Journal of Electroanalytical Chemistry 654 (2011) 102–107
Contents lists available at ScienceDirect
Journal of Electroanalytical Chemistry journal homepage: www.elsevier.com/locate/jelechem
A rotating tungsten disc electrode in concentrated strong alkaline solutions: An electroanalytical aspect Tamara Tuvic´ a, Igor Pašti a, Slavko V. Mentus a,b,⇑ a b
University of Belgrade, Faculty of Physical Chemistry, Studentski trg 12, 11158 Belgrade, Serbia Serbian Academy of Science and Arts, Knez Mihajlova 35, 11158 Belgrade, Serbia
a r t i c l e
i n f o
Article history: Received 23 November 2010 Received in revised form 11 January 2011 Accepted 17 January 2011 Available online 25 January 2011 Keywords: Anodic dissolution Hydroxide ion Rotating disc electrode Voltammetry Tungsten
a b s t r a c t Electrochemical behavior of tungsten in concentrated strong alkaline solutions (103–101 M KOH, 11 < pH < 13) was studied by means of rotating disc electrode technique. Contrary to the case of less concentrated alkaline solutions, the formation of solid anodic passive film was not evidenced. Chronoamperometric measurements evidenced a fast establishment of steady state conditions in the surface region. In spite of the complexity of overall surface electrochemistry, a linear relationship between the limiting anodic currents and the concentration of OH ions was evidenced, which may be used for easy and accurate determination of OH ion concentration at high pH values where the glass electrode loses its high reliability. Ó 2011 Elsevier B.V. All rights reserved.
1. Introduction Glass electrode presents a commonly used sensor for the determination of acidity and basicity of solutions. However, its significant ‘‘alkaline error’’ in the solutions of high pH is an always present disadvantage [1]. In the last two decades voltammetric and amperometric approaches under steady state conditions have been studied in order to develop the methods of non-potentiometric determination of concentration of OH ions [2–22]. A series of papers by Daniele and co-workers [13–17] was devoted to the voltammetric determination of hydroxide ions in aqueous solution using both gold microelectrode [13–16] and rotating disk electrode [15]. Their measurements involved the oxidation of OH ions under diffusion limiting conditions. A relatively narrow potential window, of about 0.2 V, of diffusion control of OH ion oxidation on gold electrode, being limited on the high potential side by H2O oxidation, presents an aggravating factor of this procedure. Abu-Rabi et al. [18] published recently that the current of OH ions oxidation may be shifted to the cathodic side by intense cathodic electrode pretreatment. Hence, a reliable measurement of OH concentration using gold electrode requires a careful control of the state of the surface, relating particularly to the surface coverage by gold oxide.
⇑ Corresponding author at: University of Belgrade, Faculty of Physical Chemistry, Studentski trg 12, 11158 Belgrade, Serbia. Tel./fax: +381 11 2187133. E-mail address:
[email protected] (S.V. Mentus). 1572-6657/$ - see front matter Ó 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jelechem.2011.01.018
The Compton group [19] proposed gold ultra-microelectrode arrays consisting of hundreds and thousands electrochemically active electrodes for sensing hydroxyl ions. The reproducible voltammetric waves were observed, and the linearity of peak height vs. hydroxide concentration was proven in the concentration range 50 mM up to 1 mM. Trough a series of investigations, Banks et al. [20–22] developed a nickel oxide screen printed electrodes for the electroanalytical sensing of OH ions. The sensor allows the detection of OH ions over the low micro-molar to millimolar range with a detection limit of 23 mM. According to the available data related to the electrochemical behavior of tungsten at pH values close to 7, its anodic oxidation proceeds in a manner similar to the oxidation of valve metals, namely, obeys the high-filed mechanism of the formation of passive oxide layer [23–27]. As demonstrated by Ortiz et al. [27] for a pH range 7–10, when the cyclic voltammetry (CV) experiments were performed in a quiescent solution, the oxide formation was terminated immediately upon reversing polarization direction. If, in a series of CV experiments, vertex anodic potential was enlarged with respect to the previous one, the oxide formation started at slightly lower potential with respect to the vertex potential of the previous polarization cycle. This indicated a chemical dissolution of WO3 at a measurable rate. The dissolution rate increased with the increase in pH, and the passivation almost disappeared at pH 12 [24–27]. Such an behavior is not specific for true valve metal oxides, e.g. titanium dioxide, which is quite insoluble under similar conditions [28]. As regards to the anodic dissolution of
T. Tuvic´ et al. / Journal of Electroanalytical Chemistry 654 (2011) 102–107
tungsten in highly alkaline media, it is worth mentioning that, at sufficiently high potentials, the current passes into a plateau, which is sensitive towards the rotation rate and extends to a surprisingly large potentials, up to about 8 V [27]. Davydov et al. [29–31] considered theoretically the ion fluxes accompanying the anodic dissolution of tungsten in alkaline solution, and derived the equations of limiting current under both steady state [29] and pulsed [30] conditions. Some other examples of anodic dissolution of metals, e.g. iron, copper and nickel dissolution in acidic media, leading to a mass transfer control too, were studied by Barcia et al. [32]. Although in several published papers [24,29,30] the rate of anodic dissolution of tungsten in alkaline solutions was found to depend on the base concentration, none of them did consider the question whether this behavior may be used for amperometric determination of OH ion concentration. Therefore, the subject of the present study is analytical determination of OH ion concentration in a high concentration region, above 103 M, by means of anodic dissolution of tungsten. The applicability of rotating disk (RDE) technique in both supported (K2SO4) and unsupported electrolyte solutions was studied.
103
Fig. 1. Cyclic voltammograms of tungsten rotating disc electrode in 0.08 M KOH solution supported by 0.25 M K2SO4. Polarization rate was 50 mV s1.
2. Materials and methods A tungsten rod (Alfa Aesar, 99,99%) 3.17 mm in diameter, was mounted in a cylindrical Teflon holder of a rotating electrode. The cross section surface (0.079 cm2) of tungsten rod was exposed to the electrolyte. It was polished by dry corundum particles (1 lm) before the first use. The three-electrode cell (effective volume 20 ml) with a Pt foil as a counter electrode and a saturated calomel electrode (SCE) as the reference one were used. The solutions were purged by bubbling N2 (5 N) through the cell. All the experiments were conducted at room temperature (25 ± 1)°C. The electrochemical measurements were carried out by means of the device Gamry PC14-750 equipped with Pine Rotator. The cyclic voltammetry experiments (RDE–CV) were performed at a scan rate 20 mV s1, under simultaneous electrode rotation. The cyclovoltammograms were recorded in the potential range between 1.3 and 1.2 V vs. SCE in the solutions containing K2SO4, and in the potential range between 1.3 and 1.7 V vs. SCE, in the unsupported electrolyte solutions. The rotating disk chronoamperometry was performed at fixed potentials corresponding to the limiting current of W dissolution, namely at 1.2 V vs. SCE, in the solutions containing the supporting electrolyte (K2SO4), and at 1.5 V vs. SCE, in the unsupported solutions. The starting solution was 10 ml of 0.01 M of KOH. The analyte, 0.1 M KOH, was injected in equal portions, and the transient current was registered during roughly 60 s.
Fig. 2. Cyclic voltammograms of tungsten rotating disc electrode in unsupported 0.08 M KOH solution. Polarization rate was 50 mV s1.
WO3ðsÞ þ OH ! HWO4
ð1dÞ
HWO4 þ OH ! WO2 4 þ H2 O
ð1eÞ
General dissolution reaction may be written as [33]:
3. Results and discussion According to the RDE–CV curves, presented in Figs. 1 and 2, in 0.08 M KOH solution, extensive current of tungsten electrooxidation onsets at the potential 0.4 V vs. SCE. With the progress of anodic polarization, the oxidation current makes several maxima and then passes into a flat plateau. The wavy form of current–voltage curve in the potential region that precedes the occurrence of the current plateau agrees with the well accepted multi-step mechanism of W dissolution in alkaline solutions elaborated by Kelsey [33]:
W þ 2OH ! WO2ðsÞ þ 2Hþ þ 4e
ð1aÞ
WO2ðsÞ þ OH ! WO3 HðsÞ þ e ðRDSÞ
ð1bÞ
WO3 HðsÞ þ OH ! WO3 ðsÞ þ H2 O þ e
ð1cÞ
W þ 8OH ! WO2 4 þ 4H2 O þ 6e
ð1Þ
The current oscillations corresponding to these processes are more pronounced in diluted solutions, however, their amplitude does not change with the change in concentration, and consequently, they become negligible at high hydroxide concentrations. As Figs. 1 and 2 show, upon several initial oscillations, current passes into plateau. According to the literature data [27], the potential region of the current plateau extends up to even 8.5 V vs. SCE before than a visible current of oxygen evolution appears. Contrary to the case of diluted base solutions, where passivation may occur [23,27], in the case of concentrated base solutions, the dissolution of tungsten proceeds without any passivation, which is evidenced in Figs. 1 and 2 by the fact that, at any rotation rate, almost overlapping currents are registered during anodic and cathodic scans.
T. Tuvic´ et al. / Journal of Electroanalytical Chemistry 654 (2011) 102–107
104
For a simple electrochemical process on a rotating disc electrode, during of which an electroactive species of concentration c is consumed, the limiting diffusion current depends on bulk concentration c of electroactive species, its diffusion coefficient D and the thickness of the diffusion layer, d, according to the formula which follows from the Fick’s law:
jL ¼
nFDc d
ð2Þ
The thickness of the diffusion layer, d, may be calculated by means of the Levich equation:
d ¼ 1:61D1=3 m1=6 x1=2
ð3Þ
where m represents kinematic viscosity (i.e., viscosity divided by density) and x is the angular rotation frequency (rad s1). The combination of Eqs. (2) and (3) yields the Levich equation in a new form (Eq. (4)) which predicts a linear dependence of limiting diffusion current on the square root of angular rotation frequency: 2
1
1
jL ¼ 0:62 n F D3 m6 x2 c
ð4Þ
The coefficient 0.62, c in units mol m3, and m and D in units m s1, give limiting current density in units A m2. In order to check whether the system under investigation obeys the Eq. (4), regardless which one among the reactants appearing in Eq. (1) is responsible for mass transfer control, for several fixed OH ion concentrations, the dependence of current density (taken from the region of current plateau in Figs. 1 and 2) on square root of angular frequency was plotted in Figs. 3 and 4, for supported and unsupported case, respectively. Figs. 3 and 4 indicate that the process under investigation obeys, at least qualitatively, the Eq. (4), i.e. one deals with the process which behaves as a purely mass transfer controlled one. In order to check the quantitative correspondence of Eq. (4) to the assumption that only OH ions are responsible for mass transfer limitation of anodic dissolution of tungsten, the diffusion coefficient of OH ions at infinite dilution, amounting to 5.27 105 cm2 s1 [34], and n = 1, were used to calculate the dependence jL vs. x1/2 according to the Eq. (4). As one can see in Fig. 5, the slope of the calculated dependence is considerably higher than the one determined experimentally for anodic dissolution of
Fig. 4. The Levich plots for anodic dissolution of tungsten in unsupported KOH solutions of the following concentrations (M): j – 0.02; – 0.04; – 0.06; – 0.08; – 0.10. The limiting currents, at the potential of 1.5 V, were used to make the plots.
2
Fig. 3. The Levich plots for anodic dissolution of tungsten in KOH solutions of selected concentrations (M): j – 0.02; – 0.04; – 0.06; – 0.08; – 0.10, supported by 0.25 M K2SO4. The limiting currents, at the potential of 1.2 V, were used to make the plots.
tungsten, even in unsupported solution, and this disagreement will be discussed below. What additionally may be expected on the basis of Eq. (4) for a fixed rotation rate, is a linear dependence of jL vs. concentration of reactants responsible for mass transfer limitations. From the dependencies jL vs. x1/2 presented in Figs. 3 and 4, the dependence of limiting current on OH ion concentration, for a fixed rotation rate of 1800 rpm, was derived and presented in Fig. 6. A very fair linearity may be observed in the OH concentration range covering full two decades. The curves obtained (Fig. 6), which may be considered as the calibration ones, are characteristic of high R2 factor (>0.99). However, the calculation based on Eq. (4) under the assumption that OH ion is the only diffusing species, gave again the line with much higher slope than found experimentally. In addition to the Levich analysis, the apparent reaction order with respect to OH ions was determined from the slope of the
Fig. 5. The Levich plot of rotating tungsten disc electrode in unsupported ( ) and supported (0.25 M K2SO4) ( ) 0.1 M KOH solution. Triangles (N) present the points calculated on the basis of Levich equation (4) under assumption that the diffusion of OH ions only limits the anodic dissolution of tungsten (n = 1); the diffusion coefficient of OH ions at infinite dilution, D = 5.26 105 cm2 s1, was used for the calculation.
T. Tuvic´ et al. / Journal of Electroanalytical Chemistry 654 (2011) 102–107
Fig. 6. The limiting diffusion current density in function of concentration of OH ions for tungsten rotating disc electrode in KOH solutions, for the following concentrations (M) of supporting electrolyte K2SO4: j – 0.50; – 0.25; – 0.10; – 0.05; – 0.00. Opened circles (s) present the data calculated by means of Levich equation (4) under assumption that the diffusion of OH ions only limits the anodic dissolution of tungsten (n = 1); the diffusion coefficient of OH ions at infinite dilution, D = 5.26 105 cm2 s1, was used for the calculation. A common rotation rate of 1800 rpm holds for all points.
dependence log j = f(log c(OH)) (Fig. 7, rotation rate 1800 rpm) and it was found to amount to:
dlogj ¼1 dlogðcðOH ÞÞ
ð5Þ
This confirmed the first order kinetics of anodic W dissolution with respect to OH ions, as already found by Ortiz et al. for strong alkaline solutions [35]. According to Fig. 7, this value is quite independent on the ionic strength of the solution, and agrees with the mechanism of tungsten electrodissolution proposed by Kelsey [33]. In order to examine the speed of current response to the change of OH ion concentration in the solution, the hydrodynamic
Fig. 7. The dependence log jL vs. log c(OH) for tungsten rotating disc electrode in unsupported and K2SO4 supported KOH solutions, at the following concentrations (M) of K2SO4 supporting electrolyte: j – 0.50; – 0.25; – 0.10; – 0.05; – 0.00. Empty circles present the data calculated by means of Levic equation (4) under assumption that the diffusion of OH ions only limits the anodic dissolution of tungsten (n = 1); the diffusion coefficient of OH ions at infinite dilution, D = 5.26 105 cm2 s1, was used for the calculation. A common rotation rate of 1800 rpm holds for all points.
105
Fig. 8. Hydrodynamic chronoamperometric curves of tungsten rotating disc electrode in both supported (0.5 M K2SO4) (h) and unsupported (s) KOH solution at a common rotation rate of 1800 rpm. The working potential was fixed within the region of anodic current plateau (1.5 V). The current jumps correspond to the moments when KOH portions were injected.
chronoamperometric curves were recorded upon each injection of equal volume of KOH solution to the examined solution. The solution was stirred by electrode rotation only. Some representative chronoamperometric curves are presented in Fig. 8. As one can see, upon each addition of new portion of hydroxide, the current approached a steady state value almost immediately upon solution homogenization, within a time interval of roughly one second only. If the steady state currents are plotted vs. the concentration of OH ions, linear relationship was obtained for both supported and unsupported KOH solutions. In addition, there is an excellent agreement between j = f (c(OH)) lines obtained by chronoamperometric and voltametric experiments, as presented in Fig. 9. The here described results evidence doubtless that, regardless of the mechanism of anodic dissolution, on the basis of strong linearity of limiting current on concentration, rotating tungsten disc electrode in the regime of anodic dissolution may be used as a fast,
Fig. 9. The dependence of limiting current of tungsten electrodissolution on OH concentration, in both supported (0.5 M K2SO4) (squares) and unsupported (circles) KOH solution, recorded at a rotation rate of 1800 rpm. Empty symbols present the voltammetric data, and filled symbols present the chronoamperometric data.
106
T. Tuvic´ et al. / Journal of Electroanalytical Chemistry 654 (2011) 102–107
robust and effective amperometric sensor of OH ion concentration, which does not suffer of disadvantages found for noble metal electrodes [13–18], i.e. being insensitive to both the initial surface state and the potential variations in an extremely broad range, 1– 8 V, along which the limiting current is registered. Since the reaction (1) presents in fact the tungsten electropolishing, tungsten electrode used to sense OH ions in high concentrated basic solutions presents factically the self-renewable, i.e. maintenance free sensor type. Nevertheless, the facts one need to have in mind here are: (a) the slope of the Levich function depends on the concentration of supporting electrolyte, and (b) there is a remarkable disagreement between the experimentally determined Levich slope and that calculated under assumption that mass transport of OH ions only limits the rate of tungsten anodic dissolution. The comparison of the RDE–CV recorded in the solutions containing supporting electrolyte (K2SO4) to the ones recorded in the unsupported KOH solutions reveals a remarkable effect of supporting electrolyte on the limiting current of tungsten dissolution (Fig. 5). The potential needed to reach limiting diffusion current depends on the presence of supporting electrolyte (Fig. 1). The effect of supporting electrolyte on the limiting diffusion current was already observed in the literature [12,18,24]. Anik et al. [24] reported that the variation in the concentrations of KNO3 supporting electrolyte at pH = 4 has no effect on the hydrodynamic voltammetric curves of tungsten electrode, and assumed that this holds in a broad pH range, 1–13. His conclusion may be justified by the fact that the ionic strength of the solutions was kept constant. Contrary to that, in our case, the addition of supporting electrolyte caused a considerable change in ionic strength and transport behavior of the solution, thus the observed remarkable effect is not a surprise. Formerly, the decrease in limiting diffusion current with the increase in concentration of supporting electrolyte, observed in the case of OH oxidation on gold electrode [12,18], was attributed to the decrease of mobility of OH ions due to strong ion–ion interactions. The remaining important point that deserves the attention is a remarkable lag of the measured limiting currents compared to the limiting current calculated by means of Levich equation (4) under assumption that OH ions only are responsible for mass transfer limitation of tungsten anodic dissolution (Figs. 5, 6 and 8). Obviously, the coefficient 0.62 in Eq. (4) does not apply in this case. This is the consequence of a complex nature of the reaction (1), in which not only OH ions but also tungsten atoms are reactants, and slowly moving WO2 anions (DWO4 = 0.919 105 cm2 s1 4 [34]) participate together with the OH ions in the mass transfer process in the electrolyte diffusion layer. This at least reduces to less than one the ratio of number of electrons against the number of OH ions. Considering the contribution of all ions moving through the diffusion layer, Davydov et al. [29,31] derived the equation for limiting anodic current of itungsten dissolution in h 8DWO 1 the form: jl ¼ nFDc x ð34 xÞ2ðx1Þ=ð3x4Þ , where is x ¼ D 4 and d 4ðx1Þ D = DOH. This equation predicts the factor 0.114 (with n = 6) to modify Eq. (4) in order to apply to the reaction (1) [31]. However, this factor alone does not describe exactly the course of experimentally obtained lines in Figs. 3–7. It is very probably that enlarged viscosity discussed in Ref. [32], caused by large concentration of WO2 4 ions within the diffusion layer, as well as the ionic interactions being neglected in earlier calculations [29], should be taken into consideration, too.
4. Conclusions The electrochemical behavior of tungsten electrode in 103– 10 M KOH solution, both unsupported and supported by K2SO4 1
of various concentrations, was investigated using cyclic voltammetry and chronoamperometry on rotating disc electrode. In the potential range between 0.6 V and 1 V vs. SCE, the successive oxidation steps from W(0) toward the final W(VI) state took place, at a remarkably faster rate in supported than in unsupported solutions. At the potentials higher than 0.5 V in supported, and higher than 1.5 V in unsupported solutions, anodic oxidation of tungsten produces soluble WO2 4 ions, at a constant rate controlled by mass transfer processes. Complete dissolution was confirmed to proceed as a first order reaction with respect to OH-ion. Direct linear relationship was found between anodic limiting current density and the concentration of OH-ions (R2 > 0.99). The current response of the tungsten electrode during anodic polarization under hydrodynamic conditions was found to be fast, highly reproducible and independent on the electrode pretreatment, and such determined limiting current stands in linear correlation to the OH ion concentration with the high correlation factor (R2 > 0.9998). The results of the measurements performed by RDE chronoamperometry show an excellent agreement with the ones obtained by RDE–CV measurements. The here presented results indicate that metallic tungsten electrode can be used as a fast, robust, self-renewable hydrodynamic amperometric sensor for OH ions in the solutions with high pH values. The current response of this sensor is insensitive to the potential variation in a very broad potential range. Due to the dependence of limiting current on the concentration of supporting electrolyte (i.e. on the ionic strength) the calibration procedure is needed.
Acknowledgments The authors I.P. and S.M. are grateful to the Ministry of Science and Technological Development of Republic Serbia for funding this study through the Contract No. III45014. The Serbian Academy of Science and Arts supported the investigations too, through the project ‘‘Electrocatalysis in the contemporary processes of energy conversion’’.
References [1] J. Li, T. Peng, Anal. Chim. Acta 478 (2003) 337. [2] M. Ciszkowska, Z. Stojek, S.E. Morris, J.G. Osteryoung, Anal. Chem. 64 (1992) 2372. [3] M.H. Troise Frank, G. Denuault, J. Electroanal. Chem. 354 (1993) 331. [4] Z. Stojek, M. Ciszkowska, J.G. Osteryoung, Anal. Chem. 66 (1994) 1507. [5] M. Ciszkowska, Z. Stojek, J.G. Osteryoung, J. Electroanal. Chem. 398 (1995) 49. [6] M. Perdicakis, C. Piatnicki, M. Sadik, R. Pasturaud, B. Benzakour, J. Bessiere, Anal. Chim. Acta 273 (1993) 81. [7] S. Daniele, I. Lavagnini, M.A. Baldo, F. Magno, J. Electroanal. Chem. 404 (1996) 105. [8] S. Daniele, M.A. Baldo, F. Simonetto, Anal. Chim. Acta. 331 (1996) 117. [9] M. Ciszkowska, A. Jaworski, J.G. Osteryoung, J. Electroanal. Chem. 423 (1997) 95. [10] S. Daniele, I. Lavagnini, M.A. Baldo, F. Magno, Anal. Chem. 70 (1998) 285. [11] M. Elsayed Abdelsalam, G. Denuault, M.A. Baldo, S. Daniele, J. Electroanal. Chem. 449 (1998) 5. [12] S. Daniele, M.A. Baldo, C. Bragato, G. Denuault, M. Elsayed Abdelsalam, Anal. Chem. 71 (1999) 811. [13] M.E. Abdelsalam, G. Denuault, M.A. Baldo, S. Daniele, J. Electroanal. Chem. 449 (1998) 5. [14] S. Daniele, M.A. Baldo, C. Bragato, Anal. Chem. 71 (1999) 811. [15] M.E. Abdelsalam, G. Denuault, M.A. Baldo, C. Bragato, S. Daniele, Electroanalysis 13 (2001) 289. [16] S. Daniele, M.A. Baldo, C. Bragato, Anal. Chem. 74 (2002) 3290. [17] I. Ciani, S. Daniele, J. Electroanal. Chem. 564 (2004) 133. [18] A. Abu-Rabi, D. Jašin, S. Mentus, J. Electroanal. Chem. 600 (2007) 364. ˇ oz, R.G. Compton, [19] O. Ordeig, C.E. Banks, T.J. Davies, F.J. del Campo, F.X. Mun Gold ultra-microelectrode arrays: application to the steady-state voltammetry of hydroxide ion in aqueous solution, Anal. Sci. 22 (2006) 679–683. [20] N.A. Choudhry, D.K. Kampouris, R.O. Kadara, N. Jenkinson, C.E. Banks, Anal. Methods 1 (2009) 183. [21] R.O. Kadara, N. Jenkinson, C.E. Banks, Electroanalysis 21 (2009) 2410.
T. Tuvic´ et al. / Journal of Electroanalytical Chemistry 654 (2011) 102–107 [22] P.M. Hallam, D.K. Kampouris, R.O. Kadara, N. Jenkinson, C.E. Banks, Nickel oxide screen printed electrodes for the sensing of hydroxide ions in aqueous solution, Anal. Methods 2 (2010) 1152–1155. [23] O.E. Linarez Pérez, V.C. Fuertes, M.A. Pérez, M. López Teijelo, Electrochem. Commun. 10 (2008) 433–437. [24] M. Anik, K. Osseo-Asare, J. Electrochem. Soc. 149 (2002) B224. [25] M. Anik, T. Cansizoglu, J. Appl. Electrochem. 36 (2006) 603. [26] M. Anik, Electrochim. Acta 54 (2009) 3943. [27] P.I. Ortiz, M. López Teijelo, M.C. Giordano, J. Electroanal. Chem. 243 (1988) 379. [28] R.M. Torresi, O.R. Camara, C.P. de Pauli, M.C. Giordano, Electrochim. Acta 32 (1987) 1291.
107
[29] S.H. Aityan, A.D. Davydov, B.N. Kabanov, Elektrokhimiya 8 (1972) 1391. [30] A.D. Davydov, V.S. Shaldaev, A.N. Malofeeva, I.V. Savotion, J. Appl. Electrochem. 27 (1997) 351. [31] A.D. Davydov, V.S. Krylov, G.R. Engelgardt, Elektrokhimiya 16 (1980) 192. [32] O.E. Barcia, O.R. Mattos, N. Pébère, B. Tribollet, Electrochim. Acta 41 (1996) 1385. [33] G.S. Kelsey, J. Electrochem. Soc. 124 (1977) 814. [34] P. Vany´sek, in: D. R. Linde (Ed.), Handbook of Chemistry and Physics, 84th ed., CRC Press LLC, 2003–2004 section 5; pp. 93–94. [35] P.I. Ortiz, M. López Teijelo, M.C. Giordano, An. Asoc. Quim. Argent. 74 (1986) 539–553.