Why PFAS remediation needs a new ally: real-time, low-cost sensors
- Tommaso A. Dragani
- 20 hours ago
- 7 min read
Across Europe and worldwide, utilities and industries are racing to remove PFAS from
drinking water. Yet we still have a blind spot: in most cases, we don’t see PFAS in real time.
We send samples to the lab, wait days for LC-MS/MS results, then adjust treatment. For
chemicals that move through aquifers and treatment trains every hour, this time lag is huge.
The recent review by Zha et al. in Environmental Research offers a timely and
comprehensive look at PFAS sensors, focusing on how different molecular probes –
antibodies, aptamers, synthetic small molecules and polymers – actually “recognise” PFAS
at the molecular level [Zha et al. 2025]. It’s an impressive scientific landscape. But it also
highlights a simple message: we urgently need PFAS sensors that are not only sensitive
and clever, but also robust, affordable, and deployable at scale.
Tightening standards, rising pressure on monitoring
Regulators are moving fast, and they are pushing detection limits down into the low-ng/L
range:
The recast EU Drinking Water Directive (2020/2184) requires Member States to
ensure by January 2026 that the sum of 20 “priority PFAS” does not exceed 100 ng/L,
and that “PFAS total” stays below 500 ng/L in drinking water [EC 2024].
In the United States, EPA has set enforceable Maximum Contaminant Levels as low
as 4 ng/L for PFOA and PFOS, with additional rules for other PFAS and mixtures [EPA
2025].
At the same time, the International Agency for Research on Cancer (IARC) now classifies
PFOA as carcinogenic to humans (Group 1) and PFOS as possibly carcinogenic (Group
2B), based on epidemiological, animal and mechanistic evidence [Zahm et al. 2024].
This tightening of standards, paired with the growing recognition of PFAS hazards, is
forcing water suppliers to rethink not only how they remove these compounds, but how
they track them. Treatment can no longer rely on occasional snapshots; it needs a
continuous pulse. And yet, today, routine PFAS monitoring still depends on LC-MS/MS,
extraordinary instruments, but costly, centralized and slow to deliver answers [Hu et al.,
2023].
For utilities, remediation plants and emergency responders, that delay is no longer
tolerable. They need to see PFAS as they move, not days after the fact.
What the new review tells us about PFAS sensors
The review by Zha and co-authors takes a different angle from previous sensor overviews:
instead of classifying sensors by signal type (fluorescence, colorimetric, electrochemical,
etc.), it organizes them by molecular probe – the part of the sensor that actually “grabs”
PFAS molecules [Zha et al. 2025].
They analyse four main probe families:
1. Antibodies
Extremely specific binding to PFAS (especially PFOA and PFOS), using
hydrophobic pockets and complementary electrostatic interactions in the
antibody’s variable region.
Enable ELISA assays and immunosensors with good selectivity even in
complex matrices.
Challenges: expensive and slow to produce, sensitive to temperature and
storage conditions, and often better suited to lab-based platforms than rugged
field devices.
2. Aptamers (DNA/RNA)
Single-stranded nucleic acids selected by SELEX that fold into 3D structures
forming hydrophobic pockets around fluorocarbon chains.
Can reach very low detection limits: recent electrochemiluminescence
platforms report sub-ng/L PFOA in river water with good agreement to LC-MS.
Challenges: laborious screening; sequences are still few; performance can be
strongly affected by ionic strength, pH and nucleases in real waters.
3. Synthetic micromolecules
Calixarenes, cyclodextrins and other host molecules engineered to interact
with PFAS via hydrophobic, electrostatic and fluorophilic (F···F) interactions.
Often cheap and fast, with good potential for simple fluorescence or
colorimetric tests [Chen et al. 2021].
Challenges: selectivity is limited; many devices are best suited for rapid
screening, not regulatory-grade quantification.
4. Polymers and molecularly imprinted polymers (MIPs)
Polymers whose cavities are “imprinted” with PFAS templates, giving a lock-
and-key fit combined with fluorophilic and hydrogen-bonding interactions
[Ahmadi Tabar et al. 2023].
Some MIP-based electrochemical sensors now achieve detection limits at or
below 1 ng/L for PFOA and PFOS, comparable to regulatory needs [Hafeez et
al. 2024].
Challenges: interference from PFAS homologues, reproducibility of imprints,
template removal, and performance in real, complex matrices like
wastewater or mixed industrial effluents.
The message is encouraging: the toolbox of PFAS-selective probes is getting richer, and
some technologies already reach extremely low detection limits in controlled conditions.
But for utilities and remediation operators, performance on spiked lab samples is only the
starting point.
A reality check: why sensors are still not “plug & play” for utilities
Put simply, most PFAS sensors are still stuck in the transition from beautiful prototype to
reliable field tool. Several recurring limitations emerge across the literature:
Matrix effects
Natural organic matter, co-contaminants, variable ionic strength and pH often
reduce sensitivity or distort signals. A sensor that performs at 1 ng/L in buffered
water may perform far worse in groundwater or treated effluent.
Mixtures rather than single targets
Many sensor formats are optimized for individual PFAS (often PFOA or PFOS). In
reality, utilities face complex mixtures of long- and short-chain PFAS,
transformation products and unknowns. Discrimination is difficult, and cross-
reactivity can both under- and over-estimate risk.
Calibration and standardisation
Regulatory acceptance will require robust inter-laboratory validation and
calibration protocols. For now, LC-MS/MS remains the reference method against
which sensors must be benchmarked.
Cost and complexity of deployment
Some of the most sensitive sensors rely on sophisticated nanomaterials or
instrumentation that is still closer to a research lab than to a water treatment plant
control room. Others, like surface plasmon resonance fiber-optic biosensors, have
promising detection limits (e.g. ≈210 ng/L for PFOA/PFOS in seawater) but require
careful optical setups [Cennamo et al. 2018].
In summary, we already know how to design molecular systems that “feel” PFAS with great
sensitivity. The real challenge is to turn these systems into simple, inexpensive, rugged
devices that work day after day in harsh, real-world conditions.
What we actually need: from excellent sensors to useful sensors
From the perspective of a PFAS remediation company like ASPIDIA, a “good” PFAS sensor is
not only about the lowest possible LOD. It is about enabling better decisions, faster, in three
key contexts:
1. Monitoring raw and treated water
Continuous or high-frequency monitoring at the inlet and outlet of treatment plants.
Ability to detect excursions above regulatory thresholds and send alarms in real time.
2. Controlling remediation processes
Inline monitoring inside treatment trains (e.g. before/after advanced oxidation, after adsorption columns, during bioreactors).
Data streams that allow dynamic optimisation of process conditions (flow rates, oxidant dose, regeneration cycles) and energy use.
3. Large-scale surveillance of aquifers and distribution networks
Deployable sensor nodes (possibly low-cost, semi-quantitative) for dense spatial coverage.
Integration with IoT and digital twins to understand PFAS migration, plume dynamics and the impact of mitigation strategies [Chen et al. 2021].
For these use cases, the “ideal” sensor would:
reliably detect PFAS at or below relevant regulatory limits (EU 100/500 ng/L; US 4
ng/L for key compounds); [EEA 2024]
work in complex waters (groundwater, surface water, industrial effluents) with
minimal sample preparation;
operate automatically for long periods with simple maintenance;
be affordable enough to deploy in numbers, not as a single delicate instrument in a
central lab.
How this connects to ASPIDIA’s mission
ASPIDIA is focused on removing PFAS from water, combining:
a modular cavitation–oxidation–adsorption platform (TriClean), and
AI-guided enzymatic bioremediation using engineered dehalogenase (DEHA) enzymes.
But advanced treatment without advanced monitoring risks becoming a black box: PFAS go
in, something happens, and we learn what really worked only days later, when LC-MS/MS
results arrive.
For us, the review by Zha et al. 2025 is not just an academic exercise. It provides a map of
the sensing landscape and helps to identify which directions are the most promising for
integration with real-world remediation:
Polymer and MIP-based sensors that can be coated onto electrodes or membranes
and integrated into process lines [Ahmadi Tabar et al. 2023].
Simple micromolecule or polymer films that could serve as low-cost warning layers
in distributed sensor networks [Zhou et al. 2024].
Data-rich platforms (electrochemical, optical, nanopore-based) that could be
combined with AI analytics for pattern recognition and mixture fingerprinting rather than single-compound quantification [Wei et al. 2024].
Ultimately, effective PFAS remediation is a cycle: we must detect contamination, treat it,
confirm the result, and then refine the process. Today, we are capable of treating and
verifying. What is still missing, at the scale the problem demands, is the ability to detect and
adjust in real time, while the water is moving.
A shared challenge for the PFAS community
The conclusion from the review, and from the broader literature, is both sobering and
exciting:
We can build sensors that “see” PFAS at astonishingly low levels.
But very few of them, so far, are ready to become the everyday workhorses of
utilities, remediation plants, and regulators.
For ASPIDIA, this is not a reason to wait. It is a reason to collaborate: with sensor
developers, academic groups, utilities, and technology providers who share the same vision.
Because if we want PFAS concentrations in water to go down and stay down, we first need
to see them clearly, cheaply, and continuously.
And that means turning the impressive science of PFAS sensing into something very
concrete: a small, robust device that can sit on a pipe or in a well and quietly answer, every
few minutes: “How much PFAS is there right now?”
Until that answer is easy, fast, and affordable, the work of remediation will always be one
step behind.
References
Ahmadi Tabar F, Lowdon JW, Bakhshi Sichani S, et al.
An Overview on Recent Advances in Biomimetic Sensors for the Detection of Perfluoroalkyl Substances. Sensors (Basel). 2023;24(1):130. doi:10.3390/s24010130
Cennamo N, Zeni L, Tortora P, et al.
A High Sensitivity Biosensor to detect the presence of perfluorinated compounds in environment. Talanta. 2018;178:955-961. doi:10.1016/j.talanta.2017.10.034
Chen B, Yang Z, Qu X, Zheng S, Yin D, Fu H.
Screening and Discrimination of Perfluoroalkyl Substances in Aqueous Solution Using a Luminescent Metal-Organic Framework Sensor Array. ACS Appl Mater Interfaces. 2021;13(40):47706-47716. doi:10.1021/acsami.1c15528
EC 2024. Commission Notice
Technical guidelines regarding methods of analysis for monitoring of per- and polyfluoroalkyl substances (PFAS) in water intended for human consumption. 2024. C/2024/5414. OJ C, C/2024/4910, 7.8.2024. http://data.europa.eu/eli/C/2024/4910/oj
EEA 2024. European Environment Agency.
PFAS pollution in European waters. Briefing no. 19/2024. doi: 10.2800/9324640.
EPA 2025.
Per- and Polyfluoroalkyl Substances (PFAS). Final PFAS National Primary Drinking Water Regulation. 2025. https://www.epa.gov/sdwa/and-polyfluoroalkyl-substances-pfas
Hafeez S, Khanam A, Cao H, Chaplin BP, Xu W.
Novel Conductive and Redox-Active Molecularly Imprinted Polymer for Direct Quantification of Perfluorooctanoic Acid. Environ Sci Technol Lett. 2024;11(8):871-877. doi:10.1021/acs.estlett.4c00557
Hu H, Liu M, Shen L, Zhang L, Zhu H, Wu Q.
Simultaneous determination of multiple perfluoroalkyl and polyfluoroalkyl substances in aquatic products by ultra-performance liquid chromatography-tandem mass spectrometry with automated solid-phase extraction. J Chromatogr B Analyt Technol Biomed Life Sci. 2023;1224:123736. doi:10.1016/j.jchromb.2023.123736
Wei X, Choudhary A, Wang LY, et al.
Single-molecule profiling of per- and polyfluoroalkyl substances by cyclodextrin mediated host-guest interactions within a biological nanopore. Sci Adv. 2024;10(45):eadp8134. doi:10.1126/sciadv.adp8134
Zahm S, Bonde JP, Chiu WA, et al.
Carcinogenicity of perfluorooctanoic acid and perfluorooctanesulfonic acid. Lancet Oncol. 2024;25(1):16-17. doi:10.1016/S1470-2045(23)00622-8
Zha J, Ma M, Shen Y, et al.
A critical review of sensors for detecting per- and polyfluoroalkyl substances: Focusing on diverse molecular probes. Environ Res. 2025;278:121669. doi:10.1016/j.envres.2025.121669
Zhou M, Wei D, Li J, Fan K, Wang W.
Low-cost and highly sensitive conductive polymer-based lateral flow assay for per- and polyfluoroalkyl substances detection. Innovation (Camb). 2024;6(3):100768. doi:10.1016/j.xinn.2024.100768