Tailoring the Chemical Environment of Ru–Mo Composites for Efficient Hydrogen and Oxygen Evolution Reactions

The development of cost‐effective and stable electrocatalysts that can replace the prevailing Pt‐based and Ir‐based catalysts for water splitting remains a formidable challenge. The electrocatalytic performance of a catalyst depends on its chemical composition and environment during catalysis. Herein, a Ru–Mo composite is used as an example to highlight the significance of tailoring the chemical environments of the active‐sites. This customization enhances the cathodic and anodic reactions involved in water splitting, ultimately leading to an improved full reaction efficiency. Specifically, the chemically reduced state of Ru–Mo demonstrates promising hydrogen evolution activity and exhibits low overpotentials of 48 and 34 mV at the current density of 10 mA cm−2 in acidic and alkaline electrolytes, respectively. The chemically oxidized state of Ru–Mo, derived from the same precursor, demonstrates proficient oxygen evolution activity and exhibits low overpotentials of 260 and 270 mV at 10 mA cm−2 in acidic and alkaline electrolytes, respectively. Additionally, in both cases, the Ru–Mo composites exhibit significantly improved stability under typical water‐splitting conditions. Despite the close chemical compositions of these two catalysts, they show poor performance in the counter electrode reactions, demonstrating the importance of establishing a suitable chemical environment for efficient electrocatalysis.


Introduction
Hydrogen is recognized as a promising alternative to fossil fuels due to its high gravimetric energy density and zero carbon emission after combustion. [1,2]Large-scale implementation of hydrogen has the potential to mitigate environmental issues caused by fossil fuel consumption to accelerate the transition toward sustainability. [3][6] Therefore, the world needs green hydrogen production strategies to mitigate carbon emissions.Electrochemical water splitting driven by renewable energy sources, such as solar energy, wind, or geothermal, is an appealing approach due to its carbon-free production process and scalability. [7]During water splitting, two half-reactions namely hydrogen evolution reaction (HER) and oxygen evolution reaction (OER) occur at the cathode and anode, respectively. [8]The sluggish reaction kinetics on both electrodes, particularly the anode, require high-voltage input to drive the full reaction, resulting in significant energy consumption and ultimately increasing the production cost of hydrogen.To address these challenges, highly active catalysts are essential for promoting the reaction kinetics and minimizing the energy losses.Currently, the-state-of-the-art catalysts for the overall water splitting still rely on Pt-based and Ir-based materials. [9]owever, the high cost and limited abundance of Pt and Ir pose significant challenges for their large-scale applications, motivating the discovery of cost-effective alternatives for water splitting.One promising option is the use of Ru-based catalysts.][12][13] Moreover, the cost of Ru is only about 4% of Pt, [14] making it an attractive candidate to reduce the overall cost of catalytic materials for water splitting.Unfortunately, the wide use of Ru-based catalysts in electrochemical applications is hindered by their unsatisfactory stability, as Ru is prone to dissolve into electrolytes under HER and/or OER conditions. [15]Since Ru shows lower dissolution potential compared to Ir, the dissolution rate of metallic Ru can be 1-2 orders of magnitude higher than that of Ir, whereas for Ru oxides, it can be %30 times higher than that of IrO 2 , under the same operational conditions. [15]Therefore, it is desirable to develop more stable Ru-based catalyst for electrochemical applications.
According to the Brewer-Engel theory, [16] appropriate alloying of the left-hand-side transition metals with the right-hand-side transition metals, such alloying Mo with Ru, can potentially DOI: 10.1002/sstr.202300394 The development of cost-effective and stable electrocatalysts that can replace the prevailing Pt-based and Ir-based catalysts for water splitting remains a formidable challenge.The electrocatalytic performance of a catalyst depends on its chemical composition and environment during catalysis.Herein, a Ru-Mo composite is used as an example to highlight the significance of tailoring the chemical environments of the active-sites.This customization enhances the cathodic and anodic reactions involved in water splitting, ultimately leading to an improved full reaction efficiency.Specifically, the chemically reduced state of Ru-Mo demonstrates promising hydrogen evolution activity and exhibits low overpotentials of 48 and 34 mV at the current density of 10 mA cm À2 in acidic and alkaline electrolytes, respectively.The chemically oxidized state of Ru-Mo, derived from the same precursor, demonstrates proficient oxygen evolution activity and exhibits low overpotentials of 260 and 270 mV at 10 mA cm À2 in acidic and alkaline electrolytes, respectively.Additionally, in both cases, the Ru-Mo composites exhibit significantly improved stability under typical water-splitting conditions.Despite the close chemical compositions of these two catalysts, they show poor performance in the counter electrode reactions, demonstrating the importance of establishing a suitable chemical environment for efficient electrocatalysis.
increase the number of available electron pairs and expose more d-orbitals for H adsorption during the electrochemical reaction, resulting in improved electrocatalytic performance.Hence, in this work, we introduced Ru into the same Mo precursor using imidazole ligand-assisted method and prepared two distinct RuMo composites with different chemical environments of Ru active sites via different treatments.These two RuMo catalysts exhibit improved activity/stability for HER and OER compared to the commercial Pt and Ir catalysts, respectively.Based on comprehensive physical characterizations, we confirm that the HER-active RuMo composite is composed of Ru active sites surrounded by Mo 2 C and is in a reductive state, while OER-active RuMo composite consists of Ru oxides embedded into MoO 3 and the Ru species is in an oxidized state.The sample with Ru active-sites surrounded by Mo 2 C is referred to as RuMo-Ar (treated in Ar), and it exhibits excellent catalytic performance and exceptional stability in HER.Using RuMo-Ar for HER, the overpotentials required to achieve a current density of 10 mA cm À2 are 48 and 34 mV in acidic and alkaline electrolytes, respectively.The latter sample, denoted as RuMo-Air (treated in Air), exhibits promising performance for OER.Overpotentials of only 260 and 270 mV are required to achieve the OER current density of 10 mA cm À2 using RuMo-Air in acidic and alkaline electrolytes, respectively.Notably, while RuMo-Ar shows exceptional stability in HER, it is inactive and degrades rapidly under the OER operational conditions.Conversely, RuMo-Air exhibits poor activity in HER, highlighting the importance of an appropriate chemical environment in enhancing both the catalytic performance and stability of an electrocatalyst.

Catalysts Synthesis and Morphology
As shown in Figure 1a, the precursor for both electrocatalysts, RuMo-Ar and RuMo-Air, was prepared through a hydrothermal method that incorporates Ru-and Mo-based salts in the presence of imidazole.Subsequently, annealing processes were carried out under Ar and air atmospheres to obtain the two electrocatalysts for HER and OER (detailed synthetic procedures are presented in the Supporting Information), respectively.Field-emission scanning microscopy (FESEM) was employed to investigate the morphology of the catalysts.The results showed that both RuMo-based catalysts exhibited sphere-like characteristics with a size distribution of around 100-200 nm (Figure 1b,e).Transmission electron microscopy (TEM) images were obtained to further confirm the sphere-like morphology of these two catalysts (Figure 1c,d,f,g).Note that the surface of RuMo-Air, which was synthesized through air pyrolysis, appears to be rougher.This may be due to the removal of organic moieties from the framework, which creates voids and leads to a more irregular surface.Results from energy-dispersive X-Ray spectroscopy (EDS) (Figure S7, Supporting Information) suggest that the Ru and Mo elements are uniformly dispersed in the samples of RuMo precursor, RuMo-Ar, and RuMo-Air.Additionally, the oxygen signal in RuMo-Air is much stronger than that in RuMo-Ar, indicating that RuMo-Air is primarily composed of metal oxides, while RuMo-Ar is likely a metallic composite.Moreover, the use of the imidazole ligand and the bimetallic composition were found to play critical roles in achieving the uniform spherical morphology of the catalysts.As illustrated in the FESEM images in Figure S8 (Supporting Information), samples synthesized without imidazole or Mo exhibit nonuniform morphology/particle sizes, regardless of the atmosphere in which they were prepared.To further demonstrate the advantages of using imidazole ligands and the incorporating of Mo, the electrochemically active surface area (ECSA) of all samples was estimated by measuring their double-layer capacitances (C dl ), which reflects the number of active sites. [17]It was found that the samples containing both imidazole and Mo exhibited the largest ECSA regardless of annealing atmosphere (Figure S9 and S10, Supporting Information).[20][21] It is also worth noting that the ECSA of RuMo-Air is significantly larger than that of RuMo-Ar, which is consistent with the TEM observations.It can be attributed to the removal of the imidazole during annealing in air, which likely creates a more porous structure and consequently increases the surface area.

Chemical States of the as-Prepared Catalysts
X-Ray photoelectron spectroscopy (XPS) was employed to investigate the chemical states of the obtained RuMo-based catalysts.As shown in Figure 2a-c, there are significant differences in the chemical states of Ru and Mo between the two catalysts.XPS survey spectra (Figure 2a) show discernible peaks for Ru and Mo in both catalysts, with a stronger signal intensity for oxygen observed in RuMo-Air than RuMo-Ar, which is consistent with the EDS results.In the case of RuMo-Ar, both Ru and Mo tend to exist in a low-valence state or metallic state.In contrast, Ru and Mo in RuMo-Air tend to exist in higher-valence state or form oxides.The high-resolution XPS of Mo 3d (Figure 2b) shows that Mo 2þ species dominate in RuMo-Ar, and the main peaks at 227.86 and 232.26 eV can be attributed to Mo 2 C 3d 5/2 and MoO 3 3d 5/2 , [22,23] respectively.In RuMo-Air, the Mo species shift to a higher-valence state, and the peaks at 231.74 and 233.25 eV can be assigned to MoO 3 /C 3d 5/2 and MoO 3 3d 5/2 , [24,25] respectively.The observed shift in binding energy of the MoO 3 species between RuMo-Ar and RuMo-Air can be attributed to the difference in sample processing conditions. [26,27]The chemical states of the Ru in the catalysts were also investigated (Figure 2c).In RuMo-Ar, a major peak at 461.09 eV was observed, which could be assigned to metallic Ru 3p 3/2 [28] In contrast, the binding energy of the Ru peaks deconvoluted for RuMo-Air showed a positive shift to 463.02 and 465.97 eV, which could be attributed to the characteristic peaks of RuO 2 3p 3/2 and the satellite structure, [29] respectively.These results suggest that annealing of the RuMo precursor in an O 2 -rich atmosphere can effectively transform the metallic components to their higher chemical states, while the absence of O 2 can lead to the formation of lower chemical state counterparts.Based on these XPS results, the surface elemental concentration ratios were evaluated and the Ru/Mo ratios in RuMo-Ar and RuMo-Air were found to be 0.64 and 0.88, respectively (Table S1, Supporting Information).Both values are higher than the corresponding bulk Ru/Mo ratios of 0.39 and 0.33, as measured by inductively coupled plasma-optical emission spectrometry (ICP-OES).This suggests that there is a higher concentration of Ru on the catalyst surface compared to the bulk materials, suggesting that more active sites are exposed for reactions.
The X-Ray diffraction (XRD) patterns shown in Figure 2d are consistent with the above XPS results.The Mo-associated diffraction peaks of RuMo-Ar could be indexed to the Mo 2 C phase, [30] while those of RuMo-Air could be assigned to the MoO 3 phase. [31][34][35] To further explore this, high-resolution TEM (HRTEM) was utilized to investigate the surface lattice structure of both catalysts.Regarding RuMo-Ar (Figure S11a, Supporting Information), the observed lattice spacing of 0.231 and 0.237 nm can be assigned to the (121) facet of Mo 2 C and the (100) facet of Ru, [36,37] respectively.For RuMo-Air (Figure S11b, Supporting Information), lattice spacings of 0.330 and 0.321 nm were found in the HRTEM image, corresponding to the (021) facet of MoO 3 and (110) facet of RuO 2 , [38,39] respectively.The HRTEM results suggest that both catalysts exhibit intermingling of Mo and Ru, and the crystalline phase of Mo is consistent with the corresponding XRD patterns.Taken together, we have successfully fabricated two RuMo-based catalysts with different chemical environments.One is characterized by metallic Ru surrounded by Mo 2 C and is present in a relatively reductive chemical environment, while the other one consists of RuO 2 embedded into the MoO 3 and is present in a relatively oxidative chemical environment.

HER on RuMo-Ar
][42][43] Thus, we were motivated to evaluate RuMo-Ar toward HER on the rotating disk electrode (RDE).As shown in Figure 3a,d,g, RuMo-Ar shows high HER activity in both acidic and alkaline electrolytes.The overpotentials required to reach the current density of 10 mA cm À2 are 48 and 34 mV versus RHE (all the potential reported within this work is against to RHE unless defined otherwise), respectively.This performance surpasses those of the control samples which lack imidazole or Mo (denoted as RuMo-Ar WO imi and Ru-Ar).These control samples with incomplete ingredients exhibit inferior HER performance, requiring significantly larger overpotentials to achieve the same current density as RuMo-Ar.This is particularly evident in the case of RuMo-Ar WO imi, which failed to reach the current density of 10 mA cm À2 within the selected potential window.Pt-based catalysts have been experimentally and theoretically proven to be the best HER catalyst due to their optimal H adsorption strength, particularly under acidic conditions. [3,44]The overpotentials for Pt to make a current density of 10 mA cm À2 are 17 and 38 mV in acidic and alkaline electrolytes (Figure 3a,d,g), respectively.[47] To analyze the HER reaction kinetics, Tafel slopes of RuMo-Ar and Pt/C were measured in both acidic and alkaline electrolytes.Note that it is important to use low catalyst loading to mitigate the influence of H 2 mass transportation and to allow for a better assessment of the intrinsic activity of the catalysts using Tafel analysis, particularly for catalysts with high activity. [44,45]As shown in Figure 3b,e and Figure S12 (Supporting Information), the Tafel slopes obtained for RuMo-Ar increase significantly as the catalyst loading decreases, indicating that the mass transportation effect has been diminished at the selected potential window.With a low catalyst loading of 0.004 mg cm À2 , Tafel slopes of 85.6 and 61.5 mV dec À1 were observed for RuMo-Ar in acidic and alkaline electrolyte, respectively.Both values are in the between 120 and 40 mV dec À1 , suggesting the HER rate is controlled by a mix of Volmer and Heyrovsky steps. [48]For Pt/C, a similar Tafel slope (83.2 mV dec À1 ) was observed from the alkaline electrolyte, indicating the same HER rate-determining step (RDS) with RuMo-Ar.An extremely low Tafel slope of 15.6 mV dec À1 was obtained in acidic electrolyte (Figure 3c), which reaffirms the unparalleled activity of Pt/C to HER.Such a low Tafel slope suggests that the Pt/C is still controlled by mass transportation of H 2 and it is difficult to interpretate the reaction kinetics under this condition. [49]Notably, as shown in Figure S13a,b,e and S14a,b (Supporting Information), RuMo-Ar exhibits higher exchange current density and turnover frequency (TOF) under alkaline condition compared to that in acidic electrolyte, and it is even higher than that of Pt/C in alkaline electrolyte, indicating its improved intrinsic HER activity in an alkaline electrolysis and the potential to replace Pt in alkaline electrolyzers.
Electrochemical impedance spectroscopy (EIS) was also employed to investigate the reaction kinetics for HER on RuMo-Ar.The corresponding Nyquist plots of RuMo-Ar and Pt/C are presented in Figure 3h,i, respectively.The Nyquist plot curve high-frequency region fittings were performed based on the corresponding equivalent circuits to extract the charge transfer resistance (detailed fitting results for different samples at different electrolytes are presented in Figure S15a, Supporting Information). [45]For RuMo-Ar, the obtained R ct are 1.39 and 1.79 Ω in acid and base, respectively.On the other hand, for Pt/C, the R ct of 0.62 and 1.95 Ω were obtained under the same experimental conditions.The comparable charge transfer resistances of RuMo-Ar and Pt/C in alkaline electrolyte further confirmed the fast HER kinetics of RuMo-Ar.We believe that the observed high activity of RuMo-Ar for HER can be attributed to its uniform catalyst morphology and suitable reductive chemical environment around the Ru active sites.The uniform catalystmorphology facilitates mass transfer, [50][51][52] while the reductive chemical environment of the Ru active sites optimizes the H adsorption strength and then promotes hydrogen evolution reaction kinetics. [45,53,54]tability is a crucial indicator for assessing a given catalyst's performance in a catalytic reaction.During the measurement of HER, a common issue that may arise is the detachment of the catalyst from the RDE surface.To address this issue, carbon paper was chosen as an alternative catalyst substrate for conducting accelerated durability tests (ADT).To ensure the accuracy of the ADT, the HER performance of RuMo-Ar on carbon paper was assessed, and the same activity trends as the results obtained from the RDE hold (Figure S16a,b, Supporting Information).The ADT results (Figure S17a,b, Supporting Information) demonstrate that RuMo-Ar exhibits excellent stability and outperforms Pt/C in both acidic and alkaline electrolytes.Specifically, it retained its initial activity even after 10 000 cycles, whereas Pt/C exhibited noticeable degradation only after 1000 cycles.

OER on RuMo-Air
[57] Here we assessed the RuMo-Air for OER since its Ru component is in oxidized state.As shown in Figure 4a,d, RuMo-Air demonstrates the highest activity compared to its counterparts and the-state-of-the-art Irbased catalyst for both acidic and alkaline OER.The overpotentials required for RuMo-Air to reach an OER current density of 10 mA cm À2 are 260 and 270 mV in the acidic and alkaline electrolytes, respectively.In comparison, for Ir/C, these values are 330 and 340 mV.Additionally, similar to HER, samples lacking any of the components, such as Mo element or imidazole, exhibit inferior catalytic performance for OER compared to that of RuMo-Air.Similarly, Tafel analysis was conducted using low catalyst loadings to exclude the mass transportation effect.As shown in Figure 4b,c,e,f  (82.7 mV dec À1 in acidic solution and 78.9 mV dec À1 in alkaline solution).As suggested previously, the first electron transfer step of OER as RDS results in a Tafel slope of 120 mV dec À1 .This slope decreases to 40 mV dec À1 when the RDS shifts to the second electron transfer step. [58]Thus, Tafel slopes between 120 and 40 mV dec À1 for both RuMo-Air and Ir/C may indicate a mixed RDS involving both the first and second electron transfer steps.Smaller overpotentials and larger apparent exchange current density ( j 0 ) extracted from the Tafel slopes of RuMo-Air (Figure 4g and Figure S13c-e, Supporting Information) indicate its enhanced electrocatalytic activity toward OER compared to Ir/ C. TOF is also calculated to learn about the intrinsic activity of the catalysts.As shown in Figure S14c,d (Supporting Information), the TOF values of RuMo-Air are 17.8 and 11.1 times higher than Ir/C in acidic and alkaline electrolyte, respectively, reconfirming its higher activity on OER.EIS results further confirm the high activity of RuMo-Air for OER.As shown in Figure 4h,i and Figure S15b (Supporting Information), the OER charge transfer resistances of RuMo-Air at the electrode potential of 1.5 V (vs.RHE) are 0.70 and 1.43 Ω in acidic and alkaline electrolytes.In contrast, Ir/C shows significantly higher charge transfer resistances of 28.50 and 132.00 Ω at the same electrode potential, in the acidic and alkaline electrolytes, respectively.The smaller resistances presented by RuMo-Air indicate a more favorable charge transfer at the catalyst/electrolyte interface compared to Ir/C.
In addition to the improved activity, RuMo-Air also exhibits enhanced stability compared to Ir/C under the same OER conditions (Figure S16c,d, and S17c,d, Supporting Information).Specifically, RuMo-Air can maintain its original activity in both acidic and alkaline electrolytes even after 10 000 ADT cycles.On the contrary, Ir/C shows significant performance degradation after only 4000 cycles.It has been reported that introducing other transition metals into Ru-based materials could improve the activity and stability of OER. [59,60]The foreign metal components can potentially modify the electronic structure of the catalyst, leading to optimized binding strength of oxygen intermediates and further enhanced performance.Additionally, the interaction between Ru and the other metal may prevent the oxidation of Ru to form the unstable phase of Ru (4 þδ) þ (δ > 0) during the OER process, resulting enhanced OER stability.We believe the improved OER performance observed for RuMo-Air may be attributed to the reasons mentioned above, where the presence of Mo can optimize the adsorption strength of the oxygenate intermediates on Ru active-sites, [54,61] and it can also prevent the Ru from undergoing excessive oxidation during OER.

Overall Water Splitting Based on the RuMo Composites
Due to their high activity and excellent stability in HER and OER, RuMo-Ar and RuMo-Air were then tested for the overall water splitting reaction, with the aim of exploring their potential for future practical applications.A two-electrode configuration electrolyzer was established (Figure 5a inset), with RuMo-Ar and RuMo-Air-coated carbon papers as cathode and anode, respectively.As shown in Figure 5a,c, the combination of RuMo-Ar and RuMo-Air can effectively drive water splitting, with improved performance compared to the Pt and Ir combination.Moreover, similar to the results observed in the half reactions, RuMo-Ar and RuMo-Air exhibit excellent stability in the full reaction, with no obvious decay in activity even after 10 000 cycles.Conversely, the combination of Pt/C and Ir/C experienced significant degradation after only 2,000 cycles under the same conditions.As illustrated in Figure 5b, during the continuous water splitting in acidic solution with a fixed current density of 10 mA cm À2 , the benchmark catalysts exhibited a decay rate of 3 mV h À1 while the RuMo-based catalysts maintained a lower cell voltage and exhibited a slightly smaller decay rate of 2 mV h À1 within 40 h.The RuMo-based catalysts remained stable without any noticeable change in cell voltage over time for alkaline water splitting, while the combination of Pt/C and Ir/C showed a decay rate of 2 mV h À1 .Overall, the Ru-Mo composites exhibit higher activity and better stability toward overall water splitting compared to the combination of Pt-based and Ir-based catalysts in both acidic and alkaline electrolytes.
To gain insight into the stability of the catalyst structure, we studied the morphological and electronic structures of RuMo-Ar and RuMo-Air after the water-splitting reactions (Figure S19, Supporting Information).Both the SEM and TEM images presented in Figure S20 and S21 (Supporting Information) indicate that the morphology of both samples remains largely unchanged, regardless of the electrolytes used.The corresponding EDS mapping of each sample shown in Figure S21 (Supporting Information) further confirms that the active components in the catalysts are well-preserved, and no obvious aggregations are observed after water splitting.Additionally, the chemical state of both Mo and Ru in RuMo-Ar and RuMo-Air was retained as well (Figure S22 and S23, Supporting Information).Overall, no significant changes were observed in both morphological and chemical states of the catalysts before and after the electrocatalysis, indicating that both the RuMo-Ar and RuMo-Air composites could maintain their chemical/physical integrities during HER and OER, respectively.

Catalyst Performance in the Counter Electrode Reactions
To understand the significance of appropriate chemical environments in electrocatalytic reactions, we tested the HER activity on RuMo-Air, and the OER activity on RuMo-Ar.As shown in Figure 6b,c, while there is a modest activity increase during the inducing period, RuMo-Air exhibits much lower activity compared to that of RuMo-Ar.A comparison of XPS results between RuMo-Ar and RuMo-Air under different experimental conditions reveals that the characteristic peaks for Mo and Ru in RuMo-Air shift to a lower binding energy during HER (Figure 6f,g).This shift indicates the occurrence of electrochemical reduction of the metal sites during HER, and the reduced components are more active moieties for HER.Unfortunately, the chemical state of Ru and Mo in RuMo-Air cannot be reduced to the same extent as in RuMo-Ar, at least under the HER conditions.This leads to the inferior HER performance of RuMo-Air compared to RuMo-Ar.Similarly, the OER activity of RuMo-Ar was also investigated using the same experimental method.As shown in Figure 6d,e, RuMo-Ar exhibits poor performance in OER, where the catalyst underwent significant oxidation and dissolution as the potential increased.Specifically, a strong oxidation peak appears before the onset of OER current, and there is no significant current observed after only three LSV cycles, suggesting the substantial loss of the active components from the catalyst.The XPS of the post OER RuMo-Ar (Figure 6h,i) confirms this statement as weakened or absence of Mo and Ru signals was observed.Taken together, it can be inferred that in water-splitting reaction, catalysts with low-valence state metals as the active components may be easily dissolved in the oxidation half reaction that operates at a higher potential window, leading to insufficient activity for the reaction.On the other hand, catalysts with high-valence state metal oxides as the main components may exhibit higher catalytic activity in an oxidation half reaction rather than a reduction half reaction.

Conclusion
In this work, two RuMo-based catalysts with distinct Ru-Mo chemical environments were developed by annealing the same RuMo precursor under different atmospheres.Comprehensive physical characterizations confirmed the differences in chemical environment between the two catalysts.One catalyst (RuMo-Ar) is in a chemically reduced state, while the other one (RuMo-Air) is in a chemically oxidized state.Electrochemical measurements indicated that RuMo-Ar and RuMo-Air exhibited enhanced HER and OER activity, respectively.Meanwhile, the two RuMo-based catalysts delivered high activity and good stability in the overall water-splitting reactions.Finally, performance of the catalysts in the counter electrode reaction was evaluated and inferior activity was observed, demonstrating the significance of an appropriate chemical environment in driving a catalytic reaction efficiently and stably.Overall, our work highlights the benefits of tailoring the chemical environment of bimetallic catalysts and underscores the impact of surface chemistry and chemical environment on catalytic processes.We believe that our strategy of tailoring the chemical environments within the catalysts can be applied to rationally synthesize high-performance and stable electrocatalysts.

Figure 1 .
Figure 1.a) Synthetic route of the RuMo-based electrocatalysts.SEM images of b) RuMo-Ar and e) RuMo-Air.TEM images of c,d) RuMo-Ar and f,g) RuMo-Air.

Figure 2 .
Figure 2. a) XPS survey spectra of RuMo-Ar and RuMo-Air.High-resolution XPS spectra of b) Mo 3d and c) Ru 3p for RuMo-Ar and RuMo-Air.d) XRD patterns of RuMo precursor, RuMo-Ar, and RuMo-Air, the bottom drop lines are the standard diffraction spectra of MoO 3 , carbon, and Mo 2 C. e) Illustration of the two different chemical environments.
Figure 3. a) LSV polarization curves with 95% iR correction in 0.5 M H 2 SO 4 .b,c) Tafel slopes of RuMo-Ar and Pt/C with different catalyst loadings in 0.5 M H 2 SO 4 .d) LSV polarization curves with 95% iR correction in 1 M KOH.e,f ) Tafel slopes of RuMo-Ar and Pt/C with different catalyst loadings in 1 M KOH.g) Overpotential of different catalysts at 10 mA cm À2 ("∞" means that the current density could not reach 10 mA cm À2 within the measured potential window).h) Nyquist plots for RuMo-Ar and Pt/C at À0.03 V (vs.RHE) in 0.5 M H 2 SO 4 and i) in 1 M KOH, the insets show the corresponding equivalent circuits.

Figure 4 .
Figure 4. a) LSV polarization curves with 95% iR correction in 0.5 M H 2 SO 4 .b,c) Tafel slope plots of RuMo-Air and Ir/C with different catalyst loadings in 0.5 M H 2 SO 4 .d) LSV polarization curves with 95% iR correction in 1 M KOH.e,f ) Tafel slopes of RuMo-Air and Ir/C with different catalyst loadings in 1 M KOH.g) Overpotential of different catalysts at 10 mA cm À2 ("∞" means that the current density could not reach 10 mA cm À2 within the measured potential window).Nyquist plots for RuMo-Air and Ir/C at 1.5 V (vs.RHE) in h) 0.5 M H 2 SO 4 and i) 1 M KOH; The insets show the equivalent circuits used for fitting.

Figure 5 .
Figure 5. a) LSV polarization curves tested on carbon paper in 0.5 M H 2 SO 4 ; Inset: the two-electrode system for overall water-splitting performance measurement.b) Long-term chronopotentiometry tests at 10 mA cm À2 in 0.5 M H 2 SO 4 .c) LSV polarization curves tested on carbon paper in 1 M KOH.d) Long-term chronopotentiometry tests at 10 mA cm À2 in 1 M KOH.

Figure 6 .
Figure 6.a) Illustration of RuMo-Air and RuMo-Ar used in the counter electrode reaction.LSV polarization curves comparison which use RuMo-Air as the HER catalyst in b) 0.5 M H 2 SO 4 and c) 1 M KOH.LSV polarization curves comparison which use RuMo-Ar as the OER catalyst in d) 0.5 M H 2 SO 4 and e) 1 M KOH.Comparison of XPS spectra of samples under different experimental conditions.f,h), High-resolution spectra of Mo 3d.g,i) High-resolution spectra of Ru 3p.