Multimode sensing memristor array and preparation method thereof

Abstract

The invention discloses a memristor array based on multimode sensing and a preparation method, and belongs to the field of microelectronic material devices. The structure is composed of a sensor array and a memristor array. The sensor array comprises a sensor array electrode layer and a sensor array piezoresistive layer, and the memristor array comprises a memristor array bottom electrode layer, a memristor array resistive layer and a memristor array top electrode layer; the sensor array and the memristor array are connected through a memristor array top electrode layer electrode. The sensor array receives external pressure signals, and pressure information is learned and understood through memristor current response. Meanwhile, the memristor array resistive layer has good response to light / temperature signals, can simulate the human vision / temperature perception memory process, and provides a basis for realizing a multi-mode function. And potential application prospects are shown in the fields of intelligent robots, electronic skin, artificial intelligence and the like.

Description

Technical Field

[0001] This invention belongs to the field of microelectronics technology, and in particular relates to a multimode sensing memristor array and its fabrication method. Background Technology

[0002] Artificial multimodal sensing devices simulate human senses through effects such as light, electricity, magnetism, and heat, enabling the sensing, memorization, and processing of various external stimuli. They hold significant application value in advanced electronics and artificial intelligence. Humans receive over 80% of information from the natural environment through vision, while touch provides more fundamental information about objects, including their weight, shape, temperature, texture, and material. Traditional single-modal sensing signal input has inherent limitations, easily leading to system misjudgments. In contrast, biological nervous systems achieve accurate environmental cognition through multimodal collaborative perception, providing important insights for artificial sensing systems. A memristor-based multimodal fusion architecture significantly improves target recognition accuracy through cross-modal feature extraction and information processing. Device systems with multimodal sensing and processing capabilities represent the future direction of in-memory computing systems. Summary of the Invention

[0003] The purpose of this invention is to address the problem that most devices can only process single-mode physical signals, and to provide a method for fabricating a multi-mode sensing memristor array. This sensing memory system, used for multi-modal fusion and information processing, is of great significance for intelligent robots, neuroscience, and brain-computer interfaces.

[0004] A multi-mode sensing memristor array and its fabrication method include: a substrate 1, a sensor array, and a memristor array;

[0005] The sensor array includes: a sensor array electrode layer 2 and a sensor array piezoresistive layer 3;

[0006] The memristor array includes: a top electrode layer 4 of the memristor array, a resistive switching layer 5 of the memristor array, and a bottom electrode layer 6 of the memristor array;

[0007] The sensor array electrode layer 2 is connected to the memristor array top electrode layer 6;

[0008] The sensor array electrode layer 2 and the memristor array top electrode layer 4 are both made of active metal copper or silver, with a thickness of 100±10nm.

[0009] The piezoresistive layer of the sensor array is prepared using transition metal oxide nanoribbons, germanium, indium α-selenide or gallium arsenide, with a thickness of 100±20 μm.

[0010] The bottom electrode layer 6 of the memristor array is made of an inert metal;

[0011] The bottom electrode layer 6 of the memristor array is made of platinum, gold or tungsten, and has a thickness of 100±10nm;

[0012] The memristor array resistive switching layer 5 is a transition metal oxide nanoribbon, titanium oxide, zinc oxide, gallium arsenide, graphene, methylammonium lead iodide, molybdenum disulfide or polyaniline, with a thickness of 1-3 μm.

[0013] The sensor array electrode layer 2 and the memristor array top electrode layer 4 are integrated and both are made of silver with a thickness of 100±10nm.

[0014] The piezoresistive layer 3 of the sensor array is made of orthogonal molybdenum trioxide nanoribbons (α-MoO3 NBs) with a thickness of 100±20 μm;

[0015] The memristor array resistive switching layer 5 is made of α-MoO3;

[0016] The bottom electrode layer 6 of the memristor array is made of gold (Au).

[0017] A method for fabricating a multimode sensing memristor array includes: 1) fabricating a mask according to the designed shape; 2) sequentially depositing a memristor array bottom electrode layer 6, a memristor array resistive switching layer 5, a memristor array top electrode layer 4, and a sensor array piezoresistive layer 3 on a substrate 1.

[0018] The memristor array bottom electrode layer 6, the memristor array top electrode layer 4, and the sensor array electrode layer 2 are prepared by thermal evaporation or magnetron sputtering.

[0019] The memristor array resistive switching layer 5 is prepared by magnetron sputtering, chemical vapor deposition, drop coating or spin coating.

[0020] The piezoresistive layer 3 of the sensor array is prepared by depressurization filtration, Czochralski method, chemical vapor deposition, physical vapor deposition or molecular beam epitaxy.

[0021] The application of a multimode sensing memristor array in processing pressure, light, and temperature signals is described.

[0022] This invention provides a multi-mode sensing memristor array and its fabrication method, belonging to the field of microelectronic materials and devices. Its structure consists of two parts: a sensor array and a memristor array. The sensor array includes a sensor array electrode layer and a sensor array piezoresistive layer, while the memristor array includes a memristor array bottom electrode layer, a memristor array resistive switching layer, and a memristor array top electrode layer. The sensor array and the memristor array are connected via electrodes in the top electrode layer of the memristor array. The sensor array receives external pressure signals and learns and understands pressure information through the memristor current response. Simultaneously, the memristor array resistive switching layer exhibits good response to light / temperature signals, which can simulate the human visual / temperature perception and memory process, providing a foundation for realizing multi-modal functions. It shows potential application prospects in fields such as intelligent robots, electronic skin, and artificial intelligence. Compared with existing technologies, this invention, based on a multi-mode sensing memristor array, integrates the sensor array and the memristor array, overcoming the limitation that most current devices can only process single-mode physical signals. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of a multi-mode sensing memristor array structure; in the diagram: 1. Substrate; 2. Sensor array electrode layer; 3. Sensor array piezoresistive layer; 4. Memristor array top electrode layer; 5. Memristor array resistive switching layer; 6. Memristor array bottom electrode layer.

[0024] Figure 2 It involves writing the letter "L" and a 3D spatial pressure mapping on a 3×3 sensor array of a multimode sensing memristor array.

[0025] Figure 3 It is a current curve obtained by writing the letter "L" three times consecutively on a 3×3 sensor array of a multimode sensing memristor array.

[0026] Figure 4 This is a current response curve of an ultraviolet light signal applied to the memristor array in the multi-mode sensing memristor array of the embodiment.

[0027] Figure 5 These are the current response curves of the multimode sensing memristor array under the stimulation of two light pulses and two pressure pulses, corresponding to the input combination modes of "1111", "1100" and "1000".

[0028] Figure 6 It is the current state of the multimode sensing memristor array after being stimulated by sixteen different light pulses and pressure pulses. Detailed Implementation

[0029] Example 1: A method for fabricating a multi-mode sensing memristor array.

[0030] like Figure 1The diagram shown illustrates a structure based on a multi-mode sensing memristor array used in this embodiment. The structure comprises a tactile sensor array and a memristor array, involving the fabrication of orthorhombic molybdenum trioxide nanoribbons (α-MoO3 NBs) / Ag tactile sensor arrays and Ag / α-MoO3 / Au structure memristor arrays.

[0031] A multi-mode sensing memristor array includes: a substrate 1, a sensor array, and a memristor array;

[0032] The sensor array includes: a sensor array electrode layer 2 and a sensor array piezoresistive layer 3.

[0033] The memristor array includes: a top electrode layer 4 of the memristor array, a resistive switching layer 5 of the memristor array, and a bottom electrode layer 6 of the memristor array;

[0034] The sensor array electrode layer 2 and the memristor array top electrode layer 4 are integrated and both are made of silver with a thickness of 100±10nm.

[0035] The piezoresistive layer 3 of the sensor array is made of orthogonal molybdenum trioxide nanoribbons (α-MoO3 NBs) with a thickness of 100±20 μm;

[0036] The memristor array resistive switching layer 5 is made of α-MoO3;

[0037] The bottom electrode layer 6 of the memristor array is made of gold (Au).

[0038] Example 2: A method for fabricating a multimode sensing memristor array

[0039] A method for fabricating a multimode sensing memristor array includes the following steps:

[0040] Step 1: Fabricate the bottom electrode Au of the memristor array using a thermal evaporation method.

[0041] Specifically, a grating mask is deposited on a polyimide (PI) substrate, and the bottom electrode Au of the memristor array is fabricated using a thermal evaporation vacuum deposition machine. The thermal evaporation conditions are as follows: the chamber vacuum degree is greater than 10. -4 Pa; the purity of the Au wire (0.03 g) used for vapor deposition is 99.99%; the heating current is 110 A;

[0042] Step 2: Prepare the resistive switching layer of the memristor array using the drop-coating method.

[0043] Specifically, 10 mg of α-MoO3 NBs was dissolved in 0.4 mL of ethanol to prepare a 25 mg / mL α-MoO3 ethanol solution, which was then sonicated for 15 min. The solution was then drop-coated onto a PI film with a bottom electrode Au using a syringe. After the ethanol evaporated, an α-MoO3 NBs thin film was obtained, which served as the resistive switching layer of the memristor array.

[0044] Step 3: Fabricate the memristor array top electrode Ag and the sensor array electrode layer Ag using a thermal evaporation method.

[0045] The thermal evaporation conditions are as follows: the chamber vacuum degree is greater than 10. -4 Pa; the purity of the evaporation material Ag wire (0.03 g) is 99.99%; the heating current is 90 A; specifically, the interdigitated electrode mask is covered on the PI film, and the top electrode Ag of the memristor array is prepared by using a thermal evaporation vacuum coating machine.

[0046] Step 4: Prepare the piezoresistive layer for the tactile sensor array using the reduced pressure filtration method.

[0047] Specifically, a certain amount of dried α-MoO3 NBs powder was weighed and added to 50 mL of deionized water. The mixture was ultrasonically treated for 15 minutes to ensure thorough dispersion, thus obtaining an α-MoO3 NBs aqueous dispersion. The dispersion was then filtered under reduced pressure using a sand core filter with an ethyl cellulose membrane having a diameter of 47 mm and a pore size of 220 nm. After filtration, the membrane was carefully peeled off and tightly adhered to the surface of an Ag electrode. It was then placed in a vacuum environment at 120°C and dried for 12 hours.

[0048] Figure 2 It involves writing the letter "L" and a 3D spatial pressure mapping on a 3×3 sensor array of a multimode sensing memristor array.

[0049] like Figure 2 As shown, the letter "L" is handwritten on a 3×3 sensor array of a multimode sensing memristor array. The change in pressure is converted into a change in current. The change in current is then input into the memristor array, changing the voltage across the memristor. The conductance of the memristor is set to a fixed conductance state, thereby storing the pressure information.

[0050] Figure 3 It is a current curve obtained by writing the letter "L" three times consecutively on a 3×3 sensor array of a multimode sensing memristor array.

[0051] like Figure 3As shown, the letter "L" was written three times consecutively on the 3×3 sensor array of the multimode sensing memristor array. Analysis of the current curves reveals that after each handwriting operation, the current did not return to its initial state but remained stable in the intermediate state. This indicates that the conductance information was still stored. This demonstrates that the multimode sensing memristor array can effectively sense and store tactile information.

[0052] Figure 4 This is a graph showing the current response of a memristor array when an ultraviolet light signal is applied to it in a multimode sensing memristor array.

[0053] like Figure 4 As shown, ultraviolet light stimulation with an intensity of 5.2 mW / cm² and a pulse width of 1 s was applied to the memristor array in the multimode sensing memristor array. It can be seen that upon input of the ultraviolet light signal, the memristor rapidly generates a photoresponse current, which decays to an intermediate state different from the initial state after the peak. This multimode sensing memristor array exhibits significant optical memristor characteristics, similar to the excitatory post-synaptic current (EPSC) of biological synapses.

[0054] Figure 5 It shows the current response of “1111”, “1100” and “1000” under the stimulation of two optical pulses and two pressure pulses.

[0055] like Figure 5 As shown, two ultraviolet light pulses are first applied to the memristors in the multimode sensing memristor array, and then two pressure pulses are applied to the tactile sensor to obtain the current response of "1111", "1100" and "1000", where "1" indicates that a pulse is applied and "0" indicates that no pulse is applied.

[0056] Figure 6 It is the current state of the multimode sensing memristor array after being stimulated by sixteen different light pulses and pressure pulses.

[0057] like Figure 6 As shown, after applying two ultraviolet light pulses and two pressure pulses to the multimode sensing memristor array, sixteen current states can be read by the EPSC output by the memristor.

Claims

1. A multi-mode sensing memristor array and its fabrication method, comprising: Substrate (1), sensor array and memristor array; The sensor array includes: a sensor array electrode layer (2) and a sensor array piezoresistive layer (3); The memristor array includes: a top electrode layer (4) of the memristor array, a resistive switching layer (5) of the memristor array, and a bottom electrode layer (6) of the memristor array. The sensor array electrode layer (2) is connected to the memristor array top electrode layer (6).

2. The multi-mode sensing memristor array according to claim 1, characterized in that: The sensor array electrode layer (2) and the memristor array top electrode layer (4) are both made of active metal copper or silver, with a thickness of 100±10nm.

3. A multi-mode sensing memristor array according to claim 2, characterized in that: The piezoresistive layer of the sensor array is prepared using transition metal oxide nanoribbons, germanium, indium α-selenide, or gallium arsenide, with a thickness of 100±20 μm.

4. The multi-mode sensing memristor array according to claim 3, characterized in that: The bottom electrode layer (6) of the memristor array is made of inert metal.

5. The multi-mode sensing memristor array according to claim 4, characterized in that: The memristor array resistive switching layer (5) is a transition metal oxide nanoribbon, titanium oxide, zinc oxide, gallium arsenide, graphene, methylammonium lead iodide, molybdenum disulfide, or polyaniline.

6. The multi-mode sensing memristor array according to claim 5, characterized in that: The bottom electrode layer (6) of the memristor array is made of platinum, gold or tungsten and has a thickness of 100±10nm.

7. A multi-mode sensing memristor array according to claim 6, characterized in that: The sensor array electrode layer (2) and the memristor array top electrode layer (4) are integrated and both are made of silver; The piezoresistive layer (3) of the sensor array is orthogonal molybdenum trioxide nanoribbons α-MoO3 NBs; The memristor array resistive switching layer (5) is α-MoO3 with a thickness of 1-3 μm; The bottom electrode layer (6) of the memristor array is an interdigitated electrode made of gold.

8. The method for fabricating a multimode sensing memristor array according to claim 1, characterized in that: 1) Make a mask according to the design shape; 2) On the substrate (1), deposit the bottom electrode layer (6) of the memristor array, the resistive switching layer (5) of the memristor array, the top electrode layer (4) of the memristor array, and the piezoresistive layer (3) of the sensor array in sequence.

9. The method for fabricating a multimode sensing memristor array according to claim 8, characterized in that: The memristor array bottom electrode layer (6), the memristor array top electrode layer (4), and the sensor array electrode layer (2) are prepared by thermal evaporation or magnetron sputtering. The memristor array resistive switching layer (5) is prepared by magnetron sputtering, chemical vapor deposition, drop coating or spin coating; The sensor array piezoresistive layer (3) is prepared by depressurization filtration, Czochralski method, chemical vapor deposition, physical vapor deposition or molecular beam epitaxy.

10. The use of a multimode sensing memristor array as described in claim 1 in processing pressure, light, and temperature signals.

Images