Utilization of Laser Induced Breakdown Spectroscopy and Raman spectroscopy data fusion for classification of polymer materials

presentation, poster

en

The question of sustainable solution to the disposal of used polymer materials is resonating in the world for several years now. The straightforward way of dealing with the waste polymer materials is recycling. To optimize the recycling process, automation needs to be included as much as possible. One of ongoing problems in the automation of the recycling process is the classification of polymer materials. With various colors and qualities of material included in the recycling process, the method to classify the polymer type needs to be robust, but also very fast. The joint utilization of Laser Induced Breakdown Spectroscopy (LIBS) and Raman spectroscopy data can offer a solution to this problem. In this work we show, that the fusion of data of these two spectroscopic methods reach impressive classification accuracy on colored polymer samples of several types. The main advantage of using this method is the rapidness and almost none requirements for the sample preparation. These attributes in connection to the almost perfect (99 %) classification accuracy show potential to use this method directly in situ at recycling plant.

The question of sustainable solution to the disposal of used polymer materials is resonating in the world for several years now. The straightforward way of dealing with the waste polymer materials is recycling. To optimize the recycling process, automation needs to be included as much as possible. One of ongoing problems in the automation of the recycling process is the classification of polymer materials. With various colors and qualities of material included in the recycling process, the method to classify the polymer type needs to be robust, but also very fast. The joint utilization of Laser Induced Breakdown Spectroscopy (LIBS) and Raman spectroscopy data can offer a solution to this problem. In this work we show, that the fusion of data of these two spectroscopic methods reach impressive classification accuracy on colored polymer samples of several types. The main advantage of using this method is the rapidness and almost none requirements for the sample preparation. These attributes in connection to the almost perfect (99 %) classification accuracy show potential to use this method directly in situ at recycling plant.

Laser Induced Breakdown spectroscopy; Raman spectroscopy; data fusion; classification; PCA; LDA; Random Forest; polymer recycling

15.12.2022

1