Advanced Electrocatalyst Supports for High-Temperature Proton Exchange Membrane Fuel Cells: A Comprehensive Review of Materials, Degradation Mechanisms, and Performance Metrics

Abstract

High-temperature proton exchange membrane fuel cells (HT-PEMFCs) offer distinct advantages over their low-temperature counterparts. However, their commercial viability is significantly hampered by durability challenges stemming from electrocatalyst support degradation in the corrosive phosphoric acid environment. This review provides a comprehensive analysis of advanced strategies to overcome this critical durability issue. Two main research directions are explored. The first involves engineering more robust carbon-based materials, including graphitized carbons, carbon nanostructures (nanotubes and graphene), and heteroatom-doped carbons, which enhance stability by modifying the carbon’s intrinsic structure and surface chemistry. The second direction focuses on replacing carbon entirely with intrinsically stable non-carbonaceous materials. These include metal oxides (e.g., TiO₂, SnO₂), transition metal carbides (e.g., WC, TiC), and nitrides (e.g., Nb₄N₅). For these non-carbon materials, a key focus is on overcoming their typically low electronic conductivity through strategies such as doping and the formation of multi-component composites. The analysis benchmarks the performance and durability of these advanced supports, concluding that rationally designed composite materials, which combine the strengths of different material classes, represent the most promising path toward developing next-generation, long-lasting catalysts for HT-PEMFCs.

1. Introduction

HT-PEMFCs have emerged as a compelling alternative to conventional low-temperature (LT) PEMFCs, addressing several of the key technological barriers that have impeded the widespread adoption of fuel cell technology [1,2,3]. The defining characteristic of HT-PEMFCs is their elevated operating temperature, typically in the range of 120–200 °C, which is significantly higher than the sub-100 °C operation of traditional LT-PEMFCs [4,5,6]. This fundamental difference gives rise to a suite of operational advantages that simplify system design and broaden fuel flexibility. However, these advantages come at the cost of introducing a new set of material challenges, centered on the unique and aggressive internal environment of the cell [7,8,9,10].

The key enabling component for high-temperature operation is the electrolyte, which is typically a polybenzimidazole (PBI) membrane heavily doped with concentrated phosphoric acid (PA) [11,12,13]. PBI is a thermally and mechanically robust heterocyclic polymer that acts as a matrix, while the PA serves as the proton-conducting medium [14,15]. Unlike in LT-PEMFCs, where protons are transported via water molecules (a vehicular mechanism), proton conduction in PA-doped PBI occurs through the Grotthuss or “hopping” mechanism. In this process, protons jump between adjacent phosphoric acid molecules and their corresponding anions (H₄PO₄⁺, H₂PO₄⁻) within a dynamic hydrogen-bond network [11]. This mechanism is independent of water, enabling the cell to operate in an anhydrous state at temperatures well above the boiling point of water.

While essential for sustaining high proton conductivity at elevated temperatures, the PA-rich environment in HT-PEMFCs presents a significant challenge. The internal conditions of the cells are extremely harsh, combining several detrimental factors. First, the operating temperature typically ranges from 120 to 200 °C, which accelerates degradation processes in both the membrane and the electrocatalyst layers [16,17,18]. At the same time, the presence of concentrated phosphoric acid creates a highly acidic environment that can attack catalyst supports, corrode metal components, and compromise membrane integrity [19,20]. Adding to these stresses, the cathode is subject to high oxidative potentials, often exceeding 0.8 V versus the reversible hydrogen electrode (RHE) under normal operation, with even higher spikes during transient events such as start-up and shutdown [21]. Together, these factors impose severe chemical and electrochemical stress on the materials, making long-term durability and stability of the fuel cell a formidable challenge.

This combination of factors is exceptionally corrosive to all membrane electrode assembly (MEA) components, but it poses the most severe threat to the stability of the electrocatalyst support. Furthermore, the PA itself presents challenges. Over long-term operation, PA can be lost or redistributed from the membrane to the catalyst layers, a phenomenon known as acid leaching or migration [22,23]. This can lead to a decrease in membrane conductivity and flooding of the electrode’s porous structure [24,25]. Critically, phosphate anions from the electrolyte are known to strongly adsorb onto the surface of the Pt catalyst. This adsorption blocks active sites that would otherwise be available for the oxygen reduction reaction (ORR), thereby inhibiting reaction kinetics [26]. To compensate for this poisoning effect and achieve adequate performance, current HT-PEMFCs require significantly higher Pt loadings (typically around 1.0 mg cm⁻²) compared to their low-temperature counterparts (0.1–0.4 mg cm⁻²), which has significant cost implications [4].

The electrocatalyst support plays a far more significant role than simply serving as an inert substrate; it is an active, multifunctional foundation that critically influences both the performance and durability of the fuel cell [27,28]. To meet the stringent demands of the HT-PEMFC environment, an ideal support must simultaneously combine several key properties, as follows: (1) High Surface Area: It must possess a high specific surface area to allow for the fine and uniform dispersion of the expensive Pt catalyst nanoparticles. This maximizes the electrochemically active surface area (ECSA), ensuring efficient utilization of the precious metal [28,29]. (2) Electronic Conductivity: It must provide a highly conductive pathway for electrons to travel between the catalyst nanoparticles and the current collector (bipolar plate), minimizing ohmic losses within the electrode [30]. (3) Catalyst Anchoring: It needs to form a strong interaction with the catalyst nanoparticles to anchor them securely. This interaction can range from weak physisorption (e.g., van der Waals forces), which is typical for Pt on pristine graphitic surfaces, to strong chemisorption (e.g., covalent bonding or strong SMSI), which occurs at defect sites, heteroatom-doped sites, or on the surface of materials such as metal oxides [31]. Weak interactions are insufficient to prevent the particles from detaching, migrating, and agglomerating (sintering) during prolonged operation, which is a primary mechanism of ECSA loss and performance degradation [32]. Therefore, a key goal in advanced support design, as discussed in Section 3 and Section 4, is to engineer surfaces that promote these stronger, more durable chemical bonds with the catalyst nanoparticles. (4) Porous Structure: It must form a well-defined porous network that facilitates the efficient mass transport of reactant gases (H₂ and O₂) to the triple-phase boundary (where electrolyte, catalyst, and gas meet) and allows for the effective removal of product water vapor from the cathode [33,34]. (5) Chemical and Electrochemical Stability: Crucially, it must be exceptionally stable and resistant to corrosion in the aggressive HT-PEMFC operating environment (high temperature, high potential, and concentrated phosphoric acid) [24].

To systematically evaluate the materials discussed in this review, these requirements can be codified into a set of ideal properties, as summarized in

Table 1. Ideal properties of electrocatalyst support for HT-PEMFCs.

Property	Ideal Requirement	Rationale
Electronic Conductivity	>1 S cm−1	To minimize ohmic losses and ensure efficient electron transport to/from catalyst sites.
Specific Surface Area	>100 m2 g−1	To achieve high dispersion and utilization of Pt nanoparticles, maximizing ECSA.
Porosity	Hierarchical (micro/meso/macro)	To facilitate efficient mass transport of reactant gases and removal of product water.
Chemical Stability	Inert in hot, concentrated H3PO4	To prevent chemical dissolution and structural degradation of the support material itself.
Electrochemical Stability	High corrosion resistance at >0.8 V vs. RHE	To prevent electrochemical oxidation, especially at the cathode during operation and transients.
Catalyst–Support Interaction	Strong (e.g., covalent, SMSI)	To anchor catalyst nanoparticles, preventing detachment, migration, and agglomeration.
Cost & Scalability	Low-cost, scalable synthesis	To enable economically viable, large-scale manufacturing for commercial applications.

Data compiled from source [27].

. This framework provides a benchmark against which all existing and emerging support materials can be measured, highlighting the inherent trade-offs in material design.

Despite the compelling advantages of HT-PEMFCs, their widespread commercialization has been impeded by concerns over long-term durability. The aggressive internal environment presents a formidable challenge for all MEA components, but the stability of the catalyst layer, and specifically the catalyst support, has been identified as a primary bottleneck. The degradation of the support material, typically carbon black in conventional systems, initiates a cascade of failure mechanisms that lead to irreversible performance decay.

Therefore, the development of novel, highly durable electrocatalyst supports is one of the most critical areas of research in the HT-PEMFC field. This review aims to provide a systematic and comprehensive analysis of this topic. It is structured to first define the core problem by examining the degradation mechanisms of conventional carbon supports in detail (Section 2). It will then explore the two major strategic avenues being pursued for a solution: the engineering of more robust, advanced carbon-based materials (Section 3) and the development of inherently corrosion-immune non-carbonaceous materials (Section 4). A direct comparative analysis will benchmark the performance and durability of these advanced materials (Section 5), leading to a concluding discussion on the remaining challenges and most promising future research directions for the field (Section 6).

While several excellent reviews have covered the broad topic of advanced materials for HT-PEMFCs [35,36,37], this work distinguishes itself through three key contributions. First, it offers a singular and in-depth focus specifically on the electrocatalyst support, which we identify as a primary life-limiting bottleneck for the entire membrane electrode assembly. Unlike reviews that cover all cell components, our targeted approach allows for a more detailed analysis of the support’s degradation mechanisms and the material science solutions being developed. Second, this review provides a systematic comparative analysis and direct performance benchmarking of the two major strategic avenues: advanced, engineered carbon-based supports and inherently stable non-carbonaceous supports. This direct comparison offers a clear perspective on the relative advantages and trade-offs of each approach. Finally, we synthesize the findings to highlight an overarching trend, identifying the shift towards a “composite material paradigm” as the most promising and critical future research direction for developing truly durable, next-generation catalysts. By providing this focused analysis and strategic outlook, this review fills a specific gap and serves as a valuable resource for researchers dedicated to solving the critical durability challenge in HT-PEMFCs.

2. Degradation of Conventional Carbon Black Supports

For decades, carbon black (CB), and particularly the Vulcan XC-72 grade, has been the workhorse catalyst support in the fuel cell industry. Its widespread use stems from a combination of favorable properties that are well-suited for LT-PEMFCs: a high specific surface area (typically > 200 m² g⁻¹), which is ideal for dispersing catalyst nanoparticles; good electronic conductivity; a well-established and low-cost manufacturing process [38]; and excellent processability into catalyst inks.

However, the microstructure of carbon black, which consists of aggregated spheroidal primary particles with an amorphous core and a shell of stacked, graphene-like domains, is its primary vulnerability [39,40,41]. This structure is turbostratic and non-graphitizable, meaning it contains a high density of structural defects, edge planes, and dangling bonds [42,43]. These sites are thermodynamically less stable and serve as active sites for electrochemical oxidation, a process commonly referred to as carbon corrosion [44]. Therefore, the stability of carbon supports is intrinsically linked to their structural properties, such as the degree of graphitization and surface morphology, which are the primary focus of the advanced engineering strategies discussed in Section 3.

2.1. Mechanisms of Carbon Corrosion in the HT-PEMFC Environment

The primary degradation pathway for carbon supports in the cathode of an HT-PEMFC is electrochemical oxidation, in which solid carbon is converted into carbon dioxide according to the reaction [45]:

C + 2H₂O→CO₂ + 4H⁺ + 4e⁻

Thermodynamically, this reaction is favorable at any potential above 0.207 V vs. RHE. While the kinetics are slow at room temperature, several factors within the HT-PEMFC environment combine to dramatically accelerate this destructive process. One of the most critical factors is the high operating temperature. The elevated temperatures of 120–200 °C provide the thermal energy needed to overcome the kinetic barriers of the carbon oxidation reaction, leading to significantly higher corrosion rates compared to LT-PEMFCs [24].

To investigate the electrochemical carbon corrosion in HT-PEMFCs, Hyung-Suk Oh et al. [46] employed a potentiostat to regulate the oxygen electrode potential within the 1.0–1.4 V range, aiming to simulate abnormal operational conditions in fuel cells. Figure 1a depicts the CO₂ concentration profiles during corrosion tests conducted at varying potentials. No detectable CO₂ was observed at 1.0 V, whereas trace amounts began to emerge at 1.1 V and increased with higher potentials. Typically, upon potential application, the CO₂ concentration exhibited an initial surge followed by a gradual decline, reaching baseline levels after 30 min of testing. Evidently, the escalating potential augmented the carbon corrosion reaction’s overpotential, leading to a substantial increase in CO₂ emission. These findings unequivocally demonstrate that HT-PEMFCs are susceptible to electrochemical carbon corrosion even under non-humidified conditions. Direct experimental evidence from online mass spectrometry further confirms that, under comparable electrochemical scenarios, HT-PEMFC cathodes generate significantly more CO₂ than their low-temperature counterparts, providing a quantifiable metric for enhanced corrosion [46,47].

Another major contributor is the high cathode potential. During steady-state operation, the cathode potential is typically in the range of 0.6–0.8 V vs. RHE. More critically, during transient events such as cell start-up and shutdown, or in cases of localized fuel starvation, the cathode potential can spike to values exceeding 1.0 V and even approaching 1.5 V [45]. In this high-potential regime, the rate of carbon corrosion increases exponentially, leading to rapid and severe degradation of the support structure.

Compounding these effects is the catalytic role of the Pt nanoparticles themselves. Pt is not only a catalyst for the desired oxygen reduction reaction but also an effective catalyst for the undesired carbon oxidation reaction [48]. This catalytic effect lowers the potential at which significant corrosion begins. Experimental studies using differential electrochemical mass spectrometry (DEMS) have provided clear evidence for this phenomenon. As shown in Figure 1b, for a pure carbon support, significant CO₂ evolution (a direct marker of corrosion) is observed at potentials above ~0.9 V vs. RHE. However, for a Pt-loaded carbon support (Pt/C), an additional, distinct corrosion peak appears at much lower potentials, in the range of 0.4–0.5 V vs. RHE [49]. This means that even under normal, steady-state operating conditions, the Pt catalyst is actively promoting the destruction of its own support. This creates a fundamental paradox: the very component required for cell function is also a primary accelerant of its key degradation mechanism. This realization underscores why simply improving Pt activity on a conventional carbon support is an insufficient strategy for enhancing long-term durability and necessitates a paradigm shift towards more robust support materials [50].

Figure 1. (a) CO₂ mass-spectrographic profiles show the outcomes of carbon corrosion tests conducted on HT-PEMFCs. These experiments involved running for 30 min at individual potentials (1.0 V, 1.1 V, 1.2 V, 1.3 V, and 1.4 V) at 150 °C, with non-humidified O₂ fed to the cathode [46] (reproduced with permission, Copyright 2012, Elsevier). (b) This graph illustrates how the ionic mass current for mass 44 depends on potential, with separate curves for Pt/CV (red) and CV (blue) [49] (reproduced with permission, Copyright 2017, EDP Sciences).

Figure 1. (a) CO₂ mass-spectrographic profiles show the outcomes of carbon corrosion tests conducted on HT-PEMFCs. These experiments involved running for 30 min at individual potentials (1.0 V, 1.1 V, 1.2 V, 1.3 V, and 1.4 V) at 150 °C, with non-humidified O₂ fed to the cathode [46] (reproduced with permission, Copyright 2012, Elsevier). (b) This graph illustrates how the ionic mass current for mass 44 depends on potential, with separate curves for Pt/CV (red) and CV (blue) [49] (reproduced with permission, Copyright 2017, EDP Sciences).

Finally, the highly acidic environment created by concentrated phosphoric acid in the electrolyte contributes to the overall aggressiveness of the system. Although its direct quantitative impact on carbon corrosion is less well established compared to potential and temperature effects, it remains an additional factor that exacerbates the hostile conditions faced by the carbon support in HT-PEMFCs [24,27].

In summary, the degradation of conventional carbon supports is a cascading failure process initiated by the harsh operating conditions of HT-PEMFCs. The combination of high temperature, high cathode potentials, and transient potential spikes (particularly during start-up/shutdown) provides the thermodynamic and kinetic driving forces for the electrochemical oxidation of the carbon framework. This corrosion process is further and significantly accelerated by the catalytic activity of the Pt nanoparticles themselves, creating a self-destructive feedback loop. The direct consequence of this material loss is the catastrophic structural failure of the catalyst layer, which triggers the subsequent degradation phenomena detailed in Section 2.2, including the detachment and agglomeration of Pt nanoparticles and the collapse of the porous electrode structure.

2.2. Consequences of Support Degradation

The electrochemical corrosion of the carbon support is not an isolated event. Instead, it triggers a cascade of interrelated degradation phenomena that collectively cause a catastrophic and irreversible loss of fuel cell performance [51]. As carbon is oxidized to CO₂, the physical support material is literally consumed. This leads to the thinning of the catalyst layer and the collapse of its carefully engineered porous structure. Identical location electron microscopy studies have visualized the formation and growth of cracks in the electrode layer, leading to delamination from the membrane [48,52]. Figure 2a shows the changes in the cathode catalyst layer of the fuel cell from the beginning-of-life (BOL) state to the end-of-life (EOL) state during the start-up and shutdown process. It includes TEM images in different states (the initial one on the left and the aged one on the right) and the corresponding electrochemical performance curves (the middle voltage–current density curves), presenting the deterioration of the microstructure of the catalyst layer (such as cracking) and the process of performance degradation.

With the underlying carbon support corroding away, the Pt nanoparticles anchored to it become detached. These now-mobile particles can migrate through the catalyst layer and agglomerate (or sinter) into larger, less catalytically active particles via a process known as Ostwald ripening. This process drastically reduces the total ECSA, as a larger fraction of the Pt atoms become buried within the bulk of the agglomerated particle and are no longer accessible to reactants [45,53]. Figure 2b presents the structural evolution of Pt/Pt alloys on the surface of the fuel cell catalyst support. Initially, the Pt/Pt alloys are uniformly dispersed on the support surface; during use or aging, some particles agglomerate (Pt/Pt alloys agglomerated) and some detach from the support (Pt/Pt alloys detached). Such agglomeration and detachment reduce the effective active area of the catalyst and degrade its performance, being one of the key causes of fuel cell performance decay, often associated with long-term operation, potential cycling, and other working conditions. Some detached particles may also become electrically isolated from the conductive network, rendering them completely inactive.

It is important to note that these degradation pathways for Pt nanoparticles are fundamentally the same as those observed in LT-PEMFCs, namely (i) dissolution and redeposition (Ostwald ripening), (ii) particle migration and agglomeration, and (iii) detachment from the support [54]. However, the kinetics of these processes are significantly accelerated in the HT-PEMFC environment. The elevated operating temperatures (120–200 °C) provide greater thermal energy, which increases both the dissolution rate of platinum and the surface mobility of Pt atoms, leading to much faster particle growth via sintering and Ostwald ripening compared to low-temperature operation. Furthermore, the concentrated phosphoric acid electrolyte can influence the chemistry of Pt dissolution and may form complex species with dissolved platinum ions, potentially altering the redeposition dynamics. Therefore, while support corrosion is the primary trigger, the intrinsic degradation of the Pt nanoparticles themselves is also a severely exacerbated challenge in HT-PEMFCs.

The collapse of the porous electrode structure also impairs mass transport within the catalyst layer. The reduced porosity hinders the diffusion of reactant gases, particularly oxygen, to the remaining active sites. Simultaneously, structural degradation can alter the water and phosphoric acid management within the electrode, sometimes causing localized flooding. These further block gas diffusion pathways and exacerbate mass transport losses, especially under high current density conditions. Finally, the degradation of the interconnected network of conductive carbon particles can lead to an increase in the electronic resistance of the catalyst layer, contributing to higher overall cell ohmic losses [48,55,56]. Figure 2c illustrates the microstructural and particle size evolution of the fuel cell catalyst under pristine conditions, after potential cycling, and after start–stop cycling. The pristine catalyst features uniformly dispersed, fine particles. After potential cycling and start–stop cycling, the particles gradually agglomerate and grow, with more pronounced degradation observed under start–stop cycling. The corresponding particle size distributions reveal an increase in average particle size and a broadening of the size distribution with cycling, indicating structural degradation of the catalyst. Start–stop cycling exerts the most significant impact, reducing the active surface area and thus compromising the catalytic performance, which is a critical consideration in catalyst durability studies.

Figure 2. (a) The evolution of the cathode catalyst layer in a fuel cell from BOL to EOL, observed during start-up and shutdown cycles [48] (reproduced with permission, Copyright 2024, ACS). (b) A schematic detailing Pt agglomeration and detachment on the carbon support surface [57] (reproduced with permission, Copyright 2019, Elsevier). (c) TEM images that reveal the effects of potential cycling (c2) and start/stop degradation (c3) on the cathode catalyst layer, compared to its pristine condition (c1), along with Pt particle size distributions (c4–c6) [58] (reproduced with permission, Copyright 2011, ACS).

Figure 2. (a) The evolution of the cathode catalyst layer in a fuel cell from BOL to EOL, observed during start-up and shutdown cycles [48] (reproduced with permission, Copyright 2024, ACS). (b) A schematic detailing Pt agglomeration and detachment on the carbon support surface [57] (reproduced with permission, Copyright 2019, Elsevier). (c) TEM images that reveal the effects of potential cycling (c2) and start/stop degradation (c3) on the cathode catalyst layer, compared to its pristine condition (c1), along with Pt particle size distributions (c4–c6) [58] (reproduced with permission, Copyright 2011, ACS).

2.3. Experimental Evidence and Quantification of Degradation

The severe degradation of conventional Pt/C catalysts in HT-PEMFC environments has been thoroughly demonstrated through extensive experimental investigations, particularly using accelerated stress tests (ASTs). These tests are designed to mimic the effects of thousands of operational cycles within a short time frame by repeatedly cycling the electrode potential into the high-potential region where carbon corrosion is most pronounced [47,59]. Quantitative results from these experiments highlight the significant instability of standard carbon supports under such harsh conditions.

The test results by Cleemann et al. [47] reported that in an accelerated potential cycling test (with potentials cycled between 0.9 V and 1.2 V at a scan rate of 5 mV S⁻¹), the catalyst supported on untreated Vulcan XC-72R carbon black exhibited significant performance degradation after only 579 cycles, with no recovery observed in subsequent cycles. The polarization curves in Figure 3a, measured at various cycle numbers (0, 579, 1185, 1665, 2147 cycles, etc.), clearly demonstrate a progressive decline in cell voltage at a given current density as the number of cycles increases, particularly after 579 cycles, which directly reflects the performance decay. Similarly, a study by Lauf P et al. [60] adopted a distinct AST protocol (involving 5000 potential cycles) and found that the ECSA of the Pt/Vulcan catalyst of the same type decreased by approximately 40%. As depicted in Figure 3b, for the Pt/V (Pt/Vulcan) and Pt/KB catalysts, when comparing the EOL to the BOL, the ECSA of Pt/V shows a 38% reduction, and that of Pt/KB exhibits a 27% decrease. This is in line with the research findings of Lauf P et al. Additionally, in situ measurements of carbon dioxide emissions have been utilized to directly quantify the carbon corrosion rate. Rau M et al. [49] investigated the corrosion of carbon supports under the operating conditions of HT-PEMFCs and compared the carbon oxidation behaviors of two types of carbon supports, namely carbon nanotubes (CNTs) and carbon black (Vulcan). As shown in Figure 3c, which presents the mass spectrometric current (I_ms4) of Pt/CV, it was found that under HT-PEMFC conditions, the carbon corrosion current of Pt/CV was four times that of Pt/CNT. Additionally, the two supports showed different corrosion peaks at different potentials. The presence of platinum has a significant impact on the corrosion of carbon supports, and the corrosion of carbon supports is more pronounced at high potentials (exceeding 0.8 V vs. RHE). These findings clearly indicate that traditional carbon black supports cannot achieve long-term stable operation in high-temperature proton exchange membrane fuel cells. This inherent defect highlights the urgent need for the development of advanced support materials, and relevant content will be discussed in subsequent sections.

3. Engineering Carbon for Stability: Advanced Carbon-Based Supports

Faced with the inherent instability of conventional carbon black, one major research thrust has focused not on abandoning carbon entirely, but on re-engineering its structure and chemistry to withstand the harsh HT-PEMFC environment. This approach seeks to leverage carbon’s desirable properties—high conductivity and low cost—while mitigating its susceptibility to corrosion [61]. Strategies range from improving its inherent structural robustness through graphitization to enhancing its functional stability via nanostructuring and chemical doping.

3.1. Graphitized Carbons: Ordering for Stability

The primary vulnerability of amorphous carbon black lies in its disordered structure, which is rich in high-energy defect sites that serve as initiation points for electrochemical oxidation. The most direct strategy to address this is graphitization: a high-temperature heat treatment process (typically > 1800 °C [62,63,64]) that provides the thermal energy for carbon atoms to rearrange into a more ordered, thermodynamically stable, graphitic lattice [65]. This process significantly reduces the concentration of defect sites, thereby enhancing the material’s intrinsic corrosion resistance [66,67].

The benefits of this approach have been demonstrated with remarkable clarity. Graphitized carbon (GC) supports show vastly superior durability in ASTs compared to their non-graphitized counterparts. In a pivotal study by Xue et al. [68], a Pt catalyst supported on carbon black graphitized at 1800 °C (Pt/GCB-1800) was subjected to a high-potential holding AST designed to accelerate corrosion. As shown in Figure 4a, the Pt/GCB-1800 catalyst exhibited a potential decay of only 0.5% at a high current density of 1000 mA cm⁻². In stark contrast, a standard commercial Pt/C catalyst under the same test conditions suffered a massive potential decay of 34.9%. This enhanced support stability directly translated to improved catalyst stability: Pt particle growth on the graphitized support was almost completely suppressed (0.08 nm growth) compared to the significant agglomeration observed on the commercial support (0.95 nm growth). Even less extreme heat treatments have proven effective. Cleemann et al. [47] revealed that heat-treating Vulcan XC-72R at 1000 °C enhances the stability of the carbon material and the durability of the catalyst. Specifically, Figure 4b depicts the polarization curves of PBI fuel cells during accelerated potential cycling tests. The test electrodes were fabricated using platinum nanocatalysts supported on either untreated VXC-72R carbon black or graphitized VXC-72R carbon black. The fuel cells operated with unhumidified hydrogen and air as reactants at 150 °C. For the catalyst supported on untreated carbon black, significant performance degradation occurred within the first 579 cycles, while no further performance changes were observed in the subsequent cycles from 1185 to 2147.

However, this improvement in stability comes with a trade-off. As the carbon structure becomes more compact and ordered during graphitization, the specific surface area and porosity of the material typically decrease. As illustrated in Figure 4d, the graphitized carbon black, after high-temperature heat treatment, exhibits a substantial reduction in specific surface area, dropping from over 200 m² g⁻¹ to 66 m² g⁻¹. This can lead to a lower initial dispersion of platinum nanoparticles, resulting in slightly inferior initial performance compared to amorphous carbon with a high specific surface area. Liu et al. [69] employed CV to quantitatively assess the ECSA of catalyst-coated layers (CCLs) incorporated with various catalysts after specific cycles, focusing on the AST of the carbon support. After 5000 AST cycles, all catalysts underwent progressive ECSA loss, with the rate of loss negatively correlated with the degree of graphitization of the carbon support. The ECSA loss data summarized in Figure 4c further validates this trend, demonstrating that graphitization-controlled high-temperature treatment significantly enhances the catalyst’s durability under sustained-under-stress-degradation (SUSD) cycling conditions. For applications demanding long service life, the notable improvement in durability often renders this sacrifice of initial performance an acceptable compromise [70].

3.2. Carbon Nanostructures: Leveraging Inherent Graphitic Nature

An alternative approach to achieving a more robust carbon framework is to utilize carbon allotropes that are inherently more graphitic and crystalline than amorphous carbon black. CNTs and graphene have emerged as leading candidates in this category, offering a compelling combination of high electronic conductivity, excellent mechanical strength, large surface area, and superior corrosion resistance due to their highly ordered sp2-hybridized carbon structures [71,72,73].

3.2.1. CNTs

The tubular structure of CNTs features a high aspect ratio, and their graphitized surfaces exhibit superior oxidation resistance compared to carbon black [74,75,76]. MWCNTs have demonstrated exceptional electrochemical stability. Wang et al. [77] compared the electrochemical surface oxidation behaviors of carbon black Vulcan XC-72 and MWCNTs under potentiostatic treatment (for a maximum duration of 168 h) in a simulated PEMFC cathode environment (60 °C, N₂-purged). As shown in Figure 5a,b, electrochemical characterization results at different treatment time intervals revealed that MWCNTs possess better electrochemical stability than Vulcan XC-72. Under the studied conditions, MWCNTs generated fewer surface oxides, implying their potential for enhanced corrosion resistance and durability in fuel cell applications.

The chronoamperometric curves of Vulcan XC-72 and MWCNTs (Figure 5c) indicated that, under identical conditions, the corrosion current of MWCNTs was indeed 30% lower than that of Vulcan XC-72. Figure 5c also shows that the corrosion current increased when both materials were catalyzed with Pt nanoparticles; however, the catalytic effect of Pt on MWCNTs was less pronounced than on Vulcan XC-72. To quantify the effect of oxidation treatment on Pt surface area, CV measurements were conducted on Pt catalysts supported on Vulcan XC-72 and MWCNTs in N₂-purged 0.5 M H₂SO₄ solution, under the conditions of 0.9 V and 60 °C, with data collected at different time intervals during the treatment.

Figure 5d illustrates the loss of Pt surface area over time. The results showed that after 168 h of oxidation treatment, Vulcan XC-72 experienced a nearly 80% loss of Pt surface area, whereas MWCNTs only exhibited a 37% loss. Furthermore, the majority of the Pt surface area loss for MWCNTs occurred within the first 72 h, and the loss rate became extremely slow thereafter. These findings suggest that MWCNTs could offer significantly higher durability than Vulcan XC-72.

Therefore, due to their high corrosion resistance, MWCNTs, when used as fuel cell catalyst supports, result in less Pt surface area loss and higher oxygen reduction reaction activity. Their performance advantages are even more pronounced in HT-PEMFCs. Devrim et al. [78] performed single-cell ASTs on Pt/CNT and commercial Pt/C catalysts at 160 °C. As shown in Figure 5e,f, the Pt/CNT catalyst achieved a current density of 0.30 A cm⁻² at 0.6 V, which was significantly higher than that of the Pt/C catalyst (0.20 A cm⁻²). Meanwhile, the durability of Pt/CNT was also improved: after 10,000 AST cycles, Pt/CNT retained 67% of its initial performance, compared to only 60% retention for the standard Pt/C catalyst. This indicates that CNTs are more adaptable to the operating environment of HT-PEMFCs than traditional carbon supports.

3.2.2. Graphene

Graphene, a two-dimensional sheet constituted by sp²-hybridized carbon atoms, embodies the quintessential form of graphitic structure [79,80,81]. Ji et al. [82] successfully fabricated the composite catalyst layer by a hydrothermal synthesis method based on two-dimensional graphene flakes and a spherical structure of carbon black. Subsequent accelerated degradation tests corroborated the superior durability of these catalysts when compared to their carbon black-supported analogues, including commercial benchmark catalysts. Concurrently, the graphene-based supports themselves exhibited exceptional stability under AST conditions.

Figure 6 presents experimental data from the durability testing section, providing direct electrochemical performance validation for these aforementioned findings. Specifically, Figure 6a illustrates the comparative polarization and power density curves of platinum-based catalysts supported on both graphene and conventional carbon black. These data unequivocally demonstrate that graphene-supported catalysts, owing to their distinctive structural attributes and optimized active site dispersion, yield enhanced voltage output and power density during electrochemical operation. Furthermore, Figure 6b delineates the voltage-time profiles of various membrane electrode assemblies (MEAs) subjected to AST. Herein, the MEA integrating graphene-based supported catalysts is anticipated to exhibit a markedly attenuated voltage decay trajectory, thereby directly substantiating the formidable stability of graphene-based supports under accelerated stress conditions. Finally, Figure 6c–f presents a comparative analysis of the polarization and power density curves for specific MEAs in their pristine state and following 100 h of continuous operation. This comparative evaluation is expected to reveal a quantitatively lesser degree of performance degradation for the MEA featuring graphene-supported catalysts, unequivocally reinforcing its superior durability. After the AST, a greater amount of PA in MEA0 migrates to the gas diffusion layer, with only a small quantity remaining at the electrode-membrane interface. In contrast, MEAac retains a large amount of PA at the interface, while the PA content in its gas diffusion layer is relatively low. This phenomenon is likely to increase the number of triple-phase boundaries (TPBs), thereby facilitating proton transfer [30,83]. The underlying mechanism has been described and summarized in the schematic diagrams of Figure 6g,h.

In HT-PEMFCs, graphene-based supports can achieve high performance. Alpaydin GU et al. [84] reported a Pt-Ru alloy composite catalyst supported by MWCNTs and graphene nanoplatelets (GNPs), which achieved a peak power density of 266 mW cm⁻² (Figure 7a) in HT-PEMFCs operating on reformate gas at 160 °C. Beyond serving as a support, single-layer graphene has also been ingeniously applied as a functional interlayer between the membrane and the catalyst layer. Its impermeability to molecules combined with permeability to protons helps reduce PA leakage and fuel crossover, thereby enhancing long-term durability. Chen et al. [85] coated single-layer graphene (SLG) obtained by chemical vapor deposition between the membrane and the electrode via a wet-chemical transfer method to investigate its impact on the performance and durability of polybenzimidazole membranes in HT-PEMFCs. Figure 7b illustrates that excessive PA leaching leads to degraded performance and reduced durability. The diagram involves components such as the gas diffusion layer (GDL), microporous layer (MPL), catalyst layer (CL), and PBI membrane. Protons are most readily formed and transferred at the interface of the three-phase boundary (gas, catalyst, and PA). It presents various PA leaching scenarios: no leaching, proper leaching, excessive leaching, and with a SLG barrier. From this, it can be inferred that SLG can function as a barrier to modulate PA leaching. The results presented in Figure 7c–i demonstrate that the MEA loaded with SLG exhibited a higher peak power density, a lower electrode resistance, and a larger electrochemically active surface area.

Despite these encouraging results, significant obstacles remain for the commercialization of these nanostructured carbon materials. The synthesis of high-purity, defect-free carbon nanotubes and graphene is often complex and costly, and ensuring their uniform dispersion in catalyst inks to form well-structured electrodes remains a major processing challenge [44,86].

3.3. Heteroatom-Doped Carbons: Electronic and Surface Modification

A more subtle but highly effective strategy for improving carbon supports involves chemically modifying the carbon lattice by intentionally introducing heteroatoms, such as nitrogen (N) or boron (B). This process, known as heteroatom doping, does not necessarily change the bulk structure of the carbon but profoundly alters its surface chemistry and electronic properties, leading to enhanced stability and performance. The principle behind this approach is to create stronger anchoring sites for the Pt nanoparticles, thus improving the catalyst–support interaction.

Nitrogen doping introduces diverse nitrogen configurations into the carbon lattice, including pyridinic nitrogen, pyrrolic nitrogen, and graphitic nitrogen. Notably, pyridinic nitrogen sites, characterized by distinct electron-enriched properties, function as preferential nucleation and anchoring centers for platinum nanoparticles. This not only facilitates the uniform dispersion of smaller-sized catalyst particles but, more crucially, establishes robust metal–support interactions, which effectively suppress the dissolution and migration of platinum during operation. Cai et al. [87] realized the regulation of nitrogen content by simply adjusting the annealing temperature without altering the pore size distribution of the carbon support. As depicted in Figure 8a, through optimizing the annealing temperature in the nitrogen doping process with carbon black as the substrate, a gradient distribution of nitrogen content within the range of 0.68−2.01% was successfully achieved while maintaining the pore size distribution of the carbon support, indicating a positive correlation between nitrogen doping amount and platinum dispersion.

Some theoretical studies have further revealed that specific nitrogen configurations can enhance the intrinsic corrosion resistance of adjacent carbon atoms. Li et al. [88] systematically explored the carbon corrosion mechanism of nitrogen-doped carbon supports via density functional theory (DFT) calculations, explicitly highlighting the impact of nitrogen concentration on carbon corrosion resistance. The findings demonstrated that a higher nitrogen doping concentration does not equate to better performance; an increase in nitrogen concentration (e.g., from 1.39% to 8.3%) lowers the energy barrier of the carbon corrosion reaction, resulting in intensified carbon corrosion, as shown in Figure 8b. Specifically, an excessively high nitrogen doping concentration disrupts the stability of the graphene atomic structure, rendering electronegative OH and O more prone to adsorption, thereby deteriorating carbon corrosion resistance. Consequently, precise control of nitrogen concentration is imperative during the modification of carbon supports. This finding underscores a critical principle in designing doped carbons: an optimal doping concentration exists, and over-doping can be counterproductive, as the resulting structural disruption to the carbon lattice can outweigh the benefits of enhanced catalyst anchoring.

Similarly, boron doping introduces electron-deficient sites into the carbon lattice. Kozhushner et al. [45] doped boron (B) into the carbon support to enhance its corrosion resistance and employed it as a platinum support material. The highly graphitized structure of the SB-V support is clearly observable in Figure 8c, which exhibits ordered carbon microcrystals. With the progression of AST, the performance of all systems deteriorated, as presented in Figure 8d, which illustrates the changes in normalized current density at 0.55 V for all fuel cells. After 4000 AST cycles, the boron-containing system retained over 50% of its initial current density, whereas the boron-free system retained less than 30%. Carbon loss measurement results indicated that boron-doped carbon exhibits superior resistance to electrochemical corrosion compared to unmodified carbon, thereby enhancing the stability of the catalyst.

The development of these advanced carbon materials reveals two distinct yet complementary concepts for enhancing stability. One, represented by graphitized carbon and carbon nanotubes, enhances the intrinsic stability of the carbon skeleton itself by increasing crystallinity and reducing defect sites [89,90]. The other, represented by heteroatom doping, improves the functional stability of the catalyst system by strengthening the chemical bonds between the catalyst and the support, thereby avoiding consequences such as platinum detachment caused by corrosion. Although both strategies are effective, they target different aspects of the degradation problem: highly graphitized supports, while inherently robust, may have weak physical interactions with platinum nanoparticles; doped amorphous supports, despite their ability to effectively anchor platinum, are still prone to bulk corrosion. This suggests that the most durable carbon supports in the future will likely be hybrid materials combining the above two strategies, such as nitrogen-doped carbon nanotubes or doped graphitized carbon, thereby leveraging the advantages of both a corrosion-resistant skeleton and strong chemical anchoring sites. While these strategies focus on improving the carbon support itself, an alternative approach involves combining carbon with inherently stable non-carbonaceous materials, such as metal oxides, to form composite supports, a concept that is explored in detail in the following section.

To consolidate the findings from this section and provide a clear comparative overview of the different advanced carbon-based supports,

Table 2. Comparison of properties for different carbon-based catalyst supports.

Support Type	Typical Specific Surface Area (m2/g)	Electronic Conductivity	Corrosion Resistance (Stability)	Reported ECSA/Performance Retention	Key Trade-Off/Advantage
Conventional Carbon Black (e.g., Vulcan XC-72) [46,47,60]	High (>200)	Good	Low	Low (~40–60% loss in typical ASTs)	Trade-off: High initial Pt dispersion at the cost of very poor long-term stability.
Graphitized Carbon Black (GCB) [47,68,69]	Moderate (60–130)	Good to Excellent	High	Very High (e.g., 0.5% potential decay vs. 35% for Pt/C)	Trade-off: Sacrifices some initial surface area for a dramatic improvement in durability.
Carbon Nanotubes (CNTs) [77,78]	High (Variable, >100)	Excellent	Very High	High (e.g., 67% retention after 10k cycles)	Advantage: Excellent balance of high surface area and inherent graphitic stability. Cost and dispersion can be challenging.
Graphene [82,84,85]	Very High (Theoretical max ~2600)	Excellent	Very High	High (Markedly attenuated voltage decay vs. Pt/C)	Advantage: Superior intrinsic properties, but synthesis of high-quality, defect-free material at scale is a primary challenge.

summarizes their key properties. The table benchmarks conventional carbon against graphitized carbons and carbon nanostructures, evaluating metrics such as specific surface area, corrosion resistance, and performance retention. This comparison explicitly highlights the critical trade-off between the high initial surface area, which is beneficial for catalyst dispersion, and the superior long-term durability afforded by more ordered and graphitic structures.

4. Beyond Carbon: Corrosion-Immune Non-Carbonaceous Supports

While engineering carbon has yielded significant improvements, an alternative and arguably more definitive solution to the corrosion problem is to replace carbon entirely with materials that are intrinsically stable in the demanding HT-PEMFC environment. This has led to extensive research into non-carbonaceous supports, primarily metal oxides, carbides, and nitrides [91,92]. These materials offer the promise of near-complete immunity to electrochemical oxidation but introduce new challenges, most notably in achieving sufficient electronic conductivity.

4.1. Metal Oxides: Stability at the Cost of Conductivity

Metal oxides such as titanium dioxide (TiO₂), tin dioxide (SnO₂), and tungsten trioxide (WO₃) are renowned for their excellent chemical and thermal stability. They exhibit extremely strong corrosion resistance in the acidic and oxidative environment of the cathode in HT-PEMFCs, thus emerging as primary candidates for durable catalyst supports [93,94,95]. Additionally, many of these oxides can form strong metal–support interactions (SMSI) with platinum nanoparticles. This interaction not only provides firm anchoring to prevent sintering but also modifies the electronic properties of platinum, and in some cases, even enhances catalytic activity [96].

However, the main obstacle hindering their application is their inherently low electronic conductivity. Most pure metal oxides are wide-bandgap semiconductors or insulators, unable to provide the efficient electron transport required for catalyst supports [97]. To address this challenge, research efforts have primarily focused on two strategic approaches. One promising avenue involves the incorporation of aliovalent cations into the oxide lattice, which can effectively modulate the electronic structure and enhance charge carrier mobility. Ramani et al. [98] systematically investigated the synthesis and characterization of niobium-doped titanium dioxide (Nb-TiO₂) and antimony-doped tin dioxide (Sb-SnO₂), demonstrating that substitutional doping with pentavalent cations (Nb⁵⁺ or Sb⁵⁺) introduces excess electrons into the conduction band, thereby transforming these materials into n-type semiconductors with metallic-like conductivity. This phenomenon is further corroborated by DFT calculations, which reveal a significant reduction in bandgap and the emergence of impurity states near the Fermi level (Figure 9a). Structural simulations of Sb-doped SnO₂ (Figure 9b) further illustrate how dopant-induced lattice distortions and charge redistribution contribute to enhanced electronic conductivity. Leveraging these insights, researchers have successfully developed advanced non-carbon supports that exhibit exceptional durability and electrocatalytic performance. Sun et al. [99] reported a surfactant-free wet-chemical approach for synthesizing mesoporous Nb-TiO₂ hollow spheres, which serve as an ideal substrate for the deposition of platinum nanoparticles (Pt NPs). TEM analysis (Figure 9c) confirms the formation of uniform spherical particles with a mesoporous shell structure, providing a high surface area for Pt NP dispersion and anchoring. By precisely controlling the Pt precursor-to-support ratio, the density and size distribution of Pt NPs can be optimized to maximize electrocatalytic activity. Electrochemical measurements reveal that the PtNP/Nb-TiO₂ catalyst outperforms commercial Pt/C in both activity and stability. Specifically, after 30,000 cycles of AST at high potentials, the PtNP/Nb-TiO₂ catalyst retains ~40% of its initial ECSA, while the Pt/C counterpart retains only 19% (Figure 9d). This two-fold improvement in durability underscores the effectiveness of Nb-TiO₂ as a corrosion-resistant support, mitigating both carbon oxidation and Pt nanoparticle sintering and migration.

SnO₂-based supports also show great potential. Inaba et al. [100] synthesized connected mesoporous antimony-doped tin oxides (CMSbTOs) with controlled mesopore sizes in the range of 4–11 nm and tested their performance and durability as cathode catalyst supports. Under dry conditions, the CMSbTO support with a pore size of 7.3 nm exhibits higher fuel cell performance than solid-core tin oxide-based supports, solid-core carbon supports, and mesoporous carbon supports, and even under low-humidity conditions, it has better proton conductivity within the mesopores. In addition, the CMSbTO support shows high durability under oxidative conditions.

4.2. Transition Metal Carbides: Pt-like Properties and High Stability

Transition metal carbides (TMCs) represent another class of highly promising non-carbon supports. Materials such as tungsten carbide (WC) and titanium carbide (TiC) are ceramic materials that possess an unusual and highly desirable combination of properties: they are extremely hard and refractory, with high melting points and excellent chemical stability, yet they also exhibit high electronic conductivity comparable to that of metals [101,102].

4.2.1. WC

WC has attracted significant attention due to its distinctive electronic structure, which has been characterized by “Pt-like” properties. This characteristic suggests that WC can serve not only as a stable support but also as a co-catalyst, potentially enhancing the ORR activity. Perchthaler et al. [103] compared Pt catalysts supported on WC, tungsten oxide (WOₓ), and those on high surface area carbon (HSAC). Long-term durability tests in HT-PEMFCs over 1000 h (Figure 10a) revealed that Pt/WC catalysts exhibited significantly improved durability compared to Pt/C, with WC-based supports showing lower degradation rates. Liu et al. [104] synthesized nanoscale WC via an improved solution combustion method combined with high-temperature carburization. After carburization, Pt nanoparticles were uniformly distributed on the outer surface of WC nanoparticles; an additional carbon coating on the synthesized WC surface helped stabilize Pt nanoparticles and enhance the overall conductivity of the catalyst. Electrochemical studies (Figure 10b) confirmed that WC provides a stable surface for Pt and displays excellent corrosion resistance in acidic environments. Lori et al. [105] investigated the corrosion resistance, ORR performance, electrocatalytic activity, and durability of Pt-deposited WC in comparison with Pt/C. HRTEM images showed that Pt nanoparticles aggregated after deposition, with these aggregates composed of Pt particles approximately 3–6 nm in size (Figure 10c). The results demonstrated that the specific ORR activity of Pt/WC was over seven times that of Pt/C, along with significantly enhanced durability. Relative changes in the ECSA and E_1/2 of the samples before and after AST (Figure 10d,e) further confirmed the high durability of Pt/WC.

4.2.2. TiC

TiC also possesses high electrical conductivity and thermal stability [106]. Although it has been less extensively studied in HT-PEMFCs compared to tungsten carbide, it has shown application potential. Lobato et al. [107] reported a composite titanium silicon carbide (SiC-TiC) catalyst support for HT-PEMFCs. This support exhibits extremely high heat resistance and electrochemical resistance under harsh conditions. Platinum nanoparticles were successfully deposited on the novel SiC-TiC support, which showed higher stability and less nanoparticle agglomeration compared to the more traditional Vulcan carbon-based catalyst.

Under extreme oxidative potentials, the main degradation pathway for such carbides is their conversion into corresponding metal oxides (e.g., tungsten carbide is converted to tungsten trioxide: WC→WO₃). These oxides generally have low electronic conductivity, which may lead to performance degradation. Zhu et al. [108] synthesized nanocrystalline WC with high surface area and minimal free carbon content via a polymer route. Its physical properties, including solubility in acid solutions, electronic conductivity, and thermal stability, were thoroughly investigated at two high temperatures (95 °C and 200 °C). The results demonstrated that this material exhibits low electrochemical stability in acidic media when subjected to potential cycling at potentials greater than 0.7 V. This has prompted researchers to develop core-shell structured materials. Ignaszak et al. [109] developed a core-shell material, TiC-TiO₂, as a catalyst support, where the stable oxide shell can protect the conductive carbide core, thereby combining the excellent properties of both material classes.

4.3. Transition Metal Nitrides: The Emerging Class

The third category of non-carbon materials that has attracted significant attention is transition metal nitrides. Materials such as titanium nitride (TiN) and niobium nitride (Nb₄N₅) are also ceramic compounds, featuring high electrical conductivity, high hardness, and excellent chemical inertness, which make them suitable candidates for corrosion-resistant supports.

Their potential has been strongly demonstrated in composite support systems. Liu et al. [34] developed a novel microflower-like Nb₄N₅ material and further fabricated Nb₄N₅/C composites as Pt catalyst supports for HT-PEMFCs, with the specific morphology illustrated in Figure 11a. Investigations demonstrated that this catalyst exhibits exceptional performance and durability in HT-PEMFCs. Moreover, the single-cell test results presented in Figure 11b revealed that the Pt/Nb₄N₅/C catalyst delivers superior performance. The MEA achieved an impressive peak power density of 520.48 mW cm⁻² at 150 °C. After 5000 AST cycles, the cell performance degraded by only 5.2%, which is less than half of the performance loss (11.7%) observed for commercial Pt/C catalysts under the same conditions. In this composite, the highly stable nitride particles serve as robust anchoring sites for Pt catalysts, while the carbon component ensures a continuous conductive network within the electrode. TiN, functioning as a durable and conductive coating, has also been extensively investigated for other fuel cell components such as bipolar plates. Lobato et al. [110] evaluated two new SiC-based non-carbon supports, and the XRD patterns in Figure 11c confirmed the structures of SiC and SiC-TiC. 40 wt% Pt was successfully deposited on these supports, and the characterization of their physicochemical and electrochemical properties further verified their stability in the fuel cell environment (as shown in Figure 11d).

The exploration of non-carbon supports reveals a clear and important trend in material design. The initial goal of finding a single material to replace carbon has gradually evolved. It is now evident that the most successful and promising non-carbon supports are not single-component materials, but rather intelligently designed multi-component composites [111]. This “composite material paradigm” takes full advantage of the unique strengths of different types of materials. Whether it is a doped oxide (oxide + dopant), an oxide–carbon composite (oxide + carbon), a carbide–oxide core-shell structure (carbide + oxide), or a nitride–carbon composite (nitride + carbon), the core principle is consistent: one component provides the necessary electrochemical stability and catalyst anchoring effect, while the other component provides the required electronic conductivity. This shift in design philosophy from seeking a “one-size-fits-all” material to the rational design of hierarchical and multifunctional composites represents the most promising direction for developing truly durable catalyst supports for HT-PEMFCs. It is important to note that while these materials show exceptional promise in ASTs, published long-term (>5000 h) operational data in full-scale HT-PEMFC stacks is still limited, representing a critical gap that must be bridged between fundamental research and commercial validation.

5. Comparative Analysis and Performance Benchmarking

To distill the extensive research on advanced catalyst supports into a clear and actionable summary, a direct comparison of their performance and durability is essential. This section synthesizes the quantitative data presented in the preceding sections, providing a benchmark of the current state-of-the-art and allowing for a critical assessment of the relative merits of each material strategy.

5.1. Benchmarking Methodology

The evaluation of electrocatalyst supports relies on a set of key performance indicators (KPIs) that measure both initial activity and long-term stability. In this review, the primary metrics selected for comparison include performance and durability, which together capture the essential requirements of practical application. Performance is typically quantified by the catalyst’s ability to generate power, most commonly expressed as the peak power density (mW cm⁻²), representing the maximum power output achievable by the fuel cell [112,113,114]. Another widely used measure is the current density (A cm⁻²) at a reference voltage, such as 0.6 V, which reflects the catalyst’s effectiveness under representative operating conditions [115,116,117].

Durability, on the other hand, is evaluated through ASTs designed to emulate the degradation mechanisms encountered during extended operation. The key indicators of durability include the percentage loss of ECSA following a defined AST protocol, which reveals changes in the available catalytic sites, and the overall performance loss, which may be reported as the decline in current or power density relative to the initial state [118,119]. In long-term steady-state tests, durability can also be expressed as a voltage decay rate (μV h⁻¹), providing insight into the rate of performance deterioration over time [78,120]. These KPIs together offer a comprehensive framework for benchmarking and comparing electrocatalyst supports across different studies.

It is crucial to acknowledge that making direct, perfectly normalized comparisons across different studies is challenging. Variations in experimental conditions—such as Pt loading, MEA fabrication methods, cell operating temperature and pressure, and specific AST protocols—can significantly influence the reported values. Nevertheless, by compiling and contextualizing the available data, it is possible to discern clear trends and identify the most promising material classes.

5.2. Quantitative Comparison of Support Materials

The progress in developing advanced catalyst supports is best illustrated by a direct comparison of their performance and durability against the conventional carbon black benchmark.

Table 3. Comparative performance and durability of advanced catalyst supports in HT-PEMFCs and simulated environments.

Support Material	Catalyst	Test Conditions	Initial Performance	Durability Metric	Reference(s)
Conventional Carbon
Vulcan XC-72	Pt	HT-PEMFC, AST: 1 k cycles (0.6–1.2 V)	633.8 mA cm−2 @ 0.6 V	58.35% Performance loss	[47]
Vulcan XC-72	Pt	GDE Half-cell, AST: 5k cycles	N/A	~40% ECSA loss	[60]
Commercial Carbon	Pt	PEMFC, High-potential hold	N/A	34.9% Potential decay	[68]
Advanced Carbons
Graphitized Carbon (1800 °C)	Pt	PEMFC, High-potential hold	N/A	0.5% Potential decay @ 1000 mA cm−2	[68]
Graphitized CNT (GCNT)	Pt	HT-PEMFC @ 160 °C, AST: 10 k cycles	0.36 A cm−2 @ 0.6 V	61% Performance retention	[78]
N-doped Carbon	Pt	MEA Test, AST: 30 k cycles	Peak Power Density: 0.91 W cm−2	6.9% Performance decay	[87]
Graphene/CB Composite	Pt	HT-PEMFC @ 160 °C, 100 h test	Peak Power: ~450 mW cm−2	High stability, low voltage decay	[82]
B-doped Carbon	Pt	PEMFC, AST: 4 k cycles	N/A	>50% Current retention	[45]
MWCNT-GNP	Pt-Ru	HT-PEMFC @ 160 °C (reformate)	Peak Power Density: 266 mW cm−2	N/A	[84]
S,P-doped Carbon Nitride	Pt-Co	HT-PEMFC	Enhanced ORR Activity	High stability	[25]
Non-Carbonaceous Supports
Nb-TiO2 (Carbon-free)	Pt	RDE & AST: 30 k cycles (0.6–1.4 V)	Mass Activity: 15% > Pt/C	~40% ECSA Retention (vs. 19% for Pt/C)	[99]
Nb4N5/C (Composite)	Pt	HT-PEMFC @ 150 °C, AST: 5 k cycles	Peak Power Density: 520.48 mW cm−2	5.2% Performance loss (vs. 11.7% for Pt/C)	[110]
Sb-SnO2/C (Composite)	Pt	PEMFC	Power Density: 15% > Pt/C	N/A	[96]
Tungsten Carbide (WC)	Pt	Half-cell, AST: 5 k cycles	Specific Activity: 7x > Pt/C	~60% ECSA Retention (vs. <20% for Pt/C)	[105]
SiC-TiC (Composite)	Pt	HT-PEMFC, 100 h test	N/A	High voltage stability over 100 h	[107,110]

consolidates key data from various studies discussed in this review.

The data compiled in Table 3 reveals several critical trends that define the current landscape of HT-PEMFC support materials.

First, the baseline instability of conventional carbon black is starkly evident [121,122]. With performance losses exceeding 58% and ECSA losses reaching 40% in standard ASTs, it is clear that Vulcan XC-72 and similar materials are unsuitable for applications requiring long-term durability [47].

Second, all advanced support strategies offer substantial improvements in durability. Graphitizing the carbon support dramatically reduces degradation, with potential decay dropping from 34.9% to just 0.5% in one high-potential AST [68]. This highlights the effectiveness of increasing the inherent structural stability of the carbon.

Third, a potential trade-off between initial performance and long-term durability is apparent. Highly graphitized materials, while exceptionally stable, may sometimes exhibit lower initial surface area and thus slightly lower beginning-of-life performance [123,124]. In contrast, nanostructured carbons such as CNTs and graphene offer a powerful combination of high performance and enhanced durability [125,126,127]. The Pt/GCNT catalyst, for instance, delivered the highest initial current density among the carbon-based supports listed (0.36 A cm⁻²), while still maintaining good stability [128].

Fourth, heteroatom-doped carbons and non-carbonaceous composites appear to resolve this trade-off, delivering both high initial performance and exceptional durability [121,129,130,131,132]. The Pt/N-doped Carbon support, with a peak power density of 0.91 W cm⁻² and only 6.9% degradation after 30,000 cycles, represents a major breakthrough, demonstrating that functional stability through enhanced catalyst–support interaction is a highly effective strategy [87]. Similarly, the Pt/Nb₄N₅/C composite, with its high power density and a performance loss of only 5.2%, showcases the success of the composite paradigm, where the nitride provides stability and the carbon provides conductivity [34]. These materials currently represent the frontier of high-performance, durable supports for HT-PEMFCs.

Finally, the data underscore the success of the composite material approach for non-carbonaceous supports. Materials such as Pt/Nb-TiO₂ and Pt/Nb₄N₅/C, which combine a stable oxide or nitride with a conductive component (either through doping or physical mixing with carbon), consistently outperform their individual constituents and provide a viable path to overcoming the conductivity limitations of many corrosion-immune materials.

6. Challenges, Perspectives, and Future Research Directions

The pursuit of durable and high-performance electrocatalyst supports for HT-PEMFCs has driven significant innovation in materials science and electrochemistry. The research landscape has evolved from identifying the fundamental failure of conventional carbon black to developing sophisticated, multifunctional materials that can withstand the uniquely aggressive operating environment. This review has systematically traced this evolution, establishing that while substantial progress has been made, several critical challenges must be overcome to enable the widespread commercialization of HT-PEMFC technology. The path forward lies in the rational design of hierarchical materials, the application of advanced characterization tools, and the co-development of supports for next-generation catalysts.

6.1. Summary of the State-of-the-Art

The core conclusion of this review is that the degradation of conventional carbon black supports via electrochemical corrosion is a primary life-limiting factor for HT-PEMFCs. The combination of high temperature, high cathode potential, and the catalytic activity of platinum itself creates an environment where standard carbon supports are untenable for long-term operation. In response, the scientific community has successfully pursued two major strategies: (1) Engineering Better Carbons: By modifying carbon’s structure and chemistry through graphitization, nanostructuring (CNTs, graphene), and heteroatom doping (N, B), researchers have created supports with vastly improved corrosion resistance and catalyst anchoring. However, a key challenge remains in optimizing the doping concentration, as over-doping can introduce structural defects that compromise the intrinsic stability of the carbon framework, a critical nuance for future material design. Materials such as N-doped carbon and graphitized CNTs have demonstrated remarkable durability and high performance, proving that the carbon platform can be successfully adapted for the HT-PEMFC environment [130,133,134,135,136]. (2) Moving Beyond Carbon: By developing non-carbonaceous supports based on metal oxides (e.g., TiO₂, SnO₂), carbides (e.g., WC), and nitrides (e.g., Nb₄N₅), researchers have created materials that are inherently immune to carbon corrosion. The primary challenge of low conductivity has been effectively addressed through doping and the formation of intelligent composites [95,137,138,139].

While materials from both categories, particularly N-doped carbons and nitride/oxide composites, have demonstrated impressive performance and stability in laboratory settings, no single “perfect” support material that optimally satisfies all criteria of performance, durability, cost, and scalability has yet emerged.

6.2. The Critical Role of Advanced Characterization and Computational Modeling

The significant progress in developing durable electrocatalyst supports, as detailed in this review, is fundamentally underpinned by a powerful toolkit of advanced characterization techniques and computational modeling [140]. These tools are indispensable, providing crucial insights at every stage of research, from the initial synthesis and validation of a new material to understanding the complex, dynamic mechanisms of its degradation within an operating fuel cell.

A foundational understanding of these materials is built upon a suite of ex situ characterization techniques [58]. TEM/SEM is essential for visualizing the morphology, particle size, and, most critically, the dispersion of catalyst nanoparticles on the support surface. XRD is used to determine the crystal structure, phase purity, and crystallite size, which is vital for confirming the degree of graphitization in advanced carbons or the formation of desired oxide, carbide, and nitride phases [141]. Furthermore, surface-sensitive techniques are critical: BET analysis measures the specific surface area and pore size distribution that govern catalyst loading and mass transport, while XPS confirms the surface elemental composition and chemical states, providing direct evidence of successful heteroatom doping [87]. The quantitative performance data presented throughout this review rely on electrochemical methods such as CV to measure the ECSA, RDE analysis to evaluate intrinsic kinetic activity, and ASTs to assess long-term durability.

However, these ex situ or “post-mortem” techniques only provide static snapshots of the material before or after testing. To truly understand the dynamic failure modes, it is necessary to observe them as they occur. This is why in situ and operando characterization techniques are becoming increasingly vital. These methods allow researchers to probe the catalyst layer under realistic HT-PEMFC operating conditions, providing direct, real-time insights. For example, in situ TEM can visualize the actual migration and agglomeration of Pt nanoparticles on a support during potential cycling, while synchrotron-based X-ray imaging can map the distribution and potential loss of phosphoric acid within the electrode during operation [133].

Complementing these experimental efforts, computational modeling, particularly DFT, offers powerful predictive capabilities and atomic-level understanding. DFT allows researchers to screen the stability of new material compositions, calculate the energy barriers for carbon corrosion, and model catalyst–support interactions before engaging in time-consuming synthesis. As demonstrated in this review, DFT has been instrumental in explaining how heteroatom doping strengthens catalyst anchoring and how doping metal oxides alters their electronic structure to enhance conductivity [88]. In the future, high-throughput computational screening combined with machine learning will further accelerate the discovery cycle for novel, high-performance supports.

To provide a clear overview for researchers in the field, the common synthesis and characterization techniques central to the development of advanced catalyst supports are summarized in

Table 4. Overview of common synthesis and characterization techniques for catalyst supports.

Technique	Purpose in Catalyst Support R&D	Reference(s)
Synthesis Methods
Polyol / Wet-Chemical	Solution-based synthesis of catalyst nanoparticles (e.g., Pt) for deposition onto supports.	[142,143,144]
Hydrothermal / Solvothermal	Synthesis of crystalline materials, such as oxides or nanostructures, under elevated temperature and pressure.	[145,146]
Sol-Gel	Synthesis of porous metal oxides (e.g., TiO2, SnO2) from molecular precursors.	[143,147]
Chemical Vapor Deposition (CVD)	Growth of high-quality nanostructures, such as CNTs and graphene.	[148,149]
Pyrolysis / Heat Treatment	Graphitization of carbon supports; synthesis of doped carbons and carbides from precursors.	[78,150]
Characterization Techniques
X-Ray Diffraction (XRD)	To determine the crystal structure, phase purity, and crystallite size of the support and catalyst.	[151,152]
Transmission/Scanning Electron Microscopy (TEM/SEM)	To visualize the morphology, particle size, and dispersion of the support and catalyst nanoparticles.	[153,154]
Brunauer–Emmett–Teller (BET) Analysis	To measure the specific surface area and pore size distribution of the support material.	[155,156]
X-Ray Photoelectron Spectroscopy (XPS)	To determine the surface elemental composition and chemical/oxidation states (e.g., N-doping, Pt oxidation).	[157,158]
CV	To measure the ECSA of the catalyst.	[159,160]
Rotating Disk Electrode (RDE)	To evaluate the intrinsic kinetic activity of the catalyst for ORR in a controlled half-cell environment.	[161,162]
AST	To assess the long-term durability of the catalyst and support by simulating operational stressors.	[163,164]

.

6.3. Overarching Challenges for Future Development

Despite significant scientific progress, several overarching challenges remain before advanced materials can be successfully translated from the laboratory to commercial fuel cell stacks. One major issue is durability under real-world conditions. Most available durability data are obtained through standardized ASTs, which mainly simulate high-potential cycling associated with start-up and shutdown events. However, actual fuel cell operation involves a much more complex combination of thermal cycling, load cycling, humidity fluctuations, and prolonged steady-state operation [52,165]. Validating the stability of materials under such dynamic and coupled stressors is, therefore, an essential step toward practical application.

Another critical challenge lies at the catalyst–support–electrolyte interface. The catalyst layer functions as a sophisticated three-component system, whose long-term performance depends on the delicate interactions among the catalyst, the support material, and the phosphoric acid electrolyte [166]. The poisoning of active sites by phosphate anion adsorption, as well as the uneven distribution and possible loss of phosphoric acid within the electrode, are persistent problems that must be addressed through careful support material design and interface engineering [12,167].

Cost and scalability also present substantial barriers. Many of the most promising advanced support materials—such as high-purity single-walled carbon nanotubes, chemical vapor deposition (CVD)-grown graphene, or intricate composites—are currently produced using methods that are prohibitively expensive and technically complex for large-scale manufacturing. However, the pathways to commercialization and their readiness vary significantly between these material classes.

• Graphitized carbons represent a potentially near-term solution; while the high-temperature heat treatment is energy-intensive, it is an additional process applied to an already low-cost, scaled industrial material, making it a pragmatic approach for incremental durability improvements.
• Non-carbonaceous composites, such as doped metal oxides, present a promising mid-to-long-term pathway. Although laboratory synthesis can be complex, many of the base oxides (e.g., TiO₂) are inexpensive commodity chemicals. The primary hurdle is the development of scalable synthesis routes, such as spray pyrolysis, that can produce uniform materials cost-effectively.
• In contrast, high-purity carbon nanostructures such as CNTs and graphene, despite their superior performance, currently face the most significant long-term commercialization barriers due to persistently high production costs, limiting their immediate application to niche areas pending a major manufacturing breakthrough.

Developing scalable and cost-effective production routes for these materials remains a key engineering hurdle. For non-carbonaceous supports, the challenge lies in developing scalable manufacturing routes for the promising composite materials that can achieve uniform mixing and strong interfacial contact between the stable and conductive components without incurring prohibitive costs.

Finally, the absence of standardized testing protocols and performance reporting metrics poses a further obstacle [168,169]. The lack of universally accepted AST procedures and consistent benchmarks makes it difficult to directly and rigorously compare results across different research groups, leading to fragmentation in the field [165,170]. Establishing international standards for both testing and reporting would help accelerate the discovery, validation, and deployment of superior materials.

6.4. Future Outlook and Research Recommendations

Based on the current state-of-the-art and the remaining challenges, the future of HT-PEMFC catalyst support research is expected to concentrate on several interconnected directions. A particularly promising avenue lies in the deliberate design of hierarchical, multifunctional composite supports. Rather than relying on simple two-component mixtures, future materials are likely to evolve into complex architectures where each constituent plays a distinct, strategically placed role. For example, a support particle might feature a stable oxide core to ensure robust Pt anchoring, enveloped by a conductive carbide or graphitic carbon shell to enable efficient electron transport, with nitrogen doping at the surface interface to further enhance the metal–support interaction. Such a tailored, multifunctional design could significantly improve both activity and durability.

A deeper understanding of how these advanced materials degrade under real operating conditions is equally crucial. To move beyond a trial-and-error approach in materials development, advanced in situ and operando characterization techniques will play an increasingly important role. Methods such as in situ transmission electron microscopy (TEM), synchrotron-based X-ray imaging and spectroscopy, and confocal Raman microscopy can provide direct, real-time insights into catalyst migration, support corrosion, and phosphoric acid distribution within the catalyst layer under HT-PEMFC operating conditions. These powerful tools offer invaluable feedback that will guide the rational design of more durable and effective materials.

In parallel, computational modeling and machine learning approaches will become indispensable in accelerating material discovery and optimization [10]. DFT will continue to offer fundamental insights into catalyst–support interactions and predict the stability of new material compositions. Given the vast compositional and structural design space of potential composite supports, high-throughput computational screening coupled with machine learning algorithms can efficiently identify promising candidates. These data-driven strategies are expected to significantly shorten the development cycle for novel, high-performance support materials.

Furthermore, it is crucial to recognize that the development of advanced supports is intrinsically linked to the composition of the catalyst nanoparticles they host. While this review has focused on the support, a parallel and vital strategy for enhancing both activity and durability is the use of platinum-alloy catalysts (e.g., Pt-Co, Pt-Ni, Pt-Ru). Alloying Pt with transition metals can alter its electronic structure, leading to higher intrinsic activity for the oxygen reduction reaction and improved resistance to dissolution and agglomeration. The synergistic effects between a highly stable, engineered support (such as N-doped graphene or a carbide composite) and a durable Pt-alloy nanoparticle represent a powerful approach for creating next-generation electrocatalysts. Future research should therefore increasingly focus on the co-design and optimization of both the catalyst alloy composition and the support structure to maximize these synergistic benefits and achieve the desired long-term stability for HT-PEMFCs.

Finally, with the ultimate goal of reducing cost and reliance on platinum, research must also focus on the co-development of support materials specifically suited to PGM-free catalysts, such as Fe–N–C systems. These catalysts exhibit distinct activity and stability requirements compared to Pt-based systems, and the optimal support design for Pt may not translate effectively to PGM-free alternatives. Future efforts should therefore aim to tailor support materials that complement and stabilize these emerging catalytic systems, enabling their successful deployment in HT-PEMFC applications.

7. Conclusions

The long-term durability of HT-PEMFCs is critically hindered by the electrochemical corrosion of conventional carbon black catalyst supports in the aggressive operating environment. This review has systematically analyzed the two primary strategies pursued to overcome this challenge: the re-engineering of carbon-based materials for enhanced stability and the complete replacement of carbon with intrinsically corrosion-immune non-carbonaceous alternatives.

Within the realm of advanced carbons, strategies such as graphitization, the use of inherently stable nanostructures such as CNTs and graphene, and heteroatom doping have all proven to be highly effective at mitigating corrosion and improving catalyst anchoring. For non-carbonaceous materials, such as metal oxides, carbides, and nitrides, the principal challenge of low electronic conductivity has been successfully addressed through innovative approaches such as elemental doping and the formation of conductive composites.

A critical synthesis of the current state-of-the-art reveals that no single material has emerged as a perfect, one-size-fits-all solution. Instead, the most promising and successful approaches converge on a central theme: the rational design of multifunctional, hierarchical composite materials. Whether it involves creating nitrogen-doped graphitized nanocarbons or combining stable metal oxides with conductive components, the future of durable supports lies in intelligently integrating the distinct advantages of different material classes to meet the multifaceted demands of the HT-PEMFC environment. The continued advancement of these sophisticated support materials is pivotal. By moving beyond simple, single-component systems to rationally designed composites, the field is poised to develop the next-generation electrocatalysts required to unlock the full potential of HT-PEMFCs as a durable, efficient, and commercially viable clean energy technology.