Quickstart guide to model structures and interactions of artificial molecular muscles with efficient computational methods

Abstract

Artificial molecular muscles (AMMs) represent an important group of molecular machines. Their theoretical treatment is challenging due to size, element composition, and complex interaction motifs. Moreover, experimentally determined structures often only yield insights into the covalent connectivity of atoms rather than their 3D structure. Accordingly, a reproducible computational modeling of such structures is complicated. In this work we present a standardized, mostly quantum chemical protocol on how to obtain reliable structures from scratch and to compute contraction free energies ΔG_c for daisy-chain rotaxane AMMs efficiently. In this protocol, the recently developed force-field (GFN-FF) and extended tight-binding methods (GFNn-xTB) are employed. For comparison, dispersion-corrected density functional theory (DFT-D) based reference ΔG_c were computed. In one case for which data are available, excellent agreement between theoretical and experimental ΔG_c values within 1–2 kcal mol⁻¹ is obtained.