摘要:

Reductive Coupling and Isomerization of Electrogenerated Radical Ions of cis- and trans-Isomers BY ALLEN J. BARD, VINCENT J. PUGLISI, JOHN v. KENKEL AND ANN LGMAX Department of Chemistry, The University of Texas at Austin, Austin, Texas 78712 U.S.A. Received 5th June, 1973 The electrochemical behaviour of the cis-trans-pairs : diethyl maleate (DEM), diethyl fumarate (DEF) and cis- and trans-dibenzoylethylene (CDBE and TDBE) in N,N-dimethylformamide solutions was investigated using cyclic and rotating ring-disc electrode (RRDE) voltammetry and bulk electro- lysis. In both cases the cis-isomer is reduced at mme negative potentials than the trans-isomer. Although the radical anion formed from the trans-isomer is fairly stable, that of the cis-isomer undergoes a rapid isomerization to form the trans-radical anion and also reacts more rapidly in either self- or cross-coupling reactions.The electron spin resonance (e.s.r.) spectra obtained by electrolytic reduction of either DEM or DEF has been re-interpreted as being caused by different conformations of the trans-radical anion. Identical e.s.r. spectra are also obtained from reduction of CDBE and TDBE ; this is attributed to a single conformation of TDBE radical anion. There has been much interest in the electrochemical behaviour of activated olefins (R) / \ A3 A4 0 II II Al = A4 = H, A2 = A3 = -C-OEt diethyl fumarate (DEF), 0 Al = A, = H, A3 = A4 = -C-OEt diethyl maleate (DEM), A, = A2 = A3 = H, A4 = -CN acrylonitrile (AN), Al = A4 = H, A, = A3 = -C-Ph trans-dibenzoylethylene (TDBE), A, = A, = H, A3 = A4 = -C-Ph cis-dibenzoylethylene (CDBE), 0 0 especially because these molecules undergo hydrodimerization (or reductive coupling) upon electroreduction, leading to the formation of R2H2., Previous studies of the mechanism of these reactions in dimethylformamide (DMF) solutions by rotating ring-disc electrode (RRDE) voltammetry, cyclic voltammetry, double potential-step chronocoulonietry and electron spin resonance (e.s.r.) spectroscopy on molecules such as DEF, cinnamonitrile and fumaronitrile, have shown that the reaction path 56-M 353354 REDUCTIVE COUPLING A N D ISOMERIZATION involves the initial formation of the radical anion (R-) followed by coupling (I) and protonation 2-5 : Of particular interest is the difference of behaviour of cis- and trans-isomers upon electroreduction to the radical ions with respect to isomerization, coupling and mixed coupling (where the radical anion of one olefin reacts with the parent molecule of a second olefin).The study of the comparative behaviour of the cis-trans-pair, diethyl maleate (DEM) and diethyl fumarate (DEF), is of particular interest, since the radical anions derived from these have often been assumed to behave similarly. Nelsen produced the radical anions of DEF and DEM by electroreduction in DMSO and showed that the same e.s.r. spectrum was obtained from both and that it consisted of signals from two separate species. He concluded, by comparison with the behaviour of diethyl phthalate, that these were the cis- and trans-anion radicals. However, electrochemical studies in these Iab~ratories,~~ as well as those of Il’yasov and Kargin and co-workers We report here a detailed study of these species which provide a striking example of the strong influence of structure of electrogenerated intermediates on their behaviour.We also discuss the electrochemistry and e.s.r. spectra of trans- and cis-dibenzoylethylene (TDBE and CDBE) radical anions. 2RT+R;- (1) demonstrated rapid conversion of DEM- to DEFT. EXPERIMENTAL Experimental apparatus and techniques followed previous ~tudies.~ A Tacussel Electron- ique Bipotentiostat, model Bipad 2, was used for all RRDE experiments. A Digitec digital voltmeter, model 204, and a Fairchild digital multimeter, model 7050, were used to measure the steady-state ring and disc currents simultaneously.A Wavetek function generator provided a d.c. potential ramp for voltammetric experiments recorded on either a Mosley model 2D-2 or Hewlett-Packard Autograf X-Y recorder. The Pt-Teflon RRDE had dimensions rl = 0.187 cm, r2 = 0.200 cm and r3 = 0.332 cm and was constructed by Pine Instrument Company. The maximum collection efficiency N for this RRDE is 0.5X3 Electrolysis for the purpose of removing electroactive impurities was performed at a mercury pool electrode in the RRDE cell. The reference and auxiliary electrodes were silver (Ag, R.E.) and platinum spirals, respectively, contained in compartments separated from the working electrode compartment by medium-porosity glass frits. Bulk electrolysis performed to study the isomerization process was carried out in a conventional coulometry cell utilizing a platinum-gauze working electrode, and platinum and S.C.E.auxiliary and reference electrodes, respectively. A model 170 Electrochemistry System (Princeton Applied Research Corp., Princeton, N.J.) was employed for all cyclic voltammetric and d.c. and a.c. polaro- graphic experiments. The dual electrode flow cell used porous silver disc working electrodes (Selas-Flotronics Co.) about 50 pm thick with 3 or 5 pm average pore size. They were separated by a 200 pni thick Teflon screen cloth with a 0.2 cm2 hole. Each disc working electrode was controlled with respect to its own Pt auxiliary electrode and S.C.E. reference electrode. The cell used in e.s.r. experiments was that designed for simultaneous electrochemical- e.s.r.(SEESR) work, and comprises a Pt working electrode, tungsten auxiliary electrode and silver wire reference electrode.8 A Varian model V-4502 e.s.r. spectrometer was employed. In preparing solutions, DMF, dispensed from the storage vessel into the dispensing, vessel by positive He pressure, together with supporting electrolyte and the depolarizer were subjected to a minimum of three freeze-pump-thaw cycles. The solution was then transferred to the electrochemical cells. However, in the experiment where CDBE was analyzed in the dark, the CDBE was not carried through the freeze-pump-thaw cycles, but was added directly to the contents of the electrochemical cell. In every case, acrylonitrile was added directly to the electrochemical cell.A .J . B A R D , V. J . PUGLISI, J . V. KENKEL AND A . LOMAX 355 N,N-dimethylformamide (DMF), obtained from Baker Chemical Co. or Matheson, Coleman and Bell, was purified by vacuum distillation and stored under He. Tetra-n- butylammonium iodide (TBAI) and fluoborate (TBAF), obtained from Southwestern Analytical Chemicals, Inc., were vacuum dried and stored over Drierite. DEF, obtained from K. and K. Laboratories, was sublimed prior to use. DEM, obtained from Matheson, Coleman and Bell, contained about 6 % DEF as an impurity. The DEM was purified using preparative gas chromatography using a 6 m chromatographic column, 30 % S.E. 30 on 60180 Chromosorb P and a column temperature of 175°C. TDBE, purchased from Aldrich Chemical Co., exhibited a m.p. of 108°C and was used as received.CDBE was prepared from the trans-compound by photo-isomerization. The transformation was verified by n.m.r. analysis. The resultant cis product was in the form of white and brown needles having a m.p. of 132-134°C. Sublimation yielded white needles which had the same melting point. Acrylonitrile, purchased from Matheson, Coleman and Bell Chemicals, was used as received. RESULTS THE DEFi-DEM SYSTEM of the electroreduction of the trans-isomer, DEF, in DMF solutions, have shown that it undergoes a reversible one-electron reduction to the radical anion (fig. la, Ep = - 1.41 V us S.C.E.) followed by a coupling reaction, with a rate constant of about 34 M-I s-I. DEF+e+DEF- (2) Previous studies 2* k 2 2 DEFT +(DEF) i- (3) The cyclic voltammetry of the cis-isomer, DEM, at scan rates below 5 V s-I, however, shows a somewhat broader reduction peak, with ED = --1.61 V, and with no anodic i (a) EIV L’S S.C.E.- 1 . 7 ’ -1.6 -1.8 FIG. 1.-Cyclic voltammograms of (a) diethyl fumarate (scan rate 50 mV s-’) and (b) 3 mM dierhyi maleate (scan rate 5 V s-I) in 0.1 M TBAF-DMF. 56-M*356 REDUCTIVE COUPLING AND ISOMERIZATIOK peak corresponding to this cathodic peak (fig. lb). An anodic peak is observed at the same potentials where the oxidation of DEF- occurs. At a scan rate of 50 V s-' a small anodic peak is observed where DEM- oxidation is expected. The current function for the reduction wave, ip/u*C, where ip is the peak current, u is the scan rate, and C the concentration, is 63 FA mM-l V* s-*, slightly smaller than the current functions of reversible one-electron processes of similar molecules found with this electrode, of 80 to 95.A.c. polarography of DEM (fig. 2a) shows a broad peak at FIG. 2.-A.c. polarograms of (a) 1.2 mM dietkyl maleate solution ; (b) effluent from u a l electrode flow cell for 1.2 mM DEM inlet solution at flow rate of 0.22 ml min-' ; (c) effluent from dual electrode flow cell for 0.55 mM DEM inlet solution and flow rate ca. 0.1 ml min-', a.c., 100 Hz at 10 mV potentials corresponding to the cyclic voltammetric reduction wave. The small pre-peak observed in both the cyclic voltammetry and a.c. polarography represents a small amount of DEF impurity contained in the DEM. These results can be ex- plained by assuming an initial reversible formation of DEMi followed by rapid isomerization to DEFT, DEM+e$DEM- (4) k t DEM-+DEFT The rate constant of this isomerization is about 10 s-I.A.J . B A R D , V. J . PUGLISI, J . V . KENKEL AND A . LOMAX 357 To demonstrate unequivocally that reduction of both DEF and DEM leads only to DEFT, experiments were undertaken with a dual-electrode flow cell. This system, shown in fig. 3, consists of two closely-spaced porous electrodes with the potential . Pt A ux i I iary El act r odes Sintered glass disks FIG. 3.-Dual electrode flow cell. of each electrode being controlled with respect to separate reference electrodes and with solution flowing through the electrodes from one electrode to the other. A complete description of the construction and calibration of this apparatus is avail- able.9 The experiment undertaken involves electrolysis of a substance at the upper (inlet) electrode, followed by flow of the product to the lower (outlet) electrode where a second electrolysis is carried out.The efficiency of the electrolysis, i.e., the fraction of substance electrolyzed, and the transit time between the electrodes, is governed by the flow rate of solution. The first experiment using the flow cell involved a 1.2 mM solution of DEM in DMF containing 0.1 M TBAF flowing first through the inlet electrode held at a potential where DEM reduction occurred and then to the lower electrode held at a potential where any radical anions would be oxidized, at a flow rate of 0.22 ml min-I (interelectrode transit time of 1.4 s). The porous electrodes were shown to carry out electrolysis with greater than 99 % efficiency for flow rates below 2mlmin-I.Analysis of the effluent solution by a.c. polarography (fig. 2b) showed the presence of only DEF. A similar experiment at a flow rate of less than 0.1 ml min-I gave similar results (fig. 2c). When a similar experiment was performed358 REDUCTIVE COUPLING A N D ISOMERIZATION employing a 1.4 mM solution of DEF with reduction to DEFT at the inlet electrode and oxidation at the lower electrode, again, only DEF was found in the effluent. These results show that the reduction of either DEM or DEF leads to a solution containing only DEFT and that the different species observed in e.s.r. must be different conforiners of the trans-radical anion, EtO p..;to20* - 0 2 0 - OEt -0 -0 The different reduction potentials for DEM and DEF and the rapid isomerization of DEM- leads to interesting behaviour on attempted removal of the small amount of DEF from the DEM sample by pre-electrolysis.When a sample of DEM containing DEF was electrolyzed at a platinum gauze electrode at - 1.425 V vs S.C.E., a potential corresponding to the mass transfer liniiting current plateau of DEF but positive potentials for DEM reduction for 10 min, the cyclic voltammogram of the system shows a relative increase in the height of the DEF wave (fig. 4). Continuedelectrolysis . - - t a , . I . I I . I ' I. * I 1. : ' ,, . I .I EIV us S.C.E. FIG. 4.-Cyclic voltammograms at 200 mV s-' on a Pt working electrode, (a) diethyl maleate (4 mMj small (5-8 %) diethyl fumarate impurity in TBAP (0.15 M)-DMF initially and following, (b) 10 and (c) 45 min of bulk electrolysis at - 1.425 V against S.C.E.at this potential for an additional 35 min causes a further increase in the relative height of the DEF wave (fig. 4c). This apparently anomalous behaviour can be attributed to the reaction of DEFT with DEM, Although the extent of this reaction is slight, the irreversible rapid isomerization of DEFT +DEM+DEF+DEM- (6)A . J. B A R D , V. J . PUGLISI, J . V. KENKEL AND A . LOMAX 3 59 DEM- occurs, and the net effect of reactions (6) and (5) is the conversion of DEM to DEF catalyzed by DEFT. RRDE studies on DEM were undertaken to compare with those of DEF.3 A plot of the disc current (id) against disc potential (Ed) generally follows the behaviour expected from CV studies, i.e., a small impurity wave for DEF at about -0.7 V zis Ag R.E.and a reduction wave for DEM with Ek at - 1.13 V us Ag R.E. As with disc voltammograms reported for DEF3 a current depression is observed on the reduction plateau at potentials corresponding to those where di-anion formation is expected. The plot of ring current, i,, us Ed with the ring potential E, held at -0.5 V us Ag R.E. shows an anodic current where production of DEFT and DEM- occurs ; i, decreases where the dip in id commences. These current dips have been attributed to polymer- ization initiated by generation of the dianion. The dependence of i, on E, at different values of E d is shown in fig. 5. Curve a, taken where Ed = 0 V, shows the current - 6 --- I 4 0 p .A .i 1 , - - - - - - - - . - - - - C E/V rs Ag R.E. FIG. 5.-RRDE voltammograms of diethyl maleate (3 mM) with small (5-8 %) diethyl fumarate impurity in TBAP (0.15 M ) DMF at a rotation rate of 98.6 s-' and (0) Ed = 0 V, (6) Ed = -0.9 V, (c) Ed = - 1.4 v L'S Ag R.E. caused by impurity DEF (i,)a followed by the main DEM reduction wave. When Ed is held at -0.9 V 17s Ag R.E., corresponding to DEF reduction at the disc, the ring voltammogram shows collection of DEFT, (ir)b, and shielding of the ring in DEF reduction (curve b). When E d = - 1.4 V rs Ag R.E., corresponding to DEF and DEM reduction at the disc, (curve c) the ring current, (i,)c, shows a large increase at potentials where DEFT is oxidized, but only a very slight decrea.se in the potential region where oxidation of any DEMT that reached the ring would occur ; oxidation of DEM- in this region (-0.9 to - 1.0 V rs Ag R.E.) would cause the total iR to become less cathodic or even become anodic depending upon the relative fluxes of DEF and DEM- to the ring.The observed behaviour again demonstrates the rapid isomerization of DEM T. An estimate of the relative rate of coupling of DEFT and360 REDUCTIVE COUPLING AND ISOMERIZATION DEM- can be obtained by noting the collection efficiency (NK = li,/idl) for DEF and DEM. For these concentrations and rotation rates for this RRDE, NK for a solution of DEF is 0.55If:0.01, essentially the value observed for collection of a stable species at this RRDE, and indicating negligible reaction of DEFT on transit from disc to ring.For a DEM solution, however, NK varied between 0.25 and 0.35 depending upon rotation rate and concentration. Similarly, for an equimolar mixture of DEF and DEM the collection efficiencies determined from i, with E, at potentials for DEF- oxidation and id for DEF and DEM reduction are 0.45 and 0.36 respectively. These results show that the DEM- undergoes reaction much more rapidly than DEF- does, and this reaction competes with the isomerization of DEM-. Although coupling of DEM- appears to be a possible reaction, determination of the rate and mechanism of reactions consuming DEM- requires further study. The more rapid rate of reaction of DEM- to non-electroactive products before isomerization is also shown in the dual electrode flow-cell experiments. In the experiments involving an initial solution of 1.4 mM DEF, the effluent following reducti'on and oxidation contained about one- half of the original amount of DEF.However, for experiments involving an initial solution of 1.2 mM DEM only about 5 % of the original DEM was recovered as DEF following passage through the two electrodes, and this amount was even smaller when the flow rate was decreased. Another experiment which illustrates the difference in reactivity of the trans- and cis-anion radicals is that involving the cross-coupling with acrylonitrile (AN). 3b Cyclic and RRDE voltammetry studies of the formation and reaction of DEFT shows that addition of up to 0.2 M AN (which is reduced at more negative potentials) EiV cs S.C.E. FIG. 6.-Cyclic voltammograms at 200 mV s-' at a Pt electrode, (a) diethyl maleate (3.9 mM), TBAP (0.15 M), DMF : (6) DEM (3.9 mM) after addition of acrylonitrile (0.2 M) : (c) dimethyl fumarate (3.7 mM) in the presence of DEM and AN.A .J . B A R D , V. J . PUGLISI, J . V. K E N K E L A N D A . LOMAX 36 Z causes little change in behaviour, indicating little or no reaction between AN and DEF- under these conditions. The cyclic voltammetry of DEM in the absence and presence of AN is shown in fig. 6. The addition of AN causes disappearance of the DEF- oxidation wave, showing that the DEM- reacts with AN (presumably in an ECE reaction sequence leading to cross-coupled product) in a reaction which is fast compared to the rate of isomerization. *- T H E T R A N S - A N D CIS-l,2-DIBENZOYLETHYLENE SYSTEM The cyclic voltammogram of TDBE in 0.1 M TBAI-DMF is shown in fig.7. Two reduction waves with EPIC = -0.28 and Epzc = -0.88 V us Ag R.E. are I 0 -, I \ . -. b I I I I 4 I I 8 / 8 8 / 8 8 FIG, 7.-trans-l,2-Dibenzoylethylene (0.5 mM), TBAI (0.15 M), DMF cyclic voltammograms at 200 mV s-l. observed ; the very small wave near -0.65 V is caused by some cis-compound impurity. The current function for the first wave, i,,,/u%, is 85 pA mM-l V* s-*, characteristic of a one-electron reduction. The first reduction wave shows a reversal anodic wave, with ipa/ipc = 1. When the voltage scan is continued. past the second reduction wave and then reversed, another anodic wave at +0.27 V appears. This wave is not present when scans to negative potentials only encompass the first reduc- tion wave, even when the potential is held at potentials past the peak of the first wave for 10 s before reversal.The RRDE voltammogram shows two reduction waves in the plot of id us Ed with E+ values of -0.22 and -0.88 V vs Ag R.E. The first wave shows an id/o*C of 5.4pAmM-ls-* and the limiting current of the second wave is about one-half that of the first. A scan of i, us Ed with E, = +0.2 V us Ag R.E. shows a wave corresponding to the oxidation of the species produced at the first reduction wave with N = 0.57. This N-values is slightly higher than the value for a stable disc-produced species (0.55) because of a small amount of oxidizible impurity contained in the TDBE sample. At Ed-values corresponding to the second reduction362 REDUCTIVE COUPLING AND ISOMERIZATION wave, i, decreases to about two-thirds of its value at the first wave.For E, held at 0.0 V, a larger drop in if is observed, since the product of the second reduction wave would not be oxidized. The results can be interpreted as reduction of TDBE to a stable anion radical with the reduction product in the second wave, probably the di-anion, decomposing rapidly, at least partially to the form the species oxidized at +0.27 V. Attempts to study the cis-isomer, CDBE, with the usual methods of sample preparation in the light resulted in conversion of a large fraction of the stsrting mzterial to the trans-isomer. For this reason, samples, once sublimed, were stored and studied in the dark. Cyclic voltain~nograms of CDBE are shown in fig.8. TDBE+e+TDBE- (7) EIV us Ag R.E. FIG. 8.-cis-1,2-Dibenzoylethylene (1.2 mM), TDBE (8-12 %), TBAI (0.15 M), DMF cyclic voltam- mograms at 200 mV s-l. The main reduction waves occur at -0.61 and -0.91 V us Ag R.E. The small waves at -0.27 and -0.84V can be attributed to some TDBE in the sample. The peak current function of the first reduction wave was 70 pA mM-I V* s-3, somewhat smaller than that of the trans-isomer. Neither of the CDBE reduction waves shows an associated oxidation wave on reversal ; however, the wave attributed to oxidation of TDBE- is enhanced when the potential is scanned over the first reduction wave. An additional oxidation wave appears at about +0.28 V, similar to that observed in the voltammogram of TDBE. In this case, however, this wave appears when the scans to negative potentials only encompass the first reduction wave, although this small oxidation wave is larger when scans are made to - 1 .O V, past the second reduc- tion wave.The description of the (id, Ed) curve follows the cyclic voltammetric behaviour. The curves of i, us E d show anodic currents for oxidation of the trans-isomer radical anion. In this case the collection efficiencies were smaller than for TDBE, varying from 0.34 to 0.47 depend- ing upon rotation rate and concentration. A plot of i, us E, for different values of Ed is very similar to that of fig. 5, and can be interpreted in a similar manner. Some RRDE voltammograms for CDBE are shown in fig. 9.A . J. B A R D , V . J . PUGLISI, J . V . KENKEL AND A . LOMAX 363 RRDE data was obtained for mixtures of TDBE and CDBE.For a solution con- taining 0.42mM TDBE and 0.21 mM CDBE, values of NTDBE = = 0 V)/ id(Ed = -0.45 V)], (i.e,, for TDBE reduction and oxidation) were between 0.52 and 0.57 and close to those for TDBE in the absence of CDBE. For a solution containing EIV t's Ag R.E.) FIG. 9.-RRDE voltammograms of cis-dibenzoylethylene (1.2 mM) containing a small amount (8-12 %) trans-dibenzoylethylene, TBAI (0.15 M), DMF at 98.6 s-', (a) id 1's Ed, (b) ir us Ed with Er = 0 V. 2.06 mM TDBE and 0.70 mM CDBE, NTDBE varied between 0.28 and 0.40 for an o of 202 to 27.6 s-l. When bulk electrolysis of a solution containing both isomers at -0.45 V us Ag R.E. (where only TDBE is reducible) was carried out, cyclic voltammo- grams showed that the reduction waves of both the trans- and cis- compounds de- creased in magnitude after electrolysis.The behaviour of these isomers parallels that of the fumarate-maleate pair, and the reduction of CDBE can be described as CDBE+e$CDBE- CDBE- +TDBE- with the CDBE- also undergoing reactions such as protonation, or perhaps coupling, at a rate which is fast compared to the isomerization rate. Again, the reactivity of CDBE- is much greater than that of TDBE-. Attempts to study cross-coupling of these radical anions with AN showed little difference between the isomers. Addi- tion of AN did not perturb appreciably the cyclic voltammetric behaviour of either TDBE or CDBE, although the cis-isomer showed pronounced filming in the presence of AN while the trans-isomer did not.E . S . R . OF THE DIBENZOYLETHYLENES Intra inwos electrolytic reduction of TDBE and CEBE in DMF at a platinum electrode was undertaken to compare the behaviour of these with that of the fumarate364 REDUCTIVE COUPLING A N D 1SOMERIZATION and maleate. fl’yasov and co-workers reported that both TDBE and CDBE produce the same e.s.r. spectrum on reduction and showed a poorly-resolved spectrum obtained at -20°C. Winecoff, O’Brien and Boykin lo also mention a stable e.s.r. spectrum from TDBE. The e.s.r. spectrum obtained by electrolytic reduction of TDBE in 0.1 M TBAI-DMF in a non-flowing solution is shown in fig. lOa, while that 2.5 G I C 2.5 G e----.l I D . I ’ FIG. 10.-E.s.r. spectra of (a) radical anion generated from solution containing 3.1 mM TDBE, 0.1 M TBAI, DMF ; (b) simulated spectrum using coupling constants in table 1 and linewidth of 0.1 1 G ; (c) radica1 anion generated from solution containing 2.0 mM CDBE (+ some TDBE), 0.1 M TBAI, DMF ; (d) simulated spectrum as in (b), except line width was 0.08 G.from a solution of CDBE (containing an appreciable amount of TDBE impurity) is shown in fig. 1Oc. The spectra are essentially the same, although that of the cis isomer shows a slightly narrower line width, and gives a less intense spectrum when electrolyzed in a similar manner. The simulated spectra shown in fig. 10b and d were obtained with the coupling constants shown in table 1 and line widths of 0.1 1 and 0.08 gauss, respectively. The assignment of the coupling constants to positions in the molecule was made on the basis of McLachlan molecular orbital calculations using the parameters : 3, = 1.2 ; a, = a, + 1.4 ,6, ; BCO = 1.65 pCc.l1 For this molecule only a single conformation is found for the radical anion; the electrochemical show that this is the trans- form and probably, from steric considerations conformation, results in the 0.A .J . B A R D , V . J . PUGLISI, J . V . KENKEL AND A . LOMAX 365 TABLE 1 .-COUPLING CONSTANTS OF 172-DIBENZOYLETHYLENE ANION RADICAL * no. of equivalent type of nuclei nuclei aH(L) assignment calculated * ( G ) H 2 4.55 ethylenic 4.59 H 2 0.84 para 0.96 H 4 0.74 ortho 0.92 H 4 0.25 meta - 0.35 11 The solution was DMF containing 0.1 M TBAI and either 3.1 mM TDBE or 2.0 m M CDBE+ The SEESR cell used a Pt working electrode, W counter electrode and Ag reference elect- TDBE.rode. h From McLachlan spin densities and Q = 24.0 G. DISCUSSION The results show that for these compounds the trans-form is more easily reduced than the cis-form. The measured difference in reduction potential for DEF and DEM is about 0.20 V at a platinum electrode from cyclic voftammetry, or 0.23 V from d.c. polarography, (compared to the difference of 0.15 V lound for the methyl esters by Il'yanov et aL7" by d.c. polarography). For TDBE and CDBE this differ- ence from our results is 0.33 V (compared to 0.24 V found from d.c. polarography lo and 0.27 V found by cyclic voltammetry at a mercury electrode lo). The more difficult reduction of the cis-isomer can be ascribed to the repulsion of the carbonyl groups when a negative charge is added to the molecule making this configuration energetically less favourable.This difference in energy can be estimated by assuming a rate constant for isomerization of about 100 s-l, which when coupled with the scan rate employed in the cyclic voltammetric experiments (e.g., 0.2 V s-l) and the equation for the positive shift in potential for the wave for an electrode reaction with a fast following reaction,'* yields a difference in reversible potentials of about 0.28 V for DEM and DEF. For stilbene, where this electrostatic repulsion of the carbonyl groups is absent, the E,-values of the trans- and cis-forms in DMF are almost identi- cal, - 2.15 and - 2.18 V, respectively.' The addition of the electron also decreases the bond order of the central ethylenic bond, which, when coupled with the carbonyl group repulsion, promotes rapid isomerization of the cis-form.This isomerization also occurs in stilbene, promoted by steric effects, but the rate is slower. For example, Dietz and Peover l 3 give the lifetime of the cis-stilbene radical anion as greater than I5 s. (In the presence of alkali metal ion the isomerization rate of cis-stilbene is apparently much faster, however.14) Finally, the cis-forms (DEM and CDBE) are more reactive than the trans-forms with respect to coupling or addition reactions. This can be rationalized as being caused by greater electron density residing on the central ethylenic carbons of the cis-radical ions because of the repulsion of the carbonyl groups or partial twisting around the central ethylenic bond.This effect has been noted before, for reaction of stilbene radical ions with water,13 but the difference in reactivity was not as striking as in the cases observed here. Finally, the cross electron transfer reactions, such as (6), explain how small amounts of radical ion catalyze the cis-trans conversion. These results are also consistent with those previously found for the oxidation of cis- and trans- 1,2-bis(p-N,N-dirnethylaminopyenyl)- 1 ,Zdiphenylethylene. For oxidation of either form to the stable di-cation, followed by reduction, a mixture of cis- and trans- parent results. In this case, oxidation weakens the central ethylenic bond, and steric effects cause rotation about it. Photo-isomerization has been a366 REDUCTIVE COUPLING AND ISOMERIZATION subject of interest for many years.interesting paralIel to those studies. Isomerization upon electron transfer provides an The support of this research by the National Science Foundation (GP-31414X) and We are indebted to the Robert A. Welch Foundation is gratefully acknowledged. Prof. N. L. Bauld for helpful discussions concerning this work. see, e.g., M. M. Baizer, Organic Electrochemistry (M. Dekker, N.Y., 1973). ’ W. V. Childs, J. T. Maloy, C. P. Keszthelyi and A. J. Bard, J. Electrochem. SOC., 1971,118,874. (a) V. J. Puglisi and A. J. Bard, J. Electrochem. SOC., 1972, 119, 829, 833 ; (b) 1973, 120, 241. E. Lamy, L. Nadjo, and J. M. Saveant, J. Electroanal. Chem., 1973,42, 189 ; C. P. Andrieux, L. Nadjo and J. M. Saveant, J. Electroanal. Chem., 1973, 42, 223. J . P. Zimmer, J. A. Richards, J. C. Turner and Dennis H. Evans, Anal. Chem., 1971,43, 10oO. S . F. Nelsen, Tetrahedron Letters, 1967, 39, 3795. ’ (a) A. V. Il’yasov, Yu. M. Kargin and V. Z. Kondranina, Zzv. Akad. Nauk S.S.S.R., ser. Khim., 1971, 5 , 927; (b) A. V. Il’yasov, Yu. M. Kargin, N. N. Sotnikova, V. Z. Kondranina, B. V. Mel’nikov and A. A. Vafina, Zzv. Akad. Nauk S.S.S.R., ser. Khim., 1971, 5, 932. I. B. Goldberg and A. J. Bard, J. Phys. Chem., 1971, 75, 3281. J. V. Kenkel, M.S. Thesis (The University of Texas, 1972); J. V. Kenkel and A. J. Bard, J. Electroanal. Chem., in preparation. J. Chaudhuri, R. F. Adams and M. Szwarc, J. Amer. Chem. Soc., 1971,93, 5617; G. Vincow and G. K. Fraenkel, J. Chem. Phys., 1961,34, 1333 ; R. D. Allendoerfer, J. Mag. Res., 1973,9, 140 ; A. D. McLachlan, Mol. Phys., 1960, 3, 233. l o W. F. Winecoff, F. L. O’Brien and D. W. Boykin, Jr., J. 01.9. Chem., 1973,38, 1474. ’’ J. M. Saveant and E. Vianello, Electrochim. Acta, 1967, 12, 629. l 3 R. Dietz and M. E. Peover, Disc. Faraday SOC., 1968, 45, 154. j 4 R. Chang and J. H. Markgraf, Chem. Phys. Letters, 1972, 13, 575. l 5 A. J. Bard, Pure Appl. Chem., 1971,25, 379.

ISSN:0301-7249    

DOI:10.1039/DC9735600353

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

GENERAL DISCUSSION Prof. A. R. Despik (Belgrade) said : In connection with the use of the ring-disc or double-ring rotating electrodes in kinetic investigations, I wish to report an attempt which may make the preparation of such rotating systems for mass transfer studies technically much easier than has been the case so far. MitroviC et al. reported recently * a solution of the mass-transfer problem to a patch electrode of any shape on a rotating disc. It can be shown that the solution is particularly simple when the patch has the form of a quadrangle placed perpendicularly to the streaming lines. We have extended that case to a sandwich of two quadrangle metal strips separated by a very thin insulating layer which can act as a two-electrode set (generation and detection electrodes).The sandwich can be nailed into a Teflon cylinder or cast into epoxy resin and cut and polished to give a smooth disc surface. The solution of the mass-transfer problem for such a case and the experimental veri- fication will be published shortly. This simple preparation has a particular advantage for the studies in electro- catalysis, when discs (or rings) of the desired metal are not commercially available. Prof. I. Epelboin(Paris) said : Nekrasov writes, on the second page of his paper, that it is easy to establish the valency of the metal ions which are produced as intermediates in the anodic dissolution of the metal. This is true for the case of a metal such as copper. But, as I had already reported in the C.I.T.C.E. meeting held in Moscow in 1963,' this is impossible for oxidisable metals such as Be, Mg, etc., because the theory given by Kabanov are irrele- \.ant, the anodic layer being poor in water. and was discussed during the Gordon Conference on Corrosion, held in New London (USA), in July 1972.I should appreciate the view of Albery and Nekrasov on this important question. is wrong and the measurements by Heussler This problem has been further studied Prof. H. Gerischer (Germany) said : The interesting method which Albery has introduced for the ring-disc electrode could analogously be applied to the ring-ring electrode or to the flow channel with two neighbouring electrodes in the wall of the channeL6 Especially the latter arrangement can be constructed with a very close distance between the electrodes and is simple in technical performance.On the first view the narrow area of the generating electrode in the latter techniques seems to be an advantage for measuring short transfer times and for the absolute efficiency of the mass transfer to the indicator electrode. Could you perhaps comment on the advantages and disadvantages of these different arrangements ? Dr. M. Keddam (Paris) said: I would like to make some contribution to the discussion on the paper by Albery. It is about the way the passage of current across * IUPAC meeting, Hamburg, September, 1973. I. Epelboin and M. Froment, J. Clzini. Phys., 1963, 60, 1301-09. B. N. Kabanov, Dokl. Akad. Nauk., 1958, 120, 559. Heussler, Z. Elekrrockeni., 1961, 65, 192. H. Aida, I. Epelboin and M. Garreau, J .Electr.ochem. SOC., 1971, 118, 243. K. E. Heusler and H. Schurig, 2. Phys. Chem. (Frankfurt), 1974. H. Gerischer, I. Mattes and R. Braun, J . Electoroanal. Cliem., 1965, 10, 553. 367368 G E N E R A L DISCUSSION one electrode alters the potential between the other one and the reference. In the first days of 1972 we published a paper showing this physical coupling and we proposed an equation to take it into account in ring-disc measurements. To study this effect, we used the electronic set-up shown in fig. I (part (a)). One electrode, for example the ring, is controlled by galvanostatic regulation, the disc is controlled with a potentiostat. The electrodes are made of iron. First, (part (b)), we plotted the (current, voltage) curve of the ring with zero c-nrrent on the disc (curve A). Then we did the same with a disc-current I d through the disc (curve B).Between curves A and B we observe a constant potential shift 4; which arises from the passage of - . I I to to+200/ls - current through the disc. Then we did the same in the converse situation, and we obtained the current voltage curve C for the disc with a ring current equal to zero and the curve D with a disc ring current i,. Part (c) shows the decay of ring potential after switching off the disc current (ohmic drop) and the ring current (ohmic drop and electrochemical decay). The ohmic contributions to the potentials measured between ring Vh or disc Vd and the reference electrode are conveniently represented by a linear combination VL = R:l,+ R:Id vt = RiI,+R,dId.We verified for several ring-disc electrodes of different sizes that the resistances R$ depend only on the geometry of the cell and electrode system and for a given geometry are proportional to the resistivity of the electrolyte. Moreover in all cases R; = Rf within experimental precision. C . Gabrielli, M. Keddam and H. Takenouti, J. Chim. P/IYs., 1972, 737.369 Prof. S . Bruckenstein (Buflalu, U.S.A.) said : Shabrang and I ' 9 have found that the equivalent circuit in fig. 1 represents the uncompensated ohmic potential drops that occur at a ring-disc electrode. R,,, is the cell resistance between the auxiliary electrode and the tip of the Luggin capillary, Rc represents a common uncompensated resistance between the Luggin capillary and the ring and the disc electrodes, RD an Auxiljary Electrode GENERAL DISCUSSION 'erence Elect rode rode ohmic resistance associated only with the disc electrode and RR an ohmic resistance associated only with the ring electrode.ED and RR represent the half cell potentials of the disc and ring electrodes, and also contain an impedance parameter associated with the reactions occurring on these electrodes. This equivalent circuit also explains the simple linear relations found by Gabrielli et aL3 Dr, W. J. Albery (Oxford) said : I would like to emphasise that the alternating current and ring-disc electrode experiment gives us unique information about the electrode mechanism. Given the scheme i D disc iF iR 4 electrode 4 transport + reaction the measurement of the ring current iR and its interpretation through the theory of the transport allows the measurement of the flux of material, the Faradaic current, iF.Any model for the disc electrode reaction can be tested against the data for the phase and amplitude of iF with respect to iD. This will be similar to the impedance diagrams presented in Armstrong's paper.4 This technique shows that the ring-disc electrode is not only useful for investigating Class 1 intermediates of Gerischer's classification but also Class 2, namely those that are on the electrode surface. In answer to Epelboin's question I would say that the best technique to investigate valence states is that of the split ring electrode invented by Miller and Visco.6 The reduction of an This can be compared with the flux of electrons, the disc current, iD.M. Shabrang, f. Electrochem. Soc., 1974, 121, 1%. M. Shabrang and S. Bruckenstein, J. Electrochem. SOC., 1974. C. Gabrielli, M. Keddam and H. Takenouti, J. Chim. Phys., 1972,4, 737, R. D. Armstrong, R. E. Firman and H. R. Thirsk, this Discussion. H. Gerischer, this Discussion, Introduction. ' B. Miller and R. E. Visco, f. Electrochem. Soc., 1968, 115,251.370 GENERAL DISCUSSION intermediate on one split ring and at the same time its oxidation on the other means that one does not have to worry whether the disc is behaving reproducibly or not. In answer to Gerischer, I do not think that the double ring electrode has many significant advantages compared to the ring-disc. It is sometimes claimed that because the intermediate is generated on the inner ring as opposed to the disc it is on average generated closer to the detecting electrode, and hence shorter lived inter- mediates can be studied.This is not really the case. For the ring-disc electrode the material that reaches the ring electrode is generated on the outside edge of the disc electrode. The difference between the two ge9metrie.s is whether the centre of the disc is electroactive or electro-inactive. The only advantage of making it electro- inactive (double ring) is that one does not have to drive such a large current through the generating electrode. The much greater advantage of making it electroactive (ring disc) is that the current in the centre preserves the condition of uniform accessi- bility to the generating electrode. Although the intermediate generated on the centre of the electrode is nearly all lost to the bulk of solution the mathematical description becomes much easier and one can have analytical solutions for kinetic and transient behaviour. Much the same applies to channel or tube electrodes.We have in fact done experiments with a double tube electrode and obtained titration curves in good agreement with theory. However, again the theory is more complicated, involving triple Laplace transforms, because there is no uniform accessibility on the generating electrode. An advantage of the tube electrode is that it is more robust and is more suitable for commercial applications. I thank Bruckenstein for his equivalent circuit, and, although Despic claims that ring-disc electrodes are hard to manufacture, we do not find this to be the case.Dr. W. J. Albery (Oxford) said : Electrochemical e.s.r. in which there is a well defined hydrodynamic regime connecting the electrode and the cavity, is an attractive concept. Couper and Garnett in my laboratory have constructed a tube electrode assembly in which solution flows under laminar conditions down a Spectrosil tube of diameter 1 mm. The annular eIectrode is placed immediately above the e.s.r. cavity. For this system we can solve the convective diffusion equations exactly. log ( V ~ C ~ I ~ s - ~ ) FIG. 1 .-Plot of log ( S / i ) against log ( V ) showing that the signal varies with i and V-$ ; the gradient W J. Albery and M. L. Hitchman, Ring-Disc Electrodes (Clarendon Press, 1971). of the straight line is -3.Note the wide range of flow velocity.GENERAL DISCUSSION 37 1 We have obtained good signals for the 4NO;- radical. The equation predicts that the signal, S , will be proportional to the current, i, and to V-3 where V is the volume flow rate. The figure shows that this prediction is obeyed for 4NO;-. Compared to Kastening’s apparatus I believe that we have better hydrodynamic conditions but his apparatus may be better for investigating short lived intermediates. Prof. A. J. Bard (U.S.A.) said : Goldberg and I have also carried out simultaneous electrochemical-e.s.r. (SEESR) experiments employing a cell where the counter and reference electrodes were all placed within the flat portion of the e.s.r. cell. We found that with configurations such as that described here, with the counter electrode placed outside the cell and rather distant from the working electrode, nonuniform current densities and concentrations occur across the surface of the working electrode.Under these conditions decays in e.s.r. signals occur with time, such as those observed in fig. 3 of Dohrmann’s paper, even for the generation of stable radical ions, due to redistribution of the radical ion concentration across the surface of the working electrode. also show the occurrence of such signal decays. Perhaps these resistance effects are not so important for aqueous solutions, but they certainly are large for most studies in aprotic solu- tions. Has Dohrmann performed any SEESR experiments with his cell with the production of a stable radical ion species and observed a constant e.s.r.signal with time? Digital simulation of the operation of such e.s.r. cells Prof. J. K. Dohrmann (Berlin) (communicated) : Our technique of quantitative determination of the amount of radicals in simultaneous electrochemical-e.s.r. experiments is based on the generation and deposition of Cu2+-ions by constant- current electrolysis in aqueous solution. During constant-current electrolysis at a copper electrode geometrically identical with those used for the reduction of nitro- compounds, the e.s.r. signal of the Cu2+-ions was linear with time and became con- stant when the current was switched off. With electrodes longer than 0.85 cm we found a non-linearity in the calibration plot (change in normalized e.s.r. absorption intensity against number of Cu2+-ions electrogenerated or deposited) which we ascribe to an inhomogeneity of current density distribution.Dr. E. J. Casey and Dr. C. L. Gardner (Ottawa) said : Has either of the groups represented by Dohrmann and Kastening, or any other e.s.r.-electrochemical group represented at the meeting, studied the formation and/or reactions of electrochemi- cally generated, inorganic free-radicals, particularly those postulated as intermediates i n aqueous electrochemistry? The approach to this problem in our laboratory has been as follows : first to ask how free-radical intermediates might best be produced in quantity, be encouraged to accun;ulate, be collected and measured ; then to try to achieve the likely experimental conditions.Radiation chemists have determined some of the properties of OJ, for example : the rate of decay decreases markedly as the temperature is decreased and the pH increased. In ;: reaction cell in which cold, strong KOH as anolyte is admissable to the e.s.r. cavity kia essentially a stop-flow system, e.s.r. signals for 0; are produced from several electrode materials (Cd, Ag, Pb, e.g.) during anodic oxygen evolution. A. J. Bard and 1. Goldberg, J . Pltys. Cltern., 1971, 74, 3281. I. Goldberg, A. J . Bard and S. W. Feldberg, J. Phys. Cltern., 1971, 76, 2550. J. K. Dohrmann and F. Gallusser, Ber. Bunsertges. phys. Chem., 1971, 75, 432. C. L. Gardner and E. J. Casey, Canad. J . Client., 1971, 49, 1782.3 72 GENERAL DlSCUSSION Further, rapidly-frozen samples of cold, concentrated anolyte from Cd as anode show e.s.r.spectra of 0; and another species with a narrow line at g = 2.003. The kinetics of anodic formation and decay of 0; in KOH solutions 8-12 N at temperatures -10 to -45°C has been studied in some detail.‘ Inferences were drawn concerning its origin from different surface (hydr)oxides. From e.s.r.-electrochemical studies new information obtained about cadmium oxidation and passivation includes : at the higher temperatures ( N - 15°C) 0; is the e.s.r.detectable product, while at the lower temperatures (- -45°C) 0; is observed. Reasons for this result are thought to rest in the nature of the “CdO” surface-states facing the electrolyte. Clearly, there may be experimental conditions of temperature, electrolyte com- position, surface and field-strength under which other inorganic radicals will be stable long enough in solution to be accessible by even the conventional e.s.r.techniques we have used so far. Finally, we refer to the effects of adsorbed H on the band structure of Pt-activated C cathodes which can also be observed by e . ~ . r . ~ Dr. B. Kastening (JuZich) said: We have investigated4* in some detail the formation of the superoxide ion 0; by cathodic reduction of oxygen in aqueous solutions at mercury electrodes in the presence of surfactants as well as the reactivity of this anion radicaI. It exhibits considerable stability (q: some minutes) in aIkaIine solutions, slowly decomposing by disproportionation (k= 7 M-’ s-l) to give oxygen and hydroperoxide. Although we were sure that it was present in the solutions at concentrations up to about M, we could not detect any e.s.r.signal at room temperatures. Obviously, the corresponding resonance line is, perhaps due to spin orbit inter- action, so broad as to prevent its observation. Dr. B. Kastening (JiiZich) said: Previous work by Kolb, Wirths and Gerischer6 as well as by o~rselves,~ according to which the homogeneous decomposition of nitro anion radicals is a second order reaction, has not been taken into consideration in the paper by Dohrmann et al. In the previous work it was also shown that a (first order) heterogeneous decomposition (further reduction) of the anion radical may take place even in the presence of surfactants. (The voltammetric curves, fig. 3 and 4, show clearly that the anion radical is not stable at the electrode even at the potentials of the first wave.It may be noted in this connection, that surfactants are in general considerably less effective at amalgamated than at liquid mercury electrodes.) The average diffusion layer thickness can be approximated by the relation Taking, from fig. 4, j = 1.9 mA cm-* at z = 4 (electrons/molecuIe), co = 2.2 lilM and D = 6.5 x cm2 s-l, a value of dNernst = 0.003 cm results. The reaction layer thickness would be, with k - 1 s-I7 6,-0.0025 cm. The latter is, therefore, C. L. Gardner and E. J. Casey, Cunad. J. Chem., in press. E. J. Casey, submitted to J. Electrochem. SOC. M. Bonnemay and C. Lamy, J. Electroanal. Chem., 1971,32, 183. B. Kastening and G. Kazemifard, Ber. Bunsenges.phys. Chem., 1970,74, 551. J. DiviSek and B. Kastening, 24th IUPAC meeting, Hamburg, September 1973. D. Kolb, W. Wirths and G. Gerischer, Ber. Bunsmges. phys. Chem. 1969, 73, 148. B. Kastening and S. VaviiEka, Ber. Bunsenges. phys. Chem., 1968, 72, 27 ; B. Kastening, J . Electroanal. Chem., 1970, 24,417. * B. Kastening and L. Holleck, J. Elecfroanul. Chem., 1970, 27, 355.G E N E R A L DISCUSSION 373 comparable with, and not small as compared with, the diffusion layer-as stated in the paper. One might, therefore, suspect that the first order decay as observed does not correspond to a homogeneous reaction but to the decomposition of the anion radicals diffusing back to the electrode. The boundary condition, eqn (4), for the concentration gradient of the precursor A upon interrupting the current, is presumably wrong, because the electrode will attain a rest potential in the rising part of the first wave, where a simultaneous oxida- tion and reduction of the anion radicals will take place : R-NO; -e-+R-NO, +(R-NO; + 3 e- + 3 H,O+R-NHOH + 4 OH-).This affects also the conclusions drawn from the assumption that eqn (4) applies. Prof. S . Bruckenstein (Buflalo, U.S.A.) said: Recently Miller and I have been developing a technique which complements the AC ring-disc experiment of Albery. 1-3 In our experiment, we impose a rotation speed programme given by o* = cob{ 1 + (Aco/coo)* sin 2 n . ) where represents the square root of the centre rotation speed about which the motor speed is sinusoidally modulated with a frequency f and amplitude (Ao/oo)*.The latter quantity is usually varied in the range 0 to 0.1, while f is of the order of 1-3 Hz for wo N 60 Hz. Under these conditions, the Levich equation holds and a sinusoidally modulated current is superimposed on the steady state current. Using phase sensitive detection, the sinusoidal component may be determined in the presence of huge currents arising from steady state surface processes. Since the modulated current arises only from convective-diffusion controlled processes, this procedure provides a unique way of separating such processes from purely surface controlled processes. Prof. M. Fleischmann and Dr. R. E. W. Jansson (Southampton) (communicated) : At the meeting one of us (M.F.) questioned whether it was permissible to consider the flow to be laminar in the cell described by Kastening since the configuration of the strip electrodes might be expected to disturb such a flow pattern. Since the meeting we have considered the question further and wouId comment as follows : the Reynolds number in the region of the eIectrode appears to lie in the region 2.5,<Re,< 10 and, by continuity, in the interelectrode space in the region 0.6<Re<2.5.One would therefore expect the flow to be laminar and at very low Re this flow can more or less “go round the corners”. However, behind blunt bodies there will be recirculation so that the flow approaching the next strip will not necessarily be steady (it can be unsteady without being turbulent). Since the length to gap ratio of the electrodes Z/h-8, the effects of unsteadiness could persist over a significant part of the next electrode since entry effects can persist for 10-20 gaps.Unsteadiness should lead to wrinkling of the diffusion layer and increased mass transfer. This could account partially for the experimental e.s.r. signals being greater than the theoretical, table 2 of the paper by Kastening et aZ. For comparison, a stable vortex street would be shed from a cylinder at Re N 80. Here Re is lower but at the same time we are dealing with a blunt body because the length to thickness ratio is only 4. B. Miller, M. I. Bellavance and S. Bruckenstein, Anal. Chem., 1972, 44, 1983. B. Miller and S. Bruckenstein, Anal. Chem., in press. ’ S. Bruckenstein, M. J. Bellavance and B. Miller, J . Electrochem.SOC., 1973, 120, 1351.374 GENERAL DISCUSSION Dr. W. J. Albery (Oxford) said : The RNOF- generated is similar to the photo- chemically generated intermediate with which we started this discussion. It is unstable and it can be either reduced or oxidised on the electrode. For the condition of zero current, the potential of the electrode goes to a value where the currents from the two processes, RN02 c RNO,. --+ products, are equal. As in the discussion of E* and E$ after Schiffrin’s paper this potential is concerned with the differences between the two transition states for reduction and oxidation. Although no current flows at the electrode, I agree with Kastening that at the electrode surface - e +e a[RNo,*]/a~ # 0. [ V/cm3 s-’]+ FIG. 1.-Plot of Levich equation for limiting current at the tube electrc le of tube/e.s.r.apparatus. Fleischmann’s question about turbulence can be answered by considering whether the limiting current on the generating electrodes agrees with that calculated from the convective diffusion equations. For our tube system we find good agreement. In fig. 1 our results are compared with Levich’s equation for the tube electrode where iL is in A, c , is the bulk concentration in mol rode (in cm), and Vis the flow rate in cm3 s-l. iL = 5.31 x 105n C, 0 3 23 V* I is the length of the elect- How does Kastening’s apparatus stand up to this type of test ? Dr. B. Kastening (Germany) said : Since we employed our strip electrodes under galvanosatic rather than potentiostatic conditions, we did not directly determine the limiting currents in a way similar to that mentioned by Albery.The question whether the flow was laminar or turbulent, however, can be answered from an indirect test of the limiting current. In our experiments with nitrobenzene, the e.s,r. signal levelled up to a constant intensity at about 400 pA strip electrode current. The D. J. Schiffrin, this Discussion. V. G. Levich, Physicochemical Hydrodytzamics (Prentice Hall, 1962), p. 1 IS. N.Y.GENERAL DISCUSSION 375 theoretical value resulting from eqn (14) of our paper (which corresponds to the rela- tion given by Albery) would be 590 PA. Considering the non-ideal geometry of our cell (especially with respect to the positioning of the strips in the gap between the walls), the agreement is fairly good.Under conditions of turbulent flow, one would have expected considerably larger values of the limiting current. Moreover, the Reynolds number for our experimental conditions would be considerably less than 1000; and finally, the distance between two strips is of the same size as the length of the strips and should, therefore, serve for calming down of any turbulence developing, if at all, at the back edges of the strips. Prof. J. I(. Dohrmann (Berlin) (communicated) : (1) In our electrochemical-e.s.r. study we found a first-order rate law for the homogeneous reaction of the anion radicals derived from aliphatic nitrocompounds. This is different from the results of other workers l who investigated radical anions derived from aromatic nitro- compounds.Our result indicates differences in the kinetics of the decay of the anion radicals from aliphatic and aromatic nitrocompounds, respectively, which might not be unreasonable because of the differences in spin density distribution. One other known example is the stability of the nitrobenzene radical anion compared with the instability of the 2-nitro-2-methylpropane radical anion,, both in non-aqueous solutions, the latter decomposing by C-N bond rupture. Concerning the first- order reaction, one should mention an unpublished investigation by Kolb on the reduction of 2-nitropropane by simultaneous electrochemical-e.s.r. experiments. From the dependence on time of the e.s.r. signal during the potentiostatic reoxidation of the anion radicals, a first-order rate law was deduced for the homogeneous de- composition of the anion radicals.The rate constant measured at 20°C in solutions containing 0.1 M glycine buffer, 0.5 M NaCl and 10 % ethanol was 0.16 s-' for pH> 10.5 and independent of pH. This seems to support our results. Our larger value of the rate constant might be attributed to differences in temperature and com- position of the solution. Our measurement of the current yield clearly shows that at potentials of the first wave the primary one-electron transfer is followed by secondary electron transfer reactions which probably involve the protonated adsorbed neutral radical. This would be similar to the results for nitrobenzene and corresponds to the mechanism proposed by Holleck and Kastening.6 From fig. 10 it can be seen that at the pH of 11.9 in the presence of 3 .4 ~ M TPPO the current yield increases to 0.93, thus iiidicating that the secondary electron transfer is rather slow. At pH 11.9 and with 3 . 4 ~ M TPPO the first polarographic wave of (CH,), CNO, is very close to a one-electron wave. (2) In our steady-state measurements, the flow rate parallel to the electrode was 0.1 cm s-'. Upon increasing the flow rate up to 0.4 cm s-l, the e.s.r. signal and the current increased, the ratio of both quantities being constant as would be expected for a homogeneous first-order reaction. For still larger flow rates the ratio became B. Kastening and S. Vavfitka, Ber. Bunsenges. phys. Chem., 1968, 72,27. and H. Gerischer, Ber. Bunsenges. phys. Chem., 1969, 73, 148. D. H. Geske and A.H. Maki, J . Amer. Chem. SOC., 1960,82,2671. A. K. Hoffmann, W. G. Hodgson, D. L. Maricle and W. H. Jura, J. Amer. Chem. SOC., 1964, 86, 631. D. Kolb, Diplomarbeit (Techn. Hochschule Munchen, 1966). We are indebted to Dr. Kolb for making his results available to us and for permission to quote them. R. Koopmann and H. Gerischer, Ber. Bunsenges. phys. Chem., 1966,70, 127. L. Holleck and B. Kastening, Rev. Polarog. (Kyoto), 1963, 11, 129. D. Kolb, W. Wirths376 GENERAL DISCUSSION smaller thus indicating transport of radical anions to less sensitive parts of the cell We, therefore, concluded that a flow rate of 0.1 cni s-’ was sufficiently small. (3) Our observations concerning the primary electron-transfer reaction (CH,), CNO, + e-+(CH3), CNO; 3(CH3), CNO; -+3(CH3), CNOz + 3e- (CH,), CNO; + 3e- + 3H,O-+(CH,), CNHOH f40H- ( 1 (2) (3) do not contribute appreciably to the fluxes of the nitrocompound and the anion radical at the electrode surface because the electron transfer, eqn (l), is not in equili- brium.This is clearly different from nitrobenzene. By switching the potential to less cathodic values just outside the range of potentials corresponding to the first reduction wave, we found that the reoxidation of the anion radicat (CH,), CNOi was strongly hindered and, in the presence of the inhibitor TPPO, no reoxidation was observed. It should also be noted that, contrary to nitrobenzene, the polarographic half-wave potential of the first reduction wave, ‘ E l l , , was more cathodic in the pre- sence of TPPO (A1E1l2 = -0.15 V in the presence of 3.4 x M TPPO referred to the absence of TPPO).From the slope of the (potential, time) curve observed after interrupting the current and from the double layer capacity (between 10 and 20 pF it can be estimated that the rate of reduction of the nitrocompound by double layer discharge was also negligible in most cases. The cathodic current density, - C,dE/dt, was typically less than -2 x A ern-, during the initial 0.3 s of the open-circuit decay curves and less than -6 x lo-’ A cm-2 during the main part of the decay curves. This was small enough compared with the current densities of between -2 x loA4 A cm-2 and - mol s-I for the rate of open-circuit generation of radical anions in the initiai part of the decay curve (electrode area 0.25 cm-2). This means that up to about 50 % of the amount of radical anions ascribed to the desorption effect seen at pH 11.9 (fig.8) could have been generated by double layer discharge. seem to indicate that simultaneous open-circuit reactions like A cmb2 used in the decay experiments. For a current yield of unity, we estimate an upper limit of 5 x Prof. J. I(. Dohrrnann (Berlin) (communicated) : Part of the observed deviation from a first-order decay of the radical anion (CH,),CNO;- at the pH of 1 1.9 couldalso be ascribed to generation of this species by discharge of the double layer. For a double layer capacity of 20 ,uF cm-, and a change in electrode potential of 0.5 V occurring after the interruption of the current, it is estimated that up to lo-’’ mol cm-2 of the radical anion can be generated if the discharge of the double layer proceeds by a one-electron reduction of the nitrocompound and a current yield of unity is assumed. This is half of the amount of radical anions which, on the basis of the provisional model of desorption, was ascribed to adsorbed radical anions.At present we are unable to describe the combined effects of generation by double-layer discharge and of possible desorption of radical anions in a quantitative manner. Prof. A. J. Bard (U.S.A.) said : While direct observation of short-lived radicals is possible with direct e.s.r. measurements, very unstable radicals, particularly un- charged ones, are very difficult to observe. An alternate approach in this case is to trap the radicai with a species which forms a more stable radical and observe the latter; this technique, termed “spin trapping”, has been used by Janzen and co-G E N E R A L DlSCUSSlON 377 workers for photolytically and thermally generated radicals (see e.g., E.G. Janzen, Accts. Chem. Res., 1971, 4, 31.). Mr. Richard Goodin, Prof. J. Gilbert and I at the University of Texas have recently extended this technique to electrogenerated radicals. As a spin trap, phenyl-t-butylnitrone (PBN) was employed and the trapping of phenyl radical formed on the reduction of benzenediazonium tetrafluoborate in acetonitrile at a mercury electrode was accomplished. The reaction sequence is PhNZ.+e-,Ph*+ N2 H O H 0. Ph.+PhC = NC(CH3)j-+PhZC-NC(CH3)3. (PBN) The resulting nitroxide radical is stable and was shown to give the same e.s.r. spectrum as that formed when phenyl radicals produced by the thermal decomposition of phenylazotriphenylmethane in acetonitrile are trapped with PBN.To be useful in electrochemical studies, both the spin trap and the resulting nitroxide should be nonelectroactive at potentials of the electrode reaction. Cyclic voltammetric measure- ments on PBN showed a reduction at -2.5 V and an oxidation at + 1.6 V (against Ag ref. electrode), so that it has a rather wide useful range. Several nitroxides tested showed oxidations at +0.7 to + 1.0 V and no reduction up to -2.0 V. I would like to ask Kastening how in practice, when using the multi-strip working electrode in the e.s.r. cell, does one know how to adjust the current flowing through each element? Is this done empirically? Dr.B. Kastening (Germany) said : We did not attempt to measure the potentials of the individual strip electrodes. If the electrodes are to be employed at limiting currents, then the theory predicts a maximum value of the e.s.r. signal if the largest current is applied to the first upstream electrode, with a stepwise current decrease towards the counter electrode. After the current of the first electrode was adjusted, at conditions of zero current at the other strips, to a value at which the e.s.r. signal attained its maximum intensity, the other strips were connected piece by piece, adjusting the current in each case to values at which there was no further increase of the e.s.r. signal. This mode of adjustment produced better results than any modi- fication that we tried. In the case of the SO; anion radical, where the electrodes were not operated at limiting currents and a roughly constant proportion of the SOz and SOT concentrations was favourable due to the kinetic implications, the currents of all strips were adjusted to the same value which was chosen such as to maintain a sufficient excess of SO,.Prof. R. D. Rieke (U.S.A.) said: It is not clear from the data presented by Kastening that eqn (33) has to proceed via an Sn2 reaction. It would appear that an alternant mechanism would be as follows : R -X + SO,. 3 R e + SO2 + X- (rate determining) Re + SO; -+ RSO, (diffusion controlled) This would also explain the absence of RS02* in the e.s.r. signal. Is there any evidence for the formation of R-H in your reactions as this would support the above suggested mechanism ? Dr.B. Kastening (Germany) said: As to the alternative mechanism for the sulphone synthesis proposed by Rieke, I believe that the rate-determining step of this378 GENERAL DISCUSSION mechanism is, especially for the case of simple halides, unlikely to proceed at appreci- able rates because of the rather large difference of the reduction potentials for the SO2/SO, couple (E+ - - 0.9) and RX (I?+ N - 2.2 V against SCE for butyl bromide), respectively. Moreover, we did not find, with benzyl and butyl bromides or chloro- acetic ester, products resulting from radical reactions of R., e.g., the parent hydro- carbon or coupling products. The situation may be different with reactants like &--CO-CH,Br, the reduction potential of which is fairly close to that of SO2.In this case we observed aceto- phenone as byproduct. A proof of the nucleophilic substitution mechanism, which we have proposed, would be a synthesis with an optically active halide ; such an experiment has not been carried out so far. Prof. A. J. Bard (U.S.A.) said : I would like to mention some recent experiments using the rotating ring disk electrode on the DEM system by Mr. Yeh in our labora- tories which allow estimation of the rate constants for isomerization and dimerization of DEM-0. The measurements were performed on solutions of DEM (with the small amount of DEF impurity) of 2.55, 3.23 and 3.78 mM in DMF by measuring the differences in i, for disc potentials at both the DEF and DEM waves, and for ring potentials at 0.0 V (oxidation of both DEM-• and DEF-e) and at a potential on the DEF wave (oxidation of DEM-- and reduction of DEF reaching ring). The results of these measurements at U-values of 98.6 to 305 s-I yield an isomerization rate constant, k , , for reaction (5) of about 10s-' and a dimerization rate constant, k2, for the reaction kr 2DEM-• -+ (DEM)g- of about 4 x lo4 M-' s-I. Dr. W. J. Albery (Oxford) said : In the scheme, 34M- DEF+e e DEF-. -+ +(DEF);- DEM+e DEM-. +(DEM):-~ 1 0 s - 1 ~ ( D E F H ) ~ 4 x 1 0 4 ~ - * ~ - ' has Bard considered that, when DEM is being reduced, the DEF-• formed might couple with more of its DEM-• precursor. This gives a third route : 10s-1 + DEF-• +DEM-. + )(DEF)(DEM)2-. If the coupling was fairly fast the kinetics would be first order and difficult to distin- guish from the upper route. Prof. A. J. Bard (U.S.A.) said : One must certainly consider the possibility of cross-coupling between DEM-• and DEF-• as well as the electron transfer reaction between DEM-. and DEF in the overall reaction scheme. Slight trends in the measured rate constants with rotation rate and concentration suggest that complica- ting reactions such as these must also be occurring. We are currently extending our measurements with the RRDE to solutions containing mixtures of DEF and DEM and hope to resolve this question.

ISSN:0301-7249    

DOI:10.1039/DC9735600367

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

CONCLUDING REMARKS BY J. E. B. RANDLES University of Birmingham It has fallen to my lot to make the summarising remarks for this Discussion. 1 am sure that it is best to be brief, and as it is impossible to do justice to all that has been presented there are bound to be omissions; for which I apologise. In fact 1 shall be self-indulgent to the extent of presenting to you more my personal response to this Discussion rather than attempting an objective and comprehensive summary. In his Introductory Lecture, Professor Gerischer has remarked on the large changes which have occurred in electrochemistry since the Faraday Discussion on Electrode Processes in 1947. I think that these changes were favoured by the parallel and complementary growth which took place in two closely related yet distinct fields of work.These were the structure and properties of the electrical double layer and the kinetics of electrode reactions ; together they led us to our present considerable understanding of the mechanism of electrochemical reactions. The rapid progress in these two areas was due primarily to the influence of two techniques. One was the use of the dropping mercury electrode. This had been used mainly analytically before and during the 1939-45 war but it was thereafter adopted with enthusiasm by electrochemists for their own purposes. The other technique was, of course, the use of the sophisticated electronic equipment which became increasingly available and which permitted the measurement of strongly time-dependent currents and potentials.Aided by these we took up " relaxation methods ", adding the variable " time " to the traditional electrochemical variables " current " and " potential " and the results were most gratifying. The various methods developed during these last 25 years or so are still in use : a.c. impedance measurement, cyclic voltammetry and various steps, sweeps and pulses ; indeed, all forms of chronopotentiometry and amperometry. They remain very valuable but 1 believe we have reached the stage at which they urgently need supplementing. The encouraging fact about this Dis- cussion is that it has provided so many examples of new techniques being exploited and yielding exciting results. My emphasis on the importance of new techniques at this time is linked with my view of the desirability of experimenting so far as possible in ways which allow simple and unambiguous interpretation of results.I will illustrate by an example remote enough not to cause offence. Mainly during the 1950s and early 1960s, much work was done in extending the mathematical theory of combined kinetic and diffusion control of currents at a dropping mercury electrode. The kinetic contribu- tion was extended to include all manner of combinations of the charge transfer step with chemical reactions, both monomolecular and bimolecular, preceding or following the charge transfer and sometimes with adsorption or double layer capacity currents thrown in for good measure. The main workers in this field were the Czech, Japanese and Russian schools of electrochemical kinetics and Delahay and his co-workers in the U.S.A.Now, all credit is due to them for establishing the basic theory of combined diffusion and kinetic control of electrode currents, and for reasonably simple condi- tions i t has been much used. But I am fairly certain that the more exuberant ex- tensions of the theory into situations so complex that the equations occupied several 379380 GENERAL DISCUSSION lines of typescript, were never used in practice. Parameter adjustment can always make theory agree with experiment (including experimental error), but the agreement is not unique and the interpretation cannot be unambiguous. I have a feeling that some present developments of the three variable (current, potential, time) techniques may be tending the same way.For example, the representation of a.c. impedance on the complex impedance plane as a function of frequency is a great help to interpreta- t ion when the electrode reaction involves several steps. Nevertheless, Dr. Armstrong (in his paper with Professor Thirsk and Firman) has shown that, even using complex- plane display, measurements of the as. impedance of the electrodeposition of solid metals cannot possibly distinguish between some alternative control mechanisms. If an electrodeposition process involves dissociation, diffusion and deposition of ions, followed by surface diffusion and incorporation of adatoms, I suspect that the situa- tion is too confused for detailed interpretation of electrochemical measurements to be possible, and it will not be helped by further mathematical treatment of the results.I regard critical assessment of the limitations of a method, such as that presented by Dr. Armstrong, as most valuable. In the evolution of a technique it is naturally first applied to those experimental systems and situations for which it can give clear-cut and unambiguous information. Then, as these are progressively exploited, it becomes applied in circumstances where the theoretical basis has to be extended and complicated in order to retain its validity. Eventually, interpretation of the experiments becomes clouded with uncertainty ; at that stage I think one should have the courage to stop doing them and do something else instead. The introduction of a new technique to study the same problem may be its salvation.This is being done in electrochemistry and this Discussion has provided several examples. A weakness of the familiar variables current, potential and time is that they are not specific for chemical substances (unless we regard current as specific for electrons). Spectroscopic methods of identifying and estimating reaction products and inter- mediates can obviously fill this gap. The paper by Dr. Bewick and co-workers illu- strates this well. They used reflectance spectroscopy to identify and estimate the radical COY formed as intermediate in the cathodic reduction of CO,. It was also used to follow the formation of thianthrene radical cation during the anodic oxida- tion of the molecule, a reaction also studied by Professor Kuwana et al. The paper presented by Dr.Kolb described the observation by reflectance spectroscopy of sub- monolayer up to multilayer films of metal (Zn and T1) electrodeposited on to a non- metallic substrate (ZnO). It appears possible to observe a gradual change of the optical absorption of the layer from similarity to that of individual atoms towards similarity to that of bulk metal. On the other hand, reflectance spectroscopy cannot give information about the electrodeposition on to a metal surface of ions of the same metal, and even if the substrate metal is different the interpretation of reflectance changes is difficult. The complexity of the cha-ilge of reflectance of a metal/electrolyte interfzce with potential change and ion or atom deposition, was pointed out by Dr. MacIntyre and Beck.Probably the most useful contribution that spectroscopic methods (reflectance, transmission and also Raman) can make is to the identification and estimation of reaction intermediates in the solution, close to the electrode surface, and to the observation of adsorbed organic molecules, in conjunction with capacitance measurements. Electron spin resonance is also a useful tool because of its specificity for radicals, but it is subject to experimental difficulties. If the electrochemically formed radicals are short lived, then the electrodes producing them must be located within the reson- ance cavity. This requires thzt the electrodes are small and placed in such a way asG E N E R A L DlSCUSSlON 38 1 not to cause excessive damping of the signal. The mode, and rate, of reaction of radical intermediates may be determined from signal decay curves after the electrode current is interrupted, in the manner used by Dr.Dohrman and his co-workers. An alternative method, based on the measurement of steady state concentrations in a flow system with an ingenious use of multiple electrodes, was described in the paper introduced by Dr. Kastening. The detection of 0; and 0; radicals produced in the electrolysis of strong KOH solutions at low temperature, was mentioned in the discussion by Dr. Casey. A technique which gives a new view of electrode reactions relies on the photo- excitation of an electrode. The emitted electrons thermalise and solvate within a distance of a few nm from the electrode surface and then combine with whatever electron acceptor (“ scavenger ”) they can find.It can be regarded as a more con- trolled version of radiation chemistry with the useful additional parameters : electrode current and potential. The papers by Professor Pleskov and co-workers and by Dr. Schiffrin were concerned with steady-state photo-currents due to the combination of emitted electrons with H30+, NO;, NO;, C02 and CH,Cl, and the subsequent reaction, in solution or at the electrode, of the radicals formed. Dr. Schiffrin’s study of the production and subsequent reaction of CO; gives information comple- mentary to that of Bewick et al. for the cathodic reduction of C02. Dr. Barker was a pioneer in this field and has reported, for this Discussion, on the use of very short duration laser light pulses, in order to follow rapid reactions of the radicals produced.The particular radical, in this instance, was OH- produced by the interaction of N20G. with e&. In stressing the value of new techniques, I do not want to de-value our more traditional methods. The rotating ring-disc electrode for example is a most elegant tool. Used in the conventional way (d.c.) it can give rather direct evidence of the production, by the primary electrode reactions, of soluble intermediates and also some information about the kinetics of their subsequent reaction in solution. Examples were given in the paper by Professor Nekrasov who has worked with this electrode for many years. Another veteran of the rotating ring-disc, Dr. Albery, has imposed an alternating current on the disc electrode.The resulting ring current is elegantly interpretable for a simple charge transfer reaction, but for the 2Br- + Br, reaction where both these species, and the intermediate Br-, may be absorbed, interpretation is far more difficult. Cyclic voltammetry with linear potential sweep is still useful for providing qualitative evidence of intermediates as shown in a paper presented by Dr. Eggins. Professor Conway and his co-workers described a new variant of the linear potential sweep wherein short periods of faster sweep rate were superimposed on a slower sweep with the aim of distinguishing between fast and slow adsorption, or faradaic, processes. The elegance of a piece of work depends partly on the clarity of the question posed, and partly on the appropriateness of the methods used to answer it (and, of course, on the skill with which they are applied).I remark on this in connection with with the satisfying paper presented by Professor Bard and his colleagues. Clear conclusions emerged from the combined use of simple electrochemical methods, with some assistance from e.s.r. In the verbal discussion, of this and other papers, it was evident how collation of information from diverse techniques can assist towards definite conclusions. In conclusion, I would like to say how enjoyable and encouraging I have found this Discussion. Occasionally I feel, as I believe many of us do, that Chemistry has lost its glamour, has become a late middle-aged subject. Gone are the days when wide new horizons were opening, as in the great days of organic chemistry at the382 G E N E R A L DlSCUSSfON beginning of this century, or the exciting times when quantum mechanics was explain- ing the real nature of the chemical bond. Within the realm of chemistry, only biological chemistry, I suppose, is likely to provide such excitements in the future. However, electrochemistry still has vitality and a potential for continuing usefulness. The topic of our Discussion, much of which was related to electrode reactions of organic compounds, contributes to the theoretical background of electro-synthesis which must surely have a future use in chemical industry. But in future I think we may be forced to be more selective in our choice of research topics. We should try to be answering real questions, not just exploring trivial fringes of the well known, or we should be trying to find out how best to do something that someone, say, chemical industry, really needs to do. We should refrain from barren mathematising, we should use the techniques we have where they are most appropriate and think actively about possible new ones. Then, I think, electrochemistry will retain its vitality and produce exciting ideas a s well as making itself useful to the community.

ISSN:0301-7249    

DOI:10.1039/DC9735600379

出版商:RSC    

年代:1973

数据来源: RSC

摘要:

Alber), V. J., 28, 65, 67, 73, 108, 109. 317, 369, 370, 374, 378 Angc.rstc.ia-KozIowska, H., 1 61, 210. ArcIxr, 51. D., 28 Armstrong, R. D., 244. Aylmer-Kelly, A. W. B., 96. Bard, A. J., 353, 371, 376, 378. Barker, G. C., 41, 67, 71, 72, 73, 1 13. Barrctt, M. A., 159, 235. Bergman, I., 152. Benick, A., 64, 66, 96, 1 1 0 . 113. 113. 152, 159. Bindra, P., 180. Bottura, G., 41. Brom:iri, H. F., 16. Brown. D. J., 307. Brown, 0. It., 110, 238. Bruckenstein, S., 285, 369, 373. Cantrill, P. H., 96. Capon, A., 235. Casey, E. J., 157, 371. Clarkc, J. S., 168. Cumeau, J., 285. Concialini, V., 41. Conway, B. E., 121, 152, 158,210,235, 236, 237, Davis, A. H., 317. Davison, W., 171. Despic, A. R., 199, 367. Dhar, H. P., 242. Divikk, J., 341. Dobrman, J. K., 330, 371, 375, 376. Eggins, B.R., 66, 109, 118, 276, 296, 298, 299. Eletskg, V. V., 52, 71, 72, 73. Epelboin, 1. E., 72, 202, 264, 295, 367. Field, N. J., 28. Firman, R. E., 244. Fleischmann, M., 180, 203, 205, 206, 207, 208, Galluser, F., 330. Gardner, C. L., 157, 371. Gerischer, H., 7, 207, 367. Gileadi, E., 228, 240, 242. GostiSa-Miheltic, B., 341. Gottesfeld, S., 210. Harrison, J. A., 171, 199, 203. Heineman, W. R., 16. Hcitbaum, J., 305. Ho, 1:. C., 210. 238, 239, 241, 242, 299. 294, 373. Janson, R. E. W., 373. Kastctring, B., 65, 239, 341, 372, 374, 377. Keddam, M., 166, 167, 203, 201, 238, 264, 294, Kenkcl, J. V., 353. Klinger, J., 210. Kolb, D. M., 138, 163, 166. Kuhn, A. T., 168. Kunana, T., 16, 64, 65, 66, 154, 298. Lakomov, V. I., 52, 71, 72, 73. Lestrade, J. C., 264, 295. Lomax, A., 353, MacDougall, B., 210.Mclntyre, J. 1). E., 122, 156, 158, 159, 162, 163, McKeown, D., 41. McQuillan, A. J., 167. Manousek, O., 300. Mason, A. J., 317. Nekrasov, L. N., 308. Oldfield, J. W., 180. Orville-Thomas, W. J., 168. Parsons, K., 62, 113, 240. Peck, W. F., Jr., 122, Pleskov, Yu. V., 52, 71, 72, 73. Por.ter, G., Sir, 64. PospiSil, L., 293. Pravdic, V., 201. Puglisi, V. J., 353. Rnrrdles, J. E. B., 379. Reeves, P. M., 67, 111, 199,297. Riecke, R. D., 377. Rotenberg, 2. A., 52,71, 72, 73. Rusiaa, A., 304. Schiffrin, D. J., 73,75,109, 110, 11 1, I1 2, 206. Schultze, J. W., 71, 236. Singleton, D., 180. Smith, F. R., 68, 113, 118, 119, 121. Tbirsk, H. R., 205, 244. Thompson, J., 171. Turner, A. D., 28. Tuxford, A. M., 96. Vetter, K. J., 237, 239. VlEek, A. A., 63, 200, 304. Volke, J., 300, 304. Vukovic, M., 201. Williams, M. J., 41. Wittchen, H., 330. 295, 367. 166, 167. AUTHOR INDEX* * References in heavy type indicate papers submitted for discussion. 383

ISSN:0301-7249    

DOI:10.1039/DC9735600383

出版商:RSC    

年代:1973

数据来源: RSC