Rule-based network analysis of complex reaction systems in biomass conversion

Catalytic upgradation is a method of conversion of lignocellulosic biomass to transportation fuels and chemicals. Such catalytic systems inherently have a large and complex reaction network, elucidation of which is key in reactor design. Moreover, cascades of multifunctional catalytic systems have recently been proposed for upgrading sugar monomers to transportation fuels. Here, we present rule-based systematic computational analyses for: (i) enumerating and analyzing complex catalytic biomass conversion systems, and (ii) formulation of potential synthetic routes to produce valuable biomass-derived molecules. We employ RING, a rule-based network generation and analysis tool with an english-like domain specific user-friendly front-end developed by us earlier, in conjunction with ab initio- and DFT-based thermochemistry analysis, as the computational platform for these analyses. Dehydration to acrolein, in solid acid catalyst systems, is a promising route for value addition of by-product glycerol, obtained from the trans-esterification process to produce biodiesel. Known elementary steps of solid acid catalysis, such as adsoprtion of alcohols, dehydration, desorption of alkoxides (modeled as carbenium ions), hydride shifts, decarbonylation, hydration, deprotonation, protonation of carbonyl group, etc. were input into RING as reaction rules to study the reaction system presented by Corma et al. The resultant reaction network for this system was more than 70,000 reactions in total. Subsequently, the network was analyzed in terms of extracting reaction pathways leading to products identified in the experimental data at various conditions. The analysis of these pathways further corroborated the observations and discussions made by Corma et al. Specifically, the shortest paths (and paths up to two steps longer) clearly indicated that 3-hydroxypropionaldehyde was an essential intermediate in acrolein production, obtained from glycerol upon dehydration, and hydrogen transfer or enol-keto type rearrangement. Acetone, on the other hand, necessarily required acetol or propane diol as an intermediate, with glycerol undergoing two dehydration steps and one or more rearrangement or hydrogen transfer steps. It was further identified that acetaldehyde production would be accompanied necessarily by the evolution of carbon monoxide. This case study, thus, shows that RING can be employed in the analysis of hypothesized mechanisms of complex reaction systems in a rule-based manner, starting with likely elementary steps and querying for pathways matching experimental observations. Lignocellulosic biomass upgradation pathways can potentially involve acid catalyzed hydrolysis and dehydration, metal catalyzed hydrogenation and hydrogenolysis, and base catalyzed esterification, ketonization and condensation reactions. Reaction rules describing the overall steps of such a broad spectrum of catalytic chemistries were input into RING with sugar monomers and/ or model oxygenates as reactants. Pathways were traced, in the resultant network, to identify multi-step-catalytic reaction cascades for synthesizing valuable compounds such as long chain alkanes. RING identified proposed synthesis routes in the literature and, in addition, proposed additional potential routes. These pathways can further be analyzed thermochemically to identify feasible/ infeasible and bottleneck steps, and the thermodynamically favorable order of deoxygenation steps. Application of RING in (a) rule-based systematic and exhaustive analysis of glycerol dehydration pathways in solid acid catalytic systems, and (b) formulating synthesis strategies for biomass upgradation, as well as thermochemical analysis of the identified pathways, will be presented with representative examples.