Influence of TiCl₄ treatment on back contact dye-sensitized solar cells sensitized with black dye

Abstract

To better understand why titanium tetrachloride (TiCl₄) treatment improves short circuit current, we studied its effects on back contact dye-sensitized solar cells sensitized with black dye [tri(thiocyanato)(4,4′,4″-tricarboxy-2,2′:6′,2″-terpyridine)ruthenium(II), Ru(tcterpy)(NCS)₃] using transient absorption spectroscopy and electrochemical impedance spectroscopy. We found that the TiCl₄ treatment improved short circuit current and achieved an overall energy conversion efficiency of 8.9%. The transient absorption signals did not change as a result of the treatment, suggesting that electron injection efficiency is not affected by the treatment. The impedance related to electron transport between TiO₂ particles decreased and the peak frequency of the imaginary part of the electrochemical impedance spectra assigned to electron transfer from TiO₂ to the redox couple was shifted to lower frequency by the treatment. This clearly indicates that TiCl₄ treatment improved electron transport in the nanocrystalline TiO₂ film in back contact dye-sensitized solar cells.

---

Broader context

Recently, there has been a significant increase in interest in renewable energy due to growing concerns about global climate change and potential fossil fuel shortages. Solar energy is a renewable alternative to conventional energy sources. In order to lower the fabrication cost of solar energy conversion devices, people have been investigating dye-sensitized solar cells (DSCs), which can be fabricated using screen printing and low cost materials. To improve device efficiency, a TiCl₄ treatment for nanocrystalline TiO₂ films has been used to increase the current generation within the film. Although many articles on this method have been published, it still remains unclear why the TiCl₄ treatment increases the current generation. This paper is the first attempt to clearly prove the roll the TiCl₄ treatment plays in increasing the current generation using a back contact dye sensitized solar cell configuration sensitized with black dye. This configuration is suitable for addressing the TiCl₄ treatment effect because the electron transport and electron injection are more sensitive in this configuration than in conventional DSCs sensitized with bipyridyne dyes (N3, N719). The results of our experiment suggest that the TiCl₄ treatment increases electron transport and electron lifetime within the nanocrystalline TiO₂ film.

Introduction

Dye-sensitized solar cells (DSCs) have been widely studied over the past decade.^1–3 Primarily these studies have focused on increasing the efficiency of DSCs. The best solar energy to electricity conversion efficiency (η) achieved to date in a DSC was over η = 11%.^4–6 For further improvement, many studies have been dedicated to understanding the charge transfer mechanism of DSCs using transient absorption spectroscopy,^7–9electrochemical impedance spectroscopy,^10–12 intensity modulated photocurrent spectroscopy (IMPS)/intensity modulated photovoltage spectroscopy (IMVS)^13,14 and stepped light-induced transient measurement.¹⁵ Despite such attempts, the electron transport/transfer mechanisms of DSCs are still debated due to the complexity of DSCs.

Titanium tetrachloride (TiCl₄) treatment is known to improve the short circuit current of DSCs. TiCl₄ treatment is carried out by an aqueous solution of TiCl₄ applied onto the nanocrystalline TiO₂ films and then sintering again, forming an extra TiO₂ layer onto the nanocrystalline TiO₂ film, which increases the diameter of TiO₂ particles. In previous studies, several reasons why the TiCl₄ treatment improves short circuit current were proposed such as the enhancement of dye absorption on the surface,¹⁶ improvement of electron transport properties in nanocrystalline TiO₂ films,¹⁷ and the enhancement of charge separation efficiency.^18,19

In our previous study²⁰ we examined the TiCl₄ treatment of back contact dyes-sensitized solar cells (BCDSCs), where the nanocrystalline TiO₂ film is fabricated on a glass-only substrate instead of a transparent conductive oxide (TCO) coated glass and the electrode for electron collection is located on the surface of the nanocrystalline TiO₂ film opposite to the irradiated surface.^20–25Fig. 1 illustrates the cross section of a BCDSC. For BCDSCs based on a nanocrystalline TiO₂ film with N719 dye [cis-bis(isothiocyanato)bis(2,2′-bipyridyl-4,4′-dicarboxylato)ruthenium(II)bis(tetrabutylammonium)], we found that TiCl₄ treatment improved electron transport property in the nanocrystalline TiO₂ film in the BCDSC, which was not clearly observed in recent papers on TiCl₄ treatment effect that used ordinary DSCs.^17,18 The average electron transport length in the nanocrystalline TiO₂ film in BCDSC is longer than that in a conventional DSC where the electrode is located on the same side of the irradiated surface, indicating that the electrons in BCDSCs are more sensitive to changes in electron transport. BCDSCs are therefore useful to study small changes in electron transport in the nanocrystallne TiO₂ film. Although we found that the improvement of electron transport is one of the reasons TiCl₄ treatment improves solar cell performance, there are the other possibilities, such as the enhancement of dye adsorption and electron injection efficiency.

Fig. 1 Cross-section of a typical back contact dye-sensitized solar cell (BCDSC). The back contact electrode (BCE) is located on the TiO₂ surface and electrolyte layer is between the BCE and the counter electrode (CE).

Black dye [tri(thiocyanato)(4,4′,4″-tricarboxy-2,2′:6′,2″-terpyridine)ruthenium(II), Ru(tcterpy)(NCS)₃]²⁶ is usually used for achieving high conversion efficiency in DSCs.⁵ The energy difference between the LUMO of black dye and conduction band of TiO₂ is smaller than that between N719 dye and TiO₂ and the electron injection efficiency from black dye to TiO₂ is more sensitive to the surrounding environment than that from N719 dye.²⁷ Thus, black dye is a better-suited system for studying electron injection than N719 dye. In this study, we investigated the influence of the TiCl₄ treatment on nanocrystalline TiO₂ films in BCDSC using black dye as sensitizer.

Experimental

BCDSCs were fabricated using a two-electrode sandwich cell configuration. A nanocrystalline TiO₂ film on a glass substrate was prepared using a previously reported method.²⁸ The thickness of the nanocrystalline TiO₂ film was about 25 µm, measured with a profilometer (Tokyo Seimitu: Surfcom A). A back contact electrode (BCE) of titanium (Ti) was deposited onto the nanocrystalline TiO₂ film using vacuum deposition (Nihon Sinkuu: EVD-500A).²⁰ The electrode was immersed in a dye solution (2 × 10⁻⁴ M) of black dye with deoxycholic acid (DCA: 2 × 10⁻² M) in 1 : 1 acetonitrile to tert-butyl alcohol solution for 65 h at room temperature. A platinum-coated conducting glass was used as the counter electrode (CE). The two electrodes were separated by a Surlyn spacer (50 µm thick) and sealed by heating the polymer frame. The cell was filled with electrolyte (0.6 M dimethylpropyl-imidazolium iodide, 0.05 M I₂, 0.1 M LiI, and 0.5 M tert-butylpyridine in acetonitrile) using capillary action. Three samples were fabricated with each condition to confirm reproducibility.

TiCl₄ treatment was applied to prepared TiO₂ films following a previously reported method.²⁹ Sintered nanocrystalline TiO₂ films were treated with an aqueous solution of 50 mM TiCl₄ and stored in an oven at 70 °C for 20 min, rinsed with purified water, and sintered again at 500 °C for 1 h.

To characterize the amount of dye adsorbed onto nanocrystalline TiO₂ film near the glass/TiO₂ interface, a cw Nd:YAG laser at 532 nm was focused on the glass/TiO₂ interface to measure Raman scattering using Horiba Jobin-Yvon Raman system type LabRAM HR800 (depth resolution <8 µm). Electrochemical impedance spectroscopy (EIS) was recorded over a frequency range of 10⁻²–10⁶ Hz at 298 K under illumination at an open circuit voltage. A Xe lamp (Hamamatsu Co. L 2274) was used as the light source. The irradiated photon density was adjusted by changing the distance between the BCDSC and the Xe lamp, and was almost the same as that of AM 1.5 solar simulator light. The electrical impedance spectra were characterized using Z-View software (Auto Lab, ECO Chemie B.V.).

Transient absorption spectroscopy (TAS) measurement was carried out following a previously reported method.²⁵ The light source for picosecond pump–probe transient absorption measurements was a regenerative amplifier system consisting of a Ti:sapphire laser (800 nm wavelength, 150 fs FWHM pulse width, 1.0 mJ pulse⁻¹ intensity, 1 kHz repetition, Spectra Physics, Hurricane) combined with optical parametric amplifiers (Spectra Physics, OPA-800). For a pump pulse, the OPA output at 580 nm with an intensity of several microjoules per pulse was used; and for a probe pulse, the fundamental beam at 800 nm was used. The probe beam transmitted through the film sample was detected with a Si photodetector after passing through a monochromator (Acton Research, SpectraPro-150). For nanosecond TA measurements, a Nd³⁺:YAG laser (HOYA Continuum, Surelite II) was used as the pumping light source. The repetition rate of the laser was 10 Hz. The second-harmonic pulse (532 nm) was used for excitation of the dye-sensitized TiO₂ films. A halogen lamp (100 W) was used as the probe light source. The light transmitted through sample films was detected with a Si photodiode (Hamamatsu, S-1722) after being dispersed with a monochromator (Acton Research, SpectraPro-150). The photocurrent from the detector was amplified with an amplifier (NF Electronic Instruments, 5305). Signals were processed with a digital oscilloscope (Tektronix, TDS380). The DC offset of the photocurrent from the detector was subtracted using an electric frequency filter (NF Electronic Instruments, FV-628B), and therefore small absorbance changes (<10⁻⁵) could be detected. The time resolution of the system was about 5 µs. The intensity of the laser pulse was measured with a pyroelectric energy meter (Ophir, PE25-SH-V2). All measurements were carried out at 295 K. Samples were immersed in acetonitrile.

The current–voltage characteristics were measured using a previously reported method³⁰ with a solar simulator (AM 1.5, 100 mW cm⁻², WXS-155S-10: Wacom Denso Co., Japan). A black metal mask (0.2209 cm²) was attached to the solar cells in order to prevent irradiation with scattered light.

Results and discussion

Fig. 2 shows the I–V characteristics of the BCDSC with and without the TiCl₄ treatment. As shown, short circuit current density (J_SC) was improved from 14.1 mA cm⁻² to 18.1 mA cm⁻² by the TiCl₄ treatment with no change in both open circuit voltage (V_OC) and fill factor.

Fig. 2 Current–voltage characteristics of BCDSC with and without TiCl₄ treatment.

J _SC can be described as

J_SC = I₀Φ_LHΦ_INJΦ_COLL (1)

where I₀ is the incident photon intensity, Φ_LH is the light harvest efficiency, Φ_INJ is the electron injection efficiency from the sensitizers to the TiO₂ films, and Φ_COLL is the charge collection efficiency which describes both electron transport and electron lifetime within the nanocrystalline TiO₂ film. An overall energy conversion efficiency of 8.9% was achieved for the TiCl₄-treated BCDSC. This conversion efficiency is higher than that of BCDSC using N719 dye.²⁴ Because the transmittance of the substrate was not changed by the TiCl₄ treatment, I₀ is not affected by the TiCl₄ treatment of the nanocrystalline TiO₂ films. In order to understand the mechanism of the TiCl₄ treatment, detailed studies for three factors, Φ_LH, Φ_INJ, and Φ_COLL are required.

Φ _LH is strongly related to the dye adsorption properties of the nanocrystalline TiO₂ film. The estimated amount of dye adsorption on the nanocrystalline TiO₂ films with TiCl₄ treatment (4.1 × 10⁻⁸ mol cm⁻²) was similar to that without TiCl₄ (4.5 × 10⁻⁸ mol cm⁻²). That difference is much less than the difference in short circuit current between the samples with and without TiCl₄ treatment. The estimated amount of dye adsorption using black dye is less than that using N719 dye²³ because the deoxycholic acid in the black dye solution also adsorbs to the surface of the TiO₂ film. For further study, we evaluated the amount of dye near the glass substrate by using Raman spectroscopy under a confocal microscope. The interface between glass substrate and nanocrystalline TiO₂ film (glass/TiO₂ interface) is far from the nanocrystalline TiO₂ film surface, which contacts the dye solution directly. Therefore, it is difficult for the dye to penetrate the nanocrystalline TiO₂ film all the way to the glass/TiO₂ interface. It was expected that if TiCl₄ treatment greatly reduced the pore size of the TiO₂ film, then the amount of dye attaching onto TiO₂ near the glass/TiO₂ film interface with TiCl₄ treatment would be less than that without TiCl₄ film. Fig. 3 shows the Raman spectra of BCDSC near the glass/TiO₂ interface with and without TiCl₄ treatment. The three peaks in the 1400–1600 cm⁻¹ region are assigned to pyridine vibration³¹ of the black dye and the peaks at 114 cm⁻¹ and 145 cm⁻¹ are assigned to the I₃⁻ in the electrolyte solution and the TiO₂, respectively.³² The peak at 114 cm⁻¹ is smaller with the TiCl₄ treatment than without, indicating that the amount of I₃⁻ ions in the electrolyte that fills the pores in the nanocrystalline TiO₂ film decreased with TiC₄ treatment because of a decrease in pore size. The TiCl₄ treatment deposits extra TiO₂ material (film or particles) onto the surface of the nanocrystalline TiO₂ particles. Despite the decrease in pore size in the TiCl₄-treated film, the intensity of the spectra in the 1400–1600 cm⁻¹ range are very similar with and without TiCl₄ treatment in the BCDSC, which indicates that black dyes can readily penetrate the pores of the TiO₂ film all the way to the glass/TiO₂ interface in both TiCl₄-treated and untreated nanocrystalline TiO₂ films, and the amount of dye absorption was unchanged with TiCl₄ treatment even near the glass/TiO₂ interface. These results demonstrate that there is not a large change in the LHE with or without the TiCl₄ treatment and therefore is not the main reason for J_SC enhancement in the BCDSCs.

Fig. 3 Raman spectra of BCDSC with and without TiCl₄ treatment.

Transient absorption spectroscopy (TAS) is a powerful tool for the study of electron transfer dynamics between dye molecules and TiO₂nanoparticles.^7–9 Electron injection efficiency (Φ_INJ) from sensitizer dyes to TiO₂nanoparticles in DSCs can be estimated using nanosecond time-resolved TAS because the amplitude of the TA signal is proportional to the number of black dye cations generated after electron injection into TiO₂nanoparticles.²⁷Fig. 4(a) shows the transient absorption (TA) signal of black dyes adsorbed on nanocrystalline TiO₂ films with and without TiCl₄ treatment recorded at 800 nm, which is near the peak of the black dye cation at 750 nm.^24,33 The amplitude of these signals are similar between TiCl₄ treated and untreated nanocrystalline TiO₂ films, indicating that Φ_INJ does not change with the TiCl₄ treatment of nanocrystalline TiO₂ films.

Fig. 4 (a) Nanosecond and (b) picosecond transient absorption signals of black dye adsorbed onto nanocrystalline TiO₂ films with and without TiCl₄ treatment.

The electron injection dynamics are strongly affected by the energy gap difference between the LUMO in sensitizer dye and the edge of the conduction band in TiO₂. Thus, picosecond TAS can be used to prove the modification of the surface electronic state of TiO₂ by the TICl₄ treatment. Fig. 4(b) shows temporal change of the picosecond TA signals of black dye absorbed on nanocrystalline TiO₂ films with and without TiCl₄ treatment recorded at 800 nm. The TA signal with the TiCl₄ treatment is almost the same as that without the TiCl₄ treatment, indicating that electron injection dynamics did not change with the TiCl₄ treatment in spite of the small energy gap difference between the LUMO of black dye and the conduction band of TiO₂. It is thus supposed that the TiCl₄ treatment does not significantly influence the conduction band level of the TiO₂ particles. This idea is supported by Fig. 2 which shows that the TiCl₄ treatment does not change the V_OC determined by the energy gap between the conduction band level of the TiO₂ particles and redox level of electrolyte/hole transport material. According to eqn (1), it is clear that the increase in J_SC upon the TiCl₄ treatment is not due to the enhancement of Φ_LH and Φ_INJ but mainly due to Φ_COLL.

Fig. 5(a) shows the impedance spectra of BCDSC with and without TiCl₄ treatment in the complex plane. Z₁ and Z₂ are impedances related to charge-transfer processes at the surface of the CE and at the TiO₂/dye/electrolyte interface, respectively. Z₃ is related to redox diffusion within the electrolyte.^10–12 The resistances related to each impedance (Z₁, Z₂ and Z₃) of the BCDSC are 0.63, 2.80 and 1.05 Ω cm² without TiCl₄ treatment and 0.63, 3.16 and 1.16 Ω cm² with TiCl₄ treatment, respectively. In the inset, the impedance spectrum of the BCDSC without TiCl₄ treatment shows a flat region between Z₁ and Z₂, suggesting that an additional resistance is present between Z₁ and Z₂. Hoshikawa et al. have reported the dependence of sintering temperature of TiO₂ paste on impedance spectra.³⁴ According to their study, the impedance spectra between Z₁ and Z₂ increased with decreasing sintering temperature. In general, at a sintering temperature below 500 °C the electron transport in the TiO₂ film is reduced with decreasing sintering temperature due to remaining grain boundaries between TiO₂ particles. This new impedance element was therefore assigned to electron transport properties between TiO₂nanoparticles. The impedance spectra of the BCDSC with TiCl₄ treatment show no flat region between Z₁ and Z₂. It is thus supposed that because an extra layer of TiO₂ is added to the sintered TiO₂ particles, the TiCl₄ treatment improves electron transport properties of Φ_COLL in the nanocrytalline TiO₂ film, especially between TiO₂nanoparticles. This result is consistent with our previous study of BCDSC with N719 dye, which indicated that the electron diffusion coefficient improved with TiCl₄ treatment.²⁰

Fig. 5 Electrochemical impedance spectrum of BCDSC with and without TiCl₄ treatment under 100 mW cm⁻² illumination. (a) In complex plan (Cole–Cole plot). Inset: Electrochemical impedance spectrum between Z₁ and Z₂. (b) Imaginary part of electrochemical impedance spectrum as a function of frequency.

Fig. 5(b) shows the imaginary part of the electron impedance spectra of BCDSC with and without TiCl₄ treatment as a function of frequency. The TiCl₄ treatment shifts the mid-frequency peak (Z₂) to lower frequencies. That peak is associated with the charge recombination process between the nanocrystalline TiO₂ film and hole transport material at the TiO₂/dye/electrolyte interface. Kern et al. found that the mid frequency peak in the electrochemical impedance spectra is inversely proportional to the electron lifetime in the nanocrystalline TiO₂ film.¹⁰ It is supposed that electron lifetime increases due to TiCl₄ treatment. It is clear that the TiCl₄ treatment improved Φ_COLL because of an improvement in both electron transport and electron lifetime within the TiO₂ film.

Conclusion

We investigated the influence of TiCl₄ treatment on nanocrystalline TiO₂ films using back contact dye-sensitized solar cells with black dye. Raman spectroscopy revealed that the dye penetrates all the way thorough the nanocrytalline TiO₂ film and the amount of dye absorption was unchanged in spite of a decreased pore size in the TiO₂ film due to the TiCl₄ treatement. Although electron injection from the black dye is sensitive to electronic coupling and the surrounding environment, the electron injection efficiency and the electron injection dynamics with TiCl₄ treatment were almost the same as without TiCl₄ treatment. The decrease in the impedance related to electron transport between TiO₂ particles and an increase in electron lifetime within the nanocrystalline TiO₂ film clearly indicates that the main factor responsible for the enhancement of the short circuit current is the improvement of electron transport and electron lifetime in the nanocrystalline TiO₂ film.

References

1. B. O'Regan and M. Grätzel, Nature (London), 1991, 353, 737.
2. M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Müller, P. Liska, N. Vlachopoulos and M. Grätzel, J. Am. Chem. Soc., 1993, 115, 6382.
3. M. Grätzel, Nature (London), 2001, 414, 338.
4. M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni, G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Grätzel, J. Am. Chem. Soc., 2005, 127, 16835.
5. Y. Chiba, A. Islam, Y. Watanabe, R. Komiya, N. Koide and L. Han, Jpn. J. Appl. Phys., 2006, 45, L638–L640.
6. Y. Cao, Y. Bai, Q. Yu, Y. Cheng, S. Liu, D. Shi, F. Gao and P. Wang, J. Phys. Chem. C, 2009, 113, 6290–6297.
7. J. B. Asbury, E. Hao, Y. Wang, H. N. Ghosh and T. Lian, J. Phys. Chem. B, 2001, 105, 4545.
8. R. Katoh, A. Furube, A. Berzykin, H. Arakawa and M. Tachiya, Coord. Chem. Rev., 2004, 248, 1195.
9. J. R. Durrant, S. A. Haque and E. Palomares, Chem. Commun., 2006, 3279.
10. R. Kern, R. Sastrawan, J. Ferber, R. Stangl and J. Luther, Electrochim. Acta, 2002, 47, 4213.
11. Q. Wang, J. E. Moser and M. Grätzel, J. Phys. Chem. B, 2005, 109, 14945.
12. L. Han, N. Koide, Y. Chiba, A. Islam, R. Komiya, N. Fuke, A. Fukui and R. Yamanaka, Appl. Phys. Lett., 2005, 86, 213501.
13. L. M. Peter, J. Phys. Chem. C, 2007, 111, 6601.
14. K. Zhu, N. Kopidakis, N. R. Neale, J. van de Langemaat and A. J. Frank, J. Phys. Chem. B, 2006, 110, 25174.
15. S. Nakade, T. Kanzaki, Y. Wada and S. Yanagida, Langmuir, 2005, 21, 10803.
16. L. Y. Zeng, S. Y. Dai, K. J. Wang, X. Pan, C. W. Shi and L. Guo, Chin. Phys. Lett., 2004, 21, 1835–1837.
17. C. J. Barbe, F. Arendse, P. Comte, M. Jirousek, F. Lenzmann, V. Shklover and M. Grätzel, J. Am. Ceram. Soc., 1997, 80, 3157.
18. P. M. Sommeling, B. C. O'Regan, R. R. Haswell, H. J. P. Smit, N. J. Bakker, J. J. T. Smits, J. M. Kroon and J. A. M. van Roosmalen, J. Phys. Chem. B, 2006, 110, 19191.
19. B. C. O'Regan, J. R. Durrant, P. M. Sommeling and N. J. Bakker, J. Phys. Chem. B, 2007, 111, 14001.
20. N. Fuke, A. Fukui, A. Islam, R. Komiya, R. Yamanaka, L. Han and H. Harima, J. Appl. Phys., 2008, 104, 064307.
21. J. M. Kroon, N. J. Bakker, H. J. P. Smith, A. Hinsch, S. Hore, U. Würfel, R. Sastrawan, J. R. Durrant, E. Palomares, H. Pettersson, T. Gruszecki, J. Walter, K. Skupien and G. E. Tulloch, Prog. Photovolt: Res. Appl., 2007, 15, 1.
22. N. Fuke, A. Fukui, Y. Chiba, R. Komiya, R. Yamanaka and L. Han, Jpn. J. Appl. Phys., Part 2, 2007, 46(18), L420.
23. Y. Kashiwa, Y. Yoshida and S. Hayase, Appl. Phys. Lett., 2008, 92, 033308.
24. N. Fuke, A. Fukui, R. Komiya, A. Islam, Y. Chiba, M. Yanagida, R. Yamanaka and L. Han, Chem. Mater., 2008, 20, 4974.
25. U. Würfel, M. Peters and A. Hinsch, J. Phys. Chem. C, 2008, 112, 1711.
26. M. K. Nazeeruddin and M. Grätzel, Chem. Commun., 1997, 1705.
27. R. Katoh, A. Furube, M. Kasuya, N. Fuke, N. Koide and L. Han, J. Mater. Chem., 2007, 17, 3190.
28. P. Wang, S. M. Zakeeruddin, P. Comte, R. Charvet, R. Humphry-Baker and M. Grätzel, J. Phys. Chem. B, 2003, 107, 14336.
29. S. Ito, P. Liska, P. Comte, R. L. Charvet, P. Pechy, U. Bach, L. Schmidt-Menda, S. M. Zakeeruddin and M. Grätzel, Chem. Commun., 2005, 4351.
30. N. Koide and L. Han, Rev. Sci. Instrum., 2004, 75, 2828.
31. C. P. León, L. Kador, B. Peng and M. Thelakkat, J. Phys. Chem. B, 2006, 110, 8723.
32. C. Shi, S. Dai, K. Wang, X. Pan, F. Kong and L. Hu, Vib. Spectrosc., 2005, 39, 99.
33. C. Bauer, G. Boschloo, E. Mukhtar and A. Hagfeldt, J. Phys. Chem. B, 2002, 106, 12693.
34. T. Hoshikawa, M. Yamada, R. Kikuchi and K. Eguchi, J. Electrochem. Soc., 2005, 152, E68.

---