In plain words: Can one control the motion of metal atoms "surfing" on a metallic surface and then anchor them at desired locations? The answer, perhaps surprisingly, is YES! Using a technique called STM (Scanning Tunneling Microscopy) one can literally move atoms one at a time. But if you wish to move many atoms, all with atomic precision, to create a nanodevice, such a technique might be painfully slow. Instead, we are pursuing an approach to modify surface properties from a distance and thereby encourage the diffusion of the absorbed molecules into desired patterns.

Controlling Diffusion by Tickling a Surface

Diffusion on metal surfaces is one of the most ubiqutious processes in nature with broad-reaching implications for a variety of chemical processes such as catalysis, self assembly and formation of nano-structures. The diffusive motion of atoms or molecules on metal surfaces is a result of numerous dynamical processes each occuring on different timescales. This fact poses a significant challenge for current computational efforts. A general methodology consisting of various techniques is under development which will be capable of spanning the disparate length and timescales inherent to these systems. A detailed atomistic description as is illustrated below is employed to accurately model the the small scale events and serves as the basis for a coarse-grained method.

Figure 1:   Side view of the copper (110) surface

Figure 2:   Top down view of the copper (110) surface

The latter consists of Langevin model driven by a stochastic potential of mean force fluctuating on a slower time scale than the thermal noise. A snapshot of the reduced-dimensional surface potential with a trajectory that traverses multiple adsorption sites can be seen below. Its parameters are determined from the equilibrium fluctuations observed in molecular dynamics simulations of the atomistic surface. Once determined, however, the coarse-grained model can be used for rapid determination of various properties ---such as the diffusion rate--- of an adatom on the surface. In this manner a heirarchy of interconnected tools have been developed which allow for both the illucidation and verification of new transport phenomena.

Figure 3:   Phenomenological time-dependent potential of mean force with a typical diffusive trajectory (red line) traversing multiple adsorption sites

Relevant references are:

• T. D. Shepherd and R. Hernandez; "An optimized mean first passage time approach to obtaining rates in activated processes," J. Chem. Phys. 117, 9227-9233 (2002). (doi:10.1063/1.1516590)
• J. M. Moix, T. D. Shepherd, and R. Hernandez; "A phenomenological model for surface diffusion: diffusive dynamics across incoherent stochastic aperiodic potentials," J. Phys. Chem. B 108, 19476-19482 (2004). (doi:10.1021/jp046629w)