Three cheap hygroscopic salt solutions on carbon nanotube sensors distinguish nine toxic gases, including warfare agents, without engineered receptor molecules.
(Nanowerk Spotlight) Many toxic gases, from acid chlorides to nerve agents, react strongly with water. They dissolve in it, break apart on contact, or shift into new chemical forms. If a sensor could hold a stable film of water on its surface, these varied reactions would generate different electrical signals for different gases, potentially replacing the expensive, purpose-built receptor molecules that gas sensors currently depend on.
But the idea has an obvious obstacle: water evaporates. Under normal ambient conditions, a thin aqueous coating on a sensor disappears within minutes, taking any sensing capability with it.
This practical barrier has kept water out of electronic gas-sensing technology. Instead, researchers have pursued molecular receptors, specialized molecules that bind selectively to a target gas and trigger a measurable signal.
Hexafluoroisopropanol coatings, for instance, can pick out nerve agents, while polyethyleneimine targets nitrogen dioxide. Each receptor works well for its intended target, but each must be individually designed, synthesized, and validated. Scaling a sensor array to cover many different threats means repeating that entire process for every new compound.
Previous work hinted at partial solutions. Hygroscopic cellulose fibers helped paper-based sensors absorb gaseous analytes, but only in very humid environments and with limited ability to distinguish one gas from another. Separately, researchers showed that hygroscopic salts, substances that spontaneously absorb atmospheric moisture, could form tiny liquid lenses on surfaces and capture gas molecules from the air. But nobody had combined these moisture-trapping salts with electronic sensors to build a working detection platform.
A research team at the Ulsan National Institute of Science and Technology (UNIST) in South Korea, working with collaborators at the Agency for Defense Development, has now bridged that gap. In a study published in Advanced Functional Materials (“Receptor‐Free Identification of Toxic Gases Enabled by Hygroscopic Aqueous Salt Films”), the researchers show that films made from highly hygroscopic salts maintain a stable liquid coating on carbon nanotube sensors indefinitely under ambient conditions.
These thin aqueous layers interact differently with different gases, producing electrical fingerprints that allow a simple four-sensor array to identify nine distinct toxic compounds, including chemical warfare agents, without any molecular receptors.
CNT chemiresistors with stable hygroscopic aqueous films. (a) Schematic of coupled gas–liquid–solid sensing: analytes partition into the hygroscopic aqueous salt film and either undergo aqueous-phase transformation via hydrolysis/acid–base speciation (left) or remain largely non-transforming (right), collectively modulating CNT network resistance. (b) Device structure showing a CNT film with electrodes and an SU-8 microwell filled with a hygroscopic aqueous film. (c) Optical image of a representative CNT chemiresistor and SEM image of the CNT network. Scale bars: 200 µm (optical), 2 µm (SEM). (d) Optical images showing long-term stability of 5 M H3PO4 hygroscopic aqueous film under ambient conditions for 1 year. Scale bar: 200 µm. (e) Raman spectra confirming that the hygroscopic films remain in the liquid state over one year. (f) Optical images showing that the films of 5m solutions partially recede but do not fully dry even under very low humidity (RH ≈ 0%). Scale bar: 30 µm. (Image: Reproduced with permission from Wiley-VCH Verlag)
The approach exploits a property called deliquescence relative humidity (DRH), the humidity threshold above which a salt absorbs enough atmospheric moisture to dissolve into a liquid. The team selected three substances with exceptionally low DRH values: lithium bromide (6.3%), phosphoric acid (9.5%), and lithium chloride (11.3%). Because indoor and outdoor air almost always exceeds 15% relative humidity, 5 molal solutions of these substances remain liquid on the sensor surface without active humidity control.
Optical imaging confirmed that the films persisted for over a year, and Raman spectroscopy verified they stayed in a liquid state throughout. Even in a nitrogen-filled glove box at roughly 0% humidity, the films only partially receded rather than drying out completely. When performance eventually degrades, the coating can be refreshed by rinsing and reapplying solution.
The sensors themselves are chemiresistors, devices whose electrical resistance changes when they encounter a target molecule. Each one consists of a thin film of single-walled carbon nanotubes deposited on a silicon substrate, with gold electrodes and a tiny well made of SU-8 photoresist to hold the liquid coating in place. The complete array uses four channels: one bare nanotube sensor and three coated with different hygroscopic solutions.
The paper organizes analytes into two broad categories: those that undergo chemical transformation in water and those that primarily dissolve without reacting. To test the first category, the team exposed the array to oxalyl chloride and oxalyl bromide, two compounds that hydrolyze rapidly. When these molecules enter the salt film, they break down to produce hydrochloric or hydrobromic acid, releasing hydrogen ions and halide ions that alter the electrical properties of the underlying nanotube network.
The magnitude of the resistance change depended not just on the gas but on which salt film it encountered. Oxalyl chloride produced a larger signal on the lithium bromide sensor, while oxalyl bromide triggered a stronger response on the lithium chloride sensor. The researchers attribute this cross-selectivity to the fact that introducing a new halide ion into a film already dominated by a different halide causes greater reorganization of the ionic environment near the nanotube surface.
The platform also works for gases in the second category, those that do not react with water. Dimethyl methylphosphonate (DMMP), a widely used stand-in for the nerve agent sarin, dissolves in water but does not undergo measurable hydrolysis at room temperature. Each salt film nonetheless produced a distinct electrical response.
The lithium chloride film showed a large, partially reversible resistance increase. The lithium bromide film gave a smaller, irreversible signal. The phosphoric acid film exhibited a distinctive two-phase response: resistance first decreased, then increased. Confocal Raman depth profiling revealed that these differences stem from salt-specific effects on how DMMP distributes itself between the air, the liquid film, and the nanotube surface. In a separate experiment, one sensor configuration detected DMMP at concentrations as low as 80 ppb.
After optimizing the film thickness at 5 µm, the team tested the full array against nine gases spanning a wide chemical range: ammonia (10 ppm), sulfur dioxide (76 ppm), cyclohexane (180 ppm), acetone (4000 ppm), nitrogen dioxide (11 ppm), hydrogen cyanide (5 ppm), DMMP (800 ppm), chloropicrin (2 ppm), and cyanogen chloride (3 ppm). Chloropicrin is classified as a choking agent, while hydrogen cyanide and cyanogen chloride are blood agents.
Each gas produced a unique four-channel pattern of resistance changes. A principal component analysis of the data showed nine well-separated clusters with no overlap, meaning the array reliably distinguished every tested compound. The first two principal components captured roughly 78% of the total variance. Even chemically similar warfare agents produced clearly different signatures.
The authors acknowledge practical limitations. Because the films sit exposed to ambient air, background gases could gradually contaminate them, and the first exposure to a new analyte sometimes conditions the film in ways that slightly alter subsequent responses. Repeated exposure cycles showed that after an initial conditioning step, the sensors operated with largely reversible responses for up to 10 cycles. The coatings are cheap and easily replaced, and the researchers suggest the sensors could work in a disposable format.
The architectural simplicity of this platform sets it apart. Rather than engineering a different receptor molecule for every threat, three commodity salt solutions and the universal chemistry of water generate the selectivity. The researchers propose that combining these hygroscopic-film sensors with conventional receptor-coated devices could yield hybrid arrays that exploit both analyte-water and analyte-receptor interactions.
Paired with compact electronics and pattern-recognition algorithms, such systems could move toxic gas identification out of the laboratory and into field-portable devices for environmental monitoring, industrial safety, and chemical defense.
For authors and communications departmentsclick to open
Lay summary
Prefilled posts
ORCID information
Chang Young Lee (Ulsan National Institute of Science and Technology (UNIST))
, 0000-0002-2757-8019 corresponding author
Nanowerk Newsletter
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.
