Multivariate Insights into Structure–Property Relationships in Biodiesel and Biolubricants from Vegetable Oils

The transition to bio-based fuels requires a thorough understanding of how molecular structure governs key physicochemical properties such as oxidation stability and viscosity. In this study, the combined influence of fatty acid composition and alcohol structure on biodiesel and biolubricant performance is investigated using a multivariate statistical approach. A dataset comprising 108 samples was analyzed, including 17 experimental samples produced in this work and 91 samples collected from peer-reviewed literature. Each sample was characterized by its fatty acid composition and at least one physicochemical property, namely oxidation stability (Rancimat induction time) and/or kinematic viscosity. Principal Component Analysis (PCA) was applied to identify dominant compositional patterns, followed by clustering (k-means) to define homogeneous compositional regions. Within these regions, a variance decomposition approach was used to quantify the relative contribution of alcohol type to property variability. The results show that fatty acid composition defines the primary structural framework governing oxidative stability, largely driven by the degree of unsaturation. In contrast, viscosity is more strongly influenced by the type of alcohol used, particularly in systems involving higher alcohols or polyols. The proposed approach provides a structured multivariate framework focused on analyzing oxidation stability and viscosity, enabling the systematic interpretation of the relative influence of fatty acid composition and alcohol structure on these key physicochemical properties. This study demonstrates that integrating PCA, clustering, and variance decomposition offers a robust strategy for analyzing complex bio-based systems, supporting the rational selection of feedstocks and processing conditions for the optimization of biodiesel and biolubricant formulations.