JP2018138687A - Connection texture, method for manufacturing connection texture, connection structure, and semiconductor device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a connection texture, connection structure, and semiconductor device having a connection reliability sufficient even in a temperature cycling test including high-temperature conditions.SOLUTION: A connection texture 30a comprises a nickel-containing layer 31, palladium-containing layers 33, 34, and a sintered-metal layer 32 in this order. The palladium-containing layer comprises a palladium layer 33 and an intermetallic compound layer 34 containing palladium and a metal element included in the sintered-metal layer.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a connection structure, a method for manufacturing the connection structure, a connection structure, and a semiconductor device.

  When manufacturing a semiconductor device, various bonding materials are used for bonding a semiconductor element and a lead frame (support member). Among semiconductor devices, high melting point lead solder has been used as a bonding material for bonding power semiconductors, LSIs, and the like that are operated at a high temperature of 150 ° C. or higher. In recent years, the demand for operating at a high temperature of 175 ° C. or more has been increased due to the increase in capacity and space saving of semiconductor elements, and the heat resistance and thermal conductivity are insufficient in the joint made of a high melting point lead solder layer, It has become difficult to ensure connection reliability. On the other hand, with the strengthening of RoHS regulations, a bonding material not containing lead is required.

  So far, joining of semiconductor elements using materials other than lead solder has been studied. For example, Patent Document 1 below proposes a technique for sintering silver nanoparticles at a low temperature to form a sintered silver layer. Such sintered silver is known to have high connection reliability with respect to the power cycle.

  As another material, a technique for sintering a copper particle to form a sintered copper layer has also been proposed. For example, Patent Document 2 below discloses a bonding copper paste containing cupric oxide particles and a reducing agent as a bonding material for bonding a semiconductor element and an electrode. Patent Document 3 listed below discloses a bonding material including copper nanoparticles and copper microparticles or copper submicroparticles, or both.

Japanese Patent No. 4928639 Patent No. 5006081 JP 2014-167145 A

  By the way, an electroless nickel plating process may be performed as a processing method for a semiconductor element and a substrate on which the semiconductor element is mounted. However, the electroless nickel plating surface formed by the treatment, and sintered silver as disclosed in Patent Document 1, or sintered copper as disclosed in Patent Document 2 and Patent Document 3, When bonded, it is difficult to obtain sufficient bonding strength. In addition, the electroless nickel plating surface may be further subjected to electroless gold plating, but it is sufficient for joining the surface of the electroless gold plating on the electroless nickel plating and sintered silver or sintered copper. It is difficult to obtain a good bonding strength. In high-temperature operation of a semiconductor device, it is necessary to ensure connection reliability at the junction of semiconductor elements, and solving this is an important issue.

  An object of the present invention is to provide a connection structure having sufficient connection reliability even in a temperature cycle test including a high temperature condition, a manufacturing method thereof, a connection structure, and a semiconductor device.

  As a result of intensive studies by the present inventors in order to solve the above problems, a palladium film is formed on the surface of the nickel film, and a laminate in which a metal paste is further applied thereon is prepared and sintered. Thus, the present inventors have found that a connection structure that is excellent in bonding strength at the connection interface and that can exhibit sufficient connection reliability even in a temperature cycle test including a high temperature condition can be obtained, and the present invention has been completed.

  That is, the present invention provides a connection structure including a nickel-containing layer, a palladium-containing layer, and a sintered metal layer in this order.

  A connection structure including the connection structure of the present invention can have sufficient connection reliability even in a temperature cycle test including a high temperature condition. The reason why such an effect can be obtained is considered as follows.

  The electroless nickel plating film generally contains several mass% or more of phosphorus, boron and the like. Therefore, it is easy to oxidize compared with the film whose nickel content (purity) is 99.5 mass% or more, such as an electrolytic nickel plating film or a nickel film obtained by sputtering. Thereby, it is considered that sufficient connection strength between the electroless nickel plating film surface and sintered silver or sintered copper is difficult to obtain. In order to suppress oxidation of the electroless nickel plating film surface, electroless gold plating may be applied to the surface of the electroless nickel plating film. Since the surface of the electroless nickel plating film is corroded by the reaction, even if gold and sintered silver or sintered copper are joined, the connection strength between nickel and gold is weak. I can't get it. In addition, nickel contained in the electroless nickel plating film may diffuse through the inside of the electroless gold plating film, and a nickel oxide film may be formed on the surface of the electroless gold plating film. It is considered that the joining with sintered silver or sintered copper becomes insufficient and it is difficult to obtain sufficient connection strength.

  On the other hand, when performing electroless palladium plating on an electroless nickel plating film, the surface of the electroless nickel plating film is hardly corroded, so the connection strength between the electroless nickel plating film and the electroless palladium plating film is high. Become. Moreover, even if the electroless palladium plating film has a thickness of about 3 nm, the effect of sufficiently suppressing the diffusion of nickel from the electroless nickel plating film to the surface of the electroless palladium plating film can be obtained. Therefore, nickel oxide is not formed on the surface of the electroless palladium plating film, and the bonding between the electroless palladium plating film and sintered silver or sintered copper is sufficient, and good connection strength is easily obtained. Think. From the above, the connection structure of the present invention can be obtained by using electroless nickel plating and electroless palladium plating. That is, the nickel-containing layer can be derived from an electroless nickel plating film, and the palladium-containing layer can be derived from at least an electroless palladium plating film.

  When the thickness of the electroless palladium plating film is somewhat thick (for example, more than 30 nm), a metal containing palladium and silver or copper at the interface between the electroless palladium plating film and sintered silver or sintered copper An intermetallic compound layer (a layer formed by the diffusion of palladium and silver or copper when the sintered metal layer is formed) is formed. That is, in the connection structure, the palladium-containing layer includes a palladium layer and an intermetallic compound layer containing a metal element contained in palladium and the sintered metal layer. In this case, although the electroless palladium plating film partially remains as a palladium layer, it is connected even when the electroless nickel plating film is used alone or when an electroless gold plating film is provided on the electroless nickel plating film. Strength is further improved.

  When the thickness of the electroless palladium plating film is thin to some extent (for example, 3 nm or more and 30 nm or less), nickel, palladium and silver or copper are present at the interface between the electroless nickel plating film and sintered silver or sintered copper. An intermetallic compound layer (a layer formed by the diffusion of nickel, palladium and silver or copper when the sintered metal layer is formed) is formed. That is, in the connection structure, the palladium-containing layer includes an intermetallic compound layer containing nickel, palladium, and a metal element contained in the sintered metal layer. In this case, the electroless palladium plating film does not remain alone, and better connection strength can be easily obtained as compared with the case where a part of the electroless palladium plating film remains.

  In the present invention, the sintered metal layer includes a structure derived from flaky copper particles oriented substantially parallel to the connection interface of the connection structure, and copper based on the total volume of the sintered metal layer, It is preferable to contain 65 volume% or more. Thereby, connection strength can be improved more.

  The present invention also provides a connection structure including the above connection structure between a first member and a second member. Such a connection structure has excellent connection reliability.

  The present invention further provides a semiconductor device comprising the above connection structure between a first member and a second member, wherein at least one of the first member and the second member is a semiconductor element. To do. Such a semiconductor device has excellent connection reliability. When the first member is a semiconductor element, the second member may be a metal wiring.

  The present invention is also a method for manufacturing a connection structure including a nickel-containing layer, a palladium-containing layer, and a sintered metal layer in this order, and a laminate in which a nickel coating, a palladium coating, and a metal paste are provided in this order. A manufacturing method is provided, comprising the step of sintering the body. Thereby, the connection structure excellent in the connection reliability can be obtained.

  In the production method of the present invention, the nickel coating may be formed by electroless nickel plating, and the palladium coating may be formed by electroless palladium plating. As described above, the nickel coating and the palladium coating are preferably obtained by electroless plating.

  In the production method of the present invention, the nickel coating preferably contains 80% by mass or more of nickel based on the total mass of the nickel coating. Moreover, it is preferable that a palladium film contains 94 mass% or more of palladium on the basis of the total mass of a palladium film. Thereby, connection reliability can be improved more.

  In the production method of the present invention, the thickness of the palladium coating is preferably 3 nm or more and 0.5 μm or less. Thereby, connection reliability can be improved more.

  In the production method of the present invention, the metal paste preferably contains metal particles and a dispersion medium, and the metal particles preferably contain flaky copper particles. Thereby, connection reliability can be improved more.

  ADVANTAGE OF THE INVENTION According to this invention, the connection structure which has sufficient connection reliability also in the temperature cycle test containing high temperature conditions, its manufacturing method, a connection structure, and a semiconductor device can be provided.

It is a schematic cross section which shows an example of the connection structure of this embodiment. It is a schematic cross section which shows an example of the connection structure of this embodiment. It is a SEM image which shows the cross section of the structure derived from flaky copper particle. It is a schematic cross section which shows an example of the connection structure of this embodiment. It is a schematic cross section which shows an example of the semiconductor device of this embodiment. It is a schematic cross section which shows an example of the semiconductor device of this embodiment.

  Hereinafter, a mode for carrying out the present invention (hereinafter referred to as “the present embodiment”) will be described in detail. The present invention is not limited to the following embodiments.

  Hereinafter, preferred embodiments will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals, and redundant description is omitted.

<Connection structure>
The connection structure of this embodiment is a connection structure including a nickel-containing layer, a palladium-containing layer, and a sintered metal layer in this order.

  FIG. 1 is a schematic cross-sectional view showing an example of a connection structure 30a of the present embodiment (first embodiment). The connection structure 30 a includes a nickel-containing layer 31, palladium-containing layers 33 and 34, and a sintered metal layer 32 in this order, and the palladium-containing layer is a metal contained in the palladium layer 33, palladium, and the sintered metal layer. An intermetallic compound layer 34 containing an element.

  FIG. 2 is a schematic cross-sectional view showing an example of the connection structure 30b of the present embodiment (second embodiment). The connection structure 30b includes a nickel-containing layer 31, a palladium-containing layer 35, and a sintered metal layer 32 in this order, and the palladium-containing layer includes a metal element containing nickel, palladium, and a metal element contained in the sintered metal layer. And a compound layer 35.

  The connection structure shown in FIG. 1 and FIG. 2 is obtained by sintering a laminate in which a nickel coating, a palladium coating, and a metal paste are provided in this order. More specifically, for example, after a nickel coating and a palladium coating are formed in this order on a predetermined member, the metal paste is further sintered on the palladium coating in a state where a metal paste described later is applied on the palladium coating. A metal layer is formed). A gold coating may be further formed on the surface of the palladium coating.

  The nickel coating can be formed, for example, by electroless nickel plating. The nickel coating preferably contains 80% by mass or more of nickel based on the total mass of the nickel coating. From the viewpoint of further improving the reliability, the content is more preferably 85% by mass or more, and further preferably 90% by mass or more. The nickel coating preferably contains at least one selected from the group consisting of phosphorus and boron, and more preferably contains phosphorus (preferably 20% by mass or less). When the nickel coating is formed by electroless plating, for example, phosphorus can be co-deposited by using a phosphorus-containing compound such as sodium hypophosphite as a reducing agent. An electrolytic nickel plating film can be formed. Further, by using a boron-containing compound such as dimethylamine borane, sodium borohydride, potassium borohydride as a reducing agent, boron can be co-deposited, and electroless nickel containing a nickel-boron alloy A plating film can be formed. The nickel coating can also be formed by other electrolytic plating, sputtering, or the like.

  The thickness of the nickel coating is preferably 0.2 μm or more and 20 μm or less, more preferably 0.3 μm or more and 15 μm or less, and further preferably 0.5 μm or more and 10 μm or less. By setting the thickness to 0.2 μm or more, the nickel coating sufficiently remains after sintering, and it becomes easier to obtain better connection reliability. The reason why the upper limit of the thickness is 20 μm is exclusively for economic reasons.

  The palladium film can be formed by, for example, electroless palladium plating. As electroless palladium plating, either a substitution type that does not use a reducing agent or a reduction type that uses a reducing agent may be used. Examples of the electroless palladium plating solution include MCA (trade name, manufactured by World Metal Co., Ltd.) and the like for the replacement type, and APP (trade name, manufactured by Ishihara Pharmaceutical Co., Ltd.) and the like for the reduction type. When the substitution type and the reduction type are compared, the reduction type is preferable from the viewpoint of easily suppressing voids and ensuring the covering area. The palladium film can also be formed by other electrolytic plating, sputtering, or the like.

The source of palladium in the electroless palladium plating solution is not particularly limited, and examples thereof include palladium compounds such as palladium chloride, sodium palladium chloride, ammonium palladium chloride, palladium sulfate, palladium nitrate, palladium acetate, and palladium oxide. Specifically, acidic palladium chloride “PdCl ₂ / HCl”, tetraamminepalladium nitrate “Pd (NH ₃ ) ₄ (NO ₃ ) ₂ ”, dinitrodiammine palladium “Pd (NH ₃ ) ₂ (NO ₂ ) ₂ ”, dicyano Diammine palladium “Pd (CN) ₂ (NH ₃ ) ₂ ”, dichlorotetraammine palladium “Pd (NH ₃ ) ₄ Cl ₂ ”, palladium sulfamate “Pd (NH ₂ SO ₃ ) ₂ ”, diammine palladium sulfate “Pd (NH ₃ ) ₂ SO ₄ ”, tetraamminepalladium oxalate“ Pd (NH ₃ ) ₄ C ₂ O ₄ ”, palladium sulfate“ PdSO ₄ ”and the like can be used.

  Although there is no restriction | limiting in particular as a reducing agent in an electroless palladium plating liquid, From a viewpoint of raising the palladium content in the palladium coating obtained enough, and also suppressing the thickness dispersion | variation in a palladium coating further, a formic acid compound (for example, sodium formate) is used. preferable.

  The palladium coating can contain phosphorus, boron and the like. As a reducing agent used in the electroless palladium plating solution, for example, phosphorus-containing compounds such as hypophosphorous acid and phosphorous acid; boron-containing compounds can be used to contain these elements. By using such a reducing agent, a palladium film containing a palladium-phosphorus alloy or a palladium-boron alloy can be obtained. In addition, what is necessary is just to adjust the density | concentration of reducing agent, pH, the temperature of a plating solution, etc. in order to make palladium content in a palladium film into a desired range.

  A complexing agent, a buffering agent, and the like can be added to the electroless palladium plating solution as necessary. Examples of complexing agents include ethylenediamine and tartaric acid. The buffer is not particularly limited and can be used.

  The palladium content in the palladium coating is preferably 94% by mass or more, more preferably 97% by mass or more, and further preferably 99% by mass or more, based on the total mass of the palladium coating. The upper limit of the palladium content in the palladium coating is 100% by mass. When the palladium content in the palladium coating is in the above range, an intermetallic compound of palladium and a metal element contained in the sintered metal layer (metal paste) is easily formed. Thereby, as shown in FIG. 2, it becomes easy to obtain the connection structure 30b in which the palladium layer 33 is not formed (that is, the palladium film does not remain), and the connection reliability is further improved.

  The thickness of the palladium coating is preferably 3 nm or more and 0.5 μm or less, more preferably 3 nm or more and 0.1 μm or less, and further preferably 3 nm or more and 30 nm or less. By setting the thickness to 3 nm or more, an effect of easily suppressing the diffusion of nickel to the surface of the palladium coating can be obtained. Moreover, the reason why the preferable upper limit of the thickness is set to 0.5 μm is exclusively an economical reason. The thickness of 3 nm or more and 30 nm or less is particularly preferable because the connection structure 30b is obtained in which the palladium layer 33 is not formed (that is, the palladium film does not remain) as shown in FIG. 2 when the sintered metal layer is formed. This is because the connection reliability is further improved.

  In the case where the palladium layer 33 is left as shown in FIG. 1 while maintaining connection reliability, the thickness of the palladium film may be, for example, more than 30 nm and 0.5 μm or less, or more than 30 nm and 500 nm or less. Alternatively, it may be 50 nm or more and 300 nm or less. However, only the thickness of the palladium coating does not contribute to the formability of the connection structure of FIG. 1 or FIG. 2, and the thickness is only a guide.

  The gold film can be formed, for example, by electroless gold plating. The gold content in the gold coating is preferably 99% by mass or more, more preferably 99.5% by mass or more, based on the total mass of the gold coating. The gold film may be formed by substitution-type gold plating, or reduction-type gold plating may be further performed after substitution-type gold plating. Further, substitutional reduction type gold plating (a substitutional type gold plating solution having a reducing agent in the plating solution, which can be thickened as compared with normal substitutional type gold plating as in the case of electroless gold plating. ) May be adopted. It is preferable that the gold plating film diffuses and disappears in the palladium film in the connection structure forming process and becomes a constituent component of the intermetallic compound layer together with palladium and the like. Therefore, the thickness of the gold coating is preferably 3 nm or more and 100 nm or less, more preferably 3 nm or more and 50 nm or less, and further preferably 3 nm or more and 20 nm or less.

  The content of each element in the nickel coating, palladium coating, and gold coating is observed at a magnification of 250,000 times using a transmission electron microscope (TEM) after cutting out the cross section of each coating by, for example, the ultramicrotome method. , And component analysis by energy dispersive X-ray analysis (EDX) attached to TEM.

<Nickel-containing layer>
The nickel-containing layer 31 is formed from the nickel coating when the metal paste is sintered.

  The thickness of the nickel-containing layer 31 is preferably 0.2 μm or more and 20 μm or less, more preferably 0.3 μm or more and 15 μm or less, and further preferably 0.5 μm or more and 10 μm or less. It becomes easy to obtain favorable connection reliability because the thickness is 0.2 μm or more.

  The nickel content in the nickel-containing layer 31 is preferably 80% by mass or more based on the total mass of the nickel-containing layer 31. From the viewpoint of further improving the reliability, the content is more preferably 85% by mass or more, and further preferably 90% by mass or more. The nickel-containing layer 31 preferably contains at least one selected from the group consisting of phosphorus and boron, and more preferably contains phosphorus. Thus, there is no change in the nickel content between the nickel coating (before sintering) and the nickel-containing layer (after sintering).

<Palladium-containing layer>
The palladium layer 33 which is one of the palladium-containing layers shown in FIG. 1 is formed from the palladium coating when the metal paste is sintered.

  The thickness of the palladium layer 33 is preferably more than 0 nm and not more than 0.5 μm, more preferably not less than 15 nm and not more than 0.5 μm, and further preferably not less than 15 nm and not more than 70 nm. By setting the thickness to more than 0 nm, the bonding interface is stably formed, and the reliability is easily maintained. Moreover, the upper limit of the thickness is set to 0.5 μm for an economical reason.

  The palladium content in the palladium layer 33 is preferably 96% by mass or more and 100% by mass or less based on the total mass of the palladium layer 33.

  As shown in FIG. 1, the intermetallic compound layer 34 containing palladium and a metal element contained in the sintered metal layer, which is one of the palladium-containing layers, is derived from the palladium coating and the metal paste when the metal paste is sintered. It is formed. In addition, although mentioned later, copper, silver, etc. are mentioned as a main metal element contained in a sintered metal layer.

  The thickness of the intermetallic compound layer 34 is preferably 0.1 nm or more and 50 nm or less, more preferably 0.3 nm or more and 30 nm or less, and further preferably 0.5 nm or more and 20 nm or less. When the thickness is 0.1 nm or more, it becomes easy to function as a layer for firmly connecting the palladium layer 33 and the sintered metal layer 32. On the other hand, by setting the thickness to 50 nm or less, it is possible to prevent a decrease in connection strength in the hard and brittle intermetallic compound layer 34 and to maintain a strong connection.

  The palladium content in the intermetallic compound layer 34 is preferably 50% by mass or more, and more preferably 60% by mass or more, based on the total mass of the intermetallic compound layer 34. The upper limit of the content can be 80% by mass. When the palladium content in the intermetallic compound layer 34 is in the above range, it is easy to form an intermetallic compound having high connection strength.

  As shown in FIG. 2, an intermetallic compound layer 35 containing a metal element contained in palladium, nickel, and a sintered metal layer, which is one of the palladium-containing layers, has a palladium coating, a nickel coating, and a metal when the metal paste is sintered. Formed from the paste.

  The thickness of the intermetallic compound layer 35 is preferably 0.1 nm or more and 50 nm or less, more preferably 0.3 nm or more and 25 nm or less, and further preferably 0.5 nm or more and 20 nm or less. When the thickness is 0.1 nm or more, it becomes easy to function as a layer for firmly connecting the nickel-containing layer 31 and the sintered metal layer 32. On the other hand, by setting the thickness to 30 nm or less, it is possible to prevent a decrease in connection strength in the hard and brittle intermetallic compound layer 34 and to maintain a strong connection.

  The palladium content in the intermetallic compound layer 35 is preferably 50% by mass or more based on the total mass of the intermetallic compound layer 35. The upper limit of the content can be 80% by mass. When the palladium content in the intermetallic compound layer 35 is in the above range, it is easy to form an intermetallic compound having high connection strength.

  The nickel content in the intermetallic compound layer 35 is preferably 50% by mass or less, and more preferably 30% by mass or less, based on the total mass of the intermetallic compound layer 35. The minimum of the said content can be 20 mass%. When the nickel content in the intermetallic compound layer 35 is in the above range, it is easy to form an intermetallic compound having high connection strength.

  The content of each element in the nickel-containing layer and the palladium-containing layer is, for example, cut out the cross section of each layer by the ultramicrotome method, then observed at a magnification of 250,000 times using a TEM, and component analysis by EDX attached to the TEM Can be obtained.

<Sintered metal layer>
The sintered metal layer is formed by sintering a metal paste. The sintered metal layer contains copper or silver as a main metal element. Hereinafter, the case where copper is mainly contained is demonstrated.

  The sintered metal layer can include a structure derived from flaky copper particles oriented substantially parallel to the connection interface of the connection structure. FIG. 3 is a schematic cross-sectional view showing an example of a typical morphology of the sintered metal layer in the connection structure of the present embodiment. The sintered metal layer shown in FIG. 3 includes sintered copper 1 having a structure derived from flaky copper particles (hereinafter sometimes referred to as “flaked structure”) and sintered derived from other copper particles. Including copper 2 and holes 3.

  The sintered metal layer may have a copper element ratio of 95% by mass or more, 97% by mass or more, or 98% by mass or more in the elements excluding light elements among the constituent elements. It may be 100% by mass. If the ratio of the copper element in the sintered metal layer is within the above range, the formation of intermetallic compounds or the precipitation of different elements on the metal copper crystal grain boundary can be suppressed, and the properties of the metallic copper constituting the sintered metal layer It is easy to become strong, and it is easy to obtain even better connection reliability.

  Sintered copper having a flaky structure can be formed by sintering a copper paste containing flaky copper particles. The flake shape includes plate shapes such as plate shapes and scale shapes. As a flaky structure, the ratio of the major axis to the thickness may be 5 or more. The number average diameter of the major axis of the flaky structure may be 2 μm or more, 3 μm or more, or 4 μm or more. If the shape of the flaky structure is within this range, the reinforcing effect by the flaky structure contained in the sintered metal layer is improved, and the connection structure is further improved in terms of joint strength and connection reliability.

  The major axis and thickness of the flaky structure can be determined from, for example, an SEM image of the connection structure. Below, the method of measuring the long diameter and thickness of flake-like structure from a SEM image is illustrated. Pour the connection structure with epoxy casting resin so that the entire sample is filled and cure. The cast sample is cut in the vicinity of the cross section to be observed, the cross section is cut by polishing, and CP (cross section polisher) processing is performed. The cross section of the sample is observed with a SEM apparatus at a magnification of 5000 times. A cross-sectional image of the connection structure (for example, 5000 times) is acquired, and it is a dense continuous part, which is a linear, rectangular parallelepiped, or ellipsoidal part, and the maximum length among the straight lines included in this part The longest diameter is 1 μm or more and the ratio of the longest diameter / thickness is 4 or more. A certain thing is regarded as a flaky structure, and the major axis and thickness of the flaky structure can be measured by image processing software having a length measuring function. These average values can be obtained by calculating the number average of 20 or more points selected at random.

  The copper content (volume ratio) in the sintered metal layer can be 65% by volume or more based on the total volume of the sintered metal layer. If the copper content in the sintered metal layer is within the above range, it is possible to suppress the formation of large pores in the sintered metal layer or the sparse sintered copper that connects the flake-like structures. . Therefore, if the copper content in the sintered metal layer is within the above range, sufficient thermal conductivity is obtained and the bonding strength between the member and the sintered metal layer is improved, and the connection structure is excellent in connection reliability. It will be a thing. The copper content in the sintered metal layer may be 67% by volume or more, or 70% by volume or more, based on the volume of the sintered metal layer. The content of copper in the sintered metal layer may be 90% by volume or less from the viewpoint of ease of the manufacturing process, based on the volume of the sintered metal layer.

When the composition of the material constituting the sintered metal layer is known, for example, the copper content in the sintered metal layer can be determined by the following procedure. First, cut the sintered metal layer into a rectangular parallelepiped, measure the vertical and horizontal length of the sintered metal layer with a caliper or external shape measuring device, and measure the volume of the sintered metal layer with a film thickness meter. calculate. The apparent density M ₁ (g / cm ³ ) is determined from the volume of the cut sintered metal layer and the weight of the sintered metal layer measured with a precision balance. Using the obtained M ₁ and the copper density of 8.96 g / cm ³ , the copper content (volume%) in the sintered metal layer is obtained from the following formula (X).
Copper content (volume%) in the sintered metal layer = [(M ₁ ) /8.96] × 100 (X)

<Connection structure>
The connection structure of this embodiment includes the connection structure described above between the first member and the second member.

  FIG. 4 is a schematic cross-sectional view illustrating an example of the connection structure according to the present embodiment. The connection structure 100 shown in FIG. 4 includes a sintered metal layer 4 that joins the first member 5 and the second member 6 between the first member 5 and the second member 6. . At least one of the surfaces 7a and 7b where the first member 5 and the second member 6 are in contact with the sintered metal layer 4, respectively, a nickel-containing layer and a palladium-containing layer derived from the nickel coating and the palladium coating are provided. It is formed (not shown because it is very thin), and the above connection structure appears in this region.

  Examples of the first member 5 and the second member 6 that can have (can be provided with) these metal films on the surface in contact with the metal paste include an IGBT, a diode, a Schottky barrier diode, a MOS-FET, Thyristors, logic, sensors, analog integrated circuits, LEDs, semiconductor lasers, semiconductor elements such as transmitters, lead frames, ceramic substrates with a metal plate (for example, DBC), substrates for mounting semiconductor elements such as LED packages, copper ribbons and metals Examples include metal wiring such as a frame, a block body such as a metal block, a power supply member such as a terminal, a heat sink, and a water-cooled plate. When one member of the first member 5 and the second member 6 is a semiconductor element, the other member is a metal wiring such as a metal frame, a block body having thermal conductivity and conductivity such as a metal block, or the like. There may be. In this case, after obtaining a connection structure having one member on one surface of the sintered metal layer, the other surface of the sintered metal layer is made of gold, copper, palladium-coated copper, aluminum, or the like as the other member. Wires, ribbon wires such as aluminum and solder, and clip-shaped metal materials can be connected. Further, the other surface of the sintered metal layer and the other member can be joined via solder.

  The bonding strength of the connection structure may be 10 MPa or more, 15 MPa or more, 20 MPa or more, or 30 MPa or more. The bonding strength can be measured using a universal bond tester (4000 series, manufactured by DAGE) or the like.

  In the connection structure of the present embodiment (for example, an electronic device), even when thermal stress generated by the difference in thermal expansion coefficient between bonded members is applied to the sintered metal layer (bonding layer), high connection reliability is achieved. Can be maintained. In addition, since this sintered metal layer is sufficiently composed of metallic copper connected by metal bonds, it exhibits high thermal conductivity and allows rapid heat dissipation when mounting electronic devices that generate large amounts of heat. is there. Furthermore, since the sintered metal layer is firmly bonded by metal bonding, it can exhibit excellent bonding strength with respect to the member having the metal coating described above. As described above, the connection structure of the present embodiment has a very effective property for joining electronic devices that generate a large amount of heat, such as power devices, logic, and amplifiers. A connection structure to which this is applied can tolerate higher input power and can be operated at a higher operating temperature.

<Method for manufacturing connection structure>
The connection structure of this embodiment can be manufactured by the following method, for example. As a manufacturing method of the connection structure, a laminated body in which the first member, the metal paste for bonding, and the second member are laminated in this order on the side in which the weight of the first member works is prepared. There is a method including a step of sintering the metal paste in a state of receiving the weight of the first member or the weight of the first member and a pressure of 0.01 MPa or less.

  The laminate can be prepared, for example, by providing a metal paste on a necessary portion of the second member and then placing the first member on the metal paste.

  As a method of providing the metal paste on a necessary portion of the second member, any method can be used as long as the metal paste can be deposited. Examples of such methods include screen printing, transfer printing, offset printing, jet printing, dispenser, jet dispenser, needle dispenser, comma coater, slit coater, die coater, gravure coater, slit coat, letterpress printing, intaglio printing, gravure printing. Printing, stencil printing, soft lithography, bar coating, applicator, particle deposition method, spray coater, spin coater, dip coater, electrodeposition coating, and the like can be used. The thickness of the metal paste may be 1 μm or more and 1000 μm or less, 10 μm or more and 500 μm or less, 50 μm or more and 200 μm or less, 10 μm or more and 3000 μm or less, or 15 μm or more and 500 μm or less. It may be 20 μm or more and 300 μm or less, 5 μm or more and 500 μm or less, 10 μm or more and 250 μm or less, or 15 μm or more and 150 μm or less.

  The metal paste provided on the second member may be appropriately dried from the viewpoint of suppressing flow during sintering and generation of voids. The gas atmosphere at the time of drying may be air, an oxygen-free atmosphere such as nitrogen or a rare gas, or a reducing atmosphere such as hydrogen or formic acid. The drying method may be drying at room temperature, drying by heating, or drying under reduced pressure. For heat drying or reduced pressure drying, for example, hot plate, hot air dryer, hot air heating furnace, nitrogen dryer, infrared dryer, infrared heating furnace, far infrared heating furnace, microwave heating device, laser heating device, electromagnetic A heating device, a heater heating device, a steam heating furnace, a hot plate press device, or the like can be used. The drying temperature and time may be appropriately adjusted according to the type and amount of the dispersion medium used. As the drying temperature and time, for example, the drying may be performed at 50 ° C. or higher and 180 ° C. or lower for 1 minute or longer and 120 minutes or shorter.

  Examples of the method for arranging the first member on the metal paste include a chip mounter, a flip chip bonder, a carbon or ceramic positioning jig. In addition, you may perform the above-mentioned drying process after the process of arrange | positioning a 1st member.

  A metal paste is sintered by heat-processing a laminated body. Sintering is performed by heat treatment. For the heat treatment, for example, hot plate, hot air dryer, hot air heating furnace, nitrogen dryer, infrared dryer, infrared heating furnace, far infrared heating furnace, microwave heating device, laser heating device, electromagnetic heating device, A heater heating device, a steam heating furnace, or the like can be used.

  The gas atmosphere at the time of sintering may be an oxygen-free atmosphere from the viewpoint of suppressing oxidation of the sintered body, the first member, and the second member. The gas atmosphere at the time of sintering may be a reducing atmosphere from the viewpoint of removing the surface oxides of the metal particles contained in the metal paste. Examples of the oxygen-free atmosphere include introduction of oxygen-free gas such as nitrogen and rare gas, or under vacuum. Examples of the reducing atmosphere include pure hydrogen gas, hydrogen and nitrogen mixed gas typified by forming gas, nitrogen containing formic acid gas, hydrogen and rare gas mixed gas, and rare gas containing formic acid gas. Can be mentioned.

  The maximum temperature reached during the heat treatment may be 250 ° C. or higher and 450 ° C. or lower, from the viewpoint of reducing thermal damage to the first member and the second member and improving the yield, and may be 250 ° C. or higher and 400 ° C. or lower. Or 250 ° C. or more and 350 ° C. or less. If the ultimate temperature is 200 ° C. or higher, the sintering tends to proceed sufficiently when the ultimate temperature holding time is 60 minutes or less.

  The ultimate temperature holding time may be from 1 minute to 60 minutes, from 1 minute to less than 40 minutes, or from 1 minute to 1 minute from the viewpoint of volatilizing all the dispersion medium and improving the yield. It may be less than 30 minutes.

  The pressure at the time of joining can be suitably set according to the metal content in the sintered body. For example, in the case of obtaining a sintered metal layer comprising sintered copper having the above flake-like structure, the copper content (volume ratio) in the sintered body is 65% by volume or more based on the sintered body. It can be set as conditions. However, by using the copper paste described later, the sintered metal layer (sintered copper) of the present embodiment described above can be formed without applying pressure when the laminate is sintered. In this case, sufficient bonding strength is obtained in a state where only the own weight of the first member laminated on the copper paste, or in addition to the first member's own weight, a pressure of 0.01 MPa or less, preferably 0.005 MPa or less. Can be obtained. If the pressure received during the sintering is within the above range, a special pressurizing device is not required, so that the void reduction, bonding strength and connection reliability can be further improved without impairing the yield. Examples of the method for receiving a pressure of 0.01 MPa or less for the copper paste include a method of placing a weight on the first member.

(Metal paste)
The metal paste used in the method for manufacturing a connection structure includes metal particles and a dispersion medium. A metal particle contains a copper particle or a silver particle as a main component. Hereinafter, a case where copper particles are mainly contained (copper paste) will be described.

(Metal particles)
Examples of the metal particles include sub-micro copper particles, flaky micro-copper particles, copper particles other than these, and other metal particles.

(Sub-micro copper particles)
Examples of the sub-micro copper particles include those containing copper particles having a particle size of 0.12 μm or more and 0.8 μm or less. For example, copper particles having a volume average particle size of 0.12 μm or more and 0.8 μm or less are used. it can. If the volume average particle diameter of the sub-micro copper particles is 0.12 μm or more, effects such as suppression of the synthesis cost of the sub-micro copper particles, good dispersibility, and suppression of the use amount of the surface treatment agent are easily obtained. If the volume average particle diameter of the sub-micro copper particles is 0.8 μm or less, the effect that the sinterability of the sub-micro copper particles is excellent is easily obtained. From the viewpoint of further exerting the above effects, the volume average particle diameter of the sub-micro copper particles may be 0.15 μm or more and 0.8 μm or less, 0.15 μm or more and 0.6 μm or less, and 0 It may be not less than 2 μm and not more than 0.5 μm, and may be not less than 0.3 μm and not more than 0.45 μm.

  In the present specification, the volume average particle diameter means 50% volume average particle diameter. When determining the volume average particle size of the copper particles, light scattering method particle size distribution measurement of the raw copper particles or dried copper particles from which volatile components have been removed from the copper paste, dispersed in a dispersion medium using a dispersant It can be determined by a method of measuring with an apparatus (for example, a Shimadzu nanoparticle size distribution measuring apparatus (SALD-7500 nano, manufactured by Shimadzu Corporation)). When using a light scattering particle size distribution analyzer, hexane, toluene, α-terpineol, or the like can be used as the dispersion medium.

  The sub-micro copper particles can contain 10% by mass or more of copper particles having a particle size of 0.12 μm or more and 0.8 μm or less. From the viewpoint of the sinterability of the copper paste, the sub-micro copper particles can contain 20% by mass or more, 30% by mass or more of copper particles having a particle size of 0.12 μm or more and 0.8 μm or less, It can be contained by mass%. When the content ratio of the copper particles having a particle size of 0.12 μm or more and 0.8 μm or less in the sub-micro copper particles is 20% by mass or more, the dispersibility of the copper particles is further improved, the viscosity is increased, and the paste concentration is decreased. It can be suppressed more.

  The particle size of the copper particles can be determined by the following method. The particle size of the copper particles can be calculated from an SEM image, for example. The copper particle powder is placed on a carbon tape for SEM with a spatula to obtain a sample for SEM. The sample for SEM is observed with a SEM apparatus at a magnification of 5000 times. A quadrilateral circumscribing the copper particles of this SEM image is drawn by image processing software, and one side thereof is set as the particle size of the particles.

  The content of the sub-micro copper particles may be 20% by mass or more and 90% by mass or less, 30% by mass or more and 90% by mass or less, or 35% by mass or more based on the total mass of the metal particles. 85 mass% or less may be sufficient, and 40 mass% or more and 80 mass% or less may be sufficient. When the content of the sub-micro copper particles is within the above range, the above-described sintered metal layer can be easily formed.

  The content of the sub-micro copper particles may be 20% by mass or more and 90% by mass or less based on the total mass of the sub-micro copper particles and the mass of the flaky micro-copper particles. When the content of the sub-micro copper particles is 20% by mass or more, the space between the flaky micro-copper particles can be sufficiently filled, and the above-described sintered metal layer can be easily formed. If the content of the sub-micro copper particles is 90% by mass or less, volume shrinkage when the copper paste is sintered can be sufficiently suppressed, so that the above-described sintered metal layer can be easily formed. From the viewpoint of further exerting the above effect, the content of the sub-micro copper particles is 30% by mass or more and 85% by mass or less based on the total mass of the sub-micro copper particles and the mass of the flaky micro-copper particles. It may be 35 mass% or more and 85 mass% or less, and may be 40 mass% or more and 80 mass% or less.

  The shape of the sub-micro copper particles is not particularly limited. Examples of the shape of the sub-micro copper particles include a spherical shape, a lump shape, a needle shape, a flake shape, a substantially spherical shape, and an aggregate thereof. From the viewpoints of dispersibility and filling properties, the shape of the sub-micro copper particles may be spherical, substantially spherical, or flaky. From the viewpoint of combustibility, dispersibility, miscibility with flaky microparticles, etc., Alternatively, it may be approximately spherical.

  The sub-micro copper particles may have an aspect ratio of 5 or less or 3 or less from the viewpoints of dispersibility, filling properties, and miscibility with the flaky micro particles. In this specification, “aspect ratio” indicates the long side / thickness of a particle. The measurement of the long side and thickness of the particle can be obtained, for example, from the SEM image of the particle.

  The sub-micro copper particles may be treated with a specific surface treatment agent. Examples of the specific surface treatment agent include organic acids having 8 to 16 carbon atoms. Examples of the organic acid having 8 to 16 carbon atoms include caprylic acid, methylheptanoic acid, ethylhexanoic acid, propylpentanoic acid, pelargonic acid, methyloctanoic acid, ethylheptanoic acid, propylhexanoic acid, capric acid, methylnonanoic acid, and ethyl. Octanoic acid, propylheptanoic acid, butylhexanoic acid, undecanoic acid, methyldecanoic acid, ethylnonanoic acid, propyloctanoic acid, butylheptanoic acid, lauric acid, methylundecanoic acid, ethyldecanoic acid, propylnonanoic acid, butyloctanoic acid, pentylheptanoic acid, Tridecanoic acid, methyldodecanoic acid, ethylundecanoic acid, propyldecanoic acid, butylnonanoic acid, pentyloctanoic acid, myristic acid, methyltridecanoic acid, ethyldodecanoic acid, propylundecanoic acid, butyldecanoic acid, pentylnonanoic acid, Siloctanoic acid, pentadecanoic acid, methyltetradecanoic acid, ethyltridecanoic acid, propyldodecanoic acid, butylundecanoic acid, pentyldecanoic acid, hexylnonanoic acid, palmitic acid, methylpentadecanoic acid, ethyltetradecanoic acid, propyltridecanoic acid, butyldodecanoic acid, Pentylundecanoic acid, hexyldecanoic acid, heptylnonanoic acid, methylcyclohexanecarboxylic acid, ethylcyclohexanecarboxylic acid, propylcyclohexanecarboxylic acid, butylcyclohexanecarboxylic acid, pentylcyclohexanecarboxylic acid, hexylcyclohexanecarboxylic acid, heptylcyclohexanecarboxylic acid, octylcyclohexanecarboxylic acid, Saturated fatty acids such as nonylcyclohexanecarboxylic acid; octenoic acid, nonenoic acid, methylnonenoic acid, decenoic acid Unsaturated fatty acids such as undecenoic acid, dodecenoic acid, tridecenoic acid, tetradecenoic acid, myristoleic acid, pentadecenoic acid, hexadecenoic acid, palmitoleic acid, sabienoic acid; terephthalic acid, pyromellitic acid, o-phenoxybenzoic acid, methylbenzoic acid, Examples include aromatic carboxylic acids such as ethyl benzoic acid, propyl benzoic acid, butyl benzoic acid, pentyl benzoic acid, hexyl benzoic acid, heptyl benzoic acid, octyl benzoic acid, and nonyl benzoic acid. An organic acid may be used individually by 1 type, and may be used in combination of 2 or more type. By combining such an organic acid and the sub-micro copper particles, the dispersibility of the sub-micro copper particles and the detachability of the organic acid during sintering tend to be compatible.

The treatment amount of the surface treatment agent may be an amount that adheres to a monomolecular layer to a trimolecular layer on the surface of the sub-micro copper particles. This amount includes the number of molecular layers (n) attached to the surface of the sub-micro copper particles, the specific surface area (A _p ) (unit m ² / g) of the sub-micro copper particles, and the molecular weight (M _s ) of the surface treatment agent (M _s ) It can be calculated from the unit g / mol), the minimum covering area (S _S ) (unit m ² / piece) of the surface treatment agent, and the Avogadro number (N _A ) (6.02 × 10 ²³ pieces). Specifically, the treatment amount of the surface treatment agent is the treatment amount of the surface treatment agent (% by mass) = {(n · A _p · M _s ) / (S _S · N _A + n · A _p · M _s )} * Calculated according to the equation of 100%.

The specific surface area of the sub-micro copper particles can be calculated by measuring the dried sub-micro copper particles by the BET specific surface area measurement method. Minimum coverage of the surface treatment agent, if the surface treatment agent is a straight-chain saturated fatty acids, is 2.05 × ¹⁰ -19 ^m 2/1 molecule. In the case of other surface treatment agents, for example, calculation from a molecular model, or “Chemistry and Education” (Akihiro Ueda, Juno Inafuku, Mori Kaoru, 40 (2), 1992, p114-117) It can be measured by the method described. An example of the method for quantifying the surface treatment agent is shown. The surface treatment agent can be identified by a thermal desorption gas / gas chromatograph mass spectrometer of a dry powder obtained by removing the dispersion medium from the copper paste, whereby the carbon number and molecular weight of the surface treatment agent can be determined. The carbon content of the surface treatment agent can be analyzed by carbon content analysis. Examples of the carbon analysis method include a high frequency induction furnace combustion / infrared absorption method. The amount of the surface treatment agent can be calculated by the above formula from the carbon number, molecular weight, and carbon content ratio of the identified surface treatment agent.

  The treatment amount of the surface treatment agent may be 0.07% by mass or more and 2.1% by mass or less, may be 0.10% by mass or more and 1.6% by mass or less, and may be 0.2% by mass. It may be 1.1% by mass or less.

  As the sub-micro copper particles, commercially available ones can be used. Commercially available sub-micro copper particles include, for example, CH-0200 (Mitsui Metal Mining Co., Ltd., volume average particle size 0.36 μm), HT-14 (Mitsui Metal Mining Co., Ltd., volume average particle size 0. 41 μm), CT-500 (manufactured by Mitsui Kinzoku Mining Co., Ltd., volume average particle size 0.72 μm), and Tn—Cu100 (manufactured by Taiyo Nippon Sanso Corporation, volume average particle size 0.12 μm).

(Flake micro copper particles)
Examples of the flaky micro copper particles include those containing copper particles having a maximum diameter of 1 μm or more and 20 μm or less and an aspect ratio of 4 or more. For example, the average maximum diameter is 1 μm or more and 20 μm or less, and the aspect ratio is 4 The above copper particles can be used. If the average maximum diameter and aspect ratio of the flaky micro copper particles are within the above ranges, volume shrinkage when the copper paste is sintered can be sufficiently reduced, and the above-described sintered metal layer can be easily formed. . From the standpoint of further achieving the above effects, the average maximum diameter of the flaky micro copper particles may be 1 μm or more and 10 μm or less, or 3 μm or more and 10 μm or less. The measurement of the maximum diameter and the average maximum diameter of the flaky micro-copper particles can be obtained from, for example, an SEM image of the particles, and is obtained as a major axis X and an average value Xav of the major axis described later.

  The flaky micro copper particles can contain 50% by mass or more of copper particles having a maximum diameter of 1 μm or more and 20 μm or less. From the viewpoint of orientation in the sintered metal layer, reinforcing effect, and filling property of the bonding paste, the flaky micro copper particles can contain 70% by mass or more of copper particles having a maximum diameter of 1 μm or more and 20 μm or less, and 80% by mass. % Or more, and 100 mass% can be contained. From the viewpoint of suppressing poor bonding, the flaky micro-copper particles preferably do not contain particles having a size exceeding the bonding thickness, such as particles having a maximum diameter exceeding 20 μm.

  A method of calculating the major axis X of the flaky micro copper particles from the SEM image is illustrated. The powder of flaky micro copper particles is placed on a carbon tape for SEM with a spatula to obtain a sample for SEM. The sample for SEM is observed with a SEM apparatus at a magnification of 5000 times. A rectangle circumscribing the flaky micro copper particles of the SEM image is drawn by image processing software, and the long side of the rectangle is defined as the major axis X of the particles. This measurement is performed on 50 or more flaky micro copper particles using a plurality of SEM images, and the average value Xav of the major axis is calculated.

  The aspect ratio of the flaky micro copper particles may be 4 or more, or 6 or more. If the aspect ratio is within the above range, the flaky micro-copper particles in the copper paste are oriented substantially parallel to the bonding surface, so that volume shrinkage when the copper paste is sintered can be suppressed. It becomes easy to form the sintered metal layer.

  The content of the flaky micro copper particles may be 1% by mass or more and 90% by mass or less, 10% by mass or more and 70% by mass or less, and 20% by mass based on the total mass of the metal particles. It may be 50% by mass or more. When the content of the flaky micro copper particles is within the above range, the above-described sintered metal layer can be easily formed.

  The total content of the sub-micro copper particles and the content of the flaky micro-copper particles may be 80% by mass or more based on the total mass of the metal particles. If the sum of the content of the sub-micro copper particles and the content of the flaky micro-copper particles is within the above range, the above-described sintered metal layer can be easily formed. From the viewpoint of further exerting the above effects, the total content of the sub-micro copper particles and the content of the flaky micro-copper particles may be 90% by mass or more based on the total mass of the metal particles. The mass may be 100% by mass or more.

  In the flaky micro copper particles, the presence or absence of the treatment with the surface treatment agent is not particularly limited. From the viewpoint of dispersion stability and oxidation resistance, the flaky micro copper particles may be treated with a surface treatment agent. The surface treatment agent may be removed during sintering. Examples of such a surface treatment agent include aliphatic carboxylic acids such as palmitic acid, stearic acid, arachidic acid, and oleic acid; aromatic carboxylic acids such as terephthalic acid, pyromellitic acid, and o-phenoxybenzoic acid; cetyl alcohol Aliphatic alcohols such as stearyl alcohol, isobornylcyclohexanol and tetraethylene glycol; aromatic alcohols such as p-phenylphenol; alkylamines such as octylamine, dodecylamine and stearylamine; stearonitrile and decanenitrile Examples include aliphatic nitriles; silane coupling agents such as alkylalkoxysilanes; polymer treatment materials such as polyethylene glycol, polyvinyl alcohol, polyvinylpyrrolidone, and silicone oligomers. A surface treating agent may be used individually by 1 type, and may be used in combination of 2 or more type.

  The treatment amount of the surface treatment agent may be a monomolecular layer or more on the particle surface. The treatment amount of such a surface treatment agent varies depending on the specific surface area of the flaky micro copper particles, the molecular weight of the surface treatment agent, and the minimum coating area of the surface treatment agent. The treatment amount of the surface treatment agent is usually 0.001% by mass or more. The specific surface area of the flaky micro copper particles, the molecular weight of the surface treatment agent, and the minimum coating area of the surface treatment agent can be calculated by the method described above.

  When preparing a copper paste only from the above-mentioned sub-micro copper particles, the volume shrinkage and sintering shrinkage accompanying the drying of the dispersion medium are large, so that the copper paste is easily peeled off from the adherend surface during the sintering of the copper paste, and bonding of semiconductor elements, etc. In this case, it is difficult to obtain sufficient die shear strength and connection reliability. By using the sub-micro copper particles and the flaky micro-copper particles in combination, volume shrinkage when the copper paste is sintered is suppressed, and the above-described sintered metal layer can be easily formed.

  In the copper paste of the present embodiment, the content of the micro copper particles having a maximum diameter of 1 μm to 20 μm and an aspect ratio of less than 2 included in the metal particles has a maximum diameter of 1 μm to 20 μm, and the aspect ratio Is preferably 50% by mass or less, more preferably 30% by mass or less, based on the total amount of flaky micro-copper particles having 4 or more. By limiting the content of micro copper particles having an average maximum diameter of 1 μm or more and 20 μm or less and an aspect ratio of less than 2, the flaky micro copper particles in the copper paste are oriented substantially parallel to the bonding surface. It becomes easy and volume shrinkage when a copper paste is sintered can be suppressed more effectively. Thereby, it becomes easy to form the sintered metal layer described above. The content of the micro copper particles having an average maximum diameter of 1 μm or more and 20 μm or less and an aspect ratio of less than 2 is such that the maximum diameter is 1 μm or more and 20 μm or less, and the aspect ratio is more easily obtained. May be 20% by mass or less, or 10% by mass or less based on the total amount of flaky micro copper particles having 4 or more.

  What is marketed can be used as flaky micro copper particle of this embodiment. Examples of commercially available flaky micro copper particles include MA-C025 (Mitsui Metal Mining Co., Ltd., average maximum diameter 4.1 μm), 3L3 (Fukuda Metal Foil Powder Co., Ltd., average maximum diameter 7.3 μm). ) 1110F (Mitsui Metal Mining Co., Ltd., average maximum diameter 5.8 μm), 2L3 (Fukuda Metal Foil Powder Co., Ltd., average maximum diameter 9 μm).

  In the copper paste of the present embodiment, the micro copper particles to be blended include flaky micro copper particles having a maximum diameter of 1 μm to 20 μm, an aspect ratio of 4 or more, and a maximum diameter of 1 μm to 20 μm. In addition, the micro copper particles having an aspect ratio of less than 2 and having a content of micro copper particles of 50% by mass or less, preferably 30% by mass or less based on the total amount of the flaky micro copper particles can be used. When commercially available flaky micro copper particles are used, the maximum diameter is 1 μm or more and 20 μm or less, the flaky micro copper particles have an aspect ratio of 4 or more, and the maximum diameter is 1 μm or more and 20 μm or less. The content of the micro copper particles having a ratio of less than 2 may be 50% by mass or less, preferably 30% by mass or less, based on the total amount of the flaky micro copper particles.

(Other metal particles other than copper particles)
As a metal particle, other metal particles other than the sub micro copper particle and micro copper particle which were mentioned above may be included, for example, particles, such as nickel, silver, gold | metal | money, palladium, platinum, may be included. The other metal particles may have a volume average particle size of 0.01 μm or more and 10 μm or less, 0.01 μm or more and 5 μm or less, or 0.05 μm or more and 3 μm or less. When other metal particles are contained, the content thereof may be less than 20% by mass or less than 10% by mass based on the total mass of the metal particles from the viewpoint of obtaining sufficient bondability. May be. Other metal particles may not be included. The shape of other metal particles is not particularly limited.

  By including metal particles other than copper particles, it is possible to obtain a sintered metal layer in which multiple types of metals are dissolved or dispersed, thereby improving the mechanical properties such as yield stress and fatigue strength of the sintered metal layer. Therefore, connection reliability is easy to improve. Further, by adding a plurality of kinds of metal particles, the formed sintered metal layer can easily improve the bonding strength and connection reliability with respect to a specific adherend.

(Dispersion medium)
The dispersion medium is not particularly limited, and may be volatile. Examples of the volatile dispersion medium include monovalent and polyvalent pentanol, hexanol, heptanol, octanol, decanol, ethylene glycol, diethylene glycol, propylene glycol, butylene glycol, α-terpineol, isobornylcyclohexanol (MTPH), and the like. Dihydric alcohols: ethylene glycol butyl ether, ethylene glycol phenyl ether, diethylene glycol methyl ether, diethylene glycol ethyl ether, diethylene glycol butyl ether, diethylene glycol isobutyl ether, diethylene glycol hexyl ether, triethylene glycol methyl ether, diethylene glycol dimethyl ether, diethylene glycol diethyl ether, diethylene glycol Butyl ether, diethylene glycol butyl methyl ether, diethylene glycol isopropyl methyl ether, triethylene glycol dimethyl ether, triethylene glycol butyl methyl ether, propylene glycol propyl ether, dipropylene glycol methyl ether, dipropylene glycol ethyl ether, dipropylene glycol propyl ether, dipropylene glycol Ethers such as butyl ether, dipropylene glycol dimethyl ether, tripropylene glycol methyl ether, tripropylene glycol dimethyl ether; ethylene glycol ethyl ether acetate, ethylene glycol butyl ether acetate, diethylene glycol ethyl ether acetate, diethylene glycol Esters such as coal butyl ether acetate, dipropylene glycol methyl ether acetate (DPMA), ethyl lactate, butyl lactate, γ-butyrolactone, propylene carbonate; N-methyl-2-pyrrolidone, N, N-dimethylacetamide, N, N- Acid amides such as dimethylformamide; Aliphatic hydrocarbons such as cyclohexane, octane, nonane, decane and undecane; Aromatic hydrocarbons such as benzene, toluene and xylene; Mercaptans having an alkyl group having 1 to 18 carbon atoms; Examples include mercaptans having 5 to 7 cycloalkyl groups. Examples of mercaptans having an alkyl group having 1 to 18 carbon atoms include ethyl mercaptan, n-propyl mercaptan, i-propyl mercaptan, n-butyl mercaptan, i-butyl mercaptan, t-butyl mercaptan, pentyl mercaptan, hexyl mercaptan. And dodecyl mercaptan. Examples of mercaptans having a cycloalkyl group having 5 to 7 carbon atoms include cyclopentyl mercaptan, cyclohexyl mercaptan, and cycloheptyl mercaptan.

  The content of the dispersion medium may be 5 to 50 parts by mass with 100 parts by mass of the total mass of the metal particles. If the content of the dispersion medium is within the above range, the metal paste can be adjusted to a more appropriate viscosity, and it is difficult to inhibit the sintering of the metal particles.

(Additive)
Where necessary, wetting additives such as nonionic surfactants and fluorosurfactants; antifoaming agents such as silicone oils; ion trapping agents such as inorganic ion exchangers and the like are appropriately added to the metal paste. Also good.

(Preparation of metal paste)
The metal paste may be prepared by mixing metal particles and optional additives in a dispersion medium. You may perform a stirring process after mixing of each component. Moreover, you may adjust the largest particle size of the metal particle contained in a dispersion liquid by classification operation.

  For example, in the case of the above-described copper paste, sub-micro copper particles, a surface treatment agent, and a dispersion medium are mixed in advance, and dispersion treatment is performed to prepare a dispersion liquid of sub-micro copper particles. Other metal particles and optional additives may be mixed to prepare. By setting it as such a procedure, the dispersibility of a sub micro copper particle improves, a mixing property with a micro copper particle improves, and the performance of a copper paste improves more. Classification may be performed on the sub-micro copper particle dispersion to remove aggregates.

<Semiconductor Device and Manufacturing Method of Semiconductor Device>
The semiconductor device of this embodiment is a semiconductor device that includes the connection structure described above between the first member and at least one of the first member and the second member is a semiconductor element.

  Examples of the semiconductor element include a power module including a diode, a rectifier, a thyristor, a MOS gate driver, a power switch, a power MOSFET, an IGBT, a Schottky diode, and a fast recovery diode, a transmitter, an amplifier, and an LED module. Non-semiconductor elements include lead frames, metal plate-attached ceramic substrates (for example, DBC), semiconductor element mounting base materials such as LED packages, copper ribbons, metal blocks, terminals and other power supply members, heat sinks, water-cooled plates Etc.

  In the semiconductor device of this embodiment, the die shear strength may be 10 MPa or more, 15 MPa or more, 20 MPa or more, and 30 MPa from the viewpoint of process compatibility and connection reliability. It may be the above. The die shear strength can be measured using a universal bond tester (4000 series, manufactured by DAGE) or the like.

  FIG. 5 is a schematic cross-sectional view showing an example of the semiconductor device of this embodiment. A semiconductor device 110 shown in FIG. 5 includes a semiconductor element 14 connected to a lead frame 11a via a sintered metal layer 4, and a mold resin 13 for molding them. The semiconductor element 14 is connected to the lead frame 11 b through the wire 12. A nickel-containing layer and a palladium-containing layer derived from the nickel coating and the palladium coating are formed on at least one of the surfaces where the lead frame 11a, the semiconductor element 14, and the sintered metal layer 4 are in contact (not shown). 1) The above connection structure appears in this region.

  Examples of the semiconductor device according to the present embodiment include a diode, a rectifier, a thyristor, a MOS gate driver, a power switch, a power MOSFET, an IGBT, a Schottky diode, a fast recovery diode, and a power module, a transmitter, an amplifier, and a high-intensity LED. Examples include modules and sensors.

  The semiconductor device of the present embodiment has high connection reliability during high temperature operation, improved heat dissipation due to high thermal conductivity, and high electrical conductivity, and operates in a high temperature environment, a semiconductor device using a high heat dissipation semiconductor element. It can be suitably used for a semiconductor device or the like.

  The semiconductor device can be manufactured in the same manner as the connection structure manufacturing method described above. That is, the first member, the metal paste, and the second member are laminated in this order on the side in which the weight of the first member acts, and at least one of the first member and the second member is a semiconductor. There is a method including a step of preparing a laminated body as an element and sintering the metal paste in a state of receiving the weight of the first member or the weight of the first member and a pressure of 0.01 MPa or less.

  When the second member is a semiconductor element, the above method can reduce damage to the semiconductor element when a metal wiring or a block body is bonded to the semiconductor element as the first member. A semiconductor device in which a member such as a metal wiring or a block body is bonded on a semiconductor element will be described below.

  As one embodiment of such a semiconductor device, a first electrode, a semiconductor element electrically connected to the first electrode, and a second electrode electrically connected to the semiconductor element via a metal wiring And an electrode having the sintered metal layer described above between the semiconductor element and the metal wiring and between the metal wiring and the second electrode.

  FIG. 6 is a schematic cross-sectional view showing an example of the semiconductor device. A semiconductor device 200 shown in FIG. 6 includes an insulating substrate 21 having a first electrode 22 and a second electrode 24, a semiconductor element 23 bonded to the first electrode 22 by a sintered metal layer 4, and a semiconductor. A metal wiring 25 that electrically connects the element 23 and the second electrode 24 is provided. By providing the nickel coating and the palladium coating on both sides or one side of the semiconductor element 23 in advance, and forming the sintered metal layer 4, the above-described connection structure appears and the reliability can be improved. At this time, the above-described nickel coating and palladium coating are provided on the first electrode 22 and the second electrode 24 in advance, and then the sintered metal layer 4 is formed, so that the above-described portion is also applied to the corresponding portion. A connection structure appears and reliability can be improved. The metal wiring 25 and the semiconductor element 23, and the metal wiring 25 and the second electrode 24 are joined by the sintered metal layer 4. The semiconductor element 23 is connected to the third electrode 26 through a wire 27. The semiconductor device 200 includes a copper plate 28 on the side opposite to the surface on which the electrodes and the like of the insulating substrate 21 are mounted. In the semiconductor device 200, the structure is sealed with an insulator 29. The semiconductor device 200 has one semiconductor element 23 on the first electrode 22, but may have two or more. In this case, the plurality of semiconductor elements 23 can be joined to the metal wiring 25 by the sintered metal layer 4.

  Hereinafter, the present invention will be described more specifically with reference to examples. However, the present invention is not limited to the following examples.

Example 1
<Surface treatment for joining objects>
(Process a: Pretreatment of plating)
A copper plate (19 mm × 25 mm × 3 mm) was immersed in a 50 ° C. degreasing solution Z-200 (trade name, manufactured by World Metal Co., Ltd.) for 3 minutes and washed with water for 2 minutes. Thereafter, the copper plate was immersed in a 100 g / l ammonium persulfate solution for 1 minute and washed with water for 2 minutes. The copper plate was immersed in 10% sulfuric acid for 1 minute and washed with water for 2 minutes. Subsequently, the copper plate was immersed for 5 minutes in SA-100 (trade name, manufactured by Hitachi Chemical Co., Ltd.), which is a plating activation treatment liquid at a liquid temperature of 25 ° C., and then washed with water for 2 minutes.

(Process b: Formation of electroless nickel plating film)
The copper plate which passed through the process a was immersed for 25 minutes in NIPS-100 (trade name, manufactured by Hitachi Chemical Co., Ltd.) which is an electroless nickel plating solution having a liquid temperature of 85 ° C., and then washed with water for 1 minute. The thickness of the formed electroless nickel plating film was 5 μm. The phosphorus concentration in the electroless nickel plating film was 7% by mass.

(Process c: Formation of electroless palladium plating film)
The copper plate plated with electroless nickel after the step b was immersed for 9 seconds in a palette (trade name, manufactured by Kojima Chemical Co., Ltd.) which is an electroless palladium plating solution having a liquid temperature of 55 ° C., and then washed with water for 1 minute. The thickness of the formed electroless palladium plating film was 3 nm (0.003 μm). The palladium concentration in the electroless palladium coating was about 100% by mass.

<Preparation of copper paste>
(Process d: Preparation of copper paste)
Α-Terpineol (manufactured by Wako Pure Chemical Industries, Ltd.) 5.2 g and isobornylcyclohexanol (MTPH, manufactured by Nippon Terpene Chemical Co., Ltd.) 6.8 g as a dispersion medium, and CH0200 (Mitsui Metal Mining Co., Ltd.) as sub-micro copper particles 52.8 g of 0.12 μm or more and 0.8 μm or less copper particles (made by company) was mixed in a plastic bottle and 19.6 kHz by an ultrasonic homogenizer (US-600, manufactured by Nippon Seiki Co., Ltd.). , 600 W for 1 minute to obtain a dispersion. To this dispersion, 35.2 g of MA-C025 (manufactured by Mitsui Mining & Smelting Co., Ltd., maximum content of copper particles with a maximum diameter of 1 μm to 20 μm) 100 fl. Stir until no more. The plastic bottle was tightly sealed and stirred for 2 minutes at 2000 rpm using a rotating and rotating stirrer (Planetary Vacuum Mixer ARV-310, manufactured by Sinky Co., Ltd.), and stirred for 2 minutes at 2000 rpm under reduced pressure to obtain a copper paste. .

<Production of connection structure>
(Process e: Preparation of connection structure)
On a copper plate (19 mm × 25 mm × 3 mm) subjected to electroless nickel / electroless palladium plating treatment obtained by performing steps (a) to (c), a square opening ( A metal mask having 3 rows and 3 columns (3 mm × 3 mm) was placed, and the copper paste obtained in the step (d) was applied by stencil printing using a metal squeegee. A silicon chip (chip thickness: 600 μm) having a 3 mm × 3 mm deposition surface made of nickel (formed by sputtering) was placed on the applied copper paste, and lightly pressed with tweezers. This was set in a tube furnace (manufactured by Ave Sea Co., Ltd.), and argon gas was flowed at 1 L / min to replace the air with argon gas. Thereafter, the copper plate and the silicon chip subjected to the electroless nickel / electroless palladium plating treatment by sintering at a temperature of 10 minutes and 350 ° C. for 10 minutes while flowing hydrogen gas at 300 mL / min are sintered metal layers. The connection structure joined by the above was obtained. Thereafter, the argon gas was cooled to 0.3 L / min and cooled, and the connection structure was taken out into the air at 50 ° C. or lower.

<Evaluation of connection structure>
(1) Measurement of die shear strength The joint strength of the connection structure was evaluated by die shear strength. Using a universal bond tester (4000 series, manufactured by DAGE) equipped with a 1 kN load cell, the silicon chip was pushed in the horizontal direction at a measurement speed of 500 μm / s and a measurement height of 100 μm to measure the die shear strength of the connection structure. The average value of the measured values for the eight connected structures was defined as the die shear strength. The case where the die shear strength was 30 MPa or more was A, and the case where the die shear strength was lower than 30 MPa was B. The results are shown in Table 2.

(2) Temperature cycle connection reliability test In the same manner as "(1) Die shear strength", a copper plate (19 mm x 25 mm x 3 mm) and a silicon chip with a deposition surface of nickel (4 mm x 8 mm, chip thickness: 600 µm) And a bonded structure joined with a sintered metal layer. An adhesion improver (HIMAL, manufactured by Hitachi Chemical Co., Ltd.) is applied and dried on the connection structure, and then sealed with a solid sealing material (CEL, manufactured by Hitachi Chemical Co., Ltd.), and a temperature cycle test piece is attached. Obtained. This test piece for temperature cycle is set in a temperature cycle tester (TSA-72SE-W, manufactured by Espec Corp.), low temperature side: −40 ° C., 15 minutes, room temperature: 2 minutes, high temperature side: 200 ° C., 15 minutes. The temperature cycle connection reliability test was performed under the conditions of defrosting cycle: automatic and cycle number: 1000 cycles. Using an ultrasonic flaw detector (Insight-300, manufactured by Insight Co., Ltd.), a SAT image of the bonding state of the interface between the sintered metal layer and the substrate or chip before and after the temperature cycle connection reliability test is obtained, and whether or not peeling occurs I investigated. The case where the peeled area of the joint was less than 20 area% was designated as A, and the case where it was 20 area% or more was designated as B. The results are shown in Table 2.

(3) Confirmation of presence / absence of electroless palladium plating film Presence / absence of electroless palladium plating film (palladium layer) is determined by cutting a cross section that can contain a film by an ultramicrotome method, and then transmitting electron microscope apparatus (TEM, The product was observed at a magnification of 250,000 times using a product name “JEM-2100F” manufactured by JEOL Ltd., and confirmed by component analysis using EDX attached to the TEM. The results are shown in Table 2. In the table, the fact that the thickness of the palladium layer was 0 μm indicates that the electroless palladium plating film did not remain.

(4) Analysis of the sintered metal layer-Copper content (volume ratio) in the sintered metal layer
An opening of 15 mm × 15 mm was provided in a 1 mm thick Teflon (registered trademark) plate. This Teflon (registered trademark) plate was placed on a glass plate, the opening was filled with copper paste, and the copper paste overflowing from the opening was removed with a metal squeegee. The Teflon (registered trademark) plate was removed, set in a tube furnace, and heated to 150 ° C. and kept for 1 hour while flowing argon gas at 0.3 L / min to remove the solvent. As it was, the gas was changed to 300 mL / min of hydrogen gas, the temperature was raised to 350 ° C., and sintering treatment was performed for 60 minutes to obtain a sintered metal layer. Thereafter, the argon gas was changed to 0.3 L / min and cooled, and the sintered metal layer was taken out into the air at 50 ° C. or lower. The plate-like sintered metal layer was peeled from the glass plate and polished with sandpaper (# 800) to obtain a plate-like sample having a size of 10 mm × 10 mm and a flat surface. The length, width, and thickness of the plate sample were measured, and the weight of the plate sample was measured. The density of the plate-like sample was calculated from these values, and the volume ratio of metallic copper was calculated according to the following formula.
Copper content (volume%) in sintered metal layer = density of plate-like sample (g / cm ³ ) /8.96 (g / cm ³ ) × 100 (%)

-Cross-section morphology observation The connection structure is fixed in the cup with a sample clip (Sampklip I, manufactured by Buehler), and an epoxy casting resin (Epomount, manufactured by Refinetech Co., Ltd.) is poured around until the entire sample is filled, and a vacuum desiccator The solution was allowed to stand inside and degassed by reducing the pressure for 1 minute. Then, it left still at room temperature for 10 hours, the epoxy casting resin was hardened, and the sample was prepared. The sample was cut in the vicinity of the silicon chip using Refine So Excel (Refine Tech Co., Ltd.). Using a polishing apparatus (Refine Polisher HV, manufactured by Refinetech Co., Ltd.) equipped with water-resistant abrasive paper (Carbo Mac paper, manufactured by Refinetech Co., Ltd.), a cross section was cut out to the vicinity of the center of the connection structure. The polished sample was made into a size that could be processed with an ion milling device by scraping off excess epoxy casting resin. Using an ion milling device (IM4000, manufactured by Hitachi High-Technologies Corporation) in the CP processing mode, cross-section processing was performed on the sized sample under the conditions of an argon gas flow rate of 0.07 to 0.1 cm ³ / min and a processing time of 120 minutes. This was used as a sample for SEM. The cross section of the sintered metal layer of this SEM sample was observed at an applied voltage of 15 kV using an SEM-EDX apparatus (ESEM XL30, manufactured by Philips).

-Calculation of ratio of major axis and thickness of flake-like structure A 5000 times SEM image obtained by "cross-sectional morphology observation" was read by Image J (manufactured by National Institutes of Health). The Straight Line was selected from the main window. Click from one end of the scale at the bottom of the image (scale showing 5 μm in this example) → Drag to draw a line, select [Analyze] → [Set Scale] from the main window, display the Set Scale window, and display the Known Distance "5" was entered in the box for ":", "μm" was entered in the box for Unit of length: and the [OK] button was clicked. Press the [T] key to display the ROI Manager window and check the Show All check box. The flake-like structure was specified by the same method as in (4), the line was clicked from end to end of the cross-section of the flake-like structure on the image → dragged, and the [T] key was pressed to register it in the ROI Manager window. This operation was repeated for all the flaky structures on the screen without duplication. A flaky structure protruding from the edge of the screen and having an image cut was not selected. Next, the Measure button in the ROI Manager window was pressed. Since the measured length is displayed in the Results window, it was saved in a file by [File] → [Save As]. Similarly, the length in the thickness direction of the cross section of the flaky structure on the image was measured and saved in a file. The saved file was read with Microsoft Excel, and the average length measurement result was calculated. Thus, the average of the major axis of the flaky structure and the average of the thickness of the flaky structure were obtained. Furthermore, the ratio of the number average length in the major axis direction to the number average length in the thickness direction of the flake-like structure was obtained by dividing the average major axis of the flake-like structure by the average thickness of the plate-like structure.

-Thermal conductivity The thermal diffusivity was measured by the laser flash method (LFA467, made by Netch Co., Ltd.) using the plate-like sample produced by the "copper content in the sintered metal layer" measurement. Silver at 25 ° C. is obtained by the product of this thermal diffusivity, the specific heat capacity obtained with a differential scanning calorimeter (DSC8500, manufactured by PerkinElmer) and the density determined by “(2) volume ratio of metallic copper”. The thermal conductivity [W / (m · K)] of the sintered body was calculated.

(Example 2)
In the same manner as in Example 1 except that the electroless palladium plating treatment time was changed from 9 seconds to 15 seconds in Step c of Example 1 to form an electroless palladium plating film of 5 nm (0.005 μm). An evaluation sample was prepared. The results are shown in Table 2.

(Example 3)
In step c of Example 1, except that the electroless palladium plating treatment temperature was changed to 65 ° C., the treatment time was 7 seconds, and the electroless palladium plating film was formed to 10 nm (0.01 μm). Evaluation samples were prepared in the same manner as in Example 1. The results are shown in Table 2.

(Example 4)
In step c of Example 1, except that the electroless palladium plating treatment temperature was changed to 65 ° C., the treatment time was 20 seconds, and the electroless palladium plating film was formed to 30 nm (0.03 μm). Evaluation samples were prepared in the same manner as in Example 1. The results are shown in Table 2.

(Example 5)
In Example c, except that the electroless palladium plating treatment temperature was changed to 65 ° C., the treatment time was 30 seconds, and the electroless palladium plating film was formed to a thickness of 50 nm (0.05 μm). Evaluation samples were prepared in the same manner as in Example 1. The results are shown in Table 2.

(Example 6)
In the process c of Example 1, except that the electroless palladium plating treatment temperature was changed to 65 ° C., the treatment time was 60 seconds, and the electroless palladium plating film was formed to 100 nm (0.1 μm). Evaluation samples were prepared in the same manner as in Example 1. The results are shown in Table 2.

(Example 7)
In step c of Example 1, except that the electroless palladium plating treatment temperature was changed to 65 ° C., the treatment time was 120 seconds, and an electroless palladium plating film was formed to 200 nm (0.2 μm). Evaluation samples were prepared in the same manner as in Example 1. The results are shown in Table 2.

(Example 8)
In step c of Example 1, except that the electroless palladium plating treatment temperature was changed to 65 ° C., the treatment time was 300 seconds, and the electroless palladium plating film was formed to 500 nm (0.5 μm). Evaluation samples were prepared in the same manner as in Example 1. The results are shown in Table 2.

(Comparative Example 1)
The steps (a) to (b) of Example 1 were performed. Then, the electroless nickel-plated copper plate which passed through the process b was immersed in HGS-100 (Hitachi Chemical Co., Ltd., brand name) which is a substitution gold plating solution for 10 minutes at 85 degreeC, and it washed with water for 1 minute. Thereby, a displacement gold plating film was formed to 0.1 μm on the electroless nickel plating film. Thereafter, an evaluation sample was prepared in the same manner as in step d and after in Example 1. The results are shown in Table 2.

  As shown in the results of the examples, excellent connection reliability could be expressed in any of the embodiments shown in FIGS.

  DESCRIPTION OF SYMBOLS 1 ... Sintered copper which has flake-like structure, 2 ... Sintered copper derived from copper particle, 3 ... Hole, 4 ... Sintered metal layer, 5 ... 1st member, 6 ... 2nd member, 7a, 7b: Surface in contact with the sintered metal layer 4, 11a, 11b ... Lead frame, 12 ... Wire, 13 ... Mold resin, 14 ... Semiconductor element, 100 ... Connection structure, 110 ... Semiconductor device, 21 ... Insulating substrate, 22 ... 1st electrode, 23 ... Semiconductor element, 24 ... 2nd electrode, 25 ... Metal wiring, 26 ... 3rd electrode, 27 ... Wire, 28 ... Copper plate, 29 ... Insulator, 30a, 30b ... Connection structure, 31 ... Nickel-containing layer, 32 ... sintered metal layer, 33 ... palladium layer (palladium-containing layer), 34 ... intermetallic compound layer (palladium-containing layer) containing palladium and a metal element contained in the sintered metal layer, 35 ... nickel In palladium and sintered metal layers Intermetallic compound layer containing a metallic element (palladium-containing layer), 200 ... semiconductor device.

Claims (14)

  A connection structure comprising a nickel-containing layer, a palladium-containing layer, and a sintered metal layer in this order.   The connection structure according to claim 1, wherein the nickel-containing layer is derived from an electroless nickel plating film, and the palladium-containing layer is derived from at least an electroless palladium plating film.   The connection structure according to claim 1, wherein the palladium-containing layer includes a palladium layer and an intermetallic compound layer containing a metal element contained in palladium and a sintered metal layer.   The connection structure according to claim 1, wherein the palladium-containing layer includes an intermetallic compound layer containing a metal element contained in nickel, palladium, and a sintered metal layer.   The sintered metal layer includes a structure derived from flaky copper particles oriented substantially parallel to the connection interface of the connection structure, and copper is 65 based on the total volume of the sintered metal layer. The connection structure according to any one of claims 1 to 4, which is contained by volume% or more.   A connection structure provided with the connection structure according to any one of claims 1 to 5 between the first member and the second member. Between the first member and the second member, the connection structure according to any one of claims 1 to 5 is provided,
A semiconductor device, wherein at least one of the first member and the second member is a semiconductor element.   The semiconductor device according to claim 7, wherein the first member is a semiconductor element, and the second member is a metal wiring. A method for producing a connection structure comprising a nickel-containing layer, a palladium-containing layer, and a sintered metal layer in this order,
A manufacturing method provided with the process of sintering the laminated body in which the nickel film, the palladium film, and the metal paste were provided in this order.   The manufacturing method according to claim 9, wherein the nickel coating is formed by electroless nickel plating, and the palladium coating is formed by electroless palladium plating.   The manufacturing method according to claim 9 or 10, wherein the nickel coating contains 80 mass% or more of nickel based on the total mass of the nickel coating.   The manufacturing method according to any one of claims 9 to 11, wherein the palladium coating contains 94 mass% or more of palladium based on the total mass of the palladium coating.   The manufacturing method as described in any one of Claims 9-12 whose thickness of the said palladium coat is 3 nm or more and 0.5 micrometer or less.   The manufacturing method according to any one of claims 9 to 13, wherein the metal paste includes metal particles and a dispersion medium, and the metal particles include flaky copper particles.