Ni–P synergetic deposition: electrochemically deposited highly active Ni as a catalyst for chemical deposition

Abstract

The nucleation mechanism of the preparation of a Ni–P deposit by the synergetic effect of electrochemical deposition and chemical deposition was studied using an electrochemical analysis method. Cathode polarization and chronoamperometry were performed to explore the effect of the reducing reaction (chemical deposition) on the electrochemical Ni deposition. It was found that the addition of reducing agent to the solution induced depolarization phenomena and decreased or removed the double electric layer capacitance. The chronoamperometry experimental data were analyzed and fitted according to the Scharifker–Hills rule to discover the difference in mechanism between the Ni nucleation and Ni–P nucleation. The morphology of Ni–P and the pure Ni deposits was measured by TEM and SEM demonstrating the ball-like particle shape of Ni–P and flat pure Ni deposits. Based on these experimental results, the nucleation mechanism was illustrated in which the active Ni atoms, produced by electrochemical deposition, act as a catalyst for the reduction of the nickel ions with the reducing agent NaH₂PO₂ on the cathode surface. The occurrence of the reducing reaction of the nickel ions with NaH₂PO₂ hindered the in-plane growth of crystalline nickel particles, generating the ball-like particle shape of the deposit. This work provides researchers new insights into the nucleation and growth process of synergetic deposition and offers novel ideas for the fabrication of functional materials by liquid deposition.

---

1. Introduction

Electrochemical deposition and chemical deposition, as two kinds of commonly used liquid deposition methods, can fabricate many kinds of metallic film such as Ni, Cu, Zn and Cr films, which display high brightness, magnetic properties, and excellent anti-corrosion properties and are widely used in daily life and scientific research.^1,2 Most research on liquid deposition focuses on smooth and dense films which have been already used in industrial fields. Due to an increasingly high quality of life, functional films like bio-inspired films with special wettability³ and transparent conductor films,⁴ and novel techniques like atomic layer deposition^5,6 and the synergetic deposition using electrochemical and chemical deposition^7,8 increasingly attract scientists’ attention. By controlling the atomic layer epitaxial growth and redundant deposition, monolayer films^6,9 have been prepared by electrodeposition due to facile preparation and high productivity. Accordingly, superhydrophobic materials,¹⁰ and porous materials for fuel cells,^11,12 oil–water separation,¹³ and anti-fouling¹⁴ films, and some other kinds of functional material^15,16 have been fabricated by electrodeposition or chemical deposition. With increasing emphasis on the deposition of ultrathin films^5,6,9,17 and nanostructures,^8,18 it is critically important to advance the scientific understanding of the kinetics and mechanisms of nucleation and the growth process.

The exploration of the nucleation mechanism in different deposition processes has drawn interest from many researchers in the underlying novel discoveries. The mechanism of the metal deposition process is widely researched^19,20 and the island growth mechanism of electrodeposition has been reviewed.¹⁹ The Scharifker–Hills rule²¹ is one of the most accepted models due to its high consistency between experimental data and fitting results. This theory indicates that under diffusion control, three-dimensional multiple nucleation can be divided into instantaneous and progressive processes.²¹ There are some theories related to the formation of patterned morphologies such as the unstable growth theory.^22–24 The crystallization process begins with nucleation, which plays a critical role in determining the structure and size distribution of the crystals.²⁵ For example, our group^8,18 fabricated micro/nanostructured flower-like morphologies which possessed tailored hydrophobic properties using the synergetic effect of electrochemical and chemical deposition. Herein, the electrochemical method is introduced to discuss the underlying nucleation mechanism. The initial stage of Ni–P electrodeposition is studied²⁶ where the P element is introduced by the addition of H₃PO₃ and H₃PO₄ which do not have any reducing power. The morphology difference between the pure electrochemical deposition and synergetic deposition is caused by the catalytic reduction nickel ions in solution using the reducing agent NaH₂PO₂ yielding electro-deposited active Ni atoms. This work is devoted to revealing the nature of the particle morphology produced via the synergy of electrochemical and chemical deposition, which can promote future work on functional film preparation and provide insight on how to grow nanoparticle materials.

2. Experimental

2.1 Preparation of pure Ni and Ni–P by synergetic deposition in solution

The solution preparation has been reported elsewhere.¹⁸ The pure Ni electrolyte bath contains 0.125 M NiSO₄·6H₂O, 0.05 M citric acid, 1.2 g L⁻¹ sodium dodecyl sulfate (SDS), and 0.075 g L⁻¹ butynediol (C₄H₆O₂). The concentration of Na₂SO₄ was adjusted to 0.3 M to enhance ionic strength of the electrolyte. The pH value of the electrolyte was adjusted to 5.5 using aqueous ammonia. Different amounts (0.01 M, 0.02 M, 0.03 M, 0.04 M, and 0.05 M) of NaH₂PO₂ (P) were added to the solution to investigate the influence of the reducing agent on the nucleation mechanism of Ni.

2.2 Electrochemical tests

All electrochemical tests were performed using an electrochemical workstation (Modulab ECS, Solartron Analytical Ltd, UK) with a traditional three-electrode system. A low-alloy steel sheet, a Pt sheet and an Ag/AgCl (saturated KCl) electrode were used as the working, counter and reference electrodes, respectively. The area of the working electrode is confined to 1 cm². Prior to electrochemical tests, the low-alloy steel sheets were polished and washed with deionized (DI) water. The Ag/AgCl reference electrode was connected to the electrochemical system through a Luggin capillary filled with saturated KCl solution. The system’s open circuit potential (OCP) was tested for 100 s to confirm the stability of the electrolyte. Cathode polarization tests (5 mV s⁻¹, from 0 V to −2 V vs. OCP) were performed to study the effect of reducing agent on the pure Ni deposition. Potentiostatic deposition was performed at −0.2 V vs. OCP for 100 s. Films for different deposition times were prepared in a two-electrode electrochemical cell with the same electrolyte described above. The electrochemical cell was placed in a heated water bath and constantly stirred with a magnetic stirring bar. All of the above tests were performed at a solution temperature of 70 °C. The schematic diagram of the experimental set-up is shown in ESI 1.†

2.3 Characterization

The surface morphologies of the films were examined using scanning electron microscopy (SEM, FEI Quanta 250 FEG, USA), and the chemical composition and elemental mapping of the films were detected by energy dispersive X-ray (EDX) spectroscopy. Transmission electron microscopy (TEM, FEI TF20, USA) and high resolution transmission electron microscopy (HRTEM) were used to observe the micro-morphologies, and the deposit crystal features were characterized by selected area electron diffraction (SAED). Water contact angles of the patterned films were measured at room temperature with a volume of 2 µL using an optical contact angle meter (OCA20, Germany).

3. Results and discussion

The electrochemical analytical methods were applied on low-alloy steel substrates with an exposed area of 1 cm². Cathode polarization test results are shown in Fig. 1. At a low current density, the open circuit potential (OCP, which could be observed and is equal to the value at the beginning of the linear sweep voltammetry experiment) was stable and showed no difference between the pure Ni and Ni–P polarization in the system, which was about −0.6 V vs. the reference. With the increase in applied voltage, the current density of the Ni–P system increased rapidly, in comparison with the hysteresis increase of the pure Ni system. This means that the addition of the reducing agent NaH₂PO₂ to the solution generated depolarization phenomena. These polarization phenomena could be classified into cathode polarization and anode polarization. The cathode polarization has been widely used in electrodeposition by adding an active substance to the electrolyte. These trace addition agents like saccharin, butynediol, and formaldehyde which could generate the cathode polarization phenomenon, had a huge effect on brightness and smoothness of the deposited films. Some of these additives were reducing agents such as formaldehyde. In this work, NaH₂PO₂, a kind of reducing agent, showed the opposite effect on the deposition–depolarization phenomena. The depolarization phenomenon reduced the reaction impedance of the electrochemical system obviously, which was due to the assumed reaction between the reducing agent and electrons at the cathode surface, like the reaction of H₂ with oxidizing agents.²⁷ The result was that the element phosphorous in NaH₂PO₂ was oxidized to trivalent phosphorous by the nickel ions in solution, corresponding to the phosphorous in phosphorous acid.¹⁸ If this j–E data was transferred to log|j|–E curves (shown in the inset in Fig. 1), it was shown that the pure Ni cathode polarization demonstrated a linear relationship between log|j| and E in the range of −0.7 to −0.9 V vs. the reference in comparison with no obvious linear relationship region for the Ni–P cathode polarization. This linear part of the graph was due to mass transportation as the rate-limiting step.²⁸ Steady state diffusion can be reached when the concentration gradient close to the electrode surface is constant, which was the most distinct difference between the two kinds of cathode polarization. When the voltage continually increased, side reactions, like the hydrogen reaction, could dominate the main current distribution, and the nickel ion reducing reaction would be ignored. It was clear that the addition of reducing agent to the solution could not affect the diffusion of nickel ions from the solution body to the cathode surface. So, the reducing agent was assumed to react at the cathode which directly led to the current density increase.

Fig. 1 Cathode polarization of pure Ni and the Ni–P deposit using electrolytes with different amounts of the reducing agent NaH₂PO₂. The inset shows the (linear) relationship of log|j| vs. E.

The experimental data of the chronoamperometry, shown in Fig. 2, illustrates the influence of deposition potential on the deposition process. Before the chronoamperometry test, the OCP was measured to confirm the stability of the system. So, a potential step existed from OCP to the chronoamperometry test. When the potential was selected as −0.1 V vs. OCP, the current density increased slowly and became stable when the test was conducted for about 30 s. This might be due to the fact that the electrode reaction was the rate-controlling step. When the driving force, the potential difference, was small, the rate of the electrode reaction was slower than that of the mass transport process which meant that the reactive materials at the cathode surface could not instantly be consumed to the lowest concentration level. The current–time curve at −0.2 V vs. OCP reached a maximum at ∼3 seconds with a typical Scharifker–Hills rule shape, indicating that the diffusion could be the rate-limiting step. If the potential step was too high like for −0.3 and −0.5 V vs. OCP, some side reactions, like the hydrogen reaction, might occur as shown in the cathode polarization results. The fitting results are shown in Fig. 2. The −0.2 V curve was the most integral in the test range.

Fig. 2 Variation of the current density with the deposition time at different deposition potentials for the Ni–P deposition with 0.05 M NaH₂PO₂ in the solution. The inset shows the fitting results of (j/j_m)² vs. t/t_m according to the S–H rule.

The Scharifker–Hills rule (S–H rule) was put forward in 1983 (ref. 21) and used to describe two nucleation mechanisms: (1) instantaneous and (2) progressive nucleation under diffusion-limiting conditions. In the case of instantaneous nucleation, the nucleation rate is independent of the deposition time, while this rate varies with the deposition time if the nucleation is regarded as progressive nucleation. The two equations below can describe the instantaneous (1) and the progressive (2) nucleation mechanism based on the I–t curve:

where i_m and t_m are the maxima of current and time as respective peak coordinates appearing in the chronoamperometry data. The fitting results according to the S–H rule are shown in the insets in Fig. 2 and 3. According to the S–H rule,²¹ the result of j_m²t_m after fitting was independent of the applied potential. The j_m²t_m results are listed in ESI 2† and are constant with very small fluctuation. The j_m²t_m value of −0.1 V was slightly bigger than those of the others, as the electrochemical reaction was the rate-limiting step at −0.1 V, which could not be described by the diffusion-limiting S–H based mode.

Fig. 3 j–t curves of the deposition of pure Ni and Ni–P with different reducing agent content in the electrolyte. The two stages of the j–t curve are divided by the red dashed line labeled in the figure. Insert: the fitting results of j–t experimental data by the S–H rule model.

The j–t curves, obtained from the deposition with different amounts of reducing agent in the electrolyte, are shown in Fig. 3. As labeled by the vertical, red dashed line in Fig. 3, the j–t curve could be divided into two stages. The first stage was considered as the nucleation process in the first 30 s. The nucleation process with different amounts of reducing agent could not be distinguished from each other. At the second stage, the current density increased with increasing reducing agent content in the electrolyte when the deposition time was longer than 30 s. The increase in current density was caused by the depolarization effect introduced by the addition of the reducing agent. Maybe the addition of reducing agent could reduce the solution resistance which could increase the current density at the same voltage. This is one of the possible reasons for the increase in current density. However, this possibility was eliminated by an experiment, in which a strong electrolyte was added to the solution and the current density showed no increase (as shown in ESI 3†). The current–time experimental data was fitted according to the S–H rule to give an insight into the nucleation pattern of the pure Ni and the Ni–P deposition, and the fitting results are shown in Fig. 3. The pure Ni nucleation data displayed a severe uptrend in comparison with the two nucleation models after the curve reached its maximum. The experimental data didn’t follow either of the nucleation mechanisms, and the reason was explained by Grujicic and coworkers²⁹ for the solution resistance under these deposition conditions. The S–H rule assumes that the nucleation would occur via island growth, which was not suitable for the pure Ni deposition film growth. The Ni–P nucleation data showed a relatively lower uptrend and the nucleation data for different amounts of P in solution did not show much difference between each other which was in good agreement with the first stage of the j–t curve (Fig. 3). This mismatched experimental data may be caused by some other factors that affect the fitting of the nucleation process data according to the S–H rule, such as the redistribution of metal ions in the double electrode layer. The intrinsic reason is currently under research and will be presented in our future work.

The j–t curves of pure Ni and Ni–P at different deposition times were measured (Fig. 4) to compare the curve shape of different kinds of deposition with different amounts of P in solution. Most of the current curves showed a maximum value when the test was performed for less than 3 seconds. By comparing the pure Ni and Ni–P j–t curves, the curve drop zone at the beginning of the pure Ni experiment disappeared when the Ni–P j–t curves were obtained. The drop zone shown in the pure Ni deposition process was due to the double electric layer charge process.²⁸ The disappearance of the drop zone meant that the reducing agent had an effect on the surface reaction. When other phosphorous compounds were introduced to the electrochemical deposition system, the drop zone did not disappear according to the results reported by Kurowski and coworkers.²⁶

Fig. 4 j–t curves of pure Ni and Ni–P at different deposition times: (a) pure Ni, (b) Ni–P with 0.02 M P in the electrolyte and (c) Ni–P with 0.05 M P in the electrolyte.

TEM was performed to observe the morphology and crystallization difference between the pure Ni and the Ni–P deposits, and the results are shown in Fig. 5. Fig. 5 shows one typical pure Ni deposit morphology which demonstrated the shape of a polygon with an interior angle of about 120°. The atom arrangement showed the features of a metal Ni {111} crystal where the surface atoms possessed the biggest coordination number and showed the most stable state; the assumption could be verified by the SAED results (Fig. 5). The SAED result demonstrated the good crystallization properties of pure Ni. In comparison, the Ni–P deposits were small ball-like particles featuring scattering at the copper net (Fig. 5d). This could show that the number of Ni–P nucleation sites was much higher than that of the pure Ni nucleation sites, which might be caused by the addition of the reducing agent NaH₂PO₂. By extending the deposition time, the Ni–P deposits grew along the ball particle surface. As shown in the Ni–P TEM image after deposition for 30 s (Fig. 5e), the ball particle grew bigger and the particle center was thicker than the particle edge. It illustrated that the crystal grew along any direction in comparison with that of the pure Ni deposit which grew along the interface of the substrate. From the SAED results (Fig. 5), the Ni–P deposit showed polycrystalline characteristics. The reducing agent at the electrode surface could react with the nickel ions and electrons together to affect the nucleation process. The HRTEM image exhibits many crystal particles (marked by red dashed lines in Fig. 5f).

Fig. 5 TEM images of the pure Ni morphologies deposited for 10 s (a) and 30 s (b); (c) pure Ni HRTEM image; TEM images of the Ni–P morphologies deposited for 10 s (d) and 30 s (e); (f) Ni–P HRTEM image. (1) Pure Ni SAED result; (2) and (3) Ni–P SAED results in different selected areas.

For pure Ni deposition, the Ni atoms reduced by electrons could migrate to the most appropriate place at the cathode surface such as the Ni {111} face and the Ni crystal grew along the plane edge. In the end, the morphology of the pure Ni deposit possessed a Ni {111} shape. With the same driving force for the Ni–P deposition process, the reducing agent NaH₂PO₂ could encounter the Ni atoms which were at the active positions of the crystal faces. As we all know, the reducing agent needs a catalyst when reducing the metal ions, and the atoms at different positions of the crystal possess different catalytic activities.³⁰ The Ni atoms at the active point which have not yet migrated into the most stable position could act as the catalyst for the reducing agent. Then, these active Ni atoms catalyze the reduction of nickel ions with the reducing agent in solution. This process hindered the in-plane growth of the Ni crystals, contributing to the ball particle deposition.

SEM images of the nucleation of pure Ni and Ni–P (in Fig. 6) are consistent with the TEM results. The pure Ni film (Fig. 6f) deposited for 100 s showed no characteristic morphology. The pure Ni deposit demonstrated a flat morphology, with some defects introduced by substrate residuals after polishing. All of the Ni–P films, deposited with different amounts of reducing agent in solution for 100 s, demonstrated a particle morphology, which was formed through the catalytic reduction by active Ni atoms.^8,18 The particle size did not change very obviously with the increase in P content in the solution. In the process of electrochemically reducing nickel ions by electrons, the reducing agent involved in this process accelerated the reducing process. The composition of these films was analyzed by EDS and the results are shown in ESI 4.† The content of the element P gradually increased with increasing reducing agent content in the electrolyte. According to research published before,^31,32 the chemical deposition needed a catalyst and could not occur at 70 °C or below in an acidic system. The occurrence of chemical deposition here meant that a highly active catalyst was produced during the electrochemical deposition process.

Fig. 6 SEM morphology images of the nucleation of pure Ni and Ni–P with different amounts of reducing agent (NaH₂PO₂) in the electrolyte after 100 s deposition: (a) 0.01 M; (b) 0.02 M; (c) 0.03 M; (d) 0.04 M; (e) 0.05 M; (f) 0 (pure Ni). The scale bar is 5 µm.

According to the above results and discussion, the Ni–P particle nucleation mechanism is proposed here and a schematic diagram of the nucleation is exhibited in Fig. 7. First, the nickel ions diffused to the cathode surface with the driving force afforded by an electric field. These nickel ions are reduced by cathode electrons and form nanoscale nickel particles which possess a very large specific surface area per atom. These Ni atoms at the particle surface have a high activity to act as a catalyst for the reducing reaction of nickel ions with NaH₂PO₂. When NaH₂PO₂ diffuses to these nanoparticle surfaces and encounters the active Ni atoms, the Ni atoms can catalyze the reduction of nickel ions with the reducing agent resulting in ball-like Ni–P particles at the cathode. In comparison, without any reducing agent in solution, the Ni atoms grow along the {111} crystal plane and form a Ni plane with a high degree of crystallinity.

Fig. 7 Schematic diagram of the nucleation of electrochemically deposited active Ni as the catalyst of the chemical deposition.

The introduction of the chemical deposition into the electrochemical deposition changed the nucleation process extensively. The nanoparticle nucleation of the Ni–P alloy film growth may also be very different from the pure electrochemical deposition. The morphology results of the Ni–P alloy films prepared for different deposition times are shown in Fig. 8. It was obvious that all of these films were composed of nanoscale particles. With prolonged deposition times, these nanoparticles aggregated at the cathode surface and some microscale morphologies appeared when the deposition time was longer than 20 min. The patterned film growth could be explained by the unstable growth theory.^24,33 These micro/nano-patterned alloy films were tested for their wetting behavior with water (ESI 5†). The contact angle of about 130° for a film deposited for 5 min increased to more than 150° for a film deposited for 60 min. The water droplet moved aside when the needle pressed it onto the film surface, which showed the superhydrophobic character of the film with micro/nanoscale morphology.

Fig. 8 SEM images of the Ni–P alloy films with patterned morphologies for different deposition times: (a) 5 min; (b) 10 min; (c) 30 min; (d) 60 min. The scale bar is 30 µm.

4. Conclusions

The nucleation mechanism of Ni–P deposits, formed by a synergetic effect of electrochemical and chemical deposition, was studied by cathode polarization and chronoamperometry. The depolarization phenomenon generated by the addition of reducing agent to the solution was assumed due to the reduction with reducing agent at the cathode surface. Electrochemical deposition could produce Ni atom clusters with a high specific surface at the beginning of the deposition process which can act as a catalyst for the reducing reaction of metal nickel ions with the reducing agent NaH₂PO₂. This catalytic reaction made the reducing reaction possible which would otherwise not occur at this deposition temperature. This catalytic reduction process can clearly clarify the formation mechanism of the ball-like particles formed by the synergetic process of the electrochemical and chemical deposition. Also, this catalytic reduction gives us a glimpse of the occurrence of chemical deposition at lower temperatures, which provides researchers with new ideas for catalyst preparation and particle growth mechanisms.

Acknowledgements

Projects 51475450 and 51335010 were supported by the National Nature Science Foundation of China. This material is also based upon work funded by the National Basic Research Program of China (no. 2014CB643302) and Zhejiang Provincial Innovation Team (grant no. 2011R50006).

References

1. L. Lu, M. L. Sui and K. Lu, Science, 2000, 287, 1463–1466.
2. E. Pellicer, A. Varea, S. Pane, B. J. Nelson, E. Menendez, M. Estrader, S. Surinach, M. Dolors Baro, J. Nogues and J. Sort, Adv. Funct. Mater., 2010, 20, 983–991.
3. T. Darmanin, E. T. de Givenchy, S. Amigoni and F. Guittard, Adv. Mater., 2013, 25, 1378–1394.
4. S. Ye, A. R. Rathmell, Z. Chen, I. E. Stewart and B. J. Wiley, Adv. Mater., 2014, 1–18,  DOI:10.1002/adma.201402710.
5. J. A. Switzer, Science, 2012, 338, 1300–1301.
6. Y. Liu, D. Gokcen, U. Bertocci and T. P. Moffat, Science, 2012, 338, 1327–1330.
7. Q. Yu, Z. Zeng, W. Zhao, Y. Ma, X. Wu and Q. Xue, ACS Appl. Mater. Interfaces, 2014, 6, 1053–1060.
8. Q. Yu, Z. Zeng, W. Zhao, H. Li, X. Wu and Q. Xue, Chem. Commun., 2013, 49, 2424–2426.
9. S. Brimaud and R. J. Behm, J. Am. Chem. Soc., 2013, 135, 11716–11719.
10. Y. Wu, T. Hang and Z. Yu, Chem. Commun., 2013, 49, 10391–10393.
11. A. Kloke, C. Köhler, R. Gerwig, R. Zengerle and S. Kerzenmacher, Adv. Mater., 2012, 24, 2916–2921.
12. A. Kloke, C. Köhler, R. Zengerle and S. Kerzenmacher, J. Phys. Chem. C, 2012, 116, 19689–19698.
13. K. Li, J. Ju, Z. Xue, J. Ma, L. Feng, S. Gao and L. Jiang, Nat. Commun., 2013, 4, 2276.
14. F. Wan, X. Pei, B. Yu, Q. Ye, F. Zhou and Q. Xue, ACS Appl. Mater. Interfaces, 2012, 4, 4557–4565.
15. Z. Shi and A. V. Walker, Langmuir, 2011, 27, 11292–11295.
16. P. Lu, Z. Shi and A. V. Walker, Langmuir, 2011, 27, 13022–13028.
17. X. R. Wang, S. M. Tabakman and H. J. Dai, J. Am. Chem. Soc., 2008, 130, 8152–8153.
18. Q. Yu, Z. Zeng, W. Zhao, Y. Ma, X. Wu and Q. Xue, ACS Appl. Mater. Interfaces, 2014, 6, 1053–1060.
19. G. Lian, O. Gerko, R. Aleksandar, M. H. Peter and C. S. Peter, J. Phys. D: Appl. Phys., 2011, 44, 443001.
20. A. M. Lockett, P. J. Thomas and P. O’Brien, J. Phys. Chem. C, 2012, 116, 8089–8094.
21. B. Scharifker and G. Hills, Electrochim. Acta, 1983, 28, 879–889.
22. R. Aogaki and T. Makino, Electrochim. Acta, 1981, 26, 1509–1517.
23. R. Aogaki and T. Makino, J. Chem. Phys., 1984, 81, 2164–2168.
24. A. Haghdoost and R. Pitchumani, Langmuir, 2014, 30, 4183–4191.
25. A. S. Myerson and B. L. Trout, Science, 2013, 341, 855–856.
26. A. Kurowski, J. W. Schultze and G. Staikov, Electrochem. Commun., 2002, 4, 565–569.
27. T. N. Andersen and H. Eyring, J. Phys. Chem., 1963, 67, 92–97.
28. I. Hanzu, T. Djenizian, G. F. Ortiz and P. Knauth, J. Phys. Chem. C, 2009, 113, 20568–20575.
29. D. Grujicic and B. Pesic, Electrochim. Acta, 2002, 47, 2901–2912.
30. H. G. Yang, C. H. Sun, S. Z. Qiao, J. Zou, G. Liu, S. C. Smith, H. M. Cheng and G. Q. Lu, Nature, 2008, 453, 638–641.
31. J. N. Balaraju, S. Millath Jahan, C. Anandan and K. S. Rajam, Surf. Coat. Technol., 2006, 200, 4885–4890.
32. J. N. Balaraju and K. S. Rajam, Surf. Coat. Technol., 2005, 195, 154–161.
33. R. Aogaki, K. Kitazawa, Y. Kose and K. Fueki, Electrochim. Acta, 1980, 25, 965–972.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra01027f

---