Quantitative Studies on Electrode Material Properties by Means of the

TECHNICAL NOTE pubs.acs.org/ac

Quantitative Studies on Electrode Material Properties by Means of the Cavity Microelectrode Cristina Locatelli, Alessandro Minguzzi, Alberto Vertova, Paola Cava, and Sandra Rondinini* Dipartimento di Chimica Fisica ed Elettrochimica, Universit
a degli Studi di Milano, Via Golgi, 19, 20133, Milan, Italy ABSTRACT: The determination of the number of active sites is a key issue in the evaluation of electrode materials for any electrochemical application. Nonetheless, and particularly in the case of powder materials, a commonly accepted method to determine the actual density of active sites is not yet available, mainly because a method to quantify the amount of material under investigation is missing. In this study, we propose the use of the cavity microelectrode (C-ME, i.e., a cylindrical recessed microdisk) of known volume, which enables the study of a known amount of material, thus allowing the quantitative evaluation of its properties. The validation of the method implied (i) the preparation of C-MEs with different radii and depths; (ii) the calibration of the relevant volumes by means of a “standard” powder, whose number of active sites can be easily determined by cyclic voltammetry; and (iii) their use for the quantification of specific parameters that identify the electrochemical properties of mixed IrO2-SnO2 powders. The results evidence that the quantity of charge relative to the number of pseudocapacitance sites and the currents for the oxygen evolution reaction are proportional to the cavity volumes. This strategy allows the direct comparison of different materials for their rapid and accurate screening. In addition, thanks to the small amount of material required for the sample (typically 10-100 ng), the method can be safely listed among the nondestructive techniques.

O

ne of the main topics in electrochemistry is the research for electrocatalysts for energy conversion/storage devices1 which mostly adopt membrane-electrode assemblies (MEA) or gas-diffusion layers/electrodes (GDL/GDE) and filter-press or swiss-roll cell structures. This has shifted the interest from the dimensionally stable anodes (DSAs), obtained by thermal decomposition of suitable salt precursors onto metal supports,2-4 to the preparation of electrocatalytic micro- or nanopowders.5,6 The complexity of the overall sequence of reactions on multiphase, multifunctional electrodes like GDEs and MEAs, which implies the proper contact between solid, liquid, and gas phases, calls for materials that can be easily dispersed in a multicomponent, composite matrix.7 Furthermore, finely dispersed materials guarantee high specific surface areas; this extends their applications to charge accumulation and storage (e.g., supercapacitors). It is also worthwhile to mention the investigation on the properties of materials (metal or metal oxides, carbons) used as supports for active nanoparticles. Preliminary studies on powder electrode materials are often performed by deposition onto a conductive support that presents negligible activity toward the investigated reactions. This strategy usually presents several disadvantages: (i) the support may introduce its own contribution in the electrochemical response of the powder; (ii) the use of a gluing agent (e.g., an ionconducting polymer) is needed for the adhesion of the particles onto the support; (iii) the amount of supported powder is often hard to be set and then to be evaluated; (iv) high ohmic drops may be introduced when the powder conductivity is not metallike, thus distorting the voltammetric signal, especially at increasing potential scan rates. The present communication discusses the preparation, the use, and the advantages of the cavity microelectrode as an r 2011 American Chemical Society

investigation tool for electrocatalytic powders. Cavity microelectrodes (C-MEs) are disk microelectrodes in which the metal wire is recessed to produce a cylindrical microcavity with internal glass walls and metal bottom. C-MEs can host very small amounts of powder (typically 10-7-10-8 g), as result of the filling of the cavity volume. C-MEs are easily prepared, low-cost tools and, according to the proposed procedure, their geometric parameters (radius and depth) can be accurately determined. C-MEs do not exhibit the disadvantages cited above; moreover, they offer all the advantages bound to their micrometric size: (i) the low current values, due to their small dimensions and the small amount of powder under investigation, minimize the effect of ohmic drops on the electrochemical signal; (ii) the electrochemical response is free from contributions of any gluing agent, and (iii) the contribution of the current collector (that is the microdisk at the base of the cavity) is negligible since its surface area is at least 2 orders of magnitude lower than the one of the hosted amount of powder. The only exception to item (iii) is represented by materials having a very low electronic conductivity. In this case, the electrolyte solution is much more conductive than the C-ME filling and the metal microdisk becomes the major (or the only) contributor to the electrochemical signal. For these reasons, C-MEs have been already used for the study of several materials for electrochemical applications: carbons,8 conductive polymers,9 battery materials10,11 and electrocatalysts.12-14 These studies involved the analysis of capacitive and pseudocapacitive (solid-state redox transitions) properties,12,13 and the qualitative comparison between materials over the same probe Received: November 11, 2010 Accepted: February 12, 2011 Published: March 03, 2011 2819

dx.doi.org/10.1021/ac200286q | Anal. Chem. 2011, 83, 2819–2823

Analytical Chemistry molecule for waste remediation15,16 and for energy conversion processes.14 Therefore, C-ME is a versatile tool which can be even used in the cases of gas-evolving reactions, provided that the formation of gas bubbles is avoided, and of processes controlled by the diffusion of reactants, provided that an appropriate model of the diffusion inside the cavity and between the particles is formulated. To the best of our knowledge, C-MEs have been used only for qualitative studies, whereas the quantitative analysis of the material features, i.e., the normalization of key parameters (accumulated quantity of charge, current-voltage characteristics, number of active sites) to the powder amount, has not yet been introduced. In this work, the use of C-MEs for quantitative investigations is applied and validated via the preparation of cavities having different known volumes. In our opinion, the quantitative analysis also demonstrates that C-MEs represent a very promising tool for the rapid, almost nondestructive, and in-depth study of electrode materials (e.g., rapid screening of material libraries). To this end, two kinds of electrocatalysts are used, namely, the popular and commercially available carbon-supported platinum and the mixed IrO2þSnO2 powders, which represent a promising class of electrocatalysts for oxygen evolution/reduction reactions in both aqueous17-19 and nonaqueous media (e.g., for fuel cells and lithium-air batteries20) as well as for electrooxidation of organic pollutants.21,22 IrO2 shares with RuO2 the role of best electrocatalyst for oxygen evolution reaction (OER) in acidic media and, because of its high cost, is often studied as diluted into low-cost, stable matrixes (SnO2,17,18 Ta2O5,23 Sb2O524). IrO2 is also a promising material for supercapacitors.25 The determination of the density of active sites is directly connected with the key issues of catalyst loading, cost, and natural abundance. As discussed below, C-MEs enable the evaluation of the active site density and the fraction of iridium sites that participate to the electron transfer reaction. The turnover frequency (TOF) is also determined, an intensive kinetic parameter totally independent of the amount of electrode loading or of its surface extension.

’ EXPERIMENTAL SECTION Electrochemical Measurements. Cyclic voltammetries (CVs) and chronoamperometries (CAs) are recorded using an EG/G 263A potentiostat/galvanostat (Princeton Applied Research, Oak Ridge, TN, USA) driven by Power Suite (PAR) or CorrWare (Scribner Associates Inc., Southern Pines, USA) software. A classic 3-electrode cell is used, with a Pt foil and a double-bridge saturated calomel electrode (SCE) as the counter and the reference electrodes, respectively. All the electrode potentials are quoted versus the reversible hydrogen electrode (RHE), by periodically calibrating the SCE against the RHE at different pH values. The C-ME is mounted in an upright position to avoid accidental trapping of N2 bubbles during the solution degassing (20 min). The cell is then maintained in nitrogen atmosphere. All measurements are performed in 0.5 M H2SO4. All solutions are prepared with Milli-Q grade water. Synthesis of IrO2þSnO2 Mixed Oxide Powders. IrO2þSnO2 mixed powders are prepared as follows: the tin oxide xerogel precursor is synthesized via sol-gel as described in refs 26 and 27 and dried at 120 °C for 2 h. The selected amount of IrCl3 3 3H2O (Alfa Aesar, 99.9% purity, metal bases), dissolved in the minimum amount of aqueous HCl, is diluted with 2-propanol (50 mL of solvent/g-of-precursor) and added to the SnO2

TECHNICAL NOTE

xerogel. Finally, the powder is calcined at 450 °C for 3 h after a 2 h ramp. Preparation of the C-MEs. C-MEs are prepared following the procedure reported in refs 15 and 16. Briefly, a Pt wire (Goodfellow, 99.9% purity, temper hard, Ø = 25, 50, 63 μm ( 10%) is sealed into a glass tube by means of an air/liquid petroleum gas flame. The Pt disk is first exposed and then polished using emery papers (400-800-1000-2400-4000 mesh) and alumina powders (mean particle size 0.3 and 0.05 μm). The contact between the Pt wire and the external Cu wire connector is made with graphite powder. The upper end of the glass tube is finally sealed by a silicon paste. The experimental radius, r (μm), is determined via the steadystate current recorded in 0.1 M KNO3 (Carlo Erba, p.a. 99.9% purity) in the presence of 1 mM Ru(NH3)6Cl3 (Sigma Aldrich, 98%) at 1 mV s-1 in the 0/-0.4 V vs SCE potential window, at 25 °C:28 Iss ¼ 4nFcb Dr

ð1Þ

where r (μm) is the geometric metallic disk radius, Iss (A) is the steady-state current intensity, cb (mol μm-3) and D (μm2 s-1) are the concentration and diffusion coefficient of the Ru(III) complex, respectively, n is the mol of electron/mol of reactant ratio, and F is the Faraday constant (C mol-1). The Pt wire is then progressively etched by immersion in boiling aqua regia for 5 min. The procedure is repeated until the desired depth is reached, checking the depth of the recess, L (μm), after each etching step. The estimation is performed via the steady-state limiting current of the ruthenium-hexamine complex, as above, according to the following equation,29 where L is the only unknown quantity: Iss ¼

4πnFcb Dr 2 4L þ πr

ð2Þ

The cavity is now ready to be filled with the investigated powder using the electrode as a pestle. Figure 1 schematically represents a C-ME after its preparation. Finally, the correct filling of the cavity is controlled with an optical microscope, and the electrode is ready to be used as working electrode in the voltammetric experiments. All measurements are repeated at least three times for reproducibility.

’ RESULTS AND DISCUSSION Geometric versus Operational Cavity Volumes. It has to be noted at this point that, because the unavoidable irregularities of the cavity bottom (e.g., the etching rate of the Pt disk at the glass is higher at the wall than at its center), eq 2 returns an average L value. In turn, L leads to the geometric cavity volume, Vg (μm3), that can significantly differ from the one actually filled by the target powder, dependent on the packing of the particles. This obstacle can be overcome by calibration with a powder of known properties. In detail, the cavity is filled with a Pt/C powder (28.7% Pt/Vulcan XC-70, as purchased from E-TEK) and CVs are recorded in 0.5 M H2SO4 aqueous solution between 0 and 1.4 V vs RHE, at 20 mV s-1. Assuming a narrow distribution of the particle size and a good dispersion of Pt onto the carbon support, the volume of each cavity is proportional to the Pt surface area, which, in turn, is determined by integration of the hydrogen adsorption/desorption peaks as reported in ref 30. Plotting the relevant quantities of 2820

dx.doi.org/10.1021/ac200286q |Anal. Chem. 2011, 83, 2819–2823

Analytical Chemistry

TECHNICAL NOTE

Figure 1. Schematic representation of a C-ME. The values of r and L can be tuned at will by choosing the wire diameter and the recess depth.

Table 1. Geometric and Operational Values of the Cavity Volumes, Vg and Vo, Together with the Corresponding Radius and Depth, for Different C-MEsa # C-ME

a

Figure 2. Average quantity of charge Q (C), with error bars, relevant to the hydrogen adsorption/desorption peaks for the C-MEs filled with E-TEK Pt/C as a function of the geometric volume, Vg. The Q value at Vg = 0 is relevant to the Pt microdisk at the bottom of the C-ME. The solid line represents the linear regression obtained from the weighted least-squares method. Inset: CV recorded in 0.5 M H2SO4 at 20 mV s-1, C-ME # 4.

charge, Q (C), versus Vg, a linear behavior is obtained, as shown in Figure 2 (the solid line representing the linear regression obtained from the weighted least-squares method). Note that reproducibility, here denoted by the error bars, is high and does not justify the spreading of the data, which has to be correctly ascribed to the irregularities of the C-MEs’ bottoms. Now, the volume actually occupied by the Pt/C powder in a given C-ME, i.e., the “operational volume” Vo (μm3) of the cavity, is then obtained on the basis of the linear regression by taking the relevant Q as the known quantity. The results are summarized in Table 1. As already mentioned, the differences between Vg and Vo depend on the irregularity of the cavity bottom and typically range from a few percents (C-ME 5) to a maximum of 30% (C-ME 3). Improving the etching procedure will smooth the irregularities and, hence, the observed differences. Work is in progress, and the results will be the subject of a future publication. Characterization of IrO2þSnO2 Powder. This paragraph is devoted to prove that the C-ME allows the quick, accurate, and reproducible full characterization of IrO2þSnO2 powders. The discussion is limited to the determination of the active site density and the estimation of TOF, but the use of C-MEs can

r/μm

L/μm

Vg/103 μm3

Vo/103 μm3

1

14.5 ( 1.8

30 ( 9

19.4 ( 7.4

14.3 ( 4.4

2

27.8 ( 1.8

27 ( 5

65.9 ( 14.6

86.0 ( 3.6 106.5 ( 4.4

3

36.4 ( 1.8

20 ( 3

82.3 ( 16.2

4

28.8 ( 1.8

43 ( 7

110.8 ( 22.2

90.7 ( 3.6

5

28.7 ( 1.8

45 ( 7

116.2 ( 23.3

103.6 ( 4.4

Data are listed at increasing Vgs.

be extended at will, provided that a few experimental conditions will be satisfied (see Introduction). The IrO2þSnO2 powder, selected for the characterization via C-MEs of known operational volumes, has a nominal composition of 15 mol % of IrO2. The investigation is performed by CA and CV at fixed 1.4 V vs RHE (CA) and between 0.4 and 1.4 V vs RHE (CV, ν = 2, 5, 10, and 20 mV s-1), and the quantity of charge, obtained by integrating the CV curve between 0.4 and 1.3 V vs RHE (a potential window which excludes the fast oxygen and hydrogen evolution reactions), is taken as representative of the amount of active material in the catalyst sample. As for many conductive oxides, in this potential window and in the absence of specifically adsorbing ions, the IrO2þSnO2 powder shows only capacitive and pseudocapacitive features. The latter are usually described by the following equation: IrOx ðOHÞy þ δHþ ðsolutionÞ þ δe- ðoxideÞ ¼ IrOx-δ ðOHÞy þ δ

ð3Þ

which implies the exchange of protons between the iridium site and the solution, which in turn is at the basis of the “bumps” typically observed in the CVs of IrO2 based materials, as shown in the inset of Figure 3. To determine the number of both the total sites and the most accessible ones,17 the procedure implies the recordings at different scanning rates. Since in our case the integrated quantity of charge is almost independent of v, we report only the values relative to 20 mV s-1. 2821

dx.doi.org/10.1021/ac200286q |Anal. Chem. 2011, 83, 2819–2823

Analytical Chemistry

Figure 3. Quantity of charge Q (C) calculated by integration of the voltammetric curve of IrO2þSnO2 powder as a function of the C-ME operational volume, Vo. Inset: CV recorded in 0.5 M H2SO4 at 20 mV s-1, C-ME # 4.

TECHNICAL NOTE

also studied in detail by Fierro et al.35,36 Our results in Figures 3 and 4 point to the fact that the activity of the material toward OER (i.e., the rate of oxygen evolution at constant potential) is proportional to the volume of the powder that, in turn, is proportional to the number of sites that are known to quickly exchange protons with the solution, as determined in a lower potential window, i.e., in a region in which the occurrence of pseucapacitive phenomena is very well-known.35,37 Therefore, the number of active sites toward OER is proportional (not necessarily equal) to the sites that are allowed to exchange protons with the solution. The linear relation found on Figure 4 makes this sentence incontrovertible. As first approximation, the number of active sites for the IrO2þSnO2 powder discussed in this paper is equal to the number of sites that can freely exchange protons with the solution. On this basis the turnover frequency, i.e., the intrinsic parameter usually adopted to evaluate the activity of a catalyst, is evaluated by T:O:F: ¼

Figure 4. Chronoamperometric steady-state current of IrO2þSnO2 as a function of the C-ME operational volumes, Vo.

Figure 3 shows the remarkable linear dependence of Q’s on the operational cavity volumes, that is the volume occupied by the powder. The figure demonstrates that C-MEs allow the normalization, with respect to the volume (and, if the mass density is known, also to the weight), of Q and of any other extensive parameter, thus making the direct comparison between different materials possible. For the present IrO2þSnO2 system, the specific number of sites31 average value is (7.3 ( 0.6) 10-5 mol g-1, as obtained by the slope of the Q/Vo characteristics. Notwithstanding the difficulties already envisaged in the case of gas evolving reactions, C-MEs also enable the characterization of materials under OER conditions, provided that turbulent gas evolution is avoided. This was confirmed by polarizing IrO2þSnO2 samples at relatively low overpotentials (1.4-1.6 V vs RHE) and recording the corresponding currents, until stability. Interestingly, the C-ME filling was stable up to 1.6 V (RHE), i.e., until the gas saturates the solution inside the cavity and bubbles are formed. Figure 4 shows the steady-state currents, Iss (nA), recorded on different C-MEs in dependence on the operational volumes. Once again, a linear relationship is obtained, thus proving that the oxygen evolution rate linearly depends on the volume of the powder inserted in the cavity, in agreement with the results obtained by Comninellis et al.32 This graph then represents the first quantitative evidence of OER as proton-coupled-electrontransfer reaction (PCET).33 This result was already envisaged by Trasatti34 who suggested that the most active sites for OER are also able to freely exchange protons with the solution; that is the freedom of exchanging protons with the solution is a needed property to be an active site for OER. The importance of the relation between the electrochemical surface properties (by CV) of metal oxide electrodes with their electrocatalytic behavior was

Iss ns nF

ð4Þ

where ns is the number of sites determined with each cavity, Iss is the corresponding oxygen evolution current, and n = 4 is the number of electrons per mol of O2. The mean value of 1.3 ( 0.1  10-3 s-1 at 1.4 V (RHE) seems absolutely reasonable if one considers the very low overpotential at which the TOF is evaluated. The quantitative evaluation of the material properties in terms of the specific number of sites and of the turnover frequency makes C-MEs a new, excellent tool for the rapid and deep evaluation of several electroactive materials and for a wide spectrum of reactions.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Fax: þ390250314203.

’ ACKNOWLEDGMENT The financial contributions of PUR (2009-2010) and PRIN 2008-2008N7CYL5_004 funds are gratefully acknowledged. C.L. and P.C. gratefully acknowledge the Oronzio and Niccol
o De Nora Foundation for their Research Fellowships. ’ REFERENCES (1) Bard, A. J. J. Am. Chem. Soc. 2010, 132, 7559. (2) Foti, G.; Mousty, C.; Reid, V.; Comninellis, C. Electrochim. Acta 1998, 44, 813. (3) Trasatti, S. Electrochim. Acta 2000, 45, 2377. (4) Millet, P.; Andolfatto, F.; Durand, R. Int. J. Hydrogen Energy 1996, 21, 87. (5) Marshall, A.; Børresen, B.; Hagenm, G.; Tsypkin, M.; Tunold, R. Electrochim. Acta 2006, 51, 3161. (6) Ma, L.; Sui, S.; Zhai, Y. Int. J. Hydrogen Energy 2009, 34, 678. (7) Litster, S.; McLean, G. J. Power Sources 2004, 130, 61. (8) Cachet-Vivier, C.; Vivier, V.; Cha, C. S.; Nedelec, J.-Y.; Yu, L. T. Electrochim. Acta 2001, 47, 181. (9) Vivier, V.; Belair, S.; Cachet-Vivier, C.; Cha, C. S.; Nedelec, J.-Y.; Yu, L. T. Electrochem. Commun. 2000, 2, 180. (10) Vivier, V.; Belair, S.; Cachet-Vivier, C.; Nedelec, J.-Y.; Yu, L. T. J. Power Sources 2001, 103, 61. 2822

dx.doi.org/10.1021/ac200286q |Anal. Chem. 2011, 83, 2819–2823

Analytical Chemistry

TECHNICAL NOTE

(11) Merzouki, A.; Cachet-Vivier, C.; Vivier, V.; Nedelec, J.-Y.; Yu, L. T.; Haddaoui, N.; Joubert, J.-M.; Percheron-Guegan, A. J. Power Sources 2002, 109, 281. (12) Vivier, V.; Cachet-Vivier, C.; Mezaille, S.; Wu, B. L.; Cha, C. S.; Nedelec, J.-Y.; Fedoroff, M.; Michel, D.; Yu, L. T. J. Electrochem. Soc. 2000, 147, 4252. (13) Vivier, V.; Regis, A.; Sagon, G.; Nedelec, J.-Y.; Yu, L. T.; CachetVivier, C. Electrochim. Acta 2001, 46, 907. (14) Guilminot, E.; Corcella, A.; Chatenet, M.; Maillard, F. J. Electroanal. Chem. 2007, 599, 111. (15) Vertova, A.; Barhdadi, R.; Cachet-Vivier, C.; Locatelli, C.; Minguzzi, A.; Nedelec, J.-Y.; Rondinini, S. J. Appl. Electrochem. 2008, 38, 965.  .; Locatelli, C.; Minguzzi, A.; (16) Rondinini, S.; Aricci, G.; Krpetic, Z Porta, F.; Vertova, A. Fuel Cells 2009, 3, 253. (17) Ardizzone, S.; Bianchi, C. L.; Borgese, L.; Cappelletti, G.; Locatelli, C.; Minguzzi, A.; Rondinini, S.; Vertova, A.; Ricci, P. C.; Cannas, C.; Musinu, A. J. Appl. Electrochem. 2009, 39, 2093. (18) Minguzzi, A.; Alpuche-Aviles, M. A.; Rodriguez Lopez, J.; Rondinini, S.; Bard, A. J. Anal. Chem. 2008, 80, 4055. (19) Takasu, Y.; Yoshinaga, N.; Sugimoto, W. Electrochem. Commun. 2008, 10, 668. (20) Debart, A.; Bao, J.; Armstrong, G.; Bruce, P. G. J. Power Sources 2007, 174, 1177. (21) Morozov, A.; De Battisti, A.; Ferro, S.; Martelli, G. N. Int. Patent WO 2005/014885 A1, 2005. (22) Fierro, S.; Ouattara, L.; Calderon, E. H.; Passas-Lagos, E.; Baltruschat, H.; Comninellis, Ch. Electrochim. Acta 2009, 54, 2053. (23) Ardizzone, S.; Bianchi, C. L.; Cappelletti, G.; Ionita, M.; Minguzzi, A.; Rondinini, S.; Vertova, A. J. Electroanal. Chem. 2006, 589, 160. (24) Chen, G.; Chen, X.; Yue, P. L. J. Phys. Chem. B 2002, 106, 4364. (25) Liu, D. Q.; Yu, S. H.; Son, S.-W.; Joo, S.-K. ECS Trans. 2008, 16, 103. (26) Ardizzone, S.; Cappelletti, G.; Ionita, M.; Minguzzi, A.; Rondinini, S.; Vertova, A. Electrochim. Acta 2005, 50, 4419. (27) Ionita, M.; Cappelletti, G.; Minguzzi, A.; Ardizzone, S.; Bianchi, C.; Rondinini, S.; Vertova, A. J. Nanopart. Res. 2006, 8, 653. (28) Bard, A. J.; Faulkner, L. R. Electrochemical methods: fundamentals and applications, 2nd ed.; John Wiley & Sons: New York, NY, 2001; p 174. (29) Bond, A. M.; Luscombe, D.; Oldham, K.; Zoski, C. G. J. Electroanal. Chem. 1988, 249, 1. (30) Schmidt, T. J.; Gasteiger, H. A.; Stab, G. D.; Urban, P. M.; Kolb, D. M.; Behm, R. J. J. Electrochem. Soc. 1998, 145, 2354. (31) Obtained from the slope of the regression straight line shown in Figure 3. The density of sites is calculated on the basis of the densities of IrO2 (11.66 g cm-3) and SnO2 (6.95 g cm-3), a compact sphere packing (both FCC and HC packing have the same occupied volume fraction, 0.74), and by considering that in the selected interval two redox transitions occur, namely, Ir(III)/Ir(IV) and Ir(IV)/Ir(V). This would also mean that the number of Ir atoms involved in the pseudocapacitive processes is about 10% of the total iridium content. (32) Calderon, E. H.; Katsaounis, A.; W€uthrich, R.; Mandin, P.; Foti, G.; Comninellis, C. J. Appl. Electrochem. 2009, 39, 1827. (33) Costentin, C.; Robert, M.; Saveant, J.-M. Chem. Rev. 2010, 110, PR1. (34) Trasatti, S. Electrochim. Acta 1990, 35, 263. (35) Fierro, S.; Ouattara, L.; Calderon, E. H.; Comninellis, C. Electrochem. Commun. 2008, 10, 955. (36) Fierro, S.; Nagel, T.; Baltruschat, H.; Comninellis, C. Electrochem. Commun. 2007, 9, 1969. (37) Vertova, A.; Borgese, L.; Cappelletti, G.; Locatelli, C.; Minguzzi, A.; Pezzoni, C.; Rondinini, S. J. Appl. Electrochem. 2008, 38, 973.

2823

dx.doi.org/10.1021/ac200286q |Anal. Chem. 2011, 83, 2819–2823