1 Introduction
Energy shortages and environmental pollution caused by the excessive consumption of fossil fuels have prompted people to continuously explore clean and efficient alternative energy technologies, such as hydropower, wind power, photovoltaic power generation, and hydrogen energy. As one of the important forms of hydrogen energy utilization, fuel cells (FC) have attracted significant attention because they are not limited by the Carnot cycle and can directly convert the chemical energy in fuel into electrical energy with high efficiency.
[1-2].
Among various types of fuel cells, the Proton Exchange Membrane Fuel Cell (PEMFC) has been extensively researched and applied in fields such as transportation, portable electronic devices, and stationary power generation, owing to its advantages of high power density, low operating temperature, and rapid startup.
[3-6]However, PEMFCs also face stringent requirements regarding cost, durability, performance, reliability, efficiency, and volume.
[7]Among these, durability issues represent one of the key bottlenecks limiting the commercial promotion of PEMFCs.
[8].
Factors affecting PEMFC durability include: corrosion of the carbon support
[9], oxidation, dissolution, and agglomeration of platinum (Pt) catalysts
[10], anode poisoning caused by CO
[11], and degradation of the proton exchange membrane (PEM)
[12], among others. Among these, PEM degradation is considered the primary cause of the decline in PEMFC performance and lifespan
[13].
According to different degradation mechanisms, PEM degradation is classified into three types: mechanical degradation, thermal degradation, and chemical degradation. Theoretically, PEMs primarily perform three functions in fuel cells: (1) isolating anode and cathode reactants; (2) conducting protons; (3) preventing electron penetration to achieve electrical insulation.
[14]. During long-term operation, pinholes may appear on the membrane surface
[15-16], cracks
[17], or local thinning
[18]and other defects, leading to decreased membrane performance and eventual failure. The resulting "gas crossover" phenomenon—where reactant gases (H
2 and O
2/N
2) permeate from the anode to the cathode through the PEM or vice versa
[18-19], not only reduces reactant concentration
[20], but may also induce side reactions within the membrane, generating highly reactive free radicals and releasing significant heat, thereby exacerbating membrane degradation and chemical corrosion, forming a vicious cycle.
[21].
In the actual operation of PEMFCs, various degradation mechanisms often occur simultaneously and interact with each other
[22], making the membrane's structural failure mechanisms more complex and significantly increasing the difficulty of research. Among these, gas crossover is closely related to the degradation process
[19]: gas permeation can trigger initial mechanical or chemical degradation, while damage to the membrane structure in turn enhances gas permeation
[20]. Therefore, systematically studying the degradation behavior of PEMs requires incorporating gas crossover effects, comprehensively analyzing the synergistic effects between different degradation modes, revealing their underlying mechanisms, and providing a theoretical basis and optimization strategies for improving PEMFC durability.
In recent years, multiple review papers have systematically summarized the overall durability issues of PEMFCs from a macroscopic perspective. For example, some articles have focused on analyzing the degradation mechanisms of internal components under typical operating conditions such as idle, dynamic load, and start-stop cycles
[23-24]. Other literature has focused on health monitoring and life prediction during long-term PEMFC operation, discussing the influencing factors and prediction indicators for membrane aging
[25-26]. Additionally, some articles have focused on optimization strategies for maintaining the long-term stability of membrane electrode assemblies (MEAs)
[4,27-28]. These studies provide important references for understanding fuel cell durability issues from a system-level perspective. However, there is currently a lack of systematic summarization and analysis targeting individual components within the MEA (especially the PEM) across three dimensions: degradation causes, influence mechanisms, and mitigation strategies.
Based on a systematic review of relevant domestic and international research achievements over the past two decades, this paper focuses on summarizing PEM degradation issues in PEMFCs. Combining gas crossover effects and synergistic mechanisms, it comprehensively discusses the types of PEM degradation, action mechanisms, and potential mitigation pathways. This paper aims to provide references for performance optimization and durability enhancement in the practical application of fuel cells.
2 Membrane degradation type
The PEMs used in fuel cells are typically ionic polymers. These polymers are chemically modified to introduce ionic functional groups, most commonly sulfonic acid groups (—SO
3-). Under the operating conditions of fuel cells, the ionic polymers self-assemble to form hydrophobic and hydrophilic regions with an interconnected structure, allowing only protons (H
+) to migrate unidirectionally along the hydrophilic channels, i.e., from the anode to the cathode, achieving effective proton transport functionality
[29].
The application of PEM in fuel cells dates back to the 1950s, when General Electric developed a sulfonic acid membrane based on phenol-formaldehyde; however, due to its poor proton transport performance, it failed to achieve widespread industrial adoption.
[30]Subsequently, DuPont designed and launched the perfluorosulfonic acid (PFSA) membrane, now widely known as Nafion.
®[31]This material, featuring perfluorovinyl ether side chains terminated with sulfonate groups grafted onto a polytetrafluoroethylene (PTFE) backbone, combines excellent structural stability with high proton conductivity, marking a milestone achievement in the field of PEM.
[14,30].
To further improve the thermal stability and ion conductivity of the membrane, Dow Chemical removed the fluoroether groups from the side chains of its original long side chain (LSC) PFSA to synthesize a short side chain (SSC) PFSA ionomer. Based on this, Solvay developed the Hyflon
® Ion series of ionomers (a representative product for fuel cells is Aquivion
® membrane)
[32-33], as
shown in Figure 1. 3M Company also designed SSC PFSA membrane materials with similar structures, enriching the commercialized PFSA membrane system
[34].
In recent years, aromatic non-fluorinated polymers have gradually become a powerful alternative to PFSA membranes due to their advantages of lower cost, high designability, and superior thermochemical stability. While maintaining high proton conductivity, these polymers offer lower preparation costs; representative materials include polyether ether ketone and polybenzimidazole.
[35-36]. Collier et al.
[14]and Wang et al.
[30]conducted comprehensive review studies that systematically examined the development history and performance evolution of PEM materials, providing a theoretical foundation for the design of next-generation membrane materials.
Over the past half-century, despite significant progress in the molecular design and engineering preparation of ionomer materials, PEMs still inevitably undergo performance degradation under long-term operating conditions. This failure leads to a continuous decline in fuel cell output power and efficiency, causing systematic irreversible damage.
[24]. Depending on the fuel cell operating environment and application scenarios, PEM failure modes can generally be categorized into three main types: mechanical degradation, thermal degradation, and chemical degradation.
2.1 Mechanical degradation
2.1.1 Cause
Mechanical fatigue loads and associated stresses are key factors leading to mechanical degradation failure of PEMs
[37]. Under constrained conditions, during long-term operation, PEMs undergo repeated expansion and contraction in response to periodic changes in water content and temperature, causing local stress concentration and time-dependent deformation, such as creep, plastic flow, and microcrack propagation
[38]. When the operating pressure of the fuel cell increases, the mechanical stress borne by the PEM further intensifies, making the membrane structure prone to deformation and rupture, leading to increased gas permeability, thereby providing conditions for failure modes such as thermal degradation and chemical degradation
[39]. The evolution process of PEM mechanical degradation is divided into four stages: the membrane material preparation stage, the fuel cell assembly stage, the short-term operation stage, and the long-term operation stage
[40].
Inherent material defects in the PEM itself, independent of the external environment, can induce or accelerate mechanical degradation. These defects typically originate from non-uniformities during the MEA fabrication process, such as cracking, orientation deviations in the catalyst layer (CL), delamination between the PEM and CL, aggregation of electrolyte or catalyst clusters, and uneven membrane thickness distribution.
[41]Taking cast membranes and extruded membranes as examples, the ion cluster structures formed by these two preparation methods exhibit cylindrical and spherical morphologies, respectively. Although their overall elastic moduli are similar, significant differences exist in local stress distribution due to varying degrees of nanoscale phase separation: under the same tensile strain, the peak stress in the cylindrical structure is 20%~30% lower than that in the spherical structure, making it less prone to fracture. This structural difference results in the durability test lifetime of extruded membranes being approximately 10 times that of cast membranes.
[42].
Sulfonated aromatic polymers also exhibit deficiencies in mechanical properties. For example, sulfonated polyether ether ketone (SPEEK)
[43]is prone to main chain scission during degradation, leading to shortened polymer chain length and reduced average molecular weight. Since polymer molecular weight and tensile strength are typically positively correlated
[44], the mechanical strength of SPEEK membranes significantly decreases after degradation; meanwhile, the random distribution of degradation sites causes uneven stress transfer, weakening the overall strength of the membrane.
Issues such as uneven pressure during fuel cell assembly, dimensional errors of components, and differences in the elastic modulus of materials
[45-46]can all cause irreversible damage to the PEM. During wet-dry cycling, the elastic difference between the frame material and the PEM easily leads to stress concentration in edge regions, where the harsh environment at gas inlets and outlets exacerbates local damage
[47]. The impact of assembly clamping displacement on the membrane is dualistic
[48]: on one hand, moderate compression restricts membrane expansion and delays crack propagation; on the other hand, excessive compressive deformation promotes crack growth. Studies indicate
[40,49-50]that the joints of battery components (especially the edges of gaskets) are the primary areas for external stress concentration and membrane bending deformation. If misalignment or dimensional errors exist in the assembly, high stress concentrations will occur at the bipolar plates and gaskets (
Figure 2), increasing the risk of membrane damage.
In the application of novel non-fluorinated polymer membranes (such as SPEEK membranes), assembly stress issues are equally prominent. Mechanical stress-strain analysis shows that clamping pressure causes structural deformation of the membrane and porous medium in the rib region, forming high von Mises stress
[51]. Compared to Nafion
®membranes, under identical operating conditions, SPEEK membranes thin significantly at the ribs and exhibit a weaker tendency for wrinkling deformation in the flow channel regions. Overall, the deformation degree of SPEEK membranes is more severe than that of Nafion
®membranes, and their mechanical toughness needs improvement
[51].
PEMFCs experience frequent temperature and relative humidity (RH) changes during startup and shutdown phases. Due to differences in swelling coefficients among various components in the MEA, thermal-humidity cycling causes repeated membrane deformation and stress concentration, leading to separation between the PEM and CL, as well as the formation of surface cracks and pinholes.
[52-53]. Experiments simulating vehicle start-stop conditions with dry-wet gas cycling indicate that after approximately 1200 cycles, significant delamination occurs in the PEM, CL, and gas diffusion layer (GDL).
[54]. Research using the cohesive zone model (CZM) points out that asymmetric humidity cycling accumulates plastic strain and exacerbates interface separation; the severity is related to the duration of the shutdown phase—the shorter the shutdown time, the more significant the separation.
[55].
During long-term operation, the PEM is subjected to the combined effects of periodic mechanical loads and operating condition changes, with membrane degradation being particularly severe under alternating dry/wet and cold/hot conditions
[56-57]. Coupled simulations using the Finite Element Method (FEM) and Computational Fluid Dynamics (CFD) show that as RH increases, membrane swelling deformation increases; however, at extremely low humidity, the membrane is in a compressed state where the compression effect dominates
[45]. The impact of temperature depends on the degree of membrane hydration: under high-temperature and high-humidity conditions (≥100 °C, 80% RH), hydrogen bonds easily form between water molecules and sulfonic acid groups, and the water plasticization effect causes the membrane to exhibit characteristics that are "harder but more brittle." If stress exceeds the yield value, wet membranes are prone to plastic fracture
[58]. Furthermore, under cyclic pressure loading, the fatigue life of PFSA membranes decreases exponentially as the maximum load pressure increases; increasing the pressure or enlarging the cyclic pressure ratio raises the stress level on the membrane and promotes the accumulation of plastic strain, ultimately accelerating membrane fatigue failure
[59]. During long-term operation, corrosion issues with bipolar plate materials (such as Cr) cannot be ignored. Metal cations (Cr
3+) generated by corrosion enter the membrane and induce the formation of cross-linked structures, restricting polymer chain segment motion, increasing membrane stiffness, and reducing water uptake, ultimately leading to a decline in proton transport performance
[60].
Overall, PEMs undergo mechanical degradation throughout their entire lifecycle (from fabrication and assembly to long-term operation), which is essentially a synergistic effect of multiple factors such as temperature, humidity, and pressure. Examining a single factor in isolation fails to accurately reflect the actual operating state of the membrane and may even lead to contradictory conclusions under different experimental conditions. In engineering applications, a balance must be struck between the mechanical durability and electrochemical performance of the membrane: while high hydration levels facilitate proton transport, they also induce swelling and mechanical deformation.
[45].
2.1.2 Mitigation Approach
The continuous pursuit of high performance in PEMFCs has guided the design of PEMs toward thinner and more efficient directions. This trend imposes stricter requirements on the mechanical durability of the membranes. Therefore, how to enhance the structural strength of the membranes while maintaining high proton conductivity has become a key focus of research.
A common strategy is to introduce inorganic reinforcing components (such as SiO
2, ZrO
2, TiO
2, graphene oxide, and nanoclay, etc.) into the PEM to improve its mechanical properties. These fillers enhance the membrane's elastic modulus and creep resistance to a certain extent. However, the interaction mechanism between the inorganic phase and the polymer matrix at the interface remains unclear
[40,61-62], and the reinforcement mechanism has not yet been clarified. Meanwhile, inorganic particles often distribute unevenly within the membrane and tend to agglomerate, blocking proton transport channels and thereby reducing the overall proton conductivity
[63].
Another representative reinforcement approach involves introducing a reinforced polymer skeleton into the PFSA membrane. By compounding a hardened porous PTFE layer within the membrane, in-plane expansion caused by humidity cycling is effectively suppressed, significantly enhancing resistance to mechanical degradation.
[64]. For instance, compared to standard single-layer expanded polytetrafluoroethylene (ePTFE) skeletons, a double-layer ePTFE structure provides stronger dimensional constraints (
Figure 3), more effectively limiting excessive expansion of the PFSA membrane under wet and dry conditions, exhibiting lower swelling rates and higher dimensional stability.
[61]. However, this improvement in mechanical stability comes at the cost of reduced proton transport performance: the dense reinforcement layer makes the proton transport path more tortuous, increasing transport resistance and leading to decreased conductivity.
At the system level, optimizing the assembly process of fuel cells is also an effective means to extend the mechanical life of PEMs. Uniformly distributing clamping pressure during assembly can avoid stress concentration within the membrane electrode assembly (MEA), thereby reducing crack propagation induced by external loads.
[46]When the inhibitory effect of clamping displacement is stronger than the swelling effect of the membrane, the fatigue crack propagation rate of the membrane significantly slows down. In addition, adopting designs such as inclined flow channels, triangular, or semi-elliptical flow channels can further disperse local stress and improve the fatigue resistance of the membrane.
[48]Such optimization strategies impose higher requirements on the manufacturing precision and assembly consistency of fuel cells, and the actual effectiveness for extending PEM life remains to be verified.
In recent years, non-fluorinated aromatic polymers have attracted widespread attention due to their low cost, but their mechanical durability is relatively poor
[51], making structural optimization design a research focus. The ePTFE-reinforced sulfonated aromatic polymer composite membrane Pemion
®exhibits excellent flexibility and structural integrity under hot and humid conditions, with Young's modulus, ultimate tensile stress, yield stress, and elastic modulus all significantly higher than those of conventional r-PFSA membranes, demonstrating superior mechanical properties
[58]. However, related research is still in its infancy, and the mechanical modification mechanisms of sulfonated aromatic PEMs and their intrinsic relationship with conductivity lack in-depth exploration.
Research on nanofiber composite membranes has also made positive progress in recent years. Wang et al.
[30]systematically reviewed the latest achievements in composite nanofiber-reinforced PEMs, pointing out that the introduction of nanofibers can simultaneously improve the proton conductivity, mechanical strength, and fuel tolerance of the membranes, making them potential materials for future high-performance PEMFCs
[65]. However, issues such as complex preparation processes and unclear mechanisms of proton transport at the interface have to some extent limited their large-scale application.
In summary, strategies to enhance the mechanical durability of PEMs mainly include: introducing inorganic fillers, composite reinforced skeletons, and optimizing assembly structures. These methods have improved the dimensional stability and mechanical strength of the membrane under cyclic loading to varying degrees, but are often accompanied by side effects such as reduced proton conductivity and increased in-plane resistance. To achieve a balance between performance and durability, commercial PEMs compromise between moderate mechanical strength and ion exchange capacity (IEC). Although this trade-off design accounts for both conductivity and structural stability, it inevitably carries the risk of reactant gas permeation, thereby inducing and accelerating thermal and chemical degradation issues. Future research should systematically reveal the intrinsic connection between mechanical degradation and the durability and electrochemical performance of membrane materials through multi-scale simulations and in-situ observation techniques, aiming to achieve the synergistic optimization goals of high strength, high conductivity, and low permeability for PEMs.
[66].
2.2 Thermal degradation
2.2.1 Causes
Increasing the operating temperature of PEMFCs can reduce the system cooling burden, which is beneficial for transportation applications limited by radiator volume and helps improve the energy utilization efficiency of combined heat and power systems. Furthermore, elevated temperatures weaken the adsorption of pollutants on the surfaces of catalysts and other fuel cell components, thereby improving electrode reaction kinetics.
[67]. However, increased temperatures lead to a decrease in membrane water content, making it more prone to local thinning, cracks, and pore defects, which accelerates the mechanical degradation process.
[19].
Increasing the temperature to a certain extent helps enhance proton transport and improve electrochemical reaction rates
[6], but when the temperature is too high, the water content in the membrane drops sharply, and proton conductivity decreases significantly
[68]. Generally, PEMFCs operate at 60~80 ℃ to maintain good hydration status. The glass transition temperature of typical PFSA membranes is about 80 ℃; at higher temperatures, the membrane structure is prone to local breakdown, causing micro-short circuits between the anode and cathode regions, H
2 and O
2 directly penetrate to the opposite side and spontaneously ignite on the catalyst surface, generating local hot spots that trigger thermally driven defects such as membrane melting and pinhole formation
[19].
For aromatic polymer membranes such as SPEEK, increasing the operating temperature from 30 ℃ to 80 ℃ can enhance the uniformity of water distribution and improve short-term battery performance; however, the mobility of polymer chains increases with rising temperature, leading to severe membrane swelling and thinning, with a significant decrease in deformation resistance.
[51].
Even at suitable temperatures, local hot spots can induce thermal degradation. When reactant gases are unevenly distributed, H
2, due to its small molecular volume, easily permeates across the membrane to the cathode
[34], contacts O
2 and undergoes a combustion reaction, generating local high-temperature regions and forming macroscopic pinholes
[6]. Furthermore, in H
2-air fuel cells, N
2 can permeate to the anode and accumulate in the flow channels, triggering local fuel starvation phenomena
[69], leading to corrosion of the electrode carbon support and degradation of cell performance
[70-71].
In addition to degradation risks caused by high temperatures, low-temperature operation and freeze-thaw cycles are also significant sources of thermal stress. In environments below 0 °C, water freezing within the membrane leads to volume expansion, followed by contraction upon melting during warming; these continuous volume changes induce severe mechanical damage.
[14]. Prolonged freeze-thaw cycling causes structural damage such as pinholes, cracks, and delamination on the membrane surface, significantly reducing fuel cell lifespan.
[72]. Taking Nafion
®membranes as an example, after multiple freeze-thaw cycles, although no catastrophic chemical degradation occurs, the membrane's elongation at break, ultimate tensile strength, and density all decrease. Meanwhile, the anisotropy of tensile strength, water uptake, and oxygen permeability of the membrane are reduced, indicating that irreversible rearrangement of the ionomer has occurred at the molecular scale.
[73].
2.2.2 Mitigation Approach
To improve the thermal stability and durability of PEMs, researchers have proposed various improvement strategies.
Introducing functionalized inorganic components into the membrane can enhance its water retention capacity under high-temperature and low-RH conditions, while achieving higher thermal stability and fuel barrier properties.
[30]. For instance, a composite membrane of two-dimensional graphene oxide and one-dimensional cellulose carbon nanofibers exhibits a proton conductivity approximately 5 times that of commercial Nafion
® 117 under conditions of 100 °C and 40% RH, demonstrating excellent high-temperature stability.
[74]. In Nafion
® membranes, introduced sulfonated carbon nanotubes enhance water retention capacity through surface —SO
3H groups, enabling the composite membrane to maintain continuous proton conduction channels under high-temperature and low-humidity conditions, sustaining high conductivity and a stable microstructure at 135 °C, as
Figure 4shows
[75]. However, such composite strategies face challenges such as uneven dispersion and agglomeration of inorganic particles, requiring further optimization of material compatibility.
Improving the uniformity of material distribution in the MEA can effectively reduce the temperature gradient during operation, avoiding thermal degradation caused by local hot spots.
[19]. This strategy requires comprehensive optimization and design from a systems engineering perspective, including homogenizing catalyst layer thickness, optimizing gas flow channel layout, and improving cooling structures.
The presence of product water has a direct impact on the temperature distribution. Introducing hydrophobic polymers (such as PTFE) into the GDL can enhance the heat transfer capability of water, accelerating heat conduction and dissipation. PTFE helps improve the thermal conductivity and temperature uniformity of the material: when the PTFE content is too high, the thermal conductivity decreases slightly with increasing pressure; at low PTFE content, the thermal conductivity increases with rising pressure.
[76].
In addition, gas purging and antifreeze flushing are effective measures to prevent residual water from freezing during freeze-thaw cycles, avoiding blockage of transmission channels and reducing membrane damage.
[77].
Overall, at normal operating temperatures (below 100 °C), the risk of thermal degradation in PEMFCs is relatively low unless local pinholes or gas permeation defects are present. Nevertheless, thermal stress and water management issues cannot be ignored during long-term membrane service. Compared to chemical or mechanical degradation, research on thermal degradation remains insufficient. The mechanisms under multi-physics coupled environments, particularly under conditions of wet-dry cycling, high temperature and pressure, and local hot spots, warrant further investigation. Future research should focus on the design of high-temperature-resistant membrane materials and the construction of efficient heat dissipation systems to extend the operational lifespan of PEMFCs and improve overall energy efficiency.
2.3 Chemical degradation
The severity of chemical degradation is typically higher than that of mechanical and thermal degradation, making it the most critical factor leading to proton exchange membrane (PEM) failure
[8,17]. Mechanical degradation stems from deformation fatigue caused by stress and fluctuations in temperature and humidity during fuel cell operation
[61], while thermal degradation often occurs after a significant increase in membrane gas permeability and the accumulation of local hot spots
[19]. In contrast, chemical degradation persists throughout the entire lifecycle of the PEM; its core mechanism involves the continuous attack by reactive radicals on specific sites within the ionomer structure, damaging both the polymer backbone and side chains, which leads to a dual deterioration of the membrane's electrochemical and mechanical properties
[6,8].
2.3.1 Cause
H
2O
2It is well known that H
2O
2, as a precursor to oxidative radicals, plays a key role in the chemical degradation of PEMs
[12,17,78]. During fuel cell operation, H
2O
2is generated via two pathways: one is the direct chemical reaction between crossover gases, and the other is the electrochemical reaction on the electrode surfaces
[79]. Studies have shown that both types of reactions occur at the anode and cathode
[80]. In the chemical reaction pathway, permeated H
2 and O
2 can directly contact on one side of the membrane to generate H
2O
2:
${\mathrm{H}}_{2}+{\mathrm{O}}_{2}\to {\mathrm{H}}_{2}{\mathrm{O}}_{2}$
The permeability of the membrane is closely related to its thickness; the thinner the membrane, the higher the gas permeation rate, and the shorter the time required for a significant increase in permeability to occur.
[81]. Some researchers have pointed out that the hydrogen permeation rate of Nafion
®membranes increases linearly with the reciprocal of the logarithm of the membrane thickness.
[82], and it is positively correlated with operating temperature, pressure, and RH.
[83-85].
The higher cathode potential in PEMFCs inhibits the formation of H
2O
2to some extent. Therefore, the anode is considered the primary region for H
2O
2generation
[86]. Numerical simulation results show that as the operating voltage increases, the transmembrane permeation rate of O
2accelerates, promoting the chemical generation of H
2O
2on the anode side; during this process, the generation rate of anodic H
2O
2is jointly controlled by the voltage driving force and O
2concentration
[87]. Thinning the membrane also significantly promotes the chemical and electrochemical generation of H
2O
2, wherein the yield of the chemical reaction pathway is approximately five orders of magnitude higher than that of the electrochemical pathway
[87].
On the other hand, some studies indicate that H
2O
2 mainly originates from the incomplete process of the cathode oxygen reduction reaction (ORR)
[88-90]. In the PEMFC cathode, O
2 reacts with protons to generate water, and the reaction may proceed via two pathways
[91]:
4-electron reduction pathway:
${\mathrm{O}}_{2}+4{\mathrm{H}}^{+}+4{e}^{-}\to 2{\mathrm{H}}_{2}\mathrm{O}$
2-electron reduction pathway
${\mathrm{O}}_{2}+2{\mathrm{H}}^{+}+2{e}^{-}\to {\mathrm{H}}_{2}{\mathrm{O}}_{2}$
${\mathrm{H}}_{2}{\mathrm{O}}_{2}+2{\mathrm{H}}^{+}+2{e}^{-}\to 2{\mathrm{H}}_{2}\mathrm{O}$
The 2-electron reduction pathway is considered the main source of cathodic H
2O
2 generation. However, theoretical calculations indicate that this pathway generates H only when the local potential is below 0.695 V.
2O
2[92]Under normal fuel cell operating conditions, such low potentials rarely occur on the cathode catalyst surface, indicating that this reaction pathway is difficult to dominate in practice, which contradicts the view that "the cathode is the main generation region".
To explain this discrepancy, a new research hypothesis suggests that the dissolution and migration of Pt catalysts may lead to H
2O
2generation regions shifting toward the interior of the membrane. During long-term operation, Pt ions dissolve from the catalyst layer and re-form nanoparticles within the PEM, becoming local reaction centers for generating H
2O
2[93]. The presence of Pt catalysts significantly promotes the formation of H
2O
2, even in the absence of an external electric field, where H
2/O
2mixed gases can generate H
2O
2[14,94]. Therefore, the migration and deposition of platinum may be a significant inducer of chemical degradation within the membrane.
Currently, the verification of the exact generation region of H
2O
2 remains insufficient, and there are discrepancies among conclusions from different studies. Some scholars believe that the reaction sites are located at the anode/membrane interface
[86-87], while others argue they are within the cathode catalyst layer or the Pt deposition region inside the membrane
[88-90]. Therefore, clarifying the H
2O
2 generation location within the MEA holds significant scientific and engineering importance for deeply revealing the formation mechanisms and attack pathways of free radicals, as well as understanding the chemical degradation process of PEMs.
Free radicals form separately existing H
2O
2oxidant, with oxidation strength insufficient to directly damage the polymer chemical structure
[95]. In the presence of metal cations, H
2O
2 is catalytically decomposed, generating highly oxidative free radicals that induce chemical degradation of the PEM. These multivalent metal ions (such as Fe
2+) are typically corrosion products of metal bipolar plates or end plates in fuel cells during long-term operation
[89]. The dissolution of metal ions is also a significant cause of stack durability decline
[23].
During fuel cell operation, metal ions M
z+ dispersed in the membrane or electrodes can react with H
2O
2 as follows, generating hydroxyl radicals (HO•) and hydroperoxyl radicals (HOO•)
[96]:
${\mathrm{H}}_{2}{\mathrm{O}}_{2}+{\mathrm{M}}^{z+}\to {\mathrm{M}}^{\left(z+1\right)+}+\mathrm{O}{\mathrm{H}}^{-}+\mathrm{H}\mathrm{O}\bullet $
${\mathrm{H}}_{2}{\mathrm{O}}_{2}+{\mathrm{M}}^{\left(z+1\right)+}\to {\mathrm{M}}^{z+}+{\mathrm{H}}^{+}+\mathrm{H}\mathrm{O}\mathrm{O}\bullet $
HOO• can also be further generated from HO• reacting with H2O2to form:
${\mathrm{H}}_{2}{\mathrm{O}}_{2}+\mathrm{H}\mathrm{O}\bullet \to {\mathrm{H}}_{2}\mathrm{O}+\mathrm{H}\mathrm{O}\mathrm{O}\bullet $
Some studies
[19,23,97]propose that a third radical—hydrogen radical (H•)—is also generated in the reaction system. When H
2is transported from the anode to the cathode, the following reactions may occur:
${\mathrm{H}}_{2}+\mathrm{H}\mathrm{O}\bullet \to {\mathrm{H}}_{2}\mathrm{O}+\mathrm{H}\bullet $
${\mathrm{O}}_{2}+\mathrm{H}\bullet \to \mathrm{H}\mathrm{O}\mathrm{O}\bullet $
However, H
2O
2 is not the sole source of reactive free radicals
[98-99]. H
2 and O
2 can react directly on the Pt electrode surface to generate free radicals without passing through H
2O
2 intermediates
[98]. Ion chromatography (IC) tests indicate that the fluoride emission rate (FER) in an H
2O
2 environment is significantly lower than in an H
2/O
2 environment, suggesting that oxidative free radicals causing membrane chemical degradation are primarily formed via a non-H
2O
2 decomposition mechanism; that is, H
2 and O
2 may undergo complex multi-step reactions on the Pt surface to directly generate reactive free radicals
[95]. Although not yet fully experimentally verified, this mechanism highlights the critical catalytic role of Pt in membrane degradation.
Overall, H
2O
2 catalytic decomposition is generally regarded as the main reaction pathway for generating highly oxidative radicals; the coexistence of Pt, H
2, and O
2 is a necessary condition for radical generation
[17].
During the chemical degradation of PEMs, three main types of radicals are involved: HO•, HOO•, and H•. Although all three can attack polymer chains, the reactivity of HO• is far higher than that of the other two, making it the primary oxidizing species responsible for membrane chemical degradation.
[100-101]. Gubler et al.
[102]pointed out that compared to HOO• or H•, the reactions between HO• and ionomers as well as their degradation intermediates are more destructive. Moreover, due to the relatively low bond energy of the O—H bond in H
2O
2, the extraction of hydrogen atoms from PFSA by HOO• is thermodynamically unfavorable.
[86].
Furthermore, according to the research by Wong and Kjeang
[103], transition metal ions represented by Fe
2+ are oxidized to higher valence states under the high potential of the fuel cell cathode, making them difficult to exist stably; therefore, the anode region is considered the primary area for the generation of radicals such as HO• and chemical degradation of the membrane
[12]. H
2O
2 and the formation pathways of various radicals are as shown in
Figure 5.
Radical attack sites - The chemical degradation of PFSAPEM stems from high concentrations of reactive radicals in the environment, which tend to attack the least stable sites in the ionomer structure, such as the α-carbon of aromatic rings, ether bonds, and sulfonic acid groups.
Currently, the most widely used LSC PFSA membranes have four main radical attack sites: carboxylic acid end groups, C—S bonds, ether bonds, and tertiary carbon atom sites at the main chain/side chain branching points (
Figure 6)
[104].
Curtin et al.
[31]suggest that HOO• and HO• are aggressive towards hydrogen-containing end groups remaining in the polymer, with the attack on carboxylic acid end groups regarded as the starting point of chemical degradation
[2]. Carboxylic acid end groups typically originate from polymerization reactions initiated by persulfates during the polymerization process: sulfate radical ions (SO
4-•) react with tetrafluoroethylene monomers to generate perfluoroalkyl sulfate structures bearing —OSO
3-end groups, which subsequently hydrolyze to form carboxylic acid end groups
[105].
The mechanism of the "decompression" reaction between HO• and carboxyl end groups is as follows: the radical abstracts a hydrogen atom from the end group to generate a perfluorocarbon radical; this radical combines with HO• to form an alcohol intermediate; the intermediate rearranges to produce acidic fluoride and HF; the fluoride reacts with water to generate a new carboxyl end group and release HF.
[106-107].
Rf $-\mathrm{C}{\mathrm{F}}_{2}\mathrm{C}\mathrm{O}\mathrm{O}\mathrm{H}+\mathrm{H}\mathrm{O}\bullet \to {\mathrm{R}}_{\mathrm{f}}$- $\mathrm{C}{\mathrm{F}}_{2}\bullet +\mathrm{C}{\mathrm{O}}_{2}+{\mathrm{H}}_{2}\mathrm{O}$
Rf $-\mathrm{C}{\mathrm{F}}_{2}\bullet +\mathrm{H}\mathrm{O}\bullet \to {\mathrm{R}}_{\mathrm{f}}$-CF2OH+CO2+HF
In the continuous "unzipping" cycle, CF
2 units on the main chain continuously break, generating low molecular weight fragments and releasing HF and CO
2. When the reaction advances to the junction between the main chain and the side chain, the entire side chain is stripped off
[8,108].
19F NMR and mass spectrometry (MS) confirm that signals for both carboxylic acid and sulfonic acid functional groups appear simultaneously in the side chain cleavage products
[109-110]. The final products of "unzipping" include perfluoroacetic acid, CO
2, HF, and sulfate ions
[111].
It is worth noting that carboxylic acid end groups not only originate from the polymerization process but can also be formed through radical attack on non-perfluorinated end groups (such as —CHF
2)
[107,112]. Therefore, before these unstable end groups are completely eliminated, "unzipping" is the most critical pathway leading to the chemical degradation of PFSA membranes, resulting in polymer backbone scission, HF release, and significant membrane thinning
[111,113].
Another major site of radical attack is the C—S bond on the ionomer side chain and the terminal sulfonic acid group
[2]:
Rf- $\mathrm{C}{\mathrm{F}}_{2}\mathrm{S}{\mathrm{O}}_{3}\mathrm{H}+\mathrm{H}\mathrm{O}\bullet \to {\mathrm{R}}_{\mathrm{f}}$- $\mathrm{C}{\mathrm{F}}_{2}\bullet +\mathrm{S}{\mathrm{O}}_{3}+{\mathrm{H}}_{2}\mathrm{O}$
Rf- $\mathrm{C}{\mathrm{F}}_{2}\bullet +\mathrm{H}\mathrm{O}\bullet \to {\mathrm{R}}_{\mathrm{f}}$-CF2OH→Rf-COF+HF
DFT calculations indicate that HO• attack leading to C—S bond cleavage is the most favorable reaction pathway, with an activation energy lower than that for attack on carboxylic acid end groups. On the other hand, in the Fenton reaction system of hydrated Nafion membranes, no HO• generation was observed, suggesting that H
2O
2-Fe complexes may also directly induce C—S bond cleavage
[114-115]. However, the comparability of Fenton experiments remains controversial, and experimental results may not accurately reflect degradation behavior under actual fuel cell operating conditions
[101].
HO• attack on terminal sulfonic acid groups has two hypotheses. The early model
[86]proposed that HO• abstracts a hydrogen atom from protonated sulfonic acid to generate a sulfonyl radical, releasing SO
3and forming a carbon-centered radical. However, DFT calculations show that this reaction has a high barrier (Δ
E = +2.19 eV), making it difficult to occur at 80 °C
[116]. A more likely pathway is HO• attacking the S atom in the deprotonated sulfonate group (—SO
3-), a pathway that requires only one HO• to initiate
[116]:
Rf- $\mathrm{C}{\mathrm{F}}_{2}\mathrm{S}{{\mathrm{O}}_{3}}^{-}+\mathrm{H}\mathrm{O}\bullet \to {\mathrm{R}}_{\mathrm{f}}$- $\mathrm{C}{\mathrm{F}}_{2}\bullet +\mathrm{H}\mathrm{S}{{\mathrm{O}}_{4}}^{-}$
The mechanism of H• attack on the sulfonic acid group is similar to that of HO•:
Rf- $\mathrm{C}{\mathrm{F}}_{2}\mathrm{S}{{\mathrm{O}}_{3}}^{-}+\mathrm{H}\bullet \to {\mathrm{R}}_{\mathrm{f}}$- $\mathrm{C}{\mathrm{F}}_{2}\bullet +\mathrm{H}\mathrm{S}{{\mathrm{O}}_{3}}^{-}$
H detected in the mass spectrum of the outlet gas
2SO
3 and SO
2 provide experimental support for this mechanism
[116].
The loss of sulfonic acid groups significantly affects ionomer performance. —SO
3-groups are highly hydrophilic, easily forming connected hydrophilic channels that support proton migration within the membrane
[117-118]. When these groups are converted to carboxylic acids, the energy barrier required for proton conduction increases significantly, and proton conductivity decreases
[119]. Therefore, during long-term MEA operation, the loss of —SO
3-groups is considered a major factor leading to significant performance degradation
[120].
The ether bond on the side chain is another highly sensitive reaction site. HO• can attack the carbon atoms at both ends of the ether bond, generating alcohols and hydroxysulfonic acids
[121].
19F NMR results indicate that the C—O—C bond at the end of the side chain is attacked first, and HO• prefers to attack the ether bond near the side chain terminus rather than at the main chain branch point
[122-123]. H
2O
2-mediated degradation can proceed in parallel with the HO•-induced process: within the hydrated cluster structure of sulfonic acid groups, H
2O
2directly cleaves the ether bond via hydrogen bonding interactions, forming alcohols and carboxylic acids without requiring radical intermediates
[114]. These results suggest that the PEM degradation mechanism may simultaneously involve both radical-induced and molecule-induced pathways.
The branch points on the main chain and the tertiary carbon sites in the side chains are also potential weak points; the C—F bond is susceptible to attack by H•, generating HF and carbon radicals
[116,124]. This process triggers the cleavage of adjacent ether bonds, ultimately forming carboxylic acid end groups
[86]. First-principles calculations indicate that the energy barrier for H• attacking the C—F bond on the main chain is low
[125].
Progress in integrated theoretical and experimental research: The "unzipping" reaction targeting carboxylic acid end groups is the starting point and main pathway for polymer degradation; the C—S bond is the most sensitive site facing HO• attack; H2O2 molecules can directly induce cleavage of C—S bonds and ether bonds; tertiary carbon sites are more susceptible to H• attack. Therefore, during the chemical degradation of PEMs, backbone end-group "unzipping" and radical attacks on side-chain structures jointly dominate the microscopic mechanisms of membrane aging and deterioration.
Free radical attack sites - sulfonated aromatic polymers. In view of the high cost and potential environmental hazards of PFSA membranes, sulfonated aromatic polymer PEMs have emerged. Such membranes mainly include two categories: aromatic backbone type and polystyrene backbone type, with basic chemical structures as shown in
Figure 7.
H
2/O
2Transmembrane permeation is the main driving force for radical generation in fuel cells; aromatic polymer membranes have lower gas permeability and offer advantages compared to PFSA membranes.
[101]. Aromatic units can easily introduce sulfonic acid groups via electrophilic sulfonation during synthesis, achieving diversity in molecular design and structural tunability.
[126]. Typical structures contain functional groups such as ether, ketone, sulfone, nitrile, and sulfide.
[28].
However, aromatic polymers exhibit significantly lower tolerance to oxidative radicals dominated by HO• compared to PFSA ionomers
[35], where the primary attack mode of radicals involves HO• addition to the ortho positions of alkyl or alkoxy substituents on the aromatic ring and to the meta position of sulfonic acid groups
[127]; this reaction selectivity stems from the ortho-activating effect of substituents and the meta-directing effect of sulfonic acid groups in electrophilic addition reactions. If benzylic α-H atoms are present, acid-catalyzed dehydration of hydroxycyclohexadienyl adducts induces main chain scission under endothermic conditions, leading to the loss of sulfonic acid groups. Polymers containing phenoxy-phenyl ether linkages (—Ar—O—Ar—), such as polysulfone (PSU) and polyetherketone (PEK), are susceptible to HO•
ipso attack, resulting in C—O—C bond cleavage
[127].
In polystyrene sulfonic acid (PSSA) ionomers, HO• can react with aromatic rings via two pathways (
Figure 8): addition to the aromatic ring to form a hydroxyl adduct, which is reversibly converted into a peroxyl radical upon oxidation; after elimination of HOO•, a stable hydroxylated product is generated; or direct abstraction of α-H to form a benzyl radical. Under acid catalysis, further dehydration of the hydroxyl adduct can also generate an unstable benzyl radical, inducing main chain scission.
[102]. In Process I, the final generation of phenolic compounds exhibits certain antioxidant effects, without producing new radical cations or benzyl radicals. However, the byproduct HOO• can oxidize the membrane structure or participate in HO• regeneration reactions; systematic research on this aspect is currently lacking.
The main chain of sulfonated polyphenyls and their derivatives consists entirely of aryl-aryl bonds, lacking heteroatom bonds susceptible to radical attack, thereby exhibiting optimal chemical stability.
[128-129]. However, the high rigidity of the linear main chain with para-linkages affects the structural integrity and mechanical properties of the membrane.
[130].
To overcome this defect, researchers introduced meta-para mixed structures into the main chain
[131]or introduced aryl ketone side chains
[132-133], thereby improving the mechanical properties of the materials. Another class of promising materials, sulfonated poly(phenylene benzene), features a simple synthesis route and lower cost
[130], and the rigidity of the main chain endows it with a higher glass transition temperature and excellent thermochemical stability
[134-135]. However, HO• can attack unsulfonated aromatic rings, generating phenolic hydroxyl structures that are further oxidized to form carboxylic acids. The resulting carboxylic acids condense with adjacent aromatic rings to form fluorenone structural units, which are subsequently oxidized into lactone structures, ultimately leading to polymer main chain scission and the generation of degradation products such as sulfobenzoic acid
[136].
Although this class of materials demonstrates significant advantages in production costs, systematic research on their mechanical properties, proton conductivity, and electrochemical stability remains lacking. Therefore, the practical application potential of novel aromatic polymers such as sulfonated polybenzophenone awaits verification.
The influence of the Pt band was mentioned earlier; the migration and deposition of Pt catalysts into the PEM is a potential cause of membrane chemical degradation. Over time, the deposited Pt particles aggregate and grow along the direction parallel to the PEM/CL interface, forming a continuous layer of Pt particles, known as the "Pt band".
[137]. The thickness and distribution range of the Pt band continue to expand with prolonged operation, eventually penetrating through the PEM layer and causing internal short circuits in the fuel cell.
Pt catalyst loss in the CL follows two pathways: first, Pt dissolves into Pt
2+under high potential and preferentially reduces and deposits onto large particles in the low potential region, known as electrochemical Ostwald ripening; second, H
2that permeates to the cathode directly reduces free Pt ions in the ionomer, causing Pt to deposit within the membrane near the cathode side
[138]. This process not only causes permanent loss of Pt catalyst active sites, weakening the effective electrochemical reaction area of the CL, but also alters the electric field distribution and conductivity characteristics within the membrane.
Pt bands form local electrochemical reaction interfaces within the membrane; H
2/O
2 penetrating into the PEM directly react to generate H
2O
2. The latter decomposes under Pt catalysis to produce active radicals such as HO• and HOO•, which attack active sites on the ionomer, triggering chain degradation reactions that ultimately lead to membrane thinning and reduced mechanical strength. Electron spin resonance (ESR) detected carbon radical signals within the Pt deposition region, indirectly confirming the formation of oxidizing radicals on the Pt bands. Meanwhile, the FER significantly increases with the amount of Pt deposition, indicating that the Pt band regions are active centers for chemical degradation.
[139].
The Pt band generation region depends on the internal membrane H
2/O
2relative partial pressure and permeability. Generally, Pt ions are more easily reduced and deposited in the region near the cathode side, forming a typical single-layer Pt band structure
[138]; under hydrogen-lean conditions, Pt
2+migrates from the cathode to deposit on the anode
[140]. In summary, Pt band formation is the combined result of three steps: Pt dissolution at the high potential of the cathode, diffusion and migration in the electrolyte, and reduction at the low potential of the anode.
2.3.2 Mitigation Approach
Researchers typically employ various strategies such as radical scavenging, structural modification, and composite reinforcement to suppress the chemical degradation of PEMs. Among these, the doping of radical quenchers
[141-142], the design of short-side-chain polymers
[101], and the introduction of high-stability functional groups
[143]are currently the three most widely applied and effective methods.
Doping radical scavengers into PEMs can significantly improve the oxidative stability of the membrane
[144-145]. The mechanism lies in the fact that oxidative radicals generated during operation are rapidly captured by the doped components, preventing them from attacking the polymer backbone or side chains.
Cerium-based compounds (Ce
3+/Ce
4+or CeO
2) are currently the most widely used inorganic free radical scavengers
[88,146], with an antioxidant mechanism based on reversible redox cycling.
$\mathrm{H}\mathrm{O}\bullet +\mathrm{C}{\mathrm{e}}^{3+}+{\mathrm{H}}^{+}\to \mathrm{C}{\mathrm{e}}^{4+}+{\mathrm{H}}_{2}\mathrm{O}$
$\mathrm{H}\mathrm{O}\mathrm{O}\bullet +\mathrm{C}{\mathrm{e}}^{3+}+{\mathrm{H}}^{+}\to \mathrm{C}{\mathrm{e}}^{4+}+{\mathrm{H}}_{2}{\mathrm{O}}_{2}$
${\mathrm{H}}_{2}{\mathrm{O}}_{2}+\mathrm{C}{\mathrm{e}}^{4+}\to \mathrm{C}{\mathrm{e}}^{3+}+\mathrm{H}\mathrm{O}\mathrm{O}\bullet +{\mathrm{H}}^{+}$
$\mathrm{H}\mathrm{O}\mathrm{O}\bullet +\mathrm{C}{\mathrm{e}}^{4+}\to \mathrm{C}{\mathrm{e}}^{3+}+{\mathrm{O}}_{2}+{\mathrm{H}}^{+}$
${\mathrm{H}}_{2}+2\mathrm{C}{\mathrm{e}}^{4+}\to 2\mathrm{C}{\mathrm{e}}^{3+}+2{\mathrm{H}}^{+}$
$2{\mathrm{H}}_{2}\mathrm{O}+4\mathrm{C}{\mathrm{e}}^{4+}\to 4\mathrm{C}{\mathrm{e}}^{3+}+4{\mathrm{H}}^{+}+{\mathrm{O}}_{2}$
Through Ce
3+/Ce
4+ redox cycling, PEM can continuously scavenge free radicals, H
2O
2, and permeated H
2, effectively delaying membrane chemical degradation
[145-146].
However, the introduction of CeO
2-based inorganic nanoparticles blurs the originally clear hydrophilic/hydrophobic phase interfaces in the ionomer membrane, which not only hinders the continuity of the hydrophilic phase and blocks proton transport channels, but also disrupts polymer chain orientation, leading to reduced crystallinity and mechanical strength
[147]. Therefore, organic antioxidants have gradually become an important alternative direction.
3,4-Dihydroxycinnamic acid (CA) possesses excellent HO• scavenging capability; compared to cerium-based compound-doped membranes, it can maintain the membrane's chemical stability and battery performance over the long term without sacrificing membrane structure and performance.
[148]. Recently proposed organic-inorganic synergistic antioxidant strategies combine CeO
2 with organic antioxidants to achieve dual protection. For instance, tea polyphenol (TP)-functionalized CeO
2 core-shell structures exhibit excellent antioxidant stability, ensuring the durability and high performance of PEMs; the TP shell can inhibit the dissolution, migration, and agglomeration of CeO
2 during operation, while simultaneously promoting the reduction and regeneration of Ce
4+ to Ce
3+, thereby accelerating the radical scavenging cycle.
[149]. Furthermore, membranes co-doped with the organic antioxidant alizarin (AZR) and Ce
3+ demonstrate a synergistic effect, exhibiting better chemical stability than pristine and singly doped PEMs.
[150]. Employing organic fixing agents such as alkylphosphonic acids (PFOPA, FHPA) inhibits cerium migration, maintaining membrane uniformity and proton conductivity under high cerium retention rates.
[151]. However, electrostatic attraction between the basic groups of organic antioxidants and protons may cause proton conductivity loss.
[144]. Therefore, achieving a balance between chemical durability and battery performance remains a challenge in designing efficient radical scavengers.
SSC PFSA membranes lack -OCF
2CF (CF
3) fragments (
Figure 9), compared to LSC Nafion
®membranes, exhibit higher chemical stability
[152]. SSC membranes with the same equivalent weight (EW) (e.g., Hyflon
®) show significantly lower water uptake at 100 °C than Nafion
®, demonstrating superior dimensional stability and thermomechanical strength; dry-state Hyflon
®membranes possess a higher glass transition temperature and a wider usable temperature range, showing potential for high-temperature PEMFC applications (≥100 °C)
[153-154]. The high stability of SSC membranes such as 3M and Aquivion
®is related to the absence of ether bonds and tertiary carbon structures in their side chains
[155]; although short-side-chain membranes are subject to radical attack, their degradation rate is lower
[156]. However, fully resolving the structural differences between SSC and LSC from a spectroscopic perspective remains difficult, and the microscopic stability mechanisms still require investigation
[157].
Increasing the ion exchange capacity (IEC) of PEMs can enhance proton conductivity, but it reduces polymer crystallinity, causing severe hydration swelling and even dissolution at high temperatures
[158-159]. Compared with LSC membranes, SSC membranes exhibit higher sulfonic acid group density, superior crystallinity, and thermal stability due to sulfonic acid groups being closer to the main chain, demonstrating excellent chemical stability under high-temperature and low-humidity operation
[153-154]. However, SSC Aquivion
®membranes cost approximately 40% more than Nafion, indicating that industrialization requires further optimization of synthesis routes and processing techniques to reduce costs.
Metal-organic frameworks (MOFs) are regarded as novel PEM enhancement materials due to their high proton conductivity and tunable pore structures
[160-161]. Based on Cu
2+, benzimidazole, and sulfate ligands, the MOF exhibits a dual-effect stabilization mechanism: imidazole ligands resist proton attack, while sulfate enhances resistance to anion erosion, enabling the material to maintain structural stability under strong acidic and alkaline conditions and at high temperatures of 90 °C
[162].
However, the inherent brittleness and grain boundary defects of MOFs limit their self-supported film formation
[163]. Therefore, introducing MOFs into a polymer matrix to prepare MOF/PEM composite membranes achieves synergistic performance. For example, incorporating UiO-66-NH
2into polybenzimidazole (OPBI) to form a composite membrane significantly improves chemical stability. The reinforcement mechanism stems from the shielding effect of the hydrogen bond-ion interaction network between the MOF and the polymer on the main chain, which inhibits radical diffusion
[164]. Functionalized nano-coatings/ionic liquids (ILs) synergistically embedded within the MOF framework enhance the membrane's oxidation resistance
[165-166]. The metal centers in the MOF act as radical scavenging sites, inhibiting polymer chain degradation by neutralizing radicals
[167]. Although MOF-based composite membranes exhibit excellent potential, their application is still in an early stage. Establishing structure-property correlation models to guide the design of high-stability MOF-based PEMs is an important direction for future research.
Structural optimization of the Pt-containing catalytic layer is also an important approach to mitigating chemical degradation. As the Pt loading increases, its dissolution and migration rates significantly increase, promoting radical generation and exacerbating localized chemical degradation within the membrane.
[168]. Reducing Pt loading to increase CL thickness can effectively extend the Pt migration path, causing Pt to agglomerate and coarsen before reaching the membrane, thereby alleviating Pt band formation (
Figure 10)
[168]. Furthermore, although reducing the ionomer-to-catalyst ratio (I/C) slightly decreases fuel cell performance, it can reduce radical generation and delay membrane degradation. Modeling studies indicate that moderately increasing membrane thickness or optimizing Pt band distribution (keeping it away from the cathode interface) can significantly improve durability at the cost of some performance.
[169].
2.4 Multi-degradation composite
In the actual operation of PEMFCs, there is a synergistic effect between mechanical stress and chemical reactions, which jointly accelerates the overall degradation and failure of the membrane. The failure process of the membrane material undergoes a multi-scale evolution involving molecular-scale chemical degradation, mesoscopic-scale void formation, micro-crack initiation, and macroscopic crack propagation.
[170].
Radical-induced chemical degradation forms bubbles on the membrane surface and generates pores internally, exacerbating stress concentration and causing mechanical cracks to propagate rapidly along defects. The coalescence of surface bubbles leads to crack expansion along the shortest path, while the connectivity of internal pores reduces the effective load-bearing area and increases stress concentration, thereby forming a dual acceleration mechanism
[22]. Cyclic mechanical stress causes continuous propagation of pinholes and microcracks, increasing permeation channels for reactant gases, making H
2 and O
2 more likely to contact within the membrane to generate H
2O
2, inducing radical generation and accelerating chemical degradation
[90,171-172]. This chemical-mechanical-chemical positive feedback loop forms the core mechanism of accelerated degradation (
Figure 11)
[172]. Calculation results from the PEM durability-fuel cell performance coupling model show that as membrane thickness decreases, plastic deformation, gas permeability, and FER increase significantly. Membrane thinning not only increases the chemical degradation rate but also accelerates mechanical fatigue accumulation, ultimately hastening performance decay
[172].
When chemistry and mechanical stress combine to cause membrane thinning and local cracking, the gas permeability of the membrane increases exponentially. The catalytic combustion reaction of the mixed gas releases a large amount of heat, far exceeding the cooling capacity of the stack, forming local hot spots
[6,173]. In the early stage of defect formation, capillary pressure within the micropores causes liquid water to instantly fill the pinholes, providing sealing and buffering effects in a short time
[174]. However, as pinholes and cracks continue to grow, local high temperatures (>100 ℃) lead to thermal decomposition of the polymer, ultimately forming structural defects that penetrate the PEM layer, causing fuel cell short circuits or even stack failure
[175].
Long-term combined stress tests indicate that after simultaneous application of mechanical and chemical aging, local bubbles appear on the PEM surface, but the mechanically reinforced Nafion
®membrane showed no detectable significant increase in hydrogen crossover current; whereas the ordinary membrane exhibited no obvious cracks in appearance, yet demonstrated a slight rise in hydrogen crossover current. This suggests that the mechanically reinforced structure can delay permeation failure caused by chemical degradation
[176-177]. The main impact of chemical degradation is concentrated on the unreinforced layer of the membrane surface, while mechanical reinforcement effectively inhibits crack propagation into the interior, significantly delaying fatigue crack growth
[22]. Furthermore, even when extreme mechanical strain is applied during chemical stress testing, changes in indicators such as FER remain dominated by chemical reactions. Membrane samples with different levels of mechanical reinforcement showed no differences in their response to chemical stress, further corroborating that chemical degradation is the intrinsic dominant factor in PEM failure
[176-177].
Overall, the multiple degradation mechanisms of PEMs result from complex synergistic effects among chemical, mechanical, and thermal stresses. Although improvement strategies have been proposed for individual degradation mechanisms, systematic research based on synergistic mechanisms remains limited. Currently, there is a lack of holistic lifetime prediction and mitigation models that adopt a multi-physics perspective and are oriented toward actual operating conditions.
3 Conclusion and Outlook
The degradation of PEMs constrains the large-scale application of PEMFCs as efficient and clean energy conversion technologies. This paper systematically reviews the causes of three major degradation mechanisms of PEMs (mechanical, thermal, and chemical degradation) and corresponding mitigation strategies, and discusses the synergistic accelerating effects of different degradation mechanisms on membrane failure.
In actual operation, PEM failure is not caused by a single mechanism alone. Cracks and pinholes induced by mechanical stress during processes such as assembly, freeze-thaw cycles, and temperature/humidity cycling exacerbate gas permeation, promoting H2O2 generation and chemical degradation via radical attack. Conversely, chemical degradation weakens the membrane's structural integrity and mechanical strength, further accelerating defect propagation. Although thermal degradation has a relatively minor impact under normal operating conditions, it can synergistically accelerate mechanical and chemical degradation failures under local hot spots or extreme temperature cycling.
Methods such as incorporating mechanically reinforcing materials, optimizing assembly processes, and improving membrane structures can enhance the mechanical durability of PEMs, but at the cost of proton conductivity. Improving the thermal stability of PEMs, optimizing thermal management designs, and enhancing water management are the primary directions for addressing thermal degradation. Oxidative radical attack is the main cause of chemical degradation; strategies to mitigate this include eliminating unstable end groups, developing SSC polymer membranes, and adding radical scavengers. The long-term stability and migration loss of scavengers, their impact on conductivity and mechanical properties, as well as the cost of novel materials, constitute the major development bottlenecks. Additionally, the formation of Pt bands exacerbates chemical degradation, necessitating simultaneous optimization of CL design.
The generation mechanisms of the three degradation modes, their impact on membrane performance, and existing mitigation methods are summarized in
Table 1.
Future research can proceed from the following perspectives:
First, elucidating the synergistic effects of multiple degradation mechanisms is fundamental to optimizing PEM durability. Current research mostly focuses on single degradation mechanisms, yielding conclusions that deviate significantly from problems encountered in actual operation; the underlying reason is that different mechanisms do not simply superimpose but interact with each other. There is an urgent need to develop coupled accelerated stress test (AST) methods and multi-scale, multi-dimensional models capable of simultaneously simulating mechanical stress, thermal load, and chemical erosion, thereby revealing membrane failure pathways, dominant mechanisms, and the weighting of multi-factor interactions that are closer to actual operating conditions, providing a theoretical basis for developing universally applicable comprehensive mitigation strategies.
Secondly, establishing quantitative correlations between gas permeability and degradation, along with predictive models, can help practitioners intuitively understand membrane degradation and aging. Currently, gas permeability diagnostic technologies are relatively mature; electrochemical methods or gas chromatography/mass spectrometry can accurately measure the gas permeability of PEMs. However, research on using this as a quantitative indicator for real-time monitoring and predicting membrane degradation status remains insufficient. Future efforts should focus on establishing quantitative structure-activity relationships between changes in gas permeability and microscopic structural damage to the membrane (such as decreased polymer molecular weight, loss of sulfonic acid groups, crack propagation, F-release) and macroscopic performance decay (including reduced proton conductivity, increased membrane resistance, and diminished mechanical strength). Online diagnostic and lifetime prediction technologies for membrane health status based on gas permeability monitoring will also facilitate aging detection in fuel cells.
Based on a comprehensive understanding of degradation mechanisms, it is imperative to develop novel PEM materials with high performance, high durability, and low cost. By combining the advantages of different materials—such as ePTFE providing mechanical support, SSC PFSA or modified aromatic polymers enhancing proton transport, and MOFs/nanoparticles offering radical scavenging or water retention capabilities—we can design composite membranes with gradient structures, featuring an anti-degradation reinforcement layer at the CL interface and an internal layer optimized for enhanced proton transport. This requires developing efficient, stable novel radical scavengers with minimal impact on proton transport; creating new polymer membranes beyond Nafion, such as SSC PFSA membranes and sulfonated aromatic polymer membranes, focusing on improving their oxidative stability and mechanical durability while optimizing synthesis processes to reduce costs; and addressing the film-forming properties and grain boundary resistance of MOF materials by investigating their distribution, interfacial interactions, and long-term stability within composite membranes.
Finally, integrate the design of PEMFC optimization with an intelligent control system. Explore novel operating strategies based on synergistic degradation mechanisms, such as optimized start-stop protocols, temperature/humidity/pressure control, and water/thermal management system design, to mitigate membrane degradation at the system level. Combine membrane status monitoring to develop adaptive control strategies that extend stack life.
In summary, addressing PEM degradation requires synergistic innovation spanning mechanism research, characterization techniques, material design, and system control. Deeply understanding and effectively suppressing the PEM failure cycle of crossover-degradation-crossover, while developing novel PEMs that combine high proton conductivity, excellent mechanical/chemical stability, and cost-effectiveness, along with optimizing operating strategies, is key to advancing PEMFC technology toward commercial application. Future research should place greater emphasis on complex interactions under actual operating conditions to create more reliable, economical, and durable PEMFC systems.
National Key Research and Development Program of China(2022YFE0207200)
National Natural Science Foundation of China(52236010)
PDF
