Atomic-scale identity, density, and working mechanism of the true active sites in chameleon-reconstructed cobalt-sulfide oxygen-evolution precatalysts (S-CoOOH and O-CoS)

The cobalt sulfide $\mathrm{Co_9S_8}$ grown on single-walled carbon nanotube films is a highly active oxygen-evolution precatalyst that undergoes an irreversible, pH-dependent ('chameleon-like') reconstruction: in alkaline electrolyte it converts to a sulfur-containing cobalt oxyhydroxide (S-CoOOH), in neutral-to-mildly-alkaline electrolyte to an oxygenated cobalt sulfide (O-CoS), while at pH 5 it largely remains $\mathrm{Co_9S_8}$. The 2021 operando X-ray absorption study that established this left the actual catalytic structure unresolved: turnover frequencies were computed under the assumption that every metal atom is active, the mechanistic role of the residual sulfur was only enumerated as four competing hypotheses, no atomistic O-O bond-forming mechanism was established for either reconstructed phase, and it remained unknown whether the reconstruction outcome is predictable from precursor chemistry and electrolyte conditions. This problem asks for an operando, atomic-scale account of the reconstructed catalyst - a quantified census and identification of the electrochemically active sites, a definite mechanistic assignment of what the residual sulfur does, and the reaction mechanism by which O-O bonds form on S-CoOOH and O-CoS - together with the generalization question of whether precursor-to-active-phase evolution is rule-governed. Major methodological progress has been made on adjacent systems, but no resolution of these questions exists for the reconstructed sulfide-derived phases, so the problem is open.

Background

The oxygen evolution reaction (OER, $4\ \mathrm{OH^-}\rightarrow\mathrm{O_2}+2\ \mathrm{H_2O}+4\ e^-$ in alkaline media) is the four-electron anodic half-reaction of water splitting and the principal kinetic bottleneck of renewable hydrogen production. A central difficulty in studying non-precious OER electrocatalysts is that many transition-metal compounds are not themselves the catalyst: under anodic polarization, sulfides, phosphides, and nitrides convert in situ into (oxy)hydroxide-like surface species. Such materials are therefore called precatalysts, and their electrochemical transformation is called (self-)reconstruction or self-optimization. Because the reconstruction product - not the as-synthesized material - carries the catalysis, knowing its atomic structure is a prerequisite for any mechanistic claim or intrinsic-activity benchmark.

Operando X-ray absorption fine structure (XAFS) spectroscopy probes a catalyst while it is actually working: the near-edge part (XANES) reports the average metal oxidation state, while the extended part (EXAFS) reports the coordination shells around the absorbing atom (e.g., Co-S, Co-O, and Co-Co bonds), so stepped-potential XAFS can watch bonds break and form during reconstruction. Cao et al. (Energy & Environmental Science 14, 906-915, 2021) applied operando X-ray absorption spectroscopy at the cobalt K-edge (the absorption edge for ionizing a cobalt 1s electron), i.e. Co K-edge XANES/EXAFS, supported by ex-situ diffraction, electron microscopy, and photoelectron spectroscopy, to a freestanding film of $\mathrm{Co_9S_8}$ nanosheets grown in situ on single-walled carbon nanotubes (SWCNT). They visualized an irreversible, pH-dependent reconstruction that they called chameleon-like self-optimization: in 1 M KOH (pH 14) the precursor converts to a cobalt oxyhydroxide that still contains measurable sulfur (S-CoOOH), with the Co-S coordination fading and Co-O/Co-Co shells growing as potential increases up to about 1.2 V vs. RHE (reversible hydrogen electrode, the standard reference potential in electrocatalysis); in neutral phosphate buffer (pH 7) and at pH 10 it converts instead to an oxygen-rich cobalt sulfide (O-CoS), with changes beginning near 1.1 V and saturating near 1.6 V; at pH 5 it largely remains $\mathrm{Co_9S_8}$. The reconstructed film delivers 10 mA cm$^{-2}$ at an overpotential (the extra voltage beyond the thermodynamic reaction potential) of roughly 150 mV in alkaline solution and remains stable for about 30 h (about 10 h in neutral), and blocking surface sulfur sites by Zn$^{2+}$ immersion markedly decreases the activity - evidence that the residual sulfur participates in catalysis.

Two measurement conventions matter for what remains unresolved. The turnover frequency (TOF) - the reaction rate per active site, here computed as $\mathrm{TOF}=JA/(4Fn)$ with $J$ the current density, $A$ the electrode area, and $n$ the mole number of assumed active sites - is only as meaningful as the site count $n$; the authors computed TOF maps under the explicit rough assumption that all metal atoms are electrochemically active, and the electrochemical surface area (ECSA, estimated from double-layer capacitance) does not determine the number of catalytically competent sites either. Second, the paper could demonstrate that residual sulfur matters but not why: it listed four possible atomic-scale roles drawn from prior literature - improving electrical conductivity, stabilizing high-valent Co species, directly modulating adsorption energies of reaction intermediates, and accelerating proton charge transfer - without discriminating among them. Likewise, operando XAFS gives bond- and valence-state evolution, not the atomistic pathway of O-O bond formation (e.g., which surface oxygen species couple, and whether lattice oxygen participates), and its stability demonstrations extend only over the examined windows (about 30 h in alkaline, 10 h in neutral). Neighboring literature had shown that reconstruction-derived oxyhydroxides often outperform directly synthesized ones, but why the same precursor adapts differently to different pH - the question underlying the paper's own chameleon hypothesis, which it tested only at four pH values on this single material - and whether that adaptation is predictable from precursor chemistry remained open; later reviews of reconstruction electrocatalysis kept restating the predictability question.

Problem Statement

Establish the atomic-scale structure that actually catalyzes oxygen evolution after the $\mathrm{Co_9S_8}$-SWCNT precatalyst reconstructs - i.e., replace the current picture, in which the reconstructed phases S-CoOOH (alkaline) and O-CoS (neutral) are known only at the bond/oxidation-state level and TOFs assume every metal atom is active, with a defensible active-site model of the working catalyst. The common objective is a complete account of what the active sites are, how many there are, and how they turn over; its interlocking components, as left open by the source study and still open in the literature, are: (1) determine the identity and the operando density of the true electrochemically active sites on S-CoOOH and on O-CoS under their respective pH and potential windows, by a method whose site count is validated independently of the all-metal-atoms assumption; (2) identify the chemical state of the residual sulfur in S-CoOOH and determine which of the proposed roles its beneficial effect actually operates through - electronic conductivity, stabilization of high-valent Co, modification of intermediate adsorption energies, or accelerated proton transfer - and, if more than one contributes, their relative importance, since the Zn$^{2+}$ blocking experiment shows sulfur matters but not why; (3) establish the atomistic pathway of O-O bond formation on each reconstructed phase, i.e., which surface species couple and whether lattice or adsorbate oxygen participates, consistent with measured kinetics (Tafel slopes, i.e. the slope of overpotential versus the logarithm of current density, and pH dependence) in the respective electrolytes. As the generalization dimension of the same objective: (4) determine why the same $\mathrm{Co_9S_8}$-SWCNT precatalyst self-optimizes into different true-catalyst phases under different working conditions (S-CoOOH at pH 14, O-CoS at pH 7 and 10, largely untransformed $\mathrm{Co_9S_8}$ at pH 5), and whether that chameleon-like outcome is predictable from the precatalyst chemistry and electrolyte conditions - testing the source study's hypothesis that the precatalyst transforms into its most favorable phase under the actual working conditions, beyond the four pH values it examined. Components (1)-(3) jointly constitute the mechanistic core: a quantitative site census is required to interpret the sulfur-role experiments, and both are prerequisites for a site-resolved O-O mechanism; component (4) tests whether that understanding is general. Claims of resolution must be specific to (or validated on) the reconstructed sulfide-derived phases S-CoOOH and O-CoS of this $\mathrm{Co_9S_8}$-SWCNT system, not merely to model or adjacent catalysts.

References

1. Cao, Dengfeng; Liu, Daobin; Chen, Shuangming; Moses, Oyawale Adetunji; Chen, Xingjia; Xu, Wenjie; Wu, Chuanqiang; Zheng, Lirong; Chu, Shengqi; Jiang, Hongliang; Wang, Changda; Ge, Binghui; Wu, Xiaojun; Zhang, Jing; Song, Li. 'Operando X-ray spectroscopy visualizing the chameleon-like structural reconstruction on an oxygen evolution electrocatalyst.' Energy & Environmental Science 14(2), 906-915 (2021). DOI 10.1039/d0ee02276d, https://doi.org/10.1039/d0ee02276d
2. Cao, Dengfeng; Shou, Hongwei; Chen, Shuangming; Song, Li. 'Manipulating and probing the structural self-optimization in oxygen evolution reaction catalysts.' Current Opinion in Electrochemistry 30, 100788 (2021). DOI 10.1016/j.coelec.2021.100788, https://doi.org/10.1016/j.coelec.2021.100788
3. Wang, Sihong; Jiang, Qu; Ju, Shenghong; Hsu, Chia-Shuo; Chen, Hao Ming; Zhang, Di; Song, Fang. 'Identifying the geometric catalytic active sites of crystalline cobalt oxyhydroxides for oxygen evolution reaction.' Nature Communications 13, 6650 (2022). DOI 10.1038/s41467-022-34380-9, https://doi.org/10.1038/s41467-022-34380-9
4. Qian, Haitao; Wei, Jie; Yu, Chenchen; Tang, Fei; Jiang, Wulv; Xia, Dongsheng; Gan, Lin. 'In Situ Quantification of the Active Sites, Turnover Frequency, and Stability of Ni-Fe (Oxy)hydroxides for the Oxygen Evolution Reaction.' ACS Catalysis 12(22), 14280-14289 (2022). DOI 10.1021/acscatal.2c03898, https://doi.org/10.1021/acscatal.2c03898
5. Tang, Lei; Xia, Meihan; Cao, Shiyu; Bo, Xin; Zhang, Shengbo; Zhang, Yunlong; Liu, Xiao; Zhang, Lizhi; Yu, Liang; Deng, Dehui. 'Operando identification of active sites in Co-Cr oxyhydroxide oxygen evolution electrocatalysts.' Nano Energy 101, 107562 (2022). DOI 10.1016/j.nanoen.2022.107562, https://doi.org/10.1016/j.nanoen.2022.107562
6. Obradović, Maja D.; Gojković, Snežana Lj. 'Challenges in determining the electrochemically active surface area of Ni-oxides in the oxygen evolution reaction.' Journal of Electroanalytical Chemistry 918, 116479 (2022). DOI 10.1016/j.jelechem.2022.116479, https://doi.org/10.1016/j.jelechem.2022.116479
7. Hu, Jing; Al-Salihy, Adel; Wang, Jing; Li, Xue; Fu, Yanfei; Li, Zhonghua; Han, Xijiang; Song, Bo; Xu, Ping. 'Improved Interface Charge Transfer and Redistribution in CuO-CoOOH p-n Heterojunction Nanoarray Electrocatalyst for Enhanced Oxygen Evolution Reaction.' Advanced Science 8(22), 2103314 (2021). DOI 10.1002/advs.202103314, https://doi.org/10.1002/advs.202103314
8. Liao, Hanxiao; Luo, Tao; Tan, Pengfei; Chen, Kejun; Lu, Lili; Liu, Yong; Liu, Min; Pan, Jun. 'Unveiling Role of Sulfate Ion in Nickel-Iron (oxy)Hydroxide with Enhanced Oxygen-Evolving Performance.' Advanced Functional Materials 31(38), 2102772 (2021). DOI 10.1002/adfm.202102772, https://doi.org/10.1002/adfm.202102772
9. Feng, Kun; Song, Ruru; Xu, Jiabin; Chen, Yufeng; Lu, Cheng; Li, Youyong; Hofer, Werner; Lin, Haiping; Kang, Zhenhui; Zhong, Jun. 'The S-Fe(Ni) sub-surface active sites for efficient and stable overall water splitting.' Applied Catalysis B: Environmental 325, 122365 (2023). DOI 10.1016/j.apcatb.2023.122365, https://doi.org/10.1016/j.apcatb.2023.122365
10. Luo, Fengting; Xiang, Jueting; Jiang, Junjie; Yu, Pei; Chen, Shijian. 'Constructing quasi-amorphous cobalt oxyhydroxide nanowires with synergy of anion and cation via ion exchange reconstruction for large-current-density oxygen generation.' Chemical Engineering Journal 492, 152223 (2024). DOI 10.1016/j.cej.2024.152223, https://doi.org/10.1016/j.cej.2024.152223
11. Liu, Jiao; Qiao, Wen; Zhu, Zexuan; Hu, Jingguo; Xu, Xiaoyong. 'Chameleon-Like Reconstruction on Redox Catalysts Adaptive to Alkali Water Electrolysis.' Small 18(30), 2202434 (2022). DOI 10.1002/smll.202202434, https://doi.org/10.1002/smll.202202434
12. Wang, Wanying; Wang, Ansheng; Xu, Jinchao; Li, Huan; Yu, Meng; Dong, Anqi; Li, Zhenguo; Zhao, Chunning; Cheng, Fangyi; Wang, Weichao. 'Surface reconstruction of pyrite-type transition metal sulfides during oxygen evolution reaction.' Journal of Colloid and Interface Science 657, 334-343 (2024). DOI 10.1016/j.jcis.2023.11.130, https://doi.org/10.1016/j.jcis.2023.11.130
13. Peng, Jiayu. 'Toward Predictive Data-Driven Atomistic Modeling of Electrocatalyst Stability and Surface Reconstruction.' ChemRxiv preprint (2025). DOI 10.26434/chemrxiv-2025-12mdb, https://doi.org/10.26434/chemrxiv-2025-12mdb
14. Hausmann, J. Niklas; Menezes, Prashanth W. 'Why should transition metal chalcogenides be investigated as water splitting precatalysts even though they transform into (oxyhydr)oxides?' Current Opinion in Electrochemistry 34, 100991 (2022). DOI 10.1016/j.coelec.2022.100991, https://doi.org/10.1016/j.coelec.2022.100991
15. Zhong, Haoyin; Zhang, Qi; Yu, Junchen; Zhang, Xin; Wu, Chao; Ma, Yifan; An, Hang; Wang, Hao; Zhang, Jun; Wang, Xiaopeng; Xue, Junmin. 'Fundamental Understanding of Structural Reconstruction Behaviors in Oxygen Evolution Reaction Electrocatalysts.' Advanced Energy Materials 13(31), 2301391 (2023). DOI 10.1002/aenm.202301391, https://doi.org/10.1002/aenm.202301391
16. Wang, Jian. 'Controlling dynamic reconstruction chemistry for superior oxygen-evolving catalysts.' Chem 9(7), 1645-1657 (2023). DOI 10.1016/j.chempr.2023.06.001, https://doi.org/10.1016/j.chempr.2023.06.001
17. Zhang, Shiqi; Cheng, Ying; Wang, Peng; Lei, Xuefei; You, Junhua; Guo, Rui; Zhang, Hangzhou. 'Research progress on surface reconstruction of transition metal sulfides (TMS, M=Fe, Co, Ni) as precatalysts for oxygen evolution reaction.' International Journal of Hydrogen Energy 51, 545-560 (2024). DOI 10.1016/j.ijhydene.2023.10.122, https://doi.org/10.1016/j.ijhydene.2023.10.122
18. Ma, Peiyu; Jia, Chuanyi; Zhang, Zhirong; Bao, Jun. 'Structural Reconstruction Chemistry for Alkaline Oxygen Evolution Catalysts.' ChemElectroChem 11(13), e202400076 (2024). DOI 10.1002/celc.202400076, https://doi.org/10.1002/celc.202400076
19. Xia, Chenfeng; Li, Fu-Min; He, Chaohui; Zaman, Shahid; Guo, Wei; Xia, Bao Yu. 'Structural reconstruction of electrocatalysts.' Fundamental Research 5(6), 2537-2552 (2025). DOI 10.1016/j.fmre.2024.04.017, https://doi.org/10.1016/j.fmre.2024.04.017
20. Qin, Minkai; Chen, Jiadong; Zheng, Xiaozhong; Qi, Menghui; Yang, Rui; Mao, Shanjun; Wang, Yong. 'Operando deciphering the activity origins for potential-induced reconstructed oxygen-evolving catalysts.' Applied Catalysis B: Environmental 316, 121602 (2022). DOI 10.1016/j.apcatb.2022.121602, https://doi.org/10.1016/j.apcatb.2022.121602
21. Wu, Jie; Zhang, Yuanyuan; Zhang, Bin; Li, Siwei; Xu, Ping. 'Zn-Doped CoS2 Nanoarrays for an Efficient Oxygen Evolution Reaction: Understanding the Doping Effect for a Precatalyst.' ACS Applied Materials & Interfaces 14(12), 14235-14242 (2022). DOI 10.1021/acsami.2c00455, https://doi.org/10.1021/acsami.2c00455