Electrochemical and Corrosion Characterization of TiO₂-RuO₂/Ti Electrodes Modified with WO₃

Abstract

DSA (dimensionally stable anodes)-type electrodes (TiO₂-RuO₂-WO₃/Ti) were prepared by the thermal decomposition of proper metal precursor on titanium substrate. Electrochemical and corrosion characteristics of these electrodes were determined in [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ system and Na₂SO₄ solution with the application of cyclic voltammetry and chronoamperometry. The electrode active surface area and the activity towards oxygen evolution were evaluated. Moreover, corrosion resistance of the tested electrodes was estimated by determination of basic corrosion parameters. The introduction of WO₃ into the oxide layer of the electrodes resulted in a decrease in electrode active surface area and in the number of active sites. The modified electrodes revealed lower activity towards OER but higher activity in organic oxidation. However, their stability was significantly higher in comparison with the non-modified electrode. Corrosion characteristics of the tested electrodes proved the highest corrosion resistance in the case of the electrode modified with 3% WO₃. The modification of TiO₂-RuO₂/Ti with WO₃ seems to be advantageous in the application in electrochemical and photoelectrochemical degradation of organic pollutants.

Introduction

Last years, electrochemical advanced oxidation processes (EAOPs) have been regarded as promising technologies in many different fields, e.g., degradation of organic compounds, due to their low negative environmental impacts with comparison to commonly applied processes. Especially, EAOPs applied in wastewater treatment draw attention since they ensure high efficiency and low resource consumption with simultaneous use of electrons as a “clean reagent” [1]. These techniques can be also applied in treatment of wastewater containing dyes used in textile industry.

TiO₂-RuO₂/Ti electrodes are typical dimensionally stable anodes (DSA) which are commonly applied in the oxidation of organic compounds which occurs with simultaneous evolution of oxygen [2,3,4]. DSA-type electrodes can be prepared with the application of different methods, e.g., thermal decomposition of polymeric precursors or thermal decomposition of precursor metal salts [5,6,7,8,9]. The method of preparation has an effect on the morphological and electrochemical characterization of electrodes. The abovementioned electrodes properly modified reveal the ability to produce reactive species like HO^•, RO^•, and ROO^• radicals as well as O²⁻ species in situ [10, 11] or other reactive species like SO4^-•, Cl^• radicals [12], and O₃, H₂O₂ and Cl₂ formed in an electrolytic process. These species take part in mineralization of organic pollutants but their formation strongly depends on properties of the electrode materials [10].

Titanium-supported electroactive oxide layers in DSAs usually consist of an active transition metal oxide (RuO₂, IrO₂, Co₃O₄, etc.) stabilized by a valve metal oxide (TiO₂, RuO₂, SnO₂, Ta₂O₅, etc.) [13]. The purpose of modification of an oxide layer by the addition of the third metal oxide is to modify electrochemical properties of the active layer or/and increase electrode durability. Moreover, an improvement of the electrochemical degradation of organic compounds can be achieved by its combination with a photocatalytic degradation. The photoelectrochemical process can be performed with the application of TiO₂-RuO₂/Ti electrodes. Photoexcitation of these electrodes requires irradiation from UV region due to TiO₂ wide bandgap (3.2 eV) [14]. Thus, in our previous paper [15], we suggested a modification of TiO₂-RuO₂/Ti electrodes by introduction of WO₃ to the oxide layer. The modification enabled also photoexcitation of the electrodes with irradiation from VIS region due to WO₃ bandgap of 2.8 eV. Tungsten oxide is an n-type semiconductor and is often coupled with TiO₂ [16, 17]. We proved that the electrodes Ti_0.7Ru_0.3/Ti and Ti_0.97Ru_0.03/Ti modified with WO₃ could be successfully applied in photoelectrochemical degradation of an azo dye Acid Orange 7 not only under UV but also under VIS irradiation [15]. The highest mineralization efficiency was achieved in the case of the electrodes modified with 6% WO₃. Thus, it seems to be interesting comparing the electrochemical and corrosion characterization of the modified electrodes, especially their electrochemically active surface area and stability.

The corrosion behavior of DSAs and also the kinetics of oxygen and chlorine evolution on them depend on the chemical and phase composition of the oxide layer as well as on their structure and morphology. It was proved that modification of an oxide layer by introduction of IrO₂ and Ta₂O₅ resulted in higher catalytic activity in the oxygen evolution reaction and higher corrosion resistance [18,19,20]. The higher catalytic activity can be explained by the development of the anode surface and/or by an increase in the number of active centers [18]. Although the surface does not always correlate with its electrocatalytic activity, it provides useful information on designing active DSAs with practical industrial applicability [21].

The determination of the real surface of the solid electrodes can be performed with various methods including electrochemical methods based on mass transfer processes under diffusion control, adsorption processes at the electrode surface, and measurements of the differential capacitance in the electric double-layer region [22]. The obtained results depend on the applied method and conditions of measurements.

The aim of this paper was to investigate the effect of different contents of WO₃ introduced to the oxide layer of Ti_0.7Ru_0.3O₂/Ti electrodes on electrochemical and corrosion behavior as well on the stability of these electrodes. Electrochemical characteristics of the electrodes were performed in [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ system and in the supporting electrolyte. Moreover, corrosion behavior and stability of the electrodes were investigated with the application of the electrochemical method (potentiodynamic polarization). The results of the experiments carried out for Ti_0.7-xRu_0.3-W_xO₃/Ti (x = 0.03 and 0.06) are described and compared with those of the non-modified electrode.

Experimental

Chemicals

Potassium hexacyanoferrate(II) trihydrate (K₄[Fe(CN)₆]·3H₂O, MERCK) was used in the preparation of solutions applied in electrochemical measurements. This chemical was dissolved in 0.1 mol L⁻¹ KCl and its concentration was 5 · 10⁻³ mol L⁻¹. The measurements were also carried out in the solution of Na₂SO₄ at the concentration of 0.1 mol L⁻¹ due to the fact that it was supporting electrolyte in our previous investigations [15]. All solutions were prepared with double distilled water. Purity of the chemicals was analytical grade.

Electrodes

Electrodes with nominal composition Ti_0.7-xO₂-Ru_0.3O₂-W_xO₃/Ti (x = 0, 0.03, 0.06) were prepared by thermal decomposition of chloride precursor mixtures in n-butanol. The concentrations of precursors were properly adjusted in order to obtain coatings with nominal compositions. The solution of precursors was spread several times by brushing on both sides of the Ti substrate. After each brushing, the electrodes were dried at 60 °C for 10 min and next annealed at 450 °C for 10 min. The procedure was repeated several times until the appropriate oxide film thickness was obtained. The oxide loading was 2.1 mgcm^-2. Finally, the coated electrodes were annealed at 500 °C for 3 h in air.

The composition of the tested electrodes was established taking into consideration the optimum WO₃ content in bi-component WO₃/TiO₂ photocatalysts. The content of RuO₂ was constant in order to ensure electrocatalytic activity and electrical conductivity of the tested electrodes. In order to compare the results of experiments, non-modified electrode, i.e., Ti_0.7Ru_0.3O₂/Ti, with the same geometric area was also applied in electroanalytical measurements. The geometric surface area of all tested electrodes was 2 cm². The counter electrode was in the form of Pt wire with the geometric surface area of 6.31 cm². Saturated calomel electrode was applied in all experiments as a reference electrode. The morphology of the tested electrodes was observed using S-4700 Hitachi (Japan) scanning electron microscope. Images were recorded at a magnification of × 5000.

Electrochemical Measurements

Electrochemical characterization of the tested electrodes was performed with the cyclic voltammetry method in K₄[Fe(CN)₆] and Na₂SO₄ solutions. Cyclic voltammograms were recorded at different scan rates in the range from 2 to 500 mV s⁻¹. Chronoamperometry was another method used in the electrochemical characterization of the tested electrodes. Chronoamperograms were recorded in the solution of K₄[Fe(CN)₆] for the electrooxidation at the potential at which only one electrode reaction occurs. The volume of the solutions in the electrochemical measurements was 20 mL. All solutions were purged with argon for at least 20 min before measurements in order to remove dissolved oxygen. During measurements, argon blanket was kept over the solution. For each of the tested electrodes, measurements were performed thrice.

Corrosion characterization of the tested electrodes was performed in Na₂SO₄ solution and was evaluated using electrochemical techniques such as open circuit potential (OCP) followed by potentiodynamic polarization sweep. After the working electrode was immersed, its potential was recorded as a function of time in order to determine the OCP. The OCP was measured by 1 h or unless the OCP value was established when dE/dt ≤ 1 μV s⁻¹ achieved. After the equilibrium was established, a tested electrode was cathodically and anodically polarized in the potential range of OCP ± 200 mV with the scan rate of 2 mV s⁻¹.

All electrochemical measurements were performed using the three-electrode cell connected to AUTOLAB electrochemical workstation—μAutolab III (Metrohm Autolab B.V., The Netherlands). Nova software vs. 2.1 was used in the analysis of recorded chronoamperograms and voltammetric curves and in the determination of voltammetric charges.

In order to assess the electrode stability, accelerated stability tests (AST) were performed in a three-electrode cell at room temperature. A potentiostat/galvanostat Autolab PGSTAT 302N (Metrohm Autolab B.V) was used to provide a constant anodic current density of 200 mA cm⁻². The electrolyte was 0.1 mol L⁻¹ Na₂SO₄. The electrolysis time at which the anodic potential reached the value at least 2 V higher than the value at time zero was considered as the lifetime of the tested electrode.

Results and Discussion

Electrochemical Characterization of the Electrodes in [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ System

In order to investigate the electrochemical properties of the tested electrodes, cyclic voltammograms were recorded in the solution of K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) at the tested electrodes. [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ redox system was selected as a “surface-sensitive” electrochemical response for various electrodes [23,24,25]. Comparison of the exemplary curves recorded at the scan rate of 5 mV s⁻¹ for the tested electrodes is presented in Fig. 1. Electrochemical parameters characterizing the redox system are listed in Table 1.

Fig. 1

Cyclic voltammograms recorded at the tested electrodes in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) at the scan rate of 5 mV s⁻¹

Table 1 Electrochemical parameters (I_pa and I_pc—anodic and cathodic peak current, E_pa and E_pc—anodic and cathodic peak potential, ΔE_p—difference between anodic an cathodic peak potential, E_1/2—half-wave potential) determined from cyclic voltammograms recorded in [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ redox system

The introduction of WO₃ to the oxide layer in the electrode surface results in a decrease of the anodic and cathodic peak currents (the highest—for the electrode with 6% WO₃). Simultaneously, the anodic and cathodic current ratio (I_pa/I_pc) is very close to 1 in the case of all tested electrodes and confirms the reversible character of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ redox system. The half-wave potential (E_1/2) is almost comparable for all tested electrodes with a very slight increase from 181 to 184 mV while the content of WO₃ in the oxide layer increases to 6%. However, the parameter ΔE_p (peak-to-peak separation) is by 9–11 mV higher in the case of the electrodes modified with WO₃. The obtained results prove that the electrochemical oxidation and reduction at the modified electrodes is slightly slower and less reversible in comparison with the non-modified electrode.

An effect of WO₃ on the electrode active surface (EAS) was determined. For the reversible reaction at 25 °C, the electroactive surface area can be calculated from the Randles-Sevcik equation [26]:

$$ {I}_{\mathrm{p}}=2.69\times {10}^5{n}^{3/2}{AD}^{1/2}{Cv}^{1/2} $$

(1)

where A is the electroactive surface area, I_p is the peak current, D is the diffusion coefficient of the analyte, n is the number of transferred electrons, v is the scan rate, and C is the concentration of the redox molecules in a solution, providing that diffusion coefficients of the oxidized and reduced species are known. This equation can be applied in the case of electrochemical processes which are controlled by diffusion. Thus, cyclic voltammograms were recorded at various scan rates at all tested electrodes (Fig. 2). The dependences of peak currents on the square root of the scan rate (Fig. 3) are linear and indicate that the electrochemical reaction occurring at the electrodes is a diffusion-controlled process in the scan rate from 2 to 200 mV s⁻¹ for the anodic and cathodic peaks. In order to confirm the diffusion control of the process in this scan rate range, a dependence of logI_p vs. logv was determined for anodic and cathodic peaks. In the case of all tested electrodes, these dependences are linear and described by the equations presented in Table 2. An experimental slope of 0.45–0.47 for the anodic peaks and 0.46–0.47 for the cathodic peaks is close to the value of 0.5 indicating that the electrochemical oxidation and reduction of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ redox couple is controlled by diffusion [26, 27].

Fig. 2

Exemplary cyclic voltammograms recorded at 67%TiO₂–30%RuO₂–3%WO₃/Ti electrode in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) at the scan rates in the range from 2 to 500 mV s⁻¹

Fig. 3

A plot of I_pa and I_pc against v^1/2 for the tested electrodes determined in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) at the scan rates from 2 to 200 mV s⁻¹

Table 2 Dependences of logI_p vs. logv determined for the tested electrodes in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl)

EAS of the tested electrodes was calculated from the slope of the anodic peak current vs. the square root of the scan rate (v^1/2) and for comparison from the slope of the cathodic peak current vs. v^1/2 (Table 3). The values of D_ox and D_red were 7.63 · 10⁻⁶ and 6.50 · 10⁻⁶ cm² s⁻¹, respectively [28, 29]. The values of EAS calculated from I_pa are slightly higher than those calculated from I_pc. Due to the fact that the cathodic peak splits into two peaks for scan rates higher than 50 mV s⁻¹, the EAS values calculated from I_pa seem to be more accurate. Results presented in Table 3 clearly show that introduction of WO₃ to the oxide layer in the electrode surface results in a decrease in the electroactive surface area. In the case of the modified electrode with 6% WO₃, EAS is slightly higher than its geometric area. Moreover, the roughness factor (ρ) was calculated for the tested electrodes according to the following equation [22, 30]:

$$ \rho =\frac{\mathrm{EAS}}{A_{\mathrm{geom}.}} $$

(2)

where A_geom. denotes the geometric surface area of an electrode. The tested electrodes reveal relatively low roughness factor with the highest value of 1.32 for the non-modified electrode. An increase of WO₃ content in the oxide layer to 3 and 6% results in the roughness factor of 1.18 and 1.06, respectively.

Table 3 Electroactive surface area (EAS) of the tested electrodes calculated from the dependence of I_pa and I_pc vs. v^1/2 determined in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl)

In order to confirm the obtained results, the electroactive surface area was also determined with the application of chronoamperometry method. EAS values were calculated from the Cottrell equation [23, 27, 31]:

$$ I=\frac{nFA{D}^{1/2}C}{\pi^{1/2}{t}^{1/2}} $$

(3)

where I is the current intensity and other parameters have their usual meanings. Chronoamperograms were recorded at the tested electrodes in the solution of K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) for the electrooxidation reaction at the potential at which only one electrode reaction occurs (0.25 V). Exemplary chronoamperogram is presented in Fig. 4 and the calculated values of the electroactive surface area and roughness factor for the tested electrodes are shown in Table 4. The values of EAS calculated from the chronoamperograms are slightly higher (by about 5%) than those calculated from cyclic voltammograms. However, these results confirm that the introduction of WO₃ to the oxide layer causes a decrease in the electroactive surface area by 0.25 cm²—approximately 10% per each 3% WO₃. Moreover, the roughness factor is by about 5% higher than in the case of calculations made in the previous method. This factor decreases with the introduction of WO₃ into the oxide layer by about 10% per each 3% WO₃. Similar results obtained in both methods indicate that [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ system can be applied in the evaluation of electrode active surface area in the case of TiO₂-RuO₂/Ti electrodes modified with WO₃.

Fig. 4

Chronoamperogram recorded at 64%TiO₂–30%RuO₂–6%WO₃/Ti in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) at the potential of 0.25 V vs. SCE. Inset, Cottrell’s plot of I vs. t^-1/2

Table 4 Electroactive surface area (EAS) and roughness factor (ρ) calculated for the tested electrodes from the chronoamperograms recorded at the tested electrodes in K₄[Fe(CN)₆] (5 · 10⁻³ mol L⁻¹ in 0.1 mol L⁻¹ KCl) at the potential of 0.25 V vs. SCE

The results of EAS calculations obtained for the tested electrodes can be compared with the results of electrochemical degradation of an azo dye performed on these electrodes described in the previous paper [15]. According to the previous results, introduction of WO₃ to the oxide layer of TiO₂-RuO₂/Ti electrode with RuO₂ content of 30% resulted in a decrease of the dye oxidation current observed in recorded cyclic voltammograms. The higher WO₃ content resulted in a higher decrease in the electrooxidation current, i.e., by 26% and 40% for 3 and 6% WO₃, respectively. This can be explained by a decrease in electrode active surface. Introduction of WO₃ resulted in 10% lower EAS and 10% lower roughness factor per each 3% WO₃ and twice higher decrease in peak current observed in the dye electrooxidation in comparison with the non-modified electrode.

Electrochemical Characterization of the Electrodes in Na₂SO₄ Solution

Due to the fact that electrochemical and photoelectrochemical degradation of organic compounds, especially dyes, requires the presence of the supporting electrolyte, the tested electrodes were also characterized electrochemically in the solution of Na₂SO₄. The cyclic voltammograms were recorded at the tested electrodes in the potential range from − 0.4 to 1.0 V vs. SCE, i.e., in the potential range between hydrogen and oxygen evolution, with different scan rates. The voltammetric charge derived from these voltammograms was proven to be proportional to the electroactive surface area (EAS) and corresponds to electrochemically active sites on the surface [32, 33]. Exemplary voltammograms recorded at the scan rate of 50 mV s⁻¹ are presented in Fig. 5.

Fig. 5

Cyclic voltammograms recorded at the tested electrodes in 0.1 mol L⁻¹ Na₂SO₄ electrolyte with the scan rate of 50 mV s⁻¹

The cyclic voltammograms presented in Fig. 5 have the rectangular shape characteristic for pseudo-capacitive behavior. The pseudo-capacitance may be related to transitions of Ru(II)/Ru(III), Ru(III)/Ru(IV), and Ru(IV)/Ru(VI). Moreover, this shape is also characteristic of RuO₂-based electroactive coatings [34, 35]. Cyclic voltammograms (Fig. 5) exhibit one pair of anodic and cathodic peaks which probably correspond to the redox transition between Ru(III) and Ru(IV). Although the proton concentration is lower than in acid solutions, the pseudo-capacitive behavior of ruthenium oxide can be attributed to solid state surface redox transitions (SSSRT) of ruthenium species according to the following reaction [34, 36]:

$$ {\mathrm{RuO}}_{\mathrm{x}}+{\updelta \mathrm{H}}^{+}+{\updelta \mathrm{e}}^{-}\iff {\mathrm{RuO}}_{\mathrm{x}-\updelta}{\left(\mathrm{OH}\right)}_{\updelta}\ \left(0\le \updelta \le \mathrm{x}\right). $$

(4)

The transitions proceed over the whole potential range of the electrolyte stability. Introduction of WO₃ to the oxide layer did not cause a change in voltammograms shape but the current values observed at the electrode modified with WO₃ are clearly lower. However, an increase in WO₃ content from 3 to 6% causes an increase in the observed current. This can be explained by the change of the number of active surface sites contributing to the SSSRT [37]. Thus, the cyclic voltammograms were recorded in Na₂SO₄ at scan rates in the range from 5 to 500 mV s⁻¹. Voltammetric charge determined in a potential range where gas is not evolved can be used to estimate electroactive surface area [33]. The total charge q* is the sum of anodic and cathodic charge calculated by the integration of anodic and cathodic part of cyclic voltammograms and is dependent on the scan rate. While the scan rate increases, the charge decreases to the constant value [36] due to the existence of less accessible surface sites which gradually do not participate in the electrode reaction. The total charge consists of two components: “inner” and “outer” charge. The outer charge q*_out is attributed to outer parts of the electrode surface and thus more accessible. The inner charge q*_in is related to inner parts of the electrode surface and less accessible, like pores, cracks, and grain boundaries [38, 39]. In the range of very low sweep rates, q*_in takes part in a reaction with an electrolyte. The charges q*_tot and the q*_out can be calculated according to the following equations [36]:

$$ \frac{1}{q^{\ast }}=\frac{1}{q_{\mathrm{tot}}^{\ast }}+A\sqrt{v} $$

(5)

$$ {q}^{\ast }={q}_{\mathrm{out}}^{\ast }+B\frac{1}{\sqrt{v}} $$

(6)

where q* is the integrated charge and v is the scan rate. The q*_tot value can be estimated by plotting a dependences of 1/q* vs. v^1/2 followed by its extrapolation to 0 (v → 0), while the q*_out value can be obtained from a dependence of q* vs. v^-1/2 and followed by its extrapolation to 0 (v → ∞). The inner charge q*_in can be calculated from the equation:

$$ {q}_{\mathrm{tot}}^{\ast }={q}_{\mathrm{out}}^{\ast }+{q}_{\mathrm{in}}^{\ast } $$

(7)

The dependences obtained for tested electrodes in 0.1 mol L⁻¹ Na₂SO₄ solution are presented in Figs. 6 and 7. The results of total, inner, and outer charge calculations are shown in Table 5.

Fig. 6

Reversibility of the pseudo-capacitance process for the tested electrodes

Fig. 7

Dependences of voltammetric charges on the scan rate determined for the tested electrodes in 0.1 mol L⁻¹ Na₂SO₄; A Extrapolation of q*_tot. B Extrapolation of q*_out

Table 5 Voltammetric charges calculated for the tested electrodes from cyclic voltammograms recorded in 0.1 mol L⁻¹ Na₂SO₄ at various scan rates

The ratio of anodic (q*_a) and cathodic (q*_c) charge indicates the reversibility of the redox process. In the case of the tested electrodes, q*_a/q*_c value is slightly higher than 1 (Fig. 6) and indicates almost reversible behavior of the SSSRT.

Figure 7 and Table 5 show that introduction of WO₃ to the oxide layer of the electrode decreases the number of active surface sites almost twice. The electrode with 6% WO₃ indicates slightly higher active surface area than the electrode with 3% WO₃. On the other hand, the modified electrodes exhibit higher values of q*_out/q*_tot ratio which indicate that most of the active sites are accessible. That means that the non-modified electrode has 78% of the voltammetric charge associated with the outer and more easily accessible active sites, while 22% of the voltammetric charge is due to the inner and less easily accessible active sites in pores. In the case of the modified electrodes, 93–94% of the voltammetric charge is related to the outer active sites while the remaining active sites are located in pores. It means that more than 90% of active sites are located at the outer surface of the electrode and can take part in the electrode process. Almost the same values of q*_out/q*_tot and q*_in/q*_tot determined for the electrodes modified with WO₃ prove their similar surface morphology. The morphology of the tested electrodes can be seen in SEM images presented in Fig. 8. The non-modified electrode, i.e., 70%TiO₂–30%RuO₂/Ti, shows a typical porous “cracked mud” structure with flat areas. The cracks are probably formed during the stage of solvent evaporation. The cracks are not observed in the case of the electrodes modified with WO₃. The surface of both electrodes, i.e., 67%TiO₂–30%RuO₂–3%WO₃/Ti and 64%TiO₂–30%RuO₂–6%WO₃/Ti, reveals the appearance of superficial agglomerates which tends to connect and form networks also with flat areas. This can be a reason for a decrease in the electrochemically active surface area. The morphology of these two electrodes is similar.

Fig. 8

SEM images of the electrodes; a 70%TiO₂–30%RuO₂/Ti. b 67%TiO₂–30%RuO₂–3%WO₃/Ti. c 64%TiO₂–30%RuO₂–6%WO₃/Ti

The porosity of the electrode surface area can be also estimated by the ratio of q*_in/q*_tot [13]. In the case of the electrodes modified with WO₃, the porosity factor is almost three times lower than the porosity factor of the non-modified electrode (Table 5). Introduction of higher WO₃ amount (6%) to the oxide layer did not result in the porosity change.

The obtained results clearly show that modification of RuO₂-TiO₂/Ti electrode with WO₃ has a significant effect on the electroactive surface area. The number of active sites is much lower for the modified electrodes, but they are mainly (93–94%) located at the outer surface of the electrodes which is more accessible than active sites located in pores. According to the previous results [15], the electrooxidation currents of the azo dye observed at voltammograms recorded at the tested electrodes were also lower than at the non-modified electrode. This can be explained by clearly lower number of active sites at the surface of the modified electrodes. However, the electrochemical degradation of the azo dye was more efficient at the modified electrodes taking into consideration not only discoloration but also demineralization of the dye solution. Thus, it can be assumed that indirect electrooxidation proceeds with higher efficiency at the modified electrodes than at the non-modified one.

Activity of the Electrodes towards OER

The oxygen evolution reaction (OER) on the electrode surface is important in the electrochemical degradation of organic compounds due to the formation of hydroxyl radicals as well as in electrode electrochemical reactivity. As it is described in the paper [40], oxidation of organic pollutants by electrochemically generated hydroxyl radicals is in competition with the reaction of these radicals’ anodic discharge to oxygen. The activity of these radicals strongly depends on their interactions with the electrode surface. In reactivity comparison of electrogenerated hydroxyl radicals at various anodes, one general rule is very important. The weaker interaction of hydroxyl radicals with anode surface means the lower electrochemical activity towards oxygen evolution. This results in higher chemical reactivity towards organics oxidation.

The chemical reaction of organic pollutants with electrogenerated hydroxyl radicals, i.e., physically adsorbed “active oxygen,” is related to indirect anodic oxidation which is preceded by water discharge at the electrode (M) according to the following reactions [41, 42]:

$$ {\mathrm{H}}_2\mathrm{O}+\mathrm{M}\to \mathrm{M}\left({}^{\bullet}\mathrm{OH}\right)+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$

(8)

$$ \mathrm{M}\left({}^{\bullet}\mathrm{OH}\right)\to \mathrm{M}+\frac{1}{2}{\mathrm{O}}_2+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$

(9)

The reaction of the anodic discharge of hydroxyl radicals and evolution of oxygen (9) is competitive to the indirect anodic oxidation of R which can be presented by the following reaction if anode surface weakly interacts with ^•OH [41]:

$$ \mathrm{R}+\mathrm{aM}\left({}^{\bullet}\mathrm{OH}\right)\to \mathrm{aM}+{\mathrm{mCO}}_2+\mathrm{n}\ {\mathrm{H}}_2\mathrm{O}+{\mathrm{xH}}^{+}+{\mathrm{ye}}^{-} $$

(10)

where R denotes an organic compound with m carbon atoms which requires a = (2m + n) oxygen atoms in total mineralization.

Thus, OER was investigated on the tested electrodes. The anodic polarization curves were recorded in 0.1 mol L⁻¹ Na₂SO₄ for the tested electrodes in the potential range to 1.4 V in order to avoid the O₂ bubbles disturbing the signal measurement and moreover, intense evolution of oxygen can result in a decrease in the real surface of the electrode because bubbles can temporarily stick to electrocatalyst surface or clog pores. Figure 9 shows the polarization curves recorded with a scan rate of 5 mV s⁻¹ at the tested electrodes. The current density was related to the electrochemically active surface of the electrodes (Table 3) determined in K₄[Fe(CN)₆] solution from the anodic peak currents recorded in cyclic voltammograms.

Fig. 9

Polarization curves (A) and Tafel curves (B) recorded at the tested electrodes in 0.1 mol L⁻¹ Na₂SO₄ during the oxygen evolution reaction; scan rate of 5 mV s⁻¹

Exchange current density and Tafel slopes determined for the electrode reaction can be used as two primary factors which describe the rate of charge transfer reaction. The determination of exchange current density at specified potential can be applied in the electrocatalytic properties comparison [36, 43]. Tafel slopes are related to the reaction mechanism and its change indicates changes in electrocatalytic effects. A good electrocatalyst for OER should be characterized by a high exchange current density and a low Tafel slope [43]. Figure 9 shows the differences in exchange current density related to OER. The potential at a given current density can be used in description of the activity of the electrode material. Comparing the potential determined at the current density of 1.3 mA cm⁻², the tested electrodes show a decrease in the activity in the following order: 70%TiO₂–30%RuO₂/Ti (1.339 V) > 64%TiO₂–30%RuO₂–6%WO₃/Ti (1.365 V) > 67%TiO₂–30%RuO₂–3%WO₃/Ti (1.395 V). On the other hand, Tafel plots are similar for all tested electrodes indicating the same reaction mechanism for OER. However, all polarization curves show deviations from Tafel line at higher current densities. This can suggest a change in Tafel slope but also can be related to uncompensated ohmic drop. In order to eliminate one of these two reasons, the experimental data were corrected according to ohmic drop. Uncompensated ohmic resistance drops (IR) for polarization curves were estimated according to the method described in the papers [38, 44, 45]. The influence of the ohmic drop on a polarization curve can be described by the following equation [44]:

$$ E- IR=a+b\log I $$

(11)

where a and b are Tafel constant and slope, respectively, and R is the uncompensated resistance. Differentiation of the above equation with respect to current results in the following equation:

$$ \frac{dE}{dI}=\frac{b}{I}+R $$

(12)

The plotting of ΔE/ΔI reckoned between two experimental points against 1/I, where I is the mean value of current in the same interval, should result in a straight line with a slope b and the intercept R. The calculated R was used to correct the experimental polarization curve. The exemplary polarization curves with and without IR correction are presented in Fig. 10.

Fig. 10

Tafel curves recorded at 64%TiO₂–30%RuO₂–6%WO₃/Ti electrode with and without IR correction in 0.1 mol L⁻¹ Na₂SO₄; scan rate of 5 mV s⁻¹

IR-compensated curves recorded for all tested electrodes show only one slope (Fig. 10). In acidic and alkaline media, the mechanism for OER on oxide catalysts can be described by the following reactions [45, 46]:

$$ {\displaystyle \begin{array}{ll}\mathrm{acidic}\kern0.24em \mathrm{medium}:& \mathrm{S}+{\mathrm{H}}_2\mathrm{O}\to \mathrm{S}-{\mathrm{OH}}_{\mathrm{ads}}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\\ {}\mathrm{alkaline}\kern0.24em \mathrm{medium}:& \mathrm{S}+{\mathrm{OH}}^{-}\to \mathrm{S}-{\mathrm{OH}}_{\mathrm{ads}}+{\mathrm{e}}^{-}\end{array}} $$

(13)

$$ {\displaystyle \begin{array}{ll}\mathrm{acidic}\kern0.24em \mathrm{medium}:& \mathrm{S}-{\mathrm{O}\mathrm{H}}_{\mathrm{ads}}\to \mathrm{S}-{\mathrm{O}}_{\mathrm{ads}}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-}\\ {}\mathrm{alkaline}\kern0.24em \mathrm{medium}:& \mathrm{S}-{\mathrm{O}\mathrm{H}}_{\mathrm{ads}}+{\mathrm{O}\mathrm{H}}^{-}\to \mathrm{S}-{\mathrm{O}}_{\mathrm{ads}}+{\mathrm{H}}_2\mathrm{O}+{\mathrm{e}}^{-}\end{array}} $$

(14)

$$ \mathrm{S}\hbox{--} {\mathrm{O}}_{\mathrm{ads}}\to \mathrm{S}+1/{2\mathrm{O}}_2 $$

(15)

where S is the surface active site. The first two steps (Reactions 13 and 14) are associated with electron transfer in both media. The third step (Reaction 15) is the same and involves the oxygen evolution reaction (13). Both mechanisms predict Tafel slopes of 120 mV dec⁻¹ if reaction (15) is the rate-determining step. The Tafel slope of 40 mV dec⁻¹ can be attributed to the second reaction (14). At lower overpotentials, reaction (14) is the rate-determining step of the oxygen evolution.

In the case of the supporting electrolyte (0.1 mol L⁻¹ Na₂SO₄) applied in the investigation described in this paper, the mixed mechanism between acidic and alkaline medium for OER can be predicted. IR-compensated Tafel curves show only one slope for all tested electrodes with the following values:

$$ 70\%{\mathrm{TiO}}_2-30\%{\mathrm{RuO}}_2/\mathrm{Ti}\kern0.5em b=170.6\ \mathrm{mV}\ {\mathrm{dec}}^{-1}, $$

$$ 67\%{\mathrm{TiO}}_2-30\%{\mathrm{RuO}}_2-3\%{\mathrm{WO}}_3/\mathrm{Ti}\kern0.5em b=163.5\ \mathrm{mV}\ {\mathrm{dec}}^{-1}, $$

$$ 64\%{\mathrm{TiO}}_2-30\%{\mathrm{RuO}}_2-6\%{\mathrm{WO}}_3/\mathrm{Ti}\kern0.5em b=161.6\ \mathrm{mV}\ {\mathrm{dec}}^{-1}. $$

The calculated b values are higher than 120 mV dec⁻¹. The similar values were reported for β-PbO₂ [43, 47] and IrO₂-Nb₂O₅ catalysts [37]. The values of Tafel slope b prove that reaction (13) is the rate-determining step in the case of all tested electrodes. This step includes the formation and adsorption of the first intermediate S–OH_ads [43]. Relatively high values of b calculated for the tested electrodes can be attributed to the non-stoichiometry of the oxides and the number of sites to absorb the hydroxyl ions from water [48].

The obtained results show that the electrodes modified with WO₃ reveal higher activity towards OER. The electrode with 3% WO₃ seems to be less active than the electrode with 6%. This means that the electrode with 3% of WO₃ should be more active in organic oxidation what was proved by electrochemical degradation of the azo dye investigated at the tested electrodes and described in the paper [15]. Then, the degradation of this dye calculated as a change in chemical oxygen demand (COD) and as discoloration degree was higher but it did not mean total mineralization. However, the highest efficiency of the azo dye degradation calculated as a change in TOC was achieved with the application of the electrode with 6% WO₃. A decrease in TOC value corresponds to mineralization of the dye solution and can be observed at higher potentials at which other reactive oxygen species can be formed. Moreover, the highest azo dye degradation was achieved in the photoelectrochemical process performed at the electrode modified with 6% WO₃.

Open Circuit Potential and Corrosion Measurements

In the case of oxide electrodes with TiO₂-RuO₂ layer coated on a titanium substrate by the thermal decomposition method, their anodic stability is very important in processes of electrochemical oxidation of organic pollutants present in industrial wastewater. It is known that RuO₂ undergoes serious corrosion in the oxygen evolution reaction. The corrosion resistance of the oxide electrodes can be enhanced by introduction of different metal oxides (e.g., IrO₂, SnO₂, Ta₂O₅, and Nb₂O₅) to their oxide layer [32, 49,50,51]. Deactivation of the oxide electrodes by corrosion affects significantly their lifetime. Corrosion mechanism can be attributed to the consumption of the active component. Moreover, the formation of an interlayer of passivation on the substrate can also be responsible for electrode corrosion. Due to the fact that TiO₂-RuO₂/Ti electrodes modified with WO₃ can be used in decomposition of organic compounds (e.g., dyes) in the electrochemical and photoelectrochemical oxidation under simultaneous evolution of oxygen, it is important to investigate their stability and corrosion behavior.

In order to assess electrode stability, determination of OCP was performed in the solution of 0.1 mol L⁻¹ Na₂SO₄. The lowest value of OCP was determined for 70%TiO₂–30%RuO₂/Ti electrode (0.010 V). In the case of electrodes modified with 3 and 6% WO₃, OCP values were 0.235 and 0.205 V, respectively. The increase in OCP value by about 200 mV indicates that modification of WO₃ increases the stability of the electrode. However, introduction of 6% WO₃ into the oxide layer causes slightly less OCP value in comparison with 3% WO₃. OCP values of all electrodes are positive. According to C.P. De Pauli and S. Trasatti [52], the electrode potential does not depend on its morphology but only on the surface properties of active species. Thus, significant increase in OCP value can be attributed to domination of W species in the surface equilibria.

Next, corrosion parameters were determined for the tested electrodes in the solution of 0.1 mol L⁻¹ Na₂SO₄. The corrosion was evaluated using potentiodynamic polarization sweep preceded by OCP determination. After the equilibrium was established, the sample was cathodically and anodically polarized. Polarization curves were recorded in the potential range OCP ± 200 mV with the scan rate of 2 mV s⁻¹. The intersection between the cathodic and anodic Tafel slopes defines the corrosion current density (j_corr). Exemplary polarization curves recorded for the tested electrodes in Na₂SO₄ solution are presented in Fig. 11. The electrochemical parameters (b_a—anodic Tafel slope, b_c—cathodic Tafel slope, j_corr—corrosion current density, E_corr—corrosion potential, and R_p—polarization resistance) associated with electrochemical measurements and determined from polarization curves are listed in Table 6. Corrosion current density was calculated taking into consideration electrode active surface area (EAS) determined in ferri-ferrocyanide system.

Fig. 11

Polarization curves measured for the tested electrodes in 0.1 mol L⁻¹ Na₂SO₄

Table 6 Corrosion parameters of TiO₂-RuO₂/Ti electrodes modified with WO₃

Corrosion potential (E_corr) determined from polarization curves is the lowest for the unmodified electrode. Introduction of WO₃ into the oxide layer results in an increase in E_corr value by about 200 mV. Corrosion potential is a thermodynamic value and shows corrosion tendency. Thus, the modified electrodes seem to be more resistant to corrosion than the unmodified electrode. However, WO₃ amount of 6% is less favorable than 3% because E_corr value slightly decreases to − 0.016 V but is still much higher than for the unmodified electrode.

Polarization resistance (R_p) is the corrosion parameter which is used in comparison of a material resistance with corrosion under specified conditions. Modification of TiO₂-RuO₂/Ti electrode with WO₃ causes an increase in R_p value by 1.7 and 2.3 times with an increase in WO₃ content from 3 to 6%, respectively. Higher R_p value implies higher corrosion resistance.

Corrosion current is a kinetic value and determines the corrosion rate. The highest j_corr resulting in the highest corrosion rate is observed in the case of the unmodified electrode. Introduction of 3% WO₃ to the electrode oxide layer causes significant decrease in the j_corr value by 1.7 times. The higher amount of WO₃ (6%) results in an additional decrease in j_corr by 2 times. Taking into consideration E_corr and j_corr values, it can be concluded that the electrodes modified with WO₃ exhibit significantly higher corrosion resistance in the solution of Na₂SO₄.

The stability of TiO₂-RuO₂/Ti electrodes can be attributed to concurrent reactions of RuO₂ dissolution and oxygen evolution. RuO₂ dissolution reactions are as follows [50]:

$$ {\mathrm{RuO}}_2+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{RuO}}_2-\mathrm{OH}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$

(16)

$$ {\mathrm{RuO}}_2-\mathrm{OH}\leftrightarrow {\mathrm{RuO}}_3+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$

(17)

$$ {\mathrm{RuO}}_3+{\mathrm{H}}_2\mathrm{O}\leftrightarrow {\mathrm{RuO}}_3-\mathrm{OH}+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$

(18)

$$ {\mathrm{RuO}}_3-\mathrm{OH}\leftrightarrow {\mathrm{RuO}}_4+{\mathrm{H}}^{+}+{\mathrm{e}}^{-} $$

(19)

According to the dissolution mechanism with oxygen evolution reaction, the last two reactions (18 and 19) can be replaced by one reaction:

$$ {2\mathrm{RuO}}_3\leftrightarrow {2\mathrm{RuO}}_2+{\mathrm{O}}_2 $$

(20)

In both cases, reaction (16) is regarded as the rate-determining step [50]. Taking into consideration that products of metal oxide degradation are soluble, the oxide layer on the anode surface is gradually enriched in insulating TiO₂ [51]. TiO₂ is applied as a stabilizing component of the oxide layer and additionally comes from the titanium substrate. Both processes lead to anode passivation.

In the case of TiO₂-RuO₂/Ti anodes modified with WO₃, it can be assumed that oxygen evolution occurs mainly at the WO₃ active sites which hinders RuO₂ degradation and results in higher stability of the anode.

Besides E_corr and j_corr values, such parameters as cathodic and anodic Tafel slopes were determined. The cathodic Tafel slopes b_c determined for the tested electrodes are similar but with a slight decrease with an increase in WO₃ content and indicates similar electrochemical reaction for hydrogen evolution. Moreover, the anodic Tafel slopes are also similar indicating similar mechanism of anodic dissolution of the electrode material independent on the presence of WO₃ in the oxide layer.

Similar measurements were performed at different immersion times (2 and 4 h) in Na₂SO₄ solution in order to determine its effect on corrosion resistance of the tested electrodes. Exemplary polarization curves are presented in Fig. 12. Electrochemical parameters determined from recorded polarization curves are compared in Table 7.

Fig. 12

Comparison of polarization curves recorded for 64%TiO₂–30%RuO₂–6%WO₃/Ti electrode after 2- and 4-h immersion in Na₂SO₄

Table 7 Comparison of corrosion parameters obtained during measurements after 2 and 4 h of immersion in Na₂SO₄ solution

The cathodic and anodic slopes determined for the tested electrodes from polarization curves recorded at different immersion times were similar indicating that the corrosion mechanism was the same even after 4-h immersion. However, clear changes in E_corr values were observed. In the case of all tested electrodes, E_corr was shifted towards more negative values which prove less resistance to corrosion phenomena. The unmodified electrode revealed E_corr value lower by 66 mV while the electrodes modified with WO₃ were characterized by E_corr value lower by about 170–180 mV after 4-h immersion but this value was still higher than in the case of the unmodified material. Thus, even if a decrease in E_corr value is higher for TiO₂-RuO₂-WO₃/Ti electrodes, their corrosion resistance is still higher in comparison with TiO₂-RuO₂/Ti electrode. Corrosion current is the second parameter which varied in measurements at different immersion times. In the case of all tested electrodes, its value increased with the immersion time which means higher corrosion rate. Although corrosion current increased only by 14% for the unmodified electrode and by 20% for the electrodes modified with WO₃ after 4-h immersion, the modified electrodes still reveal much lower corrosion current. Moreover, in the case of the electrode modified with 6% WO₃, corrosion current was even lower after 4-h immersion in comparison with 2-h immersion and was almost twice lower in comparison with the unmodified electrode.

Results of corrosion investigation in Na₂SO₄ solution prove that modification of the electrode material with WO₃ increases its corrosion resistance. The higher addition of WO₃ (6%) is even more advantageous.

Electrode Stability

Sufficient electrode stability is necessary in the process of electrochemical or photoelectrochemical degradation of organic pollutants. Results of accelerated stability tests (AST) are presented in Fig. 13. In the case of both modified electrodes, their potential decreases by 0.22 V during 1440 s probably due to wetting of their less accessible inner surface [32, 53]. Afterward, only small variations of the electrode potential are observed. Finally, the potential increases rapidly and the electrodes are quickly deactivated. The deactivation mechanism can be attributed to the following reasons [53, 54]:

• Formation of insulating TiO_x interlayer,

• Removal of the catalytic material by the intense gas production, and

• Oxidation of RuO₂ and formation of soluble Ru species.

Fig. 13

Changes of the electrode potential with time during AST at 0.2 A cm⁻² in Na₂SO₄ (0.1 mol L⁻¹) for TiO₂-RuO₂/Ti electrodes modified with WO₃ and the non-modified electrode

As can be seen in Fig. 13, the electrodes modified with WO₃ display longer accelerated lifetime in comparison with the non-modified electrode. The lifetime for 67%TiO₂–30%RuO₂–3%WO₃/Ti and 64%TiO₂–30%RuO₂–6%WO₃/Ti determined for the anodic potential 2 V higher than the potential at time zero was 8.93 and 10.40 h, respectively. The higher content of WO₃, the higher the lifetime is observed. The lifetime for 70%TiO₂–30%RuO₂/Ti electrode was 7.98 h. The reason for this should be the consequence of WO₃ introduction to the oxide layer of the non-modified electrode and its effect on the morphology and/or on the mechanism of activity loss. In comparison with the non-modified electrode, the electrodes with WO₃ reveal less cracked structure (Fig. 8) which might hamper the electrolyte penetration into the surface of the substrate resulting in a delay in the formation of insulating TiO₂ interlayer. The results indicate that the replacement of TiO₂ by WO₃ can improve the stability of 70%TiO₂–30%RuO₂/Ti electrode. Moreover, the higher content (6%) of WO₃ is more advantageous and results in the higher lifetime of the electrode. This confirms that the electrodes modified with WO₃ reveal higher corrosion resistance in Na₂SO₄ solution.

Conclusions

The investigation has demonstrated the effect of modification with WO₃ on electrochemical and corrosion characterization of TiO₂-RuO₂/Ti electrodes prepared by thermal decomposition of metal precursors. The introduction of WO₃ to the oxide layer of the electrodes resulted in a decrease in electrode active surface determined in [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ system. Both methods applied in EAS determination proved a 10% decrease in EAS and roughness factor per each 3% WO₃ introduced to the electrode surface layer in comparison with the non-modified electrode. This decrease in EAS of the modified electrodes corresponds to a decrease in the electrooxidation current of the azo dye observed in cyclic voltammograms previously described and recorded in the presence of the supporting electrolyte (Na₂SO₄). Thus, EAS value was also evaluated in Na₂SO₄ solution by determination of voltammetric charges which are proportional to the number of active sites on electrode surface. Analysis of the results proves almost twice decrease in the number of active sites as a result of modification. However, the electrode with 6% WO₃ shows a little higher active surface than the electrode with 3% WO₃. This is consistent with the results of determination of electrode activity towards oxygen evolution. Anodic polarization measurements in Na₂SO₄ solution proved a little higher activity of the electrode modified with 6% WO₃ than the electrode with 3% WO₃, what means that the last electrode reveals the highest activity in organic pollutants oxidation.

Taking into consideration that these electrodes are tested in the processes of organic pollutant degradation, their higher stability is highly required. The study of the open circuit potential demonstrates that the introduction of WO₃ to the electrode oxide layer results in its significantly higher stability in comparison with the non-modified electrode. The electrode modified with 3% WO₃ reveals even higher stability than the electrode with 6% WO₃. Moreover, the corrosion characteristics of the tested electrodes show that the electrode with 3% WO₃ demonstrates the highest resistance to corrosion phenomena in Na₂SO₄ solution and also the lowest corrosion resistance estimated by corrosion potential, polarization resistance, and corrosion current. Accelerated stability tests showed that a small addition of WO₃ increased the lifetime of TiO₂-RuO₂/Ti electrode.

The modification of TiO₂-RuO₂/Ti electrodes with WO₃ seems to be advantageous in their application in electrochemical and photoelectrochemical degradation of organic pollutants, e.g., azo dyes. Although introduction of WO₃ into the electrode oxide layer results in lower active surface area, its activity towards organics electrooxidation increases and its stability is higher in comparison with the non-modified electrode. The previous results of the azo dye degradation at these electrodes investigated in galvanostatic electrolysis and also with the presence of UV and VIS irradiation confirm the results described in this paper.