Descriptor-driven DFT study of pentlandite electrode materials for alkaline oxygen evolution reaction

The growing imperative to transition toward sustainable energy has driven intense interest in earth-abundant electrocatalysts for the oxygen evolution reaction (OER, 2H₂O → O₂ + 4H⁺ + 4e⁻ ), the primary bottleneck in water-splitting electrolyzers used for green hydrogen production. In alkaline electrolyzers, pentlandite ( Fe₄.₅Ni₄.₅S₈ )—an inexpensive, highly conductive chalcogenide—has demonstrated promising efficiency and stability experimentally, but the fundamental origins of these qualities remain incompletely understood. In this thesis, I develop and apply a comprehensive, descriptor-driven density functional theory (DFT) protocol within the computational hydrogen electrode (CHE) framework to unravel OER mechanisms on pentlandite surfaces. I consider four (111) surface models and quantify activity using the potential-dependent descriptor Gmax(U) . Modeling the mononuclear mechanism on the pristine (model 1) surface yields no catalytic activity, in contrast to experiment. Introducing pre-adsorbed intermediates via a surface Pourbaix diagram (SPD) (model 2) identifies Fe as the active site. A reduced cell (half-surface model) (model 3) captures the same qualitative trends at lower computational cost. Finally, introducing the experimentally observed oxidation into the model by substituting the top two sulfur layers with oxygen and applying the SPD (model 4) reveals that both Fe and Ni are active. Moreover, this extension demonstrates a higher mechanistic diversity of the OER, and Fe and Ni exhibit different rate-limiting steps depending on applied bias, determined via degree-of-span control analysis. Additionally, part of the research is dedicated to probing the M-OH vs. M-O scaling relation, i.e. a linear function that relates the adsorption energies of the two intermediates. Pentlandite data yield anomalous slopes, motivating a parametric study of the intercept ξsr . Using an in-house technique of generalized volcano plots I show that variations in ξsr substantially shift the activity peak and alter mechanistic diversity, whereas changes in the slope γsr merely modify the volcano's legs. Applying this approach to experimental adsorption energies for IrO₂ and RuO₂ successfully reproduces the superior OER activity of RuO₂ , a trend unattainable by conventional volcano analyses. Overall, this work delivers a reproducible, computationally efficient workflow that integrates surface-phase equilibria, advanced thermodynamic descriptors, and scaling-relation parametrization to guide the rational design of pentlandite and related earth-abundant OER electrocatalysts. Future extensions include expanding datasets for intercept fitting, performing molecular dynamics, grand canonical, or hybrid functional calculations, and exploring doped or mixed-metal pentlandites to further optimize activity and stability.