Structure–Property Correlation Study for Organic Photovoltaic Polymer Materials Using Data Science Approach

Abstract

A study workflow that utilizes several data science methods to apply on polymer materials databases is introduced to reveal correlations among their properties, structural information, and molecular descriptors. The data science methods used in this pipeline include the unsupervised machine learning (ML) method of self-organizing mapping (SOM) and the polymer molecular descriptor generator, both of which have been tailored to fit the polymer materials study. To demonstrate how this pipeline can be applied in this context, we used it on an organic photovoltaic (OPV) donor polymer database to investigate which properties or structural factors positively correlate with the power conversion efficiency (PCE) of OPV materials. This led us to discover that among the studied 8 properties and 11 molecular descriptors, only the photon energy loss (E loss ) and the number of fluorine atoms (nF) show strong positive correlations with PCE values, which is consistent with other verified studies. We also discovered that research trends can also be statistically visualized using our method. In our case study, we found that most of the studied OPV donor materials in the database have branched side chains and typically 7−12 non-hydrogen atoms, and high PCE materials usually have 6−9 aromatics rings as well. These results proved that the data science pipeline proposed in this study provides a fast and effective way to obtain research insights for polymer materials.

Experiment‐Oriented Machine Learning of Polymer:Non‐Fullerene Organic Solar Cells

“…Moreover, ML algorithms can instantaneously give an answer by considering big data, which is impossible for human brains. Thus, ML‐based studies are being conducted in different areas such as a lithium‐ion batteries, [ 6–8 ] thermoelectronics, [ 9–11 ] organic light‐emitting diodes, [ 12–14 ] organic photovoltaics (OPV), [ 15–33 ] and organic‐inorganic hybrid perovskite solar cells. [ 34–41 ]…”

Impact of substituents on the performance of small-molecule semiconductors in organic photovoltaic devices via regulating morphology

“…[163][164][165][166] We also note that the use of computer-aided approaches should be widely extended to accelerate the understanding of the structuremorphology-performance relationship of OPV materials. There have been already several studies in this context, [167][168][169][170][171][172][173] and research efforts in this area will aid achieving PCEs nearing the theoretical limit.…”

High-throughput virtual screening for organic electronics: a comparative study of alternative strategies

“…Because of the way ML methods are conceived, they are unlikely to provide guidance on the maximum achievable performance on each technological domain as in the examples presented in Section 4.d but they can be very powerful in discovering novel correlations and ranking the importance of the features included in any model in particular with the advances of explainable ML. 439 For example, recent studies have showed how ML can be a fast and efficient way to select effective molecular features correlated with the PCE of polymer materials, 444 and can be used to rationalise the effects that the feature's trends have on the open-circuit voltage ( V OC ), short-circuit current ( J SC ) and fill factor (FF) of OPVs. 435 …”