Citation

Wang Wenxue, Bai Longwei, Li Na, Zhao Shuo, Shi Xiaodong, Wang Peng. Selenium-doping metal phosphides as bifunctional catalyst carrier for durable lithium-sulfur batteries[J]. Chin. Chem. Lett., 2025, 36(10): 110938  

Wenxue Wang^a,1, Longwei Bai^a,1, Na Li^a,d,*, Shuo Zhao^b,*, Xiaodong Shi^c,*, Peng Wang^a,*     

As the increasing consumption of fossil energy, developing renewable energy has become increasingly significant [1–3]. To efficiently store renewable energy such as wind and solar, it is an urgent require but still great challenge to design advanced electrochemical energy storage devices [4–6]. Lithium-sulfur (Li-S) batteries have garnered wide interest for the merits of high theoretical capacity (1675 mAh/g), high energy density (2600 Wh/kg), abundant resources, and environmental friendliness [7–9]. Nevertheless, the practical development of Li-S batteries is still hindered by the electrochemical dissolution of active sulfur, serious shuttle effect of lithium polysulfides (LiPSs), and sluggish conversion reaction kinetics [10–12].

According to the previous reports, designing porous carriers with high catalytic activity for sulfur conversion reaction and strong adsorption capability for LiPSs, is deemed as bifunctional strategy to simultaneously address the above problems [13–16]. Especially, transition metal compounds decorated with porous nitrogen-doped carbon held the intrinsic advantages of facile synthesis method, high specific surface area, good electrical conductivity and high catalytic activity, delivering great promise for practical applications [17–23]. In general, the interaction between the inner metal compounds and the outer carbon layer contributes to construct a three-dimensional conductive network, which can effectively enhance the interfacial electron/charge transfer behavior, boost the electronic conductivity, and reduce the reaction energy barrier [24–26]. Additionally, the carbon shell also functions as stress buffer to maintain the structural stability of cathode materials during cycling process. However, the conventional metal compounds-based catalysts cannot realize the utilization of high sulfur-loading cathode materials, which always needs higher requirements for the active sites and catalytic activity of carrier materials. Heteroatom doping is considered as an effective approach to introduce structural defects [27], regulate the active sites as well as band structures of metal compounds-based catalysts from atomic level, which can reinforce both the catalytic activity and chemisorption capability for LiPSs [28]. As reported, 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 [29,30].

Herein, porous nitrogen-doped carbon coated Se-doping CoP composite (Se-CoP@NC) is facilely prepared with Co-based metal-organic frame as the precursor, and served as the functional sulfur-loading carrier. As demonstrated by theoretical calculations, the introduction of Se-doping sites effectively connects the interface of CoP core and NC shell to guarantee fast electron/charge transfer and efficient catalytic conversion reaction. Meanwhile, the Se-doping also enlarges the lattice spacing of CoP, promotes the chemical adsorption capability for LiPSs, and reduces the conversion reaction energy barriers of LiPSs, which is conducive to the nucleation growth of Li₂S, as well as the reduction reaction kinetics of LiPSs during discharge process. Owing to these merits, the optimized S/Se-CoP@NC cathode can deliver high reversible capacity of 779.6 mAh/g after 500 cycles at 0.5 C, and stable capacity of 805.9 mAh/g at high rate of 2 C. Even at high sulfur-loading content of 6.9 mg/cm² and lean electrolyte dosage of 7 µL/mg, it still exhibits reversible areal capacity of 4.5 mAh/cm² after 100 cycles at 0.1 C. Combined with the in-situ UV–vis results, these satisfactory battery performances are mainly ascribed to the positive inhibition effect for LiPSs shuttle behavior of Se-CoP@NC catalyst. This work will help to design the transition metal compounds-based carrier materials with high catalytic activity, and promote the applications of high sulfur-loading cathode materials in Li-S batteries.

X-ray diffraction (XRD) patterns were conducted to recognize the influences of selenization treatment on the crystalline structure of CoP@NC. The diffraction peaks of Se-CoP@NC are consistent with those of CoP@NC (JCPDF No. 29–0497, CoP) and there are no additional diffraction peaks belonging to CoSe, indicating that Se-doping does not change the crystalline structure of CoP (Fig. 1a). It is noteworthy that slight left shift is observed in the position of (011), (112) and (211) diffraction peaks of Se-CoP@NC, implying the enlarged lattice spacings. Combined with the refinement results of XRD patterns (Fig. S1 and Table S1 in Supporting information), the enlarged lattice spacing in Se-CoP@NC can be attributed to the lattice distortion, which is induced by the larger atomic radius of Se than P atom [31–33]. Scanning electron microscope (SEM) images of Se-CoP@NC (Fig. 1b) and CoP@NC (Fig. S2a in Supporting information) samples reveal the similar dodecahedral nanostructures with an average diameter of 200 nm, while transmission electron microscope (TEM) images of Se-CoP@NC (Fig. 1c) and CoP@NC (Fig. S2b in Supporting information) samples manifest the encapsulation structure of porous carbon for Se-CoP/CoP nanoparticles. Fig. 1d reveals the uniform distribution of Co, P and Se elements, and the semiquantitative energy dispersive spectrometer result suggests the successful Se-doping into CoP (Table S2 in Supporting information) [34]. Based on the high-resolution TEM images, the marked lattice spacing of Se-CoP@NC (Fig. 1e) is 0.205 nm, larger than that of CoP@NC sample (0.202 nm, Fig. S2c in Supporting information), which can be denoted as (112) crystal plane. Meanwhile, according to the corresponding electron diffraction patterns, the lattice spacing for (101) and (111) crystal planes of Se-CoP@NC (Fig. 1f) are also larger than those of CoP@NC (Fig. S2d in Supporting information), which is consistent with the XRD results. X-ray photoelectron spectroscopy (XPS) was employed to further confirm the introduction of Se-doping [35]. The survey XPS spectra verify the existence of Co, N, C, P and Se elements in Se-CoP@NC sample (Fig. 1g). The high-resolution Co 2p spectra indicates the weaker intensity of Co-P chemical bond in Se-CoP@NC samples (Fig. 1h), implying the bonding environment changes of Co element [36]. For the high-resolution P 2p spectra, the characteristic peaks of CoP@NC at 129.4 and 130.5 eV corresponding to the P-Co bonds, also delivers stronger intensity than that of Se-CoP@NC sample (Fig. S3 in Supporting information) due to the electronic state changes of Co and P atoms accompanying with the Se-doping [37]. For the high-resolution Se 3d spectra of Se-CoP@NC, the characteristic peaks around 55.3 and 56.5 eV can be assigned as Se 3d_5/2 and Se 3d_3/2, respectively, corresponding to the Se-Co existence form of Se-doping in CoP [38].

Figure 1

	Download: JPG larger image
Fig. 1. (a) XRD patterns of NC, CoP@NC and Se-CoP@NC samples. (b) SEM image, (c) HAADF and elemental mapping images, (d) TEM image, (e) high-resolution TEM image, and (f) electron diffraction pattern of Se-CoP@NC sample. (g) Survey XPS and (h) high-resolution Co 2p spectra of CoP@NC and Se-CoP@NC samples. (i) High-resolution Se 3d spectra of Se-CoP@NC sample.

To reveal the active mechanism of Se-doping, density functional theory (DFT) calculations were implemented based on the optimized crystalline models of CoP@NC, CoSe@NC and Se-CoP@NC (Figs. S4 and S5 in Supporting information). Fig. 2a presents the differential charge density distributions of CoP@NC and Se-CoP@NC models, and it is evident that the accumulated charge around the Se-doping site is closer to the nitrogen-doped carbon (NC) layer, thereby affecting the charge distribution within the interface of Se-CoP@NC [39]. Fig. 2b delivers the charge density distributions of Co element in CoP@NC and CoSe@NC models. Notable, the NC component is more closely bound to CoSe, and the charge distribution radius of Co element in CoSe@NC is larger than that of CoP, manifesting that CoSe is more likely to transfer electron to the NC component. Figs. 2c and d and Fig. S6 (Supporting information) exhibit the density of states (DOS) of each atom in CoP@NC, CoSe@NC and Se-CoP@NC models. Obviously, the DOS of CoP is only slightly changed by Se doping, while the DOS of CoSe phase is significantly changed compared with CoP. It can be concluded that Se-doping influences the charge transfer between CoP and NC component, but the existence of CoSe phase significantly promotes the charge transfer behavior. Furthermore, the corresponding Gibbs free energies for the conversion reactions from S₈ to Li₂S driven by the carrier materials of CoP@NC and Se-CoP@NC were also calculated by DFT calculations (Figs. S7 and S8 in Supporting information). Fig. 2e compares the calculated Gibbs free energies for the conversion reactions of LiPSs driven by different carrier materials. For the intermediates from S₈ to Li₂S₂, the corresponding Gibbs free energy of CoP@NC and Se-CoP@NC carriers remains negative values, implying the spontaneous occurrence of the continuous reduction reactions, which are not considered as the rate-determining step in the whole reduction process. For the intermediates from Li₂S₂ to Li₂S, the corresponding Gibbs free energy rises up, demonstrating it is the rate-determining step [40]. As a result, the Gibbs free energy change associated with the rate-determining step in Se-CoP@NC is only 1.11 eV, lower than that of CoP@NC (1.13 eV), suggesting the introduction of Se-doping reduces the reaction energy barriers and promotes the reaction kinetics [41,42].

Figure 2

	Download: JPG larger image
Fig. 2. (a) Differential charge density distributions of CoP@NC and Se-CoP@NC models. (b) Elemental valence-edge charge density distributions of CoP@NC and CoSe@NC models; Atomic density of states (spin up) of (c) CoP@NC and Se-CoP@NC models and (d) CoP@NC and CoSe@NC models. (e) Gibbs free energies for the conversion reactions of LiPSs driven by the carrier materials of CoP@NC and Se-CoP@NC.

To investigate the catalytic properties of different samples, the corresponding symmetric cells were assembled. Fig. 3a compares the CV curves of NC, CoP@NC and Se-CoP@NC based symmetric cells, in which the Se-CoP@NC electrode presents higher redox current, suggesting its enhanced conversion reaction kinetics of LiPSs [43]. Meanwhile, Se-CoP@NC symmetric cells delivers lower interface impedance, indicating faster charge transfer behavior (Fig. 3b and Table S3 in Supporting information). Tafel curves of Se-CoP@NC symmetric cells demonstrate the highest exchange current density (Fig. S9a in Supporting information), while LSV curves of Se-CoP@NC based three-electrode cells displays the highest current response (Fig. S9b in Supporting information), indicating the strengthened LiPSs reaction kinetics derived by Se-CoP@NC carrier. To disclose the inhibiting shuttle effect on the LiPSs, adsorption tests of CoP@NC and Se-CoP@NC samples were conducted. Obviously, the yellow Li₂S₆ solution becomes more transparent after aging for 6 h (Fig. S10 in Supporting information), proving that Se-CoP@NC has a stronger adsorption capability for Li₂S₆ component. Subsequently, the liquid supernatant after adsorption experiment was transferred to a sealed colorimetric dish, and the change of Li₂S₆ concentration was reflected by UV–vis absorption spectra. In details, the initial Li₂S₆ solution exhibits two significant absorption peaks at 370 nm and 420 nm, which cannot be detected after fully adsorbed by Se-CoP@NC powder (Fig. 3c). This result manifests that the introduction of selenium-doping into Se-CoP@NC composite will induce structural defects, regulate the active sites as well as band structure from atomic levels, which effectively contributes to strengthen the catalytic reactivity and chemisorption capability for LiPSs. To verify the practical availability of Se-CoP@NC carrier in Li-S batteries, S/Se-CoP@NC composite cathode was prepared, and its relevant material characterizations were presented in Figs. S11-S14 (Supporting information) [44,45]. Fig. 3d and Fig. S15 (Supporting information) respectively displays the CV curves of S/Se-CoP@NC and S/CoP@NC cathodes at different scan rates (0.05–0.25 mV/s), both of them contain two reduction peaks and one oxidation peak, corresponding to the reduction reaction from S₈ to soluble LiPSs, the conversion of soluble LiPSs to Li₂S₂/Li₂S, as well as the oxidation reaction from Li₂S₂/Li₂S to S₈. Fig. 3e presents the corresponding linear fitting plots, and the slopes of S/Se-CoP@NC-based batteries are higher than those of S/CoP@NC-based batteries, manifesting the thereby accelerating reaction kinetics. To intuitively analyze the effect of different catalysts on the reduction process, Li₂S deposition test was conducted based on constant potential discharge process. As shown in Figs. 3f and g and Fig. S16a (Supporting information), the peak current of Se-CoP@NC appears earliest (704 s), i.e., the earliest nucleation of Li₂S, suggesting Se-CoP@NC promotes the Li₂S deposition behavior. Meanwhile, the corresponding nucleation capacity of Se-CoP@NC is as high as 198.44 mAh/g, which is almost twice and five times that of CoP@NC and NC, respectively. Moreover, the Li₂S dissolution test was also performed to study the effect of different catalysts on the oxidation process. As exhibited in Figs. 3h and i and Fig. S16b (Supporting information), Se-CoP@NC shows the highest Li₂S decomposition capacity of 525.8 mAh/g and the earliest peak current response time at 1530 s. Both the Li₂S deposition and Li₂S dissolution tests of different catalysts strongly demonstrate that Se-CoP@NC delivers excellent catalytic activity for the conversion reaction of LiPSs [46].

Figure 3

Download: JPG larger image

Fig. 3. (a) CV curves of NC, CoP@NC and Se-CoP@NC based symmetric cells. (b) Electrochemical impedance profiles of CoP@NC and Se-CoP@NC based symmetric cells. (c) LiPSs adsorption test of CoP@NC and Se-CoP@NC materials, and the corresponding UV–vis spectra results. (d) CV curves of S/Se-CoP@NC cathode at different scan rates, and the corresponding. (e) Linear fitting plots of the square roots of peak currents with scan rates. Constant voltage discharge curves for Li2S nucleation process driven by (f) Se-CoP@NC and (g) CoP@NC materials; Constant voltage charge curves for Li2S decomposition process driven by (h) Se-CoP@NC and (i) CoP@NC materials.

To vividly evaluate the ability of Se-CoP@NC to prevent the shuttle effect of LiPSs, home-made Li-S devices were assembled in a sealed electrolyzer with different materials as cathodes. During the low-rate discharge process, S/NC and S/CoP@NC cathode initially releases yellow-green LiPSs, which subsequently dissolve into the liquid electrolyte (Fig. S17 in Supporting information). In contrast, the color change of electrolyte with S/Se-CoP@NC cathode is not obvious, implying its pronounced inhibiting effect on the shuttle effect of LiPSs. Furthermore, in-situ UV–vis spectroscopy was performed to monitor the changes of intermediate products in Li-S batteries during discharge process assembled with different cathodes (S/Se-CoP@NC, S/CoP@NC, S/NC) [47,48]. According to Eqs. S1-S6 (Supporting information), S₃^- specie is generated from the decomposition of S₆^2-, while S₃^2- and S₄^2- species are derived from the reduction reaction of S₃^- [49–51]. Based on the contour images of in-situ UV–vis spectra (Fig. 4), the signal intensity of LiPSs in S/Se-CoP@NC cathode during discharge process (Fig. 4a) is much weaker than those in S/CoP@NC (Fig. 4b) and S/NC cathodes (Fig. 4c) with the same absorbance level, manifesting the stronger LiPSs adsorption capability and effective inhibition of LiPSs shuttle ability of Se-CoP@NC catalyst [52–54]. Fig. 5a carefully contrasts the CV curves of different cathodes at 0.1 mV/s, and the narrower redox potential gap of S/Se-CoP@NC cathode indicates its faster conversion reaction kinetics than that of S/CoP@NC cathode. Fig. 5b compares the galvanostatic charge/discharge curves of S/Se-CoP@NC, S/CoP@NC and S/NC cathodes at 0.5 C, and the discharge capacity is divided into ΔQ₁ and ΔQ₂ based on the characteristic discharge platforms. Especially, ΔQ₁ represents the capacity from the first discharge platform, which is related to the ring-opening reaction of S₈ and the formation of soluble LiPSs. ΔQ₂ represents the capacity from the second discharge plateau, which is related to the conversion reactions from long-chain LiPSs to Li₂S₂/Li₂S [55].

Figure 4

	Download: JPG larger image
Fig. 4. In-situ UV–vis spectra of Li-S batteries with (a-c) S/Se-CoP@NC, (d-f) S/CoP@NC, and (g-i) S/NC composites as the cathodes.

Figure 5

Download: JPG larger image

Fig. 5. (a) Comparison of CV curves of S/Se-CoP@NC and S/CoP@NC cathodes at 0.1 mV/s. (b) Comparison of the initial galvanostatic charge and discharge curves of S/Se-CoP@NC, S/CoP@NC and S/NC cathodes at 0.5 C. (c) Capacity contribution ratios of different cathodes from the corresponding discharge platforms. (d) Electrochemical impedances of S/Se-CoP@NC and S/CoP@NC cathodes. (e) Charge and discharge curves of S/Se-CoP@NC cathode based on GITT test. (f) Internal resistances of different cathodes associated with standardized discharge-charge time. (g) Rate performance of S/Se-CoP@NC, S/CoP@NC and S/NC cathodes (25 µL electrolyte). (h) Long-term cycling stability of different cathodes at 0.5 C (40 µL electrolyte).

As summarized in Fig. 5c, the ΔQ₂/ΔQ₁ ratio of S/Se-CoP@NC cathode is 1.53, higher than those of S/CoP@NC (1.37) and S/NC (1.48) cathode, suggesting the dominant reduction reaction from LiPSs to Li₂S₂/Li₂S driven by Se-CoP@NC catalyst. Fig. S18 (Supporting information) reveals S/Se-CoP@NC cathode exhibits lower Tafel slopes for both oxidation and reduction process, further indicating the facilitated sulfur redox kinetics. Additionally, based on the EIS results of Li-S batteries (Fig. 5d and Table S3 in Supporting information), S/Se-CoP@NC cathode delivers smaller charge transfer resistance than that of S/CoP@NC, confirming the enhanced transportation behavior of electron, charge, and Li⁺ ion on the surface of S/Se-CoP@NC. Fig. 5e and Fig. S19 (Supporting information) further presents the charge/discharge curves of different cathodes based on the galvanostatic intermittent titration technique (GITT) test. Notably, S/Se-CoP@NC cathode holds the lowest open circuit voltage (OCV)/closed circuit voltage (CCV) gap (36.3 mV) than those of S/CoP@NC (50.2 mV) and S/NC (76.5 mV) cathodes during discharge process, demonstrating the excellent resistance to self-discharge behavior. Moreover, the dynamic internal resistances of these cathodes can be also calculated and presented in Fig. 5f, and S/Se-CoP@NC cathode has the lowest dynamic resistance values throughout the cycling process [56]. Fig. 5g evaluates rate performances of different cathodes, and S/Se-CoP@NC cathode exhibits the highest reversible capacity of 1435.4, 1018.3, 907.3 and 805.9 mAh/g at 0.2, 0.5, 1, and 2 C, respectively, demonstrating its excellent rate capability. When the current density recovers back to 0.5 C, the corresponding capacity stabilizes at 955.6 mAh/g (Fig. S20a in Supporting information), indicating the superior structural stability and reversibility of S/Se-CoP@NC cathode.

Fig. 5h displays the long-term cycling performances of different cathodes at 0.5 C. After 500 cycles, the remaining capacity of S/Se-CoP@NC, S/CoP@NC and S/NC cathodes are 779.6, 667.1 and 547.3 mAh/g, respectively. Compared with the previous reported sulfur-loading cathodes, both the reversible capacity and the corresponding capacity retention ratio of S/Se-CoP@NC cathode hold considerable advantages (Table S4 and Fig. S20b in Supporting information), manifesting that Se-CoP@NC carrier can effectively inhibit the active sulfur dissolution and LiPSs shuttle effect to keep stable cyclic behavior of Li-S batteries. Figs. S21a-d (Supporting information) visually investigates the LiPSs shuttle effect on the Li metal/separators coupled with different cathodes through the corresponding surface morphology characterization and separator color change. Detailedly, the cycled Li metal corresponding to S/Se-CoP@NC cathode suffers from the slightest corrosion, and negligible color change in separator, verifying the greatly inhibitory effect of Se-CoP@NC material on the shuttle effect of LiPSs, which can effectively promote the interface stability of Li metal anode and the sulfur-loading cathodes (Figs. S21e and f in Supporting information) to extend the service life of Li-S batteries. Furthermore, the cycling performances of Li-S batteries based on the S/Se-CoP@NC cathode under high sulfur-loading conditions were measured to testify its actual availability. As displayed in Fig. S22 (Supporting information), reversible areal capacity of 4.5 mAh/cm² can be still retained after 100 cycles at 0.1 C under extremely high sulfur-loading content of 6.9 mg/cm² and extremely lean electrolyte dosage of 7 µL/mg, demonstrating the application potential of S/Se-CoP@NC cathodes in practical Li-S batteries.

In summary, Se-doped CoP and nitrogen-doped carbon composite (Se-CoP@NC) is designed as functional sulfur-loading carrier for Li-S batteries. Based on the in-situ UV–vis spectra, Se-CoP@NC catalyst delivers positive effect on inhibiting the shuttle behavior of LiPSs and strong chemisorption capability for LiPSs. According to the theoretical calculations, the introduction of Se-doping can effectively enlarge the lattice spacing of CoP, promote the chemical adsorption capability for LiPSs, and reduce the reaction energy barriers of LiPSs, which is beneficial for the fast conversion reaction kinetics of active sulfur. As expected, the S/Se-CoP@NC cathode can display high reversible capacity of 779.6 mAh/g after 500 cycles at 0.5 C, and high capacity of 805.9 mAh/g at high rate of 2 C. Even at high sulfur-loading content of 6.9 mg/cm² and lean electrolyte dosage of 7 µL/mg, it still maintains high reversible areal capacity of 4.5 mAh/cm² after 100 cycles at 0.1 C, which is close to the requirements of practical application. These interesting results indicate that developing carrier materials with high catalytic activity can promote the application process of Li-S 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.

Wenxue Wang: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Longwei Bai: Visualization, Validation, Software, Resources. Na Li: Supervision, Project administration, Funding acquisition. Shuo Zhao: Visualization, Software, Investigation, Conceptualization. Xiaodong Shi: Writing – review & editing, Writing – original draft, Funding acquisition. Peng Wang: Writing – review & editing, Supervision, Project administration, Funding acquisition.

The authors thank the financial support from the National Natural Science Foundation of China (No. 52101250), the S&T program of Hebei (Nos. 215A4401D and 225A4404D), the Collaborative Innovation Center of Marine Science and Technology of Hainan University (No. XTCX2022HYC14), the Fundamental Research Funds for the Hebei University (No. 2021YWF11), the Science Research Project of Hebei Education Department (No. QN2024087), the Xingtai City Natural Science Foundation (No. 2023ZZ027), and the Key Laboratory of Advanced Energy Materials Chemistry (Ministry of Education), College of Chemistry, Nankai University. Additionally, this work was partially supported by the Pico Election Microscopy Center of Hainan University.

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110938.