Effect of anions and pH on the adsorption and oxidation of methanol on a platinum electrode

Effect of anions and pH on the adsorption and oxidation of methanol on a platinum electrode

529 J. Electroanal. Chem., 330 (1992) 529-540 Elsevier Sequoia S.A., Lausanne JEC 05023 Effect of anions and pH on the adsorption and oxidation of ...

676KB Sizes 1 Downloads 47 Views

529

J. Electroanal. Chem., 330 (1992) 529-540 Elsevier Sequoia S.A., Lausanne

JEC 05023

Effect of anions and pH on the adsorption and oxidation of methanol on a platinum electrode * J. Sobkowski, K. Franaszczuk and K. Dobrowolska Department of Chemistry, Warsaw University, Zwirki i Wigury 101, 02-089 Warsaw (Poland) (Received 24 October 1991; in revised form 7 January 1992)

Abstract

The influence of various anions and pH on methanol oxidation as well as its products of chemisorption on platinum has been studied by voltammetric and radiometric methods. It was found that the rate of methanol oxidation from the bulk solution was lowest for Na,CO, and highest for NaOH solutions. The influence of anions on the chemisorbed species of methanol on a platinum electrode has also been investigated.

INTRODUCTION

The reaction of methanol on a platinum electrode is usually studied in sulphuric and perchloric acid solutions as supporting electrolytes, though studies in alkaline solutions are also numerous. A review of papers on the influence of the electrolyte composition on the electro-oxidation of small organic molecules was given by, Parsons and VanderNoot [l]. Older works on this problem were reviewed by Breiter [2]. The influence of anions on the behaviour of methanol on a platinum electrode can be considered as a mutual competition for the occupation of adsorption sites on the platinum surface. Hence the question arises about the energies of adsorption of various anions compared with that of methanol on a platinum electrode. The situation is relatively simple in halogen anion solutions. It was shown directly by the radiometric method [3] that the rate of methanol adsorption decreases with increasing Cl- ion concentration in the supporting electrolyte (0.5 M H,SO,), the adsorption ceasing completely at 2 X 10m3 M Cl-. In the presence

Dedicated to Professor Roger Parsons on the occasion of his retirement Southampton and in recognition of his contributions to electrochemistry.

l

0022-0728/92/$05.00

0 1992 - Elsevier Sequoia S.A. All rights reserved

from the University of

530

of 5 x 10m5 M Br- or I- no adsorbed species derived from methanol are observed on the platinum electrode surface. However, when bromate or iodate ions are added to the solution after chemisorption of methanol, the products of reactions between these ions and chemisorbed species are resistant to oxidation even at 1.2 V vs. the normal hydrogen electrode (NHE) [3]. Direct information on the adsorption of anions can be obtained from radiometric measurements. It was found that halide ions are adsorbed much more strongly on a platinum electrode compared with H,PO;, HSO, and ClO; ions [4]. However, even for chloride ions the process of adsorption is reversible with respect to the bulk concentration of Cl- ions and the potential [5]. According to Breiter [6], the influence of anions on the oxidation of chemisorbed species derived from methanol can be attributed to a displacement of discharge of water molecules to more positive potentials. This effect increases with increasing adsorbability of the anions. The fact that the adsorption of anions is a reversible process whereas the adsorption of methanol involves the destruction of the molecule must be taken into account when considering the competition for adsorption of anions and methanol on a platinum electrode. The oxidation current of chemisorbed methanol in various media can be a measure of the adsorption of anions on a platinum electrode. Snell and Keenan [7] studied the reaction of ethanol on a platinum electrode in HNO,, H,SO, and HClO, solutions and found that the peaks of the oxidation current are dependent on the kind of acid, being lowest in HNO, and highest in HCIO, solutions. It was concluded that NO; anions are bound more strongly to the platinum surface than CIO; ions. It would thus seem that perchloric acid should be the best medium for methanol oxidation, but it was shown [8] that in concentrated solutions the perchlorate anions are reduced to chloride ions which, even in very low concentrations, strongly poison the platinum electrode surface. Promising results were obtained in alkaline solution [9,10]. It is supposed that the intermediate products of methanol oxidation are not adsorbed on the platinum surface and therefore the oxidation current is high [l]. It was also suggested [ll] that in alkaline media the surface concentration of adsorbed hydroxyl, OHads, which is the oxidant of the chemisorbed residue of methanol, is higher than that in acidic solution and therefore an increase in the oxidation rate is observed. A different mechanism of intermediate oxidation in acidic and alkaline solutions was suggested by Petrii et al. [9]. The rate-determining step in acidic solution is the reaction of intermediates with OH,,, originating from the reaction H,O + OH,,, + H++ ewhereas in alkaline solution the reaction OH-+

OH,,, + e-

is the slowest step. In this work the electrosorption of methanol and the electro-oxidation chemisorbed species have been studied in various supporting electrolytes

of its using

531

voltammetric and radiometric methods. The aim of the work was to compare different media which could be used for the oxidation of methanol in fuel cells. However, both methods are of limited value when one considers the real conditions in fuel cells. EXPERIMENTAL

The surface concentration of methanol adsorbed species was determined by the radiometric technique using methanol labelled with 14C. The apparatus, now slightly modified, and procedure have been described elsewhere 1121. The construction of the cell allows one to exchange the supporting electrolyte under constant-potential control of the electrode and without contact with the atmosphere. The measurements were carried out in 1 M HCIO,, 0.5 M H,SO,, 0.3 M H,PO,, 1 M HCl, 1 M NaClO,, 0.5 M Na,SO,, 0.3 M NaH,PO,, 0.5 M Na,CO,, 1 M KC1 and 1 M NaOH. All solutions were prepared from p.a. grade reagents and Millipore (Milli-Q Plus) water. The roughness factor of the platinized electrode, determined according to the procedure described earlier 113,141, was in every experiment equal to 75 k 5. All potentials are referred to the NHE in 0.5 M H,SO,. The rate of polarization was usually equal to 50 mV s-l. The measurements were carried out at ambient temperature (20 f 2 ’ 0. At the beginning of each experiment, the freshly obtained electrode was cycled in the 0.5 M H,SO, solution until a stable voltammogram was obtained. From this voltammogram the real surface area of the electrode was determined. Then the H,SO, solution was removed and the electrode was cycled in the given electrolyte for at least 30 min. In experiments on the anodic oxidation of chemisorbed species of methanol a washing procedure was used. After the chemisorption process had been completed, the solution containing methanol was replaced by the supporting electrolyte being studied. Six washings using oxygen-free solution were sufficient to remove all organic material from the cell except for that adsorbed irreversibly on the electrode. If the electrode potential was kept constant within the range of potentials from 0 to 0.45 V and oxygen-free solution was used, the count rates were constant irrespective of the number of washings. This means that during this procedure the chemisorbed species of methanol were not removed from the electrode surface. Starting from a potential of 0.45 V, a slow decrease in the count rate could be observed owing to oxidation of the chemisorbed layer. RESULTS AND DISCUSSION

Acidic solutions

In Figs. la-d the voltammograms of the platinum electrode in HClO,, H,SO,, H,PO, and HCI solutions in the absence and presence of methanol in the solution are shown. On the voltammograms in HClO,, H,SO, and H,PO, three characteristic peaks of methanol oxidation can be seen. There are two peaks during the

532

250.

.

200. 150. 100.

I 0.0

I 0.0 0.2

0.4 0.6 0.6 POTENTIAL

1.0 IV

0.2

0.4

1.2-d

0.6

0.0

1.0

POTENTIAL

IOO

b-

1.2

I.4

IV

d

1

-50

E : 2 I

0.0

0.0

0.2

0.4

0.6

0.6

1.0

.I

0.2

04

0.6

0.6

POTENTIAL

1.0

1.2

IA

POTENTIAL

1.2

1.4

/V

IV

Fig. 1. Cyclic voltammograms for methanol oxidation on platinum electrodes in acidic solutions: (a) 1 M HC104, (b) 0.5 M H,SO,, (c) 0.3 M H3P0,, (d) 1 M HCI. Sweep rate 50 mV s-l, c,-~,~~ = 6X lo-’ M.

anodic sweep: one (I) at about 0.85 V, i.e. at the potential corresponding to the onset of platinum surface oxidation, and the second (II) at about 1.4 V, where methanol oxidizes on the PtO electrode. During the cathodic scan only one anodic peak (III) is seen at 0.65 V, i.e. at the potential where the reduction of PtO is completed. Peak I is a result of two competitive processes. The more anodic the electrode potential, the higher is the rate of adsorbed species oxidation, but simultaneously the electrode surface is more and more covered with PtO which is inactive in the adsorbate oxidation. Much higher energy (i.e. more positive potential) is required for direct oxidation of methanol on the oxidized platinum surface (peak II) [lo]. It should be noted that on the oxidized platinum surface methanol from the bulk solution is not adsorbed. Although the potentials of methanol oxidation were the same in HClO,, H,SO, and H,PO,, the resulting current densities were not (Table 11, e.g. for peak I they are equal to 375, 270 and 175 PA cm-*, even though the same methanol concentration was maintained throughout. This corresponds to the increase in adsorption energies of the anions (Cloy <

533 TABLE 1 Potentials (E,) and current densities (jr) of the peak I maximum of methanol oxidation from bulk solutions Solution

El /mV

j, /FA cm-*

1.0 M HCIO,

860 840 860 950 920 925 330 130

315 270 175 290 260 140 130 440

0.5 M H,SO, 0.3 M H,PO, 1.0 M NaCIO, 0.5 M Na,SO, 0.3 M NaH ,PO, 0.5 M Na,CO, 1.0 M NaOH

HSO; < H,PO; [5]) and confirms the previous data [6,7] on the influence of anions on methanol and ethanol oxidation. As can be seen in Fig. Id, in HCI solution no methanol oxidation current is observed, which means that chloride ions inhibit methanol oxidation completely. The platinum surface is fully covered by strongly adsorbed chlorides in this solution. Owing to the blocking of the platinum surface by chlorides, the oxidation of the platinum electrode commences at more anodic potential in comparison with the other supporting electrolytes whose anions are adsorbed more weakly on a platinum electrode (see Table 2). Neutral and alkaline solutions

In Figs. 2a-2d the voltammograms of the platinum electrode in various neutral electrolytes in the absence and presence of methanol are shown. As can be seen, the shapes of the voltammograms in all electrolytes studied, except for KC1 solution, are similar. However, in comparison with the corresponding acidic solutions the peaks of methanol oxidation are lower. The displacement of the oxidation peak in unbuffered neutral solution as compared with acidic solution cannot be excluded owing to the acidification of the interface during anodic polarization [15]. However, this effect can be neglected at the scan rate used. In KCI, as in HCl solution, chloride ions adsorbed strongly on the platinum surface prevent the oxidation of methanol from the bulk solution. In alkaline solutions (Fig. 3) only one peak of the methanol oxidation current is observed during the positive scan. This peak in NaOH solution (440 PA cme2> is higher than those in other supporting electrolytes. This may be a result of the lower accumulation of adsorbed intermediates on the electrode surface [l] and/or the higher surface concentration of OH,,, species, which act as an oxidant in the methanol oxidation process [ 111. Oxidation of the chemisorbed species of methanol

To study the oxidation of chemisorbed species derived from methanol, the methanol from the bulk solution was removed by the washing procedure. A typical

534

6

250

4: s >

200 150

;

100

g

50

II

,_.- . . . . . ..:

-50 -100

._

,-----_..

0.’

2 k! 9 70

d!l!L Ill

-

L -06

._.-’ ,,_..._.______

,I--.-,

._....--

0.2 0.4 0.6 0.8 1.0 I.2

-0.2

POTENTIAL

1.4

-0.6

-0.2 0

0.2 0.4 0.6 0.8

POTENTIAL

-0.6

12

1.4

-0.4 -0.2

0

0.2

/V

0.4

0.6

POTENTIAL POTENTIAL

1.0

IV

0.8

1.0

1.2

/V

IV

Fig. 2. Cyclic voltammograms for methanol oxidation on platinum electrodes in neutral solutions: (a) 1 M NaCIO,, (b) 0.5 M Na$O,, (c) 0.3 M NaH2P04, (d) 1 M KCI. Sweep rate 50 mV s-l, cCH,oH = 6x 1O-2 M.

b

250 200

-0.6

-0.4

-0.2

0

0.2

POTENTIAL

0.4 IV

0.6

0.0

i

1

-0.4

-0.2

0

0.2

POTENTIAL

0.4

0.6

0.8

1.0

IV

Fig. 3. Cyclic voltammograms for methanol oxidation on platinum electrodes in alkaline solutions: (a) 1 M NaOH, (b) 0.5 M Na,CO,. Sweep rate 50 mV s-l, cCHJoH = 6X 10m2 M.

53.5

POTENTIAL

1”

Fig. 4. Cyclic voltammograms of platinum in 1 M HCIO,: stationary voltammogram; - - -, first cycle after adsorption of methanol and washing procedure. Elds = 400 mV, cCHjoH = 5 X 10e3 M, sweep rate 50 mV s-r.

voltammogram illustrating the oxidation of chemisorbed species is shown in Fig. 4 (for the sake of clarity only the curve taken in HClO, solution is presented). Only a single anodic current peak is observed during the first positive scan (one potential sweep is enough to remove the adsorbed species from the electrode surface). It appears at the end of the double-layer region at a potential before the onset of the potential of platinum oxide formation. The corresponding data for various supporting electrolytes are given in Table 2. The results suggest that, except for the solutions containing chlorides, an adsorbed water molecule, probably substantially

TABLE 2 Potentials of oxidation of chemisorbed different solutions

methanol residues (I?,,) and of platinum itself (EPto) in

Solution

E,, /mV

EptO

1.0 M HCIO, 0.5 M H,SO, 0.3 M H3P0, 1.0 M NaCIO, 0.5 M Na,SO, 0.3 M NaH,PO, 0.5 M Na,CO, 1.0 M NaOH 1.0 M HCl 1.0 M KC1

750 750 830 650 650 630 130 -30

820 900 910 700 750 700 180 80 1050 1000

/mV

536 TABLE 3 Charges involved in the oxidation of chemisorbed

speciesfQ,,> in different solutions

Solution

Q, /PC cm-*

1.0 M HCIO, 0.5 M HrSO, 0.3 M HsPO, 1.0 M NaCIO, 0.5 M NasSO, 0.3 M NaH,PO, 0.5 M Na&O, 1.0 M NaOH

85 78 47 77 41 30 20 55

deformed (dissociated) owing to the strong influence of the platinum surface, is the oxidant. The charges involved in the oxidation of the chemisorbed organic species for various supporting electrolytes are given in Table 3. The influence of different anions on the surface concentration of chemisorbed species of methanol can be directly determined by the radiometric method. The procedure is illustrated in Fig. 5. The methanol chemisorbed layer has been

TIME/h

Fig. 5. Time dependence of adsorption of methanol on platinum. Sequence of procedures: (a) thermal background counting (no r4CHs0H in solution); (b) addition of r4CHs0H, E = 1200 mV (no adsorption); (c) Eads = 400 mV; (d) exclusion of CHsOH from solution (washing with 0.5 M H,SO,); (e) solution exchange (---, 1 M HCI; -, all others); (f) oxidation of methanol residues during cycling.

531

<

0

0.60

w

NaOH

Na,SO,

H,SO,

HCIQ

H,PO,

KCI

HCI

Fig. 6. Plot of peak potentials of oxidation of (0) methanol residues and (0) platinum itself vs. composition of solution.

obtained in 0.5 M H,SO, at a constant potential of the platinum electrode, E = 0.4 V. At this potential the surface concentration of methanol residue reaches a maximum. Replacement of the H,SO, solution by another supporting electrolyte, apart from those containing chloride ions, does not influence the surface concentration of adsorbed species. The small decrease in count rate during the washing procedure, shown in Fig. 5, is caused by the removal of labelled methanol from the bulk solution (compare the increase in the count rate a --) b and the decrease d + e). In’ the solutions containing chloride ions a slow, continuous decrease in count rate with time has been observed as a consequence of the displacement of the organic residue by adsorbing Cl-. We have not noticed any influence of the pH- of the solution on this process. The values of oxidation current peak potentials which appear in the studied solutions together with the corresponding potentials for the onset of platinum surface oxidation are shown in Fig. 6. All the data presented in Fig. 6 originate from the experiments in which methanol was adsorbed in 0.5 M H,SO, and the adsorbed species were oxidized immediately after the solution had been replaced by another solution containing some different anion under consideration. Only the case of the solutions containing chloride ions is exceptional, since the intermediates are oxidized at a potential higher than the potential of platinum surface oxidation (Fig. 7). Methanol oxidation from the bulk solution is usually preceded by adsorption, but this is not so if chlorides which block the electrode surface are present.

538

-100

0

0.2

0.4

0.6

0.8 E/V

1.0

1.2

I

1.4

first cycle after adsorption of methanol Fig. 7. Cyclic voltammograms of platinum in 1 M HCI: -, and electrolyte exchange; - - -, stationary state voltammogram. Sweep rate 50 mV s-l.

Apparently, the energy of methanol adsorption is too low to displace chlorides from the electrode surface. If an adsorbed species is formed in H,SO, solution and subsequently oxidized in HCl (or KC0 solution, its oxidation occurs at about 1.2 V, which suggests another mechanism for the reaction. The surface concentrations of chemisorbed species of methanol have been determined for different supporting electrolytes and potentials. The pertinent data are presented in Fig. 8. The results given in Fig. 8 are not consistent with the data of Table 3. One can expect that the higher the surface concentration of chemisorbed species, the higher should be the charge of their oxidation. Of course, this expectation can only be made on the assumption that the chemisorbed species derived from methanol are the same in all the solutions studied. The chemisorbed species of methanol can interact with anions (see e.g. ref. 3). The number of electrons involved in the oxidation of one adsorbed species increases from 1.5 for HClO, to 1.7 for NaOH solution. This suggests that the chemisorbed product is reduced more in NaOH than in HClO, solution. Another explanation is also possible. The surface concentration of adsorbed species is higher in H,SO, solution than in HClO,, whereas the charge of their oxidation is greater for HClO, than for H,SO,. Such inconsistency can be caused by the different dependence of the surface concentration of adsorbed species of methanol and anions on potential. The maximum of the methanol chemisorbed species concentration in H,SO, is located at about 0.4 V, whereas the surface concentration of HSO; ions (and other anions) increases with potential up to the potential of platinum surface oxidation, i.e. it reaches a maximum at about 0.8 V

539

-0.6

-0.4

-0.2

0.0 Potentiol

0.2

0.4

0.6

0.8

/V

Fig. 8. Potential dependence of chemisorbed methanol residue concentration on a platinum surface in (1) 0.5 M HZSO,, (2) 1 M HCIO,, (3) 0.5 M Na,SO,, (4) 0.3 M H,PO,, (5) 1 M HCI and (6) 1 M NaOH.

[4]. It is quite possible that the low coverage of the platinum surface by HSO; does not influence seriously the surface concentration of chemisorbed species of methanol. Finally, the values of r (Fig. 8) were determined step by step until a constant count rate at a given potential had been attained. During this time (the electrode being kept at low potential) the slow reduction of perchlorate ions to chlorides, which strongly poison the electrode surface, cannot be excluded [8]. Hence the values of r can be lower in HClO, than in H,SO, solution. CONCLUSIONS

Electra-oxidation of both methanol from the bulk solution and its chemisorbed species on the platinum electrode surface is affected by the identity of the supporting electrolyte. The adsorption of anions on the electrode surface competes with methanol chemisorption: the higher the energy of anion adsorption, the lower the current density of methanol oxidation from the bulk solution. Chlorides inhibit methanol oxidation completely. In neutral solution the rate of methanol oxidation is slower than in acidic solutions of corresponding anions. In NaOH solution the rate of methanol oxidation is fastest. The chemisorbed species oxidation also depends strongly on the kind of anions. The situation is complicated by the fact that the maximum concentration of chemisorbed species of methanol is located at about 0.4 V and is connected with the destruction of the molecule, whereas the adsorption of anions increases with increasing electrode potential up to the potential of platinum surface oxidation and, moreover, it is a reversible process. In HClO, solution the slow reduction (at low potential) of perchlorate anions to chlorides, which strongly poison the

540

platinum surface [8], is possible. Also, the interaction of chemisorbed species with anions cannot be excluded. Oxidation of organic adsorbed intermediates involves the participation of OH adsorbed species formed by the discharge of adsorbed H,O or, in alkaline solution, by the reaction OH--, OH,,, + e-. Only in the case of HCI (KCI) are the potentials of oxidation for both the intermediates and the electrode surface displaced distinctly to more positive values. It seems that water molecules, as oxidant, are absent or inactive on a platinum surface covered by chloride ions and that another oxidant formed at about 1.2 V participates in the methanol oxidation process. ACKNOWLEDGMENT

This work was supported (Contract No. 520).

by the State Committee

for Scientific

Research

REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

R. Parsons and T. VanderNoot, J. Electroanal. Chem., 257 (1988) 9. M.W. Breiter, Electrochemical Processes in Fuel Cells, Springer, Berlin, 1969. J. Sobkowski and A. Wieckowski, J. Electroanal. Chem., 41 (1973) 373. G. Horanyi, E.M. Rizmayer and G. Inzelt, J. Electroanal. Chem., 93 (1978) 183. G. Horanyi and E.M. Rizmayer, J. Electroanal. Chem., 83 (1977) 367. M.W. Breiter, Discuss. Faraday Sot., 45 (1968) 79. K.D. Snell and A.G. Keenan, Electrochim. Acta, 27 (1982) 1683. G. Horanyi and G. Inzelt, J. Electroanal. Chem., 86 (1978) 215. O.A. Petrii, B.I. Podlovchenko, A.N. Frumkin and H. Lal, J. Electroanal. Chem., 10 (1965) 253. L.D. Burke and K.J. O’Dwyer, Electrochim. Acta, 35 (1990) 1821. C. Lamy, Electrochim. Acta, 29 (1984) 1581. A. Wiqckowski, J. Electrochem. Sot., 122 (1975) 252. T. Biegler, D.A.J. Rand and R. Woods, J. Electroanal. Chem., 29 (1971) 269. A. Capon and R. Parsons, J. Electroanal. Chem., 45 (1973) 205. K. Franaszczuk and J. Sobkowski, J. Electroanal. Chem., 261 (1989) 223.