revue roumaine de chimie (roumanian journal of ...

2 downloads 0 Views 333KB Size Report
Anodic behaviour of Ag in KOH solutions was investigated, at different KOH concentrations and temperatures, using cyclic voltammetry and chronoamperometry, ...
ACADEMIE R0UMA1NE (ROUMANIAN ACADEMY)

RE V U E R O U M A I N E DE C H I M I E (ROUMANIAN JOURNAL OF CHEMISTRY) TIRAGE A PART (REPRINT)

Tome 42 N° 6 Juin 1997

Vol. 42 Issue 6 June 1997

EDITURA ACADEMIEI ROMÂNE BUCUREŞTI

Revue Roumaine de Chimie, 1997, 42 (6), 429-434

ANODIC BEHAVIOUR OF SILVER ELECTRODE IN KOH SOLUTIONS DANIELA BUGHILF, MIHAI BUDA6 and LIANA ANICĂF “ Research and Design Institute for Electrical Engineering, ICPE-SA Bucharest, Splaiul Unirii 313, 74204, Bucharest, Roumania b Faculty o f Industrial Chemistry, Department of Applied Physical Chemistry and Electrochemistry, Calea Griviţei 132, 78122, Bucharest, Roumania

Received June IS, 1995

Anodic behaviour o f Ag in KOH solutions was investigated, at different KOH concentrations and temperatures, using cyclic voltammetry and chronoamperometry, as well as polarization curves. The obtained data were com­ pared to those existing in literature, and the mechanism o f Ag oxidation was verified. At low scan rates, the two main oxidation steps o f Ag were both found to be diffusion limited, having different activation energies. At KOH concentration between 4-7 M, Ag-O was found to have a minimum solubility at all studied temperatures.

INTRODUCTION Silver/silver oxide electrodes in alkaline solution have been extensively investigated because o f great interest met in the applications in high power batteries. In spite o f this, however, there are still uncertainties about the mechanism o f anodic formation and reduction o f Ag. Several mechanisms have been proposed,1-5 but none o f them can explain the complex behaviour o f silver in alkaline solution. In the anodic behaviour o f silver in alkaline solutions, there are two major oxidation steps: the first from Ag to Ag20 and the second from Ag20 to AgO. However, the first step is still not very clear, whether it implies the formation of a single layer of A g ^ preceded by OH- adsoiption, or the forma­ tion o f two successive layers o f Ag2Q. Some papers reveal that in the first step also takes place a partial dissolution o f Ag20 into Ag(OH)2 ions,6’7 other authors considering that this process is unimportant.3,4 The second oxidation step, from Ag20 to AgO was found to take place through a nucleation-growth process on the Ag20 substrate,3’4,8 being limited by the Ag+ diffusion into the A g ^ film. Some authors suggested that there exists another oxidation step, occurring in the potential region in which oxygen evolves, that implies the further oxidation o f AgO to Ag20 3.11,12 The present paper deals with the tem perature and OH concentration dependencies on this complex behaviour of Ag in alkaline solutions and intends to furnish more data for a better understand­ ing o f the above-mentioned processes.

EXPERIMENTAL The anodic behaviour of Ag in alkaline media was studied using a classical three-electrode cell, having the working electrode consisting of 99.99% Ag, a platinum counterelectrode and a saturated calomel (SCE) as reference electrode. The working electrode was a silver bar, incorporated in an insulator resin, having the apparent active area of 0 38 cm- Before each experiment, the silver electrode was polished with 600 emery paper, rinsed with distilled water, then * ished with ethanol, and then rinsed again with distilled water. The platinum counterelectrode was washed in nitric acid, -.nsed with distilled water and then dried prior to each determination. KOH solutions of 0.5 M, 1 M, 4 M, 7 M, 10 M (the exact concentrations were analytically determined by H N 03 0.1 N •.nation) and temperature values of 20, 30, 40, 50 and 60°C were used. The experiments were carried out using a PAR 173 Potentiostat, having a scan rate o f 200 mV/min, which is slow enough for the polarization curves to be considered in steady

430

Daniela Bughiu et al.

state conditions. The influence o f dissolved 0 2 was considered negligible, so that no inert gas was bubbled through the solutions. In order to record cyclic voltamogrammms, potential sweep rates between 0.109-10.9 mV/s were used. Chronoamperometric experiments were made for potential values, corresponding to the two main oxidation processes.

RESULTS AND DISCUSSION In the range of 0.5-10 M KOH concentration and 20-60°C temperature, the following experi­ ments have been performed: - polarization curves - cyclic voltammetry - chronoamperometry.

Polarization curves Anodic polarization curves were performed for KOH solutions in the range of 0.5-10 M and 20~60°C. Fig. 1 shows an example o f such a polarization curve at constant temperature for one KOH concentration. The polarization curve’s shape is similar for all studied temperatures, with small diffe­ rences that will be further discussed.

E (V) vs. SCE Fig. 1. - Typical polarization curve for Ag in KOH solutions.

From these curves one can see that in the anodic range o f potential two current peaks appear, which can be attributed to the Ag20 and AgO species. The first current peak i , appears at about 200-300 mV/SCE for the studied concentration and temperature range. This value (Epl) shifts towards negative values as the KOH concentration increases from 0.5 M to 10 M. The temperature does not influence significantly the potential corresponding to i Thus, E , is 310 mV/SCE at 20°C and 300 mV/SCE at 60°C. At constant temperature, for example Jb°C, Epl is 330 mV/SCE at 0.5 M KOH and 250 mV/SCE at 10 M KOH. The second current peak, denoted as ip2 appears in the 400-600 mV/SCE range (Ep2). Ep? shifts towards negative values as the concentration increases, and the temperature produces no significant influence on E r The current peak shows a maximum value at concentrations o f 4-7 M KOH, after which a small decreasing is observed. The first current peak shows a small linear dependence as concentration is increasing, a ten­ dency that is noticed for all studied temperatures. The values o f ipl are presented in Table 1 for all con­ centrations and temperatures used in the experiment.

431

Anodic behaviour o f silver electrode Tablei Current peaks for the first oxidation step of Ag in KOH solutions ip,, mA cm 2 temperature

l

20 30 40 50 60

0.5 M

1M

4M

7M

10 M

2.65 2.62 2.08 2.96 3.90

2.65 2.91 2.08 3.77 6,11

3.12 4.83 4.16 4.68 6.76

3.25 3.06 3.90 5.07 5.46

3.90 3.90 4.03 5.07 5.72

The second peak presents a maximum at concentrations of 4 -7 M KOH, as it was shown abo\ e; this could mean that the second step oxidation o f Ag20 to AgO is not influenced by the diffu­ sion of OH" ions through the film but by A g+ diffusion, which is consistent with other results.3,4 The current values for the second peak are presented in Table 2. Table 2 Current peaks for the second oxidation step o f Ag in KOH solutions i j , mA cm 2 Temperature °C 20 30 40 50 60

0.5 M

1M

4M

7M

10 M

4.40 2.62 6.30 2.96 10.30

6.71 2.91 13.10 3.77 17.32

11.10 4.83 18.04 4.68 28.87

10.43 3.06 15.79 5.07 29.62

8.42 3.90 13.78 5.07 25.90

The fact that the OH concentration does not significantly influence peak 1 can be explained by a competitive oxidation o f Ag either with OH- ions or water molecules, which is a normal assumption for anodization in aqueous solutions. The maximum for peak 2 can be explained by a minimum solubil­ ity of AgzO in KOH solution in the 4 -7 M range.9 Outside this concentration range the solubility of AgjO is higher, so the second peak is diminished. From the ratio between the charge involved in the second and fust peak, respectively, that was found about 0.91, one can see that the second peak corresponds to an entire oxidation o f the Ag20 formed in the first step, except, probably, the first monolayer anodic oxide that is formed in the first moments of Ag oxidation.10

Cyclic voltammetry The mechanism for the anodic oxidation o f A g+ in KOH media is not well understood yet, several papers suggesting different mechanisms, including diffusion determined steps and nucleation and growing ones. Probably, the m ost popular and understandable mechanism is that proposed by Becerra, Salvarezza and Arvia,4 also in good agreement with our experimental data. However, develop­ ing a mechanism for anodic oxidation o f Ag in KOH media, explaining all the experimental results, is an extremely difficult task because o f the systems’ complexity. Figure 2 presents cyclic voltamograms for Ag electrode in KOH solutions. They have relative­ ly similar shapes for all experimented concentrations and temperatures, with different magnitudes in the current peak. Three peaks are noticed: the first peak is assigned to OH- ions adsorption process with simultaneous formation o f an Ag20 monolayer (OH “ adsorption usually appears during anodic oxidanon o f noble metals; some authors suggest that in this step Ag forms a soluble species AgO~ 6,?).

Daniela Bughiu et al.

432

E (V) vs. SCE

E (V) vs. SCE Fig. 2a. - Typical c yc lov o ltanun o gram for Ag in KOH solutions; v = 0.0109 V s

Fig. 2b. - Typical cyclovoltammogram for Ag in KOH solution; v = 0.109 V s~'.

Several scientists affirm that a monolayer o f AgOH may form ,13-15 but unstable in aqueous solution, passing into AgzO. The possible reactions might be: OH" => OH ad

(I)

Ag + OHad => Ag20

(II)

Ag20 monolayer is a compact one (that was proved by SEM measurements) and has semicon­ ducting properties. The second peak corresponds to new Ag20 layers formation on that already existent, either because of QH~ ions, or o f water. The possible reactions for the second process might be: Ag + OH => AgzO

(III)

Ag + H20 => Ag20

(IV)

There are many discussions in literature about Ag20 dissolution as Ag(OH)2, but according to our experimental one may conclude that this process may be important especially at relatively great values o f KOH concentration and temperature. Ag20 + 4 OH- => 2 Ag(OH)2 + O2

(V)

The third peak, very narrow and prominent, corresponds to Ag20 oxidation to AgO for the deposited layer on the electrode; then, the oxygen evolution occurs. Ag20 + OH" => 2 AgO + H+

(VI)

Daniela Bughiu et al.

432

E (V) vs. SCE

E (V) vs. SCE Fig. 2a. - Typical cyclovoltammogram for Ag in KOH solutions; v = 0.0109 V s“

Fig. 2b. - Typical cyciovoltammogram for Ag in KOH solution; v = 0.109 V s-1.

Several scientists affirm that a monolayer o f AgOH may form ,11-15 but unstable in aqueous solution, passing into AgzO. The possible reactions might be: OH" => OH ad Ag + OHad => Ag20

(I) (II)

Ag20 monolayer is a compact one (that was proved by SEM measurements) and has semicon­ ducting properties. The second peak corresponds to new Ag20 layers formation on that already existent, either because of OH" ions, or o f water. The possible reactions for the second process might be: Ag + OH => Ag20

(III)

Ag + H20 => Ag20

(IV)

There are many discussions in literature about Ag20 dissolution as Ag(OH)2, but according to our experimental one may conclude that this process may be important especially at relatively great values o f KOH concentration and temperature. Ag20 + 4 OH" => 2 Ag(OH)2 + O,2-

(V)

The third peak, very narrow and prominent, corresponds to Ag20 oxidation to AgO for the deposited layer on the electrode; then, the oxygen evolution occurs. Ag20 + OH- => 2 AgO + H+

(VI)

Daniela Bughiu et al.

432

E (V) vs. SCE

E W vs- SCE

Fig. 2a. - Typical cyclovoItammogram for Ag in KOH solutions; v = 0.0109 V s_l.

Fig. 2b. - Typical cyclovoltaroxnogram for Ag in KOH solution; v = 0.109 V s- '.

Several scientists affirm that a monolayer o f AgOH may form ,13-15 but unstable in aqueous solution, passing into Ag20 . The possible reactions might be: OH- => O H ^

(I)

Ag + OHad => Ag20

(II)

AgzO monolayer is a compact one (that was proved by SEM measurements) and has semicon­ ducting properties. The second peak corresponds to new Ag20 layers formation on that already existent, either because of OH- ions, or o f water. The possible reactions for the second process might be: Ag + OH- => Ag20

(III)

Ag + H20 => Ag20

(IV)

There are many discussions in literature about Ag20 dissolution as Ag(OH)2, but according to our experimental one may conclude that this process may be important especially at relatively great values o f KOH concentration and temperature. Ag20 + 4 OH" => 2 Ag(OH)" + O2"

(V)

The third peak, very narrow and prominent, corresponds to Ag20 oxidation to AgO for the deposited layer on the electrode; then, the oxygen evolution occurs. Ag20 + OH" => 2 AgO + H+

(VI)

433

Anodic behaviour o f silver electrode

There exists information on Ag2O ţ formation in the potential domain corresponding to oxygen evolution, that might explain the supplementary subpeak that was noticed on the obtained voltamogramms. 2 AgO + OFT

A gjO j + H+

(VII)

Cyclic voltamogramms were also made at different potential sweep rates. Dependence o f ip versus Vv is linear only at small sweep rates (for both main peaks). This means that a diffusion control occurs on this domain and at higher rates a limitation is possible because o f ion transfer through layer. On the other side, the slopes ip/>/v for the second peak have a maximum at concentrations between 4 and 7 M; that means that at concentrations greater than 4 M an increased dissolution o f Ag20 as Ag(OH)~ takes place, with an adequate peak diminishing, especially at high temperatures. The values of ip/v 1/2 as a function of electrolyte concentration are presented in Table 3. Table 3 ip v_1/2 constants for the two main oxidation processes o f Ag in KOH solutions (in mA cm 2 V 1/2 s112)

(M)

t = 20°C

\ r ,/2 P2 t = 20°C

0.5 1 4 7 10

51.06 33.04 71.98 42.91 64.88

69.63 73.84 142.66 128.34 167.86

c

i

*pi V_I/2 t = 40°C

*p2 '/ ~ m t = 40°C

134.81 120.29 123 103.68 172.13

265.62 302.46 340.57

'p i v _ l /2 t = 60°C

166.52 150.18 227.62 207.63 285.87

-

364.91

i

P2 =

t

v ~ 1/2

60°C

455.74 497.34 697.34 561.81 565.38

Using the linear dependencies ln (y V 1/2) = f(l/T ), the activation energies o f the two main processes have been calculated, as is shown in Table 4. They correspond to a diffusion process, accord­ ing to their magnitude order, but with relatively different values for the two main processes. This fact might be assigned to A g+ and Ag2+ ions diffusion and not to the OH- ones. Table 4 Activation energies for the two main oxidation processes of Ag in KOH solutions (in cal per mole). The number in parenthesis represents the correlation coefficients of the experimental data C (M) 0.5 1.0 4.0 7.0 10.0

Peak 1 2470 4200 3400 4600 4260

(0.910) (0.866) (0.995) (0.978) (0.949)

Peak 2 5350 (0.929) 5400 (0.918) 4650(0.981) -

3500 (0.956)

Chronoamperometry Chronoamperogramms at potential values corresponding to the main peaks are also recorded and are presented in Figs 3 and 4. Their shape is the same for the whole experimented temperatures and concentrations domain. Chronoamperogramms corresponding to the first peak have a classical shape, that is characte­ ristic to those processes when a layer is deposited on electrode surface, though i - 11/2 dependence is not tspected. For the second peak, chronoamperogramms shape shows that the oxidation process o f AgjO to AgO takes places through a nucleation-growth mechanism. When the potential is more positive, fee —ix;m um current corresponding to nucleation-growth process appears earlier and is more pronounced.

433

Anodic behaviour o f silver electrode

There exists information on Ag20 3 formation in the potential domain corresponding to oxygen evolution, that might explain the supplementary subpeak that was noticed on the obtained voltamogramms. 2 AgO + O H ' => A gjO j + H+

(VII)

Cyclic voltamogramms were also made at different potential sweep rates. Dependence o f i versus Vv is linear only at small sweep rates (for both main peaks). This means that a diffusion control occurs on this domain and at higher rates a limitation is possible because o f ion transfer through layer. On the other side, the slopes ip/Vv for the second peak have a maximum at concentrations between 4 and 7 M; that means that at concentrations greater than 4 M an increased dissolution o f AgjO as Ag(OH)~ takes place, with an adequate peak diminishing, especially at high temperatures. The values of ip/v !/2 as a function of electrolyte concentration are presented in Table 3. Table 3 ip v~1/2 constants for the two main oxidation processes of Ag in KOH solutions (in mA cm 2 V 1/2 slr2) i

(M)

t = 20°C

i P2 \ r ' /2 t = 20°C

0.5 1 4 7 10

51.06 33.04 71.98 42.91 64.88

69.63 73.84 142.66 128.34 167.86

c

*Pi v_1/2 t = 40°C

134.81 120.29 123 103.68 172.13

'p2 t=

v_l/2 t = 60°C

'pi

40°C

265.62 302.46 340.57

166.52 150.18 227.62 207.63 285.87

-

364.91

*p2 V~V2 t = 60°C

455.74 497.34 697.34 561.81 565.38

Using the linear dependencies ln(i /v 1/2) = f(l/T ), the activation energies o f the two main processes have been calculated, as is shown in Table 4. They correspond to a diffusion process, accord­ ing to their magnitude order, but with relatively different values for the two main processes. This fact might be assigned to Ag+ and Ag2+ ions diffusion and not to the OH~ ones. Table 4 Activation energies for the two main oxidation processes of Ag in KOH solutions (in cal per mole). The number in parenthesis represents the correlation coefficients of the experimental data C (M) 0.5 1.0 4.0 7.0 10.0

Peak 1 2470 4200 3400 4600 4260

(0.910) (0.866) (0.995) (0.978) (0.949)

Peak 2 5350 (0.929) 5400 (0.918) 4650(0.981) -

3500 (0.956)

Chronoamperometry Chronoamperogramms at potential values corresponding to the main peaks are also recorded and are presented in Figs 3 and 4. Their shape is the same for the whole experimented temperatures and concentrations domain. Chronoamperogramms corresponding to the first peak have a classical shape, that is characte­ ristic to those processes when a layer is deposited on electrode surface, though i - 11/2 dependence is not espected. For the second peak, chronoamperogramms shape shows that the oxidation process o f AgjO to AgO takes places through a nucleation-growth mechanism. When the potential is more positive, fee maximum current corresponding to nucleation-growth process appears earlier and is more pronounced.

434

Daniela Bughiu et al.

t (s)

t(» ) Fig. 3. - Typical chronoamperogramms for the second oxidation step o f Ag in KOH solutions.

Fig. 4. - Typical chronoamperogramms for the first oxidation step of Ag in KOH solutions.

There are several nucleation-growth models in reference literature, one o f them being applied for this case by Becerra, Salvarezza and Arvia, who have compared the calculated values o f nuclei concentra­ tion N with those experimentally obtained by means o f SEM microphotographs analysis.3,4 The corre­ lation is not so good; however, the authors concluded that AgO occurs through 2D or 3D nucleationgrowth mechanism.

CONCLUSIONS After the experim ental that were done, the following was found: - a pronounced dissolution o f Ag20 with Ag(OH):; formation occurs, at relatively high KOH concentrations, greater than 4M, when silver oxide solubility is minimal in alkaline electrolyte and at temperatures higher than 40°C; - at potentials values at which oxygen evolves, Ag20 3 may be formed, which can be responsi­ ble for the subpeak in the reverse scan; - at low scan rates, both main steps of Ag oxidation are controlled by diffusion.

REFERENCES 1 N. Sato and Y. Shimizu, Electrochim. Acta, 1973,18, 567. 2 G. T. Burstein and R. C. Newman, Electrochim. Acta, 1980, 25, 1009. 3 G. Alonso, R. C. Salvarezza, .1. M.Vara and A. J. Arvia, Electrochim. Acta, 1990, 35, 489. 4 J. Gomez Becerra, R. C. Salvarezza and A. I. Arvia, Electrochim. Acta, 1988, 33, 1431. 5 T. P. Hoar and C. K. Dyer, Electrochim. Acta, 1972,17, 1563. 6 R. D. Giles and J. A. Harrison, J. Electroanalyt. Chem., 1970, 27, 161. 7 B. V.Tilak, R. S. Perkins, H. A. Kozlowska and B. E. Conway, Electrochim. Acta, 1972, 1 7 ,1447. 8 B. G. Pound, D. D. MacDonald and J. W. Tomlison, Electrochim. Acta, 1980, 25, 1293. 9 B. G. Pound, D. D. MacDonald and .T. W. Tomlison, Electrochim. Acta, 1979,24, 929. 10 H. Sasaki and S. Toshima, Electrochim. Acta, 1975, 20, 201. 11 N. A. Hampson, J. B, Lee and J. R. Morley, Electrochim. Acta, 1971,16, 637. 12 J. Ambrose and R. G. Barradas, Electrochim. Acta, 1974,19, 781. 13 B. G. Pound, D. D. MacDonald and J. W. Tomlison, Electrochim. Acta, 1980,25, 563. 14 J. Zerbino, M. L. Teijelo, J. R. Vilche and A. J. Arvia, Electrochim. Acta, 1985, 30, 1521. 15 M. L. Teijelo, J. O. Zerbino, J. R. Vilche and A. J. Arvia, Electrochim. Acta, 1984, 29, 939.