PFAS-free quasi-solid polymer electrolytes with enhanced performance for sustainable lithium-ion batteries

Abstract

Quasi-solid polymer electrolytes (QSPEs) integrate electrolyte and separator functions, offering mechanical flexibility, favorable ionic conductivity, and enhanced safety compared with liquid electrolytes. Most prior studies rely on PVDF-based systems that use PFAS-containing polymers and toxic solvents, raising environmental concerns. Here, we report PFAS-free QSPEs prepared from poly(ether sulfone) (PES) and poly(ethylene oxide) (PEO) using the green solvent dimethyl sulfoxide (DMSO). PES provides excellent thermal stability, robust mechanical properties, and an amorphous framework that disrupts PEO crystallinity and induces a porous structure, enabling continuous ion-conduction pathways. The resulting PES/PEO QSPEs exhibit an electrochemical stability window up to ∼4.23 V and an ionic conductivity of 6.05 × 10⁻⁴ S cm⁻¹, and stable Li stripping/plating behavior with low polarization. In Li|LiFePO₄ cells, they exhibit good capacity retention over multiple cycles, outperforming PES-only QSPEs and commercial separator Celgard 2325. Compared with PVDF-based systems, PES/PEO QSPEs deliver comparable electrochemical performance with superior thermal tolerance and the added benefit of being PFAS-free, providing a sustainable pathway toward safe and durable next-generation lithium-ion batteries.

---

Green foundation

1. This work advances green chemistry by eliminating PFAS-containing fluoropolymers from quasi-solid polymer electrolytes (QSPEs) while maintaining electrochemical performance required for lithium-ion batteries, addressing growing environmental and regulatory concerns associated with persistent fluorinated materials.

2. A PFAS-free QSPE based on poly(ether sulfone) and poly(ethylene oxide) is fabricated using dimethyl sulfoxide as a green solvent. The electrolyte delivers an ionic conductivity of ∼2.0 mS cm⁻¹, electrochemical stability up to ∼4.2 V, and stable cycling performance comparable to PVDF-based systems, with improved thermal robustness and reduced environmental persistence.

3. Further research could focus on replacing petroleum-derived polymers with bio-based alternatives, lowering lithium salt content, and developing solvent-free or energy-efficient processing routes to further reduce the environmental footprint.

Introduction

Lithium-ion batteries (LIBs) have become the dominant energy storage technology due to their high energy density, long cycle life, and wide applicability in portable electronics, electric vehicles, and renewable energy systems.¹ Nevertheless, conventional LIBs employing liquid electrolytes encounter critical safety and performance concerns, including leakage, flammability, and chemical instability, while traditional separators may further restrict electrochemical stability and cycling durability.² To address these challenges, the development of solid-state and quasi-solid-state electrolytes has attracted considerable attention. Such electrolytes can simultaneously act as ion-conducting media and separators, thereby improving safety, mechanical robustness, and long-term cycling performance. Among them, quasi-solid polymer electrolytes (QSPEs) have emerged as a particularly promising option, offering flexibility, ease of fabrication, and multifunctionality for next-generation LIBs.^3–5

Polymer electrolytes are generally classified into three types: gel polymer electrolytes (GPEs), solid polymer electrolytes (SPEs), and composite polymer electrolytes (CPEs).^6,7 GPEs are typically fabricated by incorporating liquid electrolytes as plasticizers into a polymer matrix, which enables high ionic conductivity but results in poor mechanical strength.^8,9 Due to their inability to withstand electrode–electrolyte interfacial stress, an additional separator is often required, limiting their practical application. SPEs, on the other hand, are solvent-free and provide superior safety and mechanical stability, but their room-temperature ionic conductivity is usually insufficient because of restricted polymer chain mobility. CPEs attempt to address these limitations by introducing inorganic fillers, which can improve both conductivity and mechanical properties, though their performance strongly depends on the choice and dispersion of fillers.^10,11

Distinct from these systems, QSPEs integrate the dual functions of electrolyte and separator into a single material. They are often regarded as polymer-based separator membranes due to their structural role. However, unlike conventional separators, QSPEs exhibit high porosity and excellent wettability, and can provide ionic conductivity, thereby enabling more compact and efficient battery architectures.^3–5 In the literature, QSPEs are frequently classified together with GPEs owing to the overlapping definitions, although their unique design emphasizes limited solubility in liquid electrolytes, which allows the polymer framework to remain intact while effectively immobilizing and adsorbing lithium salts. This structure not only enhances ionic transport but also ensures mechanical integrity, eliminating the need for an external separator. Consequently, QSPEs simultaneously suppress electrolyte leakage and strengthen interfacial stability, providing a balanced combination of conductivity, flexibility, and safety. These advantages make QSPEs highly attractive for the development of safe and durable LIBs, bridging the gap between conventional GPEs and SPEs.

High porosity is a key factor in polymer electrolytes for improving ionic conductivity and interfacial compatibility with electrodes. By enabling sufficient electrolyte infiltration into the pores, QSPEs effectively transform solid/solid interfaces into solid/liquid interfaces, which is particularly advantageous.¹² However, excessive porosity may compromise mechanical strength, negatively affecting battery performance. Common QSPE preparation methods include in situ formation of porous networks, electrospinning, and phase inversion methods.^13,14 Among these, the phase inversion method is relatively simple, does not require specialized instruments, and has been widely applied for fabricating porous membranes for water purification, gas separation, and energy-harvesting devices. Its simplicity and scalability make it one of the most promising techniques for industrial production and commercial manufacturing of polymer membranes.¹⁵ Within phase inversion techniques, the most commonly used approach is non-solvent-induced phase separation (NIPS). In NIPS, while the solvents are generally hazardous or environmentally unfriendly, the choice of non-solvent critically determines membrane properties such as porosity, pore-size distribution, and mechanical strength.

Widely studied polymers for phase inversion–based QSPEs mainly include poly(vinylidene fluoride) (PVDF) and its copolymer poly(vinylidene fluoride-co-hexafluoropropylene) (PVDF-HFP).^16–22 PVDF-based systems are well known for their high dielectric constant, chemical stability, and thermal resistance, making them widely used as separators for next-generation lithium-based batteries.¹⁶ For example, Gozdz et al. successfully fabricated QSPEs based on PVDF-HFP via phase inversion, which subsequently became the first commercialized PVDF-based QSPE product and has been extensively applied in polymer lithium-ion batteries.¹⁷ Compared with PVDF, PVDF-HFP exhibits lower crystallinity, swells in organic solvents such as ethylene carbonate (EC) and propylene carbonate (PC), and maintains good mechanical strength. Various inorganic fillers, including Al₂O₃, LiAlO₂, MgO, SiO₂, TiO₂, ZrO₂, and organic montmorillonite, have been incorporated into PVDF-HFP to enhance both ionic conductivity and mechanical performance.¹⁸ Alternatively, PVDF-HFP can be physically blended or chemically copolymerized with other polymers to further improve its applicability in batteries.¹⁹ For example, a poly(propylene carbonate) (PPC)/PVDF-HFP blend membrane, after immersion in 1 M LiPF₆ in DEC/EC, forms a QSPE that delivers an ionic conductivity of 1.1 × 10⁻³ S cm⁻¹, a lithium ion transference number of 0.47, and a capacity of 155 mAh g⁻¹ with 89% retention after 100 cycles at 0.2 C.²⁰ Most QSPEs prepared from PVDF-HFP exhibit similarly high ionic conductivity, along with excellent electrochemical stability and good cycling life.^21,22

Nevertheless, PVDF-based systems and their copolymers, such as PVDF-HFP, often incorporate per- and polyfluoroalkyl substances (PFAS)-containing components that may release PFAS during their lifecycle. PFAS have been widely recognized as emerging contaminants due to their exceptional environmental persistence, bioaccumulation potential, and resistance to chemical and biological degradation.^23,24 Owing to the strength of the C–F bond, PFAS are often referred to as “forever chemicals”, as they can accumulate in water, soil, wildlife, and human tissues over extended periods. Epidemiological and toxicological studies have associated PFAS exposure with adverse health effects, including developmental toxicity, immune suppression, endocrine disruption, and increased cancer risk.²⁵ Furthermore, PFAS contamination has been detected in drinking water sources worldwide, leading to increasing regulatory restrictions and proposed phase-outs in regions such as the European Union and the United States.²⁶ From a green chemistry perspective, minimizing the use of fluorinated polymers that may contribute to PFAS release during production, operation, or disposal is therefore highly desirable.²⁷ Consequently, despite the favorable electrochemical performance and mechanical robustness of PVDF-HFP, its potential environmental and health impacts represent a significant barrier to the sustainable advancement of polymer electrolyte technologies.

To overcome the environmental and health concerns associated with fluorinated polymers, recent studies have increasingly focused on developing PFAS-free or fluorine-free QSPEs with improved electrochemical stability and environmental sustainability.^28–32 For example, Gao et al. reported a polypropylene (PP) fiber-reinforced QSPE based on poly(ethylene glycol) (PEG) copolymer soaked with a LiTFSI/DOL/DME electrolyte, which enabled stable lithium-metal batteries with improved interfacial stability and suppressed dendrite growth.²⁸ Xu et al. further developed a fluorine-free QSPE via in situ radical copolymerization, enabling the formation of a lithium-oxide-rich solid electrolyte interphase and significantly improving the stability of Li metal batteries.²⁹ In addition, Chen et al. reported a composite QSPE comprising La@ZIF-8/SiO₂/PAN, fabricated by embedding SiO₂ nanoparticles into polyacrylonitrile (PAN) fiber membranes followed by surface coating with a La@ZIF-8 metal–organic framework (MOF), resulting in enhanced ionic transport and interfacial stability in lithium-metal batteries.³⁰

Among PFAS-free polymers, poly(ether sulfone) (PES) has attracted considerable attention as a sustainable alternative for electrolyte and separator materials in QSPE systems. PES exhibits excellent chemical resistance, flame retardancy, non-toxicity, and mechanical robustness, and has been widely used in water purification and biomedical applications. These characteristics are comparable to those of poly(vinylidene fluoride) (PVDF), which has traditionally dominated commercial separator technologies.³³ Unlike PVDF, PES is an amorphous polymer with a high glass transition temperature (∼225 °C), providing superior thermal stability. However, its relatively low dielectric constant has historically limited its application in battery electrolyte systems.

In the context of PES-based systems, Cui et al. fabricated a PES/PVDF composite fibrous membrane via electrospinning as a QSPE for lithium-ion batteries. The incorporation of PES reduced the crystallinity of PVDF and enhanced wettability, resulting in high ionic conductivity (∼1.69 × 10^–3 S cm⁻¹) and improved cycling stability and rate performance compared with polyolefin separators.³⁴ Liu et al. prepared PES-based QSPEs for Li–O₂ batteries via the NIPS method, where Al₂O₃ nanoparticle doping enhanced mechanical and thermal stability. After soaking in ionic liquid electrolytes, the membrane enabled stable cycling at 1000 mAh g⁻¹ for 52 cycles.³⁵ Lo et al. further demonstrated that incorporating glycerol into PES-based membranes improved the cycle performance of lithium-ion batteries, achieving ∼90% capacity retention after 100 charge–discharge cycles at 0.2 C.³⁶ Taken together, these studies highlight the growing interest in PFAS-free and environmentally benign polymer electrolyte systems and underscore the potential of PES-based materials for sustainable QSPE development.

Poly(ethylene oxide) (PEO) is another widely used polymer electrolyte matrix owing to its strong coordination ability with lithium ions, good flexibility, chemical stability, and compatibility with various lithium salts, making it highly effective for polymer-based electrolyte systems. In addition, PEO is generally regarded as a relatively green polymer due to its low toxicity and environmental friendliness. However, PEO-based electrolytes fabricated via conventional solution-casting methods typically form dense and highly crystalline films, which significantly restrict ionic conductivity.³⁷ Furthermore, the high crystallinity and limited mechanical strength of PEO make it difficult to fabricate porous membranes using the NIPS method. Introducing a porous structure can disrupt the dense packing of polymer chains, reduce crystallinity, and thereby facilitate ionic transport.

To address these limitations, PEO was blended with PFAS-free PES in this study, and porous PES/PEO membranes were successfully fabricated via the NIPS method. This strategy provides a sustainable pathway for the development of high-performance polymer electrolytes with improved ionic transport and structural stability. LiPF₆ was employed as the lithium salt due to its balanced electrochemical performance, including high ionic conductivity and stable electrode–electrolyte interfacial compatibility, as well as its widespread use in commercial lithium-ion batteries, ensuring practical relevance and comparability with existing electrolyte systems. The incorporation of PES reduces PEO crystallinity, enhances mechanical and thermal stability, and facilitates the formation of robust, porous QSPE membranes. DMSO, a green, non-protic solvent, was used to dissolve both PES and PEO, while water served as the environmentally benign non-solvent in the NIPS process. During the phase inversion process, partial dissolution of PEO increases the porosity of the PES/PEO membranes. The enhanced pore volume allows greater electrolyte uptake, which in turn facilitates ionic transport and improves overall ionic conductivity. While the remaining PEO can coordinate with lithium salts, the primary factor governing the increased conductivity is the higher porosity rather than PEO content. These effects collectively contribute to improved electrochemical performance in battery applications. Compared with PVDF- and PVDF-HFP-based QSPEs reported by others, our prepared PES/PEO membranes demonstrate comparable performance. Their PFAS-free composition, tunable porosity, and straightforward NIPS fabrication make them a sustainable and practical platform for high-performance QSPEs for lithium-ion batteries. This strategy employs environmentally friendly materials, including green PEO, PFAS-free PES, DMSO, and water, to produce QSPEs with enhanced ionic conductivity, mechanical strength, thermal stability, and electrochemical stability, highlighting the advantages of our membrane design for safe and durable LIB. To the best of our knowledge, this study pioneers the development of QSPEs derived from PES/PEO membranes, establishing a new materials platform for safe and sustainable lithium-ion batteries.

Material section

Materials

Poly(ether sulfone) (PES, 6020P, CAS No. 25608-63-3) was obtained from BASF. Poly(ethylene oxide) (PEO, M_w ≈ 35 000 g mol⁻¹, CAS No. 25322-68-3) was purchased from Alfa Aesar. Dimethyl sulfoxide (DMSO, analytical grade, CAS No. 67-68-5) was supplied by Merck. Distilled–deionized (DI) water (CAS No. 7732-18-5) was used as the nonsolvent. A 1 M solution of lithium hexafluorophosphate (LiPF₆, CAS No. 21324-40-3) in ethylene carbonate (EC, CAS No. 96-49-1)/dimethyl carbonate (DMC, CAS No. 616-38-6) (1 : 1 v/v) was purchased from Sigma-Aldrich. Lithium iron phosphate (LiFePO₄, model LFP-S01, CAS No. 15365-14-7) was supplied by Anatech Co., Ltd, Taiwan. All chemicals were used as received without further purification.

Membrane fabrication

PES and PEO were weighed at the desired weight ratios and dissolved in DMSO to prepare 15 mL of casting solution. The DMSO content in the casting solution was optimized through viscosity control to ensure stable membrane fabrication. A viscosity range of approximately 3000–9000 cP was found to be suitable for uniform blade casting. When the viscosity was too low, excessive spreading of the solution occurred on the substrate, resulting in ultrathin films with insufficient mechanical strength and structural integrity. In contrast, excessively high viscosity significantly increased shear stress during blade casting, leading to film tearing, delamination, and surface irregularities after phase inversion, thereby compromising mechanical properties. Accordingly, the final casting formulation was determined by achieving a balanced viscosity that enabled stable film formation, smooth surface morphology, and reproducible membrane thickness.

The homogeneous solution was then placed in a rotary oven at 80 °C for 48 h to ensure complete dissolution. After degassing at 80 °C, the casting solution was cooled to room temperature and uniformly spread on a glass plate using a 150 µm casting knife. The coated plate was immediately immersed in deionized water as a non-solvent to induce phase inversion at room temperature (25 °C). This temperature is widely adopted for PES-based membrane fabrication and provides stable and reproducible phase separation behavior. At 25 °C, solvent–non-solvent exchange occurs at a moderate rate, enabling uniform pore formation and smooth membrane detachment from the glass substrate. After immersion, membrane formation occurred rapidly and the film detached from the substrate within approximately one minute. The membranes were then washed in fresh deionized water at room temperature for two days, with the water renewed every 24 hours, to remove residual solvent and any loosely bound PEO. Finally, the membranes were dried in an oven at 40 °C for 2 days prior to characterization. Residual solvent removal was confirmed by TGA analysis.

Quasi-solid polymer electrolyte preparation

The dried porous membranes were soaked in 1 M LiPF₆ dissolved in EC : DMC (1 : 1 by volume) to form quasi-solid polymer electrolytes (QSPEs) in an argon-filled glove box (O₂ < 0.1 ppm, H₂O < 0.1 ppm).

Membrane characterization

The composition of the PES/PEO membranes was determined by ¹H nuclear magnetic resonance (NMR, Bruker AV-600). Approximately 10–15 mg of each membrane was dried in a vacuum oven at 50 °C for 4 h to remove residual moisture. The dried sample was dissolved in DMSO-d₆ at room temperature until complete dissolution. The solution was then transferred into an NMR tube using a glass pipette. Signal intensity was plotted against chemical shift, and the area under each peak was integrated to quantify the residual PES and PEO in the membrane. The residual fraction (wt%) was calculated using eqn (1):where A_a is the integrated peak area of PEO, A_p is the integrated peak area of PES, M_a is the molecular weight of the repeating unit of PEO (44), M_p is the molecular weight of the repeating unit of PES (232), n_a is the number of protons per repeating unit of PEO (4), and n_p is the number of protons per repeating unit of PES (8).

The morphology of the PES/PEO membranes was examined by field emission scanning electron microscopy (FESEM, Zeiss, Sigma 300). Prior to SEM analysis, the membranes were dried to remove residual moisture and then directly coated with a thin layer of platinum.

The porosity of the PES/PEO membranes was determined by cutting dried specimens (1.0 cm × 1.0 cm), measuring thickness with a micrometer to calculate the volume (V_t), and recording the dry mass (M_d) on an analytical balance. The samples were then fully immersed in DI water and left to allow complete penetration into the membrane pores. After immersion, excess surface water was gently removed with filter paper, and the membranes were immediately weighed again to obtain the wet mass (M_w). The porosity (ε) of the membranes was calculated using eqn (2):³⁸

where M_w is the wet mass of the membrane, M_d is the dry mass, V_t is the volume of the membrane, and ρ_w is the density of water.

The thermal degradation behavior of the PES/PEO membranes was analyzed by thermogravimetric analysis (TGA, Hi-Res TA 2950) under a nitrogen atmosphere at a heating rate of 10 °C min⁻¹ from room temperature to 800 °C.

The thermal characterization of the PES/PEO membranes was investigated using a Diamond Differential Scanning Calorimeter (DSC, PerkinElmer Pyris Diamond). Samples weighing 5–10 mg and sealed in aluminum cans were subjected to heating from 30 °C to 200 °C at a rate of 10 °C min⁻¹ under a nitrogen atmosphere.

The mechanical properties (ultimate tensile strength and elongation at break) of the PES/PEO membranes were evaluated using a universal testing machine (Shimadzu AGS-J) equipped with a 100 N load cell and a stretching speed of 5 mm min⁻¹. Membranes were cut into standard dumbbell-shaped specimens according to ISO 527-3. Prior to testing, the samples were dried in an oven at 50 °C to remove moisture.

The electrolyte uptake of the PES/PEO membranes was quantified using the electrolyte uptake ratio, calculated according to the following eqn (3):

where M_d and M_w are the weights of the membrane in the dry state and after electrolyte absorption, respectively, both measured inside a glove box using an electronic balance.

Electrochemical measurement

Coin cells with CR2032-type were utilized for electrochemical characterizations throughout the entire study. The ionic conductivity (σ) was evaluated via electrochemical impedance spectroscopy (EIS) in the frequency ranging from 1 Hz to 1 M Hz and the applied potential of 10 mV, using an electrochemical workstation (CHI 6081E). The membranes were soaked in the liquid electrolyte and sandwiched between two identical stainless steel (SS) electrodes in a SS|SS symmetrical cell. The ionic conductivity was calculated according to the eqn (4):where A (cm²) is the effective contact area between the membrane and electrodes. L (cm) is the thickness of membrane. R_b is the bulk resistance recorded from the EIS.

The lithium-ion transference number (t_Li⁺) was obtained by the measurement of AC impedance and DC polarization. The membranes were soaked in the liquid electrolyte and sandwiched between two identical lithium foil electrodes in a Li|Li symmetrical cell. It is calculated as the following eqn (5):³⁹

where the ΔV is supposed to a stationary value of 10 mV, I₀ and I_s measured by DC polarization measurements is the initial and steady current respectively, R₀ and R_s observed from DC polarization interpreted as the initial and steady interfacial resistance of electrolytes, respectively. Each measurement was repeated at least three times, and the reported values include error bars representing the standard deviation of repeated measurements.

The electrochemical stability window of the membrane soaked in the liquid electrolyte and sandwiched between a SS and a lithium foil (SS|Li) cell was detected by the linear sweep voltammetry (LSV) method in which the voltage increasing from 0 V to 6 V with a scan rate of 1 MV s⁻¹ was used.

The lithium stripping/deposition was evaluated using a symmetric Li|Li symmetrical cell. The tests were carried out at a current density of 0.5 mA cm⁻², calculated on the basis of the active area of the lithium disks (diameter 15.88 mm, ≈1.98 cm², corresponding to an applied current of ≈1.01 mA). Galvanostatic charge–discharge cycling was performed within a safety voltage window of ±2.0 V.

The cycling performance and rate capability of a Li|LiFePO₄ cell were evaluated, in which lithium foil served as the anode and LiFePO₄ (theoretical specific capacity 170 mAh g⁻¹) as the cathode. The in-house-fabricated PES/PEO membrane was thoroughly dried and punched into circular disks (≈19 mm diameter). Prior to assembly, the membranes were soaked in liquid electrolyte to ensure complete wetting. The total mass of the LiFePO₄ electrode was measured, and the active-material loading was used to calculate the applied current density and corresponding charge–discharge time. Galvanostatic cycling and rate-capability tests were conducted within a voltage window of 2.5–3.8 V. All cells were assembled in an argon-filled glove box with O₂ and H₂O < 0.1 ppm.

Results and discussion

Composition and structure

Prior to selecting the final compositions, a series of gradient PES/PEO ratios were evaluated to optimize membrane porosity, electrolyte uptake, and film-forming properties. The PES content was fixed at 14 wt% because this concentration provided an appropriate solution viscosity for blade casting and ensured uniform film thickness. Membranes with relatively low PEO content exhibited limited porosity and reduced electrolyte uptake. Increasing the PEO fraction promoted phase separation and pore formation, thereby enhancing electrolyte absorption. However, excessive PEO content significantly increased the solid content and viscosity of the casting solution, hindering smooth casting and resulting in poor mechanical integrity. Considering porosity, electrolyte uptake, processability, and mechanical stability, PES/PEO (14/10) and (14/18) were ultimately selected for detailed investigation.

The membrane structure is strongly governed by the casting solution composition, viscosity, and phase inversion conditions. During the NIPS process, the exchange rate between solvent (DMSO) and non-solvent (water) determines the kinetics of phase separation, which subsequently controls pore morphology and macrovoid formation. The viscosity of the casting solution plays a critical role in modulating solvent–non-solvent diffusion. Solutions with lower viscosity allow rapid solvent exchange, promoting instantaneous demixing and the formation of large macrovoids, whereas higher viscosity slows mass transfer, resulting in delayed demixing and a more compact pore structure. Therefore, tuning the PES/PEO ratio and overall solid content directly influences membrane morphology.

In the present study, PES/PEO membranes were fabricated via the NIPS method. While PES readily forms porous structures through NIPS, PEO alone cannot be processed into self-supporting membranes using this technique. PES therefore served as a structural framework to enable the formation of porous PES/PEO composite membranes. Because water was employed as the non-solvent and PEO is partially water-soluble, partial dissolution of PEO inevitably occurred during phase inversion and subsequent washing. Consequently, both washing duration and PEO molecular weight critically influence the final membrane composition and structural stability. To ensure reproducibility, membranes were washed in distilled water at room temperature under agitation for two days. Shorter washing durations resulted in incomplete removal of loosely bound or low-molecular-weight PEO, which could subsequently leach out during electrolyte soaking and lead to variability in electrochemical performance. Importantly, extending the washing period to one week or even one month did not produce additional mass loss, confirming that two days is sufficient to remove excess PEO while retaining structurally integrated polymer chains within the membrane matrix.

Regarding molecular weight selection, low-molecular-weight PEO was largely removed during washing, whereas excessively high molecular weight PEO (>50 000 g mol⁻¹) significantly increased solution viscosity and hindered uniform film formation. Considering structural stability, processability, and ionic transport requirements, PEO with a molecular weight of 35 000 g mol⁻¹ was selected as an optimal balance for membrane fabrication.

Fig. 1 shows the composition and structural characterization of the prepared PES, PES/PEO (14/10), and PES/PEO (14/18) membranes: (a) ¹H NMR spectra; SEM images of the (b) top surfaces, (c) bottom surfaces, and (d) cross-sections. As shown in Fig. 1(a), the characteristic peaks of PES were observed at 7.26 and 7.97 ppm, corresponding to the aromatic protons, while the resonance peak of PEO appeared at 3.50 ppm. By integrating the respective peak areas and substituting the values into eqn (1), the relative PEO content in the membranes was quantified. The summarized results are shown in Table 1. Interestingly, regardless of the initial PEO loading (41.7 or 56.3 wt%), the residual PEO content in the membranes consistently stabilized at approximately 23 wt%. Extending the water-washing period up to two weeks did not further decrease the PEO content, indicating that ∼23 wt% represents the stable residual fraction of PEO that remains after phase inversion. This residual PEO fraction was further corroborated by TGA analysis, which consistently confirmed the presence of ∼23 wt% PEO in the membranes.

Fig. 1 Composition and structural characterization of the prepared PES, PES/PEO (14/10), and PES/PEO (14/18) membranes: (a) ¹H NMR spectra; SEM images of (b) top surfaces, (c) bottom surfaces, and (d) cross-sections.

Table 1 Composition and PEO residual ratio determined by NMR

Sample code	PES (wt%)	PEO (wt%)	DMSO (wt%)	Calculated PEO fraction (wt%)	NMR-derived PEO residual fraction (wt%)
PES	14	0	86	0.0	0.0
PES/PEO (14/10)	14	10	76	41.7	22.8
PES/PEO (14/18)	14	18	68	56.3	23.7

---

This phenomenon suggests that a portion of PEO strongly interacts with PES, consistent with previous reports of good miscibility between the two polymers.⁴⁰ Such interactions likely account for the stabilization of residual PEO within the PES matrix, preventing its complete removal by water. Therefore, in this system, the maximum residual PEO content achievable is limited to approximately 23 wt%.

Table 2 presents the thickness and porosity values of PES, PES/PEO (14/10), and PES/PEO (14/18) membranes. The results show that the addition of PEO led to an increase in membrane thickness. This can be attributed to the fact that the PES content was kept constant across all samples, while the incorporation of PEO increased the total polymer mass, thereby producing thicker membranes.

Table 2 Thickness and porosity values of PES, PES/PEO (14/10), and PES/PEO (14/18) membranes

Sample code	Thickness (um)	Porosity (%)
PES	67	23.04
PES/PEO (14/10)	108	49.66
PES/PEO (14/18)	106	52.60

---

Regarding porosity, the phase inversion process inherently creates pores due to solvent and non-solvent exchange. When PEO was introduced, a significant portion of it was dissolved and washed out during water treatment, leaving behind additional voids in the membrane. As a result, the porosity increased markedly from 23.04% for pristine PES to 49.66% for PES/PEO (14/10). Although the residual PEO content was similar in both PES/PEO membranes (as confirmed by NMR), the higher initial loading in PES/PEO (14/18) meant that a greater fraction of PEO was removed during phase inversion, which further contributed to a slight increase in porosity to 52.60%.

The SEM images of the top surfaces, bottom surfaces, and cross-sections of the PES, PES/PEO (14/10), and PES/PEO (14/18) membranes are shown in Fig. 1(b)–(d). From Fig. 1(b), the top surfaces of both PES and PES/PEO membranes appear dense. This dense skin layer is formed because DMSO and water are highly miscible; when the polymer solution is immersed in water, non-solvent molecules rapidly penetrate the surface while the solvent diffuses outward. The imbalance in mass-transfer rates causes a sharp increase in local polymer concentration at the interface, leading to instantaneous solidification into a polymer-rich phase and formation of a compact skin layer.⁴¹

As shown in Fig. 1(c), the bottom surfaces exhibit more porous features compared with the dense top surfaces. This difference arises because solvent/non-solvent exchange is slower on the bottom side, allowing more time for polymer chains to rearrange and resulting in a less compact morphology.⁴² The PES/PEO membranes display more pores than pristine PES, consistent with partial leaching of PEO during phase inversion, which enhances porosity. The cross-sectional images in Fig. 1(d) clearly show a thin skin layer (<100 nm) at the top surface, supported by a thick porous substructure (100–200 μm), which constitutes the main body of the membrane. Within the porous region, prominent finger-like macrovoids are observed. These elongated voids originate from the large difference in solvent/non-solvent diffusion rates, which induces rapid phase separation (instantaneous demixing) and creates vertical channels typical of PES-based membranes.⁴³ A magnified view of the macrovoid interior further reveals a bi-continuous structure characterized by interconnected pores forming numerous transport pathways throughout the membrane (see Fig. S1). This interpenetrating morphology is beneficial for applications requiring high permeability.³⁶ Overall, incorporation of PEO does not significantly alter the fundamental membrane structure, although the number of pores increases due to partial leaching of PEO during water exchange.

Thermal and mechanical properties

Fig. 2 shows the thermal and mechanical properties of the prepared membranes: (a) TGA curves; (b) DSC thermograms; (c) stress–strain curves; and (d) ultimate tensile strength and elongation at break derived from (c). As shown in Fig. 2(a), the onset thermal decomposition temperatures are indicated by single arrows. The pristine PES membrane exhibits an onset decomposition temperature of approximately 476 °C. In contrast, the PES/PEO membranes display two distinct weight-loss stages. The first, occurring at a lower temperature (∼375 °C), corresponds to the decomposition of PEO, while the second, around 476 °C, corresponds to the decomposition of PES. Notably, the weight loss between the first and second stages is approximately 22.5 wt%, which closely matches the PEO content determined by NMR. These results confirm that the residual PEO content in both PES/PEO (14/10) and PES/PEO (14/18) membranes is similar, approximately 22–23 wt%. Moreover, no significant weight loss is observed between 100–200 °C, indicating the absence of residual volatile solvents. To further verify this, NMR analysis was conducted using DMF-d7 as the solvent (see Fig. S2 in the SI). The characteristic residual signal of DMSO at 2.50 ppm is not observed, confirming the absence of residual DMSO in the membranes. Meanwhile, the expected residual signals of DMF-d7 at ∼2.75 and ∼2.92 ppm (N–CH₃), as well as ∼8.0 ppm (–CHO), are clearly observed. These results collectively demonstrate that the membranes are thoroughly dried and free of solvent residues, ensuring that subsequent QSPE fabrication and electrochemical testing are not affected.

Fig. 2 Thermal and mechanical properties of the prepared membranes: (a) TGA curves; (b) DSC thermograms; (c) stress–strain curves; and (d) ultimate tensile strength and elongation at break derived from (c).

Fig. 2(b) shows the first heating scans of PES, PES/PEO (14/10), and PES/PEO (14/18) membranes obtained by DSC. DSC is commonly employed to analyze the thermal behavior of polymers, including glass transition, melting, and crystallinity. In this work, DSC was mainly used to examine whether crystalline PEO domains remained in the PES/PEO membranes. As shown in the figure, no melting peak around 60 °C, corresponding to crystalline PEO, was detected. This result indicates that the incorporated 22–23 wt% PEO exists predominantly in an amorphous state. The absence of PEO crystallinity is beneficial for lithium-ion transport, as amorphous PEO provides continuous ion-conduction pathways, which in turn enhances the overall electrochemical performance of the battery. In addition, as evidenced by the DSC analysis, no thermal transition peaks were detected up to 200 °C, confirming the exceptional thermal stability of the amorphous PES/PEO framework within this operating range.

Stress–strain curves of PES, PES/PEO (14/10), and PES/PEO (14/18) membranes are presented in Fig. 2(c), while Fig. 2(d) shows the ultimate strength and elongation at break of each membrane. The ultimate strength, defined as the maximum stress in the stress–strain curves, and the elongation at break, defined as the maximum strain, were extracted from Fig. 2(c). The PES membrane exhibits a high ultimate strength of 7.3 MPa, which decreases to 2.7 MPa upon incorporation of PEO. PEO typically acts as a plasticizer due to its low T_g, and its addition reduces the stiffness of the high-hardness PES matrix. Furthermore, as observed in the porosity analysis, membranes with higher initial PEO content lose more PEO during washing, generating additional voids that increase porosity and reduce the load-bearing area, which contributes to the further decrease in ultimate strength.

On the other hand, the PES membrane shows a relatively low elongation at break of 5.7%, whereas the addition of PEO increases it to approximately 12%. This is due to the enhanced polymer chain mobility and the additional free volume provided by PEO, which allow the membrane to deform more before fracture.

Basic electrochemical properties

The thermal stability of the prepared PES and PES/PEO membranes was first evaluated and compared with that of the commercial Celgard 2325 separator. The basic electrochemical properties of the corresponding QSPEs and Celgard 2325 are summarized in Fig. 3(a–d), including (a) thermal stability of the membranes, (b) electrolyte uptake, (c) EIS spectra, and (d) Li⁺ transference number. To examine the intrinsic thermal dimensional stability prior to electrolyte incorporation, the as-prepared PES, PES/PEO membranes, and Celgard 2325 were subjected to heat treatment. Photographs of PES, PES/PEO (14/10), PES/PEO (14/18), and Celgard 2325 after thermal treatment are shown in Fig. 3(a), demonstrating their thermal stability. Each specimen was heated from 30 °C to 240 °C for 30 min, after which the morphological changes were observed and photographed. The C2325 separator, a trilayer microporous PP/PE/PP polymer, begins to exhibit shrinkage when heated to 130 °C, which is near the melting temperature of the PE layer. At 150 °C, partial melting of the PP layers and eventual loss of the separator structure are observed, indicating the thermal limit of this commercial material. In contrast, the prepared PES and PES/PEO membranes remain intact under the same conditions. PES, being amorphous, does not exhibit melting upon heating; only slight warping is observed at 160 °C, which is likely attributed to thickness non-uniformity, rather than intrinsic thermal failure. No significant dimensional change is detected. The glass transition temperature of PES is 225 °C, and the membrane begins to soften near 240 °C; however, this softening does not lead to collapse or loss of shape, and the membrane remains structurally stable.

Fig. 3 Basic electrochemical properties of the prepared QSPEs and Celgard 2325: (a) thermal stability after heat treatment; (b) electrolyte uptake; (c) EIS spectra; and (d) Li⁺ transference number.

For PES/PEO membranes, no obvious structural changes are observed at 30 °C and 100 °C, indicating that PEO remains amorphous within the matrix (its melting temperature is approximately 60 °C). This observation is consistent with DSC results, which show no melting peak (Fig. 2(b)). The membranes remain stable up to 150 °C, and softening begins only after prolonged heating at 160 °C. At 180 °C, the membranes become visibly more transparent while retaining their circular shape and dimensions. The softening observed between 160–180 °C is attributed to the glass transition of the PES/PEO blend, representing the T_g of the polymer mixture rather than that of the individual components. Notably, even above this T_g, the membranes remain solid, and their overall structure is preserved at temperatures exceeding 200 °C. At temperatures up to 240 °C, the PES/PEO membranes exhibit slight yellowing, which is attributed to the oxidation of PEO. Overall, the excellent thermal stability of PES and PES/PEO membranes at elevated temperatures can effectively prevent short circuits, thereby enhancing the thermal safety of batteries. Compared with commercial Celgard 2325 separators, which begin to melt and lose structural integrity at ∼150 °C (melting temperature of PP), our PES and PES/PEO membranes maintain their shape and functionality up to 200 °C, demonstrating a clear advantage for high-temperature battery applications. Moreover, even compared with PVDF and PVDF-HFP membranes, whose melting temperatures are typically below 175 °C,^19–22 our membranes still exhibit significantly superior thermal stability.

The as-prepared PES and PES/PEO membranes were soaked in a liquid electrolyte of 1 M LiPF₆ in EC/DMC to obtain PES and PES/PEO QSPEs. For comparison, a commercial separator (Celgard 2325) was also immersed in the same electrolyte. Fig. 3(b) shows the electrolyte uptake as a function of time for PES, PES/PEO (14/10), PES/PEO (14/18) QSPEs, and Celgard 2325. All samples exhibited a rapid electrolyte uptake within the first 5 min, followed by a gradual stabilization between 10 and 30 min, during which only minor variations were observed.

The commercial Celgard 2325 separator exhibited a limited electrolyte uptake of 50–80%, primarily due to the low affinity of its PP/PE/PP trilayer structure toward polar organic electrolytes. In contrast, the PES and PES/PEO membranes showed significantly higher electrolyte uptakes. PES reached a maximum of approximately 295%, while the incorporation of PEO further enhanced this value, with PES/PEO (14/10) and (14/18) achieving 360% and 395%, respectively, nearly five times that of Celgard 2325. Notably, the experimentally determined porosity of Celgard 2325 (∼40%) is consistent with the manufacturer's specification of 39%. As shown in Table 2, the porosity of Celgard 2325 is higher than that of PES (23%), yet the electrolyte uptake of PES is clearly superior. This discrepancy indicates that porosity alone does not govern uptake; rather, membrane wettability plays a critical role. The inherently higher wettability of PES facilitates more effective electrolyte penetration and retention, accounting for its greater uptake despite lower porosity. Literature reports on PVDF-based polymer electrolytes have shown electrolyte uptakes of 100–250%, suggesting that the absorbed electrolyte can reach half of the polymer matrix weight.^19,21,22,44 By comparison, our PES-based QSPEs achieve slightly higher uptakes (300–400%), reflecting their more hydrophilic nature and superior wettability. Accordingly, these membranes can be classified as QSPEs, whereas Celgard 2325 does not fall into this category.

A quantitative analysis further reveals that electrolyte uptake exhibits a strong positive linear correlation with membrane porosity (see Table 2), described by the regression eqn (6):

This high R² value indicates that porosity is the primary factor governing electrolyte absorption in PES/PEO QSPEs. Although PEO chains can coordinate with lithium salts and improve interfacial compatibility, the comparable PEO content in PES/PEO (14/10) and PES/PEO (14/18) suggests that the observed increase in uptake is predominantly due to the enhanced pore volume generated during phase inversion rather than chemical interactions.

Fig. 3(c) shows the EIS spectra of PES, PES/PEO (14/10), PES/PEO (14/18) QSPEs, and Celgard 2325 at room temperature. All membranes exhibit an inclined line, and the bulk resistance for ion transport (R_b) was obtained by fitting the high-frequency region using an R_b – CPE (Constant Phase Element) equivalent circuit to minimize the influence of cable, contact, and spurious resistances (the fitted EIS spectra and the corresponding equivalent circuit diagrams are shown in Fig. S3–S6 of the SI). Compared with simple x-axis interception, equivalent circuit fitting provides a more reliable estimation of the bulk resistance by separating interfacial and parasitic resistances. The ionic conductivity (σ) was then calculated using eqn (2). The resulting σ values were 3.22 × 10⁻⁴ S cm⁻¹, 5.02 × 10⁻⁴ S cm⁻¹, 6.05 × 10⁻⁴ S cm⁻¹, and 7.06 × 10⁻⁵ S cm⁻¹ for PES, PES/PEO (14/10), PES/PEO (14/18) QSPEs, and Celgard 2325, respectively. The ionic conductivity of Celgard 2325 is comparable to the value reported in the literature (4.33 × 10⁻⁵ S cm⁻¹) when immersed in 1 M LiTFSI EC/DMC electrolyte.¹⁹

We further examined the relationship between membrane porosity (Table 2) and ionic conductivity. A linear regression analysis yielded a strong correlation, described by the following eqn (7):

The high R² value indicates that the enhanced ionic transport is primarily governed by increased electrolyte absorption within the porous structure. Since the PEO content in the PES/PEO (14/10) and PES/PEO (14/18) membranes is identical, the observed difference in conductivity between these two samples is mainly attributed to their different porosities rather than chemical effects. Therefore, the improved ionic conductivity of the PES/PEO QSPEs is predominantly associated with the increased pore volume generated during phase inversion, which allows greater electrolyte uptake and facilitates ion transport.

Fig. 3(d) shows the Li⁺ transference number determination for PES, PES/PEO (14/10), PES/PEO (14/18) QSPEs, and Celgard 2325. The insets display the corresponding chronoamperometry curves and EIS fittings used to determine the bulk and interfacial resistances. Here, R₀ and R_s denote resistance parameters obtained from EIS fitting rather than sample names. The Li⁺ transference numbers were measured in Li|Li symmetric cells and are summarized in Table 3, yielding values of 0.33, 0.37, 0.42, and 0.26 for PES, PES/PEO (14/10), PES/PEO (14/18), and Celgard 2325, respectively. The higher Li⁺ transference numbers observed for the PES/PEO membranes can be attributed to retained PEO segments that coordinate with Li⁺; however, the dominant factor governing ionic transport is the increased membrane porosity and corresponding electrolyte uptake. The porous structure provides continuous pathways for ion migration, improving overall conductivity while maintaining comparable Li⁺ mobility. Consequently, the PES/PEO (14/18) membrane demonstrates the most efficient Li⁺ transport among the samples. In contrast, the commercial Celgard 2325 separator lacks polar functional groups, resulting in limited electrolyte uptake and poor interfacial compatibility.⁴⁵ This restricts ion transport and overall electrochemical performance, with a Li⁺ transference number of only 0.26, which is consistent with reported values for PE separators (∼0.28) in 1 M LiPF₆ EC/DMC electrolytes⁴⁶ and for Celgard 2325 (∼0.27) in 1 M LiTFSI EC/DMC electrolyte.¹⁹

Table 3 Currents and resistances before and after polarization Li|Li symmetric cells in Fig. 3(d)

Sample Code	I 0 (×10^–6 A)	I S (×10^–6 A)	R 0 (Ω)	R s (Ω)	t Li^+	Error bar (±)
PES	20.11	11.85	357	422	0.33	0.02
PES/PEO (14/10)	19.96	11.85	338	399	0.37	0.03
PES/PEO (14/18)	15.35	9.29	301	382	0.42	0.05
C2325	18.05	9.53	406	485	0.26	0.03

---

Electrochemical stability and interfacial properties

Fig. 4 presents the electrochemical stability and interfacial properties of the prepared QSPEs and Celgard 2325: (a) LSV curves and (b) Li stripping/plating behavior. From the LSV curves in Fig. 4(a), the arrows mark the onset potentials, which were determined using the standard method defined as the intersection between the flat baseline prior to the oxidation peak and the tangent to the rising current region. The resulting onset potentials are 4.15, 4.23, 3.99, and 4.01 V for PES, PES/PEO (14/10), PES/PEO (14/18), and Celgard 2325, respectively. Compared with Celgard 2325, the PES and PES/PEO (14/10) QSPEs exhibit enhanced electrochemical stability, which can be attributed to their bi-continuous structure with interconnected pores, as observed by SEM. These microchannels facilitate Li⁺ transport and prevent local electron accumulation, thereby reducing overpotential and mitigating oxidative decomposition. This observation is consistent with previous studies, which have shown that pore and channel engineering can extend the electrochemical stability window of polymer electrolytes by promoting efficient ion–electron coupling and suppressing parasitic reactions.⁴⁷ When the PEO fraction is further increased to PES/PEO (14/18), excessive pore formation and elevated electrolyte uptake expand the electrode–electrolyte interfacial area. Moreover, the accumulated and agglomerated electrolyte within these enlarged pores can hinder Li⁺ transport and induce local charge accumulation, thereby increasing susceptibility to oxidative decomposition. As a result, the PES/PEO (14/18) membrane exhibits the lowest oxidation onset potential among the tested membranes.⁴⁸

Fig. 4 Electrochemical stability and interfacial properties of the prepared QSPEs and Celgard 2325: (a) LSV curves and (b) Li stripping/plating behavior.

The PES/PEO (14/18) membrane was not evaluated in full-cell tests due to its relatively lower electrochemical stability compared with the other membranes. Its higher porosity reduces mechanical robustness and destabilizes the electrode–electrolyte interface, potentially leading to rapid degradation and poor reproducibility during cycling. Considering both electrochemical and mechanical factors, full-cell testing was therefore focused on the PES and PES/PEO (14/10) membranes, which offer a more balanced combination of oxidative stability, ionic conductivity, and structural integrity, enabling reliable and meaningful assessment of battery performance. Fig. 4(b) shows the Li stripping/plating tests (voltage profiles) of PES, PES/PEO (14/10) QSPEs, and Celgard 2325 in Li|Li symmetric cells at a current density of 0.5 mA cm⁻². The voltage profiles of all three systems are initially unstable but gradually improve over repeated cycles. After the initial activation cycles, the Li|Li symmetric cells exhibited stable voltage profiles, with polarization voltages of ∼58 mV for PES and ∼55 mV for PES/PEO (14/10) QSPEs after 300 h, which are markedly lower than that of Celgard 2325 (∼258 mV) (Fig. 4(b)). This result indicates that the as-prepared QSPEs facilitate the formation of robust SEI layers on Li metal, maintain a stable interface during continuous lithium plating/stripping, and effectively retard Li dendrite growth.⁴⁹

It is noteworthy that the PES QSPEs require ∼120 h to reach a stable polarization state (down to 60 mV), whereas PES/PEO (14/10) QSPEs stabilize within only ∼40 h at <60 mV. This accelerated stabilization is primarily attributed to the increased porosity of the PES/PEO membranes, which provides well-connected pathways for efficient ion transport and higher electrolyte uptake. The enhanced ionic conductivity and improved electrode–electrolyte contact facilitate faster SEI formation and more uniform ion flux at the interface, effectively suppressing dendrite growth and local overpotentials. Consequently, the PES/PEO QSPEs achieve more stable interfacial behavior within a shorter activation period compared to pristine PES membranes.

Practical battery performance

Fig. 5 displays the practical battery performance of the prepared QSPEs and Celgard 2325: (a) rate capability and (b) cycling performance. Based on Fig. 5(a), as expected, the discharge capacity decreases with increasing C-rate. Among the tested electrolytes, the PES/PEO (14/10) QSPE consistently delivers the highest capacities, reaching 143.1, 137.8, 132.5, 121.9, 100.5, and 84.6 mAh g⁻¹ at 0.05, 0.2, 0.5, 1.0, 2.0, and 5.0 C, respectively. When the rate returns to 0.05 C, the capacity recovers to 142.8 mAh g⁻¹, demonstrating its highly reversible nature. The observed trend in rate capability is consistent with previous findings,³⁶ confirming that rate performance is dictated by ionic conductivity and Li⁺ transport pathways, with the PES/PEO (14/10) QSPE exhibiting the best overall performance. Furthermore, at low C-rates (0.05 and 0.2 C), the discharge capacities of PES QSPEs (139.2 and 134.4 mAh g⁻¹) are comparable to those of PES/PEO (14/10) and slightly higher than that of Celgard 2325 (137.6 and 133.5 mAh g⁻¹). However, at higher C-rates (≥0.5 C), the PES QSPEs exhibit a more pronounced capacity decay, falling below that of Celgard 2325. This behavior can be ascribed to their relatively low porosity, which is the lowest among the tested membranes, limiting the formation of continuous ion-conduction pathways. At low C-rates, ion migration remains sufficient, but at high C-rates, the reduced porosity and limited pore connectivity hinder fast Li⁺ transport, leading to increased polarization and capacity loss.⁵⁰

Fig. 5 Practical battery performance of the prepared QSPEs and Celgard 2325: (a) rate capability and (b) cycling performance.

Fig. 5(b) presents the long-term cycling performance of Li|LiFePO₄ cells assembled with PES, PES/PEO (14/10) QSPEs, and commercial Celgard 2325 at 0.2 C under room temperature. The initial discharge capacities of PES, PES/PEO (14/10), and Celgard 2325 are 145.9, 150.2, and 141.6 mAh g⁻¹, respectively, with the PES/PEO QSPEs exhibiting the highest value. After 100 cycles, both PES and PES/PEO QSPEs maintain superior capacity retention compared with Celgard 2325, which preserves only 77% of its initial capacity with a coulombic efficiency (CE) of 93%, likely owing to its limited electrolyte uptake and poor interfacial compatibility with the electrodes. In addition, its inherently poor wettability hinders sufficient electrolyte penetration, resulting in increased internal resistance and aggravated polarization during cycling. Consequently, the cell exhibits a more pronounced capacity decay, with the magnitude of the drop after 100 cycles being comparable to previous observations for unmodified Celgard 2325 separators.⁵¹ The PES membrane exhibits improved cycling stability (80% capacity retention, CE = 94%) due to its higher electrolyte uptake and enhanced interfacial stability. Incorporation of PEO increases membrane porosity and creates continuous ion-conduction pathways, resulting in the most stable performance among all the membranes. The PES/PEO (14/10) QSPE maintains 85% of its initial capacity with a CE of approximately 95% after 100 cycles, demonstrating superior interfacial stability and efficient ion transport compared with both pristine PES and Celgard 2325.

Based on the above literature, the key properties of representative PVDF-based and PFAS-free QSPEs are summarized in Table 4. PVDF-based systems without inorganic fillers are considered for comparison. Relative to these systems, the PES/PEO QSPEs developed in this work exhibit superior thermal stability (>200 °C, determined by TGA/DSC), compared with the typical range reported for PVDF-based QSPEs (133–175 °C), while maintaining comparable ionic conductivity. The PES/PEO membranes also display high electrolyte uptake, which can be attributed to their porous structure and improved wettability induced by the NIPS process, facilitating efficient ion transport. For PFAS-free QSPEs, reported systems based on PEG copolymer, PAN, or other fluorine-free polymers generally show ionic conductivities in the range of 0.45–1.40 mS cm⁻¹ and Li⁺ transference numbers of 0.59–0.77. In comparison, the PES/PEO QSPE developed in this work delivers an ionic conductivity of 0.605 mS cm⁻¹ together with a remarkably high electrolyte uptake of 395%, which is significantly higher than the values typically reported for PFAS-free QSPEs (commonly <210%). Moreover, the membrane demonstrates excellent thermal stability exceeding 200 °C, indicating improved structural integrity and safety under elevated temperature conditions.

Table 4 Key properties of representative PVDF-based and PFAS-free QSPEs compared with the PES/PEO QSPEs developed in this work

Materials	Liquid electrolyte	Electrolyte uptake (%)	Conductivity (mS cm^−1)	t Li^+	Thermal stability (°C)	Ref.
This work
PES/PEO	LiPF6/EC/DMC	395	0.605	0.42	200	This work
PVDF-based QSPEs
PVDF-HFP	LiTFSI/EC/DMC	181	0.246	0.39	175	19
PVDF-HFP/PPC	LiPF6/EC/DEC	—	1.18	0.47	133	20
PVDF-HFP/PMMA	LiPF6/EC/DEC	210	0.424	—	150	21
Honeycomb-like PVDF-HFP	LiPF6/EC/DMC	86.2	1.03	—	140	22
PFAS-free QSPEs
PP fiber-reinforced QSPE (PEG-based)	LiTFSI/DOL/DME	84	0.45	0.59	160	28
PVM fluorine-free polymer electrolyte	LiBOB/PC/DMC	—	1.40	0.77	167	29
La@ZIF-8/SiO2/PAN CQSE	LiTFSI/DOL/DME	207	0.676	0.67	—	30

---

In addition to its competitive electrochemical performance, the PES/PEO QSPE offers several practical and environmental advantages. The membrane can be fabricated via a straightforward NIPS process without the need for complex polymer synthesis, functionalization, or incorporation of inorganic fillers, while maintaining good flexibility and mechanical integrity. Compared with many reported PFAS-free QSPEs that rely on multi-step synthesis or composite design, this system achieves a balanced combination of properties through a simpler material design and processing route. From a green design perspective, the PES/PEO system provides environmental advantages over conventional fluorinated polymer electrolytes. The absence of fluorinated components eliminates concerns related to persistence and potential bioaccumulation associated with PFAS, while the use of industrially established polymers supports more sustainable material selection. Although a full life-cycle assessment is not yet available, the reduced synthesis complexity, elimination of fluorinated chemistry, and simplified fabrication process can be considered as practical indicators of lower environmental burden compared with state-of-the-art PFAS-free QSPEs that often require multi-step synthesis and functional modification.

Furthermore, the fabrication process is compatible with scalable manufacturing via the NIPS technique, which is widely used in industrial membrane production, enabling potential translation from laboratory-scale preparation to continuous large-area fabrication and roll-to-roll processing. The use of a DMSO/water solvent system also allows efficient solvent recovery and reuse, which is beneficial for reducing environmental burden during processing. Furthermore, the thermoplastic nature of the polymer matrix enables a possible “dissolve-and-recast” recycling strategy, offering an additional sustainability pathway compared with conventional crosslinked or fluorinated polymer electrolytes. This feature may facilitate material reprocessing and reduce waste generation at the end of service life. The intrinsic thermal and chemical stability of the PES framework further contributes to long-term operational durability, minimizing performance degradation under practical battery conditions and enhancing safety during extended cycling. Taken together, these features demonstrate that the PES/PEO QSPE developed here combines high electrochemical performance with practical processability and environmental compatibility, highlighting its potential as a greener quasi-solid polymer electrolyte system for sustainable energy storage applications.

Conclusions

In this study, PFAS-free PES/PEO QSPEs were successfully fabricated using a green solvent-based NIPS process. The synergistic effect of NIPS-induced phase separation and the subsequent partial removal of PEO creates a porous bi-continuous structure with significantly increased porosity and electrolyte uptake. This unique architecture establishes continuous ion-conduction pathways, leading to enhanced ionic conductivity, a wide electrochemical stability window, and stable lithium stripping/plating with reduced polarization. Compared to PES-only membranes and commercial separators, the PES/PEO QSPEs demonstrate superior rate capability and cycling stability in Li|LiFePO₄ cells. Notably, this system offers distinct advantages over representative PVDF-based and other PFAS-free QSPEs. It exhibits superior thermal stability compared to typical PVDF systems and higher electrolyte uptake than most reported PFAS-free alternatives, all without the need for complex polymer synthesis or inorganic fillers. Beyond performance, the industrial feasibility and circularity are key highlights. By leveraging mature NIPS technology and DMSO-based green chemistry, these membranes are highly compatible with large-scale roll-to-roll manufacturing. Furthermore, the thermoplastic nature of the PES/PEO framework allows for a ‘dissolve-and-recast’ physical recycling route, offering a sustainable and cost-effective alternative to conventional fluorinated electrolytes. Overall, this work provides a scalable, high-performance, and environmentally responsible platform for the next generation of safe lithium-ion batteries.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.

Supplementary information (SI) is available. See DOI: https://doi.org/10.1039/d6gc00553e.

Acknowledgements

We gratefully acknowledge financial support from the National Science and Technology Council (NSTC) of Taiwan. The authors also thank the Ministry of Education of Taiwan (Project: Talent and Technology Cultivation Base for Energy Battery Industry) for financial support. We additionally thank Tzu-Hsin Huang and Tzu-Hsin Hsiung from TKU for their valuable assistance in membrane preparation and electrochemical experiments. Z. F. acknowledges the support of China Experience Fund and the Stephen Slavens Faculty Scholar Endowment Fund from Oregon State University.

References

1. J. Xu, X. Cai, S. Cai, Y. Shao, C. Hu, S. Lu and S. Ding, High-Energy Lithium-Ion Batteries: Recent Progress and a Promising Future in Applications, Energy Environ. Mater., 2023, 6, e12450.
2. C. Comanescu, Ensuring Safety and Reliability: An Overview of Lithium-Ion Battery Service Assessment, Batteries, 2025, 11, 6.
3. J. Lu, Y. Chen, Y. Lei, P. Jaumaux, H. Tian and G. Wang, Quasi-Solid Gel Electrolytes for Alkali Metal Battery Applications, Nano-Micro Lett., 2025, 17, 194.
4. X. Fan, C. Zhong, J. Liu, J. Ding, Y. Deng, X. Han, L. Zhang, W. Hu, D. P. Wilkinson and J. Zhang, Opportunities of Flexible and Portable Electrochemical Devices for Energy Storage: Expanding the Spotlight onto Semi-Solid/Solid Electrolytes, Chem. Rev., 2022, 122, 17155–17239.
5. J. Lu, P. Jaumaux, T. Wang, C. Wang and G. Wang, Recent Progress in Quasi-Solid and Solid Polymer Electrolytes for Multivalent Metal-Ion Batteries, J. Mater. Chem. A, 2021, 9, 24175–24194.
6. K. S. Ngai, S. Ramesh, K. Ramesh and J. C. Juan, A Review of Polymer Electrolytes: Fundamental, Approaches and Applications, Ionics, 2016, 22, 1259–1279.
7. J. Chattopadhyay, T. S. Pathak and D. M. F. Santos, Applications of Polymer Electrolytes in Lithium-Ion Batteries: A Review, Polymers, 2023, 15, 3907.
8. X. Cheng, J. Pan, Y. Zhao, M. Liao and H. Peng, Gel Polymer Electrolytes for Electrochemical Energy Storage, Adv. Energy Mater., 2018, 8, 1702184.
9. M. Zhu, J. Wu, Y. Wang, M. Song, L. Long, S. H. Siyal, X. Yang and G. Sui, Recent Advances in Gel Polymer Electrolyte for High-Performance Lithium Batteries, J. Energy Chem., 2019, 37, 126–142.
10. J. Pan, P. Zhao, N. Wang, F. Huang and S. Dou, Research Progress in Stable Interfacial Constructions between Composite Polymer Electrolytes and Electrodes, Energy Environ. Sci., 2022, 15, 2753–2775.
11. S. Tang, W. Guo and Y. Fu, Advances in Composite Polymer Electrolytes for Lithium Batteries and Beyond, Adv. Funct. Mater., 2021, 31, 2000802.
12. S. Chai, Q. He, J. Zhou, Z. Chang, A. Pan and H. Zhou, A Solid-State Electrolytes and Electrode/Electrolyte Interfaces in Rechargeable Batteries, Adv. Energy Mater., 2024, 14, 202301268.
13. R. Ma, C. Lv and Z. Li, Amino-Modified Porous Aromatic Frameworks for Enhanced Lithium-Ion Dissociation and Transport in Polymer Electrolytes, Polymers, 2025, 17, 2401692.
14. X. Cheng and J. Bae, Recent Advancements in Fabrication, Separation, and Purification of Hierarchically Porous Polymer Membranes and Their Applications in Next-Generation Electrochemical Energy Storage Devices, Polymers, 2024, 16, 3269.
15. A. I. Osman, Z. Chen, A. M. Elgarahy, M. Farghali, I. M. A. Mohamed, A. K. Priya, H. B. Hawash and P.-S. Yap, Membrane Technology for Energy Saving: Principles, Techniques, Applications, Challenges, and Prospects, Adv. Energy Sustainability Res., 2024, 5, 2400011.
16. R. Liu, B. Yuan, S. Zhong, J. Liu, L. Dong, Y. Ji, Y. Dong, C. Yang and W. He, Poly(vinylidene fluoride) Separators for Next-Generation Lithium-Based Batteries, Nano Sel., 2021, 2, 2308–2345.
17. A. S. Gozdz, C. N. Schmutz and J. M. Tarascon, Lithium Secondary Battery Extraction Method, US Pat, 5540741, 1996.
18. A. Subramania, N. T. K. Sundaram, A. R. S. G. Priya and V. Kumar, Preparation of a Novel Composite Micro-Porous Polymer Electrolyte Membrane for High Performance Li-Ion Battery, J. Membr. Sci., 2007, 294, 8–15.
19. P. Zhang, R. Li, J. Huang, B. Liu, M. Zhou, B. Wen, Y. Xia and S. Okada, Flexible Poly(vinylidene fluoride-co-hexafluoropropylene)-Based Gel Polymer Electrolyte for High-Performance Lithium-Ion Batteries, RSC Adv., 2021, 11, 11943–11951.
20. Y. F. Liang, Y. Xia, S. Z. Zhang, X. L. Wang, X. H. Xia, C. D. Gu, J. B. Wu and J. P. Tu, A Preeminent Gel Blending Polymer Electrolyte of Poly(vinylidene fluoride-Hexafluoropropylene)-Poly(propylene carbonate) for Solid-State Lithium Ion Batteries, Electrochim. Acta, 2019, 296, 1064–1069.
21. D. Mouraliraman, N. Shaji, S. Praveen, M. Nanthagopal, C. W. Ho, M. V. Karthik, T. Kim and C. W. Lee, Thermally stable PVDF-HFP-Based Gel Polymer Electrolytes for High-Performance Lithium-Ion Batteries, Nanomaterials, 2022, 12, 1056.
22. J. Zhang, B. Sun, X. Huang, S. Chen and G. Wang, Honeycomb-Like Porous Gel Polymer Electrolyte Membrane for Lithium Ion Batteries with Enhanced Safety, Sci. Rep., 2014, 4, 6007.
23. E. Kissa, in Fluorinated Surfactants and Repellents, Marcel Dekker, New York, 2nd edn, 2001.
24. Z. Wang, J. C. DeWitt, C. P. Higgins and I. T. Cousins, A Never-Ending Story of Per- and Polyfluoroalkyl Substances (PFASs)?, Environ. Sci. Technol., 2017, 51, 2508–2518.
25. E. M. Sunderland, X. C. Hu, C. Dassuncao, A. K. Tokranov, C. C. Wagner and J. G. Allen, A Review of the Pathways of Human Exposure to Poly- and Perfluoroalkyl Substances (PFASs) and Present Understanding of Health Effects, J. Exposure Sci. Environ. Epidemiol., 2019, 29, 131–147.
26. I. T. Cousins, G. Goldenman, D. Herzke, R. Lohmann, M. Miller, C. A. Ng, S. Patton, M. Scheringer, X. Trier, L. Vierke, Z. Wang and J. C. DeWitt, The Concept of Essential Use for Determining When Uses of PFASs Can Be Phased Out, Environ. Sci.:Processes Impacts, 2019, 21, 1803–1815.
27. R. Lohmann and R. J. Letcher, The Universe of Fluorinated Polymers and Polymeric Substances and Potential Environmental Impacts and Concerns, Curr. Opin. Green Sustainable Chem., 2023, 41, 100795.
28. S. Gao, Y. Zhang, M. Ma, Z. Li, Z. Sun, M. Tian, H. Yang and P.-F. Cao, Fiber Reinforced Quasi-Solid Polymer Electrolytes Enabling Stable Li-metal Batteries, Mater. Adv., 2023, 4, 3452–3460.
29. W. Xu, L. Zhou, S. Lu, J. He, Y. Xu and L. Tian, Fluorine-Free Gel Polymer Electrolyte for Lithium Oxide-Rich Solid Electrolyte Interphase and Stable Li, Nat. Commun., 2025, 16, 9308.
30. Y. Chen, S. Li, C. Li, X. Leng, N. G. Prakash and T. J. Ko, Functional La@ZIF-8-Enhanced Composite Quasi-Solid Electrolyte for High-Performance Li-Metal Batteries, Chem. Eng. J., 2025, 508, 160932.
31. Q. Liu, Y. Dan, M. Kong, Y. Niu and G. Li, Sandwich-Structured Quasi-Solid Polymer Electrolyte Enables High-Capacity, Long-Cycling, and Dendrite-Free Lithium Metal Battery at Room Temperature, Small, 2023, 19, e2300118.
32. A. L. Mong and D. Kim, Self-Healable, Super Li-Ion Conductive, and Flexible Quasi-Solid Electrolyte for Long-Term Safe Lithium–Sulfur Batteries, J. Mater. Chem. A, 2023, 11, 6503–6521.
33. D. Shi, C. Li, Y. Yin, W. Lu, G. Li and X. Li, Application of Poly(ether sulfone)-Based Membranes in Clean Energy Technology, Chem. – Asian J., 2023, 18, e202201038.
34. W. Cui, L. Shi, W. Song, X. Wang, Z. Lin, W. Deng and Y. Ma, A Heatproof Electrospun PES/PVDF Composite Membrane as an Advanced Separator for Lithium–Ion Batteries, J. Appl. Polym. Sci., 2020, 137, 49328.
35. J. Liu, F. Song, Q. Li, J. Li, Z. Hong, C. Wang, M. Liu, L. Bai and F. Zeng, Preparation and Performance of Porous Polyethersulfone (PES)/Al₂O₃ Separator for High-Performance Lithium-Oxygen Battery, Ionics, 2021, 27, 4927.
36. M. Lo, K. C. W. Wu, J. M. Chin, L. P. Cheng and W. F. Kuan, Glycerol-Tailored Asymmetric Polyethersulfone Membranes with Uniform Ion Transport for Stable Lithium Metal Batteries, Giant, 2024, 17, 100227.
37. D. Zhang, L. Li, X. Wu, J. Wang, Q. Li, K. Pan and J. He, Research Progress and Application of PEO-Based Solid State Polymer Composite Electrolytes, Front. Energy Res., 2021, 9, 726738.
38. M. Xiao, F. Yang, S. Im, D. S. Dlamini, D. Jassby, S. Mahendra, R. Honda and E. M. V. Hoek, Characterizing Surface Porosity of Porous Membranes via Contact Angle Measurements, J. Membr. Sci. Lett., 2022, 2, 100022.
39. P. G. Bruce, J. Evans and C. A. Vincent, Conductivity and Transference Number Measurements on Polymer Electrolytes, Solid State Ionics, 1988, 28–30, 918–922.
40. G. Dreezen, Z. Fang and G. Groeninckx, Miscible Binary Blends of Poly(ethylene oxide) and Poly(ether sulphone). Part 1: Crystallization Kinetics, Morphology and Thermal Behavior, Polymer, 1999, 40, 5907.
41. X. Li, Y. Chen and Z. Wang, Fluoropolymer Membranes for Membrane Distillation and Membrane Crystallization, Polymers, 2022, 14, 5439.
42. C.-C. Ho, J. F. Su and L.-P. Cheng, Fabrication of High-Flux Asymmetric Polyethersulfone (PES) Ultrafiltration Membranes by Nonsolvent Induced Phase Separation Process: Effects of H₂O Contents in the Dope, Polymer, 2021, 217, 123451.
43. C. A. Smolders, A. J. Reuvers, R. M. Boom and I. M. Wienk, Microstructures in Phase-Inversion Membranes: Part I. Formation of Macrovoids, J. Membr. Sci., 1992, 73, 259–275.
44. J. C. Barbosa, J. P. Dias, S. Lanceros-Méndez and C. M. Costa, Recent Advances in Poly(vinylidene fluoride) and Its Copolymers for Lithium-Ion Battery Separators, Membranes, 2018, 8, 45.
45. Y. Song, L. Sheng, L. Wang, H. Xu and X. He, From Separator to Membrane: Separators Can Function More in Lithium Ion Batteries, Electrochem. Commun., 2021, 124, 106948.
46. X. Mao, Y. Li and J. Chen, Polyethylene Separator Activated by Hybrid Coating Improving Li+ Ion Transference Number and Ionic Conductivity for Li-Metal Battery, J. Power Sources, 2017, 342, 816–824.
47. J. Chen, G. Zhu, K. Jiang, H. Xu, Y. Zhao, H. Wang, P. Pang, Y. Bai, P. Wang and H. Wu, Facilely Construction of Multistage Channel in PVDF-HFP for High Ionic Conductivity and Prominent Anti-Lithium Dendrites, Chem. Eng. J., 2024, 479, 147761.
48. Y. Wang, K. Huang, P. Zhang, H. Li and H. Mi, PVDF-HFP Based Polymer Electrolytes with High Li⁺ Transference Number Enhancing the Cycling Performance and Rate Capability of Lithium Metal Batteries, Appl. Surf. Sci., 2022, 574, 51593.
49. J. Luo, C.-E. Wu, L.-Y. Su, S.-S. Huang, C.-C. Fang, Y.-S. Wu, J. Chou and N.-L. Wu, A Proof-of-Concept Graphite Anode with a Lithium Dendrite Suppressing Polymer Coating, J. Power Sources, 2018, 406, 63–69.
50. J. S. Edge, S. O'Kane, R. Prosser, N. D. Kirkaldy, A. N. Patel, A. Hales, A. Ghosh, W. Ai, J. Chen, J. Yang, S. Li, M.-C. Pang, L. Bravo Diaz, A. Tomaszewska, M. W. Marzook, K. N. Radhakrishnan, H. Wang, Y. Patel, B. Wu and G. J. Offer, Lithium Ion Battery Degradation: What You Need to Know, Phys. Chem. Chem. Phys., 2021, 23, 8200–8221.
51. B. Wang, L. Gao, Z. Yang, X. Wu, Q. Zhu, Q. Liu, F. Dogan, Y. Qin and C. Liao, Interphase Engineering Enabled by Using a Separator with Electrochemically Active Carbazole Polymers for Lithium-Ion Batteries, Polymers, 2025, 17, 1815.

---