Electrocatalytic conversion of lithium polysulfides by highly dispersed ultrafine Mo 2 C nanoparticles on hollow N-doped carbon flowers for Li-S batteries

The significant challenge in exploring novel nanostructured sulfur host materials for Li-S batteries is to simultaneously mitigate the notorious shuttle effect and catalytically enhance the redox kinetics of lithium polysulfides (LPSs). Herein, a novel ultrafine Mo 2 C nanoparticles uniformly distributed on 2D nanosheet-assembled 3D hollow nitrogen-doped carbon flowers (HNCFs) is designed. The Mo 2 C/HNCFs architecture with unique flower-like morphol-ogies not only efficiently suppressed the aggregation of 2D nanosheets but also highly distributed the ultrafine Mo 2 C nanoparticles that act as catalytic active sites for efficient adsorption and conversion of LPSs. Furthermore, the 3D hierarchical arrangement can afford ample internal space to accommodate sulfur species, large volume expansion, 3D electron pathway, and physical/chemical blockage of LPSs to reduce the loss of active materials. The Mo 2 C/HNCFs composite exhibits a high rate capability, unprecedented capacity retention of 92% over 100 cycles at 0.5 C placing Mo 2 C/HNCFs one of the best LPSs adsorbents and electrocatalysts.


| INTRODUCTION
To address the challenges of current environmental pollution and energy crisis, alternative energy technologies are urgently demanded to substitute traditional fossil fuels, for example, lithium sulfur batteries (LSBs), 1 which have been touted as the upcoming post lithium ion batteries (LIBs) due to the remarkably high theoretical specific energy density of 2600 W h kg −1 and low cost of nontoxic elemental sulfur. 2 On the other hand, tons of thousands of sulfur are produced as a by-product of the petroleum refining process 3 and by the reduction of SO 2 during flue gas purification. 4 Furthermore, the recovered sulfur from H 2 S by Clause process need to find a valuable and green usage. 5 Being sulfur cathode for LSBs is right a good choice. However, the practical application of the LSBs is hindered by the shuttle effect and the sluggish redox kinetics of lithium polysulfide (LPSs). 6,7 Recently, a great amount of research has shown that the electrochemical performances in LSBs are heavily dependent on electrocatalytic conversion of LPSs. 6,[8][9][10] The electrocatalysts during the discharge process accelerate the transformation from soluble LPSs to insoluble Li 2 S 2 /Li 2 S by lowering the overpotential. 9 As the fast conversion reactions decrease the detention time of soluble LPS (long detention times cause severe losses of active materials) thereby suppressing dissolution. 9 Meanwhile, during the charge process, electrocatalysts reduce the potential hill at initial charging originating from the phase nucleation of Li 2 S 2 /Li 2 S and improve the oxidation kinetics and sulfur utilization. [10][11][12] In this perspective, several pioneer works have been performed on various electrocatalysts, such as Pt/G, 13 Mo 2 C/CNT, 14 ZnS 1-x / rGO, 15 CP@NCNT@CoS 3 , 16 (Co−N/G), 11 h-TiN, 17 porphyrin (POF), 18 Co 3 S 4 , 10 VO 2 and VS 2 nanosheets, 19 Fe 3 C/Fe-N x , 20 CoP nanoparticles, 21 and Fe 2 O 3 @N-PC/ Mn 3 O 4 . 22 However, the investigation of polysulfide electrocatalysis is still at its early stage, with neither systematic designed electrocatalyst nor deep mechanism understanding. Henceforth, novel and unique nanostructured electrocatalysts are strongly demanded in order to establish electrocatalytic chemistry of sulfur-containing compounds for the practical application of LSBs.
Most importantly, the previously reported electrocatalysts for LSBs are either micro-sized or bulk crystals, imposing restriction on available active sites for catalysis or adsorption toward LPSs. 14,23,24 Reducing the dimensions of electrocatalysts is a potential solution to this issue. In this case, nano-sized transition metal carbides on two-dimensional (2D) structure will be attractive due to the large exposed active surface and sites for sulfur coverage and conversion. Particularly, molybdenumbased carbide (Mo 2 C) attracts a huge interest considering its strong polarity, promising conductivity, and efficient electrocatalytic ability. 25,26 Unfortunately, the high reaction temperature for the preparation of Mo 2 C induces extensive particle sintering, leading to big particle dimension and low surface areas that are not suitable for practical applications. 27 In order to mitigate sintering, Mo 2 C nanoparticles are usually dispersed on conductive supports such as carbon nanotubes or graphene nanosheets. Nevertheless, these one-dimensional (1D) 23 or 2D carbon supports 28 themselves are prone to entangle or aggregate, which may risk negating all the advantages associated with the small Mo 2 C nanoparticle size. Therefore, the controllable synthesis of small-sized, well dispersed, and electrochemically accessible ultrafine Mo 2 C nanoparticles uniformly distributed on three-dimensional (3D) structuredcarbon remains challenging at present.
Herein, we designed a novel 3D nanostructured sulfur host comprising ultrafine Mo 2 C nanoparticles on hollow N-doped carbon flowers (HNCFs) for LSBs. The uniformly distributed Mo 2 C nanoparticles act as catalytic active site for efficient adsorption and redox conversion of LPSs. Meanwhile, the HNCFs can afford ample internal space to: (a) accommodate sulfur species, (b) allow large volume expansion, (c) provide 3D electron pathway, (d) reduce the loss of active materials via physical/chemical blockage, and (e) the flowerlike architecture can effectively avoid lamellar stacking and particle aggregation. 29 Significantly, the strong interaction between the ultrafine Mo 2 C and HNCFs enhanced the redox kinetics and charge transfer of the sulfur electrode. Owing to unique structural design, the Mo 2 C/HNCFs composite delivers a capability of 1181 and 572 mA h g −1 at 0.2 and 4 C, respectively, and extraordinary capacity retention of 92% over 100 cycles at 0.5 C. The present work highlights the importance of designing the novel 3D structure and ultrafine Mo 2 C nanoparticles as efficient active sites on HNCFs for enhancing adsorption and redox conversion of LPSs.

| RESULTS AND DISCUSSION
The synthesis of ultrafine Mo 2 C nanoparticles on HNCFs is briefly outlined in Scheme 1. Colloidal silica spheres were first prepared using the classic Stöber method ( Figure S1). 30 Typically, the SiO 2 , Mo 7 O 24 6− anions, and dopamine (DA) were successively mixed in water and ethanol mixture. Afterward, the appropriate volume of aqueous ammonia was added to trigger the spontaneous polymerization of DA together with Mo 7 O 24 6− (Mo/PDA). The interaction between Mo/PDA and SiO 2 was facilitated by the catechol and amine functional groups, resulting in coating of Mo/PDA onto the surface of SiO 2 . Interestingly, Mo completely changed the growth behavior of PDA into 2D layers, and eventually leading to the formation of Mo/PDA with unique microflower morphology. Finally, SiO 2 @Mo/ PDA was calcined at 800 C in Ar and SiO 2 @Mo 2 C/HNCFs was obtained, which was then etched by NaOH solution to remove SiO 2 . The sulfur was loaded into the as-prepared Mo 2 C/HNCFs by vapor phase infusion method. Figure 1A shows the typical X-ray diffraction (XRD) pattern of Mo 2 C/HNCFs. In addition to a characteristic diffraction peak (002) of carbon at 2θ = 24 , several sharp peaks corresponding to Mo 2 C (JCPDS 77-0720) were confirmed. The morphology of the as-synthesized nanostructure Mo 2 C/HNCFs was firstly examined by field emission scanning electron microscopy (FESEM) equipped with energy dispersive spectroscopy (EDS) showing hollow carbon flowers with average diameter of~1.6 μm ( Figure 1B). Furthermore, each micro-flower was formed from the assembly of hundreds of thin 2D nanoflakes S C H E M E 1 Schematic illustration for the preparation of Mo 2 C/HNCFs, sulfur loading using vapor phase infusion method and electrochemical redox process F I G U R E 1 A, XRD pattern, B,C, FESEM images, and D,E, TEM images of Mo 2 C/HNCFs vertically aligned with sufficient space between neighboring nanoflakes ( Figure 1C). The content of Mo 2 C in the Mo 2 C/NHCFs hybrid obtained by EDS is estimated to bẽ 53 wt% ( Figure S2). The unique 3D structure of Mo 2 C/ HNCFs with numerous nanoflakes and nanovoids is also confirmed by transmission electron microscopy (TEM) ( Figure 1D). Furthermore, a large number of ultrafine particles are uniformly dispersed over the carbonaceous nanoflakes ( Figure 1E). The high-resolution TEM (HRTEM) image shows that these ultrafine particles possess an average size of~1.6 nm with an interlayer spacing of 0.23 nm, corresponding to (101) plane of Mo 2 C ( Figure 1E, inset), consistent with the XRD result. For a control sample, the XRD pattern of HNCS shows two diffraction peaks at 2θ of 24 and 44 , which could be indexed to the (002) and (101) diffraction planes of carbon ( Figure S3a), suggesting the mostly amorphous nature of the HNCS. The morphology of the HNCSs is characterized by FESEM which demonstrates the hollow spheres ( Figure S3b). The obtained Mo 2 C/HNCFs was further analyzed by X-ray photoelectron spectroscopy (XPS). The XPS survey of Mo 2 C/HNCFs displays well-resolved peaks of C, N, Mo, and O species (Figure 2A). Specifically, the deconvoluted C 1 second spectrum ( Figure 2B) can be attributed to sp 2 -hybridized C─C/C═C, C─N, and C═O species. 31 From the N1s spectrum, three N species referring to the pyridinic, pyrrolic and graphitic were observed ( Figure 2C), manifesting the multiconfiguration of N heteroatom doping in the carbon lattice. 27,32 The Mo 3d XPS spectrum is derived from Mo 2 C (3d 5/2 at 228.2 eV), MoO 2 (3d 5/2 at 229.0 eV), and MoO 3 (3d 5/2 at 232.4 eV) ( Figure 2D). The presence of a significant quantity of oxides seems to be unavoidable for carbide materials due to their oxidation at the surface upon exposure to air. 33 However, according to previous results, the oxide passivation layer on Mo 2 C may contribute to adsorb polysulfides. 34 To investigate the electrochemical properties of Mo 2 C/HNCFs, sulfur was loaded into the Mo 2 C/HNCFs by a vapor phase infusion method (S-Mo 2 C/HNCFs). SEM images show that the S-Mo 2 C/HNCFs composite well maintains the original morphology after sulfur loading ( Figures 3A,B and S4), demonstrating the homogeneous loading of sulfur into the Mo 2 C/HNCFs structure. Furthermore, EDS elemental mapping ( Figure 3C-G) and F I G U R E 2 A, XPS survey spectra, B, high-resolution XPS spectra of C 1s, C, N 1s, and D, Mo 3d of Mo 2 C/HNCFs XRD ( Figure 3H) (S 8 , JCPDS 08-0247) of S-Mo 2 C/NHCFs reveal that sulfur is evenly distributed in the Mo 2 C/ HNCFs host. The sulfur loading in S-Mo 2 C/HNCFs composite was determined by thermo-gravimetry analysis (TGA), that is, 69 wt% ( Figures 3I and S5). In comparison, the sulfur loaded onto the HNCSs as a control sample (by vapor phase infusion method), denoted as S-HNCSs, was confirmed by FESEM ( Figure S6) and XRD ( Figure S7) with 63 wt% sulfur loading as determined by TGA ( Figure S8).
To evaluate the advantages of the Mo 2 C/HNCFs as an advanced sulfur host for LSBs, 2025-type coin cells using S-Mo 2 C/HNCFs and S-HNCSs as cathodes and lithium metal as anodes were assembled. Figure 4A shows the cyclic performance of S-Mo 2 C/HNCFs and S-HNCS. The S-Mo 2 C/HNCFs electrode retained a discharge capacity of 902 mAh g −1 after 100 cycles with 92.7% capacity retention and 0.07% capacity decay per cycle merely. The Coulombic efficiency is $100%. In contrast, S-HNCS retained a discharge capacity of 494 mAh g −1 with low F I G U R E 3 A,B, FESEM images, C-G, STEM image and the corresponding element mapping, H, XRD pattern, and I, TGA curve of S-Mo 2 C/HNCFs capacity retention after 100 cycles. From Nyquist plots ( Figure 4B), it can be observed that the S-Mo 2 C/HNCFs electrode has a smaller semicircle in the high-frequency region, indicating the lower charge transfer resistance compared with S-HNCSs. Long cycling performance was investigated to demonstrate the superior lifetime of S-Mo 2 C/HNCFs at high current rate (1 C). After 300 cycles, the S-Mo 2 C/HNCFs cathode exhibits specific capacity of 585 mAh g −1 with a capacity decay of 0.08% per cycle and $100% Coulombic efficiency ( Figure 4C). However, S-HNCS retains a capacity of only 257 mAh g −1 with 0.11% capacity decay per cycle ( Figure S9). Figure 4D shows the rate performance of the S-Mo 2 C/HNCFs cathode, where the specific capacities at the current densities F I G U R E 4 A, Cycling performance and Coulombic efficiency (CE) at 0.5 C for 100 cycles, B, electrochemical impedance spectroscopy (EIS) for the S-Mo 2 C/HNCFs and S-HNCSs cathodes, C, long cycling performance and CE of the S-Mo 2 C/HNCFs cathode at 1 C for 300 cycles, D, rate performance and E, galvanostatic discharge and charge curves of the S-Mo 2 C/HNCFs cathode at different current densities, F, cycling performance and CE of S-Mo 2 C/HNCFs at 3.1 mg cm −2 areal sulfur loading for 200 cycles of 0.2, 0.5, 1, 2, and 4 C were 1181, 1037, 879, 738, and 572 mA h g −1 , respectively. The excellent capacity reversibility to 928 mA h g −1 was achieved at 0.5 C. In contrast, S-HNCS shows low rate performance at the same current densities ( Figure S10). Figure 4E shows the galvanostatic charge-discharge profiles of the S-Mo 2 C/HNCFs electrode at different current rates (0.2-4 C). Two reaction plateaus can be evidently observed in the discharge process that correspond to the reduction of sulfur into long-chain LPSs and the subsequent formation of short-chain LiPSs to Li 2 S, respectively, whereas the charge plateau is attributed to the transformation of Li 2 S 2 /Li 2 S to Li 2 S 8 /S 8 . Prominently, the galvanostatic charge/discharge voltage profile of S-Mo 2 C/HNCFs at 4 C is still obvious. However, S-HNCS cannot maintain the charge plateau even at 1 C (- Figure S11). Since high mass loading of active materials are essential for the energy density of LSBs, a thicker S-Mo 2 C/ HNCFs electrode with areal sulfur loading of 3.1 mg cm −2 is further evaluated ( Figure 4F). Upon cycling at 0.5 C, a high discharge capacity of 613 mAh g −1 is delivered, corresponding to $100% capacity retention and Coulombic efficiency. The capacity rise during the initial cycles is attributed to gradually wetting of the thicker 3D porous cathode by electrolyte which smooths the ion diffusion. 35 Comparatively, S-HNCS exhibits a discharge capacity of 282 mAh g −1 only ( Figure S12).
The excellent electrochemical properties of Mo 2 C/ HNCFs is anticipated to be derived from strong adsorption and efficient catalytic conversions of LPSs. 9,18 In this perspective, static adsorption tests of Mo 2 C/HNCFs and HNCSs for LPSs were first performed. When Mo 2 C/ HNCFs was added into a Li 2 S 4 solution, a change in color was observed ( Figure 5A); however, HNCSs could not decolorize Li 2 S 4 solution, indicating that Mo 2 C/HNCFs exhibited strong adsorption capability for Li 2 S 4 molecules. 36,37 Secondly, the electrochemical polarization was performed in order to confirm the redox kinetics of S-Mo 2 C/ HNCFs and S-HNCSs electrodes. Figure 5B shows the flatter and longer discharge plateaus for S-Mo 2 C/HNCFs with a higher discharge/charge capacity and a smaller hysteresis (ΔE = 134 mV) in comparison with S-HNCS (ΔE = 310 mV), suggesting the enhanced electrochemical performance and accelerated redox kinetics in S-Mo 2 C/ HNCFs. In addition, the improved electrochemical kinetics was observed in the discharge ( Figure 5C) and charge ( Figure 5D) processes during phase change between the soluble LPSs and insoluble Li 2 S 2 /Li 2 S. Figure 5C displays the discharge plateau, which confirms the overpotential The S-HNCSs electrode shows a large overpotential of 55 mV, suggesting that a high interfacial energy barrier existed for Li 2 S nucleation and deposition on the surface. 11,16 In contrast, the Li 2 S nucleation behavior is quite different in the case of S-Mo 2 C/HNCFs ( Figure 5C). The S-Mo 2 C/HNCFs electrode exhibits a significantly reduced interfacial energy barrier of 18 mV, indicating the facilitation of Li 2 S nucleation and deposition process ( Figure 5C). Likewise, Figure 5D shows the charge plateaus, from which the overpotential and potential barrier for the conversion of insoluble Li 2 S to soluble LPSs can be assessed. 11,12 The S-HNCSs cathode exhibits a high overpotential and potential barrier at about 2.39 V in the charging process, indicating a sluggish activation process. 12,19 However, the combination of uniformly distribution of ultrafine Mo 2 C nanoparticles with HNCFs significantly reduces the height of the potential barrier to 2.31 V ( Figure 5D). 11,19 This strong synergism is favorable for lowering the overpotential and improving capacity compared with HNCSs only. These findings provide insight into a fundamental understanding of ultrafine Mo 2 C nanoparticles on HNCFs for catalytical conversion of LPSs. Thirdly, the symmetric cells were configured using Mo 2 C/HNCFs or HNCSs as identical working and counter electrodes, within a voltage range of −0.7 to +0.7 V (vs Li/Li + ) to check its catalytic ability. 39 The fabrication of symmetric cells eliminates the lithium metal anode and hence provides direct analysis of the polysulfides redox conversion in the hierarchical electrode. As observed in Figure 5E, the Mo 2 C/HNCFs symmetric cell exhibits the higher polarization current compared with HNCSs, indicating that Mo 2 C/HNCFs provides the better catalytic performance, which is the direct evidence of polysulfide electrocatalytic conversion on Mo 2 C/HNCFs. 40 Furthermore, the EIS measured on symmetric cells revealed a lower resistance for Mo 2 C/HNCFs in comparison with HNCFs, which is another indication of the electrocatalytic effects derived from the unique flower-like 3D Mo 2 C/ HNCFs with unique uniformly dispersed ultrafine Mo 2 C nanoparticles ( Figure 5F). The robust catalytic effect of Mo 2 C/HNCFs is due to highly distributed ultrafine Mo 2 C nanoparticles that provide more catalytic active sites for efficient adsorption and conversion of LPSs. 41,42 Fourthly, cyclic voltammetry (CV) curves for S-Mo 2 C/ HNCFs and S-HNCS were performed at different scan rates ranging from 0.1 to 0.4 mV s −1 (Figures S13 and S14), which exhibit two cathodic peaks (I and II) and one anodic peak (III). The higher current density for S-Mo 2 C/ HNCFs at different scan rates ( Figure 6A,B) suggests the bigger decrease of polarization due to the participation of active Mo 2 C in the redox reactions in comparison with S-HNCS ( Figure 6C,D), consistent with discharge/charge profiles. Furthermore, with an increase in the scan rate, the S-Mo 2 C/HNCFs electrode shows a less decrease in onset potentials for peaks I and II, and III than that for S-HNCS, implying that the sulfur redox reaction on the surface of Mo 2 C/HNCFs was accelerated (0.05 vs 0.06, 0.09 vs 0.14, 0.09 vs 0.14). 17,43 The superb catalytic activity revealed by the Mo 2 C/HNCFs with regards to LPSs conversion could be further demonstrated by its lower Tafel slope that is 80 mV dec −1 compared with 134 mV dec −1 for HNCS, obtained by anaylzing the CV curve at 0.1 mV s −1 (Figure 6E,F). 44 Additionally, as shown in Figure 6G-I, all cathodic and anodic peak currents are linear with the square root of scan rates, from which the lithium diffusion performance can be estimated using the classical Randles Sevcik equation 45 : I p = 2:69 × 10 5 À Á n 1:5 S D Li + 0:5 C Li + ν 0:5 where I P is the peak current, n is the charge transfer number, S is the surface area of the electrode, D Li + is the lithium ion diffusion coefficient, C Li + is the concentration of lithium ions, and ν is the scan rate. Because n, S, and C Li + are unchanged, the slope of the curve (I p /ν 0.5 ) represents the Li + ions diffusion rate. At any conditions, the slopes of curves for S-Mo 2 C/HNCFs are higher than those for S-HNCSs ( Figure 6G-I), suggesting that S-Mo 2 C/ HNCFs show better diffusion properties compared with S-HNCSs. The enhanced Li + ions diffusion promotes the catalytic effect of Mo 2 C/HNCFs, which is critical in improving the battery performance. The catalytic effects of the ultrafine Mo 2 C nanoparticles on HNCF are schematically illustrated in Figure 7A, where the strong adsorption for LPSs is the first step of catalytic conversions. To validate the morphology integrity of Mo 2 C/HNCFs after long cycling, the electrode film from disassembled cells at fully charged status is re-examined by SEM and HRTEM. As shown in Figure 7B,C, the individual Mo 2 C/HNCFs flower is retained and the component nanoflakes are clearly discernible. Impressively, HRTEM unveils that the 2D carbon sheets are decorated with a large number of darkcontrast ultrafine Mo 2 C particles ( Figure 7D, inset), similar to the original Mo 2 C/HNCFs ( Figure 1E). There is no evidence that the long cycling results in the extensive aggregation.

| CONCLUSION
In summary, we have developed a unique Mo 2 C/HNCFs composite configured with ultrafine Mo 2 C nanoparticles uniformly dispersed on 2D nanosheets assembled 3D HNCFs and used it as multi-functional sulfur host for LSBs. Specifically, the ultrafine Mo 2 C nanoparticles afford effective anchoring and catalytic active sites to LiPSs, which can not only enhance the LiPSs-adsorption ability but also accelerate the redox kinetics of polysulfide conversion during the cycling processes. Besides, the architecture of 2D nanosheets assembled 3D HNCFs contributes to Li + transportation and electrolyte infiltration. As a result, the S-Mo 2 C/HNCFs electrode retained a discharge capacity of 902 mAh g −1 after 100 cycles with 92% capacity retention. Moreover, the Mo 2 C/HNCFs composite exhibits a high rate capability of 1181 and 572 mA h g −1 at charge/discharge rates of 0.2 and 4 C, respectively. Most impressively, the structural integrity of the hybrid flowers is largely preserved even after prolonged cycling. This work may open up a new avenue to design novel 3D host candidates with highly dispersed active sites on carbon network for advanced LSBs.

| EXPERIMENTAL
Preparation of colloidal silica particles: Colloidal silica was prepared by using the classic Stöber method. In a typical preparation, 6 mL of NH 3 ÁH 2 O was added into a mixture of 24.75 mL of water and 16.75 m L of ethanol followed by adding the ethanolic solution of TEOS. Ethanolic solution of TEOS was prepared by dissolving 2.25 mL of TEOS in 45.5 mL of ethanol. The as-prepared colloidal silica was dispersed under ultrasound treatment in 60 mL of isopropanol and 200 μL of APTES. After ultrasound treatment, the reaction was kept for 2 hours at 80 C in an oil bath. The final product was obtained by centrifugation followed by washing, and freeze-drying.
Synthesis of Mo 2 C/HNCFs: Typically, 50 mg of the asprepared amino-modified silica (SiO 2 ) particles were dispersed in 12 mL of ethanol under ultrasonication. Next ammonium molybdate tetrahydrate (185 mg) was dissolved in the mixed solvent of H 2 O (18 mL) and ethanol (8 mL) and added into the ethanolic solution of SiO 2 . Subsequently, 100 mg of dopamine dissolved in 18 mL of deionized water was added dropwise to the above solution. After 30 minutes, ammonia aqueous solution (0.5 mL) was added into the above mixture and continuously stirred at room temperature for 24 hours. After the completion of the reaction, the product was collected by centrifugation, washed with DI water and ethanol, and dried at 70 C overnight. The as-prepared solid product was carbonized at 350 C with a heating rate of 3 C min −1 for 3 hours and then increased to 800 C with a heating rate of 2 C min −1 and kept for 2 hours under an argon atmosphere. Next silica was removed from the carbonized product by treating with NaOH solution (4 M) for 4 hours at 100 C. The final product was washed with water and ethanol by vacuum filtration for several times and Mo 2 C/HNCFs was obtained after drying at 70 C. The hollow nitrogen-doped carbon spheres (HNCS) as a control sample was prepared by following the same procedure except adding the ammonium molybdate tetrahydrate.
Formation of S-Mo 2 C/HNCFs: Sulfur was loaded into Mo 2 C/HNCFs composite by vapor phase infusion method. In brief, sulfur powder and Mo 2 C/HNCFs were mixed well by grinding and then sealed it in a glass vial.
Finally heated the mixture at 300 C for 4 hours under Ar and cool it at room temperature to get S-Mo 2 C/HNCFs.
Structural characterizations: XRD patterns were recorded on a D8 FOCUS powder XRD instrument (Burker AXS, Germany) using 40 kV as tube voltage and 40 mA as tube current. FESEM equipped with EDS was performed on a Hitachi SU-70 microscope. The TEM was conducted on a JEOL JEM-2010 microscope at an accelerating voltage of 200 kV. XPS data were obtained on thermo scientific ESCALAB 250 XI electron spectrometer, using monochromatic Al Kα as exciting radiation at constant passing energy of 1486.6 Ev. TGA was conducted on NETZSCH, STA449 from room temperature to 800 C with a heating rate of 10 K min −1 in Ar.
Preparation of sulfur electrodes and electrochemical measurements: Sulfur electrode was prepared by mixing 70% S-Mo 2 C/HNCFs, 20% super P and 10% poly (vinylidene fluoride) (PVDF, Sigma-Aldrich) binder in N-methyl 2-pyrrolidone (NMP, Sigma-Aldrich) solvent. Typical, S-Mo 2 C/HNCFs and super P were mixed in a mortar by hand for long time followed by adding PVDF binder in NMP solvent to form a slurry. The slurry was then pasted on aluminum foil with the active material loading of~1.5 to 1.8 mg cm −2 and then dried at 60 C for 12 hours. The electrode films were cut into round disks with a diameter of 8 mm. The standard 2025-type stainless steel coin cells were assembled inside an Ar-filled glovebox with lithium metal foil as the negative electrode. The electrolyte was prepared by dissolving lithium bis-trifluoromethanesulfonylimide (LITFSI, 99%, Acros Organics, 1 M) and lithium nitrate (LiNO 3 , 99.9%, Alfa Aesar, 0.1 M) in 1,2-dimethoxyethane (DME, 99.5%, Alfa Aesar) and 1,3-dioxolane (DOL, 99.5%, Alfa Aesar) (1:1 ratio, by volume). The amount of electrolyte is~15 μL per mg sulfur . A LAND galvanostatic charge/discharge system was used. Stepwise rate performance was examined at different densities. The CV tests were performed on an SP-300 at different scan rates of (0.1-0.4 mV s −1 ). EIS measurements were conducted on an SP-300 at a frequency range of 100 kHz to 100 mHz.
Symmetrical cell assembly and measurements: The electrodes were made by mixing the Mo 2 C/HNCFs (without the presence of elemental sulfur) with PVDF in NMP at a weight ratio of 9:1 for symmetrical cells. The slurry was stirred well and pasted on aluminum foil with successive heating at 60 C for 12 hours. After drying, the electrode disks with a diameter of 12 mm were punched out and employed as working and reference electrodes. For the assembly of symmetrical cells, 40 μL of the electrolyte containing 0.5 mol L −1 Li 2 S 6 and 1 mol L −1 LiTFSI in tetraglyme was used. The cyclic CV tests were performed on SP-300 at a scan rate of 50 mV s −1 . EIS measurements of symmetrical cells were performed at a frequency range of 100 kHz to 100 mHz on SP-300.

ACKNOWLEDGMENT
This work was supported by National Natural Science Foundation of China (No. 21876061, 21906063 and 21805112) and LiaoNing Revitalization Talents Program (XLYC1802076).