Three-Step Synthesis of HES

We employ a three-step synthesis strategy to produce low-, medium-, and high-entropy catalysts (Figures 1a and S1a–c). The first step is Co-centered nucleation. Specifically, Co(acac)₂ reacts with thioacetamide to form CoS^* precursor in acetone as the initial Co-centered nuclei. Other components (Mn, Ni, Cu, Fe) are then sequentially introduced to produce precursors with increasing entropy levels, categorized by their configurational entropy as low-entropy sulfide (LES, Co, S), medium-entropy sulfide (MES, Co, Mn, Ni, S), and high-entropy sulfide (HES, Co, Mn, Ni, Fe, Cu, S). The Co–Mn–Ni–Fe–Cu combination was selected due to the similar atomic radii of its constituent elements with that of zinc. This compatibility facilitates the formation of a homogeneous solid solution and ensures interfacial stability with ZnS. The configurational entropy (ΔS_conf) is defined as low entropy (ΔS_conf < 1 R, where R is the universal gas constant), medium entropy (1 R < ΔS_conf < 1.5 R), and high entropy (ΔS_conf >1.5 R).^[8] Notably, in contrast to conventional Cu-based cation exchange method, which frequently results in inhomogeneous intermediates, the initial CoS* nuclei in our strategy serve as catalytic seeds that trigger the codeposition of other metals, facilitating the uniform incorporation of all elements into a homogeneous high-entropy solid solution.^{[11, 12]} In contrast, when Mn or Ni are used as initial nuclei, the formation of single-phase compounds is observed (Figures S2a,b and S3).^[11] Moreover, CoS^* nuclei critically maintain the Pa-3 symmetry in the solid solution structure (Figure 1b). As shown in the XRD results (Figure S4), the absence of Co disrupts this Pa-3 structure. The second step involves a dual-solvent assembly process. The as-prepared low- to high-entropy precursors assemble into nanorods within 0.5 h in a binary solvent system, achieving optimal morphology at 6 h. Nevertheless, extending the reaction to 10 h leads to over-assembly (Figures S5a–c). When a single solvent is used, ethanol yields nanospheres, whereas glycerol results in the formation of nanosheets (Figures S6a,b).^[13] The final step of annealing promotes grain growth and enhances the mechanical stability of the catalysts. The morphology of the HES catalyst is characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). As shown in Figure 1c, the HES nanorods exhibit a diameter of 240 nm and a length of 4–5 µm, with homogeneously distributed metal elements.

To investigate the crystal features of the as-prepared LES, MES, and HES catalysts, high-resolution TEM (HRTEM) is employed. For the HES, the formation of nanocrystalline grains indicates the development of short-range ordered structures, which is a typical feature of high-entropy solid solutions (Figures S7a,b and S8).^{[8, 14]} The synchrotron PD pattern of HES catalysts closely aligns with the simulated XRD pattern, confirming their Pa-3 symmetry structure.^[15] Furthermore, ICPMS analysis (Table S1) reveals the chemical formula of the HES catalyst is Fe_0.13Co_0.24Mn_0.12Ni_0.25Cu_0.26S, with a high configurational entropy of ΔS_conf = 1.56 R (12.96 J mol⁻¹ K⁻¹),^[8] implying its thermodynamic and electrochemical stability.^[8] The high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) results further reveal the uniform distribution of elements (Figure 1c). Among these, relatively weak Fe and Mn signals indicate higher oxidation states compared to Co, Cu, and Ni, which is consistent with the findings from the ICPMS analysis. Selected area electron diffraction (SAED) analysis shows distinct (111), (200), (211), (220), and (311) lattice planes of the HES catalysts, consistent with the synchrotron PD results (Figure 1d).

This three-step synthesis method produces nanorod catalysts with tunable entropy levels—low-, medium- (Figure 1e), and high-entropy configurations. Among these, the HES catalyst exhibits uniform metal distribution, crystallographic integrity, and high configurational entropy. The HES is set as the experimental group, with MES as the primary control, to study the impact of catalyst's entropy on sulfur utilization and cycling stability of the fast-charging AZSBs.

Catalyzing RDS to Improve Sulfur Utilization

To identify and catalyze the RDS in the solid–solid Zn–S redox process, we implement the operando synchrotron PD, operando synchrotron IRM, and operando Raman characterizations by incorporating the MES and HES catalysts into S@MES||Zn and S@HES||Zn batteries. As shown in Figure 2a, distinct wurtzite ZnS peaks corresponding to the (101), (105), and (109) planes emerge during discharge in the S@HES battery and vanish upon charging. In contrast, sphalerite ZnS is the typical discharge product in the reported AZSBs.^[13] This wurtzite ZnS phase, as reported,^[16] can reduce the charging polarization with accelerated oxidation reaction within the Zn–S redox process to improve the sulfur utilization compared to sphalerite ZnS.^[17]

Operando Raman spectroscopy provides insights into the different sulfur utilization between HES catalysts and MES catalysts. At the open-circuit voltage of the S@HES||Zn battery, both S₈ and S₈^* species are detected. During early discharge, the peak at ∼380 cm⁻¹ suggests the formation of ZnS₂^* (Figure 2b).^[18-21] Subsequently, the appearance of the peak at ∼410 cm⁻¹ indicates the formation of wurtzite ZnS.^[22] Additionally, a persistent weak peak at ∼630 cm⁻¹ is attributed to the HES catalysts. During charging, the redox reaction exhibits high reversibility with rapid conversion from wurtzite ZnS into ZnS₂^* and ultimately into S₈. Consequently, the complete reaction pathway catalyzed by HES can be summarized as: S₈/S₈^* ↔ ZnS₂^* ↔ ZnS. For comparison, operando Raman tests of the S@MES||Zn battery reveal that although MES catalysts facilitate the conversion of ZnS₂^*, weak peaks in the 200–300 cm⁻¹ range suggest the presence of irreversible intermediates, such as S₅²⁻ and S₃²⁻, during the discharge/charge cycle (Figure S9). The by-product formation in S@MES||Zn batteries reduce sulfur utilization, leading to low coulombic efficiency. In addition, other weak peaks at 600–630 cm⁻¹ may correspond to the potential leaching effects of the MES catalysts, indicating lower stability compared to the HES catalysts. Furthermore, operando synchrotron IRM spectra (Figures 2c,d) reveal that only SO₄²⁻ (ν₁) and H₂O (ν₁, ν₃) bands are detected in the S@HES|||Zn battery, indicating that HES catalysts efficiently enhance sulfur utilization. Notably, irreversible sulfur intermediates and sulfur disproportionation products, commonly reported in many AZSBs, are absent in the S@HES||Zn battery. In contrast, SO₃²⁻ species (∼970 cm⁻¹) are observed in the S@MES||Zn battery, which is identified as a by-product due to the sulfur disproportionation reaction.^[23] These findings highlight that HES catalysts optimize the reaction pathway of the solid–solid Zn–S redox process, ensuring efficient and rapid sulfur conversion. In contrast, MES catalysts lead to formation of sulfur-containing by-products, resulting in incomplete conversion of S₈ to wurtzite ZnS during discharging and decreased sulfur utilization.

These results reveal that HES catalysts exhibit high sulfur utilization of AZSBs by efficiently optimizing the reaction pathway of the solid–solid Zn–S redox process. In contrast, MES catalysts exhibit low sulfur utilization due to by-product formation. This distinction in sulfur utilization can be attributed to the high-entropy cocktail effect. Specifically, the five-component system of HES catalysts enables a diverse and synergistic distribution of active sites, significantly accelerating the solid–solid Zn–S redox process compared to the three-component system of MES catalysts.^[14]

Relationship Between Entropy and Cycling Stability

As implied in Raman results, HES catalysts exhibit higher cycling stability compared to MES catalysts because of the leaching effects in the MES.^[8] To further explore the relationship between catalysts’ entropy and cycling stability, NEXAFS, ICP-MS, and electronmicroscopic characterizations were conducted to examine the structural changes of the catalysts in S@HES and S@MES cathodes.

The NEXAFS results demonstrate that the chemical states of Mn, Co, Ni, and Fe in the S@HES cathodes are highly stable throughout cycling (Figures 3a–c and S10). In contrast, the chemical states of Mn, Co, and Ni in the S@MES cathodes undergo substantial change during cycling. Specifically, the intensities of Mn L₂ edge at 650 eV and L₃ edge at 642 eV exhibit a notable decrease, dropping to 41% of their original intensity, which indicates a significant reduction in the chemical state of Mn after 10 cycles. The Co L₂ edge at 793 eV and Co L₃ edge at 778 eV edges slightly shift toward higher energy, appearing at new positions of Co L₂ edge (ca. 794 eV) and Co L₃ edge (ca. 779 eV) edges, whereas a new Ni L₂ edge at 870 eV emerge after 10 cycles, indicating an increase of the chemical states for both Co and Ni. These observations suggest potential metal leaching in the MES catalysts, wherein the three-component system (Co, Mn, Ni) compensates for the dissolution of one metal species through oxidation state adjustments among the remaining elements to preserve overall electroneutrality.

To gain deep insight into the leaching effects, liquid ICPMS analysis of the electrolytes after 10 cycles was further conducted (Figure 3d). The study reveals Mn (II) and Co (II) leaching into the electrolyte in the S@MES||Zn battery, whereas these signals are not observed in the electrolyte of the cycled S@HES||Zn battery. In addition, ICPMS results illustrate the direct element changes in HES and MES catalysts after 10 cycles (Figures 3e,f). The configurational entropy of HES catalysts remains stable at 1.56 R, whereas the configurational entropy of MES catalysts significantly decreases from 1.08 R to 1.03 R (Table S2). TEM-EDX and corresponding mappings (Figures 3g–i) reveal the presence of “dead ZnS”^[13] in the S@MES cathodes at the charged status, referring to residual ZnS that cannot be reconverted to S during charging, which reduces the configurational entropy. Therefore, this decrease in the configurational entropy is primarily due to Mn/Co leaching, with a secondary contribution from the accumulation of ZnS residues.

To investigate the suppression capability of water splitting from low-entropy to high-entropy catalysts, CV and in situ GCMS analyses are conducted in batteries with varying entropy levels: S@LES||Zn, S@MES||Zn, and S@HES||Zn systems. The S@HES||Zn battery exhibits the highest electrochemical stability within the voltage range of 0.05–1.40 V (Figure S11), with no signals indicating leaching effects or water splitting. In contrast, distinct peaks emerged during the initial reduction reaction in the S@LES||Zn and S@MES||Zn batteries, indicating the Mn/Co leaching effects corresponding to the liquid ICPMS results. Additionally, in situ GCMS results confirm that, apart from the residual oxygen initially present in the electrolytic cell, no detectable oxygen accumulated in the S@HES||Zn batteries over five cycles at a current density of 20 A cm⁻²(Figure S12).

Overall, the above analyses confirm a correlation between high entropy and cycling stability, driven by the high-entropy effect.^[14] This means high-entropy materials can stabilize high-entropy solid solution structures and suppress leaching effects and water splitting. Figure 3j summarizes a comprehensive understanding of the impact of catalysts’ entropy on sulfur utilization and cycling stability. HES catalysts optimize the Zn–S redox pathway and maintain structural stability, achieving high sulfur utilization and cycling stability. In contrast, LES and MES catalysts exhibit inefficient pathways, leading to by-products and Mn/Co leaching, thus reducing sulfur utilization and cycling stability.

Electrochemical Performances of Fast-Charging AZSBs

To demonstrate the practical applications of the mechanisms, we evaluate the electrochemical performances of AZSBs with LES, MES, HES catalysts and without catalysts. Electrochemical CV profiles demonstrate that as the entropy increases, the reduction peaks shift to higher voltages, and the oxidation peaks shift to lower voltages, reducing the overall polarization (Figure 4a).^[24] In addition, derived from chronoamperometry (CA) results shown in Figure S13, it confirms that high entropy catalysts exhibit early onset potentials for the discharge reaction (Figure 4b). These findings underscore the rapid conversion kinetics of the solid–solid Zn–S redox process.

The in situ distribution of relaxation times (in situ DRT) results further reveal the impedance evolution of S@blank||Zn, S@MES||Zn, and S@HES||Zn batteries under various states of charge (SoCs) (Figures 4c,d, S14, and S15). At the early discharge stage (0.70 V vs. Zn²⁺/Zn), the S@HES||Zn battery exhibits a slight increase in charge transfer resistance (R_ct), demonstrating the rapid ZnS conversion process. In contrast, a steep rise in R_ct is observed during the middle and late discharge stages in the S@MES||Zn and S@blank||Zn batteries, indicating that the ZnS conversion in these two systems occurs at a lower reduction potential, resulting in higher battery polarization compared with the S@HES||Zn batteries. In addition, the similar changes in Warburg diffusion resistance (R_d) further emphasize the superior Zn–S redox kinetics of the S@HES||Zn battery.

To evaluate the practical electrochemical performance, pouch cells with various catalysts were assembled and designated as PC-blank, PC-MES, and PC-HES, as shown in Figure S16. Galvanostatic charge–discharge (GCD, Figure 4e) measurements reveal that the PC-HES battery exhibits a low polarization voltage of ΔE = 0.54 V at 0.1 C. Notably, the PC-HES system maintained a stable discharge plateau at 0.61 V, achieving a high cathode energy density of 313 Wh kg⁻¹ (based on the mass of sulfur, catalysts, carbon, and binder). During the charging process, PC-HES displays two distinct plateaus at 0.95 and 1.17 V, corresponding to the activation of the thick wurtzite ZnS layer and its subsequent oxidation to sulfur as confirmed by previous CV analyses. The coin-type S@HES||Zn battery (Figures S17 and S18) demonstrated a higher discharge plateau at 0.75 V and a single charge plateau at 1.17 V, resulting from the thinner insulating layer of the sulfur species compared to that in pouch cells. In contrast, PC-MES exhibits two separated discharge plateaus at 0.70 and 0.35 V, indicating increased polarization associated with an alternative reaction pathway. This facilitates the formation of undesirable polysulfides by-products such as S₅²⁻ and S₃²⁻ species, compromising battery performance. This observation aligns well with the results from the in situ DRT, CV, and CA analyses.

The fast-charging capability of PC-HES was evaluated at a high current rate of 4 C (Figure 4f). The cell delivered an initial reversible capacity of 770 mAh g⁻¹ and retained 577 mAh g⁻¹ after 400 cycles, corresponding to a capacity retention of 75%. This performance reflects an exceptionally low-capacity decay rate of 0.06% per cycle, as further confirmed by the GCD profiles in Figure S19. In addition to fast-charging performance, the low-rate performance and corrosion resistance of PC-HES were systematically investigated. As shown in Figure 4g, PC-HES exhibits excellent cycling stability at 0.1 C for 50 cycles, as indicated by the consistent discharge plateau at 0.61 V, demonstrating sustained catalytic activity of the HES. More importantly, as shown in Figures S20a,b, corrosion tests highlight the superior stability of the PC-HES, which maintained electrode integrity owing to its nonleaching characteristics. In contrast, conventional iodine-based AZSBs suffer from pronounced zinc corrosion. Figure S21a presents a comparative analysis of the rate performance for PC-MES and PC-HES, with corresponding GCD profiles shown in Figure S21b. The PC-HES system demonstrates stable capacity retention across a range of rates from 0.1 to 2 C, exhibiting only moderate capacity fading at the high rate of 5 C, as shown in Figure 4h. In contrast, PC-MES displays continuous capacity degradation over the 0.1–2 C range, which is attributed to the gradual leaching of Mo/Co catalytic components during cycling.

Compared with recently reported pouch-type AZSBs employing liquid redox mediators (Figure 4i and Table S3), the PC-HES system exhibits superior performance across several key metrics. Notably, it delivers a high cathode energy density of 313 Wh kg⁻¹ with a stable 400 cycles at 4 C under a practical sulfur loading of 50 mg per cell. Moreover, the system exhibits an exceptionally low capacity decay of only 0.5 mAh g⁻¹ per cycle, highlighting its strong potential for high-performance and durable energy storage applications.