Cheminformatics for Combustion of Hydrocarbons
Accurate transport and thermodynamic properties of chemical species and kinetic data of important elementary reactions involving such species are of utmost critical for molecular-level detailed simulations of complex reacting systems. One of such complex systems is the combustion of hydrocarbons. The ability to perform detailed simulations of such processes would have significant impacts on many industries as well as on our environment. To achieve this goal one must be able to develop fundamental kinetic model, which in principle consists of all elementary chemical reactions existing in the system along with their kinetic parameters and of thermodynamic and transport properties of all chemical species involved. The completeness of the kinetic model does not require any assumption regarding the reaction pathways. This becomes extremely important due the increasing interest in describing formation of not only major products but also minor pollutant by-products such as dioxin, furan, and soot. In practical combustion systems, kinetics models are in the order of thousands of elementary reactions or more and a large number of reactive intermediates. This requires an automated method that can generate reactions based on a set of known reaction classes. Once the mechanism is generated, one needs to supply thermodynamic and transport properties for each species and kinetic parameters of each reaction. The greatest challenge arises from this is that kinetic information of most of elementary reactions involved in combustion of hydrocarbons is not known either experimentally or theoretically.
Our ultimate goal is to develop detailed mechanisms for combustion of hydrocarbons and to provide accurate thermodynamic and kinetic information for these systems. Combustion of hydrocarbons such as alkanes is perhaps industrially the most important combustion system due to its use in many processes such as gas turbine engines, heater, incinerators, and hydrocracking furnaces that utilize combustion of methane and natural gases (C2H6, C3’s) to provide fuel.
Toward this end we have been advancing in two fronts, namely 1) developing both accurate and cost-effective methods for predicting thermal rate constants; and 2) developing an automated mechanism generator.Direct ab initio dynamics methods for predicting reaction rate from first-principles
We have been developing direct ab initio dynamics methods based on the variational transition state theory augmented by multi-dimentional tunneling methods for calculating thermal rate constants from first principles. The potential energy information needed for dynamical calculations is calculated directly from ab initio electronic structure theories. We are currently developing methods for calculating micro-canonical variational transition state theory rate constants, k(T), k(E), and k(E,J) based on a full quantum RRKM methodology. Currently, we are expanding our interest to study multi-channel multi-well reactions using the Master Equation method.Reaction Class Transition State Theory
Recently we introduced a new theory called
Class Transition State Theory (RC-TST) for predictions of thermal rate
constants for a large number of reactions in a given class. The RC-TST method recognizes that reactions in
a given class having the same reactive moiety, therefore their
surfaces along the reaction coordinate are very similar and thus can be
extrapolated from one to the others. Furthermore, we have shown that
given reaction class there is a linear energy relationship (LER)
barrier heights and reaction energies. Combining
both facts, the RC-TST/LER theory provides a
methodology for estimating thermal rate constants of any reaction only
reaction energy, which can be calculated from a relatively low level of
such as a semi-empirical molecular orbital method.
We have demonstrated the applicability of
this theory for several hydrogen abstraction reaction classes and
we can predict thermal rate constants within a factor of two. The RC-TST/LER method provides a much needed
tool for generating a large number of kinetic data that currently not
The key bridge that connects first-principles quantum chemistry and reaction engineering is a tool for automatically generating detailed mechanisms of complex reaction systems such as combustion of hydrocarbon fuels or atmospheric chemistry. Such mechanisms can consist of the order of hundreds of thousands of reactions and it would be extremely difficulty to ensure completeness if one is generating using some manual approach. We have been developing a new automatic complex mechanism generator based on the chemical graph theory (called COMGEN). Currently, COMGEN can generate a complete mechanism based on most reaction types reported in the literature for combustion of hydrocarbons. Molecular species are represented externally by SMILE string notation. The main advantage of using such notation is the available technology for converting SMILE string to 3D molecular structure and thus facilitate direct interface with first-principles quantum chemistry.