News & Blog


There is a wide range of air sensor technologies that can detect atmospheric gases and particles, but in this article we’ll focus on how low-cost sensors typically work. We’ll describe the key technology behind gas and particle detection in low-cost sensors, and place the RADICAL electronic gas sensor in context.

Gas detection

Low-cost gas sensors typically measure gases either through the interaction with light or the interaction with a material surface.

Light-based gas sensors typically work in one of two ways. In photoionisation sensors, gases are ionised by interaction with ultraviolet light, and the current generated by the resulting ions is converted into a gas concentration. Or else the sensor works through light absorption. In non-dispersive infrared sensors, the gas molecules absorb some of the infrared light within the sensor, and the amount of light absorbed is then converted into a gas concentration (or rather a mixing ratio).

In surface-based gas sensors, the detection comes from the chemical reaction between the target gas molecule and the surface of the sensor. For example, in electrochemical sensors, the target gas molecule reacts with an electrode material and generates an electrical current, which is translated into a gas concentration through data processing. In catalytic-type sensors, the target gas molecule reacts catalytically with a material, which increases the temperature of the material and thus also the resistance. The change in resistance translates into a gas concentration, again through data processing. And in metal oxide-type sensors, the electrical current depends upon the competition between ambient gases and the target gas molecules in trapping electrons, inducing a change in conductivity which is measured.

Air quality monitors. Credit: Centre for Research into Atmospheric Chemistry, University College Cork.

These principles are all scientifically sound and valid, but it is very difficult to manufacture a device that gives reliable measurements in all atmospheric conditions, and which is robust over time. The extent to which that is achieved depends on both the sensor technology and the device packaging technology, i.e. the engineering solutions for real-world scenarios.

Aerosol / particle detection

The measurement of aerosols, or particulate matter (PM), is a different challenge. Because these particles are not molecules in the air, but small physical objects, the traditional method of determining the levels of PM in the atmosphere is to capture them by pulling air through a filter. What is retained on the filter can be weighed and reported as a concentration in the air if the captured volume is also measured.

However, in a low-cost and compact device, this gravimetric analysis is not practical. Instead, low-cost PM sensors typically measure particle mass by scattering light off individual particles. A particle that passes through a beam of light, usually in the visible range, will scatter or block the light, or both. This disturbance of the light creates a pulse that can be recorded and counted. The number of such pulses counted in a fixed period of time represents the number of particles in the volume of air sampled. So by pulling a thin stream of air through a sensor device, the number of particles per litre of air can be counted.

In addition to counting particles, the way that each particle scatters light is related to the size of that particle, so it is also possible to estimate the volume of each of those particles and therefore the total particulate volume in a given volume of air. As Greek mathematician Archimedes showed in the third century BC, volume is related to mass through the quantity called density, so this method gives an estimate of particle concentration in units of micrograms per cubic metre (µg/m3) by converting the measured volume to mass. Of course, this requires some knowledge of the typical density of atmospheric particles, which is not a constant property.

A PurpleAir PM2.5 sensor deployed at a school in Edenderry as part of the LIFE EMERALD project. Credit: Rosin Byrne, CRAC Lab, UCC.

Again, the principle is well founded, but the difficulty is in avoiding measurement errors due to changing environmental conditions and making the sensors robust. In general, the lower cost of the device, the lower the accuracy of the measurements. The same type of optical technology can be implemented in a PM sensor that retails at €50 and in a regulatory-grade PM monitor that costs €30,000. But if accuracy is important (e.g. for regulatory measurements that need to stand up in a courtroom), then the latter can be worth the price.

Learn more: “Measuring Particulate Air Pollution in the Atmosphere” – video from the Centre for Research into Atmospheric Chemistry

Sensor packaging

Of course some devices on the market will target a range of pollutants, and contain several different types of sensor technologies in one unit, so it is necessary to distinguish between the packaging of a sensor, which is the measurement device as produced for deployment, and the sensor technology, which is the part where the sensing actually happens. Typically, the sensor technology is small and compact so it can be easily packaged within integrated devices.

How does the RADICAL sensor fit into this landscape of air sensors?

The RADICAL sensor is an entirely new idea. The pollutants targeted by traditional air quality measurements (e.g. PM, NO2, ozone) are generally abundant, so even in a very clean atmosphere they will be detectable with existing instrumentation, including low-cost devices made for the consumer market.

The sensors that are being developed by the RADICAL project target a completely different chemical species in the atmosphere. This species has a profound effect on both the quality of the air and the regeneration of the atmosphere, but which until now has only been detectable by very specialised and expensive research-grade lab equipment. It has never been measured in a wide network of locations simultaneously.

This species is in fact a family of species called radicals. Radicals are molecules with more electrons than they are “supposed” to have, and this makes them extremely reactive.

Atmospheric simulation chamber at University College Cork. Credit: UCC

These radical species are continuously formed by the interaction of sunlight with other atmospheric gases, but because the radicals are so reactive, they have very short lifetimes. For example, the OH radical, the most important radical in the lower atmosphere, only lasts less than a second in the atmosphere before it attacks other molecules and transforms into something else.

This makes the radical concentration in the atmosphere very, very low. But that does not mean radicals are unimportant. On the contrary, they are instrumental in the chemical processes that take place in the atmosphere. Without radicals, our atmosphere would take much longer to “clean” itself and remove pollutants.

Learn more: How atmospheric radicals transform the air by Prof John Wenger, UCC

How radicals transform the air (day-time). Credit: RADICAL Project

At its heart, the radical sensor we are building is based on a gas-surface reaction mechanism. In this sense it sounds a bit like the technologies mentioned above, but the similarity ends there.

Electrically, the detection is based on junctionless transistors, involving arrays of silicon nanowires coated with carefully-selected molecules that are reactive towards radicals. As the atmospheric radicals hit the surface and react with the molecules in the coating, the voltage changes across the nanowire transistor. This voltage change relates to the rate at which these radical interactions occur, and therefore the concentration of radicals in the atmosphere.

Chemically, the main challenge is how to detect a particular type of radical that is several thousand or even a million times less abundant than the next gas species in the atmosphere, and distinguish it from other highly-reactive radicals. This requires extraordinary selectivity in the detection, because any interference from other gases is likely to create a greater signal than the signal from the target radical. This challenge is overcome by using not one, but multiple transistors. Because the nanowires are so small, a single chip can contain many arrays of nanowires that can simultaneously measure a large number of signals. And if the molecules that coat the individual nanowires have different reactivities towards radicals and other gases, then the specificity of detection can be achieved through intelligent multivariate calibration and machine learning.

RADICAL sensor diagram, showing a functionalised nanowire transistor, tuned to capture atmospheric radicals. Credit: RADICAL project

The work in the RADICAL project represents an entirely new platform for gas sensing technology. And the interesting thing is that it can be adapted to any type of gas by changing the chemical coating on the nanowires and thereby tuning the reactivity towards different species. So radical detection is potentially only the beginning of the RADICAL adventure.

In terms of atmospheric science, by enabling real-world measurements of radicals alongside traditional air quality measurements, we can hugely advance our understanding of how the chemistry of the atmosphere works. It has never been possible to obtain this kind of information before and it will open up a great new era of atmospheric research, which can be translated into better air quality management and cleaner air for all.

About the author: Dr. Stig Hellebust is an atmospheric chemist and environmental scientist working the Centre for Research into Atmospheric Chemistry at University College Cork. His research interests have focused on source apportionment studies of atmospheric aerosols based on chemical fingerprints and automated detection of bioaerosols. He is currently involved in active research projects on linking air quality and health (INHALE), emissions from upland wildfires (FLARES), biological aerosols in the atmosphere (FONTANA) and airborne spread of SARS-CoV-2 (UPCOM) and this RADICAL project.

Follow our progress with RADICAL