The research field of organic semiconductors is the focus of much attention, as it forms the basis for future organic solar cells (OPVs), as well as organic field effect transistors (OFETs) and light-emitting diodes (OLEDs) – which can be used in applications such as portable solar systems and rollable displays. However, the search for the most effective organic molecules is still in its infancy.

When the possibilities for organic semiconductor materials are almost infinite (estimated to be in the order of 1033), how can scientists begin to process all this data in order to find the best solutions? Now a research team led by Professor Karsten Reuter, director of the Theory Department at the Fritz Haber Institute of the Max Planck Society, Berlin, and Dr Harald Oberhofer, Heisenberg Scholar at the Chair of Theoretical Chemistry, Technical University of Munich, are tackling this challenge with active self-learning machines.

First the scientists carry out simulations with smaller molecules to obtain information about their potential electrical conductivity. Once the principles are established, the algorithm then decides if various modifications to the molecule could lead to useful properties, or whether it is uncertain due to a lack of similar data. In both cases, it automatically requests new simulations, improves itself through the newly-generated data, considers new molecules, and goes on to repeat this procedure.

The team has already showed that this new “active learning” approach is more efficient than alternative algorithms and will save thousands of hours of research and development. It is hoped that the use of artificial intelligence in organic molecule research will quickly usher in the next generation of organic solar cells, which are cheaper to make and more flexible than conventional silicon-based units.