Trimetallic RuAuMo Thin-film Electrocatalyst Synthesized via Co-sputtering for Proton Exchange Membrane Water Electrolysis
Article information
Abstract
Proton exchange membrane water electrolysis (PEMWE) is recognized as a promising technology for producing green hydrogen without emitting pollutants. However, the high cost and scarcity of platinum as a catalyst for hydrogen evolution reaction (HER) have hindered the commercialization of PEMWE, and the alternatives to platinum have been considerably studied. In this study, trimetallic RuAuMo-based thin-film HER electrocatalysts prepared via facile co-sputtering were investigated as HER electrocatalysts. The atomic composition of the RuAuMo electrocatalyst can be easily modulated by controlling the sputtering power of each target, and the activity of RuAuMo varies with the sputtering power of Au and Mo targets. Trimetallic RuAuMo electrocatalyst with the optimum atomic composition exhibits an overpotential of 50.4 mV at a current density of –10 mA/cm2 (η10) for HER, and it outperforms monometallic Ru (84.5 mV) and bimetallic RuAu (67.7 mV) electrocatalysts. This was attributed to the strong interactions between three elements, as revealed in XRD and XPS analyses. RuAuMo thin-film deposited on porous metallic support is directly used as a cathode for PEMWE, exhibiting a high PEMWE performance with a current density of 1.39 A/cm2 at a cell voltage of 1.8 V. These results demonstrate that RuAuMo thin-film is an active HER catalyst and suggest that sputtering is an effective method for the electrode fabrication of PEMWEs.
INTRODUCTION
As a countermeasure to mitigate global warming caused by the indiscriminate use of fossil fuels, many researches have been conducted to develop the energy conversion systems connected with renewable energies such as wind and solar power [1,2]. Among the systems, water electrolysis has been recognized as one of eco-friendly energy-conversion process that produce only hydrogen and oxygen without any pollutants [3]. In addition, when water electrolysis is driven by electricity generated from renewable energy sources, and accordingly, an eco-friendly hydrogen production system can be established [4,5]. As compared to traditional water electrolysis system, membrane electrode assembly (MEA)-based configuration recently has attention owing to its compact structure, low ohmic resistance, and high hydrogen production efficiency at high operating current densities [6,7]. Among MEA-based systems, proton exchange membrane water electrolysis (PEMWE) generally exhibits higher hydrogen production rates with higher purity than anion exchange membrane water electrolysis [8–11].
Meanwhile, considering the acidic operating environment of PEMWE, a significant challenge arises from the reliance on platinum group metals as electrode materials for water electrolysis [12,13]. Specifically, Pt is recognized as the most effective electrocatalyst for the hydrogen evolution reaction (HER), a crucial half-reaction in water electrolysis [14]. However, its scarcity and high cost increase the MEA fabrication expenses, ultimately hindering the scale-up and industrial commercialization of PEMWE. Therefore, the development of non-Pt-based electrocatalysts and electrodes is essential for achieving a low-cost water electrolysis [15]. Among the alternatives to Pt, Ru has recently gained attention as an electrocatalyst due to its ~2.4 times lower price compared to Pt (Pt: $973.9/oz, Ru: $400.0/oz) [16]. Additionally, Ru exhibits a similar hydrogen bond strength (65 kcal/mol) to Pt (62 kcal/mol), potentially leading to comparable catalytic behavior [17,18]. However, pure Ru (η10: 99 mV) has a lower intrinsic HER performance than Pt (η10: 25 mV) [19], prompting numerous studies aimed at enhancing its performance and durability through alloying with elements such as Co [20,21], Fe [22], Ni [23,24], Cu [25], Mo [26], and Au [16,27–29]. Among these, the Ru-Au bimetallic catalyst is considered a promising Ru alloy catalyst for the HER because inert Au in an acidic environment can improve its activity and durability (η10: 43 mV) [29]. However, the hydrogen binding characteristics of Ru-Au are not optimal for HER, resulting in still lower HER activity than that of Pt [28]. Z. Zhang et al. [26] reported that the addition of Mo to Ru allows for fine-tuning of the hydrogen binding energy and this research findings provides inspiration for studying the Ru-Au-Mo ternary alloy.
Generally, the conventional electrode for PEMWE has been fabricated by drop-casting or spraying the catalyst powder via the following processes: (1) synthesizing nanoparticle catalysts, (2) preparing the catalyst slurry by blending with an ionomer solution, and (3) loading the slurry onto the membrane or porous transport layer (PTL) [30–32]. This method involves multiple steps and requires a relatively long fabrication time. In contrast, thin-film deposition methods, such as sputtering [33–35], chemical vapor deposition [36–38], and electrodeposition [39–41], can directly fabricate the catalyst layer with high coverage on the substrate and thin-film structures are effective in improving cell performance by increasing catalyst utilization [42]. Among these methods, sputtering allows for the formation of thin-films with high quality and reproducibility in a short processing time, which simplifies the electrode preparation step [43]. This one-pot process also provides significant advantages in the fabrication of porous transport electrodes (PTE) for PEMWE. Unlike typical PTE fabrication methods, sputtering does not require polymer binders, which enhance the electrical conductivity during water electrolysis [44]. In addition, the thin-film structure contributes to minimizing ohmic resistance and promoting the mass transport of reactants and products [45]. Moreover, the sputtering method can easily scale up the electrode area, making it suitable for water electrolysis to produce large amounts of hydrogen. Despite these advantages, the sputtering method faces significant challenges in electrode fabrication [33]; Its deposition rate is still low (< 50 nm/min), which leads to a long sputtering time for the high loading of the catalyst layer. Accordingly, successful utilization of the sputtering method for the fabrication of the catalyst eventually requires the enhancements of the activity of the catalyst even under the low loading condition, by the development of catalyst alloying as aforementioned.
In this concerns, sputtering method can easily prepare atomically uniform alloy-based catalysts when using the co-sputtering of multiple targets, where the atomic concentration of the alloy is controllable by adjusting the sputtering power condition [33]. In this study, we present a novel approach to fabricating RuAuMo thin films using a co-sputtering method onto a conductive support, with the films serving as efficient electrocatalysts for the HER in an acidic medium. By precisely controlling the sputtering power (Pmetal) of each target element, we were able to fine-tune the atomic concentrations of Ru, Au, and Mo in the thin films. The HER performance of the RuAuMo films was investigated as a function of atomic composition through electrochemical analysis, demonstrating significant enhancement in catalytic activity with optimized atomic ratios. Notably, we directly fabricated RuAuMo thin films onto a PTL for application as cathodes in PEMWE. The RuAuMo cathode exhibited superior cell performance compared to Ru and RuAu electrodes, emphasizing the synergetic effects of trimetallic incorporation in improving HER activity.
EXPERIMENTAL
RuAuMo-based thin-film formation via magnetron sputtering deposition
The RuAuMo thin-film was deposited onto Ti/TiN/Si and Ti foil substrates using radio-frequency magnetron sputtering. Before the sputtering deposition, the Ti foil was pretreated in 5 wt% oxalic acid at 70°C for 30 minutes to remove the native oxide layer. The pretreated Ti support was placed in the sputtering chamber and pressurized to 15 mTorr. The sputtering power (Pmetal) for each target element was modulated from 10 to 70 W for Ru, 15 to 60 W for Au, and 15 to 120 W for Mo, respectively. The sputtered area of the Ti foil was 1.5 cm2, and the co-sputtering was conducted for 1800 s. The RuAu film was prepared using the same process, except that the Mo target was not used. Additionally, Ti PTL (Bekaert, 250 μm, porosity: 60%) was used as the PTL to fabricate the RuAuMo-based PTE for PEMWE. The Ti PTL was also treated using the same method as the Ti foil. The optimized sputtering conditions were applied to the pretreated Ti PTL with a surface area of 4 cm2 to fabricate the RuAuMo-based PTE.
Characterization
The morphology and cross-section of the RuAuMo films on the Ti/TiN/Si substrates were analyzed using scanning electron microscopy (SEM, Inspect F50, Elecmi), while the bulk atomic composition was determined using energy dispersive spectroscopy (EDS). Grazing incidence X-ray diffraction (GIXRD, D8 Advance, Bruker) was used to investigate the crystal structures of the RuAuMo thin-film. The electronic structures of the RuAuMo films were characterized using X-ray photoelectron spectroscopy (XPS, PHI 5000 VersaProbe, UlvacPHI), and the metal loading amounts were analyzed by inductively coupled plasma mass spectrometry (ICPMS, NexION 2000, PerkinElmer).
Electrochemical measurement in half-cell system
The electrochemical properties of co-sputtered RuAuMo films on Ti foil were measured in a 1.0 M H2SO4 acidic solution. The as-prepared RuAuMo electrocatalysts were used as the working electrode, with an active area of 0.264 cm2. The counter and reference electrodes were a Pt wire and Ag/AgCl (KCl-saturated), respectively. Cyclic voltammetry (CV) was conducted to measure the HER performance of the RuAuMo-based electrocatalysts in the potential range from open circuit potential to –0.4 V vs. Ag/AgCl with a scan rate of 5 mV/s. Nyquist plots for the RuAuMo electrocatalysts were obtained from potentiostatic electrochemical impedance spectroscopy (PEIS) at –75 mV in the frequency range of 100 mHz to 100 kHz. All potentials from the measurements were converted to the reversible hydrogen electrode (RHE) scale.
MEA fabrication and PEMWE single cell operation
The MEA for PEMWE consisted of a cathode, an anode, and a proton exchange membrane (PEM, NR212, DuPont Co.) positioned between two electrodes with a zero-gap configuration. The cathode was composed of RuAuMo films deposited onto a Ti PTL. For the anode, a catalyst slurry containing IrO2 powder (Alfa Aesar, 99.9% metal base), deionized (DI) water, Nafion ionomer (Alfa Aesar, 5% Nafion®), and isopropanol was spray-coated onto the PEM. The loading amount of IrO2 was 1.0 mg/cm2. The RuAuMo-based cathode, IrO2-coated PEM, and Ti PTL anode for the anode were assembled by hot pressing for 2 minutes. The active area of the MEA was 1.0 cm2.
The PEMWE single-cell test was conducted at 80°C, with DI water pre-heated to the same temperature and injected into the anode side at a flow rate of 15 ml/min. Before investigating the single-cell performance, the cell was electrochemically activated by conducting chronoamperometry at 1.55 V for 30 minutes. After activation, the current-voltage (i-V) polarization curve was measured from the open-circuit voltage to 2.2 V to assess the performance of the PEMWE cell. Additionally, PEIS measurements were conducted at 1.5, 1.8, and 2.1 V, with a frequency range from 100 mHz to 100 kHz.
RESULTS AND DISCUSSION
Each of the elements Ru, Au, and Mo was deposited via sputtering onto a Ti/TiN/Si wafer with a fixed sputtering time of 1800 s. Fig. 1 shows the thickness of each elemental deposit as a function of Pmetal. The thicknesses of Ru, Au, and Mo increased monotonically with Pmetal, indicating that the deposition rate increases with Pmetal and varies with the type of metal target. In the cross-sectional SEM images, the Ru films fabricated at a PRu of 30 W had a thickness of 97 nm, which gradually increased to 250 nm at 90 W (Fig. 1b–d).
(a) Ru, Au and Mo films thickness according to Pmetal of each target element. Cross-section SEM images of Ru films onto Ti/TiN/Si wafer with varying PRu: (b) 30 W, (c) 70 W and (d) 90 W (sputtering time: 1800 s).
We fabricated bimetallic RuAu-based films via co-sputtering of Ru and Au targets. The PRu was fixed at 70 W, while that of PAu varied from 10 to 50 W. Fig. 2a shows the atomic composition of RuAu-based bimetallic films analyzed by EDS, where the Au ratio gradually increased with higher PAu. The HER performance of the as-prepared RuAu-bimetallic electrocatalysts was evaluated in a 1.0 M H2SO4 solution. The RuAu-bimetallic electrocatalysts were denoted as RuAu#, where # represents the PAu in watts. Fig. 2b presents the polarization curves for RuAu electrocatalysts, with the inset displaying the η10 of HER plotted against PAu. The h10 for pure Ru was 84.5 mV. At a PAu of 10 W, a decreased HER h10 was measured. However, as the PAu was further increased, the HER h10 also increased, reaching 85.0 mV for RuAu50. Although Au has good electrical conductivity, its weak hydrogen adsorption properties can deteriorate the HER performance of RuAu-bimetallic electrocatalysts [27,28]. Therefore, the HER h10 of RuAu against PAu, or Au concentration, exhibited a volcano-like tendency, with RuAu10 showing the maximum HER performance compared to bare Ru and other RuAu# samples. The Tafel plots derived from the HER polarization curves confirmed that the Tafel slopes for Ru and RuAu electrocatalysts ranged from 60 to 70 mV/dec, indicating that the HER proceeded via the Volmer–Heyrovsky mechanism (Fig. 2c) [14]. Additionally, RuAu10 exhibited a Tafel slope of 61.9 mV/dec, indicating faster reaction kinetics for HER among the electrocatalysts (RuAu30: 63.0 mV/dec and RuAu50: 67.5 mV/dec). Considering Randles circuit model, the Nyquist plot in Fig. 2d also exhibited the same tendency of HER polarization curve and Tafel plot for RuAu-bimetallic electrocatalysts. Based on these results, RuAu10, fabricated with PRu, PAu of 70 W for Ru and 10 W for Au, was selected as the optimal bimetallic electrocatalyst.
(a) Atomic concentrations of RuAu films by EDS analysis according to PAu. (b) The polarization curves of Ru and RuAu# (#: PAu) electrocatalysts at a scan rate of 5 mV/s (Inset: the HER η10). (c) Tafel plots of Ru and RuAu# electrocatalysts corresponding to polarization curves from Fig. 2b. (d) The Nyquist plots of Ru and RuAu# electrocatalysts at –75 mV. All electrochemical measurements were performed in 1.0 M H2SO4.
To further enhance the electrochemical properties of RuAu-based films, Mo was introduced as a third metal for sputtering, with varying PMo from 0 to 110 W. The RuAuMo trimetallic films were denoted as RuAu10Mo#, where # represents the PMo in watts. The top-view SEM images in Fig. 3a–d revealed slightly rough surfaces for the RuAu10Mo# films. Additionally, all RuAuMo films exhibited similar morphologies regardless of PMo. The atomic concentrations of RuAu10Mo# films were investigated by EDS, as shown in Fig. 3e. As the PMo increased, the Mo atomic ratio increased, while the Ru ratio decreased. Conversely, the Au atomic composition in RuAu10Mo# films remained in the range of 6–12%, showing minimal influence from Mo sputtering.
SEM surface images of RuAu10Mo# films according to PMo: (a) 0 W, (b) 20 W, (c) 30 W and (d) 110 W (#: PMo). (e) Atomic concentrations of RuAu10Mo# films by EDS analysis according to PMo.
Fig. 4 shows the XRD patterns of Ru and RuAuMo thin films fabricated by sputtering. For the Ru film, Ru (100), Ru (002), and Ru (101) peaks were detected at 38.3°, 42.1°, and 44.1°, respectively. The broaden peaks for Ru deposit was detected in XRD pattern, and it seems to be that the weak PRu generated the low-crystalline phase of Ru deposits with random growth [46,47]. As the Mo content in RuAuMo deposits increased, the intensities of these Ru peaks gradually decreased, and a new peak emerged at 41.3°, which was slightly shifted to 40.4°. This peak shift in the XRD pattern may have originated from the formation of the alloy structure of the trimetallic RuAuMo thin film. In previous literature, the similar XRD peak shift was reported that the behavior of RuMo alloy was owing to Ru lattice expansion [48]. In the RuAu10Mo110 sample with a Mo ratio of 44.8%, the intensity of Ru peaks was decreased while the influence of Mo crystallinity became dominant.
XRD patterns of Ti substrate, Ru and RuAu10Mo# films (#: PMo).
Meanwhile, XPS analyses of RuAu10 and RuAu10Mo# deposits were conducted in order to investigate the electronic structure of trimetallic RuAuMo thin film (Fig. 5). As shown in Fig. 5a, the Ru 3p1/2 and 3p3/2 spectra for RuAu10 and RuAu10Mo# film were deconvoluted into three peaks, which corresponded to the Ru0 peak and two Ru satellite peaks. Compared to metallic Ru (461.2 eV), the Ru0 peak of RuAu10 was shifted to a higher binding energy, indicating the presence of electronic interactions between Ru and Au. Previous literature suggests that the peak shift is due to electron transfer from Ru to Au along with the formation of the Ru-Au alloy [16,27,28]. Furthermore, when Mo species were added to RuAu via co-sputtering, the Ru0 peak in RuAu10Mo# was slightly shifted in the negative direction relative to that of RuAu10. Conversely, Fig. 5b shows that both RuAu10 and RuAu10Mo# exhibited an Au 4f peak shift compared to metallic Au (84.0 eV), confirming that the co-sputtering of Ru, Au, and Mo forms an alloy structure. Interestingly, a positive peak shift for Au 4f was observed in all samples, despite the higher electronegativity of Au (EN: 2.54) compared to Ru (EN: 2.20) and Mo (EN: 2.16). This shift may be attributed to the small fraction of Au analyzed by EDS, which resulted in a decrease in electron density of Au in RuAu10Mo# deposits [29]. The XRD and XPS analyses revealed that the sputtering method is suitable for fabricating the alloy structure of Ru, Au, and Mo.
(a) Ru 3p (red: Ru0, blue and magenta: Ru satellite, cyan: RuO2, respectively) and (b) Au 4f (red and green: Au0, respectively) XPS spectra for RuAuMo films with various PMo.
Fig. 6 presents the HER performance of trimetallic RuAuMo# electrocatalysts measured in a 1.0 M H2SO4 solution. According to the polarization curves in Fig. 6a, the bimetallic RuAu and trimetallic RuAuMo electrocatalysts demonstrate better HER activity compared to monometallic Ru in an acidic medium, indicating that alloy structures with multiple elements contribute to enhanced HER performance. Fig. 6b shows the HER h10 derived from the polarization curves. At a PMo of 30 W, the h10 was minimized to 50.4 mV. However, a further increase in PMo deteriorated the activity (η10 = 81.1 mV for RuAu10Mo110). In previous literatures, it has generally reported that the HER performance of the electrocatalyst is depending on various factors; morphology, catalyst structure, crystallinity, atomic composition, electronic structure, and support [49,50]. In this study, the RuAuMo electrocatalysts with varying PMo showed similar surface morphologies, as shown in Fig. 3. This suggests that other factors, such as surface composition, crystal structure, and electronic structure, could play a more prominent role in influencing the variations in HER performance.
Based on the results from the half-cell system, the RuAu10Mo30 electrocatalyst exhibited the best HER performance with facile charge transfer in an acidic medium and was selected as a cathode catalyst for PEMWE. In order to apply the PTE into PEMWE, the RuAu10Mo30 was deposited via sputtering onto porous Ti PTL as a support. Optimized Ru and RuAu on Ti PTL were also fabricated for comparison. The ICP-MS analysis revealed that the total amount of Ru, Au and Mo loaded on Ti PTL for RuAu10Mo30 was 30.96 μg/cm2. Fig. 7a presents the i-V curves of the PEMWE single cell with Ru, RuAu, and RuAuMo electrodes. At a cell voltage of 1.8 Vcell, the RuAu10Mo30/Ti PTL cathode achieved a current density of 1.39 A/cm2, demonstrating better PEMWE performance than Ru/Ti PTL (0.93 A/cm2) and RuAu10/Ti PTL (1.20 A/cm2). Furthermore, compared to values reported in previous literatures employing Ru-based cathodes, the performance of the trimetallic RuAuMo is competitive, when considering its low catalyst loading (Fig. 7b) [21,26,52–54].
(a) The polarization curves of PEMWE single cell with Ru/Ti PTL, RuAu10/Ti PTL and RuAu10Mo30/Ti PTL cathode. (b) PEMWE performance comparison at the cell voltage of 1.8 Vcell including Ru-based catalysts, against previous literatures.
CONCLUSIONS
In this study, trimetallic RuAuMo-based thin-film electrocatalysts for the HER were fabricated via a facile sputtering method, and their performance in a PEMWE cell was investigated. For the sputtering method, it was easy to control the atomic compositions of RuAuMo-based films by modulating the PMetal of each target element. With varying the PMo, a trade-off relationship between the Ru and Mo ratios was identified. In addition, XRD and XPS analysis confirmed that the RuAuMo films fabricated via sputtering formed an alloy structure. The trimetallic RuAuMo electrocatalyst exhibited enhanced HER activity compared to the monometallic Ru and bimetallic RuAu electrocatalysts, as the optimized RuAu10Mo30 electrocatalyst exhibited an HER h10 of 50.4 mV in a half-cell system. RuAuMo was directly fabricated onto Ti PTL for use as a cathode for PEMWE. The as-prepared RuAu10Mo30/Ti PTL cathode exhibited a current density of 1.39 A/cm2 at 1.8 Vcell. The mass activity of RuAu10Mo30/Ti PTL for PEMWE was relatively high compared to reported values in previous literature. Consequently, this study demonstrates the RuAuMo alloy thin-film electrocatalysts fabricated via sputtering is a promising cathode for PEMWE systems.
Acknowledgements
This work was supported by the National Research Foundation of Korea (NRF) grant funded by Ministry of Science and ICT (MIST) [NRF-2020M1A2A2080806] and by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) grant funded by the Ministry of Trade, Industry & Energy (MOTIE) [No. 20203010030010 and No. 20218801010030]. This work was also supported by the Korea Institute of Science and Technology (KIST) Institutional Program (2E33281).