The increased amount of CO₂ in the atmosphere is claimed to be one of the major contributors to the greenhouse effect, and will result in serious global warming issues ^[1]. Among various conversion methods, the electrochemical synthesis of high-value chemicals from CO₂ offers several advantages such as process simplicity and flexibility, and production of various organic chemicals, depending on the type of catalyst used ^{[2, 3]}. CO₂ is a stable molecule and generally produced by fossil fuel combustion and respiration. Converting CO₂ to useful chemicals at the same rate as its present production is beyond our current scientific and technological abilities ^[4]. The reduction of CO₂ involves the use of specific metal catalysts, to achieve product selectivity, and because of the sluggish kinetics of CO₂ electroreduction ^[5].

The study of CO₂ electroreduction in aqueous solutions at ambient temperature has focused on metal electrodes ^{[3, 6]}. The catalytic reduction of CO₂ to methanol was achieved over Cu under hydrothermal conditions with a methanol yield of 11.4% ^[7]. The product distributions and faradaic efficiencies obtained in the electrochemical reduction of CO₂ with Cu foams differ significantly from those obtained at smooth electropolished Cu electrodes. This is attributed to the high surface roughness, hierarchical porosity, and confinement of reactive species in the case of Cu foams. The faradaic efficiency for formic acid production at Cu foam electrodes was higher at all tested potentials, with a maximum efficiency of 37% at −1.5 V, which is the highest value obtained for the electroreduction of CO₂ to formic acid at a Cu electrode under ambient pressure ^[8]. The use of gas-diffusion electrodes (GDEs) for electrochemical reduction of CO₂ on Pb,In, and Sn under acidic conditions gave high efficiencies for formic acid (pH ≈ 2) production ^[9]. An Sn-based GDE (SGDE) showed good stability during CO₂ reduction; the faradaic efficiency for conversion of CO₂ to formate reached 18% during the initial 5 min and remained at about 12% until the end of the reduction time, i.e., 1 h ^[10]. Recently,SGDEs have attracted much attention for CO₂ reduction. Wang et al. ^{[11, 12]} reported that an SGDE with polytetrafluoroethylene as an additive gave a good electrochemical performance, because of the increased active catalyst surface area and CO₂ diffusion, and high catalyst loading, i.e., 5 mg/cm².

The deactivation of Sn-metal-based electrodes during CO₂ reduction is fast, and the reduction reaction on these electrodes requires an overpotential of at least ∼860 mV at a current density of 4−5 mA/cm² in an aqueous solution saturated with CO₂ at 0.1 kPa ^[13]. It is vital to explore the use of metal oxides in CO₂ reduction to overcome this problem, but there have been few reports of such studies. The role of metal oxides, whether as catalysts for the formation of formic acid or as precursors for the fabrication of well-structured catalysts, remains unclear. Kanan’s group ^[14] published several reports on the metal oxide effect in CO₂ reduction. The faradaic efficiency for CO₂ reduction depended greatly on the presence of SnO_x; Sn/SnO_x thin-film electrodes catalyzed the formation of CO and formic acid as the main reaction products. The faradaic efficiency for formic acid reached 30% at −0.7 V vs the normal hydrogen electrode. An important result of this study is the observation that controlling the size of tin oxide nanoparticles (NPs) on carbon supports enables overpotentials as low as ∼340 mV to be achieved for CO₂ reduction to formate, with significant enhancements in current density to over 10 mA/cm² on high- surface-area graphene supports. Reduced nanoscale tin oxide catalysts are highly stable during controlled-potential electrolysis ^[15].

The most important and valuable products of CO₂ reduction are formate and formic acid. SnO₂ shows good catalytic activity in formate production, but the electrolyte conditions greatly affect the formation of formic acid ^{[16, 17, 18, 19]}. The pH value of the electrolyte significantly affects the electrode potentials for the reduction of H₂O and CO₂ ^[20]:

An environment that is too acidic promotes hydrogen formation, and one that is too alkaline does not favor formation of formic acid. CO₂ electrolysis in a neutral or mildly alkaline environment stabilizes the oxide. The electrolyte concentration also greatly influences the formation of formic acid ^[21]. The faradaic efficiency for formic acid production in KHCO₃ (0.5 mol/L) was greater than that in K₂CO₃ (0.1 mol/L) with an Sn granule electrode in a fixed-bed reactor ^[16]. The highest achieved faradaic efficiency for formate production was 88.4% in 0.1 mol/L KHCO₃ at −1.72 V vs the saturated calomel electrode (SCE) ^[17], and the faradaic efficiency was between 65.0% and 79.9% in KHCO₃ (0.5 mol/L) ^[22].

In this work, we developed a novel SnO₂ NP catalyst with a high catalytic efficiency for CO₂ electroreduction, based on a GDE. Unlike that used in Wang’s group ^[11], the catalyst was a nanostructured tin oxide consisting of SnO₂ NPs with highly porous structures, and was synthesized using a facile hydrothermal self-assembly process. SnO₂-50/GDE (an SnO₂-based gas-diffusion electrode, where 50 indicates the 50% ethanol content of the electrolyte) was prepared by coating SnO₂ catalyst ink on a gas-diffusion carbon paper sheet. The SnO₂ catalyst ink was prepared by homogeneously mixing SnO₂ catalyst particles, 5 wt% Nafion solution, and isopropyl alcohol. The electrolyte conditions, i.e., the pH and concentration, were controlled, to enable a better understanding of the mechanisms of the effects of the electrolyte on formic acid formation and the faradaic efficiency. The SnO₂ NP catalyst morphology was examined using scanning electron microscopy (SEM). The electrochemical properties of the modified electrode, i.e.,SnO₂-50/ GDE, in CO₂ reduction were investigated thoroughly using cyclic voltammetry (CV), linear sweep voltammetry (LSV),CO₂ electrolysis, and ion chromatography. The production rate and faradaic efficiency for formate, which can be used as a liquid fuel during CO₂ reduction, were also investigated.

2 Experimental

2.1 Catalyst synthesis

An SnO₂ NP catalyst was synthesized from SnCl₄ and D-glucose monohydrate using a facile hydrothermal self- assembly process. SnCl₄ (4 mmol) was mixed with D-glucose monohydrate (10 mmol) and the mixture was dissolved in distilled water and ethanol (totally 35 mL) with stirring until a transparent solution was obtained. The mixture solution was transferred to a 100 mL Teflon-lined stainless- steel autoclave, which was sealed and kept at 180 ℃ for 24 h. The formed black powder was collected, washed several times with ethanol/water, and dried in a vacuum oven at 60 ℃ for 5 h. The obtained powder was calcined in air at 550 ℃ for 5 h, during which the black sediment gradually turned white, indicating the successful removal of carbon by oxidation in air, to give the SnO₂ NP catalyst. The catalyst is denoted by SnO₂-50, where 50 indicates that the percentage of ethanol content in the mixture solution is 50%.

2.2 Electrode preparation and electrochemical tests

For all electrochemical measurements, the SnO₂-50 NP catalyst was coated on a gas-diffusion carbon paper sheet (Toray,TGP-H-090) to form a working electrode. The catalyst ink was prepared by suspending the SnO₂-50 NP catalyst (15 mg) in a mixture of 5 wt% Nafion solution (100 mg) and 99.7 wt% isopropyl alcohol (1.4 mL; Sinopharm Chemical Reagent Co.). A catalyst-coated gas-diffusion layer on 4 cm² Toray carbon paper (TGP-H-090), denoted by SnO₂-50/GDE, was used as the working electrode. It was tested using a conventional three- electrode electrochemical H-type cell.

Electrochemical characterization was performed using a standard H-type cell (Aldrich Nafion^®117) equipped with a gas inlet and outlet, which allowed the passage of either N₂ (99.99%) or CO₂ (99.99%) through the solution, to investigate the catalyst properties and CO₂ reduction performance. A standard H-type cell, with a piece of Nafion^®117 cation- exchange membrane (H⁺ form) as a separator,SnO₂-50/GDE as the working electrode, a Pt foil electrode as the counter electrode, and an SCE as the reference electrode were used. All electrochemical measurements were performed using a CHI 660E instrument. For the CO₂ reduction measurements, aqueous electrolytes of KHCO₃ concentration from 0.1 to 1.0 mol/L were used as the functional electrolyte to obtain the desired solution concentration; CO₂ gas (99.99%) at 1 atm was bubbled through the solution for 30 min before all measurements.

The electrocatalytic activity of the SnO₂-50/GDE working electrode was tested using CV and LSV at potential scan rates of 50 and 5 mV/s, respectively, in the potential range 1.0 to −1.6 V vs the SCE. Controlled-potential electrolysis was performed using a CHI 660E electrochemical analyzer, with the same standard H-type cell. An aqueous KHCO₃ (0.5 mol/L) electrolyte was used for the CO₂ reduction measurements; CO₂ gas (99.99%) at 0.1 kPa was bubble through the solution for at least 30 min before each measurement. For long-term electrolysis measurements, a constant potential of −1.70 V vs the standard hydrogen electrode (SHE) was imposed for 28 h. For soluble product measurements, a constant potential of −1.70 V vs the SHE was imposed for 1 h, and the electrolysis currents were recorded continuously. All tests were performed at ambient temperature and pressure.

2.3 Physical characterization and reduction product determination

The crystal-phase X-ray diffraction (XRD) pattern of the SnO₂-50 NP catalyst synthesized at 180 ℃ for 24 h was obtained using a Philips PW3830 X-ray diffractometer equipped with a Cu K_α radiation (λ = 0.15406 nm) source. The intensity data were collected at 25 ℃ in the 2θ range from 0° to 90°, at a scanning rate of 1.20°/min. The morphology of the SnO₂-50 catalyst was examined using SEM (Ultra plus thermal field- emission instrument,Carl Zeiss SMT AG,Germany). The chemical composition of SnO₂-50/GDE after long-term electrolysis was determined using X-ray photoelectron spectroscopy (XPS; Kratos AXIS Ultra DLD electron spectrometer with Al K_α X-ray anode source,hv = 1486.6 eV, at 250 W and 14.0 kV). The solution after CO₂ reduction was diluted 50 times or 100 times (which is convenient for formate detection in the mobile phase) before determination of the soluble reduction products. The product solution was filtered using a filtration membrane (0.22 μm) and the formate concentrations in the electrolyte were determined directly by ion chromatography (ICS-90,Dionex,USA), using an AS144 mm × 250 mm separation column and a flow rate of 1 mL/min. The mobile phase used in instrument preparation was a mixed aqueous solution of Na₂CO₃ (3.5 mmol/L) and NaHCO₃ (1.0 mmol/L), and aqueous H₂SO₄ (20 mmol/L) was used as the regenerating liquid.

3 Results and discussion

3.1 Physical characterization of SnO₂ NP catalyst

The morphology and crystal structure (phase composition) of SnO₂ were examined using SEM and XRD, respectively. Fig. 1 shows SEM images at two magnifications of the SnO₂-50 NP catalyst, which was prepared using a solvent of ethanol : distilled water = 1 : 1. Fig. 1(a) shows that the sample consisted of a mixture of NP and nanosphere aggregates of diameter 500 nm to 1 μm with a highly porous structure. The high-resolution SEM image (Fig. 1(b)) shows that these large NPs and nanospheres have a clear three-dimensioned hierarchical structure, consisting entirely of secondary structures composed of aggregated small primary SnO₂ NPs of diameter 20-25 nm. This special morphology provides catalysts with large surface areas, and greatly affects the catalytic activity in CO₂ electroreduction ^{[23, 24]}.

Fig. 1

Fig. 1. SEM images of SnO2-50 aggregates at different magnifications. (a) ×10000; (b) ×1cjc-37-7-1081.

The crystalline structure of as-prepared SnO₂ was confirmed using XRD. The XRD pattern in Fig. 2 shows that the hierarchical structure consisted of small SnO₂ NPs with good phase purity. The increased intensity of the (110),(101),(200), and (211) diffraction peaks, from tetragonal SnO₂, show that the SnO₂ NPs had pronounced crystallinities. The main diffraction peaks from (110),(101),(200), and (211) were broad, which indicated that the SnO₂ NPs were nanocrystalline. The crystallinity was calculated to be 72% and the crystalline grain sizes of these four crystal surfaces were 14.8, 17.1, 15.4, and 15.7 nm, respectively.

Fig. 2

Fig. 2. XRD pattern of SnO2-50 prepared from the solvothermal reaction.

3.2 Effects of electrolyte concentration on electrochemical activity in catalytic CO₂ reduction

The CO₂ reduction activity at an SnO₂ electrode prepared from SnO₂-50/GDE was investigated using LSV in N₂- and CO₂-saturated KHCO₃ solution (0.5 mol/L), as shown in Fig. 3(a). The SnO₂-50/GDE showed good catalytic activity in CO₂ reduction, with a negative onset potential at −0.56 V vs the SHE, which is 130 mV more positive than that under N₂. The significant increase in the current in the high potential range under N₂ can be attributed to the reduction of hydrogen ions. However, the higher current in the presence of CO₂ can be ascribed to the reduction of CO₂ and H₂O; the enhanced current indicated reduction to formate. The maximum current density at −1.36 V vs the SHE under CO₂ was about 1.5 times that under N₂.

Fig. 3

Fig. 3. Linear sweep voltammetric (LSV) curves of SnO2/GDE-50. (a) In N2- and CO2-saturated KHCO3 (0.5 mol/L) electrolytes; (b) In CO2- saturated KHCO3 electrolyte at different concentrations.

The effects of electrolyte concentration on the electrochemical catalytic activity in CO₂ reduction using KHCO₃ solutions of various concentrations with CO₂ bubbling and the SnO₂-50/ GDE were studied using LSV. The scans (Fig. 3(b)) show that the SnO₂-50/GDE gave a positive onset potential (about −0.56 V vs the SHE) in all the tested electrolytes. Fig. 3(b) shows that when the KHCO₃ solution concentration was 0.1 mol/L, the current density was 15.5 mA/cm². The catalytic activity increased with increasing KHCO₃ solution concentration from 0.1 to 0.5 mol/L, as can be seen from the more positive onset potential and higher current density. The KHCO₃ solution (0.5 mol/L) gave the best catalytic activity for CO₂ reduction, with a maximum current density of 25.0 mA/cm², and the most positive onset potential, i.e.,−0.56 V vs the SHE. When the electrolyte concentration was further increased to 1.0 mol/L, the current density decreased sharply to 20 mA/cm², and the CO₂ reduction catalytic activity declined compared to that in KHCO₃ solution (0.5 mol/L).

Controlled-potential electrolysis was performed to investigate the effects of electrolyte concentration on the formate yield and faradaic efficiency using the SnO₂-50/GDE. The experiments were performed using CO₂-saturated KHCO₃ solution. The results (Fig. 4) show that formate production depended strongly on the electrolyte concentration. The faradaic efficiency for formate production first increased, reached a maximum value, and then decreased with increasing electrolyte concentration. The highest faradaic efficiency obtained in this research was 56%, in KHCO₃ solution (0.5 mol/L), which is significantly better at a low overpotential (−0.56 V vs the SHE) than the values reported elsewhere ^{[15, 22, 25, 26]}. High faradaic efficiencies achieved with Sn-based electrodes have previously been reported ^{[10, 15, 20]}. Formate production began when the electrolyte concentration was 0.1 mol/L; the faradaic efficiency for formate formation was 29% and the amount of formate produced was 82 mg/L. On further increasing the solution concentration from 0.1 to 0.5 mol/L, the faradaic efficiency and formate production increased; the highest faradaic efficiency was obtained using 0.5 mol/L KHCO₃; the corresponding formate production was 292 mg/L. This is consistent with the superior catalytic activity in KHCO₃ (0.5 mol/L) described above. The maximum formate production was 428 mg/L in KHCO₃ (0.7 mol/L). However, the faradaic efficiency decreased sharply with further increases in the electrolyte concentration. HCO₃⁻ is directly involved in the reduction of CO₂ to formate; when the HCO₃⁻ concentration is low,CO₂ reduction is controlled by mass trans fer. The faradaic efficiency for formate production therefore increased with increasing KHCO₃ concentration. When the concentration increased sufficiently, the formate synthesis was controlled by charge transfer, therefore further increasing the KHCO₃ solution concentration did not increase the faradaic efficiency. In addition, the pH of the system increased with increasing HCO₃⁻ concentration, leading to an improved formate synthesis potential; therefore, the faradaic efficiency at a high formate concentration decreased sharply.

Fig. 4

Fig. 4. Effects of the electrolyte concentration on formate yield and faradaic efficiency for formate production at the SnO2-50/GDE electrode in CO2-saturated KHCO3 solution.

3.3 Effects of electrolyte pH values on catalytic CO₂ reduction

The effects of the electrolyte pH were investigated at pH = 6, 7, 8.3, and 9, because high acidity promotes hydrogen formation and high alkalinity is not conducive to the formation of formic acid. A KHCO₃ solution of concentration 0.5 mol/L, which gives the optimum catalytic activity, was selected as the target electrolyte. The initial pH of the KHCO₃ solution (0.5 mol/L) was 8.3. Fig. 5(a) shows that the onset potential was more positive at pH = 8.3 than at the other pH values tested. At the most negative potential, the cathodic current density was much larger at pH = 8.3 than at other pH values, and almost twice that at pH = 6. These results suggest that the optimum electrolyte pH for electroreduction of CO₂ was 8.3; CO₂ reduction in this mildly alkaline environment maintains the oxide stability. Fig. 5(b) shows that the highest formate production obtained was 259 mg/L, at an electrolyte pH = 8.3; this agrees well with the catalytic activity determined form the LSV curves (Fig. 5(a)). For electrolytes of pH = 6 and 7, formate production, about 180 mg/L, was clearly lower than that achieved at pH = 8.3. However, further increasing the electrolyte from pH = 8.3 to 9.0 caused formate production to decrease slightly. Fig. 5(c) shows the faradaic efficiencies and the total charge Q during electrolysis for formate production as a function of electrolyte pH. The trends shown in Fig. 5(c) for the faradaic efficiency are similar to those for formate production shown in Fig. 5(b). The faradaic efficiency for formate production increased with increasing from pH = 6 to 8.3, at which it reached a maximum of 56%, and the total charge reached 126 C. As the pH value increased further, the faradaic efficiency began to drop. This may be because the alkaline environment did not benefit formate production.

Fig. 5

Fig. 5. The effect of the electrolyte pH value on the CO2 reduction: (a) LSV curves; (b) Formate production; (c) Faradaic efficiencies of formate and total charge Q on the SnO2-50/GDE electrode in CO2-saturated KHCO3 solution (0.5 mol/L).

3.4 Faradaic efficiency for long-term CO₂ electrolysis to formate

Fig. 6(a) shows the degradation rates of the faradaic efficiency for formate production on SnO₂-50/GDE for 28 h. It was previously reported ^{[3, 16, 17]} that a high faradaic efficiency was observed during initial electrolysis, but the efficiency then decreased significantly. In the first hour, the formate faradaic efficiency reached 56%, which was the highest value during the first 5 h, and the corresponding formate production was 419 mg/L. When CO₂ reduction was performed for 5 h, the formate faradaic efficiency sharply decreased to 44%, and formate production increased sharply and almost linearly to 1779 mg/L. After CO₂ electrolysis for 17 h, the faradaic efficiency decreased significantly, and was almost half that in the first hour. As the electrolysis time increased from 17 to 28 h, the faradaic efficiency and formate production both decreased steadily; the faradaic efficiency decreased to 24% after 28 h. As previously reported, the faradaic efficiency for formate production changes with electrolysis time. This can be explained by formate oxidation on the anode or contamination of the SnO₂-50/GDE cathode with impurities ^{[3, 16, 17]}. We determined the cathode composition before and after electrolysis using XPS. The results suggest that the decrease in the faradaic efficiency was caused by deposition of trace amounts of contaminant ions on the SnO₂-50/GDE surface, mainly fluoride as shown in Fig. 6(b). The presence of fluoride hindered CO₂ reduction to formate on the SnO₂-50/GDE surface. The full spectrum scan (Fig. 6(c)) showed that a large amount of fluoride, present as an impurity in the KHCO₃ electrolyte, was deposited on the electrode. The high-resolution F 1s XPS had a peak at 688.8 eV, which is consistent with F 1s bound to carbon in a straight chain.

Fig. 6

Fig. 6. (a) Variations of the faradaic efficiency and formate production for 28 h long time electrolysis; (b) High-resolution F 1s XPS spectrum of SnO2-50/GDE electrode after electrolysis in CO2-saturated KHCO3 solution (0.5 mol/L); (c) Full spectrum scanning.

4 Conclusions

Nanostructured SnO₂ particles were synthesized using a facile hydrothermal self-assembly method, and their use as a catalyst in CO₂ electroreduction was investigated. SEM showed that the SnO₂ NPs had a three-dimensioned hierarchical structure, and consisted of a mixture of NP and nanosphere aggregates of diameter 500 nm to 1 μm, with a highly porous structure. The catalytic activity and formate selectivity depended strongly on the electrolyte concentration. The results showed that the catalytic activity increased with increasing electrolyte concentration from 0.1 to 0.5 mol/L, at which the best catalytic activity was obtained. The highest faradaic efficiency obtained was 56% in KHCO₃ (0.5 mol/L). This was attributed to direct involvement of HCO₃⁻in the reaction; formate formation was controlled by mass transfer and charge transfer at low and high HCO₃⁻concentrations, respectively. The electrolyte pH affected formate production during CO₂ reduction. Our results suggest that pH = 8.3 was the optimum electrolyte pH for electroreduction of CO₂; CO₂ reduction in a mildly alkaline environment maintained oxide stability. A catalyst stability test for 28 h showed degradation of the faradaic efficiency for formate production on SnO₂-50/GDE. The decreased faradaic efficiency was caused by deposition of fluoride ions on the SnO₂-50/GDE surface, as shown by XPS. The presence of fluoride hindered the reduction of CO₂ to formate on the SnO₂-50/GDE surface.

References

[1]	C. S. Song, Catal. Today,2006, 115 :2–32.
[2]	J. Lee, Y. Kwon, R. L. Machunda H. J. Lee, Chem. Asian J.,2009, 4 :1516–1523.
[3]	Y. Hori, Modern Aspects Electrochem.,2008, 42 :89–189.
[4]	A. T. Bell, B.C. Gates, D. Ray, Basic Research Needs: Catalysis for Energy, U. S. Department of Energy, Basic Energy Sciences Workshop, 2007, 69-75.
[5]	C. M. Sánchez-Sánchez, V. Montiel, D. A. Tryk, A. Aldaz A. Fujishima, Pure Appl. Chem.,2001, 73 :1917–1927.
[6]	D. T. Whipple P. J. A. Kenis, J. Phys. Chem. Lett.,2010, 1 :3451–3458.
[7]	Z. B. Huo, M. B. Hu, X. Zeng, J. Yun F. M. Jin, Catal. Today,2012, 194 :25–29.
[8]	S. Sen, D. Liu G. T. R. Palmore, ACS Catal.,2014, 4 :3091–3095.
[9]	M. N. Mahmood, D. Masheder C. J. Harty, J. Appl. Electrochem.,1987, 17 :1159–1170.
[10]	R. L. Machunda, H. K. Ju J. Lee, Curr. Appl. Phys.,2011, 11 :986–988.
[11]	Q. N. Wang, H. Dong, H. Yu H. B. Yu, J. Power Sources,2015, 279 :1–5.
[12]	Q. N. Wang, H. Dong H. B. Yu, J. Power Sources,2014, 271 :278–284.
[13]	M. Grasemann G. Laurenczy, Energy Environ. Sci.,2012, 5 :8171–8181.
[14]	Y. H. Chen M. W. Kanan, J. Am. Chem. Soc.,2012, 134 :1986–1989.
[15]	S. Zhang, P. Kang T. J. Meyer, J. Am. Chem. Soc.,2014, 136 :1734–1737.
[16]	F. Köleli, T. Atilan, N. Palamut, A. M. Gizir, R. Aydin C. H. Hamann, J. Appl. Electrochem.,2003, 33 :447–450.
[17]	Y. Hori, H. Wakebe, T. Tsukamoto O. Koga, Electrochim. Acta,1994, 39 :1833–1839.
[18]	K. Hara, A. Kudo T. Sakata, J. Electroanal. Chem.,1995, 391 :141–147.
[19]	T. Mizuno, K. Ohta, A. Sasaki, T. Akai, M. Hirano A. Kawabe, Energy Source,1995, 17 :503–508.
[20]	W. X. Lv, R. Zhang, P. R. Gao L. X. Lei, J. Power Source,2014, 253 :276–281.
[21]	J. J. Wu, F. G. Risalvato, F. S. Ke, P. J. Pellechia X. D. Zhou, J. Electrochem. Soc.,2012, 159 :F353–F359.
[22]	Y. Hori, K. Kikuchi S. Suzuki, Chem. Lett.,1985, 14 :1695–1698.
[23]	A. Zeng, E. Liu, S. N. Tan, S. Zhang J. Gao, Electroanalysis,2002, 14 :1110–1115.
[24]	H. X. Zhang, J. Feng M. L. Zhang, Mater. Res. Bull.,2008, 43 :3221–3226.
[25]	J. Lee Y. Tak, Electrochim. Acta,2001, 46 :3015–3022.
[26]	Y. Hori, A. Murata R. Takahashi, J. Chem. Soc., Faraday Trans, 1,1989, 85 :2309–2326.