Effect of current density and grain refining agents on pulsed electrodeposition of nanocrystalline nickel

Abstract

Nanocrystalline nickel was electrodeposited on copper substrate using a Watt's bath at a temperature of 50°C. The effect of pulsed current density (up to 1·5 A cm⁻²) and role of organic additives (namely saccharin and sodium citrate) in the deposition bath was evaluated for rendering grain refinement. The deposited samples were characterised for current efficiency, colonial morphology, crystallite size, texture, coercivity and nanomechanical properties (such as Young's modulus and hardness). Results were compared with deposition using direct current, and utilising annealed Cu substrate. The smallest crystallite size of 22 nm was obtained, and the lowest coercivity equaled 22 Oe for the sample deposited using pulsed current with the peak current density 1 A cm⁻² and 10 g L⁻¹ saccharin addition. Annealing of the substrate has shown to increase the Young's modulus of electrodeposited Ni by 10·2%, which increases further by 14·4% with saccharin addition when compared to those of deposits without saccharin addition.

Keywords

Pulsed current electrodeposition Nanocrystalline nickel Complexing agent Coercivity Nanoindentation

Introduction

Nanocrystalline materials are one of the most investigated materials in current research. There are several routes to obtain nanocrystalline coatings such as pulsed laser deposition,1–3 electrodeposition,4–7 chemical vapour deposition8–10 and thermal spray.11–14 Among them, electrodeposition route is one of the convenient and economical methods to produce bulk, metallic, dense and nanocrystalline, pure metal or alloy or composite.7

In electrodeposition, there are two methods to obtain nanocrystalline materials: first by the addition of organic additives,15 like saccharin and sodium citrate in the electrolyte or using pulsed current (PC) density during deposition.15 Saccharin which is benzoic acid sulphimide decomposes to o-toluene sulphonamide and benzamide in Watts bath and gets codeposited along with nickel.16 In the case of sodium citrate addition, citrate ions react with nickel ions to form nickel citrate complexes and gets adsorbed on the surface of a cathode. Citrate ions also get adsorbed on the surface of the cathode, resulting in the inhibition of the hydrogen evolution reaction.17 Organic additives result in better surface finish and lower internal stresses by poisoning the nuclei from further growth as they are also codeposited.15,18,19 In PC deposition, a cyclic current waveform is used instead of constant current, unlike in the case of direct current (DC) deposition.5,6 Pulsed current electrodeposition provides better control of microstructure and resulting properties due to flexibility in controlling deposition parameters like current density, current on time, current off time, duty cycle (ratio of current on time and total pulse time) and pulse frequency.20 Pulsed deposition also results in lower internal stresses.21–23 One of the advantages of pulsed deposition is that higher current densities can be used. In the case of DC deposition, high current densities can result in powdery or burnt deposits. However, in the case of pulsed electrodeposition, sufficient time is available for metal ions build-up near surface during current off time. High current density causes higher nucleation rate and lower growth rate of crystallites, thus resulting in finer grain structure.24 In the case of alloy deposition, large overpotential is available for different types of ions to discharge, resulting in better control of alloy composition.25,26

Improvement in the properties through pulsed deposition, like high wear resistance,24 corrosion resistance27 and hardness,28 low coercivity,26,29,30 reduced porosity,28 lower internal stress,28 lower impurity,28 lower hydrogen content28 and better surface finish28 have already been reported.

In the case of PC electrodeposited nickel using Watt's bath, only a few researchers31,32 have investigated electrodeposits using high peak current densities. El-Sherik et. al.31 used peak current density of 0·4, 0·8 and 1·6 A cm⁻². They noticed (200) texture at higher peak current densities. They also noticed larger colonies but finer grain structure of size 100 nm without organic additives at higher current densities. Zimmerman et al.32 utilised Watt's bath with additive saccharin (5 g L⁻¹) and SiC as suspension to obtain Ni–SiC composite. They utilised peak current densities up to 0·9 A cm⁻² and obtained nickel with average crystallite size of 15·9 nm. It is therefore interesting to further characterise PC electrodeposits using Watt's bath (with and without additives) and peak current densities in a wide range.

In the present work, the effect of higher peak current densities (up to 1·5 A cm⁻²) on the resulting crystallite size and its consequent mechanical properties (hardness and elastic modulus) have been investigated. The samples were characterised for current efficiency, morphology, crystallite size, texture and coercivity. The effect of organic additives (saccharin and sodium citrate) was also studied. The overall objective of this study was to understand the structure property relationship caused primarily due to PC, organic additives and annealed substrate.

Experimental

Nickel on copper substrate was electrodeposited using a Watt's bath. Saccharin and sodium citrate were used as additives. Direct and PC at various peak current densities, current off time and duration were controlled by potentiostat 263A (Princeton Applied Research, Oak Ridge, TN, USA). The software ‘Virtual Potentiostat’ was used to control instantaneous current. The PC waveform is defined by current on time t_on, current off time t_off, duty cycle t_on/(t_on+t_off), frequency (t_on+t_off) ⁻¹ and peak current density i_p. The temperature was controlled by controlled temperature water circulating inside the double wall of a cylindrical electrolytic cell. The samples were polished up to 500 grade SiC paper, ultrasonically cleaned and degreased using acetone, before deposition.

All the electrodeposition parameters are tabulated in Table 1. The samples were electrodeposited in two stages. First the effect of DC and PC and without and with additives (of saccharin 10 g L⁻¹ and sodium citrate 25 g L⁻¹) was studied. The samples electrodeposited using DC with current density equal to 0·1 A cm⁻², PC with t_on = 2 ms, t_off = 18 ms and i_p = 1 A cm⁻², PC and 10 g L⁻¹ saccharin (termed as PC+S₁ from now on) as well as PC and 25 g L⁻¹ sodium citrate (termed as PC+S₂ from now on) samples were characterised. The total charge consumed in these individual experiments was 900 C cm⁻². Then, in the same electrolyte, another sample using same deposition condition was deposited for one-fifth of the deposition time (180 C cm⁻²).

Table 1

Overview of nickel electrodeposition parameters

Electrolyte composition
NiSO₄.7H₂O, g L⁻¹	250
NiCl₂.6H₂O, g L⁻¹	30
H₃BO₃, g L⁻¹	40
Saccharin, g L⁻¹	0, 10
Sodium citrate, g L⁻¹	0, 25
Electrodeposition
pH	4
Temperature, °C	50
Agitation of electrolyte	Magnetic stirrer at 500 rev min⁻¹
Anode	Ni
Substrate	Copper plate (13×13×2 mm)
Type of current	Direct current, pulsed current
Peak current density, A cm⁻²	0·33, 0·45, 0·67, 1, 1·5
Direct current density, A cm⁻²	0·1
On time t_on, ms	2
Off time t_off, ms	18
Total charge used, C cm⁻²	900, 180

Based on the characterisation results of these samples, experiments in second stage were conducted using 10 g L⁻¹ saccharin as additive and at various peak PC densities (i_p = 1·5, 1, 0·67, 0·45, 0·33 A cm⁻²) and for a fixed duration of 2 h. The other pulse parameters (t_on, t_off) were the same as that in the first stage.

Current efficiency was determined by measuring sample weight before and after electrodeposition and knowing the theoretical weight using Faraday's second law of electrolysis.

The X-ray diffraction (XRD) data was collected using Cu K_α radiation (λ = 0·1540598 nm), in a Rich Seifert ISO Debyeflex 2002 diffractometer after polishing samples. The samples were scanned in 2θ range of 40–100° at scan rates of 1 and 3° min⁻¹. In experiments conducted in the second stage, the scan rate used was 3° min⁻¹. Commercially available bulk nickel was annealed at 1000°C for 90 min. This annealed bulk nickel (ABN) was consequently etched with etchant (80 mL HNO₃ and 3 mL HF). X-ray diffraction data of bulk nickel were used to eliminate effect of machine broadening in other samples to evaluate crystallite size using Scherrer equation for various samples.

An atomic force microscope (AFM) (Molecular Imaging Corporation, Knoxville, TN, USA) was used to scan the as deposited surfaces for thin deposits of first stage samples. Thin deposits were used due to relatively smoother surface. The surfaces of these samples were also observed in an FEI Quanta 200HV scanning electron microscope (SEM).

Coercivity was measured using vibrating sample magnetometer (ADE EV7 VSM, Lowell, MA, USA). Coercivity of ABN was also measured for the comparison purpose.

Consequently, nanoindentation on the electrodeposited Ni substrate was performed using a triboindenter (Hysitron, Eden Prairie, MN, USA) with a 100 nm radius pyramidal Berkovich tip at a load of 1000 μN (ramped at 100 μN s⁻¹) and dwell time of 4 s at peak load. Young's modulus and hardness of the indented samples was evaluated by Oliver and Pharr approximation from the unloading section of load–displacement curve.33 Nanoindentation on the electrodeposited Ni without saccharin is taken as a reference to compare the elastic modulus and hardness of samples with saccharin addition, and deposits on annealed substrate.

Results and discussion

Since the experiments were conducted in two stages, results obtained in stage one will be discussed first. On the visual examination of samples deposited in the first stage, it was found that thick deposits (with thickness in the range 80–300 μm depending on the current efficiency) had rough surface and it was not possible to characterise them using AFM. Therefore, thin deposits (with thickness one fifth of corresponding thick samples were obtained after electrodepositing using same conditions but for one-fifth of the deposition time of thick samples) were utilised for AFM and SEM studies. In order to test the hypothesis that the properties are independent of deposition time (for deposits in μm range), both thick and thin deposits for all the conditions were characterised for current efficiency and XRD. It was observed that both current efficiency and XRD pattern matched for thick and thin deposits. Therefore, it may be inferred that AFM and SEM studies conducted using thin deposits correlates with that of thick deposits, concerning crystallite sizes and grain morphologies. Now the results for first stage experiments will be discussed in detail.

The XRD pattern obtained for thick deposits is shown in Fig. 1 within a 2θ range of 40–55°. Compared to that of bulk nickel, all the samples show peak broadening. All the deposited samples show almost equal shift in the peak towards lower angle (as indicated in figure through dotted line) due to increase in the lattice parameter.34 An increase in the lattice parameter indicates presence of tensile stresses in the deposits. It can be noted that tensile stresses generally result during electrodeposition of nickel.35 However, some researchers36 have also reported generation of compressive stresses during electrodeposition.

Comparing XRD patterns of electrodeposited samples with that of bulk nickel, texture was evident in almost all the electrodeposited samples. Direct current samples showed (200) texture. This is in agreement with the existing literature.37 In the case of PC samples, texturing tendency was slightly reduced due to reduction in crystallite size similar to the results available in the literature.37,38 It may be noted that finer the crystallite size, more random is the orientation of the grains. In the case of PC+S₁ sample, saccharin addition almost nullified the texture caused by electrodeposition. This may be explained due to codeposition of saccharin, which ceased any directional growth of nuclei and reduced the crystallite size further. However, this effect of additives in reducing texture was found less effective in the case of sodium citrate addition compared to that with saccharin addition. In the case of PC+S₂ sample, sodium citrate addition reduced the texturing tendency (evinced by strong (111) peak in Fig. 1) caused by electrodeposition. The crystallite sizes were calculated using Scherrer's equation and full width at half maximum of (111) peaks.39 The values are tabulated in Table 2. The crystallite size was found largest (101 nm) for DC sample as lower current density was used compared to peak current densities in the case of PC deposited samples. It should be noted that Scherrer's equation is valid for crystallite size smaller than ∼100 nm. The maximum grain refinement occurred for PC+S₁ sample with crystallite size equal to 22 nm. Pulsed current sample also resulted in grain refinement (from 101 to 39 nm). In the case of PC+S₂ sample, the effect of sodium citrate addition was found negative with respect to grain refinement as crystallite size was larger (61 nm) compared to PC sample (39 nm). It can be inferred from this observation that even higher amount of sodium citrate is not sufficient to restrict the grain growth. In order to confirm the crystallite size measured using XRD data, as deposited coating surfaces were scanned using AFM. The phase images obtained in tapping mode are shown in Fig. 2. The image size of DC deposited sample is 500×500 nm (due to their coarser grains). For other samples, the scan area was kept as 150×150 nm. Correspondingly, the crystallite size estimated using XRD matched with the values obtained via AFM images, as shown in Fig. 2 (i.e. DC>100 nm, PC ≈35 to 40 nm, PC+S₁≈25 nm, and PC+S₂≈60 nm).

Table 2

Physical data for samples DC, PC, PC+S₁ and PC+S₂: crystallite size for sample ABN was obtained using optical microscope

	DC	PC	PC+S₁	PC+S₂	ABN
Current efficiency, %	98	35	26	30
Crystallite size, nm	101	39	22	61	3×10⁵
Coercivity, Oe	54	74	22	62	6

Regarding the grain morphology, DC samples show elongated grains (Fig. 2a). The dark spots in the image were possibly due to impurity over the surface. In the case of PC sample, fine grains with irregular surface were also noticed (Fig. 2b). Sharp morphology of grains in the case of PC+S₁ and PC+S₂ (Fig. 2c and d) is attributed to the selective adsorption of saccharin and sodium citrate, which poison the growth sites and restrict crystallite size.37 In the case of PC+S₁ sample (Fig. 2c) elongated grains with river flow-like pattern were noticed. In the case of PC+S₂ (Fig. 2d) large (∼60 nm) spherical grains were obtained. In conjunction with a crystallite size of 61 nm obtained from XRD with sodium citrate addition confirms inefficient contribution in restricting the grain growth of electrodeposited Ni.

Since current efficiency indicates energy utilisation achieved during deposition, it becomes helpful in predicting the theoretical coating thickness as well. The current efficiency was found highest in the case of DC samples (98%). In the case of pulse deposited samples, the current efficiency was lower (35%). This was due to the fact that peak current density for pulsed deposited samples was 10 times higher than the current density used for DC deposited sample. Additionally, the presence of secondary organic additives lower the deposition efficiency owing to their insulating nature, thereby PC+S₁ and PC+S₂ have achieved marginally lower current efficiencies of 26 and 30% respectively. At a high peak current density, large overpotential for discharge of hydronium ions is also available leading to lower current efficiency.35

Since coatings become an inherent part of a component, retention of the magnetic charges on the surface can lead to interference with the signal. Hence, the current effort is also directed to achieve minimum coercivity of electrodeposited Ni. Coercivity data is tabulated in Table 2. Herzer40 established relationship between crystallite size and coercivity in a wide crystallite size range by compiling data for wide variety of materials. He found that coercivity decreases as sixth power of decrease in the crystallite size in the nanometer regime. In the micrometre range, coercivity decreased linearly with increasing crystallite size. In addition, internal stresses41 increase coercivity. Organic additives (for example saccharin) are known for reducing tensile internal stresses or even inducing compressive stresses depending on the amount codeposited.41 Considering all these findings, it is a difficult task to relate coercivity data to other physical parameters. The general observations on the coercivity data can be summarised as follows. In the case of ABN, coercivity was minimum (6 Oe) due to the absence of internal stresses and also due to very large crystallite size. Coercivity was higher in the case of PC sample (74 Oe) compared to DC sample (54 Oe). In the case of PC+S₁ sample, coercivity was lower (22 Oe) when compared to that of PC+S₂ sample (62 Oe). Lower coercivity is desired in the case of soft magnetic materials as it leads to lower hysteresis loss of magnetisation energy. Reduction in the grain size and reduction of the internal stresses have shown to decrease coercivity as observed in PC+S1.40,41 However, in the case of PC+S₂ sample, coercivity was marginally lower (62 Oe) than PC sample (74 Oe) due to larger crystallite size of PC+S₂ sample.

Since saccharin addition elicits optimum set of properties (i.e. refined crystallite size with minimum texturing), the electrodeposition experiments in second stage were conducted using various peak current densities. The results obtained in the second stage are discussed below.

The XRD results for various peak current densities on saccharin added deposits are shown in Fig. 3. The results are quite similar at different peak current densities. Sample deposited at i_p = 1 A cm⁻² showed maximum broadening of (111) peak. The crystallite sizes calculated using Scherrer's formula are tabulated in Table 3. At a very high current density, the crystallite size reduction reached its limiting value due to saturation of nucleation rate, and crystallite sizes were obtained between 22 and 34 nm. Thereby, correlation between current density and crystallite size could not be drawn. This indicates that current density variation is causing only marginal effect on refining the crystallite size, and it is primarily saccharin organic additive responsible for the restricting grain growth.

Table 3

Physical data for samples electrodeposited using pulsed current and 10 g L⁻¹ saccharin at various peak current densities

Peak current density, A cm⁻²	0.3	0.45	0.67	1.0	1.5
Current efficiency, %	97	72	57	26	26
Crystallite size, nm	27	34	33	22	30
Coercivity, Oe	33	40	35	22	34

Current efficiency decreased with an increase in the peak current density (Table 3)20,42 due to higher overpotential available for hydrogen evolution at higher peak current densities. Some researchers28,29 have plotted current efficiency versus average current density, and have concluded that current efficiency increases with increase in average current density. Coercivity decreased with decrease in crystallite size (Table 3) as found by Herzer.40

Substrate also has a strong effect in inducing the stresses to the electrodeposited samples. Hence, copper substrate is annealed at 1000°C for 90 min before electrodeposition of Ni. Electrodeposited Ni coating on a rolled copper substrate (without annealing and without saccharin addition) is shown in Fig. 4a (thickness≈50 μm) for the comparison purpose. Annealing of the copper substrate before electrodeposition of Ni shows significant reduction of cracking and porosity (from ∼6 to 2%) in the electrodeposited Ni (Fig. 4b). Coatings are thinner (∼25 μm) because the total charge density used for deposition was 180 C cm⁻² which was one-fifth that of thicker deposits. Moreover, with addition of saccharin, completely dense coating is observed (Fig. 4c) with similar coating thickness (∼25 μm).

Consequently, to evaluate the effect of electrodeposition on the annealed copper substrate, nanoindentation of the electrodeposited samples has been performed. Load–displacement response of electrodeposited Ni (as deposited and on annealed substrate) upon nanoindentation is shown in Fig. 5. Electrodeposited Ni without saccharin on rolled copper substrate shows an average Young's modulus of 186·8 GPa with a hardness of 3·8 GPa (Table 4). Since the generated tensile stresses are relaxed after annealing of substrate, the substrate does not affect the coating in sharing its residual stresses and cracking is minimised in the electrodeposited coating. Thereby the modulus and hardness of electrodeposited samples on annealed substrate increased to 208·1 GPa (10·2%) and 4·4 GPa respectively. However, with the addition of saccharin in the electrodeposited Ni (on an annealed copper substrate), the modulus and hardness remain approximately similar, i.e. Young's modulus of 213·8 GPa with a hardness of 4·1 GPa. In addition to rendering a smooth surface devoid of cracks and porosity, enhanced mechanical properties (Young's modulus and hardness) are obtained with saccharin addition in electrodeposited Ni after annealing the substrate. This effect of pile up to result ductility and uniform indentation depth is also observed in Table 4. It must be noted that electrodeposited nanocrystalline Ni coating contains microporosity (Fig. 4a). Consequently, elimination of porosity is assisted by release of substrate stresses after annealing, and restricting of stress transfer to deposit resulting a dense structure as shown in Fig. 4b and c. Hence, reduction in the porosity strongly contributes to enhanced Young's modulus (∼210 GPa) and hardness (>4·1 GPa) of the electrodeposited coatings on annealed substrate.43 In addition, presence of stresses, inhomogeneity (pores, cracks) allows deeper penetration of indenter tip resulting in enhanced depth displacement. Hence, marginally higher hardness (>4·1 GPa) results after the stresses are relaxed via annealing of copper substrate. It is expected that addition of organic saccharin should reduce the hardness of electrodeposited Ni. Effect of saccharin addition alone can be visualised by marginally lower hardness of annealed samples with saccharin addition (4·1 GPa) than that of without saccharin (4·4 GPa). However, the overall hardness is a balance between porosity/inhomogeneity, presence of residual stresses, crystallite size and saccharin content of the coating. Thereby, the hardness of all the coatings is more or less similar due to these complex effects. Despite the conventional knowledge of Young's modulus remaining constant with the composition, role of porosity and crystallite size reduction in reducing Young's modulus is observed to be prominent in the nanoregime (20–50 nm).43,44 Diminishing effect of grain boundary interaction result in Young's modulus of nanocrystalline Ni similar to that of a polycrystalline Ni (∼210 GPa).44

Table 4

Young's modulus and hardness of electrodeposited Ni

Conclusions

Nickel on copper substrate was electrodeposited using a Watt's bath using both DC and PC. Saccharin and sodium citrate were used as additives in electrolyte bath. The samples were characterised for crystallite size, morphology, texture, coercivity and current efficiency. All the electrodeposited samples showed peak broadening and shift in peak towards lower angle. The DC deposited sample showed (200) texture. The PC deposited samples showed reduced magnitude of texture. Additives further reduced the magnitude of texture. The effect of peak current density on crystallite size was not prominent and concludes a stronger role of saccharin additive in restricting grain growth in comparison to that of PC density. Regarding grain morphology, DC deposited samples showed large elongated grains due to much lower peak current density. Pulsed current drastically reduced the crystallite size (reduction from 101 to 39 nm). Sodium citrate addition increased the crystallite size (from 39 to 61 nm) because of its insufficient concentration in restricting grain growth. The current efficiency decreased with increase in peak current density due to peak potential exceeding the overpotential for hydrogen evolution over deposited nickel surface. The coercivity decreased with decrease in crystallite size. Coercivity was found lowest (∼22 Oe) for PC deposited sample in electrolyte containing saccharin. Annealing of the copper substrate has resulted enhanced Young's modulus (14·4%) and hardness (∼7·9%) owing to reduced porosity of the deposits.

References

Mesaros-Hristea

, Alm

, Popovici

E-J

, Boman

: Thin Solid Films, 2008, 516, 8431–8435.

Beke

, Giorgio

, Korösi

, Nánai

, Marine

: Thin Solid Films, 2008, 516, 4659–4664.

Chung

J-K

, Kim

W-J

, Kim

, Song

, Kim

: Thin Solid Films, 2008, 516, 4190–4193.

Moti

, Shariat

, Bahrololoom

: Mater. Chem. Phys., 2008, 111, 469–474.

, Jiang

, Huang

, Tian

: Appl. Surf. Sci., 2008, 254, 6865–6869.

Shen

, Xue

, Wang

, Liu

, Zuo

: Surf. Coat. Technol., 2008, 202, 5140–5145.

, Jiang

, Pang

, Wang

, Liang

: Surf. Coat. Technol., 2007, 201, 5925–5930.

Adachi

, Kavanagh

, Karim

: Thin Solid Films, 2008, 516, 7418–7421.

Mukhopadhyay

, Chowdhury

, Ray

: Thin Solid Films, 2008, 516, 6824–6828.

10.

Cheng

, Xu

, Chai

, Huang

, Ren

, Long

, Rutkevych

, Ostrikov

: Thin Solid Films, 2008, 516, 5991–5995.

11.

Agarwal

, Tim McKechnie , Seal

: ‘Net shape nanostructured aluminum oxide structures fabricated by plasma spray forming’, J. Therm. Spray Technol., 12, (3), 2003, .

12.

Laha

, Agarwal

, McKechnie

, Rea

, Seal

: ‘Synthesis of bulk nanostructured aluminum alloy component through vacuum plasma spray technique’, Acta Mater., 2005, 53, 5429–5438.

13.

Laha

, Liu

, Agarwal

: ‘Carbon nanotube reinforced aluminum nanocomposite via plasma and high velocity oxy-fuel spray forming: a comparative analysis’, J. Nanosci. Nanotech., 2007, 7, 515–524.

14.

Viswanathan

, Laha

, Balani

, Agarwal

, Seal

: ‘Challenges and advances in nanocomposite processing techniques’, Mater. Sci. Eng. R, 2006, R54, (5–6), 121–285.

15.

Brankovic SR , Vasiljevic

, Klemmer

, Johns

: J. Electrochem. Soc., 2005, 152, C196–C202.

16.

Dini

: ‘Electrodeposition: ‘The materials science of coatings and substrates’; 1993, Park Ridge, NJ, Noyes publications.

17.

C-Q

, Li

X-H

, Wang

Z-X

, Guo

H-J

: Trans. Nonferrous. Met. Soc. Chn, 2007, 17, 1300–1306.

18.

Ciszewski

, Posluszny

, Milczarek

, Baraniak

: Surf. Coat. Technol., 2004, 183, 127–133.

19.

Kim

, Sohn

, Joo

, Kim

, Yim

, Lee

, Kang

: Surf. Coat. Technol., 2005, 199, 43–48.

20.

Fenineche

, Kedim

, Coddet

: Surf. Coat. Technol., 1991, 48, 205–209.

21.

Yin

: Surf. Coat. Technol., 1996, 88, 162–164.

22.

, Chan

, Zhu

: Surf. Coat. Technol., 1997, 91, 220.

23.

Tang

: Electrochim. Acta, 2001, 47, 61–66.

24.

Jeong

, Gonzalez

, Palumbo

, Aust

, Erb

: Scr. Mater., 2001, 44, 493–499.

25.

Choo

RJC

, Toguri

, El-Sherik

, Erb

: J. Appl. Electrochem., 1995, 25, 384–403.

26.

Gomez

, Pellicer

, Valles

: Electrochem. Commun., 2005, 7, 275–281.

27.

Yang

, Yang

, An

, Zhang

, Tu

, Li

: Plat. Surf. Finish., 2001, 88, 116.

28.

Ghosh

, Grover

, Dey

, Totlani

: Surf. Coat. Technol., 2000, 126, 48–63.

29.

Nakanishi

, Ozaki

, Nam

H-S

, Yokoshima

, Osaka

: J. Electrochem. Soc., 2001, 148, C627–C631.

30.

Eftekhari

: Philos. Mag. Lett., 2004, 84, (9), 587–592.

31.

El-Sherik

, Erb

, Page

: Surf. Coat. Technool., 1996, 88, 70–78.

32.

Zimmerman

, Clark

, Aust

, Erb

: Mater. Lett., 2002, 52, 85–90.

33.

Pharr

, Oliver

, Brotzen

: J. Mater. Res., 1992, 7, 613.

34.

Thomas

, Loubens

, Gergaud

, Labat

: Appl. Surf. Sci., 2006, 253, 182–187.

35.

Lowenheim

F A

, ed (ed.): ‘Modern electroplating’, 3rd edn; 1974, New York, Wiley.

36.

Mishra

, Balasubramaniam

: Corros. Sci., 2004, 46, 3019–3029.

37.

Pavlatov

, Stroumbouli

, Gyftou

, Spyrellis

: J. Appl. Electrochem., 2006, 36, 385–394.

38.

Gyftou

, Stroumbouli

, Pavlatou

, Asimidis

, Spyrellis

: Electrochim Acta, 2005, 50, 4544–4550.

39.

Cullity

: ‘Elements of X-ray diffraction’, 2nd edn; 1977, Reading, MA, Addison Wesley.

40.

Herzer

: Scr. Mater., 1995, 33, (10/11), 1741–1756.

41.

Smith

, Godycki

, Lloyd

: J. Electrochem. Soc., 1961, 108, 996–998.

42.

Ganesan

, P S, Kumaraguru , Popov

: Surf. Coat. Technol., 2006, 201, 3658–3669.

43.

Chen

, Balani

, Agarwal

: Appl. Phys. Lett., 2007, 91, 031903.

44.

Mirshams

, Pothapragada

: Acta Mater., 2006, 54, 1123–1134.