Polymers represent around 25% of total waste from electronic and electric equipment. Any successful recycling process must ensure that polymer-specific functionalities are preserved, to avoid downcycling. This requires a precise characterization of particle compounds moving at high speeds on conveyor belts in processing plants. We present an investigation using imaging and point measurement spectral sensors on 23 polymers including ABS, PS, PC, PE-types, PP, PVC, PET-types, PMMA, and PTFE to assess their potential to perform under the operational conditions found in recycling facilities. The techniques applied include hyperspectral imaging sensors (HSI) to map reflectance in the visible to near infrared (VNIR), short-wave (SWIR) and mid-wave infrared (MWIR) as well as point Raman, FTIR and spectroradiometer instruments. We show that none of the sensors alone can identify all the compounds while meeting the industry operational requirements. HSI sensors successfully acquired simultaneous spatial and spectral information for certain polymer types. HSI, particularly the range between (1600–1900) nm, is suitable for specific identification of transparent and light-coloured (non-black) PC, PE-types, PP, PVC and PET-types plastics; HSI in the MWIR is able to resolve specific spectral features for certain PE-types, including black HDPE, and light-coloured ABS. Fast-acquisition Raman spectroscopy (down to 500 ms) enabled the identification of all polymers regardless their composition and presence of black pigments, however, it exhibited limited capacities in mapping applications. We therefore suggest a combination of both imaging and point measurements in a sequential design for enhanced robustness on industrial polymer identification.

中文翻译：

---

多传感器表征可改进 WEEE 回收中聚合物的识别

聚合物约占电子电气设备废物总量的 25%。任何成功的回收过程都必须确保保留聚合物特定的功能，以避免降级回收。这需要对加工厂传送带上高速移动的颗粒化合物进行精确表征。我们使用成像和点测量光谱传感器对 23 种聚合物（包括 ABS、PS、PC、PE 类型、PP、PVC、PET 类型、PMMA 和 PTFE）进行了一项调查，以评估它们在回收操作条件下的性能潜力设施。所应用的技术包括高光谱成像传感器 (HSI)，用于绘制可见光到近红外 (VNIR)、短波 (SWIR) 和中波红外 (MWIR) 的反射率，以及点拉曼、FTIR 和光谱辐射计仪器。我们表明，没有任何传感器可以单独识别所有化合物，同时满足行业操作要求。HSI 传感器成功地同时获取了某些聚合物类型的空间和光谱信息。HSI，特别是（1600-1900）nm之间的范围，适用于透明和浅色（非黑色）PC、PE类、PP、PVC和PET类塑料的特异性识别；MWIR 中的 HSI 能够解析某些 PE 类型的特定光谱特征，包括黑色 HDPE 和浅色 ABS。快速采集拉曼光谱（低至 500 毫秒）能够识别所有聚合物，无论其成分和是否存在黑色颜料，但其在测绘应用中的能力有限。因此，我们建议在顺序设计中结合成像和点测量，以增强工业聚合物识别的鲁棒性。