Minimum Spanning Tree-Based Clustering for Chemical Evaluation of Commercial Nail Polish Samples Using Spectroanalytical Data

Abstract

This study discusses potential toxic elements detection in conventional nail polish, including Cr and Pb. The noteworthy results highlight well-established potential risks of elevated Cr and Pb concentrations. These elements are not allowed in the European Union. Implementing the minimum spanning tree (MST) approach and the isolation forest algorithm effectively clustered samples. Forty-five samples were analyzed, and four clusters were identified. Two presented six samples with high concentrations of Fe (Cluster 1 with four samples) and Cr and Pb (Cluster 2 with two samples). The other 39 samples presented low concentrations of the determined elements (Co, Cr, Cu, Fe, Ni, and Pb). Cadmium, Zn, and Mn were not detected in any of the analyzed samples. Furthermore, integrating energy-dispersive x-ray fluorescence (ED-XRF) and laser-induced breakdown spectroscopy (LIBS) enabled fast direct analysis of nail polish samples, streamlining a swift and reliable data acquisition process. This research underscores the importance of ongoing vigilance and monitoring of potential health hazards associated with nail polish formulations, especially in regions with regulatory restrictions on certain elements.