Journal of
Electroanalytical Chemistry Journal of Electroanalytical Chemistry 589 (2006) 160–166 www.elsevier.com/locate/jelechem
Composite ternary SnO2–IrO2–Ta2O5 oxide electrocatalysts S. Ardizzone, C.L. Bianchi, G. Cappelletti, M. Ionita 1, A. Minguzzi, S. Rondinini *, A. Vertova Department of Physical Chemistry and Electrochemistry, The University of Milan, Via Golgi, 19, I-20133 Milan, Italy Received 9 November 2005; received in revised form 19 January 2006; accepted 7 February 2006 Available online 20 March 2006
Abstract Nanostructured, ternary SnO2–IrO2–Ta2O5 composites have been studied as promising electrocatalytic materials for oxygen evolution reaction (OER) in acid water electrolysers. Both unsupported and Ti-supported powders of the general composition Sn1xyIrxTayO2+2.5y (x = 0, 0.03, 0.07, 0.15 and y = 0, 0.07) have been prepared by the same sol–gel route, and characterized in terms of structural (XRD) and morphological (SEM, BET) features, surface composition (XPS), and electrochemical behaviour (cyclic voltammetry, CV, electrochemical impedance spectroscopy, EIS, and slow potentiodynamic techniques). The results point out a dramatic variation of crystallite size and surface area with Ir and Ta doping of the SnO2 base matrix, accompanied by an improvement of the electronic and electrocatalytic properties of the composites. 2006 Elsevier B.V. All rights reserved. Keywords: Sn–Ir–Ta oxide composites; Oxygen evolution; Sol–gel; XPS; Nanostructured materials; Electrocatalytic activity
1. Introduction Oxygen evolution in acid environment represents a very severe test for electrocatalysts. Only precious metal oxides are relatively stable. Among these, IrO2 is, in principle, the most resistant. However the application of pure iridium oxide coatings is strongly restricted by high costs and limited electrode lifetime. Consequently composite materials, where the precious compound is dispersed in a less active but more stable matrix, are being intensively studied [1–21] to offer less expensive electrodes which might show, at the same time, good electrocatalytic activity, stability toward anodic dissolution and electronic conductivity. In binary systems, the combination of active IrO2 and inert Ta2O5 is reported to exhibit good performance in *
Corresponding author. Tel.: +39 02 5031 4217; fax: +39 02 5031 4203. E-mail address:
[email protected] (S. Rondinini). 1 On leave from the Faculty of Industrial Chemistry, Applied Physical Chemistry and Electrochemistry Department, University Politechnica of Bucharest. 0022-0728/$ - see front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jelechem.2006.02.004
anodic stability and electrocatalytic activity [5–13]. The metal oxide coating consisting of Ir and Ta oxides, thermally prepared on a titanium substrate is, actually, one of the most frequently adopted catalyst for oxygen evolution in industrial electroplating processes [10]. Studies on electrocatalytic activity and durability of this type of anode have been extensively carried out by different research groups. It is generally observed that coatings with IrO2 content between 55 and 70 mol% display the highest activities for oxygen evolution. IrO2–SnO2 mixtures have also been extensively investigated with respect to oxygen evolution in acid environment also due to the similarity between the two oxide structures [2,3,19]. De Pauli et al. [2,3] have observed, in the case of samples prepared by thermal decomposition of the metal salts, a noticeable surface enrichment in IrO2 accompanied by an increase of the surface charge of the composite material in comparison with the pure oxides. Consistently, for electrodes containing more than 10% of precious metal oxide, O2 evolution from acidic solutions proceeded with kinetic parameters close to those of pure IrO2.
S. Ardizzone et al. / Journal of Electroanalytical Chemistry 589 (2006) 160–166
On the grounds of the very promising performances shown by both IrO2–SnO2 and IrO2–Ta2O5 mixtures we have recently started an investigation on the synthesis and characterization of composite-ternary systems SnO2– IrO2–Ta2O5. To the authors’ best knowledge no ternary composites based on the Sn + Ir + Ta oxides have been previously studied with respect to the oxygen evolution reaction (OER) in acid media. In the general case of composite materials, the electrocatalytic behaviour is the result of a subtle balance between different, often diverging, effects like e.g. the surface composition, the layer morphology, the crystallite sizes and the overall electrical conductance. The adopted synthetic procedure plays a key role in imposing on the final performance of the material. Recently we documented a lowtemperature sol–gel synthetic process to produce tailored nanostructured SnO2 exhibiting the typical semiconducting properties connected with the emptying/filling of electronic traps located within the band gap [22,23]. On the grounds of these previous results, the present ternary composites are prepared, by a common sol–gel path, both as Ti supported and unsupported particles by adding to the starting tin alcoxide/solvent mixture the guest ion salts. After hydrothermal growth and calcination the products are characterized for phase composition and crystallinity and surface area-porosity. Particular interest bears the assessment, by XPS analyses, of the surface state and composition to evidence possible surface segregation-enrichment of the minority components. The electrochemical response was investigated by CV, EIS, and slow potentiodynamic techniques.
161
the electrodes were kept at 0.9 V for 5 min. All potentials are referred to the reversible hydrogen electrode (RHE). Scanning electron microscopy (SEM) photographs are acquired by LEO 1430. 3. Results and discussion 3.1. Structural and morphological features The composite materials were prepared by a sol–gel process, using a tin (IV) alkoxide as the starting compound and adding, directly to the reacting mixture, the appropriate amounts of Ir and Ta salts. The conditions of the sol– gel reaction (water/alkoxide and water/alcohol ratios) were selected such as to promote the formation of materials with reduced crystallite sizes, after the calcination [24,25]. Fig. 1 reports the X-ray diffraction lines of the pure SnO2, Ir and Ta-doped SnO2 (Sn0.85Ir0.15O2, Sn0.78Ir0.15Ta0.07O2.175) materials calcined at 500 C. All the XRD patterns show only the peaks of the SnO2 cassiterite structure. In principle, the most intense peaks of SnO2 and IrO2 ð2hSnO2 ¼ 26:7 ; 2hIrO2 ¼ 28:8 Þ are sufficiently far apart to allow the formation of a separate IrO2 phase to be appreciated; however, in the present case, the general low degree of crystallinity of the samples, rules out the possibility to either confirm or exclude the presence of minority phase components. The same arguments apply also to possible phases related to the presence of Ta in the composite.
2. Experimental section
Intensity / a. u.
SnO2
The composite Ta, Ir doped SnO2 particles were obtained by room-temperature sol–gel reaction, as previously reported [22,23], starting from Sn(C4H9O)4 and adopting IrCl3 · 3H2O (Ir molar fractions = 0.03, 0.07, 0.15) and TaCl5 (constant Ta molar fraction = 0.07) as the dopant salts. The dried xerogels were thermally treated at 500 C for 2 h under oxygen flux. The Ti supported powders were prepared by dipping/drying cycles in the same precursor mixture used for the particles. XRD was performed by a Siemens D500 diffractometer, using CuKa radiation. Surface area and pore size distribution were obtained from nitrogen adsorption/desorption isotherms at 196 C in a Coulter SA 3100 Analyzer. XPS spectra were obtained by using an M-probe apparatus (Surface Science Instruments), adopting the same conditions reported previously [23]. CV determinations were performed by using AMEL 5000 Potentiostat/Galvanostat. EIS measurements were performed using Solartron Analytical Potentiostat (model 1287) and Frequency Response Analyzer (model 1260), at fixed electrode potential and at a sweeping frequency from 105 to 102 Hz. Polarization curves were recorded stepwise at 10 mV/min in the 1.4– 2.0 V potential range. At the end of the last backward scan
Sn0.85Ir0.15O2
Sn0.78Ir0.15Ta0.07O2.175
0
20
40
60
80
2θ Fig. 1. X-ray diffraction lines of pure (SnO2) and doped (Sn0.85Ir0.15O2, Sn0.78Ir0.15Ta0.07O2.175) tin oxide nanoparticles.
162
S. Ardizzone et al. / Journal of Electroanalytical Chemistry 589 (2006) 160–166
The addition of either Ir or Ta species to the mixture provokes, invariably, a decrease in crystallinity, immediately appreciable from the pattern of the diffractograms. A more direct information on the structural effects bound to the material composition can be obtained by the trend of the crystallite sizes, obtained by elaboration of the fitted Xray peaks by the Scherrer’s equation (Fig. 2a). In the absence of Ta the crystallite sizes progressively decrease with the Ir content, the variation being relevant for the higher Ir fractions. In the presence of Ta the crystallite sizes are much lower than the pure SnO2; a further decrease in the crystallite size upon addition of Ir species is also appreciable. In the case of IrO2–Ta2O5 mixtures [5,11], a decrease of the crystallite size, with the respect to pure IrO2 is reported; the effect is accompanied by Ta2O5 enrichment at the crystallite surface and consequent suppression of the crystallite growth [11]. In the present case the effect is somewhat more complicated as Ta species are not enriched at the surface of the composite. These considerations will
9
Sn1-xIrxO2 Sn0.93-xIrxTa0.07O2.175
d / nm
8
7
6
5
4 0.00
a
0.03
0.07
0.15
Ir molar fraction 100 90 80
SB.E.T. / m g
2 -1
Sn1-xIrxO2 Sn0.93-xIrxTa0.07O2.175
70 60 50
Table 1 Collection of literature ionic radii
40 0.00
b
be resumed in the following section while discussing XPS results. Solid solutions are reported to form in the case of both SnO2–IrO2 and Ta2O5–IrO2 composites [10,11,19]. On the grounds of the similarity between Ir and Ta ionic radii [5,14,19,26,27] (see Table 1) it can be suggested that, in the present case, Ta ions compete with Ir in the reticular substitution of Sn in the cassiterite lattice. The promotional mechanism of Ta in the decrease of the crystallite size can be therefore traced back to a hindered growth of the crystal phase caused by the high distortion effects introduced by the guest species in the lattice shape. Besides the mechanism, the Ta promoted reduction of the crystallite sizes is desirable to improve the electrocatalytic behaviour of the material. The OER is promoted by smaller crystallite sizes as the reaction will preferably take place at particle defects, kinks, cracks and corners. For increasingly small crystallites the active sites become closer to each other and facilitate the rearrangement step. Further Hu et al. [5] suggest that the smaller the crystallite the larger the activation of the material during the gas evolution. Fig. 2b reports the powder surface areas. The trend with the Ir content in absence of Ta species closely mirrors the behaviour of the crystallite sizes. The addition of Ta provokes a progressive increase in the surface areas which almost double at the highest Ir content. The smaller crystallite sizes determine the larger surface areas, even if, due to the sintering of the crystallites, the actual surface areas are lower than expected, the more so in the absence of Ta. The dependence of the crystallite size and surface area on the composition of the composites (Fig. 2) is closely mirrored by the particles morphology as apparent in the SEM micrographs (Fig. 3). The pure SnO2 powder (Fig. 3a) appears to be composed both by isolated pseudo-spherical particles with an average size of 20– 30 nm (Fig. 3a2) and also by irregularly shaped aggregates (Fig. 3a1), with sizes up to around 100 nm. A similar morphology can be also observed in the case of the SnO2– Ta2O5 and SnO2–IrO2 composites. The samples (Fig. 3b1 and b2 relative to SnO2–IrO2) are mainly composed by aggregates which are smaller than in the case of the pure SnO2 samples, with a size in the range 20–50 nm and an irregular shape. The ternary composites (SnO2–IrO2– Ta2O5) show, in their turn, a different aspect (Fig. 3c1 and c2): the samples are mainly composed by homodisperse, small, spherical particles with average sizes around 20 nm. The prevalent presence, in these samples, of smaller, less aggregated particles accounts well for the large surface area reported in Fig. 2b.
0.03
0.07
0.15
Ir molar fraction
Fig. 2. Dependence of: (a) crystallite sizes and (b) specific surface areas on the Ir molar content of the composites in the presence (empty circles) or absence (empty squares) of Ta dopant ion.
Ion
Charge
Coordination
˚ Ionic radius/A
Sn Ir Ta Ta
4+ 4+ 4+ 5+
VI VI VI VI
0.69 0.63 0.68 0.64
[26,27], [26,27], [26,27], [26,27],
0.83 0.71 0.74 0.72
[14,19,20] [5], 0.77[14,19] [5] [5]
S. Ardizzone et al. / Journal of Electroanalytical Chemistry 589 (2006) 160–166
a1
a2
20 nm
b1
b2
20 nm
c1
c2
20 nm
Fig. 3. SEM images of SnO2 (a1, a2); Sn0.85Ir0.15O2 (b1, b2); and Sn0.78Ir0.15Ta0.07O2.175 (c1, c2) nanoparticles.
3.2. XPS analyses Survey XPS spectra were recorded for all samples. No significant presence of impurities was observed, except for the ubiquitous carbon contaminant. In the case of the latter element, only the C 1s peak at 284.6 eV (due to –CH– species) was present. The chemical state of tin, tantalum, iridium and oxygen in the composite films and particles was examined by XPS. The Sn 3d region shows the regular doublet with peaks at 486.7 and 495.2 eV in agreement with literature data for tin oxides [28]. No significant differences could be appreciated in the binding energies of tin as an effect of the composite mixture. The peak of Ta 4f at 26.2 eV is hardly appreciable. Ta species are, apparently, impoverished at the surface. The Sn/Ta atomic ratio by XPS is about one order of magnitude higher with respect to the bulk one. The Ir 4f peak is well appreciable for iridium molar fractions > 0.03. The presence of Ta in the mixture promotes the enrichment of Ir at the surface. The promoting effect is the largest the lowest the Ir content in the mixture: 25% for Sn0.86Ir0.07Ta0.07O2.175 and 14% for Sn0.78Ir0.15Ta0.07O2.175. This effect can be related to the already discussed competition between Ta and Ir in the reticular substitution of Sn. Ta may displace Ir from the lattice and push it to the external layers of the particles. The Ir 4f peak is complex and shows the presence of more than one species. There is considerable disagreement in the literature about the nature of the components of the Ir 4f7/2 peak in the case of IrO2. Several authors
163
[17,29] attribute the main component to Ir(III) (61.6– 62.0 eV), and the second component at higher B.E. (62.3–62.8 eV) to Ir(IV). Consistently with this interpretation Da Silva et al. [17] report a majority presence of the Ir(III) oxide component for a commercial IrO2 sample while in the case of ternary IrO2 + TiO2 + PtOx composites observe only the occurrence of the Ir(IV), its surface concentration decreasing with the progressive Pt addition. Other authors, instead, attribute the same doublets respectively to Ir(IV) and to Ir in a higher oxidation state [30,31]. Peukert [32] has discussed the peaks at higher B.E. in terms of Ir(VI), and compositions of IrO2+x (x > 0) were proposed for IrO2 on the grounds of Rutherford backscattering analyses [33]. In the present work fittings of the Ir 4f7/2 peak were performed by using only Gaussian line shapes and without B.E. or FWHM constraints. The best fit of the peaks yields three components which are attributed respectively to Ir(III) at 61.7, to Ir(IV) at 62.6 and to Ir in an oxidation state higher than four at 63.6 eV. The component at the lowest B.E., i.e. Ir(III), is in any case the prevailing one (see Fig. 4a). No quantitative evaluations of the Ir 4f peak components are presented due to the relevant degree of arbitrariness inherent to calculations. The general shape of the Ir 4f peak does not undergo relevant modifications either upon modifying the composite composition or after polarization. The present results can be considered to compare well with data by Atanasoska et al. [34] suggesting, in the case of mixed RuO2/IrO2 layers, the presence of three different components in the XPS 4f region of IrO2. The oxygen 1s peak of the composites is, in its turn, complex and shows the presence of several components. In the case of pure iridium oxide the oxygen peak is generally fitted into three components, corresponding to three different oxygen species, i.e. lattice oxide, hydroxide, surface OH groups or undissociated water [30,31,35]. In the present case the situation is more complicated due to the presence of Sn (and to a minor extent of Ta) oxides or oxohydroxides. Fig. 4b reports the O 1s peak of a fresh, as prepared, composite sample. The best fit yields three components, which can be attributed respectively to lattice oxygen in SnO2 (529.9 eV, A component), hydroxide in Sn(OH)4 or lattice oxygen in IrO2 (530.7 eV, B component), OH groups in Ir(OH)4 or IrO(OH)2 plus possible surface OH species (531.9 eV, C component). An interesting aspect comes from the comparison between as prepared samples and samples submitted to polarization. Invariably, after the polarization (Fig. 4c), the oxygen peak of the composites shows larger B and C components and a pronounced tail at the high energy side (component D), which is not appreciable in the ‘‘dry’’ as prepared samples. The component at 532.9 eV can be attributed to tightly bound water in agreement with parallel data by Koetz et al. [35]. These features appear to be strictly connected with the activation of the Ti-supported electrodes after service in the oxygen evolution region (see the following).
164
S. Ardizzone et al. / Journal of Electroanalytical Chemistry 589 (2006) 160–166
Fig. 5. CV curves in NaClO4 0.1 M of: (A) Ti–SnO2; (B) Ti–Sn0.85Ir0.15O2; (C) Ti–Sn0.78Ir0.15Ta0.07O2.175 freshly prepared; (D) Ti–Sn0.78Ir0.15Ta0.07O2.175 after 4-h polarization in the oxygen evolution region; (E) as (D), recorded in HClO4 0.1 M. Scan rate: 20 mV s1. Empty arrows denote peak positions in curve C.
7
Fig. 4. XPS spectra of Ti supported composites (0.15 Ir molar fraction). (a) Ir 4f7/2,5/2 doublets relative to the different Ir spectral components, (b) oxygen 1s peak for a fresh ‘‘as prepared’’ sample and (c) oxygen 1s peak for the sample in (b) submitted to oxygen evolution reaction in acid media.
lg ( Rox / Ω cm2 mg-1)
6 5 4 3 2 1
a
0 10 9 8
Voltammetric investigations on Ti-supported composite electrodes were performed both in the charging (CV) and in the OE (slow potentiodynamic curves) regions. Capacitive and resistive features of the mixed oxides were obtained by EIS measurements. The results point to the progressive evolution from the semiconducting behaviour of the pure SnO2 phase [22,23] to the electronic conducting features of the Ir and Ir + Ta doped materials. This is evidenced in Figs. 5 and 6, where the data obtained for the three different oxide categories – namely pure tin oxide, mixed tin + iridium and tin + iridium + tantalum oxides at constant Ir content – are presented. The charge accumulation capabilities in NaClO4 0.1 M are shown in Fig. 5. The CVs are recorded in the 0.4– 1.4 V potential window, in the absence of both oxygen and hydrogen evolution reactions. Here the main process
Cdl / F cm-2 g-1
3.3. Electrochemical behaviour
7 6 5 4 3 2 1 0 0.2
0.4
0.6
b
0.8
1
1.2
1.4
1.6
E / V vs RHE
Fig. 6. Oxide resistance, Rox (a), and interfacial capacitance, Cdl (b), of Ti–SnO2 (triangles); Ti–Sn0.85Ir0.15O2 (squares); and Ti–Sn0.78Ir0.15Ta0.07O2.175 (lozenges). Parameters optimized with Zview Software (Scribner Associates Inc., Souther Pines, USA), by adopting the equivalent circuit described in [22,36].
is represented by the pseudo-capacitive protonation–deprotonation reaction: MOx ðOHÞy þ dHþ ðsolutionÞ þ de ðoxideÞ ¢ MOxd ðOHÞyþd
ð1Þ
S. Ardizzone et al. / Journal of Electroanalytical Chemistry 589 (2006) 160–166
where M represents any of the three elements, Sn, Ir and Ta. The current densities are normalized by weight of deposited composite material. Although not appreciable in Fig. 5, the Ti–SnO2 electrode presents the typical exponential I/E dependence chained to the emptying/filling of intragap monoelectronic levels [22]. This is confirmed by the dependence of the oxide resistance (Rox) (Fig. 6a) and the interfacial capacitance (Cdl) (Fig. 6b) from the applied potential. This feature is totally absent in the Sn + Ir mixed oxides, which exhibit potential independent electrical parameters, Rox and Cdl, with typical electronic conductor values. Parallelly, the almost rectangular shaped CVs, as reported for IrO2 contents higher than 15% mol [2], are characterized by current densities of up to two orders of magnitude higher than pure Ti–SnO2 electrodes. The increase in pseudo–capacitance is accompanied by electrocatalytic features for the OER in HClO4 0.1 M, with Tafel slopes of 47–48 mV and current densities between 0.1 and 2 mA cm2mg1 in the 1.5– 1.6 V potential range. This behaviour is intermediate to that observed by De Pauli and Trasatti [3] for tin iridium oxides at 10 and 20 IrO2 mol% and compares well with literature on IrO2 + Ta2O5 systems with IrO2 P 55% [7,13]. The addition of tantalum oxide to the SnO2–IrO2 system, to give the ternary Sn + Ir + Ta composite, further improves the observed properties. The mixed oxide resistance is the less affected feature (compare curves B and C in Fig. 6a), as it is governed by the iridium content. In fact, the increase in the particle conductivity, bound to the n-doping of Sn(IV) with Ta(V) [37], is expected to be lower than the one bound to Ir doping and possibly counterbalanced by the increase in the number of intergranular/interparticle connections chained to the reduction of the crystallite/particle size. Vice versa, the combined effect of reducing the particle size and lowering the phase crystallinity, specifically associated with the addition of Ta, produces an increase in defective surface sites which in turn increases the interfacial capacitance (compare curves B and C in Fig. 6b), thus doubling the capability of charge accumulation. This is also evident from the increased area of the CV characteristic (Fig. 5 curve C), which also presents two reversible humps located at 0.75–0.85 V and 0.97–1.07 V, respectively, frequently observed on high IrO2 content electrodes [2,38]. At the same time, Tafel lines show over one order of magnitude increase in current densities (at constant overvoltage) and a slight reduction of the slope to 44 mV, thus evidencing the further improvement of electrocatalytic activity. Note that, after 4-h polarization cycles in the OE region in HClO4 0.1 M, the CV curve has strikingly increased in terms of current density and hence of accumulated charge (Fig. 5 curve D), thus denoting an increase of accessible iridium sites. Although we cannot exclude that some of these effects might be connected with the activation of the surface by electrochemical treatment (removal of chemisorbed CO2, O2, etc.), the main features are consistent with the XPS data of un-cycled and cycled
165
electrodes, not only in term of surface composition, but also of Ir oxidation states and oxide hydration. In particular, the merging of the CV humps in the cycled Ti– SnIrTa electrodes points to an increase of the Ir/Ta ratio and to a further reduced Cdl/E dependence, while no significant alteration of the distribution of Ir(III), (IV) and (>IV) species seems to be connected with the electrochemical treatment, which is rather responsible of the increase of hydrated oxygen species, especially in the form of strongly chemisorbed water molecules. Interestingly, this is complementary to the behaviour observed by Koetz et al. [35] for anodic IrO2 films, which after polarization at increasing potentials evolve from the original fully hydroxylated state toward a less hydrated form with a strong Ir–O component. To our opinion this demonstrates that the hydration of the oxide layer is regulated by the OER, which drives the hydration/dehydration of the material toward the same final state. In our case, the rehydration of the oxide matrix after calcination has the beneficial effect of favouring the proton diffusion within the material and improving the reversibility of the intercalation process [39]. This is the more evident, the more acid the working conditions [38], as highlighted by the CV tests performed in HClO4 0.1 M (Fig. 5 curve E). Here two broad highly symmetric cathodic and anodic peaks centred between 0.7 and 0.9 V (RHE) confirm both the high surface content of Ir and the fairly hydrated state of the oxide layer, which, notwithstanding its fine physical porosity, becomes more easily penetrated by the current lines and exhibits a welcomed increase of the number of readily accessible electrocatalytic sites. 4. Conclusions Ternary SnO2 + IrO2 + Ta2O5 nanostructured composites and Ti-supported composite electrodes have been prepared by the same sol–gel route and characterized by ex-situ physico-chemical and in situ electrochemical techniques. Comparison with base SnO2 and binary SnO2 + IrO2 materials evidences the superior properties of the ternary mixtures and the key role of tantalum (even at low molar fraction) in expanding the surface area, improving the electronic conductance and the charge storage capacitance, promoting the surface enrichment of iridium. Preliminary results on Sn + Ir + Ta mixed oxide electrodes also highlight the excellent electrocatalytic properties for OER in acid electrolyte, even at low Ir content (15 mol%). The further striking improvement after a few hours of service life in the OE region seems to be connected with the partial re-hydration of the oxide matrix, almost completely dehydrated by the calcination step. The strongly chemisorbed water favours the proton exchange within the porous layer, a key process for both the charge storage and the oxygen evolution mechanisms. Future investigations will be aimed at further elucidating the Ta role and optimizing the composite preparation and composition.
166
S. Ardizzone et al. / Journal of Electroanalytical Chemistry 589 (2006) 160–166
Acknowledgements The financial contribution of The University of Milan (FIRST) and the EC support for a Marie Curie Fellowship within the programme Improving Human Research Potential and the Socio-Economic Knowledge Base – IHP, contract number HPMT-CT-2001-00314 are gratefully acknowledged. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]
S. Trasatti, Electrochim. Acta. 36 (1991) 225. C.P. De Pauli, S. Trasatti, J. Electroanal. Chem. 396 (1995) 161. C.P. De Pauli, S. Trasatti, J. Electroanal. Chem. 538 (2002) 145. A. Rossi, J.F.C. Boodts, J. Appl. Electrochem. 32 (2002) 735. J.M. Hu, J.Q. Zhang, H.M. Meng, C.N. Cao, J. Mater. Sci. 38 (2003) 705. J.M. Hu, H.M. Meng, J.Q. Zhang, C.N. Cao, Corr. Sci. 44 (2002) 1655. J.M. Hu, J.Q. Zhang, C.N. Cao, Int. J. Hydrog. Energy 29 (2004) 791. L.K. Xu, J.D. Scantlebury, J. Electrochem. Soc. 150 (2003) B254. L.K. Xu, J.D. Scantlebury, J. Electrochem. Soc. 150 (2003) B288. M. Morimitsu, R. Otogawa, M. Matsunaga, Electrochim. Acta 46 (2000) 401. E. Rasten, PhD thesis, NTNU Trondheim, Norway, 2001. C. Comninellis, P. Vercesi, J. Appl. Electrochem. 21 (1991) 139. L.M. Da Silva, D.V. Franco, L.A. De Faria, J.F.C. Boodts, Electrochim. Acta 49 (2004) 3977. X. Chen, G. Chen, P.L. Yue, J. Phys. Chem. B 105 (2001) 4623. G. Chen, X. Chen, P.L. Yue, J. Phys. Chem. B 106 (2002) 4364. F.I. Mattos-Costa, P. de Lima-Neto, S.A.S. Machado, L.A. Avaca, Electrochim. Acta 44 (1998) 1515. L.A. da Silva, V.A. Alves, S.C. de Castro, J.F.C. Boodts, Colloid. Surf. A: Physicochem. Eng. Aspects 170 (2000) 119. A.A.F. Grupioni, E. Arashiro, T.A.F. Lassali, Electrochim. Acta 48 (2002) 407.
[19] A. Marshall, B. Børresen, G. Hagen, M. Tsypkin, R. Tunold, Mater. Chem. Phys. 94 (2005) 226. [20] F. Montilla, E. Marallo´n, A. De Battisti, A. Benedetti, H. Yamashita, J.L. Va´zquez, J. Phys. Chem. B 108 (2004) 5044. [21] T.A.F. Lassali, J.F.C. Boodts, L.O.S. Bulho˜es, Electrochim. Acta 44 (1999) 4203. [22] S. Ardizzone, G. Cappelletti, M. Ionita, A. Minguzzi, S. Rondinini, A. Vertova, Electrochim. Acta 50 (2005) 4419. [23] M. Ionita, G. Cappelletti, A. Minguzzi, S. Ardizzone, C. Bianchi, S. Rondinini, A. Vertova, J. Nanop. Res., in press. [24] T. Boiadjieva, G. Cappelletti, S. Ardizzone, S. Rondinini, A. Vertova, Phys. Chem. Chem. Phys. 5 (2003) 1689. [25] T. Boiadjieva, G. Cappelletti, S. Ardizzone, S. Rondinini, A. Vertova, Phys. Chem. Chem. Phys. 6 (2004) 3535. [26] R.D. Shannon, Acta Cryst. A32 (1976) 751. [27] J.E. Huheey, E.A. Keiter, R.L. Keiter, Inorganic Chemistry: Principles of Structure and Reactivity, fourth ed., HarperCollins, New York, USA, 1993. [28] J.F. Moulder, W.F. Stickle, K.D. Bomben, Handbook of X-ray Photoelectron Spectroscopy, Perkin Elmer, Eden Praie, 1992. [29] M. Hara, K. Assami, K. Hashimoto, T. Matsumoto, Electrochim. Acta 28 (1983) 1073. [30] R.S. Chen, H.M. Chang, Y.S. Huang, D.S. Tsai, S. Chattopadhyay, K.H. Chen, J. Cryst. Growth 271 (2004) 105. [31] S. Yao, M. Wang, M. Madou, J. Electrochem. Soc. 148 (2001) H29. [32] M. Peuckert, Surface Sci. 144 (1984) 451. [33] C. Angelinetta, S. Trasatti, L. Atanasoska, R. Atanasoski, Z. Miesvski, Mater. Chem. Phys. 22 (1989) 231. [34] L. Atanasoska, R. Atanasoski, S. Trasatti, Vacuum 40 (1990) 91. [35] R. Koetz, H. Neff, S. Stucki, J. Electrochem. Soc. 131 (1984) 72. [36] J. Bisquert, G. Garcia-Belmonte, F. Fabregat-Santiago, A. Compte, Electrochem. Commun. 1 (1999) 429. [37] C. Wang, J. Wang, H. Chen, W. Wang, W. Su, G. Zang, P. Qi, J. Phys. D: Appl. Phys. 36 (2003) 3069. [38] S. Ardizzone, A. Carugati, S. Trasatti, J. Electroanal. Chem. 126 (1981) 287. [39] P.S. Patil, R.K. Kawar, S.B. Sadale, Appl. Surf. Sci. 249 (2005) 367.
