Designer materials and devices from computationally driven self-assembly

Dr. Eric Jankowski

University of Colorado, Boulder

Graphics processing units (GPUs) are inexpensive consumer electronics that allow massive computational power to be applied to problems of self-assembly.  Using GPUs with algorithms informed by statistical mechanics, we investigate how molecular shapes and how molecular interactions can be chosen so that materials with customized properties can be synthesized.  At the fundamental algorithm level, we show how implementing Monte Carlo simulations on a GPU gives 27x better performance per dollar and uses 13x less energy for systems of hard disks.  At the micro-device level, we show how information can be stored in reconfigurable colloidal clusters and demonstrate their fabrication.  At the material level, we detail how GPU-accelerated simulations are informing our fundamental understanding of organic solar cells, particularly the phase behavior of conjugated thiophenes mixed with fullerene derivatives.

Biography:

Dr. Jankowski is currently a Director's Fellow at the National Renewable Energy Laboratory, where he is developing modeling techniques for predicting the morphologies of organic solar cell active layers.  He earned a Ph.D. in chemical engineering from the University of Michigan in 2012 where he created computational tools for studying the self-assembly of "patchy" particles.  Dr. Jankowski's research interests include problems where computer simulations combined with statistical mechanics can be used to inform the synthesis of new materials.