Polysulfide barrier comprising bismuth selenide nanocrystals well anchored within N-doped carbon microspheres for stable Li-S batteries

INTRODUCTION

Lithium-sulfur batteries (LSBs) have garnered significant attention due to their high energy density (calculated at 2,600 Wh kg^-1), plentiful sulfur (S) resources, and low cost^[1-3]. Sulfur enables a multiple-electron (2e^-) redox activity, offering a high capacity (theoretical value of 1,675 mA h g^-1 at an average potential of approximately 2.2 V versus Li/Li^+[4,5]. However, their commercialization is hindered by several challenges, including the poor electrical conductivity of sulfur (5 × 10^-28 S m^-1) and lithium sulfide (Li₂S_x; x = 1 or 2, ~10^-11 S m^-1), the “shuttle effect” caused by the dissolution and migration of intermediate lithium polysulfide (LiPS) species (Li₂S_x, 4 ≤ x ≤ 8), significant volume change (~80%) during lithiation related to density differences between S (2.07 g cm^-3) and end product (Li₂S; 1.66 g cm^-3), and the instability of lithium metal anodes during redox processes^[6,7]. To address these issues, researchers have explored strategies such as scheming carbon-based frameworks to confine sulfur^[8-11], restraining LiPSs through physical or chemical interactions^[12,13], stabilizing lithium metal anodes^[14,15], accommodating volumetric changes^[16,17], and optimizing electrolytes^[18,19]. While these approaches have effectively improved LSB performance by suppressing polysulfide diffusion and confining polysulfides within the cathode domain, the complex synthesis of sulfur-carbon nanocomposites limits scalability. Furthermore, these processes often reduce the sulfur content in the composites, thereby diminishing energy density and electrochemical performance, posing significant barriers to large-scale implementation. An emerging strategy to address the challenges involves incorporating additional interlayer components into LSBs, either as a free-standing structure or coatings on the separator within the cell. This approach has gained significant attention due to its simplicity and practical applicability. The interlayer realizes multiple functions, including the effective confinement of polysulfides and facilitating rapid charge diffusion. To meet these requirements, the interlayer material should possess polar characteristics, offering abundant active chemisorption sites to anchor polysulfide and prevent their diffusion.

In this context, our group and others have demonstrated that separators functionalized with metal selenides exhibit promising electrochemical performance in LSBs, owing to their excellent electrocatalytic activity towards LiPS^[20-23]. Despite the extensive exploration of chalcogenides, one particularly intriguing candidate has been largely overlooked: bismuth selenide (Bi₂Se₃). This material belongs to the A₂^VB₃^VI class (A = Sb, Bi; B = Se, Te) of degenerated n-type semiconductor material, characterized by a low bandgap of 0.3 eV^[24-26]. While it has attracted significant interest in fields such as thermoelectric and optoelectronic devices, its potential application in LSBs remains unexplored^[27]. The n-type electronic behavior of Bi₂Se₃ arises from Se deficiencies, which behave similarly to electron donors. Bi₂Se₃ features a layered structure composed of covalently stacked Se-Bi-Se-Bi-Se atomic layers. This unique structure, combined with its high electrical conductivity and mutable valences of Bi and Se, offers a significant catalytic perspective.

Here, we disclose the synthesis of three-dimensional (3D) nitrogen-doped carbon microspheres embedded with bismuth selenide nanocrystals (3D Bi₂Se₃@N-C microspheres) using a scalable spray pyrolysis procedure trailed by a selenization process. The potential of these prepared microspheres as an effective polysulfide barrier to enhance reaction kinetics in Li-S batteries is systematically evaluated. The embedded Bi₂Se₃ nanocrystals within the microspheres provide abundant chemisorption sites, enabling efficient chemical captivity and transformation of LiPS species. Moreover, the N-doped carbon framework ensures rapid charge transfer, facilitating faster redox processes. Electrochemical evaluations reveal that 3D Bi₂Se₃@N-C microsphere coated separators exhibit excellent rate competence and stable cycling across different C-rates (0.1 C, 0.5 C, and 2.0 C). Notably, even at high active material loading (3.5 mg cm^-2), the cell demonstrates robust cycling stability, highlighting the effectiveness of this material design.

EXPERIMENTAL

RESULTS AND DISCUSSION

To validate the successful synthesis of 3D Bi₂Se₃@N-C microspheres, detailed physical characterizations were conducted at each step of the synthesis process. Field emission scanning electron microscopy (FE-SEM) and X-ray diffraction (XRD) analyses of the as-sprayed Bi-salt/PVP/Sucrose composite microspheres are presented in Supplementary Figure 1. The low-magnification FE-SEM micrograph [Supplementary Figure 1A] shows the formation of 3D spherical microspheres (average diameter = 1.3 µm). Additionally, the prepared powder appears uniform and free from agglomeration. The magnified FE-SEM micrograph [Supplementary Figure 1B] further reveals a smooth microsphere surface devoid of any foreign deposits, suggesting complete encapsulation of the Bi-salt within the microspheres. The XRD pattern [Supplementary Figure 1C] shows the absence of any sharp diffraction peaks, indicating the amorphous nature of the as-prepared sample.

Figure 1. (A) FE-SEM micrograph; (B) XRD profile; (C) TEM picture; and (D) HR-TEM image of the 3D Bi₂Se₃@N-C microspheres. FE-SEM: Field emission scanning electron microscopy, XRD: X-ray diffraction, TEM: transmission electron microscopy, HR-TEM: high-resolution transmission electron microscopy.

The as-sprayed Bi-salt/PVP/Sucrose composite microspheres were subjected to a selenization process under a mildly reductive atmosphere (4% of H₂/Ar) at 350 °C for 6 h. The physical characterization results of the resulting microspheres are presented in Figure 1. The FE-SEM micrograph [Figure 1A] discloses that the spherical morphology of the selenized microspheres was retained (average diameter = 1.1 µm) even after the selenization process. Furthermore, the high-magnification FE-SEM image [inset of Figure 1A] reveals significant changes in surface characteristics. Unlike the smooth surface of the as-sprayed sample, the selenized microspheres exhibit a rougher texture, with the surface covered by well-formed nanocrystals, likely Bi₂Se₃. This observation aligns with the XRD analysis shown in Figure 1B, where sharp, distinct diffraction peaks correspond to the rhombohedral crystal structure of Bi₂Se₃ with a R$$ \bar{3} $$m space group. Moreover, the average crystallite size of Bi₂Se₃ was cal. about 7.1 nm using Scherrer’s equation by considering the high-intensity reflection of the (110) plane for Bi₂Se₃. Notably, a broad peak centered in the 2θ range of 20°-25° is attributed to the carbonaceous species derived from sucrose and PVP. The presence of PVP also ensures N-doping within the carbonaceous framework owing to the presence of N-containing organic components. A low-magnification transmission electron microscopy (TEM) image [Supplementary Figure 2A] confirms the spherical morphology observed in the FE-SEM results. Furthermore, the existence of well-developed Bi₂Se₃ nanocrystals (dark patches) on the surface and within the carbon framework (grey region) of the prepared microsphere is clearly evident, as highlighted in Figure 1C. High-resolution TEM [Figure 1D] reveals Bi₂Se₃ nanocrystal with size of approximately 7 nm, which aligns with crystallite size calculated using Scherrer’s equation employing XRD pattern. Moreover, inset of Figure 1D displayed distinct lattice fringes with a separation of 0.30 nm, corresponding to the (015) diffraction plane of the Bi₂Se₃ phase. This result, along with the XRD findings, aligns with the selected area electron diffraction (SAED) pattern [Supplementary Figure 2B], which displays continuous diffraction rings corresponding to various Bi₂Se₃ planes. Elemental dot mapping images [Supplementary Figure 2C] confirm the uniform dispersal of Bi, Se, C, and N elements all over the microspheres, verifying the successful synthesis of Bi₂Se₃@N-C microspheres. Notably, the high density of N-element dots is aligned with the elemental analysis (EA) results (discussed shortly), further supporting the presence of nitrogen doping.

Based on the discussion above, Scheme 1 illustrates the detailed formation mechanism of the 3D Bi₂Se₃@N-C microspheres. Aqueous droplets, generated by the ultrasonic atomic nebulizer, contained uniformly distributed sucrose, PVP, and Bi-salt, as depicted in Scheme 1-①. These droplets were consequently gone through a vertically aligned quartz tube, stabilized at 400 °C, under a flowing Ar carrier gas. This process dried the droplets, transforming the precursor into a Bi-salt/PVP/sucrose composite [Scheme 1-②]. The resulting powders were then underwent selenization at 350 °C for 6 h. Elemental selenium was employed to generate H₂Se gas. During selenization, the PVP polymer (as a nitrogen source) and sucrose decomposed gradually, forming an N-doped C matrix, as illustrated in Scheme 1-③. This N-C framework offers structural robustness and numerous conductive corridors for electrons within the microspheres. Concurrently, the Bi-salt precursor was converted into the stoichiometric Bi₂Se₃ phase, completing the formation of the 3D Bi₂Se₃@N-C microspheres. Notably, the yield during the spray pyrolysis and subsequent selenization process was cal. approximately 51% and 55%, respectively [Supplementary Figure 3A-D]. The as-prepared 3D Bi₂Se₃@N-C microspheres were applied as multifunctional polysulfide interlayers through a slurry-coating process, as shown in Scheme 1-④.

Scheme 1. Representation of the detailed synthesis route for 3D Bi₂Se₃@N-C microspheres using spray pyrolysis technique (①-③) and their utilization as multifunctional polysulfide barrier (④). 3D: Three-dimensional.

The surface composition and bonding states within the 3D Bi₂Se₃@N-C microspheres were assessed via X-ray photoelectron spectroscopy (XPS). The survey profile presented in Figure 2A reveals the occurrence of photoelectron signals for Bi 4f, Se 3d, C 1s, and N 1s, confirming the existence of elements in the nanostructure. The Bi 4f expanded XPS spectrum in Figure 2B displays two well-fitted peaks: Bi 4f_7/2 at 157.2 eV and Bi 4f_5/2 at 162.2 eV, with a spin-orbit separation of 5 eV, indicating the presence of Bi in the +3 oxidation state within the prepared nanostructure^[28,29]. Besides, peaks at 157.8 eV and 163.0 eV correspond to the Bi₂O₃ species, which likely formed due to surface oxidation during data acquisition^[30,31]. The magnified Se 3d spectrum in Figure 2C shows two distinct peaks at 52.4 eV and 53.2 eV, corresponding to Se 3d_5/2 and Se 3d_3/2 electronic orbitals, respectively, which are attributed to Bi-Se bond in Bi₂Se₃^[32-34]. Moreover, the broad peak at 54.1 eV is attributed to the Se-O species, indicating oxidation of the specimen surface^[35,36]. The C 1s spectrum in Figure 2D shows three peaks at 284.1 eV, 285.3 eV, and 287.1 eV, associated with double bonded sp²-carbon (-C=C-), single bonded sp³-carbon (-C-N/C-C-), and polar covalent (-C-O-) bonds, respectively^[37-39]. The dominant -C=C- signal suggests a carbon matrix, while the -C-N/C-C- peak reveals N-doping within the carbonaceous architecture, which enhances the conductivity of the material and offers supplementary sites for polysulfide trapping^[40-42]. The N 1s XPS spectrum in Figure 2E further validates the N-doping, revealing three peaks at 398.1 eV, 400.5 eV, and 402.5 eV, corresponding to pyridinic, pyrrolic, and graphitic species of nitrogen, respectively^[43,38]. These findings are corroborated by EA results, presented in Supplementary Table 1, which shows that the carbon and nitrogen contents are approximately 37.5 and 6.1 wt%, respectively. This confirms the successful formation of an N-C skeleton constituting the microspheres. The thermal stability of the prepared microspheres was evaluated using the thermogravimetric (TG) curve in an air atmosphere, as shown in Figure 2F, which reveals a continuous phase alteration of the selenide phase (Bi₂Se₃) to an oxide phase (Bi₂O₃), as temperature progresses. Furthermore, the N₂ adsorption-desorption analysis of 3D Bi₂Se₃@N-C microspheres was performed to measure the porosity of the microspheres [Supplementary Figure 4A and B]. The Brunauer-Emmett-Teller (BET) surface area of the 3D Bi₂Se₃@N-C microspheres was 3 m² g^-1, suggesting low porosity of the nanostructure. The low specific surface area of the nanostructure minimizes the consumption of electrolytes by the interlayer materials, enabling the implementation of a LSB that can operate with a low E/S ratio.

Figure 2. (A) XPS survey; (B) deconvoluted Bi 4f spectrum; (C) deconvoluted Se 3d spectrum; (D) deconvoluted C 1s spectrum; (E) deconvoluted N 1s spectrum; and (F) TG profile of 3D Bi₂Se₃@N-C microspheres in an air atmosphere. XPS: X-ray photoelectron spectroscopy, TG: thermogravimetric, 3D: three-dimensional.

Before evaluating the electrochemical properties of the assembled Li-S cells, the prepared multifunctional barrier and the pristine separator were subjected to physical analysis and integrity tests. Supplementary Figure 5A demonstrates a digital image of the coated separator, which appears uniform, continuous, and free from cracks. The coated separator was sliced into round disks (ϕ = 19 mm), as displayed in Supplementary Figure 5B. The thickness of the coated functional separator [Supplementary Figure 5C] and the pristine separator [Supplementary Figure 5D] was measured using a digital thickness meter, revealing a coating layer approximately 5 µm thick. Besides, the mechanical robustness of the modified separator was evaluated through folding and bending tests, as shown in Supplementary Figure 5E. Remarkably, the coated separator regained its original shape after these tests, demonstrating excellent mechanical robustness. The active loading mass of the coated separator was determined to be 5.2 mg [Supplementary Figure 5F], corresponding to a surface mass density of approximately 0.4 mg cm^-2. The FE-SEM micrograph of the uncoated Celgard 2400 separator, presented in Supplementary Figure 6A, reveals interconnected submicron-sized porous openings. These openings facilitate the uninterrupted transport of Li⁺ ions during redox reactions and enable efficient electrolyte percolation. Conversely, the FE-SEM micrograph of the functional polysulfide barrier coated with 3D Bi₂Se₃@N-C microsphere [Supplementary Figure 6B] displays a dense and uniform coating of the microspheres. Additionally, the thickness of the coating layer was observed to be approximately 5 µm, as confirmed by Supplementary Figure 6C, which aligns well with the values obtained from the digital thickness measurements. Furthermore, the coated separator was subjected to wettability tests, as shown in Supplementary Figure 7. As observed, a smaller contact angle [15.8°, Supplementary Figure 7A] for the coated separator compared to the pristine sample [39.7°, Supplementary Figure 7B] implies better wettability. Moreover, it was observed that the electrolyte absorption capability of the coated separator is approximately 63% higher than that of the pristine separator, as evident from the digital images presented in Supplementary Figure 7C-F, which clearly indicates a higher weight of the coated separator after soaking for a predetermined time in the electrolyte.

The as-prepared 3D Bi₂Se₃@N-C microspheres were engaged as a polysulfide barrier by coating them onto a commercial Celgard separator, positioned to face the regular sulfur cathode. The assembled CR2032 coin cells were initially subjected to cyclic voltammetry (CV) tests at a scan rate of 0.1 mV s^-1, and the first scans are presented in Figure 3A. For the cell incorporating the 3D Bi₂Se₃@N-C coated separator, the CV scan reveals two sharp and intense reduction peaks, labeled R1 and R2, centered at 2.31 V and 2.0 V, respectively. In contrast, the cell paired with a pristine separator exhibits broader and less intense reduction peaks located at 2.20 V and 1.96 V, respectively. Notably, the R1 peak corresponds to the reduction process to form high-order LiPSs (Li₂S₈ or Li₂S₆), which are further converted into middle-order polysulfides. The R2 peak signifies the subsequent reduction to form insoluble Li₂S species^[44,45]. During the reverse scan, two closely spaced oxidation peaks are observed for the 3D Bi₂Se₃@N-C coated separator, representing the reverse mechanism. These peaks correspond to the oxidation of Li₂S to Li₂S₆ or Li₂S₈ at 2.38 V and further to elemental sulfur at around 2.44 V, completing the redox process^[46,47]. Conversely, the cell with the pristine separator displays two unresolved oxidation peaks at significantly higher voltage values, 2.49 V and 2.52 V. It is noteworthy to mention that the cell utilizing the 3D Bi₂Se₃@N-C modified separator demonstrates the maximum redox current values, signifying enriched redox activities. This observation aligns with the individual peak voltage values reflected in Supplementary Figure 8A and respective voltage polarization values (the difference in voltage between the O and R2 peaks). Notably, the cell featuring the 3D Bi₂Se₃@N-C coated separator exhibits a significantly lower voltage polarization (∆V = 380 mV) compared to the cell paired with the pristine separator (∆V = 530 mV). Furthermore, the first three CV cycles for both separator configurations indicate coinciding curves, reflecting promising redox processes with high reversibility. However, the extent of reversibility and kinetics is superior for the cell employing 3D Bi₂Se₃@N-C [Figure 3B] interlayer compared to the pristine separator [Supplementary Figure 9].

Figure 3. (A) CV profiles of cells paired with coated/pristine separator at 0.1 mV s^-1; (B) first three CV scans for cell paired with 3D Bi₂Se₃@N-C coated separator; (C) initial Galvanostatic charge-discharge voltage profile comparison at 0.05 C; (D) charge-discharge voltage profiles at different C-rates with 3D Bi₂Se₃@N-C coated separator; (E) rate performance at various C-rates; and (F) capacity utilization values for 3D Bi₂Se₃@N-C coated separator. CV: Cyclic voltammetry; 3D: three-dimensional.

To validate the CV results, the initial Galvanostatic charge-discharge (GCD) characteristics were examined at 0.05 C, as shown in Figure 3C. The voltage profiles for both the separator configurations exhibit similar shapes, with two distinct reduction voltage regions and a sloppy charging plateau, reflecting the multistep electrochemical processes. This observation aligns well with CV results. However, differences in the extent of the redox plateaus suggest varying extents of redox reactions within the cells. Notably, the cell incorporating the 3D Bi₂Se₃@N-C-modified separator demonstrates the lengthiest redox plateau and the lowermost polarization (ΔE). The ΔE values, presented in Supplementary Figure 8B, reveal that the 3D Bi₂Se₃@N-C coated separator exhibits significantly lower polarization (ΔE = 215 mV) compared to the cell featuring the pristine separator (ΔE = 231 mV). To gain deeper insights, two variables, Q1 and Q2, along with their relative ratio (Q2/Q1), were also evaluated. Here, Q1 measures discharge capacity (with 419 mA h g^-1 as the theoretical value) corresponding to the first reduction plateau followed by the sloppy region. Likewise, Q2 corresponds to the discharge capacity (with 1,256 mA h g^-1 as the theoretical value) linked to the second reduction plateau and its sloping region, reflecting the efficiency of the conversion reaction of LiPS to Li₂S^[21]. A high relative ratio (Q2/Q1; theoretical value: 3) indicates the conversion efficiency of the designed sample^[20]. As presented in Supplementary Figure 8B, the cell featuring the 3D Bi₂Se₃@N-C coated separator reveals a higher ratio (2.65) in contrast to the pristine separator (2.05). This finding highlights the enhanced electrocatalytic activity of the 3D Bi₂Se₃@N-C coated separator toward LiPS species. Consequently, a high initial discharge capacity of 1,132 mA h g^-1 was delivered by the cell incorporating the 3D Bi₂Se₃@N-C coated separator, equivalent to 67.6% of the theoretical value. In contrast, the cell with the pristine separator achieves only 681 mA h g^-1 or 40.7% of the theoretical limit. The significantly higher discharge capacity value of the cell utilizing 3D Bi₂Se₃@N-C coated separator underscores its ability to facilitate kinetically favored redox reactions. This performance is attributed to the efficient trapping and catalytic conversion of polysulfide species, leading to enhanced active material utilization.

Rate capability tests were conducted at various current values ranging from 0.05-2.0 C, as shown in Figure 3D and E. The GCD curves at different C-rates for cell paired with 3D Bi₂Se₃@N-C coated separator exhibit well-defined charge-discharge redox plateaus even at 2.0 C-rate, highlighting superior performance compared to the cell with the pristine separator [Supplementary Figure 10]. For the 3D Bi₂Se₃@N-C modified separator, initial capacities of 1134 (0.05 C), 936 (0.1 C), 780 (0.2 C), 650 (0.3 C), 545 (0.5 C), 465 (0.7 C), 405 (1.0 C), and 297 (2.0 C) mA h g^-1 were achieved. Contrariwise, the cell with the uncoated separator delivered significantly lower discharge capacities of 681 mA h g^-1, 495 mA h g^-1, 386 mA h g^-1, 349 mA h g^-1, 330 mA h g^-1, 321 mA h g^-1, 81 mA h g^-1, and 49 mA h g^-1, at the same C-rates, as illustrated in Figure 3E. Furthermore, when the current was reversed to 0.1 C, the cell with the 3D Bi₂Se₃@N-C coated separator retained a stable discharge capacity of 694 mA h g^-1, significantly higher than the 401 mA h g^-1 observed for the pristine separator, indicating the excellent reversibility of the redox processes. The proportion of material utilized in redox reactions was measured through capacity utilization (%) parameter, as presented in Figure 3F and Supplementary Figure 11. As expected, the cell paired with a 3D Bi₂Se₃@N-C coated separator demonstrated the maximum capacity utilization values at all C-rates compared to the cell with the pristine separator. This observation is well synchronized with the CV results, which indicated lower polarization and superior electrocatalytic conversion efficiency.

Stable cycling performance is a critical factor for the commercial viability of the prepared nanostructure. Figure 4A-C illustrates the cycling performance tests of the 3D Bi₂Se₃@N-C coated separator compared to the pristine separator. At a 0.1 C-rate [Figure 4A], 3D Bi₂Se₃@N-C-modified separator exhibits an early discharge capacity of 1,039 mA h g^-1 and retains 83% (863 mA h g^-1) after 100 continuous charge-discharge cycles, corresponding to an average capacity decay rate of just 0.16% per cycle. Moreover, the high Columbic efficiency (~99%) indicates highly reversible redox processes. Conversely, the pristine separator shows a significantly lower initial capacity of 510 mA h g^-1 (30% of theoretical) and retains only 379 mA h g^-1 (74% retention) after 100 cycles, corresponding to a higher capacity decay rate of 0.27% per cycle. Similar trends are observed at higher C-rates. For example, at 0.5 C [Figure 4B], a 3D Bi₂Se₃@N-C-coated separator achieves a discharge capacity of 440 mA h g^-1 after 200 cycles, compared to just 229 mA h g^-1 for the pristine separator cell. At an even higher C-rate of 2.0 C [Figure 4C], the discharge capacities at the end of the 500^th cycle are 219 mA h g^-1 for the 3D Bi₂Se₃@N-C coated separator and only 62 mA h g^-1 for the pristine separator. The excellent capacity retention, even after prolonged cycling, underscores the structural merits of the prepared sample. To further access practical viability, cells were assembled with high S-loading electrodes (2.5 and 3.5 mg cm^-2). The cycling performance results at 0.3 C-rate are presented in Figure 4D. Even at high sulfur loadings, the cells demonstrate stable cycling up to 200 cycles. For instance, with a sulfur loading of 2.5 mg cm^-2, the cell achieves an initial discharge capacity of 579 mA h g^-1 and retains 58% (337 mA h g^-1) of its capacity after 200 cycles. Likewise, at a sulfur loading of 3.5 mg cm^-2, the cell retains 57% of its initial capacity, further validating the applicability of the prepared nanostructure for practical applications. These results clearly validate that the manifestation of Bi₂Se₃ nanocrystals significantly enhances the catalytic transformation of anchored polysulfide species, improving active material consumption. Additionally, the N-doping within the nanostructures expedites rapid charge kinetics, resulting in faster redox activity and hence, improved electrochemical performance. The detailed working mechanism of the 3D Bi₂Se₃@N-C coated separator as a polysulfide barrier is schematically demonstrated in Figure 4E. Furthermore, the electrochemical performance results obtained in this study are summarized in Supplementary Table 2 and benchmarked against previously published works. Furthermore, high gravimetric energy density for Li-S cells utilizing different sulfur loadings and 3D Bi₂Se₃@N-C interlayers [Supplementary Table 3] again indicates the practical application potential of this strategy.

Figure 4. Cycling tests of cells featuring coated/pristine separators at (A) 0.1 C-rate; (B) 0.5 C-rate; (C) 2.0 C-rate; (D) cycling test of 3D Bi₂Se₃@N-C at 0.3 C coupled with high loading electrodes; and (E) detailed working mechanism of the 3D Bi₂Se₃@N-C modified separator. 3D: Three-dimensional.

The enhanced redox kinetics of the 3D Bi₂Se₃@N-C coated separator were analyzed by determining the diffusion coefficient ($$ D_{L i^+} $$) values. CV measurements were performed on cells with 3D Bi₂Se₃@N-C coated and pristine separator at different scan rates (0.05 mV s^-1 to 1.0 mV s^-1), as revealed in Figure 5A and B. The $$ D_{L i^+} $$ values were calculated by plotting the I_p (peak current) against the v^0.5 (scan rate square root), based on the Randles-Sevick equation^[48]:

Figure 5. $$ D_{L i^+} $$ coefficient and EIS plots for coated/pristine separators: (A and B) CV curves at different scan rates (0.05-1.0 mV s^-1); (C and D) redox peak intensity vs. square root of scan rate; (E) EIS plots of newly assembled cell; and (F) after 300 cycles at 2.0 C. CV: Cyclic voltammetry; EIS: electrochemical impedance spectroscopy.

where I_p represents the redox peak current, n is the number of electrons involved, A is the electrode surface area (cm²), C_Li is the concentration (mol L^-1) of Li-ion, and v is the voltage scan rate (Vs^-1). The CV curves for cell featuring the 3D Bi₂Se₃@N-C coated separator [Figure 5A] exhibit two distinct reduction peaks (R1/R2) and one oxidation peak (O), even at high voltage scan rates, indicating enhanced redox reaction kinetics. In contrast, the cell with the pristine separator [Figure 5B] displays poorly resolved reduction peaks at high scan rates, reflecting sluggish reaction kinetics. Additionally, the more symmetric CV curves and high current intensity observed for the 3D Bi₂Se₃@N-C coated separator highlight faster Li-ion diffusion. The $$ D_{L i^+} $$ values for the Li-S cell with a 3D Bi₂Se₃@N-C coated separator were approximately 1.5 × 10^-7 cm² s^-1, 0.3 × 10^-7 cm² s^-1, and 0.6 × 10^-7 cm² s^-1 for peaks O, R1, and R2, respectively [Figure 5C]. In comparison, the pristine separator cell showed significantly lower $$ D_{L i^+} $$ values of 2.3 × 10^-8 cm² s^-1, 0.4 × 10^-8 cm² s^-1, and 1.6 × 10^-8 cm² s^-1 for peaks O, R1, and R2, respectively [Figure 5D]. These higher $$ D_{L i^+} $$ values underscore the structural merits of the 3D Bi₂Se₃@N-C microspheres, which effectively trap and catalytically convert polysulfide while the highly conductive N-C framework facilitates electron transfer by lowering the charge-transfer resistance (R_ct), thereby boosting the electrochemical properties. To further evaluate the interfacial surface processes taking place at electrode-separator-electrolyte boundaries, Nyquist plots were plotted via electrochemical impedance spectroscopy (EIS) both before [Figure 5E] and after cycling [Figure 5F]. The data was fitted using an equivalent circuit model [Supplementary Figure 12]. Before cycling, the impedance measurements at open-circuit voltage revealed negligible solution resistance (R_s) values of 3-5 Ω for both cells, indicating a similar electrolyte-electrode interface. However, the R_ct for the cell with a 3D Bi₂Se₃@N-C coated separator was significantly lower (~45 Ω) compared to the uncoated separator (~80 Ω), suggesting enhanced Li-ion kinetics. After 500 cycles at a 2.0 C-rate, the R_ct for the 3D Bi₂Se₃@N-C coated separator cell decreases significantly to ~17 Ω, demonstrating enhanced kinetics over prolonged cycling. Conversely, the uncoated separator cell demonstrates a marginal reduction in R_ct to ~66 Ω, indicating poor diffusion kinetics primarily due to uncontrolled polysulfide migration and the formation of inactive materials that hinder reaction efficiency. This behavior is further supported by the presence of two distinct semicircles in the EIS curve of the pristine separator cell after cycling, with the high-frequency semicircle attributed to the insulating layer of solid Li₂S₂/Li₂S^[49]. In contrast, the 3D Bi₂Se₃@N-C coated separator suppressed polysulfide diffusion, as evidenced by a single semicircle in the EIS plot. These findings validate the structural superiority of the 3D Bi₂Se₃@N-C microspheres, which provide outstanding charge transfer features, diminish cell resistance, effectively trap and catalyze polysulfide molecules, and enhance sulfur consumption during prolonged cycling. The efficient trapping of the polysulfide was also confirmed through perceivable polysulfide adsorption tests [Supplementary Figure 13]. A continuous fading of solution color containing 3D Bi₂Se₃@N-C microspheres from yellow to translucent (T = 1 h), signifying effective anchoring of the LiPSs. To further validate this, the cell comprising 3D Bi₂Se₃@N-C microspheres as the coated interlayer was disassembled carefully inside the glove box after cycling at 2.0 C and the interlayer was analyzed using FE-SEM. The micrographs shown in Supplementary Figure 14A and B demonstrate that the spherical morphology of the cycled separator remained intact even after prolonged cycling (500 cycles). Furthermore, no evidence of polysulfide agglomeration was found on the spherical surface, indicating the effective electrocatalytic conversion of the trapped LiPS species.

Overall, the enhancement in electrochemical properties of cells retaining 3D Bi₂Se₃@N-C microspheres as a multifunctional polysulfide barrier firmly validates the structural supremacy of the prepared nanostructures that resulted in improved redox reaction kinetics via fast charge transfer characteristics, suppress polysulfide diffusion, and enhance active sulfur participation during electrochemical processes.

CONCLUSIONS

In conclusion, this work presents a comprehensive study on the synthesis of 3D Bi₂Se₃@N-C microspheres and their application as a multifunctional polysulfide barrier for improved LSB properties such as realistic rate capability and steady cycling performance across a range of C-rates, from low (0.1 C and 0.5 C) to high (2.0 C). The incorporation of Bi₂Se₃ nanocrystals ensures efficient anchoring and transformation of polysulfide, enhancing sulfur participation in electrochemical reactions. Furthermore, N-doping within the carbon framework expedites rapid charge transferal, accelerating redox activity. As a result, the cell equipped with a 3D Bi₂Se₃@N-C microsphere coated separator delivered remarkable performance, including high rate properties (297 mA h g^-1 at 2.0 C-rate) and prolonged stable cycling performance at different C-rates (863 mA h g^-1 after 100 cycles at 0.1 C, 440 mA h g^-1 after 200 cycles at 0.5 C, and 219 mA h g^-1 after 500 cycles at 2.0 C). Additionally, even at a loading of 3.5 mg cm^-2, the cell maintained satisfactory cycling performance, retaining 57% of its capacity after 200 charge-discharge cycles.