English

• Lithium-sulfur (Li-S) batteries are widely regarded as among the most promising candidates for next-generation renewable-energy storage, owing to their high theoretical capacity (1672 mAh/g) and gravimetric energy density (2600 Wh/kg) [1,2]. Despite these advantages, practical deployment is hindered by the shuttle of soluble long-chain lithium polysulfides (LiPSs) between the cathode and anode and by intrinsically sluggish redox kinetics, which together precipitate rapid capacity fading, shortened cycle life, and depressed Coulombic efficiency [3,4]. Prior efforts have largely focused on chemical confinement of LiPSs via electrolyte additives [5], membrane modifications [6,7] and hybridization with polar compounds [8] to suppress shuttling and enhance cyclability. However, these approaches primarily immobilize LiPSs rather than fundamentally overcoming the kinetic bottlenecks inherent to sulfur redox chemistry. Critically, incomplete reduction of LiPSs to Li₂S and inefficient re-oxidation to S₈ during cycling drive sulfur agglomeration and electrolyte depletion. Accordingly, there is an urgent need to develop highly active electrocatalysts that concurrently strengthen LiPSs adsorption and accelerate bidirectional (reduction/oxidation) conversion kinetics in Li-S batteries.

  Polyoxometalates are a diverse class of anionic, transition metal oxygen clusters (typically based on Mo, W, V, Nb, etc.) that feature exceptional multi-electron storage capacity, pronounced redox activity, and excellent stability under electrochemical operating conditions [9,10]. Owing to these attributes, polyoxometalates have been extensively investigated as versatile electron reservoirs and redox mediators to promote a broad spectrum of electrocatalytic reactions. In the context of Li-S batteries, cluster-based polyoxometalates catalysts are frequently integrated as functional interlayers or coatings on commercial polypropylene separators (PP), thereby directly participating in cell operation to catalyze the rapid, bidirectional conversion between Li₂S/Li₂S₂ and soluble LiPSs species [11,12]. Among the various polyoxometalates families, high-nuclear species (e.g., Mo₁₃₂, Mo₁₃₈, Mo₁₅₄) arise from the orderly, atomically precise assembly of metal-oxide building units [13,14]. Compared with classical polyoxometalates, these high-nuclear analogues offer a higher density of metal-oxygen assembly units, while their mixed-valent Mo^Ⅵ/Mo^Ⅴ subunits provide abundant redox-active centers and adsorption sites, thereby imparting superior catalytic potential [15–17]. Consequently, incorporating high-nuclear polyoxometalates into Li-S battery systems is expected to markedly enhance both polysulfide anchoring and catalytic conversion. However, recent studies have shown that the reductive "heteropoly blue" species formed during lithiation are highly soluble in typical electrolytes [18]. During subsequent charging, these reductive moieties tend to leach into the electrolyte, progressively degrading the high-nuclear framework. Thus, engineering polyoxometalates catalysts with suppressed solubility and robust structural integrity under operating conditions remains a pressing challenge.

  Atomically precise copper nanoclusters have attracted substantial attention in catalysis, because they not only retain the advantages of copper nanoparticles but also feature precise chemical compositions, tunable metal cores, and functional ligand shells [19,20]. Unlike conventional copper nanoparticles, the copper nanoclusters exhibited a precise number of copper atoms and various coordination environments, such as Cu-C, Cu-S and Cu-Cu interactions, which modulate the electronic structure of the metal core and, in turn, its catalytic behavior [21,22]. These clusters are stabilized by surface ligands that, beyond preserving structural integrity, strongly influence the electronic and catalytic properties of the nanoclusters [23]. For example, multiple studies have demonstrated that tailoring ligand substituents markedly alters the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) of gold clusters [24,25]. By analogy, the electronic structures of Cu(I) clusters can, in principle, be systematically tuned via ligand engineering, thereby endowing them with the potential for outstanding catalytic performance [26]. Moreover, appropriately designed ligands can engage in covalent bonding or strong coordination with complementary catalyst/host frameworks, enabling synergistic pathways that further enhance activity [27]. Despite these advantages and notable advances, atomically precise copper nanoclusters remain largely unexplored as electrocatalysts for Li-S batteries, primarily because Cu species are prone to dissolution in organic electrolytes, leading to leaching and rapid deactivation [28]. Accordingly, the central challenge lies in devising robust strategies to immobilize and stabilize copper nanoclusters under Li-S battery operating conditions, ensuring their long-term retention and catalytic functionality within the battery environment.

  Building on the above, a new cluster-to-cluster composite (abbr. as Cu₂₀−Mo₁₅₄) has been constructed by integrating an atomically precise copper nanocluster of ([(O)Cu₂₀(C₂₀H₂₄O₂)₁₂(CH₃COO)₆]) (abbr. as Cu₂₀) with a high-nuclear polyoxometalate of (Na₁₄[Mo₁₅₄O₄₆₂H₁₄(CH₃OH)₈(H₂O)₆₂·mH₂O·nCH₃OH] (m ≈ 400, n ≈ 10) (abbr. as Mo₁₅₄). The two clusters assemble predominantly via pervasive hydrogen-bonding interactions, thereby preserving the inherent structural features and catalytic sites of each component while substantially suppressing their solubility in aqueous or organic media. Given the rich redox activity of both constituents, we further elucidate the mechanisms by which the Cu₂₀−Mo₁₅₄ composite enhances Li-S battery performance when deployed as a separator interlayer. With a Cu₂₀−Mo₁₅₄ modified separator, a conventional sulfur cathode containing 68.5 wt% S delivers a high reversible capacity of 1355 mAh/g at 0.1 C, together with excellent cycling stability. Remarkably, even at a high rate of 10 C, the cell retains 317 mAh/g after 1100 cycles, indicative of outstanding high-rate capability and suggestive of flash-charging potential. Moreover, Li-S batteries equipped with the Cu₂₀−Mo₁₅₄ modified separator exhibit durable lithium protection for up to 3000 h at a current density of 1 mA/cm².

  The structural and morphological analysis of Mo₁₅₄ and Cu₂₀ are presented in Figs. S1 and S2 (Supporting information). Cu₂₀—Mo₁₅₄ was synthesized by adding an ethanol solution of Cu₂₀ nanoclusters to an aqueous solution of Mo₁₅₄ (Fig. 1a). Briefly, the mixed solution is stirred at room temperature for 0.5 h, followed by precipitation in diethyl ether, Cu₂₀−Mo₁₅₄ can be obtained in a 95% yield. Energy dispersive X-ray spectroscopy (EDS) analysis reveals that the atomic fractions of Cu and Mo are 1.26% and 20.38% (Fig. 1b), respectively. Meanwhile, inductively coupled plasma mass spectrometry (ICP-MS) results confirm that the mass percentages of Cu and Mo in the Cu₂₀−Mo₁₅₄ composite are 8.3% and 43.3%, respectively (Fig. S3 in Supporting information). Both sets of quantitative results are consistent with an approximate 2:1 molar ratio of Mo₁₅₄ to Cu₂₀ in the as-prepared Cu₂₀−Mo₁₅₄ composite. Through structural and surface elemental characterization and analysis of the Cu₂₀−Mo₁₅₄ composite, it is confirmed that the Cu₂₀−Mo₁₅₄ composite has been successfully synthesized (Fig. S4 in Supporting information). High-resolution transmission electron microscope (HRTEM) studies were displayed large particles but mostly of par reduced crystallinity because of higher weathering inside the microscope column (Fig. 1c). The FTIR spectrum of Cu₂₀−Mo₁₅₄ retains characteristic bands of both constituents (Fig. 1d). The Mo₁₅₄ component exhibits features at 1618 cm⁻¹ (m; δ(H₂O)), 978 cm⁻¹ (m), 912 cm⁻¹ (w − m; ν(Mo=O)), 810 cm⁻¹ (sh), 745 cm⁻¹ (s), 628 cm⁻¹ (s), and 555 cm⁻¹ (s), attributable to ν(Mo=O), ν(Mo−O−Mo), and δ(Mo−O) vibrations, while the Cu₂₀ cluster shows characteristic bands in excellent agreement with our previous report [29]. The composition and surface electronic structure of Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ are further investigated by X-ray photoelectron spectroscopy (XPS). As shown in Fig. 1e, the Cu 2p spectra peaks of Cu₂₀ shift to lower binding energy, from 933.0 eV to 931.0 eV (Cu 2p_3/2) and from 953.0 eV to 952.8 eV (Cu 2p_1/2) after interacting with Mo₁₅₄, consistent with Cu⁰ or Cu⁺ species [30]. The results indicate that the introduction of Mo donates electrons to the Cu sites, enhancing the overall reducibility of the Cu sites. On the contrary, compared with Mo₁₅₄ (Fig. 1f), the Mo 3d peaks assigned to Mo⁶⁺ and Mo⁵⁺ [31] in Cu₂₀−Mo₁₅₄ shift to higher binding energy, accompanied by a decreased Mo⁵⁺ ratio. Specifically, the Mo⁶⁺ 3d peaks shift from 232.5 to 232.7 eV (Mo 3d_5/2) and from 235.7 eV to 235.9 eV (Mo 3d_3/2), while the Mo⁵⁺ 3d peaks move from 234.5 eV to 234.9 eV (Mo 3d_3/2) and from 231.3 eV to 231.9 eV (Mo 3d_5/2). These changes reflect a reduced local electron density at the Mo sites. Together with the low binding energy shift of Cu 2p peaks, these results confirm net electron transfer from Mo to Cu at the Cu₂₀ and Mo₁₅₄ interface. The O 1s spectra provides further insight into interfacial interactions. For Cu₂₀ (Fig. S5a in Supporting information), a single peak at 532.2 eV is observed, corresponding to surface hydroxyl oxygen from ethinyl-estradiol. For Mo₁₅₄ (Fig. S5b in Supporting information), peaks at 530.5 and 531.6 eV are assigned to lattice oxygen and surface-adsorbed oxygen, respectively. In Cu₂₀−Mo₁₅₄, the O 1s peak at 530.6 eV corresponds to lattice oxygen, while the peak at 532.0 eV is attributed to surface-adsorbed oxygen. The hydroxyl oxygen of Cu₂₀ shifts to lower binding energy, whereas the surface-adsorbed oxygen of Mo₁₅₄ shifts to higher binding energy after Cu₂₀−Mo₁₅₄ composite material formation. These opposite shifts indicate the formation of hydrogen bonds between surface hydroxyl oxygen on Cu₂₀ and surface-adsorbed oxygen on Mo₁₅₄ [32]. Cyclic voltammetry (Fig. 1g) shows that both anodic and cathodic peak potential of Cu₂₀ are more positive than those of Mo₁₅₄, reflecting the stronger oxidizing character of Cu₂₀. Upon composite formation, hydrogen bond assisted interfacial interactions drive electron transfer from Mo₁₅₄ to Cu₂₀ [33], in agreement with the XPS results. For Mo₁₅₄, the cathodic peak current substantially exceeds the anodic counterpart, evidencing that the Mo⁶⁺ to Mo⁵⁺ reduction predominates while the reverse oxidation (Mo⁵⁺ to Mo⁶⁺) is sluggish, indicating the redox process is poorly reversible. After incorporating Cu₂₀, the anodic peak current of Cu₂₀−Mo₁₅₄ increases markedly, while the anodic peak potential remains close to that of Mo₁₅₄, indicating that Mo₁₅₄ remains the principal redox active center in the composite. Meanwhile, Cu₂₀ promotes the re-oxidation of Mo⁵⁺ to Mo⁶⁺, lowers the apparent polarization (sharper peaks with higher intensities), and thereby enhances the redox reversibility and kinetics.

  Figure 1

  

  Figure 1.  (a) Schematic illustration for the synthesis process of the Cu₂₀−Mo₁₅₄. (b) Atomic fractions on Cu₂₀−Mo₁₅₄. (c) HRTEM images of Cu₂₀−Mo₁₅₄. (d) FTIR spectra of Cu₂₀−Mo₁₅₄. XPS patterns of (e) Cu 2p and (f) Mo 3d. (g) CV curves of different materials.

  DownLoad: Full-Size Img PowerPoint

  Compared with strategies such as anode host design, electrolyte optimization, and interface engineering, separator modification has attracted particular attention owing to its simplicity and low cost. We employed the Cu₂₀−Mo₁₅₄ composite catalyst to functionalize the Li-S battery separator. This interfacial coating is designed to catalyze polysulfide conversion, accelerating both reduction and oxidation pathways to promote efficient LiPSs transformation. As illustrated in Fig. 2a, a homogeneous slurry comprising Cu₂₀−Mo₁₅₄, conductive carbon black (CCB), and polyvinylidene fluoride (PVDF) binder was prepared and cast onto one side of a commercial polypropylene (PP) separator to form a functional interlayer. The Cu₂₀−Mo₁₅₄ modification layer forms a physical barrier on the PP surface to prevent LiPSs migration from the cathode to the anode. As displayed in Fig. S6a (Supporting information), the electrolyte rapidly and completely wets the Cu₂₀−Mo₁₅₄ modified separator within 1.0 s, yielding a markedly smaller contact angle than the untreated PP separator. This enhanced wettability indicates improved interfacial compatibility between the electrolyte and the separator after Cu₂₀−Mo₁₅₄ modification. Superior electrolyte affinity mitigates interfacial Li⁺ concentration gradients during transport and thus promotes faster Li⁺ migration kinetics. In the folding/recovery tests, the Cu₂₀−Mo₁₅₄ modified separator exhibits no visible cracks or delamination (Fig. S6b in Supporting information), indicating strong interfacial adhesion and robust mechanical integrity. In addition, only a negligible amount of Cu₂₀−Mo₁₅₄ is dissolved in the electrolyte (Fig. S7 in Supporting information), and its chemical structure remained unchanged (Fig. S8 in Supporting information). Its stability in the electrolyte is favorable for the stable catalytic effect toward polysulfides on the battery separator surface. Elemental mapping by EDS (Fig. S9 in Supporting information) confirms a homogeneous distribution of Cu, Mo, O, and Na throughout the modified layer. Ionic transport was evaluated using electrochemical impedance spectroscopy (EIS) on stainless-steel symmetric cells (SS//separator//SS). The bulk resistance, taken from the high-frequency x axis intercept of the Nyquist plot, together with the separator thickness (Fig. S10 in Supporting information), was used to calculate ionic conductivity via Eq. S2 (Supporting information). At room temperature the Cu₂₀−Mo₁₅₄ modified separator exhibits the lowest resistance of 3.19 Ω (Fig. S11 in Supporting information), corresponding to Li⁺ conductivity of 0.77 mS/cm (Fig. S12 in Supporting information), which exceeds those of the Mo₁₅₄ modified (0.66 mS/cm), Cu₂₀ modified (0.18 mS/cm), and PP (0.31 mS/cm) separators. Collectively, these results demonstrate that Cu₂₀−Mo₁₅₄ modification enhances electrolyte separator interfacial wettability and facilitates Li⁺ transport, enabling high flux Li⁺ diffusion across the separator.

  Figure 2

  

  Figure 2.  (a) The schematic illustration that Cu₂₀−Mo₁₅₄ suppresses LiPSs shuttling. (b) electronic photograph of different material immersion in the Li₂S₆ solution for 9 h. (c) UV–vis spectra of different samples after immersion in the Li₂S₆ solution for 9 h. XPS analysis of (d) Cu 2p, (e) Mo 3d and (f) O 1s of Cu₂₀−Mo₁₅₄ and Cu₂₀−Mo₁₅₄@Li₂S₆. (g) XPS analysis of S 2p spectra for Li₂S₆ and Cu₂₀−Mo₁₅₄@Li₂S₆.

  DownLoad: Full-Size Img PowerPoint

  To visually assess the adsorption of Li₂S₆ by different materials, conductive carbon black, Cu₂₀, Mo₁₅₄, and Cu₂₀−Mo₁₅₄ were introduced into a 0.5 mol/L Li₂S₆ solution (1,3-dioxane/1,2-dimethoxyethane) and allowed to react (Fig. 2b). After 9 h, the supernatant for conductive carbon black no noticeable change compared with the blank. In contrast, the supernatant for Cu₂₀ turned deep black, consistent with dissolution of Cu₂₀ into the electrolyte and the concurrent adsorption of Li₂S₆, which intensified the coloration. The supernatant of Mo₁₅₄ became light yellow, indicating a modest adsorption capacity for Li₂S₆. Strikingly, Cu₂₀−Mo₁₅₄ rendered the supernatant nearly colorless, evidencing strong adsorption of Li₂S₆ and implying that the interaction between Cu₂₀ and Mo₁₅₄ suppresses Cu₂₀ dissolution in the electrolyte while enhancing Li₂S₆ capture. UV–vis spectra of the supernatants corroborate these observations, confirming the Li₂S₆ adsorption capacity in the order Cu₂₀−Mo₁₅₄ > Mo₁₅₄ > Cu₂₀ > CCB (Fig. 2c). The chemical interaction between Cu₂₀−Mo₁₅₄ and Li₂S₆ was further probed by XPS. The Cu 2p, Mo 3d and O 1s spectra reveal binding-energy shifts after Li₂S₆ adsorption indicative of electronic interactions between Cu₂₀−Mo₁₅₄ and Li₂S₆ [34,35]. This behavior is attributed to the strong electronegativity of terminal O sites, which withdraw electron density from Li (Fig. S13 in Supporting information) and modulate the electronic structure of the embedded metal centers via the "collision effect" [35,36]. Specifically, the Cu 2p peaks shift to lower binding energies after exposure to Li₂S₆, with Cu 2p_3/2 moving from 931.85 eV to 931.6 eV and Cu 2p_1/2 from 951.9 eV to 951.5 eV (Fig. 2d). In the Mo 3d region of Cu₂₀−Mo₁₅₄@Li₂S₆, both Mo⁶⁺ and Mo⁵⁺ features shift to lower binding energies, accompanied by an increased ration of Mo⁵⁺ (Fig. 2e), consistent with the partial reduction of Mo⁶⁺ upon electron gain. When Li₂S₆ adsorption, a new O 1s component emerges at 530.7 eV, assignable to Li−O bonds (Fig. 2f) [36]. For S 2p, pristine Li₂S₆ exhibits the characteristic doublets of terminal sulfur (ST⁻¹) and bridging sulfur (SB⁰) at 160.2–162.8 eV [37]. In Cu₂₀−Mo₁₅₄@Li₂S₆, these polysulfide signals are attenuated and a new peak appears at 168.4 eV, characteristic of thiosulfate (Fig. 2g). The S 2p data indicate oxidation of sulfur in Li₂S₆ (−1/0) to higher-valence oxy-sulfur species (e.g., sulfate) [38] on Cu₂₀−Mo₁₅₄. The shuttle of lithium polysulfides through the modified separator under a concentration gradient was detected using an H-type glass bottle (Fig. S14 in Supporting information). It is worth noting that the original PP separator could not prevent the diffusion of lithium polysulfides, while the penetration of polysulfides was inhibited when using the Mo₁₅₄ modified separator. However, the Cu₂₀ modified separator could not prevent the polysulfide shuttle due to the dissolution of Cu₂₀ in the electrolyte. Fortunately, there was no significant change in the right chamber using the Cu₂₀−Mo₁₅₄ modified separator, which indicates that Cu₂₀−Mo₁₅₄ has a more excellent ability to inhibit polysulfide shuttle due to the synergistic effect of physical barrier and strong chemical adsorption. Collectively, these results demonstrate that Cu₂₀−Mo₁₅₄ composite affords strong adsorption of LiPSs.

  Density functional theory (DFT) calculations were performed to elucidate the adsorption behavior of lithium polysulfides on Cu₂₀−Mo₁₅₄. Prior studies on large molybdenum-oxide clusters (Mo₈₀, Mo₁₅₄, and Mo₁₆₀) have reported that the reduced electrons are delocalized within the {Mo₅O₆} unit of Mo₁₅₄ [39], whereas the semi-closed architectures of Mo₈₀ and Mo₁₆₀ exhibit Mo 4d electrons delocalization not only between the wheel-like {Mo₅O₆} and {Mo₇O₉} units but also toward the central vertices {Mo₃O₁₄} and {Mo₈O₃₆} [16]. The reduction degree of Mo and the energetic position of Mo 3d states are known to govern electronic delocalization in polyoxomolybdate. Structurally, Mo₁₅₄ is composed of {Mo₅O₆} (abbr. as {Mo₅}) and {Mo₈O₂₆} (abbr. as {Mo₈}) building units [13]. While previous work has emphasized electron delocalization within {Mo₅}, the role of {Mo₈} has been less explored. To address this gap, we computed the adsorption energies of representative LiPSs on {Mo₈}, {Mo₅} and {Cu₃} [29] site (Fig. 3a). The optimized configurations (Fig. 3b) show that {Mo₈}, {Mo₅}, and {Cu₃} all bind LiPSs primarily via Li–O coordination. Li⁺ (Lewis acidic sites) interacts with terminal O atoms (Lewis basic sites) on the Mo–O framework or with the −OH group of ethinyl estradiol in {Cu₃}. The trend of adsorption strength for long-chain lithium polysulfides (Li₂S₈ and Li₂S₆) on different structural units is {Mo₈} > {Mo₅} > {Cu₃}, whereas for short-chain lithium polysulfides (Li₂S₄ and Li₂S₂), the adsorption strength patterns are {Mo₅} > {Mo₈} > > {Cu₃} (Fig. 3c). Long-chain lithium polysulfides possess more delocalized charge distributions [40]. The {Mo₈} structural unit provides a dense array of terminal Mo=O and bridging Mo–O–Mo sites that offer multiple Li⁺ coordination points, together with mix-valent Mo⁶⁺ and Mo⁵⁺ centers that afford notable electron accepting capacity [41]. This combination facilitates charge redistribution and stabilization of dissolved Li₂S₈/Li₂S₆, yielding stronger and more numerous binding motifs. By contrast, shorter chains (Li₂S₄ and Li₂S₂) feature more concentrated electron density at terminal S atoms [42], resulting in stronger chemical adsorption to exposed Mo sites in {Mo₅}. Such interactions promote S–S bond cleavage and nucleation, thereby favoring deep conversion of LiPSs and uniform Li₂S deposition. The {Cu₃} site engages LiPSs mainly through the hydroxyl functionality of its organic ligand, providing weaker direct Li⁺ coordination than the oxygen rich Mo–O framework. Nevertheless, the Cu d electrons can assist S–S activation and electron transfer, facilitating reversibility, albeit with limited adsorption strength. In summary, the synergistic interplay between Cu₂₀ and Mo₁₅₄ endows Cu₂₀−Mo₁₅₄ with functions of "strong anchoring" for LiPSs, thereby suppressing LiPSs shuttling.

  Figure 3

  

  Figure 3.  (a) The structure of sandwiched skeleton in Cu₂₀−Mo₁₅₄ and constituent structural unit for Mo₁₅₄ ({Mo₅}, {Mo₈}) and Cu₂₀ ({Cu₃}). (b) Optimized adsorption configurations of various LiPSs on unit of {Mo₅}, {Mo₈} and {Cu₃}. (c) Binding energies between various LiPSs on unit of {Mo₅}, {Mo₈} and {Cu₃}.

  DownLoad: Full-Size Img PowerPoint

  A symmetrical battery was assembled using Li₂S₆ to investigate the conversion reaction kinetics of LiPSs. Cyclic voltammetry (CV) of symmetric cells (Fig. 4a) shows that introducing Li₂S₆ yields pronounced redox currents compared with symmetrical cells without Li₂S₆, confirming that the current arises from the reversible polysulfide conversion in the Li₂S₆ electrolyte. At identical overpotential, Cu₂₀−Mo₁₅₄ delivers the largest current and integrated area indicative of the fastest LiPSs redox kinetics and the lowest polarization. After four cycles, the CV traces of the Cu₂₀−Mo₁₅₄ symmetric cell nearly overlap, with the forward and reverse currents remaining highly symmetric and only a slight increase in current observed (Fig. 4b), evidencing high reversibility, a stable interface, and sustained electrocatalytic activity. The XRD results (Fig. S15 in Supporting information) show that the characteristic peaks of Cu₂₀−Mo₁₅₄ remains unchanged in position and shape after 10 cycles, with no new impurity peaks detected, confirming that the structure of Cu₂₀−Mo₁₅₄ remains intact after long-term cycling. Electrochemical impedance spectroscopy (EIS) of Li₂S₆ symmetric cells equipped with Cu₂₀, Mo₁₅₄, or Cu₂₀−Mo₁₅₄ electrodes (Fig. S16a in Supporting information) shows that Cu₂₀−Mo₁₅₄ exhibits the smallest high frequency semicircle (lowest charge transfer resistance) and the weakest Warburg diffusion feature, consistent with the fastest LiPSs oxidation/reduction kinetics. This enhancement is attributed to strong multipoint Li–O interactions at the {Mo₈} and {Mo₅} sites of Mo₁₅₄, which rapidly capture and activate LiPSs, while the Cu sites provide efficient electronic pathways. The synergistic effect accelerates LiPSs conversion and markedly lowers impedance. Tafel analysis (Fig. 4c) further reveals that Cu₂₀−Mo₁₅₄ has the highest exchange current density in agreement with EIS result and a larger i₀ corresponds to a lower charge transfer resistance. The chronoamperometry curves of symmetric cells with Li₂S₆ exhibit an enhanced current response compared to cells without Li₂S₆ (Fig. S16b in Supporting information), suggesting that the current responses are primarily influenced by the lithiation and delithiation reaction pathway rather than the double-layer capacitance. Furthermore, Cu₂₀−Mo₁₅₄ electrode exhibits the highest current response, which proves the enhanced LiPSs affinity and re-dox kinetics [43]. Among the tested electrodes, Cu₂₀−Mo₁₅₄ demonstrates the highest current response, reflecting enhanced LiPSs adsorption, favorable redox kinetics, and reduced charge transfer barriers. This is mainly attributed to the multisite Li−O coordination of LiPSs on Mo₁₅₄ combined with Cu₂₀ mediated electron pathways, which facilitates S–S bond activation and accelerates electron transfer. The deposition and dissolution capabilities of Li₂S on Cu₂₀−Mo₁₅₄ and Cu₂₀ are evaluated through Faraday's law. The initial peak corresponds to instantaneous nucleation of Li₂S, followed by a decay characteristic of diffusion-controlled growth. The early segment reflects Li₂S₈ to Li₂S₄ reduction, followed by Li₂S₄ to Li₂S deposition process, and continued reduction of Li₂S₆ still present in the long tail segment. The deposition capacity of Li₂S on Cu₂₀−Mo₁₅₄ (Fig. 4d) is approximately 281.6 mAh/g by integrated of curve. By contrast, Cu₂₀ (Fig. 4g) exhibits a smaller peak and rapid decay, yielding a significantly lower integrated Li₂S capacity. These results indicate that Mo₁₅₄ furnishes abundant, uniformly distributed active sites featuring Li–O interactions on Cu₂₀−Mo₁₅₄, which promote Li₂S nucleation and enrich and catalyze LiPSs, thereby accelerating deposition kinetics. The experimentally measured potentiostatic deposition curves were compared and fitted with the classical nucleation theory model (Figs. 4e and h). The results demonstrate that the reductive deposition and growth process of lithium sulfide (Li₂S) on both Cu₂₀−Mo₁₅₄ and Cu₂₀ electrodes is dominated by the two-dimensional progressive (2DI) nucleation mechanism. This nucleation mode corresponds to the formation of a continuous film-like Li₂S deposition layer (insets of Figs. 4e and h), which is in excellent agreement with the high-density and uniformly distributed active sites on the electrode surface. Both Cu₂₀−Mo₁₅₄ and Cu₂₀ support reversible Li₂S dissolution (Figs. 4f and i), indicating effective bidirectional catalysis that facilitates both deposition and electrochemical regeneration of Li₂S. Collectively, Cu₂₀−Mo₁₅₄ couples strong chemical anchoring for LiPSs (multipoint Li–O coordination at Mo₁₅₄) with efficient electronic conduction (Cu sites), delivering rapid and reversible LiPSs conversion, reduced impedance, accelerated Li₂S nucleation/growth via 2DI pathways, and mitigated passivation features that underpin superior rate capability and cycling stability in Li-S batteries.

  Figure 4

  

  Figure 4.  (a) CV curves for Li₂S₆ symmetric cells with the Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ electrodes. (b) CV curves of Cu₂₀−Mo₁₅₄ modified separator symmetric cell at different cycles. (c) The Tafel curves at the scan rate of 5 mV/s. Potentiostatic discharge at 2.08 V on (d) Cu₂₀−Mo₁₅₄ and (g) Cu₂₀. Dimensionless transient profiles at 2.08 V for (e) Cu₂₀−Mo₁₅₄ and (h) Cu₂₀, and the insets are the corresponding SEM images after Li₂S deposition. (f) Potentiostatic charge profiles at 2.35 V for Cu₂₀−Mo₁₅₄ and (i) Cu₂₀.

  DownLoad: Full-Size Img PowerPoint

  Li-S batteries were assembled with separators modified by Cu₂₀, Mo₁₅₄, and Cu₂₀−Mo₁₅₄, and their electrochemical performance was evaluated. As shown in Fig. 5a, all three cells exhibit the characteristic of CV features Li-S batteries, including two reduction peaks during discharge and one oxidation peak during charge. The peak currents follow the order Cu₂₀−Mo₁₅₄ > Mo₁₅₄ > Cu₂₀, indicating that Cu₂₀−Mo₁₅₄ has the highest reaction rate and the greatest density of active sites. The cathodic peak potentials obey the trend of Cu₂₀−Mo₁₅₄ > Mo₁₅₄ > Cu₂₀, suggesting that Cu₂₀−Mo₁₅₄ most effectively promotes the conversion of solid sulfur to soluble long-chain and medium-chain polysulfides and affords the fastest kinetics and best reversibility for subsequent Li₂S deposition. Consistently, the anodic peak of Cu₂₀−Mo₁₅₄ appears at a lower potential than those of Mo₁₅₄ and Cu₂₀, indicating that Li₂S is more readily oxidized and dissolved. Galvanostatic charge-discharge profiles (Fig. 5b) further corroborate these trends. The high-voltage discharge plateau (Q_H) of the Cu₂₀−Mo₁₅₄ cell occurs at a higher potential and shows largest Q_H (447 mAh/g, Fig. S17 in Supporting information), reflecting stronger adsorption and activation of LiPSs, together with improved electron transport, thereby accelerating sulfur dissolution and the formation/transition of long- and medium-chain LiPSs [44]. The low-voltage discharge plateau (Q_L) is the longest and flattest for Cu₂₀−Mo₁₅₄ and performs largest Q_L (1173 mAh/g, Fig. S17), indicating the highest sulfur utilization and the fastest deposition kinetics. These results collectively demonstrate that Cu₂₀−Mo₁₅₄ delivers superior Li₂S nucleation and growth, along with deep conversion capability, relative to Mo₁₅₄ and Cu₂₀. The cycling stability of Li-S batteries using modified separators prepared from different materials was evaluated. The Cu₂₀−Mo₁₅₄ cell exhibits the highest initial capacity of 1561.5 mAh/g, surpassing that of Mo₁₅₄ (1434.9 mAh/g) and Cu₂₀ (1022 mAh/g), and it retains 762.2 mAh/g after 300 cycles at a current density of 1 C (Fig. S18 in Supporting information). To further substantiate the enhanced electrochemical performance, the rate capability of Li-S cells was measured at current densities ranging from 0.1 C to 5.0 C and then returned to 0.1 C (Fig. 5c). Specifically, the Cu₂₀−Mo₁₅₄ battery delivers discharge capacities of 1626.4, 1152.8, 1040.5, 932.7, 850, 812, 795 and 1103 mAh/g at rates of 0.1, 0.2, 0.5, 1.0, 2.0, 3.0, 5.0 and 0.1 C, respectively (Fig. 5d), outperforming the other materials modified separator. In contrast, the Cu₂₀ electrode exhibits the poorest rate performance. There is no capacity release when the current density exceeds 2 C. It is worth noting that the Cu₂₀−Mo₁₅₄ battery delivers an initial capacity of 569.3 mAh/g at an ultrahigh current density of 10 C and maintains 317.5 mAh/g after 1100 cycles (Fig. 5e), corresponding to a low average capacity fade of 0.048% per-cycle. By comparison, the Mo₁₅₄ battery starts at 564.5 mAh/g and declines to 160.6 mAh/g after 1100 cycles, corresponding to an average capacity decay rate of 0.1% per cycle. However, the Cu₂₀ electrode does not exhibit any capacity release. The Cu₂₀ battery shows negligible effective capacity, which can be attributable to the rapid dissolution of Cu₂₀ in the electrolyte. The combination of Mo₁₅₄ with Cu₂₀ suppresses Cu₂₀ dissolution, enhances cycling stability, and increases the discharge capacity. The Cu₂₀−Mo₁₅₄ interface layer can effectively adsorb and catalyze the conversion of polysulfides, thus significantly inhibiting the shuttle effect, resulting in a stable coulombic efficiency of over 99.5% throughout the 1100 cycles. The self-discharge (Fig. S19 in Supporting information) tests of batteries indicates that the cell with Cu₂₀−Mo₁₅₄ modified separator shows only a slight open-circuit voltage (OCV) decay from 2.41 V to 2.36 V after resting 160 h. For comparison, the cells with individual Cu₂₀ and Mo₁₅₄ modified separators show voltage fades from 2.41 V to 2.04 V and from 2.42 V to 2.12 V, respectively. Thereby, the batteries with separators modified by Cu₂₀−Mo₁₅₄ layer on long-term storage performance. Overall, the Cu₂₀−Mo₁₅₄ modified separator delivers superior cycling stability and rate capability by coupling strong polysulfide adsorption and catalytic conversion with robust interfacial stability, thereby outperforming the Mo₁₅₄ and Cu₂₀ as well as commercial PP separators across abroad range of operating conditions.

  Figure 5

  

  Figure 5.  (a) CV curves and (b) galvanostatic charge/discharge profiles of Li-S cells with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators. (c) Rate capabilities of Li-S cells based on polypropylene separators (PP) and with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators at different current densities. (d) Discharge-charge profiles of cell with Cu₂₀−Mo₁₅₄ modified separator at different current rates. (e) Long cycle performance of cell with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators at current densities of 10 C and coulombic efficiency of Cu₂₀−Mo₁₅₄.

  DownLoad: Full-Size Img PowerPoint

  Cyclic voltammetry was conducted at scan rates from 0.1 mV/s to 1.0 mV/s to probe Li⁺ diffusion. According to the Randles-Sevcik equation, the peak current i_p scales linearly with ν^1/2, and the slope of i_p vs. ν^1/2 reflects the Li⁺ diffusion coefficient [45]. As shown in Figs. 6a and b, Figs. S20a and b (Supporting information), the Cu₂₀−Mo₁₅₄ modified separator exhibits the largest slopes for both anodic and cathodic peaks, indicating faster interfacial Li⁺diffusion during charge and discharge. This enhanced Li⁺ transport improves the redox kinetics and overall performance of Li-S battery [46]. The electrochemical behavior of the operating battery was analyzed using the galvanostatic intermittent titration technique (GITT) under near quasi-equilibrium conditions (Fig. 6c). From the GITT transients, the internal resistance (Figs. 6d and e) and the apparent Li⁺ diffusion coefficient (Figs. S21a and b in Supporting information) were determined during both charge and discharge. As summarized in, the Cu₂₀−Mo₁₅₄ modified separator exhibits the lowest internal resistance and the highest Li⁺ diffusion coefficient across the state of charge. Meanwhile, EIS measurements reveal that the cell with the Cu₂₀−Mo₁₅₄ modified separator exhibits a lower interfacial resistance compared to those with Cu₂₀ modified and Mo₁₅₄ modified separators, and the equivalent circuit in Fig. S22 (Supporting information) is employed to simulate the obtained ESI results, evidencing the fastest diffusion and phase-transition kinetics [47]. Mechanistically, polar sites in Mo₁₅₄ (e.g., Li–O coordination) strongly anchor and enrich polysulfides, increasing the local concentration of LiPSs and suppressing the shuttle effect, while Cu sites furnish rapid electronic pathways that lower charge-transfer resistance. The resulting synergy accelerates LiPSs conversion and transport. A Li||Li symmetric cell was employed to assess the cycling durability of lithium stripping/deposition under different materials modified separators, thereby exploring the impact of the modified layer on the uniformity of lithium deposition [48]. EIS before and after constant-potential polarization (Fig. 6f, Figs. S23a and b in Supporting information) enabled determination of the Li⁺ transference number t_Li⁺ according to Eq. S3 (Supporting information). In PP separators, Li⁺ migrates predominantly as large solvation clusters, leading to sluggish diffusion. In contrast, the high electronegativity of anionic clusters can facilitate Li⁺ transport [49]. When employing Cu₂₀−Mo₁₅₄ modified separators, the ion-dipole and electrostatic interactions between Mo₁₅₄ anionic clusters and Li⁺ lead to changes in the solvation structure of Li⁺. Consequently, the high negativity of the Mo₁₅₄ surface ensures high flux diffusion of Li⁺ and achieves uniform lithium deposition. Fig. 6g shows the cycling voltage curves at 1 mA/cm² with areal capacity of 1 mA/cm². The Li||Li symmetrical cell using the PP separator experienced a short circuit after approximately 340 h. The Cu₂₀ modified separator exhibited an overpotential of 330 mV during the initial 7 h, followed by pronounced polarization fluctuations due to Cu₂₀ dissolution. By comparison, the Mo₁₅₄ modified separator consistently maintained an overpotential of 99 mV during continuous cycling. The magnified plots clearly show that the Cu₂₀−Mo₁₅₄ cell exhibits a low and stable voltage hysteresis after the initial SEI formation, with no irregular voltage spikes, fluctuations, or sudden polarization increase throughout the steady and late cycle stages. This confirms that the Cu₂₀−Mo₁₅₄ interface layer can effectively regulate uniform Li deposition, stabilize the electrode-electrolyte interface, and inhibit the growth of Li dendrites, which is the core reason for the excellent long-cycle stability of the cell. Notably, the Cu₂₀−Mo₁₅₄ modified separator sustains consistent polarization for up to 3000 h, indicating high flux Li⁺ transport and uniform Li deposition. Together, accelerated ion and electron transport, stabilized solvation, uniform deposition, and shuttle suppression underpin the superior kinetics and durability of Cu₂₀−Mo₁₅₄ enabled Li-S batteries.

  Figure 6

  

  Figure 6.  (a) CV curves of Cu₂₀−Mo₁₅₄ modified separator at various scan rates. (b) Corresponding linear fitting for the peak current vs. the square root of the scanning rate of Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄. (c) GITT profiles and corresponding reaction resistances of cells with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators during (d) charge and (e) discharge processes. (f) Measurement of t_Li⁺ using potentiostatic polarization of Li||Li symmetric cell with Cu₂₀−Mo₁₅₄ modified separator (inset: Nyquist plots of impedance before and after polarization). (g) Galvanostatic cycling of Li||Li symmetric cells with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators at 1 mA/cm² with areal capacity of 1 mA/cm² and the inset shows magnified views of the galvanostatic cycling curves for the Cu₂₀−Mo₁₅₄ modified separator cell at different cycling stages.

  DownLoad: Full-Size Img PowerPoint

  In ex-situ XPS was used to investigate the kinetics of LiPSs conversion in Cu₂₀−Mo₁₅₄ during the charge-discharge process, confirming the synergistic effect of Cu₂₀ and Mo₁₅₄ in Li-S batteries (Figs. 7a-h). As illustrated in Fig. 7b, when the discharge voltage is approximately 2.10 V, the S 2p spectrum can be deconvoluted into five distinct peaks. The peaks at 169.6, 168.3, and 166.3 eV are corresponded to sulfate, polysulfides, and thiosulfate, respectively. The peaks at 162.2 and 160.0 eV correspond to Li₂S₂ and Li₂S, respectively [50]. Upon further discharge to 1.70 V, the Li₂S₂ and Li₂S signals intensify, consistent with the complete reduction of S₈ to Li₂S₂ and Li₂S. Concurrently (Fig. 7c), the Cu 2p_3/2 and Cu 2p_1/2 peaks shift to higher binding energies as the discharge proceeds, indicating oxidation of Cu (decreased electron density), whereas the Mo 3d peaks shift to lower binding energies (Fig. 7d), accompanied by an increasing fraction of Mo⁵⁺, evidencing the reduction of Mo centers. These opposite shifts signify electron transfer from Cu₂₀ to LiPSs and Mo₁₅₄ during discharge. In contrast, during the charging process, at the fully charged state (2.80 V), the S 2p (Fig. 7f) spectrum exhibits two dominant high binding energy features assigned to sulfate and polysulfide, indicating complete oxidation of long-chain LiPSs. In the case of the Cu₂₀ spectrum (Fig. 7g), both Cu 2p_3/2 and Cu 2p_1/2 shift to lower binding energies as charging progresses, consistent with Cu reduction. While the Mo 3d spectrum (Fig. 7h) shows the conversion of Mo⁵⁺ back to Mo⁶⁺, reflecting the oxidation of Mo. These trends indicate reverse electron flow from Mo₁₅₄ and LiPSs to Cu₂₀ during charging, with Mo₁₅₄ acting primarily to oxidize LiPSs. The binding energy shifts across S, Cu, and Mo demonstrate that Cu₂₀−Mo₁₅₄ operates as a redox mediating, synergistic catalyst. Cu₂₀ donates electrons (via hydrogen bond) to activate LiPSs while Mo₁₅₄ is reduced during discharge. Mo₁₅₄ promotes LiPSs oxidation and returns electrons toward Cu₂₀ during charge. This bidirectional electron mediation accelerates polysulfide redox kinetics, underpinning the high-rate capability and long-term cycling stability observed in Li-S battery with the Cu₂₀−Mo₁₅₄ modified separator.

  Figure 7

  

  Figure 7.  (a) Galvanostatic discharge profiles and corresponding ex-situ XPS analysis for (b) S 2p, (c) Cu 2p and (d) Mo 3d spectra of the cell with Cu₂₀−Mo₁₅₄ modified separator. (e) Galvanostatic charge profiles and corresponding ex-situ XPS analysis for (f) S 2p, (g) Cu 2p and (h) Mo 3d spectra of the cell with Cu₂₀−Mo₁₅₄ modified separator.

  DownLoad: Full-Size Img PowerPoint

  The Li-S electrochemical performance using the Cu₂₀−Mo₁₅₄ modified separator is shown in Fig. 8. The Cu₂₀−Mo₁₅₄ modified layer functions as both a physical barrier and a chemical catalyst for LiPSs. Physically, it hinders LiPSs migration across the separator. Chemically, Cu₂₀ and Mo₁₅₄ act synergistically to anchor LiPSs and catalyze their reversible transformation (LiPSs ⇌ Li₂S) during cycling. Specifically, Mo₁₅₄ captures LiPSs via Li–O coordination at terminal oxygens (Mo=O), while Cu₂₀ provides additional Li–O coordination through hydroxyl groups of surface ligands. Together, these interactions effectively suppress the shuttle effect, enhance active sulfur utilization, and mitigate corrosion of the lithium anode. During discharge, electrons are transferred from Cu₂₀ to Mo₁₅₄ and LiPSs, which accelerates the conversion of solid S₈ to Li₂S and markedly improves the catalytic reduction kinetics of polysulfides. During charge, Mo₁₅₄ and LiPSs transfer electrons to Cu₂₀, promoting LiPSs oxidation, lowering internal resistance, and enhancing overall reaction efficiency. Consistent with the in situ XPS trends in Fig. 7, these reciprocal electron flows confirm that Cu₂₀ and Mo₁₅₄ operate as a cooperative redox mediator pair. Beyond catalysis, Cu₂₀ serves as an "electron buffer reservoir" for Mo₁₅₄, storing and supplying electrons to maintain an appropriate electronic density on the Mo₁₅₄ surface. This stabilizes the local electronic structure of Mo₁₅₄, reduces the likely of surface and structural degradation during operation, and enhances the stability of the Cu₂₀−Mo₁₅₄ composite. Collectively, the integrated physical obstruction, multisite chemical adsorption, bidirectional redox mediation, and electron reservoir buffering significantly improve Li-S battery performance, making this system more advantageous for practical applications.

  Figure 8

  

  Figure 8.  Schematic illustration for the catalytic mechanism of the high-performance Li-S cell equipped with Cu₂₀−Mo₁₅₄ modified separator.

  DownLoad: Full-Size Img PowerPoint

  In this study, we successfully constructed a novel synergistic catalyst Cu₂₀−Mo₁₅₄, composed of atomically precise copper nanoclusters and giant Keplerian molybdenum clusters, designed to accelerate LiPSs conversion and enhance the electrochemical performance of Li-S batteries. The synergy between Cu₂₀ and Mo₁₅₄ provides more active surface sites for LiPSs conversion, increasing the available surface area for catalytic reactions and thereby enhancing the overall catalytic efficiency. During the charging and discharging process, rapid electron transfer occurs between Cu₂₀ and Mo₁₅₄, maintaining an optimal electron density on Cu₂₀ and Mo₁₅₄, which enhancing catalytic activity and improving the structural stability of Cu₂₀−Mo₁₅₄. A high reversible capacity of 1355 mAh/g is observed at 0.1 C, and an excellent cycling stability of 317 mAh/g remains even after 1100 cycles at 10 C. It is confirmed that the synergistic effect between Cu₂₀ and Mo₁₅₄ enhances the bidirectional catalytic conversion rate of LiPSs ⇌ Li₂S. This study offers a viable route for designing efficient cluster-to-cluster synergistic catalysts, with the goal of suppressing the shuttling effect of LiPSs in Li-S batteries and realizing excellent rate capability.

  CRediT authorship contribution statement

  Caili Lv: Writing – review & editing, Methodology, Funding acquisition. Yihai Song: Data curation. Yi Feng: Data curation. Tianbo Li: Formal analysis. Yun Dong Cao: Writing – review & editing, Software, Funding acquisition. Linlin Fan: Funding acquisition, Formal analysis. Hong Liu: Writing – review & editing, Software, Formal analysis. Jiayuan Zhang: Data curation. Guanggang Gao: Writing – original draft, Methodology, Funding acquisition.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (No. 22471095), Natural Science Foundation of Shandong Province (Nos. ZR2024QE440 and ZR2024QB211), Postdoctoral Fellowship Program of CPSF (No. GZC20250015), Postdoctoral Innovation Program of Shandong Province (No. SDCX-ZG-202502050). the Jinan City "New University 20″ Project (No. 202228113), and Science and Technology Program of University of Jinan (No. XBS2404)

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2026.112867.

  
  
• 1. [1]

     C.Y. Zhang, X. Lu, X. Han, et al., J. Am. Chem. Soc. 145 (2023) 18992–19004. doi: 10.1021/jacs.3c06288

  2. [2]

     Y. Lei, C. Wu, X. Lu, et al., Angew. Chem. Int. Ed. 61 (2022) e202200384. doi: 10.1002/anie.202200384

  3. [3]

     C. Ma, W. Cui, Q. Zhang, et al., Adv. Funct. Mater. 36 (2026) e15760. doi: 10.1002/adfm.202515760

  4. [4]

     J. Li, Y. Li, T. Liu, et al., Adv. Funct. Mater. 36 (2026) 2504546. doi: 10.1002/adfm.202504546

  5. [5]

     H. Cho, J. Jung, I. Kim, et al., Nat. Commun. 16 (2025) 6805. doi: 10.1038/s41467-025-62169-z

  6. [6]

     Y. Zhang, Y. Cao, L. Fan, et al., Chem. Eng. J. 482 (2024) 148991. doi: 10.1016/j.cej.2024.148991

  7. [7]

     J.Y. Zhang, X.Y. Zhang, J. Wang, et al., Adv. Funct. Mater. 34 (2024) 2400170. doi: 10.1002/adfm.202400170

  8. [8]

     Y. Shi, S. Wang, P. Cui, et al., Sci. Bull. 70 (2025) 2094–2103. doi: 10.1016/j.scib.2025.05.005

  9. [9]

     C. Wang, X. Duan, Y. Jiang, et al., Adv. Mater. 37 (2025) 2500114. doi: 10.1002/adma.202500114

  10. [10]

      Y.J. Zhu, Y.M. Cui, N.H. Wang, et al., Inorg. Chem. 64 (2025) 19113–19119. doi: 10.1021/acs.inorgchem.5c03641

  11. [11]

      L. Ni, J. Gu, X. Jiang, et al., Angew. Chem. Int. Ed. 62 (2023) e202306528. doi: 10.1002/anie.202306528

  12. [12]

      L. Wu, Z. Jin, X. Xie, et al., Energy Storage Mater. 77 (2025) 104167.

  13. [13]

      A. Müller, J. Meyer, E. Krickemeyer, E. Diemann, Angew. Chem. Int. Ed. 35 (1996) 1206–1208. doi: 10.1002/anie.199612061

  14. [14]

      A. Müller, S.K. Das, V.P. Fedin, et al., Z. Anorg. Allg. Chem. 625 (1999) 1187–1192.

  15. [15]

      J. Liu, N. Jiang, J.M. Lin, et al., Angew. Chem. 135 (2023) e202304728. doi: 10.1002/ange.202304728

  16. [16]

      M. Hou, X. Li, C. Zhao, et al., Adv. Funct. Mater. 35 (2025) e10454. doi: 10.1002/adfm.202510454

  17. [17]

      J.M. Lin, Z.B. Mei, C. Guo, et al., J. Am. Chem. Soc. 146 (2024) 22797–22806. doi: 10.1021/jacs.4c08043

  18. [18]

      L. Ni, J. Gu, X. Jiang, et al., Angew. Chem. 135 (2023) e202306528. doi: 10.1002/ange.202306528

  19. [19]

      M. Bodiuzzaman, K. Murugesan, P. Yuan, et al., J. Am. Chem. Soc. 146 (2024) 26994–27005. doi: 10.1021/jacs.4c08688

  20. [20]

      Z. Liu, J.J. Fang, Z.Y. Wang, Y.P. Xie, X. Lu, Chem. Eur. J. 31 (2025) e202500230. doi: 10.1002/chem.202500230

  21. [21]

      N. Ma, Q. Wang, D. Zhao, et al., Chin. Chem. Lett. 37 (2026) 112356. doi: 10.1016/j.cclet.2025.112356

  22. [22]

      S. Li, N.N. Li, X.Y. Dong, S.Q. Zang, T.C.W. Mak, Chem. Rev. 124 (2024) 7262–7378. doi: 10.1021/acs.chemrev.3c00896

  23. [23]

      K. Liu, Z. Wang, Y. Xie, et al., Chin. Chem. Lett. 37 (2026) 111701. doi: 10.1016/j.cclet.2025.111701

  24. [24]

      R. Guo, R.W. Murray, J. Am. Chem. Soc. 127 (2005) 12140–12143. doi: 10.1021/ja053119m

  25. [25]

      M. Chi, J. Wei, Z. Zhao, et al., Angew. Chem. 137 (2025) e202424499. doi: 10.1002/ange.202424499

  26. [26]

      J. Zhang, Y. Wu, Y. Li, et al., Chin. Chem. Lett. 37 (2026) 111003. doi: 10.1016/j.cclet.2025.111003

  27. [27]

      X. Su, Y. Zhang, X. Dai, et al., Chin. Chem. Lett. (2026), doi: 10.1016/j.cclet.2026.112539.

  28. [28]

      K. Song, X. Wang, Z. Xie, et al., Angew. Chem. Int. Ed. 62 (2023) e202216450. doi: 10.1002/anie.202216450

  29. [29]

      Y.D. Cao, H.P. Hao, H.S. Liu, et al., Nanoscale 13 (2021) 16182–16188. doi: 10.1039/d1nr04683g

  30. [30]

      S. Zheng, D.H. Si, H. Guo, et al., ACS. Nano 19 (2025) 24130–24139. doi: 10.1021/acsnano.5c07703

  31. [31]

      G. Zhou, H. Chen, Y. Cui, Nat. Energy 7 (2022) 312–319. doi: 10.1038/s41560-022-01001-0

  32. [32]

      W. Lee, D. Um, Y. Baek, et al., Angew. Chem. Int. Ed. 64 (2025) e202506861. doi: 10.1002/anie.202506861

  33. [33]

      L. Li, Y. Li, H.Y. Zang, et al., Angew Chem. Int. Ed. 65 (2026) e17583. doi: 10.1002/anie.202517583

  34. [34]

      Y. Li, L. Wang, F. Zhang, et al., Adv. Sci. 10 (2023) 2205020. doi: 10.1002/advs.202205020

  35. [35]

      Y. Zhang, H. Guo, J. Ren, et al., Appl. Catal. B 298 (2021) 120582. doi: 10.1016/j.apcatb.2021.120582

  36. [36]

      J.L. Yang, S.X. Zhao, Y.M. Lu, et al., Nano Energy 68 (2020) 104356. doi: 10.1016/j.nanoen.2019.104356

  37. [37]

      Y. Liu, Q. Zhao, W. Du, et al., Chin. Chem. Lett. (2026), doi: 10.1016/j.cclet.2026.112608.

  38. [38]

      M. Du, P. Geng, C. Pei, et al., Angew. Chem. Int. Ed. 61 (2022) e202209350. doi: 10.1002/anie.202209350

  39. [39]

      A. Müller, C. Serain, Acc. Chem. Res. 33 (2000) 2–10. doi: 10.1021/ar9601510

  40. [40]

      X. Yao, C. Guo, C. Song, et al., Adv. Mater. 35 (2023) 2208846. doi: 10.1002/adma.202208846

  41. [41]

      H.M. Zeng, M. Cheng, W. Xuan, Y. Wei, Inorg. Chem. 64 (2025) 67–72. doi: 10.1021/acs.inorgchem.4c04168

  42. [42]

      Y. Xu, Y. Jiang, X. Yu, et al., Adv. Sci. 12 (2025) e13282. doi: 10.1002/advs.202513282

  43. [43]

      J. Yang, C. Cao, Z. Li, et al., Chem. Eng. J. 476 (2023) 146532. doi: 10.1016/j.cej.2023.146532

  44. [44]

      S. Jing, X. Peng, S. Li, et al., Chin. Chem. Lett. 36 (2025) 110732. doi: 10.1016/j.cclet.2024.110732

  45. [45]

      J. Song, S. Xia, N. Wang, et al., Adv. Mater. 37 (2025) 2418295. doi: 10.1002/adma.202418295

  46. [46]

      L. Du, J. Sun, Y. Zhang, et al., Adv. Mater. 37 (2025) e08523. doi: 10.1002/adma.202508523

  47. [47]

      Y. Liu, C. Zeng, M. Hu, et al., Adv. Energy Mater. 15 (2025) 2405960. doi: 10.1002/aenm.202405960

  48. [48]

      C. Sun, J.P. Chen, Y.L. Wu, et al., Angew. Chem. Int. Ed. 137 (2025) e202506533. doi: 10.1002/ange.202506533

  49. [49]

      Y. Liu, X. Zhou, T. Qiu, et al., Adv. Mater. 36 (2024) 2407705. doi: 10.1002/adma.202407705

  50. [50]

      M. Xu, T. Deng, C. Li, et al., Chin. Chem. Lett. 36 (2025) 110372. doi: 10.1016/j.cclet.2024.110372

• 

  Figure 1  (a) Schematic illustration for the synthesis process of the Cu₂₀−Mo₁₅₄. (b) Atomic fractions on Cu₂₀−Mo₁₅₄. (c) HRTEM images of Cu₂₀−Mo₁₅₄. (d) FTIR spectra of Cu₂₀−Mo₁₅₄. XPS patterns of (e) Cu 2p and (f) Mo 3d. (g) CV curves of different materials.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) The schematic illustration that Cu₂₀−Mo₁₅₄ suppresses LiPSs shuttling. (b) electronic photograph of different material immersion in the Li₂S₆ solution for 9 h. (c) UV–vis spectra of different samples after immersion in the Li₂S₆ solution for 9 h. XPS analysis of (d) Cu 2p, (e) Mo 3d and (f) O 1s of Cu₂₀−Mo₁₅₄ and Cu₂₀−Mo₁₅₄@Li₂S₆. (g) XPS analysis of S 2p spectra for Li₂S₆ and Cu₂₀−Mo₁₅₄@Li₂S₆.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) The structure of sandwiched skeleton in Cu₂₀−Mo₁₅₄ and constituent structural unit for Mo₁₅₄ ({Mo₅}, {Mo₈}) and Cu₂₀ ({Cu₃}). (b) Optimized adsorption configurations of various LiPSs on unit of {Mo₅}, {Mo₈} and {Cu₃}. (c) Binding energies between various LiPSs on unit of {Mo₅}, {Mo₈} and {Cu₃}.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) CV curves for Li₂S₆ symmetric cells with the Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ electrodes. (b) CV curves of Cu₂₀−Mo₁₅₄ modified separator symmetric cell at different cycles. (c) The Tafel curves at the scan rate of 5 mV/s. Potentiostatic discharge at 2.08 V on (d) Cu₂₀−Mo₁₅₄ and (g) Cu₂₀. Dimensionless transient profiles at 2.08 V for (e) Cu₂₀−Mo₁₅₄ and (h) Cu₂₀, and the insets are the corresponding SEM images after Li₂S deposition. (f) Potentiostatic charge profiles at 2.35 V for Cu₂₀−Mo₁₅₄ and (i) Cu₂₀.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) CV curves and (b) galvanostatic charge/discharge profiles of Li-S cells with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators. (c) Rate capabilities of Li-S cells based on polypropylene separators (PP) and with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators at different current densities. (d) Discharge-charge profiles of cell with Cu₂₀−Mo₁₅₄ modified separator at different current rates. (e) Long cycle performance of cell with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators at current densities of 10 C and coulombic efficiency of Cu₂₀−Mo₁₅₄.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) CV curves of Cu₂₀−Mo₁₅₄ modified separator at various scan rates. (b) Corresponding linear fitting for the peak current vs. the square root of the scanning rate of Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄. (c) GITT profiles and corresponding reaction resistances of cells with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators during (d) charge and (e) discharge processes. (f) Measurement of t_Li⁺ using potentiostatic polarization of Li||Li symmetric cell with Cu₂₀−Mo₁₅₄ modified separator (inset: Nyquist plots of impedance before and after polarization). (g) Galvanostatic cycling of Li||Li symmetric cells with Cu₂₀, Mo₁₅₄ and Cu₂₀−Mo₁₅₄ modified separators at 1 mA/cm² with areal capacity of 1 mA/cm² and the inset shows magnified views of the galvanostatic cycling curves for the Cu₂₀−Mo₁₅₄ modified separator cell at different cycling stages.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  (a) Galvanostatic discharge profiles and corresponding ex-situ XPS analysis for (b) S 2p, (c) Cu 2p and (d) Mo 3d spectra of the cell with Cu₂₀−Mo₁₅₄ modified separator. (e) Galvanostatic charge profiles and corresponding ex-situ XPS analysis for (f) S 2p, (g) Cu 2p and (h) Mo 3d spectra of the cell with Cu₂₀−Mo₁₅₄ modified separator.

  下载: 全尺寸图片 幻灯片
  

  Figure 8  Schematic illustration for the catalytic mechanism of the high-performance Li-S cell equipped with Cu₂₀−Mo₁₅₄ modified separator.

  下载: 全尺寸图片 幻灯片
•