Complex chemistry

Forecasting how chemical species transform from reactants to products has implications for the design of fuels that combust efficiently, for predicting the fate of greenhouse gases in the atmosphere, and in expediting the synthesis of new materials and drugs. Combustion is a particular challenge because it is a transient phenomenon where reactants are quickly, and perhaps violently, converted to products and large amounts of energy. This raises many interesting questions: How predictable are the mechanisms of such complex chemical reactions? How well can we predict the future of the chemistry of combustion, given an observation of an incomplete overall reaction? What features of complex chemistry limit our ability to predict the mechanism? How much information do the sequences of chemical symbols contain about the outcome of the reaction? We are working to answer these questions and more.

Mixing liquids

Natural phenomena mix matter, energy or both. The mixing of liquids is especially important; it is responsible for the molecular structure and function of biological cells, the production and processing of everyday products, and the yield of chemical reactions. For liquids that are sufficiently alike, thermodynamics makes predictions about whether the liquids will mix. However, the entropy associated with the interdiffusion of liquids is difficult to estimate, say, from molecular simulations, especially if the liquids are dissimilar or the mixing requires heat transfer. We have developed theory and computational strategies that enable the direct calculation of statistical entropy changes from the underlying molecular dynamics involved in mixing two liquids. These entropy changes are related to dynamical randomness and, yet, are consistent with the thermodynamic entropy of mixing. We are working to extend this approach from simple liquids to liquids composed of molecules with internal structure.

Disordered kinetics

Kinetics provides predictions about how fast a particular process will happen - its rate. For example, predicting the rate of making and breaking hydrogen bonds in liquid water can give insight into the role of this process in the function of biomolecules. Many such processes present a theoretical challenge to kinetics because of the intrinsic fluctuations in the rate caused by the surrounding environment; the interaction between two water molecules sensitively depends on the proximity, geometry, and behavior of the water molecules surrounding them. We are designing a general theoretical framework that applies to this broad class of rate processes, including water. This framework is being implemented as a computational algorithm to analyze the results of both computer simulations and laboratory experiments.