Recent advances in functionalized separators for shuttle-free and dendrite-free lithium/sodium-sulfur batteries

Shuttle effect of Li/Na polysulfide

The discharge process of Li/Na-sulfur batteries involves the reduction of sulfur (S₈) to polysulfides (Li₂S_n, 2 ≤ n ≤ 8) and the release of electrons and Li⁺, accompanied by solid-liquid-solid phase conversions. The typical discharge reaction of Li-S batteries is recognized as four stages as follows as shown in Figure 1B^[18].

Stage I: The solid-phase S₈ ring in the cathode is reduced to Li₂S₈. This involves the solid-to-liquid conversion and the dissolution of polysulfides in the electrolyte, with the reaction plateau voltage above 2.4 V given in

In stage II, Li₂S₈ migrates to the surface of the lithium metal anode under the influence of chemical diffusion and electric field forces and is reduced stepwise to higher-order polysulfides Li₂S_n (4 ≤ n ≤ 8) as given in

During this stage, the transformation of polysulfides occurs entirely in the electrolyte, accompanied by a voltage plateau drop from approximately 2.4 to 2.1 V.

The stage III involves the conversion of soluble Li₂S₄ to insoluble lower-order polysulfides Li₂S_n (1≤ n ≤3), expressed as

The stage IV refers to the solid-solid conversion of insoluble Li₂S₂ to Li₂S as

The last two stages, III and IV, are the slowest in terms of reaction kinetics during discharge because of the involvement of solid-state conversions^[19]. Similarly, during the discharge process of Na-S batteries, a similar solid-liquid-solid transformation from S₈ → Na₂S occurs as shown in Figure 1C, similar to that in Li-S batteries^[20]. The subsequent charging reaction of Li-S/Na-S batteries involves a multi-step reverse conversion from Li₂S/Na₂S to S₈.

Figure 1. Current issues of Li/Na-sulfur batteries. (A) Schematics of the issues faced by Li-S batteries. (B) Theoretical charging/discharging profile of Li-S battery. (C) Theoretical discharge profile of Na-S battery. (D) Schematic illustration of the redox reactions on metal anode surface. This figure is quoted with permission from Lei et al.^[4], Wu et al.^[18], Yu et al.^[20], and Cooper et al.^[23], respectively.

The “shuttle effect” in Li/Na-S batteries occurs when soluble high-order polysulfides pass through conventional separators and reach the anode, reacting with Li/Na metal to form insoluble lower-order polysulfides (Li₂S_n/Na₂S_n, 1 ≤ n ≤ 3)^[21]. During charging, the lower-order polysulfides are unable to be oxidized back into higher-order forms and return to the cathode, leading to the accumulation on the anode as “dead sulfur”. The shuttle effect of Li/Na polysulfides causes active material loss and increased irreversible capacity, reducing the practical capacity of Li/Na-sulfur batteries. Enhancing polysulfide adsorption and accelerating the conversion via physical/chemical methods are effective strategies to suppress the shuttle effect and improve battery performance. It is noteworthy that the formation and shuttle behavior of polysulfides in Na-S batteries exhibit significant differences compared to Li-S batteries. On the one hand, due to the larger ionic radius of Na⁺, the complete discharge process from S₈ to Na₂S brings more severe volume expansion to the cathode, leading to increased issues of active material detachment and degradation (157% in Na-S batteries, compared to 80% in Li-S batteries)^[22]. On the other hand, the weaker interaction between Na⁺ and sulfide anions, derived from the lower electronegativity of Na⁺, results in decreased interfacial stability and slower reaction kinetics, exacerbates the accumulation of soluble polysulfides and intensifies the shuttle effect.

Uncontrollable growth of Li/Na dendrites

For Li/Na-sulfur batteries, uneven Li/Na ion deposition results in the formation of “dead lithium” and “dead sodium”, which are unable to participate in future electrochemical reactions, and aggravate undesirable metal dendrite growth on the surface of alkali metal anodes, as shown in Figure 1D^[23]. The dendrites are prone to penetrate the separator, leading to short circuits, which not only reduce battery lifespan but also pose severe safety risks, including overheating and fire hazards^[24].

The Sand’s time model is widely applied for studying the formation and growth mechanism of metal dendrites. Taking Li-S batteries for instance, the concentration of Li⁺ near the anode surface gradually decreases as the migration of Li⁺ towards the cathode. When the Li⁺ concentration reaches near zero at a specific current density, the formation of dendrites will be initiated, which is closely related to the Sand’s time, as given in^[25]

where J represents the effective electrode current density, D denotes the ambipolar diffusion coefficient, e is the electron charge, t_a stands for the anionic transference number, and C_o refers to the initial Li salt concentration.

At high current densities, the rate of Li⁺ diffusion lags behind that of electroplating, leading to local supersaturation and uneven deposition, which triggers dendrite formation^[26]. Continuous Li⁺ deposition at the dendrite tips during cycling accelerates dendrite growth and threatens Li-S battery safety. According to Eq. (8), the nucleation time of metal dendrites inversely correlates with the current density J and directly with the Li⁺ transference number t_Li+ (which refers to 1-t_a). Thus, reducing local current density and increasing ion transference is vital for controlling dendrite growth.

In summary, during Li-S battery charging, Li⁺ deposition on the anode is greatly affected by current density, surface energy, and electrode interface non-uniformity, leading to localized lithium deposition and dendrite formation. Accumulated Li in protruding areas enhances the local electric field, promoting further Li⁺ deposition and accelerating dendrite growth. Additionally, the SEI layer formed during initial charging often has structural and compositional inconsistencies, causing uneven Li⁺ conduction, SEI rupture, and further dendrite growth, which results in irreversible capacity loss and safety risks. In Na-S batteries, the sodium dendrites share almost similar formation and growth progress to lithium dendrites. However, the lower surface and adsorption energies of Na metal make uniform nucleation more difficult. Additionally, the lower chemical and electrochemical stability of Na⁺ during high-current discharging exacerbates conductive pathway disruption and uncontrolled dendrite growth, leading to rapid degradation of battery performance. Thus, designing functional separators to inhibit sodium dendrite growth is more challenging than in Li-S batteries.

General strategies

Common optimization strategies for Li/Na-S batteries include cathode and anode modification, electrolyte optimization, and separator functionalization^[27]. Cathode improvements focus on enhancing conductivity and rigidity through physical regulation, chemical intercalation, and redox mediators, which aid sulfur redox reactions and manage volume changes^[28]. Anode modifications encompass the development of three-dimensional (3D) anodes and artificial SEI layers. Electrolyte optimization, particularly in Na-S batteries, emphasizes solid-state electrolytes to address compatibility issues with Na metal^[29].

Functionalized separator strategies

In Li/Na-S batteries, the separator contacts the sulfur cathode, Li/Na anode, and electrolyte, playing a key role in electrochemical reactions and ion transport^[30]. The strategy towards functionalized separators with enhanced mechanical strength, regulated ion and molecular transport and suppressed shuttle effects and dendrite growth^[15]. Traditional polypropylene (PP) and polyethylene (PE) separators exhibit poor mechanical properties, making them inadequate for handling electrode volume changes. The large, irregular pore sizes allow uncontrolled polysulfide diffusion, leading to capacity loss derived from the shuttle effect, and uneven ion deposition, causing dendrite growth. Current separator functionalization strategies include replacing separators with novel ones, adding functional layers, and integrating separator functionalization with electrode and electrolyte modifications^[31].

(a) Replacement of conventional separators

Polymer separators, including polyvinylidene fluoride (PVDF), polytetrafluoroethylene (PTFE), polyimide (PI), and polyacrylonitrile (PAN), are common alternatives to PP and PE separators. PVDF offers chemical stability and electrolyte wettability but lacks mechanical strength^[32]. PI provides strong mechanical properties and thermal stability, but the application is limited by the high cost^[33]. Similarly, the inferior economic efficiency of PTFE restricts its widespread application despite the stability advantages^[34]. PAN is relatively competitive due to its mechanical strength, flexibility, and wettability but still requires further functionalization^[35].

Glass fiber (GF) separators are promising separator alternatives due to their high porosity, thermal stability, electrolyte wettability, and electrochemical performance^[36]. Although multifold thicker than PP, GF offers significant advantages in mechanical strength, electrolyte uptake, and ion transport, enhancing polysulfide capture and metal ion transport, and is particularly suited for Na-S batteries due to the superior porosity^[37]. However, the thickness of GF separators, generally reaching over 100 μm, adds extra volume to the battery.

(b) Functional interlayer loading

Loading functional interlayers on commercial separators effectively regulates polysulfide diffusion and metal deposition and is efficient for achieving bifunctionality in suppressing the shuttle effect and dendrite growth. Interlayers can generally be categorized into double-sided loading and single-sided bifunctional material loading.

(i) Double-sided loading separators include sandwich and Janus types. Sandwich separators have functional materials on both the cathode and anode sides with the insulating membrane in the middle, while Janus separators consist of only two functional layers. On the cathode side, functional materials typically feature strong electronegativity, abundant polar bonds, nanoscale structures, or high catalytic activity, aiding in polysulfide adsorption and enhanced redox kinetics^[38]. On the anode side, materials with uniform microstructure and low lithium adsorption energy help reduce side reactions, redistribute metal ion flow, and suppress dendrite growth.

(ii) The single-sided loading strategy focuses on bifunctional materials with both sulfiphilic and lithiophilic/sodiophilic properties. The bifunctional materials have high polarity and strong interactions with polysulfides and metal ions, significantly contributing to suppressing the shuttle effect and dendrite growth simultaneously. Compared to double-sided modifications, single-sided bifunctional materials are more cost-effective and help increase energy density while barely bringing additive weight to batteries.

Integrated strategies

Separator modification alone cannot fully address the slow redox kinetics, volume changes, and side reactions in Li/Na-S batteries. Thus, integrated strategies that combine separator functionalization with electrode modification and electrolyte optimization are crucial^[39]. Through the combined effects of modified cathodes and functionalized separators, ion transport and reaction kinetics between the separator and cathode are accelerated, leading to better capacity retention of the battery. With the combined effects of modified metal anodes and/or artificial SEI layers along with functionalized separators, the flux of metal ions is maximally regulated, resulting in uniform metal deposition and the suppression of dendrite growth. Furthermore, the synergistic effects of optimized electrolytes and functionalized separators enhance the transport of electrons and ions, accelerate solid-liquid and liquid-phase reactions, and suppress side reactions^[40]. Despite being constrained by cost and production conditions, integrated strategies remain a vital approach to maximizing the capacity, energy density, cycle life, and safety of metal-sulfur batteries.

Inhibiting the shuttle effect of polysulfide

Physical/chemical adsorption and catalytic conversion are considered the most important functionalization strategies to suppress the polysulfide shuttle and protect both sulfur and metal electrodes, as shown in Figure 2A^[41], which enhance the adsorption capacity for polysulfides, confining polysulfides to the cathode side of the separator and accelerating the conversion rate of polysulfides. Unlike in Li-S batteries, where Li₂S is the only thermodynamically stable species after full discharge, Na_nS (with n = 1, 2, 4, 5) in Na-S batteries exhibits significant thermodynamic stability at room temperature, which may precipitate out of the electrolyte as solids during discharge. Therefore, for separators in Na-S batteries targeting the inhibition of the shuttle effect, simultaneously blocking and regulating the shuttle and precipitation behavior of polysulfides presents a new challenge and a critical requirement.

Figure 2. Requirements for inhibiting polysulfide shuttle effect. (A) Configuration diagram of a Li-S battery with a multifunctional separator. (B) Schematic illustration of physical blocking of polysulfides. (C) Schematic illustration of chemical adsorption of polysulfides. (D) Catalytic process of sulfur species conversions. (E) Illustration of the catalysis mechanism of Ni nanoparticles to polysulfides. This figure is quoted with permission from Yang et al.^[41], Wang et al.^[42], Razaq et al.^[44], He et al.^[47], and Zhang et al.^[48], respectively.

(a) Physical blocking

Physical blocking against polysulfides generally relies on electrostatic repulsion and sieving effects to achieve physical and spatial confinement of polysulfides, as shown in Figure 2B^[42]. Firstly, according to the electrostatic repulsion principle, introducing negative charges on the separator is favorable for confining S_n^2- and S_n/2^- species after ionization and dissociation to the cathode side and preventing soluble M₂S_n (4 ≤ n ≤ 8) from migrating across the separator. However, the accumulation of negatively charged polysulfides around the cathode can lead to the formation of a low-conductivity inert layer on the cathode surface and hinder metal ion plating. Therefore, it is essential to precisely regulate the charge density of the separator. The sieving effect strategy aims at constructing elective free transport of Li⁺/Na⁺ by utilizing the particle size difference between polysulfides and alkali metal ions. The diameter of Lithium and Sodium ions are only 1.52 and 2.04 Å, respectively, whereas the van der Waals diameters of higher-order polysulfides range from 10.99 to 19.19 Å which increase further when dissolved in the electrolyte^[43]. Thus, constructing artificial small-sized nanopores on commercial separators (pore sizes of 50-100 nm for PP and 1.2-2.7 μm for cellulose separators) is a reasonable strategy to isolate larger polysulfides on the cathode side while allowing the free transport of Li⁺/Na⁺.

(b) Chemical adsorption

Chemical adsorption primarily relies on Lewis acid-base interaction and polar bonding adsorption to anchor the polysulfides and prevent the migration of soluble M₂S_n, as shown in Figure 2C^[44]. Incorporating materials with Lewis acid centers that accept electron pairs on the separator can improve the formation of coordinate covalent bonds with soluble polysulfides that donate electron pairs in electrolytes, thereby inhibiting the shuttle effect. As for polar bonding adsorption, the rich polar covalent bonds between sulfur atoms of different valence states in polysulfides make it feasible for polar materials to generate strong chemical interactions with polysulfides^[45]. Thus, applying sulfiphilic polar materials is a promising choice for anchoring polysulfides through chemical adsorption and inhibiting the shuttle effect.

(c) Catalytic conversion

Enhancing the reutilization of adsorbed high-order sulfur species is critical to inhibiting the polysulfide shuttle effect. Once adsorbed, the polysulfide diffusion towards the anode is blocked, and the surface of the separator becomes the new platform for subsequent redox conversions instead of the anode. The separator interface becomes crucial for phase conversions of sulfur species and the return of converted polysulfides to the cathode^[46]. Loading sulfiphilic catalytic materials or combining them with conductive substrates on the separator can accelerate redox kinetics, improve sulfur utilization, and enhance battery capacity retention, as shown in Figure 2D and E^[47,48].

Suppressing dendrite growth

To inhibit the uncontrollable growth of lithium and sodium metal dendrites, strong lithiophilic/sodiophilic properties and uniform high-flux metal flow handling capabilities to regulate metal deposition behavior are crucial for functionalized separators in Li-S and Na-S batteries. Assisting in the formation of a stable and uniform SEI layer is also a key requirement for functionalized separators, especially for Na-S batteries.

Firstly, the separator is supposed to possess sufficient mechanical strength and puncture resistance so as to physically smash the dendrites before the penetration, as shown in Figure 3A^[24,49]. Besides, the anode volume changes due to metal ions stripping/plating and the formation and breakdown of the SEI, and the transport of active substances will exert physical stress on the separator. Thus, enhancing physical strength is feasible to reduce the puncture risk caused by continuous dendrite growth.

Figure 3. Requirements for suppressing metal dendrite growth and bifunctional separators. (A) Schematic illustration of strengthened separator countering dendrite growth. (B) Illustration of uniform lithium-ion transform. (C) Schematic illustration of lithium-ion conduction behavior in the functional separator. (D) Configuration of Li-S batteries with bifunctional separator. (E) Schematic of the working mechanism of bifunctional Ni₃S₂@C separator. This figure is quoted with permission from Liu et al.^[24], Hou et al.^[50], Cao et al.^[52], Huang et al.^[53], and Zhang et al.^[54], respectively.

Secondly, the uneven flow of metal particles is one of the major causes of lithium dendrite growth. Traditional separators with irregular pore sizes and poor electrolyte wettability lead to uneven ion deposition. Incorporating lithiophilic materials can create uniform, high-speed ion transport pathways, regulating ion migration and suppressing early-stage dendrite growth [Figure 3B]^[50,51].

Thirdly, facilitating the formation of a stable SEI layer on the anode surface is also a crucial function of the functionalized separator for suppressing lithium dendrite growth. The accumulated polysulfides on the anode not only disrupt the initially smooth plating/stripping anode surface but also form vulnerable SEI layers with a poor mechanical property that triggers continuous side reactions, metal and electrolyte consumption, and deteriorate the transport kinetics of metal ions, producing “dead Li/Na” and expediting dendrite growth, ultimately resulting in pulverization of the anode. Note that compared to lithium dendrites, sodium dendrites are relatively softer due to the lower Young's modulus of Na metal and conduct less pressure on the SEI layer. However, the SEI layer on the sodium anode is highly unstable, which exposes additional sodium nucleation sites after fracturing and exacerbates Na dendrite growth, leading to severe volume changes and internal stress compression on the anode. Therefore, functional separators are supposed to assist in forming a homogeneous structure and enhancing metal ion transfer capability to construct stable SEI layers with high ionic conductivity.

Bifunctional separators

In addition to addressing the shuttle effect, the strong influence of uncontrollable dendrite growth on the retention of capacity and cycling stability in Li-S and Na-S batteries makes it extremely important to explore bifunctional separators. Note that in this review, “monofunctional separators” refer to separators that have been specifically tested for the ability to block polysulfides, while “bifunctional separators” specifically refer to the separators that can simultaneously inhibit both the shuttle effect and dendrite growth.

The necessary capabilities of bifunctional separators are as follows [Figure 3C-E]^[52-54]: (i) Adequate mechanical intensity to withstand external pressure, mitigate interface issues, and counteract volume changes and dendrite growth; (ii) Appropriate physical/chemical polysulfide anchoring ability to inhibit the shuttle effect; (iii) Improved catalytic capability to enhance polysulfide conversion and utilization efficiency; (iv) Superior electrolyte wettability and suitable ionic and electronic conductivity for fast material transport; and (v) Enhanced metal ion adsorption and transport capabilities to regulate metal ion transport behavior and help for constructing stable SEI layers to achieve uniform metal ion deposition and prolonged cycling stability.

Physical/chemical blocking and adsorption of polysulfide

Monofunctional interlayers for suppressing shuttle effect are generally categorized into: carbon materials with excellent mechanical properties and diverse microstructures, metal-organic frameworks (MOFs)/covalent organic frameworks (COFs) and polymers with tunable microstructures and functionalization feasibility, and polar metal compounds.

(a) Carbon materials

Carbon nanomaterials encompass diverse interesting nanostructures, including 0D dots, 1D nanowires and nanotubes, 2D nanosheets, and 3D nanospheres and foam networks. Certain carbon nanomaterials, such as graphene and carbon nanotubes (CNTs), exhibit excellent conductivity that can accelerate electron and ion transport and facilitate the formation of electrostatic/chemical adsorption networks for polysulfides, thereby suppressing the shuttle effect and enhancing the utilization of the sulfur cathode^[55].

Liu et al. introduced a thin carbon layer derived from cathode inert materials into the separator, which utilized existing cathode materials and avoided additional weight, as shown in Figure 4A^[56]. Carbon materials enriched with heteroatoms such as N, O, and B play a significant role in the chemical adsorption of polar polysulfides^[57,58]. A biomass carbon doped with nitrogen and oxygen is endowed with abundant micro/mesoporous structure, grand specific surface area, ideal electrical conductivity, and the nitrogen and oxygen heteroatoms that exhibit strong chemical adsorption for lithium polysulfides (LiPSs), which synergistically ensure adequate adsorption of polysulfides^[59]. The carbon materials with rich porous structures can extend the diffusion path of polysulfides from the sulfur cathode to the metal anode and enhance the adsorption efficiency of incorporated functional materials. Ma et al. incorporated a 3D mesoporous carbon nanocage on PP as a polysulfide buffer and cathode redox conversion boosting layer which redistributes the sulfur species through the abundant mesoporous structure^[60]. CNTs provide abundant deposition sites for LiPSs due to their outstanding electrical conductivity, superior mechanical strength and chemical stability and large specific surface area derived from the 1D novel nanostructure, and are widely used as one of the preferred materials for shuttle-effect-suppressing materials. The composite interlayer of mesoporous multi-walled CNTs and ferroelectric barium titanate with high dielectric constant can serve as a “conductive scaffold” in Li-S batteries, providing sufficient conductive pathways for ion and electron transport adsorption capacity for LiPSs^[61]. Graphene is another commonly used 2D conductive carbon substrate. Jiang et al. demonstrated the effectiveness of directly using graphene oxide (GO) as a functional layer coated on PP, as shown in Figure 4B^[62]. Reduced GO (rGO), with abundant electronegative functional groups, not only easily forms electrostatic composites with other materials but also effectively blocks polysulfides through electrostatic repulsion^[63]. A thin layer of rGO and sodium lignosulfonate composite with a porous network structure directly coated on PP can achieve electrostatic repulsion polysulfides through the abundant negatively charged sulfonate groups in SL, while still allowing normal Li⁺ ion transport^[64].

Figure 4. Monofunctional separators with carbon materials, MOF/COF and polymers for Li-S batteries. (A) SEM and TEM images of PCN. (B) Schematic of Li-S cells with GO-coated separator. (C) Fabrication process of batteries with CPC separators. (D) Schematic illustrations of Li-S cells with TB-BAA/SWCNT/PP separators. This figure is quoted with permission from Liu et al.^[56], Jiang et al.^[62], Zhang et al.^[67], and Sun et al.^[69], respectively.

(b) MOF and COF materials

MOF and COF materials possess adjustable porosity, large surface areas, and highly-ordered uniform porous structures that are typically smaller than the dimensions of higher-order LiPSs, enabling MOFs/COFs to block the liquid-phase diffusion of long-chain and large-size polysulfides by sieving effect^[65]. Moreover, MOFs/COFs can not only serve as conductive substrates but also act as sacrificial templates and the MOFs can provide abundant Lewis acidic centers for strong interaction with soluble polysulfides^[66]. A carbon photon crystal (CPC) with 3D ordered hierarchical porous structure can be synthesized by templating method, as shown in Figure 4C^[67]. The CPC-modified separator leverages the hierarchical macro-meso-microporous network to act as a polysulfide trap and boost the adsorption of LiPSs and reutilization of active sulfur. Despite high surface area, tunable pore structures, and excellent thermal and chemical stability, MOF/COF materials face challenges due to the inherently fragile structures and complex synthesis processes. These factors can lead to degradation during cycling and increased costs, which are detrimental to large-scale commercial applications.

(c) Polymers and polymer additives

The inherent complex microporous and mesoporous structures of polymers are highly beneficial for spatially inhibiting the diffusion of polysulfides. Conductive functionalization techniques can further enhance the chemical adsorption capacity of polymers for polysulfides^[32]. Moreover, polymers possess a mature and controllable fabrication process, significantly reducing manufacturing costs and process complexity^[68]. Sun et al. constructed a novel zwitterionic microporous polymer (TB-BAA) on the grooved layer of single-walled CNTs (SWCNTs) as presented in Figure 4D^[69]. The ultra-thin TB-BAA top blocking layer, with zwitterionic groups, assisted in constructing constrained ion channels with lithiophilic and sulfiphilic groups, ensuring lithium-ion conductivity while simultaneously fixing polysulfides. The SWCNTs provided ample electron transport channels, reduced and homogenized the pore size of the PP membrane, and minimized the penetration of TB-BAA.

In short, inherent insulative polymers are unfavorable for enhancing the conversion efficiency of LiPSs and the redox kinetics of sulfur cathodes, making the efficiency of LiPSs blockage ability inferior to that of carbon or MOF/COF-based composites. However, the low-cost, controllable microstructure construction techniques make the development of polymer additives and polymer-based composite materials highly competitive.

(d) Common Metallic Compounds

Common metals, except for transition metals, can provide a certain degree of mechanical strength to serve as physical barriers against LiPSs and can form polar chemical bonds with non-metal elements, which enables the LiPSs immobilization of common metallic compounds^[70]. Aluminum oxides exhibit good mechanical strength and thermal conductivity and are favorable for heat redistribution during battery cycling^[71]. Lin et al. developed an interlayer based on a mesoporous alumina and graphene (MA@G) composite in Figure 5A^[72]. The mesoporous structure and high specific surface area (598 m² g^-1) of MA provide abundant exposed adsorption sites for LiPSs and have a low migration barrier for lithium, facilitating the suppression of LiPSs and the rapid transport of lithium^[72]. Some metal compounds can form nanosheets or other particular morphologies through self-assembly or liquid exfoliation, achieving larger specific surface areas and more exposed adsorption sites for LiPSs. Zhang et al. utilized ultrasonically exfoliated MgB₂ nanosheets on the surface of PP, providing d increased specific surface area and pore volume and the ability to accommodate the volume expansion associated with the conversion of lithium sulfide (Li₂S) as shown in Figure 5B^[73]. Unfortunately, due to the limited catalytic ability in the subsequent processes after immobilizing LiPSs, the application of common metal compounds in functionalized separators for Li-S batteries is significantly restricted.

Figure 5. Functional separators with common metal compounds for Li-S batteries. (A) MA-Based Interlayer. (B) Peeling process of MgB₂, rate performance and cycle performance at 0.5C of the cell using MgB₂@PP separator. This figure is quoted with permission from Lin et al.^[72], and Zhang et al.^[73], respectively.

Catalytic conversion of polysulfide

After being captured, the LiPSs are confined to the cathode side of the functional separator and participate in the subsequent conversion reactions (S₈ ⇄ Li₂S)^[74]. Therefore, the catalytic capability of the separator exerts significant impact on the conversion and reutilization efficiency of LiPSs that crucial for enhancing the redox reaction rate and utilization efficiency of the sulfur cathode. The catalytic functionalization of the separator in Li-S batteries primarily relies on transition metal compounds (TMCs) and the composites. Transition metals, located in the d-block and ds-block of the periodic table, can provide lone pair d-orbital electrons to act as nucleophiles or offer unfilled d-orbitals to act as electrophiles in chemical reactions, demonstrating strong catalytic activity. Moreover, the TMCs also exhibit strong redox properties but possess more stable chemical and thermodynamic characteristics, which is beneficial for reducing the activation energy of LiPSs conversion reactions and facilitating the rapid formation of intermediate LiPSs products^[75]. Additionally, TMCs can be easily synthesized into novel microstructures such as quantum dots and nanoribbons by simple chemical methods that further enhance the adsorption and catalytic capabilities of LiPSs^[76].

In transition metal oxides, the metal-O polar chemical bonds in transition metal oxides exhibit strong reactivity with polar LiPSs, providing abundant deposition sites for LiPSs, and demonstrate impressive catalytic activity, promoting the chemical adsorption and catalytic conversion of LiPSs^[77]. The acrylic cloth carbon (ACC)/MnO₂ hybrid material synthesized via hydrothermal method by Zuo et al. can serve as an efficient LiPSs captor and catalyst. The ACC substrate in ACC/MnO₂ provides a broad conductive network for electron and Li⁺ transport, while the polar MnO₂ offers strong chemical affinity for LiPSs and exhibits superior redox kinetics [Figure 6A]^[78]. Zhu et al. fabricated coated separators with CeO₂ nanoparticles supported on multi-walled CNTs (MWCNTs/CeO₂/PP), as shown in Figure 6B^[79]. The high affinity of CeO₂ for polysulfides effectively prevents the LiPSs transport through the separator, while the MWCNTs block polysulfide shuttling through physical adsorption and act as secondary current collectors to facilitate charge transfer^[79]. Compared to oxides, transition metal sulfides, phosphides and selenides have greater atomic numbers and can form stronger polar chemical bonds with metals, resulting in enhanced polysulfide adsorption capabilities^[80]. The interlayer with aligned hollow Co₉S₈@MOF arrays grown in situ on Celgard separators developed by He et al. helps to block LiPSs shuttling and fully utilize the sulfur cathode, which exhibits excellent long-cycle stability even under high sulfur loading of 5.6 mg cm^-2 [Figure 6C]^[81]. The MoS₂ nanosheet/PP composite separator obtained through a simple exfoliation method features strong polarity from MoS₂ that enables effective adsorption and anchoring of polysulfides as in Figure 6D. Besides, the large specific surface area of the nanosheet structure provides abundant Li⁺ adsorption sites and guarantees rapid adsorption and transport of Li^+[82]. Fan et al. designed a coral-like porous composite functional layer containing molybdenum disulfide (rGO/MoS₂/C) with large specific surface area and good electrolyte wettability^[83]. The active sites of MoS₂ effectively enhance the adsorption of LiPSs, demonstrating excellent cycling stability with a capacity decay of only 0.022% per cycle after 1,000 cycles at 2.0C. The symmetric cell with the rGO/MoS₂/C composite showed a larger CV curve enclosed area, indicating a lower energy barrier for the catalytic conversion of Li₂S_n^[83]. Zhou et al. prepared a novel interlayer composed of Fe₃Se₄/FeSe heterojunctions encapsulated by two-dimensional (2D) MXene nanosheets through the in situ growth and selenization of Fe-MOFs^[84]. The Fe₃Se₄/FeSe@MXene interlayer exhibits low adsorption energies on the surfaces of S₈, Li₂S₈, Li₂S₆, Li₂S₄, Li₂S₂, and Li₂S owing to the MXene framework and the outstanding synergistic catalytic effect of the Fe₃Se₄/FeSe heterojunctions^[84]. Wang et al. employed continuous low-temperature chemical vapor deposition to controllably grow generic graphene-MoSe₂ heterostructures (Gr-M_xSe_y) on commercial GF separators^[85]. The coupling of graphene with MoSe₂ as a polar and catalytic framework creates hybrid energy levels, enhancing the conductivity and electrocatalytic activity of the separator for sulfur redox reactions. The Gr-M_xSe_y/GF separator demonstrated high capacities of 1,118.7 mAh g^-1 at 0.2C and 552.8 mAh g^-1 at 3C^[85]. Unique structures such as quantum dots aid in the uniform and dense distribution of TMCs on the separator surface, enabling functionalization with less mass and higher chemical activity. The MoP quantum dots@N, P-doped hollow polypyrrole interlayer utilizes the excellent chemical adsorption and catalytic conversion capabilities of MoP quantum dots to suppress the LiPS shuttle effect [Figure 6E]^[86].

Figure 6. Functional separators with transition metal compounds for Li-S batteries. (A) LiPSs adsorption and catalytic effect of ACC/MnO₂. (B) Preparation of PP-MWCNTs/CeO₂ separators. (C) Long-term cycling profile at 1C of the Li-S cells with the Co₉S₈ coated separators. (D) Schematic of LiPSs adsorption of MoS₂/Celgard separator. (E) Fabrication and adsorption energies of the MPQ@NP-HCNT. This figure is quoted with permission from Zuo et al.^[78], Zhu et al.^[79], He et al.^[81], Ghazi et al.^[82], and Zhang et al.^[86], respectively.

C, N, and B atoms have more diverse valence states than O, resulting in stronger polar metal-C/N/B bonds and higher chemical stability^[87]. Transition metal carbides, nitrides (TMNs), and borides (TMBs) have relatively low average molecular masses, which hardly bring additional mass to the batteries. Compared to metal-O bonds, LiPSs are more inclined to bond with the stronger polar metal-C/N/B bonds, leading to lower adsorption and conversion energy barriers for LiPSs. Moreover, transition metal carbides, TMNs, and TMBs can easily achieve structural regulation by controlling synthesis conditions, resulting in microstructures such as nanowires and quantum dots that provide abundant anchoring sites for LiPSs and adsorption sites for Li⁺. The Mo₂C/C can act as a polar catalytic interlayer and exhibit dynamic adsorption and promoted kinetics for the reversible conversion of LiPSs^[88]. According to Pan et al.; the polar Fe₃C and doped N in Fe₃C-N-doped rGO enhanced atomic attraction for sulfur and LiPSs and achieved higher capacities at increased current densities as in Figure 7A^[89]. Ma et al. designed and synthesized a novel nitrogen-doped graphene (NG) nanosheets embedded with conductive WN_0.67 nanoneedle that enable rapid electron/ion transfer, effective anchoring of LiPSs, promotion of uniform Li₂S nucleation/growth, and improved sulfur utilization^[90]. As exhibited in Figure 7B, monodispersed Mo₂N quantum dot-modified nitrogen-doped carbon nanosheets (Mo₂N@NCNS) have also been proven to be an effective interlayer. The Mo₂N quantum dot enhances the intrinsic conductivity of the 3D NCNS network and provides abundant catalytic sites and strong polarity. Mo₂N@NCNS not only regulates the random motion of LiPSs through physical confinement and chemical adsorption, but also exhibits superior LiPSs catalytic and promotes the nucleation and deposition of Li₂S₂/Li₂S, thereby enhancing the overall electrochemical performance of Li-S batteries^[91]. Two-dimensional transition metal carbides, TMNs, or carbonitrides (MXenes) also exhibit surprising adsorption and catalytic capabilities for LiPSs due to the regular 2D sheet structure, large specific surface area, and abundant polar deposition sites^[92]. As shown in Figure 7C, Yang et al. addressed the agglomeration issue by using the electrostatic adsorption principle to self-assemble MXene quantum dots (MQDs) on the surface of g-C₃N₄ sheets to provide more active sites for capturing dissolved LiPSs^[41].

Figure 7. Functional separators with catalytic transition metal compounds for shuttle effect inhibition. (A) Cyclic-voltammetry profiles of Fe₃C-N-rGO coated separator and TEM image of Fe₃C-N-rGO. (B) Schematic illustration of Mo₂N@NCNS. (C) Preparation schematic diagram of g-C₃N₄@MQDs. (D) In situ Raman spectra in discharging processes of the cells with TiB₂/PP separators. (E) Long life cycle performance and Coulombic efficiency at 5C Li-S cells with Co₂B@CNT-separator. This figure is quoted with permission from Pan et al.^[89], Ma et al.^[91], Yang et al.^[41], Li et al.^[93], and Guan et al.^[94], respectively.

Compared to carbides and nitrides, borides are more abundant on Earth and contribute to more economical separator functionalization. The TiB₂, with a gradient distribution of conductive sites, adsorbs polysulfides via polar chemical bonds and accelerates the conversion by high catalytic properties as shown in Figure 7D^[93]. The Co₂B@CNT layer on PP developed by Guan et al. exhibited a high polysulfide trapping capability (Li₂S₆ at 11.67 mg/m²) owing to the co-adsorption of Co and B and showed a significant catalytic effect for the conversion of Li₂S₈ to Li₂S due to the catalytic Co [Figure 7E]^[94]. Due to the metalloid properties of boron and the abundant polysulfide adsorption sites, borides can form highly active planar-like polysulfide adsorption networks when uniformly dispersed on conductive substrates and exhibit excellent capabilities in suppressing the shuttle effect^[51,95,96].

In conclusion, TMCs with unique morphologies, such as quantum dots, nanowires/needles/ribbons, and nanosheets, serve as superior anchoring site doners and catalysts for LiPSs. As displayed in Table 1, the combination of TMCs with CNT, rGO, and MOF/COF frameworks aids in enhancing the mechanical strength of the separator and facilitates the construction of uniform and efficient networks for the adsorption and catalysis of LiPSs.

Double-sided separators

As mentioned before, double-sided separators for Li-S batteries generally show asymmetric structures in order to deal with shuttle effect of polysulfides and growth of lithium dendrites separately^[99]. Wang et al. coated Al₂O₃ particles on the anode side and sodium CMC on the other side of PE^[38]. The gel-polymer-inorganic separator utilizes the “polysulfide-phobic” interface formed by the CMC layer and the mechanical strength of Al₂O₃ to inhibit the growth of polysulfides and dendrites on the respective sides of the separator, as shown in Figure 8A^[38]. Chang et al. proposed a sandwich separator loaded with a composite thermal conductive layer (TCL) made of V₂O₅ microspheres and Al₂O₃ nanosheets^[100]. The V₂O₅ on the cathode side not only combines with the LiPS catalytic layer (LCL) on the surface of the S cathode to catalyze the conversion of LiPSs but also collaborates with the Al₂O₃ nanosheets on the anode side to redistribute Li⁺ flux and thermal energy. The Li-S batteries with the TCL/PP separator demonstrated excellent cycling stability, with a capacity decay rate of 0.036% per cycle over 2,000 cycles at a 1C rate as shown in Figure 8B^[100]. Dang et al. grew continuous cerium-based metal-organic framework-808 (Ce-MOF-808) on both sides of the PP separator^[101]. Due to the highly ordered nanostructure and polar groups, Ce-MOF-808 on the cathode side performs multiple functions of selective sieving, adsorption, and catalytic conversion of LiPSs. On the anode side, it promotes the rapid diffusion of Li⁺ and facilitates stable nucleation and growth of lithium [Figure 8C]^[101].

Figure 8. Double-sided bifunctional separators for Li-S batteries. (A) SEM images of the cycled Li-metal anodes disassembled from symmetrical cells with PP separator and the Al₂O₃-PE-CMC separator and polysulfide adsorption tests before and after adding CMC powder. (B) Long-term cycling performance of the LCL-TCL modified separator and the schematic of corresponding Li-S cells. (C) Configuration schematic of Li-S cells with Ce-MOF-808/PP separator, diffusion energy barrier of Li-ion of Ce-MOF-808, and the Li-Li symmetric cells voltage profiles. This figure is quoted with permission from Wang et al.^[38], Chang et al.^[100], and Dang et al.^[101], respectively.

Bifunctional interlayers on separators

Polar materials are one of the most common bifunctional materials and are generally combined with conductive substrates to further improve the surface structure of the pristine separator^[102,103]. Wang et al. applied anodic aluminum oxide (AAO) interlayers with high rigidity and non-flammability to enhance the safety of Li-S batteries. The orderly microstructure of AAO facilitates rapid transport channels for Li⁺, reducing the area-specific resistance and accelerating redox reactions^[104]. Liu et al. embedded ZrO₂ nanoparticles into a polyetherimide (PZ)-based separator using interfacial polymerization to enhance the Li⁺ transference number (0.55) and diffusion coefficient (2.16 × 10^-8 cm² s^-1) of the ZrO₂/PZ separator^[105]. The impressive bifunctionality of the ZrO₂/PZ is attributed to the pore size regulation, promoted electrostatic repulsion, and physical/chemical adsorption^[105].

TMCs exhibit excellent sulfiphilic and lithiophilic properties owing to the rich d-orbital electronic states and polar chemical bonds, making them the ideal bifunctional materials for fabricating separator interlayers. The double-sided “nanobrick wall” structure constructed by layer-by-layer self-assembly of MoS₂ and poly(diallyldimethylammonium chloride) with poly(acrylic acid) (PAA), marked as (M-P/P)₁₀, provides a high discharge capacity at a high 3C [Figure 9A]^[106]. Liu et al. prepared an ultra-small CoP (S-CoP, 25 nm) composite material embedded within CNT-modulated MOF. The S-CoP/CNT effectively promoted the conversion of Li₂S and provided ultra-long dendrite-free lithium deposition^[107]. Li et al. designed an NbB₂/rGO modified PP separator (NbB₂/rGO/PP) and revealed that the catalysis of LiPSs by NbB₂ relies on the passivated surface with the oxide layer (O₂-NbB₂), where B-O-Li and Nb-O-Li bonds significantly lower the decomposition barrier of Li₂S^[108]. The rapid conversion and reutilization of LiPSs help maintain the continuous exposure of active sites on the separator surface for Li⁺ adsorption, which is beneficial for enhancing the lithium-ion diffusion rate as displayed in Figure 9B.

Figure 9. Bifunctional separators with polar TMBs for Li-S batteries. (A) Schematic diagrams, rate performance and charge-discharge curves of cells with M-P/P)₁₀. (B) Schematic of the LiPSs conversion on the NbB₂/rGO separator and adsorption configuration of Li₂S_n on O₂-NbB₂. This figure is quoted with permission from Wu et al.^[106], and Li et al.^[108], respectively.

The TMCs with unique nanoscale microstructures can facilitate not only the highly effective entrapment of polysulfides but also the regular transport of Li⁺, such as heterostructure TMCs with built-in electric fields and 2D layered TMCs with high specific surface areas^[109]. Wang et al. designed a heterostructured MoP-MoO₂@defective porous carbon nanofibers (PCNFs) bifunctional interlayer with abundant anchoring sites as shown in Figure 10A^[110]. Benefiting from the shortened Li⁺ pathways provided by PCNFs, the heterostructured MoP-MoO₂ nanobelts exhibit excellent capabilities for intercepting and converting sulfur species and enabling long-term dendrite-free cycling of the anode^[110]. As in Figure 10B, Lv et al. synthesized a unique VSe₂/V₂C heterostructure in situ by sacrificing a V₂C precursor, effectively combining the strong affinity of V₂C for LiPSs with the high electrocatalytic activity of VSe₂^[111]. The local built-in electric field at the VSe₂/V₂C heterointerface significantly enhances electron and ion transport capabilities, ultimately improving the conversion kinetics of sulfur species and the transport kinetics of Li^+[111]. Incorporating MXene materials with high lithiophilicity and sulfiphilicity, derived from the inherent surface polar groups, provides higher conductivity and optimal flexibility to the bifunctional interlayer, significantly accelerating the cross-separator transport of Li⁺ and ensuring the integrity and stability of the Li-S batteries^[112]. For the MXene covered with TiS₂/TiO₂ heterostructure on the surface (TOS/MX/TOS), the TiO₂ nanoparticles adsorb and inhibit LiPSs, while the high-conductive and catalytic TiS₂ nanoparticles significantly promote the conversion of long-chain LiPSs into short-chain Li₂S₂/Li₂S. The TOS/MX/TOS interlayer also protects the lithium anode by preventing the undesirable deposition of polysulfides on the anode surface and attains stable cycling performance, as shown in Figure 10C^[113].

Figure 10. Bifunctional separators with MXenes and heterostructured TMBs for Li-S batteries. (A) HRTEM of MoP-MoO₂ and rate performance of the MoP-MoO₂/PCNT separator. (B) Schematic of Li-S cells with VSe₂/V₂C-CNTs-PP and energy band diagrams of V₂C and VSe₂. (C) Long-term cycling profile of the TOS/MX/TOS modified separator. This figure is quoted with permission from Wang et al.^[110], Lv et al.^[111], and Yao et al.^[113], respectively.

TMC quantum dots possess highly discrete electron and hole energy levels, resulting in more chemically active sites for the of polysulfide adsorption and Li⁺ anchoring^[114]. Dispersing TMCs on conductive substrates with high specific surface area ensures maximal uniform dispersion of the quantum dots, and facilitates comprehensive adsorption of LiPSs and further homogenizing Li⁺ flux. Lithiophilic and sulfiphilic Mo₂N, Mo₂S, and VN quantum dots have been demonstrated to effectively suppress the shuttle effect and dendrite growth when modifying NG^[115-117]. The porous NG framework serves as a 3D electronic channel and spatial reservoir for LiPSs adsorption, which not only physically confines sulfur species but also accelerates electron transfer during redox reactions and prolongs the nucleation time of lithium dendrites. Mo₂N quantum dots@NG exhibit higher Li⁺ conductivity [Figure 11A] and excellent cycling stability at 2C^[115]. In Figure 11B, Mo₂S quantum dots@nitrogen-doping reduced graphene oxides (NRGO), featuring abundant 10 nm mesopores and a high specific surface area of up to 69.18 m² g^-1, facilitates the full exposure of Mo₂S catalytic sites^[116]. Mo₂S@NRGO possesses a lower Gibbs free energy (0.016 eV) and the free energy required for the decomposition of Li₂S* (1.06 eV), which accelerate the liquid-solid conversion reaction from Li₂S₄* to Li₂S₂* and the solid-solid decomposition reaction of Li₂S, respectively. Similarly, VN quantum dots@NG, with a Li₂S dissociation energy barrier of only 0.31 eV on the VN (111) surface and the conductive network provided by the NG framework [Figure 11C], effectively achieve 1,000 h of dendrite-free stable cycling^[117].

Figure 11. Bifunctional separators with TMBs quantum dots for Li-S batteries. (A) Lithium-ion conductivity and long cycle stability at 2C of the cells with Mo₂N@NG/PP modified separators. (B) TEM, HRTEM images and LiPSs reduction energy profiles of MoS₂ QDs-NRGO. (C) Energy profiles of Li₂S dissociation process on the surface of VN (111) and voltage-time curves of Li-Li symmetric batteries with VN@NC separators. This figure is quoted with permission from Ma et al.^[115], Zhang et al.^[116], Ma et al.^[117], respectively.

Furthermore, benefiting from the uniform pore sizes, MOF/COF materials can significantly contribute to the consistent Li⁺ transport and preventing the formation of lithium dendrites^[44,118,119]. The incorporation of polymer materials in functional separators can positively impact the electrolyte affinity of the separator and assist in the protection of the metal anode^[120].

According to the summary of performance of Li-S batteries with bifunctional separators in Table 2, compared to double-sided modification, bifunctional interlayers loaded with sulfiphilic and lithiophilic materials on a single-sided separator offer greater advantages in increasing cathode sulfur loading, enhancing the energy density and safety of Li-S batteries, simplifying the process, and reducing costs. This approach represents the main strategy for bifunctional separators in current Li-S battery technology.