Black Plastic Identification for Sorting and Recycling With Deep-UV Raman Spectroscopy

We present a deep-ultraviolet (DUV) Raman spectrometer for plastic identification for sorting purposes in a recycling facility. Plastic recycling entails processing large volumes of inhomogeneous plastic waste of different colors, sizes, and polymer types. Currently, hyperspectral imaging in the near-infrared is one of the leading techniques for plastic identification for sorting purposes, but it falls short for dark or black plastics due to strong absorption. Alternatively, Raman spectroscopy could be used; however, conventional Raman technology at visible wavelengths often suffers from a strong fluorescence background that overwhelms the Raman signal, hampering the identification of black plastics in particular. In contrast, Raman spectroscopy in the DUV wavelength range is suitable for black plastics because the fluorescence emission that occurs over the near-UV and visible range will not overlap with the Raman spectrum. Here, we present a dedicated DUV setup, based on a 248.6-nm NeCu laser and designed and tested for stand-off detection of black plastic waste objects on a conveyor belt. We characterized key parameters such as focal depth and detection at different conveyor speeds, emphasizing real-world applicability under recycling factory conditions. In this work, we focus on polyethylene (PE), polypropylene (PP), and polyethylene terephthalate (PET), as these are considered commercially interesting. Analysis of the deep-UV spectra led to promising results for identification, for both stationary and moving conditions. We also demonstrate successful identification of multilayer materials (metalized or black coated) that are challenging for hyperspectral NIR techniques. We discuss advantages of this DUV Raman modality, options to overcome its current shortcomings, and its future potential for polyolefin sorting in a recycling factory.