Gas characteristics extraction method using acoustic measurement and device therefor

Abstract

To provide a method capable of extracting characteristics of an object gas by measuring physical characteristics specific to the object gas with a relatively simple device configuration, and device therefor.SOLUTION: A gas characteristics extraction method by acoustic measurement, comprises: using an acoustic resonator 100 including housings 120 and 130, acoustic wave output means 150 attached to one end of the housing 120, and sound pressure measurement means 140 disposed in the housing 130, to introduce an object gas into the acoustic resonator 100 while emitting an acoustic wave excited by a specific signal from the acoustic wave output means 150 into the acoustic resonator 100; causing the sound pressure measuring means 140 to measure sound pressure due to the acoustic wave in both a frequency domain and a time domain; and extracting characteristics of the object gas from acquired results.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a method and apparatus for characterizing a gas by acoustic measurements, and in particular to a method and apparatus for characterizing a gas by acoustic measurements in both the frequency and time domains. [Background technology]

[0002] Measuring, sensing and extracting gas properties is crucial for the identification of chemical species. Gas detectors measure certain properties such as molecular weight, density, volatility, vapor pressure, etc. Various gas properties can be extracted from the measured signal using analytical models or by combining multiple sensors measuring different properties. Such highly accurate measurements allow the identification of gas samples based on the properties of the gas molecules and are used in many fields such as agriculture, healthcare, medicine, safety, robotics, and environmental science.

[0003] There are many types of existing gas sensors, including those based on metal oxides, chemiresistors, field effect transistors, electrochemistry, surface acoustic waves (SAW), cantilevers, membranes, quartz crystals, and microchannels. When the analyte flows around the sensor element, the adsorption of the analyte molecules to the sensor element changes the behavior of the sensor element, which outputs a unique signal based on the properties of both the sensor element and the target gas. This type of sensor usually relies on the chemical affinity between the sensor element and the gas molecules, so that although the chemical selectivity can be specifically designed for a particular gas species (type of target gas), it is still difficult to determine physical parameters such as gas density (ρ) and vapor pressure in addition to concentration (C) with a simple device. In contrast, gas sensors that measure physical properties do not rely on the chemical affinity between such sensor elements and the analyte molecules, so they can be applied to any type of gas and can quantify physical parameters.

[0004] Non-Patent Documents 1 to 3 describe acoustic gas sensors using solid-state piezoelectric SAW. However, these sensors use chemical receptors to induce a frequency shift, and therefore the measurement accuracy depends on the performance of the chemical receptors. In addition, the synthesis and production of the chemical receptors is time-consuming. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] X. Qi, J. Liu, Y. Liang, J. Li, S. He, The response mechanism of surface acoustic wave gas sensors in real time, Japanese Journal of Applied Physics, 58(2019) 014001. [Non-Patent Document 2] L. Zhou, Z. Hu, P. Wang, N. Gao, B. Zhai, M. Ouyang, et al., Enhanced NO2 sensitivity of SnO2 SAW gas sensors by facet engineering, Sensors Actuators B: Chemical, 361(2022) 131735. [Non-Patent Document 3] N. Levit, D. Pestov, G. Tepper, High surface area polymer coatings for SAW-based chemical sensor applications, Sensors Actuators B: Chemical, 82(2002) 241-9. Summary of the Invention [Problem to be solved by the invention]

[0006] An object of the present invention is to provide a method and an apparatus therefor that enable extraction of characteristics of a target gas by measuring physical characteristics specific to the target gas with a relatively simple apparatus configuration. [Means for solving the problem]

[0007] The features of the present invention for achieving the above object are as follows.

[0008] [1] A method for extracting characteristics of a gas by acoustic measurement, comprising the steps of: using an acoustic resonator having a housing, an acoustic wave output means attached to one end of the housing, and a sound pressure measurement means disposed within the housing; introducing a target gas into the acoustic resonator while emitting an acoustic wave excited by a specific signal from the acoustic wave output means into the acoustic resonator; and measuring the sound pressure due to the acoustic wave in both the frequency domain and the time domain with the sound pressure measurement means; and extracting characteristics of the target gas from the results obtained in the measuring step. A method comprising: [2] The method according to [1], wherein in the measuring step, the measurement in the frequency domain includes using an acoustic wave excited by 1 / f noise having constant energy per octave, measuring a signal output from the sound pressure measuring means, and obtaining a frequency response function of the sound pressure acting on the sound pressure measuring means, and the measurement in the time domain includes using an acoustic wave excited by a sine wave of a fixed frequency, and measuring a time change in the signal output from the sound pressure measuring means. [3] The method described in [2], wherein the extracting step includes obtaining physical parameters of the target gas from a frequency response function obtained by the measurement in the frequency domain, and includes extracting features from the time change of the output signal obtained by the measurement in the time domain. [4] The method according to [3], wherein the measurement in the frequency domain includes measuring a signal output from the sound pressure measuring means when the acoustic resonator is filled with the target gas to obtain a first frequency response function, and measuring a signal output from the sound pressure measuring means when the acoustic resonator is filled with a purge gas to obtain a second frequency response function, and in the extracting step, obtaining a physical parameter of the target gas based on a resonant peak selected by comparing the first and second frequency response functions. [5] The method according to [3], wherein the measurement in the time domain includes alternately introducing the target gas and a purge gas into the acoustic resonator and measuring a time change in a signal output from the sound pressure measuring means, and in the extraction step, a feature is extracted from the time change in the output signal. [6] The method according to any of [3] to [5], further comprising applying principal component analysis or linear discriminant analysis to a data set obtained in the extraction step, the data set being composed of physical parameters of the target gas from the measurement in the frequency domain and feature quantities from the measurement in the time domain. [7] The method according to any of [1] to [6], wherein the acoustic resonator is a single-sided open-ended tube having one open end and the other closed end, and is a (1 / 4+2n / 4) wavelength resonator (n is an integer not less than 0), or a double-sided open-ended tube having both open ends, and is a (1 / 2+2n / 2) wavelength resonator (n is an integer not less than 0). [8] The method according to [7], wherein the acoustic resonator is a one-sided open tube and is a (1 / 4+2n / 4) wavelength resonator (n is an integer equal to or greater than 0). [9] The method according to [8], wherein the acoustic resonator is a quarter-wave resonator.

[10] An apparatus for extracting gas characteristics by acoustic measurement, comprising: an acoustic resonator having a housing, an acoustic wave output means attached to one end of the housing, and a sound pressure measuring means disposed within the housing, the acoustic resonator having an inlet portion for introducing a target gas into its interior, and an outlet portion capable of discharging the internal gas to the outside, the apparatus being configured to introduce a target gas into the acoustic resonator while emitting acoustic waves excited by a specific signal from the acoustic wave output means into the acoustic resonator, and to enable the sound pressure due to the acoustic waves to be measured in both the frequency domain and the time domain by the sound pressure measuring means. Effect of the Invention

[0009] According to the present invention, there is provided a method and an apparatus therefor that enables extraction of characteristics of a target gas by measuring physical characteristics specific to the target gas with a relatively simple apparatus configuration. [Brief description of the drawings]

[0010] [Figure 1] 1A and 1B are schematic cross-sectional views showing specific configuration examples of acoustic resonators that can be used in the present invention. [Diagram 2] FIG. 13 is a graph showing the results of a numerical simulation, by finite element analysis, of the resonance frequency fres of the secondary resonance of the quarter-wavelength acoustic resonator fabricated in the embodiment. [Diagram 3] FIG. 13 is a graph showing the results of a numerical simulation performed by finite element analysis of the frequency response function of the sound pressure acting on a microphone for various sound velocities c in the acoustic resonator produced in the embodiment. [Figure 4] FIG. 2 is a schematic diagram showing the configuration of a measurement system for gas measurement using the acoustic resonator produced in the examples. [Figure 5A] FIG. 13 is a diagram showing a frequency response function obtained by introducing a target gas into a measurement chamber in a frequency domain measurement in an embodiment. [Figure 5B] 5B is an enlarged view of the range enclosed by the dotted line in FIG. 5A (the range including the resonance peak of the secondary resonance). FIG. [Figure 5C]1 shows a graph plotting the relationship between the gas density ρ (X-axis) and the resonant frequency fres and the sound speed c (Y-axis) for the target gas. For both the resonant frequency fres and the sound speed c, the coefficient of determination of the fitted curve was R2=1. [Figure 6A] 5A shows an alternating cycle of frequency and time domain behavior of n-hexane for the frequency and time domain measurements of the embodiment. The frequency response functions inserted in the figure are the resonance peaks of pure nitrogen and n-hexane in the frequency range of 5.5 to 8.5 kHz for FIG. 5A. [Figure 6B] FIG. 13 is a graph showing the root mean square value of the microphone output voltage in the time domain for different gases, relating to frequency domain and time domain measurement results of an embodiment. [Figure 6C] FIG. 7 is a graph showing the microphone output voltage in the third cycle shown in FIG. 6B after subtracting the baseline value (ΔV) with respect to the measurement results in the frequency domain and the time domain of the embodiment. [Figure 6D] Normalized frequency and time domain measurements of an embodiment of the invention shown in FIG. 6C, with the response signal for water shown in the inset. [Figure 6E] 6C is a diagram for explaining the feature quantities of rise time, rise amplitude, fall time, and fall amplitude extracted from the measurement results in the time domain, using the response signal for n-hexane shown in FIG. 6C as an example, in relation to the measurement results in the frequency domain and the time domain of the embodiment. [Figure 6F] FIG. 13 is a diagram showing the results of a principal component analysis performed on a data set extracted from the frequency domain and time domain measurement results of an embodiment. [Figure 7A] FIG. 11 is a graph showing frequency response functions obtained by introducing target gases having different n-hexane concentrations into a measurement chamber, in relation to the results of concentration dependency measurements using n-hexane in an embodiment. [Figure 7B]FIG. 11 is a graph showing the root mean square value in the time domain of the microphone output voltage for target gases having different n-hexane concentrations, in relation to the results of concentration dependency measurements using n-hexane in an embodiment. [Figure 7C] 1 is a plot of the relationship between the amplitude level (ΔV) of the microphone output voltage and the n-hexane concentration in the target gas, for the results of the concentration dependency measurement using n-hexane in the embodiment. The straight line in the figure indicates the regression line. [Figure 7D] FIG. 13 is a graph showing the results of a principal component analysis performed on a data set extracted from the measurement results in the frequency domain and the time domain of a target gas having different n-hexane concentrations, regarding the results of concentration dependency measurement using n-hexane in an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, an embodiment of the present invention will be described in detail. The following description of the components may be based on a representative embodiment of the present invention, but the present invention is not limited to such an embodiment.

[0012] The method for extracting gas characteristics by acoustic measurement of the present invention uses an acoustic resonator having a housing, an acoustic wave output means attached to one end of the housing, and a sound pressure measuring means disposed within the housing.

[0013] FIG. 1 is a schematic cross-sectional view showing a specific example of the configuration of the acoustic resonator.

[0014] 1 has an upper housing 120 and a lower housing 130 via a substrate 110. The internal spaces of the upper housing 120 and the lower housing 130 are in electrical communication with each other via a hole (orifice) 112 provided in the substrate 110. Furthermore, the upper housing 120 and the lower housing 130 are provided with sealing members (O-rings) 122, 132, which ensure the sealing (sealing) between the top surface of the substrate 110 and the bottom of the upper housing 120, and between the bottom surface of the substrate 110 and the top of the lower housing 130, respectively.

[0015] A sound pressure measuring means 140 is disposed inside the lower housing 130 , and the sound pressure measuring means 140 is capable of receiving sound pressure from inside the upper housing 120 via the orifice 112 of the substrate 110 .

[0016] An acoustic wave output means 150 is attached to the upper end of the upper housing 120, i.e., the end opposite to the side where the lower housing 130 on which the sound pressure measurement means 140 is disposed is located. The portion of the acoustic wave output means 150 that emits acoustic waves is directed toward the inside of the upper housing 120. In addition, a seal member (O-ring) 124 is provided between the upper end of the upper housing 120 and the acoustic wave output means 150, ensuring the hermeticity (sealing) between the upper end of the upper housing 120 and the acoustic wave output means 150.

[0017] In addition, the upper housing 120 has openings 126 and 128 on both the left and right sides, one of the openings (e.g., opening 126) functions as an inlet for introducing a target gas into the upper housing 120 (i.e., acoustic resonator 100), and the other opening (e.g., opening 128) functions as an outlet for discharging the gas inside the upper housing 120 (acoustic resonator 100) to the outside. In other words, in the acoustic resonator 100 shown in FIG. 1, there is a room (chamber) C formed of a space including the inside of the upper housing 120 and the orifice 112 of the substrate 110, and the chamber C is configured to introduce a target gas into the inside through the opening 126 and to discharge the gas inside through the opening 128 to the outside. Hereinafter, such a chamber C is also referred to as a measurement chamber. The locations and sizes of the openings 126 and 128 are not limited to the embodiment shown in FIG. 1 and can be designed as appropriate. In addition, in the narrow sense, the measurement chamber C is the range (symbol x1 shown in FIG. 1) from the lower end of the acoustic wave output means 150 to the lower surface of the substrate 110 (the upper surface of the sound pressure measurement means 140), but in the context of the resonance wavelength of the acoustic resonator 100 described below, it should be noted that the range of the measurement chamber C takes into consideration the position of the acoustic wave generating unit 152 of the acoustic wave output means 150 (symbol x2 shown in FIG. 1) and the position of the sound pressure receiving unit 142 of the sound pressure measurement means 140 (symbol x3 shown in FIG. 1) (i.e., symbol X shown in FIG. 1). Note that, in cases where it is acceptable to exclude one or both of the values ​​of x2 and x3 in terms of the configuration (structure) of the members actually used as the acoustic wave output means 150 and the sound pressure measurement means 140, it may be possible to use, for example, the sum of the above values ​​of x1 and x2 as the substantial range of the measurement chamber C. In addition, for ease of understanding, in FIG. 1, substrate 110 is depicted as having a constant thickness; however, if the thickness of the material actually used as substrate 110 is sufficiently small compared to the above value of x1 (or the sum of x1 and x2), the effective range of measurement chamber C may be the value obtained by dividing the thickness of substrate 110 from the sum of x1 and x2 (i.e., the range from the top surface of substrate 110 to acoustic wave generating section 152 of acoustic wave output means 150).

[0018] In the acoustic resonator 100 having such a configuration, the acoustic wave output means 150 emits acoustic waves excited by a signal having a specific frequency into the acoustic resonator 100 (into the measurement chamber C), and the sound pressure measuring means 140 measures the sound pressure due to the acoustic waves, thereby enabling required acoustic measurements to be performed. In the above acoustic measurements, it is also possible to introduce gas into the measurement chamber C through the opening 126 while emitting acoustic waves from the acoustic wave output means 150 into the measurement chamber C, and measure the sound pressure due to the acoustic waves at that time with the sound pressure measuring means 140.

[0019] In the present invention, a commonly available microelectromechanical system (MEMS) microphone and speaker can be used as the sound pressure measuring means 140 and the acoustic wave output means 150, respectively. The substrate 110 may be an insulating substrate, a printed wiring board (PWB) on which conductor wiring is provided on or inside the insulating substrate, or a printed circuit board (PCB) on which electronic components are further attached. When a PCB is used as the substrate 110, it is also possible to electrically connect the sound pressure measuring means (microphone) 140 and the acoustic wave output means (speaker) 150 to it, and configure it to control their operations.

[0020] It is preferable to select a material that is unlikely to cause chemical interactions such as adsorption and desorption of gas introduced into the housing of the acoustic resonator 100 (upper housing 120 and lower housing 130). This is because the occurrence of such chemical interactions may affect the results of extraction of gas characteristics by the method of the present invention. Specifically, fluororesins such as PTFE (polytetrafluoroethylene), PFA (perfluoroalkoxyalkane), FEP (perfluoroethylenepropene copolymer), and ETFE (ethylenetetrafluoroethylene copolymer) can be suitably used as materials for the housing of the acoustic resonator 100, in particular the upper housing 120 having the measurement chamber C.

[0021] In the present invention, acoustic resonator 100 having the above-described configuration is used to measure a target gas in both the frequency and time domains, thereby achieving extraction of the characteristics of the target gas.

[0022] Here, in a representative embodiment of the present invention, the acoustic resonator 100 is preferably configured to form a 1 / 4 wavelength resonator. In the acoustic resonator 100 having the configuration shown in Fig. 1, the acoustic resonator 100 that functions as a 1 / 4 wavelength resonator can be configured by appropriately setting the value of the length X represented by the sum of the distance x1 from the lower surface of the substrate 110 to the lower end of the acoustic wave output means 150 (the part that emits acoustic waves), the distance x2 from the lower end of the acoustic wave output means 150 to the acoustic wave generating part (e.g., the diaphragm (diaphragm) of a speaker) 152 provided inside the acoustic wave output means 150, and the distance x3 from the lower surface of the substrate 110 to the sound pressure receiving part (e.g., the diaphragm (diaphragm) of a microphone) 142 provided inside the sound pressure measuring means 140. A specific configuration example of such a 1 / 4 wavelength resonator will be described in the Examples section. If the design of the device allows, acoustic resonator 100 can also be configured as a (1 / 4+2n / 4) wavelength resonator (n is an integer equal to or greater than 1), such as a 3 / 4 wavelength resonator or a 5 / 4 wavelength resonator. In addition, while acoustic resonator 100 shown in Fig. 1 is formed as a one-sided open tube (closed tube resonator) in which the top end of upper housing 120 is an open end and the bottom of lower housing 130 is a closed end, if the bottom of lower housing 130 is left open to form a double-sided open tube (open tube resonator), acoustic resonator 100 can also be configured as a 1 / 2 wavelength resonator or a (1 / 2+2n / 2) wavelength resonator (n is an integer equal to or greater than 1). That is, in one embodiment of the present invention, acoustic resonator 100 may be a single-sided open-ended tube with one open end and the other closed end, and may be a (1 / 4+2n / 4) wavelength resonator (n is an integer equal to or greater than 0), and in another embodiment, acoustic resonator 100 may be a double-sided open-ended tube with both ends open, and may be a (1 / 2+2n / 2) wavelength resonator (n is an integer equal to or greater than 0).

[0023] By using the acoustic resonator 100 configured to form the above-mentioned (1 / 4+2n / 4) wavelength resonator or (1 / 2+2n / 2) wavelength resonator (n is an integer equal to or greater than 0), more specifically, to form a 1 / 4 wavelength resonator, it becomes possible to measure the sound pressure that is significantly increased at the resonant frequency with the sound pressure measuring means 140. This resonant frequency is given by the sound velocity c and density ρ of the gas inside the acoustic resonator 100 (inside the measurement chamber C). That is, when the target gas is introduced into the acoustic resonator 100 (into the measurement chamber C) from the opening 126, the density ρ and sound velocity c of the gas in the measurement chamber C change, and the sound pressure wave emitted from the acoustic wave output means 150 changes. The change in the resonant frequency of the gas in the measurement chamber C caused by the change in the sound pressure wave can be experimentally obtained by measurement in the frequency domain, specifically, a standard 1 / f equal octave pink noise test in the audible frequency range. In addition, time-series signals obtained by measuring in the time domain under constant (fixed) frequency conditions show unique profiles that depend on the type of target gas and the concentration of specific gases contained in the target gas, so multidimensional data analysis is possible by extracting multiple feature quantities from them. Based on the multidimensional data (data set) obtained by measuring in both the frequency and time domains, it is possible to clarify the type of target gas and the concentration of specific gases contained in the target gas.

[0024] In comparison with the existing acoustic gas sensors using solid-state piezoelectric SAW as described in Non-Patent Documents 1 to 3, the device of the present invention does not use chemical receptors or the like to induce the above-mentioned frequency shift, and has a resonance frequency (f res ) The speed of sound in various gases can be measured using photoacoustic and Helmholtz resonators, but these are typically used to monitor specific pure gases with well-known properties.

[0025] In contrast, in the present invention, as specifically described in the examples below, measurements are performed on a gas prepared using the headspace gas of a sealed vial containing a specific chemical substance, and from the results, the sound speed c and density ρ can be obtained as physical parameters specific to the type of target gas. In addition, by measuring the time-series change in the time domain of the sound pressure when the target gas is introduced into the measurement chamber, multidimensional data specific to the type of target gas can be obtained, and the chemical properties of the target gas can also be obtained from the multidimensional data.

[0026] In addition, in the present invention, it is possible to extract the characteristics of the target gas by performing various statistical processes and associated machine learning, such as Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), that use physical parameters (such as sound speed and density) obtained by measurement in the frequency domain and multidimensional data obtained by measurement in the time domain to express the degree of similarity between data sets in some form based on the characteristics of the data sets.

[0027] The present invention will be described in more detail below with reference to examples. Note that the following examples are not intended to limit the present invention, but are intended to aid in understanding the present invention. EXAMPLES

[0028] [Device configuration] In this example, an apparatus (100) having a similar configuration to that described with reference to FIG. 1 was produced. A commercially available MEMS microphone (ADMP401-sparkfun) was used as the sound pressure measuring means (140), and one of the commercially available earphones (ZNS-0111-BEIY, manufactured by Boesklenn) was used as the acoustic wave output means (150). A PCB was used as the substrate (110), with an upper housing (120) made of PTFE placed on one side of the PCB, and a lower housing (130) made with a 3D printer placed on the other side of the PCB.

[0029] In the prepared device, the length of the space (measurement chamber C) inside the upper housing into which the target gas is introduced (the value of the symbol X shown in FIG. 1, i.e., the distance from the diaphragm inside the speaker to the diaphragm inside the microphone) was about 17 mm, and by dividing this length by the distance from the bottom surface of the PCB to the diaphragm of the microphone and the thickness of the PCB, the result was about 14 mm (corresponding to the distance from the top surface of the PCB to the diaphragm of the speaker). The diameter of the measurement chamber (the inner diameter of the upper housing) was 6 mm, which is the same size as the inner diameter of the O-ring (124) attached around the speaker to ensure airtightness between the speaker and the top end of the upper housing. The speaker and the O-ring were held down with a clip to ensure close contact between the two components.

[0030] With the above configuration, a quarter-wavelength acoustic resonator was formed on the upper end side of the upper housing, having an inlet portion (126) capable of introducing a target gas into the acoustic resonator (into the measurement chamber) and an exhaust portion (128) capable of exhausting gas from the inside of the measurement chamber to the outside. Here, the resonant frequency of the quarter-wave acoustic resonator is given by the following equation:

number

[0031] The results are shown in Figure 2. The f of nitrogen gas calculated by FEA is shown in Figure 2. res was 18.737kHz (c=349[m / s]). In addition, the line indicated by the symbol DD' in FIG. 2 indicates the position of the diaphragm inside the microphone, which is disposed so as to be in contact with the portion where the orifice is provided on the lower surface of the substrate.

[0032] In this embodiment, in order to more accurately obtain the sound velocity c of the target gas, the second resonance is used for the following three reasons. (1) The resonance peak is clear: Around the frequency range of the primary resonance (fundamental vibration), the resonance of the speaker's diaphragm and the Helmholtz resonance between the orifice and the lower housing volume overlap, but around the secondary resonance, there are no other significant resonances, and the resonance peak is clear. (2) Frequency change (Δf res ) is large: The second resonance has a frequency change due to changes in the gas properties that is three times larger than that of the first resonance, making it easier to distinguish between the sound speeds c of different gases. (3) It is in the audible range: f of the second resonance res is below 20 kHz, making it possible to use simple audio equipment (for example, the minimum sampling rate for DVD quality is 44.1 kHz). When the resonator is filled with a different gas, the change in density ρ and sound speed c causes the frequency f res changes. The sound speed c of an ideal gas is given by the following equation:

number

number

[0033] [Experimental Procedure] FIG. 4 is a schematic diagram showing the configuration of a measurement system for gas measurement using the above-mentioned device. In the measurement system shown in FIG. 4, the first mass flow controller (MFC1) is connected to a nitrogen line and transports headspace gas (sample gas), which is the vapor of a liquid sample contained in a vial, to a mixing chamber. The second mass flow controller (MFC2) plays a role in diluting the sample gas transported via MFC1 to a predetermined ratio by introducing pure nitrogen into the mixing chamber. The mixed gas (target gas) prepared in this way is introduced into the measurement chamber (1 / 4 wavelength acoustic resonator). The vial used was made of glass and had a volume of 20 mL, and contained 3 mL of liquid sample.

[0034] Using such a measurement system, two different acoustic wave-based measurements were performed in both the frequency and time domains to extract various information about the properties of the target gas, obtaining the density ρ, sound speed c, and other multidimensional data of the target gas.

[0035] For frequency domain measurements, to obtain the FRF, the acoustic wave was excited with 1 / f iso-octave pink noise with an amplitude of 10 mV. On the other hand, in the time domain, the acoustic wave is V in The excitation was a fixed frequency sine wave of the form × sin(2πft), where V in is the input voltage to the speaker (mV), f is the test frequency (Hz), and t is the time (s).

[0036] A PXI5406 (NI) was used for the voltage input to the speaker, and a PXI5922 (NI) was used to measure the microphone output. The bias voltage to the microphone was set to 3 V for all measurements. The typical temperature in the laboratory during the measurements was 20 °C. All measurements were performed under normal laboratory conditions to evaluate the possibility of obtaining data under a constant noise environment.

[0037] For frequency domain measurements (pink noise tests), measurements were taken at a sampling rate of 200 kHz, acquiring 200 kilosamples (kS) (i.e., data was acquired at 1 Hz), and 100 pink noise tests were averaged to reduce variations due to ambient noise sources. For time-series measurements in the time domain, measurements were performed under conditions that allow 10 Hz measurement using a standard microphone input terminal, i.e., 4.8 kS acquisition at a sampling rate of 48 kHz. In addition, in the time domain measurements, a ±500Hz band-pass filter was used to remove the effects of other unwanted noise sources. For example, at the test frequency f = 6.9 (kHz), a 6.4-7.4kHz band-pass filter was applied. After collecting the filtered data (4.8kS), the root mean square (RMS) value of the collected time series data was calculated for the change (ΔV) in the microphone output voltage at each sampling interval (100ms).

[0038] [Experimental Results] These measurement results are described in detail below.

[0039] [Frequency domain measurements: Frequency response functions of target gases prepared from various liquid samples] The liquid samples used were water (ultrapure water, prepared using an ultrapure water production system manufactured by Merck Millipore), methanol (Kanto Chemical, purity >99.8%), ethanol (Fujifilm Wako Pure Chemical Industries, purity >99.5%), ethyl acetate (Kanto Chemical, purity ≥99.5%), toluene (Wako Pure Chemical Industries, purity ≥99.5%), acetone (Sigma-Aldrich, purity ≥99.5%), and n-hexane (Fujifilm Wako Pure Chemical Industries, purity ≥96.0%).

[0040] The vapor of the above liquid sample (headspace gas in the vial) was transported to the mixing chamber by MFC1 set at 2 sccm and diluted with pure nitrogen sent from MFC2 set at 8 sccm to prepare a mixed gas (target gas) containing the sample gas at a concentration of 20% with a flow rate of 10 sccm. This was then introduced into the measurement chamber of the acoustic resonator.

[0041] For frequency domain measurements, the target gas was introduced into the measurement chamber for 4 minutes, and then the FRF was determined using pink noise. Pink noise tests were performed 100 times, with a 0.8 second delay between each pink noise test, and the FRF was averaged over approximately 3 minutes. During this time, the target gas was continuously introduced into the measurement chamber.

[0042] As shown in Figure 5A, each target gas exhibits its own unique resonance peak, and it was confirmed that the frequency and intensity of the resonance peak differs depending on the target gas. For each target gas, the maximum resonance peak is found around 6.9-7.0 kHz, which is composed of at least the following multiple resonances: The fundamental quarter-wave resonance (estimated by FEA to be 5.8 kHz), The main resonance of the speaker (measured experimentally in an open environment and found to be 6.6 kHz), and · The fundamental Helmholtz resonance between the PCB orifice and the lower housing housing the microphone (estimated by FEA to be 9kHz).

[0043] Apart from the convolution of multiple resonance peaks in the 5-9 kHz range, there is a clear resonance peak in the 17-19 kHz range, which corresponds to the secondary resonance (n=3, triple vibration) in the measurement chamber for different gases. As mentioned in the "Apparatus Configuration" section above, the secondary resonance was selected as the main resonance for determining the sound speed c of each target gas, and an enlarged view of the vicinity of this resonance peak is shown in Figure 5B. The nine graphs in Figures 5A and 5B are, from left to right based on the peak position in Figure 5B, n-Hexane, Acetone, Ethyl Acetate, Toluene, Air, Ethanol, Methanol, Water, and Pure Nitrogen.

[0044] The sharp peaks observed in the FRFs shown in Figures 5A and 5B are sometimes observed regardless of the gas flow rate, and may be attributed to acoustic noise in the laboratory environment. As shown in Figure 5B, the resonance peaks of the second resonance were clearly and accurately observed, indicating the possibility of determining the sound speed c of different gases even in a noisy environment. Note that "Air" refers to the state in which MFC1 and MFC2 are set to 0 sccm and the gas flow path is open, and under this condition, the well-known values ​​of density ρ and sound speed c of air can be obtained as reference values.

[0045] f for each target gas res was calculated by Lorentz fitting. The sound speed c of nitrogen is 349 m / s (at 20°C), and the correction length L = L from the design value of the measurement chamber length of about 14 mm using the above formula (1). eff was calculated to be 13.92 mm. A manometer (1500N, Hodaka Seisakusho) was connected to the gas flow path to check the pressure of each target gas, which showed the same value (0.5 kPa) for all target gases. This means that the pressure of each target gas is equivalent to 101.8 kPa, which is the sum of the atmospheric pressure (101.3 kPa) and the flow pressure (0.5 kPa) caused by both MFCs. Considering that the pressure in the measurement chamber is constant, the ideal gas equation can be rearranged for the gas density ρ as follows:

number

[0046] The resonant frequency f of all the target gases used in this example, including the measurements described below, res The results, including the density ρ, sound speed c, and average molar mass Mw, are shown in Table 1. The order of the target gases shown in Table 1 is based on the resonant frequency f res The resonant frequency f of the target gas compared to pure nitrogen (N2) is res The reason why decreases is because the sound speed c of each target gas decreases according to the relationship between equations (1) and (3) above, and the gas density ρ increases according to the power law (Figure 5C). (Figure 5B shows the arrow with the notation "Increasing ρ"). The relationship expected from equation (3) is ρ -0.5 and f for the actual measurement results res For both, the exponent is -0.5, indicating good agreement.

[0047] [Table 1]

[0048] [Time domain measurements: time domain response of target gases prepared from various liquid samples] The advantage of the acoustic resonator used in the present invention is that it can operate in both time and frequency domains, which provides a unique multi-dimensional data set for target gas identification. Based on this data set and some calculated parameters (e.g., Mw and ρ as shown in Table 1), target gas type can be identified. The acoustic wave equation in both space and time domain is given by:

number

number

[0049] In the time domain measurements, the mixed gas (target gas) prepared as described above was introduced into the measurement chamber of the acoustic resonator, and sampling and purging were performed alternately. In the sampling, the flow rates of MFC1 and MFC2 were 2 sccm and 8 sccm, respectively, and in the purging (pure nitrogen), the flow rates of MFC1 and MFC2 were 0 sccm and 10 sccm, respectively. Sampling and purging were performed for 4 minutes, and measurements were performed in three consecutive cycles for each target gas, resulting in measurements in an approximately steady state. The sample gas concentration in the target gas was changed by adjusting the flow rate ratio of MFC1 and MFC2 while maintaining a total flow rate of 10 sccm.

[0050] In the time domain measurements, a constant test frequency f = 6.9 kHz was used because the resonance peak shifts to lower frequencies due to the properties of the target gas (inset in Figure 6A). res = 7.46 [kHz] or less and the maximum f of n-hexane res The test frequency was chosen to cover a frequency range around =6.952 kHz. Also, at this test frequency, the power input to the speaker was minimal (V in =30 mV). Whereas a single physical parameter can be obtained from frequency domain measurements, a multi-dimensional data set can be extracted from the time domain measurement data.

[0051] Figure 6A shows an alternating cycle of frequency and time domain behavior for n-hexane. At f=6.9 kHz, the magnitude of the FRF changes when the acoustic resonator is filled with n-hexane (Gas Sampling) and nitrogen (N2 Purging). When gas sampling occurs, the amplitude of the corresponding FRF increases in the frequency domain and the microphone output voltage increases in the time domain (with a slight clipping of the microphone as it reaches its maximum voltage for n-hexane). This process is measured reversibly as an increase and decrease in the output voltage during the sampling and purging steps, respectively.

[0052] Figure 6B shows the RMS of the microphone output voltage in the time domain for different target gases. This shows that the sample gas molecules in the target gas enter and leave the acoustic resonator by sampling and purging, and is reversible for all target gases. Here, the seven graphs in Figure 6B are n-Hexane, Acetone, Ethyl Acetate, Toluene, EtOH, Methanol, and Water, in order of the change in output voltage in each cycle. The slight difference between the first cycle and the two following cycles is due to the change in the headspace gas in the vial from a static state to a dynamic state caused by the nitrogen carrier gas in the first cycle. The difference in output voltage from water, which has the lowest output voltage, to n-Hexane, which has the highest output voltage, is in the range of -3 to 500 mV, as shown in Figure 6C, after subtracting the baseline value (ΔV) of the third cycle. The correspondence between the seven graphs in Figure 6C and the types of target gas is the same as that in Figure 6B described above. It is clearly shown that the signal intensity and shape of the rise and fall times in the time domain differ for each type of target gas, and multi-dimensional data can be extracted from these differences.

[0053] To further characterize these differences, we plot the normalized responses in Figure 6D. While most of the target gases exhibited a conventional first-order response with various time constants, water vapor exhibited a very different signal, increasing for the first 20 seconds and then decreasing, resulting in a lower RMS output voltage compared to the reference nitrogen. Additionally, n-hexane (larger Mw, higher vapor pressure) exhibited the fastest rise time, while toluene (larger Mw, lower vapor pressure) exhibited the slowest rise time. This is likely due to toluene's larger Mw and lower vapor pressure compared to acetone (smaller Mw, higher vapor pressure) and ethanol (smaller Mw, lower vapor pressure).

[0054] Here, we explain the rise time, rise amplitude, fall time, and fall amplitude as features extracted from the time domain measurements (Figure 6C). The rise amplitude and rise time are defined as the amplitude level (ΔV) between 10 and 90% between 1200 and 1440 seconds and the time to reach this value, respectively. Similarly, the fall amplitude and fall time features are defined as the decrease in amplitude (ΔV) and the time required to achieve this decrease, each in the same 10-90% interval, for the 1440-1680 second purge cycle.

[0055] An outline of these terms is shown in Figure 6E, using the response signal for n-hexane shown in Figure 6C as an example. In addition, the results of Principal Component Analysis (PCA) performed using the main information extracted from the signals in both domains are shown in Figure 6F. This PCA confirmed that each sample exhibited different behavior, and that they could be distinguished by using either the frequency domain or the time domain. For example, in the frequency domain, ethanol and methanol are weighted by the sound speed c compared to other chemicals, but it is difficult to distinguish alcohols by these specific features. Also, the rise and fall times are weighted by ethyl acetate and toluene, but the rise and fall amplitudes are most related to n-hexane and acetone. This is due to the differences in vapor pressure, gas density ρ, and average molar mass Mw between these four samples. In any case, by combining the multidimensional data extracted from the measurements in the time domain and the sound speed c obtained from the measurements in the frequency domain, a multidimensional data set capable of distinguishing different target gases is obtained.

[0056] [Concentration dependent measurement] Here, we describe the measurement results in both the frequency domain and the time domain using six types of target gas with different n-hexane concentrations.

[0057] The n-hexane concentrations in the target gases used were 6%, 8%, 10%, 13%, 16%, and 20%. The method for preparing a mixed gas with an n-hexane concentration of 20% was as described above. Mixed gases with other concentrations were prepared by adjusting the flow rate ratio of MFC1 and MFC2 while maintaining a total flow rate of 10 sccm. For example, a mixed gas with an n-hexane concentration of 16% was prepared by transporting the headspace gas of a vial containing n-hexane to the mixing chamber by MFC1 set to 1.6 sccm, and diluting it with pure nitrogen sent from MFC2 set to 8.4 sccm, resulting in a flow rate of 10 sccm.

[0058] As shown in Figure 7A, in the frequency domain measurements, the density ρ of the mixed gas increases with increasing n-hexane concentration, and therefore the sound speed c decreases. Here, the six graphs in Figure 7A are, from left to right based on the peak position, 20%, 16%, 13%, 10%, 8%, and 6% n-hexane concentrations (i.e., descending order of n-hexane concentration). In the time domain measurements shown in Figure 7B, the shape of the microphone output voltage curve changed depending on the n-hexane concentration in the target gas. Specifically, when the n-hexane concentration was between 6 and 10%, the output voltage reached a maximum value at the end of the sampling step, but when the n-hexane concentration was 10% or more, the output voltage reached a maximum value at the beginning of the sampling step and then slowly decreased toward a steady value, showing a slightly different behavior. It is believed that the dynamic turbulence in the headspace inside the vial is the cause of this apparent phenomenon at different concentration levels where the vapor pressure changes. The six graphs in FIG. 7B are, in order from largest to smallest based on the amount of change in output voltage, for n-hexane concentrations of 20%, 16%, 13%, 10%, 8%, and 6% (i.e., descending order of n-hexane concentration).

[0059] Figure 7C shows the relationship between the amplitude level (ΔV) of the microphone output voltage and the n-hexane concentration C in the target gas, which shows a strong linearity (zero point indicates pure nitrogen without n-hexane, i.e. C = 0%). Thus, the linearity of ΔV vs. C can be used to determine the concentration of chemicals in the target gas. ΔV values ​​were taken at 1440 seconds for comparison (as mentioned above, data at 20% concentration was excluded because n-hexane caused clipping of the signal from the microphone). This linearity also indicates that the gas used in this experiment can be treated as an ideal gas, confirming that the system used for the measurements operates in the linear acoustic range.

[0060] The results of PCA analysis using multi-dimensional data of sound speed (frequency domain) and time domain are shown in Figure 7D (see the explanation of Figure 6E above for the explanation of the features). This shows that different concentration levels of n-hexane in the target gas can be distinguished using these features, demonstrating the advantage of the dual domain method (measurement in both frequency and time domains) based on multi-dimensional data.

[0061] [Conclusion] In this example, a quarter-wave acoustic resonator was used to measure the characteristics of different gases in both the frequency and time domains. When the target gas was introduced into the resonator, the sound speed decreased and the gas density increased compared to the gas before introduction (pure nitrogen supplied in the purge step), resulting in a shift in the frequency response in the frequency domain and an increase in sound pressure in the time domain. By measuring in the frequency domain (pink noise test), the frequency response functions and sound speeds of different target gases were accurately obtained, from which the density ρ of each target gas could be measured. In the time domain measurements, a cycle of sampling and purging different target gases was applied at a fixed frequency. A major feature of time domain analysis is the ability to extract multidimensional data. By using the multidimensional data extracted from the time domain measurement results and the sound speed obtained from the frequency domain measurement, principal component analysis was shown to be able to easily distinguish the attributes of the type of target gas and the concentration of specific gases contained in the target gas. [Industrial Applicability]

[0062] The acoustic approach of the present invention described above has a high possibility of being applied to various applications, because of the low cost and high sensitivity of recent MEMS microphones, as well as the possibility of miniaturizing and reducing the cost of various acoustic amplifiers and measuring devices. It is believed that this approach will make it possible in the future to distinguish and identify each chemical species in a mixed gas by utilizing the abundant existing acoustic-related devices and signal processing technologies.

[0063] In addition, the present invention is expected to lead to the creation of a new acoustic measurement technology system that uses a relatively small acoustic resonator, leading to the realization of gas sensing and artificial olfaction using sound. [Explanation of symbols]

[0064] 100 acoustic resonator 110 Substrate 112 Hole (orifice) 120 Upper case 122, 124 Sealing material (O-ring) 126, 128 Opening 130 Lower housing 132 Sealing material (O-ring) 140 Sound pressure measuring means (microphone) 142 Sound pressure receptor (diaphragm) 150 Acoustic wave output means (speaker) 152 Acoustic wave generating part (vibration membrane) C. Measuring chamber

Claims

1. 1. A method for extracting gas characteristics by acoustic measurements, comprising: The housing and an acoustic wave output means attached to one end of the housing; a sound pressure measuring means disposed within the housing; An acoustic resonator having a step of introducing a target gas into the acoustic resonator while emitting an acoustic wave excited by a specific signal from the acoustic wave output means into the acoustic resonator, and measuring the sound pressure caused by the acoustic wave in both the frequency domain and the time domain by the sound pressure measurement means; extracting a characteristic of the target gas from the results obtained in the measuring step; A method comprising:

2. In the measuring step, the measurement in the frequency domain includes measuring a signal output from the sound pressure measuring means using an acoustic wave excited by 1 / f noise having a constant energy per octave, and obtaining a frequency response function of the sound pressure acting on the sound pressure measuring means; The measurement in the time domain includes measuring a time change of a signal output from the sound pressure measuring means using an acoustic wave excited by a sine wave of a fixed frequency. The method of claim 1.

3. The extracting step includes: obtaining physical parameters of the target gas from a frequency response function obtained by the measurement in the frequency domain; extracting a feature from a time change of the output signal obtained by the measurement in the time domain, The method of claim 2.

4. The frequency domain measurement is measuring a signal output from the sound pressure measuring means when the acoustic resonator is filled with the target gas, and obtaining a first frequency response function; measuring a signal output from the sound pressure measuring means when the acoustic resonator is filled with a purge gas, and obtaining a second frequency response function; Including, In the extracting step, the first and second frequency response functions are compared to select a resonant peak to obtain a physical parameter of the target gas. The method of claim 3.

5. the measurement in the time domain includes alternately introducing the target gas and a purge gas into the acoustic resonator and measuring a time change in a signal output from the sound pressure measuring means; In the extracting step, a feature amount is extracted from a time change of the output signal. The method of claim 3.

6. The method of claim 3, further comprising applying principal component analysis or linear discriminant analysis to a data set obtained in the extracting step and configured from the physical parameters of the target gas from the measurement in the frequency domain and the feature quantities from the measurement in the time domain.

7. The method according to any one of claims 1 to 6, wherein the acoustic resonator is a single-ended open-ended tube having one open end and the other closed end, and is a (1 / 4 + 2n / 4) wavelength resonator (n is an integer equal to or greater than 0), or a double-ended open-ended tube having both open ends, and is a (1 / 2 + 2n / 2) wavelength resonator (n is an integer equal to or greater than 0).

8. The method according to claim 7, wherein the acoustic resonator is the one-sided open-ended tube and is a (1 / 4+2n / 4) wavelength resonator (n is an integer equal to or greater than 0).

9. The method of claim 8 , wherein the acoustic resonator is a quarter-wave resonator.

10. 1. An apparatus for extracting gas properties by acoustic measurements, comprising: The housing and an acoustic wave output means attached to one end of the housing; a sound pressure measuring means disposed within the housing; an acoustic resonator having the acoustic resonator has an inlet portion for introducing a target gas into the interior thereof and an outlet portion for discharging the gas therein to the outside; a target gas is introduced into the acoustic resonator while an acoustic wave excited by a specific signal is emitted from the acoustic wave output means into the acoustic resonator, and the sound pressure due to the acoustic wave can be measured by the sound pressure measuring means in both the frequency domain and the time domain; Device.

Images