Advanced characterization techniques for man-made vitreous fibers: a multimodal approach for hazard assessment and recycling potential

Man-made vitreous fibers (MMVFs), widely used in construction, are classified as hazardous or non-hazardous based on their chemical and geometric properties. However, the lack of rapid, in situ methods to assess their hazardousness, particularly in construction and demolition waste (CDW), has limited recycling efforts and led to misclassification. This study proposes a preliminary combined chemometric approach using Scanning Electron Microscopy (SEM), micro-X-Ray Fluorescence (XRF) and Hyperspectral microimaging (HSI) to achieve a faster and more effective MMVFs characterization. Together, these techniques facilitate the development of automated models for fiber classification. A three-step Principal Component Analysis (PCA) was carried out to distinguish among three primary fiber classes and two secondary subclasses. The proposed approach demonstrates strong predictive capability, consistent with the PCA and SEM and XRF analyses. It offers a path to obtain a fast and user-independent alternative to conventional fiber characterisation methods, with the future objective of supporting improved waste management decisions, promoting fiber recovery and reuse in line with circular economy goals. Future work will focus on refining the model and expanding its fiber classification capabilities.