Composants Pour Mesurer le Taux d’Alcool dans l’Haleine

Trouver un détecteur approprié pour la mesure d’un gaz spécifique semble à priori aisé. Afin de mesurer le taux d’alcool dans l’haleine (TAH), il faut analyser les raies d’absorption de l’éthanol. Elles sont situées autour de 9,5 µm, 8,1 µm, 7.2 µm et 3,4 µm. En théorie, il suffirait de mesurer le signal du détecteur sur une seule raie d’absorption ; toutefois, il est important de s’assurer qu’il n’y a de sensibilité croisée avec d’autres substances présentes dans l’haleine (ex. eau, méthane et éthanol).

Finding a suitable detector for a specific gas measurement still seems theoretically trivial. In order to measure the breath alcohol content (BrAC), the absorption lines of ethanol gas have to be analyzed. They are approximately 9.5 µm, 8.1 µm, 7.2 µm, and 3.4 µm. Theoretically, it would be sufficient to measure the detector signal at a single absorption line; however, it is important to make sure that a cross-sensitivity to other substances that are also in the breath (water, methane, and ethanol) does not exist.

Filter Selection
IR detectors are polychromatic. A suitable filter is required to detect a specific gas or measure a specific wavelength: Bandpass filters only allow the beams of a specific wavelength range to pass.
Thus, the success of the measurement always depends on the quality of the filter: the spectral width must line up precisely with the absorption line being measured, and the transmission of the filter should be as high as possible.
For BrAC measurements, analysis of the wavelengths 3.4 µm and 9.5 µm has become the standard.

Detector Selection
The shorter wavelength could be identified using different detector types: According to the overview, InAs, PbSe, cooled PbS and PbSe, and pyroelectric detectors are all suited for use.
Exclusive observation at 3.4 µm is critical, however, because the absorption lines of water vapor and methane gas are approximately equal.

The air in the breath contains moisture. In a worst-case scenario, it would not be possible to strictly separate ethanol and water vapor at 3.4 µm. As a greenhouse gas, methane can also be detected in the air; thus, the problem is the same.
Measurement at the wavelength of 9.5 µm is more significant, thus the selection of pyroelectric detectors is more likely.

The pyros can house up to four independent detector chips with an integrated bandpass filter in a single housing. One channel often serves as a reference, and the others are there for gas detection. The simultaneous measurement of a gas using two filters would be possible and is actually carried out as such in practice when a high dynamic range is required for measurement.