Multiscale Materials Modeling Group
Universidad de Valladolid
The development of nanotechnology has revealed the need for modeling of materials properties, devices and processes at atomic scale. Ideally one would want to solve the Schrödinger equation for the complete system under study and at the same time reproduce times and temperatures of actual processes. However, even structures in the nanometer scale involve thousands to millions of atoms, and times of at least milliseconds that are impossible to handle at a fundamental level with the present computational capabilities.
To connect experimental processes and observations with fundamental properties we resort to a multiscale simulation scheme that involve the combination of several modeling techniques, each one of them with different level of detail in the atomic description and appropriate for different time and space scales. We use in house developed codes (Kinetic Monte Carlo: DADOS), open access codes adapted for the problem under study (Classical molecular dynamics: LAMMPS) and licensed codes (ab initio: VASP).
As an example of this multiscale scheme, using classical molecular dynamics simulations we have studied the fundamental properties of the point defect known as the bond defect or IV pair and we demonstrated that accumulation of bond defects leads to the progressive amorphization of the semiconductor lattice via the introduction of five- and seven-membered rings which structurally define the amorphous matrix. The characterization of the bond defect allowed us to develop an atomistic model for semiconductor amorphization suitable to Kinetic Monte Carlo simulators. On the basis of this model, we have been able to predict the effect of implant and annealing parameters on the onset of amorphization and regrowth, and the influence of residual damage on dopant diffusion.
