ABSTRACT The amorphous structure and complex local environment of rare‐earth ions in laser glasses have long hindered the prediction of spectroscopic properties and the quantification of composition–structure–property (C–S–P) relationships. This has led to a heavy reliance on trial‐and‐error approaches in material development. In this study, we demonstrate through molecular‐dynamics simulations and experiments that the short‐ and medium‐range structure of a glass can be effectively represented as a statistical ensemble of its nearest‐neighboring congruently melting compounds (CMCs). This modeling approach remains robust even for the chemically complex SiO 2 ‐B 2 O 3 ‐BaO system containing two network formers. Using this approach, we predicted the local environment of Nd 3 + ions and 15 spectroscopic properties across three emission bands (894, 1070, and 1334 nm), with relative errors below 10%. We further developed a high‐throughput computational framework to construct a C–S–P database encompassing over 1000 compositions and generated multi‐dimensional luminescence maps. This framework enables the rapid identification of compositions that satisfy multiple performance targets, offering a rational and efficient pathway for the tailored design of laser glasses.
Dong et al. (Fri,) studied this question.