Accurate calculation of electron paramagnetic resonance parameters for molybdenum compounds

Paramagnetic molybdenum compounds are of great interest in 4d-metal inorganic chemistry and metalloenzyme catalysis. Electron paramagnetic resonance (EPR) spectroscopies that determine hyperfine coupling parameters and g values are essential tools for investigating the local and global electronic structure of these compounds. Such studies require support from accurate quantum chemical approaches to establish reliable structure–spectroscopy correlations. Here we present a curated database of 22 molecular Mo complexes with well-defined structures and EPR parameters and investigate quantum chemical approaches to determine optimal protocols for computing 95Mo hyperfine coupling constants (HFCs) and g values of Mo(V) compounds. It is shown that the SARC all-electron basis sets developed for the Douglas–Kroll–Hess (DKH) Hamiltonian can be used without any adaptation also for the exact-2-component (X2C) Hamiltonian and require no modifications to produce excellent and converged results for both HFCs and g values. The dependence of EPR parameters on the functional is studied in detail. Double-hybrid functionals and global hybrids with high percentage of exact exchange are top performers for 95Mo HFCs, with PBE0-DH achieving the best agreement with experiment. The DFT results on HFCs are compared with values obtained by coupled cluster theory with the domain-based local pair natural orbital approach (DLPNO-CCSD) and we show that the latter falls short in terms of accuracy and consistency compared to the best performing functionals for the preset set of compounds. Smaller differentiation among functionals is observed for the calculation of g tensors, with double hybrids being surpassed by several global and range-separated hybrid functionals, although PBE0-DH is still a top performer and can thus be recommended as the most reliable DFT approach overall for both valence and core properties of Mo compounds.