Computer models and simulations in phase equilibria, conformational analysis, and chemical engineering education

Public Deposited

The prediction of multicomponent adsorption equilibria from pure component data alone has been an active research area for over forty years. The Ideal Adsorbed Solution theory (IAS) has been the standard method of doing so, but contains an embedded assumption of Langmuir mixing. This assumption causes its predictions to be very poor in some cases. I present a method by which an arbitrary mixing model can be specified by the user, resulting in superior predictions to IAS if the mixing model is more reflective of physical reality than the Langmuir model. The framework in which this method is developed can easily be applied to equilibria between any two phases. Molecular squares are an exciting new class of materials that have great potential in catalysis, molecular recognition, and other commercially important applications. We performed molecular dynamics (MD) simulations of molecular squares in solution using the force fields of Sarkisov et al. for the squares and Chalaris et al. for the solvent. The simulations quantitatively predict aspects of the experimental wide-angle X-ray scattering patterns, suggesting that this force field may be useful for future simulations of binding and other applications. Chemical engineering undergraduates tend to lack proficiency in computer programming, numerical methods, and agent-based modelling. A sequence of problems is presented to aid their understanding of these topics in the context of a mass transfer course. The numerical solution of boundary value diffusion problems is demonstrated using MATLAB. Students were introduced to agent-based modelling by a Monte Carlo simulation, also in MATLAB, of cells diffusing into and reproducing within a polymeric tissue engineering scaffold. The code for solving these problems is publicly available at the group website.

Last modified
  • 08/30/2018
Date created
Resource type
Rights statement