Computational Cluster-Based Methods for Molecular Thermodynamics

Cluster-based methods in statistical mechanics express fluid properties in terms of power series in density or other state variables. Series coefficients can be determined rigorously from molecular models, and consequently these approaches can produce thermodynamic models that exactly describe the behavior that corresponds to the molecular model, over some range of conditions. This capability provides a unique route for connecting molecular-based treatments, including first-principles methods, to experiment, and in doing so it can guide the formulation of molecular models, methods, and theories; alternatively, this capability provides a modeling tool that can serve a wide range of applications in science and engineering, including diverse fields such as nanotechnology, combustion engineering, environmental science, separations, and more.

We provide an overview of work we and others have performed in recent years to make computational cluster-based methods into a viable tool for molecular thermodynamics. We consider methods for computing cluster integrals, treatment of complications such as molecular flexibility, nuclear quantum effects, and multibody interactions, and methods to extend the range of application of the general approach. We also show ways that these methods can be used in conjunction with experiment and standard molecular simulations to improve interpretation of results obtained there.