Johannes Hachmann, Department of Chemistry and Chemical Biology, Harvard University
I will introduce the Harvard Clean Energy Project (CEP) which is concerned with the in situ screening and design of carbon-based photovoltaics. Plastic solar cells are one of the promising approaches to ubiquitously establishing a source of renewable energy and providing basic electricity to rural areas in undeveloped countries. The necessary 10-15% energy conversion efficiency and lifetime to make this technology commercially viable has, however, not been achieved yet. CEP has established an automated, high-throughput framework to study millions of potential candidate structures. Our current work is focused on the relevant molecular properties of these candidates, in their own right or as polymer building blocks. We use first principles density functional theory and ideas from cheminformatics to characterize molecular motifs and assess their quality with respect to applications as electronic materials. In addition to finding specific structures with certain properties, it is the goal of CEP to illuminate and understand the structure-property relations in the domain of organic electronics. Such insights can open the door to a rational, systematic, and accelerated development of future high-performance materials. CEP is a large-scale investigation which utilizes the massive computational resource of IBM’s World Community Grid (www.worldcommunitygrid.org). In collaboration with IBM we have ported the Q-Chem program package to the BOINC environment for distributed, grid-based use on volunteer hosts. In this context, it is deployed as a screensaver application harvesting idle computing time on donor machines, similar to the seti@home, qmc@home, or folding@home projects. This cyberinfrastructure paradigm has already allowed us to characterize 5.5 million molecules of interest in about 65 million DFT calculations. The results are compiled and analyzed in an extensive reference database and will be made available for public use in 2012. Our presentation addresses the project organization and workflow, data mining, analysis, and scoring, as well as promising molecular motifs which have emerged from CEP so far.
Séminaires du RQMP – Versant Nord