7 Electro-organic Chemistry By J. H. P. UTLEY Chemistry Department Queen Mary College Mile End Road London E. ? THISreview is organised as in previous years i.e. anodic and cathodic processes are categorised according to the species believed to be discharged. The bulk of the material is concerned with preparative scale processes although some electro-analytical work which has a direct bearing on mechanistic interpretation is included. 1 Anodic Processes Anodic Reactions of Organic Anions.-Anodic Oxidation of Curboxylates to Radicals (Kobe reaction). A novel hybrid technique polaromicrotribometry has been used to investigate the electrode conditions necessary for Kolbe dimerisation.' In this method a measure of the friction between a glass plate and the platinum electrode is monitored during the construction of a Tafel plot (log current density us.anode potential). For an aqueous solution of acetate ions the coefficient of friction changes markedly at the well-known transition point where oxygen evolution is suppressed and ethane formation begins. At this point the appearance of a film on the electrode was noted and by using electrodes of large surface area sufficient of the pale yellow film was collected for analysis. It is insoluble in water bases and acids (except H2S04)but is soluble in organic solvents. It decomposes at 170°C has the composition 78.2%C 11 %H,and 10x0,and i.r. spectroscopy suggests that it is a polymer containing the unit (CH,) with n > 50. It is tempting to speculate that in aqueous solution such films are necessary for the Kolbe reaction although the Kolbe reaction is often more efficiently performed in organic solution.Oxidation of carboxylates at graphite anodes usually results in carbonium ion formation (Ann. Reports (B),1968 p. 238) but for long-chain alkanoic acids ( >C,) in acetonitrile one-electron oxidation with dimerisation predominates. This is interpreted in terms of the familiar stacking of alkyl chains favouring dimerisation and inhibiting a second electron transfer by slowing diffusion of the radical to the electrode surface after decarboxylation (Scheme 1). A review of the use of the Kolbe reaction for obtaining bifunctional compounds contains useful tables of results and a description of an ingenious cylindrical J.E. Dubois and F. Bruno Compt. rend. 1970 271C 791. D. L. Muck and E. R. Wilson J. Electrochem. Sac. 1970 117 1358. 220 J. H. P. Utley II 02CCH ,CH ,R C CH,CH2R II -+ O,CCH,CH,R 0 A Scheme 1 divided cell.3 The contents are stirred by vibration of the electrodes at 100Hz and the cell is capable of operation at 25 A dm-, thereby allowing large scale reactions to be performed. Good yields of bifunctional compounds whose chemical preparation is difficult have been obtained4 by electrolysis of yy-dimethylalkanoic acids (1) X C(Me),CH,CH,CO,H + XC(Me),(CH,),C(Me),X (1) X = CH,OH CHO C02H CN or NO2 Attempts at cyclisation via Kolbe oxidation are summarised within a general review of electrochemical methods for cyclisation.’ On mechanistic aspects of anodic reactions however the review is badly out of date.Many of the cyclisation reactions attributed to the generation of carboxyl radicals almost certainly proceed by discharge of aromatic portions of the molecules with radical cations thus formed being attacked intramolecularly by carboxylate anions. Anodic Oxidation of Carboxylates to Carbonium Ions. The products of anodic oxidation at graphite of a series of cycloalkane carboxylates have been charac- terised. For alkene and alcohol formation the intermediacy of carbonium ions is supported by the observation6 of mobility of a deuterium label originally at C-1. The extent of deuterium rearrangement is dependent on ring size being at a maximum for electrolysis of [,H ,]cycle-octane carboxylic acid.Thermal decomposition of the corresponding t-butyl peroxyester gives cyclo- octane with no detectable rearrangement of the deuterium label. The major product of anodic oxidation’ of 4-cyano-2,2-dimethylbutanoicacid is NCCH,CH,C(Me):CH (16%) with little of the Saytzeff product NCCH,CH,C CMe (2%). The scheme proposed for anodic acetamidation (Ann. Reports (B),1969,p. 220) has been confirmed and extended by an elegant series of experiments.* Electro- J. Haufe and F. Beck Chem.-Ing.-Tech. 1970 42 170. * G. Cauquis and B. Haemmerle Bull. SOC. chim. France 1970 183. J. D. Anderson J. P. Petrovich and M. M. Baizer Adv. Org. Chem. 1969 6 257. J. G. Traynham E. E. Green and R. L. Frye J. Org. Chem. 1970,35 361 1.’ G. Cauquis and B. Haemmerle Bull. SOC.chim. France 1970 190. J. M. Kornprobst A. Laurent and E. Laurent-Dieuzeide Bull. SOC.chim. France 1970 1490. Electro-organicChemistry 22 1 lysis of 1,l-dimethylbutanoic acid in acetonitrile yields in addition to amides 10% of the acid anhydride. This is rationalised according to Scheme 2. An alternative reaction of intermediates of type (2) is rearrangement to an N-acylamide (3) which may cleave to give amides (4) and (5). The importance of -2e MeCN -+ RCO 5 RCO2-R+ 4R-N=C-CH * R-N=C-CH, -co RNHCOCH + (RC0)ZO R = EtC(Me),-Scheme 2 this route for tertiary acids is confirmed by the results of electrolysis of pivalic acid in the presence of CH3I4CO2H(Scheme 3). The specific radioactivity of the N-t-butylacetamide suggests that 80% of the product is formed via the N-acy lamide.j'?0 ../ COCH RNHCOCH R-N=C-CH 4R-N -c (4) I '14c OR1 RNH 4COR1 R'-l4C (5) \O (3) R = Bu' R' = CH Scheme 3 a-Ketocarboxylic acids are oxidised' to the corresponding acylium ions upon electrolysis although in methanolic solution pyruvic acid affords a small amount of biacetyl the Kolbe product. Anodic bisdecarboxylation has pro- vided" a good route to bicyclo[3,2,2]nona-6,8-diene(6),previously obtainable with some difficulty. ' B. Wladislaw and J. P. Zimmermann J. Chem. SOC.(B) 1970 290. lo A. J. Baker A. M. Chalmers W. W. Flood D. D. MacNicol A. B. Penrose and R. A. Raphael Chem. Comm. 1970 166. J. H. P.Utley Anodic Reactions of Other Organic Anions.An authoritative review of oxidative additions of anions to alkenes has appeared." Anions which may be oxidised anodically to radicals with subsequent addition to alkenes include those of malonic esters aliphatic nitro-compounds and Grignard reagents. Organic azides also may be prepared by oxidative addition of the azide anion. The experimental work connected with anodic oxidation of Grignard reagents is included in an accountI2 of the factors influencing addition dimerisation and polymerisation (Scheme 4)at copper platinum and carbon anodes. In the absence of olefinic trapping agents oxidation' at high current density of di- methylsodiomalonate ethyl sodiomethylacetoacetate and sodiodimedone gives low yields of a confusing variety of products.R-R 6-df ? RMgBr 5R' -+ RH + RH(-H,) RCH,CH(Y)CH(Y)CH,R +- RCH,CHY --+ RCH,CH(Y)CH,CHY1 further oligomerisation Scheme 4 Anodic Reactions of Neutral Organic Molecules.-The report of a two-electron transfer from tetra-p-anisylethylene (TAE) has been qualified (cf Ann. Reports (B) 1969 p. 227). A careful study14 at the rotating disc electrode and cyclic voltammetry in both acetonitrile and methylene chloride solution shows that formation of the dication of TAE is the result of two closely spaced one-electron transfers In methylene chloride solution the separation is visible on the cyclic voltammogram and coulometry supports one-electron transfer at 0.95 V (us. S.C.E.).the second transfer being at 1.2 V. An e.s.r.signal was obtained from the solution electrolysed at 0.95 V. The corresponding dimethylamino-compound (7) is apparently oxidised with a completely reversible two-electron (p-Me N.C6 H,),C=C(C6 H,.NMe,-p) (7) transfer with no detectable separation of the steps. The anodic oxidation of the series of phenyl-substituted ethylenes from PhCH :CH to PhzC:CPh has also been investigated' by cyclic voltammetry. At preparative concentra- tions olefins can be dimerised" anodically (e.g. Scheme 5). Alternatively a ' I H. Schafer Chem.-Ing.- Tech. 1970,42 164. H. Schafer and H. Kiintzel Tetrahedron Letters 1970 3333. '' (a) R. Brettle and D. Seddon J. Chem. Soc. (a, 1970 1153; (h) R. Brettle J. G. Parkin and D. Seddon J. Chem. SOC. (C),1970 1317; (c) R. Brettle and D.Seddon J. Chem. SOC.(C) 1970,2175. l4 A. J. Bard and J. Phelps J. Electroanalyt. Chem. Interfacial Electrochem. 1970 25 2A. I5 B. L. Funt and D. G. Gray J. Electrochem. Soc. 1970 117 1020. Electro-orgunic Chemistry graphireanode 40 % current yield \ MeOH-NaCIO \ / OMe Scheme 5 mixture of two olefins may be oxidised16 at controlled potential and the mixed coupled product obtained (Scheme 6). grdphite anode PhCH :CH + CH :CHOEt M~OH-N~CIO,I PhCH(OMe)CH,CH,C(OMe)OEt 30 "/ current yield Scheme 6 The electrochemical reactions of mercury-olefin complexes have attracted attention. For cyclohexene in acetonitrile the Hg" complex may be formed directly" by using a mercury anode which is self-oxidised to Hgz+.Further oxidation of the complex and reaction between the cation formed and solvent leads to the organo-mercury compound (8) [Scheme 71. The cathodic reactions of (8) are also of some interest (see p. 238). Fleischmann and his co-workers Scheme 7 (8) have investigated the anodic reactions of a number of Hg"-olefin complexes. At high current density efficient oxidation to mixtures of carboxylic acids is achieved. The products are explicable in terms of Scheme 9 noteworthy features EtCH,CHO anode + EtCH,CO,H (33% current yield) f + H$ + EtCH(OH)CH + Et(OH)CHCH,Hg+ H,O. EtCH:CH + Hg2+ L EtCOCH EtC0,H (8 %) + CH,CO,H (44%)+ HC02H(8 %) Scheme 8 of which are the catalytic role of Hg2+ and the suppression of Kolbe oxidation of the acids produced even at +2.4 V (us.Ag/Ag+). In an attempt" to control products by using potential control to select the major electrode reaction mixtures of cyclohexene and chloride ion were oxidised at a series of potentials. The CI-/CI,reaction occurs in acetonitrile at +0.75 V (Ag/Ag+) whereas '' H. Schafer and E. Steckhan Tetrahedron Letters 1970 3835. l7 N. L. Weinberg Tetrahedron Letters 1970 4823. M. Fleischmann D. Pletcher and G. M. Race J. Chern. Snc. (B) 1970 1746. '')G. Faita M. Fleischmann and D. Pletcher J. Electroanalyt. Chern. Interjacial Electro- chern. 1970 25 455. 224 J. H. P. UtIey direct oxidation of cyclohexene demands + 2.05 V. For equimolar mixtures of chloride and cyclohexene electrolysis in acetonitrile at 0.7 V gave with 80 % current efficiency N-(2-chlorocyclohexyl)acetamide,which is also a major pro- duct of the chemical reaction between chlorine and cyclohexene in acetonitrile.At + 2.1 V the only product identified was 3-chlorocyclohexene presumably a result of the deprotonation of the cyclohexene radical ion followed by re- oxidation to the corresponding allylic carbonium ion (cf:Ann. Reports (B),1968 p. 236). Although the products are of trivial importance the results are another encouraging sign that the electrode surface conditions can be manipulated and combined with potential control to select between competing processes. An interesting consequence of a careful cyclic voitammetric study” of the anodic oxidation of 9,lO-diphenylanthracene (DPA) in benzonitrile is the observation of a reduction peak for the dication (9).The smallness of the + Nu = unspecifiednucleophile DPA2+ + Nu -+ \ / Ph (9) Ph Ph Scheme 9 corresponding oxidation peak on the return scan suggests that K in Scheme 9 is very large. Similar studies,” combined with e.s.r. spectroscopy suggest that radical-cations of aromatic hydrocarbons are more stable in methylene chloride solution than in acetonitrile. The stability of 9,lO-disubstituted anthra- cenes is not a simple function of the charge delocalising ability of the substituents. For instance the 9,1O-di-(a-naphthyl)anthraceneradical-cation is less stable than DPA presumably because the naphthyl groups are the more reactive at the para-positions. The anodic oxidation in non-aqueous solution of benzo[a]pyrene gives a mixture of benzo[a]pyrene quinones.Despite the complexity a consistent account” of the major features of the reaction has been constructed. There is compelling evidence for direct discharge of aromatic hydrocarbons prior to nuclear acetoxylation (Ann. Reports (B) 1968 p. 245). Side-chain acetoxylation however is possible by two routes depending on whether the aromatic species or an inorganic anion is discharged (Scheme 10). ’O R. Dietz and B. E. Larcambe J. Chem. SOC. (B) 1970 1369. 21 L. S. Marcoux A. Lornax,and A. J. Bard J. Amer. Chem. SOC.,1970,92 243. 22 L. Jeftic and R. N. Adams J. Amer. Chem. SOC.,1970,92 1332. Electro-0rganic Chemistry x-X’ ArCH,’ [ArCH,] Scheme 10 The difficulty in distinguishing between these processes is ill~strated~~ by the electrochemical oxidation of toluene in acetic acid with potassium acetate tetramet hylammonium nitrate and tetramethylammonium toluene-p-sulphonate as electrolytes.Voltammetric curves show that the relative ease of oxidation of the anions is NO > -0Ac > -0Ts with the tosylate ion oxidising at potentials higher than the discharge potential of toluene. However for each of the anion oxidations added toluene lowers the current suggesting mere adsorption of toluene on the electrode even at high anodic potentials. Addition of the more easily oxidised naphthalene and anthracene causes a large increase in current. Thus the balance of evidence is in favour of route (a)being the major pathway although considerable doubt remains about the relative importance of the two processes.Related on mesitylene and durene compounds the ambiguity as in mixtures hydrocarbons and nitrate ions are oxidised at similar potentials and a current drop is again observed on addition of hydrocarbon at a given potential. In methylene chloride solution the strongest nucleophiles are the electron-rich hydrocarbons themselves. With tetra-n-butylammonium tetra-fluoroborate as electrolyte durene and mesitylene give2 $ upon electro-oxidation the coupled products (10) and (1l),presumably by either of the routes drawn in Scheme 11. coy [H3CArArCH3Izf [ArCH,] f H3CArArCH3 (1 1). A r b [H,CArArCH,] :2$(from mesitylene) ArCH _&*+ ArCH, [ArCHz+] -H+ P ArCH,ArCH (from durene) (10) Scheme 11 A full account26 is available of Eberson’s mechanistic work on anodic aceta- midation of alkylaromatic compounds (cf:Ann.Reports (B) 1969 p. 225). The general conclusion is that most of the results are explained by Scheme 12. In these reactions the yield of acetamidation from hexamethylbenzene is less 23 S. D. Ross M. Finkelstein and R.C. Petersen J. Org. Chem. 1970 35 781. 24 K. Nyberg Acta Chem. Scand. 1970 24 473. 25 K. Nyberg Acta Chem. Scand. 1970 24 1609. 26 L. Eberson and B. Olofsson Acta Chem. Scand. 1969 23 2355. J. H. P.UtIey sensitive to added water than is the yield from toluene. This is interpreted as meaning that the more stable hexamethylbenzene cation is less selective than -' ArCH *[ArCH,]? -H+ ArCH,' /fArCHzCHzAr ArCH,' ArCH,NHCOMe Scheme 12 the benzyl cation thus upsetting the usual relationship between stability reac- tivity and selectivity.Oxidation at +2.0 V (us. S.C.E.)of a mixture of diphenyl- acetylene and sodium cyanide in methanol leads2' to a good and exclusive yield of the 4-cyano-derivative. Breslow and co-workers continue to use electrochemical data to estimate the magnitude of anti-aromatic destabilisation (cf:Ann. Reports (B) 1969 p. 233). They have obtained28 polarographic data for oxidation of the compounds (12) and (13) and of 1,4-naphthohydroquinone (14) [Scheme 131. As compound (13) is oxidised more easily than naphthohydroquinone it is unlikely that strain OH 0 (12) OH 0 OH OH WMe Me @ OH OH (14) Scheme 13 differences involving the bond common to the four- and six-membered rings is reflected in large changes in half-wave potential.This leaves the difference (0.27 V) between the ease of oxidation of (12;R = Ph) and naphthohydroquinone to be explained as an anti-aromatic destabilisation of the cyclobutadiene entity of at least 12 kcal mol- (0.27 x 2 eV). Results from the use of a rotating disc electrode lead29 to the conclusion that hydroquinone is oxidised in acetonitrile by a two-electron process although the method would not distinguish between an EEC and ECE process ifdeprotonation of an intermediate were rapid. An alternative view of this reaction has been 27 K. Yoshida and T. Fueno Chem. Comm. 1970 71 1.28 R. Breslow R. Grubbs and Shun-Ichi Murahashi J. Amer. Chem. Sor. 1970 92 4139. 29 V. D. Parker and L. Eberson Chem. Comm. 1970 1289. Electro-organic Chemistry 227 presented3* and the general problems of hydroquinone electro-oxidation in aprotic solvents ~urveyed.~ Phenols may be coupled anodically and further examples of the application of this technique to phenolic tetrahydroisoquinolines have been described.32 Oxidation of t-butylhydroperoxide at a carbon anode in the presence of tetracyclone leads33 to a lactone which is one of the isomers of (14a). The compound is obtained by electrolysis of a two-layer solution with Ph ( 144 only the peroxide in contact with the electrodes. Oxidation with ceric ammonium sulphate gives the same result and an analogy with the Baeyer-Villiger reaction is drawn.The e.s.r. spectrum of the di-p-anisylamine radical-cation obtained34 by electrolysis of the amine in both neutral (CH3N02) and acidic solution (CF3C02H-CH3N02).is identical with that produced by dissolution of the di-p-anisylnitroxyl radical in the acid mixture. The electrolysis product is further characterised by coulometry and U.V. spectroscopy and it is now doubtful that it was the protonated nitroxide which was prepared in previous experiments (Ann. Reports (B) 1967 p. 317). The stable solid ionic free radical (15) is as a blue-black anodic deposit from the electrolysis of o-tolidine in methylene chloride solution. L (15) The reaction schemes encountered in the electro-oxidation of aromatic amines are now well characterised and during the past year several thorough st~dies~~-~' have been described.In the absence of substituents para to the amino function benzidines are often the major product. However for NN-dimethylaniline 3" B. R. Eggins and J. Q. Chambers J. Electrochem. SOC. 1970 117 186. 3' J. Bessard G. Cauquis and D. Serve Tetrahedron Letters 1970 3103. 32 J. M. Bobbitt K. H. Weisgraber A. S. Steinfeld and S. G. Weiss J. Org. Chem. 1970 35 2884. 33 N. L. Weinberg Canad. J. Chem. 1970 48 1533. 34 G. Cauquis and D. Serve Tetrahedron Letters 1970 17. 35 H. N. Blount and T. Kuwana J. Amer. Chem. SOC. 1970 92 5773. 36 R. Hand and R. F. Nelson J. Electrochem. SOC.,1970 117 1353. " A. G. Hudson A. E.Pedler and J. C. Tatlow Tetrahedron 1970 26 3791. 38 M. Melicharek and R. F. Nelson J. Electroanalyt. Chem. Interfacial Electrochem. 1970 26 201. 39 D. W. Leedy and R. N. Adams J. Amer. Chem. SOC.,1970,92 1646. 228 J. H. P.Utley (ca. lo-' moll-') in acetonitrile or benzonitrile a significant product is (16) with the origin of the extra methylene group being something of a mystery.36 In aqueous acetone at + 1.5 V (us. S.C.E.) pentafluoroaniline is coupled3' at a platinum anode to give decafluoroazobenzene (18 %). Octafluorophenazine (17) is formed as a minor product possibly via formation and oxidation of 2-amino- nonafluorodiphen ylamine. Separate anodic oxidation of the diphenylamine (17) provides an efficient (40%) route to (17). Macroscale anodic oxidation of NN-dimethyl-p-toluidine gives3 mainly the coupled product arising from Scheme 14.+' p-Me,N.C,H,-CH p-Me,NC,H,-CH %p-Me,N-C,H,-CH,' (p-Me NC H,-CH,-) 4 Scheme 14 The scheme is supported by e.s.r. evidence for the radical-cation and coulometric and cyclic voltammetric data for the electrochemical reaction. Dimethylaminoalkenes very readily undergo electrochemical oxidation e.g.(18) is oxidised4' at -0.9 V (us. S.C.E.). Aliphatic tertiary and secondary amines are electrolysed in acetonitrile at ca. 1-OV (Ag/Ag+) resulting41 in smooth de-alkylation. For instance di-n-propylamine is converted into n-propylamine (66%). Controlled potential electrolysis [+ 1-5V (Ag/Ag+)] of NN-diphenylhydrazine in acetonitrile in the presence of perchloric acid yields42 a solution believed to (Me,N),C C(Me)C(Me) :C(NMe,) [Ph,N:NH]+ (18) (19) +-Ph2NNH2+ 2py -2e 2pyH' + Ph,N=N +-2Ph2N-N Ph,N-N=N-NPh, ____+ Scheme 15 40 J.M. Fritsch H. Weingarten and J. D. Wilson J. Amer. Chem. SOC.,1970 92 4038. 41 L. C. Portis V. V. Bhat and C. K. Mann J. Org. Chern. 1970,35 2175. 42 G. Cauquis and M. Genies Tetrahedron Letters 1970 2903. Electro-organic Chemistry 229 contain the diphenyldiazenium cation (19). In the presence of pyridine electro- lysis at 0.3 V yields tetraphenyltetrazine presumed to be formed according to Scheme 15. Cinnoline derivatives may be by two successive oxida- tions of a mixture of NN-diphenylhydrazine and styrene. The possibilities of this method are illustrated by Scheme 16.Ph I -2e Ph,NNH 7[Ph,N=NH]+ -Ph -2e A2H Ph Ph Qf4p-Q I Ph Ph Scheme 16 In practice the cations are obtained as perchlorates and sodium bicarbonate is used to ensure the irreversibility of deprotonation steps. Using 3,6-di-isobutylpiperazine-2,5-dione(20) as a model for dipeptides it has been found44 that 1,3-~ycloaddition of acetonitrile may be induced anodically albeit with low efficiency. Incorporation of CD3CN proves that the solvent is indeed involved the overall reaction being given by Scheme 17. Scheme 17 Alkyl halides may be anodically oxidised to carbonium ions (Ann. Reports (B) 1969 p. 220) and for alkyl iodides in particular several examples of the reaction have been in~estigated.~' In acetonitrile solution the carbonium ions are pro- duced and trapped efficiently to give the range of acetamides familiar from carboxylate oxidations (see p.221) and for cyclohexyl iodide additional pro- duction of an allylic acetamide. Anodic oxidation of cyclohexene in acetonitrile 43 G. Cauquis and M. Genies Tetrahedron Letters 1970 3403. 44 L. A. Simonson and C. K. Mann Tetrahedron Letters 1970 3303. " A. Laurent and R. Tardivel Compt. rend. 1970 271C 324. 230 J. H. P. Utley gives proportions of the saturated (80x)and unsaturated (20x)acetamide similar to those obtained by electrolysis of cyclohexyl iodide thus supporting Scheme 18. 1-. 1MeCN-H,O 0MecN-H,o GOMe MeCONH _II+ Scheme 18 2-Methylthiophen has been oxidised under preparative conditions to give46 in acidic methanol solution ring-opened carbonyl compounds similar to those obtained from furan.At a platinum anode at + 1.0V (us. S.C.E.)or-mercapto-phenylacetic acid oxidises not at the carboxylate group but at sulphur to yield4' the disulphide (21). In methylene chloride with aluminium trichloride electrolyte [PhCH(C02H)S-f2 (21) an aluminium anode is consumed at the rate of 1 g atom per Faraday with the ultimate formation48 of methylenebis(a1uminium dichloride) (22) in Scheme 19. In these insertion reactions an analogy is drawn between the 'AlCl' species and carbene. c1-=% C1' Cl' + A1 -[AICI] [AICI] .tCH,CI -+ CI,AICH,CI [AICI] + CI,AICH2Cl -+ CI,AlCH,AICI (22) Scheme 19 46 J. Sprogl M.Janda and M. Valentova Coll. Czech. Chem. Comm. 1970 35 148. 47 J. Artaud M. Estienne and P. Courbis Compt. rend. 1970 271C 125. 48 E. H. Mottus and M. R. Ort J. Electrochem. Sac. 1970 117 385. Electro-organic Chemistry 231 2 Cathodic Processes Cathodic Reactions of Organic Cations.-Phenyl-substituted cyclopropenyl radicals have been generated4' electrochemically from the corresponding cations. For this system further reduction to the anti-aromatic cyclopropenyl anion is difficult. Some of the key experiments are outlined in Scheme 20. Ph Ph . Ph . Ph' Ph ' Ph Ph ph%Ph H + phm 1% Ph Ph Ph 61 % Ph F H -"@D-. -+ Ph Ph Ph\ Ph 40x Scheme 20 The reviewer's suggestion (Ann.Reports (B),1968,p. 249)for the route by which toluene is formed in the cathodic reduction of the benzyl trimethylammonium ion has been substantiated.Reduction in NN-dimethylformamide solution saturated with carbon dioxide results5' in the formation of phenylacetic acid at the expense of toluene. This is good evidence for the intermediacy of the benzyl anion. A systematic investigation of the electrochemical reduction of quaternary ammonium compounds of the type RNMe,' in hexamethyl-phosphoramide (HMPA) has shown that the competitive rupture of N-C (yR' + NMe R-AMe -%R-NMe3 Ah CH,' + R-NMe (a) R = allyl t-butyl (a)and (h) R = phenyl n-butyl Scheme 21 bonds proceeds5' as in Scheme 21. One unexpected result is the formation of a substantial amount of (23) by electrolysis of PhCH,NMe,+ in HMPA.Catho- 49 T. Shono T. Toda and R. Oda Tetrahedron Letters 1970 369. 50 S. D. Ross M. Finkelstein and R. C. Petersen J. Amer. Chem. SOC., 1970 92 6003. 5' J. E. Dubois A. Monvernay and P. C. Lacaze Electrochim. Acta 1970 15 315. J. H. P.Utley (23) dic cleavage of (1 -propenyI)triphenylphosphonium bromide is very efficient affording triphenylphosphine and pr~pene.’~ Cathodic Reactions of Neutral Organic Compounds.-A romatic Compounds and AIkenes. The use of electroanalytical techniques for the measurement of otherwise inaccessible pK values has been extended (cJ:Ann. Reports (B) 1969 p. 233). Using the data of Voltz and Lotsch (Ann. Reports (B) 1969 p. 231) for polarographic reduction of substituted triarylmethyl cations combined with Deno’s pK,+ values for cation formation from alcohols the pK values for the triarylmethanes have been calculated53 via the cycle shown in Scheme 22.ROH R‘ Scheme 22 The scope of the indirect method of reduction of aromatic systems is still being investigated. Isolated olefinic bonds are less easy to reduce by this method than aromatic species. For instance ally1 benzene can be selectively reduceds4 to 2,5-dihydroallylbenzene by lithium produced electrochemically in methylamine. cT2x CH,’ abstfaction \ dimerisation aldehyde Scheme 23 ’* L. Horner I. Ertel H. D. Ruprecht and 0. Belovsk Chem. Ber. 1970 103 1582. ’’ R. Breslow and W. Chu J. Amer. Chem. SOC.,1970 92 2165. ’‘ R. A. Benkeser and S.J. Mels J. Org.Chem. 1969 34 3970. Electro-organic Chemistry 233 Nitrobenzyl halides readily lose halogen upon electroreduction in aprotic sol- vents. The major products are the corresponding dinitrobibenzyls and nitro- toluenes and in the presence of oxygen aldehydes and esters are also formed.”*56 The results are readily rationalised by Scheme 23 where halide ion loss is rapid for ortho-and para-nitro-substitution but slower for the rneta-isomer. Electrochemical methods seem well suited for the reductions of heteroaro- matic compounds. Recent examples include the reductions of 2-phenylquinoxa-line and 2,3-diphenyl-5,6-dihydropyrazineto (24) and (25) respectively.” The H H H H (24) (25) cathodic reactions of substituted phthalazines demonstrates8 how a variety of reduction products can be obtained depending on cathode potential and pH (Scheme 24).2e high pH 1 Scheme 24 The cyclic voltammogram of 2-methoxyazocine obtained” in DMF solution shows a two-electron reduction peak. The dianion is reoxidised with some difficulty the difference between cathodic and anodic peaks being about one volt. This suggests that the lh-electron dianion (26)is impressively stable. 55 J. G. Lawless D. E. Bartak and M. D. Hawley J. Amer. Chem. SOC. 1969,91,7121. 56 P. Peterson A. K. Carpenter and R. F. Nelson J. Electroanalyt. Chem. Interfacial Electrochem. 1970. 27 1. 57 J. Pinson J. P. Launay and J. Armand Compt. rend. 1970 270C 1881. 58 H. Lund and E. Th. Jensen Acta Chem. Scand. 1970,24 1867. 59 L. A.Paquette J. F. Hansen T. Kakihana and L. B. Anderson Tetrahedron Letters 1970 533. J. H. P. Utley Activated unsaturated functions including olefins are smoothly reduced to radical-anions and anions with subsequent Michael-type coupling. Reductive coupling such as this is covered in a long and comprehensive review.60 Cross- coupling of activated olefins is dependent upon cathode potential in a surprising way the reactivity of the acceptor olefin being enhanced6’ at potentials below those required for discharge. The nature of the supporting electrolyte is also important quaternary ammonium salts favouring hydrodimerisation. Horner has studied62 this dependence by incorporating the reducible species into the ammonium ion in the salts [R,N+CH2CH2.0COCH:CHPh]X-,C1 or I.X = Hydrodimerisation is favoured for the iodides and salts containing long alkyl chains i.e. conditions encouraging ion-pairing in the double layer. In the preparation of 2,3,4,5-tetraphenyl- 1,6-hexanedioic acid (27) it was found63 that the major product from the attempted hydrodimerisation of cc-phenylcinnamo-nitrile was unexpectedly the cyclic compound (28) presumably formed according to Scheme 25. Ph Ph PhCH :C(CN)Ph /” / PhCH :C(CN)Ph 3PhCHCH(CN)Ph Ph e PhC N PhJA:;;Ph i;alkali ii; HNO Ph Ph Ph (27) Scheme 25 The indirect reduction of coumarin provides a good system for asymmetric induction. The conjugate acids of optically active alkaloids are discharged cathodically with production of the chiral radicals (29).Simultaneous discharge of for example 4-methylcoumarin leads to the radical (30) and hydrogen-atom transfer completes the reduction (Scheme 26). In this example the use of sparteine leads to formation64 of (31) with a 17% optical yield. 6o M. M. Baizer and J. P. Petrovich Progress in Physical Organic Chemistry eds. A. Streitweiser and R. W. Taft Interscience London-New York 1970 Vol. 7 p. 189. 61 M. M. Baizer J. P. Petrovich and D. A. Tyszec J. Electrochem. Soc. 1970 117 173. 62 L. Horner and H. D. Ruprecht Tetrahedron Letters 1970 2803. 63 S. Wawzonek A. R. Zigman and G. R. Hansen J. Electrochem. Soc. 1970 117 1351. 64 R. N. Gourley J. Grimshaw and P. G. Millar J. Chem. Soc. (0,1970 2318. Electro-organic Chemistry R1R2R3kH % R'RZR3NH (29) Me Me (30) (31) Scheme 26 Carbonyl Compounds.At a mercury cathode carbon dioxide is reduced to malate ion (Ann. Reports (B),1969 p. 235). Similar electrolysis using a lead elec- trode and quaternary ammonium electrolytes in aqueous solution enables carbon dioxide to be converted65 into glycollate ion. A very high current efficiency (ca. 600%) is associated with both of the processes. Horner reports66 another example of asymmetric induction through the use of chiral supporting electrolytes. Optical yields in the region 5-7% are obtained in the reduction outlined in Scheme 27 with the configuration of the preferred enantiomer depending on that of the chiral salt. Stereoselectivity has also been observed in the reducti01-1~~ of conformationally biased cyclic ketones under carefully controlled conditions.For dihydroisophorone and 4-t-butylcyclohexanone reduction in neutral solution proceeds efficiently at ca. -2.4 V (vs. S.C.E.)to approximately the equilibrium mixture of axial (cu. 10%) and equatorial (ca.90 %) alcohols. However in the presence of acetic acid electroreduction at lead may be achieved at ca. -1.7 V and the epimeric alcohols produced in similar amounts. The result is rationalised in terms of chemisorption of intermediates in certain favoured conformations with protonation of anions by coadsorbed acetic acid before equilibration is reached. PhC(Me):NCH,Ph :;+ P PhCH(Me)NHCH,Ph R*NMe X-R* = (R or S)-PhCH(Me);(R or S)-PhCH(OH)CH(Me) X = I or C1 Scheme 27 65 A.Bewick and G. P. Greener Tetrahedron Letters 1970 391. 66 L. Horner and D. H. Skaletz Tetrahedron Letters 1970 3679. '' J. P. Coleman R.J. Kobylecki and J. H. P. Utley Chem. Cornm. 1971 104. J.H. P.Utley In addition to the secondary alcohol and the meso-and dl-pinacols cathodic reduction68 of o-aminoacetophenone gives rise to both stereoisomers of com- pound (32). Carbonyl reduction with aromatisation to the corresponding tri- hydroxybenzenes is the result6' of electrolysis of humulones of the type (33). (32) (33) A simple indirect electrochemical procedure for the conversion of ketones into either alcohols or N-methylamines is described7' by Benkeser and Mels the reductant being lithium produced by electrolysis of lithium chloride in methylamine.Electrolysis with dropwise addition of the ketone results in alcohol formation whereas electrolysis after the ketone-methylamine-lithium chloride mixture has stood for several hours results in efficient formation of the corresponding N-methylamine. Amides have also been reduced7' by the indirect method and the products are the corresponding alcohols or aldehydes depending on the absence or presence respectively of a proton donor. Direct electroreduction of amides to aldehydes may be achieved for certain heterocyclic systems where in aqueous acid the aldehyde function is hydrated and protected from further reduction. For instance pyridine-2-carboxamide is reduced72 at a mercury cathode to the aldehyde (54 % isolated yield).Trifluoroacetic acid is smoothly reduced73 to l,l,l-trifluoroethane at a platinum cathode at <0.3V (on the hydrogen scale). The smoothness of this reduction contrasts with the irreducibility of acetic acid in this way and points not to a direct electroreduction but to hydrogenation with molecular hydrogen produced at the platinum electrode which can also act as a catalyst. Cathodic electron transfer to the carbonyl function may initiate cleavage reactions. An example of this type of reaction is provided74 by the electrochemical cleavage of carbon-oxygen and carbon-fluorine bonds outlined in Scheme 28. Evidence for the participation of the electroactive substituent (which may also be C t N) includes the desirability of its being ortho or para to the leaving group and the formation of aldehydic and alcoholic products when the reaction is performed in the presence of acetic acid which can efficiently trap the intermediate 68 H.Lund and A. D. Thomsen Acra Chem. Scand. 1969 23 3567. 69 K. L. Schroder Tetrahedron Letters 1970 2479. 70 R. A. Benkeser and S. J. Mels J. Org. Chem. 1970 35 261. 11 R. A. Benkeser H. Watanabe S. J. Mels and M. A. Sabol J. Org. Chem. 1970 35 1210. 72 P. E. Iversen Acta Chem. Scand. 1970 24 2459. 73 R. Woods Electrochim. Acta 1970 15 815. 74 J. P. Coleman H. G. Gilde J. H. P. Utley and B. C. L. Weedon Chem. Comm. 1970 738. Electro-organic Chernis try e I 6" 6' C02 Me IC\ 0 OMe L-(34) -X- -HC\ CO2Me COz Me CO Me '0 OMe R = H;X = OMeorOAc orR = X = F Scheme 28 (34).A similar mechanism might account for the racemisation found7' for reductive cleavage of the atrolactic acid derivative in Scheme 29. Other reac- OCOPh Ph ~ I YoPh/O' -PhCO,-\ /'O Ph-C-C0,Me -% Ph-C-C 1 I t-\ /c=c\ Me Me OMe OMe PhCH(Me)CO,Me PhC(Me)CO,Me -Ph /'='\ Me OMe Scheme 29 tions in this category are the electrored~ction~~ of the ester (35) to 4-methoxy- benzyl cyanide and the preparation of methyl ketone by cathodic cleavage77 of 8-ketosulphones (Scheme 30). (35) RCOCH(R')SO,Ph RCOCHR' fe_ RCOCHR' RCOCH,R' -PhSOZ-R = Ph or PhCH,; R' = H or Me Scheme 30 R. E. Erickson and C. M. Fischer J. Org. Chem. 1970 35 1604. '6 S. Wawzonek and J. D. Fredericksen J.Electrochem. Soc. 1959 106 325. " (a) B. Lamm and B. Samuelsson Acta Chem. Scand. 1970,24 561 ; (b)B. Lamm and B. Samuelsson Chem. Comm. 1970 1010. 238 J. H. P.Utley Reductive Cleavage of Halogens. Despite much experimental work the mechanism by which organic halides may be electroreduced to the hydrocarbon is not clear if indeed there is one mechanism. The full account78 of a careful study of the reduction of the halides (36) leaves the question open. The system was chosen because much is known about the stereochemical behaviour of the PhAMe Ph X (36) relevant radical (racemisation) and carbanion (retention). At the mercury cathode in acetonitrile solution the optically pure bromide of (36) is cleaved with 26% net retention.Use of CD,CN solvent results in a 75% labelling of the cyclo- propane but whether abstraction is by radical or carbanion is not clear. Small ring compounds may be obtained by the electroreduction of dihalides probably through concerted processes (Ann. Reports (B) 1969 p. 240). Further evidence for this saggestion and a useful preparative procedure comes from the macroscale (0.05 mol) electr~lysis~~ described in Scheme 3 1. -Br-Br Br 70 % Scheme 31 The organomercury compound (37) is formeds0 in the reduction at a mercury cathode of either 3-bromo-3-methylbut-1-yne or l-bromo-3-methylbuta-1,2-diene strongly suggesting for this cleavage reaction the intermediacy of the radical (38). Another organomercury derivative obtained cathodically is (39) which is formedI7 at -1.8V (us.S.C.E.)from (8) itself the oxidation product of cyclohexene at a mercury anode an acetonitrile solution (see p. 223). Me,C:C:CH-Hg-CH :C:CMe (37) Me2C:C:CH' (38) I COMe (39) ICOMe A good example of selective electrochemical reaction is provided in the . cathodic removal8' at controlled potential of bromine from p-bromo-y-chloro- butyrophenone (40) a substrate with three electroactive functional groups. p-BrC H,CO(CH,)3 C1 (40) " J. L. Webb C. K. Mann and H. M. Walborsky J. Amer. Chem. Soc. 1970,92,2042. 79 R. Gerdil Helv. Chim. Acta 1970 53 2100. 'O J. Simonet H. Doupeux P. Martinet and D. Bretelle Bull. SOC.chim. France 1970 3930. A. J. Fry M. A. Mitnick and R. G. Reed J. Org. Chem. 1970,35 1232.Electro-organic Chemistry 239 MiscelIaneous Unsaturated Functional Groups. In the presence of suitable adjacent functional groups electroreduction of nitro-groups on aromatic rings initiates cyclisation usually through the corresponding hydro~ylamine.~~-~ A particularly efficient reaction of this type is the near quantitative formationg5 of 3-phenylbenzo- 1,2,4-triazine (Scheme 32). An interesting dimerisation reaction is associated with the reduction" at -1-4V (us. S.C.E.),at mercury of bis-(p-nitropheny1)phosphate.A blue-green solution is obtained upon consumption of 3 Faradays mol-' and e.s.r. spectro- scopy confirms the presence of the 4,4'-dinitrodiphenyl radical anion. There is no evidence from the cyclic voltammogram that the aryl groups are cleaved in a stepwise fashion in the proposed Scheme 33.(ArO),P(0)OH [(ArO) P(O)OH]'-lsolvent SH [Ar-Ar]' Ar-Ar + O:P(OH) + 2s-Ar = p-02N.C6H4-Scheme 33 Intramolecular cyclisation reminiscent of the dihalide reductions (see p. 238) is the basis of a high yield electrochemical method for the preparationa7 of epoxides from ditosyloxyalkanes. For instance ethylene oxide is formed in 85 % yield by electrolysis of TsO(CH,),OTs in dry acetonitrile. The reactions of the superoxide ion 02-with organic compounds have recently attracted attention. Alkali-metal salts of the ion are not very soluble in organic solvents but electrolysis of tetraethylammonium perchlorate in dimethylsul- phoxide saturated with oxygen leadsaa to the formation of a stable solution of the quaternary ammonium superoxide.Under these conditions O2-behaves as a nucleophile. e.g. displacing chloride from alkyl chlorides with the formation of alkylperoxyl radicals. In another it was found that the epoxide (41) resulted from reaction between electrochemically generated superoxide ion and cyclohex-2-enone. 82 M. Jubault Compt. rend. 1970 271C 1671. 83 E. Laviron and T. Lewandowsa Bull. SOC. chim. France 1970 3177. 84 H. Lund and L. G. Feoktistov Actu Chem. Scand. 1969 23 3482. 85 S. Kwee and €3. Lund Actu Chem. Scand. 1969 23,271 1. 86 K. S. V. Santhanam and A. J. Bard J. Electroanalyt. Chem. Interfacial Electrochem. 1970 25 6A. R. Gerdil Helu. Chim. Acta 1970 53 2097. M. V. Merritt and D. T. Sawyer J. Org.Chem. 1970,35,2157. 89 R. Dietz M. E. Peover and P. Rothbaum Chem.-Ing.-Tech. 1970 42 185. 240 J. H. P.Utley (41) Practical Innovations.-Cells Solvents and Reference Electrodes. Details of a versatile cell have been p~blished.’~ One particular feature of the design is that gaseous products can be collected separately from the electrodes. The electroactive range methods of purification and details of suitable reference electrodes are included in an indispensible review” of the use of aprotic solvents in electrochemistry. For propylene carbonate the preparation and stability of salt bridges,92 the useful electroactive range,93 and the properties of a calomel electrode94 have been described. In this solvent between platinum electrodes with potassium hexafluorophosphate a potential range of 6.8 V was obtained before the current density exceeded 1 mA cm-’.As part of a series of reports on amide solvents the standard potential of the Ag/AgCl electrode in N-methylformamide has been rneas~red.~’ The molten salt system aluminium trichloride-sodium chloride-potassium chloride may be used for electrochemistry at 150°C and benzene has been efficiently ~xidised~~ under these conditions albeit to carbon as a ‘coherent deposit on the anode’. 90 G. Cauquis and B. Haemmerle Bull. SOC. chim. France 1970 2000. * C. K. Mann ‘Electroanalytical Chemistry’ ed. A. J. Bard Marcel Dekker Inc. N.Y. 1969 Vol. 3. 92 H. J. McComsey and M. S. Spritzer Analyt. Letters 1970 3 427. 93 J. Courtot-Coupez and M.L’Her Bull. SOC. chim. France 1970 1631. 9A I. Fried and H. Barak J. Electroanalyt. Chem. Interfacial Electrochem. 1970 27 170. 95 M. L. Berardelli G. Pecci and B. Scrosati J. Electrochem. SOC. 1970 117 781. 96 M. Fleischmann and D. Pletcher J. Electroanalyt. Chem. Interfacial Electrochem. 1970 25 449.