Computer Simulations for Improved Atomic-Level Control and Understanding of Molecular Electronic Devices

Abstract

Molecular electronics (ME) is an emerging field with the potential to miniaturize electronic devices beyond what is possible with current top-down approaches. In ME, molecular sized building blocks self-assemble together from the bottom-up to form active electronic circuit elements. This dissertation presents results from atomic-level modeling and simulations of mechanically controllable break-junction experiments in which proof-of-concept ME devices consisting of a single molecule bridged across fractured Au nanowire (AuNW) tips are formed. Several studies are performed of molecular- and atomic-scale junctions under realistic conditions, to elucidate how environmental factors impact their behavior.

First, the ductile-to-brittle transition is probed in AuNWs as a function of aspect ratio and temperature. Porting simulation code to graphics processing units enables the analysis of a large number of independent trajectories. These simulations reveal that stochastic motion is prominent enough to occasionally cause failure behavior that is uncharacteristic of the AuNW size. Additional simulations of AuNWs in solvent demonstrate that strong Au-solvent interactions can extend the lifetime and mechanical stability of important structural motifs within elongating AuNWs, while weak interactions can adversely affect the stability of certain structures.

Next, for molecular junctions containing benzene-1,4-dithiolate (BDT) bridging two Au nanotips, the surface density of a BDT self-assembled monolayer is used to tune the formation of single- and multi-molecule junctions. Other factors such as electrode geometry and temperature also influence the number and structure of bridged molecules. Finally, through a combination of hybrid molecular dynamics-Monte Carlo simulations and quantum mechanical calculations, the conductance evolution of elongating Au-BDT-Au junctions is computed for direct comparison to experiment. The simulations closely mimic experimental protocols for forming junctions, and thus produce configurations that are more realistic than prior theoretical work. This enables new structure-conductance relationships to be defined that (i) provide a structural basis for the design of mechanically responsive devices and (ii) aid in the development of new strategies for improved control over conductance fluctuations in ME devices.