Date of Award

December 2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Environmental Engineering

First Advisor

Mahmut Ersan

Abstract

Microplastics (MPs), defined as plastic particles ranging from 1 μm to 5 mm, have emerged as one of the significant environmental contaminants due to their widespread presence in terrestrial and aquatic ecosystems. MPs originate from primary sources, such as industrial pre-production pellets and microbeads in personal care products, or secondary sources, such as the breakdown of larger plastic particles. Their persistence in the environment is due to their resistance to degradation and capacity to transport toxic substances and pathogens. These characteristics raise concerns about their ecological impacts and potential entry into food chains, ultimately affecting human health. Despite extensive research on aquatic MP contamination, understanding their distribution, behavior, and impact in terrestrial environments remains limited due to lack of robust field-deployable detection technologies.Current detection techniques, such as microscopy and spectroscopy, face significant challenges in analyzing heterogeneous matrices like soils, including interference from organic matter, moisture, and mineral content. These limitations necessitate the development of innovative, efficient, and precise detection methodologies. This study introduces hyperspectral imaging (HSI) as a novel approach to detect and quantify MPs in soil matrices. HSI integrates imaging and spectroscopy to analyze materials based on their spectral signatures, enabling the simultaneous identification of multiple MP types in a non-destructive manner. Unlike traditional techniques, HSI offers high-throughput analysis, reducing the labor intensity and variability associated with manual detection methods. The study focuses on the spectral characterization of three common polymeric plastics, polyethylene (PE), polypropylene (PP), and polystyrene (PS) and evaluates the impact of soil properties, MP characteristics, and environmental weathering on detection accuracy. Using a state-of-the-art HSI system operating in the near-infrared (NIR) spectral range of 1000–1700 nm, pristine MPs of PE, PP, and PS were analyzed to establish baseline spectral signatures. Soil samples with MPs were systematically varied in moisture content, particle size, and color to assess how these factors influence detection. The spectral data underwent preprocessing, including second-derivative smoothing and normalization, to identify distinct absorption features. Results demonstrated that HSI effectively differentiates MP types based on their unique spectral patterns. PE showed characteristic absorption bands at 1040 nm, 1207 nm, and 1413 nm, while PP exhibited more complex patterns associated with its -CH, -CH2, and -CH3 groups. PS displayed high reflectance in the NIR range, with prominent aromatic C–H bands. Soil moisture was identified as a critical factor affecting detection, as increased moisture intensified NIR reflectance, complicating the identification of MPs, particularly at lower concentrations. Larger MPs were more easily detected due to their limited interaction with soil components, while colored MPs demonstrated more pronounced spectral signals. Additionally, weathered MPs subjected to oxidative and photodegradation processes displayed altered spectral properties. Though HSI successfully detected shifts in the spectra and demonstrated its effectiveness in identifying MPs, it also highlights its potential for use in both natural and engineered systems. This study highlights the potential of HSI as an innovative tool for MP detection in complex environmental matrices. By leveraging spectral and spatial data, HSI enables precise, automated analysis while minimizing human error. The findings highlight the importance of addressing factors like soil moisture, MP size, color, and weathering in detection protocols. Standardized methods and calibration frameworks are recommended to enhance reliability and facilitate cross-study comparisons. Furthermore, integrating HSI with machine learning could improve classification accuracy and expand its applicability to diverse environmental contexts. By advancing MP detection technologies, this research contributes to better monitoring and management of plastic pollution, paving the way for more sustainable environmental practices.

Share

COinS