Influence of DFT Functionals and Solvation Models on the Prediction of Far-Infrared Spectra of Pt-Based Anticancer Drugs: Why Do Different Complexes Require Different Levels of Theory?

Computational modeling was applied
to far-infrared (FIR) spectra
of Pt-based anticancer drugs to study the hydrolysis of these important
molecules. Here, we present a study that investigates the influence
of different factorsbasis sets on non-Pt atoms, relativistic
effective core potentials (RECPs) on the Pt atom, density functional
theory (DFT) functionals, and solvation modelson the prediction
of FIR spectra of two Pt-based anticancer drugs, cisplatin and carboplatin.
Geometry optimizations and frequency calculations were performed with
a range of functionals (PBE, PBE0, M06-L, and M06-2X), Dunning’s
correlation-consisted basis sets (VDZ, VTZ, aVDZ, and aVTZ), RECPs
(VDZ-pp, VTZ-pp, aVDZ-pp, and aVTZ-pp), and solvation models (IEFPCM,
CPCM, and SMD). The best combination of the basis set/DFT functional/solvation
model was identified for each anticancer drug by comparing with experimentally
available FIR spectra. Different combinations were established for
cisplatin and carboplatin, which was rationalized by means of the
partial atomic charge scheme, ChelpG, that was utilized to study the
charge transfer between the Pt ion and ligands in both cisplatin and
carboplatin.