Skip to main content
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Dec 17;22(5):e11358. doi: 10.1002/smll.202511358

In Situ Crosslinked Diallylammonium‐Functionalized Poly(Biphenyl Alkylene) for High‐Performance Anion Exchange Membranes

Jiyoon Jung 1,2,3, Young Sang Park 1,2,3, Mo Beom Koo 1, Jinsuk Ku 1,2,3, Cheol‐Hee Ahn 2, Seung Sang Hwang 1, Hoon Taek Chung 4,, Albert S Lee 1,3,
PMCID: PMC12824562  PMID: 41406383

Abstract

The intrinsic trade‐off between ion conductivity and mechanical integrity in high‐ion exchange capacity (IEC) anion exchange membranes (AEMs) has remained a major challenge for the practical implementation of anion exchange membrane water electrolyzer (AEMWE) technology. Here, crosslinked AEMs (BP‐cross‐PDAA) is developed via free‐radical cyclopolymerization of diallylammonium‐functionalized poly(biphenyl alkylene), where the IEC (n) is precisely tuned from 2.32 meq g−1 to 3.39 meq g−1 by adjusting the content of diallyldimethylammonium comonomer. BP‐cross‐PDAA membranes exhibited excellent hydration characteristics, including hydroxide conductivity (up to 152.4 mS cm−1 at 80 °C), swelling ratio (24.2%), and alkaline stability (98.7% retention over 350 h), attributed to their robust crosslinked network structure without sacrificing their high IEC. Furthermore, membrane‐electrode assembly (MEA‐n) tests demonstrated remarkable AEMWE performance, with MEA‐3.39 achieving a peak current density of 12.39 A cm−2 at 2.0 V using a non‐platinum group metal anode, and 6.1 A cm−2 under pure water condition. MEA‐3.39 also exhibited excellent durability, with a minimal voltage degradation rate of 1.2 µV h−1 over 1,000 h at 1 A cm−2 in 1 M KOH at 50 °C.

Keywords: anion exchange membrane water electrolyzer, crosslinked anion exchange membrane, free‐radical cyclopolymerization


Crosslinked anion exchange membranes are synthesized via in situ free‐radical cyclopolymerization of diallylammonium‐functionalized poly(biphenyl alkylene). The precisely tunable ion exchange capacity (2.32–3.39 meq g−1) and robust crosslinked structure provided excellent hydration stability without compromising ion conductivity. The optimized MEA exhibited remarkable water electrolysis performance, achieving a peak current density of 12.39 A cm−2 at 2.0 V using a non‐PGM anode.

graphic file with name SMLL-22-e11358-g003.jpg

1. Introduction

By combining the high hydrogen production efficiency of proton exchange membrane water electrolyzers (PEMWEs) with the capability of employing non‐platinum group metal catalysts of liquid alkaline water electrolyzers (AWEs),[ 1 3 ] anion exchange membrane water electrolyzers (AEMWEs) have emerged as a promising technology for large‐scale, cost‐effective green hydrogen production.[ 4 , 5 ] However, the essential attributes of anion exchange membranes (AEMs) required for high‐performance AEMWEs, including high ion conductivity, mechanical robustness, dimensional stability, and long‐term alkaline endurance, have yet to be achieved concurrently.[ 6 8 ]

High ion exchange capacity (IEC) is essential to achieve high charge‐carrier density and thus high ion conductivity, which is critical for ef efficient AEM‐based electrochemical devices. However, excessively increasing IEC generally causes over‐hydration and mechanical weakening, leading to severe dimensional instability and reduced durability. Therefore, recent AEM research has focused on overcoming this intrinsic trade‐off between ion transport efficiency and mechanical stability rather than merely optimizing one property.[ 9 , 10 ] Crosslinking has been recognized as a rational strategy that enables the realization of high IEC and excellent dimensionalstability simultaneously.[ 11 15 ] Depending on the design requirements, crosslinking can be introduced during membrane fabrication (in situ crosslinking)[ 16 18 ] or through post‐functionalization techniques involving multifunctional crosslinkers,[ 19 , 20 ] thermal treatment,[ 21 23 ] or UV irradiation.[ 24 , 25 ] Common approaches include the use of di‐ or multi‐functional monomers in free‐radical polymerization,[ 26 28 ] incorporation of reactive pendant groups (e.g., allyl, vinyl, azido, thiol or epoxy) followed by chemical crosslinking,[ 29 , 30 ] or chain‐growth polymerization reactions[ 21 , 31 , 32 ] to construct crosslinked AEMs. The resulting crosslinked structure restricts polymer chain rearrangement, thereby enabling AEMs to simultaneous achieve dimensional stability and efficient hydroxide transport under high IEC conditions.

Cyclopolymerization is a unique type of polymerization in which monomers with two reactive sites undergo intramolecular cyclization during the propagation step, resulting in the formation of fused ring structures along the polymer backbones.[ 33 , 34 ] In particular, the free‐radical cyclopolymerization of diallylammonium monomers, known as Butler's cyclopolymerization, has been employed as an effective synthetic route to cationic polyelectrolytes.[ 33 , 35 37 ] The formation of the polymer backbone comprising quaternized pyrrolidinium rings imparts high charge density (i.e., high IEC) and thermal stability, while the geometrically constrained ring structure provides excellent resistance to alkaline degradation, making these polymers attractive for AEMWE applications. However, their high hydrophilicity often results in excessive water solubility, thereby limiting their direct use as AEMs without further structural reinforcement.[ 38 ]

Here, we reported a novel class of crosslinked AEMs synthesized via in situ crosslinking through free‐radical cyclopolymerization of diallylammonium‐functionalized aromatic polymer. Our crosslinking approach introduces covalent junctions through in situ cyclopolymerization of diallylammonium units, which simultaneously (i) increase the concentration of cationic sites (raising IEC), (ii) generates a 3D polymer network that physically restricts swelling, and (iii) enhance mechanical integrity even under hydrated conditions. In this approach, IEC of the membrane was precisely controlled by adjusting the content of diallyldimethylammonium comonomer introduced during the growth of poly (diallylammonium) chains, enabling the fabrication of high‐IEC AEMs with values reaching up to 3.39 meq g−1. The unique crosslinked structure, combined with the aryl ether‐free aromatic backbone and the geometrically constrained pyrrolidinium rings of the poly (diallylammonium) chains, imparted mechanical robustness and favorable membrane properties, including high water uptake, controlled swelling ratio, high hydroxide conductivity, and excellent alkaline stability. Furthermore, the AEMWE performance of the crosslinked membranes was evaluated using membrane‐electrode assembly tests with TMA‐70 as the ionomer, demonstrating excellent current density and long‐term durability, thereby, rendering them as a promising candidate for next‐generation AEMWE applications.

2. Results and Discussion

As shown in Figure 1a, poly(biphenyl alkylene) (BP‐HBrF3) was synthesized via Friedel‐Craft polyhydroxyalkylation of 7‐bromo‐1,1,1‐trifluoroheptan‐2‐one (HBrF3) with biphenyl (BP) monomers, yielding a high molecular weight (Mw ) of 314,800 g mol−1 with dispersity (Ð) of 2.88 (Figure 1b).[ 39 ] The well‐defined structure of aryl ether‐free aromatic backbones composed of biphenyl units and a bromopentyl‐tethered quaternary carbon was confirmed by both 1H and 19F nuclear magnetic resonance (NMR) spectroscopies (Figure 1c; Figure S1, Supporting Information). Subsequently, the bromo group on the side chain was converted into diallylmethylammonium group via Menshutkin reaction, affording diallylammonium‐functionalized poly(biphenyl alkylene) (BP‐DAMA). 1H NMR spectrum of the quanternized polyaromatic displays the appearance of proton signals corresponding to the allyl group (5.6 and 6.0 ppm) and the methyl group (2.9 ppm) (Figure 1d).

Figure 1.

Figure 1

a) Synthetic route for the preparation of diallylammonium‐functionalized poly(biphenyl alkylene) (BP‐DAMA). b) Gel permeation chromatography (GPC) chromatogram and c) 19F‐NMR spectrum of BP‐HBrF3. d) 1H‐NMR spectrum of BP‐DAMA.

We envisioned that cyclopolymerization of diallylmethylammonium groups on the side chains, in the presence of diallyldimethylammonium chloride (DADMAC) as a comonomer, would induce in situ crosslinking of BP‐DAMA, resulting in the formation of mechanically robust polyaromatic membranes (Figure 2a). Initially, a solution containing BP‐DAMA, DADMAC, and azobisisobutyronitrile (AIBN) as a radical initiator was cast onto a mold. Upon heating to 80 °C, thermal decomposition of AIBN generated radicals that added to one of the double bonds of the diallylammonium groups, forming a carbon‐centered radical. This radical subsequently underwent intramolecular cyclization via attack on the adjacent double bond, yielding a 5‐membered pyrrolidinium ring structure. Successive intramolecular cyclization promoted the chain growth of poly(diallylammonium) (PDAA), while the connection of DAMA groups embedded within the BP chains facilitated crosslinking, leading to the formation of crosslinked polyaromatic anion exchange membranes (BP‐cross‐PDAAs). Following detachment from the mold and immersion in 1 m KOH solution to exchange counterions with hydroxide, transparent yellow and flexible films of BP‐cross‐PDAA were obtained (Figure 2b).

Figure 2.

Figure 2

a) Formation of crosslinked AEMs (BP‐cross‐PDAA) via free‐radical cyclopolymerization. b) Photograph of BP‐cross‐PDAA‐3.39 membrane. SEM images of the c) surface and d) cross‐section of BP‐cross‐PDAA‐3.39 membrane.

The successful cyclopolymerization and network formation was verified by FT‐IR analysis, which showed that the = CH2 stretching band (≈3083 cm−1) almost disappeared after crosslinking (Figure S2, Supporting Information),[ 40 ] indicating that a near‐complete conversion of the allyl groups into cyclic structures was achieved during the cyclopolymerization process. Simultaneously, the distinct characteristic peak of N─C stretching band (≈1325 cm−1) provides compelling spectroscopic evidence for the formation of the pyrrolidinium ring within the polymer backbone.[ 41 ] In addition, as shown in Figure 2c,d, both the surface and cross‐sectional SEM images of BP‐cross‐PDAA reveal dense, defect‐free morphologies without voids or discontinuities arising from inhomogeneous crosslinking during cyclopolymerization. Furthermore, to demonstrate the successful formation of the crosslinked membrane, we evaluated the degree of crosslinking of the BP‐cross‐PDAA membranes using a gel fraction experiment. The gel fraction was determined by immersing the membranes in DMSO for 24 h to remove any soluble components, followed by complete drying in an oven to obtain the residual dry mass. When compared with the initial dry mass of the membrane, all BP‐cross‐PDAA membranes exhibited gel fractions above 87% (Table S1, Supporting Information), indicating that the membranes retain an almost entirely insoluble fraction after extraction. These results confirm the high degree of crosslinking achieved through the free‐radical cyclopolymerization process.

In our approach, precise control over the IEC was achieved by adjusting the DADMAC content during cyclopolymerization. With increasing DADMAC incorporation between crosslinking points, the theoretical IEC of BP‐cross‐PDAA is predicted to approach the maximum value characteristic of Polydiallyldimethylammonium chloride (PDADMAC) homopolymer (6.98 meq g−1). In this study, to avoid significant deterioration in mechanical strength under wet conditions, five membranes (BP‐cross‐PDAA‐n) were prepared with the IEC values (n) ranging from 2.32 meq g−1 to 3.39 meq g−1. The experimental IEC values of the membranes, determined by acid‐base titration, were in good agreement with the corresponding theoretical values (Table S1, Supporting Information).

For practical application of BP‐cross‐PDAA in AEMWEs, we evaluated the thermal and mechanical properties of the crosslinked AEMs. Thermogravimetric analysis (TGA) revealed that the BP‐cross‐PDAA membranes exhibited excellent thermal stability without noticeable decomposition up to 200 °C, followed by a distinct two‐step degradation behavior under N2 (Figure S3a, Supporting Information). The first weight‐loss step occuring ≈200 °C is attributed to the decomposition of the pyrrolidinium rings via ring‐opening and C─N bond cleavage. The second major mass‐loss stage corresponds to the thermal degradation of the aromatic polymer backbone, which represents the main structural decomposition of the membrane. Tensile tests were conducted to assess the mechanical strength of BP‐cross‐PDAA membranes under wet conditions (Figure S3b, Supporting Information). As the IEC increased, the tensile strength decreased from 13.2 MPa for BP‐cross‐PDAA‐2.32 to 5.78 MPa for BP‐cross‐PDAA‐3.39. This reduction in strength is attributed to the enhanced water adsorption of high‐IEC membranes, which leads to a decrease in effective crosslinking density.

The hydration and ion transport properties of BP‐cross‐PDAA, including water uptake, swelling ratio, KOH uptake, and hydroxide conductivity, were systematically examined to evaluate dimensional stability and ion conductivity performance (Figure  3 ; Table 1). Water uptake and swelling ratio were determined by measuring the changes in weight and length of dried membranes after immersion in water at room temperature for 24 h. As shown in Figure 3a, water uptake of BP‐cross‐PDAA increased from 20.8% to 92.1% as the IEC increased from 2.32 meq g−1 to 3.39 meq g−1. Notably, BP‐cross‐PDAA‐3.39 exhibited a swelling ratio of 24.2%, which reflects the effectiveness of the in situ cyclocrosslinking strategy in maintaining dimensional stability given its high IEC (3.39 meq g−1, OH form) and water uptake (92.1%). While high‐IEC AEMs often suffer from excessive swelling and consequent mechanical degradation, the crosslinked structure of BP‐cross‐PDAA could suppressed volumetric expansion by restricting polymer chain mobility, thereby enabling the simultaneous achievement of high water uptake and controlled swelling ratio. This swelling behavior was further evaluated in comparison with state‐of‐the‐art AEMs reported in literature (Figure 3b).[ 42 57 ] Furthermore, we measured the temperature‐dependent water uptake and swelling ratio of BP‐cross‐PDAA membranes over the range of 25–80 °C (Figure S4, Supporting Information). Both water uptake and swelling ratio increase gradually with increasing temperature for all membranes, consistent with enhanced hydration at elevated temperatures. Among the series, BP‐cross‐PDAA‐3.39 exhibits the highest water uptake (116.9%) and swelling ratio (33.3%) at 80 °C, which reflects the expected influence of its higher IEC. The moderate swelling ratios observed for the BP‐cross‐PDAA membranes at temperatures higher than 25 °C provide additional experimental support for the role of the crosslinked network in moderating volumetric expansion under hydrated conditions, even at high IEC. In addition, KOH uptake of BP‐cross‐PDAA was measured, which showed an increase from 52.6% for BP‐cross‐PDAA‐2.32 to 137.4% for BP‐cross‐PDAA‐3.39 with increasing IEC (Figure 3c).

Figure 3.

Figure 3

Characterization of BP‐cross‐PDAA‐n membranes, including water uptake, dimensional stability, and ionic conductivity as functions of IEC. a) Water uptake and swelling ratio, b) Comparison of the swelling ratio—IEC relation with representative AEMs in recent reports,[ 42 57 ] and c) KOH uptake. d) Temperature‐dependent hydroxide conductivity under fully hydrated conditions. e) Hydroxide conductivity at 80 °C versus IEC. f) Conductivity retention of BP‐cross‐PDAA‐3.39 after immersion in 1 m KOH at 60 °C.

Table 1.

Hydration and ion transport properties of BP‐cross‐PDAA membranes.

Sample Water uptake (%) b) Swelling ratio (%) c) KOH uptake (%) d) Ionic conductivity (mS cm−1) e)
25 °C 80 °C
BP‐cross‐PDAA‐2.32 20.8 (±5.05) 9.5 (±1.12) 52.6 (±4.61) 27.55 (±8.31) 90.35 (±12.90)
BP‐cross‐PDAA‐2.61 38.4 (±3.87) 14.0 (±2.81) 64.8 (±8.20) 38.42 (±6.05) 111.92 (±8.69)
BP‐cross‐PDAA‐2.88 47.9 (±4.50) 14.9 (±1.88) 81.3 (±9.10) 45.69 (±4.78) 124.58 (±6.16)
BP‐cross‐PDAA‐3.19 62.3 (±4.64) 22.2 (±1.45) 106.7 (±6.78) 50.36 (±5.51) 132.4 (±7.35)
BP‐cross‐PDAA‐3.39 92.1 (±6.11) 24.2 (±0.96) 137.4 (±9.56) 56.2 (±5.11) 152.4 (±14.70)
a)

Measured by acid‐base titration,

b,c,d)

Measured at 25 °C,

e)

Measured at 100% RH.

Hydroxide conductivity of BP‐cross‐PDAA was evaluated over a temperature range of 25 to 80 °C under fully hydrated condition (Figure 3d). All BP‐cross‐PDAA membranes exhibited enhanced hydroxide mobility with rising temperature. In addition, an increase in hydroxide conductivity of BP‐cross‐PDAA was also observed with increasing IEC from 2.32 to 3.39 across the entire temperature range. Temperature‐dependent hydroxide ion conductivity of the BP‐cross‐PDAA‐n membranes were analyzed and plotted the corresponding Arrhenius relationships (Figure S5, Supporting Information). The activation energy (Ea) was obtained in the range of 15–18 kJ mol−1. At 80 °C, BP‐cross‐PDAA‐3.39 exhibited the highest hydroxide conductivity of 152.4 mS cm−1 and, as the IEC of BP‐cross‐PDAA‐n decreased, the hydroxide conductivity decreased to 132.4 mS cm−1 (n = 3.19), 124.58 mS cm−1 (n = 2.88), 111.92 mS cm−1 (n = 2.61), and 90.35 mS cm−1 (n = 2.32) (Figure 3e). These results clearly demonstrate that hydroxide conductivity increases proportionally with IEC within the BP‐cross‐PDAA‐n series, confirming the positive IEC‐σ relationship shown in Figure S6 (Supporting Information).[ 58 63 ] The alkaline stability of BP‐cross‐PDAA was investigated by measuring hydroxide conductivity after immersion in 1 m KOH at 60 °C. As shown in Figure 3f, negligible change in conductivity (from 152.4 mS cm−1 to 150.4 mS cm−1, 98.7%) of BP‐cross‐PDAA‐3.39 was observed after immersion for 350 h. Furthermore, under intentionally accelerated degradation conditions (2 m KOH, 60 °C), the conductivity gradually decreased to 86.4% after 234 h, accompanied by a reduction (15%) in hydrated thickness (Figure S7, Supporting Information). This concurrent loss in conductivity and thickness indicates partial side‐chain scission and matrix relaxation—typical degradation pathways of quaternary ammonium hydrocarbon AEMs when exposed to highly caustic environments.[ 10 ] Together, these findings underscore that the in situ cyclopolymerization strategy affords a well‐controlled balance between hydroxide conductivity and structural stability.

These results collectively indicate that BP‐cross‐PDAA, enabled by its crosslinked structure and the geometrically constrained pyrrolidinium ring, maintains reliable ion transport properties and steady alkaline tolerance without excessive dimensional change, even at IEC values above 3.0 meq g−1. This balanced combination of conductivity retention and structural control suggests that BP‐cross‐PDAA is a viable membrane platform for high‐performance AEM applications.

To demonstrate the applicability of BP‐cross‐PDAA for AEMWE applications, the performance of membrane electrode assemblies (MEA‐n), fabricated using BP‐cross‐PDAA‐n membranes and TMA‐70 ionomer, was assessed under operating conditions of 1 m KOH at 80 °C. As shown in the linear sweep voltammetry (LSV) curves (Figure 4a,b), the current density increased with IEC, rising from 2.46 A cm−2 to 5.07 A cm−2 at 1.8 V and from 4.54 A cm−2 to 9.98 A cm−2 at 2.0 V. This improvement is attributed to the enhanced ion conductivity and increased water uptake of BP‐cross‐PDAA at higher IEC values, which collectively facilitate efficient hydroxide transport within the membrane.

Figure 4.

Figure 4

Electrochemical characterizations of AEM water electrolyzers using BP‐cross‐PDAA‐n membranes with various IEC values. a) Linear sweep voltammetry (LSV) curves at 80 °C 1 m KOH. b) Current densities at 1.8 and 2.0 V as a function of IEC. c) Electrochemical impedance spectroscopy (EIS) Nyquist plots at 1.8 and 2.1 V. d) Extracted resistances (Rohm, Rct, Rmt) at 1.8 V versus IEC. e) Chronopotentiometry of MEA‐3.39 at 1 A cm−2 in 50 °C 1 m KOH for 1,000 h. LSV curves of MEA‐3.39 with IrO2, NiFe, and Co anodes f) in 1 M KOH, and g) in pure water. h) Comparison of current density at 2.0 V with state‐of‐the‐art AEMWE systems.[ 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 ]

BP‐cross‐PDAA/TMA‐70 MEA was further investigated by electrochemical impedance spectroscopy (EIS) analysis and represented as a Nyquist plot (Figure 4c,d). Compared to the other MEAs, MEA‐2.32 exhibited much higher ohmic resistance (Rohm), charge transfer resistance (Rct), and mass transport resistance (Rmt). These significant performance improvements observed in the other MEAs are attributed to the incorporation of DACMAC into the PDAA chain, which simultaneously increases the IEC and imparts a soft, conformable nature to the BP‐cross‐PDAA membranes. This dual effect enhances hydroxide conductivity and promotes strong interfacial adhesion with the catalyst layer, thereby reducing contact resistance and enabling a well‐integrated three‐phase interface.

To evaluate the durability of MEA‐n, a chronopotentiometry (CP) test was conducted at a constant current density of 1 A cm−2 in 1 m KOH at 50 °C. As shown in Figure 4e, MEA‐3.39 exhibited no appreciable change in cell voltage over 1000 h, with a minimal voltage degradation rate of 1.2 µV h−1. This remarkable long‐term durability is attributed not only to the alkaline‐stable pyrrolidinium rings embedded within the crosslinked BP‐cross‐PDAA membrane, but also to the enhanced dimensional stability arising from its moderate swelling behavior under hydrated condition. Figure 4f,g present the LSV curves obtained using various anode catalysts (IrO2, NiFe, and Co) paired with BP‐cross‐PDAA‐3.39 membrane under different feed conditions. In 1 m KOH, current density at 1.8 V reached 5.1 A cm−2, 6.1 A cm−2, and 1.6 A cm−2 for IrO2‐, NiFe‐, Co‐based anodes, respectively, highlighting the favorable kinetics achieved with high‐activity catalysts in alkaline media (Figure 4f). Furthermore, as shown in Figure S8 (Supporting Information), the BP‐cross‐PDAA‐3.39 membrane exhibited higher current densities and lower cell resistances than the commercial PiperION, and Xion AEM‐Pention‐72–15CL membranes under identical AEMWE operating conditions. These results verify that the superior ionic transport properties of BP‐cross‐PDAA translate directly into improved single cell performance, underscoring its competitiveness for practical water electrolysis systems. Upon switching the feed to pure water (Figure 4g), the current density at 1.8 V exhibited a modest decline to 2.4 A cm−2 for IrO2, 3.1 A cm−2 for NiFe, and 0.8 A cm−2 for Co. Despite the reduction in performance, these results demonstrate the capability of BP‐cross‐PDAA for efficient AEMWE operation under pure water condition.

To highlight the robust AEMWE performance of the BP‐cross‐PDAA/TMA‐70 MEA system, we compared the current density values at 1.8 and 2.0 V as a function of IEC with those of state‐of‐the‐art AEMWE systems (Figure 4 h; Figure S9 and Table S2, Supporting Information).[ 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53 , 54 ] Under IEC values below 3.0 meq g−1, MEA‐n exhibited excellent current densities, while few studies reporting comparable performance in this IEC range. Notably, MEA‐3.39 demonstrated an outstanding current density of 12.39 A cm−2 at 2.0 V, which, to the best of our knowledge, represents the highest AEMWE performance reported in recent literatures. These results demonstrate that the rational design of BP‐cross‐PDAA membrane, when combined with appropriate catalyst systems, enables efficient hydrogen production under both alkaline and pure water conditions.

3. Conclusion

In summary, we developed a new series of crosslinked polyaromatic anion exchange membranes (BP‐cross‐PDAA) through free‐radical cyclopolymerization of diallylammonium‐functionalized poly(biphenyl alkylene) with a DADMAC comonomer. By adjusting the DADMAC content, the IEC (n) of BP‐cross‐PDAA‐n was precisely tuned from 2.32 meq g−1 to 3.39 meq g−1. The resulting membranes demonstrated excellent hydration and ion transport properties (e.g., water uptake, membrane swelling, KOH uptake, and hydroxide conductivity) as well as acceptable alkaline stability (98.7% retention over 350 h). Notably, among the BP‐cross‐PDAA series, BP‐cross‐PDAA‐3.39 simultaneously achieved high hydroxide conductivity (152.4 mS cm−1 at 80 °C) and controlled membrane swelling (24.2%), balancing the typical trade‐off between ion conductivity and mechanical integrity found in conventional AEMs. Furthermore, the AEMWE performance of BP‐cross‐PDAA membranes was evaluated by MEA test using TMA‐70 as an ionomer. Remarkably, MEA‐3.39 (BP‐cross‐PDAA‐3.39/TMA‐70 MEA) exhibited an exceptional current density of 12.39 A cm−2 at 2.0 V, which, as far as we know, surpasses all previously reported AEMWE performances. This robust performance is attributed to the structural design of the BP‐cross‐PDAA membrane, in which DADMAC was incorporated into the PDAA chain, simultaneously increasing the IEC and conferring softness and conformability to the membrane, thereby, reducing contact resistance and facilitating charge transfer. Moreover, MEA‐3.39 demonstrated excellent durability over 1000 h at 1 A cm−2 in 1 m KOH at 50 °C, with a minimal voltage degradation rate of 1.2 µV h−1, while achieving a high current density of 12.39 A cm−2 at 2.0 V in 1 m KOH with non‐platinum group metal anodes and 6.1 A cm−2 at 2.0 V under pure water operations. These outstanding AEMWE performance metrics and long‐term durability of the BP‐cross‐PDAA/TMA‐70 MEA system highlights its great potential for future integration into AEMWE energy applications.

4. Experimental Section

Materials

For membrane preparation, biphenyl (BP, >99%), diallylmethylamine (97%), diallyldimethylammonium chloride (DADMAC, >97%), butylated hydroxytoluene (BHT, 99%) and dimethyl sulfoxide‐d6 (DMSO‐d6, 99.9 atom% D) were supplied by Sigma Aldrich Chemical Co. Ltd and used as received. 7‐Bromo‐1,1,1‐trifluoroheptan‐2‐one (95%) was supplied by Habotech and used as received. Potassium hydroxide (KOH, 93%), dichloromethane (DCM, 99%), trifluoromethane sulfonic acid (TFSA, triflic acid, 99%), N‐methyl‐2‐pyrrolidone (NMP, 99%), n‐hexane (98.5%), dimethyl sulfoxide (DMSO, 99.5%), 2,2′‐azobis(2‐methylpropionitrile) (AIBN, 99%), tetrahydrofuran (THF, 99.5%), and methyl alcohol (MeOH, 99.5%) were obtained from Daejung and used as received.

For membrane electrode assembly fabrication, platinum ruthenium (Pt‐50 wt.%, Ru‐25 wt.%, 04 7371.06) on high surface area advanced carbon support, and iridium(IV) oxide (99.99 wt.%, 04 3396.06), cobalt nanopowder (99.8%, 46 347), nickel(II) chloride (NiCl2), and iron(III) chloride (FeCl3), Sodium borohydrate (NaBH4) were received from Alfa Aesar.

Synthesis of BP‐HBrF3

TFSA was used as an acid catalyst for polyhydroxyalkylation reaction using biphenyl and 7‐bromo‐1,1,1‐trifluoroheptane‐2‐one. First, biphenyl (1.34 g, 8.69 mmol) was dissolved in 8.3 mL of DCM with 7‐bromo‐1,1,1‐trifluoroheptane‐2‐one (2.37 g, 9.59 mmol). After placing the round bottle in an ice‐bath, 8.3 mL of TFSA was added slowly all at once. When the reaction was conducted for 24 h at room temperature while stirring, the reaction solution turns into a dark green color with high viscosity. After precipitating the viscous polymer fraction in excess methanol, the obtained polymer is once again dissolved in THF and precipitated in methanol to purify to obtain ivory colored BP‐HBrF3, followed by vacuum drying at 80 °C.

Synthesis of BP‐DAMA

A diallylmethylamine group was introduced into BP‐HBrF3 through the SN2 reaction of an alkyl halide and a tertiary amine. Diallylmethylamine (0.58 g, 5.22 mmol) was slowly added to BP‐HBrF3 (1 g, 2.61 mmol) dissolved in 7 mL of NMP, and the reaction mixture was stirred at 50 °C for 72 h. After the completion of the reaction was confirmed by 1H‐NMR, the reaction solution is precipitated in n‐hexane, washed several times, and vacuum dried at room temperature.

Preparation of BP‐cross‐PDAA Membrane

To prepare BP‐cross‐PDAA membrane with IEC 3.39 mmol g−1, BP‐DAMA (1.39 g, 2.82 mmol) was dissolved in 12.53 g of DMSO. DADMAC (0.41 g, 2.53 mmol) was added to the solution followed by AIBN (0.007 g) as an initiator. After sufficiently mixing the solution, it was poured onto a mold and polycyclopolymerization induced in a 80 °C oven for 72 h to obtain membranes. The fabricated membrane was immersed in 1 m KOH for one day before use to exchange the counter ion into hydroxide form. The IEC of the membrane was tuned by varing the molar ratio of DADMAC to BP‐DAMA in the polymerization mixture. Specifically, BP‐cross‐PDAA membranes with IEC values of 2.32, 2.61, 2.88, 3.19, and 3.39 meq g−1 were prepared by adjusting the feed ratio of DADMAC/ BP‐DAMA to 0, 0.25, 0.5, 0.75, 0.9 mol/mol, respectively.

Characterization and Measurements

1H‐NMR, and 19F‐NMR were measured at 25 °C using Varian Unity INOV. Fourier‐transform infrared spectroscopy (FT‐IR) spectra was obtained using PerkinElmer FT‐IR system (Spectrum‐GX) to analyze the structure of AEMs. The thermal stability of the PDAA based membranes were examined by using TA Instrument TGA 2950 with heating rate of 10 °C min−1 under N2 atmosphere. Mechanical property of the AEMs were characterized using universal tensile machine (Tinius Olsen H5K‐T). Membrane samples (4 cm × 0.5 cm) were prepared to measure the mechanical strength and tested at stretching speed of 10 mm min−1. Field Emission Scanning Electron Microscopy (FE‐SEM) images to verify morphology and shape of patterned Cu, was taken on an Inspect F50 (FEI, Korea).

Gel fraction was determined from the residual weight of the membranes after immersion in DMSO for 24 h followed by complete drying in an oven. The gel fraction was calculated using the following equation:

Gelfraction%=mr/mi100 (1)

where mi is the initial dry mass of the membranes before extraction, and mr is the dry mass of membranes after extraction and subsequent drying.

Ion exchange capacity (IEC) value was determined by the back titration. OH‐ form membranes were immersed in 0.01 m HCl solution for 24 hrs. The HCl solution was titrated with 0.01 m KOH solution after adding three drops of phenolphthalein/EtOH indicator solution. The IEC (meq g−1) was calculated based on the following equation:

IEC=VHClCHClVNaOHCNaOH/Wdry (2)

where VHCl and VNaOH are volume of HCl and NaOH, respectively. CHCl and CNaOH are the concentration of HCl and NaOH, respectively. Wdry is the weight of the dried membranes.

The swelling ratio (SR) and water uptake (WU) were evaluated after immersing the membrane in water over the temperature range of 25–80 °C for 24 hrs. After removing the membrane from the water, excess water on the surface was carefully wiped off and the weight and length was measured for comparison. After membrane was completely dried, the weight and length of the dried membrane were also measured. As the result, SR and WU were calculated according to the following formula:

SR%=LwetLdry/Ldry100 (3)
WU%=WwetWdry/Wdry100 (4)

The KOH uptake rate is also tested in the same way. After the preparation of the membrane, the membrane was immersed in 1 m KOH solution at room temperature for 24 hrs. The weight of the membrane was measured after removing the excess KOH solution on the surface of the membrane. After washing several times with water and drying, the weight of dried membrane was measured and compared. KOH uptake was calculated using the following equation.

KOHuptake=mwet,KOHmdry/mdry100 (5)

The ohmic resistance of the AEMs were measured in water and 0.1–5 m KOH solution at different temperature by two‐electrode electrochemical impedance spectroscopy (EIS, SI 1260, Solartron) over the frequency range from 10 Hz to 10 MHz with an amplitude of 20 mV. Before measurement, all OH‐ form membranes were immersed in distilled water for at least 24 hrs to equilibrate in water. Hydroxide conductivity was calculated by the following equation:

σ=L/R×W×d (6)

where L was the distance between two electrodes (cm), R was the measured ohmic resistance (Ω), W was the width of the membrane (cm) and d was the width of the membrane (cm).

Synthesis of NiFe non‐PGM Catalyst

A stock solution of 2 m metal chloride was prepared by dissolving 2.592 g NiCl2 and 3.244 FeCl3 in 10 mL deionized water to form a dark green and dark yellow solution, respectively. The 2 m NaBH4 was freshly prepared by dissolving 0.378 g powder in 5 mL deionized water. A mixed solution of 0.665 mL of NiCl2 and 0.335 mL of FeCl3 was quickly injected into the prepared NaHB4 solution with stirring at room temperature. The stirring was maintained for ≈1 min, and the black dispersion was set aside until the next day. Black Ni3Fe1 nanofoams were precipitated at the bottom of the container, and the supernatant was colorless. After freeze‐drying, 3D Ni3Fe1 nanofoams were obtained. The synthetic method was modified based on previously reported procedure.[ 57 ]

MEA Fabrication and Electrolysis Performance

PDAA AEM based MEAs (1 cm2 active area) were fabricated by the catalyst coated substrate (CCS) method. 50 wt.% PtRu/C was used as cathode and IrO2, NiFe, and Co were used as anode catalysts, respectively. Catalyst inks were prepared by dispersing each catalyst powder over the TMA‐70 ionomer. (4.65 wt.% TMA‐70 solution in n‐propanol: aqueous 5:5 wt.% co‐solution, the ionomer to carbon (I/C) weight ratio is 0.5, and for the OER catalyst layer the binder content was 10 wt.%) in an aqueous solution of n‐propanol and little amount of water, followed by ultrasonic treatment for more than 20 min with maintaining water temperature less than 35 °C to prevent catalyst agglomeration. All BP‐cross‐PDAA‐n membrane series were used thickness range of 35–40 µm to avoid artificially enhanced current densities that can occur with excessively thin membranes. Synthetic route and NMR analysis of TMA‐70 are shown in Figure S8 (Supporting Information). The prepared catalyst inks were directly sprayed onto the gas diffusion layer (GDL) set on a 70 °C pre‐heated hotplate. Ni fiber paper (Dioxide materials) were used for the cathode and anode as a porous transport layer (PTL). Cathode catalyst loading amount is 0.6 mgPt cm−2, and anode catalyst loading amount is 2.0 mgIr cm−2, 3.0 mgNiFe cm−2, 0.7 mgCo cm−2. The fabricated CCS was dried at room temperature for more than 1 h to remove residual solvent in the catalyst layers. Prior to single cell application, the PDAA based membranes were sandwiched by the fabricated electrode without hot‐pressing process. Polarization curves were obtained on Scribner electrolyzer cell with power supplied by Biologic potentiostat (SP‐200) attached with power booster (HCV‐3048) with electrolyte supplied with peristaltic pump.

Conflict of Interest

The authors declare no conflict of financial interest.

Author Contributions

J.J. Y.S.P. and M.B.K. contributed equally to this work. The manuscript was written through contributions of all authors.

Supporting information

Supporting Information

Acknowledgements

A.S. Lee acknowledges support by the Nano & Material Technology Development Program through the National Research Foundation of Korea (NRF) funded by Ministry of Science and ICT (RS‐2024‐00450102) with 50% contribution, the National Research Council of Science & Technology (NST) grant by the Korea government (MSIT) (CRC22031‐000) and the institutional program at KIST (2V10154).

Jung J., Park Y. S., Koo M. B., et al. “In Situ Crosslinked Diallylammonium‐Functionalized Poly(Biphenyl Alkylene) for High‐Performance Anion Exchange Membranes.” Small 22, no. 5 (2026): e11358. 10.1002/smll.202511358

Contributor Information

Hoon Taek Chung, Email: hchung@sejong.ac.kr.

Albert S. Lee, Email: aslee@kist.re.kr.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  • 1. Liu R. T., Xu Z. L., Li F. M., Chen F. Y., Yu J. Y., Yan Y., Chen Y., Xia B. Y., Chem. Soc. Rev. 2023, 52, 5652. [DOI] [PubMed] [Google Scholar]
  • 2. Ma L., Sui S., Zhai Y., Int. J. Hydrog. Energy 2009, 34, 678. [Google Scholar]
  • 3. Tüysüz H., Acc. Chem. Res. 2024, 57, 558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Du N., Roy C., Peach R., Turnbull M., Thiele S., Bock C., Chem. Rev. 2022, 122, 11830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Palmas S., Rodriguez J., Mais L., Mascia M., Herrando M., Vacca C. A., Curr. Opin. Electrochem. 2023, 37, 101178. [Google Scholar]
  • 6. Falcão D. S., Energies 2023, 16, 943. [Google Scholar]
  • 7. Ge X., Zhang F., Wu L., Yang Z., Xu T., Macromolecules 2022, 55, 3773. [Google Scholar]
  • 8. Li D., Motz A. R., Bae C., Fujimoto C., Yang G., Zhang F. Y., Ayers K. E., Kim Y. S., Energy Environ. Sci. 2021, 14, 3393. [Google Scholar]
  • 9. Hu C., Wang, Y. W. , Lee Y. M., Angew. Chem., Int. Ed. 2025, 64, 202418324. [DOI] [PubMed] [Google Scholar]
  • 10. Park E., Jannasch P., Miyatake K., Bae C., Noonan K., Fujimoto C., Holdcroft S., Varcoe J., Henkensmeier D., Guiver M., Kim Y. S., Chem. Soc. Rev. 2024, 53, 5704. [DOI] [PubMed] [Google Scholar]
  • 11. Clemens A. L., Jayathilake B. S., Karnes J. J., Schwartz J. J., Baker S. E., Duoss E. B., Oakdale J. S., Polym. J. 2023, 15, 1534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Hu C., Deng X., Dong X., Hong Y., Zhang Q., Liu Q., J. Membr. Sci. 2021, 619, 118806. [Google Scholar]
  • 13. Barnett A., Karnes J. J., Clemens A. L., Oakdale J. S., Molinero V., J. Phys. Chem. C 2023, 127, 5613. [Google Scholar]
  • 14. Kwon S., Rao A. H. N., Kim T. H., J. Power Sources 2018, 375, 421. [Google Scholar]
  • 15. Mandal M., Huang G., Kohl P. A., ACS Appl. Energy Mater. 2019, 2, 2447. [Google Scholar]
  • 16. Guo M., Wang Z., Xu Y., Zhu X., J. Membr. Sci. 2024, 707, 123026. [Google Scholar]
  • 17. Liu G., Wang A., Ji W., Zhang F., Wu J., Zhang T., Tang H., Zhang H., Chem. Eng. J. 2023, 454, 140046. [Google Scholar]
  • 18. Arunkumar I., Rana M. M., Jeevitha A., Gokulapriyan R., Yu J. H., Yoo D. J., ACS Sustainable Chem. Eng. 2024, 12, 9361. [Google Scholar]
  • 19. Koronka D., Mahmoud A. M. A., Miyatake K., Polym. Sci., Part A: Polym. Chem. 2019, 57, 1059. [Google Scholar]
  • 20. Wang X., Thomas A. M., Lammertink R. G. H., ACS Appl. Mater. Interfaces 2024, 16, 2593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Wu L., Pan Q., Varcoe J. R., Zhou D., Ran J., Yang Z., Xu T., J. Membr. Sci. 2015, 490, 1. [Google Scholar]
  • 22. Golubenko D., Ahmed F. E., Hilal N., Membranes 2024, 14, 138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Mahmoud A. M. A., Miyatake K., Liu F., Yadav V., Guo L., Wong C. Y., Iwataki T., Kakinuma K., Uchida M., Adv. Energy Sustainability Res. 2024, 5, 2300236. [Google Scholar]
  • 24. Clemens A. L., Cerón M. R., Ferron T. J., Mettry‐Yassa M., Hammons J. A., Jayathilake B., Molinero V., Karnes J. J., Oakdale J. S., ACS Appl. Polym. Mater. 2024, 6, 4419. [Google Scholar]
  • 25. Seo J., Kushner D. I., Hickner M. A., ACS Appl. Mater. Interfaces 2016, 8, 16656. [DOI] [PubMed] [Google Scholar]
  • 26. Lin B., Qiu L., Lu J., Yan F., Chem. Mater. 2010, 22, 6718. [Google Scholar]
  • 27. Qiu B., Lin B., Qiu L., Yan F., J. Mater. Chem. 2012, 22, 1040. [Google Scholar]
  • 28. Xu F., Qiu K., Lin B., Ren Y., Li J., Ding J., Hickner M. A., J. Membr. Sci. 2021, 637, 119629. [Google Scholar]
  • 29. Hu B., Miao L., Bai Y., Lü C., Polym. Chem. 2017, 8, 4403. [Google Scholar]
  • 30. Lowe A. B., Polym. Chem. 2010, 1, 17. [Google Scholar]
  • 31. Liu Y., Pan Q., Wang Y., Zheng C., Wua L., Xu T., Sep. Purif. Technol. 2015, 156, 226. [Google Scholar]
  • 32. He Y. B., Wu L., Pan J. F., Zhu Y., Ge X. L., Yang Z. J., Ran J., Xu T. W., J. Membr. Sci. 2016, 504, 47. [Google Scholar]
  • 33. Pasini D., Takeuchi D., Chem. Rev. 2018, 118, 8983. [DOI] [PubMed] [Google Scholar]
  • 34. Gao Y., Zhou D., Lyu J., S. A., Xu Q., Newland B., Matyjaszewski K., Tai H., Wang W., Nat. Rev. Chem. 2020, 4, 194. [DOI] [PubMed] [Google Scholar]
  • 35. Butler G. B., Angelo R. J., J. Am. Chem. Soc. 1957, 79, 3128. [Google Scholar]
  • 36. Butler G. B., J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 3451. [Google Scholar]
  • 37. Biery A. R., Knauss D. M., Mater. Today Chem. 2022, 26, 101251. [Google Scholar]
  • 38. Olsson J. S., Pham T. H., Jannasch P., Macromolecules 2017, 50, 2784. [Google Scholar]
  • 39. Lee W. H., Kim Y. S., Bae C., ACS Macro Lett. 2015, 4, 814. [DOI] [PubMed] [Google Scholar]
  • 40. Mahdavi H., Haghani E., Malakian B., React. Funct. Polym. 2006, 66, 1033. [Google Scholar]
  • 41. Chu F., Chu X., Ahang S., Zhu H., Ren Y., Han J., R, X. , Lin B., Ding J., ChemistrySelect 2019, 4, 5269. [Google Scholar]
  • 42. Chen N., Paek S. Y., Lee J. Y., Park J. H., Lee S. Y., Lee Y. M., Energy Environ. Sci. 2021, 14, 6338. [Google Scholar]
  • 43. Cha M. S., Park J. E., Kim S., Han S. H., Shin S. H., Yang S. H., Kim T. H., Yu D. M., So S., Hong Y. T., Yoon S. J., Oh S. G., Kang S. Y., Kim O. H., Park H. S., Bae B., Sung Y. E., Cho Y. H., Lee J. Y., Energy Environ. Sci. 2020, 13, 3633. [Google Scholar]
  • 44. Hu X., Huang Y., Liu L., Ju Q., Zhou X., Qiao X., Zheng Z., Li N., J. Membr. Sci. 2021, 621, 118964. [Google Scholar]
  • 45. Park H. J., Lee S. Y., Lee T. K., Kim H. J., Lee Y. M., J. Membr. Sci. 2020, 611, 118355. [Google Scholar]
  • 46. Jung J., Park Y. S., Choi G. H., Park H. J., Ahn C. H., Hwang S. S., Lee A. S., Int. J. Energy Res. 2023, 2023, 4716537. [Google Scholar]
  • 47. Jung J., Park Y. S., Hwang D. J., Choi G. H., Choi D. H., Park H. J., Ahn C. H., Hwang S. S., Lee A. S., J. Mater. Chem. A 2023, 11, 10891. [Google Scholar]
  • 48. Song W. J., Peng K., Xu W., Liu X., Zhang H. Q., Liang X., Ye B. J., Zhang H. J., Yang Z. J., Wu L., Nat. Commun. 2023, 14, 2732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Jiang Z. T., Yi G. Q., Yao X., Ma Y. C., Su X. Y., Liu Q. L., Zhang Q. G., Chem. Eng. J. 2023, 467, 143442. [Google Scholar]
  • 50. Li H., Yu N., Gellrich F., Reumert A. K., Kraglund M. R., Dong J., Aili D., Yang J., J. Membr. Sci. 2021, 633, 119418. [Google Scholar]
  • 51. Ma Y., Hu C., Yi G., Jiang Z., Su X., Liu Q., Lee J. Y., Lee S. Y., Lee Y. M., Zhang Q., Angew. Chem., Int. Ed. Engl. 2023, 62, 202311509. [DOI] [PubMed] [Google Scholar]
  • 52. Min K., Lee Y., Choi Y., Kwon O. J., Kim T. H., J. Membr. Sci. 2022, 664, 121071. [Google Scholar]
  • 53. Lim H., Jeong I., Choi J., Shin G., Kim J., Kim T. H., Park T., Appl. Surf. Sci. 2023, 610, 155601. [Google Scholar]
  • 54. Hager L., Hegelheimer M., Stonawski J., Freiberg A. T. S., Jaramillo‐Hernández C., Abellán G., Hutzler, A. H. , Böhm T., Thiele S., Kerres J., J. Mater. Chem. A 2023, 11, 22347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. hen N., Paek S. Y., Lee J. Y., Park J. H., Lee S. Y., Lee Y. M., Energy Environ. Sci. 2021, 14, 6338. [Google Scholar]
  • 56. Li D., Park E. J., Zhu W., Shi Q., Zhou Y., Tian H., Lin Y., Serov A., Zulevi B., Baca E. D., Nat. Energy 2020, 5, 378. [Google Scholar]
  • 57. Fu S., Song J., Zhu C., Xu G. L., Amine K., Sun C., Li X., Engelhard M. H., Du D., Lin Y., Nano Energy 2018, 44, 319. [Google Scholar]
  • 58. Chen W., Mandal M., Huang G., We X., He G., Kohl P. A., ACS Appl. Energy Mater. 2019, 2, 2458. [Google Scholar]
  • 59. Ma R., Kang Y., Wang T., Jiang T., Yin H., Liu C., Wei H., Ding Y., J. Membr. Sci. 2023, 678, 121667. [Google Scholar]
  • 60. Li J., Liu Q., Tian L., Ma W., Wang F., Wang Z., Zhu H., Int. J. Hydrog. Energy 2022, 47, 32262. [Google Scholar]
  • 61. Wang X., Huang Y., He M., Shen Y., Liu S., Liu L., Li N., ACS Appl. Energy Mater. 2023, 6, 11396. [Google Scholar]
  • 62. Lin B., Xu F., Chu F., Ren Y., Ding J., Yan F., J. Mater. Chem. A 2019, 7, 13275. [Google Scholar]
  • 63. Cao D., Yang F., Sheng W., Zhou Y., Zhou X., Lu Y., Nie F., Li N., Pan L., Li Y., J. Membr. Sci. 2022, 641, 119938. [Google Scholar]

Associated Data

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

Supplementary Materials

Supporting Information

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.


Articles from Small (Weinheim an Der Bergstrasse, Germany) are provided here courtesy of Wiley

RESOURCES