Machine Learning Assisted Chemical Process Parameter Mapping on Lignin Hydrogenolysis

Lignin depolymerization has been studied for decades to produce carbon-neutral chemicals/biofuels and biopolymers. Among different chemical reaction pathways, catalytic hydrogenolysis favors reactions under relatively mild conditions, while its yield of bio-oil high-value aromatic products is high. In this study, the influence parameters on lignin are discussed by process parameter mapping modeled using three machine learning algorithms based upon literature experimental data. The best R2 scores solid residue were 0.92 0.88 xgboost, respectively. importance was examined, it observed that lignin-to-solvent ratio average pore size have a larger impact results. Finally, optimal conditions predicted best-fit model, which indicates further improvements can potentially generate higher yields in industrial applications.