Self-assembled anodic TiO₂ nanotube arrays: electrolyte properties and their effect on resulting morphologies

Abstract

Self-assembled TiO₂ nanotube arrays fabricated by electrochemical anodization of titanium are of great interest having been successfully used in many applications including gas sensing, water photoelectrolysis, drug delivery and photovoltaics. Nanotube array synthesis techniques have been studied and developed through several electrolyte systems, however, the key parameters controlling self-organization of the nanotubes have remained unclear. Herein we examine nanotube array morphological growth parameters as dependent upon electrolyte conductivity and titanium concentration. Electrolyte properties establish a regime wherein the TiO₂ nanotube arrays self-assemble. Nanotube morphological parameters, including pore diameter, wall thickness and tube-to-tube spacing, are all found to increase with electrolyte conductivity. Using diethylene glycol (DEG) based electrolytes as a model, we detail how manipulation of electrolyte conductivity enables control of nanotube array morphological features.

---

Introduction

Vertically oriented TiO₂ nanotube arrays fabricated by anodization of titanium in fluoride-based electrolytes have been found to posses useful and unique properties, finding great utility in a variety of applications including sensors,^1–3dye sensitized solar cells,^4–9water photoelectrolysis,^10,11 photocatalytic reduction of CO₂ under sunlight,^12,13 and biomedical applications.^14–16 In the synthesis of TiO₂ nanotube array films it is important to achieve specific nanotube array morphological features, including pore size, length, wall thickness, and tube-to-tube spacing for enhanced device performance.^17–22 The uniformity of the self-assembled TiO₂ nanotube arrays has been found to be dependent on synthesis parameters that include the anodization voltage,^23–25 temperature,^23–25titanium purity,²⁶pore imprinting,²⁷ and the re-anodization of previously anodized samples.^28–32 Tube geometry can vary considerably with the electrolyte chemistry in which the tubes are formed.^13,17,20,33 For example, hexagonal close-packed TiO₂ nanotube arrays can be obtained in ethylene glycol electrolyte,^{17,18,23,24,34,35} while a wide variation of nanotube-to-nanotube spacing can be achieved in diethylene glycol (DEG).^21,36 In fluoride containing electrolytes electrochemical anodization leads to the Ti ion dissolving in the electrolyte as a stable form of [TiF₆]²⁻ complex;³⁷ longer anodization times lead to higher electrolyte conductivities.²⁰ Using an electrolyte of HF containing diethylene glycol as our model system, we investigate electrolyte conductivity and titanium concentration as a function of anodizing parameters including voltage, time, HF concentration, inter-electrode spacing, and solvent additives. Our study elucidates the dependence of the electrolyte conductivity on the titanium concentration, and electrolyte effect on the morphological features of the resulting nanotubes.

Experiment

Titanium foils (0.25 mm, 99.7%, Sigma-Aldrich) were cleaned with acetone, soap and isopropanol prior to anodization. The Ti foils were subjected to potentiostatic anodization, at room temperature of about 23 °C, in a two-electrode electrochemical cell with platinum foil as the counter electrode. In our experiments the sample area exposed to the electrolyte, diethylene glycol (DEG, 99.7%, Sigma-Aldrich) in combination with hydrofluoric acid (HF, 48% solution, Merck), was generally fixed at 3.0 cm² with a few noted 5.0 cm² samples. Solvent additives include formamide (FA, 99.5%, Sigma-Aldrich), N-methylformamide (NMF, Sigma-Aldrich), hexamethylphosphoramide (HMPA, 99%, Aldrich), dimethyl sulfoxide (DMSO, 99.9%, Sigma-Aldrich), acetyl acetone (Acac, 99%, Alfa Aesar), acetonitrile (AN, 99.8%, anhydrous, Sigma-Aldrich), formic acid (HCOOH, 96%, Sigma-Aldrich), acetic acid (HOAc, glacial, Baker Analyzed), tetrahydrofuran (THF, 99%, Sigma-Aldrich), propylene carbonate (PC, 99.7%, Sigma-Aldrich), methyl acetate (MA, 99%, Sigma-Aldrich), ethanol (EtOH, anhydrous, Sigma-Aldrich), and de-ionized water. The total volume of all electrolyte baths in this study was fixed at 50 ml. Conductivity of the as-anodized electrolytes was measured at 23 °C using a conductivity meter (YSI 3200, Cole-Parmer). The titanium concentration of the electrolytes used for anodization was examined by inductively coupled plasma atomic emission spectrometry (ICP-AES, Perkin-Elmer Optima 5300 ICP). Morphologies of the anodic titania nanotube array films were characterized using a field emission scanning electron microscope (FE-SEM, Leo 1530). Crystallization of the as-anodized films was investigated by glancing-angle X-ray diffractometer (GAXRD, Scintag, Inc.).

Results and discussion

Fig. 1a shows the anodization current–time behavior of titanium foils anodized in 2% HF–DEG electrolytes mixed with different additives. The continuously increasing current density as a function of anodization time is a characteristic of Ti anodization in a DEG based electrolyte, dramatically different from the anodization curves for anodization of Ti observed in either aqueous or other non-aqueous electrolytes.^20,23,32 The magnitude of the anodization current generally reflects oxide formation, which is a result of various O-containing anions in the electrolyte transporting through the oxide layer to interact with Ti⁴⁺ ions at the metal/oxide interface. The oxide growth rate is limited by ion diffusion in the electrolyte;^23,38solvent properties such as viscosity, conductivity, and molecular hydrogen bonding³⁹ are expected to play important roles in determining ionic movement in the electrolyte to the oxide layer, and hence determining characteristics of oxide growth and the anodization current.

Fig. 1 Current density–time behavior of titanium foils anodized at 60 V for 24 h in 2% HF–DEG-based electrolytes containing (a) 5% THF, DMSO, PC, HMPA, NMF, H₂O, FA; 1% formic acid, or 1% acetic acid. (b) 3%, 5%, and 9% H₂O, and 3% and 5% FA.

Fig. 1b shows the anodization current behavior in 2% HF–DEG electrolytes containing 3% and 5% formamide, and 3%, 5%, and 9% water. The current density changes significantly with the addition of highly polar solvents. For the DEG–FA bath, the anodization current increases with time reaching a maximum value before rapidly reducing after 5 h. It is generally considered that the decrease in current is due to nanotube formation, and electrolyte conductivity.^20,21 The growth rate of the titania film is limited by the migration of Ti⁴⁺ and O²⁻ ions through the oxide film.²³ By high field conduction theory,⁴⁰ the current density j is related to the voltage V drop across the barrier oxide layer by

j = A exp (BV/d)

where A and B are material dependent constants for a given temperature, d is the overall barrier layer thickness, and V/d represents the effective average field across the barrier oxide layer.

When the ohmic resistance of the electrolyte dominates that of the charge-transfer resistance at the oxide surface the overall rate of oxide formation is limited by the diffusion of ions towards the electrode surfaces. As seen in Fig. 1a and b, the anodization current amplitude is dependent upon the electrolyte conductivity, with the higher additive content resulting in higher current densities. The high electrolyte conductivity could possibly enhance the oxide dissolution process; the dissolution of the barrier oxide layer leads to larger anodization currents, facilitating ionic transport and faster nanotube growth.^32,41 For a 24 h anodization the length of nanotube array samples grown in DEG–FA baths is ∼8 µm, while the average tube length obtained from the baths containing other additives is ∼3 µm. The growth rate of nanotube films in FA containing baths is ∼0.3 µm h⁻¹, slightly higher than the average rate of other additive baths at ∼0.1 µm h⁻¹. However, the growth rates for DEG based electrolytes are much smaller than those obtained in dimethyl sulfoxide (∼1 µm h⁻¹),²⁰formamide (∼2 µm h⁻¹)³² and ethylene glycol (∼15 µm h⁻¹)²³ based electrolytes, clearly indicating that the electrolyte medium plays a key role in determining the overall reaction rate for nanotube growth.^21,23,32

For a given anodization voltage, the conductivity and the titanium concentration of the electrolyte increase with larger water content⁴² as shown in Fig. 2a. Fig. 2b shows the conductivities plotted against titanium concentrations, with a linear relation clearly seen. The results are in a good agreement with eqn (1), indicating that the electrolyte conductivity κ is proportional to the concentrations c of the constituent ions (ion i) for dilute electrolyte solutions:⁴³

κ = ∑|Z_i|Fu_ic_i = ∑λ_ic_i (1)

where Z_i is the ion i charge number, F is Faraday constant, u_i is the electric mobility of ion i, c_i is the proportionality constant and λ_i is the ionic conductivity or the molar conductivity of ion i.

Fig. 2 (a) Electrolyte conductivity and titanium concentration plotted against water content, measured from 2% HF–(3, 5, 7, and 9%) H₂O–DEG electrolytes after anodizing at 60 V for 24 h. (b) A linear plot of electrolyte conductivity and titanium concentration using data from (a). (c) Electrolyte conductivity vs.titanium concentration obtained from different anodizing conditions by varying anodization time (line i, 2–70 h, 1% HF–DEG, 60 V), varying voltage (line ii, 40–80 V, 2% HF–DEG, 24 h and line iii, 40–80 V, 2% HF–5% H₂O–DEG, 24 h), varying interelectrode spacing (line iv, 0.5–4.5 cm, 2% HF–DEG, 60 V, 24 h), and varying water content (line v, 0–9%, 2% HF–DEG, 60 V, 24 h). All data from lines (i)–(v) were obtained from using Ti area of 3.0 cm². (Line vi) Data obtained from anodizing in 2% HF–(3–9%) H₂O–DEG electrolytes (60 V and 24 h) using a larger sample area of 5.0 cm².

A linear conductivity–titanium concentration relation was observed under a variety of anodizing conditions, see Fig. 2c. The use of longer anodization durations (line i), higher voltages (line ii), and smaller electrode spacings (line iv) leads to increased titanium dissolution and increased conductivity. A significant change in these relationships was observed with the addition of water to the electrolyte. The slopes of line iii and line v (with water) are smaller than that of line ii (without water), suggesting that water significantly increased the electrolyte conductivity. Using a fixed Ti sample area of 3.0 cm², we found a nominal electrolyte conductivity of 250 µS cm⁻¹ and titanium concentration values of ∼1200 ppm. Higher water content results in decreased nanotube wall thickness and larger pore size due to the greater loss of titanium species through dissolution and/or charge consuming side reactions.³³

A linear relation between conductivity and titanium concentration was observed for a variety of different electrolyte solvent additives as demonstrated in Fig. 3a. Group 1, with conductivity values of approximately 100 µS cm⁻¹, were obtained from DEG (relative permittivity ε = 31.69) electrolytes containing THF (ε = 7.58), HMPA (ε = 29.6), PC (ε = 64.92), EtOH (ε = 24.6), HOAc (ε = 6.19), MA (ε = 6.68), HCOOH (ε = 58.5), DMSO (ε = 46.5), Acac (ε = 25.7), AN (ε = 35.9), and NMF (ε = 182.4). Group 2 and Group 3 additives, respectively water (ε = 78.39) and FA (ε = 111), had higher conductivities. It appears that the increased conductivity is due to the increased number of titanium-complex anions in the electrolyte, including Ti⁴⁺ migration at the metal/oxide interface and the chemical dissolution (oxide etching) of the TiO₂ wall.⁴³

Fig. 3 (a) Linear relationship between electrolyte conductivity and titanium concentration of the as-anodized DEG–2% HF electrolytes mixed with different solvent additives. The green solid circle line represents the linear data of 5% additives including THF, HMPA, PC, EtOH, MA, DMSO, Acac, AN, and NMF, and 1% of HCOOH and HOAc. The red half-filled square line represents the data of using water with varied content from 3% to 9%. The filled triangle line represents the linear data of FA additive with varied content from 1.5% to 5%. (b) Schematic representation of porous structure and separated nanotube array structure.

For Group 1, the relative permittivity showed no significant effect on the increased conductivity, but the conductivity was found to be directly dependent upon the increased titanium concentration. The electrolyte conductivity values in Group 1 are lower than those in Group 2 and Group 3, a behavior we attribute to the formation of ion pairs in these weakly polar solvent electrolyte mixtures.⁴⁴NMF, a solvent of high ε (182.4),⁴³ is also classified in Group 1, however, ion mobility in this electrolyte is most likely hindered by the strong self-bonding network of the NMF molecules resulting in low conductivity (Fig. 1a), and correspondingly low anodization current.⁴²

For Group 2, the electrolyte conductivity and the titanium concentration increase as a function of water content. We hypothesize that more dissociated ions are available in the electrolyte with higher water content, leading to higher conductivity. The higher conductivity, due to more free ions in the electrolyte, in turn induces more charges on the oxide layer improving Ti⁴⁺ ion extraction.³² Since a larger amount of water leads to a larger proton concentration in the electrolyte, one may assume that the oxide dissolution is enhanced leading to a greater titanium concentration in the electrolyte according to:⁴⁴

TiO₂ + 4H⁺ + 6F⁻ → TiF₆²⁻ + 2H₂O

The idea of ion dissociation dependent upon relative permittivity can also be applied to Group 3. Mixing FA (ε = 111), which is substantially more polar than water (ε = 78.39), into DEG should allow more free ions leading to increased electrolyte conductivity. However, the titanium concentrations measured in the anodized electrolytes of Group 3 are considerably lower than those of Group 1 and Group 2, suggesting that the titanium concentration is not the main factor that governs the electrolyte conductivity. We speculate that the high conductivity is probably due to the enhanced effect of ionophore dissociation.⁴⁴ The ionophores—the ionic forms arranged in crystalline state—may dissociate almost completely into free ions resulting in high electrolyte conductivity.

The electrolyte conductivity is found to have a significant effect on the degree of nanotube self-ordering, with higher electrolyte conductivities resulting in more ordered, well separated nanotube arrays, see Fig. 3a. For Group 1, conductivity range < 100 µS cm⁻¹, disordered nanotube arrays were commonly observed, see Fig. 4b. The lowest conductivity electrolytes, containing THF, HMPA, and PC provided an alumina-like nanoporous structure, see Fig. 4a. For Group 2 and Group 3, with electrolyte conductivities >100 µS cm⁻¹, more uniformly aligned and well-separated nanotube arrays were obtained, see Fig. 4c. The uniform well-aligned nanotube arrays can be prepared within limited processing windows, satisfying self-ordering conditions. The restriction is due to the conductivity of the electrolyte which strongly affects the movement of ionic species for oxide formation, while thinner oxide layers facilitate hydroxyl ion injection through the oxide layer allowing further movement of the metal/oxide interface into the metal.⁴⁵ Due to the high conductivities of Group 2 and Group 3 electrolytes the pore growth speed is sufficiently high, over each nanotube, to allow homogeneous pore deepening at the metal/oxide interface with uniform expansion of the oxide walls. In low conductivity electrolytes oxide growth is limited by ionic movement into the oxide, leading to inconsistent oxide growth and hence irregular pore formation.^23,41

Fig. 4 FE-SEM images showing top view (left) and cross-sectional view (right) of nanotubes formed in 2% HF–DEG electrolytes containing (a) 5% PC, (b) 1% formic acid, and (c) 5% H₂O, at 60 V for 24 h anodization; the inset (in c, left) is a top view image showing the bottom part of the arrays obtained after the top part of nanotube arrays was mechanically removed.

Nanotube morphological parameters such as pore size, wall thickness, and inter-tubular spacing as a function of electrolyte conductivity and anodization voltage are shown in Fig. 5a–c. We measured pore sizes close to the bottom of the nanotube arrays since chemical dissolution at the top results in the nanotube having a tapered or conical structure.^33,46 At a voltage of 40 V electrolyte conductivity showed only a slight effect on tube dimensions. At 80 V films anodized in the electrolytes containing highly polar additives were completely dissolved within 24 h. Growth at the pore bottom is determined by ionic transport of the metal cations and oxygen anions through the oxide layer either at the metal/oxide or oxide/electrolyte interfaces, as well as by the diffusion of ionic species in the electrolytes such as F⁻ and TiF₆²⁻.^29,47 The mass transport of the anionic species driven by the strong electric field accelerates the rate of oxide formation as well as dissolution at the pore bottom,^41,47 resulting in larger wall thicknesses and pore sizes as seen in Fig. 5a and b. Under high anodization voltages, the stronger electric field and local-heating-enhanced dehydration of the oxide between pores lead to nanotube separation, see Fig. 5c.⁴⁸ At lower voltages the driving force for oxide dissolution is small, hence only a small gap between pores can be obtained. Dissolution of the oxide layer between neighboring pores is strictly controlled by the field-assisted mass transport at the oxide/electrolyte interface.

Fig. 5 Nanotube morphological parameters including pore size (a), wall thickness (b), and tube-to-tube spacing (c) plotted against electrolyte conductivity, measured from the nanotube arrays anodized in 2% HF–DEG electrolytes, 24 h and using different voltages.

The anodically grown titania nanotube arrays are amorphous. Crystallization of the nanotube arrays is critical for their device application, enabling facile charge transport.^8,49 High temperature annealing in oxygen induces crystallization, with anatase phase obtained at 280 °C.⁵⁰ A previous study of ours reported that as-anodized nanotube arrays with partially crystalline anatase phase could be obtained using DEG based electrolytes at room temperature.²¹ Partial crystallization of the nanotube array films could also be seen in this study. GAXRD spectra for the nanotube array samples fabricated at room temperature in DEG-based electrolytes containing different additives are shown in Fig. 6a. Crystalline phases of small intensity could be observed for the DEG-based electrolytes containing THF, HMPA, and PC, with relatively low conductivity electrolytes (∼50 µS cm⁻¹). The magnified intensity patterns of the crystallized peaks are shown in Fig. 6b. Surprisingly, the peaks between 20° and 30° are most closely ascribed to the appearance of crystalline TiO (PDF 00-009-0240). It is possible that the phase transformation of these samples occurs in low conductivity electrolytes where the nucleation growth process is comparatively slow with, ultimately, the ability to obtain a high degree of in situcrystallization limited by the nucleation-growth kinetics.^39,51

Fig. 6 (a) GAXRD patterns of as-anodized nanotube array films formed in 2% HF–DEG electrolytes containing 5% HMPA, THF, PC, NMF, AN, Acac, DMSO, H₂O, and FA, and 1% HCOOH; all anodizations were performed at 60 V for 24 h. (b) Magnified GAXRD patterns of nanotube array samples obtained when using HMPA, THF, and PC as additives, with the anodizing conditions identical to (a).

Conclusions

In summary, we believe our model study using DEG based electrolytes provides needed insights to the formation of anodic titania nanotube arrays, elucidating electrolyte property–nanotube structure relationships. We anticipate that the results of our study will be applicable to anodization using other non-aqueous electrolytes. We find that the anodization current and its behavior are strongly affected by addition of the relatively high polar solvent additives such as water and formamide to the DEG-based electrolytes. With the exception of formamide, addition of highly polar solvents leads to high electrolyte conductivity that is a consequence of the high titanium concentration dissolving in the anodized electrolyte. A self-ordering regime as a function of electrolyte conductivity is established over a wide range of nanotube growth architectures, from alumina-like nanoporous structure to disordered and well-ordered nanotube arrays. Partial crystallization of nanotube array film during its in situgrowth is obtained with low conductivity (∼50 µS cm⁻¹) electrolytes.

Acknowledgements

The authors thank Dr Oomman K. Varghese for helpful discussions and Dr Henry Gong for ICP-AES analysis. S. Yoriya acknowledges the scholarship support under the Royal Thai Government provided by the National Metal and Materials Technology Center (MTEC), the National Science and Technology Development Agency (NSTDA), Thailand. Partial support of this work through the Department of Energy, Grant Number DE-FG36-08GO18074, is gratefully acknowledged.

References

1. O. K. Varghese, D. W. Gong, M. Paulose, K. G. Ong and C. A. Grimes, Sens. Actuators, B, 2003, 93, 338.
2. O. K. Varghese, G. K. Mor, C. A. Grimes, M. Paulose and N. Mukherjee, J. Nanosci. Nanotechnol., 2004, 4, 733.
3. S. Yoriya, H. E. Prakasam, O. K. Varghese, K. Shankar, M. Paulose, G. K. Mor, T. J. Latempa and C. A. Grimes, Sens. Lett., 2006, 4, 334.
4. G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese and C. A. Grimes, Appl. Phys. Lett., 2007, 91, 152111.
5. M. Paulose, K. Shankar, O. K. Varghese, G. K. Mor, B. Hardin and C. A. Grimes, Nanotechnology, 2006, 17, 1446.
6. K. Shankar, J. Bandara, M. Paulose, H. Wietasch, O. K. Varghese, G. K. Mor, T. J. LaTempa, M. Thelakkat and C. A. Grimes, Nano Lett., 2008, 8, 1654.
7. K. Shankar, G. K. Mor, H. E. Prakasam, O. K. Varghese and C. A. Grimes, Langmuir, 2007, 23, 12445.
8. O. K. Varghese, M. Paulose and C. A. Grimes, Nat. Nanotechnol., 2009, 4, 592.
9. K. Zhu, N. R. Neale, A. Miedaner and A. J. Frank, Nano Lett., 2007, 7, 69.
10. G. K. Mor, K. Shankar, O. K. Varghese and C. A. Grimes, J. Mater. Res., 2004, 19, 2989.
11. K. Shankar, G. K. Mor, H. E. Prakasam, S. Yoriya, M. Paulose, O. K. Varghese and C. A. Grimes, Nanotechnology, 2007, 18, 065707 , 11 pages.
12. O. K. Varghese, M. Paulose, T. J. LaTempa and C. A. Grimes, Nano Lett., 2009, 9, 731.
13. C. A. Grimes and G. K. Mor, TiO₂ Nanotube Arrays: Synthesis, Properties, and Applications, Springer, New York, 2009.
14. K. C. Popat, M. Eltgroth, T. J. La Tempa, C. A. Grimes and T. A. Desai, Small, 2007, 3, 1878.
15. K. C. Popat, L. Leoni, C. A. Grimes and T. A. Desai, Biomaterials, 2007, 28, 3188.
16. S. C. Roy, M. Paulose and C. A. Grimes, Biomaterials, 2007, 28, 4667.
17. M. Paulose, K. Shankar, S. Yoriya, H. E. Prakasam, O. K. Varghese, G. K. Mor, T. A. Latempa, A. Fitzgerald and C. A. Grimes, J. Phys. Chem. B, 2006, 110, 16179.
18. M. Paulose, H. E. Prakasam, O. K. Varghese, L. Peng, K. C. Popat, G. K. Mor, T. A. Desai and C. A. Grimes, J. Phys. Chem. C, 2007, 111, 14992.
19. M. Paulose, L. Peng, K. C. Popat, O. K. Varghese, T. J. LaTempa, N. Z. Bao, T. A. Desai and C. A. Grimes, J. Membr. Sci., 2008, 319, 199.
20. S. Yoriya, M. Paulose, O. K. Varghese, G. K. Mor and C. A. Grimes, J. Phys. Chem. C, 2007, 111, 13770.
21. S. Yoriya, G. K. Mor, S. Sharma and C. A. Grimes, J. Mater. Chem., 2008, 18, 3332.
22. Y. Alivov, M. Pandikunta, S. Nikishin and Z. Y. Fan, Nanotechnology, 2009, 20, 225602.
23. H. E. Prakasam, K. Shankar, M. Paulose, O. K. Varghese and C. A. Grimes, J. Phys. Chem. C, 2007, 111, 7235.
24. Z. B. Xie, S. Adams, D. J. Blackwood and J. Wang, Nanotechnology, 2008, 405701.
25. J. Wang and Z. Lin, J. Phys. Chem. C, 2009, 113, 4026.
26. N. K. Kuromoto, R. A. Simão and G. A. Soares, Mater. Charact., 2007, 2007, 114.
27. X. D. Wang, E. Graugnard, J. S. King, Z. L. Wang and C. J. Summers, Nano Lett., 2004, 4, 2223.
28. S. Li, G. Zhang, D. Guo, L. Yu and W. Zhang, J. Phys. Chem. C, 2009, 113, 12759.
29. O. Jessensky, F. Muller and U. Gosele, Appl. Phys. Lett., 1998, 72, 1173.
30. H. Masuda, H. Yamada, M. Satoh, H. Asoh, M. Nakao and T. Tamamura, Appl. Phys. Lett., 1997, 71, 2770.
31. Y. Shin and S. Lee, Nano Lett., 2008, 8, 3171.
32. K. Shankar, G. K. Mor, A. Fitzgerald and C. A. Grimes, J. Phys. Chem. C, 2007, 111, 21.
33. S. Berger, J. Kunze, P. Schmuki, A. T. Valota, D. J. LeClere, P. Skeldon and G. E. Thompson, J. Electrochem. Soc., 2010, 157, C18.
34. D. Wang, T. Hu, L. Hu, B. Yu, Y. Xia, F. Zhou and W. Liu, Adv. Funct. Mater., 2009, 19, 1.
35. Q. Chen, D. Xu, Z. Wu and Z. Liu, Nanotechnology, 2008, 19, 365708 , 5pages.
36. S. Yoriya and C. A. Grimes, Langmuir, 2010, 26, 417.
37. G. K. Mor, O. K. Varghese, M. Paulose, K. Shankar and C. A. Grimes, Sol. Energy Mater. Sol. Cells, 2006, 90, 2011.
38. W. Lee, R. Ji, U. Gosele and K. Nielsch, Nat. Mater., 2006, 5, 741.
39. C. M. Kinart, A. Ćwiklińska, M. Maja and W. J. Kinart, Fluid Phase Equilib., 2007, 262, 244.
40. G. E. Thompson and G. C. Wood, Nature, 1981, 290, 230.
41. Z. Su and W. Zhou, Adv. Mater., 2008, 20, 3663.
42. B. Melody, T. Kinard and P. Lessner, Electrochem. Solid-State Lett., 1998, 1, 126.
43. K. Izutsu, Electrochemistry in Nonaqueous Solutions, Wiley-VCH, New York, 2002.
44. J. R. Chipperfield, Non-Aqueous Solvents, OUP, New York, 1999.
45. Y. M. Li and L. Young, J. Electrochem. Soc., 2001, 148, B337.
46. G. K. Mor, O. K. Varghese, M. Paulose, N. Mukherjee and C. A. Grimes, J. Mater. Res., 2003, 18, 2588.
47. J. W. Diggle, T. C. Downie and C. W. Goulding, Chem. Rev., 1969, 69, 365.
48. Z. Su and W. Zhou, J. Mater. Chem., 2009, 19, 2301.
49. B.-Y. Yu, A. Tsai, S.-P. Tsai, K.-T. Wong, Y. Yang, C.-W. Chu and J.-J. Shyue, Nanotechnology, 2008, 19, 255202 , 5 pages.
50. O. K. Varghese, D. W. Gong, M. Paulose, C. A. Grimes and E. C. Dickey, J. Mater. Res., 2003, 18, 156.
51. U. Hasse, S. Fletcher and F. Scholz, J. Solid State Electrochem., 2006, 10, 833.

---