Field-Portable Microplastic Sensing in Aqueous Environments: A Perspective on Emerging Techniques.

Microplastics (MPs) have been found in aqueous environments ranging from rural ponds and lakes to the deep ocean. Despite the ubiquity of MPs, our ability to characterize MPs in the environment is limited by the lack of technologies for rapidly and accurately identifying and quantifying MPs. Although standards exist for MP sample collection and preparation, methods of MP analysis vary considerably and produce data with a broad range of data content and quality. The need for extensive analysis-specific sample preparation in current technology approaches has hindered the emergence of a single technique which can operate on aqueous samples in the field, rather than on dried laboratory preparations. In this perspective, we consider MP measurement technologies with a focus on both their eventual field-deployability and their respective data products (e.g., MP particle count, size, and/or polymer type). We present preliminary demonstrations of several prospective MP measurement techniques, with an eye towards developing a solution or solutions that can transition from the laboratory to the field. Specifically, experimental results are presented from multiple prototype systems that measure various physical properties of MPs: pyrolysis-differential mobility spectroscopy, short-wave infrared imaging, aqueous Nile Red labeling and counting, acoustophoresis, ultrasound, impedance spectroscopy, and dielectrophoresis.

Nondestructive Extraction and Identification of Microplastics from Freshwater Sport Fish Stomachs.

Microplastics were extracted from freshwater sport fish stomachs containing substantial biomass and identified using optical microscopy, scanning electron microscopy plus energy-dispersive X-ray spectroscopy (SEM/EDS), and Fourier transform infrared (FTIR) micro-spectroscopy with automated spectral mapping. An extraction method is presented that uses a negatively pressurized sieve stack and purified water to preserve plastic surface characteristics and any adsorbed persistent organic pollutants (POPs). This nondestructive extraction method for large predators' stomachs enables multiple trophic-level studies from one fish sampling event and provides other dietary and behavioral data. FTIR-identified microplastics 50-1500 µm, including polyethylene (two with plastic additive POPs), styrene acrylonitrile, polystyrene, and nylon and polyethylene terephthalate fibers 10-50 µm wide. SEM/EDS revealed characteristic surface weathering on the plastic surfaces. The nylon fibers appear to be from human fishing activities, suggesting options for management. Some particles visually identified as potential plastics were revealed by micro-spectroscopy to be mineralized, natural polyamide proteins, or nonplastic shell pieces. A low-cost, reflective sample preparation method with stable particle mounting was developed to enable automated mapping, improved FTIR throughput, and lower detection size limit. This study yielded 37 intact prey items set aside for future analyses.

Analytical precision assessment for microplastic analyses.

The need for improved microplastic (MP) data accuracy has been widely reported, but MP precision issues have been investigated less thoroughly. This work demonstrates how initial and continuing assessments of a laboratory's analytical precision can be used for establishing laboratory repeatability for MP analyses. These precision estimates can be reported along with MP results to ensure their quality and compare them meaningfully to other data. Re-analyses of reference MP samples can be used to assess and compare precision between different laboratories. A multi-lab precision exercise was demonstrated using infrared (IR) standard test methods performed on reference samples consisting of low-concentration MP spikes in both clean water and wastewater matrices. Each lab repeated their IR analyses 7 times and calculated relative standard deviations (RSD) for each detected polymer type using a standardized template. All labs' MP methods yielded generally repeatable results, though RSDs were consistently higher for lower MP counts. The reported range of total MP counts per sample was 8-33 particles, and the observed RSDs were 0.1-0.6. These RSDs were the same or lower than the expected imprecision due to random (Poisson) counting error alone, suggesting that these automated methods did not contribute any additional variability, and had slightly better reproducibility than expected for independent recounts. The wastewater matrix exhibited numerous interfering particles but did not yield more variability than the clean water matrix. The low-count design is a worst case for precision but is appropriate for some real-world sample concentrations. In practice, labs could generate separate references for precision assessment at multiple MP ranges (e.g., high, medium, and low.) The RSDs obtained from this data can be used to generate QC charts, detect changes in analyst performance, compare to Poisson error to identify additional sources of imprecision, and determine target filtration and instrumental parameters for MP analyses.