PhaseTime: A Flory-Huggins-Based Computational Framework for Predicting Time-Dependent Phase Diagrams of Reactive Polymer Blends

Abstract

Reactive polymer blends often undergo phase separation while their molecular constitution is still evolving. In such systems, stability limits and coexistence boundaries are not fixed properties of the initial mixture, but change with reaction progress, temperature, conversion, or another imposed protocol variable. This makes the interpretation and prediction of reaction-induced phase separation challenging, particularly when the components are polydisperse and the coexisting phases may differ in both composition and molecular distribution. Here we present PhaseTime, a Python framework for computing time-or protocol-dependent phase diagrams of reactive polydisperse polymer blends within Flory-Huggins-type thermodynamics. The framework combines evolving molecular-distribution models with calculations of spinodal curves, critical points, binodal curves, and cloud/shadow curves. It supports temperature-and composition-dependent interaction parameters, monodisperse reactiondependent approximations, and polydisperse Flory-Stockmayer distributions. The framework also includes parameter fitting, diagnostic, and plotting tools through both a command-line interface and a Python API. Representative calculations demonstrate idealized phase-diagram topologies, including UCST, LCST, combined UCST/LCST, hourglass, and closed-loop behavior, as well as fitting to literature data and reaction-dependent phase behavior in monodisperse and polydisperse systems. The resulting phase diagrams provide quasi-equilibrium thermodynamic references for interpreting reactioninduced phase separation and for comparison with spatially resolved models in the fast-demixing limit.