There are commercially available industrial systems for identifying and separating polymers, for instance PE from PP. However, there is a demand for analyzers that can separate within polymer classes: e.g. PE-LD from PE-HD or different polypropylenes characterised by different melting points. First, the feasibility of a reliable spectral identification was tested by extracting different PE and PP samples from an industrial recycling process, and acquiring diffuse reflectance NIR spectra using an FTIR spectrometer. The resulting spectra were then subjected to a chemometrics analysis. We successfully identified characteristic spectral features; these are determined by the chemical bonds of the material, and can be correlated to the melting points of the materials. These features were then adapted for use on a NIR hyper-spectral (HS) system, making it possible to distinguish not only different polymers, but also different types of one polymer in real-time.
[1]
R. Leitner,et al.
On-Line Classification of Synthetic Polymers Using near Infrared Spectral Imaging
,
2003
.
[2]
S. M. de Jong,et al.
Imaging spectrometry : basic principles and prospective applications
,
2001
.
[3]
Andrew R. Harvey,et al.
Spectral Imaging: Instrumentation, Applications, and Analysis
,
2000
.
[4]
Martin De Biasio,et al.
Real-time detection of flame-retardant additives in polymers and polymer blends with NIR imaging spectroscopy
,
2009,
Defense + Commercial Sensing.