Cathodic Reduction of Acetophenone in Acidic Methanol Electrode Kinetic Study of a Novel Vicinal Diether Synthesis BY MARAJ UD DIN BHATTI AND OLIVER R. BROWN * Electrochemistry Research Laboratories, Department of Physical Chemistry, The University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU Received 10th April, 1974 The polarisation curve for the reduction of acetophenone in acidic methanol shows a prewave and a mainwave; the total wave height is diffusion controlled, corresponding to a reaction involving one electron per molecule. The prewave is kinetically controlled, except of course when it contributes a large fraction of the total wave. The products formed at potentials on the shoulder of the prewave are the acetophenone pinacol dimethyl diethers. The mechanism is analogous to that which operates in acetophenone pinacol formation, the reaction which predominates in aqueous media.In the presence of acid, methanol and acetophenone are shown to form the dimethyl ketal ; a value for the equilibrium constant is determined. An earlier paper presented an electrode kinetic study of the cathodic hydro- dimerisation of acetophenone, carried out in acidic aqueous-methanol solution. The rate-determining step in pinacol formation was shown to be an irreversible dimerisation of free ketyl radicals following a reversible electron and proton transfer. However, a previous study conducted in " anhydrous " methanol containing H2S04 reported for acetophenone reduction a Tafel slope of approximately 4.6 RT/F, a value which is usually taken to indicate rate control by the first charge transfer step.A somewhat perfunctory repetition of that work lent support to that result. The present study was undertaken to examine more closely the " anhydrous " methanolic system. HCl has been employed as an alternative to H2S04 as the electrolyte. The water content of the solvent is described in the experimental section and that of the solutions in the discussion section; in general it was less than 0.2 % by volume. EXPERIMENTAL All kinetic work was performed in B.D.H. AristaR methanol, solutions being made up in a dry-bag filled with a nitrogen atmosphere. B.D.H. claim a maximum water content of 0.1 % for this solvent. Solutions of HCI in methanol were prepared simply by the uptake of gas generated by the action of B.D.H. AnalaR HzSO4 upon sodium chloride.AnalaR H2S04 was also used to make up the sulphuric acid solutions. Acetophenone (B.D.H.) was purified by distillation through a Vigreux condenser ; the first and final quarters of the distillate were discarded. Anhydrous lithium chloride was a B.D.H. product. All cells were divided by glass sinters. Counter electrodes in preliminary work were of platinum but this resulted in eventual deposition of the noble metal on the cathode when anhydrous methanolic HCI was the electrolyte ; subsequently carbon counter electrodes were adopted. Reference electrodes were separated from the Luggin capillary by a glass frit inserted between two closed taps, all in series; the tap nearer to the reference electrode contained the liquid junction.The high electrical resistance of the taps was by-passed by a.c. signals through a 1 pF capacitance connected between the reference electrode lead and a platinum wire sealed into the Luggin capillary. The Hg/Hg2C12/satd. KCI aq. reference electrode (s.c.e.) was used with HCI catholytes and Hg/Hg2S04/H2S04 aq. (1 mol dm-3) was used with H2S04 catholytes. All potentials are quoted relative to these electrodes. Solid rotating disc electrodes (r.d.e.) of lead, amalgamated gold and pyrolytic graphite 106M. D. BHATTI AND 0. R. BROWN 107 were prepared as previously described.l Most of the kinetic work, however, employed liquid mercury cathodes. Lifetimes of these dropping mercury electrodes (d.m.e.) were imposed by means of a striker operating on the capillary.Currents with values far below the mass transfer limited currents for acetophenone were obtained as the instantaneous currents at the end of the drop life (3.0 s or 5.0 s). These were recorded at 1 mV s-' on a fast-response X- Y plotter (Bryans 26000). Higher currents corresponding to more severe cathodic polarisations were obtained in the absence of mass transfer control by means of the potential step relaxation method. The drop striker initiated the potential programme which consisted of a 3.0s delay at the base potential, where no reaction occurred, followed by 32 ms at the cathodic polarisation potential before the return to the base potential. During the polarisation period, currents were recorded in a digital store (Hitek Instruments signal averager AA1) and subsequently plotted on the X- Y recorder.Acetophenone concentrations were varied by means of additions from precision microsyringes. The resistance within the capillary was directly measured whereas the total ohmic resistance (at 3.0 s) was obtained from the limiting slope of the i against Vrelation at extreme cathodic potentials. Experimental points were corrected by an appropriate voltage shift in a positive direction. The ohmic error at 5.0 s was easily derived, assuming the drop to be spherical. The Luggin capillary was placed approximately 0.4 cm from the drop. Electrode areas were easily evaluated from a measurement of the mercury flow rate. Results presented are those after correction. Mercury pool cathodes and AnalaR methanol were employed in preparative electrolyses.All kinetic experiments were performed at 22f 1°C and preparative electrolyses were carried out at 30+ 5°C. A Hewlett-Packard 185 Analyser was used for elemental analysis. N.m.r. spectra were run in deutrochloroform with tetramethylsilane internal standard on a Bruker- Spectrospin instrument at ambient temperature. Mass spectra were taken on a high resolution MS 9 instrument. Ultra-violet absorption measurements were carried out using 1 .O cm cells and a Hitachi- Perkin Elmer 124 instrument. RESULTS On lead, pyrolytic graphite or mercury-coated r.d.e. or at the d.m.e., aceto- phenone in 1 mol dm-3 HCl solution in methanol gave two waves. The total wave height varied linearly with the acetophenone concentration and with m* (cu is the r.d.e.rotation velocity) or h* (when a d.m.e. was used without the striker, h being the mercury column height). The first wave height was sensitive to traces of water in the electrolyte, especially when dilute solutions of acetophenone were used ; addition of traces of water caused a fall in the first wave. This wave height was independent of (I) or h (when no striker was used) except when the solutions were dilute and almost completely anhydrous ; then the prewave almost reached the total wave height, particularly at low (I) values. Polarisation data were essentially independent of the electrode material with the exception that the prewave on lead was distorted by the anodic oxidation of the metal. Fig. 1 shows results obtained at the d.m.e. in nominally anhydrous methanol.There is no indication of the high values of the Tafel slopes previously reported. It can be seen that the currents at the foot of the prewave give linear Tafel plots with slopes 40f 1 mV and a reaction order of approximately 1.5 with respect to aceto- phenone. Although the prewave does not give a perfect plateau on account of the commencement of the main wave, it is clear that currents in the region of the point of inflexion have an order unity. Accurate kinetic analysis of the main wave was precluded by the large value of the ohmic overpotential at the relatively high current densities involved. However, the limiting currents of the total wave essentially were the same as those obtained in aqueous-methanol solutions and can be considered to indicate a 1 F per mole process.108 CATHODIC REDUCTION OF ACETOPHENONE FIG.1 dm-3) - EJV .-Cathodic polarisation data in nominally anhydrous methanol containing HCI (1 .O mol and acetophenone (concentrations shown on the figure in mol dm-3). 0, instantaneous polarographic currents ; x , diffusion-relaxed data. N b 6 -c + 0.6 0.7 0.8 0.9 - E/V FIG 2.-The effect of progressive additions of water (amounts shown as parts per thousand by volume) on the polarisation data for the prewave given by a 0.03 mol dm-3 solution of acetophenone in an- hydrous methanol (1.0 mol dm-3 in HCl) at a mercury cathode.M. D . BHATTI AND 0. R. BROWN 109 The effect on the prewave of adding water to a 0.03 mol dm-3 acetophenone solution in nominally anhydrous methanol (1.0 mol dm-3 in HCl) is illustrated in fig.2. The addition of 1 % (by volume) water decreased prewave currents by an order of magnitude. The same phenomenon was observed at all acetophenone concentrations ; currents over the entire prewave fell markedly whereas the main wave was relatively unaffected. The net result was that the prewave fell in height and simultaneously moved towards more negative potentials until both waves merged into one distorted wave. I. I 1.2 1.3 I :L - E/V FIG. 3.-Cathodic polarisation data in 80 mole % methanol/20 % water containing H2S04 (1 .O mol dm-3) and acetophenone (concentrations shown in mol dm-3). 0, instantaneous polarographic currents ; 0, diffusion-relaxed data. Whereas the polarisation data obtained during any one experimental run were highly reproducible, differences in the intrinsic water contents of the solutions made reproducibility of the prewave polarisation data poor from one run to another.Thus measurements of the reaction order with respect to HC1 were subject to con- siderable error. Currents obtained using a nominally anhydrous methanol solution (0.1 mol dm-3 HCl and 0.9 mol dm-3 LiCl) were always considerably lower, at a given potential and acetophenone concentration, than those obtained with 1 .O mol dm-3 HCl in the nominally similar solvent but the difference factor varied widely. Polarisation experiments using the d.m.e. were repeated with 1.0 mol dm-3 methanolic sulphuric acid, and similar results were obtained. In fig. 3 are presented polarisation curves of acetophenone in 1.0 rnol dm-3 H2S04 solution prepared in the mixed solvent 80 mole % methanol 20 % water.In order to examine the nature of acidified methanol solutions of acetophenone, the main ultraviolet absorption peak of acetophenone (7c - z* transition, A = 242 nm, c - lo3 mol-' m2) was measured for mol dm-3 acetophenone solutions in110 CATHODIC REDUCTION OF ACETOPHENONE methanol, with and without HCl and water. The acidified nominally anhydrous solutions gave considerably smaller absorptions ( A ) than corresponding solutions without acid (Ao). The concentration of HCl in the range investigated (0.01 to 1.0 mol dm-3) appeared to be unimportant. In some cases the presence of HCl more than halved the acetophenone absorption but, as with the electrochemical kinetic behaviour of the prewave, traces of water had a considerable effect and merely 2 % (by volume) water caused the peak to return essentially to the height observed in unacidified “anhydrous” methanol.Fig. 4 shows a plot of relative decrease in absorbance A/(& - A ) against the volume fraction of added water. Finally, preparative electrolyses have been carried out on a solution of aceto- phenone (1.0 mol drn-,) in “ anhydrous ” methanol containing HCl (1.0 mol dm-3) at an electrode potential -0.66 V (against s.c.e.) corresponding to the shoulder of the pre-wave. Within an hour of the commencement of electrolysis white crystals X began to separate out on the cathode. After electrolysis the crystals were collected and recrystallised from methanol. The remaining catholyte was neutralised by addition of solid sodium carbonate and the solution filtered.The residue was dissolved in water and extracted with diethyl ether. The ether layer was added to the filtrate and the total volume reduced using a rotary evaporator at 100°C to obtain a further crop of crystals. The crystalline product X was examined by proton resonance spectroscopy, mass spectrometry and CH analysis. The n.m.r. spectrum consisted of a singlet at 1.46 p.p.m., another at 3.05 and an aromatic peak at 7.27 p.p.m. ; these peak areas were in the approximate ratio 1 : 1 :2. The mass spectrum at 80°C showed a small peak at 270 mass number, with progressively larger contributions at 255 and 249. However, by far the largest peak was at 135. The elemental analysis was C 79.5 %, H 8.14 %, 0 (by difference) 12.36 %.A cryoscopic determination of molecular weight was also carried out using benzene as solvent. A value of 250+ 10 % was obtained. The melting point (ca. 160°C) of the product was not sharp indicating a mixture of materials. DISCUSSION The crystalline product formed at potentials corresponding to the prewave is almost certainly a mixture of the stereoisomeric diethers, each with the formula CH3 CH3 I I I I OCH, OCH, C6Hj-C-- C C6H5. The n.m.r. spectrum can be interpreted as due to six methyl protons at 6 = 1.46, six methoxy protons at 3.05 and two phenyl groups at 7.27. The C, H, 0 figures (79.5, 8.14, 12.36 %) compare with those calculated (80.0, 8.15, 11.85 % respectively). The mass spectrum was clearly dominated by fragments resulting from a symmetrical dissociation although the parent ion (270) and fragments due to loss of methyl or methoxy groups were in evidence.The n.m.r. result and the cryoscopic measurement indicate the dimers in preference to the monomeric ethers C6H5CH(OCH3)CH3. Since the main wave was relatively unaffected by the presence of water it is reasonable to suppose that it corresponds to the formation of acetophenone pinacols.’ The prewave observed for acetophenone in acidified anhydrous methanol possesses all of the characteristics of a kinetic wave, i.e. the limiting current is determined by the rate of a homogeneous chemical reaction preceding the charge transfer step.M. D. BHATTI AND 0. R. BROWN 1 1 1 The electrochemical kinetic data obtained for the foot of the prewave indicate that the reaction mechanism is of the type already reported for the hydromerisation of acetophenone.A Tafel slope of 1.5 RT/Fand a reaction order of 1.5 are characteristic of a mechanism in which rate control is by an irreversible homogeneous bimolecular step involving species produced in an initial reversible one electron electrochemical event. However, in the present system the reaction product is not the pinacol but instead the corresponding diether. Arguing by analogy with the pinacolisation reaction it is reasonable to consider that the present reaction is a reduction of the x 2 Phc(0Me)Me + e- + Phc(0Me)Me -+ \ Me OMe / It is necessary therefore to explain the occurrence of the phenyl methoxy methyl carbonium ion in this system.It is well known that acetals can be prepared by the reaction between aldehydes and alcohols in the presence of dry HCl. Ketals too, although not usually isolated from such reactions, have been shown to form in equilibria involving the ketone and alcohol under anhydrous conditions. As acetal formations are acid-catalysed one can propose the usual type of mechanism : OH Me + H+ + R1R2C(OMe)OH MeOH / R1R2CO+H+ + (R1R2COH)+ + R1R2C - MeOH Hf + R'R2C(OMe)OH + R1R2C(OMe)OH2 - H+ \ / hemi ke t a1 0' H + (R1R2COMe)+ + \ - H20 MeOH - MeOH H2O OMe / R ~ R ~ C - H + H+ + R1R2C(OMe)2. \+/ ketal 0 Me \ SCHEME 1 The ketal of acetophenone is known; it has been prepared from acetophenone in methanol by the action of either formimido methyl ether or methyl orthoformate.6 There do not appear to have been any studies of the equilibrium between aceto- phenone and methanol although U.V.absorption has been used often to study ketone/ ketal equilibria.'112 CATHODIC REDUCTION OF ACETOPHENONE The linear plot of fig, 4 is consistent with the assumptions that (a) the equilibrium is set up rapidly in acid solution, (b) the amounts of hemiketal and other intermediates in scheme 1 are negligible, and (c) the ketal has an insignificant extinction coefficient at 242 nm. From the slope we deduce that the equilibrium constant at 22°C is PhCOCH3 + 2CH3OH + PhC(OCH3)2CH3 + H20 K = c ~ ~ ~ ~ ~ c ~ ~ ~ / c ~ = 3 x mol dm-3 and that the initial water content of the solution was slightly in excess of 0.1 volume % (the methanol manufacturer’s specified limit).-I 0 I 2 3 1, 5 6 7 6 lo3 x vol H20/vol solution FIG. 4.-Variation of the relative decrease in absorbance of acetophenone in the presence of acid with the concentration of added water. From the electrode kinetic results it is not possible to conclude whether the chemical process which limits the currents on the first wave is the formation of phenyl methoxy methyl carbonium ion from the ketal, or from the acetophenone or by a mixture of both routes (scheme 1). The effect of added water on the kinetics of reductions in the Tafel region is readily understood in terms of the pre-equilibrium + PhCOCH3 + CH30H +HI- + PhC(OCH3)CH3 + H20. It might be expected that the reaction order with respect to added water should be - 1.5 according to the theory How- ever, the solvent contains an unknown residual trace of water.Thus if the data of fig. 2 are plotted as log c~~~ against E at constant current i a linear relation is not obtained. However, if we guess values of the total residual water c,,, and plot log ( C ~ ~ ~ + C ~ , , ) against E until a linear relation is obtained then the slope of the plot should equal 2.303 RT/F. This is found to apply in fig. 5 where the potentials corresponding to a current density of 17.7 pA cm-2 are plotted against log cHz0 and against log ( c ~ ~ ~ + c,) with cres corresponding to 0.2 vol % water. This value is almost double that obtained from the U.V. measurements (fig. 4) but the methanol had less opportunity to pick up extraneous water in those experiments. Deviations from of reactions of the type described by eqn (1).\ \ OCH341430H CH3 CH3 / PhC-CH3 0.6 0.65 0.7 0.75 - E/V FIG.5.-Analysis of the potential shift with water concentration for current density = 17.7pA ciir2 from the data in fig. 2. 0, zero initial water ; 0, 2 parts per thousand water by volume. \ / OCH3 CH3 Conway and Rudd unfortunately made their measurements at lower current densities where, as fig. 3 shows, the polarisation data are distorted by the prewave so that in the region of - 1180 mV the reaction order with respect to acetophenone is unity and the Tafel plots are in fact gentle curves with a mean slope of some 52 mV. Those workers believed the pinacol formation was quantitative in that system but in fact their kinetic measurements were made under conditions where it is now evident that one reaction gradually gave way to another.Each electrode reaction separately follows the kinetic scheme first discussed by Koutecky and Hanus but in the region of mixed kinetics such behaviour is not observed owing to the influence of the preceding chemical reaction which limits the first wave.114 CATHODIC REDUCTION OF ACETOPHENONE The current yield of 2,3-dimethoxy-2,3-diphenylbutane obtained from methanolic acetophenone (0.3 mol dm-3) containing HCl (1 .O mol dm-3) at - 0.68 V was 23 %. This unexpectedly low value is undoubtedly due to the fall in the current for this reaction owing to the water it produces, and the consequent relative increase in the pinacol formation current. Other products were found to include the pinacols and there were also other gas chromatographic peaks which were probably due to the mixed dimers (mono ethers).Further work is in progress to investigate the synthetic scope of this type of reaction. M. D. B. thanks the British Council for a studentship. M. P. J. Brennan and 0. R. Brown, J.C.S. Faruday I, 1973, 69, 132. E. J. Rudd and B. E. Conway, Trans. Faraday Soc., 1971,67,440. 0. R. Brown, Disc. Faraday Sac., 1968, 45, 126. 1964), p. 296. L. Claisen, Ber., 1898, 31, 1010. M. T. Bogart and P.P. Herrera, J. Amer. Chem. SOC., 1923,45, 238. ' 0. H. Wheeler, J. Amer. Chem. SOC., 1957, 79, 4191. a J. Koutecky and V. Hanus, Coll. Czech. Chem. Comm., 1955, 20, 124. 4B. J. Cram and J. S. Hammond, Organic Chemistry (McGraw-Hill, New York, 2nd edn., Cathodic Reduction of Acetophenone in Acidic Methanol Electrode Kinetic Study of a Novel Vicinal Diether Synthesis BY MARAJ UD DIN BHATTI AND OLIVER R.BROWN * Electrochemistry Research Laboratories, Department of Physical Chemistry, The University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU Received 10th April, 1974 The polarisation curve for the reduction of acetophenone in acidic methanol shows a prewave and a mainwave; the total wave height is diffusion controlled, corresponding to a reaction involving one electron per molecule. The prewave is kinetically controlled, except of course when it contributes a large fraction of the total wave. The products formed at potentials on the shoulder of the prewave are the acetophenone pinacol dimethyl diethers. The mechanism is analogous to that which operates in acetophenone pinacol formation, the reaction which predominates in aqueous media.In the presence of acid, methanol and acetophenone are shown to form the dimethyl ketal ; a value for the equilibrium constant is determined. An earlier paper presented an electrode kinetic study of the cathodic hydro- dimerisation of acetophenone, carried out in acidic aqueous-methanol solution. The rate-determining step in pinacol formation was shown to be an irreversible dimerisation of free ketyl radicals following a reversible electron and proton transfer. However, a previous study conducted in " anhydrous " methanol containing H2S04 reported for acetophenone reduction a Tafel slope of approximately 4.6 RT/F, a value which is usually taken to indicate rate control by the first charge transfer step.A somewhat perfunctory repetition of that work lent support to that result. The present study was undertaken to examine more closely the " anhydrous " methanolic system. HCl has been employed as an alternative to H2S04 as the electrolyte. The water content of the solvent is described in the experimental section and that of the solutions in the discussion section; in general it was less than 0.2 % by volume. EXPERIMENTAL All kinetic work was performed in B.D.H. AristaR methanol, solutions being made up in a dry-bag filled with a nitrogen atmosphere. B.D.H. claim a maximum water content of 0.1 % for this solvent. Solutions of HCI in methanol were prepared simply by the uptake of gas generated by the action of B.D.H.AnalaR HzSO4 upon sodium chloride. AnalaR H2S04 was also used to make up the sulphuric acid solutions. Acetophenone (B.D.H.) was purified by distillation through a Vigreux condenser ; the first and final quarters of the distillate were discarded. Anhydrous lithium chloride was a B.D.H. product. All cells were divided by glass sinters. Counter electrodes in preliminary work were of platinum but this resulted in eventual deposition of the noble metal on the cathode when anhydrous methanolic HCI was the electrolyte ; subsequently carbon counter electrodes were adopted. Reference electrodes were separated from the Luggin capillary by a glass frit inserted between two closed taps, all in series; the tap nearer to the reference electrode contained the liquid junction.The high electrical resistance of the taps was by-passed by a.c. signals through a 1 pF capacitance connected between the reference electrode lead and a platinum wire sealed into the Luggin capillary. The Hg/Hg2C12/satd. KCI aq. reference electrode (s.c.e.) was used with HCI catholytes and Hg/Hg2S04/H2S04 aq. (1 mol dm-3) was used with H2S04 catholytes. All potentials are quoted relative to these electrodes. Solid rotating disc electrodes (r.d.e.) of lead, amalgamated gold and pyrolytic graphite 106M. D. BHATTI AND 0. R. BROWN 107 were prepared as previously described.l Most of the kinetic work, however, employed liquid mercury cathodes. Lifetimes of these dropping mercury electrodes (d.m.e.) were imposed by means of a striker operating on the capillary.Currents with values far below the mass transfer limited currents for acetophenone were obtained as the instantaneous currents at the end of the drop life (3.0 s or 5.0 s). These were recorded at 1 mV s-' on a fast-response X- Y plotter (Bryans 26000). Higher currents corresponding to more severe cathodic polarisations were obtained in the absence of mass transfer control by means of the potential step relaxation method. The drop striker initiated the potential programme which consisted of a 3.0s delay at the base potential, where no reaction occurred, followed by 32 ms at the cathodic polarisation potential before the return to the base potential. During the polarisation period, currents were recorded in a digital store (Hitek Instruments signal averager AA1) and subsequently plotted on the X- Y recorder. Acetophenone concentrations were varied by means of additions from precision microsyringes.The resistance within the capillary was directly measured whereas the total ohmic resistance (at 3.0 s) was obtained from the limiting slope of the i against Vrelation at extreme cathodic potentials. Experimental points were corrected by an appropriate voltage shift in a positive direction. The ohmic error at 5.0 s was easily derived, assuming the drop to be spherical. The Luggin capillary was placed approximately 0.4 cm from the drop. Electrode areas were easily evaluated from a measurement of the mercury flow rate. Results presented are those after correction. Mercury pool cathodes and AnalaR methanol were employed in preparative electrolyses.All kinetic experiments were performed at 22f 1°C and preparative electrolyses were carried out at 30+ 5°C. A Hewlett-Packard 185 Analyser was used for elemental analysis. N.m.r. spectra were run in deutrochloroform with tetramethylsilane internal standard on a Bruker- Spectrospin instrument at ambient temperature. Mass spectra were taken on a high resolution MS 9 instrument. Ultra-violet absorption measurements were carried out using 1 .O cm cells and a Hitachi- Perkin Elmer 124 instrument. RESULTS On lead, pyrolytic graphite or mercury-coated r.d.e. or at the d.m.e., aceto- phenone in 1 mol dm-3 HCl solution in methanol gave two waves. The total wave height varied linearly with the acetophenone concentration and with m* (cu is the r.d.e.rotation velocity) or h* (when a d.m.e. was used without the striker, h being the mercury column height). The first wave height was sensitive to traces of water in the electrolyte, especially when dilute solutions of acetophenone were used ; addition of traces of water caused a fall in the first wave. This wave height was independent of (I) or h (when no striker was used) except when the solutions were dilute and almost completely anhydrous ; then the prewave almost reached the total wave height, particularly at low (I) values. Polarisation data were essentially independent of the electrode material with the exception that the prewave on lead was distorted by the anodic oxidation of the metal. Fig. 1 shows results obtained at the d.m.e. in nominally anhydrous methanol.There is no indication of the high values of the Tafel slopes previously reported. It can be seen that the currents at the foot of the prewave give linear Tafel plots with slopes 40f 1 mV and a reaction order of approximately 1.5 with respect to aceto- phenone. Although the prewave does not give a perfect plateau on account of the commencement of the main wave, it is clear that currents in the region of the point of inflexion have an order unity. Accurate kinetic analysis of the main wave was precluded by the large value of the ohmic overpotential at the relatively high current densities involved. However, the limiting currents of the total wave essentially were the same as those obtained in aqueous-methanol solutions and can be considered to indicate a 1 F per mole process.108 CATHODIC REDUCTION OF ACETOPHENONE FIG.1 dm-3) - EJV .-Cathodic polarisation data in nominally anhydrous methanol containing HCI (1 .O mol and acetophenone (concentrations shown on the figure in mol dm-3). 0, instantaneous polarographic currents ; x , diffusion-relaxed data. N b 6 -c + 0.6 0.7 0.8 0.9 - E/V FIG 2.-The effect of progressive additions of water (amounts shown as parts per thousand by volume) on the polarisation data for the prewave given by a 0.03 mol dm-3 solution of acetophenone in an- hydrous methanol (1.0 mol dm-3 in HCl) at a mercury cathode.M. D . BHATTI AND 0. R. BROWN 109 The effect on the prewave of adding water to a 0.03 mol dm-3 acetophenone solution in nominally anhydrous methanol (1.0 mol dm-3 in HCl) is illustrated in fig.2. The addition of 1 % (by volume) water decreased prewave currents by an order of magnitude. The same phenomenon was observed at all acetophenone concentrations ; currents over the entire prewave fell markedly whereas the main wave was relatively unaffected. The net result was that the prewave fell in height and simultaneously moved towards more negative potentials until both waves merged into one distorted wave. I. I 1.2 1.3 I :L - E/V FIG. 3.-Cathodic polarisation data in 80 mole % methanol/20 % water containing H2S04 (1 .O mol dm-3) and acetophenone (concentrations shown in mol dm-3). 0, instantaneous polarographic currents ; 0, diffusion-relaxed data. Whereas the polarisation data obtained during any one experimental run were highly reproducible, differences in the intrinsic water contents of the solutions made reproducibility of the prewave polarisation data poor from one run to another.Thus measurements of the reaction order with respect to HC1 were subject to con- siderable error. Currents obtained using a nominally anhydrous methanol solution (0.1 mol dm-3 HCl and 0.9 mol dm-3 LiCl) were always considerably lower, at a given potential and acetophenone concentration, than those obtained with 1 .O mol dm-3 HCl in the nominally similar solvent but the difference factor varied widely. Polarisation experiments using the d.m.e. were repeated with 1.0 mol dm-3 methanolic sulphuric acid, and similar results were obtained. In fig. 3 are presented polarisation curves of acetophenone in 1.0 rnol dm-3 H2S04 solution prepared in the mixed solvent 80 mole % methanol 20 % water.In order to examine the nature of acidified methanol solutions of acetophenone, the main ultraviolet absorption peak of acetophenone (7c - z* transition, A = 242 nm, c - lo3 mol-' m2) was measured for mol dm-3 acetophenone solutions in110 CATHODIC REDUCTION OF ACETOPHENONE methanol, with and without HCl and water. The acidified nominally anhydrous solutions gave considerably smaller absorptions ( A ) than corresponding solutions without acid (Ao). The concentration of HCl in the range investigated (0.01 to 1.0 mol dm-3) appeared to be unimportant. In some cases the presence of HCl more than halved the acetophenone absorption but, as with the electrochemical kinetic behaviour of the prewave, traces of water had a considerable effect and merely 2 % (by volume) water caused the peak to return essentially to the height observed in unacidified “anhydrous” methanol. Fig.4 shows a plot of relative decrease in absorbance A/(& - A ) against the volume fraction of added water. Finally, preparative electrolyses have been carried out on a solution of aceto- phenone (1.0 mol drn-,) in “ anhydrous ” methanol containing HCl (1.0 mol dm-3) at an electrode potential -0.66 V (against s.c.e.) corresponding to the shoulder of the pre-wave. Within an hour of the commencement of electrolysis white crystals X began to separate out on the cathode. After electrolysis the crystals were collected and recrystallised from methanol. The remaining catholyte was neutralised by addition of solid sodium carbonate and the solution filtered.The residue was dissolved in water and extracted with diethyl ether. The ether layer was added to the filtrate and the total volume reduced using a rotary evaporator at 100°C to obtain a further crop of crystals. The crystalline product X was examined by proton resonance spectroscopy, mass spectrometry and CH analysis. The n.m.r. spectrum consisted of a singlet at 1.46 p.p.m., another at 3.05 and an aromatic peak at 7.27 p.p.m. ; these peak areas were in the approximate ratio 1 : 1 :2. The mass spectrum at 80°C showed a small peak at 270 mass number, with progressively larger contributions at 255 and 249. However, by far the largest peak was at 135. The elemental analysis was C 79.5 %, H 8.14 %, 0 (by difference) 12.36 %.A cryoscopic determination of molecular weight was also carried out using benzene as solvent. A value of 250+ 10 % was obtained. The melting point (ca. 160°C) of the product was not sharp indicating a mixture of materials. DISCUSSION The crystalline product formed at potentials corresponding to the prewave is almost certainly a mixture of the stereoisomeric diethers, each with the formula CH3 CH3 I I I I OCH, OCH, C6Hj-C-- C C6H5. The n.m.r. spectrum can be interpreted as due to six methyl protons at 6 = 1.46, six methoxy protons at 3.05 and two phenyl groups at 7.27. The C, H, 0 figures (79.5, 8.14, 12.36 %) compare with those calculated (80.0, 8.15, 11.85 % respectively). The mass spectrum was clearly dominated by fragments resulting from a symmetrical dissociation although the parent ion (270) and fragments due to loss of methyl or methoxy groups were in evidence.The n.m.r. result and the cryoscopic measurement indicate the dimers in preference to the monomeric ethers C6H5CH(OCH3)CH3. Since the main wave was relatively unaffected by the presence of water it is reasonable to suppose that it corresponds to the formation of acetophenone pinacols.’ The prewave observed for acetophenone in acidified anhydrous methanol possesses all of the characteristics of a kinetic wave, i.e. the limiting current is determined by the rate of a homogeneous chemical reaction preceding the charge transfer step.M. D. BHATTI AND 0. R. BROWN 1 1 1 The electrochemical kinetic data obtained for the foot of the prewave indicate that the reaction mechanism is of the type already reported for the hydromerisation of acetophenone.A Tafel slope of 1.5 RT/Fand a reaction order of 1.5 are characteristic of a mechanism in which rate control is by an irreversible homogeneous bimolecular step involving species produced in an initial reversible one electron electrochemical event. However, in the present system the reaction product is not the pinacol but instead the corresponding diether. Arguing by analogy with the pinacolisation reaction it is reasonable to consider that the present reaction is a reduction of the x 2 Phc(0Me)Me + e- + Phc(0Me)Me -+ \ Me OMe / It is necessary therefore to explain the occurrence of the phenyl methoxy methyl carbonium ion in this system.It is well known that acetals can be prepared by the reaction between aldehydes and alcohols in the presence of dry HCl. Ketals too, although not usually isolated from such reactions, have been shown to form in equilibria involving the ketone and alcohol under anhydrous conditions. As acetal formations are acid-catalysed one can propose the usual type of mechanism : OH Me + H+ + R1R2C(OMe)OH MeOH / R1R2CO+H+ + (R1R2COH)+ + R1R2C - MeOH Hf + R'R2C(OMe)OH + R1R2C(OMe)OH2 - H+ \ / hemi ke t a1 0' H + (R1R2COMe)+ + \ - H20 MeOH - MeOH H2O OMe / R ~ R ~ C - H + H+ + R1R2C(OMe)2. \+/ ketal 0 Me \ SCHEME 1 The ketal of acetophenone is known; it has been prepared from acetophenone in methanol by the action of either formimido methyl ether or methyl orthoformate.6 There do not appear to have been any studies of the equilibrium between aceto- phenone and methanol although U.V.absorption has been used often to study ketone/ ketal equilibria.'112 CATHODIC REDUCTION OF ACETOPHENONE The linear plot of fig, 4 is consistent with the assumptions that (a) the equilibrium is set up rapidly in acid solution, (b) the amounts of hemiketal and other intermediates in scheme 1 are negligible, and (c) the ketal has an insignificant extinction coefficient at 242 nm. From the slope we deduce that the equilibrium constant at 22°C is PhCOCH3 + 2CH3OH + PhC(OCH3)2CH3 + H20 K = c ~ ~ ~ ~ ~ c ~ ~ ~ / c ~ = 3 x mol dm-3 and that the initial water content of the solution was slightly in excess of 0.1 volume % (the methanol manufacturer’s specified limit).-I 0 I 2 3 1, 5 6 7 6 lo3 x vol H20/vol solution FIG. 4.-Variation of the relative decrease in absorbance of acetophenone in the presence of acid with the concentration of added water. From the electrode kinetic results it is not possible to conclude whether the chemical process which limits the currents on the first wave is the formation of phenyl methoxy methyl carbonium ion from the ketal, or from the acetophenone or by a mixture of both routes (scheme 1). The effect of added water on the kinetics of reductions in the Tafel region is readily understood in terms of the pre-equilibrium + PhCOCH3 + CH30H +HI- + PhC(OCH3)CH3 + H20. It might be expected that the reaction order with respect to added water should be - 1.5 according to the theory How- ever, the solvent contains an unknown residual trace of water.Thus if the data of fig. 2 are plotted as log c~~~ against E at constant current i a linear relation is not obtained. However, if we guess values of the total residual water c,,, and plot log ( C ~ ~ ~ + C ~ , , ) against E until a linear relation is obtained then the slope of the plot should equal 2.303 RT/F. This is found to apply in fig. 5 where the potentials corresponding to a current density of 17.7 pA cm-2 are plotted against log cHz0 and against log ( c ~ ~ ~ + c,) with cres corresponding to 0.2 vol % water. This value is almost double that obtained from the U.V. measurements (fig. 4) but the methanol had less opportunity to pick up extraneous water in those experiments. Deviations from of reactions of the type described by eqn (1).\ \ OCH341430H CH3 CH3 / PhC-CH3 0.6 0.65 0.7 0.75 - E/V FIG.5.-Analysis of the potential shift with water concentration for current density = 17.7pA ciir2 from the data in fig. 2. 0, zero initial water ; 0, 2 parts per thousand water by volume. \ / OCH3 CH3 Conway and Rudd unfortunately made their measurements at lower current densities where, as fig. 3 shows, the polarisation data are distorted by the prewave so that in the region of - 1180 mV the reaction order with respect to acetophenone is unity and the Tafel plots are in fact gentle curves with a mean slope of some 52 mV. Those workers believed the pinacol formation was quantitative in that system but in fact their kinetic measurements were made under conditions where it is now evident that one reaction gradually gave way to another. Each electrode reaction separately follows the kinetic scheme first discussed by Koutecky and Hanus but in the region of mixed kinetics such behaviour is not observed owing to the influence of the preceding chemical reaction which limits the first wave.114 CATHODIC REDUCTION OF ACETOPHENONE The current yield of 2,3-dimethoxy-2,3-diphenylbutane obtained from methanolic acetophenone (0.3 mol dm-3) containing HCl (1 .O mol dm-3) at - 0.68 V was 23 %. This unexpectedly low value is undoubtedly due to the fall in the current for this reaction owing to the water it produces, and the consequent relative increase in the pinacol formation current. Other products were found to include the pinacols and there were also other gas chromatographic peaks which were probably due to the mixed dimers (mono ethers). Further work is in progress to investigate the synthetic scope of this type of reaction. M. D. B. thanks the British Council for a studentship. M. P. J. Brennan and 0. R. Brown, J.C.S. Faruday I, 1973, 69, 132. E. J. Rudd and B. E. Conway, Trans. Faraday Soc., 1971,67,440. 0. R. Brown, Disc. Faraday Sac., 1968, 45, 126. 1964), p. 296. L. Claisen, Ber., 1898, 31, 1010. M. T. Bogart and P.P. Herrera, J. Amer. Chem. SOC., 1923,45, 238. ' 0. H. Wheeler, J. Amer. Chem. SOC., 1957, 79, 4191. a J. Koutecky and V. Hanus, Coll. Czech. Chem. Comm., 1955, 20, 124. 4B. J. Cram and J. S. Hammond, Organic Chemistry (McGraw-Hill, New York, 2nd edn.,