Mo doped porous Ni–Cu alloy as cathode for hydrogen evolution reaction in alkaline solution

Abstract

Molybdenum doped porous Ni–Cu cathodes are conveniently prepared by powder metallurgy in this work. The microporous structure of the porous Ni–Cu based cathodes has a certain correlation with the content of molybdenum, which consequently influences the electrocatalytic behavior as electrode materials. Electrocatalytic performance of the prepared cathodes has been investigated by hydrogen evolution reaction in 6.0 mol L⁻¹ KOH solution. It is demonstrated that the Mo doped porous Ni–Cu cathodes exhibit a superior electrocatalytic activity for HER over a binary porous Ni–Cu cathode, among which the porous Ni–Cu with 6.0 wt% Mo shows the highest intrinsic activity and an ascendant stability during HER. Although the porous Ni–Cu (with 9.0 wt% Mo) expresses the highest activity, a gradual decline of the current density for HER in long-term operation is found, which is due to the dissolution of Mo from the porous skeleton in alkaline solution.

---

1. Introduction

As a highly-anticipated method for large-scale production of hydrogen, water electrolysis has long been a main research area to meet the growing needs of this clean sustainable energy carrier. In order to minimize energy consumption during hydrogen production by water electrolysis, fundamental investigations are focused on the development of cathode materials. Although platinum shows almost perfect performance for the hydrogen evolution reaction so far, the drawback lies in its low abundance and the thus resulting high price.¹

Taking cost into consideration, relatively inexpensive materials such as nickel (Ni),² cobalt (Co),³ molybdenum (Mo),^4,5 tungsten (W),⁶ and their alloys (Ni–Co,⁷ Ni–W,⁸ Ni–Mo^8,9) have been subjects of intense researches due to their promising electro-catalytic performance. A common way to increase the integrated electro-catalytic behavior of these alloys is to introduce foreign elements (Mo,^8,9 Fe,¹⁰ Cu,^11,12etc.) or compounds (rare earths,¹³ hydroxides,¹⁴ sulfides,^15,16 nitrides,¹⁷etc.) into the prior systems, for obtaining larger surface areas or higher intrinsic activity. Mo is usually selected as an additive in assisting the main source of catalytic activity, while the root cause for explaining the improvement on electro-catalytic properties are not always consistent.

Previous kinetic studies have claimed that the introduction of Mo into Ni or Ni based alloys and subsequent alkaline leaching of the superficial Mo leads to a course surface with enhanced activity for the HER.^18,19 However, results from J. G. Highfield et al.²⁰ have shown that the synergy is expressed in the Ni–Mo alloys down to remarkably low levels of Mo (<5 at%). Co–Mo films with low Mo contents have also been found significant synergistic effect as cathodes in water electrolysis, which is achieved by the combination of hyper d-electronic phase Co and hypo d-electronic phase Mo. ²¹Amorphous Ni–Fe–Mo alloys have presented good activity and stability for HER, after incorporating Mo with binary Ni–Fe alloy,²² due to the synergistic effect between Ni–Fe and Mo. Based on these researches, it can be suggested that a certain amount of Mo adding to binary or ternary metallic materials may tailor their desirable electro-catalytic properties.

Many porous multi-component alloys are fabricated by powder metallurgy, for different application purpose, which is considered as a low-cost and highly convenient way to prepare materials with three-dimensional porous structure.^23–25 Careful design of the process (composition, sintering temperature, pore formers, etc.) allows to tune the porous morphology, the microstructure and phase constitution of the resulting material. The homogenous porous Ni/Cu (mass ratio Ni : Cu = 2 : 1), with an open porosity ranging from 23.5% to 37.8% are obtained in previous works.²⁶ Furthermore, numerous reports demonstrate the synergistic effect between Ni and Cu for Ni–Cu alloys applied as an electro-catalytic cathode for hydrogen evolution by water electrolysis.^12,27,28 The stability of the porous structure for cathode material is also important during a long-term application. For most reported porous cathodes, it is hard to ensure the adherence of the electro-active layers with the substrate while maintaining the composition. Based on the precise control of constituents and the flexible guidance on porous structure by powder metallurgy, this study extends to fabricate active molybdenum doped Ni–Cu alloy cathodes for further enhancing the electro-catalytic activity and stability during HER. The effects of the molybdenum concentration on the chemical and the electrochemical properties of porous Ni–Cu based alloy cathodes are carefully studied in this research.

The compositions, microstructures of the obtained molybdenum doped porous Ni–Cu cathodes are characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), and mercury intrusion porosimetry (MIP) method. The electrochemical properties of porous Ni–Cu(Mo) are evaluated by linear sweep voltammetry (LSV), impedance spectroscopy (EIS) and cyclic voltammetry (CV) measurements. The chemical stability in potassium hydroxide solution is broadly analyzed by energy dispersive X-ray spectroscopy (EDS) and inductively coupled plasma atomic emission spectroscopy (ICP).

2. Experimental procedures

2.1 Preparation of cathodes

Commercially available nickel, copper and molybdenum powders with median diameter of 8.2 μm, 19.5 μm and 5.6 μm, respectively are mixed to obtain a uniform ternary mixture 2Ni–Cu–xMo (where x = 3.0, 6.0 and 9.0 wt%). Compacts with dimensions of 55.0 mm × 9.0 mm × (1 ± 0.1 mm) are cold pressed by uniaxial pressing under 100 MPa. Sintering is performed in the furnace with vacuum of 1.0 × 10⁻⁴ Pa, and the compacts are buried in alumina hollow balls to keep uniform heat conduction. The heating rate is controlled at 5 °C per minute; the temperature is hold at 700, 800, 1000 and 1100 °C for 1 h followed by natural cooling to room temperature.

All the samples used for following tests were rinsed with deionized water in an ultrasonic cleaner for 30 min, then degreased in anhydrous alcohol, and finally dried up in vacuum oven at 75 °C for 2 hours prior to use.

2.2 Composition and pore structure characterization

The phase and crystal structure of the prepared porous Ni–Cu based samples were analyzed by XRD (D/max 2550VB with Cu Kα radiation). The morphology of the material was characterized by SEM (JSM-6360LV, 20 kV). Elemental analysis of local area was determined by an energy-dispersive X-ray spectrometer attached to the SEM. The average pore size was determined by the use of mercury intrusion (Pore Master 33GT, Quantachrome Instruments). Archimedes principle was employed to measure the open porosity of porous samples. The N₂ permeate flux and maximum pore size of porous structure was measured through bubble test.

2.3 Electrochemical measurements

All the electrochemical tests were performed in a standard three-electrode, two-compartment cell, with a deaerated 6.0 M KOH electrolyte solution. The auxiliary electrode was a platinum foil, and a saturated calomel electrode (SCE) connected to the cell via a salt bridge was used as the reference electrode. The apparent exposed surface area of the working electrode was 1.0 cm². The Luggin capillary was arranged as close as possible to the working electrode in order to minimize IR drop in electrolytes. Prior to the electrochemical measurements the cathodes were immersed in 6.0 M KOH solution at 25 °C for 5 h, till a steady open-circuit-potential (OCP) value was obtained. Steady-state CV curves were recorded in 6.0 mol L⁻¹ KOH solution at a series of scan rates of 0.001, 0.005, 0.010, 0.020, 0.030 and 0.040 V s⁻¹. Electrochemical impedance spectroscopy (EIS) measurements at selected potentials were carried out in the frequency range of 10⁻² to 10⁵ Hz, with a 5 mV sinusoidal perturbation. Inductively coupled plasma (ICP) atomic emission spectrometry was adopted to determine the concentration of ions in the immersion system.

3. Results and discussion

3.1 Characterization

Fig. 1 shows the cross-section morphology of the investigated porous cathodes, and the corresponding parameters of porous structure are collected in Table 1. Abundant and interconnected pores can be observed from the cross-section micrographs of all cathodes in Fig. 1. Results in Table 1 show that the porous structure of Mo doped porous Ni–Cu is slightly influenced by the content of Mo. The average pore size of Ni–Cu based cathodes decreases from 7.5 μm to 5.5 μm as the Mo content increases from zero to 9.0 wt%, while the open porosity maintains around 31.4% for all the cathodes. It indicates that the connectivity of pores is enlarged after adding the third element Mo, which can be related to the diffusion kinetics of Ni(Cu) and Ni(Mo) systems. Earlier researches^29,30 have confirmed that the plenty of pores formed in the interface of Ni/Cu diffusion-couple are caused by the Kirkendall effect of partial diffusion, due to the huge difference of diffusion coefficient between Ni and Cu. For Ni/Cu mixed powder compact, it is assumed that there were numerous micro Ni/Cu diffusion-couples during sintering at elevated temperature. Besides the interstitial pores, the Kirkendall effect between Ni and Cu particles will result in greatly increased porosity of the three dimensional porous Ni–Cu alloy. Nevertheless, the third element brought to the starting binary components will result in some adjustment to the prior porous structure.

Fig. 1 The cross-section morphology of the Mo doped porous Ni–Cu cathode: (a) with 0 wt% Mo; (b) with 3.0 wt% Mo; (c) with 6.0 wt% Mo; (d) with 9.0 wt% Mo (for simplicity, the four cathodes are designated as S₀, S_I (3.0 wt%), S_II (6.0 wt%) and S_III (9.0 wt%), respectively).

Table 1 The porous parameters of the investigated porous Ni–Cu alloy cathodes

	Open porosity	Average pore size (μm)	N2 permeate flux (m^3 m^−2 kPa^−1 h^−1)
S0	37.6%	7.5	76.2
SI (3.0 wt%)	37.4%	6.7	75.1
SII (6.0 wt%)	37.2%	5.9	75.7
SIII (9.0 wt%)	37.3%	5.6	76.3

---

Studies on diffusion kinetics of Ni (Mo) systems have found that the interdiffusion coefficient varies very little when the composition of Mo below 25 at% at temperatures between 1050 to 1225 °C, which further indicates that there is not much difference in the vacancy concentration at different content of Mo incorporated to the Ni–Cu solid solution.^31,32 It can well explain the nearly constant open porosity for Mo-doped porous Ni–Cu, although the Mo content is changed. Nevertheless, the Ni/Cu diffusion areas in the sintering compact can be divided into small pieces by the diffusion of Mo atoms, and spaces for existing large pores are partly occupied by the formed small pieces, thus decreases the average pore size of the porous Ni–Cu. As shown in Table 1, the average pore size of porous Ni–Cu(Mo) is decrease with Mo content increasing, the N₂ permeate flux of the porous Ni–Cu(Mo) have not exhibited a corresponding decline as the Mo content increases.

The XRD spectrograms of porous Ni–Cu based alloys with different Mo contents are displayed in Fig. 2. The strong diffraction peaks of 2θ around 44.00°, 51.26° and 75.43° belong to Ni–Cu alloy substrate, and Mo atoms have diffused into the solid solution of Ni–Cu phase. With increasing content of Mo, the diffraction peaks slightly shift to lower diffraction angles. It is due to the lattice expansion caused by the dissolving Mo atoms, since the atomic radius of Mo is larger than that of Ni and Cu. For cathode S_III, the Mo phase has also been detected by its main peak at 2θ = 40.50°. The EBSD analysis in Fig. 3 confirms that not all the Mo particles have diffused into the Ni–Cu solid solution in cathode S_III (9.0 wt%), while the cathodes with lower content of Mo (S₀, S_I and S_II) all present a homogenous single phase after sintering.

Fig. 2 XRD patterns of sintered porous Ni–Cu based cathodes with different Mo content: S₀ (0 wt% Mo), S_I (3.0 wt% Mo), S_II (6.0 wt% Mo) and S_III (9.0 wt% Mo).

Fig. 3 SEM-BSE images and EDS analysis of Mo doped porous Ni–Cu cathodes (cross section): (a) S₀; (b) S_I (3.0 wt%); (c) S_II (6.0 wt%) and (d) S_III (9.0 wt%).

Furthermore, the exact compositions of sample S_I and S_II are examined by ICP, after wholly dissolved with concentrated nitric acid. The Mo contents are measured to be 2.97 ± 0.03 wt% and 5.94 ± 0.02 wt%, respectively. Results certify that the Mo have completely dissolved into the Ni–Cu solid solution after sintering, thus can not be detected by XRD.

3.2 Steady-state polarization curves

Fig. 4a reports the linear sweep voltammetry curves for the HER on different Mo doped porous Ni–Cu cathodes in 6.0 M KOH solution at 25 °C, and recorded Tafel curves are shown in Fig. 4b. The corresponding kinetic parameters (the apparent exchange current density j₀(app), Tafel slope b, and the over-potentials at current density of 50 mA cm⁻², η₅₀) have been presented in Table 2.

Fig. 4 (a) The polarization curves and (b) corresponding Tafel plots of porous Ni–Cu based cathodes in 6.0 M KOH solution (scan rate 0.01 V s⁻¹).

Table 2 HER kinetic parameters obtained by analysis of the curves presented in Fig. 3. Results for flake Ni and porous Ni–Cu are also presented for comparison

Cathode	E onset (V)	Tafel slope (mV dec^−1)	η 50 (mV)	j 0(app) (mA cm^−2)	R f	j 0(int) (mA cm^−2)
Flake Ni	−1.47	101	396	7.51 × 10^−3	18.7	4.02 × 10^−4
S0	−1.39	197	241	0.96	385	2.49 × 10^−3
SI (3.0 wt%)	−1.33	141	168	1.93	550	3.51 × 10^−3
SII (6.0 wt%)	−1.30	112	135	2.57	637	4.03 × 10^−3
SIII (9.0 wt%)	−1.25	125	127	2.76	715	3.86 × 10^−3

---

As can be seen from Fig. 4a and Table 2, all Mo doped porous Ni–Cu alloys exhibit the enhanced activities for HER in terms of a more positive onset potential (E_onset) and much lower hydrogen over-potentials, in comparison to the porous Ni–Cu alloy. Among the four cathodes under study, the porous Ni–Cu (9.0 wt%-Mo) shows the highest activity, with an over-potential of nearly 0.127 V at 50 mA cm⁻² and 25 °C.

In order to obtain information on the intrinsic activity of the investigated porous Ni–Cu based cathodes for the HER, the actual exchange current density j₀(act) should be determined taking into account the real electrochemically active surface area. Taking that the average double layer capacitance of a smooth metal surface is 20 μF cm⁻²,³³ the mean value of the double layer capacitance of the electrodes can be calculated as C_dl = ∂j_dl,ave/∂sr (μF cm⁻²) (j_dl,ave and sr represents the average of capacitive current and the scan rate, respectively), the roughness factor can be calculated as R_f = C_dl/20. Calculated roughness factor and the actual exchange current density for the investigated cathodes, are also presented in Table 2. Porous structure parameters in Table 1 indicate that different actual surface area will be provided by the cathodes. For porous materials, when the open porosity is constant, the actual surface area should be increase with average pore size deceasing. As shown in Table 2, the roughness factor for porous Ni–Cu cathode is approximately 20 times higher than dense Ni flake, and it continues to increase after incorporation of Mo. For cathode S_III (9.0 wt%), the R_f is nearly 1.86 times higher than that of S₀. High roughness factor indicates a high volumetric density of active sites on the electrodes, and the increased number of electroactive surface sites greatly enhances the apparent electrocatalytic activity of the cathodes.³⁴ From the calculated exchange current density j₀, as well as the over-potential value under the same current density 0.05 A cm⁻², cathode S_III (9.0 wt%) exhibits the best overall activity, while cathode S_II (6.0 wt%) presents the highest intrinsic activity. The intrinsic j₀ (exchange current density) for cathode S_II (6.0 wt%) is about 4.02 × 10⁻³ A cm⁻², which is 10 times higher than that for dense Ni flake at the same electrolysis condition, and 1.62 times higher than binary porous Ni–Cu.

Moreover, the Tafel curves recorded on Mo doped porous Ni–Cu cathodes (Fig. 4b) exhibit classical Tafel behavior, clearly indicating that the HER can be described using the Tafel equation.¹⁷ As shown in Table 2, the Tafel slopes for HER are found to be 141, 112 and 125 mV dec⁻¹ for cathodes S_I, S_II and S_III, respectively. Mo doped porous Ni–Cu cathodes exhibit lower Tafel slopes and higher apparent exchange current densities compared with binary porous Ni–Cu. Obviously, the electrochemical behavior is related to the concentration of Mo incorporated to the Ni–Cu solution. Mo is completely resolved to the binary Ni–Cu solid in cathode S_I, S_II, and the solid solubility for Mo in cathode S_III is calculated to be 6.65 wt% after the designed sintering process. It indicates that the Tafel slope for the investigated cathodes can be minimized as the content of Mo to the binary Ni–Cu doped at 6.0 wt%, which can explain the relatively higher intrinsic exchange current density of HER on the cathode S_II.

It follows that the primary contribution to the enhancement of the electrocatalytic activity of the Mo doped porous Ni–Cu cathode toward HER arises not only from the increase of the actual specific area, but also a significant synergism catalytic effect is present. The synergism is optimally expressed as the Mo content in porous Ni–Cu(Mo) solid solution reaches about 6.0 wt%, according to the results. The excessive Mo is effectively redundant for a critical lattice spacing of Ni–Cu(Mo), causing a progressive loss of synergism in the alloy cathodes.

About improvements in electrocatalytic performance regarding Ni base cathodes modificated by the third constituent with proper concentration, explanations in many reported studies are ultimately switched to the Brewer–Engel valence bond theory, initially introduced by Jaksic et al.³⁵ According to this theory, Bocutti et al.³⁶ interpret that the synergism exhibited on Ni–La (0.3 at% La in LaNi5/Ni) or Ni–Ce (0.9 at% Ce in MmNi3.4Co0.8Al0.8/Ni) combines the d⁸-orbital of Ni (which can promote the discharge of H₂O) and the semi-empty d¹-orbital of La or Ce (which can conciliate the step of hydrogen adsorption). Fotis Paloukis et al.³⁴ also proposed that the strength of the metal–H and H₂O–metal interactions plays an important role in the hydrogen evolution reaction, the latter should be strong enough to split H₂O while the former interaction no need to be excessively strong, otherwise the hydrogen desorption would be restrained.¹⁸ Thus, the electrocatalytic activity of the investigated Mo doped porous Ni–Cu depends not only on the geometry of cathode, but also on the chemical composition. Additionally, Ni and Cu locate on the left side of the well-known “volcano plots”, while Mo on the right side. From this point of view, the incorporation of Mo element into the Ni–Cu solid solution will moderately enhance the H adsorption on the surface of metal electrode, which results in an intermediate metal–H interaction during the hydrogen evolution reaction. The comparatively lower onset potential and the smaller Tafel slope of Mo doped porous Ni–Cu than the binary Ni–Cu cathode demonstrate that the presence of Mo in the Ni–Cu lattice favors proton adsorption kinetics. The proper electronic interaction state between Ni–Cu and Mo turns out to be 6.0 wt% Mo in porous Ni–Cu(Mo) cathode.

Compared with multi-component cathode reported previously, the intrinsic activity for S_II is superior than NiTi, which has an intrinsic j₀ of 1.94 × 10⁻⁵ A cm⁻² in 1.0 M NaOH at 25 °C;³⁷ and also exceed the Ni–Co alloy powder electrode with the apparent j₀ of 1.2–1.4 × 10⁻³ A cm⁻² in 6.0 M KOH at 25 °C,³⁸ as well as the electrodeposited Ni overlayers on Mo polycrystalline, which is reported as an apparent j₀ for 0.9 × 10⁻³ A cm⁻² in 0.1 M NaOH at 25 °C.³⁴ Additionally, the activity of cathode S_II is comparable to the electrodeposited Ni–W (the apparent j₀ is 6.5 × 10⁻³ A cm⁻² in 6.0 M KOH at 25 °C),³⁹ and Ni–Co–Zn (the apparent j₀ is 5.5–6.31 × 10⁻³ A cm⁻² in 30 wt% KOH at 30 °C).⁴⁰

3.3 AC impedance measurements

More information in terms of the electrode/electrolyte interface of the investigated cathodes, involving surface charge, distribution of electro-catalytic centers, hydrogen adsorption and number of electrochemical active centers, has been provided by electrochemical impedance spectroscopy.

The experimental complex-plane plots and the phase-frequency plots for three Mo-doped porous Ni–Cu cathodes obtained at various potentials are displayed in Fig. 5. The presented Nyquist and Bode spectra show the existence of two distinctive time constants for all cathodes, and the absence of Warburg impedance indicates that mass transport is rapid enough so that the reaction is kinetically controlled.¹⁵ The similarity among the response of all cathodes suggests a similar mechanism for hydrogen evolution. The two-time constant serial equivalent circuit model (2TS) presented in Fig. 5d shows the appropriate simulated result to the EIS response of three cathodes during HER.

Fig. 5 (a–c) EIS spectra recorded on the Mo doped porous Ni–Cu cathodes at various potentials, in 6.0 M KOH solution (25 °C): (a) with 3.0 wt% Mo (b) with 6.0 wt% Mo (c) with 9.0 wt% Mo. Symbols are experimental data and solid lines are data fitted by the 2TS model. (d) Electrical equivalent circuit used to model the system of all Mo doped porous Ni–Cu cathodes investigated with EIS.

According to this model, the resistance element R_s is attributed to the uncompensated solution resistance; the high-frequency time constant, independent of the potential, described by the R_p and C_p connected in parallel, is related to the porosity of electrode, whereas the potential dependent time constant is related to the kinetics of the HER (R_ct and CPEct connected in parallel).⁹

The fitting curves are presented by the solid lines in the Fig. 5a–c, while the values of the impedance parameters for porous Ni–Cu based cathodes are listed in Table 3. As shown in Table 3, the charge transfer resistance (R_ct) reduced greatly while the constant phase element (CPE) increases slightly with increasing Mo concentration in porous Ni–Cu based cathodes at the same electrolysis potential, which are related to the onset of faradaic reaction of HER, and also indicates the increasing electrocatalytic activity of the porous cathodes.

Table 3 The fitted parameter values of the equivalent circuit for HER at Mo doped porous Ni–Cu cathodes in 6.0 mol dm⁻³ KOH at 25 °C

		R s (Ω)	C (F cm^−2)	R (Ω)	CPE (F cm^−2)	n	R (Ω)
SI	−1.25	0.95	1.42 × 10^−3	0.64	9.68 × 10^−3	0.86	67.9
	−1.30	0.96	1.20 × 10^−3	0.71	8.45 × 10^−3	0.89	21.9
	−1.35	0.95	6.63 × 10^−3	0.66	3.97 × 10^−3	0.82	1.3
SII	−1.25	0.99	1.59 × 10^−3	0.44	1.20 × 10^−2	0.84	41.0
	−1.30	0.99	1.49 × 10^−3	0.47	1.12 × 10^−2	0.85	16.3
	−1.35	0.99	1.36 × 10^−3	0.50	1.03 × 10^−2	0.89	5.3
SIII	−1.2	0.96	2.73 × 10^−3	0.32	2.34 × 10^−2	0.84	56.3
	−1.25	0.80	3.18 × 10^−3	0.28	2.25 × 10^−2	0.88	19.0
	−1.30	0.81	3.30 × 10^−3	0.27	2.13 × 10^−2	0.89	10.9

---

R _ct decreases remarkably at higher potentials for all cathodes, and a lower value corresponds to a faster reaction rate, at E = 1.3 V, the R_ct fall below 22 Ω for all cathodes. It often means that the charge transfer process dominates the impedance response as the potential increases and the hydrogen evolution reaction is controlled by Heyrovsky step.⁴⁰ The values of n lower than 1.0 is most likely arise from the uneven charging of the double layer due to microscopic surface roughness,¹⁸ thus the CPE behaves as non-ideal capacitor in the present case. The fitted values for capacitance C^′_dl are also listed in Table 3. The C^′_dl values determined from EIS are all higher than the C_dl values determined from CV. This is an indication that the C^′_dl contains contributions from pseudo-capacitance, probably arising from the adsorption of hydrogen on the surface.¹⁵

3.4 Stability test of cathodes for long-time electrolysis

Accelerated degradation studies have been carried out to further evaluate the stability of Mo doped porous Ni–Cu cathodes during HER in alkaline solution. Cyclic voltammetric (CV) sweeps between −1.15 V to −1.80 V have been performed continuously for 800 cycles in 6.0 M KOH solutions at 25 °C.

As shown in Fig. 6a, after 800 CV cycles, the potential required to achieve current densities of 0.025 A cm⁻² and 0.12 A cm⁻² for three cathodes all decreased by less than 1 mV and 10 mV, respectively. While the values for binary porous Ni–Cu cathode and flake Ni is 9.5 mV (at 0.12 A cm⁻²) and 126.0 mV (at 0.025 A cm⁻²), respectively, as shown in Fig. 6b and c. The corresponding current density decays for different cathodes during 800 CV cycles are collected in Table 4. Results clearly indicate that the Mo doped porous Ni–Cu cathodes have maintained their porous structure, as well as the electro-catalytic activity during this process in an alkaline environment.

Fig. 6 The cyclic voltammertry curves of investigated cathodes at the scan rate of 1 mV s⁻¹ for 100 and 800 cycles in 6.0 M KOH solutions at 25 °C: (a) Mo doped porous Ni–Cu cathodes; (b) binary porous Ni–Cu as cathode; (c) flake Ni as cathode.

Table 4 Current density decay of the investigated cathodes during 800 cycles at different potentials

Potential (V)	Flake Ni	Porous Ni–Cu	SI	SII	SIII
−1.35	34.2%	3.5%	1.1%	1.1%	1.7%
−1.52	35.5%	3.7%	1.6%	1.9%	2.1%
−1.70	36.6%	3.8%	1.8%	2.1%	2.3%

---

Fig. 7 shows the chronovoltammetry measurements of Mo doped porous Ni–Cu cathodes at the constant electrolysis potential of −1.6 V. After five hours, a nearly stationary state can be seen for the porous Ni–Cu cathodes containing low content of Mo (≤6.0 wt%) during the HER, which indicates that their reactivity are maintained along the observed period. Nevertheless, the current density of the dense Ni flake cathode shows a significant and sustained decline during the whole electrolysis, which has been generally attributed to the subsurface Ni hydride formed on Ni cathode,²⁰ suppressing further adsorption between Ni and active hydrogen.

Fig. 7 Chronovoltammetry measurements on different cathodes for more than 62 hours at a constant potential of −1.6 V.

In addition, although the current density of cathode S_III is overall higher than that observed for the other cathodes with lower Mo content, a gradually dropping appears after 40 hours. The loss in catalytic activity for cathode S_III (9.0 wt%) must be due to the removal of Mo species when in contact with the potassium hydroxide solution, as suggested by ICP analysis (Table 5). As shown in Table 5, the Mo concentration in electrolyte solution after 62 hours electrolysis for cathode S_III (9.0 wt%) is measured to be 6.75 mg L⁻¹, significantly higher than the dissolved Mo concentration of other two systems. Even though there appears the decay of electro-catalytic activity, the porous structure still remains the interconnected morphology after the long-term electrolysis, as shown in Fig. 8. It has exactly reflected the greatest advantage of the porous alloy cathodes fabricated by powder metallurgy.

Table 5 Concentrations of metal ions in the KOH solution after 62 hours electrolysis on different cathodes at the potential of −1.6 V (“/”: it has a too low content that can't be detected by ICP)

Cathodes	Ni (mg L^−1)	Cu (mg L^−1)	Mo (mg L^−1)
SI	—	—	0.47
SII	—	—	0.69
SIII	—	—	6.75

---

Fig. 8 SEM micrograph and EDS analysis of the skeleton for porous Ni–Cu (9 wt% Mo) cathode after 62 h electrolysis at −1.6 V in 6.0 mol L⁻¹ KOH.

It can be demonstrated that the Mo incorporated in Ni–Cu solid solution presents more conductive for maintaining the stability of electro-catalytic during HER. Although the moderate Mo doped to Ni–Cu solid solution has obviously improved the electro-catalytic activity owing to the synergism between Ni–Cu and Mo, and it can not be ruled out that the mixed Mo phase has synergistic effect with Ni–Cu phase in cathode S_III (9.0 wt%), and more research is needed to design for interpreting other underlying information.

4. Conclusion

The electro-catalytic behavior of three-dimensional porous Ni–Cu cathodes doped by Mo via powder metallurgy is investigated during hydrogen evolution reaction. Results show that opportune amounts of Mo incorporated to the porous Ni–Cu solid solution will greatly increase its overall electro-catalytic performance, and the increase in electrochemical activity of the investigated cathodes is due to both the synergism among Mo and Ni–Cu, as well as the increased actual exposed reactive sites, compared with the binary porous Ni–Cu cathode.

Moreover, the incorporation of Mo with lower content (≤6.0 wt%) imparts higher stability in hydrogen evolution to the porous Ni–Cu alloy without compromising corrosion resistance in alkaline solution. A single-phase ternary Ni–Cu(Mo) can not be obtained for cathode S_III (9.0 wt%) under the same preparation technology, in which the un-dissolved Mo particles, uniformly dispersed in the porous Ni–Cu skeleton, accounting for a resultant loss of activity over a long-term electrolysis. This research will open new possibilities for the development of diverse electrodes for electrochemical applications. The following work is necessary to optimize the porous structure as well as to modify the surface to obtain further improvement in the HER activity.

Acknowledgements

This research was financially supported by the State Science and Technology Support Program (2012BAC02B05) and the Natural Science Fund of Hunan Province (12JJ4044).

References

1. M. Ledendecker, G. Clavel and M. Antonietti, et al. , Adv. Funct. Mater., 2015, 25, 393–399.
2. C. A. Marozzi and A. C. Chialvo, Electrochim. Acta, 2001, 46, 861–866.
3. S. Cobo, J. Heidkamp and P. A. Jacques, et al. , Nat. Mater., 2012, 11, 802–807.
4. R. F. de Souza, G. Loget and J. C. Padilha, et al. , Electrochem. Commun., 2008, 10, 1673–1675.
5. M. D. Scanlon, X. Bian and H. Vrubel, et al. , Phys. Chem. Chem. Phys., 2013, 15, 2847–2857.
6. H. Zheng, J. Huang and W. Wang, et al. , Electrochem. Commun., 2005, 7, 1045–1049.
7. A. N. Correia, S. A. S. Machado and L. A. Avaca, Electrochem. Commun., 1999, 1, 600–604.
8. C. Fan, D. L. Piron and A. Sleb, et al. , J. Electrochem. Soc., 1994, 141, 382–387.
9. L. Chen and A. Lasia, J. Electrochem. Soc., 1992, 139, 3458–3464.
10. N.-F. Elisa, C. Zhiwen and O. Sasha, J. Mol. Catal. A: Chem., 2005, 226, 179–197.
11. R. Solmaz, A. Doner and G. Kardas, Electrochem. Commun., 2008, 10, 1909–1911.
12. K. Ngamlerdpokin and N. Tantavichet, Int. J. Hydrogen Energy, 2014, 39, 2505–2515.
13. Z. Zheng, N. Li and C.-Q. Wang, et al. , J. Power Sources, 2013, 230, 10–14.
14. J. R. C. Salgado, M. H. S. Andrade and J. C. P. Silva, et al. , Electrochim. Acta, 2002, 47, 1997–2004.
15. D. Merki, H. Vrubel and L. Rovelli, et al. , Chem. Sci., 2012, 3(8), 2515–2525.
16. D. Merki and X. Hu, Energy Environ. Sci., 2011, 4, 3878–3888.
17. W. F. Chen, K. Sasaki and C. Ma, et al. , Angew. Chem., Int. Ed., 2012, 51(25), 6131–6135.
18. J. M. Jaksic, M. V. Vojnovic and N. V. Krstajic, Electrochim. Acta, 2000, 45, 4151–4158.
19. B. E. Conway and L. Bai, Int. J. Hydrogen Energy, 1986, 11, 533–540.
20. J. G. Highfield, E. Claude and K. Oguro, Electrochim. Acta, 1999, 44, 2805–2814.
21. P. Paunovic, O. Popovski and A. Dimitrov, et al. , Electrochim. Acta, 2007, 52, 4640–4648.
22. W. Hu, Y. Zhang and D. Song, et al. , Mater. Chem. Phys., 1995, 41, 141–145.
23. G. Ryan, A. Pandit and D. P. Apatsidis, Biomaterials, 2006, 27, 2651–2670.
24. X. T. Sun, Z. X. Kang and X. L. Zhang, et al. , Electrochim. Acta, 2011, 56, 6389–6396.
25. G. Chen, K.-D. Liss and P. Cao, Acta Mater., 2014, 67, 32–44.
26. L. Yu, Y. Jiang and Y. He, et al. , Mater. Chem. Phys., 2015, 163, 355–361.
27. S. H. Ahn, H.-Y. Park and I. Choi, Int. J. Hydrogen Energy, 2013, 38, 13493–13501.
28. R. Solmaz, A. Doner and G. Kardas, Int. J. Hydrogen Energy, 2009, 34, 2089–2094.
29. R. A. Masumura, B. B. Rath and C. S. Pande, Acta Mater., 2002, 50, 4535–4544.
30. I. D. Choi, D. K. Matlock and D. L. Olson, Mater. Sci. Eng., A, 1990, 124, L15–L18.
31. V. D. Divya, S. S. K. Balam and U. Ramamurty, et al. , Scr. Mater., 2010, 62, 621–624.
32. M. S. A. Karunaratne and R. C. Reed, Defect Diffus. Forum, 2005, 237, 420.
33. M. Xia, T. Lei and N. Lv, Int. J. Hydrogen Energy, 2014, 39, 4794–4802.
34. F. Paloukis, S. Zafeiratos and V. Drakopoulos, et al. , Electrochim. Acta, 2008, 53, 8015–8025.
35. M. M. Jaksic, Electrochim. Acta, 1984, 25, 1539–1550.
36. R. Bocutti, M. J. Saeki and A. O. Florentino, et al. , Int. J. Hydrogen Energy, 2000, 25, 1051–1058.
37. K. Andrea, V. Nicolae, B. Waltraut and D. Narcis, Int. J. Hydrogen Energy, 2007, 32, 3258–3265.
38. P. Elumalai, H. N. Vasan and N. Munichandraiah, et al. , J. Appl. Electrochem., 2002, 32, 1005–1010.
39. C. Fan, D. L. Piron, A. Sleb and P. Paradis, J. Electrochem. Soc., 1994, 141, 382–387.
40. I. Herraiz-Cardona, E. Ortega and V. Pérez-Herranz, Electrochim. Acta, 2011, 56, 1308–1315.

---