Machine Learning Analysis and Discovery of Zero-Dimensional ns(2) Metal Halides toward Enhanced Photoluminescence Quantum Yield

https://doi.org/10.1021/acs.chemmater.1c02725

The dependence of photoluminescence quantum yield (PLQY) on the crystal structure of existing zero-dimensional ns2 metal halides is analyzed with the help of principal component analysis and random forest methods. The primary role of the distance between metal ions in different compounds is revealed, and the influence of other structural features such as metal-halogen distance and the distortion of metal-halogen polyhedrons are quantified. Accordingly, the two previously unknown Sb3+-based zero-dimensional metal halides were synthesized to verify the obtained model. Experimental studies of the two compounds demonstrated good agreement with the predictions, and the PLQY of (C10H16N)2SbCl5 is found to be 96.5%. Via machine learning analysis, we demonstrate that concentration quenching is the main factor that determines PLQY for all s2 ion metal halides, which will accelerate the discovery of new luminescence metal halides.