Welding method of power module and power module

Abstract

The invention discloses a welding method of a power module and the power module, the welding method comprises the following steps: providing a substrate and a to-be-welded object, the surface of the substrate comprises a first metal layer, and the surface of the to-be-welded object comprises a second metal layer; arranging a welding layer on the first metal layer; the surface of the welding layer and the surface of the second metal layer are oppositely arranged after being subjected to surface activation treatment, and a power module is formed through thermocompression bonding; the coplanarity of the surface of the first metal layer, the surface of the welding layer and the surface of the second metal layer is smaller than or equal to 15 nm, and the roughness Ra is smaller than or equal to 10 nm. The welding layer is a copper plating layer and comprises isometric crystals and / or columnar crystals; the grain size of the isometric crystal and / or the columnar crystal is less than or equal to 2.5 microns, and the (111) texture accounts for 65% or more. Through microstructure regulation and control and surface activity treatment on the welding layer, direct bonding of the power module is achieved, and the method has the advantages of being high in welding strength, simple in process, high in reliability and the like.

Description

Technical Field

[0001] This invention relates to the field of electronic integrated module manufacturing, and particularly to a welding method for a power module and a power module. Background Technology

[0002] With the rapid development of power electronic integrated module technology, integrated modules are constantly evolving towards higher power, miniaturization, and higher integration. In electronic packaging, chip interconnect materials play a crucial role in electrical contact, physical protection, heat dissipation, and mechanical support. Currently, the main interconnect materials used include lead-tin alloys, tin-based materials, and gold-based materials, such as 95Pb-5Sn solder, AuSi solder, AuGe solder, Au-AuSn eutectic solder, Ni-Sn3.5Ag, and (CuNi)-Sn3.5Ag solder. However, these traditional materials have significant limitations: high-lead solders and gold-based solders belong to the category of hard solders, and their high melting points lead to large residual stresses inside the joints. This not only reduces the reliability of electronic devices in extreme environments such as high temperatures but also causes serious environmental pollution problems. On the other hand, although lead-free tin-based solder has improved environmental performance, its connection time is relatively long and it is prone to forming hard and brittle intermetallic compounds during the connection process. After high-temperature aging, cracks often occur at the interface between the solder and the substrate. At the same time, its electrical and thermal conductivity is also difficult to meet the requirements of high-performance devices.

[0003] In recent years, sintered silver has attracted widespread attention as a novel packaging interconnect material. This material possesses unique "low-temperature connection, high-temperature service" characteristics, enabling sintering interconnects at temperatures far below the melting point of silver (typically below 300°C), while the sintered interconnect layer can theoretically withstand high-temperature operating environments exceeding 600°C. Sintered silver not only exhibits excellent electrical and thermal conductivity but also demonstrates good mechanical properties, making it a research hotspot in the field of chip interconnection. Furthermore, the interconnect interface formed using sintered silver effectively avoids solder remelting problems caused by repeated welding, solves process challenges such as substrate warpage, and ensures the compatibility of subsequent high-temperature substrate interconnect interface materials with welding processes, thereby significantly improving the reliability of substrate interface connections. This technological breakthrough provides a new solution for the reliable packaging of high-power-density electronic devices.

[0004] However, existing power device or power module packaging structure soldering technologies (including solder and sintered silver technology) still have the following drawbacks:

[0005] (1) Silver-based sintering slurry materials are expensive; during the sintering process, the slurry suffers from problems such as dimensional shrinkage and overflow, which affect the precision and insulation performance.

[0006] (2) The silver-based sintering paste process requires a silver layer to be pre-placed on the surface of the copper-clad ceramic substrate. However, the resulting silver / copper interface and the silver / pre-placed silver interface in the paste have reliability issues such as silver migration, high-temperature oxidation, and dewetting of the plating.

[0007] (3) The use of solder and sintering paste processes results in a multi-layer metal interface structure, large stray inductance, and high product loss.

[0008] Therefore, existing technologies still need improvement and development. Summary of the Invention

[0009] In view of the shortcomings of the prior art, the present invention provides a welding method and a power module, which aims to replace the original solder and sintering materials by using the technology of coating structure regulation and surface activity treatment to promote hot pressing bonding, thereby solving the problems of high cost and unstable performance of power modules prepared by the existing solder process.

[0010] The technical solution of the present invention is as follows:

[0011] Firstly, a method for welding a power module is provided, comprising the following steps:

[0012] A substrate is provided, the surface of which includes a first metal layer;

[0013] A workpiece is provided, the surface of which includes a second metal layer;

[0014] A welding layer is formed on the first metal layer;

[0015] The surfaces of the weld layer and the second metal layer are subjected to surface activation treatment;

[0016] The surfaces of the welding layer and the second metal layer are positioned opposite each other and formed into a power module by hot-press bonding.

[0017] Wherein, the coplanarity of the surface of the first metal layer, the surface of the welding layer and the surface of the second metal layer is ≤15nm and the roughness Ra is ≤10nm;

[0018] The welding layer is a copper plating layer, comprising: equiaxed crystals and / or columnar crystals;

[0019] The equiaxed and / or columnar crystals have a grain size ≤ 2.5 μm and a (111) texture ratio of more than 65%.

[0020] In a preferred embodiment, the equiaxed crystals and / or columnar crystals contain nanotwins.

[0021] In a preferred embodiment, before the step of setting the welding layer on the first metal layer, the method further includes: performing a surface planarization treatment on the first metal layer;

[0022] And / or, before the step of performing surface activation treatment on the surface of the weld layer and the surface of the second metal layer, the method further includes: performing surface planarization treatment on the surface of the weld layer and the surface of the second metal layer.

[0023] In a further preferred embodiment, the surface planarization treatment method is selected from one or more of electropolishing, mechanical grinding, mirror finishing, and chemical mechanical grinding.

[0024] In a preferred embodiment, the surface activation treatment method is chemical treatment and / or plasma treatment;

[0025] The chemical treatment uses small molecules of carboxylic acids and / or small molecules of alcohols;

[0026] The plasma treatment uses one or more of N2, O2, and Ar.

[0027] In a further preferred embodiment, the carboxylic acid molecule is selected from one or more of formic acid, acetic acid, citric acid, and ascorbic acid; and / or, the alcohol molecule is ethanol or glycerol.

[0028] In a preferred embodiment, the hot-press bonding is performed in a vacuum or ambient atmosphere, wherein the ambient atmosphere is selected from the atmosphere, an inert gas, or a reducing gas.

[0029] And / or, the temperature of the hot-press bonding is 150–300°C, and the pressure of the hot-press bonding is 10–60 MPa.

[0030] In a preferred embodiment, the thickness of the welding layer is 10–100 μm.

[0031] In a preferred embodiment, the materials of the first metal layer and the second metal layer are independently selected from one or more of gold, silver, copper, and nickel.

[0032] In a second aspect, a power module is provided, which is welded using the welding method described in the first aspect.

[0033] Beneficial effects: This invention provides a welding method for power modules, which has the following advantages compared with the prior art:

[0034] (1) It can achieve direct bonding, which simplifies the packaging of power modules and features high welding strength, simple process and high reliability.

[0035] (2) It can reduce the bonding temperature and shorten the bonding time, and can be applied to power devices and module packaging to replace traditional solder or silver sintering technology, greatly reducing the cost of welding process and reducing flux volatilization, which has the characteristics of energy saving and environmental protection.

[0036] (3) The solderability of the power module has been optimized, and it also has high thermal conductivity, heat dissipation and resistance to mechanical deformation, which improves the life of the power module. Attached Figure Description

[0037] Figure 1 This is the cross-sectional morphology of the copper plating layer prepared in Example 1.

[0038] Figure 2 This is a flowchart of the welding method in Example 1.

[0039] Figure 3 This is an ultrasonic microscope characterization image of the welded assembly after welding in Example 1.

[0040] Figure 4 This is a schematic diagram of the power module after welding in Examples 1 to 3; wherein, 1 is a lead wire, 2 is a copper foil, 3 is the first welding layer, 4 is a power chip, 5 is the second welding layer, 6 is a copper-clad ceramic substrate, 7 is the third welding layer, and 8 is a heat sink. Detailed Implementation

[0041] This invention provides a welding method for a power module and a power module. To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention is further described in detail below. It should be understood that the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. Furthermore, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for clearly describing a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed. The component numbers used herein, such as "first," "second," etc., are only used to distinguish the described objects and do not have any sequential or technical meaning. All instruments and reagents used are commercially available products.

[0042] This invention provides a welding method for a power module, comprising the following steps:

[0043] A substrate is provided, the surface of which includes a first metal layer;

[0044] A workpiece is provided, the surface of which includes a second metal layer;

[0045] A welding layer is formed on the first metal layer;

[0046] The surfaces of the weld layer and the second metal layer are subjected to surface activation treatment;

[0047] The surfaces of the welding layer and the second metal layer are positioned opposite each other and formed into a power module by hot-press bonding.

[0048] Wherein, the coplanarity of the surface of the first metal layer, the surface of the welding layer and the surface of the second metal layer is ≤15nm and the roughness Ra is ≤10nm;

[0049] The welding layer is a copper plating layer, comprising: equiaxed crystals and / or columnar crystals;

[0050] The equiaxed and / or columnar crystals have a grain size ≤ 2.5 μm and a (111) texture ratio of more than 65%.

[0051] Specifically, this invention electroplats a special copper coating on the first metal layer of the substrate surface, controlling the grain size of the equiaxed and / or columnar crystals in the copper coating to be ≤2.5μm. The reduction in grain size can accelerate grain boundary diffusion. At the same time, the proportion of the (111) texture of the equiaxed and / or columnar crystals is controlled to be above 65%. The diffusion coefficient of copper atoms on the (111) crystal plane is much higher than that of other crystal planes (e.g., the diffusion coefficient of Cu (111) crystal plane is 1.5 to 3 times that of (100) crystal plane), which can achieve hot-press bonding in a lower temperature range (150 to 300℃). Furthermore, controlling the coplanarity and roughness of the copper coating surface and the second metal layer surface of the workpiece, and performing surface activity treatment, can improve the connection strength and ensure the welding effect. Through the above improvements, this invention reduces the bonding temperature and shortens the bonding time, and can replace traditional solder or silver sintering technology in the packaging of power modules. It greatly reduces the welding process cost, reduces flux volatilization, and has the advantages of high welding strength, simple process, high reliability, energy saving and environmental protection.

[0052] Specifically, the grain size of equiaxed crystals and / or columnar crystals is ≤2.5μm, meaning that the grain diameter of equiaxed crystals is ≤2.5μm and the lateral grain size of columnar crystals is ≤2.5μm. Preferably, the grain diameter of equiaxed crystals is between 0.2 and 2.5 micrometers; the lateral grain size of columnar crystals is between 0.2 and 2.5 micrometers. The height of the columnar crystals is related to the thickness of the copper plating layer and is dependent on actual operation, but has little impact on the welding effect; therefore, it is not limited here.

[0053] In one embodiment, the equiaxed and / or columnar crystals contain nanotwins. High-density nanotwins can give copper tensile strength more than ten times higher than ordinary pure copper and electrical conductivity comparable to oxygen-free high-purity copper, and are also very effective in improving the material's electrical properties, plasticity, and fatigue resistance. Nanotwin boundaries can effectively hinder dislocation movement, achieving metal strengthening. At the same time, nanotwin boundaries have less obstruction to electron heat dissipation and a much smaller impact on electrical conductivity than other grain boundaries and defects. Therefore, the presence of nanotwins in the equiaxed and / or columnar crystals is more conducive to achieving low-temperature welding. The lamellar orientation of the nanotwins, viewed from the film surface, can be horizontally distributed, vertically distributed, or randomly anisotropic.

[0054] In one embodiment, the substrate is selected from one or more of copper foil, power chip, copper-clad ceramic substrate and heat sink. The first metal layer on the surface of the substrate has the following characteristics: (1) large area, ranging from 2 square millimeters to 50 square centimeters, (2) large initial surface roughness, with an initial roughness Ra greater than 600 nm, and no special control over the surface crystal orientation and grain size.

[0055] In one embodiment, before the step of forming a welding layer on the first metal layer, the method further includes: performing a surface planarization treatment on the first metal layer.

[0056] In a more specific embodiment, before the step of setting the welding layer on the first metal layer, the method further includes: sequentially performing pickling, surface planarization treatment, and deionized water cleaning on the first metal layer.

[0057] The pickling process involves soaking and stirring in a 5-10% (w / v) sulfuric acid (H2SO4) solution or soaking with ultrasonic waves.

[0058] In one embodiment, before the step of performing surface activation treatment on the surface of the weld layer and the surface of the second metal layer, the method further includes performing surface planarization treatment on the surface of the weld layer and the surface of the second metal layer.

[0059] In a more specific embodiment, the surface planarization method is selected from one or more of electropolishing, mechanical polishing, mirror finishing, and chemical mechanical polishing (CMP).

[0060] The electropolishing process involves placing the substrate as the anode in an electropolishing solution, applying a voltage of 2–12V for 5–300 seconds.

[0061] Mechanical grinding processes use traditional grinding equipment or milling to grind the surface, reducing the roughness to below 1000nm.

[0062] The mirror-finish rolling process improves surface quality, with a roll roughness Ra < 10 nm. Precise control of the rolling force within 50–100 MPa and the reduction within 0.5–2 μm are crucial; excessive reduction can lead to microcracks in the matrix. Rolling introduces prestress at the bonding interface, which helps accelerate metal diffusion bonding on both sides of the interface.

[0063] The chemical mechanical polishing process involves placing the substrate under the polishing head, using a pressure of 1–5 psi, a polishing pad rotation speed of 50–500 r / min, and a polishing slurry flow rate of 60–120 mL / min.

[0064] The above-mentioned surface planarization methods can be used alternately until the coplanarity is within 15nm and the roughness Ra is reduced to below 10nm. For example, for the surface planarization of copper-clad substrates, machining, CMP, and deionized water cleaning can be selected.

[0065] In one embodiment, the surface activation treatment method is chemical treatment and / or plasma treatment;

[0066] The chemical treatment uses small molecules of carboxylic acids and / or alcohols;

[0067] The plasma treatment uses one or more of N2, O2, and Ar.

[0068] In a more specific embodiment, the carboxylic acid small molecule is selected from one or more of formic acid, acetic acid, citric acid, and ascorbic acid; and / or, the alcohol small molecule is ethanol or glycerol. Specifically, the present invention uses acid and / or alcohol small molecules for surface-active treatment, and the treated material can be used for hot-press bonding without washing. During hot-press bonding (150–300°C), the carboxylic acid and / or alcohol small molecules can undergo significant decomposition, thereby playing a reducing role.

[0069] In one embodiment, the thermocompression bonding is performed in a vacuum or ambient atmosphere, the ambient atmosphere being selected from the atmosphere, an inert gas, or a reducing gas.

[0070] In one embodiment, the hot-press bonding temperature is 150–300°C, and the hot-press bonding pressure is 10–60 MPa. Preferably, the hot-press bonding temperature is 180–280°C, and the hot-press bonding pressure is 20–40 MPa.

[0071] In one embodiment, the thickness of the weld layer is 10–100 μm.

[0072] In one embodiment, the materials of the first metal layer and the second metal layer are independently selected from one or more of gold, silver, copper, and nickel.

[0073] In one embodiment, a power module is provided, which is welded using the welding method described in the first aspect.

[0074] In one embodiment, the step of forming a welding layer on the first metal layer specifically includes: forming a welding layer on the first metal layer by electrodeposition;

[0075] The electrodeposition method specifically includes:

[0076] The plating solution is placed in an electroplating tank. A phosphorus-copper anode (phosphorus content 0.03–0.075 wt%) is immersed in the solution, with the first metal layer serving as the cathode. The flow rate of the plating solution is 100–1000 rpm, the temperature is 5–50°C, and the current density is 50–90 mA / cm². 2 Electroplating time is 20–600 min.

[0077] The plating solution is prepared with water and includes: 0.3-0.6 mol / L copper salt, 0.5-1.5 mmol / L chloride ions, 0.2-0.7 mol / L sulfuric acid, and 5-200 ppm additives.

[0078] The copper salt may be obtained from one or more of copper sulfate, copper methanesulfonate, and copper aminosulfonate.

[0079] The chloride ions can be obtained from one or more of hydrochloric acid, sodium chloride (NaCl), and potassium chloride (KCl);

[0080] The sulfuric acid can be obtained by diluting 96-98 wt% concentrated sulfuric acid (H2SO4);

[0081] The additive is a grain refiner and / or a twinning accelerator; the grain refiner is sodium polydithiopropane sulfate (SPS), and the twinning accelerator is gelatin; preferably, the concentration of sodium polydithiopropane sulfate is 1-20 ppm, and the concentration of gelatin is 10-150 ppm; more preferably, the concentration of sodium polydithiopropane sulfate is 2-10 ppm, and the concentration of gelatin is 25-100 ppm.

[0082] The method for controlling the flow rate of the plating solution is selected from one or more of magnetic stirring, mechanical stirring, bubbling and blowing, and ultrasound.

[0083] In one specific embodiment, a method for welding a power module includes the following steps:

[0084] The invention provides copper foil, a power chip, a copper-clad ceramic substrate, and a heat sink; wherein the copper-clad ceramic substrate includes a ceramic sheet and a copper foil sheet located on the surface of the ceramic sheet, the copper foil sheet includes a first copper foil sheet and a second copper foil sheet, the first copper foil sheet and the second copper foil sheet being separately disposed on opposite sides of the ceramic sheet; the power chip and the heat sink have a metal layer on their surfaces;

[0085] One side of the copper foil is subjected to surface planarization and surface activation treatment;

[0086] A first welding layer is formed by electrodeposition on one side of the copper foil that has undergone surface planarization and surface activation treatment;

[0087] The side of the first welding layer away from the copper foil is subjected to surface planarization and surface activation treatment;

[0088] The side of the first welding layer away from the copper foil and the side of the power chip are arranged opposite each other, and a first welding assembly is formed by hot pressing bonding;

[0089] The side of the first copper foil away from the ceramic sheet is subjected to surface planarization and surface activation treatment;

[0090] A second welding layer is formed on the side of the first copper foil away from the ceramic sheet by electrodeposition;

[0091] The side of the second welding layer away from the first copper foil is subjected to surface planarization and surface activation treatment;

[0092] The side of the second welding layer away from the first copper foil is positioned opposite the side of the power chip away from the first welding layer, and the second welding assembly is formed by hot-press bonding.

[0093] The side of the second copper foil away from the ceramic sheet is subjected to surface planarization and surface activation treatment;

[0094] A third welding layer is formed on the side of the second copper foil away from the ceramic sheet by electrodeposition;

[0095] The side of the third welding layer away from the second copper foil and the side of the heat sink are subjected to surface planarization and surface activation treatment.

[0096] The third welding layer, the side away from the second copper foil, and the side of the heat sink that has undergone surface planarization and surface activation treatment are arranged opposite each other, and a power module is formed by hot pressing bonding.

[0097] Among them, the coplanarity of all surfaces that have undergone surface planarization treatment is ≤15nm and the roughness Ra is ≤10nm;

[0098] The first welding layer, the second welding layer and the third welding layer are all copper plating layers, and the copper plating layers include: equiaxed crystals and / or columnar crystals;

[0099] The equiaxed and / or columnar crystals have a grain size ≤ 2.5 μm and a (111) texture ratio of more than 65%.

[0100] Specifically, the power module welding method provided by this invention is mainly applied in the following three welding scenarios, replacing the original solder and sintering materials:

[0101] (1) The connection between the upper circuit copper layer surface of the copper-clad ceramic substrate and the power chip is usually achieved by coating with tin-based solder or sintering material and then reflow soldering or hot pressing sintering.

[0102] (2) The connection between the copper foil required for wire bonding between the upper surface of the power chip and the connecting substrate frame involves first connecting a layer of copper foil to the upper surface of the chip using a nano-silver film, and then performing wire bonding on the copper foil surface. The pre-placed nano-silver film is connected and fixed between the copper foil and the upper surface of the chip by pressure sintering. The above-mentioned copper layer process scheme on the chip is called... (Die Top System).

[0103] (3) Connection between power devices / modules and heat dissipation substrate. Currently, there are various materials that can achieve this, such as thermal grease, ceramic insulating thermal pads, tin-based solder, silver (or copper) conductive adhesive and silver (or copper) sintered paste, etc.

[0104] Based on the same inventive concept, this embodiment of the invention provides a power module, which is welded using the welding method described above.

[0105] In one implementation, such as Figure 4 As shown, the power module, from top to bottom, includes: a copper foil 2 for connection with the lead wire 1, a first solder layer 3, a power chip 4, a second solder layer 5, a copper-clad ceramic substrate 6, a third solder layer 7, and a heat sink 8. The copper foil 2 and the power chip 4 are soldered together via the first solder layer 3; the power chip 4 and the copper-clad ceramic substrate 6 are soldered together via the second solder layer 5; and the copper-clad ceramic substrate 6 and the heat sink 8 are soldered together via the third solder layer 7. The first solder layer 3, the second solder layer 5, and the third solder layer 7 are all copper plating layers, and the copper plating layers include equiaxed crystals and / or columnar crystals; the grain size of the equiaxed crystals and / or columnar crystals is ≤2.5μm, and the (111) texture accounts for more than 65%.

[0106] The present invention will be further described below through specific embodiments.

[0107] Example 1

[0108] This embodiment provides a method for welding a copper-clad ceramic substrate and a power chip, as detailed below:

[0109] Step 1: Substrate Preparation. The substrate is a copper-clad ceramic substrate. The substrate surface is rough and contains numerous microscopic steps. First, the roughness Ra of the metal layer on the substrate surface is reduced to below 1000 nm using machining or precision rolling. The following operations are performed sequentially: immersion in 8wt% sulfuric acid (H2SO4) solution for 10 min; ultrasonic cleaning: ultrasonic power 300W, time 5 min; electrolytic polishing: the substrate is connected as the anode in an electrolytic polishing device, using an electrolytic polishing solution with a phosphoric acid:ethanol volume ratio of 1.1, applying a voltage of 8V until the surface roughness Ra is reduced to 5 nm and the coplanarity to 10 nm; finally, rinsing with deionized water.

[0110] Step 2: Electrodeposition. Prepare the plating solution using the following component ratio and ensure thorough and uniform dispersion: 0.3 mol / L copper sulfate, 0.5 mmol / L hydrochloric acid, 0.7 mol / L sulfuric acid, 50 ppm gelatin, 5 ppm SPS, and the remainder is pure water.

[0111] The plating solution was placed in an electroplating tank. The substrate treated in step 1 was immersed in the solution as the cathode and the phosphorus-copper anode (phosphorus content 0.03wt%). The flow rate of the plating solution was controlled at 500 rpm using mechanical stirring. The solution temperature was 25℃, and the current density was 50 mA / cm². 2 The electroplating time is 60 minutes.

[0112] Step 3: Preparation before welding. The copper plating layer obtained in Step 2 is subjected to surface planarization treatment. First, electropolishing is performed using a mixed solution of 85wt% phosphoric acid and deionized water (5wt% phosphoric acid to deionized water volume ratio of 3:1) at a voltage of 2.5V. After electropolishing, the surface roughness is reduced to below 30nm. Then, mirror-finish rolling is performed with a roll roughness <5nm. The rolling force is precisely controlled at 50MPa and the reduction is controlled at 0.5μm until the surface roughness Ra is reduced to 5nm and the coplanarity is reduced to 10nm. Subsequently, activation treatment is performed. Both the substrate and the power chip surface are rinsed with 5wt% dilute hydrochloric acid for 5s, and the substrate surface is coated with 1μL of 0.1mol / L citric acid.

[0113] Step 4: Hot-press welding. The copper plating on the copper-clad ceramic substrate treated in Step 3 is used as the welding layer and directly hot-pressed with the power semiconductor chip (the surface metal layer is Ni (1.2μm)-Au (0.07μm)). The pressure is 20MPa, the temperature is 280℃, and the bonding time is 10min. The atmosphere is a formic acid reducing atmosphere.

[0114] Example 1 is Figure 4Regarding the second welding layer 5, the copper plating layer prepared in step 2 is shown below. Figure 1 The flowchart for this embodiment can be found in [link to flowchart]. Figure 2 The ultrasonic microscopy characterization of the welded assembly after welding is shown in the figure. Figure 3 After hot-press bonding, the weld strength was tested, and the average shear strength was 54.8 MPa.

[0115] Depend on Figure 1 It can be seen that the grains in the lower substrate (copper-clad ceramic substrate) are relatively large, and the grain boundaries and subgrain boundaries are relatively sparse; while the grain boundaries and twin boundaries of the upper copper plating layer are denser, and the microstructure is basically columnar crystals containing high-density twins. Due to the influence of the substrate, the copper plating layer near the lower part of the substrate has a small number of small grains, most of which are less than 1 micrometer in diameter. There are also high-density twin boundaries inside the columnar crystals, and the twin lamellae are basically horizontal. The preferred orientation of these grains containing nanotwins is the (111) direction, the texture accounts for more than 90%, and the proportion of other crystal planes is less than 10%.

[0116] Comparative Example 1

[0117] This comparative example provides a method for welding a copper-clad ceramic substrate and a power chip. The only difference between this method and Example 1 is that gelatin is not added in step 2. The specific differences are as follows:

[0118] Step 2: Electrodeposition. Prepare the electroplating solution using the following component ratio and ensure it is thoroughly mixed: 0.3 mol / L copper sulfate, 0.5 mmol / L hydrochloric acid, 0.7 mol / L sulfuric acid, 5 ppm SPS, and the remainder is pure water.

[0119] The plating solution was placed in an electroplating tank. The substrate treated in step 1 was immersed in the solution as the cathode and the phosphorus-copper anode (phosphorus content 0.03wt%). The flow rate of the plating solution was controlled at 500 rpm using mechanical stirring. The solution temperature was 25℃, and the current density was 50 mA / cm². 2 The electroplating time is 60 minutes.

[0120] The copper coating prepared in step 2 of this comparative example has a grain size of approximately 2 μm and a (111) texture accounting for 50%. Under the same hot-press bonding conditions as in Example 1, the weld exhibited numerous pores and cracks. The average shear strength was 15.2 MPa.

[0121] Comparative Example 2

[0122] This comparative example provides a method for bonding a copper-clad ceramic substrate and a power chip. The only difference between this method and Example 1 is that SPS is not added in step 2. The specific differences are as follows:

[0123] Step 2: Electrodeposition. Prepare the electroplating solution using the following component ratio and ensure it is thoroughly mixed: 0.3 mol / L copper sulfate, 0.5 mmol / L hydrochloric acid, 0.7 mol / L sulfuric acid, 50 ppm gelatin, and the remainder is pure water.

[0124] The plating solution was placed in an electroplating tank. The substrate treated in step 1 was immersed in the solution as the cathode and the phosphorus-copper anode (phosphorus content 0.03wt%). The flow rate of the plating solution was controlled at 500 rpm using mechanical stirring. The solution temperature was 25℃, and the current density was 50 mA / cm². 2 The electroplating time is 60 minutes.

[0125] The copper coating prepared in step 2 of this comparative example has a grain size of approximately 4 μm and a (111) texture accounting for 80%. Under the same hot-press bonding conditions as in Example 1, the weld exhibited numerous pores and cracks. The average shear strength was 35.8 MPa.

[0126] Example 2

[0127] This embodiment provides a welding method for a copper-clad ceramic substrate and a heat sink, as detailed below:

[0128] Step 1: Substrate Preparation. The substrate is a copper-clad ceramic substrate. The following operations are performed sequentially: acid washing by soaking in 5wt% sulfuric acid (H2SO4) solution for 10 min; first CMP rough polishing: pressure of 5 psi, polishing pad rotation speed of 300 r / min, polishing fluid flow rate controlled at 80 mL / min, until the surface roughness Ra is reduced to below 500 nm; then a second CMP is performed, with a pressure of 1 psi, polishing pad rotation speed of 100 r / min, polishing fluid flow rate controlled at 60 mL / min, until the surface roughness Ra is reduced to 10 nm and the coplanarity is reduced to 10 nm.

[0129] Step 2: Electrodeposition. Prepare the plating solution using the following component ratio and ensure it is thoroughly mixed: copper methanesulfonate 0.6 mol / L, sodium chloride 1.5 mmol / L, sulfuric acid 0.7 mol / L, gelatin 100 ppm, and the remainder is pure water.

[0130] The plating solution was placed in an electroplating tank. The substrate treated in step 1 was immersed in the solution as the cathode and the phosphorus copper anode (phosphorus content 0.03wt%). The flow rate of the plating solution was controlled at 600 rpm using mechanical stirring, and 100W ultrasonic stirring was added. The plating solution temperature was 35℃, and the current density was 70 mA / cm². 2 The electroplating time was 120 min. The obtained copper coating was columnar crystals with a preferred orientation of (111), the (111) texture accounted for 80%, and the average transverse size of the columnar crystals was 1.5 micrometers.

[0131] Step 3: Preparation before soldering. The copper plating and heat sink (copper surface) prepared in Step 2 are subjected to surface planarization treatment until the surface roughness Ra is reduced to 5 nm and the coplanarity to 10 nm, followed by surface activation treatment. The surface planarization of the copper plating on the copper-clad ceramic substrate is performed using chemical mechanical polishing (CMP), with a pressure of 1 psi, a polishing pad rotation speed of 100 r / min, and a polishing slurry flow rate of 60 mL / min. The surface planarization of the heat sink is performed sequentially using mechanical grinding and CMP. After reducing the surface roughness to 1000 nm using a fine grinding process, CMP is used for finishing. The CMP pressure is 2 psi, the polishing pad rotation speed is 100 r / min, and the polishing slurry flow rate is controlled at 60 mL / min.

[0132] After surface planarization, the copper plating and heat sink on the copper-clad ceramic substrate are subjected to surface activation treatment. The surface activation treatment process is O2 plasma treatment with a power of 150W and a treatment time of 10s.

[0133] Step 4: Hot-press welding. The copper plating on the copper-clad ceramic substrate treated in Step 3 is used as the welding layer and aligned with the heat sink treated in Step 3. Direct hot-press diffusion welding is performed at a pressure of 40 MPa, a temperature of 250℃, and a welding time of 30 minutes. Post-weld annealing is then carried out for 60 minutes in a vacuum environment. The resulting welded assembly exhibits good weld quality, with no pores or cracks in the weld seam, and a weld strength of 70 MPa.

[0134] Example 2 is Figure 4 The principle for the third bonding layer 7 is similar to that of the first bonding layer 3 and the second bonding layer 5. However, it is worth noting that because traditional copper-clad ceramic substrates and heat sinks have rough surfaces and very large grains, and the bonding area is very large (approximately 20-50 square centimeters), surface planarization here requires chemical mechanical polishing to quickly obtain a flat surface (coplanarity ≤10nm) before large-area diffusion bonding can be performed. In particular, the bonding time and annealing time are also increased accordingly to ensure good bonding results.

[0135] Example 3

[0136] This embodiment provides a method for welding copper foil and power chips, as detailed below:

[0137] Step 1: Substrate Preparation. The substrate is a copper foil with a surface roughness Ra of 30 μm. The following operations are performed sequentially: pickling by soaking in 10 wt% sulfuric acid (H2SO4) solution for 10 min; mirror finishing rolling: roll roughness of 5 nm, rolling force of 70 MPa, reduction controlled at 1 μm, until the surface roughness Ra is reduced to 5 nm and the coplanarity is reduced to 10 nm.

[0138] Step 2: Electrodeposition. Prepare the plating solution using the following component ratio and ensure it is thoroughly mixed: 0.6 mol / L copper sulfate, 1 mmol / L sodium chloride, 0.2 mol / L sulfuric acid, 100 ppm gelatin, 10 ppm SPS, and the remainder is pure water.

[0139] The plating solution was placed in an electroplating tank. The substrate treated in step 1 was immersed in the solution as the cathode and the phosphorus-copper anode (phosphorus content 0.075wt%). The flow rate of the plating solution was controlled at 600rpm using magnetic stirring. The solution temperature was 15℃, and the current density was 90mA / cm². 2 The electroplating time was 30 min. The obtained copper coating was equiaxed crystal with (111) preferred orientation, with (111) texture accounting for 85% and an average grain size of 0.5 μm.

[0140] Step 3: Preparation before welding. The copper plating obtained in Step 2 is subjected to surface planarization treatment. First, electropolishing is performed using a mixed solution of 85wt% phosphoric acid and deionized water (5wt% phosphoric acid to deionized water volume ratio 3:1). The copper foil is placed at the anode and connected to the electropolishing device, with the polishing voltage controlled at 2V. After electropolishing, the roughness is reduced to below 30nm. Next, mirror-finish rolling is performed with a roll roughness of 5nm, a rolling force of 70MPa, and a reduction controlled at 1μm. This reduces the surface roughness of the copper plating on the copper foil to approximately 5nm and the coplanarity to 10nm. A low-speed saw is then used to cut the entire copper foil, including the copper plating, into 2mm*3mm square sheets.

[0141] The copper plating and power chip were then subjected to surface activation treatment using a two-step Ar and N2 plasma activation process. The Ar plasma treatment was performed at a power of 250W for 100 seconds, while the N2 plasma treatment was performed at a power of 350W for 60 seconds. Subsequently, a chemical treatment was carried out using a mixture of formic acid and glycerol. A 1:1 volume ratio mixture of formic acid and glycerol was prepared, and 2 microliters of this mixture were coated onto the surface of the power chip.

[0142] Step 4: Hot-press welding. The copper plating on the copper foil treated in Step 3 is used as the welding layer and aligned with the power chip treated in Step 3. Direct hot-press diffusion welding is performed at a pressure of 30 MPa, a temperature of 180℃, and a welding time of 40 minutes in an N2 atmosphere. The resulting power module exhibits excellent welding performance, with no pores or cracks in the weld seam.

[0143] Example 3 is Figure 4In the case of the first weld layer 3, the copper foil planarization process uses a precision rolling process, which is simple to operate and highly efficient. It also introduces additional stress as a driving force for grain growth during bonding, thus improving bonding quality. On the other hand, unlike the second weld layer 5 and the third weld layer 7, the copper foil involves cutting, requiring strict control of the cutting temperature to prevent excessive temperature from causing changes in the bond structure and pre-structure. A mixture of formic acid and glycerol is used in the chemical treatment. Besides removing oxidation, formic acid and glycerol act as reducing agents during heating, promoting interfacial diffusion bonding.

[0144] Example 4

[0145] This embodiment 4 provides a method for welding a copper-clad ceramic substrate and a power chip. The only difference between this method and embodiment 1 is that the amount of gelatin and SPS used in step 2 is reduced. The specific differences are as follows:

[0146] Step 2: Electrodeposition. Prepare the electroplating solution using the following component ratio and ensure it is thoroughly mixed: copper sulfate 0.3 mol / L, hydrochloric acid 0.5 mmol / L, sulfuric acid 0.7 mol / L, gelatin 25 ppm, SPS 2 ppm, and the remainder is pure water.

[0147] The copper plating obtained in step 2 of this embodiment has a grain size of approximately 2 μm and a (111) texture accounting for 65%. Under the same hot-press bonding conditions as in Example 1, the average shear strength is 40.5 MPa.

[0148] Example 5

[0149] This embodiment provides a welding method for a copper-clad ceramic substrate and a heat sink, which differs from Embodiment 2 in that the surface planarization treatment in steps 1 and 3 is different, as detailed below:

[0150] Step 1: Substrate Preparation. The substrate is a copper-clad ceramic substrate. The following operations are performed sequentially: pickling with 5wt% sulfuric acid (H2SO4) solution for 10 min, mechanical grinding, electropolishing, and mirror finishing rolling. The electropolishing solution is a mixture of 85wt% phosphoric acid and deionized water (volume ratio of 5wt% phosphoric acid to deionized water is 3:1), the polishing voltage is 5V, and the time is 300s. For mirror finishing rolling, a roll roughness of 5nm is selected, the rolling force is precisely controlled at 100MPa, and the reduction is controlled at 1.5μm, until the surface roughness Ra is reduced to 5nm and the coplanarity is reduced to 10nm.

[0151] Step 3: Preparation before welding. The copper plating and heat sink prepared in Step 2 are subjected to surface planarization treatment until the surface roughness Ra is reduced to 10 nm and the coplanarity is reduced to 10 nm, followed by surface activation treatment. The surface planarization treatment of the copper plating on the copper-clad ceramic substrate involves sequentially immersing in 5 wt% sulfuric acid (H2SO4) for 10 min for pickling, mechanical grinding, electrolytic polishing, and mirror-finish rolling. The electrolytic polishing solution is a mixture of 85 wt% phosphoric acid and deionized water (volume ratio of 5 wt% phosphoric acid to deionized water is 3:1), with a polishing voltage of 5 V and a time of 300 s. Mirror-finish rolling uses rolls with a roughness of 5 nm, precisely controlling the rolling force at 100 MPa and the reduction at 1.5 μm. The surface planarization treatment of the heat sink involves sequentially immersing in 5 wt% sulfuric acid (H2SO4) for 10 min for pickling, mechanical grinding, and electrolytic polishing. The electropolishing solution consisted of a mixture of 85 wt% phosphoric acid and deionized water (5 wt% phosphoric acid to deionized water in a volume ratio of 3:1). The polishing voltage was 5 V, and the time was 300 s. Mechanical grinding and electropolishing were repeated until the surface roughness Ra and coplanarity were reduced to 10 nm.

[0152] After surface planarization, the copper plating and heat sink on the copper-clad ceramic substrate are subjected to surface activation treatment. The surface activation treatment process is O2 plasma treatment with a power of 150W and a treatment time of 10s.

[0153] The remaining steps are the same as in Example 2, and the weld strength of the obtained welded assembly is 50 MPa.

[0154] Example 6

[0155] This embodiment provides a method for welding copper foil and power chips, which differs from Embodiment 3 in that the surface activity treatment in step 3 is different, as detailed below:

[0156] Step 3: Preparation before welding. The copper plating obtained in Step 2 is subjected to surface planarization treatment. First, electropolishing is performed using a mixed solution of 85wt% phosphoric acid and deionized water (5wt% phosphoric acid to deionized water volume ratio 3:1). The copper foil is placed at the anode and connected to the electropolishing device, with the polishing voltage controlled at 2V. After electropolishing, the roughness is reduced to below 30nm. Next, mirror-finish rolling is performed with a roll roughness of 5nm, a rolling force of 70MPa, and a reduction controlled at 1μm. This reduces the surface roughness of the copper plating on the copper foil to approximately 5nm and the coplanarity to 10nm. A low-speed saw is then used to cut the entire copper foil, including the copper plating, into 2mm*3mm square sheets.

[0157] The copper plating and power chip were then subjected to surface activation treatment using a two-step Ar and N2 plasma activation process. The Ar plasma treatment was performed at a power of 550W for 200 seconds, while the N2 plasma treatment was performed at a power of 450W for 30 seconds. Subsequently, a chemical treatment was carried out using a mixture of ascorbic acid and ethanol. A 1:1 volume ratio mixture of ascorbic acid and ethanol was prepared, and 5 microliters of this mixture were coated onto the surface of the power chip.

[0158] The obtained power module has good welding effect, and the weld is free of holes and cracks.

[0159] In summary, due to the rough surface and very large grains of copper-clad ceramic substrates, traditional processes cannot directly use diffusion welding technology for connection. However, this invention first obtains a flat surface (coplanarity ≤15nm, roughness ≤10nm) through surface planarization treatment to ensure good contact between the upper and lower interfaces. Then, the coating structure is controlled through electrodeposition process and component concentration adjustment to prepare a copper coating with high (111) selectivity (above 65%) and grain size ≤2.5μm. Since the diffusion coefficient of copper atoms on the (111) crystal plane is much higher than that on other crystal planes (e.g., the diffusion coefficient of Cu (111) crystal plane is 1.5 to 3 times that of (100) crystal plane), and the reduction of grain size can also accelerate grain boundary diffusion, the diffusion welding temperature can be reduced and the diffusion time shortened, thereby achieving low-temperature hot-press welding.

[0160] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for welding a power module, characterized in that, Including the following steps: A substrate is provided, the surface of which includes a first metal layer; A workpiece is provided, the surface of which includes a second metal layer; A welding layer is formed on the first metal layer; The surfaces of the weld layer and the second metal layer are subjected to surface activation treatment; The surfaces of the welding layer and the second metal layer are positioned opposite each other and formed into a power module by hot-press bonding. Wherein, the coplanarity of the surface of the first metal layer, the surface of the welding layer and the surface of the second metal layer is ≤15nm and the roughness Ra is ≤10nm; The welding layer is a copper plating layer, comprising: equiaxed crystals and / or columnar crystals; The equiaxed and / or columnar crystals have a grain size ≤ 2.5 μm and a (111) texture ratio of more than 65%.

2. The welding method according to claim 1, characterized in that, The equiaxed crystals and / or columnar crystals contain nanotwins.

3. The welding method according to claim 1, characterized in that, Before the step of setting the welding layer on the first metal layer, the method further includes: performing a surface planarization treatment on the first metal layer; And / or, before the step of performing surface activation treatment on the surface of the weld layer and the surface of the second metal layer, the method further includes: performing surface planarization treatment on the surface of the weld layer and the surface of the second metal layer.

4. The welding method according to claim 3, characterized in that, The surface planarization method is selected from one or more of electropolishing, mechanical grinding, mirror finishing, and chemical mechanical grinding.

5. The welding method according to claim 1, characterized in that, The surface activation treatment method is chemical treatment and / or plasma treatment; The chemical treatment uses small molecules of carboxylic acids and / or small molecules of alcohols; The plasma treatment uses one or more of N2, O2, and Ar.

6. The welding method according to claim 5, characterized in that, The carboxylic acid molecules are selected from one or more of formic acid, acetic acid, citric acid and ascorbic acid; and / or, the alcohol molecules are ethanol or glycerol.

7. The welding method according to claim 1, characterized in that, The hot-press bonding is carried out in a vacuum or ambient atmosphere, wherein the ambient atmosphere is selected from the atmosphere, an inert gas, or a reducing gas; And / or, the temperature of the hot-press bonding is 150–300°C, and the pressure of the hot-press bonding is 10–60 MPa.

8. The welding method according to claim 1, characterized in that, The thickness of the weld layer is 10–100 μm.

9. The welding method according to claim 1, characterized in that, The materials of the first metal layer and the second metal layer are independently selected from one or more of gold, silver, copper, and nickel.

10. A power module, characterized in that, It is obtained by welding using the welding method described in any one of claims 1 to 8.

Images