Molecular conductance calculations on single-molecule junctions using projection-based density functional embedding

Single-Molecule Junctions (SMJs) are key platforms for the exploration of electron transport at the molecular scale. In this work, we present a method that employs different exchange-correlation density functionals for the molecule and the electrode domains in an SMJ, allowing us to choose the optimal one for each part. This is accomplished using a formally exact projection-based DFT-in-DFT embedding technique combined with the non-equilibrium Green's function (NEGF) method to predict the zero-bias conductance. The effectiveness of the approach is illustrated through transport calculations on SMJs with benzene-1,4-diamine (BDA) and its tetramethylated and tetrafluorinated variants using the CAM-B3LYP range-separated hybrid functional for the embedded molecule and the PBE functional for the electrodes. The findings indicate a substantial improvement in the accuracy of the predicted zero-bias conductance compared to traditional modeling using the PBE functional across the entire system. The causes for the noted improvement are demonstrated through the examination of alterations in the energy levels of the embedded molecule, along with variations in the electrode-molecule interactions.