Predicting standard reduction potential for anticancer Au(III)-complexes: A DFT study

Abstract

The Au(III) complexes are considered promising for cancer treatment given their structural and electronic similarities to Pt(II) complexes. Different from Pt(II) complexes, the Au(III) compounds are much less stable under physiological conditions, due to their high Au3+/Au+ reduction potential. Therefore, great effort has been done to develop more stable Au(III) complexes in order to explore and improve their biological activity. Density functional theory (DFT) methods were herein applied to calculate the standard reduction potential of nine Au(III) organometallic complexes of the type [Au3+(R-C^N^C)L]n, using available experimental data as benchmark. Overall, the DFT methods lead to satisfactory results, with absolute error lower than 170mV. The CAM-B3LYP functional in combination with the SMD solvation model was superior to wB97xD and B3LYP results, with absolute error of 82mV relative to ferrocene/ferrocenium (Fc+/Fc) redox couple. However, in spite of the larger error found for B3LYP results, the qualitative trend was closer to the observed one, which allowed the proposal of a scaling model using the experimental reduction potential as dependent variable and the calculated reaction Gibbs free energy as independent variable. The linear regression was statistically acceptable (R2=0.8) at B3LYP level and lead to an average error of only 35mV. Besides, the variable-temperature H-atom addition/abstraction (VT-HAA) approach was applied for the redox couples with non-equivalent charges (complexes containing chloride as ligand), leading to a significant improvement in the reduction potential prediction. The mean absolute error was only 87mV without any scaling procedure, which is much lower than that found for standard approach, 144mV.