Introduction
In odor sensing based upon QCM[1] sensing film is crucial for both sensor sensitivity and selectivity. The viscoelastic property of the sensing film greatly influences the response of the QCM sensor. The shear wave generated from QCM cannot penetrate deeply into the viscous sensing film. However, many appropriate sensing films with interesting properties are viscous. We would like to understand the behavior of the viscous sensing film although the behavior of the rigid sensing film is well understood.
Method
A QCM has an equivalent circuit derived from Mason equivalent circuit[2] as shown in fig. 1. Equations (1) and (2) indicate the acoustic impedance (Z3e) due to acoustic loading as a function of film thickness (h) film density (ρA1) and complex wave velocity (νA1). To extract QCM equivalent circuit component, we used a network analyzer to measure conductance from a QCM. Then, we apply the curve fitting technique to extract R1, L1, and C1 from the conductance curve[3]. This technique was used instead of a conventional motional admittance method (measure the frequency of maximum conductance). Because all the data were used for fitting the curve, this method is more robust and less susceptible to noise than conventional motional admittance method.
The dip-coating technique was used for coating viscous film on QCM (9MHz, AT-cut). The QCMs were immersed into glycerol solution and then pulled-up with various speeds, which results in different film thicknesses.
Results and Conclusions
To replicate odor deposition into sensing film. We were using a small-size dispenser to deposit droplets of liquid onto QCM. Water droplet was shot onto the glycerol film as on a QCM. The preliminary result is shown in fig.2. The frequency was increased after each vapor exposure, whereas the resistance was decreased, we sometimes encountered increase in frequency due to vapor exposure at the highly viscous environment. We tried to understand its mechanism using the following analysis.
The study was firstly done with liquid loading which represents viscous film. Only a singles side of QCM should be in contact with liquid. We calculated resonance frequency shift and resistance change as a function of water film thickness as is shown in fig. 3. Our calculation result based upon Mason equivalent circuit roughly agreed with both Kanazawa’s equation[4] and the experiment when the film thickness became large.
To estimate film thickness and dynamic viscosity of the film, Mason equivalent circuit was used. The optimization was performed to obtain the dynamic viscosity and film thickness using the experimental data obtained from various film thicknesses. Fig. 4 shows the experimental data points at estimated film thickness and calculated curve. The simulation of Mason circuit shows good agreement with experimental data. Moreover, the simulation result also suggests that the frequency increase occurs due to liquid loading at a certain condition.
We will perform the experiment on depositing tiny liquid droplets into sensing film before the conference.
References:
[1] K. Nakamura, T. Nakamoto, and T. Moriizumi, Classification and evaluation of sensing films for QCM odor sensors by steady-state sensor response measurement, Sens. Actuators B, vol. 69, no. 3, pp. 295–301 (2000); doi: 10.1016/S0925-4005(00)00510-4
[2] T. Nakamoto, and T. Moriizumi, A Theory of a Quartz Crystal Microbalance Based upon a Mason Equivalent Circuit, Japanese Journal of Applied Physics, vol. 29, no. 5, pp. 963-969 (1990); doi: 10.1143/JJAP.29.963
[3] S. N. songkhla and T. Nakamoto, Signal Processing of Vector Network Analyzer Measurement for Quartz Crystal Microbalance with Viscous Damping, IEEE Sensors Journal, vol. 19, no. 22, pp. 10386 - 10392 (2019); doi: 10.1109/JSEN.2019.2930733
[4] K. K. Kanazawa and J. G. Gordon, Frequency of a quartz microbalance in contact with liquid, Anal. Chem, vol. 57, no. 8, pp.1770–1771 (1985); doi: 10.1021/ac00285a062
Figure 1
