Introduction

Cathodic evolution of hydrogen at metal electrodes is one of the most significant electrochemical processes, especially important with respect to rapid development of proton exchange membrane (PEM) fuel cell and battery technologies. On the other hand, nickel makes one of the best non-noble catalyst materials suitable for hydrogen evolution reaction (HER) in alkaline media, primarily due to its high corrosion resistance at high pH values [1-5]. Nickel foams are made by chemical vapour deposition (CVD) [6], or electrochemical or electroless deposition method, typically with a polyurethane foam precursor [7]. Such-produced catalyst is characterized by high porosity and specific surface area, and also by good electrical conductivity, high corrosion resistance and superior mechanical properties [7, 8]. Significant improvement of electrocatalytic properties of bare nickel foam (e.g. towards HER or oxidation of aliphatic alcohols) could be realized through surface deposition of nano-structured noble metals (e.g. Pd, Ru, Rh, Pt or their binary/ternary alloys). The above might be performed by means of electrodeposition, spontaneous deposition [8-10] or by chemical reduction processes, where the latter are typically facilitated with NaBH4, ethylene glycol, hydrazine or their compositions [11-15].

In this communication, Pd-activated nickel foam catalyst material was prepared via spontaneous deposition method [8], to produce cathode having superior HER performance for alkaline environments. The choice of palladium results from the fact that Pd element itself is well known to exhibit high electrochemical activity towards hydrogen absorption and evolution processes [16-21] with an exchange current density value of 1 × 10−3 A cm−2, which is very close to that exhibited by Pt [18]. These properties of palladium along with highly porous, large surface area metallic Ni foam structure [22] make a Pd-modified nickel foam composite material a very good candidate for studying the HER behaviour.

Experimental

All solutions were prepared by means of a Direct-Q 3 UV ultra-pure water purification system from Millipore (18.2 MΩ cm water resistivity). A 0.1-M NaOH supporting solution was prepared from AESAR, 99.996 % NaOH pellets. An electrochemical cell, made of Pyrex glass, was used during the course of this work. The cell comprised three electrodes: a Ni-foam-based working electrode (WE) in a central part, a reversible Pd (0.5-mm-diameter wire of 99.9 % purity, Aldrich) hydrogen electrode (RHE) as reference and a Pt (1.0-mm-diameter wire of 99.9998 % purity, Johnson Matthey, Inc.) counter electrode (CE), both placed in separate compartments. The procedures for cleaning the cell and preparation of the CE/RHE electrodes for the measurements were as previously described in Pierozynski and Smoczynski [23], Pierozynski [24] and Pierozynski [25]. Before conducting the HER experiments, each Ni foam electrode was also activated in 0.1 M NaOH by cathodic polarization, carried out at 20 mA for 300 s in order to maintain surface area reproducibility [26].

Nickel foam was supplied by MTI Corporation (>99.99 % Ni, thickness 1.6 mm, surface density 346 g m−2, porosity ≥95 %). All electrodes were 1 cm × 1 cm, for which pure Ni foam accounted for about 2.6 cm2 of geometrical area and 0.030 g (~88 cm2 g−1). Spontaneous deposition of Pd catalyst on nickel foam samples was performed according to the corresponding description given in Verlato et al. [8]. Hence, freshly cut foam samples were subjected to acetone and CH2Cl2 wash (15 min + ultrasonication), following air drying and acid etching in 2 M HCl (15 min at 60 °C). Then, spontaneous deposition of metal catalyst was realized by dipping pretreated foam electrodes in 0.005 M PdCl2 (pH 1.0, t dep. = 15–300 s, temperature 25 ± 1 °C) to produce Pd-activated Ni foam catalyst material. Working electrode assemblies were made by mechanically attaching Ni foam/Pd-modified Ni foam samples to 0.5-mm-diameter Ni wire (Aldrich, ≥99.9 %). In addition, shrinkable PE or PTFE sleeves were used to provide Pd wire insulation within the electrochemical cell.

A.c. impedance spectroscopy and quasi steady state polarization techniques were employed during the course of this work. All measurements were performed over the temperature range of 20–60 °C (controlled by laboratory water bath) by means of the Solartron 12,608 W Full Electrochemical System, consisting of 1260 frequency response analyser (FRA) and 1287 electrochemical interface (EI) for an a.c. signal of 5 mV and the frequency range sweep between 1.0 × 105 and 0.5 × 10−1 Hz. The instruments were controlled by ZPlot 2.9 or Corrware 2.9 software for Windows (Scribner Associates, Inc.), whereas data analysis was performed with ZView 2.9 (Corrview 2.9) software package. The impedance spectra were fitted by means of a complex, non-linear, least squares immitance fitting program, LEVM 6, written by J.R. Macdonald [27]. In addition, quasi-potentiostatic cathodic polarization experiments (recorded at a scan rate of 0.5 mV s−1) for the HER were carried out at all examined Ni foam electrodes. Also, an appropriate correction was introduced [28], in order to account for a small but significant temperature shift of the Pd RHE over the studied temperature range of 20–60 °C.

Spectroscopic characterization of Pd-activated Ni foam electrodes was performed by means of Quanta FEG 250 scanning electron microscope (SEM). In addition, powder X-ray diffraction (XRD) technique was employed to determine average size of Pd crystallite domains for the selected Pd-modified Ni foam samples. The XRD spectra were recorded by means of Siemens D500 powder diffractometer with Cu Kα radiation (λ = 1.5418 Å, U = 38 kV, I = 30 mA) in a standard Bragg-Brentano mode without sample spinning. Experimentally recorded data were analysed by means of the phase analysis XRAYAN software.

Results and Discussion

SEM and XRD Characterizations of Pd-Modified Ni Foam Electrode

Figure 1a, b illustrates the effect of spontaneous deposition of palladium at a very low level (~0.1 wt% Pd) on the MTI foam, recorded for the magnifications of × 100 and × 15,000, respectively. Hence, at the magnification of × 15,000, high-density and homogeneously distributed small Pd nuclei could clearly be seen in Fig. 1b. The SEM-approximated Pd grain size value came to 10.0 ± 1.0 nm. The above was performed through utilization of the Image Analysis Program (NIS-Elements Basic Research on Nikon), based on an analogous procedure to that described in detail in Smoczyński et al. [29]. In addition, the powder-XRD-calculated average Pd grain size value came to 7.0 ± 0.7 nm (see Fig. 2 above). Measurements were carried out based on the Scherrer’s method, which correlates the size of crystallite domains with relative widening of diffraction peaks (see Pecharsky and Zavalij [30] for further details).

Fig. 1
figure 1

a SEM micrograph picture of Pd-modified Ni foam surface (~0.1 wt% Pd), taken at × 100 magnification; b As in Fig. 1a, but taken at × 15,000 magnification

Fig. 2
figure 2

XRD pattern for spontaneously deposited Pd element on nickel foam substrate, where observed diffraction lines correspond to the following sequence of fcc indices: (111), (200), (220), (311) and (222) for both Ni and Pd elements

A.c. Impedance Behaviour of HER on Pure and Pd-Modified Ni Foam Materials in 0.1 M NaOH

A.c. impedance characterization of the HER on pure Ni foam and catalyst (Pd)-modified Ni foam electrodes in 0.1 M NaOH is shown in Fig. 3 and Tables 1, 2 and 3. Hence, the impedance-examined, both unmodified and Pd-activated Ni foam electrodes exhibited single, “depressed” semicircles (a single-step charge-transfer reaction) at all examined potentials and reaction temperatures, in the explored frequency range (please note that a high-frequency semicircle electrode porosity response, which is typically observed in alkaline media, was practically indiscernible). Examples of Nyquist impedance plots, recorded for pure and the Pd-modified nickel foam electrodes at −100 mV versus RHE, are shown in Fig. 3 below. The overpotential dependence of Faradaic reaction resistance (R ct) and double-layer capacitance (C dl) parameters for the HER, examined over the temperature range of 20–60 °C at pure nickel foam (derived based on a constant phase element (CPE)-modified Randles equivalent circuit model shown in Fig. 4) is presented in Table 1. The CPE element was used in the circuit in order to account for the capacitance dispersion [31, 32] effect, represented by somewhat distorted semicircles in the Nyquist impedance plots.

Fig. 3
figure 3

Complex-plane impedance plots for the HER on pure and Pd-modified Ni foam electrode surfaces in contact with 0.1 M NaOH, recorded at room temperature for −100 mV (vs RHE). The solid lines correspond to representation of the data according to the equivalent circuit shown in Fig. 4

Table 1 Electrochemical parameters for the HER on cathodically activated pure Ni foam electrode in contact with 0.1 M NaOH, studied over the temperature range of 20–60 °C
Table 2 Electrochemical parameters for the HER on Pd-modified Ni foam electrode in contact with 0.1 M NaOH, studied over the temperature range of 20–60 °C
Table 3 Exchange current density, Tafel slope and activation energy electrochemical parameters for the HER, obtained on pure Ni foam and Pd-modified Ni foam electrodes in contact with 0.1 M NaOH over the temperature range of 20–60 °C
Fig. 4
figure 4

Equivalent circuit model used for fitting the impedance data for both pure and Pd-modified Ni foam electrodes, obtained in 0.1 M NaOH. The circuit includes a constant phase element (CPE) for distributed capacitance; R ct and C dl (as CPEdl) elements correspond to the HER charge-transfer resistance and double-layer capacitance components, and R sol is solution resistance

Thus, for the cathodically activated Ni foam electrodes, the recorded R ct parameter examined at 20 °C decreased from 13.653 Ω g at −50 mV to 0.162 Ω g at the potential of −400 mV versus RHE. Simultaneously, the C dl parameter significantly reduced from 10,959 to 3,806 μF g−1 sφ1−1 (2.9 times) for the same potential range. The latter effect most likely results from partial blocking of electrochemically active electrode surface by freshly formed H2 bubbles. The above might easily be visualized for a complex Ni foam electrode structure (Fig. 1a, b), especially at significant overpotentials. Also, when the capacitance value of 10,959 μF g−1 sφ1−1 (recorded for electrode weight of 35.0 mg at −50-mV RHE) is referred to that commonly used value of 20 μF cm−2 in literature for smooth and homogeneous surfaces [33, 34], electrochemically active surface area of the Ni foam electrode could roughly be estimated at 19.2 cm2 (or 548 cm2 g−1). Similar impedance behaviour (with a single-step charge-transfer process) has recently been presented by Grden et al. [35] for the HER characterization of Incofoam® Ni foam commercial product in 0.5-M KOH solution. There, the recorded at −200-mV R ct parameter value for an etched foam sample came to 43 Ω (~1.30 Ω g), which is significantly higher than 0.877 Ω g recorded for analogous conditions in Table 1. Simultaneously, the impedance-approximated [35] specific surface area for the etched Incofoam® sample came to 241 cm2 g−1, as compared to that of 548 cm2 g−1 for the MTI foam examined in this work. These results suggest that the latter nickel foam provides significantly enhanced electrochemical properties towards the HER (most likely due to its unique porous structure), as compared to the former one. In another recent work by van Drunen et al. [36], cyclic-voltammetry-derived values of electrochemically active surface area for analogous Ni foam material were on the order of 250–450 cm2 g−1. Other important options for the fabrication of highly-porous, three-dimensional Ni or Ni-originated matrices include powder-based [37, 38] and electrochemical deposition processes [10, 39]. As all options are significant, nickel foam comparatively provides a fully commercialized technology with highly reproducible material having superior mechanical and electrical properties.

Table 1 also presents the temperature dependence of the R ct and C dl electrochemical parameters, studied over the temperature range of 20–60 °C. Hence, for the two temperature extremes, 20 and 60 °C, the recorded charge-transfer resistance exhibited significant reduction, correspondingly from 13.653 to 0.844 Ω g (by 16.2 times) at −50 mV and from 0.162 to 0.089 Ω g (by 1.8 times) at −400 mV. Furthermore, the C dl parameter recorded at individual overpotentials kept slowly rising upon temperature increase (see Table 1). The above could be explained in terms of extended access to the catalytic surface within the porous electrode structure at elevated temperatures. In addition, higher temperatures should considerably facilitate hydrogen bubble removal.

On the other hand, the recorded charge transfer resistance parameter for the Pd-modified Ni foam electrode at 20 °C ranged from 1.526 to 0.128 Ω g for the overpotential range of 50–400-mV RHE (see Table 2). The above means considerable reduction of the R ct parameter (as compared to those R ct values recorded for the unmodified Ni foam surface), namely by 8.9 times and 1.3 times at −50 and −400 mV, respectively. In addition, deposition of catalytic Pd nano-structure resulted in significant enhancement of electrochemically available surface area, as compared to that of the nickel foam baseline material. Thus, for the Pd-activated Ni foam, the recorded values of the C dl parameter at −50-mV RHE came to 204,554 μF g−1 sφ2−1 (about 10, 228 cm2 g−1). The above is ~18.7 times greater than that recorded for the baseline Ni foam cathode. Understandably, for the catalyst-modified nickel foam material, the C dl parameter exhibited similar (decreasing) overpotential dependence to that recorded for the baseline Ni foam electrode. However, the recorded C dl ratio for the −50/−400-mV pair came to ~1.3, which was radically smaller than that recorded for the unmodified Ni foam electrode above (2.9 times). In other words, on the Pd-modified nickel foam, the HER predominantly proceeds on the catalyst-covered outer parts of the foam structure, so that the removal of hydrogen bubbles from the catalyst becomes strongly facilitated.

Correspondingly, Table 2 above shows the temperature dependence of the R ct and C dl parameters, examined over the temperature range of 20–60 °C for the Pd-modified Ni foam electrode. Thus, the recorded (at 20 and 60 °C) charge-transfer resistance exhibited significant reduction, respectively from 1.526 to 0.151 Ω g (by 10.1 times) at −50 mV and from 0.128 to 0.078 Ω g (by 1.6 times) at −400 mV. Then, the C dl parameter recorded at specific overpotentials also continued increasing upon rising temperature (see Table 2 and compare with the corresponding results presented in Table 1 above).

Finally, dimensionless φ 1 and φ 2 parameters of the CPE circuit (see Fig. 4 and Tables 1 and 2), where φ determines the constant phase angle in the complex-plane plot, (0 ≤ φ ≤ 1) varied between 0.83–0.97 and 0.82–0.94, respectively.

Based on the linear relationship − log R ct versus overpotential (for kinetically controlled reactions), exhibited here over the overpotential range of 50–400 mV versus RHE (see Fig. 5 above), the exchange current-densities for the HER were calculated based on the Butler-Volmer equation and through utilization of the relation between the exchange current density (j 0) and the R ct parameter for overpotential approaching zero value [5, 40-42]. Hence, the impedance-derived values of the j 0 parameter came to 1.2 × 10−6 A cm−2 (at 20 °C) and 2.0 × 10−5 A cm−2 (at 60 °C) for the unmodified Ni foam, whereas significantly increased j 0 values, namely, 1.3 × 10−5 A cm−2 (at 20 °C) and 1.4 × 10−4 A cm−2 (at 60 °C), were recorded for the Pd-modified nickel foam catalyst material (see Table 3 for details). The room-temperature-derived HER exchange current density value (1.3 × 10−5 A cm−2) for the Pd-activated nickel foam is significantly higher than the j 0 for bulk Ni (6.3 × 10−6 A cm−2), but dramatically lower than the corresponding bulk j 0 parameter value for Pd element (1 × 10−3 A cm−2) [18]. However, it should be understood that electrocatalytic activity of a catalyst material strongly depends on the extent of its electrochemically active surface area and nano-structural properties. One should also note that nano-scaled catalytic structures typically demonstrate far superior electrocatalytic activities to those exhibited by the respective bulk catalyst materials. Thus, highly modifiable (e.g. by deposition of variable, but negligible amount of Pd), large surface area porous Ni foam could potentially produce exceptionally active and relatively inexpensive HER catalyst.

Fig. 5
figure 5

Linear plots of − log R ct in function of potential (vs RHE), obtained for the HER performed on pure and Pd-modified Ni foam electrodes in 0.1-M NaOH solution, recorded at 20 and 60 °C. Symbols represent experimental results and lines are data fits

In addition, Fig. 6 presents − log R ct versus T −1 Arrhenius-type plots, constructed based on the R ct = f(T) impedance results, presented for the overpotential range of 100–400 mV versus RHE for the pure and the Pd-activated Ni foam catalyst electrodes. Thus-obtained experimental electrochemical energies of activation, E A [kJ mol−1] for the HER on the Pd-modified nickel foam, came significantly lower (by ~25 % at −100 mV and by 20 % at −400 mV) than those for the bare Ni foam electrode (Table 3).

Fig. 6
figure 6

Linear plots of − log R ct versus T −1 for the HER performed on pure and Pd-modified Ni foam electrodes in 0.1-M NaOH solution, at the stated overpotential values

HER Characterization by Steady State Tafel Polarization Plots

The kinetic results discussed above are in good agreement with these of the potentiostatic Tafel polarizations, presented for the pure and the Pd-modified Ni foam HER catalysts (performed at 20 and 60 °C) in Fig. 7. In fact, radical enhancement of the HER performance upon introduction of catalytic amount of Pd into Ni foam structure could clearly be observed over the kinetically controlled, low-overpotential region in this figure. The recorded cathodic Tafel slopes (parameter b c in Table 3) exhibited significant temperature dependence (see Conway [43] for details) for both examined electrode materials. Hence, for the temperature range of 20–60 °C, the Tafel slope changed from −137 to −222 mV dec−1 and from −65 to −97 mV dec−1 for the baseline Ni foam and the Pd-modified Ni foam catalyst materials, respectively. At the same time, considerable reduction of the cathodic Tafel slope for the palladium-activated Ni foam came in-line with significant facilitation of the exchange current density parameter. Thus, the Tafel-based values of the j 0 parameter for the HER reached 9.8 × 10−7 (at 20 °C) and 1.2 × 10−5 A cm−2 (at 60 °C) for the unmodified Ni foam, and 8.6 × 10−6 (at 20 °C) and 5.3 × 10−5 A cm−2 (at 60 °C) for the Pd-modified nickel foam electrode (see other details in Table 3). These results are in a fairly good agreement with the impedance-derived values of the j 0 parameter. However, the recorded difference between the Tafel-calculated and the impedance-derived values of the exchange current density parameter in Table 3 (especially perceptible for the Pd-modified Ni foam electrode) could possibly result from somewhat inferior quality of the Tafel fit for this case. It should be stated here that all calculated in this work exchange current-densities and cathodic Tafel slopes for the HER are within the range that is commonly quoted for Ni-porous entities in literature. Hence, Dominguez-Crespo et al. have recently reported [38] Tafel slopes in the range 152–214 mV dec−1 and optimum j 0 value of 6.3 × 10−5 A cm−2 for highly porous (La, Ce)/Ni electrodes. Furthermore, Vazquez-Gomez et al. [10] reported for cathodically deposited porous Ni layer onto Ni disk electrode a Tafel slope of 138 mV dec−1 and a respective j 0 value equal to 3.2 × 10−4 A cm−2, whereas further Ru or Ir modifications (by spontaneous metal depositions) resulted there (generally) in reduction of Tafel slopes and a dramatic increase (by 7–38 times) of the j 0 parameter (see Table 3 in Vazquez-Gomez [10] and compare with the results presented above). Similar trends for the recorded b c and j 0 parameters (including their temperature dependence) were also reported [39] for three-dimensional porous Ni structures, fabricated over the so-called double-template electrochemical deposition process (see Table 2, Herraiz-Cardona [39]).

Fig. 7
figure 7

Quasi-potentiostatic cathodic Tafel polarization curves (recorded at a rate of 0.5 mV s−1) for the HER on pure and Pd-modified Ni foam electrode surfaces, carried out in 0.1-M NaOH solution at 20 and 60 °C (appropriate iR correction was made based on the solution resistance derived from the impedance measurements)

In addition, as originally argued by Conway [43] (based on non-linear T dependence of the b c Tafel slope), the recorded charge transfer coefficient α linearly decreased with rising temperature [44] for both examined HER catalyst materials (see Fig. 8 above).

Fig. 8
figure 8

Temperature dependence of experimental Tafel charge transfer coefficient for the HER on pure and Pd-modified foam electrodes in contact with 0.1-M NaOH solution

Conclusions

Palladium nano-deposit (at ~0.1 wt%) on the surface of cathodically activated Ni foam material significantly enhanced catalytic activity of baseline foam material towards cathodic evolution of hydrogen in 0.1-M NaOH solution. The above was specifically observed over kinetically controlled, low-overpotential range, being a result of superior HER activity of a catalytic Pd additive, in addition to the resultant, extensive modification of electrochemically active surface for this catalyst material.

Both Tafel slopes, as well as charge transfer coefficient, exhibited temperature-dependent behaviour. Significant facilitation of the HER performance for the Pd-modified Ni foam material was revealed through considerably reduced Tafel slopes and electrochemical energies of activation, and by substantially increased values of exchange current density parameter, as compared to those obtained for the unmodified nickel foam cathode. Finally, the results obtained in this work indicated substantial opportunities for Ni-foam-modified cathode materials in commercial alkaline water electrolysers.