A new approach to the spectral analysis of liquid membrane oscillators by Gábor transformation.

Liquid membrane oscillators very frequently have an irregular oscillatory behavior. Fourier transformation cannot be used for these nonstationary oscillations to establish their power spectra. This important point seems to be overlooked in the field of chemical oscillators.

On the possibility of molecular recognition of taste substances studied by Gábor analysis of oscillations.

A liquid membrane oscillator containing nitromethane as membrane material has been investigated. The influence of substances responsible for taste belonging to four classes (sweetness, saltiness, sourness and bitterness) on oscillation patterns of liquid membrane oscillators with cationic surfactant benzyldimethyltetradecylammonium chloride (BDMTAC) was examined. A new approach based on Gábor transformation is proposed for obtaining the power spectra of the observed oscillating signals and for establishing "fingerprints" of the investigated substances.