Towards recycling of textiles: fast detection of contaminated waste fabric by near-infrared spectroscopy.

Since the implementation of EU Waste Framework Directive, which mandates the separate collection of textile waste from 2025 onwards, there has been an increasing need for effective systems for textile waste collection, sorting, and recycling, together with reliable approaches to assess the chemical safety of textile products. The present study explores near-infrared (NIR) spectroscopy, combined with chemometric tools, proposed as a non-destructive rapid approach to identify the presence of contaminants on textile samples. Representative contaminants were applied to pre-consumer virgin polyester, virgin cotton and post-consumer polyester fabrics. This combined strategy succeeded in distinguishing contaminated textiles from uncontaminated ones, supported by statistical analysis with principal components (PCA), enhanced with feature selection by Minimum Redundancy Maximum Relevance (mRMR) and supervised classification using k-Nearest Neighbors (kNN). This NIR-based integrated method, implemented on a miniaturized NIR spectrometer (MicroNIR), is a promising solution to fast screen contaminated textile waste and produce safe secondary raw materials, reducing analysis time and costs of textile recycling workflows.