Artificial neural network-based QSPR study on absorption maxima of organic dyes for dye-sensitised solar cells

Abstract

A quantitative structure–property relationship study was performed between descriptors representing the molecular structures and the absorption maxima (λ_max) of organic dyes for dye-sensitised solar cells. The entire set of 70 dyes was divided into a training set of 53 dyes and a test set of 17 dyes according to Kennard and Stones algorithm. Seven descriptors were selected on the training set by genetic algorithm. Based on these seven descriptors, a nonlinear model with the squared correlation coefficient R ² = 0.991 was developed by using artificial neural networks. The reliability of the proposed model was validated through the test set. All descriptors involved in the model were derived solely from the chemical structures of the dyes, which makes the model very useful to estimate the λ_max of the dyes before they are actually synthesised.

Keywords:

• absorption maxima
• QSPR
• organic dyes
• dye-sensitised solar cells
• artificial neural network

Acknowledgements

This work was supported by the Foundation of Wuhan Textile University (No. 2009003), the Natural Science Foundation of Hubei Province (No. 2008CDB261) and China (No. 51003082), the Key Project of Science and Technology Research of Ministry of Education (No. 208089) and the Educational Commission of Hubei Province (Q20101606). The authors gratefully wish to express their thanks to the reviewers for critically reviewing the manuscript and making important suggestions.