Method for producing ceramic circuit board

Abstract

A method for producing a ceramic circuit board according to an embodiment of the present invention includes: an application step for applying a resist for imparting a pattern shape to a metal plate of a bonded body in which a ceramic substrate and the metal plate are bonded by the intermediary of a bonding layer; a first etching step for etching the metal plate a plurality of times in the depth direction so as to expose the bonding layer; a second etching step for etching the bonding layer exposed by the first etching step; and a third etching step for etching a side surface of the metal plate formed by the first etching step so as to form a bonding layer protruding portion. In the application step, it is preferable that the resist is applied so as to satisfy 0.2P ≤ W ≤ 0.6P, where P is the shortest distance between metal plates adjacent to each other after the third etching step and W is the gap between resists applied to the metal plates.

Description

Manufacturing method for ceramic circuit board

[0001] The embodiments described below generally relate to a method for manufacturing a ceramic circuit board.

[0002] In recent years, as industrial equipment has become more sophisticated, the power modules mounted on them have become increasingly powerful. This has led to an increase in the power output of semiconductor elements. The guaranteed operating temperature for semiconductor elements is 125°C to 150°C, but this may rise to 175°C or higher in the future. Ceramic circuit boards are used as circuit boards on which semiconductor elements are mounted. Examples of ceramic substrates include silicon nitride substrates, aluminum nitride substrates, and aluminum oxide substrates.

[0003] For example, Japanese Patent No. 6789955 (Patent Document 1) discloses a ceramic circuit board in which a ceramic substrate and a metal plate are bonded via a bonding layer. In the ceramic circuit board described in Patent Document 1, the size and hardness of the protruding portion of the bonding layer are controlled. Furthermore, Japanese Patent No. 7332588 (Patent Document 2) discloses a ceramic circuit board in which the width of the inclined portion is 0.1 to 0.5 times the thickness of the metal plate. As a result, Patent Documents 1 and 2 provide ceramic circuit boards with excellent characteristics in a temperature cycling test (TCT).

[0004] Patent Document 6 describes a metal-ceramic circuit board in which a recess is formed in the lower portion of the metal circuit board, carved out so as to curve inward on the side surface of the circuit pattern, and the distance between the upper ends of the linear portions of the opposing side surfaces of adjacent circuit patterns is greater than the distance between the lower ends of the linear portions. Ｂ The metal-ceramic circuit board has a ratio of the thickness T of the circuit pattern to the thickness T of the circuit pattern of 0.3 or more, more preferably 0.8 or more, and most preferably 1 or more and 2 or more.

[0005] An etching process is used to control the size of the protruding portion of the bonding layer and the slope shape of the side surface of the metal plate. For example, etching processes are shown in Japanese Patent No. 7278215 (Patent Document 3) and Japanese Patent No. 7596515 (Patent Document 5). In Patent Document 3, etching is performed by combining the etching process of the bonding layer containing Ag and Cu, the etching process of the titanium nitride layer, and chemical polishing. In Patent Document 5, etching is performed on the bonding layer in which the brazing material does not contain Ag.

[0006] Japanese Patent No. 6789955 Japanese Patent No. 7332588 Japanese Patent No. 7278215 Japanese Patent Laid-Open No. 2007-134563 Japanese Patent No. 7596515 Japanese Patent No. 6566586

[0007] There is a demand for miniaturization of semiconductor modules. To miniaturize modules, it is also necessary to miniaturize ceramic circuit substrates. Reducing the spacing between metal plates is an effective way to achieve miniaturization while still ensuring the mounting area for semiconductor elements, etc. However, when attempting to impart a specified inclined structure to the side of the metal plate in order to improve the TCT characteristics, it was difficult to reduce the spacing between the metal plates.

[0008] For example, Japanese Patent Laid-Open Publication No. 2007-134563 (Patent Document 4) discloses a method for manufacturing a ceramic circuit board in which metal plates that have been half-etched in advance are joined. In Patent Document 4, the gap between the metal plates can be narrowed by etching the areas that have been half-etched in advance. However, since the metal plates must be half-etched in advance, it cannot necessarily be said that mass production is good. As such, conventional etching processes have not been able to achieve both the inclined structure on the side surfaces of the metal plates and the narrow gap between the metal plates.

[0009] The problem to be solved by the present invention is to provide a method for manufacturing a ceramic circuit board in which the intervals between the metal plates are narrowed.

[0010] The manufacturing method of the ceramic circuit board according to the embodiment is characterized by comprising a coating step of applying a resist to impart a pattern shape to the metal plate of a bonded body in which a ceramic substrate and a metal plate are bonded via a bonding layer; a first etching step of etching the metal plate multiple times in the depth direction to expose the bonding layer; a second etching step of etching the bonding layer exposed by the first etching step; and a third etching step of etching the side surface of the metal plate formed by the first etching step to form an overhanging portion of the bonding layer.

[0011] FIG. 1 is a side cross-sectional view showing a bonded body and the like to explain an example of an etching step in a manufacturing method for a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing an example of a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing another example of a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing an example of an inclined portion of a front metal plate in a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing an example of a contact point of an inclined portion of a front metal plate in a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing a protrusion width of a front metal plate in a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing another example of an inclined portion of a front metal plate in a ceramic circuit board according to an embodiment. FIG. 1 is a side cross-sectional view showing an example of a front metal plate having a step portion in a ceramic circuit board according to an embodiment. FIG. 2 is a side cross-sectional view showing another example of a front metal plate in a ceramic circuit board according to an embodiment. Embodiment

[0012] The manufacturing method of the ceramic circuit board according to the embodiment is characterized by comprising: a coating step of applying a resist to impart a pattern shape to the metal plate of a bonded assembly in which a ceramic substrate and a metal plate are bonded via a bonding layer; a first etching step of etching the metal plate multiple times in the depth direction to expose the bonding layer; a second etching step of etching the bonding layer exposed by the first etching step; and a third etching step of etching the side surface of the metal plate formed by the first etching step to form an overhanging portion of the bonding layer.

[0013] An example of a method for manufacturing a ceramic circuit board according to an embodiment is shown in Figure 1. In the figure, reference numeral 1 denotes a bonded body, reference numeral 2 denotes a ceramic substrate, reference numeral 3 denotes a front metal plate among the metal plates, reference numeral 4 denotes a rear metal plate among the metal plates, reference numeral 5 denotes a bonding layer, and reference numeral 14 denotes a resist. Figure 1 is a cross-sectional side view (X-Z cross-section) illustrating a bonded body and the like to explain the process of forming inclined portions 6, 7 between adjacent front metal plates 31, 32. In Figures 1 to 9, the plane of the ceramic substrate 2 is defined as the X-axis and Y-axis, and the direction perpendicular to the X-axis and Y-axis is defined as the Z-axis (thickness direction).

[0014] The bonded body is formed by bonding a ceramic substrate 2 to a metal plate 3 and a ceramic substrate 2 to a metal plate 4 via a bonding layer 5. In FIG. 1 , metal plates 3 and 4 are bonded to both sides of the ceramic substrate 2. The bonded body according to the embodiment may have metal plates bonded to both sides or only one side. For convenience, the metal plate bonded to the front surface (top surface in the drawing) is referred to as the front metal plate 3, and the metal plate bonded to the back surface (bottom surface in the drawing) is referred to as the back metal plate 4.

[0015] First, a step is performed to prepare a bonded body in which the ceramic substrate 2 and the metal plates 3 and 4 are bonded via the bonding layer 5. Next, a step is performed to apply a resist to the front metal plate 3 of the bonded body in order to impart a pattern shape.

[0016] FIG. 1( a ) shows a process of applying resist 14 to the joined body. This process also shows a process of applying resist 14 to the front metal plate 3 of the joined body. The resist 14 is an etching resist. The resist 14 is provided at locations on the front metal plate 3 where a circuit pattern is desired to be applied. Furthermore, in locations where T≧P is desired, the resist 14 is positioned so that the gap W between the resists 14 is within a range of 0.2 to 0.6 times the shortest distance P. Therefore, it is preferable that the gap W between the resists 14 is within a range of 0.2 to 0.6 times the shortest distance P between adjacent front metal plates 31 and 32 after the third etching step shown in FIG. 1( d ) described below. In other words, it is preferable that the resist 14 is applied so that 0.2P≦W≦0.6P is satisfied immediately before removing the resist 14 after the third etching step. Note that, in locations where T≧P is not required, the gap W between the resists 14 is optional.

[0017] A first etching step is performed in which the front metal plate 3 is etched multiple times in the depth direction to expose the bonding layer 5. Figures 1(b) and 1(c) illustrate the first etching step.

[0018] FIG. 1(b) shows a process of etching the gaps between the resists 14. FIG. 1(b) shows a process of etching a portion of the thickness T of the front metal plate 3. A method of etching a portion of the thickness of the front metal plate 3, but not all of it, is called half-etching. The depth of each half-etching is preferably within a range of 0.2 to 0.7 times the thickness T of the front metal plate 3. The depth of the first half-etching serves to form the second inclined portion 7 of the front metal plate 3, which will be described later. Furthermore, by making the gaps between the resists 14 within a range of 0.2 to 0.6 times the shortest distance P between adjacent front metal plates after manufacturing, the shortest distance P between the front metal plates formed by the first half-etching can be T≧P. This allows the final shape to satisfy T≧P, or even T>P, even when etching is performed in stages.

[0019] For example, when the metal plates 3 and 4 are copper plates, the etching solution may be an etching solution containing copper chloride or ferric chloride.

[0020] FIG. 1(c) shows a second etching step performed on a bonded body that has been subjected to a first half-etching. FIG. 1 illustrates a step of etching the front metal plate 3 twice to expose the bonding layer 5. In the manufacturing method according to the embodiment, the number of times the front metal plate 3 is etched is not limited to two, i.e., the number of inclined portions is not limited to two, and etching may be performed three or more times. FIG. 1(c) shows a step of etching until the bonding layer 5 is exposed. In addition, the first etching step may etch only a portion of the bonding layer 5.

[0021] For example, if two inclined portions are formed in two etching steps to expose the bonding layer 5, the second inclined portion 7 is formed in the first half-etching step, and the first inclined portion 6 is formed in the second half-etching step. Furthermore, if three inclined portions are formed in three etching steps to expose the bonding layer 5, the second inclined portion 7 is formed in the first half-etching step, the third inclined portion 12 is formed in the second half-etching step, and the first inclined portion 6 is formed in the third half-etching step. The upper limit of the number of half-etching steps is not particularly limited, but preferably five or fewer. Increasing the number of etching steps may result in a loss of further effectiveness and may even increase costs. Therefore, the number of first etching steps is preferably two to five, more preferably two or three. Although the description assumes that one etching step is performed per inclined portion, this is not limited to this case. Multiple etching steps may be performed per inclined portion.

[0022] By varying the etching conditions for the first and second etching steps, the angle θ1 of the first inclined portion 5 and the angle θ2 of the second inclined portion 6 can be made different. Etching conditions include the etching time, the etching solution (etchant concentration and components), and the presence or absence of ultrasound. For example, when using the same etching solution, shortening the second etching time compared to the first etching time allows the angle θ1 of the first inclined portion 6 and the angle θ2 of the second inclined portion 7 to satisfy the equation (1) described below. In other words, when the first etching step involves multiple etchings, the amount of metal plate etched away decreases as the etching progresses. For example, the amount of front metal plate 3 etched away by the second etching step is reduced compared to the first etching step, and the amount of front metal plate 3 etched away by the third etching step is reduced compared to the first etching step. This allows the inclined portion structure described in the embodiment to be achieved. Furthermore, when half etching is performed three or more times, the angles can be controlled by varying the etching conditions for each step.

[0023] It is also preferable that the number of times the etching bath is changed corresponds to the number of times etching is performed. When changing the etching conditions, changing the etching bath makes it easy to manage the etching solution. Therefore, when changing the etching conditions, it is preferable to change the etching bath.

[0024] The etching rate may also be changed within a single etching bath. If the etching conditions are changed once within a single bath, the number of etching cycles will be two. For example, by changing the concentration, temperature, and spray pressure of the etching solution between the first and second halves of etching within a single etching bath, the etching rate can be changed midway through etching. In such cases, the number of etching cycles will be multiple, not just one.

[0025] Next, a second etching step is performed to etch the bonding layer 5, and a third etching step is performed to etch the side surface of the front metal plate 3 and form a protruding portion of the bonding layer. The second etching step and the third etching step may be a single etching step or may be separate etching steps.

[0026] FIG. 1(d) illustrates the second and third etching steps. The second etching step is a step of etching the bonding layer 5. When the bonding layer 5 is an active metal bonding layer, the etching method described in Patent Document 3 can be used. When the brazing material does not contain Ag, the etching method described in Patent Document 5 can be used. When a silicon nitride substrate and a copper plate are bonded using a Ti-containing active metal brazing material, a titanium nitride layer is formed on the surface of the silicon nitride substrate as the bonding layer 5. Because the titanium nitride layer is conductive, removing it by etching can prevent poor electrical connection between adjacent copper plates. Therefore, since the bonding layer 5 consists of at least two layers, a titanium nitride layer and another layer, the etching of the bonding layer 5 may be performed in two or more stages. The etching step of the bonding layer 5 also serves to control the size of the protruding portion of the bonding layer on the front metal plates 31 and 32.

[0027] The third etching step is a step of etching the side surface of the front metal plate 3. By etching the side surface of the front metal plate 3, it is possible to control the length of the protruding portion of the bonding layer. Furthermore, when the front metal plate 3 is a copper plate, it is possible to mainly etch the side surface of the copper plate by using ferric chloride or cupric chloride as the etching solution in the third etching step.

[0028] The first etching step, the second etching step, and the third etching step can be performed using one or more of batch processing, continuous processing, and single-wafer processing. Batch processing is a method in which bonded bodies coated with resist 14 are placed in a basket and repeatedly immersed and removed from an etching tank one by one. Continuous processing is a method in which bonded bodies coated with resist 14 are placed one by one on a moving mechanism and transported within the etching tank by the moving mechanism. Examples of a moving mechanism for transporting multiple bonded bodies within the etching tank include a belt conveyor. Furthermore, single-wafer processing is a method in which bonded bodies coated with resist 14 are placed on a transport tray, and the transport tray carrying the bonded bodies is transported between multiple processing chambers on a belt conveyor. Furthermore, the back metal plate 4 may also be subjected to the first etching step, the second etching step, and the third etching step in the same manner as the front metal plate 3, or may be performed simultaneously with the etching of the front metal plate 3.

[0029] At least one of a chemical polishing step and a water washing step may be provided between multiple etching steps. When providing a front metal plate 31 having a step 13 as shown in Fig. 8, a method is available in which a front metal plate 3 having a step 13 is previously bonded. Alternatively, the bonded front metal plate 3 may be subjected to etching or cutting to form the step 13.

[0030] The ceramic substrate 2 is preferably one selected from the group consisting of a silicon nitride substrate, an aluminum nitride substrate, an aluminum oxide substrate, and a zirconium oxide substrate.

[0031] The thermal conductivity of silicon nitride substrates is 50 W / m·K or more, and even 80 W / m·K or more. Furthermore, the three-point bending strength of silicon nitride substrates is 600 MPa or more, and even 700 MPa or more. The thermal conductivity of aluminum nitride substrates is 150 W / m·K or more, and even 200 W / m·K or more. The three-point bending strength of aluminum nitride substrates is approximately 300 to 450 MPa. The three-point bending strength of aluminum oxide substrates is approximately 300 to 450 MPa, but aluminum oxide substrates are less expensive than other substrates. Furthermore, the thermal conductivity of aluminum oxide substrates is approximately 20 to 30 W / m·K. The three-point bending strength of zirconium oxide substrates is high, approximately 550 MPa, but the thermal conductivity is approximately 30 to 50 W / m·K.

[0032] The thickness of the ceramic substrate is preferably in the range of 0.2 mm to 3 mm, and more preferably 0.2 mm to 1 mm. If the thickness of the ceramic substrate is less than 0.2 mm, the strength may be insufficient. If the thickness of the ceramic substrate exceeds 3 mm, the ceramic substrate may become a thermal resistor and the heat dissipation performance may be reduced. Furthermore, since silicon nitride substrates have high strength, the thickness can be set to the range of 0.2 mm to 1.0 mm, and even 0.2 mm to 0.5 mm. From the viewpoint of thinning the substrate, it is preferable to use a silicon nitride substrate.

[0033] The metal plates 3 and 4 are preferably copper plates (including copper alloy plates) or aluminum plates (including aluminum alloy plates). Furthermore, the copper plates are preferably oxygen-free copper plates. Oxygen-free copper has a copper purity of 99.96 wt% or higher, as specified in JIS-H-3100. The thermal conductivity of copper is approximately 400 W / m·K, while the thermal conductivity of aluminum is approximately 240 W / m·K. Because copper has a higher thermal conductivity than aluminum, it improves heat dissipation. Furthermore, the aluminum plates are preferably pure aluminum. Pure aluminum is specified in JIS-H-4000. JIS-H-4000 corresponds to ISO 6361. JIS-H-3100 corresponds to ISO 197 and other standards. For this reason, it is preferable that the ceramic substrate 2 be a silicon nitride substrate with a thickness of 1.0 mm or less, and the metal plates 3 and 4 be copper plates with a thickness of 0.5 mm or more.

[0034] The bonding layer 5 is preferably an active metal bonding layer. When the metal plates 3 and 4 are copper plates, the bonding layer 5 is preferably an active metal bonding layer containing either Ag or Cu as a main component. The active metal may be one or more selected from titanium (Ti), zirconium (Zr), hafnium (Hf), and niobium (Nb).

[0035] The composition of the active metal brazing material containing Ag or Cu as the main component is 0 mass % or more and 60 mass % or less of Ag (silver), 15 mass % or more and 70 mass % or less of Cu (copper), Ti (titanium) or TiH ２ It is preferable that the titanium hydride is contained in an amount of 1 mass % or more and 15 mass % or less. ２ When both Ag and Cu are used, the total content is preferably in the range of 1% by mass to 15% by mass. When both Ag and Cu are used, the Ag content is preferably in the range of 20% by mass to 60% by mass and the Cu content is preferably in the range of 15% by mass to 40% by mass.

[0036] If necessary, one or both of Sn (tin) and In (indium) may be contained in an amount of 1 mass % to 50 mass %. ２ The content of is preferably in the range of 1% by mass to 15% by mass. If necessary, C (carbon) may be contained in an amount of 0.1% by mass to 2% by mass.

[0037] The ratio of the active metal brazing material composition is calculated by taking the total of the mixed raw materials as 100% by mass. For example, when the active metal brazing material is composed of three elements, Ag, Cu, and Ti, the ratio is Ag + Cu + Ti = 100% by mass. ２ , and In, Ag + Cu + TiH ２ In addition, when the alloy is composed of five elements, Ag, Cu, Ti, Sn, and C, the total weight of the alloy is Ag+Cu+Ti+Sn+C=100% by mass.

[0038] Ag or Cu is a component that serves as the base material of the brazing material. Sn or In has the effect of lowering the melting point of the brazing material. C (carbon) has the effect of controlling the fluidity of the brazing material and controlling the structure of the bonding layer 5 by reacting with other components. Therefore, examples of components of the brazing material include Ag-Cu-Ti, Ag-Cu-Sn-Ti, Ag-Cu-Ti-C, Ag-Cu-Sn-Ti-C, Ag-Ti, Cu-Ti, Ag-Sn-Ti, Cu-Sn-Ti, Ag-Ti-C, Cu-Ti-C, Ag-Sn-Ti-C, and Cu-Sn-Ti-C. In addition, In may be used instead of Sn. Furthermore, both Sn and In may be used.

[0039] The active metal brazing filler metal may contain 0.1 mass % to 10 mass % of one or more elements selected from tungsten (W), molybdenum (Mo), and rhenium (Re). Tungsten, molybdenum, and rhenium can control the fluidity of the active metal brazing filler metal. Magnesium (Mg) may also be added to the active metal brazing filler metal.

[0040] When the metal plates 3 and 4 are aluminum plates, the bonding layer 5 is preferably an active metal bonding layer primarily composed of Al. The active metal may be one or more selected from silicon (Si) and magnesium (Mg). The active metal brazing material primarily composed of Al is preferably an Al-Si or Al-Mg brazing material. The content of one or both of Si and Mg in the active metal brazing material is preferably within the range of 0.1% by mass to 20% by mass. The active metal brazing material is made into a paste to prepare an active metal brazing material paste. The active metal brazing material paste is applied to the ceramic substrate 2, and the metal plates are placed on top of it. When bonding the metal plates 3 and 4 to both sides, the active metal brazing material paste is applied to both sides, and the metal plates 3 and 4 are then placed.

[0041] Thereafter, the ceramic substrate 2 and the metal plates 3, 4 are heated and bonded at a temperature in the range of 600°C to 960°C, thereby obtaining a bonded body. The bonded body is subjected to an etching process, thereby producing the ceramic circuit board 1 according to the embodiment. Also, a direct bonding method may be used instead of the active metal bonding method.

[0042] The ceramic substrate 2 may also be a large substrate with a long side of 100 mm or more. A large ceramic substrate 2 to which large metal plates 3 and 4 are bonded is called a large bonded body. A large bonded body can be produced in multiple pieces. Multi-cavity production means that the large bonded body is divided into multiple ceramic circuit substrates 2. Multi-cavity production is excellent for mass production.

[0043] The ceramic circuit board 1 according to the embodiment can be obtained by etching a large bonded body and then dividing it into a plurality of ceramic circuit boards 1. In the method of bonding a pre-half-etched metal plate to a large ceramic substrate as in Patent Document 4, alignment is time-consuming. For this reason, the method using a pre-half-etched metal plate cannot necessarily be said to be excellent for mass production.

[0044] Furthermore, the ceramic circuit board 1 obtained by the manufacturing method of the ceramic circuit board 1 according to the embodiment has a portion where T≧P, where T is the thickness of each front metal plate 3 of the plurality of metal plates and P is the shortest distance between adjacent front metal plates 31, 32 among the plurality of metal plates, and the side of the front metal plate that faces the adjacent front metal plate 32 has at least a first inclined portion 6 and a second inclined portion 7 that have different inclination angles in that order from the ceramic substrate 2 side, and the protrusion width 9 of the upper end portion 8 of adjacent front metal plates 3 can be ±40 μm or less.

[0045] 2 to 5 show an example of a ceramic circuit board according to an embodiment. In the drawings, reference numeral 1 denotes a ceramic circuit board, reference numeral 2 denotes a ceramic substrate, reference numeral 3 denotes a front metal plate among the metal plates, reference numerals 31 and 32 denote front metal plates among the front metal plate 3, reference numeral 4 denotes a rear metal plate among the metal plates, reference numeral 5 denotes a bonding layer, reference numeral 6 denotes a first inclined portion, reference numeral 7 denotes a second inclined portion, reference numeral 8 denotes an upper end portion of the front metal plate 3, reference numeral 9 denotes a protruding portion of the upper end portion 8, reference numeral 10 denotes a contact point of the inclined portions, reference numeral 10 denotes a contact line of the inclined portions (the contact point of the inclined portions in the side cross-sectional view), reference numeral 11 denotes a difference in height between the contact points 10 of the inclined portions, reference numeral T denotes the thickness of the front metal plate 3, reference numeral P denotes the shortest distance between adjacent front metal plates 3, reference numeral θ1 denotes the inclination angle of the first inclined portion 6, reference numeral θ2 denotes the inclination angle of the second inclined portion 7, and reference numeral D denotes the combined width of the inclined portions 6 and 7.

[0046] For convenience, the metal plate to which a circuit pattern is imparted is referred to as the front metal plate 3, and the metal plate used as a heat sink is referred to as the back metal plate 4. In the ceramic circuit board 2 according to the embodiment, a circuit pattern may also be imparted to the back metal plate 4. Furthermore, although two front metal plates 3 are arranged in one direction (X-axis direction) in the examples shown in FIGS. 1 to 5, there may be three or more front metal plates 3 arranged in one direction. The same applies when a circuit pattern is imparted to the back metal plate 4.

[0047] The ceramic circuit board 1 according to the embodiment has multiple front metal plates 31, 32 bonded to at least one surface of a ceramic substrate 2 (front metal plate 3 in FIGS. 1 to 5 ). The front metal plates 3 are arranged in a one-to-multiple, multiple-to-one, or multiple-to-multiple configuration in the X-axis and Y-axis directions. When the thickness of the front metal plate 3 is T and the shortest distance between adjacent front metal plates 31, 32 is P, there is a location where T≧P. The shortest distance P between adjacent front metal plates 31, 32 is the distance between the ends of the front metal plate 3 that contact the bonding layer 5 (or the ceramic substrate 2). The shortest distance P is not limited to the lower end of the metal plate as shown in FIG. 2 , but may be the upper end 8 or other location. The shortest distance P between adjacent front metal plates 31, 32 is sometimes simply referred to as the shortest distance P.

[0048] The locations where T≧P indicate locations where the shortest distance P between adjacent front metal plates 31, 32 is the same or is smaller than the thickness T of the front metal plates 31, 32. By having a location where the shortest distance P is smaller than the metal plate thickness T, the bonding area of ​​the front metal plate 3 can be secured. Furthermore, it is possible to reduce the size of the ceramic circuit board 1 compared to when the bonding area of ​​the front metal plate 3 is the same. Furthermore, when the size of the ceramic substrate 2 is the same, the bonding area of ​​the front metal plate 3 can be increased.

[0049] The thickness T of the front metal plate 3 is the thickness of a single metal plate. When multiple front metal plates are stacked, the thickness T is the thickness of the front metal plate 3 bonded to the ceramic substrate 2 or the bonding layer 5. The side of the front metal plate 31 facing the adjacent front metal plate 32 has at least a first inclined portion 6 and a second inclined portion 7, which have different inclination angles, in that order from the ceramic substrate 2 side. The side surfaces of the adjacent front metal plates 31, 32 have a first inclined portion 6 and a second inclined portion 7. The first inclined portion 6 and the second inclined portion 7 need only differ in inclination angle by 1° or more. In the plane of FIGS. 1 to 9, the side of the front metal plate 31 facing the adjacent front metal plate 32 is the right side of the front metal plate 31. In the plane of FIGS. 1 to 5, the side of the front metal plate 32 facing the adjacent front metal plate 31 is the left side of the front metal plate 32.

[0050] The tangent line (contact point in the side cross section) between the first inclined portion 6 and the second inclined portion 7 is the inclined portion contact point 10. The inclined portion closer to the ceramic substrate 2 (or the bonding layer 5) is the first inclined portion 6. The inclined portion closer to the upper end portion 8 of the front metal plate 3 is the second inclined portion 7. The first inclined portion 6 and the second inclined portion 7 have a substantially linear shape when viewed in side cross section. At least one of the first inclined portion 6 and the second inclined portion 7 may have an R-shape when viewed in side cross section. The first inclined portion 6 and the second inclined portion 7 may be an inclined portion that combines a substantially linear shape and an R-shape when viewed in side cross section.

[0051] The protrusion width 9 of the upper end portions 8 (shown in FIG. 4 ) of adjacent front metal plates 31, 32 is ±40 μm or less. The protrusion width 9 of the upper end portions 8 of the front metal plates 31, 32 is the sum of the length (−40 μm) by which the upper end portions 8 protrude from the line in the Z-axis direction passing through the inclined portion contact point 10 of the front metal plate 31 and the length (+40 μm) by which the upper end portions 8 are recessed from the line in the Z-axis direction passing through the inclined portion contact point 10. In other words, |protrusion width|≦40. As shown in FIG. 6( a), if the upper end portions 8 extend beyond the line in the Z-axis direction passing through the inclined portion contact point 10 toward the adjacent front metal plate (positive direction of the X-axis), the protrusion width 9 is a negative value. As shown in FIG. 6( b), if the upper end portions 8 extend beyond the line in the Z-axis direction passing through the inclined portion contact point 10 toward the opposite side of the adjacent front metal plate (negative direction of the X-axis), the protrusion width 9 is a positive value. 6 shows only the inclined portion on the right side for convenience, but the same can be considered with the Z-axis line as the reference when the inclined portion is on the left side. Also, when the upper end 8 is on the Z-axis line passing through the inclined portion contact point 10 of the front metal plate 31, the protrusion portion of the front metal plate 31 is 0 (zero).

[0052] The protrusion width 9 of the upper end portion 8 is a value that depends on the angle θ2. Therefore, a protrusion width 9 of ±40 μm or less indicates that the angle θ2 is close to 90°. This allows the width between the upper end portions 8 of adjacent front metal plates 31, 32 to be wide, even if T≧P, thereby suppressing discharge and maintaining insulation. For this reason, the protrusion width 9 is preferably ±40 μm or less, and even more preferably ±20 μm or less. Furthermore, it is preferable that the protrusion width 9 does not extend beyond the lower end of the front metal plate 3 in the X-axis direction. This is because if the protrusion width 9 extends along the X-axis beyond the lower end of the front metal plate 3, the gap between the upper end portion 8 of the front metal plate 31 and the upper end portion 8 of the front metal plate 32 will be narrowed, potentially reducing insulation.

[0053] The shortest distance P, the presence or absence of the inclined portions 6, 7, and the protrusion width 9 of the upper end portions 8 of the front metal plates 31, 32 are observed in a cross-sectional SEM photograph of the ceramic circuit board 1. A cross section perpendicular to the thickness direction of the adjacent front metal plates 31, 32 is observed. The SEM photograph used is a magnification of 50 times.

[0054] The upper end portions 8 of adjacent front metal plates 31, 32 preferably have a shape with a radius of curvature of 15 μm or less. Having a radius of curvature of 15 μm or less means that the upper end portions 8 of the front metal plates 31, 32 are almost completely curved, allowing for a wider flat surface on the surface of the front metal plates 31, 32. A wider flat surface allows for a larger mounting area for semiconductor devices and the like. Therefore, the upper end portions 8 of adjacent front metal plates 31, 32 preferably have a radius of curvature of 15 μm or less, and even 7 μm or less. The radius of curvature is measured using the aforementioned SEM photograph (50x magnification). The angles θ1 and θ2, the protrusion width 9, the radius of curvature, and the like may be calculated using imaging software. For example, ImageJ can be used as imaging software.

[0055] It is preferable that the angle θ1 of the first inclined portion 6 and the angle θ2 of the second inclined portion 7 satisfy the following formula (1). The angle θ1 shown in FIG. 4 is the angle between the bottom surface of the front metal plate 31 and the first inclined portion 6. The angle θ2 is the angle between the upper side (surface) of the metal plate 31 and the second inclined portion 7. Note that the front metal plate 32 is not shown in FIG. 4. θ1<θ2 (1) It is also preferable that the angle of the first inclined portion 6 is in the range of 20° to 75°, and the angle of the second inclined portion 7 is in the range of 80° to 110°.

[0056] The first inclined portion 6 is an inclined portion starting from the lower end of the side surface of the front metal plate 3 that is in contact with the ceramic substrate 2 or the bonding layer 5. The second inclined portion 7 is an inclined portion starting from the upper end 8 of the front metal plate 3. When the angle θ1 of the first inclined portion 6 and the angle θ2 of the second inclined portion 7 satisfy the above formula (1), the thermal stress relaxation effect of the lower end of the side surface of the front metal plate 3 is improved.

[0057] The angle θ1 of the first inclined portion 6 is preferably within a range of 20° to 75°. The angle θ1 of the first inclined portion 6 is sometimes simply referred to as the angle θ1. By setting the angle θ1 to 20° to 75°, it is possible to achieve both a stress relaxation effect and a reduction in the combined width D of the inclined portions 6 and 7. For this reason, the angle θ1 is preferably within a range of 20° to 75°, and more preferably within a range of 30° to 60°.

[0058] The angle θ2 of the second inclined portion 7 is preferably within a range of 80° to 110°. The angle θ2 of the second inclined portion 7 is sometimes simply referred to as the angle θ2. By setting the angle θ2 within a range of 80° to 110°, it becomes easier to control the protrusion width 9 to ±40 μm or less. It also becomes easier to control the radius of curvature of the upper end portion 8 of the front metal plate 3 to 15 μm or less. For this reason, the angle θ2 is preferably within a range of 80° to 110°, and more preferably within a range of 85° to 100°.

[0059] The angles θ1 and θ2 are measured using the SEM photograph (50x magnification) described above. A straight line is drawn at the location of the first inclined portion 6 and the second inclined portion 7, and the respective angles are determined. In this case, even if there are small irregularities (e.g., irregularities of 5 μm or less) on the surfaces of the first inclined portion 6 and the second inclined portion 7, they are treated as straight lines.

[0060] Even when the inclined portions 6 and 7 have an R-shape, a straight line is drawn at the location of the first inclined portion 6 and the second inclined portion 7 using an SEM photograph (50x magnification) to determine the respective angles.

[0061] It is preferable that the ratio D / T of the combined width D of the inclined portions 6 and 7 of the front metal plate 3 to the thickness T of the front metal plate 3 satisfies the following formula (2): D / T≦0.5 (2)

[0062] The combined width D of the inclined portions 6 and 7 of the front metal plate 3 is sometimes simply referred to as width D. Width D is the distance from the intersection of a line in the X-axis direction passing through the top end 8 of the front metal plate 31 and a line in the Z-axis direction passing through the edge of the bottom end of the front metal plate 31 to the top end 8 of the front metal plate 31. A ratio D / T of 0.5 or less indicates that width D is 1 / 2 or less of the thickness T of the front metal plate 3. While the lower limit of the ratio D / T is not limited, it is preferable that the ratio D / T be 0.1 or greater. If the ratio D / T is less than 0.1, the small width D may make it difficult to provide the first inclined portion 6 and the second inclined portion 7. Therefore, even if width D is reduced to 0.1 or greater and 0.5 or less relative to the thickness T of the front metal plate 3, the number of locations satisfying T≧P can be increased. Width D is measured using the SEM photograph (50x magnification) described above.

[0063] It is preferable that the thickness T of the front metal plate 3 satisfy the following formula (3): T≧0.5 mm (3) Increasing the thickness of the front metal plate 3 can improve heat dissipation and current-carrying capacity. For this reason, the thickness T of the front metal plate 3 is preferably 0.5 mm or more, and more preferably 0.8 mm or more. Because the front metal plate 3 has the first inclined portion 6 and the second inclined portion 7 with different inclination angles, a location where T≧P can be achieved can be created even if the front metal plate 3 is thick. Note that the upper limit of the thickness T of the front metal plate 3 is not particularly limited, but is preferably 5 mm or less. For this reason, the thickness T of the front metal plate 3 is preferably within a range of 0.5 mm to 5 mm, and more preferably 0.8 mm to 3 mm.

[0064] When the point of contact between the first inclined portion 6 and the second inclined portion 7 is defined as the inclined portion contact point 10, the height difference between the inclined portion contact points 10 between adjacent front metal plates 31, 32 is preferably within a range of 100 μm. The height difference between the inclined portion contact points 10 between adjacent front metal plates 31, 32 is sometimes simply referred to as the height difference between the inclined portion contact points 10. The inclined portion contact point 10, which is the point of contact between the first inclined portion 6 and the second inclined portion 7, is observed in the SEM photograph (50x magnification) described above. A line is drawn in the X-axis direction from the inclined portion contact point 10 of the front metal plate 31 to the inclined portion contact point 10 of the adjacent front metal plate 32. The height difference between the line in the X-axis direction related to the front metal plate 31 and the line in the X-axis direction related to the front metal plate 32 is defined as the height difference 11 of the inclined portion contact points 10. For example, the difference in height (thickness direction of the front metal plate 3) between a line in the X-axis direction passing through the inclined portion contact point 10 of the front metal plate 31 and a line in the X-axis direction passing through the inclined portion contact point 10 of the front metal plate 32 is measured.

[0065] When the height difference 11 of the inclined portion contact points 10 is 100 μm or less, this indicates that the side shapes of the adjacent front metal plates 31, 32 are similar. This makes it possible to improve the TCT characteristics even if there are locations where T≧P. The similarity of the side shapes of the adjacent front metal plates 31, 32 makes it possible to homogenize the effects of thermal expansion. For this reason, it is preferable that the height difference 11 of the inclined portion contact points 10 be 100 μm or less, and more preferably within the range of 0 μm to 50 μm.

[0066] The inclined portions 6 and 7 may each have an R-shape as shown in Fig. 9. In this case, it is preferable that the inclined portion contact point 10 is a point where the radius of curvature of the first inclined portion 6 and the second inclined portion 7 differ by 50 µm or more. It is preferable that the radius of curvature of the first inclined portion 6 and the radius of curvature of the second inclined portion 7 satisfy the following formula (4): radius of curvature of first inclined portion 6 ≥ radius of curvature of second inclined portion 7 (4)

[0067] The radius of curvature of the first inclined portion 6 is preferably in the range of 600 μm to 850 μm, and the radius of curvature of the second inclined portion 7 is preferably in the range of 350 μm to 550 μm. Note that the front metal plate 32 is not shown in FIG. 9 .

[0068] When the radius of curvature of the first inclined portion 6 is large within the range of 600 μm or more and 850 μm or less, the first inclined portion 6 has a gentler slope than the second inclined portion 7. Therefore, the angle θ1 of the first inclined portion 6 can also be made small. On the other hand, when the radius of curvature of the second inclined portion 7 is large within the range of 350 μm or more and 550 μm or less, the slope becomes steep (close to vertical), and the angle θ2 of the second inclined portion 7 becomes large.

[0069] Therefore, by satisfying the above formula (4), the above formula (1) is satisfied as described above, and it is possible to improve the thermal stress relaxation effect at the lower end of the side surface of the front metal plate 3. The method for measuring the radius of curvature was to use an SEM photograph (50x magnification) to draw circles along the slopes of the R parts of the first inclined portion 6 and the second inclined portion 7, and the difference in the radii of the circles was taken as the radius of curvature.

[0070] The number of inclined portions on the side surface of the front metal plate 3 is not limited to two, and may be three or more. Figure 7 illustrates a front metal plate 31 having three inclined portions. In the figure, reference numeral 31 denotes the front metal plate, reference numeral 6 denotes the first inclined portion, reference numeral 7 denotes the second inclined portion, and reference numeral 12 denotes the third inclined portion. When there are three inclined portions, there are two inclined portion contact points 10: the contact point between the first inclined portion 6 and the third inclined portion 12, and the contact point between the second inclined portion 7 and the third inclined portion 12. Even when there are multiple inclined portion contact points 10, it is preferable that the height difference 11 between the inclined portion contact points 10 is 100 μm or less. The angle of the third inclined portion 12 is indicated by θ3. The angle θ3 is the angle formed between the line in the X-axis direction passing through the tangent point 10 between the first inclined portion 6 and the third inclined portion 12 and the tangent point with the third inclined portion 12, and is greater than angle θ1 and smaller than angle θ2.

[0071] Although not shown, when a fourth inclined portion is provided on the front metal plate 3, the fourth inclined portion is provided on the second inclined portion 7 side of the third inclined portion 12. The inclination angle of the fourth inclined portion is defined as θ4, which is the angle formed between the line in the X-axis direction passing through the junction of the third inclined portion 12 and the fourth inclined portion and the junction of the fourth inclined portion. As the number of inclined portions becomes n, the angle of the inclined portion is represented by θn.

[0072] Among the multiple front metal plates 3, at least one front metal plate 3 may have a step portion on its surface. Figure 8 shows an example of a front metal plate 31 having a step portion. In the figure, reference numeral 31 denotes the front metal plate, reference numeral 6 denotes the first inclined portion, reference numeral 7 denotes the second inclined portion, reference numeral 8 denotes the upper end of the front metal plate 31, reference numeral 13 denotes the step portion (protrusion), and reference numeral T denotes the thickness of the front metal plate 31. The step portion 13 is integrated with the front metal plate 31. The step portion 13 is not formed by joining a separate metal plate to the front metal plate 31. Providing the step portion 13 can improve the heat dissipation and current-carrying capacity of that portion. Furthermore, the thickness T of the front metal plate 31 having the step portion 13 is the thickness up to the upper end 8 of the front metal plate 31, which is the starting point of the second inclined portion 7. The side of the step portion 13 itself may or may not have an inclined portion.

[0073] The range of 50% to 100% of the shortest distance P between adjacent front metal plates 31, 32 may be such that T>P is satisfied. The total length in the Y-axis direction of the side surfaces of the front metal plates 31, 32 adjacent in the X-axis direction that face each other is taken as 100%, and this shows the proportion of the length in the Y-axis direction of the side surfaces that satisfy T>P. This means that the following formula (5) is satisfied: 50%≦(length of side surface that satisfies T>P / total length of side surface facing adjacent surface)×100≦100% ... (5)

[0074] The length of the side surface that satisfies T > P refers to the length of the side surface that satisfies T > P among the lengths in the Y-axis direction of the side surface of the front metal plate 31 that faces the adjacent front metal plate 32 in the X-axis direction when viewed from above the ceramic substrate 2. The length of the side surface facing the adjacent surface refers to the length in the Y-axis direction of the side surface of the front metal plate 31 that faces the adjacent front metal plate 32. Note that the side surface of the front metal plate 31 that faces the adjacent front metal plate 32 is not necessarily straight when viewed from above, and may be bent. In such cases, the length in the Y-axis direction of the side surface of the front metal plate 31 that faces the adjacent front metal plate 32 includes the length of the bent portion. The same applies to the case where the front metal plate 31 faces the adjacent front metal plate in the Y-axis direction as to the case where the front metal plate faces the adjacent front metal plate in the X-axis direction. By increasing the number of side surfaces for which the shortest distance P between adjacent front metal plates 31, 32 satisfies T > P in relation to the thickness T of the front metal plate 3, the ceramic circuit substrate 1 can be made smaller.

[0075] When viewing the X-Z cross section of the ceramic substrate 2 in an SEM image (50x magnification), the distance between adjacent front metal plates 31, 32, i.e., the inter-pattern distance S, varies depending on the position in the thickness direction. The proportion of the inter-pattern distance S that satisfies S<T with respect to the thickness of the front metal plate 3 may be 25% or more and 100% or less. In other words, the proportion of areas where the inter-pattern distance S is wider than the metal plate thickness T may be within the range of 30% or more and 100% or less. By setting the proportion that satisfies S<T to 25% or more and 100% or less, the heat dissipation properties of the front metal plate 3 can be improved. This makes it possible to achieve a miniaturization of the ceramic circuit substrate 1.

[0076] It is preferable that the ratio Mm / Mc of the area Mc of one surface of the ceramic substrate 2 to the total area Mm of the front metal plate 3 bonded to that surface satisfies the following formula (6): Mm / Mc≧0.8 (6) Mm is the total area of ​​the bonding surfaces of the front metal plate 3 bonded to one surface of the ceramic substrate 2. An Mm / Mc ratio of 0.8 or more indicates that the bonding area of ​​the front metal plate 3 is 80% or more. This means that even if the bonding area of ​​the front metal plate 3 is set to 80% or more, a location that satisfies T≧P can be created. Because a location that satisfies T≧P can be created while increasing the bonding area of ​​the front metal plate 3, the ceramic circuit substrate 1 can be made smaller than conventional ceramic circuit substrates to which metal plates of the same size or area are bonded.

[0077] The ceramic substrate 2 and the front metal plate 3 are bonded via a bonding layer 5, and the bonding layer 5 preferably has a protruding portion that protrudes from the end of the front metal plate 3. By having the bonding layer 5 protrude from the end of the front metal plate 3 (the lower end of the metal plate), the thermal stress relaxation effect is enhanced. The protruding length of the protruding portion of the bonding layer is preferably within the range of 10 μm to 150 μm. If the protruding length is less than 10 μm, the stress relaxation effect may be insufficient. If the protruding length exceeds 150 μm, poor conductivity with the adjacent metal plate may occur.

[0078] Of the side surfaces of the front metal plate 3 of the ceramic substrate 2, the side surface that does not face the adjacent front metal plate may or may not be provided with a plurality of inclined portions with different inclination angles. Inclined portions may be provided on both surfaces of the front metal plate 3 and the back metal plate 4. What is effective is to control the inclination of the side surface that faces the adjacent metal plate.

[0079] (Examples) (Examples 1 to 6, Comparative Examples 1 and 2) A silicon nitride substrate with a thermal conductivity of 90 W / m·K and a three-point bending strength of 650 MPa was prepared as the ceramic substrate 2. An active metal brazing filler metal containing Ti was used as the brazing filler metal. Copper plates (front copper plate and back copper plate) with a thickness of 0.5 mm or more were used as the metal plates 3 and 4. Bonded bodies 1 to 4 were prepared in which copper plates were bonded to both sides of a silicon nitride substrate via an active metal bonding layer. Of bonded bodies 1 to 4, bonded body 4 had a step portion 13 on the front copper plate. The sizes of the bonded bodies are as shown in Table 1. The bonded bodies were large bonded bodies with the long sides of the ceramic substrates being 100 mm or more.

[0080]

[0081] Next, an etching process was performed on the bonded bodies 1 to 4. In the etching process of the examples, in the first etching process, half-etching of the front copper plate was repeated two to three times to expose the bonding layer 5. The depth of each half-etching was repeated within a range of 0.2 to 0.7 times the thickness of the front metal plate. Thereafter, in the second etching process, the bonding layer 5 was etched. In the third etching process, the side surface of the front copper plate was etched to adjust the size of the protruding portion of the bonding layer. In the comparative example, the bonding layer 5 was exposed in one copper plate etching process. The gaps between the resists 14 were as shown in Table 2.

[0082]

[0083] Furthermore, a ceramic circuit substrate was fabricated by an etching process. The following were measured for the resulting ceramic circuit substrate: the shortest distance P between adjacent metal plates, the protrusion width 9, the radius of curvature of the upper end of the front metal plate, the angle θ1 of the first inclined portion, the angle θ2 of the second inclined portion, the angle θ3 of the third inclined portion, the difference in height between the inclined portion contact points, the combined width D of the multiple inclined portions formed on the front copper plate / the metal plate thickness T, the percentage of cases where the shortest distance P between adjacent metal plates is T>P, and the size of the protrusion portion of the bonding layer. The ratio Mm / Mc, where Mc is the area of ​​one surface of the ceramic substrate and Mm is the total area of ​​the metal plates bonded to that surface, was also determined. The results are shown in Tables 3 and 4. The protrusion width in Table 3 has a lower limit of 0 (zero) when the protrusion width value is positive, and an upper limit of 0 (zero) when the protrusion width value is negative.

[0084]

[0085]

[0086] As can be seen from Tables 3 and 4, in the examples, ceramic circuit boards were produced in which the side surfaces satisfying T>P accounted for 50% or more of the entire side surfaces. Furthermore, in examples 1, 2, 3, 5, and 6, in which half etching was performed twice, a first inclined portion 6 and a second inclined portion 7 were formed. Furthermore, in example 4, in which half etching was performed three times, a first inclined portion 6, a second inclined portion 7, and a third inclined portion 12 were formed.

[0087] In addition, Comparative Example 1 does not have any portion where T>P. This is because the bonding layer was exposed by a single etching. In Comparative Example 1, the shortest distance P was large, so Mm / Mc could not be made 0.8 or more. In Comparative Example 2, although T>P was satisfied, the first inclined portion and the second inclined portion with different angles were not provided.

[0088] Examples 3 and 4 and Comparative Examples 1 and 2 use the bonded body 3. In Examples 3 and 4, the number of locations where T>P is increased, which makes it possible to increase the bonding area of ​​the copper plates 3 and 4 to the ceramic substrate 2. In other words, this shows that if the bonding area of ​​the copper plates 3 and 4 is the same, it is possible to reduce the size of the ceramic circuit substrate 1.

[0089] The obtained ceramic circuit substrate was divided to obtain a plurality of ceramic circuit substrates 1. The variation in shape of the plurality of ceramic circuit substrates 1 after division was 10% or less (including 0%). It was found that the example is a manufacturing method suitable for mass production and suitable for multiple-cavity production.

[0090] Next, TCT was performed on the ceramic circuit substrates of the examples and comparative examples after they had been divided into multiple ceramic circuit substrates. TCT was performed 3,000 cycles, with one cycle consisting of -40°C x 30 minutes → room temperature x 10 minutes → 170°C x 30 minutes → room temperature x 10 minutes. The ceramic circuit substrates after TCT were observed for defects. The results are shown in Table 5.

[0091]

[0092] As can be seen from Table 5, the TCT characteristics of the examples were good. This shows that durability is good even when the region where T>P is increased. Comparative Example 1 had good TCT characteristics because T<P. Furthermore, the TCT characteristics of Comparative Example 2 deteriorated. This is because Comparative Example 2 does not have a first inclined portion and a second inclined portion with different inclination angles. Therefore, it can be seen that the examples can improve the TCT characteristics by increasing the region where T≧P and even T>P.

[0093] Although several embodiments of the present invention have been described above, these embodiments are presented by way of example only and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, modifications, etc. can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, as well as within the scope of the invention and its equivalents as set forth in the claims. Furthermore, the above-described embodiments can be implemented in combination with each other.

Claims

1. A method for manufacturing a ceramic circuit board, comprising: a coating step of coating a resist to impart a pattern to a metal plate of a bonded assembly in which a ceramic substrate and a metal plate are bonded via a bonding layer; a first etching step of etching the metal plate multiple times in the depth direction to expose the bonding layer; a second etching step of etching the bonding layer exposed by the first etching step; and a third etching step of etching the side surface of the metal plate formed by the first etching step to form an overhanging portion of the bonding layer.

2. A method for manufacturing a ceramic circuit board as described in claim 1, wherein the resist is applied in the coating step so that 0.2P≦W≦0.6P holds true, where P is the shortest distance between adjacent metal plates after the third etching step and W is the gap between the resists applied to the metal plates.

3. The method for manufacturing a ceramic circuit board according to claim 1 or 2, wherein the first etching step is performed two or three times.

4. A method for manufacturing a ceramic circuit board according to any one of claims 1 to 3, wherein the first etching step involves multiple etchings, and the more etching progresses, the less of the metal plate is removed.

5. A method for manufacturing a ceramic circuit board according to any one of claims 1 to 4, wherein the obtained ceramic circuit board has a portion where T≧P, where T is the thickness of each of the plurality of metal plates constituting the ceramic circuit board and P is the shortest distance between adjacent metal plates among the plurality of metal plates, and the side surfaces of the plurality of metal plates that face the adjacent metal plates have at least a first inclined portion and a second inclined portion with different inclination angles, in that order from the ceramic substrate side, and the protrusion width of the upper ends of the adjacent metal plates is ±40 μm or less.

6. The method for manufacturing a ceramic circuit board according to claim 5, wherein the radius of curvature of the upper ends of adjacent metal plates in the ceramic circuit board is 15 μm or less.

7. A method for manufacturing a ceramic circuit board according to claim 6, wherein the angle of the first inclined portion of the ceramic circuit board is within the range of 20° to 75°, and the angle of the second inclined portion is within the range of 80° to 110°.

8. A method for manufacturing a ceramic circuit board as described in claim 7, wherein the ratio D / T of the combined width D of the first inclined portion and the second inclined portion to the thickness T of each of the metal plates is 0.5 or less.

9. A method for manufacturing a ceramic circuit board according to any one of claims 5 to 8, wherein the thickness T of each of the metal plates in the ceramic circuit board is 0.5 mm or more.

10. A method for manufacturing a ceramic circuit board as described in claim 9, wherein the ratio of the length of the side surface where adjacent metal plates face each other in the Y-axis direction that satisfies T>P is within the range of 50% to 100%.

11. A method for manufacturing a ceramic circuit board according to any one of claims 1 to 10, wherein the ceramic substrate is a silicon nitride substrate having a thickness of 1.0 mm or less, and the metal plate is a copper plate having a thickness of 0.5 mm or more.

12. A method for manufacturing a ceramic circuit board according to any one of claims 1 to 11, wherein the long side of the ceramic substrate is 100 mm or more.

13. A method for manufacturing a ceramic circuit board, further comprising the step of dividing the ceramic circuit board according to claim 12 into a plurality of ceramic circuit boards.

Images