CoS₂ nanoparticles anchored on N,B co-doped porous carbon for high-rate and long-life lithium–sulfur batteries

Wei Dong*^a, Zhaomeng Wu^b, Xiaochen Xu*^a, Su Li^c and Yang Ji^d
aLiaoning Provincial Key Laboratory of Energy Storage and Utilization, Yingkou 115014, China. E-mail: lgddongwei@163.com; xuxiaochen@yku.edu.cn
^bCollege of Material Science and Engineering, Liaoning Technical University, Fuxin 123000, China
^cMianyang High-tech Exencell New Energy Technology CO., Ltd., Mianyang 621000, China
^dLiaoGang Holding (Yingkou) Co., Ltd., No. 4 Branch, Yingkou 115014, China

First published on 22nd August 2025

---

---

1. Introduction

Lithium–sulfur (Li–S) batteries are considered one of the best options for next-generation batteries due to their high energy density and low cost. The theoretical gravimetric energy density of up to 2600 Wh kg⁻¹ is an order of magnitude higher than that of traditional lithium-ion battery systems.^1,2 Besides, thanks to the abundant resources, low cost and environmentally friendly nature of sulfur, the practical application of lithium–sulfur batteries has gained more competitive advantages. However, there are three major challenges in the positive electrode of lithium–sulfur batteries that limit their development. (1) The low electrical conductivity of sulfur and its products leads to slow electrochemical reactions.³ (2) The volume change during the electrochemical process has a negative impact on safety and durability.⁴ (3) The soluble intermediate product lithium polysulfides (LiPSs) generated during the reaction diffuses to the negative electrode to form irreversible products, that is, the “shuttle effect”, which seriously affects the sulfur utilization rate and coulombic efficiency.⁵

The conductivity and diverse structure of carbon materials can effectively solve the problems of low conductivity and volume change of sulfur.⁶ At the same time, by reasonably designing the structure of carbon materials, it is also expected to alleviate the shuttle effect of polysulfides through physical confinement methods.^7–10 However, the limited physical confinement of carbon materials and the weak van der Waals force between them and polysulfides are not sufficient to alleviate the shuttle of LiPSs.¹¹ Therefore, chemical methods are further introduced into the sulfur-host system of the positive electrode to inhibit the shuttle of polysulfides. Currently, there are mainly two strategies. One is to dope heteroatoms (N,¹² O,¹³ P,¹⁴ S,¹⁵ Fe,¹⁶ Co,¹⁷ etc.) or modify the surface of carbon materials by functionalization (-O-groups,^18,19 etc.). The other is to load polar compounds with significant ability to immobilize LiPSs, such as metal nitrides, metal oxides, and metal sulfides.²⁰ By modifying or compounding carbon materials, the polar adsorption and catalytic conversion of polysulfides can be significantly improved. Different from electronegative atoms such as N, O, S and P, which are particularly effective in capturing LiPSs through the formation of X⋯Li₂S_n interactions by lone-pair electrons, boron atoms are electropositive in the carbon skeleton, resulting in the chemisorption of LiPS anions (Sx²⁻).²¹ Therefore, if N and B atoms are simultaneously introduced into carbon, the chemisorption of LiPSs will be greatly increased due to the synergistic effect between the formed N⋯Li₂S_n and Li₂S_n⋯B.²² Zhu et al.²³ prepared N,B-co-doped porous carbon fibers (BNCNFs) and applied them in lithium–sulfur batteries. Li–N and B–S bonds provide strong constraints for LiPSs. The dual-fixation mechanism can effectively capture LiPSs within the positive electrode and reduce the shuttle effect. In addition, the obvious N–B and NB bonds in N,B co-doped carbon can enhance the polarization degree of the carbon surface and strengthen the anchoring effect on LiPS. Based on the above advantages, the cathode of N,B co-doped carbon exhibits excellent electrochemical performance. At a current density of 1C, the initial discharge specific capacity is 1054.7 mAh g⁻¹, and after 1000 cycles, it still maintains 433.7 mAh g⁻¹ with a decay rate of 0.058%. At a high current density of 5C, the initial capacity is 612.2 mAh g⁻¹, and after 600 cycles, the decay rate is 0.085%. Although the modification strategy plays a great role in improving the shuttle effect, there are still problems of slow kinetics due to the limited doping amount or the number of polar sites. Loading transition metal compounds (TMCs) can also improve the polar adsorption and catalytic conversion of polysulfides, such as transition metal oxides (TMOs), transition metal chalcogenides (TMDCs, such as S and Se), transition metal nitrides (TMNs), and transition metal carbides (TMCBs).²⁴ Transition metal sulfides have a high affinity for sulfur-containing substances and can provide good chemical anchoring for polysulfides. Their sulfur atoms also have high electronegativity and can capture electrons from transition metals, thus becoming stable intermediate reaction active sites.²⁵ Therefore, transition metal sulfides have been widely studied. In addition, metal sulfides usually have better conductivity than metal oxides, which is beneficial for the electrochemical conversion of sulfide electrodes. For example, Xu et al.²⁶ prepared a carbon cloth covered with CoS₂ nanoparticles (CC-CoS₂) to act a three-dimensional (3D) current collector, which has high electrical conductivity, high sulfur loading and high polysulfide capture ability. Therefore, the Al@S/AB@CC-CoS₂ battery with a sulfur loading of 1.2 mg⁻¹ exhibits high-rate performance (823 mAh g⁻¹ at 4C) and excellent capacity retention (with a decay of approximately 0.021% per cycle at 4C over 1000 cycles).

When multifunctional TMCs and heteroatom-doped carbon are assembled together as parts with different physical and chemical properties to construct a cooperative interface, the absorption rate and conversion ability of LiPSs at the interface are greatly improved. In this study, innovatively, CoS₂/porous boron–nitrogen co-doped carbon (CoS₂/BNC) was prepared using a boron–nitrogen co-doped porous carbon precursor. This heterogeneous CoS₂/BNC synergistically enhanced the chemical adsorption and redox kinetics of LiPSs from multiple dimensions, different from the single action mechanism, effectively suppressing the shuttle effect of LiPSs. Meanwhile, the hollow structure was designed as a sulfur storage chamber, achieving the dual goals of suppressing sulfur volume expansion and maximizing sulfur utilization, breaking through the limitations of traditional structures in single performance optimization. Based on the above advantages, the CoS₂/BNC/S lithium–sulfur battery exhibits excellent performance: the CoS₂/BNC/S lithium–sulfur battery has a high initial discharge capacity (1639.1 mAh g⁻¹ under the condition of 0.1C) and a long cycle life (at a current density of 2C, after 500 cycles, the capacity still remains 752.7 mAh g⁻¹, the capacity decay per cycle is only 0.048%, and the coulombic efficiency is maintained at 97.8%), and its comprehensive electrochemical performance is significantly improved compared with existing similar materials. This study pioneeringly provides a new idea and method for constructing metal sulfide/boron–nitrogen-doped porous carbon heterostructures to improve the performance of lithium–sulfur batteries, opening up a new path for material design in this field.

2. Experimental

2.1. Preparation of experimental samples

2.2. Material characterization

The structure and phase of the materials are characterized using an X-ray diffractometer (XRD, XRD-6000, Cu Kα). The surface morphology is observed by using a field-emission scanning electron microscope (SEM, JSM-7800F). The finer internal structure of the materials is observed by using a transmission electron microscope (TEM, JEOL2100F). The content and type of elements are quantitatively analyzed by using an X-ray photoelectron spectroscopy (XPS) instrument. The specific surface area and pore size distribution are measured by using a nitrogen adsorption specific surface area tester, Autosorb-iQ3. The specific surface area is determined by the Brunauer–Emmett–Teller (BET) method, and the pore size is calculated by the Barrett–Joyner–Halenda (BJH) method.

2.3. Electrochemical measurement

The preparation of electrode slurry adopts the mass ratio of active material, conductive agent carbon black, and binder polyvinylidene fluoride of 8:1:1. Use N-methylpyrrolidone to adjust the viscosity of the slurry, and coat it with aluminum foil as the current collector. After coating, transfer it to a vacuum drying oven and dry it at 60 °C. Use Cellgard-2400 as the separator, a pure lithium sheet as the counter electrode, and 1 M lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) dissolved in DOL and DME at a volume ratio of 1:1 as the electrolyte to assemble the battery in a glove box filled with high-purity argon. The battery case adopts a button battery case CR-2025 to assemble a button battery. Replace the electrode material with the working electrode of the battery and the counter electrode, replace the electrolyte with 0.2 M Li₂S₆ solution, and assemble a symmetrical battery. Use the electrode material as the working electrode, a pure lithium sheet as the counter electrode, Cellgard-2400 as the separator, and 0.2 M Li₂S₈ solution as the electrolyte to assemble a deposition experiment battery. Constant-current charge–discharge tests are carried out on a NEWARE charge–discharge tester, and the test voltage is 1.7–2.8 V. Cyclic voltammetry and alternating-current impedance tests are carried out on a Shanghai Chenhua electrochemical workstation CHI660E. The voltage of the cyclic voltammetry test is 1.7–2.8 V, and the frequency of the alternating-current impedance test is 0.01 Hz–100 kHz. The deposition kinetics test is carried out on an electrochemical workstation. The specific operation is as follows: under the current density of 0.1C, conduct a constant-current discharge test on the battery. After the voltage drops to 2.06 V, immediately adjust the parameters to make it conduct a constant-voltage discharge under the voltage condition of 2.05 V until the current is less than 0.01 mA.

3. Results and analysis

Fig. 1(a) shows the synthesis process of AC, BNC and CoS₂/BNC. During the asphalt carbonization process, urea and boric acid are used as nitrogen and boron sources, respectively, to achieve boron–nitrogen co-doping, and then CoS₂ nanoparticles are loaded through hydrothermal treatment.

Fig. 1(b)–(d) show the SEM images of AC, BNC and CoS₂/BNC. As shown in Fig. 1(b), AC contains a large number of pore structures, and the pore structure is similar to that of termite nests. It not only has an interconnected loose and porous channel system but also consists of tiny particles. This structure is formed with silica as a template, and the diameter of the pores ranges from several nanometers to dozens of nanometers. Nanoparticles are distributed on the surface of BNC porous carbon. As shown in Fig. 1(c), after nitrogen and boron elements are doped into the asphalt porous carbon, respectively, the interconnected loose and porous channels are still maintained, and the pore structure is significantly increased, and the pores become larger, forming a looser “sponge”. This structure is beneficial for the contact between the electrolyte and the electrode and shortens the ion/electron transmission path.²⁷ As shown in Fig. 1(d), after the hydrothermal treatment, the CoS₂ nanoparticles loaded on the porous carbon can be observed. Fig. 1(e)–(j) show the elemental distribution of CoS₂/BNC in the mapping area, which proves the existence of B, N, C, Co and S elements and their uniform distribution in the CoS₂/BNC sample. This indicates that during the sulfurization process, CoS₂ and BNC porous carbon are successfully combined.

Fig. 1(k)–(n) show the TEM image of CoS₂/BNC. It can be seen from the figure that nanoparticles with a diameter of 40–70 nm are distributed on the BNC porous carbon, which is consistent with the SEM image. Fig. 1(m) shows a high-resolution TEM (HRTEM) image. The calculated spacing of the lattice fringes of the nanoparticles is 0.278 nm, corresponding to the (200) crystal plane of CoS₂, indicating that the nanoparticles on the porous carbon are CoS₂, further confirming that the CoS₂ nanoparticles have been successfully loaded on the BNC porous carbon.²⁸

The XRD diffraction patterns of AC, BNC and CoS₂/BNC are shown in Fig. 2(a). It can be seen from the figure that AC and BNC both have two broad diffraction peaks at 25.4° and 42.7°, corresponding to the (002) and (100) crystal planes of carbon materials, respectively, indicating that the samples have an amorphous structure.²⁹ According to the Bragg equation, the average layer spacing is 0.357 and 0.36 nm, respectively. For CoS₂/BNC, the broad diffraction peak at around 26° is attributed to the boron–nitrogen co-doped asphalt porous carbon. The diffraction peaks at 32.2°, 36.2°, 46.2° and 55.1° correspond well to the CoS₂ card (JCPDS 41-1471), corresponding to the (200), (210), (220) and (311) crystal planes of CoS₂, respectively. According to the Scherrer formula, the average grain size of CoS₂ is approximately 50 nm, which is consistent with the TEM results, indicating that the CoS₂ composite boron–nitrogen co-doped asphalt porous carbon material was successfully prepared.³⁰ The XRD diffraction patterns of AC/S, BNC/S and CoS₂/BNC/S after sulfur loading are shown in Fig. 2(b). All three samples have obvious sulfur peaks, and the characteristic peaks correspond well to the S card (JCPDS 08-0247), proving the existence of sulfur in the composite materials.

The N₂ adsorption–desorption curves and pore size distributions of AC, BNC and CoS₂/BNC are shown in Fig. 2(c) and (d). It can be seen from Fig. 2(c) that the curves of P/P₀ of AC, BNC and CoS₂/BNC show a rising trend with a small slope in the range of 0.15–0.9, indicating that there are certain mesopores (2–50 nm) in the materials, showing the Langmuir type III adsorption characteristics of mesoporous materials and the existence of a typical H3 hysteresis loop. The pore size distribution curve calculated using the BJH model is shown in Fig. 2(d). The pore size distribution indicates that most of the pores are mainly around 31 nm. This porous structure can not only accelerate sulfur impregnation but also facilitate the diffusion of the electrolyte and the transport of lithium ions during the electrochemical reaction process. Table 1 shows the specific surface area and pore volume of three materials. Among them, CoS₂/BNC has the largest specific surface area (208.2 m² g⁻¹) and the largest pore volume (1.549 cm³ g⁻¹), which makes it superior to AC and BNC. The sufficient pores provide channels for electrolyte penetration and ion diffusion, which is the structural basis for high-rate performance.

---

Fig. 3(a) shows the full-range X-ray photoelectron spectra (XPS) of AC, BNC and CoS₂/BNC samples. It can be seen from the figure that AC has diffraction peaks at 285 and 532 eV, which correspond to C 1s and O 1s, respectively, and the elemental contents are 86.51 at% and 13.49 at%, respectively. BNC has diffraction peaks at 191, 285, 400 and 532 eV, which correspond to B 1s, C 1s, N 1s and O 1s, respectively, and the elemental contents are 4.39 at%, 76.69 at%, 7.45 at% and 11.2 at%, respectively. In the full-range XPS spectra of CoS₂/BNC, there are six characteristic peaks at 165.9, 232.5, 284.8, 399.9, 531.4 eV and 778.7 eV, which correspond to S 2p, B 1s, C 1s, N 1s, O 1s and Co 2p, respectively. The elemental contents are 4.61 at%, 3.27 at%, 74.87 at%, 4.63 at%, 12.01 at% and 0.62 at%, respectively. Deconvolution of the high-resolution spectra is performed to further analyze the interface composition. Fig. 3(b) shows the high-resolution spectrum of the B 1s peak of BNC. It can be seen from the figure that the B element has three different chemical forms, with the peaks at 190.5, 191.5 and 192.5 eV, which correspond to B–C, B–N and B–O, respectively. Among them, the B–N bond can improve the adsorption capacity of the material for polysulfides and promote the redox reaction process in lithium–sulfur batteries. The existence of these B–N and C–N bonds confirms the successful synthesis of the characteristic B and N co-doped structure. Fig. 3(c) shows the high-resolution N 1s spectrum of BNC. The peaks at 398.2, 399, 400.3 and 401.5 eV correspond to N–B, pyridinic N, pyrrolic N and graphitic N, respectively. Due to the Lewis acid–base interaction between the lone-pair electrons of the N atom and Li⁺ in LiPS, pyridine N and pyrrole N have relatively high binding energy for LiPS. They introduce many defects and active sites, thereby increasing the electronic conductivity. The co-doping of electropositive B atoms and electronegative N atoms in the carbon matrix can improve the electronic conductivity and provide sufficient active sites for capturing soluble polysulfides.

Fig. 3(d) shows the high-resolution Co 2p spectrum of CoS₂/BNC. The Co³⁺ peaks at 779.6 and 794.7 eV correspond to the Co–S bond, indicating the successful preparation of CoS₂. The Co²⁺ at 781.5 and 796.6 eV corresponds to the Co-O bond, and the peaks at 786.9 and 802.0 eV are the two satellite peaks of Co²⁺. Fig. 3(e) shows the high-resolution S 2p spectrum. The two peaks of S 2p at 163.7 and 164.9 eV correspond to S 2p_3/2 and S 2p_1/2, respectively, further indicating the existence of CoS₂. The peak at 169.1 eV corresponds to oxidized sulfur. XPS analysis shows the successful preparation of the CoS₂/BNC material, which is consistent with the XRD analysis.

To compare the adsorption properties of AC, BNC and CoS₂/BNC materials, an adsorption test was carried out. Fig. 4(a) shows the photos of the adsorption experiment. The three materials were added to the Li₂S₆ solution, respectively, and the change in the color of each liquid over time was observed. After standing for 2 hours, the solution with the AC sample added was still dark yellow. The Li₂S₆ with BNC added showed a color change from dark to light yellow, while the color of Li₂S₆ with CoS₂/BNC added was almost transparent. After standing for 12 hours, the solution with the AC sample added was still dark yellow, indicating that the adsorption of BNC and CoS₂/BNC materials was stronger than that of AC. In order to conduct an in-depth analysis of the chemisorption process, ultraviolet-visible spectroscopy tests were carried out on the samples before and after adsorption. The UV-vis spectrum shows a strong absorption band in the 300–400 nm region, which can be attributed to S₄²⁻ and S₆²⁻ polysulfide substances. The absorbance of the CoS₂/BNC sample is lower than that of the BNC and AC samples. It proves the superior adsorption ability of the synergistic effect between heteroatoms and transition metals in CoS₂/BNC. The adsorption of BNC was stronger than that of AC materials due to the simultaneous introduction of electronegative N atoms and electropositive B atoms, and this synergistic effect could simultaneously increase the chemisorption of Li cations and polysulfide anions. The CoS₂/BNC material had the strongest adsorption capacity because transition-metal sulfides can also enhance the polar adsorption of polysulfides, further improving the adsorption of the material.

To study the doping of boron and nitrogen elements and the catalytic rate of CoS₂ nanoparticles on the conversion of polysulfides to Li₂S, the liquid–solid conversion reaction rate of lithium polysulfide by AC, BNC and CoS₂/BNC was tested (Li₂S nucleation experiment), as shown in Fig. 4(b)–(d). It can be clearly seen from the figures that the Li₂S nucleation reaction on BNC (803 s) occurred earlier than that on AC (1479 s), and the nucleation time of CoS₂/BNC (360 s) was earlier than that of AC and BNC. Moreover, the nucleation current density of Li₂S on CoS₂/BNC (187.6 mA g⁻¹) and BNC (143.1 mA g⁻¹) was significantly higher than that on AC (78.4 mA g⁻¹). In addition, the deposition capacity of Li₂S on CoS₂/BNC (366.9 mAh g⁻¹) was the highest, better than that of AC (199.8 mAh g⁻¹) and BNC (315.5 mAh g⁻¹). The above results showed that the doping of boron and nitrogen elements could catalyze the reduction reaction to a certain extent, and after loading CoS₂ nanoparticles, it could further catalyze the nucleation and growth of Li₂S, promote the redox reaction kinetics, and enhance the speed of polysulfide catalytic conversion and the catalytic reduction reaction.

Fig. 5(a) shows the volt-ampere characteristic curves of three kinds of cathode materials, AC/S, BNC/S and CoS₂/BNC/S, at a scanning rate of 0.2 mV s⁻¹. It can be seen from the figure that during the reduction process of the three composite materials, there are two reduction reaction peaks, located at ∼1.95 V and ∼2.25 V, respectively. During the oxidation process, there are also two oxidation reaction peaks, located at ∼2.35 V and ∼2.43 V, respectively. The difference from the reduction peaks is that the two oxidation peaks in this process are partially overlapped. During the initial discharge, the first peak at 2.25 V is attributed to the breakage of cyclic sulfur molecules and the formation of high-order polysulfides (Li₂S_n, 4 ≤ n ≤ 8), and these polysulfides are highly soluble in the electrolyte; at ∼1.95 V, the second peak represents the further reduction of high-order polysulfides to insoluble low-order lithium sulfides (Li₂S₂, Li₂S). The subsequent anode scanning shows two partially overlapped oxidation peaks at 2.35 V and 2.43 V, which correspond to the reversible conversion from lithium sulfide to polysulfide and then to elemental sulfur, respectively. The two reduction peaks of AC/S appear at ∼1.9 V and ∼2.23 V, and the oxidation peaks appear at ∼2.35 V and ∼2.46 V. The reduction peaks of BNC/S appear at 1.99 V and 2.27 V, and the oxidation peaks appear at ∼2.35 V and ∼2.43 V. The difference between the oxidation peak and the reduction peak of BNC/S (0.08 V) is significantly smaller than that of AC/S (0.12 V), indicating smaller polarization and the rapid conversion of soluble products (LiS_x, x > 2) to insoluble products (LiS₂) during the discharge process. Compared with AC/S and BNC/S, the Li–S battery based on CoS₂/BNC/S has a more positive reduction peak and a more negative oxidation peak potential, with a difference of 0.07 V, and has the largest peak current and the smallest polarization voltage difference. This indicates that the Li–S battery modified by the CoS₂ catalyst has lower electrochemical reaction polarization, which shows that the catalytic effect of the CoS₂ nanoparticles further improves the redox kinetics in the electrochemical process and has a stronger catalytic ability for the redox conversion of polysulfides. According to the classical Randles–Sevcik equation,³¹ the diffusion coefficient of lithium ions can be indirectly calculated to compare the reaction kinetics between different batteries.

Among them, D represents the lithium-ion diffusion coefficient (cm² s⁻¹), I_P represents the peak current, n represents the number of electrons involved in the reaction (for Li–S batteries, n = 2), A represents the geometric area of the electrode (cm²), C refers to the lithium-ion concentration (mol L⁻¹), and v represents the scanning rate (V s⁻¹). It can be obtained that the diffusion coefficient of CoS₂/BNC/S is 1.4303 × 10⁻¹², which is higher than 4.8081 × 10⁻¹³ of AC/S and 5.4303 × 10⁻¹³ of BNC/S. The above results indicate that the CoS₂-BNC composite material has the best conductivity and lithium-ion diffusion coefficient, thereby improving the redox reaction kinetics.

To study the impedance of lithium ions in different materials, electrochemical impedance spectroscopy (EIS) measurements were carried out.³² Fig. 5(b) shows the electrochemical impedance spectra of the three materials, AC/S, BNC/S and CoS₂/BNC/S. Fitting was performed using the equivalent circuit diagram in the figure. It can be seen from the figure that the EIS spectra of the three materials are basically similar, all presenting typical Nyquist curves. The Nyquist curves are composed of a semicircle in the medium-high-frequency region and an oblique line. Among them, the semicircle in the high-frequency region corresponds to the lithium-ion transfer resistance (R_ct) in the positive electrode, the oblique line in the low-frequency region corresponds to the diffusion resistance (Warburg impedance) caused by ion diffusion, and the intersection of the EIS curve and the horizontal axis in the high-frequency region is represented as the system resistance (R_s), corresponding to the resistance of the electrolyte. The R_ct is an important parameter in electrochemical systems, which affects the charge–discharge performance, energy conversion efficiency, cycling stability of batteries, etc. The R_ct of CoS₂/BNC/S is 5.15 Ω, which is smaller than that of BNC/S (5.62 Ω) and AC/S (16.46 Ω). This is because metal sulfides and boron–nitrogen doping promotes the increase in the conductivity of pitch carbon and the reduction of R_ct.

The constant-current charge–discharge curves of AC/S, BNC/S and CoS₂/BNC/S at 0.1C are shown in Fig. 6(a). During the discharging process, these three materials have two distinct discharge plateaus at ∼2.3 V and ∼2.1 V.^33,34 This is because the outer-layer electrons of S have a multi-electron structure, and the reduction reaction proceeds in multiple-step electrode reactions during the reduction process, generating different lithium polysulfide intermediates. Among them, the ∼2.3 V plateau corresponds to the reduction of the S₈ cyclic molecule to soluble long-chain polysulfides Li₂S_n (6 ≤ n ≤ 8). Subsequently, the “steep” oblique line represents the liquid–liquid conversion reaction, corresponding to the conversion of soluble long-chain polysulfides to short-chain polysulfides Li₂S_n (3 ≤ n ≤ 5). After further lithiation, a relatively long plateau with a voltage of ∼2.1 V appears, corresponding to the two-phase reduction process of the conversion of soluble short-chain polysulfides to solid short-chain sulfides Li₂S₂/Li₂S. The decreasing part below 2.1 V corresponds to the transformation process from insoluble Li₂S₂ to Li₂S. The charging process is the reverse of the discharging process, but it has only one charging plateau. This is mainly because the reaction polarization during the oxidation process is large, resulting in the superposition of two oxidation plateaus. As shown in the figure, the initial discharge capacities of AC/S, BNC/S and CoS₂/BNC/S at a current density of 0.2C are 1038.4, 1424.7 and 1639.1 mAh g⁻¹, respectively, and the charging capacities are 998.0, 1324.6 and 1624.1 mAh g⁻¹, respectively. The specific discharge capacities at the around 2.3 V plateau (Q_H) and the around 2.1 V plateau (Q_L) for the initial discharge were 311 mAh g⁻¹ and 727 mAh g⁻¹ for AC/S and 444 mAh g⁻¹ and 980 mAh g⁻¹ for BNC/S. The Q_L/Q_H ratios were 2.34 and 2.20 for AC/S and BNC/S, respectively. While the Q_H and Q_L of CoS₂/BNC/S are 483 and 1156, respectively, the Q_L/Q_H ratios were 2.39. A larger Q_L/Q_H capacity ratio suggests more efficient active material conversion, with CoS₂/BNC/S exhibiting the highest initial discharge capacity and Q_L/Q_H ratio, suggesting the highest utilization of active material when further reducing soluble lithium polysulfides (Li₂S_x, 4 ≤ x ≤ 8) to insoluble lithium sulfides (Li₂S₂, Li₂S). The voltage difference of the charge–discharge plateau of BNC/S is 0.19 V, which is significantly smaller than 0.24 V of AC/S. The polarization is small, resulting in a high sulfur utilization rate, indicating that the B/N co-doped porous carbon materials promote the redox reaction kinetics of lithium–sulfur batteries. CoS₂/BNC has a minimum potential difference of 0.18 V and the minimum polarization, indicating that the conductivity of the boron–nitrogen co-doped carbon and the ability to catalyze the conversion of polysulfides are further improved after loading CoS₂ nanoparticles.

Fig. 6(b) shows a partial enlarged view of the conversion of Li₂S₂/Li₂S to polysulfides in the charging curves of AC/S, BNC/S and CoS₂/BNC/S. It can be seen from the figure that the potential difference of the AC/S electrode material is only 48.5 mV. Compared with AC/S, larger over-potentials (89.8 mV and 85 mV) were observed in BNC/S and CoS₂/BNC/S, respectively. This indicates that the precipitation amounts of Li₂S₂/Li₂S in the BNC/S and CoS₂/BNC/S electrode materials are larger and they have higher conversion rates of polysulfide capacity. Moreover, there is little difference between BNC/S and CoS₂/BNC/S. Fig. 6(c) shows a partially enlarged view of the conversion of polysulfides to Li₂S₂ in the discharge curves of AC/S, BNC/S and CoS₂/BNC/S. It can be seen from the figure that the potential differences of the electrode materials of AC/S, BNC/S and CoS₂/BNC/S are 56.4, 44.5 and 16.3 mV, respectively. It shows that under the action of loading CoS₂ and co-doping of boron and nitrogen, soluble polysulfides (Li₂S₄) are more easily converted into insoluble polysulfides (Li₂S₂ and Li₂S).

The rate performance curves of AC/S, BNC/S and CoS₂/BNC/S are shown in Fig. 6(d). It can be seen that when the current density of AC/S is 0.2, 0.5, 1, 2 and 4C, the average discharge specific capacities are 894.1, 738.7, 677.5, 604.8 and 226.3 mAh g⁻¹, respectively. When the current density returns to 0.2C, the discharge specific capacity returns to 759.4 mAh g⁻¹. The performance of the pure asphalt carbon material is poor at high current densities. BNC/S shows better rate performance, the average discharge specific capacities are 1109.7, 918.4, 838.4, 759.5 and 689.7 mAh g⁻¹, respectively, and the rate performance is significantly improved. When the current density returns to 0.2C, the discharge specific capacity of BNC/S returns to 933.4 mAh g⁻¹. CoS₂/BNC/S exhibits the best rate performance and long-cycle stability, and the average specific capacities are 1409.8, 1091.1, 948.8, 840.2 and 743.7 mAh g⁻¹, respectively. When the current density returns to 0.2C, the average discharge capacity is 1061.4 mAh g⁻¹, indicating that the CoS/BNC/S composite electrode structure is relatively stable.

Further comparison of the charge–discharge curves of AC/S, BNC/S and CoS₂/BNC/S at current densities from 0.2C to 4C is shown in Fig. 6(e). Compared with AC/S and BNC/S, CoS₂/BNC/S shows longer and more stable voltage plateaus, and the voltage difference between the charge–discharge plateaus is smaller, indicating that the CoS₂/BNC/S composite electrode exhibits low polarization and fast reaction kinetics and can relieve the shuttle effect. In addition, at all the detected C rates, the charge–discharge curves of CoS₂/BNC/S show a longer secondary discharge plateau. At all the inspected C rates, the specific capacities of the first and second discharge plateaus (Q_H:Q_L) of the CoS₂/BNC/S battery are closer to the theoretical value of 1:3, indicating that with the help of the catalyst, the conversion and deposition of LiPSs are more complete, further proving that CoS₂/BNC/S promotes the conversion process of polysulfides.

The cycle performance curves of AC/S, BNC/S and CoS₂/BNC/S at 0.5C are shown in Fig. 7(a). It can be seen from the curves that the initial discharge capacities of AC/S, BNC/S and CoS₂/BNC/S are 875.6, 1334.6 and 1586.1 mAh g⁻¹, respectively, and the remaining capacities after 100 cycles are 662.8, 871.1 and 928.6 mAh g⁻¹, respectively. The capacity of boron–nitrogen co-doped asphalt porous carbon is higher than that of asphalt porous carbon, showing good cycle stability and high sulfur utilization rate. It indicates that boron–nitrogen co-doped porous carbon can improve the conductivity of the positive electrode material and the anchoring performance of polysulfides. This is mainly because BNC/S has a relatively high specific surface area, and the shuttle effect of polysulfides is alleviated through the synergistic effect of physical and chemical adsorption. Compared with BNC/S, the initial discharge specific capacity and the capacity after 100 cycles of CoS₂/BNC/S are further improved. This improvement in cycle performance is mainly attributed to the loaded CoS₂ nanoparticles. The existence of CoS₂ accelerates the reaction between soluble LiPSs and insoluble Li₂S₂/Li₂S by adsorbing LiPSs on the surface of the conductive carbon substrate, significantly improving the utilization rate of active sulfur. CoS₂/BNC/S shows excellent cycle performance, and the coulombic efficiency is close to 100%. This electrochemical performance difference is mainly due to the strong chemical interaction between CoS₂ nanoparticles and polysulfides, further improving the adsorption capacity and catalytic conversion ability of polysulfides. The cycle performance curves of AC/S, BNC/S and CoS₂/BNC/S at 1C are shown in Fig. 7(b). It can be seen from the curves that the initial discharge capacities of AC/S, BNC/S and CoS₂/BNC/S are 750.1, 1319.6 and 1350.3 mAh g⁻¹ respectively, and the remaining capacities after 100 cycles are 652.6, 798.6 and 855.3 mAh g⁻¹ respectively. At a current density of 1C, the cycle performance of BNC/S is also better than that of AC/S, proving the inhibitory effect of boron- and nitrogen-doping on polysulfides and effectively relieving the impact of the shuttle effect. The cycle performance of CoS₂/BNC/S is improved compared with that of BNC/S, further verifying the catalytic effect of CoS₂.

To evaluate the long-cycle stability of the AC/S, BNC/S and CoS₂/BNC/S composite electrodes, a 500 cycle test was carried out at a current density of 2C, as shown in Fig. 7(c). The initial discharge capacity of AC/S is 713.6 mAh g⁻¹. After 500 cycles, the remaining capacity of AC/S is 393.7 mAh g⁻¹, and the capacity attenuation rate per cycle is 0.089%. The initial discharge capacity of BNC/S is 831.1 mAh g⁻¹. After 500 cycles, the remaining capacity of BNC/S is 597.3 mAh g⁻¹, and the capacity attenuation rate per cycle is 0.056%. The CoS₂/BNC/S positive electrode material has a high initial capacity of 996.5 mAh g⁻¹. After 500 cycles, the capacity is still 752.7 mAh g⁻¹, and the retention rate is 75.5%. This retention rate corresponds to a capacity fade of 0.048% per cycle. Compared with AC/S and BNC/S materials, the CoS₂/BNC/S material significantly improves cycle stability and long-cycle life.

Table 2 summarizes the electrochemical performance of the CoS₂/BNC/S electrode and other doped carbon materials loaded with transition metal sulfides. It is evident that the CoS₂/BNC/S electrode has obvious advantages in terms of high rate capability and cycling stability.

---

The low sulfur content in the positive electrode will lead to a decrease in the actual energy density of the battery, and overcoming this problem is crucial for releasing the application potential of lithium–sulfur batteries. A high-loading battery was assembled with CoS₂/BNC/S as the positive electrode (sulfur loading of 5 mg cm⁻²) and subjected to long-cycle tests. As shown in Fig. S1, the CoS₂/BNC/S electrode exhibits stable cycling performance under high sulfur loading and lean electrolyte conditions. After cycling 100 times at 0.5C, the average reversible capacity is 757 mAh g⁻¹, and the average capacity loss per cycle is 0.174%.

4. Conclusion

In this work, CoS₂/porous boron–nitrogen co-doped carbon (CoS₂/BNC) heterogeneous porous carbon was prepared for the positive electrode of lithium–sulfur batteries. This material has abundant pore structures, and appropriate specific surface area and pore size distribution, which are beneficial to the electrolyte and ion transport. Compared with AC and BNC, CoS₂/BNC has the best catalytic effect on the adsorption of polysulfides and the liquid–solid conversion reaction. This is due to the synergy of boron and nitrogen atoms and CoS₂ nanoparticles, which can promote redox kinetics, accelerate the conversion of polysulfides, and reduce the reaction polarization. Electrochemical tests show that the CoS₂/BNC/S positive electrode material exhibits excellent performance in different tests. The volt-ampere characteristic curve shows that CoS₂ nanoparticles can improve the redox kinetics. In the constant-current charge–discharge test, the initial discharge capacity is high at 0.2C, the potential difference of the charge–discharge platform is small, and the polarization is small. The performance at different current densities is better than that of AC/S and BNC/S. In the rate performance test, it has the best rate performance and stability at 0.2–4C. In the cycle performance test, the capacity and coulombic efficiency are excellent after 500 cycles at 2C. In conclusion, the CoS₂/BNC material solves the polysulfide shuttle problem through its unique structure and synergy, improves the electrochemical performance of lithium sulfur batteries, and provides reference and ideas for the development of high-performance lithium–sulfur batteries.

Author contributions

Wei Dong: conceptualization, writing – review & editing, and funding acquisition. Zhaomeng Wu: writing – original draft. Xiaochen Xu: visualization. Su Li: methodology. Yang Ji: investigation.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data availability section has been verified. The content is accurate, and all external links are functional.

Acknowledgements

This work was supported by the Foundation of Liaoning Provincial Key Laboratory of Energy Storage and Utilization (CNWK202405 and CNNK202422) and the Basic Scientific Research Project of Liaoning Provincial Department of Education (JYTMS20230070).

References

1. G. Li, Z. Chen and J. Lu, Chem, 2018, 4, 3–7 CAS .
2. A. Rosenman, E. Markevich, G. Salitra, D. Aurbach, A. Garsuch and F. Chesneau, Adv. Energy Mater., 2015, 5, 1500212 CrossRef .
3. Z. Xiao, Z. Yang, L. Zhou, L. Zhang and R. Wang, ACS Appl. Mater. Interfaces, 2017, 9, 18845–18855 CrossRef CAS PubMed .
4. Z. Ma, F. Jing, Y. Fan, J. Li, Y. Zhao and G. Shao, J. Alloys Compd., 2019, 789, 71–79 CrossRef CAS .
5. H. Shi, W. Lv, C. Zhang, D. Wang, G. Ling, Y. He, F. Kang and Q. Yang, Adv. Funct. Mater., 2018, 28, 1800508 CrossRef .
6. H. Xu, Y. Liu, Q. Bai and R. Wu, J. Mater. Chem. A, 2019, 7, 3558–3562 RSC .
7. X. Ji, K. Lee and L. Nazar, Nat. Mater., 2009, 8, 500–506 CrossRef CAS PubMed .
8. Z. Seh, Y. Sun, Q. Zhang and Y. Cui, Chem. Soc. Rev., 2016, 45, 5605–5634 RSC .
9. C. Li, Z. Wang, Q. Wang and D. Gu, Rare Met., 2017, 36, 365–380 CrossRef CAS .
10. Y. Song, X. Long, Z. Luo, C. Guo, C. Geng, Q. Ouyang, Z. Han, G. Zhou and J. Shao, ACS Appl. Mater. Interfaces, 2022, 14, 32183–32195 CrossRef CAS PubMed .
11. M. Zheng, Z. Luo, Y. Song, M. Zhou, C. Guo, Y. Shi, L. Li, Q. Sun, B. Shi, Z. Yi, F. Su, J. Shao and G. Zhou, J. Power Sources, 2023, 580, 233445 CrossRef CAS .
12. J. Wu, Z. Pan, Y. Zhang, B. Wang and H. Peng, J. Mater. Chem. A, 2018, 6, 12932–12944 RSC .
13. S. Zhang, Inorg. Chem. Front., 2015, 2, 1059–1069 RSC .
14. X. Ren, Z. Liu, M. Zhang, D. Li, S. Yuan and C. Lu, ChemElectroChem, 2021, 8, 3457–3471 CrossRef CAS .
15. X. Liu, J. Huang, Q. Zhang and L. Mai, Chem. Eng. J., 2017, 29, 1601759 Search PubMed .
16. J. Zhang, Z. Li, Y. Chen, S. Gao and X. Lou, Angew. Chem., 2018, 130, 11110–11114 CrossRef .
17. L. You, H. Tang, P. Wang, W. Song, P. Ji, W. Xu, C. Feng and J. Liu, J. Alloys Compd., 2019, 795, 531–539 CrossRef .
18. J. Paraknowitsch and A. Thomas, Energy Environ. Sci., 2013, 6, 2839–2855 RSC .
19. W. Wang, X. Wang, J. Shan, L. Yue, Z. Shao, L. Chen, D. Lua and Y. Li, Energy Environ. Sci., 2023, 16, 2669–2683 RSC .
20. Y. Song, S. Zhu, X. Long, Z. Luo, Q. Sun, C. Geng, H. Li, Z. Han, Q. Ouyang, G. Zhou and J. Shao, Electrochim. Acta, 2023, 441, 141875 CrossRef CAS .
21. J. Wang and W. Han, Adv. Funct. Mater., 2022, 32, 2107166 CrossRef CAS .
22. H. Ci, M. Wang, Z. Sun, C. Wei, J. Cai, C. Lu, G. Cui, Z. Liu and J. Sun, J. Energy Chem., 2022, 66, 474–482 CrossRef CAS .
23. J. Zhu, R. Pitcheri, T. Kang, C. Jiao, Y. Guo, J. Li and Y. Qiu, J. Electroanal. Chem., 2019, 833, 151–159 CrossRef CAS .
24. S. Ng, M. Y. Lau and W. Ong, Adv. Mater., 2021, 33, 2008654 CrossRef CAS PubMed .
25. N. Li, T. Meng, L. Ma, H. Zhang, J. Yao, M. Xu, C. Li and J. Jiang, Nano-Micro Lett., 2020, 12, 145 CrossRef CAS PubMed .
26. J. Xu, Adv. Sci., 2021, 8(16), 2101019 CrossRef CAS PubMed .
27. S. Xu, D. McOwen, L. Zhang, G. Hitz, C. Wang, Z. Ma, C. Chen, W. Luo, J. Dai, Y. Kuang, E. Hitz, K. Fu, Y. Gong, E. Wachsman and L. Hu, Energy Storage Mater., 2018, 15, 458–464 CrossRef .
28. S. Seo, D. Park, S. Park and D. Kim, Adv. Funct. Mater., 2019, 29, 1903712 CrossRef .
29. C. Toh, H. Zhang, J. Lin, A. Mayorov, Y. Wang, C. Orofeo, D. Ferry, H. Andersen, N. Kakenov, Z. Guo, I. Abidi, H. Sims, K. Suenaga, S. Pantelides and B. Özyilmaz, Nature, 2020, 577, 199–203 CrossRef CAS PubMed .
30. C. Yang, Y. Li, W. Peng, F. Zhang and X. Fan, Chem. Eng. J., 2022, 427, 131792 CrossRef CAS .
31. W. Feng, Y. Niu, X. Zheng, W. Su, J. Chen and L. Zhang, J. Mater. Sci.: Mater. Electron., 2023, 34, 779 CrossRef CAS .
32. Y. Zhu, Y. Zuo, X. Jiao, R. Manjunatha, E. Ezeigwe, W. Yan and J. Zhang, Carbon Energy, 2022, 5, e249 CrossRef .
33. Y. Luo, Z. Fang, S. Duan, H. Wu, H. Liu, Y. Zhao, K. Wang, Q. Li, S. Fan, Z. Zheng, W. Duan, Y. Zhang and J. Wang, Angew. Chem., 2023, 62, e202215802 CrossRef CAS PubMed .
34. Z. Yang, R. Li and Z. Deng, ACS Appl. Mater. Interfaces, 2018, 10, 13519–13527 CrossRef CAS PubMed .
35. C. Chen, S. Lin, Y. Wu, J. Su, C. Cheng, P. Cheng, Y. Ting and S. Lu, Chem. Eng. J., 2022, 431, 133924 CrossRef CAS .
36. H. You, F. Liu, H. Wang, Z. Wang, X. Wang, B. Zhang, K. Tang, J. Ren and R. Wang, ACS Sustainable Chem. Eng., 2025, 13, 2154–2163 CrossRef CAS .
37. X. Zhang, Y. Cui, Y. Zhong, D. Wang, W. Tang, X. Wang, X. Xia, C. Gu and J. Tu, J. Alloys Compd., 2019, 551, 219–226 CAS .
38. H. Zhao, J. Wu, T. Chen, P. Yan, W. Yao, X. Ma, Y. Sun, W. Wang and M. Shi, J. Energy Storage, 2024, 101, 113903 CrossRef .

This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2025