Active layer, device, array and visual system based on light control free radical bipolar doping

Abstract

The invention discloses an active layer based on light regulation free radical bipolar doping, a device, an array and a visual system. The problems that existing neuromorphic vision hardware is low in integration level, single in function and difficult to meet complex requirements are solved. The active layer comprises an organic semiconductor material and a free radical dopant, and wavelength selective reversible bipolar doping can be realized: p-type doping is presented and excitatory response is generated under 808 nm irradiation, and n-type doping is presented and inhibitory response is generated under 450 nm irradiation. Based on the active layer, an optical reconfigurable bipolar neuromorphic device with a simple structure is constructed, and the device has dynamic synaptic characteristics and light-controlled bipolar photoelectric response at the same time. The device is used as a pixel unit, can be integrated on a large scale to form a neuromorphic visual array, and forms a complete system with a signal reading and processing unit. The system can synchronously execute multiple visual functions of motion detection, spatio-temporal information memory and contrast enhancement feature extraction at a sensing front end.

Description

Technical Field

[0001] This invention relates to the field of neuromorphic devices and biomimetic visual sensors, specifically to an active layer, device, array, and visual system based on photo-controlled free radical bipolar doping. Background Technology

[0002] Neuromorphic vision aims to mimic biological visual systems, integrating perception, memory, and processing functions into the sensing front end to achieve high-efficiency, low-latency parallel information processing. This provides a crucial hardware development path for next-generation machine vision, such as autonomous driving and industrial inspection. However, most current neuromorphic vision hardware is limited to single functions, such as motion detection or visual memory, and struggles to meet the demands of complex and dynamic environments. Addressing these challenges requires neuromorphic vision hardware that directly integrates multiple perception and processing functions at the array level. Therefore, developing such multifunctional hardware within a unified system is crucial for advancing machine vision towards greater autonomy, adaptability, and intelligence.

[0003] The key to building such highly integrated hardware platforms lies in device units with multi-response characteristics. Among these, time-varying synaptic properties and optically reconfigurable excitation-inhibition bipolar response capabilities are particularly crucial: the former is responsible for encoding motion information and realizing visual memory, while the latter enables the hardware implementation of retinal-like central-peripheral receptive fields, thereby efficiently extracting edge features and enhancing visual information. Currently, methods for integrating these two functions into a single device mainly rely on interface or defect-state capture mechanisms, but face three fundamental challenges. First, the multi-layered heterogeneous structure design required for interface capture significantly increases process complexity, hindering large-scale integration. Second, reliance on randomly distributed defect-state capture results in poor response controllability, severely limiting the uniformity and reliability of the hardware platform. Finally, the inherently weak optical response severely degrades during pixel miniaturization, hindering effective array-level detection. These limitations collectively impede the progress of large-scale, multifunctional neuromorphic vision systems. Summary of the Invention

[0004] This invention aims to overcome the technical challenges of existing neuromorphic vision hardware, such as difficulties in integration and limited functionality, and provides an innovative solution covering the entire chain from method and device to system. The core of this invention lies in proposing and utilizing a novel strategy of photo-controlled free radical bipolar doping. Through a simple material system and device structure, it achieves the goal of integrating dynamic synaptic properties and photo-reconstructed bipolar response in a single device, and successfully constructs a large-scale, multifunctional integrated neuromorphic vision system.

[0005] The technical problem solved by this invention is achieved through the following technical solution:

[0006] A first aspect of the present invention is to provide an active layer based on photo-controlled radical bipolar doping, comprising:

[0007] Organic semiconductor materials; and

[0008] Free radical dopants dispersed in the organic semiconductor material;

[0009] The active layer is configured such that: under illumination of a first wavelength, the free radical dopant captures photogenerated charges, causing the active layer to exhibit a first conductivity type doped state and generating a photoresponse of a first polarity; under illumination of a second wavelength, the free radical dopant releases the captured charges, causing the active layer to exhibit a second conductivity type doped state opposite to the first conductivity type and generating a photoresponse of a second polarity opposite to the first polarity.

[0010] Furthermore, the first wavelength is located in the near-infrared band, and the second wavelength is located in the visible light band; preferably, the first wavelength is 808 nm, and the second wavelength is 450 nm.

[0011] Further, the organic semiconductor material is a donor-acceptor type conjugated polymer; preferably, the organic semiconductor material is a DPP-DTT polymer having the structure shown in Formula I:

[0012]

[0013] Formula I

[0014] Furthermore, the radical dopant is a galvanoxy radical; preferably, the galvanoxy radical has the structure shown in Formula II:

[0015]

[0016] Formula II

[0017] Furthermore, the free radical dopant in the active layer has a mass fraction of 1% to 20%, preferably 5% to 15%.

[0018] Furthermore, the organic semiconductor material is preferably a DPP-DTT polymer as shown in Formula I, which has good light absorption and charge transport capabilities. The free radical dopant is preferably a Galvinoxyl free radical as shown in Formula II, which has stable unpaired electrons and can serve as a highly efficient photogenerated charge trapping and releasing center.

[0019] The preparation method of the material system includes: dissolving the organic semiconductor in a suitable solvent to form a solution; adding the free radical dopant to the solution and mixing thoroughly to form a homogeneous blend solution. This blend solution can be used to subsequently fabricate the active layer of a device.

[0020] A second aspect of the present invention is to provide an optically reconfigurable bipolar neuromorphic device based on the above-mentioned active layer.

[0021] The device has a sandwich structure, comprising, in sequence: a substrate, a bottom electrode, an active layer, and a top electrode. The substrate can be a rigid (e.g., glass) or flexible (e.g., polyimide) transparent substrate. The bottom electrode is a transparent conductive electrode, such as indium tin oxide (ITO). The active layer is formed by coating (e.g., spin coating, blade coating) or vapor deposition of the aforementioned blend solution containing organic semiconductors (e.g., DPP-DTT) and free radical dopant (e.g., Galvinoxyl). This layer is the functional core for achieving photo-controlled bipolar doping. The top electrode is a metal electrode, such as silver (Ag), gold (Au), or aluminum (Al).

[0022] A typical and optimized device structure is: glass / ITO / active layer (DPP-DTT:Galvinoxyl) / Ag.

[0023] The device operates as follows: Under illumination at approximately 808 nm, Galvinoxyl radicals tend to capture photogenerated electrons, leading to a relative increase in the hole concentration in the active layer, manifesting as p-type doping, and the device generates an excitatory (positive) photocurrent. Under illumination at approximately 450 nm, the captured electrons are released and recombine with holes, resulting in a relative increase in the electron concentration in the active layer, manifesting as n-type doping, and the device generates a suppressive (negative) photocurrent. This process is highly reversible and reconfigurable.

[0024] A third aspect of the present invention is to provide a neuromorphic visual array comprising a plurality of pixel units arranged in an array, wherein each pixel unit is the aforementioned optically reconfigurable bipolar neuromorphic device; preferably, the pixel size of the array is 256×256 or more.

[0025] The optically reconfigurable bipolar neuromorphic device of this invention can be fabricated into a neuromorphic visual array and integrated onto the backplane of a thin-film transistor (TFT) via solution processing. This method only requires arraying the exposed electrodes on the TFT backplane, eliminating the need for photolithographic patterning of the photosensitive layer, ultimately resulting in an arrayed image sensing array device. Furthermore, the size of the pixel units in the arrayed device can be controlled by altering the shape and size of the exposed electrodes on the TFT backplane, thereby adjusting the image resolution.

[0026] A fourth aspect of the present invention is to provide a fully integrated, multifunctional neuromorphic visual system.

[0027] The system integrates the aforementioned neuromorphic visual array, signal readout circuitry (such as a readout integrated circuit), and image receiving unit. Crucially, due to the inherent bipolar response and synaptic characteristics of the underlying pixels, this system can achieve the execution of various biomimetic visual functions at the hardware level without complex software algorithms, primarily including:

[0028] Dynamic motion detection: Directly utilizes the synaptic response characteristics of pixels to extract motion trajectories in real time.

[0029] Spatiotemporal information memory: short-term storage of visual information is achieved by utilizing the charge capture / release dynamics of pixels.

[0030] Contrast-enhanced edge extraction: By programming and controlling the excitation / inhibition states of different regions of the array, the center-peripheral receptive field of the retina is simulated at the sensing end, and the enhanced edge feature map is directly output.

[0031] A fifth aspect of the present invention provides an imaging or visual information processing method employing the aforementioned neuromorphic visual system, comprising the steps of:

[0032] The aforementioned neuromorphic visual array is used to receive light signals from dynamic visual scenes.

[0033] By optically modulating the bipolar doping effect of the free radical dopant, concurrent excitatory or inhibitory neuromorphic response signals are generated in the array.

[0034] The response signal is processed to obtain motion information, memory information, and feature information of the scene.

[0035] Advantages and beneficial effects of the present invention:

[0036] 1. This invention is the first to utilize the photo-controlled charge capture / release of free radicals to achieve bipolar doping, unifying two key neuromorphic properties, synaptic plasticity and bipolar photoresponse, into a single, simple material system.

[0037] 2. The neuromorphic vision system of the present invention includes a pixel array composed of multiple optically reconfigurable bipolar neuromorphic devices, wherein each device achieves wavelength-selective reversible bipolar doping by introducing galvanoxy radicals as dopants, thereby achieving the integration of dynamic synaptic characteristics and optically reconfigurable bipolar response.

[0038] 3. The neuromorphic vision system of this invention can simultaneously perform multiple visual functions, including motion detection, spatiotemporal information memory, and contrast enhancement feature extraction, at the sensing front end. The device structure of this invention is simplified, highly uniform, and has a strong light response, achieving large-scale (256×256) array integration, providing a hardware solution for machine vision in complex dynamic scenarios such as autonomous driving and industrial inspection. Attached Figure Description

[0039] Figure 1 The UV-Vis absorption spectrum is shown for the Galvinoxyl radical and DPP-DTT polymer semiconductor blend system prepared in Example 1 of this invention.

[0040] Figure 2 The electron paramagnetic resonance spectrum is that of the Galvinoxyl free radical and DPP-DTT polymer semiconductor blend system prepared in Example 1 of this invention.

[0041] Figure 3 This is a schematic cross-sectional view of the optically reconfigurable bipolar neuromorphic device in Embodiment 2 of the present invention.

[0042] Figure 4 The figure shows the excitatory postsynaptic current response curve generated by the device in Example 2 under stimulation by an 808 nm light pulse sequence.

[0043] Figure 5 The image shows the inhibitory postsynaptic current response curve generated by the device in Example 2 under stimulation by a 450 nm light pulse sequence.

[0044] Figure 6 This is a schematic diagram of a single pixel cross-section of a neuromorphic visual array.

[0045] Figure 7 This is a hardware architecture block diagram of the complete neuromorphic visual system constructed according to Embodiment 3 of the present invention.

[0046] Figure 8 This is a schematic diagram showing the results of dynamic motion detection using the system in Example 3.

[0047] Figure 9 The figure shows the experimental results of using the spatiotemporal information memory function of the system in Example 3.

[0048] Figure 10 The images show a comparison of the results of contrast enhancement edge extraction using the system in Example 3. The left image shows the result before edge extraction, and the right image shows the result after edge extraction. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the described embodiments are intended to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of the invention.

[0050] Example 1: Preparation and characterization of Galvinoxyl radical-DPP-DTT polymer semiconductor blend system

[0051] 1. Material preparation: The organic semiconductor material is a DPP-DTT polymer with the structure of Formula I, and its number average molecular weight (Mn) is about 24 kDa; the dopant is a Galvinoxyl radical (GX) with the structure of Formula II, and its purity is >98%; the solvent is chlorobenzene, which has been treated with anhydrous and oxygen-free methods.

[0052] 2. Preparation of the blend solution: Dissolve DPP-DTT solid in chlorobenzene at a concentration of 10 mg / mL and stir at 50°C until completely dissolved (approximately 12 hours). Then, in a nitrogen glove box, add the pre-prepared GX chlorobenzene solution (60 mg / mL) to the DPP-DTT main solution at a specific volume ratio, ensuring that the mass fraction of GX in the solid content is 10 wt%. Continue stirring the mixed solution for 1 hour to ensure homogeneous mixing.

[0053] 3. Thin film preparation and spectroscopic characterization: The above-mentioned blend solution was spin-coated (2000 rpm, 30 s) onto a pretreated quartz substrate to form a thin film with a thickness of approximately 200 nm. The film was immediately placed on an 80°C hot plate for annealing for 15 minutes to remove residual solvent.

[0054] Ultraviolet-Visible-NearInfrared Absorption Spectroscopy Test: such as Figure 1 As shown, the absorption peak of the pure DPP-DTT film is located at approximately 808 nm. The absorption peak of the pure GX film is located at approximately 450 nm.

[0055] Electron paramagnetic resonance (EPR) testing: such as Figure 2 As shown, pure DPP-DTT films show no EPR signal. After GX doping, a strong EPR signal appears at g=2.003, which is consistent with the signal of the GX standard, confirming that GX radicals exist in the blend film in an active form with unpaired electrons, without dimerization or deactivation.

[0056] Example 2: Fabrication and Performance Testing of Optically Reconfigurable Bipolar Neuromorphic Devices

[0057] Device structure: such as Figure 3 As shown, a sandwich structure is adopted, specifically: glass / indium tin oxide (ITO) / active layer (DPP-DTT:GX (10 wt%), 200 nm) / silver (Ag, 100 nm).

[0058] The device fabrication steps are as follows:

[0059] (1) Cleaning of the substrate: The substrate is a glass substrate with ITO. Place the substrate on a shelf and clean it in sequence with soapy water, deionized water, acetone and isopropanol, twice each time for half an hour each time. Before use, blow away the residual solvent on the substrate with N2 and clean it with plasma for 1 minute to further remove the residual organic matter on the substrate surface.

[0060] (2) Processing of the photosensitive layer: The concentration of the active layer solution was 10 mg / mL, the solvent was chlorobenzene, and the mass ratio of DPP-DTT to GX was 1:0.1. The solution was stirred for at least three hours. During spin coating, the rotation speed was 2000 r / min and the time was 30 s, with 18 μL taken each time. After spin coating of the active layer, it was annealed on a hot plate at 80℃ for 15 min.

[0061] (3) Thermal evaporation of hole transport layer: Place the substrate with the spin-coated active layer onto the mask plate, and the evaporation conditions are that the vacuum chamber pressure is less than 2×10. -4 The deposition rate was less than 0.2 nm / s, and the deposition thickness was 100 nm. The effective area of ​​the final photodetector was 0.04 cm². 2 .

[0062] Photoelectric performance testing:

[0063] Photocurrent response: Under a bias voltage of -200mV, 808 nm and 450 nm laser pulses were controlled using a signal generator (power density 100 mW / cm²). 2 The effective area of ​​the irradiation device. For example... Figure 4-5 As shown, the device generates a positive photocurrent (excitatory response) to 808 nm light and a negative photocurrent (inhibitory response) to 450 nm light, demonstrating its photoreconstruction bipolar response capability. Synaptic characteristics: After applying pulsed light, the device generates a photocurrent spike followed by an exponentially decaying afterglow current, simulating the short-term memory (STP) behavior of biological synapses, such as... Figure 4-5 As shown.

[0064] Example 3: Fabrication and Functional Demonstration of a Fully Integrated Neuromorphic Visual System

[0065] System Integration: This embodiment demonstrates a method for fabricating an arrayed neuromorphic vision device. The neuromorphic vision array is fabricated using a thin-film transistor (TFT) backplane purchased from LinkZill, model SC-T-G256-2401 (GHS1), with 256×256 pixel units and an active area size of 32×32mm. The device structure is also a sandwich structure, and the other materials and fabrication methods are the same as in Embodiment 2. After the neuromorphic vision array is fabricated, a signal readout device purchased from LinkZill is used to acquire the image obtained by the neuromorphic device. The system architecture is shown in... Figure 7 .

[0066] Function demonstration:

[0067] Demonstration 1: Dynamic Motion Detection: A moving, patterned light (808 nm) is placed in front of the array. (Example:...) Figure 8 As shown, the system directly extracts and displays the motion trajectory of the pattern in real time.

[0068] Demonstration 2: Spatiotemporal Information Memory: Using a mask, different numerical sequences are sequentially presented on an array. The final output image contains all sequence information, such as... Figure 9 As shown.

[0069] Demonstration 3: Contrast Enhancement Edge Extraction: Comparing excitation (background)-excitation (local) with excitation (background)-inhibition (local), the latter constructs a retinal-like center-surround receptive field, with significantly enhanced contrast and sharper edges. (Shown in...) Figure 10 .

[0070] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An active layer based on photo-controlled free radical bipolar doping, characterized in that, Organic semiconductor materials; and Free radical dopants dispersed in the organic semiconductor material; The active layer is configured such that: under illumination of a first wavelength, the free radical dopant captures photogenerated charges, causing the active layer to exhibit a first conductivity type doped state and generating a photoresponse of a first polarity; under illumination of a second wavelength, the free radical dopant releases the captured charges, causing the active layer to exhibit a second conductivity type doped state opposite to the first conductivity type and generating a photoresponse of a second polarity opposite to the first polarity.

2. The active layer according to claim 1, characterized in that, The first wavelength is located in the near-infrared band, and the second wavelength is located in the visible light band; preferably, the first wavelength is 808 nm and the second wavelength is 450 nm.

3. The active layer according to claim 1 or 2, characterized in that, The organic semiconductor material is a donor-acceptor type conjugated polymer; preferably, the organic semiconductor material is a DPP-DTT polymer having the structure shown in Formula I: Formula I.

4. The active layer according to any one of claims 1-3, characterized in that, The free radical dopant is a galvanoxy radical; preferably, the galvanoxy radical has the structure shown in Formula II: Formula II.

5. The active layer according to any one of claims 1-4, characterized in that, The free radical dopant in the active layer has a mass fraction of 1% to 20%, preferably 5% to 15%.

6. An optically reconfigurable bipolar neuromorphic device, characterized in that, include: Base; The bottom electrode is disposed on the substrate; An active layer as described in any one of claims 1-5 is disposed on the bottom electrode; as well as The top electrode is disposed on the active layer.

7. The optically reconfigurable bipolar neuromorphic device according to claim 6, characterized in that, The substrate is a glass substrate or a flexible polymer substrate; the bottom electrode is a transparent conductive oxide electrode, preferably an indium tin oxide electrode; the top electrode is a metal electrode, preferably a silver electrode.

8. A neuromorphic visual array, characterized in that, It includes a plurality of pixel units arranged in an array, wherein each pixel unit is an optically reconfigurable bipolar neuromorphic device as described in claim 6 or 7; preferably, the pixel size of the array is 256×256 or more.

9. A neuromorphic visual system, characterized in that, include: The neuromorphic visual array as described in claim 8 is used to receive optical signals and generate neuromorphic response signals; A signal readout circuit electrically connected to the neuromorphic visual array is used to read the response signal; as well as An image processing unit, which is communicatively connected to the signal readout circuit, is used to process the response signal to obtain visual information.

10. A method for processing imaging or visual information, characterized in that, The neuromorphic visual system of claim 8 includes the following steps: The neuromorphic visual array described in claim 8 is used to receive light signals from dynamic visual scenes. By optically modulating the bipolar doping effect of the free radical dopant, concurrent excitatory or inhibitory neuromorphic response signals are generated in the array. The response signal is processed to obtain motion information, memory information, and feature information of the scene.

Images