PITA Fiscal Year 2009 Projects - Product and Process Design and Optimization

Multi-Scale Modeling of Phase Change Materials

Principal Investigators: Alan McGaughey, Jim Bain

Phase change materials (PCMs), which exhibit strikingly different electrical and optical properties in their crystalline and amorphous phases, have attracted significant interest as a central component in next-generation data storage devices. Progress towards the use of PCMs has been slow due to the need to consider both electrical and thermal transport phenomena in understanding their complex behavior. To date, significant effort has only focused on the electrical transport.

In the proposed work, we will build a computational framework for modeling and understanding thermal transport in the phase change materials GeSb and GeSbTe. Using input from electronic structure calculations and by validating against experimental data, we will develop an interatomic potential for Ge-Sb-Te systems, which can be used to predict the energies and forces between sets of atoms. The interatomic potential will be implemented into a molecular dynamics computer code capable of modeling systems with millions of atoms. The thermal conductivities of amorphous and crystalline GeSb and GeSbTe, needed in device-level thermal management models, will be predicted. We will then correlate the atomic structures of the two phases to the onset of phase change. The associated activation energies, critical inputs to the nucleation models currently used to predict device performance, will be predicted.

The computational work will be performed in collaboration with Intel Research Pittsburgh. They will provide the necessary computational resources and benefit by studying how we take advantage of these resources. They will work directly with us to improve how we (i) develop our computer programs and (ii) manage the large volumes of generated data.