Unraveling the mechanism of a novel laser-induced graphene for adsorptive fuel desulfurization: Box-Behnken process optimization and statistical physics modeling

In this work, laser-induced graphene (LIG), synthesized directly from polyimide via ultraviolet laser irradiation was for the first time investigated an adsorbent for deep adsorptive desulfurization (ADS) to remove thiophene (TP) and benzothiophene (BT) from a model fuel. Characterization results of the LIG revealed it to be a three-dimensional, defect-rich, hierarchically porous graphene architecture with a specific surface area of 186.71 m²/g, providing appreciable accessible adsorption sites. Process conditions affecting adsorption performance were optimized using response surface methodology based on a Box–Behnken design with statistical analysis confirming that quadratic model adequately represented the adsorption process and reliably predicted optimal conditions. Under the optimized parameters, LIG demonstrated high removal efficiency for both sulfur compounds, with BT exhibiting a stronger adsorption affinity than TP. The adsorption process shows that LIG has higher affinity toward BT than TP, reaching equilibrium within 60 mins. Deeper mechanistic insight was obtained through advanced statistical physics modeling which suggested heterogeneous energy distributions, horizontal molecular orientation, and a multi-docking adsorption mechanism governed primarily by active site density rather than steric occupancy. The process is predominantly governed by physisorption as signaled by adsorption energies values below 40 kJ/mol. Furthermore, the regeneration studies demonstrated that LIG retains over 70% of its desulfurization efficiency after multiple cycles, signifying its structural stability and reuse potential. The obtained results confirmed LIG as a promising material to be a reliable adsorbent for ADS while also illustrating the significance of combining classical isotherms and statistical physics approaches to gain insight into the adsorption mechanisms.