# Threshold-Filtered Kinetic Monte Carlo Simulation for Real-Time Simulation and Control of Biomass Fractionation

## TL;DR

This paper introduces a faster simulation method for modeling complex biomass reactions, improving efficiency without losing accuracy.

## Contribution

A threshold-filtered kinetic Monte Carlo framework that accelerates lignin fractionation simulations by excluding low-probability reactions.

## Key findings

- The model reduces CPU time by orders of magnitude while preserving accuracy in lignin molar mass and S/G ratio predictions.
- The method enables efficient, data-independent modeling of complex reaction networks in lignin valorization systems.

## Abstract

For lignin and other polymer reaction systems, reaction
kinetics
are inherently complex. Individual bond-cleavage reactions exhibit
a wide distribution of activation energies due to structural heterogeneity
among β-O-4 linkages. In conventional kinetic
Monte Carlo (kMC) simulations, reactions with high activation energies,
whose probabilities are extremely low, are still evaluated at every
step, leading to substantial computational cost with negligible impact
on system evolution. To overcome this limitation, we developed a threshold-filtered
kMC framework that accelerates multiscale lignin fractionation simulations
by excluding kinetically irrelevant events using an Arrhenius-type
activation energy threshold. The model preserves its fidelity while
reducing CPU time by orders of magnitude compared with conventional
approaches. It accurately predicts the evolution of lignin molar masses
and S/G ratios under varying reaction conditions. This strategy enables
efficient, data-independent modeling of complex reaction networks
and establishes a scalable tool for process-level optimization and
control in lignin valorization systems.