CONCLUSIONS A new general method has been developed for the thiylation of diacetylene in liquid ammonia for the preparation of enyne sulfides containing various substituents and functional groups. LITERATURE CITED I. 2. 3. 4. 5. 6. 7. 8. 9.

M . F . Postakovskii and A. V. Bogdanova, The Chemistry of Acetylene [in Russian], Nauka~ Moscow (1971), ~ p. 127. R.N. Kudyakova and A. N. Volkov, Zh. Org. Khim., 13, 934 (1977), A . N . Volkov, R. N. Kudyakova, and B. A. Trofimov, Zh. Org. Khim., 15, 1554 (1979). H. Smith, Organic Reactions in Liquid Ammonia, New York--London (1963), p. 8. J . J . Lagowski, Pure and Appl. Chem., 25, 429 (1971). N . V . Elsanov and A. A. Petrov, Opt. Spektrosk., 16, 797 (1964). E . N . Prilezhaeva, G. S. Vasil'ev, and V. N. Petrov, Izv. Akad. Nauk SSSR, Set. Khim., 2217 (1967). V . S . Bogdanov, T. M. Ushakova, A. N. Volkov, and A. V. Bogdanova, in: The Chemistry of Acetylene [in Russian], Nauka, Moscow (1968), p. 403. M . F . Shostakovskii, N. V. Komarov, and T. D. Burnashova, Izv. Akad. Nauk SSSR, Ser. Khim., 629 (1968).

ROLE OF COMPONENTS OF THE REACTION MEDIUM IN REACTIONS OF ELECTROCHEMICALLY GENERATED RADICAL ANIONS. 2.

ELECTROCHEMICAL HYDRODIMERIZATION OF

~-THIOPHENE ALDEHYDE IN ORGANIC SOLVENTS* V. P. Gul'tyai, L. M. Korotaeva, A. S. Mendkovich, and I. V. Proskurovskaya

UDC 541.13:541.515:547.733

The most advantageous of the existing methods for preparing aromatic pinacols comprises electrochemical hydrodimerization of the corresponding aldehydes or ketones [2~ 3]~ The yield of pinacols obtained by electrolysis of aromatic carbonyl compounds in aqueous organic solvents depends considerably on the nature of the inert electrolyte, e.g., varying from 40 to 70% in the case of the reduction of acetophenone [3]~ At the same time, the yield of pinacols can be increased sharply, e.g., to 95% in the case of 2,3-diphenyl2,3-butanediol [4], by electrochemically reducing such compounds in nonaqueous media. The best results for the electrochemical synthesis of aromatic pinacols are obtained by using nonaqueous polar solvents (DMF or MeCN) and ammonium salts as supporting electrolytes [5]. This conclusion, however, is not in accord with the results obtained by electrolyzing acetophenone in dry acetonitrile [6]. It is known [7, 8] that the synthesis of pinacols by electrochemical reduction of thiophene carbonyl derivatives is hampered by the formation of considerable amounts of resinous products. Only comparatively recently has a yield of 56% been achieved for pinacol synthesis from ~-acetylthiophene in 70% aqueous DMF containg AcOK as supporting electrolyte [9]. The reductive dimerization of ~-benzoylthiophene in MeCN containing 0.i mole of tetraethylammonium perchlorate and added organic acids gives a pinacol yield of 68% [i0]. In the present work we have investigated the influence of the composition of the reaction medium on the previously unstudied electrochemical hydrodimerization of ~-thiophene aldehyde (I) to 1,2-di(a-thienyl)-l,2-ethanediol (II). The latter has previously been prepared by enzymatic reduction of ~,~'-thienoin [ii]. *See [i] for communication 1 of this series. N. D. Zelinskii Institute of Organic Chemistry, Academy of Sciences of the USSR, Moscow. Translated from Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya, No. 4, pp. 834838, April, 1981. Original article submitted June 3, 1980. 0568-5230/81/3004-0615507.50

9 1981 Plenum Publishing Corporation

615

TABLE i. Dependence of Electrochemical Reduction of aThiophene Aldehyde to 1,2-Di(a-thienyl)-l,2-ethanediol on Composition of Reaction Medium Expt.

Supporting electrolyte

Solvent

Yield of (9) based on (I) reacted,

% t

2 3 4 5 6 7 8 9 t0 tl t2 t3 I4 15 16 t7 t8 t9 20

DMF

90~ DMF+ 10% H20 DMF MeCN

90% MeCN + 10% HzO

Ms

0,t M TEAP 0,1 M YBAP 0,t M TBAP + 0,025 M LiCIQ 0,i M TBAP + 0,05 M LiC1Q 0,t M TBAP + 0,1 M LiC1Q 0,t'MLiCIO~ 0,t M LiNOa 0,f M LiNOa 0d M TEAP 0,1 M TBAP 0,t M LiNOa + 0,3 M AcOH 0,t M TEAP 0,1 M TBAP 0,1 M TEAP + 0,05 M LiCIQ 0,t 71// TBAP + 0,05 M LiC10~ 0,t M TBAP + 0,t M LiCIO~ 0,t M LiCIO,, 0,t M TEAP 0,1 M TBAP 0,t M TEAP + 0,2 M thiophenol

t5 7 45 57 80 64 60 68 36 37 50 t5 17 35 53 59 64 48 54 27

EXPERIMENTAL Electrolysis was performed in MeCN which had been distilled over PaOs and in a~hydrous DMF which had been contacted with calcined K2COa for 1 day and then stored over a 4 A molecular sieve for 3 days before being distilled under vacuum. Preparative electrolysis was carried out using a P-5848 potentiostat and a 25-mi cell with a ceramic diaphragm; the cathode was a pool of mercury on the bottom of the cell (area ~30 cm=), the anode was a Pt gauze, and the reference electrode was a saturated calomel electrode. Reduction of 0.56 g of the aldehyde (I) was effected using a catholyte comprising a 0.i M solution of a lithium or tetraalkylammonium salt in DMF for MeCN containing various additives (Table I). The water content of the electrolysis solution was ~1% (except in experiments 8, 9, I0, 18, and 19). Electrolysis was effected for 3-4 h at --1.7 V (initial current 100-130 mA) until the current dropped to 15-20 mA. The variation in the concentration of (I) in the course of electrolysis was monitored by GLC using an LKhM-SMD-5 instrument with a 3 m• mm column of 5% XE-60 on Chromaton N-AW-DMCS. On completion of electrolysis, the catholyte was treated with i0 ml of Ac20 and 5 ml of pyridine, separated from the mercury, and distilled under vacuum after i0 h to remove the excess Ac20 and pyridine and most of the DMF. The residue was treated with water and extracted with ether. The extract was washed with saturated aqueous NaHCOa and water, and dried with anhydrous MgSO~. After removing the ether, the residue, comprising a mixture of diastereomeric 1,2-di(a-thienyl)-l,2-ethanediol diacetates,* was analyzed by GLC on a 1 m• mm column of 3% OV-17 on Chromosorb W using oterphenyl as standard. The overall yields of the diastereomers of (II) were calculated on the basis of the amount of (I) reacted. DISCUSSION OF RESULTS Our experiments on the electrochemical reduction of (I) in DMF and MeCN containing tetrabutyl- and tetraethylammonium perchlorate (TBAP and TEAP) show that the reaction is accompanied by considerable resin formation and that the yield of (If), identified in diacetate form, is no more than 17% (see Table i, experiments I, 2, 12, and 13), which is not in accord with recommendations regarding the optimum conditions for electrochemical hydrodimerization of carbonyl compounds [5]. The addition of lithium salts proved to have an unexpected effect on the yield of dimeric product. Even relatively small amounts of LiCIO~ result (see Table i, experiments 3 and 4) in a sharp increase in the yield of (II), the maximum yield in DMF solutions being observed iN the presence of equal amounts of LiCIO~ and TBAP (experiment 5), when the *The physicochemical constants and spectra of the diastereomeric diacetates of pinacol (II) will be given in a subsequent communication.

616

formation of resinous products is also reduced appreciably. The presence of the lithium salt in MeCN solutions has a similar but less pronounced effect (see Table I, experiments 14, 15, and 16). It is to be noted particularly that the use of lithium salts as inert electrolytes (experiments 6, 7, and 17) results in an appreciable increase in the yield of (II) compared with the runs using only the tetraalkylammonium salts as supporting electrolytes (experiments i, 2, 12, and 13). It is known [12] that the addition of proton donors (in particular, water or phenols) to nonaqueous solvents may result in protonation of the radical anions formed in the first stage of the reduction of a carbonyl compound, thereby forming free radicals capable of undergoing further reduction to the corresponding alcohol. However~ protonation of the radical anion of (I) does not take place even when the DMF solution contains ~50% water [13]. In view of this, it would be of interest to study the effect of water, and also of thiophenol and AcOH, on the electrochemical hydrodimerization of (I). The absence of thienyl alcohol in the products obtained by electrolyzing solutions containing additives of this type (see Table i) provides further confirmation of the fact that the radical anions of (I) are not protonated by weak proton donors during the electrode process. Electrolysis using the quaternary ammonium salts as supporting electrolytes in the presence of 10% water gives (II) in yields of up to 37% in DMF (see Table !, experiments 9 and i0) and up to 54% in MeCN (experiments 18 and 19), the effect being similar in principle to that of Li§ An analogous picture is observed when 10% water is added to a solution of LiN03 in DMP (experiment 8), but the yield of (II) is somewhat reduced if 0.3 mole of AcOH is present in the electrolysis solution containing the lithium salt (experiment ii). Electrolysis in a MeCN solution (0.i M TEAP) containing thiophenol results in a certain increase in the yield of (II) (experiments 12 and 20) and in the formation of other unidentified products, the thiophenol content decreasing in the course of electrolysis. Diphenyl disulfide was detected in the electrolysis products by GLC, its yield being up to 40% based on thiopheno! reacted. The formation of this disulfide under reducing conditions provides unequivocal evidence that radical processes are taking place. In the boundary-orbital approximation, it can be assumed that the reactivity of a radical anion having an electron in the lowest vacant orbital will be largely dete~nnined by the density distribution of just this unpaired electrono It would be logical to suppose, therefore, that the variations in pinacol yield occurring when the composition of the solution is varied are due to the redistribution of electron density in the radical anion. According to ESR data [14-16] obtained under conditions which exclude formation of ion pairs or hydration of the radical anion (e.g., in DMF or DMSO containing tetraalkylammonium salts), the spin-density distribution must correspond to that predicted by quantum-chemical calculations. In the case of aromatic and unsaturated compounds with electron-acceptor substituents (such as C=O and NO2), this means increased electron density in the aromatic ring or the conjugated double bonds and decreased electron density in the polar group [1416]. This kind of electron-density distribution in the radical anions of aromatic carbonyl compounds will assist reactions involving the C atoms of the aromatic ring. It is obviously this kind of reaction which results in resin formation during the electrochemical reduction of aromatic carbonyl compounds in nonaqueous aprotic organic solvents. The addition of water, which has considerably greater solvating power than DMF with respect to anions, to a DMF solution will increase the electron density in the car~onyl group of the radical anions [14]. The formation of ion pairs with metal cations will have a similar effect [15]. Consequently, increasing the concentration of water and Li + during the electrolysis of (I) in DMF or MeCN will increase the yield of (II) and decrease the yield of resinous products. It should be borne in mind that the formation of ion pairs between the radical anions and Li+ may also accelerate the dimerization reaction by reducing the effective negative charge on the reactant species and thus reducing the electrostatic repulsion between them. However, surface reactions involving the ion pairs [17] or disproportionation of triplet associates [18], probably resulting in the formation of resinous products, prevents the yield of (II) from exceeding 65% in Li salt solutions. The presence of sufficient TBAP in such solutions (see Table I, experiment 5) substantially increases the yield of (II), evidently due to displacement of ion pairs or electrochemical reaction products from the electrode surface by surface-active tetrabutylammonium cations.

617

Thus, in contrast to existing views on the electrochemical hydrodimerization of carbonyl compounds in nonaqueous media [5], the above data suggest that the extent to which the reactions of the electrochemically generated radical anions of these compounds are directed towards dimerization depends not so much on the proton-donor properties of the medium as on the redistribution of electron density in the boundary orbital of the anion radicals due to the formation of solvates, ion pairs, or complexes with components of the medium. The authors wish to thank A. M. Moiseenkov for discussing the results of the present work. CONCLUSIONS I. Conditions have been found under which 1,2'di(a-thienyl)-l,2-ethanediol pared electrochemically from a-thiophene aldehyde in yields up to 80%.

can be pre-

2. The reactions of the electrochemically generated radical anions of a-thiophene aldehyde are directed by solvation and by the formation of ion pairs with components of the electrolytic solution. LITERATURE CITED i. 2. 3. 4. 5. 6. 7. 8. 9. i0. ii. 12. 13.

V. P. Gul'tyai and A. M. Moiseenkov, Zh. Org. Khim., 16, 1026 (1980). A. P. Tomilov, M. Ya. Fioshin, and V. A. Smirnov, Electrochemical Synthesis of Organic Substances [in Russian], Khimiya, Moscow (1976), pp. 205-209. J. H. Stocker and R. M. Jenevein, J. Org. Chem., 33, 294 (1968). J . H . Stocker and R. M. Jenevein, Coll. Czech. Chem. Commun., 36, 925 (1971). W. van Tilborg and C. J. Smit, Tetrahedron Lett., 3651 (1977). E. M. Abbot, A. J. Bellamy, and J. Kerr, Chem. Ind., 828 (1974). C. Caullet, J. M. Bessin, and J. C. Bodard, Compt. Rend., 261, 1848 (1965). C. Caullet, M. Salaun, and M. Hebert, Compt. Rend., 264C, 228 (1967). E. V. Kryukova and A. P. Tomilov, Elektrokhimiya, ~, 869 (1968). P. Foulatier, J. P. Sala~n, and C. Caullet, Compt. Rend., 279C, 779 (1974). J. Deschamps, W. King, and F. J. Nord, J. Org. Chem., 14, 184 (1949). C. K. Mann and K. K. Barnes (editors), Electrochemical Reactions in Nonaqueous Systems, Marcel Dekker (1970). V. P. Gul'tyai, S. G. Mairanovskii, T. Ya. Rubinskii, and N. P. Rodionov, Elektrokhimiya,

16, 370 (1980). 14. 15. 16.

J. GendeZl, J. H. Freed, and G. K. Frankel, J. Chem. Phys., 37, 2832 (1962). N. Hirota, J. Chem. Phys., 37, 1884 (1962). T. M. McKinney, in: Electroanalytical Chemistry (ed. by A. Bard), Vol. i0, New York

17. 18.

A. Bewick and D. J. Brown, J. Chem. Soc. Perkin Trans. 2, 99 (1977). A. Lasia, J. Electroanal. Chem., 102, 117 (1979).

(1977).

618