Self-repairing chip and preparation method thereof

By introducing a combination of gradient nanoporous glass substrate, oxide alloy capsule and infrared laser array into the self-healing chip, real-time stress monitoring and rapid repair are achieved, solving the problem of lag in repair response in the prior art and improving the reliability and lifespan of the chip under extreme conditions.

CN121693172APending Publication Date: 2026-03-17SHENZHEN NTEK TESTING TECH +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511829593.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-05
Publication Date
2026-03-17

AI Technical Summary

Technical Problem

Existing self-healing chips cannot achieve real-time repair under extreme temperatures and mechanical stress, resulting in rapid microcrack propagation. Existing repair agents have a delayed response and insufficient heat resistance, affecting chip reliability.

Method used

By employing a combination of gradient nanoporous glass substrate, oxide alloy capsule structure, and infrared laser array, stress distribution is monitored in real time by stress sensor, and the capsule structure is triggered by infrared laser to repair microcracks before they appear. Combined with graphene film and copper wiring layer to provide conductive channels, rapid repair is achieved.

Benefits of technology

The ability to repair microcracks within 0.05 seconds significantly improves the reliability and lifespan of chips under extreme conditions, enhancing the stability and maintenance economy of aerospace and automotive electronic equipment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121693172A_ABST
    Figure CN121693172A_ABST
Patent Text Reader

Abstract

The invention discloses a self-repairing chip and a preparation method thereof, and the preparation method comprises the steps: carrying out the preliminary etching treatment and laser etching treatment of a glass substrate, and obtaining a substrate with a flow channel network; the method comprises the following steps: mixing an oxide alloy with an initiator to form core slurry, and performing pulse laser deposition treatment on the core slurry to form a capsule structure coated with an inner layer and an outer layer; in a vacuum environment, filling the capsule structure into the flow channel network, depositing a graphene film on the upper surface of the filled flow channel network, and electroplating a copper wiring layer to obtain an integrated structure; and bonding a stress sensor array and an infrared laser array at a preset interval on the edge of the integrated structure to obtain the self-repairing chip. The problem of self-repairing failure caused by advanced liquefaction loss of a repairing agent and passive mechanical triggering lag in the prior art is solved, and the continuous service life of the chip is prolonged under the conditions of temperature circulation of-65 DEG C to 250 DEG C and high-frequency vibration impact.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of semiconductor device manufacturing technology, and in particular to a self-healing chip and its preparation method. Background Technology

[0002] In the modern electronics industry, glass-based chips serve as the carrier and interconnect components for electronic devices, applied in aerospace, automotive electronics, and consumer products. However, under extreme temperature cycling (e.g., -65°C to 250°C) and complex mechanical stresses (e.g., vibration, shock, bending), glass-based chips are highly susceptible to microcracks. These microcracks originate from material fatigue or thermal expansion coefficient mismatch, gradually propagating into the conductive path, causing a sharp increase in resistance or even open-circuit failure, ultimately leading to chip malfunction. Once formed, microcracks can propagate at a rate of micrometers per second, making real-time intervention difficult with existing passive protection methods, seriously threatening system reliability.

[0003] In existing technologies, crack healing is mainly achieved by filling ceramic capsules with a low-melting-point tin-bismuth alloy. When a microcrack propagates to the capsule, the tin-bismuth alloy melts and flows to fill the crack, forming a temporary conductive bridge. However, the melting point of the tin-bismuth alloy is only 139°C, which prematurely liquefies and fails when the chip operating temperature exceeds 400°C. This causes the repair agent to be lost before the crack appears, resulting in a severely delayed repair response and insufficient heat resistance. At the same time, capsule rupture depends on the mechanical propagation of the crack, and the response time is usually several seconds to several minutes, which is far behind the instantaneous propagation process of the microcrack, making real-time repair impossible.

[0004] Therefore, a new type of self-healing chip is urgently needed to solve the above problems. Summary of the Invention

[0005] The purpose of this application is to provide a self-healing chip and its preparation method, thereby solving the technical problem that the self-healing chip in the prior art has a delayed repair response and cannot repair in real time.

[0006] To achieve this objective, the present application adopts the following technical solution: A method for fabricating a self-healing chip, comprising: A glass substrate is subjected to preliminary etching to form a gradient nanoporous glass substrate, and the gradient nanoporous glass substrate is subjected to laser etching to obtain a substrate with a flow channel network. An oxide alloy is mixed with an initiator to form a core slurry, and the core slurry is subjected to pulsed laser deposition to form a capsule structure with inner and outer layers. In a vacuum environment, the capsule structure is filled into the flow channel network, and a graphene film is deposited on the upper surface of the filled flow channel network and a copper wiring layer is electroplated to obtain an integrated structure. The self-healing chip is obtained by bonding a stress sensor array and an infrared laser array with a preset spacing to the edge of the integrated structure. The stress sensor array is connected to the analysis module, which analyzes the stress distribution on the chip and feeds it back to the control module. The infrared laser array receives commands from the control module and emits laser light to irradiate the capsule structure.

[0007] Furthermore, the oxide alloy is an oxide alloy of indium, gallium and zinc, with indium accounting for 60-65% of the molar ratio, gallium accounting for 20-25% of the molar ratio, and zinc accounting for 10-15% of the molar ratio.

[0008] Further, the step of performing preliminary etching on the glass substrate to form a gradient nanoporous glass substrate includes: After the high-purity borosilicate glass substrate is ultrasonically cleaned with deionized water and dried, it is placed in a sealed reaction chamber. Selective etching is performed by introducing a vapor phase etchant containing hydrofluoric acid into the sealed reaction chamber, controlling the etching time to be 15-20 minutes and the etching temperature to be 45-50°C, to obtain a gradient nanoporous glass substrate. The gradient nanoporous glass substrate comprises a surface layer, a middle layer, and a bottom layer. The surface layer has a pore size of 50-55 nm, the middle layer has a pore size of 150-155 nm, and the bottom layer has a pore size of 80-85 nm.

[0009] Further, the step of performing laser etching on the gradient nanoporous glass substrate to obtain a substrate with a flow channel network includes: A continuous semiconductor laser is passed through the gradient nanoporous glass substrate for full-field scanning irradiation. Selective fluorescent labeling is induced based on the refractive index difference between the surface pore size and the middle pore size to obtain a labeled substrate with a three-dimensional positioning grid. The marking substrate with the three-dimensional positioning grid is subjected to a carbon dioxide laser for zoned preheating treatment, reducing the viscosity of the surface glass to 10. 6 ~10 7 Pa·s and the middle layer pores remain intact, resulting in a softened marked substrate; The softened marked substrate was etched using a femtosecond laser in a three-dimensional fractal scanning manner, with an energy density of 0.1~0.2 pJ / μm. 2 A substrate with a tree-like fractal microchannel network is obtained by dynamically shaping the beam into a tree-like fractal path through a spatial light modulator with a pulse width of 150~160 femtoseconds and a repetition frequency of 1~2MHz.

[0010] Further, the step of mixing the oxide alloy with an initiator to form a core slurry, and then performing pulsed laser deposition on the core slurry to form a capsule structure with inner and outer layers, includes: Indium gallium zinc oxide alloy was mixed with gold nanorod-polydopamine composite initiator, wherein the aspect ratio of the gold nanorod-polydopamine composite initiator was 10:1, to obtain core slurry; The core slurry surface is coated with a boron nitride inner layer to obtain an intermediate capsule structure with a boron nitride inner layer coating; Plasma sintering is performed on the boron nitride inner layer of the intermediate capsule structure to form a silicon carbide outer shell, thus obtaining the capsule structure. The capsule structure is subjected to carrier dispersion treatment, so that the capsule is suspended in a perfluoropolyether carrier with a capsule concentration of 30~35 vol%, resulting in a capsule structure with inner and outer coatings.

[0011] Further, the step of filling the capsule structure into the flow channel network under vacuum includes: In 10 -3 ~10 -4 In a vacuum environment of Pa, a capsule structure suspended in a perfluoropolyether carrier is uniformly injected into a flow channel network at an injection rate of 5 μL / min using a microfluidic injection device to obtain a flow channel network substrate. The flow channel network substrate is subjected to low-temperature plasma polymerization treatment by introducing fluorinated monomer gas to obtain a microchannel structure including a fluorinated protective film.

[0012] Further, the step of depositing a graphene film and electroplating a copper wiring layer on the upper surface of the filled flow channel network to obtain the integrated structure includes: The pattern of the preset mold is photolithographically etched on the upper surface of the filled flow channel network, and a graphene film is deposited in the defined area of ​​the pattern using chemical vapor deposition to obtain a graphene-covered substrate. The integrated structure is obtained by electroplating a thick copper wiring layer on the upper surface of the graphene-covered substrate.

[0013] Further, the step of bonding a stress sensor array and an infrared laser array with a preset spacing at the edges of the integrated structure to obtain the self-healing chip includes: The edges of the integrated structure are subjected to reactive ion etching to form a bonding platform. A stress sensor array with a bonding spacing of 40~50μm is bonded to the bonding platform and electrically connected to the graphene thin film wiring layer. An infrared laser array is bonded to the bonding platform with a spacing of 40~50μm, and the infrared laser array is electrically connected to the graphene thin film wiring layer to form an assembly structure. The graphene film wiring layer and stress sensor array of the assembly structure are connected to the analysis module signal, and the output of the analysis module is connected to the input of the infrared laser array to obtain the self-healing chip.

[0014] This application also discloses a self-healing chip, which is prepared using the self-healing chip preparation method described in any of the above claims, wherein the self-healing chip comprises: A glass substrate having a gradient nanoporous layer, a flow channel network within the glass substrate having the gradient nanoporous layer, a capsule structure filled within the flow channel network, the capsule structure having an inner boron nitride coating and an outer silicon carbide coating, an oxide alloy repair material encapsulated within the capsule, and chip pads for connecting chips provided on the flow channel network; An integrated structure is disposed on the upper surface of a glass substrate with gradient nanopores, including a graphene film deposited on the upper surface of a flow channel network and a wiring layer electroplated on the graphene film. The edge of the integrated structure is bonded with a stress sensor array and an infrared laser array with a preset spacing. The stress sensor array is connected to the analysis module to monitor the stress distribution on the chip and feed the data back to the analysis module. The output of the analysis module is connected to the input of the infrared laser array and can issue control commands based on the stress distribution data. After receiving the control command, the infrared laser array can directionally emit laser light to the capsule structure to trigger a self-healing reaction.

[0015] Furthermore, the flow channel network is a tree-like fractal microchannel network with a three-dimensional tree-like fractal structure. The width of the main flow channel of the flow channel network is 1~1.2μm, the width of the branch flow channel is 0.6~0.9μm, and the width of the terminal flow channel is 0.3~0.5μm. The capsule structure is uniformly distributed in the terminal flow channel.

[0016] Compared with the prior art, this application has the following beneficial effects: This application discloses a self-healing chip and its fabrication method. A stress sensor array monitors the local stress distribution of the chip in real time. An analysis module calculates the risk coordinates and feeds them back to the control module when stress anomalies first appear. An infrared laser array receives a command and precisely irradiates the target capsule area with pulsed lasers, causing the oxide alloy capsule with inner and outer layers to selectively rupture within 0.05 seconds. With the assistance of an initiator, the core slurry rapidly melts and is injected into potential or existing microcracks along a gradient nanopore channel network at a flow rate greater than 15 μm / ms. A graphene film and a copper wiring layer work together to provide highly conductive and thermally conductive channels to accelerate heat transfer and alloy solidification. Finally, repair is completed before the crack propagates to the conductive path. This solves the self-healing failure problem in existing technologies caused by premature liquefaction and loss of the repair agent and the lag of passive mechanical triggering. It extends the chip's continuous service life under temperature cycling from -65℃ to 250℃ and high-frequency vibration and shock conditions, significantly improving system stability and maintenance economy in high-reliability fields such as aerospace and automotive electronics. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] The structures, proportions, sizes, etc., shown in the accompanying drawings are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed in the specification, and are not intended to limit the implementation conditions of this application. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size should still fall within the scope of the technical content disclosed in this application, provided that they do not affect the effects and purposes that this application can produce.

[0019] Figure 1 This is a schematic diagram illustrating the overall steps of the self-healing chip fabrication method. Figure 2 This is a schematic diagram of the self-healing chip.

[0020] The figures are labeled as follows: 1. Glass substrate; 2. Gradient nanopore layer; 3. Flow channel network; 4. Wiring layer; 5. Chip pad. Detailed Implementation

[0021] To make the inventive objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] In the description of this application, it should be understood that the terms "upper," "lower," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. It should be noted that when a component is considered to be "connected" to another component, it can be directly connected to the other component or there may be a component centrally located at the same time.

[0023] The technical solution of this application will be further described below with reference to the accompanying drawings and specific embodiments.

[0024] refer to Figure 1 This application provides a method for fabricating a self-healing chip, comprising: S1: Heating the first epoxy resin matrix yields a low-viscosity epoxy resin solution; In step S1, a high-purity borosilicate glass substrate with a thickness of 100 μm is selected as the starting material. Selective etching is performed in a cleanroom environment using hydrofluoric acid vapor phase deposition. The hydrofluoric acid gas reacts controllably with the silicon-oxygen bonds on the glass surface. The etching time is controlled at 15 minutes, and the temperature is maintained at 45°C, thereby regulating the etching depth and pore size distribution, forming a dense nanoporous layer with a pore size of approximately 50 nm on the substrate surface. This layer primarily provides high-density adsorption sites to enhance the adhesion of subsequent fillers. The middle layer gradually transitions to a transition pore size of 150 nm through etching expansion. This gradient design optimizes the resistance gradient of fluid transport, providing a dynamic guiding channel for laser etching and preventing sidewall damage caused by energy diffusion in the uniform pore structure. The bottom layer has anchoring holes with a pore size of 80 nm. These anchoring holes, like microscopic anchors, are embedded deep into the substrate, improving the overall mechanical interlocking strength and ensuring the chip maintains structural stability under high-temperature pressing or mechanical vibration environments, forming a gradient nanoporous glass substrate. The glass substrate was placed inside a vacuum chamber to eliminate air interference and oxidation risks. An 800nm ​​femtosecond laser was used as the etching tool, with a laser speed of 0.1pJ / μm. 2The energy density is used for three-dimensional scanning etching along a preset digital model with a pulse width of 150 femtoseconds. The short pulse design ensures that the heat-affected zone is less than 0.5μm, thus achieving a cold processing effect. After etching, a tree-like fractal microchannel network with a width of approximately 1μm and a depth of 5μm is formed inside the substrate. The network coverage exceeds 95%, and the branch angle is optimized to 60°, ensuring uniform flow and retention of the repair medium at the branch ends. During the etching process, the 150nm pore size in the middle layer of the gradient nanopore serves as a dynamic guiding channel for the laser focus, preventing energy diffusion from increasing the sidewall roughness and achieving sub-micron level etching precision. After etching, nitrogen gas is introduced for rapid cooling to release residual thermal stress and prevent oxidation. The depth error is controlled within ±0.1μm. At this point, the microchannel network and the substrate form an integrated vertical structure, reserving a highly interconnected path for low-resistance injection of the capsule suspension.

[0025] S2: Mix the oxide alloy with an initiator to form a core slurry, and perform pulsed laser deposition on the core slurry to form a capsule structure with inner and outer layers; In step S2, an indium gallium zinc oxide alloy is selected as the core material, with an atomic ratio controlled at In:Ga:Zn = 60%:25%:15%, stabilizing the alloy's melting point at around 15°C to maintain its liquid fluidity at room temperature. The zinc content exceeding 10% significantly enhances the chemical bonding affinity with the glass substrate, preventing the repair medium from detaching from the substrate after triggering and causing repair failure. The melting process is carried out in a high vacuum environment, with the vacuum level maintained at 10... -6 To eliminate the risk of oxygen contamination, the alloy raw materials, after multiple zone melting purification processes, were placed in a quartz crucible and gradually heated to complete liquefaction using radio frequency induction heating. Simultaneously, the melt temperature fluctuation was monitored to ensure it did not exceed ±0.5℃, preventing phase separation. A gold nanorod-polydopamine composite initiator with an aspect ratio of 10:1 was selected and uniformly dispersed in the liquid alloy. The surface plasmon resonance effect of the gold nanorods was utilized to absorb near-infrared light energy, while the polydopamine layer provided biocompatibility and adhesion, preventing nanorod aggregation in the slurry. The core slurry was placed on a rotating target stage in an independent reaction chamber, with the chamber vacuum also maintained at 10℃. -6 Pa is used to avoid oxidation. For the inner coating, boron nitride is used as the target material, and the pulsed laser wavelength is selected as 248 nm with a pulse energy density of 0.5 J / cm². 2 The repetition frequency was 100Hz, and layer-by-layer deposition was performed under a substrate temperature controlled at 200℃, with a deposition thickness accurate to 20nm. The inner boron nitride layer provided excellent thermal conductivity (over 300W / m·K), ensuring rapid heat conduction to the core alloy during repair and preventing external corrosion of the liquid metal. The target material was switched to silicon carbide, and the laser parameters were adjusted to an energy density of 0.8J / cm². 2At a frequency of 150Hz and a substrate temperature of 300℃, a 30nm thick silicon carbide shell is formed through plasma-assisted sintering. This outer layer increases the heat resistance of the capsule from 400℃ to 550℃, sufficient to withstand the instantaneous thermal shock of the chip during high-temperature pressing or laser triggering. A small amount of argon gas is introduced as a plasma source during the sintering process to enhance particle mobility, resulting in a dense, non-porous shell with a surface roughness controlled within 2nm. Cross-sectional analysis using transmission electron microscopy confirms a clear inner-outer boundary and the absence of a diffusion layer. After preparation, the capsule is transferred to a perfluoropolyether carrier to form a 30vol% injection slurry. The high wettability and low volatility of this carrier ensure that the capsule remains suspended and does not deposit during subsequent vacuum filling, while eliminating the potential ion residue hazards associated with traditional chemical initiators.

[0026] S3: In a vacuum environment, the capsule structure is filled into the flow channel network, a graphene film is deposited on the upper surface of the filled flow channel network and a copper wiring layer is electroplated to obtain an integrated structure. In step S3, the capsule suspension is used as the injection slurry and injected into the dendritic microchannel network at a low speed of 5 μL / min using an injection pump. The injection process utilizes femtosecond laser etching to achieve submicron precision sidewall smoothness (heat-affected zone less than 0.5 μm) to prevent eddy deposition or local blockage at bends. After filling, an 808 nm near-infrared pre-scan is applied to verify channel patency. After the filling stage, a 10 nm thick fluorinated protective film is formed at the channel inlet using low-temperature plasma polymerization. Graphene film deposition and copper wiring layer electroplating are performed on the sealed substrate surface. A patterned mold is transferred using nanoimprint lithography with a linewidth of 200 nm, an imprinting pressure of 50 MPa, and a temperature of 120 °C, ensuring the mold nodes are vertically aligned with the underlying microchannels. A single-layer graphene film is deposited as a resistance sensing layer using chemical vapor deposition. The graphene mesh is embedded in the subsequent redistribution layer (RDL) and electrically connected to the peripheral pins. Then, copper electroplating is performed to form a 2μm thick wiring layer. The electroplating process is carried out on the filled and sealed surface to obtain an integrated structure. In the integrated structure, the microfluidic network acts as a self-healing reserve, and the graphene film and the copper wiring layer achieve electrical interconnection.

[0027] S4: Bond a stress sensor array and an infrared laser array with a preset spacing to the edge of the integrated structure to obtain the self-healing chip.

[0028] In step S4, using the integrated RDL chip as the main body, a MEMS stress sensor array is batch-etched and bonded to the edge of the glass substrate using microelectromechanical systems (MEMS) technology. The sensor spacing is controlled at 50 μm, forming a high-density stress monitoring network. Optimized spacing improves edge space utilization and enables real-time detection of minute stress fluctuations, avoiding the blind spot problem caused by traditional single sensors. The convolutional neural network AI module chip is then mounted onto the RDL pins using flip-chip bonding. Copper interconnects connect the stress sensor array, graphene grid, and near-infrared laser array, forming a complete closed-loop feedback system. The entire assembly process is performed under a nitrogen protective atmosphere, with the bonding temperature strictly controlled below 250°C to prevent thermal damage to the underlying capsule structure and ensure the capsule's integrity and photothermal response performance. The AI ​​module's training dataset covers 100,000 stress simulation scenarios, enabling efficient analysis of the stress distribution on the chip. When a risk threshold greater than 50 MPa is detected, the AI ​​module immediately generates control commands and feeds them back to the control module. The control module then drives an infrared laser array to irradiate the capsule structure with a laser wavelength optimized to 808 nm and a pulse duration of only 0.1 seconds. This short-duration, high-energy irradiation fully utilizes the heat resistance of the outer silicon carbide shell and the thermal conductivity of the inner boron nitride layer, causing the indium gallium zinc oxide alloy in the core slurry to melt rapidly and be released to the damaged area, achieving self-repair. After repair, the graphene mesh provides real-time feedback on the impedance change rate. If the change rate is less than 0.5%, the AI ​​module automatically terminates the laser irradiation to avoid energy waste caused by over-repair. The entire system has a response speed of 0.5 seconds and a conductivity recovery rate of up to 99.5%. Compared to existing technologies, it can predict latent cracks 10-30 seconds in advance, thus systematically solving the damage problem of advanced glass-based chips. The bonding of the infrared laser array also employs MEMS (Micro-Electro-Mechanical Systems) technology. It is arranged parallel to the stress sensor array at the substrate edge, with the laser spacing matched to the sensor spacing at 50 μm. This ensures the laser focus is precisely aligned with the capsule positions within the microfluidic network, achieving an irradiation accuracy error of less than 0.2 μm. After receiving commands from the control module, the laser array can selectively activate specific lasers based on the stress distribution map analyzed by AI, irradiating only high-risk areas, further improving energy efficiency and targeted repair.A stress sensor array continuously monitors and inputs the data into the AI ​​module at a sampling rate of 100kHz. The AI ​​uses a deep learning algorithm to analyze the stress gradient and predict the crack propagation path. The control module acts as a bridge to convert the prediction results into laser commands. The infrared laser array then irradiates and triggers the capsule to rupture and release the repair medium. The repair medium rapidly diffuses along the dendritic microchannel network, filling the crack and restoring the conductive path. The graphene film acts as a feedback layer to monitor the repair effect in real time, forming a dynamic and adaptive repair cycle. This achieves the unity of proactive prevention and immediate repair. The low-temperature characteristics of the bonding process protect the integrity of the gradient nanopores and the micro-nano structure of the capsule, avoiding secondary damage induced by thermal stress. The resulting self-healing chip functionally realizes stress analysis, feedback control, and laser irradiation mechanisms.

[0029] Specifically, when the chip experiences mechanical stress or thermal shock during use, leading to microcracks, the stress sensor array detects localized stress anomalies. Sensors with a sensitivity of 0.01 MPa, arranged in a 50 μm spacing array, capture the stress distribution and transmit the data to the analysis module via the copper wiring layer in the RDL at a 100 kHz sampling rate. This module uses a convolutional neural network AI to analyze the stress distribution on the chip. When a risk area exceeding 50 MPa is detected, it immediately feeds back to the control module. The control module then issues a command to activate the infrared laser array, which precisely targets the damaged area and emits 808 nm near-infrared laser light. With an irradiation time of 0.1 seconds, the laser energy is efficiently absorbed by the gold nanorod-polydopamine composite initiator in the capsule structure, triggering a photothermal effect that rapidly melts the indium gallium zinc oxide alloy core into a liquid state. The melting point of only 15°C ensures rapid flow, while the excellent thermal conductivity of the boron nitride inner layer evenly diffuses the heat. The silicon carbide outer layer ruptures at a heat resistance of 550°C, releasing the liquid alloy. This liquid alloy rapidly migrates along the branches of the dendritic fractal microchannel network to the crack, where it utilizes the high affinity of zinc to chemically bond with the glass substrate, filling and solidifying to repair the crack. During the repair process, the graphene film acts as a resistance sensing layer to monitor impedance changes in real time with a resolution of 0.1%. When the impedance change rate recovers to less than 0.5%, the feedback signal terminates the laser irradiation. The entire process has a response speed of only 0.5 seconds and a conductivity recovery rate of up to 99.5%. The repair medium is evenly distributed without dead corners. The AI ​​module's predictive ability can identify hidden damage 10-30 seconds in advance, preventing crack propagation. This closed-loop system seamlessly integrates sensing, analysis, triggering, and repair, solving the problem of brittle failure in traditional chips and achieving active self-healing.

[0030] In another embodiment, this application selects a comparative example for performance testing, and the test results are shown in Figure 1 below: Table 1: The tests were conducted using the same glass substrate (100μm borosilicate glass), the same lamination process (300℃, 50MPa), and the same damage model (50MPa local stress-induced 1μm crack). Performance indicators were statistically analyzed after 1000 thermal cycles from -65℃ to 250℃. This embodiment employs gradient nanopores, dendritic microchannels (1μm width), double-shell capsules (boron nitride 20nm / silicon carbide 30nm), photothermal triggering (gold nanorods-polydopamine), a MEMS sensor array (50μm spacing), and its AI closed-loop control system to achieve an internalized repair structure and an active prediction mechanism.

[0031] Comparative Example 1 had the following differences: the core was a tin-bismuth alloy (conductivity 8.7 MS / m, melting point 139℃); a single alumina shell (thickness 50 nm, heat resistance 380℃); no internal microchannels, the repair agent diffused through the surface; and passive thermal triggering. The results showed that the capsule rupture rate reached 38% during compression at 300℃, the oxidation of the tin-bismuth alloy at high temperatures led to a 15% decrease in conductivity, the repair was delayed by 28 seconds (requiring the crack to heat up to 139℃), and the final conductivity recovery rate was only 85%.

[0032] Comparative Example 2 differed in the following conditions: the repair agent was liquid gallium-indium alloy; a 5μm micro-hole array was etched on the surface (occupying 60% of the wiring area); a peroxide chemical initiator coating was applied; there was no sensor array, and the initiator was exposed through crack propagation. The results showed that the micro-hole structure limited the wiring density to 1800 lines / mm. 2 The initiator residue increased the corrosion rate of copper wiring by 42%, and the repair took 35 seconds (the crack extended to the initiator layer). The conductivity recovery rate after repair was 78%.

[0033] In summary, this embodiment, with its silicon carbide outer shell and boron nitride inner layer, increases the heat resistance threshold from ≤400℃ to 550℃, while reducing the compression breakage rate to 0%; 1μm internal microchannels replace surface micropores, achieving a wiring density of 10klines / mm. 2 MEMS+AI enables crack incubation period capture (predicting 10-30 seconds in advance), photothermal triggering replaces chemical initiators, and response time is compressed to 0.5 seconds; InGaZnO alloy + zinc element bonding + dendritic fractal network ensure uniform filling, conductivity recovery rate of 99.5%, and no ion contamination.

[0034] In one embodiment, the oxide alloy is an oxide alloy of indium, gallium, and zinc, wherein, calculated by molar ratio, indium accounts for 60-65%, gallium accounts for 20-25%, and zinc accounts for 10-15%.

[0035] In one embodiment, the step of performing a preliminary etching process on the glass substrate to form a gradient nanoporous glass substrate includes: After the high-purity borosilicate glass substrate is ultrasonically cleaned with deionized water and dried, it is placed in a sealed reaction chamber. Selective etching is performed by introducing a vapor phase etchant containing hydrofluoric acid into the sealed reaction chamber, controlling the etching time to be 15-20 minutes and the etching temperature to be 45-50°C, to obtain a gradient nanoporous glass substrate. The gradient nanoporous glass substrate comprises a surface layer, a middle layer, and a bottom layer. The surface layer has a pore size of 50-55 nm, the middle layer has a pore size of 150-155 nm, and the bottom layer has a pore size of 80-85 nm.

[0036] In this embodiment, a high-purity borosilicate glass substrate with a thickness of 100 μm was selected as the starting material. It underwent ultrasonic cleaning with deionized water in a cleanroom environment to remove surface contaminants. Subsequently, it was dried to ensure the substrate was dry and residue-free. The substrate was then placed in a sealed reaction chamber, and a vapor-phase etchant containing hydrofluoric acid was introduced into the chamber for selective etching. By controlling the etching time to 15–20 minutes and the temperature to 45–50 °C, phase separation and selective dissolution of the borosilicate glass were achieved. This resulted in the formation of dense nanopores with a pore size of 50–55 nm on the substrate surface, transitional pores extending to a pore size of 150–155 nm in the middle layer, and anchoring pores converging to a pore size of 80–85 nm at the bottom layer. This gradient pore structure design utilizes the difference in etching rates between the borate and silicate phases in the borosilicate glass, enhancing the mechanical bonding strength between the subsequent microfluidic network and the substrate. This provides multi-scale channels for the vacuum filling of the capsule structure, allowing the capsules to be uniformly distributed during injection and avoiding aggregation.

[0037] In one embodiment, the step of laser etching the gradient nanoporous glass substrate to obtain a substrate with a flow channel network includes: A continuous semiconductor laser is passed through the gradient nanoporous glass substrate for full-field scanning irradiation. Selective fluorescent labeling is induced based on the refractive index difference between the surface pore size and the middle pore size to obtain a labeled substrate with a three-dimensional positioning grid. The marking substrate with the three-dimensional positioning grid is subjected to a carbon dioxide laser for zoned preheating treatment, reducing the viscosity of the surface glass to 10. 6 ~10 7 Pa·s and the middle layer pores remain intact, resulting in a softened marked substrate; The softened marked substrate was etched using a femtosecond laser in a three-dimensional fractal scanning manner, with an energy density of 0.1~0.2 pJ / μm. 2 A substrate with a tree-like fractal microchannel network is obtained by dynamically shaping the beam into a tree-like fractal path through a spatial light modulator with a pulse width of 150~160 femtoseconds and a repetition frequency of 1~2MHz.

[0038] In this embodiment, a gradient nanoporous glass substrate is used as the workpiece and placed inside a vacuum chamber. A continuous semiconductor laser is passed through it for full-field scanning irradiation. The difference in refractive index between the surface pore size (50-55 nm) and the middle pore size (150-155 nm) induces selective fluorescent labeling, thereby generating a labeled substrate with a three-dimensional positioning grid. Then, the labeled substrate is subjected to a carbon dioxide laser for zoned preheating treatment, reducing the viscosity of the surface glass to 10. 6 ~10 7 Pa·s, while the middle layer pores remain intact, resulting in a softened labeled substrate. The small pore size of the dense nanopores on the surface leads to a high refractive index of the glass medium (approximately 1.50~1.52), while the larger pore size in the middle layer results in a local decrease in glass density and a relatively low refractive index (approximately 1.45~1.47). This difference in refractive index will produce interface reflection and refractive gradient effects under laser irradiation, thereby inducing selective scattering and fluorescence enhancement of laser photons at the middle layer pores. Because the larger pore size in the middle layer allows more laser energy to be captured and interact with borosilicate defect sites in the glass (such as Si-O bond breakage or boron-oxygen vacancies), the intensity of the excited fluorescence signal (mainly blue-green light in the 450~550nm band) is much higher than that of the surface and bottom layers, thus achieving three-dimensional selective fluorescence labeling of the middle layer pores. At the same time, the high refractive index interface of the surface layer further amplifies this contrast, forming a clear fluorescence intensity gradient distribution. These fluorescent labeled points are captured by a high-speed CCD imaging system as a virtual three-dimensional positioning grid for subsequent laser etching. Three-dimensional fractal scanning etching is performed using femtosecond lasers with energy density controlled at 0.1~0.2pJ / μm², pulse width at 150~160 femtoseconds, and repetition frequency at 1~2MHz. The beam is dynamically shaped into a tree-like fractal path using a spatial light modulator to form a substrate with a tree-like fractal microchannel network.

[0039] In one embodiment, the step of mixing an oxide alloy with an initiator to form a core slurry, and then performing pulsed laser deposition on the core slurry to form a capsule structure with inner and outer layers, includes: Indium gallium zinc oxide alloy was mixed with gold nanorod-polydopamine composite initiator, wherein the aspect ratio of the gold nanorod-polydopamine composite initiator was 10:1, to obtain core slurry; The core slurry surface is coated with a boron nitride inner layer to obtain an intermediate capsule structure with a boron nitride inner layer coating; Plasma sintering is performed on the boron nitride inner layer of the intermediate capsule structure to form a silicon carbide outer shell, thus obtaining the capsule structure. The capsule structure is subjected to carrier dispersion treatment, so that the capsule is suspended in a perfluoropolyether carrier with a capsule concentration of 30~35 vol%, resulting in a capsule structure with inner and outer coatings.

[0040] In this embodiment, an indium gallium zinc oxide alloy (indium 60-65%, gallium 20-25%, zinc 10-15%) was uniformly mixed with a gold nanorod-polydopamine composite initiator with an aspect ratio of 10:1 to obtain a core slurry. The melting point of this alloy was controlled at around 15°C. A boron nitride inner layer was deposited on the surface of the slurry to form an intermediate capsule structure with a boron nitride inner layer, providing excellent thermal conductivity and chemical inertness. The boron nitride inner layer of the intermediate structure was subjected to plasma sintering to form a silicon carbide shell, achieving a gradient ceramic protection with a heat resistance temperature increased from the traditional 400°C to 550°C. The entire synthesis process was carried out under a vacuum of 10... -6 To avoid oxidative contamination, the zinc content exceeding 10% enhances the chemical bonding affinity to glass. After preparation, the capsules are dispersed in a carrier to suspend them in a perfluoropolyether carrier at a concentration of 30-35 vol%, resulting in a capsule structure with inner and outer layers. The fractal topology of the tree network ensures uniform residence of the capsules at the branch ends, thereby constructing a photothermal responsive self-healing reservoir.

[0041] In one embodiment, the step of filling the capsule structure into the flow channel network under vacuum includes: In 10 -3 ~10 -4 In a vacuum environment of Pa, a capsule structure suspended in a perfluoropolyether carrier is uniformly injected into a flow channel network at an injection rate of 5 μL / min using a microfluidic injection device to obtain a flow channel network substrate. The flow channel network substrate is subjected to low-temperature plasma polymerization treatment by introducing fluorinated monomer gas to obtain a microchannel structure including a fluorinated protective film.

[0042] In this embodiment, a capsule suspension is used as the injection medium, at 10 -3 ~10 -4 In a high-vacuum environment, a dendritic fractal microchannel network is uniformly injected at a rate of 5 μL / min using a microfluidic injection device. The capillary effect of the fractal structure and the anchoring effect of the gradient nanopores at the bottom layer (80-85 nm) achieve bubble-free filling with a coverage rate exceeding 98%. After filling, a fluorinated monomer gas is introduced for low-temperature plasma polymerization, forming a 10 nm thick fluorinated protective film at the channel inlet to prevent carrier volatilization. This embodiment utilizes the submicron precision sidewall smoothness achieved by femtosecond laser etching to prevent eddy deposition of the suspension at bends. The high wettability of the perfluoropolyether carrier ensures stable suspension of the capsule at a depth of 5 μm. The entire microchannel system transforms into a closed, self-healing medium, and the filled channel network substrate exhibits high reliability and sealing.

[0043] In one embodiment, the step of depositing a graphene film and electroplating a copper wiring layer on the upper surface of the filled channel network to obtain the integrated structure includes: The pattern of the preset mold is photolithographically etched on the upper surface of the filled flow channel network, and a graphene film is deposited in the defined area of ​​the pattern using chemical vapor deposition to obtain a graphene-covered substrate. The integrated structure is obtained by electroplating a thick copper wiring layer on the upper surface of the graphene-covered substrate.

[0044] In this embodiment, a sealed glass substrate is used as the substrate. A pre-defined pattern is created on the upper surface of the substrate using photolithography. A single-layer graphene film is then deposited in the defined area of ​​the pattern using chemical vapor deposition as a resistance sensing layer, resulting in a graphene-covered substrate. Electroplating is then performed on the upper surface of the substrate to form a thick copper wiring layer, resulting in an integrated structure. Nanoimprint lithography is used to transfer the patterned mold at a linewidth of 200 nm, an imprinting pressure of 50 MPa, and a temperature of 120 °C, ensuring that the grid nodes are perpendicularly aligned with the underlying microchannels. The flatness of the substrate after filling is used as the imprinting reference, avoiding alignment errors inherent in traditional photolithography. The high sensitivity of graphene enables a resistance change resolution of 0.1%, and the redistribution layer density reaches 10 k lines / mm. 2 It is compatible with 1nm process. After integration, the grid impedance uniformity error is verified to be less than 0.5% through probe testing. At this time, the monitoring grid and microfluidic network form a vertically connected sensing-repair coupling structure, providing a high-fidelity signal channel for real-time data acquisition of the analysis module. It breaks through the limitation of planar integration on vertical space occupation and ensures that the integrated structure has excellent electrical performance and mechanical stability.

[0045] In one embodiment, the step of bonding a stress sensor array and an infrared laser array with a predetermined spacing at the edges of the integrated structure to obtain the self-healing chip includes: The edges of the integrated structure are subjected to reactive ion etching to form a bonding platform. A stress sensor array with a bonding spacing of 40~50μm is bonded to the bonding platform and electrically connected to the graphene thin film wiring layer. An infrared laser array is bonded to the bonding platform with a spacing of 40~50μm, and the infrared laser array is electrically connected to the graphene thin film wiring layer to form an assembly structure. The graphene film wiring layer and stress sensor array of the assembly structure are connected to the analysis module signal, and the output of the analysis module is connected to the input of the infrared laser array to obtain the self-healing chip.

[0046] In this embodiment, reactive ion etching is performed on the edge of the integrated structure to form a bonding platform. A stress sensor array with a spacing of 40-50 μm is bonded on this platform and electrically connected to the graphene thin film wiring layer. Subsequently, an infrared laser array is bonded at the same spacing and electrically connected to the graphene thin film wiring layer to form an assembly structure. The graphene thin film wiring layer and stress sensor array are connected to the analysis module signal, and the output of the analysis module is connected to the input of the infrared laser array to obtain a self-healing chip. In this embodiment, MEMS sensors are batch etched and bonded using microelectromechanical systems (MEMS) technology, achieving a sensitivity of 0.01 MPa. A convolutional neural network analysis module is mounted by flip-chip bonding, and a 100 kHz sampling rate data input is achieved. The training dataset of the analysis module covers 100,000 stress simulations. When a risk greater than 50 MPa is detected, laser irradiation is automatically triggered for 0.1 seconds. The process terminates when the graphene feedback impedance change rate is less than 0.5% after repair. The entire assembly is carried out under nitrogen protection with a bonding temperature below 250°C to protect the integrity of the capsule. Verification was performed by 1000 cycles of testing from -65°C to 250°C without failure. At this time, the chip response speed is 0.5 seconds and the conductivity recovery rate is 99.5%. Compared with the prior art, damage is predicted 10-30 seconds earlier, systematically solving the problem of hidden cracks in glass-based chips and ensuring a complete closed loop of self-healing function.

[0047] In one embodiment, reference Figure 2 The present invention also discloses a self-healing chip, which is prepared by the self-healing chip preparation method described in any of the above claims, comprising a glass substrate 1, wherein the glass substrate 1 has a gradient nanopore layer 2, the glass substrate 1 with the gradient nanopore layer 2 has a flow channel network 3, the flow channel network 3 is filled with a capsule structure, the capsule structure has an inner boron nitride coating and an outer silicon carbide coating, the capsule encapsulates an oxide alloy repair material, and the flow channel network 3 is provided with chip pads 5 for connecting the chip; an integrated structure is disposed on the surface of the glass substrate 1 with gradient nanopores. The integrated structure includes a graphene film deposited on the surface of the flow channel network 3 and a wiring layer 4 electroplated on the graphene film; the edges of the integrated structure are bonded with a stress sensor array and an infrared laser array with a preset spacing; wherein, the stress sensor array is signal-connected to the analysis module to monitor the stress distribution of the chip and feed the data back to the analysis module; the output end of the analysis module is connected to the input end of the infrared laser array, and can issue control commands based on the stress distribution data; after receiving the control command, the infrared laser array can directionally emit lasers to the capsule structure to trigger a self-healing reaction.

[0048] In this embodiment, the working principle is as follows: when the chip is subjected to mechanical stress, the stress sensor array captures stress distribution data in real time and transmits it to the analysis module. The analysis module assesses the damage risk through a preset algorithm. Once the threshold is exceeded, it instructs the infrared laser array to precisely irradiate the capsule structure in the corresponding flow channel network 3. The laser energy is absorbed by the gold nanorod-polydopamine composite initiator inside the capsule, triggering a local photothermal effect that melts the oxide alloy and rapidly flows along the dendritic fractal microfluidic network 3 to fill the crack, achieving self-repair of micron-level damage. At the same time, the graphene film, as a highly sensitive resistive layer, monitors the impedance change before and after repair to verify the effect and terminate the laser irradiation, forming a closed-loop feedback system. The combination of the gradient nanoporous glass substrate and the dendritic fractal flow channel network 3 provides highly connected multi-scale channels, making the repair material uniformly distributed and with low flow resistance, avoiding the clogging or agglomeration problems of traditional uniform channels, and improving the filling coverage to over 98%. The double-coating structure of the boron nitride inner layer and the silicon carbide outer layer gives the capsule a heat resistance temperature of up to 550℃ and strong chemical inertness, preventing alloy leakage during high-temperature processes. The increased zinc content enhances the bonding affinity with the glass substrate 1, improving mechanical stability. Integration of the graphene film and copper wiring layer 4 achieves high-density wiring up to 10k lines / mm², compatible with advanced processes. A resistance resolution of 0.1% supports precise monitoring. The 40-50μm spacing between the edge-bonded sensor and laser array ensures high spatial resolution stress mapping and directional triggering. The overall response speed reaches 0.5 seconds, with a conductivity recovery rate of 99.5%. It has passed 1000 cycles of testing from -65℃ to 250℃ without failure. Compared to existing technologies, it predicts latent cracks 10-30 seconds earlier, solving the chronic problem of brittle fracture in glass-based chips.

[0049] In one embodiment, the flow channel network is a tree-like fractal microchannel network with a three-dimensional tree-like fractal structure. The width of the main flow channel of the flow channel network is 1~1.2μm, the width of the branch flow channel is 0.6~0.9μm, and the width of the terminal flow channel is 0.3~0.5μm. The capsule structure is uniformly distributed in the terminal flow channel.

[0050] In this embodiment, the working principle is based on fluid mechanics and fractal topology optimization. When the infrared laser triggers the capsule to rupture, the molten oxide alloy diffuses stepwise from the trunk to the branches under the capillary effect driven by the dendritic network. The high surface tension of the narrow channel at the end ensures that the repair material is accurately deposited at the damage point, avoiding large-scale flow waste. The bottom 80nm anchoring holes of the gradient nanopores assist in capsule fixation and prevent displacement caused by vibration or thermal expansion and contraction. The laser energy gradually decays along the network path to achieve uniform heating. After repair, the alloy solidifies to form a continuous phase chemically bonded to the substrate. The graphene film provides real-time feedback impedance recovery for closed-loop control. The 60° branch angle of the dendritic fractal path optimizes fluid distribution, with a coverage rate of over 95%. Compared with straight channels, it reduces flow resistance by 70%. An injection speed of only 5μL / min can achieve bubble-free filling. The capsule concentration of 30vol% is uniformly distributed at the ends to avoid agglomeration. The multi-level width tapering design matches the capsule diameter to within 800nm, ensuring 100% end-capsule dwell time and freeing up trunk space for signal transmission. The heat-affected zone is less than 0.5μm to prevent sidewall damage. Furthermore, combined with the sub-micron precision of femtosecond laser etching, the depth error is ±0.1μm, and the sidewall roughness is within 5nm, supporting high vacuum 10... -3 ~10 -4 Pa filling and low-temperature plasma polymerized fluorinated protective film sealing significantly extend chip life and are compatible with 1nm process integration.

[0051] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for preparing a self-repairing chip, characterized in that, The method comprises the following steps: performing a preliminary etching treatment on a glass substrate to form a gradient nanopore glass substrate, performing a laser etching treatment on the gradient nanopore glass substrate to obtain a substrate with a flow channel network; mixing an oxide alloy and an initiator to form a core slurry, performing a pulsed laser deposition treatment on the core slurry to form a capsule structure with an inner layer and an outer layer; filling the capsule structure into the flow channel network in a vacuum environment, depositing a graphene film on the upper surface of the filled flow channel network and electroplating a copper wiring layer to obtain an integrated structure; bonding a stress sensor array and an infrared laser array with a predetermined spacing to the edge of the integrated structure to obtain the self-repairing chip; wherein the stress sensor array is connected to an analysis module, the analysis module is used to analyze the stress distribution of the chip and feed back to a control module, and the infrared laser array is used to receive instructions from the control module and emit laser to irradiate the capsule structure.

2. The method of claim 1, wherein the self-repairing chip is prepared by the steps of: The oxide alloy is an oxide alloy of indium, gallium and zinc, and the molar ratio of indium is 60-65%, the molar ratio of gallium is 20-25%, and the molar ratio of zinc is 10-15%.

3. The method of claim 1, wherein the self-repairing chip is prepared by the steps of: The step of performing a preliminary etching treatment on a glass substrate to form a gradient nanopore glass substrate comprises: After the high-purity borosilicate glass substrate is subjected to deionized water ultrasonic cleaning treatment and drying treatment, it is placed in a sealed reaction chamber; a vapor phase etchant containing hydrofluoric acid is introduced into the sealed reaction chamber to perform a selective etching treatment, the etching time is controlled to be 15-20 minutes, and the etching temperature is controlled to be 45-50°C to obtain a gradient nanopore glass substrate; wherein the gradient nanopore glass substrate comprises a surface layer, a middle layer and a bottom layer, the surface layer has a pore size of 50-55 nm, the middle layer has a pore size of 150-155 nm, and the bottom layer has a pore size of 80-85 nm.

4. The method of claim 3, wherein the self-repairing chip is prepared by the steps of: The step of performing a laser etching treatment on the gradient nanopore glass substrate to obtain a substrate with a flow channel network comprises: a continuous semiconductor laser is introduced into the gradient nanopore glass substrate to perform full-field scanning irradiation, and based on the difference in refractive index between the surface layer and the middle layer, a selective fluorescent label is induced to obtain a labeled substrate with a three-dimensional positioning grid; The mark substrate with three-dimensional positioning grid is preheated by carbon dioxide laser to make the surface layer glass viscosity reduce to 10 6 ~10 7 Pa·s and the middle layer hole remains complete, and a softened mark substrate is obtained. The softened state marking substrate is subjected to three-dimensional fractal scanning etching by a femtosecond laser, and the energy density is 0.1-0.2 pJ / μm 2 The light beam is dynamically shaped into a tree fractal path by a spatial light modulator, and a substrate with a tree fractal micro-channel network is obtained.

5. The method of claim 1, wherein the self-repairing chip is prepared by the steps of: The step of mixing an oxide alloy and an initiator to form a core slurry, and performing a pulsed laser deposition treatment on the core slurry to form a capsule structure with an inner layer and an outer layer comprises: The indium gallium zinc oxide alloy and the gold nanorod-polydopamine composite initiator are mixed, wherein the length-diameter ratio of the gold nanorod-polydopamine composite initiator is 10:1 to obtain a core slurry; The surface of the core slurry is coated with a boron nitride inner layer to obtain an intermediate capsule structure with a boron nitride inner layer coating; The boron nitride inner layer of the intermediate capsule structure is subjected to plasma sintering treatment to form a silicon carbide shell outside the boron nitride inner layer to obtain a capsule structure; The capsule structure is subjected to carrier dispersion treatment to suspend the capsules in a perfluoropolyether carrier, and the concentration of the capsules in the carrier is 30-35 vol% to obtain a capsule structure with an inner layer and an outer layer.

6. The method of claim 1, wherein the self-repairing chip is prepared by the steps of: The step of filling the capsule structure into the flow channel network in a vacuum environment comprises: In 10 -3 ~10 -4 The capsule structure suspended in the perfluoropolyether carrier is uniformly injected into the flow channel network at an injection speed of 5 μL / min through a microfluidic injection device in a vacuum environment of Pa, to obtain a flow channel network substrate. The micro-channel network substrate is subjected to low-temperature plasma polymerization treatment by introducing fluorine-containing monomer gas to obtain a micro-channel structure including a fluorinated protective film.

7. The method of claim 1, wherein the self-repairing chip is prepared by a process comprising: The step of depositing a graphene film on the upper surface of the filled micro-channel network and electroplating a copper wiring layer to obtain an integrated structure includes: The upper surface of the filled micro-channel network is subjected to photoetching to form a predetermined mold pattern, and a graphene film is deposited on the limited area of the pattern by chemical vapor deposition to obtain a graphene-covered substrate. The upper surface of the graphene-covered substrate is subjected to electroplating thick copper wiring layer treatment to obtain the integrated structure.

8. The method of claim 1, wherein the self-repairing chip is prepared by a method comprising: The step of bonding a stress sensor array and an infrared laser array with a predetermined interval to the edge of the integrated structure to obtain the self-repairing chip includes: The edge of the integrated structure is subjected to reactive ion etching treatment to form a bonding platform, and a stress sensor array with an interval of 40-50 μm is bonded to the bonding platform and electrically connected to the graphene film wiring layer. The bonding platform is bonded to an infrared laser array with an interval of 40-50 μm, and the infrared laser array is electrically connected to the graphene film wiring layer to form an assembled structure. The graphene film wiring layer and the stress sensor array of the assembled structure are connected to the signal of the analysis module, and the output of the analysis module is connected to the input of the infrared laser array to obtain the self-repairing chip.

9. A self-healing chip, characterized by The self-repairing chip is prepared by the method of any one of claims 1-8, and the self-repairing chip includes: A glass substrate having a gradient nanopore layer, the glass substrate with the gradient nanopore layer is provided with a micro-channel network, the micro-channel network is filled with a capsule structure, the capsule structure has an inner layer of boron nitride coating and an outer layer of silicon carbide coating, the capsule encapsulates an oxide alloy repair material, and the micro-channel network is provided with a chip pad connected to a chip; An integrated structure provided on the upper surface of the glass substrate with the gradient nanopore layer, including a graphene film deposited on the upper surface of the micro-channel network and a wiring layer electroplated on the graphene film; The edge of the integrated structure is bonded to a stress sensor array and an infrared laser array with a predetermined interval; The stress sensor array is connected to the signal of the analysis module for monitoring the stress distribution on the chip and feeding back data to the analysis module; the output of the analysis module is connected to the input of the infrared laser array, which can issue control instructions according to the stress distribution data; and the infrared laser array can emit laser light to the capsule structure in a directional manner after receiving the control instructions to trigger a self-repairing reaction.

10. The self-healing chip of claim 9, wherein, The micro-channel network is a tree-shaped fractal micro-channel network with a three-dimensional tree-shaped fractal structure, the main channel width of the micro-channel network is 1-1.2 μm, the branch channel width is 0.6-0.9 μm, and the terminal channel width is 0.3-0.5 μm, and the capsule structures are uniformly distributed in the terminal channels.