Delta Algorithm for Developing Models of Binary Amorphous Systems from Diffraction Data

Abstract

A new algorithm was proposed for generating models of low-temperature two-component amorphous systems from partial pair correlation functions (PPCFs). In this algorithm, the pair force functions are consecutively given small increments in the form of a delta function so as to attain better agreement between the model and diffraction (objective) PPCFs. The delta algorithm was used to construct models of amorphous Fe2Tb and Ni63.7Zr36.3 . The results indicate that the static structure of the amorphous system can be assessed very accurately. However, the problem of deriving the force functions from the PPCFs has no unique solution. Accordingly, dynamic properties such as autocorrelation functions and densities of vibrational states also cannot be determined uniquely.