New Quality Control and Sorting Strategies for Polyolefins (PP-PE) Recycling

The aim of this work was the evaluation of a new sensing technology, based on hyperspectral imaging, for identification of plastics in waste streams, with particular reference to polyolefins (polyethylene, PE and polypropylene, PP). It is well known that it is difficult to separate mixed plastics with similar density and surface charge by conventional technologies, such as gravity or electrostatic separations. Starting from a complex waste sample coming from automotive shredder residue (ASR), first different contaminants (wood, rubber, stone, metal, etc.) have been eliminated and then plastics have been subjected to different stages of gravity separation, producing five classes of densities, representative of polyethylene (HDPE and LDPE) and polypropylene. Reflectance spectra in the visible-near infrared field (400-1000 nm) have been acquired by an hyperspectral imaging system for different plastic density fractions sorted by color. Results showed as the spectra of plastics from the different density fractions are correlated not only with particle color, but also with their density and hence with their nature, i.e. PE and PP. Such results demonstrated as the use of hyperspectral imaging for polyolefins identification is promising and should be further investigated for its implementation at industriallevel for on-line sorting and/or quality assessment purposes.