Sensing and calculating multifunctional self-calibration chip and preparation method thereof

Abstract

The application is a kind of self-calibration chip with sensing and calculating multifunction and its preparation method, which comprises an array structure composed of multiple multifunctional composite units and a control and feedback unit connected with the array structure; the multifunctional composite unit has the ability of tactile perception, brain-like calculation, light emission and light detection in the same structure; the multifunctional composite unit is used to generate an electric signal under the action of external stress, realize brain-like calculation based on the electric signal, generate a light signal under the electric drive, receive external light signals and convert them into electric signals; the multifunctional composite units realize information transmission through the emission and reception of light signals and realize coupling connection through electric signals, thereby constructing an array interconnection network; the control and feedback unit is used to process signals and adjust the multifunctional composite units according to the processing results. The application realizes the high integration of sensing, calculation and active detection functions, realizes collaborative processing based on information interaction between array units, and realizes self-calibration.

Description

Technical Field

[0001] This invention relates to the field of neuromorphic device technology, specifically to a self-calibrating chip with multiple functions including sensing, computing, and display, and its fabrication method. Background Technology

[0002] In the fields of intelligent sensing and human-computer interaction, tactile perception, visual perception, and edge intelligent computing are gradually becoming important fundamental capabilities for realizing robot systems, intelligent terminals, and neuromorphic devices. Especially in flexible electronic skin, intelligent sensor arrays, and autonomous sensing systems, devices typically need to simultaneously possess multiple functions such as external environment perception, signal processing, and information feedback in order to achieve shape recognition, state judgment, and dynamic response of target objects.

[0003] In existing technologies, sensing systems typically adopt a discrete architecture, consisting of a sensing unit, a signal acquisition circuit, a data processing unit, and a display or feedback module. The functional modules are connected through external circuits to achieve hierarchical processing and transmission of signals. However, this architecture has problems such as complex structure, low system integration, and long signal transmission paths in practical applications, which can easily lead to high system latency and make it difficult to meet the requirements of next-generation edge intelligent systems for miniaturization, high integration, and real-time processing capabilities.

[0004] To address the aforementioned issues, existing technologies have provided improved solutions. For example, patent CN121586512A discloses a single-chip active array for tactile or stress sensing and its application. This solution integrates stress sensing and digital signal conversion functions on a single chip, enabling each sensing node to have a certain information processing capability, thereby improving the system's anti-interference performance and achieving effective acquisition of tactile or stress information.

[0005] However, the aforementioned solutions primarily focus on single-modal sensing and signal processing, and still have shortcomings in multimodal information fusion, integrated sensing-computing-display, and inter-unit collaborative capabilities. In particular, they lack a collaborative processing mechanism based on information interaction between array units, and their system adaptive adjustment and autonomous calibration capabilities still need improvement. Therefore, how to construct a multifunctional chip with multimodal sensing capabilities, achieving a high degree of integration of sensing, computing, and active detection functions, while simultaneously enabling collaborative processing based on information interaction between array units and possessing autonomous calibration capabilities, has become a pressing technical problem to be solved in this field. Summary of the Invention

[0006] This invention provides a multi-functional autonomous calibration chip and its fabrication method to overcome the technical problems of existing tactile sensing systems, such as functional separation, single perception dimension, low system integration, and difficulty in achieving edge intelligent processing and autonomous calibration.

[0007] To achieve the above-mentioned technical objectives and effects, the present invention is implemented through the following technical solution: A multi-functional autonomous calibration chip with sensing, computing, and display capabilities, comprising: An array structure consisting of multiple multifunctional composite units, and a control and feedback unit connected to the array structure; The multifunctional composite unit is a device unit with a unified physical structure, which has the ability to realize tactile perception, brain-like computing, light emission and light detection in the same structure; The multifunctional composite unit is used to generate electrical signals under external stress and realize brain-like computing based on the electrical signals. At the same time, it generates optical signals under electric drive and receives external optical signals and converts them into electrical signals. The multifunctional composite units transmit information through the transmission and reception of optical signals and are coupled together through electrical signals, thereby constructing an array interconnection network. The control and feedback unit is used to process the signal and adjust the multifunctional composite unit according to the processing result to achieve autonomous calibration. The tactile perception is achieved through changes in electrical signals caused by external stress, and is used to modulate the brain-like computing process.

[0008] Furthermore, the multifunctional composite units can transmit information over short distances between adjacent units or across units between non-adjacent units, thereby enabling direct or indirect optical communication between the units.

[0009] Furthermore, the optical signal is transmitted in the array structure through a waveguide structure, which includes at least one of a planar waveguide, a ridge waveguide, or a photonic crystal waveguide.

[0010] Furthermore, the electrical signal coupling connection between the multifunctional composite units includes at least one of shared electrode connection, local interconnection line or global interconnection network.

[0011] Furthermore, the multifunctional composite units in the array structure constitute a two-dimensional or three-dimensional topology and support at least one information interaction mode among neighborhood connection, full connection or sparse connection.

[0012] Furthermore, the multifunctional composite unit generates an adjustable current response based on the input signal, which varies with the amplitude, frequency, or time interval of the input pulse to simulate the excitatory and inhibitory behaviors of the synapse.

[0013] Furthermore, the multifunctional composite unit provides active illumination in the light-emitting mode and receives light signals emitted by external sources or other units in the light-detection mode, thereby achieving multiplexing of sensing and communication. Switching between the light-emitting mode and the light-detection mode is achieved through electrical signal control.

[0014] Furthermore, the control and feedback unit dynamically adjusts the operating state of the multifunctional composite unit based on the array output signal to achieve autonomous parameter calibration; wherein: The dynamic adjustment is based on the error between the output signal and the target signal; The autonomous calibration includes at least one of adjusting the threshold, conductivity state, response intensity, or time characteristics; The threshold or conductivity state is adjusted by applying an electrical pulse or an optical pulse.

[0015] Furthermore, the multifunctional composite unit is based on a semiconductor material, including GaN, (In,Ga)N, or a multi-quantum well structure formed therefrom, and a p-(Al,Ga)N electron blocking layer is disposed in the multifunctional composite unit.

[0016] A method for fabricating a multi-functional autonomous calibration chip with sensing and computing capabilities, the method comprising: The multifunctional composite unit uses GaN devices. First, a GaN epitaxial wafer is fabricated. Then, the GaN epitaxial wafer is divided to obtain several GaN devices, wherein: The fabrication process of GaN epitaxial wafers is as follows: Step S101: Prepare a Si substrate, place it in a container containing acetone and isopropanol, and ultrasonically clean it separately. Step S102: The Si substrate is then placed in an MOCVD growth chamber and annealed for a period of time to remove the surface natural oxide layer. Step S103: Finally, GaN epitaxial wafers for the self-calibration chip used in sensing, computing, and display functions are grown using epitaxial technology. The specific process is as follows: Step S1031: Epitaxially grow an AlN nucleation layer of a certain thickness and a (Al,Ga)N multilayer buffer layer of a certain thickness to reduce dislocation density; Step S1032: Continue to grow an undoped GaN layer of a certain thickness, and epitaxially grow a silicon-doped n-type GaN layer of a certain thickness to form a conductive channel; Step S1033: Epitaxially construct several periods of (In,Ga)N / GaN multi-quantum-well layers; Step S1034: Grow a magnesium-doped p-(Al,Ga)N electron blocking layer of a certain thickness to suppress carrier leakage; Step S1035: Sequentially construct a p-GaN layer of a certain thickness and a p-GaN contact layer of a certain thickness with heavy Mg doping to complete the p-region construction and thus complete the preparation of the GaN epitaxial wafer; The process of segmenting GaN epitaxial wafers is as follows: Step S201: Take the prepared GaN epitaxial wafer and perform electrode deposition on its p-GaN surface. The electrode deposition process adopts magnetron sputtering, specifically as follows: First, deposit a certain thickness of Ti, then deposit a certain thickness of Al, then deposit a certain thickness of Ni, and finally deposit a certain thickness of Au to form a Ti / Al / Ni / Au electrode structure, so as to achieve low contact resistance, good adhesion, excellent thermal stability and reliable electrical performance, thereby forming a stable ohmic structure. Step S202: Perform device dicing on the epitaxial wafer after electrode fabrication, specifically including the following steps: Step S2021: Use laser scribing technology to cut the epitaxial wafer into multiple standardized device units; Step S2022: Perform an electrochemical stripping process on the diced epitaxial wafer to separate each device unit from the substrate and obtain independent GaN devices; Step S2023: Clean and passivate the separated device units.

[0017] The beneficial effects of this invention are: (1) The present invention integrates tactile perception, brain-like computing and light emission and light detection functions in the same multifunctional composite unit, realizing an integrated structure of perception-computation-display, breaking through the traditional discrete architecture of multiple devices, and significantly improving the system integration and functional synergy.

[0018] (2) The present invention has tactile sensing capability and directly generates electrical signal output under external pressure or stress, without the need for complex external signal conversion circuits, thereby simplifying the system structure and reducing the complexity of the signal link.

[0019] (3) The multifunctional composite unit in this invention has brain-like computing characteristics. It realizes synaptic plasticity through the nonlinear change of device conductance or current response, thereby enabling information processing and storage at the device level, improving the edge computing capability of the system and reducing dependence on external processing units.

[0020] (4) The present invention integrates light emission and light detection functions in the same device to realize an active optical sensing mechanism, that is, to illuminate the target object by emitting light through the device and to receive the reflected light signal, thereby obtaining the spatial shape and surface feature information of the target object and improving the sensing dimension and recognition accuracy.

[0021] (5) In this invention, information is transmitted between units in the multifunctional composite unit array through the transmission and reception of optical signals, and an array interconnection network is constructed by combining electrical signal coupling, so that each unit can perform information interaction and collaborative work, thereby realizing distributed information processing and brain-like computing functions, significantly improving the overall computing efficiency and scalability of the system.

[0022] (6) The present invention constructs a closed-loop adjustment mechanism through a control and feedback unit, which can adjust the working state of the device in real time according to the tactile signal and the optical feedback signal, thereby achieving autonomous calibration, effectively compensating for device drift, environmental changes and differences between devices, and improving system stability and consistency.

[0023] (7) The present invention is based on the epitaxial system of GaN / (In,Ga)N multiple quantum wells and p-(Al,Ga)N electron blocking layer, which enables the multifunctional composite unit to have the response capability from ultraviolet to visible light band, and at the same time has high temperature resistance and radiation resistance characteristics, making it suitable for multimodal sensing and computing scenarios in complex or extreme environments.

[0024] (8) This invention achieves high integration of multifunctional composite units and interconnection between units in the same array structure, thereby constructing an intelligent system architecture that integrates sensing, computing, communication and feedback, thereby further improving the miniaturization and functional integration of the system. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of a GaN epitaxial wafer in an embodiment of the present invention; Figure 2 This is a schematic diagram of the chip array in an embodiment of the present invention; Figure 3 This is a functional block diagram in an embodiment of the present invention; Figure 4 This is a current data graph for detecting external light in an embodiment of the present invention. Detailed Implementation

[0026] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0027] A multi-functional autonomous calibration chip with sensing, computing, and display capabilities, comprising: An array structure consisting of multiple multifunctional composite units, and a control and feedback unit connected to the array structure; The multifunctional composite unit array is composed of multiple multifunctional composite units. Each multifunctional composite unit is a unified physical structure device unit, and simultaneously possesses tactile sensing function, brain-like computing function, light emission function, and light detection function in the same structure. The multifunctional composite unit is used to generate electrical signals under external stress to achieve tactile perception, and to achieve brain-like computing based on the electrical state of the electrical signal modulation device; at the same time, it generates light signals under electric drive to actively illuminate the target object, and receives the light signals reflected by the target object and converts them into electrical signals, thereby achieving active optical perception. In the multifunctional composite unit array, information is transmitted between units through the transmission and reception of optical signals, and coupled together through electrical signals, thereby constructing an array interconnection network with information interaction capabilities. Based on the interconnection structure, each unit can collaboratively complete information processing and realize distributed brain-like computing functions.

[0028] The control and feedback unit is used to process the electrical signal and the reflected light signal, and dynamically adjust the working parameters of the multifunctional composite unit according to the processing results, thereby forming a closed-loop feedback mechanism to realize the chip's autonomous calibration and autonomous control.

[0029] In one embodiment of this application, the tactile sensing function of the multifunctional composite unit is realized based on the piezoelectric effect. When the device is subjected to external pressure or stress, it generates a corresponding electrical signal output, thereby realizing the direct perception of tactile information.

[0030] The tactile perception can also be achieved based on the piezoresistive effect, the capacitive effect, or other electromechanical coupling mechanisms.

[0031] In one embodiment of this application, the multifunctional composite unit exhibits synaptic plasticity characteristics through nonlinear changes in its conductance or current response, enabling the device to modulate and store input signals, thereby realizing neuromorphic computing functions at the device level.

[0032] The synaptic-like properties include short-term plasticity, long-term plasticity, or threshold modulation behavior.

[0033] In one embodiment of this application, the multifunctional composite unit generates an optical signal under electrically driven conditions, and its light emission mode is electroluminescence, which is used to actively illuminate the target object.

[0034] The emission wavelengths include ultraviolet light, visible light, or near-infrared light.

[0035] In one embodiment of this application, the multifunctional composite unit realizes the photodetection function based on the photoelectric effect, converting the received optical signal into an electrical signal for output.

[0036] The optical signal mechanism includes the photoconductive effect and the photovoltaic effect.

[0037] In one embodiment of this application, the multifunctional composite unit simultaneously realizes the light-emitting function and the detection function in the same device structure, thereby achieving device-level optical integration.

[0038] In one embodiment of this application, the multifunctional composite unit actively illuminates the target object by emitting its own light and achieves active optical sensing by receiving the light signal reflected by the target object.

[0039] The optical signal includes intensity information, change information, or spatial distribution information.

[0040] In one embodiment of this application, the multifunctional composite units achieve neighborhood communication or remote communication through the optical signal and achieve auxiliary coupling through electrical connection, thereby forming a scalable array interconnect structure.

[0041] In one embodiment of this application, the control and feedback unit analyzes and acquires the shape information of the target object based on the reflected light signal.

[0042] The information is processed simultaneously by one or more units, and the morphological information includes the contour features, surface structure, or spatial distribution information of the target object.

[0043] In one embodiment of this application, the control and feedback unit adjusts the driving voltage, current, or bias conditions of the multifunctional composite unit according to the processing result.

[0044] The adjustment can be continuous or discrete.

[0045] In one embodiment of this application, the chip achieves autonomous calibration through a closed-loop adjustment mechanism based on feedback signals, thereby improving the stability and consistency of the system.

[0046] The autonomous calibration is used to compensate for device drift, environmental changes, or differences between devices.

[0047] In one embodiment of this application, the multifunctional composite unit is based on a semiconductor material; preferably, the semiconductor material includes GaN, (In,Ga)N or a multi-quantum well structure formed therefrom, and the multifunctional composite unit is provided with a p-(Al,Ga)N electron blocking layer.

[0048] The semiconductor materials also include other material systems with piezoelectric effect and photoelectric response characteristics.

[0049] In one embodiment of this application, the composite unit array is a two-dimensional array structure.

[0050] The array structure can be a regular array or an irregular array.

[0051] A method for fabricating a multifunctional autonomous calibration chip with sensing, computing, and display functions is disclosed. In this embodiment, the multifunctional composite unit uses GaN devices. First, a GaN epitaxial wafer is prepared, and then the GaN epitaxial wafer is divided to obtain several GaN devices.

[0052] The fabrication process of GaN epitaxial wafers is as follows: A Si(111) substrate is prepared and placed in a container filled with acetone and isopropanol, and ultrasonically cleaned for 5 minutes each. The substrate is then placed in an MOCVD growth chamber and annealed at 900℃ for several tens of minutes to remove the surface natural oxide layer. Finally, GaN epitaxial wafers for a self-calibrating chip with multiple functions of sensing, computing, and display are grown using epitaxial technology. The specific steps include: The first step involves epitaxially growing an AlN nucleation layer approximately 330 nm thick and a (Al,Ga)N multilayer buffer layer approximately 600 nm thick to reduce dislocation density. The second step involves growing an undoped GaN layer approximately 800 nm thick and epitaxially growing a silicon-doped n-type GaN layer approximately 2800 nm thick to form a conductive channel.

[0053] The third step involves epitaxially growing a (In,Ga)N / GaN multi-quantum-well layer with nine cycles, each cycle having a thickness of approximately 3 / 5 nm and a growth temperature of approximately 750 °C. The fourth step is to grow a magnesium-doped p-(Al,Ga)N electron blocking layer approximately 30 nm thick to suppress carrier leakage. The fifth step involves sequentially layering a 60nm thick p-GaN layer and a 20nm heavily doped Mg p-GaN contact layer to complete the p-region construction. This completes the fabrication of the GaN epitaxial wafer, whose layered structure is shown below. Figure 1 As shown.

[0054] For example, the process of dividing a GaN epitaxial wafer is as follows: The prepared GaN epitaxial wafer is taken, and electrode deposition is performed on its p-GaN surface using magnetron sputtering. First, a Ti layer with a thickness of approximately 20 nm is deposited; second, an Al layer with a thickness of approximately 120 nm is deposited; then, a Ni layer with a thickness of approximately 60 nm is deposited; finally, an Au layer with a thickness of approximately 200 nm is deposited, forming a Ti / Al / Ni / Au electrode structure. This achieves low contact resistance, good adhesion, excellent thermal stability, and reliable electrical performance, thereby forming a stable ohmic structure. Then, the epitaxial wafer with the completed electrode preparation is divided into devices; specifically, the following steps are included: The first step involves using laser scribing to cut the epitaxial wafer into multiple pieces with an area of ​​approximately 1 mm. 2 Standardized device units; The second step is to perform an electrochemical stripping process on the diced epitaxial wafer to separate each device unit from the substrate and obtain independent GaN devices. The third step is to clean and passivate the separated device units. Thus, the electrode fabrication and device dicing processes are completed, resulting in multiple independent GaN devices, forming a structure like... Figure 2 The chip array shown.

[0055] Thus, the self-calibrating chip with multiple functions including sensing, computing, and display was completed, and its functional block diagram is as follows. Figure 3 As shown.

[0056] For example, the current data graph used as a detection function is as follows: Figure 4 As shown, its functional characteristics are verified.

[0057] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-functional autonomous calibration chip with sensing, calculation, and display capabilities, characterized in that, The chip includes: An array structure consisting of multiple multifunctional composite units, and a control and feedback unit connected to the array structure; The multifunctional composite unit is a device unit with a unified physical structure, which has the ability to realize tactile perception, brain-like computing, light emission and light detection in the same structure; The multifunctional composite unit is used to generate electrical signals under external stress and realize brain-like computing based on the electrical signals. At the same time, it generates optical signals under electric drive and receives external optical signals and converts them into electrical signals. The multifunctional composite units transmit information through the transmission and reception of optical signals and are coupled together through electrical signals, thereby constructing an array interconnection network. The control and feedback unit is used to process the signal and adjust the multifunctional composite unit according to the processing result to achieve autonomous calibration. The tactile perception is achieved through changes in electrical signals caused by external stress, and is used to modulate the brain-like computing process.

2. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The multifunctional composite units can achieve direct or indirect optical communication between units by transmitting short-distance information between adjacent units or by transmitting cross-unit information between non-adjacent units.

3. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The optical signal is transmitted in the array structure through a waveguide structure, which includes at least one of a planar waveguide, a ridge waveguide, or a photonic crystal waveguide.

4. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The electrical signal coupling connection between the multifunctional composite units includes at least one of shared electrode connection, local interconnection line or global interconnection network.

5. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The multifunctional composite units in the array structure constitute a two-dimensional or three-dimensional topology and support at least one information interaction mode among neighborhood connection, full connection or sparse connection.

6. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The multifunctional composite unit generates an adjustable current response based on the input signal. The current response varies with the amplitude, frequency, or time interval of the input pulse to simulate the excitatory and inhibitory behaviors of the synapse.

7. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The multifunctional composite unit provides active illumination in the light-emitting mode and receives light signals emitted by external sources or other units in the light-detection mode, thereby achieving multiplexing of sensing and communication. Switching between the light-emitting mode and the light-detection mode is achieved through electrical signal control.

8. The self-calibrating chip with multiple functions including sensing, computing, and display as described in claim 1, characterized in that, The control and feedback unit dynamically adjusts the operating state of the multifunctional composite unit based on the array output signal to achieve autonomous parameter calibration; wherein: The dynamic adjustment is based on the error between the output signal and the target signal; The autonomous calibration includes at least one of adjusting the threshold, conductivity state, response intensity, or time characteristics; The threshold or conductivity state is adjusted by applying an electrical pulse or an optical pulse.

9. The self-calibrating chip with multiple functions of sensing, computing, and display as described in claim 1, and its fabrication method, characterized in that, The multifunctional composite unit is based on semiconductor materials, including GaN, (In,Ga)N, or a multi-quantum well structure formed therefrom, and a p-(Al,Ga)N electron blocking layer is disposed in the multifunctional composite unit.

10. A method for fabricating a multifunctional autonomous calibration chip with sensing and display capabilities according to any one of claims 1-9, characterized in that, The method includes: The multifunctional composite unit uses GaN devices. First, a GaN epitaxial wafer is fabricated. Then, the GaN epitaxial wafer is divided to obtain several GaN devices, wherein: The fabrication process of GaN epitaxial wafers is as follows: Step S101: Prepare a Si substrate, place it in a container containing acetone and isopropanol, and ultrasonically clean it separately. Step S102: The Si substrate is then placed in an MOCVD growth chamber and annealed for a period of time to remove the surface natural oxide layer. Step S103: Finally, GaN epitaxial wafers for the self-calibration chip used in sensing, computing, and display functions are grown using epitaxial technology. The specific process is as follows: Step S1031: Epitaxially grow an AlN nucleation layer of a certain thickness and a (Al,Ga)N multilayer buffer layer of a certain thickness to reduce dislocation density; Step S1032: Continue to grow an undoped GaN layer of a certain thickness, and epitaxially grow a silicon-doped n-type GaN layer of a certain thickness to form a conductive channel; Step S1033: Epitaxially construct several periods of (In,Ga)N / GaN multi-quantum-well layers; Step S1034: Grow a magnesium-doped p-(Al,Ga)N electron blocking layer of a certain thickness to suppress carrier leakage; Step S1035: Sequentially construct a p-GaN layer of a certain thickness and a p-GaN contact layer of a certain thickness with heavy Mg doping to complete the p-region construction and thus complete the preparation of the GaN epitaxial wafer; The process of segmenting GaN epitaxial wafers is as follows: Step S201: Take the prepared GaN epitaxial wafer and perform electrode deposition on its p-GaN surface. The electrode deposition process adopts magnetron sputtering, specifically as follows: First, deposit a certain thickness of Ti, then deposit a certain thickness of Al, then deposit a certain thickness of Ni, and finally deposit a certain thickness of Au to form a Ti / Al / Ni / Au electrode structure, so as to achieve low contact resistance, good adhesion, excellent thermal stability and reliable electrical performance, thereby forming a stable ohmic structure. Step S202: Perform device dicing on the epitaxial wafer after electrode fabrication, specifically including the following steps: Step S2021: Use laser scribing technology to cut the epitaxial wafer into multiple standardized device units; Step S2022: Perform an electrochemical stripping process on the diced epitaxial wafer to separate each device unit from the substrate and obtain independent GaN devices; Step S2023: Clean and passivate the separated device units.

Images