Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Aug 22;65(42):e2958917. doi: 10.1002/anie.2958917

Ultrathin Membrane With Ion‐Specific Gating Enables Interfacial Decoupling of Conductivity–Selectivity for Aqueous Flow Batteries

Di Mu 1, Yu Xiong 2, Feiran Wang 1, Jiafeng Lei 1, Fei Ai 1, Jun Fan 2, Yi‐Chun Lu 1,✉
PMCID: PMC13653613  PMID: 42631394

ABSTRACT

Ion‐exchange membranes (IEMs) are essential for efficient and durable flow batteries, yet their inherent conductivity–selectivity trade‐off severely limits flow battery performance. Most conventional strategies to improve selectivity rely on regulating ion diffusion within ion‐exchange channels through size exclusion and electrostatic repulsion, inevitably compromising ionic conductivity. Here we shift the paradigm to ion partitioning at the membrane–electrolyte interface by establishing ion‐specific gating that selectively intercepts active species while preserving rapid charge carrier transport. The developed ultrathin (29 ± 2 µm) polar oxygen‐vacancy‐enhanced (POVE) membrane suppresses polysulfide crossover through strong chemisorption and a strengthened hydration layer, delivering over 3.4 times the selectivity of the base membrane at comparable conductivity and over 24.8 times the selectivity of commercial Nafion membranes. In the polysulfide–ferrocyanide flow battery, it achieved 82% energy efficiency at 40 mA cm−2 and sustained over 2000 cycles with negligible capacity decay, far exceeding 286 cycles for the base membrane and < 100 cycles for Nafion. This work decouples conductivity from selectivity via a scalable interfacial ion‐specific gating strategy, offering a transformative framework for developing affordable, high‐performance IEMs.

Keywords: chemisorption, conductivity–selectivity trade‐off, flow batteries, membranes, interface

An interfacial ion‐specific gating strategy shifts membrane design from ion diffusion within ion‐exchange channels to ion partitioning at the membrane–electrolyte interface, thereby decoupling selectivity from conductivity. The strategy underscores the pivotal yet underexplored role of the membrane–electrolyte interface, providing a broadly applicable framework for high‐performance membranes.

graphic file with name ANIE-65-e2958917-g003.webp

1. Introduction

Ion‐exchange membranes (IEMs) are critical components in a broad array of electrochemical systems, including flow batteries [1, 2], fuel cells [3, 4], seawater desalination [5], and water electrolysis [6]. Among these applications, flow batteries impose one of the most demanding requirements on IEMs, necessitating efficient conduction of charge carriers while rigorously blocking crossover of active species. In most flow battery systems, IEMs are the key components that determine the energy efficiency and cycle life, such as polysulfide‐based flow batteries (PBFBs) and vanadium flow batteries (VFBs) [7, 8]. However, a persistent and fundamental challenge of IEMs—the trade‐off between ionic conductivity and selectivity [9, 10]—has long hindered the commercial deployment of flow batteries. For instance, PBFBs possess extremely low chemical costs owing to the high earth abundance of sulfur, while the commercial membrane (e.g., Nafion) fails to provide sufficient potassium ion (K+) conductivity and effectively block small (∼9Å) polysulfide anions (Sx 2−) crossover [11, 12, 13], leading to low energy efficiency (< 70% at 40 mA cm−2) and poor cycle life (< 100 cycles) [14]. Therefore, overcoming the fundamental conductivity–selectivity trade‐off in IEMs is critical to realizing the flow battery's full potential as a low‐cost, durable grid‐scale storage technology.

Ion transport through IEMs encompasses two key processes: ion partitioning at the membrane–electrolyte interface and subsequent diffusion within ion‐exchange channels (Figure 1a) [15]. Conventional strategies to improve ionic selectivity of IEMs primarily focus on engineering the bulk membrane matrix to suppress active species crossover through structural densification of the polymer network and prolongation of tortuous ion transport pathways. Typical approaches include chemical crosslinking, polymer blending, and filler incorporation [16, 17, 18]. For example, incorporating triple tertiary amine‐grafted poly(2,6‐dimethyl‐1,4‐phenylene oxide) (PPO‐TTA) with sulfonated poly(ether ether ketone) (SPEEK) effectively mitigates vanadium ions crossover [17]. Ye et al. reported a hybrid membrane with superhydrophilic TiO2 nanotubes in Nafion 212 (N212) matrix, which showed higher coulombic efficiency and improved cycling stability  than the N212 membrane in VFBs [19]. Another charge‐reinforced selective membrane was applied in PBFBs, which significantly suppressed notorious polysulfide shuttling and extended the cycle life by establishing an enhanced electrostatic repulsion layer and reducing the diameter of the ion transport channels [20]. However, these modification approaches primarily regulate the ion diffusion within ion‐exchange channels through nonspecific mechanisms, like size exclusion and electrostatic repulsion (Figure 1b, left), restricting both active species and charge carriers. As a result, efforts to mitigate crossover inevitably impede the very conduction passways for charge carriers, increasing the area‐specific resistance (ASR) of IEMs and amplifying ohmic losses in flow battery applications. Such methods remain constrained by the conductivity–selectivity trade‐off, delivering limited improvements in overall battery performance (Figure 1c).

FIGURE 1.

FIGURE 1

Open in a new tab

Conductivity–selectivity trade‐off and different design strategies for IEMs. (a) Schematic illustration of the free energy landscape of ion transport through IEMs; (b) Comparison of strategies for enhancing selectivity in IEMs: conventional strategies regulate ion diffusion within ion‐exchange channels (left), and our ion‐specific gating strategy selectively blocks redox‐active species at the membrane–electrolyte interface, preserving rapid conduction of charge carriers (right); (c) Different conductivity–selectivity relationships between conventional strategies and interfacial ion‐specific gating strategy.

Shifting the paradigm from nonspecific bulk‐phase constraints to precise interfacial control, where ion partitioning predominantly occurs, represents a pivotal yet underexplored direction that holds promise for decoupling ionic selectivity from conductivity in IEMs. Biological ion channels provide a compelling blueprint: they achieve rapid transport of target ions with exceptional ionic selectivity through specific chemical interactions at binding sites localized at the membrane surface [21, 22]. Inspired by this natural paradigm, we developed an interfacial ion‐specific gating strategy that precisely targets the membrane–electrolyte interface to selectively intercept active species before they penetrate the membrane's internal transport network, while allowing unimpeded conduction of charge carriers (Figure 1b, right). Specifically, we fabricated an ultrathin (29 ± 2 µm) polar oxygen‐vacancy‐enhanced (POVE) membrane by coating Magnéli‐phase titanium suboxide (Ti4O7) layers onto a low‐cost and scalable SPEEK base substrate. The abundant oxygen vacancies in Ti4O7 [23, 24, 25] promoted strong polysulfide adsorption and strengthened the interfacial hydration layer, effectively blocking crossover while minimally impeding K+ transport. Consequently, this POVE membrane achieved over 3.4 times the selectivity of the base SPEEK membrane with comparable K+ conductivity, and over 24.8 times the selectivity of commercial Nafion membranes. In the polysulfide–ferrocyanide (S–Fe) flow battery, the POVE membrane delivered an energy efficiency of 82% at 40 mA cm−2 and sustained over 2,000 cycles with negligible capacity decay, far exceeding 286 cycles for the SPEEK base membrane and < 100 cycles for Nafion. At an estimated cost of 22.6–25.8 US$ m−2, which is less than one‐fiftieth that of Nafion 117 (N117) [20, 26, 27, 28] on an average cost basis, the strategy not only decouples the longstanding conductivity–selectivity trade‐off but also offers a scalable route to affordable and high‐performance IEMs for large‐scale flow battery deployment.

2. Results and Discussion

2.1. Macroscopic Physicochemical Properties Highlight Interfacial Decoupling of Conductivity and Selectivity

Ion transport through IEMs encounters interfacial ion partitioning, where ions rearrange and redistribute energy, forming an electric double layer at the membrane–electrolyte interface; subsequently, these ions migrate through inner hydrated nanochannels, traversing a heterogeneous free energy landscape composed of multiple local barriers [15]. Ionic selectivity arises from differences in the overall free energy profiles governing partitioning and transport for various species. However, in traditional IEMs, the high ionic strength and chemical complexity of flow battery electrolytes weaken interfacial discrimination, and the excessive swelling of hydrated nanochannels inevitably conduct charge carriers along with active species through the shared transport pathways, thereby coupling conductivity with selectivity. Our interfacial ion‐specific gating strategy addresses this by engineering a tailored interfacial layer that preferentially anchors active species while adding negligible penalties to charge carriers conduction, thus decoupling the two properties.

We fabricated the ultrathin (29 ± 2 µm) POVE membrane using low‐cost, easily prepared SPEEK as the base membrane and coating it with Magnéli‐phase Ti4O7, whose oxygen vacancies confer strong polysulfide affinity [23, 24, 25]. Bare SPEEK membrane with comparable thickness, carbon‐based Ketjen black‐decorated SPEEK (S@KB) membrane, and commonly used N212 and N117 are included for comparison (Figure S1). The cross‐sectional scanning electron microscopy (SEM) images and digital photographs revealed uniform, dense morphologies for Nafion series and SPEEK base membranes, while S@KB and POVE membranes exhibited sandwich architectures with functional layers coating on both sides (Figure 2a). Energy‐dispersive x‐ray spectroscopy (EDS) mapping further confirmed the elemental distributions of these ultrathin layers (Figures S2 and S3).

FIGURE 2.

FIGURE 2

Open in a new tab

Macroscopic physicochemical properties of different IEMs. (a) Cross‐sectional SEM images; (b) Solution uptake in various aqueous solutions; (c) Swelling ratios in various aqueous solutions; (d) Permeability of [Fe(CN)6]3−; (e) Permeability of S2 2−; (f) Area‐specific resistance (ASR) and ionic conductivity (IC) in 1 M KOH solution; (g) K+/S2 2− selectivity. The shallow ranges and error bars represent the standard deviation from at least three independent measurements for each membrane and condition.

Solution uptake is vital for forming hydrated, interconnected ion‐transport pathways, but excessive swelling can expand channels and reduce ionic selectivity. We noted that all membranes showed lower electrolyte uptake and swelling ratios in 1 M KOH and 1 M KCl aqueous solutions compared to deionized (DI) water (Figure 2b,c, and Table S1), and their swelling ratios decreased progressively with increasing saline/alkaline concentrations (Figure S4). This concentration‐dependent deswelling could be attributed to the reduced osmotic pressure and enhanced screening of the fixed sulfonate charges at higher external solution concentrations, which together suppress water uptake and expansion of the hydrated ionic domains [14]. N212 and N117 showed relatively low uptake but high swelling ratios in DI water, 1 M KOH, and 1 M KCl, attributable to their phase‐separated structure from rigid hydrophobic backbones and pendant hydrophilic side chains [14, 15, 29]. In contrast, the SPEEK membrane exhibited higher solution uptake with moderate swelling ratios. The S@KB and POVE membranes further showed slightly lower swelling ratios compared to the SPEEK substrate membrane while largely retaining the internal water channel density (Figure S5), suggesting that surface coatings limited electrolyte‐induced dimensional expansion without substantially disrupting the ion‐transport structures within the SPEEK substrate.

Ionic permeability was quantified using UV–visible spectroscopy in an H‐cell configuration (Figures S6 and S7). For posolytes, ferricyanide ions ([Fe(CN)6]3−) and ferrocyanide ions ([Fe(CN)6]4−) permeated rapidly through N212 (55 ± 3 µm, post‐preparation), and tripling the membrane thickness to N117 (183 ± 5 µm, post‐preparation) suppressed the accumulated concentration to about one‐third (Figure S8a,b). While Nafion series (including N212 and N117) still showed markedly higher ionic permeability than the ultrathin SPEEK‐based IEMs (including SPEEK, S@KB, and POVE), in which the crossover of [Fe(CN)6]3− and [Fe(CN)6]4− remained undetectable over the testing period (Figure 2d and Figure S9). These results underscored the excellent blocking ability of our prepared ultrathin SPEEK‐based membranes toward active materials in posolyte. In terms of S2 2−, a similar thickness‐dependent permeation trend was also observed in conventional Nafion series membranes: the thinner N212 exhibited nearly three times the flux of N117 and failed prematurely due to severe crossover (Figure S8c). However, despite being only one‐sixth as thick as N117, the SPEEK membrane reduced permeation to one‐third that of N117. Incorporation of ion‐specific interfacial layers further suppressed S2 2− crossover, with S@KB and POVE membranes reducing permeation to about half and one‐third of the SPEEK substrate membrane, respectively, at comparable thickness. Notably, the POVE membrane showed the smallest S2 2− permeation rate (2.25 × 10−8 cm2 min−1) among all membranes, substantially lower than those of Nafion membranes (107.26 × 10−8 cm2 min−1 for N117 and 140.31 × 10−8 cm2 min−1 for N212; Figure 2e). The relative permeation rates of S4 2− across different membranes also followed the same ranking order as S2 2− (Figure S8d). Among all tested ion species, S2 2− exhibited the highest permeability, which markedly exceeded that of S4 2− and was about an order of magnitude higher than those of [Fe(CN)6]3− and [Fe(CN)6]4− (Figure S9 and Table S2), making it a direct and sensitive metric for evaluating polysulfide‐blocking efficacy of IEMs.

Area‐specific resistance (ASR) and ionic conductivity (IC) were measured via electrochemical impedance spectroscopy (EIS) in 1 M KOH and 1 M KCl, respectively (Figure 2f and Figures S10 and S11). Although the ionic conductivity of the SPEEK base substrate (6.0 mS cm−1 in KOH) was slightly lower compared to Nafion membranes (9.9 mS cm−1 for N212 and 10.4 mS cm−1 for N117 in KOH), its markedly reduced thickness enabled a comparable area‐specific resistance to that of N212 (0.55 Ω cm2 for N212 versus 0.57 Ω cm2 for SPEEK in KOH) and only one‐third that of N117 (1.82 Ω cm2 in KOH). The interfacial coatings introduced marginal area‐specific resistance penalties, maintaining the intrinsic low‐resistance advantages of the ultrathin SPEEK membrane. The ionic conductivities of S@KB and POVE membranes further confirmed the negligible impact on K+ transport. Consequently, the K+/S2 2− selectivity (Figure 2g and Figure S12; Table S3) was substantially enhanced in the ultrathin ion‐specific gating membranes: the POVE membrane reached 2.4 × 105 S min cm−3 in KOH and 1.6 × 105 S min cm−3 in KCl, corresponding to over 3.4 times the ionic selectivity compared to the SPEEK base membrane while maintaining near‐identical conductivity, and over 24.8 times the ionic selectivity compared to widely used Nafion membranes.

Long‐term stability is essential for the practical deployment of flow batteries [30, 31]. The prepared membranes exhibited balanced mechanical rigidity, superior thermal integrity across operational temperatures, and robust chemical stability after 60 days in posolytes and negolytes (Figures S13–S16 and Note S1), supporting their reliability for extended cycling in flow batteries.

2.2. Microscopic Mechanisms Reveal the Role of Strong Adsorption and Strengthened Hydration Layer in Gating

To elucidate how interfacial ion‐specific gating achieves this decoupling function, we probed microscopic structures of various IEMs. Small‐angle x‐ray scattering (SAXS) of fully hydrated membranes showed continuous ion‐exchange channels in all SPEEK‐based variants, with nearly unchanged characteristic peaks at 3.23–3.07 nm−1 (Figure S17). This confirmed that our surface modifications leave the internal nanostructure largely intact, attributing the observed conductivity–selectivity decoupling to enhanced ion‐specific gating at the interface rather than bulk‐phase alterations. Figure 3a schematically illustrates the interfacial differences between S@KB and POVE membranes. The S@KB membrane relies on weak van der Waals–driven physisorption between porous carbon and polar polysulfide ions at the outer Helmholtz plane. In contrast, the Magnéli‐phase Ti4O7 on the POVE membrane promotes strong chemisorption at the inner Helmholtz plane [23, 24, 25], coupled with a strengthened hydration layer that further suppresses polysulfide crossover.

FIGURE 3.

FIGURE 3

Open in a new tab

Microscopic comparison of ion‐specific gating membranes with different polysulfide adsorption and hydration layer architectures. (a) Comparison of different electrolyte–membrane interfaces between POVE and S@KB membranes (purple particles for Ti4O7, black particles for KB and yellow particles for Sx 2−); (b) Density functional theory (DFT) calculations of polysulfide adsorption energies of Ti4O7 and KB; (c) UV–visible spectra of polysulfide ions before (original profile) and after soaking with Ti4O7 and KB, respectively; (d) XPS S 2p high‐resolution spectra of S@KB and POVE membrane after soaking in polysulfide solution; (e) The water contact angles of S@KB and POVE membranes; (f) ATR–FTIR spectra of fully hydrated S@KB and POVE membranes during dehydration for 0, 6, and 12 min (from darkest to lightest color); (g) XPS O 1s high‐resolution spectra of S@KB and POVE membrane after soaking in polysulfide solution.

Ti4O7 belongs to the sub‐stoichiometric titanium oxides, bridging TiO2 and Ti2O3 with oxygen‐vacancy‐driven shear planes of Ti–O octahedra (Figure S18). This endows Ti4O7 with metallic conductivity comparable to that of carbon‐based materials (Figure S19) [32, 33], which is crucial for low‐resistance gating of IEMs. Theoretical calculations confirmed strong chemical binding between Ti4O7 and polysulfides. Ti4O7 showed much stronger polysulfide adsorption than KB, with adsorption energies of −3.25 and −3.41 eV for K2S2 and K2S4, compared with −0.71 and −0.81 eV for KB, respectively (Figure 3b and Figure S20). The corresponding shorter substrate–S distances on Ti4O7 (2.45 and 2.57 Å) than KB (4.64 and 5.92 Å) further confirmed the stronger interfacial interactions (Table S4 and Note S2). Moreover, the high diffusion energy barriers restricted the diffusion of adsorbed polysulfides, and ab initio molecular dynamics (AIMD) simulations demonstrated the stable polysulfide binding under hydrated conditions (Figures S21–S23 and Note S3). Consistent with theoretical calculations, experimental adsorption tests also demonstrated strong polysulfide affinity of Ti4O7: the characteristic UV–visible peaks for S2 2− (300 nm) and S4 2− (375 nm) [20] diminished markedly with Ti4O7, but only slightly attenuated with KB (Figure 3c). These results can be attributed to the different adsorption mechanisms between functional materials and polysulfides. Unlike KB's surface‐area‐dependent physisorption, polysulfide adsorption by Ti4O7 arises from chemical binding at abundant active sites [23, 34]. High‐resolution x‐ray photoelectron spectroscopy (XPS) S 2p spectra of the POVE membrane after polysulfide immersion showed chemisorption peaks at 161.9 and 162.8 eV, while absent in the S@KB membrane (only sulfonate group signals at 167.9 and 169 eV; Figure 3d) [35, 36, 37]. The strong interaction facilitates the rapid immobilization of polysulfide ions, preventing further crossover.

The strengthened hydration layer on the POVE membrane adds a second gating mechanism. Ions partially dehydrate and reorganize hydration shells at the interface, with associated energy barriers sensitive to ionic size, hydration free energy and surface hydrophilicity [38, 39, 40, 41]. Compared with K+, Sx 2− with larger hydrated dimensions are expected to experience overall higher penalties for their structural rearrangement and partial dehydration when traversing the highly structured hydration layer, thereby further lowering the possibility of Sx 2− penetrating through the membrane [14, 42]. As shown in Figure 3e, the contact angle was only 53° for the POVE membrane, reflecting the hydrophilic nature of Ti4O7; whereas an increased value of 98° for the S@KB membrane, indicating its surface hydrophobicity. Compared with S@KB membrane, attenuated total reflectance Fourier–transform infrared (ATR–FTIR) spectra of fully hydrated POVE membrane exhibited pronounced redshifts in both the O–H stretching and H–O–H bending vibration (Figure 3f). Besides, a prominent increased peak appeared at 3238 cm−1 in the POVE membrane. During the dehydration process, the POVE membrane showed significantly smaller blueshifts in both characteristic peaks, especially in the O–H stretching vibration, which were only an 8 cm−1 blueshift to 3372 cm−1 for POVE while a 38 cm−1 blueshift to 3416 cm−1 for the S@KB membrane. These spectral changes indicated a strengthened and stable hydrogen‐bond network of the POVE membrane, consistent with the formation of a more structured hydration layer. XPS O 1s spectra also showed Ti−OH peak, which corroborates the stronger interfacial hydration environment in the POVE membrane (Figure 3g). The POVE's more negative surface charge in negolyte also enhanced electrostatic repulsion at the shear plane (Figure S24).

2.3. Efficient and Durable Performance in Aqueous S–Fe Flow Batteries Demonstrates Practical Impact

We evaluated the membranes in S–Fe flow batteries employing S2 2−/S4 2− (−0.46 V versus V SHE ) as the negolyte redox couple and [Fe(CN)6]4−/[Fe(CN)6]3− (0.45 V versus V SHE ) as the posolyte redox couple (Figure 4a) [43, 44, 45]. S–Fe flow battery with N117 displayed nearly twice the resistance than that of SPEEK‐based membranes (Figure 4b and Figure S25). As a result, N117 showed significantly high overpotentials across all current densities, limiting the maximum stable operation to 120 mA cm−2 (Figure 4c). Beyond this, the discharge capacity collapsed mainly due to the rapid crossover‐induced sulfur accumulation and resistance‐driven failure (Figure S26). In contrast, the S–Fe flow batteries equipped with our prepared S@KB and POVE membranes sustained up to 200 mA cm−2, with discharge capacities consistently exceeding N117 and N212 from 20 to 200 mA cm−2 (Figure 4d and Figure S27). The flow batteries assembled with S@KB and POVE membranes reached the energy efficiency of 81% and 82% at 40 mA cm−2, respectively, which was comparable to the SPEEK base membrane, versus only 69% for N117 (Figure S28). Furthermore, power densities of the S–Fe flow batteries improved markedly (Figure S29): N117 peaked at 63.5 mW cm−2 (@130 mA cm−2), while POVE membrane achieved 109.4 mW cm−2 (@210 mA cm−2) and S@KB membrane reached 104.6 mW cm−2 (@200 mA cm−2), respectively. The POVE membrane's edge at higher densities than the S@KB membrane is likely attributable to stronger chemisorption anchoring polysulfides more effectively under harsh conditions, thereby extending the operational range. The open‐circuit voltage (OCV) decay test reflects the overall crossover behavior under continuous‐flow conditions. The flow battery with N117 sustained only 153 h before a rapid voltage drop, and N212 exhibited even faster voltage decay, failing within just 46 h (Figure 4e and Figure S30). These results reflected the severe polysulfide shuttling and thickness dependence in both Nafion membranes. In contrast, the S–Fe flow battery with the POVE membrane maintained high voltage for over 600 h, confirming its superior crossover‐blocking ability compared to the SPEEK base membrane (289 h). As a result, the flow battery with POVE membrane delivered exceptional cycling stability, sustaining over 2000 cycles with over 99.9997% capacity retention per cycle (Figure 4f and Figure S31), far outperforming batteries using SPEEK base substrate (286 cycles) and S@KB membrane (∼1000 cycles). By contrast, Nafion membranes failed much earlier, which was 98 cycles for N117 and 61 cycles for N212. We further explored the applicability of the POVE membrane to other polysulfide‐based flow battery systems, such as polysulfide‐polyiodide (S–I) flow battery (Figure S32). Owing to its low membrane resistance and effective suppression of polysulfide crossover, the POVE membrane enabled reversible operation over current densities ranging from 20 to 80 mA cm−2 and showed stable cycling performance over 250 cycles at 40 mA cm−2, with an average coulombic efficiency exceeding 99.6% and an average energy efficiency over 82.6%. Post‐cycling characterizations (Figures S33–S36 and Note S4) showed the minimal structural degradation and limited surface deposition of the POVE membrane, supporting the long‐term durability of the ultrathin ion‐specific gating IEMs.

FIGURE 4.

FIGURE 4

Open in a new tab

Efficient and durable aqueous S–Fe flow battery performance. (a) Schematic illustration of aqueous S–Fe flow battery assembled with IEM; (b) Nyquist plots of various IEMs by EIS test; (c) Voltage–current profiles of the S–Fe flow battery with N117 membrane, the color change indicates increasing current density from 20 mA cm−2 (lightest) to 200 mA cm−2 (darkest) with a 20 mA cm−2 interval; (d) Voltage−current profiles of the S–Fe flow battery with POVE membrane, the color change indicates increasing current density from 20 mA cm−2 (lightest) to 200 mA cm−2 (darkest) with a 20 mA cm−2 interval; (e) Open‐circuit voltage (OCV) curves of various IEMs; (f) Long‐term cycling stability of various IEMs under the current density of @40 mA cm−2; (g) The comparison of the membrane price and cycle number of N117 and POVE membranes (error bars represent the price range of IEMs); (h) Cycling performance of the S–Fe flow battery with POVE membrane compared with recently reported PBFBs [13, 14, 20, 43, 44, 45, 46, 47, 48, 49, 50].

A cost–performance comparison emphasizes the impact of the ultrathin POVE membrane (Figure 4g, Tables S5 and S6). With a production cost of 22.6–25.8 US$ m−2, the POVE membrane enabled stable operation for over 2000 cycles. In stark contrast, N117 membranes (500−2200 US$ m−2) [20, 26, 27, 28], with a cost of over 50‐fold higher on an average basis, failed the operation after just 98 cycles. Figure 4h and Table S7 compare the S–Fe flow battery performance using the POVE membrane with that of other recently reported PBFBs. The POVE membrane exhibited a rare combination of high energy efficiency and long‐term cycling stability, underscoring the effectiveness of this interfacial ion‐specific gating strategy in overcoming membrane limitations of PBFBs. The strategy highlights the pivotal role of the membrane–electrolyte interface and offers a scalable route to affordable IEMs in diverse energy systems, facilitating grid‐scale energy storage for renewable energy integration.

3. Conclusion

We have shown that shifting the design from nonspecific regulation within bulk‐phase ion‐exchange channels to ion partition at the membrane–electrolyte interface offers a promising design route to addressing the longstanding conductivity–selectivity trade‐off in IEMs. An interfacial ion‐specific gating strategy was rigorously validated in PBFBs, where low K+/S2 2−selectivity led to sulfur precipitation, membrane degradation, and eventual battery failure. By designing ultrathin (29 ± 2 µm) low‐cost SPEEK‐based membranes with tailored interfacial ion‐specific gating, we achieved precise control over ion‐selective transport with ionic conductivity largely intact. The POVE membrane anchors polysulfides through strong chemisorption and a strengthened hydration layer, outperforming the S@KB membrane's weak physisorption as well as its base SPEEK membrane. As a result, the POVE membranes delivered over 3.4 times the K+/S2 2− selectivity of its base membrane with comparable conductivity, and over 24.8 times K+/S2 2− selectivity relative to commercial Nafion membranes. In the S–Fe flow battery, the POVE membrane enabled stable operation at current densities up to 200 mA cm−2, delivered approximately 82% energy efficiency at 40 mA cm−2, and sustained over 2000 cycles with negligible capacity fade (∼1000 cycles for S@KB and <100 cycles for both N117 and N212). The POVE achieved an estimated cost of 22.6–25.8 US$ m−2, illustrating the practicality of this interfacial ion‐specific gating strategy that not only breaks the selectivity–conductivity trade‐off but also offers a simple, effective, and broadly applicable framework for high‐performance, low‐cost IEMs. Beyond this detailed example in PBFBs, the strategy also demonstrates the pivotal yet underexplored role of the membrane–electrolyte interface, which offers a versatile design route to enhance electrochemical energy storage performance.

Author Contributions

Di Mu: conceptualization, methodology, data curation, investigation, validation, formal analysis, visualization, writing – original draft, writing – review & editing. Yu Xiong: methodology, formal analysis, writing – review & editing. Feiran Wang: methodology, investigation, writing – review & editing. Jiafeng Lei: investigation, methodology, writing – review & editing. Fei Ai: methodology, writing – review & editing. Jun Fan: methodology, writing – review & editing, funding acquisition. Yi‐Chun Lu: conceptualization, writing – review & editing, supervision, funding acquisition, project administration.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Experimental methods and other data that support the findings of this study are available in the accompanying Supporting Information. The authors have cited additional references within the Supporting Information [51–66].Supporting File: anie74329‐sup‐0001‐SuppMat.pdf.

ANIE-65-e2958917-s001.pdf (3.3MB, pdf)

Acknowledgments

The work described in this paper was supported by grants from the Research Grant Council (RGC) of the Hong Kong Special Administrative Region, China (project nos. RFS2223‐4S03, CUHK 14302823, and CUHK 14310124), City University of Hong Kong Projects 7006111 and 7020112, and a grant from the National Natural Science Foundation of China (22525806). Y.‐C.L. acknowledges the support from the Xplorer Prize by New Cornerstone Science Foundation.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. The experimental results are presented in the supplemental information.

References

  • 1. Dunn B., Kamath H., and Tarascon J.‐M., “Electrical Energy Storage for the Grid: A Battery of Choices,” Science 334 (2011): 928–935, 10.1126/science.1212741. [DOI] [PubMed] [Google Scholar]
  • 2. Park M., Ryu J., Wang W., and Cho J., “Material Design and Engineering of Next‐Generation Flow‐Battery Technologies,” Nature Reviews Materials 2 (2017): 16080, 10.1038/natrevmats.2016.80. [DOI] [Google Scholar]
  • 3. Jiao K., Xuan J., Du Q., et al., “Designing the Next Generation of Proton‐Exchange Membrane Fuel Cells,” Nature 595 (2021): 361–369, 10.1038/s41586-021-03482-7. [DOI] [PubMed] [Google Scholar]
  • 4. Debe M. K., “Electrocatalyst Approaches and Challenges for Automotive Fuel Cells,” Nature 486 (2012): 43–51, 10.1038/nature11115. [DOI] [PubMed] [Google Scholar]
  • 5. Srimuk P., Su X., Yoon J., Aurbach D., and Presser V., “Charge‐Transfer Materials for Electrochemical Water Desalination, Ion Separation and the Recovery of Elements,” Nature Reviews Materials 5 (2020): 517–538, 10.1038/s41578-020-0193-1. [DOI] [Google Scholar]
  • 6. Yu L., Ning M., Wang Y., Yuan C., and Ren Z., “Direct Seawater Electrolysis for Hydrogen Production,” Nature Reviews Materials 10 (2025): 857–873, 10.1038/s41578-025-00826-x. [DOI] [Google Scholar]
  • 7. Li Z. and Lu Y.‐C., “Material Design of Aqueous Redox Flow Batteries: Fundamental Challenges and Mitigation Strategies,” Advanced Materials 32 (2020): 2002132, 10.1002/adma.202002132. [DOI] [PubMed] [Google Scholar]
  • 8. Qian P., Zhou W., Zhang Y., Chao D., and Song M., “Review and Perspectives of Sulfonated Poly(Ether Ether Ketone) Proton Exchange Membrane for Vanadium Flow Batteries,” Energy & Fuels 37 (2023): 17681–17707, 10.1021/acs.energyfuels.3c02373. [DOI] [Google Scholar]
  • 9. Dai Q., Liu Z., Huang L., et al., “Thin‐Film Composite Membrane Breaking the Trade‐Off Between Conductivity and Selectivity for a Flow Battery,” Nature Communications 11 (2020): 13, 10.1038/s41467-019-13704-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Park H. B., Kamcev J., Robeson L. M., Elimelech M., and Freeman B. D., “Maximizing the Right Stuff: The Trade‐Off Between Membrane Permeability and Selectivity,” Science 356 (2017): eaab0530, 10.1126/science.aab0530. [DOI] [PubMed] [Google Scholar]
  • 11. Ding M., Fu H., Lou X., et al., “A Stable and Energy‐Dense Polysulfide/Permanganate Flow Battery,” ACS Nano 17 (2023): 16252–16263, 10.1021/acsnano.3c06273. [DOI] [PubMed] [Google Scholar]
  • 12. Lei J., Yao Y., Huang Y., and Lu Y.‐C., “A Highly Reversible Low‐Cost Aqueous Sulfur–Manganese Redox Flow Battery,” ACS Energy Letters 8 (2023): 429–435, 10.1021/acsenergylett.2c02524. [DOI] [Google Scholar]
  • 13. Chen B., Huang H., Lin J., et al., “Doping Engineering of M‐N‐C Electrocatalyst Based Membrane‐Electrode Assembly for High‐Performance Aqueous Polysulfides Redox Flow Batteries,” Advanced Science 10 (2023): 2206949, 10.1002/advs.202206949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Wang F., Luo S., Lei J., et al., “Nonfluorinated Membrane With a Decentralized Ion‐Transport Network Enables Efficient and Sustainable Polysulfide Redox Flow Batteries,” Science Advances 11 (2025): eaea0032, 10.1126/sciadv.aea0032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Wang F., Mu D., and Lu Y.‐C., “Ion‐Conducting Membranes for Long‐Duration Energy Storage,” ACS Energy Letters 10 (2025): 3096–3111, 10.1021/acsenergylett.5c00934. [DOI] [Google Scholar]
  • 16. Xu Z., Tang H., and Li N., “Enhanced Proton/Iron Permselectivity of Sulfonated Poly(Ether Ether Ketone) Membrane Functionalized With Basic Pendant Groups during Electrodialysis,” Journal of Membrane Science 610 (2020): 118227, 10.1016/j.memsci.2020.118227. [DOI] [Google Scholar]
  • 17. Zhang H., Yan X., Gao L., et al., “Novel Triple Tertiary Amine Polymer‐Based Hydrogen Bond Network Inducing Highly Efficient Proton‐Conducting Channels of Amphoteric Membranes for High‐Performance Vanadium Redox Flow Battery,” ACS Applied Materials & Interfaces 11 (2019): 5003–5014, 10.1021/acsami.8b18617. [DOI] [PubMed] [Google Scholar]
  • 18. Xia Y., Cao H., Xu F., et al., “Polymeric Membranes With Aligned Zeolite Nanosheets for Sustainable Energy Storage,” Nature Sustainability 5 (2022): 1080–1091, 10.1038/s41893-022-00974-w. [DOI] [Google Scholar]
  • 19. Ye J., Zhao X., Ma Y., et al., “Hybrid Membranes Dispersed With Superhydrophilic TiO2 Nanotubes Toward Ultra‐Stable and High‐Performance Vanadium Redox Flow Batteries,” Advanced Energy Materials 10 (2020): 1904041, 10.1002/aenm.201904041. [DOI] [Google Scholar]
  • 20. Li Z. and Lu Y.‐C., “Polysulfide‐Based Redox Flow Batteries With Long Life and Low Levelized Cost Enabled by Charge‐Reinforced Ion‐Selective Membranes,” Nature Energy 6 (2021): 517–528, 10.1038/s41560-021-00804-x. [DOI] [Google Scholar]
  • 21. Violet C., Ball A., Heiranian M., et al., “Designing Membranes With Specific Binding Sites for Selective Ion Separations,” Nature Water 2 (2024): 706–718, 10.1038/s44221-024-00279-6. [DOI] [Google Scholar]
  • 22. Kopec W., Köpfer D. A., Vickery O. N., et al., “Direct Knock‐on of Desolvated Ions Governs Strict Ion Selectivity in K+ Channels,” Nature Chemistry 10 (2018): 813–820, 10.1038/s41557-018-0105-9. [DOI] [PubMed] [Google Scholar]
  • 23. Pang Q., Kundu D., Cuisinier M., and Nazar L. F., “Surface‐Enhanced Redox Chemistry of Polysulphides on a Metallic and Polar Host for Lithium‐Sulphur Batteries,” Nature Communications 5 (2014): 4759, 10.1038/ncomms5759. [DOI] [PubMed] [Google Scholar]
  • 24. Wang F., Ding X., Shi R., et al., “Facile Synthesis of Ti4O7 on Hollow Carbon Spheres With Enhanced Polysulfide Binding for High‐Performance Lithium–Sulfur Batteries,” Journal of Materials Chemistry A 7 (2019): 10494–10504, 10.1039/C9TA00544G. [DOI] [Google Scholar]
  • 25. Liu M., Jhulki S., Sun Z., Magasinski A., Hendrix C., and Yushin G., “Atom‐Economic Synthesis of Magnéli Phase Ti4O7 Microspheres for Improved Sulfur Cathodes for Li–S Batteries,” Nano Energy 79 (2021): 105428, 10.1016/j.nanoen.2020.105428. [DOI] [Google Scholar]
  • 26. Oliot M., Galier S., Roux de Balmann H., and Bergel A., “Ion Transport in Microbial Fuel Cells: Key Roles, Theory and Critical Review,” Applied Energy 183 (2016): 1682–1704, 10.1016/j.apenergy.2016.09.043. [DOI] [Google Scholar]
  • 27. Prifti H., Parasuraman A., Winardi S., Lim T. M., and Skyllas‐Kazacos M., “Membranes for Redox Flow Battery Applications,” Membranes 2 (2012): 275–306, 10.3390/membranes2020275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Elham O. S. J., Kamarudin S. K., Shaari N., Zainoodin A. M., Zakaria Z., and Yusof M. R., “Development of Low‐Cost Nafion‐Lignin Composite Conductive Membranes for Application in Direct Methanol Fuel Cells,” Journal of Environmental Chemical Engineering 12 (2024): 111514, 10.1016/j.jece.2023.111514. [DOI] [Google Scholar]
  • 29. Mauritz K. A. and Moore R. B., “State of Understanding of Nafion,” Chemical Reviews 104 (2004): 4535–4586, 10.1021/cr0207123. [DOI] [PubMed] [Google Scholar]
  • 30. Zhang C., Yuan Z., and Li X., “Designing Better Flow Batteries: An Overview on Fifty Years' Research,” ACS Energy Letters 9 (2024): 3456–3473, 10.1021/acsenergylett.4c00773. [DOI] [Google Scholar]
  • 31. Fan H., Liu K., Zhang X., et al., “Spatial Structure Regulation Towards Armor‐Clad Five‐Membered Pyrroline Nitroxides Catholyte for Long‐Life Aqueous Organic Redox Flow Batteries,” eScience 4 (2024): 100202, 10.1016/j.esci.2023.100202. [DOI] [Google Scholar]
  • 32. Domaschke M., Zhou X., Wergen L., et al., “Magnéli‐Phases in Anatase Strongly Promote Cocatalyst‐Free Photocatalytic Hydrogen Evolution,” ACS Catalysis 9 (2019): 3627–3632, 10.1021/acscatal.9b00578. [DOI] [Google Scholar]
  • 33. Tao X., Wang J., Ying Z., et al., “Strong Sulfur Binding With Conducting Magnéli‐Phase TinO2n–1 Nanomaterials for Improving Lithium–Sulfur Batteries,” Nano Letters 14 (2014): 5288–5294, 10.1021/nl502331f. [DOI] [PubMed] [Google Scholar]
  • 34. Mei S., Jafta C. J., Lauermann I., et al., “Porous Ti4O7 Particles With Interconnected‐Pore Structure as a High‐Efficiency Polysulfide Mediator for Lithium–Sulfur Batteries,” Advanced Functional Materials 27 (2017): 1701176, 10.1002/adfm.201701176. [DOI] [Google Scholar]
  • 35. Wei P., Sui Y., Li X., et al., “Sandwich‐Structure PI/SPEEK/PI Proton Exchange Membrane Developed for Achieving the High Durability on Excellent Proton Conductivity and Stability,” Journal of Membrane Science 644 (2022): 120116, 10.1016/j.memsci.2021.120116. [DOI] [Google Scholar]
  • 36. Nasluzov V., Shor A., Romanchenko A., Tomashevich Y., and Mikhlin Y., “DFT + U and Low‐Temperature XPS Studies of Fe‐Depleted Chalcopyrite (CuFeS2) Surfaces: A Focus on Polysulfide Species,” Journal of Physical Chemistry C 123 (2019): 21031–21041, 10.1021/acs.jpcc.9b06127. [DOI] [Google Scholar]
  • 37. Guo Y., Chen Y., Lim G. J. H., et al., “Reversible Alkaline Sulfur Cathode Based on Six‐Electron Electrochemistry for Advanced Aqueous Sulfur Batteries,” ACS Nano 1 9 (2025): 15522–15536, 10.1021/acsnano.4c16835. [DOI] [PubMed] [Google Scholar]
  • 38. Razmjou A., Asadnia M., Hosseini E., Habibnejad Korayem A., and Chen V., “Design Principles of Ion Selective Nanostructured Membranes for the Extraction of Lithium Ions,” Nature Communications 10 (2019): 5793, 10.1038/s41467-019-13648-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Wei Z., Wang Y., Hong H., et al., “Long‐Life Aqueous Zinc‐Iodine Flow Batteries Enabled by Selectively Intercepting Hydrated Ions,” Nature Communications 16 (2025): 9301, 10.1038/s41467-025-64344-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Lu C., Hu C., Chen Z., et al., “Dehydration‐Enhanced Ion‐Pore Interactions Dominate Anion Transport and Selectivity in Nanochannels,” Science Advances 9 (2023): eadf8412, 10.1126/sciadv.adf8412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Ershov A., Xu H., Li Y., Tong T., and Epsztein R., “Role of Ion Dehydration in Ion–Ion Selectivity of Dense Membranes,” Environmental Science & Technology 59 (2025): 17997–18009, 10.1021/acs.est.5c04303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Epsztein R., Shaulsky E., Qin M., and Elimelech M., “Activation Behavior for Ion Permeation in Ion‐Exchange Membranes: Role of Ion Dehydration in Selective Transport,” Journal of Membrane Science 580 (2019): 316–326, 10.1016/j.memsci.2019.02.009. [DOI] [Google Scholar]
  • 43. Wei X., Xia G.‐G., Kirby B., et al., “An Aqueous Redox Flow Battery Based on Neutral Alkali Metal Ferri/Ferrocyanide and Polysulfide Electrolytes,” Journal of the Electrochemical Society 163 (2016): A5150–A5153, 10.1149/2.0221601jes. [DOI] [Google Scholar]
  • 44. Lei J., Zhang Y., Yao Y., et al., “An Active and Durable Molecular Catalyst for Aqueous Polysulfide‐Based Redox Flow Batteries,” Nature Energy 8 (2023): 1355–1364, 10.1038/s41560-023-01370-0. [DOI] [Google Scholar]
  • 45. Lou X., Fu H., Xu J., et al., “Cost‐Effective Membrane and Advanced Electrode for Stable Polysulfide‐Ferricyanide Flow Battery,” Energy Material Advances 2022 (2022): 9865618, 10.34133/2022/9865618. [DOI] [Google Scholar]
  • 46. Li Z., Weng G., Zou Q., Cong G., and Lu Y.‐C., “A High‐Energy and Low‐Cost Polysulfide/Iodide Redox Flow Battery,” Nano Energy 30 (2016): 283–292, 10.1016/j.nanoen.2016.09.043. [DOI] [Google Scholar]
  • 47. Ma D., Hu B., Wu W., et al., “Highly Active Nanostructured CoS2/CoS Heterojunction Electrocatalysts for Aqueous Polysulfide/Iodide Redox Flow Batteries,” Nature Communications 10 (2019): 3367, 10.1038/s41467-019-11176-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Sarfaraz Khabbaz M., Thakur A., Maity D., et al., “A Polysulfide/Ferricyanide Redox Flow Battery With Extended Cycling,” Journal of Materials Chemistry A 13 (2025): 29610–29620, 10.1039/d5ta05404d. [DOI] [Google Scholar]
  • 49. Sreenath S., Nayanthara P. S., Pawar C. M., et al., “An Aqueous Polysulfide Redox Flow Battery With a Semi‐Fluorinated Cation Exchange Membrane,” Energy Advances 3 (2024): 203–214, 10.1039/d3ya00509g. [DOI] [Google Scholar]
  • 50. Hou Q., Cao H., Xia Y., et al., “Rigid Framework‐Constrained Membranes: A Key to Long‐Life Polysulfide/Polyiodide Redox Flow Batteries,” ACS Applied Materials & Interfaces 17 (2025): 40508–40516, 10.1021/acsami.5c07898. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Experimental methods and other data that support the findings of this study are available in the accompanying Supporting Information. The authors have cited additional references within the Supporting Information [51–66].Supporting File: anie74329‐sup‐0001‐SuppMat.pdf.

ANIE-65-e2958917-s001.pdf (3.3MB, pdf)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. The experimental results are presented in the supplemental information.

Articles from Angewandte Chemie (International Ed. in English) are provided here courtesy of Wiley

RESOURCES