Primacy of lattice distortion over strain in platinum fuel cell nanoalloy catalysts

Abstract

Alloying platinum with early or late transition metals enhances its intrinsic activity toward the oxygen reduction reaction for proton exchange membrane fuel cells (PEMFCs), according to strain and ligand effects. However, these alloying effects disappear following the dissolution of non-noble metal component(s) during operation, leading to PEMFC performance degradation. In this study, we investigate PtNi nanoalloys across critical stages of their development (viz. as-synthesized, post-electrochemical activation, after membrane electrode assembly fabrication, and following accelerated stress testing) using a comprehensive set of ex situ, in situ, operando, and post mortem characterization techniques, which allows assessing the contributions from alloying and structural effects. Our results reveal that local lattice distortion, rather than global strain and ligand effects, is an important factor effectively contributing to both catalytic activity and durability in PEMFC. These finding challenges conventional electrocatalyst design strategies and validates the defect-engineering strategy for advanced fuel cell applications, independently from transition metal(s) retention.

---

Broader context

The transition to a low-carbon economy hinges on the deployment of clean energy technologies, with hydrogen fuel cells offering a promising solution for zero-emission power in transport and other sectors. However, the widespread adoption of proton exchange membrane fuel cells (PEMFCs) is limited by their reliance on costly and scarce platinum catalysts. While alloying platinum with other transitions metals like nickel has improved performance, these benefits often vanish under real-world conditions due to metal dissolution and structural degradation. This study tackles a central challenge in catalyst design: how to maintain high performance and stability without depending on unstable alloying elements. By examining the microstructure of platinum–nickel catalysts across all stages of development and operation, the research reveals that local lattice distortion, rather than traditional alloying (strain or ligand) effects, plays an important role in controlling catalytic activity and durability. This finding shifts the paradigm in electrocatalyst design, offering a pathway to more robust and efficient PEMFCs that are less reliant on vulnerable alloying strategies. The insights open new avenues for defect engineering and catalyst optimization, with implications for cleaner transportation technologies and sustainable energy systems worldwide.

Introduction

Hydrogen fuel cells are expected to play a key role in achieving decarbonization targets across various sectors, such as transportation when implemented in fuel cell electric vehicles (FCEVs). Among the different technologies of fuel cells, proton exchange membrane fuel cells (PEMFCs) have emerged as the most promising candidate for efficient conversion of hydrogen chemical energy to electricity. However, despite their high technological maturity, PEMFCs affordability, performance, and durability remain to a great extent limited by the required high amount of costly platinum (Pt) electrocatalyst at the cathode where the sluggish oxygen reduction reaction (ORR) occurs.

Considerable research efforts have been made over the past 20 years to develop active and stable Pt-based ORR catalysts to decrease the cost of PEMFCs, and comprehensive literature reviews exist on this prolific subject.^1–6 A general conclusion is that among various strategies explored, the one consisting of alloying Pt with early or late transition metal(s) (Ni, Co, Cu, etc.) does not only lead to the most significant enhancement of Pt intrinsic activity for the ORR (due to the so-called strain, ligand and ensemble effects⁷), but also enables a better utilization of Pt due to its possible concentration specifically at the catalyst surface. However, the most active materials prepared following this strategy are still facing important bottlenecks. The main one must be the ineluctable dissolution of the non-noble metal element from the Pt alloy, not only in the conditions relevant to the PEMFC cathode, but even during the membrane electrode assembly (MEA) fabrication.^8–11 Consequently, practical Pt bimetallic catalyst surfaces are in fact dealloyed beforehand. This results in core–shell nanoparticles in which the transition metal fraction is protected from dissolution by a thick Pt shell of ∼1.2 nm,¹² in which lattice strain is the only alloying effect contributing to electrocatalytic activity enhancement.¹³ In fact, in Pt alloy catalysts, the ligand effect operates only over a very short range. Experimental and theoretical work showed the range of the ligand effect in Pt alloy catalysts for the ORR is on the order of the first few atomic layers (∼1–3 layers or ∼0.5–1 nm). Deeper subsurface atoms do not measurably modify catalytic sites’ d-band and thus have little influence on ORR activity.⁷

Complementarily to the alloying effects mentioned above, the major role of local surface structure in heterogeneous (electro)catalysis is known for long-time.¹⁴ In the specific case of the ORR, numerous studies have observed and/or predicted the effects of surface crystallographic orientation^15,16 (more generally the catalytic site coordination^17–19) and local lattice distortion (local strain or microstrain)^20–26 on ORR kinetics in model liquid electrolyte. In this case, the implementation of structural disorder (grain boundaries, stacking faults, atomic vacancies, surface roughness etc.) allows the local tuning of such properties, leading to the emergence of catalytic sites with desirable coordination and/or local strain. Still, the viability of such an alternative strategy (because not relying on the difficult retention of highly unstable transition metal elements in the conditions of the ORR) has never been validated in PEMFC.

In this contribution, a series of carbon-supported, dealloyed PtNi/C electrocatalysts with controlled degrees of lattice strain and distortion are synthesized by varying the instrumental pH of a microwave-assisted polyol process. The impact of the materials’ initial physico-chemistry on their electrocatalytic activity and stability for the ORR is investigated throughout the key steps of catalyst development, (viz. as-synthesized, before and after electrochemical activation, after MEA manufacturing, and accelerated stress testing) in liquid electrolyte with the rotating-disk electrode (RDE) technique and in a PEMFC single-cell. Overall, the interpretation of electrochemical measurements obtained using ex situ, in situ, operando and post mortem techniques, including high-resolution transmission (scanning) electron microscopy possibly coupled with energy dispersive X-ray spectroscopy (HR(S)TEM)/X-EDS, electrochemical on-line inductively coupled plasma mass spectrometry (ICP-MS), and especially synchrotron wide-angle X-ray scattering (WAXS, see Fig. 1) reveals the important contribution of local lattice distortion over global lattice strain and ligand effect in controlling the performance of operating Pt-based fuel cell cathode catalysts. Especially, the lattice distortion content dynamics observed during both electrode preparation and operation results in desirable performance and stability for the most monocrystalline and high surface area particles synthesized, independently from its transition metal content.

Fig. 1 Establishing the structure–activity–stability relationship for the ORR across key stages of fuel cell catalysts development by means of ex situ, in situ and operando wide angle X-ray scattering. Because the structure and chemistry of fuel cell electrocatalysts largely depend on their environment and operation history, this study investigates structure–activity-stability trends for the ORR by probing the microstructure of materials as synthetized (ex situ in powder capillary), in 0.1 M HClO₄ liquid electrolyte (in situ in thin-film cell configuration) and operando in PEMFC by WAXS. A particular attention is dedicated to quantify the impact of the so-called catalyst activation and MEA fabrication on the initial activity, as well as the impact of long-term operation by means of accelerated stress tests.

Results and discussion

A series of six carbon-supported, PtNi/C catalysts (denoted as PtNi-X in the following) were obtained by controlling the initial instrumental pH (set to X = 6.3, 8, 9.5, 11 and 12) of a ‘one pot’ microwave-assisted polyol method (see Supplementary Information for details). As shown in Fig. 2a–c and in Fig. S1 and Table S1 of the SI, and in agreement with a previous contribution²⁷ the pH of the synthesis medium is found to control the particle growth mechanism, notably the initial formation of either Ni-rich nuclei (low pH) or Pt-rich nuclei (high pH), followed by the formation of a Pt-rich shell (low pH) or a Ni-rich shell (high pH). Dealloying was performed by acid treatment in 1 M H₂SO₄ for 20 h, leading to ‘hollow’ to ‘skeletal’ particles with increasing synthesis pH (see Fig. S2). Moreover, X-EDS linescans (Fig. S2) reveal the formation of Pt-rich overlayers after dealloying. The Pt shell thickness is found superior to 1 nm, thus exceeding the range of the ligand effect possibly induced by subsurface Ni-rich regions.⁷ HRTEM and associated fast Fourier transform images in Fig. 2c also reveal an increase in particle polycrystallinity with the synthesis pH, with near monocrystalline particle for PtNi-6.3 sample in contrast to PtNi-9.5 which appears clearly polycrystalline.

Fig. 2 Controlling the morphology and structure of PtNi/C catalysts by varying the pH of the synthesis medium. (a) TEM images of the PtNi/C catalysts; (b) proposed particle formation mechanisms depending on the instrumental pH of the synthesis medium; (c) HRTEM images with associated Fourier transform for PtNi-6.3 and PtNi-9.5 catalysts; (d) comparison of the global lattice strain vs. Ni content trend with the theoretical Vegard's law; (e) crystallite size and (f) microstrain values derived from Rietveld refinement of the WAXS patterns plotted for the catalysts as a function of the synthesis pH. In (e) and (f), the dotted horizontal lines indicate the values for the Pt/C benchmark.

The microstructures of the as-synthesized electrocatalysts were investigated through the refinement of ex situ WAXS using the Rietveld method (see SI). As shown in Fig. S3, all the materials present a unique face-centred cubic (fcc) phase, confirming the formation of disordered PtNi alloys. As shown in Fig. 2d, the refined lattice strain values of the PtNi/C catalysts compared to a Pt/C benchmark from Johnson Matthey (Pt-JM) vary in good agreement with Vegard's law considering their Ni at% estimated from ex situ ICP-MS. The refined crystallite sizes span between 4.7 and 7.5 nm for all PtNi/C materials (see in Fig. 2e), which remains slightly higher than that of Pt/C-JM (2.8 nm), but largely below their respective TEM sizes. As shown in Fig. 2f, varying the synthesis pH allows control of the initial degree of microstrain present in the various PtNi/C catalysts, ranging from 128%% (%% means the value has been multiplied by 10⁴) for PtNi-6.3 to 228%% for PtNi-9.5. Thus, only a mildly positive impact of microstrain is expected for the PtNi-6.3 catalyst, and a maximal contribution for the PtNi-9.5 catalyst. However, and as shown in the following, initial ex situ measurements of microstrain are not necessarily relevant to predict electrocatalytic activity.

Indeed, the electrochemical properties of the various catalysts were first screened in 0.1 M HClO₄ using the RDE technique (see SI), after an electrochemical activation procedure of 50 potential cycles (triangular profile) between 0.05 and 1.23 V vs. RHE at 500 mV s⁻¹ in N₂-saturated electrolyte (Fig. 3a). As highlighted by several research groups,^8,28–31 and monitored here by means of in situ WAXS and on-line ICP-MS, this activation protocol triggers significant microstructural and chemical changes for the PtNi catalysts (data shown only for the two most active PtNi samples and Pt benchmark), such as lattice strain relaxation (Fig. 3b), distinct microstrain evolutions (Fig. 3c) and both Pt and Ni metal dissolution (Fig. 3d). After activation, the CO_ads stripping method was used to determine their electrochemically active surface area (ECSA). As shown in Fig. 3e, and in agreement with previous contributions,^{21,22,24,32,33} the position and shape of the CO_ads oxidation profiles are largely dependent on the PtNi/C structure and chemistry. Notably, the feature at ∼0.7 V vs. RHE associated with CO_ads oxidation in the presence of grain boundaries, is minimal for the only partially aggregated 40 wt% Pt-JM and with various relative intensities for PtNi catalysts, confirming the insights from HRTEM images regarding varying degree of polycrystallinity as a function of synthesis pH. The derived Pt specific surface areas of the PtNi catalysts follow an overall decreasing trend with the synthesis pH, with the highest value of 57 ± 0.7 m² g_Pt⁻¹ obtained for PtNi-6.3, approaching the Pt-JM benchmark measured as 61 ± 5 m² g_Pt⁻¹ (Fig. 3f). The Tafel plot for the ORR displayed in Fig. 3g shows an overall enhancement of the mass-normalized ORR kinetic activity (MA) for the PtNi catalysts over the Pt-JM benchmark, with the greatest enhancement factors at 0.95 V vs. RHE of ca. 2.5 and 2.1 for PtNi-6.3 and PtNi-9.5, respectively (note the PtNi-6.3 sample exceeds 1.0 A mg_Pt⁻¹ at 0.90 V vs. RHE). These MA values can be rationalized by the product of the specific surface areas in Fig. 3f and the Pt surface-normalized ORR kinetic activities (SA) in Fig. 3h. Interestingly, the trend observed in SA does not follow the trends from ex situ measurements of either the strain or microstrain values (Fig. 2d and f). Such discrepancy in structure–activity trends from ex situ characterization is resolved in the light of the in situ characterization in Fig. 3b and c, in which the microstrain retention observed for PtNi-6.3 over PtNi-9.5 despite both undergoing lattice strain relaxation translates into a consistent ‘true’ structure-ORR activity relationship, where the initial ORR activity follows a linear relationship with the surface distortion (SD) descriptor, introduced in a previous contribution,²⁵ but not the global strain (see in Fig. S4). Moreover, to rule out the possibility of the ligand effect contributing to the ORR activity, the ratios between the integrated charges for H_UPD and CO_ads stripping (Q_CO/2Q_H) were calculated from the voltammograms in Fig. 3e. As shown in Fig. S4, all Q_CO/2Q_H values are below 0.9, which confirms (i) the absence of a thin Pt-skin on top of Ni-rich subsurface (thus the absence of ligand effect)^7,34 and (ii) the presence of lattice-distorted surfaces³² for PtNi-6.3 and PtNi-9.5, with low Q_CO/2Q_H values of 0.83 and 0.81, respectively.

Fig. 3 Activation-activity-stability in 0.1 M HClO₄. (a) Electrode potential profile; (b) lattice constant; (c) microstrain values and (d) metal dissolution rates for Pt-JM (black) PtNi-6.3 (green) and PtNi-9.5 (cyan) catalysts derived from in situ WAXS and on-line ICP-MS as a function of time during electrochemical activation; (e) CO_ads stripping voltammograms; (f) Pt specific surface area derived from CO_ads stripping experiments; (g) Pt mass-normalized and (h) Pt surface-normalized ORR Tafel plot with insert showing the activity at E= 0.95 V vs. RHE for all the catalysts; (i) Pt surface-normalized activity evolution (left axis, square symbols) during 25 000 AST potential cycles and microstrain values (right axis, circle symbols) during 5000 AST cycles from in situ WAXS. The AST comprised square wave potential cycles between 0.60 (3 s) and 1.0 V vs. RHE (3 s). The dotted line in (e) indicates the features at E ∼ 0.7 V vs. RHE associated with the presence of grain boundaries. Error bars correspond to the standard deviations of the different parameters obtained.

The stability of Pt-JM, PtNi-6.3 and PtNi-9.5 catalysts was screened first in 0.1 M HClO₄ at 25 °C, using an accelerated stress test (AST) protocol consisting of 25 000 square wave potential cycles between 0.60 (3s) and 1.0 V vs. RHE (3s). Electrochemical characterizations were performed at beginning of test on ‘fresh’ materials, and after 1000 and 25 000 potential cycles. First, it must be noted that, under these conditions, most of the activity variations occur within the first 1000 potential cycles (see in Fig. 3i, left axis). This is rationalized from on-line ICP-MS measurements conducted during the first 1000 cycles of the AST (see Fig. S5), which show this number of cycles is enough to dissolve most of the Ni from the surface of PtNi samples. The SA values confirm the dramatic stability difference between the two PtNi materials, as the PtNi-9.5 catalyst loses 43 and 51% of its initial specific activity after 1000 and 25 000 cycles, respectively, in total contrast with the PtNi-6.3 catalyst which gains 27 and 17% of specific activity, resulting in a 2.7-fold MA enhancement retention over the Pt-JM benchmark. Fig. S6 gives a breakdown of the ORR activity values in terms of MA and ECSA evolutions, showing the SA increase for PtNi-6.3 after AST induces an increase in MA at near constant ECSA. In contrast, both SA and MA decreased for PtNi-9.5, and Pt-JM only underwent a loss of ECSA. In situ WAXS was performed during the first 5000 cycles of the AST in a separated experiment, and the extracted microstrain values are displayed in Fig. 3i (right axis). Clearly, the trends in microstrain determined during electrochemical activation in Fig. 3c continue during the AST, with a singular increase in the case of PtNi-6.3. Considering the global strain relaxation and crystallite size increase observed for all catalysts during AST (see in Fig. S7) which have detrimental impact on ORR activity, this singular simultaneous increase in microstrain and ORR activity during AST for PtNi-6.3 confirms the important role of lattice distortion in controlling both initial and long-term activity in liquid electrolyte at ambient temperature.

Catalyst powders for Pt-JM, PtNi-6.3 and PtNi-9.5 were integrated into membrane electrode assemblies (MEAs) featuring a Pt loading of ∼0.2 mg_Pt cm⁻² at the cathode and by using Nafion® NRE-212 membranes and commercial 0.5 mg_Pt cm⁻² gas diffusion electrodes (GDEs, Baltic Fuel Cells) at the anode (see more details in the SI). Both the microstructure and the performance of the MEAs were investigated in PEMFC by means of operando WAXS and dedicated fuel cell tests, respectively, in separate experiments (see SI). Note an additional Pt/C benchmark from Tanaka (Pt-TKK) featuring a Pt loading of 20 wt% on carbon was used for fair comparison with the ∼17.5 wt% PtNi materials (see Table S1) in terms of catalyst layer thicknesses, but the following conclusions are qualitatively similar when compared to the 40 wt% Pt-JM, as shown in Fig. S8. The MEA break-in (activation) in PEMFC consisted of voltage cycling (10 s at open circuit voltage followed by 1 min at 0.1 V) for a duration of 1 h. The PEMFC was operated at 80 °C and 100% RH under H₂/air at anode/cathode, respectively. In agreement with the literature,^8,9,31 both MEA manufacturing (sonication of the catalyst ink, hot pressing etc.) and electrochemical activation in PEMFC triggered lattice strain relaxation (Fig. 4a), crystallite growth (Fig. 4b) and microstrain content variations (Fig. 4c) in the catalyst layer compared to the as-synthesized material. Moreover, the trends obtained in PEMFC qualitatively agree with the results previously obtained in 0.1 M HClO₄ electrolyte at 25 °C, except the amplitude variations of the different parameters being overall larger in the fuel cell (square vs. star markers after activation in Fig. 4a–c), especially for the bimetallic catalysts. Strikingly, the singular microstrain retention observed in liquid electrolyte for the PtNi-6.3 catalyst is also observed in PEMFC, as its value even slightly increases during the MEA manufacturing and break-in steps. The ECSA of the activated cathodes was measured using the desorption charge of underpotentially deposited hydrogen (H_UPD) during cyclic voltammetry experiments in H₂/N₂ configuration. Note that the use of this method is highly controversial for bimetallic^34–36 and surface lattice-distorted³² systems, but was preferred to the CO_ads stripping method in PEMFC for safety reasons. Thus, the reader is invited to consider the ECSA values reported in the insert of Fig. 4d as indicative only. It is important to note, however, that by simply considering the pronounced crystallite size growth observed in PEMFC after MEA manufacturing and activation, the ECSA decay of 24%, 7% and 25% for Pt-JM, PtNi-6.3 and PtNi-9.5, respectively, are expected compared to the values measured in a liquid environment (supposing spherical shape for the crystallites). This supports why PtNi-6.3 shows ECSA comparable to Pt-JM benchmark in PEMFC.

Fig. 4 MEA manufacturing, activation, activity and stability in PEMFC. Evolution of the (a) lattice constant, (b) crystallite size and (c) microstrain from values measured ex situ on catalyst powders to values measured in PEMFC operando (square symbols) before and after the electrochemical activation. The star symbols in (a–c) represent the values obtained after activation in thin-film configuration in liquid electrolyte for comparison. (d) Cyclic voltammograms under nitrogen-fed cathode and associated ECSA values derived from integration of the hydrogen desorption charge; (e) Tafel plots for the ORR under air-fed cathode; (f) Tafel plot for the ORR under O₂-fed cathode as a function of the fuel cell current density corrected from H₂ crossover; (g) ECSA values derived from integration of the hydrogen desorption charge before (blue) and after AST (red); (h) surface-normalized activity at 0.90 V before (blue) and after AST (red); (i) mass-normalized activity at 0.9 V before (blue) and after AST (red). The cell was operated at 80 °C and 100% RH, and the voltage is corrected from the cell high frequency resistance measured at each point. Nafion® NRE-212 membranes were used and the Pt loadings for the electrodes were 0.5/0.2 mg_Pt cm⁻² for anode/cathode, respectively. The AST comprised 5000 (3 s–3 s) square wave potential cycles between 0.60 and 1.0 V cell voltage under H₂/N₂, 55% RH, and at T = 95 °C. The activity and stability of an additional 20 wt% Pt/C benchmark from Tanaka (Pt-TKK) was also investigated for better comparison with the ∼17 wt% PtNi/C materials.

The performance of the catalysts was examined in PEMFC under both air- and O₂-fed cathode for ‘practical’ and kinetics ORR activity measurements, respectively. Under air-fed conditions, the Tafel plots in Fig. 4e show a higher voltage for PtNi-6.3 over all the current density range (see also linear scale in Fig. S9). Quantitatively, PtNi-6.3 material shows near 1.3-fold current density enhancement at 0.6 V compared to both Pt/C-TKK and PtNi-9.5, or ∼200 mV voltage gain at 1 A cm_MEA⁻². The MA and SA for the ORR were extracted at 0.9 V from the polarization curves recorded in O₂-fed cathode plotted as a function of H₂ crossover-free current density as suggested by Gasteiger et al.³⁷ The results (fresh catalysts in Fig. 4h and i) indicate a near 2.9-fold and 2.4-fold enhancement in MA and SA for both PtNi-6.3 compared to the Pt/C-TKK, in very close agreement with RDE technique in liquid electrolyte at 0.95 V vs. RHE vs. Pt-JM.

The stability of the catalysts was evaluated in the PEMFC during a short AST comprising 5000 (3 s–3 s) square wave potential cycles between 0.60 and 1.0 V cell voltage under H₂/N₂, 55% RH at 95 °C. The polarization curves recorded after the AST with air-fed cathode (Fig. S9) show that PtNi-6.3 keeps a 1.2-fold current density enhancement compared to Pt-TKK as well as 130 mV gain at 1 A cm_MEA⁻². The evolutions of the ECSA, SA and MA in Fig. 4g–i, show the highest values are maintained for the PtNi-6.3 catalyst. Notably, the singular SA increase for PtNi-6.3 observed after AST in liquid electrolyte is also observed in PEMFC.

Finally, the electrocatalytic performance and stability for the ORR observed in a PEMFC single cell are interpreted in the light of the results of in situ and operando WAXS as well as post mortem STEM-HAADF/X-EDS performed during and after AST, respectively. The same short AST (5000 cycles, 3s–3s) was performed in liquid electrolyte at ambient temperature for comparison, and the results are displayed in Fig. 5. The lattice constant evolution for the different catalysts in Fig. 5a–c clearly indicates strain relaxation for the PtNi materials (while the lattice of pure Pt benchmark remains unchanged), likely caused by Ni dissolution as previously detected by on-line ICP-MS. Clearly, the strain relaxation is more pronounced in PEMFC conditions compared to that occurring in liquid electrolyte, and the lattice constant after 5000 cycles approaches or almost equals that of pure Pt for PtNi-6.3 or PtNi-9.5, respectively. Consequently, only little or no contribution from alloying through the strain effect is expected in the ORR activity after AST. Additionally, pronounced crystallite size increase is observed for the PtNi materials compared to the Pt benchmark in Fig. 5d–f. Crystallite growth during PEMFC catalyst operation has for long time been associated mainly with Pt dissolution and redeposition (Ostwald electrochemical ripening process), but particle aggregation and coalescence have been recently highlighted as the most influent in operating PEMFC.³⁸ In fact, and as confirmed by post mortem STEM-HAADF in Fig. 5g and h, the main cause of ECSA loss in PEMFC appears to be the collapse of the hollow structure, which does not occur in liquid electrolyte at ambient temperature. Thus, contrary to conventional Pt-bimetallic catalysts the deactivation of this particular catalyst seems to be driven mostly by morphological changes (and associated ECSA loss) rather than dealloying and the associated loss of alloying effects. Moreover, X-EDS linescans in Fig. 5g and h further confirm the depletion of Ni from PtNi-6.3 surface after both ASTs, with the thickness of the Pt-rich shell increasing from ∼1.1 nm (Fig. S2) before AST to ∼1.2 nm and to ∼1.6 nm after AST in liquid electrolyte and in PEMFC, respectively. This further rule out a possible contribution of the ligand effect from subsurface Ni, and leave the increasing microstrain values in Fig. 5i–k as the only parameter able to explain the increase of intrinsic ORR activity despite the loss of alloying effects.

Fig. 5 Microstructural and morphological evolution during accelerated stress tests in 0.1 M HClO₄ at 25 °C vs. in PEMFC at 95 °C from in situ/operando WAXS and post mortem STEM-HAADF/X-EDS. (a)–(c) Lattice constant; (d)–(f) crystallite size and (i)–(k) microstrain values for Pt-JM, PtNi-6.3 and PtNi-9.5 catalyst as a function of AST cycle number. The AST consisted of 5000 3s–3s square wave potential cycles between 0.60 and 1.0 V (either vs. RHE or cell voltage) in N₂-saturated 0.1 M HClO₄ at 25 °C (filled markers) or in PEMFC operated in H₂/N₂, 55% RH at 95 °C (hollow markers). STEM-HAADF/X-EDS images of the PtNi-6.3 catalyst after AST in (g) 0.1 M HClO₄ and in (h) PEMFC.

Conclusions

This study employed a large set of ex situ, in situ, operando and post mortem characterization techniques to highlight the fundamental role of local lattice distortion over global strain in governing the activity and stability of bimetallic Pt catalysts in the PEMFC cathode. Notably, by probing the microstructure of the catalysts throughout the key steps of their development and integration into model and practical electrodes, the results revealed complex but ubiquitous lattice distortion dynamics from early to long-term operation, mostly caused by transition metal dissolution. However, among the variety of catalyst structures investigated, the increase of lattice distortion was found to possibly compensate for the unavoidable loss of lattice strain effects through dealloying, and to positively impact ORR performance in the operating PEMFC.

These insights suggest an alternative design paradigm focused on inducing local lattice distortions rather than global strain, thereby decoupling catalyst performance from the retention of transition metals. This approach paves the way for the development of more robust and efficient fuel cell catalysts. Future better understanding and identification of stable source(s) of lattice distortion, as well as strategies to stabilize particle morphology³⁹ could certainly further enhance the interest and practical relevance of such an approach for more performing and reliable fuel cell systems.

Author contributions

Material synthesis. A. G. and A. T. S. Ex situ/in situ/operando X-ray scattering experiments. Conception: R. C. and J. D.; execution: A. G., M. E., M. S., A. T. S., R. B., V. V., M. M., R. C., and J. D.; analysis: R. C. RDE measurements. Execution and analysis: A. G. Electrochemical on-line ICP-MS experiments. Conception: C. A. C.-R.; execution: C. A. C.-R. and A. G.; analysis: C. A. C.-R. Fuel cell testing. Conception: M. D. and F. L. Execution and analysis: A. G. and R. C. Electron microscopy. Execution and analysis: M. S. and C. R.

R. C., J. D. and D. J. supervised the work. R. C. and A. G. wrote the manuscript with input from all authors.

Conflicts of interest

The authors declare no competing interests.

Data availability

All data supporting the findings in this study are available within the paper and the supplementary information (SI). Supplementary information: Methods, Fig. S1–S9, eqn (S1)–(S10), Table S1. See DOI: https://doi.org/10.1039/d5ee04563k.

Acknowledgements

The authors thank the ESRF for providing beamtime at ID31 and especially H. Isern and F. Russello for their support at the beamline. The authors acknowledge the Balard Analysis and Characterization Platform (PAC Balard) facilities for technical support. R. C. gratefully acknowledge financial support from the French National Research Agency through the HOLYCAT project (grant number no. ANR-22-CE05-0007) and R. C. and D. J. acknowledge PEMFC95 project funded by “France 2030” operated by the French National Research Agency (grant ANR-22-PEHY-0005).

References

1. Y.-J. Wang, et al., Carbon-Supported Pt-Based Alloy Electrocatalysts for the Oxygen Reduction Reaction in Polymer Electrolyte Membrane Fuel Cells: Particle Size, Shape, and Composition Manipulation and Their Impact to Activity, Chem. Rev., 2015, 115, 3433–3467.
2. M. Shao, Q. Chang, J.-P. Dodelet and R. Chenitz, Recent Advances in Electrocatalysts for Oxygen Reduction Reaction, Chem. Rev., 2016, 116, 3594–3657.
3. H. Mistry, A. S. Varela, S. Kühl, P. Strasser and B. R. Cuenya, Nanostructured electrocatalysts with tunable activity and selectivity, Nat. Rev. Mater., 2016, 1, 16009.
4. A. Kulkarni, S. Siahrostami, A. Patel and J. K. Nørskov, Understanding Catalytic Activity Trends in the Oxygen Reduction Reaction, Chem. Rev., 2018, 118, 2302–2312.
5. Z. Zhao, et al., Theory-guided design of catalytic materials using scaling relationships and reactivity descriptors, Nat. Rev. Mater., 2019, 4, 792–804.
6. Q. Yun, et al., Recent Progress on Phase Engineering of Nanomaterials, Chem. Rev., 2023, 123, 13489–13692.
7. C. Lim, A. R. Fairhurst, B. J. Ransom, D. Haering and V. R. Stamenkovic, Role of Transition Metals in Pt Alloy Catalysts for the Oxygen Reduction Reaction, ACS Catal., 2023, 13, 14874–14893.
8. R. Chattot, et al., Break-In Bad: On the Conditioning of Fuel Cell Nanoalloy Catalysts, ACS Catal., 2022, 12, 15675–15685.
9. M. Ronovský, et al., Origins of Nanoalloy Catalysts Degradation during Membrane Electrode Assembly Fabrication, ACS Energy Lett., 2024, 9, 5251–5258.
10. C. Roiron, V. Martin, K. Kumar, L. Dubau and F. Maillard, Assessing Pt and Ni dissolution mechanism and kinetics of shape-controlled oxygen reduction nanocatalysts, Electrochim. Acta, 2024, 477, 143760.
11. C. Roiron, et al., Tracking the Early-life of PtNi/C Shape-Controlled Catalysts upon their Integration in PEMFC, J. Electrochem. Soc., 2024, 171, 114504.
12. B. Han, et al., Record activity and stability of dealloyed bimetallic catalysts for proton exchange membrane fuel cells, Energy Environ. Sci., 2015, 8, 258–266.
13. P. Strasser, et al., Lattice-strain control of the activity in dealloyed core-shell fuel cell catalysts, Nat. Chem., 2010, 2, 454–460.
14. E. M. Hofer and H. E. Hintermann, Correlation between the catalytic activity of Raney nickel and its structure, Trans. Faraday Soc., 1964, 60, 1457.
15. N. M. Markovic and P. N. Ross, Surface science studies of model fuel cell electrocatalysts, Surf. Sci. Rep., 2002, 45, 117–229.
16. V. R. Stamenkovic, et al., Improved oxygen reduction activity on Pt3Ni(111) via increased surface site availability, Science, 2007, 315, 493–497.
17. F. Calle-Vallejo, et al., Finding optimal surface sites on heterogeneous catalysts by counting nearest neighbors, Science, 2015, 350, 185–189.
18. F. Calle-vallejo, M. D. Pohl and A. S. Bandarenka, Quantitative Coordination-Activity Relations for the Design of Enhanced Pt catalysts for CO Electro-Oxidation, 2017.
19. F. Calle-Vallejo, et al., Why conclusions from platinum model surfaces do not necessarily lead to enhanced nanoparticle catalysts for the oxygen reduction reaction, Chem. Sci., 2017, 8, 2283–2289.
20. L. Dubau, et al., Defects do Catalysis: CO Monolayer Oxidation and Oxygen Reduction Reaction on Hollow PtNi/C Nanoparticles, ACS Catal., 2016, 6, 4673–4684.
21. R. Chattot, et al., Beyond Strain and Ligand Effects: Microstrain-Induced Enhancement of the Oxygen Reduction Reaction Kinetics on Various PtNi/C Nanostructures, ACS Catal., 2017, 7, 398–408.
22. L. Dubau, J. Nelayah, T. Asset, R. Chattot and F. Maillard, Implementing structural disorder as a promising direction for improving the stability of PtNi/C nanoparticles, ACS Catal., 2017, 7(4), 3072–3081.
23. O. Le Bacq, et al., Effect of Atomic Vacancies on the Structure and the Electrocatalytic Activity of Pt-rich/C Nanoparticles: A Combined Experimental and Density Functional Theory Study, ChemCatChem, 2017, 9, 2324–2338.
24. T. Asset, et al., Elucidating the Mechanisms Driving the Aging of Porous Hollow PtNi/C Nanoparticles by Means of COads Stripping, ACS Appl. Mater. Interfaces, 2017, 9, 25298–25307.
25. R. Chattot, et al., Surface distortion as a unifying concept and descriptor in oxygen reduction reaction electrocatalysis, Nat. Mater., 2018, 17, 827–833.
26. M. Li, et al., Ultrafine jagged platinum nanowires enable ultrahigh mass activity for the oxygen reduction reaction, Science, 2016, 354, 1414–1419.
27. R. Chattot, et al., Atomic-Scale Snapshots of the Formation and Growth of Hollow PtNi/C Nanocatalysts, Nano Lett., 2017, 17, 2447–2453.
28. M. Gatalo, et al., Importance of Chemical Activation and the Effect of Low Operation Voltage on the Performance of Pt-Alloy Fuel Cell Electrocatalysts, ACS Appl. Energy Mater., 2022, 5, 8862–8877.
29. C. A. Campos-Roldán, et al., Structure Dynamics of Carbon-Supported Platinum-Neodymium Nanoalloys during the Oxygen Reduction Reaction, ACS Catal., 2023, 7417–7427,  DOI:10.1021/acscatal.3c00371.
30. C. A. Campos-Roldán, et al., Stabilization of Carbon-Supported Platinum–Rare Earth Nanoalloys during Electrochemical Activation, ACS Catal., 2023, 13, 13319–13324.
31. M. Kostelec, M. Gatalo and N. Hodnik, Fundamental and Practical Aspects of Break-In/Conditioning of Proton Exchange Membrane Fuel Cells, Chem. Rec., 2024, 202400114.
32. R. Chattot, et al., Disclosing Pt-Bimetallic Alloy Nanoparticle Surface Lattice Distortion with Electrochemical Probes, ACS Energy Lett., 2020, 5, 162–169.
33. R. Chattot, et al., Building Practical Descriptors for Defect Engineering of Electrocatalytic Materials, ACS Catal., 2020, 10, 9046–9056.
34. D. F. van der Vliet, et al., Unique Electrochemical Adsorption Properties of Pt-Skin Surfaces, Angew. Chem., 2012, 124, 3193–3196.
35. H. Schulenburg, J. Durst, E. Müller, A. Wokaun and G. G. Scherer, Real surface area measurements of Pt3Co/C catalysts, J. Electroanal. Chem., 2010, 642, 52–60.
36. S. Rudi, C. Cui, L. Gan and P. Strasser, Comparative Study of the Electrocatalytically Active Surface Areas (ECSAs) of Pt Alloy Nanoparticles Evaluated by Hupdand CO-stripping voltammetry, Electrocatalysis, 2014, 5, 408–418.
37. H. A. Gasteiger, S. S. Kocha, B. Sompalli and F. T. Wagner, Activity benchmarks and requirements for Pt, Pt-alloy, and non-Pt oxygen reduction catalysts for PEMFCs, Appl. Catal., B, 2005, 56, 9–35.
38. I. Martens, R. Chattot and J. Drnec, Decoupling catalyst aggregation, ripening, and coalescence processes inside operating fuel cells, J. Power Sources, 2022, 521, 230851.
39. V. Beermann, et al., Rh-Doped Pt–Ni Octahedral Nanoparticles: Understanding the Correlation between Elemental Distribution, Oxygen Reduction Reaction, and Shape Stability, Nano Lett., 2016, 16, 1719–1725.

---