Advancements in Non-Precious Metal Catalysts for High-Temperature Proton-Exchange Membrane Fuel Cells: A Comprehensive Review

Abstract

High-Temperature Proton-Exchange Membrane Fuel Cells (HT-PEMFCs) represent a promising clean energy technology and are valued for their fuel flexibility and simplified balance of plant. Their commercialization, however, is critically hindered by the prohibitive cost and resource scarcity of platinum-group metal (PGM) catalysts. The challenge is amplified in the phosphoric acid (PA) electrolyte of HT-PEMFCs, where the severe anion poisoning of PGM active sites necessitates impractically high catalyst loadings. This review addresses the urgent need for cost-effective alternatives by providing a comprehensive assessment of recent advancements in non-precious metal (NPM) catalysts for the oxygen reduction reaction (ORR) in HT-PEMFCs. It systematically explores synthesis strategies and structure–performance relationships for emerging catalyst classes, including transition metal compounds, metal–nitrogen–carbon (M-N-C) materials, and metal-free heteroatom-doped carbons. A significant focus is placed on M-N-C catalysts, particularly those with atomically dispersed Fe-Nx active sites, which have emerged as the most viable replacements for platinum due to their high intrinsic activity and notable tolerance to phosphate poisoning. This review critically analyzes key challenges that impede practical application, such as the trade-off between catalyst activity and stability, mass transport limitations in thick electrodes, and long-term degradation in the harsh PA environment. Finally, it outlines future research directions, emphasizing the need for a synergistic approach that integrates computational modeling with advanced operando characterization to guide the rational design of durable, high-performance catalysts and electrode architectures, thereby accelerating the path to commercial viability for HT-PEMFC technology.

1. Introduction

The combustion of fossil fuels is largely responsible for about 90 percent of worldwide CO₂ emissions. As illustrated in Figure 1, the CO₂ emissions hit an all-time high of 37.8 billion tons CO₂ in 2024. Fossil fuels are harmful to the environment and are in short supply. However, the worldwide demand for energy has increased by 2.2% in 2024, significantly higher than the annual average of 1.3% observed between 2013 and 2023 [1,2,3,4]. Great efforts have been made to research and develop clean and sustainable energy technologies such as photovoltaics and fuel cells, which have become a global priority [5]. Fuel cells are among the most potential energy conversion technologies, converting chemical energy into electricity through electrochemical reactions. Fuel cells are capable of continuously producing electricity with superior performance and near-zero greenhouse gas emissions, serve as a potential replacement for conventional electricity and heat production, and can be used in an extensive variety of applications, such as household devices, automobile vehicles, and megawatt power plants [6,7,8,9]. Fuel cells are classified primarily based on anions/cations and/or the liquid/solid nature of the electrolyte into polymer electrolyte membrane (PEM) fuel cells, direct methanol fuel cells (DMFCs), alkaline fuel cells (AFCs), phosphoric acid fuel cells (PAFCs), molten carbonate fuel cells (MCFCs), and solid oxide fuel cells (SOFCs). Among these fuel cells, the PEMFC has the highest energy density, simple start-up, and promising clean energy technology. PEMFCs are categorized into two different types based on their working temperature: Low-Temperature PEMFCs (LT-PEMFCs) operate between 70 and 95 °C, and high-temperature PEMFCs (HT-PEMFCs) operate between 150 and 250 °C [10,11,12].

Generally, LT-PEMFCs operate with perfluoro sulfonic acid (PFSA) membranes (e.g., Nafion) as PEMs [13,14]. Hydrogen or hydrogen-rich mixtures of gases are used as fuel for the anode and air or oxygen as fuel for the cathode, and both fuels should be in continuous supply. In particular, LT-PEMFCs are susceptible to impurity and consequently require high-purity hydrogen (99.9999%), and the cost is very high [15]. Hydrogen fuel can be produced using a variety of resources, including biomass, fossil fuels, and water electrolysis [16,17,18]. Furthermore, the lack of transportation infrastructure for hydrogen transfer and refueling is one of the main obstacles preventing the commercial marketing of hydrogen-fuel-cell automobiles and other applications. While methanol is more cost-effective than the hydrogen and much simpler to transport and store, using it in LT-PEMFCs with reformed hydrogen is difficult because its carbon monoxide content must be less than 10 ppm to prevent serious active site poisoning [19,20,21,22,23,24]. Beyond that, its temperature has to be decreased from its reforming temperature, which is >200 °C, to the operational condition of LT-PEMFCs (~85 °C) (Figure 2a) [25,26,27]. Initially the reformate gas is cooled down by condenser I, and the methanol from the reformate gas is separated by cold trap I. The reformate gas is then cleaned through a scrubbing bottle. After that, the reformate gas enters the methanation chamber to lower the carbon monoxide concentration. Condenser II further cools the reformatted gas across from the methanation chamber, which has a much lower carbon monoxide concentration; additionally, vapor separation occurs in cold trap II. Because of the complicated nature of these devices with cooling, vapor formation, and carbon monoxide cleaning requirements, there is a great need for fuel cells working at elevated temperature with improved CO tolerance, without the requirement for carbon monoxide purification and thermal utilization for evaporation [28,29]. HT-PEMFCs generally provide greater energy efficiency with a wider range of fuel alternatives and can employ low-noble metal and non-noble metal catalysts. A PEMFC working at 100–300 °C is capable of (i) improving the kinetic reaction with a low loading of noble metal and non-noble metal catalysts in the hydrogen oxidation reaction (HOR) and oxygen reduction reaction (ORR); (ii) enhancing tolerance to catalyst poisoning by CO, NO_x, sulfur oxide, and hydrogen sulfide, with possibilities involving low-purity hydrogen; (iii) facilitating the transfer of heat without liquid water complications; and (iv) improving the thermodynamic property of the generated heat [30,31,32,33,34,35,36,37]. In HT-PEMFCs, a mixture of methanol and water can be employed as fuel, which is injected into an evaporator prior to being transported to the reformer (Figure 2b). The operation of steam reforming subsequently transforms this fuel mix into a hydrogen-rich gas along with high carbon monoxide content compared with LT-PEMFCs, due to the greater CO-tolerant property of HT-PEMFCs. Additionally, this technology makes it possible to heat the evaporator by using the heat which is generated by the HT-PEMFC [38,39,40,41]. To satisfy these specifications, liquid water-free HT-PEMFCs have been established with a proton-conducting acid which has an extreme boiling temperature (such as PA and phosphate) enclosed in a thermally stable proton-conducting polymer membrane to a maximum of 400 °C [10,42].

Table 1. Comparative overview of Low-Temperature (LT) and High-Temperature (HT) PEMFCs.

Feature	LT-PEMFC	HT-PEMFC
Operating Temperature	50–100 °C	120–200 °C
Electrolyte	Perfluorosulfonic acid (PFSA) polymer (e.g., Nafion)	Phosphoric acid (PA)-doped polymer (e.g., Polybenzimidazole, PBI)
Proton Conduction Mechanism	Vehicular (via H2O molecules)	Grotthuss (hopping via PA network)
Water Management	Requires external humidification; prone to flooding/dehydration	No external humidification needed; simplified water removal (vapor phase)
Thermal Management	Requires large cooling system due to small ΔT	Simplified with smaller cooling system due to large ΔT; high-quality waste heat
Fuel Impurity Tolerance (CO)	Very low (<10 ppm)	Very high (1–3%)
Typical Cathode Catalyst	Pt/C, Pt-alloy/C	Pt/C, Pt-alloy/C
Typical Cathode Pt Loading	0.1–0.4 mg cm−2	>1.0 mg cm−2
Key Advantages	Fast start-up and high power density	Fuel flexibility, simplified BoP, CHP potential, and enhanced kinetics
Key Challenges	Water management, CO poisoning, and high-purity H2 requirement	Catalyst poisoning by phosphate, long-term durability, and acid management

provides a concise comparison of the key features of LT-PEMFCs and HT-PEMFCs, highlighting the fundamental differences that motivate the distinct research and development trajectories for each technology.

These combined advantages make HT-PEMFCs a particularly compelling technology for applications where fuel flexibility, system simplicity, and robustness are paramount, such as in heavy-duty vehicles, maritime shipping, aviation, and stationary power generation [10]. They represent a strategic technological pathway for sectors where the stringent fuel purity and complex water management requirements of LT-PEMFCs pose significant economic and logistical barriers.

1.1. Working Principles and Electrochemistry of HT-PEMFCs

The structure of an HT-PEMFC basically looks the same as the traditional PEMFC, as it possesses a polymeric proton-conducting membrane positioned within an anode and cathode supporting layer, but the type of membrane used here varies (e.g., PA-PBI) (Figure 3). While the HT-PEMFC operates between 150 and 200 °C, hydrogen gas is supplied to the anode surface without humidification, and under the action of a Pt catalyst, it decomposes into protons and electrons (Equation (1)), whereas the proton and electron travel towards the cathode via a PEM and an outer circuit. In the meantime, oxygen flows to the cathode catalyst surface, and water is generated when protons and electrons react with oxygen under the action of the catalyst through the process of ORR (Equation (2)).

Anode: 2H₂ → 4H⁺ + 4e⁻

(1)

Cathode: O₂ + 4H⁺ + 4e⁻ → 2H₂O

(2)

Overall: O₂ + 2H₂ → 2H₂O

(3)

Two primary perspectives can be used to present the kinetic study of the HOR and ORR for HT-PEMFCs: the reaction routes and how they impact the response rate. Figure 4a–c show the HOR and ORR pathways [43,44]. In the specific process of the HOR, hydrogen flows into the catalyst via the catalyst layer (CL) pores and adsorbs on the active areas, weakening the H-H bond strength before breaking into the adsorbed H atom [45]. Under the influence of the water molecules and the electrode’s driving strength, the H atom loses one electron to become a proton [46,47,48]. HORs are split into two categories: the Tafel–Volmer process consisting of Tafel and Volmer steps (Equations (4) and (5)), and the Heyrovsky–Volmer process, which includes Heyrovsky and Volmer steps (Equations (6) and (7)); these steps are known as the Heyrovsky–Tafel model [49,50].

HOR: H₂ 2H⁺ ⇌ + 2e⁻

(4)

Tafel step: H₂ → 2H*

(5)

Heyrovsky step: H₂ → H* + H⁺ +e⁻

(6)

Volmer step: H* → H⁺ + e⁻

(7)

In HT-PEMFCs, the ORR is typically broken down into three primary processes: (i) oxygen initially diffuses and adsorbs over the catalyst’s surface; (ii) electrons move toward the adsorbed oxygen molecules via the electrode, and during this time, the O-O bond breaks down; and (iii) the resultant substance exits from the electrolyte [52,53]. Additionally, the ORR at increased temperatures may be classified into two primary categories, equivalent to LT-PEMFCs, two-electron transfer (Equation (8)) and four-electron transfer (Equation (9)), as illustrated in Figure 4b,c. Due to partial reduction of O₂ to H₂O₂, the former lowers the catalyst selectivity and is detrimental to the performance of HT-PEMFCs. In contrast, the latter facilitates the ORR process by reducing O₂ into H₂O. There are many oxygen-containing substances involved in the ORR process in the two-electron pathway, such as OOH*, O*, and OH* [49,54]. The overall reaction rate is determined by the voltage needed to convert the intermediates [51,52].

ORR:
Two-electron transfer: O₂ + 2H⁺ + 2e⁻ → H₂O₂

(8)

Four-electron transfer: O₂ + 4H⁺ + 4e⁻ → 2H₂O

(9)

The Butler–Volmer (B–V) equation, activation energy, and activation overpotential are frequently employed to assess the kinetic rate in HT-PEMFCs [55]. The energy threshold that should be crossed during the reactant-to-product conversion process is known as activation energy, and it is mostly influenced by the three-phase area. In principle, the faster the reaction rate, the lower the activation energy. The activation overpotential’s value is mostly determined by the reactive substance’s activity, which is determined by certain conditions. Along with the previous two factors, a theoretical base for the predictions of kinetic changes is provided by the Butler–Volmer (B–V) equation (Equation (10)). This might be acquired through exchange current density calculation processes.

$$j=j_0\left[\mathit{exp}\left({\displaystyle \frac{{\alpha }_{a}nFƞ}{RT}}\right)-exp\left(-{\displaystyle \frac{{\alpha }_{c}nFƞ}{RT}}\right)\right]$$

(10)

where j₀—exchange current density; α_a—anodic charge transfer coefficient; and α_c—cathodic charge transfer coefficient.

1.2. Problem Associates with PGM Catalyst in HT-PEMFCs

For decades, platinum (Pt) and its alloys, typically dispersed as nanoparticles on a carbon support (Pt/C), have been the undisputed state-of-the-art electrocatalysts for PEMFCs. One of the primary causes of HT-PEMFC performance losses is the degradation process of platinum (Pt) and platinum-group metal (PGM) catalyst, which is similar to LT-PEMFCs (Figure 5a). Meanwhile, the existence of PA at higher operational temperatures complicates catalyst degradation [56,57]. The acidic atmosphere of HT-PEMFCs leads the platinum particles to gradually dissolve into Pt²⁺ ions and then redeposit on already existing Pt nanoparticles, creating large-diameter nanoparticles and reducing the amount of active surface area [38,56,58]. The dissolved nanoparticles migrate to other regions of the MEA that are inaccessible to reactant gases [59,60]. The action of dissolution, migration, and redeposition of Pt nanoparticles that results in the formation of nanoparticles is called Ostwald ripening [61,62,63]. Furthermore, the neighboring platinum nanoparticles diffuse to create larger nanoparticles, and this process is called platinum agglomeration [64]. This process occurs more quickly whenever the size of particles is smaller because of the larger Gibbs free energy [65]. The gradual increase in the size of platinum particles will reduce the electrochemical surface area (ECSA) and performance of fuel cells. The Pt particles are dissolved according to the following equations [66,67] under open circuit voltage (OCV) conditions:

Pt + H₂O → PtO + 2H⁺ + 2e⁻

(11)

PtO + 2H⁺ + 2e⁻ ↔ Pt²⁺ + H₂O

(12)

Another cause of degradation in HT-PEMFCs is carbon support corrosion, which may lead to the separation and aggregation of catalyst particles [68,69]. In particular, during the operation of HT-PEMFCs, the bombardment of certain energetic free radicals causes the corrosion of the carbon support. Water with a high potential frequently produces these energetic free radicals. When the potential rises, these free radicals oxidize to more stable oxides. This specific phenomenon of carbon corrosion is shown in the following equation [10]:

C* + H₂O → C-OH + H⁺ + e⁻ (E = 0.20 V vs. RHE)

(13)

C-OH → C = O + H⁺ + e⁻ (E = 0.80 V vs. RHE)

(14)

C-OH + H₂O → C* + CO₂ + 3H⁺ + 3 e⁻ (E = 0.95 V vs. RHE)

(15)

Another potential problem with HT-PEMFCs is phosphate adsorption over the Pt catalyst, which hinders the sites of reaction and inhibits the HOR and ORR processes [70,71,72,73]. Phosphate molecule adsorption on the cathode electrode is influenced by temperature and voltage potential. Figure 5b is a schematic diagram of potential dependency. Adsorption of hydrogen exists just below 300 mV, while adsorption of oxygen does above 800mV. Phosphate adsorption occurs between voltages of 300 and 800 mV, and this voltage range diminishes as the temperature increases [70,74]. Although phosphate adsorption also takes place at the anode electrode, it does not poison the catalyst and might potentially help to stabilize hydrogen adsorption. CO in the anode gaseous supply can boost phosphate adsorption, and at the same time, the existence of water reduces coverage [72,75].

Figure 5. Degradation mechanisms in fuel-cell catalysts. (a) Pt nanoparticle deterioration on carbon support under operational stress [76]. Reproduced with permission. Copyright 2014, Elsevier. (b) Voltage-dependent adsorbate deposition behavior on Pt surfaces. Reprinted/adapted with permission from Ref. [70]. Copyright 2016, American Chemical Society.

Figure 5. Degradation mechanisms in fuel-cell catalysts. (a) Pt nanoparticle deterioration on carbon support under operational stress [76]. Reproduced with permission. Copyright 2014, Elsevier. (b) Voltage-dependent adsorbate deposition behavior on Pt surfaces. Reprinted/adapted with permission from Ref. [70]. Copyright 2016, American Chemical Society.

The platinum problem in High-Temperature Proton-Exchange Membrane Fuel Cells (HT-PEMFCs) is surprisingly complicated, as the higher operating temperature enhances the intrinsic kinetic activity of platinum (Pt), but this benefit is largely exceeded by severe poisoning effects from the phosphate (PA) electrolyte [77]. Phosphate anions strongly adsorb onto the Pt surface, blocking active sites and inhibiting the oxygen reduction reaction (ORR) [78], which necessitates much higher platinum loadings in HT-PEMFCs—often exceeding 1.0 mg of Pt per square centimeter—compared with the 0.1–0.4 mg_Ptcm⁻² typical for the low-temperature counterparts. This counterintuitive requirement negates the kinetic advantage of high temperature and significantly increases costs, making the development of non-precious metal (NPM) catalysts not only desirable but essential to the economic viability of HT-PEMFC technology. Despite advances in NPM catalysts with improved activity and durability, their commercial adoption remains limited by critical challenges such as uncertainties in large-scale synthesis, batch-to-batch reproducibility, and long-term stability under real operating conditions. Additionally, the cost-effectiveness of these catalysts must be evaluated not only in terms of material costs but also with regard to electrode processing, system integration, and maintenance. This review provides a comprehensive assessment of NPM catalysts by analyzing their electrochemical performance under HT-PEMFC conditions while addressing challenges related to scale-up, manufacturing consistency, and techno-economic viability. By bridging the gap between laboratory-scale development and commercial application, this work contributes to advancing low-cost, sustainable, high-performance fuel-cell technologies, ultimately fostering the development of cost-effective, durable, and efficient HT-PEMFC systems.

2. Types of Non-Precious Metal (NPM) Catalysts

The high loading of Pt-group metal catalysts is required to increase the possibility of using HT-PEMFCs for real-world applications because of the phosphate chemisorption of Pt-based catalysts at cathode electrodes. The expensive price, limited availability, and inadequate catalytic area in a PA environment has motivated research and development of NPM catalysts for HT-PEMFCs. NPM catalysts with affordable price, easily obtainable and widely available materials, and simplicity in fabrication have received significant interest in recent years. A timeline chart illustrating the advancements in NPM catalysts for HT-PEMFCs is shown in Figure 6. In 2014, the first report of an NPM catalyst made for ORR applications to use as a cathode electrode in HT-PEMFCs was published [79,80,81]. Over a decade, NPM catalysts have been studied for HT-PEMFCs and have shown substantial enhancement in performance, long-term stability, and operational efficiency. Above all, the NPM catalysts’ stability in HT-PEMFCs is significantly better than in LT-PEMFCs [82,83,84,85].

2.1. Transition Metal Catalysts

The high Pt loading required at the cathode to overcome the slow kinetics of the ORR is a major challenge for conventional HT-PEMFCs. Consequently, developing NPM catalysts is crucial to reducing the reliance on Pt. Among these, earth-abundant transition metals like Mn, Fe, Co, and Ni are promising candidates due to their affordability, stability, and good ORR activity [86,87,88]. An early and effective approach involved creating active sites by pyrolyzing nitrogen-containing polymers with a metal precursor. For instance, Zelenay et al. [79] developed a notable PANI-Fe-C catalyst by polymerizing aniline on a carbon support in the presence of an iron precursor, followed by high-temperature pyrolysis and an acid leaching step to remove unstable species. The resulting high-surface-area catalyst (845 m²/g, Figure 7a–c) demonstrated superior ORR performance and excellent tolerance to phosphate ions in 5.0 M PA, even outperforming a conventional Pt/C catalyst (Figure 7d).

Building on these foundational strategies, subsequent research has focused on creating specific and highly stable active phases, such as metal carbides, to enhance durability and performance. Li et al. [82] exemplified this by preparing Fe₃C nanoparticles encapsulated in graphitic shells via a one-step pyrolysis method using ferrocene and cyanamide. They found that a pyrolysis temperature of 700 °C was optimal for creating hollow microspheres where active Fe₃C nanoparticles were protected by a thin graphitic layer (Figure 8a). This protective shell proved critical to durability, as shown by a 2000-cycle accelerated stress test (AST) in a half-cell (Figure 8b), with Transmission Electron Microscopy (TEM) analysis confirming minimal structural change post-test (Figure 8c). When tested in an HT-PEMFC, the catalyst’s performance significantly improved by raising the operating temperature from 120 °C to 180 °C, achieving a peak power density of 60 mW/cm², and showing reasonable long-term stability (Figure 8d).

A more advanced strategy for creating uniformly distributed active sites involves using Metal–Organic Frameworks (MOFs) as sacrificial templates. This method offers precise control over the final catalyst’s structure and porosity. Illustrating this, Strickland et al. [89] developed an FePhen@MOF-ArNH₃ electrocatalyst that showed bamboo-like carbon nanotube (CNT) morphology (Figure 9a) and minimal activity loss compared with Pt/C in a PA environment (Figure 9b). To understand the catalyst’s structural and electronic properties and its interaction with phosphate anions, the team employed in situ X-ray absorption spectroscopy (XAS) and the Δμ method (Figure 9d). This detailed characterization supported the observation of excellent single-cell HT-PEMFC performance at 200 °C in both pure oxygen and air (Figure 9c).

Pushing the boundaries of catalyst design further, another key direction is maximizing atom efficiency through the development of single-atom catalysts (SACs). These catalysts aim to create the highest possible density of active sites. In a pioneering study, Yu et al. [90] synthesized microporous/mesoporous catalysts (LEDFe5-NH₃) using an organometallic EDTA-Fe complex as a precursor (Figure 10a). Characterization revealed that iron was distributed as both single atoms (Fe-Nₓ sites) and nanoparticles within the N-doped carbon matrix (Figure 10b). This unique structure, combining high porosity with highly effective single-atom Fe-Nₓ sites, resulted in outstanding ORR activity and remarkable durability, withstanding 30,000 potential cycles in half-cell tests (Figure 10c). When integrated into an HT-PEMFC, the catalyst also showed excellent performance, attributed to its superior active site density compared with traditional Pt/C (Figure 10d).

2.2. Metal–Nitrogen–Carbon (M-N-C) Catalysts

Among the various PGM-free catalysts explored for HT-PEMFCs, those based on M-N-C have emerged as the most viable alternatives to conventional Pt/C. In particular, catalysts featuring atomically dispersed Fe-Nₓ moieties have demonstrated significant potential, attracting considerable research interest in recent years [85,91,92,93]. Pioneering studies by Gokhale et al. and Hu et al. [94,95] in 2018 first demonstrated the application of Fe-N-C cathodes in HT-PEMFCs (Figure 11a–d). Their findings revealed a common challenge: while performance was comparable to Pt/C at low current densities (<100 mA cm⁻²), significant voltage losses occurred at higher densities. This drop was attributed to mass transport limitations, a consequence of the catalyst’s lower volumetric activity which necessitated thick catalyst layers (CLs) with high loadings (e.g., 2 to 7.8 mg cm⁻²), which impede oxygen flow.

The critical role of mass transport was further highlighted by subsequent research. For instance, Byeon et al. [97] observed very poor performance from an Fe-N-C catalyst (Figure 11e,f), which they linked to substantial mass transport resistance caused by an exceptionally high catalyst loading of 21 mg cm⁻². In a more detailed comparative study, Bevilacqua et al. [98] employed electrochemical impedance spectroscopy (EIS) and distribution of relaxation times (DRT) analysis. They quantified that an Fe-N-C-based membrane electrode assembly (MEA) exhibited 2.5 times greater resistance than its Pt/C and PtCo/C counterparts. This was ascribed to a combination of restricted O₂ diffusion and hindered access of the PA electrolyte to the active Fe-N_x sites.

Beyond initial performance, the long-term durability of Fe-N-C catalysts, particularly the stability of the carbon support and active sites, is a major area of investigation. To this end, Schmies et al. [96] investigated the stability of four distinct Fe-N-C catalysts with varying carbon support properties. A crucial finding was the dramatic voltage drop—up to 27%—within the first 24 h of a 90 h constant load test. Post-mortem and operando analyses identified two primary degradation mechanisms: the corrosion of the carbon support and the deactivation or loss of active Fe-N_x sites, likely via dissolution into metal clusters or oxides. Interestingly, DRT analysis also detected a temporary activation effect attributed to improved electrolyte distribution, but this slight improvement was ultimately overshadowed by the continuous degradation of active sites.

Collectively, these studies illustrate a clear research trajectory. While Fe-N-C catalysts are promising, overcoming challenges related to mass transport and long-term durability is essential [99,100]. The detailed electrochemical and physical characterization methods employed in these research studies are paving the way for the rational design of more robust and high-performance Fe-N-C catalysts for HT-PEMFCs.

2.3. P-Doped Metal–Nitrogen–Carbon (M-N-C) Catalysts

The loading of Fe in the catalyst can currently reach up to 5 to 10 weight percent, which is challenging to increase further due to four pyridinic/pyrrolic N atoms and a minimum of ten C atoms being required to stabilize a single Fe atom. To extensively boost the ORR activity and TOF of Fe-N₄ sites is very important. Substituting a Cu atom next to Fe-N₄ may alter the electronic configuration of core Fe and considerably accelerate the ORR. These successful methods of controlling the configuration and functionality of single-atom catalysts by nearby metal atoms or heteroatoms inspired many researchers. Wei et al. also demonstrated that P-doping the Fe-N-C catalyst enhances the ORR activity along with providing high tolerance against small molecules, e.g., SO_x, NO_x, and PO_x (Figure 12). The synthesis process of the 3d-Fe-PNC catalyst is shown in Figure 12a. The P-doping process significantly improves the catalytic activity because of the synergistic effect of P to N that provide a more stable active region for the ORR. As a result of the above advantages in contrast with 3D-Fe-NC and Pt/C, 3D-Fe-PNC demonstrated superior ORR activity in alkaline/acidic solutions (Figure 12b–e). Furthermore, 3D-Fe-PNC exhibited a potent anti-poisoning activity when exposed to several anions, including SO_x, NO_x, and PO_x [101,102,103]. Sun et al. established a novel single-pot process to obtain P-doped Fe-N-C (Fe-NCP) catalysts, utilizing an Fe precursor, formamide (FA), and red phosphorous as starting substances (Figure 13a,b). When red phosphorous reacts with FA, it forms a chemical link to the condensation product, which additionally acts as a multidentate ligand to efficiently bind Fe atoms. An atomically scattered Fe-N₄ site with a nearby P atom was produced resulting from the pyrolyzing procedure. The Fe-NCP catalyst outperforms the Fe-NC catalyst and Pt/C in HT-PEMFCs at 160 °C (357 mW/cm²; Figure 13c,d). In addition to increasing the ORR activity, the nearby P might effectively raise the charge density of Fe and support the phosphate tolerance of Fe-N₄, which would also prevent Fe-N₄ poisoning [104]. According to the above studies, the P-doping-against-phosphate approach offers a novel catalyst design that could be a potential pathway and makes it possible to effectively promote the active site in an environment of PA-based HT-PEMFCs.

2.4. Multi-Metallic M-N-C Systems

Introducing an additional metal species to create bi-metallic (typically expressed as M1M2-N-C) or multi-metallic active sites is a preferable way in order to modify the electronic configuration and enhance the efficiency of M-N-C catalysts for ORR performance [105,106,107]. Computational calculation studies on combinations of Fe and Co reveal that Fe-N₄ and Co-N₄ moieties possess significant binding strength with O₂, and it estimated that they represent more active catalysts for ORR [108,109,110,111]. Therefore, very few studies had been conducted on M1M2-N-C. Jiang et al. synthesized dual atom (Fe, Ni)–nitrogen-doped carbon (FeNi-N-C) effectively through the simple pyrolyzing of Fe, Ni co-doped ZIF-8 (Figure 14a,b). The obtained FeNi0.25-N-C catalyst was used for ORR performance, and it showed superior activity in the PA electrolyte compared with Pt/C (Figure 14c,d), due to exceptional tolerance of phosphate ions by Fe and Ni active regions. A systematic evaluation shows that Ni improved the ORR activity and longevity of Fe-N-C in two ways. On the one hand, Ni-N and Fe-N moieties work together to improve the ORR, particularly in the high ORR overpotential region. On the other hand, Ni serves as a catalyst to facilitate graphitization and strengthen the stability of the FeNi-N-C catalyst throughout the high-temperature carbonization process. FeNi-NC has been shown to have a wide range of possible uses as an NPM catalyst in HT-PEMFCs [112,113].

The FeSn-N-C catalyst was examined in 0.5M H₃PO₄ acid medium by Buschermohle et al. in contrast to monometallic-N-C (Sn-N-C and Fe-N-C) (Figure 15). The synthesized catalyst has substantially greater nitrogen content (14–20 at. %), primarily consisting of ORR active nitrogen species such as pyridinic-N, pyrrolic-N, and graphitic-N, compared with commercial Fe-N-C (7 at. %). The obtained catalyst exhibits greater ORR activity and AST stability compared with commercial Fe-N-C. An in-depth study of degradation analysis using Grazing Incidence X-Ray Diffraction (GIXRD) and X-ray absorption spectroscopy (XAS) showed that the degradation of the active site in the catalyst was due to diminished pyridinic and pyrrolic bonded nitrogen. Since there was no metallic particle formation and loss of metal, it was suspected that non-metallic nitrogen had been lost. It is suggested that dual-atom catalysts (DACs) show synergistic effects, such as the electronic impact associated with metals [114,115,116,117]. Sn may detrimentally impact the electronic conditions of redox-active Fe regions or participate as a redox-active component to facilitate the process [77], hence creating an opportunity to improve NPC material efficiency for HT-PEMFCs.

2.5. Advanced Carbon Supports

To overcome the electrochemical oxidation (i.e., corrosion) inherent in conventional carbon blacks, research has shifted towards supports with more ordered, graphitic structures [118,119]. Materials such as highly graphitized carbons, carbon nanotubes (CNTs), and graphene offer intrinsically higher resistance to the corrosive conditions of HT-PEMFCs, providing a stable anchor for catalytic active sites.

Graphene, in particular, has been leveraged as a robust support for single-atom catalysts. For instance, Jiang et al. [120] developed an iron single-atom catalyst on a graphene support (FeSA-G) via a modified one-pot pyrolysis method (Figure 16a,b). This catalyst achieved a high single-atom Fe loading of 7.7 wt.% and demonstrated ORR performance comparable to commercial Pt/C (Figure 16c,d), but with the significant advantage of superior tolerance to phosphate ions. When tested in an HT-PEMFC with a PA/PBI/SiO₂ hybrid membrane at 230 °C, the FeSA-G cathode (0.3 mg/cm² loading) delivered a peak power density of 325 mW/cm², outperforming the Pt/C cathode (1.0 mg/cm² loading), which reached 313 mW/cm² (Figure 16e). Furthermore, the FeSA-G catalyst exhibited superior long-term durability (Figure 16f), highlighting its promise as a next-generation catalyst. Following a similar strategy, Cheng et al. [121] utilized highly conductive CNTs as the support material. They successfully synthesized a catalyst with iron single atoms uniformly dispersed on CNTs (FeSA-CNTs) at a loading of 3.5 wt.% (Figure 17a,b). The resulting catalyst showed excellent ORR activity, which was attributed to the combination of highly active atomic sites, inherent phosphate resistance, and the exceptional conductivity of the CNT network. In cell testing at 240 °C, the FeSA-CNT cathode delivered a peak power density of 266 mW/cm² and demonstrated excellent long-term operational stability (Figure 17c,d).

Building on the potential of CNT supports, Wang et al. [122] introduced a novel bimetallic design, creating an atomically distributed Fe-Cu catalyst on N-doped CNTs (FeCu/N-CNTs) (Figure 18a–c). Remarkably, their work revealed a paradigm-shifting finding: contrary to its well-known poisoning effect on PGM catalysts, the phosphate anion actually enhances the ORR performance of the FeCu/N-CNT catalyst. Both experimental and theoretical analyses suggested that the strong binding of phosphate to the Cu atoms electronically modulates the neighboring Fe atoms, thereby promoting their catalytic activity. An HT-PEMFC equipped with this cathode delivered a high power density of 302 mW/cm² at 230 °C and showed significantly better long-term stability than Pt/C (Figure 18d,e). This discovery opens an entirely new avenue for designing highly efficient and durable NPM catalysts that can uniquely leverage the chemical environment of PA-based HT-PEMFCs.

2.6. Metal-Free Heteroatom-Doped Carbon Catalysts

In acidic environments, NPM catalysts are unstable due to several consequences of degrading factors, including anion/proton binding, corrosion, and the generation of H₂O₂ [123,124,125]. The primary undesirable aspect of the ORR process corresponds to radical oxygen molecules that are produced by metal ions, even in trace amounts of peroxide [126], which is speculated to speed up the deterioration of membrane [127]. Additionally, a wide range of different small molecules, including SO_x, CO, and NO_x, have been discovered to poison catalysts [128]. PO_x is an additional poison molecule which originates from PA-PBI (an HT-PEMFC membrane) and could negatively impact ORR activity by phosphate ion adsorption on catalyst surfaces [129].

Regarding metal-free catalysts, only a few studies have been published on tolerance to withstand small molecular toxins, which could be a crucial factor for many potential applications. Based on theoretical research the doping of P and N could be promising for the ORR because it can result in highly confined states near the Fermi level [130,131,132]. When compared with N, P is bigger in atomic size and has lower electronegativity, and doped P alters the surrounding electron and spin density of N-doped C via its 3p lone-pair electrons and facilitates the adsorption of oxygen molecules by allowing its valence electrons to migrate towards its unfilled 3d orbitals [93]. This may be advantageous for establishing high tolerance upon small poisonous substances. This concept was experimentally validated by Wei et al. [133], who synthesized a P,N-doped carbon (PNC) material through the pyrolysis of a self-assembled supramolecular precursor (Figure 19a). The resulting metal-free PNC catalyst exhibited not only excellent ORR performance and stability but also, most importantly, a robust anti-poisoning ability. As demonstrated in poisoning tests (Figure 19b–e), the PNC catalyst maintained stable performance in the presence of SOx, NOx, and POx. In stark contrast, control samples of N-doped carbon (NC), P-doped carbon (PC), and even commercial Pt/C all showed significant performance degradation under the same conditions. The authors attributed the high catalytic activity of PNC to the abundance of graphitic edge-defects created by dual doping.

Table 2. Quantitative performance and durability comparison of leading NPM catalysts in HT-PEMFC single-cell tests.

Catalyst Type	Cathode Loading (mg cm−2)	Anode Catalyst	Membrane	Temp. (°C)	Test Conditions	Peak Power Density (mW cm−2)	Performance Metric	Durability/Degradation	Ref.
Fe/C-700	6.3	Pt/C	PA-doped PBI membrane	120–180	H2/O2, 2.0 bar	60 @ 160 °C	0.44 A cm−2 @ 0.8 V	130 h	[82]
BP-Fe-N-C	7.8	Pt/C	PA-doped PBI membrane	160	H2/O2, 20 mL min−1	700	0.57 A cm−2 @ 0.6 V	34% loss @ 0.6 V after 400 h	[94]
FePhen@MOF-ArNH3	2	Commercial Advent A1100W Pt GDE	PBI membrane	200	H2/O2	137	0.55 A cm−2 @ 0.6 V	-	[89]
Fe-N-C	2	Pt/C	PBI membrane	160	H2/O2, 500/400 sccm	-	0.65 V @ 0.2 A cm−2	250 h	[95]
Fe-N-C	21	Pt/C	PA-doped BASF membrane	150	H2/O2, 200/600 mL min−1	20	20 mA cm−2 @ 0.6 V	-	[97]
FeSA-G	0.3FeSA	Pt/C	SiO2 nanoparticle-doped PA/PBI membrane	230	H2/O2	325	353 mA cm−2 @ 0.5 V	16% loss @ 0.5 V after 100 h	[120]
LEDFe5-NH3	3.8	Pt/C	PA-doped PBI membrane	150	H2/O2, 1.5 bar	260	1260 mA cm−2 @ 0.2 V and 78 mA cm−2 @ 0.7 V	-	[90]
FeCu/N-CNTs	4 (0.12FeCu)	Pt/C	SiO2 nanoparticle-doped PA/PBI membrane	230	H2/O2, 100 mL min−1	302	320 mA cm−2 @ 0.5 V	No loss @ 0.5 V after 100 h	[122]
FeSA/HP	4 (0.13Fe)	Pt/C	SiO2 nanoparticle-doped PA/PBI membrane	240	Anhydrous H2/O2, 150/100 sccm	266	365 mA cm−2 @ 0.5 V	12% loss @ 0.5 V after 100 h	[121]
Fe-NCP	2.5	Pt/C	PA-doped PBI membrane	160	H2/O2	357	-	-	[104]
CoFe-N-SiOCa	3	Pt/C	PA-doped PBI membrane	160	H2/O2, 1.5/9.5	34	-	-	[134]

summarizes the performance of recently reported NPM advanced electrocatalyst in HT-PEMFCs. This comparative summary highlights several critical trends. A wide range of peak power densities have been reported, from 20 mW cm⁻² to an impressive 700 mW cm⁻² for a BP-Fe-N-C catalyst, though this required a very high cathode loading of 7.8 mg cm⁻². This exemplifies the recurring issue that achieving high performance necessitates thick catalyst layers, which can impede mass transport. More recent advancements, however, showcase a promising path forward. Catalysts on advanced supports like graphene or CNTs (e.g., FeSA-G, FeCu/N-CNTs) and those with P-doping (Fe-NCP) demonstrate excellent power densities (266–357 mW cm⁻²) and enhanced stability at high operating temperatures. The bimetallic FeCu/N-CNT system is particularly noteworthy for its outstanding durability and unique synergistic interaction with the phosphate electrolyte. The data also underscore the difficulty in making direct comparisons due to variations in test protocols across studies. Nevertheless, the clear progression towards more durable and sophisticated catalyst architectures is encouraging for future development.

Overall, these results provide strong evidence that well-designed metal-free catalysts can effectively address the critical challenge of poison tolerance in HT-PEMFCs. The performance of the PNC catalyst demonstrates a promising pathway for developing next-generation cathodes that are intrinsically durable in chemically demanding operating environments.

3. Challenges and Future Perspectives

Currently, despite a lot of progress in the development of NPM catalysts for HT-PEMFCs, there are still a number of important challenges to overcome. Further development of this technology mainly depends on overcoming these challenges via distinctive scientific methods and rigorous engineering. This section summarizes the main challenges and suggests potential approaches to further study that might lead to a breakthrough.

3.1. Current Challenges in NPM Catalyst Development for HT-PEMFCs

The research scenario for NPM catalysts in HT-PEMFCs is currently characterized by four main challenges:

(1)

Limitation between Activity and Stability: An ongoing and essential problem in the fabrication of M-N-C catalysts is the contrary interaction between the initial activity and stability. The high-temperature pyrolysis process is necessary to synthesize the highly active and graphitic M-N_x sites, often due to carbon support being more sensitive to electrochemical corrosion [135]. On the other hand, the synthesis conditions provide more durable catalysts, but amorphous carbon support might not provide sufficient active sites [130,136]. One of the main objectives of the field is to overcome this limitation in order to produce a catalyst that is both extremely active and extremely durable.

(2)

Mass Transport and Electrode Engineering: As mentioned earlier, thick catalyst layers are required to obtain sufficient performance due to the lower volumetric activity of NPMs in comparison to Pt [137]. This leads to a series of mass transit-related engineering issues. A high thickness and porous layer will severely hinder the transport of oxygen toward the active sites and the outflow of water vapor, resulting in concentration losses and reduction in the power density of fuel cells, particularly under high load [92]. Additionally, regulating PA ionomer diffusion within this thick coating layer is essential, as too little acid leads to inefficient proton conductivity and poor active sites, while a surplus amount of acid (flooding) blocks the pores and stops the diffusion of gas [137,138]. This is an electrode-level technical challenge and equally significant as the synthesis of catalyst materials [139].

(3)

Understanding and Mitigating Degradation in PA: Fe-N-C catalysts have been admired because of their ability to withstand phosphate poisoning, which involves reversibly blocking active sites, although they are not impervious to the general harshness of the HT-PEMFC environment [89]. The corrosive properties of concentrated PA, high electrode potential, and high temperature (160–200 °C) can all speed up basic degradation processes such as carbon support corrosion and dissolution of the active metal centers [140]. The mechanism of degradation is a complicated, interconnected phenomenon in which the PA electrolyte, the Polybenzimidazole (PBI) membrane, and the catalyst all interact with one another and degrade over time [141]. Developing an extensive understanding of these interconnected system-level failure mechanisms requires designing an extremely long-lasting MEA.

(4)

Scalability and Standardization: There is an enormous gap between laboratory-scale innovations and industrial-scale production. A lot of sophisticated synthesis techniques, which produce high-performance catalysts, such as multi-step templating or Chemical Vapor Deposition (CVD), are complicated and might not be commercially scalable [137]. For commercial transformation, developing more straightforward and reliable synthesis methods is essential. At the same time, the absence of widely accepted, standardized processes for MEA development and long-term testing makes it very challenging to consistently evaluate performance data from various research groups and determine which catalyst candidates show the greatest promise for future development [137].

3.2. Innovative Approaches and Future Research Directions

Addressing these complex issues is necessary to achieve a conceptual change from theoretical, experimental approaches to a more logical, system-level approach. The development of the field lies is at the correlation of advancement in the materials science, chemical engineering, and computational science, where the in-depth analysis of failure mechanisms directly impacts the design of new materials and electrodes.

Rational catalyst design for NPM materials benefits greatly from computational modeling and advanced in situ characterization techniques. Because the synthesis of M–N–C catalysts involves a vast array of variables—such as precursors, compositions, and temperatures—it is impractical to rely solely on extensive experimental screening. Instead, sophisticated computational tools, particularly DFT, can be employed to evaluate the stability of potential M–N–C structures in PA environments and to predict their oxygen binding energies [141]. These theoretical insights help narrow down the most promising material candidates and guide experimental efforts more efficiently, significantly accelerating the discovery of optimal catalysts.

At the same time, understanding the true nature of active sites and degradation mechanisms in NPM catalysts for HT-PEMFCs requires advanced in situ and operando characterization methods to observe catalytic activity under real operating conditions. For example, operando electrochemical mass spectrometry can detect volatile decomposition products in real time, providing insights into catalyst degradation pathways, while synchrotron-based X-ray absorption spectroscopy (XAS) enables detailed investigation of metal center coordination and oxidation states under applied potentials [142,143]. Despite these advancements, the field still lacks standardized post-mortem analysis to confirm degradation routes under realistic HT-PEMFC conditions. Techniques such as X-ray Photoelectron Spectroscopy (XPS) for tracking surface functional groups, XAS for probing the local structure of Fe/Co sites, HR-TEM for identifying morphological changes, Mössbauer spectroscopy for differentiating Fe oxidation states, and ICP-MS for quantifying metal leaching are essential tools for understanding failure modes in NPM catalysts. The insights gained from these techniques clarify the fundamental causes of performance loss and provide guidance for the rational redesign of next-generation catalysts.

• Urgent Need for Standardized Protocols:

We highlight the urgent need for international consensus on testing conditions for HT-PEMFCs, analogous to the standardized protocols developed for LT-PEMFCs by organizations such as DOE, EU Horizon 2020, and IUPAC. Standardization would improve data reproducibility, facilitate cross-laboratory comparisons, and accelerate catalyst development for real-world applications.

• Proposed Framework for Future Benchmarking:

We propose a tentative benchmarking framework for NPM catalysts in HT-PEMFCs, which includes the following:

(i)

Standard operating window: Temperature = 160–180 °C, pressure = 1 atm, and RH = 0–20%, using doped PBI-type membranes.

(ii)

Electrode architecture: Standardized catalyst loading (e.g., 1.0 ± 0.2 mg/cm²), gas diffusion layer, and ink composition.

(iii)

Electrochemical protocols: Report initial activity via polarization curves and ORR onset potential, and assess stability via 100 h chronoamperometry or accelerated stress tests (ASTs).

(iv)

Post-mortem analysis: Encourage the use of XPS, XAS, and HR-TEM to understand degradation mechanisms.

(v)

Data reporting format: Include both gravimetric and areal performance metrics (mA/mg, mA/cm²), membrane resistance, and cell voltage at defined current densities.

4. Conclusions and Remarks

The search for a clean and sustainable energy system has drawn considerable attention to high-performance energy conversion technologies, among which HT-PEMFCs stand out for their distinct benefits in terms of fuel flexibility and system simplicity. However, the use of expensive and rare PGM catalysts has limited their application. The use of PGM catalysts is even worse in the HT-PEMFC environment due to extreme catalyst poisoning with the PA electrolyte. This review has demonstrated the remarkable progress made in the development of NPM catalysts as a significant solution to this challenge.

The key findings of this comprehensive analysis of M-N-C catalysts, especially those made of iron, is that they have become the apparent leading materials to replace the platinum. They are particularly well-suited for HT-PEMFCs because of their remarkable intrinsic activity for the ORR and, more importantly, their basic robustness to phosphate poisoning. The development of catalysts with high densities of atomically distributed active sites within hierarchically porous structures has been made possible by advancements in synthesis techniques, particularly those that use Metal–Organic Frameworks (MOFs) as self-sacrificing templates. This has resulted in encouraging initial performance in single-cell devices.

Despite all of these impressive achievements, we still face severe challenges to commercialization. Under the severe operating circumstances of an HT-PEMFC, the most significant obstacle in the present NPM catalysts is limited long-term durability. It is necessary to carefully address the interrelated failure mechanisms such as poor mass transport within the unavoidably thick catalyst layers, active site degradation by demetallation and radical attack, and carbon support corrosion. In addition, real-world application requires bridging the gap between complex lab-scale synthesis and scalable, economical manufacturing processes.

The future of NPM catalyst development will be driven by a synergistic integration of advanced computational modelling to guide the identification of materials, advanced operando assessment methods to uncover the true mechanisms of degradation, and novel-electrode engineering to get around mass transport constraints. A particularly interesting approach is the development of innovative composite catalysts, which combine the better stability of corrosion-resi stant substrates such as MXenes and transition metal carbides with the high activity of M-N-C catalysts.