Anisotropic optical memory and its preparation method and application

Abstract

This invention discloses an anisotropic optical memory, its fabrication method, and its applications, belonging to the field of visual sensing technology. The optical memory provided by this invention includes a first electrode, a semiconductor nanowire array, and a second electrode; the semiconductor nanowire array is grown on the surface of the first electrode; the semiconductor nanowire array has gradient-distributed defects and a bandgap of 2-4 eV; the second electrode is disposed on the surface of the semiconductor nanowire array opposite to the first electrode and parallel to the first electrode. The optical memory provided by this invention has anisotropic optical storage characteristics, which can effectively improve its viewing angle and depth sensing function, realizing multifunctional integrated design. This invention also provides the fabrication method and applications of the above-mentioned optical memory.

Description

Technical Field

[0001] This invention relates to the field of visual sensing technology, and in particular to an anisotropic optical memory, its fabrication method, and its applications. Background Technology

[0002] With the development of AI (Artificial Intelligence), 5G (5th Generation Mobile Communication Technology), and other emerging technologies, the demand for industry applications in the intelligent vision market is constantly upgrading. In particular, the need for wide field-of-view (FOV) depth perception and motion localization in complex environments is entering a phase of rapid growth, primarily involving fields such as autonomous driving and drones. However, due to hardware and architectural limitations, even the most advanced machine vision systems currently cannot perform wide field-of-view (FOV) depth perception and motion information processing efficiently or compactly, making it difficult to achieve effective and accurate environmental interaction. Traditional machine vision architectures, relying on multi-camera setups (wide-angle lenses, stereo cameras) and additional post-processors for wide-angle image fusion, depth analysis, and motion processing, suffer from fundamental limitations in energy efficiency and system compactness. These von Neumann architectures introduce significant latency and resource overhead, making them unsuitable for resource-constrained edge applications.

[0003] To address these challenges, researchers often draw inspiration from biological vision systems to achieve more efficient computational paradigms and higher area utilization, hoping to replace vision systems based on traditional von Neumann architectures. Emerging neuromorphic devices have attracted widespread attention due to their potential advantages, and these devices are expected to replace hardware in areas such as wide FOV detection, depth sensing, or motion processing. In wide FOV sensing, human eye-inspired wide FOV curved sensor arrays have already achieved wide FOV imaging and simple preprocessing. However, the manufacturing process of curved sensor arrays is complex, resulting in partial image distortion, and they are not yet effective in handling complex processing tasks such as depth sensing. In depth sensing, current cutting-edge neuromorphic technologies employ binocular or Time-of-Flight (ToF) sensors integrated with resistive random access memory (RRAM)-based processing units. However, current depth sensing hardware systems still primarily rely on near-sensor computing strategies, with limitations in integration and efficiency. More importantly, these systems can only achieve single functions and have not yet integrated the required wide FOV depth sensing capabilities into a single unit, making it difficult to meet the needs of compact, low-power edge applications.

[0004] In recent years, neuromorphic vision devices based on the in-sensor computing paradigm have emerged, enabling efficient static image preprocessing and motion preprocessing tasks. These devices significantly simplify hardware design, thereby greatly reducing power consumption and latency. However, state-of-the-art emerging neuromorphic vision sensors can only acquire two-dimensional images, lack depth perception and estimation capabilities, and have narrow detection fields of view, which greatly limits their applications.

[0005] Therefore, for applications such as modern edge autonomous systems (e.g., autonomous driving, unmanned aerial vehicles, smart homes, and patrol robots), there is a need for compact and efficient machine vision solutions that can achieve wide-angle field of view, depth perception, and motion positioning (including direction and angle recognition). There is an urgent need to propose a new type of anisotropic sensor-in-the-loop computing device to achieve multifunctional integrated functionality. Summary of the Invention

[0006] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes an anisotropic optical memory that can effectively improve its viewing angle and depth sensing capabilities, achieving multifunctional integrated functionality.

[0007] The present invention also provides a method for fabricating the above-mentioned optical memory.

[0008] The present invention also provides applications of the above-mentioned optical memory.

[0009] According to an embodiment of a first aspect of the present invention, an anisotropic optical memory is provided, the optical memory comprising a first electrode, a semiconductor nanowire array, and a second electrode; the semiconductor nanowire array is grown on the surface of the first electrode; the semiconductor nanowire array exhibits gradient-distributed defects and has a bandgap of 2~4 eV; the second electrode is disposed on the surface of the semiconductor nanowire array opposite to the first electrode and parallel to the first electrode.

[0010] The optical memory according to embodiments of the present invention has at least the following beneficial effects: The optical memory provided by this invention is a two-terminal neuromorphic device with excellent non-volatile optical storage performance and sensitivity to different angles. Furthermore, it enables in-situ information processing at the sensing end, such as depth perception, making it suitable for artificial intelligence vision systems and achieving integrated sensing, storage, and computing. Specifically, The optical storage mechanism of the aforementioned optical memory is mainly based on the electron capture / decapture mechanism of interface defects. Therefore, the defect concentration and its distribution state largely determine the storage performance of the optical memory. In addition, the polarization, reflection, and refraction of light in the semiconductor also affect the anisotropic optical storage performance of the optical memory to a certain extent. This invention discovers that there is a synergistic effect between the gradient distribution of defects and the existence form of the nanowire array. The combination of the two makes the resulting optical memory exhibit behaviors such as illumination angle dependence, multi-level non-volatility, and light-controlled resistor switching. Based on a single planar device, it can achieve multi-functional integrated integration of depth sensing and motion positioning with a wide field of view (FOV) of not less than 140°, significantly reducing hardware complexity and power consumption. It has great application potential in edge computing that requires real-time motion sensing, such as autonomous driving, smart homes, and patrol robots.

[0011] In summary, the optical memory provided by this invention uses a semiconductor nanowire array as the photosensitive functional layer, possessing anisotropic optical storage capabilities. Therefore, it constructs a planar device with anisotropic photoresistive switching characteristics, which can be further applied to wide-angle depth sensing and motion localization. Compared to traditional hardware that relies on multiple lenses, wide-field-of-view (FOV) cameras, and complex processors, we can achieve wide-FOV depth sensing using a single device, significantly reducing hardware complexity and power consumption.

[0012] According to some embodiments of the present invention, the optical memory further includes a substrate, and the first electrode is disposed on the surface of the substrate. Further, the substrate is a transparent glass substrate.

[0013] According to some embodiments of the present invention, the thickness of the first electrode is >50 nm. For example, it can be 80 nm, 100 nm, 150 nm, 160 nm, 200 nm, 250 nm; or a range of values ​​composed of any two of the above points.

[0014] According to some embodiments of the present invention, the material of the first electrode includes at least one of a transparent conductive metal oxide and a metal. Wherein, The transparent conductive metal oxide includes ITO (indium tin oxide).

[0015] The metal includes at least one of Ti, Pt, Cr, W, Mo, Au, and Pd.

[0016] According to some embodiments of the present invention, the optical memory further includes a seed layer disposed between the first electrode and the semiconductor nanowire array. The seed layer and the semiconductor nanowire array are made of the same material.

[0017] According to some embodiments of the present invention, in the semiconductor nanowire array, the length of a single semiconductor nanowire is 300 nm to 1 μm. For example, it can be 300 nm, 400 nm, 500 nm, 600 nm, 700 nm, 800 nm, 900 nm, 1 μm; or a range of values ​​composed of any two of the above points.

[0018] According to some embodiments of the present invention, in the semiconductor nanowire array, the diameter of a single semiconductor nanowire is 40~120nm. For example, it can be 40nm, 50nm, 60nm, 70nm, 80nm, 90nm, 100nm, 110nm, 120nm; or a range of values ​​composed of any two of the above points.

[0019] According to some embodiments of the present invention, in the semiconductor nanowire array, 1 μm 2 The number of semiconductor nanowires in the projected area is 200-300.

[0020] According to some embodiments of the present invention, in the semiconductor nanowire array, the concentration of the defects gradually decreases from the interface to the end of the semiconductor nanowire array. The interface is the interface where the semiconductor nanowire array contacts the surface of the first electrode.

[0021] According to some embodiments of the present invention, individual semiconductor nanowires are nearly conical from the interface to the end of the semiconductor nanowire array.

[0022] According to some embodiments of the present invention, defects in the semiconductor nanowire array include oxygen vacancies.

[0023] According to some embodiments of the present invention, defects in the semiconductor nanowire array tend to be distributed on the surface of individual nanowires. That is, from the perspective of individual nanowires, there is also a defect gradient from the core to the surface.

[0024] According to some embodiments of the present invention, the bandgap of the semiconductor nanowire array is 3.2~3.5 eV. For example, it can be 3.2 eV, 3.23 eV, 3.3 eV, 3.4 eV, 3.5 eV; or a range of values ​​consisting of any two of the above points.

[0025] According to some embodiments of the present invention, the semiconductor nanowire array is made of at least one of zinc oxide (approximately 3.37 eV) and titanium dioxide (3~3.2 eV; crystal form has an influence). When the material of the semiconductor nanowire array is selected from zinc oxide, the nanowire array grows vertically in layers along the

[002] direction from the surface of the first electrode.

[0026] According to some embodiments of the present invention, the semiconductor nanowire array is anisotropic.

[0027] According to some embodiments of the present invention, the defects in the semiconductor nanowire array are oxygen vacancy defects. The concentration and distribution of these defects are related.

[0028] According to some embodiments of the present invention, the optical memory further includes a semiconductor nanosheet disposed between the semiconductor nanowire array and the second electrode. The semiconductor nanosheet and the semiconductor nanowire array are made of the same material.

[0029] According to some embodiments of the present invention, the area of ​​a single second electrode is (25~200μm)×(25~200μm); for example, it can be 25μm×25μm, 50μm×50μm, 100μm×100μm or 200μm×200μm.

[0030] According to some embodiments of the present invention, the thickness of the second electrode is 15~40nm; for example, it can be 15nm, 20nm, 25nm, 30nm, 35nm, 40nm; or a range of values ​​composed of any two of the above points.

[0031] According to some embodiments of the present invention, the second electrode includes a chromium layer and a gold layer sequentially stacked from the surface of the semiconductor nanowire array.

[0032] According to some embodiments of the present invention, in the second electrode, the thickness of the chromium layer is 5-10 nm; the thickness of the gold layer is 15-40 nm. Specifically, the thickness of the chromium layer is 5 nm, 8 nm, 10 nm, or a range of any two of the above values; the thickness of the gold layer is 15 nm, 20 nm, 25 nm, 30 nm, 35 nm, 40 nm; or a range of any two of the above values.

[0033] According to some embodiments of the present invention, the viewing angle of the optical memory is ≥140°. For example, it can be 140°, 150°, 160°, 170°, 180°; or a range of values ​​composed of any two of the above points.

[0034] According to some embodiments of the present invention, the optical memory has depth sensing capabilities.

[0035] According to some embodiments of the present invention, the optical memory includes a visual sensor.

[0036] According to an embodiment of a second aspect of the present invention, a method for fabricating an optical memory provided in an embodiment of a first aspect of the present invention is provided, the method comprising the following steps: S1. The semiconductor nanowire array is grown on the surface of the first electrode; S2. The second electrode is grown on the surface of the semiconductor nanowire array opposite to the first electrode.

[0037] Since the preparation method adopts all the technical solutions of the optical memory in the above embodiments, it has at least all the beneficial effects brought about by the technical solutions in the above embodiments.

[0038] According to some embodiments of the present invention, step S1 further includes cleaning and drying the first electrode before setting the semiconductor nanowire array. The cleaning method includes sequential ultrasonic cleaning with acetone, ultrasonic cleaning with isopropanol, and ultrasonic cleaning with distilled water. Each ultrasonic cleaning session lasts 5-10 minutes.

[0039] The drying method includes drying with nitrogen gas.

[0040] According to some embodiments of the present invention, the semiconductor nanowire array is made of zinc oxide, and step S1 includes the following sub-steps: S1a. Spin-coat the surface of the first electrode with an ethanol solution of zinc salt and anneal it to obtain a seed layer; S1b. The seed layer is suspended and immersed in a zinc oxide precursor solution and reacted; wherein... The seed layer faces the bottom of the precursor solution; The solutes in the precursor solution include zinc nitrate and C6H. 12 N4. (Chinese name: hexamethylenetetramine, or hexamethylenetetramine, CAS: 100-97-0).

[0041] Because the orientation of the seed layer is set, the growth of the semiconductor nanowire array is a stacked growth, which is equivalent to being carried out under the traction of gravity. The generated semiconductor nanowire array is almost perpendicular to the first electrode (occasionally some nanowires deviate from the vertical direction, with a deviation angle between ±30°). Furthermore, each intermediate layer of the stacked growth is parallel to the first electrode and the second electrode.

[0042] Furthermore, in the above-described growth method, the individual semiconductor nanowires in the resulting semiconductor nanowire array are nearly conical. This results in numerous defects on the sides.

[0043] Therefore, the preparation method of the zinc oxide nanowire array is simple, low-cost, and highly efficient.

[0044] According to some embodiments of the present invention, in step S1a, the zinc salt includes at least one of zinc acetate, zinc gluconate and zinc citrate.

[0045] According to some embodiments of the present invention, in step S1a, the ethanol solution further includes an alkanolamine. This can suppress premature hydrolysis of the zinc salt to form hydroxide precipitate, improve the stability of the ethanol solution, and also facilitate uniform film formation of the seed layer.

[0046] According to some embodiments of the present invention, the alkanolamine includes at least one of ethanolamine and diethanolamine.

[0047] According to some embodiments of the present invention, in the ethanol solution, the molar ratio of the alkanolamine to the zinc salt is 1:1 to 1.2.

[0048] According to some embodiments of the present invention, in step S1a, the concentration of the ethanol solution is 0.01-0.02M. For example, it can specifically be 0.01M, 0.015M, 0.02M; or a range of values ​​composed of any two of the above points.

[0049] According to some embodiments of the present invention, in step S1a, the amount of ethanol solution used is ≥8.5 μL / cm³. 2 For example, it could be 9 μL / cm 2 10μL / cm 2 11μL / cm 2 12μL / cm 2 13μL / cm 2 14μL / cm 2 15μL / cm 2 Or a range of values ​​consisting of any two of the above points. This dosage is calculated based on the amount of material fed in. In actual production, some liquid will be splashed out during the spin coating process. Within the above range, the ethanol solution is sufficient to cover the first electrode and form a uniform coating.

[0050] According to some embodiments of the present invention, in step S1a, the spin coating speed is 3000-4000 rpm. Specifically, it can be 3000 rpm, 3200 rpm, 3400 rpm, 3600 rpm, 3800 rpm, or 4000 rpm; or a range of values ​​consisting of any two of the above points. Within the above speed range, sufficient centrifugal force can be provided to ensure uniform thickness of the resulting coated film.

[0051] According to some embodiments of the present invention, in step S1a, the spin coating time is 30-60 seconds. Specifically, it can be 30 seconds, 40 seconds, 50 seconds, or 60 seconds; or a range of values ​​consisting of any two of the above values. Within the above value range, a film with uniform thickness can be obtained.

[0052] According to some embodiments of the present invention, in step S1a, the thickness of the seed layer is 40-60 nm. For example, it can be 40 nm, 50 nm, 60 nm; or a range of values ​​composed of any two of the above points.

[0053] According to some embodiments of the present invention, in step S1a, the heating rate of the annealing is 5-10℃ / min. For example, it can be 5℃ / min, 8℃ / min, 10℃ / min; or a range of values ​​consisting of any two of the above points.

[0054] According to some embodiments of the present invention, in step S1a, the annealing temperature is 260-300°C. For example, it can specifically be 260°C, 270°C, 280°C, 290°C, 300°C; or a range of values ​​consisting of any two of the above points.

[0055] According to some embodiments of the present invention, in step S1a, the annealing time is 15-20 minutes. For example, it can be 15 minutes, 18 minutes, 20 minutes; or a range of values ​​composed of any two of the above points.

[0056] According to some embodiments of the present invention, in step S1b, the method for preparing the zinc oxide precursor solution includes mixing an aqueous solution of zinc nitrate and C6H... 12 N4 mixture. Among them, The zinc nitrate, water and C6H 12 The N4 ratio is as follows: 1 mmol: 50-100 mL: 1-2 mmol. For example, it can be 1 mmol: 80 mL: 1.5 mmol. The zinc nitrate is anhydrous zinc nitrate or zinc nitrate hexahydrate. This concentration ratio and the reaction time in step S1b affect the growth density of the semiconductor nanowire array and the thickness of individual semiconductor nanowires to a certain extent.

[0057] According to some embodiments of the present invention, in step S1b, the reaction temperature is 80~120℃. Specifically, it can be 80℃, 90℃, 100℃, 110℃, 120℃; or a range of any two of the above values. That is, the present invention prepares zinc oxide nanowire arrays using a low-temperature solution growth method.

[0058] According to some embodiments of the present invention, in step S1b, the reaction time is 15 to 60 minutes. For example, it can be 15 minutes, 20 minutes, 25 minutes, 30 minutes, 40 minutes, 50 minutes, or 60 minutes; or a range of values ​​composed of any two of the above points.

[0059] According to some embodiments of the present invention, step S1b further includes washing the product obtained after the reaction with water and drying it. This removes surface impurities from the obtained product. The drying method includes nitrogen blowing.

[0060] According to some embodiments of the present invention, in step S2, the second electrode is disposed by electron beam evaporation.

[0061] According to an embodiment of the third aspect of the present invention, an optical memory provided by the first aspect of the present invention, or an optical memory prepared by the preparation method provided by the second aspect of the present invention, is provided for application in the fields of autonomous driving, drones, and smart homes.

[0062] Since the application adopts all the technical solutions of the optical memory or preparation method of the above embodiments, it has at least all the beneficial effects brought about by the technical solutions of the above embodiments.

[0063] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. Attached Figure Description

[0064] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a schematic diagram of the structure of the visual sensor obtained in Embodiment 1 of the present invention; Figure 2 This is a flowchart of the preparation process in step S1 of Embodiment 1 of the present invention; Figure 3 The visual sensor obtained in Embodiment 1 of the present invention operates at 200 light intensities of 0.92 mW / cm². 2 1.10mW cm 2 1.32mW cm 2 1.50mW cm 2 and 1.72mW cm 2 Continuous optical pulses (pulse width 8ms, pulse interval 2ms) trigger long duration enhancement (LTP) and 200 consecutive voltage pulses ( The long duration rejection (LTD) process is triggered at 2V, with a read voltage of 0.01V and an illumination angle of 45°. Figure 4The graphs show the changes in photocurrent over time of the visual sensor obtained in Embodiment 1 of the present invention after being stimulated by 100 light pulses (pulse width: 8ms, pulse interval: 2ms) at three different illumination angles (a) 60°, (b) 45° and (c) 20°; and (d) the photocurrent increment as a function of light intensity at the three angles. Figure 5 The visual sensor obtained in Embodiment 1 of the present invention was subjected to 100 consecutive light pulses (pulse width: 8ms, pulse interval: 2ms, light intensity: 1.10mW cm⁻¹) at three illumination angles. 2 ) and 100 voltage pulses ( LTP and LTD behavior excited at 2V, 100ms, using a read voltage of 0.01V; Figure 6 (a) A scene diagram showing a moving object (bird) approaching a drone equipped with the visual sensor obtained in Embodiment 1 of the present invention from 20° / 160°, 45° / 135° and 60° / 120°, and a scene diagram showing light input approaching the visual sensor at different angles and intensities; and (b) a functional relationship between light intensity and the distance between the visual sensor and the light source, corresponding to three distances (x1, x2, x3) in the scene. Figure 7 In Embodiment 1 of the present invention, a moving object is continuously input with three light intensities of P1 = 1.06 mW / cm² from different angles: (a) 20° / 160°, (b) 45° / 135°, and (c) 60° / 120°. -2 P2 = 1.14 mW cm -2 and P3 = 1.23mW cm -2 The current storage state at time; and (d) the corresponding discrimination output current of the moving object at different distances (x1, x2 and x3); Figure 8 According to Figures 6-7 The depth value is calculated from the distance detected and the approach angle. Figure 9 The following are the (a) XRD patterns and (b) photocurrent responses of the zinc oxide nanowire arrays obtained in Example 1 and Comparative Example 1 of this invention (illumination time 1 s, light intensity 0.86 mW cm⁻¹). -2 Comparison chart; Figure 10 Example 1 of the present invention (a) (illumination time is 0.1s, light intensity is 0.86mW / cm²) -2 (a) Example 2 (illumination time 10s, light intensity 0.86mW / cm²) -2 Comparison of photocurrent response of the obtained visual sensors.

[0065] Figure 11 This is a schematic diagram showing the depth of X-ray penetration of the zinc oxide nanowires obtained in Example 1 caused by XPS testing at different incident angles, and a comparison of high-resolution XPS spectra of O 1s at measurement angles of 20°, 45° and 60°.

[0066] Figure 12 This is the high-resolution XPS spectrum of the zinc oxide nanowire array obtained in Example 1 of the present invention at a measurement angle of 90°.

[0067] Figure 13 This is a SEM image of the zinc oxide nanowire array obtained in Example 1 of the present invention.

[0068] Figure 14 This is a TEM image of the zinc oxide nanowire array obtained in Example 1 of the present invention.

[0069] Figure 15 This is a SEM image of the zinc oxide nanowire array obtained in Example 2 of the present invention. Detailed Implementation

[0070] The following will describe the concept and technical effects of the present invention clearly and completely with reference to embodiments, so as to fully understand the purpose, features and effects of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention.

[0071] Example 1 refer to Figure 2 The process described in this example involves fabricating an optical memory, specifically a visual sensor. The specific steps are as follows: S1. A semiconductor nanowire array made of zinc oxide is grown on the surface of the first electrode as a photosensitive functional layer; wherein, the first electrode is an ITO film with a thickness of 160 nm disposed on a transparent glass substrate (substrate); the first electrode is ultrasonically cleaned sequentially with acetone, isopropanol (IPA) and distilled water before use, and then dried with nitrogen gas flow; the duration of each ultrasonic cleaning is 5 min; the total duration of ultrasonic cleaning is 15 min.

[0072] The fabrication process of semiconductor nanowire arrays is as follows: S1a. Use a pipette to apply 30 µL of ethanol solution to the surface of the ITO membrane (approximately 13.3 μL / cm). 2The solution was spin-coated at 3000 rpm for 40 seconds. The ethanol solution contained 0.01 M ZnAc2 (AR, 99.0%, Aladdin, CAS: 557-34-6, zinc acetate) and 0.01 M ethanolamine.

[0073] Subsequently, the ITO film was placed in a muffle furnace and heated to 280°C at a heating rate of 10°C / min and annealed for 20 min to obtain a ZnO seed layer with a thickness of ~50 nm.

[0074] S1b. Add 1 mmol Zn(NO3)2 6H₂O (AR, 99.0%, Aladdin, CAS: 10196-18-6) was dissolved in 50 mL of deionized water, and 1 mmol of C₆H₂O was added. 12 N4 (89%, Alfa, CAS: 100-97-0) was used as a complexing agent to prepare ZnO precursor solution.

[0075] The product obtained in step S1a is suspended and soaked in the precursor solution, ensuring that the seed layer faces the bottom of the precursor solution. Then grow at 90℃ for 30 minutes; Finally, the sample surface was rinsed with plenty of pure water to remove surface impurities, and the surface moisture was dried using a nitrogen gun. Due to gravity, we obtained a uniform and highly upright array of zinc oxide nanowires. The zinc oxide nanowires obtained in this step have a diameter of approximately 50 nm and a length of approximately 500 nm.

[0076] S2. Growth of the second electrode: With the assistance of a mask, a second electrode is deposited on the surface of the zinc oxide nanowire array opposite to the first electrode using electron beam evaporation; wherein, The mask pattern consists of squares with a length of 200 μm, arranged in an array with a spacing of 300 μm between adjacent squares; the resulting second electrode is an array of squares with a side length of 200 μm.

[0077] The second electrode consists of a 5nm Cr layer and a 30nm Au layer superimposed from the zinc oxide nanowire array.

[0078] The structural diagram of the vision sensor fabricated in this example is shown below. Figure 1 As shown, a visual sensor device with storage at both ends of Au / zinc oxide nanowire array / ITO was constructed.

[0079] Example 2 This example demonstrates the fabrication of a visual sensor, which differs from Example 1 in that: In step S1b, the growth time is 60 min; the length of the obtained zinc oxide nanowires is about 750 nm, and nanosheets are grown on the top of the nanowire array.

[0080] Comparative Example 1 This example demonstrates the fabrication of a visual sensor, which differs from Example 1 in that: In step S1b, the product obtained in step S1a is placed directly at the bottom of the precursor solution, with the seed layer facing the surface of the precursor solution.

[0081] Test case This example tests the physicochemical properties of the vision sensors obtained in the embodiments and comparative examples. Specifically, XRD patterns, SEM, and TEM results of the zinc oxide nanowire array show that, in the preparation process provided in Example 1, due to the special orientation of the seed layer in step S1b, the zinc oxide nanowire array exhibits significant anisotropy during growth due to its own gravitational pull. Specifically, it grows vertically and uniformly along the

[002] direction, with each layer parallel to the first / second electrode. Along the growth direction, each individual semiconductor nanowire is nearly conical (i.e., gradually thins), measuring 1 μm. 2 The number of semiconductor nanowires in the projected area is 200-300; SEM and TEM images of the zinc oxide nanowire array obtained in Example 1 are shown below. Figure 13 and 14 As shown. In Example 2, the growth rate of the zinc oxide nanowire array gradually slows down due to the gradual consumption of the precursor, giving the crystal nuclei sufficient time to grow laterally, thereby forming semiconductor nanosheets at the top of the semiconductor nanowire array, specifically as follows. Figure 15 As shown in Comparative Example 1, the gravitational pull is absent, and the crystals in the nanowires grow isotropically without a clear orientation. This means that the layered growth in the nanowires cannot be guaranteed to be parallel to the ZnO seed layer.

[0082] Because zinc oxide nanowire arrays grow almost vertically, the depth reached by X-rays varies depending on the angle at which XPS testing is performed. Since XPS is a surface and shallow surface test, it can be used to test the surface defect distribution at different depths within the zinc oxide nanowire array. The XPS spectra of O 1s show that the percentages of oxygen vacancies at angles of 20°, 45°, 60°, and 90° are 16.5%, 23.9%, 29.4%, and 39.4%, respectively. This demonstrates that oxygen vacancy defects in the zinc oxide nanowire array are gradient-distributed, specifically decreasing in concentration from the seed layer to the end of the array. Specific test results are as follows: Figures 11-12 As shown.

[0083] This example also tests the visual sensing performance of the vision sensors obtained in the embodiments and comparative examples. Specific test items and results are shown below: The non-volatile optical storage performance of the neuromorphic vision sensor provided in the examples and comparative examples was tested. Photoelectric performance was tested in room temperature air using a semiconductor analyzer (Keithley 2636B). The first electrode was grounded, and a voltage of 0.01V was applied to the second electrode as the readout voltage. A pulse generator (Tektronix, AFG 31000) was used to precisely control the width, interval, and number of light pulses from the ultraviolet laser source; unless otherwise specified, the wavelength of the source was 380nm.

[0084] The optical storage performance of the visual sensors obtained in Example 1 and Comparative Example 1 is as follows: Figure 9 As shown, when subjected to continuous illumination for 1 second (light intensity at 45° is 1.1 mW / cm²), -2 In Example 1 (Device 1), the current of the vision sensor increases continuously over time and then stabilizes over a long period of time; while in Comparative Example 1 (Device 2), there is almost no change in current, that is, there is almost no light response phenomenon.

[0085] The test results of the multi-level optical storage behavior of the vision sensor obtained in Examples 1 and 2 (Device 3) are as follows: Figure 10 As shown. Specifically, light (380nm, 0.86mW cm⁻¹) was applied to both. -2 The results showed that after a single 0.1s (pulsed light, 100Hz) illumination was applied to the visual sensor obtained in Example 1, the current increased linearly with time, and it exhibited obvious non-volatility after the light was removed; while after a single 10s current was applied to the device obtained in Example 2, the current increased slowly with time, and it had some storage performance after the light was removed, but its non-volatility performance was slightly worse than that of Example 1.

[0086] The test results of the optically triggered LTP and electronically controlled LTD of the vision sensor obtained in Example 1 are as follows: Figure 3 As shown. During the optically triggered LTP (Long-Term Enhancement) characteristic test, 200 consecutive optical pulses were used, with a pulse width and interval of 8ms and 2ms, respectively. During the electronically controlled LTD (Long-Term Suppression) characteristic test, 200 constant voltage pulses were applied, with a pulse amplitude of... 2V. The results show that the vision sensor obtained in this embodiment exhibits highly linear LTP (Long-Term Duration Enhancement) with 200 different states during LTP testing. Furthermore, the rate of resistance change during LTP can be observed with different light intensities (0.92 mW cm⁻¹). -2 1.10mW cm -2 1.32mW cm-2 1.50mW cm -2 1.72mW cm -2 The resistance change rate is higher under higher light intensity by adjusting the resistance.

[0087] The visual sensor obtained in Example 1 exhibits LTP behavior and retention characteristics induced by light of different intensities at different illumination angles (20°, 45°, 60°), such as... Figure 4 As shown. When 100 consecutive light pulses (pulse width: 8ms, light intensity: 0.78~1.41mW cm⁻¹) are applied... -2 The visual sensor exhibits highly linear conductance enhancement under continuous light pulses, and maintains good conductivity even after the light is removed. The photocurrent increment of the visual sensor obtained in Example 1 increases sequentially from 45°, 20° to 60°, demonstrating that the visual sensor obtained in this embodiment of the invention has obvious anisotropic photocontrolled resistive switching behavior under different illumination angles.

[0088] Figure 5 The image shows the visual sensor obtained in Example 1 under different illumination angles (60°, 45°, 20°) with 100 consecutive light pulses (light intensity: 1.1 mW cm⁻¹). 2 The LTP process under ) and the application of 100 consecutive voltage pulses (amplitude: The LTD process was performed at 2V (duration: 100ms). Near-linear current response was observed at all three angles, with photocurrent modulation ranging from 60° to 20° to 45°. Under reverse bias, the current state gradually returned to its initial high-resistivity state. Thus, the vision sensor obtained in this embodiment exhibits anisotropy under different illumination angles, as well as multi-stage non-volatile photocontrolled resistive switching and LTP behavior; therefore, it can be used for wide FOV depth sensing.

[0089] Based on a single vision sensor at different angles (60° / 120°, 45° / 135° and 20° / 160°) and a certain light intensity (0.78~1.41W cm⁻¹), -2 The light pulse applied to the device has different photocurrent responses within a certain range, thus the angle can be determined based on its current response. Furthermore, the distance (x) from the moving object to the device has a certain exponential relationship with the light intensity (P) of the light pulse applied to the device. The depth value at a certain distance and angle can be calculated using the formula (d=xsinθ, d: depth, θ: angle between the object and the device plane). In addition, the visual sensor provided by this invention has multi-level non-volatile light properties, thus the above theory can be used to demonstrate angle recognition and depth perception.

[0090] Figure 6This describes a real-world scenario where a bird approaches a drone equipped with a visual sensor at three distances (x1 = 1.56 m, x2 = 1.2 m, and x3 = 0.92 m; these distances are merely illustrative and not strictly limited in actual use) at three different angles (20° / 160°, 45° / 135°, and 60° / 120°). As the bird approaches the drone from different angles, the current response of the visual sensor varies with the distance from the bird to the sensor (x1 to x3). Figure 6 (b) shows the relationship between the measured distances from the light source to the device (X1, X2, and X3) and the input light intensity of the device. The comparison shows that the actual distance and the measured distance are directly proportional; this is equivalent to testing the light intensity input to the device from the light source at different distances, which range from 0 to 5.2 mm. Figure 7 As shown in (ac), the distances of the moving object (bird) at x1, x2, and x3 correspond to the input intensity P1 (1.06 mW cm⁻¹) incident on the visual sensor. 2 P2 (1.14mW cm) 2 ) and P3 (1.23mW cm 2 The corresponding discrimination output current at different angles and distances (coordinates x, θ) is as follows: Figure 7 As shown in (d), the corresponding distance and angle can be obtained from the output current. This vision sensor's ability to sense distance and angle allows it to accurately estimate the depth of an object using a depth calculation formula, such as... Figure 8 As shown, nine distinguishable depths are displayed at different distances and angles. Based on the above tests, it is clear that the visual sensor provided by this invention has a viewing angle greater than 140° and can simultaneously perceive depth.

[0091] In summary, the visual sensor provided by this invention can simultaneously achieve depth and wide-angle sensing, and is expected to be widely used in autonomous driving and drones.

[0092] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments, and various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention. Furthermore, the embodiments of the present invention and the features thereof can be combined with each other unless otherwise specified.

Claims

1. An anisotropic optical memory, characterized in that, The optical memory includes a first electrode, a semiconductor nanowire array, and a second electrode. The semiconductor nanowire array is grown on the surface of the first electrode; the semiconductor nanowire array has defects with a gradient distribution and a band gap of 2~4 eV; The second electrode is disposed on the surface of the semiconductor nanowire array opposite to the first electrode and parallel to the first electrode.

2. The optical memory of claim 1, wherein, In the semiconductor nanowire array, the length of a single semiconductor nanowire is 300 nm to 1 μm; and / or, the diameter of a single semiconductor nanowire in the semiconductor nanowire array is 40 to 120 nm.

3. The optical memory of claim 1, wherein, The thickness of the first electrode is >50 nm; and / or the material of the first electrode includes at least one of transparent conductive metal oxide and metal.

4. The optical memory of claim 1, wherein, The thickness of the second electrode is 15~40nm; And / or, the area of ​​a single second electrode is (25~200μm)×(25~200μm); And / or, the second electrode includes a chromium layer and a gold layer sequentially stacked from the surface of the semiconductor nanowire array.

5. The optical memory according to any one of claims 1 to 4, characterized in that The bandgap of the semiconductor nanowire array is 3.2~3.5eV; and / or the material of the semiconductor nanowire array includes at least one of zinc oxide and titanium dioxide.

6. The optical memory according to any one of claims 1 to 4, wherein The optical memory has a viewing angle ≥140°; and / or, the optical memory has a depth sensing function.

7. A method of producing an optical memory as claimed in any one of claims 1-6, characterized in that The preparation method includes the following steps: S1. The semiconductor nanowire array is grown on the surface of the first electrode; S2. The second electrode is grown on the surface of the semiconductor nanowire array opposite to the first electrode.

8. The preparation method according to claim 7, characterized in that, The semiconductor nanowire array is made of zinc oxide, and step S1 includes the following sub-steps: S1a. Spin-coat an ethanol solution of zinc salt onto the surface of the first electrode and anneal it to obtain a zinc oxide seed layer; S1b. The seed layer is suspended and immersed in a zinc oxide precursor solution and reacted; wherein... The seed layer faces the bottom of the precursor solution; The solute of the precursor solution includes zinc nitrate and C6H 12 N4.

9. The preparation method according to claim 8, characterized in that, In step S1b, the reaction temperature is 80~120℃; and / or the reaction duration is 15~60min.

10. An optical memory as described in any one of claims 1 to 6, or an optical memory prepared by any one of claims 7 to 9, in the fields of autonomous driving, drones, and smart homes.

Images