A digital recognition memristor and a method for preparing the same

Abstract

The present disclosure provides a digital recognition memristor and a preparation method thereof, wherein the digital recognition memristor comprises: a bottom electrode, a dielectric layer and a top electrode; wherein the dielectric layer is located between the bottom electrode and the top electrode, the dielectric layer is a heterojunction formed by zinc oxide nanoparticles and copper oxide nanowires, the zinc oxide nanoparticles are attached to the copper oxide nanowires, the heterojunction serves as an oxygen vacancy aggregation layer, connects the top electrode and the bottom electrode to form a conductive path, and makes the conductance form a gradual process, showing synaptic characteristics. Compared with the existing memristors that are mostly used for storage, the scheme described in the present disclosure has the advantages of simple preparation, special microscopic morphology, and the function of simulating human brain synapses, which is realized in handwritten digital recognition. The prepared memristor has great potential in neuromorphic computing and provides a foundation for future emerging computer systems.

Description

Technical Field

[0001] The present invention belongs to the technical field of memristor design, and in particular, relates to a digital recognition memristor and a preparation method thereof. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] As the fourth type of passive device besides resistors, capacitors and inductors, memristors have become the most promising candidate for building the next generation of high-performance memory and realizing on-chip neuromorphic computing due to their advantages such as simple structure, small size, multi-level storage and integrated storage and computing. According to the change of resistance state, memristors can be divided into digital memristors and analog memristors. Digital memristors can mutate between two resistance states, while analog memristors can continuously change between multiple resistance states. This gradual conductivity phenomenon can simulate the synaptic behavior of organisms and provide the possibility of building artificial electronic synapses and realizing neuromorphic computing.

[0004] As we all know, neuromorphic computing system is a new computing mode that can achieve parallelism, high efficiency and low power consumption, and is an important means to build the next generation of computer architecture. At present, although many materials have been proven to have resistive switching properties, including organic materials, perovskite materials and metal oxides, these materials often do not have the characteristics of neuromorphic computing, which limits their development in future artificial intelligence and neural networks, resulting in existing memristors being mostly used for storage and unable to play a role in neural networks. Summary of the invention

[0005] In order to solve the above problems, the present disclosure provides a digital recognition memristor and a preparation method thereof. The scheme provides a digital recognition memristor based on zinc oxide nanoparticles / copper oxide nanowires heterojunction. The memristor of ZnO NPs / CuO NWs heterojunction is prepared by solution method and spin coating method, which can effectively exhibit artificial synaptic characteristics. Its performance can be applied to handwritten digital recognition networks and can achieve a high accuracy.

[0006] According to a first aspect of an embodiment of the present disclosure, a digital recognition memristor is provided, comprising: a bottom electrode, a dielectric layer and a top electrode; wherein the dielectric layer is located between the bottom electrode and the top electrode, the dielectric layer is a heterojunction formed by zinc oxide nanoparticles and copper oxide nanowires, the zinc oxide nanoparticles are attached to the copper oxide nanowires, the heterojunction serves as an oxygen vacancy aggregation layer, connects the top electrode and the bottom electrode to form a conductive path, so that the conductivity forms a gradual process, exhibiting synaptic characteristics.

[0007] Furthermore, the heterojunction acts as an oxygen vacancy aggregation layer, and the resistive switching behavior is controlled by regulating the migration of oxygen vacancies.

[0008] Furthermore, when a positive voltage is applied to the top electrode, the oxygen vacancies in the dielectric layer gather at the top electrode and diffuse from the top electrode to the bottom electrode to form oxygen vacancy conductive filaments, and the conductive filaments have the characteristic of orderly growing along the nanowire structure; at the same time, electrons transfer from the cathode to the anode, the energy band bends, and the conductivity value increases continuously.

[0009] Furthermore, when negative pressure is applied to the top electrode, the oxygen vacancy conductive filaments begin to break from the thinner bottom electrode under the action of the internal electric field. At the same time, electrons return from the anode to the cathode, causing a continuous decrease in conductivity.

[0010] Furthermore, the synaptic characteristic is specifically manifested as a continuous increase and decrease in conductance value.

[0011] According to a second aspect of an embodiment of the present disclosure, a method for preparing a digital identification memristor is provided, comprising:

[0012] The Cu / CuO NWs structure was obtained based on the chemical solution method;

[0013] The ZnO NPs solution was prepared by a hot solvent method;

[0014] Based on the obtained Cu / CuO NWs structure and ZnO NPs solution, a digital recognition memristor based on ZnO nanoparticles / copper oxide nanowire heterojunction was prepared by spin coating.

[0015] Furthermore, the Cu / CuO NWs structure is prepared based on the chemical solution method, specifically:

[0016] Dissolve sodium hydroxide and ammonium persulfate in deionized water simultaneously, perform ultrasonication until they are completely dissolved, and obtain a mixed solution for later use;

[0017] The copper foil was placed in the prepared mixed solution until the copper foil turned blue, and then rinsed with deionized water and dried to obtain a Cu / Cu(OH)2NWs structure;

[0018] The Cu(OH)2NWs were heated to obtain the Cu / CuO NWs structure;

[0019] or,

[0020] The ZnO NPs solution is prepared by the hot solvent method, specifically:

[0021] Dissolve zinc acetate dihydrate in ethanol and stir to dissolve it under heating to obtain a zinc acetate solution for later use;

[0022] Dissolve potassium hydroxide in ethanol and ultrasonicate to obtain a colorless and transparent solution, which is potassium hydroxide solution and is ready for use;

[0023] The potassium hydroxide solution is slowly dripped into the zinc acetate solution under conditions of increasing temperature and vigorous stirring to carry out a hot solvent reaction to obtain a ZnO NPs solution.

[0024] Furthermore, the copper foil is placed in the mixed solution for 10-30 minutes;

[0025] or,

[0026] The conditions for drying the Cu(OH)2NWs are: maintaining at 55-85°C for 10-30min;

[0027] or,

[0028] The growth conditions of the CuO NWs are: 140-180° C. for 120-180 min;

[0029] or,

[0030] The zinc acetate dihydrate is dissolved in ethanol and stirred at a temperature of 60-100°C;

[0031] or,

[0032] The conditions of the hot solvent reaction are: maintaining at 60-100° C. for 90-120 min.

[0033] Furthermore, the digital recognition memristor based on zinc oxide nanoparticle / copper oxide nanowire heterojunction is prepared by spin coating, specifically:

[0034] The ZnO NPs solution was dropped onto CuO NWs, spin-coated and then thermally cured to obtain the Cu / CuO NWs / ZnO NPs structure.

[0035] A top electrode was prepared on the CuO NWs / ZnO NPs heterojunction to obtain a "Cu / CuO NWs / ZnO NPs / top electrode" structure, realizing the preparation of a digital recognition memristor.

[0036] Furthermore, the top electrode of the digital recognition memristor is prepared by magnetron sputtering;

[0037] or,

[0038] The magnetron sputtering parameters are as follows: the base pressure is 5.0×10 -3 Pa, working pressure is 0.6-1.0Pa, sputtering temperature is 20-50°C, gas flow rate is 10-30sccm, sputtering energy is 50-80W, pre-sputtering time is 1-2min and sputtering time is 20-60min.

[0039] or,

[0040] When spin coating ZnO NPs, the process parameters of spin coating are: spin coating at a rotation speed of 2500-4500 rpm for 20-60 seconds, and then heating on a hot plate at 70-90° C. for 5-30 minutes.

[0041] Compared with the prior art, the beneficial effects of the present invention are:

[0042] The present disclosure provides a digital recognition memristor and a preparation method thereof. The scheme uses a solution method and a spin coating method to prepare a ZnO NPs / CuO NWs heterojunction memristor exhibiting artificial synaptic characteristics, and its performance can be applied to a handwritten digital recognition network with an accuracy rate of up to 93%. Compared with existing digital memristors, the present disclosure has the advantages of simple preparation, special microscopic morphology, and can simulate the function of human brain synapses. It has great potential in neuromorphic computing and provides a foundation for future emerging computer systems.

[0043] Advantages of additional aspects of the present disclosure will be given in part in the following description and in part will become apparent from the following description or will be learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings constituting a part of the present disclosure are used to provide a further understanding of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure.

[0045] FIG. 1( a ) is a schematic diagram of the structure of a digital recognition memristor of a zinc oxide nanoparticle / copper oxide nanowire heterojunction described in an embodiment of the present disclosure;

[0046] FIG1( b ) is a SEM image of a digital recognition memristor of a zinc oxide nanoparticle / copper oxide nanowire heterojunction described in an embodiment of the present disclosure;

[0047] Figure 2 TEM characterization image of the ZnO NPs / CuO NWs heterojunction prepared in the embodiments of the present disclosure;

[0048] FIG3( a ) shows the change of the current of the ZnO NPs / CuONWs heterojunction-based memristor prepared under continuous positive voltage scanning in the embodiment of the present disclosure;

[0049] FIG3( b ) shows the change of the current of the ZnO NPs / CuONWs heterojunction-based memristor prepared under continuous negative voltage scanning as described in the embodiment of the present disclosure;

[0050] Figure 4A neural network model diagram for digital recognition combined with the prepared ZnO NPs / CuO NWs-based memristor described in the embodiments of the present disclosure;

[0051] FIG5( a ) is a graph showing the accuracy of the digital recognition memristor prepared in the embodiment of the present disclosure in handwritten digital recognition;

[0052] FIG5( b ) is a confusion matrix diagram of the digital recognition memristor prepared as described in an embodiment of the present disclosure in handwritten digital recognition. DETAILED DESCRIPTION

[0053] The present disclosure is further described below in conjunction with the accompanying drawings and embodiments.

[0054] It should be noted that the following detailed descriptions are all illustrative and are intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present disclosure belongs.

[0055] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0056] In the absence of conflict, the embodiments in the present disclosure and the features in the embodiments may be combined with each other.

[0057] Embodiment 1:

[0058] The purpose of this embodiment is to provide a digital recognition memristor.

[0059] The preparation of the resistive layer of a memristor can be divided into physical methods and chemical methods. Physical methods mainly include: magnetron sputtering, thermal evaporation, pulsed laser deposition and other technologies; chemical methods include: hydrothermal method, solution method, spin coating method, chemical vapor deposition and the like. Compared with physical methods, chemical methods often have the advantages of being simple and easy to operate, low cost, and easy to synthesize on a large scale. For example, patent document CN113540346A discloses a memristor based on organic / inorganic composite perovskite and a preparation method thereof, which prepares a template layer on a substrate, then uses spin coating to prepare an organic / inorganic perovskite as a resistive layer, and finally uses vacuum deposition to prepare a top electrode; a memristor with a bottom electrode, a template layer, a resistive layer, and a top electrode structure is obtained.

[0060] Based on the problems existing in the prior art, the present disclosure provides a digital recognition memristor, comprising: a bottom electrode, a dielectric layer and a top electrode; wherein the dielectric layer is located between the bottom electrode and the top electrode, the dielectric layer is a heterojunction composed of zinc oxide nanoparticles (ZnO NPs) and copper oxide nanowires (CuO NWs), the zinc oxide nanoparticles are attached to the copper oxide nanowires, the heterojunction serves as an oxygen vacancy aggregation layer, connects the top electrode and the bottom electrode to form a conductive path, so that the conductivity forms a gradual change process, showing synaptic characteristics.

[0061] Furthermore, the heterojunction acts as an oxygen vacancy aggregation layer, and the resistive switching behavior is controlled by regulating the migration of oxygen vacancies.

[0062] Furthermore, when a positive voltage is applied to the top electrode, the oxygen vacancies in the dielectric layer gather at the top electrode and diffuse from the top electrode to the bottom electrode, forming oxygen vacancy conductive filaments, and the conductive filaments can grow orderly along the nanowire structure; at the same time, electrons transfer from the cathode to the anode, the energy band bends, and the conductivity value increases continuously.

[0063] Furthermore, when negative voltage is applied to the top electrode, electrons return from the anode to the cathode. At the same time, under the action of the internal electric field, the oxygen vacancy conductive filaments begin to break from the thinner bottom electrode, resulting in a continuous decrease in conductivity.

[0064] Furthermore, the synaptic characteristic is specifically manifested as a continuous increase and decrease in conductance value.

[0065] In one or more embodiments, the CuO NWs exist in the form of nanowires, and the ZnO NPs are attached to the copper oxide nanowires in the form of nanoscale particles, thereby forming a digital recognition memristor with a "top electrode / ZnO NPs / CuO NWs / Cu" structure.

[0066] In one or more embodiments, the bottom electrode comprises any one of copper foils with a thickness of 0.1 mm, 0.2 mm, and 0.5 mm. The copper foil with a thickness of 0.2 mm is preferred. Compared with the copper foil with a thickness of 0.1 mm, the copper foil with a thickness of 0.2 mm is not easily damaged, and the copper oxide nanowires grown thereon are more fully than those grown with the copper foil with a thickness of 0.5 mm.

[0067] In one or more embodiments, the top electrode includes any one or more of aluminum, copper, silver, and titanium.

[0068] Furthermore, the working principle of the digital recognition memristor is as follows: the ZnO NPs / CuO NWs heterojunction acts as an oxygen vacancy aggregation layer, and the resistive switching behavior is controlled by regulating the migration of oxygen vacancies. When a positive voltage is applied to the top electrode, the oxygen vacancies in the dielectric layer are gathered at the top electrode and diffuse from the top electrode to the bottom electrode, forming oxygen vacancy conductive filaments, and the conductive filaments can grow in an orderly manner along the nanowire structure, thereby enhancing the stability of the conductive filaments. At the same time, electrons gradually transfer from the cathode to the anode, and the energy band bends, causing a continuous increase in the device's conductance. Conversely, when a negative voltage is applied to the top electrode, electrons gradually return from the anode to the cathode, and at the same time, under the action of the internal electric field, the oxygen vacancy conductive filaments begin to gradually break from the thinner bottom electrode, and the device's conductance value continuously decreases. The process of continuous increase and decrease in conductance is the basic characteristic of the synapse. In addition, when positive / negative voltage pulses are continuously applied to the top electrode, a continuous increase / decrease in conductance occurs due to the migration of oxygen vacancies, which corresponds to long-term potentiation and long-term depression behaviors in synaptic plasticity.

[0069] Embodiment 2:

[0070] The purpose of this embodiment is to provide a method for preparing a digital recognition memristor.

[0071] A method for preparing a digital recognition memristor, comprising:

[0072] The Cu / CuO NWs structure was obtained based on the chemical solution method;

[0073] The ZnO NPs solution was prepared by a hot solvent method;

[0074] Based on the obtained Cu / CuO NWs structure and ZnO NPs solution, a digital recognition memristor based on ZnO nanoparticles / copper oxide nanowire heterojunction was prepared by spin coating.

[0075] Furthermore, the Cu / CuO NWs structure is prepared based on the chemical solution method, specifically:

[0076] Dissolve sodium hydroxide and ammonium persulfate in deionized water simultaneously, perform ultrasonication until they are completely dissolved, and obtain a mixed solution for later use;

[0077] The copper foil was placed in the prepared mixed solution until the copper foil turned blue, and then rinsed with deionized water and dried to obtain a Cu / Cu(OH)2NWs structure;

[0078] The Cu(OH)2NWs were heated to obtain the Cu / CuO NWs structure;

[0079] or,

[0080] The ZnO NPs solution is prepared by the hot solvent method, specifically:

[0081] Dissolve zinc acetate dihydrate in ethanol and stir to dissolve it under heating to obtain a zinc acetate solution for later use;

[0082] Dissolve potassium hydroxide in ethanol and ultrasonicate to obtain a colorless and transparent solution, which is potassium hydroxide solution and is ready for use;

[0083] The potassium hydroxide solution is slowly dripped into the zinc acetate solution under conditions of increasing temperature and vigorous stirring to carry out a hot solvent reaction to obtain a ZnO NPs solution.

[0084] In one or more embodiments, the copper foil is placed in the mixed solution for 10-30 minutes;

[0085] or,

[0086] The conditions for drying the Cu(OH)2NWs are: maintaining at 55-85°C for 10-30min;

[0087] or,

[0088] The growth conditions of the CuO NWs are: 140-180° C. for 120-180 min;

[0089] or,

[0090] The zinc acetate dihydrate is dissolved in ethanol and stirred at a temperature of 60-100°C;

[0091] or,

[0092] The conditions of the hot solvent reaction are: maintaining at 60-100° C. for 90-120 min.

[0093] In one or more embodiments, the digital recognition memristor based on zinc oxide nanoparticle / copper oxide nanowire heterojunction is prepared by spin coating, specifically:

[0094] The ZnO NPs solution was dropped onto CuO NWs, spin-coated and then thermally cured to obtain the Cu / CuO NWs / ZnO NPs structure.

[0095] A top electrode was prepared on the CuO NWs / ZnO NPs heterojunction to obtain a "Cu / CuO NWs / ZnO NPs / top electrode" structure, realizing the preparation of a digital recognition memristor.

[0096] In one or more embodiments, the top electrode of the digital recognition memristor is prepared by magnetron sputtering;

[0097] or,

[0098] The magnetron sputtering parameters are as follows: the base pressure is 5.0×10-3 Pa, working pressure is 0.6-1.0Pa, sputtering temperature is 20-50°C, gas flow rate is 10-30sccm, sputtering energy is 50-80W, pre-sputtering time is 1-2min and sputtering time is 20-60min.

[0099] or,

[0100] When spin coating ZnO NPs, the process parameters of spin coating are: spin coating at a rotation speed of 2500-4500 rpm for 20-60 seconds, and then heating on a hot plate at 70-90° C. for 5-30 minutes.

[0101] In one or more embodiments, the zinc oxide nanoparticle / copper oxide nanowire heterojunction-based memristor is applied in the field of neuromorphic computing, such as digital recognition.

[0102] Specifically, for ease of understanding, the solution disclosed in the present disclosure is described in detail below with reference to the accompanying drawings:

[0103] A method for preparing a digital recognition memristor, the specific steps are as follows:

[0104] A. Preparation of CuO NWs by chemical solution method:

[0105] 1) Dissolve 3.2 g of sodium hydroxide and 0.918 g of ammonium persulfate in 30 ml of deionized water at the same time, and perform ultrasonic treatment until they are completely dissolved to obtain a mixed solution for later use.

[0106] 2) Place a 2cm*2cm copper foil vertically into the prepared mixed solution and react for 10 minutes until the copper foil turns blue, rinse with deionized water and dry at 65°C for 10 minutes to obtain a "Cu / Cu(OH)2NWs" structure;

[0107] 3) heating the Cu(OH)2NWs at 140°C for 120 min to obtain a “Cu / CuO NWs” structure;

[0108] B. Preparation of ZnO NPs solution by hot solvent method:

[0109] 1) Dissolve 0.4895 g of zinc acetate dihydrate in 21 ml of ethanol at 60° C. and stir to dissolve to obtain a zinc acetate solution for later use.

[0110] 2) Dissolve 0.243 g potassium hydroxide in 11.5 ml ethanol and ultrasonicate to obtain a colorless and transparent solution, which is the potassium hydroxide solution and is set aside;

[0111] 3) dropping the potassium hydroxide solution into the zinc acetate solution under vigorous stirring at 60° C. for 10 min, and performing a hot solvent reaction for 100 min to obtain a colorless and transparent ZnO NPs solution;

[0112] C. Digital recognition memristor fabricated by spin coating of zinc oxide nanoparticles / copper oxide nanowire heterojunction:

[0113] 1) 70 μl of the prepared ZnO NPs solution was dropped onto the surface of the CuO NWs, and the solution was spin-coated at 3500 rpm for 40 s, followed by heating on a hot plate at 80° C. for 5 min to obtain a “Cu / CuO NWs / ZnO NPs” structure.

[0114] D. Magnetron sputtering top electrode:

[0115] A 40 nm thick Al top electrode was deposited by DC sputtering in an argon environment. The parameters of magnetron sputtering are as follows: a. The base pressure is 5.0×10 -3 Pa; b, working pressure is 0.9Pa; c, sputtering temperature is 25℃; d, gas flow rate is 20sccm; e, sputtering energy is 70W; f, pre-sputtering time is 100s; g, sputtering time is 40min. The digital recognition memristor of "Cu / CuO NWs / ZnO NPs / Al" structure was obtained.

[0116] Embodiment three:

[0117] The purpose of this embodiment is to provide a method for preparing a digital recognition memristor.

[0118] A method for preparing a digital recognition memristor, the specific steps are as follows:

[0119] A. Preparation of CuO NWs by chemical solution method:

[0120] 1) Dissolve 3.2 g of sodium hydroxide and 0.918 g of ammonium persulfate in 40 ml of deionized water at the same time, and ultrasonicate until completely dissolved to obtain a mixed solution for later use.

[0121] 2) Place a 1cm*1cm copper foil vertically into the prepared mixed solution and react for 15 minutes until the copper foil turns blue, rinse with deionized water and dry at 55°C for 20 minutes to obtain a "Cu / Cu(OH)2NWs" structure;

[0122] 3) heating the Cu(OH)2NWs at 150°C for 130 min to obtain a “Cu / CuO NWs” structure;

[0123] B. Preparation of ZnO NPs solution by hot solvent method:

[0124] 1) Dissolve 0.4895 g of zinc acetate dihydrate in 42 ml of ethanol at 70° C. and stir to dissolve to obtain a zinc acetate solution for later use.

[0125] 2) Dissolve 0.243 g potassium hydroxide in 23 ml ethanol and ultrasonicate to obtain a colorless and transparent solution, which is the potassium hydroxide solution and is set aside;

[0126] 3) dropping the potassium hydroxide solution into the zinc acetate solution under vigorous stirring at 70° C. within 5 minutes, and performing a hot solvent reaction for 90 minutes to obtain a colorless and transparent ZnO NPs solution;

[0127] C. Digital recognition memristor fabricated by spin coating of zinc oxide nanoparticles / copper oxide nanowire heterojunction:

[0128] 1) 50 μl of the prepared ZnO NPs solution was dropped onto the surface of the CuO NWs, and the solution was spin-coated at 3000 rpm for 30 s, followed by heating on a hot plate at 70° C. for 10 min to obtain a “Cu / CuO NWs / ZnO NPs” structure.

[0129] D. Magnetron sputtering top electrode:

[0130] A 40 nm thick Al top electrode was deposited by DC sputtering in an argon environment. The parameters of magnetron sputtering are as follows: a. The base pressure is 5.0×10 -3 Pa; b, working pressure is 1.0Pa; c, sputtering temperature is 35℃; d, gas flow rate is 10sccm; e, sputtering energy is 50W; f, pre-sputtering time is 120s; g, sputtering time is 50min. The digital recognition memristor of "Cu / CuO NWs / ZnO NPs / Al" structure was obtained.

[0131] Embodiment 4:

[0132] The purpose of this embodiment is to provide a method for preparing a digital recognition memristor.

[0133] A method for preparing a digital recognition memristor, the specific steps are as follows:

[0134] A. Preparation of CuO NWs by chemical solution method:

[0135] 1) Dissolve 3.2 g of sodium hydroxide and 0.918 g of ammonium persulfate in 50 ml of deionized water at the same time, and ultrasonicate until completely dissolved to obtain a mixed solution for later use.

[0136] 2) A 3cm*3cm copper foil was placed vertically in the prepared mixed solution and reacted for 20 minutes until the copper foil turned blue. After rinsing with deionized water, it was dried at 70°C for 25 minutes to obtain a "Cu / Cu(OH)2NWs" structure;

[0137] 3) heating the Cu(OH)2NWs at 160°C for 140 min to obtain a “Cu / CuO NWs” structure;

[0138] B. Preparation of ZnO NPs solution by hot solvent method:

[0139] 1) Dissolve 0.4895 g of zinc acetate dihydrate in 10.5 ml of ethanol at 80° C. and stir to dissolve to obtain a zinc acetate solution for later use.

[0140] 2) Dissolve 0.243 g potassium hydroxide in 5.75 ml ethanol and ultrasonicate to obtain a colorless and transparent solution, which is the potassium hydroxide solution and is set aside;

[0141] 3) dropping the potassium hydroxide solution into the zinc acetate solution under vigorous stirring at 80° C. for 15 min, and performing a hot solvent reaction for 110 min to obtain a colorless and transparent ZnO NPs solution;

[0142] C. Digital recognition memristor fabricated by spin coating of zinc oxide nanoparticles / copper oxide nanowire heterojunction:

[0143] 1) 100 μl of the prepared ZnO NPs solution was dropped onto the surface of the CuO NWs, and the solution was spin-coated at 4000 rpm for 50 s, followed by heating on a hot plate at 90° C. for 20 min to obtain a “Cu / CuO NWs / ZnO NPs” structure.

[0144] D. Magnetron sputtering top electrode:

[0145] A 40 nm thick Al top electrode was deposited by DC sputtering in an argon environment. The parameters of magnetron sputtering are as follows: a. The base pressure is 5.0×10 -3 Pa; b, working pressure is 0.6Pa; c, sputtering temperature is 20℃; d, gas flow rate is 30sccm; e, sputtering energy is 80W; f, pre-sputtering time is 90s; g, sputtering time is 60min. The digital recognition memristor of "Cu / CuO NWs / ZnO NPs / Al" structure was obtained.

[0146] Embodiment five:

[0147] The purpose of this embodiment is to provide a method for preparing a digital recognition memristor.

[0148] A method for preparing a digital recognition memristor, the specific steps are as follows:

[0149] A. Preparation of CuO NWs by chemical solution method:

[0150] 1) Dissolve 3.2 g of sodium hydroxide and 0.918 g of ammonium persulfate in 60 ml of deionized water at the same time, and ultrasonicate until completely dissolved to obtain a mixed solution for later use.

[0151] 2) Place a 4cm*4cm copper foil vertically into the prepared mixed solution and react for 30 minutes until the copper foil turns blue, rinse with deionized water and dry at 80°C for 30 minutes to obtain a "Cu / Cu(OH)2NWs" structure;

[0152] 3) heating the Cu(OH)2NWs at 180°C for 180 min to obtain a “Cu / CuO NWs” structure;

[0153] B. Preparation of ZnO NPs solution by hot solvent method:

[0154] 1) Dissolve 0.4895 g of zinc acetate dihydrate in 30 ml of ethanol at 100° C. and stir to dissolve to obtain a zinc acetate solution for later use.

[0155] 2) Dissolve 0.243 g potassium hydroxide in 25 ml ethanol and ultrasonicate to obtain a colorless and transparent solution, which is the potassium hydroxide solution and is set aside;

[0156] 3) dropping the potassium hydroxide solution into the zinc acetate solution under vigorous stirring at 100° C. for 20 minutes, and performing a hot solvent reaction for 120 minutes to obtain a colorless and transparent ZnO NPs solution;

[0157] C. Digital recognition memristor fabricated by spin coating of zinc oxide nanoparticles / copper oxide nanowire heterojunction:

[0158] 1) 150 μl of the prepared ZnO NPs solution was dropped onto the surface of the CuO NWs, and the solution was spin-coated at 4500 rpm for 60 s, followed by heating on a hot plate at 85° C. for 30 min to obtain a “Cu / CuO NWs / ZnO NPs” structure.

[0159] D. Magnetron sputtering top electrode:

[0160] A 40 nm thick Al top electrode was deposited by DC sputtering in an argon environment. The parameters of magnetron sputtering are as follows: a. The base pressure is 5.0×10 -3Pa; b, working pressure is 0.8Pa; c, sputtering temperature is 40℃; d, gas flow rate is 25sccm; e, sputtering energy is 65W; f, pre-sputtering time is 50s; g, sputtering time is 20min. The digital recognition memristor of "Cu / CuO NWs / ZnO NPs / Al" structure was obtained.

[0161] Furthermore, in order to facilitate understanding of the digital identification memristor of the "Cu / CuO NWs / ZnO NPs / Al" structure prepared in the above embodiment, the present disclosure provides a structural schematic diagram of the memristor, as shown in the left figure in Figure 1, wherein the digital identification memristor includes: an Al top electrode; ZnO NPs; CuO NWs; and a Cu bottom electrode; wherein the heterojunction of the ZnO NPs and CuO NWs is located between the Al top electrode and the Cu bottom electrode, and the CuO NWs is a linear structure, and the ZnO NPs are nanoscale particles attached to the CuO NWs.

[0162] Furthermore, in order to verify the characteristics of the digital recognition memristor based on zinc oxide nanoparticle / copper oxide nanowire heterojunction prepared in the present invention for application in neural networks, the present invention takes Example 2 as an example to test the micromorphology of the prepared ZnONPs / CuO NWs heterojunction and the synaptic performance and application results of the memristor, as shown in the figure below.

[0163] FIG1( b ) is a cross-sectional SEM of the ZnO NPs / CuO NWs heterojunction prepared in Example 2, wherein the length of the CuO NWs is about 6 μm and the diameter is about 20 nm.

[0164] Figure 2 This is a TEM characterization image of the ZnO NPs / CuO NWs heterojunction prepared in Example 2. It can be seen from the figure that the synthesized ZnO NPs are attached to the CuO NWs.

[0165] Figures 3(a) and 3(b) show that the ZnO NPs / CuO NWs heterojunction-based device prepared in Example 2 was subjected to 5 consecutive positive and negative 1V voltage scans to observe the change in its current. It can be seen from the figure that the current of the ZnO NPs / CuO NWs heterojunction-based device has a gradual increase or decrease trend rather than a sudden increase or decrease. When positive voltage is applied continuously, the current gradually increases and each time is higher than the previous current. After five consecutive applications, the current increases by ~0.2μA compared to the initial value, which corresponds to the memory process of human brain synapses and has a high memory efficiency. Conversely, after five consecutive applications of negative voltage, the current gradually decreases and each time is slightly smaller than the previous current value. The total change can reach ~6μA. This phenomenon can simulate the forgetting behavior of human brain synapses.

[0166] Figure 4 In order to verify whether the device has efficient neuromorphic computing characteristics, a neural network model combined with the ZnONPs / CuO NWs heterojunction memristor prepared in Example 2 was constructed. As can be seen from the figure: the model is a three-layer neural network, which is an input layer of 784 neurons, a hidden layer and an output layer of 10 neurons. The weights of each layer of neurons are updated by the change of conductance, and the weight change in the neural network is realized by modulating the conductance value with positive and negative voltages.

[0167] Figure 5 (a) and Figure 5 (b) show the recognition accuracy and confusion matrix of the digital recognition memristor of the ZnO NPs / CuO NWs heterojunction prepared in Example 2. It can be seen from the figure that after 1000 iterations of training, the digital recognition accuracy of the digital recognition memristor reached 93%, which is almost close to the recognition accuracy under the ideal state (98%), and the probability of digital recognition of 0-9 can reach more than 50%. It proves that the digital recognition memristor of the ZnO NPs / CuO NWs heterojunction has great development potential in neuromorphic computing, and has opened up a new path for the subsequent combination of memristors and neural networks.

[0168] The digital recognition memristor and the preparation method thereof provided in the above-mentioned embodiment can be realized and have broad application prospects.

[0169] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. For those skilled in the art, the present disclosure may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A digital recognition memristor, characterized in that: include: A bottom electrode, a dielectric layer and a top electrode; wherein the dielectric layer is located between the bottom electrode and the top electrode, the dielectric layer is a heterojunction formed by zinc oxide nanoparticles and copper oxide nanowires, the zinc oxide nanoparticles are attached to the copper oxide nanowires, the heterojunction serves as an oxygen vacancy aggregation layer, connects the top electrode and the bottom electrode to form a conductive path, so that the conductivity forms a gradual process, showing synaptic characteristics; The heterojunction acts as an oxygen vacancy aggregation layer, and the resistive switching behavior is controlled by regulating the migration of oxygen vacancies; The copper oxide nanowires CuO NWs exist in the form of nanowires, and the zinc oxide nanoparticles ZnO NPs are attached to the copper oxide nanowires in the form of nanoscale particles, forming a digital recognition memristor with a "top electrode / ZnO NPs / CuO NWs / Cu" structure; A neural network model combined with ZnO NPs / CuO NWs heterojunction memristor was constructed. The model was a three-layer neural network, including an input layer, a hidden layer, and an output layer. The weights of neurons in each layer were updated by changes in conductance, and the conductance values ​​were modulated by positive and negative voltages.

2. A digital recognition memristor as claimed in claim 1, characterized in that: When a positive voltage is applied to the top electrode, the oxygen vacancies in the dielectric layer gather at the top electrode and gradually diffuse from the top electrode to the bottom electrode, forming oxygen vacancy conductive filaments. The conductive filaments can grow orderly along the nanowire structure. At the same time, electrons transfer from the cathode to the anode, the energy band bends, and the conductivity value increases continuously.

3. A digital recognition memristor as claimed in claim 1, characterized in that: When negative voltage is applied to the top electrode, electrons return from the anode to the cathode. At the same time, under the action of the internal electric field, the oxygen vacancy conductive filaments begin to break from the thinner bottom electrode, resulting in a continuous decrease in conductivity.

4. A digital recognition memristor as claimed in claim 1, characterized in that: The synaptic characteristics are specifically manifested as a continuous increase and decrease in conductance value.

5. A method for preparing a digital recognition memristor, characterized in that: include: The Cu / CuO NWs structure was obtained based on the chemical solution method; The ZnO NPs solution was prepared by a hot solvent method; Based on the obtained Cu / CuO NWs structure and ZnO NPs solution, a digital recognition memristor based on ZnO nanoparticle / CuO nanowire heterojunction was prepared by spin coating. The heterojunction acts as an oxygen vacancy aggregation layer, and the resistive switching behavior is controlled by regulating the migration of oxygen vacancies; The copper oxide nanowires CuO NWs exist in the form of nanowires, and the zinc oxide nanoparticles ZnO NPs are attached to the copper oxide nanowires in the form of nanoscale particles, forming a digital recognition memristor with a "top electrode / ZnO NPs / CuO NWs / Cu" structure; A neural network model combined with ZnO NPs / CuO NWs heterojunction memristor was constructed. The model was a three-layer neural network, including an input layer, a hidden layer, and an output layer. The weights of neurons in each layer were updated by changes in conductance, and the conductance values ​​were modulated by positive and negative voltages.

6. The method for preparing a digital recognition memristor according to claim 5, characterized in that: The Cu / CuO NWs structure is prepared based on the chemical solution method. Dissolve sodium hydroxide and ammonium persulfate in deionized water simultaneously, perform ultrasonication until they are completely dissolved, and obtain a mixed solution for later use; The copper foil was placed in the prepared mixed solution until the copper foil turned blue, and then rinsed with deionized water and dried to obtain a Cu / Cu(OH)2NWs structure; The Cu(OH)2NWs were heated to obtain the Cu / CuO NWs structure; or, The ZnO NPs solution is prepared by the hot solvent method, specifically: Dissolve zinc acetate dihydrate in ethanol and stir to dissolve it under heating to obtain a zinc acetate solution for later use; Dissolve potassium hydroxide in ethanol and ultrasonicate to obtain a colorless and transparent solution, which is the potassium hydroxide solution for later use; The potassium hydroxide solution is slowly dripped into the zinc acetate solution under conditions of heating and vigorous stirring to perform a hot solvent reaction to obtain a ZnO NPs solution.

7. A method for preparing a digital recognition memristor as claimed in claim 6, characterized in that: The copper foil is placed in the mixed solution for 10-30 minutes; or, The conditions for drying the Cu(OH)2NWs are: maintaining at 55-85°C for 10-30min; or, The growth conditions of the CuO NWs are: 140-180° C. for 120-180 min; or, The zinc acetate dihydrate is dissolved in ethanol and stirred at a temperature of 60-100°C; or, The conditions of the hot solvent reaction are: maintaining at 60-100° C. for 90-120 min.

8. The method for preparing a digital recognition memristor according to claim 5, characterized in that: The digital recognition memristor based on zinc oxide nanoparticle / copper oxide nanowire heterojunction is prepared by spin coating method, specifically: The ZnO NPs solution was dropped onto CuO NWs, and then spin-coated and thermally cured to obtain the Cu / CuO NWs / ZnO NPs structure; A top electrode was prepared on the CuO NWs / ZnO NPs heterojunction to obtain a "Cu / CuO NWs / ZnO NPs / top electrode" structure, thereby realizing the preparation of a digital recognition memristor.

9. The method for preparing a digital recognition memristor according to claim 5, characterized in that: The top electrode of the digital recognition memristor is prepared by magnetron sputtering; or, The magnetron sputtering parameters are as follows: the base pressure is 5.0×10 -3 Pa, working pressure is 0.6-1.0Pa, sputtering temperature is 20-50℃, gas flow rate is 10-30sccm, sputtering energy is 50-80W, pre-sputtering time is 1-2min and sputtering time is 20-60min; or, When spin coating ZnO NPs, the process parameters of spin coating are: spin coating at a rotation speed of 2500-4500 rpm for 20-60 seconds, and then heating on a hot plate at 70-90° C. for 5-30 minutes.

Images