Contents
| Introduction |
| Experimental Materials Preparation of catalysts
|
| Results and discussion |
| Conclusion |
| Declaration of competing interests |
| Acknowledgements |
| Data availability statement |
| Online appendix |
| References |
1 Introduction
The increasing consumption of fossil fuels and the worsening living environment have driven the exploration of environmentally friendly and sustainable energy sources as alternatives to traditional fossil fuels [
1]. Hydrogen energy has attracted great attention from researchers due to its high energy density and environmental friendliness [
2]. Hydrogen is regarded as the most promising substitute, and hydrogen production by water electrolysis is an important means to achieve industrialized and low-cost hydrogen generation [
3‒
4]. However, its sluggish reaction kinetics need to be overcome by means of effective electrocatalysts [
5‒
6]. Water electrolysis involves two half-reactions, namely hydrogen evolution reaction (HER) and oxygen evolution reaction (OER). As for such two reactions, OER involves multi-electron transfer steps and has a relatively higher energy barrier, which greatly hinders the large-scale hydrogen production on an industrial scale [
7–
9]. Currently, noble metal-based catalysts, such as Pt/C and IrO
2/RuO
2, can achieve high electrocatalytic efficiencies and are the benchmark catalysts for HER and OER, respectively [
10]. However, the high cost and scarcity have hindered their widespread application and driven people to seek cheaper and more efficient electrocatalysts [
11–
14]. In addition, most of these catalysts are powdered materials and usually need to be combined with a polymer binder (Nafion) before loading onto the electrodes, which inevitably increases the electrode interface impedance and reduces the catalytic activity [
15]. Powdered catalysts are also prone to detachment from the electrode surface during the gas evolution, thereby significantly reducing the cycling stability of catalysts [
16–
19]. Therefore, it is of vital importance to design and develop self-supported catalysts with a low noble metal content for HER and OER.
In recent years, a great number of first-row transition metal alloy catalysts have been widely studied due to their abundant terrestrial resources, low cost and excellent electrocatalytic activity, such as Fe
2Ni [
20], NiZn [
21], NiCo [
22], and ZnNiCo [
23] alloys. Metal–organic frameworks (MOFs), a novel class of porous crystalline materials [
24‒
25], are composed of a variety of organic ligands and metal centers, with a wide range of applications such as water electrolysis [
26], gas storage [
27], and metal–air batteries [
28]. Due to the flexible tunability and well-defined structure of MOFs, they have become promising models for the design of catalysts at the molecular level. However, when directly used as electrocatalysts for OER and HER, pristine MOFs usually only exhibit single catalytic activity and inefficient electrochemical performance, thus failing to achieve the expected result. Researchers have made numerous attempts to improve the catalytic efficiency of MOFs. One of the most effective methods is to integrate a small amount of noble metals (Ru or Ir) with high electrocatalytic activity into MOF precursors that have the advantage of low cost. For instance, Sun et al. introduced single-atom dispersed Ru into Ni‒benzene dicarboxylic acid (BDC), and the synthesized catalyst exhibited excellent catalytic activity for HER under all pH values [
1]. Li et al. used Ru-modified CoNi‒1,3,5-benzenetricarboxylic acid (BTC) for the preparation of catalysts, achieving the outstanding electrocatalytic performance in both KOH and seawater environments [
2]. A multimetallic MOF-derived high-entropy alloy (HEA) catalyst, CC@FeCoNiMoRu-HEA/C, designed by Hu et al. has excellent OER catalytic activity and strong durability while reducing the content of Ru and lowering the economic cost [
12].
In this study, Ru-modified NiFe‒MOF nanosheets (NSs) were synthesized on nickel foam (NF) through a one-step hydrothermal method via spontaneous reduction–oxidation (redox) reactions and coordination reactions. The Ru@NiFe‒MOF/NF catalyst was subsequently prepared via the annealing treatment. The introduction of NiFe‒MOFs promoted the dispersion of Ru throughout the framework, providing the possibility of the improvement in the electrocatalytic performance of bifunctional Ru-based catalysts. Experimental studies show that the introduction of Ru atoms has changed the electronic structure of Ni atoms through metal–ligand interactions, thereby enhancing the intrinsic electrocatalytic activity and enhancing the electrocatalytic performance of Ru@NiFe‒MOF/NF for HER and OER. In addition, Ru@NiFe‒MOF/NF also has great potential for hydrogen production via water electrolysis in an alkaline environment due to its excellent activity and stability. The excellent electrocatalytic activity of Ru@NiFe‒MOF/NF was thoroughly investigated through various experiments and structural characterizations. The design and synthesis of Ru@NiFe‒MOF/NF provide new insights and research directions, paving the way for continuous hydrogen production in practical situations.
2 Experimental
2.1 Materials
NF (1 mm in thickness) was purchased from Suzhou Shuerte Industrial Technology Co., Ltd. Ruthenium chloride hydrate (RuCl3·xH2O, guaranteed reagent (GR)) was obtained from Sinopharm Chemical Reagent Co., Ltd. Hydrochloric acid (HCl, 36 wt.%) was sourced from Beijing Beihua Fine Chemicals Co., Ltd. Potassium hydroxide (KOH, 85%) was acquired from Aladdin Co., Ltd. Nickel nitrate hexahydrate (Ni(NO3)2·6H2O, analytical reagent (AR)) was purchased from Sinopharm Chemical Reagent Co., Ltd. Iron(III) nitrate nonahydrate (Fe(NO3)3·9H2O, AR) was supplied by Xilong Scientific Co., Ltd. Cobalt nitrate hexahydrate (Co(NO3)2·6H2O, AR) was acquired from Shanghai Zhanyun Chemical Co., Ltd. Terephthalic acid (PTA, purity 99%) was obtained from Aladdin Co., Ltd. Absolute ethanol (ethanol absolute, density of 0.7893 g·mL−1) was sourced from Tianjin Yufutai Chemical Reagent Co., Ltd. N,N-dimethylformamide (DMF, AR). The commercial 20 wt.% Pt/C catalyst was procured from Shanghai Hesen Electric Co., Ltd. The commercial RuO2 catalyst (purity 99.95%) was purchased from Alfa Aesar Co., Ltd. The Nafion solution (5 wt.%) was acquired from InnoChem Co., Ltd. Deionized water (DIW) used throughout the entire experiment was self-made in the laboratory, and all chemicals and solvents were not subjected to further purification.
2.2 Preparation of catalysts
First, NF was cut into small pieces followed by ultrasonic cleaning with 3 mol·L−1 HCl for 15 min to remove surface oxides and organic contaminants. Then they were rinsed with DIW and dried. Next, a homogeneous solution was prepared, which included ethanol (1.5 mL), DIW (1.5 mL), PTA (80 mg), DMF (30 mL), and Ni(NO3)2·6H2O (0.140 g). Subsequently, the cleaned NF was immersed in this solution and transferred to a 100 mL polytetrafluoroethylene (PTFE)-lined autoclave. After maintaining at 125 °C for 12 h to allow the growth of MOFs on the NF substrate, the resulting product was rinsed with DIW and dried at 50 °C for 8 h. Ni‒MOF/NF was finally obtained after annealing at 350 °C for 3 h under an argon atmosphere.
For the preparation of NiFe‒MOF/NF, 0.164 g Fe(NO3)3·9H2O was added. For Ru@Ni‒MOF/NF, 0.003 g RuCl3·xH2O was added. For Ru@Fe‒MOF/NF, Ni(NO3)2·6H2O was replaced with 0.164 g Fe(NO3)3·9H2O based on the preparation procedure of Ru@Ni‒MOF/NF. For Ru@NiFe‒MOF/NF, 0.164 g Fe(NO3)3·9H2O was additionally incorporated into the synthesis of Ru@Ni‒MOF/NF. For Ru@NiCo‒MOF/NF, 0.118 g Co(NO3)2·6H2O was additionally added in the preparation of Ru@Ni‒MOF/NF. All other experimental conditions remain the same.
3 Results and discussion
The synthesis process of Ru@NiFe‒MOF/NF is illustrated in . In this process, NF not only serves as a conductive substrate supporting the catalyst, but also acts as a reducing agent to effectively reduce Ru
3+ ions to metallic Ru due to the favorable chemical process (Δ
E =
E(Ru
3+/Ru) −
E(Ni
2+/Ni) > 0, where
E(Ni
2+/Ni) = −0.257 V and
E(Ru
3+/Ru) = +0.704 V) [
29]. Meanwhile, using Ni(NO
3)
2·6H
2O and Fe(NO
3)
3·9H
2O as metal sources and PTA as a ligand, ultrathin NiFe‒MOF NS arrays were grown on NF. Benefiting from the two simultaneous reaction processes, Ru nanoparticles (NPs) were embedded into NiFe‒MOFs, followed by
in-situ self-assembling and growth on the surface of NF, and the resulting product is hereafter denoted as Ru@NiFe‒MOF/NF.
Fig.1(a) shows the scanning electron microscopy (SEM) image of three-dimensional (3D) NiFe‒MOF NSs with smooth surfaces assembled on NF. The micron-scale spherical aggregates on the NF are NiFe‒MOFs, which exhibit an interconnected network that facilitates electrolyte penetration and, in particularly, enables more Ru active sites to be exposed. Surprisingly, upon the incorporation of Ru3+, the original smooth NiFe‒MOF NSs develop surface roughening, as evidenced in Fig.1(b) along with Fig. S1 (included by ESM of Appendix). This structural evolution suggests the critical role of Ru NPs in modulating the morphology of MOFs. As shown in Fig. S2 (included by ESM of Appendix), disordered ultrathin Ni‒MOF NS arrays grew on Ru@Ni‒MOF/NF, maintaining their structural integrity without detectable surface roughening upon the incorporation of Ru3+. Fe‒MOFs did not exist in the form of ultrathin NS arrays on NF, but as particles rich in defects, as shown in Fig. S3 (included by ESM of Appendix). Notably, NiCo‒MOFs exhibit symmetrical flowers on both sides. It was precisely this unique structure that provided abundant attachment sites for Ru3+. However, the insufficient interparticle connectivity within NiCo‒MOFs impedes charge/mass transport dynamics, thereby adversely affecting the OER/HER kinetics to some extent, as shown in Fig. S4 (included by ESM of Appendix).
In addition, transmission electron microscopy (TEM) images of Ru@NiFe‒MOF/NF show the uniform distribution of Ru NPs with an average diameter of approximately 4 nm in Fig.1(c) along with Fig. S5 (included by ESM of Appendix). The high-resolution transmission electron microscopy (HRTEM) image of Ru@NiFe‒MOF/NF in Fig.1(d) shows lattice fringes of the Ru (1 0 0) plane with a lattice spacing of 0.205 nm. Importantly, the strong interfacial interaction between Ru NPs and NiFe‒MOFs is evident from the clear interfacial structure observed. As reported in previous works, the performance of heterogeneous catalysis is strongly influenced by the surface and interface properties, because the interface structure can modify the active sites to adjust the bonding strength of reaction intermediates, thereby improving the sluggish water electrolysis reaction. Furthermore, the selected-area electron diffraction (SAED) pattern shown in Fig. S6 (included by ESM of Appendix) indicates that Ru@NiFe‒MOF/NF exhibits a polycrystalline structure with the (1 1 1) plane of NiO, the (1 0 1) plane of Ru, and the (2 0 0) plane of Ni.
The existence of the element Ru was further confirmed by the analyses on high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy-dispersive X-ray spectroscopy (EDS) images of Ru@NiFe‒MOF/NF in Fig.1(e)‒1(j), indicating that the elements Ni, Fe, Ru, C and O were uniformly distributed (Fig. S7 (included by ESM of Appendix)). Based on the EDS analysis, the elemental contents of C, O, Fe, Ni, and Ru in the Ru@NiFe‒MOF/NF catalyst are summarized in Table S1 (included by ESM of Appendix).
X-ray diffraction (XRD) patterns of Ru@Ni‒MOF/NF, Ru@Fe‒MOF/NF, Ru@NiFe‒MOF/NF, and Ru@NiCo‒MOF/NF are shown in Fig.2(a). For such four catalysts, the diffraction peaks at 2
θ values of 44.54°, 51.9°, and 76.42° correspond to (1 1 1), (2 0 0), and (2 2 0) crystal planes of Ni (JCPDS No. 87-0712), respectively [
30–
33]. Due to the strong signal of the NF substrate along with the small diameter and low loading amount of NPs for metal such as Ru, there are no obvious characteristic peaks, consistent with findings reported in most previous studies.
X-ray photoelectron spectroscopy (XPS) results reflect the compositions of materials and valence states of various elements. The XPS survey spectrum in Fig.2(b) is consistent with results of the element distribution, also indicating that there are four elements in Ru@NiFe‒MOF/NF, i.e., Ru, Ni, Fe, and O, which aligns well with results of the EDS analysis. To further investigate the chemical bonding states of Ru, Ni, Fe, and O, we analyzed Ru@NiFe‒MOF/NF based on high-resolution XPS spectra. As shown by the Fe 2p XPS spectrum in Fig.2(c), two characteristic peaks appear near 713.2 and 723.0 eV, corresponding to Fe 2p
3/2 and Fe 2p
1/2, respectively [
34‒
35], indicating the presence of both Fe
3+ and Fe
2+ in the catalyst. In Fig.2(d), it is observed from the Ru 3p spectrum that there are Ru 3p
3/2 and Ru 3p
1/2 peaks at 463.2 and 485.4 eV, respectively, which are attributed to the presence of metallic Ru [
36–
38]. Peaks at 465.5 and 487.6 eV can be ascribed to the Ru−O bond, indicating that Ru NPs can be stably anchored to the ligand framework through the formation of interfacial bonds between the edge of Ru NPs and the carboxyl oxygen of NiFe‒MOFs, which will significantly enhance the long-term catalytic stability. Ru
n+ peaks with binding energies located at 483.6 and 488.4 eV may be formed due to the surface oxidation of the catalyst when it is exposed to air. In the Ni 2p XPS spectrum shown in Fig.2(e), two strong peaks at 873.4 and 855.5 eV correspond to Ni 2p
1/2 and Ni 2p
3/2, respectively, indicating the presence of Ni
2+ in Ru@NiFe‒MOF/NF. Meanwhile, other two peaks at 852.7 and 870.1 eV are attributed to metallic Ni 2p of the NF substrate [
39–
42]. Surprisingly, the O 1s spectrum shows a typical characteristic peak of Ru−O at 530.9 eV, as shown in Fig.2(f), which indicates the formation of the Ru−O bond. The elemental contents of C, O, Fe, Ni, and Ru in the Ru@NiFe‒MOF/NF catalyst based on XPS analysis are also revealed in Table S2 (included by ESM of Appendix).
By using inductively coupled plasma-atomic emission spectrometry (ICP-OES), the mass loadings of Ru on Ru@NiFe‒MOF/NF were calculated to be about 0.03 wt.% when considering the entire electrode (Table S3 included by ESM of Appendix).
Fig.3(a) presents OER 95%
IR-compensated linear sweep voltammetry (LSV) curves of Ru@Ni‒MOF/NF, Ru@Fe‒MOF/NF, Ru@NiFe‒MOF/NF, Ru@NiCo‒MOF/NF, Ni‒MOF/NF, and RuO
2/NF. Among them, the target catalyst, Ru@NiFe‒MOF/NF, exhibited the best OER performance. At a current density of 10 mA·cm
−2, its overpotential was 240 mV, which was lower than those of Ru@Fe‒MOF/NF (260 mV), Ru@Ni‒MOF/NF (270 mV), Ru@NiCo‒MOF/NF (280 mV), Ni‒MOF/NF (310 mV), and RuO
2/NF (250 mV). In addition, the overpotential of Ru@NiFe‒MOF/NF was 280 mV at a current density of 50 mA·cm
−2 while 320 mV at 100 mA·cm
−2. For comparison, representative of state-of-the-art Ru-based and transition metal-based OER electrocatalysts reported in the last three years, with overpotentials below 350 mV at 10 mA·cm
−2 in 1 mol·L
−1 KOH, are selected, as shown in Tab.1 [
43–
52]. It is detected that the resulting Ru@NiFe‒MOF/NF in this work has superior catalytic performance.
In addition, the Tafel slope of each catalyst was calculated to investigate the reaction kinetics. As shown in Fig.3(b), the Tafel slope of Ru@NiFe‒MOF/NF is 43 mV·dec−1, which is lower than those of Ru@Fe‒MOF/NF (44 mV·dec−1), Ru@Ni‒MOF/NF (79 mV·dec−1), Ru@NiCo‒MOF/NF (81 mV·dec−1), Ni‒MOF/NF (85 mV·dec−1), and RuO2/NF (281 mV·dec−1), indicating that Ru@NiFe‒MOF/NF has the best OER catalytic activity in this work.
The overpotentials of Ru@Ni‒MOF/NF, Ru@Fe‒MOF/NF, Ru@NiFe‒MOF/NF, and Ru@NiCo‒MOF/NF at different current densities are shown in Fig.3(c).
To further elucidate effective active sites on the catalyst surface, we performed a detailed investigation on catalysts using cyclic voltammetry (CV) at the scan rate ranging from 20 to 120 mV·s
−1, and the results are revealed in Figs. S8‒S11 (included by ESM of Appendix). We also estimated the double-layer capacitance (
Cdl) values of catalysts. The
Cdl is usually employed as an indicator to evaluate the electrochemical active surface area (ECSA) of a catalyst [
53]. As depicted in Fig.3(d), the
Cdl value of Ru@NiFe‒MOF/NF is 3.58 mF·cm
−2, which is higher than those of Ru@Fe‒MOF/NF (2.71 mF·cm
−2), Ru@Ni‒MOF/NF (2.70 mF·cm
−2), and Ru@NiCo‒MOF/NF (2 mF·cm
−2). It is widely recognized that the ECSA is directly proportional to the
Cdl. For a catalyst, a higher
Cdl implies a larger ECSA and a bigger number of active sites exposed on the interface, thereby promoting the enhancement of OER.
To gain a deeper understanding of the catalytic activity of Ru@NiFe‒MOF/NF, electrochemical impedance spectroscopy (EIS) was conducted to further analyze the charge transfer resistance (
Rct) between the catalyst and electrolyte. As shown in Fig.3(e), Ru@NiFe‒MOF/NF has the smallest
Rct, meaning that it has the fastest catalytic reaction kinetics. This may be due to its excellent electrical conductivity, synergistic electronic coupling effect among different metal elements, and 3D highly open hierarchical porous structure, which help Ru@NiFe‒MOF/NF effectively achieve rapid charge transfer [
54–
56].
Electrocatalysts are required to maintain good stability during practical applications. Thus we studied the electrochemical long-term stability of Ru@NiFe‒MOF/NF for the OER process in 1 mol·L−1 KOH electrolyte using chronopotentiometry (CP). As shown in Fig.3(f), there was no significant change in the voltage of Ru@NiFe‒MOF/NF when it worked for 40 h at a current density of 10 mA·cm−2, demonstrating its high stability for the OER process.
Similarly, we investigated HER performances of different samples in 1 mol·L
−1 KOH solution and conducted electrochemical tests in a three-electrode system. Fig.4(a) presents LSV curves of Ru@Ni‒MOF/NF, Ru@Fe‒MOF/NF, Ru@NiFe‒MOF/NF, Ru@NiCo‒MOF/NF, Ni‒MOF/NF, and Pt/C/NF. It is detected that among them, the target catalyst, Ru@NiFe‒MOF/NF, exhibited the best HER performance. At a current density of 10 mA·cm
−2, its overpotential was 84 mV, obviously lower than those of Ru@Fe‒MOF/NF (102 mV), Ru@Ni‒MOF/NF (195 mV), Ru@NiCo‒MOF/NF (226 mV), Ni‒MOF/NF (293 mV), and Pt/C/NF (137 mV). In addition, the overpotential of Ru@NiFe‒MOF/NF was 163 mV at a current density of 50 mA·cm
−2 while 214 mV at 100 mA·cm
−2. The comparison results show that the catalytic performance of the Ru@NiFe‒MOF/NF catalyst is superior to those of many reported Ru-based and transition metal-based HER catalysts. As shown in Tab.2 [
24,
47,
57–
64], the selected catalysts are typical representatives of state-of-the-art Ru-based and transition metal-based electrocatalysts for HER reported in the last few years.
As is known to all, reaction kinetics can provide important basis for understanding HER pathways on electrocatalysts. As shown in Fig.4(b), the Tafel slope of Ru@NiFe‒MOF/NF is 95 mV·dec−1, which is lower than those of Ru@Fe‒MOF/NF (125 mV·dec−1), Ru@Ni‒MOF/NF (115 mV·dec−1), Ru@NiCo‒MOF/NF (96 mV·dec−1), Ni‒MOF/NF (108 mV·dec−1), and Pt/C/NF (140 mV·dec−1).
The overpotentials of Ru@Ni‒MOF/NF, Ru@Fe‒MOF/NF, Ru@NiFe‒MOF/NF, and Ru@NiCo‒MOF/NF at different current densities are also presented in Fig.4(c).
Long-term durability is another important parameter for the practical application of catalysts. As shown in Fig.4(d), Ru@NiFe‒MOF/NF underwent long-term electrocatalytic HER testing at a constant current density of 10 mA·cm−2. The CP curve was recorded over 40 h, with no significant change in the voltage during the testing process, indicating that Ru@NiFe‒MOF/NF has excellent stability during the HER process.
4 Conclusion
In summary, the Ru@NiFe‒MOF/NF catalyst has been prepared using a straightforward one-step hydrothermal method followed by thermal annealing. The synergistic interaction between Ru NPs and NiFe‒MOFs endows Ru@Ni‒MOF/NF with excellent catalytic performance. In 1 mol·L−1 KOH, the OER only requires an overpotential of 240 mV to reach a current density of 10 mA·cm−2, while the HER needs 84 mV to reach 10 mA·cm−2. Meanwhile, this catalyst also exhibits excellent long-term durability. Whether it is for OER or HER, Ru@NiFe‒MOF/NF can work stably for 40 h at a current density of 10 mA·cm−2 without significant voltage fluctuations. Its excellent electrochemical performance can be attributed to three key factors. Firstly, the 3D NS structure effectively facilitates the mitigation of the NP agglomeration and promotes the penetration and transport of the electrolyte. Secondly, the valence states of Ru, Ni, and Fe are diverse, providing sufficient active sites for redox reactions and improving the electrical conductivity of the material. Moreover, Fe3+ is also an effective promoter for the electrochemical OER catalytic reaction. Thirdly, the large specific surface area of the electrode material can provide more open channels and increase the accessibility of active sites, thereby enhancing the overall electrochemical performance of the material. This study not only provides a new type of alkaline water electrolysis catalyst with excellent catalytic performance, but also opens a new avenue for the rational design of other catalysts.