Modeling diffusion of nanocars on a Cu (110) surface

Abstract

Nanoscale devices and machines that can be externally controlled or programmed promise revolutionary technological improvements. One example of such machines are nanocars, which are organic supramolecular structures (typically between 200–2000 Da) designed to achieve controlled molecular motion on atomically smooth surfaces. Spurred by a recent global competition where such nanocars had to race each other, interest in this nascent area has recently increased. However, the design space of nanocars is large, and a thorough understanding of how their structure affects their motion on surfaces is lacking. In this work, we investigated the diffusion of nine large organic molecules on a Cu (110) surface using classical simulation methods and transition state theory (TST). We find that, as expected, these molecules tended to diffuse more slowly as their molecular weight and attraction to the surface increases. However, these two parameters do not give a complete picture of surface diffusion. Thus we defined a structural parameter, elevation weighted density, based on the geometry of the molecule that interacts with the surface. We show that this parameter is a good predictor of surface diffusion, as demonstrated by its high rank correlation with TST free energy barriers and with diffusion coefficients calculated using classical simulations. We further discuss design strategies to tune the diffusion performance of nanocars.