Machine Vision for Optoelectronic Perovskite Film Characterization

We present our research for fast and reliable bandgap and absorption quality value extraction for triple-cation perovskite thin films from sample scans. To this end, thin film samples were synthesized via blade-coating and their photoluminescence, ultraviolet-visible spectra and the film thickness were collected. For qualitative evaluation of the absorption of perovskite films for use in photovoltaic modules we propose a method of computing a dimensionless figure of merit we called the Area Under Absorption Coefficient (AUAC). Our approach consists of perform regression tasks aimed at predicting the properties of interest using machine learning methods, namely convolutional neural networks. Similar methods have been reported in literature before [1,2,3], while the novelty lies in using very simple imaging techniques, as well as the concept and prediction of the AUAC.<br/>This work demonstrates the usability of simple imaging techniques to characterize repeatedly fabricated experimental samples while requiring only a feasibly acquirable initial amount of data. Our reported method can help speed up time consuming material optimizations by reducing lab time spent on recurrent characterizations, nicely synergizes with high throughput production lines and could be adapted for quick extraction of other optoelectrical quantities.<br/><br/><br/>[1] Srivastava, M., Howard, J. M., Gong, T., Rebello Sousa Dias, M., & Leite, M. S. (2021). Machine Learning Roadmap for Perovskite Photovoltaics. The Journal of Physical Chemistry Letters, 12(32), 7866–7877. doi:10.1021/acs.jpclett.1c01961<br/>[2] Taherimakhsousi, N., MacLeod, B. P., Parlane, F. G. L., Morrissey, T. D., Booker, E. P., Dettelbach, K. E., & Berlinguette, C. P. (2020). Quantifying defects in thin films using machine vision. npj Computational Materials, 6(1), 111. doi:10.1038/s41524-020-00380-w<br/>[3] Chenyu Zhang, Jie Feng, Luis Rangel DaCosta, Paul.M. Voyles, Atomic resolution convergent beam electron diffraction analysis using convolutional neural networks, Ultramicroscopy, Volume 210, 2020, 112921, ISSN 0304-3991, https://doi.org/10.1016/j.ultramic.2019.112921.