nature.com

Synergy of single atoms and sulfur vacancies for advanced polysulfide–iodide redox flow battery

Abstract

Aqueous redox flow batteries (RFBs) incorporating polysulfide/iodide chemistries have received considerable attention due to their safety, high scalability, and cost-effectiveness. However, the sluggish redox kinetics restricted their output energy efficiency and power density. Here we designed a defective MoS2 nanosheets supported Co single-atom catalyst that accelerated the transformation of S2−/Sx2− and I−/I3− redox couples, hence endow the derived polysulfide–iodide RFB with an initial energy efficiency (EE) of 87.9% and an overpotential of 113 mV with an average EE 80.4% at 20 mA cm−2 and 50% state-of-charge for 50 cycles, and a maximal power density of 95.7 mW cm−2 for an extended cycling life exceeding 850 cycles at 10 mA cm−2 and 10% state-of-charge. In situ experimental and theoretical analyses elucidate that Co single atoms induce the generation of abundant sulfur vacancies in MoS2 via a phase transition process, which synergistically contributed to the enhanced adsorption of reactants and key reaction intermediates and improved charge transfer, resulting in the enhanced RFB performance.

Introduction

The global challenge of energy crises underscores the importance of addressing the massive depletion of fossil fuels. In this case, aqueous redox flow batteries (RFBs) have received great attention due to their high scalability, design flexibility, capability to decouple power and energy, and enhanced safety, showing an excellent suitability for grid integration to alleviate the intermittent and fluctuating nature inherent in renewable energy sources1,2. Notably, the sulfide/polysulfide (S2−/Sx2−) and iodide/triiodide (I−/I3−) redox couples exhibit high solubility and cost-effectiveness3,4; and thus, the polysulfide/iodide–based RFBs (SIRFBs) become an appealing candidate5,6,7. However, the multistep charge transfer reactions within the S2−/Sx2− and I−/I3− couples on electrode result in elevated polarization resistance and poor kinetic reversibility8,9, which induce slow adsorption behavior, limited operational lifespan, and diminished energy efficiency (EE), thereby impeding the widespread adoption of SIRFBs10,11. Therefore, designing efficient and stable catalytic electrodes to accelerate the transformation of S2−/Sx2− and I−/I3− couples remains challenging but crucial for practical deployment of SIRFBs.

Many metal sulfides12,13,14,15,16,17,18, such as CoS2, CoS2@CoS, Cu7S4, NiCo2S4, and CuFeS2, had shown substantial potential applications in SIRFBs to boost the S2−/Sx2− and I−/I3− redox couples with suppressed side reactions. Recently, the two-dimensional molybdenum disulfide (MoS2) has received considerable attention in electrocatalysis and battery fields because of its unique electronic structures and defective properties19,20,21. Particularly, MoS2 was able to improve the polysulfide shuttle process for Li–S batteries22 and accelerate the I−/I3− redox reaction in Zn–I2 RFBs23, respectively; while its application in SIRFB has not been reported yet. Meanwhile, the intrinsic activity of MoS2 still needs to be further improved to meet the high-power demand of industrial RFBs. In recent years, single atom catalysts (SACs) have attracted widespread attention in heterogeneous catalysis and electrocatalysis fields due to the merits of maximal atom utilization, low-coordination environment, and abundant unoccupied orbitals, etc.24,25,26. This concept provides new opportunities for exploring high-performance electrocatalysts for RFBs. The single atoms anchoring onto MoS2 surface can not only introduce new-type and highly active catalytic sites, but also regulate the defective structure and electronic property of MoS2 substrate, which will synergistically optimize the comprehensive performance of SIRFBs. However, the application of SA-doped MoS2 in SIRFBs has not yet been reported, presenting an opportunity to contribute novel insights and address existing gaps in the research field.

Given the above-mentioned considerations, a reasonable design and application of Co SA doped S vacancies-containing MoS2 (CoSA-VS/MoS2) in the SIRFB is firstly proposed. The derived SIRFB achieved an EE of 87.9% at 20 mA cm−2, which is higher than that of the reported CoS2/CoS (71.6%)12, Cu2CoGeS4 (77.2%)13, Cu7S4 (78.5%)14, CuFeS2 (79.6%)18, etc. Besides, the battery exhibited a peak power density of 95.7 mW cm−2 and an average EE of 76.5% at 30 mA cm−2 within 50 cycles, greater than those of the reported Cu7S4/CNT (84.6 mW cm−2, 66%)16 and NiCo2S4@CP (82.4 mW cm−2, 72.4%)17 catalysts. It also demonstrated a cycling life of approximately 850 cycles during continuous operation at 10 mA cm−2 with a 10% state of charge (SOC) and a low overpotential of 113 mV at 20 mA cm−2. Signally, the initial EE of 93.1% could be almost fully recovered after refreshing the electrolytes (200th and 600th cycles). And, more details can be found in Supplementary Table 1. Those suggest the advantage of the designed SIRFB with CoSA-VS/MoS2. The in situ Raman and in situ UV–vis experiments, catalytic tests, and density functional theory (DFT) calculations reveal that the introduction of CoSA to MoS2 stimulated a phase transformation from trigonal prismatic to metallic octahedral and triggered the formation of vacancy defects. Consequently, CoSA and Vs sites synergistically optimized the interface electronic structure, promoted the reactant adsorption capacity, and accelerated the kinetics of S2−/Sx2− and I−/I3− redox couples, simultaneously.

Results

Synthesis and characterization

The CoSA-VS/MoS2 hybrids on graphite felt (GF) were synthesized using a one-step solvothermal in situ growth strategy, following the annealing activation process, as illustrated in Fig. 1a, with detailed information provided in “Methods” section. The scanning electron microscopy (SEM) images of GF, MoS2, and CoSA-VS/MoS2 are depicted in Supplementary Fig. 1 and Fig. 1b, revealing the uniform distribution of nanosheet-like arrays with a lateral size of approximately 350 to 450 nm on the pristine GF surface, which endows fast charge transport and robust mechanical stability. Of note, the introduction of CoSA did not change the morphology of MoS2. Meantime, these nanosheets are interconnected to each other thus forming a porous structure, leading to more active edge sites exposed and effective mass transportability. Further morphological analyses using transmission electron microscopy (TEM), scanning transmission electron microscopy (STEM) and energy-dispersive X-ray spectroscopy (EDX) mapping are presented in Fig. 1c–e, demonstrating a nanosheet-like architecture with Co, Mo, and S evenly distributed, which corroborates well with the SEM results.

Fig. 1: Microstructural characterizations of the CoSA-VS/MoS2 nanosheets.

figure 1

a Scheme diagram for the CoSA-VS/MoS2 synthetic process. b SEM, c TEM, d STEM, e EDX, f, g AC-HAADF-STEM images of CoSA-VS/MoS2 (yellow sphere: S atom, red sphere: Co atom, blue sphere: Mo atom).

Full size image

X-ray powder diffraction (XRD) measurements and Raman characterizations were simultaneously conducted on GF, MoS2, and CoSA-VS/MoS2, and the results are presented in Supplementary Fig. 2. Clearly, MoS2 and CoSA-VS/MoS2 display characteristic diffraction peaks at 14.4° and 35.8°, corresponding to the (002) and (102) planes of pristine MoS2 (PDF#37-1492)27, respectively (Supplementary Fig. 2a and Supplementary Note 1). The weakened (002) peak in CoSA-VS/MoS2 is attributed to the defect enrichment arise from the introduction of Co atoms28,29. Additionally, the Raman spectra reveal the characteristic MoS2 peaks of E2g1 (379 cm−1) and A1g (407 cm−1) in CoSA-VS/MoS2 (Supplementary Fig. 2b). High-resolution TEM (HRTEM) analysis of CoSA-VS/MoS2 (Supplementary Fig. 2c) indicates the (002) plane of 2H-MoS2 (PDF#37-1492) with a lattice spacing of 0.616 nm, conforming the XRD result. Moreover, the spherical aberration-corrected high-angle annular dark-field STEM (AC-HAADF-STEM) is decidedly showed the detailed interfacial with hexagonal lattice structure in the synthesized CoSA-VS/MoS2, and the metallic octahedral 1T-MoS2 appears after Co SA-doping (Fig. 1f, g). This implies that the introduction of Co contributes the arrangement of S atoms and facilitates the phase formation of 1T-MoS230, resulting in the creation of abundant VS to boost the reaction intermediate of S2−/Sx2− and I−/I3− redox couples adsorbing on adjacent S sites, which will be further verified by theoretical analyses. Furthermore, as can be seen from Supplementary Fig. 3, the marked dark spots in MoS2 lattice represent the single Co atoms, as evidenced by the corresponding cross-sectional intensity spectra.

The X-ray photoelectron spectroscopy (XPS) and X-ray absorption fine structure (XAFS) results of CoSA-VS/MoS2 are presented in Fig. 2, which aim to further scrutinize its chemical compositions and bonding states. XPS provides direct evidence of the phase conversion from 2H to 1T after the introduction of Co atoms31. Supplementary Fig. 4a indicates the signals of Co, S, and Mo in the CoSA-VS/MoS2, which agrees well with EDX mapping. In addition, the signal of Co was absent in MoS2 (Supplementary Fig. 4b). The atomic ratio of Mo and Co species in the CoSA-VS/MoS2 is determined to be approximately 34:1 (Supplementary Table 2). The Co 2p spectrum is deconvoluted into two peaks corresponding to Co2+ 2p3/2 (781.3 eV) and Co2+ 2p1/2 (797.4 eV), and two satellite peaks at 786.7 and 803.6 eV (Fig. 2a)32. In the Mo 3d XPS spectrum of CoSA-VS/MoS2 (Fig. 2b), two distinctive characteristic peaks at 229.9 and 233.2 eV correspond to Mo 3d5/2 and Mo 3d3/2, respectively, indicating the presence of Mo4+ (2H-MoS2)33. Concurrently, two predominant peaks at 229.3 and 232.6 eV can be indexed to the Mo4+ (1T-MoS2)33. In addition, two peaks at 234 and 236 eV are assigned to Mo6+, indicative of the formation of VS34; and a peak at 226.6 eV originates from the S 2s signal27,35. Compared to MoS2, a notable negative shift is observed in the CoSA-VS/MoS2, further verifying the generation of 1T-MoS2 by SA-doping31,36. Importantly, this can be further explained by theoretical calculations (Supplementary Figs. 5 and 6). Regarding the S 2p spectrum of CoSA-VS/MoS2 (Fig. 2c), two peaks at 162.0 and 163.1 eV are assigned to S 2p3/2 and S 2p1/2 of 1T MoS237, respectively. Besides, two peaks at 162.5 and 163.8 eV, which are assigned to S 2p3/2 and S 2p1/2, respectively, represent 2H-MoS237. Similar negative shift in S 2p spectrum of the CoSA-VS/MoS2 are attributed to the introduction of Co atoms, which triggers the formation of VS and the lattice distortion, and thus evokes changes in bond energy27. Overall, the anchoring of Co atoms can effectively regulate the atom arrangement of MoS2, thereby resulting in VS and phase transformation, which is in accordance with AC-HAADF-STEM and XRD analyses.

Fig. 2: Analysis of XPS and XAFS spectra.

figure 2

a Co 2p XPS spectra of CoSA-VS/MoS2. b, c Mo 3d and S 2p XPS spectra of CoSA-VS/MoS2 and MoS2. d, e XANES spectra extracted from the Co K-edge and the corresponding FT-EXAFS at the R space of CoSA-VS/MoS2 and the reference samples. f FT-EXAFS fitting curve at the R space of CoSA-VS/MoS2 at the Co K-edge (yellow sphere: S atom, red sphere: Co atom, blue sphere: Mo atom). g Wavelet transform contour plots of CoSA-VS/MoS2 and the reference samples.

Full size image

XAFS analyses were further conducted to obtain deeper insights into the electronic structure and coordination characteristics of Co species in CoSA-VS/MoS2. Figure 2d illustrates the Co K-edge X-ray absorption near-edge structure (XANES) spectrum of CoSA-VS/MoS2, along with the reference spectra of CoS2, CoO, and Co foil. The absorption edge of CoSA-VS/MoS2 approaches to that of CoO, indicative of positive charged Co atoms with an average valence state approximately +2, consistent with the XPS results38,39,40. Coordination environment of Co atoms in CoSA-VS/MoS2 is further elucidated through Fourier-transformed k-weighted extended X-ray absorption fine structure (FT-EXAFS) spectra (Fig. 2e), where the prominent peak at 1.70 Å can be attributed to the Co–S bond, slightly lower than that of CoS2 (1.81 Å). This is due to the existence of VS around CoSA in CoSA-VS/MoS241, which is evidenced by the fitting results (Fig. 2f). Clearly, no characteristic Co–Co bond is observed in CoSA-VS/MoS2 (Fig. 2e), confirming that Co atoms are isolated dispersed. In addition, quantitative fitting results also identify that the CoSA is coordinated to approximately five S atoms, while one S atom loss thus forming the VS (Fig. 2f and Supplementary Table 3), which is in accordance with the previous report41 and XPS analyses. Meanwhile, the wavelet transform contour plot of CoSA-VS/MoS2 reveals a maximum intensity at ~1.70 Å (Fig. 2g), corresponding to Co–S, further conforms the FT-EXAFS results38,42. These findings provide compelling evidence of the isolated dispersion of Co atoms and the formation of VS in CoSA-VS/MoS2.

Adsorption and electrocatalytic activity

The adsorption behavior of Sx2− and I3− redox species plays a pivotal role in the electrocatalytic process (Supplementary Note 2). In this case, the combination of Co SA and VS can optimize the electronic structure of CoSA-VS/MoS2, and thereby promoting the adsorption and conversion kinetics of reactants and key intermediates31,34,43,44. Accordingly, the ultraviolet-visible (UV–vis) absorption spectra and calibration curve between absorption and solution concentration of Sx2− and I3− with embedding GF, MoS2, and CoSA-VS/MoS2 electrode are shown in Fig. 3a, b, and the analyses were determined in the H-type cell with static diffusion tests via monitored the electrolyte side after 12 h (Fig. 3c, d). Compared to GF and MoS2, CoSA-VS/MoS2 exhibits more pronounced adsorption capacity toward I3− (Fig. 3a and Supplementary Table 4). This is further validated by the UV–vis absorption spectra where CoSA-VS/MoS2 shows a significant weaker peak intensity at 290 and 352 nm associated with I3− ions in comparison with the counterparts, in line with calibration curve between absorption intensity and solution concentration (Fig. 3a). Similarly, CoSA-VS/MoS2 displays a higher absorptivity for S2− (260 nm), S22− (310 nm) and S42−/S22− (370 nm) anions than that of GF and MoS2 (Fig. 3b)4,5, further affirmed by the recorded UV–vis spectra. Besides, the adsorption and permeability phenomena of I3− and Sx2− redox species are intuitively observed in the H-type cells with static diffusion tests (Fig. 3c, d), respectively, which utilized the Nafion membranes (SEN-G115 and SEN-K117) to separate the deionized water chamber (colorless) and electrolyte chamber (yellow). Clearly, the CoSA-VS/MoS2 and MoS2 electrode in the I3− electrolyte have stronger adsorption capacity and restraint for I3− shuttle than GF after 12 h resting, while in the Sx2− electrolyte after resting for 12 h, CoSA-VS/MoS2 electrode exhibits relatively better adsorption capacity for Sx2− redox species than those of MoS2 and GF.

Fig. 3: Adsorption and cyclic voltammetry characterizations.

figure 3

a, b UV–vis spectra of the I3− and S22− solutions with GF, MoS2, and CoSA-VS/MoS2 impregnation for 12 h, and the inset is the calibration curve between absorption intensity and concentration (at 25 ± 2 °C, in the environmental atmosphere). c, d Digital images of I3− and Sx2− diffusion and adsorption behavior in I3− and S22− solution with GF, MoS2, and CoSA-VS/MoS2 impregnation for 12 h (left chamber: deionized water; right chamber: 5.2 mM NaI + 1.3 mM I2 or 12 mM Na2S2). e, f CV curves of the catalysts recorded in 0.1 M NaI3 + 0.5 M NaCl solution and 0.06 M Na2S + 0.02 M S + 0.5 M NaCl solution, respectively, at a scan rate of 50 mV s−1.

Full size image

Then, the electrocatalytic activities of CoSA-VS/MoS2 toward S2−/Sx2− and I−/I3− redox reactions were investigated by cyclic voltammetry (CV) measurements. The pair of peaks corresponding to the redox reactions of I−/I3− couples exhibit a peak-to-peak separation (Epp) of 0.24 V (Fig. 3e and Supplementary Table 5), lower than those of MoS2 (0.29 V) and GF (0.32 V). This suggests that CoSA-VS/MoS2 owns a higher activity towards the I−/I3− conversion as compared to MoS2 and GF. Meanwhile, CoSA-VS/MoS2 possesses a larger peak current density compared to GF and MoS2 (Fig. 3e)45. This further signifies that CoSA-VS/MoS2 is active for the redox reaction of I−/I3−. Furthermore, the JOx/|JRed| ratio of 0.68 for CoSA-VS/MoS2 indicates its better reversibility to I−/I3– redox reaction than those of MoS2 (0.61) and GF (0.63). Besides, CoSA-VS/MoS2 significantly enhances the electrochemical reactivity of the polysulfide couple with prominent redox peaks (Fig. 3f). The three reduction peaks imply that polysulfide may be transformed to Sx2− (x = 2–8) species on CoSA-VS/MoS246,47. In addition, the Nyquist plots indicate that CoSA-VS/MoS2 possesses an improved S2−/Sx2− conversion kinetics (Supplementary Fig. 7, Supplementary Table 6 and Supplementary Note 3). The above findings reveal that CoSA-VS/MoS2 can effectively catalyze the transformation of S2−/Sx2− and I−/I3− couples. Besides, the diffusion-controlled S2−/Sx2− and I−/I3− redox reaction processes on CoSA-VS/MoS2 surface were evidenced by their linear relationship between the peak currents and the square root of the scan rate (ν1/2, Supplementary Figs. 8 and 9). Additionally, according to the Berzins–Delahay equation, the order of the calculated I− diffusion coefficient (D) follows: GF < MoS2 < CoSA-VS/MoS2, in accordance with the battery performances. Furthermore, CoSA-VS/MoS2 presents stable activities toward I−/I3− and S2−/Sx2− redox reactions over 100 cycles (Supplementary Fig. 10).

The reaction mechanism of S2−/Sx2− and I−/I3− conversion was further monitored by in situ UV–vis experiments (Supplementary Figs. 11 and 12). The characteristic peaks corresponding S42−/S22− (362 nm) and S22− (331 nm) are detected (Supplementary Fig. 12a), evidencing that the occurrence of S2− ↔ Sx2− conversion4. Of note, the intensity of S42−/S22− and S42− reduces during the oxidation process and increases at the reduction reaction (Supplementary Fig. 12b). Similarly, the characteristic peaks of I3− (288 and 354 nm) are observed (Supplementary Fig. 12c), indicating the polyiodides response in the electrolyte for I− ↔ I3− conversion48,49; while the intensity of I3− increases during the oxidation process and reduces at the reduction reaction (Supplementary Fig. 12d).

Density functional theory (DFT) calculations

To reveal the catalytic enhancement of CoSA-VS/MoS2, DFT simulations were performed (Supplementary Note 4). The total density of states (DOS) of two models evidence that CoSA and VS collectively induce spin polarization states (Fig. 4a)50, resulting in higher conductivity and affinity toward intermediates compared with MoS251,52,53. This guarantees a fast charge transfer of redox reactions, accounting for enhanced catalytic activity of CoSA-VS/MoS254,55. Further, CoSA-VS/MoS2 has lower adsorption energy toward I−, I2, I3−, S2−, S22−, S42−, S62−, and S82− intermediates as compared to those of MoS2 (Fig. 4b, Supplementary Fig. 13 and Supplementary Data 1), demonstrating that CoSA-VS/MoS2 present a higher affinity to those reaction intermediates (Supplementary Table 7), which is in line with the adsorption behaviors and UV–vis analyses (Fig. 3a, b).

Fig. 4: Theoretical analyses of CoSA-VS/MoS2 and MoS2.

figure 4

a Calculated DOS of CoSA-VS/MoS2 and MoS2, with aligned Fermi level. b The calculated adsorption energy of I-containing and S-containing species on catalytic surface. c The optimized intermediate adsorption configurations for CoSA-VS/MoS2. d, e Free-energy diagrams of I and S intermediates.

Full size image

The transformation of I−/I3− can be divided into four steps, while the S2−/Sx2− couple includes six steps, as illustrated in Fig. 4c, Supplementary Fig. 14 and Supplementary Data 1; where the elementary reaction of * → *I−, *I2 → *I3−, and *S2− → *S22− are the corresponding potential-determining steps (PDSs). Clearly, CoSA-VS/MoS2 shows PDS energy barrier of 0.14 eV in the elementary reaction of * → *I− (Fig. 4d), significantly lower than MoS2 (0.54 eV) in the elementary reaction of *I2 → *I3−, implying an energy-favorable process. Interestingly, it is found that the PDS energy barrier corresponds to S2−/Sx2− on CoSA-VS/MoS2 is 0.38 eV (Fig. 4e), smaller than that on MoS2 (0.68 eV), leading to the high performance of CoSA-VS/MoS2 towards polysulfide species transformation (Fig. 4c). Similarly, the DFT calculation of Co doped in MoS2 without VS (CoSA/MoS2) and pristine VS in MoS2 (VS/MoS2) models also have been implemented (Supplementary Fig. 15), the corresponding PDSs energy barrier of I−/I3− (0.34 eV) and S2−/Sx2− (0.48 eV) redox in CoSA/MoS2, and the PDSs energy barrier of I−/I3− (0.46 eV) and S2−/Sx2− (0.45 eV) redox in VS/MoS2 are all higher than those of CoSA-VS/MoS2 and below the MoS2. Therefore, both experimental results and DFT simulations confirm that CoSA sites effectively promote the activity of I−/I3− and S2−/Sx2− redox, and thus favors the SIRFB performance.

Redox flow battery performance

A schematic representation of the proposed SIRFB configuration is depicted in Fig. 5a. SIRFB with CoSA-VS/MoS2 as anodic and cathodic catalyst was assembled, and the 2.0 M Na2S2 and 2.0 M NaI + 0.5 M I2 were adopted as anolyte and catholyte, respectively (Supplementary Fig. 16). A two-layer separator consisting of G115 and K117 was used to enhance Na+ cation transport and restrict the crossover of active species56. Electrochemical impedance spectroscopy (EIS) spectra indicate that the charge transfer resistance (Rct) on CoSA-VS/MoS2 is 0.76 Ω cm−2 (Fig. 5b and Supplementary Fig. 17), smaller than that on GF (18.81 Ω cm−2) and MoS2 (0.85 Ω cm−2). Of note, the solution resistance (Rs) on CoSA-VS/MoS2 is 3.42 Ω cm−2 (Supplementary Table 8), lower than those on GF (4.67 Ω cm−2) and MoS2 (3.88 Ω cm−2). These results evidence that CoSA-VS/MoS2 is capable of facilitating charge transfer and promoting diffusion dynamics12, which is consistent with the CV experiments and DOS analyses. Figure 5c and Supplementary Fig. 18 depict the first charge and discharge curves of SIRFBs recorded at 15 mA cm−2 associated with 100% SOC. Evidently, the SIRFB with CoSA-VS/MoS2 exhibits the highest capacity of 131.2 mAh, surpassing those of MoS2 (117 mAh) and GF (93.1 mAh). Besides, the first charge and discharge curves extracted at a current density of 20 mA cm−2 with discharge to 50% SOC (Fig. 5d) demonstrate that only a small polarization voltage difference of 113 mV is observed on CoSA-VS/MoS2, distinctly lower than those of MoS2 (211 mV) and GF (640 mV). The small overpotential signifies that the water splitting is effectively suppressed to avoid the generation of O2/H2, and the reaction kinetics is greatly accelerated as well57.

Fig. 5: Electrochemical performance of aqueous SIRFBs with CoSA-VS/MoS2.

figure 5

a Scheme of the proposed SIRFB configuration. b EIS plots of the SIRFBs with GF, MoS2, and CoSA-VS/MoS2. c, d The first charge and discharge curves recorded at 15 mA cm−2 with discharged to 0.1 V (voltage cutoff) or 100% SOC (capacity cutoff) and then charged to 100% SOC (capacity cutoff) and 20 mA cm−2 with discharge to 50% SOC, respectively. e VE plots at different operational current densities with 50% SOC. f Discharge curves and the corresponding power densities with 50% SOC. g Charge and discharge polarization curves obtained at different current densities with 50% SOC. h The EE taken from galvanostatic cycling test at 20 mA cm−2 with 50% SOC.

Full size image

Further, in situ Raman spectrometry was conducted to monitor the S2−/Sx2− and I−/I3− conversion on CoSA-VS/MoS2 in SIRFB. At the discharging progress, the intensity of S22− peak (489 cm−1) reduces while the intensity of S42− peak increases accompanied by the peak of S42− (193 cm−1) appears58; then those (S42−) peak intensities show opposite trend in the charging progress (Supplementary Fig. 19a, b). Supplementary Fig. 19c, d presents the peak intensity of I3− (113 cm−1) reduces in the discharging progress and increases in the charging progress59. These findings conform to the in situ UV–vis results.

The EE of CoSA-VS/MoS2 derived SIRFB estimated from the first charge and discharge curve is 87.9% (Fig. 5d), larger than those of MoS2 (77.3%) and GF (61.2%). Further, the voltage efficiency (VE) was calculated according to the equation (VE = EE/CE, CE represents coulombic efficiency) and the results are depicted in Fig. 5e and Supplementary Figs. 20–22. Notably, CoSA-VS/MoS2 shows higher VE values over a wide current density range as compared to GF and MoS2 (Fig. 5e), conforming the change trend of CE and EE (Supplementary Figs. 21 and 22). The CE consistently remains approximately 100% at evaluated current densities (Supplementary Fig. 22 and Supplementary Note 5). Meanwhile, the calculated EE of CoSA-VS/MoS2 declines from 89.9% at 15 mA cm−2 to 55.4% at 100 mA cm−2, attributing to the increased overpotential at larger current densities. The power density of SIRFB with CoSA-VS/MoS2 reaches as high as 95.7 mW cm−2 (Fig. 5f), outperforming those of GF (69.1 mW cm−2) and MoS2 (83.5 mW cm−2). The open-circuit voltages of the SIRFB with GF, MoS2, and CoSA-VS/MoS2 are close to the theoretical value of 1.02 V12 (Supplementary Fig. 23). In addition, the voltage gap between the charge and discharge curves of the SIRFB with CoSA-VS/MoS2 is smaller than those observed in other SIRFBs under the same current density (Fig. 5g), suggesting a better redox reversibility.

Figure 5h illustrates the galvanostatic cycling performance of the SIRFBs operating at 20 mA cm−2 with 50% SOC, which confirms that CoSA-VS/MoS2 possesses higher EEs relative to others at series of current densities, originating from the reduced charge/discharge overpotentials (Supplementary Fig. 24). Importantly, the SIRFB using CoSA-VS/MoS2 achieves an EE of 78.0% at 20 mA cm−2 after 50 cycles, corresponding to 88.7% retention. At the same cases, the observed EE and retention for GF and MoS2 are 61.3% and 77.3%, 51.7% and 49.4%, respectively, both of them are inferior to CoSA-VS/MoS2. Similarly, the VE and CE of CoSA-VS/MoS2 are above 78.6% and 89.6% after 50 cycles at 20 and 30 mA cm−2 (Supplementary Figs. 25 and 26). Especially, the average EE of CoSA-VS/MoS2 reaches 76.5% at a high current density of 30 mA cm−2, higher than those of Cu7S4/CNT16 and NiCo2S417. Additionally, the EE retention is 87.1% at 30 mA cm−2 after 50 cycles (Supplementary Fig. 26).

To evaluate the durability of CoSA-VS/MoS2 after the galvanostatic cycling test, the I-side electrodes were examined (Supplementary Fig. 27), and the results reveal that yellow substance is clearly observed on GF and MoS2. By contrast, there is no I2 residue exist on the surface of CoSA-VS/MoS2, corroborating that CoSA-VS/MoS2 exhibits better activity towards I−/I3− transformation and thus inhibits the I2 production. The Co content in sample CoSA-VS/MoS2 is determined to be 1.62 wt% according to the inductively coupled plasma atomic emission spectrometry (ICP-OES); while the Co content is found to be 1.59 wt% after 50 cycles at 20 mA cm−2 with the 50% SOC, indicating that scarcely any Co is dissolved during the cycling test. Additional SEM images and XPS analyses of CoSA-VS/MoS2 after 50 cycles show a tiny variation for both negative electrode and positive electrode, indicative of its high structural and chemical stability (Supplementary Figs. 28, 29 and Supplementary Note 6). Moreover, the SIRFB using CoSA-VS/MoS2 can work stably exceed 850 cycles with an EE retention of 62.0% at a 10 mA cm−2 charge/discharge current density with 10% SOC (Supplementary Fig. 30 and Supplementary Note 7), which is greater than those of GF and MoS2. Meanwhile, the initial improved performance of GF is due to the increased roughness and active sites during charge/discharge processes (Supplementary Fig. 31 and Supplementary Note 8), while the activity decay occurred because of the precipitation of I2, enveloping GF and further reduced the reaction area of the electrode60. In addition, CoSA-VS/MoS2 gives an EE of 93.1% when operates at 10 mA cm−2, which is superior to the recently best-reported results (Supplementary Table 1). The EE values of the SIRFB based on CoSA-VS/MoS2 electrode with increasing current densities (from 10 to 100 mA cm−2) at the 50% SOC are fall between 56 and 93% (Supplementary Fig. 32). Overall, the SIRFB with CoSA-VS/MoS2 shows competitive advantage in terms of power density, EE, CE, and VE as compared with those of other metal sulfides12,13,14 used in SIRFBs (Supplementary Fig. 33 and Supplementary Table 1). Meanwhile, this obtained SIRFB with CoSA-VS/MoS2 also show competitive advantages as compared to the vanadium/zinc-bromine redox batteries61,62.

Besides, we prepared defective MoS2 (H-VS/MoS2, annealed in H2 atmosphere) to assess the synergistic effect of CoSA sites and abundant VS (Supplementary Fig. 34 and Supplementary Note 9). As expected, the performance of SIRFB with H-VS/MoS2 falls between that of MoS2 and CoSA-VS/MoS2, strongly confirming that the combination of CoSA sites and abundant VS contributes SIRFB enhancement (Supplementary Fig. 35, Supplementary Table 9 and Supplementary Note 10). These findings further emphasize the vital role of single-atom doping in governing the S2−/Sx2− and I−/I3− redox kinetic performance.

Discussion

In this study, we fabricated the self-supporting CoSA-VS/MoS2 nanosheets on GF substrate by introducing CoSA sites into MoS2 lattice, associating with the formation of VS and triggering metallic 1T phase transformation. The SIRFB employing CoSA-VS/MoS2 demonstrates an EE and a low overpotential of 87.9% and 113 mV at 20 mA cm−2, respectively, accompanying with EE retention remains stable above 88.7% after 50 cycles. Moreover, the SIRFB delivers a power density of 95.7 mW cm−2 along with a long operational life of approximately 850 cycles at 10 mA cm−2 with 10% SOC. The in situ Raman, in situ UV–vis experiments and DFT calculations indicate that CoSA-VS/MoS2 memorably enhances the electrocatalytic activities toward the S2−/Sx2− and I−/I3− redox reactions by optimizing the adsorption behaviors of intermediates and reducing the energy barriers. This work presents ingenious and resultful tools to explore high performance single-atom doped MoS2 for sundry advanced SIRFBs application, and more memorably, gains insights into the polysulfide/iodide chemistries.

Methods

Materials

Cobalt nitrate (Co(NO3)2 ∙ 6H2O, ＞99.5%) and sodium sulfide nonahydrate (Na2S ∙ 9H2O, 98%) were received from Aladdin. Thiourea (CH4N2S, ≥99.0%), ammonium molybdate ((NH4)6Mo7O24 ∙ 4H2O, ≥99.0%), sodium iodide (NaI, 99.5%), sodium chloride (NaCl, ≥99.5%), sublimed sulfur (S, AR), and sulfuric acid (H2SO4, 95.0～98.0%) were purchased from Sinopharm Chemical Reagent Co., Ltd. Iodine (I2, 99.8%) was purchased from Shanghai Titan Scientific Co., Ltd. The Nafion membrane (SEN-G115 (thickness: 127 μm) was used in S side and the Nafion membrane SEN-K117 (thickness: 175 μm)) was used in I side, both of them were obtained from Suzhou Sinero Technology Co., Ltd., and their lateral dimension is 10*10 cm. GF was from Beijing Jinglong Carbon Technology Co., Ltd. The stoichiometric ratios of S and Na2S were mixed at 25 ± 2 °C to yield sodium polysulfide (Na2Sx) electrolytes. Likewise, sodium triiodide (NaI3) electrolytes were obtained according to the addition of stoichiometric ratios of NaI and I2 at 25 ± 2 °C.

Synthesis of electrode materials

The preparation of the CoSA-VS/MoS2 involved the initial treatment of the commercially obtained GF, which had immersed in 6 M H2SO4 for approximately 12 h, followed by thorough washing with deionized (DI) water until the flushing water reached a near-neutral (pH ≈ 7). Subsequently, the treated GF was heat treated (300 °C, 2 h). The CoSA-VS/MoS2 was synthesized using an in situ one-step hydrothermal method. In this procedure, a mixture containing 0.1765 g of (NH4)6Mo7O24 ∙ 4H2O, 0.228 g of CH4N2S, and 0.0291 g of Co(NO3)2 ∙ 6H2O dissolved in 35 mL of DI water was uniformly dispersed. The resultant mixture was transferred to a 50-mL Teflon-lined stainless-steel reactor, where the GF (1.3 cm × 1.3 cm × 3.0 mm) was immersed in the precursor solution. The hydrothermal treatment was performed at 200 °C for 20 h, yielding a black product. Afterward, the product was washed with DI water and freeze dried for 24 h, and then it was annealed at 400 °C for 2 h under a nitrogen (N2) atmosphere. Subsequently, the CoSA-VS/MoS2-loaded GF underwent ultrasonic cleaning with water until no discernible CoSA-VS/MoS2 particles were observed in the wastewater stream. The final product was then freeze dried for 24 h. The MoS2 electrode was synthesized using a similar approach, except for the addition of a cobalt source.

Electrochemical measurements

CV tests were executed in a three-electrode configuration on a CHI 760E electrochemical workstation with the Ag/AgCl electrode and Pt sheet worked as the reference electrode and the counter electrode, respectively. The working electrodes was established by the catalysts (CoSA-VS/MoS2, 10 mg) dispersed in the mixed solution that included 500 μL of DI water, 500 μL of isopropanol, and 20 μL of 5 wt% Nafion, forming a homogeneous ink. Thereafter, 10 μL of the formed ink (catalyst-mixed solution) was loaded on a glassy carbon electrode (diameter of 3 mm), and the electrode was dried at 25 ± 2 °C. The anolyte was 0.02 M S + 0.06 M Na2S + 0.5 M NaCl; the catholyte was 0.1 M NaI3 + 0.5 M NaCl, respectively. The EIS measurement was performed at a frequency range of 0.01 Hz to 1 MHz (5 mV ac oscillation), operate on the open-circuit voltage stable after 60 s, potentiostatic and the number of data points is 12 (per decade of frequency).

Ex situ test

GF, MoS2, and CoSA-VS/MoS2 of equivalent mass were soaked in a 5 mL containing iodide or polysulfide aqueous solution with 5.2 mM NaI + 1.3 mM I2 or 12 mM Na2S2 in a side of H-type cells (another side filled in 5 ml DI water), separately, and the admixtures were adequately resting reaction for 12 h to achieve thorough adsorption. The electrolytes were compared and recorded by UV–vis spectra test. The post-electrolysis electrode was washed by DI water and then dried; after that, the electrode was characterized by SEM, XPS and Raman tests. All preparation and testing process under ambient atmosphere at 25 ± 2 °C.

In situ test

The in situ UV–vis spectrometry measurements were conducted using CoSA-VS/MoS2 electrode performed in a single cell at a scan rate of 10 mV s−1 from 300 to 500 nm (–1.0 V ~ 0.1 V) or from 250 to 400 nm (0.3 V ~ 0.8 V), and the aqueous solution of 0.02 M S + 0.06 M Na2S + 0.5 M NaCl or 0.1 M NaI3 + 0.5 M NaCl, respectively. In situ Raman spectrometry measurements were conducted using CoSA-VS/MoS2 electrode performed in SIRFB (at 15 mA cm−2 with 100% SOC) from 100 to 600 cm−1 or from 70 to 210 cm−1, and the electrolyte volume was 5 mL for anolyte (2.0 M Na2S2) and catholyte (2.0 M NaI + 0.5 M I2), respectively. All preparation and testing process under ambient atmosphere at 25 ± 2 °C.

Assembly of the aqueous SIRFB

This study used the configuration of an aqueous SIRFB (Supplementary Fig. 16). The flow cell pack was procured from Taizhou TiSing Technology Co., Ltd., which resembled previous research reports3,5,12. The flow cell featured interdigitated flow fields with a geometric active area of 3.24 cm−2 (1.8 cm × 1.8 cm). Commercially available Nafion membranes (G115 and K117) were used as separators. The Nafion membranes were treated with 5% H2O2 at 80 °C for 1 h, followed by immersion in 5% H2SO4 at 80 °C for an additional 1 h. Subsequently, a 1 M NaOH aqueous solution transformed the H-type (proton conductive) Nafion membranes to Na-type (Na+ conductive) at 80 °C for 2 h. After each step, the membranes were rinsed with DI water for 30 min to eliminate residual chemicals. GF, MoS2, and CoSA-VS/MoS2 electrodes were used as positive and negative electrodes. The electrolytes were prepared at 25 ± 2 °C under continuous nitrogen gas bubbling for 30 min to prevent oxidation before use. Two half-cell bodies, incorporating the GF, MoS2, and CoSA-VS/MoS2 electrodes as positive and negative electrodes, with the Nafion membrane (G115 and K117) in between, were assembled in ambient air. A peristaltic pump (BT100LC, Baoding Rongbai Precision Pump Manufacturing Co., Ltd.) was used to facilitate the circulation of electrolytes through the flow cell and reservoirs. A silicon tube (with an inner diameter of 3.1 mm) was also used to circulate the electrolyte throughout the system.

Flow cell tests

The SIRFB cells were galvanostatically characterized using a battery testing system (NEWARE, Shenzhen Electronics Co., Ltd). The electrolyte volume was 5 mL for catholyte (2.0 M NaI + 0.5 M I2) and anolyte (2.0 M Na2S2), and controlled the electrolyte flow rate at 10 mL min−1 for all experiments. The theoretical capacity is a nominal capacity of 134 mAh, which was calculated by the catholyte (the iodide part, 5 mL of 2.0 M NaI + 0.5 M I2). The capacity-limiting of the full cell was a controlling factor for charge process, and set up the galvanostatically discharged to 0.1 V (voltage cutoff) or 50% SOC and charged to 50% SOC (capacity cutoff) for the flow cell. Note that the charge process of Fig. 5c was set to 100% SOC (capacity cutoff) after discharged to 0.1 V (voltage cutoff) or 100% SOC (capacity cutoff). In addition, the flow cell was operated at current densities ranging from 10 to 100 mA cm−2. During the cycle test, the N2 had filled in electrolytes. The extended cycling experiment was conducted at 10, 20, and 30 mA cm−2.

Characterization

XRD patterns were obtained using a Rigaku D/max 2500 diffractometer with Cu Kα radiation (λ = 0.15418 nm) in the 2θ range of 10°–80° with a scan rate of 10° min−1. Field-emission gun SEM (FEG-SEM) was performed using the Verios 460 L FEI instrument for observation. TEM and mapping images were obtained using an FEI Talos F200X S/TEM with an FEG. X-ray absorption fine structure (XAFS) measurements were performed at Beamline 8-ID, National Synchrotron Light Source II, Brookhaven National Laboratory. Co K-edge absorption spectra were acquired using a cryogenically cooled double crystal Si (1111) monochromator. The spectra were recorded in the fluorescence mode of the monochromator equipped with a passivated implanted planar silicon detector. XPS experiments were performed on a Kratos AXIS Ultra DLD system with Al Kα radiation as the X-ray source. Raman spectroscopy was performed using an HR confocal micro-Raman spectrometer (HORIBA EVOLUTION). The UV–vis absorption spectrum was obtained using a UV-1800PC spectrophotometer (Shanghai Jinghua Technology Instrument Co., Ltd.). The Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES) was executed by Thermo Fisher ICPA 6300.

Computational methods

All DFT calculations were performed using the Vienna Ab initio Simulation Package. The Perdew–Burke–Ernzerhof functional with generalized gradient approximation was used to handle the exchange–correlation interactions among electrons. The plane-wave cutoff energy was set to 450 eV, and the convergence criteria for energy and forces were set to 10−5 eV and 0.02 eV Å−1, respectively. Surface calculations of CoSA-VS/MoS2 (002) and MoS2 (002) were performed using a 5 × 5 × 1 supercell model, which allowed for the relaxation of all atoms during calculations. A vacuum layer with a thickness of 15 Å in the z-direction was included to prevent interactions among different cells. A 2 × 2 × 1 Monkhorst–Pack k-point grid was also used for the aforementioned structures. In addition, for H2S, H2, I2, and Na2S molecules, calculations were performed using only the gamma point in a 10 × 10 × 10-Å box. VASPKIT software was utilized for the preprocessing and postprocessing of the computational data.

Data availability

Source data are provided with this paper.

References

Chao, D. et al. Roadmap for advanced aqueous batteries: from design of materials to applications. Sci. Adv. 6, eaba4098 (2020).

ADSCASPubMedPubMed CentralGoogle Scholar

Yang, Z. et al. Alkaline benzoquinone aqueous flow battery for large-scale storage of electrical energy. Adv. Energy Mater. 8, 1702056 (2018).

Google Scholar

Su, L., Badel, A. F., Cao, C., Hinricher, J. J. & Brushett, F. R. Toward an inexpensive aqueous polysulfide–polyiodide redox flow battery. Ind. Eng. Chem. Res. 56, 9783–9792 (2017).

CASGoogle Scholar

Chen, Q. et al. High-reversibility sulfur anode for advanced aqueous battery. Adv. Mater. 36, 2309038 (2024).

CASGoogle Scholar

Li, Z., Weng, G., Zou, Q., Cong, G. & Lu, Y.-C. A high-energy and low-cost polysulfide/iodide redox flow battery. Nano Energy 30, 283–292 (2016).

CASGoogle Scholar

Zhang, S. et al. Recent progress in polysulfide redox-flow batteries. Batter. Supercaps 2, 627–637 (2019).

CASMATHGoogle Scholar

Lei, J. et al. An active and durable molecular catalyst for aqueous polysulfide-based redox flow batteries. Nat. Energy 8, 1355–1364 (2023).

ADSCASMATHGoogle Scholar

Castañeda, L. F., Walsh, F. C., Nava, J. L. & Ponce de León, C. Graphite felt as a versatile electrode material: properties, reaction environment, performance and applications. Electrochim. Acta 258, 1115–1139 (2017).

Google Scholar

Liu, W., Lu, W., Zhang, H. & Li, X. Aqueous flow batteries: research and development. Chem. Eur. J. 25, 1649–1664 (2019).

CASPubMedMATHGoogle Scholar

Li, Z. & Lu, Y.-C. Material design of aqueous redox flow batteries: fundamental challenges and mitigation strategies. Adv. Mater. 32, 2002132 (2020).

CASGoogle Scholar

Lu, G. et al. Cathode materials for halide-based aqueous redox flow batteries: recent progress and future perspectives. Nanoscale 15, 4250–4260 (2023).

CASPubMedMATHGoogle Scholar

Ma, D. et al. Highly active nanostructured CoS2/CoS heterojunction electrocatalysts for aqueous polysulfide/iodide redox flow batteries. Nat. Commun. 10, 3367 (2019).

ADSPubMedPubMed CentralMATHGoogle Scholar

Zhu, Y. et al. Cu2CoGeS4 nanocrystals for high performance aqueous polysulfide/iodide redox flow batteries: enhanced selectively towards the electrocatalytic conversion of polysulfides. Sustain. Energy Fuels 4, 2892–2899 (2020).

CASMATHGoogle Scholar

Zai, J. et al. Sandwiched Cu7S4@graphite felt electrode for high performance aqueous polysulfide/iodide redox flow batteries: enhanced cycling stability and electrocatalytic dynamics of polysulfides. Mater. Chem. Phys. 250, 123143 (2020).

CASMATHGoogle Scholar

Gao, M. et al. Successive ionic layer adsorption and reaction–deposited copper sulfide electrocatalyst for high-power polysulfide-based aqueous flow batteries. Mater. Today Energy 18, 100540 (2020).

CASGoogle Scholar

Qin, Y., Li, X., Liu, W. & Lei, X. High-performance aqueous polysulfide-iodide flow battery realized by an efficient bifunctional catalyst based on copper sulfide. Mater. Today Energy 21, 100746 (2021).

CASGoogle Scholar

Liu, J. et al. A highly-stable bifunctional NiCo2S4 nanoarray@carbon paper electrode for aqueous polysulfide/iodide redox flow battery. J. Power Sources 561, 232607 (2023).

CASGoogle Scholar

Tsega, T. T., Zai, J., Lai, C., Li, X. & Qian, X. Earth-abundant CuFeS2 nanocrystals@graphite felt electrode for high performance aqueous polysulfide/iodide redox flow batteries. Chin. J. Struct. Chem. 43, 100192 (2024).

CASGoogle Scholar

Wang, H.-E. et al. Sulfur-deficient MoS2 grown inside hollow mesoporous carbon as a functional polysulfide mediator. J. Mater. Chem. A 7, 12068–12074 (2019).

CASGoogle Scholar

Kibsgaard, J., Chen, Z., Reinecke, B. N. & Jaramillo, T. F. Engineering the surface structure of MoS2 to preferentially expose active edge sites for electrocatalysis. Nat. Mater. 11, 963–969 (2012).

ADSCASPubMedGoogle Scholar

Wang, L. et al. The reformation of catalyst: from a trial-and-error synthesis to rational design. Nano Res. 17, 3261–3301 (2024).

ADSCASMATHGoogle Scholar

Liu, F. et al. Phosphorus doping of 3D structural MoS2 to promote catalytic activity for lithium-sulfur batteries. Chem. Eng. J. 431, 133923 (2022).

CASMATHGoogle Scholar

Liu, J. et al. MoS2-modified graphite felt as a high performance electrode material for zinc–polyiodide redox flow batteries. Inorg. Chem. Front. 6, 731–735 (2019).

ADSCASMATHGoogle Scholar

Wang, Z. et al. Single-atomic Co-B2N2 sites anchored on carbon nanotube arrays promote lithium polysulfide conversion in lithium–sulfur batteries. Carbon Energy 5, e306 (2023).

CASGoogle Scholar

Zhao, S. et al. A universal seeding strategy to synthesize single atom catalysts on 2D materials for electrocatalytic applications. Adv. Funct. Mater. 30, 1906157 (2020).

CASGoogle Scholar

Liu, X. et al. Advances in the synthesis strategies of carbon⁃based single⁃atom catalysts and their electrochemical applications. China Powder Sci. Technol. 30, 35–45 (2024).

MATHGoogle Scholar

Liang, Z. et al. Co doped MoS2 as cocatalyst considerably improved photocatalytic hydrogen evolution of g-C3N4 in an alkalescent environment. Chem. Eng. J. 421, 130016 (2021).

ADSCASMATHGoogle Scholar

Qin, X., Ke, P., Wang, A. & Kim, K. H. Microstructure, mechanical and tribological behaviors of MoS2-Ti composite coatings deposited by a hybrid HIPIMS method. Surf. Coat. Technol. 228, 275–281 (2013).

CASGoogle Scholar

Gong, X. et al. Thermally stable hierarchical nanostructures of ultrathin MoS2 nanosheet-coated CeO2 hollow spheres as catalyst for ammonia decomposition. Inorg. Chem. 55, 3992–3999 (2016).

CASPubMedMATHGoogle Scholar

Yu, Y. et al. High phase-purity 1T′-MoS2- and 1T′-MoSe2-layered crystals. Nat. Chem. 10, 638–643 (2018).

ADSCASPubMedMATHGoogle Scholar

Liu, W. et al. Cobalt-doping of molybdenum disulfide for enhanced catalytic polysulfide conversion in lithium–sulfur batteries. ACS Nano 15, 7491–7499 (2021).

CASPubMedMATHGoogle Scholar

Guo, C. et al. Constructing CoO/Co3S4 heterostructures embedded in N-doped carbon frameworks for high-performance sodium-ion batteries. Adv. Funct. Mater. 29, 1901925 (2019).

Google Scholar

Liu, Q. et al. Electron-doped 1T-MoS2 via interface engineering for enhanced electrocatalytic hydrogen evolution. Chem. Mater. 29, 4738–4744 (2017).

CASMATHGoogle Scholar

Zhang, J. et al. Cobalt-modulated molybdenum–dinitrogen interaction in MoS2 for catalyzing ammonia synthesis. J. Am. Chem. Soc. 141, 19269–19275 (2019).

CASPubMedGoogle Scholar

Wang, T. et al. Multilevel heterostructure of MoS2/GDYO for lithium-ion batteries. Adv. Funct. Mater. 33, 2308470 (2023).

CASGoogle Scholar

Luo, Z. et al. Chemically activating MoS2 via spontaneous atomic palladium interfacial doping towards efficient hydrogen evolution. Nat. Commun. 9, 2120 (2018).

ADSPubMedPubMed CentralMATHGoogle Scholar

Li, X. et al. One-step hydrothermal synthesis of high-percentage 1T-phase MoS2 quantum dots for remarkably enhanced visible-light-driven photocatalytic h2 evolution. Appl. Catal. B 243, 76–85 (2019).

CASGoogle Scholar

Liu, H.-J., Zhang, S., Chai, Y.-M. & Dong, B. Ligand modulation of active sites to promote cobalt-doped 1T-MoS2 electrocatalytic hydrogen evolution in alkaline media. Angew. Chem. Int. Ed. 62, e202313845 (2023).

CASGoogle Scholar

Shi, Z. et al. Phase-dependent growth of Pt on MoS2 for highly efficient H2 evolution. Nature 621, 300–305 (2023).

ADSCASPubMedMATHGoogle Scholar

Gan, T. & Wang, D. Atomically dispersed materials: ideal catalysts in atomic era. Nano Res. 17, 18–38 (2024).

ADSCASMATHGoogle Scholar

Liu, G. et al. MoS2 monolayer catalyst doped with isolated Co atoms for the hydrodeoxygenation reaction. Nat. Chem. 9, 810–816 (2017).

CASPubMedMATHGoogle Scholar

Duan, H. et al. Single-atom-layer catalysis in a MoS2 monolayer activated by long-range ferromagnetism for the hydrogen evolution reaction: beyond single-atom catalysis. Angew. Chem. Int. Ed. 60, 7251–7258 (2021).

CASMATHGoogle Scholar

Song, H. et al. Cooperative catalytic Mo-S-Co heterojunctions with sulfur vacancies for kinetically boosted lithium-sulfur battery. Chem. Eng. J. 450, 138115 (2022).

CASMATHGoogle Scholar

Tsai, C. et al. Electrochemical generation of sulfur vacancies in the basal plane of MoS2 for hydrogen evolution. Nat. Commun. 8, 15113 (2017).

ADSPubMedPubMed CentralMATHGoogle Scholar

Huang, S. et al. Synergistically enhanced electrochemical performance of Ni3S4–PtX (X = Fe, Ni) heteronanorods as heterogeneous catalysts in dye-sensitized solar cells. ACS Appl. Mater. Interfaces 9, 27607–27617 (2017).

ADSCASPubMedGoogle Scholar

Zou, Q. et al. Cation-directed selective polysulfide stabilization in alkali metal–sulfur batteries. J. Am. Chem. Soc. 140, 10740–10748 (2018).

CASPubMedGoogle Scholar

Zou, Q. & Lu, Y.-C. Solvent-dictated lithium sulfur redox reactions: an operando UV–vis spectroscopic study. J. Phys. Chem. Lett. 7, 1518–1525 (2016).

CASPubMedMATHGoogle Scholar

Yan, W. et al. A tripartite synergistic optimization strategy for zinc-iodine batteries. Nat. Commun. 15, 9702 (2024).

CASPubMedPubMed CentralMATHGoogle Scholar

Zhang, S.-J. et al. Polyiodide confinement by starch enables shuttle-free Zn–iodine batteries. Adv. Mater. 34, 2201716 (2022).

CASGoogle Scholar

Wang, Y., Li, S. & Yi, J. Electronic and magnetic properties of Co doped MoS2 monolayer. Sci. Rep. 6, 24153 (2016).

ADSCASPubMedPubMed CentralGoogle Scholar

Yang, M., Ren, X., Hu, L., Guo, W. & Zhan, J. Facet-controlled activation of persulfate by goethite for tetracycline degradation in aqueous solution. Chem. Eng. J. 412, 128628 (2021).

CASMATHGoogle Scholar

Ning, S. et al. Co0−Coδ+ interface double-site-mediated C−C coupling for the photothermal conversion of CO2 into light olefins. Angew. Chem. Int. Ed. 62, e202302253 (2023).

CASGoogle Scholar

Liu, Z. et al. Cobalt-doped molybdenum disulfide for efficient sulfite activation to remove as(iii): preparation, efficacy, and mechanisms. J. Hazard. Mater. 452, 131311 (2023).

CASPubMedGoogle Scholar

Li, Q. et al. Forming atom–vacancy interface on the MoS2 catalyst for efficient hydrodeoxygenation reactions. Small Methods 3, 1800315 (2019).

Google Scholar

Wu, H. et al. Dual‐single‐atoms of Pt–Co boost sulfur redox kinetics for ultrafast Li–S batteries. Carbon Energy 6, e422 (2023).

MATHGoogle Scholar

Li, Z. & Lu, Y.-C. Polysulfide-based redox flow batteries with long life and low levelized cost enabled by charge-reinforced ion-selective membranes. Nat. Energy 6, 517–528 (2021).

ADSCASMATHGoogle Scholar

Wang, J. et al. Ambient ammonia synthesis via palladium-catalyzed electrohydrogenation of dinitrogen at low overpotential. Nat. Commun. 9, 1795 (2018).

ADSPubMedPubMed CentralMATHGoogle Scholar

Xue, L. et al. In situ/operando Raman techniques in lithium–sulfur batteries. Small Struct. 3, 2100170 (2022).

Google Scholar

Ma, L. et al. Long-lasting zinc–iodine batteries with ultrahigh areal capacity and boosted rate capability enabled by nickel single-atom electrocatalysts. Nano Lett. 23, 5272–5280 (2023).

ADSCASPubMedMATHGoogle Scholar

Dan, X. et al. Rare earth doped CeO2 uniformly enwraps graphite felt as high current density electrode for all vanadium redox flow battery. J. Power Sources 628, 235921 (2025).

CASGoogle Scholar

Li, Z. et al. Air-breathing aqueous sulfur flow battery for ultralow-cost long-duration electrical storage. Joule 1, 306–327 (2017).

CASMATHGoogle Scholar

Tang, L., Li, T., Lu, W. & Li, X. Lamella-like electrode with high Br2-entrapping capability and activity enabled by adsorption and spatial confinement effects for bromine-based flow battery. Sci. Bull. 67, 1362–1371 (2022).

CASMATHGoogle Scholar

Download references

Acknowledgements

This work was financially supported by the Guangxi Natural Science Fund for Distinguished Young Scholars (2024GXNSFFA010008), the National Natural Science Foundation of China (22469002, 22109118, 51971157, and 22275166). We also thank Jin Shen (Specreation Instruments Co., Ltd.) for help and discussions regarding XAS measurements and data analyses in this study.

Author information

Author notes

These authors contributed equally: Zhigui Wang, Guolong Lu.

Authors and Affiliations

Key Laboratory of Carbon Materials of Zhejiang Province, College of Chemistry and Materials Engineering, Wenzhou University, Wenzhou, 325035, China

Zhigui Wang, Guolong Lu & Ge Meng

Guangxi Key Laboratory of Processing for Non-ferrous Metals and Featured Materials, School of Chemistry and Chemical Engineering, Guangxi University, Nanning, 530004, Guangxi, China

Zhigui Wang, Guolong Lu, Tianran Wei, Haoxiang Cai, Yanhong Feng & Xijun Liu

School of Materials Science and Engineering, Lanzhou Jiaotong University, Lanzhou, 730070, China

Tianran Wei & Ke Chu

ShenSi Lab, Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Longhua District, Shenzhen, 518110, China

Jun Luo

School of Ecology and Environmental Science, Yunnan University, Kunming, 650504, China

Guangzhi Hu

Department of Chemistry, Tsinghua University, 100084, Beijing, China

Dingsheng Wang

Authors

Zhigui Wang

View author publications