XTBDFT: Automated workflow for conformer searching of minima and transition states powered by extended tight binding and density functional theory
Sibo Lin, Mohamed Elanany, Motaz Khawaji
Abstract
Recent developments have greatly increased accuracy, applicability, and accessibility of extended tight binding theory (XTB), allowing for rapid semi-empirical quantum mechanical calculations of molecular potential energy surfaces. For finer thermochemical accuracy, density functional theory (DFT) calculations are still required on stationary points. Such multi-level simulations present a steep learning curve, and trained researchers must spend significant time monitoring calculations and managing the workflow. Herein, we present XTBDFT, the first automated workflow between all freely licensed engines for XTB, DFT, and helper modules. In this version, the specific engines are GFN-XTB/CREST, NWChem, and GoodVibes, respectively. Example applications of this workflow is presented, finding the lowest energy conformer of a metal complex with multiple hindered degrees of internal rotation.