Chemometric Mapping of Phenolic Diversity in Lamiaceae Extracts Highlights Solvent-Dependent Patterns

Introduction: Deep eutectic solvents (DES) and natural deep eutectic solvents (NADES) have emerged as sustainable alternatives to conventional organic solvents for the extraction of phenolic compounds from aromatic plants. Multivariate chemometric tools provide a powerful means of evaluating complex extraction datasets and identifying patterns associated with phenolic selectivity across different solvent systems.

Methods: Multivariate chemometric analysis was applied to evaluate the relationships among phenolic compound profiles of ten Lamiaceae species extracted with five solvent systems — two hydrophobic DES, two NADES, and 70% ethanol — yielding a dataset of 50 samples and 12 quantified phenolic compounds. Prior to analysis, data were standardized by z-score normalization. Principal Component Analysis (PCA), Hierarchical Cluster Analysis (HCA), and heatmap visualization were performed to assess sample grouping and identify key discriminating variables.

Results and Discussion: PCA revealed that the first two principal components accounted for 41.8% and 23.7% of total variance, respectively, with a cumulative explained variance of 65.5%. The score plot highlighted sample grouping that was more closely associated with extraction solvent than with botanical origin. The menthol/methyl salicylate DES formed a well-separated cluster associated with high rosmarinic acid, chlorogenic acid, and quercetin loadings, while NADES and ethanol extracts showed partial overlap, which may reflect similarities in their polarity ranges. HCA similarly showed two primary clusters broadly corresponding to hydrophobic DES and hydrophilic extraction systems. Loadings analysis identified rosmarinic acid, chlorogenic acid, and quercetin as principal discriminating variables.

Conclusions: These results highlight solvent-dependent patterns in the phenolic profiles of Lamiaceae extracts and demonstrate the utility of chemometric tools in characterizing phenolic fingerprints and support rational solvent selection for targeted extraction from complex plant matrices.

Acknowledgments: This work is supported by Provincial Secretariat for Higher Education and Scientific Research, Province of Vojvodina (Grant No. 003874431 2025 09418 003 000 000 001).