Gas sensing, storage and calculation integrated heterogeneous memristor for renal function detection and intelligent mask system
By constructing a heterogeneous memristor using MXene and NiMo-P heterostructures, gas sensing, storage, and computing functions are integrated into one device, solving the problems of high energy consumption, large latency, and difficulty in flexible integration of existing gas sensors. This enables low-power, low-latency real-time kidney function monitoring and early warning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-27
- Publication Date
- 2026-04-14
AI Technical Summary
Existing gas sensors are mostly based on a single material system and have not been effectively adapted to in-memory computing technology, resulting in high energy consumption and large latency. Furthermore, rigid substrates are difficult to meet the integration requirements of flexible wearables, which limits the application of real-time online health monitoring.
A heterojunction is constructed using MXene and NiMo-P composite materials to form a pn heterojunction, which integrates gas sensing, information storage and computing functions. Multimodal sensing and reconfigurable computing are achieved through the dynamic changes of the depletion layer at the heterojunction interface, and it is integrated on a flexible substrate.
It achieves low-power, low-latency real-time multimodal gas monitoring, supports early warning of kidney disease, has high sensitivity and selectivity, and is suitable for flexible wearable devices.
Smart Images
Figure CN121845585A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of flexible electronics and intelligent medical sensing technology, and in particular to a gas-sensing and storage-computing integrated heterogeneous memristor and intelligent mask system for kidney function detection. Background Technology
[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.
[0003] With the deep integration of artificial intelligence and Internet of Things technologies, intelligent health monitoring is rapidly developing towards miniaturization, low power consumption, and high integration. Against this backdrop, wearable breath analysis technology has attracted significant attention due to its potential for non-invasive and continuous monitoring.
[0004] Abnormal concentrations of ammonia (NH3) in exhaled human breath are important volatile biomarkers of kidney damage and failure. Real-time, dynamic monitoring of NH3 is expected to enable early warning and follow-up management of kidney diseases, which has significant clinical and health management implications.
[0005] However, while traditional breath detection methods (such as gas chromatography-mass spectrometry) offer high accuracy, they typically rely on large instruments, skilled operators, and controlled laboratory environments, limiting their widespread application. Furthermore, most existing gas sensors are based on single-material systems and have not yet effectively utilized the relaxation effect compatible with in-memory computing. Under the current von Neumann architecture, which separates sensing, storage, and computation, systems require frequent data migrations, leading to significant energy consumption and increased processing latency, making it difficult to meet the needs of routine, real-time online personal health monitoring.
[0006] To overcome these limitations, the "integrated sensing, storage, and computing" architecture has become an important development direction. Among them, gas memristors, as an emerging neuromorphic sensing device, can combine gas-sensitive characteristics with memristor behavior, realizing gas signal sensing, information storage, and in-situ computing simultaneously in a single device. This fundamentally avoids data transfer, greatly reduces system power consumption and response latency, and has significant innovation and technological advantages.
[0007] However, existing research on gas memristors is mostly limited to single gas detection or fixed computation modes, lacking the ability to fuse and sense multimodal dynamic signals and failing to achieve dynamic reconstruction of computational paradigms. Furthermore, most devices are based on rigid substrates, which makes it difficult to meet the requirements of flexibility, fit, and biocompatibility for flexible wearable integration, thus limiting their practical application in mobile health monitoring. Summary of the Invention
[0008] To address the aforementioned issues, this invention proposes a gas-sensing, storage-computing integrated heterojunction memristor and smart mask system for kidney function detection. The system uses a gas-sensing, storage-computing integrated heterojunction memristor as its core, constructing a heterojunction of MXene and NiMo-P composite materials. By manipulating the band structure through the heterojunction, a pure gas sensor is extended into a novel gas memristor with storage-computing capabilities. This emphasizes its innovative architecture, integrating multimodal sensing, information storage, and reconfigurable computing at a single device level. It is used for real-time monitoring of exhaled gases and early disease warning, fundamentally breaking through the traditional paradigm of separating sensing and computation; achieving real-time monitoring of exhaled gases and early warning for kidney function diseases.
[0009] In some implementations, the following technical solutions are adopted: A gas-sensing, storage, and computing integrated heterojunction memristor for kidney function detection comprises, from bottom to top: a flexible substrate, interdigitated electrodes, and an MXene@NiMo-P nanosphere film; wherein MXene is a p-type semiconductor and NiMo-P is an n-type semiconductor, forming a pn heterojunction. The interface depletion layer of the pn heterojunction widens when gas is detected, causing a change in the internal current; the width recovers when an electrical pulse is applied, enabling reconfigurable control of the internal current.
[0010] As a further embodiment, the method for preparing the MXene@NiMo-P nanoflower-shaped thin film includes: LiF was uniformly dispersed in HCl solution as a fluorine source, and Ti3AlC2 powder was added. After the reaction was completed, the mixture was centrifuged and washed multiple times. The resulting black precipitate was dried and ground to obtain monolayer MXene powder. The monolayer MXene powder was ultrasonically exfoliated in deionized water to prepare a monolayer MXene colloidal solution. The MXene colloidal solution was mixed with polymethyl methacrylate microspheres in a predetermined ratio. MXene nanosheets were spontaneously and uniformly coated on the surface of the polymethyl methacrylate microspheres by electrostatic adsorption. After stirring, centrifuging, washing, and drying the mixed solution, PMMA@MXene composite microspheres were obtained. The PMMA@MXene composite microspheres were uniformly dispersed in a mixed solvent composed of ethanol and ethylene glycol. Nickel nitrate hexahydrate, sodium molybdate dihydrate, and urea were added sequentially as reaction precursors. The reaction precursors were subjected to ultrasonic and magnetic stirring and then transferred to a high-pressure reactor lined with polytetrafluoroethylene for reaction. After the reaction was completed, the MXene@NiMo-LDH intermediate was obtained by cooling, centrifugation, washing, and drying. The MXene@NiMo-LDH intermediate and sodium hypophosphite were placed in two ceramic boats respectively, and phosphating was carried out under continuous argon gas protection by heating, finally obtaining the MXene@NiMo-P composite material with a hollow nanoflower structure.
[0011] As a further approach, the MXene@NiMo-P composite material with hollow nanoflower-shaped structure is uniformly dispersed in anhydrous ethanol to form a slurry. The slurry is then deposited onto a pre-patterned flexible interdigitated electrode using a spin coating method. The coated film is then heat-treated to remove residual solvent and enhance the film-substrate adhesion, ultimately forming an MXene@NiMo-P composite thin film memristor.
[0012] In other embodiments, the following technical solutions are adopted: A reconfigurable neuromorphic olfactory sensing system, comprising: The flexible sensing computing unit, including the aforementioned heterogeneous memristor, is used to sense exhaled gas signals and output them in the form of analog current signals. A signal conditioning and encoding unit, connected to the sensing and computing unit, is used to receive the analog current signal output by the sensing and computing unit, preprocess it, and encode it into a pulse sequence. The neuromorphic computing unit, connected to the signal conditioning and encoding unit, includes a reservoir computing network and a spiking neural network. The neuromorphic computing unit receives the pulse sequence, identifies the gas type through the spiking neural network, and then identifies the gas concentration through the reservoir computing network. The decision-making and execution unit, connected to the neuromorphic computing unit, is used to generate early warning signals based on the gas type and gas concentration identification results, and transmit them to the user terminal via Bluetooth module.
[0013] As a further approach, in gas sensing mode, gas molecules interact with the pn heterojunction in the heterojunction memristor, altering the heterojunction's band structure and causing analog, non-volatile, or volatile changes in the device's conductance. This is used to simulate the short-term and long-term plasticity of biological synapses, at which point the memristor exhibits synaptic characteristics under gas stimulation. After sensing pulse sequence signals of different gases, the gas type is identified using a pulse neural network based on the different waveforms of the pulse sequences of different types of gases. For the identified gas types, the gas concentration is further identified using a reservoir computing network based on the different current change ranges of different gas concentrations.
[0014] As a further approach, in electrical processing mode, the applied voltage pulse can cause the pn heterojunction band structure to reset.
[0015] As a further measure, if the concentration of the exhaled gas exceeds a set threshold, an alarm signal is triggered.
[0016] In other embodiments, the following technical solutions are adopted: A smart mask system includes: a breathing box, the breathing box including a reconfigurable neuromorphic olfactory sensing system as described in any one of claims 4-7; the breathing box is embedded in the interlayer of the mask to ensure that the sensing system is exposed to the path of the exhaled airflow, and the breathing box is provided with a one-way inlet valve and a one-way outlet valve to isolate external environmental interference.
[0017] As a further option, it also includes: The intelligent terminal communicates with the breathing box and is used to display gas detection results and receive alarm signals.
[0018] In other embodiments, the following technical solutions are adopted: Application of a reconfigurable neuromorphic olfactory sensing system in personal health monitoring, early warning of renal failure, or screening for respiratory diseases.
[0019] Compared with the prior art, the beneficial effects of the present invention are: (1) This invention overturns the von Neumann architecture followed by traditional electronic olfactory systems, integrating the three major functions of gas sensing, information processing, and storage into a single MXene@NiMo-P heterostructure memristor. As the core sensing and computing unit, this device can dynamically reconstruct its functional role according to the type of input signal (gas stimulation or electrical pulse stimulation): in gas mode, it exhibits synaptic characteristics and supports reservoir computing; in electrical pulse mode, it exhibits neuronal characteristics and supports spiking neural network computing. This on-demand reconstructing at the hardware level fundamentally eliminates the energy consumption and delay caused by data transfer between separate units, realizing low-power, low-latency in-sensor computing.
[0020] (2) This invention seamlessly integrates two advanced neuromorphic computing paradigms, reservoir computing and spiking neural networks, on the same hardware platform, and achieves integrated sensing, storage, and computing within the sensor, greatly reducing the computational load on the microcontroller and realizing the symbiosis of the two neural networks. Among them, the reservoir computing network is good at handling the spatiotemporal characteristics of gas concentration, while the spiking neural network is good at identifying gas types. This hybrid neural network architecture enables the system to process multimodal gas information (such as gas type and concentration) in parallel and efficiently, providing more comprehensive and accurate environmental perception and intelligent analysis capabilities, and greatly enhancing the system's flexibility and adaptability when dealing with complex and variable gas environments.
[0021] (3) The core sensitive material of this invention adopts a uniquely designed hollow nanoflower-shaped MXene@NiMo-P heterojunction. This structure not only greatly increases the specific surface area and improves the adsorption efficiency of gas molecules (such as NH3), but also allows the depletion layer width at the heterojunction interface to dynamically change with gas adsorption / desorption or electrical pulse stimulation, thereby achieving reconfigurable control of the conductive state. Through band engineering and phosphating optimization, the material possesses excellent charge transport dynamics, which lays a solid physical foundation for the device to achieve high sensitivity, high selectivity, and reliable dual-mode reconfiguration characteristics.
[0022] (4) This invention successfully integrates a reconfigurable neuromorphic olfactory sensing system onto a flexible wearable mask platform, constructing a complete "perception-decision-execution" application closed loop. Experiments have shown that the system can still work stably in a bent state and can display gas concentration, risk level, and issue warnings in real time via a mobile terminal. This plug-and-play integrated design realizes a complete function from high-risk gas detection to early warning of diseases (such as kidney failure), providing a highly practical and innovative solution for fields such as smart healthcare, personal health monitoring, and industrial safety.
[0023] Other features and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description
[0024] Figure 1 This is a physical image of the intelligent mask system in an embodiment of the present invention; Figure 2 This is a schematic diagram of the breathing valve control on the mask in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the preparation process of MXene@NiMo-P nanoflower-shaped materials in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the fabrication process of the reconfigurable neuromorphic olfactory sensing system in an embodiment of the present invention. Figure 5 This is a schematic diagram of the structural composition of the intelligent mask system in an embodiment of the present invention; Figure 6 The image shown is the XRD pattern of the MXene@NiMo-P composite material in the embodiment of the present invention. Figure 7 Here is a SEM image of the MXene@NiMo-P composite material in an embodiment of the present invention; Figure 8 This is an EDS image of the MXene@NiMo-P composite material in an embodiment of the present invention; Figure 9This is the synaptic response recovery time curve of the device in gas mode in an embodiment of the present invention; Figure 10 This is a curve showing the synaptic double-pulse facilitation behavior of the device in gas mode in an embodiment of the present invention; Figure 11 The LTP-LTD curves of the device in gas mode at different times in the embodiments of the present invention; Figure 12 These are the pulse frequency curves of the device in gas mode in the embodiments of the present invention; Figure 13 These are test curves of the device in gas mode in embodiments of the present invention; Figure 14 This is a binary test curve of the device in gas mode in an embodiment of the present invention. Figure 15 The neuronal characteristics of the device in the electrical pulse mode in the embodiments of the present invention; Figure 16 This is a schematic diagram of a neural network in an embodiment of the present invention; Figure 17 This is a schematic diagram of the mobile terminal display interface in an embodiment of the present invention. Detailed Implementation
[0025] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.
[0026] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.
[0027] Example 1 In one or more embodiments, a gas-sensing, storage, and computing integrated heterojunction for renal function detection is disclosed. The heterojunction comprises, from bottom to top, a flexible substrate, interdigitated electrodes, and an MXene@NiMo-P nanosphere film. MXene is a p-type semiconductor, and NiMo-P is an n-type semiconductor, forming a pn heterojunction. The width of the depletion layer at the interface of the pn heterojunction dynamically changes with gas adsorption or electrical stimulation. That is, when gas is detected, the width of the depletion layer widens, thereby causing a change in the internal current. When an electrical pulse is applied, the width recovers, achieving reconfigurable control of the internal current.
[0028] Combination Figure 3 The preparation process of MXene@NiMo-P nanoflower-shaped thin films includes: Step 1: Prepare monolayer MXene material; Specifically, LiF (lithium fluoride) was uniformly dispersed as a fluorine source in a 9 mol / L HCl (hydrogen chloride or hydrochloric acid) solution, and continuously stirred to form a homogeneous etching agent. Subsequently, under ice-water bath protection, Ti3AlC2 powder, a MAX phase ceramic precursor, was slowly added to the mixture, with the mass ratio of LiF to Ti3AlC2 precisely controlled at 1:1. The mixed reaction system was transferred to a 30-50℃ constant temperature water bath and mechanically stirred continuously for 24 hours under inert gas protection to fully etch and remove the Al atomic layer and obtain a monolayer MXene (Ti3C2T). x ).
[0029] After the reaction, the product was washed repeatedly by centrifugation with a large amount of deionized water until the pH of the supernatant was close to neutral to remove residual acid and ionic byproducts. Finally, the resulting black precipitate was dried in a vacuum drying oven at 60°C for 12 hours, and then ground to obtain a multilayer MXene powder suitable for subsequent experiments.
[0030] As an alternative implementation method, Ti3AlC2 powder is added and the reaction is carried out in a water bath at 40°C.
[0031] Step 2: Prepare 3D MXene nanospheres; The multilayer MXene powder obtained in step 1 was ultrasonically exfoliated in deionized water to obtain a uniformly dispersed monolayer MXene colloidal solution (1 mg / mL). Subsequently, monodisperse polymethyl methacrylate (PMMA) microspheres (approximately 800 nm in diameter) were mixed with the above MXene colloidal solution at a mass ratio of 9-11:1. As an alternative embodiment, PMMA template microspheres were mixed with multilayer MXene at a mass ratio of 10:1.
[0032] Leveraging the potential difference between the abundant negatively charged functional groups (-O, -OH) on the MXene surface and the PMMA microsphere surface, MXene nanosheets spontaneously and uniformly coat the PMMA microsphere surface through electrostatic adsorption. After slowly stirring the mixed solution for 12 hours, the composite microspheres were collected by centrifugation (3500 rpm, 5 min) and washed several times with anhydrous ethanol to remove physically adsorbed MXene fragments. Finally, the obtained PMMA@MXene composite microspheres were vacuum-dried overnight at 60 °C for subsequent use.
[0033] In this embodiment, PMMA was used as a template to fabricate MXene into three-dimensional nanospheres, which increased the surface area and facilitated gas adsorption reactions.
[0034] Step 3: Synthesize MXene@NiMo-P composite material; 50 mg of the PMMA@MXene composite microspheres prepared in step 3 were uniformly dispersed in 50 mL of a mixed solvent (volume ratio 1:1) of ethanol and ethylene glycol. Subsequently, 2 mmol of nickel nitrate hexahydrate (Ni(NO3)2·6H2O), 2 mmol of sodium molybdate dihydrate (Na2MoO4·2H2O), and 0.3 g of urea were added sequentially as reaction precursors. Urea served both as an alkali source to promote the growth of layered hydrogen peroxide (LDH) and as a structure-directing agent in the reaction. After forming a homogeneous precursor solution through ultrasonic and magnetic stirring for 30 minutes, it was transferred to a 100 mL high-pressure reactor lined with polytetrafluoroethylene (PTFE) and reacted in a forced-air drying oven at 100-140 °C for 4 hours. Alternatively, after transferring to the high-pressure reactor, the reaction was carried out in a forced-air drying oven at 120 °C.
[0035] After the reaction, the mixture was naturally cooled to room temperature. The resulting product was centrifuged and washed several times with alternating layers of deionized water and anhydrous ethanol. It was then vacuum-dried at 60°C for 12 hours to obtain the MXene@NiMo-LDH intermediate. Phosphating was subsequently performed to optimize the material's conductivity and catalytic activity. The obtained MXene@NiMo-LDH and 150 mg of sodium hypophosphite (NaH2PO2) were placed in two ceramic boats, with the NaH2PO2 positioned upstream of a tube furnace. Under a continuous argon flow (50 sccm), the furnace temperature was increased to 400°C-500°C (for example, 450°C) at a rate of 5°C / min and maintained for 2 hours for the phosphating reaction. During this process, the PMMA template was completely decomposed and removed, and NiMo-LDH was partially converted into phosphide (NiMo-P), ultimately yielding the MXene@NiMo-P composite material with a hollow nanofloral structure.
[0036] Combination Figure 4 The MXene@NiMo-P nanosphere powder synthesized in step 3 was dispersed in anhydrous ethanol to prepare a homogeneous slurry with a concentration of 5 mg / mL. This slurry was then deposited onto pre-patterned flexible interdigitated electrodes (interpolation spacing of 10-50 nm) using spin coating: first, the slurry was spin-coated at 1000 rpm for 10 seconds to spread, then the spin coating speed was increased to 3000 rpm for 30 seconds to form a uniformly thick (approximately 2-3 μm) sensitive film. Subsequently, the coated film was heat-treated at 80°C for 1 hour to remove residual solvent and enhance the film-substrate adhesion, ultimately forming an MXene@NiMo-P composite thin-film memristor as the core sensing and computing unit.
[0037] As an alternative example, the spin coating speed is 3000 rpm and the spin coating time is 30 seconds.
[0038] Example 2 Based on the gas-sensing, storage-computing, and heterogeneous memristor for kidney function detection disclosed in Embodiment 1, a reconfigurable neuromorphic olfactory sensing system is further disclosed, specifically including: (1) Flexible sensing computing unit, including the MXene@NiMo-P heterostructure memristor disclosed in Example 1, used to sense exhaled gas signals and output them in the form of analog current signals; Among them, the MXene@NiMo-P heterostructure memristor exhibits synaptic characteristics under gas stimulation and neuronal characteristics under electrical pulse stimulation (electrical pulse stimulation is formed by applying different voltages to the device in the form of pulses).
[0039] (2) Signal conditioning and encoding unit, connected to the sensing and computing unit, is used to receive the analog current signal output by the sensing and computing unit, preprocess it and encode it into a pulse sequence; the purpose of encoding it into a pulse sequence is to resist interference and transmit it over long distances, convert the analog signal into a digital pulse sequence on-site, and then transmit it to the neuromorphic computing unit through the circuit.
[0040] (3) A neuromorphic computing unit, connected to a signal conditioning and coding unit, includes a reservoir computing network and a spiking neural network. The neuromorphic computing unit receives the pulse sequence, identifies the gas type through the spiking neural network, and then identifies the gas concentration through the reservoir computing network. (4) Decision and execution unit, connected to neuromorphic computing unit, is used to generate early warning signals based on the gas type and gas concentration identification results, and transmit them to the user terminal via Bluetooth module.
[0041] In this embodiment, the working mode of the reconfigurable neuromorphic olfactory sensing system can be dynamically switched, specifically including: In gas sensing mode, gas molecules interact with the pn heterojunction in the heterojunction memristor, changing the band structure of the heterojunction and causing analog, non-volatile, or volatile changes in the device's conductance. This is used to simulate the short-term and long-term plasticity of biological synapses. At this time, the memristor exhibits synaptic characteristics under gas stimulation. After sensing the pulse sequence signals of different gases, the pulse sequence waveforms of different types of gases are different, and the gas type is identified using a pulse neural network, outputting the current characteristic bending curve. For the identified gas type, the spatiotemporal characteristics of the gas concentration are further identified using a reservoir computing network based on the different current change ranges of different gas concentrations.
[0042] In the electrical processing mode, applying a voltage pulse can cause the heterojunction band structure to reset.
[0043] It should be noted that both spiking neural networks and reservoir computing networks are existing network structures, and they need to be trained with a large number of gases of known types and concentrations before they can be applied.
[0044] This embodiment integrates gas sensing, information processing, and storage functions into a single MXene@NiMo-P heterostructure memristor. The combination of MXene and NiMo-P is chosen primarily to construct a composite sensing material with a dynamically tunable pn heterojunction, thereby achieving dual-mode reconstruction of gas sensing and neuromorphic computing in a single device. MXene, as a p-type substrate, provides a high specific surface area and a good conductive network; NiMo-P, as an n-type active material, possesses excellent gas adsorption and catalytic properties. The depletion layer at the heterojunction interface formed by these two materials dynamically changes with gas adsorption or electrical pulse stimulation, enabling the device to exhibit synaptic plasticity in gas mode and neuronal threshold excitation characteristics in electrical pulse mode. This structure not only significantly improves the sensitivity and selectivity of gas detection but also lays the material foundation for realizing a low-power, low-latency olfactory sensing system integrating sensing, storage, and computing.
[0045] As the core sensing and computing unit, this device can dynamically reconstruct its functional role based on the type of input signal (gas stimulation or electrical pulse stimulation): in gas mode, it exhibits synaptic characteristics, supporting reservoir computing; in electrical pulse mode, it exhibits neuron characteristics, supporting spiking neural network computing. This on-demand reconstructing at the hardware level fundamentally eliminates the energy consumption and latency caused by data transfer between separate units, achieving low-power, low-latency in-sensor computing.
[0046] Based on this, as a further implementation method, a smart mask system is disclosed, combining... Figure 1 , Figure 2 and Figure 5 The system includes a breathing box containing the aforementioned sensing system, which is reliably connected to a custom-designed micro-signal processing circuit board via conductive silver paste. This circuit board includes a microcontroller (MCU), a Bluetooth Low Energy (BLE) communication chip, a micro-lithium battery, and a smart terminal (smartphone). The microcontroller is connected to both the sensing system and the micro-lithium battery. The microcontroller also communicates with the smart terminal via the Bluetooth Low Energy communication chip. The sensing system transmits the gas type and concentration detection results to the microcontroller, which can then provide early warnings for diseases such as kidney failure based on the detection results, and send the warning results to the smart terminal.
[0047] The entire system is cleverly embedded in the interlayer of a special mask, ensuring that the sensing unit is exposed to the path of the exhaled airflow. To prevent leakage during exhalation and facilitate inhalation, one-way inlet and outlet valves are prepared on the surface of the mask to control the exhalation and inhalation effects. Figure 2 The left side is the breathing valve state when it is straight and not exhaling, while the right side is curved and is the breathing valve state when exhaling. The baffle can effectively prevent other gases besides exhaled gas from contacting the sensor.
[0048] In this embodiment, to demonstrate the crystallization effect of the MXene@NiMo-P composite material, XRD (X-ray diffraction) testing was performed, and the test results are as follows: Figure 6 As shown, this demonstrates the successful synthesis of the MXene@NiMo-P composite material.
[0049] To demonstrate the microstructure of the MXene@NiMo-P composite material, SEM (scanning electron microscopy) tests were performed. The test results are as follows: Figure 7 As shown, the MXene@NiMo-P composite material consists of microspheres with uniformly sized nanosheets loaded on its surface. The results of SEM and EDS (energy dispersive spectroscopy) tests were compared. Figure 8 As shown in the figure, the nanosheet structure NiMo-P is NiMo-P nanosheet.
[0050] To demonstrate the response of the MXene@NiMo-P sensor Restore performance and respond. Resume testing. Test results are as follows: Figure 9 As shown, the response time is 1.1 s and the recovery time is 6.9 s, which means that the MXene@NiMo-P sensor has the characteristics of fast response and recovery.
[0051] An experiment on gas synaptic characteristic detection based on the MXene@NiMo-P sensor was conducted. The specific test conditions were: ammonia pulse detection experiment with ammonia gas at a concentration of 100 ppm in a gas chamber at a temperature of 25℃ and a humidity of 40%. Figure 10 As shown, the device exhibits facilitated behavior under synaptic double-pulse conditions in gas mode. Figure 11 As shown, through multiple gas pulse tests, the synaptic properties transition from short-term plasticity to long-term plasticity with increasing number of tests. The results of pulse tests at different frequencies and with different gas concentrations are shown below. Figure 12 and Figure 13 As shown, the peak current increases while the decay rate decreases with increasing frequency and concentration.
[0052] An experiment was conducted to train a gas binary 4-bit code based on an MXene@NiMo-P sensor. The specific test conditions were: ammonia pulse detection experiments were performed in a gas chamber with a concentration of 100 ppm ammonia at 25℃ and 40% humidity. The applied and stopped pulses were defined as "1" and "0" in binary to simulate a time-series test. The test results are as follows: Figure 14 As shown.
[0053] An experiment was conducted to detect the electrical neuron characteristics based on the MXene@NiMo-P sensor. The specific test conditions were: pulse detection with a pulse width of 0.3 seconds using a 10V current. Figure 15 As shown, after multiple electrical pulse tests, the device exhibits neuron-like rapid response characteristics. In continuous testing, when the current signal exceeds a certain threshold, it simulates the threshold firing characteristics of a neuron.
[0054] To enhance the system's intelligence and accuracy, the above test results were combined with a spiking neural network and a reservoir computing network. The neural network diagram is shown below. Figure 16 As shown, when gas is detected, the spiking neural network can quickly distinguish the type of gas, and the reservoir computing network can effectively identify the gas concentration.
[0055] To more clearly define the system's feedback when it detects exhaled gas signals Figure 17 This demonstrates an example of how the system displays an interface on a mobile device via Bluetooth when gas is detected.
[0056] In further embodiments, a reconfigurable neuromorphic olfactory sensing system as described above is also provided for monitoring human exhaled gases, combining Bluetooth communication with a mobile terminal, and realizing early warning of diseases (such as kidney failure), applicable to fields such as personal health monitoring, smart healthcare and remote diagnosis.
[0057] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.
Claims
1. A gas-sensing, storage-computing, heterogeneous memristor for kidney function detection, characterized in that, The heterojunction memristor comprises, from bottom to top, a flexible substrate, interdigitated electrodes, and an MXene@NiMo-P nanosphere film; wherein MXene is a p-type semiconductor and NiMo-P is an n-type semiconductor, forming a pn heterojunction. The interface depletion layer of the pn heterojunction widens when gas is detected, causing a change in the internal current; the width recovers when an electrical pulse is applied, enabling reconfigurable control of the internal current.
2. The gas-sensing, storage-computing, heterogeneous memristor for kidney function detection as described in claim 1, characterized in that, The preparation method of the MXene@NiMo-P nanoflower-shaped thin film includes: LiF was uniformly dispersed in HCl solution as a fluorine source, and Ti3AlC2 powder was added. After the reaction was completed, the mixture was centrifuged and washed multiple times. The resulting black precipitate was dried and ground to obtain monolayer MXene powder. The monolayer MXene powder was ultrasonically exfoliated in deionized water to prepare a monolayer MXene colloidal solution. The MXene colloidal solution was mixed with polymethyl methacrylate microspheres in a predetermined ratio. MXene nanosheets were spontaneously and uniformly coated on the surface of the polymethyl methacrylate microspheres by electrostatic adsorption. After stirring, centrifuging, washing, and drying the mixed solution, PMMA@MXene composite microspheres were obtained. The PMMA@MXene composite microspheres were uniformly dispersed in a mixed solvent composed of ethanol and ethylene glycol. Nickel nitrate hexahydrate, sodium molybdate dihydrate, and urea were added sequentially as reaction precursors. The reaction precursors were subjected to ultrasonic and magnetic stirring and then transferred to a high-pressure reactor lined with polytetrafluoroethylene for reaction. After the reaction was completed, the MXene@NiMo-LDH intermediate was obtained by cooling, centrifugation, washing, and drying. The MXene@NiMo-LDH intermediate and sodium hypophosphite were placed in two ceramic boats respectively, and phosphating was carried out under continuous argon gas protection by heating, finally obtaining the MXene@NiMo-P composite material with a hollow nanoflower structure.
3. A gas-sensing, storage-computing, heterogeneous memristor for kidney function detection as described in claim 2, characterized in that, The MXene@NiMo-P composite material with hollow nanoflower-shaped structure was uniformly dispersed in anhydrous ethanol to form a slurry. The slurry was then deposited onto a pre-patterned flexible interdigitated electrode using a spin coating method. The coated film was then heat-treated to remove residual solvent and enhance the film-substrate adhesion, ultimately forming an MXene@NiMo-P composite thin film memristor.
4. A reconfigurable neuromorphic olfactory sensing system, characterized in that, include: A flexible sensing computing unit, including the heterogeneous memristor as described in any one of claims 1-3, is used to sense exhaled gas signals and output them in the form of analog current signals. A signal conditioning and encoding unit, connected to the sensing and computing unit, is used to receive the analog current signal output by the sensing and computing unit, preprocess it, and encode it into a pulse sequence. The neuromorphic computing unit, connected to the signal conditioning and encoding unit, includes a reservoir computing network and a spiking neural network. The neuromorphic computing unit receives the pulse sequence, identifies the gas type through the spiking neural network, and then identifies the gas concentration through the reservoir computing network. The decision-making and execution unit, connected to the neuromorphic computing unit, is used to generate early warning signals based on the gas type and gas concentration identification results, and transmit them to the user terminal via Bluetooth module.
5. A reconfigurable neuromorphic olfactory sensing system as described in claim 4, characterized in that, In gas sensing mode, gas molecules interact with the pn heterojunction in the heterojunction memristor, changing the band structure of the heterojunction and causing analog, non-volatile, or volatile changes in the device's conductance. This is used to simulate the short-term and long-term plasticity of biological synapses. At this time, the memristor exhibits synaptic characteristics under gas stimulation. After sensing the pulse sequence signals of different gases, the gas type is identified using a pulse neural network based on the different pulse sequence waveforms of different types of gases. For the identified gas types, the gas concentration is further identified using a reservoir computing network based on the different current change ranges of different gas concentrations.
6. A reconfigurable neuromorphic olfactory sensing system as described in claim 4, characterized in that, In the electrical processing mode, the applied voltage pulse can cause the pn heterojunction band structure to reset.
7. A reconfigurable neuromorphic olfactory sensing system as described in claim 4, characterized in that, If the concentration of the exhaled gas exceeds a set threshold, an alarm signal is triggered.
8. A smart mask system, characterized in that, include: A breathing box, comprising the reconfigurable neuromorphic olfactory sensing system according to any one of claims 4-7; the breathing box is embedded in the interlayer of a mask to ensure that the sensing system is exposed to the path of exhaled airflow, and the breathing box is provided with a one-way inlet valve and a one-way outlet valve to isolate external environmental interference.
9. The intelligent mask system as described in claim 8, characterized in that, Also includes: The intelligent terminal communicates with the breathing box and is used to display gas detection results and receive alarm signals.
10. The application of a reconfigurable neuromorphic olfactory sensing system as described in any one of claims 4-6 in personal health monitoring, early warning of renal failure, or screening for respiratory diseases.