QCM sensor detection system

Abstract

A QCM sensor detection system, comprising: a machine body, a dynamic gas distribution and enrichment module, a QCM sensor detection module, a control module, and a power supply module. The dynamic gas distribution and enrichment module, the QCM sensor detection module, the control module, and the power supply module are all installed on the machine body. The dynamic gas distribution and enrichment module is connected to the QCM sensor detection module via a gas path. The power supply module and the control module are connected to the dynamic gas distribution and enrichment module and the QCM sensor detection module via wires and are used for power supply and control.

Description

A QCM sensor detection system Technical Field

[0001] This invention relates to the field of detection technology, and in particular to a QCM sensor detection system. Background Technology

[0002] Gases, as biomarkers, can be used in various detection scenarios. With the development of science and technology, higher demands are placed on gas detection technology in fields such as environmental monitoring, food safety, and medical testing. Currently, gas detection technologies are mostly based on laboratory instrumental analysis methods, such as gas chromatography-mass spectrometry (GC-MS) and ion mobility spectrometry (IMP). While these technologies offer extremely high detection accuracy and selectivity, they suffer from drawbacks such as expensive equipment, complex operation procedures, and time-consuming processes, preventing real-time and rapid gas detection. Electronic nose devices based on metal-oxide-semiconductor (MOS) sensor arrays can achieve rapid and efficient sample detection; however, MOS sensors require high-temperature operating environments and have relatively poor selectivity. QCM sensor detection systems based on quartz crystal microbalance (QCM) sensors can achieve rapid sample detection at room temperature and possess extremely high detection accuracy.

[0003] QCM sensors possess mass-sensitive characteristics, converting minute mass changes caused by odor molecules adsorbed on the electrode surface into frequency changes to reflect sample information, achieving detection accuracy down to the nanogram level. The Sauerbrey equation explains the fundamental principle of QCM sensors: there is a linear relationship between the frequency change of the quartz crystal and the mass change adsorbed on its surface.

[0004] Where Δf is the change in the resonant frequency of the quartz crystal, the negative sign indicates a negative correlation, f0 is the inherent oscillation frequency of the quartz crystal, A is the effective working area of ​​the quartz crystal resonator, and μ q ρ is the shear modulus of quartz crystal. q Let m be the density of the quartz crystal, m be the mass of the quartz crystal, and Δm be the change in mass of the quartz crystal electrode surface. When Δm / m << 1, the above equation is the Sauerbrey equation.

[0005] QCM sensors require the preparation and modification of sensitive materials to achieve high selectivity and high sensitivity. In addition, to improve the detection performance of the QCM detection system, modules such as the gas distribution system, temperature and humidity control also need to be designed. Summary of the Invention

[0006] To address this, the present invention proposes a QCM sensor detection system that can rapidly detect the volatile odor of a sample under room temperature conditions, and has advantages such as high sensitivity, good selectivity, and minimal interference from temperature and humidity.

[0007] To achieve the objectives of this invention, the following technical solution is adopted:

[0008] A QCM sensor detection system includes a body, a dynamic gas blending enrichment module, a QCM sensor detection module, a control module, and a power supply module, wherein: the dynamic gas blending enrichment module, the QCM sensor detection module, the control module, and the power supply module are all mounted on the body; the dynamic gas blending enrichment module is connected to the QCM sensor detection module through a gas path; the power supply module and the control module are connected to the dynamic gas blending enrichment module and the QCM sensor detection module through wires for power supply and control.

[0009] The QCM sensor detection system includes a dynamic gas enrichment module comprising a gas source balancing gas path, a sample volatile odor enrichment gas path, a low-flow baseline acquisition gas path, a humidity control mixed gas detection gas path, and a high-flow cleaning gas path.

[0010] The QCM sensor detection system, wherein the airflow path of the gas source balance gas path includes: nitrogen gas is released from the main valve of the nitrogen cylinder, passes sequentially through the input and output ports of the pressure reducing valve, the second and third ports of the first two-position three-way solenoid valve, the second and third ports of the second two-position three-way solenoid valve, the filter, and then is discharged into the air.

[0011] The QCM sensor detection system, wherein the gas flow path of the sample volatile odor enrichment gas path includes: nitrogen gas flow released from the main valve of the nitrogen cylinder, passing sequentially through the input and output ports of the pressure reducing valve, the second and first ports of the first two-position three-way solenoid valve, the input and output ports of the first mass flow controller, the second and third ports of the fourth two-position three-way solenoid valve, the input and output ports of the sample chamber, the third and second ports of the fifth two-position three-way solenoid valve, and the input port of the aluminum foil gas sampling bag.

[0012] The QCM sensor detection system, wherein the airflow path of the low-flow baseline acquisition gas path includes: nitrogen gas being released from the main valve of the nitrogen cylinder, passing sequentially through the input and output ports of the pressure reducing valve, the second and third ports of the first two-position three-way solenoid valve, the second and first ports of the second two-position three-way solenoid valve, the input and output ports of the second mass flow controller, the third and second ports of the third two-position three-way solenoid valve, the input and output ports of the gas detection chamber, and a filter, before being discharged into the air.

[0013] The QCM sensor detection system, wherein the airflow path of the humidity control mixed gas detection gas path includes: nitrogen gas being released from the main valve of the nitrogen cylinder, passing sequentially through the input and output ports of the pressure reducing valve, the second and first ports of the first two-position three-way solenoid valve, the input and output ports of the first mass flow controller, the second and first ports of the fourth two-position three-way solenoid valve, the input and output ports of the electrolyte tank, the first and second ports of the fifth two-position three-way solenoid valve, the input and output ports of the aluminum foil gas sampling bag, the first and second ports of the third two-position three-way solenoid valve, the input and output ports of the gas detection chamber, and a filter, before being discharged into the air.

[0014] The QCM sensor detection system, wherein the airflow path of the high-flow cleaning gas path includes: nitrogen gas being released from the main valve of the nitrogen cylinder, passing sequentially through the input and output ports of the pressure reducing valve, the second and third ports of the first two-position three-way solenoid valve, the second and first ports of the second two-position three-way solenoid valve, the input and output ports of the second mass flow controller, the third and second ports of the third two-position three-way solenoid valve, the input and output ports of the gas detection chamber, and the filter, and then being discharged into the air.

[0015] The QCM sensor detection system includes a QCM detection module comprising a gas detection chamber, a constant temperature heating stage, a QCM sensor, a crystal oscillator test box, and a quartz crystal microbalance.

[0016] The QCM sensor detection system described herein includes the following working processes: gas source balancing and debugging stage, sample enrichment stage, baseline acquisition stage, gas mixing detection stage, baseline recovery stage, and data storage and processing stage.

[0017] The QCM sensor detection system described above includes the following steps for processing the collected sample odor information:

[0018] (1) Data preprocessing was performed on the collected sample odor frequency signals, and the SG five-point smoothing algorithm was used to process the raw data;

[0019] (2) Extract features from the smoothed frequency signal curve to obtain the eigenvalue matrix.

[0020] (3) Input the eigenvalue matrix into the machine learning model for training and learning, and test and verify to obtain a reliable sample detection model.

[0021] The QCM sensor detection system, wherein:

[0022] The frequency data preprocessing SG smoothing algorithm in step (1) is as follows:

[0023] The filter window width is n = 2m + 1, and the measurement points are f = (-m, -m + 1, ..., -1, 0, 1, ..., m - 1, m). Formula (1) is used to fit the data points in the window. The n equations form a system of k linear equations. If n > k, the equation has a solution. The fitting parameter A is determined by the least squares method.

[0024] The QCM sensor detection system, wherein:

[0025] Take two adjacent points before and after each current data point, and use a cubic polynomial. To approximate the result, the coefficients a0, a1, a2, and a3 are determined using the least squares method, resulting in the five-point cubic smoothing formula:

[0026] in f i The improvement value. Attached Figure Description

[0027] Figure 1 is a schematic diagram of the QCM sensor detection system;

[0028] Figure 2 shows the hardware circuit wiring diagram of the detection system;

[0029] Figure 3 shows the composite material modification diagram of the QCM sensor electrode surface;

[0030] Figure 4 shows the equivalent circuit model of the AT-type quartz crystal resonator;

[0031] Figure 5 shows the software interface of the control module;

[0032] Figure 6 shows the response curve of the detection system for detecting the volatile odor of the sample.

[0033] Figure 7 is a schematic diagram of the QCM sensor detection system. Detailed Implementation

[0034] The specific embodiments of the present invention will be described in detail below with reference to Figures 1-7. These embodiments are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Obviously, the embodiments described in this invention are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0035] The terms "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of the invention include the specific features, structures, or characteristics described in connection with that embodiment. Therefore, the terms "comprising," "including," "having," and variations thereof in this specification mean "including but not limited to," unless otherwise specifically emphasized.

[0036] As shown in Figure 1, the QCM sensor detection system includes a main body (not shown in the figure), a dynamic gas blending and enrichment module, a QCM sensor detection module, a control module, and a power supply module. The dynamic gas blending and enrichment module, QCM sensor detection module, control module, and power supply module are all mounted on the main body. The system's gas source is high-purity nitrogen, which is compressed and stored in a steel cylinder. The dynamic gas blending and enrichment module and the QCM sensor detection module are connected together via silicone tubing. The power supply module and control module are connected to the dynamic gas blending and enrichment module and the QCM sensor detection module via wires, providing power and control functions. As shown in Figure 7, the gas path and circuit sections of the detection system are designed independently, located on the upper and lower layers of the device, respectively. Separating the gas path and circuit sections avoids mutual interference and reduces the failure rate of the detection system. When the system is running and detecting samples, factors such as volatile test gases in the gas path and moisture in the humidity control scheme may cause circuit failures, resulting in short circuits or even damage to components. The circuit section has significant heat dissipation due to the presence of the power module. Separating it from the gas path section prevents additional temperature field factors from interfering with sample detection. The open design of the system is also for this reason. In addition, separating the gas path and circuit section facilitates the maintenance and repair of the detection system, and the circuit wiring is more standardized, neat, and aesthetically pleasing, combining both practicality and aesthetics.

[0037] The dynamic gas mixing and enrichment module includes a gas source balancing gas path, a sample volatile odor enrichment gas path, a low-flow baseline acquisition gas path, a humidity control mixed gas detection gas path, and a high-flow cleaning gas path.

[0038] The gas source balancing circuit is to prevent the system from being under negative pressure and causing airflow impact that could damage the components. The nitrogen flow is released from the main valve of nitrogen cylinder 1, and passes sequentially through the input and output ports of pressure reducing valve 2, the second and third ports of the first two-position three-way solenoid valve 3, the second and third ports of the second two-position three-way solenoid valve 12, and filter 11, before being discharged into the air.

[0039] The sample odor enrichment gas path is for preparing test samples. Based on the volatility of the test sample, this invention selects an aluminum foil gas sampling bag 9 of appropriate capacity and sets the flow rate of the first mass flow controller 4 to determine the enrichment preparation time for the test sample. First, the sample is placed in the sample chamber 7, and the input valve of the aluminum foil gas sampling bag 9 is opened by rotating it 180° counterclockwise. It is confirmed that the output valve of the gas sampling bag 9 is in the closed state. The nitrogen gas flow is released from the main valve of the nitrogen cylinder 1 and passes sequentially through the input and output ports of the pressure reducing valve 2, the second and first ports of the first two-position three-way solenoid valve 3, the input and output ports of the first mass flow controller 4, the second and third ports of the fourth two-position three-way solenoid valve 5, the input and output ports of the sample chamber 7, the third and second ports of the fifth two-position three-way solenoid valve 8, and the input port of the aluminum foil gas sampling bag 9, thereby enriching the sample odor in the gas sampling bag 9.

[0040] Before sample testing, baseline acquisition is required to eliminate interference from system factors such as airflow disturbance. For this purpose, the present invention sets up a low-flow baseline acquisition gas path. In addition, in order to ensure the consistency of detection conditions, the second mass flow controller 13 is set to a low-flow mode (flow rate ≤ 1000 ml / min) to keep the flow range consistent with that of the first mass flow controller 4. The nitrogen gas flow is released from the main valve of the nitrogen cylinder 1, and passes sequentially through the input and output ports of the pressure reducing valve 2, the second and third ports of the first two-position three-way solenoid valve 3, the second and first ports of the second two-position three-way solenoid valve 12, the input and output ports of the second mass flow controller 13, the third and second ports of the third two-position three-way solenoid valve 14, the input and output ports of the gas detection chamber 16, and the filter 19, and then is discharged into the air.

[0041] In the humidity-controlled mixed gas detection gas path, this invention introduces an electrolyte pool (different saturated salt solutions correspond to different humidity levels in the headspace vapor, see Table 1) into the sample detection gas path for humidity control to verify the system's resistance to humidity interference. The output valve of the aluminum foil gas sampling bag 9 is opened by rotating it counterclockwise by 180°. Nitrogen gas is released from the main valve of the nitrogen cylinder 1, passing sequentially through the input and output ports of the pressure reducing valve 2, the second and first ports of the first two-position three-way solenoid valve 3, the input and output ports of the first mass flow controller 4, the second and first ports of the fourth two-position three-way solenoid valve 5, the input and output ports of the electrolyte pool 6, the first and second ports of the fifth two-position three-way solenoid valve 8, the input and output ports of the aluminum foil gas sampling bag 9, the first and second ports of the third two-position three-way solenoid valve 14, the input and output ports of the gas detection chamber 16, and the filter 19, before being discharged into the air.

[0042] To expedite the desorption of odor molecules from the sensor after sample testing, a high-flow-rate purge gas path was implemented. The second mass flow controller 13 was set to high-flow mode (flow rate ≤ 3000 ml / min), and nitrogen gas was released from the main valve of nitrogen cylinder 1. The gas flow passed sequentially through the input and output ports of pressure reducing valve 2, the second and third ports of the first two-position three-way solenoid valve 3, the second and first ports of the second two-position three-way solenoid valve 12, the input and output ports of the second mass flow controller 13, the third and second ports of the third two-position three-way solenoid valve 14, the input and output ports of the gas detection chamber 16, and the filter 19, before being discharged into the air.

[0043] The QCM detection module includes a gas detection chamber 16, a constant temperature heating stage 17, a QCM sensor 18, a crystal oscillator test box 15, and a quartz crystal microbalance 20. The QCM sensor 18 is installed inside the gas detection chamber 16, which is placed on the constant temperature heating stage 17. The constant temperature heating stage 17 is used for temperature control during detection. The QCM sensor 18, the crystal oscillator test box 15, and the quartz crystal microbalance 20 are electrically connected in sequence. The crystal oscillator test box 15 drives the QCM sensor 18 to generate an oscillation signal. Combined with the frequency measuring instrument of the quartz crystal microbalance 20, the frequency change of the QCM sensor 18 can be detected in real time, thereby reflecting the information of the sample being tested.

[0044] As shown in Figure 3, the QCM sensor 18 is modified with a PVC / ZIF-8@MIPs composite sensing material. ZIF-8 is a typical MOF 23 material with advantages such as good thermal stability, high porosity, and large specific area. MIPs 24 have high selectivity and affinity for template molecules (target molecules), enabling selective adsorption of target molecules. Polyvinyl chloride (PVC) 22 is an inert and soluble polymer that can be used as an adhesive to fix ZIF-8@MIPs to the electrode surface. Therefore, the sensor possesses the excellent properties of both metal-organic frameworks (MOFs) and molecularly imprinted polymers (MIPs). The novel composite material, with the rhombic dodecahedral 23 ZIF-8 as the core, provides more attachment sites for MIPs. The MIPs encapsulate ZIF-8 to form MIPs clusters, and the entire structure is firmly fixed to the electrode surface by the spider web-like PVC.

[0045] Among the aforementioned composite sensitive modification materials, ZIF-8 is a major representative of metal-organic framework (MOF) zeolite imidazole frameworks, composed of Zn 2+As a node, imidazole acts as a connecting bridge, coordinating and assembling into novel inorganic-organic hybrid materials. The three-dimensional porous structure endows them with an ultra-large specific surface area. MIPs are prepared based on molecular imprinting technology (MIT). The preparation of molecular imprinted polymers (MIPs) employs a bulk polymerization method, using the target molecule as a template molecule. Functional monomers are screened and pre-assembled with it to form a template molecule-functional monomer complex. A crosslinking agent, initiator, and porogen are then added to thermally initiate the polymerization reaction, fixing the structure of the pre-assembled template molecule-functional monomer complex. The complex is then washed with an eluent to remove the template molecule, thus completing the preparation of a three-dimensional cavity polymer material (MIP) with the target molecule. The three-dimensional cavity is completely complementary to the shape, size, and chemical bond energy of the target molecule, resulting in the high affinity and strong selectivity of MIPs for the target molecule. The introduction of polyvinyl chloride (PVC) increases the adhesion of the sensitive modification film to the QCM sensor electrode. However, the amount of PVC added must be carefully controlled. Excessive PVC will cause a large number of cavities in the MIPs to be buried and increase the relative mass of the sensitive modification film, greatly affecting the selectivity and sensitivity of the sensor.

[0046] The mass ratio of ZIF-8, MIPs, and PVC in the composite sensitive modification material is 2:5:1 during synthesis. The preparation method is as follows: First, prepare a dichloromethane solution with PVC as the base material. The concentration of this solution is 1 mg / ml, that is, if the amount of PVC material is 1 mg, then the amount of dichloromethane solution is 1 ml. Add the ZIF-8 and MIPs powders in the above ratio into the dichloromethane solution with PVC as the base material (concentration of 1 mg / ml), and then mix them evenly by sonication for 15 min. Use a micro-syringe to extract the mixture and drop it onto the quartz crystal body of the QCM sensor. After the dichloromethane solution evaporates, the QCM sensor modified with the composite sensitive modification material is formed.

[0047] In this invention, the correctness of the Sauerbrey equation can only be guaranteed if the mass of the rigid adhesion layer formed by the sensitive membrane of the QCM sensor does not exceed 2% of the mass of the quartz crystal itself. Otherwise, the QCM sensor will be difficult to start oscillating or the quality factor Q will decrease. Based on the mass of the quartz crystal body, the final drop amount of the composite sensitive modification material was determined to be 15 μL. The prepared QCM sensor has a sensitivity of 2 Hz / ppm and a detection limit (LOD) of <1 Hz / ppm. In the range of 1-200 ppm, the response to target gas molecules is much higher than that to non-target molecules, and the repeatability index Re≥98%. This QCM sensor combines the properties of metal-organic framework ZIF-8 with its numerous pores and large specific surface area and the high selectivity and affinity of molecularly imprinted polymers, effectively realizing rapid real-time detection of samples at room temperature.

[0048] In gas phase detection, the QCM sensor often employs an AT-type quartz crystal resonator. Its equivalent circuit model (Figure 4) can be simplified to a circuit composed of resistors, capacitors, and inductors. According to oscillation circuit theory, when the imaginary part and phase of the total impedance Z of the equivalent circuit are both zero, the equivalent circuit resonates. At this time, the series resonant frequency f is... s Parallel resonant frequency f p Maximum impedance frequency f zmax and minimum impedance frequency f zmin for:

[0049] In the formula, C0 is the static capacitance, C is the dynamic capacitance, L is the dynamic inductance, and Q is the quality factor. Because the resonant frequency f of the QCM cannot be obtained through direct measurement... r and anti-resonant frequency f a The value is such that the commonly used approach is to use the series resonant frequency f. s and minimum impedance frequency f zmin Replacing the resonant frequency f r Using the parallel resonant frequency f p and the maximum impedance frequency f zmax Replace the anti-resonant frequency f a .

[0050] The control module is used for the overall control of the QCM sensor detection system, and includes a data acquisition card 10, a relay (not shown in the figure), and an industrial control computer 21. The control module software for the QCM sensor detection system is written using Microsoft Visual Studio 2010 software. The written control software is embedded in the industrial control computer 21, and all the necessary operating environments for the system are configured. Communication between the control module program and the data acquisition card 10 is established to call the various functions of the acquisition card. The `int OpenUsbV12(void)` and `int CloseUsbV12(void)` functions are called to open and close the data acquisition card 10; the `int DASingleOutV12(int chan, int value)` function is called to set the flow rate of the first mass flow controller 4 (0-1000ml / min) and the second mass flow controller 13 (0-3000ml / min); the `int ADContinuV12(int ad_mod, int chan, int gain, int Num_Sample, int Rate_Sample, float*databuf)` function is called to provide flow rate feedback for the first mass flow controller 4 (0-1000ml / min) and the second mass flow controller 13 (0-3000ml / min); and the `int DoSetV12(unsigned char chan, unsigned char ...` function is called to provide flow rate feedback for the first mass flow controller 4 (0-1000ml / min) and the second mass flow controller 13 (0-3000ml / min). The operation function controls the energization and de-energization of the relay (not shown in the figure), thereby controlling the pneumatic circuit switching of two-position three-way solenoid valves 3, 5, 8, 12, and 14 (unenergized state: the first port of the solenoid valve is closed, and the second and third ports are connected; energized state: the third port of the solenoid valve is closed, and the first and second ports are connected).

[0051] The power module includes two 24V power supplies and one 15V power supply. After power calculation for each load, the two 24V power supplies provide power to the two-position three-way solenoid valves 3, 5, 8, 12, and 14, the industrial control all-in-one computer 21, and the constant temperature heating platform 17, respectively. The 15V power supply provides power to the mass flow controllers 4 and 13.

[0052] As shown in Figure 2, the hardware circuit layout of the QCM sensor detection system is as follows: the industrial control all-in-one computer 21 communicates with the data acquisition card 10, the quartz crystal microbalance 20, and the constant temperature heating stage 17 via USB. The OUT3-OUT7 ports of the data acquisition card 10 control the on / off state of five relays (not shown in the figure), thereby controlling the working state of five two-position three-way solenoid valves 3, 5, 8, 12, and 14. The AD1 and AD2 ports of the data acquisition card 10 are connected to the SET ports of two mass flow controllers 4 and 13, and the flow rate value is set by inputting a 0-5V analog voltage. The FB ports of the mass flow controllers 4 and 13 are connected to the DA1 and DA2 ports of the data acquisition card 10, and output a 0-5V analog voltage to provide feedback on the real-time flow rate of the mass flow meter. The two pins of the QCM sensor 18 are connected to the crystal oscillator box 15, and the crystal oscillator box 15 is connected to the COM port of the quartz crystal microbalance 20. This module converts the minute mass changes caused by the odor molecules adsorbed on the surface of the QCM sensor into frequency changes and transmits them to the industrial control all-in-one computer.

[0053] As shown in Figure 5, the QCM sensor detection system control software includes the flow value setting and real-time flow feedback functions of two mass flow controllers 4 and 13, the function of controlling different working processes of the QCM sensor detection system by five two-position three-way solenoid valves 3, 5, 8, 12, and 14, and the start and stop functions of the QCM sensor detection system.

[0054] As shown in Figure 6, the response curve of the detection system for detecting the volatile odor of the sample begins with the baseline acquisition stage. The injection of the target gas causes the QCM sensor to generate a response signal (Δf and Δm are negatively correlated, so the curve goes down) until it stabilizes. The injection of N2 begins the cleaning stage, and odor molecules are desorbed from the sensor surface. The curve returns to the baseline state. After saving the sample response data, the detection of the sample ends.

[0055] The working process of the QCM sensor detection system includes: gas source balancing and debugging stage, sample enrichment stage, baseline acquisition stage, gas mixing detection stage, baseline recovery stage, and data storage and processing stage.

[0056] During the gas source balancing and debugging phase, all two-position three-way solenoid valves are in an unenergized state (unenergized state: the second and third ports of the solenoid valve are connected; energized state: the first and second ports of the solenoid valve are connected, the same below). The main valve of nitrogen cylinder 1 is opened, and the pressure reducing valve 2 is adjusted to output a suitable flow rate of nitrogen gas. The nitrogen gas flows through the second and third ports of the first two-position three-way solenoid valve 3, the second and third ports of the second two-position three-way solenoid valve 12, and filter 11 before being discharged into the air. During the sample enrichment phase, the first two-position three-way solenoid valve 3 is energized, while the other two-position three-way solenoid valves are unenergized. The sample is placed in the sample chamber, and the input valve of the aluminum foil gas bag 9 is opened. The nitrogen gas flows through the second and first ports of the first two-position three-way solenoid valve 3, the input and output ports of the first mass flow controller 4, the second and third ports of the fourth two-position three-way solenoid valve 5, the sample chamber 7, and the fifth two-position three-way solenoid valve 8. Third, the second port and the input valve of the aluminum foil gas bag 9 enter the aluminum foil gas bag. The flow rate and enrichment stage time of the first mass flow controller 4 are set according to the volume of the gas bag. When the gas bag 9 is full, the real-time flow output of the first mass flow controller 4 will be abnormal (the flow rate setting of the first mass flow controller 4 requires an input voltage of 0-5V, corresponding to a flow rate of 0-1000ml / min. The airflow passing through the mass flow controller 4 will feed back a voltage of 0-5V, corresponding to a flow rate of 0-1000ml / min. The input and feedback voltages of the mass flow controller 4 are set and received by a data acquisition card. When the gas bag is full, the air pressure at the output port of the mass flow controller 4 will increase. As a result, the real-time flow output of the mass flow controller 4 will be abnormal. This can be used as a sign that the gas bag is full, that is, the sample enrichment stage is completed). At this time, the input valve of the aluminum foil gas bag 9 is closed, and the sample odor enrichment is completed.

[0057] During the baseline acquisition phase, the second two-position three-way solenoid valve 12 is energized, while the other two-position three-way solenoid valves are de-energized. Nitrogen gas flows through the second and third ports of the first two-position three-way solenoid valve 3, the second and first ports of the second two-position three-way solenoid valve 12, the input and output ports of the second mass flow controller 13, the third and second ports of the third two-position three-way solenoid valve 14, the input and output ports of the gas detection chamber 16, and the filter 19 before being discharged into the air. During this phase, the nitrogen gas flows through the gas detection chamber 16. If the frequency signal fluctuation of the QCM sensor 18 inside the detection chamber 16 is within ±1Hz, the baseline acquisition phase is considered complete.

[0058] During the gas mixing detection phase, the first, third, fourth, and fifth two-position three-way solenoid valves 3, 14, 5, and 8 are all energized, while the second two-position three-way solenoid valve 12 is de-energized. Based on the required humidity conditions, the appropriate electrolyte solution is added to the electrolyte tank 6 to control humidity (humidity significantly affects the QCM sensor signal; therefore, a humidity control scheme needs to be added to this system to test the sensor's resistance to humidity interference; different electrolyte solutions produce different headspace humidity levels, providing different humidity control environments). The input and output valves of the aluminum foil gas sampling bag 9 are opened, and nitrogen gas flows through the second and first ports of the first two-position three-way solenoid valve 3. A mass flow controller consists of a mass flow controller with input and output ports 4, a second two-position three-way solenoid valve 5 with second and first ports, an electrolyte tank 6, a second two-position three-way solenoid valve 8 with first and second ports, an aluminum foil gas bag 9 with input and output ports, a third two-position three-way solenoid valve 14 with first and second ports, a gas detection chamber 16 with input and output ports, and a filter 19. High-purity nitrogen gas is used as the carrier gas to send a sample with a certain humidity and volatile odor into the gas detection chamber 16 for detection. The surface of the QCM sensor 18 (installed in the gas detection chamber) generates a frequency signal due to the change in the adsorption mass of odor molecules, thus completing the mixed gas detection stage.

[0059] During the baseline recovery phase, the second two-position three-way solenoid valve 12 is energized, while the other two-position three-way solenoid valves are de-energized. The second mass flow controller 13 is set to high flow mode (3000 ml / min). Nitrogen gas flows through the second and third ports of the first two-position three-way solenoid valve 3, the second and first ports of the second two-position three-way solenoid valve 12, the input and output ports of the second mass flow controller 13, the third and second ports of the third two-position three-way solenoid valve 14, the input and output ports of the gas detection chamber 16, and the filter 19 before being discharged into the air. The nitrogen gas flows through the gas detection chamber 16 to purge it. The baseline recovery phase is considered complete when the frequency signal of the QCM sensor 18 inside the detection chamber recovers to 95% of the baseline signal and the signal fluctuation is within ±1 Hz.

[0060] In the data storage and processing stage, the computer stores and processes the sample odor frequency signal to obtain the sample detection results.

[0061] The performance evaluation of the QCM sensor detection system includes the system's test sensitivity, the system's limit of detection (LOD), and the system's test repeatability. This evaluation process uses test gases of different concentration gradients prepared from chemically pure reagents (analytical grade ≥99%). The preparation process is as follows: First, the saturated vapor concentration c of the different substances (analytical grade) used for testing is calculated. 饱和蒸气浓度 :

[0062] Among them, P 蒸气压—Vapor pressure of the substance, mmHg; t—Temperature, ℃; A, B, and C are constants, which can be found in Table 2.

[0063] Where D is the vapor density, mg / m³. 3 M—Molecular weight, g / mol; R—Gas constant, 8.314 Pa·m 3 / (K·mol); T—Kelvin temperature, K.

[0064] Among them, c 饱和蒸气浓度 —Saturated vapor concentration, ppm.

[0065] Using the saturated vapor concentration as a threshold, any concentration c' below the saturated vapor concentration is prepared according to the following formula, and the pure liquid sample is injected into the container at room temperature and allowed to stand and evaporate:

[0066] Where, c'—concentration of the sample to be tested, ppm; ρ—density of the liquid sample, g / mL; T—temperature inside the container (298K at 25℃), K; V 注入 —Volume of pure sample injected into the container, μL; M—Molecular weight of the substance, g / mol; V s —The volume of the container, in L.

[0067] (1) The test sensitivity of the system is defined as the ratio of the change in frequency to the corresponding concentration when the response reaches equilibrium:

[0068] Where S is the system's test sensitivity, Δf is the frequency change, and Δc is the corresponding concentration.

[0069] (2) The system's lowest detection limit (LOD) is determined by three times the signal-to-noise ratio:

[0070] Where S represents the system's test sensitivity.

[0071] (3) The test repeatability (Re) of the system is used to evaluate the consistency of the system's response to the same concentration of target gas:

[0072] Where x i In response to the frequency value at which equilibrium is reached, The average value of the response frequency values ​​from multiple tests is given when the test samples are at the same concentration gradient, k ≥ 6 (Re ≥ 95%).

[0073] The QCM sensor detection system processes the collected sample odor information in the following steps:

[0074] (1) Data preprocessing was performed on the collected sample odor frequency signals, and the SG five-point smoothing algorithm was used to process the raw data;

[0075] (2) Extract features from the smoothed frequency signal curve to obtain the eigenvalue matrix.

[0076] (3) Input the eigenvalue matrix into the machine learning model for training and learning, and test and verify to obtain a reliable sample detection model.

[0077] The frequency data preprocessing SG smoothing algorithm in step (1) is as follows:

[0078] Where f is the sensor frequency response value, and A = [a0, a1, a2, ..., a...]. k-1 ] is the coefficient matrix, The improved value of the sensor frequency response is given by the filter window width n = 2m + 1, and the measurement points are f = (-m, -m + 1, ..., -1, 0, 1, ..., m - 1, m). The data points within the window are fitted using the k-1 degree polynomial of the above formula. The n equations form a system of k linear equations. If n > k, the equation has a solution. The fitting parameter A is determined by the least squares method.

[0079] In step (1), the SG five-point cubic smoothing algorithm is used. The system uses n equidistant points, and takes two adjacent points before and after each current data point, using a cubic polynomial. To approximate the result, the coefficients a0, a1, a2, and a3 are determined using the least squares method, resulting in the five-point cubic smoothing formula:

[0080] in f i The improved value requires n≥5 in the formula. The calculation of endpoints: use the formula for the 1st and 2nd points. For the (n-1)th and nth points, use the formula Use the formula for the remaining points. Perform smoothing processing.

[0081] In step (2), the following feature values ​​of the frequency signal curve are extracted: relative change value RCV, integral value INV, and average differential value ADC.

[0082] RCV = maxf(t) i )-minf(t i )

[0083] Where, f(t) i Let f(t) be the frequency response of the sensor in the i-th second, and max f(t)i ) represents the maximum value of the frequency response within n seconds during the sample injection detection period, min f(t) i () represents the minimum frequency response within n seconds during sample introduction and detection (sampling frequency can be 1Hz). This represents the total time when the frequency response curve reaches its peak value.

[0084] In step (3), the machine learning model is determined according to the sample testing requirements, and the quantitative and qualitative analyses correspond to the regression model and the classification model, respectively.

[0085] This invention enables rapid and targeted identification of characteristic volatile odors in samples at room temperature with extremely high detection accuracy. Through the dynamic gas enrichment module, electrolyte cell, and constant temperature heating stage, it can flexibly simulate test scenarios under complex interfering gases, varying humidity, and temperature conditions, thus providing the detection system with better reliability and applicability.

[0086] Table 1 shows the relative humidity of saturated vapor of different electrolyte salt solutions at room temperature (25℃).

[0087] Table 2

[0088] Table 2 shows the saturated vapor pressure parameters of some substances.

[0089] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A QCM sensor detection system, comprising a machine body, a dynamic gas enrichment module, a QCM sensor detection module, a control module and a power module, characterized in that: The dynamic gas enrichment module, the QCM sensor detection module, the control module and the power module are installed on the machine body; the dynamic gas enrichment module is connected with the QCM sensor detection module through a gas path; the power module and the control module are connected with the dynamic gas enrichment module and the QCM sensor detection module through wires for power supply and control.

2. The QCM sensor detection system of claim 1, wherein: The dynamic gas enrichment module comprises a gas source balance gas path, a sample volatile odor enrichment gas path, a low-flow baseline collection gas path, a humidity control mixed gas detection gas path and a high-flow cleaning gas path.

3. The QCM sensor detection system of claim 2, wherein: The gas flow route of the gas source balance gas path comprises: nitrogen flow is released from the nitrogen cylinder main valve, sequentially passes through the input port and the output port of the pressure reducing valve, the second port and the third port of the first two-position three-way electromagnetic valve, the second port and the third port of the second two-position three-way electromagnetic valve, a filter, and then is discharged into the air.

4. The QCM sensor detection system of claim 1, wherein: The QCM detection module comprises a gas detection chamber, a constant temperature heating table, a QCM sensor, a crystal vibration test box and a quartz crystal microbalance.

5. The QCM sensor detection system according to one of claims 1 to 4, characterized in that: The working process of the QCM sensor detection system comprises: a gas source balance debugging stage, a sample enrichment stage, a baseline collection stage, a mixed gas detection stage, a baseline recovery stage and a data storage and processing stage.

Images