Scholarlink

<p>The quality of polyethylene (PE) recyclates is crucial for their use in new applications. However, the presence of aged PE products in the recycling stream can reduce PE recyclate quality. To minimize the need for virgin material or additional stabilizers to counteract this quality loss, reliable identification and removal of heavily aged materials is essential for sustainably improving the quality of PE recyclates. In this study, the potential of near-infrared (NIR) hyperspectral imaging for detecting aging and degradation in PE was investigated. For this purpose, PE samples were artificially aged by UV irradiation and their degradation was assessed through mechanical testing. The strain at break from these tests served as the criterion for assigning samples to degradation classes. NIR hyperspectral data of the samples were acquired using a&#xa0;laboratory NIR setup. Separability was then evaluated using classification models based on multivariate data analysis (Partial Least Squares Discriminant Analysis) and a&#xa0;machine learning algorithm (Support Vector Machine), which were subsequently compared. The results show that the machine learning model enables highly accurate discrimination between strongly and less degraded PE samples. This demonstrates that established NIR sorting systems could be expanded beyond polymer-type identification to include targeted detection of degradation, representing an important step toward producing higher-quality recyclates and improving resource efficiency.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Alt oder wertvoll? – Detektion von gealtertem Polyethylen mittels NIR-Hyperspektraltechnik für hochwertiges Kunststoffrecycling

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月26日

查看全文
  • Jutta Geier,
  • Chiara Barretta,
  • Mario Messiha,
  • Katarina Marković,
  • Márton Bredács,
  • Florian Arbeiter,
  • Eric Helfer,
  • Lisa Meinhart,
  • Gernot Oreski

摘要

The quality of polyethylene (PE) recyclates is crucial for their use in new applications. However, the presence of aged PE products in the recycling stream can reduce PE recyclate quality. To minimize the need for virgin material or additional stabilizers to counteract this quality loss, reliable identification and removal of heavily aged materials is essential for sustainably improving the quality of PE recyclates. In this study, the potential of near-infrared (NIR) hyperspectral imaging for detecting aging and degradation in PE was investigated. For this purpose, PE samples were artificially aged by UV irradiation and their degradation was assessed through mechanical testing. The strain at break from these tests served as the criterion for assigning samples to degradation classes. NIR hyperspectral data of the samples were acquired using a laboratory NIR setup. Separability was then evaluated using classification models based on multivariate data analysis (Partial Least Squares Discriminant Analysis) and a machine learning algorithm (Support Vector Machine), which were subsequently compared. The results show that the machine learning model enables highly accurate discrimination between strongly and less degraded PE samples. This demonstrates that established NIR sorting systems could be expanded beyond polymer-type identification to include targeted detection of degradation, representing an important step toward producing higher-quality recyclates and improving resource efficiency.