Unravelling Structure–Property–Performance Relationships in Nanocomposite Membranes Using Principal Component Analysis

This study investigates the structure–property–performance relationships of advanced membrane materials for crystal violet (CV) dye removal using Principal Component Analysis (PCA). A dataset of electrospun nanocomposite membranes, including pristine, PES/PVDF-blended, and nanofiller-modified variants incorporating ZIF-67, graphene oxide (GO), and graphene nanoplatelets (GNPs) was analysed. PCA reduced dataset complexity and revealed two dominant components explaining 76.50% of the total variance. The first component (PC1, 54.66%) represented a performance axis driven by viscosity, porosity, pore size, flux, and rejection, while the second component (PC2, 21.84%) captured a wettability–processing axis governed by contact angle and flow rate. Based on the dominant PCA loadings, two reduced PCA-derived descriptors were formulated: the Rheology–Structure–Transport Index (RSTI) and the Hydrophilicity–Processing Index (HPI). These descriptors provide compact, physically interpretable summaries of the main PC1 and PC2 tendencies within the analysed dataset, rather than externally validated predictive models. The results show that high membrane performance can be achieved through distinct physicochemical pathways depending on nanofiller type. The proposed PCA-based framework offers a robust and interpretable tool for analysing complex membrane datasets and provides valuable guidance for the rational design and optimisation of advanced membranes for wastewater treatment.