  Nano-scale Simulations Help Build Materials of the Future
How fast and how small will the next generation of electronic components and computers be? Will we be able to manipulate biological cells with nano- and micro-machines? The answers to both these questions depend on our ability to build materials — and predict their failure points — at smaller and smaller scales. Faculty in the Department of Mechanical Engineering are now developing atomic level simulations to model the deformation and failure of materials at scales approaching 109 meters, the nano-scale. Researchers have had to create new simulation methods to address this critical area of nano- and micro-scale design, as traditional continuum-based modeling tools become obsolete at such small scales. Some electronic components, or Micro-Electro-Mechanical Systems (MEMS), for example, rely upon microscopic layers of aluminum deposited on silicon for the creation of conductive paths. The question of how small we can make these components while keeping them structurally stable against temperature changes or forces can only be addressed at the atomic level.
Assistant Professor Ken Gall and Associate Professor Martin L. Dunn recently received funding from the National Science Foundation and Sandia National Laboratories to examine the atomic level deformation mechanics of aluminum-silicon and gold-silicon material systems for multi-layer MEMS. The technological challenge is understanding and quantifying deformation mechanics in joined thin films at length scales approaching several microns (106 meters). Based on Professor Gall's previous work using the Embedded Atom Method, atomic level simulations will be constructed for sections of these MEMS. The simulations consider the atoms as point entities that may interact through adjacent repulsive or attractive forces, and an embedding energy that models multiple atom interactions. The original structures are built by specifying the location of all atoms within the solid. The contrived atomic MEMS structures can then be subjected to a wide variety of external factors such as temperature changes or applied velocities, simulating the loading conditions that actual MEMS will experience. The atomic simulations also permit alteration of the local structure of the materials, allowing the structure to be tailored until an optimal material is obtained. Although atom-based simulations have always been a curiosity of materials researchers, they have not been feasible due to a lack of computing speed. To model the deformation of a 0.1-micron (105 meter) cube, for example, the attractive-repulsive force and energy interactions must be calculated between nearly 15 million atoms. Although the fastest computers today can handle atomic level calculations for one million atom simulations, computing resources must be further improved to conduct simulations at significantly larger sizes. Ironically, atomic level simulations of structural nano-circuits may improve the very computer chips that make atomic level simulations possible in the first place.  Engineering Home |
