ELECTROCATALYST FOR OXYGEN EVOLUTION REACTIONS

Description

FIELD

The specification relates generally to a catalyst for an electrochemical reaction, and more particularly to an electrocatalyst for an oxygen evolution reaction.

BACKGROUND

In the current energy system, traditional fossil fuel sources provide for the bulk of demand, with associated CO₂ emissions that threaten the climate. The capture and utilization of CO₂ offers a route to non-fossil fuels and feedstocks for sustainable society. The electrocatalytic conversion of CO₂ is of interest in view of advantages in selectivity, controllable reaction and mild reaction conditions. However, further development of electrocatalytic CO₂ reduction technology is limited by high operating costs and low catalyst stability—factors that increase capital costs.

SUMMARY

According to an aspect of the present specification an example electrochemical system includes: a cathode including a first electrocatalyst configured to catalyze a reduction reaction of carbon dioxide to produce water; an anode including a second electrocatalyst comprising ruthenium doped iridium oxide, the second electrocatalyst configured to catalyze an oxygen evolution reaction to produce diatomic oxygen from the water; an electrolyte connecting the cathode and the anode; and an electricity source configured to apply an electrical current across the cathode and the anode to catalyze the reduction and oxygen evolution reactions.

According to another aspect of the present specification, an example method includes: providing an electrochemical cell comprising: a cathode including a first electrocatalyst; an anode including a second electrocatalyst comprising ruthenium (Ru) doped iridium oxide (IrOx); an electrolyte connecting the cathode and the anode; and an electricity source configured to apply an electrical current across the cathode and the anode; providing carbon dioxide at the cathode; and applying, via the electricity source, the electrical current to the electrochemical cell, and in response: catalyzing, at the cathode, a reduction reaction of the carbon dioxide to produce water; and catalyzing, at the anode, an oxygen evolution reaction to produce diatomic oxygen from the water.

According to another aspect of the present specification, an example catalyst for an oxygen evolution reaction includes ruthenium (Ru) doped iridium oxide (IrO_x).

BRIEF DESCRIPTION OF DRAWINGS

The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

Implementations are described with reference to the following figures, in which:

FIG. 1 depicts a schematic block diagram of an example electrochemical cell including an ruthenium doped iridium oxide anodic electrocatalyst, in accordance with the present disclosure.

FIG. 2A depicts a flowchart of an example method of preparing ruthenium doped iridium oxide electrocatalyst for use as an anode in an electrochemical cell.

FIG. 2B depicts a schematic diagram of an example performance of the method of FIG. 2A.

FIG. 3 is a schematic block diagram of an experimental electrolyzer setup.

FIGS. 4A and 4B depict GI-XRD patterns of IrO_x and Ru/IrO_x before and after stability tests.

FIG. 5 depicts energy dispersive spectroscopy (EDS) images and spectrum of the Ru/IrO_x sample.

FIG. 6 depicts aberration corrected scanning transmission electron microscopy high angle angular dark field (AC-STEM-HAADF) of Ru0.05 mg/IrO_x before and after a 12 h reaction in 500 mA cm⁻² neutral conditions.

FIG. 7 depicts scanning electron microscope (SEM) images of IrO_x (A-B) and IrO_x/Ru (C-D) before vs after the stability test under 500 mA cm⁻².

FIGS. 8A and 8B depict x-ray photoelectron spectroscopy (XPS) percentage results of Ru-0.05 mg/IrO_x-EG before vs after reaction, and IrO_x-EG before vs after reaction.

FIGS. 9A-9D depict: (FIG. 9A) Ir 4f XPS results of IrO_x before vs after stability test under 500 mA cm⁻²; (FIG. 9B) Ir 4f XPS results of IrO_x/Ru before vs after stability test under 500 mA cm⁻²; (FIG. 9C) O 1s XPS results of IrO_x before vs after stability test under 500 mA cm⁻²; (FIG. 9D) O1s XPS results of Ru/IrO_x before vs after stability test under 500 mA cm⁻².

FIG. 10 depicts a scanning transmission X-ray microscopy (STXM) oxygen K-edge X-ray absorption spectra of Ru-0.05 mg/IrO_x-EG before vs after reaction, and IrO_x-EG before/after reaction.

FIGS. 11A-11D depict: (FIG. 11A) Ir L₃-edge XANES before vs after reaction. (FIG. 11B) FT-EXAFS spectra of Ir foil reference, IrO_x before vs after reaction, and Ru/IrO_x before vs after reaction; (FIGS. 11C-11D) Ru K-edge XANES and FT-EXAFS spectra of Ru foil reference and Ru/IrO_x before vs after reaction.

FIGS. 12A-12H depict: (FIG. 12A) Ir foil reference L-edge (FIG. 12B) Ru foil reference K-edge EXAFS (points) and the curvefit (line); (FIG. 12C) IrO_x before reaction (FIG. 12D) IrO_x after reaction (FIG. 12E) Ru/IrO_x before reaction (FIG. 12F) Ru/IrO_x after reaction; Ru K-edge EXAFS (points) and the curvefit (line) for (FIG. 12G) Ru/IrO_x before reaction (FIG. 12H) Ru/IrO_x after reaction, shown in k³ weighted k-space.

FIGS. 13A-13B depict different electrodeposited IrO_x LSV polarization curves and I-T chronoamperometry curves with corresponding work potentials at 100 mA cm⁻².

FIGS. 14A-14D depict LSV and PT results for IrO_x with/without EG treatment.

FIGS. 15A-15H depict ECSA for IrO_x without and with EG treatment.

FIGS. 16A-16D depict OER performance of Ru doped IrO_x Chronopotentiometry curves during 12 h runs using different Ru loads in Ru/IrO_x at 10 mA cm⁻², 200 mA cm⁻², and 500 mA cm⁻².

FIGS. 17A-17B depicts Chronopotentiometry (PT) curves during 12 h runs with different loads of Ru in Ru/IrO_x at 10 mA cm⁻² and 200 mA cm⁻².

FIG. 18 depicts Chronopotentiometry (PT) curves during 12 h runs using Ru/IrO_x with different Ru loading amounts at 500 mA cm⁻².

FIGS. 19A and 19B depict LSV curves normalized by geometric area for samples with different Ru loads before vs after the 12 h stability test under 500 mA cm⁻² and ECSA analysis of Ru/IrO_x with different loading amount vs IrO_x, respectively.

FIGS. 20A-20D depict CV curves of IrO_x-Ru 0.005 mg.

FIGS. 21A-21D depict CV curves of IrO_x-Ru 0.01 mg.

FIGS. 22A-22D depict CV curves of IrO_x-Ru 0.05 mg.

FIGS. 23A-23D depict CV curves of IrO_x-Ru 1 mg.

FIGS. 24A-24D depicts double-layer capacitance (Cdl) represented by curve slope.

FIGS. 25A-25H depict (FIGS. 25A-25D) LSV curves normalized by geometric area; and (FIGS. 25E-25H) LSV curves normalized by ECSA for specific activity.

FIG. 26 depicts OER performance comparison between Ru-0.05 mg/IrO_x and other reported materials at pH˜7 and the chronopotentiometric curves of Ru-0.05 mg/IrO_x at a current density of 500 and 200 mA cm⁻².

FIGS. 27A-27D depicts: (FIG. 27A) Different Ru_x/IrO₂ loading systems at various sites. Arrows are used to indicate four specific sites for Ir doping (FIG. 27B) Calculated formation energy of an oxygen vacancy for different Ru_x/IrO₂ loading systems at various sites. A negative value indicates that vacancy formation is energetically favorable. The model systems include pure IrO₂, Ru_6%(a-c)/IrO₂, Ru_13%(a,b)/IrO₂, and Ru_19%/IrO₂ (FIG. 27C) Volcano plot of the OER overpotential as a function of adsorption free energies, ΔG_HO* andΔG_O*, for reaction intermediates. (FIG. 27D) Theoretical overpotential, η^OER, for the OER.

FIG. 28 depicts free energy diagrams of IrO₂, and Ru_13%(a)/IrO₂ depicting the adsorbate evolution mechanism (AEM) and lattice-oxygen-mediated mechanism (LOM) for the OER at U=0 V, and U=1.23 V standard potential, pH=0, and T=298 K. Key intermediate structures for both AEM and LOM pathways are shown on top. Atom colors are indicated.

FIG. 29 depicts the chronopotentiometric curve of the CO₂ reduction MEA electrolyzer using Ru-0.05 mg/IrO_x as OER anode and Ag cathode as CO₂ RR cathode. Inset shows the digital photographs of the membrane electrode assembly (MEA) device.

FIG. 30 depicts the GC result of CO product.

FIG. 31 depicts an example flowchart of a method for carbon dioxide reduction, in accordance with the present disclosure.

DETAILED DESCRIPTION

Electrocatalytic CO₂ conversion into value-added chemicals could help decarbonize global supply chains for sustainable society. Electrocatalytic CO₂ reduction reaction (CO₂RR) in KHCOs electrolyte shows promise, but is limited by poor stability of the anodic oxygen evolution reaction (OER) catalyst when operating at industrially-relevant current density (stability generally less than 100 hours, when operating >200 mA cm⁻²). As described herein, a Ru doped IrO_x catalyst (Ru/IrO_x) for OER in KHCO₃ electrolyte applicable to high current density conditions is discussed. Using density functional theory (DFT) calculations, we show that the introduction of Ru modifies the electronic structure of the catalyst surface, resulting in a lowering the associated overpotential, thereby enhancing the OER activity. Utilizing combined experimental and computational results, we demonstrate that the stability of IrO₂ is sensitive to different incorporations of Ru. The integrated system achieves a full cell voltage of 3.9 V at 200 mA cm⁻² for 480 hours in membrane electrode assembly.

FIG. 1 depicts an example electrochemical system 100 for implementing the electrochemical reduction of carbon dioxide (CO₂). The electrochemical system 100 includes a cathode 102 configured to support a reduction reaction of the CO₂ to produce water, an anode 104 configured to support an oxygen evolution reaction (OER) to produce diatomic oxygen (O₂) from the water, an electrolyte 106 fluidly connecting the cathode 102 and the anode 104, and an electricity source 108 configured to apply an electrical current across the system 100 to induce the reduction and oxygen evolution reactions at the cathode 102 and the anode 104, respectively. In some examples, the system 100 may further include a membrane 110 separate the anodic and cathodic sub-cells of the system 100, thereby forming a membrane electrode assembly (MEA).

In particular, the electrochemical system 100 may preferably an electrocatalytic system, in which the reactions at each of the electrodes (i.e., the cathode 102 and the anode 104) include electrocatalysts configured to catalyze the reduction and oxygen evolutions. Accordingly, an electrolyte 106 may be selected to stabilize the catalytic reactions at the electrodes. For example, the electrolyte may be potassium bicarbonate (KHCO₃), which may stabilize the CO₂ reduction at the cathode 102 by offering lower hydrogen generation than that in acidic conditions and higher carbon utilization efficiency than that of alkaline conditions.

However, the anodic oxygen evolution reaction involves multi-electron transfer and sluggish reaction kinetics, resulting in high overpotential—factors that are exacerbated by operating at neutral conditions. The stability of many OER catalysts in neutral KHCO3 electrolytes is low, typically under 100 hours at current densities at or above the industrial relevant threshold of ˜200 mA/cm².

Currently, iridium oxide (IrO_x) is widely used as the standard OER catalyst for CO₂ reduction reaction processes. Another commonly used catalyst for OER is ruthenium oxide (RuO₂). Both RuO₂ and IrO₂ exhibit excellent OER catalytic activity in acidic and alkaline electrolytes (RuO₂ is slightly higher than IrO₂) and are generally considered the benchmark for OER catalysts. However, the stability of these catalysts is a barrier to their practical application. It has been demonstrated that both RuO₂ and IrO₂ experience dissolution and activity decay at high anodic potentials, with RuO₂ exhibiting a faster dissolution rate than IrO₂. During the OER process, (Ru⁴⁺)O₂ transforms into RuO₂(OH)₂ and then undergoes protonation to form highly oxidized species (Ru⁸⁺)O₄. However, these highly oxidized species are unstable and easily dissolve from the anode, reducing catalytic activity and risking transport to, and poisoning of, the cathode. Similarly, IrO₂ also experiences a similar phenomenon during the OER process, forming unstable high oxidation state species (Ir⁶⁺)O₃. Several studies have confirmed that although the activity of IrO₂ may be slightly lower than that of RuO₂, IrO₂ exhibits significantly higher stability than RuO₂. Neutral conditions further exacerbate the overpotential and stability challenges associated with these leading OER catalysts.

In accordance with the present disclosure, Ru³⁺ ions were introduced as single-atoms into IrO_x-based catalysts, which demonstrated that the stability of IrO₂ can be controlled by the concentration of these Ru sites. As a result, the Ru doped IrO_x catalyst (Ru/IrO_x), exhibits excellent OER activity and long-term stability (>100 hours) in a neutral electrolyte of 0.5 M KHCOs at pH=7.3 and high current densities of 200 and 500 mA·cm⁻². Further, density functional theory (DFT) calculations were used to investigate the nature of catalytic active sites and associated reaction mechanisms. To replicate experimental samples, DFT models investigate various Ru contents between 6% and 19%, as a theoretical atomic percentage. The results show that the Ru modification alters the electronic structure of the catalyst surface, with the Ru_13%(a)/IrO₂ system exhibiting the lowest overpotential, with Ir remaining as the active site, demonstrating enhanced OER activity and stability. A stability of 480 hours under a constant current density of 200 mA·cm⁻² in a membrane electrode assembly (MEA) paired with CO₂ reduction was also demonstrated.

Referring to FIG. 2A, an example method 200 of preparing the ruthenium doped iridium oxide electrocatalyst for use as an anode in an electrocatalytic cell is depicted.

At block 205, IrO_x/Ti electrodes are formed via anodic electrodeposition.

According to an example implementation, before the electrodeposition process, a Titanium (Ti) Fiber Felt was cut into 1 cm*2 cm or 3 cm*3 cm size and then washed with acetone, dilute HCl solution, and water, respectively. The IrO_x electrodeposition solution adopted the recipe reported by Zhang et al. (“Electrodeposited nanometer-size IrO₂/Ti electrodes with 0.3 mg IrO₂ cm⁻² for sludge dewatering electrolysers”): Oxalic acid dihydrate (1 mmol) was combined with 30 ml of water containing 0.2 mmol of IrCl₃. The mixture was stirred for 10 minutes, and then K₂CO₃ (5 mmol) was added to raise the pH to 10-10.5. Next, the solution was diluted to a total volume of 50 ml, resulting in a final concentration of iridium ions of 4 mmol L⁻¹. This solution was maintained at a temperature of 40° C. for approximately 5 days or longer to achieve stabilization. Subsequently, it was stored as a stock solution at a temperature of 4° C. The electrodeposition was conducted using a three-electrode system, in which the Ti Fiber Felt served as the working electrode with a 1*1 cm² or 3*3 cm² working area. The reference electrode and counter electrodes were an Ag/AgCl electrode and a platinum wire, respectively. The electrochemical technique used for electrodeposition was the potentiometric method. The potentiometric electrodeposition was conducted in 25 mL electrodeposition solution with the Ti Fiber Felt as substrate under a potential of 0.0016 A (versus Ag/AgCl) for 15, 30, 60, and 120 min. Finally, the samples were heat-treated in a tube furnace for 1 h at 500° C. under air.

At block 210, an RuCl₃ solution is prepared using ethylene glycol (EG) as solvent. EG acts as a reducing agent in the synthesis to reduce Ru³⁺ ions. For example, 2 or 40 mL solutions having a concentration of x mg mL⁻¹ of RuCl₃ may be prepared for different quantities of x (e.g., x=0, 0.005, 0.01, 0.05, 1) according to a desired subsequent quantity of Ru atoms doped into the IrO_x.

At block 215, the Ru/IrO_x is synthesized by a galvanic replacement reaction. For example, the Ti Fiber Felt with IrO_x nanoparticles was immersed into the solution prepared at block 210 under 80° C. for 15 min. That is, the Ir atoms may be replaced by the Ru³⁺ ions at certain sites, as driven by the higher reduction potential of the ruthenium. The concentration of Ru may also be carefully controlled in the solution (i.e., by way of quantity of x as given at block 210), to promote the doping of the Ru into the IrO_x as single atoms while minimizing the formation of nanoparticles (i.e., via the formation of Ru-Ru metallic bonds). That is, if the concentration of Ru is sufficiently high, the Ru ions may be doped sufficiently close to promote such nanoparticle formation. Accordingly, the concentration of Ru may be such that the nanoparticle formation is minimized, while being sufficiently high so as to achieve the stabilization of the IrO_x anodic electrocatalyst in the OER as described herein. For example, the Ru solution may preferably have a concentration of Ru of about 0.05 mg mL⁻¹ to about 1 mg mL⁻¹, and further preferably about 1 mg mL⁻¹. The resulting Ru/IrO_x may have a concentration of Ru of less than 1% by weight.

After washing with isopropanol and water, the Ti Fiber Felt with IrO_x was dried by air flow. Ti mesh acts as a conductive electrode and does not participate in the OER reaction.

FIG. 2B depicts a schematic diagram of an example performance of the method 200 to prepare the ruthenium doped iridium oxide electrocatalyst for use as an anode in an electrocatalytic cell.

Materials Characterization of Catalysts

To unravel the mechanism of Ru/IrO_x and IrO_x in OER, a variety of material characterizations were conducted both before and after a stability test. Ru/IrO_x nanoparticles were prepared through a spontaneous galvanic replacement reaction (e.g., as described in the method 200), which was driven by the reduction potential difference between RuCl₃ and Ir.

In particular, the OER performance of Ru/IrO_x catalysts was evaluated in membrane electrode assembly (MEA) electrolyzers coupling with the CO₂ reduction reaction. A schematic diagram of the electrolyzer setup is shown in FIG. 3. A potentiostat with a current booster (10 A) was used to apply the current. To facilitate the electrochemical reactions, a commercially available CO₂ MEA electrolyzer was utilized. The MEA electrolyzer consisted of flow field plates with a serpentine-shaped flow field of 5 cm², serving as the anode and cathode, enabling a continuous supply of 0.5 M KHCO₃ anolyte and humidified CO₂ to their respective electrodes. As for the specific components, the cathode utilized a silver (Ag) gas diffusion electrode (GDE), the anode employed a Ti Fiber Felt with Ru/IrO_x catalyst as described herein, and a physically separated anion exchange membrane (Sustaninion membrane) separated the anode and cathode. To ensure even distribution of electrical current, the cathode electrodes were securely taped to the stainless-steel flow field plate using a copper frame. The electrolyzer bolts were appropriately tightened with equal compression torque. Prior to conducting the electrochemical testing, the anion exchange membrane was activated in 1 M KOH solution for more than 24 hours. Once the electrolyzer assembly was completed, the anolyte (0.5 M KHCO₃) flowed through the anode at a constant rate of 18 mL/min using a peristaltic pump with silicone tubing to circulate the anolyte, while the humidified CO₂ was supplied from the gas diffusion electrode (GDL) at a constant flow rate of 50 standard cubic centimeters per minute (sccm). The OER was then initiated by applying a constant current density (100, 200, and 500 mA cm⁻²) for long term stability tests. The corresponding cell potentials for the current densities of interest were recorded with continuous monitoring. The full-cell potentials were reported without IR correction. Gas products were analyzed using gas chromatography (GC) to determine the gas product yield. For each current density tested, the gas products were collected upon complete stabilization of the cell voltage at least three times.

The grazing incidence X-ray diffraction (GI-XRD) measurement was first carried out to verify the crystalline structures of the synthesized Ru/IrO_x and IrO_x samples on Ti fiber felt. The characteristic peaks in the XRD pattern of Ru/IrO_x, as can be seen in FIGS. 4A and 4B, can be ascribed to IrO₂ in the rutile phase. No Ru peak appeared in the GI-XRD spectrum of Ru/IrO_x, indicating no Ru crystalline phase formed in the sample. The crystal phase remains largely unchanged before versus after reaction in Ru-0.05 mg/IrO_x and IrO_x. The GI-XRD measurements were obtained using a diffractometer with Cu Kα radiation (1.54 Å) at an incidence angle of 0.3°.

The energy dispersive spectroscopy (EDS) images and spectrum of the Ru/IrO_x sample (FIG. 5) reveal a homogeneous distribution of Ru, Ir, and O atoms, demonstrating the co-existence and homogenous distribution of Ru and Ir atoms in the Ru/IrO_x. The aberration corrected scanning transmission electron microscopy high angle angular dark field (AC-STEM-HAADF) images (FIG. 6) show the Ru/IrO_x and IrO_x sample features for surface reconstruction before vs after the reaction for 12 h in 500 mA cm⁻² neutral conditions. AC-STEM-HAADF images were taken on an FEI Titan 80-300 HB TEM equipped with energy-dispersive X-ray spectroscopy (EDX) at 80 kV. The AC-STEM-HAADF images also indicated that the Ru/IrO_x electrocatalyst reconstructed during OER, from 4-fold symmetry to 6-fold symmetry—a structural change to which we attribute improved catalyst activity and stability. After the OER reaction, the surface structure of the IrO_x sample undergoes distortion, which might be a potential cause for the decrease in stability.

As seen from the scanning electron microscope (SEM) images (FIG. 7), no obvious change was observed in the morphology of Ir after its galvanic replacement reaction with RuCl₃. SEM was also used to observe the morphology of the catalysts before vs after the stability test. The Ru/IrO_x nanoparticles almost kept their initial morphology, while many nanoparticles of IrO_x peeled off from the substrate after the stability test. This may be the reason for the decreasing activity of IrO_x. It was reported that the IrO_x electrocatalyst would suffer irreversible oxidation and reconstruction during the OER process. Scanning electron microscopy (SEM) images were captured with a working accelerating voltage of 10 kV.

X-ray photoelectron spectroscopy (XPS) was measured for both samples before vs after the stability test. For unstable catalysts (IrO_x & Ru-0.005 mg/IrO_x), Ir⁰ components decreased but high valence Ir⁴⁺ and lattice oxygen greatly increased after the stability test. For stable catalysts (Ru-0.05 mg/IrO_x & Ru-1 mg/IrO_x), Ir⁰ components increased but high valence Ir⁴⁺ and lattice oxygen were almost unchanged (FIGS. 8A-8B, and 9A-9D). This agrees well with the electrochemical results since the high valence Ir and lattice O are responsible for the high activity of IrO_x. The lattice oxygen increase leads to the instability in IrO_x. In contrast, there is a small decrease of the high valence Ir⁴⁺ from 29% to 24% in the Ru/IrO_x catalyst after the stability test, while the lattice oxygen was almost unchanged. Usually, the abundance of lattice oxygen tends to increase during the stability test and cause activity degradation, but this was not observed after the stability test for the Ru/IrO_x sample. X-ray photoelectron spectroscopy (XPS) measurements were carried out on a microprobe with a monochromatic Al Kα X-ray source (1486.6 eV). The obtained spectra were calibrated using the C 1s line. The catalyst was sonicated from the carbon paper in IPA solution. Then the solvent was drop casted onto ultrathin lacey carbon TEM grids for imaging. Some residue carbon paper is hard to separate from the samples.

The oxygen K-edge X-ray absorption (XAS) spectra were measured using synchrotron scanning transmission X-ray microscopy (STXM) (FIG. 10). It is worth noting that XPS reveals the presence of Ir⁴⁺, whereas STXM XAS does not show it, suggesting that Ir⁴⁺ may exist on the catalyst surface or the Ir valence edges (M and N edges) have low sensitivity for STXM transmission measurement. Oxygen K-edge XAS spectra measure the transition from oxygen 1s orbitals to unoccupied orbitals formed by the interactions between oxygen 2p and metal d orbitals. Thus, these oxygen K-edge XAS spectra reflect the covalency in the metal-oxygen bond. The change of 1,2,3 peaks intensity may reflect a higher Ir-O covalency in Ru/IrO_x after the reaction relative to that in Ru/IrO_x before the reaction. The higher Ir-O covalency in Ru/IrO_x after the reaction suggests a lower amount of unoccupied 3d orbitals in Ir, indicating the valence state of Ir decreased in Ru-0.05 mg/IrO_x after the reaction. By contrast, a lower Ir-O covalency was observed in IrO_x, suggesting a higher amount of unoccupied 3d orbitals in Ir and thus a higher valence state of Ir in IrO_x after the sublayer reaction. This contrast in Ir-O covalency may be caused by a Ru induced mechanism during OER that does not occur for bare IrO_x. This interpretation is consistent with the XPS analysis. Both Ru/IrO_x and IrO_x show different oxygen K-edge peaks than commercial IrO₂. This indicates oxygen vacancies in both Ru/IrO_x and IrO_x samples. Moreover, STXM XAS does not indicate the presence of RuO₂, whereas extended X-ray absorption fine structure (EXAFS) exhibits Ru-O bonding.

This discrepancy could be attributed to sensitivity issues at the O K-edge for low-concentration Ru doping. An extremely low Ru content of 0.1 wt % was determined by inductively coupled plasma mass spectrometry (ICP-MS) analysis. Inductively coupled plasma mass spectrometry (ICP-MS) analyses of Ru and Ir abundances were carried out using He as a collision cell gas, and In and Bi as internal standards to correct for instrument drift. Primary ICP calibration standards were from Aristar VWR Chemicals BDH. Analysis of secondary standards confirmed instrument accuracy to within 3%. Detection limits and background equivalent concentrations were <3 ppt (parts-per-trillion) for both Ru and Ir.

To further reveal the coordination structure, oxidation state, bond configurations, and changes in bulk structure and chemical environment of Ir and Ru in IrO_x and Ru-0.05 mg/IrO_x, X-ray absorption spectroscopy (XAS) was carried out before and after the OER experiment. The X-ray absorption near edge structure (XANES) of the Ir L3-edge and Ru K-edge region and the extended X-ray absorption fine structure (EXAFS) measurements of the Ir L₃-edge and were conducted on Ir and Ru foil (FIGS. 11A-11D and 12A-12H). The k-range (0-12) used in the Fourier Transformation (FT) was chosen to exclude the regions of noise at high-k. The corresponding FT R-space profiles (without phase correction) was also analyzed. XANES spectra at the Ir L₃-edge of IrO_x and Ru-0.05 mg/IrO_x catalysts before vs after the reaction are compared with the standard Ir foil (FIG. 11A). The energy of the Ir white-line peak is correlated with the symmetry of the bulk structure's central atom (O coordination); higher symmetry leads to an energy right shift, while lower symmetry causes a shift to the left. Combined with AC-STEM-HAADF images, the Ru/IrO_x electrocatalyst undergoes reconstruction during the OER process. Increased symmetry (FIG. 6) results in a stabilized bulk structure and increased active sites, aligning with the observed rightward shift in the white-line peak for Ru/IrO_x after the reaction.

On the other hand, the post-reaction IrO_x structure experiences distortion and decreased symmetry, exhibiting an unstable structure (FIG. 6), consistent with the leftward shift in the white-line peak for IrO_x after the reaction. In addition, high energy represents high oxidation state (FIG. 11A). Ir was reduced in IrO_x after the reaction while Ir was oxidized in Ru/IrO_x after the reaction (FIG. 11A). The FT-EXAFS fitting results of the Ir L₃-edge show that the Ir-O bond in IrO_x after reaction exhibits a slight “elongation” (˜2.1 Å) compared with the bond length of Ir-O bond before reaction (˜1.9 Å), meanwhile Ir-Ir peak appeared in IrO_x samples after reaction (FIG. 11B). Table 1 further illustrates this:

In comparison with IrO_x samples, the Ru-0.05 mg/IrO_x samples before vs after the reaction reveal a single dominant peak Ir-O, the Ir-O bond remains nearly unchanged at ˜2.0 Å and coordination number ˜ 5.9 (FIG. 11B and Table 1), which shows excellent structural stability. These findings align with the XPS analysis, indicate that the increase in lattice oxygen (leading to an abundance of oxygen vacancies) exacerbates instability. Conversely, for the Ru/IrO_x sample, the measured lattice oxygen remains largely unaffected, affirming the stability of the Ru/IrO_x samples during the OER. The Ru K-edge XANES spectra of Ru-0.05 mg/IrO_x catalysts before vs after the reaction as well as the Ru foil for reference (FIG. 11C). The white line of Ru/IrO_x shows an apparent shift to higher energy for the adsorption edge (E₀) compared to that of Ru foil, indicating that Ru in Ru/IrO_x has a charge transfer with IrO_x support and carries a positive charge. The Ru was oxidized in Ru/IrO_x after the reaction (FIG. 11C). Moreover, the Fourier transform (FT) of the k³-weighted EXAFS curve for the Ru K-edge of the Ru/IrO_x catalyst exhibited distinctive characteristics (FIG. 11D). Prior to the reaction, the FT-EXAFS curve showed Ru-O at 2.15 Å coordination, after the 12-hour 500 mA cm⁻² reaction (FIG. 11D and Table 2).

The Ru/IrO_x sample displayed a single dominant Ru-O peak at 1.87 Å after the reaction (FIG. 11D and Table 2). This could suggest the electrocatalyst undergoes irreversible oxidation and reconstruction during the OER process. This observation suggests that Ru atoms have only one coordination as Ru-O, which contrasts with the Ru-Ru coordination in Ru foil at 2.7 A (FIG. 11D and Table 2). Combined with the EXAFS analysis and AC-HAADF-STEM images, the single-atom dispersion of Ru atoms with localized coordination in the Ru/IrO_x could be synergistically confirmed. Through the characterizations, it indicates the Ru single atoms doping can prevent the instability originated from the lattice oxygen activation in IrO_x.

XAS measurements were carried out at the 20-BM and 20-ID-C beamline. The measurements at the Ir L-edge and Ru K-edge were performed in fluorescence mode using a Lytle detector. The Ir L-edge and Ru K-edge XANES and EXAFS data were analyzed and treated using the software package Athena. The EXAFS data was fitted using the software package Artemis. Ir and Ru foil was applied for reference and calibration samples. In this fitting data, CN represents the coordination numbers of identical atoms; R is assigned as the interatomic distance; δ² denotes the Debye-Waller factors; and R factor is the goodness of fit. The fitting parameters strictly comply with all experimental requirements.

Electrocatalytic OER Performance

OER Performance of Different Loads and Ethylene Glycol (EG) treatment of IrO_x

To determine the catalytic performance of different loads of electrodeposited IrO_x, the OER performance of the IrO_x (900 s, 1800 s, 3600 s, 7200 s) catalyst was recorded in an Ar saturated 0.5 M KHCOs electrolyte. As shown by the linear sweep voltammetry (LSV) curves, the catalysts subjected to IrO_x deposition for 3600 s and 7200 s exhibit a smaller potential value (2.1V) at 100 mA cm⁻² than that of the catalysts with deposition times of 900 s and 1800 s (FIGS. 13A-13B), which indicates better OER activity. The chronoamperometry test under the corresponding potentials of 100 mA cm⁻² results of three catalysts (1800 s, 3600 s, and 7200 s) show favorable stability and activity throughout the 30-min test (FIGS. 13A-13B). Therefore, both the catalysts deposited for 2 hours and 1 hour demonstrate satisfactory performance. However, given that Ir is an expensive noble metal and the deposition amount is relatively high for two hours, the 1-hour-deposited IrO_x catalyst was selected for subsequent experiments.

To evaluate the effect of EG on the samples, we immersed the IrO_x in EG. We used 1 h IrO_x immersed in x mg/ml EG (x=0, 0.005, 0.01, 0.05, 1) for 15 min to dope the

Ru atoms. The LSV curves (FIGS. 14A-14B) show that the IrO_x catalyst exhibits a reduced potential at 200 mA cm⁻² from 3.6V to 3.2V following EG treatment. The chronopotentiometry (FIGS. 14C-14D) and electrochemically active surface area (ECSA) (FIGS. 15A-15H) results show that both catalysts performance demonstrate relatively good stability for 12 h runs at 10 mA cm⁻² but deteriorates rapidly for 12 h runs at high current densities of 200 mA cm⁻² and 500 mA cm⁻², while the IrO_x catalyst treated with EG exhibits a significantly slower decay rate compared to the catalyst without EG treatment. These results suggest an enhancement in catalyst performance through EG soaking. OER performance of different loading of Ru in Ru/IrO_x

To improve IrO_x stability under a large current density, some Ru atoms are doped into the IrO_x. Increased ruthenium (Ru) doping does not significantly enhance the initial OER performance of IrO_x regarding the overpotential. The overpotential (n) at 10 mA cm⁻² of IrO_x and Ru/IrO_x all around 600 mV (ERHE˜1.83V), at 200 mA cm⁻² all around 1.97V (ERHE˜3.2V) (FIGS. 14A-14D and 16A-16D). However, Ru doping does improve the stability of IrO_x. The results of chronopotentiometry for 12 h runs using different loading amounts of Ru in Ru/IrO_x and different current densities (10, 200 and 500 mA cm⁻²) (FIGS. 16A-16D, 17A-17B, 18), shows that all samples are stable under 10 mA cm⁻². However, at higher current densities, only Ru-0.05 mg/IrO_x and Ru-1 mg/IrO_x exhibit no apparent potential increase during the 12 h stability test under all current densities. Hence, Ru-0.05 mg/IrO_x and Ru-1 mg/IrO_x demonstrate the greatest stability (FIGS. 16A-16D, 17A-17B).

The Ru doping not only improves the stability, but also the activity for different loadings of Ru/IrO_x and IrO_x (FIG. 19A). The analysis of activity for various Ru/IrO_x shows better initial activity than that of IrO_x (FIG. 19A). In particular, the OER activity of Ru-0.05 mg/IrO_x slightly increases after the 12 h stability test under 500 mA cm⁻², while the activity of Ru-1 mg/IrO_x slightly decreases, and other samples degrade severely after the stability test (FIG. 19A).

The ECSA (FIG. 19B and FIGS. 20A-20D, 21A-21D, 22A-22D, 23A-23D, 24A-24D) of IrO_x decreased after the stability test, whereas the ECSA of Ru0.005 mg/IrO_x and Ru0.01 mg/IrO_x are nearly unchanged and Ru-0.05 mg/IrO_x and Ru-1 mg/IrO_x increased after the stability test. Activity analysis (FIGS. 25A-25H) lends support to this observation. The Ru-1 mg/IrO_x sample showed the highest specific activity, whereas the IrO_x activity decreased.

Although it seems reasonable to attribute the stability change to the ECSA, the reason for the difference in ECSA change between these two sets of samples (IrO_x VS Ru/IrO_x) is still unknown. There are two main OER mechanisms, i.e., the lattice oxygen mechanism (LOM) and adsorbate evolution mechanism (AEM). Usually, the catalysts working under LOM shows better activity but poorer stability than the catalysts working under AEM. For the IrO_x-based catalyst, LOM may be the main reason for its poor long-term stability, specifically, the lattice oxygen of IrO_x participates in the OER reaction and exchanges with the oxygen from water. It is reasonable to attribute the decrease in ECSA of IrO_x to the LOM, but the reason for the increase of ECSA of Ru/IrO_x before versus after the reaction is still unclear. Another question is whether ECSA is the only reason for the activity change and difference in stability. To answer these questions, the geometrical activity was normalized to ECSA. The specific activity of Ru/IrO_x-based on ECSA is much higher than that of IrO_x after the stability test (FIGS. 25A-25H). This observation means that the difference in ECSA is not the only reason for the ongoing difference activity between IrO_x and Ru/IrO_x. We conclude that the structure of the intrinsic active sites may differ in IrO_x vs. Ru/IrO_x.

Electrochemical tests were measured on Metrohm Autolab PGSTAT 204 electrochemical workstations at room temperature with IR corrected. The OER stability and activity performance of the IrO_x and Ru/IrO_x catalyst with different Ru loading amounts in neutral conditions (0.5M KHCO₃ electrolytes) were systematically studied in a single compartment electrolytic cell. A three-electrode system was fabricated with the prepared IrO_x-based materials, platinum wire, and Ag/AgCl (in saturated KCl) electrode serving as the working electrode, the counter electrode, and the reference electrode, respectively. The synthesized Ti Fiber Felt with Ru/IrO_x catalyst directly served as the working electrode with an electrode holder. The surface area of the working electrode was controlled at 1 cm² in 25ml 0.5M KHCO₃ solution under an Ar gas environment. Linear sweep voltammetry (LSV) under a select potential range with 0.4V iR compensation was used for overpotential and activity measurement of the Ru/IrO_x catalytic OER reaction. After that, the chronoamperometry (I-T) test was used for stability measurements under the over potential at 10 mA and 100 mA for 30 min or 60 min. The Ru/IrO_x catalytic OER performance was evaluated using chronopotentiometry (P-T) tests under a selective potential for 12, 100, and 200 hours. Before the OER test, LSV curves were performed until the polarization curves achieved steady state. The electrochemically active surface area (ECSA) measurements were conducted using cyclic voltammetry (CV) scans at different rates from 10 to 50 mV s⁻¹ with a 10 mV s⁻¹ increment speed from 0.5 V_RHE to 0.7 V_RHE; within this potential range, the Faradic process is excluded. All the potential values are presented in RHE unless otherwise stated using the following equation (1):

E RHE = E Ag / AgCl + 0 . 1 ⁢ 97 ⁢ V + 0.059 × pH ( 1 )

The ESCA is calculated by the following equation (2):

A ECSA = Specific ⁢ capacitance 40 ⁢ µFcm - 2 ⁢ cm ECSA - 2 ( 2 )

The overpotential (η) is calculated from equation (3):

η ⁡ ( V ) = - 1.23 ⁢ V ( 3 )

The specific capacitance of the sample was obtained by CV. It was carried out at different scan rates in the range of 0.5 V_RHE to 0.7 V_RHE. 40 mF cm⁻² is the specific capacitance of a flat surface for metallic and semiconducting materials with 1 cm² of the real surface area in the aqueous electrolyte.

OER Performance of Ru/IrO_x with Respect to Long Term Stability

It is evident that both Ru-0.05 mg-IrO_x and Ru-1 mg/IrO_x exhibit excellent stability and activity, with a minor discrepancy at 200 Ma cm⁻² and 500 mA cm⁻². Given the precious metal nature of Ru, we opted for the lower Ru loading mount, Ru-0.05 mg/IrO_x, for conducting the long-term stability experiments. The electrochemical stability for neutral OER was evaluated (FIG. 26). The galvanostatic curves of Ru-0.05 mg/IrO_x were obtained at a current density of 200 and 500 mA cm⁻². As shown, The Ru-0.05 mg/IrO_x performance is relatively steady for 200 h at a constant current density of 200 mA cm⁻², and for about 100 h at a constant current density of 500 mA cm⁻². Furthermore, this OER catalyst developed in this study exhibits work potentials of only 2.5V and 2.8V for current densities of 200 mAcm⁻² and 500 mAcm⁻², respectively. This OER stability outperforms the previous OER performance in a neutral electrolyte under a large current density in the literature (FIG. 26 and Table 3).

Computational Results and Discussion

To better understand the reasons behind high activity and stability of Ru_x/IrO₂ samples in our experimental section, we utilized DFT calculations. Our integrated analysis of EXAFS=and AC-HAADF-STEM images confirms the single-atom dispersion of Ru atoms with localized coordination within the Ru/IrO_2-x structure. Therefore, to capture the influence of coordination environment, we constructed various model structures by incorporating different numbers of Ru single atoms doped into the IrO_2-x (110) rutile crystal structure (FIG. 27A).

The studied model structures include various content of atomic percentages of Ru, including 6%, 13% and 19%. Of note, the percentages in our computational models may not directly mirror the experimental compositions. However, constructing a variety of models for atomic configurations aids in elucidating the potential coordination chemistry that impacts the stability and activity of Rux-IrO_2-x. Ru_6%-IrO₂ includes one Ru single atom in the IrO₂ slab, while Ru_13%-IrO₂ and Ru_19%-IrO₂ include two and three Ru single atoms in the IrO₂ slab, respectively. For Ru_6%-IrO₂, we investigated three probable sites for Ru atoms, indicated as Ru_6%(a)-IrO₂ and Ru_6%(b)-IrO₂ and Ru_6%(c)-IrO₂ in FIG. 27A, each of which corresponds to a different location for the doped Ru atom, either on the surface or subsurface of the IrO₂ slab. Specifically, in Ru_6%(a)-IrO₂, the Ru atom is a sublayer atom located below the Ir atom active site. In Ru_6%(b)-IrO₂, the Ru atom is a surface atom located next to the Ir atom active site. In Ru_6%(c)-IrO₂, the Ru atom is a surface atom, and it serves as the active site.

For Ru_13%-IrO₂, we examined two different model structures where (Ru_13%(a)-IrO₂ and Ru_13%(b)-IrO₂) correspond to different arrangements of Ru atoms within the IrO₂ slab. In Ru_13%(a)-IrO₂, one Ru atom is a surface atom located next to the Ir atom active site, while another Ru atom is a sublayer atom positioned below the Ir atom active site. In Ru_13%(b)-IrO₂, one Ru atom is a sublayer atom below the Ir atom active site, and another Ru atom is a sublayer atom next to this Ru sublayer atom. For Ru_19%-IrO₂, we examined only one possibility due to the configuration it presents. In this structure, the two Ru atoms on the surface and the Ru atom in the sublayer are the nearest neighbor atoms to Ir at the surface for this amount of Ru.

According to our XPS findings, Ru_x/IrO_x samples maintaining their lattice oxygen content after OER (e.g. Ru-0.05 mg/IrO_x and Ru-1 mg/IrO_x) exhibit greater stability. The rise in lattice oxygen content correlates directly with an increase in oxygen vacancies. Consequently, it is apparent that samples like IrO₂, which display a substantial increase in lattice oxygen content following OER testing, have undergone the formation of numerous oxygen vacancies, resulting in low OER stability. To investigate the stability in more detail, we calculated the formation energy of an oxygen vacancy (ΔE_f) in various model structures constructed for Rux-IrO_2-x in FIG. 27A using DFT calculations (FIG. 27B). We then compared these with the vacancy formation energy of pristine IrO₂.

The computational results presented in this study were performed using the Vienna Ab initio Simulation Package (VASP, version 5.4.4) in conjunction with the Atomic Simulation Environment (ASE) Interface. All spin-polarized density functional theory (DFT) calculations utilized the PBE exchange-correlation functional. The electron-ion interactions were described using projector-augmented wave (PAW) potentials.

The (110) surface of IrO₂ was employed for surface calculations. The tetragonal P4₂/mnm space group bulk structure of IrO₂, was utilized from Materials Project identifier mp-2723. to construct the IrO₂ (110) surface. First, the lattice parameters of the bulk structure were optimized with a plane-wave cutoff of 520 eV and a Monkhorst-Pack k-point mesh of (5×5×7). Subsequently, a (1×2×4) IrO₂ (110) surface was constructed based on the optimized bulk structure, with a 20Å vacuum spacing between repeating structures in the z-direction. To simulate the Ru-doped model structures, Ir atoms from the surface and subsurface layer(s) were substituted with Ru, with a total atomic percentage ranging from 6 to 19%. The surface geometry was optimized using a plane-wave cutoff of 520 eV and a Monkhorst-Pack k-point mesh of (4×4×1). The convergence criteria for electronic and ionic iterations in geometry optimizations were set to 10⁻⁶ eV and 0.03 eV/A, respectively.

The vacancy formation energy is evaluated using eq. (4):

ΔE f = E slab + vacancy - ( E slab - 1 / 2 ⁢ µO 2 ) ( 4 )

• • where, ΔE_f represents the formation energy of the oxygen vacancy, E_slab+vacancy is the total energy of the slab model with an oxygen vacancy, E_slab is the total energy of the perfect slab without a vacancy, and 1/2μO₂ is the chemical potential of oxygen expressed as 1/2E(O₂)=E_H2O−E_H2. The equation essentially quantifies the energy required to form an oxygen vacancy. A lower formation energy indicates a more stable oxygen vacancy. If ΔE_f is negative, it implies that the formation of the oxygen vacancy is energetically favorable.

In evaluating the free energy for both gas-phase molecules and adsorbed species in eachstep of the OER, we considered zero-point energy (ZPE) and entropic(S) corrections obtained from a vibrational analysis performed at 298.15 K. Gas-phase molecules and adsorbed species were analyzed under the Ideal-gas oscillator limit and Harmonic limit, respectively. Following this analysis, the free energy change is expressed as:

Δ ⁢ G i = Δ ⁢ E 1 + Δ ⁢ ZPE - T ⁢ Δ ⁢ S ( 5 )

• • where ΔE_i denotes the energy difference between the reactant and product.

The theoretical overpotential, denoted as η^OER, is defined as,

η OER = max [ Δ ⁢ G 1 , Δ ⁢ G 2 , Δ ⁢ G 3 , Δ ⁢ G 4 ] / e - 1.23 V ( 6 )

By considering the difference between ΔG_O* and ΔG_HO* as a distinctive descriptor for OER activity, the theoretical overpotential (under standard conditions: T=298.15 K, P=1 bar, pH=0) can be expressed as:

η OER = { max [ ( Δ ⁢ G O * - Δ ⁢ G HO * ) , 3.2 eV - ( Δ ⁢ G O * - Δ ⁢ G HO * ) ] / e } - 1.23 V ( 7 )

Despite being a thermodynamic quantity, this equation reliably correlates with experimentally determined overpotentials.

The tendency of the catalyst to form an oxygen vacancy (less formation energy) is an indication of a less stable structure. The stability of IrO₂ is sensitive to different incorporations of Ru (FIG. 27B). IrO₂ has a negative oxygen vacancy formation energy of ΔE_f=−0.27 eV, indicating a facile loss of oxygen from site 1 (FIG. 27B). However, any amount of Ru incorporated into IrO₂ shifts the oxygen vacancy formation energies toward more positive values ranging from 0.1 to 0.39 eV when a vacancy is formed in site 1 (FIG. 27B). When subsurface oxygen vacancies are considered in Ru_x/IrO₂ structures, the oxygen vacancy formation energy is decreased indicating more tendency to form subsurface oxygen vacancies. The fact that incorporating any amount of Ru (Ru_x/IrO₂) shifts the oxygen vacancy formation energy towards more positive values indicates higher stability for Ru_x/IrO₂ structures, leading to less leaching of the Ir⁺⁴ into the solution during the OER in comparison with pure IrO₂. This analysis offers valuable insights into understanding the interactions between Ru dopants and IrO₂ at different crystallographic sites, leading to tailoring the IrO₂ catalytic properties toward better OER performance.

As stated earlier, it is established that catalysts operating under LOM exhibit superior activity, but inferior stability compared to those operating under AEM. Previous experimental and computational studies on IrO₂ and RuO₂ catalysts have demonstrated the LOM as a likely pathway for OER. To understand how the OER activity changes across different Ru_x/IrO_2-x model structures, we investigated both AEM and LOM mechanisms. We first did the analysis for the AEM mechanism and calculated the adsorption free energies of OER intermediates, ΔG_HO*, ΔG_O*, and ΔG_HOO*. Computational hydrogen electrode (CHE) in conjunction with descriptor-based analysis was used to identify trends in OER activity across different model structures in FIG. 27A. The free energy difference of the key intermediates ΔG_HO* and ΔG_O* is used as the OER activity descriptor. Theoretical overpotential was calculated for all different model structures which allowed us to establish the OER activity volcano plot (FIG. 27C).

The color bar in FIG. 27C shows the theoretical overpotential with the most active catalyst located in the red region where overpotential is minimized. Different examined structures were studied to draw the catalytic activity trends (FIG. 27A). Our findings reveal that Ir remains the most active site in all Ru_x/IrO_2-x structures. When a surface Ru atom is considered as the active OER catalytic site (Ru_6%(c)/IrO₂) in any of the structures studied, the calculated overpotential is exceptionally high, reaching 0.71 V (FIG. 27D). The results in FIGS. 27C-27D indicate that even minimal percentage of Ru atoms (e.g. Ru_6%(b)/IrO₂) modifies the electronic structure of the adjacent Ir sites, lowering the OER overpotential and thereby increasing activity. Among the studied systems, Ru_13%(a)/IrO₂ structure displays the least overpotential (0.38 V, FIGS. 27C-27D). In this configuration, Ru single atoms are dispersed across both the surface and the sublayer (as depicted in FIG. 27A), thereby altering the electronic structure of the adjacent Ir atom, which acts as an active site.

We further investigated the LOM mechanism on our most active catalyst surface, i.e., Ru_13%(a)/IrO₂ (FIG. 28). Our results show that the reaction mechanism may involve a combination of both LOM and AEM, with LOM potentially playing a role in initiating the reaction by facilitating the transfer of lattice oxygen atoms, which then proceed to participate in adsorbate evolution on the catalyst surface. FIG. 28 displays the OER free energy diagrams on IrO₂ and Ru_13%(a)/IrO₂ at U=0.0 V and U=1.23 V, considering both AEM and LOM mechanisms. LOM relies on the participation of lattice oxygen atoms, whereas AEM involves the adsorption and evolution of reactive OER intermediates on the catalyst surface. Thus, we further calculated the adsorption energy of the key intermediate steps for LOM, ΔG_HO*, ΔG_2HO*, and ΔG_2O*.

Taking the AEM pathway on IrO₂ at U =0.0 V, we find that the initial step involves the exothermic adsorption of HO* on the IrO₂ surface (FIG. 28). The subsequent oxidation of HO* to O* is an uphill process with a thermodynamic barrier of 1.63 eV, followed by an oxidation of additional H₂O molecule and O* to form HOO* with a barrier of 2.93 eV. The free energy diagram at U=1.23 V, shows that the most uphill step (also known as thermodynamic barrier) for the OER following the AEM is associated with the oxidation of O* to HOO*. The free energy diagram for the LOM pathway starts with the adsorption of H₂O(I) and a subsequent oxidation leading to the formation of HO*. The next step in LOM pathway diverges from AEM as another H2O molecule needs to be oxidized and form the second HO* in a second active site, followed by an oxidation of both HO's to 2O*, resulting in the formation of O₂(g). The thermodynamic barrier in LOM is energetically competitive with its counterpart in the AEM pathway. However, the occurrence of two exothermic steps at the onset implies that LOM could potentially offer a more advantageous route in IrO₂ when contrasted with AEM. This observation aligns with recent experimental findings reported for IrO₂ using in-situ operando experiments and explains why IrO₂ has a significant increase in lattice oxygen content after OER (XPS analysis).

For the Ru_13%(a)/IrO₂ system (FIG. 28), the LOM and AEM mechanisms closely mirror that of IrO₂, sharing the same key intermediates. The most apparent difference in the LOM pathway is that the introduction of a Ru single atom weakens the adsorption free energy of the second HO* on Ir sites, which in turn results in a lower overpotential. In the AEM, the highest thermodynamic barrier still remains to be the oxidation of O* to HOO*. In the LOM pathway, there is a slightly exothermic adsorption of HO* with an energy gain of 0.076 eV. However, subsequent steps in the LOM pathway are uphill, resembling the energy profile observed in the AEM pathway. Notably, the thermodynamic barrier at U=1.23 V for oxidation of 2HO* to 2O* is slightly higher than the corresponding barrier in the IrO₂ system.

At U=1.23 V for the AEM, going from O++H2O to HOO* is accompanied by 1.70 eV and 2.02 eV on IrO₂ and Ru_13%(a)/IrO₂, respectively. Conversely, the LOM mechanism has thermodynamic barriers of 1.78 eV and 1.9 eV on IrO₂ and Ru_13%(a)/IrO₂, for oxidation of 2HO* to 2O*. These results suggest that while there is a competition between AEM and LOM in both IrO₂ and Ru_13%(a)/IrO₂, LOM is potentially less favorable when Ru is doped into the IrO₂. This entails the higher stability for Ru_x/IrO_2-x catalysts which is consistent with experimental observations from XPS results where the lattice oxygen content (an indicator of abundant oxygen vacancies) is less impacted or remains unchanged for after OER test.

OER performance of Ru/IrO_x electrocatalyst for CO₂ reduction in MEA systems

We found Ru-0.05 mg/IrO_x to offer better long-term stability at a current density of 200 mA cm⁻² instead of 500 mA cm⁻² and employed 200 mA cm⁻² for long-duration MEA testing. The OER performance of the Ru-0.05 mg/IrO_x catalyst for CO₂ reduction was validated in a MEA electrolyzer, as described with respect to FIG. 3. An anion exchange membrane was employed. The MEA was set with an anodic 3 cm² Ru-0.05 mg/IrO_x catalyst and 2.5 cm² Ag GDE as the cathode. The MEA exhibited a cell voltage of 3.9 V at 200 mA cm⁻². Ru-0.05 mg/IrO_x exhibited excellent stability during CO₂ reduction under large current density in MEA, with no apparent increase in cell voltage over 500 hours (FIG. 29) of CO production [at standard temperature and pressure (STP)]. That is, FIG. 29 demonstrates the electrocatalytic stability in MEA with CO₂ RR paired with OER. CO is the exclusive product of silver-catalyzed CO₂ reduction at these conditions, and the generation of CO was monitored using online GC analysis (FIG. 30).

The product of the silver-catalyzed reduction of CO₂ is CO, so a standard curve for CO gas is first established using a standard gas. In detail, different concentrations were prepared by diluting the CO with CO₂ using mass flow controllers. During the electrocatalytic CO₂ reduction, the gas products flow into the GC with the input CO₂ gas in the online tube. In experiments, CO was detected by the thermal conductivity detector (TCD).

This MEA stability of OER coupling with CO₂ reduction in a neutral electrolyte outperforms the previous reported stability in a neutral electrolyte under a large current density (Table S4). These findings combined confirm excellent OER activity and stability using the Ru-0.05 mg/IrO_x catalyst under neutral conditions and at industrially-relevant current densities.

FIG. 31 depicts a flowchart of an example method 3100 for carbon dioxide reduction, in accordance with the present disclosure.

At block 3105, an electrochemical system is provided. For example, the system 100 may be provided as the electrochemical system. In particular, the electrochemical system includes a cathode including a first electrocatalyst, such as an Ag gas diffusion electrode, and an anode, including a second electrocatalyst, in particular a ruthenium doped iridium oxide. The electrochemical system further includes an electrolyte connecting the cathode and the anode, and an electricity source configured to apply an electrical current across the cathode and the anode to drive the electrochemical reaction. The electrolyte may be a neutral electrolyte, such as KCOs to stabilize the cathodic first electrocatalyst. In some examples, the electrochemical system may include a membrane separating the cathodic and anodic sub-cells to form an MEA.

At block 3110, carbon dioxide is provided or introduced to the cathode. For example, gaseous CO₂ may be flowed into the cathode.

At block 3115, the electricity source is activated to apply an electrical current to the electrochemical system. Preferably, the electrical current may be applied at a current density of about 200 mA cm⁻² to about 500 mA cm⁻². Further, the electricity source may be configured to apply the electrical current for at least 100 hours to obtain an industrially significant application of the CO₂RR. In particular, for the current density of 500 mA cm⁻², the electrical current may be applied for at least 100 hours, while for the current density of 200 mA cm⁻², the electrical current may be applied for at least 200 hours, and in some examples, for at least 400 hours.

At block 3120, as a result of the application of the electrical current at block 3115, a reduction reaction of the carbon dioxide is catalyzed at the cathode. Carbon products such as carbon monoxide, methane, or ethylene may be produced in addition to water.

At block 3125, as a result of the application of the electrical current at block 3115, an oxygen evolution reaction of the water is catalyzed at the anode. In particular, the oxygen evolution results in the production of diatomic oxygen (O₂).

Thus, overall, the method 3100 promotes the conversion of CO₂ to O₂.

As described herein, an approach to design oxygen evolution reaction catalysts for application in neutral CO₂ RR systems is provided—by incorporating Ru single atoms in IrO_x—that exhibit high activity and stability. The resulting Ru_x/IrO_x catalyst achieves low work potentials of 2.5V and 2.8V for current densities of 200 mAcm⁻² and 500 mA cm⁻², and stability of 200 h and 100 h respectively. DFT calculations reveal that Ru enhances the electronic structure of the catalyst surface in two keyways: firstly, by reducing the likelihood of oxygen vacancy formation, thus increasing the stability of IrO₂; and secondly, by boosting activity, resulting in the lowest observed overpotential in the Ru_13%(a)/IrO₂ system. We demonstrated the stability of the Ru-0.05 mg/IrO_x catalyst for 480 hours in a AEM base MEA system at 200 mA cm⁻², attesting to the exceptional OER activity and stability of the Ru/IrO_x catalyst under both neutral electrolyte conditions and at industrially-relevant current densities. These findings advance efforts to develop commercial application of OER catalysts in a KHCO₃ electrolyte, and remove a barrier to the adoption of CO₂ electrocatalytic production of chemical feedstocks and fuels.

The scope of the claims should not be limited by the embodiments set forth in the above examples but should be given the broadest interpretation consistent with the description as a whole.