Pulse plating of Pd–Ag alloy films from deep eutectic solvents

Introduction

Palladium and its alloys are important for industrial applications because of its wear and corrosion resistance and high catalytic activity. Pd–Ag coatings are used as contact materials in the electrical engineering industry and as decorative coatings in the jewellery industry. Other applications of Pd–Ag alloys that have been intensively studied in the recent years include H₂ sensing (detecting) [1-5], renewable hydrogen storage material [6] and catalysts for the alcohol oxidation reaction [7,8].

Electrodeposition is well known for being an economical and uncomplicated method for creating metal and alloy films. The composition, morphology and thickness of the films can be varied through the applied current density or deposition potential. In aqueous solutions, the electrodeposition of Pd is strongly dependent on the pH value of the electrolyte. At a low pH, the hydrogen evolution reaction takes place during Pd electrodeposition, causing hydrogen embrittlement problems. Non-aqueous solutions with wide electrochemical windows and good conductivities could resolve this problem. In recent years, moisture and air-stable deep eutectic solvents (DES) [9,10] have attracted intense interest as alternative electrolyte systems for the electrodeposition of various transition metals (Ag, Cu, Co, Ni, Pd, Sn, In, Ru or Zn) [11-22] and alloys (CoSm, ZnNi, ZnCo or CuSn) [23-27], because of their ease of handling and the interesting physical–chemical properties of the electrolyte solutions. In comparison to organic solvents [28-30] DES solutions based on choline chloride (ChCl) and urea (UA) are inexpensive, non-toxic and the individual chemicals are widely available. Due to the relatively high viscosity (∼42 mPa s at 70°C) and moderate electrical conductivity (∼8.4 mS cm⁻¹ at 70°C) [13] of ChCl/UA/PdCl₂ based DES this type of electrolyte requires increased temperatures (60–90°C) and enhanced electrolyte agitation to provide sufficient mass transport of metal ions to the electrode surface [12].

Beside galvanostatic deposition, pulse current deposition (PP) is one of the techniques commonly used in electroplating. Pulse plating has several advantages over conventional plating [31-34]. It is known that morphology, microstructure, hardness, ductility, porosity and the surface roughness of electrodeposits are impacted by the process parameters [31-33] and that PP can refine such deposition properties. PP also yields a finer homogenous surface appearance because it is possible to achieve higher instantaneous current densities during electrodeposition. Until now, there have been relatively few publications of electrodeposition from DES electrolytes under PP conditions [12,19,33,34].

On the other hand, the complexity of reducing metal complex ions in ionic liquids and DES is increased due to the relatively slow charge transfer kinetics [12,22]. There have only been a few publications that have reported about electroplating from DES with organic additives [12,18,20,33,35-38]. The very distinct and complex structure of the double layer at the interface of the electrode and the DES solution (or ionic liquid) may affect the electrodeposition process. It has been proven both experimentally and theoretically that the ionic components of such solutions are highly structured, forming several layers within the ‘double layer’ [39-41]. Depending on electrode polarisation, arrangement and conformation of the electrolyte species can change, resulting in reduction processes with high overpotentials. The addition of surface active compounds and the implementation of pulse current have been proposed to overcome these drawbacks and affect the rate of the charge transfer reaction and the structure of the electrical double layer, at the interface of the electrode and DES solution [12,19,34,37,38].

In this study, the electrodeposition of Pd–Ag alloys was investigated in a choline chloride/urea based DES with and without a surface active compound (sulfosalicylic acid dihydrate – SSS), at ambient atmosphere and intense hydrodynamic conditions (rotating electrode and stirred electrolyte solution). The electrochemical reduction of Pd and Ag was studied using cyclic voltammetry (CV). Thin films of Pd–Ag alloys were electrodeposited with pulse plating on copper rod substrates. The effects of various Ag concentrations within the Pd electrolytes and working with or without a surface active compound on layer thickness, composition and morphology of the produced Pd–Ag alloy electrodeposits were examined.

Materials and methods

Chemicals. The simple electrolyte solution was a deep eutectic mixture of choline chloride (ChCl) (Sigma, 98%) and urea (UA) (AppliChem, 99%) with a molar ratio of 1:2. The mixture was processed at 70°C under partial vacuum for 4 h to remove traces of amines and to reduce the moiety of water in the solution. PdCl₂ (Alfa Aesar, 99%) was directly dissolved in the initial DES solution. The AgCl (Alfa Aesar, 99%) was dissolved directly into DES/PdCl₂ solution at 50°C while stirring with a magnetic stir bar. The metal salt concentrations were as follows: 5 g L⁻¹ Pd(II) and 0.3, 0.6, 1.2 and 2.4 g L⁻¹ Ag(I). 0.5 g L⁻¹ sulfosalicylic acid dihydrate (SSS) (Sigma, 98%) was added to some of the electrolyte solutions as a surface active compound. 0.1 M KOH (Sigma-Aldrich, p.a.) and 0.3 M CH₃OH (Sigma-Aldrich, p.a.) were used for the methanol oxidation reaction.

Apparatus. All experiments were performed at 70 ± 1°C under ambient atmospheric conditions. The electrochemical experiments were conducted with a model SP-150 potentiostat/galvanostat (BioLogic) and controlled by EC-Lab Software (BioLogic). A three-electrode cell (V = 25 ml) was used for electrochemical experiments. For CV the working electrode (WE) was a glassy carbon (GC) disc electrode (A = 0.07 cm²) and the counter electrode (CE) was a palladium plate (A = 0.7 cm²). The quasi-reference electrode (RE) was a platinum wire (Ø = 1 mm, 99.9%) immersed directly in the electrolyte solution. A Ag/AgCl RE was used for the methanol oxidation. PP experiments were conducted on copper rods (Ø = 4 mm; Length = 40 mm; A = 1.28 cm²). Cu was chosen as the substrate for this work because it is often used as a base material for electronic and jewellery applications, Cu is easy to handle and is suitable for investigating the physical properties of the PdAg depositions. A palladium plate (Alfa Aesar, 99,99%, A = 0.7 cm²) was used as a soluble anode. The square-wave current pulses used in the PP experiments had an average current density of j_av = 0.2 mA cm⁻², on-/off-times of t_on/t_off  = 1 s/1 s and a duty cycle of θ = 50%. The current parameters were optimised in a previous work [42]. All experiments were performed under intensified hydrodynamic conditions. The rotating electrode (RDE) (Radiometer Analytical) rotated at 800 rpm. Additionally, the electrolyte solution was stirred by a magnetic stirrer (MR) (Heidolph Instruments) with a bar magnet at a rotational rate of 200 rpm.

The surface morphology and chemical composition of the electrodeposits were determined using high resolution scanning electron microscope (SEM, Gemini SEM 300, Zeiss) with energy dispersive spectroscopy (EDS) operated at 15 kV. The layer thickness of the produced Pd–Ag films was determined using energy dispersive X-ray fluorescence spectrometry (EDXRF, Fischerscope XDAL, Fischer). For phase analyses, X-ray diffraction (XRD) was performed with Co Kα radiation (D8 Discover, Bruker AXS, tube parameters 40 kV, 40 mA, 1D detector Lynxeye XE) and the Powder-diffraction file (PDF) database 2014 was used for phase assignment. Crystallite sizes were evaluated using line-profile analyses with consideration of microstrains and device-related line-broadening effects (software Topas, Bruker AXS).

The hardness of the coatings was measured with a Fischer H100c XYp Indentor: Vickers Norm: DIN EN ISO 14577.

The adhesiveness of the layers was tested using DIN 58196–6 from the ‘Deutsche Norm’, ‘Optical coatings – Part 6: Testing of the adhesion with a tape (tape test)’, July 1995.

Results and discussion

Cyclic voltammetry

Figure 1(a) presents cyclic voltammograms recorded on GC electrodes for the initial electrolyte solution (ChCl/UA/SSS) with the addition of Pd(II), Ag(I) and Pd(II) + Ag(I) ions at 70°C. The reduction of Pd(II) to Pd takes place at −0.59 V and the anodic peak for dissolution of pure Pd deposit at +0.33 V. The reduction of Ag(I) takes place at −0.86 V and the anodic dissolution peak of pure Ag deposit occurs at −0.17 V. In contrast, Tai et al. reported that in the ionic liquid 1-Ethyl-3-metylimidazolium chloride/tetra-fluoroborate (Emim-Cl-BF₄) the reduction of Ag(I) proceeds at a more positive potential compared to the reduction of Pd(II) [43]. The Ag(I) reduction peak had a smaller current density than Pd(II) and therefore indicates that the Ag deposition from a pure Ag electrolyte is inhibited. The observed nucleation loop in the voltammograms is typical for electrodeposition processes requiring high nucleation overpotentials [43]. At a scan rate of 50 mV s⁻¹ it was not possible to detect two separated reduction peaks for Pd(II) and Ag(I) in the voltammogram for the Pd–Ag alloy (Figure 1(a), dotted line). The reduction peak for the Pd–Ag alloy occurred at approximately −0.8 V. In the oxidation part of the Pd–Ag curve progression, the anodic peak corresponding to pure Pd stripping decreased significantly (→ a2) to +0.12 V and a new ‘peak shoulder’ appeared at approximately +0.53 V (→ a3). Furthermore, the anodic peak for the dissolution of pure Ag is replaced by an anodic peak at a further positive potential (→ a1) of +0.04 V and its peak current density is larger than that of the anodic peak from pure Ag(I) (Figure 1(a), dashed line). Most likely peaks a1 and a2 are the dissolution peaks of Ag and Pd from the mixed Pd–Ag crystallites. These changes in the voltammetric features indicate the formation of Pd–Ag alloy and are in agreement with the investigations of Tai et al. [43].

OPEN IN VIEWER

Figure 1

(a) CVs of Ag(I), Pd(II) and Pd(II) + Ag(I) in DES and (b) CVs of Pd(II) with different concentrations of Ag(I) in DES at GC electrode, T = 70°C and scan rate = 50 mV s⁻¹.

Voltammograms from Pd–Ag electrolytes with reduced Ag(I) concentrations (0.3 and 0.6 g L⁻¹) exhibited only one distinct anodic peak at approximately +0.15 V (Figure 1(b), solid and dashed line). The addition of 1.2 g L⁻¹ Ag(I) to the Pd electrolyte revealed in the anodic part of the curve a second peak (small shoulder around −0.03 V, see Figure 1(b), dotted line). It is predicted that this peak corresponds to the peak (a1) shown in Figure 1(a). Increasing the Ag(I) concentration from 0.3 g L⁻¹–1.2 g L⁻¹ shifted the reduction potential positively and enhanced the cathodic peak current for the Pd–Ag alloy deposition. Further increases to the Ag(I) concentration from 1.2 g L⁻¹– 2.4 g L⁻¹ did not further affect the deposition potential for the Pd–Ag alloy formation but led to the rise of peak (a3) in Figure 1(a).

The electrochemical deposition features of Pd–Ag alloys from the investigated DES system in this study are comparable to that from the Emim-Cl-BF₄ reported by Tai et al. [42]. In fact, there was also observed an appearance of three peaks in the anodic potential scan.

Electrodeposition and characterisation (SEM)

Electrolyte without additive. The surface morphology of the Pd–Ag coatings with a relatively high Ag content (50 wt-% Ag) were examined in the SEM at three different positions along the vertical direction of each sample: 2.5, 4.5 and 6.5 mm from the upper front face of the cylindrical substrates. This was necessary because layers with high Ag content deposited a striped pattern and the surface colour changed from dark to light-grey. It is assumed that the horizontal stripes on the cylindrical substrates are also related with the rotation rate of the electrode and agitation of the viscous electrolyte. Near the upper front face (2.5 mm distance) the surface morphology was relatively rough, coarse-grained and populated with distinct dendrites. As the distance from the front face increased the surface morphology changed from dendrite rich to a more compact and dense Pd–Ag layer (see image (a) + (b) in Figure 2). The consistent layer composition in various sample areas is something to highlight. Pd–Ag layers with 20 wt-% Ag were matt, uniform grey in colour and had a coarse globular-grained structure (results not shown here).

OPEN IN VIEWER

Figure 2

Surface morphology of two Pd–Ag coatings from DES without (a,b) and with (c,d) organic additive at two different surface sites: (a) 4.5 mm distance to front face, Pd–Ag (50:50 wt-%), (b) 6.5 mm distance, Pd–Ag (50:50 wt-%), (c) 4.5 mm distance, Pd–Ag (60:40 wt-%), (d) 6.5 mm distance, Pd–Ag (60:40 wt-%).

Electrolyte with Additive. It is well known that conventional aqueous electrolytes for electrolysis contain numerous chemical components (such as: organic additives, inhibitors, conduction salts or surfactants) to improve the overall performance of the electrolyte system. In this study, sulfosalicylic acid dihydrate (SSS) was added to the DES to improve the metal deposition rate, stabilise the coordination environment of the Pd(II) ions and change the electrode polarisation during electrolysis. After adding 0.5 g L⁻¹ SSS to the electrolyte, layers were produced with reduced dendrites. The dendrites found near the front face were significantly reduced in size in comparison to the deposits from DES without additive. The other areas of the samples had fewer coarse-grains and were more compact and homogeneous (compare image (b) with (d) in Figure 2). The presence of the selected organic additive decreased the Ag content in the Pd–Ag deposits from 50 wt-% to about 40 wt-% Ag. The layer composition across the sample surface remained at about 40 wt-% Ag.

These results demonstrate that an optimisation of the deposition parameters is necessary to substantially improve the metal deposition rate and quality of Pd–Ag coatings electrodeposited from non-aqueous DES solutions. This includes: working electrode polarisation (addition of surface active compound), electrolyte agitation (mass transport enhancement) and current mode (pulse plating).

Phase composition (XRD)

Figure 3 displays the diffractograms of two Pd–Ag films, electrodeposited using pulse plating from DES with a low Ag(I) concentration (0.6 g L⁻¹, solid line) and a relatively high Ag(I) concentration (2.4 g L⁻¹, dashed line). Besides the peaks from the Cu-substrate, additional peaks for face-centered cubic (fcc) solid-solution phases could be observed. As Pd, Ag and Cu can all form fcc solid solutions with each other, a clear phase assignment for these thin Pd–Ag films is not possible with XRD. However, the shift of the lattice parameters can be evaluated. In Table 1, the lattice parameters are given for the fcc lattice of pure Pd, Ag and Cu are summarised along with the lattice parameters of the detected fcc phases and values for calculated crystallite sizes.

OPEN IN VIEWER

Figure 3

XRD diagrams of Pd–Ag films electrodeposited deposited from DES with 0.6 g L⁻¹ Ag(I) (solid line) and 2.4 g L⁻¹ Ag(I) (dashed line).

OPEN IN VIEWER

Table 1

Assigned phases and their lattice parameters and crystallite sizes.

Composition Pd:Ag (wt-%)	Layer thickness (µm)	Phase	Lattice parameter	Crystallite size (nm)
	–	Cu (fcc)	0.3615	–
	–	Pd (fcc)	0.3890	–
	–	Ag (fcc)	0.4086	–
76:25	0.4	Pd–Ag(–Cu)	0.3807	15 ± 1
45:52	1.2	Pd–Ag(–Cu)	0.3819	41 ± 4
		Ag–Pd(–Cu)	0.4017	33 ± 4

Note: Pd, Ag and Cu phases are given for comparison and were not identified in the samples.

The Pd–Ag layers deposited from a low Ag(I) concentration electrolyte generated only one additional Pd-rich fcc phase other than the fcc Cu substrate phase. Its lattice parameter is smaller than both the single fcc phases of Pd and Ag. There was both a Pd-rich and Ag-rich phase detectable from the Pd–Ag deposited from the high Ag(I) concentration electrolyte. The Pd-rich and Ag-rich fcc phases lattice parameters are smaller than those of single phase Pd and Ag (see Table 1). The crystallite size of these two fcc phases in the Ag-rich deposit are about 2 times larger than those from the Pd-rich alloy deposited from the low Ag(I) concentration electrolyte. The collected XRD data proves the existence of Pd–Ag alloys.

Physical and chemical properties of the coatings

Layer Thickness (EDXRF). The average Pd–Ag layer thickness for each coating was measured at several points along the sample. The thickness of the Pd–Ag alloys was affected by the Ag concentration in the DES solution. The distribution of the layer thickness proved to be more homogenous when produced with the selected organic additive. Figure 4 graphically presents the average layer thickness of electrodeposits from DES solutions with SSS and a variation of the Ag(I) concentration. Generally, increasing the Ag(I) concentration in simple DES solution increased the layer thickness of the Pd–Ag deposits. As expected from CV data, Ag(I) concentrations above 1.2 g L⁻¹ did not significantly change the composition of the electrodeposits (Figure 3). Overall, there was a trend towards thicker metal films from electrolytes with both metals in comparison to the monometallic Pd electrolyte. It can be assumed that there is a synergetic effect (co-deposition) between Ag(I) and Pd(II) ions near the electrode surface resulting in an enhanced metal deposition rate in comparison to the monometallic deposition of Pd. The calculated current efficiency for the Pd–Ag alloys deposition was about 45%.

OPEN IN VIEWER

Figure 4

Variation of the Pd–Ag alloy composition with dependence on the Ag(I) concentration in the simple DES.

Methanol oxidation. The typical curve for the methanol oxidation on Pt wire is shown in Figure 5 (solid line). In comparison with this result it can be seen that such process didn't take place on Pd–Ag coatings deposited from DES. The reduction peaks caused after surface oxidation during the positive scan could be observed in the cathodic part of the curves, at approximately −0.45 V for both investigated alloys. With increased Ag content in the alloy (dotted line) two additional peaks were observed in the cathodic part of the curve. They are due to electroreduction of oxygen containing Ag(I) species [44].

OPEN IN VIEWER

Figure 5

Methanol oxidation at room temperature. Solid curve – pure Pt; dashed curve Pd-Ag 20 wt-%; doted curve Pd–Ag 50 wt-%.

Adhesiveness. The adhesiveness of the layers was proved using a tape test (DIN 58196-6). After removing the tape from the samples and sticking them to a microscope glass slide it was observed that all layers showed high adhesiveness – only some dendrites from the end of the samples remained on the tape.

Hardness. The hardness of the layers deposited from DES was comparable to such alloys deposited from water based electrolytes – at about 450 HV for Pd–Ag10 wt-% and 350 HV for Pd–Ag50 wt-%.

Current efficiency. The Pd–Ag layers are very thin, that is why the real current efficiency could only be estimated. The theoretical calculated current efficiency of the alloys is about 45% and is lower in comparison to that from water based electrolytes. It was calculated using the followed equation: where η = current efficiency, θ = layer thickness, ρ = density (calculated for each alloy), c = electrochemical equivalent, i = current density and t = deposition time.

Experiments to investigate the conductivity, wear and corrosion of the deposits are planned for the near future.

In comparison to layers produced from water solutions, it is not yet possible to deposit bright glossy Pd–Ag layers from DES, independent from the deposition conditions. The thickness of all samples was also thinner when compared to Pd–Ag depositions from water electrolyte. Identical Pd–Ag phases have been observed in Pd–Ag layers deposited from both DES and based water electrolytes [44].

Conclusions

Pd–Ag alloys have been successfully electroplated using pulse plating from choline chloride/urea based DES with and without an organic additive at 70°C under ambient atmosphere and intensive hydrodynamic conditions.