Journal Archive

The bath chemistry consisted of trisodium hexabromoiridate (Na₃[Ir(III)Br₆]) and nickel sulfate hexahydrate (NiSO₄·6H₂O). All chemical reagents were analytical grade. The bath solution was purged with pure nitrogen for 10 min before turning on the current. A nitrogen blanket was passed over the solution during electrodeposition. In all experiments, a magnetic stir bar was used to stir the solution. Copper foam was supplied by Taili Foam Metal Corporation, China (>99.99% copper, porosity ≥98%). The copper foam samples (10 mm × 10 mm × 0.3 mm) were first immersed in acetone for 40 min and ultrasonically cleaned for 10 min. The samples were then immersed in a mixture solution of nitric acid (HNO₃)·H₂O (volume ratio = 1:1) at room temperature for 5–10 s in a fume hood. Finally, the samples were washed with pure water and then dried in air. A conventional three-electrode cell was used for electrodeposition of iridium-nickel thin films on a copper foam sheet as the working electrode, platinum foil as the counter-electrode and a reference Ag/AgCl 3 M KCl electrode. The galvanostatical electrodeposition and electrochemical processes were carried out in the cell, using a CHI 660E electrochemical workstation (CH Instruments Inc, China) coupled to a computer with specific data acquisition software installed. The solution pH was examined by a PHS-3C pH meter and was adjusted to the desired value by adding 5.0 M NaOH while being stirred. The volume of the electrolyte in the cell was approximately 15 ml. The solution temperature was controlled by a HH-501 thermostatic bath (Jintan Baita Xinbao Instrument Factory, China). The mass change of the specimens before and after deposition was recorded by an analytical balance (FA2004B, resolution 0.1 mg). Table I shows the composition of the plating bath and the deposition parameters, three specimens were prepared in the same electrolyte, the first specimen #1 was deposited with 1 min, then the second specimen #2 for 3 min and the last specimen #3 for 5 min.

The CV of iridium in the bath of 26 mM Na₃[Ir(III)Br₆], nickel in the bath of 26 mM NiSO₄·6H₂O and iridium-nickel in the bath of 26 mM Na₃[Ir(III)Br₆] and 26 mM NiSO₄·6H₂O were performed, using an electrochemical workstation (CHI 660E). The electrocatalysts on copper sheets with diameter of 1 cm served as the working electrodes. The platinum foil with diameter of 1 cm served as the counter-electrode and a reference Ag/AgCl 3 M KCl electrode was used. The solutions for CV tests of iridium-nickel, nickel and iridium films were electrolytes at room temperature with a scanning rate of 10 mV s^–1.

The microstructure and morphology of the surface of the films were observed by SEM (SUPRA 55, Sapphire and ZEISS Sigma, ZEISS Microscopy, Germany). The chemical composition of the film was determined by X-ray EDS using x-act detector (Oxford Instruments plc, UK).

The chemical composition and elemental states of the top surface of the iridium-nickel thin film on copper foam before HER performance testing were performed by X-ray photoelectron spectroscopy (XPS), (ESCALAB^TM 250Xi, Thermo Fisher Scientific, USA) with an ultrahigh vacuum of 3.3 × 10^–8 Pa base pressure. The top surface of the iridium-nickel thin film on copper foam was irradiated with an aluminium Kα monochromatic source of 1486.6 eV, and the outcome electrons were analysed by a spherical capacitor analyser using the slit aperture of 0.8 mm. The carbon signal for C 1s at 285 eV was taken as an energy reference for the measured peaks. In order to identify the elements on the film surface, a low-resolution survey spectrum was firstly taken over a wide energy range of 0–1350 eV, with a pass energy of 150.0 eV at increments of 1.0 eV step^–1. High-resolution spectra were acquired with a pass energy of 30.0 eV at small increments of 0.05 eV step^–1, to allow precise determination of the position of the peaks and their shapes. Curve-fitting was done with Gaussian–Lorentzian functions, using the XPSPEAK 4.1 software. Two fitting parameters, the position of the peak and its full width at half-maximum (FWHM), were fixed within less than about ± 0.2 eV.

The LSV was tested in an alkaline solution at a scanning rate of 5 mV s^–1 using a three-electrode system. Nitrogen was bubbled through the 1.0 M potassium hydroxide solution for 1 min before all electrochemical measurements were done. The iridium-nickel electrocatalysts deposited on the copper foam electrode served as the working electrode. The graphite rod and Hg/Hg₂Cl₂ were used as the counter and reference electrodes, respectively. The catalytic activities of iridium-nickel thin films with different loadings and solution temperatures were compared by LSV testing. Furthermore, the potentials were transferred to a reversible hydrogen electrode (RHE) potential. The formula for calibrating the saturated calomel electrode (SCE) to RHE is as follows (Equation (i)):

(i)

The CV of an iridium-nickel thin film was performed in alkaline solution using an electrochemical workstation (CHI 660E). The iridium-nickel thin film on the copper foam electrode served as the working electrode. In this work, the electrochemically active surface area (ECSA) of iridium-nickel catalysts were compared on a relative scale using the capacitance of the electrochemical double layer (C_dl) on the electrode-electrolyte interface. This comparison is validated as ECSA is proportional to C_dl. C_dl is obtained by the following procedure reported by McCrory et al. (36), which assumes that the measured current at the non-Faradaic region i_C is due to the charging of the double-layer capacitor. Thus, the current i_C is proportional to the scanning rate v_scan (Equation (ii)):

(ii)

CV measurements were conducted at different scanning rates ranging from 10 mV s^–1 to 160 mV s^–1. The capacitor charging current i_C at 0.05 V is plotted against the scanning rate ν_scan, which shows good linear dependence. Then, C_dl is obtained from the slope of the linear fitting. The relative ECSA of the iridium-nickel catalysts with different loadings are compared by their C_dl values.

Potential values were corrected with the ohmic (iR) drop by the electrolyte resistance obtained from EIS measurements taken in a potentiation mode in the frequency range from 100,000 Hz to 0.01 Hz. The solution temperature was controlled in the range of 20–60°C.

In order to better understand the electrodeposition mechanism, the CV measured in electrolytes containing nickel, iridium and both iridium and nickel salts are shown in Figure 1. The reduction peaks of iridium, nickel and iridium-nickel were at −0.56 V, −0.37 V and −0.54 V due to the reduction of Ni(II) to Ni⁰ and Ir(III) to Ir⁰, respectively. It can be found that the reduction potentials of iridium-nickel and iridium are very close and their peak currents are much higher than that of nickel, which may lead to much more iridium content in the iridium-nickel deposits. Without the addition of complexing agent in the electrolyte, the reduction potential of Ni²⁺ ions is higher than that of [Ir(III)Br₆]^3– ions. Therefore, the deposition of nickel was preferable during the electrodeposition process. The oxidation peak for iridium is present at –0.24 V vs. Ag/AgCl. The oxidation peak was mainly attributed to iridium oxidation, Ir0 → Ir(III). However, the oxidation peak for iridium-nickel is present at around –0.24~–0.15 V vs. Ag/AgCl because the iridium-nickel codeposited film is suggested to be alloyed. There is no oxidation peak for nickel at negative potential. For iridium-nickel film, the current peak at –0.54 V was due to the hydrogen adsorption, while the current peak at around –0.24~–0.15 V was due to the reoxidation of the codeposited deposits. For iridium film, the current peak at –0.56 V was observed due to the hydrogen adsorption, which is higher than the current peak at –0.23 V due to hydrogen desorption. These findings indicate that iridium and iridium-nickel deposits have a significant facility for hydrogen incorporation into the plated deposits (37). The deposition processes of iridium and iridium-nickel films are accompanied by a large amount of hydrogen evolution.

Fig. 1.

Figure 2 shows the plots of the mass gain of the specimens deposited in one electrolyte with different deposition times. There is a nonlinear dependence of the mass gain against deposition time. According to CV curves, nickel is preferentially deposited because the deposition potential of nickel is higher than that of iridium. Therefore, the deposition rate for the first specimen, about 2.3 mg min^–1 is larger than that of the others. Subsequently, the deposition rate kept stable, and the nickel content in the films decreased. The slope of the second plot is around 0.35, indicating a deposition rate of about 0.35 mg min^–1. The atomic composition of the film by EDS is listed in Table II. The iridium content in the deposits increased remarkably from 1 min to 5 min. This result is in agreement with the above discussion.

Fig. 2.

Figure 3 shows the surface morphology of iridium-nickel electrocatalysts and copper foam. Figure 3(a) shows a large number of dendritic structures for copper foam with diameters 30–60 nm. In Figures 3(b) and 3(c), many pores and hollowed topography can be observed. The grain boundary of the copper foam is clearly visible because of the thin layer. With the increase of deposition time, the thickness of the film also increases. It can be clearly seen that the iridium-nickel thin film is attached to the surface of the substrate (see Figure 3(d)). The cross structure and many pores of copper foam can increase the active area of the thin film, which is advantageous for hydrogen evolution performance. Figure 4 shows the EDS spectrum of the iridium-nickel electrocatalysts with different loadings. The chemical composition of iridium and nickel in the thin films is shown in Table II. EDS elemental mapping to probe nickel and iridium presence for Ir₈₀Ni₂₀ and the distribution on the substrate surface is shown in Figure 5. A large amount of copper is evenly distributed on the surface. It can be observed that the amount of iridium is larger than that of nickel in the image. Ir₈₀Ni₂₀ thin film was almost completely covered on the copper foam surface. The phase and crystallographic structure of the films on copper foam were determined by X-ray diffraction (XRD), however the signals of the films were not detected due to small loading, the information of copper foam was only present. The copper foam was composed of polycrystalline structure.

Fig. 3.

Fig. 4.

Fig. 5.

The chemical composition and elemental states of Ir₈₀Ni₂₀ thin films were deeply analysed by XPS technique. The atomic composition of Ir₈₀Ni₂₀ thin films on copper foam is listed in Table III. Figure 6 shows the XPS depth profile for the top surface of iridium-nickel thin films on copper foam. The elements copper, iridium, carbon, oxygen and bromine were determined on the top surface of as-deposited film. Unfortunately, the signal nickel element was not detected, probably due to the low quantities in the film. The large amounts of carbon and oxygen contents were attributed to ordinary adsorption from the environment (see Table III), significant amounts of oxides formed on the surface of the electrode during electrodeposition. The bromine signal was mainly from the electrolytes. The coverage of the Ir₈₀Ni₂₀ thin film on copper foam was not so perfect that the signal copper was determined. Figure 7 shows the high-resolution XPS spectra of as-deposited Ir₈₀Ni₂₀ thin film on copper foam. The binding energies of iridium were located at 63.6 eV and 60.6 eV for Ir⁰ 4f5/2 and 4f7/2, respectively. It was indicated that the film was composed of the metallic state of iridium, although the nickel signal was not detected (Figure 7(a)). The binding energies of Ir⁰ 4f5/2 and 4f7/2 have weak shifts in iridium-nickel thin films in contrast with pure iridium (63.8 eV and 60.8 eV) (38). This is ascribed to the incorporation of nickel into the electronic structures of iridium, demonstrating the changes in the alloyed phase, in turn enhancing the catalytic performance (39–41). In the O-1s spectrum (see Figure 7(b)), the main peak at 530.5 eV is attributed to O 1s of the oxides, and the peaks at 531.3 eV and 531.9 eV are usually ascribed to surface species, such as hydroxyls or absorbed water of the film. An additional broader feature peak is present at a higher binding energy of ~533.2 eV.

Fig. 6.

Fig. 7.

The electrocatalytic activities for HER of iridium-nickel thin films with different loadings were investigated in 1.0 M KOH solutions. Figure 8 presents the iR-corrected LSV curves and Tafel slopes of the bare copper foam and iridium-nickel samples. As anticipated, the bare copper foam displays a relatively low catalytic activity which requires an overpotential of 502.5 mV to drive a current density of 10 mA cm^–2. In contrast, the iridium-nickel catalyst exhibits excellent catalytic activity, demonstrating a negligible onset potential (8.3–18.3 mV) at 1 mA cm^–2 for hydrogen evolution in the electrolyte (see Table IV), which are much lower than the onset potential of copper foam. Here, the onset potential should always be defined on the basis of a specific current density, where the Tafel constant can be considered as the onset potential of HER (42, 43). From an electrochemical point of view, the Tafel constant becomes complementary to the Tafel slope.

Fig. 8.

In order to obtain a current density of 10 mA cm^–2, Ir₄₂Ni₅₈, Ir₈₀Ni₂₀ and Ir₈₈Ni₁₂ thin films require overpotential of 76.4 mV, 60 mV and 95.4 mV, respectively (see Table IV). These values are already twice the overpotential of 34 mV at the current density of 10 mA cm^–2 for the state-of-the-art Pt/C catalyst tested in the same electrolyte. To shed light on insights about the reaction kinetics, a detailed Tafel analysis has been performed. The Tafel equation is as follows (44) (Equation (iii)):

(iii)

where j is the current density, j₀ is the exchange current density (i.e. a constant at η = 0 V) and b is the Tafel slope. The equation indicates that excellent catalysts should have both low Tafel slopes and high exchange current densities. The potential-dependency of current density j is related to the interfacial electrocatalytic reaction n, as the following (Equation (iv)):

(iv)

where n is the number of electrons, F is the Faraday’s constant (96,500 mol C^–1). Because the current density j is potential-dependent, ν is also potential-dependent and consisted of three elementary steps as the following Equations (v)–(vii):

Initial discharge or Volmer step:

(v)

Atom + ion or Heyrovsky step:

(vi)

Atom + atom or Tafel step:

(vii)

The above elementary steps lead to two mechanisms: Volmer-Heyrovsky and Volmer-Tafel. Three rate determining steps, Volmer, Heyrovsky and Tafel are possible for the above two mechanisms. The linear portions of the Tafel plots were fitted to the Tafel equation, the Tafel slope values for Ir₄₂Ni₅₈, Ir₈₀Ni₂₀ and Ir₈₈Ni₁₂ thin films were 49 mV dec^–1, 40 mV dec^–1 and 43 mV dec^–1 respectively, indicating that the Ir₈₀Ni₂₀ electrocatalyst has much higher intrinsic activity than other catalysts for HER while still being less active than the Pt/C catalyst (28mV dec^–1) (45). The Tafel slopes indicate the HER process of iridium-nickel film follows a Volmer-Heyrovsky mechanism, where electrochemical desorption of hydrogen is regarded as the rate-limiting step, i.e. the HER rate is determined by both H₂O discharge and desorption of H from the catalyst surface (46–48). The exchange current density (j₀) of iridium-nickel thin films and copper foam were calculated by the Tafel extrapolation method (see Table IV), which reflects the catalytic activity of the electrode material under the reaction thermodynamic equilibrium conditions. The j₀ value of the Ir₈₀Ni₂₀ thin film was 87.6% of the Pt/C catalyst (0.75 mg cm^–2, 15 wt% Pt) (45). Therefore, an iridium-nickel thin film has highly efficient electrocatalytic activity for HER. In addition, a summary of the hydrogen evolution performance of electrocatalysts in alkaline solutions is listed in detail in Table V (49–81). It can be found that the Tafel slope of the iridium-nickel thin film is low, indicating iridium-nickel thin film has excellent electrocatalytic performance.

The ECSA of the catalyst is proportional to the electrochemical double-layer capacitance (C_dl). As shown in Figure 9, the C_dl values for Ir₄₂Ni₅₈, Ir₈₀Ni₂₀ and Ir₈₈Ni₁₂ were 59.82 mF cm^–2, 132.91 mF cm^–2 and 70.03 mF cm^–2, respectively. It indicates that the high hydrogen evolution performance of Ir₈₀Ni₂₀ catalyst was mainly due to the high exposure of effective active sites. On the other hand, the scanning range of –0.15~0.85V for Ir₈₀Ni₂₀ catalyst is larger than the range of –0.25~0.65 V for other catalysts. In our previous publication (32), the initial CV curve of Ir₈₀Ni₂₀ catalyst was measured at a scanning rate of 10 mV s^–1. It was found that iridium oxides are electrochemically formed at high positive potential on the surface of Ir₈₀Ni₂₀ thin film during a positive scanning direction, however the formation of iridium oxides cannot easily be reduced to the metal state (32). The formed iridium oxides could result in a significant decrease of HER activity. Therefore, when the scanning range of Ir₄₂Ni₅₈ and Ir₈₈Ni₁₂ thin films was shifted from –0.25 V to 0.65 V, the obtained C_dl values should be valid. Therefore, it is inferred that the hydrogen evolution performance of Ir₈₈Ni₁₂ film is better than that of Ir₄₂Ni₅₈ film, which might be attributed to the increase in iridium content of the film or electrode surface defects, resulting in increasing the number of effective active sites.

Fig. 9.

The large active surface area could be related to the porous structure and hollow architecture with crossed branch structure, which results in a significantly enhanced catalytic activity. The rough texture and the porous structure of copper foam facilitate fast mass transport for the enhanced reaction kinetics (82). However, Ir₈₈Ni₁₂ thin film with a good hydrogen evolution performance has a larger electrochemical surface area than Ir₄₂Ni₅₈ film. The electrocatalytic activity of iridium-nickel catalysts show a loading dependence.

Electrochemical impedance measurements were performed to further investigate the reaction kinetics of the HER process under the experimental conditions. Nyquist plots of copper foam, Ir₈₀Ni₂₂, iridium and nickel thin films as a function of overpotential are shown in Figure 10. The preparation of the iridium and nickel electrocatalysts was addressed (32). According to alternating current (AC) circuit theory, impedance spectra obtained for a given electrochemical system can be correlated to one or more equivalent circuit (83). Thus, different equivalent circuits were suggested to model the present data and the relevant model with the minimum number of electrical elements. The model of Ir₈₀Ni₂₂ film consists of the solution resistance (R_s), the low frequency time constant characterising the double-layer capacitance (C_dl) and charge transfer resistance (R_ct). The potential dependencies of the obtained data are shown in Table VI. Due to surface heterogeneity of solid electrodes resulting from surface roughness and formation of porous layers, a constant phase element (CPE) is commonly used to replace the capacitance (C) in a real electrochemical process, which mainly depends on a non-ideal capacitance behaviour (84, 85). R_ct values of iridium film and copper foam are 72.34 Ω cm² and 76.64 Ω cm², respectively. While the charge transfer resistance of nickel and Ir₈₀Ni₂₂ films are large, about 2642 Ω cm² and 1312 Ω cm², indicating that the R_ct values of iridium film and copper foam are lower than those of nickel and iridium-nickel films. According to the XPS data, there is no iridium oxide on the catalyst surface for Ir₈₀Ni₂₂ thin films. It can be inferred that the iridium thin film was composed of metallic state. On the other hand, copper foam was immersed in nitric acid solution to activate it before the experiment. The absence of oxides in copper foam may result in a charge transfer resistance that is less than other electrocatalysts as a result. For nickel and Ir₈₀Ni₂₂ thin films, the top surface might be composed of some nickel oxides. The surface of nickel-rich nickel-iridium thin films consisted of lots of nickel oxides, the amount of nickel oxides was much more than that of metal nickel (unpublished data). Therefore, there is a contradiction here. The electrocatalytic performance of the thin film involves various factors, such as surface chemical substances, the number of catalytic active sites per unit area, and the electronic effect of the thin film metal. At the cathode, the process of hydrogen reduction for hydrogen gas requires energy to remove electrons from the metal electrode and connect electrons to protons to produce hydrogen. Therefore, the process of transferring electrons from the electrode to the hydrogen ions in the liquid phase has a certain resistance, which is a charge transfer resistance. Iridium electrode with a low resistance could accelerate the electron transfer during the electrocatalytic reaction.

Fig. 10.

Figure 11 shows the polarisation curves and Tafel curves of the Ir₈₈Ni₁₂ film at different temperatures. As shown in Figure 11(a), the electrode has the best hydrogen evolution performance at 60°C, with only an overpotential of 186 mV to obtain a current density of 30 mA cm^–2. At the temperature of 30°C, an overpotential of 212 mV is required. The performance of hydrogen evolution improves from 30°C to 60°C. Interestingly, the hydrogen evolution performance of the catalyst has decreased from 20°C to 30°C. The effect of temperature is not obvious, and the curves are very close. This result can also be derived from the Tafel slopes (see Figure 11(b)). From 20°C to 60°C, the Tafel slope is 46 mV dec^–1, 56 mV dec^–1, 52 mV dec^–1, 43 mV dec^–1 and 40 mV dec^–1, respectively. The comparison of the HER catalytic performance at different temperatures is shown in Table VII.

Fig. 11.

To investigate the essence of the improvement of HER activity of the iridium-nickel electrocatalyst, the apparent activation energies (E_a) of the film were determined via the following Arrhenius equations (86) (Equations (viii)–(x)):

(viii)

(ix)

(x)

where j₀ is exchange current density (A cm^–2), F is the Faraday’s constant, k is Kohlrausch coefficient (dimensionless), c is the concentration of reactant (constant), E_a is the apparent activation energy (J mol^–1), T is the temperature (K), and R is the gas constant (8.314 J mol^–1 K^–1). The linear relationship between logj₀ and 1/T for Ir₈₈Ni₁₂ thin film is displayed in Figure 12. According to Equations (viii)–(x), the apparent activation energy of Ir₈₈Ni₁₂ thin film electrocatalyst is calculated as 7.1 kJ mol^–1, by the slopes of lines. Compared with the nickel cathode with about 40 kJ mol^–1 in standard electrolyte (87), this result indicates that the Ir₈₈Ni₁₂ thin films can remarkably reduce E_a for HER and accordingly result in higher electrocatalytic activity. Hence, the codeposition process of nickel and iridium species on the copper foam provides a large number of active centres for hydrogen adsorption, with the synergetic effect giving electronic structure suitable for HER.

Fig. 12.

AC impedance characterisation of HER on the Ir₈₈Ni₁₂ electrode in 1.0 M KOH with different temperatures is shown in Figure 13. Nyquist plots of Ir₈₈Ni₁₂ film as a function of overpotential are shown in Figure 13(a). Impedance spectra obtained for a given electrochemical system can be correlated to an equivalent circuit (Figure 13(b)). The temperature dependence of R_ct and C_dl parameters for the HER of Ir₈₈Ni₁₂ film examined at the temperature range of 20–60°C is present in Table VIII. The Ir₈₈Ni₁₂ electrode exhibited single, ‘depressed’ semicircles (a single-step charge-transfer reaction) at all reaction temperatures, in the explored frequency range, it is noted that a high-frequency semicircle electrode porosity response, which is typically observed in alkaline media, was practically indiscernible (34). The recorded R_s parameter decreased from 1.572 Ω cm² at 30°C to 1.038 Ω cm² at 60°C. Simultaneously, the R_ct parameter significantly reduced from 48.34 Ω cm² to 14.75 Ω cm² for the same temperature range (see Table VIII). The lower Faraday resistance of the Ir₈₈Ni₁₂ electrode surface accelerates the electron transfer during the electrocatalytic reaction. The C_dl parameter was significantly reduced from 0.02515 F cm^–2 to 0.02885 F cm^–2. The effect most likely results from partial blocking of electrochemically active electrode surface by fresh hydrogen bubbles (34).

Fig. 13.

Apart from the catalytic activity, stability is another important requirement of catalysts for the HER system. In this case, the long-term stability of Ir₈₀Ni₂₀ film electrocatalyst is assessed in 1.0 M KOH at constant current densities of 10 mA cm^–2 for 10 h (see Figure 14(a)). A slight increase in the overpotential has been observed in the V-t curve. The result of the long-term hydrogen evolution tests exhibited excellent electrocatalystic stability in alkaline solution. Polarisation curves recorded after 400 cycles testing for Ir₈₈Ni₁₂ film indicate that there is a little decay while the overpotential exceeds 0.068 V, instead, a slight improvement of the electrocatalytic activity of the electrode at low overpotential (see Figure 14(b)). This slight increase in the catalytic activity was possibly owing to the reduction of surface oxides during the initial period of hydrogen evolution, while the observed increase in overpotential as shown in Figure 14(a) could be a result of increased mass exchange resistance due to the continuous gas bubbling (88).

Fig. 14.