JPH0426958B2 - - Google Patents

Description

Detailed Description of the Invention [Field of Application of the Invention] The present invention relates to a method for electrically, thermally, and mechanically joining two or more dissimilar members having mutually different coefficients of thermal expansion;
In particular, the present invention relates to a soldering method that is effective in reducing electrical disconnection and mechanical damage at joints.

[Background of the invention]

Solder materials whose main components are lead and tin are used to electrically connect dissimilar components with different thermal expansion coefficients or materials.
It is used as a material for thermal or mechanical bonding in a wide range of technical fields, especially in the electronic industry.

An example of conventionally known joining in this type of technical field will be described below. (1) U.S. Patent No. 3,429,040 discloses a preferred form of mechanically and electrically bonding a semiconductor substrate onto a conductor of a dielectric substrate by a method called Controlled Collapse Bonding. A solder composition of about 5 to 40 weight percent tin and 95 to 6 weight percent lead is used to reflow the interconnect. (2)IEEE
Trans.on Parts, Hybrids, and Packaging,
“Thermal Fatigue Failure of Soft” by SKKang et al. in PHP-13, 318 (1977)
Soldered Contacts to Silicon Power
In a paper titled ``Transistors,'' a silicon substrate was mounted on a copper plate soldered with silver to a steel plate using lead-based solder (95% lead-5% tin solder).
The die bonding structure has been started. (3) JP-A-51-130285 discloses that a semiconductor chip having a nickel layer on the adhesive surface and a metal support are
Solder bonded die bonded structures consisting of weight percent silver, 1.5-4.5 weight percent tin, and 93.5-97.5 weight percent lead have been initiated. (4) IEEE Trans.
on Components, Hybrids, and Manufacture
“Temperature cycling of
In the paper titled ``HIC Thin-Film Solder Connections'', a package structure is disclosed in which metal terminals are fixed and integrated with a ceramic board for hybrid integrated circuit devices using 60% tin-40% lead solder.(5) Solid State Technology, July, 54,
“Mechanical Design of Chip Components for
In a paper titled “Flip” and Short Beam-Lead Mounting, a two-terminal capacitor element was
A structure is disclosed in which wiring is soldered onto an alumina substrate using 60/40 tin-lead solder. (6) JP-A-58-39047 discloses that the atomic ratio of lead/tin after adhesive heat treatment is about 99.5/0.5 to 70/30 on a fine ohmic contact electrode formed on a semiconductor substrate. A semiconductor device is disclosed in which metal foil terminals patterned on an organic film are fixedly integrated using solder. (7)Electronic technology, No.
23, No. 7, 88 (1981), Arino et al. discloses the composition and melting temperature of various solder materials whose main components are lead and tin, and in particular It is disclosed that a solder material having a composition of 50% by weight lead-50% by weight tin and a solder material having a composition of 30% by weight lead-70% by weight tin are used for semiconductor assembly.

The reasons why lead-tin solder is used in the above technical field are (1) lead and tin are relatively inexpensive and easy to alloy and process, which is economically advantageous; (2) the melting point of the solder material is is 183-327℃ and other gold-based, silver-based,
(3) Lead-tin solder is lower than other metal solders, and has many advantages in the joining process, such as being able to join without damaging parts around the joint; (3) lead-tin solder is lower than other metal solders; This is because it is soft, has excellent plastic deformability, and is considered to be excellent as a thermal strain absorbing carrier. However, among the lead-tin solders, the most commonly used solder materials are lead-based solder materials such as 95% lead-5% tin solder, or 40% lead-60% by weight. It was a eutectic solder material such as that represented by weight% tin solder. The reason for this is that lead-based solders, e.g.
As evidenced by reference to the binary alloy phase diagram described in the publication entitled “Constitution of Binary Alloys” by Max Hansen, published by McGrow-Hill Book Company, Inc., p. 1106 (1958), soft In this stress field, the α solid solution (a small amount of tin in solid solution in lead), which is rich in plastic deformation, becomes the β solid solution (a small amount of lead in solid solution in tin), which is hard and difficult to deform plastically.
The solder material as a whole is more susceptible to plastic deformation, absorbs thermal stress or strain due to the difference in thermal expansion coefficient of the parts to be joined, and maintains the electrical and mechanical performance of the joint. Because it was considered advantageous. On the other hand, eutectic solder
As is clear with reference to the above publication by Hansen, the composition is eutectic (37 wt.% lead - 63 wt.%
It exhibits a eutectic structure of the above α solid solution and β solid solution consisting of fine crystal grains. In this case, since the eutectic solder has fine crystal grains, it tends to cause plastic deformation due to grain boundary interface slip and diffusive creep action, such as transaction of
SW in the ASM, Vol.61, 300 (1968)
“Suprplasticity” by Zehr and WABackofen
It exhibits a superplastic phenomenon as disclosed in the paper entitled "In Lead-Tin Alloys", and absorbs the thermal stress or strain caused by the difference in thermal expansion coefficient between the joined parts, and the electrical and This is because it was thought to be advantageous in maintaining thermal and mechanical performance.

The above-mentioned prior art examples (1) to (6) also disclose that a lead-based solder material or a eutectic solder material can be used. On the other hand, the prior art (1) uses a solder composition of 40 wt% tin and 60 wt% lead, and (7) uses a 50 wt% lead-50 wt% tin solder and 30 wt% lead-
It is disclosed that solder materials in composition ranges other than lead-based or eutectic solder materials may be used, such as 70% by weight tin solder. However, solder materials that belong to composition ranges other than lead-based or eutectic solders are considered to be inferior to lead-based or eutectic solders in terms of reliability, and common sense based on conventional knowledge should avoid using them. was common. That is, in the composition range outside the lead-based system or the eutectic system, (1) for example, 60 wt% lead - 40 wt% tin solder, as the abundance ratio of the primary α solid solution changes, the eutectic occurs at the grain boundaries between the α solid solutions. (2) The presence of a primary α solid solution, which inhibits superplastic action, in the fine eutectic structure reduces grain boundary deformation performance, and (2) for example, In the 30 wt% lead-70 wt% tin solder, a relatively hard primary β solid solution exists in the fine eutectic structure that inhibits superplastic action, reducing the plastic deformation performance due to grain boundary sliding, resulting in poor connection. This is because there was a concern that the solder material itself that carries the connection part or the connected members arranged in series with the connection part may be damaged. In other words, solder materials belonging to composition ranges other than lead-based or eutectic systems cannot be used for the purpose of metallurgically joining, physically integrating, or electrically connecting dissimilar components. Technical example
It was a material capable of acting as disclosed in (1) or (7). However, from the perspective of reducing physical failures that result from the introduction of stress or strain to the solder material that serves as the joint and the accumulation of fatigue due to the thermal cycles that an integrated product of dissimilar parts undergoes during use, It had not been considered a material that could meet that purpose. In addition, in addition to this background, there has not been sufficient study of the bonding process for the purpose of improving reliability when using solder materials belonging to a composition range other than lead-based or eutectic, and it is not commercially viable. The reality is that no process has been established.

[Purpose of the invention]

An object of the present invention is to provide a joining method that can improve the thermal fatigue resistance of a joined part of an integrated product formed by metallurgically joining dissimilar members or dissimilar members having mutually different coefficients of thermal expansion.

Another object of the present invention is to provide a new joining method in which the metallurgical process for joining is improved using a compositionally adjusted solder material as a means of improving the fatigue resistance of the joint.

Another object of the present invention is to obtain an integrated product in which the cooling rate or temperature profile after melting of the solder material is adjusted using a solder material belonging to a composition range other than lead-based or eutectic solder material. The object of the present invention is to provide a new joining method.

Another object of the present invention is to provide a novel metallurgical joining method for obtaining a metal structure that can suppress the progress of fatigue of a solder layer that serves as a joint as much as possible.

[Summary of the invention]

The solder joining method of the present invention is characterized in that tin contains 35% or more by weight.
A solder material containing 80% by weight or less (excluding 60% by weight or more and less than 65% by weight) and the remainder consisting essentially of lead, or a solder material in which a metal as a third element is added to the same solder material is a different type of solder material. It is characterized by having a metallurgical process in which the solder material is interposed between members or different members having different thermal expansion coefficients, and the solder material is melted and then cooled at a rate of 125° C./min or less.

As a result of various studies conducted by the present inventors, it was found that the thermal fatigue strength of an integrated product made by joining dissimilar parts using a solder material mainly composed of lead and tin has an extremely clear compositional dependence, which is particularly desirable. It was confirmed that the integrated product bonded under metallurgical process conditions exhibits extremely excellent thermal fatigue life characteristics when the solder composition is 50 wt% lead-50 wt% tin and 25 wt% lead-75 wt% tin. did. The preferred metallurgical process conditions mentioned here are solder materials belonging to a composition range other than lead-based or eutectic solders, more preferably lead in a range of 40 wt% or more and less than 65 wt%, and tin in a range of 35 wt% or more and 60 wt%. Solder material with less than 20 to 35% lead by weight
A solder material containing 65 to 80% by weight of tin is interposed between dissimilar components, and after melting this solder material, the solder melt is melted at a selected speed for at least a period of time until the solder melt is completely solidified. That is, cooling at a rate of 125° C./min or less.

The reason for choosing such a metallurgical process is, firstly, to increase the fracture strength or elastic stress range of the solder material itself, and secondly, to introduce a metal structure that is resistant to plastic deformation or suppresses plastic deformation into the solder layer. That's true. By achieving these points, the rigidity of the solder layer, which is the softest member in an integrated product and where stress concentration and associated plastic deformation are noticeable, is increased, and the stress in the same layer is dispersed to reduce plastic deformation. The aim is to reduce the amount of carbon dioxide and improve fatigue life performance.

[Embodiments of the invention]

Next, embodiments of the present invention will be described in further detail with reference to the drawings.

FIG. 1 relates to a first embodiment of the present invention.
Silicon transistor substrate 3 (thermal expansion coefficient 3.5×10 ^-6 /
12 is a graph showing the thermal fatigue life of the joint of the integrated product of copper supporting plate 2 (16.5×10 ^-6 /°C) and copper support plate 2 (16.5×10 −6 /°C).
This thermal fatigue life data (marked with ○) is -
Temperature change from 55℃ to +150℃ (temperature change width:
The life is defined as the number of repetitions when the thermal resistance, which is read sequentially while repeatedly applying 205 deg., reaches 1.5 times the initial value, and the life values of 10 samples are plotted on a Weibull chart. It expresses the average lifespan determined from the straight line obtained.
Here, the thermal resistance is calculated from the silicon substrate 3 to the copper support plate 2.
is the steady-state thermal resistance of the heat conduction path leading to , which should rise due to fatigue cracks in the solder layer 1 of the integral. The integrated product has a thickness of 100μ on a support plate 2 made of a nickel-plated copper plate with a thickness of 3mm.
Silicon substrate 3 [13
mm×13mm, thickness 250μm, joint surface; chrome (0.1μ
m) - Nickel (0.6 μm) - Silver (2 μm) multilayer vapor-deposited metallization] is placed and heated in a hydrogen atmosphere to melt the solder sheet, at least the solid phase of the molten solder is completed. Period up to 40±10
It was obtained by cooling at a rate of °C/min. The solder sheet used was made of an alloy selected from a composition range consisting of 35 to 59% by weight of tin with the balance being substantially lead;
It is an alloy selected from a composition range of 65 to 80% by weight, the balance being substantially lead. The heat treatment in the above-mentioned metallurgical process is controlled so that the maximum temperature reached is approximately 50 degrees higher than the liquid point of the solder material, and so that the temperature reached is maintained for approximately 5 minutes.
In addition, it should be selected and used in consideration of the fact that the metallurgical bond between the highest temperature solder material and the parts to be joined is completely achieved, and the constituent components of the parts to be joined are mixed into the molten solder. Any temperature above the liquidus point of the solder material can be selected. In the same sense, the retention time can also be adjusted arbitrarily.

Referring to FIG. 1, the thermal fatigue life of the integrated silicon substrate 3 and copper support plate 2 soldered together has a clear composition dependence, and the tin concentration in the solder layer 1 is 50% by weight. It shows a life characteristic that has a maximum value at 75% by weight. In addition, when a solder material having the composition range of the above example is used, the life value is 500.
recorded more than once. In FIG. 1, as a comparative example, the life characteristics of an integrated product produced through a similar metallurgical process using a solder material having a composition selected from the range of the above-mentioned examples are indicated by black circles. As is clear from the comparison with the integrated products of the above-mentioned Example compositions, the integrated products of these comparative examples reach the end of their lifespans at a stage where the number of cycles is small. In particular, in comparison with integrated products using lead-based or eutectic solders, which have been commonly used to join dissimilar parts, all integrated products using the solder material of the above example composition have a long lifespan. ing. From the above disclosure, it is understood that the metallurgical process of this embodiment has an extremely effective effect in improving the fatigue life of the integrated product. Although the reasons for the above-mentioned composition dependence of life and the preferable life characteristics in the example composition range are not necessarily sufficiently clear, the present inventors believe that the reasons are due to the following mechanism based on the results of studies conducted so far. I guess.
The mechanism will be explained below.

Figure 2 relates to the second embodiment, and shows an alumina substrate (width
15mm, length 30mm, thickness 0.6mm, thermal expansion coefficient 7.5×
10 ^-6 /℃) 4 and brass terminal member (width 1 mm, length 13
Shear strength characteristics (curves A ^, B, C) is shown. Curve A is -25℃, curve B is
20℃, and the same C is the measured value at 100℃. Referring to the same figure, it is observed that the shear rupture strength is compositionally dependent, with maximum values occurring when the tin composition in the solder is 50% and 75% by weight. The strength decreases greatly when the ambient temperature is low and as the ambient temperature increases, but it shows approximately the same composition dependence at any temperature.
This tendency has a clear correspondence with the life characteristics shown in FIG. 1, and is understood to suggest that the degree of reinforcement of the solder layer 1 is reflected to a large extent in the fatigue life characteristics. For comparison, the figure shows solder 1 in an integrated product of an alumina substrate 4 obtained by cooling at a cooling rate of approximately 200°C/min after melting the solder and a brass terminal 5 plated with nickel with a thickness of 3 μm.
An example of shear rupture strength characteristics (curve D) is shown. In this case, for example, Material Testing Technology, Vol. 25, No. 1, 31
Similar to the tensile strength obtained in the tensile test of rolled copper sheets disclosed in the paper titled "Effect of tensile speed on tensile strength and elongation of Pb-Sn binary alloy" by Nishihata et al. (1980). However, curves A, B, and C have different resistances. Therefore, another important point that FIG. 2 means is that the solder layer 1 obtained by rapid cooling is not reinforced even if it is a solder material having the composition of the above example. Therefore, the solder material having the composition of the above example is strengthened.

The integrated product is made by interposing a mixed paste of rosin-based flux and solder powder between the terminal 5 and the alumina substrate 4, and heating it in the air to melt the solder powder and raise it from the liquidus point of the solder material. It was obtained by holding at a temperature approximately 50°C higher and then cooling at a rate of 60±10°C/min until at least the solidification of the molten solder was completed, and curves A, B, and and C are data obtained from a tensile test (tensile speed 2 mm/min) under controlled temperature. A bonding pad (width 1 mm, length 1.5 mm) is formed on the alumina substrate 4 of this bonding area, and the molybdenum fired metallized layer is plated with nickel to a thickness of about 3 μm.The amount of solder supplied to the bonding area is controlled by the metallurgical process. The thickness of the solder layer after soldering is adjusted to approximately 100 μm. Even in this case, the results of the thermal fatigue life characteristics of the solder layer 1 were similar to those shown in FIG. 1 in terms of trends with respect to composition.

Figure 3 shows solder materials selected from the compositions of the above embodiments (50 wt% lead-50 wt% tin: curve A, 25
This is the relationship between the cooling rate after melting the solder and the shear rupture strength (at room temperature) in an integrated product of the alumina substrate 4 and the terminal 5 obtained using the curve B): lead by weight - 75% by weight. Referring to the same figure, the shear strength is large in the region where the cooling rate is low, and shows a small value when cooling at high speed, but stable high strength is obtained at approximately 125°C/min or less. It is understood that Note that even if the composition was other than the above-mentioned 50% by weight tin or 75% by weight tin, the solder materials in the Examples range had similar cooling rate dependence.

As mentioned above, the first reason for the favorable life characteristics is that the solder material
It is understood that this is a point that itself is strengthened.

Next, the second reason for exhibiting preferable life characteristics will be explained.

FIGS. 4a and 4b show preferred metallurgical processes for 50 wt% lead-50 wt% tin solder material and 25 wt% lead-75 wt% tin solder material selected as representative examples of solder materials in the above embodiment composition range, respectively. (Cooling rate 40℃/
Figures c and d show the metal structure after undergoing a metallurgical process (cooling rate of 150° C./min), which is not preferred in the present invention. Referring to figure a, the solder layer exhibits a structure in which a relatively large grain size eutectic surrounds α solid solution primary crystals that have grown to large grain sizes, whereas in figure c, the solder layer has a relatively small grain size. It has a structure in which an α solid solution primary crystal is surrounded by a eutectic consisting of fine grains. Referring to figure b, in the solder layer, the large-grained eutectic surrounds the large-grown β-solid solution primary crystal, whereas in figure d, the small-grained β-solid solution primary is surrounded by fine-grained symbiotic crystals. It represents an organization.

As mentioned above, a eutectic consisting of fine crystal grains is thought to be easily deformed due to superplasticity, but as shown in Figure 4a, the α solid solution and β solid solution that make up the eutectic vary in their crystal grain size. Increasing the eutectic leads to a decrease in the plastic deformation performance of the eutectic itself, and also has the effect of suppressing the plastic deformation of the α solid solution primary crystal. On the other hand, when a eutectic with a fine grain size exists as shown in c in the same figure, the deformation performance of the eutectic itself is maintained, and as a result, the effect of suppressing the plastic deformation of the primary α solid solution is also reduced. On the other hand, when a large-grown β solid solution pair primary crystal is surrounded by a eutectic with a large grain size as shown in Figure 4b, not only is the β solid solution difficult to plastically deform, but the eutectic itself also loses its plastic deformation performance. On the other hand, as shown in Figure d, the β solid solution primary crystals with a relatively small particle size are surrounded by eutectic particles with a fine particle size.
It contributes to maintaining the plastic deformation performance of the entire solder layer. That is, under favorable metallurgical process conditions, the yield strength of the solder layer, in other words, the rigidity, is increased, and the stress or strain concentrated in the solder layer is appropriately dispersed to the members 2, 3, 4, and 5 to be joined, and the final It is understood that as a result of this, the accumulation of fatigue due to plastic deformation is suppressed. The reason why the fatigue life characteristics are inferior in a composition range other than the composition range of the above example where the lead content is higher, such as 95 wt% lead-5 wt% tin solder, is because the presence ratio of α solid solution is overwhelmingly high in the solder layer. This is because the overall plastic deformation performance is maintained, and the reason why the fatigue life characteristics decrease in the case of a composition with an extremely high tin content, such as 5% lead-95% tin solder, is because the lead concentration is low. This is understood to be because the β solid solution has properties close to pure tin metal.

To summarize the above description, the second reason above is:
Solder layer 1 obtained through a preferred metallurgical process
This is due to the fact that the crystal grain size constituting the eutectic is large, and the deformation performance of the eutectic itself is reduced, resulting in an increase in yield strength and stress dispersion.

In the present invention, the solder material is used for the purpose of electrically, thermally, or mechanically joining dissimilar members with different coefficients of thermal expansion, and in any of these cases, the solder material and the members to be joined are It is desirable that the metallurgical bond be as strong as possible.
Since it is metallurgical joining, it is natural that alloying occurs between the solder material and at least the outermost layer constituent material of the members to be joined in the main part of the joining part. This means that metals other than lead and tin as a third component are inevitably dissolved or mixed into the solder material after joining is completed. Further, it may be preferable to add metals other than lead or tin to the solder material for the purpose of improving the wettability of the solder material and suppressing contamination of constituent components of the parts to be joined. A problem to be worried about in this case is that the fatigue resistance of the solder material itself deteriorates due to the addition of the third component. This point will be explained below.

FIG. 5 relates to the third embodiment,
A semiconductor substrate 31 made of silicon as an integrated circuit (hereinafter abbreviated as IC) chip (width 6 mm, length 7 mm, thickness 0.4 mm) is made of soda glass (thermal expansion coefficient 9×10 ^-6 /°C) as an induction-resistant substrate. Controlled Collapse using minute solder group 11 on board 6
2 is a graph showing the thermal fatigue life of a joint part of an integrated structure for a liquid crystal display device that is electrically and mechanically connected by a method known as a bonding method. This thermal fatigue life data shows that when the IC chip 31 is energized intermittently and the chip 31 is repeatedly subjected to temperature changes from 50°C to 125°C (temperature change width 75 degrees), the chip 31 is It is the number of power cycles when the circuit function as an IC is lost, and is expressed as the average value of 15 to 20 samples. The figure shows that tin is 35 to 59% by weight or 65 to 80% by weight;
In the case of the example composition using micro solder 11 selected from a composition range in which copper is 1.0% by weight and the balance is substantially lead (○ mark), and in other compositions, copper is 1.0% by weight.
The results are shown for a comparative example composition (● mark) using a small amount of solder containing % by weight. Referring to the figure, the power cycle life of the integrated product has a life characteristic that has a maximum value when the tin concentration of the solder layer 11 is 50% and 75% by weight, and it depends on the composition in the same way as in Figure 1. It has a sexual nature. From this disclosure,
Even when copper is added as a third component to the micro solder 11, the composition dependence of fatigue life has the same tendency as when no copper is added, and in the case of the example composition range, the composition dependence of the fatigue life is similar to that of the comparative example. It is recognized that the composition has better life characteristics than that of lead-based or eutectic compositions.

In the third embodiment, the IC chip substrate has chromium (0.1 μm)-copper (0.6 μm)-nickel (0.3 μm) selectively formed on the aluminum wiring pattern.
)
A metal layer made of laminated metal and chromium (0.1 μm)-copper (2 μm) selectively formed on the glass substrate 6
A minute solder 11 is interposed between the laminated metal patterns,
Vapor that utilizes the latent heat of vaporization of fluorocarbon liquid
After melting a small amount of solder 11 by heating to a temperature approximately 50°C higher than the liquidus point of the solder material by the Cdendensation Soldering method, the solder material is heated for a period of 15±5°C/min at least until solidification of the solder is completed. It was obtained by cooling at high speed. In such metallurgical processes, silver, silver,
gold, palladium, nickel, antimony, zinc,
Even when bismuth, indium, cadmium, arsenic, and gallium are added, or when these metals are added in combination, the same compositional dependence as described above is observed, especially when it comes to gold, silver, and palladium. It was confirmed that the product life could be further extended compared to the case without additives. Further, even if the third additive metal is included, fatigue life performance will not be impaired if a preferable metallurgical process of cooling at a rate of 125° C./min or less is performed.

In the embodiments of the present invention, bonding methods for semiconductors and metals, ceramics or glass inorganic derivatives, and metals, particularly for electronic devices, have been disclosed as preferred embodiments. However, according to the preferred metallurgical process of the present invention, it is possible to bond organic resins such as glass epoxy, paper phenol, and epoxy with metals, inorganic derivatives, and semiconductors, or to bond organic resins with each other, metals with each other, and derivatives with each other. Improved fatigue life properties can be obtained in bonding or integrated bonding of any two or more organic resins, metals, derivatives, semiconductors, or even in technical fields other than the electronics industry. In addition, in metallurgical processes, gases such as hydrogen, nitrogen, helium, argon, and carbon dioxide gas, and vapors such as fluorocarbons, can be used in controlled atmospheres or in air without any problems. Without,
From this point of view, heat treatment in a furnace, heat treatment on a hot plate, beam irradiation heat treatment such as infrared rays, laser beam or electron beam, immersion heat treatment, vapor heat treatment, etc.
Any method such as condensation soldering heat treatment or parallel gap soldering heat treatment can be applied, and it is also possible to use different fluxes depending on the performance that the integrated product should have. In the present invention, dissimilar members refer to members having mutually different physical property values, such as thermal expansion coefficient, thermal conductivity, mechanical properties, and electrical properties. Therefore,
For example, even if the member to be soldered is a molybdenum plate, a single molybdenum plate and a molybdenum plate bonded with silver solder to a copper plate belong to the range of different types of members. Furthermore, even if it is a composite member made by joining a copper plate and a steel plate, those made by diffusion bonding and those joined by brazing metal belong to the range of different types of members.

〔Effect of the invention〕

The present invention discloses an improved metallurgical process for electrically, thermally, and mechanically joining dissimilar members having mutually different coefficients of thermal expansion. The effects and advantages of the present invention will be shown below.

(1) In the present invention, the integrated product obtained through the preferred metallurgical process has further improved thermal fatigue life performance of the joint. This is mainly due to the improvement in the strength of the solder material itself and the effect of suppressing plastic deformation, and contributes dramatically to improving the quality and reliability of the integrated member.

(2) For example, in the case of an integrated product of a large semiconductor substrate and a metal support member, in order to maintain its thermal fatigue life performance, it is recommended to interpose an intermediate member such as molybdenum that alleviates the difference in thermal expansion coefficient. As always, the preferred metallurgical process of the present invention significantly improves fatigue life performance, thereby reducing the need for intermediate components. This contributes to reducing the number of parts. Furthermore, in the case of an integrated product of a heat generating member or a heat receiving member and a heat absorbing/radiating or heat transmitting member, the interposition of an intermediate member can be avoided, which also contributes to improvement of heat conduction performance.

(3) The solder material is strengthened by the metallurgical process of the present invention, and can be directly soldered and joined even when joining members with a large difference in coefficient of thermal expansion. Therefore, there is no need to strictly select the expansion coefficient of the members to be joined, increasing the degree of freedom in designing the integrated product.

(4) In the metallurgical process of the present invention, highly reliable electrical, thermal, and mechanical bonding can be realized by using a general alloy containing lead and tin as main components as the bonding member. From this point of view, it is a great contribution to the economic aspect.

[Brief explanation of drawings]

FIG. 1 is a graph showing the fatigue life characteristics of an embodiment of the present invention, FIGS. 2 and 3 are graphs showing the shear rupture strength of the solder material, and FIG. 4 is a graph showing the improved metallurgical process and the improved metallurgical process. FIG. 5, which is a diagram showing the difference in the metal structure of the solder material obtained by the metallurgical process, is a graph showing the fatigue life characteristics of another example. 1...Solder layer, 2...Copper support plate, 3...Silicon base.

Claims (1)

[Claims] 1. A solder material containing 35% by weight or more and 60% by weight or less of tin, with the remainder being substantially lead, is interposed between dissimilar members, and after melting the solder material, the solder material is heated at 125°C/ A method for soldering dissimilar parts, characterized by passing through a metallurgical process of cooling at a rate of less than 1 minute. 2. The method of soldering dissimilar members according to claim 1, wherein the dissimilar members have mutually different coefficients of thermal expansion. 3 In claim 1, at least the period until solidification of the molten solder material is completed.
A method for soldering dissimilar parts, characterized by cooling at a rate of 125°C/min or less.