Researchers at Michigan State University (MSU) have demonstrated a biological approach to PFAS detection that uses the neural circuitry of locusts to identify and distinguish multiple forever chemicals.
The study recorded neural activity from the insects’ olfactory systems after exposure to several per- and polyfluoroalkyl substances (PFAS), including perfluorooctanesulfonic acid (PFOS).
The researchers found that different PFAS generated distinct patterns of activity in the locust brain. These neural signatures allowed the team to classify seven PFAS compounds with 87% accuracy and distinguish different concentration ranges with 84% accuracy. Environmental-level PFOS was classified with 100% accuracy in the study’s tests.
The result points to a potential new form of PFAS detection that could eventually complement conventional laboratory techniques. The Michigan State team is now investigating whether the approach can work with environmental samples and could ultimately be developed into a compact sensor for field monitoring.
Why PFAS detection remains difficult
PFAS are a large group of synthetic chemicals used in applications including non-stick products, waterproof materials and firefighting foams. Their strong chemical bonds make many PFAS highly resistant to degradation, allowing them to persist in soil, water and biological systems.
Monitoring these compounds is therefore an important part of understanding and managing contamination. Conventional PFAS detection commonly relies on analytical techniques such as liquid chromatography coupled with mass spectrometry.
These methods can provide highly detailed chemical measurements, but they require sophisticated equipment, trained personnel and laboratory infrastructure.
The challenge becomes greater when monitoring needs to move outside controlled laboratory environments or when multiple PFAS need to be screened rapidly.
Turning locust neurons into chemical sensors
The Michigan State researchers approached the problem from a different direction: instead of engineering an entirely artificial chemical sensor, they used an existing biological sensing system.
Locusts have a sophisticated olfactory network capable of converting chemical stimuli into patterns of electrical activity. The researchers exposed the insects to PFAS and recorded activity in the antennal lobe, a brain region involved in processing odour information.
The resulting neural signals were not identical. Each PFAS produced its own temporal and spatial pattern of neuronal activity, effectively creating a chemical signature that could be analysed computationally.
The study involved seven PFAS compounds and recordings from dozens of neurons. Statistical analysis showed that the population-level neural responses could separate the different chemicals, while machine-learning models were used to classify PFAS and concentration levels.
Detecting PFOS at environmental concentrations
One of the more significant tests involved PFOS, a persistent and widely detected PFAS compound that is particularly challenging for gas-sensing approaches because of its very low volatility.
The researchers tested PFOS at concentrations designed to represent environmental conditions. The lowest concentration corresponded to approximately 34 parts per trillion (PPT) in the vapour phase.
The locust neural system was able to distinguish PFOS responses from a clean-air control and differentiate between the tested concentrations. Classification of the environmental PFOS samples reached 100% in the reported experiments.
The findings are notable because PFAS are synthetic chemicals and have not been part of the evolutionary environment in which insect olfactory systems developed.
The researchers therefore do not yet have a complete explanation for why the locust neural circuitry responds so clearly to these compounds.
From laboratory experiment to field sensor
The work remains a proof of concept rather than a ready-to-use PFAS detection technology.
The experiments relied on controlled chemical exposures, neural recordings and computational analysis, so further development will be needed before the system could operate as a practical environmental monitoring device.
The researchers are now examining environmental samples to establish whether the biological sensing approach can identify PFAS outside controlled laboratory conditions.
The longer-term objective is to combine biological olfaction, neural recording and machine learning in a portable platform capable of rapidly screening for several PFAS compounds.
Such a system would not necessarily replace laboratory mass spectrometry, but could provide a faster way to identify potential contamination and determine where more detailed chemical analysis is warranted.
The study describes the work as the first biological olfaction-based platform designed specifically for broad PFAS detection.
