GENERAL DISCUSSION Dr. W. Paik and Prof. J. O'M. Bockris (University OfPennsylvania) (communicated) We agree with Stedman that in interpreting ellipsometric data at the metal-solution interface it is necessary to take into account the electroreflectance effect. We doubt at present that there is a significant effect of pressure on the refractive index of the double layer. However Stedmann has assumed following Hansen that as the electron concentration in the metal surface changes due to the potential change there will be a monotonic shift of the NM and KM parameters along the ;1 axis and this assumption clearly does have difficulties near to the optical edge. In work in press,l having aims similar to those of Stedmann we have examined the wave-length dependence of 611 but have used the electron theory of metals to relate the changes in optical properties to frequency (using Br- adsorption on Au thus eliminating the limitation which may exist in her work).In this way we have found a fairly consistent agreement between the observed quantities of those expected on the basis of our calculations of the electroreflectance effect. Dr. R. Parsons (University of Bristol) said Stedman mentioned in her paper the use of ellipsometry to determine points of zero charge. However because there is more often specific adsorption than not this method will not be very useful in general. Perhaps it might be a useful method to determine the point at which the charge on the diffuse layer is zero. I would be interested in her comments and in particular whether she has succeeded in determining points of zero charge this way.Dr. M. Stedman (Nat. Phys. Lab. Teddington) said In reply to Parsons I agree that the presence of specific adsorption would complicate if not preclude the use of optical methods to determine potentials of zero charge. Even in the absence of adsorption it would be difficult to determine with certainty the contribution of the diffuse layer. The remark in my paper was not intended as a practical proposal but was a passing thought which might trigger further ideas. I have not made any attempts to determine potentials of zero charge. Dr. B. Cahan (Case Western Reserve University) said There is some question as to the appropriateness of the use of the Hansen-Prostak model in the calculations used to generate the graphs in fig. 1 of the paper of Stedmann et al.It has been pointed out by several authors (McIntyre Bewick) at this Symposium that this treatment does not give a fit for metals other than Au. We have shown that the magnitude of the effect agrees numerically with the calculated value only at the absorption edge and Stedmann has pointed out that even the sign of the effect is wrong. Since the primary assumption used in treatment is thermodynamically unsound it appears that the agreement at the absorption edge of gold is fortuitous. Therefore there is neither an empirical nor a theoretical basis for the use of this theory for these calculations. The magnitude of the expected changes in reflectivity due to the double layer may be somewhat overstated. While it is true that a AA of -0.120' is well within W. Paik and J. O'M.Bockris to be published. W. Paik M. A. Genshaw and J. O'M. Bockris J. Phys. Chem. 1970,74,4266. 85 86 GENERAL DISCUSSION the resolution obtainable ellipsometrically on a relative basis this is not true for an absolute measurement. With present surface handling techniques it is impossible to prepare two specimens of the same material that agree within this tolerance. The reference state used to determine the above-mentioned 0.120" is a filmless state (i.e, without any double layer) and therefore experimentally unrealizable. A real measure- ment will involve smaller changes than the extremes used here and these effects should be correspondingly lower. This is not to imply that the optical changes in the double layer are unimportant. On the contrary as out recent glancing angle reflection experiments have shown (see discussion on our paper) we are vitally interested in this question but believe that these measurements can and should be made with a system in which the effects of the substrate do not interfere.Dr. M. Stedman (N.P.L. Teddington) said I agree with Cahan that in the light of contributions that have been made to this symposium one would no longer choose the Hansen-Prostak model for predicting the contribution of the metal to the optical behaviour of electrodes. The figure of -0.120' for AA was an estimate of the optical effect of changes in the compression or density of the compact layer for the system aqueous NaF/Hg. It relates to the compression of 17 % derived by Hills and Payne for the surface charge changing from - 10 to +20 pC cm-2. Accordingly AA was computed for a system with compressed film (n = 1.405 d = 4 4 relative to that with an uncompressed film (n = 1.334).The latter is conceptually present though having the same refractive index as the bulk solution one may neglect it for the purposes of computation. It was in this sense that I called the reference state filmless. Therefore the figure for Ad is a true contribution to a real measurement in which the surface charge changes as stated above. Since the magnitude of com- pression is typical of the values derived by Hills and Payne for various electrolytes I conclude that changes in the compact layer may make important contributions to double layer optical effects. The main uncertainty in my calculation lies in estimat- ing the refractive index of the compressed layer.In conclusion most of the optical effects computed in my paper are contributions to the total observable effect of particular features of double layer structure. Individually they are experimentally unrealizable but relate to real systems when summed. Dr. A. T. Kuhn (University of Salford) said Work is in progress under the super- vision of Prof. Orville-Thomas and Dr. A. T. Kuhn at the University of Salford to measure the vibrational frequencies of species adsorbed at platinum electrodes in aqueous solution. Calculations were made with the computer programme prepared at N.P.L. by Dr. M. Stedman to ascertain the required measurement sensitivity to investigate the interaction of infra-red radiation with a partial monolayer of adsorbed species. The results indicated that absorption of radiation by surface species would be low but adequate separation of absorption bands of interest from the background spectrum could be made by observing the reflected radiation by phase-sensitive detection while the electrode potential was modulated by an a.c.wave. The results also gave some indication of the tolerable range of film thicknesses for platinum sputtered on a germanium hemi-cylinder at which penetration of the radiation into the electrolyte would occur while undergoing internal reflection at the platinum/electrolyte boundary. Platinum films have been prepared on germanium substrates with suitable thickness and resistive characteristics. With such films the absorption spectrum of water has been recorded using internal reflection spectro- scopy. GENERAL DISCUSSION 87 Dr.M. A. Genshaw (Elkhart Indiana) said I would with regard to the paper by Stedman suggest a possible method of distinguishing optical changes occurring inside of the metal from those due to changes on the solution side. This is to make kinetic measurements of the optical changes. To stimulate work in this direction I offer the following data which I obtained while at the University of Pennsylvania. A Spectra Physics model 132 He-Ne laser was used as the source in a Rudolph 200-E ellipsometer. The stability of the laser intensity was 3.15 %. The angle of incidence was 65" outside the cell or about 71.5" in the solution. A Motorola MRD-300 phototransistor was used as a photo diode with a operational amplifier with a megohm resistor in the feedback loop to amplify the output (1 V = 1 PA).V (S.C.E.) FIG. 1.-Modulation of A and intensity as a function of potential at 500 Hz and 18" offset. A Hewlett-Packard Wave Analyzer model 302A was used to separate and measure the light modulation. The d.c. amplifier output was 0.15 V/deg. in P (polarizer setting) with the polarizer at 18" from extinction the analyzer at extinction and 0.32 V/deg. at 45". The r.m.s. noise in the output was 5 pV with a d.c. level of 1.3 V whrch is equivalent to 33 pdeg. in P or 66 pdeg. in A. The electrode was modulated with a 100 mV p-p signal and the potential controlled with a Wenking potentiostat. The working electrode was platinum in pH 0.4 1 M H,SO solution. Plots of the potential dependence of the modulation are given in fig. 1 and 2. A strong dependence on potential is noted with a maximum at about 0.5V.The modulation decreases markedly on the oxide-covered surface with the usual hysteresis in oxide reduction being observed. Comparison of fig. 1 and 2 is difficult due to the change in both frequency and offset angle but the apparent increase in the A 88 GENERAL DISCUSSION modulation would indicate that the intensity modulation may be dominant over the A modulation. The frequency dependence is illustrated in fig. 3. Linearity is not observed in linear or full logarithmic plots. This would implicate kinetic control 150 > I- bo 3 120 E C 2 90 -0 4 60 -. 30 120 C .r( 0 V (S.C.E.) FIG. 2.-Modulation of A and intensity as a function of potential at 3000 Hz and 45" offset. of the process causing the modulation. The marked decrease on the oxide-covered surface would suggest that the electron density inside the metal is not the doininant factor in the electromodulation effect as the double-layer capacity and hence the I I00 frequency in Hz FIG.3.-Frequency dependence of electromodulation at 0.6 V and 45" offset. charge introduced into the metal on modulating does not decrease when the oxide phase is formed. I acknowledge the assistance of Mr. 2. Nagy and Mrs. N. El. Nadori in making these measurements. Dr. M. Stedman (N.P.L. Teddington) said I thank McIntyre for pointing out the general relationships between reflectance changes that apply at an angle of GENERAL DISCUSSION 89 incidence of 45”; their simplicity may well make them useful diagnostically. However various considerations enter into the choice of conditions for observing effects due to changes in the metal or in the electrolyte and I believe that computed predictions of the type illustrated in my fig.1 (but based on acceptable models) will still be valuable. The optical constants for Hg used in my calculations were taken from Faber and Smith and appear to be the most reliable available. The refractive index is indeed higher than predicted by Drude theory but the difference is probably The comment that MA theory predicts an ER effect for Hg of sign opposite to that used by myself is interesting. Applied to the theoretical curves in my fig. 3 this would tend to reduce the slope of the total optical effect and improve agreement with experiment. The crucial test would be in the behaviour over a range of wave- length. In reply to Genshaw I agree that time dependence can be a useful diagnostic in analyzing changes due to several concurrent processes.For this reason I used potential stepping in the NaF/Hg experiments to obviate the effects of impurity adsorption. Some double layer processes will be very fast and I doubt if kinetic measurements can provide a complete distinction between changes in the metal and changes on the solution side. Dr. W. J. Plieth (Free University Berlin) said The reflectivity measurements of Br- adsorption shown in fig. 6a and b of the paper of Barrett and Parsons show a steady decrease of reflectivity with increasing potential in the potential range investi- gated. The investigations of Bagotzky Vassiliev Weber and Pirtskhalava and of Balashova and Kazarinov show a constant maximum coverage in the same potential region.The concept of partial charge 5 * has to be considered for the adsorption of Br- ions on platinum. In this case the adsorbed Br- ions are characterized by a partial charge Br”l caused by a partial transfer of electrons to the platinum electrode. The partial charge A- 1 characterizes the probability of finding an electron in the electron system of the adsorbed ion. This partial charge value is combined with an equilibrium distribution function of the solvation energies.’ The partial charge transfer coefficient A increases with the potential. Therefore the properties of the adsorption layer change from the properties of a pure ionic layer to those of a layer of Br atoms. This change should be accompanied by an increase of optical absorption in the adsorption layer.In this case the change in reflectivity in the investigated potential region would not be in contradiction to the otherwise observed constant surface coverage. Barrett and Parsons assume that a change in the properties of the adsorbed film is fast compared with the time constant of the adsorption step. They explain the relative slow change in the reflectivity by the change in the surface coverage. The rearrangement of the solvate molecules on the electrode surface and in the solvation shell of an ion is rate determining for the adsorption of an ion.8 The electronic equilibrium is adjusted during the greater part of the process. The rearrangement of the solvate molecules in the adsorption T. E. Faber and N. V. Smith J. Opt. SOC. Amer. 1968,58,102. N. V. Smith A h .Phys. 1967,16 629. V. S. Bagotzky Yu. B. Vassiliev J. Weber J. N. Pirtskhalava J. Electroanal. Chem. 1970,27 31. N. A. Balashova V. E. Kazarinov Electroanal. Chem. ed. A. J. Bard (Marcel Dekker New York) 1969,3,135. W. Lorenz G. Salit 2. phys. Chem. 1961,218,259; 2. phys. Chem. N.E 1961,29,390. W. J. Plieth K. J. Vetter 2. phys. Chern. N.F. 1968 61 282. K. J. Vetter W. J. Plieth 2. phys. Chem. N.F. 1969 65 181. ti G. Salid W. Lorenz 2. phys. Chern. N.F. 1961,29,408. 90 GENERAL DISCUSSION Iayer is rate determining for a change in the structure of the adsorption layer.' The electronic equilibrium remains adjusted over the whole process. Therefore the rate of a change in the structure and in the charge of the adsorption layer should be of the same order a the rate of the adsorption step.The rate should not be as fast as the authors assume and the decrease in reflectivity between 0 and 1 V is more likely to be due to as change in the structure and in the charge of the adsorption layer than to an increase in surface coverage. It is of interest that optical measurements allow such observations. Dr. B. Cahan (Case Western Reserve University) said Part of the difficulty experienced by Parsons in fitting optical constants to the adsorbed layer of hydrogen on platinum may be resolved by a consideration of some (as yet) unpublished data which I obtained during studies on sputtered Pt-film electrodes. Information on absorbed hydrogen in the metal was obtained by two techniques one involved a bending cantilever beam originally developed far studying the interfacial tension of solid electrodes and the other used conventional voltammetry.The bending beam was a 1 x 10 cm sheet of # 00 cover glass on which was sputtered a thin film of platinum. This substrate is so flexible that even the small forces involved in interfacial changes with potential are sufficient to cause deflections of the free end of the cantilever of 20-30 fringes when measured interferometrically. When FIG. 1.-Two voltammetric sweeps from 0.5 to 0.05 V and back for a thin-film Pt electrode in 1 N HC104. Sweep A starts after a 2-min rest at 0.5 V. Sweep B was the tenth sweep in a con- t inuous series. Sweep speed 25 V/s ; vertical sensitivity 50 ma/cm ; horizontal sensitivity 50 mv/cm. the potential was lowered to the potential at which hydrogen coverage is large the beam was deflected through several thousand fringes.This displacement was rever- sible but 2-3 s were required for complete recovery. Effects of this magnitude could not have been caused by surface forces only and are strong evidence of an absorption into the bulk metal causing severe stresses to be set up in it. The time constant W. J. Plieth Z.pliys. Chem. N.F. 1969 57,:178. GENERAL DISCUSSION 91 associated with the positional recovery agreed well with Bockris’ data for the solubility and diffusion coefficient of hydrogen in platinum. When studying the voltammetry of hydrogen on bulk platinum two current peaks occur before Hz evolution. The position of the more anodic of these two peaks is observed to change as a function of sweep rate and history. With thin films of Pt these peaks are sharpened and it can be seen that shift is only apparent and is caused by a replacement of the first peak with a new one.The figure shows traces from two sweeps from the double-layer region down to hydrogen and back. The only difference between the two is the time the potential was held in the D.L. region. Trace A was run after standing at 0.5 V for 2 min while trace B was the tenth consecutive sweep. While not shown on the figure the crossovers at 350 mV on the cathodic branch and 600mV on the anodic branch are typical isosbestic points which usually implies the gradual replacement of one peak by another too close to it to be completely resolved. The time constant associated with this change- over from A to B is about 1-2 s. Both traces originate from a potential (0.5V) where no hydrogen should be present on the surface.Yet the 50 mV difference between the two first peaks means that the first step in the h.e.r. has been changed by the recent history of the electrode. The implication is that the mechanism has been altered because of a change in the nature of the surface of the Pt. It is unlikely at these potentials that this change has been the result of migration of the platinum surface atoms. It is most likely that the change is induced by the presence of hydrogen below the surface of the metal. Once deposited inside the metal it is then free to diffuse into the bulk but in a thin film rapid saturation occurs. This diffusion of hydrogen into the metal and the resultant stress of the platinum is consistent with the results obtained with the bending beam.Any change in the nature of the platinum surface sufficient to cause a change in the mechanism of the h.e.r. should alter the structure of the surface. This will result in a change of the optical constants of platinum to a depth commensurate with the depth of penetration of the hydrogen. This can explain the inability of Parsons to fit the optical behaviour of the hydrogen layer to any reasonable set of optical constants for a thin layer since the bulk optical constants of the substrate may be changed by the internal hydrogen. The above concept also provides an alternative point of view to that of Bewick regarding the strongly and wealky bound hydrogen. If the H is really inside the surface and can diffuse in or out slowly the slow step may not be the H:dS2H+ but may simply be the slow diffusion of the dissolved hydrogen to the surface.Prof. E. Yeager (Cleveland) said Multiple specular reflection has been used by Dr. T. Takamura Dr. K. Takamura and myself to examine the specific adsorption of halide ions on gold. The experimental arrangement is similar to that used in an earlier paper except that the number of reflections* has been reduced to - 13. The Cl- concentration dependence of the reflectivity is shown in fig. 1 for gold in 0.2 M HC104 in the presence of various concentrations of added C1- at a potential of 0.80V (Ag AgC1) at which voltammetry measurements have shown C1- to be T. Takamura K. Takamura W. Nippe and E. Yeager J. Electrochem. SOC. 1970 117 626. *The use of such a large number of reflections can lead to substantial errors in the absolute value of the reflectivity change and in its wavelenth dependence as pointed out in the earlier paper Cahan Horkans and myself.At a given wavelength however the relative dependence of reflec- follow adsorption and desorption of various species at the electrode 92 10 8 6 - 4 - 2 - 0 - I I I - - A A Y ” - I - I I I Fra. 1 .-ReIat&/e reflectivity change of gold in 0.2 M HC104+ C1- at + 0.8 V against &AgCl at 5600 A. adsorbed. The ordinate represents the relative change in reflectivity between this potential and 0.0 V (Ag AgCl) where specific adsorption of C1- does not occur minus the corresponding change in the absence of Cl- ions in solution. A linear relation- ship exists between the surface concentration of specifically adsorbed halide anion and the reflectivity.Thus the curve fig. 1 corresponds to an adsorption isotherm 12 10 8 $ 6 4 a 4 2 0 20 40 6 0 AQ W/cm2) FIG. 2.-Reflectivity changes against charge for halide adsorption. AQ = Q2- el where Q2 = anodic Q in presence of halide anion and Ql is without halide anion. AR = R2 where Rl R2 are defined analogous to the Q. Ro = highest value on each ( R E ) curve. GENERAL DISCUSSION I I I I I I 93 0 I0 20 30 40 50 time (s) FIG. 3.-Time dependence of reflectivity of AU. Initial potential = -0.3 V; final potential = 3-0.8 V (solid line). Dashed line reverse potential step (time scale shifted +5 s). Electrolyte 0.2 M HC104 ; C1- molarity a 6.3 x ; b 3.1 x lo-’. Wavelength = 560 nm. Insufficient points are presently available at low concentrations to check the mathe- matical form of the isotherm and furthermore the concentrations of Cl- listed for low values may be in some small error because of the possibility of residual C1- in the perchlorate electrolyte.Fig. 2 indicates plots of the reflectivity changes due to the adsorption of the halide ions as a function of their charge as determined from integration of the voltammetry mrves. The concentration of halide anion was held constant as the potential was varied to change the halide adsorption. The linearity of these plots is noted. The 8 6 I 2 0 2 4 6 8 dt s+ FIG. 4.-Reflectivity against dtirne with potential steps. Conditions same as for fig. 3. C1- molarity a 6 . 3 ~ b 3.1 x lo5 ; c 1 . 2 5 ~ 94 GENERAL DISCUSSION increase of slope in going from C1- to I- is in accord with the fact that the I- is more strongly adsorbed and thus interacts more strongly with the surface orbitals e.g.the surface 5d or 6s orbitals. The rate of adsorption of the halide anions has been examined by stepping the potential between values at which halide adsorption does and does not occur. The solid curves in fig. 3 reveal two parts a fast component which is the same as the total change in solutions without C1- ions and a slow component which depends on the C1- concentration with its rate increasing with Cl- concentration. The plot of the slow component of the reflectivity against the square-root of time yields linear plots (fig. 4). When the potential is stepped in the opposite direction resulting in desorption of the halide anion the reflectivity recovers the original value virtually instantaneously within the time resolution of the apparatus (-0.1 s).This indicates that the de- sorption is not diffusion controlled and further that the reflectivity is not caused 6 0 10 r3 (A3) FIG. 5.-Adsorption potential against r3 for anions. by an abnormal concentration of C1- ions in the diffusion layer. Further if the reflectivity were infiuenced to an appreciable extent by refractive index changes in the ionic double layer then the reflectivity should be sufficiently sensitive to detect also the abnormal Cl- concentration in the diffusion layer since the product of the layer thickness times the average refractive index deviation from the bulk value should be comparable for the ionic double layer and the diffusion layer under these conditions. A comparison of the (reflectivity potential) curves with the voltamnictry curves in the presence of halide anions reveals that the change in reflectivity is a maximum GENERAL DISCUSSION 95 at the peak in the voltammetry curve.This is readily understood from the equation for the adsorption current density where R is reflectivity E is potential and q is the charge of the adsorbed ions. Since dR/dq is essentially constant and a linear voltage sweep is used iad is directly pro- portional to dR/dE. The potential of the maximum in the (dR/dE E) curves at high halide concentra- tion with reversible adsorption is some function of the adsorption free energy with the nature of the function dependent on the type of adsorption isotherm. If the anion adsorption depends on the polarizability of the anion then this adsorption potential should be a function of the polarizability and in turn the cube of the ionic radius since the polarizability is proportional to r3.The plot of the adsorption potential against r3 using Pauling radii Since only a small range of radii (1.81-2.16 A) is covered in this plot other functions of Y might prove dso nearly as linear within the precision of the plot. There is relatively little doubt that other factors besides polarizability need to be considered. These results indicate that reflectivity measurements readily lend themselves to the study of halide adsorp- tion on metals such as gold.* is linear (see fig. 5). Prof. B. E. Conway (University of Ottawa) said In relation to the dependence of the optical parameters A and $ on coverage of Pt electrodes by surface oxide referred to in the papers of Parsons Bewick and McIntyre it is of interest to compare the structural details of the form of the electrochemical charging curve for oxide formation examined by Kozlowska and I in very pure solutions with the potential dependence of A.In fig. 1 are shown successive anodic (current (i) potential (V)) profiles at Pt in highly purified 1 M aq. H2S04 taken up to progressively higher anodic termination potentials ; the corresponding cathodic reduction curves are also shown. The curves were obtained potentiodynamically and are hence differential charging curves since the charge for surface oxide formation (or reduction) is given by Q = lidt = ](i/s)dV where s is the rate of potential sweep. In fig. 1 is also shown the integral charge-potential relation expressed in terms of the ratio Q,/& where QH is the charge required to form a monolayer of adsorbed H in the cathodic region (+0.03 to +0.40 V).Several features are of interest (a) the differential (i V ) curves show appreciable structure with two pronounced and one less pronounced peak current being evident. That these are not due to intrinsic heterogeneity of the surface is indicated by the fact that only one peak is observed in cathodic reduction of the surface oxide except when the oxide is formed at high potentials and reduced in a fast sweep.2 However at slower sweeps from less anodic potentials (up to 1.1 V) analysis of the shapes of the reduction curves (fig. 1) indicates some participation of a second species particularly when effects of holding the anodic termination potential at a fixed constant value for various times are examined.Similar structure is observed with single crystal surfaces of Pt. The oxidation behaviour corresponding to the anodic peaks can be represented in terms of successive stages of oxidation corresponding to coverage of Pt sites by * The support of this research by the U.S. Ofice of Naval Research is acknowledged. L. Pauling The Nature ofthe Chemical Bond 3rd ed. (Cornell Univ. Press Ithaca 1960) p. 5 14. D. Gilroy and B. E. Conway Can. J. Chem. 1968,46 875. 96 GENERAL DISCUSSION " OH " then by " 0 " and finally phase oxide formation by place exchange amongst Pt and " 0 " species in the surface beyond ca. 1.1 V. Even at 1.1 V beyond the second main peak (fig. l) it is surprising that a charge of only ca. 1 electron per Pt atom has been passed.The successive stages of Pt oxidation therefore probably involve initially Pt + H,O-+PtOH + Hf + e ; followed by PtOH + P t - d + H+ + e. Since the appearance of peaks normally corresponds to half coverage by the species being electrochemisorbed over the potential range concerned a species (one 0 per 2 Pt sites) competing with OH for coverage of the Pt sites is required to account for the observation of the second peak already at a degree of surface oxidation corres- Pt \Pt ponding to only le per Pt site. *1 1.5 3 0) \ 8 1.0 3.5 0 FIG. 1.-Differential charging curves for oxidation and reduction of Pt electrode surface in cyclic voltammetry up to various anodic termination potentials. Integral oxidation charge curve is also shown in terms of the ratio Qo/QH as a function of anodic potential E (N.H.E.) 1 N H2S04 25°C.(b) The curves of fig. 1 also indicate a progressive irreversibility corresponding to hysteresis between the anodic and cathodic curves. The material initially laid down (first two curves of fig. 1) is evidently reversibly reduced and is probably simply an electrochemisorbed OH radical. However oxidation of the surface up to higher potentials is accompanied by increasing irreversibility. These effects are presumably connected with formation of the surface oxide in growing clumps and the reduction becomes more difficult as the clumps become larger and the fraction of edge sites at N. Sat0 and M. Cohen J. Electrochem. Soc. 1964 111 512. B. E. Conway N. Marincic D. Gilroy and E. J. Rudd J. Electrochem. SOC. 1963 113 1144. GENERAL DISCUSSION 97 which reduction may be more facile becomes smaller (cf.Everett's theory of hysteresis in sorption and desorption at micro-porous substances). The role of edge sites in reduction was suggested as a basis for electrochemical periodicity in reactions at Pt. Fig. 1 shows that reduction from the highest anodic termination potentials is a potential (N.H.E.) FIG. 2.-Changes of ellipsometric parameter A for surface oxidation of Pt referred to horizontal base line (dotted curve) or more correctly sloping bassline (full curve). A is shown as a function of potential (N.H.E.) and surface charge expressed as Qo/QH. 4 = 70.75 ; A = 6328 8 ; sweep rate = 100 mV s-l. relatively slow process and in the cathodic sweeps it is evident that an appreciable excursion to relatively cathodic potentials is required before the current in the cathodic sweep becomes itself cathodic.4 = 70.75' ; A = 6328 8 ; s = 100 mV s-' ; + 0.05 +1.30 V. FIG. 3.-Relation between changes of A in surface oxidation of Pt and differential charging behaviour for surface oxidation from cyclic voltammetry. D. H. Everett Trans. Faruduy SOC. 1954,50 187; 1955,51 1551. J. Wojtowicz N. Marincic and B. E. Conway J. Gem. Phys. 1968,48,4333. s4-4 98 GENERAL DISCUSSION Fig. 2 and 3 shows the relation between changes of A (6A) at Pt and the potential and surface oxidation charge Qo/QH examined by Dr. Laliberte in our laboratory No structure corresponding to fig. 1 is seen in this plot except beyond 1.05-1.1 V (le per Pt) where the (A V ) or (A QJQ,) relation takes a different slope. Also changes of A evidently exactly follow changes of Q determined electrochemically so that there is no basis for distinction of an initial adsorbed “ 0 ” species between 0.8 and 0.95 V from that formed at higher potentials as has been claimed and forcibly argued.2 The only optical distinction and this is a clear one is between the properties of the oxide layer below 1.05 V and above that potential.The state “ 02Pt ” (ix. one 0 per 2 Pt sites) therefore seems to be an electrochemically and physically significant one. These observations and conclusions are consistent with the ARIR results mentioned by Mclntyre and Kolb in this Symposium. Miss M. A. Barrett and Dr. R. Parsons (University of Bristol) (communicated) A study of the coulometric curves in Conway’s discussion remarks and those using both sweeps and stepped potential sequences by Icenhower et aZ.,3 as well as the various optical results suggests to us that the relatively sharp transition at about 1.1 V is typical of sweeps whereas only a slight curvature of the relevant quantities occurs if all oxidation is at a constant potential.For R, the change of slope during potential sweeps is greater than would be predicted by the coulometric curve especially at the longer wavelengths. Thus a plot against coulometrically-determined oxide would show the opposite trend from the delta curve shown by Conway. This confirms a change of optical properties. Another relevant observation i s that oxida- tion at an intermediate potential followed by oxidation at a higher potential yields a smaller dip in R than when the oxidation is carried out entirely at the higher potential. A. K. N. Reddy M. A. Genshaw and J. O’M. Bockris J . Chem. Phys. 1968,48,671. R. Greef J. Chem. Phys. 1969 51 3148 ; cf. M. A. Genshaw and J. O’M. Bockris. J. Cltem. Phys. 1969 51 3149. D. E. Icenhower H. B. Urbach and J. H. Harrison J. Electrochem. SOC. 1970,117,1500.