Harnessing the power of computational science to accurately predict material properties and rapidly design and deploy new materials is one of the urgent technological challenges of the 21st century. The first step in any computational materials design effort is the generation of realistic structural models, a task that is rendered much more difficult in the case of complex (i.e., molecular, polymeric, multicomponent, nanostructured, etc.) disordered solids. The investigators expect to substantially advance the state of the art, and surmount traditional challenges associated with identifying non-global potential energy minima for materials produced under non-thermodynamic conditions and aligning simulation and growth process timescales, by developing novel algorithms for linking growth conditions and characterization information to atomic model simulation, and for mapping fabrication conditions to desired properties. The success of this goal hinges upon the novel approach to combining state of the art computational techniques (AIMD, HRMC) with modern optimization algorithms (PSO, ANN), in conjunction with specialized experimental characterization techniques, implemented by a uniquely qualified team of experts in materials growth, characterization, and simulation.
