JPS59169694A - Solder and joining method thereof - Google Patents

Abstract

PURPOSE:To concentrate the strain in the thermally deforming stage of solder and to obtain a good bonded part by using the solder formed by adhering a low melting tin-lead alloy having a high content of tin on the surface of a high melting lead-tin alloy having a high content of lead and performing soldering at the adequate temp. between both m.p. CONSTITUTION:Solder 8 is manufactured by adhering alloys 8b, 8c, which has the compsn. consisting of >=60% tin and the balance lead and impurities and has a low m.p. and high yield stress, on a part or the whole of the surface of an alloy sheet 8a, which has the compsn. consisting of >=70% lead and the balance tin and impurities and has a high m.p. and yield stress. Such solder 8 is placed between a silicon plate 9 plated thereon with a nickel layer 10 and a copper plate 11, etc. of a semiconductor device, and both plates are soldered at the suitable temp. between the high m.p. of the main sheet 8a and the low m.p. of the auxiliary sheets 8a, 8b. The stress concentration in the thermal deforming stage of the solder 8 is thus prevented. The resistance to thermal fatigue of the solder is improved by controlling the thickness of the solder.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a solder used for adhering components such as semiconductor devices and a method for adhering the same.

[Background of the invention]

It is well known that conventional semiconductor devices are constructed by bonding materials with a low coefficient of linear expansion, such as silicon and ceramics, and materials with a high coefficient of linear expansion, such as copper and iron, using an adhesive such as solder. This is Notooshi. Such a semiconductor device has the disadvantage that the reliability of the semiconductor device is lowered because it is damaged due to temperature changes.

As mentioned above, with solder containing lead and tin as main components, the thermal fatigue strength decreases due to the formation of alloy metals during soldering. This mechanism will be explained with reference to FIGS. 1 and 2.

As shown in FIG. 1, when a pair of plates 1a and 1b are soldered with solder 2 whose main components are lead and tin,
At the boundary between a, lb and the solder 2, an alloy layer 3 with a thickness of about 5 μm is formed from the component elements of the plate 1a+lb and the component elements of the solder 2.
a and 3b are generated.

For example, when the plates 1a, lb are made of copper, the alloy layers 3a, 3b are Cu3Sn or Cu6Sns.

The element in the lead-tin solder that participates in the formation of the alloy layers 3a and 3b is only tin, and it is reported in the literature (Japan Institute of Metals Bulletin No. 21 No. 8) that in most cases lead is not involved. , p6
281 is clear. Therefore, the alloy layer 3a,
At the boundary between solder 3b and solder 2, there is an alloy layer 4 with a reduced tin content.
a+4b is generated.

On the other hand, in solder with a lead content of 70 or more, the yield stress decreases as the tin content decreases. The relationship between this yield stress, that is, the yield stress generated when a shear strain of 1 inch is applied to the solder, and the proportion of lead is shown in FIG. 2. As shown in this figure, the yield stress of an alloy with a composition of 701 or more lead and the remainder tin is 0.9 to 2 K4 f /
The yield stress of an alloy with a composition consisting of 401 or more tin and the balance lead is 1.2~ZOKpf/■
It turns out that it is 2. Further, when thermal strain occurs in such a solder-bonded structure (see FIG. 1), the strain concentrates on the alloy layers 4a and 4b, which have a low tin content, making these portions susceptible to thermal fatigue fracture.

Furthermore, even during subsequent operation, soldering causes an expansion of the constituent elements of tin and the plate within the solid, resulting in the growth of an alloy layer. This phenomenon is called tin cracking, and may have an adverse effect on the metal structure of the solder, resulting in a decrease in strength.

As mentioned above, thermal fatigue failure of solder joints in semiconductor devices is caused by
In most cases, solder 2 and alloy layers 3a and 3 shown in FIG.
This occurs in the alloy layers 4a and 4b at the boundary with b. Possible means for preventing such thermal fatigue failure include reducing temperature changes in the semiconductor device and creating a structure in which thermal distortion of the solder is less likely to occur. However, the former means for reducing temperature changes requires a device that keeps the ambient temperature of the semiconductor device constant and a component that improves the efficiency of dissipating heat generated from the semiconductor device.

On the other hand, as a means to prevent the latter thermal strain from occurring, it is effective to increase the thickness of the solder. The reason for this will be explained with reference to FIG. 3, which shows a state in which a plate 5 with a small coefficient of linear expansion and a plate 6 with a large coefficient of linear expansion are bonded together via solder 7 and then heated. It is. Before heating, the right end surfaces of plates 5 and 6 were formed on the same plane, but after heating, due to the difference in thermal expansion coefficient between both plates 5 and 6, a step difference Δt( (constant). In this case, the strain γ of the solder 7 is expressed by the following (1).

γ=Δt/h (1) According to equation (1), the larger the thickness h of the solder 7, the smaller the strain γ of the solder 7. However, in order to increase the thickness h of the solder 7, simply increasing the amount of the solder 7 does not achieve the objective. This is because the melted solder 7a flows out due to the weight of the solder 7 and the weight of the upper plate 5.

Therefore, a method has been proposed in which the lower plate 6 is provided with protrusions to ensure the thickness of the solder 7 mechanically (% opening/closing 5
3-1399741゜Since this method requires a process for machining the protrusions, and press working is usually used for this process, it is very difficult to provide the protrusions on brittle materials such as semiconductors, such as silicon and ceramics. Have difficulty.

[Purpose of the invention]

In view of the above, the present invention aims to prevent strain concentration during thermal deformation of solder, freely control the thickness of solder, and improve the thermal fatigue strength of solder.

[Summary of the invention]

In order to achieve the above object, the present invention has a high melting point alloy having a composition of 70 or more lead and the remainder being tin or impurities associated with tin. The solder is produced by layering a low melting point alloy having a composition of lead and impurities associated with lead with the remainder being lead or impurities associated with lead. Furthermore, soldering using such solder is carried out at an appropriate temperature between the melting points of the high melting point alloy and the low melting point alloy.

[Embodiments of the invention]

Embodiments of the present invention will be described below with reference to the drawings.

In FIG. 4, reference numeral 8 denotes solder formed in the form of a sheet, and this solder 8 is constructed by adhering sub-sheets 8b and 8C to both sides of a main sheet 8a by appropriate means such as pressure bonding, plating, vapor deposition, etc. ing. The main sea) 8a is an alloy with a high melting point of 260 to 327C and a small yield stress,
For example, lead: 70%.

Tin: Made by rolling an alloy having a composition of 30%, the sub-sea) 8b and 8c have a low melting point of 183 to 232C.
and alloys with high yield stress, such as lead: 4 (1, tin =
It is made by rolling an alloy with a composition of 60 mm.

FIG. 5 shows a first embodiment of the solder bonding method of the present invention. That is, a silicon plate 9 with a low coefficient of linear expansion and a copper plate 11 with a high coefficient of linear expansion, which have been plated with a nickel layer 10 in advance, are applied to the solder 8. A semiconductor device can be manufactured by bonding the silicon plate 9 and the copper plate 11 by soldering at a temperature between the low melting points of C) 8b and 8C.

In this case, between the nickel layer 10 and the sub-sea) 8b,
Nis an, Njs Sn4, Nis S-
An alloy layer 12a mainly composed of
Between the copper plate 11 and the sub-sheet 8c, there is Cu38nlC.
An alloy layer 12b containing u6 Sns as a main component is produced. The tin contained in both these alloy layers 12a and 12b is
Since they are supplied from the sub-sheets 8b and 8c, the main y
The component composition of -18a remains unchanged. In this way, since strain concentration due to a decrease in the tin content does not occur in any part of the main sheet 8a, the thermal fatigue strength does not decrease, so that the reliability of the semiconductor device can be improved.

Conventionally, when bonding a pair of plates with solder, it was difficult to increase the thickness of the solder to about 100 μm or more, but in this embodiment, the thickness of the main sheet 8a with a high melting point can be ensured. Therefore, the solder 8 can be set to have an arbitrary thickness, and a uniform thickness can be easily obtained. Furthermore, when the thickness of the solder 8 is increased, the strain r of the solder becomes smaller according to the equation (1), so that the thermal fatigue strength of the solder 8 can be improved.

The second embodiment shown in FIG. 6 has a spherical solder 15 formed by bonding a low melting point solder 15b to the surface of a spherical portion 15a formed of the high melting point solder, and a pair of plates 13, 14.
Soldering was performed at a temperature between the two melting points.

By using the spherical solder 15 in this manner, the solder thickness can be ensured by the spherical diameter, thereby preventing strain concentration during thermal deformation of the solder 15 and improving the thermal fatigue strength of the solder 15.

In this case, in order to bond the specific points 17a, 17b of the plate 13 and the specific points 16a, 16b of the plate 14 with the spherical solder 15, first the specific points 16a of the lower plate 14 are bonded.

Each spherical solder 15 is placed on the solder 16b, and soldering is performed at a temperature between the melting points of the high melting point solder 15a and the low melting point solder 15b. Then, it is cooled and the spherical solder 15 is attached to the lower plate 14.
After fixing, the specific points 17a and 17b of the upper plate 13 are respectively placed on the spherical portions 15a of the respective spherical solders 15, and soldering is performed again at the above temperature.

In the third embodiment shown in FIG. 7, a pair of plates 13 and 14 are bonded to the second plate using solder 18 formed by bonding an arbitrary number of juxtaposed spherical high melting point solders 18a with low melting point solder 18b.
It is soldered in the same manner as the embodiment (FIG. 6). When such solder 18 is used, a solder thickness equal to the diameter of spherical solder 18a can be ensured, so that the thermal fatigue strength of solder 18 can be improved.

〔Effect of the invention〕

As described above, according to the present invention, the thermal fatigue strength of the solder can be improved by preventing strain concentration during thermal deformation of the solder and freely controlling the solder thickness. Therefore, if the present invention is applied to a semiconductor device, its reliability can be improved.

[Brief explanation of the drawing]

Figures 1 and 3 are explanatory diagrams of the mechanism of thermal strain in conventional solder, Figure 2 is a diagram showing the relationship between the component composition and yield strength of solder consisting of lead and tin, and Figure 4 is a diagram showing the relationship between the composition and yield strength of solder consisting of lead and tin. 5 to 7 are cross-sectional views showing an embodiment of the solder bonding method of the present invention. 8.15.16...Solder, 8 a * 15 a
+ 18 a...High melting point alloy, 8b, 8C, 15
b, 18b...low melting point alloy. Figure 1 fJ z Figure rb gIi Table Figure 3 Figure 4 Figure 5 Figure 7 Figure 7

Claims (1)

[Claims] 1. Part or all of the surface of an alloy with a composition consisting of 70 or more lead and the remainder impurities associated with tin, with 60 or more tin and the remainder lead or A solder characterized in that an alloy having a composition consisting of lead and accompanying impurities is bonded by an appropriate means. 2. Part or all of the surface of a high melting point alloy with a composition consisting of 70 or more parts of lead and the balance of tin or impurities associated with tin, with 60 or more parts of tin and the balance of lead or impurities associated with lead. A solder produced by adhering a low melting point alloy with impurities by an appropriate means is soldered at an appropriate temperature between the melting points of the high melting point alloy and the low melting point alloy. Solder bonding method.