General Theory of Inhomogeneous Systems, Based on Maximum Entropy

Abstract

Inhomogeneous systems are modeled by use of the maximum-entropy method so that only explicit properties of their structure are taken into account, with no arbitrary assumptions. A phase diagram describes the evolution of one inhomogeneous system into another. It contains first- and second-order transitions, with lattice animals, percolation, and a lattice gas as special cases. Two order parameters are identified. A tentative mapping onto the Potts model results in continuously varying critical exponents. New methods, including Monte Carlo, are proposed for maximum-entropy problems.