The rapid growth in global energy demand has raised concerns over fossil fuel depletion and environmental pollution, driving the pursuit of clean and sustainable energy solutions. Among various emerging technologies, hydrogen fuel cells have attracted significant attention for their high efficiency and low emissions. In particular, proton exchange membrane fuel cells (PEMFCs) are considered one of the most promising candidates for transportation and stationary power generation applications due to their high power density, low operating temperature, and ease of system integration. However, the performance and durability of PEMFCs are highly sensitive to air contaminants, which can lead to significant degradation under real-world operating conditions. Therefore, understanding the impact of airborne impurities on PEMFC performance and developing effective mitigation strategies are essential for ensuring their reliable and long-term operation.

Extensive studies have been conducted to address this issue, primarily focusing on fuel cell catalyst material improvements and membrane modifications [1–16]. For instance, Kim et al. [14] investigated the effects of ionomer content and relative humidity on cell performance, finding that higher humidity improved cell voltage, while excessive ionomer content adversely affected performance. Laribi et al. [15] examined the influence of air temperature on membrane water uptake and reported that excessive water accumulation can lead to flooding and performance loss. Wang et al. [17] evaluated multiple parameters, including humidification levels, stoichiometric ratios, and operating temperatures, concluding that the air stoichiometric ratio had the most significant impact on cell performance. Zhao et al. [17] further investigated operating conditions such as cell temperature, backpressure, and relative humidity, observing that elevated temperatures led to membrane dehydration and increased resistance, while higher back pressure reduced ohmic and activation losses. Additionally, cathode flooding and carbon corrosion were found to be strongly correlated with local water content near the outlet.

The adverse effects of airborne contaminants in both the anode and cathode have also been widely studied [18–21]. Shi et al. [18] demonstrated that CO and H₂S contamination at the anode led to considerable performance losses, with CO causing more severe voltage drops but offering better recoverability. Zhai et al. [19–21] explored various cathode contaminants, including bromomethane, acetylene, and acetonitrile, highlighting the role of surface adsorption in hindering mass transport and reducing catalyst activity. Some contaminants exhibited partial recoverability under neat air, while others caused irreversible damage. Reshetenko et al. [22] focused on halogenated and sulfuric compounds, reporting that even low concentrations of CH3Br led to irreversible performance loss due to platinum agglomeration and reduced electrochemical surface area. Further studies specifically investigated the effects of SO₂ and NO₂ contamination in the cathode air stream [23–40]. Prithi et al. [23] reported that SO₂ adsorption on Pt catalyst sites reduced performance, with partial recovery attributed to H₂SO₄ hydrolysis at high current densities. Nagahara et al. [25] found that sulfur-based pollutants caused more severe degradation than nitrogen-based contaminants. Fu et al. [26] employed cyclic voltammetry (CV) to detect the oxidation of sulfur species at high potentials, resulting in electrochemical surface area loss. Dorn et al. [34] confirmed catalyst layer degradation and MEA thinning through SEM analysis. Jing et al. [35] investigated the interaction between SO₂ and NO₂ mixtures, revealing that NO₂ preferentially adsorbs on the catalyst surface, thereby inhibiting SO₂ adsorption. Their study evaluated three mixture ratios (NO₂:SO₂ = 3:1, 1:1, and 1:3) to examine the synergistic or antagonistic effects of these combined gases. Similarly, Lin et al. [42] investigated NH₃/NO_x co-contamination and reported that NO_x dominated the initial performance decay, while NH₃ contributed more significantly to long-term degradation, highlighting that mixed gas effects are not simply additive. More recently, Lin et al. [43] systematically examined NO₂ contamination under low-concentration conditions and found that NO₂ poisoning was largely reversible under clean air recovery. Also, Reshetenko [44] conducted experiments under low SO₂ concentrations and comprehensively analysed the corresponding recovery procedures.

Although previous studies have provided valuable insights into the individual effects of various airborne contaminants, including SO₂ and NO₂, on PEMFC performance, most investigations have focused on single-species exposure or limited mixture ratios under low concentration. The interaction mechanisms and combined effects of these contaminants under varying concentration ratios remain insufficiently understood, particularly under dry operating conditions that are frequently encountered in real-world applications such as automotive and portable systems. Moreover, the recovery characteristics following exposure to mixed contaminants have not been comprehensively characterized, leaving a critical knowledge gap for practical fuel cell durability management. While it is known that these contaminants degrade cell performance through catalyst surface adsorption, mass transport hindrance, and electrochemical surface area loss, their interaction mechanisms when introduced as mixed gases have not been sufficiently explored. Furthermore, the comparative severity of SO₂ and NO₂ poisoning, as well as their potential synergistic or antagonistic effects on fuel cell degradation and recovery, is not fully understood.

Accordingly, this study aims to systematically examine the combined effects of SO₂ and NO₂ mixtures on PEMFC performance. To accelerate the degradation process and effectively evaluate the impact of contaminants within a practical experimental timeframe, the total impurity concentration was set at 100 ppm for all mixture ratios. By employing electrochemical diagnostic techniques such as cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), the research seeks to clarify the degradation mechanisms, quantify changes in electrochemical properties, and provide insights into the recoverability of PEMFC performance under mixed-gas exposure.

This study systematically investigated the combined effects of nitrogen dioxide (NO₂) and sulfur dioxide (SO₂) contamination on the performance of proton exchange membrane fuel cells (PEMFCs) under controlled operating conditions. All experiments were designed to simulate accelerated degradation scenarios by maintaining a relatively high total contaminant concentration of 100 ppm, enabling the evaluation of performance loss and recovery behaviour within a practical time frame.

Fig. 2 shows the polarization curves reflecting the performance degradation and recovery characteristics of the PEMFC under various NO₂ and SO₂ mixing ratios. In these experiments, the total concentration of contaminants was fixed at 100 ppm for all cases to enable a direct comparison of the individual and combined effects of the two gases. The tested conditions include (a) 100 ppm NO₂, (b) 75 ppm NO₂:25 ppm SO₂, (c) 50 ppm NO₂:50 ppm SO₂, (d) 25 ppm NO₂:75 ppm SO₂, and (e) 100 ppm SO₂. The results clearly indicate that the presence of NO₂ leads to a faster and more severe decline in cell performance compared to that of SO₂ under constant total concentration of contaminants of 100 ppm. As the concentration of NO₂ in the mixture increased, a more rapid voltage degradation was observed, particularly in the high-current-density region. In contrast, SO₂-rich conditions exhibited comparatively gradual degradation behavior. This trend suggests that NO₂ has a more aggressive poisoning effect on the cathode catalyst layer, likely due to its strong oxidative nature and surface adsorption characteristics, which impede the oxygen reduction reaction (ORR) more effectively than SO₂ under the tested conditions. These findings are consistent with previous studies reporting the significant impact of NO₂ on PEMFC performance, further emphasizing the need for effective contaminant mitigation strategies in fuel cell systems operating in polluted air environments.

Fig. 3 illustrates the temporal evolution of cell voltage during contamination and recovery phases under various NO₂ and SO₂ mixing ratios, conducted at a constant current density of 0.8 A cm^–2. Upon introducing 100 ppm of NO₂, the cell voltage rapidly decreased from 0.642 V to 0 V, followed by recovery to 0.62 V after switching back to neat air. A similar voltage profile was observed under 75 ppm NO₂:25 ppm SO₂ system with the voltage dropping to 0 V and subsequently recovering to 0.61 V. For the 50 ppm NO₂:50 ppm SO₂ system and 25 ppm NO₂:75 ppm SO₂ system, the minimum voltages recorded were 0.32 V and 0.37 V, respectively, with partial recovery to 0.58 V and 0.57 V after air purging. In the case of 100 ppm of SO₂, the cell voltage decreased to 0.45 V and recovered up to 0.56 V. These trends confirm that NO₂-rich conditions induce a more immediate and complete performance loss, primarily due to the rapid adsorption of NO₂ molecules onto the platinum catalyst surface, effectively blocking the oxygen reduction reaction (ORR) sites [35].

In contrast, mixtures containing higher SO₂ concentrations exhibited a more gradual voltage decay, as the relatively slower adsorption kinetics of SO₂ allowed for partial retention of catalytic activity during exposure. As summarized in Table 1, the corresponding recovery rates were 96.57%, 95.01%, 90.34%, 88.78%, and 87.27% for 100 ppm NO₂, 75 ppm NO₂:25 ppm SO₂, 50 ppm NO₂:50 ppm SO₂, 25 ppm NO₂:75 ppm SO₂, and 100 ppm SO₂, respectively. The recovery rate clearly decreased with increasing SO₂ concentration in the contaminant mixture.

This behavior is attributed to the distinct adsorption characteristics of NO₂ and SO₂ on Pt catalysts [34,35]. NO₂ forms relatively weaker and more reversible surface species, which can be readily desorbed or oxidized during the recovery phase [35]. In contrast, SO₂ generates strongly adsorbed sulfur species, which persist on the catalyst surface even after neat air purging, leading to residual catalyst poisoning. These findings align with previous reports indicating that prolonged SO₂ exposure not only increases charge transfer resistance but also contributes to irreversible catalyst layer degradation and MEA thinning [26,28,38]. EIS analysis further supported this conclusion, revealing elevated charge transfer resistance values in cells exposed to higher SO₂ concentrations, even after recovery, indicative of incomplete electrochemical surface restoration. Consequently, the recovery efficiency deteriorates progressively as the SO₂ content increases in the contaminant mixture [29,38].

Fig. 4 illustrates the cyclic voltammetry (CV) profiles obtained for various contaminant ratios of NO₂ and SO₂ introduced at the cathode side of the proton exchange membrane fuel cell (PEMFC). The investigated conditions include five different gas mixtures: (a) 100 ppm NO₂, (b) 75 ppm NO₂:25 ppm SO₂, (c) 50 ppm NO₂:50 ppm SO₂, (d) 25 ppm NO₂:75 ppm SO₂, and (e) 100 ppm SO₂. For each contamination scenario, CV measurements were conducted both prior to contaminant exposure and following the electrochemical recovery process, enabling a direct assessment of contaminant adsorption and subsequent removal behaviour on the Pt catalyst surface. Notably, the adsorption of NO₂ onto the Pt catalyst was found to be largely reversible under the applied recovery conditions. As evidenced by the data presented in Table 1, recovery efficiencies of 96.57% and 95.01% were achieved in specific cases, indicating that the majority of NO₂ species adsorbed during contamination were effectively desorbed during the recovery process. This high recovery efficiency highlights the comparatively moderate binding strength of NO₂ on the Pt surface, which allows for efficient removal through electrochemical cleaning techniques. These findings are particularly significant when considering the operational stability of PEMFC systems in environments where NO₂ contamination is prevalent, as they suggest that performance losses induced by NO₂ can be substantially mitigated through appropriate recovery protocols. Moreover, the comparison of recovery behaviour across different contaminant ratios provides valuable insights into the competitive adsorption dynamics between NO₂ and SO₂ species on the cathode catalyst layer, which will be further discussed in subsequent sections.

In all CV profiles obtained under the various contamination conditions, oxidation peaks appearing around 0.75 V were consistently observed, which are typically associated with the formation of Pt-oxide layers on the catalyst surface. These oxidation features serve as important indicators of surface oxidation states and catalytic activity changes induced by contaminant adsorption. Additionally, the emergence of distinct peaks near 1.1 V was noted, corresponding to the presence of residual sulfur species that remained adsorbed on the Pt catalyst surface, even after the recovery process. The persistence of these sulfur-related oxidation peaks suggests incomplete desorption of sulfur-containing compounds, reflecting the strong adsorption affinity of sulfur species for the Pt surface.

As the concentration of SO₂ in the contaminant mixture increased, several notable changes in the CV profiles were observed. In particular, enhanced hydrogen desorption peaks appeared in the region of approximately 0.2 V, indicating alterations in the hydrogen adsorption/desorption characteristics of the catalyst surface due to sulfur adsorption. Furthermore, more prominent Pt reduction peaks were detected near 0.8 V under higher SO₂ concentrations, which reflect modifications in the reduction behaviour of surface Pt oxides in the presence of adsorbed sulfur species.

Collectively, these electrochemical features suggest that SO₂ exerts a substantially more pronounced influence on the catalyst surface than NO₂. This conclusion is further supported by the observation that the characteristic oxidation peaks near 0.75 V diminished significantly as the SO₂ ratio increased in the contaminant gas mixture, indicating the progressive deactivation of active Pt sites by adsorbed sulfur species. These findings are consistent with previous studies [35,38], which have reported the dominant poisoning effect of SO₂ over NO₂ in PEMFC cathode environments. The strong and persistent adsorption of sulfur compounds not only inhibits the formation of Pt oxides but also impedes hydrogen adsorption and desorption processes, ultimately leading to severe degradation of electrochemical surface properties and fuel cell performance.

To further elucidate the impact of impurities on the electrochemical properties of the catalyst layer, the electrochemically active surface area (ECSA) of the Pt catalyst was calculated based on the hydrogen adsorption/desorption charge obtained from cyclic voltammetry (CV) measurements. The ECSA was estimated using the following equation:

where Q_H is the integration charge in the hydrogen adsorption/desorption region (mC), r is the charge required to reduce a monolayer of protons on Pt (210 μC cm^–2), and L is the Pt loading in the electrode (0.4 mg cm^–2).

Collectively, these electrochemical features suggest that SO₂ exerts a substantially more pronounced influence on the catalyst surface than NO₂. This conclusion is further supported by the observation that the characteristic oxidation peaks near 0.75 V diminished significantly as the SO₂ ratio increased in the contaminant gas mixture, indicating the progressive deactivation of active Pt sites by adsorbed sulfur species. These findings are consistent with previous studies [35,38], which have reported the dominant poisoning effect of SO₂ over NO₂ in PEMFC cathode environments. The strong and persistent adsorption of sulfur compounds not only inhibits the formation of Pt oxides but also impedes hydrogen adsorption and desorption processes, ultimately leading to severe degradation of electrochemical surface properties and fuel cell performance.

Based on this calculation, the ECSA values were compared between initial and recovery processes under various contamination conditions. As shown in Fig. 5 the ECSA decreased progressively with increasing SO₂ content in the contaminant mixtures, and this decrease persisted even after the recovery phase, indicating irreversible degradation of catalyst sites primarily under SO₂-rich conditions. In addition, CV profiles obtained under SO₂-rich conditions exhibited pronounced oxidation peaks near 1.0 V, which are attributed to strongly adsorbed sulfide species on the Pt surface. These persistent peaks were not removed during the recovery process, providing further electrochemical evidence of the irreversible binding of sulfur-containing compounds that contribute catalyst deactivation.

To quantify the extent of catalyst poisoning, an adsorption index was defined based on the hydrogen underpotential deposition (HUPD) charge, obtained by integrating the anodic (dehydrogenation) peak within the potential range of 0.05–0.35 V (vs. RHE). The residual poisoning fraction (θ) was subsequently fitted using the competitive Langmuir isotherm model, expressed as:

where K_NO2 and K_SO2 represent the adsorption affinity constants of NO₂ and SO₂, respectively, C_NO2 and C_SO2 denote their concentrations. Using a scan rate, Pt loading, and an assumed single charge listed above, the fitting results yielded K_NO2 and K_SO2, as 4.6 × 10^–4 ppm^–1 and 1.6 × 10^–3 ppm^–1, respectively. The K_SO2 is approximately 3.5 times larger than K_NO2. These findings demonstrate that the adsorption affinity of SO₂ is substantially stronger than that of NO₂, indicating that SO₂ is the dominant poisoning species and successfully outcompetes NO₂ for active adsorption sites on the platinum catalyst, even when present in a gas mixture.

From the competitive Langmuir analysis, the adsorption equilibrium constant for NO₂ obtained from the 100-ppm experiments was obtained above. Accordingly, the steady-state poisoning fraction is given by equation (2) was approximately 0.044, which lies well below saturation. Under the Langmuir assumption that adsorption equilibrium constant is concentration-independent at constant temperature, extrapolation to 1 ppm yields the poisoning fraction approximately 4.6 × 10^–4, about 100 times smaller than that of 100 ppm. Also the poisoning fraction of SO₂ obtained from equation (2) at concentrations of 100 ppm and 1 ppm are 0.138 and 1.59 × 10^–3, respectively, which shows similar result from those of NO₂.

Fig. 6 presents a comprehensive comparison of electrochemical impedance spectroscopy (EIS) results obtained under three distinct operational conditions: the initial pristine state, during exposure to contaminant gases, and following the recovery process. The introduction of NO₂ and SO₂ contaminants into the cathode stream resulted in a noticeable increase in the activation resistance component, as evidenced by the enlarged semicircle in the high-frequency region of the Nyquist plots. This increase in resistance is attributed to catalyst poisoning effects, where adsorbed contaminant species hinder the electrochemical reactions at the catalyst surface by occupying active sites essential for oxygen reduction. Notably in Fig. 6(a) and Fig. 6(b), which correspond to the conditions of 100 ppm NO₂ and 75 ppm NO₂:25 ppm SO₂ exposure, respectively, the resistance values observed after the recovery process nearly returned to their initial, pre-contamination levels. This behavior indicates a high recovery efficiency for these cases, consistent with the CV analysis results and recovery efficiency values listed in Table 1. The effective desorption of NO₂ and limited SO₂ content in these mixtures likely contributed to the minimal residual degradation of the catalyst layer, enabling the near-complete restoration of electrochemical performance.

In contrast, the EIS profiles shown in Fig. 6(c), Fig. 6(d), and Fig. 6(e) -representing higher SO₂ concentration conditions-exhibited significantly elevated resistance values after the recovery process when compared to the initial state. This persistent increase in resistance suggests the incomplete desorption of sulfur-containing species from the Pt catalyst surface, even after the applied recovery procedure. The strong adsorption affinity and chemisorption characteristics of sulfur compounds likely resulted in irreversible occupation of active catalytic sites, thereby causing residual catalyst degradation and impeding the full recovery of electrochemical performance. These findings align with the CV data and further reinforce the conclusion that SO₂ exerts a more detrimental and lasting effect on PEMFC cathode catalyst layers than NO₂, particularly under mixed-contaminant and SO₂-rich conditions.

Based on the equivalent circuit model, the EIS fitting results reveal the impact of NO₂ :SO₂ mixtures on PEMFC performance. The significant increase in charge transfer resistance (Rct) under NO₂-rich conditions indicates that NO₂ is a primary cause of kinetic inhibition. Conversely, SO₂-rich mixtures lead to a more modest increase in R_ct but a notable rise in solution resistance (R_s), suggesting that SO₂ may also affect the membrane’s ohmic properties. The partial recovery of resistance values after exposure shows that the poisoning is reversible under the given conditions. The fittings for the EIS results are listed in Table 2.

In this study, the individual and combined effects of nitrogen dioxide (NO₂) and sulfur dioxide (SO₂) impurities on the performance decay and recovery behavior of a PEMFC were systematically investigated while maintaining a constant total contaminant concentration of 100 ppm. A series of performance evaluations, including cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), provided valuable insights into the distinct poisoning mechanisms and recovery characteristics associated with each contaminant.

A sharper and more immediate decline in cell performance was observed with increasing NO₂ concentration. This rapid degradation was primarily attributed to the fast and strong adsorption of NO₂ molecules onto the Pt catalyst surface, leading to a temporary deactivation of active sites essential for the oxygen reduction reaction (ORR). Despite this aggressive initial poisoning effect, recovery analysis revealed that NO₂ contamination was largely reversible under the applied recovery protocols. High recovery efficiencies restored electrochemical active surface area, and significantly reduced charge transfer resistance after recovery indicated effective desorption of NO₂ from the catalyst surface.

In contrast, SO₂-induced degradation exhibited a slower but more persistent progression. The gradual decline in performance was linked to the formation of stable and strongly bound sulfur species on the Pt catalyst, which were difficult to remove even after the recovery process. In addition, ECSA analysis confirmed that this irreversible degradation was accompanied by a progressive loss of electrochemical active surface area with increasing SO₂ content, even after the recovery phase. CV measurements also exhibited distinct oxidation peaks near 1.0 V under SO₂-rich conditions, indicating the presence of strongly adsorbed sulfide species on the Pt surface. These persistent peaks were not removed during the recovery process, providing further electrochemical evidence of irreversible sulfur poisoning. This irreversible catalyst poisoning was confirmed by incomplete recovery rates, diminished CV peak areas, and sustained high charge transfer resistance values in EIS measurements, particularly under SO₂-rich conditions. This conclusion is consistent with the comprehensive analysis of recovery procedures conducted by Reshetenko [44].

Collectively, these findings underscore the significant influence of impurity species composition on PEMFC operational stability and long-term performance. The results suggest that while NO₂ contamination effects can be effectively mitigated through appropriate recovery procedures, SO₂ poses a more critical challenge due to its irreversible binding behavior. This study provides important baseline data for understanding contaminant-induced degradation phenomena in PEMFCs and emphasizes the necessity of developing advanced catalyst materials or enhanced recovery strategies, particularly for sulfur-containing environments. Future work should focus on investigating contaminant interactions at elevated operating conditions, long-term exposure effects, and the application of catalyst modifications or protective layers to improve resistance against irreversible poisoning.