The Clean Energy Project

The Harvard Clean Energy Project (CEP) was a high-throughput in-silico hunt for the next generation of organic photovoltaic materials — molecules that could be printed into cheap, flexible solar cells. Conceived in 2008 by Alán Aspuru-Guzik's group at Harvard, CEP became one of the largest computational chemistry efforts ever undertaken: 150 million density-functional-theory calculations on 2.3 million candidate molecules, all crunched by World Community Grid volunteers over six years.

The challenge was scale. Each candidate molecule needed quantum-mechanical evaluation of its frontier orbital energies (HOMO/LUMO levels) to estimate its solar-cell performance. A single such calculation takes hours on a CPU core; 150 million of them would have required centuries of dedicated cluster time. Volunteer computing turned that into a six-year campaign distributed across hundreds of thousands of personal computers.

The output is the Harvard Clean Energy Project Database (CEPDB), released to the public in June 2013. The database is still actively used by computational chemists worldwide as a benchmark and starting point for new materials searches; it inspired the broader field of high-throughput computational materials discovery and helped seed today's machine-learning-driven materials informatics.

The project's main results were published in 2011 in The Journal of Physical Chemistry Letters. The Clean Energy Project completed its main work in 2014; today it stands as a proof that volunteer computing could match — and in some dimensions exceed — what dedicated supercomputer time could deliver, and that the model could work for chemistry as well as it had for biology, physics, and astronomy.