DE102017128308B4 - Process for the production of a metal-ceramic substrate - Google Patents

Abstract

A method for producing a metal-ceramic substrate (1), in particular a copper-ceramic substrate, comprising: providing a first metal layer (11), in particular a first copper layer, and a second metal layer (12), in particular a second copper layer, wherein the first metal layer (11) and / or the second metal layer (12) is admixed with a means for influencing the grain size, and - connecting the first metal layer (11) and / or the second metal layer (12) with a ceramic layer to form the metal -Ceramic substrate (1) in an AMB connection method, the first metal layer (11) and the second metal layer (12) for connection to the ceramic layer each being provided as individual sheets or as a composite in which the first metal layer (11 ) is joined to the second metal layer (12), a portion of the agent for influencing the grain size in the first metal layer (11) being different from an ant Part of the means for influencing the grain size in the second metal layer (12) differs in such a way that in the finished copper-ceramic substrate (1) the first metal layer (11) has a first grain size and the second metal layer (12) has a different size from the first grain size have a differing second grain size, the first metal layer (11) facing away from the ceramic layer (10) and the second metal layer (12) facing the ceramic layer (10) in the finished state, the proportion of the agent for influencing the grain size in the first metal layer (11) is greater than the proportion of the agent for influencing the grain size in the second metal layer (12) and the second metal layer (12) is thicker than the first metal layer (11).

Description

The present invention relates to a method for producing a metal-ceramic substrate.

Metal-ceramic substrates are well known as supports for electrical or electronic components. Such metal-ceramic substrates typically comprise a ceramic layer with a metallization, the ceramic layer being structured for insulation and the metallization for forming conductor tracks and connection points for the electrical or electronic components. In addition, from the DE 10 2015 224 464 A1 a copper-ceramic substrate is known in which the copper layer, ie the metallization, has a first layer with an averaged first grain size and a second layer with an averaged second grain size. The averaged first grain size differs from the averaged second grain size. This configuration enables a metallization in which the copper layer facing the ceramic layer is coarser than that of the copper layer facing away from the ceramic layer.

This is advantageous in that the larger grain size on the ceramic layer means that a lower proof stress and thus an improved resistance to temperature changes can be set for the copper-ceramic substrate. The result is a reduced likelihood of delamination between the metallization made of copper and the ceramic layer when exposed to heat, in particular on the interface between copper and ceramic. At the same time, the fine-grained outside of the metallization is advantageous for further processing with electrical or electro-optical systems, the connection of electrical or electronic components to the copper layer and the overall visual impression.

For the formation of the different averaged grain sizes in the DE 10 2015 224 464 A1 proposed a method in which a temperature treatment and / or different copper materials are provided. The different copper materials are specified as OFE-Cu, as Cu-ETP or as OF-Cu, i.e. as copper materials, which can be produced by electrolytic refining and differ depending on the quality of the electrolytic refining, for example also using vacuum deoxidation are low in oxygen.

From the US 2014 0 284 088 A1 A circuit board is also known in which the metallization on the surface consists of a first part and a second part arranged above the first part. It is provided that the first part differs from the second part in terms of grain size.

From the US 2014 0 126 155 A1 is known a solder material for bonding a metal layer to a ceramic carrier. Here the grain size in the solder material is specifically set. In the EP 1 498 946 B1 a metal composite is connected to the ceramic layer by means of a solder material.

Based on this background, the present invention sees it as an object to further improve the metal-ceramic substrates known from the prior art and the process for their production, in particular with regard to a first or second grain size in the metallization. The method should also be designed to be simpler and less expensive.

This object is achieved by a method for producing a metal-ceramic substrate according to claim 1. Further advantages and features of the invention result from the subclaims and the description and the accompanying figures.

• - Bereitstellen einer ersten Metallschicht, insbesondere einer ersten Kupferschicht, und einer zweiten Metallschicht, insbesondere einer zweiten Kuperschicht, wobei der ersten Metallschicht und/oder der zweiten Metallschicht ein Mittel zur Beeinflussung der Korngröße beigemischt ist, und
• - Verbinden der ersten und/oder der zweiten Metallschicht mit einer Keramikschicht zur Bildung des Metall-Keramik-Substrats,

wobei sich ein Anteil des Mittels zur Beeinflussung der Korngröße in der ersten Metallschicht von einem Anteil des Mittels zur Beeinflussung der Korngröße in der zweiten Metallschicht derart unterscheidet, dass im gefertigten Kupfer-Keramik-Substrat die erste Metallschicht eine erste Korngröße und die zweite Metallschicht eine sich von der ersten Korngröße unterscheidende zweite Korngröße aufweisen.According to the invention, a method for producing a metal-ceramic substrate, in particular a copper-ceramic substrate, is provided, comprising:

• - Providing a first metal layer, in particular a first copper layer, and a second metal layer, in particular a second copper layer, with a means for influencing the grain size being added to the first metal layer and / or the second metal layer, and
• Connecting the first and / or the second metal layer to a ceramic layer to form the metal-ceramic substrate,

wherein a proportion of the agent for influencing the grain size in the first metal layer differs from a proportion of the agent for influencing the grain size in the second metal layer in such a way that in the copper-ceramic substrate produced the first metal layer has a first grain size and the second metal layer one have a second grain size different from the first grain size.

In contrast to the prior art, it is provided according to the invention that the grain size is influenced or specifically adjusted by adding proportions of an agent for influencing the grain size in the first metal layer or the second metal layer. As a result, the desired and optimized grain size in the first or second can be optimized for the application Adjust the metal layer as controlled as possible. For this purpose, the agent for influencing the grain size is preferably added to the first metal layer or the second metal layer in the molten state, ie in the form of a first metal melt and a second metal melt. In particular, it is provided that the means for influencing is selected such that the addition of the agent on the one hand influences the grain growth and on the other hand other properties of the first metal layer or the second metal layer, such as, for. B. a degree of purity or an electrical conductivity, are influenced only insignificantly, preferably so that there are no differences within a tolerance range (for example, determined by usual material fluctuations or measurement errors).

The means for influencing the grain size is preferably designed in such a way that, with increasing amount of the agent in the first or second metal layer, the average grain size in the first metal layer and the second metal layer in the manufactured metal-ceramic substrate decreases. However, it is also alternatively or additionally provided that the proportion of the first molten metal differs from the proportion of the second molten metal by the type or composition of the agent for influencing, e.g. B. in terms of their material composition or its material. Furthermore, it is an alternative embodiment of the invention that the first metal layer and the second metal layer made of the same metal material, e.g. B. OFE-Cu, exist and differ only with regard to the proportions of the agent for influencing the grain size. This can, for. B. equivalent copper can advantageously be used to form the first and second metal layers. Alternatively, the first metal layer and the second metal layer additionally differ in terms of the metal material, e.g. B. is the first metal layer made of OFE-Cu and the second metal layer made of OF-Cu.

It is also conceivable that only the first metal layer or only the second metal layer is added with the agent for influencing the grain size and the proportion of the agent for influencing the grain size disappears in the first or the second metal layer. The amount of agents for influencing the grain size in the first metal layer is preferably 8 to 15,000 times, preferably 15 to 2000 times and particularly preferably 30-100 times as large as the amount of agents for influencing the grain size in the second metal layer. Furthermore, it is preferably provided that the amount of agents for influencing the grain size in the second metal layer or the second metal layer is less than 20 ppm (by weight), preferably less than 15 ppm (weight) and particularly preferably less than 10 ppm ( Wt.). The first grain size or the second grain size is to be understood in particular as an average grain size in the metal of the first or second metal layer. It is advantageously provided that both the first metal layer and the second metal layer are mixed with a proportion of means for influencing the grain size, since this also results in the coarser-grained layer, ie. h, especially in the second metal layer, can advantageously avoid excessive grain growth.

According to the invention, it is provided that the first metal layer and the second metal layer for connection to the ceramic layer are each provided as individual sheets or as a composite in which the first metal layer is already joined to the second metal layer. For example, the first metal layer and the second metal layer are each rolled individually and / or together. It is preferably provided here that the composite of first and second metal layers has a thickness measured in the layer direction, which is less than 5 mm, preferably less than 3 mm and particularly preferably less than 1 mm. Furthermore, when the first metal layer and the second metal layer are joined together, a desired ratio between a first layer thickness of the first metal layer measured in the layer direction and a second layer thickness of the second metal layer measured in the layer direction can be set. The first metal layer is preferably thicker than the second metal layer, preferably 1 to 15 times as thick, preferably 2 to 10 times as thick and particularly preferably 3 to 8 times as thick, or the second metal layer is thicker than the first metal layer , preferably 1 to 15 times as thick, preferably 2 to 10 times as thick and particularly preferably 3 to 8 times as thick.

It is particularly preferably provided that the desired grain size is only realized by connecting the first metal layer or the second metal layer to the ceramic layer. In other words, the connection method used to produce the metal-ceramic substrate initiates and / or finalizes the grain growth, so that the first grain size or the second grain size of the manufactured metal-ceramic substrate assumes the desired size. It is also conceivable that grain growth is already initiated or prepared with the method for joining the first metal layer and the second metal layer into a composite. It is apparent to the person skilled in the art that in addition to the first metal layer and the second metal layer, further metal layers can be provided in the metallization. For example, a sandwich structure is conceivable in which a third metal layer is arranged between the first and second metal layers and has a grain size that lies between that of the first and second layers.

It is particularly preferably provided that the metal layer structure is made of copper, i. H. the first metal layer is a first copper layer and the second metal layer is a second copper layer. It is also conceivable that the metal composition in the first metal layer differs from the metal composition in the second metal layer. Furthermore, it is conceivable that the first metal layer and the second metal layer are made of a copper alloy or aluminum. Finally, the metal layers are advantageously in the annealed condition prior to joining, i. H. they are not or only partially recrystallized after the rolling process.

The ceramic from which the ceramic layer is formed can be, for example, Al ₂ O ₃ , Si ₃ N ₄ , AIN or an HPSX ceramic (ie a ceramic with an Al ₂ O ₃ matrix which has an x percentage of ZrO ₂ comprises, for example Al ₂ O ₃ with 9% ZrO ₂ = HPS9 or Al ₂ O ₃ with 25% ZrO ₂ = HPS25). The ceramic layer preferably has a comparatively high electrical insulation strength, preferably of more than 5 kV / mm, particularly preferably of more than 10, 20 or even more than 30 kV / mm and / or a high thermal conductivity, preferably of more than 10 W. / mK, particularly preferably of more than 20 or even more than 60 W / mK, such as, for example, technical ceramics or organic insulation materials filled with heat-conducting materials.

mit einer Streckgrenze R_e, einer Startspannung σ₀, einem Korngrenzwiderstand K und einer Korngröße d_k), die eine in der Metallschicht vorliegende innere Spannung in Verbindung setzt mit einer Korngröße des Gefüges, ein vergleichsweises grobes Gefüge in der zweiten Metallschicht realisieren, wodurch mit Vorteil ein Spannungsniveau in einem Anbindungsbereich, in dem die zweite Metallschicht an das Trägerelement angebunden ist, reduziert wird.According to a further embodiment of the present invention, it is provided that in the finished state the first metal layer faces away from the ceramic layer and the second metal layer faces towards the ceramic layer, the proportion of the agent for influencing the grain size in the first metal layer being greater than the proportion of the By means of influencing the second metal layer. This advantageously enables a metal-ceramic substrate to be realized in which the first metal layer facing away from the ceramic layer is finer-grained than the second metal layer facing the ceramic layer. Correspondingly, the design of the first metal layer with a smaller average first grain size in comparison to the second metal layer prevents a structure that is too coarse in the first metal layer from making it difficult to connect electronic components, in particular in thin wire bonding. In addition, a comparatively fine structure simplifies automatic optical inspection (AOI), ie testing the metal-ceramic substrates after they have been manufactured. At the same time, a grain size gradient is formed along a layer direction, along which the first metal layer and the second metal layer are layered on the ceramic layer, when the metal layer structure is connected. In the sense of the Hall-Petch relationship ( $R_e={\sigma }_0+\frac{K}{\sqrt{d_k}};$

with a yield strength R _e , a starting stress σ ₀ , a grain limit resistance K and a grain size d _k ), which combines an internal stress present in the metal layer with a grain size of the structure, to realize a comparatively coarse structure in the second metal layer, thus with The advantage of a voltage level in a connection area in which the second metal layer is connected to the carrier element is reduced.

In particular, the person skilled in the art understands by a fine structure a metal layer with a comparatively small average grain size, in particular with an expansion or dimension of less than 100 μm, and by a coarse structure a metal layer with a comparatively large average grain size, preferably with an expansion or Dimension of more than 100 µm. Furthermore, it is preferably provided that in the first metal layer the first average grain size and in the second metal layer the second average grain size, in particular in a direction running parallel and / or perpendicular to the layer direction, is essentially constant. This can advantageously be achieved by a uniform distribution of the means for influencing the grain size. However, it is also conceivable that by means of a targeted local distribution of the agent for influencing the grain size, a desired grain size distribution can be achieved within the first metal layer or the second metal layer.

• - für Titan kleiner als 2,1 Gew.-%
• - für Zirkonium kleiner als 0,17 Gew.-%
• - für Hafnium kleiner als 1,1 Gew.-%
• - für Chrom kleiner als 0,73 Gew.-% und
• - für Niobium kleiner als 0,15 Gew.-%.

alle jedoch einzeln oder zusammen kleiner als 0,01 Gew.-%. Beispielsweise beträgt bei einem Cu-Gehalt von mehr als 99,99 Gew.-% in einem ersten Bespiel ein Cr-Anteil weniger als 0,01 Gew.-% und in einem zweiten Beispiel ein Nb-Anteil weniger als 0,01 Gew-%. Dabei handelt es sich bei dem Kupfer im zweiten Beispiel vorzugsweise um ETP-KupferIt is preferably provided that the means for influencing the grain size comprises a grain refining agent, in particular a grain refining agent containing titanium, zirconium, hafnium, chromium and / or niobium. The grain refining agent is preferably an alloy metal with a comparatively high oxygen affinity and the lowest possible marginal solubility. As a result, the required amount of grain refining agent can be kept as small as possible, so that the metal of the first metal layer and the second metal layer is not contaminated in such a way that the first metal layer and the second metal layer are not suitable as metallization in power electronics. In particular, it has been found that in the case of the grain refiners B, Ca, Fe, Cr, and zirconium arsenide, at least a few parts per thousand must be added in order to decisively influence the grain growth, so that after their addition in the case of copper as metal for the first and the second metal layer no more Cu-OFE or Cu-PHC is present. in the In contrast, it has been found that alloy metals such as titanium, zirconium, hafnium, chromium and / or niobium are suitable in small amounts for the formation of the desired grain sizes. The percentage by weight of the grain refining agent in the first or second metal layer is preferably less than 2.5% by weight, preferably less than 1.5% by weight and particularly preferably less than 1% by weight or even less than 0 , 5% by weight. For example, the proportion of the grain refining agent in the first metal layer or the second metal layer, in particular for a first metal layer and a second metal layer, is made of Cu-OFE

• - for titanium less than 2.1% by weight
• - for zirconium less than 0.17% by weight
• - for hafnium less than 1.1% by weight
• - for chromium less than 0.73 wt .-% and
• - for niobium less than 0.15 wt .-%.

however, all individually or together less than 0.01% by weight. For example, with a Cu content of more than 99.99% by weight in a first example, a Cr content is less than 0.01% by weight and in a second example an Nb content is less than 0.01% by weight. %. The copper in the second example is preferably ETP copper

A “DCB method” (direct copper bond technology) is understood by the person skilled in the art to mean such a method, for example for connecting metal layers or sheets (for example copper sheets or foils) to one another and / or to ceramic or ceramic layers serves, namely using metal or copper sheets or metal or copper foils, which have a layer or a coating (melting layer) of a chemical compound of the metal and a reactive gas, preferably oxygen, on its surface sides. In this example in the U.S. Patent 37 44 120 or in the DE-PS 23 19 854 The described method forms this layer or this coating (melting layer) a eutectic with a melting temperature below the melting temperature of the metal (e.g. copper), so that it can be connected to one another by placing the film on the ceramic and by heating all layers, and by melting the metal or copper essentially only in the area of the melting layer or oxide layer.

• - Oxidieren einer Kupferfolie derart, dass sich eine gleichmäßige Kupferoxidschicht ergibt;
• - Auflegen des Kupferfolie auf die Keramikschicht;
• - Erhitzen des Verbundes auf eine Prozesstemperatur zwischen etwa 1025 bis 1083°C, z. B. auf ca. 1071 °C;
• - Abkühlen auf Raumtemperatur.

In particular, the DCB method then z. B. the following process steps:

• - Oxidizing a copper foil so that there is a uniform copper oxide layer;
• - placing the copper foil on the ceramic layer;
• - Heating the composite to a process temperature between about 1025 to 1083 ° C, z. B. to about 1071 ° C;
• - cooling to room temperature.

If a DCB process is used, the metal can preferably take up a comparatively high proportion of oxygen in the first but also the second metal layer, in particular the first copper layer and the second copper layer, whereupon inert oxides melting at grain boundaries form, which impede grain growth and so on Formation of a smaller grain (ie a metal with a smaller average grain size) favors.

In a further embodiment of the present invention, it is provided that the means for influencing the grain size comprises crystallization nuclei. In particular, the crystallization nuclei are disperse, preferably finely disperse, particles which are added to the metal of the first metal layer or the second metal layer. The particles are preferably selected such that the particles have a higher melting point than the metal, in particular than the copper, of the first or the second metal layer and cannot be dissolved in the metal, in particular in the copper. It is also advantageous if a particle surface is wetted with the metal, in particular the copper, in the course of the illustration, be it in a molten state or by diffusion. Particularly advantageous for this have been particles made of a ceramic, such as. B. Al ₂ O ₃ , Si ₃ N ₄ or AIN. It is preferably provided that the nuclei are evenly distributed in the first metal layer and the second metal layer in order to avoid agglomeration of the nuclei.

It is expedient for the crystallization nuclei to be added to the first metal layer and / or the second metal layer together with an active metal. For example, the active metal is Ti, Cr, Zr, Hf, Nb, V, Ce, La or a comparable active metal. It is preferably provided that the active metals react with the dispersive particles so that the proportion of active metals in the finished metal-ceramic substrate is reduced.

According to the invention, it is provided that the first metal layer, the second metal layer and the ceramic layer are connected to one another in an AMB connection method, in particular in a common AMB connection method. The use of crystallization nuclei has the advantage that one does not necessarily have to rely on the oxygen of the DCB process to control the grain size growth. Under a Active soldering process, ie "active metal brazing (AMB)" process, e.g. B. for connecting metal layers or metal foils, in particular also copper layers or copper foils with ceramic material, is to be understood as a method which is also used specifically for the production of metal-ceramic substrates. A connection between a metal foil, for example copper foil, and a ceramic substrate, for example an aluminum nitride ceramic, is produced at a temperature between approx. 650-1000 ° C. using a hard solder, which in addition to a main component such as copper, silver and / or gold also contains an active metal. This active metal, which is, for example, at least one element from the group Hf, Ti, Zr, Nb, Ce, creates a connection between the solder and the ceramic by chemical reaction, while the connection between the solder and the metal is a metallic brazing connection .

It is preferably provided that the first metal layer and the second metal layer are joined together in a DCB process to form a composite and the composite is bonded to the ceramic layer. The connection of the first metal layer to the second metal layer can preferably take place independently of the connection to the ceramic layer. According to the invention, the composite is connected to the ceramic layer using an AMB method. The method for joining the first metal layer and the second metal layer is particularly preferably coordinated with that for connecting the composite to the ceramic layer, in order in this way to set the desired first grain size and second grain size in the finished metal-ceramic substrate.

It is preferably provided that the attached metal layer structure is at least partially etched to form a metal surface structure. As a result, a metal surface structuring, such as eg. B. conductor tracks or pads, on a component side of the metal-ceramic substrate, which allows the use of the metal-ceramic substrate as a circuit board or printed circuit board.

It is expediently provided that the first average grain size is smaller than the second average grain size. The second average grain size is preferably between 160-2,000 μm, preferably between 200-1,000 μm and particularly preferably between 250-500 μm. Furthermore, it is preferably provided that the first average grain size is between 50 and 200 μm, preferably between 70 and 150 μm and particularly preferably between 80 and 120 μm. The average grain size is determined as an average over the grains distributed along a plane perpendicular to the layer direction.

In a further embodiment of the present invention it is provided that the second metal layer is thicker than the first metal layer, preferably 1.1 to 15 times as thick, preferably 2 to 10 times as thick and particularly preferably 3 to 8 times as thick thick. This provides the second metal layer with its larger first average grain size with a corresponding space, while the first metal layer takes into account that the finer grain in the first metal layer requires less space.

It is preferably provided that a first average grain size essentially corresponds to a first layer thickness of the first metal layer. This advantageously achieves a maximum possible first average grain size. It is obvious to the person skilled in the art that “essentially” includes in particular those first average grain sizes that, for example, within a tolerance of - ± 20% or preferably - ± 10%, particularly preferably - ± 5% with the thickness of the layer in the direction of the layer primary metal layer match.

• 1a und 1b: ein Verfahren zur Herstellung eines Metall-Keramik-Substrats gemäß einer ersten beispielhaften Ausführungsform der vorliegende Erfindung und
• 2a und 2b: ein Verfahren zur Herstellung eines Metall-Keramik-Substrats gemäß einer ersten beispielhaften Ausführungsform der vorliegende Erfindung und 3a und 3b ein Trägersubstrate gemäß einer zweiten und dritten bevorzugten Ausführungsform der vorliegende. Erfindung in einer Schnittansicht

Further advantages and features result from the following description of preferred embodiments of the object according to the invention with reference to the attached figures. It shows:

• 1a and 1b : A method of manufacturing a metal-ceramic substrate according to a first exemplary embodiment of the present invention and
• 2a and 2 B : A method of manufacturing a metal-ceramic substrate according to a first exemplary embodiment of the present invention and 3a and 3b a carrier substrate according to a second and third preferred embodiment of the present. Invention in a sectional view

In the 1a and 1b is a schematic of a method of manufacturing a metal-ceramic substrate 1 according to a first exemplary embodiment of the present invention. Such metal-ceramic substrates 1 serve as a carrier for electrical or electronic components. For the electrically conductive connection of the individual electrical or electronic components, there are in particular metallizations for forming conductor tracks on a ceramic surface or top 15 the ceramic layer 10th intended. In particular, the metallization is structured in order to provide conductor tracks or fastening points for the electronic or electrical components. It has proven advantageous here if the metallization applied to the ceramic has a first metal layer 11 and a second metal layer 12th having a first grain size in the first metal view 11 , ie a first average grain size of the metal in the first metal layer 11 , of a second grain size, ie a second average grain size of the metal in the second metal layer 12th , differs. In particular, it is provided that the first metal layer 11 and the second metal layer 12th to form the metallization on the ceramic layer, a composite, in particular a sandwich structure or a semi-finished metal product 2nd , form, ie in the finished state are the first metal layer 11 and the second metal layer 12th in a perpendicular to the ceramic surface 15 extending stack direction S stacked or stacked. In the 1b is a metal-ceramic substrate 1 shown in which such a composite of the first metal layer both on the underside and on the top 11 and the second metal layer 12th is arranged.

Here is the second metal layer 12th the ceramic layer 10th facing and the first metal layer 11 the ceramic layer 1 facing away, in particular an average first grain size in the first metal layer 11 is smaller than the average second grain size in the second metal layer 12th . By the opposite of the second metal layer 12th finer grain design of the first metal layer 11 it is advantageously easier to implement an electronic or electrical component to the metallization or the composite in comparison to a coarse-grained metallization. In addition, a visually detectable, more homogeneous or finer cover layer can be used for the metallization on the ceramic layer 10th provide. At the same time, it is advantageously possible due to the opposite of the first metal layer 11 coarser second metal layer 12th to design the connection of the metallization and the ceramic layer in such a way that, despite the different coefficients of thermal expansion, it is optimized with regard to its resistance to temperature changes. That is, the second grain size is preferably selected such that the mechanical stress, which is caused by different thermal expansion coefficients, is kept as low as possible. This advantageously increases the service life of the metal-ceramic substrate 1 .

For targeted control of the first grain size or the second grain size, in particular a ratio between the first grain size and the second grain size, it is particularly provided that the first metal layer 11 and / or the second metal layer 12th a means of influencing the grain size is added or admixed in the manufacturing process. In particular, the means for influencing the grain size are each introduced into a metal melt, in particular a copper melt. Metal melts with different proportions of agents for influencing the grain size preferably each become a first metal layer 11 , in particular in the form of a first metal sheet, and a second metal layer 12th , in particular in the form of a second metal sheet, which in turn forms a semifinished metal product or composite 2nd be joined together.
Due to the different proportions of means for influencing the grain sizes, the first grain size or the second grain size can be controlled in a controlled manner, in particular depending on the application. It is also conceivable that the second grain size is approximately a second layer thickness D2 the second metal layer 12th corresponds. It is also provided that the second layer thickness D2 is greater than a first layer thickness D1 the first metal layer 11 . It is also conceivable, for example, that the first grain size assumes a value of less than 200 μm, preferably 100 μm, and / or the second grain size assumes a value of more than 160 μm, preferably between 200 and 1,000 μm. In the in the 1a and 1b In the illustrated embodiment, the means for influencing the grain size is a grain refining agent, the proportion of the grain refining agent in the first metal layer 11 the proportion of the grain refining agent in the second metal layer 12th differs. A proportion by weight is in the first metal layer 11 or the second metal layer 12th to understand. In particular, it is provided that the first metal layer 11 has a larger proportion of grain refining agent than the second metal layer 12th . It has proven to be particularly advantageous if the grain refining agent is an alloy metal with a comparatively low edge solubility in copper and a comparatively high oxygen affinity, such as, for. B. titanium, zirconium, hafnium, chromium and / or niobium.

The grain refining agent is preferably added or admixed to a first metal melt or second metal melt, the first metal layer being formed from the first metal melt 11 or the second metal layer is formed from the second metal melt 12th becomes. Subsequently, titanium, zirconium, hafnium, chromium and / or niobium is deposited in the annealing process at the respective grain boundaries in the first metal layer 11 and the second metal layer 12th attached, for example by means of an annealing process after sheet metal forming. Preferably the first metal layer 11 and the second metal layer 12th as it is in 1a is shown, assembled by means of a DCB method. In the DCB process, a comparatively high proportion of oxygen is generated from the first metal layer due to the process 11 and / or second metal layer 12th recorded, whereupon preferably high-melting inert oxides form at the grain boundaries, which hinder grain growth, ie develop a grain-refining effect. In this way, the grain size in the first metal layer can advantageously be achieved by means of the grain refining agents 11 or the second metal layer 12th influence. Furthermore, it is conceivable that the grain refining agent is in the first metal layer 11 of that in the second metal layer 12th differs in terms of its material composition.

In the 2a and 2 B is a schematic of a method of manufacturing a metal-ceramic substrate 1 according to a second exemplary embodiment of the present invention. In particular, the second embodiment differs from the first embodiment in that, in the second embodiment, crystallization nuclei act as a means for influencing the grain size 5 be used. In particular, the crystallization nuclei comprise disperse particles, those of the metal melt of the first metal layer 11 or the metal melt of the second metal layer 12th in different amounts, so that their proportions in the first metal layer 11 or the second metal layer 12th to influence. In particular, these particles are selected such that they have a higher melting point than the metal of the first metal layer 11 or the second metal layer 12th and not in the metal of the first metal layer 11 or the second metal layer 12th to solve. At the same time, the surface of the particles should be wetted with copper in the course of the illustration, be it in the course of a melting process or by diffusion. For this purpose, as particles, ie as seed particles, in particular ceramic particles, such as. B. Al ₂ O ₃ , Si ₃ N ₄ or AIN particles, proven in the metal of the first metal layer 11 or the second metal layer 12th be introduced, preferably with an active metal such. B. Ti, Cr, Zr, Hf, Nb, V, Ce, La or a comparable active metal. In particular, it is provided that the disperse particles react with the active metal, so that the proportion of the active metal is reduced in the finished metal-copper substrate.

It is preferably provided that the crystallization nuclei 5 in the first metal layer 11 and / or the second metal layer 12th or the first molten metal and / or second molten metal are evenly distributed. A uniform distribution is understood to mean, in particular, a distribution in which the nuclei do not occur locally, ie agglomerate, ie the distance between two nuclei in the first or second molten metal is essentially constant. It is further provided that the amount of crystallization nuclei which are added to the first and / or the second metal melt is less than 20 ppm, preferably less than 15 ppm and particularly preferably less than 10 ppm. It has surprisingly been found that the grain size of the metal in the first metal layer is already apparent with such small amounts of crystallization nuclei 11 or the second metal layer 12th can be limited to a maximum of 200 µm. The nuclei 1 lie after the representation of the first metal layer 11 or second metal layer 12th then as two-layer particles in a metal matrix, in particular a copper matrix, and a chemical analysis would only detect a contamination of the metal material of a few ppm. In other words: the use of the crystallization nuclei allows the controlled setting of a maximum grain size in the first or the second metal layer without any significant contamination of the first metal layer 11 or the second metal layer 12th in the manufactured metal-ceramic substrate. That is, a possible contamination is within the range of what would be expected as a fluctuation in the composition of the metal raw material. Joining the first metal layer 11 and the second metal layer 12th into a network 2nd (please refer 2a) to the ceramic layer 10th can be done using an AMB connection method, since here the oxygen content is not absolutely necessary to adjust the grain size compared to using the grain refining agent to adjust the grain size.

Reference list

1

Metal-ceramic substrate

2nd

Semi-finished metal product

5

Seed of crystallization

10th

Ceramic layer

11

first layer of metal

12th

second metal layer

15

Ceramic top

S

Stacking direction

D1

first layer thickness

D2

second layer thickness

Claims (6)

A method for producing a metal-ceramic substrate (1), in particular a copper-ceramic substrate, comprising: providing a first metal layer (11), in particular a first copper layer, and a second metal layer (12), in particular a second copper layer, wherein the first metal layer (11) and / or the second metal layer (12) is admixed with a means for influencing the grain size, and - connecting the first metal layer (11) and / or the second metal layer (12) to a Ceramic layer for forming the metal-ceramic substrate (1) in an AMB connection method, wherein the first metal layer (11) and the second metal layer (12) for connection to the ceramic layer are each provided as individual sheets or as a composite in which the first metal layer (11) is joined together with the second metal layer (12), a portion of the agent for influencing the grain size in the first metal layer (11) differing from a portion of the agent for influencing the grain size in the second metal layer (12) distinguishes that in the finished copper-ceramic substrate (1) the first metal layer (11) has a first grain size and the second metal layer (12) has a second grain size that differs from the first grain size, the first metal layer (11) in the finished state faces away from the ceramic layer (10) and the second metal layer (12) faces the ceramic layer (10), the proportion of the agent s for influencing the grain size in the first metal layer (11) is greater than the proportion of the means for influencing the grain size in the second metal layer (12) and the second metal layer (12) is thicker than the first metal layer (11). Procedure according to Claim 1 , wherein the agent for influencing the grain size comprises a grain refining agent, in particular a grain refining agent containing titanium, zirconium, hafnium, chromium and / or niobium. Method according to one of the preceding claims, wherein the means for influencing the grain size comprises crystallization nuclei (5). Procedure according to Claim 3 , wherein the nuclei of the first metal layer (11) and / or the second metal layer (12) are added together with an active metal. Procedure according to Claim 3 or 4th , wherein the first metal layer (11), the second metal layer (12) and the ceramic layer (10) are connected to one another in an AMB connection method, in particular in a common AMB connection method. Method according to one of the Claims 1 to 4th The first metal layer (11) and the second metal layer (12) are joined together in a DCB process to form a composite, in particular a semifinished metal product (2), and the composite is bonded to the ceramic layer (10).

Images