A critical review on molecular modeling approaches for green solvent screening in the extractive removal of aromatic sulfur-containing, aromatic nitrogen-containing, and aromatic hydrocarbon compounds from fossil fuels

As global energy demand rises, emissions of pollutants such as sulfur-containing, nitrogen-containing, and aromatic hydrocarbon compounds from fossil fuels combustion have increased. Considering their adverse effects on the environment and human health, removal of these contaminants has become essential. Liquid-liquid extraction has emerged as a promising technique for this purpose. However, conventional organic solvents used in this process often suffer from high volatility, toxicity, and flammability. Ionic liquids (ILs) and deep eutectic solvents (DESs) offer green alternatives that can overcome these limitations. Experimental approaches for identifying suitable green solvents are typically costly and time-consuming, making molecular modeling techniques increasingly important for solvent screening. Computational methods such as Molecular Dynamics (MD) simulations, COnductor-like Screening MOdel (COSMO)-based models, and Quantitative Structure-Property Relationship (QSPR) modeling have proven effective in understanding and predicting solute-solvent interactions. This review discusses these three modeling approaches in the context of desulfurization, denitrogenation, and aromatic compound removal using ILs and DESs. The findings highlight the value of computational tools in selecting green solvents and optimizing process efficiency, thereby reducing reliance on experimental trials. This study also identifies key research gaps by comparing the strengths and limitations of various solvent screening approaches in the extractive removal of aromatic sulfur-containing, aromatic nitrogen-containing, and aromatic hydrocarbon compounds.