Selective Recognition of Acetone in Air Against Hydrogen By Impedance Measurement of Two-Terminal Electrochemical Sensors Based on Ionic Liquids

Abstract

Introduction Ionic Liquids (ILs) are molten salts at room temperature and have both solvent and electrolyte properties. ILs have been actively studied for use in electrochemical gas sensors, thanks to their gas absorption property, electrochemical stability, and low evaporation loss. For example, gas sensors based on the changes in electrical double layer capacitance or in bulk impedance induced by gas dissolution in ionic liquids have been reported [1, 2]. However, there are few reports on the selective recognition of volatile organic compounds such as metabolites in human exhaled air against interfering gases. In this study, we demonstrated selective recognitions of acetone, which is a typical low-molecular-weight metabolic gas in human exhaled air, against hydrogen, which is a metabolite of enterobacteria, by electrochemical impedance measurement using ionic liquids. Method The interdigitated electrodes made of Cr/Pt shown in Fig. 1 were fabricated on a glass substrate. The 3-nm-thick chromium film acting as the adhesion layer was deposited by electron beam evaporation technique, which was followed by the pattern definition and evaporation of 100-nm-thick platinum electrode acting as the main electrical conduction layer. The droplet of 3-μL 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide, a kind of ionic liquids, was casted onto the electrodes, which completed the fabrication of a two-terminal electrochemical sensor. The impedance measurements were performed under various atmospheres. The following shows the sequence of atmospheres under which the impedance responses of the fabricated sensors were measured. ⅰ) Dry air (3 min) ⅱ) (case Ⅰ) 25-ppm H2 in dry air (3 min)ⅱ) (case Ⅱ) 25-ppm CH3COCH3 in dry air (3 min) ⅲ) dry air (3 min)In the second process ii), case I or case II was chosen from two target gases; hydrogen and acetone. The gas flow rate was kept constant at 200 sccm. Impedance measurements were performed using an LCR meter (Keysight E4980A) with DC bias of 0.0 V and AC amplitude of 10 mV. The measurement frequency was varied from 300 kHz to 20 Hz. Results Fig. 2 shows the measured sensor response at (a) 200 kHz, (b) 30 Hz, and (c) 350 Hz. The sensor response is defined as impedance change ΔZ re due to target molecules normalized by the impedance in dry air Z re(Air), where Z re represents the real part of the impedance. The equation used to extract the sensor response is shown in Fig. 2. We noticed that the sensor response was strongly dependent on the measured frequency. Z re decreases at 200 kHz in both the CH3COCH3 and H2 atmospheres (Fig. 2(a)), whereas it is increased by the target molecules at 30 Hz (Fig. 2(b)). The directions of the impedance changes were the same for both the target molecules. On the other hand, at 350 Hz (Fig. 2(c)), Z re decreases in the CH3COCH3 atmosphere, although it stays constant in the H2 atmosphere. This clearly demonstrates that CH3COCH3 can be selectively detected against H2 by impedance measurement with an appropriate measurement frequency. Discussion and Conclusion We consider that the equivalent circuit of the present system includes an element related with the bulk IL resistance and one related with the electrical double layer at the electrode/IL interfaces. The measurement of the impedance phase angle reveals that the phase angle is close to 0 degrees at 200 kHz and close to -90 degrees at 30 Hz. Therefore, we consider that high-frequency (200 kHz) Z re is affected by the bulk IL resistance change induced by dissolved gaseous molecules and that low-frequency (30 Hz) Z re is influenced by the change in electrical double layer characteristics. On the other hand, at medium frequencies; namely 350 Hz in this study, the gas-induced effects both on bulk IL resistance and on the electric double layer characteristics causing Z re modulation may be compensated with each other, and no sensor response was observed under the H2 atmosphere as shown in Fig. 2(c). We will extract the equivalent circuit model for the system used in this experiment and examine how the parameters of circuit elements change in the H2 and CH3COCH3 atmospheres in comparison with the standard dry air atmosphere. By investigating the frequency dependence of the impact of each parameter change on Z re, we will clarify the mechanism of Z re constancy in the H2 atmosphere at medium frequencies.