No abstract

This study explores the feasibility of using the molecular dynamics computational technique to predict the thermal conductivity of solid thin films in the direction perpendicular to the film plane. The results show that thermal conductivity, as expected from thin-film experimental data and theoretical predictions, decreases as film thickness is reduced. In the large-size limit, this method yields thermal conductivities which asymptote to a value comparable to experimental data. The calculations modestly overpredict thermal conductivity, probably due to the use of a too-steep intermolecular potential. Most interestingly, an unusual wave effect is revealed for thin film thermal conductivity. This effect may be a manifestation of phonon wave interference analogous to the interference of light which determines the radiative properties of thin films.It is also found that there are some temperature and computational domain size limitations on the applicability of molecular dynamics to the study of solid systems. A regime map is developed which delineates the conditions necessary for molecular dynamics to produce physically meaningful results. This work shows that molecular dynamics, applied under the correct conditions, is a viable tool for calculating the thermal conductivity of solid thin films. More generally, this work demonstrates the potential of molecular dynamics for ascertaining microscale thermophysical properties in more complex structures. INTRODUCTIONMolecular dynamics (MD) is a computational method which simulates the real behavior of materials and calculates physical properties of these materials by simultaneously solving the equations of motion for a system of atoms interacting with a given potential. The computational work on anharmonic one-dimensional chains of atoms by Fermi, Pasta, and Ulam (Fermi et al., 1965) in the 1950s was the earliest contribution to the field of MD. This pioneering research was followed by other critical MD studies, including those of Alder and Wainwright (1960), Gibson et al. (1960), andRahman (1964). A lack of sufficient computational power limited these and other early simulations to systems with a very small number of atoms.In the past two decades, however, the number of MD studies has skyrocketed due to rapid developments in computer speed and memory. It is now possible, using parallel computation, to model systems on the order of a million atoms (Hoover et al., 1990). Still, the spatial domains treated by these "large" simulations remain very small. Even when periodic boundary conditions are used, constraints on size can complicate the calculation of bulk properties. The limitations of MD in simulating bulk materials can be turned to advantage for novel nanometer-scale materials such as buckyballs and buckytubes, highly nanoporous and ultrathin films, and quantum wires and dots. In particular, solid thin films are key components in integrated-circuit transistors and quantum-well lasers, and porous thin films of materials with favorable optical properties may p...

No abstract

No abstract

No abstract

No abstract

No abstract

No abstract