Advanced Strategies for Mitigating Catalyst Poisoning in Low and High Temperature Proton Exchange Membrane Fuel Cells: Recent Progress and Perspectives

Abstract

Catalyst poisoning remains a persistent barrier to the efficiency and longevity of electrocatalytic energy conversion devices, namely fuel cells. To address this challenge, this review provides a systematic investigation of recent advancements in mitigation strategies, with particular emphasis on surface engineering, alloying, and combined approaches. Notable developments include the rational design of Pt-alloy catalysts with enhanced CO, H₂S, and H₃PO₄ tolerance as well as the implementation of anti-poisoning molecular architectures and carbon-based protective layers. These methods collectively show considerable promise for improving catalytic activity by fine-tuning electronic structures and minimizing interactions with undesired adsorbates. In addition to presenting a comprehensive overview of the current progress, this review identifies promising future directions, guiding the design and realization of robust, poison-tolerant catalysts crucial for sustainable energy technologies.

1. Introduction

With their exceptional efficiency, high power density, rapid start-up, and lightweight design, hydrogen fuel cells are poised to transform the automotive and portable device sectors, offering a major step forward toward cleaner and more efficient energy solutions. Despite the significant promise of large-scale pure hydrogen production, it remains cost-prohibitive. As a result, approximately 95% of industrial hydrogen continues to be generated through economically viable methods such as natural gas steam reforming and coal gasification. In the former, natural gas reacts with steam to produce hydrogen, carbon monoxide, and trace amounts of carbon dioxide, whereas in the latter, coal is converted into syngas (a mixture of hydrogen and carbon monoxide), which is subsequently processed to extract hydrogen. These processes predominate largely because of their cost-effectiveness and well-established infrastructure.

However, hydrogen streams derived from these methods inevitably contain impurities including CO, H₂S, and NH₃, which can enter a fuel cell’s anode, degrade key materials, and curtail system performance. This contamination constrains the durability and commercial appeal of fuel cell electric vehicles. To further advance fuel cell technology, it is therefore critical to develop cells that tolerate these common impurities. Among them, CO and H₂S pose the greatest challenges. Strong CO adsorption onto transition metals severely compromises catalytic processes in many applications. Specifically, in fuel cell anodes, CO competes with hydrogen for active sites on platinum, hindering the dissociation of H₂ into atomic hydrogen. Similarly, hydrogen sulfide (H₂S) exhibits a strong affinity for precious metals. When H₂S interacts with platinum in an anode catalyst, sulfide formation ensues, diminishing the catalyst’s activity and negatively impacting the overall fuel cell performance.

Low temperature proton exchange membrane fuel cells (LT-PEMFCs), regarded as a valuable environmentally friendly technology for the next generation, have been facing the challenge of the aforementioned poisoning [1]. In LT-PEMFCs, the hydrogen oxidation reaction (HOR) occurs at the anode, while the oxygen reduction reaction (ORR) occurs at the cathode [2,3,4]. In general, the hydrogen from reformed natural gas serves as the fuel source for HOR, and even a small amount of carbon monoxide (CO) and hydrogen sulfide (H₂S) in the parts per million (ppm) range within this gas can lead to catalyst poisoning and a subsequent reduction in fuel cell performance [5,6,7,8,9,10,11]. Considering the present technical constraints in producing pure hydrogen, which is derived from water electrolysis [12], as a fuel source, the utilization of reformed hydrogen is necessity. As a result, the development of catalysts that are tolerant to poisoning has emerged as a top priority [13,14]. Thus, numerous studies are currently underway to address this critical challenge [15,16,17,18,19].

Phosphoric acid fuel cells (PAFCs) or high temperature PEMFCs (HT-PEMFCs) are a class of fuel cells in which a phosphoric acid-impregnated membrane serves as the proton-conducting medium [20]. A notable advantage of HT-PEMFCs is their ability to operate at elevated temperatures exceeding 200 °C, which minimizes the susceptibility to poisoning by impurities in the fuel—a common limitation observed in LT-PEMFCs. However, HT-PEMFCs face a significant challenge due to the adsorption of bulky anionic phosphate species on catalyst surfaces, such as platinum, which are critical for oxygen adsorption. This adsorption can result in catalyst poisoning and performance degradation. Phosphate ions, characterized by their large size and strong binding affinity, have a pronounced detrimental effect on catalytic activity, even when adsorbed in small quantities. This phenomenon represents a critical and persistent limitation in the operation of HT-PEMFCs. In response, considerable research efforts are being directed toward developing strategies to mitigate phosphate-induced poisoning including the modification of catalyst surfaces and the incorporation of additional physical barriers. Such advancements are essential to enhance the durability and efficiency of HT-PEMFCs, thereby addressing one of the key challenges in their practical application.

This issue is not exclusive to FCs; it extends to various devices and chemical processes employing catalysts [21,22,23]. In conclusion, our ultimate goal is to develop catalysts with high catalytic activity that remain resistant to poisoning. Therefore, this paper aims to provide an extensive review of past research and propose a path for future exploration. The method for designing the anti-poisoning catalysts to be discussed was categorized into three main approaches: surface engineering, alloy development, and combination strategies.

While there have been previous reviews covering specific aspects of catalyst poisoning or targeting certain types of fuel cells, few have offered a comprehensive comparative analysis across LT-PEMFC and HT-PEMFC system. In particular, most existing works have focused on either surface engineering or alloy-based strategies alone, leaving a gap in discussions of combined, synergistic approaches. Therefore, the present review is both timely and urgent, as it consolidates diverse anti-poisoning methodologies—from molecular coatings to core–shell alloys—and examines how these can be adapted or optimized to address real-world operational challenges in a broad range of fuel cell technologies. We anticipate that our critical examination of recent advances, along with our proposed future directions, will guide researchers in designing more robust, multi-tolerant electrocatalysts. Ultimately, we hope that this review will serve as a catalyst itself, inspiring further innovations aimed at sustaining high-performance fuel cells in increasingly demanding applications.

2. Overview of State-of-the-Art Poison-Tolerant Electrocatalysts

2.1. Surface Engineering

Surface engineering aims to tailor the catalytic surface to improve the overall performance and tolerance toward common poisoning species [24,25]. Broadly, this strategy involves modifying a catalyst’s outermost layers through doping with heteroatoms, coating with protective shells, or constructing core–shell architectures. By adjusting the surface composition and atomic arrangement, researchers can influence key parameters such as the adsorption energy, charge transfer, and local electronic structure. Common materials used in surface engineering include noble metals (e.g., Pt, Au), transition-metal carbides/nitrides (e.g., TiWC, TiWN), and various organic/inorganic coatings (e.g., carbon shells, molecular layers). These engineered surfaces find particular utility in electrocatalysts for fuel cells, where even trace amounts of contaminants like CO or H₂S can severely degrade activity. Applications span low-temperature-PEMFCs and HT-PEMFCs, all of which demand robust catalysts that remain active over extended operation. In this section, we explore how distinct surface-engineering methods—ranging from nitridation processes to molecular canopy-like structures—can mitigate poisoning effects and enhance catalyst longevity under realistic electrochemical conditions.

Aaron Garg et al. [26] presented the synthesis of early transition-metal nitride nanoparticles (TMN NPs) with a coating of thin noble-metal (NM) layers [27,28,29,30,31]. The production of Pt/TiWN core–shell NPs involves the nitridation process of Pt/TiWC NPs supported on carbon black (CB) under NH₃ flow at 800 °C. A novel high-temperature self-assembly technique was devised to generate noble metal (NM)/transition-metal carbide (TMC) nanoparticles. This method entails carburizing silica-encapsulated transition metal oxide nanoparticles, which are coated with NM salt precursors.

The powder X-ray diffraction (PXRD) pattern showed that the Pt/TiWC NPs initially aligned closely with that of the face-centered cubic (fcc) WC, featuring a calculated lattice parameter of 0.422 nm (Figure 1a). The scarcity of reflections associated with fcc Pt implies a close interaction between Pt and the TiWC surface, preventing the formation of separate particles. To confirm the preservation of the core–shell structure in Pt/TiWN, scanning transmission electron microscopy combined with energy dispersive X-ray spectroscopy (STEM-EDX) was employed. Due to a weak Pt signal, STEM-EDX mapping could only distinctly discern Pt shells that were a few monolayers thick (Figure 1b).

The X-ray photoelectron spectroscopy (XPS) spectra of the core–shell nitrides and carbides, with Pt coverages ranging, exhibited distinct patterns in both the W 4f and Pt 4f binding energies (refer to Figure 2a,b) For all cases, the W 4f energies exhibited higher energy shifts for nitrides compared with carbides with equivalent Pt loading, while the W 4f and Pt 4f peak binding energies decreased with increasing Pt coverage. These alterations, termed core level shifts (CLSs), indicate bonding interactions between the Pt and W atoms, leading to a charge redistribution impacting the core levels. Although the cause of CLSs remains uncertain, they suggest significant changes in Pt’s electronic and chemical properties due to the presence of the transition-metal carbide (TMC) or nitride (TMN) core. CLSs offer insights into potential modifications in Pt reactivity, with larger CLSs in Pt/TiWN indicating weaker adsorbate binding compared with Pt/TiWC. Overall, these findings manifest the influence of core–shell composition on the electronic structure and reactivity of Pt.

The HOR activity results demonstrate the CO tolerance of core–shell catalysts (Figure 2c). In the absence of CO, all materials demonstrated exceptional HOR catalytic activity, requiring overpotentials below 50 mV for mass transfer limitation. Upon introducing 1000 ppm of CO, Pt_comm displayed suppressed HOR activity until the potentials surpassed 0.5 V. In contrast, both core–shell materials readily recovered their activity at potentials as low as 0.1 V, showcasing noteworthy CO tolerance. The observed CO tolerance can be ascribed to a combination of reduced CO binding energy and OH adsorption onto accessible W sites at low potentials. Significantly, consistent with the XPS CLSs data, Pt/TiWN exhibited superior CO tolerance compared with Pt/TiWC, displaying higher initial activity over Pt_comm and Pt/TiWC. This indicates weakened CO binding, facilitating CO oxidation even at 0.025 V, and the catalyst surface remained active without complete deactivation under these conditions.

A comparison of the pre-peak maxima of CO stripping voltammograms for Pt/TiWN and Pt/TiWC (Figure 2d) at 0.274 and 0.309 V, respectively, revealed that Pt/TiWN exhibited weaker binding to CO than Pt/TiWC. Additionally, the disparity in pre-peak area implies that Pt/TiWN has a higher proportion of weakly bound CO compared with Pt/TiWC (44% vs. 23%). These findings provide direct evidence of the distinctive catalytic properties arising from the utilization of a nitride core to modulate the electronic structure of Pt.

To address the issue of surface poisoning in HT-PEMFCs MEA, Kara et al. [32] introduced an electrocatalyst that eliminated noble metals at the cathode electrode interface. They employed non-platinum group metal (non-PGM) compounds with metal-organic framework (MOF) precursors for oxygen reduction [33,34,35]. Preliminary studies demonstrated the improved resistance to anion poisoning in various electrolytes, showcasing the oxygen reduction reaction (ORR) activity of Fe-based non-PGMs. The same group investigated the use of a MOF-supported Fe-based non-PGM (FePhen@MOF-ArNH₃) as a potential electrocatalyst for ORR in HT-PEM fuel cells.

In contrast to FePhen@MOF-ArNH₃ (Figure 3a), the introduction of only 10 mM H₃PO₄ into the electrolyte for Pt/C led to a slight decrease in the limiting current magnitude. This phenomenon suggests a hindrance to O₂ reaching the electrode surface. Furthermore, compared with the profile without H₃PO₄, the half-wave potential (E_1/2), a pseudo-kinetic parameter reflecting ORR activity, shifted cathodically by approximately 150 mV (0.87 vs. 0.72 V).

Tafel plots from the polarization data (Figure 3b) revealed that Pt/C experienced a significant increase in overvoltage in the presence of 10 mM H₃PO₄, indicating susceptibility to phosphate anion adsorption. In contrast, FePhen@MOF-ArNH₃ showed negligible changes in Tafel performance, suggesting resistance to phosphoric acid poisoning. This clearly demonstrates that the active sites of FePhen@MOF-ArNH₃ remain unaffected in the presence of phosphate anions.

To investigate adsorbate coverage on the electrocatalyst surface, the study employed the equation below and monitored the adsorbate coverage on FePhen@MOF-ArNH₃.

$$\Delta \mu =\mu (V,Ar100mMHPO)-\mu (0.3V,Ar)$$

(1)

A comparison of the experimental $\Delta \mu$ spectra (Figure 3c) with the theoretical models (Figure 3e) representative of phosphate anion adsorption on an Fe₆ cluster revealed negligible $\Delta \mu$ amplitude (<0.005 |$\Delta \mu$|) in the presence of H₃PO₄, indicating an absence of phosphate anion adsorption. This observation suggests that Fe particles are subsurface to graphitic layers, preventing direct contact with the electrolyte and hindering phosphate anion adsorption. These results corroborate the earlier claim of the absence of direct Fe−N coordination on the FePhen@MOF-Ar-NH₃ catalyst. While the catalyst exhibited immunity to phosphate adsorption at room temperature, the authors emphasize the necessity for additional investigation under operational conditions. This is particularly crucial due to potential alterations in the electrolyte structure at elevated phosphate concentrations during fuel cell operation.

Tao Wang et al. [36] proposed a new strategy to enhance the resilience of local Pt active sites during H₂ oxidation, providing high tolerance to CO and H₂S [37,38,39,40,41]. Researchers introduced an organic molecular architecture using 2,6-diacetylpridine (DAcPy), a pyridine derivative with two carbonyl groups. DAcPy was reduced under the reaction condition and strongly bound to the Pt surface through tridentate coordination, forming a canopy-like structure. This architecture allowed Pt atoms to remain accessible for small-sized H₂ while effectively blocking relatively larger CO and H₂S molecules.

Comparing the two HOR polarization curves (Figure 4a,b), Figure 4a shows that the bare Pt/C is highly susceptible to H₂S poisoning. As the Na₂S concentration increased, the bare Pt/C was completely poisoned, resulting in the absence of HOR current. In contrast, the addition of DAcPy to the solution could enhance the H₂S tolerance by spontaneous adsorption of DAcPy onto the Pt surface (Figure 4b). The XPS results demonstrate successful blocking of sulfur poisoning by DAcPy (Figure 4c). After the addition of DAcPy, no observable signal of the S element was detected on the Pt surface modified by DAcPy [42,43].

Density functional theory (DFT) calculations were conducted to confirm the adsorption of DAcPy on Pt (Figure 4d,e). In its pristine state, DAcPy tends to adsorb in parallel to the surface through van der Waals forces due to steric hindrance stemming from its acetyl groups, with an adsorption energy of 1.40 eV. It is lower than the adsorption energy of Py (1.61 eV) due to the N atom adsorption. However, under reductive adsorption, the energy dramatically increases to 3.50 eV, which is dramatically higher than that of CO (1.91 eV) and H₂S (1.12 eV). In this reductive state, DAcPy accepts two proton–electron pairs, facilitating the reduction of the carbonyl groups to hydroxyl groups, thus forming Pt–C–OH bonds with Pt. This process is thermodynamically favorable, with a reaction energy of up to 2.10 eV. In this configuration, the reduced DAcPy adsorbs through three coordination sites (two Pt–C bonds and one Pt–N bond) with the pyridine ring in a tilted orientation (Figure 4d), allowing it to strongly compete with H₂S and CO adsorption. Calculated adsorption energies for AcPy and its reduced form were 1.22 and 2.51 eV, respectively, which corresponded to the superior sulfur tolerance of DAcPy over AcPy.

When DAcPy adsorbed onto the Pt surface in a tilted orientation with a pyridine ring (Figure 4e), it formed a wedge-shaped space beneath it, creating a barrier against CO and H₂S during the hydrogen oxidation reaction (HOR). This configuration established a height-limited space with a 2.41 Å gap between Pt and the pyridine ring. Within this space, small-sized H₂ (2.30 Å) could approach isolated Pt atoms beneath the pyridine ring (depicted as light green spheres), while larger CO (2.65 Å) and H₂S (4.47 Å) were impeded by the pyridine ring. This novel design marks the pioneering instance of creating molecular architectures for Pt-based catalysts to resist CO and H₂S by strategically managing steric hindrance at local active sites. Notably, this approach avoids dependence on highly ordered and compact 2D molecular patterns, hinting at its potential versatility in mitigating various poisoning species beyond CO and H₂S in the context of HOR.

Mengmeng Sun et al. [44] employed an ammonia borane (NH₃BH₃, AB) dehydration method to modify PtRu@h-BN/C core–shell electrocatalysts [45]. By utilizing the PtRu-catalyzed polymerization of ammonia borane and subsequent pyrolysis into graphitic h-BN in the presence of NH₃, the researchers successfully enhanced the CO tolerance of PtRu/C nanocatalysts through encapsulation with few-layer h-BN shells.

Based on the TEM and HR-TEM images (Figure 5a,b), well-dispersed nanoparticles ranging in size from 3 to 6 nm were observed. Additionally, a significant portion of these NPs was surrounded by few-layer graphitic shells. Despite treatment at 600 °C in flowing NH₃, the particle sizes of the PtRu@h-BN/C catalysts remained similar to those of the untreated PtRu/C catalyst.

HS-LCIS (Figure 5c) analysis showed that a significant portion of PtRu NPs were fully encapsulated within h-BN covers, forming PtRu@h-BN core–shell structures. This encapsulation was further confirmed through the benzothiazole adsorption test (Figure 5d,e). As expected, benzothiazole molecules exhibited strong adsorption on the surface of the PtRu/C catalyst (Figure 5d). In contrast, due to the large size of the molecules, benzothiazole diffusion through the h-BN shells was hindered, allowing only selective adsorption on the bare PtRu surface (Figure 5e).

In a CO tolerance durability experiment, the PtRu@h-BN/C catalyst demonstrated strong CO-tolerance in the fuel cell applications. As shown in Figure 6a,b, the single cell employing the PtRu/C anode catalyst manifested noticeable CO poisoning in contrast to the cell utilizing the PtRu@h-BN/C anode.

The current densities (Figure 6c,d) showed that PtRu@h-BN/C catalyst exhibited a minimal decrease in current density at 0.9 V, far less than the PtRu/C catalyst. In the voltage range of 0.4 to 0.8 V, the PtRu/C catalyst showed a 60% decrease, while the PtRu@h-BN/C catalyst achieved a reduction below 20%. These findings highlight the superior catalytic performance and enhanced CO tolerance of the PtRu@h-BN/C anode catalyst.

In this section, we explored various surface-engineering strategies for modifying the outer layers of catalysts to control the adsorption characteristics of reactants and poisonous species. By implementing techniques such as core–shell architectures, nitride/carbon coatings, and molecular canopy-like layers [26,32,36,44], researchers can effectively mitigate the detrimental effects of CO, H₂S, and phosphate ions, thus enhancing both the durability and stability of fuel cells.

Nevertheless, surface engineering alone may not be sufficient for all operating conditions. In many cases, it is coupled with alloying (Section 2.2) in order to exploit synergistic effects. The following section will illustrate how combining different metal elements fine-tunes the electronic structure to achieve superior catalytic performance and enhanced tolerance to common poisoning species.

2.2. Alloying

Alloying represents a pivotal technique in heterogeneous catalysis, wherein combining different metals can yield synergistic effects unattainable by single-metal catalysts alone [46,47,48,49,50,51]. By incorporating second or third metal elements (e.g., Fe, Ru, Ni, Co, Ag), scientists can tune parameters such as the d-band center and the adsorption energies of reaction intermediates. This approach often improves catalytic activity, enhances selectivity, and bolsters resistance to poisoning species like CO or phosphate anions. Practical synthesis routes include one-pot solvothermal reactions, galvanic exchange, and high-temperature annealing, each tailored to control alloy composition and nanoparticle morphology. These alloyed catalysts have proven invaluable in fuel cell applications—particularly at the anode for hydrogen oxidation and at the cathode for oxygen reduction—where high durability and poisoning resilience are paramount. In this section, we introduce fundamental alloying concepts and illustrate their importance through recent examples of ternary and core–shell electrocatalysts, highlighting improvements in CO tolerance, long-term stability, and overall electrocatalytic efficiency.

Zhao et al. [52] modified a ternary electrocatalyst with surface Au-decorated PtFe nanocrystals (NCs) [53]. The resulting PtFeAu catalyst exhibited enhanced resistance to CO poisoning compared with bare Pt and binary PtFe. Surface engineering likely creates its distinctive electronic structure.

A ternary electrocatalyst comprising surface Au-decorated PtFe nanocrystals (NCs) (Figure 7a) was synthesized using a one-pot solvothermal method to form binary PtFe NCs, followed by a cation redox reaction in the presence of Au³⁺. The structural analysis was evaluated by HR-TEM and XRD (Figure 7b–d), confirming successful Au alloying. The TEM and HR-TEM images (Figure 7b,c) demonstrated the mono-dispersity of the binary PtFe NCs with an average particle size of around 3.2 nm. Subsequent galvanic reaction between Fe and Au³⁺ preserved the size and morphology of the resulting PtFeAu NCs, where the discontinuity in the PtFeAu lattice indicated the introduction of Au atoms at the surface.

In the XPS and CO stripping experiments, a significant negative shift was observed. The XPS (Figure 7e) findings confirmed the transfer of electron density from Pt to Au/Fe, resulting in a reduction in the Fermi level and the repositioning of the d-band center. Consequently, the binding strength of CO on the Pt site was weakened. CO stripping experiments (Figure 7f) revealed that CO stripping on the PtFeAu surface was the most facile reaction, which noted the weakened binding of CO on the Pt centers.

Chronoamperometric (CA) measurements and accelerated durability tests (Figure 7g,h) showed that ternary PtFeAu electrocatalysts outperformed PtFe NCs and Pt/C catalysts in current density and ECSA.

Lee et al. [54] reported research on fabricating a PtRuNi/C ternary electrocatalyst with a composition gradient shell using a combination of high-temperature heat treatment and the polydopamine (PDA) protective coating method [55,56,57]. Their work highlighted the effectiveness of protective coating layers in creating core–shell structures with a Pt-rich shell for optimal ORR catalysis. The protective coating method was found to prevent particle sintering during high-temperature treatment (600–800 °C).

The MEA performance showed (Figure 8a) that the development of PtRuNi/C ternary electrocatalysts is crucial for achieving high CO tolerance. The addition of Ni to the PtRu/C binary electrocatalyst resulted in enhanced CO tolerance in the ternary catalysts. The higher heat treatment temperature correlated with an increase in CO tolerance levels; however, the catalyst’s CO tolerance began to diminish at 700 °C, where Ru segregation was initiated. Notably, the catalyst treated at 800 °C exhibited the lowest CO resistance. Thus, it is imperative to subject the particles to heat treatment at a temperature that promotes a high alloying degree without inducing Ru segregation.

Table 1. Comparison of the CO tolerance of different electrocatalysts in a fuel cell. Reproduced with permission. Copyright 2021, Elsevier [54].

Anode Catalyst	Metal Loading (mg cm−2)	Membrane	Operating Temperature (°C)	CO Gas (ppm)	Decreasing Ratio at 0.6 V (%)
PtSn/C	0.3	N112	80	10	30
PtPd/C	0.3	N212	80	10	50
PtMo/C	0.5	N212	70	25	27
PtRu/C	0.2	N117	80	10	27
PtRu/C	0.4	N212	75	10	30
PtNi/C	1.0	N212	75	30	57
PtRuNi/C	0.4	N212	75	10	11

summarizes the CO tolerance of various electrocatalysts from the literature, comparing them with the Pt₁Ru₁Ni_0.75/C-PDA-650 catalyst in a fuel cell. The Pt₁Ru₁Ni_0.75/C-PDA-650 catalyst demonstrated significant CO resistance, showing only an 11% performance reduction at 0.6 V in the presence of CO, outperforming other catalysts. This improved CO tolerance was attributed to the enhanced ligand effect.

The XPS spectra of Pt 4f in Pt₁Ru₁Ni_0.75/C-PDA-650 and Pt₁Ru₁/C (Figure 8b) exhibited a negative shift in binding energies. This shift implies an electron transfer from Ni and Ru to Pt, leading to a reduction in the Pt-CO binding energy. As a result, the Pt₁Ru₁Ni_0.75/C-PDA-650 catalyst showed enhanced CO tolerance.

Jingyu et al. [58] presented an innovative surface-engineering technique involving the application of PtBi alloy shells to coat Pt/Pt-alloys [59]. Despite the PtBi alloys’ downshifted d-band center, the introduction of Bi induced interlayer tensile strain, adversely affecting ORR electrocatalysis. However, the incorporation of large Bi atoms in the Pt overlayers resulted in in-plane shearing, causing surface compression and shortened Pt@Pt bonds (Figure 9a). Through annealing, solid-solution alloys were transformed into ordered intermetallic Pt-alloy NPs, and both materials were coated with PtBi alloy shells using a leach-embed-rearrangement process. The results inferred that the addition of PtBi led to improved activity as well as stability for all materials. Particularly notable was the exceptional ORR performance of FePt@PtBi, demonstrating a high resistance to methanol and CO poisoning.

The XRD patterns (Figure 9b) of FePt@PtBi and FePt revealed characteristic peaks for tetragonal FePt, intermetallic phases, and ordered L1₀-FePt, with similar (110)/(111) intensity ratios indicating highly ordered crystal structures. High angle annular dark field imaging-STEM (HAADF-STEM) images (Figure 9c) of FePt@PtBi depicted a 2–3 layer-thick PtBi shell covering the ordered intermetallic core with (110) planes of L1₀-FePt, and the FFT patterns confirmed the core’s face-centered tetragonal (fct) structure.

The same research group demonstrated the resistance of FePt@PtBi to poisoning in the presence of CH₃OH or CO. When exposed to CH₃OH (Figure 9d), FePt@PtBi maintained a consistent current response, in contrast with the commercial Pt/C electrocatalyst, which suffered a 70% reduction in current. Similarly, the introduction of CO (Figure 9e) had minimal impact on FePt@PtBi, while the commercial Pt/C electrocatalyst experienced an approximate 50% decline in the initial current.

Jang et al. [60] synthesized carbon-supported PtNi alloying nanoparticles (NPs) with varying thicknesses of carbon shells by controlling the concentration of oleylamine [61,62,63,64,65]. The oleylamine carbonization process was revealed through TEM images (Figure 10a), employing a high-temperature heat treatment in a H₂-reductive atmosphere. This surface coating induced a geometric barrier effect in the catalyst. As the concentration of oleylamine increased, the thickness of the carbon shell also increased (Figure 10b). The higher standard deviation in thicker carbon shells indicates an uneven surface state in the metal nanoparticles (NPs), negatively impacting the electrochemical reaction by causing the loss of certain active sites.

Half-cell test results confirmed that the carbon shells acted as geometric barriers, preventing phosphate anion adsorption (Figure 11a). For PtNi@Cx, the carbon shells acted as spacers, restraining the adhesion of phosphate ions to the Pt surface due to the distance between the two materials. The oxygen reduction reaction (ORR) test results illustrated this phenomenon (Figure 11b,c), presenting scenarios with and without a carbon shell, respectively. Catalysts exhibited a similar reduction in ORR activity when exposed to oleylamine, dependent on the presence of phosphoric acid. In contrast, PtNi nanoparticles with carbon shells consistently recovered their ORR activities, irrespective of the thickness of the carbon shells. This implies the existence of a geometric barrier effect caused by the carbon shells. PtNi@C₂’s superior catalytic performance in high-temperature proton exchange membrane fuel cells resulted from the synergistic effects of its geometry and electronics, enabled by a simple synthesis method.

In high-temperature PEMFCs, sustaining the molecular sieve effect is challenging due to the detachment of organic molecules from the metal nanoparticles. Sourabh et al. [66] proposed a solution involving the creation of a durable carbon molecular sieve layer (MSL) on Pt-based nanoparticles, thus maintaining the particle size and preventing metal dissolution [67]. This MSL enhances the ORR activity in phosphoric acid by reducing phosphate anion adsorption and facilitating selective O₂ permeation, addressing challenges in long-term HT-PEMFC operation.

The HR-TEM images (Figure 12) illustrate the formation of ultrathin carbon shells on the Pt nanoparticles in the Pt@MSL catalysts through annealing in various gas atmospheres. Notably, annealing with 20% H₂/N₂ gas induced carbon etching, removing most carbon shells. The precisely controlled carbon MSLs are anticipated to evoke distinct phosphate poisoning effects in electrochemical tests.

A comparison of the ORR performance between the commercial Pt/C catalyst and Pt@MSL7 (Figure 13a–d) was conducted to observe the molecular sieve effect by carbon MSLs. Due to decreased active sites from phosphate anion absorption, the commercial Pt/C catalyst portrayed reduced ORR activity in the H₃PO₄-mixed HClO₄ solution. Conversely, the Pt@MSL catalysts demonstrated interesting changes in their electrochemical properties. The ORR activity in both solutions enhanced with increasing carbon shell porosity from Pt@MSL7-0 to Pt@MSL7-20. The difference in ORR performance between electrolyte solutions increased with larger carbon shell pores, suggesting a pronounced molecular sieve effect. Smaller differences in ORR activity correlated with higher molecular sieving effects.

The Pt@MSL7-0 catalyst, with minimal mass activity for ORR in the H₃PO₄-mixed HClO₄ solution, exhibited a notable molecular sieving effect (~90%). This suggests the effective prevention of catalyst poisoning by anions through the high-density carbon shell encapsulating the Pt nanoparticles. In contrast, the Pt@MSL7-10 catalyst displayed heightened mass activity and an appropriate molecular sieving effect value (44%) in the H₃PO₄-mixed solution, indicating selective O₂ gas permeation through small carbon pores (defects) with the effective prevention of phosphate anion adsorption. Considered optimized among the prepared Pt@MSL catalysts, Pt@MSL7-10 demonstrated improved ORR activity, slightly lower than Pt@MSL7-20.

Jeong-Hoon et al. [68] investigated Pt-based phosphide catalysts for high-temperature PEMFCs (HT-PEMFCs) [69,70]. The catalysts were prepared using trioctylphosphine (TOP) as the phosphorus source, and the synthesis involved Ketjen black (KB) carbon. Unlike conventional methods using phosphine gas or tristrimethylphosphine, this study employed TOP as both a phosphorus source and a coordinating solvent. The PtP₂ phase synthesis followed a unique conversion-based mechanism, setting it apart from previous strategies.

The synthesis process involved the initial conversion of Pt salt into Pt atoms, followed by their reaction with dissociated P atoms to yield the PtP₂ phase (Figure 14a). To obtain a pure PtP₂ phase, the P concentration was adjusted while keeping the Pt concentration constant. The procedure included vacuum treatment, heating at 390 °C for 10 h, washing with an ethanol and toluene (50:50) solution to remove unreacted TOP, and drying for enhanced crystallization, resulting in the catalyst PtP₂/C-ASP.

TEM revealed well-dispersed PtP₂ particles with a 2–3 nm diameter on the carbon surface of PtP₂/C-ASP (Figure 14b), lacking long-range ordering as observed in the HR-TEM analysis. No diffraction spots were observed in the fast Fourier transform (FFT) pattern (Figure 14c,d). The amorphous PtP₂ particle in PtP₂/C-ASP, revealed by HAADF and energy-dispersive X-ray (EDX) imaging (Figure 14e), contained both Pt and P atoms. Upon heat treatment, a slight increase in the PtP₂ particle size to approximately 3.9 nm was observed in PtP₂/C-800 (Figure 14f), accompanied by the development of crystalline ordering (Figure 14g,h). The inverse FFT patterns (Figure 14i,j) indicated polycrystalline ordering in PtP₂/C-800, which was supported by EDX images confirming the presence of the PtP₂ crystalline phase (Figure 14k).

The authors employed DFT calculations to elucidate the resistance of PtP₂ catalysts to phosphate poisoning and their exceptional activity. The results demonstrated stronger phosphate adsorption on PtP₂ compared with Pt (Figure 15a), with P atoms attracting HPO₄⁻ anions, preserving Pt sites. Contrary to common assumptions, materials with stronger phosphate binding, such as PtP_2, proved advantageous. The Pourbaix diagram of the surface (Figure 15b) revealed the formation of a protective layer of phosphoric oxide-based oligomers on PtP₂ at oxidative potentials, ensuring the maintenance of active Pt sites during HPO₄⁻ adsorption.

Here, we discuss how alloying different metal components—such as Fe, Ru, Ni, Co, and others—can improve the catalytic properties beyond those of single-metal systems. By tuning the d-band center and modulating the adsorption energies of key intermediates, alloy catalysts exhibit increased activity, stability, and resistance to poisoning agents like CO and phosphates [52,54,58,60,66,68].

However, surface engineering and alloying each addresses different facets of catalyst design. While alloying manipulates the core and near-surface electronic structure, surface-engineering techniques provide targeted modifications at the atomic or molecular level. In the next section, we highlight how these two approaches can be combined to achieve even greater enhancements in catalyst performance under demanding fuel cell conditions.

2.3. Combined Approach

A growing trend in catalyst design involves merging both surface engineering and alloying approaches to optimize performance [71,72,73,74]. In these dual strategies, alloying adjusts the bulk and near-surface electronic structure, while targeted surface modifications—such as shell growth, protective coatings, or molecular-level design—fine-tune the active site accessibility and selectivity. By simultaneously exploiting the “bulk” electronic effects of alloys and the “local” control from surface coatings, researchers have achieved heightened activity and robust poison resistance, even under severe operating conditions found in high-temperature or direct methanol fuel cells. This combined methodology also facilitates the development of catalysts tailored to specific operational needs, such as selective blocking of large contaminant molecules, while still permitting smaller reactants to reach active sites. In this section, we review recent advances illustrating how the interplay between alloy composition and surface functionalization can produce cutting-edge electrocatalysts with outstanding CO, H₂S, and phosphate-tolerance, ultimately underscoring the significance of integrated design principles in next-generation fuel cell systems.

A remarkable investigation was reported by Daisuke et al. [75], who proposed replacing RuO_2.1 nanosheets with Ru nanosheets to enhance the CO tolerance of PtRu/C [76]. Through metallization, the Ru nanosheets established close contact with the surface of the PtRu nanoparticles while preserving the high surface area of Pt, aiming to boost the catalyst’s CO tolerance.

Figure 16a–c shows the mean size of the PtRu nanoparticles in Ru(ns)-PtRu/C (4.3 ± 0.4 nm) was similar to that of the original PtRu/C (4.2 ± 0.3 nm). This indicates that the metallization process does not lead to an increase in the size of the PtRu nanoparticles. Moreover, Ru(ns)-PtRu/C showed an identical fcc(220) reflex and width (d-spacing of 0.1369 nm) compared with PtRu/C and RuO_2.1ns-PtRu/C (Figure 16d). This similarity suggests that the alloying state of PtRu nanoparticles was consistent among all catalysts.

Comparing the hydrodynamic voltammograms of the HOR contingent on the CO (Figure 17a,b), it is evident that the behavior of HOR in pure H₂ was comparable among PtRu/C, RuO_2.1ns-PtRu/C, and Ru(ns)-PtRu/C. This observation underscores that the introduction of metallic Ru nanosheets does not impede the activity of HOR. Conversely, when HOR was conducted in the presence of CO, the HOR current for all catalysts was lower than that observed in pure H₂ due to CO poisoning. Upon closer examination, RuO_2.1ns-PtRu/C exhibited a higher HOR current than PtRu/C, and Ru(ns)-PtRu/C surpassed RuO_2.1ns-PtRu/C in HOR current. These findings suggest that integrating metallic Ru nanosheets effectively enhanced the catalyst’s CO tolerance.

Similarly, the chronoamperogram results in pure H₂ (Figure 17c) indicate that RuO_2.1ns PtRu/C and Ru(ns)-PtRu/C displayed comparable or slightly higher HOR activity than PtRu/C. As anticipated, a sharp initial decrease in HOR activity for all catalysts was observed in 300 ppm CO/H₂ due to CO accumulation (Figure 17d), followed by a quasi-steady state. At 5 h, Ru(ns)-PtRu/C exhibited 1.3 times higher HOR activity than PtRu/C. The normalized current decay ratio (Figure 17e) illustrates that Ru(ns)-PtRu/C mitigated CO poisoning in the initial 60 min. The steady-state current at 3 h for Ru(ns)-PtRu/C was in contrast to PtRu/C’s continuous decrease even after 5 h, suggesting that the improved CO tolerance due to metallic Ru nanosheet modification was most likely from the suppression of CO adsorption on the Pt surface.

Furthermore, the HOR activity after ADT (

Table 2. HOR activity measured before and after the durability test of catalysts from chronoamperometry (at 20 mV vs. RHE). Reproduced with permission. Copyright 2016, IOP [75].

HOR Activity/A (g-PtRu)−1
Catalyst	Initial	After 1000 Cycles	After 3000 Cycles
PtRu/C	89	64	55
RuO2.1ns-PtRu/C	111	86	75
Ru(ns)-PtRu/C	124	112	97

) demonstrated superior retention in HOR activity for Ru(ns)-PtRu/C after both 1000 and 3000 cycles. In conclusion, the superior activity and stability of HOR in the presence of CO are due to the Ru nanosheets effectively suppressing the adsorption of CO on the catalyst surface.

Yezhou Hu et al. [77] introduced a one-step pyrolysis method for synthesizing Pt-Fe intermetallic compounds with in situ N-doped carbon encapsulation [78,79]. The resulting O-Pt-Fe@NC/C demonstrated improved ORR activity and long-term stability in 0.1 M HClO₄, along with enhanced tolerance to CO, SO_x, and PO_x compared with Pt/C, attributed to the protective carbon layer. O-Pt-Fe@NC/C, as a cathode catalyst in HT-PEMFCs, exhibited significantly enhanced maximum power density.

O-Pt-Fe@NC/C was synthesized via one-step pyrolysis (Figure 18), involving the reduction of metal salts and the transformation of dicyandiamide into graphitic carbon nitride (g-C₃N₄) at temperatures below 550 °C. As the temperature increased to 800 °C, the decomposition of g-C₃N₄ occurred, forming an ultrathin N-doped carbon layer catalyzed by Pt-Fe nanoparticles and coating their surface, while simultaneously promoting the formation of the ordered Pt-Fe intermetallic phase.

Comparing the XRD patterns (Figure 19a) of Pt/C, the disordered Pt-Fe alloy (D-Pt-Fe/C), and O-Pt-Fe@NC/C, both O-Pt-Fe@NC/C and D-Pt-Fe/C exhibited reflex shifts to higher angles, signifying lattice contraction from the introduced Fe. The TEM images (Figure 19b) revealed well-dispersed nanoparticles (average size ~3.8 nm) on the carbon support, demonstrating effective size control by the carbon layer, which resisted Ostwald ripening or coalescence during high-temperature treatment. The HR-TEM images (Figure 19c) depicted Pt-Fe nanoparticles encapsulated in a ~0.7 nm thick carbon shell, suggesting a strong coupling effect. Pores in the carbon layer (Figure 19d) resulted from N doping and CN_x gases released during the C₃N₄ heat treatment. The EDS elemental mapping (Figure 19e–j) confirmed the presence of an N-doped carbon shell adjacent to nanoparticles. XPS validated the existence of N in the carbon shell, indicating an interface confinement effect between the N-doped carbon and Pt-Fe ordered intermetallic nanoparticles.

In the CO stripping tests (Figure 20a,b), O-Pt-Fe@NC/C exhibited superior anti-poisoning capability with the lowest CO oxidation peak potential and minimal half-wave decay in the presence of SO₃²⁻ compared with the others. The authors attributed this effect to nitrogen introduction. Nitrogen’s higher electronegativity changes the local charge on carbon, which improves the charge transfer and reduces the adsorption of poisonous molecules.

Additionally, the carbon layer acts as a physical protective barrier, effectively isolating nanoparticles from contact with poisonous molecules. Furthermore, O-Pt-Fe@NC/C was systematically compared with previously reported findings (Figure 20c). The results highlighted O-Pt-Fe@NC/C as a promising material for HT-PEMFCs operating at elevated temperatures such as 160 °C.

Douglas et al. [80] improved CO tolerance in low-temperature PEM hydrogen oxidation by utilizing a Pt/C and WO₃ powder mixture [81,82,83]. The authors elucidated the CO tolerance enhancement mechanism in the Pt/WO₃ system, focusing on the role of W^V/W^VI in the bifunctional mechanism and WO₃’s influence on CO coverage and binding modes to Pt. This investigation provides insights into potential electronic effects.

The TEM images illustrated the sizes of the synthesized WO₃ nano-platelets, Pt particles, and mixed catalyst electrodes (Figure 21a–c). In the mixed state, there was an interdispersion of the Pt/C and WO₃ components, with a stacking of WO₃ nanoplatelets. The XPS analysis (Figure 21d) revealed a ~0.2 eV shift in the Pt^4f electron peak position, indicating a potential two-phase boundary or equilibration of the electronic energy levels in the Pt/C/WO₃ catalyst mixture. The result addressed the possibility of W soluble species depositing on the platinum surface during the measurements, considering the stability of WO₃ electrodes toward dissolution under operation conditions.

The authors proposed a mechanism of catalytic cycle for CO oxidation by WO₃ (Figure 21e), which involves W^VI as the active redox species oxidizing CO_ads,Pt in the presence of water (reaction a) to form CO₂ and tungsten bronze. The rapid electrochemical oxidation of tungsten bronze (reaction b) occurs at potentials exceeding ~0.3 VRHE, completing the catalytic cycle. Despite the thermodynamic feasibility of the electrochemical oxidation of WV above ~0.2 VRHE, the slow electron transfer kinetics between 0.2 and 0.3 VRHE make reaction “b” the rate-determining step in this potential range.

To investigate the oxidation state of tungsten oxide and its potential influence on enhanced CO tolerance in Pt/C/WO₃ catalysts, in situ Raman spectroscopy was conducted (Figure 22a,b). The Pt/C/WO₃ electrode exhibited peaks at 271, 714, and 807 cm⁻¹ at open-circuit potential, corresponding to δ(O−W−O) bending and W−O stretching modes of WO₃. Reducing potential led to peak disappearance, indicating the formation of hydrogenated tungsten bronze. At 0.1–0.25 VRHE, a peak at 820 cm⁻¹ suggested partially reduced tungsten centers, while potentials above 0.3 VRHE revealed fully oxidized WO₃ (W^VI). The upward shift in peak position at lower potentials signified a stronger W-O bond in the hydrogenated tungsten bronze than in WO₃, attributed to the increased covalency of the WV-O compared with W^VI-O bonds (Figure 22b). This effect may enhance CO oxidation at positive potentials. The reduced peak area in more hydrogenated tungsten bronze was attributed to increased coloration, reducing scattered light as well as lowering the optical skin depth in more conductive tungsten bronze compared with semiconducting WO₃. Although the comparison of peak areas is qualitative, it strongly suggests the presence of distinct phases at different potentials. Additionally, Pt is known to induce oxidation state changes in WO₃ through H₂ spillover at room temperature, a phenomenon directly applicable to the Pt/C/WO₃ electrochemical system in this study.

In situ ATR-IR spectroscopy was employed to investigate the CO oxidation behavior of Pt/C catalysts with and without WO₃. Spectra (Figure 22c) revealed characteristic Pt-CO features with bands at 2000–2100 cm⁻¹ and 1820–1920 cm⁻¹, corresponding to linear (COL) and bridge-bonded (COB) CO configurations on Pt. At an initial potential of 0.04 VRHE, the broad and asymmetric COL band suggested multiple CO adsorption environments, similar to the carbon-supported Pt catalysts. Incremental potential steps from 0.04 to 1.09 VRHE resulted in shifts and intensity reductions in both the COL and COB bands, indicating CO removal from the surface, with a surface structure sensitivity observed in the Pt-CO oxidation process. While the Pt/C/WO₃ catalyst displayed a similar behavior to Pt/C, differences were observed in the shape and position of the linear-bonded CO band (Figure 22d). Pt/C/WO₃ showed a primary peak at a lower wavenumber (2009 cm⁻¹) with shoulders at 2027 and 2053 cm⁻¹, indicating a unique Pt−CO bonding environment induced by WO₃. The lower wavenumbers suggest an electronic/polarization effect or reduced dipole−dipole coupling, implying lower coverage or higher mobility. The intensity decreased with increasing potential, and preferential removal of the lower wavenumber feature mirrored the Pt/C behavior.

Therefore, this study emphasizes WO₃’s effectiveness in improving CO tolerance in Pt/C catalysts, particularly in the potential range 0.3 ≥ E ≥ 0.4 V_RHE. This improvement makes them suitable for applications with an anodic overpotential, such as in direct methanol fuel cells, by revealing WO₃‘s electronic effect on CO binding to Pt.

In this section, we presented examples of merging surface engineering and alloying methods to create advanced electrocatalysts—such as alloy core–shell structures with protective coatings or molecular modifications [75,77,80]. These combined strategies allow for the precise tuning of both bulk electronic effects and local surface chemistry, resulting in outstanding stability and poison resistance in severe operating environments.

As we move forward to the conclusion, we will integrate the key insights from Section 2.1, Section 2.2 and Section 2.3, discussing how the interplay between surface modifications and alloy design drives next-generation catalysts with excellent durability, activity, and tolerance in diverse fuel cell systems.

3. Converging Anti-Poisoning Strategies

Table 3. A comparison of previously proposed mitigation strategies categorized by composition, method, poisoning molecules, mechanism, and performance.

#	Catalyst Composition	Synthesis/Key Method	Target Poison	Key Mechanism for Poison Mitigation	1. Performance Gains /2. Highlights	Reference
1	Pt/TiWN	Nitridation of Pt/TiWC (800 °C in NH3)	CO	Core–shell synergy (Pt shell on transition-metal nitride), lowered CO binding energy	1. Rapid HOR recovery at ~0.1 V in presence of CO 2. Higher CO tolerance than commercial Pt (Ptcomm)	Aaron Garg et al. [26]
2	FePhen@MOF-ArNH3 (Fe-based N-PGM)	MOF-based pyrolysis (Fe-Phen@MOF) with ammonia treatment	Phosphate (H3PO4)	Subsurface Fe encapsulation; minimal phosphate adsorption on Fe sites	1. Negligible Tafel slope changes under phosphate 2. Enhanced ORR stability and reduced anion poisoning	Kara et al. [32]
3	Pt/C + DAcPy	In situ adsorption/reduction of 2,6-diacetylpyridine (DAcPy)	CO, H2S	“Molecular canopy” (tridentate binding); size-selective blocking of CO/H2S while allowing H2 access	1. Nearly complete prevention of sulfur poisoning 2. No detectable S on Pt surface by XPS	Tao Wang et al. [36]
4	PtRu@h-BN/C	Ammonia borane (AB) polymerization + pyrolysis; formation of few-layer h-BN shell	CO	h-BN encapsulation partially blocks CO from PtRu surface	1. Significantly improved CO tolerance in single-cell tests 2. Only ~20% drop vs. ~60% drop with PtRu/C	Mengmeng Sun et al. [44]
5	PtFeAu	One-pot solvothermal synthesis + partial galvanic replacement	CO	Au decoration shifts electron density away from Pt: => weaker CO binding	1. Largest negative shift in CO stripping 2. High ECSA and stable chronoamperometry	Zhao et al. [52]
6	Pt1Ru1Ni0.75/C-PDA-650	High-temp (600–800 °C) heat treatment + polydopamine coating to form gradient shells	CO	Strong ligand effect (Ni and Ru → Pt): downshift of Pt-CO binding energy	1. Only 11% performance loss at 0.6 V under CO 2. Superior CO tolerance compared with PtRu/C	Lee et al. [54]
7	FePt@PtBi	Leach-embed-rearrangement to create PtBi shells on ordered FePt intermetallic core	CO, CH3OH	Compressive strain + larger Bi atoms: d-band shift, weakened adsorbate binding	1. ~70% higher current retention vs. commercial Pt/C 2. High methanol and CO tolerance	Jingyu et al. [58]
8	PtNi@Cx	Oleylamine carbonization + H2 reduction forming carbon shells of variable thickness	Phosphate (H3PO4)	Geometric barrier: carbon shell prevents phosphate adsorption on PtNi surface	1. ORR activity quickly recovers even in phosphoric acid 2. Maintains active sites for prolonged operation	Jang et al. [60]
9	Pt@MSL (Pt with carbon molecular sieve layer)	Annealing in different gas atmospheres to form controllable carbon shells	Phosphate (H3PO4)	Selective O2 permeation via dense carbon shell; blocks anion diffusion	1. Increased ORR activity with high molecular-sieving effect 2. Up to ~90% poison-blocking efficiency	Sourabh et al. [66]
10	PtP2	Trioctylphosphine (TOP) route; Pt salt converted to PtP2 phase (amorphous or crystalline)	Phosphate (H3PO4)	Strong P−HPO4− interactions preserve Pt sites: stable at higher potentials	1. Maintains ORR activity under phosphate poisoning 2. Protective phosphoric oxide oligomers on PtP2 surface	Jeong-Hoon et al. [68]
11	Ru(ns)-PtRu/C	Metallization of RuO2.1 nanosheets to form metallic Ru nanosheets on PtRu/C	CO	Metallic Ru nanosheets in close contact with Pt: suppressed CO adsorption	1. 1.3× higher HOR current after 5 h vs. PtRu/C 2. Superior CO tolerance and durability	Daisuke et al. [75]
12	O-Pt-Fe@NC/C	One-step pyrolysis: metal salts + dicyandiamide => N-doped carbon shell + ordered PtFe alloy	CO, SOx, POx	N doping + thin carbon encapsulation: electron modulation + physical blocking	1. Excellent multi-poison tolerance (CO, SO32−, PO43−) 2. Enhanced performance as HT-PEMFCs cathode	Yezhou Hu et al. [77]
13	Pt/C + WO3	Simple physical mixing of Pt/C with WO3 nanoplatelets	CO	W5+/W6+ redox cycle in WO3: catalytic CO oxidation (tungsten bronze)	1. Improved CO removal at ~0.3 VRHE 2. Potential for direct methanol fuel cell applications	Douglas et al. [80]

highlights that, despite the diversity in catalyst compositions, synthesis strategies, and target poisoning species, all of the studies featured shared the same overarching goal: mitigating catalyst deactivation due to adsorbed contaminants. By comparing parameters such as catalyst materials, surface engineering methods, alloying compositions, and resulting performance metrics, we can see how seemingly unrelated research has ultimately converged on controlling the surface adsorption energies and modulating local electronic structures. This unified perspective underscores that every approach—ranging from molecular coatings on Pt to carbide or nitride core–shell architectures—seeks to preserve active sites from poisoning by CO, H₂S, phosphate anions, or other foulants. In doing so, each study contributes a complementary piece to the larger puzzle of designing robust, long-lived catalysts for fuel cell technologies and related electrochemical applications.

4. Conclusions and Perspectives

Catalyst poisoning represents a critical challenge in electrochemical reactions, significantly impeding catalytic activity and accelerating the development of stable electrocatalysts for PEMFCs. Comprehensive investigations in this review identified three primary strategies for mitigating catalyst poisoning:

• Surface engineering: This approach involves direct surface modifications and the implementation of protective barriers to manipulate catalyst geometric properties and enhance poisoning resistance.
• Alloying: A strategic method that modifies the electronic structure of catalytic materials to improve their tolerance to poisoning agents.
• Combined approaches: Generation of synergistic effects by combining multiple mitigation strategies to optimize catalyst performance.

Among these approaches, integrated methodologies have demonstrated the most promising results in developing high-performance, poison-resistant catalysts. The effectiveness of combined strategies highlights the potential for advanced catalyst design that can overcome chronic limitations in electrochemical systems. The current research progress necessitates a critical transition from laboratory-scale investigations to various evaluations under realistic operational conditions. Critical considerations for future development include:

• Rigorous and standardized characterization protocols of catalyst performance under industrially relevant environments;
• Development of advanced in situ monitoring techniques;
• Comprehensive understanding of long-term catalyst stability and degradation mechanisms.

Sustained research efforts and interdisciplinary collaborations are essential to address the complex challenges of catalyst poisoning. Future investigations should focus on the development of robust methodologies that can reliably mitigate the poisoning effects and achieve the severe requirements of large-scale commercial applications in PEMFCs. These advancements will pave the way for the reliable and efficient large-scale commercialization of PEMFCs in automotive and stationary power applications.

Although the present review primarily focused on LT-PEMFCs and HT-PEMFCs, the anti-poisoning strategies outlined here including surface engineering, alloying, and combined approaches can be extended to a broad range of electrochemical energy conversion and storage devices. For instance, catalytic materials used in CO₂ reduction systems, electrolyzers, and metal–air batteries can face analogous poisoning challenges arising from contaminants, by-products, or strongly adsorbing intermediates [12,84,85]. The design principles highlighted in this review, particularly those involving the manipulation of d-band centers, the selective blocking of large adsorbates, and the integration of protective or functional coatings, are readily transferrable to other platforms that require sustained catalytic activity over prolonged operation.

Advanced in situ monitoring techniques such as operando X-ray absorption spectroscopy and high-resolution electron microscopy can be leveraged not only to examine degradation pathways in fuel cells, but also to elucidate poisoning mechanisms in diverse electrochemical systems [80]. By doing so, researchers can expedite the transition of promising anti-poisoning methods from laboratory proof-of-concept to large-scale deployment. Moreover, interdisciplinary collaborations encompassing materials science, chemical engineering, and computational modeling will be instrumental in realizing catalyst designs that exhibit high selectivity and durability across multiple applications [36,68]. This synergy offers significant potential for accelerating innovation beyond the confines of fuel cell technologies, paving the way for robust catalysts that can withstand both conventional and emerging contaminants in numerous electrochemical processes.

In summary, while this review has underscored the urgency and feasibility of combating catalyst poisoning in LT-PEMFCs and HT-PEMFCs, it also serves as a roadmap for developing broadly applicable solutions that address the multifaceted demands of modern electrochemical energy conversion and storage. The lessons learned here, such as tuning surface geometry, employing alloying for electronic effects, and combining multiple strategies for synergistic gains, can inform the next generation of catalysts designed to sustain activity and minimize deactivation in any high-performance electrochemical environment.