Calculation of electric field gradients in Cd(II) model complexes of the CueR protein metal site

With this work we first test various DFT functionals against CCSD(T) for calculation of EFGs at the position of Cd(II) in a very small model system, Cd(SCH₃)₂. Moreover, the available basis sets in ADF are tested in terms of basis set convergence, and the effect of including relativistic effects using the scalar relativistic and spin orbit ZORA Hamiltonians is explored. The results indicate that an error of up to around 10% on the calculated EFG may be expected using spin-orbit ZORA and the BHandHLYP functional with a locally dense basis set. Next, this method was applied to model systems of the CueR protein, aiming to interpret ¹¹¹Ag-PAC spectroscopic data. Note that ¹¹¹Ag decays to ¹¹¹Cd on which the PAC data are recorded. Surprisingly, model systems truncated - as is often done - at the first C-C bond from the central Cd(II) are inadequate in size, and larger model systems must be employed to achieve reliable EFG calculations. The calculated EFGs agree well with experimental PAC data, and indicate that shortly after the nuclear decay the structure relaxes from linear two-coordinate AgS₂ in the native protein, to a structure (or structures) where Cd(II) recruits additional ligands such as backbone carbonyl oxygens to achieve higher coordination number(s).