Metal-ceramic composite

Abstract

The present invention relates to a metal-ceramic composite containing a ceramic substrate, which has a front face and a rear face and contains aluminum nitride, and a metal coat on the front face of the ceramic substrate, wherein: the metal coat has at least one recess and a surface of the ceramic substrate in the vicinity of the recess is exposed; the surface of the ceramic substrate exposed by means of the recess has grooves; and the grooves have an average groove distance of 100 nm to 500 nm and an average groove depth of 25 nm to 250 nm.

Description

[0001] DESCRIPTION

[0002] Metal-ceramic composite

[0003] The present invention relates to a metal-ceramic composite which can be used as a ceramic circuit carrier in power electronics.

[0004] In power electronics, printed circuit boards that carry power components such as MOSFETs must be designed for high currents and be able to dissipate waste heat quickly.

[0005] Since ceramic materials such as aluminum oxide, aluminum nitride and silicon nitride have a significantly higher thermal conductivity than polymers used to manufacture conventional printed circuit boards, ceramic circuit boards are often used in power modules.

[0006] An overview of ceramic substrates used in power electronics for the production of ceramic circuit boards can be found, for example, in the following publication:

[0007] N. Chasserio et al., “Ceramic Substrates for High-Temperature Electronic Integration,” Journal of Electronic Materials, Volume 38 (2009), pp. 164-174

[0008] Ceramic substrates based on aluminum nitride exhibit very high thermal conductivity combined with sufficiently high mechanical strength and are therefore very well suited for applications in power electronics.

[0009] A ceramic circuit board contains a ceramic substrate coated with a metal layer on at least one of its sides, usually on both sides. The semiconductor components are applied to one of these metal layers in the final module, while the metal layer on the opposite side of the ceramic substrate can be thermally connected to a heat sink, for example. The ceramic substrate electrically insulates the metal layers from each other.

[0010] The production of a metallized ceramic substrate functioning as a ceramic circuit board, as is known to those skilled in the art, takes place, for example, by bringing the front and back of the ceramic substrate into contact with a metal foil (e.g. a copper or aluminum foil) and bonding them together. The bonding of the metal foils is achieved, for example, by eutectic bonding or active metal brazing (AMB). If the metal foil is a copper foil, the eutectic bonding is also referred to as the DCB or DBC process (DCB: "Direct Copper Bonding" -, DBC: "Direct Bonded Copper"). In the case of aluminum foil, the term "DAB" (Direct Aluminum Bonding) is also used for eutectic bonding. A metallized ceramic substrate produced using a DCB or AM B process is occasionally referred to as a DCB substrate (alternatively: DBC substrate) or AM B substrate.

[0011] For metallization using the DCB process, aluminum nitride substrates, for example, are first surface oxidized. The copper foil is then applied to this thin aluminum oxide layer on the surface of the aluminum nitride substrate using eutectic bonding.

[0012] For metallization by the active soldering process, active solders are used which, in addition to a main component such as Cu, Ag or Au, usually contain one or more elements that can react with the ceramic to form an adhesion-promoting reaction layer (see, for example, Chapter 8.2.4.3 (“Active Soldering”), pages 203-204, in Brevier Technische Keramik, Verband der Keramischen Industrie eV, 2003, Fahner Verlag). Examples of reactive elements used are hafnium (Hf), titanium (Ti), zirconium (Zr), niobium (Nb), cerium (Ce), tantalum (Ta) and vanadium (V). When metallizing an aluminum nitride substrate by active soldering, the reaction layer contains, for example, a nitride of the reactive element (A. Pönicke et al., “Active Soldering of Copper with Aluminum Nitride and Silicon Nitride Ceramics”, Keramische Zeitschrift, 63(5), 2011, 334-342).

[0013] D. Liu et al., "Fabrication of AIN / Cu composite structure via laser surface metallization assisted direct bonding technology," Journal of Manufacturing Processes, Vol. 119, 2024, pp. 640-648, describe a method for bonding a copper foil to an aluminum nitride ceramic substrate. In this method, the surface of the ceramic substrate is irradiated with a laser emitting pulses with a pulse duration in the nanosecond range. The irradiation parameters are selected such that a significant portion of the surface aluminum nitride is decomposed to form metallic aluminum according to the following reaction equation (1):

[0014] 2 AIN -> 2 AI + N2(1)

[0015] The laser treatment forms a metallic aluminum film on the surface of the aluminum nitride, which is used for the direct bonding of a copper foil. The copper foil is applied to the metallized surface of the ceramic substrate and bonded to the ceramic substrate under pressure and at a temperature of 620°C.

[0016] The metal layer carrying the semiconductor components has one or more recesses and is therefore also referred to as a structured metal coating. The structuring is used to form, for example, metallic conductor tracks on the ceramic substrate. Adjacent conductor tracks are spatially separated by the recesses and are therefore electrically insulated from one another. The structuring of AM B substrates can, for example, be carried out in a two-stage process. In a first step, the metal layer is removed in defined areas (e.g. using a first etching solution) and then, in a second step, the adhesion-promoting layer resulting from the active solder process is removed (e.g. using a second etching solution). By removing the metal layer and, if applicable, the adhesion-promoting layer, the ceramic substrate is exposed in defined areas.

[0017] After structuring the metal layer, further wet-chemical treatment steps (e.g., removing the etching mask or applying another metal to the structured metal layer) may be required. These treatment steps can lead to the adsorption of undesired chemical components, particularly ionic compounds (e.g., salts), which may adversely affect the electrical properties of the final power module, onto the exposed ceramic surface. The undesired chemical components can be removed by treating the exposed ceramic surface with suitable cleaning fluids. Ionic residues are removed, for example, by rinsing the ceramic surface with water. It would be advantageous if the ceramic surface was easily wettable by water.

[0018] Modules based on ceramic circuit carriers used in power electronics can be encapsulated, for example, during packaging by embedding the power module in a potting compound. The potting compound or a precursor of the potting compound is applied, for example, using an organic solvent. To ensure the best possible bond between the potting compound and the exposed ceramic surface, it would be advantageous if the ceramic surface was easily wettable by the organic solvent. Providing surfaces that exhibit good wettability with both water and nonpolar organic liquids is fundamentally challenging.

[0019] As mentioned above, aluminum nitride-based ceramic substrates are used as circuit carriers in power electronics due to their very high thermal conductivity combined with high mechanical strength. To maximize their potential as circuit carriers, it would be desirable for the surface of the aluminum nitride substrate exposed after patterning of the metal coating to allow efficient removal of ionic residues by rinsing with water. Optionally, it would also be desirable for the ceramic surface to allow efficient application of a potting compound in an organic solvent.

[0020] One object of the present invention is to provide a metallized aluminum nitride-containing ceramic substrate from whose exposed ceramic surface, ionic residues resulting from production can be removed as easily and efficiently as possible. Optionally, the exposed ceramic surface should be readily wettable by organic solvents.

[0021] The object is achieved by a metal-ceramic composite comprising a ceramic substrate having a front side and a back side and containing aluminum nitride, a metal coating present on the front side of the ceramic substrate, wherein the metal coating has at least one recess and a surface of the ceramic substrate is exposed in the region of the recess, wherein the exposed surface of the ceramic substrate has grooves Rfine, wherein the grooves Rfine have an average groove spacing of 100 nm to 500 nm and an average groove depth of 25 nm to 250 nm.

[0022] Within the scope of the present invention, it was surprisingly discovered that the surface of the aluminum nitride substrate is readily wettable by both water and organic liquids if the surface has the groove structure according to the invention. Water wettability enables efficient removal of unwanted ionic residues from the ceramic surface. Wettability by organic solvents enables efficient application of a potting compound.

[0023] A groove is an elongated depression in the surface of the ceramic substrate.

[0024] As described in more detail below, the groove structure according to the invention is obtained in the surface of the aluminum nitride ceramic substrate by a suitable treatment with an ultrashort pulse laser (pulse duration in the range of pico- or femtoseconds).

[0025] The groove spacing and groove depth are determined using scanning electron micrographs showing the grooves Rfine in cross-section (see “Measurement methods”).

[0026] In an exemplary embodiment, the grooves Rfine have an average groove pitch of 200 nm to 400 nm and / or an average groove depth of 50 nm to 200 nm.

[0027] Optionally, the surface of the ceramic substrate exposed in the area of ​​the metal coating recess can also have grooves R in addition to the grooves Rfine (hereinafter also referred to as “fine grooves”). grO b (“coarse grooves”), with the grooves R grO b have an average groove spacing of 600 nm to 1000 nm. The grooves R grO b, for example, have an average groove depth of 25 nm to 500 nm, more preferably 100 nm to 400 nm. The average groove pitch and the average groove depth of the coarse grooves R grOb are determined in the same way as for the fine grooves Rfein-

[0028] For example, the grooves Rfine and, if present, the grooves R grO b an average groove length of at least 1 .0 pm, more preferably at least 1 .5 pm.

[0029] For example, adjacent grooves run parallel to each other or deviate from a parallel alignment by no more than 20°, preferably no more than 10°.

[0030] For example, at least 20%, more preferably at least 40%, particularly preferably at least 60% of the total area of ​​the surface of the ceramic substrate exposed in the region of the recess has the grooves Rfein according to the invention.

[0031] For example, the following relationship is fulfilled: 0.2 X Ages - A(Rfine) + A(Rcoarse) where

[0032] Ages is the total area of ​​the exposed surface of the ceramic substrate in the region of the recess,

[0033] A(Rfine) is the area of ​​the surface of the ceramic substrate exposed in the region of the recess, which has the grooves Rfine,

[0034] A(Rg ro b) is the area of ​​the surface of the ceramic substrate exposed in the region of the recess, which contains the grooves R grO b.

[0035] Regarding the upper limit, the following applies: A(Rf ejn ) + A(R grO b) A ges

[0036] The following applies preferably: A(Rf ein ) + A(R grO b) s 0.4 x A ges ; even more preferably: A(Rf ejn ) + A(R grO b) s 0.6 x A ges

[0037] For example, A(Rf ejn ) and A(R grO b) the following condition:

[0038] 0 SA(Rcoarse) / A(Rfine) — 5, provided that A(Rf ejn ) + A(R grO b) A ges where A(Rfine), A(R grO b) and A ges each have the meaning given above. If A(R grOb) / A(Rf ejn ) = 0, then the surface of the ceramic substrate exposed in the area of ​​the recess has the fine grooves Rfine, but no coarse grooves R grO b on.

[0039] For example, the ceramic substrate contains the aluminum nitride in a proportion of at least 70 wt%, more preferably at least 80 wt%. Optionally, the ceramic substrate can be surface oxidized, i.e., the surface of the ceramic substrate comprises an aluminum oxide and / or an aluminum oxynitride in addition to the aluminum nitride or instead of the aluminum nitride.

[0040] Optionally, the ceramic substrate may contain one or more metal oxides. These were added, for example, as sintering aids during production of the ceramic substrate. For example, the ceramic substrate contains one or more of the following oxides: one or more alkaline earth metal oxides such as magnesium oxide, one or more transition metal oxides (e.g., one or more rare earth oxides such as yttrium oxide or erbium oxide), a silicon oxide (e.g., SiO2), or a silicate. The ceramic substrate has, for example, a thickness in the range of 0.1 mm to 1.5 mm, more preferably 0.2 mm to 1.0 mm.

[0041] On the front side of the ceramic substrate, there is a metal coating that has at least one recess, so that a surface of the ceramic substrate is exposed through the recess. This metal coating is also referred to below as a structured metal coating. Semiconductor components can be mounted on the structured metal coating.

[0042] For example, the metal coating is separated by the recess into at least two metallic conductor tracks. The distance between adjacent metallic conductor tracks separated by the recess is, for example, at least 0.3 mm, e.g., 0.3 mm to 2 mm. For example, the recess is linear and has a width of at least 0.3 mm, e.g., 0.3 mm to 2 mm.

[0043] Optionally, a metal coating is also present on the back of the ceramic substrate. This back metal coating can optionally also have at least one recess through which a surface of the ceramic substrate is exposed. To achieve the most efficient heat dissipation possible, it may be preferable for the back metal coating not to have such a recess.

[0044] The metal coating present on the front side and optionally the back side of the ceramic substrate is, for example, a copper coating or an aluminum coating. The metal coating has a thickness in the range of 0.05 mm to 1.5 mm, for example, and more preferably 0.2 mm to 0.8 mm.

[0045] If the metal coating is a copper coating, it has, for example, a copper content of at least 97% by weight, more preferably at least 99% by weight.

[0046] If the metal coating is an aluminum coating, it has, for example, an aluminum content of at least 97% by weight, more preferably at least 99% by weight.

[0047] The metal coating can be applied to the front and, if appropriate, the back of the ceramic substrate using methods known to those skilled in the art. For example, a metal foil (e.g., a copper or aluminum foil) is bonded to the front of the ceramic substrate by active soldering.

[0048] In active soldering, for example, a bond is created between the metal foil and the ceramic substrate using an active solder at a temperature of approximately 600-1000°C. Due to their alloy composition, active solders are capable of wetting non-metallic, inorganic materials such as ceramic substrates. In addition to a main component such as copper, silver, and / or gold, active solders also contain one or more active metals such as Hf, Ti, Zr, Nb, V, Ta, or Ce, which can react with the ceramic substrate to form a reaction layer.

[0049] Preferably, a reaction layer resulting from the active soldering is present between the metal coating and the front side of the ceramic substrate. The reaction layer contains, for example, one or more ERS elements selected from Hf, Ti, Zr, Nb, V, Ta, and Ce, preferably selected from Hf, Ti, Zr, Nb, and Ce, more preferably selected from Hf, Ti, and Zr. The ERS element in the reaction layer is particularly preferably titanium. For example, the ERS elements are present in the reaction layer in the form of a nitride and / or oxynitride and / or aluminide. For example, the reaction layer contains the ERS elements in a total amount of at least 50 wt%. For example, the reaction layer contains the nitrides and / or oxynitrides and / or aluminides of the ERS elements in a total amount of at least 70 wt%, more preferably at least 85 wt%. In a semiconductor module for power electronics, the migration of silver can lead to problems.It may therefore be preferred that the reaction layer contains silver in a proportion of not more than 5 wt%, more preferably not more than 1 wt%, or is even silver-free.

[0050] Exposing a surface of the ceramic substrate, for example, occurs in several steps. First, the metal coating is removed, for example, by etching, and the reaction layer formed during active soldering is exposed. Subsequently, the exposed reaction layer is removed. Preferably, the exposed reaction layer is removed using an ultrashort pulse laser with a pulse duration in the picosecond or femtosecond range (e.g., an IR picosecond or femtosecond laser).

[0051] In order to electrically isolate the areas of the metal coating separated by the recess, it would be sufficient to carry out the material removal with the pulsed laser only for such a time or under such conditions until the electrically conductive material has been essentially completely removed in the treated areas, but a modification of the surface of the ceramic substrate is essentially avoided.

[0052] Within the scope of the present invention, however, the pulsed laser treatment not only exposes the surface of the ceramic substrate, but also modifies this surface. As a result of this modification, the surface of the ceramic substrate has grooves Rfine in the region of the recess, with the grooves Rfine having an average groove spacing of 100 nm to 500 nm and an average groove depth of 25 nm to 250 nm.

[0053] For the production of a metal-ceramic substrate having the groove structure according to the invention, it has proven advantageous if the pulses of the laser beam have a pulse duration of < 100 picoseconds (e.g. 1 ps to 100 ps or 1 fs to < 1000 fs) and a pulse energy in the range of 5 pJ to 50 pJ and the laser beam is guided over the surface to be processed at a scanning speed such that the ratio of scanning speed (in mm / s) to pulse energy (in pJ) is at least 30 mm / (s x pJ), for example 30 to 4000 mm / (s x pJ), more preferably at least 50 mm / (s x pJ), for example 50 to 3000 mm / (s x pJ).

[0054] Within the scope of the present invention, it is also possible to first expose the ceramic substrate surface by a single- or multi-stage etching process and then to treat the exposed ceramic substrate surface with the pulsed laser beam until the groove structure according to the invention is realized.

[0055] However, for reasons of process efficiency, it may be preferable to use the pulsed laser both for exposing the ceramic substrate surface and for modifying this surface.

[0056] An exposed surface of the ceramic substrate exhibiting the groove structure according to the invention is readily wettable by both water and organic liquids. Water wettability enables efficient removal of unwanted ionic residues from the ceramic surface. Wettability by organic solvents enables efficient application of a potting compound.

[0057] If the pulsed laser is used both to expose the ceramic substrate surface and to modify this surface, it is advantageous to remove the reaction layer resulting from the active soldering process and exposed by etching to at least such an extent that only a very small amount of this reaction layer remains on the surface of the ceramic substrate. This very small amount of remaining reaction layer is not sufficient to form a continuous layer. Rather, the remaining reaction layer exists in the form of individual islands, and the exposed surface of the ceramic substrate lies between the islands of the remaining reaction layer. For the cleaning properties of the exposed surface with water as a cleaning agent, it has proven advantageous if the reaction layer is removed as completely as possible by treatment with the USP laser.In an exemplary embodiment, one or more ERS elements selected from Hf, Ti, Zr, Nb, V, Ta, and Ce, preferably selected from Hf, Ti, Zr, Nb, and Ce, more preferably selected from Hf, Ti, and Zr, and particularly preferably Ti, are present in the region of the recess of the metal coating in a total concentration of not more than 1.0 wt.%, more preferably not more than 0.6 wt.%, for example in the form of a nitride and / or oxynitride and / or aluminide. The concentration of the ERS elements is determined by energy-dispersive X-ray spectroscopy (EDX) coupled with scanning electron microscopy (SEM-EDX).

[0058] Alternatively, the ultrashort pulse laser treatment can be performed in such a way that the reaction layer resulting from the active soldering process and exposed by etching is completely removed, so that none of the ERS elements (Hf, Ti, Zr, Nb, V, Ta, and Ce) can be detected in the area of ​​the recess in the metal coating by energy-dispersive X-ray spectroscopy (EDX) coupled with scanning electron microscopy (SEM-EDX). If complete removal of the reaction layer is desired, it is advantageous if the exposed surface of the ceramic substrate is not exposed to the ultrashort pulse laser for too long, as this can lead to further material removal and a reduction in the thickness of the ceramic substrate. An excessively thin ceramic substrate can have a detrimental effect on its dielectric strength.For example, the treatment with the ultrashort pulse laser is carried out such that the thickness of the ceramic substrate is reduced by less than 10% during the ultrashort pulse laser treatment. Preferably, the treatment with the ultrashort pulse laser is carried out such that the thickness of the ceramic substrate does not change during the ultrashort pulse laser treatment or is reduced by at least less than 7%, more preferably less than 5%, and even more preferably less than 3%. This can have a beneficial effect on the dielectric strength of the ceramic substrate. The thickness of the ceramic substrate before and after treatment with the ultrashort pulse laser can be determined using an SEM image of a cross-section of the metal-ceramic composite. In an exemplary embodiment, the following applies:

[0059] (dK1 - dK2) / dK1 < 0.10 where dK1 is the thickness of the ceramic substrate in a region where the metal coating is present on the ceramic substrate, dK2 is the thickness of the ceramic substrate in the region where the recess of the metal coating is present.

[0060] The thickness dK1 takes into account the ceramic substrate, not the metal coating on top. Furthermore, dK1 > dK2

[0061] Preferably: (dK1 - dK2) / dK1 < 0.07

[0062] Even more preferred is: (dK1 - dK2) / dK1 < 0.03

[0063] The thicknesses dK1 and dK2 of the ceramic substrate can be determined using an SEM image of a cross-section of the metal-ceramic composite.

[0064] The present invention also relates to a semiconductor module comprising the metal-ceramic composite described above and one or more semiconductor components.

[0065] Optionally, the semiconductor module contains a potting compound, which contacts the surface of the ceramic substrate of the metal-ceramic composite exposed by the recess. Potting compounds for electronic components are known to those skilled in the art. The potting compound contains, for example, a polymer (e.g., a thermoplastic or thermosetting polymer). For example, the potting compound contains an optionally cured epoxy resin or silicone resin, a polyurethane, or an inorganic cement (e.g., a phosphate cement).

[0066] Measurement methods

[0067] Average groove spacing and average groove depth

[0068] SEM images are taken that show the grooves in cross-section. The groove spacing and groove depths are determined from the profile of the surface of the ceramic substrate in the SEM image. An example cross-sectional profile of a grooved surface of a ceramic substrate in an SEM image is illustrated in Figure 1. Figure 1 shows a surface profile 1 with grooves 2 in cross-section. Figure 1 also shows

[0069] - vertical lines vGmin, which run in the vertical direction (relative to the image plane of the SEM image) through the local minima 3 of the surface profile,

[0070] - horizontal straight lines hGmin, which run in the horizontal direction (relative to the image plane of the SEM image) through the local minima 3 of the surface profile,

[0071] - horizontal lines hG max, which run in the horizontal direction (relative to the image plane of the SEM image) through the local maxima 4 of the surface profile.

[0072] The groove spacing is determined using the vertical line vGmin and the groove depth is determined using the horizontal lines hGmin and hGmax.

[0073] A local minimum is the lowest point in the cross-section of a groove. Each local minimum is adjacent to two local maxima.

[0074] The groove pitch (i.e. the distance between adjacent grooves) and the mean groove pitch are determined as follows:

[0075] In the profile of the surface of the ceramic substrate, straight lines are drawn that run vertically (relative to the image plane of the SEM image) through the local minima of the surface profile. In Figure 1, these are the straight lines vGmin. The groove spacing is the distance between adjacent straight lines vGmin. The mean groove spacing is the arithmetic mean of the groove spacings. At least 50 groove spacings are considered to determine the mean groove spacing.

[0076] The depth of a groove (groove depth) and the average groove depth are determined as follows:

[0077] In the profile of the surface of the ceramic substrate, straight lines are drawn that run horizontally (relative to the image plane of the SEM image) through the local minima (see straight line hGmin in Figure 1) and local maxima (see straight line hGmax in Figure 1) of the surface profile. The depth T of an individual groove (groove depth) is determined from the following relationship:

[0078] T = 0.5 x (di + d2) where di is the distance between the vertical line passing through the local minimum (hGmin in Figure 1) and the vertical line passing through the first adjacent local maximum (hG ma x in Figure 1), d2 is the distance between the vertical line passing through the local minimum (hLmin in Figure 1) and the vertical line passing through the second adjacent local maximum (hGmax in Figure 1).

[0079] The mean groove depth is the arithmetic mean of the groove depths. At least 50 groove depths are considered to determine the mean groove depth. The mean groove depth is determined based on the grooves that were already considered for determining the mean groove pitch.

[0080] To take an SEM image showing the grooves in cross-section, the ceramic substrate is cut perpendicular to the groove direction. Cutting of the ceramic substrate is performed using a focused ion beam (FIB). A liquid metal ion gallium ion source (Ga +) is used to bombard the sample surface with an intense focused ion beam, thus achieving precise local material removal and cutting the ceramic substrate. Before cutting with the focused ion beam, the sample surface is coated with a tungsten layer up to 2 μm thick. The resulting tungsten layer protects the sample surface from radiation damage caused by the Ga ion beam. The sample is then tilted 45° from the horizontal position for sectioning with the focused ion beam. In the FIB column, the elemental Ga is evaporated and converted to Ga + ionized. The Ga ions are then accelerated by an electric field of 30 kV and focused onto a specific point on the sample surface by an electrostatic lens.

[0081] An SEM image is taken of the cut ceramic substrate, showing the grooves in cross-section. A scanning electron microscope (Gemini Ultra 55, ZEISS Ltd) with a silicon drift EDX detector (Ultimax 100 from Oxford Instruments) and analysis software (AZtec, Oxford Instruments) were used for the examination. To prepare for the examination, the sample surface is first coated with a very thin (a few nm thick) layer of carbon (for example, using the SCD 005 sputter coater with CEA 035 carbon evaporation supply, BalTec AG). The sample is then placed in the sample chamber, and the chamber is vacuumized. The SEM is then examined. The following settings were used: magnification: 100x, 500x, and 1000x, accelerating voltage = 15 kV

[0082] The length of a groove is determined using a SEM image taken in a top view of the surface of the ceramic substrate. The mean groove length is the arithmetic mean of the groove lengths. At least 50 groove lengths are considered to determine the mean groove length. The mean groove length is preferably determined using the grooves considered for determining the mean groove spacing.

[0083] Composition of the adhesion promoter layer

[0084] The composition of the adhesion-promoting layer is determined by energy-dispersive X-ray spectroscopy (EDX) coupled with scanning electron microscopy (SEM-EDX).

[0085] In SEM-EDX, a focused primary electron beam is scanned point by point across the sample surface. The scattered electrons are recorded by a detector, with the number of electrons per pixel producing a microscopic image of the sample surface in grayscale. In addition, the primary electron beam excites the sample to emit characteristic X-rays, allowing the elements in the sample and their weight fraction to be determined by analyzing the energy spectrum with an EDX detector.

[0086] A scanning electron microscope (e.g., Gemini Ultra 55, ZEISS Ltd) with a silicon drift EDX detector (e.g., Ultimax 100, Oxford Instruments) and analysis software (e.g., AZtec, Oxford Instruments) are preferably used for the examination. To prepare for the examination, the sample surface is first coated with a very thin (a few nm thick) layer of carbon (e.g., using the SCD 005 sputter coater with CEA 035 carbon evaporation supply, BalTec AG). The sample is then placed in the sample chamber, and the chamber is placed under vacuum. Subsequently, the examination is carried out using SEM-EDX. The following settings are preferably used: magnification: 100x, 500x, and 1000x; acceleration voltage = 15 kV. A separate EDX spectrum is recorded for each point measured on the sample surface.All recorded EDX spectra are processed by the analysis software to determine a quantitative chemical composition for each point. This allows the average elemental content (e.g., the active metal) at the sample surface to be quantitatively determined. The elemental content is determined both in atomic percent and in weight percent, with the total amount representing 100%. The analysis is preferably performed at 100x magnification.

[0087] SEM-EDX allows the composition of the adhesion promoter layer to be determined both qualitatively (detection of specific elements and phases, e.g., a metal nitride phase present in the adhesion promoter layer) and quantitatively. The measurement is performed, for example, at at least 10 locations on the adhesion promoter layer.

[0088] Examples

[0089] Six aluminum nitride starting substrates from one manufacturing batch were used, which had matching dimensions (174 mm x 139 mm x 0.32 mm) and a matching dielectric strength of 18 kV / mm.

[0090] The aluminum nitride substrates were metallized using identical active soldering processes under the conditions described below.

[0091] An active solder paste was screen-printed onto one side of the ceramic substrate, measuring 168 mm x 130 mm, and pre-dried for 15 minutes at 125°C. The active solder paste consisted of 67 weight percent copper powder, 19.8 weight percent tin powder, 3.7 weight percent titanium hydride, and 9.5 weight percent of an organic vehicle. The paste thickness after pre-drying was 25 + / - 5 μm. A copper foil measuring 170 mm x 132 mm x 0.3 mm was then placed on the pre-dried paste. The resulting assembly was then flipped over, the paste was similarly screen-printed onto the opposite side of the ceramic substrate, pre-dried, and coated with a copper foil to create a sandwich assembly.The sandwich assembly was loaded with a 1 kg weight, fired at a maximum temperature of 910°C for 20 minutes, and then cooled to room temperature to obtain an unstructured metal-ceramic composite. Due to the active brazing process, an adhesion-promoting reaction layer exists between the metal coating and the ceramic substrate. This layer contains titanium (e.g., in the form of a nitride). In each of the six copper-ceramic composites, the copper coating was structured in an identical manner using an etching treatment.

[0092] First, etching masks were applied to the copper layers of the copper-ceramic composites. The unmasked areas of the copper coatings were then wet-chemically etched. A copper chloride solution containing hydrochloric acid and hydrogen peroxide (copper ion content = 160 g / l) was used as the etching solution. Etching was carried out at a temperature of 50°C and a spray pressure of 2.8 bar. The unmasked copper coating was removed by etching. However, the reaction layer resulting from the active soldering process was not removed by the copper chloride solution.

[0093] The removal of the reaction layer resulting from the active soldering process and exposed by etching was carried out under different conditions.

[0094] In Comparative Example VB1, the exposed reaction layer was removed with an etching solution containing ammonium fluoride, fluoroboric acid and hydrogen peroxide.

[0095] In Examples EB1 to EB5 according to the invention, the exposed reaction layer was removed by treatment with a picosecond ultrashort pulse laser. The laser pulses had a pulse frequency of 2000 kHz. The scanning speed of the laser beam and pulse energy were varied as shown in Table 1.

[0096] Following removal of the reaction layer, the respective etch masks of Examples EB1-EB5 and Comparative Example VB1 were removed in a stripping system using a 2.5% sodium hydroxide solution. In each of the examples, the NaOH solution also came into contact with the exposed surface of the aluminum nitride substrate, leaving ionic residues on the ceramic surface. To remove ionic residues from the ceramic surface, all samples were thoroughly rinsed with the same amount of water for the same time.

[0097] Any ionic residues remaining on the sample surface after rinsing with water were detected by conductivity measurement. For this purpose, each sample was placed in an equal amount of distilled water. The water was homogenized by a stirrer above the respective sample surface. It was checked whether the electrical conductivity of the water increased within 30 minutes. If the electrical conductivity did not increase or increased only slightly, this indicates that the ionic residues were essentially completely removed from the surface of the aluminum nitride ceramic substrate after the first rinse with water.If, however, a significant increase in electrical conductivity is observed, this indicates that significant amounts of ionic residues remain on the surface of the aluminum nitride ceramic substrate after the first rinsing with water and only gradually detach from the ceramic surface in the water bath.

[0098] The contact angles of the exposed ceramic surfaces of examples EB1 to EB5 and VB1, as well as the surface of the original substrate, were determined. The measurements were performed using the OCA 15EC contact angle measuring device from Dataphysics. Distilled water and diiodomethane were used for the determination. Contact angles were measured on at least 20 drops.

[0099] For the aluminum nitride starting substrate, a contact angle of 24.6° was obtained for water and 37.9° for diiodomethane. The exposed ceramic surfaces of the inventive examples EB1 to EB5 were fully wettable with both water and diiodomethane (i.e., contact angle > 0°). In comparative example VB1, a contact angle of 21.3° was obtained for water and 35.1° for diiodomethane.

[0100] For each of the samples from Examples EB1 to EB5 and VB1, a top-view SEM image of the exposed surface of the ceramic substrate was taken to check for the presence of a groove structure. A top-view SEM image of the surface of the aluminum nitride starting substrate was also taken. The SEM images of the ceramic surfaces of Examples EB1 to EB5 according to the invention each showed a groove structure, while no grooves could be identified in the SEM images of the ceramic substrate of Comparative Example VB1 and the starting substrate.

[0101] To further characterize the grooves in the ceramic surfaces of inventive examples EB1 to EB5, the ceramic substrates were sectioned perpendicular to the groove direction using a focused ion beam. SEM images were taken of these sample cross-sections. The average groove spacing and average groove depth were determined from the SEM images. The surfaces of the ceramic substrates exposed by treatment with the ultrashort pulse laser each exhibited a groove structure, with the grooves each having an average groove spacing in the range of 100-500 nm and an average groove depth in the range of 25 to 250 nm.

[0102] The results are summarized in Table 1.

[0103] Table 1 : Laser settings, surface structures and conductivity test results Through treatment with the pulsed laser, the surface structure of the aluminum nitride substrates was modified. Compared to the original substrate and the substrate exposed by etching, they exhibit a grooved surface finish that allows for improved wettability with water. This significantly improves the removal of ionic contaminants resulting from the manufacturing processes. This enables shorter process times (rinsing steps) during production and fewer bath changes in subsequent steps (e.g., metallization baths). Furthermore, wettability with nonpolar liquids is also significantly improved, improving the contact of the ceramic with encapsulating compounds such as silicones.

Claims

CLAIMS 1. A metal-ceramic composite comprising a ceramic substrate having a front side and a back side and containing aluminum nitride, a metal coating present on the front side of the ceramic substrate, wherein the metal coating has at least one recess and a surface of the ceramic substrate is exposed in the region of the recess, wherein the exposed surface of the ceramic substrate has grooves Rfine, wherein the grooves Rfine have an average groove spacing of 100 nm to 500 nm and an average groove depth of 25 nm to 250 nm, wherein the average groove spacing and the average groove depth are determined on scanning electron micrographs of cross sections of the grooves Rfine according to the method specified in the description.

2. The metal-ceramic composite according to claim 1, wherein the grooves Rfein have an average groove pitch of 200 nm to 400 nm and an average groove depth of 50 nm to 200 nm.

3. The metal-ceramic composite according to any one of the preceding claims, wherein the metal coating is a copper or aluminum coating.

4. The metal-ceramic composite according to any one of the preceding claims, wherein a reaction layer is present between the ceramic substrate and the metal coating, the reaction layer containing one or more elements ERS selected from Ti, Hf, Zr, Nb, V, Ta and Ce.

5. The metal-ceramic composite according to one of the preceding claims, wherein in the region of the recess one or more elements ERS, selected from Hf, Ti, Zr, Nb, V, Ta and Ce, are present in a total concentration of < 1 wt%, wherein the determination is carried out by energy-dispersive X-ray spectroscopy coupled with scanning electron microscopy (SEM-EDX).

6. The metal-ceramic composite according to any one of the preceding claims, wherein the ceramic substrate contains the aluminum nitride in a proportion of at least 70% by weight.

7. A semiconductor module comprising the metal-ceramic composite according to any one of claims 1-6, one or more semiconductor components.

8. The semiconductor module of claim 7, further comprising: a potting compound, wherein the potting compound contacts the surface of the ceramic substrate of the metal-ceramic composite exposed by the recess.

Images