Memristor for adenovirus concentration detection and preparation method thereof

Abstract

The invention belongs to the technical field of external diagnosis devices, and particularly relates to a memristor for adenovirus concentration detection and a preparation method of the memristor. The memristor comprises a bottom electrode layer, a metal oxide layer, a functional layer and an upper electrode layer which are sequentially formed from bottom to top in the thickness direction, the bottom electrode layer is an FTO conductive glass layer, the metal oxide layer is tantalum pentoxide, the functional layer is a gelatin-beta cyclodextrin composite film, and the upper electrode layer is a conductive electrode layer. The gelatin-beta cyclodextrin composite film with the memristive effect serves as a functional layer, after the film makes contact with adenoviruses, the film adsorbs virus particles, then the unique resistance change characteristic of the memristor is caused, the memristive behavior of the memristor is further fed back, and the memristive effect of the memristor is improved. The method is used for recording and analyzing the electrochemical or bio-electricity signal characteristics of adenovirus solutions with different concentrations, the adenovirus solutions with different concentrations are effectively distinguished according to the electric signal difference of the adenovirus solution and the control solution, the effect is obvious, and the repeatability is good.

Description

Technical Field

[0001] This invention belongs to the field of external diagnostic device technology, specifically relating to a memristor for adenovirus concentration detection and its preparation method. Background Technology

[0002] Adenoviruses are a class of non-enveloped viruses with double-stranded DNA genomes, widely distributed in nature, capable of infecting various tissues including the respiratory, digestive, urinary, and central nervous systems. Especially in children, immunocompromised individuals, and critically ill patients, adenovirus infection can cause severe complications such as pneumonia, meningitis, encephalitis, and myocarditis, even leading to death. In recent years, with the emergence of viral variants and the diversification of transmission routes, the harm caused by adenoviruses in public health emergencies, hospital-acquired infections, and neurological diseases has become increasingly prominent, posing higher demands on clinical diagnosis and prevention.

[0003] Currently, the mainstream detection methods for adenovirus include nucleic acid detection (such as PCR and RT-qPCR), immunological detection (such as ELISA and lateral chromatography strips), and viral culture. While PCR-based detections are highly sensitive, they require highly specialized equipment, operating environments, and skilled personnel, and are time-consuming and costly. Immunological methods, although simple and portable, suffer from insufficient sensitivity, false negatives, and the inability to perform quantitative analysis. Viral culture methods take several days, making them unsuitable for rapid clinical diagnosis and emergency control. Therefore, existing detection technologies struggle to simultaneously meet the combined demands of high sensitivity, speed, portability, and low cost. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a memristor for adenovirus concentration detection and its preparation method. This invention utilizes a tantalum pentoxide (Ta2O5) and gelatin-β-cyclodextrin composite film with memristive effects as the functional layer. Since β-cyclodextrin can adsorb adenovirus particles, the memristor can interact with biomolecules (such as antibodies). Through biofunctionalization, the specific binding between antibodies and antigens can cause changes in the memristor's resistance, further reflecting its memristive behavior. This is used to record and analyze the electrochemical or bioelectrical signal characteristics of tumor cells. Based on the differences in electrical signals between tumor cells and normal cells, different types of tumor cells can be effectively distinguished, with significant effects and good reproducibility.

[0005] The present invention solves the above-mentioned technical problems through the following technical solutions.

[0006] The first objective of this invention is to provide a memristor for adenovirus concentration detection, wherein the memristor comprises, from bottom to top, a bottom electrode layer, a metal oxide layer, a functional layer and a top electrode layer along the thickness direction. The bottom electrode layer is an FTO conductive glass layer, the metal oxide layer is tantalum pentoxide, the functional layer is a gelatin-β-cyclodextrin composite film, and the top electrode layer is a conductive electrode layer.

[0007] Furthermore, the thickness of the gelatin-β-cyclodextrin composite film is 50 nm to 100 nm.

[0008] Furthermore, the thickness of the metal oxide layer is 300 nm to 350 nm.

[0009] Furthermore, the bottom electrode layer is an FTO electrode layer with a thickness of 150nm to 200nm; the top electrode layer is an Ag electrode layer with a thickness of 100nm to 150nm.

[0010] A second objective of this invention is to provide a method for preparing the memristor for adenovirus concentration detection described above, comprising the following steps: S1. A metal oxide layer is deposited on the bottom electrode substrate by magnetron sputtering.

[0011] S2. A gelatin-β-cyclodextrin composite film functional layer is coated onto the metal oxide layer using a spin coating method.

[0012] S3. Cover the functional layer with a metal mask with openings, and deposit an electrode layer on the metal mask by DC sputtering to obtain a memristor for adenovirus concentration detection.

[0013] Furthermore, in the magnetron sputtering process, the target-substrate distance is 9cm to 11cm, and the vacuum level is 3×10⁻⁶. -4 Pa~4×10 -4 Pa, working gas is argon, sputtering pressure is 0.7 Pa to 0.8 Pa, sputtering power is 120 W / cm². 2 The sputtering time is 40 min to 90 min.

[0014] Furthermore, during the spin coating process, a gelatin-β-cyclodextrin precursor solution is dropped onto the surface of the metal oxide layer, with a drop volume of 120 μL to 300 μL. The coating is first performed at 400 rpm to 600 rpm for 15 to 25 seconds, and then at 1800 rpm to 2400 rpm for 25 to 35 seconds. After coating, the film is cross-linked with glutaraldehyde vapor for 7 to 9 hours to enhance the structural stability of the gelatin-β-cyclodextrin composite film.

[0015] Furthermore, the preparation method of the gelatin-β-cyclodextrin precursor solution includes the following steps: Add water to gelatin and stir until completely dissolved. Then add β-cyclodextrin and stir until completely dissolved. Let stand overnight to obtain a gelatin-β-cyclodextrin precursor solution. The mass ratio of gelatin to β-cyclodextrin is 1:0.4-0.5, and the volume ratio of gelatin to water is 0.4g-0.5g:10mL.

[0016] Furthermore, in the DC sputtering process, the target-substrate distance is 9cm to 11cm, and the vacuum level is 2×10⁻⁶. -4 Pa~3×10 -4 Pa, working gas is argon, sputtering pressure is 0.7 Pa to 0.8 Pa, sputtering power is 60 W / cm². 2 The sputtering time is 15 min to 25 min.

[0017] Furthermore, the metal mask has an array of openings, and the upper electrode layer is deposited on all the openings.

[0018] Compared with the prior art, the present invention has the following advantages: (1) This invention comprises a bottom electrode layer, a metal oxide layer, a functional layer, and a top electrode layer arranged sequentially from bottom to top. The metal oxide layer is tantalum pentoxide (Ta2O5), the functional layer is a gelatin-β-cyclodextrin composite film, and the top electrode layer is a conductive electrode layer. Ta2O5, as a high dielectric constant material, has abundant oxygen vacancies and strong ion migration channels. When a positive bias voltage is applied, the top electrode Ag is electrochemically oxidized to Ag. + Ions migrate downwards under the influence of an electric field, accumulating along oxygen vacancy paths in Ta₂O₅ and gradually reducing to Ag atoms, forming locally conductive filaments that penetrate the oxide layer. This allows the device to transition from a high-resistivity (HRS) state to a low-resistivity (LRS) state. Reverse bias or localized thermal effects can cause the filament neck to break or ion diffusion to dissolve back, restoring the device to a high-resistivity state and exhibiting typical reversible resistive switching characteristics. In this process, the current conduction mechanism is dominated by Schottky emission in the low-field region, Hopping and Fowler-Nordbeim emission in the medium-to-high-field region, and metallic conductivity in the on-state. The gelatin-β-cyclodextrin composite film plays a crucial regulatory and enhancing role in this system; its abundant hydroxyl and amino groups can modulate the polarization environment of the Ta₂O₅ surface and capture or release Ag. + This effectively stabilizes the formation and breakage process of conductive filaments. Simultaneously, the hydrophobic cavity structure of β-cyclodextrin can form a strong non-specific binding with the hydrophobic regions on the virus surface, causing changes in the interfacial charge distribution and local electric field response of the device with varying virus concentrations, thus leading to detectable changes in the memristor's conductance. Higher virus concentrations result in more viral proteins adsorbed at the gelatin-β-cyclodextrin composite film interface, significantly enhancing interfacial polarization and charge barriers, partially shielding the local electric field, and increasing Ag... +Migration and the formation of conductive filaments are inhibited, leading to a decrease in LRS conductivity. Simultaneously, residual charge and polarization effects in the virus adsorption layer prevent some filaments from completely breaking during resetting, thus increasing HRS conductivity. Under the combined effect of these two factors, the overall on / off ratio (HRS / LRS) of the device decreases with increasing virus concentration. Conversely, when the virus concentration is low, interfacial interference is weak, and Ag... + The migration path is smoother, the formation and breakage of filaments are clearer and more controllable, and the device can recover a larger conductance difference. Therefore, while maintaining stable bipolar resistive switching characteristics, this memristor can reflect the virus concentration through the dynamic change of the on / off ratio, achieving sensitive and quantifiable electrical detection and analysis.

[0019] (2) The memristor provided by the present invention has different electrical response characteristics to virus solutions of different concentrations. This difference is due to the interfacial polarization and charge barrier changes caused by virus adsorption, thereby regulating Ag. + Migration and conductive filament formation process. Using memristors as sensors and combining them with detection circuits to capture electrical signals of adenovirus solutions of different concentrations, and analyzing the concentration of adenovirus solutions based on the differences in electrical signals, has the following advantages: (1) High sensitivity, as memristors are highly sensitive to weak changes in electrical signals, making them suitable for detecting minute differences in the electrical characteristics of virus concentration; (2) Small size, with the potential to be combined with microfluidic chips to achieve high-throughput detection; (3) The ability to dynamically record signals, with the potential for real-time analysis and monitoring; (4) Better integration with artificial neural networks, which can improve the efficiency and accuracy of virus concentration detection. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the structure of the memristor of the present invention.

[0021] Figure 2 Memristor characteristics analysis of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of this invention before testing. Figure 2 In the figure, a represents the typical IV curve of the memristor device before testing; b represents the retention characteristics analysis of the HRS and LRS of the memristor device after 100 cycles.

[0022] Figure 3 The image shows the test response of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of this invention to adenovirus solutions of different concentrations in deionized water medium. Figure 3 In the diagram, 'a' represents the response of the memristor device to deionized aqueous solution; 'b' to 'd' represent the responses of the memristor device to the detection of adenovirus at different concentrations in the deionized aqueous solution medium.

[0023] Figure 4This is a test response diagram of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of the present invention to adenovirus solutions of different concentrations in PBS medium. Figure 4 In the diagram, a represents the response of the memristor device to the PBS solution; b represents the response of the memristor device to the detection of adenovirus at different concentrations in the PBS solution medium.

[0024] Figure 5 The image shows the test response of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of this invention to adenovirus solutions of different concentrations in artificial cerebrospinal fluid medium. Figure 5 In the diagram, a represents the response of the memristor device to the artificial cerebrospinal fluid solution; b to c represent the responses of the memristor device to the detection of adenovirus at different concentrations in the artificial cerebrospinal fluid solution medium; and d represents the confusion matrix for determining the presence of adenovirus in the artificial cerebrospinal fluid using a machine learning algorithm. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] It should be noted that the technical terms used in this invention are only for the purpose of describing specific embodiments and are not intended to limit the scope of protection of this invention. Unless otherwise specified, all raw materials, reagents, instruments and equipment used in the following embodiments of this invention can be purchased from the market or prepared by existing methods.

[0027] On one hand, the present invention provides a memristor for adenovirus concentration detection. The memristor, along its thickness direction, comprises, from bottom to top, a bottom electrode layer 1, a metal oxide layer 2, a functional layer 3, and a top electrode layer 4. The bottom electrode layer 1 is an FTO conductive glass layer, the metal oxide layer 2 is tantalum pentoxide, the functional layer 3 is a gelatin-β-cyclodextrin composite film, and the top electrode layer 4 is a conductive electrode layer. The metal oxide layer 2 has oxygen vacancies and ion migration channels. When the top electrode is electrochemically oxidized into metal ions and migrates downwards under the action of an electric field, it forms localized conductive filaments penetrating the metal oxide layer. The active functional groups of the gelatin-β-cyclodextrin composite film regulate the surface polarization environment of the metal oxide layer 2 and capture or release metal ions, thereby causing the formation and breakage of the conductive filaments.

[0028] In this invention, the memristor is composed of a bottom electrode layer 1, a metal oxide layer 2, a functional layer 3, and a top electrode layer, arranged sequentially from bottom to top. The bottom electrode layer 1 is FTO conductive glass, the metal oxide layer 2 is tantalum pentoxide (Ta2O5), the functional layer 3 is a gelatin-β-cyclodextrin composite film, and the top electrode layer 4 is a conductive electrode layer. In some preferred embodiments, the bottom electrode layer 1 is an FTO electrode layer, and the top electrode layer 4 is an Ag electrode layer. Ta2O5, as a high dielectric constant material, has abundant oxygen vacancies and strong ion migration channels. When a positive bias is applied, the top electrode Ag is electrochemically oxidized to Ag. + Ions migrate downwards under the influence of an electric field, accumulating along oxygen vacancy paths in Ta₂O₅ and gradually reducing to Ag atoms, forming locally conductive filaments that penetrate the oxide layer. This allows the device to transition from a high-resistivity (HRS) state to a low-resistivity (LRS) state. Reverse bias or localized thermal effects can cause the filament neck to break or ion diffusion to dissolve back, restoring the device to a high-resistivity state and exhibiting typical reversible resistive switching characteristics. In this process, the current conduction mechanism is dominated by Schottky emission in the low-field region, Hopping and Fowler-Nordbeim emission in the medium-to-high-field region, and metallic conductivity in the on-state. The gelatin-β-cyclodextrin composite film plays a crucial regulatory and enhancing role in this system; its abundant hydroxyl and amino groups can modulate the polarization environment of the Ta₂O₅ surface and capture or release Ag. + This effectively stabilizes the formation and breakage process of conductive filaments. When memristors are used to detect adenovirus, the hydrophobic cavity structure of β-cyclodextrin can form a strong non-specific binding with the hydrophobic structural regions on the virus surface, causing the interfacial charge distribution and local electric field response of the device to change with the virus concentration, thereby causing a detectable change in the memristor's conductance. The higher the virus concentration, the more viral proteins adsorbed on the gelatin-β-cyclodextrin composite film interface, the more the interfacial polarization and charge barrier are significantly enhanced, and the local electric field is partially shielded. + Migration and the formation of conductive filaments are inhibited, leading to a decrease in LRS conductivity. Simultaneously, residual charge and polarization effects in the virus adsorption layer prevent some filaments from completely breaking during resetting, thus increasing HRS conductivity. Under the combined effect of these two factors, the overall HRS / LRS of the device decreases with increasing virus concentration. Conversely, when the virus concentration is low, interfacial interference is weak, and Ag... + The migration path is smoother, the formation and breakage of filaments are clearer and more controllable, and the device can recover a larger conductance difference. Therefore, while maintaining stable bipolar resistive switching characteristics, this memristor can reflect the virus concentration through the dynamic change of the on / off ratio, achieving sensitive and quantifiable electrical detection and analysis.

[0029] In a specific embodiment, the thickness of the gelatin-β-cyclodextrin composite film is 50 nm to 100 nm. The thickness of the metal oxide layer is 300 nm to 350 nm. The thickness of the bottom electrode layer is 150 nm to 200 nm; and the thickness of the top electrode layer is 100 nm to 150 nm.

[0030] The memristor provided by this invention uses an FTO electrode layer as the bottom electrode layer, tantalum pentoxide as the metal oxide layer, a gelatin-β-cyclodextrin composite film as the functional layer, and an Ag electrode layer as the top electrode layer, forming an Ag / gelatin-β-cyclodextrin / Ta₂O₅ / FTO memristor. The Ta₂O₅ layer is rich in oxygen vacancies and serves as a dielectric layer for ion migration and conductive channel formation. It can form a Schottky barrier at the interface with the FTO electrode layer, enhancing non-volatility and data retention. The Ag electrode layer, as the active electrode, readily undergoes electrochemical oxidation under an electric field, with Ag atoms easily oxidized to Ag₂. + Ions, via Ag + The migration of ions and the movement of oxygen vacancies in Ta₂O₅ work together to form and break conductive filaments, thereby achieving reversible changes in resistance. The gelatin-β-cyclodextrin composite film, rich in hydroxyl and amino groups, can modulate the polarization environment of the Ta₂O₅ surface and capture or release Ag. + This effectively stabilizes the formation and breakage process of the conductive filaments, thereby regulating and enhancing the performance of the memristor. The Ag electrode layer is resistant to oxidation and corrosion, improving the long-term reliability of the device and reducing its power consumption due to its low resistivity.

[0031] In this invention, the performance of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor is highly dependent on the thickness of each layer. By precisely controlling the thickness matching and interface engineering, reversible changes in resistance can be achieved. Through biofunctionalization and its unique resistance change characteristics, the specific binding between antibodies and antigens can cause changes in the resistance of the memristor, which is further fed back to its memristor behavior. This is used to record and analyze the electrochemical or bioelectrical signal characteristics of adenovirus solutions of different concentrations. Based on the differences in electrical signals, the adenovirus concentration can be effectively distinguished. If the thickness of the FTO electrode layer is too thin (<150nm), it may lead to poor contact with the Ta2O5 layer, increase the interface resistance, affect the electron injection efficiency, easily cause local breakdown, and reduce the stability of the device. If it is too thick (>200nm), it increases the overall size of the device, which is not conducive to high-density integration and also introduces thermal stress, affecting the crystal quality of the Ta2O5 layer. If the Ta2O5 layer is too thin (<300nm): the electric field strength is too high, the conductive filaments (Ag) form rapidly but are difficult to control, the oxygen vacancy distribution is limited, and multiple filaments compete for space, reducing durability. If it is too thick (>350nm): the voltage required to drive ion migration increases, energy consumption increases, the conductive filament path is too long, and the on / off ratio decreases. If the gelatin-β-cyclodextrin composite film layer is too thin (<50nm): it will lead to insufficient interface recognition sites, weak virus adsorption capacity, and ineffective Ag regulation. + Migration; Excessive thickness (>100nm): Causes electric field attenuation and restricts ion migration, making it difficult to form conductive filaments. Both of these factors weaken the electrical response sensitivity and stability of the device. If the Ag electrode layer is too thin (<100nm): The Ag ion source is limited, and Ag is depleted after multiple cycles, leading to device failure. Ion migration rate is limited, and sensitivity is reduced. Excessive thickness (>150nm) can easily cause cracking of the gelatin-β-cyclodextrin composite film layer.

[0032] On the other hand, the present invention provides a method for preparing the memristor for adenovirus concentration detection, comprising the following steps: S1. A metal oxide layer is deposited on the bottom electrode substrate by magnetron sputtering.

[0033] In the magnetron sputtering process, the target-substrate distance is 9cm to 11cm, and the vacuum level is 3×10⁻⁶. -4 Pa~4×10 - 4 Pa, working gas is argon, sputtering pressure is 0.7 Pa to 0.8 Pa, sputtering power is 120 W / cm². 2 The sputtering time is 40 min to 90 min.

[0034] S2. A gelatin-β-cyclodextrin composite film functional layer is coated onto the metal oxide layer using a spin coating method.

[0035] In the spin coating process, a gelatin-β-cyclodextrin precursor solution is dropped onto the surface of the metal oxide layer, with a drop volume of 120μL to 300μL. The coating is first spin-coated at 400rpm to 600rpm for 15s to 25s, and then at 1800rpm to 2400rpm for 25s to 35s. After coating, the film is cross-linked with glutaraldehyde vapor for 7h to 9h to enhance the structural stability of the gelatin-β-cyclodextrin composite film.

[0036] A method for preparing a gelatin-β-cyclodextrin precursor solution includes the following steps: Add water to gelatin and stir at 60°C for 1 hour until completely dissolved. Then add β-cyclodextrin and stir at 60°C for 1 hour until completely dissolved. Let stand overnight to obtain a gelatin-β-cyclodextrin precursor solution. The mass ratio of gelatin to β-cyclodextrin is 1:0.4-0.5, and the volume ratio of gelatin to water is 0.4g-0.5g:10mL.

[0037] S3. Cover the functional layer with a metal mask with openings, and deposit an electrode layer on the metal mask by DC sputtering to obtain a memristor for adenovirus concentration detection.

[0038] In the DC sputtering process, the target-substrate distance is 9cm to 11cm, and the vacuum level is 2×10⁻⁶. -4 Pa~3×10 - 4 Pa, working gas is argon, sputtering pressure is 0.7 Pa to 0.8 Pa, sputtering power is 60 W / cm². 2 The sputtering time is 15 min to 25 min.

[0039] In a specific embodiment, the metal mask has an array of openings, and the upper electrode layer is deposited on all the openings. The diameter of the openings can be 0.8 mm to 1.5 mm.

[0040] This invention uses sputtering to fabricate memristors, resulting in uniform and dense films of the bottom electrode layer, metal oxide layer, functional layer, and top electrode layer, with strong adhesion and stable and highly repeatable process.

[0041] The memristor provided by this invention exhibits different electrical response characteristics to virus solutions of varying concentrations. These differences stem from interfacial polarization and charge barrier changes caused by virus adsorption, thereby modulating the Ag... +Migration and conductive filament formation process. Using memristors as sensors and combining them with detection circuits to capture electrical signals of adenovirus solutions of different concentrations, and analyzing the concentration of adenovirus solutions based on the differences in electrical signals, has the following advantages: (1) High sensitivity, as memristors are highly sensitive to weak changes in electrical signals, making them suitable for detecting minute differences in the electrical characteristics of virus concentration; (2) Small size, with the potential to be combined with microfluidic chips to achieve high-throughput detection; (3) The ability to dynamically record signals, with the potential for real-time analysis and monitoring; (4) Better integration with artificial neural networks, which can improve the efficiency and accuracy of virus concentration detection.

[0042] The following specific examples will provide further explanation.

[0043] Example 1 A memristor for adenovirus concentration detection, such as Figure 1 As shown, the memristor, along its thickness direction, comprises, from bottom to top, an FTO electrode layer, a Ta2O5 oxide layer, a gelatin-β-cyclodextrin composite film functional layer, and an Ag electrode layer; the thickness of the FTO electrode layer is 160 nm, the thickness of the Ta2O5 oxide layer is 300 nm, the thickness of the gelatin-β-cyclodextrin composite film functional layer is 80 nm, and the thickness of the Ag electrode layer is 120 nm.

[0044] The above-mentioned method for preparing a memristor for adenovirus concentration detection includes the following steps: S1. Cleaning the substrate: Soak the glass substrate in deionized water, alcohol, acetone, alcohol, and deionized water in sequence for 30 minutes to completely clean the surface oil. After drying the glass substrate with N2, place it in the magnetron sputtering chamber as a substrate.

[0045] S2. A Ta2O5 oxide layer is deposited on the FTO electrode layer using magnetron sputtering: A Ta2O5 compound target is mounted on the magnetron sputtering gun, with a Ta:O atomic ratio of 2:5. The target-substrate distance is set to 11 cm, and the background vacuum of the sputtering chamber is evacuated to 4 × 10⁻⁶. -4 The sputtering pressure was 0.8 Pa, and the sputtering power was 120 W / cm². Argon gas with a purity of 99.999% was introduced as the working gas. 2 The sputtering time was 60 min, and a Ta2O5 oxide layer with a thickness of 300 nm was obtained.

[0046] S3. Prepare a gelatin-β-cyclodextrin precursor solution by adding 0.4 g of gelatin and 10 mL of deionized water, and stirring at 60°C for 1 hour until completely dissolved. After complete dissolution, add 0.16 g of β-cyclodextrin to the gelatin solution and continue stirring at 60°C for 1 hour until completely dissolved. Let the resulting mixture stand at room temperature overnight for later use. A gelatin-β-cyclodextrin composite film is prepared on the surface of Ta2O5 using a spin-coating process. 200 μL of the gelatin-β-cyclodextrin precursor solution is taken using a pipette, and the spin-coating program is set to 500 rpm for 20 seconds and 2000 rpm for 30 seconds. The coated substrate is placed in a 37°C oven and cross-linked with glutaraldehyde vapor for 8 hours to enhance the structural stability of the gelatin-β-cyclodextrin composite film, obtaining a gelatin-β-cyclodextrin composite film layer with a thickness of 80 nm.

[0047] S4. Deposit an Ag electrode layer on the gelatin-β-cyclodextrin composite film using magnetron sputtering: Cover the surface of the gelatin-β-cyclodextrin composite film layer with a metal mask with a aperture of 1 mm, deposit a circular Ag electrode using DC sputtering, mount an Ag metal target on the magnetron sputtering DC target, set the target-substrate distance to 11 cm, and evacuate the sputtering chamber to a background vacuum of 3 × 10⁻⁶. -4 The sputtering pressure was 0.8 Pa, and the sputtering power was 60 W / cm². Argon gas with a purity of 99.999% was introduced as the working gas. 2 The sputtering time was 15 min to obtain an Ag electrode layer with a thickness of 120 nm. Silver wires were fixed on the Ag electrode layer and Ti electrode layer respectively using silver paste to obtain a memristor, named Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor.

[0048] Example 2 A memristor for adenovirus concentration detection, such as Figure 1 As shown, the memristor, along its thickness direction, comprises, from bottom to top, an FTO electrode layer, a Ta2O5 oxide layer, a gelatin-β-cyclodextrin composite film functional layer, and an Ag electrode layer; the thickness of the FTO electrode layer is 160 nm, the thickness of the Ta2O5 oxide layer is 350 nm, the thickness of the gelatin-β-cyclodextrin composite film functional layer is 50 nm, and the thickness of the Ag electrode layer is 120 nm.

[0049] The above-mentioned method for preparing a memristor for adenovirus concentration detection includes the following steps: S1. Cleaning the substrate: Soak the glass substrate in deionized water, alcohol, acetone, alcohol, and deionized water in sequence for 30 minutes to completely clean the surface oil. After drying the glass substrate with N2, place it in the magnetron sputtering chamber as a substrate.

[0050] S2. A Ta2O5 oxide layer is deposited on the FTO electrode layer using magnetron sputtering: A Ta2O5 compound target is mounted on the magnetron sputtering gun, with a Ta:O atomic ratio of 2:5. The target-substrate distance is set to 10 cm, and the background vacuum of the sputtering chamber is evacuated to 4 × 10⁻⁶. -4 The sputtering pressure was 0.8 Pa, and the sputtering power was 120 W / cm². Argon gas with a purity of 99.999% was introduced as the working gas. 2 The sputtering time was 70 min, and a Ta2O5 oxide layer with a thickness of 350 nm was obtained.

[0051] S3. Prepare a gelatin-β-cyclodextrin precursor solution by adding 0.4 g of gelatin and 10 mL of deionized water, and stirring at 60°C for 1 hour until completely dissolved. After complete dissolution, add 0.16 g of β-cyclodextrin to the gelatin solution and continue stirring at 60°C for 1 hour until completely dissolved. Let the resulting mixture stand at room temperature overnight for later use. A gelatin-β-cyclodextrin composite film is prepared on the surface of Ta2O5 using a spin-coating process. 125 μL of the gelatin-β-cyclodextrin precursor solution is taken using a pipette, and the spin-coating program is set to 500 rpm for 20 seconds and 2000 rpm for 30 seconds. The coated substrate is placed in a 37°C oven and crosslinked with glutaraldehyde vapor for 8 hours to enhance the structural stability of the gelatin-β-cyclodextrin composite film, obtaining a gelatin-β-cyclodextrin composite film layer with a thickness of 50 nm.

[0052] S4. Deposit an Ag electrode layer on the gelatin-β-cyclodextrin composite film using magnetron sputtering: Cover the surface of the gelatin-β-cyclodextrin composite film layer with a metal mask with a aperture of 1 mm, deposit a circular Ag electrode using DC sputtering, mount an Ag metal target on the magnetron sputtering DC target, set the target-substrate distance to 11 cm, and evacuate the sputtering chamber to a background vacuum of 3 × 10⁻⁶. -4 The sputtering pressure was 0.8 Pa, and the sputtering power was 60 W / cm². Argon gas with a purity of 99.999% was introduced as the working gas. 2 The sputtering time was 15 min to obtain an Ag electrode layer with a thickness of 120 nm. Silver wires were fixed on the Ag electrode layer and Ti electrode layer respectively using silver paste to obtain a memristor, named Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor.

[0053] Example 3 A memristor for adenovirus concentration detection, such as Figure 1As shown, the memristor comprises, from bottom to top, an FTO electrode layer, a Ta2O5 oxide layer, a gelatin-β-cyclodextrin composite film functional layer, and an Ag electrode layer along the thickness direction; the thickness of the FTO electrode layer is 160 nm, the thickness of the Ta2O5 oxide layer is 300 nm, the thickness of the gelatin-β-cyclodextrin composite film functional layer is 100 nm, and the thickness of the Ag electrode layer is 120 nm.

[0054] The above-mentioned method for preparing a memristor for adenovirus concentration detection includes the following steps: S1. Cleaning the substrate: Soak the glass substrate in deionized water, alcohol, acetone, alcohol, and deionized water in sequence for 30 minutes to completely clean the surface oil. After drying the glass substrate with N2, place it in the magnetron sputtering chamber as a substrate.

[0055] S2. A Ta2O5 oxide layer is deposited on the FTO electrode layer using magnetron sputtering: A Ta2O5 compound target is mounted on the magnetron sputtering gun, with a Ta:O atomic ratio of 2:5. The target-substrate distance is set to 11 cm, and the background vacuum of the sputtering chamber is evacuated to 4 × 10⁻⁶. -4 The sputtering pressure was 0.7 Pa, and the sputtering power was 120 W / cm². Argon gas with a purity of 99.999% was introduced as the working gas. 2 The sputtering time was 90 min, and a Ta2O5 oxide layer with a thickness of 300 nm was obtained.

[0056] S3. Prepare a gelatin-β-cyclodextrin precursor solution by adding 0.4 g of gelatin and 10 mL of deionized water, and stirring at 60°C for 1 hour until completely dissolved. After complete dissolution, add 0.16 g of β-cyclodextrin to the gelatin solution and continue stirring at 60°C for 1 hour until completely dissolved. Let the resulting mixture stand at room temperature overnight for later use. A gelatin-β-cyclodextrin composite film is prepared on the surface of Ta2O5 using a spin-coating process. 250 μL of the gelatin-β-cyclodextrin precursor solution is taken using a pipette, and the spin-coating program is set to 500 rpm for 20 seconds and 2000 rpm for 30 seconds. The coated substrate is placed in a 37°C oven and cross-linked with glutaraldehyde vapor for 8 hours to enhance the structural stability of the gelatin-β-cyclodextrin composite film, obtaining a gelatin-β-cyclodextrin composite film layer with a thickness of 100 nm.

[0057] S4. Deposit an Ag electrode layer on the gelatin-β-cyclodextrin composite film using magnetron sputtering: Cover the surface of the gelatin-β-cyclodextrin composite film layer with a metal mask with a aperture of 1 mm, deposit a circular Ag electrode using DC sputtering, mount an Ag metal target on the magnetron sputtering DC target, set the target-substrate distance to 11 cm, and evacuate the sputtering chamber to a background vacuum of 3 × 10⁻⁶. -4The working gas is argon gas with a purity of 99.999%, the sputtering pressure is 0.7 Pa, and the sputtering power is 60 W / cm². 2 The sputtering time was 20 min to obtain an Ag electrode layer with a thickness of 120 nm. Silver wires were fixed on the Ag electrode layer and Ti electrode layer respectively using silver paste to obtain a memristor, named Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor.

[0058] Since the memristors prepared in Examples 1 to 3 have similar structures and essentially the same performance, Example 1 was used as the subject for application in detecting adenovirus solution concentration in different solution media. The results are as follows.

[0059] When preparing test samples, the adenovirus concentration needs to be identified. The identification method is as follows: The initial concentration of the adenovirus sample (HANBIO, Shanghai, China) is 1×10⁻⁶. 12 VP / ml. Virus titration was performed as follows: the absorbance of the virus suspension was measured at a wavelength of 260 nm, and this absorbance value corresponds to the concentration of virus particles. An optical density value (OD260) of 1.0 is equivalent to 1 × 10⁻⁶ VP / ml. 12 Each virus particle was collected and stored at 4°C. A virus lysis buffer was prepared, containing 0.1% SDS, 10 mM Tris-HCl (pH 7.4), and 1 mM EDTA. Three 1.5 mL centrifuge tubes were prepared, and the virus suspension was diluted with VLB buffer at ratios of 1:50, 1:25, and 1:10, respectively. The mixture was incubated in a 55°C water bath for 15 minutes, vortexed every 5 minutes, and then centrifuged at 13,000 rpm for 2 minutes. The supernatant was transferred to a clean centrifuge tube, and the OD260 value was measured using a spectrophotometer, with VLB buffer as a blank control. Each sample was measured three times, and the average OD260 value was used to calculate the virus concentration using the formula: Virus concentration (VP / mL) = OD260 × dilution factor × 1.1 × 10⁻⁶. 12 The quantified virus suspension was then diluted with deionized water, PBS, or ACSF to obtain the working concentration (VP / mL) required for subsequent experiments.

[0060] In this invention, by combining the XGBoost algorithm and based on the electrical response of a memristor to adenovirus solutions of different concentrations, the presence of adenovirus in the solution can be quickly and effectively determined. Specific method:

[0061] First, feature extraction is performed, extracting multiple statistical and physical features from the resistive switching curve, including turn-on voltage, reset voltage, HRS / LRS conductance, on / off ratio, nonlinearity index, hysteresis area, and dynamic current change rate, to construct a multidimensional feature matrix to characterize the electrical response mode under different virus concentration conditions.

[0062] Among them, the multi-dimensional feature extraction method based on specified features (applicable to the -3V to 3V scan voltage and current data scenario, i.e., 0V→3V→0V→-3V→0V) first prepares the data foundation by collecting target device test data containing sample identifiers, -3V to 3V scan voltage values, corresponding current measurement values, and binary labels. The device is divided into high-resistance and low-resistance states based on its conduction / cutoff characteristics at different voltages. The average current in the corresponding range is taken as the HRS and LRS values, and the HRS / LRS ratio is calculated. Then, data rows with empty or abnormal voltage, HRS, and LRS values ​​are deleted to form a standardized raw dataset. This dataset is then divided into multiple independent sample data subsets based on the sample identifiers. Next, seven types of features are extracted for each sample data subset, including the total number of data rows, the number of effective voltage values, and the effective HRS / LRS ratio. The basic statistical characteristics of the number of S values ​​are analyzed, and the distribution characteristics of the mean, median, and standard deviation of the HRS / LRS ratio are calculated. Specific voltage characteristics of the HRS / LRS values ​​and their difference at +0.5V and -0.5V are extracted. The relationship characteristics between the Pearson correlation coefficient, linear fitting slope and intercept, and trapezoidal area of ​​the V-HRS / LRS curve are calculated. The classification proportion characteristics of the HRS / LRS ratio in different intervals are statistically analyzed. Univariate statistical characteristics of the mean, standard deviation, and average ratio of HRS and LRS are calculated. The identification label characteristics of sample identifiers and binary labels are extracted. Finally, all specified features of each sample are integrated in the format of "sample row - feature column" to form a multidimensional feature matrix with the dimension of "sample number × total number of specified features", which is used for subsequent data analysis and modeling tasks.

[0063] Secondly, during the model training phase, the XGBoost algorithm is used to construct the classifier. Its core idea is to achieve efficient fitting of nonlinear features through iterative weighted combination of multiple weak classification trees. To obtain optimal performance, a hyperparameter optimization strategy combining grid search and five-fold cross-validation is used to systematically tune parameters such as learning rate, maximum depth, subsampling ratio, and L1 / L2 regularization coefficients to prevent overfitting and improve generalization ability.

[0064] Among them, based on the "multidimensional feature matrix in the -3V to 3V scan voltage and current data scenario", the XGBoost binary classification model for adapter device qualification classification, combined with specified hyperparameters (RANDOM_SEED=42, OUTER_SPLITS=5, INNER_SPLITS=4, N_TRIALS=80, EARLY_STOP=200, N_PERM=500, N_BOOT=1000, TOPK_SHAP=25), adopts a nested strategy of "outer-layer rigorous verification + inner-layer Optuna parameter tuning". First, the multidimensional feature matrix is ​​used as input (excluding the Sample identifier column and Binary class). The label column is the 0 / 1 dependent variable y. The feature column does not need to be standardized. It is encapsulated with XGBoost's DMatrix and random_state=42 is fixed to ensure reproducibility. The outer 5-fold cross-validation is fixed (divided into groups, 1 fold is the outer test set and 4 fold is the inner training set, looped 5 times) and the inner 4-fold cross-validation (adjusted automatically according to the number of groups). The range of parameters to be optimized is set to 7 items such as learning rate [0.01, 0.2] and maximum depth [3, 10]. 80 Bayesian optimization searches are performed through Optuna (with the goal of maximizing the mean F1 score of the inner cross-validation, EARLY_STOP=200 is set for early stopping, and the optimal hyperparameters of the inner layer are selected according to the priority of reg_lambda→max_depth→subsample). The optimal hyperparameters of the inner layer are then substituted into the model. After retraining on the "inner training set", the performance is evaluated on the "outer test set". The 95% confidence interval of the performance index is calculated based on 1000 Bootstrap samplings (the performance is stable if the span is <10%). The significance of the 25 core features of TOPK_SHAP is verified through 500 permutation tests (the feature proportion with p value <0.05 is >80% if it is effective). At the same time, the difference between the F1 scores of the inner and outer training sets and the validation set is monitored (overfitting is judged and the parameter range is adjusted with thresholds of 15% and 10% respectively). Finally, the optimal hyperparameters of the inner layer, the ranking of parameter importance, the performance index of the outer layer, and the SHAP plot of the core features are output. In practical applications, newly collected data is input into the model trained based on the "mean of the optimal hyperparameters of the outer layer" after feature extraction, so as to realize device classification and performance anomaly diagnosis.

[0065] Finally, after the model training is complete, it is independently tested using outer-layer cross-validation. All of the above data processing procedures were implemented and parameter-tuned using Python.

[0066] The Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 was operated under no-load conditions. The voltage was scanned from 0V to 3V to 0V to -3V to 0V, and the current was the output. Its IV curve was continuously monitored. After the IV curve of the memristor stabilized under no-load conditions, 1 μL of virus solution was dropped onto the surface of the memristor to obtain the IV curve for testing.

[0067] Figure 3 The image shows the test response of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of this invention to adenovirus solutions of different concentrations in deionized water medium. Figure 3 In the diagram, (a) represents the device's response to a deionized aqueous solution; (b) to (d) represent the device's response to the detection of adenovirus at different concentrations in a deionized aqueous solution. For example... Figure 2 As shown, this device exhibits very stable memristor characteristics and retention characteristics.

[0068] Figure 3 The image shows the test response of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of this invention to adenovirus solutions of different concentrations in deionized water medium. Figure 3 In the diagram, 'a' represents the response of the memristor device to deionized aqueous solution; 'b' to 'd' represent the responses of the memristor device to the detection of adenovirus at different concentrations in the deionized aqueous solution medium. For example... Figure 3 As shown, the prepared device exhibits significant differences in response to different concentrations of adenovirus in deionized water medium. The on / off ratio decreases with increasing virus concentration, especially when the concentration exceeds 10⁻⁶. 5 This trend is particularly pronounced after VP / ml.

[0069] Figure 4 This is a test response diagram of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of the present invention to adenovirus solutions of different concentrations in PBS medium. Figure 4 In the diagram, 'a' represents the response of the memristor device to the PBS solution; 'b' represents the response of the memristor device to the detection of different concentrations of adenovirus in the PBS solution medium. For example... Figure 4 As shown, it can be seen that in PBS medium, when the concentration exceeds 10... 5 After VP / ml, the prepared device still retained the response to different concentrations of adenovirus, and the trend was the same as in the aqueous medium.

[0070] Figure 5 The image shows the test response of the Ag / gelatin-β-cyclodextrin / Ta2O5 / FTO memristor prepared in Example 1 of this invention to adenovirus solutions of different concentrations in artificial cerebrospinal fluid medium. Figure 5In the diagram, 'a' represents the response of the memristor device to the artificial cerebrospinal fluid solution; 'b' to 'c' represent the responses of the memristor device to the detection of different concentrations of adenovirus in the artificial cerebrospinal fluid solution; and 'd' represents the confusion matrix used in conjunction with a machine learning algorithm to determine the presence of adenovirus in the artificial cerebrospinal fluid. For example... Figure 5 As shown, it can be seen that in artificial cerebrospinal fluid media, when the concentration exceeds 10... 5 After VP / ml, the fabricated device still retained its response to different concentrations of adenovirus, and the trend was the same as in the aqueous medium. Combined with the XGBoost algorithm, the presence of 10 in the artificial cerebrospinal fluid medium can be quickly determined. 5 Adenovirus at concentrations of VP / ml and above has a sensitivity of 0.989 and a specificity of 0.758.

[0071] It should be noted that when numerical ranges are involved in this invention, it should be understood that both endpoints of each numerical range and any value between the two endpoints can be selected. Since the steps and methods used are the same as in the embodiments, preferred embodiments are described here to avoid redundancy. Although preferred embodiments of the invention have been described, those skilled in the art, once they understand the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this invention.

[0072] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A memristor for adenovirus concentration detection, characterized in that, The memristor, along its thickness direction, comprises, from bottom to top, a bottom electrode layer, a metal oxide layer, a functional layer, and a top electrode layer. The bottom electrode layer is an FTO conductive glass layer, the metal oxide layer is tantalum pentoxide, the functional layer is a gelatin-β-cyclodextrin composite film, and the top electrode layer is a conductive electrode layer.

2. The memristor for adenovirus concentration detection according to claim 1, characterized in that, The thickness of the gelatin-β-cyclodextrin composite film is 50 nm to 100 nm.

3. The memristor for adenovirus concentration detection according to claim 1, characterized in that, The thickness of the metal oxide layer is 300 nm to 350 nm.

4. The memristor for adenovirus concentration detection according to claim 1, characterized in that, The bottom electrode layer is an FTO electrode layer with a thickness of 150nm to 200nm; the top electrode layer is an Ag electrode layer with a thickness of 100nm to 150nm.

5. A method for preparing a memristor for adenovirus concentration detection according to any one of claims 1 to 4, characterized in that, Includes the following steps: A metal oxide layer is deposited on the bottom electrode layer substrate by magnetron sputtering. A gelatin-β-cyclodextrin composite film functional layer was coated onto a metal oxide layer using a spin coating method. A metal mask with openings is covered on the functional layer, and an electrode layer is deposited on the metal mask by DC sputtering to obtain a memristor for adenovirus concentration detection.

6. The method for preparing a memristor for adenovirus concentration detection according to claim 1, characterized in that, During the magnetron sputtering process, the target-substrate distance is 9cm to 11cm, and the vacuum level is 3×10⁻⁶. -4 Pa~4×10 -4 Pa, working gas is argon, sputtering pressure is 0.7 Pa to 0.8 Pa, sputtering power is 120 W / cm². 2 The sputtering time is 40 min to 90 min.

7. The method for preparing a memristor for adenovirus concentration detection according to claim 6, characterized in that, During the spin coating process, a gelatin-β-cyclodextrin precursor solution is dropped onto the surface of the metal oxide layer, with a drop volume of 120μL to 300μL. The coating is first spin-coated at 400rpm to 600rpm for 15s to 25s, and then at 1800rpm to 2400rpm for 25s to 35s. After coating, the coating is cross-linked with glutaraldehyde vapor for 7h to 9h.

8. The method for preparing a memristor for adenovirus concentration detection according to claim 7, characterized in that, A method for preparing a gelatin-β-cyclodextrin precursor solution includes the following steps: Add water to gelatin and stir until completely dissolved. Then add β-cyclodextrin and stir until completely dissolved. Let stand overnight to obtain a gelatin-β-cyclodextrin precursor solution. The mass ratio of gelatin to β-cyclodextrin is 1:0.4 to 0.5, and the ratio of gelatin to water is 0.4 g to 0.5 g: 10 mL.

9. The method for preparing a memristor for adenovirus concentration detection according to claim 6, characterized in that, During the DC sputtering process, the target-substrate distance is 9cm to 11cm, and the vacuum level is 2×10⁻⁶. -4 Pa~3×10 -4 Pa, working gas is argon, sputtering pressure is 0.7 Pa to 0.8 Pa, sputtering power is 60 W / cm². 2 The sputtering time is 15 min to 25 min.

10. The method for preparing a memristor for adenovirus concentration detection according to claim 6, characterized in that, The metal mask has an array of openings, and the upper electrode layer is deposited on all the openings.

Images