Surface Hydrides on Fe2P Electrocatalyst Reduce CO2 at Low Overpotential: Steering Selectivity to Ethylene Glycol.

Development of efficient electrocatalysts for the CO2 reduction reaction (CO2RR) to multicarbon products has been constrained by high overpotentials and poor selectivity. Here, we introduce iron phosphide (Fe2P) as an earth-abundant catalyst for the CO2RR to mainly C2-C4 products with a total CO2RR Faradaic efficiency of 53% at 0 V vs RHE. Carbon product selectivity is tuned in favor of ethylene glycol formation with increasing negative bias at the expense of C3-C4 products. Both Grand Canonical-DFT (GC-DFT) calculations and experiments reveal that *formate, not *CO, is the initial intermediate formed from surface phosphino-hydrides and that the latter form ionic hydrides at both surface phosphorus atoms (H@Ps) and P-reconstructed Fe3 hollow sites (H@P*). Binding of these surface hydrides weakens with negative bias (reactivity increases), which accounts for both the shift to C2 products over higher C-C coupling products and the increase in the H2 evolution reaction (HER) rate. GC-DFT predicts that phosphino-hydrides convert *formate to *formaldehyde, the key intermediate for C-C coupling, whereas hydrogen atoms on Fe generate tightly bound *CO via sequential PCET reactions to H2O. GC-DFT predicts the peak in CO2RR current density near -0.1 V is due to a local maximum in the binding affinity of *formate and *formaldehyde at this bias, which together with the more labile C2 product affinity, accounts for the shift to ethylene glycol and away from C3-C4 products. Consistent with these predictions, addition of exogenous CO is shown to block all carbon product formation and lower the HER rate. These results demonstrate that the formation of ionic hydrides and their binding affinity, as modulated by the applied potential, controls the carbon product distribution. This knowledge provides new insight into the influence of hydride speciation and applied bias on the chemical reaction mechanism of CO2RR that is relevant to all transition metal phosphides.

Dataset of theoretical multinary perovskite oxides.

Perovskite oxides (ternary chemical formula ABO3) are a diverse class of materials with applications including heterogeneous catalysis, solid-oxide fuel cells, thermochemical conversion, and oxygen transport membranes. However, their multicomponent (chemical formula [Formula: see text]) chemical space is underexplored due to the immense number of possible compositions. To expand the number of computed [Formula: see text] compounds we report a dataset of 66,516 theoretical multinary oxides, 59,708 of which are perovskites. First, 69,407 [Formula: see text] compositions were generated in the a-b+a- Glazer tilting mode using the computationally-inexpensive Structure Prediction and Diagnostic Software (SPuDS) program. Next, we optimized these structures with density functional theory (DFT) using parameters compatible with the Materials Project (MP) database. Our dataset contains these optimized structures and their formation (ΔHf) and decomposition enthalpies (ΔHd) computed relative to MP tabulated elemental references and competing phases, respectively. This dataset can be mined, used to train machine learning models, and rapidly and systematically expanded by optimizing more SPuDS-generated [Formula: see text] perovskite structures using MP-compatible DFT calculations.

Bond-Valence Parameterization for the Accurate Description of DFT Energetics.

We report a bond-valence method (BVM) parameterization framework that captures density functional theory (DFT)-computed relative stabilities using the BVM global instability index (GII). We benchmarked our framework against a dataset of 188 experimentally observed ABO3 perovskite oxides, each of which was generated in 11 unique Glazer octahedral tilt systems and optimized using DFT. Our constrained minimization procedure minimizes the GIIs of the 188 perovskite ground state structures predicted by DFT while enforcing a linear correlation between the GIIs and DFT energies of all 2068 competing structures. GIIs based on BVM parameters determined using our framework correctly identified the DFT ground state perovskite structure in 135 of 188 compositions or one of the two lowest energy structures in 152 of 188 compositions. Using the most common approach to parameterize BVM, which minimizes the root-mean-square deviation of the BVM site discrepancy factors, GIIs correctly identified the DFT ground state perovskite structure in only 41 of 188 compositions. Our new parameterization framework is therefore a marked improvement over the existing procedure and an important first step toward BVM-based structure generation protocols that reproduce DFT.

Predicting Spinel Disorder and Its Effect on Oxygen Transport Kinetics in Hercynite.

The iron aluminate spinel hercynite (FeAl2O4) is a promising redox material for solar thermochemical hydrogen production (STCH). Although it has a high H2 production capacity, the kinetics of its oxidation and reduction may be too slow to be practical for STCH. However, our results suggest that Fe-rich hercynite may have substantially faster redox kinetics, which could make hercynite competitive with other materials for STCH. We used density functional theory to investigate the origin of hercynite's slow kinetic behavior and show that it arises from the high activation barrier of 2.46 eV for oxygen vacancy (VO) diffusion in normal hercynite. To model the effect of disorder caused by spinel inversion, we examined 11 of the most common cation arrangements and found a near 1:1 correlation between the diffusion barrier and VO formation energy, both of which decrease by 0.6 eV for each additional nearest-neighbor Fe atom. To examine this trend, we used integrated crystal orbital Hamilton population (ICOHP) analysis to estimate the difference in the metal-oxygen bond strengths of cations neighboring VO and the diffusion transition state. The ICOHP predicted bond strengths correlate to both the diffusion barrier and VO formation energy. We also computed the effect of the charge state of the oxygen vacancy and found that positively charged vacancies are stable at low Fermi energies and have a diffusion barrier of only 0.79 eV, 1.67 eV lower than that of the neutral vacancy, demonstrating that stabilizing these charged vacancies may enable faster oxidation and reduction kinetics in hercynite. We show that uncompensated Fe antisite defects, which are present in Fe-rich hercynite, provide redox flexibility that stabilizes the charged VO and thereby increase the rate of VO diffusion. Finally, we predict that at higher VO concentrations the diffusion barrier depends on the relative positions of the vacancies and decreases when they are next-nearest neighbors.