Microplastic recovery experiments: Beads vs. fragments - What difference does it make?

The positive quality control aspect of microplastic research, i.e. recovery control is often overlooked and is generally considered to be less important than negative quality control, i.e. blank control.

Recovery control is determined by putting a known number of microplastics into a sample, performing your protocol, and determining how many of those microplastics made it from the matrix onto the final analysis substrate. The consequent loss is then corrected for (Fig. 1).

Fig. 1 - Typical outlay of a recovery experiment: From spiking with microplastic beads, to full protocol execution, to the final recovery estimation. Figure reproduced from Deimante-Deimantovica et al. (2022).

Commonly, microplastic recovery rates have been established by spiking with microplastic beads in a narrow size range. The resultsing recovery rate would then be applied to all recovered microplastics in the sample, regardless of their size.

While this method has been widely adopted in the community, it has alsos been criticized because it does not discriminate between large and small microplastics, which have been speculated to demonstrate wastely diffrerent recovery rates -theorizing that small microplastics are more mobile and thus harder to recover than their larger counterparts.

During 2022 to 2023, I staged an experiment to test whether this hypothesis was true, and found that small microplastic fragments from to 5 µm, were not harder to recover than equivalents up to 500 µm in diameter. However, I did find that with increasing filter saturation, the analytical recovery of smaller microplastics would actually appear to decrease.

We figured that this observed loss was not linked to whether or not the particles were acutally present on the filter, but that agglomeration of the particles hindered their identification, causing bias towards larger particles (Fig. 2). However, as long as the filter was not saturated above 5% area coverage, there was no noticable difference in the recovery of larger or smaller microplastics.

Fig. 2 - When more particles are present on a substrate, sample agglomeration is more likely to occur, leading to an observed, relative decrease in the analytical recovery of small microplastics due to particle agglomeration. Figure reproduced from Hagelskjær et al. (2023).

This data then supported the validity of using microplastics beads in a narrow size range, given that there was no difference in the microplastic recovery.

However, for the sake of this experiment, the protocol was very simple; the particles were vacuum filtered onto a filter surface and flushed with water into a vial, and repeated 5 times (Fig. 3).

Fig. 3 - illustration of simple recovery experiment. Figure reproduced from Hagelskjær et al. (2023).

For true environmental samples, much more intricate sample pretreatment steps are incorporated into the protocol. These include: digestive treatment under heating, magnetic stirring, 'bubbling', ultrasonication, etc.

Environmental microplastics are likely fragilized from exposure to weathering effects, in particular from photodegradation. Therefore, it is not hard to imagine that sample pretreatment may lead to particle breakup of the indigenous microplastics - an effect that cannot be oberserved using virgin plastic microbeads.

In my latest study I used EasyMP™ (fragments 10-100 µm, Polyethylene, Red) for staging the recovery experiment of my protocol, which invloved oxidant/acid digeston as well as ultrasonication. Here, we determined a positive recovery rate (up to almost 150%) for microplastics < 15 µm in area-equivalent diameter (Fig. 4).

Fig. 4 - Microplastic recovery rate (n = 4) within size groups, demonstrating positive recovery for microplastics < 15 µm. This observation has been interpreted as the results of particle breakup during sample pretreatment. Figure reproduced from Hagelskjær et al. (2024).

Our theory of particle breakup was supported by another independent piece of evidence; the presence of of PVDF microplastic particles - the polymer type from which the filter membranes used for particle transfer, were made from. These filters were exposed to HCL and H2O2 in combination with ultrasonication for only a few minutes, which was enough to cause fragmentation of this otherwise inert polymer type. When we replaced the PVDF membranes with nitrocellulose equivalents, and analyzed the sample which demonstrated the highest number of PVDF particles (> n = 1000 in the subsample), we didn't find a single PVDF particle in the sample! We thus, had two independent indicators, that the particles were fragmentating due to sample manipulation, ultimately leading to an overestimation of the smaller microplastic fraction by up to 150%.

Particle breakup from sample treatment, is a subject that has been largely ignored although reported as early as 2017. Conversely, it's an effect that is highly important to determine and correct for, as small microplastics < 10 µm are considered to have the highest implications for human health. We do therefore, not wish to overestimate the fraction of these in true environmental samples.

Using miroplastics beads for recovery experiments may thus be misleading, unless applied in the context of benevolent sample manipulation.

I believe it's time to advance our microplastic recovery experiments to the next level. If you are studying fragments in environmetal samples, use fragments. On the other hand, if you are studying fibers, use fibers! I haven't investigated this myself, but it's not hard to imagine that a long fiber may break into several shorter pieces, during sample manipulation - and some polymer types may be more resitant than others.

If you wish to elevate the relevance of your microplastic research, we produce and sell custom microplastic reference materials, including fragments and fibers. Do not hesitate to reach out to me personally, for questions, inquiries or comments.

Oskar Hagelskjær, Ph.D.

Founder and CEO, Microplastic Solution

oskar@microplasticsolution.com