Chemical structure effects on coal pyrolyzates and reactions by using large-scale reactive molecular dynamics

The relationship between the coal chemical structure and its thermal reactivity is vital to understand coal pyrolysis behaviors. In order to explore chemical structure effects on pyrolysis process, five large-scale coal models of different ranks were constructed and simulated with ReaxFF MD simulations by a combined approach of high performance computing and cheminformatics based reaction analysis in this work. The qualitative temperature mapping results between ReaxFF MD simulations and thermogravimetry experiments were obtained for the first time through the detected covalent bond breaking, which suggests a promising scheme to map the simulation results to the real world. Importantly, the typical structures in coal can be used as the indicators to predict pyrolysis stages, weight loss profiles and major pyrolyzate distributions from the atomistic level. The starting temperature of coal thermal decomposition is anchored by the parameters of falO in 13C NMR representing alkyl ether amounts; and the largest phenol tar generation links closely to the faP NMR parameter; meanwhile methoxy groups determine the initial generation of CH3 radicals and CH4 at relatively low temperature. Additionally, the dynamic profiles of Car-Car bonds and the second increasing trend for CH3 radicals have strong relationship with recombination reactions. With the reasonable coal structures, the large-scale ReaxFF MD simulation alone can complement experimental observation comprehensively to understand the complex coal thermal chemistry and used as a preliminary screening approach to select the coal type or rank for industrial utilization.