Article https://doi.org/10.

1038/s41467-025-59450-6

Sustainable and cost-efficient hydrogen

production using platinum clusters at
minimal loading
Received: 27 August 2024 Hongliang Zeng1,6, Zheng Chen 2,6, Qiu Jiang 1,3 , Qingtian Zhong1, Yuan Ji1,
Yizhen Chen4, Jiawei Li 1, Chunxiao Liu 1, Runhao Zhang1, Jialin Tang1,
Accepted: 23 April 2025
Xiaoxia Xiong1, Zhongyue Zhang1, Zhaoyang Chen1, Yizhou Dai1, Chengbo Li1,
Yinfang Chen1, Donghao Zhao1, Xu Li 1, Tingting Zheng 1, Xin Xu 2,5 &
Chuan Xia 1,3
Check for updates
1234567890():,;
1234567890():,;

Proton exchange membrane water electrolysis stands as a promising tech-

nology for sustainable hydrogen production, although its viability hinges on
minimizing platinum (Pt) usage without sacrificing catalytic efficiency. Central
to this challenge is enhancing the intrinsic activity of Pt while ensuring the
stability of the catalyst. We herein present a Mo2TiC2 MXene-supported Pt
nanocluster catalyst (Mo2TiC2-PtNC) that requires a minimal Pt content
(36 μg cm−2) to function, yet remains highly active and stable. Operando
spectroscopy and theoretical simulation provide evidence for anomalous
charge transfer from the MXene substrate to PtNC, thus generating highly
efficient electron-rich Pt sites for robust hydrogen evolution. When incorpo-
rated into a proton exchange membrane electrolyzer, the catalyst affords
more than 8700 h at 200 mA cm−2 under ambient temperature with a decay
rate of just 2.2 μV h−1. All the performance metrics of the present Mo2TiC2-PtNC
catalysts are on par with or even surpass those of current hydrogen evolution
electrocatalysts under identical operation conditions, thereby challenging the
monopoly of high-loading Pt/C-20% in the current electrolyzer design.

The escalating importance of renewable green hydrogen as an energy costs. The DOE’s 2026 target mandates a total Pt-group metal (PGM)
vector represents an undeniable facet of the evolving energy loading below 0.5 mg cm−2 for PEM devices6. Currently, the Pt content
landscape1,2. A pressing target in the U.S. Department of Energy’s in the cathode alone falls within the range of 0.4–0.6 mg cm−2 7. This
(DOE’s) Energy Earthshots initiative is to reduce the cost of clean highlights the urgent need to implement low Pt-group metal cathodes,
hydrogen to $1 per kilogram in a decade, known as the “111” target3. thereby reserving the PGM budget for iridium (Ir) incorporation into
Central to realizing this vision is the application of proton exchange the anode (typically more than 5 times the Pt content)7–12. Despite
membrane (PEM) water electrolysis technology, noted for its low progress in developing nonprecious materials for cost efficiency13–15, Pt
electrical resistance, high current density, high H2 purity, and swift remains the material of choice for facilitating the acidic hydrogen
device response4,5. However, the viability of its commercialization evolution reaction (HER) in current commercial PEM water electro-
within the hydrogen market is contingent upon the reduction in capital lyzers, owing to its unmatched intrinsic activity and stability16–19.

1
School of Materials and Energy, University of Electronic Science and Technology of China, Chengdu, P. R. China. 2Department of Chemistry, Shanghai Key
Laboratory of Molecular Catalysis and Innovative Materials, MOE Key Laboratory of Computational Physical Sciences, Fudan University, Shanghai, P. R. China.
3
Yangtze Delta Region Institute (Huzhou), University of Electronic Science and Technology of China, Huzhou, Zhejiang, P. R. China. 4Department of Chemical
Engineering, University of California, Davis, CA, USA. 5Hefei National Laboratory, Hefei, P. R. China. 6These authors contributed equally: Hongliang Zeng,
Zheng Chen. e-mail: [email protected]; [email protected]; [email protected]

Nature Communications | (2025)16:4314 1

Article https://doi.org/10.1038/s41467-025-59450-6

The HER on metal surfaces such as Pt involves protons from a selectively removed from the parent Mo2TiAlC2 powder (illustrated in
solution combined with electrons at an electrode. This HER reaction Supplementary Figs. 1–3). Following this, the structure was inter-
proceeds in two main steps: forming chemisorbed hydrogen atoms on calated with the organic molecule tetrabutylammonium hydroxide
the electrode, followed by H2 gas evolution. This simple mechanism (TBAOH), thereby yielding the few-layered Mo2TiC2 MXene29,31. The
allows enhancing the HER through two main strategies: maximizing process was completed with a thorough washing step to remove any
the intrinsic catalytic efficiency and optimizing the exposure of active residual TBAOH. The layered structure of Mo2TiC2 was evident from
sites on the catalyst. Nørskov et al. showed the intrinsic HER activity in the scanning electron microscopy (SEM) image (Fig. 1b). The interlayer
a “volcano plot” using the binding free energy (ΔGH) of hydrogen spacing of Mo2TiC2 was calculated to be 1.8 nm (inset of Fig. 1b) from
(*H)20. They found the highest activity when ΔGH was near zero, transmission electron microscopy (TEM), which corresponds to the
demonstrating that the optimal electrocatalytic sites for the HER (002) plane from the X-ray diffraction (XRD) results (Fig. 1f). Finally, by
should bind *H slightly weaker than metals such as Pd, Rh, or Pt. This employing Mo2TiC2 as a support, we facilitated the deposition of PtNC
could correspond to a more negative chemical state than metallic via a hydrothermal protocol, incorporating chloroplatinic acid hex-
states21. On the other hand, a surface-to-volume ratio can be achieved, ahydrate (H2PtCl6·6H2O) and Mo2TiC2 in the process (see Methods for
for instance, by downsizing bulk catalysts to nanoclusters (NCs) and additional details). Quantitative analysis via inductively coupled
even to isolated single atoms (SAs). This approach is particularly plasma optical emission spectroscopy (ICP-OES) indicated that the as-
attractive because it leverages nearly all available platinum atoms, prepared Mo2TiC2-PtNC catalyst had a Pt loading of 3.6 ± 0.1 wt%
thereby optimizing their effectiveness and enhancing the overall (Supplementary Table 1).
efficiency22–24. However, this strategy also exposes more low- After the hydrothermal reaction, TEM and high-angle annular
coordinate Pt atoms, which have been shown to strongly adsorb *H, dark-field scanning TEM (HAADF-STEM) images (Fig. 1c–e) confirmed
consequently reducing the intrinsic activity of Pt. Balancing the metal that the PtNC were homogeneously sized, with an average diameter of
particle size and maintaining the desired binding energy are critical for approximately 1.7 nm (Fig. 1d). Furthermore, the results from the high-
developing effective HER catalysts. Moreover, stability is another key resolution HAADF-STEM image revealed that the exposed crystal facet
factor that must be considered in the utilization of Pt for the HER, as of the experimentally prepared PtNC is the (111) facet (Fig. 1e). The two-
extended electrolysis or a higher current increases the likelihood of Pt dimensional crystal structure of Mo2TiC2 was also well maintained
agglomeration and thus the degradation of active sites. (Fig. 1f), suggesting that the incorporation of the Pt precursor into the
Given that the electronic properties of Pt can be fine-tuned by precursor solution had no discernible effect on the morphology of the
selective bonding with proper supports25, and that Pt agglomeration MXene structure. For comparison, we also synthesized MXene-
can be mitigated through spatial confinement or geometric supported Pt single atoms (Mo2TiC2-PtSA, 2.7 wt% Pt) and reduced
shielding26, we envisioned that the intrinsic activity and stability of Pt graphene oxide (rGO)-supported PtNC (rGO-PtNC, Pt size ~1.5 nm,
nanoclusters (PtNC) could be concurrently modulated by designing 4.1 wt% Pt) as controls. The HAADF-STEM images, in conjunction with
supported Pt islands surrounded by functional groups that have high energy-dispersive spectroscopy (EDS) elemental mapping, unveiled a
diffusion barriers to isolate these Pt islands. MXenes, first reported by uniform dispersion of PtSA or PtNC across the support structure for the
Gogotsi et al. in 201027–29, are a class of well-known two-dimensional controls (Supplementary Figs. 4–5). Multiple spectra, including XRD
transition metal carbides or nitrides with layered structures that may and Raman spectroscopy, revealed that both Mo2TiC2-PtSA and
serve as an ideal model to validate our hypothesis. In MXenes, early Mo2TiC2-PtNC exhibited peak patterns similar to those of Mo2TiC2
transition metals (M layers) are sandwiched between conductive car- MXene, indicating the structural stability of the MXene substrate
bon layers, with abundant surface termination groups, such as O and (Supplementary Figs. 6–7). Additionally, PtSA and PtNC on the support
OH, on the outer M layers30. Pursuing this line of inquiry, we synthe- were distinguished using CO adsorption diffuse reflectance infrared
sized a highly stable and efficient electrocatalyst composed of PtNC Fourier transform spectroscopy (DRIFTS; see details in Methods). The
supported on MXene (Mo2TiC2-PtNC) for use in PEM water electrolysis. spectra primarily displayed a strong band at 2105 cm−1 for Mo2TiC2-
From our experimental and theoretical evidence, we found anomalous PtSA and 2051 cm−1 for Mo2TiC2-PtNC, further substantiating the dis-
charge transfer from the MXene substrate to PtNC, which generated tinctions in the electronic structure between the two samples (Sup-
highly efficient electron-rich Pt sites for robust hydrogen evolution. plementary Fig. 8)32.
The as-synthesized Mo2TiC2-PtNC matches the performance of com- To reveal the electronic structure of the Pt species within the
mercial Pt/C-20%, requiring only an overpotential of 13 ± 3.6 mV to catalysts, we employed X-ray absorption fine structure (XAFS) and
reach 10 mA cm−2 while maintaining stability at this current density for X-ray photoelectron spectroscopy (XPS) analyses. The normalized
more than 280 h. Impressively, our Mo2TiC2-PtNC-based device show- X-ray absorption near-edge structure (XANES) depicted in Fig. 1g
cased a robust capability to electrolyze pure water splitting for more indicated that the white line intensity of the Mo2TiC2-PtNC catalysts was
than 8700 h at 200 mA cm−2 under ambient temperature, when cou- even lower than that of the bulk Pt foil, suggesting a possibly reduced
pled with a commercial oxygen evolution catalyst (IrO2) in a PEM oxidation state. Figure 1h shows the k3-weighted Fourier transform of
electrolyzer. Notably, the device delivered a high current density of the extended X-ray absorption fine structure (EXAFS) at the Pt L3-edge.
1 A cm−2 at a low average applied voltage of ~1.65 V and maintained We observed a dominant peak at 1.72 Å for Mo2TiC2-PtSA, which cor-
steady operation for over 4800 h under real-world operating condi- responded to the Pt-O bond, and no Pt-Pt bonds were observed. This
tions (80 °C, 1 bar). Notably, the Pt loading on the cathode in our result confirmed that Pt atoms were dispersed as single sites in
assembled PEM device is merely 36 μg cm−2, far less than one-tenth of Mo2TiC2-PtSA, and the evident Pt-O coordination was consistent with
the loading in commercial design. Our Mo2TiC2-PtNC catalyst holds previous reports33,34. Since the first shell scattering peak in the Pt L3-
great promise as a viable alternative to commercial Pt/C-20%, repre- edge of Mo₂TiC₂-PtNC is located within the same distance region as the
senting a prime candidate for satisfying the operational demands of Pt-Pt scattering peak of Pt foil (R = 1.5–3.2 Å, which corresponds to
next-generation PEM devices. metal-metal bonding)35, and the difference in the main peak position is
less than 0.1 Å36,37, distinguishing Pt-Mo and Pt-Pt coordination signals
Results and discussion solely through observation of the EXAFS spectrum is challenging. We
We adopted a straightforward hydrothermal method to synthesize the subsequently performed fitting of the L3-edge EXAFS of the Mo2TiC2-
Mo2TiC2-PtNC catalysts (see details in the Methods and Fig. 1a). The first PtNC and Pt foil. Compared with those of the Pt foil (Supplementary
step involved the synthesis of multi-layered Mo2TiC2 (denoted as Fig. 9 and Supplementary Table 2), the EXAFS fitting results of
Mo2TiC2-M) via the HF etching method, in which the Al layers were Mo2TiC2-PtNC (Supplementary Table 3) reveal a Pt-Mo coordination

Nature Communications | (2025)16:4314 2

Article https://doi.org/10.1038/s41467-025-59450-6

Fig. 1 | Synthesis and structural characterization of the catalyst. a Schematic of PtNC, and Pt foil. R is the interatomic distance. i High-resolution XPS spectra (Pt 4f)
the synthesis process of Mo2TiC2-PtNC. b SEM and TEM images of few-layer of Mo2TiC2-PtSA, Mo2TiC2-PtNC, rGO-PtNC and Pt foil. (The circles represent raw
Mo2TiC2. c TEM image of Mo2TiC2-PtNC. d Size distribution of the Pt nanoclusters. data, and the lines represent fitting data and base line.) j DEMS measurements of
e HADDF-STEM image of the nanoclusters. f XRD spectra of Mo2TiAlC2, Mo2TiC2, Mo2TiC2-PtSA, Mo2TiC2-PtNC, rGO-PtNC and Pt/C-20%, the onset potentials were
and Mo2TiC2-PtNC. g Normalized XANES spectra at the Pt L3-edge of PtO2, Mo2TiC2- determined at points where the signal-to-noise ratio exceeded 5. Source data are
PtSA, Mo2TiC2-PtNC, rGO-PtNC and Pt foil. h k3-weighted Fourier transform of EXAFS provided as a Source Data file.
spectra derived from the EXAFS spectra of PtO2, Mo2TiC2-PtSA, Mo2TiC2-PtNC, rGO-

Nature Communications | (2025)16:4314 3

Article https://doi.org/10.1038/s41467-025-59450-6

signal at 2.69 Å, which aligns well with the fitting results of Pt-Mo consistent with that of Pt/C-20%, following the same mechanism
coordination in other articles35–37. Notably, the absence of a Pt-O signal referred to as the Tafel step. To further describe the elementary
in the Mo2TiC2-PtNC spectrum suggested that the PtNC on the MXene reaction in the HER process on Mo2TiC2-PtNC, we conducted operando
surface was directly bonded to the MXene lattice rather than to surface electrochemical impedance spectroscopy (EIS) experiments (Fig. 2e
oxygen functionalities. Interestingly, the high-resolution Pt 4f XPS and Supplementary Figs. 15, 16). Specifically, we observed that
spectrum revealed that the Mo2TiC2-PtNC catalyst featured a negatively Mo2TiC2-PtNC exhibited only one phase angle, indicative of the same
shifted binding energy (70.5 eV and 73.8 eV for the Pt 4f 7/2 and 4f 5/2 Volmer-Tafel kinetic process that characterizes commercial Pt/C-20%.
doublets, respectively) in comparison to that of the bulk Pt foil (71 eV This finding stands in stark contrast to the two-phase angles typically
and 74.3 eV)38, in good agreement with the XANES analysis. To validate observed in the Volmer-Heyrovsky kinetic process46,47.
the robustness of our data, we also calibrated our XPS spectra using Si We further quantitatively analyzed HER activity in terms of mass
2p lines as the internal standard. These results were consistent with activity and turnover frequency (TOF, normalized to the active site
those obtained using the C 1s calibration method (Supplementary density) (Fig. 2f). Specifically, at an overpotential of 60 mV, the mass
Fig. 10). This phenomenon is intriguing, as Pt on most carbon or oxide activity of Mo2TiC2-PtNC was estimated to be 3.3 ± 1.31 A mg−1, which
substrates generally exists in an oxidized or metallic state (Fig. 1i and was substantially greater than that of Pt/C-20% (0.72 ± 0.19 A mg−1) and
Supplementary Fig. 11)39–41. As expected, Pt in Mo2TiC2-PtSA shows a Mo2TiC2-PtSA (0.35 ± 0.07 A mg−1). At the same overpotential, a sig-
more positively shifted binding energy than does bulk Pt, which aligns nificant difference in the TOF was observed. Specifically, the TOF of
well with previous reports18,34,42–45. These findings underscore the size Mo2TiC2-PtNC was found to be 9.45 ± 3.71 s−1. This value not only out-
effect on the electronic state of Pt. In contrast, rGO-PtNC displays more performs the benchmark Pt/C-20% (1.82 ± 0.48 s−1) but also surpasses
positive binding energy than Pt foil does, highlighting the unique role Mo2TiC2-PtSA (0.36 ± 0.07 s−1). Notably, the TOF represents a lower
of the MXene substrate. Taken together, the above results demon- bound for the true activity of Mo2TiC2-PtNC because a subset of the Pt
strate that the direct bonding of PtNC with the MXene lattice results in atoms remains inaccessible within the bulk, coupled with the con-
anomalous electron transfer to Pt atoms, creating an electron-rich Pt straints posed by mass transport limitations48.
surface. Furthermore, the high-resolution Mo 3d XPS spectrum shows In addition to the above performance, maintaining robust elec-
that Mo2TiC2-PtNC displays more pronounced high-valence Mo signals trolytic stability is the key. We further evaluated the stability of
than Mo2TiC2 does (Supplementary Fig. 12), further confirming elec- Mo2TiC2-PtNC and found that it maintained a hydrogen Faradaic effi-
tron transfer from Mo to PtNC. This shifts the d-band center of Pt ciency (FE) close to 100% across different current densities (Supple-
downwards from −3.62 (Pt foil) to −3.77 eV (Mo2TiC2-PtNC) (Supple- mentary Fig. 17). Additionally, stable hydrogen production was
mentary Fig. 13), resulting in closer thermal-neutral binding of *H achieved through electrolysis at 10 mA cm−2 for over 280 h (Fig. 2g).
(ΔGH = 0). Furthermore, we employed operando differential electro- Accelerated stability tests revealed virtually no decay after 10,000
chemical mass spectrometry (DEMS) to assess the HER capabilities of cycles (Supplementary Fig. 18). Structural characterization via XRD,
the different catalysts (Fig. 1j). Impressively, the Mo2TiC2-PtNC catalyst SEM, XPS, TEM and XAFS (XANES and EXAFS) further demonstrated
manifested a much lower onset potential than did its counterparts, that the structure was almost unchanged after the durability test
drawing closer to the commercial benchmark Pt/C-20% catalyst. Sim- (Supplementary Figs. 19, 20). Consequently, the aforementioned
ply replacing the MXene substrate with carbon (rGO-PtNC, with similar attributes of Mo2TiC2-PtNC, including its low overpotential, reduced
Pt dimensions and loading) or decreasing the particle size from PtNC to Tafel slope, and slow decay rate, were found to surpass those of most
PtSA (Mo2TiC2-PtSA, with similar Pt loading) results in much lower HER previously reported Pt-based catalysts (Supplementary Fig. 21 and
activity than Mo2TiC2-PtNC does. This finding implies that the HER Supplementary Table 5).
activity is closely dependent on the synergistic effect between the Pt To explore the impact of the substrate on the loaded PtNC, the
metal and MXene substrates. We deduce that the intrinsic activity of HER mechanism was further analyzed utilizing the equivalent circuit
the Mo2TiC2-PtNC catalyst is significantly influenced by the unique depicted in Supplementary Fig. 22. This approach allowed for a
electron-rich electronic structure of the Pt center due to Pt-MXene detailed examination of the underlying electrochemical processes,
interactions. providing insights into the specific interactions and dynamics that
The HER performances of the different catalysts were then eval- govern the reaction kinetics. The intermediate coverage on the cata-
uated in detail using a three-electrode setup with Hg/Hg2SO4 as the lyst surface can be represented by two parallel components (Cφ and R2,
reference electrode and a graphite rod as the counter electrode in H2- which represent the hydrogen adsorption pseudo-capacitance and
saturated 0.5 M H2SO4. Prior to the experimental tests, the reference resistance, respectively) in the equivalent circuit49. As illustrated in
electrode was meticulously calibrated to a reversible hydrogen elec- Supplementary Figs. 23, 24, the adsorption charges of hydrogen on the
trode (RHE). Linear sweep voltammetry (LSV) curves revealed that the surfaces of Mo2TiC2-PtNC and rGO-PtNC were obtained through EIS
as-prepared Mo2TiC2-PtNC catalyst exhibited effective HER activity fitting. Notably, Mo2TiC2-PtNC exhibited more hydrogen adsorption
with an overpotential of only 13 ± 3.6 mV at 10 mA cm−2, similar to the charges (750 μC) than did rGO-PtNC (624 μC). This result indicates a
benchmark Pt/C-20% (14.7 ± 0.6 mV at 10 mA cm−2; Fig. 2a). To further much greater surface intermediate coverage on Mo2TiC2-PtNC, despite
compare the intrinsic HER activity of Mo2TiC2-PtNC with that of the its quite similar Pt dimensions and content to those of rGO-PtNC. Such a
benchmark material, we measured the hydrogen underpotential difference highlights the critical role of electron enrichment on the Pt
deposition (H-UPD) features to exclude the contribution of different surface of Mo2TiC2-PtNC.
electrochemical surface area (ECSA, Fig. 2b, c). In a clear demonstra- Additionally, we compared the in situ attenuated total reflection
tion of electrocatalytic performance, a comparison of the LSV curves, surface-enhanced infrared absorption spectroscopy (ATR-SEIRAS) of
normalized using H-UPD, revealed a substantial difference between Mo2TiC2-PtSA, rGO-PtNC, and Mo2TiC2-PtNC under different applied
the two materials. Specifically, Mo2TiC2-PtNC showed an intrinsic per- potentials (0.1 to −0.1 V) (Supplementary Fig. 25). Among these cata-
formance of 1.55 ± 0.4 mA cm−2ECSA (the specific value in Fig. 2c is lysts, Mo2TiC2-PtNC exhibited the highest vO-H wavenumber, indicating
1.99 mA cm−2ECSA) which was 4.8 times greater than that of the the weakest degree of hydrogen bonding in its interfacial water50,51.
benchmark Pt/C-20% of 0.32 ± 0.04 mA cm−2ECSA, measured at an This characteristic facilitates more efficient intermediate transport,
overpotential of 31 mV. Additionally, Mo2TiC2-PtNC demonstrated a thereby increasing the reaction kinetics. We further investigated the
Tafel slope of 24 mV dec−1, a value that was comparable to the Tafel effect of the substrate by comparing the activity and stability of
slope observed for Pt/C-20% (Fig. 2d). This similarity in the Tafel slopes Mo2TiC2-PtNC and rGO-PtNC. A comparison was made under identical
suggested that the kinetic rate-step process of Mo2TiC2-PtNC was conditions, with both substrates having similar particle sizes of PtNC

Nature Communications | (2025)16:4314 4

Article https://doi.org/10.1038/s41467-025-59450-6

Fig. 2 | Electrocatalytic performance of the catalysts. a The iR-corrected polar- Supplementary Table 4, respectively). b H-UPD region of Mo2TiC2-PtNC and Pt/C-
ization curves of glass carbon (GC), Mo2TiC2 (R = 3.378 Ω), Mo2TiC2-PtSA (R = 2.18 20% obtained from the cyclic voltammetry curve (Scan rate: 50 mV s−1). c ECSA-
Ω), Mo2TiC2-PtNC (R = 1.763 Ω) and Pt/C-20% (R = 1.433 Ω) acquired using a graphite normalized (H-UPD) HER polarization curves of Mo2TiC2-PtNC and Pt/C-20% in
rod as the counter electrode in 0.5 M H2SO4 (GCE electrode surface area: 1.13 cm2, 0.5 M H2SO4. d Tafel slopes of Mo2TiC2, Mo2TiC2-PtSA Mo2TiC2-PtNC and Pt/C-20%.
catalyst loading: 1 mg cm−2, rotation rate: 1600 rpm, pH value of the electrolyte e Operando EIS tests of Mo2TiC2-PtNC. f Calculated TOF and mass activity values at
(H2SO4): 0.43 ± 0.02, and all measurements were conducted at ambient tempera- an overpotential of 60 mV in 0.5 M H2SO4. g Stability test of Mo2TiC2-PtNC through
ture. All error bars in Fig. 2 represent the standard deviation. All potentials are 100% chronopotentiometry at a current density of 10 mA cm−2 in 0.5 M H2SO4. Source
iR compensated, and all non-iR corrected data and corresponding resistance values data are provided as a Source Data file.
for three independent samples are shown in Supplementary Fig. 14 and

and Pt loading. As shown in Supplementary Figs. 26, 27, the perfor- (Fig. 3b) was employed here due to the absence of a Pt-O signal for
mance of rGO-PtNC was found to be inferior, and its stability at a cur- Mo2TiC2-PtNC from experimental observations (Fig. 1h). DFT calcula-
rent density of 10 mA cm−2 clearly decreased over a period of 70 h. tions revealed that the two different models resulted in opposite
These results collectively affirm that the high structural stability of charge transfer directions (Fig. 3a, b). In contrast to the commonly
Mo2TiC2-PtNC can be attributed to the unique interaction between the observed charge transfer from the metal to the support, the latter
Pt and MXene supports. model showed a significant charge transfer reversal from the support
Density functional theory (DFT) calculations were performed to to the metal, which is consistent with the XPS observations (Fig. 1i).
elucidate the origin of the high stability and high HER activity of Thus, both the experimental and theoretical findings corroborate the
Mo2TiC2-PtNC, where unique metal-support interactions play a pivotal hypothesis that PtNC binds to outer Mo sites, which aligns with the fact
role, as indicated by comparisons with controls. Considering the that TiC2 layers are sandwiched between Mo layers in Mo2TiC230.
computational consumption and experimental size of PtNC, a Pt31 Moreover, this model showed a much stronger metal-support inter-
model was chosen for simulation, which mostly exposes the (111) facet, action (−2.81 eV per interacting atom; Fig. 3b) than that of the model in
as revealed by the previous HAADF-STEM image in Fig. 1e. Although which Pt31 bound to the surface oxygen (−0.79 eV per interacting atom;
the supported model involving metal binding to surface oxygen Fig. 3a), which even surpassed the binding strength between Pt31 and
(Fig. 3a) is a common alternative, a model in which Pt31 binds to Mo the Pt (111) surface (−2.11 eV per interacting atom; Fig. 3c). This result

Nature Communications | (2025)16:4314 5

Article https://doi.org/10.1038/s41467-025-59450-6

Fig. 3 | DFT calculations for elucidating the stability and activity. The models for including LC-Pt and MoC-Pt, respectively. e Free energy diagram of hydrogen
PtNC are supported on the (a) surface oxygen of Mo2TiC2, (b) Mo of Mo2TiC2, and evolution at zero potential and pH 0 for the Pt (111) surface and Mo2TiC2-PtNC,
(c) Pt (111) surface. ΔEMSI: strength of the metal-support interaction. d The d-elec- including LC-Pt and MoC-Pt. Atomic coordinate dataset is provided as Supple-
tron structure of on-surface Pt atoms on the Pt (111) surface and Mo2TiC2-PtNC, mentary Data 1, other source data are provided as a Source Data file.

indicates that Mo2TiC2 with oxygen functional groups can act as a membrane electrode assembly (MEA) specifically for the electrolysis of
bifunctional heteroenergetic support. The Mo-block anchors and pure water (Fig. 4a). A commercial PEM device (Pt/C-20%||IrO2) with a
secures the PtNC in place, resulting in anomalous charge transfer, while loading of 500 μg cm−2 Pt was included for comparison. Although we
the O-block isolates the PtNC, preventing particle migration and coa- lack the specific expertise and equipment to fabricate industrial-grade,
lescence and inhibiting Ostwald ripening26. Therefore, the unique high-quality membrane electrodes with minimal ionic resistance
structure of Mo2TiC2-PtNC enables the achievement of high stability for between the cathode and anode, this does not compromise the validity
nanoscale PtNC, which is consistent with experimental observa- of the performance comparison. We sandwiched the Nafion 117
tions (Fig. 2g). membrane between Mo2TiC2-PtNC and commercial IrO2 for PEM water
As indicated by the charge transfer observed in both the experi- electrolysis, with a loading of 36 μg cm−2 Pt (1 mg cm−2 catalyst, see
mental (Fig. 1i) and theoretical (Fig. 3b) results, the unique metal- Supplementary Fig. 29 for Pt load optimization). We assessed stable
support interaction also significantly tunes the electronic structure of electrolytic hydrogen production at room temperature. Although the
PtNC and thus the HER activity. Figure 3d shows the d-electron struc- Pt loading in our case is substantially lower than that in commercial
tures of the surface Pt atoms on Mo2TiC2-PtNC, which included low- designs, the performance of our device, when directly compared,
coordinate Pt (LC-Pt) and Mo-coordinate Pt (MoC-Pt). Compared with aligns with the levels observed in contemporary commercial designs,
that of the surface Pt on the Pt (111) slab, the d-band center of LC-Pt is operating under an equivalent bias voltage (Fig. 4b). Moreover, our
upshifted as expected, whereas that of MoC-Pt is downshifted. This is Mo2TiC2-PtNC based cell was able to electrolyze hydrogen for more
consistent with the downshift in the d-band center of Mo2TiC2-PtNC than 8700 h at 200 mA cm−2. This was achieved with an overpotential
observed via XPS (−3.77 eV) compared with that of the Pt foil (−3.62 eV; of ~1.97 V and a degradation rate of only 2.2 μV h−1 (Fig. 4c), which
Supplementary Fig. 13). A downshift of the d-band center weakens the matches the DOE’s 2026 target (2.3 μV h−1).
binding of *H and moves it toward the optimal value, which is subse- In addition, the stability of such devices was better than that of
quently evidenced by the calculated adsorption-free energy of *H recently reported precious metal catalysts (Supplementary Table 6).
(Fig. 3e). The weakened *H binding increases the intrinsic HER activity We further conducted electrolysis experiments on the Mo2TiC2-PtNC
by accelerating the Tafel steps, which has been identified as the rate- based cell under industrial working conditions (1 A cm−2, 1 bar at 80
determining step for both Pt/C and Mo2TiC2-PtNC (Fig. 2e and Sup- °C). By constantly regulating the parameters during the assembly
plementary Fig. 15). Conversely, for the model with PtNC supported on process of the device, the impedance between the cathode and anode
graphene instead of Mo2TiC2-PtNC (Supplementary Fig. 28), Pt atoms at in our device can be reduced to 15–20 mΩ cm−2; as a result, the device
the interface were found to bind *H as strongly as LC-Pt. Therefore, the can run stably for hydrogen production for 4800 h at 1 A cm−2 (Fig. 4c).
metal-support interaction of Mo2TiC2-PtNC enables us to overcome the According to the calculation from DOE’s suggestion (see details in
disadvantage of decreasing the Pt size, where increasing the exposure Note 1), note that the clean electricity price of $0.03 per kWh, as
of LC-Pt would strengthen *H binding and decrease the intrinsic HER proposed by the DOE, was employed in our calculation52,53, only US$
activity. 1.327 was required to produce 1 kg H2 with this device, well below the
Finally, to assess the capacity of our Mo2TiC2-PtNC to electrolyze DOE goal of 2026 (<US$ 2 kg−1 H2). Compared with other membrane
water in an industrial setting, we utilized PEM devices assembled with a electrode assembly water electrolyzers (MEAWEs) with cathodic PGM

Nature Communications | (2025)16:4314 6

Article https://doi.org/10.1038/s41467-025-59450-6

Fig. 4 | Performance of PEM water electrolyzers (PEMWE) devices. a Photograph unexpected power off. d Photograph of the 49 cm2 device and MEA membrane
and schematic of the PEMWE device. b Chronopotentiometry test of a 4 cm2 Pt/C- together with the Mo2TiC2-PtNC cathode and IrO2 anode layers. e Long-term sta-
20%||IrO2 PEMWE device at 200 mA cm−2 (25 °C, ambient pressure), compared with bility of the 49 cm2 device at ambient pressure. All the performance tests of the
a Mo2TiC2-PtNC device. c Long-term electrolytic test of PEMs assembled with aforementioned PEMWE devices were conducted without iR correction. Source
Mo2TiC2-PtNC at 200 mA cm−2 and 1 A cm−2 using commercial IrO2 as the anode data are provided as a Source Data file.
catalyst, at 25 °C and 80 °C, respectively. Discontinuities in voltage arise from the

catalysts, the observed superiority of Mo2TiC2-PtNC as a cathode cat- commercial Pt/C-20%. Notably, our Mo2TiC2-PtNC catalyst demon-
alyst underscores its potential for industrial applications (Table 1). This strated a robust capacity for hydrogen production at 200 mA cm−2 over
potential is further emphasized by our ability to achieve gram-scale 8700 h in a PEM device, and even under industrial conditions (1 A cm−2,
synthesis of the catalyst (Supplementary Fig. 30). To demonstrate the 1 bar at 80 °C), it maintained stability for more than 4800 hours.
availability of our catalyst in mass hydrogen generation, we also con- Therefore, we identified a catalyst with the potential to replace com-
ducted a constant electrolytic hydrogen production test on an mercial Pt/C-20% for water electrolysis in industrial-scale PEM devices.
extended electrode area (49 cm2) of the Mo2TiC2-PtNC based cell, This advancement could help reduce catalyst costs and support the
which outputs over 4.09-liter pure hydrogen per hour, and the stability broader adoption of PEM reactors and grid implementations.
exceeded 3600 h (Fig. 4d, e). The findings reported here could con-
tribute to future developments in large-area membrane electrode Methods
assembly and reactor implementation, highlighting its potential for Materials synthesis
industrial applications. Chemicals. Ethanol, Nafion 117 perfluorinated resin solution (5 wt%),
In conclusion, to alleviate the usage of Pt-group precious metals in chloroplatinic acid hexahydrate (H2PtCl6·6H2O) and tetra-
acidic water electrolysis, we developed a catalyst (Mo2TiC2-PtNC), fea- butylammonium hydroxide (C16H37NO) were purchased from Macklin.
turing a low Pt mass loading on MXenes. The interaction between Pt Graphene oxide (GO) was purchased from Suzhou Tanfeng Graphene
and the support induces electron enrichment on the Pt surface, Technology Co., Ltd. Pt/C (nominally 20 wt% on carbon black) was
enabling thermo-neutral hydrogen adsorption. Consequently, the cat- purchased from Johnson Matthey. All reagents were used without
alyst achieved electrolytic performance and stability on par with further purification. All aqueous solutions were prepared with

Nature Communications | (2025)16:4314 7

Article https://doi.org/10.1038/s41467-025-59450-6

Table 1 | Performance metrics of different MEAWEs with acceleration voltage of 200 kV. X-ray photoelectron spectroscopy
cathodic Pt group metal (PGM) catalysts (XPS) was performed on an Axis Supra photoelectron spectrometer
using an exciting source of Al Kα radiation (1486.6 eV), and the binding
Cathodic Current Stability Degradation
PGM load- density (h) rate energy of the C 1s peak (284.8 eV) was selected as the actual reference.
ing (mA cm−2) (μV h−1) HADDF–STEM and EDS elemental mapping were performed on a
(μg cm−2) Themis Z field-emission transmission electron microscope at an
This work 36 200 8700+ 2.2a acceleration voltage of 200 kV. The Pt concentrations of all the sam-
1000 4800+ / ples were measured using inductively coupled plasma optical emission
PEMWEs67–69 40–500 100–1000 48–500 83.33–200 spectroscopy (ICP-OES, Avio 500). In situ diffuse reflectance Fourier
AEMWEs70–72 14–278 270–1000 20–2000 40–1600 transform infrared spectroscopy (DRIFTS) experiments were con-
Neutral >47 100–150 40 700–3500 ducted on a Thermo Scientific Nicolet iS50 FTIR spectrometer with
MEAWEs50,73 ZnSe as the prismatic window at room temperature. XAS measure-
a
Owing to an unforeseen power interruption at approximately 3840 h, the degradation rate was ments at the Pt L3-edge of the samples were carried out on the BL14W1
determined based on the initial continuous operation of 3840 h. beamline of the Shanghai Synchrotron Radiation Facility, operated in
fluorescent mode on all samples. XAFS data processing and fitting
were carried out with Demeter software.
Millipore water (resistivity of 18.2 MΩ cm). The detailed parameters are
provided in Supplementary Table 7. CO DRIFTS. The adsorption behavior of CO on the Mo2TiC2-PtSA and
Mo2TiC2-PtNC catalysts was studied by diffuse reflectance Fourier
Synthesis of multilayer Mo2TiC2 MXene. Mo2TiAlC2 MAX-phase transform infrared spectroscopy (DRIFTS). DRIFTS spectra were col-
powder (2.0 g) was slowly added to an HF (40%, 20 ml) solution in a lected on a Thermo Scientific Nicolet iS50 FTIR spectrometer at a
Teflon beaker under continuous stirring for 120 h at 55 °C. The product resolution of 8 cm−1. The infrared cell was first partially filled with inert
was centrifuged and washed several times in argon-saturated water KBr powder and filled with catalysts on the KBr holder. For fresh cat-
until the pH approached 6 or 7. The suspension was freeze-dried at −59 alysts, pretreatment at 100 °C and Ar for 30 min was performed to
°C for 48 h to obtain dry multilayer Mo2TiC2 powders. remove impurities (H2O) that may be adsorbed on the surface. Then,
the background spectrum was recorded after the reaction cell was
Synthesis of fewer-layer Mo2TiC2 MXene. Multilayer Mo2TiC2 MXene cooled to 25 °C. Then, CO adsorption was conducted with a mixture of
(1.0 g) was added to 20 ml of organic solvent (54–56 wt% TBAOH 10% CO/Ar (15 ml min−1) and Ar (30 ml min−1) for 20 min. Furthermore,
((C4H9)4NOH)) and stirred at 50 °C for 24 h. The resulting mixture was gas-phase CO in the reaction cell was removed by Ar purging at a flow
centrifuged and washed three times in deionized water to separate the rate of 30 ml min−1, after which the desorption spectrum of the
MXene from TBAOH. After the upper layer was emptied, 100 ml of catalyst-adsorbed CO was recorded. CO adsorption experiments were
deionized water was added, and the mixture was sonicated for 1 h in an carried out at 25 °C.
ice bath, followed by centrifugation for 1 h at 3500 rpm (~1370 × g) to
obtain fewer-layer MXenes with uniform dispersion. In situ DEMS tests. For the in situ DEMS test, 2 mg of each catalyst was
mixed with 5 wt% Nafion (20 µl) in 980 µl of ethanol and then sonicated
Synthesis of Mo2TiC2-PtNC. The sample was prepared using a stan- for 20 min to form a homogeneous solution. A 0.01 ml of ink was drop-
dard hydrothermal technique. Specifically, 50 mg of few-layered cast onto a glassy carbon (diameter of 0.3 cm) working electrode. The
Mo2TiC2 MXene was uniformly dispersed in 25 ml of deionized water electrodes were dried at room temperature for at least 30 min. In situ
with 7.65 mg of H2PtCl6·6H2O. This mixture was then subjected to differential electrochemical mass spectrometry (DEMS) was per-
ultrasonic agitation to achieve a homogenous solution. The mixture formed using a custom capillary electrochemical mass spectrometer
was transferred to a 50 ml Teflon-lined autoclave and reacted at 180 °C single cell. A capillary was inserted into the in situ cell, close to the side
for 0.5 h. After naturally cooling to room temperature, the products above the working electrode. The gaseous product was introduced
were washed and centrifuged three times with deionized water to into a DEMS sensor (PrismaPro). Linear sweep voltammetry was per-
obtain the Mo2TiC2-PtNC catalysts. formed on the cathode at a scan rate of 1 mV s−1. A photograph of the
in situ DEMS setup is provided in Supplementary Fig. 31.
Synthesis of Mo2TiC2-PtSA and rGO-PtNC. The synthesis of Mo2TiC2-
PtSA was performed via a hydrothermal method similar to that used for In situ ATR-SEIRAS test. 4 mg of each catalyst was mixed with 40 μl of
Mo2TiC2-PtNC. First, 5.5 mg of [Pt(NH3)4](NO3)2 was added to a 50 ml Nafion (5 wt%) and 960 μl of ethanol and then sonicated for 20 min to
solution of 50 mg of Mo2TiC2 and heated at 140 °C for 1 h to obtain form a homogeneous solution. A 0.238 ml of ink was dropped on a
Mo2TiC2-PtSA after centrifugation and washing. The preparation of gold-plated silicon crystal (usable area of 0.95 cm2) with a typical cat-
rGO-PtNC was achieved by a simple impregnation method. Specifically, alyst loading (~1 mg cm−2). For constant potential testing, the potential
7.65 mg of H2PtCl6·6H2O was added to 25 ml of aqueous solution was varied from 0.1 to −0.1 V versus RHE, and the infrared spectral data
containing 50 mg of GO to form a uniform dispersion. The mixture was were recorded. A photograph of the in situ ATR-SEIRAS setup is pro-
further freeze-dried for 48 h at −59 °C, Following this, it was trans- vided in Supplementary Fig. 32.
ferred to a tube furnace and reacted at 450 °C under a 5% H2/Ar All the detailed configurations of the in situ electrochemical cells
atmosphere for 2 h, resulting in the rGO-PtNC product. used are provided in Supplementary Table 8.

Characterizations Electrochemical measurements

Scanning electron microscopy (SEM) was performed on a Gemini- Calibration of the reference electrode and conversion to RHE scale.
SEM300 electron microscope. Powder X-ray diffraction (PXRD) was A mercury-mercurous sulfate reference electrode (Hg/Hg2SO4) was
performed using Cu Kα radiation (λ = 1.54178 Å) by a Philips X’Pert Pro used in all the measurements. The calibration of the reference elec-
Super diffractometer. Raman spectra were obtained using a LabRAM trode was performed in the standard three-electrode configuration.
HR laser Raman analyzer with an excitation wavelength of 785 nm. Polished platinum foil was used as the working and counter electrode,
Transmission electron microscopy (TEM) was performed on a Tecnai and Hg/Hg2SO4 was used as the reference electrode. The electrolyte
G2 F20 S-Twin TMP transmission electron microscope with an was purged and saturated with high-purity H2 before measurement

Nature Communications | (2025)16:4314 8

Article https://doi.org/10.1038/s41467-025-59450-6

and was kept flowing during the calibration process. Cyclic voltam- curves. Calculate ECSA using the following Eq. (2).
metry (CV) was then run at a scan rate of 1 mV s−1, and the average
potential at which the current crossed zero was taken as the thermo- SQ ðA
V Þ=vðV s1 Þ
ECSAPt = 2 Þ
ð2Þ
dynamic potential for the hydrogen electrode reactions. In our 210ðμc cmPt
experiment, in 0.5 M H2SO4, the zero current point was calculated to
be 0.697 V. where SQ is the integral area of the hydrogen desorption charge region
Therefore, E(RHE) = E(Hg/Hg2SO4) + 0.697 V in the CV curve. v is the scan rate and 210 μC cmPt–2 for a monolayer
H-UPD at Pt.
Measurement of the electrode material. All the electrochemical The TOF values were calculated based on the number of surface
performance tests were performed at room temperature by a Bio- active Pt atoms in each catalyst (the TOF value of Mo2TiC2-PtSA was
Logic VMP3 electrochemical workstation with a typical three- calculated from the mass of Pt) on the electrode according the fol-
electrode system. A glass carbon electrode with a diameter of lowing Eq. (3).
1.2 cm was selected as the working electrode, Hg/HgSO4 was used as
1
the reference electrode, and a graphite rod was used as the counter 1 iðAÞ=FðC mol Þ
electrode. An H-type cell was used as the electrolytic cell (a Nafion TOF =
ð3Þ
2 SQ ðA V Þ=vðV s1 ÞFðC mol 1 Þ
membrane was used to separate the anode and cathode chambers).
The Hg/HgSO4 electrode was calibrated in a H2-saturated 0.5 M where i represents the current recorded from the LSV curves.
H2SO4 electrolyte. The preparation, storage and pH values of the The mass activity was determined using the following Eq. (4).
electrolyte (0.5 M H2SO4, pH value: 0.43 ± 0.02) are provided in
Supplementary Fig. 33 and Note 2. iðAÞ
Mass activity = ð4Þ
To prepare the working electrode ink, 4 mg of each catalyst was mPt ðmgÞ
mixed with 40 μl of Nafion (5 wt%) and 960 μl of ethanol and then
sonicated for 20 min to form a homogeneous solution. A 0.283 ml of where mPt is the mass of Pt in each catalyst on the electrode.
ink was dropped on glass carbon (area of 1.13 cm2) with a typical cat-
alyst loading (~1 mg cm−2). Additionally, the cathode chamber was Performance of the membrane electrode devices. To prepare the
separated from the anode chamber (counter electrode: graphite rod) membrane electrode device, the anodic catalyst IrO2 was selected for
by a Nafion 117 membrane (the activation process of the Nafion the oxygen evolution reaction (OER), and Pt/C (20 wt%, Johnson Mat-
membrane is provided in Note 3). For the performance test, the they) or Mo2TiC2-PtNC was used as the cathode HER catalyst. For
cathode chamber was placed on a magnetic stirring table (~1600 rpm) commercial design, a slurry composed of Pt/C, Nafion (5 wt%) and
to facilitate the rapid desorption of H2 gas bubbles. Linear sweep ethanol was evenly sprayed on both sides of the proton exchange
voltammetry was carried out in 0.5 M H2SO4 at a scan rate of 5 mV s−1, membrane (Nafion 117), with mass loading of 1 mg cm−2 Ir and
deaerated with Ar. Cyclic voltammetry was performed in the potential 0.5 mg cm−2 Pt. The PEM device with Mo2TiC2-PtNC catalyst is prepared
window from 0.15 V to −1.5 V (versus RHE) for 10000 cycles (scan rate in the same procedure and the mass loadings of IrO2 and Mo2TiC2-PtNC
of 50 mV s−1). Electrochemical resistance measurements were per- were controlled to be 1 mg cm−2 Ir and 36 μg cm−2 Pt, respectively. After
formed at the OCV with the frequency range from 0.01 Hz to 106 Hz drying, the membrane electrode was hot-pressed (100 °C) at a pres-
(operando EIS tests were performed at potentials ranging from 0.02 V sure of 2 MPa for 10 min. Titanium felt and carbon paper were used as
to −0.05 V (versus RHE), and the resistance values were calculated the anode and cathode gas diffusion layers, and a peristaltic pump
from the high frequency intercepts with the X-axis on the Nyquist plot. (2.5 ml min−1) was used for pure water circulation to the anodic side. A
The Tafel test was performed by chronoamperometry54, which is more Fumasep FS-990-PK membrane was used to fabricate the 49 cm2 PEM
accurate than the Tafel value obtained from the polarization curve. device following the same preparation procedures described pre-
Cyclic voltammetry and potential-time stability tests were conducted viously. All the cell tests were performed at ambient pressure, and
with a catalyst loading of ~1 mg cm−2 on 1 cm2 carbon paper. none of the measured cell voltages were iR compensated. The details
Additionally, the constant potential test was conducted using the of the catalyst loading calculation and the geometric dimensions of the
“Chronoamperometry” function in BioLogic VMP3, whereas the con- electrodes are provided in Note 4.
stant current test was performed using the “Chronopotentiometry” In a PEM water electrolyzer test, a DC current power supply
function in BioLogic VMP3. Data acquisition and processing were (DCPS0614, 30 V/20 A) was employed as a constant potentiostat. Sta-
carried out using the corresponding software, “EC-Lab”, via a USB bility tests were performed in constant current mode, with data col-
serial connection. lected and processed using the “DC Power Supply” software connected
The FE of Mo2TiC2-PtNC was calculated using the following Eq. (1). via a USB serial port.

n
F
nðH 2 Þ
Computational methods
FE =
100% ð1Þ The Vienna ab initio simulation package (VASP)55–57 was used for all
i
t
density functional theory (DFT) calculations. In these simulations, the
valence electrons were defined as follows: the 1s electron in H, the 2s
where n represents the number of electron transfer involved in the and 2p electrons in C and O, the 3d and 4s electrons in Ti, the 4d, 4p,
HER, F denotes the Faraday constant (96,485 C mol−1), nðH 2 Þ denotes and 5s electrons Mo, and the 5d and 6s electrons in Pt58. A plane-wave
the number of moles of H2 detected by gas chromatography (GC) basis set was employed with a kinetic energy cutoff of 450 eV. Core
during HER electrolysis (using GC to detect the volume of H2, applying electrons were treated using the projector augmented-wave (PAW)
the ideal gas equation to convert to nðH 2 Þ ), i represents the applied method59. Monkhorst–Pack meshes60 of 2 × 2 × 1 sampling in the Bril-
current, and t represents the reaction time of HER electrocatalysis. louin zone were employed for the slab models.
The ECSA of Pt was estimated by measuring the H-UPD char- For pristine Mo2TiC2-O2, a 6 × 6 supercell with one layered
acteristics. CV measurements were performed in a three-electrode structure was chosen. For pristine graphene, an 8 × 8 supercell was
system (with an Ar-saturated 0.5 M H2SO4 solution) collected from 0 to chosen. For pristine Pt(111), a 6 × 6 supercell with four layers was
0.6 V versus RHE at a scan rate of 50 mV s−1. Calculated from the region chosen. Except for the bottom two layers of the Pt(111)-based cases, all
of hydrogen desorption charge (~0–0.5 V versus RHE) on the CV the atoms were relaxed. Convergence of the geometry optimization

Nature Communications | (2025)16:4314 9

Article https://doi.org/10.1038/s41467-025-59450-6

was assumed when the force on each atom was less than 0.02 eV Å−161. Note 2|Preparation and storage of the electrolyte
The Perdew-Burke-Ernzerhof (PBE) functional, within the generalized The 0.5 M H2SO4 electrolyte was freshly prepared before each use
gradient approximation (GGA), was employed for all calculations62. using the following procedure. First, 486.4 ml of deionized water was
Dispersion interactions were accounted for using the DFT + D3 measured and transferred into a 500 ml volumetric flask. Next, 13.6 ml
method with Becke-Jonson (BJ) damping63,64. of concentrated H2SO4 (98%) was carefully measured using a 20 ml
The free energy of the adsorbed state was calculated using Eq. (5): pipette and slowly added to the volumetric flask while gently swirling
the solution to ensure thorough mixing. The solution was then allowed
ΔGH = ΔEH + ΔEZPE  TΔS ð5Þ to cool to room temperature. Once cooled, deionized water was gra-
dually added to the volumetric flask until the solution’s meniscus
reached the calibration mark. The solution was then sealed and stored
where ΔEH is the adsorption energy of hydrogen, ΔEZPE is the differ-
in a cool, dry place for use.
ence corresponding to the zero point energy between the adsorbed
state and the gas phase, and TΔS is the term corresponding to the
Note 3|Activation process of the Nafion 117 membrane
entropy correction of hydron adsorption. As reported in previous
The Nafion 117 membrane, with a thickness of 183 μm, was used
work, a value of ΔEZPE -TΔS of 0.24 eV is employed here65.
when fabricating the three-electrode and MEA device and under-
In the free energy diagram of the HER, the Gibbs free energy
went an activation process prior to use: the membrane was first
change (ΔG) of the proton-coupled electron transfer (PCET) step was
treated in a 5% hydrogen peroxide solution at 80 °C for 1 h to
calculated using the computational hydrogen electrode (CHE)
remove organic impurities. It was then rinsed by soaking in deio-
model20,66, which defines the chemical potential of the proton-electron
nized water for 30 minutes to eliminate any residual peroxide. The
pair as half of the chemical potential of hydrogen.
membrane was subsequently treated in a 5% dilute sulfuric acid
solution at 80 °C for 1 h to increase its proton conductivity, fol-
Note 1 | Techno-economic analysis
lowed by a final rinse in deionized water for 30 min to remove any
Considering that the cost of noble metals in the cathode accounts for a
remaining acid.
high proportion of the total MEA cost (0.4–0.6 mg cm−2), cost
accounting of the cathodic noble metals is necessary. The price of the
Note 4|Details of the catalyst loading calculation and geometric
noble metal (Pt) was obtained from the Johnson Matthey Price Charts.
dimensions of the electrodes
1) The cathodic noble metal content of our PEM electrolyzer is
In a three-electrode test, the catalyst slurry concentration (c) was
0.036 mg cm−2, which is much lower than the reported cathode plati-
4 mg ml−1, and the glassy carbon electrode area (s) was 1.13 cm2 (dia-
num load (0.4–0.6 mg cm−2). Our cathodic platinum metal cost is only
meter of 1.2 cm). A catalyst drop volume (v) of 0.283 ml was applied,
$0.01181 cm−2.
resulting in a catalyst loading (m) on the electrode calculated as
2) The energy efficiency of a PEM electrolyzer can be calculated by
the following equation:
m = v
c=s  1 mg cm2

1:23 V 1:23 V In situ DEMS tests, for a catalyst slurry concentration of 2 mg ml−1, a
Energy efficiency = = × 100% = 74:5% ð6Þ
U cell 1:65V glassy carbon electrode area of 0.0707 cm2 (diameter of 0.3 cm), and a
catalyst drop volume of 0.01 ml, the catalyst loading (excessive load-
where 1.23 V represents the theoretical energy of the products and ing of the catalyst on such an electrode can cause catalyst detachment,
U cell is the cell voltage (V) required to deliver a current density leading to unreliable experimental results) was calculated as
of 1 A cm−2.
3) The energy consumption of a PEM electrolyzer is calculated by m = v
c=s  0:283 mg cm2
the following equation:
In situ ATR-SEIRAS test, for a catalyst slurry concentration of 4 mg ml−1,
U cell I cell t 1:65V × 4A × 4000h a usable gold-plated silicon crystal area of 0.95 cm2 (diameter of
Energy consumption = = 1.1 cm), and a catalyst drop volume of 0.238 ml, the catalyst loading
mH 2 596:98 g ð7Þ
1 was calculated as
= 44:22 kWh kg H2
m = v
c=s  1 mg cm2
where I cell is the current delivered (A), t is the operation time (h), and
mH 2 is the mass of hydrogen produced in a t duration, which can be
calculated by Faraday’s laws of electrolysis:
Data availability
All data were available in the main text or the supplementary mate-
I cell × t 4A × 4000 × 3600s rials. Source data are provided with this paper.
mH 2 = × M H2 = × 2g=mol = 596:98g ð8Þ
z ×F 2 × 96485:3C=mol
References
z in Eq. (8) is the number of electrons transferred to produce one 1. van Renssen, S. The hydrogen solution? Nat. Clim. Chang. 10,
hydrogen molecule, M H 2 is the relative molecular mass (2 g mol−1). 799–801 (2020).
The energy efficiency and energy consumption of our PEM elec- 2. Dresselhaus, M. S. & Thomas, I. L. Alternative energy technologies.
trolyzer are superior to those of the targets (69% and 48 kWh kg−1H2 for Nature 414, 332–337 (2001).
2026) set by the US Department of Energy. 3. Hydrogen Shot|Department of Energy. https://www.energy.gov/
4) Cost of H2 per kilogram of hydrogen: eere/fuelcells/hydrogen-shot.
Cost (H2/kg) = energy consumption × electricity price 4. Terlouw, T., Bauer, C., McKenna, R. & Mazzotti, M. Large-scale
=44.22 kWh/kg H2 × $0.03/kWh hydrogen production via water electrolysis: a techno-economic
=$1.327/kg H2 and environmental assessment. Energy Environ. Sci. 15,
Our H2 production cost is much lower than the target ($2/kg H2 3583–3602 (2022).
for 2026) set by the US Department of Energy.

Nature Communications | (2025)16:4314 10

Article https://doi.org/10.1038/s41467-025-59450-6

5. King, L. A. et al. A non-precious metal hydrogen catalyst in a com- 29. Naguib, M. et al. Two-dimensional nanocrystals produced by
mercial polymer electrolyte membrane electrolyser. Nat. Nano- exfoliation of Ti3AlC2. Adv. Mater. 23, 4248–4253 (2011).
technol. 14, 1071–1074 (2019). 30. Anasori, B. et al. Two-dimensional, ordered, double transition
6. Technical Targets for Proton Exchange Membrane Electrolysis | metals carbides (MXenes). ACS Nano 9, 9507–9516 (2015).
Department of Energy. https://www.energy.gov/eere/fuelcells/ 31. Naguib, M., Unocic, R. R., Armstrong, B. L. & Nanda, J. Large-scale
technical-targets-proton-exchange-membrane-electrolysis. delamination of multi-layers transition metal carbides and carbo-
7. Yu, H. & Yi, B. Hydrogen for energy storage and hydrogen pro- nitrides “MXenes”. Dalton Trans. 44, 9353–9358 (2015).
duction from electrolysis. Chin. J. Eng. Sci. 20, 58 (2018). 32. Bazin, P. et al. FT-IR study of CO adsorption on Pt/CeO2: char-
8. Minke, C., Suermann, M., Bensmann, B. & Hanke-Rauschenbach, R. acterisation and structural rearrangement of small Pt particles.
Is iridium demand a potential bottleneck in the realization of large- Phys. Chem. Chem. Phys. 7, 187–194 (2005).
scale PEM water electrolysis? Int J. Hydrog. Energy 46, 33. Zhao, D. et al. MXene (Ti3C2) vacancy-confined single-atom catalyst
23581–23590 (2021). for efficient functionalization of CO2. J. Am. Chem. Soc. 141,
9. An, L. et al. Recent development of oxygen evolution electro- 4086–4093 (2019).
catalysts in acidic environment. Adv. Mater. 33, 2006328 (2021). 34. Zhang, J. et al. Single platinum atoms immobilized on an MXene as
10. Chen, Z. et al. Advances in oxygen evolution electrocatalysts for an efficient catalyst for the hydrogen evolution reaction. Nat. Catal.
proton exchange membrane water electrolyzers. Adv. Energy 1, 985–992 (2018).
Mater. 12, 2103670 (2022). 35. Li, Z. et al. Direct methane activation by atomically thin platinum
11. Liang, Q. & Li, D. Activating localized lattice oxygen for durable nanolayers on two-dimensional metal carbides. Nat. Catal. 4,
acidic water oxidation. Chem. Catal. 1, 506–508 (2021). 882–891 (2021).
12. Tymoczko, J., Calle-Vallejo, F., Schuhmann, W. & Bandarenka, A. S. 36. Mao, J. et al. Design of ultrathin Pt-Mo-Ni nanowire catalysts for
Making the hydrogen evolution reaction in polymer electrolyte ethanol electrooxidation. Sci Adv 3 (2017).
membrane electrolysers even faster. Nat. Commun. 7, 37. Zhang, X. et al. A stable low-temperature H2-production catalyst by
10990 (2016). crowding Pt on α-MoC. Nature 589, 396–401 (2021).
13. Pan, S. et al. Efficient and stable noble-metal-free catalyst for acidic 38. Moulder, J. F., Stickle, W. F., Sobol, P. E., Bomben, K. D. &
water oxidation. Nat. Commun. 13, 2294 (2022). Chastain, J. Handbook of X-Ray Photoelectron Spectroscopy A
14. Staszak-Jirkovský, J. et al. Design of active and stable Co–Mo–Sx Reference Book of Standard Spectra for Identification and
chalcogels as pH-universal catalysts for the hydrogen evolution Interpretation of XPS Data (Physical Electronics Division, Perkin-
reaction. Nat. Mater. 15, 197–203 (2016). Elmer Corporation, 1979).
15. Zhang, X.-L. et al. Efficient acidic hydrogen evolution in proton 39. Zhao, Y. et al. Modulating Pt-O-Pt atomic clusters with isolated
exchange membrane electrolyzers over a sulfur-doped marcasite- cobalt atoms for enhanced hydrogen evolution catalysis. Nat.
type electrocatalyst. Sci. Adv. 9, eadh2885 (2023). Commun. 13, 2430 (2022).
16. Yang, H. et al. Metastable-phase platinum oxide for clarifying the 40. Xiong, M. et al. In situ tuning of electronic structure of catalysts
Pt–O active site for the hydrogen evolution reaction. Energy using controllable hydrogen spillover for enhanced selectivity. Nat.
Environ. Sci. 16, 574–583 (2023). Commun. 11, 4773 (2020).
17. Subbaraman, R. et al. Enhancing hydrogen evolution activity in 41. Wang, X. et al. Atomic-precision Pt6 nanoclusters for enhanced
water splitting by tailoring Li+-Ni(OH)2-Pt interfaces. Science 334, hydrogen electro-oxidation. Nat. Commun. 13, 1596 (2022).
1256–1260 (2011). 42. Zhou, K. L. et al. Platinum single-atom catalyst coupled with tran-
18. Shi, Y. et al. Electronic metal–support interaction modulates single- sition metal/metal oxide heterostructure for accelerating alkaline
atom platinum catalysis for hydrogen evolution reaction. Nat. hydrogen evolution reaction. Nat. Commun. 12, 3783 (2021).
Commun. 12, 3021 (2021). 43. Cheng, N. et al. Platinum single-atom and cluster catalysis of the
19. Hansen, J. N. et al. Is there anything better than Pt for HER? ACS hydrogen evolution reaction. Nat. Commun. 7, 13638 (2016).
Energy Lett. 6, 1175–1180 (2021). 44. Liu, D. et al. Atomically dispersed platinum supported on curved
20. Nørskov, J. K. et al. Trends in the exchange current for hydrogen carbon supports for efficient electrocatalytic hydrogen evolution.
evolution. J. Electrochem Soc. 152, J23 (2005). Nat. Energy 4, 512–518 (2019).
21. Yan, Q.-Q. et al. Reversing the charge transfer between platinum 45. Fang, S. et al. Uncovering near-free platinum single-atom dynamics
and sulfur-doped carbon support for electrocatalytic hydrogen during electrochemical hydrogen evolution reaction. Nat. Com-
evolution. Nat. Commun. 10, 4977 (2019). mun. 11, 1029 (2020).
22. Wang, A., Li, J. & Zhang, T. Heterogeneous single-atom catalysis. 46. Yan, P. et al. One stone five birds” plasma activation strategy
Nat. Rev. Chem. 2, 65–81 (2018). synergistic with Ru single atoms doping boosting the hydrogen
23. Liu, L. & Corma, A. Metal catalysts for heterogeneous catalysis: from evolution performance of metal hydroxide. Adv. Funct. Mater. 33,
single atoms to nanoclusters and nanoparticles. Chem. Rev. 118, 2301343 (2023).
4981–5079 (2018). 47. Chen, W. et al. Deciphering the alternating synergy between
24. Chen, Y. et al. Single-atom catalysts: synthetic strategies and interlayer Pt single-atom and NiFe layered double hydroxide for
electrochemical applications. Joule 2, 1242–1264 (2018). overall water splitting. Energy Environ. Sci. 14, 6428 (2021).
25. Rong, H., Ji, S., Zhang, J., Wang, D. & Li, Y. Synthetic strategies of 48. Zalitis, C. M., Kramer, D., Sharman, J., Wright, E. & Kucernak, A. R. Pt
supported atomic clusters for heterogeneous catalysis. Nat. Com- nano-particle performance for PEFC reactions at low catalyst
mun. 11, 5884 (2020). loading and high reactant mass transport. ECS Trans. 58,
26. Hu, S. & Li, W.-X. Sabatier principle of metal-support interaction for 39–47 (2013).
design of ultrastable metal nanocatalysts. Science 374, 49. Li, J. et al. Ethylene-glycol ligand environment facilitates highly
1360–1365 (2021). efficient hydrogen evolution of Pt/CoP through proton concentra-
27. Naguib, M., Mochalin, V. N., Barsoum, M. W. & Gogotsi, Y. 25th tion and hydrogen spillover. Energy Environ. Sci. 12, 2298–2304
anniversary article: mxenes: a new family of two-dimensional (2019).
materials. Adv. Mater. 26, 992–1005 (2014). 50. Sun, K. et al. Interfacial water engineering boosts neutral water
28. Gogotsi, Y. & Anasori, B. The rise of MXenes. ACS Nano 13, reduction. Nat. Commun. 13, 6260 (2022).
8491–8494 (2019).

Nature Communications | (2025)16:4314 11

Article https://doi.org/10.1038/s41467-025-59450-6

51. Li, P. et al. Hydrogen bond network connectivity in the electric Acknowledgements
double layer dominates the kinetic pH effect in hydrogen electro- The authors acknowledge the National Key Research and Development
catalysis on Pt. Nat. Catal. 5, 900–911 (2022). Program of China (2024YFB4105700, 2022YFA1504402), National Nat-
52. Hao, S. et al. Torsion strained iridium oxide for efficient acidic water ural Science Foundation of China (22405035, 52171201, 22233002,
oxidation in proton exchange membrane electrolyzers. Nat. Nano- 22103014), Natural Science Foundation of Sichuan Province
technol. 16, 1371–1377 (2021). (2025NSFJQ0017, 24NSFSC5779), the Innovation Program for Quantum
53. Shi, Z. et al. Customized reaction route for ruthenium oxide towards Science and Technology (2021ZD0303305), the China Postdoctoral
stabilized water oxidation in high-performance PEM electrolyzers. Science Foundation funded project (2022M710601), and the Huzhou
Nat. Commun. 14, 843 (2023). Science and Technology Bureau (2023GZ02). We thank beamline
54. Anantharaj, S. et al. The pitfalls of using potentiodynamic polar- BL14W1 of the Shanghai Synchrotron Radiation Facility for providing the
ization curves for tafel analysis in electrocatalytic water splitting. facilities.
ACS Energy Lett. 6, 1607–1611 (2021).
55. Kresse, G. & Hafner, J. Ab initio molecular dynamics for open-shell Author contributions
transition metals. Phys. Rev. B 48, 13115–13118 (1993). The project was conceptualized by C.X. and supervised by C.X., X.X. and
56. Kresse, G. & Furthmüller, J. Efficiency of ab initio total energy cal- Q.J. H.Z. prepared the catalysts. H.Z., Q.Z., J.L. and C.L. performed the
culations for metals and semiconductors using a plane-wave basis catalytic tests. H.Z., Y.J., R.Z., Z.Z., Z.X., Y.C. and D.Z. performed the
set. Comput. Mater. Sci. 6, 15–50 (1996). catalyst characterizations. X.Xiong, Zhao.C., Y.D. and C.L. performed the
57. Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio XAFS measurements. Yi.C., J.T., T.Z. and X.L. helped in the analysis of the
total-energy calculations using a plane-wave basis set. Phys. Rev. B. data. Z.C. and X.X. performed the DFT calculations. C.X., Q.J., X.X., Z.C.
54, 11169–11186 (1996). and H.Z. wrote the paper with input from all the authors. All the authors
58. Niu, W. et al. Pb-rich Cu grain boundary sites for selective CO-to-n- discussed the results and commented on the manuscript.
propanol electroconversion. Nat. Commun. 14, 4882 (2023).
59. Blochl, P. E. Projector augmented-wave method. Phys. Rev. B 50, Competing interests
17953–17979 (1994). A China provisional patent application (202310633252.7) based on the
60. Monkhorst, H. J. & Pack, J. D. Special points for Brillouin-zone technology described in this work was filed in July 2023 by C.X., H.Z. and
integrations. Phys. Rev. B 13, 5188–5192 (1976). Q.J. at the University of Electronic Science and Technology of China. The
61. Sun, S. et al. Potassium-promoted limestone for preferential direct other authors declare no competing interests.
hydrogenation of carbonates in integrated CO2 capture and utili-
zation. JACS Au 4, 72–79 (2024). Additional information
62. Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient Supplementary information The online version contains
approximation made simple. Phys. Rev. Lett. 77, 3865–3868 (1996). supplementary material available at
63. Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and https://doi.org/10.1038/s41467-025-59450-6.
accurate ab initio parametrization of density functional dispersion
correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132, Correspondence and requests for materials should be addressed to
154104 (2010). Qiu Jiang, Xin Xu or Chuan Xia.
64. Grimme, S., Ehrlich, S. & Goerigk, L. Effect of the damping function
in dispersion corrected density functional theory. J. Comput. Chem. Peer review information Nature Communications thanks Min-Rui Gao,
32, 1456–1465 (2011). Xiong Peng and the other anonymous reviewer(s) for their contribution
65. Chen, Z. et al. Accurate descriptions of molecule-surface interac- to the peer review of this work. A peer review file is available.
tions in electrocatalytic CO2 reduction on the copper surfaces. Nat.
Commun. 14, 936 (2023). Reprints and permissions information is available at
66. Nørskov, J. K. et al. Origin of the overpotential for oxygen reduction http://www.nature.com/reprints
at a fuel-cell cathode. J. Phys. Chem. B 108, 17886–17892 (2004).
67. Dong, S. et al. Overall design of anode with gradient ordered Publisher’s note Springer Nature remains neutral with regard to
structure with low iridium loading for proton exchange membrane jurisdictional claims in published maps and institutional affiliations.
water electrolysis. Nano Lett. 22, 9434–9440 (2022).
68. Liu, W. et al. Single‐layer platinum cluster catalyst for efficient Open Access This article is licensed under a Creative Commons
hydrogen electro‐production. Adv. Funct. Mater. 33, Attribution-NonCommercial-NoDerivatives 4.0 International License,
2212752 (2023). which permits any non-commercial use, sharing, distribution and
69. Shi, Z. et al. Phase-dependent growth of Pt on MoS2 for highly reproduction in any medium or format, as long as you give appropriate
efficient H2 evolution. Nature 621, 300–305 (2023). credit to the original author(s) and the source, provide a link to the
70. Gao, L. et al. Engineering a local potassium cation concentrated Creative Commons licence, and indicate if you modified the licensed
microenvironment toward the ampere-level current density material. You do not have permission under this licence to share adapted
hydrogen evolution reaction. Energy Environ. Sci. 16, material derived from this article or parts of it. The images or other third
285–294 (2023). party material in this article are included in the article’s Creative
71. Zhao, J. et al. Activating Ru-O-Co interaction on the a-Co(OH)2 @Ru Commons licence, unless indicated otherwise in a credit line to the
interface for accelerating the volmer step of alkaline hydrogen material. If material is not included in the article’s Creative Commons
evolution. Small Methods 7, 2201362 (2023). licence and your intended use is not permitted by statutory regulation or
72. Zhang, T. et al. Pinpointing the axial ligand effect on platinum exceeds the permitted use, you will need to obtain permission directly
single-atom-catalyst towards efficient alkaline hydrogen evolution from the copyright holder. To view a copy of this licence, visit http://
reaction. Nat. Commun. 13, 6875 (2022). creativecommons.org/licenses/by-nc-nd/4.0/.
73. Zheng, X. et al. Tailoring a local acid-like microenvironment for
efficient neutral hydrogen evolution. Nat. Commun. 14, © The Author(s) 2025
4209 (2023).

Nature Communications | (2025)16:4314 12