Renewable organic materials are an essential block for sustainable economic growth while mitigating climate change. Industrial biomass utilization as an alternative to fossil fuels and products is an essential first step towards manufacturing these renewable materials. A promising feedstock for biofuel and biomaterial is the aromatic heteropolymer lignin, found abundantly in plant secondary cell walls. Lignin phenolic structures offer unique functionality for biomaterial applications, however this functionality needs to be liberated from the native lignin polymer before use. Real-world lignin depolymerization has been carried out through various thermochemical treatments, hydrolysis, pyrolysis, hydrogenolysis etc. Most widely and industrially applied depolymerization technique is the pyrolysis of lignin. To simulate the depolymerization process <i>in silico,</i> prior studies have applied reactive force fields to a limited number of lignin topologies that are unrepresentative of native lignin. Recent advances in lignin modeling and recently developed neural network potentials enable reactive molecular simulation for a broad sampling of native lignin structures. Starting from stochastically generated lignin structures, we carry out high-temperature lignin simulations with reactive force fields in LAMMPS, and compare the results with the recently developed neural network approach called the Accurate NeurAl networK engINe for Molecular Energies (abbreviated as ANAKIN-ME or ANI). Based on an evaluation of the bond-breakage sequences observed through simulation, we observe different reaction pathways emphasized through both techniques. The varied reaction pathways results in different end product distributions, which offer suggestions for how pyrolysis conditions can be tuned to improve the value of the final products.

Understanding depolymerization of lignin through neutral network based reactive molecular dynamics simulation

ACS Fall 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.