Efficient and Automated Generation of Orthogonal Atmospheres for the Characterization of Low-Cost Gas Sensor Systems in Air Quality Monitoring

Abstract

This article presents a compact continuous-flow automaton for the metrological characterization of an array of low-cost gas sensor systems (up to 17 devices) used in air quality monitoring. The automaton readily generates homogeneous gas mixtures of carbon monoxide (CO), nitrogen monoxide (NO) and dioxide (NO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ), and ozone (O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> ) in the parts-per-billion range (ppb, nmol mol <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$^{-1}$ </tex-math></inline-formula> ) that are stable, that is, dispersion of 1 ppb (NO, NO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> , O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> ) and 9 ppb (CO), with response times of 2 min (CO, NO, O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> ) and 30 min (NO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ). The resulting mixtures, which are traceable to the International System of Units (SI) due to the usage of calibrated high-grade reference instruments, can be humidified [0%–60% relative humidity (RH)] and, in addition, the devices under test (DUTs) can be heated (to temperatures between 5 °C and 30 °C) to systematically simulate different atmospheric environments. The application of fractional factorial designs makes the protocol efficient and leads to orthogonal variables. With the presented installation, ten low-cost gas sensor systems are calibrated and an uncertainty estimation is performed. The average values of the relative standard uncertainties across all DUTs are estimated as 52% (CO), 61% (NO), 22% (NO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ), and 21% (O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> ).