Due to the high demand for clean and renewable energy sources, electrochemical water spitting for (H₂) production is one of the most attractive and challenging tasks [1–3]. Furthermore, water electrolysis is one of the most efficient procedures for storing huge volumes of intermittently generated renewable electricity in the form of hydrogen gas [4–7]. However, an efficient electrocatalyst is necessary for large-scale applications owing to the slow reaction kinetics and complex electron transfer process [8–10]. To date, researchers have reported a number of catalysts for water-splitting reactions. For example, iridium (Ir) based electrocatalysts showed remarkable activity and stability toward oxygen evolution reaction (OER) [11–13], and platinum (Pt) based electrocatalysts are considered the best catalyst for hydrogen evolution reaction (HER) [14–16]. Although non-precious metal catalysts reported by some research teams have also played an important role, their activity or stability is still far behind that of noble metal-based catalysts in practical applications [17–19].

In practice, for proton exchange membrane (PEM) water electrolysis, conventional catalysts are based on iridium oxides because of the severe reaction environment on the OER side (low pH, high potential, and high O₂ concentration) [20,21]. Nonetheless, because Ir is one of the rarest elements on earth, the anode is not only the most inefficient component but also the primary cause of the electrolyzer's high cost [22,23]. The scarcity of Ir makes it difficult to meet the 2025 goal of the US Department of Energy (DOE) (H₂ production < US ＄ 2/kg) [12,24]. Thus, designing advanced electrocatalysts with low or no Ir content is imperative [25]. Ruthenium (Ru) has been proven to possess suitable adsorption energies for H_ads and O_ads, which gives an opportunity to develop highly efficient HER and OER catalysts [26–28], and Ru with a lower price, higher earth abundance was regarded as a substitute for Ir [29]. Recently, Ru nanoparticles and mixed Ru oxide, such as SrRuO₃, Cr_0.6Ru_0.4O₂, Co-RuO₂, and Ni/Co-doped RuO₂ typically present good OER or HER activity, but poor solubility resistance of these Ru-based electrocatalysts under harsh anodic and acidic conditions, making it unrealistic for practical utilization due to the limited lifespan [30–33]. Therefore, the development of a highly active and stable bifunctional RuO₂-based electrocatalyst has been a challenge for overall water-splitting. In order to resolve these problems, it was an effective strategy that the incorporation of other metal atoms into the RuO₂ could optimize the electronic structure to improve their activity and stability.

Based on the above-described results, herein, we developed high crystallinity RuO₂ catalysts with the incorporation of Mn, for use as a bifunctional electrocatalyst for OER and HER in acid and alkaline, presenting a high electrocatalytic activity and long-term stability. The strong electronic interaction between the incorporation of Mn and RuO₂ endows them with high OER and HER activity and stability. In an acid electrolyte, the overpotentials of OER and HER are only 200 mV and 20 mV at a current density of 10 mA/cm². Furthermore, the two-electrode PEM electrolyzer assembled by the Mn-RuO₂ requires a cell voltage as low as 1.50 V to achieve 10 mA/cm² for acid overall water-splitting as well as the long-term durability of 50 h at 50 mA/cm², which is superior to the state-of-the-art Ru-based catalyst. In addition, the Mn-RuO₂ exhibits low overpotentials of 37 and 220 mV for HER and OER in an alkaline electrolyte, the assembled alkaline electrolyzer similarly shows low cell voltages of 1.49 V at 10 mA/cm². These findings provide new insight into the development of highly efficient and stable catalysts for overall water-splitting in both alkaline and acidic electrolytes.

The synthesis process of Mn-RuO₂ is shown in Fig. 1a. Firstly, carbon-supported RuMn nanoparticles (RuMn/C) were obtained by thermal reduction of metal precursors adopted on a carbon black support at high temperatures under an H₂/Ar atmosphere (Fig. S1 in Supporting information). Carbon support during this process can effectively prevent nanoparticle agglomeration [34,35]. Subsequently, RuMn/C was ensured to fully oxidize to RuMnO_x, and to remove the carbon support. Finally, the unstable Mn species were removed by acid treatment to obtain the final catalysts Mn-RuO₂. The experimental details are provided in Supporting Information. As shown in Fig. 1b, X-ray diffraction (XRD) investigated the crystalline of these samples. The XRD pattern of the Mn-RuO₂ is well constant with the RuO₂ (JCPDS No. 40–1290) with good crystallinity [36], suggesting that the incorporation of Mn did not affect the RuO₂ lattice structure. Inductively Coupled Plasma Mass Spectrometry (ICP-MS) indicated the content of Mn and Ru in Mn-RuO₂ are 4 and 71.2 wt%. The morphological and structural details of RuMn/C, Mn-RuO₂, and RuO₂ were performed by transmission electron microscopy (TEM). The HADDF-STEM image of RuMn/C (Fig. 1c) shows its ultra-small size and the average of nanoparticles is 2.7 nm. Mn-RuO₂ nanoparticles were uniform in size, with an average diameter of about 6 nm, as shown in Fig. 1d, attributed to the alloy precursor's homogeneous particle size. Besides, the average size of RuO₂ nanoparticles is near 6 nm (Fig. S2 in Supporting information), indicating that the introduction of Mn essentially does not affect the size of RuO₂. Furthermore, the high-resolution TEM (HR-TEM) images (Fig. 1e) revealed that Mn-RuO₂ nanoparticles exhibited high crystallinity and distinct lattices. For the HR-TEM image of Mn-RuO₂, we observed crystalline structure with lattice fringe spacing of 0.317 nm and 0.252 nm, which can be ascribed to the (110) and (101) lattice planes of RuO₂. From the HR-TEM image of as-prepared RuO₂ (Fig. 1f), we also observed its high crystallinity and distinct lattices like Mn-RuO₂, and the lattice fringe spacing of the same lattice planes did not change. Energy-dispersive spectroscopic (EDS) elemental mapping (Fig. 1g) showed the homogeneous distribution of Mn, Ru, and O elements across the nanoparticles. These results confirmed Mn atoms were successfully incorporated into RuO₂.

To understand the electronic structure of Mn-RuO₂, the surface compositions and chemical states were investigated by X-ray photoelectron spectroscopy (XPS) measurements. The full-scan XPS survey spectrum verifies the expected elements of Mn, Ru, and O in Mn-RuO₂ nanoparticles (Fig. S3 in Supporting information), and revealed an Mn atomic ratio of 3.22 at% (Mn: Ru ratio ~1:10) in Mn-RuO₂. In the high-resolution Mn 2p XPS spectrum (Fig. 2a), two peaks at 652.0 eV and 641.3 eV were assigned to Mn 2p_1/2 and Mn 2p_3/2 of Mn³⁺, respectively [37,38]. Owing to Ru 3d_3/2 regions inherently overlapping with the C 1s spectra (Fig. S4 in Supporting information), the Ru 3p spectrum was examined for the oxidation state of Ru. For the Ru 3p XPS spectrum (Fig. 2b), two sets of doublet peaks located at 462.9 eV and 484.6 eV with two satellite peaks, can be attributed to Ru⁴⁺ [39–42]. The binding energy of Ru 3p_3/2 for Mn-RuO₂ had a positive shift (0.3 eV and 0.4 eV) in contrast to as-prepared RuO₂ and commercial RuO₂ (Com-RuO₂), indicating exits the strong electronic interaction between the incorporation of Mn and RuO₂. In addition, the peaks of Ru⁴⁺ 3p_3/2 with higher binding energy possibly suggested Ru ^{> 4+} species exiting [30]. Based on previous studies, Ru species with a high valance can enhance the OER activity [43]. In the O 1s spectrum of Mn-RuO₂ (Fig. 2c), the peaks the peaks located at about 529.2 eV, 530.5 eV, and 532.3 eV were ascribed to M-O (lattice oxygen), OH⁻/oxygen vacancy, and absorbed H₂O [44,45]. For Mn-RuO₂, the three peaks display a positive shift to higher binding energy, indicating that O grabs fewer electrons from the metal elements with doping Mn atoms. Furthermore, we employed ultraviolet photoemission spectroscopy (UPS) to investigate the energy level differences between Mn-RuO₂ and RuO₂ (Fig. 2d). Since the secondary electron cut-off edges for Mn-RuO₂ and RuO₂ are 17.11 eV and 16.75 eV, the values of work function for Mn-RuO₂ and RuO₂ are 4.11 eV and 4.47 eV. From a fundamental point of view, the materials' work function is an effective parameter that characterizes their electron donation ability. Mn-RuO₂ with the lower work function suggested the smaller energy barrier of electron transfer from the surface of electrocatalysts to reactants and intermediates, accelerating the HER and OER processes [46–48].

The OER performance of Mn-RuO₂ and contrast catalysts are investigated by a three-electrode system in 0.5 mol/L H₂SO₄. As shown in Fig. 3a, linear sweep voltammetry (LSV) curves showed that as-prepared RuO₂ exhibits better OER activity compared to Com-RuO₂, and delivers low overpotentials of 210 mV to achieve 10 mA/cm². For Mn-RuO₂, the incorporation of Mn showed an obvious enhancement in the OER activity of as-prepared RuO₂ and the overpotentials of 200 mV to achieve 10 mA/cm² (Fig. 3b). Besides RuO₂ and Com-RuO₂, Mn-RuO₂ also showed higher OER activity than its precursors RuMn/C (Fig. S5 in Supporting information) and MnRuO_x (Fig. S6 in Supporting information). The OER performance as a function of Ru amounts was calculated. Among them (Fig. 3c), Mn-RuO₂ delivered the highest mass activity of 956 A g/Ru, which is about 9.5-fold higher than that of Com-RuO₂ (103 A g/Ru), revealing the superior utilization of Ru in Mn-RuO₂. After adjusting the molar ratios of Mn and Ru (Fig. S7 in Supporting information) and oxidation temperature (Fig. S8 in Supporting information), the Mn-RuO₂ (Mn: 4 wt%, oxidation temperature: 500 ℃) catalysts displayed the highest OER catalytic activity. In addition, the Tafel slope of Mn-RuO₂ was as low as 56.1 mV/dec (Fig. 3d), which was lower than that of as-prepared RuO₂ (69.1 mV/dec) and Com-RuO₂ (99.2 mV/dec), implying a much more rapid kinetic process in Ru sites after regulating the electronic structures of RuO₂ by introducing Mn atoms. Similarly, electrochemical impedance spectroscopy (EIS) of Mn-RuO₂ showed the lowest charge transfer resistance in comparison with RuO₂ and Com-RuO₂ (Fig. 3e), indicating a higher reaction rate, and faster charge transfer [49]. To better understand the origin of the high OER performance of Mn-RuO₂, the double-layer capacitance (C_dl) and electrochemically active surface area (ECSA) were also calculated by cyclic voltammetry (CV) measurements (Fig. 3f). The Mn-RuO₂ catalysts had the highest C_dl and ECSA (Fig. S9 in Supporting information), which was 2-fold higher than RuO₂ and 10.7-fold higher than Com-RuO₂, suggesting a more electrochemical active surface area to expose abundant active sites with the incorporation of Mn. The activity and kinetics of Mn-RuO₂ were compared with those of recently reported Ru-based catalysts, suggesting that Mn-RuO₂ possesses better activity and faster OER kinetics than most of the reported Ru-based catalysts (Fig. 3g and Table S1 in Supporting information). Furthermore, Fe, Co, and Ni doping also had an identical effect that can enhance the activity of RuO₂ (Fig. S10 in Supporting information). In addition to the high activity, stability is a critical indicator for evaluating the OER performance in an acid medium. We investigated OER durability at 10 mA/cm² in 0.5 mol/L H₂SO₄ (Fig. 3h). The stability of current state-of-the-art noble RuO₂ catalysts is still a major concern because they quickly form acidic soluble RuO₄ species with higher oxidation states in highly oxidative and acidic environments [50,51]. Indeed, our result confirmed that the Com-RuO₂ catalyst suffers from fast performance decay with quickly elevated potentials within a few hours. Impressively, the catalytic stability of Mn-RuO₂ is far greater than that of commercial RuO₂. As a result, Mn-RuO₂ could continuously catalyze the OER for 100 h and only decay 30 mV. Furthermore, we systematically characterized post-catalysis Mn-RuO₂. HR-TEM (Fig. S11 in Supporting information) and XRD (Fig. S12 in Supporting information) results after the durability show structures of catalysts were retained well and no distinct structure reconstruction. The above results indicate the addition of Mn atoms can effectively improve the activity and stability of RuO₂ in acid electrolytes.

To further confirm the different catalytic performances, we explore the OER mechanism on various catalysts by performing DFT calculations. Our TEM observations indicated that the (110) facet was the most exposed facet for RuO₂ and Mn-RuO₂. Therefore, the slab (110) surface models of Mn-RuO₂ and RuO₂ were constructed to simulate the catalysts obtained in the experiment (Fig. S13 in Supporting information). Fig. 4a showed the Ru-O bond length of Mn-RuO₂ and RuO₂, indicating the doping of Mn evoked longer Ru-O bond length in favor of OER activity. Meanwhile, the free energy of OER with consecutive coupled proton-electron transfer steps was calculated to evaluate the OER performance for RuO₂ and doped Mn-RuO₂ with the adsorbate evolution mechanism (AEM, Fig. 4b). The formation of *OOH was found to be the rate-determining step (RDS) for the two catalysts, and the Mn-RuO₂ exhibited a lower free energy barrier compared with the RuO₂, which showed a smaller overpotential by 440 mV (Fig. 4c). DFT result indicates that the doping of Mn into RuO₂ can enhance the OER activity in acid media.

In general, RuO₂ has lower HER activity under acidic conditions. However, besides the outstanding OER catalytic performance, Mn-RuO₂ can also exhibit superior HER performances in acid media. As shown in Fig. 5a, although the activity is not as good as commercial Pt/C (Com-Pt/C), Mn-RuO₂ displays excellent HER performance with a lower overpotential of 20 mV at 10 mA/cm², much smaller than that of RuO₂ (48 mV) and Com-RuO₂ (167 mV, Fig. 5b). Moreover, the Tafel slope is only 25.5 mV/dec for Mn-RuO₂ much lower than that of as-prepared RuO₂ and Com-RuO₂ (Fig. 5c), and it has the lowest charge transfer resistance (Fig. S14 in Supporting information), indicating the superior reaction kinetics process of Mn-RuO₂ for HER. Encouraged by the superior OER and HER activity of Mn-RuO₂, a two-electrode PEM configuration was constructed by utilizing the bifunctional Mn-RuO₂ as both the cathode and the anode in acidic media (Mn-RuO₂//Mn-RuO₂). For comparison, the electrolyzers are based on commercial 20 wt% Pt/C and commercial RuO₂ (Pt/C//Com-RuO₂), commercial 20 wt% Pt/C and as-prepared RuO₂ (Pt/C//RuO₂) as references. Notably, the Mn-RuO₂//Mn-RuO₂ displayed excellent activity toward overall water-splitting in acid electrolytes and the electrolyzer's potentials were only required 1.50 V and 1.56 V at 10 and 50 mA/cm², respectively (Fig. 5d). However, to attach such current density, the cell potentials were 1.54 V and 1.59 V for Pt/C//RuO₂, 1.63 V and 1.71 V for Pt/C//Com-RuO₂, respectively. Additionally, the long-term stability of Mn-RuO₂//Mn-RuO₂ for overall water-splitting was explored by chronopotentiometry tests at 50 mA/cm² (Fig. 5e). The Mn-RuO₂//Mn-RuO₂ electrolyzer exhibited superior durability and can operate stably for 50 h with only a decay of 30 mV. In particular, we measured the Faradaic efficiency of Mn-RuO₂//Mn-RuO₂ electrolyzer for H₂ and O₂ evolution (Fig. S15 in Supporting information) in acid, the measured amount of O₂ and H₂ varies with time at a given current (50 mA); the molar ratio of H₂ and O₂ is close to 2:1. As a result, the final Faradaic efficiency of cathode and anode is both around 99%. The activity and stability of Mn-RuO₂//Mn-RuO₂ were better than other reported bifunctional electrocatalysts for overall water-splitting in acid (Table S2 in Supporting information). These results indicate the high potential of the Mn-RuO₂ catalyst for acid overall water electrolysis.

Based on the high HER and OER activity of Mn-RuO₂ in acid, we further evaluated Mn-RuO₂ electrocatalytic performance toward HER, OER, and overall water-splitting in alkaline electrolytes. As shown in Fig. 6a, the Mn-RuO₂ showed excellent OER activity in 1.0 mol/L KOH with an overpotential of 220 mV to deliver 10 mA/cm², in contrast to the RuO₂ (252 mV) and Com-RuO₂ (274 mV, Fig. 6b). Additionally, Mn-RuO₂ possesses the fastest electrocatalytic kinetics with the smallest Tafel slope of 59.7 mV/dec (Fig. 6c) and charge-transfer resistances (Fig. S16 in Supporting information). Meanwhile, as shown in Fig. 6d, the Mn-RuO₂ exhibited a much lower HER overpotential of 37 mV at 10 mA/cm², which was superior to that of the RuO₂ (69 mV) and Com-RuO₂ (101 mV, Fig. 6e). Furthermore, the kinetic process was revealed by the Tafel slope and EIS. The Mn-RuO₂ showed the lowest Tafel slope (40.8 mV/dec, Fig. 6f) and the smallest charge-transfer resistances (Fig. S17 in Supporting information) compared to RuO₂ and Com-RuO₂. To better understand the origin of the high OER and HER performance of Mn-RuO₂, we employed CV measurements calculated to the C_dl and ECSA (Fig. S18 in Supporting information), Mn-RuO₂ exhibited a considerably high C_dl value of 43.2 mF/cm², which corresponds to an ECSA value of 1234.2 cm² (Fig. S19 in Supporting information), was 3-fold higher than RuO₂ and 7-fold higher than Com-RuO₂, indicating Mn-RuO₂ also exposed a more electrochemical active surface area in alkaline electrolytes compared to other catalysts. Because of the significant HER and OER activity in 1.0 mol/L KOH, the overall water-splitting activity of the Mn-RuO₂ catalyst was evaluated with a two-electrode configuration. As shown in Fig. 6g, the Mn-RuO₂//Mn-RuO₂ showed overall water-splitting superior activity, only needing 1.48 V and 1.56 V at 10 and 50 mA/cm², respectively, which was much lower than the references Pt/C//Com-RuO₂ (1.58 V and 1.75 V) and Pt/C//RuO₂ (1.54 V and 1.61 V) at the same current densities. Impressively, the overall water-splitting activity of Mn-RuO₂//Mn-RuO₂ was higher than most of the recently reported electrocatalysts for overall water-splitting in alkaline (Table S3 in Supporting information). We measured the Faradaic efficiency of Mn-RuO₂//Mn-RuO₂ electrolyzer for H₂ and O₂ evolution (Fig. S20 in Supporting information) in alkaline, the molar ratio of H₂ and O₂ is close to 2:1. And the final Faradaic efficiency is both around 99%. Correspondingly, Mn-RuO₂//Mn-RuO₂ can deliver superior durability even with the reaction time prolonging to 50 h at 50 mA/cm² with only decay of 10 mV (Fig. 6h). The above results indicate that Mn-RuO₂ can be one of the most promising Ru-based catalysts for practical alkaline water electrolysis.

In summary, Mn-RuO₂ nanocrystals are successfully synthesized as bifunctional electrocatalysts for overall water-splitting in acid and alkaline. The as-prepared Mn-RuO₂ catalysts show excellent OER activity in acid electrolytes, resulting in an overpotential of 200 mV at 10 mA/cm², which outperforms most Ru-based catalysts. DFT result indicates that the doping of Mn into RuO₂ can enhance the OER activity. Meanwhile, the Mn-RuO₂ also possesses outstanding HER activity, leading to an overpotential of 20 mV at 10 mA/cm². Furthermore, the two-electrode PEM electrolyzer assembled by the Mn-RuO₂, yielding 10 mA/cm² at 1.5 V and the long-term durability of 50 h at 50 mA/cm², which is superior to the state-of-the-art Ru-based catalyst. In addition, the Mn-RuO₂ also exhibits high activity for HER and OER in alkaline media, and the assembled alkaline electrolyzer by Mn-RuO₂ shows low cell voltages of 1.49 V at 10 mA/cm². Such remarkable performance of the Mn-RuO₂ was attributed to the incorporation of Mn modulating the electronic structure of RuO₂, exposing more electrochemically active areas, and accelerating HER and OER processes. This work provides an effective strategy for designing highly active and stable Ru oxides-based electrocatalysts toward OER and HER in acid and alkaline.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Jing Cao: Conceptualization, Data curation, Formal analysis, Investigation, Software, Supervision, Validation, Writing – original draft, Writing – review & editing. Dezheng Zhang: Software. Bianqing Ren: Investigation. Ping Song: Investigation, Software, Writing – review & editing. Weilin Xu: Conceptualization, Funding acquisition, Writing – review & editing.

This work was supported by the Key Research and Development Program sponsored by the Ministry of Science and Technology (MOST, Nos. 2022YFB4002000, 2022YFA1203400) and the National Natural Science Foundation of China (Nos. 21925205, 22072145, 22372155, 22005294 and 22102172).

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109863.