In the pursuit of standardization

The field of analytical microplastic science is far from standardized and consequently some of us (inevitably) have different opinions on how 'things should be done'.

One of my previous blog posts on how to report blanks in micoplastic research, opened a discussion on whether blanks should be subtracted at all? Currently, there are no certified, standardized methods - and certainly none that encompass the diversity of all sample matrices - but, it won't be long before the ISO releases a finalized methodology standard for the analysis of microplastics in drinking water (ISO/DIS 16094-2).

Last week, during the 3rd PlasticTrace consortium meeting, we had the opportunity to present and engage in discussions with some of the leading experts in microplastics research. A key topic that emerged was the handling of blank data—a critical aspect of ensuring robust and reliable results.

There was broad agreement on the importance of procedural blanks, with a shared understanding that more blank data typically leads to better quality control. However, a divergence in perspectives arose when addressing the question: What should be done with the blank data once collected?

From our perspective, comparable procedural blank data should be subtracted from the true sample data. Although both datasets are subject to some degree of uncertainty, microplastics detected in the blanks represent contamination unrelated to the investigated environment. Subtracting this contamination data moves us closer to an accurate representation of the environmental microplastic concentrations. We also emphasize the importance of maintaining access to the original raw data to ensure transparency and traceability in the adjustment process.

One counterargument we encountered suggested that, because of inherent uncertainties, subtracting blank data from true sample data could be considered subtracting "random from random," thereby rendering the process meaningless. However, if we accept the premise that all data contains some level of randomness, one could argue that removing one source of randomness from another may reduce variability and enhance precision.

That said, it's important to acknowledge that not all authorities share this view. For instance, the ISO 16094-2 standard advises against blank subtraction. While our stance on this matter reflects our approach to enhancing analytical accuracy, we respect the differing perspectives within the scientific community and remain committed to advancing microplastics research in collaboration with experts worldwide. As an alternative to blank subtraction, laboratories are adviced to establish a reporting limit (RL) by analyzing ten blanks under the same protocol. The mean microplastic content of these ten blanks, plus '3 standard deviation', is commonly recognized as the Reporting Limit (RL) (Fig. 1).

In this artificial scenario, true samples with microplastic counts above n = 32.5 would be accepted, while samples with findings below n = 32.5, would be classified as 'below reporting limit'.

As a standardized method for detecting microplastics in drinking water is soon to be established, and to ensure we are prepared to analyze microplastics according to the latest standards, we are implementing this methodology for analyzing drinking water.

In early 2025, we will set and release the RL for our protocol to detect microplastics ranging from 1-20 µm in drinking water.

We are making every effort to remain at the forefront and generate reliable and valuable microplastic data!

Oskar Hagelskjær

Founder and CEO

Microplastic Solution