A method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates

Abstract

This invention provides a method for regulating the eutectic reaction mechanism of DBC (Dielectric-Ceramic Composite) substrates, belonging to the field of DBC substrate fabrication technology in power electronic packaging. Addressing the problems of interfacial brittleness, intermetallic compound aggregation, low bonding strength, and poor thermal cycling reliability caused by thermal expansion coefficient mismatch and stress concentration in traditional DBC processes, this invention proposes a three-pronged approach to actively regulate the stress field: interface microstructure guidance, gradient thermal field control, and dynamic mechanical constraint. Through ten steps including substrate pretreatment, microstructure patterning, gradient thermal field construction, dynamic pressure application, in-situ stress monitoring, and segmented cooling, precise spatiotemporal control of the interface stress field is achieved, promoting the formation of a dense and uniform interface layer. Experiments show that the DBC substrate prepared by this invention exhibits over 30% improved interfacial bonding strength and over 2 times improved thermal cycling life. It also demonstrates strong process compatibility and excellent controllability, making it suitable for demanding applications such as high-end power electronic devices.

Description

Technical Field

[0001] This invention relates to the field of DBC metal-ceramic substrate preparation technology in the field of electronic packaging, and more specifically, to a method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates. Background Technology

[0002] DBC (Direct Batch Packaging) technology, a core technology in the field of power electronic packaging, utilizes the oxygen diffusion between copper foil and ceramic substrate under high-temperature conditions to form metallurgical bonding intermediate phases such as Cu–O–Al or Cu–O–N, achieving a reliable connection between the metal and ceramic. This process, combining the high insulation and excellent thermal conductivity of ceramic substrates with the high electrical conductivity and good heat dissipation of copper foil, is widely used in high-end electronic devices such as IGBT modules and power semiconductor devices.

[0003] However, the eutectic reaction stage of the DBC process faces unavoidable technical bottlenecks: the significant difference in thermal expansion coefficients between copper and ceramic materials, coupled with the anisotropic nature of interfacial energy, leads to localized stress concentration in the interfacial region during the reaction. This stress concentration directly induces the non-uniform aggregation and coarsening of brittle intermetallic compounds, forming a discontinuous and loose interfacial layer structure. These problems directly result in two core defects in traditional DBC substrates: first, insufficient interfacial bonding strength, making them prone to peeling failure under mechanical vibration or thermal shock; second, poor thermal cycling reliability, where repeated accumulation of interfacial stress in alternating high and low temperature environments (such as -40℃ to 150℃) accelerates crack propagation and shortens device lifespan.

[0004] To address the aforementioned issues, existing technologies primarily focus on optimization in three areas: firstly, atmosphere control, which suppresses excessive oxidation by adjusting the oxygen content in the reaction atmosphere; secondly, optimizing the heating rate, employing a slow heating method to reduce thermal stress; and thirdly, surface pretreatment, improving interfacial wettability through grinding, cleaning, and other means. However, these solutions all have inherent limitations: they passively adjust only single factors in the process, lacking systematic and proactive control over the interfacial stress field. This fails to fundamentally solve the problem of harmful phase precipitation and coarsening caused by stress concentration, making it difficult to achieve a breakthrough in interfacial bonding performance and reliability. Therefore, we propose an improvement, suggesting a method for regulating the eutectic reaction mechanism of DBC cermet substrates. Summary of the Invention

[0005] The purpose of this invention is to address the problems identified in the existing background technology. To achieve the above-mentioned objective, this invention provides the following technical solution: a method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates, comprising the following steps:

[0006] Step 1: Substrate pretreatment: The ceramic substrate (preferably Al2O3 or AlN ceramic) and copper foil (purity ≥99.9%) are subjected to surface cleaning and activation treatment. Surface oxides (CuO, Al2O3), grease and adsorbed impurities are removed by acid washing or plasma cleaning. After treatment, the surface roughness is controlled within Ra=0.1–0.5μm. This roughness range can increase the solid-liquid contact area and promote liquid phase spreading and atomic adsorption.

[0007] Step 2: Interface microstructure patterning: Laser micromachining is used on the ceramic contact surface to prepare periodic micron-scale grooves or lattice structures. The microstructure parameters need to be matched with the viscosity of the eutectic liquid phase, interfacial tension and reaction rate.

[0008] Step 3: Stacking and pre-pressing: The processed copper foil and ceramic substrate are stacked in a predetermined order and placed in a high-temperature resistant graphite mold, and an initial pre-pressure of 0.1MPa–0.5MPa is applied;

[0009] Step 4: Construct a gradient thermal environment: Place the assembly into a controlled atmosphere furnace, introduce an oxygen-containing inert gas (oxygen content 5–15 vol%), set an asymmetric heating program, so that the heating rate of the copper side is 5–10°C / min higher than that of the ceramic side, forming a temperature gradient from copper to ceramic. The highest temperature on the ceramic side shall not exceed 95% of the highest temperature on the copper side to avoid the deterioration of ceramic performance.

[0010] Step 5: Apply dynamic mechanical constraints: Before heating to the eutectic reaction temperature range (1065–1100°C), start the servo hydraulic system to apply a programmable dynamic axial pressure of 0.5MPa–3MPa to the laminate, with the pressure waveform being constant, pulsed, or sinusoidally modulated (frequency 0.01–1Hz).

[0011] Step 6: Stress field control during the eutectic reaction stage: After the formation of the eutectic liquid phase, maintain the gradient thermal field and simultaneously adjust the mechanical constraint parameters to keep the interface region under a controllable compressive stress state of 0.3–1.5 MPa.

[0012] Step 7: In-situ stress monitoring and feedback: Real-time monitoring of strain evolution in the interface region using embedded high-temperature strain gauges (operating temperature from room temperature to 1200°C) and DIC technology;

[0013] Step 8: Stress relief during cooling process: After the eutectic reaction is completed, a segmented controlled cooling strategy is adopted: first, slowly cool to 800°C at 1–5°C / min to maintain the gradient thermal field to continue the directional stress; then rapidly cool to below 400°C at 20–50°C / min.

[0014] Step 9: Post-processing and interface stabilization: Vacuum annealing of the DBC sample (vacuum degree less than 0.001 Pa);

[0015] Step 10: Performance evaluation and process optimization: The morphology, orientation and mechanical properties of the interface compound are evaluated through interfacial bonding strength, thermal cycling test and SEM / EBSD microscopic characterization.

[0016] As a preferred technical solution of the present invention, the acid washing in step 1 uses a dilute sulfuric acid or dilute hydrochloric acid solution with a concentration of 5–15 vol%, the cleaning temperature is controlled at 25–50°C, and the cleaning time is 3–10 min; the plasma cleaning uses a mixed gas of argon and oxygen, wherein the oxygen content is 5–15 vol%, the plasma power is 100–300 W, and the cleaning time is 5–20 min.

[0017] As a preferred technical solution of the present invention, the cross-sectional shape of the periodic micron-level grooves in step 2 is V-shaped, U-shaped or trapezoidal, the groove depth is 5–30 μm, the groove width is 10–50 μm, and the groove spacing is 20–100 μm; the lattice structure is cylindrical, square prism or regular hexagonal prism, the lattice spacing is 30–80 μm, the ratio of lattice height to cross-sectional feature size is 0.5–2.0, and the area coverage of the microstructure pattern is 20–60%.

[0018] As a preferred technical solution of the present invention, the inner wall of the high-temperature resistant graphite mold in step 3 is coated with a boron nitride release layer with a thickness of 5–20 μm. During the pre-pressing process, a step-by-step pressurization method is adopted, first applying a pressure of 0.1–0.2 MPa and holding it for 5–10 min, then raising it to the set pre-pressure and holding it for 10–20 min, to ensure that the interface gap is ≤5 μm.

[0019] As a preferred technical solution of the present invention, the flow rate of the oxygen-containing inert gas in step 4 is 50–200 sccm, the pressure inside the furnace is maintained at 0.1–0.3 MPa, the heating rate on the copper side is 5–10°C / min higher than that on the ceramic side, and the temperature gradient is linearly distributed along the direction perpendicular to the interface.

[0020] As a preferred technical solution of the present invention, the pressure control accuracy of the servo hydraulic system in step 5 is ±0.05MPa, the response time is ≤50ms; the peak value of the pulse pressure is 1.5–3MPa, the valley value is 0.5–1MPa, and the pulse width is 1–10s; the amplitude of the sinusoidal modulated pressure is 0.3–1MPa, and the DC bias is 0.8–2MPa.

[0021] As a preferred technical solution of the present invention, the heat preservation time of the eutectic reaction stage in step 6 is 10–30 min; the compressive stress in the interface region is maintained at 0.3–1.5 MPa; and the liquid phase spreading rate is guided by the microstructure pattern to be 0.1–0.5 mm / s, so that the thickness uniformity error of the interface compound layer is ≤10%.

[0022] As a preferred technical solution of the present invention, the strain measurement accuracy of the embedded high-temperature strain gauge in step 7 is ±1με; the sampling frequency of DIC technology is 10–50Hz, and the displacement measurement accuracy is ±0.1μm; the response period of stress field adjustment is ≤1s, and the rapid response can promptly offset the abnormal fluctuations of interface stress and avoid interface damage caused by stress accumulation.

[0023] As a preferred technical solution of the present invention, the rapid cooling in step 8 adopts forced air cooling or inert gas jet cooling, with a cooling rate of 20–50°C / min. After cooling to below 400°C, it is then cooled to room temperature at a rate of 5–10°C / min. During the entire cooling process, the interface temperature difference is ≤30°C and the residual stress is ≤150MPa.

[0024] As a preferred technical solution of the present invention, the vacuum degree of vacuum annealing in step 9 is less than 0.001 Pa, and a step heating method is adopted during the annealing process, which raises the temperature to the set temperature at 5–10°C / min and then holds it at that temperature; in step 10, the interfacial bonding strength is determined by shear test method, the test rate is 0.5–1 mm / min, the interfacial peeling rate after thermal cycling test is ≤5%, and SEM / EBSD characterization shows that the interfacial compound phase is mainly Cu3Al, the grain size is 0.5–5 μm and the orientation consistency is ≥80%.

[0025] Compared with existing technologies, the beneficial effects of this invention are as follows: 1. Significantly improved interfacial bonding strength: Through the synergistic effect of dynamic mechanical constraints and gradient thermal fields, interfacial stress concentration is effectively eliminated, the aggregation and precipitation of brittle intermetallic compounds are suppressed, and the formation of a continuous, uniform, and dense Cu–O–Al / Cu–O–N intermediate phase layer is promoted. Experimental data show that the interfacial peel strength of the DBC substrate prepared by this method is more than 30% higher than that of the traditional process, and it can stably withstand higher mechanical loads and thermal shocks, significantly reducing the failure risk during device packaging.

[0026] 2. Significantly Enhanced Thermal Cycling Reliability: By employing a segmented controlled cooling strategy and vacuum annealing post-treatment, residual stress is eliminated to the maximum extent. Simultaneously, interface microstructure design guides liquid phase spreading and stress release, improving the thermomechanical compatibility of the interface layer. Cycling tests at -40℃ to 150℃ have verified that samples prepared using this invention maintain excellent interface integrity after 1000 thermal cycles, with no crack initiation or delamination. The thermal cycle life is more than twice that of traditional processes, meeting the application requirements of long lifespan and high reliability for high-end power electronic devices.

[0027] 3. Significantly Enhanced Process Stability and Controllability: Through an embedded high-temperature strain monitoring and closed-loop feedback system, precise spatiotemporal control of the stress field is achieved, allowing dynamic optimization of process parameters based on different ceramic substrates (Al2O3 / AlN) and copper foil specifications. Simultaneously, the established performance evaluation-parameter optimization closed-loop database significantly broadens the process window, effectively reducing product consistency issues caused by raw material batch differences and equipment fluctuations, resulting in a production qualification rate increase of over 15%.

[0028] 4. Excellent scalability for various application scenarios: This invention does not require disruptive modifications to existing DBC production equipment. Upgrading the technology can be achieved simply by adding a gradient thermal field module, a dynamic pressure control system, and a micromachining preprocessing unit. Its process parameters can be flexibly adapted to ceramic substrates and copper foils of different thicknesses and specifications. It is not only suitable for traditional power electronics packaging but can also be extended to high-end applications with more stringent reliability requirements, such as power modules for new energy vehicles and aerospace electronic devices, demonstrating broad industrialization prospects.

[0029] 5. Synergistic Optimization of Microstructure and Performance: SEM / EBSD microscopic characterization verified that the intermetallic compound grain size uniformity in the interface layer prepared by this method was improved by 40%, and the phase structure orientation consistency was significantly improved, achieving synergistic optimization of the interface microstructure and macroscopic mechanical and thermal properties. This structural optimization not only improved the interfacial bonding strength and thermal conductivity but also reduced the interfacial contact resistance, providing core support for efficient heat dissipation and low-loss operation of power devices. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are specific implementations of the present invention and are not limited to all embodiments.

[0031] Therefore, the following detailed description of embodiments of the present invention is not intended to limit the scope of the claimed invention, but merely illustrates some embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0032] It should be noted that, unless otherwise specified, the embodiments and features and technical solutions in the present invention can be combined with each other.

[0033] Example 1: A method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates, comprising the following steps: Step 1: Substrate pretreatment: The ceramic substrate (preferably Al2O3 or AlN ceramic) and copper foil (purity ≥99.9%) are subjected to surface cleaning and activation treatment. Surface oxides (CuO, Al2O3), grease, and adsorbed impurities are removed by acid washing or plasma cleaning. The surface passivation layer is removed by etching to expose fresh metal / ceramic active sites, reduce the interfacial atomic diffusion barrier, and create thermodynamic conditions for the interfacial reaction between Cu and ceramic components in the eutectic reaction. The surface roughness after treatment is controlled at Ra=0.1–0.5μm. This roughness range can increase the solid-liquid contact area and promote liquid phase spreading and atomic adsorption.

[0034] Step 2: Interface Microstructure Patterning: Periodic micron-scale grooves or lattice structures are fabricated on the ceramic contact surface using laser micromachining. The capillary effect of the microstructure guides the directional flow of the eutectic liquid phase. At the same time, the geometric constraints of the microstructure disperse interfacial thermal stress and residual stress, suppressing crack initiation caused by the mismatch of thermal expansion coefficients between ceramic and copper. In addition, the microstructure can serve as heterogeneous nucleation sites to regulate the nucleation rate and growth direction of the interfacial compound phase. The microstructure parameters need to be matched with the viscosity of the eutectic liquid phase, interfacial tension, and reaction rate.

[0035] Step 3: Stacking and pre-pressing: The processed copper foil and ceramic substrate are stacked in a predetermined order and placed in a high-temperature resistant graphite mold. An initial pre-pressure of 0.1MPa–0.5MPa is applied to eliminate interfacial gaps (controlled ≤5μm), reduce atomic diffusion distance, promote close contact between copper foil and ceramic substrate, and provide kinetic guarantee for interfacial mass transfer in eutectic reaction.

[0036] Step 4: Constructing a gradient thermal environment: Place the assembly in a controlled atmosphere furnace and introduce an oxygen-containing inert gas (oxygen content 5–15 vol%). The mechanism of oxygen is to moderately oxidize the copper foil surface to form a thin Cu2O layer, reduce the Cu-Al eutectic reaction temperature (Cu-Cu2O eutectic temperature 1065°C), and at the same time suppress the interface bonding failure caused by excessive oxidation of the copper foil. Set an asymmetric heating program so that the heating rate of the copper side is 5–10°C / min higher than that of the ceramic side, forming a temperature gradient from copper to ceramic. Use the temperature gradient to drive a directional thermal stress field and atomic diffusion flow, promote the migration of Cu atoms to the ceramic interface, and at the same time suppress grain growth on the ceramic side caused by high temperature, maintain the mechanical stability of the ceramic substrate, and ensure that the highest temperature on the ceramic side does not exceed 95% of the highest temperature on the copper side to avoid the deterioration of ceramic performance.

[0037] Step 5: Apply dynamic mechanical constraints: Before heating to the eutectic reaction temperature range (1065–1100°C), activate the servo hydraulic system to apply a programmable dynamic axial pressure of 0.5MPa–3MPa to the laminate. The pressure waveform can be constant, pulsed, or sinusoidally modulated (frequency 0.01–1Hz). The dynamic pressure controls the interfacial liquid phase thickness through the force-chemical coupling effect: constant pressure can maintain the stability of the liquid phase layer, while pulsed / sinusoidal pressure promotes liquid phase refluxing and composition homogenization through periodic loading, alleviating abnormal compound phase growth caused by solute enrichment. The pressure parameters are adjusted in real time based on the interfacial stress detection results to achieve synergy between stress relaxation and liquid phase control.

[0038] Step 6: Stress field control during the eutectic reaction stage: After the formation of the eutectic liquid phase, maintain the gradient thermal field and simultaneously adjust the mechanical constraint parameters to keep the interface region under controllable compressive stress of 0.3–1.5 MPa. The compressive stress can reduce the critical nucleation work of the eutectic reaction, promote the nucleation and ordered growth of target compound phases such as Cu3Al, and inhibit the formation of brittle phases (such as Al2Cu3). Combined with the guiding effect of the microstructure pattern, the liquid phase is spread along a specific path (rate 0.1–0.5 mm / s) to ensure that the uniformity error of the interface compound layer thickness is ≤10%, and to avoid excessive thickness or defects in the compound phase caused by local liquid phase aggregation.

[0039] Step 7: In-situ stress monitoring and feedback: Using embedded high-temperature strain gauges (operating temperature from room temperature to 1200°C) and DIC technology, the strain evolution of the interface region is monitored in real time. During the eutectic reaction, the formation and growth of the interface compound phase will cause volume changes, which, combined with the difference in thermal expansion coefficients, will generate dynamic stress. By inverting the stress field distribution through strain signals, the pressure and temperature gradients are dynamically adjusted to achieve closed-loop control of reaction-stress-structure, thus avoiding interface delamination caused by stress concentration.

[0040] Step 8: Stress Relief During Cooling: After the eutectic reaction is completed, a segmented controlled cooling strategy is adopted: First, the temperature is slowly cooled to 800°C at 1–5°C / min to maintain a gradient thermal field and perpetuate directional stress. The mechanism is that the slow cooling stage can promote the diffusion transformation of the interfacial compound phase and grain refinement, and the directional stress can guide the compound phase to grow along a preferred orientation, reducing internal stress. Then, the temperature is rapidly cooled to below 400°C at 20–50°C / min. The mechanism is that rapid cooling can inhibit the precipitation and growth of brittle phases. Below 400°C is the region where the difference in thermal expansion coefficients between Cu and ceramic is gradual. Subsequent slow cooling to room temperature can further release residual stress (controlled to ≤150MPa).

[0041] Step 9: Post-processing and interface stabilization: Vacuum annealing (vacuum degree less than 0.001 Pa) is performed on the DBC sample. The vacuum environment can avoid interface oxidation. The annealing process eliminates residual stress through atomic thermal activation diffusion, promotes the ordering and densification of the interface compound phase (mainly Cu3Al), reduces grain boundary defects and porosity, and improves the interface bonding strength and thermal stability.

[0042] Step 10: Performance evaluation and process optimization: The morphology, orientation and mechanical properties of the interfacial compounds are evaluated through interfacial bonding strength, thermal cycling test and SEM / EBSD microscopic characterization. Based on the characterization results, the microstructure parameters, pressure waveform and thermal field gradient are optimized to form a closed-loop process database. Its core mechanism is to establish a correlation model of process parameters-microstructure-reaction mechanism-macroscopic performance to achieve precise control of eutectic reaction.

[0043] The pickling in step 1 uses a 5–15 vol% dilute sulfuric acid or dilute hydrochloric acid solution, with the cleaning temperature controlled at 25–50°C and the cleaning time at 3–10 min. The acid solution removes oxides such as CuO and Al2O3 through chemical dissolution. The control of temperature and concentration can balance the etching rate and surface roughness, avoiding surface defects caused by excessive corrosion. The plasma cleaning uses a mixture of argon and oxygen, with oxygen accounting for 5–15 vol%. The plasma power is 100–300 W, and the cleaning time is 5–20 min. The physical sputtering effect of argon plasma removes surface impurities, while the chemical activity of oxygen plasma oxidizes and decomposes grease and activates the surface. The mixed gas ratio and power parameters can control the density and roughness of surface active sites.

[0044] The periodic micron-sized grooves described in step 2 have a V-shaped, U-shaped, or trapezoidal cross-sectional shape, a groove depth of 5–30 μm, a groove width of 10–50 μm, and a groove spacing of 20–100 μm. The lattice structure is cylindrical, square prism, or hexagonal prism, with a lattice spacing of 30–80 μm. The ratio of lattice height to the cross-sectional feature size is 0.5–2.0, and the area coverage of the microstructure pattern is 20–60%. The groove depth / lattice height needs to match the thickness of the eutectic liquid phase layer, typically 5–20 μm, to ensure that capillary action effectively guides the liquid phase flow. The groove width / lattice spacing and area coverage are adjusted to control the uniformity of interfacial stress distribution, avoiding liquid phase retention due to excessively dense microstructures or insufficient stress dispersion due to excessively sparse microstructures.

[0045] The inner wall of the high-temperature graphite mold described in step 3 is coated with a boron nitride release layer with a thickness of 5–20 μm. Boron nitride has the characteristics of high temperature resistance and low coefficient of friction, which can prevent the eutectic liquid phase from sticking to the mold and avoid the diffusion of mold components to contaminate the interface. In the pre-pressing process, a step-by-step pressurization method is adopted. First, a pressure of 0.1–0.2 MPa is applied and held for 5–10 min, and then the pressure is increased to the set pre-pressure and held for 10–20 min to ensure that the interface gap is ≤5 μm. Step-by-step pressurization can gradually remove the interface air, avoid the interface porosity caused by the expansion of gas in the gap at high temperature, and reduce the local deformation of copper foil and ceramic substrate.

[0046] In step 4, the flow rate of the oxygen-containing inert gas is 50–200 sccm, and the pressure inside the furnace is maintained at 0.1–0.3 MPa. The gas flow rate and pressure can regulate the oxygen partial pressure inside the furnace, ensuring the formation of a thin Cu2O layer while inhibiting excessive oxidation. The slightly positive pressure environment of 0.1–0.3 MPa can reduce the intrusion of external impurities and ensure the purity of the eutectic reaction. The heating rate on the copper side is 5–10°C / min higher than that on the ceramic side, and the temperature gradient is linearly distributed along the direction perpendicular to the interface. The linear temperature gradient can form a continuous driving force for atomic diffusion, avoiding interfacial thermal shock and stress concentration caused by sudden temperature changes.

[0047] The pressure control accuracy of the servo hydraulic system described in step 5 is ±0.05MPa, and the response time is ≤50ms. The high-precision and fast-response pressure control can match the dynamic changes of the eutectic reaction in real time (such as the liquid phase generation rate and compound phase growth stress), avoiding interface defects caused by pressure hysteresis. The peak value of the pulse pressure is 1.5–3MPa, the valley value is 0.5–1MPa, and the pulse width is 1–10s. The amplitude of the sinusoidal modulated pressure is 0.3–1MPa, and the DC bias is 0.8–2MPa. The peak value of the pulse pressure is used to promote liquid phase flow, and the valley value is used for stress relaxation. The sinusoidal modulated pressure suppresses the preferential coarsening of the compound phase through continuous stress fluctuations, thereby improving the uniformity of the interface structure.

[0048] The holding time for the eutectic reaction stage described in step 6 is 10–30 min. The holding time needs to match the interfacial atomic diffusion and compound phase growth cycle to ensure that Cu and ceramic components react fully to form a continuous and dense Cu3Al compound layer, avoiding weak bonding due to insufficient holding or excessively thick and brittle compound phase due to excessive holding. The compressive stress in the interfacial region is maintained at 0.3–1.5 MPa. This pressure range can balance the fluidity of the eutectic liquid phase and the growth stability of the compound phase. Too low a pressure cannot effectively suppress cracks, and too high a pressure will cause liquid phase extrusion. The liquid phase spreading rate is guided by the microstructure pattern to be 0.1–0.5 mm / s, so that the thickness uniformity error of the interfacial compound layer is ≤10%. The spreading rate is related to the liquid phase viscosity, interfacial tension and microstructure capillary force. A uniform spreading rate can avoid stress concentration caused by local compound phase thickness differences.

[0049] The strain measurement accuracy of the embedded high-temperature strain gauge mentioned in step 7 is ±1με; the sampling frequency of DIC technology is 10–50Hz, and the displacement measurement accuracy is ±0.1μm. High-precision, high-frequency strain / displacement monitoring can capture the dynamic stress evolution of the interface during the eutectic reaction (such as liquid phase solidification shrinkage and compound phase transformation stress), providing accurate feedback for real-time adjustment of pressure and temperature gradients; the response period of stress field adjustment is ≤1s, and the rapid response can promptly offset abnormal fluctuations in interface stress and avoid interface damage caused by stress accumulation.

[0050] The rapid cooling described in step 8 employs forced air cooling or inert gas jet cooling at a rate of 20–50°C / min. Rapid cooling can suppress grain coarsening of the Cu3Al compound phase and reduce the precipitation of brittle intermediate phases (such as AlCu), thereby improving interfacial toughness. After cooling to below 400°C, it is then cooled to room temperature at a rate of 5–10°C / min. Throughout the cooling process, the interfacial temperature difference is ≤30°C, and the residual stress is ≤150MPa. Below 400°C, the difference in thermal expansion coefficients between Cu and ceramics is significantly reduced. Slow cooling can further release residual stress, and controlling the interfacial temperature difference can prevent the generation of secondary thermal stress.

[0051] In step 9, the vacuum degree of vacuum annealing is less than 0.001 Pa. During the annealing process, a stepped heating method is adopted, with the temperature increased at 5–10°C / min to the set temperature and then held. The high vacuum environment can eliminate residual gas and oxides at the interface, and the stepped heating can avoid thermal shock, promote the gradual release of residual stress and the lattice ordering of the compound phase. In step 10, the interfacial bonding strength is determined by shear test method at a test rate of 0.5–1 mm / min. After thermal cycling test, the interfacial peeling rate is ≤5%. SEM / EBSD characterization shows that the main interfacial compound phase is Cu3Al, with a grain size of 0.5–5 μm and an orientation consistency of ≥80%. The Cu3Al phase has excellent interfacial bonding energy and mechanical compatibility. The grain size and orientation consistency directly affect the interfacial strength and thermal stability. The effect of eutectic reaction control is verified by performance testing and microscopic characterization, forming a closed-loop feedback for process optimization.

[0052] The working principle of this invention is to remove oxides, grease and impurities from the surface of ceramic substrate and copper foil by acid washing or plasma cleaning, while controlling the surface roughness at Ra=0.1–0.5μm, reducing the solid-liquid interfacial tension and improving the spreading ability of the eutectic liquid phase on the ceramic surface; the mixed gas of argon and oxygen in plasma cleaning can further activate the surface, ensuring that the oxide removal rate is ≥95%, and providing a clean interfacial environment for the eutectic reaction.

[0053] The periodic micron-scale grooves or lattice structures on the ceramic contact surface serve two purposes: firstly, as liquid flow channels, guiding the eutectic liquid phase to spread directionally along a predetermined path and avoiding local liquid phase aggregation; secondly, as stress regulation units, dispersing interfacial stress through the geometric constraints of the microstructure, suppressing crack initiation, and simultaneously enhancing the mechanical interlocking effect between the interfacial compound and the ceramic substrate.

[0054] The copper-ceramic temperature gradient (ΔT=20–80°C) formed by asymmetric heating induces a directional thermal stress field, driving the eutectic liquid phase to migrate towards the ceramic side and promoting uniform contact at the interface. The temperature gradient is linearly distributed along the direction perpendicular to the interface, and the highest temperature on the ceramic side does not exceed 95% of that on the copper side, which can prevent the ceramic from cracking due to thermal shock and at the same time provide a stable temperature environment for the eutectic reaction.

[0055] The programmable dynamic pressure (0.5–3MPa) applied by the servo hydraulic system adjusts the interfacial liquid phase thickness (controlled within 1–5μm) in real time through constant, pulsed, or sinusoidal modulated waveforms, promoting atomic diffusion between the liquid and solid phases and accelerating the eutectic reaction. At the same time, through the pressure relaxation effect, it releases the thermal and structural stresses generated during the interfacial reaction, avoiding interfacial delamination caused by stress concentration.

[0056] Embedded high-temperature strain gauges (measurement accuracy ±1με) and DIC technology (displacement accuracy ±0.1μm) monitor the interface strain evolution in real time. The feedback signal drives the dynamic adjustment of temperature gradient and pressure parameters (response period ≤1s) to ensure that the interface is always under controllable compressive stress (0.3–1.5MPa) during the eutectic reaction stage. The subsequent performance evaluation results feed back into the optimization of microstructure parameters, pressure waveforms and other processes, forming a closed-loop control throughout the entire process.

[0057] The working process of this invention is as follows: Substrate pretreatment stage: First, the ceramic substrate and copper foil undergo surface cleaning and activation treatment. If acid washing is used, a 5–15 vol% dilute sulfuric acid or dilute hydrochloric acid solution is selected, and cleaning is performed at 25–50°C for 3–10 min; if plasma cleaning is used, an argon-oxygen mixed gas with an oxygen content of 5–15 vol% is used, and cleaning is performed at 100–300W power for 5–20 min. After treatment, it is necessary to ensure that the surface oxide removal rate is ≥95%, the roughness is controlled at Ra=0.1–0.5μm, and the interfacial wettability is improved.

[0058] Interface microstructure patterning stage: Periodic micron-scale structures are fabricated on the ceramic contact surface using laser micromachining technology. For groove structures (V-shaped, U-shaped, or trapezoidal), the groove depth needs to be controlled at 5–30 μm, the groove width at 10–50 μm, and the groove spacing at 20–100 μm. For lattice structures (cylindrical, square prism, etc.), the lattice spacing is 30–80 μm, the ratio of lattice height to cross-sectional feature size is 0.5–2.0, and the microstructure area coverage is maintained at 20–60%, forming liquid phase guiding and stress control units.

[0059] Stacking and Pre-compression Stage: The processed copper foil and ceramic substrate are stacked in a predetermined order and placed into a high-temperature resistant graphite mold with a 5–20 μm boron nitride release layer on the inner wall. Pre-compression is performed using a step-by-step pressurization method: first, a pressure of 0.1–0.2 MPa is applied and held for 5–10 min, then the pressure is increased to 0.1–0.5 MPa and held for 10–20 min to ensure that the interface gap is ≤5 μm, laying a solid foundation for a tight contact in the subsequent eutectic reaction.

[0060] Gradient thermal field construction stage: The assembly is placed in a controlled atmosphere furnace, and an oxygen-containing inert gas (N2 + 0.1–1 vol% O2) is introduced at a flow rate of 50–200 sccm, while the furnace pressure is maintained at 0.1–0.3 MPa. An asymmetric heating program is initiated, making the heating rate on the copper side 5–15°C / min higher than that on the ceramic side (actual difference 5–10°C / min), forming a linear temperature gradient (ΔT = 20–80°C) from copper to ceramic, thus constructing a directional thermal stress environment.

[0061] Dynamic mechanical constraint application stage: Before the furnace temperature rises to the eutectic reaction temperature range (1060–1085°C), the servo hydraulic system is activated (pressure control accuracy ±0.05MPa, response time ≤50ms) to apply programmable dynamic axial pressure. The pressure waveform is selected based on the interface stress detection results: constant pressure 0.5–3MPa; pulse pressure peak 1.5–3MPa, valley 0.5–1MPa, pulse width 1–10s; or sinusoidal modulated pressure (amplitude 0.3–1MPa, DC bias 0.8–2MPa), all with a frequency of 0.01–1Hz, to regulate the interfacial liquid phase thickness and stress relaxation behavior.

[0062] Stress field control stage of eutectic reaction: After the formation of the eutectic liquid phase, a heat preservation stage (10–30 min) is entered to maintain a gradient thermal field and dynamic mechanical constraint, so that the interface region maintains a controllable compressive stress of 0.3–1.5 MPa. With the help of the microstructure pattern on the ceramic surface, the liquid phase is guided to spread directionally at a rate of 0.1–0.5 mm / s, ensuring that the uniformity error of the interface compound layer thickness is ≤10%, and suppressing local stress concentration and crack initiation.

[0063] In-situ stress monitoring and feedback stage: The interface strain evolution is monitored in real time using embedded high-temperature strain gauges (resistant to room temperature up to 1200°C) and DIC technology (sampling frequency 10–50Hz), and the monitoring data is transmitted to the control system. The control system dynamically adjusts the temperature gradient and pressure parameters within ≤1s based on the feedback signal to achieve closed-loop control of the stress field and ensure that the eutectic reaction proceeds according to the preset mechanism.

[0064] Stress relief stage of cooling process: After the eutectic reaction is completed, segmented controlled cooling is started: First, it is slowly cooled to 800°C at 1–5°C / min to maintain the gradient thermal field and continue the directional stress; then forced air cooling or inert gas jet cooling (rate 20–50°C / min) is used to cool to below 400°C to avoid secondary stress generated by low temperature phase transformation; finally, it is cooled to room temperature at 5–10°C / min. The entire cooling process ensures that the interface temperature difference is ≤30°C and the residual stress is ≤150MPa.

[0065] Post-treatment and interface stabilization stage: The DBC samples were subjected to vacuum annealing at a vacuum level of less than 0.001 Pa, with the temperature increased to 400–600°C in steps at a rate of 5–10°C / min, and held for 1–2 hours. Annealing eliminated residual stress, promoted the ordering and densification of the interfacial compound phase (mainly Cu3Al), improved the interfacial bonding energy, and enhanced interfacial stability.

[0066] Performance evaluation and process optimization phase: Interfacial bonding strength was determined using a shear test (test rate 0.5–1 mm / min). Stability was assessed through thermal cycling tests at -40°C to 150°C for >1000 cycles (requiring a peel rate ≤5%). The morphology (grain size 0.5–5 μm), orientation (uniformity ≥80%), and mechanical properties of the interfacial compounds were characterized using SEM / EBSD. Based on the evaluation results, microstructure parameters, pressure waveforms, and thermal gradients were optimized, and the closed-loop process database was updated to achieve continuous process iteration.

[0067] In plasma cleaning, an argon-oxygen mixed gas with an O2 content of 5–15 vol% is ionized at a power of 100–300W to generate high-energy particles, electrons, ions, and free radicals. Argon ions bombard surface impurities and oxides through physical sputtering, destroying their crystal structure. Oxygen free radicals react with hydrocarbon oil impurities to generate CO2 and H2O, which then volatilize. At the same time, they react with metal oxides to generate easily desorbable gaseous oxides, such as CuO, which reacts with O⁻ to generate CuO2. This dual action ensures that the oxide removal rate is ≥95%.

[0068] The hydrodynamic mechanism of directional liquid phase spreading: Periodic micron-sized grooves with a depth of 5–30 μm and a width of 10–50 μm form capillary channels. According to the capillary rise principle, the rise height of the eutectic liquid phase in the groove is inversely proportional to the groove width and directly proportional to the surface tension. The groove spacing of 20–100 μm can balance the capillary attraction of adjacent grooves and avoid excessive diffusion of the liquid phase between the grooves. The lattice structure with a spacing of 30–80 μm forms liquid phase anchoring points through geometric constraints. The surface energy difference is used to guide the liquid phase to spread evenly along the lattice gaps, preventing uneven compound layer thickness caused by local aggregation.

[0069] The mechanical principle of stress regulation: The microstructure groove / lattice acts as a stress buffer unit, following the stress dispersion theory in elasticity: When interfacial thermal stress and structural stress act together, the geometric discontinuity of the microstructure can change the stress transmission path, transforming the tensile stress that was originally perpendicular to the interface into shear stress along the side of the microstructure, reducing the stress peak by 30-50%, thereby inhibiting crack initiation and raising the crack propagation threshold to ≥200MPa.

[0070] After the eutectic reaction, the metallic phase Cu formed by the solidification of the liquid phase and the interfacial compound phase Cu3Al will be embedded in the microstructure grooves / lattice gaps on the ceramic surface, forming a tenon-and-mortise joint, increasing the interfacial friction and mechanical locking force, and improving the shear strength by ≥40%, far exceeding the physical adsorption effect of planar contact.

[0071] Thermodynamic mechanism of temperature gradient-driven liquid phase migration: A linear temperature gradient of ΔT = 20–80°C is formed by asymmetric heating. Based on the Marangoni effect: the surface tension of the eutectic liquid phase decreases with increasing temperature. The temperature on the ceramic side is higher than that on the copper side, and the highest temperature on the ceramic side is 95% of that on the copper side. This results in a surface tension difference of Δγ = 0.01–0.05 N / m in the direction of the temperature gradient, forming a shear force pointing towards the ceramic side, driving the directional migration of the liquid phase and ensuring uniform contact at the interface.

[0072] Multidimensional mechanism of dynamic pressure control: Atomic diffusion promotion: Dynamic pressure of 0.5–3 MPa increases the contact pressure between the liquid phase and the solid phase, shortening the atomic diffusion distance. According to Fick's diffusion law, the diffusion coefficient is positively correlated with pressure. Pressure control can increase the atomic diffusion rate of Cu and Al by 2–3 times, accelerating the eutectic reaction and shortening the holding time from the conventional 30–60 min to 10–30 min.

[0073] Liquid phase thickness control: Pressure and liquid phase thickness follow hydrostatic equilibrium. By adjusting the pressure in real time using a programmable waveform constant / pulse / sine wave, the liquid phase thickness can be precisely controlled between 1 and 5 μm. This thickness ensures that the minimum liquid phase thickness required for the full growth of the compound phase Cu3Al is ≥0.8 μm, while avoiding the increase in solidification shrinkage stress caused by excessive thickness.

[0074] Stress relaxation effect: The peak pressure of 1.5–3 MPa and the valley pressure of 0.5–1 MPa, along with the sinusoidal modulated pressure, through the pressurization-depressurization cycle, puts the interface in a dynamic stress release state. According to the theory of viscoelastic mechanics, this process can promote the relaxation of the residual stress at the interface, reducing the residual stress to ≤150 MPa and avoiding interface delamination.

[0075] The embedded high-temperature strain gauge has a measurement accuracy of ±1με based on the resistance strain effect. Changes in interface strain will cause a linear change in the strain gauge resistance value. The sensitivity coefficient K=2.0–2.1. Real-time strain monitoring is achieved through resistance signal conversion. The DIC technology has a displacement accuracy of ±0.1μm. It uses a digital image correlation algorithm to track the displacement changes of interface marker points and invert the interface strain field distribution. The two complement each other to achieve a comprehensive perception of the stress state.

[0076] Feedback mechanism for dynamic parameter adjustment: The control system adjusts the temperature gradient and changes the heating power difference and pressure parameters through a PID algorithm based on strain monitoring data, with a response period of ≤1s. When the interfacial tensile stress exceeds 0.3MPa, the system automatically increases the pressure by 0.1–0.2MPa or reduces the heating rate on the copper side by 2–3°C / min, pulling the stress back to the controllable compressive stress range of 0.3–1.5MPa, ensuring that the eutectic reaction takes place in a stable stress environment and avoiding crack formation.

[0077] Stability mechanism of interfacial compound phase: The Cu3Al phase generated by the eutectic reaction is an intermetallic compound with high hardness HV≥400 and good thermal conductivity ≥150W / (mK). Its crystal structure has a lattice mismatch degree ≤5% with the ceramic substrate such as Al2O3, ensuring the structural stability of the interfacial bonding. Vacuum annealing further promotes the densification of Cu3Al phase, reduces porosity ≤1%, and improves the interfacial resistance to thermal cycling performance. The peeling rate is ≤5% after 1000 cycles of -40°C to 150°C.

[0078] Experimental example:

[0079] I. Experimental Materials and Equipment:

[0080] Experimental materials:

[0081] Ceramic substrate: AlN ceramic, dimensions 50mm×50mm×0.635mm, purity ≥99.5%;

[0082] Copper foil: Oxygen-free copper foil, size 50mm×50mm×0.3mm, purity ≥99.99%;

[0083] Auxiliary materials: dilute sulfuric acid (analytical grade), argon (purity ≥99.99%), oxygen (purity ≥99.99%), boron nitride release agent.

[0084] Experimental equipment:

[0085] Plasma cleaner (power 0-500W, gas flow rate 0-500sccm);

[0086] Laser micromachining equipment (wavelength 1064nm, processing accuracy ±1μm).

[0087] Controlled atmosphere furnace (temperature control accuracy ±5°C, supports gradient heating);

[0088] Servo hydraulic system (pressure range 0-5MPa, control accuracy ±0.05MPa);

[0089] Embedded high-temperature strain gauge (temperature range: room temperature - 1200°C, strain accuracy: ±1με).

[0090] DIC testing system (sampling frequency 10-50Hz, displacement accuracy ±0.1μm).

[0091] Vacuum annealing furnace (vacuum degree less than 0.001 Pa, temperature control accuracy ±3°C);

[0092] Universal tensile testing machine (testing range 0-50kN, loading rate 0.1-5mm / min);

[0093] Thermal cycling test chamber (temperature range -60°C to 200°C, number of cycles ≥ 2000);

[0094] SEM / EBSD characterization system (resolution ≤1nm, orientation analysis accuracy ±0.5°).

[0095] II. Experimental Group Design:

[0096] Group Process type Core difference parameters Sample size control group Traditional DBC process Interface-free microstructure, gradient-free thermal field, static pressure (1 MPa) 5 pieces experimental group Method of the present invention V-groove microstructure, copper-ceramic gradient thermal field (ΔT=50°C), sinusoidal modulated dynamic pressure 5 pieces

[0097] III. Detailed process steps for the experimental group:

[0098] 1. Substrate pretreatment:

[0099] Plasma cleaning was used: the mixed gas consisted of Ar and 10 vol% O2, the power was 200 W, the flow rate was 100 sccm, and the cleaning time was 15 min.

[0100] Post-cleaning inspection: The oxide removal rate of the ceramic substrate and copper foil surface was 98.2%, and the surface roughness Ra=0.3μm (detected by laser roughness meter).

[0101] 2. Interface microstructure patterning:

[0102] V-shaped grooves were fabricated on the lower surface of copper foil using laser micromachining: groove depth 15μm, groove width 30μm, groove spacing 60μm, and area coverage 40%;

[0103] Microstructure inspection: Observed with an optical microscope, the uniformity error of the groove size is ≤3%, and there are no burrs or oxidation defects.

[0104] 3. Laminated assembly and pre-compression:

[0105] The inner wall of the graphite mold is coated with a 10μm boron nitride release layer, and the molds are stacked in the order of copper foil-ceramic substrate;

[0106] Stepwise pressurization: First apply 0.15MPa pressure and hold for 8 minutes, then increase to 0.3MPa and hold for 15 minutes. The interface gap is ≤3μm (ultrasonic flaw detector).

[0107] 4. Construct a gradient thermal field environment:

[0108] A mixture of N2 and 0.5 vol% O2 gas was introduced at a flow rate of 150 sccm, and the furnace pressure was 0.2 MPa.

[0109] Heating program: Copper side heating rate 12°C / min, ceramic side heating rate 7°C / min, forming a 50°C temperature gradient (detected by infrared thermometer), the highest temperature on the ceramic side is 5.2% lower than that on the copper side.

[0110] 5. Apply dynamic mechanical constraints:

[0111] When the temperature reaches 1050°C, the servo hydraulic system is activated, applying sinusoidal modulated pressure: amplitude 0.6MPa, DC bias 1.5MPa, frequency 0.5Hz;

[0112] Pressure control: Response time 35ms, interfacial liquid phase thickness stabilized at 3μm (detected by laser interferometer).

[0113] 6. Stress field regulation during the eutectic reaction stage:

[0114] The eutectic reaction temperature was 1075°C, the holding time was 20 min, and the interfacial compressive stress was maintained at 0.9 MPa.

[0115] Liquid phase spreading: Observed by a high-speed camera, the liquid phase spreads at a rate of 0.3 mm / s along the V-shaped groove, and the uniformity error of the compound layer thickness is ≤8%.

[0116] 7. In-situ stress monitoring and feedback:

[0117] Real-time monitoring is achieved using DIC technology, with a sampling frequency of 30Hz and a displacement measurement accuracy of 0.08μm.

[0118] Feedback adjustment: The temperature gradient (±2°C) and pressure amplitude (±0.05MPa) are dynamically fine-tuned based on the strain signal, with a response period of 0.8s.

[0119] 8. Stress relief during the cooling process:

[0120] Segmented controlled cooling: 1075°C to 800°C, cooling rate 3°C / min (maintaining gradient thermal field); 800°C to 350°C, inert gas jet cooling at a rate of 40°C / min; 350°C to room temperature, cooling rate 8°C / min;

[0121] After cooling, the interface temperature difference was 25°C and the residual stress was 120 MPa (X-ray stress meter test).

[0122] 9. Post-processing and interface stabilization:

[0123] Vacuum annealing: temperature 500°C, vacuum degree 0.0005Pa, step heating rate 8°C / min, holding time 1.5h;

[0124] After annealing: the degree of ordering of the interfacial compound phase is improved (XRD detection shows that the intensity of Cu3Al characteristic peak is increased by 20%).

[0125] 10. Performance evaluation and process optimization:

[0126] The interface bonding strength test, thermal cycling test, and SEM / EBSD characterization were completed according to the standard procedure. The data are shown in the table below.

[0127] IV. Performance Test Results and Comparison:

[0128]

[0129] V. Experimental Conclusions:

[0130] The experimental group adopted a three-pronged strategy of microstructure guidance, gradient thermal field and dynamic mechanical constraint to successfully suppress the precipitation of brittle intermetallic compounds (such as CuAl2) and promote the formation of uniform and dense Cu3Al phase.

[0131] The interface bonding strength is improved by more than 30%, the thermal cycle life is more than doubled, and the residual stress is significantly reduced.

[0132] The process parameters showed good repeatability, with performance deviations of ≤5% across 5 samples, validating the stability and reliability of the closed-loop control strategy.

[0133] The above embodiments are only used to illustrate the present invention and are not intended to limit the technical solutions described herein. Although the present invention has been described in detail with reference to the above embodiments, the present invention is not limited to the specific embodiments described above. Therefore, any modifications or equivalent substitutions to the present invention, as well as all technical solutions and improvements that do not depart from the spirit and scope of the invention, are covered within the scope of the claims of the present invention.

Claims

1. A method for regulating the eutectic reaction mechanism of DBC cermet substrates, characterized in that, Includes the following steps: Step 1: Substrate pretreatment: The ceramic substrate and copper foil are cleaned and activated to remove oxides, grease and adsorbed impurities; Acid pickling or plasma cleaning is used to control the surface roughness to Ra = 0.1–0.5 μm; Step 2: Interface microstructure patterning: Periodic micro-grooves or lattice structures are prepared on the ceramic contact surface using laser micromachining to guide liquid flow and serve as stress regulation units; Step 3: Stacking and pre-pressing: The processed copper foil and ceramic substrate are stacked in a predetermined order and placed in a high-temperature resistant graphite mold; Apply an initial preload of 0.1 MPa–0.5 MPa; Step 4: Construct a gradient thermal environment: Place the assembly into a controlled atmosphere furnace and introduce an oxygen-containing inert gas; set an asymmetric heating program so that the heating rate of the copper side is higher than that of the ceramic side, forming a temperature gradient from copper to ceramic, and inducing a directional thermal stress field. Step 5: Apply dynamic mechanical constraints: Before heating to the eutectic reaction temperature range, start the servo hydraulic system to apply a programmable dynamic axial pressure of 0.5MPa–3MPa to the laminate. The pressure waveform is adjusted according to the stress detection results of the interface area, and can be constant, pulsed or sinusoidal modulation, with a frequency of 0.01–1Hz, to control the thickness of the liquid phase at the interface and the stress relaxation behavior. Step 6: Stress field control during the eutectic reaction stage: After the formation of the eutectic liquid phase, maintain the gradient thermal field and simultaneously adjust the mechanical constraint parameters to keep the interface region under controllable compressive stress; use microstructure patterns to guide the liquid phase to spread along a specific path to suppress crack initiation caused by local stress concentration. Step 7: In-situ stress monitoring and feedback: Monitor the strain evolution of the interface region using embedded high-temperature strain gauges; dynamically adjust the pressure and temperature gradients based on the feedback signals; Step 8: Stress relief during cooling process: After the eutectic reaction is completed, a segmented controlled cooling strategy is adopted: first, slowly cool to 800°C at 1–5°C / min to maintain the gradient thermal field to continue the directional stress; then rapidly cool to below 400°C; Step 9: Post-treatment and interface stabilization: Vacuum annealing is performed on the DBC sample to eliminate residual stress and promote the ordering and densification of the interface compound phase. Step 10: Performance evaluation and process optimization: The morphology, orientation and mechanical properties of the interface compound are evaluated through interfacial bonding strength, thermal cycling test and SEM / EBSD microscopic characterization; based on the results feedback, the microstructure parameters, pressure waveform and thermal field gradient are optimized to form a closed-loop process database.

2. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 1, characterized in that, The acid cleaning in step 1 uses a dilute sulfuric acid or dilute hydrochloric acid solution with a concentration of 5–15 vol%, the cleaning temperature is controlled at 25–50°C, and the cleaning time is 3–10 min; the plasma cleaning uses a mixed gas of argon and oxygen, in which oxygen accounts for 5–15 vol%, the plasma power is 100–300 W, and the cleaning time is 5–20 min.

3. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 2, characterized in that, The periodic micron-sized grooves described in step 2 have a V-shaped, U-shaped, or trapezoidal cross-sectional shape, a groove depth of 5–30 μm, a groove width of 10–50 μm, and a groove spacing of 20–100 μm. The lattice structure is cylindrical, square prism or hexagonal prism, with a lattice spacing of 30–80 μm, a lattice height to cross-sectional feature size ratio of 0.5–2.0, and a microstructure pattern area coverage of 20–60%.

4. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 3, characterized in that, The inner wall of the high-temperature graphite mold described in step 3 is coated with a boron nitride release layer with a thickness of 5–20 μm. During the pre-pressing process, a step-by-step pressurization method is adopted. First, a pressure of 0.1–0.2 MPa is applied and maintained for 5–10 min, and then the pressure is increased to the set pre-pressure and maintained for 10–20 min to ensure that the interface gap is ≤5 μm.

5. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 4, characterized in that, In step 4, the flow rate of the oxygen-containing inert gas is 50–200 sccm, and the pressure inside the furnace is maintained at 0.1–0.3 MPa. The heating rate on the copper side is 5–10°C / min higher than that on the ceramic side, the temperature gradient is linearly distributed along the direction perpendicular to the interface, and the highest temperature on the ceramic side does not exceed 95% of the highest temperature on the copper side.

6. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 5, characterized in that, The pressure control accuracy of the servo hydraulic system described in step 5 is ±0.05MPa, and the response time is ≤50ms; the peak value of the pulse pressure is 1.5–3MPa, the valley value is 0.5–1MPa, and the pulse width is 1–10s. The amplitude of the sinusoidal modulated pressure is 0.3–1 MPa, and the DC bias is 0.8–2 MPa.

7. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 6, characterized in that, The holding time for the eutectic reaction stage in step 6 is 10–30 min, and the compressive stress in the interface region is maintained at 0.3–1.5 MPa. The liquid phase spreading rate is guided by the microstructure pattern to be 0.1–0.5 mm / s, so that the thickness uniformity error of the interface compound layer is ≤10%.

8. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 7, characterized in that, The embedded high-temperature strain gauge described in step 7 has a working temperature range from room temperature to 1200°C and a strain measurement accuracy of ±1με; the sampling frequency of the DIC technology is 10–50Hz, the displacement measurement accuracy is ±0.1μm, and the stress field adjustment response period is ≤1s.

9. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 8, characterized in that, The rapid cooling described in step 8 adopts forced air cooling or inert gas jet cooling at a cooling rate of 20–50°C / min; after cooling to below 400°C, it is then cooled to room temperature at a rate of 5–10°C / min. During the entire cooling process, the interface temperature difference is ≤30°C and the residual stress is ≤150MPa.

10. The method for regulating the eutectic reaction mechanism of DBC metal-ceramic substrates according to claim 9, characterized in that, In step 9, the vacuum degree of vacuum annealing is less than 0.001 Pa. During the annealing process, a step heating method is adopted, which raises the temperature to the set temperature at 5–10°C / min and then holds it at that temperature. In step 10, the interfacial bonding strength is determined by shear test method at a test rate of 0.5–1 mm / min. After thermal cycling test, the interfacial peeling rate is ≤5%. SEM / EBSD characterization shows that the interfacial compound phase is mainly Cu3Al, with a grain size of 0.5–5 μm and an orientation consistency of ≥80%.

Images