Improved high throughput differentiation of polyolefins (POs) can improve recycling rates and resin quality, as more homogeneous single streams result in consistent and higher value post-consumer resin (PCR) feedstocks.  Near infrared (NIR) spectroscopy is the industry standard for identifying and sorting of mixed plastic wastes, however, finer structural classifications of POs is challenging as information about the chain composition, conformation or topology of polyolefins cannot readily be interpreted from NIR data. Robust, low throughput polymer characterization methods that determine molecular weight, branching distribution, thermal properties, and comonomer content of POs yield important information that affect the final PCR properties, but cannot be done online at recycling facilities. To address this challenge, we have developed a library of NIR spectra and multimodal physical measurements for commercial and standard POs and have correlated the structural information from these slower measurement techniques to NIR spectra    through machine learning (ML) models. Using functional principal component analysis (fPCA), we were able to differentiate structural classifications of POs, such as high density polyethylenes (HDPE), low density polyethylenes (LDPE), and linear low density polyethylenes (LLDPE) by relating the first two functional principal components (functional PCs 1 and 2). Furthermore, we were also able to correlate the fPCA data to measured material properties, such as density, crystallinity, and branch content from offline polymer analytical measurements. Ultimately, we aim to expand our ML database to extract infrared chemical signatures to properties such as molar mass, branching content and melting temperature for “smarter” NIR based classification of polyolefins, validate on double blinded polymer samples and blends, and develop transferable tools to enable recyclers to improve characterization at their facilities.

Improving classification of commercial polyolefins via correlation of near-infrared spectra to bulk properties

Where thousands of chemistry professionals meet to share ideas and advance scientific and technical knowledge. Spring 2023 abstracts, posters, and presentations will launch at the end of March.

ACS Spring 2023

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.