Generation of cyanogen from the decomposition of several nitrogen

Generation of cyanogen from the decomposition of several nitrogen-containing aromatics on platinum(111). J. R. Kingsley, and John C. Hemminger. Langmu...
3 downloads 0 Views 653KB Size
Langmuir 1986,2,460-464

460 I

2 00

1

3 00

LOO

Wavelength in m

I

Figure 8. Absorption spectra of PSA-Na+, solubilized in DDAC/cyclohexane, below (- - -) and above (-) the operational cmc.

pendent on the surfactants or water concentration, always exponential. The fluorescence decay curve of sodium 1pyrenesulfonate (PSA-Na+), however, is for several R values two exponential a t surfactant concentrations (S) below the operational cmc and becomes exponential above the cmc. Additionally for the latter probe, a strong broadening of the absorption spectrum is observed at low S, while a normal spectrum is obtained above the cmc (Figure 8). The ammonium group can only interact with the arene if a close contact is possible. This is only possible for NAA and PSA-Na+, and not for NBA. If the aromatic moiety of NBA should interact then the polar COOH group is too far away from the polar region of the reverse micelle, which is unfavorable for enthalpic reasons. Also, from quenching and energy transfer studied3it is known that the distance between the aromatic moiety of NBA and the polar micellar core is larger than in the case of NAA. The reason why a t S above the cmc the decay curve of PSA-Na+ becomes exponential (100% q,)while for NAA still 15% of the short decay time is present, is probably partly due to the fact that the interaction of the head group of DDAC with the ionic sulfonate group of PSA-Na+ is stronger than with the neutral acid group of NAA. (13)Gelade, E.; De Schryver,F. C. J. Am. Chem. SOC.1984,106,5871.

Conclusions The observed differences in 'H and 13CNMR and absorption spectra of NHCl and NMeCl can only be explained by specific interactions between the positive charged tetraalkylammonium group and the ?r-electrons of the naphthalene chromophore. Due to alkylation of the ammonium group the solvation and therefore the solubilization in water is decreased, resulting in intermolecular aggregate formation. From VPO and the two exponential fluorescence decay of NMeC1, an equilibrium between optimal stabilized dimers and monomers is proposed. The excited state of the dimer is only a few nanoseconds, due to the perturbation of the ?r-electron cloud by the positively charged ammonium group. Calculation of the activation energies of deactivation of both components corroborates the proposed interaction. As could be expected, the dimer formation is entropically unfavorable but occurs due to enthalpic stabilization. In inverse micelles, this kind of interaction enabled us to get information on the aggregation mechanism of ionic surfactant molecules in apolar media.12 In aqueous micellar solution a specific interaction between arenes and the ammonium head group of CTAB has often been proposed to explain the higher binding constant and the higher surface solubilization of arenes in this system compared to SDS solutions. This result suggests, however, that if such an interaction occurs it must be very weak, because the fluorescence decay curves of 1methylpyrene are perfectly exponential in both systems and the calculated decay times ~ o m p a r a b l e .Therefore ~ it is suggested that for aqueous CTAB micelles, compared to SDS micelles, the difference in interphase, water penetration, micellar structure and chain packing and folding are probably more important than the weak interaction between arenes and the ammonium head groups. Acknowledgment. We thank the NFWO and IWONL for a fellowship to E.G., A.V., and K.V. The FKFO, University Research Fund, ERO, and Agfa are thanked for financial support to the laboratory. S. Toppet and W. Vranckx are thanked for the NMR analysis and the VPO measurements, respectively. Registry No. NMeC1, 14564-86-4;NHC1, 552-46-5; NAA, 86-87-3;PSA-Na+,59323-54-5;NBA, 781-74-8;DDAC, 3401-74-9.

Generation of Cyanogen from the Decomposition of Several Nitrogen-Containing Aromatics on Pt(11 1) J. R. Kingsley and John C. Hemminger*t Department of Chemistry, University of California-Irvine, Irvine, California 9271 7 Received December 16, 1985. In Final Form: March 11, 1986 The thermal decomposition of several nitrogen-containingaromatics on Pt(111)results in desorption products of H2, HCN, and C2N2. C2N2from triazine decomposition desorbs at 740 K, which is the same temperature for C2N2desorption following adsorption of cyanogen. All of the other aromatics evolve C2N2 at a significantly higher temperature (830-990K) which corresponds to the onset of decomposition of the dehydrogenated aromatic ring. The similar temperature for C2N2production from triazine and cyanogen indicates a macrocyclic model of the B state of cyanogen adsorbed on Pt(ll1) which has been previously suggested is likely. 1. Introduction Recently there have been several studies of the bonding and chemistry of nitrogen-containing molecules on the 'Alfred P. Sloan Research Fellow.

0743-7463/86/2402-0460$01.50/0

surfaces of a variety of single-crystal metals. CH3CN,' CH3NC,l HCN,%3C2N2,1,2k-10 C2H8N2,I2 and C4N4H6"have (1) Hemminger, J. C.; Muethrties, E. L.; Somorjai, G. A. J.Am. Chem. SOC.1979, 101,62.

0 1986 American Chemical Society

Langmuir, Vol. 2, No. 4, 1986 461

Decomposition of Nitrogen-Containing Aromatics on Pt(ll1)

>

+ - -

to-

z w I-

21 H

w

> -a

1 W

a I

I

I

I

I

I

I

I

400

500

400

I

700 800 900 1000 1100 SURFACE TEMPERATURE ( K ) Figure 1. Typical TDS spectrum of cyanogen (mass 52) following adsorption of 20 langmuirs of cyanogen at 286 K on a Pt(ll1) 300

600

500

600

700

800

900 1000

1100

SURFACE TEMPERATURE ( K )

Figure 2. H2,HCN, and C2N2TDS spectra after adsorption of 1langmuir of quinoxaline at 293 K on Pt(ll1). Heating rate was 8 K/s.

crystal. Heating rate was 8 K/s.

been the objects of studies on various surfaces of Pt, Ni, and Cu. The thermal chemistry of C2N2has been studied previously on Pt(lOO),S$JOPt(ll0),26 Pt(lll),a12and a stepped Pt surface with Pt(ll1) terraces.' Only C2N2is observed as a desorption product from these surfaces (see Figure 1as an example of cyanogen desorption from Pt(ll1)). In all of these cases a t least two desorption states are seen (designated a and &). On several of these surfaces a second j3 state is seen and designated is2. The a state desorbs in the temperature range of 370 (for Pt(ll1))to 460 K (for Pt(ll0)). The PI state desorbs in the range 680-690 K. The is2 state desorbs in the range 750 K to above 800 K. The present interpretation of the a state is that it is due to molecularly adsorbed cyanogen. Two interpretations of the j3 states have been put forward. Bridge, Marbrow, and Lambert2 have suggested that the j3 states are due to a recombination of CN groups, while Netzer5 has proposed the existence of a paracyanogen-like overlayer. Similar high-temperature states are observed for cyanogen adsorption on other transition-metal surfaces such as Ni(lll).l Two suggestions have appeared in the literature for possible (CN)= polymeric structures for the j3 states of cyanogen on transition metals.lp5 The paracyanogen-like structure (1) has been suggested by Netzer et al., while

'."

I

11

1

NI'=' I

,$(

Hemminger et al. have suggested the existence of a cyan(2) Bridge, M. E.; Marbrow, R. A.; Lambert, R. M. Surf. Sci. 1976,57, 415. (3)Bridge, M.E.;Lambert,R. M.J. Catal. 1977,46, 143. (4) Solymosi, F.;Kiss, J. Surf. Sci. 1981,108,368. (5)Netzer, F. P.Surf. Sci. 1976,6I,343. (6)Bridge, M. E.;Lambert, R. M. Surf. Sci. 1977,63, 316. (7) Netzer, F. P.;Wille, R. A. Surf. Sci. 1978, 74,547. (8)Hoffman, W.; Batel, E.; Neteer, F. P. J . Catal. 1979,60,316. (9) Conrad, H.; Kuppers, J.; Nitschke, F.; Netzer, F. P. J. Catal. 1978, 52, 186. (10)Conrad, H.; Kuppers, J.; Nitschke, F.; Netzer, F. P. Chem. Phys. Lett. 1977,46,571. (11) Dahlgren, D.; Hemminger, J. C. Surf. Sci. 1982,120, 456. (12) Kingsley, J. R.; Dahlgren, D.; Hemminger, J. C. Surf. Sci. 1984, 139,417.

ogen trimer on Ni(ll1) which may be extended into a macrocyclic structure (2). In an effort to develop an understanding of the chemistry that should be expected for species like 1 and 2, we have studied the thermal' chemistry of a number of nitrogen-containing aromatic molecules which decompose a t high temperatures to give cyanogen when adsorbed on Pt(ll1) (see Figure 6 for the names and structures of the molecules studied in this report). The decomposition chemistry of triazine may be expected to be similar to that of macrocycle 2 whereas quinoxaline chemistry would be expected to be close to that of the paracyanogen structure 1. As we will see the cyanogen evolution from triazine is unique among the molecular adsorbates we have studied. The other chemical analogues to quinoxaline and triazine in this study will act as a check as to the uniqueness of the thermal chemistry of triazine. 2. Experimental Section All experiments were performed in a UHV chamber equipped with low-energy electron diffraction (LEED), Auger electron spectroscopy (AES), and a quadrupole mass spectrometer in conjunction with a custom-built data system, all of which have been described elsewhere.I2 In thermal desorption experiments up to 16 masses can be monitored simultaneously. All gas exposures were performed in the pressure range of 1 X lo4 to 5 X lo-' torr. Pressures were read directly from the ion gauge without correction for sensitivity relative to NZ. All reagents were purchased from Aldrich Chemical Co. AU reagents were used without further purification with the exception of quinoxaline. The quinoxaline showed signs of oxidation and was vacuumsublimed at room temperature to obtain a pure sample. The piperazine was handled under an inert atmosphere to avoid air contamination. All reagents were introduced into the vacuum chamber from a stainless steel gas manifold through sapphiresealed leak valves. If necessary, the manifold was heated to raise the vapor pressure of the reagent being used. After exposures were made, the system was allowed to pump for as long as an hour to reduce the background of each reagent, as all are very "sticky" molecules. 3. Results 3.1. Quinoxaline on Pt(ll1). Exposures of quinoxaline

were made a t 283 K. After adsorption no evidence was observed for ordering of the overlayer either before or after slow warming of the crystal. Figure 2 shows typical thermal desorption spectra (TDS) of H2, HCN, and C2N2following a 1-langmuir exposure of quinoxaline. The single cyanogen desorption peak at 960 K occurs a t a much higher temperature than any cyanogen desorption previously observed from a Pt surface. Con-

462 Langmuir, Vol. 2, No. 4, 1986

Kingsley and Hemminger

SURFACE TEMPERATURE (K) Figure 3. H2,HCN, and CzNzTDS spectra after adsorption of 1 langmuir of triazine at 273 K on Pt(ll1). Heating rate was 8

K/s.

300

500 600 700 aoo 900 1000 1 1 0 0 SURFACE TEMPERATURE ( K )

400

Figure 5. H2,HCN, and CzNzTDS spectra after adsorption of 2 langmuir of pyridazine at 300 K on Pt(ll1). Heating rate was 8 K/s.

600

700

800

900

1000

SURFACE TEMPERATURE (K)

Figure 4. Effect of strongly bonded oxygen on the TDS spectrum of cyanogen after adsorption of 2 langmuirs of triazine at 273 K on Pt(ll1): (A) adsorption on clean surface; (B)crystal heated to 873 K for 1 min in 1 X lo4 torr of O2 before adsorption of triazine; (C)subsequent run after crystal had been flashed once to 800 K. Trace C has been enlarged slightly so as to not obscure trace A. Heating rate was 8 K/s. firmation that the observed cyanogen desorption was coming from the crystal and not from the heating wires or other source was obtained by monitoring the nitrogen Auger signal from the surface. This desorption peak, along with those for H2 and HCN, does not show any temperature dependence on exposure. 3.2. Triazine on Pt( 11 1). Figure 3 shows typical desorption spectra for H2,HCN, and C2N2following exposure of 2 langmuirs of triazine at 273 K. The peak temperatures for all three products showed no exposure dependence. No evidence of ordering of the adsorbate was seen at any exposure level or upon a slow heating. Monitoring of the carbon and nitrogen Auger signals before and after desorption indicates that the triazine is completely removed from the surface with heating to 1300 K. The C272:N391 Auger signal ratio after exposure was found to be approximately 1:l. This ratio is higher than what would be expected (0.7) and may be indicative of some amount of electron beam damage to the triazine. In several experiments with triazine, a second cyanogen peak was observed a t 680 K. This second peak varied in size from run to run relative to the peak a t 740 K. It was found that this state is the result of residual amounts of oxygen in the bulk of the Pt crystal left over from an oxygen cleaning. Figure 4 shows the results of an oxygen cleaning exposure at 873 K and 1 X lo4 torr of oxygen for 1min and subsequent adsorption of triazine. It should be noted that in this experiment the crystal was not flashed to high temperature after the oxygen treatment to remove

residual oxygen as is typically done in cleaning of the crystal. As can be seen, residual oxygen causes total population of the 680 K peak. Subsequent exposures and desorptions from this crystal show a gradual shift of cyanogen population to the 740 K peak, leaving only a shoulder at 680 K. Careful cleaning and high-temperature annealing of the crystal resulted in complete elimination of this shoulder. H2and HCN were relatively unaffected by residual dissolved oxygen. Similar effects of impurities have been observed for C2N2adsorption.13 Coadsorption of triazine with D2resulted in no incorporation of deuterium into either the HCN or the hightemperature H2 peak. Preadsorption of 5 langmuirs of cyanogen with 2 langmuirs of triazine had little effect on the HCN desorption spectrum. The cyanogen desorption spectrum showed an additional peak a t 370 K, which is the same as is seen for simple desorption of just cyanogen on Pt(111).8J2The hydrogen spectrum displayed a loss of the low-temperature background peak. This hydrogen was most likely bound up by the cyanogen on the surface at the time of adsorption and was removed in the higher temperature H2 peak. This effect has been seen previously in cyanogen-hydrogen coadsorption experiments, where it is demonstrated that a partially hydrogenated surface species is created.12 3.3. Pyridazine on Pt( 111). Exposures of pyridazine were made at 300 K. Again, no ordered structures were observed following adsorption on Pt(ll1). Figure 5 shows typical thermal desorption spectra (TDS) of H2,HCN, and C2N2following a 2-langmuir exposure of pyridazine. As in the case of quinoxaline the C2N2desorption peak is observed at a temperature (833 K) significantly above the P state for cyanogen adsorption on Pt(ll1). An interesting side note is that we did not observe nitrogen desorption following adsorption of pyridazine. This is in marked contrast with results for the decomposition of dimethyltetrazine on Pt(ll1) which results in both low-temperature N2 production, indicative of direct release of molecular nitrogen during the decomposition, and a higher temperature N2 product, indicative of N atom recombination." 3.4. Other Nitrogen Aromatics on Pt(ll1). The results just described for quinoxaline and pyridazine decomposition on Pt(ll1) are characteristic of what we have observed for a variety of similar nitrogen-containing aromatic analogues adsorbed on Pt(ll1). Figure 6 collects the results we have obtained for the decomposition of pyri(13) Kingsley, J. R.; Lloyd, K. G.; Hemminger, J. C., unpublished results.

Decomposition of Nitrogen-Containing Aromatics on Pt(ll1) CYANOGEN TRIAZINE

CZNZ

Jr:

PYRIMIDINE

0

PIPERAZINE

[:I

PYRAZINE

300 400

500

600

700

800

900

SURFACE TEMPERATURE

1000 1100

(K)

Figure 6. Desorption temperature and relative abundance for Hz, HCN, and CzNz for each of the molecules in this study. Relative abundances are indicated by the height of the bar and are not corrected for sensitivity and transmission factors. The relative amounts are normalized to the largest peak for each molecule but not to each other. midine, pyrazine, piperazine, and quinazoline along with the results just described for pyridazine, quinoxaline, triazine, and cyanogen. Figure 6 indicates the peak temperatures and relative abundances for the decomposition products H2, HCN, and C2N2. It is clearly seen that the molecules we have studied can be put into three classes: single-ring compounds with two nitrogens, naphthalenic compounds with two nitrogens, and triazine. All of the single-ring compounds have very similar decomposition behavior. Piperazine, the saturated analogue of pyrazine, undergoes dehydrogenation a t low temperatures (possibly upon adsorption). After the initial low-temperature dehydrogenation, the chemistry of piperazine is nearly identical with that of pyrazine. Quinoxaline and quinazoline show H2, HCN, and C2N2production a t very similar temperatures. Both of these classes of molecules result in C2N2 production a t temperatures significantly above that observed for @ state cyanogen adsorption on Pt(ll1). In contrast, triazine decomposes a t a significantly lower temperature resulting in C2N2production a t 740 K. This is the same temperature as that for the p state of cyanogen. 4. Discussion The high temperature a t which cyanogen is released from quinoxaline and quinazoline adsorbed on Pt(ll1) suggests that the naphthalene-like ring structure of these compounds is extremely stable, even when it has been dehydrogenated. Loss of cyanogen is not the only method by which this stable ring structure is broken. As is seen in sections 3.1 and 3.3, a significant amount of HCN is seen to evolve during the course of the experiment. The evolution of HCN occurs in direct competition with loss of hydrogen a t around 700 K. When all of the hydrogen has been lost, ring cleavage is no longer favorable and the molecules that have not lost their CN groups are stable up to 960 K where cyanogen is produced. The stability of the aromatic backbone of quinoxaline and quinazoline is further evidenced by two additional experiments. First we have adsorbed quinoxaline on a Pt(ll1) surface with small amounts of residual oxygen in the crystal. Under these conditions CO is produced and observed as a sharp peak in the TDS spectrum of mass 28 coincident with the cyanogen (960 K). Similar production of CO is observed when naphthalene is adsorbed on a partially oxidized Pt(ll1)surface. In the naphthalene case the CO is observed a t a temperature of 925 K. Thus we believe that the high-temperature cyhnogen from qui-

Langmuir, Vol. 2, No. 4, 1986 463 noxaline is produced at a temperature that is characteristic of the decomposition of the naphthalene-like aromatic ring structure which has been completely dehydrogenated a t temperatures below 800 K (see Figure 2 for the temperature of hydrogen evolution). After decomposition of the aromatic ring the carbon residue which is left behind forms graphite islands, as evidenced by observation of rings in the LEED pattern characteristic of graphite overlayers. Another indication of the stability of the quinoxaline backbone is the high temperature a t which HCN is evolved. Both triazine and the adsorption of HCN directly on Pt(ll1) give HCN desorption at a temperature of 510 K.3 This is also true of HCN production from dimethyltetrazine." This is in contrast to the higher temperature (740 K) required to produce HCN from quinoxaline and quinazoline. If Netzer's suggestion of a paracyanogen-like overlayer 1 is correct, one would envision chemistry similar to that of the cyanogen group in quinoxaline. That being the case, one would expect that the desorption temperature for cyanogen from such a structure might resemble that for quinoxaline. This is not what is observed. As was mentioned in the Introduction, the desorption state of cyanogen from Pt(ll1) is observed a t 680 K. Similar behavior has been observed for the single-ring compounds which contain only two nitrogens. The decomposition of the ring structure is first indicated by HCN production a t temperatures above 625 K, and C2N2is not observed until temperatures above 830 K (well above the temperature of the cyanogen p state). Both of these classes of molecules are in direct contrast to triazine. The molecule triazine provides a model of the repeating unit of the macrocyclic model 2 for the p state of cyanogen in which the hydrogens replace the connecting bonds to other units. The bonding in triazine and the macrocycle shown earlier is such that formation of cyanogen must result from either decomposition to isolated CN followed by recombination or a concerted formation of cyanogen involving CN groups from different triazine-likeunits. Our results with triazine show that cyanogen is observed at the same temperature as the p state produced by cyanogen adsorption. The desorption spectra following triazine exposure suggest a relatively simple picture of the thermal chemistry of triazine on Pt. Triazine is thought to be molecularly adsorbed on Pt(ll1) a t room temperature. Upon heating, two hydrogen atoms are removed a t 485 K. At slightly higher temperatures a molecule of HCN is split out of the aromatic ring. Finally at 740 K, cyanogen is desorbed a t the same temperature as the p state of cyanogen. 5. Summary The production of C2Nz from the naphthalene-like molecules quinoxaline and quinazoline as well as from the single-ring compounds with two nitrogens occurs a t temperatures significantly above the temperature of the cyanogen p state. Of the nitrogen-containing molecules that we have studied on Pt(ll1) only triazine results in C2N2desorption a t a temperature similar to that of the cyanogen 0 state. We interpret the evolution of cyanogen from quinoxaline and quinazoline at 960 K as the onset of fragmentation of the naphthalene-like aromatic ring structure after complete dehydrogenation. The extremely high desorption temperature (960 K) of cyanogen observed from quinoxaline adsorbed on Pt(ll1) suggests that the paracyanogen (1) model for the 0 state of cyanogen on Pt(ll1) is unlikely. On the other hand, the desorption temperature of cyanogen from triazine (similar structurally to the macrocyclic

464

Langmuir 1986,2,464-468

model for the /3 state) is identical with that of the /3 state. I t is, however, possible that this is still due to the recombination of isolated CN groups formed after initial release of HCN. Study of these systems by surface sP~troscoPic probes will be necessary to develop a clearer picture of the cyanogen /3 state.

Acknowledgment. This work has been supported by the Office of Naval Research. Registry No. C2N2,460-19-5;Pt, 7440-06-4;triazine, 290-87-9; pyridazine, 289-80-5;pyrimidine, 289-95-2;piperazine, 110-85-0; pyrazine, 290-37-9; quinazoline, 253-82-7; quinoxaline, 91-19-0; dimethyltetrazine,1558-23-2.

Carbon Monoxide Adsorption on a Platinum Electrode Studied by Polarization-Modulated FT-IR Reflection-Absorption Spectroscopy. 2. CO Adsorbed at a Potential in the Hydrogen Region and Its Oxidation in Acids K. Kunimatsu,t H. Seki,* W. G. Golden, J. G. Gordon 11, and M. R. Philpott IBM Almaden Research Center, San Jose, California 95120-6099 Received December 20, 1985 The carbon monoxide layer on a platinum electrode, which is adsorbed at 0.05 V relative to a normal hydrogen electrode (NHE) in 0.5 M sulfuric acid, and ita oxidation to carbon dioxide at higher electrode potentials have been studied by electrochemicaland in situ Fourier transform infrared reflection-absorption spectroscopy (FT-IRW). Polarization-modulatedFT-IRRASwas used to measure the vibrational spectra of adsorbed carbon monoxide as well as the evolved COz as a function of electrode potential. It is shown that the dominant surface species is linearly adsorbed CO but that the bridge-bonded species is oxidized first at about 0.20 V, giving rise to a decrease in the linear C-O stretching frequency along with a broadening of the band. Oxidation of the linearly adsorbed CO begins at 0.35 V, producing a further, sharp decrease in the C-0 stretching frequency as well as a considerable broadening of the band. It is concluded that the oxidation of the CO adlayer produced at 0.05 V occurs randomly throughout the layer, and not on the edges, which is characteristic of CO adsorbed at 0.4 V. It is proposed that the difference in behavior of these two kinds of adsorbed CO is due to crystallographic modification of the platinum surface when the CO is adsorbed at 0.05 V in the hydrogen region resulting in a higher density of bridge-bonded CO.

Introduction In a previous paper we reported an in situ IR spectroscopic study of CO adsorbed on a smooth platinum electrode in acids, the adsorption being initiated a t a potential of 0.4 V (NHE) in the double-layer The bandcenter position, intensity, and line width for the C-0 stretch were measured as a function of electrode potential. The potential was varied from ca. 0.05 V (NHE) well into the oxidation region. The evolution of C02 during oxidation was also monitored spectroscopically. It was concluded that the linearly adsorbed CO (linear CO(a)) was the predominant surface species, the concentration of bridge-bonded CO species (bridged CO(a)) being very low, consistent with other electrochemical studies of CO on both p o l y c r y ~ t a l l i n eand ~ ~ ~ i n g l e - c r y s t a lplatinum ~~~ electrodes. On the basis of the spectroscopic data obtained it was concluded that the electrooxidation of the adsorbed CO layer occurred primarily a t the edges of CO islands and involved water molecules adsorbed adjacent to the CO sites. The present study focuses on the adsorption and oxidation of CO on the platinum electrode when initial CO adsorption occurs in the hydrogen potential region. There have been some differences in the 1iterature"l4 concerning the nature of the surface species produced when CO is adsorbed at a potential in the hydrogen region. Additional information regarding the structure of this adlayer can now + IBM visiting scientist (1983-1984); on leave from the Research Institute for Catalysis, Hokkaido University, Sapporo 060, Japan.

0743-7463/86/2402-0464$01.50/0

be obtained by using the polarization-modulated FT-IRRAS.'

Experimental Section The technique of polarization-modulated (PM)FT-IRRAS as employed in this work and the experimental details in relation to the electrochemical cell have been described in previous rep0rts.'~J~9'~ The electrolyte is 0.5 M sulfuric acid prepared from the organic-free water (Barnstead) and sulfuric acid (double distilled from Vycor, GFS Chemicals). The platinum electrode was prepared by sealing a 99.95% pure Pt electrode disk (20mm (1) Golden, W. G.; Kunimatsu, K.; Seki, H . J.Phys. Chem. 1984,88, 1275. (2) Kunimatsu, K.; Golden, W. G.; Seki, H.; Philpott, M.R. Langmuir 1985, 1, 245. (3) Breiter, M.W. J. Electroanal. Chem. 1979, 101, 329. (4) Gilman, S. J. Phys. Chem. 1963, 67, 78. (5) Brummer, S. B.; Ford, J. I. J. Phys. Chem. 1965, 69, 1355. (6) Beden, B.; Blimes, S.; Lamy, C.; Leper, J. M.J.Electroanal. Chem. 1984,164, 129. (7) Furuya, N.; Motoo, T. Proceedings of the Fall Conference of Japan Electrochemical Society, 1982; p 16. (8) Czerwinskin,A.; Sobkowski, J. J. Electroanal. Chem. 1978,91,47. (9) Breiter, M. W. J.Phys. Chem. 1969, 73, 3283. (10) Breiter, M. W. J. Phys. Chem. 1968, 72, 1305. (11) Breiter, M.W. J.Electroanal. Chem. 1985, 180, 25. (12) Grambow, L.; Bruckenetein, S. Electrochim. Acta 1977,22,377. (13) Stonehart, P. Electrochim. Acta 1973, IS,63. (14) Kazarinov, W. E.; Andreev, W. N.; Tysyachnaya, G. J. Elektrokhimiya 1972,6, 927. (15) Golden, W. G.; Saperstein, D. D. J. Electron Spectrosc. 1983,30, 43. (16) Seki, H.; Kunimatsu, K.; Golden, W. G. Appl. Spectrosc. 1985, 39, 437.

0 1986 American Chemical Society