Atomistic Modeling with Supercomputers and Nanoscale Materials

In a world where technology has evolved so rapidly a great focus has been placed on nanoscale materials. These materials possess greater properties than their bulk material. Properties such as superior strength, resistance to cracking, high temperature applications and almost everything you can think of. The problem is often that these materials are too small to study in conventional ways such as a scanning electron microscope (SEM) or a transmission electron microscope (TEM).  The desire to create these superior materials has led to the development of atomistic modeling programs for supercomputers.

One such notable program is called GULP (General Lattice Utility Program) published by Julian Gale at the Nanochemstry Research Institute, Curtin University, Western Australia. These computer programs account for many millions of interactions occurring on the atomic scale. The programs generally allow for a verity of input parameters that give rise to specific situations materials might encounter. Programs like GULP work by calculating Hessian Cycle and using maximization and minimization functions to achieve what is called a relaxed structure. The relaxed structure is a very accurate estimation of what the material might look like if it was subjected to certain conditions.  

Several important values arise from these calculations such as the energy change from the original structure to the relaxed structure. These types of values can be used to calculate binding energies, lattice expansion and many other properties. Nanoscale modeling programs allow the user to input defects including Frenkel and Schottky defects. These defects are considered point defects and generally replace (substitution) a specific cation in the crystal lattice. Creating defects is very useful because it allows the user to study the behavior of a material if the product is contaminated with a defect during formation of the product.

Due to the fact that these materials are of the nanoscale (10-9 m) impurities are a common problem.  Nanoscale programs are generally run on a Linux based system and considerable user skills are required to operate the system.  Due to many possible millions of reactions taking place within the model powerful computers called supercomputers are need to run the calculations. Still even with supercomputers it can take weeks to run a single calculation with a structure composed of more than 500 atoms. The term “supercomputer” refers to a system which possesses at least 2 quad core cpu’s and 20 gigabytes of ram. Supercomputers are used for calculation-intensive tasks, generally quantum physical parameters and biological applications and aerodynamics. Many input parameters contain millions of variables or more and extensive studying is required to understand how do design input files for these programs.