Bonded body, ceramic circuit board, and semiconductor device

Abstract

A bonded body according to an embodiment includes a ceramic substrate, a copper plate, and a bonding layer. The bonding layer is disposed on at least one surface of the ceramic substrate and bonds the ceramic substrate and the copper plate. The bonding layer contains a Ti reaction layer and a plurality of first alloys. The Ti reaction layer contains titanium nitride or titanium oxide as a main component. The plurality of first alloys are located between the Ti reaction layer and the copper plate. Each of the plurality of first alloys contains at least one type of alloy selected from the group consisting of a Cu-Sn alloy and a Cu-In alloy. The plurality of first alloys have mutually different Sn concentrations or In concentrations. According to the embodiment, the warping amount can be reduced, and the temperature rising speed and the temperature falling speed of the bonding process can be improved. In an embodiment, a silicon nitride substrate is suitable as a ceramic substrate.

Description

[0001] This application is a divisional application of the application with application number "202180053012.9", application date October 7, 2021, and invention name "Joint body, ceramic circuit substrate and semiconductor device". Technical Field

[0002] The embodiments described below relate to a bonded body, a ceramic circuit substrate, and a semiconductor device. Background Art

[0003] The joint body of a ceramic substrate and a copper plate is used as a circuit substrate for mounting semiconductor elements and the like. International Publication No. 2018 / 021472 (Patent Document 1) discloses a ceramic copper circuit substrate in which a ceramic substrate and a copper plate are joined. In Patent Document 1, a brazing material containing Ag, Cu, Ti, etc. is used in the joining layer. In Patent Document 1, the bonding layer is 1×10 - 3 Heat bonding is performed at a pressure of less than 100 Pa.

[0004] Such 1×10 -3 The pressure below Pa is called vacuum. In addition, the bonding method using Ti is called active metal bonding. Ti is easily nitrided or oxidized because it is an active metal. In the active metal bonding method, bonding is performed in a vacuum to prevent Ti from nitriding and oxidizing before bonding. To bond in a vacuum, the inside of the bonding device must be vacuumed. The temperature is increased after the vacuum is formed, so only batch processing can be performed in the heating bonding process. In the heating bonding process, each batch includes four steps: vacuuming, heating, bonding, and cooling. Each batch takes more than 24 hours. Therefore, mass productivity is not good.

[0005] On the other hand, in International Publication No. 2018 / 199060 (Patent Document 2), a heating bonding process is disclosed using a continuous furnace. In the embodiment of Patent Document 2, bonding is performed at a furnace length of 3m and a conveying speed of 10cm / minute. Including the holding time in the middle, the heating bonding can be performed in about 2 hours.

[0006] Prior art literature

[0007] Patent Literature

[0008] Patent Document 1: International Publication No. 2018 / 021472

[0009] Patent Document 2: International Publication No. 2018 / 199060 Summary of the invention

[0010] Problem that the invention aims to solve

[0011] In Patent Document 1 and Patent Document 2, an Ag-Cu-Ti based joining brazing material containing more than 40 mass % of Ag is used. Ag and Cu are components that form a eutectic. The solidification temperature of the AgCu eutectic is about 780°C. The joining temperature of the active metal joining method is 800°C to 950°C. The solidification temperature of the AgCu eutectic is about 780°C, so it begins to solidify during the joining process. In the joining process, after being maintained at the joining temperature, it is gradually cooled to room temperature. If the solidification temperature of the compound present in the joining layer is high, the thermal stress increases during the cooling process. The increase in thermal stress leads to an increase in the warping amount of the joint body.

[0012] In order to cope with such problems, the present invention provides a joined body in which a compound having a low solidification temperature or phase transition temperature exists in a joining layer.

[0013] Means used to solve problems

[0014] The joint body involved in the embodiment comprises a ceramic substrate, a copper plate and a joining layer. The joining layer is arranged on at least one side of the ceramic substrate and joins the ceramic substrate and the copper plate. The joining layer contains a Ti reaction layer and a plurality of first alloys. The Ti reaction layer contains titanium nitride or titanium oxide as a main component. The plurality of first alloys are located between the Ti reaction layer and the copper plate. The plurality of first alloys respectively contain one or more selected from Cu-Sn alloy and Cu-In alloy. The plurality of first alloys have different Sn concentrations or In concentrations. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram showing an example of a joined body according to the embodiment.

[0016] Figure 2 It is a schematic cross-sectional view showing an example of a joining layer of a joined body according to an embodiment.

[0017] Figure 3 This is a ternary phase diagram showing an example of the composition of a CuSnTi alloy.

[0018] Figure 4 It is a schematic diagram showing an example of the ceramic circuit board according to the embodiment.

[0019] Figure 5 It is a schematic diagram showing an example of a semiconductor device according to an embodiment. DETAILED DESCRIPTION

[0020] The joint body involved in the embodiment comprises a ceramic substrate, a copper plate and a joining layer. The joining layer is arranged on at least one side of the ceramic substrate to join the ceramic substrate and the copper plate. The joining layer contains a Ti reaction layer and a plurality of first alloys. The Ti reaction layer contains titanium nitride or titanium oxide as a main component. The plurality of first alloys are located between the Ti reaction layer and the copper plate. The plurality of first alloys respectively contain one or more selected from Cu-Sn alloy and Cu-In alloy. The plurality of first alloys have different Sn concentrations or In concentrations.

[0021] Figure 1 It is a schematic diagram showing an example of a joined body according to the embodiment. Figure 2 It is a schematic cross-sectional view showing an example of a joining layer of a joined body according to an embodiment. Figure 1 and Figure 2 In the figure, 1 is a bonded body, 2 is a ceramic substrate, 3 is a copper plate, 4 is a bonding layer, 5 is a Ti reaction layer, 6 is a first alloy, 7 is a second alloy, and 8 is a titanium silicide particle. Figure 2 In FIG. 1 , the first alloy 6 is indicated by a dotted line. The second alloy 7 is indicated by a filled dot. The titanium silicide particles 8 are indicated by a black color. Figure 1 In the illustrated joint body 1, copper plates 3 are provided on both sides of a ceramic substrate 2 via a joint layer 4. In addition, the longitudinal and transverse dimensions of the ceramic substrate 2 are respectively the same as the longitudinal and transverse dimensions of the copper plates 3. The joint body involved in the embodiment is not limited to such a form. The joint body may also have a structure in which the copper plate 3 is provided only on one side of the ceramic substrate 2. The longitudinal and transverse dimensions of the ceramic substrate 2 may also be different from the longitudinal and transverse dimensions of the copper plates 3.

[0022] As the ceramic substrate 2, silicon nitride substrate, aluminum nitride substrate, aluminum oxide substrate, Alusil high silicon heat-resistant aluminum alloy substrate, etc. can be cited. Alusil high silicon heat-resistant aluminum alloy substrate is a ceramic substrate mixed with aluminum oxide and zirconium oxide. The thickness of the ceramic substrate 2 is preferably 0.1 mm or more and 1 mm or less. If the substrate thickness is less than 0.1 mm, there is a possibility of strength reduction. In addition, if the substrate thickness is greater than 1 mm, there is a possibility that the ceramic substrate becomes a thermal resistor, which reduces the heat dissipation of the joint.

[0023] In addition, the three-point bending strength of the silicon nitride substrate is preferably 600 MPa or more. In addition, the thermal conductivity is preferably 80 W / m·K or more. By increasing the strength of the silicon nitride substrate, the substrate thickness can be reduced. Therefore, the three-point bending strength of the silicon nitride substrate is preferably 600 MPa or more, more preferably 700 MPa or more. The substrate thickness of the silicon nitride substrate can be reduced to less than 0.40 mm, and further reduced to less than 0.30 mm.

[0024] The three-point bending strength of the aluminum nitride substrate is in the range of 300 to 450 MPa. On the other hand, the thermal conductivity of the aluminum nitride substrate is 160 W / m·K or more. Since the strength of the aluminum nitride substrate is low, the substrate thickness is preferably 0.60 mm or more.

[0025] The three-point bending strength of an alumina substrate is in the range of 300 to 450 MPa, but the alumina substrate is inexpensive. In addition, the three-point bending strength of an alusil high-silicon heat-resistant aluminum alloy substrate is as high as about 550 MPa, but the thermal conductivity is in the range of 30 to 50 W / m·K.

[0026] As the ceramic substrate 2, it is preferably any one of a silicon nitride substrate and an aluminum nitride substrate. Silicon nitride substrates and aluminum nitride substrates are nitride ceramic substrates. Nitride ceramics react with an active metal brazing material containing Ti to form a reaction layer with titanium nitride as the main component. In addition, oxide ceramics react with an active metal brazing material containing Ti to form a reaction layer with titanium oxide as the main component. Oxide ceramics are aluminum oxide substrates and alusil high silicon heat-resistant aluminum alloy substrates. The Ti reaction layer 5 is a layer with titanium nitride as the main component or a layer with titanium oxide as the main component. A layer with titanium nitride as the main component or a layer with titanium oxide as the main component can be formed by the ceramic substrate 2 reacting with the Ti of the active metal brazing material.

[0027] It is preferred to arrange copper plates 3 on both sides of the ceramic substrate 2. By joining copper plates on both sides, warping of the joint body can be suppressed. As the copper plate 3, a pure copper plate or a copper alloy plate can be used. The copper plate 3 is preferably oxygen-free copper. As shown in JIS-H-3100 (ISO1337, etc.), the copper purity of oxygen-free copper is 99.96% by mass or more. Oxygen-free copper is a kind of pure copper. The copper plate 3 can be used as a circuit part or a heat sink. By thickening the copper plate 3, the current carrying capacity and heat dissipation can be improved. Therefore, the thickness of the copper plate 3 is preferably 0.6 mm or more, and more preferably 0.8 mm or more.

[0028] The bonding layer 4 includes a Ti reaction layer 5 whose main component is titanium nitride or titanium oxide. The Ti reaction layer 5 can be formed by using the so-called active metal bonding method. When the ceramic substrate 2 is a nitride-based ceramic, the main component of the Ti reaction layer 5 is titanium nitride (TiN). The main component of the Ti reaction layer 5, titanium nitride, may also contain a compound other than the one in which the atomic ratio of titanium and nitrogen is 1:1. When the ceramic substrate 2 is an oxide-based ceramic, the main component of the Ti reaction layer 5 is titanium oxide (TiO 2 The main component of the Ti reaction layer 5, titanium oxide, may also contain TiO 2 、TiO、Ti 2 O 3 Compounds other than those in which the atomic ratio of titanium to oxygen is 1:2.

[0029] The Ti reaction layer 5 refers to a region formed on the surface of the ceramic substrate 2 in which the Ti content is 50 mass % or more. In addition, it is preferred that the Ti reaction layer 5 is present in an area of ​​40% or more and 100% or less per unit length of 200 μm on the surface of the ceramic substrate 2. The so-called 40% or more per unit length of 200 μm means that when observing the cross section of the bonding layer 4, the Ti reaction layer 5 is formed in a total of 40 μm or more in 200 μm on the surface of the ceramic substrate 2.

[0030] The bonding layer 4 contains a plurality of first alloys 6 located between the Ti reaction layer 5 and the copper plate 3. The composition ratios of the plurality of first alloys 6 are different from each other. Specifically, the plurality of first alloys 6 have different Sn concentrations or In concentrations. The plurality of first alloys respectively contain one or more selected from Cu-Sn alloys and Cu-In alloys. Furthermore, all Cu-Sn alloys or Cu-In alloys present in the bonding layer 4 do not need to have different Sn concentrations or In concentrations. Two or more first alloys 6 having different Sn concentrations or In concentrations may be present in the bonding layer 4. Below, the example in which the first alloy 6 is a Cu-Sn compound is mainly described. In the following examples, the Cu-Sn compound may also be replaced by a Cu-In compound.

[0031] Furthermore, the term "different Sn concentrations" means that when the total of the Cu content and the Sn content is set to 100 atomic %, the Sn concentrations are different from each other by 2 atomic % or more. The term "different In concentrations" means that when the total of the Cu content and the In content is set to 100 atomic %, the In concentrations are different from each other by 2 atomic % or more.

[0032] The atomic ratio of Cu and Sn can be measured by point analysis of Energy Dispersive X-ray Spectroscopy (EDX). EDX is sometimes also called EDS. Here, SEM-EDX is used as EDX. As SEM, a field emission scanning electron microscope (FE-SEM) can also be used.

[0033] As the FE-SEM, JSM-7200F manufactured by JEOL Ltd. or an apparatus having equivalent performance thereto can be used, and as the EDX, EX-74600U4L2Q manufactured by JEOL Ltd. or an apparatus having equivalent performance thereto can be used.

[0034] FE-SEM was used to observe the cross section of the bonding layer 4. The cross section is a plane parallel to the thickness direction. The thickness direction is perpendicular to the surface of the ceramic substrate 2 and parallel to the direction connecting the ceramic substrate 2 and the copper plate 3. The measurement conditions of FE-SEM were set to an acceleration voltage of 15 kV and a magnification of 3000 times. The field of view was set to the thickness of the bonding layer × 40 μm in the width direction. The width direction was parallel to the surface of the ceramic substrate 2 and perpendicular to the thickness direction.

[0035] The measurement conditions of EDX are set to 50 scans and a dwell time of 0.2ms. The so-called dwell time refers to the measurement speed for each pixel. When performing regional analysis of EDX, the settings are as follows: the number of pixels read is 256 horizontally and 198 vertically, the number of detection counts is 3700 to 4100 cps (Count Per Second), and the quantitative image is 5×5 bit / point. Point analysis is the analysis of the measurement results using 1 detection count (1 place). Regional analysis is the analysis of the measurement results using multiple detection counts. In addition, regional analysis is sometimes also called surface analysis.

[0036] EDX-based analysis can also be performed after the elements that constitute the bonding layer are investigated in advance and the specific elements are identified. For example, when Cu (copper), Sn (tin), Ti (titanium), Si (silicon), O (oxygen), and N (nitrogen) are detected as elements that constitute the bonding layer, these elements are set as specific elements. By mapping specific elements using the results of regional analysis, the area ratio can be calculated. When the thickness of the bonding layer is 30 μm, the field of view area is 30 μm in the thickness direction × 40 μm in the width direction.

[0037] In the case of SEM-EDX mapping, the functions provided by SEM-EDX are used. In the case of no mapping function provided, imaging software is used for mapping.

[0038] The area ratio can be calculated by performing EDX regional analysis on three regions and using the average value. The area of ​​one region (field of view area) is the thickness of the bonding layer × 40 μm in the width direction. The field of view is measured corresponding to the bonding layer 4. Analyze any three adjacent regions from the cross section of the bonding layer 4. The bonding layer 4 is the range from the boundary between the ceramic substrate 2 and the bonding layer 4 to the boundary between the bonding layer 4 and the copper plate 3.

[0039] The boundary between the ceramic substrate 2 and the bonding layer 4 is the bonding interface between the surface of the ceramic substrate 2 and the Ti reaction layer 5. For example, when a silicon nitride substrate is used, the bonding interface between the silicon nitride substrate and the titanium nitride layer is the boundary between the ceramic substrate 2 and the bonding layer 4. Figure 2 The dashed line BL1 shown is the boundary between the ceramic substrate 2 and the bonding layer 4 .

[0040] The boundary between the bonding layer 4 and the copper plate 3 is defined based on the amount of Ti. Near the boundary between the bonding layer 4 and the copper plate 3, the amount of Ti gradually decreases from the bonding layer 4 toward the surface of the copper plate 3. The boundary is defined based on the area where the amount of Ti is less than 1 atomic % and is continuously formed in the width direction of 50 μm. When there are multiple such places, the boundary between the bonding layer 4 and the copper plate 3 is defined based on the place closest to the bonding layer 4 (ceramic substrate 2) among the multiple places. Figure 2 In the figure, the solid line SL between the bonding layer 4 and the copper plate 3 represents a collection of points where the Ti content is 1 atomic %. As an example, near the solid line SL, through the parts P1 and P2, a region where the Ti content is less than 1 atomic % is continuously formed in the width direction of 50 μm. The part P1 is located on the ceramic substrate 2 side compared to the part P2. Therefore, the dotted line BL2 based on the part P1 can be defined as the boundary between the bonding layer 4 and the copper plate 3.

[0041] EDX point analysis was used to determine the amount of Ti at the boundary between the bonding layer 4 and the copper plate 3. Point analysis was performed at 5 μm intervals to determine the amount of Ti. In addition, the area where the amount of Ti was 1 atomic % or less also included an area where the amount of Ti was 0 atomic % (below the detection limit).

[0042] The first alloy 6 refers to an alloy in which the total of Cu and the first element is 50 atomic % or more when the total of Cu, the first element and Ti is set to 100 atomic %. The first element is one or two selected from Sn and In. For example, with respect to a Cu-Sn alloy, when the total of Cu, Sn and Ti is set to 100 atomic %, the total of Cu and Sn is 50 atomic % or more. With respect to a Cu-In alloy, when the total of Cu, In and Ti is set to 100 atomic %, the total of Cu and In is 50 atomic % or more. The composition of the first alloy 6 can be obtained by EDX point analysis.

[0043] The second alloy 7 refers to an alloy in which the total of Ti and the first element is 50 atomic % or more when the total of Cu, the first element and Ti is set to 100 atomic %. For example, with respect to a Ti-Sn alloy, when the total of Cu, Sn and Ti is set to 100 atomic %, the total of Ti and Sn is 50 atomic % or more. With respect to a Ti-In alloy, when the total of Cu, In and Ti is set to 100 atomic %, the total of Ti and In is 50 atomic % or more. The composition of the second alloy 7 can be obtained by EDX point analysis.

[0044] The Cu-Sn alloy may be a state where Cu and Sn are mixed, or may contain a compound of Cu and Sn. The compound includes an intermetallic compound. It may also be a state where Ti is mixed in the Cu-Sn alloy or a state where Ti is solid-dissolved. The first alloy 6 includes an alloy having a melting point of 400° C. or higher and 600° C. or lower.

[0045] The melting point of the first alloy 6 can be investigated from the solidification point of the alloy. The so-called solidification point refers to the temperature at which a liquid becomes a solid. In other words, the temperature at which the bonding solder melts into a liquid phase and becomes a solid by cooling. In simple terms, the solidification point of the compound can also be investigated using a phase diagram. In addition, the bonding layer 4 can be cut out from the bonded body 1 and the temperature at which it melts by slowly heating can be measured.

[0046] The melting point of the first alloy 6 can be measured from the DSC curve. The DSC curve is the result of applying heat to the sample using a differential scanning calorimeter (DSC) to determine the presence or absence of an endothermic reaction and an exothermic reaction. If an endothermic reaction or an exothermic reaction occurs, a peak is generated in the DSC curve. The peak in the negative direction indicates that an endothermic reaction occurs. The peak in the positive direction indicates that an exothermic reaction occurs. An endothermic reaction indicates that melting, decomposition, etc. of the sample occurs. In addition, an exothermic reaction indicates that the formation or solidification of a compound (including an alloy) occurs by the mutual reaction of the constituent elements of the sample. The larger the peak, the greater the heat of reaction. Here, the peak in the negative direction is called an endothermic peak, and the peak in the positive direction is called an exothermic peak. In addition, the top of the peak is called the peak top. The difference between the maximum point and the minimum point of the peak is called the peak height.

[0047] For example, since the endothermic peak is a peak in the negative direction, the DSC curve drops near the endothermic peak and then gradually rises. Near the endothermic peak, the DSC curve has a maximum point → a minimum point → a maximum point. The minimum point (the point where the temperature drops the most in the negative direction) is the peak top. In addition, the larger value of the maximum point on the low temperature side and the high temperature side of the minimum point is taken as the maximum point (the point where the temperature rises the most in the positive direction). The peak height is the value obtained by subtracting the minimum point from the maximum point.

[0048] Since the exothermic peak is a peak in the positive direction, the DSC curve rises near the exothermic peak and then gradually decreases. Near the exothermic peak, the DSC curve has a minimum point → maximum point → minimum point. The maximum point (the point where the maximum rise is in the positive direction) is the peak top. In addition, the smaller value of the minimum point on the low temperature side and the high temperature side of the maximum point is the minimum point (the point where the maximum drop is in the negative direction). The peak height is the value obtained by subtracting the minimum point from the maximum point.

[0049] Furthermore, sometimes the end point of an endothermic peak (negative peak) is regarded as an exothermic peak (positive peak). Here, the end point is also counted as an exothermic peak. In addition, by drawing a baseline, the respective peak heights can also be obtained. In addition, in the DSC curve, in the vertical axis, changes above 0.02 mW / mg are counted as peaks. In other words, changes below 0.02 mW / mg are not counted as peaks.

[0050] In the DSC curve in the cooling process, an exothermic peak is preferably detected in the range of more than 600°C and less than 900°C. The exothermic peak in this range becomes the main body of the solidification reaction. Moreover, it is preferred to have an exothermic peak in the range of more than 400°C and less than 600°C. As mentioned above, the exothermic reaction indicates that the formation or solidification of a compound (including an alloy) occurs by the reaction of the solder components with each other. In other words, the exothermic peak in the range of more than 400°C and less than 600°C indicates that the formation or solidification reaction of a compound occurs within the temperature range. So-called solidification is a reaction accompanied by a phase change from a liquid to a solid. In addition, the formation of a compound is sometimes accompanied by a phase change from a solid to a solid with a different crystal structure. By causing such a phase change in multiple stages, the stress during cooling can be relaxed. If the solder composition is as described later, the phase change used to form the compound becomes the main reason for the occurrence of the exothermic peak. In other words, if the first alloy 6 is present, an exothermic peak can occur in the range of 400°C to 600°C of the DSC curve in the cooling process. By having exothermic peaks in the range of 400°C to 600°C and in the range of more than 600°C to 900°C, the reaction accompanied by phase change can be multi-stage. Here, the temperature at which the phase change from liquid to solid occurs is called the solidification point. The temperature at which the phase change to a solid having a different crystal structure occurs is called the phase change point.

[0051] Furthermore, when there are a plurality of alloys (including compounds) having different composition ratios, the exothermic peaks of these alloys may overlap. In other words, even when only one exothermic peak is detected, there may be a plurality of alloys (including compounds) having different composition ratios.

[0052] The solidification point is the temperature at which the liquid phase changes to the solid phase. The phase transition point is the temperature at which the solid phase changes to a solid having a different crystal structure. By solidifying or transforming in the range of 400°C to 600°C, the temperature at which thermal stress occurs can be lowered. The solidification point of the conventional AgCu eutectic compound is about 780°C. In the cooling process, the compound solidified at 780°C generates thermal stress before returning to room temperature. By having the first alloy 6 having a solidification point or phase transition point in the range of 400°C to 600°C, the thermal stress before returning to room temperature after solidification or phase transition at 400°C to 600°C can be reduced.

[0053] Since the multiple first alloys 6 have different composition ratios, their respective solidification points or phase transformation points are also different from each other. In the range of 400°C to 600°C, there are multiple first alloys 6 with different solidification points or phase transformation points. As a result, the solidification or phase transformation of the bonding layer can be performed in multiple stages. As a result, thermal stress can be further reduced. Thermal stress is generated in the process of the bonding brazing material melting into a liquid phase and then cooling to room temperature through solidification. In particular, it is important to reduce the thermal stress from solidification to cooling to room temperature. By causing the solidification or phase transformation of the first alloy 6 to occur in multiple stages, the thermal stress from solidification to cooling to room temperature can be reduced. If the solidification point or phase transformation point of the first alloy 6 is lower than 400°C, there is a possibility that the solidification point or phase transformation point of the bonding layer 4 is lowered, and the temperature cycle test (TCT) characteristics are reduced. If the solidification point or phase transformation point of the first alloy 6 is as high as more than 600°C, there is a possibility that the thermal stress increases. Therefore, the solidification point or phase transformation point of the first alloy 6 is preferably 400° C. to 600° C., more preferably 450° C. to 550° C. In addition, even if the exothermic peaks of the DSC curve in the temperature drop process at 400° C. to 600° C. overlap, the presence of a plurality of first alloys 6 having different composition ratios can provide a multi-stage solidification or phase transformation effect.

[0054] In addition, it is preferred that the first alloy 6 contains one or more selected from Cu-Sn compounds and Cu-In compounds. In the compound, the constituent elements are bonded to each other at the atomic level. As compounds, intermetallic compounds can be listed. As alloys, solid solutions, eutectics, etc. can be listed. Compounds are distinguished from solid solutions and eutectics. It is possible to specify the phase structure according to the crystal structure and composition. For example, in a binary alloy of CuSn, when the Sn amount is less than 9 atomic %, the alloy is mainly a solid solution. When the Sn amount is more than 10 atomic %, the alloy easily forms intermetallic compounds.

[0055] For at least a portion of the plurality of first alloys 6, the Sn content is preferably in the range of 3 atomic % to 30 atomic %. If the Sn content is in the range of 3 atomic % to 30 atomic %, it is easy to control the solidification point or phase transition point. If the Sn amount is less than 3 atomic %, the solidification point or phase transition point cannot be sufficiently lowered. In addition, if the Sn amount exceeds 30 atomic %, there is a possibility that the solidification point or phase transition point is excessively lowered. Therefore, the Sn content is preferably 3 atomic % to 30 atomic %. In addition, the first alloy 6 having a Sn content of 3 atomic % to 30 atomic % is prone to form a Cu-Sn compound.

[0056] When EDX point analysis is performed on a region of a Cu-Sn alloy having more Cu in EDX regional analysis, preferably more than 80% of the region is a Cu-Sn alloy having a Sn content of more than 3 atomic % and less than 30 atomic %. EDX point analysis is performed more than 10 times. The so-called region having more Cu refers to a region having more Cu by comparing the atomic % of Cu and the atomic % of Ti through EDX regional analysis.

[0057] In addition, the first alloy 6 may also contain less than 2 atomic % of metal components other than Cu and Sn. Metal components other than Cu and Sn are metal components or impurities that constitute the joining brazing material. As constituent components of the joining brazing material other than Cu and Sn, Ti (titanium) and C (carbon) can be listed as described later. For example, the solidification point of the TiSn alloy exceeds 600°C. If the amount of Ti in the Cu-Sn alloy increases, there is a possibility that the solidification point or phase transformation point will not reach the range of more than 400°C and less than 600°C. Therefore, in at least a portion of the plurality of first alloys 6, the amount of Ti is preferably less than 2 atomic %, more preferably less than 1 atomic %. In the determination of the amount of Ti in the Cu-Sn compound, EDX point analysis is used. As components other than metal components, oxygen and nitrogen can be listed. The first alloy 6 may also contain less than 10 atomic % of Cu, Sn and components other than metal components.

[0058] If the bonding solder contains Ag, AgCu eutectic is easily formed. The melting point of AgCu eutectic is about 780°C. Therefore, as a constituent component of the bonding solder, it is preferred that Ag is not contained. In other words, it is preferred that the bonding layer 4 does not contain Ag. As a bonding solder, Ag-Cu-Sn-Ti with Ag as the main component is sometimes used. If Ag is the main component, AgCu or AgSn is formed. The proportion of the CuSn alloy contained in the bonding layer 4 is reduced. Therefore, the possibility of forming a Cu-Sn alloy with a different composition is reduced. From this point of view, it is preferred that the bonding layer 4 does not contain Ag.

[0059] Preferably, a portion of the plurality of first alloys 6 is a Cu-Sn alloy having a Sn content of 3 atomic % to 9 atomic %, and another portion of the plurality of first alloys 6 is a Cu-Sn alloy having a Sn content of 10 atomic % to 30 atomic %.

[0060] Here, the Cu-Sn alloy with Sn content of 3 atomic % or more and 9 atomic % or less is referred to as the first Cu-Sn alloy, and the Cu-Sn alloy with Sn content of 10 atomic % or more and 30 atomic % or less is referred to as the second Cu-Sn alloy. The first Cu-Sn alloy is preferably a Cu-Sn compound. The second Cu-Sn alloy is preferably a Cu-Sn compound. The compound is because it is easy to perform multi-stage solidification or phase transformation.

[0061] The bonding layer 4 preferably contains the second alloy 7. The second alloy 7 contains one or two selected from Ti-Sn alloy and Ti-In alloy. The Ti-Sn alloy or Ti-In alloy may also contain an intermetallic compound. For example, as the intermetallic compound of the Ti-Sn alloy, SnTi 3 Sn 3 Ti 5 Sn 5 Ti 6 One or more of the following.

[0062] In the average values ​​measured at 3 locations of the bonding layer 4, the total area of ​​the plurality of first alloys 6 is preferably greater than the total area of ​​the second alloy 7. That is, the total area of ​​the Cu-Sn alloy or the Cu-In alloy is preferably greater than the total area of ​​the Ti-Sn alloy or the Ti-In alloy. The viewing area is set to the thickness of the bonding layer × 40 μm in the width direction. The measurement in one viewing area can also be performed multiple times. In addition, when measuring multiple times, each area is set in a non-overlapping manner.

[0063] In addition, at least a portion of the Cu-Sn alloy and the Ti-Sn alloy detected by EDX point analysis of the bonding layer 4 are preferably within a region surrounded by (97, 3, 0), (60, 40, 0), (2, 40, 58), and (39, 3, 58) in the ternary phase diagram of Cu, Sn, and Ti. This region is referred to as the first composition region.

[0064] Figure 3 It is the ternary phase diagram of Cu, Sn and Ti. Figure 3 The concentration of each element is expressed in atomic %.

[0065] First, the regional analysis function of SEM-EDX is used to obtain image data through element mapping. Through the mapping function, EDX point analysis is performed on the places judged as the first alloy 6 and the second alloy 7. According to element mapping, the concentration difference of a specific element can be represented by brightness. The brightness of the area with high concentration is high, and the brightness of the area with low concentration is low. For example, as long as Ti is specified as a specific element, the concentration of Ti can be visualized. By adopting this function, the first alloy 6 and the second alloy 7 present in the field of view area can be distinguished. That is, by specifying the first element as a specific element, the concentration of these elements is visualized, and the first alloy 6 present in the field of view area can be distinguished. In addition, by specifying Ti as a specific element, the concentration of these elements is visualized, and the second alloy 7 present in the field of view area can be distinguished. When EDX point analysis is performed on the places judged as the first alloy 6 and the second alloy 7, it is preferred that the composition is within the range of the first composition area. There may also be alloys outside the range of the first composition area, but the characteristics can be improved by the alloy composition being within the range of the first composition area.

[0066] More preferably, when performing EDX point analysis, the composition of at least a portion of the Cu-Sn alloy is within the region surrounded by (97, 3, 0), (77, 23, 0), (52, 23, 25), and (72, 3, 25) in the ternary phase diagram of Cu, Sn, and Ti. This region is referred to as the second composition region.

[0067] The first alloy 6 having a composition other than the second composition region may also exist in the bonding layer 4. On the other hand, if the first alloy 6 having a composition in the second composition region is more, the characteristics can be further improved. In addition, when the Cu-Sn alloy is observed by element mapping based on SEM-EDX regional analysis and the Cu-Sn alloy is subjected to EDX point analysis, it is preferred that more than 80% of the analyzed places have a composition within the range of the second composition region. Furthermore, in the EDX point analysis, more than 10 points are analyzed at places separated by more than 3 μm from each other.

[0068] The Ti-Sn alloy detected by EDX point analysis of the bonding layer 4 is preferably within the area surrounded by (41, 23, 36), (24, 40, 36), (2, 40, 58), and (19, 23, 58) in the ternary phase diagram of Cu, Sn, and Ti. This area is called the third composition area. When the bonding layer 4 contains multiple Ti-Sn alloys, the composition of at least a portion of the multiple Ti-Sn alloys is preferably within the range of the third composition area. On the other hand, if there are more Ti-Sn alloys having a composition within the range of the third composition area, the characteristics can be further improved. In addition, when the Ti-Sn alloy is observed by element mapping based on SEM-EDX regional analysis and the Ti-Sn alloy is subjected to EDX point analysis, it is preferred that more than 80% of the analyzed area have a composition within the range of the third composition area. Furthermore, in the EDX point analysis, more than 10 points are analyzed at locations separated from each other by more than 3μm.

[0069] The Ti reaction layer 5 preferably contains titanium nitride particles with an average particle size of less than 50nm. In addition, it is preferred that Cu and the first element exist at the grain boundaries of the titanium nitride particles. The size of the titanium nitride particles located in the Ti reaction layer 5 can be analyzed using a scanning transmission electron microscope (STEM). An enlarged photograph of the cross-section of the bonding layer can be obtained by STEM. In the outer edge of the titanium nitride particles reflected in the enlarged photograph, the distance between the two farthest points is taken as the particle size. The particle size is measured for any 30 particles. Their average value is taken as the average particle size. By making the average particle size as small as less than 50nm, Cu and the first element can easily enter the grain boundaries. By having Cu and the first element in the grain boundaries of the titanium nitride particles, the solidification or phase change of the bonding layer 4 can be further multi-staged.

[0070] It is preferred that the amount of Cu in the Ti reaction layer 5 is within a range of 0.5 atomic % and above and 5 atomic % or below. If the amount of Cu in the Ti reaction layer 5 is within this range, thermal stress can be reduced without reducing the bonding strength. If the amount of Cu in the Ti reaction layer 5 is less than 0.5 atomic %, the effect of reducing thermal stress cannot be fully obtained due to the small amount of Cu. If the amount of Cu exceeds 5 atomic %, there is a possibility that the bonding strength is reduced due to insufficient strength of the Ti reaction layer 5.

[0071] The amount of the first element in the Ti reaction layer 5 is preferably 0.05 atomic % or more and 2 atomic % or less. If the amount of the first element in the Ti reaction layer 5 is within this range, thermal stress can be reduced without reducing the bonding strength. In addition, Cu and Sn in the Ti reaction layer 5 are easily alloyed. By forming an alloy (preferably a compound) with Cu and Sn in the Ti reaction layer 5, the occurrence of thermal stress can be suppressed. The amount of Cu and the first element in the Ti reaction layer 5 refers to the amount present in the grain boundaries of the titanium nitride particles.

[0072] As titanium nitride in the Ti reaction layer 5, TiN and Ti 2 N on both sides. By the presence of TiN and Ti 2 N, the solidification of the layer mainly composed of titanium nitride can further occur in multiple stages. The thickness of the Ti reaction layer 5 is preferably 1 μm or less. More preferably, the thickness of the Ti reaction layer 5 is 0.3 μm or less. By thinning the Ti reaction layer 5, Cu and the first element can be easily present at the grain boundary.

[0073] The thickness of the Ti reaction layer 5 can be analyzed by a transmission electron microscope (TEM). The thickness of the Ti reaction layer 5 is measured based on the interface between the Ti reaction layer 5 and the ceramic substrate. TEM-EDX is used to analyze the composition of the Ti reaction layer 5.

[0074] In the average value obtained by performing EDX regional analysis on the three regions of the bonding layer 4, the total area of ​​the first alloy 6 is preferably more than 50% and less than 95% of the area of ​​the three regions. The field of view area (the size of one region) can be set to the bonding layer thickness × 40μm in the width direction. If the total area ratio of the Cu-Sn alloy is within the range of more than 50% and less than 95%, the thermal stress of the bonding layer 4 can be reduced. If the total area ratio of the Cu-Sn alloy is less than 50%, the existence ratio of the Cu-Sn alloy is small, so there is a possibility that the effect of reducing thermal stress is insufficient. In addition, if the total area ratio of the Cu-Sn alloy is as high as more than 95%, there is a possibility that the solidification point or phase transformation point of the bonding layer 4 is too low. If the solidification point or phase transformation point of the bonding layer 4 decreases, there is a possibility that the TCT characteristics of the joint 1 decrease. Therefore, the total area ratio of the Cu-Sn alloy is preferably more than 50% and less than 95%, and more preferably more than 70% and less than 90%. Furthermore, no matter which three regions in the bonding layer 4 are measured, the total area of ​​the Cu—Sn alloy is preferably within a range of 50% to 95%.

[0075] In the average value obtained by performing EDX regional analysis on the three regions of the bonding layer 4, the total area of ​​the Ti-Sn alloy is preferably in the range of 5% or more and 30% or less of the area of ​​the three regions. The total area of ​​the titanium silicide is preferably in the range of 0.5% or more and 15% or less of the area of ​​the three regions. In addition, the area ratio of other than this is preferably in the range of 10% or less. As material structures other than this, Ti reaction layer, carbide, etc. can be listed.

[0076] In the measurement of the area ratio, the regional analysis function of SEM-EDX is used to obtain image data through element mapping. The mapping function attached to the regional analysis function of SEM-EDX is used. When the mapping function is not attached, electron probe microanalysis (EPMA) can also be used. In addition, image analysis software can be used to process the image data obtained by SEM-EDX. As the image analysis software, Image-J or equivalent software can be used.

[0077] The viewing area can be set to the thickness of the bonding layer × 40μm in the width direction. The measurement conditions for regional analysis are as described above. The area ratio is calculated using the element distribution map obtained by regional analysis. For the element distribution map, the click size is set to 5×5bit / point. When using image analysis software such as Image-J, binarization is performed.

[0078] The thinking method of binarization is as follows. First, set the area of ​​the element x to be detected to Sx. It can be calculated by Sx = ((the number of pixels in the extraction area based on the binarization of the x element) / (the total number of pixels in the image)) × 100. By changing the element x to be detected to Ti, Si, N, and C, the area can be calculated respectively.

[0079] For example, the area ratio of the Ti-Sn alloy can be calculated as follows. As mentioned above, in the second composition region of the Cu-Sn alloy (the first alloy 6), the Ti amount is greater than 0 atomic % and less than 25 atomic %. Therefore, the area where the Ti amount is less than 25 atomic % is excluded. The element distribution map can be represented by a brightness corresponding to the Ti concentration. By adopting this function, the area where the Ti amount is less than 25 atomic % can be excluded. By binarizing the image of the element distribution map in which the area where the Ti amount is less than 25 atomic % is removed, the area of ​​the Ti element and the area of ​​the Si element are measured. The area of ​​the Ti-Sn alloy is set to S Sn-Ti , set the area of ​​the Ti element to S Ti , set the area of ​​Si element to S Si , set the area of ​​N element to S N , set the area of ​​the C element to S C By S Sn-Ti =S Ti -(S Si +S N +S C ) can be used to determine the area of ​​the Ti-Sn alloy. The Si in the bonding layer is titanium silicide. The N (nitrogen) in the bonding layer is titanium nitride. If the boundary between the silicon nitride substrate surface and the bonding layer is curved, silicon nitride may also be detected. In addition, the C (carbon) in the bonding layer is carbon or titanium carbide. Ti , S Si , S N , S C Alternatively, only Ti, Si, N, and C may be mapped, the resulting images may be binarized, and calculations may be performed using these images.

[0080] The area ratio of Cu-Sn alloy is set to S Cu-Sn . S Cu-Sn By removing S from the region where the Ti content is 0 atomic % or more and 25 atomic %, Si , S N , S C Come and ask for it.

[0081] Ti element is sometimes contained in both Cu-Sn alloy and Ti-Sn alloy. Cu-Sn alloy and Ti-Sn alloy can be distinguished based on 25 atomic % of Ti. The bonding layer 4 may also contain carbon. The carbon present in the bonding layer 4 may be a single carbon substance or a carbide. By adding carbon, the fluidity of the bonding brazing material can be controlled. Thus, the distribution of the Cu-Sn alloy in the bonding layer can be easily controlled.

[0082] As materials other than the Cu-Sn alloy present in the bonding layer 4, titanium nitride (TiN), Ti-Sn alloy, and titanium silicide can be listed. In addition, when carbon is contained, titanium carbide (TiC) may also be present. In addition, the components constituting the bonding layer 4 may also exist as a metal single substance. Furthermore, the alloy and the metal single substance can be distinguished from the electron diffraction image of the TEM.

[0083] The bonding layer 4 preferably contains titanium silicide particles 8 having a major diameter of 0.5 μm or more and 6 μm or less. In the field of view area of ​​the bonding layer 4 = bonding layer thickness × 200 μm in the width direction, the total area of ​​the titanium silicide particles 8 having a major diameter of 0.5 μm or more and 6 μm or less is preferably 1% or more and 15% or less of the field of view area.

[0084] In the determination of the long diameter of the titanium silicide particle 8, an SEM photograph is used. In the outer edge of the titanium silicide particle 8 shown in the enlarged photograph of the cross section of the bonding layer 4, the distance between the two points farthest apart is taken as the long diameter. The titanium silicide particles 8 present in the bonding layer 4 can be identified by the mapping function included in the regional analysis function of the SEM-EDX. Specifically, Ti and Si are designated as specific elements, and the presence of these elements is visualized. The area where both Ti and Si exist corresponds to the titanium silicide particle 8.

[0085] When titanium silicide particles 8 are present, preferably 80% or more and 100% or less of the titanium silicide particles 8 are present within 1 μm from the Ti reaction layer 5. If the Ti reaction layer 5 is not formed or the Ti of the Ti-Sn alloy reacts with the silicon nitride substrate, titanium silicide particles are formed. When the titanium silicide particles 8 are present near the Ti reaction layer 5, Cu and the first element easily enter the grain boundaries of the titanium nitride particles.

[0086] If the major diameter of the titanium silicide particles 8 exceeds 6 μm or the area ratio exceeds 15%, there is a possibility that the formation of the Ti reaction layer 5 is insufficient.

[0087] The thickness of the bonding layer 4 is preferably 5 μm or more and 60 μm or less. Within this range, the area ratio of the Cu-Sn alloy can be easily controlled. The thickness of the bonding layer 4 is the distance from the boundary between the ceramic substrate 2 and the Ti reaction layer 5 to the boundary between the bonding layer 4 and the copper plate 3.

[0088] The concentration (atomic %) of the first element in the boundary between the bonding layer 4 and the copper plate 3 is referred to as the first concentration C1. The concentration (atomic %) of the first element in the region of the copper plate 3 10 μm away from the boundary between the bonding layer 4 and the copper plate 3 is referred to as the second concentration C2. Preferably, the ratio C2 / C1 of the second concentration C2 to the first concentration C1 is 0.60 or less.

[0089] In addition, the concentration (atomic %) of the first element in the region of the copper plate 3 at a distance of 20 μm from the boundary between the bonding layer 4 and the copper plate 3 is referred to as the third concentration C3. The ratio C3 / C1 of the third concentration C3 to the first concentration C1 is preferably 0.40 or less. When a pure copper plate is used as the copper plate 3, the second concentration C2 and the third concentration C3 represent the diffusion amount of the first element into the copper plate 3.

[0090] When Sn is used in the bonding layer 4, the concentration of Sn is used as the concentration of the first element. When In is used in the bonding layer 4, the concentration of In is used as the concentration of the first element. When both Sn and In are used in the bonding layer 4, the total of the concentration of Sn and the concentration of In is used as the concentration of the first element.

[0091] For example, when the bonding layer 4 contains a Cu-Sn alloy, the first concentration is described as C1 Sn , record the second concentration as C2 Sn , record the third concentration as C3 Sn The so-called C2 Sn / C1 Sn ≤0.60 or C3 Sn / C1 Sn ≤0.40, indicating that the diffusion of Sn into the copper plate is suppressed.

[0092] By having the first alloy 6 with different composition ratios in the bonding layer 4, the diffusion amount of the first element into the copper plate 3 can be reduced. In addition, the diffusion distance of the first element from the above-mentioned boundary is preferably less than 50 μm. The so-called diffusion area refers to the distance from the above-mentioned boundary to the point where the Sn amount is 0 mass% (below the detection limit). In addition, if the bonding method using a continuous furnace is used as described later, the first concentration can be less than 8 atomic %. The second concentration C2 and the third concentration C3 in the copper plate 3 can be analyzed by EDX. Line analysis is used in the analysis.

[0093] By reducing the amount of Sn diffused into the copper plate 3, the heat resistance of the copper plate 3 can be improved. For example, if Sn diffuses into the copper plate 3, a CuSn alloy is formed in the copper plate. The CuSn alloy has a lower melting point than the copper plate. Therefore, if the CuSn alloy in the copper plate increases, the copper plate is easily deformed due to heat. There is a possibility that the TCT characteristics of the joint body will decrease. Although Sn is used for illustration here, the same is true for In.

[0094] By having the above-mentioned structure, the amount of voids in the bonding layer can be reduced. By reducing the amount of Sn diffusion (or In diffusion) to the copper plate, the amount of voids in the bonding layer can be made to be greater than 0% and less than 3% by volume. The amount of voids in the bonding layer is calculated by ultrasonic flaw detection (SAT). The probe for flaw detection is brought into contact with the surface of the bonded body 1, and ultrasonic waves are transmitted toward the bonding layer 4. The voids present in the bonding layer can be investigated from the reception results of the reflected waves. In addition, the volume ratio of the void amount can be calculated from the reception results. By reducing the amount of voids in the bonding layer, the bonding strength of the copper plate can be improved. The bonding strength can be measured by a peel test. The bonding strength of the copper plate can be set to be greater than 15 kN / mm, and further set to be greater than 25 kN / mm.

[0095] The above-described bonded body is suitable for use in a ceramic circuit substrate.

[0096] Figure 4 It is a schematic diagram showing an example of the ceramic circuit board according to the embodiment. Figure 4 In the figure, 10 is a ceramic circuit substrate, 11 is a circuit portion, and 12 is a heat sink. Figure 4 The ceramic circuit board 10 shown can be produced by processing the copper plate 3 on the front side of the bonded body 1 into the circuit portion 11 and processing the copper plate 3 on the back side into the heat sink 12. The copper plate 3 with a circuit shape is used as the circuit portion 11. Figure 4 In the embodiment, two circuit parts 11 are provided. Furthermore, the ceramic circuit substrate 10 involved in the embodiment is not limited to such a structure. The number and size of the circuit parts 11 can be appropriately changed as needed. In addition, Figure 4 In the embodiment, the copper plate 3 on the back side is used as the heat sink 12, but the copper plates 3 on both sides may be given a circuit structure. As required, an inclined shape may be given to the side surface of the circuit portion 11 or the side surface of the heat sink 12. As required, a protrusion protruding from the side end of the circuit portion 11 may be provided on the bonding layer 4 between the ceramic substrate 2 and the circuit portion 11. A protrusion protruding from the side end of the heat sink 12 may also be provided on the bonding layer 4 between the ceramic substrate 2 and the heat sink 12. It is preferred to use an etching process to impart the circuit structure and the inclined shape to the side surface of the copper plate.

[0097] The ceramic circuit substrate according to the embodiment is suitable for a semiconductor device.

[0098] Figure 5 It is a schematic diagram showing an example of a semiconductor device according to an embodiment. Figure 5 In the figure, 20 is a semiconductor device, 21 is a semiconductor element, and 22 is a lead frame. Figure 5In the semiconductor device 20 shown, a semiconductor element 21 is mounted on one of the two circuit portions 11. A lead frame 22 is connected to the other of the two circuit portions 11. The semiconductor device 20 according to the embodiment is not limited to such a structure. The number and size of the circuit portions 11, the number and size of the semiconductor elements 21, etc. can be appropriately changed as needed.

[0099] Next, the manufacturing method of the bonded body according to the embodiment will be described. As long as the bonded body according to the embodiment has the above-mentioned structure, the manufacturing method thereof is not limited. Hereinafter, an example of a method for obtaining a bonded body with a high yield will be described.

[0100] First, prepare a ceramic substrate 2. Examples of the ceramic substrate 2 include a silicon nitride substrate, an aluminum nitride substrate, an aluminum oxide substrate, and an alusil substrate. The alusil substrate is a substrate formed by mixing aluminum oxide and zirconium oxide.

[0101] A pure copper plate or a copper alloy plate can be used as the copper plate 3. The copper plate is preferably oxygen-free copper. Oxygen-free copper is copper having a purity of 99.96 mass % or more as shown in JIS-H-3100.

[0102] Next, prepare a bonding brazing material. Preferably, the bonding brazing material contains 50 atomic % or more of Cu, 5 atomic % or more and 40 atomic % or less of Ti, 2 atomic % or more and 35 atomic % or less of the first element, and 0 atomic % or more and 12 atomic % or less of carbon. This is the ratio when the total of Cu, Ti, the first element and C is set to 100 atomic %. Ti can also be in the form of titanium hydride (TiH 2 ) is added. It is effective not to add Ag to the bonding solder.

[0103] The properties of the bonding brazing material used in manufacturing the bonded body are preferably investigated in advance by DSC. The DSC curve of the bonding brazing material is measured by setting the heating rate in the heating process to 20°C / min and the cooling rate in the cooling process to 20°C / min.

[0104] As DSC, the TGA-DSC simultaneous thermal analysis device STA449-F3-Jupiter manufactured by NETZSCH or a device with equivalent performance can be used. The measurement is performed by dropping an appropriate amount of solder into an alumina container in an Ar (argon) gas flow. Cover the alumina container with a lid and perform the measurement. Since the measurement is performed in an Ar atmosphere, it is necessary to prevent the solder from reacting with the atmosphere. Furthermore, the amount of dripping (mg) is measured in advance using a balance. Use a sample of 15 mg or more.

[0105] In the heating process, the temperature is raised from room temperature to 950°C at a heating rate of 20°C / min. Next, a holding process is performed at 950°C for 20 minutes. In the cooling process, the temperature is lowered from 950°C to room temperature at a cooling rate of 20°C / min. The temperature curve for obtaining the DSC curve is composed of a heating process, a holding process, and a cooling process. Hereinafter, the DSC curve based on the cooling process at a cooling rate of 20°C / min is sometimes referred to as the DSC curve of the cooling process. In addition, the curve with a difference of 0.02mW / mg or more in heat flow is regarded as a peak.

[0106] If there is a first alloy 6 having a solidification point or a phase transformation point of 400°C or more and 600°C or less, an exothermic peak can be detected in the range of 400°C or more and 600°C or less of the DSC curve of the cooling process. When multiple exothermic peaks are detected between 400°C and 600°C of the DSC curve of the cooling process, the temperature at which the largest peak appears is taken as the solidification point or phase transformation point of the bonding layer 4.

[0107] The average particle size D of the Cu powder used as the brazing material 50 The particle size is preferably 12.0 μm or less, and more preferably 10.0 μm or less. Ti powder or TiH 2 The average particle size of the powder is D 50 The average particle size D of Sn powder or In powder is preferably 6.0 μm or less, and more preferably 4.0 μm or less. 50 It is preferably 16.0 μm or less, and more preferably 14.0 μm or less. The average particle size D of the C powder 50 The average particle size D of the Ag powder is preferably 6.0 μm or less, and more preferably 4.0 μm or less. 50 It is preferably 3.0 μm or less, and more preferably 2.0 μm or less. By controlling the particle size of the powder, the reaction of each powder can be made uniform.

[0108] The average particle size D of the Cu powder is preferably 50 Smaller than the average particle size D of Sn powder or In powder 50 As mentioned above, Cu and the first element become the base materials of the brazing material composition. The first element has a lower melting point than Cu. By increasing the particle size of Sn powder or In powder, the reaction of Cu powder and the powder of the first element can be made uniform. Thus, even if the heating rate or the cooling rate is increased, the ceramic substrate 2 and the copper plate 3 can be well bonded.

[0109] The process of bonding the ceramic substrate and the copper plate is performed by using the above bonding brazing material. The bonding brazing material is mixed with an organic substance to prepare a brazing paste. The brazing paste is applied to the surface of the ceramic substrate 2 (or the copper plate 3) to form a brazing paste layer. The copper plate 3 (or the ceramic substrate 2) is arranged on the brazing paste layer.

[0110] The heating bonding process can be performed in a nitrogen atmosphere. The so-called nitrogen atmosphere is an atmosphere in which nitrogen is 70 vol% or more. The nitrogen content of the nitrogen atmosphere is preferably 70 vol% or more, and more preferably 85 vol% or more and 100 vol% or less. As components other than nitrogen in the nitrogen atmosphere, inert gas or air can be cited. The heating bonding in the nitrogen atmosphere can be performed under various conditions such as normal pressure, reduced pressure or pressurized pressure.

[0111] In the heating process in the heating bonding process, the heating rate is set to be 30°C / min or more. The upper limit of the heating rate is not particularly limited, but is preferably 100°C / min or less. If the heating rate is as high as more than 100°C / min, there is a possibility that the bonding property is reduced. Therefore, the heating rate is preferably 30°C / min or more and 100°C / min or less, and more preferably 40°C / min or more and 70°C / min or less.

[0112] The bonding temperature can be set at 750°C or higher. The bonding temperature is preferably 1000°C or lower. If the bonding temperature is as high as more than 1000°C, the bonding temperature approaches the melting point of copper (1085°C). As a result, there is a possibility that the copper plate will deform. Therefore, the bonding temperature is preferably 750°C or higher and 1000°C or lower, and more preferably 800°C or higher and 950°C or lower.

[0113] In the heating and joining process, the joining temperature can be maintained for more than 10 minutes. The holding time at the joining temperature is referred to as the heating holding time. The heating holding time is preferably more than 10 minutes and less than 100 minutes. If the heating holding time is less than 10 minutes, there is a possibility that the time for the solder to melt and solidify is insufficient. If the heating holding time is longer than 100 minutes, there is a possibility that the alloy distribution in the joining layer 4 is difficult to control.

[0114] After the heating holding time is over, a cooling process is performed. The cooling process is a process of cooling from the joining temperature to room temperature. The cooling rate can be set at more than 30°C / minute. In the previous active metal joining method, the cooling rate was about 5°C / minute. By adopting a joining brazing material having a DSC curve as described above, joining can be performed even if the cooling rate is increased. There is no particular limit on the upper limit of the cooling rate, but it is preferably less than 100°C / minute. If it is as high as more than 100°C / minute, there is a possibility of decreased bondability. Therefore, the cooling rate is preferably more than 30°C / minute and less than 100°C / minute, and more preferably more than 30°C / minute and less than 70°C / minute. In addition, in order to increase the heating rate and the cooling rate, a continuous furnace is preferably used. The continuous furnace can perform heat treatment in a nitrogen atmosphere. Since there is no need to evacuate, the heating rate and the cooling rate can be increased.

[0115] By increasing the cooling rate, the distribution of the Cu-Sn alloy in the bonding layer 4 can be controlled. This is because by rapidly cooling the bonding brazing material after it melts, an alloy that matches the solidification point or phase transition point can be formed. In addition, by the presence of Ti that does not form the Ti reaction layer 5, Ti-Sn alloy and titanium silicide particles can also be formed. Rapid cooling can suppress unnecessary diffusion into the copper plate.

[0116] The difference between the heating rate and the cooling rate is preferably less than 20°C / min. In other words, it is preferred to satisfy |heating rate - cooling rate|≤20°C / min. In the brazing material layer, melting reaction and solidification reaction are generated by heating. These reactions occur in the heating process and the cooling process. By reducing the difference between the heating rate and the cooling rate, the stress generated in the reaction can be uniformized. Thus, undesirable conditions such as warping can be suppressed.

[0117] If necessary, the heat bonding step may be performed while placing a load on the bonded body 1 .

[0118] The bonded structure 1 can be manufactured by the above-described bonding process. Furthermore, since the thermal stress of the bonding layer 4 can be reduced, the warping of the bonded structure 1 can be reduced.

[0119] The obtained bonded body 1 is etched to be processed into a ceramic circuit board 10. A plurality of bonded bodies 1 can be processed simultaneously by scribing the bonded body 1 as necessary.

[0120] (Example)

[0121] (Examples 1 to 8, Comparative Example 1)

[0122] The bonding brazing materials shown in Tables 1 and 2 were prepared. Table 1 shows the particle size of the raw material powder, and Table 2 shows the composition ratio.

[0123] Table 1

[0124]

[0125] Table 2

[0126]

[0127] The solder pastes were prepared by mixing the solder components of the examples and comparative examples with an organic binder, and the DSC curves of the solder pastes were measured using a differential scanning calorimeter.

[0128] As DSC, a TGA-DSC simultaneous thermal analyzer STA449-F3-Jupiter manufactured by NETZSCH was used. The measurement was performed by dropping an appropriate amount of brazing material into an alumina container in an Ar gas flow. The temperature program was set to a heating rate of 20°C / min, 950°C×20 minutes, and a cooling rate of 20°C / min.

[0129] The temperature at which the exothermic peak in the DSC curve of the temperature drop step was detected was measured. Table 3 shows the temperature at which the maximum peak appeared among the exothermic peaks of 400° C. or higher and 600° C. or lower in the DSC curve of the temperature drop step.

[0130] Table 3

[0131]

[0132] As can be seen from Table 3, for the bonding brazing materials involved in the examples, the temperature of the exothermic peak in the cooling process of the DSC curve is within the range of 400°C to 600°C. In contrast, in Comparative Example 1, the temperature of the exothermic peak exceeds 600°C. No exothermic peak is detected within the range of 400°C to 600°C. In addition, for the bonding brazing materials involved in the examples, an exothermic peak is detected even within the range of more than 600°C to 900°C.

[0133] Next, a silicon nitride substrate was prepared as a ceramic substrate. A silicon nitride substrate having a thermal conductivity of 90 W / m·K, a three-point bending strength of 600 MPa, a length of 300 mm×width of 200 mm×a thickness of 0.32 mm was used as silicon nitride substrate 1. In addition, a silicon nitride substrate having a thermal conductivity of 85 W / m·K, a three-point bending strength of 650 MPa, a length of 300 mm×width of 200 mm×a thickness of 0.25 mm was used as silicon nitride substrate 2.

[0134] Oxygen-free copper of 300 mm in length, 200 mm in width, and 0.5 mm in thickness was used as copper plate 1. Oxygen-free copper of 300 mm in length, 200 mm in width, and 0.8 mm in thickness was used as copper plate 2. 30 μm of brazing paste was applied to both surfaces of the ceramic substrate, and copper plates were placed on each surface.

[0135] Next, a heating bonding process was performed. The bonding atmosphere was unified into a nitrogen atmosphere. In addition, the bonding conditions were set within the range of 850°C to 920°C × 30 minutes to 50 minutes, using a continuous furnace. The heating rate and cooling rate were set to the conditions shown in Table 4. The difference between the heating rate and the cooling rate was calculated by |heating rate - cooling rate|.

[0136] Table 4

[0137]

[0138] The joint body was manufactured through the above process. The joint layer section of the obtained joint body was observed to investigate the presence of Cu-Sn alloys with different composition ratios, the average particle size of titanium nitride in the Ti reaction layer, and the amount of Cu and Sn in the Ti reaction layer. In addition, when there were Cu-Sn alloys with different composition ratios, the presence of the first Cu-Sn alloy with a Sn amount of 3 atomic % or more and 9 atomic % or less and the presence of the second Cu-Sn alloy with a Sn amount of 10 atomic % or more and 30 atomic % or less were investigated.

[0139] In addition, the area ratio of Cu-Sn alloy and the area ratio of Ti-Sn alloy were also investigated. In addition, the major diameter and area ratio of titanium silicide were also investigated. In the calculation of the area ratio, the EDX element distribution was used with imaging software. Figure 2 The detailed method is as described above. The results are shown in Tables 5 to 8.

[0140] Table 5

[0141]

[0142] Table 6

[0143]

[0144] Table 7

[0145]

[0146] Table 8

[0147]

[0148] As can be seen from Tables 5 to 8, the joint bodies according to the examples satisfy the preferred conditions. The Cu-Sn alloy forms the main component of the joint layer. In addition, it is observed that the first Cu-Sn alloy and the second Cu-Sn alloy are intermetallic compounds.

[0149] EDX point analysis was performed at 10 arbitrary locations in the Cu-Sn alloy region. Also, EDX point analysis was performed at 10 arbitrary locations in the Ti-Sn alloy region. Figure 3 The results are shown in the ternary phase diagram of , and they are all within the range of the first composition region. In addition, in the EDX point analysis results of the Cu-Sn alloy, the composition of the Cu-Sn alloy of more than 80% and less than 100% is within the range of the second composition region. In the EDX point analysis results of the Ti-Sn alloy of more than 80% and less than 100%, the composition of the Ti-Sn alloy of more than 80% and less than 100% is within the range of the third composition region.

[0150] The "others" in the area ratio of Table 8 are regions where a Ti reaction layer, carbide, or metal component is separately melted as a single substance.

[0151] The thickness of the Ti reaction layer in the joint body according to the embodiment is 1 μm or less. In addition, TiN and Ti 2 N both sides.

[0152] In contrast, no Cu-Sn alloy with a different composition ratio was observed in Comparative Example 1. A small amount of Cu-Sn alloy was observed in Comparative Example 1, but no Cu-Sn alloy with a different composition was observed. In Comparative Example 1, it is believed that the AgCu eutectic or AgSn became the main component of the bonding layer.

[0153] Next, 100 bonded bodies according to each of the examples and comparative examples were manufactured, and the bonding strength and the amount of warpage were measured.

[0154] The bonding strength is measured by a peel test. Specifically, in each embodiment and comparative example, a sample for the peel test is prepared. As a sample, a long copper plate is bonded to a ceramic substrate. At this time, the bonding is performed in a manner that one end of the copper plate protrudes from the ceramic substrate. The peel strength is measured by vertically stretching the protruding copper plate.

[0155] In addition, the warpage is the warpage of the long side of the joint body. The example with more than 90 joint bodies and a warpage of less than 0.3 mm is regarded as "best". The example with a warpage of less than 0.3 mm among 70 to 89 joint bodies is regarded as "good". The example with more than 31 joint bodies with a warpage exceeding 0.3 mm is regarded as "bad". Table 9 shows the results.

[0156] Table 9

[0157]

[0158] As can be seen from Table 9, the bonding strength of the joint body involved in the embodiment is equivalent to that of the joint body involved in Comparative Example 1 using Ag-containing solder. Even if the size of the joint body is increased to more than 200 mm, the warping amount can be suppressed. Even if the heating rate and the cooling rate are increased, the warping amount can be suppressed. Therefore, it is known that it is a joint body with good mass production. Correspondingly, in Comparative Example 1, multiple joint bodies with warping amounts exceeding 0.3 mm are produced. This is because no Cu-Sn alloy with different composition ratios is formed.

[0159] The amount of Sn in the copper plate of the joint body involved in the embodiment and the comparative example was investigated. The amount of Sn was measured by analyzing the cross section of the joint body with EDX. The boundary between the joint layer and the copper plate was defined as described above. The amount of Sn in the boundary between the joint layer and the copper plate was set as Sn-1, the amount of Sn at a distance of 10 μm from the boundary was set as Sn-2, and the amount of Sn at a distance of 20 μm from the boundary was set as Sn-3. Sn-2 / Sn-1 and Sn-3 / Sn-1 were calculated.

[0160] It was confirmed whether the point farthest from the boundary in the region where Sn was detected was within 50 μm. The case within 50 μm was marked as "0", and the case exceeding 50 μm was marked as "×".

[0161] The presence or absence of voids in the bonding layer was investigated. The presence or absence of voids was investigated by ultrasonic flaw detection (SAT). The volume ratio of voids was 0% or more and 0.3% or less as "◎", 0.4% or more and 1% or less as "〇", 2% or more and 3% or less as "△", and 4% or more as "×". The results are shown in Table 10.

[0162] Table 10

[0163]

[0164] From Tables 9 and 10, it is known that the diffusion of Sn into the copper plate is suppressed in the joint bodies involved in the examples. In addition, it is known that the void area is also small. In particular, in Examples 6 to 8 in which the void area is reduced, the joint strength is as high as 25 kN / mm or more.

[0165] In contrast, in the comparative example, the diffusion amount of Sn into the copper plate is large, and the void area ratio is also large.

[0166] Several embodiments of the present invention have been described above, but these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other ways, and various omissions, substitutions, and changes can be made without departing from the scope of the main purpose of the invention. These embodiments and their modifications are included in the scope and main purpose of the invention, and are also included in the invention described in the claims and their equivalents. In addition, the above-mentioned embodiments can be implemented in a combination with each other.

[0167] Explanation of symbols

[0168] 1-joint

[0169] 2-Ceramic substrate

[0170] 3-Copper Coin

[0171] 4 - Bonding layer

[0172] 5-Ti reaction layer

[0173] 6-Cu-Sn alloy or Cu-In alloy

[0174] 7-Ti-Sn alloy or Ti-In alloy

[0175] 8-Titanium silicide particles

[0176] 10-Ceramic circuit substrate

[0177] 11-Circuit Department

[0178] 12-Heat sink

Claims

1. A conjugate, It is characterized in that It has: Ceramic substrate, Copper plate, and a bonding layer disposed on at least one side of the ceramic substrate and bonding the ceramic substrate and the copper plate; The bonding layer contains: A Ti reaction layer containing titanium nitride or titanium oxide as a main component, and a first alloy containing a Cu-Sn alloy, a second alloy containing a Ti-Sn alloy, and titanium silicide located between the Ti reaction layer and the copper plate, And the bonding layer does not contain Ag; In the average values ​​obtained by performing EDX regional analysis on three regions of the bonding layer with a thickness of 40 μm and a width direction of 40 μm, when the total area of ​​the three regions is set as 100%, The area ratio of the first alloy is 50% or more and 95% or less, The area ratio of the second alloy is 5% or more and 30% or less, The area ratio of the titanium silicide is greater than or equal to 0.5% and less than or equal to 15%, The plurality of first alloys and the second alloy of the bonding layer detected by EDX point analysis include alloys that are within a region surrounded by (97, 3, 0), (60, 40, 0), (2, 40, 58), and (39, 3, 58) in the ternary phase diagram of Cu, Sn, and Ti.

2. The conjugate according to claim 1, It is characterized in that The area ratio of the first alloy is 90% or less.

3. The conjugate according to claim 2, It is characterized in that In an average value obtained by performing EDX regional analysis on the three regions, an area ratio of the region other than the first alloy, the second alloy, and the titanium silicide is 10% or less.

4. The conjugate according to any one of claims 1 to 3, It is characterized in that The plurality of first alloys detected by EDX point analysis of the bonding layer include alloys within a region surrounded by (97, 3, 0), (77, 23, 0), (52, 23, 25), and (72, 3, 25) in a ternary phase diagram of Cu, Sn, and Ti.

5. The conjugate according to any one of claims 1 to 3, It is characterized in that The second alloy of the bonding layer detected by EDX point analysis includes an alloy located in a region surrounded by (41, 23, 36), (24, 40, 36), (2, 40, 58), and (19, 23, 58) in the ternary phase diagram of Cu, Sn, and Ti.

6. The conjugate according to any one of claims 1 to 3, It is characterized in that The Ti reaction layer contains titanium nitride particles with an average particle size of 50 nm or less. Cu and Sn exist in the grain boundaries of the titanium nitride particles.

7. The conjugate according to claim 5, It is characterized in that The Ti reaction layer contains titanium nitride particles with an average particle size of 50 nm or less. Cu and Sn exist in the grain boundaries of the titanium nitride particles.

8. The conjugate according to any one of claims 1 to 3, It is characterized in that The bonding layer contains titanium silicide particles having a major diameter of 0.5 μm or more and 6 μm or less.

9. The conjugate according to claim 7, It is characterized in that The bonding layer contains titanium silicide particles having a major diameter of 0.5 μm or more and 6 μm or less.

10. The conjugate according to any one of claims 1 to 3, It is characterized in that The joining layer further contains carbon.

11. The conjugate according to claim 9, It is characterized in that The joining layer further contains carbon.

12. The conjugate according to any one of claims 1 to 3, It is characterized in that The thickness of the Ti reaction layer is less than 1 μm.

13. The conjugate according to claim 11, It is characterized in that The thickness of the Ti reaction layer is less than 1 μm.

14. The conjugate according to any one of claims 1 to 3, It is characterized in that Sn exists in the boundary between the bonding layer and the copper plate, A ratio C2 / C1 of a second concentration C2 of Sn in a region of the copper plate 10 μm away from the boundary to a first concentration C1 of Sn in the boundary is 0.60 or less.

15. The conjugate according to claim 13, It is characterized in that Sn exists in the boundary between the bonding layer and the copper plate, A ratio C2 / C1 of a second concentration C2 of Sn in a region of the copper plate 10 μm away from the boundary to a first concentration C1 of Sn in the boundary is 0.60 or less.

16. A ceramic circuit substrate, It is characterized in that The device comprises the joined body according to any one of claims 1 to 15.

17. A semiconductor device, It is characterized in that It has: The ceramic circuit substrate according to claim 16, and A semiconductor element is mounted on the ceramic circuit substrate.

Images