Assessing the Lost Fraction: Diversity, Abundance, and Mass of Microplastics (1-300 µm) in Aquatic Systems

Summary

This protocol describes a methodology to be applied to aquatic samples collected on filters to detect, identify, and quantify micron-sized (1-300 µm) MPs. Raman microspectroscopy can identify the polymeric chemical structure of MP particles and quantify their abundances in terms of the number of particles and their mass in water samples.

Abstract

The protocol presented here enables the quantification of microplastics (MPs) as small as ~1 µm in diameter, accurate identification of polymer types, and estimation of particle volume, critically allowing for the calculation of MP mass. Representative results from samples collected in the Great South Bay (GSB), NY, showed that particles within the 1-6 µm equivalent spherical diameter (ESD) range were the most abundant, with approximately 75% of particles measuring less than 5 µm. Notably, the pre-sieving step failed to yield any particles larger than 60 µm, suggesting that large MPs were rare at the coastal sites sampled. Prior to filtration, a chemical oxidation step was used to remove organic debris, which facilitated the filtration of larger water volumes (>1 L) from discrete bottle-collected samples and reduced filter clogging. While this approach significantly improved filtration efficiency, aspects of the methodology still require refinement to reduce the total time required for sample preparation and data analysis. Raman microspectroscopy and associated data processing remain time-intensive, particularly for accurately analyzing particles smaller than 10 µm in complex environmental matrices. Ongoing efforts are focused on minimizing analytical uncertainties, optimizing the trade-off between particle counting accuracy and processing time, and reducing artifacts in the detection workflow.

Introduction

Microplastics (MPs) are a highly heterogeneous group of particles that vary widely in size, shape, color, density, chemical composition, and other physical properties, making their identification and quantification particularly challenging¹. A variety of analytical methods have been employed in MP surveys, from visual techniques such as light microscopy² to more advanced chemical approaches like pyrolysis-gas chromatography³. Light microscopy is a commonly used method for characterizing larger MPs (typically 0.5-5 mm) due to its simplicity, speed, and cost-effectiveness¹. It enables the visual identification and quantification of plastic-like particles while providing information on surface morphology and structure². However, this technique does not yield chemical information about the chemical composition and relies on the analyst's expertise. Misidentification is a well-documented limitation of visual analysis, with reported error rates ranging from 20% to 70%, particularly for small, transparent particles. Such errors can significantly overestimate MP concentrations in environmental samples⁴.

On the other hand, pyrolysis-gas chromatography (GC) combined with mass spectrometry (MS) can provide detailed information about the chemical composition of MPs by analyzing thermal degradation products³. When paired with a thermal desorption step, it can also detect plastic additives during the same analysis⁵. However, this method requires manual pre-selection of plastic-like particles and their careful placement into the pyrolysis system, resulting in low sample throughput¹.

To overcome the limitations of traditional methods, automated techniques that enable both identification and quantification of MPs are becoming increasingly essential. Raman and Fourier-transform infrared (FTIR) spectroscopy provide accurate polymer identification through their unique spectral fingerprints^6,7,8. These vibrational spectroscopy methods detect molecular vibrations induced by light, laser for Raman and infrared for FTIR, producing spectra that reflect specific molecular structures⁷. Both techniques have significantly advanced the detection and characterization of small MP particles (<300 µm) in various environmental matrices⁸. They reduce false positives by chemically confirming plastic-like particles, minimize false negatives, and are non-destructive. When coupled with microscopy (microspectroscopy), these techniques allow the identification and quantification of polymeric particles smaller than 0.3 mm (with µFTIR detecting down to ~10 µm and µRaman to ~0.3 µm)⁸. Each method, however, has limitations: Raman offers higher spatial resolution and enhanced detection for certain polymers (e.g., polystyrene), but is more susceptible to fluorescence interference and typically requires longer acquisition times. Nevertheless, Raman microspectroscopy can detect particles smaller than 0.3 µm, covering a broader microplastic size range than FTIR imaging^8,9.

The primary tool for the work described here is a confocal Raman microspectrophotometer available within SoMAS' NAno-Raman Molecular Imaging Laboratory (NARMIL). The protocol has been optimized to detect, identify, and quantify micro-sized MPs (<300 µm) on filters from water samples that usually are excluded when plankton or manta nets are used to sample marine plastics. Also, a Chemi-oxidation process is necessary to reduce the presence of non-plastic residue (organic debris). This minimizes analytical interference prior to Raman microspectroscopic analysis and, importantly, prevents the clogging of filter pores, allowing the filtration of larger volumes of water. Using this technology, it is possible to detect MP particles in the sub-micron to millimeter size fraction collected on filters from water samples, and it enables the calculation of specific MP concentrations, particle sizes, and masses.

Protocol

This protocol has been modified from the original version⁹. The protocol is a guidance for the analysis of aquatic samples (drinking water, river, lake, coastal, or open ocean) to extract, detect, and quantify MPs (<300 µm) on filters. The reagents and equipment used are listed in the Table of Materials.

1. Quality control

1. Ensure to filter all reagents through a 0.22 µm membrane filter before use.
2. Ensure the personnel wear cotton lab coats (no synthetic materials) and wash their hands before and between sample handling.
3. Rinse the filtration equipment with 0.22 µm-filtered distilled water before and between filtrations. Cover with clean aluminum foil when not in use.
4. Field blanks: Fill rinsed sample bottles with 0.22 µm-filtered distilled water and process like environmental samples.
5. Laboratory blanks: Process 1 L of 0.22 µm-filtered distilled water through Al₂O₃ filters alongside each sample batch.

2. Collection and preparation of natural water samples for MP quantification

1. Pre-rinse all sample bottles (including Niskin and glass containers) with 0.22 µm-filtered distilled water to minimize contamination.
2. Discrete water column samples (>1 L) from aquatic systems can be collected using standard sampling methods, e.g., glass bottles (surface water) or rosette-mounted polyvinyl chloride (PVC) Niskin bottles (water column) and transferred directly into glass bottles.
3. Prefilter collected water using 200 µm and 60 µm metal sieves to remove organic debris and separate MPs into size fractions.
   NOTE: Retained particles can be identified directly under the microscope.
4. Store the water that passes through the 60 µm metal sieve in glass bottles. Later, filter this fraction onto aluminum oxide (Al₂O₃) membrane filters (25 mm diameter, 0.2 µm pore size).

3. Chemi-oxidation of water samples

NOTE: To minimize analytical interference from abundant biogenic particles, samples are cleaned by chemi-oxidation prior to Raman microspectroscopic analysis.

1. Combine 100 mL of 30% (v/v) hydrogen peroxide with 1 L of sample in a glass container and mix thoroughly.
   CAUTION: Handle H₂O₂ with care - strong oxidizer, may cause skin and eye irritation. Wear gloves and goggles, and follow lab safety procedures.
2. Cover the container with a lid or aluminum foil and place it in an oven at 60 °C for ~10 h.
3. Let the sample cool for 10 min, then proceed to filtration while still warm.

4. Filtration of water samples

NOTE: Due to the filter's inherent brittleness, care should be taken not to bend the Al₂O₃ membrane filters during mounting and removal by exclusively handling the filter by the annular polypropylene ring bonded to the membrane.

1. In a laminar flow hood, filter 1 L samples and blanks using an all-glass filtration system with a sintered glass base and metal clamp.
2. After filtration, rinse the filters with 0.22 µm-filtered distilled water.
3. Mount the filter on a microscope slide without a coverslip for Raman analysis.

5. Raman microspectroscopic analysis of filtered MP particles

NOTE: A confocal Raman microspectrophotometer and its software are used as an analytical method to identify and quantify MP particles. The instrument is configured with a modified upright epifluorescence microscope, a computer-controlled motorized x-y-z stage with 0.1 µm step size in all dimensions, four laser lines (457/514 nm Ar+ ion laser, 633 nm He/Ne laser, 785 nm diode laser), and a 1,040 × 256 peltier-cooled CCD detector.

1. Using the image acquisition feature in the Raman software, set up the instrument configuration for MPs analysis of aquatic samples.
   1. Laser: 633 nm He/Ne at ~9.3 mW laser power (via 50× objective).
   2. Grating: 1200 lines/mm.
   3. Spectral range: 200-2500 cm^-1.
   4. Exposure time: 0.7 s/spectrum.
2. Set up a grid within a region of interest (ROI) by taking spectra every 1 µm in the x and y dimensions.
3. Define each ROI as a 100 µm x 100 µm grid in the camera image (10,000 µm²).
   NOTE: ROI area can be modified; for example, smaller areas will reduce the total time of analysis.
4. Analyze ~40 ROIs per filter, randomly selected using a cross-pattern model to represent MP spatial distribution across the entire exposed filter area.
5. After automated data collection, use the image acquisition feature in the software to produce detailed 2-D chemical maps of each ROI.
6. From each chemical map, identify and quantify MP particles using component analysis (non-negative least squares correlation method) according to software capabilities.
7. Compare spectra with reference libraries. Accept matches with score >60 (internal library score 0-100) as confirmed MPs¹⁰.

6. Filter-based quantitative enumeration of MP particles

1. After totaling the number of particles from ~40 ROIs, calculate the number of MP particles per liter of sample (N), according to equation 1¹¹.
       (1)
   where Af is the total exposed filter area (mm²), n is the total number of MP particles counted summed across all examined ROI, b the number of ROI examined, M the area size of one ROI (mm²), and V the volume (l) of the water filtered from the sample.
2. Establish the analytical uncertainty of the total number of particles (N) in each sample by calculating the standard error from all ROI counted per sample.
3. Report the total number of particles per L as N ± 1S.E.

7. Particle size, volume, and mass calculations

NOTE: Determine the dimensions, area, and equivalent spherical diameter (ESD) of each particle with the aid of an image analysis software.

1. Determine the area by counting the total plastic-positive pixels assigned to each particle image.
2. Use this particle area (Ap) to calculate the ESD by applying equation 2¹¹:
       (2)
3. Calculate an idealized ellipsoid volume estimate for each particle using equation 3¹².
       (3)
4. Calculate the masses (M) of each plastic polymer in the MP pool using volume estimates and published specific densities (ρ; g cm³) of each plastic polymer (i.e., M = ρV). Mass is usually reported in µg/L
   NOTE: To estimate the uncertainty propagated in the mass calculation (Mass relative error, RE = SD/mean) for each particle, analytical uncertainty in both the volume algorithm and mean density is considered by applying the following equation¹³:
       (4)
   Where R.E. vol is the relative error (S.E. of the slope/slope) of the linear regression slope (3-D and "2-D volume")⁹. R.E. den is the relative variations (S.E. / mean) in the range of reported density values for each polymer contained in a specific particle.

Results

Here, results obtained using the most recent version of the methodology are presented. The method has been modified from its original application^9,14. Samples used in this article are part of a project to assess MP pollution in three main coastal water bodies of Long Island, NY: Great South Bay, the Peconic Estuary, and Shinnecock Bay. The results presented here serve as a proof-of-concept for our improved method rather than a systematic assessment of water quali...

Discussion

This methodology enables the quantification of microplastic (MP) particles as small as 1 µm from water samples collected on filters, allowing for detailed measurements of particle abundance, chemical composition, size, and mass. However, it is essential to recognize and address sources of uncertainty inherent in this approach, particularly when analyzing particles smaller than 10 µm in complex environmental matrices.

The current lack of standardized procedures for determi...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Aluminum foil	any commercial brand
Analytical stainless-steel sieves	Retsch GmbH, Germany	DIN 4188
Anodisc Filters	Whatman
Filter holder	FisherBrand-Millipore Sigma	XX1012542
Glass Bottles	FisherBrand-Millipore Sigma	02-912-313	Any glass container works for the sampling process
Glassware funnel	FisherBrand-Millipore Sigma	XX1012514
Hidrogen Peroxide 30 %	Innovating Sciences	7722-84-1
inVia confocal Raman microspectrophotometer	Renishaw
Laminar Flowhood	Labconco or any other laminar hood fro laboratoy use
Metal clamp	FisherBrand-Millipore Sigma	XX1012503
Metal forceps	FisherBrand-Millipore Sigma	13-820-061
Microscope slides	Thermo Fisher	https://www.thermofisher.com/
Nitrile gloves	Kimtech	55080-54
Nitrocelulose membrane fliters 0.22	GVS filter technology	1214898
Radial heat Oven	Lab Line Instruments	3609
Vacuum Pump	Welch	25468-01 A
WiRE 5.2	Renishaw