[0028] The internal size of the FR4 container used in the present invention is 7.5×7.5mm 2 To cover the sensor detection area. The height of the container is 1.6 mm, which provides enough space for adjusting the volume of the liquid to be measured. The width and length of the container (i.e. c1 and c2) are respectively larger than the width of the detection area sensor (4.1×4.6mm 2 ) Large 3.4 and 2.9mm to ensure that the liquid to be tested is exposed to the electric field on the side of the container.
[0029] The parameter study of the influence of the volume of the liquid to be measured on the frequency and decibels of the reflection notch is carried out. To evaluate the different volumes of the liquid to be measured in the sensor, Figure 5 The measurement result is displayed. Evaluate the different volumes of S11 for 0.02, 0.04, 0.06, 0.08 and 0.1 mL deionized (DI) water placed on the detection area. Deionized water was chosen because of its high dielectric constant, which is suitable for limiting testing. Picture 9 It is shown that by increasing the volume of the sample liquid from 0.02 to 0.1 mL, the resonance frequency decreases and the S11 value increases. In the initial region (0.02mL to 0.08mL), the resonance frequency changes with the volume of the liquid to be measured. In the subsequent region (0.08mL to 0.1mL), the resonance frequency is constant. With the further increase of the liquid volume, the change of S11 can be ignored. The electric radiation field is concentrated on the copper surface of the sensor. Cover the liquid sample and further increase the volume of the liquid to be tested (> 0.08mL) will not affect electromagnetic energy, and the effective dielectric constant will not change. Therefore, for volume sensing applications, the operation must be performed in the first region, and for liquid characterization, the second region should be used. Once a certain volume is reached, the resonance frequency becomes constant. The saturation point is defined as the minimum volume required for liquid sensing applications. Use a micropipette to obtain the correct volume of the liquid to be tested (> 0.08mL), and test different volumes of the liquid to be tested, indicating that the sensor can successfully detect liquids with a dielectric constant of 1 to 80. The sensor exhibits the same behavior for all liquid volumes to be measured.
[0030] Four liquid measurement results:
[0031] The dielectric constant of liquid composites is measured in the frequency range of 100-375MHz. Keep the temperature at room temperature (25°C). The test liquid used in the experiment is deionized (DI) water (ε'=80, ε"=9.2), methanol (ε'=35, ε"=5.13), ethanol (ε'=22, ε"= 6.6), and butanol (ε'=17, ε'=7.21). Put these sample materials (> 0.08mL) was loaded onto the detection area and characterized to verify the proposed sensor design. Image 6 The results of the calculated parameter responses of different liquids with high dielectric constant are shown. The results revealed clear differences between the different liquids. The dielectric properties of the liquid to be tested are affected by temperature. Regarding the repeatability of the measurement and the stability of the liquid material to be tested, the liquid sample to be tested should be stored in a thermostat (25°C) closet. During the measurement, the laboratory environment should be maintained at a steady state of 25°C and the air-water heat exchanger equipment should be heated. The measurement should be carried out in a short time to avoid changes due to temperature.
[0032] The air is evaluated to calibrate the measurement value of the empty detection area, and the corresponding liquids of the other four materials are filled in the area. The results show that for an empty, air-filled detection area, the measured resonance frequency of the OCSRR sensor is 330MHz, and the return loss is 62dB. Use a micropipette to inject and control the liquid to be tested. The change of S parameter is recorded and transmitted to the computer within 30 seconds. For methanol and deionized water samples, the resonance frequencies are 231 and 204 MHz, and the return loss is 37.41 and 40.53 dB, respectively. For ethanol and butanol samples, the resonance frequencies are 264 and 284 MHz, and the return loss is 34.31 and 29.56 dB, respectively. A difference of 126 MHz was observed between the resonance frequencies of DI water and air (empty channel), and a difference of 99 MHz was observed between the resonance frequencies of methanol and air.
[0033] ANSYS High Frequency Structure Simulator (HFSS) 18.2 is used as a simulation tool to extract S11 scattering parameters. HFSS software is a full-wave frequency domain three-dimensional electromagnetic field solver. It uses the gold standard precision finite element method (FEM) to model the structure and calculate the electromagnetic field in the frequency domain. The finite element method integral equation hybrid technology is based on HFSS and domain decomposition method (DDM), and is used to solve the local area of high geometric detail in complex materials. Then, HFSS 18.2 uses a field-based adaptive grid to obtain the final grid. As mentioned earlier, a constant complex permittivity value is used to model the properties of the liquid materials to be tested, water, methanol, ethanol, and butanol used in the simulation. The simulation is shifted in frequency (MHz), and the S11 reflection coefficient (dB) in the four cases results in the plot as a fitted curve format and together with the measurement data. The computer simulation of the ε'function in the MHz range is Figure 8 Is depicted as a dashed line. Picture 9 The loss tangent function in dB depicted in is represented by a dashed line. All geometric parameters of the proposed sensor. EM simulation is consistent with the measurement of pure liquid to be measured in the region of interest.
[0034] At the operating resonance frequency, the transmission coefficient S21 is approximately -0.1 to -0.2dB (independent of the liquid sample). At the resonance frequency band, the power transmission from port 1 to port 2 is almost 99%. Even if the sensor is loaded with the liquid to be measured, S11 is about -30 to -60dB for the four liquids to be measured. The power loss of S21 in the resonance region is about -0.2dB. For high-loss liquid detection, almost no loss operation, and the liquid to be tested absorbs very little power. This feature is suitable for biomedical applications.
[0035] In order to simplify the calculation, a set of exponential functions (5) and (6) are then used to derive the measurement results between the dielectric constant and the resonance frequency (in MHz) and the loss tangent and the complex dielectric constant of the liquid, respectively. The equation that characterizes the measurement results is as follows:
[0036] ε'(MHz)=11845exp(-1×f 0 (MHz)/40.61)+0.8419 (5)
[0037] loss tangent(S 11 (dB))=9.946exp(S 11 (dB)/9.345)-0.0168 (6)
[0038] Implementation results:
[0039] In order to verify the proposed method for estimating the complex permittivity and accuracy of the liquid to be tested, another experiment was performed using a DI water/ethanol mixture in the range of 10%-100%, with a step size of 10% for the liquid The concentration of ethanol. sample. In the sample, 0% ethanol means 100% deionized water. Fill the container with various concentrations of ethanol and compare with DI water alone. Picture 9 Shows the measured frequency response obtained using a vector network analyzer after loading the sensor with concentration materials with different relative complex permittivities. The changing mixture concentration corresponds to an equivalent change in resonance frequency and amplitude.
[0040] For ethanol concentrations of 100% and 10%, respectively, the resonance frequency changed from fo=264.3MHz (-34.15dB) to fo=211.2MHz (-40.92dB), which proves the deviation of the dielectric constant of 53.1MHz (6.77dB). To simplify the problem, use (5) and (6) to obtain measurement data to approximate the frequency shift of the sample, the variation of the dB factor and the interdependence between the complex permittivity. The results show that the estimated value of the complex permittivity calculated using the proposed method is consistent with the literature. However, small differences between measured values and data have been observed in the literature with high ethanol scores. Based on repeatability testing, a standard deviation of 3 MHz was observed in the 100% ethanol measurement. Calculate the 3MHz change value of the liquid ethanol to be tested from (5) to obtain a dielectric constant of 1.59, and its ε'is 23.59. If there is 5% uncertainty in the literature.