The field of organic photovoltaics (OPV) has gone through a great deal of turmoil and disruption over the past few years.  Concerns about reproducibility, long-term stability, competing excitement of new materials (i.e. metal-halide perovskites), and the sheer complexity of multiple intertwined variables weigh heavily on researchers in the area.  The race to a single number, a high power conversion efficiency, has obscured the effort put forward by entire teams of researchers, their creativity, and their fundamental insights into OPV and related devices and materials. These challenges facing the field do not, however, diminish the immense potential of being able to print or spray-coat inexpensive and lightweight rolls of plastic solar cells for applications worldwide. What to do, next, then?  In this talk, we describe our forays into applied machine learning, combined with experimental design, towards optimization of OPV to rapidly increase the rate of screening of devices, by simultaneously manipulating several variables. Rapid screening in this fashion provides insights through exploration of design space that is typically overlooked in traditional empirical screening, and can open new avenues for research in OPV by systematically and simultaneously manipulating several variables through design of experiments methodologies.

Optimization of organic solar cell processing parameters: Combining machine learning and experimental design

255th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.