Title:	Non-adiabatic molecular dynamics simulations of molecular motors
DNr:	SNIC 2016/1-483
Project Type:	SNIC Medium Compute
Principal Investigator:	Bo Durbeej <bo.durbeej@liu.se>
Affiliation:	Linköpings universitet
Duration:	2016-12-01 – 2017-12-01
Classification:	10407 21001 10405
Homepage:	http://www.liu.se/forskning/foass/bo-durbeej
Keywords:

Abstract

Molecular motors are molecules that can perform work by absorbing energy and converting the energy into directed mechanical motion like rotation around a chemical bond. Because of their ability to execute a number of useful functions, such as rotating objects thousand-fold heavier than themselves and acting as wheels in perfectly manoeuvrable nanocars, it has long been recognized that molecular motors have enormous potential for a wide variety of applications in nanotechnology and medicine. Fittingly, the 2016 Nobel Prize in Chemistry was deservedly awarded to three scientists who have made outstanding contributions to the design and synthesis of three different types of molecular motors. In the last few years, we have initiated a line of research unique to Swedish academia in which computational techniques in theoretical chemistry are used to design more powerful molecular motors. Specifically, we have focused on finding ways to improve the performance of one particular class of synthetic light-driven molecular motors that, under ambient conditions, are able to achieve rotary motion in the MHz regime. In this endeavor, we have used quantum chemical methods to identify strategies for accelerating the rate-limiting thermal steps that together with photochemical steps make up the rotary cycles of these motors. Indeed, through our work, we have discovered a number of steric, electronic and conformational approaches to bring the rotational frequencies of the motors into completely new territory, beyond the MHz regime (see, for example, RSC Adv. 2014, 4, 10240; Phys. Chem. Chem. Phys. 2015, 17, 21740; ChemPhysChem 2016, 17, 3399). Such rate acceleration is a prerequisite for many of the applications currently envisioned for molecular motors. However, in order to improve the rotational frequencies of the motors even further, it is important to additionally gain an understanding of how the photochemical steps of the rotary cycles can be made faster. Furthermore, it is also of interest to design entirely new motors that produce rotary motion in a purely photochemical fashion, without the intermediacy of slower thermal steps. The present project aims to make progress toward both these goals. From the viewpoint of computational resources, this is more demanding than our previous work, because the excited-state potential energy surfaces that underlie the photochemical steps cannot be mapped through static quantum chemical calculations alone. Rather, since the excited-state surfaces are much flatter than the ground-state surfaces governing the thermal steps, meaningful and reliable modeling of the photochemical steps requires that first-principles non-adiabatic molecular dynamics (NAMD) simulations are performed. Through such simulations, it is possible to both predict how long the photochemical steps take and estimate the corresponding quantum yields, which are key variables in assessing the motor performance. By carrying out NAMD simulations for a variety of different motors, this project will provide very valuable guidelines for the design of new molecular motors better capable of executing useful tasks in nanotechnology and medicine than existing ones.