Gold autodeactivation during oxygen electroreduction studied by electrochemical impedance spectroscopy

Gold autodeactivation during oxygen electroreduction studied by electrochemical impedance spectroscopy

Journal of Electroanalytical Chemistry 683 (2012) 21–24 Contents lists available at SciVerse ScienceDirect Journal of Electroanalytical Chemistry jo...

444KB Sizes 0 Downloads 28 Views

Journal of Electroanalytical Chemistry 683 (2012) 21–24

Contents lists available at SciVerse ScienceDirect

Journal of Electroanalytical Chemistry journal homepage: www.elsevier.com/locate/jelechem

Short Communication

Gold autodeactivation during oxygen electroreduction studied by electrochemical impedance spectroscopy Oleg Tripachev a, Vera Bogdanovskaya a, Mikhail Tarasevich a, Viktor Andoralov a,b,⇑ a b

Laboratory of Electrocatalysis and Fuel Cells, A.N. Frumkin Institute of Physical Chemistry and Electrochemistry RAS, Leninsky Prospect 31, RU-119071 Moscow, Russia Biomedical Science Laboratory, Faculty of Health and Society, Malmö University, Södra Förstadsgatan 101, SE-20506 Malmö, Sweden

a r t i c l e

i n f o

Article history: Received 15 February 2012 Received in revised form 19 July 2012 Accepted 25 July 2012 Available online 2 August 2012

a b s t r a c t The deactivation of a polycrystalline gold electrode is observed during oxygen electroreduction reaction (ORR) in basic medium. At that, the cause of the process is chemical decomposition of the ORR intermediate and blocking of active sites of the electrode surface by hydroxyl radical-like species. The deactivation mechanism is discussed. Ó 2012 Elsevier B.V. All rights reserved.

Keywords: Autodeactivation Oxygen reduction reaction Impedance Superoxide Gold Catalysis

1. Introduction The mechanism and kinetics of the oxygen reduction reaction (ORR) depend on a number of factors and may be affected by a secondary process in electrocatalysis. However, causes of the affection are usually quite various. For example, these are adsorption (desorption) of blocking species on (from) an electrode surface [1], dissolving of one of complex system components [2], transformation of a crystalline structure of the electrode surface [3] and so on. It also concerns the ORR on a polycrystalline gold (p-Au) electrode. Perspective gold based catalysts for a Fuel Cell technology [4] have to demonstrate rate of the ORR as higher as possible. On the other hand, at present gold materials are very popular for bioelectrochemical [5] and bioanalytical purposes [6,7]. In this case the ORR often plays a negative role. To understand and control such systems, it is important to know a mechanism of a process which can influence on the activity of the gold electrode. As was shown in early works, activity of the p-Au towards the ORR in the alkaline medium strong depends on a method of the electrode pretreatment [3,8]. The authors suppose that it might be connected with differences in the structure and composition

⇑ Corresponding author at: Laboratory of Electrocatalysis and Fuel Cells, A.N. Frumkin Institute of Physical Chemistry and Electrochemistry RAS, Leninsky Prospect 31, RU-119071 Moscow, Russia. Tel.: +46 40 665 7545; fax: +46 40 665 8100, +7 495 952 0846. E-mail addresses: [email protected], [email protected] (V. Andoralov). 1572-6657/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jelechem.2012.07.030

of the electrode surface. Activity of the p-Au electrode may be significantly changed by reaction with hydroxyl radicals [9]. At that the selective inactivation of catalytic sites takes place [10]. Here we illustrate some aspects of the p-Au electrode autodeactivation during the ORR for the first time. It is studied by voltammetry and electrochemical impedance spectroscopy (EIS). The main goal of the work was determination and description of the ORR pathways and the p-Au electrode autodeactivation mechanism. The mechanism of the process is proposed. 2. Experimental section 2.1. Materials Solutions were prepared using NaOH (99.99%) from Sigma (St. Louis, MO, USA), H2O2 without stabilizers was from Institute of General and Inorganic Chemistry (Moscow, Russia), H2SO4, K2HPO4, KH2PO4, CH3COONa, CH3COOH were from ChimMed (Moscow, Russia). Water (18 MX cm) was purified with a UVOI‘‘MF’’-1812-NA system from Mediana Filter (Moscow, Russia). Highly purified gasses of O2 (99.999%) and Ar (99.999%) were used. All reagents and materials were analytical grade. 2.2. Instrumentation A three electrode electrochemical cell was used. The cell consisted of chemical-resistant glass. The counter electrode was separated from the working electrode space. A polycrystalline gold

22

O. Tripachev et al. / Journal of Electroanalytical Chemistry 683 (2012) 21–24

cylinder (99.999%) of 5 mm in diameter was hermetically pressed in a Teflon tube and used as the working rotating disk electrode. Before measurements the electrode was polished in series with water pastes of Buehler consisting of aluminium oxide with particle size 5 lm and 0.05 lm respectively. Then the electrode was treated with an ultrasound bath for 2 min, further, the electrode was rinsed with water. A platinum mesh was used as the counter electrode. The reference electrodes were a Ag|AgCl|KClsat and a Hg|HgO|NaOH electrodes. All potentials were recalculated and presented vs. NHE. All measurements were done at constant temperature of 20 °C. Electrochemical measurements were preformed using potentiostat Solartron 1287A (Farnborough, UK) and frequency response analyzer Solartron 1255B (Farnborough, UK). Electrochemical Impedance was measured with potential perturbation amplitude of 5 mV with logarithmic law (10 dots per decade) from 20 kHz up to 30 mHz.

3. Results and discussion A typical voltammogram of the ORR on the p-Au electrode in alkaline medium represents two waves of catalysis (Fig. 1A, curve 2) [3,8,11,12]. The first wave is located in the high-potential region. As was shown by a number of investigations the main process in the potential region is the ORR to peroxoanion (HO 2 ). The second wave of catalysis appears in the low-potential region and corresponds to the consecutive ORR to water (Fig. 1A, curve 2) [8,11]. This is in agreement with voltammogram of HO 2 reduction on

the p-Au electrode. There is one wave of catalysis on the curve (Fig. 1A, curve 1) and the potential range of the wave coincides with the second wave of the ORR. Fig. 1B shows the linear sweep voltammograms (LSVs) of the ORR which were measured on the p-Au disk electrode after different time of electrode polarization at 0.05 V without rotation. As we can see from Fig. 1B deactivation of the electrode occurs during the ORR. The electrode may be again activated by applying of a significantly low potential (below 1.0 V vs. NHE). The fresh electrode does not lose the initial activity both in the case of holding at the open circuit potential in O2 saturated solution and during cathodic polarization at Ar saturated solution. Special attention should be focused on that the electrode deactivation does not occur in the presence of HO 2 in the deaerated solution. These experiments clearly verify that the deactivation is not due to contaminants. Fig. 1 also shows that the limiting current of the first wave is close to the theoretical value calculated for transfer of two electrons at the same conditions. In this case the ORR occurs to HO 2 [3]. Rate-determining steps (RDSs) on the fresh electrode were defined in acid and basic medium using the pH dependence of ORR half-wave potential (Fig. 2). In the weak-acid medium, nature of RDS changes from H+ dependent to H+ independent. The transition from one mechanism to another occurs around pH 4 that is close to the pKa of superoxide (4.9). It was shown by analysis of Tafel plots that in acid medium the rds of the ORR on the p-Au electrode includes transfer of one electron [13,14]. In this case we can propose the following equation for the rds of the ORR main pathway in acid medium on the p-Au electrode:

O2ðadsÞ þ Hþ þ e ! HO2ðadsÞ

Increasing of pH conducts to dissociation of surface superoxidelike species. The RDS of the ORR in basic medium may be described with the following equation:

0

j / mA cm-2

ð1Þ

-1

1

0.20

-2

0.15

2

/ V

-3

0.10

1/ 2

A

E

-4 -1.0

-0.5

0.05

0.0

E / V (vs . NHE) 0.00 -0.05

0

0

j / mA cm-2

4

2

4

6

8

10

12

14

pH Fig. 2. pH dependence of half-wave potential for high-potential ORR wave. Conditions: fresh p-Au electrode, 1500 rpm, O2, 5 mV s1.

3 2

-1

1

-

2e -2 -0,6

-0,4

-0,2

B 0,0

0,2

E / V (vs . NHE) Fig. 1. (A) Linear sweep voltammograms of the ORR on fresh p-Au electrode in the presence of (1) -HO 2 and (2) -O2. (B) Linear sweep voltammograms of the ORR on pAu electrode after different time of polarization at 0.050 V in O2 saturated solution: (1) 0 min, (2) 15 min, (3) 30 min, and (4) 5 h. Conditions: 1 M NaOH, 1500 rpm, cathodic direction of scanning with rate of 5 mV s1.

Scheme 1. Scheme of two-electron ORR to HO 2 in basic medium on p-Au electrode. kI (kV) – diffusion rate constant of O2 (HO 2 ), kII – constant of one-electron O2    reduction to O 2ðadsÞ , kIII – constant of one-electron O2ðadsÞ reduction to HO2 , kIV (k IV) – constant of the intermediate decomposition on deactivated (activated) electrode, k0 – back processes constants.

23

O. Tripachev et al. / Journal of Electroanalytical Chemistry 683 (2012) 21–24

O2ðadsÞ þ e ! O2ðadsÞ

ð2Þ

HO 2

The ORR to with the intermediate formation might be described by Scheme 1 [15]. As we can see from the scheme the  ORR starts from O 2ðadsÞ formation. The presence of O2ðadsÞ on gold electrode surface was experimentally shown by Raman spectroscopy and RRDE methods [16,17]. The intermediate can decompose or be electrochemically reduced (Scheme 1). EIS measurements were carried out on the deactivated p-Au electrode which demonstrated stable electrocatalytic activity towards O2. The Nyquist plots measured at different potentials are shown in Fig. 3. As we can see the structure of the impedance spectrum is simple (one arc) in the high-potential region and corresponds with the ORR to HO 2 (Fig. 3A). Impedance spectra were also measured at different polarization time in the high-potential region. Resistance of charge transfer (Rct) was obtained by the data fitting using respective circuit (Fig. 3A). The resistance grows during of the electrode polarization. The growing stops approximately in 6 h. When polarization of the electrode goes up, the inductive element appears in the spectrum structure (Fig. 3B–D). Diffusion short-circuit Warburg impedance (Ws) starts to be clear at potentials lower than 0.28 V [18]. This element concludes a parameter (Tw) which is suitable for diffusion coefficient determination. From the Eq. (3) for Ws it is clear why the element behaves as a capacitive element and how the element relates with Tw.

Z Ws ¼

Rd  tanhð½jxT W 0:5 Þ

ð3Þ

ðjxT W Þ0:5

where Rd is the diffusion resistance, j is the imaginary unit, and x is the angular frequency. Experimental spectra with ‘‘capacitive-like’’ arc of Ws were fitted and Tw parameter was evaluated for different conditions (Table 1). As we can see Tw does not depend on applied potential but the parameter is inversely proportional to rotation rate of disk electrode (4). 2

TW ¼

l 2:59  m1=3 1  ¼ w D D1=3

ð4Þ

where l is the effective double layer thickness, v is the kinematic viscosity, w is the rotating rate (rad s1), and D is the diffusion coefficient. The value of 1.86  105 cm2 s1 for oxygen diffusion coefficient was obtained based on the fitting data. The coefficient is in good agreement with literature data [19]. The inductive element (Fig. 3(C,D)) is connected with long life intermediate species (O 2ðadsÞ ) formation during the ORR [17]. The intermediate adsorbs on the electrode surface that results in adsorptive impedance appearing. The impedance can behave both as inductive or capacitive element depending on what the main process of the intermediate transformation is [14,17,18]. The main process which is responsible for the adsorptive impedance was defined by analysis of low frequency phase angle dependence on electrode potential. There are phase angle – potential diagrams plotted using the data of theoretical calculations (mathematical Table 1 Values of Tw parameter of diffusion impedance. Rotation rate (rpm) Fig. 3. Nyquist plots of electrochemical impedance and respective equivalent circuits of the ORR on deactivated p-Au electrode at different electrode potentials (vs. NHE). Points – experimental data, lines – fitting using the equivalent circuits. Le – inductivity of conductive system, La – adsorptive inductivity, Rs – solution resistance, Rct – resistance of charge transfer, Ra – adsorptive resistance, Ws – Warburg short-circuit impedance, CPE – constant phase element. Conditions: 1 M NaOH, 1500 rpm, O2.

(E/V) 0.35 0.38 0.43 0.48

780 0.121 – – –

1150 0.167 0.140 0.113 0.149 Mean 0.142 SD 15%

1500 0.225 – – –

1850 0.340 – – –

24

O. Tripachev et al. / Journal of Electroanalytical Chemistry 683 (2012) 21–24

20

de

co

mp

os

10

40

А itio

30

n

Phase

Phase

15

5 0 -5

-0,5

-0,4

exp

eri

20

me

nta

ld

ata

10

uc

d

ore

ctr

ele

-10

n tio

B

0 -0,3

-0,2

-0,5

η/V

-0,4

-0,3

-0,2

E/V

Fig. 4. Phase angle dependencies on overvoltage or electrode potential at frequencies of adsorptive impedance. (A) Theoretical calculated dependencies of the ORR with decomposition (at 1.5 mHz) and electroreduction (at 100 mHz) of intermediate. (B) Experimental data (at 30 mHz) of the ORR on the stabilized p-Au electrode. Conditions: 1 M NaOH, 1500 rpm, O2.

principle was from [17]) and experimental results in Fig. 4. The experimental data are well-correlated (Fig. 4B) with the curve calculated for the decomposition pathway (Fig. 4A) kI ? kII ? kIV ? kV (Scheme 1). The theoretical curve (Fig. 4A) of the electroreduction pathway kI ? kII ? kIII ? kV (Scheme 1) is principal different with the experimental data by the character. Thus the comparison of experimental data and model calculations lets us conclude that in this case the main pathway of the intermediate reorganization is decomposition [16,20]. The measurements of impedance at different concentrations of  HO 2 showed that spectrum structure does not depend on HO2 concentration. These results indicate, that rate of desorption of the product is much higher than back process and process of its formation. So, the deactivation of the electrode is connected with the chemical decomposition of the intermediate (O 2ðadsÞ ). Mechanism of the decomposition process can be described in the following way:

2O2ðadsÞ þ H2 O ! O2 þ HO2 þ HO

ð5Þ

However, the decomposition process can also occur in agreement with Eq. (6). In this case adsorbed hydroxyl radical-like species can block active sites of the p-Au electrode. As result total rate of the ORR decreases.

2O2ðadsÞ þ 2H2 O ! O2 þ 2HOðadsÞ þ 2HO

ð6Þ

The process (6) is more complex than (5) because one of two peroxo-bonds mast be broken during the process (6). Probability of the process (6) on the p-Au is lower than (5) therefore rate of the deactivation also is low. Moreover, we believe that the reaction (6) takes place just on certain active sites and when all of the sites are blocked, deactivation of the p-Au electrode stops (Fig. 1(B)) [9,10]. The decreasing of the ORR current might be connected with lessening of real surface area of the electrode as well. The real surface area was measured using oxide layer formation [21]. The roughness factor of the electrode was the same for both activated and deactivated states. This means that the deactivation is not due to gold smoothing. 4. Conclusion The deactivation of the p-Au electrode towards the ORR was studied. Pathways of the ORR and the mechanism of the deactivation were proposed to be based on obtained results. It was shown that the deactivation directly connected with the ORR, so this process can be called as gold autodeactivation during the ORR. Acknowledgements The work has been financially supported by the Russian Foundation for Basic Research (Project 10-03-00236). Dr. G. Zhutaeva

and O. Lozovaya are gratefully acknowledged for helpful discussion. References [1] T.J. Schmidt, U.A. Paulus, H.A. Gasteiger, R.J. Behm, The oxygen reduction reaction on a Pt/carbon fuel cell catalyst in the presence of chloride anions, J. Electroanal. Chem. 508 (2001) 41–47. [2] D.V. Novikov, M.R. Tarasevich, V.A. Bogdanovskaya, V.M. Andoralov, G.V. Zhutaeva, Peculiarities of the behavior of a PdCoPt/C cathodic catalyst in an acidic environment: formation of core–shell structures, Prot. Met. Phys. Chem. Surf. 46 (2010) 444–451. [3] C. Paliteiro, (1 0 0)-Type behavior of polycrystalline gold towards O2 reduction, Electrochim. Acta 39 (1994) 1633–1639. [4] W. Chen, S. Chen, Oxygen electroreduction catalyzed by gold nanoclusters: strong core size effects, Angew. Chem. Int. Ed. 48 (2009) 4386–4389. [5] S. Shleev, T. Ruzgas, Transistor-like behavior of a fungal laccase, Angew. Chem. Int. Ed. 47 (2008) 7270–7274. [6] A. Bardea, E. Katz, A.F. Buckmann, I. Willner, NAD(+)-dependent enzyme electrodes: electrical contact of cofactor-dependent enzymes and electrodes, J. Am. Chem. Soc. 119 (1997) 9114–9119. [7] P. Scodeller, V. Flexer, R. Szamocki, E.J. Calvo, N. Tognalli, H. Troiani, A. Fainstein, Wired-enzyme core–shell Au nanoparticle biosensor, J. Am. Chem. Soc. 130 (2008) 12690–12697. [8] B.G. Podlibner, L.N. Nekrasov, Sov. Electrochem. 51 (1969) 340–343. [9] A.M. Nowicka, U. Hasse, M. Hermes, F. Scholz, Hydroxyl radicals attack metallic gold, Angew. Chem. Int. Ed. 122 (2010) 1079–1081. [10] A.M. Nowicka, U. Hasse, G. Sievers, M. Donten, Z. Stojek, S. Fletcher, F. Scholz, Selective knockout of gold active sites, Angew. Chem. Int. Ed. 122 (2010) 3070– 3073. [11] M.R. Tarasevich, K.A. Radyushkina, V.Yu. Filinovskiy, R.Kh. Burshteyn, Russ. J. Electrochem. 61 (1970) 1451–1468. [12] K. Kajii, T. Ohsaka, F. Kitamura, Crucial effect of rubidium cation on catalytic activity of a polycrystalline gold electrode toward oxygen reduction reaction in alkaline media, Electrochem. Commun. 12 (2010) 970–972. [13] M.A. Genshaw, A. Damjanovic, J.O’M. Bockris, Hydrogen peroxide formation in oxygen reduction at gold electrodes: I. Acid solution, J. Electroanal. Chem. 15 (1967) 163–172. [14] V.M. Andoralov, M.R. Tarasevich, O.V. Tripachev, Oxygen reduction reaction on polycrystalline gold. Pathways of hydrogen peroxide transformation in the acidic medium, Russ. J. Electrochem. 47 (2011) 1327–1336. [15] S. Strbac, R.R. Adzic, The influence of OH chemisorption on the catalytic properties of gold single crystal surfaces for oxygen reduction in alkaline solutions, J. Electroanal. Chem. 403 (1996) 169–181. [16] J. Kim, A.A. Gewirth, Mechanism of oxygen electroreduction on gold surfaces in basic media, J. Phys. Chem. B 110 (2006) 2565–2571. [17] M. Itagaki, H. Hasegawa, K. Watanabe, T. Hachiya, Electroreduction mechanism of oxygen investigated by electrochemical impedance spectroscopy, J. Electroanal. Chem. 557 (2003) 59–73. [18] A. Lasia, in: B.I. Conway, J.O’M. Bockris, R.E. White (Eds.), Modern aspects of electrochemistry, 32, Kluwer Academic Publishers, New York, Boston, Dordrecht, London, Moscow, 2002, pp. 143–238. [19] M. Chatenet, M. Aurousseau, R. Durand, Electrochemical measurement of the oxygen diffusivity and solubility in concentrated alkaline media on rotating ring-disk and disk electrodes–application to industrial chlorine-soda electrolyte, Electrochim. Acta 45 (2000) 2823–2827. [20] R.W. Zurilla, R.K. Sen, E. Yeager, The kinetics of the oxygen reduction reaction on gold in alkaline solution, J. Electrochem. Soc. 125 (1978) 1103– 1109. [21] S. Trasatti, O.A. Petrii, Real surface area measurements in electrochemistry, Pure Appl. Chem. 63 (1991) 711–734.