New Technique for Recycling Plastics

@@ Every year, over 15 million tons of plastics are produced in Japan. Not many different kinds�@(types) of plastics, but also the same plastics of different characteristics or composition, in other words, plastics of different grades are manufactured. In recent years, along with other increasing environmental problems, plastic waste recycling has been one of the essential issues. In case of the plastic wastes recycling, improving the characteristics of the recycled materials is expected. At the same time, the plastic wastes should be sorted according to their types and what is more, the same kinds of plastics are to be discriminated according to their grades. This is mainly due to the economical reason. The values of the plastic goods manufactured vary tremendously depending on the grades of the recycled plastic materials used. For example, high-density polyethylene (HDPE) is more than twice as expensive as low-density polyethylene (LDPE).
@@ To discriminate different kinds of plastics, a non-destructive measurement by near infrared spectra is known as an effective technique. We have already succeeded in developing a technique to discriminate more than 50 different kinds of plastic patterns. However, because the discrimination of the grades by the near-infrared spectra measurement had been considered difficult until recently, a medium infrared spectra measurement had been the main method applied for that purpose. On the other hand, non-destructive measurements by the medium infrared spectra are not as easy compared with near-infrared spectra measurements, thus, a rapid discriminating technique using near-infrared spectra measurements has been desired. On the basis of this understanding, using polyethylene as a sample plastic, we have investigated the possibility of the grade discrimination of HDPE and LDPE by combining the near infrared reflection spectra measurements with the neural network analyses.

@@ The near infrared reflection spectra in the 1.1~2.2ƒÊm wavelength region on the PE films and sheets of 14 kinds (HDPE 6, LDPE 8) of sample materials are measured. The derived spectral data were learned by the neural network and the discriminating tests were conducted. As a result, the spectra of 14 kinds of PE were discriminated 100% to HDPE and LDPE. The results of the principal component analyses showed that HDPE and LDPE were distinctively dispersed in the space. Therefore, it became clear that HDPE and LDPE are able to discriminate by combining near infrared spectra measurements with chemometrics analyses.

@@ Furthermore, this technique not only can be applied to discriminate HD and LD, but also enables to determine a quantitative density. Moreover, the application of this technique has the possibility to discriminate the grades of polymers other than PE.