Ultrasound-assisted electrodeposition of nickel: Effect of ultrasonic power on the characteristics of thin coatings

Surface & Coatings Technology 264 (2015) 49–59

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat

Ultrasound-assisted electrodeposition of nickel: Effect of ultrasonic power on the characteristics of thin coatings Ignacio Tudela a,b,⁎, Yi Zhang b, Madan Pal b, Ian Kerr b, Timothy J. Mason c, Andrew J. Cobley a,⁎⁎ a b c

Daido Metal Co., Ltd., The European Headquarters, Winterhay Lane, Ilminster, TA19 9PH, UK The Functional Materials Applied Research Group, Faculty of Health and Life Sciences, Coventry University, Priory Street, Coventry CV1 5FB, UK The Sonochemistry Centre, Faculty of Health and Life Sciences, Coventry University, Priory Street, Coventry CV1 5FB, UK

a r t i c l e

i n f o

Article history: Received 5 November 2014 Accepted in revised form 9 January 2015 Available online 17 January 2015 Keywords: Nickel Watts bath Electroplating Electrodeposition Ultrasound

a b s t r a c t The effect of ultrasonic power on the characteristics of low-frequency ultrasound-assisted electrodeposited Ni coatings from an additive-free Watts bath has been evaluated by different methods. XRD analysis showed that, while mechanical agitation favoured the electrocrystallization of Ni in the [211] direction, ultrasound promoted the electrodeposition of Ni with a [100] preferred orientation. FIB-SEM images of the surface of Ni deposits not only indicated that the surface structure agreed to some extent with the XRD results, but also that ultrasound refined, to a certain extent, some of the grains of the surface of the coatings. FIB-SEM images of the cross-section of the coatings confirmed this effect of ultrasound on the microstructure of the deposits. Such change in the microstructure of Ni, along with work-hardening by ultrasound, resulted in an increase in the hardness of the deposits. The characteristics of the deposits depended on the ultrasonic power employed, and it was found that Ni coatings electrodeposited using an ultrasonic power of 0.124 W/cm3 presented the higher proportion of crystals with a [100] preferred orientation, the highest degree of grain refinement in the surface and the highest microhardness values. Nevertheless, these deposits also presented visible erosion marks on the surface of the coatings due to the formation of transient bubble structures near the surface of the cathode during the electrodeposition. These erosion marks might be considered the main drawback to the use of ultrasound during the electrodeposition. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Since Bird first described the formation of ‘a crust of metallic nickel on the negative electrode, often of a silvery lustre on the surface immediately applied to the platinum’ from a bath consisting of NiCl2 and NiSO4 nearly 180 years ago [1], a wide variety of studies have been focused on the electrodeposition of Ni as it is one of the most common metal plating processes in industry [2]. Its importance in terms of economic and commercial impact in the form of metal and salts annually consumed by the electroplating industry has been roughly estimated around 100,000 tonnes worldwide [3]. Among the different Ni electroplating processes currently employed in industry the Watts bath [4] has grown to become the most widespread Ni electroplating process with little modification. This type of bath not only produces high quality deposits, but it also is a very efficient process as the cathode

⁎ Correspondence to: I. Tudela, The Functional Materials Applied Research Group, Faculty of Health and Life Sciences, Coventry University, Priory Street, Coventry CV1 5FB, UK. Tel.: + 44 7521160565. ⁎⁎ Corresponding author. Tel.: +44 7706955901. E-mail addresses: [email protected] (I. Tudela), [email protected] (A.J. Cobley).

http://dx.doi.org/10.1016/j.surfcoat.2015.01.020 0257-8972/© 2015 Elsevier B.V. All rights reserved.

current efficiency for general Ni Watts bath formulations generally remains around 90–97% [3]. In the last 20 years, the electrodeposition of thin Ni films has received a renewed attention from the research community [5,6]. Recent studies have been focused on the addition of different additives such as saccharin [7,8] and the use of novel plating methods such as pulse plating [9,10] in order to produce novel functional Ni coatings for different applications. The use of ultrasound in electrochemical processes [11] and electroplating in particular [12] has also been reported to improve the electrodeposition process itself and the characteristics of Ni deposits (enhancement of residual stress [13], wear resistance [14], fatigue strength [15] and hardness [16]). In this sense, Kobayasi et al. [17] found that the frequency in the low-frequency range could play a key role on improving charge transfer reaction and modifying crystal orientation (no effect = silent conditions b 100 kHz b 28 kHz b 45 kHz = highest effect) of Ni coatings electrodeposited from a Watts bath. Jensen et al. [18] studied the effect of high-frequency ultrasound (1000 kHz) on Ni deposits produced form a modified Watts bath with some surfactants and other additives (sodium lauryl sulphate, naphthalene trisulphonic acid and butyne diole) and observed that, although high-frequency ultrasound had a beneficial effect in levelling when electrodepositing Ni in deep grooves, it also had an apparently undesired effect in terms of

50

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

pitting over the surface and the appearance of a fringe pattern. This fringe pattern was not only power dependant (it was less severe at high-power, high-frequency ultrasonic irradiation), but it also was affected by the nature of the ultrasonic field (travelling/standing wave). Touyeras et al. [19] also studied the effect of high-frequency ultrasound (300 kHz, 500 kHz and 800 kHz) at different powers (from 0 to 15 W) on the electrodeposition of Ni from a Watts bath (with/without an unspecified brightener). In this case, the authors reported that the grain size of the deposit varied as a function of ultrasonic power for each frequency (the effect being more evident in the presence of a brightener). These researchers also showed how the distribution of pressure nodes/ antinodes along the substrate being plated had a strong influence on the surface morphology and hardness of the coating. Sulitanu et al. [20] evaluated the effect of high-frequency ultrasound (2000 kHz) at different powers (1 to 10 kW/m2) in a sulphate bath with a brightener. In this latter case, increasing ultrasonic power did not only seem to have a grain refinement effect, but also an enhancement of the kinetics of the electrodeposition process in terms of higher limiting currents. Nevertheless, low-power, high-frequency ultrasound was the best way to reduce the roughness of the deposits. It can be seen that none of the studies on the use of low-frequency ultrasound (≤ 100 kHz) systematically studied the effect that the ultrasonic power could have on the characteristics (visual appearance, crystal orientation of the deposits, surface morphology, grain structure and hardness) of electrodeposited Ni coatings [13–17]. In this sense, the effect of ultrasonic power on the characteristics of the Ni coatings electrodeposited under low-frequency ultrasound may be completely different than the same under high-frequency ultrasound, especially if one takes into account how different ultrasonic cavitation is (both mechanical and chemical effects) depending on the operating frequency [21,22]. Therefore, due to the lack of studies of how ultrasonic power may affect the characteristics of Ni coatings electrodeposited under low-frequency ultrasound, we here present a study focused on the effect of low-frequency ultrasound on the characteristics of thin Ni coatings electrodeposited from a Watts bath currently used in industry. The effect of the ultrasonic power on the visual appearance, crystal orientation, surface morphology, grain structure and hardness of the Ni deposits was evaluated, showing that ultrasonic power had an effect on all these properties. This paper discusses all these responses in detail and defines how low-frequency ultrasound (and the occurrence and nature of ultrasonic cavitation near the surface being plated) influences the characteristics of ultrasoundassisted electrodeposited thin Ni coatings. 2. Materials and methods 2.1. Experimental set-up For this study, an additive-free Watts bath was chosen as the plating solution (Table 1). This Watts bath, which is currently used in industry for preparing thin Ni coatings analogous to the ones here presented, is a kinetics-controlled process with a cathode current efficiency higher

Table 1 Ni Watts process used in the present study.

than 90% when operated at a current density of 4 A/dm2. This means that current plating rate is not affected by any enhancement in mass transport from the bulk solution to the cathode/electrolyte interface by mechanical or ultrasonic agitation, as demonstrated by Hyde and Compton [23] for different electrodeposition processes carried out in highly concentrated and highly conductive plating baths. C106 Cu substrates (5 × 2 × 0.12 cm, 99.9% of Cu) were used as cathodes with an approximate active area of around 4 cm2 (2 × 2 cm with the back side masked), while Ni anodes (7 × 1.4 × 0.05 cm) with an approximate active area of 20 cm2 were fabricated from 201 Ni sheets (99.0% of Ni). The plating time was 14 min in order to produce Ni coatings with a thickness of around 5–6 μm in the central area of the active surface after considering edge build-up near the edges of the active area of the cathode. In order to achieve a good adhesion between the substrate and the electrodeposited coating, the substrates were vapour-degreased for 15 min in a Dürr Ecoclean degreaser and the cathode surface was activated with an anodic acid etching process (Cu substrate acting as an anode in a solution of 30% by volume of HCl at 3 A/dm2 for 90 s) right before the electrodeposition process. All the electrodeposition experiments were conducted in a 600 mL beaker containing 500 mL of the plating solution immersed in an ultrasonic bath as shown in Fig. 1. The beaker was always placed in the centre of the bath at a controlled depth (around 11 cm between the bottom of the beaker and the surface of the water) with a constant water level (around 2 cm between the edge of the ultrasonic bath and the surface of the water) in the ultrasonic bath to ensure the reproducibility of the experiments. The distance between the cathode and the anode was around 8 cm. The bath was a QS12 ultrasonic bath operating at a frequency of around 32–38 kHz (ultrasonic transducer power: 200 W, heating power: 300 W, working capacity: 12.5 L) provided by Ultrawave. This ultrasonic bath had a built-in thermostat, enabling the control of temperature up to 70 °C. The QS12 ultrasonic bath was calibrated by the calorimetric method [24–26]: for ultrasonic output powers of 60%, 80% and 100%, the estimated ultrasonic power inside the 600 mL beaker, once immersed in the ultrasonic bath and placed in the designated area, was 0.011, 0.124 and 0.180 W/cm3, respectively. This set-up (ultrasonic bath) was chosen instead of a different one based on an ultrasonic horn due to different reasons [27]: • Ultrasonic baths are widely available at a much lower cost than horns (Langevin transducers such as those used in submersible transducer units widely used in industry for different purposes share the same basic design than those used in ultrasonic baths). • Cavitation phenomena are less violent and more uniformly distributed within an ultrasonic bath due to lower attenuation of ultrasound by cavitation than in horn-like systems.

An IPS2010 power supply unit (0 to 20 V, 1 to 10 A) from ISO-TECH was used as the rectifier, while a CAT R18 85 W overhead stirrer (110 to 2000 rpm) equipped with a 3-point propeller shaft (50 mm wide) was used in the electroplating experiments conducted under mechanical agitation. Ni deposits were produced under five different agitation conditions: silent/still (absence of agitation/ultrasound), mechanical agitation at 300 rpm, and ultrasonic irradiation at 0.011, 0.124 and 0.180 W/cm3.

Bath composition NiSO4·6H2O NiCl2·6H2O H3BO3

290 g/L 50 g/L 30 g/L

Plating conditions pH Temperature Current density

3.2 50 °C 4 A/dm2

2.2. Characterisation of the coatings Different methods were used to characterise the electroplated Ni coatings. X-ray diffraction (XRD) analysis was performed on the coatings with a Bruker D8 ADVANCE equipment to determine the effect of ultrasound on the growth direction of the crystals during the electrocrystallization, while detailed characterisation of the surface morphology and coating structure of the Ni deposits was carried

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

51

Fig. 1. Diagram showing front (A), lateral (B) and top (C) views of the experimental set-up used in the present study. Numbered elements are: 1) Cu cathode, 2) Ni anode, 3) overhead stirrer (when required), 4) 600 mL beaker containing 500 mL of Watts bath, 5) ultrasonic bath, and 6) ultrasonic transducers.

out with a FEI Nova 600 Nanolab Dualbeam Focused Ion BeamScanning Electron Microscope (FIB-SEM) system. Microhardness tests were performed on the cross-section of the deposits with a MicroWiZhard HM 221 system from Mitutoyo to evaluate the effect that the ultrasonic power could have on the hardness of the electroplated material. The tested samples were horizontally cut

near the horizontal symmetry axis of the deposit, mounted on epoxy resin and thoroughly polished. The microhardness was then evaluated in five random locations around the central area of the cross-section to avoid any effect of the Cu substrate on the measurements. A load of 2 g-force was applied for 10 s during the microhardness tests.

Fig. 2. Visual appearance of Ni coatings electrodeposited on Cu substrates under different conditions: A) silent/still, B) mechanical agitation at 300 rpm, and ultrasound at C) 0.011 W/cm3, D) 0.124 W/cm3 and E) 0.180 W/cm3. Moisture (water) stains can also be observed near the edges in some of the samples.

52

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

3. Results and discussions 3.1. Surface finish and appearance of anodes Fig. 2 shows some examples of Ni coatings produced under the five different agitation conditions studied. Under silent/still conditions and mechanical agitation, smooth and uniform deposits with good surface finish were obtained with a slightly clearer grey colour near the edges and corners of the coated Cu substrate due to edge build-up (edge build-up occurred in these areas due to locally higher current densities). Some evidence of edge build-up was also observed in the Ni coatings electrodeposited under ultrasound. However in this case the surface finish of the deposits was not as uniform as that of Ni coatings produced in the absence of ultrasound, as erosion marks due to the presence of cavitating bubbles were clearly observed in some areas of the Ni deposits. Nevertheless, while in those Ni coatings electrodeposited at either 0.011 or 0.180 W/cm3 the erosion marks were very small compared to the coated area of the Cu substrate, in those Ni deposits produced at 0.124 W/cm3 the marks were more prominent, indicating the presence of much more aggressive cavitation phenomena near the surface of the Cu substrates at this particular ultrasonic power. In all cases when ultrasound was applied, the appearance of the marks was quite random. Sonication also had a significant effect on the appearance of Ni anodes after being continuously used for the production of several samples (Fig. 3). When no ultrasound was applied during plating, uniformly-distributed corrosion pitting was observed. However, pitting distribution slightly changed when ultrasonic irradiation at 0.011 W/cm3 was introduced in the plating process as some discontinuities in the corrosion marks were observed especially in the lower half of the anode surface. This discontinuity became even more evident at higher ultrasonic powers reaching a point where three pressure antinodes were clearly seen on the surface of the anode. In fact, the cavitation activity with the ultrasonic power can be qualitatively estimated by the observation of this erosion of the anodes continuously used to produce different samples: at 0.011 W/cm3, the ultrasonic power was so low that only two pressure antinodes (antinode 1 at the bottom and antinode 2 on the central area of the anode) were noticed in the erosion marks observed in the anode. The antinodes became more evident at 0.124 W/cm3, where some erosion marks indicating the presence of randomly localised, aggressive bubble streamers were observed around them. A third pressure antinode

(antinode 3) forming near the liquid level at this power was also noticed, although it was not as evident as the other antinodes. Some erratic marks between antinodes 2 and 3 were also observed in anodes used at 0.124 W/cm3. The third antinode is clearly seen at 0.180 W/cm3, with well-defined limits within the three pressure antinodes. It is interesting to note that, while erosion marks for antinodes 1, 2 and 3 were broader and more uniform in the anodes used at 0.180 W/cm3 compared to the anodes used at 0.124 W/cm3 (as it would be expected considering that the ultrasonic power was higher), localised areas corresponding to antinodes 1 and 2 showed significantly higher erosion in the anodes used at 0.124 W/cm3 than the anodes used at 0.180 W/cm3. The distance between the consecutive antinodes was roughly around 2.1 cm, which is consistent with the establishment of a standing acoustic field inside the beaker used in the present study: by roughly approximating that the electrolyte at 50 °C has similar properties to water then, for a standing ultrasonic field with a frequency of around 32–38 kHz the wavelength or λ would be around 4.06–4.40 cm, and therefore the distance between nodes (λ/2) would roughly be around 2.03–2.20 cm. These antinode marks observed in the Ni anodes used for consecutive plating experiments (more than 3 h) are analogous to the observations made by Jensen et al. [18] and Touyeras et al. [19] on their Ni deposits (appearance of a fringe pattern in the former and different surface morphologies in the latter), as all of these different features observed in these various studies related to different Watts formulations and reflect the spatial distribution of pressure nodes/antinodes within the bath. On the other hand, the reason for the erratic and random erosion marks formed on both cathode and anode at lower-medium powers, especially at 0.124 W/cm3, is the formation of random local bubble structures attached to the surface of both cathode and anode similar to those observed by Krefting et al. in a large ultrasonic system operated at 40 kHz [28]. In their work, the authors found that these bubble structures and their effect on the surface of solids were quite random compared to freely existing structures such as jellyfish structures which were highly reproducible. In the present case, a higher ultrasonic power (0.180 W/cm3) yielded higher pressures at the antinodes, especially in the symmetry axis of the beaker, which led to the formation of larger clusters of bubbles that could ‘attract’ and remove those bubbles that would form random structures in solid surfaces and would erode the surface of the deposit as in the samples plated at 0.124 and 0.011 W/cm3 to a lesser extent.

Fig. 3. Visual appearance of Ni anodes after being used for more than 3 h at different ultrasonic powers: A) 0 W/cm3 — still, B) 0.011 W/cm3, C) 0.124 W/cm3 and D) 0.180 W/cm3.

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

53

Fig. 4. A) XRD spectra of Ni coatings electrodeposited on Cu substrates under different conditions. Peaks related to the Cu substrate are also marked. Scan rate = 0.1°/s. B) RTC(hkl) estimated for the different crystal planes observed in Ni coatings electrodeposited under different conditions. C) RTC[hkl] estimated for the different orientations related to the crystal planes observed in Ni coatings electrodeposited under different conditions.

3.2. Crystal orientation In order to have a deeper understanding of the effects of ultrasound in the properties of Ni coatings electroplated from a Watts bath, XRD analysis was conducted on the samples plated under different agitation conditions. The 2θ scans (Fig. 4A) showed that, under silent still/conditions, about the same height of (111) and (200) crystal planes were obtained, with small peaks related to (220), (311) and (222) planes. Relatively higher peaks were observed for the (111) and (311) planes in coatings electrodeposited under mechanical agitation, along with a relative decrease in the intensity of the (200) planes. Although Ni coatings electrodeposited under ultrasound at 0.011 W/cm3 also presented high intensities for (111) crystal planes, those were relatively lower in Ni deposits produced at 0.124 W/cm3 and 0.180 W/cm3, which presented significantly high peaks for (200) crystals. For a better quantification of the different crystal planes, the ‘Relative Texture Coefficient’ (RTC) method developed by Bérubé and L'Espérance [29] was employed. This method, which is commonly used to quantify crystal planes in electrodeposited metal-based coatings [30–34], yields normalized and quantitative data of the different crystal planes observed in a sample and eliminates the roughness effect of the deposits analysed. The Relative Texture Coefficient or RTC(hkl) for a (hkl) crystal plane is defined as: IðhklÞ =IðhklÞ;P RTCðhklÞ ¼ 100  X5 I =I 1 ðhklÞ ðhklÞ;P

ð1Þ

where I(hkl) is the intensity of the reflection for the (hkl) crystal plane in the analysed sample and I(hkl),P is the intensity of the reflection for the same crystal plane in a standard Ni powder sample with random orientation. The denominator in Eq. (1) is the sum of the relation between I(hkl) and I(hkl),P for all the different crystal planes observed in the XRD spectra, which for the case of Ni are (111) for 2θ ≈ 44.50°, (200) for 2θ ≈ 51.85°, (220) for 2θ ≈ 76.38°, (311) for 2θ ≈ 92.94° and (222) for 2θ ≈ 98.45° [35]. RTC(hkl) values were estimated for all the Ni crystal planes observed in each 2θ scan (Fig. 4B). In all the deposits, RTC(200) always reached the highest value, although its presence would vary from 30% in Ni electrodeposits produced under mechanical agitation to 50% in Ni coatings electrodeposited under ultrasound at 0.124 W/cm3. Regarding (220) planes, RTC(220) values always remained lower than 10% in all cases, whereas the (111), (311) and (222) crystal planes seemed to be linked to each other: taking the RTC(111), RTC(311) and RTC(222) values of the Ni coating electrodeposited under silent/still conditions as a reference, it was observed that, in all the Ni deposits where RTC(111) was higher (Ni coatings under either mechanical agitation or ultrasound at 0.011 W/cm3), RTC(311) and RTC(222) would also increase proportionately, while for all those deposits where RTC(111) was lower (Ni deposits under ultrasound at either 0.124 W/cm3 or 0.180 W/cm3), a proportional decrease was observed for RTC(311) and RTC(222). This increase/decrease would inversely be connected to the RTC(200) values. Electrocrystallization on [111] and [311] directions can be associated to a dispersed [211] preferred orientation [36–38], implying that (111), (311) and (222) crystal planes may be attributed to Ni grains with a [211] orientation. On the other hand, (200) and (220) crystal planes

54

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

depict Ni crystals growing with [100] and [110] orientations [32], respectively. Therefore, RTC[hkl] values (Fig. 4C) analogous to RTC(hkl) were obtained by defining: RTC½100 ¼ RTCð100Þ

ð2Þ

RTC½110 ¼ RTCð220Þ

ð3Þ

RTC½211 ¼ RTCð111Þ þ RTCð311Þ þ RTCð222Þ :

ð4Þ

As expected from the RTC(hkl) previously commented, there was an inverse relationship between the [100] and [211] orientations: the higher the RTC[100] is, the lower the RTC[211] gets, and vice versa. The preferential orientation of Ni crystals in the electroplated coating is under the influence of the so-called ‘inhibition effect’ [39,40]. Electrodeposition of Ni with no inhibition of crystal growth yields deposits with a [100] preferred orientation, commonly known as the ‘free’ growth mode. Amblard et al. related the inhibition of the electrocrystallization of Ni, particularly the [100] growth mode, by the presence of different species in the electrolyte–cathode interface [41–44]: • Crystal growth in [110] direction is observed when inhibition by adsorbed atomic hydrogen (Hads) covering the surface is expected. • [211] orientation is the least inhibited by the presence of colloidal and/or precipitated Ni(OH)2 near the electrolyte–cathode interface. • Electrocrystallization in [210] direction is caused by the massive presence of gaseous hydrogen (H2) due to massive hydrogen evolution at the cathode. These inhibiting species would however not be incorporated to the final deposit, as they would be “sufficiently stable on certain crystal faces to modify growth while at the same time being sufficiently unstable so as to allow decomposition and/or release so as not to be codeposited” [45]. Ni coatings produced under silent/still conditions in the present study had a combined [100] and [211] preferred orientation, with a slightly higher predominance of the latter. Significantly higher RTC[211] and lower RTC[100] values were observed for the Ni deposits electrodeposited under mechanical agitation, suggesting that the agitation of the solution ‘promoted’ the formation of [211] textures instead of the ‘free’ [100] growth mode. Such inhibition effect on the free growth mode, as suggested by Amblard et al. [41–44], is due to the presence of colloidal and precipitated Ni(OH)2 near the electrolyte– cathode interface, which can be related to a localised increase of the pH near the interface [46]. The cause for this alkalization near the electrode surface is the local discharge and/or adsorption of hydrogen [41–44]. An enhancement in the discharge of hydrogen due to a lower cathode current efficiency would explain the local increase in pH and the presence of Ni(OH)2 near the surface. In this sense, different authors have recently reported a slight decrease in the cathode current efficiency and an increase in hydrogen evolution with mechanical agitation in systems apparently under kinetic/mixed-controlled conditions [47,48]. This decrease in efficiency with agitation, although it can be quite small depending on the working conditions, may have a significant effect on the crystal structure. The other possible explanation for the increase in pH is the adsorption of hydrogen in the cathode surface [41], a phenomenon that could be enhanced by an increase in agitation, as recently suggested in different recent works [7,49]. Regarding Ni crystals with a [110], their presence was fairly small, with RTC[110] values of around 6–9%. Ni deposits produced under ultrasound at 0.011 W/cm3 presented similar results as those observed for coatings produced under mechanical agitation, although RTC[100] and RTC[211] were not as low and high, respectively. The key for this similarity between Ni coatings produced under either mechanical agitation or ultrasound at 0.011 W/cm3 is

‘agitation’: while in the first case the electrolyte is agitated by the mechanical stirring action of the overhead stirrer, in the second case the electrolyte is also agitated by the ‘acoustic streaming’ effect induced by the ultrasonic field set in the bath. One would assume that, at a higher ultrasonic power (0.124 W/cm3), an even more evident increase in the value of RTC[211] would be obtained, as stronger acoustic streaming is expected. But in this case, the intensity of the cavitation activity near the surface also takes a key role, as the presence of Ni crystals with preferred [100] is increased to a point where the [211] orientation is no longer predominant (RTC[100] N RTC[211] for Ni deposits produced under ultrasound at 0.124 W/cm3). These were the deposits that presented a highest rate of erosion marks due to the presence of cavitation in the form of randomly-attached bubble structures near the surface. A further increase in the ultrasonic field (0.180 W/cm3) may reduce the formation of such bubble structures near the surface, but the overall increase in the cavitation activity in the plating solution would still keep a high presence of crystals with preferred [100] orientation in the deposit. In addition, as in the Ni deposits produced under silent conditions, ultrasound did not have much effect on the electrocrystallization of Ni in the [110] direction, as RTC[110] values again remained around 6–9%. These results agree to some extent with the observations made by Kobayasi et al. [17], as they also noticed a similar trend regarding the crystal orientation with different ultrasonic frequencies: i) increase in the proportion of Ni crystals with a [100] orientation, ii) decrease in the proportion of Ni crystals with a [211] orientation, and iii) a low, fairly constant proportion of Ni crystals with a [110] orientation. In their case though, they suggested that this finding was due to the hypothetical effect that ultrasound could have on the activation energies of the surface diffusion of Ni adions to the different crystal faces, ignoring the ‘inhibition effect’ caused by the presence of inhibiting species in the cathode–electrolyte interface. Overall, these results suggest that the presence of cavitating bubbles near the cathode surface counteracts the effect of agitating the plating solution, which would apparently reduce the adsorption of hydrogen on the surface of the cathode. Local alkalization of the electrolyte near the cathode–electrolyte interface would be reduced, which would result in reduced formation of Ni(OH)2 near the surface, and therefore, a less inhibited electrocrystallization of Ni with higher presence of Ni crystals with [100] orientation. Nevertheless, the effect of cavitation phenomena near the electrode on the double layer must not be discarded. The presence of cavitating bubbles may disturb the formation of a double layer structure on the electrode surface, and it could result in either a decrease in the formation of precipitated and colloidal Ni(OH)2 or its removal from the cathode–electrolyte interface. 3.3. Surface morphology and microstructure FIB-SEM analysis was performed to observe any significant effect of ultrasound on the structure of the Ni deposits produced under ultrasound. Tilted FIB-SEM images of the surface of Ni deposits electroplated under different conditions are shown in Fig. 5. Irregular grooves and nodule-shaped structures with a cauliflower-like appearance were noticed in all the deposits. This latter morphology is quite common in Ni electroplating, as it has previously been reported in many previous different studies [50,51]. Nevertheless, some variations were observed in the coatings plated under the different conditions: for the coatings plated in the absence of ultrasound, irregular grooves were clearly observed all over the surface, more prominently in those coatings plated under mechanical agitation, while the presence of nodule-shaped structures was minimal. Such nodule-shaped structures were far more evident in the coatings electroplated under ultrasound at 0.011 and 0.180 W/cm3, where irregular grooves were harder to locate. For the coatings electroplated under ultrasound at 0.180 W/cm3 though, the nodule-shaped structures were not as numerous as in the deposits produced at the other ultrasonic powers.

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

55

Fig. 5. Tilted FIB-SEM images of the surface of Ni coatings electrodeposited on Cu substrates under different conditions: A) silent/still, B) mechanical agitation at 300 rpm, and ultrasound at C) 0.011 W/cm3, D) 0.124 W/cm3 and E) 0.180 W/cm3.

All of these results agree, to some extent, with previous studies carried out by Touyeras et al. [19]. In this latter case, the authors observed that, in the areas of their Ni deposits corresponding to the existence of

pressure nodes, the surface morphology was very similar to that of our Ni deposits produced with mechanical agitation, as very similar grooves were clearly seen over that area. This ‘coincidence’ is caused

Fig. 6. High magnification SEM images of the surface of Ni coatings electrodeposited on Cu substrates under different conditions: A) silent/still, B) mechanical agitation at 300 rpm, and ultrasound at C) 0.011 W/cm3, D) 0.124 W/cm3 and E) 0.180 W/cm3.

56

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

by the fact that, in pressure nodes, although they would not expect any cavitation at all, stirring of the solution is still quite important, and hence the surface morphology with high formation of grooves. A massive presence of nodules was noticed in the areas of their Ni deposits corresponding to the existence of pressure antinodes. These nodules seemed quite similar to the nodular structures observed over the surface of the ultrasonic-assisted electrodeposited Ni coatings here presented. In our case though, the nodules were not as massive and prominent as those observed by Touyeras et al. [19]. This could be caused by the frequency they applied (high-frequency ultrasound at either 300, 500 or 800 kHz) and the consequent changes in cavitation phenomena [52] (less mechanical effects/more chemical effects, less violent collapse of bubbles, etc.), although the differences in the Watts formulation used, the presence of a brightener and plating operational parameters (current density, temperature, etc.) could also have an effect. Fig. 6 displays high-magnification FIB-SEM images of the coatings plated under the different agitation conditions, which reflect somehow the results obtained in the XRD analysis. Although the relation between grain orientation and surface morphology may sound rather subjective, large Ni grains with an irregular structure similar to that of crystals with a [100] preferred orientation reported by other authors [31,32,37,53] were routinely seen in the coatings produced under silent/still conditions, along with smaller grains with a regular pyramidal structure and binary symmetry similar to that of crystals with a clearly defined [211] orientation reported elsewhere [31,32,37], agreeing to some extent with the XRD analysis results previously reported in this same paper. Although the assumed [211] textures were rather small, the apparent [100] structures showed a wide range of sizes similar to those previously reported by Vicenzo and Cavalloti [53]. The surface morphology apparently changed for the deposits plated under mechanical agitation, as the presence of grains with a structure similar to that of Ni deposits with a [211] preferred orientation reported by Pavlatou and co-workers [31,32,37] became more evident, while crystals with apparent [100] and [211] orientations were observed in the coatings plated under ultrasound at 0.011 W/cm3. However, fewer grains with a structure similar to that of Ni crystals with a [211] orientation were noticed in the surface of the deposits plated under ultrasound at 0.124 and 0.180 W/cm3. In these latter cases, the largest [100] structures were smaller than the largest ones of the same type noticed under silent/ still conditions, and the proportion of very small crystals seemed to be greater, especially in the deposits plated at 0.124 W/cm3. The presence of these very fine crystals in the surface of coatings deposited under ultrasound roughly gives an idea of the grain refinement that may be achieved by ultrasound during the electrodeposition process. Very few grains with a pseudo-pentagonal crystal symmetry resembling crystals with a [110] orientation such as those reported by other authors [31,32, 37] were observed in all the samples evaluated, agreeing to some extent with the low RTC[110] values previously estimated from the XRD spectra here presented. Further evaluation of the cross-section of the coatings electrodeposited under silent/still conditions, mechanical agitation and under ultrasound was performed to check the effect of ultrasonic irradiation on the microstructure of Ni coatings (Fig. 7). Whereas the Ni coatings electrodeposited under silent/still conditions exhibited a characteristic columnar structure mainly consisting of large crystals, the Ni deposits produced under mechanical agitation consisted of a combination of columnar crystals and smaller grains. Related to the latter, both columnar crystals and smaller grains apparently exhibited a higher aspect ratio than the large columnar crystals observed in the deposits produced under silent/still conditions. Regarding the cross-section of Ni deposits produced under ultrasound, Ni coatings electrodeposited at 0.011 W/cm3 exhibited a columnar structure quite similar to that of Ni deposits produced under silent/still conditions, although quite a few small grains could be noticed in the coatings produced with the presence of ultrasound. The presence of small grains was more

noticeable in the Ni coatings electrodeposited under ultrasound at either 0.124 W/cm3 or 0.180 W/cm3, as a very significant proportion of very small grains were present along with some large columnar crystals. Grain refinement and structure modification by ultrasound have already been widely suggested in the past [54,55], although most of the works available in the literature related to electrodeposited Ni coatings only show either some evidence of the grain refinement effect by ultrasound on the surface of the coatings [20,56,57] or did not show any real evidence at all [19]. Regarding this, Lampke et al.'s backscattered electron beam diffraction (EBSD) results showed that, whereas large Ni crystals were still noticeable in Ni deposits produced under ultrasound, many small crystals could be observed, resulting in an overall grain refinement when compared with Ni coatings electrodeposited in the absence of ultrasound [58]. A very similar effect in the grain structure was observed in the present study, as the use of ultrasound during the electrodeposition, particularly at high powers, resulted in a more fragmented structure where columnar crystals were observed alternating with very small grains. 3.4. Hardness Microhardness tests were performed on the different Ni deposits (Fig. 8) in order to initially observe any effect that the previously suggested modification of the structure of the deposit by ultrasound may have on the mechanical properties of the material. The coatings plated under silent/still conditions showed the lowest microhardness values, quite similar to the values measured for the deposits plated under mechanical agitation, while the coatings produced under ultrasound exhibited higher hardness. Ni deposits plated at 0.124 W/cm3 presented the highest microhardness (27% more than electrodeposited Ni under silent/still conditions), followed by those plated at 0.180 W/cm3 (10% more than electrodeposited Ni under silent/still conditions) and 0.011 W/cm3 (around 5% more than electrodeposited Ni under silent/ still conditions). Ni crystals with the [100] preferred orientation are commonly associated to a ductile behaviour [59,60] with less brittleness [61] and lower values of internal stress and hardness [62], while the [211] growth mode is generally related to higher values of hardness [63]. However, no significant increase in hardness was observed in the deposits plated under mechanical agitation in the present study, which exhibited a higher proportion of [211] textures compared to the coatings produced under silent/still conditions, which had a similar proportion of crystals with [100] and [211] preferred orientations. On the other hand, the increase in hardness of the coatings deposited under ultrasound is concurrent with previous studies by Prasad et al. [16] (from 235 HV to 332 HV in 50-micron-thick Ni coatings deposited on mild steel cathodes in a Watts bath at 4 A/dm2 in a 22 kHz ultrasonic bath), Zanella et al. [56] (from ≈ 300 HV to ≈ 320 HV in Ni coatings plated onto a low carbon steel substrate in a Watts bath at 2 A/dm2 under undefined ultrasonic conditions), and García-Lecina et al. [57] (from ≈ 250 HV to ≈ 300 HV in Ni coatings plated onto mild steel cathodes in a Watts bath at 5 A/dm2 with a 24 kHz ultrasonic horn). In this latter case, the authors pointed out that the increase in hardness would be due to a change in the structure of the deposit, as previously suggested by Lampke et al. [58]. The higher presence of smaller grains on the surface of Ni coatings electrodeposited under ultrasound at 0.124 W/cm3 and 0.180 W/cm3 and the grain size refinement and fragmentation of the microstructure observed in those same deposits would therefore be responsible to some extent for the increase in hardness observed in the thin Ni coatings electrodeposited under ultrasound at higher power. Nevertheless, structure modification was also observed in the Ni coatings electrodeposited under mechanical agitation here presented, although these deposits did not exhibit any significant increase in hardness. This implies that the presence of ultrasound during the electrodeposition must have another effect which may result in the

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

57

Fig. 7. FIB-SEM images of the cross-section of Ni coatings electrodeposited on Cu substrates under different conditions: A) silent/still, B) mechanical agitation at 300 rpm, and ultrasound at C) 0.011 W/cm3, D) 0.124 W/cm3 and E) 0.180 W/cm3.

increase in hardness here and elsewhere reported. In this sense, Walker and Walker suggested that work hardening during electrodeposition could also occur due to the mechanical action exerted by ultrasonic cavitation near the cathode [64]. Work hardening by cavitation near the surface would also explain why the Ni coatings electrodeposited at 0.124 W/cm 3 , the ones with erosion marks caused by random bubble structures attached to the surface,

exhibited a significantly higher hardness than those electrodeposited at 0.180 W/cm3. 4. Conclusions The surface finish of the electrodeposited coatings showed that the presence of ultrasonic cavitation near the electrode affects the surface

58

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59

systems or low-frequency, high-power systems such as horns), can be a concerning issue when using ultrasonic systems more suitable to be scaled-up to large plating baths (i.e. Langevin-like transducers such as those used in submersible transducer units widely used in industry). Acknowledgements The authors acknowledge the Technology Strategy Board and the Engineering and Physical Sciences Research Council in United Kingdom for their funding through the Knowledge Transfer Partnership scheme. The authors would also like to thank Dr M. Lesley Wears from the University of Exeter and Dr Geoff West from Loughborough University for their assistance with XRD and FIB-SEM analysis, respectively. I.T. wishes to thank Prof. Andreas Bund and his group at Technische Universität Ilmenau for hosting him during his PhD studies. The authors would also like to thank Dr M. Camargo and Dr R. Griesler from Technische Universität Ilmenau for insightful discussions on XRD analysis and crystal orientation. References Fig. 8. Microhardness values measured for Ni coatings electrodeposited on Cu substrates under different conditions. Load = 2 g-force, load time = 10 s.

finish of the coatings, provoking the appearance of local marks where bubble spots were easily observed. Ultrasound also modified the orientation of the deposit, counteracting the effect that the agitation of the solution may have on the preferred growth mode. Related to this, XRD analysis showed that, while the agitation of the solution favoured the presence of [211] textures, the presence of ultrasonic cavitation promotes the electrocrystallization of Ni in the [100] preferred orientation. This trend was also observed in FIB-SEM images of the surface of the different deposits, which also pointed a significant presence of very small crystals under ultrasound at higher ultrasonic powers. FIB-SEM images of the cross-section of the coatings confirmed this effect of ultrasound on modifying the microstructure of Ni deposits, resulting in a more fragmented metal matrix with less columnar crystals and more refined grains. Such effect of ultrasound in refining the grain size, along with its work-hardening effect during the electrodeposition, resulted in an increase in the hardness of the Ni deposits. The combination of an increased hardness of the Ni deposits plated under ultrasound with the higher proportion of [100] textures that are associated to a ductile behaviour with less brittleness would be in no doubt of interest for the production of functional electroplated Ni deposits in industry, especially in novel areas such as energy/photovoltaic applications [65]. Although the effect of ultrasound on the characteristics of the electrodeposited Ni coatings becomes more evident at higher powers, it was found that an intermediate power, 0.124 W/cm3, had the highest effect. This was not due to a higher overall cavitation activity in the plating bath (the calorimetry method used to characterise how much ultrasound was introduced to the plating vessel used here and the antinode marks observed in the Ni anodes used would indicate otherwise), but due to the formation of random bubble structures attached to the electrode surface, which caused a higher modification of the characteristics of the electrodeposited Ni coatings, including the appearance of local marks where bubble spots were easily observed. These random local bubble structures attached to the surface of the cathode had an undesired impact on the visual appearance and uniformity of the coating. Regarding this, the present paper highlights how important the nature of the ‘real’ cavitation activity near the surface of the cathode can be (even more than the ultrasonic power itself) when scaling-up the ultrasound-assisted electrodeposition of metals. In this sense, the appearance of the random local bubble structures previously mentioned, which may not really occur or remain unnoticed in small laboratory set-ups (particularly when using high-frequency systems, low-power

[1] G. Bird, Phil. Trans. R. Soc. London 127 (1837) 37–45. [2] Y.D. Gamburg, G. Zangari, Theory and Practice of Metal Electrodeposition, Springer, New York-Dordrecht-Heidelberg-London, 2011. [3] G.A. Di Bari, in: M. Schlesinger, M. Paunovic (Eds.), Modern Electroplating, 5th ed.John Wiley & Sons, Hoboken, 2010. [4] O.P. Watts, Trans. Am. Electrochem. Soc. 29 (1916) 395–403. [5] J.-d. Li, P. Zhang, Y.-h. Wu, Y.-s. Liu, M. Xuan, Microsyst. Technol. 15 (2009) 505–510. [6] M. Boubatra, A. Azizi, G. Schmerber, A. Dinia, Ionics 18 (2012) 425–432. [7] E. Moti', M.H. Shariat, M.E. Bahrololoom, Mater. Chem. Phys. 111 (2008) 469–474. [8] A.M. Rashidi, A. Amadeh, Surf. Coat. Technol. 204 (2009) 353–358. [9] R. Mishra, R. Balasubramaniam, Corros. Sci. 46 (2004) 3019–3029. [10] Y. Xuetao, W. Yu, S. Dongbai, Y. Hongying, Surf. Coat. Technol. 202 (2008) 1895–1903. [11] T.J. Mason, J.P. Lorimer, D.J. Walton, Ultrasonics 28 (1990) 333–337. [12] Hihn, et al., in: B.G. Pollet (Ed.), Power Ultrasound in Electrochemistry: From Versatile Laboratory Tool to Engineering Solution, John Wiley & Sons, Chichester, 2011. [13] P.B.S.N.V. Prasad, R. Vasudevan, S.K. Seshadri, J. Mater. Sci. Lett. 11 (1992) 1424–1425. [14] P.B.S.N.V. Prasad, R. Vasudevan, S.K. Seshadri, J. Mater. Sci. Lett. 12 (1993) 902–903. [15] P.B.S.N.V. Prasad, S. Ahila, R. Vasudevan, S.K. Seshadri, J. Mater. Sci. Lett. 13 (1994) 15–16. [16] P.B.S.N.V. Prasad, R. Vasudevan, S.K. Seshadri, S. Ahila, Mater. Lett. 17 (1993) 357–359. [17] K. Kobayasi, A. Chiba, N. Minami, Ultrasonics 38 (2000) 676–681. [18] J.A.D. Jensen, P. Pocwiardowski, P.O.A. Persson, L. Hultman, P. Møller, Chem. Phys. Lett. 368 (2003) 732–737. [19] F. Touyeras, J.Y. Hihn, X. Bourgoin, B. Jacques, L. Hallez, V. Branger, Ultrason. Sonochem. 12 (2005) 13–19. [20] N. Sulitanu, C. Pirghie, I. Sandu, Rev. Chim. 58 (2007) 20–24. [21] T.J. Mason, A.J. Cobley, J.E. Graves, D. Morgan, Ultrason. Sonochem. 18 (2011) 226–230. [22] K.V.B. Tran, T. Kimura, T. Kondo, S. Koda, Ultrason. Sonochem. 21 (2014) 716–721. [23] M.E. Hyde, R.G. Compton, J. Electroanal. Chem. 531 (2002) 19–24. [24] T.J. Mason, Practical Sonochemistry: User's Guide to Applications in Chemistry and Chemical Engineering, Ellis Horword Ltd., Chichester, 1991. [25] Ratoarinoro, F. Contamine, A.M. Wilhelm, J. Berlan, H. Delmas, Ultrason. Sonochem. 2 (1995) S43–S47. [26] O. Louisnard, J. González-García, in: H. Feng, G. Barbosa-Canovas, J. Weiss (Eds.), Ultrasound Technologies for Food and Bioprocessing, Springer, New York-DordrechtHeidelberg-London, 2011. [27] I. Tudela, Y. Zhang, M. Pal, I. Kerr, A.J. Cobley, Surf. Coat. Technol. 259 (2014) 363–373. [28] D. Krefting, R. Mettin, W. Lauterborn, Ultrason. Sonochem. 11 (2004) 119–123. [29] L.Ph. Bérubé, G. L'Espérance, J. Electrochem. Soc. 136 (1989) 2314–2315. [30] S. Surviliene, V. Jasulaitiene, A. Lisowska-Oleksiak, V.A. Safonov, J. Appl. Electrochem. 35 (2005) 9–15. [31] E.A. Pavlatou, N. Spyrellis, Russ. J. Electrochem. 44 (2008) 745–754. [32] S. Spanou, E.A. Pavlatou, N. Spyrellis, Electrochim. Acta 54 (2009) 2547–2555. [33] C.N. Panagopoulos, E.P. Georgiou, A. Tsopani, L. Piperi, Appl. Surf. Sci. 257 (2011) 4769–4773. [34] M.K. Camargo, U. Schmidt, R. Grieseler, M. Wilke, A. Bund, J. Electrochem. Soc. 161 (2014) D168–D175. [35] Inorganic Crystal Structure Database code 64989, Fachinformationszentrum Karlsruhe (Germany) and National Institute of Standards and Technology (United States of America). [36] J. Amblard, M. Froment, G. Maurin, Electrodepos. Surf. Treat. 2 (1974) 205–222. [37] E.A. Pavlatou, M. Raptakis, N. Spyrellis, Surf. Coat. Technol. 201 (2007) 4571–4577. [38] S. Spanou, E.A. Pavlatou, J. Appl. Electrochem. 40 (2010) 1325–1336. [39] H. Fischer, Electrodepos. Surf. Treat. 1 (1973) 239–251.

I. Tudela et al. / Surface & Coatings Technology 264 (2015) 49–59 [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53]

H. Fischer, Electrodepos. Surf. Treat. 1 (1973) 319–337. J. Amblard, M. Froment, N. Spyrellis, Surf. Technol. 5 (1977) 205–234. J. Amblard, M. Froment, Faraday Symp. Chem. Soc. 12 (1977) 136–144. J. Amblard, I. Epelboin, M. Froment, G. Maurin, J. Appl. Electrochem. 9 (1979) 233–242. J. Amblard, M. Froment, G. Maurin, N. Spyrellis, E. Trevisan-Souteyrand, Electrochim. Acta 28 (1983) 909–915. C.B. Nielsen, A. Horsewell, M.J.L. Østergård, J. Appl. Electrochem. 27 (1997) 839–845. A.G. Ives, J.W. Edington, G.P. Rothwell, Electrochim. Acta 15 (1970) 1797–1801. B. Bozzini, Mater. Chem. Phys. 66 (2000) 278–285. X.-l. Ren, Q.-f. Wei, Z. Liu, J. Liu, Trans. Nonferrous Metals Soc. China 22 (2012) 467–475. E. Moti', M.H. Shariat, M.E. Bahrololoom, J. Appl. Electrochem. 38 (2008) 605–612. J. Kubisztal, A. Budniok, A. Lasia, Int. J. Hydrogen Energy 32 (2007) 1211–1218. Y.-w. Song, D.-y. Shan, R.-s. Chen, E.-h. Han, Trans. Nonferrous Metals Soc. China 18 (2008) s339–s343. T.J. Mason, J.P. Lorimer, Applied Sonochemistry: The Use of Power Ultrasound in Chemistry and Processing, Wiley-VCH, Weinheim, 2002. A. Vicenzo, P.L. Cavalloti, Russ. J. Electrochem. 44 (2008) 716–727.

59

[54] C.T. Walker, R. Walker, Electrodepos. Surf. Treat. 1 (1973) 457–469. [55] V. Sáez, M.D. Esclapez, A.J. Frías-Ferrer, P. Bonete, I. Tudela, M.I. Díez-García, J. González-García, Ultrason. Sonochem. 18 (2011) 873–880. [56] C. Zanella, M. Lekka, P.L. Bonora, Surf. Eng. 26 (2010) 511–518. [57] E. García-Lecina, I. García-Urrutia, J.A. Díez, J. Morgiel, P. Indyka, Surf. Coat. Technol. 206 (2012) 2998–3005. [58] T. Lampke, D. Dietrich, A. Leopold, G. Alisch, B. Wielage, Surf. Coat. Technol. 202 (2008) 3967–3974. [59] I. Matsui, Y. Takigawa, T. Uesugi, K. Higashi, Mater. Lett. 99 (2013) 65–67. [60] I. Matsui, Y. Takigawa, T. Uesugi, K. Higashi, Mater. Sci. Eng. A 578 (2013) 318–322. [61] F. Denise, H. Leidheiser Jr., J. Electrochem. Soc. 100 (1953) 490–495. [62] D.J. Evans, Trans. Faraday Soc. 54 (1958) 1086–1091. [63] N.A. Badarulzaman, A.A. Mohamad, S. Puwadaria, Z.A. Ahmad, J. Coat. Technol. Res. 7 (2010) 815–820. [64] R. Walker, C.T. Walker, Nature 244 (1973) 141–142. [65] N.P. Klochko, G.S. Khrypunov, V.R. Kopach, I.I. Tyukhov, K.S. Klepikova, M.V. Kirichenko, V.M. Lyubov, Sol. Energy 98 (2013) 384–391.