Polar Mesoporous Carbon-Coated Hybrid Separator Enabling Reversible Anchoring of Polysulfides in Li-S Cells
Article information
Abstract
Despite tremendous efforts to develop lithium-sulfur (Li-S) batteries as next-generation energy-storage systems, their commercialization has been hindered by critical challenges such as the polysulfide shuttle effect and poor cycling stability. In this study, polar mesoporous carbon (p-MC) featuring abundant polar functional groups and high porosity is synthesized via pyrolysis at 400°C, and a functional hybrid separator is then fabricated by uniformly coating p-MC onto the surface of a polyethylene (PE) membrane to mitigate polysulfide diffusion toward the lithium anode. A Li-S cell incorporating p-MC-coated PE separator (p-MC/PE) is systematically compared with two control cells: one employing a conventional PE separator and the other utilizing a nonpolar mesoporous carbon-coated separator (n-MC/PE). Remarkably, the p-MC/PE cell exhibits twice the capacity retention, enhanced rate capability, and reduced self-discharge after 300 cycles compared with the control PE cell. In addition, the p-MC/PE cell retains 83.1% of its initial capacity after 300 cycles, significantly outperforming the n-MC/PE cell. The exceptional performance of the p-MC/PE separator is attributed to the strong polar-polar interactions between the lithium polysulfides and functional groups on the p-MC surface. The synergy between polar functional groups and well-interconnected mesopores facilitates the capture of polysulfides through both chemical interactions and physical confinement.
INTRODUCTION
Lithium-sulfur (Li-S) batteries have emerged as leading candidates in next-generation energy-storage systems because of their exceptional theoretical capacity, cost efficiency, and environmental advantages [14]. With an impressive theoretical specific energy of approximately 2600 Wh kg–1, they significantly surpass the energy density of conventional lithium-ion batteries, making them ideal for a wide array of applications ranging from portable electronics to electric vehicles and large-scale energy-storage solutions [5,6]. Moreover, sulfur, which is a key component of Li-S batteries, is both abundantly available and inexpensive; it reduces production costs and enhances the economic viability of these batteries for widespread application [7‒9].
The low toxicity and environmental benefits of sulfur further bolster the sustainability of Li-S batteries, rendering them promising next-generation electrochemical energy-storage systems [10‒12]. Despite their potential, Li-S batteries are plagued by persistent challenges such as limited cycle life and lithium dendrite formation. The primary cause of sulfur cathode degradation during repeated cycling is the diffusion of lithium polysulfides (LiPSs) toward the Li anode, which triggers irreversible reactions with Li, leading to capacity reduction and performance degradation [13‒15].
Researchers have explored multiple strategies to address these issues, including the development of advanced sulfur host materials, interlayers, and functional separators [16‒19]. Mesoporous carbon hosts are particularly promising because of their large surface area and porosity, which facilitate efficient electrode reactions and accommodate volume expansion during cycling [20‒22]. However, as the sulfur loading increases, the effectiveness of these carbon host materials diminishes, and their complex, time-consuming synthesis processes limit their practicability and scalability. Although thick interlayers can extend the cycle life of Li-S cells, they significantly reduce the overall energy density, limiting their commercial viability [23‒26].
In this regard, an effective strategy is the modification of the separator, which can suppress polysulfide migration and mitigate the shuttle effect without compromising energy density. Various porous carbon materials, such as mesoporous carbon, carbon nanotubes (CNTs), and graphene, have been widely studied for separator modification [27‒31]. Carbon materials with high electrical conductivity generally have a weak chemical affinity for polysulfides owing to their nonpolar surfaces. Although they can physically trap soluble polysulfides, they cannot chemically or stably bind polar polysulfides during cycling [32‒34]. To overcome this, researchers have doped carbon materials with heteroatoms, such as sulfur and nitrogen, to create surface polarity, thereby improving their chemical affinity for polysulfides [35‒37]. In addition, highly polar metal oxides have been integrated with nonpolar carbons on separators, inducing a synergistic effect to capture soluble polysulfides more efficiently [38,39]. However, these approaches typically involve complex and expensive additional processes. Consequently, there is an increasing demand for more efficient and scalable solutions that can suppress polysulfide migration while maintaining high sulfur utilization and overall energy density.
In this study, we developed a novel carbon material as a viable solution for preventing polysulfide migration via two key mechanisms: physical confinement and chemical interactions within the cathode region. This material has a large surface area (>1000 m2 g–1), high porosity (>1.8 cm3 g–1), and abundant polar functional groups. Its unique structure allows the strong chemical anchoring of polysulfides, significantly increasing its polysulfide absorption capacity. We first produced a silica/carbon composite at a low temperature of 400°C using sucrose as the carbon precursor and silica nanoparticles as a template, followed by silica etching with hydrofluoric acid to obtain polar mesoporous carbon. As the pyrolysis temperature increases, carbon-rich graphitic structures with a honeycomb pattern are typically formed, reducing the O/C and H/C atomic ratios and increasing the mechanical strength of the material [40‒43]. In this study, we selected a relatively low temperature of 400°C, which allowed the material to retain sufficient mechanical strength to preserve its mesoporous structure, while maintaining an ample amount of oxygen-containing functional groups. This innovative material, which is capable of both physical absorption and chemical adsorption, was integrated onto the surface of a polyethylene (PE) separator. In addition, because it can be synthesized at low temperatures and does not require additional processes such as heteroatom doping or mixing of polar metal oxides to generate polar sites, it is both cost-effective and easy to manufacture. The Li-S cell equipped with this functional separator exhibited exceptional results with regard to cycle life and rate capability. This approach proved to be highly effective for preventing polysulfide diffusion and mitigating intrinsic self-discharge in Li-S cells.
EXPERIMENTAL
Synthesis and characterization of as-prepared material
Polar mesoporous carbon (p-MC) was prepared via a nanotemplating approach using sucrose (DAEJUNG, Korea) as the carbon precursor. The sucrose solution (6.25 g in 100 mL of triply distilled water) was acidified with 1 mL of 85% phosphoric acid (SAMCHUN, Korea). Subsequently, 21 mL of a 10-nm silica colloid (30% in water, Thermo Fisher Scientific, USA) was added. The mixture was dried at 100°C for 6 h, caramelized at 160°C for 6 h, and carbonized at 400°C for 6 h under argon flow. The silica template was removed via etching with 10% hydrofluoric acid (DAEJUNG, Korea), followed by extensive water washing and vacuum drying at 130°C. The resulting p-MC incorporated phosphorus moieties, which were derived from the phosphoric acid catalyst used during sucrose dehydration.
The porous structure of p-MC was characterized via nitrogen adsorption‒desorption isotherms obtained using a Tristar II 3020 analyzer (Micromeritics, USA). The specific surface areas and pore volumes were calculated using the Brunauer‒Emmett‒Teller (BET) and Barrett‒Joyner‒Halenda (BJH) methods, respectively. The morphology of p-MC was examined using field-emission scanning electron microscopy (FE-SEM, Apreo S HiVac, Thermo Fisher Scientific Czech). The surface functional groups and their interactions with the polysulfides were analyzed using X-ray photoelectron spectroscopy (XPS; NEXSA G2, Thermo Fisher Scientific, USA). The elemental composition (C, H, O, and P) of p-MC was determined via elemental analysis (FLASH 2000 Series, Thermo Fisher Scientific, USA) and inductively coupled plasma optical emission spectroscopy (ICP-OES, 5100 SVDV, Agilent Technologies, USA). The crystallinity of the carbon matrix in the porous carbon materials was investigated via X-ray diffraction (XRD) using an X-ray diffractometer (EMPYREAN, Malvern Panalytical BV, UK). The interlayer spacing and crystallite size along the c-axis (Lc) were determined from the (002) diffraction peak.
Fabrication and electrochemical evaluation of hybrid separator
A p-MC-coated PE-based separator (p-MC/PE) was fabricated by preparing a slurry on a 0.5 g solids basis. Typically, 0.1 g of poly(vinylidene fluoride) (PVDF) was dissolved in 7 mL of N-methyl-2-pyrrolidone (NMP; JUNSEI, Japan), and then 0.4 g of p-MC powder was added, corresponding to a p-MC:PVDF weight ratio of 8:2. The resulting slurry was evenly coated onto a 12 μm PE separator (SK Information Electronic Technology, Korea) using a doctor blade to achieve a coating loading of 2.5 mg cm–2. The coated separator was then dried in a vacuum oven at 25°C for 24 h, yielding a coating thickness of ~3 μm. A sulfur-based cathode, comprising sulfur, carbon nanotubes (CNTs; Betterial, Korea), and PVDF in a 60:20:5 weight ratio, was prepared and coated onto a carbon-coated aluminum foil current collector (Dongwon Systems, Korea), which was then dried at 40°C for 24 h to yield a sulfur loading density of 4 mg cm–2.
A coin-type Li-S cell consisting of a sulfur cathode, a lithium anode, and an organic electrolyte was assembled with the p-MC/PE separator in an argon-filled glovebox. For comparison, a control cell was fabricated with the PE separator instead of the p-MC/PE separator. The electrolyte (ENCHEM, Korea) consisted of 1.0 M lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) and 0.3 M lithium nitrate (LiNO3) in a 3:1:1 (v/v/v) mixture of 1,2-dimethoxyethane (DME), diglyme (DGM), and 1,3-dioxolane (DOL). The first cycling of the cells was performed at 0.1 C (1 C = 1675 mA g–1) between 1.8 and 2.8 V. Subsequently, the cycling stability and rate capability were evaluated at 25°C. Cyclic voltammetry (CV) measurements were conducted using a potentiostat (VMP3, BioLogic, France) to assess the cell kinetics. The contact angles of the PE and p-MC/PE separators with the electrolyte were measured to qualitatively assess their wettability.
The morphologies of the p-MC/PE separator and Li metal anode before and after cycling were examined using FE-SEM along with energy-dispersive X-ray spectroscopy (EDS). For a direct and fair comparison, nonpolar mesoporous carbon (n-MC) with a similar microstructure was synthesized at 900°C, and an n-MC/PE separator was fabricated and evaluated following the same procedures.
RESULTS AND DISCUSSION
Materials synthesis and characterization
Nonpolar carbons are commonly used as coating layers for fabricating separators that can mitigate polysulfide diffusion [27‒31]. However, owing to their low physical adsorption capacity, they are ineffective for stable adsorption of polar polysulfides [32‒34]. To overcome this limitation, nonpolar carbons have been doped with heteroatoms (e.g., N, S) and combined with various polar metal oxides [38‒39].
In this study, we used a polar mesoporous carbonaceous material with numerous polar functional groups as a novel separator. We hypothesized that this polar mesoporous carbonaceous material on the PE separator would not only function as a reservoir for polysulfides because of its high porosity but also anchor polysulfides securely via strong interactions with its surface polar functional groups (Fig. 1). To test this, a polar mesoporous carbon material (p-MC) with a high pore volume and large surface area was synthesized using a nano-replication technique, followed by pyrolysis under an argon atmosphere. The temperature was carefully selected to ensure that the carbon walls had sufficient mechanical strength to maintain the mesoporous structure, while preserving ample surface polar functional groups. Higher temperatures promote the development of carbon-rich graphitic layers and reduce both the O/C and H/C atomic ratios [40‒43]; however, at temperatures below 350°C, the mechanical strength of the carbon walls is insufficient, leading to structural collapse and a severe reduction in the surface area and pore volume (Fig. S1). Notably, the synthesis of high-quality mesoporous carbon at temperatures below 350°C is extremely challenging. In this study, we selected 400°C as the optimal pyrolysis temperature to balance mechanical strength and functional-group retention, which resulted in a stable mesoporous carbon structure with desirable surface characteristics.
Schematic of the Li-S battery configuration featuring the novel polar mesoporous carbon-coated separator (p-MC/PE). The p-MC/PE separator, with polar functional groups, effectively confines lithium polysulfides within the cathode region by capturing them through chemical interactions and physical absorption.
As shown in Fig. 2a, the as-synthesized p-MC has a highly porous architecture and is characterized by well-defined spherical mesopores directly replicated from the silica templates. Most of the mesopores appeared to be randomly distributed and separated by carbon walls. However, in some areas, mesopores merged into larger structures because of the aggregation of colloidal silica during the synthesis process. The carbon matrix of p-MC exhibited a broad (002) diffraction peak in its XRD profile, implying a highly disordered structure (Fig. 2b).
Characterization of the synthesized polar mesoporous carbon (p-MC). (a) SEM images of p-MC showing a highly porous structure with well-defined spherical mesopores. (b) XRD pattern of the p-MC. (c) N2 adsorption-desorption isotherms of p-MC, showing a BET surface area of 1069.47 m2 g–1 and pore volume of 1.84 cm3 g–1, with the corresponding pore size distribution curve (inset) derived using the BJH method. XPS spectra of the p-MC material showing (d) C 1s with deconvoluted peaks corresponding to C–C, C–OH, C=O, and COOH groups, (e) O 1s with peaks for C–OH and C=O (carbonyl, carboxyl), and (f) P 2p with peaks for pyrophosphate and metaphosphate groups.
Fig. 2c presents the N2 adsorption‒desorption isotherms of p-MC with the corresponding pore size distribution (inset of fig. 2c). The as-synthesized p-MC exhibited a marked increase in the adsorption curve in the relative pressure (p/p0) range of 0.80 to 0.95, which is attributed to nitrogen condensation within the mesopores, accompanied by a broad hysteresis loop—a characteristic feature of mesoporous materials [44]. The pore size distribution of p-MC, centered at 16 nm, reflects the successful replication of the silica templates. In addition, p-MC exhibited a large BET surface area of 1070 m2 g–1 and a substantial pore volume of 1.84 cm3 g–1, as determined using the BJH method. These properties indicate that p-MC provides numerous sites for polysulfide adsorption and can store a large number of polysulfides within its abundant pores, potentially capturing them as they traverse the separator.
Elemental analysis of p-MC was performed to quantitatively assess the remaining heteroatoms such as oxygen and hydrogen. The H/C and O/C atomic ratios of p-MC were determined to be 0.44 and 0.17, respectively, suggesting the presence of numerous polar functional groups. Typically, carbonaceous materials pyrolyzed at 400°C exhibit H/C ratios between 0.4 and 0.7 and O/C ratios ranging from 0.15 to 0.25 [40‒43]. The relatively low H/C and O/C ratios of p-MC imply that the carbonization process was highly efficient under the synthesis conditions employed in this study. Moreover, ICP-OES analysis indicated that the residual phosphorus content in the p-MC was 3.3%, corresponding to a P/C ratio of 0.015. SEM/EDS analysis revealed that O and P were uniformly distributed within the p-MC particles (Fig. S2), indicating a homogeneous composition throughout the particle structure. Residual heteroatoms, such as H, O, and P, were located at the edges of the carbon layers with a honeycomb structure, forming polar functional groups.
XPS was performed to determine the specific functional groups associated with these elements. The C 1s XPS profile in Fig. 2d exhibits four distinct peaks at 284.9, 286.4, 287.0, and 288.0 eV. The peak at 284.9 eV corresponds to the covalent C–C bonds in the graphitic honeycomb structures, whereas those at 286.4, 287.0, and 288.0 eV are assigned to the C–OH, C=O, and COOH groups, respectively, which are present at the edges of the graphene layers (Fig. S3) [43,45,46]. The O 1s XPS profile in Fig. 2e exhibits two peaks centered at 531.7 and 533.3 eV, corresponding to the O=C and C–OH bonds, respectively [43,45,46]. The P 2p XPS profile in Fig. 2f exhibits two 2p3/2 peaks centered at 132.7 and 134.9 eV, revealing the presence of the pyrophosphate and metaphosphate groups, respectively [46,47]. These abundant polar functional groups—particularly hydroxyl and carbonyl groups—exhibit strong chemical interactions with polysulfides, effectively anchoring them and thereby enhancing the cycle life of batteries [48‒50].
A p-MC/PE functional separator was prepared by coating a mixture of p-MC and PVDF onto a PE separator. In the surface SEM image of the p-MC coating layer on the p-MC/PE separator, the p-MC particles appeared to be uniformly distributed across the surface of the PE separator (Fig. 3a). As is typical of slurry coatings with microparticles, interstitial voids were clearly visible within the coating layer. These voids are filled with electrolyte solutions, facilitating seamless ion transport, similar to the function of ceramic separators used in current Li-ion batteries. In the cross-sectional image of the p-MC/PE separator, the p-MC coating layer appeared to be uniformly cast with a thickness of approximately 3 μm and maintained close contact with the PE substrate (Fig. 3b). The p-MC/PE separator exhibited a significantly lower contact angle of 12.87° (Fig. 3d) compared with the PE separator (28.43°, Fig. 3c) when the electrolyte solution was dropped onto the surface. This suggests that the p-MC/PE separator has superior wettability toward the electrolyte solution, promoting a more uniform distribution of the electrolyte within the cell.
(a) Surface and (b) cross-sectional SEM images of the p-MC/PE separator, showing a coating thickness of 3 µm. Contact-angle measurements of (c) the PE separator and (d) the p-MC/PE separator, with an electrolyte droplet demonstrating the improved wettability of the p-MC/PE separator toward the electrolyte.
Electrochemical evaluation of p-MC/PE hybrid separator
The electrochemical performance of the p-MC/PE separator was evaluated in coin-type Li-S cells consisting of a sulfur cathode, a Li metal anode, and an organic electrolyte. For comparison, a control cell with a PE separator was fabricated following the same procedure. After cell assembly, the Li-S cells were discharged and charged at a low current density of 0.1 C. Both cells exhibited typical charge/discharge curves for Li-S batteries (Fig. 4a). The control cell with the PE separator exhibited a discharge capacity of 1030 mAh g–1, while that with the p-MC/PE separator (p-MC/PE cell) exhibited a discharge capacity of 1137 mAh g–1, corresponding to an increase of 107 mAh g–1. When solid sulfur (S8) is completely discharged to Li2S, the oxidation state of sulfur decreases from 0 to −2. Accordingly, the average oxidation state of sulfur in the PE cell at the end of the discharge process was calculated to be −1.23, corresponding to an average polysulfide chain length (x) of 1.626 in Sx2–. In contrast, the average oxidation state of sulfur in the p-MC/PE cell reached −1.36, corresponding to a shorter average chain length (x) of 1.471 in Sx2–. This indicated that the p-MC/PE separator entrapped the migrating polysulfides, preventing them from diffusing toward the lithium anode and confining them to the cathode. This allowed further reduction of polysulfides, leading to deeper reduction of the active sulfur species and significantly improving sulfur utilization.
lectrochemical performance comparison of Li-S cells with the pristine PE and p-MC/PE separators. (a) First charge/discharge potential profiles at 0.1 C for the pristine PE and p-MC/PE cells. (b) Cycling stability and coulombic efficiency of the pristine PE and p-MC/PE cells at 0.25 C over 300 cycles. (c) Voltage profiles of the pristine PE cell at different cycles (1st, 100th, 200th, 300th) showing significant voltage drop with cycling. (d) Voltage profiles of the p-MC/PE cell at the same cycles, showing stable performance with minimal voltage drop. (e) First-plateau discharge capacity as a function of cycle number, comparing the faster decline for the pristine PE cell with the more stable p-MC/PE cell.
Fig. 4b illustrates the cycling behavior of the two cells tested at 0.25 C after the initial cycle. The cell with the PE separator delivered a discharge capacity of 373 mAh g–1 after 300 cycles, corresponding to capacity retention of 47.6% and an average Coulombic efficiency of 98.1%. Notably, the p-MC/PE cell achieved a notably higher capacity of 722 mAh g–1 after 300 cycles—1.93 times higher than that of the pristine PE cell—along with an improved capacity retention of 83.1% and an average Coulombic efficiency exceeding 99.2%. These results indicate that the p-MC/PE separator markedly reduces the active-material loss and improves the cycling durability of the Li-S cell by confining the diffusing polysulfides. Furthermore, these findings suggest that the p-MC/PE separator is a reliable solution for ensuring the long-term reliability of Li-S cells.
The cycling behaviors of the two Li-S cells were thoroughly examined at the 1st, 100th, 200th, and 300th cycles (Fig. 4c and 4d). In the case of the control cell, considerable capacity fading was observed before 100 cycles. During cycling, the mid-voltage of the charging curves increased from 2.285 to 2.319 V. A voltage spike occurred at the beginning of the charging curve, indicating that a higher overpotential is needed to oxidize the solid discharge products during charging. The highly reactive polysulfides degrade the solid–electrolyte interphase (SEI) layer on the Li anode, leading to continuous decomposition of the electrolyte on the Li surface, which ultimately results in electrolyte depletion. When the electrolyte becomes insufficient, the oxidation of solid discharge products becomes more challenging as the chemical reactions between polysulfides, which are coupled with the electrode reactions, slow down. In contrast, the p-MC/PE cell maintained steady voltage profiles over 300 cycles, with no significant changes in the capacity or mid-voltage. This indicates that the p-MC/PE separator minimizes the side reactions at the Li metal surface, allowing the Li-S cell to operate reliably over extended periods of cycling. In Li-S cells, active-material loss is typically associated with irreversible chemical reactions between the soluble polysulfides and lithium, which result in capacity reduction during repeated cycles. The first discharge plateau is used as an indicator of the available active material [51,52]. For the control cell, the first-plateau discharge capacity decreased sharply during cycling (Fig. 4e). In contrast, the p-MC/PE separator decelerated the decline in the first-plateau discharge capacity (Fig. 4e). These results indicate that the p-MC/PE separator successfully confined the active materials within the cathode area, reducing the active-material loss.
Rate testing was sequentially performed at discharge current densities of 0.2, 0.5, 1.0, and 2.0 C (Fig. 5a). The control cell exhibited a significant decrease in the specific capacity from 914 mAh g–1 at 0.2 C to 285 mAh g–1 at 1 C and further to 188 mAh g–1 at 2 C. In comparison, the p-MC/PE cell consistently exhibited better performance, with a specific capacity of 920 mAh g–1 at 0.2 C, 747 mAh g–1 at 1 C, and 612 mAh g–1 at 2 C. Compared with the specific capacity obtained at 0.2 C, the p-MC/PE cell retained a remarkable 66.5% of its initial capacity at 2.0 C, in stark contrast to the mere 20.6% capacity retention recorded for the control cell. Fig. 5b and 5c present the voltage profiles of the control and p-MC/PE cells, respectively, at various current densities. For the control cell, the discharge mid-voltage exhibited a pronounced reduction from 2.08 to 1.95 V as the current density increased from 0.2 C to 1 C, whereas for the p-MC/PE cell, the discharge mid-voltage decreased marginally from 2.08 to 2.02 V. In the control cell, the polysulfides are broadly distributed in the electrolyte and spread from the cathode to the anode. Consequently, when an electrode reaction occurs at the cathode, the long transport distance of the active species increases in the mass-transfer overpotential. In contrast, the well-confined polysulfides within the cathode region of the p-MC/PE cell shortened the transport distance and accelerated polysulfide reduction, thus explaining the observed high mid-voltage and superior discharge capacity at a given C-rate.
Rate-capability comparison between Li-S cells with pristine PE and p-MC/PE separators. (a) Discharge capacities of the pristine PE and p-MC/PE cells at various current densities (0.2 C, 0.5 C, 1 C, and 2 C) over 50 cycles. (b) Voltage profiles of the pristine PE cell at different current densities (0.2 C, 0.5 C, 1 C, and 2 C). (c) Voltage profiles of the p-MC/PE cell at the same current densities, demonstrating more stable discharge behavior. (d, e) CV curves of (d) the pristine PE and (e) p-MC/PE cells over the first 10 cycles. (f) Oxidation capacity vs. cutoff voltage for pristine PE and p-MC/PE cells at the 1st and 10th cycles, indicating better kinetics in the p-MC/PE cell.
Furthermore, the ratio of the second-plateau capacity to the first-plateau capacity was thoroughly investigated with an increase in the current density from 0.2 C to 1 C. For the control cell, this ratio decreased significantly from 1.78 to 1.19, whereas for the p-MC/PE cell, it decreased marginally from 2.05 to 1.75. The length of the second plateau is typically affected by factors such as the location and distribution of polysulfides, the type of carbon material, and the electrolyte-to-sulfur (E/S) ratio. The second plateau is associated with the reduction of mid-chain polysulfides; for rapid reduction to occur, chemical reactions must occur rapidly between the polysulfides—particularly those coupled with active species such as S3- and S4- [53,54]. In the control cell, the spreading of polysulfides from the cathode to the lithium anode surface hindered the rapid occurrence of chemical reactions between them, significantly reducing the length of the second plateau at high rates. In contrast, in the p-MC/PE cell, most of the polysulfides remained confined within the cathode region, enabling the occurrence of faster chemical reactions, which resulted in a limited reduction in the length of the second plateau. This confinement led to a high ratio of 1.75 at 1 C for the p-MC/PE cell compared with the lower ratio of 1.19 for the control cell.
The electrochemical behavior of the Li-S cells was further evaluated via CV to clarify the impact of the functional separator on the cell performance. The CV profiles of the control and p-MC/PE cells over multiple cycles were compared (Fig. 5d and 5e, respectively). The control cell exhibited appreciable changes in the CV curves over repeated cycles. Specifically, pronounced changes were observed with regard to the shape, peak potential, and peak current of the redox waves. In contrast, the p-MC/PE cell exhibited remarkable electrochemical reversibility, as evidenced by the consistent shapes and positions of the redox peaks throughout cycling. This was due to the good confinement of the polysulfides by p-MC/PE. The outstanding reversibility and minimal degradation observed agree with the cycle performance metrics described above, implying that p-MC plays a critical role in prolonging the operational lifespan and maintaining cell integrity. Notably, the oxidation peak of the p-MC/PE cell appeared at a lower potential of 2.30 V in the 10th cycle than that of the control cell (2.49 V). This shift indicates accelerated oxidation kinetics in the p-MC/PE cell relative to those in the control cell. To quantitatively evaluate this kinetic difference, the specific oxidation capacities were determined by integrating the CV curves at three cutoff voltages: 2.4, 2.5, and 2.6 V (Fig. 5f). At a cutoff voltage of 2.6 V, the p-MC/PE cell achieved a slightly higher oxidation capacity of 1002 mAh g–1 than the control cell (942 mAh g–1). More pronounced differences were observed at a low cutoff voltage of 2.4 V, where the oxidation capacity of the p-MC/PE cell reached 785 mAh g–1, significantly surpassing that of the control cell (596 mAh g–1). These results confirm that the p-MC/PE cell can charge more rapidly than the control cell.
Post-mortem analysis
After 100 cycles, both the cells were discharged to 1.8 V and disassembled to examine the surfaces of the respective separators. The digital image of the anode side of the PE separator (Fig. S4a) showed that its surface was covered with yellowish materials that were transferred from the Li metal surface during cell disassembly. The corresponding SEM image shown in Fig. 6a further highlights the presence of numerous particulate precipitates on the anode side. EDS elemental mapping (inset of Fig. 6a) revealed a pronounced sulfur signal, indicating that these precipitates were likely the result of polysulfide diffusion from the cathode to the anode during cell operation. The cathode side of the PE separator exhibited a significantly different morphology, characterized by sparsely scattered particles (Fig. 6b) and a markedly reduced sulfur signal (inset of Fig. 6b). These results were due to insoluble polysulfides, such as Li2S and Li2S2, which were evident in the SEM analysis. Figure 6c presents an SEM image of the Li metal anode surface collected from the control cell after 100 cycles, which shows a highly rough and porous surface texture on the Li metal. This porous structure results from the formation of precipitates due to irreversible reactions between the Li metal anode and soluble polysulfides [55,56]. The anode side of the p-MC/PE separator exhibited an almost pristine surface (Fig. S4b), in contrast to that of the pristine PE separator (Fig. S4a). The corresponding SEM image in Fig. 6d confirms that the anodic side is clean, with only a few visible precipitates, indicating that the formation of a passivation layer due to polysulfide diffusion was significantly suppressed. The EDS sulfur mapping image (inset of Fig. 6d) shows a weak sulfur signal, suggesting that the p-MC/PE separator completely blocked the diffusion of polysulfides to the Li metal anode. Figure 6e presents an SEM image of the cathode side of the p-MC/PE separator, which is nearly identical to the SEM image obtained before cycling, indicating that the coated p-MC layer remained intact and structurally stable during repeated cycling. In contrast to the anode side, the EDS sulfur mapping image of the cathode side exhibited a uniform and strong sulfur signal (inset of Fig. 6e), suggesting that the p-MC/PE separator effectively captured soluble polysulfides. These contrasting results highlight the efficacy of the p-MC coating layer in preventing polysulfide diffusion. In contrast to the control cell, the Li metal surface in the p-MC/PE cell appeared smoother and less damaged (Fig. 6f), with only a faint sulfur signal detected in the EDS elemental mapping (inset of Fig. 6f). These observations confirm that the degradation of the Li metal surface by polysulfide attack was minimal, as only a trace amount of polysulfides reached the Li metal anode through the p-MC/PE separator.
SEM and EDS sulfur elemental mapping images of the separator and lithium metal anode surfaces collected from pristine PE and p-MC/PE cells after 100 cycles. (a) SEM image of the pristine PE separator (anode side), (b) SEM image of the pristine PE separator (cathode side), (c) SEM image of the lithium surface in the pristine PE cell, (d) SEM image of the p-MC/PE separator (anode side), (e) SEM image of the p-MC/PE separator (cathode side), and (f) SEM image of the lithium surface in the p-MC/PE cell. The insets in the SEM images show the corresponding S mapping. XPS spectra of the cycled p-MC/PE separator, collected at the end of discharge after 100 cycles, showing (g) S 2p with deconvoluted peaks for polythionate, thiosulfate, bridge sulfur, and terminal sulfur species, (h) O 1s with peaks for SO3− and C=O (carbonyl, carboxyl), and (i) P 2p with peaks for pyrophosphate.
The S 2p3/2 XPS characterization of the cycled p-MC/PE separator (Fig. 6g) revealed that the peaks corresponding to terminal sulfur (162.9 eV) and bridge sulfur (163.7 eV) shifted to higher energy values relative to those of Li2S6 (terminal: 161.6 eV; bridge: 163.1 eV) [57‒59]. This shift is attributed to the interaction between the polar functional groups of p-MC and the polar LiPS species [48‒50,60,61]. Moreover, the thiosulfate (O3S-S, 2p3/2 = 166.7 eV) and polythionate (O3S-Sx-S, 2p3/2 = 168.8 eV) signals arose from the interaction between the LiPSs and −OH groups on the surface of the p-MC, as follows [50,61‒65]:
Thiosulfate : LiPSs + −OH → [O3S-S]2−
Polythionate : LiPSs + [O3S-S]2− → [O3S-Sx-SO3]2− + Li2Sy (x ≥ 2, y < 3).
The thiosulfate groups attached to the carbon wall function as intrinsic anchors, capturing soluble polysulfides and facilitating their conversion into polythionates and insoluble short-chain polysulfides. This process markedly slowed the migration of polysulfides, mitigating the shuttle effect. After cycling, the O 1s XPS profile (Fig. 6h) exhibited the disappearance of the −OH group, along with the emergence of a new peak at 531.5 eV, which is linked to thiosulfate or polythionate species [66‒68]. This provides strong evidence for the formation of thiosulfate and polythionate described above.
The binding energy of the C=O group shifted to a slightly higher value (by 0.4 eV), supporting its chemical interaction with the lithium ions (Fig. S5). This shi+ stems from the electron-donating effect of oxygen to Li⁺, which reduces the electron cloud density and results in a higher binding energy [48–50,60,61]. The P 2p3/2 XPS profile of the cycled p-MC/PE separator (Fig. 6i) revealed that the peak corresponding to metaphosphate disappeared, whereas that corresponding to pyrophosphate shifted to a higher binding energy, that is, from 132.7 to 133.4 eV. The disappearance of the metaphosphate peak may have resulted from its conversion to pyrophosphate during cell operation. The shift in the pyrophosphate peak confirms the chemical interaction between pyrophosphate and Li⁺ (Fig. S5). This can be explained by the electron-donating effect of oxygen in P=O to Li+, which reduced the electron cloud density around phosphorus and shifted the P 2p peak to a higher binding energy. Overall, XPS analysis revealed that the LiPS species were entrapped via chemical interactions with the p-MC/PE separator during cycling, which is consistent with the sorption test results of the p-MC/PE separator in the Li2S8 solution (Fig. S6).
Efficacy of polar functional groups on p-MC
To further investigate the efficacy of the chemical interactions of p-MC with LiPSs, nonpolar mesoporous carbon (n-MC) with a lower concentration of heteroatoms than p-MC was synthesized at 900°C using the procedure described in Section 2.1. The n-MC had a similar microstructure to p-MC in terms of pore volume and pore size distribution (Fig. S7a), with a slight increase in the Lc value of the carbon walls from 7.84 Å for p-MC to 9.47 Å for n-MC (Fig. S7b) [69]. In cycling stability tests (Fig. 7a), the p-MC/PE separator outperformed its n-MC/PE counterpart, retaining 83.1% of its initial capacity after 300 cycles compared with the 63.0% capacity retention exhibited by the latter. This indicates that p-MC interacts more strongly with LiPSs, confining them to the cathode region. As the temperature increased from 400 to 900°C, the H/C and O/C ratios decreased from 0.44 to 0.21 and from 0.17 to 0.03, respectively (Table 1), indicating the removal of heteroatoms from the carbon structure. The higher temperature resulted in a more carbon-rich structure with fewer oxygen- and hydrogen-containing groups. The C 1s and O 1s XPS profiles (Fig. S7c and S7d, respectively) indicated that in contrast to p-MC, which retained abundant functional groups such as–OH, C=O, and COOH, n-MC contained only trace amounts of –OH groups. The lack of polar functional groups in n-MC resulted in weak interactions with LiPSs, contributing to more severe capacity degradation over multiple cycles owing to its inert surface properties.
(a) Cycle-performance comparison of Li-S cells using the n-MC/PE separator and the p-MC/PE separator over 300 cycles. (b) Self-discharge behavior of pristine PE and p-MC/PE cells after a 3-d rest period at room temperature, showing a considerably lower capacity loss for the p-MC/PE cell.
Compositions and structural properties of the as-synthesized p-MC and n-MC materials. The H/C, O/C, and P/C ratios were determined via elemental analysis (H, C, and O) and ICP-OES (P). The crystallinity parameters, including the Lc and d (interlayer spacing) values, were estimated from the (002) diffraction peak.
Aside from their limited cycling longevity, Li-S batteries suffer from self-discharge and reduced shelf life due to the reactivity of LiPSs, which delays their commercialization. Room-temperature storage tests were conducted immediately after cell assembly. The p-MC/PE cell exhibited a considerably more stable open-circuit voltage (OCV) of approximately 2.4 V over 10 d, whereas the control cell exhibited a gradual decline in the OCV after 80 h (Fig. S8). In addition, the self-discharge behavior of the Li-S cells was monitored by allowing them to rest for 3 d at 25°C after 30 cycles. As shown in Fig. 7b, the control cell exhibited a capacity loss of 35.35%. In contrast, the capacity loss of the p-MC/PE cell decreased markedly to 8.45%. These results confirm that the p-MC/PE separator securely immobilized LiPSs on the cathode side, mitigating self-discharge.
CONCLUSIONS
We developed a novel polar mesoporous carbon (p-MC) material functionalized with polar groups at a pyrolysis temperature of 400°C to address the critical challenge of polysulfide migration in Li-S batteries. The p-MC material, with diverse polar functional groups such as carbonyl, carboxyl, hydroxyl, and pyrophosphate, which were designed to interact strongly with LiPSs, exhibited high porosity (1.84 cm3 g–1) and a large surface area (1070 m2 g–1). When integrated into a PE separator, it effectively suppressed polysulfide diffusion, resulting in significantly enhanced sulfur utilization, reduced self-discharge, and excellent cycling stability. An Li-S cell with the p-MC/PE separator retained 83.1% of its initial capacity after 300 cycles, which was almost double the capacity retention achieved using a conventional PE separator. Furthermore, the p-MC/PE cell demonstrated superior cyclability relative to the cell with the nonpolar mesoporous carbon (n-MC/PE), which lacked the strong polar interactions necessary for effective polysulfide confinement, leading to greater capacity degradation.
The effectiveness of the p-MC/PE separator was also reflected in its ability to minimize self-discharge, with a considerably more stable OCV over prolonged storage and a capacity loss of only 8.45% compared with 35.35% for the control cell. Additionally, the p-MC/PE cell exhibited an excellent rate capability, retaining 66.5% of its initial capacity at 2.0 C, compared with 20.6% for the control cell. The superior efficiency of the p-MC/PE separator is attributed to the strong polar-polar interactions between the LiPSs and the functional groups on the surface of the p-MC, as well as its highly porous structure. These properties led to the efficient confinement of polysulfides via both chemical bonding and physical absorption, as confirmed by XPS, SEM, and EDS analyses. These findings underscore the considerable potential of the p-MC/PE separator for substantially improving the long-term stability, rate capability, and self-discharge characteristics of Li- S batteries. The scalability, affordability, and simplicity of the fabrication process further highlight the practicability of this approach for real-world applications. This study offers a promising solution to address the major technical bottlenecks hindering the successful commercialization of Li-S batteries by leveraging the critical roles of polar functional groups and mesoporous structures.
Notes
ACKNOWLEDGEMENTS
This work was supported by the Education and Research Promotion Program (2025) of KOREATECH and Individual Basic Science and Engineering Research Program funded by the Ministry of Education through the National Research Foundation of Korea (Grant No. 2021R1F1A1062040). The authors express their gratitude to the Cooperative Equipment Center at KOREATECH for assistance with FE-SEM, FIB-SEM, ICP-OES and XPS analyses.