Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# Boride-derived oxygen-evolution catalysts

## Abstract

Metal borides/borates have been considered promising as oxygen evolution reaction catalysts; however, to date, there is a dearth of evidence of long-term stability at practical current densities. Here we report a phase composition modulation approach to fabricate effective borides/borates-based catalysts. We find that metal borides in-situ formed metal borates are responsible for their high activity. This knowledge prompts us to synthesize NiFe-Boride, and to use it as a templating precursor to form an active NiFe-Borate catalyst. This boride-derived oxide catalyzes oxygen evolution with an overpotential of 167 mV at 10 mA/cm2 in 1 M KOH electrolyte and requires a record-low overpotential of 460 mV to maintain water splitting performance for over 400 h at current density of 1 A/cm2. We couple the catalyst with CO reduction in an alkaline membrane electrode assembly electrolyser, reporting stable C2H4 electrosynthesis at current density 200 mA/cm2 for over 80 h.

## Introduction

The design of active, stable, and cost-effective catalysts for the oxygen evolution reaction (OER) is critical to improving the full process efficiency in electrocatalytic systems, such as water splitting and CO2/CO reduction reaction (CO2/CORR) coupled with OER1,2,3,4. In these systems, approaches that use alkaline electrolytes have the benefit of increasing efficiency toward CO2RR and CORR by suppressing the competing hydrogen evolution reaction (HER).

While there have been improvements in the activity of metal phosphides, sulfides, selenides, and borides/borates5,6,7,8, achieving high electrocatalytic activity (>1 A/cm2) united with long-term stability (>100 h) remains an unmet need9,10. The lack of highly active and stable OER catalysts limits the performance and stability of electrosynthesis processes that rely on an OER anode (i.e., CO2RR/CORR electrolysers) to production rates <100 mA/cm2 and durations of <100 h. It remains an important task for the field to increase catalytic activity while achieving long-term durable and earth-abundant OER catalysts that operate at alkaline pH.

Metal borides/borates exhibit promising OER performance in high-pH regimes: incorporating highly electronegative boron diminishes the metal oxidation reaction energy barrier under a positive bias and assists with charge transfer11,12. To date, the durability of metal borides has been limited to <60 h at a modest current density of ~20 mA/cm2,13,14,15. Thus far, to improve the activity and stability, significant efforts have been invested in charge-transfer enhancement and electronic structure optimization; however, we found herein that metal borides are prone to convert into metal borate under OER conditions, and thus to behave as the operando catalyst16,17,18.

This knowledge prompted us to design metal borides as precursors to construct, in-situ, the metal borates that would endow the catalysts with activity and stability.

We focused on a NiFe-Boride compound and probed its phase transformation using operando X-ray spectroscopy during OER. We witnessed the in-situ transformation of NiFe-Boride to NiB4O7 and FeBO3 phases on the catalyst surface during electrochemical activation. These results accord with DFT calculations which reveal that NiB4O7 species facilitate the step *O - > *OOH and enhance electrocatalytic activity toward OER.

Experimentally, we achieve an overpotential of 167 mV at 10 mA/cm2 and stable water splitting performance of 400 h at 1 A/cm2. Encouraged by its high activity and stability, we implement NiFe-Boride as the OER catalyst in a CORR membrane-electrode assembly (MEA) electrolyser that uses CO and H2O to produce C2H419. The CORR electrolyser based on the NiFe-Boride OER catalyst exhibited stable performance at 200 mA/cm2 for over 80 h—maintaining a C2H4 energy efficiency of 19%—whereas this value is 17% for the CORR electrolyser that relies on a noble IrO2 OER catalyst.

## Results

### Insight from computational studies into stable and active phases

Noting that metal borides undergo phase transformation at highly oxidative potentials, we sought to calculate Pourbaix diagrams to evaluate the phase composition of NiFe-Boride ternary system at high potentials in alkaline electrolytes. Compared to the Pourbaix diagram of NiFe (Fig. 1a), we observe two new stable phases, FeBO3 and NiB4O7, under a working potential (U = 1.4—2 V vs. RHE) at electrolyte pH (pH = 10—14) of alkaline OER in the Pourbaix diagram of NiFe-Boride (Fig. 1b), indicating that boron ions promote the new phases formation. The atomic configurations of these two new phases are shown in the Fig. 1c.

We carried out density functional theory (DFT) calculations to further explore the effects of these in-situ formed phases on the OER activity of NiFe catalyst. We first calculated the reaction energy of elementary steps and the overpotential for the OER based on the 4e-mechanism proposed by Nørskov et al. for water oxidation20,21,22 with the reaction pathway:

$$\ast +{{{{{{\rm{OH}}}}}}}^{-}\to \ast {{{{{\rm{OH}}}}}}+{e}^{-}$$
(1)
$$\ast {{{{{\rm{OH}}}}}}+{{{{{{\rm{OH}}}}}}}^{-}\to \ast {{{{{\rm{O}}}}}}+{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}+{e}^{-}$$
(2)
$$\ast {{{{{\rm{O}}}}}}+{{{{{{\rm{OH}}}}}}}^{-}\to \ast {{{{{\rm{OOH}}}}}}+{{{{{{\rm{e}}}}}}}^{-}$$
(3)
$$\ast {{{{{\rm{OOH}}}}}}+{{{{{{\rm{OH}}}}}}}^{-}\to {{{{{{\rm{O}}}}}}}_{2}+{{{{{{\rm{H}}}}}}}_{2}{{{{{\rm{O}}}}}}+{{{{{{\rm{e}}}}}}}^{-}$$
(4)

We cleaved the (100) facet of these two new phases to study the adsorption energy of various intermediates for OER, and thus to determine the theoretical overpotentials as an indicator for catalytic activity (Fig. 1d and Supplementary Fig. 1). The largest reaction energy (Step 2, deprotonation of *OH) was identified as the potential-determining step (PDS)23, as shown in Fig. 1e, and the reaction energy was similar for both FeBO3 (0.53 V) and NiOOH (0.56 V) at 1.23 V. In contrast, NiB4O7 shows a thermodynamic overpotential of 0.43 V, lowest among NiOOH, NiFeOOH and FeBO3, which suggests that NiB4O7 phases are the active sites in NiFe-Boride that enable superior catalytic OER performance. We observed that the NiB4O7 shifts the PDS to the *OOH formation step at 1.23 V. We also obtained the similar calculation results at 1.40 V (Fig. 1e and Supplementary Fig. 2).

To account for other possible reaction mechanisms, we further studied the intramolecular oxygen coupling mechanism24: we found that Ni and Fe borate have increased activity compared to Ni and Fe (oxy)hydroxide (Supplementary Fig. 3). We also considered more sophisticated activity descriptors for multiple-electron processes, exploring the use of electrochemical-step symmetry index (ESSI) and the PDS at zero overpotential (Gmax)25,26,27: these results indicated higher OER activities of these NiFe-borate phases (Supplementary Table 1). Taken together, these results indicate that boron addition enables formation of new active phases, especially the Ni phase, in the pristine framework, thereby enhancing electrocatalytic activity toward OER.

### Catalyst synthesis and characterization

To test these hypotheses experimentally, we synthesized NiFe-Boride catalysts by a chemical reduction method (see Methods for details). Transmission electron microscopy (TEM), scanning electron microscopy (SEM), and X-ray diffraction (XRD) were conducted to investigate the structure, composition, and morphology of the NiFe-Boride catalyst. TEM and SEM images (Supplementary Figs. 4 and 5) revealed that catalysts the consisted of agglomerated nanoparticles. Elemental mapping analysis indicated that Ni, Fe, B, and O elements distribute homogeneously, without obvious separation or aggregation (Fig. 2a and Supplementary Fig. 6). A broad XRD diffraction peak between 40° and 50° indicates low crystallinity in the catalyst (Fig. 2b).

We further carried out X-ray photoelectron spectroscopy (XPS) to study the chemical state of the NiFe-Boride catalyst. XPS survey spectra show the presence of Ni, Fe, B, and O, which agrees well with the elemental mapping results (Supplementary Fig. 7). The B 1 s spectrum obtained from the pristine sample was comprised of two peaks located at 192 eV and 187 eV, which are assigned to B-O and B0/B-M species (Fig. 2c), respectively, indicating boride. The Ni and Fe K-edge X-ray absorption near-edge structure (XANES, Fig. 2d, e) revealed that NiFe-Boride possesses an electronic structure that differs from that of the control samples, indicating that the incorporation of boride leads to phase and structural changes. To probe the origins of these changes, we obtained a Fourier transform (FT) of the extended X-ray absorption fine structure (EXAFS, Fig. 2f, g) spectrum. NiFe-Boride showed two FT features: (1) the peak at ~1.5 Å represents the single-scattering path of M-O and (2) the peak at ~2.5 Å results from the single-scattering path of the closest neighboring transition metals around the metal cations (M-M) (the locations of the peaks match the locations of those in the NiFe control sample)28. The altered electronic structure suggests that the boride ions were incorporated into the NiFe framework via the formation of BOx and NiFeBx species. Further investigations of the M-B peak remain a topic of interest.

Using inductively coupled plasma optical emission spectrometry (ICP-OES) analysis, we determined the molar ratio of Ni:Fe to be 1:1. Further analysis of the as-deposited system with the aid of energy-dispersive X-ray spectroscopy (EDS) indicated a 1:1 atomic ratio of Ni:Fe (Supplementary Fig. 8).

### In-situ and ex-situ characterization

Using in-situ synchrotron radiation X-ray diffraction (SRXRD), we tracked changes in the crystal structure of catalysts during the OER process. We found that, prior to OER, the phase structure of the catalyst was NiFeBO4 (see Supplementary Fig. 9 for the model structure). We found that FeBO3 and NiB4O7 were formed under an applied potential (Fig. 3a, the standard patterns are provided in the Supplementary Fig. 10), consistent with in-situ formation of new phases in the NiFe-Boride catalyst during OER. We then carried out in-situ XAS to monitor the local electronic structure changes of the metal sites during the OER. The Ni and Fe K-edge XANES (Fig. 3b, c) spectra exhibited a positive energy shift, i.e., the valence states of Ni and Fe increased with the applied potential29. We analyzed the absorption edge energies in order to characterize the Ni/Fe valence states. Before OER, the valence states of Ni and Fe in NiFe-Boride were 1.91 and 2.73, respectively. During OER, the valence states of Ni and Fe increased to 1.98 and 2.95, respectively (Supplementary Figs. 1113). The ex-situ L-edge XAS results were in good agreement with this change: the reacted catalyst – compared to the pristine one – had a higher Ni3+ ratio and a similar Fe2+/3+ ratio. The similarity in Fe2+/3+ ratios indicates that Fe ions return to their initial states upon release of the applied potential. These results accord with the higher Ni3+ ratio in the reacted catalyst observed via O K-edge XAS (Supplementary Fig. 14).

We further analyzed the EXAFS spectra (Fig. 3d, e) to clarify the local structural variations associated with the Ni and Fe sites30,31. Under open circuit potential (OCP) in KOH electrolyte, the FT peaks demonstrated features similar to those obtained from the as-prepared catalysts. Once a positive bias was applied, the progressive FT peaks show that the original phase accomplishes some structural transformations. We conducted EXAFS fitting (Supplementary Tables 2–3 and Supplementary Figs. 15, 16) to determine the bond lengths of M-O and M-M (M is Ni or Fe). The interatomic distances of Ni-O and Ni-Ni(Fe) in as-prepared NiFe-Boride are 2.04 Å and 3.09 Å, respectively, while those measured during OER are 2.02 Å and 2.98 Å. The interatomic distances of Fe-O and Fe-Fe(Ni) in as-prepared NiFe-Boride are 2.23 Å and 3.22 Å, while those measured during OER are 1.92 and 3.05 Å. As oxidation and phase transformation progress, the M-O and M-M bonds become shorter than those at OCP. These results are in line with the generation of transition metal active sites with high valence29. The electronic structure of NiFe-Boride is different than that of Ni-Boride and Fe-Boride, indicating the presence of electronic communication between the Ni and Fe in M-M bonds (Supplementary Fig. 17).

We carried out XPS to study the local evolution of chemical state and electronic structure before vs. after OER, with all peaks calibrated using C 1 s (Supplementary Fig. 18). High-resolution XPS in Ni 2p region showed two main peaks of Ni2+ 2p3/2 and Ni2+ 2p1/2 located at 855.4 and 873.3 eV (Fig. 3f), accompanied by satellite peaks centered at 861.2 and 879.4 eV32. In the Fe 2p region, two main peaks located at 710.7 and 724.6 eV were assigned to Fe 2p3/2 and Fe 2p1/2 (Fig. 3g). We noted minor peaks centered at ~851.9 and 706.1 eV in the Ni 2p and Fe 2p spectra of fresh catalyst, suggesting the existence of Ni-B and Fe-B bonds in NiFe-Boride (Fig. 3f, g)33, consistent with EXAFS. The B 1 s spectrum consists of two peaks corresponding to B-O and B0/B-M species. Upon completion of electrochemical activation during OER, negative chemical shifts in the binding energy of Ni 2p and Fe 2p peaks were observed compared to pristine NiFe-Boride. These negative shifts indicated the modulation of the surface electronic structure through the oxidation process, in accordance with the positive shift in the peak position of B-O bonds (Supplementary Fig. 19). Harsh oxidative conditions during OER converted the M-B species into oxides and hydroxides, evidenced by the disappearance of Ni-B, Fe-B, and B0/B-M signals.

In sum, NiFe-Boride acts as a precursor that phase transforms at highly oxidative potentials, leading to the formation of stable FeBO3 and NiB4O7 phases during the OER process.

### Electrocatalytic performance

The OER performance of NiFe-Boride and controls on Ni foam was examined in 1 M KOH electrolyte using a three-electrode cell at room temperature (see Methods). As shown in linear sweep voltammetry (LSV, Fig. 4a) curves, the overpotential of NiFe-Boride catalyst at 10 mA/cm2,34 and without iR compensation is 167 mV. This is lower than the 200 mV and 189 mV obtained for sol-gel NiFe and IrO2, respectively (Supplementary Table 4). Rotating ring disk measurements show a similar trend: NiFe-Boride enables higher-rate water oxidation performance than in control catalysts (Supplementary Fig. 20). To seek insight into the effect of in-situ active sites on performance, we compared the catalytic activity between annealed and pristine catalysts. The annealed catalysts have lower performance, consistent with an activity-promoter role for the in-situ formed active sites (Supplementary Figs. 21-22). A Ni:Fe ratio of 1:1 (Supplementary Fig. 23 and Table 5) leads to the lowest overpotential. We determined the Tafel slope (Fig. 4b), finding 25 mV/dec for the NiFe-Boride catalyst, lower than for IrO2 (41 mV/dec) and NiFe (66 mV/dec). The lower Tafel slope value indicating rate-determining step is shifted from the *O step to the *OOH formation step35,36. Compared to benchmark OER catalysts from literature, NiFe-Boride shows higher catalytic performance in terms of overpotential at 10 mA/cm2 and Tafel slope in the alkaline solution (Fig. 4c). From Nyquist plots based on electrochemical impedance spectroscopy (EIS, Fig. 4d) at applied potential 1.45 V vs. RHE reveal that NiFe-Boride exhibits a lower charge-transfer resistance (Rct) than does either NiFe or IrO2.

We then used electrochemically active surface area (ECSA) obtained via the double-layer capacitance (Cdl, Supplementary Fig. 24) technique to estimate a normalized current density37. This for NiFe-Boride is 7.5 mA/cm2, >2x greater than that in IrO2, and 4x greater than in NiFe catalysts, at 1.45 V vs. RHE (Supplementary Fig. 25). IR-corrected LSV excludes the effect of ionic conductivity on performance: the NiFe-Boride catalyst showed an iR-corrected overpotential of 160 mV at 10 mA/cm2 on Ni foam (Supplementary Fig. 26), which is 35 and 23 mV lower than for NiFe and IrO2.

We measured the stability of the best-performing NiFe-Boride catalyst at 20 mA/cm2 and found that the increase in potential was less than 10 mV after 100 h of continuous water splitting (Fig. 4e), more stable than the control and reference samples (Fig. 4f). We obtained stability at 1 A/cm2, finding that catalysts enable stable operation of 400 h with an overpotential of 460 mV without degradation (Fig. 4g). B 1 s XPS results indicate that the B concentration gradually decreases over the first 100 h but then remains constant over the ensuing 300 h (Supplementary Fig. 27). The Faradaic efficiency (FE) toward O2 remains 98 ± 2% over the first 400 h, excluding a major current contribution associated with materials corrosion.

To investigate further the potential relevance of these catalysts to future systems, we implemented NiFe-Boride as the OER catalyst in a CORR MEA electrolyser that produces C2H4 from CO and H2O (Supplementary Fig. 28). The CORR electrolyser exhibited stable performance at 200 mA/cm2 for over 80 h, maintaining a cell potential of −2.85 V, a C2H4 FE of 45%, and a C2H4 energy efficiency of 19%. Under similar operating conditions, the cell potential for the IrO2 OER catalyst-based electrolyser was −3.06 V, which is ~200 mV higher than in the NiFe-Boride OER catalyst-based system (Fig. 4h and Supplementary Fig. 29), corresponding to a C2H4 energy efficiency of 17%. Using XPS and EDS line scanning, we found that Ni, Fe, and B elemental signals remained following the stability study (Supplementary Figs. 30 and 31)., i.e., the borate remained in the NiFe.

## Discussion

Starting from DFT calculations, we saw that the phase transformation of NiFe-Boride to NiFe-Borate can be activated under OER potentials in OER. The in-situ formed new phase (NiB4O7) can function as active site to facilitate *O - > *OOH, enhancing the electrocatalytic activity of NiFe-Boride in OER. We evaluated the electrochemical performance of the catalyst and found that NiFe-Boride possesses a Tafel slope of 25 mV/dec and an overpotential of 167 mV at current density 10 mA/cm2 in 1 M KOH solution, outperforming benchmark NiFe oxide. The in-situ formed NiFe-Borate phases are confirmed via operando spectroscopies. The catalyst demonstrates stable performance over 400 h at 1 A/cm2 in water splitting. We further showcased 80 h of stable C2H4 electrosynthesis using a NiFe-Boride OER catalyst in a CORR MEA electrolyser, with a ~200 mV lower cell potential compared to MEA electrolysers based IrO2 OER catalysts.

## Methods

### I.1. Chemicals

Iron (III) chloride (FeCl3), nickel (II) chloride hexahygrate (NiCl2·6H2O), Sodium borohydride (NaBH4), ethanol (≥99.5%) and acetone (≥99 %) were purchased from Sigma-Aldrich. Iridium (IV) oxide dehydrate (99.99%) was purchased from Alfa Aesar®. All chemicals were used without further purification.

### I.2. Synthesis of NiFe-Boride and control catalysts

Th NiFe-Boride catalysts were synthesized using a modified chemical reduction method. NiCl2·6H2O (332.8 mg) and anhydrous FeCl3 (227.1 mg) were first dissolved in deionized water (DI, 2 mL) in a vial. A solution of NaBH4 (2 mL, 5 M) in water was prepared in a separate vial. All solutions mentioned above were cooled in an ice bath for 10 min to prevent uncontrolled hydrolysis and consideration which may lead to the formation of the precipitate. The Ni and Fe precursors were then added dropwise to NaBH4 to form a black solution. Finally, the prepared solution was further washed and centrifuged with water and ethanol three times before vacuum drying. The samples of different Ni:Fe ratios were synthesized by changing the Ni and Fe precursors concentration.

NiFe catalysts were synthesized using a modified aqueous sol-gel technique. NiCl2·6H2O (2.1 mmol) and anhydrous FeCl3 (0.35 mmol) were first dissolved in ethanol (2 mL) in a vial. A solution of deionized water (DI) (0.18 mL) in ethanol (2 mL) was prepared in a separate vial. All solutions mentioned above were cooled in an ice bath for 2 h to prevent uncontrolled hydrolysis and consideration, which may lead to the formation of precipitate rather than gel formation. The Ni and Fe precursors were then mixed with an ethanol-water mixture to form a clear solution. Propylene oxide (≈ 1 mL) was then slowly added to form a gel. Finally, the prepared NiFe wet-gel was aged for 24 h to promote network formation and immersed in acetone for 5 days before the gel was vacuum dried.

### I.3. Characterization

The morphologies of the catalysts were characterized by SEM (Hitachi S-4800) imaging, and energy-dispersive X-ray spectroscopy mapping operated at 1.5 kV was used to characterize the surfaces of all samples. TEM images and corresponding elemental mapping analysis were conducted using a JEM-2100F. The electronic structure measurements were carried out using XPS (ESCALAB 250Xi). To determine the best Ni/Fe ratio in the NiFe-Boride system, we first dissolved 50 mg of the catalyst into 5 ml aqua regia through microwave digestion. Then, we took out 1 ml of sample and diluted it to 10 mL to test ICP-OES (5110).

In-situ XAS was carried out in the same condition as electrocatalytic characterization case using a modified single cell with an opening sealed by Kapton tape. The XAS signal was collected by Vortex detector in 9BM beamline of Advanced Photon Source (APS) in Argonne National Laboratory. X-ray absorption near-edge spectra (XANES) of Ni and Fe K-edge were collected by silicon drift detector at ambient air in 9BM beamline of Advanced Photon Source (APS) located in the Argonne National Laboratory.

In-situ SRXRD measurements on NiFe-Boride in a typical three-electrode setup as the same condition in electrochemical characterization case were conducted at beamline 01C2 at National Synchrotron Radiation Research Center using a specially designed acrylic container with a 1 mL of gap, sealed by Kepton tape. 16 keV of the incident X-ray (λ = 0.7749 Å) was allowed to perpendicularly transmit through the sample and electrolyte, so that the signal of synchrotron X-ray diffraction of transmission mode would be collected in a large Debye-Scherrer camera and processed using the GSAS-II software 45 to obtain plots of intensity versus 2θ.

### I.4. Catalysts loaded on the electrodes

To load the catalyst on the Ni foam (thickness: 1.6 mm, Sigma), 20 mg of catalyst was dispersed in a mixture consisting of 2 mL water and 2 mL ethanol, followed by the addition of 100 μL Nafion solution. The suspension was sonicated for 30 min to prepare a homogeneous ink. Ni foam with a fixed area of 0.5 × 0.5 cm2 coated with water-resistant silicone glue was drop-casted with 20 μL of the catalyst ink. Ag/AgCl (with saturated KCl as the filling solution) and platinum foil were used as the reference and counter electrodes, respectively.

### I.5. Electrochemical characterization

Electrochemical measurements were performed using a three-electrode system connected to an electrochemical workstation (Autolab PGSTAT302N), equipped with built-in EIS analyzer. The working electrode was Ni foam (thickness: 1.6 mm, Sigma), which was prepared by dispersing 20 mg of catalyst in a mixture containing 1 mL water and 1 mL ethanol, followed by the addition of 100 μL Nafion solution. The suspension was sonicated for 30 min to prepare a homogeneous ink. Cyclic voltammetry (CV) measurements at 50 mV/s were performed prior to recording LSV at 1 mV/s for each sample. EIS measurements were conducted in a static solution at 1.45 V (vs. RHE). The amplitude of the sinusoidal wave was 10 mV, and the frequency scan range was between 100 kHz and 0.01 Hz. All experiments were performed at ambient temperature (23 ± 2 °C), and electrode potential was converted to the RHE scale using Eq. (5)

$$E\left({vs}.{RHE}\right)=E({vs}.{Ag}/{AgCl})+{E}_{{Ag}/{AgCl}}({vs}.{NHE})+0.059\times {pH}$$
(5)

### MEA test

The MEA cell is a complete CO electrolyser (Dioxide Materials) consisted of a titanium anode flow field, 904 L Stainless Steel cathode flow field, and associated nuts, bolts and insulating kit. The geometric area of each flow field is 5 cm2, of which 45% is the channel while the rest 55% is the land area. The catalyst electrode was attached on the cathode by a copper tape, which was protected by insulating Kapton tapes to avoid electrical contact with membranes or electrolytes. A Sustainion® AEM membrane (Fumasep) was activated in 1 M aqueous KOH solution for 24 h, washed with water prior to use. The anode consisted of NiFe-Boride supported on titanium felt was prepared by a dip-coating and thermal decomposition method. The cathode is the high surface area Cu catalyst on hydrophobic carbon paper. Aqueous 2 M KOH electrolyte was used as the anolyte and was circulated using a peristaltic pump. The flow rate of the anolyte was kept at 20 mL/min. The flow rate of the CO gas flowing into the cathodic gas chamber was kept at 80 sccm by a mass flow controller. The CORR was initiated by applying a constant current density of 200 mA/cm2, and the current density was kept constant over the course of electrolysis.

### IR correction

All the polarization curves on different supports were corrected for ohmic losses (including the wiring, substrate, catalyst, and solution resistances). The iR-corrected data were given by the following equation:

$${E}_{{Co}{rrected}}=E-{iR}$$
(6)

where R is the series resistance of measurement, which can be obtained from an EIS Nyquist plot as the first intercept of the main arc with the axis.

### ECSA

The ECSA of each catalyst was determined by measuring the electrochemical double-layer capacitances (Cdl) from the scan rate CV-dependence plot. The CV cycle potential window was between 0.7 and 0.8 V vs. RHE, and the scan rates were 20, 40, 60, 80, 100, and 120 mV/s. The Cdl was estimated $$\triangle j=\frac{1}{2}(j_{{charge}}-{J}_{{off\; charge}})$$ at the average potential in the selected range against the scan rates. The slope of the linear fit was calculated as the Cdl. A specific capacitance of 40 μF/cm2 was used here. The ECSA of the catalyst was calculated by using the following equation,

$${ECSA}=\frac{{C}_{{dl}}}{40\mu F/{{cm}}^{2}}{{cm}}_{{ECSA}}^{2}$$
(7)

### DFT calculations

We performed the DFT calculations using the Vienna Ab Initio Simulation Package (VASP) code38,39. The exchange-correlation energy was modelled using Perdew-Burke-Ernzerhof (PBE) functional40 within the generalized gradient approximation (GGA). The projector augmented-wave (PAW) pseudo-potentials41 were used to describe the ionic cores. The cut-off energy of 450 eV was adopted after a series of tests. A Methfessel–Paxton smearing of 0.05 eV to the orbital occupation was applied for the geometry optimization and total energy computations. In all calculations, atoms at all positions were assumed to possess Hellmann-Feynman forces lower than 0.02 eV/Å, and the electronic iteration convergence criterion was 10-5 eV using the normal algorithm. Pourbaix diagrams were calculated using Atomic Simulation Environment (ASE)42 with input formation energy values from Materials project43. Surface models for NiB4O7 and FeBO3 were simulated by five-layer (2 × 2) NiB4O7 (100) and FeBO3 (100) supercells (Supplementary Fig. 32), accompanied with a sufficient vacuum gap of 15 Å. Structural optimizations were performed on all modified slab models. We considered 4-steps intermediates adsorption to simulate the OER process (details in below & Supplementary Fig. 33). During adsorption calculations, the top two layers were fully relaxed, while the other layers were fixed at the tested lattice position.

The reaction energies (ΔG) were obtained by

$$\varDelta {{{{{\rm{G}}}}}}=\varDelta {{{{{{\rm{E}}}}}}}^{{{{{{\rm{DFT}}}}}}}+\varDelta {{{{{\rm{ZPE}}}}}}-{{{{{\rm{T}}}}}}\varDelta {{{{{\rm{S}}}}}}$$
(8)

where ΔEDFT is the reaction energy calculated from DFT; ΔZPE is the zero-point energy, which was neglected here; ΔS is the change in entropy, and for gaseous molecules ΔS values are obtained from the standard database in the NIST web-book44.

Taking the reaction (1) as an example, the reaction energy is equals to

$$\varDelta {{{{{{\rm{G}}}}}}}^{1}={{{{{\rm{G}}}}}}(\ast {{{{{\rm{OH}}}}}})\mbox{-}{{{{{\rm{G}}}}}}(\ast )\mbox{-}G({{{{{{\rm{OH}}}}}}}^{-})$$
(9)

where G(*OH) is the total energy of *OH adsorption configuration; G(*) is the energy of catalyst surface; G(OH-) is the energy of hydroxyl ion, to calculate this value, we assume the equilibrium

$${H}_{2}{{{{{\rm{O}}}}}}\leftrightarrow {H}^{+}+{{OH}}^{-}$$
(10)

which relates the chemical potentials as

$${\mu }_{{H}^{+}}+{\mu }_{{{OH}}^{-}}={\mu }_{{H}_{2}{{{{{\rm{O}}}}}}}$$
(11)

Then,

$${\mu }_{{H}^{+}}+{\mu }_{{e}^{-}}+{\mu }_{{{OH}}^{-}}-{\mu }_{{e}^{-}}={\mu }_{{H}_{2}{{{{{\rm{O}}}}}}}$$
(12)

Thus,

$${\mu }_{{{OH}}^{-}}-{\mu }_{{e}^{-}}={\mu }_{{H}_{2}{{{{{\rm{O}}}}}}}-({\mu }_{{H}^{+}}+{\mu }_{{e}^{-}})$$
(13)

here, $${\mu }_{{H}^{+}}+{\mu }_{{e}^{-}}$$ can be calculated using computational hydrogen electrode (CHE) model developed by Nørskov and co-workers45,46. Then, the theoretical overpotential $$\eta$$ for OER can be calculated by

$$\eta ={{{{{\rm{max }}}}}}\left(\Delta {G}^{i}\right)-1.23$$
(14)

$$\Delta {G}^{i}$$ stands for the reaction energy for reaction (8) – (11), i = 1, 2, 3, 4.

Electrochemical-step symmetry index (ESSI) and maximum energy Gmax can act as a sophisticated activity descriptors for multiple-electron processes. We calculated ESSI using the following formula,

$${ESSI}=\frac{1}{n}\mathop{\sum }\limits_{k=1}^{n}\left(\frac{\triangle {G}_{k}^{+}}{e}\right)$$
(15)

where $$\triangle {G}_{k}^{+}$$ means all positive values of reaction energies, while negative values are not accounted.

The Gmax is the free-energy spanning from intermediate with the smallest free energy to the intermediate with the highest energy during OER.

## Data availability

The data supporting this study are available within the paper and the Supplementary Information. The source data of DFT calculations are provided in this paper. All other relevant source data are available from the corresponding authors upon reasonable request. Source data are provided with this paper.

## References

1. Yu, Y., Shi, Y. & Zhang, B. Synergetic transformation of solid inorganic–organic hybrids into advanced nanomaterials for catalytic water splitting. Acc. Chem. Res. 51, 1711–1721 (2018).

2. You, B. & Sun, Y. Innovative strategies for electrocatalytic water splitting. Acc. Chem. Res. 51, 1571–1580 (2018).

3. Yang, M. Q., Wang, J., Wu, H. & Ho, G. W. Noble metal-free nanocatalysts with vacancies for electrochemical water splitting. Small 14, e1703323 (2018).

4. Wang, Y. & Yan, D. El Hankari, S., Zou, Y. & Wang, S. Recent progress on layered double hydroxides and their derivatives for electrocatalytic water splitting. Adv. Sci. 5, 1800064 (2018).

5. Li, H. et al. Earth‐abundant iron diboride (FeB2) nanoparticles as highly active bifunctional electrocatalysts for overall water splitting. Adv. Energy Mater. 7, 1700513 (2017).

6. Yu, F. et al. High-performance bifunctional porous non-noble metal phosphide catalyst for overall water splitting. Nat. Commun. 9, 2551 (2018).

7. Panda, C. et al. From a molecular 2Fe-2Se precursor to a highly efficient Iron diselenide electrocatalyst for overall water splitting. Angew. Chem. Int. Ed. 56, 10506–10510 (2017).

8. Yang, Y. et al. MoS2–Ni3S2 heteronanorods as efficient and stable bifunctional electrocatalysts for overall water splitting. ACS Catal. 7, 2357–2366 (2017).

9. Shiva Kumar, S. & Himabindu, V. Hydrogen production by PEM water electrolysis – A review. Mater. Sci. Energy Technol. 2, 442–454 (2019).

10. Jouny, M., Luc, W. & Jiao, F. General techno-economic analysis of CO2 electrolysis systems. Ind. Eng. Chem. Res. 57, 2165–2177 (2018).

11. Bediako, D. K., Surendranath, Y. & Nocera, D. G. Mechanistic studies of the oxygen evolution reaction mediated by a nickel-borate thin film electrocatalyst. J. Am. Chem. Soc. 135, 3662–3674 (2013).

12. Masa, J. et al. Amorphous cobalt boride (Co2B) as a highly efficient nonprecious catalyst for electrochemical water splitting: oxygen and hydrogen evolution. Adv. Energy Mater. 6, 1502313 (2016).

13. Nsanzimana, J. M. V. et al. Ultrathin amorphous iron–nickel boride nanosheets for highly efficient electrocatalytic oxygen production. Chem. Eur. J. 24, 18502–18511 (2018).

14. Su, L. et al. Borate-ion intercalated NiFe layered double hydroxide to simultaneously boost mass transport and charge transfer for catalysis of water oxidation. J. Colloid Interface Sci. 528, 36–44 (2018).

15. Gupta, S., Patel, M. K., Miotello, A. & Patel, N. Metal boride‐based catalysts for electrochemical water‐splitting: A review. Adv. Funct. Mater. 30, 1906481 (2020).

16. Mabayoje, O., Shoola, A., Wygant, B. R. & Mullins, C. B. The role of anions in metal chalcogenide oxygen evolution catalysis: electrodeposited thin films of nickel sulfide as “pre-catalysts”. ACS Energy Lett. 1, 195–201 (2016).

17. Nsanzimana, J. M. V. et al. Facile synthesis of amorphous ternary metal borides–reduced graphene oxide hybrid with superior oxygen evolution activity. ACS Appl. Mater. Interfaces 11, 846–855 (2019).

18. Xu, X., Song, F. & Hu, X. A nickel iron diselenide-derived efficient oxygen-evolution catalyst. Nat. Commun. 7, 1–7 (2016).

19. De Luna, P. et al. What would it take for renewably powered electrosynthesis to displace petrochemical processes? Science 364, 6438 (2019).

20. Bajdich, M., Garcia-Mota, M., Vojvodic, A., Norskov, J. K. & Bell, A. T. Theoretical investigation of the activity of cobalt oxides for the electrochemical oxidation of water. J. Am. Chem. Soc. 135, 13521–13530 (2013).

21. Dickens, C. F., Kirk, C. & Nørskov, J. K. Insights into the electrochemical oxygen evolution reaction with ab Initio calculations and microkinetic modeling: Beyond the limiting potential volcano. J. Phys. Chem. C. 123, 18960–18977 (2019).

22. Rossmeisl, J., Logadottir, A. & Nørskov, J. K. Electrolysis of water on (oxidized) metal surfaces. Chem. Phys. 319, 178–184 (2005).

23. Koper, M. T. Analysis of electrocatalytic reaction schemes: distinction between rate-determining and potential-determining steps. J. Solid State Electrochem 17, 339–344 (2013).

24. Romain, S., Bozoglian, F., Sala, X. & Llobet, A. Oxygen− oxygen bond formation by the Ru-Hbpp water oxidation catalyst occurs solely via an intramolecular reaction pathway. J. Am. Chem. Soc. 131, 2768–2769 (2009).

25. Govindarajan, N., García-Lastra, J. M., Meijer, E. J. & Calle-Vallejo, F. Does the breaking of adsorption-energy scaling relations guarantee enhanced electrocatalysis? Curr. Opin. Electrochem 8, 110–117 (2018).

26. Exner, K. S. A universal descriptor for the screening of electrode materials for multiple-electron processes: beyond the thermodynamic overpotential. ACS Catal. 10, 12607–12617 (2020).

27. Exner, K. S. Why approximating electrocatalytic activity by a single free‐energy change is insufficient. Electrochim. Acta 375, 137975 (2021).

28. Masa, J. et al. Ultrathin high surface area nickel boride (NixB) nanosheets as highly efficient electrocatalyst for oxygen evolution. Adv. Energy Mater. 7, 1700381 (2017).

29. Görlin, M. et al. Oxygen evolution reaction dynamics, faradaic charge efficiency, and the active metal redox states of Ni–Fe oxide water splitting electrocatalysts. J. Am. Chem. Soc. 138, 5603–5614 (2016).

30. Wang, H. Y. et al. Ni3+‐inducedformation of active NiOOH on the spinel Ni–Co oxide surface for efficient oxygen evolution reaction. Adv. Energy Mater. 5, 1500091 (2015).

31. Hung, S.-F. et al. In situ spatially coherent identification of phosphide-based catalysts: crystallographic latching for highly efficient overall water electrolysis. ACS Energy Lett. 4, 2813–2820 (2019).

32. Weng, B. et al. A layered Na1−xNiyFe1−yO2 double oxide oxygen evolution reaction electrocatalyst for highly efficient water-splitting. Energy Environ. Sci. 10, 121–128 (2017).

33. Hong, W., Sun, S., Kong, Y., Hu, Y. & Chen, G. NixFe1−xB nanoparticle self-modified nanosheets as efficient bifunctional electrocatalysts for water splitting: experiments and theories. J. Mater. Chem. A 8, 7360–7367 (2020).

34. Wang, X. et al. Self-constructed multiple plasmonic hotspots on an individual fractal to amplify broadband hot electron generation. ACS Nano 15, 10553–10564 (2021).

35. Marshall, A. T. & Vaisson-Béthune, L. Avoid the quasi-equilibrium assumption when evaluating the electrocatalytic oxygen evolution reaction mechanism by Tafel slope analysis. Electrochem. Commun. 61, 23–26 (2015).

36. Sun, S. et al. Switch of the rate-determining step of water oxidation by spin-selected electron transfer in spinel oxides. Chem. Mater. 31, 8106–8111 (2019).

37. Li, H. et al. Amorphous nickel-cobalt complexes hybridized with 1T-phase molybdenum disulfide via hydrazine-induced phase transformation for water splitting. Nat. Commun. 8, 15377 (2017).

38. Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558 (1993).

39. Kresse, G. & Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169 (1996).

40. Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865 (1996).

41. Payne, M. C., Teter, M. P., Allan, D. C., Arias, T. & Joannopoulos, A. J. Iterative minimization techniques for ab initio total-energy calculations: molecular dynamics and conjugate gradients. Rev. Mod. Phys. 64, 1045 (1992).

42. Larsen, A. H. et al. The atomic simulation environment-a Python library for working with atoms. J. Phys.: Condens. Matter 29, 273002 (2017).

43. Jain, A., Castelli, I. E., Hautier, G., Bailey, D. H. & Jacobsen, K. W. Performance of genetic algorithms in search for water splitting perovskites. J. Mater. Sci. 48, 6519–6534 (2013).

44. Chase, M. W. NIST‐JANAF thermochemical tables for oxygen fluorides. J. Phys. Chem. Ref. Data. 25, 551–603 (1996).

45. Rossmeisl, J., Qu, Z.-W., Zhu, H., Kroes, G.-J. & Nørskov, J. K. Electrolysis of water on oxide surfaces. J. Electroanal. Chem. 607, 83–89 (2007).

46. Nørskov, J. K. et al. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J. Phys. Chem. B 108, 17886–17892 (2004).

47. Cai, W. et al. Amorphous versus crystalline in water oxidation catalysis: A case study of NiFe alloy. Nano Lett. 20, 4278–4285 (2020).

48. Zou, Y. et al. 3D hierarchical heterostructure assembled by NiFe LDH/(NiFe)Sx on biomass-derived hollow carbon microtubes as bifunctional electrocatalysts for overall water splitting. Electrochim. Acta 348, 136339 (2020).

49. Chen, G. et al. An amorphous nickel-iron-based electrocatalyst with unusual local structures for ultrafast oxygen evolution reaction. Adv. Mater. 31, 1900883 (2019).

50. Zhou, Y. et al. Exceptional performance of hierarchical Ni-Fe (hydr)oxide@NiCu electrocatalysts for water splitting. Adv. Mater. 31, 1806769 (2019).

51. Teng, X. et al. Self-Growing NiFe-Based hybrid nanosheet arrays on Ni nanowires for overall water aplitting. ACS Appl. Energy Mater. 2, 5465–5471 (2019).

52. Qiu, Z., Tai, C.-W., Niklasson, G. A. & Edvinsson, T. Direct observation of active catalyst surface phases and the effect of dynamic self-optimization in NiFe-layered double hydroxides for alkaline water splitting. Energy Environ. Sci. 12, 572–581 (2019).

53. Zhang, P. et al. Dendritic core-shell nickel-iron-copper metal/metal oxide electrode for efficient electrocatalytic water oxidation. Nat. Commun. 9, 1–10 (2018). 381.

54. Zhang, J. et al. Single-atom Au/NiFe layered double hydroxide electrocatalyst: probing the origin of activity for oxygen evolution reaction. J. Am. Chem. Soc. 140, 3876–3879 (2018).

55. Dong, C., Kou, T., Gao, H., Peng, Z. & Zhang, Z. Eutectic-Derived mesoporous Ni-Fe-O nanowire network catalyzing oxygen evolution and overall water splitting. Adv. Energy Mater. 8, 1701347 (2018).

56. Zhang, H. et al. Bifunctional heterostructure assembly of NiFe LDH nanosheets on NiCoP nanowires for highly efficient and stable overall water splitting. Adv. Funct. Mater. 28, 1706847 (2018).

57. Cai, Z. et al. Introducing Fe(2+) into Nickel-Iron layered double hydroxide: local structure modulated water oxidation activity. Angew. Chem. 57, 9392–9396 (2018).

58. He, T. et al. Synthesis of amorphous boride nanosheets by the chemical reduction of Prussian blue analogs for efficient water electrolysis. J. Mater. Chem. A 6, 23289–23294 (2018).

## Acknowledgements

The N.W. and H.Y.L. acknowledge support from the National Natural Science Foundation of China (NSFC No. 51771132). The authors acknowledge funding from Canada’s Natural Gas Innovation Fund, the Natural Sciences and Engineering Research Council (NSERC) of Canada, the Natural Resources Canada Clean Growth Program, and the Ontario Research Fund—Research Excellence program. All DFT computations were performed on the Niagara supercomputer of the SciNet HPC Consortium. SciNet is funded by the Canada Foundation for Innovation, the Government of Ontario, Ontario Research Fund Research Excellence Program, and the University of Toronto. The authors thank C.-L. Dong and Y.-R. Lu for technical support at the 20 A beamline in NSRRC. The authors acknowledge Taiwan Light Source (TLS) beamline 17C1, 20A1 and Taiwan Photon Source (TPS) beamline 44A1 of NSRRC for providing synchrotron beamtime.

## Author information

Authors

### Contributions

H.Y.L. and E.H.S. supervised the project. N.W. conceived the idea and carried out the experiments. N.W., H.Y.L., and E.H.S. wrote the paper. A.X. P.O., and Z.W. carried out the DFT computations. Y.X. and J.A. performed the in-situ XAS measurements. M.H. and K.Y. assisted in electrochemical experiments. Z.W. A.S. A.V., and S-F.H. polished the paper. A.O. integrated the OER catalyst into the CORR MEA electrolyser and tested the C2H4 production performance. S-F.H., A.O., and Z.W. assisted in data analysis and manuscript writing. All authors discussed the results and assisted during manuscript preparation.

### Corresponding authors

Correspondence to Yongchang Liu, Edward H. Sargent or Hongyan Liang.

## Ethics declarations

### Competing interests

The authors declare no competing interests.

Peer review information Nature Communications thanks Abdoulaye Djire, Kai Exner, Justus Masa, and the other, anonymous, reviewers for their contributions to the peer review of this work.

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

## Rights and permissions

Reprints and Permissions

Wang, N., Xu, A., Ou, P. et al. Boride-derived oxygen-evolution catalysts. Nat Commun 12, 6089 (2021). https://doi.org/10.1038/s41467-021-26307-7

• Accepted:

• Published:

• DOI: https://doi.org/10.1038/s41467-021-26307-7

• ### Shining Light on Anion-Mixed Nanocatalysts for Efficient Water Electrolysis: Fundamentals, Progress, and Perspectives

• Yaoda Liu
• Paranthaman Vijayakumar
• Zhengfei Dai

Nano-Micro Letters (2022)