Measuring Microplastics in Remote and Pristine Environments

Plastic pollution is a significant issue globally. It is estimated that 19–23 million metric tons of plastic waste end up in water sources yearly (1–2). In particular, microplastics (MPs), which are plastic particles less than 5 millimeters in size, are increasingly recognized as pervasive environmental pollutants (3). In the atmosphere, they are transported through complex processes involving wind, precipitation, and turbulence (1,3–4). Originating from sources like urban dust, road traffic, industrial emissions, and even oceanic sprays, MPs are carried over long distances, crossing continents and reaching remote regions, including polar areas and mountain peaks (3,4). They can settle out of the atmosphere via dry deposition or be scavenged by rain and snow, contributing to their accumulation in terrestrial and aquatic ecosystems (4). This atmospheric transport poses potential risks to ecosystems and human health, as these particles may carry adsorbed pollutants or harmful additives (4).

Aleksandra “Sasha” Karapetrova is a graduate student at the University of California, Riverside and Win Cowger is the research director at the Moore Institute for Plastic Pollution Research, and they are exploring this issue. Recently, they investigated atmospheric deposits of MPs found in remote snow areas along the western coast of the United States using µ-Fourier transform infrared (µFT-IR) spectroscopy and Open Specy software. Spectroscopy recently sat down with them to talk about their research, which offers new insights into the long-range transport of MPs.

Aleksandra “Sasha” Karapetrova (left) of the University of California, Riverside, and Win Cowger (right) of the Moore Institute for Plastic Pollution Research. Photo Credits: Sasha Karapetrova and Win Cowger.

Could you summarize the key objectives of your study on microplastic (MP) transport in the atmosphere and their detection in snow?

Sasha Karapetrova (SK): The key objective of this study was to determine the concentration, particle size, polymer type, and particle shape of MPs in North American snowpacks, which are large reservoirs for freshwater in the West. We wanted to gain a better understanding of MP atmospheric transport and subsequent deposition in the snowpack by characterizing MPs in places that they have not been studied before. We can close the knowledge gaps in the atmospheric MP cycle, specifically the part of the cycle where atmospheric MPs are depositing into hydrologically significant environments. MP pollution is a part of all our lives and disproportionately affects some more than others. As an environmental toxicologist, I am interested in researching fate and transport of contaminants and how this may impact public health exposure risks.

In your study, you took samples from 11 sites across three main locations, Mount Hood National Forest, Inyo National Forest, and the Seward Peninsula. How did you choose the 11 sampling sites across western coastal North America, and what factors influenced the selection of these locations and sites?

SK: The snow sampling was a concerted effort by many people. Funding for traveling for the fieldwork was provided by the Consortium of Universities for the Advancement of Hydrologic Science, Inc. (CUAHSI) Pathfinder grant. I wasn’t sure what our detection probability was, so we sampled different types of snow of various ages and proximity to anthropogenic activity. Accessibility to the sites was important for safety reasons and to sample undisturbed snow. All sites were away from avalanche-prone areas and were safely accessible with the gear that we were carrying. It was also important to sample snowpacks that were valuable water resources for millions of people, so we could understand how much contamination might be present in freshwater sources.

At the time of sampling, there were no studies in North America looking at snow. There had been a paper (5) that just came out on rain in Utah–Colorado area (Plastic rain in Protected Areas of the United States), so it felt dicey sampling only one type of site. As a result, we asked a biologist, Jason Clark, to collect samples while doing a beaver survey on the Seward Peninsula. This site was one of the most remote sites in the data set, and it was meant to provide a reference site. I had spent many summers working at Mt. Hood, and I was familiar with the terrain and longevity of snow persistence through the summer. Additionally, the cone-like shape of the volcano presented a potentially interesting question about wind direction and aspect effects on MP concentration in snow. Lastly, the California sites were chosen based on accessibility, such as familiarity with ski touring routes and the UC Santa Barbara and the Cold Regions Research and Engineering Laboratory Energy Site (CUES), an instrument-abundant snow monitoring station, maintained by Dr. Ned Bair and the late Professor Emeritus Jeff Dozier. Everyone was so supportive in teaching me how to sample snow, and there were many conversations over coffee on how to do this the best way.

Can you explain how the integration of linear array µ-Fourier transform spectroscopy (µFT-IR) and batch spectral analysis using Open Specy software enhanced the detection and characterization of MPs?

SK: From the point of view of an environmental scientist, it made this work possible. There were millions of particles on a single filter, and we needed to be able to determine which ones are minerals, organic matter, and which ones are plastic. One sampling event can have anywhere between 2–20 filters, and these filters needed to be analyzed in a timely fashion.

Win Cowger (WC): Identifying small microplastic particles 50–500 µm is often extremely labor-intensive and is inaccurate without verification via spectroscopy. Hyperspectral imaging (HSI) allowed thousands of particles’ spectra to be measured in just 1 h and 30 min compared to weeks of time using a manual approach. Once you have the hyperspectral images, you need a way to automatically analyze them because there are hundreds of thousands of spectra per sample. There wasn’t a commercially available routine for doing this with the iN10 MX instrument, so we created one in Open Specy. For reference, Open Specy is an open-source spectroscopy software and spectral library for FT-IR and Raman spectroscopies.

What size MPs did you measure and is µFT-IR limited in the size of particles that can be analyzed?

SK: The size of the MPs that we measured ranged from 50 µm in Feret minimum diameter (the shortest length of the particle that can be drawn through the center of the particle) to more than that. The quality of the spectra can diminish as the detection size becomes smaller, but this is instrument-specific and typically does not go below 10 µm. For smaller than 10 µm, Raman spectroscopy is typically used using a point-and-shoot method. There are some mapping Raman capabilities that are developing and being implemented.

WC: The pixel size of the µFT-IR spectra we have is 25 µm. You need at least two pixels to be confident in particle identification because artificial particles can be found because of background noise, but this is often just a single pixel at a time. This is where the 50-µm threshold size quoted above comes from.

What measures did you implement to ensure minimal contamination during sample collection and analysis, and how did you evaluate the recovery rate?

SK: In the field, we wore wool, purple-dyed cotton laboratory coats (to help identify contamination in our samples) if weight wasn’t an issue, and all sampling gear had neither paint nor plastic. When samples were brought to the laboratory, the metal bottles were wiped down externally so that any surface dust didn’t contaminate the sample, and we worked in a laminar flow hood that had also been cleaned before every filtering session. When samples are brought to the Moore Institute for Plastic Pollution Research (MIPPR), they have strict contamination prevention and cleaning protocols; a positive pressure and filtered air laboratory, and similar non-plastic clothing rules in the laboratory as we had in the field. To monitor contamination, we collected field blanks, laminar flow hood blanks, and MIPPR collected blanks in their laboratory as well. This is critical because MPs are everywhere and the amounts of MPs in our samples could be affected by orders of magnitude from contamination.

The recovery rate mimicked every step of sample processing, but we doped our sample with known amounts of plastic. This helped us understand how much of the plastic in our field samples we are recovering—is there more or less than what we are measuring, or is what we are measuring an accurate representation of what is in the environment? These types of experiments are critical for every MP study to understand the efficacy of the method being used.

What trends or patterns were observed in the concentration of MPs across fresh snowfall, months-old surface snow, and snowpack stratifications?

SK: We noticed that there were statistically significant differences among the snowpack layer samples at the CUES site. The highest concentration of MPs found in a sample was a crust layer that was formed during the longest dry period in California recorded history (Jan–Feb 2022). We hypothesized that MPs accumulated in the air at that time and gravitationally deposited into the snowpack. Snowfall events create a “washout” effect, and the lowest concentrations were found in samples at the end of a prolonged snowfall event or in fresh snowfall. This finding was important because many snow studies do not specify the stratification that occurs in the snowpack, and it is important to distinguish the time period that is being sampled or being represented by a concentration. One can think of a snowpack like layers of a cake, some layers being less dense than others (sponge vs. butter cream) and having different concentrations of MPs (different amounts of sprinkles).

Another sample that had high concentrations of MPs was a remote summer snow site that ended up not being so remote–there was a significant amount of foot travel by summer alpine climbers, all of whom wear a lot of gear that is plastic-based.

Sites with low-anthropogenic activity had statistically significant different polymer content than sites that were near anthropogenic activity. This is important because it means that different polymer types of MPs have different modes of transport. Microplastic transport is determined by the shape, size, and density of the particle. Polystyrene fragments are a type of plastic that we found at remote sites at higher concentrations than sites with localized anthropogenic activity. This suggests that polystyrene could be transported long distances, along with polyacrylamide/amide and polyvinyl esters fragments. Meanwhile, some typically heavier polymer plastics seem to be not as susceptible to long range transport, like polyester/terephthalates, polyethylene, and polypropylene fragments.

Based on your findings, what can be inferred about the long-range transport of MPs in the atmosphere and their deposition in remote environments?

SK: We can infer that fragment-shaped MPs of lighter polymer types are susceptible to long-range transport, and specifically polystyrene, polyamide, and polyvinylester fragments in the 50–100 µm size range of MPs. This may vary at different sites on a global scale, as well as for MP fibers (different shape) and other size ranges. In this study, we detected more fibers at our sites with more localized anthropogenic activity than at our remote sites. At different sites around the world, there have been different patterns of pollution or anthropogenic activity. The patterns we are finding at the moment are most like MP cloud studies and remote European Alps studies.

How does your study contribute to understanding the role of snowfall in MP transport and deposition, particularly in pristine or less-studied areas?

SK: This is a tough question, because I’m not sure how the paper has been received yet and what kind of conclusions folks are pulling from the study!

I think it contributes to our understanding that within a cloud, MPs can nucleate ice to form snowflakes and that MPs below a cloud can be scavenged as well. The reality is that MPs are in our air, that the atmosphere is a growing reservoir of MP pollution, and that MPs can be scavenged from the air. Snow is a special medium that allows us to detect this phenomenon, and that if MPs are depositing in remote areas, it is happening at a larger scale in urban and suburban areas. It is also important to note that this phenomenon is cyclical and that all our environments, no matter how remote, are connected and dependent on one another. When the snow melts in the spring, these MPs will be further transported in surface runoff into the soil or watersheds. Eventually, this could be in somebody’s drinking water. If the MP is resuspended, it will then have some residence time in our atmosphere and be transported and potentially deposited elsewhere.

Ultimately, I hope this provides data for modelers and other scientists trying to build understanding of MPs in the atmosphere and their deposition into terrestrial and aquatic environments. I hope it answers questions the public might have about the plastic cycle, how pervasive MPs are, and the cyclical and wide-transport nature of the contaminant. I hope this study also brings attention to the amount of plastic that is worn as outdoor gear and how that is contributing to the number of MPs we find in “protected” areas such as national parks, wilderness areas, and other frequented outdoor attractions.

What advancements or follow-up studies do you envision for refining detection methods or further exploring the atmospheric transport of MPs?

SK: We are currently working on developing deposition rate calculations, identifying differences and similarities between sampling seasons, and modeling and extrapolating the data to develop an understanding of what point sources could be contributing to atmospheric MPs, making sure the method holds up for samples from other matrices like drinking water, streams, lakes, sediments, and live organism-related samples. We also would like to expand and try to understand which contaminants are being co-transported with MPs. It would also be fascinating to be involved in the development of measuring MPs in the sub-50-µm range.

WC: We are working to advance the Open Specy algorithm diligently. The spectral reference library often isn’t robust enough for the vast diversity of natural and synthetic materials. We continue to expand and refine the spectral reference library used in this study and now have over 40,000 reference spectra. Additionally, we developed a new way to smooth hyperspectral images to enhance the particle spectra and their size and shape measurement because across a spectral image, the roughness can create artifacts. Recently, we added the ability to automatically measure the average particle color in the visual image using an overlay of the visual and hyperspectral image. We will be submitting a manuscript on these advancements early next year.

References

1. Bergmann, M.; Collard, F.; Fabres, J.; et al. Plastic Pollution in the Arctic. Nat. Rev. Earth Environ. 2022, 3, 323–337. DOI: 10.1038/s43017-022-00279-8
2. Borrelle, S. B.; Ringma, J.; Law, K. L.; et al. Predicted Growth in Plastic Waste Exceeds Efforts to Mitigate Plastic Pollution. Science 2021, 369 (6510), 1515–1518. DOI: 10.1126/science.aba3656
3. Berard, A. Microplastics Impact Cloud Formation, Likely Affecting Weather and Climate. PSU.edu. Available at: https://www.psu.edu/news/research/story/microplastics-impact-cloud-formation-likely-affecting-weather-and-climate (accessed 2024-12-06).
4. Karapetrova, A.; Cowger, W.; Michell, A.; et al. Exploring Microplastic Distribution in Western North American Snow. J. Hazard. Mat. 2024, 480, 136126. DOI: 10.1016/j.jhazmat.2024.136126
5. Brahney, J.; Hallerud, M.; Heim, E.; Hahnenberger, M.; Sukumaran, S. Plastic Rain in Protected Areas of the United States. Science 2020, 368 (6496), 1257–1260. DOI: 10.1126/science.aaz5819