Solvatochromic rationalization of co-solvent effects on the synthesis of biodiesel

Biodiesel is a promising fuel due to its environmental benign nature. However, there are lots of challenges inherent in the quantum and quality of synthesized biodiesel due to the influence of multi-reaction conditions and co-solvents effects on biodiesel synthesis. This work however investigates the effect of selected solvents on solvent solvatochromic properties. Empirical models were developed to establish quantitative relationships between the solvatochromic properties on biodiesel yield both for homogeneous and heterogeneous catalysis. For the later, IBM SPSS was used to perform regression analysis, yielding a regression model with an R² value of 0.985. The results indicate that biodiesel yield increases with higher solvent dipolarity/polarizability (π*), acidity (α), and dielectric constant (ϵr), while it decreases with higher normalization value (ETN), refractive index (n), and basicity (β). For the homogeneous catalyst, machine learning was employed using Google Colab Jupyter Notebook, resulting in a regression model with an R² of 1.00, indicating perfect predictive accuracy. In contrast to the heterogeneous catalyst model, ETN, β, and ϵr had the most significant positive effects on biodiesel yield, whereas acidity (α), dipolarity (π*), and refractive index (n) negatively impacted biodiesel yield.The findings highlight distinct solvent effects depending on the catalyst type and provide a systematic framework for solvent selection in biodiesel production. The developed models can be applied to predict biodiesel yield for various solvents, enabling the selection of optimal co-solvents for enhanced biodiesel efficiency and sustainability.