Dresden 2020 – scientific programme

The DPG Spring Meeting in Dresden had to be cancelled! Read more ...

Parts | Days | Selection | Search | Updates | Downloads | Help

O: Fachverband Oberflächenphysik

O 42: Focus Session: Innovation in Machine learning PRocEsses for Surface Science (IMPRESS)

O 42.1: Invited Talk

Tuesday, March 17, 2020, 10:30–11:00, TRE Phy

Exploring the Design Space of Organic Semiconductors with Machine Learning — •Harald Oberhofer — Chair for Theoretical Chemistry, Technical University Munich

Organic electronics---in the form of field effect transistors, light emitting diodes, or solar cells---are slowly finding their use in everyday consumer devices. So far though, one of the main challenges holding back their wide-scale adoption are their low intrinsic charge carrier mobilities. Improving these is usually attempted by structural tuning of a promising compound family, thereby relying on intuition, experience, or simply trial and error. While sometimes quite successful, such incremental changes only lead to a local exploration of the vast chemical space of possible molecules, potentially overlooking many interesting materials.

In contrast, modern data mining strategies allow the extraction of general design rules through the systematic evaluation of large compound databases. Starting from an analysis of such a database consisting of >64.000 molecular crystals we evaluate the impact of molecular scaffolds and side groups on the charge transport properties of each crystal contained in our database to reveal statistically reliable, general design criteria. A visualization of the chemical space contained in our dataset highlights large gaps in the experimentally covered range of synthesized organic materials. Adopting an active learning strategy we venture to fill these gaps to uncover promising new organic semiconductor materials.