| Title |
Near-infrared-based determination of mass-based material flow compositions in mechanical recycling of post-consumer plastics: Technical feasibility enables novel applications |
| ID_Doc |
24789 |
| Authors |
Kroell, N; Chen, XZ; Küppers, B; Lorenzo, J; Maghmoumi, A; Schlaak, M; Thor, E; Nordmann, C; Greiff, K |
| Title |
Near-infrared-based determination of mass-based material flow compositions in mechanical recycling of post-consumer plastics: Technical feasibility enables novel applications |
| Year |
2023 |
| Published |
|
| DOI |
10.1016/j.resconrec.2023.106873 |
| Abstract |
Mass-based material flow compositions (MFCOs) are crucial to assess and optimize mechanical plastic recycling processes. While MFCOs are determined by manual sorting analysis today, in the future MFCOs could be determined inline through near-infrared-based material flow characterization. This study aims to quantify the accuracy of near-infrared-based MFCO determinations to assess its technical feasibility. Binary mixtures of plastic flakes and post-consumer packaging were pixel-based classified at different material flow presentations, and mass-based MFCOs were predicted from the resulting false-color data using different data processing techniques. The results show high correlations between near-infrared-based false-color data and mass-based MFCOs. Through regression models and data aggregation, it was possible to predict mass-based MFCOs with mean absolute errors of 0.5% and 1.0% and R2-scores of 99.9% and 99.4% for plastic flakes and packaging, respectively, across all material flow presentations. The demonstrated technical feasibility thus paves the way for new sensor technology applications in plastic recycling. |
| Author Keywords |
Sensor -based material flow characterization; Sensor -based sorting; Automated quality control; Lightweight packaging waste; Machine learning; Circular economy |
| Index Keywords |
Index Keywords |
| Document Type |
Other |
| Open Access |
Open Access |
| Source |
Science Citation Index Expanded (SCI-EXPANDED) |
| EID |
WOS:000995133200001 |
| WoS Category |
Engineering, Environmental; Environmental Sciences |
| Research Area |
Engineering; Environmental Sciences & Ecology |
| PDF |
https://doi.org/10.1016/j.resconrec.2023.106873
|