Atomically dispersed Pt and Fe sites and Pt-Fe nanoparticles for durable proton exchange membrane fuel cells

Proton exchange membrane fuel cells (PEMFCs) are promising clean energy devices but are limited by the cost and durability of Pt-based oxygen reduction reaction (ORR) catalysts. Achieving the fuel cell cost target (~US$30 kW−1) requires Pt loadings below 0.125 mg cm−2, but reducing Pt increases oxygen transport resistance and can compromise durability. While advances in Pt-based catalysts have improved mass activity in liquid cells, these gains often do not translate to fuel cell performance. Pt-group-metal-free (PGM-free) Me–N–C catalysts with single transition-metal atoms (for example, Fe) in nitrogen-doped carbon show promise but suffer poor durability. Previous hybrid strategies combining small amounts of Pt with Me–N–C either enhanced activity but suffered durability losses or improved stability without increasing activity. This study addresses these challenges by designing a hybrid electrocatalyst featuring atomically dispersed Pt and Fe single atoms and Pt–Fe alloy nanoparticles to enhance activity and durability at ultralow Pt loadings.

Prior work on Me–N–C catalysts (PGM-free) established high ORR activity but inadequate durability in fuel cells. Early studies used Me–N–C as supports for Pt-based catalysts to improve stability. Liu et al. reported a hybrid with ultralow Pt (2–3 wt%) comprising Pt–Co alloy nanoparticles on Co–N–C that achieved high ORR activity (1.77 A mgPt−1 at 0.9 V, iR-free) but significant activity loss (83% after 30,000 cycles, 45% after 22 h hold). Jaouen and co-workers showed adding 1–2 wt% Pt to Fe–N–C improved stability but not activity. These findings suggest small Pt additions can boost performance but durability remains a concern, motivating the development of hybrid catalysts that combine multiple active sites to synergistically enhance both activity and durability.

Synthesis: Fe-doped ZIF-8 was prepared from Zn(NO3)2·6H2O, FeSO4·7H2O, and 2-methylimidazole in methanol, washed, dried, and pyrolyzed in Ar at 1,000 °C for 1 h to form Fe–N–C. For Pt–Fe–N–C, Pt(II) acetylacetonate and 1,10-phenanthroline were mixed in ethanol, combined with Fe-doped ZIF-8, dried, ball-milled, then subjected to NH3 (900 °C, 15 min) followed by Ar (1,000 °C, 1 h) to yield the catalyst with ~1.7 wt% Pt and ~2.0 wt% Fe. A Pt–N–C reference was similarly prepared on ZIF-8 with NH3/Ar heat treatments (Pt ~2.3 wt%). Characterization: Morphology and atomic dispersion were examined by TEM and HAADF-STEM with EELS; crystal structure by XRD; bulk composition by ICP-MS; surface chemistry by XPS; textural properties by BET; local coordination by XAS (XANES/EXAFS) at Pt L3 and Fe K edges; Fe species by 57Fe Mössbauer; carbon support integrity by micro-Raman. Post-durability STEM/EELS/EDX probed structural evolution and Pt shell formation. Electrochemical RRDE: Catalyst inks (water:isopropanol 4:1 with Nafion) were drop-cast on a rotating-ring disk electrode. Measurements in 0.1 M HClO4 at 1,600 rpm evaluated ORR polarization, H2O2 yield and electron transfer number; durability by potential cycling (0.6–1.0 V) at 50 mV s−1. Fuel cell testing: MEAs (5 cm2) used Nafion HP membrane. Cathodes: Pt–Fe–N–C (0.88 mg cm−2 catalyst, 0.015 mgPt cm−2), Fe–N–C (3.5 mg cm−2), Pt–N–C (0.1 mgPt cm−2). Anode: Pt/C 0.1 mgPt cm−2. Conditions: 80 °C, 100% RH. Polarization in H2/O2 at 1.5/2.5 bar abs and in H2/air; cutoff 2.0 A cm−2. Durability: DOE protocol square-wave cycling between 0.6 and 0.95 V (3 s holds) up to 100,000 cycles; chronoamperometry at 0.6 V in H2/air or H2/O2 at 1 bar. Effluent water analyzed for F− by ion chromatography. Theory: DFT (VASP, RPBE-GGA, D3 dispersion, spin polarization; GGA+U for Fe) modeled ORR on candidate active sites: Pt–N,C; Pt–N2C2; Fe–N,C; Fe–Pt dual-site; and core–shell PtML/PtFe(111) (also 2–3 ML). Gibbs free energy diagrams constructed via computational hydrogen electrode at URHE=0.9 V; H2O2 reduction on PtML/PtFe(111) vs Pt(111) evaluated.

• Structure: Pt–Fe–N–C comprises abundant atomically dispersed Pt and Fe single atoms on N-doped carbon plus Pt–Fe nanoparticles. Small nanoparticles are ordered intermetallic PtFe; larger ones are disordered PtFex (1<x<3). XAS shows Pt–Fe bonding (Pt–Fe peak at ~2.23 Å) and Fe coordination consistent with alloys and Fe–N,C species; Mössbauer identifies singlet (Pt–Fe alloy) and doublets (Fe(II)N4) components.
• RRDE activity: Half-wave potential shifted from 0.765 V (Fe–N–C) to 0.909 V (Pt–Fe–N–C). Pt mass activity (RRDE, 0.1 M HClO4, 0.9 V, no iR correction) reached 1.74 A mg−1 (~10× Pt/C at 0.18 A mg−1). H2O2 yield max 2.4% with electron transfer number ~4.
• Electrolyte durability: Only 14 mV half-wave potential loss after 40,000 cycles (0.6–1.0 V). STEM shows preserved single atoms and ordered nanoparticles; Pt-shell formation observed.
• Fuel cell performance (H2/O2): With 0.015 mgPt cm−2 Pt–Fe–N–C cathode, peak power density 1.08 W cm−2 at 2.0 A cm−2; Pt/C (0.1 mgPt cm−2) reached 1.37 W cm−2. In H2/air, Pt–Fe–N–C outperformed Fe–N–C across current densities (peak 0.55 vs 0.33 W cm−2).
• Fuel cell mass activity (calibrated to 1 bar abs): 0.77 A mg−1 (H2/O2) and 0.74 A mg−1 (H2/air) at 0.9 Vi-free, ~3.7× Pt/C (0.21 A mg−1) and 1.75× DOE 2025 target (0.44 A mg−1).
• Fuel cell durability: After 100,000 square-wave cycles (0.6/0.95 V), MA at 0.9 Vi-free decreased slightly to 0.75 A mg−1 (−3%) and power density at 2.0 A cm−2 to 1.03 W cm−2 (−5%), surpassing DOE durability goals (<40% MA loss after 30,000 cycles). Chronoamperometry at 0.6 V: negligible current decay over 206 h in H2/air; ~5% drop over 210 h in H2/O2. Pt/C showed 8% drop in 120 h; Fe–N–C showed 73% drop in 43 h.
• Degradation indicators: F− emission in effluent lower for Pt–Fe–N–C than Fe–N–C and close to Pt/C after extended operation, indicating mitigated membrane/ionomer degradation.
• Structural evolution under cycling: Most nanoparticles formed stable PtFe@Pt core–shell with intermetallic core; only ~7% of larger disordered particles evolved into percolated, Pt-rich structures due to Fe leaching.
• Theory: DFT indicates Pt–N,C single-atom sites and PtML/PtFe(111) core–shell surfaces have low thermodynamic barriers for ORR (*OH removal 0.17–0.18 eV) and outperform Pt(111). PtML/PtFe(111) also facilitates H2O2 reduction with a lower barrier (0.18 eV) than Pt(111) (0.34 eV), supporting reduced peroxide accumulation.

The study demonstrates that combining multiple active site types—Pt single atoms (Pt–N,C), Fe single atoms (Fe–N,C), and Pt–Fe alloy nanoparticles that evolve into PtFe@Pt core–shell—synergistically enhances Pt utilization, ORR activity, and durability at ultralow Pt loading. Pt–N,C and PtFe@Pt provide highly active ORR sites with favorable *OH desorption energetics, while Fe–N,C sites contribute activity but are known peroxide producers. The proximity of PtFe@Pt nanoparticles enables rapid H2O2 reduction, lowering peroxide concentration and mitigating membrane and ionomer degradation, as reflected by reduced fluoride emissions. Durable core–shell formation preserves intermetallic cores and resists nanoparticle coarsening, maintaining performance through 100,000 cycles. Although overall power at very high current densities remains below Pt/C due to lower Pt loading, the mass activity and stability strongly surpass benchmarks and DOE targets, validating the hybrid-site strategy for low-Pt PEMFC cathodes.

A hybrid ORR electrocatalyst with ultralow Pt loading (~1.7 wt%) comprising atomically dispersed Pt and Fe single atoms and Pt–Fe alloy nanoparticles was synthesized and comprehensively characterized. In fuel cells, it achieved 0.77 A mg−1 at 0.9 Vi-free (1 bar) and 1.08 W cm−2 at 2.0 A cm−2 with only 0.015 mgPt cm−2 at the cathode. Durability was exceptional, with 97% activity retention after 100,000 cycles and stable current at 0.6 V over >200 h. DFT and experiments indicate active sites include Pt–N4Cx, Fe–N4Cx, and PtFe@Pt, where the latter both enhances ORR and reduces H2O2, contributing to membrane stability. These results highlight synergistic active-site interactions as a pathway to active, durable, and low-Pt PEMFC catalysts. Future work may optimize nanoparticle size/ordering, tune single-atom coordination environments, and engineer spatial distributions to further improve high-current-density performance and robustness.

Despite high mass activity and durability, the fuel cell with Pt–Fe–N–C could not match Pt/C performance at very high current densities, attributed to the much lower Pt loading at the cathode. A small fraction (~7%) of larger disordered PtFex nanoparticles evolved into percolated, Pt-rich structures via Fe leaching under cycling. Performance and durability were evaluated under specific temperatures, humidities, and gas pressures; translation to varying operating conditions and scale-up may require further validation.