A regression-driven study of entropy growth in graph-theoretic models of lithium manganese arsenide

This work is an in-depth study of the topological complexity of Lithium Manganese Arsenide (LiMnAs) through the application of chemical graph theory. By representing the molecular structure in graph form, a number of Zagreb-type topological indices were computed to quantify the connectivity and geometry of the extended LiMnAs lattice. Corresponding Shannon entropies were obtained from the indices for measuring the irregularity and information content in the molecular graph. A logarithmic regression model was utilized to determine the quantitative relationship between every index and its corresponding measure of entropy. The findings show strong logarithmic fits, confirming the application of topological indices as good descriptors of structural entropy. In addition, violin plots were utilized for the visualization of indices and measures of entropy, and the subtleties in structural behavior were highlighted. These results help in the topological characterization of LiMnAs and pave the way for its usage in materials physics and science.