The Use of Computational Quantum Chemistry in Applied Thermodynamics

University Of Delaware, Newark DE

Investigators

Abstract

ABSTRACT CTS-0083709 S. Sandler University of Delaware The unifying theme of the research is the use of modern computational chemistry to improve predictive methods used in the chemical, biochemical and environmental industries. Modern computational quantum mechanics has progressed rapidly in the last decade, both in terms of accuracy and speed, largely as a result of the availability of large-scale, highly parallel computers. To a large extent, the chemical engineering community has not taken advantage of this progress. This project is directed to the application and utilization of computational quantum chemistry in three areas of molecular thermodynamics. The first area of application of computational quantum mechanics is to the improvement of group contribution methods. It is well known that such methods suffer from several defects. The most important is the proximity effect in which the behavior of a functional group is affected by neighboring groups on the same molecule. This is a violation of the basic group contribution concept that a functional group should behave the same independent of the molecule of which it is a part. Here one generalize's a successful, but so far restricted hybrid quantum mechanics/group contribution model that had been developed that corrects the properties and interactions of each group based on how its charge and dipole moment vary from a reference state due to neighboring groups on the same molecule. The second area is the application of quantum chemistry methods to compute the interaction energy landscape between two molecules, and then to use this information in a interaction potential function together with nonadditive multibody effects in Monte Carlo computer simulation. Computer simulation techniques have provided a great deal of qualitative insight into molecular level phenomena. However, the quantitative accuracy of simulations have been poor unless the empirical parameters in the effective two-body potentials have been fit to experimental data. The goal of the this part of the research is to determine whether the phase behavior of mixtures of interest to chemical engineers can be accurately computed using state-of-the-art multibody potential models. The third research area involves the use of the results of combined quantum chemistry/computer simulation calculations for mixtures mentioned above to test the underlying theoretical basis of activity coefficient models traditionally used by chemical engineers. The goalis to improve upon these models, and to provide a method of determining the parameters in these models in the absence of experimental data.