1977 1831Electrochemical Difunctionalisation of Adamantane and Further Oxid-ation of Substituted AdamantanesBy Alan Bewick, Gary J. Edwards, Stephen R. Jones, and John M. Mellor," Department of Chemistry, TheAnodic oxidation of adamantane in trifluoroacetic acid at a potential shown by cyclic voltammetry to be in the secondwave, gives after hydrolytic work up adamantane-I ,3-diol (70%). Similar oxidation in acetonitrile gives A"'-adamantane-l,3-diyl) bisacetarnide (58%). Electroanalytical techniques are used to define the oxidation mechan-ism, and oxidations of some substituted adamantanes at high anodic potentials are described.University, Southampton SO9 5N HTHERE are few methods available for direct introductionof polyfunctionality into saturated hydr0carbons.l Wehave previously reported studies of controlled electro-lysis conducted in the first oxidation wave.Here wereport in full the results of oxidation in the secondanodic wave, which for the first time establish thatdirect electrochemical difunctionalisation of hydro-carbons is possible in preparatively significant yields.RESULTSAs in earlier studies we have examined hydrocarbonsexpected to give stable primary products of oxidation. Byusing adamantanes, the formation of olefins, capable ofrapid further oxidation, is avoided.Volammetric data from linear potential sweep measure-ments for a series of adamantanes are reported in Table 1.Long-lived transient intermediates, for example cationradicals or dications produced in the anodic sweep, werenot detected: in the reverse cathodic sweep no wave wasobserved. For adamantane in trifluoroacetic acid threeseparate oxidation waves were detected, whereas in aceto-nitrile only two waves were observed.The results of preparative electrochemical oxidation ofadamantane in the first anodic wave both in trifluoroaceticacid and in acetonitrile have been de~cribed.~.~ Table 2shows the results of preparative electrolysis in the secondanodic wave.Product yields were determined either byg.1.c. or by isolation of products (see Experimental section).Preliminary report, A. Bewick, G. J. Edwards, S. R. Jones,and J. M. Mellor, Tetrahedron Letters, 1976, 631.G. J . Edwards, S. R. Jones, and J . M. Mellor, J.C.S. Perkin11, 1977, 505.The nature of the oxidative process in the second wavewas investigated both by coulometry and by additionalvoltammetric studies.Oxidation of adamantane in aceto-nitrile at 2.4 V (the first anodic wave)TABLE 1Voltammetric data for substitutedCompound SolventAdamantane MeCNl-t-Butyladamantane MeCN1,3-Dimethyladarnantane MeCNN-( l-Adamanty1)acetamide MeCNagainst an AglAg+adamantanesE,I,'/V Ep122/V2.38 2.862.33 2.812.42 2.871.90Adamantane CF,CO,H 1.86 2.35(2.75)Adamantan-1-01 CF3C0,H 2.30 2.763,5-Dimethyladamantan-l-o1 CF,CO,H 2.28 2.67Adamantanone CF,CO,H 1.75 2.55N- (l-Adamantyl) acetamide CF,CO,H 2.34 2.901- Adamantylamine CF,CO,H 2.38 2.671- Adamantylammonium CF,CO,H 2.46 2.89tetrafluoroboratea In acetonitrile the reference electrode was AgJAgNO,(0.01~) in MeCN.In trifluoroacetic acid the reference elec-trode was a silver wire in CF,C02H. Sweep rate 0.1 V s-l([substrate] 10 m M ) ; electrolyte Bun,NBF,. Dissolution ofalcohols in CF3C0,H leads to rapid formation of the tri-fluoroacetates.( 0 . 0 1 ~ ) electrode in acetonitrile led to a linear plot of currentveysus Faradays mol-l, which only deviated from linearitytowards the end of electrolysis. The slope of the earlylinear portion gave an n value of 2.0-2.2 Faradays mol-l.Similarly in the second wave a t 3.0 V, linearity was ob-served for >50% electrolysis and the n value was 4.0-4.33 V. R. Koch and L. L. Miller, J . Amer. Chem. Soc., 1973, 96,86311832 J.C.S. Perkin 1Faradays mol-l.More closely linear plots were observedfor electrolyses in trifluoroacetic acid, where the back-ground current was very low. In oxidation of adamantanen values were determined from these good linear plots.Oxidation a t 1.8 V (first anodic wave) (silver wire referenceelectrode) gave an n value of 1.9-2.0 Faradays mol-l andat 2.3 V (second anodic wave) a value of 3.8-4.0 Faradaysmol-l. In a separate experiment 1.9 Faradays mol-1 werepassed a t 1.8 V and then a further 2 Faradays mol-1 a t2.3 V. In both anodic regions linear plots indicating thepassage of 2 Faradays mol-1 in each wave were obtained.TABLE 2Electrochemical oxidation of adamantane and derivatives aPotentialCompound Solvent (V) Products (yo)Adamantane MeCN 3.0 (3) (6); (5) (58)Adamantane MeCN 3.0 (31 (24); ( 5 ) (20)Adamantane CF,CO,H 2.3 (gj (12j; (10)i2):(111 (511j12j i 5 j ; (13)(1)(44)Adamantane CF,CO,Hd 2.3 (9) (4); (11) (70)Adamantan-2-01 CF,CO,H 2.3 (13) (11); (12)Adamantan- 1-01 CF,CO,H 2.3 (11) (50)N - ( l-Adamanty1)- MeCN 2.8 (3) (83); 5N-( 1-Adamanty1)- CF,C02H 2.4 (9) (8); (11) (68)acetamide ((3)acetamideAdamantan-1-01 CF,C02He 2.3 (5) (80)Background electrolyte Bun4NBF,.Yields determinedby g.1.c. analysis. Yields of isolated products; backgroundelectrolyte Me4NBF4. For details see Experimental section.Work-up with MeCN-H+ (see Experimental section).These conclusions were substantiated by further volt-ammetric experiments. In cyclic volammetry linear plotspassing through the origin of a peak current density versus(sweep rate)* plot were obtained for both first and secondanodic waves in acetonitrile.Comparison of the values ofthe current function VF [equation (i)] with that obtainedIIA cm-2(v*/V* sd) x (clmol 1-1) VF = (i)for NN-bis-( 2,4,6-trimethoxyphenyl)methylamine (1) , acompound known to undergo reversible one-electronoxidation, showed that the transfer of two electrons wasassociated with each wave for adamantane [VF values 2.9for adamantane (first wave) ; 2.9 for adamantane (secondwave); 1.3 for (l)]. Similar waves in trifluoroacetic acidshowed plots of current density versus (sweep rate)* whichwere linear, and had the same slope, passing through theorigin.Similar conclusions were derived from studies at arotating disc electrode.Values of the current functionR F [equation (ii)] were obtained for rotation speedsIIA cm-2(o)/rad) s-*) x (clmol 1-l)R F = (ii)between 500 and 5000 rev. min-l. Comparison of theR F value for the standard (1) (0.09) with those of the firstwave (0.24) and the second wave (0.45) for adamantane inacetonitrile, making some allowance 4 for the difference indiffusion coefficients, indicates that the transfer of twoelectrons is associated with the first wave, and the transferof an additional two with the second.The oxidations of a number of substituted adamantaneswere examined with regard to synthetic utility. Volt-ammetric results are shown in Table 1 and product studiesin Table 2, and further voltammetric and coulometricexperiments are noted in the Discussion section.DISCUSSIONCoulometry and voltammetry establish that foradamantane in either trifluoroacetic acid or acetonitrileanodic oxidation leads to removal of two electrons inthe first wave and a further two in the second. Earlierpreparative studies 2 9 3 have shown that in the first wavethe trifluoroacetate of adamantan-1-01 (2) is the majorproduct in trifluoroacetic acid, and N-( 1-adamanty1)-acetamide (3) the major product in acetonitrile. Oxid-ation of adamantane in trifluoroacetic acid in the second(1)(2) X = 02C*CF,(3) X = NHAc(6) X = N:CMe(9) X = OH(10)(4) X = Y = @C*CF3(5) X=Y=NHAc(7) X =Y=N:?Me(8) X = NHAC, Y = 02C*CF3(11) X = Y = OHOH OH(12) (13)wave at 2.3 V gives a high yield of the bistrifluoroacetate(4), and similarly, after work-up of the solution fromoxidation of adamantane in acetonitrile in the secondwave at 3.0 V, the diamide (5) is obtained (see Table 2).These results are the first electrochemical difunctionalis-ations of saturated hydrocarbons to be reported.Interestingly, in trifluoroacetic acid higher yields of di-substituted products are obtained by passage of 2Faradays mol-1 at 1.8 V and then a further 2 Faradaysmol-l at 2.3 V.Two aspects of this work deservediscussion : the mechanistic aspects of the oxidation inthe second wave, and the generality of a process whichpermits either direct introduction of two functionalgroups into a hydrocarbon, or further functionalisationof a monofunctionalised substrate.In the case of N-(l-adamanty1)acetamide (3) at 2.8 V,although oxidation of the substrate occurs, as shown bythe large increase in electrode current on addition of theamide, isolation of the products leads largely to un-changed amide (83%), and the diamide (5) is only4 V.D. Parker, Electrochim. Acta, 1973, 18, 5191977obtained in trace quantities ((3%). This shows thatthe diamide (5) is not obtained in the second waveoxidation of adamantane v i a the monoamide (3) asintermediate. This is substantiated by voltammetry ofthe amide (3), which shows peak positions considerablydisplaced from those of adamantane. A simple explan-ation of these results is that oxidation of the amide (3)electrochemically leads to the l-adamantyl cation byoxidation of the amide side chain. Capture of thiscation by solvent then leads to the nitrilium ion (6), andwith water this gives back the amide (3).Overall theprocess represents an oxidation of acetonitrile. Inoxidation of adamantane in the second wave the crucialintermediate is the nitrilium ion (6). In the rigorousabsence of water no amide can be formed and hence theelectroactive species in the second wave must be themononitrilium ion (6), which on further oxidation givesthe dinitrilium ion (7). Quenching with water thengives the diamide (5). The difference between thepotentials of the first and second waves of adamantaneis explained by the deactivation of further electronwithdrawal afforded by the positively charged nitriliumsubstituent.This analysis accords well with the earlierobservation 3 that in the first wave adamantane isoxidised to give a nitrilium ion in acetonitrile.Oxidation of adamantane in trifluoroacetic acid ismechanistically more straightforward. In the secondwave the product of oxidation of the first wave, l-ada-mantyl trifluoroacetate, is further oxidised to givedirectly the diester (4). Adamantan-1-01 can be electro-lysed in trifluoroacetic acid to give the diester (4).Again the difference between the potentials of the firstand second waves of adamantane in trifluoroacetic acidis explained by the deactivation effect of the substituent,here the trifluoroacetoxy-group. We note a confirm-ation of this analysis in the similar EplZ values fromcyclic voltammetry of the second wave of adamantaneand the first wave of adamantan-1-01 in trifluoroaceticacid (i.e.l-adamantyl trifluoroacetate). In such casessuccessive functionalisation is possible by appropriateincrease in the applied potential. Polyamides areobtained in dry nitrile solvents, polyesters in trifluoro-acetic acid.Although the amide (3) is oxidised in acetonitrile byloss of electrons from the functional group, we thought itpossible that in trifluoroacetic acid protonation of thefunctional group might lead to preferential oxidation ofthe hydrocarbon moiety. Oxidations of some sub-stituted adamantanes have been r e p ~ r t e d , ~ ? ~ and bothoxidation of the substituent and difunctionalisation havebeen observed.Oxidation of the amide (3) at 2.4 V intrifluoroace t ic acid containing trifluoroacet ic anhydride(5%) gave the diester (4). The formation of this esterwas explained when the amide (3) was observed todissolve in the above solvent in the absence of an appliedcurrent. Solvolytic cleavage gave the ester (2) inF. Vincent, R. Tardivel, and P. Mison, Tetrahedron, 1976,32,R. N . Lacey, J . Chem. SOC., 1960, 1633.1681.high yield. Precedent for solvolytic conversion of anamide into ester exists.6 Trifluoroacetamides undergoready solvolysis.7 Solvolyses of derivatives of N-t-but ylacetamide and of N- (l-adaman tyl) acetamide aremarkedly facilitated by the stability of the tertiarycarbocation formed. Here it permits efficient passagefrom a bridgehead amide to bridgehead alcohol.Electrolysis of the amide (3) in trifluoroacetic acidalone gave the amide ester (8) in poor yield.Furtherelectrolyses of other substituted adamantanes aredescribed in the Experimental section. Oxidation ofadamantanols is an efficient route to diesters and di-amides IThe major problems in the electrochemical poly-functionalisation of saturated hydrocarbons are (a) theneed to use high anode potentials and (b) the efficientcapture of reaction intermediates by nucleophilicsubstitution to prevent olefin formation and hencecomplex secondary oxidation processes. This workshows that in both acetonitrile and trifluoroacetic acidthe problems associated with the use of high anodepotentials are not important. By using bridged hydro-carbons, unable to form olefins, the second constraint isremoved and hence difunctionalisation proceeds in asatisfactory manner. To make such polyfunctionalis-ations more general requires methods for more efficientcapture of intermediate carbocations.EXPERIMENTALGeneral experimental details and electrochemical tech-niques 2 have been reported in previous papers.Pro-cedures for purification of solvents and the electrolyteand of adamantane have been described. Adamantan-1-01(Cambrian), adamantan-2-01 (Aldrich) , l-adamantylaminehydrochloride (Cambrian), and adamantan-%one (Aldrich)were used as supplied. Preparations of l-t-butylada-mantane, N-( l-adamantyl)acetamide, and 3,5-dimethyl-adamantan- 1-01 have been described. 2*Electrolysis of Adarnantane in Acetonitrile at 3.0 V.-Adamantane (71 mg) was oxidised at 3.0 V in acetonitrilewith tetramethylammonium tetrafluoroborate (0.IM) aselectrolyte. In the absence of added adamantane abackground current of 6 mA was recorded. Addition ofadamantane led to a current of 50 mA which fell to 8 mA.Water (0.2 ml) was then added to the anolyte and thesolvent removed under reduced pressure. The residue waswashed with water and then dichloromethane, and theaqueous layer was further extracted continuously withdichloromethane. The organic extracts were dried andevaporated to leave a white solid (90 mg). Preparativet.1.c. (ethyl acetate as eluant) afforded N-( l-adamanty1)-acetamide (3) (23 mg), identical with an authentic sampleand NN'-( adamantane- 1,S-diyl) bisacetamide (5) (25 mg),which was recrystallised from ethyl acetate, m.p.226-227" (lit.,@ 226-227'), 6 1.60 (2 H, t), 1.88 (s) and 1.96br(total 14 H), 2.22br (2 H), 2.28 (2 H, s), and 5.27br (NH);vn/e 250 (60%, ,'H+), 207 (100, M - C,H,O), 193 (61, M -57), 149 (22), 148 (21), 136 (33), 108 (22), and 43 (48);vmx. (CHC1,) 3 400, 1 660, 1 520, and 1 300 cm-1.J. B. Hendrickson, R. Bergeron, A. Giga, and D. Sternbach,J . Amer. Chem. SOC., 1973, 95, 3412.S. R. Jones and J . M. Mellor, J.C.S. Perkin I, 1976, 2576.H. Stetter and C . Wulff, Chem. Ber., 1960, 03, 1366J.C.S. Perkin IIn a further experiment product yields were determineddirectly by g.1.c. analysis.Thus adamantane (29 mg)was oxidised at 3.0 V with tetrabutylammonium tetra-fluoroborate as electrolyte until the current had fallen tothe background value. Water (0.2 ml) was added. Directg.1.c. analysis (with added internal standard) indicatedthe formation of N-( l-adamanty1)acetamide (3) (6%) andNN'-(adamantane- 1,3-diyl)bisacetamide (5) (58%).Electrolysis of Adamantane in Trijluoroacetic A cad at2.3 V.-Adamantane (47.8 mg) was oxidised a t 2.3 V intrifluoroacetic acid with added trifluoroacetic anhydride(1%) and tetrabutylammonium tetrafluoroborate ( 0 . 1 ~ ) aselectrolyte. In the absence of added adamantane a back-ground current of 0.6 mA was recorded. Addition ofadamantane led to a current of 55 mA which fell to 5 mA.After electrolysis the anolyte was poured onto ice-watercovered with ether and the acidity was quenched by slowaddition of sodium hydrogen carbonate. The ethersolution was separated, dried (MgSO,) , and evaporated,and the residue was hydrolysed as previously described.8G.1.c.analysis of the hydrolysate showed the presence ofadamantan-1-01 ( 12y0), adamantan-2-01 (2y0), adamantane-lf3-diol (51 yo), adamantane- 1,4-diol (mixture of isomers ;5%), and adamantane-l,2-diol (1%) (by comparison withauthentic samples).In a further experiment adamantane (52 mg) wasoxidised under similar conditions but at 1.8 V for thepassage of 2 Faradays mol-l and then a t 2.3 V for thepassage of a further 2 Faradays mol-l. The initial currentwas 29 mA and the final current 1.7 mA (background 0.1mA).Work-up and hydrolysis as above afforded amixture of alcohols separated by preparative t.1.c. Elutionwith ethyl acetate gave adamantan-1-01 (2 mg, 4%) andadamantane- 1,S-diol (45 mg, 70%). Recrystallisation fromethyl acetate-ethyl alcohol (9 : 1) afforded pure adamantane-1,3-diol, m.p. 330" (1it.,lo 325-330") ; 6 [(CD,),SO] 1.49(12 H, m), 2.12 (2 H, s), and 4.37 (2 H, s); m/e 168 (19%,M+), 150 (4, M - H,O), 112 ( 8 ) , 111 (loo), 110 ( 5 ) , 109 (4),108 (a), and 95 (16); Y,. (KBr) 3 220, 2 930, 2 850,1580, 1443, 1 327, 1295, 1200, 1 131, and 1 024 cm-l.Other, minor products of the previous experiment werenot observed here.Electrolysis of Adamantan- l-ol in l'rij7uoroucetic A cid at2.3 V.-Adamantan-1-01 (100 mg) was oxidised a t 2.3 Vin trifluoroacetic acid with added trifluoroacetic anhydride(1%) and tetrabutylammonium tetrafluoroborate ( 0 .1 ~ ) aselectrolyte. An undivided preparative cell was used. Theinitial current (42 mA) fell to 8 mA for the passage of2 Faradays mol-l. The potential was pulsed to 1.0 Vanodic for 1 in every 30 s. After electrolysis the solutionwas worked up to give a mixture of products analysed byg.1.c. Before hydrolysis no alcohols were observed and theonly products were adamantane (a trace), the bis trifluoro-acetate of adamantane-1,3-diol (!joy0), and some l-ada-mantyl trifluoroacetate which was not oxidised.Electrolysis of Adamantan-2-01 in Tripuoroacetic A cid at2.3 V.-Adamantan-2-01 (49 mg) was oxidised at 2.3 V intrifluoroacetic acid with added trifluoroacetic anhydride(1 %) and tetrabutylammonium tetrafluoroborate (0.ZM) aselectrolyte. The initial current (14.7 mA) had fallen to3.3 mA at the end of electrolysis. The potential waspulsed to 1.0 V anodic for 1 in every 30 s. After electro-lysis the anolyte was worked up to give a mixture ofproducts, which were separated by preparative t .l.c.Elution with ethyl acetate afforded adamantane- 1,2-diol(7 mg, 11%) and a mixture of adamantane-1,4-diols (23 mg,44y0), identified by comparison with authentic samplesobtained by reduction (LiAlH,) of 5-hydroxyadamantanone.Trijluoroacetolysis of N-( l-Adamanty1)acetamide (3) .-N-( 1-Adamanty1)acetamide (3) (25 mg) dissolved in tri-fluoroacetic acid-trifluoroacetic anhydride ( 19 : 1) was setaside a t room temperature for 3 h.G.1.c. analysis thenshowed the presence of a little unchanged amide butsubstantial formation of l-adamantyl trifluoroacetate ( 1)( >80yo). In trifluoroacetic acid alone after 3 h g.1.c.analysis showed little (< 5%) formation of l-adamantyltrifluoroacetate (1).Electrolysis of N- ( l-Adamanty1)acetamide (3) in Trijluoro-acetic Acid at 2.4 V.-N-( l-Adamanty1)acetamide (3) (176mg) was oxidised a t 2.4 V in trifluoroacetic acid with addedtrifluoroacetic anhydride ( 1 %) and tetrabutylammoniumtetrafluoroborate as electrolyte. After electrolysis in which2 Faradays mol-l were passed and the current fell from28 to 0.7 mA the solution was worked up as above. G.1.c.analysis showed formation of (2) and (4) as the majorproducts.The products were hydrolysed as describedabove and the hydrolysis products were analysed by g.1.c.and subsequently crystallised. Analysis showed thepresence of adamantan- l-ol (8%) and adamantane- 1,3-diol(68%). Recrystallisation afforded pure adamantane-l,3-diol.Electrolysis of N-( l-Adamanty1)acetamide (3) in Aceto-nitrile at 2.8 V.-N-( l-Adamanty1)acetamide (17.7 mg) wasoxidised a t 2.8 V in acetonitrile with tetrabutylammoniumtetrafluoroborate as electrolyte. Electrolysis as describedabove and work-up provided a sample for g.1.c. analysis.This showed that the starting material was largely un-changed (83% recovery) and little diamide (<3%) wasobtained even after the passage of 2 Faradays mol-l.Electrolysis of Adamantan- 1-02 in Trijluoroacetic Acid at2.3 V and Reaction of the Products with Acetonitrile.-Adamantan-1-01 (208 mg) was oxidised a t 2.3 V in trifluoro-acetic acid as above. After the passage of 2 Faradaysmol-l work-up was effected by extraction with ether from asolution of the anolyte in ice-water. The acid was neutral-ised by slow addition of sodium carbonate. The residualyellow oil was heated a t 95 "C with stirring for 48 h withacetonitrile (4 ml), trifluoroacetic acid ( 8 ml), sulphuric acid(98%; 0.8 ml), and water (0.1 ml). The cold solution waspoured into water (80 ml) and chloroform (80 ml), sodiumcarbonate was added, and the aqueous layer was furtherextracted with chloroform. The combined chloroformextracts were dried and evaporated and the solid residuewas analysed by g.1.c. and recrystallised. Analysis showedthe presence of NN'-(adamantane- 1,3-diyl)bisacetamide ( 5 )as major product (80%). Recrystallisation from ethylacetate afforded pure diamide ( 5 ) , m.p. 224-225" (144 mg,42%).We thank the S.R.C. for financial support.[7/284 Received, 16th February, 19771H. W. Geluk and J . L. M. A. Schlatmann, Tetrahedron, 1968,24, 5369