The increasing demand for electric vehicles, hybrid electric vehicles and portable electronic devices impulse the development of advanced energy-storage technologies [1-4]. Li-S battery has been considered as the next-generation energy storage system because of the high theoretical specific capacity at 1675 mAh/g and high energy density 2600 Wh/kg [5]. However, there are many challenges to overcome before large-scale commercialization: (1) the severe "shuttle effect" of polysulfides; (2) the large volume expansion of sulfur change to the Li₂S; (3) the low conductivity of sulfur [6].

To address these problems, a large number of materials have been made and applied for lithium sulfur batteries. Nazar et al. first utilized CMK-3 carbon material as cathode material to adsorb and limit the diffusion of polysulfides [7]. Inspired by this work, carbon materials such as carbon nanotubes and graphene have been applied to lithium-sulfur batteries [8]. However, the interaction between nonpolar carbon and polar polysulfides was too weak to effectively immobilize of polysulfides. In subsequence, polar materials with better adsorption ability such as oxides [9], sulfides [10], carbides [11] and nitrides [12] were applied to lithium-sulfur batteries. Furthermore, the researchers found that most polar materials have a certain catalytic, which can accelerate the conversion of polysulfide ions. Recently, Professor Yang and others researchers have synthesized heterostructures, such as TiO₂/TiN [13], VO₂/VN [14], combining the advantages of different materials to trap and convert the polysulfides. The performance of the battery has been improved further. Nevertheless, the research on composite host materials is still in its infancy. There are still many unknown areas to be explored in this field. We know that the interface of composite is an important part of this system and the properties of the interface have an important influence on the electrical and other properties of heterostructures [15]. The research in this area has not been drawn much attention and the mechanism of the interface of composite host materials in lithium sulfur battery needs to be explored to guide the design of heterostructures more rationally.

MXene has great potential in lithium-sulfur batteries due to good conductivity and 2D structure [16-19]. However, MXene has the disadvantages such as relative lower catalytic activity for polysulfide conversion, which limit its further application. Co₂B has good catalytic activity for fast polysulfide kinetic conversion [21]. Therefore in this work, Co₂B@MXene composite is designed and explored its application for Li-S battery. Combining DFT theoretical calculations, spectroscopic analysis and electrochemical characterizations, we firstly reveal the interfacial electron interaction mechanism between Co₂B and MXene, which enhance the catalytic activity of Co₂B. As a result, the cell with Co₂B@MXene achieves a high initial discharge specific capacity of 1577 mAh/g at 0.1 C and exhibits 597 mAh/g at 5 C. Moreover, the cell also demonstrates a good long-cycling life and achieves high energy efficiency over 82.3% at 2 C (2000 cycles). In addition, the cell can achieve 5.2 mAh/cm² (0.2 C) under high sulfur loading of 5.1 mg/cm².

The procedure for the synthesis of Co₂B@MXene composites was illustrated in Fig. 1a. 1 mmol Co(NO₃)₂ was added into 50 mL deionized water (DI). And 5 mL MXene (5 mg/mL) was dropped into Co(NO₃)₂ solution and stirred for 30 min. Then, 10 mL NaBH₄ solution (1 mol/L) was added and maintained at 60 ℃ for overnight under stirring. The product was collected by centrifuged, washed and dried in a vacuum overnight. Moreover, the detail experimental was listed in Supporting information. The schematic diagram of Co₂B@MXene was shown in Fig. 1b, and Fig. 1c illustrated the effect of Co₂B@MXene separator in Li-S batteries which can immobilize the polysulfide and suppress the "shuttle effect".

X-ray diffraction patterns of Co₂B@MXene and MXene were shown in Fig. 2a. The XRD of MXene can be found the (002) peak, which attributes to the characteristic for the Ti₃C₂T_x. On the contrary, it was observed that the Co₂B@MXene has no obvious characteristic peak, demonstrating the amorphous substance of Co₂B [20]. The X-ray photoelectron spectroscopy (XPS) analysis was performed to probe the chemical composition and valence state of the elements. The Co 2p high-resolution spectrum of Co₂B and Co₂B@MXene were performed as shown in Figs. 2b and c. Compared with the peak of Co 2p in the Co₂B, the peak of Co 2p in the Co₂B@MXene moved higher binding energy. Notably, the peak of Ti 2p in the Co₂B@MXene was shifted lower binding energy compared with Ti 2p of the MXene in Fig. S1 (Supporting information). These results indicated the chemical bonding of Co₂B to MXene and the electron transfer from Co₂B to MXene [1, 15]. In addition, the electron transfer made Co centers in Co₂B more positively charged, promoting the attraction of more polysulfides during the reaction. More XPS information about B 1s was shown in Fig. S2 (Supporting information). Peaks for B 1s level with BE of 192.1 and 188.2 eV agreed very well with the B-O state and Co-B state, which were typical species in transition metal borides. The XPS peak position of B 1s matched with the previous report [21], which proved that the Co₂B material was prepared successfully. Furthermore, Raman spectroscopy of MXene and Co₂B@MXene were exhibited in Fig. 2d. The Raman spectroscopy of MXene nanosheets had three broad peaks at around 200, 400 and 600 cm⁻¹ owing to the vibration from Ti₃C₂T_x [22]. Moreover, the Raman spectra of Co₂B@MXene peaks present a little excursion compared with MXene, which attributed to the chemical bonding of Co₂B with MXene. Furthermore, the precise atomic ratio of composition was obtained by the ICP analysis and the constitution was determined as the Co₂B (Table S2 in Supporting information).

More morphology and structural characterization of Co₂B@MXene were represented in Fig. 3. As shown in Figs. 3a–c, it was observed that the Co₂B@MXene still remained an ultra-thin homogeneous layer structure, and there was no significant aggregation phenomenon. The Co₂B existed on the surface of MXene and between MXene layers in Fig. S15 (Supporting information). In addition, transmission electron microscopy (TEM) and elemental mapping were used to gain more information about the Co₂B@MXene. Co₂B nanoparticles were grown on MXene as shown in Fig. 3d. The high-resolution TEM image revealed Co₂B was amorphous, which was consistent with the XRD result and a clear interface between Co₂B and MXene can be noticed. Moreover, EDX elemental mapping images of Co₂B@MXene also illustrated the existence of element Co, Ti, B and the uniform dispersion of Co, B and Ti atoms on the Co₂B@MXene nanostructure.

To gain insight into the electron behavior of Co₂B@MXene interface and the interfacial effect on the conversion for polysulfides, a series of theoretical calculations were implemented as shown in Fig. 4. As shown in Fig. 4e, the charge density difference of the Co₂B@ MXene confirmed the electron migration from Co₂B to MXene. The electrons (yellow) crowd at MXene while holes (wathet blue) were converged at the Co₂B. Further, a model was established to probe the reason for electron transfer in Fig. 4d. The Fermi level (E_F) of MXene was -2.39 eV much lower than the E_F of Co₂B (E_F = -1.12 eV). On account of the energy level difference between Co₂B and MXene, electrons naturally shifted from Co₂B to MXene through the interface and produce a built-in electric field [23]. To further demonstrate the catalytic ability for polysulfides, we analyzed the Li₂S₄ adsorption on the surface of pure Co₂B and the Co₂B of Co₂B@MXene. Comparing the S-S bond length of Li₂S₄ before and after adsorption with Co₂B and Co₂B@MXene, it exhibited that the Co₂B and Co₂B@MXene induced a longer S₁-S₂ and S₃-S₄ bond length of Li₂S₄ in Figs. 4a–c and Table S3 (Supporting information). The longer bond length meant a weaker bond energy and the weakened S-S bond facilitated the decomposition of polysulfides [24]. The S₁-S₂ and S₃-S₄ bond were more prolonged on the surface Co₂B@MXene than that of Co₂B which indicated the Co₂B@MXene had stronger catalytic ability than that of Co₂B. Furthermore, the adsorption energy of Li₂S₄ on Co₂B and Co₂B@MXene were -1.418 eV and -2.66 eV in Fig. S13 (Supporting information), respectively. The interfacial electronic interaction between Co₂B and MXene can also promote the adsorption ability of polysulfides, which was helpful to restrain their shuttling. These phenomena indicated that the influence of interfacial electronic interaction, which can improve the catalytic capacity of Co₂B in Co₂B@MXene and promote the conversion of polysulfides.

To verify the stability and safety of the Co₂B@MXene separator, we performed a series of measurements to test its performance. As shown in Fig. S4 (Supporting information), we carried out the bendable exhibition for the Co₂B@MXene separator, and the Co₂B@MXene did not fall off before and after folding. It confirmed that the Co₂B@MXene separator had a good mechanical property. To further explore the adsorption ability of Co₂B@MXene with polysulfides, the visual permeation tests were performed in Fig. S5 (Supporting information). The left compartment was injected into the 0.03 mol/L Li₂S₆ solution, and the right compartment was infused into the pure electrolyte solution, while the separator was installed between these two compartments. The Li₂S₆ diffused through PP separator within 30 min easily in Fig. S5b, and Li₂S₆ diffused through the Co₂B separator, MXene separator within 1 h, however, the Li₂S₆ did not permeate the Co₂B@MXene separator. Even after 6 h, the solution of right compartment with Co₂B@MXene separator still remained clear. This result demonstrated the Co₂B@MXene had a well blocking ability to polysulfides. Moreover, the infiltrating angle between the separator and electrolyte was also an important index to evaluate the infiltration property [25]. The infiltration property tests with different separators were shown in Fig. S6 (Supporting information). The contact angle of Co₂B@MXene separator was 5.6°, which was the least compared to the contact angle of PP, Co₂B and MXene separator, which revealed the Co₂B@MXene separator had the best wettability with the minimum surface tension.

To further systematically evaluate the electrochemical performance of Co₂B@MXene separator in Li-S batteries, more electrochemical properties of cell were tested with coin-type CR2025 cells. The sulfur content of CNT/S composites was 73.53 wt% in Fig. S8 (Supporting information) by thermogravimetry analysis (TGA) and the sulfur loading mass was 1.2–1.5 mg/cm². In addition, electrochemical properties of Co₂B@MXene with different thickness were further investigated to identify the optimal coating thickness. SEM images of Co₂B@MXene with different thickness were shown in Figs. S7a-c (Supporting information), and the cell of Co₂B@MXene with the 14.7 μm thickness emerged the best cycling performance. Therefore, we utilized Co₂B@MXene separator with the 14.7 μm thickness for the following characterizations. Moreover, the electrochemical performance with different ratios of Co₂B to MXene were shown in Fig. S14 (Supporting information). The cyclic voltammetric (CV) curves of cells with Co₂B@MXene, MXene, Co₂B and PP separator were measured at the voltage window of 1.7–2.8 V with a scan rate of 0.1 mV/s (Fig. S9 in Supporting information). The cell with Co₂B@MXene separator delivered two cathodic peaks at 2.32 and 2.05 V during the discharge measurements which representing the reaction from S₈ to soluble Li₂S_n (Li₂S_n, 4 < n < 8) and the further reduction product of Li₂S₂/Li₂S [26]. The cell with MXene separator exhibited two cathodic peaks at 2.30 and 2.02 V, the cell with Co₂B separator showed two cathodic peaks at 2.30 and 2.04 V and the cell with PP separator showed two cathodic peaks at 2.28 and 2.01 V. The cell with Co₂B@MXene separator showed a distinguishable positive shift for the two cathodic peaks and higher peak current, revealing the rapid transformation of S₈ to soluble polysulfides and insoluble products (Li₂S). The cell with Co₂B@MXene separator exhibited two anodic peaks at 2.34 V during the charge measurements, which were associated with the oxidation of Li₂S to S₈. The anodic peak of MXene, Co₂B and PP separator located at 2.39, 2.39 and 2.41 V, respectively. At the same time, the anodic peak of Co₂B@MXene also displayed distinct negative shifts, expounding that the Co₂B@MXene facilitates the oxidation of Li₂S to S₈. The CV curve of the cell with Co₂B@MXene separator had a companion peak (B₂) at 2.39 V near the anodic peak (B₁) at 2.34 V in the charging process, which can indicate the smaller polarization comparing the anodic peaks position of cells with MXene, Co₂B and PP separator. The redox reaction kinetics in the cell with Co₂B@MXene was more favorable than the cell with MXene and Co₂B, suggesting fast charge transfer in the Co₂B@MXene cell compared with the MXene and Co₂B cell. To gain insight into the electrocatalytic effect of Co₂B@MXene on the electrochemical transformation of polysulfides, a set of kinetics experimentations were carried out. The CV curve of the symmetrical cell of Co₂B /Li₂S₆ and Co₂B@MXene/Li₂S₆ showed obvious response currents in Fig. S10a (Supporting information), which proved that Co₂B and Co₂B@MXene can accelerate polysulfide transformation [16]. It is worth noting that Co₂B@MXene has the largest response current, indicating that Co₂B@MXene has a stronger catalytic activity. Furthermore, the CV curve of Co₂B@MXene without Li₂S₆ (Fig. S10b in Supporting information) displayed a minimal current behavior, eliminating the influence of Co₂B@MXene itself. This phenomenon was consistent with the theoretical results and the interfacial electronic interaction can enhance the catalytic capacity and accelerate the redox kinetics.

To further evaluate the electrochemical performance characterization of the cell with Co₂B@MXene separators, some measures were carried out. Fig. 5a exhibited the initial galvanostatic charge/discharge curves of the cell with Co₂B@MXene separator, Co₂B separator, MXene separator and PP separator at 0.1 C (1 C=1675 mAh/g) with a voltage window of 1.7–2.8 V. The discharging profile had two reduction plateaus, demonstrating the conversion of sulfur to Li₂S_n (4 < n < 8) and further transformation to insoluble Li₂S₂/Li₂S ultimately. The charging curve showed a long oxidation plateau, indicating the oxidation process of Li₂S to S₈. Moreover, the cell with Co₂B@MXene separator displayed the minimum polarization (ΔE₁ = 0.139), the cell with MXene separator was ΔE₂ = 0.202, the cell with Co₂B separator was ΔE₄ = 0.167 and the cell with PP separator was ΔE₃ = 0.245, respectively. This result agreed well the above CV profiles (Fig. 5a). In addition, the cell with Co₂B@MXene separator showed an initial discharge capacity of 1577 mAh/g at 0.1C, while the cell with MXene, Co₂B and PP separator exhibited discharge capacity of 1240, 1150 and 1090 mAh/g, respectively. These results demonstrated that Co₂B@MXene can reduce the polarization and delay the "shuttle effect" for increasing the specific capacity.

To further investigate the electrochemical performance of Co₂B@MXene under high rate condition, the tests were carried out as shown in Fig. 5b. The cell with Co₂B@MXene separator exhibited good rate performance with the current range from 0.1 C to 5 C. When the current density at 0.1, 0.2, 0.5, 1.0, 2.0 and 5.0 C, the cell with Co₂B@MXene separator delivered steady cycling discharge specific capacities of 1566, 1374, 1169, 917, 775 and 597 mAh/g, individually. Meanwhile, the cell with MXene, Co₂B and PP separator delivered capacities of 1322, 1112, 923, 740, 566, 478 mAh/g, 1160, 1058, 902, 769, 596, 502 mAh/g and 1118, 895, 665, 509, 397 and 209 mAh/g at 0.1, 0.2, 0.5, 1, 2 and 5 C, respectively. Even at 5C current density, the cell with Co₂B@MXene separator also displayed an outstanding high discharge capacity of 597 mAh/g, illustrating that the Co₂B@MXene promoted the utilization of sulfur and retarded the "shuttle effect" effectively.

To detect the cycling stability of the cell with different separators, cycling performances were carried out at 0.5 C current density in Fig. 5c. The cell with Co₂B@MXene exhibited an initial discharge capacity at 1276 mAh/g and presented a high capacity retention of 1061 mAh/g after 100 cycles with the 0.168% per cycle average capacity decay rate. On the contrast, the discharge specific capacity of the cell with MXene, Co₂B and PP separator were 956 mAh/g (0.323% decay rate), 901 mAh/g (0.286% decay rate) and 670 mAh/g (0.414% decay rate), individually. This result indicated the Co₂B@MXene can improve the cycling stability and enhance specific capacity effectively. As shown in Fig. 5d, the areal capacity of the cell with Co₂B@MXene was 5.2 mAh/cm² and reserved 4.1 mAh/cm² after 100 cycles at 0.2 C with the sulfur mass loading of 5.1 mg/cm². It should be noted that the cell with Co₂B@MXene maintained good cycle stability than cell with MXene, Co₂B and PP separator under various areal density condition, demonstrating that the Co₂B@MXene not only enhanced the reaction kinetics of polysulfide but also promoted the sulfur utilization during cycling. Moreover, the long-life cycling performance of lithium-sulfur battery with Co₂B@MXene was measured at 2 C in Fig. 5e, which displayed an initial discharge specific capacity of 786 mAh/g. The capacity decay rate of the cell with Co₂B@MXene was 0.0088% per cycle after 2000 cycling, revealing an excellent long-term cycling property. Coulombic efficiency of the cell with Co₂B@MXene remained over 98%, indicating high reversible conversion reaction of polysulfide.

To further explore the effect of Co₂B@MXene on electrochemical kinetics before and after cycling, the electrochemical impedance spectroscopy (EIS) were measured in Fig. S11 (Supporting information). The cell with Co₂B@MXene separator exhibited the smallest polarization resistance as shown in Fig. S11a. Besides, the change of the impedance of Co₂B@MXene was significantly smaller than that of others, suggesting Co₂B@MXene can promote the conversion of polysulfides. These results also indicated that Co₂B@MXene not only immobilized the polysulfide, but also enhanced rapid conversion reaction of polysulfides.

The more electrochemical characteristics of different separators were measured in the Fig. S12 (Supporting information). The lithium ion conductivity was an index to evaluate Li⁺ ion transport performance, and the Li⁺ ion conductivity of different separators were measured in Fig. S12a and Table S1 (Supporting information). The lithium ion conductivity was calculated by σ = l/(R_b×A), according to the semicircular loop in high frequency region of EIS curves. The conductivity of Co₂B@MXene separator, PP separator, Co₂B separator and MXene separator were 0.88 mS/cm, 0.598 mS/cm, 0.74 mS/cm and 0.761 mS/cm, separately. To further study the self-discharge behavior of the cell with PP, MXene, Co₂B and Co₂B@MXene separators, the open circuit voltage (OCV) was carried out in Fig. S12b. The cell with PP, MXene and Co₂B separator appeared severe self-discharge and the OCV were 2.29 V, 2.36 V and 2.34 V after 20 h. On the contrast, the OCV of cell with Co₂B@MXene separator was 2.38 V, proving the Co₂B@MXene suppressed the self-discharge efficiently. Besides, Li⁺ transference was another index to estimate separator property. Li⁺ transfer number can be acquired as shown in Figs. S12c-e through the time-current measurement. The Li⁺ transfer numbers with PP, MXene, Co₂B and Co₂B@MXene separator were 0.715, 0.675, 0.708 and 0.733, respectively. The Co₂B@MXene did not reduce the Li⁺ transport property.

In summary, we designed and fabricated a new Co₂B@MXene composite as interlayer material to enhance the conversion of polysulfides and to acquire atomic-level understanding of polysulfides catalytic conversion behavior on heterogeneous material. The theoretical calculations and experiments reveal that the interfacial electronic interaction between two materials can enhance the catalytic ability and point out that the two materials with different Fermi energy levels was the intrinsic reason for the spontaneous electron transfer. The cell with Co₂B@MXene delivers excellent discharge specific capacity and extraordinarily rate performance. The cell with Co₂B@MXene exhibits the discharge capacity of 1577 mAh/g at 0.1 C and maintains an impressive rate capability with 597 mAh/g at 5 C. Moreover, the capacity decay of the cell with Co₂B@MXene is only 0.0088% per cycle at 2 C over 2000 cycles. Even with the sulfur content of 5.1 mg/cm², stable cycling can still be maintained over 100 cycles. This work provides a deep understanding of the effect of interfacial interaction and inspires the design and research of advanced catalytic materials for lithium-sulfur batteries.

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.

The work was supported by the State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology (No. 2019DX13).

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.cclet.2020.12.051.