Microscopic Theory of Softness in Supercooled Liquids

We introduce a new measure of the structure liquid which is softness mean-field potential developed by us earlier. find that this sensitive to small changes in structure. then study its correlation with supercooled dynamics. The involves wide range liquids (fragile, strong, attractive, repulsive, and active) predicts some universal behaviours like linearly proportional temperature inversely activation barrier dynamics system dependent proportionality constants. write down master equation between parameter show indeed when scaled parameters data collapse plotted against softness. fragile strong dependence whereas much weaker dependence. also connect present earlier studies involving machine learning (ML) thus providing theoretical framework for understanding ML results.