English

• Tremendous efforts have been made to develop earth-abundant OER electrocatalysts, including transition metal alloys, oxides, and hydroxides, for alkaline water electrolysis. In particular, NiFe-based catalysts, mainly in the form of NiFe oxides and hydroxides, have been intensively investigated for their excellent performance in alkaline media [1–6]. Although nanosized NiFe-based materials exhibit high activity with low overpotential for the OER, the stability of nanostructured catalysts commonly suffers from rapid degradation under high current density operating conditions [7–10]. To further improve the activity and stability of NiFe-based OER catalysts, the strategy of third-element-doping has been developed and intensively studied [11]. The integration of high-valence-state transition metal elements, such as Co^2+~4+ and Cr^3+~6+, into NiFe (oxy)hydroxides has been proven to be able to enhance catalytic activity [5,8,12–20].

  Recently, transition metal alloys have attracted increasing attention for use in fabricating monolithic electrodes for the OER because of their easy preparation, simple treatment processes, and promising long-term stability [21–24]. An electrochemical activation step is commonly used to convert the alloy surface into the (oxy)hydroxide species, thus enhancing the OER performance [1,24–31]. Recently, medium-entropy alloys (MEAs) that contain three or four metallic elements have emerged as a new class of materials that exhibit improved mechanical strength and excellent corrosion resistance [32–36]. Fu et al. [35] designed an FCC single-phase nanocrystalline Co₂₅Ni₂₅Fe₂₅Al_7.5Cu_17.5. This FCC structured MEA exhibited a compressive yield strength of 1975 MPa with a hardness of 454 Hv, which was more robust than any face-centered cubic (FCC) alloy reported in the previous literature. Quiambao et al. [36] designed a Ni₃₃Cr₁₆Fe₁₇Ru₁₉Mo₉W₆ that forms a thick passive layer rapidly during potentiostatic processes, which showed a strong corrosion resistance in acid media. However, the application of MEAs for fabricating monolithic OER electrodes has not been well explored.

  Here, we demonstrate that MEAs (with the elements Fe, Co, Cr, and Ni) can be used for fabricating monolithic OER electrodes and exhibit excellent electrochemical performance in terms of catalytic activity and long-term stability. The MEAs underwent surface reconstruction during the electrochemical activation process. This surface reconstruction transformed the MEA surface into a rough layer containing medium-entropy (oxy)hydroxides. As a result, the electroactivated FeCoCrNi alloy exhibited a remarkably low overpotential of 237 mV at 10 mA/cm² while maintaining excellent stability for 2000 h of long-term testing at 1 A/cm².

  The MEA monolithic OER electrodes were fabricated by vacuum melting metal ingots several times under an argon atmosphere (Fig. 1a). The crystalline MEA structures were confirmed by XRD (Fig. 1b). For the alloys containing Ni (FeCrNi, FeCoCrNi, CoCrNi), three strong diffraction peaks located at 44.51°, 51.91° and 76.41° can be observed, which are indexed to the (111), (200) and (220) planes of face-centered cubic (FCC) Ni (JCPDS No. 04–0850) [24], respectively. We can observe that all the diffraction peaks in the CoCrNi MEA sample have shifted slightly to higher angles compared to the FeCrNi MEA sample. This indicates that the lattice constant and interplanar spacing have become larger when the Co atoms are replaced by Fe atoms, which have a larger atomic radius. Moreover, the slight broadening of the peak around 44° for all the MEA samples could be attributed to lattice distortion within the MEA structures. These observations collectively confirm the successful synthesis of the various MEA compositions investigated in this study. For the FeCoCr alloy, three strong diffraction peaks were located at 44.76°, 47.57°, and 75.94°. These peaks can be attributed to the (002), (101) and (110) planes of hexagonal Co (JCPDS No. 05–0727). Other weak peak diffractions can be attributed to Co₇Fe₃ (JCPDS No. 50–0795).

  Figure 1

  

  Figure 1.  (a) Schematic of MEAs as OER catalysts synthesized by vacuum melting. (b) XRD patterns of the initial MEAs.

  DownLoad: Full-Size Img PowerPoint

  SEM images (Fig. S2 in Supporting information) show that all the initial MEAs had relatively smooth surfaces before the electrochemical activation process. After 200 cycles of CV activation (0‒0.6 V vs. Hg/HgO), small holes with diameters of approximately 200 nm started to appear on the alloy surface, and irregular nanoparticles constituted the alloy surface (Fig. 2). The surface alterations are a result of dealloying, a process in which more easily dissolvable metals within alloys are selectively removed. Almost no original planes can be found in EA-FeCoCr or EA-CoCrNi. However, EA-FeCrNi and EA-FeCoCrNi have different corrosion traces. As shown in Figs. 2e and g, the original smooth planes of the surface partly remain in EA-FeCrNi and EA-FeCoCrNi. The differences in surface morphologies between the NiFe-based and non-NiFe-based MEAs imply varying levels of corrosion resistance among these alloys. Furthermore, energy dispersive X-ray spectroscopy (EDS) mapping images (Fig. S2 and Fig. 2) for both the original alloys and electrochemically active alloys clearly demonstrate that the elemental distribution is uniform throughout the measured region.

  Figure 2

  

  Figure 2.  SEM and corresponding SEM‒EDS mapping images of the electrochemically activated MEAs. (a, b) EA-FeCoCr. (c, d) EA-CoCrNi. (e, f) EA-FeCrNi. (g, h) EA-FeCoCrNi.

  DownLoad: Full-Size Img PowerPoint

  Linear sweep voltammetry (LSV) polarization curves and chronopotentiometric tests were performed for the MEAs and EA-MEAs to evaluate the OER performance of the MEAs (Figs. 3a-d). Electrochemical activation has a positive effect on MEAs, resulting in a decrease in overpotentials for EA-MEAs of 10–30 mV compared to their original value. Among all the EA-MEA electrodes, EA-FeCrNi exhibited the lowest overpotential of 286 mV. The overpotential for EA-FeCoCrNi was reduced to 301 mV at 10 mA/cm² (normalized by the geometrical area of the electrode) after activation. Due to the outstanding corrosion resistance of the quaternary medium-entropy alloy, EA-FeCoCrNi underwent limited reconstruction (Fig. 2g) and exhibited normal activity after CV activation. A chronopotentiometric test was run at a high current density of 1 A/cm², and all four EA-MEAs exhibited no significant decay for more than 1800 h (Fig. 3g), which is close to practical application. The oxygen bubbles production condition for Post-FeCoCrNi was recorded in Video S1 (Supporting information). The ions leaching condition during chronopotentiometric test was also monitored by ICP-AES (Table S7 in Supporting information). However, the transition metal ions will deposit with OH^-. The ICP-AES results may fail to reflect the true ions leaching conditions. The EA-MEAs were simplified as Post-FeCoCr, Post-CoCrNi, Post-FeCrNi and Post-FeCoCrNi after the stability test. The overpotentials of the Post-MEAs were further improved after the chronopotentiometric test. Post-FeCoCrNi exhibited the smallest overpotential of 237 mV after complete reconstruction (Fig. S3g in Supporting information). Fig. 3e shows the activities of all MEAs are improved by electro-activation and long-term chronopotentiometry test. And the LSV and CV curves without iR compensate are provided in supporting information (Figs. S9 and S10 in Supporting information).

  Figure 3

  

  Figure 3.  Electrochemical OER activity evaluations. (a-d) LSV polarization curves of various catalysts in 1 mol/L KOH solution at a scan rate of 10 mV/s. (e) Overpotentials of various catalysts at a current density of 10 mA/cm². (f) Tafel plots of various catalysts derived from the polarization curves in (a-d) LSV polarization curves of MEAs after 200 CV cycles. (g) Chronopotentiometric curves of various catalysts in 1 mol/L KOH at a current density of 1 A/cm² (without iR correction).

  DownLoad: Full-Size Img PowerPoint

  Fig. 3f shows the Tafel plots of the EA-MEAs, which were used to probe the reaction kinetics. The Tafel slope of EA-FeCoCrNi was determined to be 39.9 mV/dec, which was observed to decrease to 37.9 mV/dec for EA-FeCrNi. This decrease in Tafel slopes indicates that there was an acceleration in the reaction kinetics, and there could have been a change in the rate-limiting step (RLS).

  To compare the relative electrochemically active surface areas (ECSAs) of the different EA-MEAs, we also estimated the double-layer capacitance (C_dl) of the four electrochemically activated samples by measuring the charging current in a potential window free of Faradaic processes. The plots of Δj = (j_a − j_c)/2 at 0.82 V (vs. RHE) against the scan rate (Fig. S4 in Supporting information) show that the samples have slightly different slopes and thus C_dl and ECSA values (Table S2 in Supporting information). The EA-FeCrNi sample has the lowest C_dl among all the EA-MEAs, indicating that it has the smallest electrochemically active surface area. However, EA-FeCrNi still presented the highest current density after ECSA normalization (Figs. 4f-h), which further confirmed that its intrinsic activity was the highest among all the EA-MEAs. And EA-FeCrNi also shows the smallest R_ct (Fig 4i). However, EA-FeCoCrNi suffers relatively small reconstructions due to the tough nature of its medium-entropy alloy materials, which leads to poor activity after CV activation. And after the chronopotentiometric test, Post-FeCoCrNi gets largest ECSA value and highest current density after ECSA normalization (Figs. S8b-d in Supporting information) among all Post-MEAs. Post-FeCoCrNi also shows the smallest R_ct (Fig. S8a).

  Figure 4

  

  Figure 4.  Electrocatalytic properties of the electrodes for the OER. (a-d) pH-dependent LSV polarization curves of various catalysts. (a) EA-FeCoCr, (b) EA-CoCrNi, (c) EA-FeCrNi and (d) EA-FeCoCrNi. (e) Specific OER activity at 1.55 V vs. RHE after iR correction as a function of pH. (f) Capacitive current density differences (Δj = (j_a − j_c)/2 at 0.82 V versus RHE as a function of scan rate. The linear slope is equivalent to that of C_dl. (g) The corresponding LSV polarization curves normalized to the specific activities of various catalysts by ECSAs. (h) Current density values at an overpotential of 300 mV. (i) Nyquist plots of various catalysts at 1.52 V versus RHE in 1 mol/L KOH solution.

  DownLoad: Full-Size Img PowerPoint

  To explore how the differences in electronic structures among MEAs could impact OER kinetics and mechanisms, we measured OER activity in KOH at various pH values (from 13 to 14) via LSV polarization curves (Figs. 4a-d). At higher pH levels, the OER current for EA-FeCoCr increased at the same potential on the RHE scale. However, EA-CoCrNi, EA-FeCrNi and EA-FeCoCrNi showed only mild increases in the OER current at the same potential on the RHE scale (Fig. 4e). This suggests a mismatch in electron transfer kinetics and hydroxide affinity for different alloys. The Ni-based alloys exhibited a much weaker correlation between pH and OER activity, while the EA-FeCoCr alloy exhibited a stronger correlation between pH and OER activity. The strong pH dependence of the OER activity is related to nonconcerted proton−electron transfer processes with lattice oxygen as the active site [37,38]. The CV curves are tested to show the OER performance of Post-MEAs at different temperature (Fig. S11 in Supporting information). All Post-MEA electrodes performances are improved under higher temperature. Besides the temperature, the Fe³⁺ ions in solution can also affect the OER performance of Ni-based and Co-based materials. The Fe³⁺ ions influence is also discussed in supporting information (Fig. S12 and Table S6 in Supporting information).

  The results of the activity evaluations indicate a significant improvement in the MEAs after the stability test (Figs. 3a-d). Further characterizations have been applied to determine the evolution of the interatomic electronic structure of multimetal species and reconstruction of catalyst surfaces.

  The XRD patterns for the Post-MEAs, as shown in Fig. 5a, show no significant changes in peak positions, and the crystal structures of all the alloys in the bulk phase remain unchanged. SEM images (Fig. S3 in Supporting information) demonstrated that the Post-MEA surfaces underwent deep reconstruction, and no original smooth planes were found. Additionally, EDS mapping images (Fig. S3) reveal a uniform distribution of all the elements throughout the region measured Post-MEA. No obvious segregation or decrease in certain metal elements can be recognized from the EDS mapping images.

  Figure 5

  

  Figure 5.  Physicochemical property characterization. (a) XRD patterns of the electrochemically activated MEAs. (b) Raman spectra of the electrochemically activated MEAs.

  DownLoad: Full-Size Img PowerPoint

  Raman spectroscopy provides more sensitive insights into the surface evolution of MEAs. Raman experiments (Fig. 5b) revealed Raman peaks that corresponded to δ(Ni^Ⅲ-O) (470 cm⁻¹) and ν(Ni^Ⅲ-O) (550 cm⁻¹) after Ni-based Post-MEAs underwent a stability test, resulting in the formation of NiOOH, a mixture of γ-NiOOH and β-NiOOH [39–45]. The intensity ratio of the doublet peaks at 470 cm⁻¹ and 550 cm⁻¹ was further used to quantify the active β-NiOOH phase. A stronger peak at 550 cm^-1 indicates a higher content of β-NiOOH, which is a more active intermediate for the OER. Two obvious Raman bands at 312 cm⁻¹ and 677 cm⁻¹ caused by surface disordered Fe^Ⅲ-OOH species [1,40] were detected only for Post-FeCrNi. We speculated that the Raman peaks of FeOOH in the reconstructed layers became invisible due to overlap with the strong bands of NiOOH and the dissolution of Fe during the OER. A broad feature previously assigned to ν(O—O) of the active oxygen species Ni-OO⁻ was also observed in the region of 850–1200 cm^{−1 [39,]} for all the samples containing Ni. Post-FeCoCr exhibited two broad Raman peaks at approximately 560 cm⁻¹ and 485 cm⁻¹. Co₃O₄ exhibited multiple peaks in the region of 485−691 cm⁻¹ [46], and two new peaks at 517 and 560 cm⁻¹ appeared after oxidation; these peaks were assigned to CoOOH [46,47]. According to Raman spectroscopy, the Post-FeCoCr surface mainly contained Co^Ⅲ-OOH, and Co₃O₄ partly oxidized by Raman laser. The main active species after the reconstruction are NiOOH for the Ni-based MEAs and CoOOH for the non-Ni-based MEA. The CoOOH on the Post-FeCoCr MEA follows the proton-coupled electron-transfer route [48] and shows higher pH dependence, while the Ni-based MEAs exhibit lower pH dependence and follow a different path, which provides higher activity. The key difference lies in the band gap between the M-3d and O-2p orbitals. A larger band gap hinders electron movement between the two bands and the triggering of the LOM pathway [37,38,48].

  To further confirm the reconstruction layer structure, we tested the powder dropping after ultrasonic oscillation by high resolution transmission electron microscopy (HRTEM) from Post-FeCoCrNi. HRTEM image and corresponding selected area electron diffraction (SEAD) image (Fig. 6a) showed the reconstructed specie was amorphous structure. And Cr element leached from active specie layer seriously according to HRTEM-STEM related EDS mapping image (Fig. 6b). Considering the improved activity and reconstructed structure, we supposed that amorphization and self-evolved element ratio optimization during reconstruction are responsible for the higher performance.

  Figure 6

  

  Figure 6.  (a) HRTEM image for Post-FeCoCrNi nanoparticle dropping after ultrasonic oscillation and corresponding SEAD image. (b) HAADF-STEM image, corresponding EDS element mapping showing the distribution of Ni, Fe, Co, Cr and O. (c) XPS spectra for all elements (Fe 2p, Co 2p, Cr 2p, Ni 2p and O 1s) in the FeCoCrNi alloy and peak fitting analysis of the MEAs before and after the chronopotentiometric test.

  DownLoad: Full-Size Img PowerPoint

  In addition, the chemical composition and oxidation state of the MEAs were characterized using X-ray photoelectron spectroscopy (XPS) before and after the stability test. Prior to the test, the original MEAs exhibited metallic features with inevitable surface oxidation in air. The fine-scan Ni 2p spectrum from the original Ni-based alloy surface (Fig. 6c, Figs. S6 and S7 in Supporting information), consisting of two spin–orbital doublets of Ni 2p_3/2 (855.5 eV), Ni 2p_1/2 (872.9 eV) and two satellites (abbreviated as ''sat.''), indicated that Ni²⁺ was present [1,40], while the minor peaks at 852.7 and 870.2 eV were assigned to metallic Ni. Similarly, the deconvoluted peaks located at 710.6 and 723.3 eV and two satellites in the Fe 2p profile can be attributed to Fe³⁺ [40]. The peaks located at 706.6 and 719.8 eV are assigned to metallic Fe [1]. For Co-containing alloys (Fig. 6, Figs. S5 and S6 in Supporting information), the Co 2p_3/2 peak can be fitted to the Co³⁺ and Co²⁺ components at binding energies of 780.8 and 781.4 eV, respectively, which are consistent with the Raman results. The peaks located at 778.2 and 793.2 eV are assigned to metallic Co [24]. For Cr-containing alloys (Fig. 6 and Figs. S5-S7), the deconvoluted peaks located at 576.7 and 586.4 eV in the Cr 2p profile can be attributed to Cr³⁺ [24,40], and the peaks located at 573.7 and 582.9 eV are assigned to metallic Cr. We also collected O 1s XPS spectra (Fig. 6 and Figs. S5-S7), which showed three characteristic peaks at 529.8, 531.3, and 532.6 eV, marked as M-O, M-OH, and absorbed H₂O, respectively [24].

  After the chronopotentiometric test, all the metal peak signals disappeared, and deconvoluted peaks attributed to Ni³⁺ (located at 856.6 and 874.3 eV) appeared, which was in line with the formation of NiOOH surface species detected by Raman spectroscopy. For Co-containing materials, the Co³⁺ species have stronger intensities in the Co 2p profile after the stability test. Moreover, the Fe and Co peak positions showed a slight positive shift after oxidation. The Cr³⁺ 2p profile can hardly be distinguished from the collected signal noise, which means that the Cr atoms leached from the surface of the alloys during the long-term OER test, which is consistent with the EDS mapping of HRTEM-STEM for Post-FeCoCrNi (Fig. 6b). Only in the XPS spectra of Post-CoCrNi (Fig. S6), a weak signal corresponding to Cr³⁺ could be detected. The composition and structure of alloy catalysts undergo dynamic evolution during OER process, with enhanced electrocatalytic performance observed after structural reconstruction. The reconstruction process is highly complex and involves the formation of amorphous structures, which makes it challenging to predict catalytic performance through theoretical models based on free energy calculations [49–51]. However, the reconfiguration outcome is closely associated with the precursors' structures and compositions. By employing a rational design and control of the metal ratios, the highly active monolithic MEA electrocatalysts are obtained.

  In summary, the in situ transformed NiFe-based oxyhydroxide nanoparticles on the surface increased the number of active species and the specific surface area of the NiFe-based MEAs. The leaching of unstable elements (e.g., Cr) facilitates the formation of a stable interface. Therefore, both activity and stability can be improved during the self-evolution process. As a result, the FeCoCrNi MEA electrode, which benefits from the favorable formation of NiFe oxyhydroxide species and reconstructed stable surface, offers a low overpotential (237 mV) and excellent stability (over 2000 h@1 A/cm²) for the OER. This represents a significant improvement over the state-of-the-art IrO₂ catalyst in 1 mol/L KOH. We demonstrated that monolithic medium-entropy alloy electrodes exhibit improved performance during OER testing. This strategy can be extended to a variety of alloy materials and reactions, allowing for the development of catalysts that are suitable for real industry applications under harsh working conditions.

  Declaration of competing interest

  The authors declare no conflicts of interest.

  CRediT authorship contribution statement

  Xiaoke Xi: Writing – original draft. Xinpeng Li: Writing – original draft. Yang Liu: Writing – review & editing. Yucheng Zhang: Writing – original draft. Linmei Li: Writing – review & editing. Jianming Li: Writing – original draft. Xu Jin: Writing – original draft. Shuhong Jiao: Writing – review & editing. Zhanwu Lei: Writing – review & editing. Ruiguo Cao: Writing – review & editing.

  Acknowledgments

  This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (No. XDB0450302), the National Natural Science Foundation of China (Nos. 52072358, 51902304, 22209162, U21A2082), the Fundamental Research Funds for the Central Universities (Nos. YD2060002043, WK2060000048), and the Hefei Municipal Natural Science Foundation (No. BJ2060000042). The authors appreciate the financial support from the R & D Department of PetroChina. Material characterizations in this work were carried out at the Instruments Center for Physical Science, University of Science and Technology of China.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110535 .

  
  
• 1. [1]

     C.W. Liang, P.C. Zou, A. Nairan, et al., Energy Environ. Sci. 13 (2020) 86–95. doi: 10.1039/c9ee02388g

  2. [2]

     K.Y. Zhu, X.F. Zhu, W.S. Yang, Angew. Chem. Int. Ed. 58 (2019) 1252–1265. doi: 10.1002/anie.201802923

  3. [3]

     S. Jo, W.B. Park, K.B. Lee, et al., Appl. Catal. B 317 (2022) 121685. doi: 10.1016/j.apcatb.2022.121685

  4. [4]

     S.X. Liu, H.W. Zhang, E.L. Hu, et al., J. Mater. Chem. A 9 (2021) 23697–23702. doi: 10.1039/d1ta06263h

  5. [5]

     M.H. Wang, Z.X. Lou, X. Wu, et al., Small 18 (2022) 2200303. doi: 10.1002/smll.202200303

  6. [6]

     C.G. Kuai, Z.R. Xu, C. Xi, et al., Nat. Catal. 3 (2020) 743–753. doi: 10.1038/s41929-020-0496-z

  7. [7]

     S.W. Im, H. Ahn, E.S. Park, et al., ACS Appl. Energy Mater. 4 (2021) 595–601. doi: 10.1021/acsaem.0c02476

  8. [8]

     Y. Yang, L.N. Dang, M.J. Shearer, et al., Adv. Energy Mater. 8 (2018) 17031189.

  9. [9]

     H.X. Li, C.Y. Zhang, W.J. Xiang, et al., Chem. Eng. J. 452 (2023) 139104. doi: 10.1016/j.cej.2022.139104

  10. [10]

      W. Cheng, J.F. He, K.E. Dettelbach, et al., Chem 4 (2018) 821–832. doi: 10.1016/j.chempr.2017.12.030

  11. [11]

      B. Zhang, L. Wang, Z. Cao, et al., Nat. Catal. 3 (2020) 985–992. doi: 10.1038/s41929-020-00525-6

  12. [12]

      N.K. Dang, J.N. Tiwari, S. Sultan, et al., Chem. Eng. J. 404 (2021) 126513. doi: 10.1016/j.cej.2020.126513

  13. [13]

      R.N. Singh, J.P. Singh, B. Lal, et al., Electrochim. Acta 51 (2006) 5515–5523. doi: 10.1016/j.electacta.2006.02.028

  14. [14]

      C. Schwanke, H.S. Stein, L.F. Xi, et al., Sci. Rep. 7 (2017) 44192. doi: 10.1038/srep44192

  15. [15]

      B. Zhang, X.L. Zheng, O. Voznyy, et al., Science 352 (2016) 333–337. doi: 10.1126/science.aaf1525

  16. [16]

      X. Bo, Y.B. Li, X.J. Chen, et al., J. Power Sources 402 (2018) 381–387. doi: 10.1016/j.jpowsour.2018.09.063

  17. [17]

      T.H. Shen, L. Spillane, J. Vavra, et al., J. Am. Chem. Soc. 142 (2020) 15876–15883. doi: 10.1021/jacs.0c06268

  18. [18]

      L. Reith, J.N. Hausmann, S. Mebs, et al., Adv. Energy Mater. 13 (2023) 2203886. doi: 10.1002/aenm.202203886

  19. [19]

      J.X. Guo, Y. Zheng, Z.P. Hu, et al., Nat. Energy 8 (2023) 264–272.

  20. [20]

      J.Z. Huang, H.Y. Sheng, R.D. Ross, et al., Nat. Commun. 12 (2021) 3036. doi: 10.1038/s41467-021-23390-8

  21. [21]

      L. Magnier, G. Cossard, V. Martin, et al., Nat. Mater. 23 (2024) 252–261. doi: 10.1038/s41563-023-01744-5

  22. [22]

      A.A.H. Tajuddin, M. Wakisaka, T. Ohto, et al., Adv. Mater. 35 (2023) 2207466. doi: 10.1002/adma.202207466

  23. [23]

      L. Tao, M.Z. Sun, Y. Zhou, et al., J. Am. Chem. Soc. 144 (2022) 10582–10590. doi: 10.1021/jacs.2c03544

  24. [24]

      N. Zhang, X.B. Feng, D.W. Rao, et al., Nat. Commun. 11 (2020) 4066. doi: 10.1038/s41467-020-17934-7

  25. [25]

      J.T. Mefford, A.R. Akbashev, M. Kang, et al., Nature 593 (2021) 67–73. doi: 10.1038/s41586-021-03454-x

  26. [26]

      C.H. Zhao, N. Li, R.Z. Zhang, et al., ACS Appl. Mater. Interfaces 11 (2019) 47858–47867. doi: 10.1021/acsami.9b13834

  27. [27]

      L. Magnier, G. Cossard, V. Martin, et al., Nat. Mater. 23 (2024) 252–261. doi: 10.1038/s41563-023-01744-5

  28. [28]

      H.L. Jiang, Q. He, X. Li, X.Y. Su, et al., Adv. Mater. 31 (2019) 1805127. doi: 10.1002/adma.201805127

  29. [29]

      W.K. Hu, Q. Liu, T.X. Lv, et al., Electrochim. Acta 381 (2021) 138276. doi: 10.1016/j.electacta.2021.138276

  30. [30]

      R. Duan, Y.J. Li, S. Wang, et al., J. Phys. Chem. Lett. 13 (2022) 1446–1452. doi: 10.1021/acs.jpclett.1c04136

  31. [31]

      L.M. Wang, D.X. Liu, Z.F. Zhang, et al., J. Alloys Compd. 934 (2023) 167846. doi: 10.1016/j.jallcom.2022.167846

  32. [32]

      S. Moniri, Y. Yang, J. Ding, et al., Nature 624 (2023) 564–569. doi: 10.1038/s41586-023-06785-z

  33. [33]

      G.H. Han, M.G. Li, H. Liu, et al., Adv. Mater. 34 (2022) 2202943. doi: 10.1002/adma.202202943

  34. [34]

      S.S. Sohn, A.K.D. Silva, Y. Ikeda, et al., Adv. Mater. 31 (2019) 1807142. doi: 10.1002/adma.201807142

  35. [35]

      Z. Fu, W. Chen, H. Wen, et al., Acta Mater. 107 (2016) 59–71. doi: 10.3390/f7030059

  36. [36]

      K.F. Quiambao, S.J. McDonnell, D.K. Schreiber, et al., Acta Mater. 164 (2019) 362–376. doi: 10.1016/j.actamat.2018.10.026

  37. [37]

      L.J. Zhang, Y.M. Yang, M.A. Ziaee, et al., ACS Appl. Mater. Interfaces 10 (2018) 9460–9467. doi: 10.1021/acsami.8b00211

  38. [38]

      A. Grimaud, O. Diaz-Morales, B. Han, et al., Nat. Chem. 9 (2017) 457–465. doi: 10.1038/nchem.2695

  39. [39]

      B.J. Trzesniewski, O. Diaz-Morales, D.A. Vermaas, et al., J. Am. Chem. Soc. 137 (2015) 15112–15121. doi: 10.1021/jacs.5b06814

  40. [40]

      X. Bo, R.K. Hocking, S. Zhou, et al., Energy Environ. Sci. 13 (2020) 4225–4237. doi: 10.1039/d0ee01609h

  41. [41]

      Y.K. Bai, Y. Wu, X.C. Zhou, et al., Nat. Commun. 13 (2022) 6094. doi: 10.1038/s41467-022-33846-0

  42. [42]

      S. Lee, K. Banjac, M. Lingenfelder, et al., Angew. Chem. Int. Ed. 58 (2019) 10295–10299. doi: 10.1002/anie.201903200

  43. [43]

      J.X. Kang, Y.F. Xue, J. Yang, et al., J. Am. Chem. Soc. 144 (2022) 8969–8976. doi: 10.1021/jacs.1c13523

  44. [44]

      J.X. Kang, X.Y. Qiu, Q. Hu, et al., Nat. Catal. 4 (2021) 1050–1058. doi: 10.1038/s41929-021-00715-w

  45. [45]

      J.X. Kang, G. Liu, Q. Hu, et al., J. Am. Chem. Soc. 145 (2023) 25143–25149. doi: 10.1021/jacs.3c05688

  46. [46]

      B.S. Yeo, A.T. Bell, J. Am. Chem. Soc. 133 (2011) 5587–5593. doi: 10.1021/ja200559j

  47. [47]

      A. Moysiadou, S. Lee, C.S. Hsu, et al., J. Am. Chem. Soc. 10 (2017) 2190–2200.

  48. [48]

      W.T. Hong, K.A. Stoerzinger, Y.L. Lee, et al., Energy Environ. Sci. 13 (2020) 86–95. doi: 10.5749/vergstudglobasia.6.1.0086

  49. [49]

      L.K. Gao, X. Cui, C.D. Sewell, et al., Chem. Soc. Rev. 50 (2021) 8428. doi: 10.1039/d0cs00962h

  50. [50]

      H. Ding, H.F. Liu, W.S. Chu, et al., Chem. Rev. 121 (2021) 13174–13212. doi: 10.1021/acs.chemrev.1c00234

  51. [51]

      L. Liu, A. Corma, Nat. Rev. Chem. 5 (2021) 256–276. doi: 10.1038/s41570-021-00255-8

• 

  Figure 1  (a) Schematic of MEAs as OER catalysts synthesized by vacuum melting. (b) XRD patterns of the initial MEAs.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  SEM and corresponding SEM‒EDS mapping images of the electrochemically activated MEAs. (a, b) EA-FeCoCr. (c, d) EA-CoCrNi. (e, f) EA-FeCrNi. (g, h) EA-FeCoCrNi.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Electrochemical OER activity evaluations. (a-d) LSV polarization curves of various catalysts in 1 mol/L KOH solution at a scan rate of 10 mV/s. (e) Overpotentials of various catalysts at a current density of 10 mA/cm². (f) Tafel plots of various catalysts derived from the polarization curves in (a-d) LSV polarization curves of MEAs after 200 CV cycles. (g) Chronopotentiometric curves of various catalysts in 1 mol/L KOH at a current density of 1 A/cm² (without iR correction).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Electrocatalytic properties of the electrodes for the OER. (a-d) pH-dependent LSV polarization curves of various catalysts. (a) EA-FeCoCr, (b) EA-CoCrNi, (c) EA-FeCrNi and (d) EA-FeCoCrNi. (e) Specific OER activity at 1.55 V vs. RHE after iR correction as a function of pH. (f) Capacitive current density differences (Δj = (j_a − j_c)/2 at 0.82 V versus RHE as a function of scan rate. The linear slope is equivalent to that of C_dl. (g) The corresponding LSV polarization curves normalized to the specific activities of various catalysts by ECSAs. (h) Current density values at an overpotential of 300 mV. (i) Nyquist plots of various catalysts at 1.52 V versus RHE in 1 mol/L KOH solution.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Physicochemical property characterization. (a) XRD patterns of the electrochemically activated MEAs. (b) Raman spectra of the electrochemically activated MEAs.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) HRTEM image for Post-FeCoCrNi nanoparticle dropping after ultrasonic oscillation and corresponding SEAD image. (b) HAADF-STEM image, corresponding EDS element mapping showing the distribution of Ni, Fe, Co, Cr and O. (c) XPS spectra for all elements (Fe 2p, Co 2p, Cr 2p, Ni 2p and O 1s) in the FeCoCrNi alloy and peak fitting analysis of the MEAs before and after the chronopotentiometric test.

  下载: 全尺寸图片 幻灯片
•