Sulfur-doped cobalt–nitrogen–carbon materials for efficient oxygen electrocatalysis

Abstract

The oxygen reduction reaction (ORR) is central to clean energy technologies such as metal–air batteries, but its sluggish kinetics typically rely on precious metal catalysts. Herein, a sulfur-functionalized cobalt–nitrogen–carbon catalyst (S@Co–N–C) was successfully synthesized via a thiourea-assisted pyrolysis strategy using a two-dimensional (2D) zeolitic imidazolate framework (ZIF) as the precursor. Experimental characterization revealed that S-doping effectively modulated electronic structures of Co–N₄ sites, significantly enhancing the intrinsic ORR activity of the Co–N–C material. In 0.1 M KOH, S@Co–N–C exhibited a half-wave potential (E_1/2) of 0.895 V (vs. RHE), surpassing that of commercial Pt/C (20 wt%; 0.870 V vs. RHE). Density functional theory (DFT) calculations confirmed that the introduction of S atoms optimized the d-band center of Co sites and reduced the *OH desorption energy barrier, thereby accelerating ORR kinetics. Furthermore, a Zn–air battery assembled with S@Co–N–C delivered a peak power density of 210 mW cm⁻², outperforming the Pt/C + RuO₂ benchmark (140 mW cm⁻²). S@Co–N–C also demonstrated superior stability for both the ORR and Zn–air battery compared to the control sample Co–N–C and commercial benchmark. This study provides new insights into the design of non-precious metal ORR catalysts with high stability and elucidates the critical role of S-doping in M–N–C materials.

---

1. Introduction

Metal–air batteries have garnered extensive attention as promising energy storage devices due to their high safety, environmental friendliness, and superior specific energy.^1–6 Central to their operation is the oxygen reduction reaction (ORR), a core electrochemical process critical for energy conversion and storage technologies like fuel cells and metal–air batteries.^7–14 However, the practical efficiency of these technologies is severely hampered by the sluggish kinetics and high overpotential inherent to the ORR.^5,15–19 Platinum-based catalysts currently represent the benchmark for ORR activity, while their prohibitively high cost and scarcity present significant barriers to large-scale commercialization.^20–24 Consequently, the development of high-efficiency, stable, and low-cost non-precious metal catalysts has emerged as a paramount research objective.^25–28 Among the promising alternatives, transition metal–nitrogen–carbon (M–N–C, where M = Fe, Co, Mn, etc.) catalysts, which usually feature M–N₄ coordination structures, have been extensively explored.^29–32 The M–N₄ structure mimics the active sites in biological enzymes and metal macrocyclic compounds (such as porphyrins), providing well-defined and uniform active centers.³³ Moreover, the electronic structure of the central metal atom and nitrogen coordination environment can be effectively regulated to optimize the adsorption energy of reaction intermediates, break linear scaling relationships, and achieve excellent intrinsic activity.^34–36 Critically, the performance of the M–N–C catalyst is profoundly dependent on their precise local coordination environment and electronic structure.^37–40 Therefore, designing M–N–C catalysts that simultaneously achieve high efficiency and long-term stability remains a formidable challenge.

Zeolitic imidazolate frameworks (ZIFs), a prominent subclass of metal–organic frameworks (MOFs), combine the inherent advantages of zeolite-like structures and MOFs, such as high specific surface area and tunable pore size, with their utility as excellent templates for preparing M–N–C catalysts through pyrolysis at high temperature.^41–47 Among various ZIFs, ZIF-L has emerged as a significant research focus with distinct advantages including an environmentally benign aqueous synthesis route, a unique two-dimensional (2D) morphology, and abundant surface ligand-unsaturated sites.^19,48–50 The characteristic 2D structure offers multiple decisive benefits for electrocatalysis, which exposes a greater number of active sites, shortens mass/charge transfer pathways, and features a highly porous layered architecture.^51–54 To further optimize the intrinsic ORR activity within M–N–C catalysts, introducing secondary heteroatoms is an effective strategy, which induces charge redistribution, optimizing interactions with adsorbed molecular O₂ and key oxygenated intermediates (e.g., *O and *OH).^9,55–57 Specifically, sulfur doping stands out as a powerful approach for electronic structure modulation, leading to significant ORR performance enhancement.^18,58–60 Incorporating sulfur functional groups, such as thiophene-S and oxidized-S, effectively tailors the electronic distribution around single-atom M–N_x active sites, thereby directly improving their intrinsic catalytic activity.^61,62 However, the underlying mechanism governing this improvement remains ambiguous. Particularly for ZIF-L-derived carbon systems, the dynamic interplay between S-functional groups and Co–N_x moieties in the 4e⁻ ORR pathway has not been quantitatively correlated with experimental observations.

In this work, we synthesized a sulfur-functionalized Co–N–C catalyst (S@Co–N–C) using a sulfurization-assisted pyrolysis strategy, with ZIF-L as the sacrificial template and thiourea as the sulfur source. Through systematic experimental characterization and theoretical calculations, we comprehensively investigated the formation of sulfur functional groups and their precise regulatory mechanism on the intrinsic activity of the Co sites, elucidating their critical role in enhancing ORR performance. A key finding is that sulfur incorporation significantly enriches the structural diversity of the catalytic material, generating additional active sites that collectively accelerate ORR kinetics. Consequently, the S@Co–N–C catalyst demonstrates exceptional ORR performance in alkaline electrolytes, characterized by a favorable half-wave potential (E_1/2) and a low Tafel slope, alongside excellent stability and methanol tolerance. To uncover the origin of this enhanced efficiency, density functional theory (DFT) calculations were performed. These calculations revealed that sulfur atom incorporation effectively modulates the electron density of the metal center and optimizes the adsorption energies of key reaction intermediates, thereby facilitating the overall reaction kinetics. This work not only provides novel and fundamental insights into the design of high-efficiency ORR catalysts but also establishes a robust theoretical foundation for the strategic application of sulfur doping in advanced electrocatalysis.

2. Experimental section

2.1 Synthesis of ZIF-L

A solution containing Zn(NO₃)₂ (1 mmol, 0.30 g) and Co(NO₃)₂ (1 mmol, 0.29 g) in 40 mL deionized water was prepared. Separately, 2-methylimidazole (16 mmol, 1.31 g) was dissolved in 40 mL deionized water via ultrasonication. The metal ion solution was then added dropwise to the 2-methylimidazole solution under continuous stirring. After complete addition, the resulting mixture was stirred for an additional 2 h. The suspension was collected by centrifugation (12 000 rpm, 3 min, 20 °C), washed three times with deionized water, and vacuum-dried at 60 °C for 12 h to afford a purple powder, denoted as ZIF-L.

2.2 Synthesis of S@ZIF-L

A solution containing Zn(NO₃)₂ (1 mmol, 0.30 g) and Co(NO₃)₂ (1 mmol, 0.29 g) in 40 mL deionized water was prepared. Separately, a ligand mixture of 2-methylimidazole (14 mmol, 1.15 g) and thiourea (TA, 1 mmol, 0.075 g; 2 mmol, 0.15 g; 4 mmol, 0.30 g) in 14 : 1, 7 : 1 and 7 : 2 molar ratios was dissolved in 40 mL deionized water via ultrasonication until complete dissolution. The metal ion solution was then added dropwise to the ligand mixture solution under continuous stirring. The mixture was stirred for an additional 2 h to obtain a purple powder, denoted as 0.5 × TA@ZIF-L, TA@ZIF-L and 2 × TA@ZIF-L. A similar procedure was carried out by replacing TA with 2,4-dimethylthiazole (DT) and trithiocyanuric acid (TTA), and the resulting products were denoted as DT@ZIF-L and TTA@ZIF-L.

2.3 Synthesis of Co–N–C and S@Co–N–C

To prevent structural collapse during pyrolysis, a stepwise heating protocol was employed. Under an inert argon atmosphere, the precursors ZIF-L and S@ZIF-L were heated to 600 °C at a ramp rate of 2 °C min⁻¹, held for 3 h, and then heated to 950 °C at 3 °C min⁻¹ and held for another 3 h. The resulting black powders were denoted as Co–N–C and S@Co–N–C, respectively. A similar procedure was carried out for other S@ZIF-L precursors. Subsequently, the pyrolyzed S@Co–N–C product was leached in HCl solution to remove residual particles along with inactive and unstable species. This acid-washed material is referred to as washed S@Co–N–C.

3. Results and discussion

3.1 Characterization of Co–N–C and S@Co–N–C

Fig. 1a illustrates the synthesis procedure of S@Co–N–C. Scanning electron microscopy (SEM) images confirmed the formation of 2D leaf-like ZIF-L nanoparticles with dimensions of 4–5 µm (Fig. S1). Subsequent introduction of thiourea into the synthesis system yielded TA@ZIF-L. SEM analysis revealed that TA@ZIF-L retained the original leaf-like morphology and size (∼4–5 µm) (Fig. 1b), indicating that the incorporation of sulfur at this stage does not alter the main structure and morphology of the precursor. Furthermore, samples with different S loadings were prepared, and it was found that the S loading did not affect the morphology of the composite material (Fig. S2). This result was further supported by powder X-ray diffraction (XRD) patterns, which showed identical peaks for ZIF-L and TA@ZIF-L, confirming the preserved crystal structure despite the sulfur incorporation (Fig. S3 and S4). Following pyrolysis, distinct morphological differences emerged between the catalysts. Co–N–C adopted an irregular block-like morphology with carbon nanotubes (CNTs) formed on its surface (Fig. S5). In contrast, S@Co–N–C and washed S@Co–N–C remarkably preserved the leaf-like morphology of its ZIF-L precursor (Fig. 1c, d, and S6). Significantly, S-doping effectively suppressed CNT formation on S@Co–N–C and promoted a more uniform dispersion of Co nanoparticles on its surface. Transmission electron microscopy (TEM) images verified the presence of aggregated Co particles on S@Co–N–C and washed S@Co–N–C (Fig. 1e and S7). The high-resolution TEM (HRTEM) image revealed distinct lattice fringes with a spacing of 0.18 nm within S@Co–N–C, corresponding to the (111) plane of metallic Co, which was encapsulated by layered graphitic carbon with a spacing of 0.32 nm (Fig. 1f). Critically, elemental mapping confirmed the homogeneous spatial distribution of Co, N, and S throughout the S@Co–N–C structure (Fig. 1g). Inductively coupled plasma optical emission spectrometry (ICP-OES) analysis revealed that the Co and S content in S@Co–N–C is 24.3 wt% and 0.41 wt%, respectively, further confirming the presence of S (Table S1).

Fig. 1 (a) Synthesis procedure of S@Co–N–C. SEM images of TA@ZIF-L (b) and S@Co–N–C (c and d). (e) TEM image, (f) HRTEM image, (g) HAADF-STEM image, and corresponding N, S, and Co elemental mappings of S@Co–N–C.

Brunauer–Emmett–Teller (BET) analysis confirmed micro/mesoporous structures in both Co–N–C, S@Co–N–C and washed S@Co–N–C catalysts. S@Co–N–C exhibited a specific surface area of 384.0 m² g⁻¹ (Fig. S8a) and displayed a pore size distribution centered around 3.02 nm, indicative of its mesoporous character (Fig. S8b). Washed S@Co–N–C exhibited a specific surface area of 555.8 m² g⁻¹ (Fig. S9a) and displayed a pore size distribution centered around 3.13 nm (Fig. S9b). This distribution further revealed the coexistence of micropores and mesopores within the material. Micropores contribute to enhanced stability and selectivity by providing abundant active sites, while mesopores increase surface area, reaction site accessibility, active site dispersion, and mass transport efficiency. The synergistic combination of micro- and mesopores thus optimizes adsorption capacity and mass transfer pathways, collectively boosting catalytic ORR performance. In contrast, Co–N–C showed a significantly lower specific surface area of 185.3 m² g⁻¹ (Fig. S10a). The pore size distribution of Co–N–C also showed distinct mesoporous structures but with a larger average pore size of 6.11 nm (Fig. S10b). Critically, the introduction of sulfur during synthesis led to a substantial increase in the mesoporous structure and a higher surface area compared to the undoped Co–N–C.

Raman spectroscopy confirmed carbon structures of catalysts, revealing characteristic D-band (defect carbon, 1340 cm⁻¹) and G-band (graphitic carbon, 1580 cm⁻¹) peaks (Fig. S11). The higher intensity ratio for S@Co–N–C (I_D/I_G = 0.98) and washed S@Co–N–C (I_D/I_G = 1.02) compared to Co–N–C (I_D/I_G = 0.81) indicates an increased graphitic carbon content in the sulfur-doped material (Fig. S11 and S12).⁶³ The increase in graphitic carbon content significantly enhances both the activity and stability of the catalyst by improving structural stability and synergistically optimizing the electronic structure of Co active centers in combination with S doping. XRD analysis identified metallic Co in catalysts. XRD patterns exhibited peaks at 44.21° and 51.51°, corresponding to the (111) and (200) planes of metallic Co (Fig. 2a).⁶⁴ X-ray photoelectron spectroscopy (XPS) survey spectra confirmed the presence of Co, N, and C in both materials, with an additional distinct S signal observed in S@Co–N–C and washed S@Co–N–C, unequivocally confirming successful sulfur incorporation (Fig. 2b and S13). High-resolution XPS spectra provided further chemical state analysis. The C 1s XPS spectrum displayed three peaks at 284.4 eV (C–C), 285.8 eV (C–N), and 289.8 eV (C–O) (Fig. 2c).⁶⁵ The high-resolution S 2p XPS spectrum of S@Co–N–C revealed peaks at 163.3 eV, 164.8 eV, and 168.8 eV, attributed to C–S–C (2p_3/2), C–S–C (2p_1/2), and SO_x species, respectively (Fig. 2d).⁶⁶ In the N 1s XPS spectrum of S@Co–N–C, distinct peaks corresponding to pyridinic N (398.7 eV), graphitic N (400.9 eV), and oxidized N (403.9 eV) were observed (Fig. 2e).⁶⁷ The Co 2p XPS spectrum of S@Co–N–C showed peaks at 778.4 eV (metallic Co), 780.7 eV (Co–C_xN_y), and 784.2 eV (Co–N_x) (Fig. 2f).⁶⁸

Fig. 2 (a) XRD patterns, (b) full survey XPS spectra, and high resolution XPS spectra of (c) C 1s, (d) S 2p, (e) N 1s, and (f) Co 2p of S@Co–N–C and Co–N–C. (g) Normalized Co K-edge XANES and (h) k₃-weighted FT-EXAFS spectra (without phase correction) of Co foil, CoPc, Co(OH)₂, S@Co–N–C and washed S@Co–N–C. (i) FT-EXAFS fitting curve of washed S@Co–N–C.

Fig. 2g shows the X-ray absorption near-edge structure (XANES) spectra at the Co K-edge of S@Co–N–C before and after washing with reference samples including Co foil and Co(OH)₂. S@Co–N–C primarily exists in the form of metallic Co. In contrast, the normalized Co K-edge XANES of the washed S@Co–N–C exhibits an absorption edge similar to that of Co(OH)₂, indicating a Co valence state close to +2 (Fig. 2g). As shown in Fig. 3h, the FT-EXAFS analysis of washed S@Co–N–C reveals a main peak at 2.13 Å and a shoulder peak at 1.37 Å, which may be attributed to Co–Co and Co–N scattering, respectively. Due to the low doping amount of S, the Co–S signal might be overshadowed by the dominant Co–Co signal. The EXAFS fitting results for the S@Co–N–C sample indicate the presence of a Co–Co coordination path (Fig. S14). After acid washing, the Co–N₃S coordination signal was clearly revealed in the EXAFS fitting due to the removal of metal nanoparticles. Furthermore, the EXAFS fitting results confirm that washed S@Co–N–C has a Co–N coordination number and average bond length of 3.07 and 1.817 Å, respectively, while the corresponding values for Co–S are 0.76 and 2.553 Å (Fig. 2i and S15). Therefore, the doping of S transforms a Co–N₄ site into a Co–N₃S site.

Fig. 3 (a) CV data of S@Co–N–C measured in N₂- and O₂-saturated 0.1 M KOH solutions. (b) ORR LSV data, (c) comparison of E_1/2, (d) Tafel slopes, (e) calculated n and H₂O₂ values, (h) methanol resistance tests, and (i) stability tests of S@Co–N–C, Co–N–C, and Pt/C. (f) ORR LSV data at different rotation speeds and (g) K–L plots of S@Co–N–C.

3.2 Electrocatalytic ORR studies of S@Co–N–C

The electrocatalytic ORR performance was evaluated using a rotating ring-disk electrode (RRDE) in an O₂-saturated 0.1 M KOH electrolyte. Cyclic voltammetry (CV) tests conducted in N₂- and O₂-saturated 0.1 M KOH solutions revealed a distinct O₂ reduction peak for S@Co–N–C under O₂ conditions compared to the N₂ baseline, confirming its electrocatalytic ORR activity (Fig. 3a). Linear sweep voltammetry (LSV) measurements at 1600 rpm demonstrated that S@Co–N–C exhibited a half-wave potential (E_1/2) of 0.895 V (vs. RHE), outperforming the benchmark commercial Pt/C (20 wt%) with a value of 0.870 V (vs. RHE). LSV curves of samples with different S loadings were measured, and the sample with 2-methylimidazole and thiourea in a 7 : 1 molar ratio exhibited the optimal ORR performance (Fig. S16). Besides, the half-wave potential of S@Co–N–C is much larger than that of Co–N–C (0.860 V (vs. RHE)), washed S@Co–N–C (0.875 V (vs. RHE)), DT@Co–N–C (0.865 V (vs. RHE)), and TTA@Co–N–C (0.860 V (vs. RHE)) (Fig. 3b and S17–S19). Fig. 3c provides a comparative overview of E_1/2 values for these catalysts. During acid washing, Co nanoparticles were removed from the S@Co–N–C sample, which disrupted the synergistic effect of Co nanoparticles and Co–N₃S active sites and consequently led to a partial performance decline. The Tafel slope of S@Co–N–C (83 mV dec⁻¹) is also lower than that of Co–N–C (112 mV dec⁻¹), Pt/C (98 mV dec⁻¹), washed S@Co–N–C (86 mV dec⁻¹), DT@Co–N–C (107 mV dec⁻¹), and TTA@Co–N–C (130 mV dec⁻¹), verifying the fast ORR kinetics (Fig. 3d, S20, and S21). The above results indicate that the synergistic effect between Co nanoparticles and sulfur enhances the ORR performance. The electron transfer number (n) and H₂O₂ yield (%) were calculated based on the current density measured during RRDE experiments. S@Co–N–C, washed S@Co–N–C, and Co–N–C exhibited average n values of 3.82, 3.65, and 3.70, respectively, across a wide potential range of 0.2–0.8 V (vs. RHE), approaching the n value of 3.98 observed for Pt/C. This result indicates a dominant four-electron (4e⁻) ORR pathway for S@Co–N–C, washed S@Co–N–C, and Co–N–C. Correspondingly, the H₂O₂ yield of Pt/C, S@Co–N–C, washed S@Co–N–C, and Co–N–C within the same potential range is ∼1%, ∼9%, ∼20%, and ∼14%, respectively (Fig. 3e and S22). LSV data for S@Co–N–C and washed S@Co–N–C were also recorded at various rotation speeds (Fig. 3f and S23). Koutecky–Levich (K–L) analysis of the j⁻¹vs. ω^−1/2 plots yielded an n value of 3.98 and 3.75 for S@Co–N–C and washed S@Co–N–C, consistent with RRDE results (Fig. 3g and S24). S@Co–N–C and washed S@Co–N–C demonstrated superior methanol tolerance compared to Pt/C, as evidenced by the minimal current density change upon methanol introduction into the electrolyte (Fig. 3h and S25). Furthermore, the stability of S@Co–N–C was assessed at 1600 rpm under electrolytic conditions, revealing only a 3% loss in current density after 10 hours. This performance significantly surpassed those of Pt/C and Co–N–C, which showed losses of 20% and 23%, respectively (Fig. 3i), highlighting the critical role of sulfur incorporation in enhancing ORR stability. Moreover, the sample maintained its original morphology after the stability test (Fig. S26).

3.3 Electrocatalytic OER and Zn–air battery studies of S@Co–N–C

The oxygen evolution reaction (OER) performance of S@Co–N–C was evaluated using a glassy carbon electrode (GCE) in 1.0 M KOH solution. S@Co–N–C exhibited an overpotential (η) of 335 mV at j = 10 mA cm⁻² and a Tafel slope of 101 mV dec⁻¹. Differential pulse voltammetry (DPV) analysis revealed distinct oxidation peaks for S@Co–N–C, corresponding to Co^III/II and Co^IV/III oxidation peaks (Fig. 4b). For comparison, commercial RuO₂ under identical conditions showed an η of 345 mV and a Tafel slope of 99 mV dec⁻¹ (Fig. 4a and c). To evaluate the electrochemical active surface area (ECSA), the electrochemical double-layer capacitance (C_dl) was measured. The C_dl value, proportional to ECSA, was derived from CV data in a non-faradaic potential region at various scan rates (Fig. S27). S@Co–N–C exhibited a higher C_dl (8.8 mF) than RuO₂ (5.7 mF), indicating a larger ECSA and more exposed active sites (Fig. S28). Overall, S@Co–N–C demonstrated excellent bifunctional electrocatalytic activity for both the ORR and OER.

Fig. 4 (a) LSV data and (c) Tafel plots of S@Co–N–C and RuO₂. (b) DPV of S@Co–N–C. (d) Schematic illustration of a rechargeable Zn–air battery. (e) Open circuit voltage, (f) discharge polarization curves and the corresponding power density curves, (g) discharge curves at j = 20.0 mA cm⁻², and (h) discharge–charge cycling data at j = 2.0 mA cm⁻² for Zn–air batteries constructed with S@Co–N–C and Pt/C + RuO₂. (i) Long-time cycling test of a Zn–air battery measured at j = 2.0 mA cm⁻² with S@Co–N–C.

Given the superior bifunctional ORR/OER performance of S@Co–N–C, its application in rechargeable Zn–air batteries was comprehensively evaluated under alkaline conditions. A Zn–air battery was assembled using a Zn plate anode, a catalyst-coated carbon cloth air cathode (Fig. 4d), and an electrolyte containing Zn(OAc)₂·2H₂O (0.2 M) and KOH (6.0 M). A mixture of commercial Pt/C and RuO₂ served as the benchmark cathode catalyst. The S@Co–N–C-based Zn–air battery exhibited an open-circuit voltage of 1.49 V, surpassing that of the Pt/C + RuO₂ benchmark (1.47 V) (Fig. 4e). Furthermore, S@Co–N–C delivered a higher peak discharge power density of 210 mW cm⁻² compared to Pt/C + RuO₂ (140 mW cm⁻²) (Fig. 4f). At a discharge current density of 20.0 mA cm⁻², S@Co–N–C demonstrated a higher specific capacity of 844 mW cm⁻²versus 703 mW cm⁻² for Pt/C + RuO₂ (Fig. 4g). Long-term stability was assessed via galvanostatic discharge–charge cycling at j = 2.0 mA cm⁻². The S@Co–N–C-based battery exhibited a smaller charge–discharge voltage gap (0.70 V) compared to the Pt/C + RuO₂ battery (0.79 V) (Fig. 4h). The S@Co–N–C-based battery also maintained a stable charge–discharge voltage for over 160 hours (Fig. 4i).

3.4 Theoretical calculations

To further investigate the synergistic effect of Co sites and sulfur functional groups on ORR activity, density functional theory (DFT) calculations were performed. Atomic structure models of Co–N₄ and Co–N₃S were constructed (Fig. 5a and b). In the Co–N₄ structure, Co atom coordinates with four N atoms in a square-planar geometry. Substituting one nitrogen atom with sulfur forms the Co–N₃S structure, where Co coordinates with three N atoms and one S atom. This modification significantly alters the electronic distribution at the Co site.

Fig. 5 (a) Model structure and (a₁) charge density difference diagrams for Co–N₃S. (b) Model structure and (b₁) charge density difference diagrams for Co–N₄. Model structures of (a₂) Co–N₃S–OO, (a₃) Co–N₃S–OOH, (a₄) Co–N₃S–OH, (a₅) Co–N₃S–O, (b₂) Co–N₄–OO, (b₃) Co–N₄–OOH, (b₄) Co–N₄–OH, and (b₅) Co–N₄–O intermediates. (c) Free energy diagrams of the ORR for Co–N₃S and Co–N₄ at 1.23 V.

Differential charge density maps illustrate electron redistribution. In Co–N₄, electrons primarily transfer from Co to the N atoms, resulting in a higher oxidation state for Co (Fig. 5b₁). Conversely, in Co–N₃S, sulfur incorporation reduces charge transfer from Co to N. Electrons from S transfer through the carbon skeleton to the Co center, populating its antibonding orbitals and thereby weakening the adsorption of ORR intermediates (Fig. 5a₁). This electron redistribution lowers the oxidation state of Co, facilitating electron acceptance during the ORR and improving catalytic efficiency.

Free energy calculations for the 4e⁻ ORR pathway (O₂ → *OOH → *O → *OH → H₂O) identified the *OH desorption step as the potential-determining step, exhibiting the highest energy barrier (Fig. 5a₂–a₅ and b₂–b₅). DFT calculations showed an overpotential of 0.70 V for Co–N₄, which decreases significantly to 0.48 V for Co–N₃S (Fig. 5c). This reduction in overpotential corresponds to markedly enhanced ORR activity for the sulfur-modified structure. Collectively, both experimental and theoretical results confirm that sulfur functional groups regulate the electronic structure of Co–N₄ sites, moderately weakening intermediate adsorption and consequently improving ORR kinetics.

4. Conclusions

In summary, the sulfur-functionalized Co–N–C catalyst was successfully synthesized via thiourea-assisted pyrolysis of ZIF-L precursors. S-doping effectively preserved the 2D leaf-like morphology of the precursor while modulating the electronic structure of Co centers through the formation of Co–N₃S active sites. Experimental results demonstrated that S@Co–N–C exhibits outstanding ORR activity (E_1/2 = 0.895 V vs. RHE) and stability (only 3% current decay after 10 hours) in alkaline media, surpassing the performance of commercial Pt/C. DFT calculations, which further elucidated that sulfur incorporation weakened the adsorption strength of ORR intermediates and thereby significantly enhanced the catalytic kinetics. Moreover, S@Co–N–C delivered a high peak power density (210 mW cm⁻²) and demonstrated long-term cycling stability (>160 hours) in Zn–air batteries, highlighting its practical potential. This work provides theoretical insights for designing S-doped catalysts and paves the way for developing efficient, low-cost energy conversion materials. Future studies could explore the universality of S-doping in other transition metal systems and optimize the synergistic effects within multi-element catalysts.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5se01222h.

Acknowledgements

We are grateful for support from the National Natural Science Foundation of China (22178213 and 22478237) and Fundamental Research Funds for the Central Universities (GK202309002).

References

1. R. Wang, Y. Wu, Y. Niu, Q. Yang, H. Li, Y. Song, B. Zhong, L. Yang, T. Chen, Z. Wu and X. Guo, Chin. J. Chem., 2024, 42, 2056–2065.
2. C.-X. Zhao, J.-N. Liu, J. Wang, D. Ren, B.-Q. Li and Q. Zhang, Chem. Soc. Rev., 2021, 50, 7745–7778.
3. Z. Liang, H. Lei, H. Zheng, H.-Y. Wang, W. Zhang and R. Cao, Chem. Soc. Rev., 2025, 54, 5248–5291.
4. F. Qian, L. Peng, Y. Zhuang, L. Liu and Q. Chen, Chin. J. Chem. Eng., 2023, 61, 140–146.
5. L. Song, Y. Liu, J.-J. Wang, R. Wu, J.-S. Dang, H. Zhang, W. Zhang, R. Cao and H. Zheng, Energy Mater., 2025, 5, 500139.
6. Y. Zhou, R. Li, Z. Lv, J. Liu, H. Zhou and C. Xu, Chin. J. Chem. Eng., 2022, 43, 2–13.
7. Y.-J. Wang, W. Long, L. Wang, R. Yuan, A. Ignaszak, B. Fang and D. P. Wilkinson, Energy Environ. Sci., 2018, 11, 258–275.
8. M. Shao, Q. Chang, J.-P. Dodelet and R. Chenitz, Chem. Rev., 2016, 116, 3594–3657.
9. B. Natarajan, P. Kannan, G. Maduraiveeran and A. S. Alnaser, Adv. Colloid Interface Sci., 2025, 343, 103557.
10. K. Wei, X. Wang and J. Ge, Energy Mater., 2023, 3, 300051.
11. Y. Mou, J. Zhang, H. Qin, X. Li, Z. Zeng, R. Zhang, Z. Liang and R. Cao, Chem. Commun., 2025, 61, 1878–1881.
12. J. Zhang, W. Suo, Y. Han, Y. Cao, Y. Xu, M. Wang, Z. Liang, Y. Wang, H. Zheng and R. Cao, J. Mater. Chem. A, 2025, 13, 669–679.
13. Y. Cao, Y. Mou, J. Zhang, R. Zhang and Z. Liang, Mater. Today Catal., 2024, 4, 100044.
14. J. Chen, L. Li, L. Yang, C. Chen, S. Wang, Y. Huang and D. Cao, Chin. J. Chem. Eng., 2022, 43, 161–168.
15. Z. Liang, N. Kong, C. Yang, W. Zhang, H. Zheng, H. Lin and R. Cao, Angew. Chem., Int. Ed., 2021, 60, 12759–12764.
16. Y. Wang, T. Yang, X. Fan, Z. Bao, A. Tayal, H. Tan, M. Shi, Z. Liang, W. Zhang, H. Lin, R. Cao, Z. Huang and H. Zheng, Angew. Chem., Int. Ed., 2024, 63, e202313034.
17. Z. Liang, G. Zhou, H. Tan, Y. Mou, J. Zhang, H. Guo, S. Yang, H. Lei, H. Zheng, W. Zhang, H. Lin and R. Cao, Adv. Mater., 2024, 36, 2408094.
18. H.-C. Li, P.-C. Ji, Y. Teng, H.-L. Jia and M.-Y. Guan, Sustainable Energy Fuels, 2023, 7, 1127–1134.
19. J. Zhang, Z. Liang, Z. Li, S. Qu, Y. Cao, H. Zheng and R. Cao, Sustainable Energy Fuels, 2025, 9, 3152–3181.
20. Y. Song, K. Ji, H. Duan and M. Shao, Exploration, 2021, 1, 20210050.
21. Z. Ma, Z. P. Cano, A. Yu, Z. Chen, G. Jiang, X. Fu, L. Yang, T. Wu, Z. Bai and J. Lu, Angew. Chem., Int. Ed., 2020, 59, 18334–18348.
22. X. Chen, Z. Zhang, Y. Chen, R. Xu, C. Song, T. Yuan, W. Tang, X. Gao, N. Wang and L. Cui, Energy Mater., 2023, 3, 300031.
23. M. Wang, X. Feng, S. Li, Y. Ma, Y. Peng, S. Yang, Y. Liu, H. Lei, J. Dang, W. Zhang, R. Cao and H. Zheng, Adv. Funct. Mater., 2024, 34, 2410439.
24. S. Liu, H. Tan, G. Dai, S. Xiong, Y. Zhao and B. Li, Energy Mater., 2025, 5, 500144.
25. Z. Shi, W. Yang, Y. Gu, T. Liao and Z. Sun, Adv. Sci., 2020, 7, 2001069.
26. N. Cheng, L. Ren, X. Xu, Y. Du and S. X. Dou, Adv. Energy Mater., 2018, 8, 1801257.
27. G. Fang, J. Gao, J. Lv, H. Jia, H. Li, W. Liu, G. Xie, Z. Chen, Y. Huang, Q. Yuan, X. Liu, X. Lin, S. Sun and H.-J. Qiu, Appl. Catal., B, 2020, 268, 118431.
28. Y. Wu, Y. Yang, T. Gao, X. Chen, Z. Huang, J. Zhang, Y. Guo and D. Xiao, Sustainable Energy Fuels, 2020, 4, 4117–4125.
29. L. Peng, J. Yang, Y. Yang, F. Qian, Q. Wang, D. Sun-Waterhouse, L. Shang, T. Zhang and G. I. N. Waterhouse, Adv. Mater., 2022, 34, 2202544.
30. J. Li, W. Xia, J. Tang, Y. Gao, C. Jiang, Y. Jia, T. Chen, Z. Hou, R. Qi, D. Jiang, T. Asahi, X. Xu, T. Wang, J. He and Y. Yamauchi, J. Am. Chem. Soc., 2022, 144, 9280–9291.
31. J. Yang, W. Liu, M. Xu, X. Liu, H. Qi, L. Zhang, X. Yang, S. Niu, D. Zhou, Y. Liu, Y. Su, J.-F. Li, Z.-Q. Tian, W. Zhou, A. Wang and T. Zhang, J. Am. Chem. Soc., 2021, 143, 14530–14539.
32. Y. Zhang, L. Yin, Z. Luo, X. Zhuge, P. Wei, Z. Song and K. Luo, Sustainable Energy Fuels, 2023, 7, 3276–3283.
33. Z. Huang, M. Li, X. Yang, T. Zhang, X. Wang, W. Song, J. Zhang, H. Wang, Y. Chen, J. Ding and W. Hu, J. Am. Chem. Soc., 2024, 146, 24842–24854.
34. S. Shu, T. Song, C. Wang, H. Dai and L. Duan, Angew. Chem., Int. Ed., 2024, 63, e202405650.
35. W. Song, Z. Wen, X. Wang, K. Qian, T. Zhang, H. Wang, J. Ding and W. Hu, Nat. Commun., 2025, 16, 2795.
36. S. Wu, S. Jiang, S.-Q. Liu, X. Tan, N. Chen, J.-L. Luo, S. H. Mushrif, K. Cadien and Z. Li, Energy Environ. Sci., 2023, 16, 3576–3586.
37. N. Li, L. Li, J. Xia, M. Arif, S. Zhou, F. Yin, G. He and H. Chen, J. Mater. Sci. Technol., 2023, 139, 224–231.
38. X. Liu, J. Wu, Z. Luo, P. Liu, Y. Tian, X. Wang and H. Li, ACS Appl. Mater. Interfaces, 2023, 15, 9240–9249.
39. X. Xie, H. Peng, K. Sun, W. Li, A. Liang, G. Ma, Z. Lei and Y. Xu, Adv. Funct. Mater., 2024, 34, 2316037.
40. Y. Cao, J. Zhang, J. Han, G. Li, W. Ma, H. Zheng, Z. Liang and R. Cao, Chem.–Eur. J., 2025, 31, e202501464.
41. X. Li, Y. Zhao, Y. Yang and S. Gao, Nano Energy, 2019, 62, 628–637.
42. M. Zhang, Q. Dai, H. Zheng, M. Chen and L. Dai, Adv. Mater., 2018, 30, 1705431.
43. R. Su, Y. Ma, L. Liu, Q. Wu, D. Fu, Y. Li, H. Lin, X. Wei, M. S. Siddique, J. Chen and X.-L. Wu, J. Cleaner Prod., 2025, 486, 144548.
44. J. Chen, L. Li, L. Yang, C. Chen, S. Wang, Y. Huang and D. Cao, Chin. J. Chem. Eng., 2022, 43, 161–168.
45. A. Rezaei, S. Aber, E. Asghari and S. Elmas, Fuel, 2024, 374, 132508.
46. S. Qiang, J. Chen, S. Huang, H. Xu, X. Zhuo, H. Zhou, A. Yuan, H. Zhou and Y. Qiao, J. Colloid Interface Sci., 2026, 701, 138722.
47. Y. Shen, S. He, Y. Zhuang, S. Huang, C. Meng, A. Yuan, W. Miao and H. Zhou, ACS Appl. Nano Mater., 2023, 6, 16873–16881.
48. W. Wu, F. Gao, T. Shen, Y. Du, W. Guo, J. Mao, J. Ye, Y. Zhan, C. Wang and G. Huang, Chem. Eng. Sci., 2025, 305, 121152.
49. Y. Kong, H. Li, B. Li, D. Li, J. Guo, Q. Zhou, Z. Wang, W. Ma and J. Yuan, J. Energy Chem., 2025, 109, 1–7.
50. J. Choi, S. Im, J. Choi, S. Surendran, D. J. Moon, J. Y. Kim, J. K. Kim and U. Sim, Energy Mater., 2024, 4, 400020.
51. J.-F. Gu, J. Wang, C. Wang, J. Li, C. Chen, N. Zhang, X.-Y. Xu and S. Chaemchuen, J. Colloid Interface Sci., 2025, 684, 159–169.
52. Y. Guan, Y. Li, S. Luo, X. Ren, L. Deng, L. Sun, H. Mi, P. Zhang and J. Liu, Appl. Catal., B, 2019, 256, 117871.
53. F. Guo, Y. He, H. Zeng, H. Liu, D. Yang, H. Chen, H. Li and Y. Liu, Colloids Surf., A, 2022, 648, 129417.
54. Y. Peng, S. Li, M. Wang, X. Xiong, J. Dang, W. Zhang, R. Cao and H. Zheng, J. Colloid Interface Sci., 2024, 658, 518–527.
55. Y. Wang, C. Xie, D. Liu, X. Huang, J. Huo and S. Wang, ACS Appl. Mater. Interfaces, 2016, 8, 18652–18657.
56. X. L. Zhang, X. Zhang, L. C. Zhang, Z. G. Huang, F. Fang, J. J. Hu, Y. X. Yang, M. X. Gao, H. G. Pan and Y. F. Liu, Mater. Today Nano, 2022, 18, 100200.
57. Z. Liang, J. Zhang, W. Suo, H. Zheng, Y. Wang and R. Cao, J. Alloys Compd., 2024, 971, 172775.
58. H. Liu, L. Jiang, Y. Sun, J. Khan, B. Feng, J. Xiao, H. Zhang, H. Xie, L. Li, S. Wang and L. Han, Adv. Energy Mater., 2023, 13, 2301223.
59. S. Ibraheem, S. Chen, J. Li, W. Li, X. Gao, Q. Wang and Z. Wei, ACS Appl. Mater. Interfaces, 2019, 11, 699–705.
60. J. Zhang, Y. Mou, W. Suo, S. Yang, J. Shen, H. Xu, Z. Zeng, R. Zhang, Z. Liang, Y. Wang, H. Zheng, J. Cao and R. Cao, Adv. Funct. Mater., 2025, 35, 2417621.
61. G. Li, Y. Tang, T. Fu, Y. Xiang, Z. Xiong, Y. Si, C. Guo and Z. Jiang, Chem. Eng. J., 2022, 429, 132174.
62. L. Wu, R. Zhao, G. Du, H. Wang, M. Hou, W. Zhang, P. Sun and T. Chen, Green Energy Environ., 2023, 8, 1693–1702.
63. Z. Liang, J. Zhang, H. Zheng and R. Cao, Chem. Commun., 2024, 60, 2216–2219.
64. T. Wang, C. Yang, Y. Liu, M. Yang, X. Li, Y. He, H. Li, H. Chen and Z. Lin, Nano Lett., 2020, 20, 5639–5645.
65. H. Chen, C. Meng, Z. Jiao, A. Yuan and H. Zhou, Energy Fuels, 2025, 39, 4069–4078.
66. C. Yang, H. Jin, C. Cui, J. Li, J. Wang, K. Amine, J. Lu and S. Wang, Nano Energy, 2018, 54, 192–199.
67. K. Im, D. Kim, J.-H. Jang, J. Kim and S. J. Yoo, Appl. Catal., B, 2020, 260, 118192.
68. S. Huang, Y. Shen, A. Li, H. Zhou, Y. Qiao, A. Yuan, H. Zhou and S. Zheng, Int. J. Miner., Metall. Mater., 2025 DOI:10.1007/s12613-025-3190-y.

---