Ceramic-copper composite and method for manufacturing ceramic-copper composite

Abstract

The ceramic-copper composite (100) of the present application is a flat plate-shaped ceramic-copper composite (100) having a ceramic layer (1), a copper layer (2), and a solder layer (3) present between the ceramic layer (1) and the copper layer (2) and containing Sn or In and Ag, a concave-convex portion being formed on the copper layer (2) side of the solder layer (3), and a plurality of Cu-rich phases (4) being present in a state of being spaced apart from each other in at least one convex portion (6).

Description

TECHNICAL FIELD

[0001] The present application relates to a ceramic-copper composite and a method for manufacturing a ceramic-copper composite. BACKGROUND

[0002] Various developments have been made on ceramic-copper composites so far. As such a technology, for example, the technology described in Patent Literature 1 is known. In Patent Literature 1, regarding a manufacturing method of a ceramic circuit substrate, the joining conditions of performing heat treatment at a temperature of 800°C for 15 minutes in a vacuum atmosphere in a state where Cu plates are pressed and contacted on the front and back surfaces of an aluminum nitride substrate on which a solder paste is printed are described (paragraph 0034 and the like of Patent Literature 1).

[0003] PRIOR ART DOCUMENTS

[0004] PATENT LITERATURE

[0005] Patent Literature 1: Japanese Patent Application Laid-Open No. 2005-101415 SUMMARY

[0006] PROBLEMS TO BE SOLVED BY THE INVENTION

[0007] However, the inventors of the present application have found out that the ceramic-copper composite described in Patent Literature 1 has room for improvement in thermal cycle characteristics.

[0008] MEANS FOR SOLVING THE PROBLEMS

[0009] The inventors of the present application have further conducted research and as a result, have found that by forming a plurality of Cu-rich phases in a region of the solder layer that becomes convex toward the copper layer side, it is possible to improve and enhance the thermal cycle characteristics of the ceramic-copper composite, thereby completing the present application.

[0010] According to the present application, there is provided a ceramic-copper composite which is a flat plate-shaped ceramic-copper composite, has a ceramic layer, a copper layer, and a solder layer, the solder layer is present between the ceramic layer and the copper layer and contains Sn or In and Ag,

[0011] In at least one of the cross sections when the ceramic-copper composite is cut along a plane perpendicular to the main surface thereof, a concave-convex portion is formed on the copper layer side of the solder layer, and in at least one convex portion, a plurality of Cu-rich phases exist in a state of being spaced apart from each other.

[0012] Further, according to the present application, there is provided a manufacturing method of a flat plate-shaped ceramic-copper composite, which includes:

[0013] a step of preparing a flat plate-shaped laminate in which a ceramic layer, a solder containing Ag, and a copper layer are laminated; and

[0014] The joining step is performed by heating the layered body at a temperature of 750°C to 900°C at a temperature increase rate of 20°C / min or more and 150°C / min or less under a nitrogen atmosphere,

[0015] The flat ceramic-copper composite has the ceramic layer, the copper layer, and a solder layer containing Ag, which is present between the ceramic layer and the copper layer.

[0016] Effects of the Invention

[0017] According to the present invention, a ceramic-copper composite having excellent thermal cycle characteristics and a method for producing the same can be provided. BRIEF DESCRIPTION OF DRAWINGS

[0018] [ Figure 1 ] is a joining cross-sectional view of an outer portion in one example of the ceramic-copper composite according to the present embodiment.

[0019] [ Figure 2 ] is an enlarged view of an α region of Figure 1 .

[0020] [ Figure 3 ] is an enlarged view of a β region of Figure 2 .

[0021] [ Figure 4 ] is an SEM image of the ceramic-copper composite of Example 1.

[0022] [ Figure 5 ] is an SEM image of the ceramic-copper composite of Example 1.

[0023] [ Figure 6 ] is a binarized image of the SEM image of the ceramic-copper composite of Example 1.

[0024] [ Figure 7 ] is an SEM image of the ceramic-copper composite of Example 2.

[0025] [ Figure 8 ] is an SEM image of the ceramic-copper composite of Example 2.

[0026] [ Figure 9 ] is a binarized image of the SEM image of the ceramic-copper composite of Example 2.

[0027] [ Figure 10 ] is an SEM image of the ceramic-copper composite of Example 3.

[0028] [ Figure 11 ] is an SEM image of the ceramic-copper composite of Example 3.

[0029] [ Figure 12A binarized image of the SEM image of the ceramic-copper composite of Example 3.

[0030] [ Figure 13 An SEM image of the ceramic-copper composite of Example 4.

[0031] [ Figure 14 An SEM image of the ceramic-copper composite of Example 4.

[0032] [ Figure 15 A binarized image of the SEM image of the ceramic-copper composite of Example 4.

[0033] [ Figure 16 An SEM image of the ceramic-copper composite of Example 5.

[0034] [ Figure 17 An SEM image of the ceramic-copper composite of Example 5.

[0035] [ Figure 18 A binarized image of the SEM image of the ceramic-copper composite of Example 5.

[0036] [ Figure 19 An SEM image of the ceramic-copper composite of Comparative Example 1.

[0037] [ Figure 20 An SEM image of the ceramic-copper composite of Comparative Example 1.

[0038] [ Figure 21 A binarized image of the SEM image of the ceramic-copper composite of Comparative Example 1.

[0039] [ Figure 22 An SEM image of the ceramic-copper composite of Comparative Example 2.

[0040] [ Figure 23 An SEM image of the ceramic-copper composite of Comparative Example 2.

[0041] [ Figure 24 A binarized image of the SEM image of the ceramic-copper composite of Comparative Example 2. DETAILED DESCRIPTION

[0042] Hereinafter, embodiments of the present application will be described with reference to the drawings. Note that the same components are denoted by the same reference numerals throughout the drawings, and description thereof will not be repeated. In addition, the drawings are schematic views, and the dimensional ratios are not necessarily the same as actual ones.

[0043] Note that in the present embodiment, the directions of front, back, left, right, up and down are defined as illustrated in the drawings. However, this is defined for the sake of convenience in simply explaining the relative relationship of the constituent elements. Therefore, the directions at the time of manufacturing and at the time of use of the product implementing the present application are not limited.

[0044] The ceramic-copper composite of the present embodiment is outlined.

[0045] The ceramic-copper composite is a flat plate-shaped member that has a ceramic layer, a copper layer, and a solder layer, the solder layer being present between the ceramic layer and the copper layer and containing Sn or In and Ag.

[0046] In at least one of the cut surfaces when the ceramic-copper composite is cut along a surface perpendicular to the main surface thereof, a concavo-convex portion is formed on the copper layer side of the solder layer, and in at least one convex portion, a plurality of Cu-rich phases exist in a state of being spaced apart from each other.

[0047] According to the present inventors' insight, it was found that by forming a plurality of Cu-rich phases in the region of the solder layer that becomes convex toward the copper layer side, the thermal cycle characteristics of the ceramic-copper composite can be improved.

[0048] Although the detailed mechanism is not certain, it is believed that by the presence of a plurality of Cu-rich phases in the convex portion, the peeling strength at the interface between the solder layer and the copper layer is improved, and as a result, the occurrence of peeling therebetween after a thermal cycle test is suppressed, and thus the thermal cycle characteristics are improved.

[0049] In addition, with such a structure, it is possible to obtain by appropriately selecting the joining conditions of the ceramic-copper composite, such as, for example, under a nitrogen atmosphere, using a condition of rapid heating and rapid cooling, setting the heating peak temperature at a high level, and the like. Although the detailed mechanism is not certain, it is presumed that the interface of the Ag-rich phase of the solder layer moves toward the inside of the copper layer, a convex-shaped Ag-rich phase is formed, and inside the convex portion composed of the Ag-rich phase, Cu atoms diffused from the copper layer or Cu atoms contained in the solder form a plurality of Cu-rich phases.

[0050] By using the ceramic-copper composite of the present embodiment to form a circuit pattern on the copper layer, it is possible to realize a ceramic circuit substrate having a copper heat sink and a copper circuit board on both sides of the ceramic layer via the solder layer. Such a ceramic circuit substrate can improve the thermal cycle characteristics.

[0051] By using such a ceramic circuit substrate, it is possible to realize an electronic component module that has a ceramic circuit substrate, an electronic component provided on the copper circuit board of the ceramic circuit substrate, and a heat sink provided on the copper heat sink of the ceramic circuit substrate. By using the above-described ceramic circuit substrate, it is possible to improve the connection reliability of the electronic component module.

[0052] The ceramic-copper composite 100 can be applied to various uses, and as one of the uses, can be applied to a power module for a vehicle that requires strict reliability.

[0053] Hereinafter, the ceramic-copper composite of the present embodiment will be described in detail.

[0054] Figure 1 is an enlarged view of the α region of Figure 2 is an enlarged view of the α region of Figure 1 is an enlarged view of the β region of Figure 3 is an enlarged view of the β region of Figure 2 is an enlarged view of the β region of Figure 1 In the drawing, black color represents the ceramic layer 1, gray color represents the copper layer 2, and white color represents the solder layer 3. Figure 3 In the drawing, in the solder layer 3, white color represents the Ag-rich phase 5, and gray color represents the Cu-rich phase 4.

[0055] Figure 1 The ceramic-copper composite 100 shown in the drawing has a structure in which the ceramic layer 1, the solder layer 3, and the copper layer 2 are sequentially stacked.

[0056] As the ceramic layer 1, for example, nitride-based ceramics such as silicon nitride and aluminum nitride, oxide-based ceramics such as alumina and zirconia, carbide-based ceramics such as silicon carbide, and boride-based ceramics such as lanthanum boride can be used. Among them, from the viewpoint of metal joint property, non-oxide-based ceramics such as aluminum nitride and silicon nitride are suitable, and from the viewpoint of excellent mechanical strength and breaking toughness, silicon nitride is further preferred.

[0057] The thickness of the ceramic layer 1 is not particularly limited, and is usually around 0.1 mm to 3.0 mm, and particularly in view of heat dissipation property and thermal resistance reduction, it is preferably 0.2 mm or less, and more preferably 0.25 mm or less to 1.0 mm or less.

[0058] In the present specification, "~" indicates that the upper limit value and the lower limit value are included, unless otherwise specified.

[0059] The copper layer 2 is not particularly limited, and for example, can be formed of a copper plate. The material used in the copper plate is preferably pure copper. The thickness of the copper plate is not particularly limited, and is usually 0.1 mm to 1.5 mm, and particularly in view of heat dissipation property, it is preferably 0.3 mm or more, and more preferably 0.5 mm or more.

[0060] The method of forming a circuit pattern on the copper layer 2 can use, for example, a method of forming an etching mask and performing etching treatment after joining a metal plate (for example, a copper plate) to the ceramic layer 1 using the solder layer 3, thereby forming a circuit pattern.

[0061] The solder layer 3 is composed of a solder containing Sn or In and Ag.

[0062] The filler material can be constituted of at least any one of: Ag and Sn; Ag and In; or Ag, Sn, and In. The filler material can contain other elements in addition to these elements, and for example, can contain Cu, Ti, or the like.

[0063] At least one of the cross sections of the ceramic-copper composite 100 when cut along a face perpendicular to the main face thereof Figures 1-3 ) is formed with a concave-convex portion on the copper layer 2 side of the filler material layer 3. It is constituted in such a manner that a plurality of Cu-rich phases 4 exist in a state of being spaced apart from each other within at least one convex portion 6 in the concave-convex portion.

[0064] Figures 1-3 The cross section of the ceramic-copper composite 100 when cut along a face perpendicular to the main face thereof

[0065] The convex portion 6 of the filler material layer 3 can be defined, for example, in a cross-sectional view, as a portion that exceeds a position of 1 / 2 of the average thickness of the filler material layer 3 in a perpendicular direction from the interface of the ceramic layer 1 and the filler material layer 3 toward the copper layer 2 side. In addition, the concave portion of the concave-convex portion of the filler material layer 3 can be defined as a portion that does not exceed a position of 1 / 2 of the average thickness of the filler material layer 3.

[0066] Here, regarding the average thickness of the filler material layer 3, the thickness from the interface of the ceramic layer 1 and the filler material layer 3 to the interface of the filler material layer 3 and the copper layer 2 can be measured by image observation described below, using the average of the maximum thickness and the minimum thickness thereof.

[0067] The image observation in the present embodiment can be performed as follows.

[0068] Using a scanning electron microscope, a SEM image is obtained at an arbitrary position of the joint interface of the cross section of the ceramic-copper composite. Using the obtained SEM image, a 40 μm x 60 μm region of 3 fields of view is observed arbitrarily in the vicinity of the center of the central plate thickness cross section in the long side direction.

[0069] Regarding the values calculated from the observation fields of view, an average value can be used that averages the values measured by the observation of the 3 fields of view.

[0070] By the image observation described above, the length in the plate width direction of the observation field of view, the length of the line segment of the interface portion of the filler material layer 3 and the copper layer 2 are found, and as shown in Figure 3 , are set to L1, L2, respectively.

[0071] At this time, the ratio of L2 to L1 (L2 / L1) can be, for example, 1.25 to 5.0, more preferably 1.30 to 3.0. It is presumed that by making L2 / L1 be the above lower limit value or more, the interface joint of the filler material layer 3 and the copper layer 2 is improved, and thus the thermal cycle characteristics can be improved.

[0072] The solder layer 3 can have a plurality of protrusions 6 that intrude into the interior of the copper layer 2. The plurality of protrusions 6 can be formed at least near the end portion of the ceramic-copper composite 100, and can be formed both near the end portion and in the interior.

[0073] In addition, the number of protrusions 6 of the solder layer 3 is not particularly limited, and can be configured so that the number of protrusions is greater in the outer region near the end portion than in the interior. Thus, when the ceramic-copper composite 100 is viewed from a direction perpendicular to the main surface, peeling, cracking, and the like during a thermal cycle test can be suppressed near the outer periphery.

[0074] The Cu-rich phase 4 present in the protrusions 6 of the solder layer 3 can be configured in various shapes such as a substantially spherical shape, a substantially elliptical shape, and the like in a cross-sectional view, and at least one or more of them can be formed in a striped shape. They can be used alone, or two or more of them can be used in combination. In addition, a plurality of Cu-rich phases 4 in a striped shape can be present in the same protrusion 6, and they can also be arranged at intervals from each other to form a striped pattern.

[0075] The aspect ratio of one Cu-rich phase 4 configured in a striped shape can be, for example, greater than 1 and 10 or less, and preferably 2 or more and 9 or less. Thus, the thermal cycle characteristics can be improved.

[0076] The number density of the Cu-rich phase 4 in the protrusions 6 of the solder layer 3 is, for example, 0.1 pieces / μm 2 or more and 20 pieces / μm 2 or less. It is preferable to be 0.2 pieces / μm 2 or more and 10 pieces / μm 2 or less. By making the number density of the Cu-rich phase 4 be the above lower limit value or more, the thermal cycle characteristics can be improved.

[0077] Regarding the number density of the Cu-rich phase 4, a graph obtained by binarizing the SEM image used in the above image observation using the image analysis software GIMP2 (threshold value 90) is used, the area of the solder layer 3 is measured using the image analysis software Image-pro plus, and the number density of the Cu-rich phase 4 (pieces / μm 2 ) is calculated using the following equation.

[0078] Equation: [Number density of Cu-rich phase 4] = [Number of Cu-rich phase 4] / [Area of solder layer 3]

[0079] The Cu-rich phase in the solder layer 3 can include, for example, a fine phase of 0.3 μm or more and less than 1.0 μm, and a coarse phase of 1.0 μm or more and 7.5 μm or less. By the presence of the fine phase and the coarse phase, the Cu-rich phase 4 can be present in the protrusions 6, and the number density of the Cu-rich phase 4 in the protrusions 6 can be improved.

[0080] The solder layer 3 only needs to be configured to have at least a plurality of Cu-rich phases 4 in the protrusion 6, and the area outside the protrusion 6 may or may not have Cu-rich phases 4.

[0081] The solder layer 3 is composed of an Ag-rich phase 5. The Ag-rich phase 5 can be continuously formed in the protrusions 6 of the solder layer 3. Inside the Ag-rich phase 5 present in the protrusions 6, multiple Cu-rich phases 4 can be discontinuously distributed.

[0082] Furthermore, at least a portion of the outer edge of the protrusion 6, i.e., the portion of the protrusion 6 that contacts the interface between the solder layer 3 and the copper layer 2, the Ag-rich phase 5 can be formed, or it can be formed continuously. That is, in the protrusion 6, the Cu-rich phase 4 can be discretely present inside the region surrounded by the Ag-rich phase 5.

[0083] Cu-rich phase 4 can be defined as a Cu-dominant solid solution. The Cu content in Cu-rich phase 4 of protrusion 6 can be, for example, 80% by mass or more.

[0084] In the Ag-rich phase 5, Ag-rich can be defined as a solid solution mainly containing Ag. The Ag content in the Ag-rich phase 5 of the protrusion 6 can be, for example, 80% by mass or more.

[0085] The content ratio of Cu and Ag was quantitatively analyzed using an electron beam excited X-ray analyzer (EPMA).

[0086] The diffusion distance of Ag, which is a component of the solder layer 3, is, for example, 10 μm to 50 μm, preferably 15 μm to 45 μm.

[0087] Here, the diffusion distance of Ag can be defined as the distance between the surface of ceramic layer 1 and the furthest part of Ag that diffuses from the surface of ceramic layer 1 to the surface of copper layer 2 (in a direction perpendicular to the main surface of the ceramic-copper composite 100). The diffusion distance of Ag is not necessarily consistent with the thickness of the continuous solder layer.

[0088] The diffusion distance of Ag is determined by observing three randomly selected 500x magnification fields (within a 250μm range in the horizontal direction of the bonding interface) on a cross-section (the thickest section at the center of the long side) of the ceramic-copper composite 100 using a scanning electron microscope. The maximum diffusion distance of Ag measured in each field is the largest distance.

[0089] Additionally, the ceramic-copper composite 100 may have a Ti-containing bonding layer existing between the ceramic layer 1 and the solder layer 3.

[0090] The bonding porosity of such a ceramic-copper composite 100 can be configured to be, for example, 1.0% or less. This helps to suppress the delamination of the copper plate during thermal cycling.

[0091] Alternatively, the solder layer 3 may be configured without a climbing section formed by a portion of it climbing along the side of the copper layer 2.

[0092] Next, the manufacturing method of the ceramic-copper composite 100 will be described.

[0093] One method for manufacturing a ceramic-copper composite 100 includes: a step of preparing a flat laminate in which a ceramic layer, a solder containing Ag, and a copper layer are stacked; and a bonding step for the laminate. Thus, a flat ceramic-copper composite 100 is obtained, the ceramic-copper composite 100 comprising a ceramic layer 1, a copper layer 2, and a solder layer 3, the solder layer 3 being present between the ceramic layer 1 and the solder layer 3 and containing Ag.

[0094] Regarding the Ag / Cu ratio used in solder bonding, increasing the Ag powder proportion compared to the 72% by mass:28% eutectic composition of Ag and Cu prevents the coarsening of the Cu-rich phase, resulting in a continuous Ag-rich solder layer structure. Furthermore, if the Ag powder content is high and the Cu powder content is low, the Ag powder will not completely dissolve during bonding, leaving behind gaps. Additionally, the Sn or In in the solder powder reduces the contact angle between the solder and the ceramic substrate and improves the solder's wettability. Insufficient Sn or In reduces wettability with the ceramic substrate, potentially leading to poor bonding; excessive In In causes discontinuity of the Ag-rich phase in the solder layer due to the Cu-rich phase, becoming the starting point for solder fracture and potentially reducing the thermal cycling characteristics of the ceramic circuit board.

[0095] Therefore, the following mixing ratios can be used for Ag powder, Cu powder, and Sn powder or In powder: Ag powder is 85.0 to 95.0 parts by weight, preferably 88.0 to 92.0 parts by weight, more preferably 88.5 to 91.0 parts by weight; Cu powder is 5.0 to 13.0 parts by weight, preferably 6.0 to 12.0 parts by weight, more preferably 7.0 to 11.0 parts by weight; and Sn powder or In powder is 0.4 to 2.0 parts by weight, preferably 0.5 to 1.5 parts by weight.

[0096] As a method for mixing brazing filler metal raw materials, it is preferable to combine metal powder with organic solvents and binders, and mix them using a sand mixer, a rotary mixer, a planetary mixer, a three-roll mill, etc., to form a paste. Typically, organic solvents such as methyl cellosolve, ethyl cellosolve, isophorone, toluene, ethyl acetate, terpineol, diethylene glycol monobutyl ether, and texanol can be used, while binders such as polyisobutyl methacrylate, ethyl cellulose, methyl cellulose, and acrylic resins can be used.

[0097] Methods for applying solder paste to both sides of a ceramic substrate include roller coating, screen printing, and transfer printing. Screen printing is preferred for uniform solder coating. In screen printing, to ensure uniform solder paste coating, the viscosity of the solder paste is preferably controlled to be 5–20 Pa·s. By adjusting the amount of organic solvent in the solder paste to 5–17% by mass and the binder content to 2–8% by mass, a solder paste with excellent printability can be obtained.

[0098] In the above-described bonding process, ceramic layer 1 and solder layer 3 are bonded together using solder. In this bonding process, the laminate can be heated under a nitrogen atmosphere at a heating rate of 20°C / min to 150°C / min, and then heated at a temperature of 750°C to 900°C.

[0099] The lower limit of the above-mentioned heating rate is 20°C / minute or more, preferably 25°C / minute or more, and more preferably 30°C / minute or more. It should be noted that the heating rate can be set as the average heating rate from 20°C to 750°C.

[0100] In addition, the peak temperature of the heating temperature in the joining process can be 750°C or higher, preferably 800°C or higher, and more preferably 820°C or higher.

[0101] Methods for bonding ceramic circuit boards under a vacuum atmosphere are generally known.

[0102] However, rapid heating above 20°C / minute is not possible in a vacuum atmosphere; the typical heating rate is 5°C / minute.

[0103] In contrast, in this embodiment, by selecting nitrogen from inert gases as the heat medium, the aforementioned rapid heating can be achieved.

[0104] In addition, in the bonding process, the laminate can be cooled at a cooling rate of 10°C / min to 100°C / min after heating.

[0105] The lower limit of the aforementioned cooling rate is, for example, 10°C / minute or more, preferably 13°C / minute or more, and more preferably 15°C / minute or more. It should be noted that the cooling rate can be set as the average cooling rate from 750°C to 20°C.

[0106] In the above-mentioned joining process, the oxygen concentration of the nitrogen atmosphere is, for example, 500 ppm or less, preferably 200 ppm or less, and more preferably 120 ppm or less. This ensures a good connection at the ends of the ceramic-copper composite 100.

[0107] In this embodiment, as an example, the heating in the joining process can be performed using a nitrogen furnace to heat the structure obtained by overlapping two or more flat laminates. By using a nitrogen furnace, the aforementioned rapid heating, rapid cooling, and nitrogen atmosphere conditions can be appropriately controlled.

[0108] In order to form a circuit pattern on the copper layer 2 of the ceramic-copper composite 100, an etching resist can be applied to the copper layer 2 and then etched.

[0109] There are no particular restrictions on the etching resist; for example, commonly used UV-curable or thermosetting etching resists can be used. There are also no particular restrictions on the coating method of the etching resist; known coating methods such as screen printing can be employed.

[0110] Electroless Ni plating, Au flash plating, or displacement-type Ag plating can be applied to the surface of the copper layer 2 with the circuit pattern as needed. Alternatively, the surface can be smoothed by grinding, physical polishing, or chemical polishing without electroplating, and then a rust inhibitor can be applied.

[0111] The embodiments of the present invention have been described above, but these are merely examples, and various configurations other than those described can be employed. Furthermore, the present invention is not limited to the embodiments described above; modifications and improvements within the scope of achieving the objectives of the present invention are included in the present invention.

[0112] Example

[0113] The present invention will now be described in detail with reference to embodiments, but the present invention is not limited in any way by these embodiments.

[0114] <Fabrication of Ceramic-Copper Composites>

[0115] (Example 1)

[0116] On the two main surfaces of a 0.32 mm thick silicon nitride substrate, a coating of 8 mg / cm² was applied using screen printing. 2 The active metal solder is coated in a manner that, relative to Ag powder (manufactured by Fukuda Metal Foil Powder Industry Co., Ltd.: Ag-HWQ), has an average particle size D50 of 2.5 μm and a specific surface area of ​​0.4 m². 2 89.5 parts by weight (g), Cu powder (manufactured by Fukuda Metal Foil Powder Industry Co., Ltd.: Cu-HWQ, average particle size D50: 3.0 μm, specific surface area 0.4 m²). 2 9.5 parts by weight (g), Sn powder (manufactured by Fukuda Metal Foil Powder Industry Co., Ltd.: Sn-HPN, average particle size D50: 3μm, specific surface area 0.1m²). 2For a total of 100 parts by weight of 1.0 parts by weight, it contains 3.5 parts by weight of titanium hydride powder (manufactured by TOHO TECHNICAL Co., Ltd.: TCH-100).

[0117] Then, a circuit forming metal plate is superimposed on one side of the silicon nitride substrate, and a heat sink forming metal plate (both are C1020 oxygen-free copper plates with a thickness of 0.5 mm) is superimposed on the other side to obtain a laminate.

[0118] Using a nitrogen heating furnace, the obtained laminate is heated from 25°C to 870°C at a heating rate of 90°C / min in a nitrogen atmosphere with an oxygen concentration of 100 ppm, held at 870°C (bonding temperature) for 18 minutes (bonding time), and then cooled to 25°C at a cooling rate of 15°C / min (bonding process).

[0119] An etching resist is printed on the bonded copper plate, and a circuit pattern is formed by etching with ferric chloride solution to obtain a ceramic-copper composite.

[0120] (Examples 2-5)

[0121] The bonding process was changed to the bonding conditions in Table 1, and otherwise operated in the same manner as in Example 1 to obtain a ceramic-copper composite.

[0122] (Compare Examples 1 and 2)

[0123] In the heating furnace, at 1.0 × 10 -3 In a vacuum below Pa, the bonding process was changed to the bonding conditions in Table 1. Otherwise, the process was carried out in the same manner as in Example 1 to obtain a ceramic-copper composite.

[0124] [Table 1]

[0125]

[0126] The obtained ceramic-copper composite was evaluated based on the following evaluation criteria.

[0127] (SEM image observation)

[0128] Cross-sections of the ceramic-copper composites of Examples 1-5 and Comparative Examples 1 and 2 were observed using a scanning electron microscope to obtain SEM images. In the SEM images, black represents the silicon nitride substrate, white represents the solder layer, and gray represents the copper plate.

[0129] In Table 1, Cu-rich tissue 1 represents the image obtained by binarizing a 500x magnified SEM image using the image analysis software GIMP2 (threshold 90), Cu-rich tissue 2 represents the image obtained by binarizing a 2000x magnified SEM image using the image analysis software GIMP2 (threshold 90), and Cu-rich tissue 3 represents the image obtained by binarizing a 2000x magnified SEM image using the image analysis software GIMP2 (threshold 90).

[0130] <Observation of Cu-rich phase in solder layer microstructure>

[0131] The Cu-rich phase in the solder layer microstructure was evaluated as follows: Using a scanning electron microscope (JEOL JSM-6380), a 40 μm x 60 μm field of view was observed at any position on the bonding interface of the ceramic-copper composite cross-section, at 2000x magnification, in three fields of view near the center of the thick cross-section along the long side. Using this method, Cu-rich phases with a particle size of 0.1 μm or larger could be observed. Regarding the magnification, if the magnification is too high, the field of view narrows, making it impossible to observe a sufficient number of Cu-rich phases; conversely, if the magnification is too low, Cu-rich phases smaller than 1 μm cannot be observed. Therefore, 2000x was set as the magnification.

[0132] (Tissue shape)

[0133] Based on the above image observations, the presence or absence of unevenness in the solder layer structure and the presence or absence of Cu-rich phases within the uneven areas are evaluated according to the following criteria.

[0134] Regarding the average thickness of the solder layer, the thickness from the interface between the silicon nitride substrate and the solder layer to the interface between the solder layer and the copper plate was measured, and the average of the maximum and minimum thicknesses was used. Measurements were performed in three fields of view, and the average of the three fields of view was used.

[0135] In the region of the solder layer shown in the SEM image, a portion extending beyond 1 / 2 of the average thickness of the solder layer in a vertical direction from the interface between the silicon nitride substrate and the solder layer toward the copper plate side is judged as having a protrusion in the solder layer, and a portion extending beyond 1 / 2 of the average thickness of the solder layer is judged as having a recess in the solder layer.

[0136] In addition, when there are protrusions in the solder layer, the Cu-rich phase is observed in the internal region of the protrusion, and its presence and shape are evaluated.

[0137] (Fine phase, coarse phase)

[0138] Based on the above image observations, the presence or absence of fine phases (greater than 0.3 μm and less than 1.0 μm) and coarse phases (greater than 1.0 μm and less than 7.5 μm) was investigated. The results are shown in Table 1.

[0139] In addition, the proportion of Cu in the Cu-rich phase was quantitatively analyzed using an electron beam excited X-ray analysis apparatus (EPMA).

[0140] (L1, L2)

[0141] Based on the above image observations, the length L1 of the board width direction in the field of view and the length L2 of the line segment at the interface between the solder layer and the copper layer were measured, and the ratio L1 / L2 was calculated. The results are shown in Table 1.

[0142] <Evaluation of Cu-rich Phases in Solder Layer Microstructure>

[0143] The SEM images obtained by the above method were binarized using the image analysis software GIMP2 (threshold 90). The resulting images were then analyzed to determine the aspect ratio and number density of Cu-rich phases.

[0144] The software used for image analysis is Image-ProPlus, an image processing software manufactured by MediaCybernetics.

[0145] Regarding the Cu-rich phase in the solder layer microstructure, the grain size varies widely, ranging from 0.3 μm to several μm. The fine Cu-rich phases with grain sizes greater than 0.1 μm and less than 1.0 μm, observable by scanning electron microscopy, were analyzed based on aspect ratio and number density. To assess the average size of the Cu-rich phases, the centroid diameter of all observed Cu-rich phases was measured in three fields of view and taken as the average value.

[0146] In addition, regarding the number density of Cu-rich phases, the area of ​​the solder layer structure was measured using Image-pro plus image analysis software manufactured by MediaCybernetics, and the number density of Cu-rich phases was calculated using the following formula (I).

[0147] Cu-rich phase number density (units / μm) 2 = Number of Cu-rich phases / Area of ​​solder layer structure ... (I)

[0148] (Ag diffusion distance)

[0149] After cutting the ceramic-copper composite, embedding it in resin, and grinding the cross-section, the reflected electron images of three random fields of view (a range of 250 μm in the horizontal direction of the bonding interface) were captured at 500x magnification using a scanning electron microscope. The shortest distance (μm) between the silicon nitride substrate and the position of Ag closest to the surface of the copper plate was measured.

[0150] (Joint porosity)

[0151] The area of ​​the bonding voids in the ceramic-copper composite, as observed by an ultrasonic flaw detector (manufactured by Hitachi Power Solution Co., Ltd.: ES5000), was measured and divided by the area of ​​the circuit pattern to calculate the bonding void ratio (area %).

[0152] (Thermal cycling test)

[0153] The ceramic-copper composite was held on a hot plate at 350°C for 5 minutes, then at 25°C for 5 minutes, then in a mixed solvent of ethanol and dry ice at -78°C for 5 minutes, then at 25°C for 5 minutes. This thermal cycle was repeated for 15 cycles.

[0154] Then, the copper plate, solder layer and nitride layer are removed by etching. The horizontal cracks generated on the surface of the ceramic substrate are scanned and stored in the database using a scanner at a resolution of 600dpi×600dpi. The horizontal crack area is divided by the circuit pattern area to calculate the crack rate (%) after thermal cycling.

[0155] Compared with Comparative Examples 1 and 2, the ceramic-copper composites of Examples 1 to 5 showed a lower crack rate after thermal cycling in the thermal cycling test, thus demonstrating excellent thermal cycling characteristics.

[0156] This application claims priority based on Japanese Patent Application No. 2020-009305, filed on January 23, 2020, the entire disclosure of which is incorporated herein by reference.

[0157] Explanation of reference numerals in the attached figures

[0158] 1. Ceramic layer

[0159] 2 Copper Layer

[0160] 3. Brazing filler layer

[0161] 4 Cu-rich phase

[0162] 5 Ag-rich phase

[0163] 6 convex part

[0164] 100 Ceramic-Copper Composite

Claims

1. A ceramic-copper composite, which is a flat ceramic-copper composite having a ceramic layer, a copper layer, and a solder layer, wherein the solder layer exists between the ceramic layer and the copper layer and contains Sn or In and Ag. In at least one of the cut surfaces when the ceramic-copper composite is cut along a plane perpendicular to its main surface, The Ag in the solder layer moves into the interior of the copper layer, forming an uneven portion on the copper layer side of the solder layer. The protrusion in the above-mentioned uneven portion is the portion extending beyond 1 / 2 of the average thickness of the solder layer in a vertical direction from the interface between the ceramic layer and the solder layer toward the copper layer side. Within at least one of the protrusions, multiple Cu-rich phases exist spaced apart from each other within the region surrounded by Ag-rich phases. The multiple Cu-rich phases are arranged in a striped pattern, spaced apart from each other. The Cu content in the Cu-rich phase present in the protrusion is 80% by mass or more. The Ag content in the Ag-rich phase present in the protrusion is 80% by mass or more.

2. The ceramic-copper composite as described in claim 1, wherein, The aspect ratio of the Cu-rich phase is greater than 1 and less than 10.

3. The ceramic-copper composite as described in claim 1 or 2, wherein, The number density of the Cu-rich phase is 0.1 phases / μm. 2 Above 20 / μm 2 the following.

4. The ceramic-copper composite as described in claim 1 or 2, wherein, In the field of view of the cross section, the ratio (L2 / L1) of the length L2 of the line segment at the interface between the solder layer and the copper layer to the length L1 in the width direction of the field of view is 1.25 or more and 5.0 or less.

5. The ceramic-copper composite as described in claim 1 or 2, wherein, The Cu-rich phase comprises a fine phase of 0.3 μm or larger and less than 1.0 μm, and a coarse phase of 1.0 μm or larger and less than 7.5 μm.

Images