Mid-infrared acetone gas sensors using substrate-integrated hollow waveguides augmented by advanced preconcentrators.

The identification of volatile organic compounds (VOC) such as acetone, a relevant biomarker for diabetes mellitus in exhaled human breath has become essential for early disease diagnostics, prognosis, and monitoring of metabolic responses to pharmacological interventions. Gas chromatography coupled with mass spectrometry (GC-MS) is considered as the gold standard for breath analysis. However, its inability to offer point-of-care monitoring limits its applicability in clinical environments. Optical techniques based on absorption in the mid-infrared range (2.5 to 25 μm) appear as promising alternatives due to their inherent selectivity, potential for miniaturization, portability, and direct analysis with rapid response. Relevant biomarkers present in exhaled breath, are usually observed in the ppb to ppm (v/v) concentration regime. For sensitivity enhancement of optical sensing techniques, appropriate preconcentration schemes are required prior to the optical detection. The present study describes a method combining a multi-channel substrate-integrated preconcentration module (a.k.a., muciPRECON) for enhancing and optimizing the detection of acetone via a mid-infrared photonic sensing system. The sensor system comprises a compact Fourier Transform Infrared (FTIR) spectrometer, a technique that enables detailed infrared spectral analysis, coupled to a substrate-integrated hollow waveguide (iHWG) simultaneously acting as a gas cell and as a photon conduit. Preconcentation experiments were from acetone/nitrogen gas mixtures in the concentration range of 5-100 ppm at - 10 °C followed by desorption at a temperature of 100 °C and direct injection into the IR sensing system Thus obtained acetone spectra were quantified evaluating a molecule-specific vibrational absorption peak area in the spectral window 1260-1170 cm. After extensive screening, Tenax was identified as superior sorbent material providing an enrichment factor of up to 153-times, a limit of detection (LOD) of 0.118 ppm, and a limit of quantification (LOQ) of 0.393 ppm. These results are indeed promising for practical applications, especially since acetone concentrations usually vary between 0.3 and 0.9 ppmv within the exhaled breath of healthy individuals, and in individuals with diabetes, acetone concentrations are typically around 1.7 to 3.0 ppmv. Consequently, the developed systems have the necessary sensitivity and accuracy to detect acetone levels that are in the relevant physiological range indicating their potential use in future real-world scenarios.