Using Handheld NIR Spectrometry in the Classification of Polyolefin Films

Plastic waste and its management are major environmental challenges. Near-infrared (NIR) spectroscopy has emerged over time as a cutting-edge technology for classifying and separating plastic waste, particularly polyolefins, which dominate the packaging industry. While NIR spectroscopy has proven effective for sorting polyolefin plastics, multilayer films present additional hurdles. One challenge is the difficulty of classifying thin films, which produce weaker spectra. The use of reflective backgrounds like copper or aluminum can enhance spectral quality, increasing the accuracy of identification. A recent study explored how different reflective materials—such as copper, aluminum, gold, silver, Teflon, and white tile—affect the classification of multilayer polyolefin films. By using a handheld NIR spectrometer, researchers aimed to optimize the classification process, making it possible to differentiate between polyolefin and non-polyolefin films with a single measurement, even for thin films. Their findings could contribute to better recycling processes, helping to address the growing plastic waste problem. Spectroscopy spoke to Hana Stipanovic, corresponding author of the paper resulting from this study, about this research.

The abstract of your paper (1) begins “The low thickness of plastic films poses a challenge when using near-infrared (NIR) spectroscopy as it affects the spectral quality and classification.” Please elaborate on these challenges and their ramifications.

The low thickness of plastic films poses a challenge in the application of near-infrared (NIR) handheld spectroscopy by impacting the spectral quality which eventually affects the classification accuracy as well as reproducibility of the results. This is crucial in the classification of the plastic materials for recycling, as it demands both speed and accuracy of the measurement. The thinner the film, the weaker the spectroscopic signal, resulting in a lower signal-to-noise ratio.

Why do you believe, as you state later in your paper, that NIR spectroscopy has proven to be the state-of-the-art technology for plastic waste classification and separation, as opposed to other spectroscopic techniques?

Mainly because of its combination of speed, accuracy, safety and robustness which also enabled it to be the main technology in automated sorting facility because of the ability to handle high-throughput and complex waste streams, such as plastic waste with the goal of sorting of different types of plastics found in the waste.

Are there any drawbacks to using NIR as opposed to other vibrational spectroscopy techniques?

Yes; while NIR spectroscopy offers many advantages for waste classification and sorting, when it comes to plastic waste, the primary drawback is the challenge of classifying dark colored plastics. This is because carbon black, often used for black coloring of the plastics, fully absorbs NIR light hampering the spectral analysis and eventually classification and sorting of black colored plastics.

The material measured for this study was described as a 2D plastic packaging fraction (film) originating from the eject stream of the NIR sorting step of a material recovery facility (MRF) in Austria. Why did you choose this material?

This material was selected because it aligns without requirements for thin plastic materials and is derived from post-consumer waste. One of our project’s goals is to explore methods for sorting challenging post-consumer plastic materials, with a particular focus on polyolefin plastics, to enable their use in chemical recycling. Chemical recycling offers an alternative to mechanical recycling for certain plastic types, such as mixed polyolefins. However, achieving this requires high input material purity, making the work presented in this publication a crucial step toward that goal.

Briefly summarize your findings that you discuss in your article, and the conclusions you came to after reviewing these findings.

The main findings of the publication are the method development for the rapid classification of waste plastic multilayer polyolefin films using an NIR handheld spectrometer, while investigating the impact of different measuring backgrounds, including Teflon, aluminum, copper, silver and gold, on the classification. Given that the materials used were thin plastic films, the study explores how these backgrounds could enhance spectral quality and ultimately improve spectral accuracy, contributing to a more reliable classification method. Based on the findings, we could conclude that it is possible to develop a method using an NIR handheld spectrometer to classify polyolefin plastic films and distinguish them from non-polyolefin films. Additionally, the use of metallic backgrounds significantly enhanced spectral quality, enabling method development and accurate classification.

Do you anticipate similar results with different plastic materials?

Yes; in addition to the results presented in this publication, experiments were performed using four different plastic film fractions, which also yielded high accuracy. However, as this was not the primary focus of our work, we did not investigate further. Nevertheless, these findings contributed to the development of a method for classifying various types of plastic materials found in light packaging waste collection, including both thin and thicker and monolayer and multilayer plastic packaging. This method has been applied in-house for analyzing the composition of the lightweight packaging waste in Austria.

What advantages does the portability of the instrument you used provide?

The portability of the instrument used offers flexibility in its usage, allowing it to be used outside the traditional laboratory settings, such as in automated waste sorting plants to test the sample of the input material, support training of automated NIR systems or to support manual waste sorting processes. It also can enhance accessibility of the measurements and improve overall efficiency of the sorting processes.

What difficulties did you encounter in your work, specifically sampling and analytical challenges?

The sampling was conducted as a separate part of our project, and we essentially received the samples from them. If you are interested in the process of sampling for the experiments and overall gaining of plastic waste from municipal waste I would invite you to read the publication “Recovery of plastic packaging from mixed municipal solid waste. A case study from Austria” from Blasenbauer and associates (2). Among the remaining challenges, the weak signal from thin films posed a challenge which was then resolved with the usage of the metallic backgrounds. Another was also the influence of those backgrounds which also could not have been fully ruled out but was the lowest when using metallic backgrounds and did not seem to influence the classification results, especially after performing pre-processing steps.

How did you process the spectroscopic data to obtain the results you were looking for?

The processing steps involved the scatter-correction method, specifically standard normal variate (SNV), and the spectral derivative, utilizing the Savitzky-Golay second derivative with five smoothing points. These steps were selected after extensive testing of various pre-processing method combinations, as this combination achieved the highest classification accuracy.

Were there any factors that might affect the accuracy of your findings?

Sure; multilayer plastic films are highly engineered, sophisticated products that can consist of multiple layers. In the publication, the classification focused specifically on the two outer layers, without exploring the composition of the inner layers. This approach was taken because elements like additives within the inner layers were outside the scope of the publication. While they could potentially influence the classification, no such impact was observed in our results. Another influencing factor is the human element. Since the spectrometer is handheld, its position can vary depending on how the user holds it. These positional changes could naturally impact the results.  

How do you imagine the results of your study can/will be applied?

The results are expected to serve as supporting analytics for training of automated NIR sorting systems. The goal is to enhance training by providing classification results for all plastic materials available, without excluding any due to technical challenges in classification. Additionally, applications beyond plastic classification are anticipated, particularly in scenarios where improved spectral quality is required due to the low thickness of the materials.

Are there any next steps in this research?

Yes, based on the results demonstrating the potential of using an NIR handheld spectrometer for classification and sorting of complex waste materials, further experiments and developments are planned. These efforts aim to enhance classification and sorting capabilities for other complex waste streams, addressing the current obstacles and proposing solutions while also defining where are the limits of handheld spectrometers in the area.

References

1. Stipanovic, H.; Arth, P.; Koinig, G.; Kuhn, N.; Lederer, J.; Blasenbauer, D.; Lipp, A. M.;Tischberger-Aldrian, A. Influence of Different Measuring Backgrounds on the Classification of Multilayer Polyolefin Films Using a Near-Infrared Handheld Spectrometer. Appl. Spectrosc. 2024, 37028241307034. DOI: 10.1177/00037028241307034

2. Blasenbauer, D.; Lipp, A. M.; Fellner, J.; Tischberger-Aldrian, A.; Stipanović, H.; Lederer, J. Recovery of Plastic Packaging from Mixed Municipal Solid Waste. A Case Study from Austria. Waste Manag. 2024, 180, 9–22. DOI: 10.1016/j.wasman.2024.02.040

Hana Stipanovic is a research assistant and a PhD candidate at the Chair of Waste Processing Technology and Waste Management at Montanuniversitaet Leoben, Austria.

She earned her bachelor’s degree in the field of Mining Engineering and her master’s degree in Waste Management and Disposal from the University of Zagreb, Croatia. Afterwards, she continued to work in waste management before starting her PhD journey. In her PhD she explores the material-based characterization of various types of waste using NIR spectroscopy with focus on plastic and textile wastes. This work combines her background in mining engineering, where NIR spectroscopy is important mineral processing technology, with her expertise in waste management and sorting for recycling. Through her work, she aims to contribute to more efficient and sustainable waste sorting solutions, while also identifying the limitations of NIR spectroscopy in waste sorting and exploring potential solutions to overcome these challenges.