Ir-O-Mn embedded in porous nanosheets enhances charge transfer in low-iridium PEM electrolyzers

W4406000695 · 수집 2026-09-26 12:32:49

T11690 Advanced battery technologies research 이차전지

CoT 라벨링

라벨링 전

PDF 파일

자동 다운로드한 PDF PDF 보기

기본 정보

DOI 제목 발행년도 저널 URL 페이지 열기 PDF URL PDF 열기 발행기관 피인용수 라이선스

추가 정보 초록 본문

Article https://doi.org/10.1038/s41467-024-54646-8
# Ir-O-Mn embedded in porous nanosheets enhances charge transfer in low-iridium PEM electrolyzers

Received: 22 March 2024

Accepted: 19 November 2024

Check for updates

Dawei Wang 1,4, Fangxu Lin 1,4, Heng Luo1, Jinhui Zhou1, Wenshu Zhang1, Lu Li1,
Yi Wei 1, Qinghua Zhang 2, Lin Gu 2, Yanfei Wang3, Mingchuan Luo1,
Fan Lv 1 & Shaojun Guo 1

Using metal oxides to disperse iridium (Ir) in the anode layer proves effective
for lowering Ir loading in proton exchange membrane water electrolyzers
(PEMWE). However, the reported low-Ir-based catalysts still suffer from
unsatisfying electrolytic efficiency and durability under practical industrial
working conditions, mainly due to insufficient catalytic activity and mass
transport in the catalyst layer. Herein we report a class of porous heterogeneous nanosheet catalyst with abundant Ir-O-Mn bonds, achieving a notable
mass activity of 4 A mgIr−1 for oxygen evolution reaction at an overpotential of

300 mV, which is 150.6 times higher than that of commercial IrO2. Ir-O-Mn
bonds are unraveled to serve as efficient charge-transfer channels between insitu electrochemically-formed IrOx clusters and MnOx matrix, fostering the
generation and stabilization of highly active Ir 3+ species. Notably, Ir/MnOxbased PEMWE demonstrates comparable performance under 10-fold lower Ir
loading (0.2 mgIr cm −2 ), taking a low cell voltage of 1.63 V to deliver 1 A cm −2 for
over 300 h, which positions it among the elite of low Ir-based PEMWEs.

The proton exchange membrane water electrolyzers (PEMWE) have
emerged as one of the most significant techniques for the green
hydrogen synthesis 1–4 . However, the large amount of iridium (Ir) used
in current PEMWE devices (400 kg GW −1 ) has been widely considered
as the main impediment to the large-scale and sustainable deployment
of PEMWEs 5–9 . Cooperating iridium and metal oxide-based support
materials serves as an effective strategy for the elevated atomic utilization of Ir to enhance the activity and stability of oxygen evolution
reaction (OER) with the decrease of Ir loading 10–13 . Though efficient
supported Ir catalysts have been reported, realizing both satisfactory
activity and durability under industrially high current density with a
low Ir loading (<0.3 mgIr cm −2 ) still remains a big challenge because the
limited number of active sites with insufficient intrinsic activity usually
necessitate higher electrolytic voltage for achieving industrial current
density, leading to the conversion of Ir into a more readily dissoluble

higher oxidation state 14–18 . Furthermore, the practical PEMWE devices
appeal harsher demands to OER catalysts with the capacity of rapid
gas-liquid mass transfer and good interface conductivity 19,20 .

Herein we report an in-situ reduction and phase-separation
method for synthesizing the Ir/MnOx catalyst with Ir nanoclusters
embedded within two-dimensional porous MnOx matrix. We reveal
that the abundant Ir-O-Mn bonds near the reconstructed IrOx/MnOx
interfaces, confirmed by in-situ X-ray absorption fine structures (XAFS)
and surface-enhanced Raman spectroscopy (SERS), can provide efficient charge transfer channels for the generation and stabilization of
active Ir 3+ species. Because of the optimal d-band structure of lowvalence Ir state, the IrOx/MnOx exhibits comparable OER activity and
stability, with a notable mass activity of 4 A mgIr−1 at the overpotential

of 300 mV, which is 150.6 times higher than that of IrO2 and a comparable S-number of 6.8 × 10 5 in all the reported Ir-metal-based

1School of Materials Science and Engineering, Peking University, Beijing, China. 2Beijing National Laboratory for Condensed Matter Physics, Institute of Physics
Chinese Academy of Science, Beijing, China. 3 Petrochemical Research Institute, PetroChina, Beijing, China. 4 These authors contributed equally: Dawei Wang,
Fangxu Lin. [e-mail: lvfan@pku.edu.cn; guosj@pku.edu.cn](mailto:lvfan@pku.edu.cn)

catalysts. Notably, the PEMWE device utilizing Ir/MnOx as anodic catalyst layer with a low 0.2 mgIr cm −2 loading achieves a high mass
activity of 6.69 A mgIr−1 @ 1.7 V and a stable operation for 300 h at a

current density of 1 A cm −2 with a very high S-number of 1.48 × 10 8 .
Such impressive performances in PEMWE are not only attributable to
the abundant and enduring Ir 3+ species, but also stem from the
enhanced mass transfer and optimized interfacial contact enabled by
the porous and ultra-thin 2D nanosheet structure. This synergy
underscores the formidable potential of the material for PEMWE under
stringent operating conditions, as evidenced by the comparably low
decay rate (1.24 mV per 1000 cycles at 2 A cm −2 ) in accelerated stress
tests (ASTs).

Results and discussion
The heterointerface-rich Ir/MnOx catalyst was synthesized by the use of
an advanced in-situ reduction and phase-separation method with
iridium-manganese bimetallic oxide (Ir-Mn BO) nanosheets as the precursors. Transmission electron microscopy (TEM, Supplementary Fig. 1)
and atomic force microscopy (Supplementary Fig. 2) images of the asformed nanosheets indicate that they show the spatially cross-linking
morphology with an ultrathin thickness of mere 1.0 nm. High-angle
annular dark-field scanning transmission electron microscopy (HAADFSTEM, Supplementary Fig. 3) with the energy dispersive spectrometer
(EDS) elemental mappings (Supplementary Fig. 4) confirms the homogeneous atomic-level dispersion of Ir throughout the Ir-Mn BO lattice
even under a rather high Ir content up to 33 at.% (Supplementary Fig. 5).
During the following annealing process under H2/Ar atmosphere
(Fig. 1a), the uniformly-distributed high-valence (near +4) Ir ions within
the lattice of Ir-Mn BO (Supplementary Fig. 6) experience sufficient
reduction to generate Ir atoms near the surface. Driven by spontaneous
formation of oxygen vacancies, the as-formed Ir atoms generally diffuse
outward and gather as Ir nanoclusters, thus creating abundant heterointerfaces between metal and oxide matrix. This hypothetic phaseseparation process was verified by X-ray diffraction patterns (XRD,
Fig. 1b; Supplementary Fig. 7) that the original characteristic peaks of
Mn2O3 phase for Ir-Mn BO vanish after annealing while the typical signals of fcc-Ir (located at 40.7° and 69.1°, assigned to (111) and (220)
planes, respectively) emerge for Ir/MnOx. Moreover, an easilyoverlooked widened diffraction peak around 60.0° attributed to the
(224) facets of Mn3O4 suggests a phase transformation from the crude
Mn2O3 matrix to near amorphous Mn3O4 phase induced by the H2/Ar
treatment, further confirmed by X-ray photoelectron spectroscopy
(XPS) spectra with decreased Mn 2p binding energy and increased
splitting degree of Mn 3s peaks (Supplementary Fig. 8) as well as the
Raman spectra with dominant Mn-O stretching vibration peak exclusively indexing the Mn3O4 phase (Fig. 1c; Supplementary Fig. 9).

Because of the structural robustness of MnOx matrix, the ultrathin
two-dimensional morphology with reticular stereo configuration
remains for Ir/MnOx after 300 °C annealing (Fig. 1d) while its average
thickness is measured to be around 1.67 nm (Supplementary Fig. 10), a
little bit thicker than that of the precursor. High-resolution HAADFSTEM image (Fig. 1e) and corresponding EDS elemental mappings
(Fig. 1f) further prove the formation of Ir nanoclusters with an average
size of approximately 1.25 nm (Supplementary Fig. 11). The specific insitu separation of Ir nanoclusters inevitably induces chaotic atomic
arrangement, leading to abundant heterointerfaces and local pores
within Ir/MnOx (Fig. 1e). A d-spacing, measured as 0.191 nm, can be
ascribed to (200) facets of fcc Ir, well consistent with the fast Fourier
transform (FFT) pattern (Fig. 1g, obtained from the selected region in
Supplementary Fig. 12) observed from [011] zone axis. In addition, an
extra pair of reciprocal spots also show up, representative as (224)
planes of Mn3O4, which is in accord with the XRD, XPS and Raman
results discussed above. Further analysisonFFT pattern provides more
specific facet information, reflecting an exposed (1–10) plane of Mn3O4
(Supplementary Fig. 13). To clarify the detailed structure of the

heterointerface, STEM-EDS mappings (Supplementary Fig. 14) as well
as HAADF-STEM image (Fig. 1h) from the side view of the Ir/MnOx were
captured, mutually delineating the distribution of Ir clusters deeply
embedded within the MnOx substrate lattice, which implies intensive
coupling with rather large interface areas. More advanced, clear lattice
fringes can be seen for both Mn3O4 matrix and Ir clusters in Fig. 1h and
Supplementary Fig. 15. Since the {220} plane of Mn3O4 share a similar
d-spacing with that of {111} facets of Ir, one can explicitly observe the
coherent grain boundary near the interfaces, speculating well-fitted
epitaxial growth of Ir clusters derived from the phase-separation synthetic method, underscoring the uniform existence of strong interfacial couplings. Besides, the atomic ratio of Mn to Ir is determined to
be 67.1 to 32.9 for Ir/MnOx, revealing no obvious composition variation
before and after annealing (Supplementary Figs. 16 and 17). As such,
the well-designed Ir/MnOx catalyst containing reticularly porous
matrix and rich heterointerface-anchored Ir nanoclusters with high
loading was successfully obtained.

To delve deeper into the support-metal interaction between insitu formed Ir nanoclusters and Mn oxide substrate near the heterointerfaces, a reference sample, Ir+MnOx, was prepared by synthesizing
free-standing Ir nanoclusters with similar size, and then physically
loading them onto Mn oxide nanosheets (Supplementary
Figs. 18 and 19). XPS was subsequently employed to analyze the chemical state of Ir in both Ir+MnOx and Ir/MnOx. We find that Ir/MnOx has
a lower electron density of Ir (Supplementary Fig. 20), and also has the
stronger electronic interaction between MnOx and Ir due to the coupling heterointerfaces than that of Ir+MnOx. Furthermore, control
experiments with different annealing temperatures were carried out
(Fig. 2a) to investigate the physicochemical nature of such intensified
charge transfer wherein a linear negative relation was found between
the Ir 4f7/2 binding energy and the as-applied temperature till 300 °C
(Supplementary Fig. 21). Interestingly, when the annealing temperature continues to rise to 350 °C, the binding energy of Ir 4f7/2 for Ir/
MnOx keeps invariable and finally stabilizes at 61.38 eV, still higher
than those of Ir+MnOx (61.31 eV) and free-standing Ir clusters
(61.29 eV). Given the fact that comparably high binding energy cannot be achieved through simply physical loading process (Supplementary Fig. 22), one can conceive that the stubbornly high Ir 4f7/2
binding energy stems from extra electronic donation of Ir except for
conventional surface oxidation. Ultraviolet photoelectron spectroscopy (UPS) supports the same conclusion with obviously elevated
working function for Ir/MnOx (Fig. 2b) 21,22 . A hypothetic Ir-O-Mn bond
at the heterointerface was deduced as effective electron transfer
channel (The inset of Fig. 2a), and insufficient reduction of Ir can be
ascribed to unquenchable oxygen coordination derived from MnOx
matrix.

The elaborate fitting of extended X-ray absorption fine structures
(EXAFS) was deployed in both k space and R space for precisely
unveiling the local coordination mode of as-concepted Ir-O-Mn bonds
(Supplementary Fig. 23). As demonstrated in Fig. 2c, in addition to
metallic Ir-Ir bonds, Ir/MnOx also presents a dominant Ir-O scattering
path, the lengths of which follow the order of Ir/MnOx > Ir+MnOx > IrO2
(Fig. 2d; Supplementary Table 1). Such prolonged Ir-O distance indicates
asymmetrical coordination of oxygen atoms, implying extra connection
of O with another transition metal, confirming the existence of Ir-O-Mn
bonds. In addition, the identical scattering paths of Ir-O coordination to
IrO2 were applied to fit Ir/MnOx, with a rather high coordination number
(CN) of 1.8 in total, while an average equivalent path was adopted for Ir
+MnOx with lower CN of 0.4 (Supplementary Table 1). More delicate
information about atom dispersion and bonding circumstance can be
obtained by introducing additional k-space resolution. Compared to Ir
foil, both Ir+MnOx and Ir/MnOx present more pronounced low-k oscillations (Supplementary Fig. 23), indicating abundant Ir-O coordination.
In addition, the wavelet transformation (WT) patterns illustrate that
although the second shell domain in Ir/MnOx (marked by yellow dash

5 nm 5 nm

5 nm 50 5 nm

nm

Fig. 1 | Synthesis and characterizations ofIr/MnOx. a Schematic illustration of the
synthetic approach for Ir/MnOx. b XRD pattern of Ir/MnOx. c Raman spectrum of Ir/
MnOx. d Dark-field STEM image of Ir/MnOx. e HAADF-STEM image of the top view of

line) has a similar R value to that of Ir-Ir scattering in Ir foil (marked by
white dash line), their k values are quite different, which could be
attributed to the weaker ability of Mn, as a low-atomic element, to
scatter X-ray compared to Ir 23–26 (Fig. 2e). Overall, the EXAFS analysis
together with XPS and UPS data, collectively ensure the abundant
existence of Ir-O-Mn bonds near the heterointerfaces.

Ir/MnOx. The boundaries of heterointerfaces and local pores are marked by pink
and blue dash lines, respectively. f STEM-EDS elemental mappings of Ir/MnOx.
g Typical FFT pattern of Ir/MnOx. h HAADF-STEM image of the side view of Ir/MnOx.

The OER performances of Ir/MnOx and other reference catalysts in
acid medium were evaluated in acidic media using a rotating disk
electrode (RDE; Supplementary Fig. 24). The iR-corrected linear sweep
voltammetry (LSV) curves reveal that Ir/MnOx demonstrates a lower
onset potential of 1.43 V versus RHE than Ir/C (1.47 V) and IrO2 (1.54 V)
(Fig. 3a). Furthermore, the Tafel slope of Ir/MnOx was calculated to be

## a 62.2 b Cut-off region Ir/MnOx e 4 Ir/MnOx

IrO2 (62.21 eV)

Ir/MnOx
Ir+MnOx

## e

62.2

62.1

61.7

61.5

Cut-off region

IrO2 (62.21 eV)

Ir-Mn BO (62.08 eV)

Ir/MnOx

Ir+MnOx (61.31 eV)

61.3
150 200 250 300 350

Annealing temperature (°C)

## c d

IrO2

Ir/MnOx

Ir foil/3

0 2 4 6

(Å)

0 5 10 15 20

Binding energy (eV)

17.33

_WF=h -EB,max_ 16.73 _EB, max_

14 15 16 17 18

1
0 4 8 12 16

k (Å -1 )

Ir foil

1
0 4 8 12 16

k (Å -1 )

1.99

1.98

1.97

1.96

1.95

Binding energy (eV)

IrO2
Ir+MnOx
Ir/MnOx

Fig. 2 | Verification of Ir-O-Mn bonds as electron transfer channels for Ir/MnOx.
a Binding energy of XPS Ir 4f7/2 peak as a function of the annealing temperature of
the precursor with a 95% confidence interval. The inset indicates robust Ir-O-Mn
bonds generated near the heterointerfaces after annealing. b UPS spectra of Ir/

42.5 mV dec −1, lower than those of Ir/C (56.3 mV dec −1 ) and IrO2
(60.7 mV dec −1 ), demonstrating a considerably accelerated OER kinetics
(Fig. 3b). As such, Ir/MnOx exhibits an overpotential of mere 238 mV at a
current density of 10 mA cm −2, roughly 49 mV and 121 mV smaller than
that of commercial Ir/C and IrO2, respectively (Fig. 3c). The mass activity
(MA) of Ir/MnOx even reaches 4 A mgIr−1 at the overpotential of 300 mV,

(MA) of Ir/MnOx even reaches 4 A mgIr−1 at the overpotential of 300 mV,

approximately 16.0 and 150.6 times higher than those of Ir/C
(0.253 A mgIr−1) and IrO2 (0.027 A mgIr−1). The Ir/MnOx also delivers

comparable OER activity in neutral electrolyte (Supplementary Fig. 25).
In addition, the comprehensive electrochemical impedance spectroscopy (EIS) of Ir/MnOx and commercial IrO2 anode across the varying
potential reveals that Ir/MnOx has a notably lower interface charge
transfer resistance (Rct), facilitating a more expeditious electrocatalytic
kinetics during the OER process (Supplementary Fig. 26).

Chronopotentiometry (CP) tests were implemented at a constant
current density of 10 mA cm −2 to evaluate the catalytic stability of Ir/
MnOx. As presented in Fig. 3d, the potential required for Ir/MnOx to
reach 10 mA cm −2 shows negligible elevation during a 200-h test. On
the contrary, Ir/C experiences a rapid increase of overpotential and
collapses in just one day. The origin of the as-promoted OER stability
was unraveled by quantitatively monitoring the leaching amount of Ir
during CP test by the use of inductively coupled plasma mass spectrometer (ICP-MS, Fig. 3e). Ir/C shows severe Ir leaching while the
dissolution rate of Ir in Ir/MnOx is much slower with its stability
number of 6.8 × 10 5, 7.67 times higher than that of Ir/C (8.9 × 10 4 ) 27 .
Moreover, TEM images and EDS mappings after CP test further reveal
the morphological and compositional robustness of Ir/MnOx with no
obvious aggregation of either MnOx nanosheets or Ir nanoclusters
(Supplementary Fig. 27) and only slight changes of Mn/Ir elemental
ratio (Supplementary Figs. 28 and 29), far surpassing the commercial
Ir/C with terrible ripening (Supplementary Fig. 30). Taking the mass
activity and S-number as two main indicators for OER activity and
stability, respectively, Ir/MnOx catalyst with strong coupling effects

MnOx and Ir+MnOx. c Comparison of R-space fitting results of Ir L3-edge EXAFS
spectra among different Ir-based samples. d Ir-O bond length comparison among
different Ir-based samples. e The corresponding wavelet transform patterns of Ir/
MnOx and Ir foil.

induced by Ir-O-Mn bonds undoubtedly stands on the top among the
state-of-the-art OER catalysts (Supplementary Table 2).

To delve into the mechanisms underlying the enhancement of
OER activity and stability, XPS (Supplementary Fig. 31a, b) 10,28,29 and
X-ray absorption near-edge structure (XANES, Supplementary
Fig. 31c, d) 30–33 analysis were conducted to probe the chemical state
changes of Ir/MnOx after the high current density and long-term
electrolysis. Electrochemical oxidation of metallic Ir toward IrOx
occurs rapidly during OER, but it is worth noting that its valence state
has been effectively controlled as low as +3.12. Meantime, the valence
state of Mn rises from +2.75 to nearly +3.48. These phenomena imply
that both Ir and Mn elements inevitably experience the oxidation
under high OER potentials. Due to the effective electron buffer effect
derived from the MnOx substrate, the low-valence Ir 3+ sites with higher
intrinsic OER activity are generated and well maintained.

EXAFS analysis was also carried out to uncover the variation of
local coordination configuration of the catalyst before and after
reconstruction. The EXAFS fitting and wavelet transform patterns
results show that, for Ir, the coordination configuration changes from
fcc-Ir metallic bonds to IrO2-type covalent bonds, implying in-situ
electrochemical oxidation (Supplementary Figs. 32, 34a–c), while for
Mn, its coordination environment does not change significantly,
meaning that its phase remains unchanged during electrochemical
reconstruction (Supplementary Figs. 33, 34e–g). Moreover, the Mn-O
bond lengths, both prior to and subsequent to the reconstruction, are
observed to be slightly shorter than those in Mn3O4 (Supplementary
Table 3). This can be attributed to the compressive strain introduced
into the ultrathin MnOx nanosheets by the in-situ precipitation of Ir
clusters at the interface, which aligns well with the rightward shift of
the corresponding Mn oxide peaks observed in the XRD pattern
(Fig. 1b). Furthermore, the left/right shifting of Ir-(O)-M/Mn-(O)-M
scattering intensity maximum in wavelet transform patterns paralleling with the increase/decrease of Ir-O/Mn-O bond length persuasively

−1) and IrO2 (0.027 A mgIr

Fig. 3 | OER electrochemical performances of Ir/MnOx at room temperature in
0.1 M HClO4 (pH = 1 ± 0.1). a Polarization curves of Ir/MnOx, Ir/C and IrO2 with 90%
iR-drop compensation, with dashed lines representing corresponding polarization
curves without iR-drop compensation. b Tafel slopes of different catalysts. c Mass
activity at the overpotential of 300 mV and the overpotential at 10 mA cm −2 of Ir/

proves the general existence of Ir-O-Mn, ensuring the strong interfacial
coupling in IrOx/MnOx material (Supplementary Fig. 34c, d, g, h).

In addition, the reconstruction evolution of the material was also
characterized using in-situ SERS (Fig. 4a and Supplementary Fig. 35a).
The spectral bands at 360 cm −1 and 650 cm −1 are attributed to the
skeletal vibration and Mn-O stretching vibration of Mn oxide, respectively. Notably, the positions of these two peaks remain unchanged
with increasing voltage and prolonged electrolysis time, confirming
that the Mn oxide matrix undergoes negligible phase transition before
and after electrolysis 34,35 . Additionally, the spectral bands at 548 cm −1

and 790 cm −1 correspond to Ir-O-Ir stretching and Ir-O out-of-plane
vibration, respectively 36,37 . By analyzing the intensity trend of the Ir-OIr stretching peak (Supplementary Fig. 35b, c), we demonstrate that Ir
undergoes oxidative transformation during electrolysis and rapidly
reconstructs to a stable state.

Furthermore, in-situ XAFS studies were also carried out to clarify
the reconstruction process of Ir/MnOx with varying voltage. An analysis of valence states using XANES spectra reveals that as the potential
increases, both Ir and Mn undergo a limited degree of increase in
valence state (Fig. 4b, c). Then, wavelet transformation of EXAFS
provides a more intuitive view of real-time structural evolution, indicating that as the potential rises, Ir transforms from a metallic to an
oxidized state, while the coordination structure of Mn remains almost
unchanged (Supplementary Fig. 36). Moreover, the obvious lower
second maximum intensity in k (≈9 Å −1 ) of Ir L3 edge over Ir/MnOx
compared to that of IrO2 (k ≈ 13 Å −1 ) in the WT pattern verifies the
general existence of Ir-O-Mn during the reconstruction process. Subsequently, the EXAFS of Ir with notable changes in coordination
environment was fitted (Supplementary Fig. 37 and Supplementary
Table 4), demonstrating that as potential increased, the average
lengths of Ir-O and Ir-O-Ir bonds shortened significantly, accompanied
by an increase in coordination number (CN). Conversely, Ir-Ir bonds
were elongated, paralleling with a decrease of CN, confirming the insitu transformation of Ir clusters into IrOx clusters (Supplementary

MnOx, Ir/C and IrO2. d Chronopotentiometry (CP) curves of Ir/MnOx and Ir/C. The
inset shows the calculated S-number. e Metal loading preservation percentage of Ir/
MnOx and Ir/C during CP test. All error bars in Fig. 3 represent the standard
deviation.

Fig. 38). In summary, the active structure of the catalyst can be
described as electrochemically in-situ oxidized IrOx clusters embedded within MnOx, with the active sites being the abundant and longlasting Ir 3+ species derived from the enduring electron transfer from
the MnOx matrix.

Based on the aforementioned reconstructed structure, density
functional theory (DFT) calculation was further performed to evidence
the electron-feeding effect toward the promoted OER activity and
stability for IrOx/MnOx, the model of which was constructed derived
from the experimental observation (Supplementary Fig. 39) by
anchoring individual IrOx clusters onto the surface of MnOx (Supplementary Fig. 40, Supplementary Data 1 and 2). The charge density
difference of IrOx/MnOx model indicates obvious electron cloud
deviation from Mn to Ir near the heterointerfaces (Fig. 4d; Supplementary Fig. 41), proving the as-characterized electron feeding effect.
The projected density of states clearly reveal that the Ir d orbit substantially overlaps with Mn d orbit, suggesting more prominent covalent nature and thus stronger electronic interaction between Ir and Mn
atoms (Supplementary Fig. 42) 26 . Bader charge of Ir in both individual
IrOx and IrOx/MnOx were further calculated to predict their adsorption
behaviors toward multiple oxygenated intermediates during OER. As
one can conclude from Supplementary Fig. 43, IrOx/MnOx showcases
lower Bader charge of Ir under both bare and various adsorption states
than individual IrOx, well consistent with the experimental results that
MnOx matrix contributes to elevated electron density of Ir and thereby
helps realize weaker adsorption of all the OER-involved intermediates.
A step further, the charge-transfer energy defined by the energy gap
between d-band center of Ir and p-band center of O was also calculated
to evaluate the adsorption strengths (Supplementary Fig. 44), where
the higher values of IrOx/MnOx indicate a less overlapping between its
Ir 5 d and O 2p orbits, stressing its optimally modulated d-band
structure with optimally weakened adsorptions 12 .

The Gibbs free energy diagrams along OER reaction coordinate of
the two IrOx-based models were obtained under classic adsorption

*OOH

PDS = 0.74 eV

0.5

0.0

-0.5

-1.0

H2O

IrOx
IrOx/MnOx

PDS = 0.51 eV

*OH
*O

Weak adsorption U = 1.23 V

Reaction coordinate

Fig. 4 | In-situ spectroscopic and theoretical study ofIrOx/MnOx for its catalytic
activity and stability enhancement. a In-situ SERS spectraof Ir/MnOx (IrOx/MnOx)
varying with potential. b Ir white-line peak position as a function of the oxidation
states of Ir for Ir/MnOx (IrOx/MnOx) in-situ operated at different potentials, with Ir
foil (0) and IrO2 (+4) as standards. c Mn K-edge energy as a function of the oxidation
states of Mn for Ir/MnOx (IrOx/MnOx) in-situ operated at different potentials, with

evolution mechanism (AEM) pathway (Supplementary Fig. 45). Subject
to the over-strong binding of *OOH, individual IrOx suffers from
excessive energy barrier of potential determining step (PDS)
( * OOH → * + H + +e − +O2) under both U = 0 V vs. RHE (0.74 eV, Supplementary Fig. 46) and U = 1.23 V vs. RHE (1.97 eV, Fig. 4e) 38 . However,
due to the considerably weakened adsorption energies of multiple
oxygenated intermediates, particularly the * O species, the deprotonation process of * OOH gets well accelerated on IrOx/MnOx, resulting in
its PDS transposing to the conversion from * O to * OOH and its energy
barrier decreasing to 0.51 eV (1.74 eV) under U = 1.23 V (0 V) vs. RHE.
Both the alternation of PDS as well as the dramatically reduced energy
barrier originate from the elevated surface reactivity of Ir through
electron feeding effect and mutually account for the elevated OER
activity of IrOx/MnOx. To explore the origins of promoted OER stability
of IrOx/MnOx, the over-oxidation and leaching process of Ir in both
IrOx and IrOx/MnOx were investigated. In contrast with IrOx (Supplementary Fig. 47), the strong-coupling of IrOx and MnOx in IrOx/MnOx
introduces an extra kinetics-controlled transition state evolution
(Supplementary Fig. 48), which increases the formation energy of Ir
vacancy, effectively suppressing the over-oxidation and leaching of Ir
and thus endowing IrOx/MnOx with excellent OER durability (Fig. 4f).
To conclude, theoretical calculation ensures the indispensable role of
Ir-O-Mn bond-derived electron feeding effect in generating and stabilizing highly active low-valence Ir species for acidic OER
electrocatalysis 12 .

A membrane electrode assembly was constructed (Supplementary Fig. 49) by employing Ir/MnOx as the anodic catalyst for OER and
Pt/C as the cathodic catalyst for HER, to evaluate its behaviors in
practical PEMWE devices.As shown in Fig. 5a, even with mereone tenth
of Ir loading, Ir/MnOx exhibits better performance as its cell voltage
demanded to deliver an industrial current density of 1 A cm −2 reaches
as low as 1.63 V, farther surpassing its commercial counterparts (IrO2

Mn3O4 (+2.67) and MnO2 (+4) as standards. d Side panel of charge density difference on IrOx/MnOx. The yellow and cyan isosurfaces show the charge gain and lost
regions, respectively. e The Gibbs free energy diagrams of the four-electron OER
process on the Ir sites of IrOx/MnOx and individual IrOx under the applied potential
of 1.23 V vs. RHE. f The Gibbs free energy diagrams of the over-oxidation and
leaching of Ir in IrOx/MnOx and IrOx.

with 2 mgIr cm −2 and 0.2 mgIr cm −2, donated as IrO2-2 and IrO2-0.2,
respectively). It is worth noting that Ir/MnOx remains the linear
transcending mode even in the high-current-density region, indicative
of its efficient mass transfer derived from the interlaced porous
structure with densely distributed clusters 39 . Besides, within the range
of 1.5 V to 1.8 V vs. RHE, Ir/MnOx exhibits much higher mass activity
than IrO2-0.2 and IrO2-2, showing promising OER behaviors in practical
devices with desirably low Ir loading (Fig. 5b; Supplementary Table 5),
which makes it a most outstanding candidate among other reported
OER catalysts for PEMWEs (Supplementary Table 6).

Unlike the RDE tests, which mainly reflect the kinetics, the
PEMWE voltage is also influenced by the system’s ohmic and masstransfer resistance. Therefore, overpotential analysis was further
carried out to discern the contributions of kinetic, ohmic, and masstransport properties separately (Fig. 5c; Supplementary
Figs. 50 and 51) 19,20,40 . Substantial reduction of approximately 14.5% in
kinetic overpotential at 2 A cm −2 was observed for Ir/MnOx relative to
IrO2-0.2, indicating the noteworthily enhanced OER kinetics over Ir/
MnOx. Moreover, Ir/MnOx demonstrates a significantly decreased
ohmic overpotential, roughly 29.6% lower than that of IrO2-0.2
(Supplementary Fig. 52, Supplementary Table 7), implying the porous nanosheet structure effectively mitigates the interfacial contact
between the catalyst layer and proton exchange membrane 40 (Supplementary Fig. 53). More importantly, Ir/MnOx exhibits significant
advantages in terms of mass-transport overpotential compared to
IrO2-0.2, particularly at high current densities, with mere 47.2-mV
increase at current density of 2 A cm −2, 73.0% lower than that of IrO20.2. The prominently triggered mass transfer of Ir/MnOx can be
attributed to its reticularly porous structure that enables smooth
flow of reactants and products (Supplementary Fig. 55). Therefore,
the PEMWE device based on Ir/MnOx demonstrates comparable
operation efficiency (Supplementary Fig. 54, Supplementary Table 8)

Fig. 5 | Electrochemical performances of Ir/MnOx on PEMWE devices.
a Polarization curves for Ir/MnOx (with 0.2 mg cm −2 Ir loading), IrO2-0.2 and IrO2-2.
“0.2” and “2” represents Ir loading. b Mass activities of Ir/MnOx and IrO2-2 at various
cell voltage. c Various overpotential contributions to the polarization curves of Ir/

and lower cost per kilogram of H2 (Supplementary Fig. 56) compared
to IrO2-2 and IrO2-0.2 (Supplementary Fig. 57).

The long-term stability is of utmost importance for practical
application of OER catalysts. PEMWE based on Ir/MnOx with an Ir
loading of 0.2 mg cm −2 exhibits stability over 300 h at 1 A cm −2 . In
contrast, IrO2-0.2 suffers from rapid decay after a mere 50-h operation
(Fig. 5d). Over the 300-h operating period, the average energy consumption of Ir/MnOx is quite low, calculated as 43.75 kWh kg −1 H2
(equated to 3.94 kWh m −3 H2 at standard temperature and pressure,
details in Supporting Information) 41 . ICP-MS was also employed to
analyze the ion leaching during PEMWE stability test. The results
indicate minor loss of only 1.81% Ir and 0.42% Mn leaching during 300h test with a high S-number of Ir/MnOx of 1.48 × 10 8 . In addition to
better resistance against ion leaching, the comparable stability of Ir/
MnOx also stems from its notable structural stability, allowing the
preservation of its advantageous architecture (Supplementary Fig. 58)
to realize low ohmic and charge transfer resistance (Supplementary
Fig. 59, Supplementary Table 7). The AST under various Ir loadings
were also used to evaluate the stability of Ir/MnOx-based PEMWE
(Supplementary Fig. 60), showing that Ir/MnOx-based PEMWE still
exhibits excellent durability at lower Ir loading (0.1 mgIr cm −2 ) with an
extremely low decay rate (1.275 mV increase per 1000 cycles at
2 A cm −2 ), much better than that of commercial IrO2 with a higher
loading of 2 mgIr cm −2 (1.365 mV per 1000 cycles at 2 A cm −2 ).

In summary, we reported a highly efficient two-dimensional Ir/
MnOx catalyst with strong-coupling hetero-interfaces and high capacity of mass/charge transfer as an anode catalyst to achieve practical
durable low-iridium PEMWE. Advanced characterizations and theoretical calculations reveal that abundant Ir-O-Mn near the heterointerfaces of the reconstructed IrOx/MnOx catalyst can realize distinct
charge transfer to effectively address the over-oxidation issue of Ir 3+,
resulting in a comparable mass activity (4 A mgIr−1 at the overpotential

MnOx and IrO2-0.2 at different current density. d Chronopotentiometry curves at
current density of 1.0 A cm −2 of Ir/MnOx and IrO2-0.2. All error bars in Fig. 5
represent the standard deviation.

of 300 mV) and a significant S-number (6.8 × 10 5 ). Besides its excellent
kinetic advantages, the unique spatially interlaced porous nanosheet
structure of Ir/MnOx as a catalyst layer confers notable electron and
mass transfer with 29.6% and 73.0% of corresponding overpotential
lower than that of commercial IrO2 at a high working current of
2 A cm −2 in practical PEMWEs. Notably, PEMWE utilizing Ir/MnOx as
anode catalysts with mere 0.2 mgIr cm −2 loading achieves a high mass
activity of 6.69 A mgIr−1 at 1.7 V and can operate stably at an industrial

current density of 1 A cm −2 for 300 h, presenting the top level among
the reported low-Ir-loading electrolyzers.

Methods
Chemicals
Hexachloroiridum acid hydrate (H2IrCl6·H2O, Ir ≥ 35%) was obtained
from Adamas-beta. Manganese chloride (MnCl2) was obtained from
Macklin. Potassium nitrate (KNO3), potassium hydroxide (KOH, 95%)
and perchloric acid (HClO4, 70%) were obtained from Aladdin Reagent
Co. Ltd. Nafion’D-521 dispersion (5 wt.%) was bought from Alfa Aesar.
Isopropanol, ethanol, and cyclohexane were all supplied by Beijing
Tongguang Fine Chemicals Company. Commercial Ir/C (20 wt% of
2–4 nm Ir nanoparticles on carbon black) and IrO2 for RDE testing were
purchased from Premetek Company. IrO2 for PEMWE testing was
purchased from Cotrun New Energy. Manganese acetate Hydrate
(MnAc⋅4H2O) was purchased from 3A Materials. Carbon paper (29 BC)
was purchased from SIGRACET. Carbon cloth was purchased from
Toray. All reagents were used as received without further purification.
The water used in all experiments was ultrapure (18.2 MΩ cm).

Synthesis of Ir-Mn bimetallic oxide precursor
200 μL of 0.1 g mL −1 solution of H2IrCl6·H2O was added to 10 mL of
0.1 M KOH. The mixture was left undisturbed until the solution changed from brown to a nearly colorless state. Next, the resulting liquid

was transferred to a 40 mL glass vial and placed in an oil bath at 80 °C.
The reactant’s color turned light blue during this process. Once the
solution was turned into the aforementioned light blue color, 1 mL of a
10 mg mL −1 solution of MnCl2 was added drop by drop into the mixture
under stirring. The reaction mixture was then kept at 80 °C with continuous stirring for 2 h. The obtained black product was washed three
times with deionized water and once with ethanol, and finally was
dispersed in ethanol.

Synthesis of Ir/MnOx catalyst
The total obtained precursor was mixed with 10 mg carbon black
under ambient sonication for 1 h in ethanol. The resulting product was
collected by centrifugation and dried at 60 °C in an oven for 3 h.
Subsequently, it was calcined in H2/Ar at 300 °C for 1 h to obtain the Ir/
MnOx. To obtain the Ir/MnOx-T, the calcination step was performed at
temperature of T with other conditions remaining unchanged.

Synthesis of Ir + MnOx
Ir clusters and MnOx were synthesized, respectively. As for Ir clusters
of Ir+MnOx, 250 mg CTAB was dissolved in 50 mL ethylene glycol
(EG) and ultrasound for 1 h. Then, 600 μL of 0.1 g/mL solution of
H2IrCl6·H2O was dropwise added to above mixture under vigorous
stirring. Then, the obtained mixture was heat up to 160 °C and
refluxed for 3 h. Finally, the suspension was washed with a mixture of
deionized water and ethanol. After that it was dispersed in ethanol.
With regard to MnOx, MnOx was electrodeposited on carbon cloth.
Before that, the carbon cloth was electrochemically treated in a
three-electrode cell using 0.5 M KNO3 electrolyte. A piece of graphite
foil and a saturated calomel electrode (SCE) were used as the counter
electrode and reference electrode, respectively. A constant potential
of 1.85 V vs. SCE was applied to the working electrode (carbon cloth)
for 30 min to enhance the wettability of the electrode. The as-treated
electrode was further scanned using cyclic voltammetry from −1.0 to
1.0 V vs. SCE for 50 cycles at a scan rate of 100 mV s −1 to recover the
conductivity of the carbon cloth. The electrodeposition of MnOx on
electrochemically treated carbon cloth was conducted in a threeelectrode cell containing 0.3 M manganese acetate (MnAc·4H2O),
using a constant current density of 20 mA cm −2 for 2 min. The prepared ethanol solution of Ir clusters was dropped onto MnOx at room
temperature.

Synthesis of Ir/MnOx anode for the durability test
Appropriate amount of precursor without Nafion ionomer was sprayed
onto carbon paper (1.0 × 1.5 cm 2 ). Subsequently, it was calcined in H2/
Ar at 300 °C for 1 h to obtain Ir/MnOx anode. The specific catalyst
loading was calculated to be 0.056 mgIr cm −2 (based on ICP-OES).

Electrochemical measurements in a three-electrode system
The electrochemical measurements of OER were conducted in a
three-electrode system, utilizing a RDE apparatus (Pine Research
Instrumentation, the United States) connected to CHI750e electrochemical workstation (Shanghai Chenhua Instrument Corporation,
China). In the three-electrode setup, a glassy carbon RDE (area:
0.196 cm 2 ) coated with catalysts served as the working electrode,
while a Pt foil (dimensions: 1.0 × 1.5 cm 2 ) functioned as the counter
electrode, and a SCE acted as the reference electrode. 2 mg of Ir/
MnOx was dispersed in 1 mL of a mixture of ultrapure water, isopropanol and Nafion’D-521 solution (volume ratio = 1:1: 0.04) by
sonication for 1 h to get a catalyst ink. The same procedure was followed to prepare a catalyst ink of commercial Ir/C and IrO2. Next, the
ink was dropped onto the glassy carbon RDE, and the specific catalyst
loading was calculated to be 0.056 mgIr cm −2 for each catalyst (based
on ICP-OES). For electrochemical assessments in acidic media, fast
cyclic voltammograms (CVs, 500 mV/s, 0.05−1.4 V, in 0.1 M N2-saturated HClO4 solution (freshly prepared and promptly utilized (pH ~

1)), at a rotating speed of 1600 rpm) were conducted to activate the
catalysts until stable CVs were obtained, and then slow CVs at 50 mV/
s from 0.05 V to 1.4 V (in 0.1 M N2-saturated HClO4 solution, without
rotating) were recorded. OER polarization curves were recorded by
LSV in 0.1 M O2-saturated HClO4 solution at the scan rate of 5 mV/s
with 90% iR drop compensation at a rotating speed of 2000 rpm. The
onset potential was defined as the intersection potential of the tangent line and the baseline 42 . EIS measurements were recorded across
the varying potential and a frequency range from 100,000 to
0.01 Hz. Chronopotentiometry (CP) based on three-electrode system
was utilized to assess the stability of electrocatalysts at a constant
current density of 10 mA cm −2 in 0.1 M HClO4 solution. In the system,
Ir/MnOx anode was used as working electrode, Pt foil (1.0 × 1.5 cm 2 )
was used as counter electrode and SCE was used as reference electrode. The dissolution of metal elements from the catalyst into the
electrolyte solution during the durability test was quantified by collecting 2 mL electrolyte at regular intervals for the ICP-MS test.
Subsequently, 2 mL of 0.1 M HClO4 solution was added to the system.
The final calculating results was based on the cumulative dissolution
of metal elements.

Every time before using reference electrodes to conduct electrochemical tests, we calibrate the reference electrodes using the following method 43,44 . The calibration process was carried out in an
electrolyte saturated with high-purity hydrogen, using two Pt plates as
the working electrode and counter electrode, respectively. Cyclic
voltammetry (CV) scans were conducted at a rate of 1 mV/s, and the
thermodynamic potential for the hydrogen electrode reactions was
determined as the average of the two potentials at which the current
crossed zero during these scans, which was used to scale the potential
to reversible hydrogen electrode (RHE).

Synthesis of Ir/MnOx anode for PEMWE
Appropriate amount of precursor without Nafion ionomer was sprayed
onto Pt-plated titanium felt with a porosity of ~50% and a thickness of
0.25 mm. Subsequently, it was calcined in H2/Ar at 300 °C for 1 h to
obtain Ir/MnOx anode.

Electrochemical measurements in PEM electrolyzer
Appropriate amount of Ir/MnOx precursor or IrO2 was sprayed onto Ptplated titanium felt with a porosity of ~50% and a thickness of 0.25 mm,
to prepare anodes using our designed automatic airbrush. Pt/C
(70 wt.%) was used as the cathode catalyst. The ionomer to catalyst (I/
C) ratio at cathode is controlled to be 0.25. The prepared cathode
catalyst’s ink was sprayed onto Nafion membrane (N117) using our
designed automatic airbrush. X-ray fluorescence (XRF) was used to
evaluate the loading of catalyst and uniformity of the catalyst layer.
The Nafion perfluorosulfonic acid membrane was treated by 5 wt%
H2O2, 0.5 M H2SO4 at 80 °C for 1 h each step to remove impurities and
protonate the sulfonic-acid groups before usage. Pt-plated sintered
titanium with a porosity of ≈50% and a thickness of 0.25 mm was used
as porous transport layers (PTL) at cathode. Anode and cathodic catalyst coated membrane were assembled in homemade single cell
hardware with titanium parallel flow field on cathode and platinized
titanium parallel flow field on anode. The PEM electrolyzer was operated at the temperature of 80 °C using distilled water as the electrolyte
solution under the flowing rate of 40 mL min −1 cm −2 . The EIS was carried out at 1 A cm −2 by applying an alternated signal of 50 mA over the
frequency range from 10 kHz to 1 Hz. The stability of the PEM electrolyzer was evaluated by measuring chronopotentiometry at 1 A cm −2 .
The dissolution of metal elements in the anode solution during the
300 h test was quantified by ICP-MS.

Characterizations
TEM was performed using a HITACHI HT7700 instrument, operated at
an accelerating voltage of 100 kV. XRD patterns were acquired using a

Rigaku D/max-2500 powder diffractometer, employing Cu-Kα radiation with a wavelength of 0.154 nm. HAADF-STEM and energydispersive X-ray spectroscopy (EDS) element mappings were carried
out on an aberration-corrected JEOL 2200FS STEM/TEM microscope,
which was operated at 300 kV and equipped with a Bruker-AXS SDD
detector. The composition and concentration of the catalysts were
analyzed using an inductively coupled plasma optical emission spectrometer (ICP-OES), specifically the Agilent 5110 model. The mass
loading on CCM is determined by X-ray Fluorescence Spectrometer.
The dissolution of metal elements from the anode catalyst into the
electrolyte solution during the test was quantified by ICP-MS. UPS and
XPS measurements were conducted on an XPS AXIS Ultra DLD (Kratos
Analytical).

Raman spectra were obtained using a HORIBA-Xplora Plus confocal microprobe Raman system. A 50× objective with an 8 mm
working distance was utilized for the measurements. The excitation
source was a 532 nm He-Ne laser with an approximately 6 mW of
power. Calibration of the Raman frequencies was performed with a Si
wafer spectrum. Each displayed spectrum was collected over a 120 s
acquisition time and represents the average of two measurements. Insitu Raman experiments were performed in a membrane separation
flow cell using a CHI760E electrochemical workstation, using a counter
electrode and a reference electrode, in an electrolyte solution with a
volume of approximately 20 mL. The in-situ XAS experiments were
conducted under the same conditions as those used for the electrocatalytic characterization, employing a single cell that had its opening
securely sealed with Kapton tape 45 .

S-number calculation
The S-number was calculated based on the following equation:

S � number =

nO2

nIr

ð1Þ

where nO2 is the amount of oxygen molecules produced during the
durability test (mol), according to the Faraday’s laws of electrolysis:

nO2 = j × z × s × F t

ð2Þ

G(U) = G(0 V) - neU, where e represents the elementary charge of an
electron, n is the number of proton-electron pairs transferred, and U is
the applied potential. The Gibbs free energy (ΔG) was determined
using the equation provided below:

ΔG = ΔE + ΔEZPE            - TΔS ð3Þ

where the value of ΔE, ΔEZPE and ΔS represents the variations in DFT
energy, zero-point energy and entropy at a temperature of 298.15 K,
respectively 46,47 .

Overpotential analysis
The cell voltage, Ecell, consists of the reversible cell potential E 0 rev and

three primary overpotentials: kinetic overpotential (ηkin), ohmic
overpotential (ηΩ), and mass transport overpotential (ηmt). Given
that the hydrogen evolution reaction (HER) exhibits more preferable
kinetics and mass transport under current testing conditions compared to the OER, the overpotential analysis focuses solely on the
OER side.

Ecell = E 0 rev + η kin + η Ω + η mt ð4Þ

The temperature-dependent, standard reversible potential for the
water splitting reaction is computed using the following formula:

E 0 rev = 1:2291 � 0:0008456 × ð T � 298:15 Þ ð5Þ

Under the current electrolyzer testing conditions, the voltage is
calculated to be 1.168 V, with a thermoneutral voltage of 1.42 V.

EIS was utilized to assess the high-frequency resistance (HFR),
which signifies the overall electronic resistance of the cell. Consequently, the ohmic overpotential (ηΩ) can be determined as follows:

ηΩ = i × Rtot = i × HFR ð6Þ

The kinetic overpotential is derived using the Tafel model, where
the Tafel slope b and exchange current density i0 are the key kinetic
parameters. Under the assumption of a non-polarizable HER, the entire
kinetic overpotential of the cell is determined by OER. The Tafel slope
b for OER is given as 2:303 × RT= 4F, where R is the ideal gas constant, T
is the temperature, F is Faraday’s constant.

-
i
ηkin = b × log ð7Þ
i0

The mass transport encompasses the transfer of gases and liquids
within the porous transport layer (PTL) and catalytic layer (CL), as well
as the ionic transport within the CLs. In this research, it is determined
by deducting the reversible cell potential, along with kinetic and ohmic
overpotentials, from the observed cell potential 40 .

Operation efficiency to produce H2
The operation efficiency to produce H2 was calculated based on the
following equation:

H2 power out
Operation efficiency= ð8Þ
Electrolyzer power

In the equation above, electrolyzer power wascalculated based on
the following equation:

Electrolyzer power = U × j ð9Þ

where U is the voltage of PEMWE electrolyzer (V), j is the current
density (A cm −2 ).

where j is the current density (A cm −2 ), s is the geometry area of the
electrode (cm 2 ), t is the electrolysis time (s), z is the number of electrons transferred to produce an oxygen molecule (=4), F is the Faraday
constant (96484 C mol −1 ). nIr is the total amount of Ir dissolved in the
electrolyte (mol).

Theoretical computation
We utilized the Vienna Ab initio Simulation Package (VASP) for conducting all DFT computations, adopting the generalized gradient
approximation (GGA) framework with the Perdew-Burke-Ernzerhof
(PBE) functional. In our approach, we selected the projected augmented wave (PAW) potentials to represent the ionic cores, while
accounting for the valence electrons through a plane wave basis set,
imposing a kinetic energy cutoff of 450 eV. To address the highly
localized 3 d states in Mn, the DFT + U method was employed with
specific parameters of U - J = 3.9 eV. Additionally, the spin-polarization
effect was taken into account. The transition states were identified
using the climbing image nudged elastic band (CI-NEB) method,
adhering to a consistent convergence criterion. The free energy
changes (ΔG) for each fundamental step in the OER were estimated
using the computational hydrogen electrode (CHE) model. Here, the
chemical potential aligns with half the energy of gaseous H2 at 0 V
relative to the reversible hydrogen electrode (RHE). During an electron
transfer step, the electrode potential (U relative to RHE) is accounted
for by subtracting eU from the free energy. This can be expressed as

H2 power out was calculated based on the following equation:

3. King, L. A. et al. A non-precious metal hydrogen catalyst in a commercial polymer electrolyte membrane electrolyser. Nat. Nanotechnol. 14, 1071–1074 (2019).
4. Chen, S. et al. Reconstructed Ir-O-Mo species with strong Brønsted
acidity for acidic water oxidation. Nat. Commun. 14, 4127 (2023).
5. Liu, R.-T. et al. Recent advances in proton exchange membrane
water electrolysis. Chem. Soc. Rev. 52, 5652–5683 (2023).
6. Minke, C., Suermann, M., Bensmann, B. & Hanke-Rauschenbach, R. Is
iridium demand a potential bottleneck in the realization of large-scale
PEMwaterelectrolysis? Int.J.HydrogenEnergy 46, 23581–23590(2021).
7. Chen, Z. et al. Advances in oxygen evolution electrocatalysts for
proton exchange membrane water electrolyzers. Adv. Energy
Mater. 12, 2103670 (2022).
8. Yin, J. et al. Iridium single atoms coupling with oxygen vacancies
boosts oxygen evolution reaction in acid media. J. Am. Chem. Soc.
142, 18378–18386 (2020).
9. Zheng, Y.-R. et al. Monitoring oxygen production on mass-selected
iridium-tantalum oxide electrocatalysts. Nat. Energy 7, 55–64 (2022).
10. Kim, E.-J. et al. Stabilizing role of Mo in TiO2-MoOx supported Ir
catalyst toward oxygen evolution reaction. Appl. Catal. B Environ.
280, 119433 (2021).
11. Lee, C. et al. Catalyst-support interactions in Zr2ON2-supported IrOx
electrocatalysts to break the trade-off relationship between the
activity and stability in the acidic oxygen evolution reaction. Adv.
Funct. Mater. 33, 2301557 (2023).
12. Shi, Z. et al. Enhanced acidic water oxidation by dynamic migration
of oxygen species at the Ir/Nb2O5-x Catalyst/Support Interfaces.
Angew. Chem. Int. Ed. 61, e202212341 (2022).
13. Ge, S. et al. A robust chromium-iridium oxide catalyst for highcurrent-density acidic oxygen evolution in proton exchange membrane electrolyzers. Energy Environ. Sci. 16, 3734–3742 (2023).
14. BalaKrishnan, A. et al. Direct detection of surface species formed on
iridium electrocatalysts during the oxygen evolution reaction.
Angew. Chem. Int. Ed. 60, 21396–21403 (2021).
15. Li, T. et al. Atomic-scale insights into surface species of electrocatalysts in three dimensions. Nat. Catal. 1, 300–305 (2018).
16. Lončar, A., Escalera-López, D., Cherevko, S. & Hodnik, N. Interrelationships between oxygen evolution and iridium dissolution
mechanisms. Angew. Chem. Int. Ed. 61, e202114437 (2022).
17. Kasian, O., Grote, J.-P., Geiger, S., Cherevko, S. & Mayrhofer, K. J. J.
The common intermediates of oxygen evolution and dissolution
reactions during water electrolysis on iridium. Angew. Chem. Int.
Ed. 57, 2488–2491 (2018).
18. Lv, H. et al. N/C doped nano-size IrO2 catalyst of high activity and
stability in proton exchange membrane water electrolysis. Int. J.
Hydrogen Energy 48, 16949–16957 (2023).
19. Dong, S. et al. Overall design of anode with gradient ordered
structure with low iridium loading for proton exchange membrane
water electrolysis. Nano Lett. 22, 9434–9440 (2022).
20. Pham, C. V., Escalera-López, D., Mayrhofer, K., Cherevko, S. &
Thiele, S. Essentials of high performance water electrolyzers – from
catalyst layer materials to electrode engineering. Adv. Energy
Mater. 11, 2101998 (2021).
21. Zheng, X. et al. Managing grains and interfaces via ligand anchoring
enables 22.3%-efficiency inverted perovskite solar cells. Nat.
Energy 5, 131–140 (2020).
22. He, D. et al. Active electron density modulation of Co3O4-based
catalysts enhances their oxygen evolution performance. Angew.
Chem. Int. Ed. 59, 6929–6935 (2020).
23. Chang, J. et al. Improving Pd-N-C fuel cell electrocatalysts through
fluorination-driven rearrangements of local coordination environment. Nat. Energy 6, 1144–1153 (2021).
24. Liu, H. et al. Iridium doped pyrochlore ruthenates for efficient and
durable electrocatalytic oxygen evolution in acidic media. Small 18,
2202513 (2022).

H2 power out =

j × MH2 × q

n × F

ð10Þ

where j is the current density (A cm −2 ), MH2 is molar mass of H2
(2 g mol −1 ), F is the Faraday constant (96484 C mol −1 ), n is the the
number of electrons transferred to produce a hydrogen molecule (=2),
q is the heating value of H2 where the lower heating value
(LHV = 1.185 × 10 5 J g −1 ) and the higher heating value
(HHV = 1.43 × 10 5 J g −1 ) are applied to calculate operation efficiency,
respectively.

Energy consumption per kilogram of H2 calculation
The Energy consumption per kilogram of H2 (kWh kg −1 H2) was calculated based on the following equation:

Energy consumption per kilogram of H2 = U × j × A × t ð11Þ

1000 × mH2

where U is the voltage of PEMWE electrolyzer (V), j is the current
density (A cm −2 ), A is the effective working area of electrolyzer (cm 2 ), t
is the electrolysis time (h). The average energy consumption per kilogram of H2 is calculated by integrating U with respect to t. mH2 is mass
of H2 (kg) based on the following equation:

mH2 =

j × A × t × MH2 ð12Þ

1000 × n × F

where j is the current density (A cm −2 ), A is the effective working area
of electrolyzer (cm 2 ), t is the electrolysis time (s), MH2 is molar mass
of H2 (2 g mol −1 ), n is the the number of electrons transferred to
produce a hydrogen molecule (=2), F is the Faraday constant
(96484 C mol -1 ).

Energy consumption per cubic meter of H2 calculation
The Energy consumption per kilogram of H2 (kWh m −3 H2) was calculated based on the following equation:

Energy consumption per cubic meter of H2 =

ð13Þ
Energy consumption per kilogram of H2 × ρ

where ρ is the density of H2 (0.09 kg m −3 ) at standard temperature and
pressure.

Energy cost per kilogram of H2 calculation
Energy cost per kilogram of H2 (USD kg −1 H2) was calculated based on
the following equation:

Cost per kilogram of H2 =
ð14Þ
energy consumption per kilogram of H2 × electricity bill

where electricity bill obtained from the previous reports is USD
0.02 kWh −1 .

Data availability
The data that support the findings of this study are available in the
supplementary material of this article. Source data are provided with
this paper.

References
1. Lagadec, M. F. & Grimaud, A. Water electrolysers with closed and
open electrochemical systems. Nat. Mater. 19, 1140–1150 (2020).
2. Kibsgaard, J. & Chorkendorff, I. Considerations for the scaling-up of
water splitting catalysts. Nat. Energy 4, 430–433 (2019).

25. Zhang, X.-L. et al. Efficient acidic hydrogen evolution in proton
exchange membrane electrolyzers over a sulfur-doped marcasitetype electrocatalyst. Sci. Adv. 9, eadh2885 (2023).
26. Zhu, Y. et al. Iridium single atoms incorporated in Co3O4 efficiently
catalyze the oxygen evolution in acidic conditions. Nat. Commun.
13, 7754 (2022).
27. Geiger, S. et al. The stability number as a metric for electrocatalyst
stability benchmarking. Nat. Catal. 1, 508–515 (2018).
28. Weng, Y. et al. High-valence-manganese driven strong anchoring of
iridium species for robust acidic water oxidation. Adv. Sci. 10,
2205920 (2023).
29. Nie, Z. et al. Catalytic kinetics regulation for enhanced electrochemical nitrogen oxidation by Ru-Nanoclusters-coupled Mn3O4
catalysts decorated with atomically dispersed Ru atoms. Adv.
Mater. 34, 2108180 (2022).
30. Shi, Z. et al. Confined Ir single sites with triggered lattice oxygen
redox: toward boosted and sustained water oxidation catalysis.
Joule 5, 2164–2176 (2021).
31. Nong, H. N. et al. A unique oxygen ligand environment facilitates
water oxidation in hole-doped IrNiOx core–shell electrocatalysts.
Nat. Catal. 1, 841–851 (2018).
32. Chang, H.-W. et al. Nanoflaky MnO2/functionalized carbon nanotubes for supercapacitors: an in situ X-ray absorption spectroscopic
investigation. Nanoscale 7, 1725–1735 (2015).
33. Li, A. et al. Enhancing the stability of cobalt spinel oxide towards
sustainable oxygen evolution in acid. Nat. Catal. 5, 109–118 (2022).
34. Bernardini, S. et al. Raman spectra of natural manganese oxides. J
Raman Spectrosc. 50, 873–888 (2019).
35. Guo, W. et al. Mismatching integration-enabled strains and defects
engineering in LDH microstructure for high-rate and long-life
charge storage. Nat. Commun. 13, 1409 (2022).
36. Pavlovic, Z. et al. Probing the structure of a water-oxidizing anodic
iridium oxide catalyst using Raman Spectroscopy. ACS Catal. 6,
8098–8105 (2016).
37. Chen, A.-Y. et al. Ultrafine iridium nanoparticles prepared without a
surfactant for the acidic oxygen evolution reaction. Mater. Chem.
Front. 7, 4900–4907 (2023).
38. Kwon, J. et al. Tailored electronic structure of ir in high entropy alloy
for highly active and durable bifunctional electrocatalyst for water
splitting under an acidic environment. Adv. Mater. 35,
2300091 (2023).
39. Retuerto, M. et al. Highly active and stable OER electrocatalysts
derived from Sr2MIrO6 for proton exchange membrane water
electrolyzers. Nat. Commun. 13, 7935 (2022).
40. Chatterjee, S. et al. Nanoporous iridium nanosheets for polymer
electrolyte membrane electrolysis. Adv. Energy Mater. 11,
2101438 (2021).
41. Xu, J. et al. IrOx·nH2O with lattice water–assisted oxygen exchange
for high-performance proton exchange membrane water electrolyzers. Sci. Adv. 9, eadh1718 (2023).
42. Wang, H. et al. High-valence oxides for high performance oxygen
evolution electrocatalysis. Adv. Sci. 10, 2301706 (2023).
43. Liang, Y. et al. Co3O4 nanocrystals on graphene as a synergistic
catalyst for oxygen reduction reaction. Nat. Mater. 10,
780–786 (2011).
44. Fan, L. et al. Atomically isolated nickel species anchored on graphitized carbon for efficient hydrogen evolution electrocatalysis.
Nat. Commun. 7, 10667 (2016).
45. Wang, N. et al. Boride-derived oxygen-evolution catalysts. Nat.
Commun. 12, 6089 (2021).
46. Xu, Y. et al. Constructing asymmetric dual active sites through
symbiotic effect for achieving efficient and selective photoreduction of CO2 to C2H4. Energy Environ. Sci. 17, 5060–5069 (2024).

47. Zhai, J. et al. Photo-thermal coupling to enhance CO2 hydrogenation toward CH4 over Ru/MnO/Mn3O4. Nat. Commun. 15,
1109 (2024).

Acknowledgements
This study was financially supported by National Natural Science Foundation of China (52333010 (S.G.)), National Science Fund for Distinguished Young Scholars (52025133 (S.G.)), Beijing Outstanding Young
Scientist Program (JWZQ20240102004 (S.G.)), National Natural Science
Foundation of China (22309004 (F.Lv.), 22479002 (F.Lv.)), China National
Petroleum Corporation-Peking University Strategic Cooperation Project
of Fundamental Research (S.G.), Tencent Foundation through the
XPLORER PRIZE (S.G.). The authors thank the BL11B, BL14W1, and BL20U
photoemission photoendstations in the Shanghai Synchrotron Radiation
Facility (SSRF) and the 1W1B beamline at Beijing Synchrotron Radiation
Facility (BSRF) for the help with XAFS characterizations.

Author contributions
D.W. and F.Lv. designed research; D.W. performed research; H.L., J.Z.,
W.Z., and Y.Wang. offered assistance in experimental operation; Y.Wei,
Q.Z., and L.G. contributed analytic tools; L.L. performed computation;
D.W. wrote the first draft of the paper; F.Lin, F.Lv., M.L., and S.G. revised
the paper.

Competing interests
The authors declare no competing interests.

Additional information
Supplementary information The online version contains
supplementary material available at
[https://doi.org/10.1038/s41467-024-54646-8.](https://doi.org/10.1038/s41467-024-54646-8)

Correspondence and requests for materials should be addressed to
Fan Lv or Shaojun Guo.

Peer review information Nature Communications thanks Prashanth
Menezes, Dipak Shinde and the other anonymous reviewer(s) for their
contribution to the peer review of this work. A peer review file is available.

Reprints and permissions information is available at
[http://www.nature.com/reprints](http://www.nature.com/reprints)

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons
Attribution-NonCommercial-NoDerivatives 4.0 International License,
which permits any non-commercial use, sharing, distribution and
reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if you modified the licensed
material. You do not have permission under this licence toshare adapted
material derived from this article or parts of it. The images or other third
party material in this article are included in the article’s Creative
Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons
licence and your intended use is not permitted by statutory regulation or
exceeds the permitted use, you will need to obtain permission directly
[from the copyright holder. To view a copy of this licence, visit http://](http://creativecommons.org/licenses/by-nc-nd/4.0/)
[creativecommons.org/licenses/by-nc-nd/4.0/.](http://creativecommons.org/licenses/by-nc-nd/4.0/)

본문 출처 PDF: PDF 보기 · 다른 PDF는 위 [PDF 업로드]로 올릴 수 있습니다.

활용가능성

이력 0건

아직 편집 이력이 없습니다. 저장할 때마다 이전 내용이 여기에 백업됩니다.