Data driven enhancement of mid-infrared non-linear optical properties of quaternary and ternary chalcogenides

In the present paper, we primarily aim to understand the trend in the nonlinear optical (NLO) properties of some ternary and quaternary chalcogenides compounds for the mid-infrared (mid-IR) coherent light generation using data-mining techniques. We have noticed through the literature an important number of data have been produced for NLO properties of chalcogenides. Some of the challenges in searching through discrete data include the difficulty of analyzing large amounts of data, understanding the correlations among various properties, and using the correlations to better understand the underlying physics of the system. Utilizing a multivariate analysis, the data can be examined so that trends and correlations become apparent. A multiple selection criteria approach is proposed to screen the data rationality. Additionally, we demonstrated that it is possible to predict new materials that achieve a balance between the band gap and second harmonic generation (SHG) effect using a Partial Least Square model and existing knowledge on NLO materials, whether experimental or theoretical. However, predicting the band gap (Eg) solely based on the number of periods and groups in the periodic table poses challenges due to variations in stoichiometric coefficients and compound phases. To address this, future work could include additional parameters such as the molar mass (M) to better represent the identity of each compound.