Dual domain acoustic olfactory discriminator

Acoustic transduction combining a quarter wavelength resonator, a speaker and a microphone is used for the detection and identification of gas properties such as n-hexane, acetone and ethanol. As a target gas flows into the resonator, the density and the speed of sound of the gas in the resonator change, causing a shift in the acoustic pressure waves manifesting from the speaker. Resonance frequency curves of each gas were experimentally obtained using a standard 1/f equal octave pink noise test over the audible frequency range. The speed of sound of each gas was analytically determined from the obtained resonance frequency. As the flow concentration of a target gas increases, the speed of sound decreases as the gas density increases. Time series signals at a fixed frequency exhibit unique profiles for each gas and concentration. The theoretical limit of detection for n-hexane in the time domain was calculated to be in the order of several tens of ppm. Whilst the frequency domain data obtains a direct physical parameter, time domain data enables multi-dimensional data analysis, relaying decisive data for artificial olfaction. Principal component analysis (PCA) reveals unique attributes and discrimination per gas species and concentration based on multi-dimensional data obtained through the dual domain measurements. This study demonstrates that the device can adequately identify gases at concentrations of at least several thousand ppm. This approach may provide a new platform as a mobile gas discriminator coupled to artificial olfaction with the added benefit of audio capabilities and signal processing techniques.