Nano cantilever beam array and preparation method thereof, resonant array gas sensor

Abstract

A nano cantilever beam array and a preparation method thereof, and a resonant array gas sensor, the nano cantilever beam array comprising a substrate and a cantilever beam array, each cantilever beam of the cantilever beam array comprising: a support layer, arranged on the substrate, one end in contact with the substrate forming a U-shaped cantilever beam support arm, the end of the support layer away from the U-shaped cantilever beam support arm being suspended on the substrate; a metal conductive layer, arranged on the support layer; and a sensitive layer, arranged on the metal conductive layer away from the U-shaped cantilever beam support arm, on which a sensitive unit is arranged. The present invention adopts a cantilever beam structure with double arm support, which can effectively improve the sensitivity of the sensor; the cantilever beam array with a double arm support structure has multiple sensitive units, which can respond to a variety of chemical gases quickly and sensitively.

Description

Technical Field

[0001] The invention belongs to the field of gas sensors, and in particular relates to a nano cantilever beam array and a preparation method thereof, and a resonant array gas sensor. Background Art

[0002] The identification and quantitative analysis of individual components in multi-component volatile organic compounds (VOCs) mixtures has always been a hot topic of research. It is possible to solve this problem by developing easy-to-use, portable, miniaturized and durable chemical sensor arrays with distributed local characteristics. Such portable detectors can implement, in-situ, online and continuous monitoring in specific processes. Traditional chemical analysis techniques are generally based on chromatography and spectroscopy, which monitor multi-component gases by monitoring processes such as electron-photon interactions, light absorption, luminescence or electrochemistry. In recent years, the emergence of resistive sensors, electrochemical sensors, and mass sensors based on quartz crystal microbalances, surface acoustic waves and microcantilevers has promoted the development of a new generation of chemical analysis instruments. In particular, resonant mass sensors based on nanoelectromechanical systems (NEMS) have shown ultra-high sensitivity: reaching zeptograms (10 -21 g) level, at a standard atmospheric pressure, reaching attogram (10 -18 g) level mass sensitivity makes it possible to detect extremely trace amounts of chemical gases.

[0003] One obstacle currently hindering the practical development and widespread use of NEMS cantilever devices is the difficulty in achieving sensitive displacement conduction at the nanoscale. In addition to the initial challenge of manufacturing ultra-small mechanical devices, the successful implementation of NEMS involves the dual challenges of achieving very high-frequency displacement sensing while achieving extremely high sub-nanometer resolution. Displacement sensing methods commonly used in MEMS are not suitable for NEMS. For example, the efficiency of capacitive detection drops sharply at the nanoscale, and the signal is often overwhelmed by uncontrollable parasitic effects. For optical detection, diffraction effects become significant when the device size is scaled to well below the illumination wavelength used, and the optical equipment is often large and difficult to achieve portability. Summary of the invention

[0004] In view of this, one of the main objectives of the present invention is to provide a nano-cantilever array and a preparation method thereof, and a resonant array gas sensor, in order to at least partially solve at least one of the above-mentioned technical problems.

[0005] In order to achieve the above object, as one aspect of the present invention, a nano cantilever beam array is provided, comprising a substrate and a cantilever beam array, wherein each cantilever beam of the cantilever beam array comprises:

[0006] The support layer is arranged on the substrate, and one end in contact with the substrate forms a U-shaped cantilever beam support arm, and one end of the support layer away from the U-shaped cantilever beam support arm is suspended on the substrate;

[0007] A metal conductive layer is disposed on the support layer; and

[0008] The sensitive layer is arranged on the metal conductive layer at one side away from the supporting arm of the U-shaped cantilever beam, and a sensitive unit is arranged on the layer.

[0009] As another aspect of the present invention, a method for preparing the nano-cantilever array as described above is also provided, comprising:

[0010] preparing a support layer on a substrate;

[0011] preparing a metal conductive layer on the support layer;

[0012] preparing a cantilever beam pattern on the metal conductive layer;

[0013] The substrate is etched to release the cantilever beam;

[0014] A sensitive layer is prepared on the metal conductive layer to obtain the nano cantilever beam array.

[0015] As another aspect of the present invention, a resonant array gas sensor is provided, comprising:

[0016] The nanocantilever array as described above or the nanocantilever array prepared by the method as described above is used to specifically adsorb target gas molecules and convert chemical signals into thermal vibration signals of the nanocantilever array itself;

[0017] A driving and reading circuit is connected to the U-shaped cantilever support arm of the nanocantilever array, and is used to drive the nanocantilever array to resonate and measure the change of its natural frequency during the gas adsorption process;

[0018] A signal processing unit, connected to the driving and reading circuits, is used to record the change signal of the natural frequency, perform data preprocessing and feature extraction, and then pattern recognition uses the calibrated data to perform data evaluation and output the results to the response and output unit;

[0019] The response and output unit is connected to the signal processing unit and is used to output the composition and specific concentration values ​​of the multi-component mixed gas.

[0020] Based on the above technical solutions, it can be seen that the nano-cantilever beam array and its preparation method, and the resonant array gas sensor of the present invention have at least one of the following advantages over the prior art:

[0021] (1) The cantilever beam structure with double-arm support can effectively improve the sensitivity of the sensor;

[0022] (2) The cantilever beam array with a double-arm support structure has multiple sensitive units and can respond to a variety of chemical gases quickly and sensitively;

[0023] (3) Using a combination of thermoelastic actuation and piezoresistance as the driving and detection systems and implementing them in the same circuit, the sensor has both excellent portability and fast reading capabilities;

[0024] (4) An artificial neural network method based on a multi-layer perceptron (MLP) model combined with a back propagation (BP) algorithm is used to sense, accurately distinguish and quantify the concentrations of multiple gases, ensuring continuous and real-time analysis capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] For a better understanding of the embodiments of the present invention, reference is made to the accompanying drawings, in which:

[0026] Figure 1 A schematic diagram of the structure of a nano cantilever beam array in an embodiment of the present invention;

[0027] Figure 2 It is a schematic diagram of the structure of a single cantilever beam in an embodiment of the present invention;

[0028] Figure 3 Schematic diagram of the structure of the driving and reading circuit in an embodiment of the present invention;

[0029] Figure 4 A schematic diagram of a multi-layer perceptron quantized concentration estimation model with multiple layers, multiple inputs and a single output in an embodiment of the present invention;

[0030] Figure 5a-5f 1 is a flow chart of the preparation of a nanocantilever array in an embodiment of the present invention.

[0031] Description of reference numerals:

[0032] 10 - substrate; 11 - support layer; 12 - metal conductive layer; 13 - sensitive layer; 14 - U-shaped cantilever beam support arm; 15 - connection electrode. DETAILED DESCRIPTION

[0033] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

[0034] The ultra-sensitive micro-cantilever array gas sensor based on NEMS of the present invention uses a cantilever array as a sensing element and is applied to the response of the sensor array through a pattern recognition algorithm. In order to achieve the problem of identification and quantitative analysis of each component in a multi-component volatile gas (VOCs) mixture, an ultra-sensitive micro-cantilever array gas sensor based on NEMS is provided. The cantilever array gas sensor vibrates the cantilever array through a driving circuit. When the sensitive layer of the cantilever array gas sensor adsorbs the gas, its mass increases, thereby causing a change in the resonant frequency. The response of each cantilever to the gas to be measured is measured through a readout circuit. Since each cantilever responds to multiple gases, a pattern recognition algorithm is used to sense the concentrations of multiple gases and accurately distinguish and quantify them.

[0035] The present invention discloses a nano cantilever beam array, comprising a substrate and a cantilever beam array, wherein each cantilever beam of the cantilever beam array comprises:

[0036] The support layer is arranged on the substrate, and one end in contact with the substrate forms a U-shaped cantilever beam support arm, and one end of the support layer away from the U-shaped cantilever beam support arm is suspended on the substrate;

[0037] A metal conductive layer is disposed on the support layer; and

[0038] The sensitive layer is arranged on the metal conductive layer at one side away from the supporting arm of the U-shaped cantilever beam, and a sensitive unit is arranged on the layer.

[0039] In some embodiments of the present invention, the material used for the support layer includes at least one of Si3N4 or 3C-SiC, and the thickness is 60 to 100 nm.

[0040] In some embodiments of the present invention, the material used for the metal conductive layer includes at least one of gold, platinum, and copper, and has a thickness of 20 to 40 nm.

[0041] In some embodiments of the present invention, the material used in the sensitive layer includes at least one of a non-conductive polymer, a zeolite, a metal organic framework, a calixarene, and a nanocomposite material;

[0042] In some embodiments of the present invention, different sensitive units in the nanocantilever array are used to identify the same or different gases.

[0043] The present invention also discloses a method for preparing the nano cantilever beam array as described above, comprising:

[0044] preparing a support layer on a substrate;

[0045] preparing a metal conductive layer on the support layer;

[0046] preparing a cantilever beam pattern on the metal conductive layer;

[0047] The substrate is etched to release the cantilever beam;

[0048] A sensitive layer is prepared on the metal conductive layer to obtain the nano cantilever beam array.

[0049] The present invention also discloses a resonant array gas sensor, comprising:

[0050] The nanocantilever array as described above or the nanocantilever array prepared by the method as described above is used to specifically adsorb target gas molecules and convert chemical signals into thermal vibration signals of the nanocantilever array itself;

[0051] A driving and reading circuit is connected to the U-shaped cantilever support arm of the nanocantilever array, and is used to drive the nanocantilever array to resonate and measure the change of its natural frequency during the gas adsorption process;

[0052] A signal processing unit, connected to the driving and reading circuits, is used to record the change signal of the natural frequency, perform data preprocessing and feature extraction, and then pattern recognition uses the calibrated data to perform data evaluation and output the results to the response and output unit;

[0053] The response and output unit is connected to the signal processing unit and is used to output the composition and specific concentration values ​​of the multi-component mixed gas.

[0054] In some embodiments of the present invention, the driving and reading circuit includes a first circuit and a second circuit; the first circuit is used to generate a reference signal Δω of the lock-in amplifier; the second circuit is used to generate a driving electrical signal ω of the nanocantilever beam array d And the output signal of the nanocantilever is sent to the lock-in amplifier for detection.

[0055] In some embodiments of the present invention, the first circuit includes an AC driving voltage, a first directional coupler, a frequency multiplier, an AC bias voltage, a second directional coupler, a mixer, a phase-locked amplifier, and a first filter; the AC driving voltage generates a radio frequency voltage after passing through the first directional coupler and the frequency multiplier, and the AC bias voltage outputs an AC voltage and a radio frequency voltage of a local oscillator frequency through the second directional coupler, enters the mixer, and outputs an intermediate frequency voltage that enters the phase-locked amplifier;

[0056] In some embodiments of the present invention, the second circuit includes an AC driving voltage, a first directional coupler, an AC bias voltage, a second directional coupler, a second filter and a frequency amplifier; the AC driving voltage passing through the first directional coupler and the AC bias voltage passing through the second directional coupler successively enter the second filter and the frequency amplifier and then enter the nanocantilever array, and the output signal of the nanocantilever array is input into the phase-locked amplifier through the second filter and the frequency amplifier for detection.

[0057] In some embodiments of the present invention, the signal processing unit includes a multi-layer perceptron quantization model, which includes multiple multi-input single-output perceptrons; each multi-input single-output perceptron includes an input layer, a hidden layer and an output layer; the model is used to perform back-propagation learning on a large data set between a sensor array data matrix and a target concentration matrix, and the learned parameters are stored in a weight matrix and a bias matrix; finally, based on these parameters, the cantilever beam resonant gas sensor can estimate the concentration of multiple unknown gas samples.

[0058] The technical solution of the present invention is further described below by means of specific embodiments in combination with the accompanying drawings. It should be noted that the following specific embodiments are only used as examples, and the protection scope of the present invention is not limited thereto.

[0059] This embodiment provides a nanoelectromechanical system resonant array gas sensor, including a nanocantilever array composed of nanocantilever beams, a driving and readout circuit, a signal processing unit, and a response and output unit.

[0060] Nanocantilever arrays are used to specifically adsorb target gas molecules and convert chemical signals into their own thermal vibration signals;

[0061] A driving and reading circuit is connected to the U-shaped cantilever support arm of the nanocantilever array, and is used to drive the nanocantilever array to resonate and measure the change of its natural frequency during the gas adsorption process;

[0062] A signal processing unit, connected to the driving and reading circuits, is used to record the signal, perform data preprocessing and feature extraction, and then pattern recognition uses the calibrated data for data evaluation and outputs the results to the response and output unit;

[0063] The response and output unit is connected to the signal processing unit and is used to output and display the composition and concentration values ​​of various detected gases.

[0064] like Figure 1 As shown, the nanocantilever array consists of several nanoscale cantilevers.

[0065] like Figure 2 As shown, the specific dimensions of the cantilever beam in this embodiment are: the total length l of the cantilever beam is 2.7 μm, the total width w is 0.8 μm, the support arm length l1 is 1.5 μm, and the width w1 is 0.2 μm. In some embodiments, the total length l is 2.0-3.4 μm, the total width w is 0.6-1.0 μm, the support arm length l1 is 1.2-1.8 μm, and the width w1 is 0.15-0.25 μm.

[0066] The cantilever beam adopts a two-arm support structure including a substrate 10, a support layer 11, a metal conductive layer 12, a sensitive layer 13, a cantilever beam support arm 14 and a connecting electrode 15. Compared with a rectangular cantilever beam of the same size (length 2.7 μm, width 0.8 μm), this structure not only has a larger deflection at the free end, but also has a smaller mass, which can effectively improve the sensitivity of a single micro cantilever beam.

[0067] The support layer 11 is made of low-stress Si3N4 with a thickness of 70 nm, serving as a vibrating cantilever beam of the resonator. In some embodiments, the thickness of the support layer 11 is 60-100 nm.

[0068] The metal conductive layer 12 is made of gold and has a thickness of 30 nm. It plays a dual role, serving as an etching mask for silicon nitride and as a piezoresistive strain sensor for displacement conduction of mechanical movement. In some embodiments, the thickness of the metal conductive layer is 20-40 nm.

[0069] The material used for the sensitive layer 13 is PDMS (polydimethylsiloxane), and any other thin film material sensitive to the gas to be detected can also be used to selectively adsorb the target gas;

[0070] Each cantilever beam acts as a sensitive unit and can be coated with different gas-sensitive materials, which can specifically respond to different gases. Therefore, the combination of multiple sensitive units achieves a fast and highly sensitive response to multiple gases.

[0071] like Figure 3 As shown, the driving and reading circuit in this embodiment uses a combination of thermoelastic driving and piezoresistive mixing. This circuit mainly consists of two loops. The first loop is mainly used to generate a reference signal Δω of the phase-locked amplifier. The AC driving voltage is ω d / 2 frequency is output to the first directional coupler, and then passes through the frequency multiplier to output the radio frequency (RF) voltage. b (ω b =ω d -Δω) outputs an AC voltage of the local oscillator (LO) frequency through a second directional coupler. The RF and LO are combined through a commercial mixer to output an intermediate frequency (IF) (<30MHz) voltage with a frequency of Δω as the reference signal for the phase-locked amplifier.

[0072] In the second loop, the AC driving voltage of the first directional coupler and the AC bias voltage of the second directional coupler pass through the high-pass second filter and the frequency amplifier and then are at the frequency ω dTemperature changes are generated at the cantilever beam, causing it to vibrate. Since the output signal of the nanocantilever beam array is small enough (<100kHz), the falling mixed frequency component will be attenuated to the minimum by the parallel capacitor. Then it is output to the first low-pass filter to remove the residual carrier and upper sideband, and then the signal is amplified by an amplifier with a magnification of 1000, and finally the signal IN is generated and sent to the phase-locked amplifier for detection.

[0073] It should be noted that in circuit design, a relatively large RF capacitor (C RF =6nF) and grounded to ensure that both the drive voltage and the bias voltage drop primarily across the metal loop of the resonator and not elsewhere in the circuit. In addition, the low-pass filter ensures that only the downmixed signal is transmitted to the low-noise amplifier instead of the RF drive and bias voltage.

[0074] The signal processing unit uses a multi-layer perceptron quantization (MLP) model combined with a back propagation algorithm (BP) to sense the concentrations of multiple gases and achieve accurate resolution and quantification. The model based on the multi-layer perceptron (MLP) is composed of multiple multi-input single-output (MMISO) MLPs. Each MLP consists of three parts: an input layer, a hidden layer, and an output layer. This model is used to perform back propagation learning on a large data set between the sensor array data matrix and the target concentration matrix, and the learned parameters are stored in a weight matrix and a bias matrix. Finally, based on these parameters, the cantilever beam resonant gas sensor can estimate the concentrations of multiple unknown gas samples.

[0075] Assuming that the concentration of q gases is estimated, q multi-input single-output MLP neural networks are required. For each individual MLP (MLP1, MLP2, ..., MLPq), it contains m input neurons and one output neuron. At the same time, the MLP of each gas is independently learned through the back-propagation algorithm: the input data set is placed in the input layer of the MLP, forward propagated to the output layer through the MLP, the output is compared with the required target set, and the error is calculated. These error sources are then back-propagated (through neuron tracing), and the weight factors and biases are corrected for each neuron connection. Repeat this training process until the output of each training input set converges to the corresponding target set. Through this learning method, the parameters obtained are the weight matrix connected between the input layer and the hidden layer, the bias matrix of the hidden layer, the weight matrix between the hidden layer and the output layer, and the bias matrix of the output layer. Based on these parameters, the sensor can estimate the concentration of multiple gas samples. Figure 4 A schematic diagram showing the structure of the MLP concentration estimation model for MMISO.

[0076] The process of measuring multi-component mixed gas by nanocantilever array resonant gas sensor is as follows: through sampling, filtering and pretreatment, the analyte is guided to the flow cell where the nanocantilever array is located. Due to the adsorption of specific gas molecules by the sensitive substance coated on the surface of the cantilever, the mass of the beam changes, causing the change of the natural frequency of the resonant beam. Then, the change of the natural frequency during the gas adsorption process is measured through the readout circuit. The detected electrical signal passes through the signal processing unit, and after data preprocessing and feature extraction, the measurement is evaluated using the calibrated data according to the pattern recognition algorithm, and the results are output to the response and output unit, which outputs the composition and concentration values ​​of various detected gases.

[0077] The preparation process of the nanoscale micro-cantilever resonant sensor in this embodiment is simpler than the traditional method of preparing micro-cantilever, and the specific steps are as follows:

[0078] like Figure 5a As shown, firstly, a double-sided polished SOI silicon wafer with an n-type doped (100) crystal plane is used as a substrate 10, and a silicon nitride film with a target thickness of 70 nm is formed on the silicon substrate 10 by a low pressure chemical vapor deposition (LPCVD) method to form a support layer 11;

[0079] like Figure 5b As shown, a metal conductive layer 12 with a thickness of 30 nm is formed on the support layer 11 by electron beam evaporation;

[0080] like Figure 5c As shown, the metal film on the surface of the support layer 11 is patterned using electron beam exposure (DBL) and lift-off processes;

[0081] like Figure 5d As shown, anisotropic plasma etching (DRIE) is used to etch the support layer 11 by using the upper patterned metal conductive layer 12 as a mask;

[0082] like Figure 5e As shown, a photoresist with a target thickness of about 1.5 μm is applied to the back of the silicon wafer. The photoresist is used as a mask to photoetch a pattern on the back that can form a micro-cantilever beam area. The micro-cantilever beam is released by isotropic silicon undercut etching;

[0083] like Figure 5f As shown, the released device is subjected to surface functionalization treatment, so that its surface is covered with a gas sensitive layer thin film material, such as polydimethylsiloxane (PDMS), as the sensitive layer 13. In order to improve the adhesion of the material film, hexamethyldisilane (HDMS) can be used to treat the chip surface first.

[0084] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A resonant array gas sensor, comprising: Nano cantilever array, used to specifically adsorb target gas molecules and convert chemical signals into their own thermal vibration signals; A driving and reading circuit is connected to the U-shaped cantilever support arm of the nanocantilever array, and is used to drive the nanocantilever array to resonate and measure the change of its natural frequency during the gas adsorption process; A signal processing unit, connected to the driving and reading circuits, is used to record the change signal of the natural frequency, perform data preprocessing and feature extraction, and then pattern recognition uses the calibrated data to perform data evaluation and output the results to the response and output unit; The response and output unit is connected to the signal processing unit and is used to output the composition and specific concentration values ​​of the multi-component mixed gas; Wherein, the nano cantilever beam array comprises a substrate and a cantilever beam array, and each cantilever beam of the cantilever beam array comprises: The support layer is arranged on the substrate, and one end in contact with the substrate forms a U-shaped cantilever beam support arm, and one end of the support layer away from the U-shaped cantilever beam support arm is suspended on the substrate; A metal conductive layer is disposed on the support layer; and The sensitive layer is arranged on the metal conductive layer at a side away from the supporting arm of the U-shaped cantilever beam, and a sensitive unit is arranged on the layer; The driving and reading circuit comprises a first circuit and a second circuit; the first circuit is used to generate a reference signal of the lock-in amplifier The second circuit is used to generate a driving electrical signal for the nano cantilever beam array. and sending the output signal of the nanocantilever to a lock-in amplifier for detection; The first circuit includes an AC driving voltage, a first directional coupler, a frequency multiplier, an AC bias voltage, a second directional coupler, a mixer, a phase-locked amplifier and a first filter; the AC driving voltage generates a radio frequency voltage after passing through the first directional coupler and the frequency multiplier, and the AC bias voltage outputs an AC voltage and a radio frequency voltage of a local oscillator frequency through the second directional coupler, enters the mixer, and outputs an intermediate frequency voltage that enters the phase-locked amplifier; The second circuit includes an AC driving voltage, a first directional coupler, an AC bias voltage, a second directional coupler, a second filter and a frequency amplifier; the AC driving voltage passing through the first directional coupler and the AC bias voltage passing through the second directional coupler sequentially enter the second filter and the frequency amplifier and then enter the nano cantilever array, and the output signal of the nano cantilever array is input into the phase-locked amplifier through the second filter and the frequency amplifier for detection.

2. The resonant array gas sensor according to claim 1, characterized in that: The material used for the support layer includes at least one of Si3N4 and 3C-SiC, and the thickness is 60 to 100 nm.

3. The resonant array gas sensor according to claim 1, characterized in that: The metal conductive layer is made of at least one of gold, platinum and copper, and has a thickness of 20 to 40 nm.

4. The resonant array gas sensor according to claim 1, characterized in that: Different sensitive units in the nanocantilever array are used to identify the same or different gases.

5. The resonant array gas sensor according to claim 1, characterized in that: The material used in the sensitive layer includes at least one of non-conductive polymer, zeolite, metal organic framework, calixarene and nano-composite material.

6. The resonant array gas sensor according to claim 1, characterized in that: The signal processing unit includes a multi-layer perceptron quantization model, which includes multiple multi-input single-output perceptrons; each multi-input single-output perceptron includes an input layer, a hidden layer and an output layer; the model is used to perform back-propagation learning on a large data set between a sensor array data matrix and a target concentration matrix, and the learned parameters are stored in a weight matrix and a bias matrix; finally, based on these parameters, the cantilever beam resonant gas sensor can estimate the concentration of multiple unknown gas samples.

7. The resonant array gas sensor according to claim 1, characterized in that: The nano cantilever array is prepared by the following method: preparing a support layer on a substrate; preparing a metal conductive layer on the support layer; preparing a cantilever beam pattern on the metal conductive layer; The substrate is etched to release the cantilever beam; A sensitive layer is prepared on the metal conductive layer to obtain the nano cantilever beam array.

Images