English

• Lithium-sulfur batteries (LSBs) have captured significant attention in the field of electrochemical energy storage devices for the high theoretical specific capacity (1675 mAh/g) [1–3]. However, the practical application of LSBs is still limited by the notorious shuttle effect and the sluggish reaction kinetics of lithium polysulfides (Li₂S_x, 4 ≤ x ≤ 6; LiPSs) during the cycling process, which always leads to inferior cycling stability and poor rate capability [4–6].

  Generally, the capacity contribution originating from the conversion reaction between LiPSs and Li₂S can achieve 75% [7], and facilitating this conversion reaction is of significance to develop high-performance LSBs, which is severely restricted by the electrochemical inertia and intrinsic insulation property of LiPSs [8,9]. Therefore, designing catalytic sulfur-carrier materials is an urgent need to promote the conversion reaction kinetics from LiPSs to Li₂S [10–12]. Modulating the active sites and band structure configuration of sulfur-host materials is crucial factor to enhance the catalytic activity and chemical adsorption behavior for LiPSs at the atomic level [13], which can be rationally realized by introducing structural defects, mainly including heteroatom-vacancy and heteroatom doping [14–17]. As demonstrated in the previous reports, the oxygen-defect sites in the crystal structure of TiO₂ effectively improved the adsorption capability of LiPSs [18,19], while the introduced iron atoms into nitrogen-doped graphene firmly bonded with the LiPSs through the interaction of Fe-S bond [20]. Additionally, the introduction of nitrogen-doping into CoP greatly weakened the bond strength between S-S bond in Li₂S₄ and Li-S bond in Li₂S, which contributed to facilitate the dissolution behavior of Li₂S and ultimately boost the conversion reaction kinetics of LiPSs [21,22]. Though the structural defects induced by the single heteroatom-vacancy or single heteroatom-doping can improve the adsorption capability and reaction kinetics of LiPSs to a certain extent, it still cannot meet the high-requirements of high-sulfur loading cathodes for the practical application of LSBs. Therefore, it is relentless demand but great challenge to elaborately design the advanced sulfur-host catalysts with more significant catalytic activity, stronger polysulfide adsorption capability and stronger inhibition of polysulfide shuttle ability [23–25].

  In this work, double defect of boron-doping and phosphorus-vacancy is elaborately realized in the B-MoP_1-x@NC catalyst by soaking the MoP@NC powder into the NaBH₄ solution. As demonstrated by the materials characterization and theoretical calculations, the dual-defect coordination in B-MoP_1-x@NC can effectively guarantee the enlarged lattice spacing, the enhanced adsorption capability for LiPSs, the reduced lithium diffusion energy barriers and the low Gibbs free energy for the conversion reactions of LiPSs. Benefiting from these merits, B-MoP_1-x@NC-based electrode delivers the lowest Li₂S oxidation overpotential, the highest lithium diffusivity and the smallest Tafel slope, which is conducive to promote the conversion reaction kinetics of LiPSs and improve the corresponding specific capacity, rate capability as well as cycling stability. In details, the assembled Li-S/B-MoP_1-x@NC cells exhibit reversible capacity of 753 mAh/g at 0.5 C after 300 cycles, and high specific capacity of 621 mAh/g at high rate of 2 C. Even at the high-loading condition (5.1 mg/cm²), S/B-MoP_1-x@NC cathode also can achieve a high capacity of 520 mAh/g after 100 cycles. Based on the in-situ UV–vis spectra, the superior battery performance can be attributed to the effective inhibition of polysulfides shuttle and strong adsorption capability for polysulfides induced by the double-defect in B-MoP_1-x@NC catalyst. This study makes it possible to design dual-defect catalysts for LSBs and promote the basic research on the application of high-performance catalytic sulfur-host cathode materials.

  The research subject of B-MoP_1-x@NC was prepared by calcining the PPy-PMo₁₂ precursor and immersing the calcined product into NaBH₄ solution. The scanning electron microscope (SEM) (Fig. S1a in Supporting information) and transmission electron microscope (TEM) images (Fig. 1a) verify the uniform spherical sea urchin-like morphology of B-MoP_1-x@NC nanoparticles, and the High-Angle Annular Dark Field (HADDF) and corresponding element mapping images confirm the uniform distribution of Mo/P/B elements on the surface of B-MoP_1-x@NC sample (Fig. 1b). The high-resolution transmission electron microscope (HRTEM) image represents the high crystallinity region of B-MoP_1-x@NC sample (Fig. 1c), which can be assigned to the characteristic (101) peak of MoP. The corresponding interplanar spacing of B-MoP_1-x@NC is 0.25 nm, larger than the value (0.22 nm) of standard MoP crystal (PDF#24–0771) owing to the existence of phosphorus vacancies, which can induce the rearrangement of adjacent atoms in the crystal structure of B-MoP_1-x@NC, and promote the catalytic performance as well as adsorption behavior for LiPSs [26]. As clearly demonstrated by the scanning transmission electron microscopy (STEM) image with dual spherical aberration correction (Fig. 1d), the P atoms are darker than Mo atoms due to the different atomic number, and the locally disordered arrangements of P and Mo atoms imply the introduction of phosphorus defects (Figs. 1e and f). X-ray diffraction (XRD) patterns validate the crystal structure difference of B-MoP_1-x@NC and MoP@NC samples lies in the slight shifts of diffraction angle (2θ) caused by the introduction of phosphorus vacancies (Figs. 1g and h). According to the Bragg's formula, the leftward shift of 2θ indicates the enlarged lattice spacing in the crystal structure of B-MoP_1-x@NC, which can be corroborated by the corresponding HRTEM image (Fig. 1c) [27]. To investigate the characteristics of NC carrier, Raman spectra of B-MoP_1-x@NC, MoP@NC and NC samples were conducted. As shown in Fig. S1b (Supporting information), two distinct peaks corresponding to the D-band and G-band of carbon materials can be observed, and the higher intensity of D-band in these samples manifests the amorphous structure of NC components [28].

  Figure 1

  

  Figure 1.  (a) TEM image, (b) STEM image and the corresponding element mapping images, and (c) HRTEM image of B-MoP_1-x@NC sample. (d) STEM-HADDF image and the corresponding P vacancy evidences of B-MoP_1-x@NC sample. (e) The intensity of Mo and P atomic columns along red dotted line in (d). (f) The theoretical model of MoP. (g) XRD patterns of MoP@NC and B-MoP_1-x@NC samples. (h) Magnified image of (101) diffraction peaks of MoP@NC and B-MoP_1-x@NC samples; High-resolution of (i) Mo 3d, (j) P 2p and (k) B 1s spectra of MoP@NC and B-MoP_1-x@NC samples. (l) EPR signals of MoP@NC and B-MoP_1-x@NC samples.

  DownLoad: Full-Size Img PowerPoint

  X-ray photoelectron spectroscopy (XPS) were employed to study the chemical valences of constituent elements in B-MoP_1-x@NC and MoP@NC samples. Notably, the high-resolution Mo 3d spectra of both B-MoP_1-x@NC and MoP@NC samples (Fig. 1i) deliver the presence of Mo-P bonds, while the higher binding energies and oxidation states of Mo 3d_5/2 and Mo 3d_3/2 in MoP@NC than those in B-MoP_1-x@NC reflect the influences of phosphorus vacancies on the valence states of Mo atoms. According to the corresponding high-resolution P 2p spectra (Fig. 1j), both the intensity and binding energy of Mo-P bonds in B-MoP_1-x@NC sample are weaker than those in MoP@NC samples, suggesting the generation of phosphorus vacancies in B-MoP_1-x@NC. Fig. 1k exhibits the high-resolution B 1s spectra in B-MoP_1-x@NC, and a distinct peak at 187.8 eV demonstrates the successful B-doping in the crystal structure of MoP. Additionally, the high-resolution C 1s and N 1s spectra (Figs. S2a and b in Supporting information) verify the same N-doping forms in B-MoP_1-x@NC and MoP@NC samples [29]. Fig. S2c (Supporting information) presents the high-resolution spectrum of S 2p for the S/B-MoP_1-x@NC material, which can be categorized into three distinct characteristic peaks. The peak observed at 164.5 eV is attributed to the Mo-S bond, while the peak at 165.7 eV indicates the presence of sulfur or S₈ within the composite. Additionally, the peak at 172 eV corresponds to the formation of SO₄²⁻. During heat treatment, a certain amount of free hydroxyl groups may react with S₈ to generate SO₄²⁻ [30]. Moreover, electron paramagnetic resonance (EPR) test was conducted to further check the attendance of phosphorus defects. Compared to the flat EPR signal in MoP@NC (Fig. 1l), B-MoP_1-x@NC sample delivers a lone electron peak with G value of 1.925 resulting from the partial removal of phosphorus atoms and the residual phosphorus vacancies [28].

  The LiPSs adsorption capability of NC, MoP@NC and B-MoP_1-x@NC samples was evaluated by soaking them into the Li₂S₆ solution and standing for 6 h. As a result, the initial brown Li₂S₆ solution is almost transformed into a colorless solution with the addition of B-MoP_1-x@NC powder (Fig. S3a in Supporting information), and the corresponding UV–vis spectra of B-MoP_1-x@NC-Li₂S₆ solution holds the weakest peak intensity (Fig. S3b in Supporting information). These phenomenon manifests the adsorption capability of B-MoP_1-x@NC for Li₂S₆ is stronger than that of NC and MoP@NC powder, which contributes to mitigate the shuttle effect of polysulfide during cycling process for LSBs. After carefully analyzing the high-resolution Mo 3d spectra of B-MoP_1-x@NC-Li₂S₆ (Fig. 2a), an emerging peak at 226 eV is identified as the Mo-S bond, which can be attributed to the chemisorption effect of B-MoP_1-x@NC for polysulfide species [31,32]. Density Functional Theory (DFT) calculations were conducted to quantify the adsorption capability of different LiPSs on the surface of MoP@NC and B-MoP_1-x@NC models (Fig. 2b) [33]. As summarized in Fig. S3c (Supporting information), the corresponding adsorption energies of B-MoP_1-x@NC for S₈, Li₂S₈, Li₂S₆, Li₂S₄, Li₂S₂ and Li₂S components are stronger than those of MoP@NC, which is consistent with the experimental results (Figs. S3a and b). Fig. 2c displays the differential charge distributions of Li₂S₆ adsorbed on the (101) crystal surface of MoP@NC and B-MoP_1-x@NC, in which the yellow and cyan regions mean the electron accumulation and electron depletion regions [34]. Notably, the reduced charge density and weak Li-S bonds around the cyan isosurfaces of B-MoP_1-x@NC effectively favor the conversion reaction of Li₂S₆ towards lower-order LiPSs. Meanwhile, the yellow isosurface means a strengthened charge density, suggesting the strong interaction between B-MoP_1-x@NC and Li atoms of Li₂S₆ [35–37].

  Figure 2

  

  Figure 2.  (a) High-resolution Mo 3d spectra of the pristine B-MoP_1-x@NC catalyst and the composite of B-MoP_1-x@NC-Li₂S₆ sample. (b) LiPSs adsorption models on the surface of MoP@NC and B-MoP_1-x@NC catalysts. (c) Differential charge distributions of Li₂S₆ adsorbed on the (101) crystal plane of MoP@NC and B-MoP_1-x@NC (yellow regions: electron accumulation; cyan regions: electron depletion). (d) Li⁺ ions diffusion energy barriers, (e) Li₂S₆ cluster migration energy barriers, and (f) Li₂S cluster decomposition energy barriers on the (101) crystal plane of MoP@NC and B-MoP_1-x@NC catalysts. (g) The calculated Gibbs free energy for the conversion reactions of LiPSs driven by MoP@NC and B-MoP_1-x@NC catalysts.

  DownLoad: Full-Size Img PowerPoint

  DFT calculations was also employed to assess the migration kinetics of Li⁺ and Li₂S₆ on the (101) plane of MoP@NC and B-MoP_1-x@NC. Consequently, the Li⁺ ions (Fig. 2d) and Li₂S₆ cluster (Fig. 2e) diffusion energy barrier in B-MoP_1-x@NC are calculated to be 0.25 and 0.29 eV, respectively. Both of them are lower than those in MoP@NC (0.54 and 0.43 eV), which can significantly facilitate the diffusion behavior of Li⁺ and Li₂S₆ cluster. In addition, the corresponding energy barriers for the Li₂S decomposition reaction of MoP@NC and B-MoP_1-x@NC are 0.76 and 0.49 eV (Fig. 2f), respectively, which can effectively promote the redox reaction reversibility within Li₂S and LiPSs on the surface of B-MoP_1-x@NC. Fig. 2g displays the calculated Gibbs free energies for the conversion reactions of LiPSs driven by MoP@NC and B-MoP_1-x@NC catalysts. Notably, the corresponding Gibbs free energies of B-MoP_1-x@NC catalyst throughout the conversion reactions of S₈ → Li₂S₈ (−2.98 eV), Li₂S₈ → Li₂S₆ (−0.52 eV), Li₂S₆ → Li₂S₄ (−0.40 eV), Li₂S₄→Li₂S₂ (0.68 eV), and Li₂S₂→Li₂S (1.38 eV), are lower than those of MoP@NC catalyst, indicating the reduced conversion reaction barriers and the facilitated conversion reaction kinetics [38,39]. It is worth mentioning that the highest Gibbs free energy (1.38 eV) of B-MoP_1-x@NC catalyst for the conversion reaction from Li₂S₂ to Li₂S, which is the rate-determining step among the conversion reactions of LiPSs [40,41].

  Figs. 3a and b schematically illustrates the internal action mechanism of B-doping and P-defect in B-MoP_1-x on the conversion reaction kinetics of LiPSs. For the MoP@NC catalyst, the spatial effect of central subshell hinders effective hybridization between the Mo 3d orbitals in MoP and the S 3p orbitals in LiPSs. However, the interaction between LiPSs and MoP@NC primarily relies on the orbital coupling (referred to as s-p hybridization) between Li 1s orbitals in LiPSs and P 2p orbitals in MoP, which falls short in binding LiPSs. To enhance the adsorption capability for LiPSs, simultaneous introduction of B-doping and P-vacancy (P_v) with dual-site coordination is employed to provide sulfurophilic and lithiophilic sites. Meanwhile, the existence of P-vacancy also enhances the bond interaction between s-p hybridization. Fig. 3c delivers the influences of band engineering on the P-band elevation and s-p bond hybridization, demonstrating the center of P-band in B-MoP_1-x@NC is higher than that in MoP@NC. Moreover, the bonding state (s-p) and anti-bonding state (s-p*) arise with an increase in the center of the P-band, leading to an upward shift of s-p*. Notably, occupying fewer anti-bonding states corresponds to a higher energy level for s-p*, implying B-MoP_1-x@NC exhibits more efficient s-p hybridization than MoP@NC. In brief, the introduction of B-doping embellishes the alignment of atomic surface orbitals with LiPSs, which in turn changes the orientation of S-coupling states. Figs. 3d and e illustrate the density of states (DOS) of MoP@NC and B-MoP_1-x@NC, revealing the B-doping and P-vacancy effectively alters the energy band structure [42]. Additionally, B-dopants exhibit a sp hybridization state with vacant orbitals that are oriented at right angles to the plane (Fig. 3f). In contrast to the unbefitting orbital orientation (P_x and P_y) of S atom in MoP@NC, the vertical vacancy orbital allows for enhanced electron coupling with the intermediate S atoms [43,44].

  Figure 3

  

  Figure 3.  Schematic illustration the differences in action mechanism between (a) MoP and (b) B-MoP_1-x catalysts on the conversion reaction kinetics of LiPSs. (c) The influences of band engineering on the P-band elevation and s-p hybridization. Density of states (DOS) of (d) MoP@NC and (e) B-MoP_1-x@NC. (f) The alignment of atomic surface orbitals with LiPSs and the orientation of S-coupling modified by the introduction of boron dopants.

  DownLoad: Full-Size Img PowerPoint

  To investigate the oxidation behavior of Li₂S in different catalysts, the corresponding linear sweep voltammetry curves were conducted with B-MoP_1-x@NC/MoP@NC/NC samples as the working electrode, Pt foil as the counter electrode, Ag/AgCl as the reference electrode, and Li₂S-containing methanol solution as the electrolyte. Consequently, B-MoP_1-x@NC electrode displays the highest response current and the lowest initial potential (Fig. S4a in Supporting information), manifesting the lowest Li₂S oxidation overpotential and enhanced reaction kinetics. The catalytic activity of different sulfur carrier materials was determined by testing the cyclic voltammetry (CV) curves of the corresponding B-MoP_1-x@NC/MoP@NC/NC-based symmetrical cells with Li₂S₆ as the LiPSs within −1.0 V to 1.0 V. As a result, B-MoP_1-x@NC electrode presents the highest peak current density and the largest enclosed CV area (Fig. S4b in Supporting information), suggesting the fast polysulfide conversion reaction and high catalytic activity of B-MoP_1-x@NC material. Meanwhile, the charge transfer behavior at the interface of different catalysts and Li₂S₆-based electrolyte can be disclosed by testing the electrochemical impedance spectroscopy (EIS) curves of symmetrical cells. As expected, B-MoP_1-x@NC electrode exhibits the smallest charge transfer resistance (Fig. S4c in Supporting information), indicating the facilitated interfacial reaction kinetics. Furthermore, the corresponding Tafel curves of B-MoP_1-x@NC/MoP@NC/NC-based symmetrical cells also can reveal the enhanced LiPSs conversion reactions (S₈ to Li₂S) catalyzed by B-MoP_1-x@NC owing to its highest current density (Fig. S4d in Supporting information).

  To verify the effects of double defect engineering on the electrochemical performances of practical LSBs, the sulfur is elaborately fixed into the carrier materials by solid phase melting method to prepare the cathodes of S/B-MoP_1-x@NC, S/MoP@NC and S@NC [45]. As demonstrated by Fig. S5 (Supporting information), high specific surface area and abundant pore structure are conducive to the efficient sulfur fixation process, and the fixed sulfur always plugs the original pore structure of carrier material. As derived from the thermogravimetric (TG) analysis curve of S/B-MoP_1-x@NC sample (Fig. S6 in Supporting information), the calculated sulfur content is about 70 wt%, testifying the intrinsic structural advantages as sulfur carrier. Fig. 4a displays the cyclic voltammetry (CV) curves of LSBs with S/B-MoP_1-x@NC, S/MoP@NC and S/NC as the cathodes. Obviously, two cathodic peaks around 2.3 and 2.0 V represent the stepwise reduction reactions from S₈ to Li₂S_x (4 ≤ x ≤ 6) and eventually transformed into Li₂S₂/Li₂S. The anodic peak around 2.4 V signifies the corresponding oxidation reaction process. Additionally, the enhanced redox kinetics of S/B-MoP_1-x@NC are evident in the higher peak current, larger enclosed area, and smaller voltage polarization in CV curves, which contributes to promote the conversion process of LiPSs. The influences of lithium diffusion rate on the LiPSs conversion kinetics were further analyzed by testing the CV curves at various scan rates (Fig. S7 in Supporting information). According to the Eqs. S1 and S2 (Supporting information), there is a linear correlation between the corresponding peak currents (logi) and the square root of scan rate (logv), and the calculated b values are closed to 0.5, manifesting the predominance of diffusion-controlled reaction mechanism for LSBs (Figs. S8a-c in Supporting information). Combined with the Eq. S3 (Supporting information), the lithium diffusivity (D_Li) can be well calculated [46] As summarized in Fig. 4b, the D_Li values corresponding to the redox peaks in S/B-MoP_1-x@NC cathode, are higher than those in S/MoP@NC and S/NC cathodes, which conduces to reinforce the catalytic performance and inhibit the dissolution of active sulfur component into the liquid electrolyte.

  Figure 4

  

  Figure 4.  (a) CV curves of S/B-MoP_1-x@NC, S/MoP@NC and S/NC cathodes at 0.1 mV/s. (b) The calculated diffusion coefficients of Li⁺ ions corresponding to redox peaks A, C₁ and C₂. Tafel slopes calculated based on the (c) oxidation peaks and (d) reduction peaks in CV curves of S/B-MoP_1-x@NC and S/MoP@NC cathodes. (e) Li₂S nucleation/growth/precipitation behavior and (f) Li₂S dissolution behavior of B-MoP_1-x@NC/MoP@NC/NC electrodes. (g) Rate performances of S/B-MoP_1-x@NC, S/MoP@NC and S/NC cathodes (0.2–2 C). (h) Galvanostatic charge/discharge curves corresponding to the rate performance of S/B-MoP_1-x@NC cathode. (i) Comparisons of cyclic stability of S/B-MoP_1-x@NC and S/MoP@NC cathodes at 0.5 C.

  DownLoad: Full-Size Img PowerPoint

  The Tafel curves of LSBs were also conducted to uncover the effects of electrocatalysis behavior on the conversion reaction kinetics of LiPSs [47]. Notably, the Tafel slope for both oxidation and reduction reactions of S/B-MoP_1-x@NC cathode is much smaller than that of S/MoP@NC cathode (Figs. 4c and d), proving the structural advantages of dual-defect engineering strategy, which can be demonstrated by the corresponding EIS spectra (Fig. S8d in Supporting information) and galvanostatic intermittent titration technique (GITT) curves (Fig. S9 in Supporting information) of LSBs. Meanwhile, the Li₂S nucleation/growth/precipitation behavior on the carrier materials plays a key role in determining the conversion reaction kinetics of LiPSs [48]. Thus, Li₂S nucleation test was implemented to evaluate the reaction kinetics of Li₂S on the surface of NC, MoP@NC and B-MoP_1-x@NC electrodes. As presented in Fig. 4e, the precipitation capacity of Li₂S on B-MoP_1-x@NC electrode is 311.33 mAh/g, higher than those on NC (231.39 mAh/g) and MoP@NC (299.63 mAh/g) electrodes after constant potential discharge process for 8000 s. Moreover, Li₂S dissolution test was further carried out and shown in Fig. 4f. The calculated capacity of B-MoP_1-x@NC electrode corresponding to the dissolution behavior of Li₂S is 641.28 mAh/g, much higher than those of MoP@NC (328.39 mAh/g) and NC (441.98 mAh/g) electrodes, certifying B-MoP_1-x@NC has the strongest catalytic capability to synergistically facilitate the redox reactions of LiPSs intermediates. The constant-current charge/discharge curves of LSBs at 0.2 C were investigated to estimate the capacity contribution and polarization potentials of different sulfur-loading cathodes. The reversible discharge capacity of S/B-MoP_1-x@NC cathode can achieve 1178 mAh/g at 0.2 C (Fig. S10a in Supporting information), which is higher than that of S/MoP@NC (1024 mAh/g) and S/NC (926 mAh/g) cathodes, and the polarization potential (redox potential difference) of S/B-MoP_1-x@NC cathode is 218 mV (Fig. S10b in Supporting information), which is smaller than that of S/MoP@NC (238 mV) and S/NC (259 mV) cathodes, implying the higher energy efficiency of S/B-MoP_1-x@NC cathode. Meanwhile, the specific capacity contribution derived from the high and low discharge platforms in the cathodes, well correspond to the conversion reaction from S₈ to Li₂S_x (4 ≤ x ≤ 6, Q_H) and conversion reaction from Li₂S_x (4 ≤ x ≤ 6) to Li₂S₂/Li₂S (Q_L), respectively. As summarized in Fig. S10c (Supporting information), the calculated Q_L ratio of S/B-MoP_1-x@NC cathode is 64%, higher than that of S/MoP@NC (60%) and S/NC (58%), manifesting the thorough conversion reaction of LiPSs driven by B-MoP_1-x@NC catalyst.

  The rate performances of different sulfur-loading cathodes are also supplemented to verify their adaptability and compatibility to different current densities. As demonstrated, S/B-MoP_1-x@NC cathode delivers the highest specific capacity of 1244, 879, 758 and 621 mAh/g at the rates of 0.2, 0.5, 1.0 and 2.0 C (Figs. 4g and h, Fig. S11 in Supporting information), and significant capacity recovery capability of 863 mAh/g at 0.5 C. The cycling stability of S/B-MoP_1-x@NC and S/MoP@NC cathodes at different current densities were comprehensively compared. Within 100 cycles, the reversible capacity of S/B-MoP_1-x@NC cathode is 887 mAh/g with a capacity retention ratio of 81% (Figs. S12a and S13 in Supporting information), higher than that of S/MoP@NC cathode (71%) [34]. Within 300 cycles, S/B-MoP_1-x@NC cathode delivers a reversible capacity of 753 mAh/g at 0.5 C, higher than that of S/MoP@NC cathode (Fig. 4i), which can be attributed to the restrained shuttle effect of LiPSs. Fig. S12b (Supporting information) investigates the cycling performances of S/B-MoP_1-x@NC, S/MoP@NC, and S/NC cathodes with high sulfur-loading content of 5.1 mg/cm² at 0.5 C, and the specific capacity of S/B-MoP_1-x@NC cathode decreases from 654 mAh/g to 520 mAh/g within 100 cycles, more stable than that of other cathodes.

  Additionally, the positive inhibition effect of B-MoP_1-x@NC catalyst for the LiPSs during cycling process can be feasibly characterized by the in-situ UV–vis spectra. Fig. S14a (Supporting information) displays the schematic diagram of battery test model with cuvette spectrometer as the reaction unit, and the characteristic peaks of LiPSs intermediates in the liquid electrolyte can be accurately detected. According to the contour images of in-situ UV–vis spectra, the signal intensity of LiPSs in Li-S/MoP@NC cells (Fig. S14b in Supporting information) is much stronger than that in Li-S/B-MoP_1-x@NC cells (Fig. S14c in Supporting information) during cycling process under the same absorbance level, demonstrating the effective suppression of LiPSs shuttle effect and stronger adsorption behavior for LiPSs by the dual-defect coordination environment in B-MoP_1-x@NC catalyst [22,49]. Fig. S15 (Supporting information) shows the off-line visualization test for the LiPSs shuttle effect in S/NC, S/MoP@NC and S/B-MoP_1-x@NC cathodes during the discharge process. Notably, the liquid electrolyte in Li-S/B-MoP_1-x@NC cells is yellowish at the final stage of discharge process, manifesting the strong chemical adsorption capability of B-MoP_1-x@NC for LiPSs to further inhibit the shuttle effect. In sharp contrast, the liquid electrolytes in Li-S/MoP@NC cells and Li-S/NC cells change from colorless to light yellow due to spontaneous dissolution of active sulfur component and self-driven shuttle behavior of LiPSs [50]. Furthermore, the surface morphology of Li foil as well as separators coupled with S/B-MoP_1-x@NC and S/MoP@NC cathodes is also compared to directly judge the erosion effect induced by the shuttle effect of LiPSs. Evidently, the separator with relatively light color change removed from the cycled Li-S/B-MoP_1-x@NC cells (Fig. S16 in Supporting information), certifying that the dual-defect coordination design effectively inhibits the polysulfide shuttle, and reduces the corrosion of separators by polysulfides [51–53].

  In summary, an elaborate sulfur-host precursor (B-MoP_1-x@NC) with dual-defect configuration of B-doping and P-vacancy is comprehensively investigated as advanced cathode material (S/B-MoP_1-x@NC) for LSBs. Both the experimental data and DFT calculations demonstrate that the B-doping and P-vacancy effectively modulate the electronic structure of MoP, and facilitate its chemical affinity with LiPSs to availably suppress the shuttle effect. As expected, B-MoP_1-x@NC-based electrode holds low Li₂S oxidation potential, high lithium diffusivity and small Tafel slope, greatly contributing to enhance the conversion reaction kinetics of LiPSs. Specifically, the optimized S/B-MoP_1-x@NC cathode maintains high reversible capacity of 887 mAh/g at 0.5 C after 100 cycles with a high-capacity retention ratio of 81%, and high capacity of 621 mAh/g at 2 C, indicating stable cyclic performance and good rate capability. As further disclosed by the results of in-situ UV–vis spectra, the excellent electrochemical performance can be ascribed to the introduction of double-defect coordination in B-MoP_1-x@NC catalyst to greatly strengthen the chemical adsorption capability for LiPSs, and effectively inhibit the shuttle behavior of LiPSs. This work may help to open new window for the design and application of advanced catalytic sulfur-carrier materials for next-generation metal-sulfur batteries.

  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.

  CRediT authorship contribution statement

  Na Li: Writing – original draft, Supervision, Funding acquisition, Conceptualization. Wenxue Wang: Writing – original draft, Methodology, Investigation, Data curation. Peng Wang: Writing – review & editing, Supervision, Project administration, Funding acquisition. Zhanying Sun: Visualization, Software, Resources, Methodology. Xinlong Tian: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition. Xiaodong Shi: Writing – review & editing, Writing – original draft, Project administration, Funding acquisition, Formal analysis.

  Acknowledgments

  The authors thank the financial support from National Natural Science Foundation of China (No. 52101250), Hebei Provincial Natural Science Foundation (Nos. E2021208031 and B2021208069), S & T program of Hebei (Nos. 215A4401D and 225A4404D), Research Fund of the Innovation Platform for Academicians of Hainan Province (No. YSPTZX202315), Collaborative Innovation Center of Marine Science and Technology of Hainan University (No. XTCX2022HYC14), Fundamental Research Funds for the Hebei University (No. 2021YWF11), Science Research Project of Hebei Education Department (No. QN2024087), and Xingtai City Natural Science Foundation (No. 2023ZZ027). Additionally, this work is partially supported by the Pico Election Microscopy Center of Hainan University, and High-Performance Computing Platform of Nanjing University of Aeronautics and Astronautics.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110731.

  
  
• 1. [1]

     H.J. Li, J.N. Feng, T. Liu, et al., Energy Stor. Mater. 71 (2024) 103664.

  2. [2]

     R.L. Liu, Z.Y. Wei, L.L. Peng, et al., Nature 626 (2024) 98–104. doi: 10.1038/s41586-023-06918-4

  3. [3]

     Y.T. Liu, L.H. Xu, Y.Q. Yu, et al., Joule 7 (2023) 2074–2091. doi: 10.1016/j.joule.2023.07.013

  4. [4]

     B. Qin, X.M. Zhao, Q. Wang, et al., Energy Stor. Mater. 55 (2023) 445–454.

  5. [5]

     T. Yang, Y. Niu, Q. Liu, M. Xu, Energy Environ. Sci. 5 (2023) 119–140. doi: 10.1007/s11146-022-09899-9

  6. [6]

     S.Y. Zhou, J. Shi, S.G. Liu, et al., Nature 621 (2023) 75–81. doi: 10.1038/s41586-023-06326-8

  7. [7]

     J.T. Kim, A. Rao, H.Y. Nie, et al., Nat. Commun. 14 (2023) 6404. doi: 10.1038/s41467-023-42109-5

  8. [8]

     X.Y. Zhou, Y.C. Cui, X. Huang, et al., Small 19 (2023) 2301545. doi: 10.1002/smll.202301545

  9. [9]

     W.Q. Yao, J. Xu, L.B. Ma, et al., Adv. Mater. 35 (2023) 2212116. doi: 10.1002/adma.202212116

  10. [10]

      Z.X. Shi, Y.F. Ding, Q. Zhang, J.Y. Sun, Adv. Energy Mater. 12 (2022) 2201056. doi: 10.1002/aenm.202201056

  11. [11]

      P. Wang, B.J. Xi, S.L. Xiong, Acc. Chem. Res. 57 (2024) 15563–15571.

  12. [12]

      Q.B. Jiang, H.F. Xu, K. San Hui, et al., Angew. Chem. 63 (2024) e202408474. doi: 10.1002/anie.202408474

  13. [13]

      C.H. Zhao, B. Jiang, Y. Huang, et al., Energy Environ. Sci. 16 (2023) 5490–5499. doi: 10.1039/d3ee01774e

  14. [14]

      Y. Song, H. Li, J.F. Li, et al., J. Energy Chem. 87 (2023) 51–60. doi: 10.1051/parasite/2023051

  15. [15]

      L.P. Chen, Y.H. Xu, G.Q. Cao, et al., Adv. Funct. Mater. 32 (2022) 2107838. doi: 10.1002/adfm.202107838

  16. [16]

      H.M. Wang, H. Yuan, W.W. Wang, et al., Adv. Mater. 36 (2024) 2307741. doi: 10.1002/adma.202307741

  17. [17]

      S. Yu, Y.G. Zhang, S. Yang, et al., Chin. Chem. Lett. 34 (2023) 107911. doi: 10.1016/j.cclet.2022.107911

  18. [18]

      S.Y. Hu, T.S. Wang, B.B. Lu, et al., Adv. Mater. 34 (2022) 2204147. doi: 10.1002/adma.202204147

  19. [19]

      L. Ma, Y.Q. Zhang, S. Zhang, et al., Adv. Funct. Mater. 33 (2023) 2305788. doi: 10.1002/adfm.202305788

  20. [20]

      Y. Zhang, Y. Qiu, L.S. Fan, et al., Energy Stor. Mater. 63 (2023) 103026.

  21. [21]

      H.P. Wang, N. Li, J.F. Sun, P. Wang, J. Colloid Interf. Sci. 665 (2024) 702–710. doi: 10.1016/j.jcis.2024.03.165

  22. [22]

      C.H. Zhao, Y. Huang, B. Jiang, et al., Adv. Energy Mater. 14 (2024) 2302586. doi: 10.1002/aenm.202302586

  23. [23]

      J.H. Li, Z.Y. Wang, K.X. Shi, et al., Adv. Energy Mater. 14 (2024) 2303546. doi: 10.1002/aenm.202303546

  24. [24]

      Z.N. Li, I. Sami, J. Yang, et al., Nature Energy 8 (2023) 84–93. doi: 10.1038/s41560-022-01175-7

  25. [25]

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

  26. [26]

      A. Zhou, D.S. Wang, Y. Li, Microstructures 33 (2022) 35.

  27. [27]

      Y.Z. Liang, N. Song, Z. Zhang, et al., Adv. Mater. 34 (2022) 2202673. doi: 10.1002/adma.202202673

  28. [28]

      S.Y. Hu, X.Y. Huang, L. Zhang, et al., Adv. Funct. Mater. 33 (2023) 2214161. doi: 10.1002/adfm.202214161

  29. [29]

      H. Wang, F. Fu, M. Huang, et al., Nano Mater. Sci. 5 (2023) 141–160. doi: 10.1016/j.nanoms.2022.01.002

  30. [30]

      L.P. Lv, C.F. Guo, W.W. Sun, Y. Wang, Small 15 (2018) 1804338.

  31. [31]

      Y.P. Liu, A. Chatterjee, P. Rusch, et al., ACS Nano 15 (2021) 15047–15056. doi: 10.1021/acsnano.1c05344

  32. [32]

      Z.L. Zheng, L. Yu, M. Gao, et al., Nat. Commun. 11 (2020) 3315. doi: 10.1038/s41467-020-17199-0

  33. [33]

      D.W. Yang, M.Y. Li, X.J. Zheng, et al., ACS Nano 16 (2022) 11102–11114. doi: 10.1021/acsnano.2c03788

  34. [34]

      H. Zhang, M.T. Zhang, R.Y. Liu, et al., Nat. Commun. 15 (2024) 1–13.

  35. [35]

      J.W. Wang, S.F. Cao, L.K. Yang, et al., Nanoscale 13 (2021) 16487–16498. doi: 10.1039/d1nr04879a

  36. [36]

      Q. Yang, S.Y. Shen, Z.Y. Han, et al., Adv. Mater. 36 (2024) 2405790.

  37. [37]

      X. Gao, Z.A. Yu, J.Y. Wang, et al., Proc. Natl. Acad. Sci. 120 (2023) e2301260120. doi: 10.1073/pnas.2301260120

  38. [38]

      Y.J. Li, W.Y. Wang, B. Zhang, et al., Nano Lett 21 (2021) 6656–6663. doi: 10.1021/acs.nanolett.1c02161

  39. [39]

      S.Q. Li, Z.Y. Fan, Energy Stor. Mater. 34 (2021) 107–127.

  40. [40]

      J.D. Shen, Z.W. Liang, T.T. Gu, et al., Energy Environ. Sci. 17 (2024) 6034–6045. doi: 10.1039/d4ee01885k

  41. [41]

      X.X. Sun, S.K. Liu, W.W. Sun, C.M. Zheng, Chin. Chem. Lett. 34 (2023) 107501. doi: 10.1016/j.cclet.2022.05.015

  42. [42]

      Z.J. Lao, Z.Y. Han, J.B. Ma, et al., Adv. Mater. 36 (2024) 2309024. doi: 10.1002/adma.202309024

  43. [43]

      R.L. Li, Z. Bai, W.S. Hou, et al., Chin. Chem. Lett. 34 (2023) 108263. doi: 10.1016/j.cclet.2023.108263

  44. [44]

      Z.Y. Han, R.H. Gao, T.S. Wang, et al., Nature Catal 6 (2023) 1073–1086. doi: 10.1038/s41929-023-01041-z

  45. [45]

      X. Zhang, X.Y. Li, Y.Z. Zhang, et al., Adv. Funct. Mater. 33 (2023) 2302624. doi: 10.1002/adfm.202302624

  46. [46]

      Z.L. Wu, S.X. Chen, L. Wang, et al., Energy Stor. Mater. 38 (2021) 381–388.

  47. [47]

      G.H. Al-S, J.W. Zhu, W. Zhang, et al., Chin. Chem. Lett. 34 (2023) 108190. doi: 10.1016/j.cclet.2023.108190

  48. [48]

      P.L. Feng, W.S. Hou, Z. Bai, et al., Chin. Chem. Lett. 34 (2023) 107427. doi: 10.1016/j.cclet.2022.04.025

  49. [49]

      D. Zhang, T.F. Duan, Y.X. Luo, et al., Adv. Funct. Mater. 33 (2023) 2306578. doi: 10.1002/adfm.202306578

  50. [50]

      Y.X. Song, Y. Shi, J. Wan, et al., Energy Environ. Sci. 12 (2019) 2496–2506. doi: 10.1039/c9ee00578a

  51. [51]

      F. Li, L. Wang, G.M. Qu, et al., Chin. Chem. Lett. 33 (2022) 3909–3915. doi: 10.1016/j.cclet.2021.11.046

  52. [52]

      Y.Q. Qi, N. Li, K. Zhang, et al., Adv. Mater. 34 (2022) 2204810. doi: 10.1002/adma.202204810

  53. [53]

      K. Zhang, X. Li, Y. Yang, et al., Adv. Funct. Mater. 33 (2023) 2212759. doi: 10.1002/adfm.202212759

• 

  Figure 1  (a) TEM image, (b) STEM image and the corresponding element mapping images, and (c) HRTEM image of B-MoP_1-x@NC sample. (d) STEM-HADDF image and the corresponding P vacancy evidences of B-MoP_1-x@NC sample. (e) The intensity of Mo and P atomic columns along red dotted line in (d). (f) The theoretical model of MoP. (g) XRD patterns of MoP@NC and B-MoP_1-x@NC samples. (h) Magnified image of (101) diffraction peaks of MoP@NC and B-MoP_1-x@NC samples; High-resolution of (i) Mo 3d, (j) P 2p and (k) B 1s spectra of MoP@NC and B-MoP_1-x@NC samples. (l) EPR signals of MoP@NC and B-MoP_1-x@NC samples.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) High-resolution Mo 3d spectra of the pristine B-MoP_1-x@NC catalyst and the composite of B-MoP_1-x@NC-Li₂S₆ sample. (b) LiPSs adsorption models on the surface of MoP@NC and B-MoP_1-x@NC catalysts. (c) Differential charge distributions of Li₂S₆ adsorbed on the (101) crystal plane of MoP@NC and B-MoP_1-x@NC (yellow regions: electron accumulation; cyan regions: electron depletion). (d) Li⁺ ions diffusion energy barriers, (e) Li₂S₆ cluster migration energy barriers, and (f) Li₂S cluster decomposition energy barriers on the (101) crystal plane of MoP@NC and B-MoP_1-x@NC catalysts. (g) The calculated Gibbs free energy for the conversion reactions of LiPSs driven by MoP@NC and B-MoP_1-x@NC catalysts.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Schematic illustration the differences in action mechanism between (a) MoP and (b) B-MoP_1-x catalysts on the conversion reaction kinetics of LiPSs. (c) The influences of band engineering on the P-band elevation and s-p hybridization. Density of states (DOS) of (d) MoP@NC and (e) B-MoP_1-x@NC. (f) The alignment of atomic surface orbitals with LiPSs and the orientation of S-coupling modified by the introduction of boron dopants.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) CV curves of S/B-MoP_1-x@NC, S/MoP@NC and S/NC cathodes at 0.1 mV/s. (b) The calculated diffusion coefficients of Li⁺ ions corresponding to redox peaks A, C₁ and C₂. Tafel slopes calculated based on the (c) oxidation peaks and (d) reduction peaks in CV curves of S/B-MoP_1-x@NC and S/MoP@NC cathodes. (e) Li₂S nucleation/growth/precipitation behavior and (f) Li₂S dissolution behavior of B-MoP_1-x@NC/MoP@NC/NC electrodes. (g) Rate performances of S/B-MoP_1-x@NC, S/MoP@NC and S/NC cathodes (0.2–2 C). (h) Galvanostatic charge/discharge curves corresponding to the rate performance of S/B-MoP_1-x@NC cathode. (i) Comparisons of cyclic stability of S/B-MoP_1-x@NC and S/MoP@NC cathodes at 0.5 C.

  下载: 全尺寸图片 幻灯片
•