Spin Transitions in Single Atom Catalyst (Fe-N-C) by Quantum Monte Carlo Method

The performance of Fe-N-C based single-atom catalysts (SACs) with isolated Fe atoms dispersed on nitrogen-doped carbon matrix in the oxygen reduction and evolution reactions is closely tied to transitions between high-spin (HS) and intermediate-spin (IS) state. Accurate prediction of adsorption energies in these systems requires precise quantum mechanical methods beyond GGA-density functional theory (DFT). We combine DFT with diffusion Monte Carlo (DMC) calculations to determine the most stable spin states of the above reactions. While both methods predict consistent spin-state preferences from IS to HS, there are significant discrepancies in the absolute energy values, leading to large differences in the adsorption energies of intermediates such as OH* on the iron center.



Our findings show the sensitivity of adsorption energy calculations to the different computational approaches, where DMC provides more reliable results than standard DFT functionals. To check the accuracy of DMC within the single-determinant framework, we will also report tests of other ab-initio methods such as selected configuration interaction, coupled cluster, or multi-determinant DMC to capture the complex electronic structure of the system. This study exposes the limitations of conventional methods in strongly correlated electronic systems with spin transitions and recommends the most suitable DFT functionals, emphasizing the need for advanced quantum techniques for accurate modeling at exascale.