Large-scale reactive molecular dynamics simulation and kinetic modeling of high-temperature pyrolysis of the Gloeocapsomorphaprisca microfossils.

The ability to predict accurately the thermal conversion of complex carbonaceous materials is of value in both petroleum exploration and refining operations. Modeling the thermal cracking of kerogen under basinal heating conditions improves the predrill prediction of oil and gas yields and quality, thereby ultimately lowering the exploration risk. Modeling the chemical structure and reactivity of asphaltene from petroleum vacuum residues enables prediction of coke formation and properties in refinery processes, thereby lowering operating cost. The chemical structure-chemical yield modeling (CS-CYM) developed by Freund et al. is more rigorous, time-consuming, and requires a great deal of chemical insight into reaction network and reaction kinetics. The present work explores the applicability of a more fundamental atomistic simulation using the quantum mechanically based reactive force field to predict the product yield and overall kinetics of decomposition of two biopolymers, namely, the Kukersite and Gutternberg. Reactive molecular dynamics (RMD) simulations were performed on systems consisting of 10(4) to 10(5) atoms at different densities and temperatures to derive the overall kinetic parameters and a lumped kinetic model for pyrolysis. The kinetic parameters derived from the simulated pyrolysis of an individual component and the mixture of all four components in Guttenberg reveal the role of cross-talk between the fragments and enhanced reactivity of component A by radicals from other components. The Arrhenius extrapolation of the model yields reasonable prediction for the overall barrier for cracking. Because simulations were run at very high temperature (T > 1500 K) to study cracking within the simulation time of up to 1 ns, it, however, led to the entropically favored ethylene formation as a dominant decomposition route. Future work will focus on evaluating the applicability of accelerated reactive MD approaches to study cracking.