hubbard, J F Wolfe J. Org. Chem. 47 1036 (1982) 44. R K Anderson, G W H Cheeseman J. Chem. Soc., Perkin Trans. 1 129 (1974) 45. O N Chupakhin, G L Rusinov, D G Beresnev, V N Charushin, H Neunhoeffer J. Heterocycl. Chem. 38 901 (2001) 46. V L Rusinov, O N Chupakhin Zh. Org. Khim. 34 327 (1998) c 47. D G Beresnev, G L Rusinov, O N Chupakhin, H Neunhoeffer Mendeleev Commun. 58 (2000) 48. V N Charushin, L N Naumova, M G Ponizovskii, V G Baklykov, O N Chupakhin Khim. Geterotsikl. Soedin. 1260 (1987) b 49. P D Besedin, A V Gulevskaya, A F Pozharskii Mendeleev Commun. 150 (2000) 50. O N Chupakhin, Y A Azev, S G Alexeev, S V Shorshnev, E Tsoi, V N Charushin Mendeleev Commun. 151 (1992) 51. M C Schroeder, J S Kiely J. Heterocycl. Chem. 25 1769 (1988) 52. A Rykowski, M Mojzych, Z Karczmarzyk Heterocycles 53 2175 (2000) 53. M G Ponizovsky, A M Boguslavsky,M I Kodess, V N Charushin, O N Chupakhin Mendeleev Commun. 68 (2002) 54. V N Charushin, S K Kotovskaya, N M Perova, O N Chupakhin Mendeleev Commun. 54 (2001) 55. M Armengol, J A Joule J. Chem. Soc., Perkin Trans. 1 154 (2001) 56. S V Litvinenko, Yu M Volovenko, V I Savich, F S Babichev Khim. Geterotsikl. Soedin. 107 (1992) b 57. E B Gorbunov, P V Plekhanov, in Aktual'nye Problemy Organiche- skoi Khimii (Tez. Dokl. V Molodezhnoi Shkoly-Konferentsii po Organicheskoi Khimii), Ekaterinburg, 2002 [Topical Problems of Organic Chemistry (Abstract of Reports of the Vth Youth School- Conference on Organic Chemistry), Ekaterinburg, 2002] p. 354 58. M Makosza, H Hoser Heterocycles 37 1701 (1994) 59. M Makosza, K Wojciechowski Tetrahedron Lett. 25 4791 (1984) 60. K Wojciechowski Pol. J. Chem 66 1121 (1992) 61. K Wojciechowski Pol. J. Chem, 71 1375 (1997) 62. K Wojciechowski Tetrahedron 49 7277 (1993) 63. G A Kraus, N Selvakumar Synlett 845 (1998) 64. G A Kraus, N Selvakumar J. Org. Chem. 63 9846 (1998) 65. N Moskalev, M Makosza Tetrahedron Lett. 40 5395 (1999) 66. N Moskalev, M Makosza Heterocycles 52 533 (2000) 67. K Wojciechowski, M Makosza Tetrahedron Lett. 25 4793 (1984) 68. M Makosza, W Danikievicz, K Wojciechowski Liebigs Ann. Chem. 203 (1988) 69. T V RajanBabu, B L Chenard,M A Petti J. Org. Chem. 51 1704 (1986) a�Russ. Chem. Bull., Int. Ed. (Engl. Transl.) b�Chem. Heterocycl. Compd. (Engl. Transl.) c�Russ. J. Org. Chem. (Engl. Transl.) d�Dokl. Chem. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (9) 721 ± 739 (2002) 1,3- and 1,4-Substituted tetrazolium salts S V Voitekhovich, P N Gaponik, O A Ivashkevich Contents I. Introduction II. Methods for the synthesis of tetrazolium salts III. Physicochemical properties and structures of tetrazolium salts IV. Reactions of tetrazolium salts V. Applications of tetrazolium salts VI. Conclusion Abstract. physicochemical synthesis, the on data published The The published data on the synthesis, physicochemical properties, and 1,3-(1,3,5)- of reactions and structures properties, structures and reactions of 1,3-(1,3,5)- and 1,4-(1,4,5)- 1,4-(1,4,5)- substituted and systematised are salts tetrazolium substituted tetrazolium salts are systematised and generalised. generalised. Their the in compounds starting as applications Their applications as starting compounds in the preparative preparative chemistry branches other some and derivatives heterocyclic of chemistry of heterocyclic derivatives and some other branches of of science bibliography The reviewed.are technology and science and technology are reviewed. The bibliography includes includes 122 references 122 references. I. Introduction Compounds containing positively charged tetrazole rings have been known since the end of the 19th century, since the first description of 2,3-substituted tetrazolium salts 1. Later, these compounds were given considerable attention in connection with their reduction into brightly coloured formazane under the action of various reagents.At present, practical applications of 2,3-sub- stituted tetrazolium salts are based on chemical, biochemical, electrochemical and radiochemical reduction. These compounds proved to be especially useful in biochemical, microbiological, histochemical and cytochemical studies, analytical chemistry and clinical analysis as well as components of ionising radiation monitors and photosensitive materials, germination indicators of grain crops and phase-transfer catalysts.1±5 Other types of tet- razolium salts which differ in the positions of substituents at the nitrogen atoms have not received much attention until recently. The situation has changed in the past decade owing to the appearance of a great number of publications devoted to 1,3- (2) and 1,4-substituted (3) tetrazolium salts.R1 R1 R1 + R2N N X7 R2N X7 N X7 + + N R2N NR3 1 NR3 N N 3 N NR3 2 S V Voitekhovich, P N Gaponik, O A Ivashkevich Research Institute of Physical and Chemical Problems, Belarus State University, Leningradskaya ul. 14, 220050 Minsk, Belarus. Fax (37-517) 226 46 96. Tel. (37-517) 209 51 98. E-mail: voitekhovich@bsu.by; azole@tut.by (S V Voitekhovich). Tel. (37-517) 209 51 91. E-mail: gaponic@bsu.by (P N Gaponik). Tel. (37-517) 209 52 54. E-mail: fhp@bsu.by (O A Ivashkevich) Received 5 June 2002 Uspekhi Khimii 71 (9) 819 ± 839 (2002); translated by R L Birnova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n09ABEH000739 721 721 729 731 737 738 In many respects, this is due to the development of synthetic methods including selective procedures, which make possible direct synthesis of tetrazolium salts 2 and 3 with predetermined compositions and structures which can be use in organic chemistry and some other branches of science and technology.However, systematised data on the synthesis, properties and applications of these compounds are still absent in the literature. Previously published reviews 1± 5 deal with the description of mostly 2,3- substituted tetrazolium salts. The reviews devoted to the chem- istry of tetrazoles 6±8 and some of its particular aspects 9 ±11 published in the past decade provide separate and, some- times,8, 10, 11 very scarce data on 1,3- and 1,4-substituted tetrazo- lium salts. The aim of the present review is to generalise the information on 1,3- and 1,4-substituted tetrazolium salts over the past 10 ± 15 years in order to highlight current trends in their synthesis with special emphasis on the availability of tetrazolium salts and their efficacy in organic synthesis.Earlier publications that have not been cited in other reviews and monographs are also surveyed. Depending on the nature of the substituent at C(5), tetrazo- lium salts 2 and 3 can be divided into two groups, viz., 1,3- and 1,4- disubstituted tetrazolium salts containing a labile hydrogen atom at the carbon atom of the heterocycle and 1,3,5- and 1,4,5- trisubstituted tetrazolium salts where R1= H. Yet another class of 1,3,5-trisubstituted tetrazolium salts, viz., compound 2 with the substituent R1 represented by an anionic group [O7, NH7, 7C(CN)2, etc.], are referred to as mesoionic tetrazoles.12 As the behaviour of these internal salts differs drastically from that of ordinary tetrazolium salts, they are not considered in the present review.II. Methods for the synthesis of tetrazolium salts All currently known approaches to the synthesis of 1,3- and 1,4- substituted tetrazolium salts can be classified into two main groups: (1) transformations of tetrazoles and (2) heterocyclisation of polynitrogenous linear structures. The first group includes a broad range of processes, in the first place, direct exhaustive alkylation of tetrazoles at the endocyclic nitrogen atoms, i.e., quaternisation. Reactions involving exocyclic atoms of mesoionic tetrazoles and 1,4-dihydrotetrazoles and reactions affecting both endo- and exocyclic atoms of functional- ised tetrazoles leading to the formation of annelated tetrazolium salts are also related to this group.Conversions of some tetrazo- lium salts into other ones do not have as a rule any preparative722 significance and are considered in the section devoted to the chemical properties of tetrazolium salts. The second group includes few heterocyclisation reactions the main advantage of which is that they exclude the use of tetrazoles in the synthesis of tetrazolium salts. Quaternisation of tetrazoles is the most popular and simple procedure for the synthesis of 1,3- and 1,4-substituted tetrazolium salts.This method is based on two approaches which differ in the nature of the alkylating reagents and the reaction media. The practical significance of quaternisation reactions has increased considerably in the past decade owing to the ready availavility of diverse tetrazole derivatives.6±10 1. Quaternisation of tetrazoles in organic media In the early 1950's, it was shown for the first time that N-sub- stituted tetrazoles, particularly 1-aryl-5-methyltetrazoles, react with methyl iodide to yield quaternary salts.1 Later, it was established that quaternisation is largely a general process which involves all isomericN-substituted tetrazoles 13 ± 37 including fused ones;16, 20, 24, 34 the role of alkylating reagents is played by alkyl halides, alkyl sulfates, oxonium salts and related compounds of which methylating and ethylating reagents 13 ± 37 as well as a-hal- ogeno carbonyl compounds (bromoacetone 18, 35, 37 and phenacyl bromides 21, 27, 35, 37) are studied most extensively.The mecha- nisms of quaternisation of N-substituted tetrazoles still remain obscure. It can be assumed, however, that similar to other azoles, quaternisation of tetrazoles occurs by an SN2 mechanism. The azole molecule, which plays the role of a nucleophilic reagent with respect to alkyl halide, passes into a quaternised form. This reaction is accompanied by the Walden inversion.38, 39 Quaternisation of tetrazoles for the synthesis of tetrazolium salts poses certain problems.In the first place, this is due to the non-selectivity of this reaction due to the ambident character of the tetrazolium ring 3, 6, 7 and the necessity to separate the mixtures of isomeric tetrazolium salts formed. Thus quaternization of 1-mono- and 1,5-disubstituted tetrazoles 4 yields a mixture of isomeric tetrazolium salts 2 and 3.13, 15 ± 25, 27, 28, 31, 32 ± 37 R3 N N + X7 R1 N N N N R3X 3 R2 R1 N N + R3 N N 4 R2 R1 X7 N N 2 R2 In the case of 1-monosubstituted tetrazoles 4 (R1=H), the predominant reaction is the attack at the N(4) atom 13, 15, 18, 21, 28, 33 manifesting the highest basicity (proton affinity) irrespective of the nature of the R2 substituent.6, 8, 40 The presence of a substitutent at C(5) increases the content of the 1,3- substituted tetrazolium salt 2 which becomes predominant in the presence of bulky substituents R1.Individual salts can be isolated from the mixtures owing to differences in the solubilities of the isomers. In most cases, 1,4,5-isomers are soluble in water and lower alcohols less readily than 1,3,5-isomers. Sometimes, indi- vidual salts 3 precipitate immediately.14, 22, 25, 32 This approach is largely used in the synthesis of 1,4-substituted tetrazolium salts (Table 1). The use of 2-mono- and 2,5-disubstituted tetrazoles for the synthesis of 1,3-substituted tetrazolium salts seems to be more appropriate, since the quaternisation of the latter proceeds selectively at the N(4) atom (Table 2).However, by reason of the lower basicity of this type of tetrazole,41 their quaternisation requires more drastic conditions (elevated temperatures and S V Voitekhovich, P N Gaponik, O A Ivashkevich sometimes high pressure) than that for tetrazoles 4.3, 17, 18, 21, 23, 27 ± 29, 35 ± 37 The use of tetrazole and its 5-monosubstituted derivatives 5 as starting compounds in the synthesis of tetrazolium salts is rather attractive, since the methods for the preparation of compounds 5 have been well developed.42 Nevertheless, exhaustive alkylation of this type of tetrazole has been studied for a limited number of compounds only. Thus the methylation of 5-methyltetrazole with an excess of methyl iodide (40 8C, 10 days) yields a mixture of 1,4,5- and 1,3,5-trimethyltetrazolium iodides in a 68 : 32 ratio.14 Later, it was found that tetrazoles with substituents in position 5 react with an excess of dimethyl sulfate upon heating to give 1,3- (6, X=MeSO¡4 ) and 1,4-dimethyl-5-R-tetrazolium methyl sul- fates (7, X=MeSO¡4 ) (Table 3).32, 33 The ratio of the isomers formed depends on the nature of the substituent at the C(5) atom; owing to steric factors, the proportion of the 1,3,5-isomer 6 increases in the following order: R=Me, Et, Ph, But (19%, 31%, 55% and 91%, respectively).32 The isolation of the main isomer is achieved by precipitation of the corresponding perchlo- rates 32 and picrates 33 (see Table 3).To conclude, special mention should be made of a study 26 which represents an approach to the mathematical estimation of the feasibility of quaternisation and alkylation of some azoles including tetrazoles.The authors attempted to correlate the sum of normalised potentials with pKa values of azoles and demon- strated the applicability of these principles in several reactions. 2. Quaternisation of tetrazoles in acidic media The methods for the synthesis of tetrazolium salts based on quaternisation of tetrazoles with alcohols in acidic media were proposed in the past decade 22, 30, 33, 36, 43 ± 49 and are being inten- sively developed at present. The studies into the alkylation of 5-R- tetrazoles in strong acids which, contrary to the reactions occur- ring in neutral media, proceed selectively at the N(2) atom, formed the basis for these methods.50 ± 52 Kinetic studies and an analysis of reaction mechanisms revealed that the selectivity is due to block- ing of the most nucleophilic N(4) atom because of its protona- tion.53, 54 Later, it was found that the regioselectivity of the tetrazole alkylation in acidic media remains in the quaternisation.This allowed development of efficient procedures for the synthesis of tetrazolium salts and enlargement of their range by virtue of quaternising reagents that made it possible to introduce branched substituents onto the `pyridine' nitrogen atom. It is now estab- lished that quaternisation of tetrazoles can be effected by alcohols that generate stable carbocations (e.g., tert-butyl 17, 22, 33, 36, 45 ± 49 and diacetone alcohols,46, 47 2,5-dimethylhexane-2,5-diol,36 iso- propyl alcohol,49 1-phenylethanol,49 adamantan-1-ol 43 and a-fer- rocenyl-substituted alcohols) in acidic media.44 These reactions are usually carried out in strong mineral acids (e.g., perchloric, sulfuric, tetrafluoroboric) 33, 36, 45 ± 49 or biphasic systems consist- ing of an organic solvent (ether, dichloromethane) and an acid.17, 22, 44 The direction of quaternisation cannot always be predicted unequivocally, since it depends on many factors, such as the nature of the substrate and the alkylating reagent, medium acidity and reaction time.1,3-Disubstituted tetrazolium salts 8 and 9 were obtained in high yields by quaternisation of 1-monosubstituted tetrazoles with tert-butyl 36, 45, 47 and isopropyl 49 alcohols in 72% perchloric acid at 20 8C and contained no admixtures of the corresponding 1,4- isomers.ButOH N N But +N N ClO¡4HClO4, 1 h N N N N R R 8 (63% ± 95%) R=Me, Et, But, Bn, cyclo-C6H11.1,3- and 1,4-Substituted tetrazolium salts Table 1. The synthesis of 1,4-substituted tetrazolium salts by quaternisation of 1-mono- and 1,5-disubstituted tetrazoles. N R1 N 1) R3X 2) Y7 R2N N R1 Me Me Et Pri Pri Me CH2But CH2But Bn Pri CH2But Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH Me2NCH=CH MeS MeS HMe Me Me Me Me Me NH2 NH2 CH2COMe Me Me Me Me Me But Me Me Me Ph Ph Ph Ph 4-MeC6H4 4-PhC6H4 4-MeOC6H4 4-FC6H4 4-FC6H4 4-ClC6H4 4-ClC6H4 4-BrC6H4 4-BrC6H4 4-IC6H4 4-IC6H4 Me But Me Me Me Prn Pri Bui cyclo-C6H11 CH2COMe Me Me H CH2 COMe MeI MeI BrCH2COMe Me2SO4 Me2SO4 Me Me Ph C6H4NO2-3 N a With decomposition; b M.p.of the corresponding semihydrate; c the reaction was carried out in the presence of AgBF4; d the reaction was carried out under high pressure. R1 N R3 + N Y7 R2N N R3X R2 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 MeI Me2SO4 Me2SO4 Me2SO4 Me2SO4 Et2SO4 Me2SO4 Et2SO4 Me2SO4 Et2SO4 Me2SO4 Et2SO4 Me2SO4 Me3OBF4 BrCH2COPh BrCH2COC6H4Br-4 BrCH2COC6H4NO2-4 BrCH2COPh BrCH2COPh BrCH2COPh BrCH2COPh Me2SO4 BrCH2COPh BrCH2COMe CH2COMe Me Me Me Me2SO4 Me N+N N Me2SO4 Me N+N N N Et3OBF4 NEt +N N N Quaternisation conditions Y Yield (%) t /h T /8C 216 1111111111424 44448484848 50 59 51 41 28 21 41 22 44 30 44 58 60 69 85 55 61 22 65 28 70 32 84 43 69 95 90 70 35 79 84 79 75 14 88 33 35 43 65 78 83 IClO4 ClO4 ClO4 PF6 BF4 ClO4 BF4 BPh4 PF6 PF6 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 BF4 BF4 Br Br Br Br Br Br Br Br Br BF4 IIBr ClO4 ClO4 20 ± 25 120 120 120 120 120 120 120 120 120 120 80 100 80 80 80 80 90 80 90 80 90 80 90 70 ± 80 40 80 70 ± 75 70 ± 75 70 ± 75 70 ± 75 70 ± 75 70 ± 75 20 ± 25 90 50 ± 55 50 50 50 35 35 0.5 12 96 96 120 72 48 24 548 8288.c 5.5 d 5.5 d 6.d 15 15 75 82 1 BF4 80 82 1 BF4 20 20 65 BF4 723 Ref.M.p. /8C 18 22 22 22 22 22 22 22 22 22 22 25 25 25 25 25 25 25 25 25 25 25 25 25 17 17 21 27 27 27 27 27 27 35 35 37 18 18 18 23 23 266 205 ± 206 146 ± 147 155 ± 157 177 ± 178 112 ± 114 185 ± 186 166 ± 168 207 ± 210 a 144 ± 145 237 a 160 ± 161 160 ± 161 154 ± 155 166 ± 167 153 ± 154 168 ± 169 116 ± 117 193 ± 194 126 ± 127 208 ± 209 146 ± 147 209 ± 210 157 ± 158 141 ± 143 121 ± 123 145 ± 148 170 ± 171 130 ± 135 156 ± 157 157 ± 158 155 182 ± 184 240 a 220 a, b 124 ± 127 130 162 ± 163 157 ± 158 159 ± 160 135 ± 137 24 131 ± 135 24 267 ± 269 24 185 ± 187724 Table 2.The synthesis of 1,3-substituted tetrazolium salts by quaternisation of 2-mono- and 2,5-disubstituted tetrazoles. N N R1 R1 + 1) R3X 2) Y7 NR2 NR2 Y7 R3N N N N R3X R2 R1 Me Me Me Me Me Ph Ph Me Me Me Me Me Me H CH2 COPh CH2COMe CH2CO2Me CH2CO2Me CH2COMe CH2CO2Me CH2COMe CH2COMe NH2 NH2 HHHMeH CH2 COC6H4Br-4 H CH2 COC6H4NO2-4 Me Ph C6H4Me-4 C6H4OMe-4 C6H4Cl-4 Ph Ph Ph Ph CH2COPh Ph Ph Ph Ph C6H4NO2-4 C6H4Me-4 C6H4OMe-4 C6H4Cl-4 H CH2 COMe Me C6H4OMe-4 C6H4Me-4 C6H4Me-3 Ph C6H4I-4 C6H4Br-3 C6H4NO2-3 C6H4NO2-4 CH2COMe Me Me Me Me Me Me Me Me BF4 BF4 BF4 BF4 BF4 BF4 BF4 Br Br Br Br Br Br Br Br Br Br ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 FSO3 IClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 ClO4 a The reaction was carried out in the presence of AgBF4; b the reaction was carried out under high pressure.BrCH2COPh BrCH2COPh BrCH2COC6H4Br-4 BrCH2COMe BrCH2COMe BrCH2COPh BrCH2COMe BrCH2COMe BrCH2COPh BrCH2COPh BrCH2COC6H4Br-4 BrCH2COC6H4NO2-4 BrCH2COPh Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 MeSO3F MeI Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Me2SO4 Table 3.The synthesis of 1,3- and 1,4-dimethyl-5-R-tetrazolium salts 6 and 7 by quaternisation of 5-R-tetrazoles 4. Me X7 1) Me2SO4 or MeI Me + N N N 2) HX N+ N N R R +R X7 N N N5 Me N Me N NH 7 6 Quaternisation conditions X R Salt Alkylating reagent T /8C 40 90 40 90 90 90 IClO4 IClO4 2,4,6-(NO2)3C6H2O ClO4 Me But Me Me Me Et 667777 MeI Me2SO4 MeI Me2SO4 Me2SO4 Me2SO4 a With decomposition.S V Voitekhovich, P N Gaponik, O A Ivashkevich Y Quaternisation conditions t /h T /8C 60 60 60 60 60 60 60 60 90 60 60 60 60 20 20 20 20 80 80 80 80 80 80 80 80 20 50 60 60 60 60 60 60 60 60 14 a 10 a 10 a 10 a 7 a 14 a 14 a 52.9 2 a 2 a 2 a 22222888888880.17 0.23 b 1.25 1.25 1.25 1.25 1.25 1.25 1.25 1.25 Yield (%) t /h 26 72 51 69 35 60 240 1240 121 Ref. M.p. /8C Yield (%) 37 37 37 37 37 37 37 35 35 21 21 21 21 21 21 21 21 29 29 29 29 29 29 29 29 18 18 23 23 23 23 23 23 23 23 159 ± 161 123 ± 124 107 ± 109 oil 145 210 ± 211 145 140 ± 170 230 138 ± 140 171 ± 172 153 ± 154 142 ± 146 121 ± 122 135 ± 137 132 ± 134 115 ± 117 209 ± 210 224 ± 226 228 ± 230 255 ± 257 272 ± 273 243 ± 245 230 ± 232 265 ± 267 84 ± 86 144 144 ± 145 142 ± 144 103 ± 104 129 ± 131 140 ± 142 165 138 ± 139 204 57 81 82 57 60 80 60 45 60 69 55 62 57 91 78 82 96 85 86 87 90 87 82 88 91 92 46 68 73 62 78 77 69 87 87 Ref.M.p. /8C 14 32 14 32 33 32 266 a 149 ± 150 130 a 204 ± 205 164 ± 166 144 ± 146725 1,3- and 1,4-Substituted tetrazolium salts But PriOH N N Pri +N N But + ClO¡¦4N N N ButOH HClO4, 24 O�� N N N + N N N N ClO¡¦ ClO47 + 4 72% HClO4, 1 h N N N N N N Ph Ph 9 (37%) 12 11 C6H4X C6H4X C6H4X X X Ratio of isomers 12 : 11 Ratio of isomers 12 : 11 100 : 0 100 : 0 100 : 0 2-CO2H 4-OMe N NN N The regioselectivity of quaternisation is achieved by virtue of complete protonation of 1-R-tetrazole at the most nucleophilic nitrogen atom N(4) with the formation of the 1-R-4H-tetrazolium cation 10 in which the N(2) and N(3) atoms are susceptible to electrophilic attack.The attack at the N(2) atom is sterically hindered due to the presence of a substituent at the N(1) atom. Therefore, the carbocation generated from the alcohol, attacks the tetrazolium cation at the N(3) atom.45 90 : 10 87 : 13 86 : 14 82 : 18 60 : 40 50 : 50 3-Me 3-OMe H3-Br 4-COMe 3-NO2 4-OC6H4 H + H+ [R1]+ R1 O R1OH 7H2O H + + Under certain conditions, individual 1,4-disubstituted tetra- zolium salts 13 and 14 can also be prepared by quaternisation of 1-R-tetrazoles with a-ferrocenyl-substituted 44 or diacetone alco- hols.46, 47 R1 R1 N HN HN N N N N N [R1]+ + + + CH(Fc)R2 R1 R1 FcCH(R2)OH HClO4 N N N N N N N N 7H+ N N N N X7 HX, CH2Cl2 N N ClO¡¦ ClO¡¦ R2 R2 R2 4 R2 4 4 N N 13 (48% ¡À 94%) ClO¡¦ 10 R1=Bn, Ph, 4-MeOC6H4, 4-BrC6H4, 4-NO2C6H4; R2=H, Me, Ph; Fc=C5H4FeC5H5; X=ClO4, BF4.+ CMe2CH2COMe R R Me2C(OH)CH2COMe N N N N ClO¡¦ HClO4 4 The assumption of a reaction of the azolium cation with the carbocation is unusual at first glance; however, the results of quantum-chemical calculations 45, 53 suggest that such a reaction is possible, since the N(2) and N(3) atoms of the 1-R-4H-tetrazolium cation 10 retain sufficiently high p-electron density.N N N N 14 (63% ¡À 94%) R=Me, Bn, cyclo-C6H11, Ph, 4-MeOC6H4, 3-NO2C6H4, 2-HO2CC6H4, 2-HO-4(5)-NO2C6H3. Because of steric hindrances, the aforementioned isomerisa- tion of tetrazolium salts is not realised in the quaternisation of 1,5- disubstituted tetrazoles, which allows selective preparation of 1,3,5-trisubstituted tetrazolium salts 15 (Table 4). 1,3-Disubstituted tetrazolium salts 16 (R2=H) are formed in Selective preparation of 1,3-disubstituted tetrazolium salts demands special conditions to be met, which ensure complete protonation of the starting tetrazoles 45 and exclude the possibility of isomerisation of 1,3-substituted tetrazolium salts into the corresponding 1,4-isomers in the course of the reaction.45, 55 Moreover, quaternisation occurs smoothly only in the case of 1-alkyltetrazoles.45 1,4-Disubstituted tetrazolium salts 11 are formed as reaction products in the quaternisation of some 1-aryltetrazoles with tert-butyl alcohol.Their proportion increases in the presence of electron-withdrawing groups in the aryl substituent,36, 45, 47, 49 which is explained 49 by lower basicity of 1-aryltetrazoles in comparison with their alkyl analogues.41 high yields upon quaternisation of 2-monosubstituted tetrazoles with alcohols in the presence of acids (Table 5).The yields of 1,3,5-trisubstituted derivatives in the reaction of 2,5-disubstituted Table 4. The synthesis of 1,3,5-trisubstituted tetrazolium salts 15 by quaternisation of 1,5-disubstituted tetrazoles with alcohols in acidic media. N N N R3OH N + R3 X7 R2 R2 HX N N NR1 15 NR1 Ref. Yield (%) Quaternisation conditions R3 R2 R1 M.p. /8C t /h HX 17 17 22 22 30 30 30 46 131 ¡À 133 154 ¡À 155 129 ¡À 131 149 ¡À 150 b 151 ¡À 152 140 ¡À 141 50 b 115 ¡À 117 67 64 48 52 40 46 40 75 15 15 15 15 48 48 48 12 Me But Me Me Me Me CH2=CH Me But But But But But But But CMe2CH2C(O)Me 54% HBF4 a 54% HBF4 a 54% HBF4 a 54% HBF4 a 48% HBF4 40% HClO4 48% HBF4 72% HClO4 MeS MeS Pri Me Me Me Me Me a Solution in diethyl ether; b with decomposition.726 Table 5.The synthesis of 1,3- and 1,3,5-substituted tetrazolium salts 16 by quaternisation of 2-mono- and 2,5-disubstituted tetrazoles with alcohols in acidic media. N N N R1 N R2 R1 HHHHH ButH ButH Bu H Bu HHH Bu HHHHCH2=CH CH2=CH Me a With decomposition; b Ad is 1-adamantyl; c isolated as pechlorate. Me Me Et Et Pri But CH2=CHCH2 CH2=CHCH2 CH2=CMe CH2BrCHMe MeC(O)CH2CH2 Me Et Pri Ad Me But Et tetrazoles with tert-butyl alcohol in the presence of HBF4 (see Ref. 30) do not exceed 50% (see Table 5), apparently due to steric hindrances preventing the introduction of the substituent into position N(4).The synthesis of the salt 17 shows that quaternisation can successfully involve diols, which opens up new opportunities for the selective synthesis of binuclear tetrazolium salts.36 N N + N N Pri Pri 1,3-Di-tert-butyl-5-R-tetrazolium salts 18a ± d were prepared by exhaustive alkylation of tetrazole (5a) and 5-methyltetrazole (5b) with tert-butyl alcohol in the presence of strong acids.33, 45 The selectivity of alkylation is explained by the intermediate formation of 2-tert-butyl-5-R-tetrazoles 19a,b in which only the NH N ButOH R N N HClO4 or H2SO4 5a,b Compound 18 abcd R3 N + R3OH N X7 N R1 HX N 16 R2R3 R2 But But But Buttt But Butt CMe2CH2C(O)Me CMe2CH2C(O)Me CMe2CH2C(O)Me Ad b But But But HClO4 Me2CCH2CH2CMe2 OH OH Me2CCH2CH2CMe2 N N 2 ClO¡ N N 4 + +N N N N Pri 17 (44%) But N N N N ButOH R R + HX N N N N But But 18a ± d 19a,b Yield (%) Ref.X R 45 45 33 33 80 59 62 10 ClO4 ClO4 2,4,6-(NO2)3C6H2 2,4,6-(NO2)3C6H2 HMe HMe Quaternisation conditions HX 72% HClO4 48% HBF4 72% HClO4 48% HBF4 72% HClO4 72% HClO4 72% HClO4 48% HBF4 72% HClO4 72% HClO4 72% HClO4 72% HClO4 72% HClO4 72% HClO4 74% H2SO4 48% HBF4 48% HBF4 48% HBF4 HN Ph N5c N(4) atom is accessible for quaternisation. In the case of 5-phenyl- tetrazole (5c), the expected tetrazolium salts are not formed, presumably due to steric hindrances. In this case, 2-tert-butyl-5- phenyltetrazole (19c) 33 or its protonated form, viz., 3-tert-butyl-5- phenyl-1H-tetrazolium perchlorate (20), isolated as the alky- lation product.45 It may be expected that the methods for the synthesis of tetrazolium salts based on quaternisation of tetrazoles in the presence of strong acids will be developed further.Their attrac- tiveness is due both to the availability of the starting compounds and the selectivity and simplicity of the reactions. The use of novel functionalised reagents in the quaternisation reaction will enable one to extend the range of available tetrazolium salts. X7 3. Miscellaneous methods based on the transformations of tetrazole derivatives Disubstituted tetrazoles containing reactive groups at the C(5) atom react with bifunctional reagents to give annelated tetrazo- lium salts.It was shown, in particular,56 that the condensation of 1-substituted 5-aminotetrazoles 21 with b-diketones, b-chloro- vinyl ketones and aldehydes as well as with malonaldehyde acetals yields 1-substituted tetrazolo[1,5-a]pyrimidinium salts 22a,b. This reaction proceeds at 20 8C or upon short-term heating of solu- tions of the reactants in acids (e.g., acetic, trifluoroacetic, perchloric) or in ethanol in the presence of perchloric acid. S V Voitekhovich, P N Gaponik, O A Ivashkevich Yield (%) t /days 23232323222222 c 95 56 80 72 96 74 89 62 88 75 70 94 63 70 9 21 3 4550 33 33 N H2SO4 Ph N N ButOH 19c N NH HClO4 Ph N20 Ref. M.p./8C a 48 30 48 30 48 48 48 30 48 48 48 46 46 46 43 30 30 30 150 a 155 ± 156 133 ± 135 136 ± 137 130 ± 131 120 ± 122 143 ± 145 105 ± 107 145 a 78 ± 80 97 ± 99 133 ± 135 79 ± 81 69 ± 71 240 ± 241 50138 ± 140 a 130 ± 132 NN But N ClO¡4+ N But1,3- and 1,4-Substituted tetrazolium salts R4 R4 + N R3 N HClO4 N N R3 N + ClO¡ N 4 X N N R2 H2N N R2 O R1 21 R1 22a (26% ± 91%) N N HClO4 N +(EtO)2CHCH2CH(OEt)2 N H2N Me N+ N ClO¡ N 4 N N Me 22b (69%) R1=Me, CH=CH2, CH2CH=CH2, Bn, Ph, HO2CCH2; R2=H, Me, Ph; R3=H, Me; R4=H, Me, Ph; R3±R4=(CH2)4; X=Cl, OH.2-Substituted 5-aminotetrazoles 23 generate 2-substituted tetrazolo[1,5-a]pyrimidinium salts 24 under analogous condi- tions.56 R4 R4 ClO¡4+ N R3 R3 N HClO4 X N N NR1 + NR1 N N R2 H2N R2 O N 24 (22% ± 98%) 23 R1=Me, CH2CH=CH2, Bn; R2=H, Me, Ph, CO2H; R3=H, Me; R4=H, Me, Ph; R3±R4=(CH2)4; X =Cl, OH. 1-Aryl-5-mercaptotetrazoles 25a,b react with a- and b-halo- geno ketones with heating to give tetrazolylthioalkyl ketones which undergo cyclisation under the action of sulfuric acid in the presence of HClO4 to give thiazolo[3,2-d ]tetrazolium (26) and tetrazolo[5,4-b][1,3]thiazinium perchlorates 27, respectively.7, 57 C6H4R1-4 N N THF, D SH + Hal(CH2)nC(O)R2 7HHal N N Ph 25a,b N N+ S N R1=H n=1 N ClO¡4C6H4R1-4 H2SO4 N N S (CH2)nC(O)R2 R2 26 (31% ± 87%) C6H4R1-4 HClO4 N N (12% ± 78%) N N S + n=2 N N ClO¡4R2 27 (29% ± 51%) R1=H(25a), Cl (25b); 26: R1=H: R2=But, Et, Ad, Ph, 4-MeOC6H4, 4-NO2C6H4, 4-C6H5C6H4, 4-ClC6H4, 4-BrC6H4; 27: R1=H, Cl: R2=Ph, 4-ClC6H4, 4-BrC6H4, 4-FC6H4.Treatment of 5-mercaptotetrazoles 25a and 28 substituted by epoxy bromides in position N(1) with heating affords 6-hydroxy- 1-R-6,7-dihydro-5H-tetrazolo[5,4-b][1,3]thiazinium bromides 29 in relatively low yields.57 727 R1 Br7 N N R1 R4 S + N N CHBr R2 N N MeCOEt, D SH + R3 N R5 R3 O R2 N 25a, 28 R4 R5 OH 29 (8% ± 46%) R1=Ph, cyclo-C6H11; R2, R3, R4, R5=H, Me.The reaction of 1-phenyltetrazol-5-ylsulfenyl chloride with 3,3-dimethylbut-1-ene in nitromethane in the presence of lithium perchlorate yields 7,8-dihydro-6,6,7-trimethyl-2-phenyl-6H-tetra- zolo[4,5-b][1,3]thiazinium perchlorate (30). It has been shown that the heterocyclic compound 30 is formed in the AdE reaction of an alkene with sulfenyl chloride due to the tandem rearrangement ± cyclisation occurring by a p-route.58 The reaction of this sulfenyl chloride with styrene or 1-phenylpropene yields 2,3-dihydrothi- azolo[3,2-d ]tetrazolium salts 31a,b.59 Ph N N ButCH=CH2 S + N ClO¡4 N Ph Me N N LiClO4 Me Me SCl MeNO2 N 30 (31%) N Ph PhCH=CHR N N ClO¡4+ S N N R Ph 31a,b R=H(a, 47%), Me (b, 52%).1,3,5- and 1,4,5-Trisubstituted tetrazolium salts can be pre- pared in more than 70% yields from mesoionic tetrazoles 32 12, 60, 61 and 1,4-dihydrotetrazoles 33,17, 62 in which the exocyclic atoms are readily involved in alkylation and protonation. EtO C6H4R-4 N X =O; R=H, Me, Cl BF¡ Et3OBF4 N 4 N +N X7 C6H4R-4 C6H4R-4 N H(Ph)HN N N +N Ph N X=NH, NPh; R=H C6H4R-4 BF¡4 N HBF4 N +N 32 Ph X=O, NH, NPh. NHMe X =NMe; R=Ph + Me X N N HBF4 PhBF¡4N N Me R N NN N SMe 33 + X=S; R =Me Me N N Me2SO4 MeMeSO¡4N N X=NMe, S. Analogous reactions with fused tetrazole derivatives, with pyrrolotetrazoles 34 and 35 in particular, also proceed smoothly.63728 Ph N N N N 34 Me N N + Me N 7N 35 Special mention should be made of the reaction of readily available 1,3-diphenyltetrazolium-5-olate 12 (36) with phosphorus pentachloride 61 and trifluoromethanesulfonic anhydride 64 which yields functionalised tetrazolium salts 37 and 38.Ph (CF3SO2)2O O7 Ph NN N +N 36 Ph PCl5, 90 ¡À 100 8C Table 6. The synthesis of 1,4,5-trisubstituted tetrazolium salts by the reaction of azides with nitrilium salts. R2 + R1 N N X7 N N R3 R1 Pri ButCH2 Me Me Me Pri Pri Me Me cyclo-C6H11 But Ph Ph Ph Ph Ph Ph Ph Ph cyclo-C3H5 All (E)-MeCH=CHCH2 (E)-ButCH=CHCH2 (E)-PhCH=CHCH2 (E)-MeCH=CH (Z)-EtCH=CH a With decomposition. Ph N N HClO4 ClO¡¦ N +N 4 Me (87%) Ph HClO4 N N + N Me N (60%) 2 CF3SO¡¦3N N N O + N N N Ph 37 (100%) Cl Ph N BF¡¦4 N HBF4 N +N 38 (90%) Ph R3 R2 Me Me Me Me Ph ButCH2 Ad Bn Et Bn Bn Me Bn Et Et Et Me Bn Ph Me Me Me Me Me Me Me Me Pri Me Me Me Me Me Me Et Me Me Me Me Et Et Et Ph Ph Ph Me Me Me Me Me Me Me S V Voitekhovich, P N Gaponik, O A Ivashkevich 4.Heterocyclisation reactions The main advantage of the synthesis of tetrazolium salts by heterocyclisation is that this excludes preliminary synthesis of tetrazole derivatives. Reactions of alkyl and aryl azides with nitrilium salts which proceed as [3+2]-cycloaddition and result in 1,4,5-trisubstituted tetrazolium salts in good yields have the greatest preparative significance (Table 6).22, 65, 66 Ph + R1 C N R2 + R3 N N N ClO¡¦4X7 R2 R1 N N N R3 Ph N +N Ph An intramolecular version of this reaction is also known.It was shown31 that cyclisation of the in situ generated N-methyl-o- azidonitrilium triflates 39 is a rather convenient and efficient procedure for the synthesis of annelated 1,4,5-trisubstituted tetrazolium salts 40. MeSO3CF3 (H2C)n CN N3 n=1¡À3. Yield (%) XBF4 BF4 FSO3 FSO3 �ºF3SO3 BF4 FSO3 FSO3 FSO3 FSO3 FSO3 FSO3 FSO3 FSO3 SbCl6 BF4 SbCl6 SbCl6 SbCl6 PF6 PF6 PF6 PF6 PF6 PF6 PF6 71 52 41 80 87 77 32 71 58 60 60 75 70 55 55 53 78 65 52 97 51 48 57 11 48 60 + 7 R2 R1 N+N X7 N+ X7 N N R3 F3CSO¡¦3 (H2C)n + Me N C N3 N+ 3 Me (H2C)nNN NF3CSO¡¦ 40 (43% ¡À 75%) 39 Ref.M.p. /8C 100 ¡À 101 166 ¡À 168 205 ¡À 206 155 ¡À 158 210 ¡À 211 a 144 129 ¡À 130 oil 118 ¡À 119 128 ¡À 129 oil 145 ¡À 148 135 165 ¡À 167 204 130 176 156 172 219 100 ¡À 101 93 133 ¡À 134 165 ¡À 170 190 ¡À 192 77 ¡À 79 22 22 22 65 22 22 22 65 65 65 65 65 65 65 65 65 65 65 65 66 66 66 66 66 66 661,3- and 1,4-Substituted tetrazolium salts Heterocyclisation reactions can also be employed for the synthesis of 1,3,5-trisubstituted tetrazolium salts.The use of these reactions is limited due to the low availability of the starting reagents; it has been reported, however,7 that the hydrazinium salt 41 is converted into 1-benzyl-3,5-diphenyltetrazolium iodide (42) under the action of a base. CH2Ph N NH NH Ph Ph I7 N NEt3 + I7 Ph N N+ Ph N CH2 41 H2N Ph 42 (60%) 1,3-Diaza-2-azoniaallene salts 43a,b [Ar=2,4,6-Cl3C6H2 (a), 4-ClC6H4 (b)] generated in situ from N-chlorotriazenes 44a,b react with carbodiimides to give dihydrotetrazolium derivatives 45 which are convert into 5-alkylamino-1,3-diaryltetrazolium salts 46 upon heating.67, 68SbCl5 or KPF6 Ar N N N Ar CH2Cl2,778 8C 44a,b Cl Ar + R4 N C N CR1R2CH2R3 N N N CH2Cl2,778 to 23 8C Ar X7 43a,b Ar Ar X7 R4 N N N N MeCN, 81 8C, 3 h N + + 7R1R2C CR3H N N H N N Ar Ar X7 R1R2 46 (50% ± 77%) N R4 HH R3 45 (53% ± 57%) R4 Ar R3 R2 R1 X i H i H Prcyclo-C6H11 But H Prcyclo-C6H11 SbCl6 SbCl6 SbCl6 PF6 PF6 2,4,6-Cl3C6H2 2,4,6-Cl3C6H2 2,4,6-Cl3C6H2 4-ClC6H4 4-ClC6H4 H Me H (CH2)4 Me Me H Me H (CH2)4 The use of cyanamides in this reaction leads to 5-dialkyl- amino-1,3-diaryltetrazolium salts 47.67, 68 Ar N Ar N N R1R2N C N 43a 44a N+ SbCl¡6C CH2Cl2, 3 h 778 to 23 8C NR1R2 Ar N N+ Ar N SbCl¡6N 47 NR1R2 Yield (%) R2 R1 71 H 34 Me 61 Pri 65 Pri But Me Me Pri 729 III. Physicochemical properties and structures of tetrazolium salts In the individual state and under normal conditions, tetrazolium salts are stable crystalline compounds (except for the salts containing the MeSO¡4 anion) frequently isolated as viscous fluids (see, e.g., Refs 23, 28, 32).1,4-Dialkyltetrazolium perchlorates manifest higher melting temperatures than the 1,3-isomers.45 Certain 1,3-dialkyltetrazolium perchlorates possess rather low (for ionic compounds) melting temperatures (*70 ± 80 8C, see, e.g., Refs 45, 46, 48). This allows the preparation of ionic fluids possessing unique properties presently established for other azolium salts.69, 70 1. NMR spectra NMR spectroscopy is one of the main techniques for the identi- fication of tetrazolium salts. The chemical shift (CS) of the endocyclic carbon atom in the 13C NMR spectra strongly depends on the isomer type.The signals for carbon C(5) in 1,3- substituted tetrazolium salts 2 are downfield shifted by 7 ± 16 ppm in comparison with the corresponding signals for 1,4-isomers 71, 72 (Table 7). It is noteworthy that the nature of the substituents at the nitrogen atoms of the ring has little effect on the CS, being equal to *147 ± 151 ppm for 1,3-disubstituted, and *138 ± 142 ppm for 1,4-disubstituted tetrazolium salts. In the case of trisubstituted derivatives, the magnitude of the CS depends critically on the nature of the substituent at C(5). The signal of the C(5) atom characteristic of all 13C NMR spectra is shifted upfield in the following series of substituents: Pri>Me>NH2>Cl> Br>I.15, 18, 21, 22, 28, 61, 71 ± 75 A comparative study 71, 72 of the NMR spectra of tetrazolium salts and related tetrazoles revealed that the quaternisation at the a-nitrogen atom of N-substituted tetrazoles little affects the magnitude of the CS of the C(5) atom (*1 ± 4 ppm), while the quaternisation of the b-atom significantly changes it as can be evidenced from the downfield shift of 7 ± 10 ppm.The slight change in the CS observed upon quaternisation of the N(4) atom is due to the oppositely directed effect of two factors, viz., the electronegative group N += (the downfield shift) and the decrease in the order of theN7C bond in the cation formed as a result of polarisation (the upfield shift). The 1H NMR spectra of tetrazolium salts non-substituted in position 5 are characterised by a downfield shift of the signal for the proton at the endocyclic carbon atom and a significant (by *20 Hz) increase in the spin ± spin coupling constant (SSC) 1J(13C71H) in comparison with the starting com- pounds.23, 30, 44, 71, 72 The CS value of the proton at C(5) strongly depends on the structure of the salt and the type of solvent. Chemical shifts of isomeric 1,3- and 1,4-disubstituted tetrazolium salts differ by *0.5 ± 1 ppm and even more.30, 43 ± 49 Thus the CS of 1,4-dialkyltetrazolium perchlorates and tetrafluoroborates are 11.1 ± 11.9 ppm, whereas those of 1,3-isomers are 10.2 ± 10.9 ppm [in (CD3)2SO].On going from DMSO to MeCN, the signal for the proton in the 1H NMR spectrum is shifted upfield (10.6 ± 10.9 ppm for 1,4-substituted and 9.6 ± 10.1 ppm for 1,3- substituted derivatives).30, 43 ± 49 Identification of tetrazolium salts on the basis of 1H NMR spectra often makes use of characteristic CS of the protons of the substituent owing to magnetic equivalence (or non-equivalence) of the groups.Thus the CS of the protons in the alkyl groups (Me, Et, But) in positions N(1) and N(4) differ considerably from those of the analogous substituents at the atoms C(5) and N(3).3, 14, 17, 18, 21, 35, 36, 49, 71 NMR spectroscopy allows both identification of tetrazolium salts and an analysis of the conjugation of the tetrazole fragment with the substituent. Thus the change in the CS of the carbon atoms of the vinyl group in the 13C NMR spectra upon quaterni- sation of vinyltetrazoles allows one to conclude that there is a decrease in the p,p-conjugation in N-vinyltetrazolium salts and an increase in the p,p-conjugation in C-vinyltetrazolium salts as compared to the starting tetrazoles.28 These data are consistent with the results of quantum-chemical calculations of geometric730 Table 7.The chemical shift values of the signals of C(5) atoms of 1,3- (2) and 1,4-substituted tetrazolium (3) salts determined from 13C NMRspectra (d). R1 X7 R2 N N +N N R3 2 R1 NH2 NH2 CH2=CH MeH H Ph HSO4 H2SO4 H Et Ph BF4 (CD3)2CO Me Me Me CH2COMe H Ph Ph BF4 (CD3)2SO Ph Ph Ph Ph NH2 N3 Cl BrI Ph Ph BF4 (CD3)2SO Hg EtO EtS MeS Me EtH Me CH2 =CH and electronic structures of these compounds.Also, correlations were found between the difference in CS of the carbon atoms of the vinyl group and the difference in the calculated effective charges on these atoms. 28 In recent years, 15N and 14N NMR spectroscopies (see Refs 28, 71 ± 79) have become a popular approach to the solution of various structural problems related to tetrazolium salts. Thus this method allowed reliable determination of the direction of quaternisation 28, 34 and protonation 78, 79 of N-substituted tetra- zoles. The most significant shift in the NMR spectrum (up to Table 8. The chemical shift values of the signals of endocyclic nitrogen atoms of tetrazolium salts determined from 15N NMR spectra (d).Compound R1 + R2 N R3 N X7 N N R1 N R2 N X7 + N N R3 R1 N + N X7 N N R1 + R3 R2 N N X7 N N 3 R3 R2 Me CH2COPh Me Me Ph Ph Ph Ph Ph Ph Ph Me Me Me Ph Ph Ph Me Me Me R3 R2 R1 Me CH2=CH Me H CH2 =CH Me Me H Me Bu NH2 H Me Me Me Me Me Me CH2=CH MeS NH2 H Ph Ph BF 7129.8 Ph Ph Ph Ph Ph Me Ph Ph Ph Ph Ph H Cl N3 Hg EtO EtS H 7 7 BF4 7 7 BF4 7 7 BF4 3-Me 4-Me 3-Et 4-Et 7 7 BF4 S V Voitekhovich, P N Gaponik, O A Ivashkevich Solvent X Compound 2 3 Cl Br MeSO4 I (CD3)2SO (CD3)2SO CD3OD D2O (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO BF4 BF4 BF4 BF4 147.3 148.3 158.1 158.7 162.7 161.7 147.5 157.7 155.9 153.9 143.2 117.3 183.3 160.6 162.7 162.4 777 (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO CD3OD BF4 BF4 BF4 Cl ClO4 ClO4 MeSO4 115 ppm) upon transition from tetrazoles to tetrazolium salts is observed for the signal of the nitrogen atom directly involved in quaternisation.The CS values of endocyclic nitrogen atoms depend substantially on the type of the salt and the nature of the substituents (Table 8). An analysis of 15N and 14N NMR spectra of some 1,3,5-trisubstituted tetrazolium salts prompted the con- clusion that the positive charge of the cation is predominantly localised on the N(3) atom,28, 74 ± 77 which is consistent with the X-ray diffraction data.33, 48 N(4) N(3) N(2) N(1) XMeSO4 MeSO4 Cl HSO4 7148.1 7142.5 7182.0 7147.5 717.5 714.4 727.7 717.0 717.5 722.2 727.7 711.5 7148.1 7127.5 7182.0 7130.5 MeSO4 Cl Cl 7148.1 7148.6 7179.1 4 776.1 771.1 7107.4 773.3 773.3 797.6 759.7 7112.7 783.8 765.8 7101.2 7100.9 7109.0 792.4 790.7 797.5 787.2 7103.4 793.2 796.8 717.9 716.5 730.6 733.5 722.4 731.8 726.9 736.8 726.8 720.3 BF4 BF4 BF4 BF4 BF4 HSO4 7132.7 7153.1 7118.4 7158.4 7139.8 7144.6 778.4 7165.8 781.0 7154.6 795.3 712.8 784.7 714.8 742.4 739.7 743.6 739.1 7132.2 7129.0 7132.6 7128.6 Ref. 138.2 140.3 148.5 148.9 149.4 154.7 7777777777152.7 154.9 142.1 15 15 73 35 28 18 60 61 61 60 60 60 60 74 74 73 22 22 28 Ref.Solvent 28 28 73 40 CD3OD CD3OD (CD3)2SO H2SO4 28 73 73 76 76 76 76 74 74 40 CD3OD (CD3)2SO (CD3)2SO CD3CN (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO CD3CN 34 34 34 34 (CD3)2SO (CD3)2SO (CD3)2SO (CD3)2SO1,3- and 1,4-Substituted tetrazolium salts 2. X-Ray diffraction analysis By now, crystal structures of eight tetrazolium salts, viz., 5-(1,5- diphenyl-3-formazanyl)-1,3-diphenyltetrazolium chloride (48),3 1,3-di-tert-butyl-5-methyltetrazolium picrate (18d),33 5-(4-chloro- phenyl)-1-methyl-3-phenyl- (49),29 1,3-di-tert-butyl- (50),48 1-tert- butyl-3-isopropenyl- (51),48 1-(2,2-dimethyl-4-oxopentyl)-4-me- thyltetrazolium perchlorates (52),46 tetrazolo[4,5-b][1,3]thiazi- nium- (30) 58 and thiazolo[3,2-d ]tetrazolium perchlorates 31b, have been established.59 Me N C N NHPh PhNPh But N N N N 2,4,6-(NO2)3C6H2O7 Cl7 + + N N N N Ph But 18d 48 C6H4Cl-4 But Me N N N N ClO¡ ClO¡ 4 4 + + N N N N But 50 Ph 49 But N N ClO¡4+ + ClO¡ CMe2CH2CMe 4 Me N N N N O Me N N H2C 52 51 The tetrazole rings in tetrazolium salts and other tetrazole derivatives are planar.The parameters of the tetrazole ring depend critically on the type of the salt. Thus the C(5)7N(1) [1.317(4) A] and C(5)7N(4) [1.315(4) A] bond lengths in the cation 52 are practically identical and intermediate between the double and ordinary bond lengths, N(2)7N(3) [1.276(4) A] being the shortest bond and close to a double bond.The values of the bond angles at the N(2) and N(3) atoms practically coincide with that in a regular pentagon, while those at the N(1) and N(4) atoms approximate it. These data suggest that the structure of the 1,4-disubstituted tetrazole ring is close to that of a symmetrical ring with delocal- isation of the excess positive charge over the N(4)7C(5)7N(1) fragment.46 A similar structure is characteristic of the tetrazole ring of fused 1,4,5-trisubstituted tetrazolium salts 30 (see Ref. 58) and 31b (see Ref . 59). The heterocycles of 1,3-substituted tetrazolium salts are non- symmetrical.The N(2)7N(3) [1.287(3) ± 1.299(3) A] bond in the cations 18d, 48 ± 51 is the shortest bond; the N(4)7C(5) [1.300(4) ± 1.314(3) A] bond length differs from it only insignif- icantly. This suggests that the positive charge of 1,3-substituted tetrazolium salts is predominantly localised on the N(3) atom rather than on the N(1) atom which represents the quaternisation site.48 The non-coincidence of the quaternisation site and the site of localisation of the positive charge in the case of some tetrazo- lium salts was established experimentally: the quaternisation of N-substituted tetrazoles differing in the positions of substituents gave identical compounds.30, 48 The exocyclic N7C bonds in the cations under study lie in the plane of the ring.Their lengths are rather similar and differ only insignificantly from the length of the corresponding ordinary bond N7C with the exception of the N(3)7C(Me) bond in the cation 51. Its length [1.440 A] is some- what smaller than that of an ordinary bond, while its torsion angles deviate by no more than 10.6 8 from 0 8 or 180 8, which is suggestive of the conjugation of the p-systems of the tetrazole ring with the C=C bond.48 IV. Reactions of tetrazolium salts Systematic studies of the chemical properties of 1,3- and 1,4- substituted tetrazolium salts were not carried out until recently. This section is devoted to the consideration of the most important 731 electrophilic and nucleophilic substitution reactions of tetrazo- lium salts at the carbon atom of the heterocycle and other transformations under the action of various reagents.1. Substitution reactions at the carbon atom of the heterocycle Only very few cases of substitution of the hydrogen atom in position C(5) of 1-monosubstituted tetrazoles were described before the 1980's. These include deuterium ± hydrogen exchange, lithiation, mercuration and bromination.72, 80, 81 It was believed that the problems encountered in the electrophilic substitution of hydrogen were caused by the deactivation of position 5 by annular nitrogen atoms.82 With regard to disubstituted tetrazolium salts, the only known case was the isotope exchange in alkaline and acid media. + R1 N N R2 7H+ N N D 7 + + D+ R1 R1 N N N N R1 N N R2 R2 N N N N R2 N N 53 It is assumed that this exchange proceeds through the inter- tetrazolium mediate formation of ylide 53 or its isoelectronic analogue, carbene.It was shown that the kinetic CH-acidity changes in the following order: 2-R-tetrazoles<1-R-tetrazoles<1,3-disubsti- tuted tetrazolium salts<1,4-disubstituted salts.80, 81, 83 ± 85 The first successful attempts at aminomethylation 86, 87 and iodination of 1-R-tetrazoles were reported in the late 1980's ± early 1990's.88, 89 The experimental data suggest that these reactions proceed through the intermediate formation of an ylide, which accounts for the introduction of a substituent into position 5. This reaction begins with an attack of an electrophilic reagent at the most nucleophilic N(4) atom and the formation of a tetrazolium cation, which is further deprotonated into the intermediate ylide (carbene) A stabilised by intramolecular rearrangement.R R H H N N E+ N N N 7H+ E +N N N R R R E 7 N N N N N N N N E E +N N N N A E=I, CH2NAlk2; R=Ar, Alk, H. Simultaneously electrophilic substitution at the C(5) atom of tetrazolium salts was reported. It was shown, in particular, that the substitution of lithium for hydrogen in 2,3-diaryltetrazolium salts occurs under the action of tert-butyllithium or lithium bis(trimethylsilyl)amide.90, 91 The condensation of 1,3-diphenylte- trazolium tetrafluoroborate with a diazonium salt in the presence of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), which favours the formation of a mesoionic carbene, viz., 1,3-diphenyltetrazolylene (54), has been described.60 Ph Ph BF¡4+ N N DBU N N 4-Me2NC6H4N N BF¡4 + 730 8C N N N N Ph Ph 54732 Ph BF¡4N N N N C6H4NMe2-4 + N N Ph (69%) It was found that 1,3-diaryltetrazolium salts undergo mercu- ration upon heating with mercury(II) acetate in dimethyl sulfoxide to yield bis(tetrazolio)mercury(II) salts 55 the mercury atom in which is easily replaced by a halogen.60 Ar1 Ar1 Ar1 N N N Hg(OAc)2 N N N X2 Hg + + + N N N N N 2BF¡4 Ar2 Ar2 Ar2 N BF¡455 (56% ± 82%) Ar1 N N X + N Ar2 4 N BF¡ (54% ± 78%) Ar1=Ph, 4-MeC6H4; Ar2=Ph, 4-MeOC6H4; X=Cl, Br, I.Nucleophilic substitution at the C(5) atom of tetrazolium salts has been studied in more detail and is often employed in the synthesis of mesoionic tetrazoles. Thus it was shown that the ethoxy 92 and the methylthio groups 73 and the chlorine atom 60 of 1,3,5-trisubstituted tetrazolium salts can be replaced by sulfide, hydroxide and selenide ions. R1 R1 N N N N NaYH or Na2Y Y7 X + + N N N N Z7 R2 R2 R1, R2=Me, Ph, 4-MeC6H4, 4-ClC6H4; X=Cl, OEt, SMe; Y=O, S, Se; Z=BF4, MeSO4. 5-Ethoxy-1,3-diphenyltetrazolium tetrafluoroborate (56) reacts with C-nucleophiles, e.g., malonic dinitrile,93 lithium and sodium cyclopentanedienide,94, 95 indenide, fluorenide and cyclo- penta[def]phenanthrenide 95 to give mesoionic tetrazoles 57 and 58a ± d.Ph CN N N CH2(CN)2 7 + N NEt3, MeCN N CN Ph Ph 57 (52%) N N OEt + R3 R2 N Ph R1 N BF¡ M+ Ph 4 R2 7 N 56 N R1 R4 7 + N N R3 Ph R4 58a ± d (52% ± 76%) M=Li, Na; R1=R2=R3=R4 = H (a); R1=R2=H,R3±R4=CH=CHCH=CH (b); R1±R2=CH=CHCH=CH, R3±R4=CH=CHCH=CH (c); R1±R4= =CH7CH=C7CH=CH7C7CH=CH7CH= (d). Mesoionic tetrazolium amides 59 were prepared by the reaction of 5-chloro-1,3-diphenyltetrazolium tetrafluoroborate 38 with N-nucleophiles, such as aliphatic amines,96 aniline, hydroxylamine, N-mono- and N,N-disubstituted hydrazines.60 S V Voitekhovich, P N Gaponik, O A Ivashkevich The reaction of the salt 38 with ammonia yielded bistetrazolio- amide 60 as the main product;60 the reaction with hydrazine and arylenediamines 97 gave tetrapolar mesoionic bistetrazoles 61, 62.Ph N RNH2 N N7 R + N N Ph 59 (37% ± 85%) R=OH, NHPh, NMePh, Ph NPh2, NHC6H4Me-4, Me, C6H11, Ph, 4-MeOC6H4 Ph Ph N N N N N N NH3 Cl + N 7 + + N N N N N N BF¡ Ph 4 Ph Ph 4 38 BF¡ 60 (73%) Ph PhN N N2H4 N N 7 7N N + + N N N N Ph Ph 61 (62%) Ph Ph Ph N N N 7 N N H2N X NH2 N X N7 Cl + + N N N N +N N N Ph Ph BF¡ 62 (38% ± 80%) Ph 4 38 , , , X=5-Azido-1,3-diphenyltetrazolium tetrafluoroborate (63) pre- pared by the reaction of sodium azide with the salt 38 has rich synthetic potential. The attack of strong nucleophiles, e.g., Ph N NaOH N O7 + N N Ph 64 (87%) Ph NH N N BF¡4N + N N Ph 65 (86%) Ph Ph Ph N Ph N NH2 N N N N 7 N N N N N3 N + + DBU N N N N N N BF¡ Ph 4 Ph 66 (27%) 63 Ph N N PPh3 N PPh3 + N N NBF¡ N 4 Ph 67 (93%) Ph N 7 N N X N N + NaX N N Ph 68 X=CN (93%), SO2C6H4Me-4 (46%).733 1,3- and 1,4-Substituted tetrazolium salts Me PhNH C N R=Me H2O 7N2 C6H2(NO2)3-2,4,6 O 75 (40%) ButNH C NHPh 2,4,6-(NO2)3C6H2O7 R=But NHEt2 7N2 +NEt2 sodium hydroxide or amines at the endocyclic carbon atom, yields substitution products.Weaker nucleophiles, such as the cyanide ion, p-toluenesulfonate anion, triphenylphosphine, etc., attack the terminal nitrogen atom to give the addition products 67 and 68.61 The nucleophilic substitution at the C(5) atom in 1,4,5- trisubstituted tetrazolium salts has been described exclusively for 1,4-dimethyl-5-(methylthio)tetrazolium salts.It was shown that they react with bases to give 1,4-dihydro-1,4-dimethyltetrazoles 69 and 70 (yields 27%± 40%).7, 17, 62 76 (50%) R=Me, But. O SMe+ Me Me Me Me NaOH N N N N X7 N N N N 69 NR SMe It has been reported 13 that consecutive treatment of 1-(ethoxycarbonylmethyl)-4-methyltetrazolium tetrafluorobor- ate (77) with triethylamine and then with benzylamine, phenyl- hydrazine or hydrogen sulfide yields hydantoins 78 ± 81. Unfortunately, neither the methodology of these reactions, nor the yields of the target products have been reported.13 + Me Me Me Me + RNH2 N N N N Et3N N N Me EtO2CCH2 X7 N N BF¡4N N N 70 N77 X=BF4, MeSO4; R=Me, PhNH.O Me N PhNHNH2 2. The action of bases N NHPh HN 78 O O Bn N Me N BnNH2 N Me N Bn + HN HN 80 79 O Me N H2S S NH 81 The majority of currently known tetrazolium salts are sensitive to bases. A base attacks predominantly the two carbon atoms with the lowest electron density, viz., the C(5) atom and the carbon atoms in the a-position relative to the nitrogen atoms of the heterocycle, which leads to the opening of the tetrazole ring, dealkylation or self-condensation. The salts devoid of substituents at the C(5) atom are the most susceptible to bases. The abstraction of the hydrogen atom under the action of bases is accompanied by the formation of heteroaromatic carbenes, viz., tetrazolidenes. The latter, in turn, undergo fragmentation the direction of which depends on the type of the salt used.It was shown 13, 98 that 1,4- disubstituted tetrazolium salts are split into molecular nitrogen and carbodiimides 71 under the action of tertiary amines at 20 8C. Along with other reactions, this reaction is used in the synthesis of carbodiimides (see Ref. 99) and is the sole procedure in the synthesis of, e.g., vinylcarbodiimides.13 R1 R1 1-R-4-(2-Methyl-4-oxopentan-2-yl)tetrazolium perchlorates 14 evolve nitrogen in the presence of bases to be transformed into 4,4-dimethyl-1-R-6-methylidene-3,4,5,6-tetrahydropyrimi- din-2(1H)-ones 82 and 4,4,6-trimethyl-1-R-3,4-dihydropyrimi- din-2(1H)-ones 83, the latter being predominant.47 N N + N N B7 R CMe2CH2COMe DMSO or NaOH N N H R2 N C N R1 + 7HB 7N2 N N ClO¡ N N 7N2 4 71 N N 14 R2 R2 Me Me Me Me HN HN + O O N N Me R R The carbodiimides formed, being highly reactive com- pounds,99 react with an excess of the base or the solvent, or undergo intramolecular cyclisation. Thus the reaction of 1-ethyl- 4-(p-tolyl)tetrazolium tetrafluoroborate (72) with diethylpro- pynylamine in the presence of water gives N-ethyl-N0-(p-tolyl)- urea (73).98 83 82 + R=Ph, Bn.Et C6H4Me-4 MeC CNEt2, H2O N N 7N2 N N BF¡4 EtNHCONHC6H4Me-4 73 72 3-Monosubstituted quinazoline-2,4(1H,3H)-diones 85 are formed in high yields on dissolution of 4-alkyl-1-(2-carboxyphe- nyl)tetrazolium perchlorates 84 in dimethyl sulfoxide.47 O CO2H R DMSO N + 4-Alkyl-1-phenyltetrazolium picrates 74 are decomposed slowly in the presence of water to yield trisubstituted urea 75; after treatment with diethylamine, they are converted into guani- dinium picrate 76.33 R 7N2 N N O + ClO¡4Ph R NH N N N N 84 2,4,6-(NO2)3C6H2O7 85 (90% ± 98%) N N 74 R=But, CMe2CH2COMe.734 perchlorates 4-Alkyl-1-(2-hydroxyaryl)tetrazolium 86 undergo recyclisation into 2-alkylaminobenzooxazoles 87 under the action of bases.47 Y OH O 1) DMSO Y 2) NaOH NHR + R N N N Z Z ClO¡487 (82% ± 95%) N N 86 R=But, CMe2CH2COMe; Y=H, NO2; Z=H,NO2.Preparation of 2-methylamino-5,7-dinitrobenzooxazole (88) is an example of a successful conversion of readily available 1-monosubstituted tetrazoles into benzooxazoles without the intermediate isolation of the tetrazolium salt.100 NO2 NO2 OH O 1) Me2SO4 2) H2O, 60 8C NHMe N N N O2N O2N 88 (46%) N N 1-Aryl-4-ferrocenylalkyltetrazolium salts undergo dealkyla- tion by aqueous ammonia or sodium carbonate at 20 8C to yield 1-aryltetrazoles and ferrocenylalkyl alcohols.1-Ferrocenyl- methyl-4-phenyltetrazolium tetrafluoroborate (89) undergoes the Stevens rearrangement in the presence of potassium tert- butoxide in dioxane to give 1,5-disubstituted tetrazole 90.44 Ph Ph Ph N N N BF¡4ButOK, dioxane N N N CH2Fc + 7KBF4,7ButOH N N N N N N 90 (52%) CH2Fc CH2Fc 89 Fc=C5H4FeC5H5. 1,3-Disubstituted tetrazolium salts react with strong bases to give 1,3-disubstituted 3-cyanotriazenes 91a,b.60, 82 Compound 91a is decomposed under the action of the hydroxide ion resulting in diazomethane and methylcyanamide anion.R R N N N KOH or DBU N X7 + N 7HX N N N R R RN 7 R=Me, OH7 CN N CH2N2+MeNCN N R 91a,b R=Me (a), Ph (b); X=Cl, BF4. 3,5-Diaryl-1-methyltetrazolium salts 92 are resistant to the action of Et3N in toluene at 20 8C, but react vigorously with sodium ethoxide in benzene or toluene to yield a complex mixture of reaction products among which biaryls are predominant.29 Me Me ClO¡4 N N 4-XC6H4 EtONa N N 4-XC6H4 + 7 + RH 7N2 N N EtO N N C6H4Y-4 C6H4Y-4 92 C6H4Y-4 +R R 4-XC6H4 Y=H: X=H, Me, MeO, Cl; X=H: Y=H, NO2, Me, MeO, Cl; R=Ph, C6H4Me-4. 1,4,5-Trisubstituted tetrazolium salts 93a,b undergo N-deal- kylation 98 or fragmentation in the presence of bases,25, 101 and ring-opening products always contain alkyl or aryl azide.Me I7 + Me N NN N 93a,b RN3+MeCONHMe N N 93b+ Py N N PhN3 + R=Me (a), Ph (b). 1,4,5-Trialkyltetrazolium salts 94 undergo deprotonation under the action of sodium hydride in tetrahydrofuran. This reaction has been studied in sufficiently great detail, since it allows one to prepare 1,4-dihydro-5-alkylidenetetrazoles 95, the starting compounds in the synthesis of iminoaziridines, tetrazines, and other nitrogen-containing heterocyclic systems.66, 102 Me ClO¡4 N R1 N + N R2 N Me 94 D 7N2 hn 7N2 AlkN3 ArN3 7N2 R1, R2=H, Alk. Some interesting conversions of N-acyl- and N-phenacyl- tetrazolium salts under the action of bases have been described.21, 27, 35, 103, 104 It was shown that tetrazolium salts can be converted into ylides, fused tetrazoles, and other heterocyclic molecules.The direction of these reactions depends on the nature S V Voitekhovich, P N Gaponik, O A Ivashkevich Me OH KOH Me Py R N NN N Py N N Py Ph Et3N 7N2 N Py Me NN N Py N N Py NHMe Me N R1 NaH N 7NaClO4,7H2 N R2 N Me 95 Me R1 R2 N N N N N N R2 Me Me R1 N MeR1 N Me R2 R2 Me R1 N N N N N N N Alk MeR1 N Ar R2Me Me N N N N1,3- and 1,4-Substituted tetrazolium salts of the reagent and the substituents in the substrate. Thus 1-alkyl- 5-methyl-4-phenacyltetrazolium bromides 96 are converted into ylides 97 under the action of aqueous solutions of potassium carbonate.27 7 CH2COC6H4X-4 CHCOC6H4X-4 N N K2CO3, H2O, 0 8C Br7 Me Me N N+ N N N+ N 96 R 97 (77% ± 92%) R R=Me, Prn, Pri, Bui, CH2But, cyclo-C6H11; X=H, Br, NO2.Similar conversions were observed in the case of 1,3-di- and 1,3,5-trisubstituted tetrazolium salts containing a phenacyl group at N(1) or N(3) atoms.21 1,5-Dimethyl-4-phenacyltetrazolium salts 98 undergo recyclisation into 1,2,5-trisubstituted imidazol- 4-ones 99 upon heating with aqueous sodium hydrogencarbon- ate.103 Analogous salts 100 yield 3a,6-dihydro-3H-imidazo- [1,2-d ]tetrazoles 101 in a reaction with aqueous ammonia.104 CH2COAr O N N N + NaHCO3, H2O, 80 8C Me Me Br7 N N N Ar 98 Me Me99 (48% ± 49%) Ar=Ph, C6H4Br-4.Ph CH2COPh N + R1 N N NH3, H2O, 0 8C N R1 R2 N X7 N N N N 101 (72% ± 98%) R2 100 R1=Me, Et, Pr, Pri, cyclo-C6H11, Ph, Bn, CH2C6H4NO2-4; R2=Me, Et, Bn; X=Br, I, ClO4, SbCl6, PF6. 1-Acetylmethyl-5-amino-3- (102a) and -4-methyltetrazolium bromides (102b) undergo cyclisation into 6-acetyl-2,5-dimethyl- 2H- (103a) and 6-acetyl-3,5-dimethyl-3H-imidazo[1,2-d ]tetra- zoles (103b) under the action of acetic anhydride and triethyl- amine.35 Ac CH2Ac Ac2O, Et3N Me N N N + 100 ± 110 8C, 0.5 h N N NH2 N N Br7N N Me Me 103a,b 102a,b 103a: 3-Me (86%); 103b: 4-Me (51%). Table 9. The relative stability of tetrazolium cations (data from ab initio quantum-chemical calculations).Method Compound Total energy of 1,4,5-isomer /kJ mol71 Protonated 1H,5H-tetrazole 7560438.6 7564290.9 7567204.6 7567545.4 7567573.5 7567554.4 7653363.5 71180387.0 STO-3G 3-21G 6-31G 6-31G*//6-31G 6-31G** MP2/6-31G** 6-31G GIAO-CHF GIAO-CHF 7 71180310.3 Protonated 1-methyl-5H-tetrazole N(1)-Methyltetrazolo[1,5-a]pyridine N(3)-Methyltetrazolo[1,5-a]pyridine a Original author's data. 735 3. The action of acids The heterocycle in tetrazolium salts is usually resistant to acids. In reactions with acids, the modification affects only the substituents and is usually followed by their elimination and migration into another position of the ring.It was found that 1-R-3-tert-butyltetrazolium perchlorates are slowly converted into 1-R-4-tert-butyltetrazolium perchlorates on storage in perchloric acid.45, 47, 55 But + But + N N N N HClO4 ClO¡ ClO¡ 4 4 N N N N R R (65% ± 95%) R=Me, Et,But, cyclo-C6H11, Ph, 2-HO2CC6H4, 2-HO-5-NO2C6H3, 2-HO-4-NO2C6H3. Higher thermodynamic stability of 1,4-disubstituted tetrazo- lium salts in comparison with their 1,3-isomers is the driving force of this process as can be evidenced from the results of quantum- chemical calculations of the heats of formation and total energies of isomeric tetrazolium cations carried out with the use of semi- empirical 28, 45 and ab initio methods (Table 9) with full optimisa- tion of the cation geometry.1-tert-Butyl-3-R-tetrazolium salts do not undergo isomerisa- tion in perchloric acid, but partly lose their tert-butyl groups.45 The conversion of 1-R-3-tert-butyltetrazolium salts into the corresponding 1,4-isomers is accompanied by a partial loss of the tert-butyl group.45, 55 Nothwithstanding, storage of 1-R-3- tert-butyltetrazolium salts with perchloric acid has been employed for the synthesis of 1-R-4-tert-butyltetrazolium salts.45, 47 In contrast to 1,3-disubstituted tetrazolium salts, 1,3,5-trisub- stituted derivatives do not undergo isomerisation under similar conditions but are subject to dealkylation.45 This property of 1,3,5-trisubstituted tetrazolium salts was used in the synthesis of mono- and binuclear 1,5-disubstituted tetrazoles where alkyl substituents are introduced into position N(1) of the tetrazole ring of 2,5-disubstituted tetrazoles with subsequent elimination of the tert-butyl substituent (the yields of the target products are 78%± 94%).36, 106 But N N AlkX N N ButOH R R N N H2SO4 N HN But N N N N HCl, 100 8C + R R 7CH2=CMe2,7HX N N N N X7 Alk Alk Ref.DE /kJ mol71 Total energy of 1,3,5-isomer /kJ mol71 7560416.0 7564267.5 7567189.5 7567537.7 7567566.1 7567543.8 7653350.6 7 105 105 105 105 see a see a 78 34 34 22.6 23.4 15.1 7.7 7.4 10.6 12.9 76.7 76.7736 Boiling of 1-alkyl-4-aryl-5-(2-dimethylaminovinyl)tetrazo- lium perchlorates 104 in hydrochloric acid leads to the modifica- tion of the substituent at the C(5) atom.25 R R N N N N HCl CH2CHO CH CHNMe2 + + N N 7CO 100 8C N N ClO47 ClO¡4C6H4X-4 C6H4X-4104R N N Me + N ClO¡4NC6H4X-4 R=Me, Et; X=H, Me, F, Cl, Br, I.Treatment of 3-acetylmethyl-1,5-dimethyltetrazolium chlor- ide (105) with a nitrating mixture results in mesoionic tetrazole 106.107 NO2 7 AcCH2 N N N N O2N HNO3 Me Me + + H2SO4 N N N N Cl7 Me Me 106 (35%) 105 1,4,5- (107) and 1,3,5-trisubstituted tetrazolium salts (108) undergo cyclisation into 1H- and 2H-pyrrolotetrazoles 109 and 110 upon heating in acetic acid in the presence of sodium acetate.37 R3 R3 R4 R4 N N N N NaOAc, AcOH O 100 ± 110 8C N X7 N R1 R1 N +NR2 R2 107 109 (52% ± 86%) R3 R3 R4 R4 N N NaOAc, AcOH N N + + O 100 ± 110 8C N R2 R2 X7 N N 7NR1 R1 110 (10% ± 80%) 108 R1=MeCO, CO2Me; R2=Me, Ph; R3=H, Me; R4=Me, Ph, 4-BrC6H4; X = BF4, MeSO4.4. The action of reducing and oxidising reagents Tetrazolium salts are highly resistant to oxidising and reducing reagents; their heterocycles are not opened when treated with these reagents. Thus the oxidation of 1,3-diaryl-5-mercaptotetra- zolium salts with concentrated nitric acid leads merely to the elimination of the substituent from position C(5) and the for- mation of 1,3-diaryltetrazolium salts,60 while treatment of 5-azido-1,3-diphenyltetrazolium tetrafluoroborate (63) with hydroiodic acid or sodium sulfite results in the reduction of the azido group to the amino group.61 At the same time, 1,4,5- trialkyl(aryl)tetrazolium salts are reduced to trisubstituted di- hydrotetrazoles 111 with sodium borohydride in ethanol.These salts undergo alkylation in reactions with organolithium com- pounds and vinylmagnesium bromide to give tetrasubstituted dihydrotetrazoles 112.65 S V Voitekhovich, P N Gaponik, O A Ivashkevich R2 R1 N NaBH4 N R3 N N R2 111 (45% ± 90%) + R1 N N R3 R4 R2 N N X7 R1 R4M N N R3 N N 112 (25% ± 80%) R1=Me, Et, Oh; R2=Ph, Me, cyclo-C6H11, But; R3=Me, Et, Bn, Ph; CH;M=Li, MgBr. R4=Me, Ph, PhC C, CH2 5. Thermolysis The published data on the thermal stability of tetrazolium salts are rather scarce. It was shown 6, 14 that of 1,3- and 1,4-dimethyl-5-R- tetrazolium iodides, the 1,4,5-isomers are more stable.Their decomposition begins at temperatures above 265 8C (R=Me) and 130 8C (R=Ph), whereas that of 1,3,5-isomers occurs even at 130 and 70 8C, respectively. N-Demethylation is the main process of thermolysis leading to the formation of disubstituted tetra- zoles.6, 14 Thermolysis of 1,3,5-trisubstituted derivative 67 is accompanied by the evolution of nitrogen and the formation of the salt 113 in 94% yield.61 Ph Ph BF¡4 BF¡4N N N N 180 8C N PPh3 N N N PPh3 + + 7N2 N N N N Ph 113 67 Ph Recently, the thermal stabilities of 1-methyl-4-phenyl-, 1,4,5- trimethyl-, 1,3-di-tert-butyl-5-methyl- and 1,3-di-tert-butyltetra- zolium picrates have been studied using differential scanning calorimetry and complex thermal analysis (heating rate 5 8C per min in a nitrogen atmosphere).33 Among those, 1,4,5-trimethyl- tetrazolium picrate [initial temperature of intense decomposition (Td) 215 8C] manifested the highest stability.1-Methyl-4-phenyl- tetrazolium picrate (Td=100 8C) appeared to be the least stable, which may be attributed to the easy abstraction of proton from the carbon atom of the heterocycle. Special mention should be made of the enhanced thermal resistance of poly-N,N0-dimethyl-5- vinyltetrazolium perchlorate in comparison with poly-N-methyl- 5-vinyltetrazole. According to complex thermal analysis data, the decomposition of polymeric tetrazolium perchlorate is observed at temperatures above 275 8C, whereas the non-quaternised polymer is decomposed at 230 8C.108 2-Methyl-5-phenyltetrazole is the main reaction product in the attempted synthesis of tetrazolium salts by quaternisation of 1-methyl-5-phenyltetrazole with methyl iodide at 130 8C.6 The resulting isomerisation is due to the specificity of thermolysis of the initially formed tetrazolium salt 114; its decomposition yields predominantly 2,5-disubstituted tetrazole.6 Ph Ph Ph MeI Me Me N N N N N N 130 8C, 10 h + 7MeI I7 N N N N N N Me 114 Me The formation of tetrazolium complexes is usually the key step in the isomerisation of N-substituted tetrazoles, which occurs under the action of alkylating reagents 6, 109 or acids.43, 55, 110, 111 6.Photolysis The photochemical conversions of some azido-, azo- and triazo- 1,3-diaryltetrazolium derivatives have been studied recent- ly.112 ± 114 It was found 112, 113 that the tetrazole rings of the1,3- and 1,4-Substituted tetrazolium salts compounds under study are resistant to UV irradiation.Among other conversions, only the reactions of 5-azido-1,3-diaryltetra- zolium salts, which yield hardly accessible tricyclic mesoionic compounds 115 and 116, deserve special mention. R1 1) hn, MeCN or MeOH 2) NaOH, H2O R2 R2 N N N3 + N N BF¡4Ph R1=H, Me; R2=H N + R1 N Ph 7N NN115 (20% ± 35%) Me N Me R1 =R2=Me N 7 Me N N + N N Ph 116 (29%) UV Irradiation of 1,3-diphenyl-5-phenyltetrazolium tetra- fluoroborate is accompanied by the evolution of nitrogen and the formation of 1,3-diphenyltetrazolium tetrafluoroborate and benzene.114 Ph Ph N N hn, MeOH N N N NPh + + N N N BF¡ N 4 BF¡4 7PhH 7N2 Ph Ph V.Applications of tetrazolium salts Until recently, 1,3- and 1,4-substituted tetrazolium salts have not been sufficiently investigated due to their unavailability and data concerning their applications are virtually absent from the liter- ature. However, the development of procedures for their synthesis have given strong impetus to their investigations, particularly in those areas where 2,3-substituted tetrazolium salts (e.g., phase- transfer catalysts, reagents for the separation of metals, etc.) were conventionally used before. There is evidence of successful appli- cations of 3-tert-butyl-1,5-dimethyltetrazolium perchlorate (117) as a phase-transfer catalyst in oxidation reactions catalysed by the permanganate ion.115 Me Me N ClO¡4 N+N N But 117 It is of note that 1,3-substituted tetrazolium salts have a number of advantages over 2,3-derivatives in such reactions.116 First, these methods allow one to obtain 1,3,5-trisubstituted tetrazolium salts with various types of substituents at the carbon and nitrogen atoms of the tetrazole ring and thus to vary their physicochemical properties (e.g., thermal stability, solubility, etc.).The methods for the synthesis of 2,3-disubstituted tetra- zolium salts exclude the introduction of alkyl substituents to the nitrogen atoms of the tetrazole ring, which limits the array of available derivatives.1 ±3 Of no less importance is the relative stability of 1,3-substituted tetrazolium salts in reductive media; 737 therefore, they can be used in phase-transfer reduction reactions, whereas 2,3-substituted derivatives are converted into formazanes even under the action of weak reducing agents.1±3 Recently, a series of poly-5-vinyltetrazolium salts 118 prepared by exhaustive methylation of poly-5-vinyltetrazole have been examined as sorbents for the extraction of palladium ions from various solutions in the form of PdCl2¡ 4 .These include an efficient adsorbent for the selective extraction of palladium ions from spent palladation electrolytes.108 It was found that 1,4-dimethyl- 5-[(2-oxopropyl)thio]tetrazolium triflate (119) inhibits transglu- taminase,117 and 5-(2-anilinovinyl)-1-(4-chlorophenyl)-4-ethyl- tetrazolium iodide (120) manifests antiarrhythmic activity.118 CH2 CH2 SCH2COMe + Me Me Me X7 N N N N + Me N N CF3SO¡3 N N n 119 118 X=BF¡4 , ClO¡4 , MeSO¡4 NHPh CH CH+ Et 4-ClC6H4 N N I7 N N 120 The addition of 1-aryl- and 1-cyclohexyl-4,5-dimethyltetrazo- lium chlorides 121 to photographic dyes decreases the light sensitivity of photomaterials and enhances image sharpness.119 1,4-Bis(2-decanoylaminoethyl)tetrazolium methylsulfate and per- chlorate 122 have been patented as corrosion inhibitors and disinfectants.120 Tetrazolo[1,5-a]pyridinium salts 123, 124 have found application as sensitisers in the recording layers of electro- photographic materials.56, 121 (CH2)2NHC(O)C9H19 Me N N N N X7 Cl7 Me + + N N N N (CH2)2NHC(O)C9H19 122 121 R X=ClO4, MeSO4 R=cyclo-C6H11, 2-ClC6H4, 3,4-Me2C6H3 CH CHC6H4NMe2-4 CH CHC6H4NMe2-4 + R3 N N + R3 N N NN N R1 ClO¡4 ClO¡4N R2 N N R2 R1 123 124 R1=Me, Bn; R2=H, Me, Ph, CH=CHC6H4NMe2-4; R3=H,Me R1=Me, CH2CH=CH2, Ph; R2=H, Me, Ph, CH=CHC6H4NMe2-4; R3=H,Me However, 1,3- and 1,4-substituted tetrazolium salts present the greatest interest as starting materials in the synthesis of various (including hardly accessible) heterocyclic systems. Condensation reactions of these salts are employed in the synthesis of annelated heterocycles.These reactions are specific and often the only way to the synthesis of imidazo- 35, 103 and pyrrolotetrazoles.37 It was shown that recyclisation of 1,4-substituted tetrazolium salts is a convenient procedure for the synthesis of three-, five- and six- membered nitrogenous heterocycles, such as iminoaziridines and tetrazines,66, 102 imidazolones,103 benzooxazoles and quinazoline- diones.47, 100 The conversions of substituents at the carbon atom of the heterocycles of 1,3,5- and 1,4,5-trisubstituted tetrazolium738 salts can be used in the high-yield synthesis of mesoionic tetra- zoles.21, 27, 60, 94, 95, 112 ± 114 The use of tetrazolium salts for the generation of N-heteroaromatic carbenes,44, 59, 122 which occupy a special place in the chemistry of carbenes by virtue of their peculiar structures and reactivities, especially holds great prom- ise.83, 122 VI.Conclusion Significant progress has been achieved in the synthesis and investigation of physicochemical properties of 1,3-(1,3,5)- and 1,4-(1,4,5)-substituted tetrazolium salts in the past 10 ± 15 years. The interest in these studies is largely due to the development of methods for selective functionalisation of the tetrazole ring, which permits direct introduction of substituents on the carbon and nitrogen atoms of the heterocycle, and the recently discovered isomerisation of 1,3-disubstituted tetrazolium salts into the cor- responding 1,4-disubstituted derivatives. In turn, the availability of tetrazolium salts has stimulated the search for novel areas of their applications, particularly for synthetic purposes.Recyclisa- tion and condensation reactions of tetrazolium salts and trans- formations of substituents at the carbon atom of the heterocycle proved to be useful tools in the synthesis of various organic compounds. Elucidation of the synthetic potentials of tetrazolium salts presents considerable interest for further progress in prepa- rative chemistry of nitrogen-containing heterocyclic systems and organic chemistry on the whole. In-depth studies into the proper- ties of novel 1,3-(1,3,5)- and 1,4-(1,4,5)-substituted tetrazolium salts will make it possible to extend the range of their practical applications, e.g., the design of polynitrogenous ionogenic poly- mers and ionic fluids resembling those of other azolium salts and possessing unique properties.Of no less importance are further studies into tetrazole quaternisation and possible involvement of functionalised C- and N-tetrazole derivatives and various alkylat- ing reagents in this reaction. The knowledge of regularities under- lying the formation of tetrazolium cations will enhance the potentialities of quaternisation and protonation reactions for the activation of the tetrazole ring in the substitution reactions occurring at the endocyclic carbon atom. References 1. A W Nineham Chem. Rev. 55 355 (1955) 2. W D Hooper Rev. Pure Appl. Chem. 19 221 (1969) 3. A B Zhivich, G I Koldobskii, V A Ostrovskii Khim. Geterotsikl. Soedin. 1587 (1990) a 4.A K Pikaev, Z K Kriminskaya Usp. Khim. 67 745 (1998) [Russ. Chem. Rev. 67 671 (1998)] 5. D S Daniel Chem. Appl. Leuco Dyes 207 (1997); Chem. Abstr. 127 163 093 (1997) 6. R N Butler, in Comprehensive Heterocyclic Chemistry II Vol. 4 (Eds A R Katritzky, C W Rees, E F V Scriven) (Oxford; New York: Pergamon Press, 1996) p. 621 7. H R Meier, H Heimgartner, in Methoden der Organischen Chemie (Houben-Weyl) Bd. E8d (Stuttgart: Georg Thieme, 1994) p. 664 8. G I Koldobskii, V A Ostrovskii Usp. Khim. 63 847 (1994) [Russ. Chem. Rev. 63 797 (1994)] 9. D Moderhack J. Prakt. Chem.-Chem. Ztg. 340 687 (1998) 10. P N Gaponik, in Khimicheskie Problemy Sozdaniya Novykh Materi- alov i Tekhnologii (Chemical Problems of Creation of New Materials and Technologies) (Ed.V V Sviridov) (Minsk: Belorussian University, 1998) p. 185 11. V A Ostrovskii, A O Koren Heterocycles 53 1421 (2000) 12. C G Newton, C A Ramsden Tetrahedron 38 2965 (1982) 13. O M Zimmerman, R A Olofson Tetrahedron Lett. 3453 (1970) 14. T Isida, S Kozima, S-i Fujimori, K Sisido Bull. Chem. Soc. Jpn. 45 1471 (1972) 15. A KoÈ nnecke, E Lippmann, E Kleinpeter Tetrahedron 33 1399 (1977) 16. R Faure, J P Galy, E J Vincent, J Elguero Magn. Res. Chem. 11 507 (1978) 17. H Quast, L Bieber Chem. Ber. 114 3253 (1981) S V Voitekhovich, P N Gaponik, O A Ivashkevich 18. V V Semenov, V S Bogdanov, B S El'yanov, L G Mel'nikova, S A Shevelev, V M Zhulin, A A Fainzil'berg Khim. Geterotsikl. Soedin. 1118 (1982) a 19. H Quast, U Nahr Chem.Ber. 116 3427 (1983) 20. A Messmer, G Hajos, J Fleischer, M Czugler Monatsch. Chem. 116 1227 (1985) 21. DModerhack, A Lembcke J. Chem. Soc., Perkin Trans. 1 1157 (1986) 22. H Quast, L Bieber, G Meichner Liebigs Ann. Chem. 469 (1987) 23. I Yu Shirobokov,M V Chekushina, V A Ostrovskii, G I Koldobskii, G B Erusalimskii Khim. Geterotsikl. Soedin. 502 (1988) a 24. A Messmer, G Hajos, Z Juhasz-Riedl, P Sohar J. Org. Chem. 53 973 (1988) 25. G W Fischer J. Prakt. Chem. 331 885 (1989) 26. M Begtrup, P Larsen Acta Chem. Scand., Ser. B 44 1050 (1990) 27. D Moderhack, D-O Bode J. Chem. Soc., Perkin Trans. 1 1483 (1992) 28. P N Gaponik, O A Ivashkevich, V N Naumenko, T B Kovalyova, T N Andreeva Spectrochim. Acta, Part A 49 135 (1993) 29. RNButler, J P Duffy, E P Nibhradaigh, P McArdile,DCunningham J.Chem. Res. (S) 216 (1994) 30. P N Gaponik, Yu V Grigor'ev, T N Andreeva, I I Maruda Khim. Geterotsikl. Soedin. 915 (1995) a 31. H Quast, J Balthasar, A Fuss, U Nahr,W Nudling Liebigs Ann. Chem. 671 (1997) 32. P N Gaponik, V P Karavai, Yu V Grigor'ev Dokl. Akad. Nauk Bel. 41 (5) 66 (1997) 33. P N Gaponik, S V Voitekhovich, O A Ivashkevich, A S Lyakhov, A A Govorova Khim. Geterotsikl. Soedin. 657 (1998) a 34. P Cmoch, J W Wiench, L Stefaniak, J Sitkowski J. Mol. Struct. 477 119 (1999) 35. D Moderhack, B Holtmann J. Prakt. Chem.-Chem. Ztg. 342 591 (2000) 36. P N Gaponik, S V Voitekhovich, A S Lyakhov Khim. Geterotsikl. Soedin. 387 (2000) a 37. D Moderhack,D Decker, B Holtmann J. Chem.Soc., Perkin Trans 1 720 (2001) 38. A F Pozharskii Teoreticheskie Osnovy Khimii Geterotsiklov (Theoretical Foundations of Heterocyclic Chemistry) (Moscow: Khimiya, 1985) 39. I P Gol'dshtein, L D Khamaganova, A N Fedotov, E S Domnina Zh. Fiz. Khim. 65 959 (1991) b 40. V N Naumenko, A O Koren, P N Gaponik Magn. Res. Chem. 30 558 (1992) 41. V N Naumenko, P N Gaponik, A O Koren', M M Degtyarik Izv. Akad. Nauk Bel., Ser. Khim. Nauk (1) 64 (1993) 42. V Yu Zubarev, V A Ostrovskii Khim. Geterotsikl. Soedin. 867 (2000) a 43. V V Saraev, E L Golod Zh. Org. Khim. 33 629 (1997) c 44. V I Boev, E M Krasnikova, A I Moskalenko, E I Pil'ko, L V Snegur, V N Babin, Yu S Nekrasov Zh. Obshch. Khim. 67 1386 (1997) d 45. P N Gaponik, S V Voitekhovich, I I Maruda, A A Kulak, O A Ivashkevich Pol.J. Chem. 72 2247 (1998) 46. P N Gaponik, S V Voitekhovich, A S Lyakhov, I I Maruda Khim. Geterotsikl. Soedin. 1222 (1999) a 47. S V Voitekhovich, P N Gaponik, A S Lyakhov, O A Ivashkevich Pol. J. Chem. 75 253 (2001) 48. S V Voitekhovich, P N Gaponik, A S Lyakhov, O A Ivashkevich Khim. Geterotsikl. Soedin. 1035 (2001) a 49. S V Voitekhovich, P N Gaponik, O A Ivashkevich Izv. Nat. Akad. Nauk Bel., Ser. Khim. Nauk (1) 62 (2002) 50. USSR P. 1 544 772; Byull. Izobret. (17) 3 (1990) 51. A O Koren', P N Gaponik Khim. Geterotsikl. Soedin. 1280 (1991) a 52. S V Voitekhovich, P N Gaponik, A O Koren Mendeleev Commun. 41 (1997) 53. A O Koren, P N Gaponik, V A Ostrovskii Int. J. Chem. Kinet. 25 1043 (1993) 54.A O Koren, P N Gaponik, V A Ostrovskii Int. J. Chem. Kinet. 27 919 (1995) 55. P N Gaponik, S V Voitekhovich Zh. Org. Khim. 34 788 (1998) c 56. I P Bachkovskii, V A Chuiguk Ukr. Khim. Zh. 53 319 (1987) 57. H Alper, R W Stout J. Heterocycl. Chem. 10 569 (1973) 58. G N Borisova, A V Borisov, I V Bodrikov, V K Bel'skii, A I Lutsenko, V A Smit, G A Kutyrev Zh. Org. Khim. 30 760 (1994) c1,3- and 1,4-Substituted tetrazolium salts 59. A V Borisov, V K Bel'skii, G N Borisova, V K Osmanov, Zh V Matsulevich, T V Goncharova Khim. Geterotsikl. Soedin. 763 (2001) a 60. S Araki, Y Wanibe, F Uno, A Morikawa, K Yamamoto, K Chiba, Y Butsugan Chem. Ber. 126 1149 (1993) 61. S Araki, K Yamamoto,M Yagi, T Inoe, H Fukagawa, H Hattori, H Yamamura,M Kawai, Y Butsugan Eur. J.Org. Chem. 121 (1998) 62. H Quast, A Fuss, U Nahr Chem. Ber. 118 2164 (1985) 63. D Moderhack, D Decker, B Holtmann J. Chem. Soc., Perkin Trans. 1 729 (2001) 64. S Araki, J Mizuya, Y Butsugan J. Chem. Soc., Perkin Trans. 1 2439 (1985) 65. B Carbony, R Carrie Tetrahedron 40 4115 (1984) 66. H Quast, T Hergenrother Liebigs Ann. Chem. 581 (1992) 67. W Wirschun,M Winkler,K Lutz, J C Jochims J. Chem. Soc., Perkin Trans. 1 1755 (1998) 68. W Wirschun J. Prakt. Chem.-Chem. Ztg. 340 300 (1998) 69. K R Seddon Kinet. Katal. 37 743 (1996) e 70. P Wasserscheid, W Keim Angew. Chem. 112 3926 (2000) 71. P N Gaponik, O A Ivashkevich Izbrannye Nauchnye Trudy BGU, Khimiya (Selected Proceedings of Belorussian State University) Vol.5 (Minsk: Belorussian State University, 2001) p. 353 72. P N Gaponik, Doctoral Thesis in Chemical Sciences, Research Institute for Physical Chemical Problems of Belorussian State University, Minsk, 2000 73. W Bocian, J Jazwinski, W Kozminski, L Stefaniak, G A Webb J. Chem. Soc., Perkin Trans. 2 1327 (1994) 74. W Bocian, J Jazwinski, W Kozminski, L Stefaniak, G A Webb Magn. Res. Chem. 32 284 (1994) 75. J Jazwinski, W Kozminski, L Stefaniak, G A Webb Magn. Res. Chem. 32 746 (1994) 76. J Jazwinski, O Staszewska, L Stefaniak, S Araki, G A Webb J. Mol. Struct. 523 103 (2000) 77. J Jazwinski Bull. Pol. Acad. Sci., Chem. 48 35 (2000) 78. R M Claramunt, D Sanz, G Boyer, J Catalan, J L G de Paz, J Elguero Magn. Res. Chem. 31 791 (1993) 79.R M Claramunt, D Sanz, C Lopez, J A Jimenez, M L Jimeno, J Elguero, A Fruchier Magn. Res. Chem. 35 35 (1997) 80. G I Koldobskii, V A Ostrovskii, V S Poplavskii Khim. Geterotsikl. Soedin. 1299 (1981) a 81. R N Butler, in Comprehensive Heterocyclic Chemistry Vol. 5 (Eds A R Katritzky, C W Rees) (Oxford; New York: Pergamon Press, 1984) p. 791 82. M R Grimmett, in Comprehensive Organic Chemistry Vol. 4 (Eds D Barton, W D Ollis) (Oxford: Pergamon Press, 1979) 83. O P Shvaika, N I Korotkikh, A F Aslanov Khim. Geterotsikl. Soedin. 1155 (1992) a 84. P N Gaponik, V N Naumenko, Yu V Grigor'ev, T A Madzievskaya Vest. Belorus. Univ., Ser. 2 (1) 9 (1995) 85. L I Belen'kii, L D Chuvylkin Khim. Geterotsikl. Soedin. 1535 (1996) a 86. V P Karavai, P N Gaponik Khim. Geterotsikl. Soedin. 564 (1985) a 87. V P Karavai, P N Gaponik Khim. Geterotsikl. Soedin. 66 (1991) a 88. P N Gaponik, Yu V Grigor'ev, A O Koren' Khim. Geterotsikl. Soedin. 1699 (1988) a 89. Yu V Grigor'ev, P N Gaponik Izv. Akad. Nauk. Bel., Ser. Khim. Nauk (1) 73 (1992) 90. R H Lowack, R Weiss J. Am. Chem. Soc. 112 333 (1990) 91. R Weiss, R H Lowack Angew. Chem. 103 1183 (1991) 92. R N Hanley,W D Ollis, Ch A Ramsden J. Chem. Soc., Perkin Trans. 1 741 (1979) 93. R N Hanley,W D Ollis, Ch A Ramsden, I S Smith J. Chem. Soc., Perkin Trans. 1 744 (1979) 94. S Araki, Y Butsugan J. Chem. Soc., Chem. Commun. 789 (1983) 95. S Araki, Y Butsugan J. Chem. Soc., Perkin Trans. 1 2545 (1984) 96. J Jazwinski Pol. J. Chem. 73 1719 (1999) 97. S Araki, K Yamamoto, T Inoe, K Fujimoto, H Yamamura, M Kawai, Y Butsugan, J-K Zhou, E Eichhorn, A Rieker, M Huber J. Chem. Soc., Perkin Trans. 2 985 (1999) 98. G Scherowsky, I Pickardt Chem. Ber. 116 186 (1983) 99. M Mikolajczyk, P Kielbasinski Tetrahedron 37 233 (1981) 739 100. S V Voitekhovich, P N Gaponik Khim. Geterotsikl. Soedin. 1141 (1997) a 101. E G Kovalev, V A Anufriev, G L Rusinov Khim. Geterotsikl. Soedin. 1691 (1990) a 102. H Quast, A Fuss,W Nudling Eur. J. Org. Chem. 317 (1998) 103. D Moderhack, A Lembcke Chem.-Zeit. 109 432 (1985) 104. D Moderhack, D-O Bode, D Shomburg Chem. Ber. 126 129 (1993) 105. O My, J L G de Paz,M Yanez J. Phys. Chem. 90 5597 (1986) 106. A O Koren, P N Gaponik, O A Ivashkevich, T B Kovalyova Mendeleev Commun. 10 (1995) 107. V V Semenov, S A Shevelev, L G Mel'nikova Mendeleev Commun. 58 (1993) 108. S V Voitekhovich, P N Gaponik, O A Ivashkevich, in Vtoroi Vserossiiskii Karginskii Simpozium `Khimiya i Fizika Polimerov v Nachale XXI Veka' (Tez. Dokl.), Chernogolovka, 2000 [The Second Kargin All-Russian Symposium `Chemistry and Physics of Polymers in the Beginning of the XXIth Century' (Abstracts of Reports), Chernogolovka, 2000] Vol. 1, p. 1 109. A I Podgurskii, S G Zlotin, M M Krayushkin, O A Luk'yanov Izv. Akad. Nauk SSSR, Ser. Khim. 236 (1985) f 110. A V Sachivko, V P Tverdokhlebov, I V Tselinskii Zh. Org. Khim. 22 1112 (1986) c 111. S V Voitekhovich, P N Gaponik Izv. Akad. Nauk. Bel., Ser. Khim. Nauk 68 (2) (2000) 112. S Araki, H Hattori, N Shimizi, K Ogawa, H Yamamura, M Kawai J. Heterocycl. Chem. 36 863 (1999) 113. S Araki, H Hattori, H Yamamura,M Kawai J. Heterocycl. Chem. 37 1129 (2000) 114. S Araki, H Hattori, K Ogawa,M Kuzuya, T Inoe, H Yamamura, M Kawai J. Chem. Soc., Perkin Trans. 1 2476 (2001) 115. Yu V Grigor'ev, Author's Abstracts of Candidate Thesis in Chemical Sciences, St Petersburg State Technical Institute (Technical University), St-Petersburg, 1998 116. G I Koldobskii, A B Zhivich, V A Ostrovskii Zh. Obshch. Khim. 62 3 (1992) d 117. Eur. P. 411 894; Chem. Abstr. 115 29 338 (1991) 118. US P. 5 183 813; Chem. Abstr. 119 108 996 (1993) 119. Jpn. P. 05 27 368; Chem. Abstr. 119 59 638 (1993) 120. BRD P. 4 218 585; Chem. Abstr. 120 107 029 (1994) 121. USSR P. 934 443; Byull. Izobret. (21) 220 (1982) 122. W A Herrmann, C KoÈ cher Angew. Chem., Int. Ed. Engl. 36 2162 (1997) a�Chem. Heterocycl. Compd. (Engl. Transl.) b�Russ. J. Phys. Chem. (Engl. Transl.) c�Russ. J. Org. Chem. (Engl. Transl.) d�Russ. J. Gen. Chem. (Engl. Transl.) e�Kinet. Catal. (Engl. Transl.) f�Russ. Chem. Bull., Int. Ed. (En

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (9) 741 ± 760 (2002) Supramolecular chemistry of cucurbiturils O A Gerasko, D G Samsonenko, V P Fedin Contents I. Introduction II. Host ± guest cucurbit[6]uril compounds III. Supramolecular compounds of cucurbit[6]uril with metal aqua complexes IV. Supramolecular compounds of cucurbit[6]uril with metal cluster aqua complexes V. Cucurbit[n]urils (n=5, 7 ± 10) Abstract. supramolecular of design the of principles main The The main principles of the design of supramolecular compounds starting from organic macrocyclic cavitands, compounds starting from organic macrocyclic cavitands, viz., cucurbit[ The considered. are 10), ± 5 [C n]urils ]urils [C6H6N4O2]n (n=5 ± 10), are considered. The presence of the hydrophobic inner cavity along with polar presence of the hydrophobic inner cavity along with polar carbonyl groups forming the cucurbituril portals are responsible carbonyl groups forming the cucurbituril portals are responsible for high specificity of the formation of host ± guest complexes.The for high specificity of the formation of host ± guest complexes. The unique ability of cucurbiturils to act as synthetic molecular unique ability of cucurbiturils to act as synthetic molecular containers is discussed. In these containers, bimolecular reactions containers is discussed. In these containers, bimolecular reactions between specially selected guests proceed with high regio- and between specially selected guests proceed with high regio- and stereoselectivity. The review surveys new data on the directed stereoselectivity.The review surveys new data on the directed construction inorganic ± organic supramolecular of construction of supramolecular organic ± inorganic compounds compounds through hydrogen of network extensive an of formation the through the formation of an extensive network of hydrogen bonds bonds between and cucurbiturils of atoms oxygen portal the between the portal oxygen atoms of cucurbiturils and water water molecules bibliography The complexes. aqua metal of molecules of metal aqua complexes. The bibliography includes includes 108 references 108 references. I. Introduction Supramolecular chemistry is a new extensively evolving area of research focused on the association of individual molecules giving rise to more complex chemical systems formed through intermo- lecular (non-covalent) interactions.1± 3 Supramolecular chemistry concerns various phenomena in different fields of science (organic, coordination and physical chemistry, biology, physics) and has a broad spectrum of possibilities due to the mutual enrichment of these areas of scientific exploration.Supramolecu- lar chemistry deals with supramolecular ensembles, which are spontaneously self-assembled from complementary (i.e., charac- terised by geometric and chemical compatibility) fragments in much the same fashion as spontaneous self-assembly of very complicated spatial structures in leaving cells. The directed con- struction of such systems and the design of highly ordered supra- molecular compounds with desired structures and properties from molecular building blocks are fundamental problems of modern chemistry.O A Gerasko, D G Samsonenko, V P Fedin Institute of Inorganic Chemistry, Siberian Branch of the Russian Academy of Sciences, ul. Acad. Lavrent'eva 3, 630090 Novosibirsk, Russian Federation. Fax (7-383) 234 44 89. Tel. (7-383) 234 42 53. E-mail: olager@che.nsk.su (O A Gerasko), denis@che.nsk.su (D G Samsonenko). Tel. (7-383) 234 42 53. E-mail: cluster@che.nsk.su (V P Fedin) Received 25 June 2002 Uspekhi Khimii 71 (9) 840 ± 861 (2002); translated by T N Safonova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n09ABEH000748 Intermolecular interactions holding the fragments together, viz., Coulombic interactions, hydrogen bonds and van der Waals interactions, are, on the whole, much weaker than the covalent bonds in the molecule itself.To increase the strength of these bonds, their directionality and specificity, it is necessary to use rather large molecular building blocks capable of creating an extensive network of bonds due to a large contact interface. Examples of compounds with a large number of bonded non- covalent interactions are host ± guest complexes in which the inner cavity of the host molecule (or cavitand) 4 can include a guest molecule of suitable size. Numerous complexes containing such macrocyclic organic molecules as crown ethers (1), cryptands (2), calixarenes (3) or cyclodextrins (4) as the host are presently available. The encapsulation of inert gases, metal ions or organic molecules in the cavities of these molecules gives rise to stable compounds having unusual structures and exhibiting interesting properties.The present review surveys the synthesis and structures of supramolecular compounds prepared with the use of macro- cyclic cavitands having the trivial name cucurbiturils. O O O N O O O1 R R OH OHHO OH HO OH R 3 R O O O N O O O 2 RR 741 742 746 751 756742 HO O OH O HO O OH OH O OH O O HO OH HO OH O OH HOO OH OH O HO O OHO HO O OH 4 Cucurbituril with composition C36H36N24O12 [cucurbit[6]uril (5)] consisting of six methylene-linked glycoluril fragments has been studied the most.This compound is the first example of cucurbiturils. It was prepared by Behrend et al.5 in 1905 by condensation of formaldehyde with glycoluril (condensation product of urea and glyoxal) in acidic medium. At that time, however, the procedures available did not allow one to determine correctly its composition and structure. O O O O O O O NH HN N N N N N N N N NNN N HCl, H2O +HCHO H2SO4, D NNN N NH HN N NN N N N O NN O O O O O O 5 In 1981, Freeman, Mock and Shih 6 reproduced the synthesis developed by Behrend and obtained a colourless crystalline compound whose structure was established by X-ray diffraction analysis. Cucurbituril 5 is a barrel-shaped macrocyclic cavitand containing carbonyl oxygen atoms (portals) at the top and bottom of the barrel.This cavitand received its trivial name for the visual similarity of its molecular shape to Cucurbitaceae (pumpkin family). According to the nomenclature of Chemical Abstracts, the systematic name of this compound is dodecahydro- 1H,4H,14H,17H-2,16 : 3,15-dimethano-5H, 6H, 7H,8H,9H,10H,- 11H,12H,13H,18H,19H,20H,21H,22H,23H,24H,25H,26H-2,3,- 4a,5a,6a,7a,8a,9a,10a,11a,12a,13a,15,16,17a,18a,19a, 20a, 21a,- 22a,23a,24a,25a,26a-tetraoxaazabispentaleno [1000,6000 : 500,600,700] - cycloocta(1,2,3-cd : 5,6,7-c 0d 0)dipentaleno-1,4,6,8,10,12,14,17,- 19,21,23,25-dodecane. Hereinafter, cucurbiturils consisting of n glycoluril fragments will be abbreviated to Qn or referred to as cucurbit[n]uril.The dimensions of the inner cavity of cucurbit[6]uril (5) (the height is 6 Aand the inner diameter is 5.5 A) allow the molecule to accommodate small guest organic molecules or ions and the portals formed by the carbonyl groups (the diameters of the portals are 4 A) can bind cations. The structure determination of Q6 gave impetus to investigations of its ability to act as a macro- cyclic cavitand. Cucurbit[6]uril (5) whose cavity is similar in size to those of a-cyclodextrin and 18-crown-6 has higher negative charges on the donor oxygen atoms resulting in the enhancement of stability of its adducts with positively charged ions.7, 8 Yet another difference between cucurbituril Q6 and other cavitands is its structural rigidity.Thus, this cucurbituril virtually retains its shape upon the inclusion of different guests and, consequently, exhibits higher selectivity in the formation of inclusion com- pounds.7 Cucurbit[6]uril (5) is a colourless crystalline compound, which is insoluble in water and organic solvents but is readily soluble in some mineral acids (HCl, H2SO4, CF3 SO3H), carboxylic acids O A Gerasko, D G Samsonenko, V P Fedin (for example, HCOOH) and aqueous solutions of alkali salts. Cucurbit[6]uril serves as a convenient starting compound for the preparation of various supramolecular compounds due to its unique structure, the simplicity of the synthesis procedure and thermal stability (it does not decompose upon heating to 400 8C; see Ref.9). Presently, cucurbit[6]uril is a readily available reagent (in particular, it can be purchased from Merck). The results of studies devoted to the synthesis, examination of structures and properties of supramolecular compounds of cucur- biturils and investigations of thermodynamic and kinetic aspects of formation of these systems are summarised in reviews.7, 10 ± 13 The present review surveys the recent advances in the chemistry of cucurbiturils achieved, primarily, in the latter half of the 1990s. Considerable attention is given to the studies aimed at the directed construction of supramolecular organic ± inorganic hybrid mate- rials, which have been performed in the last five years. II. Host ± guest cucurbit[6]uril compounds The presence of the rather rigid inner cavity is responsible for the ability of cucurbit[6]uril (5) to accommodate small guest mole- cules. The formation of inclusion compounds was established by crystallographic methods as well as by different physicochemical methods because absorption, fluorescence and NMR spectra of guests are changed as the guest molecules go from the solvent environment (generally, from weakly acidic aqueous solutions) to the non-polar cavity of cucurbituril.Cucurbit[6]uril (5) forms stable inclusion compounds with amines, diamines, alkylammo- nium ions, benzylammonium ions and dye molecules. 1. Inclusion compounds with alkylammonium ions Inclusion compounds of cucurbit[6]uril (5) with alkylammonium ions have received the most study.The hydrocarbon radical of the alkylammonium cation is encapsulated in the cavity of Q6 displacing the solvent molecules and is held in this cavity through van der Waals forces, whereas the nitrogen atom is fixed at the portal through Coulombic interactions of the cation with the oxygen atoms of the carbonyl groups. As a result, high specificity of binding of alkylammonium ions to cucurbit[6]uril is achieved. R R N+H2 NH +2 X-Ray diffraction analysis of the supramolecular compound of Q6 with p-xylylenediammonium chloride carried out by Free- man 14 in 1984 showed that the hydrocarbon fragment of the diammonium cation H3N+CH2C6H4CH2N+H3 is located in the cavity, whereas two protonated nitrogen atoms are located at the portals of the macrocycle.The formation of host ± guest com- pounds with composition 1 : 1 was also evidenced by X-ray diffraction data and solution NMR spectroscopic studies of supramolecular compounds of cucurbit[6]uril with alkyldiamines containing four, five or six methylene units between the nitrogen atoms.14 ± 16 The incorporation of alkylammonium cations into cucurbit- [6]uril was examined by solution NMR spectroscopy in the 1980 ± 1990s.15 ± 20 It was found that the most stable inclusion compounds with cations of primary amines H(CH2)nNHá3 and diamines H3N+(CH2)nNHá3 are formed at n=4 and n=5, 6, respectively. An increase in the length of the hydrocarbon chain either hinders coordination of both amino groups of the alkyldi- amine to the portals of cucurbit[6]uril or causes the hydrophobic alkyl fragment of the amine to interact with the second hydro- philic portals.A decrease in the length of the hydrocarbon chain leads to a decrease in the number of van der Waals contacts as a result of which two ammonium groups cannot be simultaneouslySupramolecular chemistry of cucurbiturils logK 7.0 6.0 123 5.0 4.0 3.0 2.0 1.0 0 2 4 6 8 n Figure 1. Logarithm of the formation constant (K) of inclusion com- pounds of cucurbit[6]uril with amines vs. the number of CH2 groups in the chain; (1) alkylamines, (2) alkydiamines, (3) amino alcohols. located at the portals. The plots of the formation constants of inclusion compounds with composition 1 : 1 versus the chain length for alkylamines,17 alkyldiamines 17 and amino alcohols 18 are shown in Fig.1. In the case of amino acids, no influence of the number of methylene groups on the stability constants was observed.18 The PhNH(CH2)6NH(CH2)4NH2 ligand containing three amino groups was specially synthesised and its binding with cucurbit[6]uril (5) was studied.19 An important characteristic feature of this guest molecule is the presence of two nitrogen atoms involved in the diaminoalkyl fragment. Their basicity is 106 times as high as that of the nitrogen atom bound to the phenyl group. In addition, the formation constant of the supramolecular compound of the hexanediammonium cation with cucurbit[6]uril is approximately 100 times larger than that for the butanediam- monium ion, i.e., different fragments of the triamine behave differently in binding to Q6.In aqueous solutions, the constant pKa of the nitrogen atom of the aniline fragment in this triamine is 4.7 and it is increased to 6.7 in the presence of stoichiometric amounts of Q6. The behaviour of each region of the triamine depends on pH. In acidic solutions, the hexanediammonium fragment (more complementary to the cavity in size) is located inside the cavitand. By contrast, the butanediammonium end of the triamine is bound to Q6 at pH>6.7 because the nitrogen atom of the aniline fragment is deprotonated. This system is a molecular `switch', i.e., Q6 can be moved along the triamine molecule by changing the pH.H H H H H H H H H H H H H H H pH>6.7 + N H + +N H N pH<6.7 H H H H H H H H H H H H H H H H H H H H H H H H H H H N H N+ H H N +H H H H H H H H H H H H H H H H NMR spectroscopic studies demonstrated that the rate of formation of adducts with cucurbit[6]uril (5) depends on the diameter of the ammonium molecules and it is independent of thermodynamic stability of the resulting complexes.20 Linear 743 hydrocarbon chains with a diameter that is no larger than the diameter of the portals of Q6, readily penetrate into the cavity, whereas the inclusion of cyclic or branched hydrocarbons is, apparently, accompanied by deviation of the oxygen atoms at the portals from their normal positions. Studies devoted to investigations of the thermodynamic and kinetic aspects of the inclusion of bulky ligands 21 ± 23 demonstrated that the formation of host ± guest supramolecular compounds depends, in particular, on the acidity of the medium.The formation of compounds with cucurbit[6]uril is the first example of the pH-controlled kinetics of encapsulation of organic guests into the host cavity. The inclusion of a protonated guest into the cavity of the cavitand is slow due to formation of an intermediate upon interaction of the guest with the polar oxygen atoms at the portals, whereas an unprotonated molecule is encapsulated directly in the cavity and the rate constant of this reaction is 20 times larger.23 High specificity of binding of Q6 with ammonium ions made it possible to perform unique azide ± alkyne cyclisation in which both starting components, viz., propargylamine and azidoethyl- amine, are included in the cavity of cucurbit[6]uril and cyclisation proceeds within the cavitand.24, 25 H H H N H H N HCO2H, H2O N R +N +N R0 H H H H H H N N H H H H+ R0 N N + N R H H H H H H H Under standard conditions, azide ± alkyne cyclisation pro- ceeds in solutions rather slowly (k0*1076 mol71 s71) to give two isomeric products in equal amounts.Due to interactions of the ammonium groups of the starting reagents with the opposite portals of Q6, only a certain arrangement of the components in the cavity is possible resulting in regiospecificity of this reaction (only the 1,4-addition product is generated).Besides, the reaction is substantially accelerated in the presence of catalytic amounts of cucurbit[6]uril (k0*1072 mol71 s71). 2. Inclusion of dye and other guest molecules Examples of inclusion of other guest molecules into the cavity of cucurbit[6]uril (5) are few in number. The sizes of the cavity allow tetrahydrothiophene, tetrahydrofuran, benzene, cyclopentanone and furan to be accommodated, which was proved by shifts of the signals in the solutionNMRspectra of these compounds.17, 26 The para-substituted benzene ring can also be included into the cavity of cucurbit[6]uril (if the substituents are directed toward the portals). However, the size of this ring is somewhat larger than the size of the cavity resulting in distortions of both the guest and host molecules.Calorimetric titration showed that the stability of inclusion compounds of Q6 with aliphatic alcohols as well as with acids and nitriles is unaffected by the presence of polar groups and the number of methylene units in the guest molecule.27 Various dye molecules form stable inclusion compounds with cucurbit[6]uril.7, 28 ± 30 Most of the dyes are hydrophobic and their molecules (or molecular fragments of suitable size) can interact rather strongly with the inner cavity of the cavitand. In some cases, the encapsulation in cucurbit[6]uril leads to substantial stabilisation of dye molecules. For example, Phenol Blue (7OC6H4NC6H4NMe2) decomposes in a 0.1M HCl solution during 3 min.If this dye is incorporated into cyclodextrin, it decomposes under the same conditions during 7 min. If this dye is744 encapsulated in Q6, it is stable for 7 h.28 Higher stabilities of inclusion compounds of dye molecules with cucurbit[6]uril in dilute solutions of formic acid as compared to inclusion com- pounds with c b-cyclodextrin were also confirmed by spectropho- tometric measurements.29 The formation of inclusion compounds of cucurbit[6]uril with dyes is used for example, in textile industry for efficient absorption of dissolved, dispersed or emulsified dyes from industrial wastes.7, 30 Cucurbit[6]uril can bind surfactants and polyethylene gly- cols,31 fluorescent compounds,32 and volatile organic compounds from the gas phase.33 The 129Xe and 1H NMR spectroscopic data provided evidence for the reversible inclusion of xenone into Q6 and the formation of a 1 : 1 complex in an aqueous solution.34, 35 In spite of the absence of Coulombic or hydrophobic interactions, the stability of this compound is comparable with the stabilities of Q6 complexes with alkylammonium ions.3. Synthesis and structures of nanosized rotaxanes The ability of cucurbit[6]uril (5) to form host ± guest compounds with alkyldiammonium salts provided the basis for the directed construction of nanosized rotaxanes (coordination polymers), which are not only of interest from the viewpoint of structural chemistry but also possess a series of unique properties due to which they show promise as materials, for example, for molecular electronics.12, 13 Rotaxanes are supramolecular compounds in which cyclic molecules (beads) are threaded onto linear molecules containing bulky terminal groups, which keep the beads from sliding off.36 Analogous structures devoid of bulky terminal groups are called pseudorotaxanes.Such macrocycles as crypt- ands, crown ethers, cyclodextrins and calixarenes are rather widely used in the synthesis of rotaxanes.37 Compounds contain- ing the Q6 molecule as a bead were first prepared and structurally characterised byK Kim and co-workers.38 Two protonated nitro- gen atoms of the RN+H2(CH2)4N+H2R molecule [R=(CH2)3NHC6H3(NO2)2-2,4] are fixed at both portals of cucurbit[6]uril thus holding a chain of four polymethylene groups in its cavity, whereas two bulky 2,4-dinitrophenyl groups are located at the ends of the molecule.In addition, a pseudorotaxane compound was prepared in the cited study by the replacement of the 2,4-dinitrophenyl groups with the carbamate groups. In this case, Q6 does not move along the chain due to rather strong interactions between the protonated nitrogen atoms and the carbonyl groups at the portals of cucurbit[6]uril. In crystals, these pseudorotaxanes are linked in chains through hydrogen bonds between the carbamate groups of the adjacent fragments to form one-dimensional polymers. The formation of pseudorotaxanes in solutions containing diammonium ions and cucurbit[6]uril has been proved by differ- ent methods, viz., by IR and NMR spectroscopy and mass spectrometry.39 ± 44 It has been demonstrated that pseudorotax- anes can act as molecular switches, i.e., an increase in pH of solutions, as in the case of diamines, leads to deprotonation of the ammonium nitrogen atoms responsible for binding with Q6 and causes the movement of the macrocycle along the chain.In some cases, however, not only an increase in acidity of the medium but also heating are required for the reverse reaction to occur.43 The use of fluorenyltriamine as the guest molecule made it possible to prepare a unique highly sensitive molecular switch in which the reversible movement of Q6 along the chain is accompanied by changes in the colour of the solution and fluorescence intensity.44 At pH 1.0, a yellow fluorescent solution of the Q6 complex with the hexanediamine fragment of fluorenyltriamine is formed.As the pH is increased to 7, Q6 moves to the butanediamine fragment of the guest molecule due to deprotonation of the aniline nitrogen atom. This process is accompanied by a change in the colour of the solution to violet and the fluorescence intensity decreases. The use of cucurbit[6]uril as a catalyst for azide ± alkyne cyclisation within the cavity made it possible to prepare a polyrotaxane with desired composition 45 from which the Q6 molecules were removed by increasing the pH of the medium. As a result, oligotriazoles O A Gerasko, D G Samsonenko, V P Fedin containing up to nine triazole-linked monomeric units were obtained.46 K Kim and co-workers 12, 13 developed an original approach to the synthesis of cucurbituril-containing polyrotaxanes, which allowed them to prepare first coordination-linked polymers.Initially, Q6 are threaded onto a small organic molecule with the diamine hydrocarbon chain containing suitable terminal func- tional groups. Then these fragments are linked to form a coordi- nation polymer through bonds between the donor atoms of the terminal groups of the resulting pseudorotaxanes and the metal atoms. The above-described method has particular advantages over the previously known procedures giving rise to polyrotaxanes in which such macrocycles as cyclodextrins or crown ethers are threaded onto organic polymers. First, the resulting polyrotax- anes are characterised by high structural regularity, i.e., cyclic beads are present in each repeated structural unit of the polymer chain.Second, these polyrotaxanes were obtained as single crystals and characterised by X-ray diffraction analysis, whereas single crystals of cyclodextrin-containing polyrotaxanes suitable for X-ray diffraction studies are difficult to prepare because of the existence of numerous isomers with different orientations of the cone-like macrocycle. Coordination of the donor atoms of the terminal groups of pseudorotaxanes to transition metal atoms [Cu(II), Co(II), Ni(II), Ag(I), Cd(II)] gives rise to one- and two-dimensional polyrotax- anes depending on the coordination number of the metal atom, the structure of pseudorotaxane, the size and the coordination ability of the counterion of the starting metal salt. The products of the reactions of different metal salts with the pseudorotaxanes (PRnm) are shown in Fig.2. In these compounds, the cucurbit[6]- uril molecule accommodates the molecule of dipyridylamines 6 or 7, which contain chains of n methylene units (n=4, 5) and the para- (m=4) or meta-substituted (m=3) terminal pyridine rings. + + N N CH2NH2(CH2)nNH2CH2 6 n=4, 5; m=4. + + CH2NH2(CH2)nNH2CH2 7 N N n=4, 5; m=3. The reactions of pseudorotaxanes of the PR44 (n=4, m=4) and PR43 (n=4, m=3) types with aqueous solutions of cobalt(II) nitrates 47 and nickel(II) nitrates 48 afforded chain poly- mers. In these polymers, pseudorotaxanes can be in either the trans or cis positions at the octahedrally coordinated metal atoms to form linear chains (PR44) (see Fig.2, structure A) and zigzag chains (PR43) (see Fig. 2, structure B), respectively. The size of the repeated zigzag fragment in the structure B is larger than 23 A. The reaction of PR44 with Cu(NO3)2 also gave rise to zigzag chains where two pseudorotaxane molecules are coordinated to the copper atom in a square-pyramidal environment.49 Silver(I) compounds tend to form linear two-coordinate complexes. The reaction of PR44 with silver tosylate [Ag(MeC6H4SO3)] resulted in the trans coordination of two pseudorotaxane molecules to the metal atom giving rise to a virtually linear chain structure.50 One- dimensional helical polymers were synthesised by the reactions of AgNO3 (see Ref.51) and Cd(NO3)2 (see Ref. 47) with pseudo- rotaxane PR53 in which the organic molecule encapsulated in cucurbituril contains five methylene groups and the meta-sub- stituted terminal pyridine rings. Two silver ions and two pseudo- rotaxanes coordinated to these ions comprise one turn of the helix (see Fig. 2, structure C), the length of one fragment being equal to 17.9 A. In the compound with cadmium, each turn of the helix (50.5 A) consists of four pseudorotaxanes PR53 and four metal ions (see Fig. 2, structure D). The space between the chains in polyrotaxanes is filled with water molecules and anions linkedSupramolecular chemistry of cucurbiturils n A n G F is the schematic representation of an inclusion compound; .is the metal ion. H Figure 2. Formation of one-, two- and three-dimensional polyrotaxanes A± J (see the text) in the reactions of pseudorotaxanes with solutions of metal salts. through an extensive network of hydrogen bonds resulting in additional stabilisation of the structure. Unlike the above-described reaction with silver tosylate giving rise to a chain structure, the reaction of the pseudorotaxane PR44 with an aqueous solution of silver nitrate afforded a two-dimen- sional polyrotaxane.50 The reaction with the use of another counterion afforded a polymeric structure in which the cucurbi- turil beads are threaded onto a network of face-sharing chair- shaped hexagons with silver ions occupying each vertex (see Fig.2, structure E). Each face consists of two silver ions coordi- nated by two nitrogen atoms of the terminal pyridine rings of the pseudorotaxane. The length of the face is 20.9 A and the average distance between two opposite angles is 38.0 A. In the crystal, these two-dimensional polyrotaxane networks form superim- posed interpenetrating layers with complete linkage of the hex- agons. The resulting compound is the first example of polycatenane ± polyrotaxane networks (catenanes are compounds consisting of interpenetrating rings). When the positions of the nitrogen atoms of the terminal pyridine rings of the pseudorotax- ane are changed with the simultaneous change of the counterion, the reaction of PR43 with (CF3SO3) affords a two-dimensional N +NH2 + H2N NN + H2N +NH2 N n B C N +NH2 + H2N N Cu(NO3)2 N + H2N +NH2 N 745 D E J I polyrotaxane of another type, which can be described as networks consisting of squares with cucurbituril molecules threaded onto all sides of these squares (see Fig.2, structure F).47 The networks are packed in layers (the distance between the layers is 12.7 A). In this case, the catenane structure is not formed. A compound prepared by the reaction of the pseudorotaxane PR43 with copper nitrate Cu(NO3)2 has an analogous layered structure.47 The cis-coordination of pseudorotaxanes to the metal atoms in a planar-square coordination makes it possible to prepare the so-called molecular necklaces in which several cyclic molecules are threaded onto a larger-size ring.Thus, the coordination of the ethylenediamine ligand in the [Pt(en)(NO3)2] complex (en is ethylenediamine) is such that pseudorotaxanes can occupy only cis positions on reaction with this complex to give nanocycles with three (PR44) (see Fig. 2, structure G) or four (PR43) (see Fig. 2, structure H) threaded cucurbit[6]uril molecules.52 A molecular necklace consisting of four rotaxane fragments was also prepared in the reaction of copper(II) nitrate with a polyrotaxane, whose angular form is dictated by the phenanthroline derivative that links two Q6 molecules.53 The ring is formed upon cis coordina- tion of the copper atom by the nitrogen atoms of the terminal Scheme 1 Cu(OH2)3 N N + + H2N NH2 +NH2 HN N N Cu Cu N N + NH H2N + + H2N NH2 N N Cu(OH2)3746 pyridine groups of two angular pseudorotaxanes.The distance between the metal atoms in these nanocycles is *20 A (Scheme 1). The use of molecular necklaces as large building blocks for the design of supramolecular compounds is of obvious interest. The reaction of PR53 with Cu(NO3)2 in the presence of oxalate ions afforded a supramolecular structure in which molecular necklaces consisting of six copper-bound pseudorotaxanes are linked to each other through the coordinated oxalate ions (see Fig. 2, structure I).53 Each planar ring is bound to six other rings giving rise to two types of cavities in the polymer structure.The cavities of one type are located within the hexagons (the diameter is 5 A), whereas the cavities of another type formed by oxalate ions are located between the hexagons (the diameter is 14 A). These cavities form channels occupied by the NO¡3 ions and water molecules. The sizes of the channels are sufficiently large and allow NO¡3 ions to be replaced by the PF¡6 and tosylate ions in aqueous solutions. The replacement by bulkier anions, such as tetraphenylborate, does not occur, which is indicative of the rather high selectivity of the polyrotaxane compound in these reactions. Three-dimensional polyrotaxanes can be prepared upon coor- dination of the terminal groups of pseudorotaxanes by lanthanide atoms, which typically have large coordination numbers.54 Under hydrothermal conditions, the reaction of the pseudorotaxane PR43 containing the terminal 3-carboxyphenyl groups with Tb(NO3)3 afforded a supramolecular compound with a three- dimensional coordination network as the framework.Each lattice node is occupied by two Tb3+ cations (see Fig. 2, structure J; one solid circle in each lattice node corresponds to two terbium atoms) coordinated by six pseudorotaxane molecules through the car- boxylate groups, with four carboxylate groups serving as bridges between the terbium atoms. As a result of a change in the position of the carboxylate group in the benzene ring, the reaction of Tb(NO3)3 with the pseudorotaxane PR44 yielded a two-dimen- sional polymer in which each terbium centre is coordinated by three pseudorotaxane fragments.The polycatenane-polyrotaxane structure of this compound is analogous to the structure of the two-dimensional polyrotaxane formed in the reaction of AgNO3 with a pseudorotaxane containing the terminal pyridine groups (see Fig. 2, structure E). Cucurbit[6]uril (5) has found an interesting application in the chemistry of dendrimers. In the last decade, these three-dimen- sional organic compounds have attracted the attention of R R R R NH R R HN HN HN NH NH R H R N NH R R NH N N HN N N R NH H R N R N N N N R NH N NHNH N N R N N N R NH N N R N R N N N NH N N HN HNHN R N N N N R N R H H N N N R N N N H NH R R NH NH NH R NH NH NH R R NH R R R R HN R NH2 HN R= .O O A Gerasko, D G Samsonenko, V P Fedin researchers because of their successful use in various fields of science and technology (medicine, catalysis and agriculture).55 In the reaction of Q6 with a dendrimer containing terminal alkyldiammonium groups, the macrocycles are threaded onto these terminal groups to give a dendrimer with terminal pseudo- rotaxane fragments. The use of a bulky bead as a component of the terminal groups imparts rigidity to the structure of the dendrimer. The resulting compound has a globular structure and can hold guests within its cavity. If the pHof the solution is increased, Q6 is removed and the dendrimer becomes conformationally flexible and releases the guest molecule.This reversible inclusion of molecules opens up new possibilities for using dendrimers for the transport, for example, of pharmaceuticals. Hence, rotaxanes and pseudorotaxanes based on cucur- bit[6]uril (5) have proved to be convenient building blocks for the construction of supramolecular compounds. Cucurbit[6]uril is rigidly fixed at the hydrocarbon chain through Coulombic and van der Waals interactions and the resulting rotaxanes and pseudorotaxanes are linked to each other through bonds between the donor atoms of the terminal groups and the metal atoms to form nanosized supramolecular compounds. III. Supramolecular compounds of cucurbit[6]uril with metal aqua complexes The ability of cucurbit[6]uril (5) to bind metal ions in aqueous solutions was discovered by Behrend and co-workers in 1905.5 The presence of more polar oxygen atoms in cucurbit[6]uril as compared to those in crown ethers or cryptands is responsible for the formation of stronger interactions between these oxygen atoms and metal cations.The formation constants of metal complexes with Q6 are much higher than the formation constants of complexes with the above-mentioned organic macrocycles.8 Besides, the rigidity of Q6 hinders its conformational changes upon complex formation, in contrast, for example, to metal complexes with crown ethers.56 The metal cations are coordinated by the oxygen atoms located in the planes of the portals of cucurbit[6]uril rather than are included into the cavity of the cavitand.Hence, due to the unique structure of cucurbit[6]uril, it can form not only inclusion compounds with organic molecules but also mixed organic ± inorganic compounds with metal cations. In the late 1990s, interesting supramolecular compounds of Q6 with the sodium, potassium, rubidium and cesium cations were prepared.26, 57 ± 59 Cucurbit[6]uril is insoluble in water but its solubility is noticeably increased in the presence of salts of these metals. These compounds were obtained as single crystals by slow diffusion of organic solvents into aqueous solutions and charac- terised by X-ray diffraction analysis. Metal aqua complexes are bound to the carbonyl oxygen atoms of the Q6 portals to form lids, which efficiently close the cavitand from both sides.In the compound of cucurbit[6]uril with a sodium aqua complex (8), each portal of the macrocycle is coordinated to two metal cations linked to each other through a bridging water molecule (the symbol ( designates the inclusion into the cavity of the cavitand).26 {[Na4(H2O)10](C4H8O(C36H36N24O12)}(SO4)2 . 10H2O 8 Each sodium atom is bound to two carbonyl oxygen atoms of Q6 and three water molecules. Due to formation of supramolec- ular compounds of the type `barrel with two lids', the cavity of the cavitand can include and hold the tetrahydrofuran molecule (Fig. 3). However, an increase in acidity leads to protonation of the carbonyl groups at the Q6 portals, removal of the lids and release of the guest from the cavity of the cavitand. After the addition of Na2CO3, tetrahydrofuran is again included into the cavity because the barrel is again efficiently closed by the lids.According to the 1H NMR spectroscopic data, other guests, suchSupramolecular chemistry of cucurbiturils C N O Na Figure 3. Structure of the supramolecule {[Na4(H2O)10](C4H8O(C36H36N24O12)}(SO4)2 . 10H2O (8). The cavity of the cavitand encapsulates the tetrahdyrofuran molecule. as benzene, cyclopentanone and furan, behave analogously. The supramolecular compound of Q6 with the sodium aqua complex is the first example of a molecular container into which guests can be reversibly included through the formation and cleavage of the bonds between the metal aqua complex and macrocycle.In the compounds with potassium (9) 57 or rubidium (10),58 the portals of Q6 are also closed by the bridging binuclear aqua complexes serving as the lids. {[K2(OH)2](C4H8O(C36H36N24O12)} . 18H2O 9 {[Rb2(OH)2(CH3OH)2(H2O)2](C36H36N24O12)} . 17H2O 10 Each metal atom is coordinated by four oxygen atoms at the portals of two Q6 molecules (two oxygen atoms from each molecule) (Fig. 4). This leads to the formation of hybrid N C O Rb Figure 4. Structure of the supramolecular compound {[Rb2(OH)2(CH3OH)2(H2O)2](C36H36N24O12)} . 17H2O (10) (view along the axis a). 747 b a Figure 5. Formation of the honeycomb-like structure in the supramole- cule 9 (view along the axis c).organic ± inorganic polymers consisting of alternating metal aqua complexes and cucurbit[6]uril. The structures of these compounds are stabilised by van der Waals interactions due to complementarity of the chain folds, i.e., the convexities of one chain match the concavities of the adjacent chains. The polymer chain surrounded by three adjacent chains is shifted by half of the translation along the axis c to form a honeycomb-like packing containing large channels (Fig. 5). The resulting compounds are promising for the design of porous materials, which find use in separation processes, supramolecular catalysis and optoelec- tronics. In the honeycomb-like structure of the compound with the rubidium ion 10, the diameter of the channels filled with water molecules of solvation is *10 A and the volume of these chains comprises 23% of the total volume of the structure.58 Upon the addition of THF or ethylenediamine, the molecules of these compounds are included only into the cavity of the cavitand, the channels remaining unoccupied.57, 58 Each portal of cucurbit[6]uril in the supramolecular com- pound {[Cs2(H2O)2](H2O(C36H36N24O12)}Cl2 .2H2O,59 like those in the above-described compounds with sodium (8), potas- sium (9) and rubidium (10), is coordinated to the metal aqua complexes.However, due to the larger ionic radius of the cesium cation, the barrel can be closed by a mononuclear cesium complex serving as the lid. The cesium atom is bound to four carbonyl oxygen atoms of cucurbituril and three water molecules.The addition of tetrahydrofuran gave rise to the inclusion compound {[Cs(H2O)2Cl](C4H8O(C36H36N24O12)} .5H2O (11) in which the barrel is closed only by one lid, whereas the oxygen atom of the guest THF molecule is coordinated to the cesium ion (Fig. 6). In this case, the inclusion of tetrahydrofuran, as in the case of the sodium complex 8, is also reversible and depends on the acidity of the medium. The supramolecular compound of cucurbit[6]uril with the calcium aqua complex {[Ca(HSO4)2]2(C36H36N24O12)} . 13H2O is the first example of the structurally characterised cucurbituril compound. The structure of this compound, which was estab- lished by Freeman and co-workers 6 in 1981 and refined by Freeman at a later time,14 is a polymer consisting of the alternat- ing Q6 molecules linked to the calcium cations through the carbonyl groups.In the presence of methanol, the coordinated CH3OH molecule is included into the cavity of the cavitand, the shape and packing of the supramolecules remaining virtually unchanged.60 The structure of the fragment of the chain in the {[Ca(H2O)3(HSO4)(CH3OH)]2(C36H36N24O12)}(HSO4)2 .4H2O complex (12) is shown in Fig. 7. Each Q6 molecule is bound to four calcium cations and each calcium cation is bound to two Q6748 Cl O Cs N C O O Figure 6. Structure of the supramolecule {[Cs(H2O)2Cl](C4H8O(C36H36N24O12)} .5H2O (11). The cavity of the cavitand encapsulates the tetrahdyrofuran molecule.molecules (one carbonyl group of one molecule and two carbonyl groups of another molecule). In addition to the carbonyl groups of the Q6 molecules, the calcium ions are coordinated by three water molecules, the hydrosulfate anion, and the methanol molecule whose methyl group is located in the hydrophobic cavity of the cavitand. c b Figure 8. Formation of molecular cylinders (view along the a axis) in the supramolecular compound 12. Water molecules of crystallisation and sulfate anions are omitted. O A Gerasko, D G Samsonenko, V P Fedin Figure 7. Fragment of the chain of the compound of cucurbit[6]uril with the calcium aqua complex {[Ca(H2O)3(HSO4)(CH3OH)]2(C36H36N24O12)}(HSO4)2 .4H2O (12). The cavity of the cavitand encapsulates the methyl group of MeOH.In the crystals, the polymer molecules form molecular cylin- ders linked to each other through hydrogen bonds and the channels are occupied by the water molecules of crystallisation and HSO¡4 anions (Fig. 8).Supramolecular chemistry of cucurbiturils Unlike the compounds of cucurbit[6]uril (5) with aqua com- plexes of Groups 1 and 2 metals in which there are direct interactions between the metal atoms and oxygen atoms at the portals of Q6, coordination of Group 13 metals by the macrocycle occurs through hydrogen bonding between the portal oxygen atoms and coordinated water molecules of the aqua complexes. It should be noted that in the latter type of compounds, Q6 acts as an outer-sphere ligand. In supramolecular compounds with cucurbi- t[6]uril, distortions in the aqua complexes are much smaller than those in the case of insertions of the ligands into the first coordination sphere.Consequently, the structural information for compounds with Q6 corresponds better to the compositions and structures of aqua complexes that are actually present in solution. In the compounds of cucurbit[6]uril with aqua and chloro aqua complexes of indium and aluminium, the metal complexes are located between the Q6 molecules and are linked to each other through a complex network of hydrogen bonds involving the portal oxygen atoms and water molecules. The complex trans- [InCl2(H2O)4]+ and cis- and trans-[InCl4(H2O)2]7 ions were isolated and structurally characterised in the investigation of the reaction of cucurbit[6]uril with indium nitrate in hydrochloric acid solutions.61 In the compound with composition {[InCl2(H2O)4]3(Cl(C36H36N24O12)}2 Cl .4H2O(13), the macro- cyclic molecules form a hexagonal one-layer packing (Fig.9), the crystallographic positions occupied by the macrocycles in this structure having the highest symmetry D6h possible for cucurbit- [6]uril. The cavity of the cavitand contains the Cl7 anion disordered over 12 positions. The structure of the (H3O)3{[InCl4(H2O)2]3(C36H36N24O12)2} . 17H2O compound is composed of both the trans and cis isomers of the [InCl4(H2O)2]7 C N O In Cl Cl b a Figure 9. Crystal structure of {[InCl2(H2O)4]3(2 Cl(C36H36N24O12)}Cl .4H2O (13).749 complex. In this crystal structure, the Q6 molecules form a parquet-like packing. The reactions of Q6 with aluminium chlor- ide and indium nitrate in aqueous solutions afforded supramolecular compounds with the aqua complexes [Al(H2O)6]3+ and [In(H2O)6]3+ with compositions {[Al(H2O)6]. .(C36H36N24O12)}Cl3 . 18H2O and {[In(H2O)6](C36H36N24O12)}. .(NO3)3 .9H2O, respectively.61 In the structure of the former compound, the centres of the Q6 molecules follow the body- centered law. In the structure of the latter compound, the molecules of the macrocycle form a parquet-like packing. In studies of the M3+ ± HCl ±Q6 systems, the isostructural compounds with composition (H7O3)4{[MCl4]2(C36H36N24O12)}Cl2 . nH2O, where (M=Ga, n=2;62M=Fe, n=363), were obtained.In the structures of these complexes, the Q6 molecules alternating with the [MCl4]7 anions are also linked through a system of hydrogen bonds between the portal oxygen atoms and water molecules. The reactions of lanthanide(III) nitrates and sulfates with Q6 in aqueous solutions afforded supramolecular compounds in which (according to the X-ray diffraction data) the aqua com- plexes are bound to the macrocycle through both the direct coordination of the metal atom by the oxygen atoms and hydro- gen bonding between the coordinated water molecules and portal oxygen atoms of cucurbit[6]uril.64 In the resulting compounds, the aqua complex : cucurbit[6]uril ratio is 1 : 1, 2 : 2 or 2 : 3. In com- plexes 14 ± 17 with composition 1 : 1, the metal atom together with its ligand environment forms the lid that closes the Q6 molecule from one side.{[La(H2O)6(SO4)](C36H36N24O12)}(NO3) . 12H2O 14 {[Gd(NO3)(C2H5OH)(H2O)3](C36H36N24O12)}(NO3)2 . 5.5H2O 15 {[Ho(NO3)(H2O)4](C36H36N24O12)}(NO3)2 .7H2O 16 {[Yb(NO3)(H2O)4](C36H36N24O12)}(NO3)2 . 62O 17 ON C Ce Figure 10. Fragment of the {[Ce(H2O)5]2(C36H36N24O12)2}Br6 . 26H2O structure (18): formation of the sandwich-type compound.O A Gerasko, D G Samsonenko, V P Fedin 750 In the {[Ce(H2O)5]2(C36H36N24O12)2}Br6 . 26H2O complex (18) with composition 2 : 2, two cerium aqua ions are coordinated by two Q6 molecules to form a sandwich-type compound (Fig. 10). The crystal structures of compounds 19 65 and 20 64 with N C composition 2 : 3 can be described as packings of nanosized triple-decker sandwiches built from alternating Q6 molecules and metal aqua complexes (Fig.11). O Gd {[Sm(H2O)4]2(C36H36N24O12)3}Br6 . 44H2O 19 {[Gd(H2O)4]2(C36H36N24O12)3}Br6 . 45H2O 20 The shape of this supramolecule can be approximated by a skewed cylinder whose largest van der Waals size is 33 A. In the crystals of lanthanide(III) compounds, the supramolecules are linked in chains or layers or packed in stacks through hydrogen bonding. In these structures, large channels and cavities are occupied by water molecules of crystallisation and counterions. The possibility of using Q6 for crystallisation of polynuclear aqua complexes was exemplified by the preparation of the mixed uranium(VI) oxochloride complex.66 The supramolecular com- pound (21), which is composed of the tetranuclear uranium aqua complex and cucurbit[6]uril (5), is formed through hydrogen bonding between the coordinated uranium atoms, water mole- cules and carbonyl oxygen atoms of the Q6 portals (Fig.12). {[(UO2)4O2Cl4(H2O)6](C36H36N24O12)} . 5H2O 21 Figure 11. Fragment of the {[Gd(H2O)4]2(C36H36N24O12)3}Br6 . 45H2O structure (20): the triple-decker sandwich. In the crystals, the Q6 molecules and aqua complexes form alternating layers linked to each other through a system of hydrogen bonds. In the aqua complexes, the uranium atoms of four uranyl groups are virtually in a single plane. N C O Cl U Figure 12. Fragment of the structure of the {[(UO2)4O2Cl4(H2O)6](C36H36N24O12)} .5H2O compound (21).Supramolecular chemistry of cucurbiturils By comparing the modes of binding of cucurbit[6]uril (5) to different metals, several conclusions can be made.Thus, the carbonyl oxygen atoms of the Q6 portals are directly coordinated to the metal atoms in supramolecular compounds with alkali, alkaline earth and rare-earth metals. These most electropositive metals have rather large ionic radii (rion=0.97, 1.73, 1.03 and 1.14 A for Na+, Cs+, Ca2+ and La3+, respectively 67) and their binding to the ligands should be considered as Coulombic (ion ± dipole) interactions. In these cases, cucurbit[6]uril, which is a weak donor but contains polarisedC=Ogroups, competes with other ligands for site in the coordination sphere about the metal atom.Group 13 metal ions have smaller radii (rion=0.76 and 0.50 A for In3+ and Al3+, respectively 67) and their binding to the ligands is predominantly covalent. The larger charges and smaller sizes of Group 13 metal cations lead to an increase in acidity of the coordinated water molecules of aqua ions. Consequently, the protons of the water molecules coordinated to these metal cations can form much stronger hydrogen bonds with the polarisedC=O groups due to which the inclusion of water into the first coordi- nation sphere is more favourable than the inclusion of Q6. If the metal ± ligand interaction is predominantly Coulombic, Q6 com- petes with water for sites in the inner coordination sphere of the metal atom.If this interaction is predominantly covalent, the structure is determined primarily by a system of hydrogen bonds between the aqua ligands and cucurbit[6]uril (5) as the outer- sphere ligand. IV. Supramolecular compounds of cucurbit[6]uril with metal cluster aqua complexes The use of bulky fragments, which retain their geometry and can form extensive networks of hydrogen bonds due to a large contact surface, is an important condition for the directed construction of nanosized complexes or supramolecular compounds. Molybde- num and tungsten cluster aqua complexes are suitable large molecular building blocks for the construction of supramolecular compounds with cucurbit[6]uril (5).Studies of interactions of Q6 with trinuclear thio and seleno complexes of these metals and their heterometallic derivatives led to the discovery of a new class of supramolecular compounds. In aqueous solutions, cluster thio and seleno complexes of molybdenum and tungsten produce the stable trinuclear aqua ions [M13 Y4(H2O)9]4+ (M1=Mo or W; Y=S or Se) in which the metal atoms are coordinated by one m3-bridging chalcogen atom and three m2-bridging chalcogen atoms.68 ± 71 The coordinatively unsaturated m2-bridging chalco- gen atoms are readily coordinated to transition and post-tran- sition metal atoms.72 In this case, the trinuclear cluster (a cube without one vertex) is transformed into the cubane complex M2M13 Y4 and the m2-bridging chalcogen atoms are transformed into the m3-bridging atoms.+ M2 M2M13 Y4 M13 Y4 M1=Mo, W; M2 is the transition or main-group metal atom; Y=S, Se. Each M1 atom of the cluster core (M13 Y4 or M2M13 Y4) is coordinated by three water molecules. Six water molecules are in cis positions with respect to the m3-bridging chalcogen atom in the trinuclear aqua complexes or with respect to the m2-bridging chalcogen atom, which links only the molybdenum (tungsten) atoms, in the cubane heterometallic complexes. These six aqua ligands are virtually in a single plane and are well suited in size and symmetry to six carbonyl groups at each portal of Q6. In addition, the cluster aqua complexes [M13 Y4(H2O)9]4+ have rather high positive charges due to which they exhibit pronounced acidic properties and act as good donors in hydrogen bonding, whereas the oxygen atoms of the polarised carbonyl groups of Q6 are potential acceptors.This geometric and functional (donor ± ac- ceptor hydrogen bond) complementarity gives rise to an extensive network of hydrogen bonds between the macrocycles and the triangular (cubane) aqua complexes. In spite of the fact that the energy of each individual hydrogen bond is low, the presence of a system of these bonds facilitates the formation of stable supra- molecular compounds. The use of Q6 for crystallisation of chalcogenide cluster aqua complexes made it possible for the first time to isolate and structurally characterise more than 20 supramolecular com- pounds.The compounds were crystallised by slow evaporation (in air) of solutions of cluster aqua complexes and Q6 in dilute hydrochloric acid. The resulting crystals are insoluble or weakly soluble in aqueous solutions of hydrochloric acid in which the starting reagents are readily soluble. If the cluster : cucurbit[6]uril ratio in supramolecular compounds is 1 : 1, only one portal of Q6 is closed by the cluster complex (structural type `barrel with one lid'). If the cluster : cucurbit[6]uril ratio is 2 : 1, both portals are closed (structural type `barrel with two lids') (Fig. 13). The distances between the carbonyl oxygen atoms of the macrocycle and aqua ligands in both structural types vary in the range of 2.5 ± 3.1 A.If the lids are coordinated, the cavity of the cavitand C b C S(Se) Figure 13. Supramolecular compounds of cucurbit[6]uril with metal clus- ter aqua complexes: (a) barrel with one lid; (b) barrel with two lids. a S(Se) O Mo(W) N O Mo(W) N 751752 becomes isolated and small molecules can be held in the molecular container. The sizes of supramolecules closed by two lids are 25 ± 30 A. A wide variety of available supramolecular compounds of Q6 with molybdenum and tungsten cluster aqua complexes is asso- ciated with the nature of the chalcogenide cluster, the presence of guests in solution and the concentration of hydrochloric acid in the aqueous solutions in which the reactions proceed. The reactions of Q6 with the trinuclear tungsten sulfide aqua complex [W3S4(H2O)9]4+ in 2M HCl afforded products with different compositions and structures depending on the presence of pyridine in the system.The supramolecule {[W3S4(H2O)7Cl2](C36H36N24O12)}Cl2 . 10H2O (22) belongs to the structural type `barrel with one lid' in which only one portal of Q6 is closed by the cluster complex through a system of complementary hydrogen bonds between six carbonyl groups and six aqua ligands.73 In the crystal, the supramolecules are linked in zigzag chains through additional hydrogen bonds between the aqua ligands of the cluster and the portal oxygen atoms of Q6 (Fig. 14). Under the same conditions, the reaction performed in the presence of pyridine led to destruction of the chain structure to form the layered compound {[W3S4(H2O)8Cl](PyH( N C C N OW S Figure 14.Fragment of the zigzag chain in {[W3S4(H2O)7Cl2](C36H36N24O12)}Cl2 . 10H2O (22). O S W O A Gerasko, D G Samsonenko, V P Fedin C36H36N24O12)}Cl4 . 10H2O.74 In this compound, the cavity of the Q6 molecule includes the pyridinium cation and both portals are closed by the trinuclear cluster complexes serving as the lids. In the crystal, these supramolecules are packed in layers. Upon the inclusion of the pyridinium cation into the cavity of Q6, both the guest and host molecules adapt to each other in size resulting in distortions of their structures. Thus, the Q6 molecule becomes slightly extended, whereas the pyridine molecule slightly con- tracts.However, the inclusion of a guest does not necessarily lead to such dramatic changes in the structure and packing of supra- molecules. The conditions of the preparation of selenium-con- taining tungsten complexes 23 75 and 24 76 with cucurbit[6]uril also differ only by the presence of pyridine in the reaction mixture. {[W3Se4(H2O)8Cl]2(C36H36N24O12)}Cl6 . 12H2O 23 {[W3Se4(H2O)6Cl3]2(PyH(C36H36N24O12)}Cl3 . 18H2O 24 Both compounds have chain structures in which the supra- molecules of the structural type described as a barrel with two lids are linked to each other through short (compared to the sum of the van der Waals radii of 3.9 A) nonbonded Se_Se interactions (3.59 ± 3.72 A). The cavity of cucurbit[6]uril in compound 24 contains the pyridinium cation (Fig.15). The chalcogen ± chalco- N C O W Se Se W O Figure 15. Fragment of the chain structure in {[W3Se4(H2O)6Cl3]2(PyH(C36H36N24O12)}Cl3 . 18H2O (24). The cav- ity of cucurbit[6]uril accommodates the pyridinium cation.Supramolecular chemistry of cucurbiturilsa b NC OSe Mo 25 Figure 16. Comparison of the chain structures in {[Mo3Se4(H2O)8Cl]2(C36H36N24O12)}Cl6 . 16H2O (25) (matrix) and {[Mo6HgSe8(H2O)14Cl4](C36H36N24O12)}Cl4 . 14H2O (26) (intercalate). gen interactions between adjacent clusters typical of trinuclear chalcogenide clusters of transition metals are of importance in the structure formation of these complexes.77 The stronger bonding between the selenide atoms (Se_Se) as compared to that between the sulfur atoms (S_S) [according to the structural data available in the Cambridge Structural Database for {M3Y4_Y4M3} com- pounds, where Y=Se or S (see Ref.78)] facilitates the retention of chain structures of the supramolecular compounds of Q6 with selenium-containing tungsten clusters upon the inclusion of a guest molecule. The structures of the selenium-containing com- plexes differ only by the mode of packing of the chains. In the crystal of compound 23, the chains are packed to form a hexagonal motif, whereas these chains in the crystal of compound 24 form a distorted square packing. The chalcogen ± chalcogen interactions between the adjacent cluster fragments M3Y4 of supramolecular compounds are anal- ogous to interactions between the adjacent layers formed by the chalcogen atoms in the layered transition metal dichalcogenides MY2 (Y=S, Se).75 An important property of the layered com- pounds is their ability to incorporate molecules or atoms between the layers giving rise to intercalates.A comparison of the structures of chain supramolecular compounds 25 78 and 26 75, 79 shows that these compounds are regarded as a matrix and intercalate from the structural viewpoint. Thus, the mercury atom is included between the chalcogenide clusters M3Y4, the N C O Mo Hg Se Mo 26 principal parameters of the supramolecules remaining virtually unchanged (Fig. 16). {[Mo3Se4(H2O)8Cl]2(C36H36N24O12)}Cl6 .16H2O 25 {[Mo6HgSe8(H2O)14Cl4](C36H36N24O12)}Cl4 . 14H2O 26 Therefore, the chemistry of supramolecular compounds of chalcogenide clusters is related to the chemistry of layered com- pounds more closely than one would expect. In both cases, the nonbonded chalcogen ± chalcogen interactions play an important role in the structure formation. Yet another important factor that has an effect on the formation of supramolecular compounds of Q6 with metal cluster aqua complexes is the concentration of hydrochloric acid in dilute solutions in which these reactions proceed. In solutions of HCl, cluster aqua complexes give different mixed chloro aqua com- plexes [M3Y4(H2O)97xClx](47x)+ (Y=Se, S) whose composi- tions depend on the nature of the metal and chalcogen atoms.As demonstrated above, six cis-aqua ligands of the cluster are involved in complementary binding in the compounds of Q6 with Mo andWaqua complexes, whereas up to three Cl7 anions can be located in trans positions (for example, in the supra- molecular compounds 27 80 and 24).76 753754 (H3O)2.5{[W3S4(H2O)6Cl3]2(C36H36N24O12)}Cl4.5 . 19H2O 27 {[W3Se4(H2O)6Cl3]2(PyH(C36H36N24O12)}Cl3 . 18H2O 24 An increase in the concentration of hydrochloric acid leads to the replacement of the water molecules in aqua complexes by the halide ligands resulting in a sharp weakening of the cluster ± cu- curbit[6]uril interaction. First, the cluster complex loses a positive charge, i.e., its acidity and ability to act as a donor in hydrogen bonding are decreased.Second, coordination of more than four Cl7 ligands to the cluster core requires the replacement of the cis- aqua ligands responsible for binding of the cluster complex to the portals of the macrocycle. Here, crystallisation of a particular form depends on other factors (Y_Y interactions, packing factors, etc.). Therefore, cucurbit[6]uril in HCl solutions acts as an outer-sphere ligand with respect to the cluster complex and forms hydrogen bonds only with some of the possible forms of [M3Y4(H2O)97xClx](47x)+. 2 The latter assumption is exemplified by the data on the structure of the (H3O)2[Mo3Se4(H2O)4Cl5]2(C36H36N24O12) . 15H2O compound (28) prepared from a 6M HCl solution.78 This compound contains triangular clusters in the [Mo3Se4(H2O)4Cl5]7 anionic form. Unlike the chain compound 25 (see Fig.16), which crystallises from a 2M HCl solution and contains only one chloride ion per trinuclear cluster, the Mo atoms in the compound 28 are coordinated by three or four Cl7 ligands in cis positions with respect to m3-Se. As a consequence, the Q6 molecules and the chloro aqua complexes are not involved in a system of complementary hydrogen bonds. The chloro aqua complexes are linked in the dimeric cluster aggregates {Mo3Se4(H2O)4Cl5}2¡ through short nonbonded Se_Se interac- tions. In the crystals, these dimers are packed in layers, which alternate with layers consisting of Q6 molecules. Analogous dimers are present in the isostructural (H3O)4{[M3S7Cl6]2(C36H36N24O12)} .8H2O compounds (29, M=Mo or W).81 The [M3S7]4+ cluster core in which the metal atoms are coordinated by the m3-S ligand and three disulfide ligands m2-S2 exhibits a much higher affinity for halide ions than the [M3S4]4+ core.82 In a 3M HCl solution, each metal atom is coordinated by two chloride ligands due to which hydrogen bonding with Q6 becomes impossible.In this case, the structure formation depends primarily on nonbond S_S interactions between the bridging atoms of the disulfide ligands resulting in dimerisation of the [M3S7Cl6]27 anions. The crystal packing consists of alternating layers of dimeric cluster anions and Q6 molecules. The heterometallic molybdenum and tungsten aqua com- plexes [LmM0M3Y4(H2O)9]n+ (Y=Se or S) also interact with cucurbit[6]uril to give supramolecular adducts belonging to `bar- rels with one lid' or `barrel with two lids'.In these aqua complexes, the bridging chalcogen atoms are coordinated to the transition and post-transition metal atoms and lose the ability to be involved in nonbonded Y_Y interactions. In the crystals, the supramole- cules of heterometallic aqua complexes with cucurbit[6]uril are linked either in chains through additional hydrogen bonds, as in compounds 30 83 and 31 84 (Fig. 17), or in layers, as in supra- molecular compound 32.85 {[Mo3(NiCl)S4(H2O)8Cl](PyH(C36H36N24O12)}Cl3 . 14.5H2O 30 {[Mo3(PdCl)Se4(H2O)7Cl2](C36H36N24O12)}Cl .7H2O 31 {[Mo3(PdCl)S4(H2O)6Cl3](PyH(C36H36N24O12)}Cl .14H2O 32 O A Gerasko, D G Samsonenko, V P Fedin Cl Se Mo Pd O C N Figure 17. Fragment of the zigzag chain in {[Mo3(PdCl)Se4(H2O)7Cl2](C36H36N24O12)}Cl .7H2O (31). From a comparison of compounds 31 and 32 in which only one portal of the macrocycle is closed by the cluster lid, it follows that the inclusion of the pyridinium cation into the cavity of Q6 does not lead to substantial changes in the structures of the supramolecules, i.e., not only the cluster : cucurbit[6]uril ratio but also the total charge (+1) remain unchanged. Only the crystal packing undergoes essential changes. The supramolecules 31 form zigzag chains through hydrogen bonds between the aqua ligands of the cluster and the oxygen atoms at the portals of Q6 of the adjacent adduct, whereas the addition of pyridine gives rise to dimers packed in layers.The use of Q6 for crystallisation of the reaction products of the palladium complex [Mo3(PdCl)S4(H2O)9]3+ with P(III)- or As(III)-containing n-donor ligands made it possible to prepare and structurally characterise the {[Mo3(PdE(OH)3)S4(H2O)6Cl3]2. .(C36H36N24O12)}Cl2 . nH2O compounds [E=P, n=20 (33a); E=As, n=19 (33b)].86 In these compounds, the Q6 molecules are closed by two cluster lids. The palladium atoms of these clusters are coordinated by the inorganic P(OH)3 or As(OH)3 ligands (Fig. 18). These are the first examples of the structurally characterised tribasic forms of phosphorous and arsenous acids. Heterometallic indium- and tin-containing supramolecular compounds of cucurbit[6]uril 34 87 and 35 78 belong to the struc- tural type described as `barrels with two lids'.755 Supramolecular chemistry of cucurbiturils O P O Cl Pd S Mo Cl Mo CN C N Figure 19. Structure of the supramolecule {[Mo3O4(H2O)6Cl3]2(C36H36N24O12)}2+ (36). Figure 18.Structure of the supramolecule {[Mo3(PdP(OH)3)S4(H2O)6Cl3]2(C36H36N24O12)}2+ (33a). {[Nb2S4(H2O)8](C36H36N24O12)}Cl4 . 15H2O (37) In the compound of Q6 with the binuclear niobium aqua complex (Fig. 20),89 the portal of cucurbit[6]uril is incompletely closed by {[W3(InCl3)S4(H2O)9]2(C36H36N24O12)}Cl4 . 28H2O 34 {[Mo3(SnCl3)Se4(H2O)6Cl3][Mo3(SnCl3)Se4(H2O)7Cl2]. .(C36H36N24O12)}Cl . 26H2O 35 Among the characteristic features of the crystal structure of the compound 34, noteworthy are large channels (*7 A in diameter) occupied by water molecules of solvation and Cl7 counterions. Crystallisation of the tin-containing molybdenum aqua complex with Q6 made it possible for the first time to isolate and structurally characterise the mixed chloro aqua complexes [Mo3(SnCl3)Se4(H2O)6Cl3] and [Mo3(SnCl3)Se4(H2O)7Cl2]+. O S Nb Hence, as a result of the good compatibility of the sizes and symmetry of the carbonyl groups at the portals of Q6 with those of the aqua ligands of the triangular (cubane) molybdenum and tungsten chalcogenide aqua complexes, extensive networks of hydrogen bonds are formed and the macrocycle becomes effi- ciently closed by the cluster lids.C N Figure 20. Fragment of the chain structure in {[Nb2S4(H2O)8](C36H36N24O12)}Cl4 . 15H2O (37). As in the case of the thio and seleno analogues, the trinuclear molybdenum oxo cluster [Mo3O4]4+ also interacts with Q6 through hydrogen bonds to give the supramolecular compound {[Mo3O4(H2O)6Cl3]2(C36H36N24O12)}Cl2 . 14H2O (36). In this compound, the portals are closed by two lids formed by the cluster cations (Fig. 19).88 However, unlike the compounds of Q6 with chalcogenide complexes, the plane of the Mo3 triangle in the supramolecule 36 is tilted to the plane of the portal of the macrocycle (the tilting angle is 18.5 8) and the cluster cation is shifted to one end of the portal resulting in the nonequivalence of the hydrogen bonds that are formed.These differences in the structures are attributed to the smaller size of the oxo cluster as compared to the sizes of the chalcogenide clusters and also to the higher acidity of the oxo complex.756 the binuclear cluster and the latter is shifted to one end of the portal because of spatial incompatibility. In this case, only two water molecules trans-coordinated to the niobium atom are bound directly to the portal oxygen atoms of two Q6 molecules. Several more water molecules present in the hydrophilic region of the portals are involved in additional hydrogen bonds between the aqua complex and macrocycle. In the crystal, the alternating cluster niobium aqua complexes and Q6 molecules are linked in chains.V. Cucurbit[n]urils (n=5, 7 ± 10) In addition to the above-considered macrocyclic cavitand cucur- bit[6]uril (5) consisting of six methylene-bound glycoluril frag- ments, there are also cucurbiturils, which are composed of a smaller or larger number of these fragments and have the shape of barrels with smaller or larger diameters. Crystals of the cyclic pentamer were first prepared by the reaction of dimethylglycoluril with formaldehyde and structurally characterised in 1992.90 The decamethylcucurbit[5]uril has a slightly distorted symmetry D5h and its methyl groups are located at the periphery of the molecule. In the cited study, the nomen- clature for cucurbiturils was proposed by analogy with that used for calixarenes. According to this nomenclature, the number of glycoluril (or substituted glycoluril) fragments is denoted by the figure in brackets and the number of substituents is represented by the prefix.In 2000, K Kim and co-workers 91 performed the reaction of glycoluril with formaldehyde in 9M sulfuric acid and obtained a mixture of cyclisation products (n=5 ± 11) from which five-, seven- and eight-membered homologues of cucurbituril were isolated in very low yields. These compounds were structurally characterised. A comparison of their structural parameters (Table 1) demonstrates that the diameters of the cavity and portals of cucurbiturils increase as the number of glycoluril fragments in the rings is increased. A twofold increase in the inner diameter of the cavity on going from cucurbit[5]uril to cucurbit[8]uril leads to more than a fivefold increase in the volume of the cavity.Table 1. Comparison of the structural parameters of cucurbit[n]urils. Cucurbit[n]uril Cavity diameter Portal diameter Cavity volume /A /A /A3 Molecule height /A 9.1 9.1 9.1 9.1 4.4 5.5 7.3 8.8 2.4 4.0 5.4 6.9 n=5 n=6 n=7 n=8 82 164 279 479 Day et al.92 studied the reactions of glycoluril with form- aldehyde in H2SO4, HCl or HBF4 solutions by electrospray mass spectrometry and 13C NMR spectroscopy and established the existence of cucurbit[n]urils with n=5 ± 16 in solution. Based on the results of investigations into the effects of the nature of acid and its concentration, the concentrations of other reagents and the reaction temperature, a mechanism for the formation of different cucurbit[n]urils was proposed and the procedure for their syn- thesis optimised.Cucurbit[n]urils with n=5 ± 10 were isolated in pure form (the yields were 8%, 46%, 24%, 12%, 85% and 5%, respectively). Individual cucurbit[n]urils with larger numbers of glycoluril fragments are presently unavailable. According to the results of theoretical calculations (density functional theory),93 compounds containing six or seven glyco- luril fragments are the most stable homologues of cucurbiturils, and decamethylcucurbit[5]uril is the most stable methyl-substi- tuted derivative. These results agree with both the above- described experimental data and the distortions of the NCN O A Gerasko, D G Samsonenko, V P Fedin angles at the carbon atoms of the bridging methylene groups that link the glycoluril fragments upon assembly.As demonstrated above (see Section II), the presence of the hydrophobic cavity along with the polarised carbonyl groups at the portals of cucurbit[6]uril (5) are responsible for the formation of numerous inclusion compounds. The smaller cavity in cucurbi- turils allows small molecules to be encapsulated and securely held, whereas larger-size molecular containers can accommodate bulk- ier guest molecules. Besides, an increase in the number of the oxygen atoms at the portals is favourable for the formation of a large number of hydrogen bonds and, correspondingly, more stable supramolecular compounds.4 Decamethylcucurbit[5]uril can accommodate a molecule of nitric acid. Due to the fact that the molecular size of nitric acid matches the volume of the cavity, this guest molecule can form rather short contacts and be held in the cavity through both Coulombic interactions with the carbon atoms and hydrogen bonding with the carbonyl oxygen atoms of the macrocycle.90 It was proved by electrospray mass spectrometric study of the ammonium acetate ± decamethylcucurbit[5]uril system that the N2, O2, methanol and acetonitrile molecules can be included into the cavity, held in the cavity through the formation of the NHá lids and released upon the removal of these lids. This ability of the rather rigid macrocycle to accommodate small molecules is of interest for purification or separation of gas mixtures.94 The cavity volume of cucurbit[5]uril is insufficient for the inclusion of aromatic molecules.However, according to the data from solution 1H NMR spectroscopy, the reaction with aliphatic hexylamine afforded an inclusion compound.95 Hexylamine and 1,6-diaminohexane are not encapsulated into decamethylcucur- bit[5]uril. According to the results of X-ray diffraction analysis, the terminal ammonium groups of 1,6-diaminohexane are involved in hydrogen bonds with the portal oxygen atoms of the adjacent macrocycles to form chains. The water solubility of cucurbit[5]uril and decamethylcucur- bit[5]uril as well as the solubility of cucurbit[6]uril are substan- tially increased upon the addition of salts of ammonium, alkali and alkaline earth metals due to formation of complexes with cations,96 decamethylcucurbit[5]uril giving alkali metal complexes with composition 1 : 1.97 Cucurbit[5]uril and decamethylcucurbit- [5]uril serve as ligands in transition metal complexes (M2+ and M3+).The formation constants of these complexes were deter- mined by different experimental methods.98 It has been also reported that decamethylcucurbit[5]uril in solution exhibits high selectivity toward Pb2+ cations.97 No compounds of cucurbit- [5]urils with metal cations have been structurally characterised. Due to the larger cavity volumes in cucurbiturils consisting of seven or eight glycoluril fragments, these compounds can encap- sulate bulky molecules whose sizes are too large to be included into compounds of Q6.Thus, Q7 forms inclusion compounds with o-carborane,99 with the 2,3-diazabicyclo[2.2.2]oct-2-ene molecule, whose size matches excellently the cavity size and fills virtually the entire cavity resulting in a stable supramolecular compound.100 The cavity sizes of cucurbit[n]urils (n=8, 10) allows smaller macrocycles to be included giving rise to rather rare compounds. Heating of an aqueous solution of Q8 with 1,4,7,10-tetraazacy- clododecane (cyclen) and 1,4,8,10-tetraazacyclotetradecane (cyclam) afforded inclusion compounds, which have been struc- turally characterised. These compounds are stabilised primarily through van der Waals interactions between the inner macrocycle and the cavity of the outer macrocycle.101 The small and large macrocycles are inclined to one another (the angle between the equatorial planes is*38 8).The ability of tetraazamacrocycles to coordinate transition metal ions (these complexes are known as catalysts, for example, of epoxidation or DNA hydrolysis) was used in the reactions of the resulting inclusion compounds with Cu2+ and Zn2+ salts. The compound containing [Cu(cyclen). .(H2O)]2+ in the cavity of Q8 was studied by X-ray diffraction analysis. In this compound, the macrocycles are virtually parallel to each other (Fig. 21) and the copper atom present in the innerSupramolecular chemistry of cucurbiturils N O C Figure 21.Structure of the compound containing [Cu(cyclen)(H2O)]2+ in the cavity of cucurbit[8]uril. The copper atom is represented by a large solid circle. macrocycle additionally coordinates the water molecule, which is located in the plane of the portal of Q8 and can be replaced by other ligands. It should be noted that the cyclen coordinated by the metal atom is not encapsulated into cucurbit[8]uril because the formation of the complex with the metal atom enhances its conformational rigidity and hinders its inclusion into the cavity of Q8. Recently, the unique compound of cucurbit[5]uril encapsu- lated into cucurbit[10]uril was prepared and characterised by X-ray diffraction analysis (Fig. 22).102 The molecules of the macrocycles are inclined to each other (the angle between their equatorial planes is 64 8).This fact along with the 1H and 13C NMR spectroscopic data provide evidence that the macro- cycles in solution undergo free rotation with respect to each other (molecular analogue of a gyroscope). The cavity of Q5 contains the chloride anion and there is a system of hydrogen bonds between the portal oxygen atoms of both macrocycles and water molecules serving as the lids of the inner pentamer (see Fig. 22). It is believed that Q5 serves as a template for the assembly of larger- size cucurbiturils. Cucurbituril Q10 is as yet unavailable as an individual compound. Figure 22. Structure of the inclusion compound of cucurbit[5]uril in cucurbit[10]uril. A compound of cucurbit[7]uril was prepared in which the inorganic cis-[SnCl4(OH2)2] complex is completely encapsulated in the cavity of the macrocycle.103, 104 A compound of Q7 with the trinuclear anionic chromium complex [Cr3O10]27 has been syn- thesised.In this compound, the molecules of the macrocycle are linked in hexagons to form large pores with a diameter of*17 A. The Q7 molecules, trinuclear anions and coordinated water 757 molecules are linked through an extensive network of hydrogen bonds.104 Like cucurbit[6]uril, larger macrocycles form stable inclusion compounds with organic molecules whose protonated nitrogen atoms are linked to the polarised oxygen atoms at the portals through Coulombic interactions. Due to the larger sizes of the cavities, such bulky molecules as 2,6-bis(4,5-dihydro-1H-imida- zol-2-yl)naphthalene can be accommodated in these macrocycles.Inclusion compounds with compositions 1 : 1 and 2 : 1 were prepared with cucurbit[7]uril and cucurbit[8]uril, respectively.91 The results of quantum-chemical calculations (density functional theory) of the former compound demonstrated that not only van der Waals interactions and the hydrophobic effect but also hydrogen bonding between the protonated dihydroimidazole substituents and the portal oxygen atoms of Q7 play an important role in the formation of inclusion compounds.105 Two such bulky molecules are located in an even larger inner cavity of Q8. According to the X-ray diffraction data, the imidazole substitu- ents are located outside and are bound to the portals of Q8.Two naphthalene rings within the cavity are located parallel to each other at a distance of *3.4 A, which is indicative of p ± p interactions between these rings (Fig. 23).91 Figure 23. Structure of the inclusion compound of two 2,6-bis(4,5-dihy- dro-1H-imidazol-2-yl)naphthalene molecules in cucurbit[8]uril. The encapsulation of two or more guest molecules into the cavity of the macromolecule is of great interest in view of the unique possibility of studying new types of stereoisomerism, bimolecular reactions and the behaviour of the molecules in the microenvironment. The Q8 molecule possessing a very large cavity provides rich possibilities for investigating such interac- tions.K Kim and co-workers 106 were the first to study the selective encapsulation of two different guests into the cavity of one host. Interactions between two guests are the driving forces for the formation of supramolecular compounds. Two organic mole- cules, viz., the electron-deficient molecule A and molecule B acting as the electron donor, are accommodated in the cavity of Q8 [A is the N,N0-dimethyl-4,40-bipyridinium cation (methylviologen MeN+C10H8N+Me) and B is 2,6-dihydroxynaphthalene or 1,4- dihydroxybenzene]. It should be noted that inclusion compounds with only either molecules A or molecules B were not obtained. The data from 1H NMR, emission and UV spectroscopy are unambiguously indicative of strong charge-transfer interactions between the guest molecules due to their tight contact within the cavity.This interaction is much stronger than the interaction758+ + H3N hn + H3N + + NH3 H3N 38 Q8 between these guest molecules if they are not encapsulated in the cavity of the cavitand. The resulting inclusion compounds are rather stable, which enables one to isolate and structurally characterise these compounds. For instance, the structure of the compound of Q8 with the unsymmetrical guest, viz., the cation of carboxybenzylviologen CH3N+C10H8N+CH2C6H4CO2H, was established by X-ray diffraction analysis (Fig. 24). The methyl- pyridine fragment of this cation and the naphthalene ring of 2,6- dihydroxynaphthalene serving as the electron donor are located in the cavity virtually parallel to each other at a distance of 3.4 A.This confirms the occurrence of a strong interaction, which was found by spectroscopic methods. The ability of cucurbit[8]uril to accommodate two closely spaced guests located in particular orientations allows one to use this compound as the synthetic molecular container in which bimolecular reactions between specially selected guests can pro- ceed with high regio- and stereoselectivity. Thus, photodimerisa- tion between two molecules of 4,40-diammoniostilbene dichloride (38) encapsulated into the cavity of Q8 proceeded with a much higher rate and high stereroselectivity (0.5 h, the ratio syn : anti> 95 : 5) as compared, for example, with this reaction in the presence of g-cyclodextrin (72 h, the ratio syn : anti*4 : 1).107 According to the 1H NMR spectroscopic data, Q8 forms a stable inclusion compound with two molecules 38 in solution even if the starting reagents are taken in a ratio of 1 : 1.In this case, the guests can occupy only particular positions, viz., with the parallel orientation of the alkene groups involved in photodimerisation (Scheme 2). Owing to rather rigid structures and the ability to include various molecules and ions, cucurbiturils are very attractive compounds both as synthetic receptors and convenient building Figure 24. Encapsulation of `heteroguests', viz., electron-deficient (car- boxybenzylmethylviologen) and electron-donating (2,6-dihydroxynaph- thalene) molecules in cucurbit[8]uril.O A Gerasko, D G Samsonenko, V P Fedin Scheme 2 +NH3 + + H3N NH3 + + NH3 + + NH3 H3N anti synblocks for the construction of supramolecular materials. The detailed study of supramolecular compounds based on the known cucurbiturils is hindered, in particular, by their very low solubilities in water and organic solvents. Because of this, the most promising line of investigation of the chemistry of these com- pounds is the development of procedures for the preparation of soluble derivatives by the insertion of substituents containing different functional groups. One approach to the solution of this problem involves the synthesis of cucurbiturils by condensation of substituted glycolurils whose bridging CH groups contain either alkyl or aryl substituents instead of hydrogen atoms.However, decamethylcucurbit[5]uril containing peripheral methyl groups, which was synthesised from dimethylglycoluril,90 also appeared to be a poorly soluble compound. The synthesis of cucurbiturils consisting of five or six cyclohexanoglycoluril fragments 108 is the first success after years of failed attempts. The resulting com- pounds were reliably characterised by different methods, among them X-ray diffraction analysis. The cyclohexane fragments formed by two carbon atoms of glycoluril and four methylene units are located at the periphery of cucurbiturils. The sizes of the cavities and portals are only slightly different from those in the unsubstituted analogues. These compounds are readily soluble in water, methanol and DMSO and moderately soluble in ethanol, DMFand acetonitrile. Good solubility in organic solvents enables one to use cyclohexanocucurbit[n]urils (n=5, 6) as membrane ion-selective electrodes.Cyclohexanocucurbit[6]uril dissolved in water can be used for highly selective isolation of acetylcholine (neurotransmitter).108 * * * To summarise, the recent advances in the chemistry of cucurbit- urils surveyed in the present review provide evidence that these macrocyclic organic molecules can be considered as molecular building blocks for the directed construction of nanosized highly ordered supramolecular compounds. The advantages of cucurbit- urils over other macrocyclic cavitands are the presence of a rather rigid hydrophobic inner cavity along with hydrophilic portals, which can form supramolecular compounds with metal aqua complexes through systems of hydrogen bonds.Large cucur- bit[n]urils (n=7 ± 10), which became readily available in recent years, show considerable promise. Thus, large molecular contain- ers can encapsulate bulkier guests, and the inclusion of two or more molecules provides a unique possibility for studying new types of stereoisomerism, bimolecular reactions and molecular behaviour in a microenvironment. This review will hopefully be found interesting and useful and contribute to the progress of the chemistry of cucurbiturils. This review has been written with the financial support of the Russian Foundation for Basic Research (Project Nos 01-03- 32789 and 02-03-32604) and INTAS (Grant No.2346). References 1. J-M Lehn Angew. Chem., Int. Ed. Engl. 27 89 (1988) 2. J-M Lehn Supramolecular Chemistry. Concepts and Perspectives (Weinheim: VCH, 1995) 3. J-M Lehn Science 227 849 (1985)Supramolecular chemistry of cucurbiturils 4. J R Moran, S Karbach,D J Cram J. Am. Chem. Soc. 104 5856 (1982) 5. R Behrend, E Meyer, F Rusche Liebigs Ann. Chem. 339 1 (1905) 6. W A Freeman, W L Mock, N-Y Shih J. Am. Chem. Soc. 103 7367 (1981) 7. P Cintas J. Incl. Phenom. Mol. Rec. Chem. 17 205 (1994) 8. H-J Buschmann, E Cleve, E Schollmeyer Inorg. Chim. Acta 193 93 (1992) 9. P Germain, J M Letoffe,M P Merlin,H J Buschmann Thermochim. Acta 315 87 (1998) 10. W L Mock, in Comprehensive Supramolecular Chemistry Vol. 2 (Ed.F VoÈ gtle) (Oxford: Pergamon Press, 1996) p. 477 11. H-J Buschmann, L Mutihac, K Jansen J. Incl. Phenom. Macro. 39 1 (2001) 12. K M Park, J Heo, S-G Roh, Y-M Jeon, D Whang, K Kim Mol. Cryst. Liq. Cryst. 327 65 (1999) 13. J Heo, S-Y Kim, S-G Roh, K M Park, G-J Park, D Whang, K Kim Mol. Cryst. Liq. Cryst. 342 29 (2000) 14. W A Freeman Acta Crystallogr., Sect. B 40 382 (1984) 15. W L Mock, N-Y Shih J. Org. Chem. 48 3618 (1983) 16. W L Mock, N-Y Shih J. Am. Chem. Soc. 110 4706 (1988) 17. W L Mock, N-Y Shih J. Org. Chem. 51 4440 (1986) 18. H-J Buschmann, K Jansen, E Schollmeyer Thermochim. Acta 317 95 (1988) 19. W L Mock, J Pierpont J. Chem. Soc., Chem. Commun. 1509 (1990) 20.W L Mock, N-Y Shih, J. Am. Chem. Soc. 111 2697 (1989) 21. R Hoffmann,W Knoche, C Fenn, H-J Buschmann J. Chem. Soc., Faraday Trans. 90 1507 (1994) 22. R Neugebauer, W Knoche J. Chem. Soc., Perkin Trans. 2 529 (1998) 23. C Marquez,W M Nau Angew. Chem., Int. Ed. Engl. 40 3155 (2001) 24. W L Mock, T A Irra, J P Wepsiec, T L Manimaran J. Org. Chem. 48 3619 (1983) 25. W L Mock, T A Irra, J P Wepsiec,M Adhya J. Org. Chem. 54 5302 (1989) 26. Y M Jeon, J Kim, D Whang, K Kim J. Am. Chem. Soc. 118 9790 (1996) 27. H-J Buschmann, K Jansen, E Schollmeyer Thermochim. Acta 346 33 (2000) 28. H-J Buschmann, E Schollmeyer J. Incl. Phenom. 14 91 (1992) 29. H-J Buschmann, E Schollmeyer J. Incl. Phenom. 29 167 (1997) 30. S Karcher, A Kornmuller,M Jekel Water Sci.Technol. 40 425 (1999) 31. H-J Buschmann, K Jansen, E Schollmeyer J. Incl. Phenom. Macro. 37 231 (2000) 32. H-J Buschmann, T Wolff J. Photochem. Photobiol., A Chem. 121 99 (1999) 33. D A Dantz, C Meschke, H-J Buschmann, E Schollmeyer Supramol. Chem. 9 79 (1998) 34. M El Haouaj,M Luhmer, Y H Ko, K Kim, K Bartik J. Chem. Soc., Perkin Trans. 2 804 (2001) 35. M El Haouaj, Y H Ko,M Luhmer, K Kim, K Bartik J. Chem. Soc., Perkin Trans. 2 2104 (2001) 36. G Schill Catenanes, Rotaxanes, Knots (New York: Academic Press, 1971) 37. I G Panova, I N Topchieva Usp. Khim. 70 28 (2001) [Russ. Chem. Rev. 70 23 (2001)] 38. Y-M Jeon, D Whang, J Kim, K Kim Chem. Lett. 503 (1996) 39. H-J Buschmann, A Wego, E Schollmeyer,D Dopp Supramol. Chem. 11 225 (2000) 40.C Meschke, H-J Buschmann, E Schollmeyer Polymer 40 945 (1999) 41. C Meschke, H-J Buschmann, E Schollmeyer Macromol. Rapid Commun. 19 59 (1998) 42. D Tuncel, J H G Steinke Chem. Commun. 253 (2001) 43. J W Lee, K Kim, K Kim Chem. Commun. 1042 (2001) 44. S I Jun, J W Lee, S Sakamoto, K Yamaguchi, K Kim Tetrahedron Lett. 41 471 (2000) 45. D Tuncel, J H G Steinke Chem. Commun. 1509 (1999) 46. T C Krasia, J H G Steinke Chem. Commun. 22 (2002) 47. K-M Park, D Whang, E Lee, J Heo, K Kim Chem. Eur. J. 8 498 (2002) 48. S-G Roh, K-M Park, G-J Park, S Sakamoto, K Yamaguchi,K Kim Angew. Chem., Int. Ed. Engl. 38 638 (1999) 49. D Whang, Y-M Jeon, J Heo, K Kim J. Am. Chem. Soc. 118 11 333 (1996) 50. D Whang, K Kim J. Am. Chem.Soc. 119 451 (1997) 51. D Whang, J Heo, C-A Kim, K Kim Chem. Commun. 2361 (1997) 759 52. D Whang, K-M Park, J Heo, V P Ashton,K Kim J. Am. Chem. Soc. 120 4899 (1998) 53. E Lee, J Kim, J Heo,D Whang,K KimAngew. Chem., Int. Ed. Engl. 40 399 (2001) 54. E Lee, J Heo, K Kim Angew. Chem., Int. Ed. Engl. 39 2699 (2000) 55. J W Lee, Y H Ko, S-H Park, K Yamaguchi, K Kim Angew. Chem., Int. Ed. Engl. 40 746 (2001) 56. M Dobler Chimia 38 415 (1984) 57. J Heo, J Kim, D Whang, K Kim Inorg. Chim. Acta 297 307 (2000) 641 (1999) 78 (1998) V P Fedin Koord. Khim. 27 12 (2001) a V P Fedin Eur. J. Inorg. Chem. 167 (2001) Crystallogr., Sect. E 57 m33 (2001) V Klegg Zh. Strukt. Khim. 42 384 (2001) b M N Sokolov, V P Fedin, J G Platas, R Hernandez-Molina, 58. J Heo, S-Y Kim, D Whang, K Kim Angew.Chem., Int. Ed. Engl. 38 59. D Whang, J Heo, J H Park, K Kim Angew. Chem., Int. Ed. Engl. 37 60. D G Samsonenko, A A Sharonova,M N Sokolov, A V Virovets, 61. D G Samsonenko,M N Sokolov, A V Virovets, N V Pervukhina, 62. A V Virovets, D G Samsonenko, M N Sokolov, V P Fedin Acta 63. A V Virovets, D G Samsonenko, D N Dybtsev, V P Fedin, 64. D G Samsonenko, J Lipkowski, O A Gerasko, A V Virovets, A Mederos Eur. J. Inorg. Chem. 2380 (2002) 65. D G Samsonenko, O A Gerasko, Ya Lipkovski, A V Virovets, V P Fedin Izv. Akad. Nauk, Ser. Khim. (10) 1763 (2002) c 66. O A Gerasko, D G Samsonenko, A A Sharonova, A V Virovets, Ya Lipkovski, V P Fedin Izv. Akad. Nauk, Ser. Khim. 325 (2002) c 67. S S Batsanov Zh.Neorg. Khim. 36 3015 (1991) d 68. T Saito, in Early Transition Metal Clusters with p-Donor Ligands (Ed. M H Chisholm) (New York: VCH, 1995) p. 63 69. T Saito Adv. Inorg. Chem. 44 45 (1996) 70. R Hernandez-Molina, A G Sykes J. Chem. Soc., Dalton Trans. 3137 (1999) 71. C D Borman, V P Fedin,M C Hong, G J Lamprecht, C H Kwak, C A Routledge, D M Saysell, A G Sykes Pure. Appl. Chem. 67 305 (1995) 72. R Hernandez-Molina,M N Sokolov, A G Sykes Acc. Chem. Res. 34 223 (2001) 73. V P Fedin, A V Virovets, M N Sokolov, D N Dybtsev, O A Gerasko, W Clegg Inorg. Chem. 39 2227 (2000) 74. D N Dybtsev, O A Gerasko, A V Virovets,M N Sokolov, V P Fedin Inorg. Chem. Commun. 3 345 (2000) 75. M N Sokolov, A V Virovets, D N Dybtsev, O A Gerasko, V P Fedin, R Hernandez-Molina, W Clegg, A G Sykes Angew. Chem., Int. Ed. Engl. 39 1659 (2000) 76. O A Gerasko, A V Virovets, M N Sokolov, D N Dybtsev, A V Gerasimenko,D Fenske,V P Fedin Izv. Akad. Nauk, Ser. Khim. (10) 1654 (2002) c 77. A V Virovets, N V Podberezskaya Zh. Strukt. Khim. 34 150 (1993) b 78. V P Fedin, M N Sokolov, D N Dybtsev, O A Gerasko, A V Virovets, D Fenske Inorg. Chim. Acta 331 31 (2002) 79. V P Fedin, M Sokolov, G J Lamprecht, R Hernandez-Molina, M-S Seo, A V Virovets, W Clegg, A G Sykes Inorg. Chem. 40 6598 (2001) 80. D N Dybtsev, Candidate Thesis in Chemical Sciences, Institute of Inorganic Chemistry, Siberian Branch of Russian Academy of 81. O A Gerasko, A V Virovets, D N Dybtsev, V Klegg, V P Fedin 82. D M Saysell, V P Fedin, G J Lamprecht,M N Sokolov, A G Sykes 83. V P Fedin, V Gramlich, M Worle, T Weber Inorg. Chem. 40 1074 84. M N Sokolov, D N Dybtsev, A V Virovets, K Khegechvailer, 85. D N Dybtsev, O A Gerasko, A V Virovets,M N Sokolov, T Veber, 86. M N Sokolov, A V Virovets, D N Dybtsev, E V Chubarova, 87. M N Sokolov,D N Dybtsev,A V Virovets, V Klegg,V P Fedin Izv. 88. D G Samsonenko, A V Virovets, A A Sharonova, V P Fedin, Sciences, Novosibirsk, 2002 Koord. Khim. 26 512 (2000) a Inorg. Chem. 36 2982 (1997) (2001) V P Fedin Izv. Akad. Nauk, Ser. Khim. 1905 (2000) b V P Fedin Zh. Neorg. Khim. 46 908 (2001) d V P Fedin, D Fenske Inorg. Chem. 40 4816 (2001) Akad. Nauk, Ser. Khim. 1092 (2001) b D Fenske Izv. Akad. Nauk, Ser. Khim. 475 (2001) bO A Gerasko, D G Samsonenko, V P Fedin 760 89. V P Fedin, A V Virovets, D N Dybtsev, O A Gerasko, K Hegetschweiler, M R J Elsegood,W Clegg Inorg. Chim. Acta 304 301(2000) 90. A Flinn, G C Hough, J F Stoddart, D J Williams Angew. Chem., Int. Ed. Engl. 31 1475 (1992) 91. J Kim, I-S Jung, S-Y Kim, E Lee, J-K Kang, S Sakamoto, K Yamaguchi, K Kim J. Am. Chem. Soc. 122 540 (2000) 92. A Day, A P Arnold, R J Blanch, B Snushall J. Org. Chem. 66 8094 (2002) 93. K S Oh, J Yoon, K S Kim J. Phys. Chem. B 105 9726 (2001) 94. K A Kellersberger, J D Anderson, S M Ward, K E Krakowiak, D V Dearden J. Am. Chem. Soc. 123 11 316 (2001) 95. K Jansen, H-J Buschmann, A Wego, D Dopp, C Mayer, H-J Drexler, H-J Holdt, E Schollmeyer J. Incl. Phenom. 96. H J Buschmann, E Cleve, K Jansen, A Wego, E Schollmeyer 97. X X Zhang, K E Krakowiak, G P Xue, J S Bradshaw, R M Izatt 98. H J Buschmann, E Cleve, K Jansen, E Schollmeyer Anal. Chim. 99. R Blanch, A Arnold, A Sleeman, T,White, A Day, in Proceedings Macrocycl. Chem. 39 357 (2001) J. Incl. Phenom. Macrocycl. Chem. 40 117 (2001) Ind. Eng. Chem. Res. 39 3516 (2000) Acta 437 157 (2001) of the World Chemistry Congress, Brisbane, 2001 p. A32 100. C Marquez,W M Nau Angew. Chem., Int. Ed. Engl. 40 4387 (2001) 101. S-Y Kim, I-S Jung, E Lee, J Kim, S Sakamoto, K Yamaguchi, K Kim Angew. Chem., Int. Ed. Engl. 40 2119 (2001) 102. A I Day, R J Blanch, A P Arnold, S Lorenzo, G R Lewis, I Dance Angew. Chem., Int. Ed. Engl. 41 275 (2002) 103. S Lorenzo, A Day, D Craig, R J Blanch, A P Arnold, I Dance Cryst. Eng. Commun. 1 (2001) 104. S Lorenzo, A Day, I Dance, P Turner, R Blanch, A Arnold, J McMurtrie, in Proceedings of the World Chemistry Congress, Brisbane, 2001 p. OFB12 105. K-C Zhang, T-W Mu, L Liu, Q-X Guo Chinese J. Chem. 19 558 (2001) 106. H J Kim, J Heo,W S Jeon, E Lee, J Kim, S Sakamoto, K Yamaguchi, K Kim Angew. Chem., Int. Ed. Engl. 40 1526 (2001) 107. S Y Jon, Y H Ko, S H Park, H-J Kim, K Kim Chem. Commun. 1938 (2001) 108. J Zhao, H-J Kim, J Oh, S-Y Kim, J W Lee, S Sakamoto, K Yamaguchi, K Kim Angew. Chem., Int. Ed. Engl. 40 4233 (2001) a�Russ. J. Coord. Chem. (Engl. Transl.) b�Russ. J. Struct. Chem. (Engl. Transl.) c�Russ. Chem. Bull., Int. Ed. (Engl. Transl.) d�Russ. J. Inorg. Chem. (Engl.

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (9) 761 ± 774 (2002) Proton-conducting electrolyte membranes based on aromatic condensation polymers A L Rusanov, D Yu Likhatchev, KMuÈ llen Contents I. Introduction II. Sulfonated aromatic condensation polymers and membranes based on them III. Alkylsulfonated aromatic condensation polymers and proton-conducting electrolyte membranes based on them IV. Proton-conducting electrolyte membranes based on acid ± base polymer complexes V. Fuel cell applications of proton-conducting membranes based on aromatic condensation polymers Abstract. development of field the in investigations of results The The results of investigations in the field of development and applications of proton-conducting electrolyte membranes and applications of proton-conducting electrolyte membranes based analysed are polymers condensation aromatic on based on aromatic condensation polymers are analysed and and summarised.of properties the to paid is attention Primary summarised. Primary attention is paid to the properties of the the starting polymers, such as the thermal stability, water uptake and starting polymers, such as the thermal stability, water uptake and proton preparation the to approaches General conductivity. proton conductivity. General approaches to the preparation of of aromatic proton high with polymers condensation aromatic condensation polymers with high proton conductivity conductivity are from synthesis sulfonation, including considered, are considered, including sulfonation, synthesis from monomers monomers containing of incorporation groups, acid sulfonic containing sulfonic acid groups, incorporation of alkylsulfonated alkylsulfonated substituents polymer basic ± acid of formation and substituents and formation of acid ± basic polymer complexes.complexes. The references 115 includes bibliography The bibliography includes 115 references. I. Introduction Advanced technologies of the 21st century are thought to be associated with wide use of fuel cells that allow fuels to be utilised in the fuel cell based engines with much higher efficiency com- pared to internal combustion engines.1 Currently, fuel cells are used as power sources in both stationary and mobile (motor vehicles, buses, locomotives, etc.) engines. The most important components of fuel cells are proton-conducting, ion-containing membranes which can operate under severe conditions.The operating conditions include, first, high operating temperatures (sometimes, they exceed 100 8C) and, second, the medium that can be chemically active towards the material the membranes are made of due to the use of aggressive fuels (e.g., methanol and products of its partial oxidation), oxidants (e.g., oxygen) and catalysts, and due to the generation of active radicals at electrodes (especially, at cathode).1 Of particular interest are new proton-conducting ion- exchange membranes based on solid polymer electrolytes.2, 3 A L Rusanov A N Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, ul.Vavilova 28, 119991 Moscow, Russian Federation. Fax (7-095) 135 50 85. Tel. (7-095) 135 63 72. E-mail: alrus@ineos.ac.ru D Yu Likhatchev Materials Research Institute, UNAM, Cirquito Exterior s/n, CU, Apdo Postal 70-360 Coyoacan, 04510 Mexico City, Mexico. Fax (52-555) 616 12 01. Tel. (52-555) 622 45 87. E-mail: likhach@servidor.unam.mx KMuÈ llen Max-Planck-Institute of Polymer Research, Ackermanweg 10, 55128 Mainz, Germany. Fax (49-613) 137 91 00. Tel. (49-613) 137 91 50. E-mail: muellen@mpip-mainz.mpg.de Received 10 June 2002 Uspekhi Khimii 71 (9) 862 ± 877 (2002); translated by A M Raevsky #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n09ABEH000740 761 762 767 771 772 The idea of using organic cation-exchange membranes as solid electrolytes in electrochemical cells was first materialised in 1959.4 Currently, the solid polymer electrolyte fuel cells (SPEFCs) are considered to be the most promising candidate systems out of all types of fuel cells.5±8 These devices are frequently called also as polymer electrolyte membrane or proton exchange membrane fuel cells (PEMFCs).The structure of a PEMFC is schematically shown in Fig. 1. A polymer electrolyte membrane is sandwiched between two noncorrosive porous electrodes. Electrochemical stacks can be fabricated by mechanically compressing individual assemblies across electrically conducting separators. Generally, PEMFCs require humidified gases, viz., hydrogen and oxygen (or air) as fuels for their operation. The electro- chemical reactions occurring at electrodes are as follows: 2H++2e7; at anode: H2 H2O; at cathode: 0.5O2+2H++2e7 H2O+Q1+Q2, overall: H2+0.5O2 where Q1 is the electrical energy and Q2 is the heat energy.Membranes for PEMFCs 9 must possess some specific proper- ties, such as: 7 low permeability with respect to oxygen and hydrogen (to prevent membrane breakdown); 7 high swelling (to provide efficient dissociation of acids and to form a hydrated ionic phase within the thickness of the membrane); 7 good water uptakes at temperatures above 100 8C; Cathode Anode H+ H2 O2 H+H+ H2O H+ Catalytic layer Polymer electrolyte membrane Figure 1. A scheme of a PEMFC.762 7 high ion-exchange capacity sufficient to provide a conduc- tivity of the order of 1071 S cm71 at temperatures near 50 8C; 7 high proton conductivity at temperatures above 100 and below 0 8C; 7 high chemical and mechanical stability for long-term operation under severe conditions (over 2000 h for electric vehicle applications).{ In addition, ion-conducting films must be coated onto catalyst particles with ease.And of course, such membranes have to meet all requirements placed upon commercial products in the fuel cell market. As a rule, proton-conducting polymers are prepared using polymer electrolytes containing negatively charged groups. The polymer electrolytes with sulfate and phosphate groups are most widely used. In dry state, they are rigid-chain, proton-conducting systems.As the content of water increases, the proton conductiv- ity of hydrated polymer electrolytes substantially increases and can be as high as 1072 ± 1071 S cm71. The firstPEMFC(a 1 kWpower plant) used in an operational system was built at General Electric Co. (USA).10 The polymer membrane was made of poly(styrenesulfonic acid) (PSSA) CH CH2 SO3H . n This type of PEMFC was used for primary power sources for the Gemini (USA) spacecrafts in the mid-1960s. The lifetime of these PEMFCs was limited due to the degra- dation of the PSSA membrane under the action of HO2 .radicals. Research on fluorinated polymers 11 ± 14 led to considerable advances in the development of polymer electrolytes. The most widely used fluorinated polymers are prepared by copolymerisa- tion of tetrafluoroethylene and perfluorinated vinyl ethers of the type F, SO3 CF(CF3) O CF2CF2 F2C CF O CF2 accompanied by hydrolysis of fluorosulfonic acid groups.The basic perfluorinated chains of such polymers determine high chemical and thermal stability, while the side chains possess the properties of strong acids. Perfluorinated electrolyte membranes with the general for- mula CF CF2 CF2)x (CF2 O (CF2)n SO3H (O CF2 CF)m are also widely used.x m n Membrane 21 ± 5 2 ± 5 2 10.1 0.3 0 6 ± 10 3 ± 10 1.5 ± 14 3 ± 10 Nafion Flemion Aciplex-S Dow membrane An improved PEMFC for the Gemini spacecrafts was based on a perfluorinated Nafion membrane.This membrane possesses substantially improved characteristics compared to the PSSA membranes; particular types of Nafion membranes are charac- terised by lifetimes of 50 000 h. Different types of Nafion mem- branes have different equivalent masses (grammes of polymer per mole H+), namely, 1200 (Nafion 120), 1100 (Nafion 117 and Nafion 115) and 1000 (Nafion 105). { Optimum durability is ten years. A L Rusanov, D Yu Likhatchev, KMuÈ llen Perfluorinated membranes (Dowmembrane) were developed at Dow Chemical Co. (USA). They have equivalent masses of 800 ± 850 and a dry state thickness of*5 mm. The Flemion membranes with equivalent masses*1000 were developed at Asahi Glass Co. (Japan).5 Aciplex-S membranes developed at Asahi Chemical Industry (Japan) are characterised by equivalent masses of 1000 ± 1200. All the membranes mentioned above, as well as the Neo- septa-F (Tokuyama, Japan) and Gore-Select (W L Gore and Associates, Inc., USA) membranes possess high proton conduc- tivity (1072 ± 1071 S cm71) at water uptakes up to 15 H2O molecules per SO3Hgroup and are characterised by good thermal, chemical and mechanical properties.On the other hand, these membranes are poor ionic conductors at reduced humidity and/or elevated temperatures. For instance, the conductivity of fully hydrated Nafion membranes at room temperature reaches 1072 S cm71; however, it dramatically decreases at 100 8C be- cause of the loss of absorbed water in the membranes. In addition, such membranes tend to undergo chemical degradation at ele- vated temperatures.Finally, their fabrication is rather expensive. Therefore, the development of new solid polymer electrolytes, which combine sufficient electrochemical characteristics and low cost, is of current interest. A promising way of attacking this problem involves preparation of membranes based on aromatic condensation polymers (ACPs). The chemistry of ACPs was characterised by considerable progress in the 1960 ± 1990s.15 ± 26 ACPs have some advantages that make them practically attractive: 7 ACPs are cheaper than perfluorinated polymers, and some of them are commercially available; 7 ACPs containing polar groups have high water uptakes over a wide temperature range; 7 decomposition of ACPs can be to a great extent suppressed by proper molecular design; 7 ACPs are easily recycled by conventional methods.A number of reviews concerning the development of proton- conducting membranes based on polymer electrolytes are avail- able.1, 7, 8, 27 ± 30 They contain information on the advanced mate- rials, their electrochemical properties, water uptakes and thermal stabilities. However, rapid accumulation of newly obtained results gives an impetus to further generalisation of information in this field. In the last decade, research into PEMFCs has been most intensively carried out in the following avenues: 7 development of sulfonated aromatic condensation poly- mers (ACPs) and membranes based on them; 7 development of alkylsulfonated ACPs and membranes based on them and 7 development of acid ± basic polymer complexes and mem- branes based on them.II. Sulfonated aromatic condensation polymers and membranes based on them Aromatic polymers containing sulfonic acid groups can be pre- pared by sulfonation of high-molecular-mass ACPs or their frag- ments and by condensation of monomers containing sulfonic acid groups. 1. Sulfonation of high-molecular-mass aromatic condensation polymers The simplest and most widely used method for the synthesis of sulfonated ACPs involves sulfonation of different classes of polymers, such as poly(1,4-phenylenes),31 ± 33 poly(p-xylyl- ene),34, 35 poly(1,4-oxyphenylenes),36 poly(ether ether ketones) (PEEK),37 ± 44 poly(arylene ether sulfones),3, 45 ± 52 poly(phenylene sulfides),53 poly(phenylquinoxalines),54 ± 56 poly(benzimid- azoles) 57 and some other ACPs.The chemical structures of sulfonated poly(4-phenoxybenzoyl-1,4-phenylene) (S-PPBP) (1), poly(p-xylylene) (S-PPX) (2), poly(phenylene sulfide) (S-PPS) (3),Proton-conducting electrolyte membranes based on aromatic condensation polymers poly(phenylene oxide) (S-PPO) (4), poly(ether ether ketone) (S- PEEK) (5), poly(ether ether sulfone) (S-PEES) (6), arylsulfonated poly(benzimidazole) (S-PBI) (7) and sulfonated poly(phenylqui- noxaline) (S-PPQ) (8) are shown below. O 1 OO NNSO3H ACPs are sulfonated using known sulfonating agents.58 ± 60 In particular, PEEK can be sulfonated with concentrated sulfuric acid,61 chlorosulfonic acid,62 SO3 (either pure or taken in a mixture),42, 48, 62, 63 a mixture of methanesulfonic acid with con- centrated sulfuric acid 64 and acetyl sulfate.65, 66 Sulfonation of ACPs was systematically studied taking a number of polymers (first of all, PEEK and PPBP) as examples.7 It was shown that sulfonation with chlorosulfonic or fuming sulfuric acid is sometimes accompanied by degradation of these polymers.The sulfonation rate of ACPs in sulfuric acid can be controlled by varying the reaction time and the acid concentra- tion.67 This technique allows preparation of target ACPs with sulfonation degrees ranging from 30% to 100% without chemical degradation or cross-linking of the polymers.68 However, it should be noted that direct sulfonation reactions cannot be used for preparation of random sulfonated copolymers and sulfona- tion levels of less than 30%, since sulfonation in sulfuric acid occurs under heterogeneous conditions due to high viscosity of the reaction solutions.40, 61 For this reason, preparation of random copolymers requires the duration of the dissolution process to be shortened to 1 h.The dependences of the degree of sulfonation of PEEK and PPBP 23 on the reaction time at room temperature are shown in Fig. 2. SO3H HO3S CH22 S C O (HO3S)m 3 n OC O 5 SO3H OS O O 6 SO3H NNSO3H 7 SO3H SO3H N O N 8 CH2 n O n n 4 (HO3S)m nn n NN n 400 The solubility of polymers changes as the degree of sulfona- tion increases. For instance, S-PEEK containing 30 mol.% sul- fonic acid groups can be dissolved in DMF, DMSO and N-methyl-2-pyrrolidone (N-MP); at 70% sulfonation, the poly- mer is soluble in methanol and at 100%, in water.Non-sulfonated PPBP is soluble in conventional chlorinated solvents (e.g., chloro- form and dichloromethane), whereas S-PPBP with 30% sulfona- tion is insoluble in these solvents; however, the polymer can be dissolved in DMF,DMSOandN-MP. At sulfonation levels above 65%, S-PPBP swells in methanol and water. Sulfonation of PEEK in concentrated sulfuric acid at room temperature is accompanied by incorporation of no more than one sulfonic acid group into each repeating unit of the poly- mer.62, 65, 69 From the results of FT-IR spectroscopy studies it follows that PEEK is sulfonated at the phenylene ring between the ether groups.Sulfonation of PPBP occurs at the p-position of the terminal phenoxy group. Tsuchida et al.70, 71 reported the synthesis of poly(thiopheny- lenesulfonic acid) containing up to two sulfonic acid groups per repeating unit. Polymerisation of 4-(methylsulfinyl)diphenyl sul- fide in sulfuric acid upon heating or in the presence of SO3 resulted in a sulfonated poly(sulfonium cation), which was then converted into the corresponding sulfonated poly(phenylene sulfide). Degree of sulfonation (mol.%) a 100 80 60 40 20 b 100 80 60 40 20 0 200 Figure 2. Degree of sulfonation of PEEK (a) and PPBP (b) as a function of reaction time at room temperature.7 Sulfonation is a kind of electrophilic substitution reaction; therefore, it strongly depends on the nature of substituents in the rings, namely, electron-donor substituents favour the reaction, whereas electron-acceptor substituents do not.For instance, in the case of PPBP the terminal phenyl ring of the side chain can be sulfonated under mild conditions that are comparable with the sulfonation conditions of PEEK. In contrast to this, the phenyl ring of the substituent in poly(4-benzoyl)-1,4-phenylene (PBP), which contains an electron-acceptor carbonyl group, cannot be sulfonated under these conditions.7 The sulfonation level of PPBP and PEEK reaches nearly 80% within 100 h. The degree of sulfonation of PPBP is saturated at 85%, whereas that of PEEK can be as high as 100%.This is thought to be due to steric hindrances to further sulfonation of PPBP in a viscous sulfuric acid solution. 763 600 t /h764 S m=1¡À2. The course of sulfonation was controlled by varying the reaction time, the temperature and/or by adding SO3. Polymer electrolytes thus obtained are soluble in water and methanol and can form transparent films. Novel polymer electrolytes exhibiting high proton conduc- tivity (higher than 1072 S cm71) were prepared by sulfonation of poly(ether sulfone) (PES).72, 73 In these polymers the protons of the sulfonic acid groups are partially replaced by metal ions (Mg, Ti, Al, Ln), which leads to extension of the durability of the electrolytes.2. Synthesis of aromatic condensation polymers based on sulfonated monomers Sulfonated ACPs are prepared both by direct sulfonation and by the polycondensation and polycyclocondensation of sulfonated compounds. In particular, sulfonated PEEK were prepared by the reac- tions of sulfonated hydroquinone with difluoro-substituted aro- matic compounds containing carbonyl groups:74 nHO SO3K O, R= CO An analogous procedure was employed in recent studies 75, 76 on the synthesis of poly(arylene ether sulfones) in the reactions of sulfonated 4,40-dichlorodiphenyl sulfone with various bisphenols. n Cl HO3S S Me O HSO¡¦4 + SO3/H2SO4 S S Me n HSO¡¦4 + NaCl S S NaOH Me (SO3H)m (SO3H)m n 9 H+ S S (SO3H)m (SO3Na)m 2n 2n MP, C6H6 OH+n F F R K2CO3, 200 8C HCl R O n SO3K R O O n SO3H C C .O O SO3H OS Cl+nHO Ar OH O A L Rusanov, D Yu Likhatchev, KMu�� llen SO3H O O Ar O SO n HO3S Ar= . , The use of m-aminophenol as an additive along with bis- phenols allowed the preparation of poly(arylene ether sulfones) with terminal amino groups.77 Sulfonated poly(thiophenylene sulfones) were prepared by the interaction of sulfonated 4,40-difluorodiphenyl sulfone with 4,40- dimercaptobenzophenone.78 Not only homopolymers, but also copolymers were obtained (in the latter case, a fraction of sulfonated 4,40-difluorodiphenyl sulfone was replaced by non- sulfonated monomer) (Scheme 1). Using this approach, one can not only prepare polymers with regular arrangement of sulfonic acid groups, but sometimes introduce a larger number of sulfonic acid groups into the ACP macromolecules compared to the sulfonation of ACPs in the last stage of the synthesis.Sulfonated monomers were also used for the synthesis of sulfonated polyimides.79, 80 In particular, a sodium salt of sulfo- nated bis-4-[(3-aminophenoxy)phenyl]phenylphosphine oxide was used for the preparation of sulfonated polyimides (Scheme 2).79 Of considerable interest is the use of 4,40-diamino-2,20-diphe- nylsulfonic acid 80 ¡À 82 produced on a semi-industrial scale as a sulfonated monomer for the preparation of polyimides. The reactions of a mixture of this monomer and 4,40-diaminodiphenyl- methane and 4,40-diaminodiphenyloxide with diphenyloxido- 3,30,4,40-tetracarboxylic acid dianhydride resulted in sulfonated polyimides 80 with the following formulae: O SO3H O O N N HO3S O l O O O O N N R O O m n R=O, CH2.Poly(naphthylimides) containing six-membered imide rings in the backbones 80 ¡À 86O SO3H ON N O O HO3S n are characterised by substantially improved chemical resistance compared to analogous poly(phthalimides).87 ¡À 89 Almost all studies on the synthesis of poly(naphthylimides) based on 4,40-diamino-2,20-diphenylenesulfonic acid were aimed at preparing copolymers with controlled properties that can be varied over a wide range. Poly(naphthylimides) can be synthesised from sulfonated diamines, e.g., 4,40-diaminodiphenylamino-2-sulfonic acid.90 HN NH2 H2N HO3SProton-conducting electrolyte membranes based on aromatic condensation polymers O mF SO NaO3S HO3S O n H2N O NO 3.Properties of sulfonated aromatic condensation polymers The most important properties of sulfonated ACPs are their thermal stability, water uptake and proton conductivity. PEMFCs and other electrochemical devices operating in the temperature range 100 ± 200 8C require polymer electrolyte mem- branes characterised by fast proton transfer. Operation of PEMFCs at elevated temperatures has a number of advantages. It causes an increase in the rates of the fuel cell reactions and reduces catalyst poisoning with absorbed carbon monoxide, thus reducing the demand for catalysts.Thermal stability of proton-conducting polymer electrolyte membranes is an important characteristic, which determines the possibility of their use for fuel cell applications. The thermal stability of polymer membranes based on S-PPBP was studied 7 by heating the samples followed by elemen- tal analysis (thermogravimetric analysis, or TGA, at a heating rate of 10 deg min71 in nitrogen atmosphere) (Fig. 3). According to the results of TGA studies, S-PPBP showed a mass loss of nearly *20% in the temperature range between 250 and 400 8C, which corresponds to the decomposition of sulfonic acid groups. Dm (%) 100 80 60 100 Figure 3. TGA curve of S-PPBP with 80 mol.% sulfonation level.7 SO3H OS F+ k F O HO3SSO3Na O O S S O O k SO3H O O S S O O k O O P NH2+n EtO HO SO3Na O CF3 C O N CF3 O 500 300 T /8C OS F +n HS O S m S m O CF3 CCF3 O O O P SO3Na The dependence of the degradation temperature, Td, of S-PPBP and S-PEEK on the degree of sulfonation is presented in Fig.4. Degradation of sulfonated polymers was observed between 250 and 350 8C, i.e., at temperatures that are much lower than the decomposition temperatures of non-sulfonated PPBP and PEEK. As the degree of sulfonation increased, the degradation temperatures decreased from 500 down to 300 8C for S-PEEK and from 500 down to 250 8Cfor S-PPBP. The results of elemental analysis of the residues indicated a dramatic (nearly tenfold) decrease in the sulfur content in the polymers after heating at temperatures above 400 8C.These data confirm that thermal decomposition occurs by desulfonation mechanism. No thermal decomposition of sulfonated polymers was observed at temperatures below 200 8C. This means that the thermal stabilities of the polymers are sufficient for fuel cell applications even at high sulfonation levels.7 Td /8C 500 300 100 0 20 60 Sulfonation level (mol.%) Figure 4. Degradation temperature of S-PPBP (1) and S-PEEK (2) as function of sulfonation level.7 SH O S SO OSO O OEt OH O n 40 765 Scheme 1 DMAA K2CO3H+ n S n Scheme 2 o-DCB,MP N2, 180 8C 12 100 80766 Other proton-conducting polymer electrolytes based on sulfo- nated aromatic polymers also show the onset of thermal degrada- tion at temperatures between 200 and 400 8C.Desulfonation of arylsulfonic acids occurs with ease on heating their aqueous solutions to 100 ± 175 8C. Therefore, desulfonation imposes lim- itations on the thermal stability of sulfonated aromatic polymer electrolytes. Mention may be made that the presence of bulky substituents attached to the phenyl rings can to some extent favour an increase in the temperature of the onset of thermal degradation. According to Tsuchida et al.,70 highly sulfonated poly(phe- nylene sulfide) exhibits higher thermal stability compared to other sulfonated aromatic polymer electrolytes.This conclusion was based on the results of a TGA study of the thermal stability of poly(thiophenylenesulfonic acid) with different degrees of sulfo- nation. The degradation temperature of a highly sulfonated polymer (degree of sulfonation m=2.0) is 265 8C, which is 125 8C higher than that of the low sulfonated polymer (m=0.6). The C7S bond in the highly sulfonated polymer is stronger due to the presence of two electron-acceptor sulfonic acid substituents attached to each benzene ring. The initial mass loss of this polymer at 265 ± 380 8C is only 13%, which corresponds to the loss of two water molecules per repeating unit. Therefore, the desulfonation reaction in this polymer slows down upon introduction of electron-acceptor. Water is carried out in the fuel cells via humidified gas streams (H2, O2) and enters electrodes as a result of gas diffusion.A mixture of liquid water and water vapours passes through each electrode towards the electrode/electrolyte interface and crosses it, thus assisting the hydration of the electrolyte membranes. Oxygen reduction at the cathode provides an additional source of water in the electrolyte. Water transport in the membrane occurs in two ways, viz., due to electro-osmotic drag of water by proton transfer from anode to cathode and due to diffusion of water molecules down concen- tration gradients that build up. Optimum hydration level of electrolyte membranes is a key factor for normal fuel cell operation: if the electrolyte membrane is too dry, its conductivity decreases, whereas an excess of water in thebrane can lead to cathode flooding.In both cases the fuel cell performance reduces. Absorption of water vapour by polymer films prepared from S-PEEK and S-PPBP was studied by placing the films in atmo- spheres with different humidities and measuring the equilibrium water content. The results obtained were found to be close to those reported in an analogous study of Nafion membranes.10 The dependence of the water uptakes of the S-PEEK and S-PPBP films on the relative humidity at room temperature is shown in Fig. 5. Assuming that water activity and water content in the mem- brane obey the Raoult law, the activity coefficient of water in the [H2O] (mass%) 30 1234 20 100 100 75 50 25 Relative humidity (%) Figure 5.Water uptake of S-PPBP (1 ± 3) and S-PEEK (4) at room temperature as function of relative humidity.7 Concentration of SO3H groups in the polymer (mol.%): 30 (1), 65 (2), 80 (3) and 65 (4). A L Rusanov, D Yu Likhatchev, KMuÈ llen polymer is larger than unity at relative humidities exceeding a particular value. The equilibrium content of water in S-PEEK and S-PPBP increases as the sulfonation level increases. At relative humidities in the range from 0% to 50% (first region), a relatively small increase in the water uptake is observed, whereas an increase in the relative humidity from 50% to 100% (second region) leads to a much greater increase in the water uptake. The first region corresponds to water uptake due to solvation of the proton and sulfonate ions.During solvation, water is involved in the interaction with ionic components of the polymer. These interactions overcome the tendency of the polymer to exclude water due to its hydrophobic nature and resistance to swelling.7 The second region corresponds to the uptake of water involved in polymer swelling. The content of water in S-PPBP (65 mol.% sulfonation) is larger than in S-PEEK with the same sulfonation level. At a relative humidity of 100% and room temperature, the content of water in S-PPBP and S-PEEK is 8.7 and 2.5 molecules per sulfonic acid group, respectively. Picnometric measurements showed that the densities of the polymers with a sulfonation level of 65 mol.% were 1.338 (S-PEEK) and 1.373 g cm73 (S-PPBP). According to the results obtained by scanning electron microscopy, both polymers exhib- ited very close characteristics of their surfaces and fracture surfaces.The difference in water uptake between S-PEEK and S-PPBP can be explained by the flexibility of the phenoxybenzoyl group in the side chain of S-PPBP, which favours water permeation into the polymer and water absorption in the terminal sulfonic acid groups. The water uptake of S-PPBP is comparable with that of Nafion membranes. DTA studies revealed a rather strong interaction between the water molecules in sulfonated hydrocarbon polymers and their sulfonic acid groups, which leads to high proton conductivities at high temperatures and low humidities.The proton conductivity of sulfonated poly(phenylene sulfide) is 1075 S cm71 at room temperature and a relative humidity of 30%. It exponentially increases as the relative humidity increases and reaches a value of 261072 S cm71 at 94% humidity (Fig. 6). In this case, the content of water in the polymer is 10.3 molecules per sulfonic acid group. The maximum conductivity of sulfonated poly(phenylene sulfide) (m=2.0) at 80 8C was 4.561072 S cm71. Experiments 7 on water absorption in the S-PEEK and S-PPBP films showed that the proton conductivities of the films containing equilibrium amounts of absorbed water depend on the relative humidity. Figure 7 presents the plots of the proton conductivities of S-PPBP and S-PEEK with different sulfonation levels as a function of the relative humidity.As can be seen, the proton conductivities of the films increase with the relative log s (S cm71) 0 72 74 76 100 80 60 40Relative humidity (%) Figure 6. Proton conductivity of compound 9 (m=2.0) at room temper- ature as function of relative humidity.70Proton-conducting electrolyte membranes based on aromatic condensation polymers log s (S cm71) 72 74 76 78 710 12345 712 80 40 20 100 0 60 Relative humidity (%) Figure 7. Proton conductivity of S-PEEK (1) and S-PPBP (2 ± 5) with different sulfonation levels as a function of relative humidity at room temperature.7 Sulfonation level (mol.%): 65 (1), 30 (2), 65 (3), 80 (4) and 85 (5). humidity and water uptake and can be as high as 1075 S cm71 (for S-PEEK). The proton conductivities of S-PEEK and S-PPBP with equal degrees of sulfonation (65 mol.%) at a 100% relative humidity can be compared using the plots shown in Fig.8. As can be seen, the proton conductivity and water uptake of S-PPBP are much higher than those of S-PEEK. Moreover, the proton conductivity of S-PEEK dramatically decreases at temperatures above 100 8C, whereas that of S-PPBP is much less temperature dependent. Sulfonated poly(phenylene sulfide) and S-PPBP exhibit stable proton conductivities at elevated temperatures. For this reason, they are considered as promising polymers for creation of proton- conducting electrolyte membranes to be used at elevated temper- atures and low humidities.On the contrary, the conductivity of perfluorinated polymer electrolytes usually appreciably decreases with increasing temper- ature, that is, the conductivity of such electrolytes at 80 8C is by an order of magnitude lower than at 60 8C. Perfluorinated polymer membranes become less conducting at high temperatures, since the loss of water causes the channels to collapse, thus making proton transport more difficult. In particular, the proton conductivity of Nafion membranes at temperatures above 100 8C dramatically decreases due to their dehydration. In Fig. 9, we present the temperature dependences of the proton conductivity of S-PEEK with a sulfonation degree of 85 mol.% at different relative humidity values. log s (S cm71) 0 1 72 74 2 76 78 710 712 3.0 2.6 103T71 /K71 2.4 2.8 Figure 8.Temperature dependences of proton conductivity of S-PPBP (1) and S-PEEK (2) with the same degrees of sulfonation (65 mol.%) at a relative humidity of 100%.7 767 log s (S cm71) 0 3 71 2 72 1 732.8 3.1 2.9 3.0 103 T71 /K71 Figure 9. Temperature dependence of proton conductivity of an S-PEEK membrane with sulfonation level of 85% at a relative humidity of 50% (1), 70% (2) and 90% (3).7 Similarly to Nafion membranes, the proton conductivity of S-PEEK substantially decreases as the humidity decreases.74 The dependence of proton conductivity on humidity reflects a ten- dency of S-PEEK to absorb water vapours.This can be attributed to the `liquid' proton conductivity mechanism when protons are transported in the form of hydronium ions through water-filled pores in the membrane.30 S-PEEK samples exhibited a slight increase in conductivity with increasing temperature at all relative humidities (50%, 70% and 90%). This can be due to the strong interaction between the sulfonic acid groups and the absorbed water molecules. Proton-conducting polymer electrolyte membranes based on ACPs such as S-PPBP and sulfonated poly(phenylene sulfide) contain rather large amounts of bound water. This seems to be the reason for such a salient feature of these membranes as an increased proton conductivity at high temperatures and/or low humidities.This conclusion was confirmed by the results of differential scanning calorimetry (DSC) studies of these systems.7 III. Alkylsulfonated aromatic condensation polymers and proton-conducting electrolyte membranes based on them The major drawback of sulfonated proton-conducting polymer electrolytes is their degradation at 200 ± 400 8C due to desulfona- tion.By introducing alkylsulfonated substituents into the macro- molecules of aromatic polymers one can prepare thermostable proton-conducting polymers. Their electrochemical properties can be controlled by varying the number of substituents and the length of alkyl chains. The water uptake and proton conductivity of alkylsulfonated polymers are close to those of sulfonated electrolytes that exhibit high thermal and chemical stability and mechanical strength.Poly(p-phenyleneterephthalamido-N-propylsulfonate) and poly(p-phenyleneterephthalamido-N-benzylsulfonate) were syn- thesised based on related polyamides containing reactive NH groups.91 The polyamides were modified by treating with NaH and DMSO91, 92 and the polyanion obtained was introduced in the reaction with 1,3-propane sultone (Scheme 3). An analogous approach was employed in the modification of poly(benzimidazoles) (PBI).93 ± 100 Yet another synthetic route to sulfonated PBI involves treat- ment of the above-mentioned polyanion with 4-bromobenzyl- sulfonate resulting in poly[2,2 0-m-phenylene-bi(N-benzylsulfo- nato)benzoimidazolo-5,5 0-diyl] (Scheme 4). Compared to the starting polymers, alkylsulfonated PBI are more soluble in polar organic solvents (DMAA or DMSO).The solubility depends on the degree of alkylsulfonation. The degree of alkylsulfonation as a function of the 1,3- propane sultone : PBI ratio is plotted in Fig. 10. The degree of alkylsulfonation of NH groups in PBI was estimated based on the results of 1H NMR study and elemental analysis. This parameter can be controlled with ease by varying768 Degree of alkyl- O O NaH S S Na+ CH¡¦ H3C CH3 H3C 2 O O O O S Na+ CH¡¦ H3C 2 C C CHN NH n O O OC C C HN NH x Scheme 4 N N DMAA, LiH HN HN n N N N77 N n O SO2 N N N N n SO3H HO3S CH2Br HO3S N N N N n SO3H HO3S the 1,3-propane sultone : PBI ratio.For instance, the alkylsulfo- nation level can be as high as 60 mol.% at a 1,3-propane sultone : PBI ratio of 5.0. 80 60 40 200 10.0 5.0 2.5 7.5 Ratio [1,3-propane sultone] : [PBI] Figure 10. Degree of alkylsulfonation of PBI as function of 1,3-propane sultone : PBI ratio.7 sulfonation (mol.%) A L Rusanov, D Yu Likhatchev, KMu�� llen O O 7 C7 N N n O C N N y SO3H HO3S An attempt to synthesise ethylphosphorylated PBI using the above-mentioned treatment of PBI at NH groups (Scheme 5) was reported.7 The substitution reaction at the NH sites of benzimi- dazole rings was performed successfully, but the resulting polymer was insoluble in organic solvents. The reason can be aggregation of phosphoric acid groups during the substitution reaction.Ethylphosphorylated PBI exhibited high proton conductivity (1073 S cm71) even in the pellet form. According to the results obtained, the presence of polar phosphoric acid groups enhances the proton conductivity of the polymer electrolytes. Alkylsulfonation and arylsulfonation of the starting aromatic polymers was aimed at increasing their water uptakes and proton conductivities while retaining high thermal stabilities. The poly- mers obtained were studied by TGA in inert and oxidative atmospheres.96 The parent PBI exhibits a very high thermal stability. In an inert atmosphere, the onset of its degradation occurs at 650 8C, a 5%mass loss is observed at 700 8C, while more than 80% of the initial mass is retained at 800 8C.Introduction of substituents that are not conjugated with the polymer backbones reduces the degradation temperature in an inert medium, which is consistent with the expectations. The degradation of poly[2,2 0-m- phenylene-bi(N-benzylsulfonato)benzimidazolo-5,5 0-diyl] with 22% substitution begins at 480 8C while the onset of mass loss process of poly-{2,2 0-m-phenylene-bi[N-(3-propylsulfo)benz- imidazolo-5,5 0-diyl]} with a substitution level of 54% is observed at 450 8C. After removal of the substituents the degradation slows down so nearly 50%¡À 60% of the initial mass is retained at 800 8C. The degradation of PBI in an oxidative atmosphere (dry air) begins at 520 8C, which is nearly 100 8C lower than the degrada- tion temperature of this polymer in inert medium.N N HN HN N N 7N N7NN PO(OH)2 Scheme 3 SO2 Scheme 5 LiH DMAA, 85 8C n Cl(CH2)2PO(OH)2, NEt3 DMAA, 21 8C n NNPO(OH)2 nProton-conducting electrolyte membranes based on aromatic condensation polymers The degradation temperatures of substituted PBI in oxidative media are close. For all polymers, the mass loss in air is much larger than in nitrogen and the amount of the residual char is much smaller. This is first of all due to the lower stability of the starting PBI in dry air and to some extent to the introduction of substituents. In an inert atmosphere, poly(p-phenylene terephthalamide) (PPTA) is stable below 550 8C. Rapid mass loss of the polymer (up to 50% of the initial mass) begins at 600 8C.After modifica- tion with propylsulfonate side groups (a 66% substitution) the polymer is stable below 400 8C; only 40% of the initial mass is retained at 800 8C. The benzylsulfonated derivative of PPTA with 66% substitution level is more thermally stable compared to the propylsulfonated derivative; its degradation begins at 470 8C. A decrease in mass of the sample down to 50% of initial mass is observed at 800 8C. The degradation temperature of PPTA in a dry air atmosphere is 70 8C lower than in nitrogen atmosphere.96 Comparison of degradation processes of benzylsulfonated PPTA with a 66% substitution in air and in nitrogen showed that the degradation in air begins at a lower temperature.The major difference in behaviour is that the mass loss is larger while after the initial mass loss it is much smaller at high temperatures, which is due to oxidative degradation of the polymer chains. Introduc- tion of substituents into aromatic polymers reduces their thermal stability irrespective of the medium in which degradation occurs. This is the expected manner of changes in properties, since the side groups, especially sulfonic acid groups, are not stabilised by conjugation with the polymer backbones. Gieselman and Reynolds concluded 96 that the benzylsulfo- nate side group is more stable than the propylsulfonate group irrespective of the structure of the polymer backbone. This suggests that the side group cleavage point is not only at the N7C bond.A TGA study of benzylsulfonated PBI with 75% degree of sulfonation in air at a heating rate of 1 deg min71 showed 97 that introduction of benzylsulfonate groups into the polymer reduces its thermal stability. In this case, thermal degradation begins at 370 8C while the mass loss in the temper- ature range 370 ± 420 8C is attributed to the degradation of the sulfonic acid groups. The degradation mechanism of these poly- mer electrolytes seems to be very complex, since the results of TGA studies are affected by the residual water, impurities, sulfonation level and measurement conditions. In air, arylsulfonated PBI is stable up to 350 8C (see Ref. 97), while benzylsulfonated PBI is stable up to 500 8C (see Ref.96). These results are hard to compare because of different degrees of sulfonation of the PBI samples under study. One can assume that benzylsulfonated PBI is less stable than propylsulfonated PBI due to the presence of weak aryl7S bond. In fact, the degradation temperature of benzylsulfonated PBI is comparable with those of polymer electrolytes obtained by sulfonation with sulfuric acid. The thermal stability of anhydrous propylsulfonated PBI (PBI-PS) in a nitrogen atmosphere was studied 7 by TGA at a heating rate of 5 deg min71. Prior to analysis, all samples were vacuum oven dried at 60 8C for 48 h. However, this polymer is hygroscopic and again rapidly absorbs water after vacuum oven drying. Because of this, it was dried in situ and then differential thermal analysis was immediately performed.In contrast to PBI, degradation of PBI-PS was observed in the temperature range 400 ± 450 8C. The decomposition temperature of PBI-PS decreases as the degree of alkylsulfonation increases to 400 8C (Fig. 11); however, it is higher than the degradation temperature of perfluorinated polymer electrolytes (nearly 280 8C). Degradation of PBI-PS was studied by elemental analysis and FT-IR spectroscopy. It was found that the intensities of SO stretching vibrations decreased after heating the PBI-PS samples above 400 8C for 1 h. These results are close to those reported by Gieselman and Reynolds 96 who found that the degradation of PBI-PS occurs due to desulfonation. Hence, alkylsulfonated PBI is more thermally stable than sulfonated aromatic polymer 769 Td /8C 800 600 400 80 40 20 60 0 Alkylsulfonation level (mol.%) Figure 11.Decomposition temperature of PBI-s function of alkylsul- fonation level.7 electrolytes characterised by degradation temperatures lying between 200 and 350 8C. The thermal stability of alkylsulfonated polymer electrolytes can be attributed to strong chemical bond between the alkyl and the sulfonic acid groups. The introduction of alkylsulfonic acid groups into thermostable polymers involving alkane sultone is one of the most important approaches to the preparation of thermostable proton-conducting polymer electro- lytes. Introduction of arylsulfonic and alkylsulfonic acid groups into aromatic polymers induces water absorption and makes them more hygroscopic.The water uptake of PBI-PS was determined 7 by measuring the mass of the polymer before and after hydration. In Fig. 12, we present the dependence of the water uptake of PBI-PS on the relative humidity. As can be seen, the water uptake changes with the relative humidity. The equilibrium water uptake of PBI-PS increases as the relative humidity and degree of alkylsulfonation increases. The water uptake of PBI-PS with an alkylsulfonation level of 73.1 mol.% is 11.3 H2O molecules per SO3H group at room temperature and a relative humidity of 90% (cf. 11.0 H2O molecules per SO3H group for Nafion 117 mem- branes under the same conditions). Water uptake /H2O molecules per SO3H group 105 123 0 75 50 25 Relative humidity (%) Figure 12.Water uptake of PBI-PS as a function of relative humidity at alkylsulfonation levels of 49.3 mol.% (1), 61.5 mol.% (2) and 73.1 mol.% (3).7This procedure was also employed for the synthesis of butylsulfonated and (methyl)propylsulfonated PBI (PBI-BS and PBI-MPS, respectively) via butane sultone and methylpropane sultone. The water uptakes of these polymers differ from that of PBI-PS and are 19.5 (PBI-BS) and 27.5 (PBI-MPS) H2O mole- cules per SO3H group at a relative humidity of 90%. The water uptakes of alkylsulfonated PBI depend on the length of alkyl chains and on the degree of chain branching, that is, as the chain length and the degree of alkyl chain branching increase, the water uptake also increases.This is thought to be associated with the770 greater flexibility of long alkyl chains and with larger amount of water absorbed in the cavities between the branched chains. The specific role of the absorbed water in polymer electrolytes and the physical state of the water absorbed by polymer electro- lytes were studied by IR 101 and 1H NMR spectroscopy (low- temperature relaxation time measurements) 102 and DSC.7 The DSC curve of a hydrated PBI-PS film (73.1 mol.%) containing 11.3 H2O molecules per SO3H group is shown in Fig. 13. T1 Endo Exo T2 50 7150 0 T /8C 7100 750 Figure 13. DSC curve of hydrous PBI-PS (73.1 mol.%) film containing 11.3 H2O molecules per sulfonic acid group.7 T1 is the freezing temperature (736.6 8C) and T2 is the melting temper- ature (721.6 8C).The DSC curve of anhydrous PBI-PS exhibited no peaks, whereas the DSC curve of hydrated PBI-PS exhibited two peaks corresponding to phase transitions of absorbed water at 736.6 and 721.6 8C that were attributed to the freezing and melting temperatures of the absorbed water, respectively. A study of hydrated Nafion membranes under the same conditions revealed a phase transition at 0 8C. These results indicate that the adsorbed water in the Nafion membrane is bound to a lesser extent compared to PBI-PS which can exist in the hydrated state even at elevated temperatures. Wet PBI-PS films possess no electron conduction despite the fact that the main polymer chains are conjugated.To elucidate the nature of charge carriers in PBI-PS, the conductivity of PBI-PS films containing H2O and D2O was measured.7 The results of measurements are presented in Fig. 14. As can be seen, the conductivity of the films containing water increased with increas- ing water uptake and was higher than that of the PBI-PS films containing D2O in the same temperature range. This points that the charge carrier in hydrated PBI-PS is a proton (hydronium ion).The temperature dependence of proton conductivity of PBI- PS containing the equilibrium amount of water is shown in 72 1 74 2 76 3.0 2.8 2.6 103T71 /K71 2.4 Figure 14. Temperature dependence of proton conductivity of PBI-PS (61.5 mol.%) films containing H2O (1) and D2O (2).7 log s (S cm71) log s (S cm71) 72 74 log s (S cm71) 72 74 76 782.2 Figure 15.Temperature dependences of proton conductivity of PBI-PS films with the same water uptake (48%) and different degrees of sulfonation (a) and with the same degree of sulfonation (73.1%) and different water uptakes (b);7 (a): degree of sulfonation (mol.%): 49.3 (1), 61.5 (2) and 73.1 (3); (b): water uptake: 11.2 (1), 25.0 (2), 29.0 (3) and 48.0 (4). Fig. 15. Hydrated PBI-PS exhibits a high proton conductivity at room temperature. The conductivity of a PBI-PS sample contain- ing 3.1 H2O molecules per SO3H group reached 1075 S cm71 at 80 8C and decreased slightly at higher temperatures due to a small loss of water (*10 mass %).The conductivity of a PBI-PS film containing more than 5.2 H2O molecules per SO3H group increased as the temperature increased and was as high as 1073 S cm71 at temperatures above 100 8C. The proton conduc- tivity of a PBI-PS film containing 11.3 H2O molecules per SO3H group was*1073 S cm71. The water uptake of PBI-PS films placed in an atmosphere with a relative humidity of 90% was comparable with that of Nafion membranes. The proton conductivity of Nafion mem- branes was as high as 1073 S cm71 at room temperature; how- ever, it decreased due to the loss of absorbed water at temperatures above 100 8C. In contrast to this, hydrated PBI-PS exhibited a high proton conductivity at temperatures above 100 8C. The large water uptake and high proton conductivity of PBI- PS at temperatures above 100 8C are due to the specific properties of the polymer and the physical state of absorbed water.The proton conductivity of benzylsulfonated PBI at different values of relative humidity has been studied.102 It was found that the proton conductivity increases as the degree of substitution increases. The polymer with a 75% substitution level exhibited a conductivity of *1072 S cm71 at 40 8C and a relative humidity of 100%. The results obtained in the above-mentioned studies suggest that alkylsulfonated aromatic polymer electrolytes exhibit suffi- cient thermal stabilities for fuel cell applications at 80 8C(a typical operating temperature for perfluorinated polymer electrolyte membranes).The water uptakes and proton conductivities of these polymers are close to the corresponding values for perfluori- nated polymer electrolytes at temperatures below 80 8C but are larger than the latter at temperatures above 80 8C. The absorbed water molecules are more strongly bound to alkylsulfonated rather than perfluorinated polymers. One can assume that this is related to the difference in the absorption mechanisms and to the physical state of the absorbed water in PBI-PS and perfluorinated polymer electrolytes. A L Rusanov, D Yu Likhatchev, KMuÈ llen a 123 b 1234 2.6 3.0 103T71 /KProton-conducting electrolyte membranes based on aromatic condensation polymers IV. Proton-conducting electrolyte membranes based on acid ± base polymer complexes Proton-conducting membranes used in PEMFC operate under severe conditions (see above).Recently, complexes of basic polymers with strong acids have attracted a considerable interest. Such complexes are characterised by stable electrochemical prop- erties and large water uptakes at high temperatures. Recently, new proton-conducting polymer electrolyte mem- branes based on PBI ± orthophosphoric acid complexes have been proposed for use in PEMFCs.103 ± 106 N N H3PO4 HN NH n 10 The most important advantages of this polymer electrolyte over perfluorinated polymer electrolytes and other acid ± basic polymer complexes are that PBI/H3PO4 possesses conductivity even at low activity of water and high thermal stability of these systems.The materials based on these complexes are expected to operate over a wide range from room to high temperature in both humid and dry gas. Such complexes are prepared by immersing PBI films into phosphoric acid solutions. In particular, the preparation of PBI ± strong acid complexes by immersion of PBI films into solutions of strong acids in methanol was reported.107, 108 The absorption level of strong acid molecules increased with an increase in the concentration of the strong acid and reached up to 2.9 molecules per repeating unit for polymer complexes 10. An IR spectroscopy study of the complexes revealed that the acid molecules, except for H3PO4, protonate the nitrogen atoms in the imidazole ring.Phosphoric acid (H3PO4) is incapable of protonating the imidazole groups in PBI but interacts with them via the formation of strong hydrogen bonds between NH and OH groups. PBI films doped with phosphoric acid were prepared by immersion of PBI films in aqueous solutions of phosphoric acid for at least 16 h.103 ± 106 Upon equilibration in an 11 M H3PO4 solution a doping level of *5 phosphoric acid molecules per repeating unit of the polymer was achieved. Anhydrous sulfonated aromatic polymers are highly brittle. Recently,109 new materials with high mechanical strength were reported. They were prepared using polymer blending technique by combining PBI and sulfonated polymers (S-PEEK or ortho- sulfonated polysulfone 11). Such polymer blends exhibit high proton conductivities, moderate swelling values and high thermal stabilities.The specific interaction of SO3H groups with basic nitrogen atoms was confirmed by FT-IR spectroscopy. The acid ± base interaction between the sulfonated polymer and PBI provided a material with high mechanical strength and thermal stability. The thermal stability of PBI ± strong acid polymer complexes was studied by TGA and DTA. Figure 16 presents the TGA curves of polybenzimidazole and its complexes with strong acids. As can be seen, PBI exhibits an extremely high thermal stability over the entire temperature range. Small mass losses by all samples at temperatures below 200 8C are due to the loss of water and solvent present in the membranes.Typical proton-conducting polymer electrolytes undergo con- siderable degradation in the temperature range under study. SO3H O Me O O S C O Me n 11 771 Dm (%) 100 80 60 40 12345 20 0 300 200 100 400 T /8C Figure 16. TGA curves of PBI (1) and its complexes with H3PO4 (2), H2SO4 (3), MeSO3H (4) and EtSO3H (5).7 A decrease in the degradation temperature of polymer complexes 10 was expected because of the complexation of acid molecules which easily corrode and oxidise the polymer macromolecules. However, no degradation was observed in nitrogen atmosphere. At the same time, thermal decomposition of PBI complexes with H2SO4, MeSO3H and EtSO3H begins at 330, 240 and 220 8C, respectively. After thermal decomposition of these polymer com- plexes in the temperature range 220 ± 400 8C the residues were 50% of the initial masses of the samples.Therefore, complexation of PBI with H2SO4, MeSO3H and EtSO3H results in a loss of thermal stability. The decomposition of complexes is first of all due to elimination of acid molecules. This assumption was confirmed by the results of elemental analysis. At temperatures above 400 8C, the PBI chains gradually decompose under the action of high temperatures and strong acids. Complexes 10 are thermally stable up to 500 8C. It was found that treatment of PBI with a phosphoric acid solution (27 mass %) improved the thermal stability of the polymer.110 This was associated with the formation of benzimidazonium cations. Samms et al.104 studied the thermal stability of polymer complexes 10 and showed that these complexes are promising for use as polymer electrolytes in the hydrogen ± air and methanol fuel cells.To simulate the operating conditions in a high-temperature PEMFC, the polymer complexes 10 were coated with platinum black, doped with phosphoric acid (4.8 H3PO4 molecules per repeating unit of PBI) and heated in an atmosphere of nitrogen and 5% hydrogen or in air in a TGA analyser. The degradation products were identified by mass spectrometry. In all cases the mass loss below 400 8C was found to be due to the loss of water. In addition, it was found that polymer complexes 10 coated with platinum black are thermally stable up to 600 8C. Variation of the conductivity of polymer complexes 10 as a function of water vapour activity, temperature and acid doping level was studied.106 It was shown that the conductivity of heavily doped complexes (500 mol.%) is nearly twice as high as that of the film doped to 338 mol.% at the same temperature and humidity. For instance, the conductivity of PBI doped with 500 mol.% H3PO4 (5 H3PO4 molecules per repeating unit of PBI) is 3.561072 S cm71 at 190 8C and a water vapour activity of 0.1.Raising the temperature and water vapour activity causes an increase in the conductivity of the polymers irrespective of the doping level of PBI with phosphoric acid. In addition, it was found that crossover of methanol molecules through the polymer com- plexes 10 is by an order of magnitude smaller than in the case of N N HN HN n772 perfluorinated polymer electrolytes and that the mechanical strength of such complexes is three orders of magnitude higher compared to that of Nafion membranes. The proton conductivity of PBI polymer complexes prepared by the interaction of PBI with methanol solutions of strong acids was studied.7 The temperature dependences of the conductivities of anhydrous PBI ¡À strong acid polymer complexes are shown in Fig.17 a. log s (S cm71) 75 76 77 78 79 710 75 76 77 78 2.6 Figure 17. Temperature dependence of proton conductivity of anhydrous (a) and hydrous (b) PBI complexes withH3PO4 (1),H2SO4 (2), EtSO3H(3) and MeSO3H (4);7 (a): doping level, acid molecules per PBI unit: 2.0 (1), 1.8 (2), 2.0 (3), 1.9 (4); (b): water uptake (mass %): 13 (1,4), 19 (2) and 26 (3).All anhydrous polymer complexes of PBI with strong acids possess a proton conductivity of the order of 1076 ¡À 1079 S cm71 at 100 8C. The conductivity of polymer complexes 10 can be as high as 1075 S cm71 at 160 8C, whereas other PBI ¡À acid com- plexes showed a decrease in the conductivity at temperatures above 80 8C. These results point to high thermal stability of polymer complexes 10. To prepare hydrated systems, the films of PBI ¡À strong acid polymer complexes were placed in a desiccator with a relative humidity of 90% for 72 h. The water uptakes of the complexes were 13 mass%¡À26 mass %. The proton conductivity of the hydrous PBI ¡À strong acid polymer complexes was found to be nearly an order of magnitude higher than the conductivity of anhydrous polymer complexes (Fig.17 b). This difference can be explained by the improvement of charge carrier generation in the absorbed water. Changes in the proton conductivity at room temperature are especially remarkable. Figure 18 presents the temperature dependences of the conductivity of anhydrous com- plexes 10 with different acid contents. As can be seen, the conductivity of polymer complexes 10 increases with the concen- tration of H3PO4. The temperature dependence of the conductivity of polymer complexes 10 is quite different: in this temperature range the conductivity is low. This suggests that two H3PO4 molecules quantitatively react with the PBI units containing two imidazole groups.As a consequence, an excess of H3PO4 determines the necessary proton conductivity. A study of PBI/H3PO4 polymer complexes by FT-IR spectroscopy showed that the spectra exhib- ited three characteristic absorption maxima near 1090 cm71 (HPO2¡¦ 4 ), 1008 cm71 (POH) and 970 cm71 (H2PO¡¦4 ).111 ¡À 114 As a 1234 b 1234 103T71 /K 3.0 A L Rusanov, D Yu Likhatchev, KMu�� llen log s (S cm71) 74 54321 75 76 77 78 79 710 3.0 2.6 103T71 /K Figure 18. Temperature dependences of proton conductivity of an- hydrous PBI/H3PO4 complexes containing 1.4 (1), 2.0 (2), 2.7 (3), 2.3 (4) and 2.9 (5) H3PO4 molecules per PBI unit.7 4 4 the concentration of H3PO4 increases, the intensity of the absorp- tion maxima of HPO2¡¦ and H2PO¡¦4 increases.This ssts that proton conductivity can occur by the Grotthus mechanism115 involving an exchange of protons between H3PO4 and PO2¡¦ or H2PO¡¦4 . V. Fuel cell applications of proton-conducting membranes based on aromatic condensation polymers Two blend polymer electrolytes containing acid and basic func- tional groups (90 mass% PEEK and 10 mass% PBI or 95 mass% PES and 5 mass% PBI) were applied in H2/O2 fuel cells. The current-vs.-voltage curves of the membranes in the fuel cells were comparable with that of Nafion 112 membranes.109 Fuel cell tests of membranes based on sulfonated PES showed 7 a cell voltage of 550 mV at a current density of 700 mA cm72 (atmospheric pressure, humidified gases, 70 8C).No significant loss of membrane performance was observed after long-term operation (1000 h) under fuel cell conditions. The maximum power of fuel cells with S-PPBP membranes reaches 0.3 W cm72 at a current density of 800 mA cm72. The conductivity of the electrolyte membranes was 361073 S cm71; the membrane thickness and surface area were 0.01 cm and 3.15 cm2, respectively. The maximum power of fuel cells H2/O2 and CH3OH/O2 with membranes based on polymer complexes 10 (see Ref. 7) was as high as 0.25 W cm72 at a current density of 700 mA cm72. The electrical resistance of electrolyte membranes was 0.4 O, the thickness and surface area of the membranes were 0.01 cm and 1 cm2, respectively, and the doping level was 500 mol.%.The measured electrical resistance of the cell was equivalent to a conductivity of 0.025 S cm71. It was found that the electrical resistance of the fuel cell is independent of the water content in the gas (water produced at cathode is sufficient for maintaining the necessary conductivity of the electrolyte). This type of fuel cells was characterised by continuous operation at a current density of 200 mA cm72 over a period of 200 h (and for longer times) without reduction of the membrane performance. The power ofCH3OH/O2 fuel cells at 200 8C and atmospheric pressure reached 0.1 W cm72 at a current density of 250 ¡À 500 mA cm72. The conductivity of membranes operating under these conditions remained constant in the temperature range 30 ¡À 140 8C.* * * To summarise, the aforesaid shows that the aromatic condensa- tion polymers can be thought to be candidates for fuel cellProton-conducting electrolyte membranes based on aromatic condensation polymers applications: the structure of ACP-based polymer electrolytes can be modified with ease; these polymer electrolytes possess large water uptakes and high proton conductivities at high temper- atures and low humidities, as well as sufficient thermal and chemical stability. Further investigations to design proton-con- ducting materials exhibiting long-term thermal stability and mechanical strength, capable of operating at high temperatures without humidification and further improvement of fuel cells are required. References 1.J A Kerres J. Membr. Sci. 185 3 (2001) 2. H Riedinger, W Faul J. Membr. Sci. 36 5 (1988) 3. R Nolte, K Ledjeff,M Bauer, R Mulhaupt J. Membr. Sci. 83 211 (1993) 4. K B Prater J. Power Sources 51 129 (1994) 5. S Watkins, in Fuel Cell Systems (Eds L G Blumen,M N Mugerwa) (New York: Plenum, 1993) p. 493 6. J C Lasseques, in Proton Conductors: Solids Membranes and Gels (Ed. P Colombon) (Cambridge: Cambridge University Press, 1992) p. 311 7. M Rikukawa, K Sanui Prog. Polym. Sci. 25 1463 (2000) 8. O Savadogo J. New Matter. Electrochem. Syst. 1 66 (1998) 9. M Higuchi, N Minoura, T Kinoshita Chem. Lett. 2 227 (1994) 10. T A Zawodzinski, C Derouin, S Radzinski, R J Sherman, U T Smith, T E Springer, S Gottesfeld J.Electrochem. Soc. 140 1041 (1993) 11. M Yamabe, H Migake, in Organofluorine Chemistry. Principles and Commercial Applications (Eds R E Bomks, B E Smart, J C Tatlow) (New York: Plenum, 1994) p. 403 12. Yu E Kirsh, S A Smirnov, Yu M Popkov, S F Timashev Usp. Khim. 59 970 (1990) [Russ. Chem. Rev. 59 560 (1990)] 13. W G Grot Macromol. Symp. 82 161 (1994) 14. A Steck, in Proceedings of the 1st International Symposium on New Materials for Fuel Cell Systems, Membrane Materials in Fuel Cells (EdsO Savadogo, P R Roberge, T N Vezirogly) (Canada: Montreal, 1995) p. 74 15. A S Hay Adv. Polym. Sci. 4 496 (1967) 16. S Maiti, B Mandal Prog. Polym. Sci. 12 111 (1986) 17. H Lee, D Stoffey, K Neville, in New Linear Polymers (New York: McGraw-Hill, 1967) p.374 18. A H Frazer, in High Temperature Resistant Polymers (New York: Interscience, 1968) p. 338 19. V V Korshak Termostoikie Polimery (Thermally Stable Polymers) (Moscow: Khimiya, 1969) 20. P E Cassidy, in Thermally Stable Polymers (New York: Marcel Dekker, 1980) p. 163 21. J P Critchley,W WWright, in Heat-Resistant Polymers (New York: Plenum, 1983) p. 259 22. K-U BuÈ hler Spezialplaste (Berlin: Akademie-Verlag, 1978) 23. A L Rusanov, D S Tugushi, V V Korshak Uspekhi Khimii Poligeteroarilenov (Progress in the Chemistry of Heteroarylenes) (Tbilisi: Tbilisi State University, 1988) 24. V V Korshak, A L Rusanov Vysokomol. Soedin., Ser. A 26 3 (1984) a 25. A L Rusanov Vysokomol.Soedin., Ser. A 28 1571 (1986) a 26. A D Schluter J. Polym. Sci., Polym. Chem. 39 1533 (2001) 27. E D Power, G A Serad High Performance Polymers: Origin and Development (New York: Elsevier, 1986) 28. R R Luise, in Applications of High Temperature Polymers (Boca Raton, FL: CRC Press, 1996) p. 272 29. K D Kreuer J. Membr. Sci. 185 29 (2001) 30. K D Kreuer Chem. Mater. 8 610 (1996) 31. Z Qi, P G Pickup Chem. Commun. 1 15 (1998) 32. A D Child, J R Reynolds Macromolecules 27 1975 (1994) 33. T Kobayashi, T Rikukawa, K Sanui, N Ogata Solid State Ion. 106 219 (1998) 34. V A Sochilin, A V Pebalk, V I Semenov, I E Kardash Vysokomol. Soedin., Ser. A 33 1536 (1991) a 35. V A Sochilin, A V Pebalk, V I Semenov,M A Sevast'yanov, I E Kardash Vysokomol.Soedin., Ser. A 35 1480 (1993) a 36. A J Chalk, A S Hay J. Polym. Sci. A 7 5843 (1982) 37. I L Bredas, R R Chomce, R Silbey Phys. Rev. B 26 5843 (1982) 773 38. H Kobayashi, H Tomita, H Moriyama, A Kobayashi, T Watanabe J. Am. Chem. Soc. 116 3153 (1994) 39. F Wong, J Roovers Macromolecules 26 5295 (1993) 40. C Bailly, D J Williams, F F Karasz,W J McNight Polymer 28 1009 (1987) 41. X Jin,M T Bishop, T S Ellis, F E Karazz Br. Polym. J. 17 4 (1985) 42. MI Litter, C S Marvel J. Polym. Sci., Polym. Chem. Ed. 23 2205 (1985) 43. J Devaux, D Delimoy, D Dauost, R Legras, J P Mercier, C Strazielle, E Nield Polymer 26 322 (1985) 44. T Ogawa,CS Marwel J. Polym. Sci., Polym. Chem. Ed. 23 1231 (1985) 45. US P. 4 625 000; Chem Abstr.106 85 292 (1987) 46. US P. 4 413 106; Chem Abstr. 100 35 025 (1983) 47. A Noshay, L M Robeson J. Appl. Polym. Sci. 20 1885 (1976) 48. B C Johnson, I Yilgor, C Tran, MIqbal, J P Wightman, D R Lloyd, J E McGrath J. Polym. Sci., Polym. Chem. Ed. 22 721 (1984) 49. C Mottet, A Revillon, P Le Perchec,M E Lauro, A Guyot Polym. Bull. 8 511 (1982) 50. C Arnold Jr, R A Assinic J. Membr. Sci. 38 71 (1988) 51. P Zschocke, D Guellmalz J. Membr. Sci. 22 325 (1985) 52. P Genova-Dimitrova, B Baradie, D Foscallo, C Poisington, J Y Somchez J. Membr. Sci. 185 59 (1976) 53. Z Qi, M C Lefebre, P G Pickup J. Electrochem. 459 9 (1998) 54. Russ. P. 1 819 418; Byull. Izobret. (14) 61 (1992) 55. N M Belomoina, A L Rusanov, N A Yanul', Yu E Kirsh Vysokomol.Soedin., Ser.B 38 355 (1996) a 56. Yu E Kirsh, N A Yanul', N M Belomoina, A L Rusanov Elektrokhimiya 32 169 (1996) b 57. US P. 4 634 530; Chem. Abstr. 106 157 701 (1987) 58. X L Wei, Y Z Wang, S M Long, C Bobeczko, A J Epstein J. Am. Chem. Soc. 118 2545 (1996) 59. E E Gilbert, in Sulfonation and Related Reactions (New York: Wiley-Interscience, 1965) p. 78 60. E E Gilbert Sul'firovanie Organicheskikh Soedinenii (Sulfonation and Related Reactions) (Translated into Russian; Moscow: Khimiya, 1969) 61. X Jin,M T Bishop, T S Ellis, F E Karasz Br. Polym. J. 17 4 (1985) 62. J Lee, C S Marvel J. Polym. Sci., Polym. Chem. Ed. 22 295 (1984) 63. B C Johnson, I Ylgor, M Iqbal, J P Wrightman, D R Lloyd, J E McGrath J. Polym. Sci., Polym. Chem. Ed.22 72 (1984) 64. C Bially,D Williams, F E Karasz,W J MacKnight Polymer 28 1009 (1987) 65. W A Thaler J. Polym. Sci. 20 875 (1982) 66. W A Thaler Macromolecules 16 623 (1983) 67. M T Bishop, F E Karasz, P S Russo, K H Langley Macromolecules 18 86 (1985) 68. DDevaux,DDelimoy,DDaoust, R Legras, J P Mercier, C Strazielle, E Neild Polymer 26 1994 (1985) 69. T Ogawa, C S Marvel J. Polym. Sci., Polym. Chem. Ed. 23 1231 (1985) 70. K Miyatake, E Shouji, K Yamamoto, E Tsuchida Macromolecules 30 2941 (1997) 71. K Miyatake, H Iyotani, K Yamamoto, E Tsuchida Macromolecules 29 6969 (1996) 72. Jpn. P. 11-116 679; Chem. Abstr. 130 325 763 (1999) 73. Jpn. P. 11-067 224; Chem. Abstr. 130 239 961 (1999) 74. N Venkatasubramanion, D Deon, F E Arnold Polym.Prepr. 37 (1) 354 (1996) 75. M Ueda, H Toyota, T Ochi, J Sugiyama, K Yonetake, T Masuko, T Teremoto J. Polym. Sci., Polym. Chem. Ed. 31 853 (1993) 76. J E McGrath, R Formato, R Covar,W Harrison, J B Mechant Polym. Prepr. 41 (1) (2000) 77. J Mecham, H K Shobha, F Wong, W Harrison, J E McGrath Polym. Prepr. 41 (2) 1388 (2000) 78. F Wong, J Mecham,W Harrison, J E McGrath Polym. Prepr. 41 (2) 1401 (2000) 79. N K Shobha,M Somkarapandion, T E Glass, J E McGrath Polym. Prepr. 41 (2) 1298 (2001) 80. S Faure, R Mercier,M Pineri, B Sillion, in The 4th European Technical Symposium on Polyimides and Other High Performance Polymers (France: Montpellier, 1996) p. 414 81. C Genies, R Mercier, B Sillion, R Petioud, N Cornet, G Gebel, M Pineri Polymer 42 5097 (2001) 82. N Gunduz, J E McGrath Polym. Prepr. 41 (2) 1565 (2000)A L Rusanov, D Yu Likhatchev, KMuÈ llen 774 83. C Genies, R Mercier, B Sillion, N Cornet, G Gebel,M Pineri Polymer 42 359 (2001) 84. Y Zhang,M Litt, R F Savinell, J S Wainright, J Vendramun Polym. Prepr. 41 (2) 1561 (2000) 85. G I Timofeeva, I I Ponomarev, A R Khokhlov, R Mercier, B Sillion Macromol. Symp. 106 345 (1996) 86. I I Ponomarev, O G Nikol'skii, Yu A Volkova, A V Zakharov Vysokomol. Soedin., Ser. A 36 1185 (1994) a 87. A L Rusanov Usp. Khim. 61 815 (1992) [Russ. Chem. Rev. 61 449 (1992)] 88. A L Rusanov Adv. Polym. Sci. 111 116 (1994) 89. A L Rusanov, L B Elshina, E G Bulycheva, K Myullen Vysokomol. Soedin., Ser. A 41 2 (1999) a 90. V A Solomin, E N Lakh, B A Zhubanov Vysokomol. Soedinen., Ser. A 34 274 (1992) a 91. M B Gieselman, J R Reynolds Macromolecules 23 3188 (1990) 92. M Tarayanagi Pure Appl. Chem. 55 819 (1983) 93. US P. 4 814 399; Chem. Abstr. 111 1 956 601 (1989) 94. T D Domg, F E Arnold Polym. Prepr. 33 912 (1992) 95. M B Gieselman, J R Reynolds Macromolecules 25 4832 (1992) 96. M B Gieselman, J R Reynolds Macromolecules 26 5633 (1993) 97. X Glipa,M E Haddad,D J Jones, J Roziere Solid State Ion. 97 323 (1997) 98. K Tsuruhara, K Hara,M Kawahara, M Rikukawa, K Sanui, N Ogata Electrochim. Acta 45 1223 (2000) 99. M Kawahara, M Rikukawa, K Sanui Polym. Adv. Technol. 11 544 (2000) 100. M Kawahara, M Rikukawa, K Sanui, N Ogata, in Proceedings of the 6th International Symposium on Polymer Electrolytes (Extended Abstracts), Hayawa, 1998 p. 98 101. H Yoshida, T Hatakeyama, H Hatakeyama Polymer 31 693 (1990) 102. M Folk Can. J. Chem. 58 1495 (1980) 103. J-T Wang, S Wasmus, R F Savinell J. Electrochem. Soc. 143 1233 (1996) 104. S R Samms, S Wasmus,R F Savinell J. Electrochem. Soc. 143 1225 (1996) 105. R F Savinell, E Yeager, D Tryk, U Landau, J S Wainright, D Weng, K Lux, M Litt, C Rogers J. Electrochem. Soc. 141 L46 (1994) 106. J S Wainright, J-T Wang, D Weng, R F Savinell,M Litt J. Electrochem. Soc. 142 L121 (1995) 107. M Kawahara, J Morita,M Rikukawa, K Sanui, N Ogata Electrochim. Acta 45 1395 (2000) 108. M Kawahara,M Rikukawa,K Sanui,N Ogata Solid State Ion. 136 1193 (2000) 109. J Kerres, A Ullrich, F Meier, T Haring Solid State Ion. 125 243 (1999) 110. E D Powers,G A Serad, in High Performance Polymers: Origin and Development (New York: Elsevier, 1986) p. 355 111. J R Stevens,W Wieczorek, D Raducha, K R Jeffrey Solid State Ion. 97 347 (1997) 112. W Wieczorek, J R Stevens Polymer 38 2057 (1997) 113. R Tanaka, H Yamamoto, S Kawamura, T Iwase Electrochim. Acta 40 2421 (1995) 114. M F Daniel, B Destbat, F Cruege, O Trinquet, J C Lassegues Solid State Ion. 28 ± 30 637 (1988) 115. N Agmon Chem. Phys. Lett. 244 456 (1995) a�Polym. Sci. (Engl. Transl.) b�Russ. J. Electrochem. (Engl. Tran

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC

摘要:

Russian Chemical Reviews 71 (9) 775 ± 787 (2002) Chemical modification of electrolytes for lithium batteries V N Afanas'ev, A G Grechin Contents I. Introduction II. Liquid-phase electrolytes III. Polymeric electrolytes IV. Electrolytic solvosystems for lithium batteries and mechanisms of charge transfer in solutions Abstract. electro- chemically modifying to approaches Modern Modern approaches to modifying chemically electro- lytes for lithium batteries are analysed with the aim of optimising lytes for lithium batteries are analysed with the aim of optimising the solid and liquid-phase in processes charge-transfer the charge-transfer processes in liquid-phase and solid (polymeric) (polymeric) media. of properties transport of regularities main The media. The main regularities of transport properties of lithium lithium electrolyte ions (encapsulated) complex containing solutions electrolyte solutions containing complex (encapsulated) ions in in aprotic for prospects The discussed. are polymers and solvents aprotic solvents and polymers are discussed.The prospects for the the development chain the with solvosystems electrolytic of development of electrolytic solvosystems with the chain (iono- (iono- tropic) ions lithium to respect with conduction of mechanism tropic) mechanism of conduction with respect to lithium ions are are outlined. The bibliography includes 126 references outlined. The bibliography includes 126 references. I. Introduction Direct energy conversion from chemical to electrical forms is a central problem of modern science and technology.At present, active research is carried out on the development of primary and secondary lithium cells with characteristics (high energy density, low self-discharge rate, applicability at low temperatures, high reliability, etc.) that should surpass those of traditional electro- chemical systems 1±6 and displace them in future. The research on lithium chemical power sources (CPS) was, first of all, stimulated by the progress in such industrial fields as electronics, new communication systems and transport (electrical vehicles). Lith- ium cells and rechargeable batteries are increasingly applied as autonomous power sources in various domestic, medical and computer devices. Special-purpose lithium CPS are used in night-vision devices, guidance systems, space equipment, and new weapon systems.At present, the yearly production of lithium CPS for special equipment and military branches is *30 million cells, which is 3 ± 4 times higher compared with other industrial branches.7 Elaboration of highly efficient lithium CPS is largely aimed at the development of electrolytic systems with high conductivities (1073± 1072 S cm71) in a wide temperature range (from 750 to +70 8C) and chemical and electrochemical stabilities with respect to lithium and cathodic materials. Such systems should also enable sufficiently fast and reversible electrode processes. Insuffi- cient conductivity of liquid-phase and, particularly, solid poly- V N Afanas'ev, A G Grechin Institute of Solution Chemistry, Russian Academy of Sciences, ul.Akademicheskaya 1, 153045 Ivanovo, Russian Federation. Fax (7-093) 237 85 09. Tel. (7-093) 237 85 19. E-mail: vna@isc-ras.ru (V N Afanas'ev), agg@ihnr.polytech.ivanovo.su (A G Grechin) Received 27 May 2002 Uspekhi Khimii 71 (9) 878 ± 892 (2002); translated by T Ya Safonova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n09ABEH000737 775 776 781 783 meric lithium electrolytes is a key factor impeding the development of CPS with lithium electrodes.8 ±11 The importance of solving this problem is determined by the fact that the internal resistance of a power source directly governs its main performance characteristics such as the power density and the discharge current.Most approaches to choosing optimum electrolytic composi- tions for lithium CPS were based on theoretical concepts of the ion-migration mechanism of charge transfer.12 ± 14 Hence, the aim of the latest studies was to create such conditions that would both enhance the migration rates of electrolyte ions as the main charge carriers in solution and in polymers and increase their concen- trations. For this purpose, mixed low-viscosity solvents with sufficiently low dielectric constants were used as the electrolytic systems, macrocyclic ligands which can encapsulate lithium cati- ons and different anions were added, lithium salts with low lattice energies which involve large anions with delocalised charges were synthesised, etc.It is known that a chain (prototropic) mechanism of conduction is brought about in aqueous solutions of acids and bases used as the electrolytes for traditional CPS, which is responsible for the conductivities of these solutions being 1 ± 2 orders of magnitude higher as compared with solutions with the migration mechanism of charge transfer. It is also known that the ionotropic mechanism of charge transfer can involve not only protons in water and protic solvents but also some other ions in aprotic media. In this connection, the latest studies have consid- ered a possibility of developing highly conductive nonaqueous electrolytes with a conduction mechanism different from the ion- migration one, as well as designing CPS based on such electrolytes with improved performance characteristics.This survey analyses and generalises the results of studies published in the past 10 ± 15 years and devoted to modifying chemically the conventional electrolytes applied in lithium CPS with the aim of optimising the charge transfer processes in liquid- phase and solid (polymeric) media. The analysis was carried out from the standpoint of dividing the electrolytes according to their phase compositions, viz., liquid-phase and solid (polymeric) electrolytes. The latter type traditionally involves biphasic electro- lytes, i.e., gels, composites, etc. The survey also includes new systems which do not fit the traditional classification (e.g., electro- lytes with polymeric anions and ionic liquids).That is why it seemed reasonable to analyse the results of studies of different polymeric electrolytes according to the examples of concrete chemical modifications that were most interesting from our view- point, rather than to consider in detail the classes of electrolytes.776 II. Liquid-phase electrolytes The main factors that affect the charge migration rate in electro- lytes are the solvent viscosity and the processes of solvation and association. One of the first approaches to optimising electrolytic systems in lithium CPS consisted in the use of complex lithium compounds with low lattice energies, which involve weakly solvated and easily polarisable anions. This provided solutions with enhanced conductivities and higher concentrations of elec- trolyte ions.Within this approach, lithium electrolytes with coordinatively saturated fluorine-containing anions AsF¡6 , PF¡6 , BF¡4 and also with ClO¡4 , Ph4 B7, all of which exhibited suffi- ciently high electrochemical stability, were extensively used.2, 8 ± 11, 15 ± 17 Moreover, the use of fluorine-containing salts allowed one to reduce the corrosion of cathodes with aluminium current leads at high positive potentials. However, the perform- ance of power sources with the electrolytes mentioned entailed quite a number of problems. For example, lithium perchlorate is thermally unstable, and its solutions are explosive, especially, in ether-like solvents; lithium hexafluorophosphate can decompose in solution to give poorly soluble lithium fluoride and a Lewis acid PF5, which initiates polymerisation of cyclic ethers; hexafluoro- arsenate ions are reduced electrochemically to give AsF5 and As0 which are environmentally hazardous; lithium tetrafluoroborate solutions do not provide a sufficient conductivity level and, moreover, BF¡4 ions initiate polymerisation of cyclic ethers.Therefore, the search for new lithium electrolytes which would exhibit better compatibility with solvents and electrode materials is still in progress. 3 , B10Cl¡10, B12Cl¡12 4 turned out to be more stable but their solutions Lithium salts with anions such as CF3SO¡ and BMe¡ possessed low conductivities. Recently, new lithium electrolytes with large anions and minor polarisability have been proposed: Li[CF3SO2]2N (1) and the corresponding cyclic imides Li[N(SO2)(CF2)n(SO2)] (n=1± 3) (2), lithium methanide Li(CF3SO2]3C (3), the family of lithium chelatoborates of the general formula Li[BR2] (4 ± 6) (where bidentate ligands R are aliphatic or aromatic diols), and the corresponding lithium phosphates Li[PR3] which are highly stable chemically and electrochemically, well soluble (>1 mol litre71) and weakly associated.9, 18 ± 25 The charge distribution in anions strongly affects the electrolyte association.With addition of electron- acceptor substituents such as F, CF3, COOR, SO2OR and nitro- gen atoms, which can delocalise the negative charge in anions, the association constants of lithium electrolytes decreased by several orders of magnitude. The effects mentioned favour the increase in the conductivity and extend the electrochemical window of solvents.Thus the specific conductivity of 1Msolutions of lithium F3CSO2 SO2CF3 N7 Li+ 1 F3CSO2 Li+SO2CF3 7SO2CF3 3O OB7 O O Li+ 5 (CF2)n SO2 O2S N 7 Li+ 2 F F F F O O B7 O F F O Li+ F F 4 F O SO2 O7 B O O Li+ F SO2 6 V N Afanas'ev, A G Grechin chelatoborates Li[B(C6HxF47xO2)2] (1<x<4) in 1,2-dime- thoxyethane (DME) increased by a factor of 4.4 at 25 8C and by a factor of 2.4 at 745 8C with substitution of fluorine atoms for four hydrogen atoms. A linear correlation between the limiting potentials of anodic oxidation of lithium borates and the energies of higher occupied molecular orbitals of anions, which was estimated using semiempirical quantum-mechanical methods AM1 and MNDO,9, 19 indicates that the electrochemical stability of these electrolytes depends on the degree of delocalisation of the negative charge in anions.Lithium CPS most often employ cyclic and linear ethers and esters, viz., DME, propylene carbonate (PC), g-butyrolactone (g-BL), tetrahydrofuran (THF) and inorganic sulfur compounds (SO2, SOCl2, SO2Cl2) as the solvents.2, 26 ± 28 Solutions of lithium salts in such solvents demonstrated sufficiently high conductivities but often exhibited an enhanced activity with respect to both the lithium anode and cathodes.To enhance the stability of electrode materials, it was proposed to add dialkyl carbonates to solutions. In this case, due to a protective film formed on the anode and consisting mainly of solvent oxidation products, the cycling stability substantially increased. At present, the electrolytes for lithium power sources involve, as a rule, binary and ternary compositions based on PC or ethylene carbonate with addition of a certain low-viscosity solvent (DME, THF, etc.). At low temperatures, mixed solvents with low viscosities and sufficiently high dielectric constants proved to be most efficient.9, 10 More- over, the possibility of using certain strongly associated and sufficiently viscous liquids, viz., sulfones,29 ± 31 sulfamides R1R2NSO2NR3R4 and glymes (dimethyl ethers of polyethylene glycol) CH3O(CH2CH2O)nCH3 (see Refs 32 ± 34), which allow preparing solutions of lithium salts with conductivities of an order of magnitude of (0.8 ± 2.5)61073 S cm71, was discussed.1. Electrolytes with molecularly encapsulated lithium ions The maximum conductivity and prevention of concentration polarisation during the discharge, were achieved if concentrated (1.5 ± 3 M) electrolyte solutions were used in lithium CPS. How- ever under these conditions, many lithium salts in solvents of low and medium dielectric constants dissociated incompletely to form ion pairs and other associates which cannot conduct current.35 Insufficient solubility of lithium electrolytes also limited their practical application.The solubility and conductivity of lithium electrolytes could be enhanced by introducing macrocyclic com- pounds which contain oxygen or nitrogen atoms possessing appreciable affinity towards lithium ions into solutions. Several groups of mono- and bicyclic oxo- and aza(oxo)-ligands which form stable lithium complexes with the 1 : 1 composition are known. These comprise crown ethers: 12-crown-4 (7), 15-crown-5 (8), 18-crown-6 (9), 1,4,7-trioxa-10-azacyclododecane (10), N-benzyl-1,4,7-trioxa-10-azacyclododecane (11), etc.; crypt- ands: cryptand 211 (12), cryptand 221 (13), cryptand 222 (14); cryptand 222D (15), cryptand 221D (16), etc.; calixarenes: 25,27;26,28-bis[1,4,7,10,13,16-hexaoxa(hexadecamethylene)]calix- [4]arene (17), 4,10,16,22-tetra-tert-butyl-25,26,27,28-tetrakis- (ethoxycarbonylmethoxy)calix[4]arene (18), etc.36 ± 43 Tricyclic ligands with cylindrical cavities can also form binuclear complexes with the composition 2 : 1 at high Li+: L ratios.42 Among all macrocyclic compounds known, cryptands form the most stable inclusion complexes (cryptates) with alkali metal ions.Such ligands have a spheroidal cavities that can encapsulate cations of suitable sizes, virtually screening the latter from interactions with solvent molecules and counterions.36, 37, 43 ± 63 Due to their unique complexing properties, cryptands and crown- ethers favour the dissolution of metal salts and weaken their association in low-polarity media. The solubility increases due to a decrease in the free energy of ions as a result of their encapsulation in the ligand's internal cavity MáL Má solv .solv á LsolvChemical modification of electrolytes for lithium batteries O O O O O O O O O 7 8 O O HN O O O O O O O 10 9 O N O N N O O O O O 12 11 O O O O O N N N O O N O O O O 13 14 (CH2)9CH3 (CH2)9CH3 O O O O O N N O O N N O O O O 15 16 O O O O But O O 4 O O O O O O O O O17 18 This effect and the solvation of hydrophobic groups of the ligand by organic solvent molecules compete with the increase in the free energy of ions as their radii increase upon complexing. By the example of cryptand 14, it was shown that complexing of salts with macrocyclic ligands results in a decrease in the solubility of salts well soluble in this solvent and its increase for poorly soluble salts. For a given cryptate salt, the solubility increased with the transition to less polar media in which this salt is more strongly associated.Table 1. Standard enthalpies of formation for lithium electrolytes in acetonitrile (AN) and propylene carbonate (PC) at 298.15 K.45 ¡À 49 Electrolyte LiBF4 [Li(7)]BF4 [Li(8)]BF4 [Li(10)]BF4 [Li(11)]BF4 [Li(14)]BF4 [Li(18)]BF4 LiClO4 [Li(7)]ClO4 [Li(11)]ClO4 [Li(14)]ClO4 [Li(18)]ClO4 LiCF3SO3 [Li(7)]CF3SO3 [Li(8)]CF3SO3 [Li(10)]CF3SO3 [Li(11)]CF3SO3 [Li(14)]CF3SO3 [Li(18)]CF3SO3 LiAsF6 [Li(7)]AsF6 [Li(8)]AsF6 [Li(10)]AsF6 [Li(11)]AsF6 [Li(18)]AsF6 The strategy of choosing complex electrolytes for lithium CPS, which has been put forward in Refs 45 ¡À 49, is based on studying the processes of complex formation and solvation in lithium salt solutions.The experimentally determined enthalpies of dissolu- tion DH for a number of lithium electrolytes in acetonitrile and PC (Table 1) showed that, as a rule, a lithium ion in the free state is much more strongly solvated (DH<0) than that incorporated into a complex with macrocyclic ligands (DH>0). A decrease in the degree of solvation due to complexing, which leads to an increase in the conductivity, was observed in less polar media. A number of lithium electrolytes with complex cations involving crown ethers, cryptands and calix[4]arenes as the ligands were isolated as solids.Studies of solvation of simple (M+X7) and complex ([ML]+X7) electrolytes as well as ligands (L) allowed the estimating of the changes in the enthalpy for solid-state complex- ing processes MX(s)+L(s) These values can be considered as the characteristics of complex formation in media with low dielectric constants.45 ¡À 49 As was shown, the association of a complex cation with an anion in cryptand-containing solid electrolytes is weaker as compared with electrolytes containing crown ethers and calixarenes. For lithium salts (LiX), the coordination process becomes less exo- thermic with a decrease in the polarisability in the following sequence of anions: I7>ClO¡¦4 >Br7>CF3SO¡¦3 >AsF¡¦6 > BF¡¦4 .In propylene carbonate at 298.15 K, the complexing of Li+ with crown ethers resulted in a substantial increase in the ionic mobility (Table 2). Similar results were obtained for lithium complexes with 9 in PC and 2-cyanopyridine and complexes with 9 and 14 in a PC¡À dichloromethane mixed solvent.50, 51 Studies of the molar conductivity (L) of a 1 M LiBF4 solution in g-BL as a function of the [L] : [LiBF4] ratio [L designates crown ethers 8 and 9 and dibenzo-18-crown-6 (19)] have shown 52 that L increases with an increase in the crown ether content and reaches its maximum at the equimolar ratio. An electrolyte based on crown ether 8 which forms the most stable complex with a lithium DH /kJ mol71 AN 714.57 4.74 4.18 0.85 11.33 6.7 721.0 743.26 1.27 10.23 20.9 79.17 715.99 15.69 9.17 14.66 18.96 18.4 714.0 718.45 7.17 10.62 3.72 8.75 722.5 MLX(s).777 PC 715.55 9.69 7.29 3.73 12.35 DDD6.51 9.62 DD 712.50 D13.34 18.74 22.50 DD 715.14 14.92 13.19 4.84 13.91 D778 Table 2. Limiting mobility (S cm2 mol71) of Li+ and [LiL]+ cations at 298.15 K. Ref. Cation Ref. Cation L L LiLá LiLá Solvent�nitromethane Solvent�PC 58 58 41.4 29.3 Li+ [Li(7)]+ [Li(8)]+ [Li(9)]+ [Li(10)]+ [Li(11)]+ 46 46 46 50 46 46 Li+ [Li(8)]+ Solvent�AN69.5 68.68 65.10 63.20 60.0 45.6 45.1 64 61 61 64 64 64 64 Li+ [Li(7)]+ [Li(8)]+ [Li(13)]+ [Li(14)]+ [Li(15)]+ [Li(16)]+ 7.86 10.02 10.50 8.41 10.25 11.99 Solvent�methanol 39.61 37.53 38.20 38.60 Li+ [Li(7)]+ [Li(8)]+ [Li(14)]+ 62 61 61 62 ion exhibited the maximum conductivity (1.161072 S cm71).The use of 8 enhanced the battery power. A possibility of using substances capable of forming charge-transfer complexes (CTC) as the functional additives to electrolytes was also considered.52 It was found that for solutions containing CTCof metal lithium with hexamethylphosphoric triamide (HMPA) Li+[(CH3)2N]3PO7 in g-BL (lithium was added in excess, the ratio [HMPA] : [g-BL]= 1 : 2), the conductivity increased by three orders of magnitude as compared with the initial solution without CTC and reached 4.261072 S cm71.The potential drop vs. current density dependence has shown that the transport properties of a complex electrolyte are better compared with the ordinary one. It was assumed that in the presence of CTC in solution the charge transfer can be realised by the chain mechanism, where the discharge of an ion is not preceded by its desolvation. O O O O O HN NH O O O O O 19 20 Most of the studies observed a correlation between the stability of complex lithium ions and the increase in the conduc- tivity of the solution. Thus it was found 53 that the addition of crown ethers 7 and 9 to LiAsF6 solutions inDMEand 9 to LiClO4 solutions in DME has practically no effect on the conductivity in the range of salt concentration from 1074 to 561072 mol litre71.The result obtained can be explained by the weak interaction between lithium ions and ligands due to strong solvation of ions by the solvent molecules with the chelate structure. The addition of crown ethers to NaClO4 solutions in DME resulted in for- mation of more stable complexes of sodium ions with ligands, thus decreasing the electrolyte association and increasing its conduc- tivity. For more strongly associated electrolytes, the addition of macrocyclic ligands led to a greater increase in conductivity as compared with less associated electrolytes. For instance, the molar conductivity of a LiClO4 solution in a cyclic ether, namely, 1,3-dioxolane, in the absence of macrocyclic compounds was lower as compared with its solutions in DME due to the stronger association of the salt in the former case.At the same time, the conductivities of 1,3-dioxolane solutions of LiClO4 containing equimolar additions of 7 ± 9 with cLiClO4>1073 mol litre71 were substantially higher compared with DME solutions. The greatest increase in conductivity was observed upon addition of crown ether 8 (Fig. 1). It was concluded 54 that the extent of the complex- logL (S cm2 mol71) 1.0 0.5 0.0 70.5 73 Figure 1. Dependence of the molar conductivity of LiClO4 and LiClO4 + crown-ether (1 : 1) solutions in 1,3-dioxolane on the salt concentration at 298 K:54 (1) LiClO4, (2) LiClO4+7, (3) LiClO4+9, (4) LiClO4+8. ing-induced increase in conductivity should be directly propor- tional to the increase in the solution dielectric constant (e) observed in low-polar media due to the higher dipole moments of ion pairs comprising complex cations as compared with ordinary ion pairs.Conductometric and spectroscopic studies of dissociation of salts [LiL]+[ZPh3]7 and [LiL]+[A]7 (Z=C, Ge and Sn, L is cryptand 12 and A is an anion of a CH-acid) in solvents of low polarity have shown that the introduction of a cryptand to solutions of organolithium compounds resulted in the formation of `cryptand-separated' ion pairs with light absorption maxima independent of the solvent nature.55, 56 In nonpolar solvents, the ion pairs comprising anions with strongly delocalised charges (Ph3C7) were less strongly associated.In more polar media, the inverse order of anion stability (with respect to ion pairs) due to the stronger ion ± dipole interaction of localised charges with solvent molecules was found. Thus, in polar solvents, the electro- lytes with charges localised on anions are associated to a lesser degree. For lithium salts containing anions of CH-acids with strongly delocalised charges, the dissociation constants (Kd) remained virtually unchanged with the formation of cryptates. The degree of association changed only for salts containing anions with localised charges. On going from THF (e=7.6, Kd=1076 mol litre71) to cyclohexylamine (e=4.7), and N-methylmorpho- line, the dissociation constants of cryptate-containing ion pairs decreased by 3 ± 4 orders of magnitude, whereas the transition to benzene (e=2.3) resulted in their decrease by 9 orders of magnitude.Moreover, the constants depended linearly on the reciprocal dielectric constant of the medium. Aseries of studies 57 ± 64 dealt with conductivity of lithium salts in media with similar dielectric constants, viz., acetonitrile (AN) and dimethylformamide (DMF), methanol and nitromethane (NM), with the aim of elucidating how the molecular structure of the solvent and the presence of macrocyclic ligands affect the transport properties of solvents. It was shown that the limiting V N Afanas'ev, A G Grechin 4 2 3 1 71 logc (mol litre71) 72Chemical modification of electrolytes for lithium batteries molar conductivity (L0) of a complex electrolyte [LiL]Pi [L is diaza-18-crown-6 (20), Pi is picrate ion] inANis smaller thanL0 of a simple LiPi salt.However, with an increase in the LiPi concen- tration (at the constant L concentration of 10.261073 mol litre71), the molar conductivity of an electrolyte containing 20 became higher than without this additive. At a constant salt concentration, the molar conductivity increased with an increase in the L concentration. At the same time, the molar conductivity of KPi in AN with addition of 20 (c20=5.261073 mol litre71) was lower and decreased with an increase in the ligand content.57 The results obtained were analysed with account of the following equilibria: K1 (C+S)+(A7S)+L C++A7+S+LK5 (C+LA7)S where C is the solute,Ais a counterion, S is a solvent molecule and L is a ligand.Assuming that K2>K5 and taking into account the strong association of LiPi in acetonitrile (K5=103 kg mol71), the increase in conductivity can be explained by dissociation of nonconducting ion pairs, which is caused by complex formation (K2). For KPi, it was concluded that equilibria 3, 4 and 5 weakly affect equilibria 1 and 2 due to insignificant association of the salt (K5=77 kg mol71). It was also found that the addition of L to solutions of KPi and LiPi in MeOH did not affect the value of L due to the specific solvation of the ligand by solvent with the formation ds. However, upon addition of an alkyl- substituted diazacrown ether RL (R=Alk), which can form complexes with cations, the conductivity of solutions decreased.The addition of crown ether 8 to LiPi and LiClO4 solution in MeOH and DMF, in which these electrolytes are weakly associ- ated, did not affect the value of L. For strongly associated LiPi and LiClO4 in AN and NM (Fig. 2), the introduction of macro- L /S cm2 mol71 140 120 100 2 80 60 40 200 20 10 Figure 2. Dependence of the molar conductivity of lithium picrate LiPi solutions in (1, 2) acetonitrile and (3, 4) nitromethane on the salt concentration at 298.15 K.58, 59 Concentration of additive 8: (1) 46.4361074 (3) 89.0561074 mol litre71, (2, 4) in the absence of 8. K2 (C+SA7) K3 K4 (C+L)S +(A7S) 1 3 4 30 104c /mol litre71 cyclic ligand 8 enhanced the conductivity, decreased the associa- tion constants, and made the decay in L which occurs with an increase in electrolyte concentration less pronounced.58, 59 In an AN+NM mixed solvent, the association constants were practi- cally independent of its composition (as might be expected for isodielectric systems); in the absence of 8, the values of constants substantially differed from one another and depended nonlinearly on the mixed solvent composition.60 Studies of the molar conductivity of solutions of lithium, sodium and potassium perchlorates in MeOH and AN containing crown ethers 7, 8 and cryptands 13, 14 have shown that the association constants, mobilities and Stokes radii of [ML]+ complex ions are practically independent of the nature of the metal ion involved.As a rule, the association constants for complex ions were lower compared with `free' ions; however, L0([ML]+X7)4L0(MX) (Table 3).61 ± 64 It was assumed that the formation of complexes levels off the differences in the degrees of solvation, charge densities and sizes of complex ions, so that their mobilities depend solely on the size of the ligand attached. A deviation from this rule observed for large cryptate ions was attributed to the changes in their degrees of solvation, because for `free' alkali ions L0 increases as their solvation weakens Table 3. Molar conductivity values at infinite dilution and association constants for lithium electrolytes at 298.15 K. Electrolyte L0 /S cm2 mol71 Solvent�AN 173.220.02 172.300.01 166.20.2 167.010.01 163.660.01 149.41 148.89 127.67 LiClO4 [Li(7)]ClO4 [Li(8)]ClO4 [Li(13)]ClO4 [Li(14)]ClO4 [Li(15)]ClO4 [Li(16)]ClO4 [Li(17)]ClO4 Solvent�nitromethane LiClO4 [Li(8)]ClO4 LiPi [Li(8)]Pi 110.60.2 97.030.2 86.4 73.310.06 Solvent�DMF LiClO4 [Li(8)]ClO4 LiPi [Li(8)]Pi 73.340.04 70.320.2 58.330.05 57.90.2 Solvent�methanol 110.520.03 108.500.03 109.170.02 109.550.05 LiClO4 [Li(7)]ClO4 [Li(8)]ClO4 [Li(14)]ClO4 [Li(17)]ClO4 97.18 Solvent�PC 26.75 28.91 27.30 23.52 28.57 31.06 30.83 32.70 24.27 27.14 26.73 LiClO4 [Li(7)]ClO4 [Li(9)]ClO4 [Li(17)]ClO4 LiBF4 [Li(8)]BF4 [Li(10)]BF4 [Li(11)]BF4 LiCF3SO3 [Li(8)]CF3SO3 [Li(10)]CF3SO3 779 Ref.Ka /litre mol71 64 64 64 64 52 230.3 59 7.970.01 61 151 61 8.810.04 5.20.1 15.900.06 13.770.07 16.98 2052 60 13.10.3 1.66106 60 60 60 84.70.5 59 91 59 101 59 12 161 59 251 5961 61 7.9670.001 2.40.004 152 6362 13.97 1.2 ��0.48 0.29 ���1.91 �� 50 46 50 62 46 46 46 46 46 46 46780 (Li+<Na+<K+) and Stokes radii increase. Similar results were reported in Ref. 65. The addition of dicyclohexyl-18- crown-6 to ethanol solutions of NaCl, KCl and RbCl (c=1074± 1073 mol litre71) resulted in complete dissociation of the electrolytes. At the same time, the values of L0 decreased.For the system KI+crown ethers in AN, theL0 values were lower than in solutions without ligands, decreasing with a decrease in the r1/r2 ratio, where r1 is the ionic radius and r2 is the cavity radius. It is of note that electrolytes based on [K(12-crown-4)]I and [K(15-crown-5)]I (r1/r2>1) in AN exhibited association con- stants of 284 and 235 litre mol71, respectively, whereas the original electrolyte KI was completely dissociated. The explan- ation of this effect invokes the fact that the effective radii of complex ions are smaller than those of solvated (`free') ions. It is known that an increase in association constants due to complex formation is most pronounced for multicharged strongly solvated ions, e.g., halides of rare-earth elements with the cations strongly bound with organic ligands partially substituting their solvate shells.66 The effect of the addition of crown ethers 9 and 19 and cryptand 14 on the thermodynamics of electrolytic dissociation and the conduction of KSCN in such solvents as AN, DMSO, PC and their binary mixtures with chlorobenzene was also studied.67, 68 It was found that throughout the range of mixed solvent concentrations (e=5 ± 66), the association constants are smaller for electrolytes with complex cations than for simple salts. However, the conductivity in the systems studied (cKSCN=1073 ±1074 mol litre71) increased with the addition of macrocyclic compounds as a result of a decrease in the degree of electrolyte association only for mixed solvents enriched with the nonpolar component.In individual solvents with relatively high dielectric constants (AN, DMSO and PC) and binary mixtures with high contents of these solvents, the degree of electrolytic dissociation of potassium thiocyanate was high; however, the conduction decreased upon addition of crown ethers and crypt- ands. This effect was attributed to the increase in the cation size with the formation of complexes comprising macrocyclic ligands, which reduced the cation mobility. In addition to the change in the ion size, the nature of the ligand attached also plays a certain role. It was shown 69 that, as compared with L0 of free ions in AN, a decrease in L0(ML+) (M+=Na+, K+) is more pronounced for complex ions containing macrocyclic ligand 9 with two tert-butyl substituents than for complex ions with similar but larger ligands with four tert-butyl substituents.The explanation of these results invoked the fact that in the latter case complex cations [ML]+ were less stable and retained their association with anions. The molar conductivity of potassium chloride solutions with additions of cryptand 14 was calculated for different interaction potentials using the Monte-Carlo method.70 For KCl solutions with concentrations of 0.10 and 0.15 mol litre71 at 298 K, the values of L were found to be 128.82 and 126.52 S cm2 mol71, respectively, whereas for the same solutions with additions of cryptand 14 the L values turned out to be substantially smaller, viz., 71.5 and 65.9 S cm2 mol71, respectively.For a constant cryptand-to-salt concentration ratio, the molar conductivity decreased as the concentration increased in the range from 0.03 to 0.15 mol litre71. If the overall interaction potential took into account the hydrodynamic component in addition to the Cou- lomb attraction and repulsion forces, an adequate agreement of calculated and experimental L values was observed. However, the authors failed to derive a generalised interaction potential which would allow one to adequately simulate both the structural (neutron diffraction spectra) and dynamic (conduction) proper- ties of the system studied. 2. Lithium electrolytes with molecularly encapsulated anions By binding anions of lithium salts into complexes, it is possible to reduce not only the conductivity of a solution (due to a decrease in the degree of ion association), but also the transport number of lithium.Most of the known ligands capable of forming anionic complexes cannot be used in lithium CPS, because they either coordinate anions by forming hydrogen bonds or comprise positively charged fragments (or metal atoms) acting as Lewis acids. Recently, new neutral ligands, viz., cyclic and linear aza- compounds, in which the electron-acceptor groups CF3SO2 were substituted for hydrogen atoms in NH groups [octakis(trifluor- omethylsulfonyl)pentaethylene hexamine (21), hexakis(trifluoro- methylsulfonyl)hexacyclene (22), etc.] were studied.71 ± 73 Nitrogen atoms of nonsubstituted polyamines display electron- donor properties, which favour thewith cations.Upon introduction of electron-acceptor groups, these ligands become the receptors of anions due to the local positive charge that appears on the nitrogen atoms as the electronic density shifts to the substituent. The formation of complexes by linear ligands and anions was confirmed by the NEXAFS method (X-ray absorption spectra in solution). For complexes with cyclic ligands isolated in the solid state, this was confirmed by X-ray diffraction technique. It was shown that, for an equimolar concentration ratio of ligands and lithium halides (LiCl, LiBr, LiI, c=0.1 or 0.2 mol litre71), the conductivity of THF solutions increased by 1 ± 3 orders of magnitude (Table 4).The conductivity also increased with an increase in the number of sulfonamide groups NSO2CF3 for both cyclic and linear ligands. For LiBr and LiCl solutions in THF, the conductivity increased with an increase in the ligand concentration, however, to a lesser extent; the conductivity of a THF solution of lithium iodide remained virtually unchanged as the ligand-to-salt concentration ratio changed from 1 : 1 to 1 : 2. The greatest increase in conductivity with the introduction of macrocyclic ligands was observed for solutions of the most strongly associated salt, viz., lithium chloride, which exhibited the lowest conductivity. It was noticed that, upon introduction of macrocyclic ligands comprising six and seven sulfonamide groups in a ring, the conductivity of lithium halide solutions in THF increased to practically the same value (*1.661073 S cm71), although their solutions without addi- tives exhibited quite different conductivities.Yet another group of anionic receptors comprises boron- containing ligands with fluorinated aryl and alkyl groups: boranes [(C6F5)3B (23), etc.], borates [(C6F5O)3B (24), etc.], boronates [(C6H3F)O2B(C6H3F2) (25), etc.].74 ± 77 As compared with borane (23), borate (24) has a higher solubility in solvents of low polarity Table 4. The effect of the addition of equimolar amounts of boron- and nitrogen-containing anionic receptors on the conductivity of lithium electrolytes at 298.15 K.71 ± 77 Electrolyte Solvent�DME 0.2 M LiCl 0.2 M LiCl+0.2 M (24) 0.2 M LiCl+0.2 M (23) 0.2 M LiI 0.2 M LiI+0.2 M (24) 0.2 M LiI+0.2 M (23) 0.2 M LiCF3COO 0.2 M LiCF3COO+0.2 M (24) 0.2 M LiCF3COO+0.2 M (23) 0.2 M LiCF3COO+0.2 M (25) 0.8 M LiF+0.8 M (24) 0.8 M LiF+0.8 M (23) Solvent�THF 0.1 M LiCl 0.1 M LiCl+0.1 M (21) 0.2 M LiCl 0.2 M LiCl+0.2 M (22) 0.1 M LiBr 0.1 M LiBr+0.1 M (21) V N Afanas'ev, A G Grechin s /S cm71 5.061076 3.061073 2.661073 7.361074 3.261073 2.261073 3.361075 3.361073 3.261073 1.261073 6.461073 6.661073 7.561076 1.661073 1.661076 1.461073 2.961075 1.861073Chemical modification of electrolytes for lithium batteries R R N N R N N R NR2 NR221, R=CF3SO2 F F B F 3 F F 23 F F O F B O25 (e.g., their solubilities in dimethyl carbonate at room temperature were >1 and 0.3 mol litre71, respectively).The complexing of these ligands with halide ions was also confirmed by X-ray absorption spectra of solutions (NEXAFS). The addition of boron-containing ligands to lithium halide solutions in DME substantially increased the conductivity and solubility (see Table 4). For example, lithium fluoride was practically insoluble in solvents with low dielectric constants; however, the addition of a compound 23 or 24 (1 : 1) to its 1 M solution in DME yielded solutions with conductivities of (6.2 ± 6.8)61073 S cm71. The conductivity was shown to increase with an increase in the number of fluorine atoms in the ligand molecules.Boron-containing ligands, especially boranes, are electrochemically stable and compatible with the lithium anode. Thus for the 1 M LiF+1 M 23 solution in a mixed PC ± ethylene carbonate ± dimethyl carbo- nate (1 : 1 : 3) solution, the range of working potentials reached 5.0 V. Compared with such electrolytes as LiPF6 and LiAsF6, lithium halides are cheaper and less toxic. Taking into account that the starting reagents for the synthesis of boron-containing anionic receptors are also inexpensive and of a low toxicity, electrolytes with complex anions have the prospects of gaining a wide use in future. III. Polymeric electrolytes Currently, keen attention is drawn to the safety of application and utilisation of waste power sources. In this respect, metallic lithium is an environmentally clean material as compared with lead and cadmium.The safety of lithium CPS for the environment and human beings is ensured by the correct choice of nonaqueous electrolyte solutions. Numerous studies in the field of polymeric lithium electrolytes were largely aimed at solving the problems of enhancing the stability of the electrolytic system with respect to the lithium anode and increasing the safety and reliability of the performance of lithium batteries used for energising domestic devices.78 The main drawback of polymeric electrolytes is their low conductivities at room temperature. Under the conditions men- tioned, these electrolytes are mostly in a quasicrystalline state (s=1076 ± 1077 S cm71).The conductivity increases only with the transition to the amorphous state (as the temperature rises). To improve the characteristics of polymeric electrolytic systems, different additives were extensively used, along with the proce- dures of modifying the polymer structure itself (which allows, e.g., the reduction of the glass transition temperature). The addition of R R N N R N N R N N R R 22, R=CF3SO2 F F O B F 3 F F 24 781 plasticisers that, as a rule, represent electrolyte solutions in low- molecular aprotic solvents led to formation of hybrid biphasic gel- electrolytes with acceptable conductivities. However, they were highly active with respect to lithium electrodes. Additions of macrocyclic compounds weaken the association of ions in non- polar polymeric matrices, thus increasing the conductivity.The effect of cryptand 14 additives on the conductivity of sodium salt of an oxysiloxane polymer was studied.79 It was found that with addition of the cryptand, the conductivity increased by a factor of 15 irrespective of temperature and concentration and was 1.761075 S cm71. It was concluded that, in the general case, the conductivity should increase with an increase in the equili- brium constant of the formation of cation ± cryptand complexes and an increase in the association constant of the polymeric electrolyte. This agrees with the conclusions made in the studies of solutions of liquid-phase lithium electrolytes.Cations are free charge carriers despite their sufficiently strong solvation by the polymer, and a cryptand additive should not cause any substantial conductivity rise. This was confirmed in several studies.80 ± 84 Additions of cryptand 14 or crown ether 8 to polymeric electro- lytes based on NaSO3CF3, NaBF4, NaI and NaSCN solutions in amorphous polyethylene oxide were shown to reduce their sol- ubilities and condutivities.80, 81 Using spectroscopic methods, it was shown that association of these salts is insignificant and all polymer ± salt complexes studied were amorphous. At the same time, the solubility and conductivity of NaSO3CH3 which is poorly soluble in the polymer increased with the addition of the crown ether and cryptand but did not reach the values observed for the other electrolytes of this system.An analysis of parameters of the Vogel ± Tamman ± Fulcher (VTF) equation which describes the temperature dependence of conductivity showed that the addition of the cryptand decreases the number of charge carriers for electrolytes with weakly associated NaSO3CF3, NaBF4, NaI and NaSCN salts, increasing it only for electrolytes containing the more strongly associated salt NaSO3CH3. Solid electrolytes based on Li[CF3SO2N(CH2)3OCH3] com- plexes with cryptands 12 ± 14 and crown ethers 7 ± 9 and Li[(CF3SO2)2N] complexes with 9 were also studied (Fig. 3).83, 84 It was assumed that amorphous phases are formed due to the difference in the radii of the macrocycle cavity and the cation.The method of differential scanning calorimetry has shown that the stable macrocycle ± salt complexes [ML]+X7 (M is an alkali metal, L is a crown ether or cryptand), in which the sizes of the cation and the ligand's (7, 12) cavity match most closely one another, form only crystalline phases. A series of amorphous complexes of macrocyclic compounds and lithium salts were logs (S cm71) 72 73 74 123456 75 76 3.2 3.0 2.8 103 T71 /K71 2.6 Figure 3. Temperature dependence of the specific conductivity (s) of (1) [Li(14)][CF3SO2N(CH2)3OCH3], amorphous complexes: (2) [Li(13)][CF3SO2N(CH2)3OCH3], (3) [Li(12)][CF3SO2N(CH2)3OCH3], (4) [Li(9)][CF3SO2N(CH2)3OCH3], (5) [Li(8)][CF3SO2N(CH2)3OCH3] and (6) [Li(7)][CF3SO2N(CH2)3OCH3].81782 synthesised where the size ratio mentioned was violated.Com- plexes with cryptands passed into the amorphous state at lower temperatures as compared with complexes with crown ethers. The presence of asymmetric anions favoured the glass transition.84 Amorphous samples with the lowest glass transition temperature exhibited the greatest conductivity (1074.5 S cm71). An analysis of IR and Raman spectra has shown that cryptands most effectively screen lithium cations. The parameters of the VTF equation have shown that cryptate complexes exhibit higher concentrations of charge carriers as compared with complexes with crown ethers (see Fig. 3). An increase in conductivity at room temperature by 2 orders of magnitude (up to 1074 ± 1073 S cm71) for polymeric electro- lytes based on LiClO4 and polyethylene oxide observed upon introduction of crown ether 8 was explained 85 by the formation of stable complexes of crown ethers with lithium ions.However, the addition of 1,4,7,10,13-pentaoxa(13)orthocyclophane(benzo-15- crown-5) had practically no effect on the conductivity in the system mentioned, because the formation of complexes with lithium ions was complicated. The solid polymeric electrolytes synthesised were well compatible with the lithium anode and stable during long-term storage. New synthetic methods for polymeric and liquid electrolytes of the series of nitrogen-containing compounds, viz., aromatic amines, diarylamines and organosilicon compounds with ammo- nium groups and their practical applications as the additives have been reported.86 Organosilicon derivatives of 4,4 0-bipyridyl were referred to as most promising.The additives mentioned almost double the conductivity of polymeric electrolytes and prevent the capacity fade during cycling. The mobility of ions in polymeric electrolytes is directly associated with the mobility of polymeric chain fragments, which is the main factor limiting the conductivity. An approach to enhancement of the conductivity and simultaneous increase in the transport number of lithium ion is the use of compounds which act as Lewis acids and can interact with anions. For example, good results were obtained where cyclic boron-containing compounds 26 (n=1 ± 3, 7.2), particularly, 2,4,6-tris(2-methoxyethoxy)bor- oxine were introduced into polymeric electrolytes based on LiSO3CF3 and poly(methyl methacrylate).87 This was explained by the effective interaction of oligoether oxygen-containing fragments with lithium ions and the anion-acceptor nature of boroxine rings.Introduction of 11-acryloyloxy-3,6,9-trioxaun- decyl biphenyl-2,2 0-diyl borate (27) into solutions of strongly associated electrolytes of LiCl and CF3CO2Li in both polar and low-polar media, as well as of electrolytes formed by LiBF4 and 1 in DME and dimethyl carbonate (DMC) increased the conduc- tivity.88 In contrast, introduction of the same additive to solutions of well dissociated salts, namely, LiBF4 and 1 in polar solvents resulted in a decrease in the ionic conduction, which was explained by a decrease in viscosity (Table 5).It is of note that the conductivity increased for all gel-polymeric electrolytes, because the additive introduced was built-in into the polymer structure in such a way that the microviscosity of the solution surrounding the ions remained unchanged. O O(CH2CH2O)nCH3 CH3(OCH2CH2)nO B B O O BO(CH2CH2O)nCH3 26 O O B (OCH2CH2)4 O O 27 V N Afanas'ev, A G Grechin Table 5. The effect of the addition of equimolar amounts of 27 on the conductivity of lithium electrolyte solutions (cm=0.2 mol kg71) at 298.15 K.88 Electrolyte Salt Conductivity (mS cm71) in a medium of DMC PC g-BL DME LiCF3COO 5.2 1.2 without addition 0.33 0.82 0.024 <5.61074 2.8 0.19 with addition LiBF4 0.19 2.5 4.3 3.1 without addition 2.7 2.2 1.561073 0.027 with addition 0.25 0.42 3.0 6.9 4.3 3.3 Li(CF3SO2)2N without addition 3.0 2.4 with addition In a number of studies, it was suggested to use lithium salts with polymeric anions. Thus for electrolytes based on acrylates of a co-polymer of ethylene oxide and propylene oxide and contain- ing a lithium salt with a polyimide anion, viz., lithium poly(5-oxo- 3-oxa-4-trifluoromethyl-1,1,2,2,4-pentafluoropentamethylene- sulfonimide) (28), the conductivity was of an order of magnitude of 1075 S cm2 and the transport number of lithium >0.7 substantially exceeded the value obtained for polymeric electro- lytes with a `monomeric' lithium salt 1.89 Polyanionic electrolytes prepared by introducing Lewis bases (imidazole) into the poly- meric structure (the `base in chain' method) were studied.90 Plasticisation of the corresponding lithium polyimide salts (29a, b) by chloroaluminates (1 : 1) and AlCl3 made it possible to substantially decrease the glass transition temperature (<720 8C) and obtain liquid-phase systems with the electrolyte conductivity values currently best for polyanionic electrolytes.Thus the conductivity of lithium poly(carbamine sulfonylimide) (29a) reached the maximum value upon addition of 20% LiAlCl4 (1072.9 S cm71 at 25 8C) (Fig. 4), whereas, for a similar complex with the 1,3-dichlorosulfonyl-1,3-dilithiocarbamide (29b), the maximum conductivity value obtained upon introduction of 40% LiAlCl4 did not exceed 1073.3 S cm71. However, the pres- ence of Lewis acids (AlCl3, BF3 ) in the system can initiate corrosion of the power source materials and other undesirable electrochemical processes.In this connection, it was proposed to `fix' a Lewis acid directly in the polymer chain. For this purpose, the 7[O7B7(Ph)7O7([CH2]2O)n] (n=2 ± 23) polymers syn- thesised by the `acid in chain' method were transformed into polymeric anions by complexation with lithium salt anions representing Lewis acids (e.g., Ph7, CN7).91 The largest con- ductivity values were obtained for short polyether chains ([CH2]2O)n which separated the anionic centres in the polymer. The polymers thus synthesised were reported to exhibit lower chemical activities and better compatibility with cathodic and anodic materials as compared with their analogues obtained by the `base in chain' method.It is expected that higher values of ionic conduction can be reached by using lithium compounds with anions of a higher basicity, such as methoxide and formate ions, as well as with the methyl anion (CH¡3 ). Li+ O O 7 S N C CF CF2 O CF2 n O CF3 28 O O 7 7 ClSO2 ClSO2 7 7 C C SO2Cl SO2 Cl n NLi+ NLi+ NLi+ NLi+ 29a 29bChemical modification of electrolytes for lithium batteries logs (S cm71) 298.15K 72 73 I 74 75 76 123456 77 II 78 3.5 2.5 103 T71 /K71 3 Figure 4. Temperature dependence of the specific conductivity of poly- meric electrolytes based on lithium complexes with polyanions, viz., (17x)29a ± AlCl3 (1 : 1) ± xLiAlCl4 (see Ref.90). Arrows designate the conductivity of (I ) an electrolyte without LiAlCl4 at 298.15 K and (II ) an electrolyte containing 20 mol.% LiAlCl4 at the glass transition temper- ature; x: (1) 0, (2) 0.1, (3) 0.2, (4) 0.3, (5) 0.4, (6) 0.5. Polymeric electrolytes with relatively high conductivities can be obtained not only in the amorphous state but also in the crystalline state. The following `rigid' polymers were synthesised: poly(vinylene carbonate) (30) and poly(2-oxo-1,3-dioxolane-4,5- diyl oxalate) (31).92 In contrast to most polymeric electrolytes known, the addition of a lithium salt (LiCF3SO3) to these polymers decreased the glass transition temperature and the maximum conductivity was reached at a higher salt concentration (the molar concentration ratio of the salt to the monomeric units of polymers was 1 : 1).The crystalline structure of these polymers did not prevent the ionic transport, because the mobility of ions in the polymer is independent of the segmental mobility of polymeric chains. O(O)C C(O)O n n O O O O O O31 30 In a number of cases, solid polymeric electrolytes based on amorphous co-polymers of acrylonitrile with the maximum con- centration of lithium salt were found 93 to exhibit unusually high conductivity, viz.,*1073 S cm71. Moreover, the conductivity of systems studied was shown to be either virtually independent of the temperature or have a weakly pronounced inverse temperature dependence as compared with the Arrhenius-type dependence.Measurements of charge transfer numbers showed that it is mainly cations that are involved in the ionic transport. An analysis of IR spectra of electrolytes confirmed the presence of substantial amounts of ion pairs, triplets and more complex associates of lithium salts. Based on these results, the authors proposed that highly concentrated macromolecular polymeric solutions have a special structure which involves ionic clusters interacting with one another. The ionic transport in such solutions proceeds by an unusual, low-energy mechanism. The use of extremely high (close to the solubility limit) concentrations of lithium salts in polymeric matrices was assumed to be promising for the development of highly conductive solid polymeric electrolytes.783 Recently, great success was achieved in studying the so-called ionic liquids, viz., salts with unusually low melting points, which remain in the liquid state even at room temperature. Such systems are of great interest as the electrolytes for lithium CPS. Based on lithium salts and Lewis acids (AlCl3), the following liquid electro- lyte systems with conductivities of >1073 S cm71 at room temperature were synthesised: 0.33 LiCF3SO3 ± 0.67 AlCl3,94 0.4 LiSCN ± 0.6 AlCl3.95 It was assumed that, due to ion ± mole- cule interactions, the following complex anions are formed in these systems: [AlnCl3n+1]7, [AlnCl3n7m+1(SCN)m]7, etc., and the melt structure changes, which makes crystallisation difficult and is favourable for the system to retain its liquid or glassy state.Moreover, the charge transfer mechanism is virtually independent of the structural relaxation of the medium (the charge is trans- ferred preferentially by cations). The above methods for modifying chemically the electrolytic systems are different; however, they have a common theoretical basis determined by the ion-migration (Stokes) mechanism of charge transfer. These methods were largely applied to dilute solutions of lithium electrolytes in aprotic solvents and polymers. Concentrated electrolyte solutions in aprotic solvents, which are used in lithium CPS, still remain insufficiently studied.The results obtained to date 9, 53, 54 suggested that, in low-polar solvents, the greatest increase in conductivity upon addition of macrocyclic compounds takes place at concentrations below the molar con- ductivity minimum point, i.e., where ion pairs prevail in solution. For higher concentrations (in the `anomalous' range), the con- ductivity increased to a lesser extent, apparently, due to the changes in the nature and composition of ions. For strongly associated electrolytes in solvents with low and moderate dielectric constants and polymers, the conductivity increases mainly due to the weakened association and increased solubility. The most pronounced weakening in the ion association was observed with the formation of lithium ± cryptand complexes as compared with lithium ± crown ether complexes.As a rule, an increase in conductivity in these media is proportional to the increase in the stability constant of the complex ions formed. With an increase in ligand concentration, conductivity increases and reaches a maximum at the equimolar Li :L ratio. In more polar solvents, where the electrolyte is sufficiently soluble and weakly associated, and the size of the ligand attached exceeds that of the solvating shell, the mobility of ions can decrease due to an increase in their Stokes radii. For electrolytes in polar media, in which the ions are strongly solvated (e.g., for lithium salt solutions in propylene carbonate, Ka&0), the conductivity can increase as a result of weakening in solvation, and, correspondingly, a decrease in the `free' (solvated) ion radius with the formation of a complex with a macrocyclic ligand.At the same time, in certain cases, a substantial weakening in the lithium ion solvation upon its complexing can increase the association constants for the electro- lytes involving complex cations as compared with ordinary electrolytes. IV. Electrolytic solvosystems for lithium batteries and mechanisms of charge transfer in solutions To understand the reasons for the electrochemical behaviour of electrolytic systems of different nature in wide ranges of electro- lyte concentrations and temperatures, it is necessary to gain an insight into the nature of charge carriers and the mechanism of charge transfer.It is known that the high conductivity typical of concentrated aqueous solutions of bases and acids used as the electrolytes in conventional CPS is caused by the chain mechanism of conduction.13, 96, 97 This mechanism is distinguished by the exchange interaction between the electrolyte ions which corre- spond to the lyonium and lyate ions of the solvent and the solvent molecules (H2O, H2SO4, H2SeO4, H3PO4, HF). Thus in an aqueous solution, hydroxonium ions (hydrated protons) and water molecules exchange protons784 2H2O H3O++OH7, H3O+ +H2O H2O+ H3O+. Several approaches were proposed for estimating different contributions to the charge-transfer mechanism in solu- tions.13, 66, 98 ¡À 103 However, to date, no sufficiently reliable criteria and methods were proposed for elucidation of the nature of charge carriers, which makes the studies of the conduction mechanism in nonaqueous media quite difficult.104 The greater complications arise for concentrated electrolyte solutions for which the concepts of the Debye ¡ÀHu�� ckel ¡À Onsager theory and the law of independence of ions' motions (the Kohlrausch law) are not fulfilled.One of the approaches to assessing the conduction mechanism in electrolyte solutions consists in comparing the activation energies of the conduction and viscous flow processes. For example, within the framework of this approach, it was assumed that the charge transfer in the LiCl ¡À AlCl3¡ÀCH3NO2 ¡À SOCl2 system (see Refs 27, 28) with the electrolyte concentration >1 mol litre71 occurs by both the ion-migration and chain mechanisms.The latter mechanism provided high conductivity values during the lithium battery discharge irrespective of the observed increase in the solution concentration and viscosity. The conduction mechanism may be characterised by the transport number of ions (ti), which however is insufficient to confirm the chain mechanism of conduction. Transport numbers of ions for most binary aqueous solutions of salts insignificantly deviate from the average value of 0.5; in aqueous solutions of acids and bases, tH+ and tOH7 are substantially higher, viz., from 0.75 to 0.85. In nonaqueous solvents, transport numbers strongly depend on the differences in the degrees of solvation for cations and anions.For example, the greatest part of solvation energy for lithium salts in PC (up to 80%) falls to cations.17 Hence, as a rule, tLi+ is smaller than the transport number of the corresponding anion and is equal to 0.3 ¡À 0.5. The known attempts to analyse the experimental results on conductivities and the solution structure ambiguously assessed these data and often led to unorthodox conclusions. Thus it was suggested 105, 106 that electrolyte solu- tions be considered as impurity conductors in which the charge is transferred by electrons and holes rather than by ions. Based on high-frequestudies, electrolyte solutions were concluded to have ionic lattices like those in crystals and the presence of a long- range order was assumed to be a key condition for enhancing the mobility of charge transfer in condensed media.The possibility of realising the chain mechanism of conduction in solutions is largely determined by the structure and the degree of self-ionisation of the solvent, which depends (in protic solvents) on the strength of bridging hydrogen bonds between mole- cules.96, 97 For instance, in liquid ammonium the constant of self- ionisation (at 750 8C) was found to be *10723 due to a weak N. . .H. . .N bond and a low e, and the mobility of ions formed upon autoprotolysis was not anomalously high.107, 108 Such an order of magnitude is typical of self-ionisation constants of most nonassociated aprotic solvents.Nonetheless, the charge transfer by the chain mechanism can also occur in aprotic media. In this case, anomalous mobility values were observed for lyonium and(or) lyate ions formed upon self-ionisation of solute associates. n¡¦m of In binary liquid systems formed by protic (H2SO4, H3PO4) acids and proton-containing organic bases (RH) (such as acet- amide, N-methylacetamide, N,N-dimethylacetamide, etc.), in the range of high acid concentrations, the charge was shown to be transferred by the chain mechanism as in individual acids.109 ¡À 113 Bearing in mind that, for binary systems of organic bases with strong aprotic acids (MXn) (SbCl3, FeCl3, SnCl4, SbCl5), the conductivity isotherm has a maximum, it may be assumed that the charge transfer in these systems also occurs by the chain mechanism in the range of high acid concentrations.The compo- nents of these systems are prone to autoionisation; hence, the acid ¡À base interaction occurs between the lyonium ion MX�¢ the acid and the lyate ion R7 of the base to give a nonelectrolyte MXn7mRm. This results in the increase in the concentration of the V N Afanas'ev, A G Grechin4 lyate ions of the acidMXp¡¦ n�¢1 and the lyonium ions of the baseRH�¢2in the solution. It is the mobility of these ions that determines the conduction of solutions. For chlorides (solvated chloride-containing ions , e.g., SbCl¡¦ or AlCl¡¦4 ) in ultra-concentrated solutions based on antimony trichloride and aluminium chloride, the following halogenotropic mechanism of conduction was postulated:109 ¡À 111, 114, 115 AlCl3+AlCl¡¦4 , AlCl¡¦4 +AlCl3 SbCl3+SbCl¡¦4 .SbCl¡¦4 +SbCl3 For bromides and iodides (polyiodides, I¡¦3 , Br¡¦3 ), PCl5 and PBr5 in liquid bromine and iodine, the same mechanism was suggested.116 ¡À 118 I¡¦ I2+ I¡¦ 3 +I2 3 ; Br2+Br¡¦3 . Br¡¦3 +Br2 It was shown 119 that the conduction of salt solutions in low- polar solvents in a wide range of concentration (from 0.1 mol litre71 to fused salts) is largely determined by self-ionisa- tion (autosolvation) of the electrolyte nAK (AK)n K(AK)�¢m +A(AK)¡¦n¡¦1¡¦m , m and A(AK)¡¦n¡¦1¡¦m are the autosolvated cation and where K(AK)�¢ anion, respectively. According to concepts put forward,116 ¡À 118 ions in concen- trated solutions form aggregates which are involved in the charge transfer.Moreover, with an increase in electrolyte concentration, upon passing its minimum at 0.01 ¡À 0.1 mol litre71, the molar conductivity of low-polar solutions was shown to increase due to the increased number of such aggregates, irrespective of the simultaneous increase in the solution viscosity. A hypothesis that an electrolyte in a concentrated solution passes into a state similar to that of a melt, which was put forward in the studies under discussion, was confirmed by the results of studying the molten ZnBr2 ¡À AlBr3 and SbBr3 ¡À AlBr3 systems. With an increase in salt concentration, the molar conductivity values of these melts (with a correction made for viscosity) increased up to a point where the conductivity of the individual liquid electrolyte is reached.According to this hypothesis, for any concentration of an electro- lyte solution, its molar conductivity (with a correction made for its viscosity) cannot exceed the corresponding value of a hypothetic molten individual electrolyte under equal conditions. The anom- alous conductivity can be observed not only for solvents with low dielectric constants but also in polar solvents at sufficiently high electrolyte concentrations. For both cases, a similar type of the dependence of the molar conductivity on the concentration was suggested. Similar conclusions were drawn in Refs 32 ¡À 34. Based on the results of studying thermodynamic and transport properties of lithium electrolytes in a wide range of concentrations in solvents of different types (glymes, sulfamides, AN, g-BL, PC and their mixtures), it was concluded that at least two conduction mecha- nisms should be taken into account, viz., the ion-migration mechanism, which occurs at low electrolyte concentrations, and the chain mechanism (charge transfer accompanied by rotation of complexes-associates), which occurs at high concentrations.It was assumed that the state of a highly concentrated electrolyte solution is similar to that of a fused salt. The chain mechanism of conduction can occur in both low-polar and polar aprotic solvents with sufficiently high electrolyte concentrations.In highly con- centrated solutions, as the distance between ions decreases, the chain mechanism prevails and the difference in the conductivity values for solvents with low and high e values decreases. A low viscosity of the medium and a high solvation power of the solvent, which weaken the ion association and facilitate reorientation of dipoles in the electric field, favour the charge transfer to occur by the chain mechanism.Chemical modification of electrolytes for lithium batteries However, for the chain mechanism (in contrast to the ion- 4 migration mechanism), many properties of electrolytic systems (viscosity, solvation, association) can exert the opposite effects on the conduction. For example, despite the high viscosity of anhydrous sulfuric acid, the mobilities of the lyonium H3SOá and the lyateHSO¡3 ions are comparable with the mobilities ofH+ and OH7 ions in water; in aqueous solutions, the mobility of protons substantially exceeds those of other ions, even if the hydration energy ofH+ is the highest; among the large number of electrolytes completely dissociated in aqueous solutions, only acid and base solutions, i.e., electrolytes that comprise the ions formed by the self-ionisation (autoprotolysis) of the solvent have anom- alously high conductivity values.The prototropic conduction decreases with an introduction of both electrolytes and non- electrolytes into these solutions,97 which is usually attributed to the hydration of ions and molecules accompanied by the changes in the water structure built by hydrogen bonds.Such a structure is best suited for the chain transfer of protons. At the same time, it was found 120 that the additivity of the molar conductivity is typical of the whole composition range of an isomolar HCl+KCl mixture in methanol. A similar L vs. composition dependence was also observed for an aqueous HCl+LiCl system (cLi48 mol litre71). Taking into account these results, it was concluded that the prototropic mechanism of proton transfer is retained even for high concentrations of aprotic electrolytes. It should be also noted that, in contrast to other ions, the mobilities of H+ and OH7 increase with an increase in the pressure.96 The temperature coefficient of conduction is positive for both aqueous solutions of acids and bases and other electro- lytes with the usual ion-migration mechanism of conduction.The effect of the temperature on the structure of hydrogen bonds in water is apparently less significant compared with the effecsure and chemical factors (solvation). A comparison of differ- ent physicochemical parameters of water and other liquids in a wide temperature range have shown that water retains a suffi- ciently distinguishable structure up to the critical point.121 An analysis of the conduction data for different kinds of aqueous and nonaqueous electrolyte solutions have shown that the chain mechanism of conduction is realised in aprotic media if, first, the chemical nature of the electrolyte is similar to that of the solvent (solutes comprise lyonium or lyate ions of the solvent) and, second, at high concentrations, the solvent and/or the electrolyte form a polymeric structure with intermolecular bonds.The presence of a certain solution structure is often charac- terised by high viscosity values (e.g., H2SO4, H3PO4, H2SeO4, concentrated electrolyte solutions). On the other hand, such associated liquids as water, (HF)n, SbCl3, etc., in which the charge is transferred by the corresponding ions by the chain mechanism, exhibit relatively low viscosities. Evidently, under conditions necessary for the chain mechanism to occur, an additional decrease in the viscosity should facilitate the charge transfer. To develop nonaqueous electrolytic solvosystems with the chain mechanism of conduction, it is necessary to carry out a quest for new-in-principle lithium electrolytes and nonaqueous media which will match the requirements described above.One of the approaches in such a research is to develop electrolytic solvosys- tems based on organolithium compounds. Organolithium com- pounds pertain to the group of the so-called electron-deficient structures, like organoelement compounds of the Groups II and III on the Periodic System (R2Be, R3B, R3Al).122, 123 At the same time, organolithium compounds behave in reactions as the reagents with a deficit of electrons and react in the same manner as carbanions. The deficit of electrons manifests itself in a strong association of molecules nRLi>(RLi)n (n=2 ± 6) by means of bridge bonds formed by the electrons of the C7Li bond and free 2p orbitals of the lithium atoms.It was assumed 124 that the nature of . . . Li7C. . . Li7C. . . `lithium' bonds in the associates of organolithium compounds is similar to that of hydrogen bonds. 785 A possibility that lithium ions of an electrolyte Li+X7 and a solvent LiR are transferred by a mechanism similar to that of proton transfer in water, i.e., LiR+Li2R+ Li2R+ +LiR should be attributed to dissociation (self-ionisation) of the asso- ciate (RLi)n (a dimer, in the simplest case) Li2R+ +R7 . Li2R2 The compound LiR should preferentially be of the nucleo- philic nature, i.e., have the affinity to lithium ions, and the lithium salt should contain a weakly basic anion (X7) for these processes to be realised.It is known that the complexes formed by aliphatic and aromatic organolithium compounds with lithium halides, which are less reactive than the original organolithium compounds, are stable in air and more handy in applications. The ability to form complexes depends on the nature of the radical (R), the lithium salt anion and the solvent. It is also of note that polyacetylene films (s=10712± 1079 S cm71) doped with organolithium com- pounds had an anomalously high conductivity (20 ± 160 S cm71) 125 typical of electron-conducting polymers and charge-transfer complexes.126 In this connection, it seems inter- esting to explore the possibility of using organolithium com- pounds as one of the components of the medium (solvent) for electrolytes in lithium CPS.* * * In the past decade, the main research on electrolytic systems for lithium power sources was directed to modifying chemically the electrolytes and solvents traditional for this field. A key problem for these studies is the enhancement of conductivity of lithium salt solutions in aprotic solvents and polymers in order to optimise the electrochemical processes and improve the performance of lithium CPS.The advances in the field of physics and chemistry of solutions, chemistry of polymers and coordination chemistry allow one to understand and explain the specific features of charge transfer in condensed media and serve as the theoretical basis for the development of new electrolytic systems.The most promising research directions involve the studies of highly concentrated solutions (including ionic liquids), the use of principles of molecular encapsulation of ions by macrocyclic ligands, and the quest for systems with nontraditional mecha- nisms of ionic transport. However, it is clear that the possibility of enhancing the conductivity within the framework of the approaches based on the ion-migration charge-transfer mecha- nism, e.g., by reducing the solvent viscosity and the degree of ionic solvation (increasing their mobility), or decreasing the degree of association of the electrolyte (increasing the number of charge carriers) are limited. This review was financially supported by the Commission of the Russian Academy of Sciences on the research carried out by junior scientists (the 6th competition-expertise, Project No.160) and by the Russian Foundation for Basic Research (Project TsKP No. 00-05-401-31). References 1. I A Kedrinskii, V E Dmitrenko, I I Grudyanov Litievye Istochniki Toka (Lithium Batteries) (Moscow: Energoatomizdat, 1992) 2. A G Demakhin, V M Ovsyannikov, S M Ponomarenko Elektrolit- nye Sistemy Litievykh KhIT (Electrolytic Systems of Lithium Chemical Batteries) (Saratov: Saratov State University, 1993) 3. V S Bagotskii, A M Skundin Elektrokhimiya 34 732 (1998) a 4. A M Skundin Elektrokhim. Energetika 1 5 (2001) 5. S Megahed, B Scrosati Electrochem. Soc. Interface 1 34 (1995) 6.A G Richie J. Power Sources 96 1 (2001)786 7. Analiz i Osnovnye Tendentsii Razvitiya Mirovogo i Rossiiskogo Rynka Litievykh Istochnikov Toka (Analysis and Basic Trends of the World and Russian Markets of Lithium Batteries) (Moscow: Institute of Industrial Development, Informelektro, 1996) 8. J Barthel, H-J Gores, in Chemistry of Nonaqueous Solutions (Eds G Mamantov, A I Popov) (New York: VCH, 1994) p. 1 9. J Barthel, H J Gores, in Handbook of Battery Materials (Ed. J O Besenhard) (New York: VCH, 1999) p. 457 10. M Salomon Pure Appl. Chem. 70 1905 (1998) 11. M Salomon, B Scrosati Gazz. Chim. Ital. 126 415 (1996) 12. Yu Ya Fialkov, A N Zhitomirskii, Yu A Tarasenko Fizicheskaya Khimiya Nevodnykh Rastvorov (Physical Chemistry of Nonaqueous Solutions) (Leningrad: Khimiya, 1973) 13.Yu A Karapetyan, V N Eichis Fiziko-Khimicheskie Svoistva Elektrolitnykh Nevodnykh Rastvorov (Physicochemical Properties of Electrolytic Nonaqueous Solutions) (Moscow: Khimiya, 1989) 14. J Barthel, H Krienke, W Kunz Physical Chemistry of Electrolyte Solutions. Modern Aspects (Darmstadt; New York: Steinkopff ± Springer, 1998) 15. V N Afanas'ev, I E Korolev, in Termodinamika Rastvorov Elektroli- tov. Sbornik Nauchnykh Trudov (Thermodynamics of Electrolyte Solutions. Collection of Scientific Proceedings) (Ivanovo: Institute of Chemistry of Nonaqueous Solutions, Russian Academy of Science, 1993) p. 24 16. V N Afanasyev, L A Zyatkova J. Chem. Eng. Data 41 1315 (1996) 17. V N Afanas'ev, L A Zyat'kova, E Yu Tyunina,M D Chekunova Elektrokhimiya 37 56 (2001) a 18.H J Gores, J Barthel Pure Appl. Chem. 67 919 (1995) 19. J Barthel, H J Gores, A Schmid Pure Appl. Chem., 71 1705 (1999) 20. J Barthel,M Wuhr, R Buestrich, H J Gores J. Electrochem. Soc. 142 21. J Barthel, R Buestrich, E Carl, H J Gores J. Electrochem. Soc. 143 22. J Barthel, R Buestrich, E Carl, H J Gores J. Electrochem. Soc. 143 2527 (1995) 3565 (1996) 3572 (1996) 23. J Barthel, R Buestrich, H J Gores, M Schmidt,M Wuhr J. Electrochem. Soc. 144 3866 (1997) 24. J Barthel,M Schmidt,H J Gores J. Electrochem. Soc. 145 L17 (1998) 25. J Barthel, A Schmid, H J Gores J. Electrochem. Soc. 147 21 (2000) 26. I E Korolev, V N Afanas'ev Stabil'nye Elektrolity na Osnove Tsikli- cheskikh Efirov dlya Litievykh KhIT Vysokoi Udel'noi Moshchnosti, Ivanovo, 1991 (Stable Electrolytes Based on Cyclic Ethers for Lithium Chemical Batteries of High Power Density, Ivanovo, 1991); article deposited at the VINITI, No 2083-V91, Moscow, 1991 27.A G Demakhin, Doctoral Thesis in Chemical Sciences, Research Institute of Chemistry, Saratov State University, Saratov, 1999 28. A G Demakhin, N N Kuznetsov, Yu A Rogachev Elektrokhimiya 29 203 (1993) a 29. V S Kolosnitsyn, Doctoral Thesis of Chemical Sciences, Institute of Organic Chemistry, Urals Branch of the Russian Academy of Sciences, Ufa, 1990 30. NV Slobodchikova, Candidate Thesis of Chemical Sciences, Institute of Organic Chemistry, Urals Scientific Centre of the Russian Academy of Sciences, Ufa, 2001 31.V S Kolosnitsyn, N V Slobodchikova, L V Sheina Zh. Fiz. Khim. 75 430 (2001) b 32. D Brouillette, G Perron, J E Desnoyers J. Solution Chem. 27 151 (1998) 33. Y Choquette, G Brisard, M Parent, D Brouillette, G Perron, J Desnoyers, M Armand, D Gravel, N Slougui J. Electrochem. Soc. 145 3500 (1998) 34. D Brouillette, G Perron, J E Desnoyers Electrochim. Acta 44 4721 (1999) 35. A I Mishustin Zh. Fiz. Khim. 70 836 (1996) b 36. B G Cox, H Schneider Coordination and Transport Properties of Macrocyclic Compounds in Solution (Amsterdam: Elsevier, 1992) 37. J-M Lehn Supramolecular Chemistry. Concepts and Perspectives (Weinheim: VCH, 1995) V N Afanas'ev, A G Grechin 38. I P Shestakova, L V Darda, O V Ivanov Osnovnye Perspektivnye Napravleniya Ispol'zovaniya Kriptandov v Nauchnykh Issledovaniyakh i Tekhnologii.Seriya `Reaktivy i Osobo Chistye Veshchestva' (Obzornaya Informatsiya IREA) [Basic Advanced Applications of Cryptands in Research and Technology. Series `Reagent and Extra Pure Compound' (Reviews of IREA)] (Moscow: NIITEKhIM, 1988) 39. Ya Z Voloshin, in XVI Mendeleevskii S'ezd po Obshchei i Prikladnoi Khimii (Referaty Dokladov i Soobshchenii), Moskva, 1998 [The XVIth Mendeleev Congress on General and Applied Chemistry (Abstracts of Reports), Moscow, 1998] p. 60 40. W R Gilkerson,M D Jackson J. Am. Chem. Soc. 104 1218 (1982) 41. U Olsher, R M Izatt, J S Bradshaw, N K Dalley Chem. Rev. 91 137 (1991) 42. M Formica, V Fusi,M Micheloni, R Pontellini, P Romani Coord.Chem. Rev. 84 347 (1999) 43. V N Afanas'ev, A G Grechin, in Problemy Khimii Rastvorov i Tekh- nologii Zhidkofaznykh Materialov (Sbornik Nauchnykh Trudov) [Problems of Solution Chemistry and Technology of Liquid-Phase Materials (Collection of Research Works)] (Ivanovo: Institute of Chemistry of Nonaqueous Solutions, Russian Academy of Sciences, 2001) p. 119 44. B G Cox, H Schneider Pure Appl. Chem. 61 171 (1989) 45. A F Danil de Namor, M A Llosa, J C Ng J. Phys. Chem. 98 11796 46. A F Danil de Namor, M A Llosa,M Salomon Pure Appl. Chem. 67 47. A F Danil de Namor, J C Ng,M Salomon J. Phys. Chem. 100 14 485 48. A F Danil de Namor, M L Zapata-Ormachea, N Al Rawi J. Phys. (1994) 1095 (1995) (1996) Chem. B 101 6772 (1997) 49.A F Danil de Namor, L E Danil Salazar, R A P Schulz J. Chem. Soc., Faraday Trans. 94 3111 (1998) 50. M Salomon, G T Hefter Pure Appl. Chem. 65 1533 (1993) 51. M Salomon J. Electroanal. Chem. 355 265 (1993) 52. E V Kuz'minskii, V D Prisyazhnyi, E O Berezhnoi, N B Golub Elektrokhimiya 34 528 (1998) a 53. C Chen,W Wallace, E M Eyring, S Petrucci J. Phys. Chem. 88 5445 (1984) 54. MXu,NInoue, EMEyring, S Petrucci J. Phys. Chem. 92 2789 (1988) 55. I S Antipin, A N Vedernikov, A I Konovalov Dokl. Akad. Nauk SSSR 283 139 (1985) c 56. I S Antipin, A N Vedernikov, A I Konovalov Zh. Org. Khim. 25 3 (1989) d 57. A D'Aprano, B Sesta J. Phys. Chem. 91 2415 (1987) 58. A D'Aprano, B Sesta, A Princi J. Electroanal. Chem. 361 135 (1993) 59.A D'Aprano, B Sesta, A Princi, C Filippi, M Iammarino J. Electroanal. Chem. 365 119 (1994) 60. A D'Aprano, M Iammarino, V Mauro, B Sesta J. Electroanal. Chem. 392 27 (1995) 61. A D'Aprano,M Salomon, V Mauro J. Solution Chem. 24 685 (1995) 62. A D'Aprano, J Vicens, Z Asfari, M Salomon,M Iammarino J. Solution Chem. 25 955 (1996) 63. A D'Aprano, B Sesta, A Capalbi, M Iammarino, V Mauro J. Electroanal. Chem. 403 257 (1996) 64. A D'Aprano, B Sesta, V Mauro, M Salomon J. Incl. Phenom. Macrocycl. Chem. 35 451 (1999) 65. M Della Monica, A Ceglie, A Agostiano Electrochim. Acta 29 161 (1984) 66. Yu Ya Fialkov Rastvoritel' kak Sredstvo Upravleniya Khimicheskim Protsessom (The Solvent as Control Means of Chemical Processes) 67. V V Kryukov, Yu A Tarasenko, Yu Ya Fialkov, V L Chumak (Leningrad: Khimiya, 1990) Ukr.Khim. Zh. 51 34 (1985) 68. V V Kryukov, Yu A Tarasenko, Yu Ya Fialkov, V L Chumak Ukr. Khim. Zh. 52 207 (1986) 69. L Tusek-Bozic, B Bozic Electrochim. Acta 30 789 (1985) 70. M Jardat, O Bernard, C Treiner, P Turg J. Phys. Chem. B 103 8462 (1999) 71. H S Lee, X Q Yang, J McBreen, Y Okamoto, L S Choi Electrochim. Acta 40 2353 (1995)Chemical modification of electrolytes for lithium batteries 72. H S Lee, X Q Yang, J McBreen, L S Choi, Y Okamoto J. Electrochem. Soc. 143 3825 (1996) 73. H S Lee, X Sun, X Q Yang, J McBreen, J H Callahan, L S Choi J. Electrochem. Soc. 146 9 (2000) 74. H S Lee, X Q Yang, C L Xiang, J McBreen, L S Choi J. Electrochem. Soc. 145 2813 (1998) 75. X Sun, H S Lee, X Q Yang, J McBreen J.Electrochem. Soc. 146 3655 (1999) 76. J McBreen, H S Lee, X Q Yang, X Sun J. Power Sources 89 163 (2000) 77. H S Lee, X Q Yang, X Sun, J McBreen J. Power Sources 97 ± 98 566 (2001) 78. F B Dias, L Plomp, J B J Veldhuis J. Power Sources 88 169 (2000) 79. M C Lonergan, M A Ratner, D F Shriver J. Am. Chem. Soc. 117 2344 (1995) 80. K E Doan, B J Heyen, M A Ratner, D F Shriver Chem. Mater. 2 539 (1990) 81. R E A Dillon, D F Shriver Chem. Mater. 11 3296 (1999) 82. R E A Dillon, D F Shriver Solid State Ion. 133 239 (2000) 83. R E A Dillon, C L Stern, D F Shriver Solid State Ion. 133 247 (2000) 84. R E A Dillon, D F Shriver Solid State Ion. 132 93 (2000) 85. O V Yarmolenko, A E Ukshe, I K Yakushenko, T I Movchan, O N Efimov Elektrokhimiya 32 508 (1996) a 86.O V Chervakov, E M Shembel', N I Globa, I M Maksyuta, N D Zaderei, A G Rybalko, Yu V Polishchuk, O V Kolomets, in Litievye Istochniki Toka (Materialy VI Mezhdunarodnoi Konferen- tsii), Novocherkassk, 2000 [Lithium Batteries (Proceedings of the VIth International Conference), Novocherkassk, 2000] p. 89 87. M A Mehta, T Fujinami, S Inoue, K Matsushita, T Miwa, T Inoue Electrochim. Acta 45 1175 (2000) 88. T Hirakimoto,M Nishiura,M Watanabe Electrochim. Acta 46 1609 (2001) 89. M Watanabe,H Tokuda, S Muto Electrochim. Acta 46 1478 (2001) 90. S-S Zhang, Z Chang, K Xu, C A Angell Electrochim. Acta 45 1229 (2000) 91. X Sun, C A Angell Electrochim. Acta 46 1467 (2001) 92. X Wei, D F Shriver Chem.Mater. 10 2307 (1998) 93. V M Zhukovskii, O V Bushkova, B I Lirova, A L Kruglyashov Elektrokhim. Energetika 1 43 (2001) 94. N Liu, C A Angell Solid State Ionics 86 ± 88 467 (1996) 95. X Wei, D F Shriver Solid State Ionics 133 233 (2000) 96. J O'M Bockris, B E Conway (Eds) Modern Aspects of Electro- 97. T Erdey-Gru Transport Phenomena in Aqueous Solution (Budapest: 98. Yu A Karapetyan, A D Krysenko, Yu Ya Fialkov Ukr. Khim. Zh. 99. Yu A Karapetyan, A D Krysenko, Yu Ya Fialkov Zh. Fiz. Khim. chemistry (London: Butterworths, 1964) Akademiai Kiado, 1974) 42 1210 (1976) 57 2982 (1983) b 100. Yu A Karapetyan, A D Krysenko Ukr. Khim. Zh. 49 1106 (1983) 101. Yu A Karapetyan, A D Krysenko, Yu Ya Fialkov Ukr. Khim. Zh. 51 1172 (1985) 102. Yu A Karapetyan, A A Rudnitskaya Ukr. Khim. Zh. 54 713 (1988) 103. V A Barbash, Yu A Karapetyan, A D Krysenko Fiziko-Khimiche- skoe Issledovanie Svoistv Individual'nykh H- i L-Kislot, Kiev, 1986 (Physicochemical Investigation of the Properties of H- and L-Acids, Kiev, 1986); article deposited at the Ukrainian NIINTI No 1860, Kiev, 1986 104. A M Kolker, L P Safonova, in Eksperimental'nye Metody Khimii Rastvorov. Densimetriya, Viskozimetriya, Konduktometriya (Experimental Methods of Solution Chemistry. Densimetry, Viscometry, Conductometry) (Moscow: Nauka, 1997) p. 91 105. V I Ermakov, V M Chembai Elektroprovodnost' Mnogokompo- nentnykh Rastvorov Elektrolitov (Conductivity of Multi-Component Electrolyte Solutions) (Moscow: D I Mendeleev Russian Chemical Technological University, 1995) 106. V I Ermakov, in Eksperimental'nye Metody Khimii Rastvorov. Spektroskopiya i Kalorimetriya (Experimental Methods of Solution Chemistry. Spectroscopy and Calorimetry) (Moscow: Nauka, 1995) p. 154 107. V Gutmann Z. Phys. Chem. 203 250 (1954) 108. V Gutmann Chemische Funktionslehre (Vien; New York: Springer, 1971) 787 109. V P Basov, Yu A Karapetyan, A D Krysenko, P M Stadniichuk, Yu Ya Fialkov Dokl. Akad. Nauk SSSR 211 851 (1973) c 110. V P Basov, Yu A Karapetyan, A D Krysenko, G M Romanenko Izv. Vyssh. Uchebn. Zaved., Khim. Khim. Tekhnol. 16 875 (1973) 111. V P Basov, Yu A Karapetyan, A D Krysenko, G M Romanenko, Yu Ya Fialkov Zh. Org. Khim. 45 1428 (1975) d 112. Yu A Karapetyan Ukr. Khim. Zh. 40 1205 (1974) 113. A N Zhitomirskii, P M Stadniichuk Zh. Fiz. Khim. 48 2246 (1974) b 114. A Reger, E Peled, E Gileadi J. Phys. Chem. 87 873 (1979) 115. E Peled, A Mitavski, A Reger, E Gileadi J. Electroanal. Chem. 75 677 (1977) 116. E Ya Gorenbein Zh. Fiz. Khim. 20 881 (1946) b 117. E Ya Gorenbein Zh. Obshch. Khim. 20 547 (1946) e 118. A E Gorenbein, E Ya Gorenbein Zh. Fiz. Khim. 49 371 (1975) b 119. A N Ivashkevich, Doctoral Thesis in Chemical Sciences, Institute of Chemistry of Nonaqueous Solutions, Russian Academy of Sciences, Ivanovo, 1994 120. I E Korolev, V N Afanas'ev, in Rastvory �Elektrolitnye Sistemy (Mezhvuzovskii Sbornik Nauchnykh Trudov) [Solutions as Electro- lyte Systems (Intercollegiate Proceedings)] (Ivanovo: Ivanovo Chemical Technology Institute, 1988) p. 83 121. Y Marcus, in Proceedings of the VIIth International Conference `The Problems of Solvation and Complex Formation in Solutions', Ivanovo, 1998 p. 16 122. A N Nesmeyanov, K A Kocheshkov (Eds) Metody Elementoorga- nicheskoi Khimii. Litii, Natrii, Kalii, Rubidii, Tsezii (The Methods of Organoelement Chemistry. Lithium, Sodium, Potassium, Rubidium, Caesium) (Moscow: Nauka, 1971) 123. B J Wakefield The Chemistry of Organolithium Compounds (Oxford: Pergamon Press, 1974) 124. A N Rodionov, D N Shigorin, T V Talalaeva, K A Kocheshkov Dokl. Akad. Nauk SSSR 123 113 (1958) c 125. B Francois, C Mathis, R Nuffer, A Rudatskira Mol. Cryst. Liq. Cryst. 117 113 (1985) 126. Elektrokhimiya Polimerov (Electrochemistry of Polymers) (Moscow: Nauka, 1990) a�Russ. J. Electrochem. (Engl. Transl.) b�Russ. J. Phys. Chem. (Engl. Transl.) c�Dokl. Chem. (Engl. Transl.) d�Russ. J. Org. Chem. (Engl. Transl.) e�Russ. J. Gen. Chem. (En

ISSN:0036-021X    

出版商:RSC    

年代:2002

数据来源: RSC