Preparation method of resistivity self-adapting regulated conductive copper paste

Abstract

This invention discloses a method for preparing conductive copper paste with adaptive resistivity control, relating to the field of electronic conductive materials technology. The invention involves in-situ polydopamine coating of copper powder in a multi-level pH gradient buffer system to prepare an antioxidant copper powder premix; simultaneously, it synthesizes thermally responsive conductive microspheres with a core-shell structure; subsequently, the two are compounded and dispersed with an organic carrier to obtain an uncured slurry; microscopic characteristic data of the slurry are collected, and the initial curing process is intelligently predicted using a deep belief network with an attention mechanism and an improved sparrow search algorithm; closed-loop correction is performed through micro-trial and error and an improved particle swarm optimization algorithm to finally determine the optimal curing parameters for curing the main slurry. This invention not only solves the problem of easy oxidation of copper paste but also achieves accurate and adaptive control of the resistivity of a single-formulation slurry, improving batch deviation control rate and product consistency and reliability.

Description

Technical Field

[0001] This invention belongs to the field of electronic conductive materials technology, and specifically relates to a method for preparing conductive copper paste with adaptive resistivity control. Background Technology

[0002] With the rapid development of printed electronics, flexible circuits, and smart wearable devices, the market demand for conductive pastes, as key basic materials, is growing daily. Currently, conductive silver paste dominates due to its excellent conductivity and stability, but the high cost of silver limits its application in large-scale, low-cost electronic products. Copper, as a low-cost alternative to silver, has become a research hotspot in the field of conductive pastes due to its similar conductivity and extremely high cost-effectiveness. Developing high-performance, low-cost, and process-adaptable conductive copper pastes is of great significance for reducing the manufacturing cost of electronic products and promoting the industrialization of printed electronics. Summary of the Invention

[0003] Existing conductive copper paste preparation technologies typically involve the physical cleaning of copper powder, the formulation of an organic carrier, and the mechanical mixing of the two. For example, a common preparation method involves acid washing to remove the surface oxide layer of commercially available flake copper powder, followed by mixing with epoxy resin, a curing agent, and a solvent, and then dispersing the mixture using a three-roll mill to obtain a paste. In use, a fixed high-temperature sintering process (e.g., above 180°C) is usually employed to achieve conductive connections between the copper powder particles. Some improved technologies attempt to introduce silver-coated copper powder or add reducing agents to alleviate copper oxidation, but they generally still follow a production model combining a fixed formula and a fixed process.

[0004] This reveals the following problems with existing technologies: 1. Copper powder is highly susceptible to oxidation, and simple physical cleaning cannot provide long-term protection, resulting in a short slurry storage period and rapid decay of conductivity after curing; 2. The resistivity of existing copper pastes is essentially fixed once the formula is determined, which cannot meet the needs of certain applications requiring specific medium to high resistance values ​​(such as heating films and embedded resistors); 3. Due to batch fluctuations in raw materials (such as differences in particle size and surface energy) and environmental factors, the use of fixed curing processes often leads to large resistivity deviations in the final product and poor batch consistency, making it difficult to meet the stringent requirements of precision electronic manufacturing for material performance stability.

[0005] (a) Technical problems to be solved To address the problems in related technologies, this invention provides a method for preparing conductive copper paste with adaptive resistivity control, thereby overcoming the aforementioned technical problems existing in the prior art.

[0006] (II) Technical Solution To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: This invention provides a method for preparing conductive copper paste with adaptive resistivity control, comprising the following steps: S1. In a multi-level pH gradient buffer system, micron-sized flake copper powder and nano-sized spherical copper powder are cascaded for surface modification and in-situ polydopamine coating is carried out to prepare an antioxidant copper powder premix. S2. Polymer microspheres are synthesized using microfluidics or emulsion polymerization as the core. A nanoscale conductive metal shell is constructed on the surface of the core through surface sensitization and activation treatment, as well as chemical plating technology, to obtain thermally responsive conductive microspheres. S3. The antioxidant copper powder premix in S1 is combined with the thermally responsive conductive microspheres in S2 and degassed and premixed with an organic carrier to obtain a premix; the premix is ​​subjected to high-shear three-roll milling to obtain an uncured conductive copper paste. S4. Obtain the interface parameters of the antioxidant copper powder premix, the thermodynamic phase transition parameters of the thermally responsive conductive microspheres, and the rheological viscosity data of the uncured conductive copper paste in S3 to obtain the characteristic data of the uncured conductive copper paste; obtain the target resistivity. S5. Input the characteristic data of the uncured conductive copper paste and the real-time required resistivity into the pre-trained curing process mapping model to obtain the initial curing process parameters. S6. Take a sample of the uncured conductive copper paste from S3, cure it based on the initial curing process parameters, and measure the actual resistivity. Determine whether the error between the actual resistivity and the target resistivity is within the error threshold. If so, use the initial curing process parameters as the final curing process to cure the remaining uncured conductive copper paste. Otherwise, an optimal curing process is found through an optimization algorithm; the optimal curing process is then used as the final curing process to cure the remaining uncured conductive copper paste. Preferably, step S1 includes the following steps: S11. Construct a modified buffer system with a multi-level pH gradient; the modified buffer system with a multi-level pH gradient includes a primary washing solution, an intermediate activation solution, and a final coating solution; the primary washing solution is a mixture of dilute H2SO4 and ethanol; the intermediate activation solution is a weakly acidic solution containing potassium fluorinated titanate; the final coating solution is composed of tris(hydroxymethyl)aminomethane, dopamine hydrochloride, and ammonium persulfate oxidation initiator; S12. Mix the flake copper powder and spherical copper powder, and pass them through the primary cleaning solution and the intermediate activation solution in sequence to obtain the treated copper powder. S13. The copper powder treated in S12 is suspended in the final coating solution in S11; the intensity change of characteristic peaks on the surface of the copper powder is monitored to characterize the growth thickness of the polydopamine film. S14. When the growth thickness of the polydopamine film reaches the preset film thickness threshold, a terminator is immediately added to quench the polymerization reaction. After centrifugation and vacuum drying, an antioxidant copper powder premix is ​​obtained. Preferably, step S2 includes the following steps: Monodisperse core microspheres were prepared using polyethylene glycol, polycaprolactone, or low-melting-point polystyrene microspheres as substrates; the onset temperature of their melting endothermic peak was determined and recorded as the disintegration threshold temperature. S22. Sensitize the nucleospheres obtained in S21 to adsorb Sn²⁺ ions; then generate Pd in ​​situ via a redox reaction. 0 Catalyzing the crystal nucleus yields the activated nucleus; S23. Prepare a chemical silver plating solution, wherein the chemical silver plating solution includes silver nitrate as the main salt, ethylenediamine complexing agent, potassium sodium tartrate reducing agent, and polyvinylpyrrolidone dispersant; under ultrasonic assistance, introduce the activated nucleus from S22 into the plating solution, control the reaction temperature below the disintegration threshold temperature and pH, so that silver ions are reduced and deposited on the surface of the activated nucleus to form a dense silver shell layer, thereby obtaining thermally responsive conductive microspheres. Preferably, step S23 includes the following steps: S231. Establish the shell growth rate equation; S232. Based on the shell growth rate equation, the reduction rate of free silver ions is controlled by adding reducing agent in steps; the consumption rate of silver ions in the plating solution is monitored in real time; when the silver ion conversion rate reaches more than 95% and the absorbance of the mixed solution no longer changes, the shell construction is determined to be complete. Preferably, step S3 includes the following steps: S31. Bisphenol A type epoxy resin, phenolic resin, solvent, leveling agent and thixotropic agent are mixed in a mass ratio and stirred to dissolve, forming a homogeneous organic carrier. S32. The antioxidant copper powder premix prepared in S1, the thermally responsive conductive microspheres prepared in S2, the homogeneous organic carrier prepared in S31, and the latent curing agent are put into a planetary vacuum mixer for degassing and premixing to obtain the premix. S33. Feed the premix from S32 into a three-roll mill; set the speed ratio of the feed roller, middle roller, and discharge roller; and perform grinding using a graded grinding strategy. S34. During the grinding process in S33, an infrared thermal imager is used to monitor the roller surface temperature, and a cooling water circulation system is used to control the slurry temperature to always be below the disintegration threshold temperature of the thermally responsive microspheres. T c Subtracting 20°C, the final product is an uncured conductive copper paste; Preferably, step S5 includes the following steps: S51. Based on the DBN neural network model, construct an initial curing process mapping model; set the maximum number of training iterations and initial weight parameters for the initial curing process mapping model; S52. Collect the interface parameters of the antioxidant copper powder premix, the thermodynamic phase change parameters of the thermally responsive conductive microspheres, and the rheological viscosity data of the uncured conductive copper paste in S3, as well as the corresponding historical required resistivity, from the historical production batch data to obtain historical production batch characteristic data; obtain the curing process parameters corresponding to the historical required resistivity; use the curing process parameters corresponding to the historical required resistivity as tags for the historical production batch characteristic data to obtain historical tag data. S53. The initial curing process mapping model is repeatedly trained using historical production batch feature data and historical tag data. During the training process, an optimization algorithm is combined, and the weight parameters of the initial curing process mapping model are found based on the maximum number of training iterations to obtain the optimal solution. The optimal solution is used as the weight parameters of the initial curing process mapping model to obtain the pre-trained curing process mapping model. The characteristic data of the uncured conductive copper paste and the real-time required resistivity are input into the pre-trained curing process mapping model to obtain the initial curing process parameters. Preferably, step S53 includes the following steps: S531. Initialize the sparrow population location, where each individual sparrow represents a set of potential neural network weight parameters; introduce Tent chaotic mapping to initialize the population; S532. Calculate the fitness value of each sparrow individual, wherein the fitness value is the reciprocal of the mean square error between the model's predicted curing process parameters and historical tag data; filter the sparrow individuals based on their fitness values ​​to obtain discoverer sparrows and follower sparrows; S533. An adaptive Cauchy-Gaussian mutation strategy is introduced to update the position of the discoverer sparrow; at the same time, the sine and cosine algorithms are used to perturb and update the position of the sparrow that joins; the updated sparrow population position is obtained. S534. Iteratively update the population position until the maximum number of iterations is reached, then output the weight parameters corresponding to the globally optimal sparrow position. Preferably, step S6 includes the following steps: S61. Take the uncured conductive copper paste sample from S3, coat it onto the test substrate, place it in a rapid annealing furnace, and heat and cure it according to the initial curing process parameters; after cooling, use a four-probe tester to measure the actual resistivity. S62. Construct the objective function; the objective function is the difference between the actual resistivity and the target resistivity; when the error between the actual resistivity and the target resistivity is ≥ the error threshold, set a search space that includes curing parameters such as heating rate, isothermal temperature, and isothermal time; based on the objective function and combined with the particle swarm optimization algorithm, find the optimal curing parameters. The optimal curing process is used as the final curing process to cure the remaining uncured conductive copper paste. Preferably, S62 includes the following steps: S621. Construct a particle set; Based on the search space and particle set of the solidified parameters, construct an initial particle position set, and use the position of each particle in the initial particle position set as a combination of heating rate, isothermal temperature, and isothermal time; Set the maximum number of optimization iterations; S622. Perform iterative update operations on the particles in the particle set, and each update is within a preset search space; and calculate the fitness value of the position of each particle in the particle set according to the objective function, update the position of each particle in the particle set, and obtain the best individual particle and the global best particle in the particle set in each iteration. The update rule is that if the actual resistivity is greater than or equal to the target resistivity, the algorithm adjusts the parameters along the gradient direction of increasing temperature or increasing time; if the actual resistivity is less than the target resistivity, the algorithm adjusts the parameters along the gradient direction of decreasing temperature or decreasing heating rate. S623. Repeat S622. When the difference between the actual resistivity and the target resistivity is less than the error threshold or the maximum optimization parameter is reached, stop the iteration and use the global best particle as the optimal parameter combination.

[0007] (III) Beneficial Effects The present invention has the following beneficial effects: This invention achieves nanoscale control over the thickness of the polydopamine coating layer on the surface of copper powder by constructing a multi-level pH gradient buffer system and combining it with in-situ Raman spectroscopy monitoring. This dense and controllable-thickness polydopamine coating layer not only effectively isolates oxygen and significantly improves the antioxidant capacity of copper powder, but also optimizes the interfacial compatibility between copper powder and organic carrier through the rich functional groups introduced on the surface. In conjunction with the introduced core-shell structured thermally responsive conductive microspheres, which act as site-blocking barriers at low temperatures and disintegrate to form conductive silver bridges at high temperatures, this invention endows the slurry with the ability to achieve resistivity control across orders of magnitude simply by temperature control without changing the formulation, greatly expanding the application window of the material.

[0008] This invention innovatively introduces a deep learning-based intelligent mapping and closed-loop feedback mechanism for curing processes. By constructing a deep belief network that integrates a multi-head attention mechanism and optimizing model weights using an improved sparrow search algorithm, it achieves prediction from the microscopic properties of materials to the optimal macroscopic curing process. This method can keenly capture the nonlinear impact of subtle fluctuations in raw materials on the final performance and automatically recommend compensatory curing parameters. Compared with the traditional experience-based trial and error method, this intelligent control strategy significantly improves the efficiency and accuracy of process design and effectively solves the problem of unstable product performance caused by batch differences.

[0009] This invention establishes a closed-loop correction process that includes micro-trial and error, intelligent optimization, and main implementation. It utilizes a particle swarm optimization algorithm to perform secondary optimization of the initial process predicted by the model based on measured data, eliminating residual model errors and interference from the actual environment. This improves the resistivity deviation control rate and batch consistency of the final product, far exceeding the control level of traditional methods. This strategy, combining standardized material preparation with intelligent process control, ensures high reliability and consistency of product performance, making it particularly suitable for precision electronics manufacturing where extremely high resistivity accuracy is required.

[0010] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, the drawings can be obtained from these drawings without creative effort.

[0012] Figure 1 This is a schematic flowchart of a method for preparing conductive copper paste with adaptive resistivity control according to the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the invention, and not all embodiments. Based on the embodiments of the invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the invention.

[0014] To resolve the above issues, please refer to [link / reference]. Figure 1 This invention discloses a method for preparing conductive copper paste with adaptive resistivity control, comprising the following steps: S1. In a multi-level pH gradient buffer system, micron-sized flake copper powder and nano-sized spherical copper powder are cascaded for surface modification and in-situ polydopamine coating is carried out to prepare an antioxidant copper powder premix. S2. Polymer microspheres are synthesized using microfluidics or emulsion polymerization as the core. A nanoscale conductive metal shell is constructed on the surface of the core through surface sensitization and activation treatment, as well as chemical plating technology, to obtain thermally responsive conductive microspheres. S3. The antioxidant copper powder premix in S1 is combined with the thermally responsive conductive microspheres in S2 and degassed and premixed with an organic carrier to obtain a premix; the premix is ​​subjected to high-shear three-roll milling to obtain an uncured conductive copper paste. S4. Obtain the interface parameters of the antioxidant copper powder premix, the thermodynamic phase transition parameters of the thermally responsive conductive microspheres, and the rheological viscosity data of the uncured conductive copper paste in S3 to obtain the characteristic data of the uncured conductive copper paste; obtain the target resistivity. S5. Input the characteristic data of the uncured conductive copper paste and the real-time required resistivity into the pre-trained curing process mapping model to obtain the initial curing process parameters. S6. Take a sample of the uncured conductive copper paste from S3, cure it based on the initial curing process parameters, and measure the actual resistivity. Determine whether the error between the actual resistivity and the target resistivity is within the error threshold. If so, use the initial curing process parameters as the final curing process to cure the remaining uncured conductive copper paste. Otherwise, an optimal curing process is found through an optimization algorithm; the optimal curing process is then used as the final curing process to cure the remaining uncured conductive copper paste. The above embodiments prepare an antioxidant copper powder premix by in-situ polydopamine coating of copper powder in a multi-level pH gradient buffer system; simultaneously synthesize thermally responsive conductive microspheres with a core-shell structure; then, the two are compounded and dispersed with an organic carrier to obtain an uncured slurry; the microscopic characteristic data of the slurry are collected, and the initial curing process is intelligently predicted using a deep belief network with an attention mechanism and an improved sparrow search algorithm; and closed-loop correction is performed through micro-trial and error and an improved particle swarm optimization algorithm to finally determine the optimal curing parameters for curing the main slurry; this invention not only solves the problem of easy oxidation of copper paste, but also achieves accurate and adaptive control of the resistivity of a single-formulation slurry, improving the batch deviation control rate and the consistency and reliability of the product.

[0015] The above embodiment S1 includes the following steps: S11. Construct a modified buffer system with a multi-level pH gradient; the modified buffer system with a multi-level pH gradient includes a primary washing solution, an intermediate activation solution, and a final coating solution; the primary washing solution is a mixture of dilute H2SO4 and ethanol; the intermediate activation solution is a weakly acidic solution containing potassium fluorinated titanate; the final coating solution consists of 10-50 mM tris(hydroxymethyl)aminomethane (Tris), 2-8 mg / mL dopamine hydrochloride, and 1-5 mM ammonium persulfate oxidizing initiator, with a dynamic pH adjustment range of 8.0-9.0; In specific implementation, the above embodiment S11 is as follows: This embodiment adopts a gradient buffer system optimized for the surface characteristics of copper powder; the primary cleaning solution is a mixture of 5% dilute sulfuric acid and anhydrous ethanol at a volume ratio of 1:1, used to quickly remove the original oxide layer (CuO / Cu2O) and grease on the surface of copper powder; the intermediate activation solution is a weakly acidic solution containing 0.5% potassium fluorotitanate, used to introduce micro-etching pits on the copper surface to increase the mechanical interlocking points for subsequent coating; the core of the final coating solution lies in the stable pH control capability of the Tris buffer, because the self-polymerization rate of dopamine is extremely sensitive to pH; in this embodiment, an aqueous solution containing 25 mM Tris and 4 mg / mL dopamine hydrochloride is prepared, and 2 mM ammonium persulfate is added as an initiator to accelerate the initial nucleation of the polymerization reaction. At the same time, dilute NaOH solution is added dropwise using an online pH meter to strictly lock the pH value at 8.5±0.05 to ensure that the polymer layer is uniform and dense; S12. Mix flake copper powder and spherical copper powder at a mass ratio of 7:3 to 8:2, and then pass them through a primary cleaning solution to remove the surface oxide layer and through a secondary activation solution to introduce surface active sites to obtain the treated copper powder. In specific implementation, the above embodiment S12 is as follows: flake copper powder (average particle size of 5μm, flake thickness of 0.5μm) is selected as the conductive body, and spherical nano copper powder (average particle size of 100nm) is selected as the filler, with a mass ratio of 7:3; this strategy of using small particles of flake copper powder and large particles of spherical nano copper powder aims to construct a high-density packing structure; the processing is carried out in a reaction vessel with ultrasonic assistance, with primary cleaning for 10 minutes and intermediate activation for 5 minutes, and between each step, the copper powder is washed with deionized water until neutral to prevent acid residue from corroding the copper powder; S13. The copper powder treated in S12 is suspended in the final coating solution in S11; the intensity change of characteristic peaks on the surface of the copper powder is monitored by in-situ Raman spectroscopy to characterize the growth thickness of the polydopamine film. In specific implementation, the above embodiment S13 specifically involves: dispersing the pretreated wet copper powder in the final coating solution, setting the mechanical stirring speed to 400 rpm; inserting the in-situ Raman spectrometer probe below the liquid surface to collect the spectral signal of the copper powder surface in real time; focusing on monitoring the 1580 cm⁻¹.-1 (G band, representing sp² hybrid carbon structure) and 1350 cm -1 The peak intensity of (D band, representing defect structure) is measured; as the polymerization reaction proceeds, the intensity of these two characteristic peaks increases linearly or quasi-linearly with the increase of film thickness; a standard curve of "Raman peak intensity-film thickness" is pre-established in the system; S14. When the growth thickness of the polydopamine film reaches the preset film thickness threshold, a terminator is immediately added to quench the polymerization reaction. After centrifugation and vacuum drying, an antioxidant copper powder premix Cu@PDA is obtained. At the same time, the temperature integral and pH change curves during the modification process are recorded as environmental context data for modeling in S4. In specific implementation, the above embodiment S14 specifically involves: the method for setting the preset film thickness threshold is to determine the inverse relationship between the copper powder oxidation weight gain rate and the PDA film thickness through a salt spray accelerated aging experiment, and to determine that the minimum film thickness to meet the 6-month antioxidant lifespan is 8 nm; secondly, based on the electron tunneling effect formula... J ∝exp(− βd )(in J Represents the tunneling current density; ∝ indicates proportional to, β The attenuation constant is based on the electronic insulation properties of PDA materials; d The maximum critical thickness for electron penetration of the PDA insulating layer was calculated to be 12 nm. The intersection range of the two [8 nm, 12 nm] was taken as the preset film thickness threshold, and in this embodiment, the median value of 10 nm was taken as the control target. When the Raman signal-calculated film thickness reached 10 nm, the system automatically triggered the feeding pump to inject a stop solution containing 5 g of ascorbic acid. Ascorbic acid has strong reducing properties and can instantly consume oxygen free radicals in the system, quenching the further oxidation and polymerization of dopamine and preventing the film layer from becoming too thick and causing a decrease in conductivity. Subsequently, centrifugation was performed, followed by washing with ethanol three times and drying in a vacuum oven at 60 °C for 12 hours. The Cu@PDA powder prepared in this embodiment was dark brown and showed no obvious oxidation discoloration after being placed in the air for 6 months, proving the effectiveness of the coating layer. The above embodiments achieved nanoscale precision control over the thickness of the polydopamine coating layer on the copper powder surface by constructing a multi-level pH gradient buffer system and combining it with in-situ Raman spectroscopy monitoring. This dense and controllable-thickness organic layer not only effectively isolates oxygen and significantly improves the oxidation resistance of the copper powder, but also optimizes the interfacial compatibility between the copper powder and the organic carrier through the abundant functional groups introduced on the surface, laying the foundation for obtaining a stable conductive network in the future.

[0016] The above embodiment S2 includes the following steps: S21. Using polyethylene glycol (PEG), polycaprolactone (PCL), or low-melting-point polystyrene microspheres as the substrate, monodisperse core microspheres with a particle size of 1-3 μm are prepared by soap-free emulsion polymerization or melt dispersion method; the onset temperature of the melting endothermic peak is determined by differential scanning calorimetry (DSC) and recorded as the disintegration threshold temperature. T c ; In specific implementation, the above embodiment S21 is as follows: In this embodiment, polyethylene wax (PE Wax) with a melting point range of 128℃-132℃ is selected as the core material; PE powder is pulverized by an air jet mill, and microspheres with a particle size distribution of 1.5μm-2.5μm are screened using an air jet classifier; to enhance its surface hydrophilicity for subsequent chemical plating, the PE microspheres are placed in a plasma treatment machine and treated at 200W power for 15 minutes in an oxygen atmosphere to introduce hydroxyl and carboxyl functional groups; DSC testing shows that its endothermic peak onset temperature is 130℃, i.e. T c =130℃; S22. The nucleospheres obtained in S21 are sensitized in a SnCl2 / HCl colloidal solution to adsorb Sn²⁺ ions; subsequently, they are activated in a PdCl2 / HCl solution to generate Pd in ​​situ through a redox reaction. 0 Catalyzing the crystal nucleus yields the activated nucleus; In specific implementation, Example S22 is as follows: the sensitization solution is formulated as 10 g / L SnCl2·2H2O + 40 mL / L concentrated HCl; the activation solution is formulated as 0.5 g / L PdCl2 + 10 mL / L concentrated HCl; 50 g of PE microspheres are first ultrasonically stirred in the sensitization solution for 30 minutes, filtered and washed with water, and then stirred in the activation solution for 30 minutes; at this time, the Sn²⁺ adsorbed on the surface of the microspheres reduces Pd²⁺ to catalytically active Pd. 0 The nanocrystalline nuclei, with the microspheres changing from white to grayish-black in appearance, provide nucleation centers for chemical silver plating; S23. Prepare a chemical silver plating solution, wherein the chemical silver plating solution comprises silver nitrate as the main salt, ethylenediamine complexing agent, potassium sodium tartrate reducing agent, and polyvinylpyrrolidone (PVP) dispersant; under ultrasonic assistance, introduce the nucleus activated in S22 into the plating solution, and control the reaction temperature below the disintegration threshold temperature. T c Furthermore, with a pH value of 10-12, silver ions are reduced and deposited on the surface of the activated nucleus, forming a dense silver shell layer with a thickness of 30-80 nm. The above embodiment S23 includes the following steps: S231. Establish a shell growth rate equation; the shell growth rate equation includes the concentration of silver ions and the concentration of reducing agent; the shell growth rate equation is as follows: ;in,k Represents the reaction rate constant. a The reaction order representing the concentration of silver ions. b The reaction order indicates the concentration of the reducing agent. Ea Indicates the activation energy of the reaction. R This represents the ideal gas constant, approximately equal to 8.314 J / (mol·K); T Indicates the thermodynamic temperature of the reaction; Ag + Red indicates the concentration of silver ions, and Red indicates the concentration of reducing agent. In specific implementation, the above embodiment S231 specifically refers to: the parameters k , a、b The method for obtaining Ag involves using a controlled variable method combined with micro-thermodynamic testing; firstly, the temperature and reducing agent concentration are fixed, and then the Ag concentration is changed. + The reaction order a≈1.0 was obtained by monitoring the logarithm of the initial reaction rate and the logarithm of the concentration, and fitting the slope of the straight line. Similarly, the reaction order of the reducing agent was measured. b ≈0.8; secondly, isoconcentration reactions were carried out at different temperatures, using the Arrhenius equation lnk=lnA− Ea / RT fits the reaction rate constant k With temperature T The activation energy was calculated based on the relationship. Ea ≈45kJ / mol and pre-exponential factor A This leads to the derivation of the reaction rate constant in this system. k Approximately 1.5 × 10 3 ; S232. Based on the shell growth rate equation, the reduction rate of free silver ions is controlled by adding reducing agent in steps to prevent self-nucleation reaction in the solution and ensure that the silver layer grows only on the surface of the nucleus. The consumption rate of silver ions in the plating solution is monitored in real time using a quartz crystal microbalance (QCM) or a UV-Vis spectrophotometer. When the silver ion conversion rate reaches more than 95% and the absorbance of the mixed solution no longer changes, the shell construction is considered complete. In specific implementation, the above embodiment S23 is as follows: the chemical silver plating solution is divided into solution A (AgNO3 10g / L + ethylenediamine 20mL / L) and solution B (potassium sodium tartrate 30g / L + PVP K30 2g / L + NaOH to adjust pH to 11); the activated nucleus from S22 is dispersed in solution B, and the temperature is raised to 40℃ (far lower than 10℃). T cUnder mechanical stirring, solution A was slowly added dropwise at a rate of 5 mL / min; PVP, as a dispersant, effectively prevented microsphere aggregation; the reaction lasted for about 60 minutes until the Ag⁺ concentration in the supernatant was detected to be below 50 ppm; SEM scanning electron microscopy revealed that the surface of the activated core was covered with a dense, continuous layer of nano-silver particles with an average shell thickness of about 60 nm, which are thermally responsive conductive microspheres; the absorbance of the mixture no longer changing means that, during the continuous monitoring period, the absorbance fluctuation amplitude at the characteristic wavelength (e.g., 420 nm) is less than the preset absorbance fluctuation amplitude threshold (the preset absorbance fluctuation amplitude threshold in this embodiment is 0.005 AU), which indicates that the reaction has reached its endpoint and the shell construction is complete; The above embodiments successfully prepared thermally responsive switch microspheres with a core-shell structure through precise control of the chemical plating process kinetics. At low temperatures, the microspheres are complete physical spheres, acting as site-occupying barriers. Once the temperature exceeds the disintegration threshold temperature, the core melts and collapses, and the original silver shell layer breaks and is released, forming highly conductive silver bridges between the copper particles. This design gives the slurry an inherent adaptive mechanism, allowing the resistivity to be controlled by temperature without changing the formulation, thus broadening the application window of the material. The above embodiment S3 includes the following steps: S31. Bisphenol A type epoxy resin, phenolic resin, solvent (terpineol, butyl carbitol acetate), leveling agent and thixotropic agent are mixed in a mass ratio of (40-60):(10-20):(20-30):(1-5):(1-5), and stirred at 80°C to dissolve and form a homogeneous organic carrier. In specific implementation, the above embodiment S31 is as follows: 40g of bisphenol A type epoxy resin E-51, 15g of phenolic resin, 25g of terpineol, and 15g of butyl carbitol acetate are added to a stirred tank; the mixture is heated to 80°C and stirred to dissolve; after cooling to room temperature, 3g of BYK-333 leveling agent and 2g of fumed silica (Aerosil R972) are added as thixotropic agents, and the mixture is stirred evenly; this carrier system has good thixotropic properties, which can ensure the fluidity during screen printing and prevent the lines from collapsing after printing; S32. The antioxidant copper powder premix Cu@PDA prepared in S1, the thermally responsive conductive microspheres prepared in S2, the homogeneous organic carrier and the latent curing agent prepared in S31 are put into a planetary vacuum mixer for degassing and premixing to obtain the premix. In specific implementation, the above embodiment S32 is as follows: Weigh the materials according to the following formula: 720g Cu@PDA copper powder, 80g Ag@PE conditioning microspheres, 185g organic carrier, 10g micronized dicyandiamide (latent curing agent), and 5g 2-methylimidazole (accelerator); place the materials in a planetary mixer, set the vacuum degree to -0.098MPa, the revolution / rotation speed to 800rpm, and stir for 15 minutes to obtain a well-wetted coarse slurry; S33. Feed the premix from S32 into a three-roll mill; set the speed ratio of the feed roller, middle roller, and discharge roller to 1:3:9; adopt a graded grinding strategy, with the first pass gap set to 50μm, the second pass gap to 20μm, and the third pass gap to 5μm; S34. During the grinding process in S33, an infrared thermal imager is used to monitor the roller surface temperature, and a cooling water circulation system is used to control the slurry temperature to always be below the disintegration threshold temperature of the thermally responsive microspheres. T c The temperature was reduced by 20°C to prevent the microspheres from pre-disintegrating during the grinding stage, resulting in an uncured conductive copper paste. In specific implementation, the above embodiment S34 is as follows: Three-roll milling is the key to dispersion, but excessive shear heat will cause the thermally responsive conductive microspheres to melt and fail prematurely; in this embodiment, the cooling water circulation inside the rollers is activated, and the water temperature is set at 15°C; an infrared thermometer monitors and displays in real time that the maximum temperature of the slurry during the milling process is controlled below 45°C, which is far lower than the temperature of the microspheres. T c (130℃); The final slurry fineness was less than 12μm as measured by a scraper fineness gauge, and SEM showed that the microsphere structure remained intact. The above embodiments solve the problem of easy failure of thermosensitive microspheres under strong shear force field by strictly controlling the temperature of the grinding process; ensure that the microspheres maintain the complete "core-shell" structure in the uncured state, thereby ensuring that the thermal response switch function can be accurately triggered in the subsequent curing step, which is the process basis for achieving precise control of resistivity.

[0017] The above embodiment S5 includes the following steps: S51. Based on the DBN neural network model, construct an initial curing process mapping model; set the maximum number of training iterations and initial weight parameters for the initial curing process mapping model; In specific implementation, the above embodiment S51 is as follows: This embodiment selects DBN as the feature extractor and introduces a multi-head attention mechanism; it utilizes the nonlinear mapping capability of the DBN to process high-dimensional material property data; at the same time, it introduces a multi-head attention mechanism to capture the long-range dependencies between different features (such as microsphere disintegration threshold temperature, slurry viscosity, and target resistivity), and assigns higher weights to key influencing factors; it constructs an initial curing process mapping model: it sets an input layer (dimension 12, corresponding to material and rheological features), a hidden layer composed of 3 layers of restricted Boltzmann machines (RBM), a multi-head attention layer (8 heads), and an output layer (dimension 3, corresponding to heating rate, isothermal temperature, and isothermal time); it sets the maximum number of training iterations to 1000 times, the initial learning rate to 0.01, and the weight parameters to be initialized through Xavier to obtain the initial curing process mapping model; S52. Collect the interface parameters of the antioxidant copper powder premix, the thermodynamic phase change parameters of the thermally responsive conductive microspheres, and the rheological viscosity data of the uncured conductive copper paste in S3, as well as the corresponding historical required resistivity, from the historical production batch data to obtain historical production batch characteristic data; obtain the curing process parameters corresponding to the historical required resistivity; use the curing process parameters corresponding to the historical required resistivity as tags for the historical production batch characteristic data to obtain historical tag data. In specific implementation, the above embodiment S52 specifically involves: collecting production data from 500 batches over the past 3 years; data preprocessing includes: outlier removal, using the 3σ principle to remove extreme data caused by equipment failure; normalization processing, mapping all feature data to the [0,1] interval to eliminate the influence of dimensions; data augmentation, generating virtual samples by introducing small random noise to expand the dataset and prevent overfitting; ensuring the authenticity and reliability of the data is the cornerstone of model accuracy; S53. The initial curing process mapping model is repeatedly trained using historical production batch feature data and historical tag data. During the training process, an optimization algorithm is combined, and the weight parameters of the initial curing process mapping model are found based on the maximum number of training iterations to obtain the optimal solution. The optimal solution is used as the weight parameters of the initial curing process mapping model to obtain the pre-trained curing process mapping model. The characteristic data of the uncured conductive copper paste and the real-time required resistivity are input into the pre-trained curing process mapping model to obtain the initial curing process parameters. The above embodiment S53 includes the following steps: S531. Initialize the sparrow population location, with each individual sparrow representing a set of potential neural network weight parameters; introduce Tent chaotic mapping to initialize the population; enhance the diversity and ergodicity of the population; In specific implementation, the initialization of the above-mentioned Tent chaotic mapping is as follows: using the Tent mapping function Generates a chaotic sequence, in which μ Setting the value to 2 ensures a completely chaotic state. According to the mathematical definition of the Tent map, to ensure that the iterative value always falls within the interval [0,1] and the system exhibits chaos, the parameter... μ Must satisfy 1< μ ≤2; when μ When the maximum value is 2, the Lyapunov exponent of the system reaches its maximum, exhibiting the strongest chaotic characteristics. The resulting sequence is most evenly distributed and ergodic within the [0,1] interval, minimizing the initial population clustering in a few regions of the solution space, thus significantly improving the global search capability and convergence efficiency of subsequent optimization algorithms. First, an initial value is randomly generated. z 0 The process iterates to generate a set of uniformly distributed chaotic variable sequences, and then maps this sequence to the value space of the neural network weight parameters (such as [-1, 1]) through a linear transformation, which serves as the initial position of the sparrow population. Compared with traditional random initialization, Tent chaotic mapping can make the initial population cover the solution space more evenly, effectively avoiding the algorithm from getting stuck in local optima in the early stage. S532. Calculate the fitness value of each individual sparrow, where the fitness value is the reciprocal of the mean square error between the model's predicted curing process parameters and historical tag data; filter the individual sparrows based on their fitness values ​​to obtain discoverer sparrows and follower sparrows; specifically, the top 20% of individuals with the best fitness values ​​are selected as discoverers, responsible for global search, and the rest are selected as followers. S533. An adaptive Cauchy-Gaussian mutation strategy is introduced to update the position of the discoverer sparrow in order to escape local optima; at the same time, the sine-cosine algorithm (SCA) is used to perturb and update the position of the sparrow that joins, thereby improving the global search capability; and the updated sparrow population position is obtained. In specific implementation, the above embodiment S533 specifically refers to: the adaptive Cauchy-Gaussian mutation strategy, that is, introducing a mixed perturbation term to the position of the discoverer sparrow, which is composed of a linear combination of random numbers following a Cauchy distribution and random numbers following a Gaussian distribution, and its coefficient adaptively decays with the number of iterations; for the discoverer sparrow, a mutation operator is used. ;in, X new Indicates the updated location of the discoverer sparrow; X best This indicates the position of the globally optimal sparrow in the current population, i.e., the best solution found so far. δ 1. δ 2. Adaptive decay with the number of iterations, let the total number of iterations be... T maxThe current iteration number is t The attenuation formula is: δ ( t )= δ (0)*(1-(t / T); if set T max =100, δ 1. Initial value 0.5, δ 2. Initial value 0.3; when t When =50, δ 1 decays to 0.25, δ 2 decays to 0.15; at the end of the iteration (t=100), both decay to 0; this design allows the algorithm to explore the global with large perturbations in the early stage, and to perform a fine local search without perturbations when approaching the optimal solution in the later stage. Cauchy (0,1) represents the large step size characteristic of random number Cauchy variation that follows a standard Cauchy distribution (position parameter is 0, scale parameter is 1), which helps to escape local extrema; the small step size characteristic of Gaussian variation helps to refine local search. For the joiner, the sparrow, the SCA algorithm formula is used. Position perturbation is performed, utilizing the periodic oscillation characteristic of the sine function to guide participants in a spiral search near the global optimum, thereby enhancing the algorithm's global exploration capability; among which, X new1 Indicates the position of the newly added sparrow; X i Indicates the position of the current participant, the sparrow; X best This indicates the position of the globally optimal sparrow in the current population; r 1 is a random weighting factor (scalar) used to control the overall step size of the movement. In the actual SCA algorithm, r 1. Decrease with the number of iterations, the decreasing rule is as described above. δ The algorithm emphasizes exploration in the early stages and development in the later stages; r2 is a random number uniformly distributed in the range [0, 2π]; r3 is a random weight uniformly distributed in the range [0, 2]; the new position in the SCA algorithm formula is obtained by adding a step size modulated by a sine function and proportional to the distance to the optimal point to the current position; this mechanism allows the participants to search around the global optimal solution in an oscillating manner, which can be guided by the information of the optimal solution, and can maintain the diversity and globality of the search through the periodicity of the sine function and random parameters; S534. Iteratively update the population position until the maximum number of iterations is reached, then output the weight parameters corresponding to the globally optimal sparrow position. In specific implementation, the above embodiment S53 is as follows: The resistivity of the conductive copper paste in this application exhibits highly nonlinear and step-like characteristics as the curing process changes (i.e., the resistivity drops sharply near the microsphere disintegration temperature). Traditional BP neural networks or simple regression models are difficult to capture this abrupt change, and are prone to generating large prediction errors near the microsphere disintegration temperature. To this end, this application constructs a DBN model that integrates a multi-head attention mechanism and uses an improved sparrow search algorithm (ISSA) to optimize its weights. By introducing an attention mechanism, higher weights can be automatically assigned to the microsphere disintegration threshold temperature and curing temperature, allowing the model to focus on the nonlinear fitting of the phase transition region rather than being overwhelmed by a large amount of smooth linear data. Multiple local minima exist between resistivity and process parameters (e.g., similar resistivity may occur at high temperatures for short periods and at low temperatures for long periods), making traditional algorithms prone to getting trapped in local optima. ISSA significantly improves global optimization capabilities through Tent chaotic initialization and Cauchy-Gaussian mutation, ensuring the model can find the globally optimal weight configuration. As historical batch data is continuously trained, the model weights are dynamically adjusted by the ISSA algorithm, enabling the model to learn the impact of small fluctuations in raw materials (such as copper powder particle size distribution drift) on the final resistivity, thus achieving adaptive prediction. The above embodiments construct a deep learning model that integrates an attention mechanism and optimize the model parameters using an improved sparrow search algorithm, thereby realizing the mapping from the microscopic properties of materials to the macroscopic curing process. This data-driven approach overcomes the blindness of the traditional experience-based trial and error method and can intelligently recommend the best process based on the subtle differences in each batch of slurry (such as viscosity fluctuations and microsphere disintegration threshold temperature drift), significantly improving the efficiency and accuracy of process design. The above embodiment S6 includes the following steps: S61. Take the uncured conductive copper paste sample from S3, coat it onto the test substrate, place it in a rapid annealing furnace, and heat and cure it according to the initial curing process parameters; after cooling, use a four-probe tester to measure the actual resistivity. In specific implementation, the above embodiment S61 is as follows: the sample is a trace amount, accounting for 0.1% to 0.5% of the total mass of the conductive copper paste obtained in step S3. 0.5g of paste is taken and a strip with a length of 10cm and a width of 2mm is prepared on the PI film using a wire rod coater; it is placed in an infrared rapid annealing furnace, and the heating program is set to be consistent with the initial parameters output in S5; after curing, a smart four-probe resistance tester is used to measure at multiple points, and the average value is taken as the actual resistivity; S62. Construct the objective function; the objective function is the difference between the actual resistivity and the target resistivity; when the error between the actual resistivity and the target resistivity is ≥ the error threshold, set a search space that includes curing parameters such as heating rate, isothermal temperature, and isothermal time; based on the objective function and combined with the particle swarm optimization algorithm, find the optimal curing parameters. The optimal curing process is used as the final curing process to cure the remaining uncured conductive copper paste. The above embodiment S62 includes the following steps: S621. Construct a particle set, and set the size of the particle set to... g The particle set is then represented as: ,in, h i Represents the first in the set of particles i One particle; based on the search space and particle set of the solidified parameters, construct the initial position set of the particles. ,in, u i Represents the first particle in the set of particles i The initial position of each particle is used as the combination of heating rate, isothermal temperature, and isothermal time; the maximum number of optimization iterations is set. S622. Perform iterative update operations on the particles in the particle set, and each update is within a preset search space; and calculate the fitness value of the position of each particle in the particle set according to the objective function, update the position of each particle in the particle set according to the fitness value from high to low, and obtain the best individual particle in the particle set and the global best particle in each iteration. The update rule is as follows: if the actual resistivity is greater than the target resistivity, the algorithm adjusts the parameters along the gradient direction of increasing temperature or extending time to promote further disintegration of microspheres in S2 and sintering of copper powder in S1; if the actual resistivity is less than the target resistivity, the algorithm adjusts the parameters along the gradient direction of decreasing temperature or slowing down the heating rate to retain some of the occupancy effect of microspheres. During the update process, the positions of particles with fitness in the top 5% are retained, and the middle positions of the positions of particles with fitness in the top 46% (i.e., 45% of the positions enter the next round) are also retained. The remaining 50% are obtained through random particle movement. The best individual particle is the particle with the highest fitness value in each round of iteration, and the global best particle is the particle with the highest fitness value in all rounds of iteration. S623. Repeat S622. When the difference between the actual resistivity and the target resistivity is less than the error threshold or the maximum optimization parameter is reached, stop the iteration and use the global best particle as the optimal parameter combination. In specific implementation, the above embodiment S623 specifically involves: adopting an improved strategy of survival of the fittest combined with random perturbation, which preserves the genes of the superior population while introducing sufficient randomness to prevent premature convergence; for this application, this strategy can quickly find the curing parameter that matches the target resistivity among many process parameters, especially for nonlinear resistivity change regions, the optimization efficiency is 30% higher than that of traditional PSO; The above embodiments establish a closed-loop feedback correction mechanism through micro-trial and error and intelligent optimization algorithm; it can eliminate residual errors in the S5 model prediction and the influence of the actual environment (such as furnace temperature deviation); through the process of actual measurement and adjustment for further confirmation, it ensures that the curing process parameters finally output to the production line have been rigorously verified, thereby ensuring the reliability of product performance. To further verify the technical effect of the method described in this invention, the following examples and comparative examples were set up for performance testing. The performance testing method was as follows: resistivity was measured using a four-probe tester (such as RTS-9) under a standard environment of 25±1℃ and 50±5%RH on the cured coating sample. Oxidation resistance was evaluated by measuring the resistivity change rate (ΔR) after placing the sample in a constant temperature and humidity test chamber (85℃ / 85%RH) for 500 hours. Batch consistency was characterized by calculating the coefficient of variation (CV value) of the resistivity of 10 consecutive production batches of samples under the same formula. The test objects included: Example 1: Conductive copper paste prepared and controlled by the entire process of S1-S7 of this invention; the target resistivity was set to 1.0×10⁻⁶. −4 Ω*cm (low-resistance wire mode); Example 2: Conductive copper paste prepared and controlled using the entire process S1-S7 of this invention; Target resistivity set to 5.0×10 Ω*cm. −3 Ω*cm (high resistance heating mode); Comparative Example 1: Ordinary physically mixed copper paste (without S1 coating, without S2 microspheres) was used and cured using a fixed process (160℃ / 30min); Comparative Example 2: The paste of this invention was used, but the intelligent control of S4-S6 was not implemented, and the empirical process (160℃ / 30min) was directly used for curing. The results are shown in Table 1 below: Test Project Example 1 Example 2 Comparative Example 1 Comparative Example 2 Target resistivity (Ω*cm) <![CDATA[1.0*10 -4 ]]> <![CDATA[5.0*10 -3 ]]> Pursue the lowest <![CDATA[1.0*10 -4 ]]> Measured resistivity (Ω*cm) <![CDATA[0.98*10 -4 ]]> <![CDATA[5.1*10 -3 ]]> <![CDATA[2.5*10 -3 ]]> <![CDATA[1.4*10 -4 ]]> Deviation -2.0% -2.0% none 12.0% Antioxidant properties (85℃ / 85%RH 500h) ΔR < 10% ΔR < 12% ΔR > 200% (failure) ΔR < 10% Batch consistency (CV value) 3.5% 4.1% 15.6% 12.8% It can be seen that the measured resistivity of Examples 1 and 2 is highly consistent with the target value, with the deviation controlled within ±2%, proving the effectiveness of the intelligent control strategy of S4-S6 in this invention. In contrast, although Comparative Example 2 used the same materials, due to the lack of process fine-tuning for the current batch, the resistivity deviation reached 12%, which could not meet the requirements of precision electronics. The conductivity of Example 1 is much better than that of Comparative Example 1 (ordinary copper paste), and it has excellent anti-oxidation performance (ΔR<10%), which is attributed to the dual protection mechanism of PDA coating in S1 and silver layer release by microspheres in S2. From the coefficient of variation (CV value), the batch stability of the method of this invention (CV<5%) is significantly better than that of the traditional method (CV>12%), indicating that the intelligent feedback system can effectively offset the impact of raw material fluctuations and is suitable for large-scale industrial production.

[0018] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0019] The preferred embodiments of the invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A method for preparing conductive copper paste with adaptive resistivity control, characterized in that, Includes the following steps: S1. In a multi-level pH gradient buffer system, micron-sized flake copper powder and nano-sized spherical copper powder are cascaded for surface modification and in-situ polydopamine coating is carried out to prepare an antioxidant copper powder premix. S2. Polymer microspheres are synthesized using microfluidics or emulsion polymerization as the core. A nanoscale conductive metal shell is constructed on the surface of the core through surface sensitization and activation treatment, as well as chemical plating technology, to obtain thermally responsive conductive microspheres. S3. The antioxidant copper powder premix in S1 is combined with the thermally responsive conductive microspheres in S2 and degassed and premixed with an organic carrier to obtain a premix; the premix is ​​subjected to high-shear three-roll milling to obtain an uncured conductive copper paste. S4. Obtain the interface parameters of the antioxidant copper powder premix, the thermodynamic phase transition parameters of the thermally responsive conductive microspheres, and the rheological viscosity data of the uncured conductive copper paste in S3 to obtain the characteristic data of the uncured conductive copper paste; obtain the target resistivity. S5. Input the characteristic data of the uncured conductive copper paste and the real-time required resistivity into the pre-trained curing process mapping model to obtain the initial curing process parameters. S6. Take a sample of the uncured conductive copper paste from S3, cure it based on the initial curing process parameters, and measure the actual resistivity. Determine whether the error between the actual resistivity and the target resistivity is within the error threshold. If so, use the initial curing process parameters as the final curing process to cure the remaining uncured conductive copper paste. Otherwise, an optimal curing process is found through an optimization algorithm; the optimal curing process is then used as the final curing process to cure the remaining uncured conductive copper paste.

2. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 1, characterized in that, S1 includes the following steps: S11. Construct a modified buffer system with a multi-level pH gradient; the modified buffer system with a multi-level pH gradient includes a primary washing solution, an intermediate activation solution, and a final coating solution; the primary washing solution is a mixture of dilute H2SO4 and ethanol; the intermediate activation solution is a weakly acidic solution containing potassium fluorinated titanate; the final coating solution is composed of tris(hydroxymethyl)aminomethane, dopamine hydrochloride, and ammonium persulfate oxidation initiator; S12. Mix the flake copper powder and spherical copper powder, and pass them through the primary cleaning solution and the intermediate activation solution in sequence to obtain the treated copper powder. S13. The copper powder treated in S12 is suspended in the final coating solution in S11; the intensity change of characteristic peaks on the surface of the copper powder is monitored to characterize the growth thickness of the polydopamine film. S14. When the growth thickness of the polydopamine film reaches the preset film thickness threshold, a terminator is immediately added to quench the polymerization reaction. After centrifugation and vacuum drying, an antioxidant copper powder premix is ​​obtained.

3. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 1, characterized in that, S2 includes the following steps: S21. Using polyethylene glycol, polycaprolactone, or low-melting-point polystyrene microspheres as the substrate, prepare monodisperse core microspheres; determine the onset temperature of their melting endothermic peak, and record it as the disintegration threshold temperature; S22. Sensitize the nucleospheres obtained in S21 to adsorb Sn²⁺ ions; then generate Pd in ​​situ via a redox reaction. 0 Catalyzing the crystal nucleus yields the activated nucleus; S23. Prepare a chemical silver plating solution, wherein the chemical silver plating solution includes silver nitrate as the main salt, ethylenediamine complexing agent, potassium sodium tartrate reducing agent, and polyvinylpyrrolidone dispersant; under ultrasonic assistance, introduce the activated nucleus from S22 into the plating solution, control the reaction temperature below the disintegration threshold temperature and pH, so that silver ions are reduced and deposited on the surface of the activated nucleus to form a dense silver shell layer, thereby obtaining thermally responsive conductive microspheres.

4. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 3, characterized in that, S23 includes the following steps: S231. Establish the shell growth rate equation; S232. Based on the shell growth rate equation, the reduction rate of free silver ions is controlled by adding reducing agent in steps; the consumption rate of silver ions in the plating solution is monitored in real time; when the silver ion conversion rate reaches more than 95% and the absorbance of the mixed solution no longer changes, the shell construction is determined to be complete.

5. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 1, characterized in that, S3 includes the following steps: S31. Bisphenol A type epoxy resin, phenolic resin, solvent, leveling agent and thixotropic agent are mixed in a mass ratio and stirred to dissolve, forming a homogeneous organic carrier. S32. The antioxidant copper powder premix prepared in S1, the thermally responsive conductive microspheres prepared in S2, the homogeneous organic carrier prepared in S31, and the latent curing agent are put into a planetary vacuum mixer for degassing and premixing to obtain the premix. S33. Feed the premix from S32 into a three-roll mill; set the speed ratio of the feed roller, middle roller, and discharge roller; and perform grinding using a graded grinding strategy. S34. During the grinding process in S33, an infrared thermal imager is used to monitor the roller surface temperature, and a cooling water circulation system is used to control the slurry temperature to always be below the disintegration threshold temperature of the thermally responsive microspheres. T c Subtracting 20°C, the final product is an uncured conductive copper paste.

6. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 1, characterized in that, S5 includes the following steps: S51. Based on the DBN neural network model, construct an initial curing process mapping model; set the maximum number of training iterations and initial weight parameters for the initial curing process mapping model; S52. Collect the interface parameters of the antioxidant copper powder premix, the thermodynamic phase change parameters of the thermally responsive conductive microspheres, and the rheological viscosity data of the uncured conductive copper paste in S3, as well as the corresponding historical required resistivity, from the historical production batch data to obtain historical production batch characteristic data; obtain the curing process parameters corresponding to the historical required resistivity; use the curing process parameters corresponding to the historical required resistivity as tags for the historical production batch characteristic data to obtain historical tag data. S53. The initial curing process mapping model is repeatedly trained using historical production batch feature data and historical tag data. During the training process, an optimization algorithm is combined, and the weight parameters of the initial curing process mapping model are found based on the maximum number of training iterations to obtain the optimal solution. The optimal solution is used as the weight parameters of the initial curing process mapping model to obtain the pre-trained curing process mapping model. The characteristic data of the uncured conductive copper paste and the required resistivity in real time are input into the pre-trained curing process mapping model to obtain the initial curing process parameters.

7. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 6, characterized in that, S53 includes the following steps: S531. Initialize the sparrow population location, where each individual sparrow represents a set of potential neural network weight parameters; introduce Tent chaotic mapping to initialize the population; S532. Calculate the fitness value of each sparrow individual, wherein the fitness value is the reciprocal of the mean square error between the model's predicted curing process parameters and historical tag data; filter the sparrow individuals based on their fitness values ​​to obtain discoverer sparrows and follower sparrows; S533. An adaptive Cauchy-Gaussian mutation strategy is introduced to update the position of the discoverer sparrow; at the same time, the sine and cosine algorithms are used to perturb and update the position of the sparrow that joins; the updated sparrow population position is obtained. S534. Iteratively update the population position until the maximum number of iterations is reached, then output the weight parameters corresponding to the globally optimal sparrow position.

8. The method for preparing conductive copper paste with adaptive resistivity control as described in claim 1, characterized in that, S6 includes the following steps: S61. Take the uncured conductive copper paste sample from S3, coat it onto the test substrate, place it in a rapid annealing furnace, and heat and cure it according to the initial curing process parameters; after cooling, use a four-probe tester to measure the actual resistivity. S62. Construct the objective function; the objective function is the difference between the actual resistivity and the target resistivity; when the error between the actual resistivity and the target resistivity is ≥ the error threshold, set a search space that includes curing parameters such as heating rate, isothermal temperature, and isothermal time; based on the objective function and combined with the particle swarm optimization algorithm, find the optimal curing parameters. The optimal curing process is used as the final curing process to cure the remaining uncured conductive copper paste.

9. The method for preparing conductive copper paste with adaptive resistivity control according to claim 1, characterized in that, S62 includes the following steps: S621. Construct a particle set; Based on the search space of the solidified parameters and the particle set, construct an initial particle position set, and use the position of each particle in the initial particle position set as a combination of heating rate, isothermal temperature, and isothermal time; set the maximum optimization parameter; S622. Perform iterative update operations on the particles in the particle set, and each update is within a preset search space; and calculate the fitness value of the position of each particle in the particle set according to the objective function, update the position of each particle in the particle set, and obtain the best individual particle and the global best particle in the particle set in each iteration. The update rule is that if the actual resistivity is greater than or equal to the target resistivity, the algorithm adjusts the parameters along the gradient direction of increasing temperature or increasing time; if the actual resistivity is less than the target resistivity, the algorithm adjusts the parameters along the gradient direction of decreasing temperature or decreasing heating rate. S623. Repeat S622. When the difference between the actual resistivity and the target resistivity is less than the error threshold or the maximum optimization parameter is reached, stop the iteration and use the global best particle as the optimal parameter combination.

10. A conductive copper paste, characterized in that, Conductive copper paste is prepared using the resistivity adaptive control method described in any one of claims 1-9.

Images