JPH07249631A - Manufacture of solder bumps and solder ball and manufacture of semiconductor device - Google Patents

Abstract

PURPOSE:To provide a method of forming solder bumps or solder balls of a desired size. CONSTITUTION:A solder ball forming member 10 having a flat surface 10a and a plurality of recessed parts 12 arranged in a specified pattern is prepared. The recessed parts 12 are filled with solder paste 14 composed of solder grains 16 and flux 15. The solder ball forming member 10 is heated, and solder grains 16 contained in the solder paste 14 in each of the recessed part 12 are fused. By the effect of surface tension, round solder balls 20 are formed. A member 22 on which solder bumps are to be formed is made to relatively approach the solder ball forming member 10. The heated solder balls 20 are transferred from the solder ball forming member 10 to the member 22 on which solder bumps are to be formed, in the same pattern as the pattern of the recessed parts 12.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of manufacturing a semiconductor device, for example, a method of manufacturing a solder bump for efficiently forming a solder bump on an electrode pad of a semiconductor chip. More specifically, the present invention relates to a solder bump manufacturing method for forming solder bumps by forming solder balls from the solder paste and a solder ball manufacturing method for forming solder balls from the solder paste.

[0002]

2. Description of the Related Art LSIs, VLSIs and the like in which electronic circuits and electronic components are integrated on a semiconductor chip are often used in information processing apparatuses for processing a large amount of information at high speed. In order to attach a semiconductor chip on which an electronic circuit is formed to, for example, a ceramic substrate, a solder bump is provided on the semiconductor chip or the ceramic substrate, and by welding the solder bump, the semiconductor chip is fixed to the ceramic substrate and electrically connected. be able to.

Conventionally, a solder paste is printed on a ceramic substrate, the leads attached to the semiconductor chip are placed on the solder paste of the ceramic substrate, and the solder is melted by heating, so that the semiconductor chip is mechanically attached to the ceramic substrate. And was electrically connected. In order to print the solder paste on the ceramic substrate, for example, a metal mask having openings of a predetermined pattern is used, the solder paste is placed on the metal mask, and the squeegee is moved along the surface of the metal mask while soldering is performed. Print the paste on a ceramic substrate.

However, when electronic circuits on a semiconductor chip are integrated, the number of leads serving as input / output terminals increases and the pitch between leads becomes smaller. In that case, the amount of solder paste to be printed becomes small, and it becomes difficult to print the solder paste. That is, the solder paste usually has a particle size of 30-
It consists of 50 μm solder particles and a flux containing pine resin, activator, organic solvent and the like, and the viscosity of the solder paste is as high as about 20 to 30 × 10 ⁴ cp at room temperature. Therefore, when the size of the opening of the metal mask is reduced to about 100 μm, the solder paste is clogged, the paste does not smoothly pass through the opening of the metal mask, and the ceramic substrate cannot be printed.

In the recent flip-chip structure, solder bumps are provided on the semiconductor chip without mounting the leads on the semiconductor chip, and the semiconductor chip is directly attached to the ceramic substrate. Also in this case, when the size of the solder bump is reduced, it becomes difficult to print the solder paste. Therefore, it has been proposed to form small solder bumps by forming solder bumps substantially only with solder without using a solder paste. As such a method, a vacuum deposition method and a solder plating method are proposed. There is.

The vacuum vapor deposition method is disclosed in, for example, Japanese Patent Laid-Open No. 4-014.
No. 834, the solder is heated in a vacuum to evaporate and the evaporated solder is attached. In the method described in the above publication, a transfer plate such as quartz or glass that is transparent and has excellent heat resistance, but has poor solder wettability, and a metal mask having openings of a predetermined pattern are used. The transfer plate is arranged immediately above the metal mask, and the solder vapor that has passed through the opening of the metal mask is deposited on the transfer plate. Therefore, solder bumps are formed on the transfer plate in a pattern corresponding to the openings of the metal mask.

Next, the semiconductor chip is heated and brought close to the transfer plate, so that the solder bumps of the transfer plate are transferred to the semiconductor chip. In this way, solder bumps are formed in a predetermined pattern on the semiconductor chip. By adopting such a method, the solder bumps can be formed with high accuracy, but there is a problem that a considerably long evaporation time is required to form the solder bumps having a desired thickness, for example, a thickness of several tens of μm. there were.

In the solder plating method, after the surface of the semiconductor chip is covered with a resist, the resist at the position where the pump is formed is opened to perform solder plating. However, there is a problem in that the characteristics are likely to be deteriorated due to the mechanical change and the composition deviation of the deposited solder is large.

The solder bumps are formed not only on the semiconductor chip but also on the ceramic substrate, the ceramic package, or the like. Also, in some applications, solder balls are provided as a lump of solder substantially without the use of solder paste, and such solder balls may be used to form solder bumps.
However, heretofore, there has been no method for easily manufacturing solder balls, or there has been a large variation in the size of the solder balls.

[0010]

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method of manufacturing a solder bump which can easily form a solder bump having a desired size. Another object of the present invention is to provide a method of manufacturing a solder ball that can easily form a solder ball having a desired size.

[0011]

A method of manufacturing a solder bump according to the present invention provides a solder ball forming member 10 having a flat surface 10a and a plurality of recesses 12 arranged in a predetermined pattern on the flat surface. The solder particles 16 and the flux 15 are prepared in the recess 12 of the solder ball forming member 10.
Is filled with the solder paste 14, and the solder ball forming member 10 is heated to melt the solder particles 16 contained in the solder paste 14 in each recess 12 to form a solder ball 20 rounded by surface tension. The solder ball 20 heated by bringing the member 22 on which the solder bumps are to be formed relatively close to the solder ball forming member 10 is the same as the pattern of the recess 12 on the member 22 on which the solder bumps are to be formed. Characterized by transferring in a pattern.

[0012]

In the above structure, the recesses of the solder ball forming member have substantially the same shape and volume, so that when the solder paste is filled in the recesses of the solder ball forming member, the volumes of the solder paste in all the recesses are mutually equal. equal. Solder particles are uniformly distributed in the solder paste. When the solder ball forming member is heated to a temperature equal to or higher than the melting point of the solder, the melted solder particles are rounded by the surface tension, and solder balls having substantially the same volume are formed in all the concave portions. The solder balls thus formed are separated from the flux, and only the solder balls can be taken out from each recess. Therefore, when the member on which the solder bumps are to be formed is brought relatively close to the solder ball forming member, the heated solder balls are transferred from the solder ball forming member to the member on which the solder bumps are to be formed in the same pattern as the concave pattern. Will be done. In this way, a solder bump having a desired size according to the volume of the recess of the solder ball forming member is formed at a predetermined position on the member on which the solder bump is to be formed.

Preferably, at least the portion including the flat surface of the solder ball forming member is made of a material having a solder wettability inferior to the solder bump forming portion of the member on which the solder bump is to be formed. As a result, the solder ball is easily transferred from the solder ball forming member to the member on which the solder bump is to be formed.

Preferably, the step of filling the solder paste in the concave portion of the solder ball forming member comprises moving the squeegee along the flat surface of the solder ball forming member. As a result, the solder paste can be filled in the recesses of the solder ball forming member most easily, and the volumes of the solder paste in all the recesses become equal to each other.

Preferably, as one of the features of the transfer of the solder balls, the concave portions of the solder ball forming member are partially formed from the surface of the solder ball forming member by the solder balls formed in the concave portions in the solder ball forming step. It is defined to have a protruding shape. Therefore, when the member on which the solder bumps are to be formed is brought relatively close to the solder ball forming member from the flat surface side, the protruding solder balls adhere to the members on which the solder bumps are to be formed and are easily transferred.

The transfer of the solder balls can also be carried out immediately after forming the solder balls while the solder balls are still heated. Alternatively, the step of cooling the solder ball forming member after the step of forming the solder ball and the step of transferring the solder ball may be included, and the step of transferring the solder ball may include heating the solder ball forming member. . Thereby, the formation of the solder balls and the transfer of the solder balls can be performed in different stages.

Preferably, the recess of the solder ball forming member has a contour shape on the flat surface, a bottom portion, and a depth from the flat surface to the bottom portion, and the size of the contour shape is formed. Larger than the size of the solder ball to be. As a result, the solder ball is shaped so as to partially project from the surface of the solder ball forming member. This can be expressed as that the depth of the recess is smaller than the diameter of the solder ball.

Preferably, the contour shape of the flat surface of the recess is substantially circular or polygonal. Preferably, the recess has a deepest portion at a defined position, and the solder ball is formed at a position corresponding to the deepest portion. The contour shape decreases as the depth of the recess increases. Alternatively, the contour shape gradually decreases as the depth of the recess increases.

Preferably, the solder ball forming member is composed of one plate. Alternatively, the solder ball forming member is composed of a plurality of stacked plates, and at least one of the plurality of plates has a through hole, and the through hole forms a part of the recess. This allows the shape of the recess to change as the depth increases.

Preferably, as another feature of the transfer of the solder ball, the concave portion of the solder ball forming member connects the contour shape of the flat surface, the bottom portion, and the annular surface connecting the flat surface and the bottom portion. A first side wall and a second side plate on which the solder ball forming member is removably overlapped, and the first plate is a through hole that forms the flat surface and the annular side wall of the recess. And the surface of the second plate forms the bottom of the recess. Therefore, the step of transferring the solder balls includes the step of moving the second plate away from the first plate while the solder balls are held by the first plate,
And relatively bringing the member on which the solder bumps are to be formed closer to the surface of the first plate opposite to the flat surface. In this way, the solder ball can be transferred from the surface of the solder ball forming member opposite to the surface having the recess.

In this case, it is preferable that the step of transferring the solder balls further includes a step of extruding the solder balls held by the first plate toward a member on which the solder bumps are to be formed.

Further, in the method for manufacturing a solder ball according to the present invention, a solder ball forming member having a flat surface and a plurality of concave portions arranged in a predetermined pattern on the flat surface is prepared, and the solder ball forming member is prepared. The concave portion of is filled with a solder paste composed of solder particles and flux, and the solder ball forming member is heated to melt the solder particles contained in the solder paste in each concave portion to form a rounded solder ball by surface tension. It is characterized in that the solder ball forming member is cooled and the solder ball is taken out from the concave portion of the solder ball forming member.

Also in this structure, the recesses of the solder ball forming member have substantially the same shape and volume,
Therefore, the volumes of the solder paste filled in the recesses are equal to each other. When the solder ball forming member is heated, the molten solder particles are rounded by the surface tension to form the solder balls having substantially the same volume in all the recesses. The solder balls thus formed are separated from the flux. Therefore, the solder ball forming member can be cooled, the solder balls can be made into individual lumps, and the solder balls can be taken out from each recess. Flux adheres to the outside of the solder ball, but by cleaning it,
Flux can be removed from the solder balls.

[0024]

FIG. 1 is a diagram showing a first embodiment of the present invention.
In the method of manufacturing a solder bump according to the present invention, first, as shown in FIG.
Prepare 0. The solder ball forming member 10 is made of a stainless steel plate (SUS304) having excellent heat resistance, and a plurality of recesses 12 having the same shape are formed on the flat surface 10a.
They are arranged in a matrix with a pitch of 0 μm. Each concave portion 12 has a contour shape of 200 μm in diameter on the surface 10a.
Is circular and has a depth of 50 μm. The shape of the concave portion 12 has a smaller diameter as the depth increases. The recess 12 is formed by cutting the surface 10 a of the solder ball forming member 10.

The solder ball forming member 10 is made of stainless steel,
It is preferably formed of a material having poor solder wettability such as chromium, glass, ceramics or titanium, or the surface thereof is preferably coated with a material having poor solder wettability such as chromium, titanium or SiO ₂ . In the example,
After forming the recess 12, the surface 10a of the solder ball forming member 10 was plated with black chrome (Cr) (abbreviated as BCR).

3 and 4 show a recess 12 formed by etching in a flat surface 10a of a solder ball forming member 10 made of a stainless steel plate. in this case,
A resist (not shown) is applied to the flat surface 10a of the solder ball forming member 10, an opening having a predetermined pattern is provided in the resist by exposure and development, and etching is performed using a ferric chloride solution as an etchant. In FIG. 4, the recesses 12 are provided in a matrix.

FIG. 5 shows an example in which the solder ball forming member 10 is formed by stacking a relatively rigid flat stainless steel plate 10b and a stainless steel plate 10c having a through hole 10d corresponding to the recess 12. FIG. Figure 6
FIG. 3 is a diagram showing an example in which the solder ball forming member 10 is formed by stacking a relatively rigid flat stainless steel plate 10 b and a stainless steel plate 10 c having a rounded through hole 10 e corresponding to the recess 12. is there. In FIGS. 5 and 6, the thickness of the plate 10c corresponds to the depth of the recess 12. This ensures the depth of the recess 12 of the solder ball forming member 10.

As shown in FIGS. 1A and 1B,
After the solder ball forming member 10 is prepared, the solder paste 14 is filled in the recess 12 of the solder ball forming member 10. For this purpose, as shown in (A), the solder paste 14 is placed on the flat surface 10 a of the solder ball forming member 10.

The solder paste 14 comprises solder particles 16 and a flux 15, and the flux 15 contains pine resin, an activator, an organic solvent and the like. The solder particles 16 are arranged in the solder paste 14 at a very high density, so that the solder components are uniformly distributed in the solder paste 14. In order to arrange the solder particles 16 with high density, the solder particles 16 include components of various sizes. For example, in the examples, out of all the solder particles 16, about 60% of particles are 20 μm or less, about 30% of particles of 30 μm or less, and about 5% of particles of 40 μm or less. In addition, the solder particles 1 in the solder paste 14
The ratio of 6 is about 90 weight percent and about 44
Volume percent. The solder was made of Pb-5Sn (melting point: 300 to 314 ° C.). It should be noted that solder having other configurations can also be used. For example,
Sn, Pb, In, Bi, Ag, Zn, Au, Ge, S
A solder containing any one of i and Sb as a constituent element can be used.

The squeegee 1 is used to fill the solder paste 14.
8 is moved along the flat surface 10a of the solder ball forming member 10 in the direction of arrow A, for example. Since the recesses 12 of the solder ball forming member 10 all have the same volume, the solder paste 14 in the recesses 12 also has the same volume. Since the solder particles 16 are substantially evenly distributed in the solder paste 14, the recess 12
The volumes of the solder particles 16 in are also substantially equal.

Next, as shown in FIG. 1C, the solder ball forming member 10 is heated to a temperature equal to or higher than the melting point of the solder (for example, 35).
Heat to 0 ° C). The solder particles 16 in the solder paste 14 are melted and the solder is curled by the surface tension to form one solder ball 20 in each recess 12.
The flux 15 is separated from the solder balls 20. When the solder ball forming member 10 is heated, the solvent in the flux 15 evaporates, and the activator or the like is partially decomposed. Further, in order to cause such evaporation of the solvent, preheating can be performed at a low temperature.

Next, as shown in FIG. 1D, the member 22 on which the solder bumps are to be formed is brought relatively close to the solder ball forming member 10, and the member 22 on which the solder bumps are to be formed is heated. The solder balls 20 are abutted, and the solder balls 20 are transferred from the solder ball forming member 10 to the members 22 on which solder bumps are to be formed. Since the solder ball forming member 10 is formed of or coated with a material having a solder wettability lower than that of the member 22 on which the solder bumps are to be formed, the solder ball forming member 10 includes the solder balls 20.
Is easily transferred (moved) to the member 22 on which the solder bump is to be formed.

In this case, the solder balls 20 remain arranged on the solder ball forming member 10 without changing their arrangement, and therefore the members 22 on which the solder bumps should be formed in the same pattern as the arrangement pattern of the recesses 12. Is transcribed to. In this way, small solder bumps of, for example, 100 μm or less arranged in order on the semiconductor chip could be obtained.

In the embodiment, the member 22 on which the solder bump is to be formed is a semiconductor chip formed as an LSI or VLSI, and is provided with an electrode pad 24 connected to an integrated circuit (not shown). The recesses 12 of the solder ball forming member 10 are provided in the same pattern as the electrode pads 24 of the member 22 on which solder bumps are to be formed. That is, the electrode pad 24 is formed on the silicon (Si) substrate by 300
It is formed with a μm pitch. The electrode pad 24 has a diameter of 100 μm, for example. The electrode pad 24 is formed on a silicon substrate and has a thickness of 1000Å / 5000Å of gold (Au).
And a layer of nickel (Ni) and have excellent solderability.

The member 22 on which the solder bumps are to be formed is properly aligned with the solder ball forming member 10,
The solder balls 20 adhere to the electrode pads 24. Thus, as shown in FIG. 1E, the solder balls 20 become the solder bumps of the member 22.

In the embodiment of FIG. 1, the solder balls 2
At the time of 0 transfer, the member 22 on which the solder bump is to be formed is brought relatively close to the solder ball forming member 10 from the flat surface 10a side. Therefore, in order to enable the transfer of the molten solder ball 20, the solder ball 20 partially protrudes outward from the flat surface 10a and can directly contact the member 22 on which the solder bump is to be formed. Need to be. By defining the shape of the recess 12 as described below, the solder ball 20 partially protrudes from the flat surface 10a to the outside.

As mentioned above, the proportion of the solder particles 16 in the solder paste 14 is about 44 volume percent. Therefore, the solder ball 20 has a flat surface 10a.
In order to partially project outward from the recess 1,
2 must be a relatively shallow groove, and the size of the contour of the recess 12 on the flat surface 10a must be larger than the size of the solder ball 20 to be formed.

FIG. 2 is a diagram showing a second embodiment of the present invention. As in the first embodiment, as shown in FIGS. 2A to 2C, the solder ball forming member 10 is prepared, and the solder paste 14 is applied to the recess 12 of the solder ball forming member 10.
And the solder ball forming member 10 is heated to form the solder balls 20.

Next, as shown in FIG. 2D, after the step of forming the solder balls 20, but before the step of transferring the solder balls 20, the solder ball forming member 10 is cooled, and in some cases, The flux 15 is washed with a solvent. Next, as shown in FIG. 2E, the solder balls 20 are transferred from the solder ball forming member 10 to the members 22 on which solder bumps are to be formed. At this time, the solder ball forming member 10 is heated again.

7 to 10 show a member (semiconductor chip) 22 on which solder bumps thus manufactured are to be formed.
It is a figure which shows the example of the semiconductor package containing. FIG. 7 shows a semiconductor package called a pin grid array (PGA), in which an electrode pattern or electrode pad 42 is provided on the surface of the ceramic substrate 40, and pins 44 extend from the lower surface of the ceramic substrate 40. Electrode pad 42
Is connected to a pin 44, and the PGA can be attached to the printed circuit board by attaching the pin 44 to the printed circuit board, for example.

The member (semiconductor chip) 22 on which the solder bumps are to be formed is placed on the PGA so that the solder bumps (solder balls) 20 are located on the electrode pads 42, and the solder bumps are formed by heating. The (solder ball) 20 is welded to the electrode pad 42. Finally, the resin 46 is used for sealing.

FIG. 8 shows a semiconductor package called a ball grid array (BGA), in which balls 48 are provided instead of the pins 44 on the lower surface of the PGA. Other configurations are similar to those of the PGA of FIG. 7. Ball 48
It is welded to the printed circuit board. Therefore, this ball 48
Is also a type of solder bump, and can be manufactured by the manufacturing method of FIGS. 1 and 2 and a manufacturing method described later. In this case, for example, the recesses 12 having a diameter of 1000 μm and a depth of 500 μm are formed at a pitch of 1270 μm. Sn-37Pb, for example, is used as the solder, and in this case, it is heated to 215 ° C. to form the solder ball 20.

FIG. 9 shows a multi-chip module (MCM)
It is a semiconductor package called. The MCM includes a ceramic substrate 40 similar to the ceramic substrate 40 of FIG. 7 and a plurality of members (semiconductor chips) 22 on which solder bumps are to be formed. Also in this case, the solder bumps (solder balls) 20 are welded to the electrode pads 42 and the resin 4
6 is sealed. The MCM may be composed of a ceramic substrate 40 similar to the ceramic substrate 40 of FIG.

FIG. 10 shows a quad flat package (Q
It is a semiconductor package called FP). The solder bumps (solder balls) 20 of the member (semiconductor chip) 22 on which the solder bumps are to be formed are welded to the leads 50 and sealed with the resin 46.

In FIG. 11, a member (semiconductor chip) 22 on which solder bumps are to be formed is printed circuit board (PWB) 54.
It is a figure which shows the example attached to. The printed circuit board 54 has electrode pads 56 and solder bumps (solder balls).
20 is welded to the electrode pad 56 of the printed circuit board 54. The solder bumps (solder balls) 20 may be provided on the printed circuit board 54.

12 and 13 are views showing an example in which the solder balls 20 are provided on the TAB component 58 to form solder bumps. This TAB component 58 corresponds to the member 22 on which the solder bumps of FIGS.
0 is formed on the TAB component 58 according to the manufacturing method shown in FIGS. The TAB component 58 is a plastic tape 60 having a central opening 60 a with leads 62 attached thereto, and the solder balls 20 are attached to the leads 62. The semiconductor chip 64 having the electrode pad 66 is placed on the lead 62 in the central opening 60a of the TAB component 58 and attached to the TAB component 58 by welding the solder balls 20. The semiconductor chip 64 is sealed with resin 68.

The recess 12 of the solder ball forming member 10 shown in FIGS. 1 to 6 is larger than the diameter of the solder ball 20 in which the contour shape on the flat surface 10a is formed.
And the depth was shallow as a whole. As shown in FIG. 14, the bottom of such a recess 12 tends to be generally flat, and the deepest portion may not be located at a fixed position. In forming the ball 20, the weight of the molten solder is
The solder ball 20 is formed in the deepest part. If the position of the deepest part of the recess 12 is not constant, the position of the solder ball 20 will vary within the recess 12. Therefore, when the solder balls 20 are transferred,
The position of the solder ball 20 is displaced with respect to the electrode pad 24 of the member 22 on which the solder bump is to be formed.

The molten solder ball 20 is used as the electrode pad 2
4 has a property that self-alignment is performed, and the melted solder ball 20 generally adheres to the center of the electrode pad 24. Therefore, even if the position of the solder ball 20 is misaligned, there is no problem in many cases. . However, if the displacement of the solder balls 20 is large, the solder balls 20 may not be properly attached to the electrode pads 24.
Therefore, it is preferable to form the deepest portion in the recess 12. When the recess 12 is formed by machining, the recess 12 is ground so that a predetermined position becomes deep.

FIGS. 15 to 24 are views showing an example in which the deepest portion is intentionally provided in the recess 12 so that the solder ball 20 is formed at the center of the recess 12. 15 and 16
The solder ball forming member 10 on the flat surface 1
0a becomes <100> silicon plate, and the flat surface 10a is anisotropically etched to form the recess 12
Was formed. In the etching, since a resist having a circular opening was used, a conical recess 12 was formed, and the deepest portion D of the bottom was at the center of the recess 12.

In FIGS. 17 and 18, the solder ball forming member 10 is similarly made of a silicon plate, and the flat surface 10a is anisotropically etched. In the etching, since a resist having a square opening was used, a quadrangular pyramidal recess 12 was formed.

In FIGS. 19 and 20, the solder ball forming member 10 is made of a stainless steel plate, and the flat surface 10a is etched twice to form the recess 12. In the first etching, a resist having a small opening was used, and in the second etching, a resist having a larger opening at the same position as the opening in the first etching was used. Therefore, the recess 12 has a two-stage hole, and the deepest part is formed at the center of each recess 12.

In FIGS. 21 and 22, a plurality of stainless steel plates 10f, 10g and 10h on which the solder ball forming member 10 is stacked are used. The first-layer plate 10f is relatively rigid and flat, the second-layer plate 10g has a small hole 10i, and the third-layer plate 10h is a large hole 10j.
have. Therefore, each recess 12 has a small hole 1 at the bottom.
0i and a large hole 10j in the upper part are formed as a two-stage hole, and the deepest part is formed at the center of each recess 12.
The plurality of stainless steel plates 10f, 10g, 10h are integrated by pressure and heat treatment. Alternatively, they can be integrated by projection welding.

In FIG. 23 and FIG. 24, the solder ball forming member 10 is constructed by stacking a stainless steel plate 10k and a plurality of resin films 10l, 10m and 10n. The stainless steel plate 10k is relatively rigid and flat, and the resin films 10l, 10m, and 10n have holes 10o, 10p, and 10q that become larger in order. Each recess 12 is composed of these holes 10o, 10p, 10q.
The deepest part is formed in the central part of the.

In one example for obtaining the configuration of FIGS. 23 and 24, the resin films 10l, 10m, 10n are first individually formed with holes 10o, 10p, 10q and then overlaid. In another example for obtaining the configuration of FIGS. 23 and 24, the resin films 10l, 10m, and 10n are each a photocurable resin, the resin film 10l is first applied, masked, and irradiated with light. . The resin part exposed to light is cured, and the part not exposed to light is melted by washing to form the hole 10o. Next, a resin film 10m is applied and light is similarly irradiated to form a hole 10p. Then, the resin film 1
0n is applied and light is similarly irradiated to form the hole 10q.

FIG. 25 is a diagram showing a third embodiment of the present invention. As shown in (A), first, a solder ball forming member 10 having a recess 12 is prepared. In this embodiment, the solder ball forming member 10 includes first and second stainless steel plates 10r, 1 removably stacked together.
It consists of 0s. The first plate 10r has a flat surface 10a and a through hole 10t forming an annular side wall of the recess 12. The second plate 10s is substantially flat and forms the bottom of the recess 12.

As in the previous embodiment, the solder ball forming member 10 is formed of a material having poor solder wettability, or the surface thereof is coated with a material having poor solder wettability. Next, as shown in FIGS. 25A and 25B, the squeegee 18 is used to fill the recess 12 with the solder paste 14.

Next, as shown in FIG. 25C, the solder ball forming member 10 is heated to a temperature equal to or higher than the melting point of the solder,
The solder particles 16 in the solder paste 14 are melted, and the solder is curled by surface tension to form one solder ball 20 in each recess 12. The flux 15 is separated from the solder balls 20.

In this embodiment, it is not necessary to form the recess 12 so that the solder ball 20 projects from the flat surface 10a. Therefore, the solder ball 20 having a desired size can be formed by increasing the depth of the recess 12 without increasing the contour shape of the flat surface 10a of the recess 12. Therefore, the recesses 12 can be arranged at a smaller pitch.

The solder ball 20 is in contact with the annular side wall portion of the recess 12 (the through hole 10t of the first plate 10r), and when the temperature decreases, the solder ball 20 moves to the annular side wall portion of the recess 12 (the first plate 10r). It is held in the through hole 10t) of 10r.
Therefore, as shown in FIG. 25D, the second plate 10s of the solder ball forming member 10 is moved away from the first plate 10r. When the flux 15 is washed, the solder balls 2
Only 0 is held in the annular side wall of the recess 12 (the through hole 10t of the first plate 10r).

Therefore, as shown in FIG.
A member (semiconductor chip) 22 on which a solder bump is to be formed is brought relatively close to the surface of the first plate 10r opposite to the flat surface 10a, and the member 22 on which a solder bump is to be formed is heated to the solder ball 20. Hit against. The first plate 10r of the solder ball forming member 10 is heated at an appropriate time. Therefore, the solder balls 20 are transferred from the first plate 10r of the solder ball forming member 10 to the member 22 on which solder bumps are to be formed.

Furthermore, a push plate 72 is used. The pressing plate 72 has a protrusion 74 suitable for entering into the concave portion 12 of the solder ball forming member 10 (the through hole 10t of the first plate 10r). By moving toward the plate 10r, the protrusions 74 push out the solder balls 20 and press the solder balls 20 against the member (semiconductor chip) 22 on which the solder bumps are to be formed, so that the solder balls 20 are securely attached. Make it a solder bump.

FIG. 26 is a diagram showing a fourth embodiment of the present invention. This embodiment shows a method of manufacturing the solder balls 20. As shown in FIG. 26A, first, the solder ball forming member 10 having the recess 12 is prepared.
Similar to the previous embodiment, the solder ball forming member 10 comprises first and second stainless steel plates 10r, 10s removably superposed. The first plate 10r has a flat surface 10a and a through hole 10t forming an annular side wall of the recess 12. The second plate 10s is substantially flat and forms the bottom of the recess 12.

As in the previous embodiment, the solder ball forming member 10 is formed of a material having poor solder wettability, or its surface is coated with a material having poor solder wettability. Next, as shown in FIGS. 26A and 26B, the recess 12 is filled with the solder paste 14 using the squeegee 18.

Next, as shown in FIG. 26C, the solder ball forming member 10 is heated to a temperature equal to or higher than the melting point of the solder,
The solder particles 16 in the solder paste 14 are melted, and the solder is curled by surface tension to form one solder ball 20 in each recess 12. The size of the solder ball 20 is formed according to the shape of the recess 12. If the shape of the recess 12 is constant, the size of the solder ball 20 formed is constant.

Therefore, the second solder ball forming member 10
The first plate 10r of the solder ball forming member 10 is cooled away from the first plate 10r of the solder ball forming member 10 by the flux 1
Wash 5. In this case, the shape of the recess 12 of the solder ball forming member 10 is set so that the solder ball 20 in a lump state that has been cooled is not tightly held in the recess 12.

Therefore, when the flux 15 is washed, the solder balls 20 roll out from the recesses 12, and therefore the solder balls 20 shown in FIG.
As shown in (E), the solder balls 20 are placed in a container 76.
To collect. If the solder balls 20 do not easily roll out of the recess 12, a push plate 72 is used. The pressing plate 72 has a protrusion 74 suitable for entering into the concave portion 12 of the solder ball forming member 10 (the through hole 10t of the first plate 10r). The protrusion 74 pushes the solder ball 20 by moving it toward the plate 10r.

FIG. 27 is a diagram showing a fifth embodiment of the present invention. This embodiment also shows a method of manufacturing the solder balls 20. As shown in FIG. 27A, first, the solder ball forming member 10 having the recess 12 is prepared.
The solder ball forming member 10 is formed of a material having poor solder wettability, or its surface is coated with a material having poor solder wettability. Next, as shown in FIGS. 27A and 27B, the squeegee 18 is used to solder 1
4 is filled in the recess 12.

Next, as shown in FIG. 27C, the solder ball forming member 10 is heated to a temperature equal to or higher than the melting point of the solder,
The solder particles 16 in the solder paste 14 are melted, and the solder is curled by surface tension to form one solder ball 20 in each recess 12. The size of the solder ball 20 is formed according to the shape of the recess 12. If the shape of the recess 12 is constant, the size of the solder ball 20 formed is constant.

Therefore, the solder ball forming member 10 is cooled and the flux 15 is washed. Then, the solder balls 20 roll out from the recesses 12, so that the solder balls 20 are collected in the container 76 as shown in FIG.
In this case, since the shape of the recess 12 is larger and the depth thereof is smaller than that of the solder ball 20, the solder ball 20 can be easily taken out.

Solder ball 20 manufactured in this way
Can be used to form solder bumps for any electrical / electronic device. For example, it can be used to form the BGA ball 48 shown in FIG. Although the solder balls 20 may be arranged one by one in the BGA, the solder balls 20 may be collectively arranged in the BGA by using a ball arranging jig 80 as shown in FIG. 28, for example. The ball placement jig 80 has holes 82 in a pattern corresponding to the positions of the balls 48 of the BGA, and the size of the holes 82 is slightly larger than the size of the solder balls 20.
Therefore, when the ball placement jig 80 is placed on the BGA and the solder balls 20 are rolled on the surface of the ball placement jig 80, the solder balls 20 enter the holes 82 one by one. Therefore, when the ball arranging jig 80 or the BGA is heated, the solder balls 20 are welded to the BGA. According to the present invention, the solder ball 20 having a small variation in size can be obtained, and thus such a ball arranging jig 80 can be used.

[0071]

As described above, according to the present invention,
The solder bumps can be formed by a relatively simple method, and thus a bump manufacturing method suitable for mass production can be established.

[Brief description of drawings]

FIG. 1 is a diagram showing a first embodiment of the present invention.

FIG. 2 is a diagram showing a second embodiment of the present invention.

FIG. 3 is a cross-sectional view showing another example of the solder ball forming member.

FIG. 4 is a plan view of the solder ball forming member of FIG.

FIG. 5 is a cross-sectional view showing another example of the solder ball forming member.

FIG. 6 is a cross-sectional view showing another example of the solder ball forming member.

FIG. 7 is a cross-sectional view showing an example of PGA using solder bumps.

FIG. 8 is a cross-sectional view showing an example of a BGA using solder bumps.

FIG. 9 is a cross-sectional view showing an example of an MCM using solder bumps.

FIG. 10 is a cross-sectional view showing an example of QFP using solder bumps.

FIG. 11 is a cross-sectional view showing an example of a printed circuit board using solder bumps.

FIG. 12 is a cross-sectional view showing an example of a TAB component using solder bumps.

13 is a sectional view showing an example in which a semiconductor chip is attached to the TAB component of FIG.

FIG. 14 is a cross-sectional view showing an extreme example of solder ball formation positions.

FIG. 15 is a cross-sectional view showing an example of a solder ball forming member having a deepest portion.

16 is a plan view of the solder ball forming member of FIG.

FIG. 17 is a plan view showing another example of the solder ball forming member.

FIG. 18 is a cross-sectional view taken along the line XVIII-XVIII in FIG.

FIG. 19 is a plan view showing another example of the solder ball forming member.

20 is a cross-sectional view taken along the line XX-XX of FIG.

FIG. 21 is a plan view showing another example of the solder ball forming member.

22 is a cross-sectional view taken along the line XXII-XXII of FIG.

FIG. 23 is a plan view showing another example of the solder ball forming member.

24 is a cross-sectional view taken along the line XXIV-XXIV of FIG. 23.

FIG. 25 is a diagram showing a third embodiment of the present invention.

FIG. 26 is a diagram showing a fourth embodiment of the present invention.

FIG. 27 is a diagram showing a fifth embodiment of the present invention.

FIG. 28 is a partial cross-sectional perspective view showing a ball placement axis.

[Explanation of symbols]

 10 ... Solder ball forming member 12 ... Recessed portion 14 ... Solder paste 15 ... Flux 16 ... Solder particles 18 ... Squeegee 20 ... Solder ball 22 ... Member to form solder bump

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Nobuo Kamehara, Nobuo Kamehara, 1015 Kamiodanaka, Nakahara-ku, Kawasaki, Kanagawa, Fujitsu Limited (72) Inventor, Rikuro Sonoda, 1015, Kamiodanaka, Nakahara-ku, Kawasaki, Kanagawa 72) Inventor Ichiro Yamaguchi 1015 Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture, Fujitsu Limited (72) Inventor Kazuhiko Mito, 1015, Kamikodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Fujitsu Limited (72) Inventor, Yukiki Otake Kanagawa Prefecture 1015 Kamiodanaka, Nakahara-ku, Kawasaki-shi Fujitsu Limited (72) Inventor Junichi Kasai 1015 Kamikodanaka, Nakahara-ku, Kawasaki-shi, Kanagawa Kanagawa Prefecture (72) Masataka Mizukoshi 1015 Kamedota-anaka, Nakahara-ku, Kawasaki-shi, Kanagawa Within Fujitsu Limited

Claims (20)

[Claims] 1. A solder ball forming member (10) having a flat surface (10a) and a plurality of recesses (12) arranged in a predetermined pattern on the flat surface is prepared. Solder particles (1
6) and a flux (15) are filled with a solder paste (14), and the solder ball forming member (10) is heated to form each recess (1).
Solder particles (16) contained in the solder paste (14) in 2) are melted to form a rounded solder ball (20) due to surface tension, and a solder ball is formed on a member (22) on which a solder bump is to be formed. Manufacture of solder bumps, characterized in that a solder ball (20) heated relatively close to a member (10) is transferred from the solder ball forming member (10) to a member (22) on which the solder bump is to be formed. Method. 2. A member (22) for forming a solder bump.
Is a semiconductor chip, a semiconductor package, a printed circuit board, or a TAB component, and at least a portion of the solder ball forming member (10) including the flat surface (10a) forms the solder bump. Power element (22)
The solder bump manufacturing method according to claim 1, wherein the solder bump forming portion is made of a material having a solder wettability inferior to that of the solder bump forming portion. 3. A member (22) for forming a solder bump.
Is an IC chip, and the method for manufacturing a solder bump according to claim 1, wherein 4. A member (22) for forming a solder bump.
Is a lead on a TAB tape, The method for manufacturing a solder bump according to claim 1, wherein 5. The step of filling the solder paste (14) into the recess (12) of the solder ball forming member (10) comprises:
The method of claim 1, wherein the squeegee (18) is moved along the flat surface of the solder ball forming member. 6. The surface (10a) of the solder ball forming member (10) is obtained by forming the recess (12) of the solder ball forming member (10) in the solder ball forming step. The step of transferring the solder balls (20) is determined such that the members (22) on which the solder bumps are to be formed are relatively projected from the flat surface side to the solder ball forming members. The method for manufacturing a solder bump according to claim 1, characterized in that the method comprises: 7. A step of cooling the solder ball forming member (10) after the step of forming the solder ball and before the step of transferring the solder ball, wherein the step of transferring the solder ball includes forming the solder ball. A method according to claim 6, characterized in that it comprises heating at least one of the member (10) and the member (22) on which the solder balls are to be formed. 8. The concave portion (12) of the solder ball forming member (10) has a contour shape on the flat surface (10a), a bottom portion, and a depth from the flat surface to the bottom portion. Solder balls (2
The solder bump manufacturing method according to claim 6, wherein the solder bump is larger than 0). 9. The method of manufacturing a solder bump according to claim 8, wherein the depth of the recess (12) is smaller than the diameter of the solder ball. 10. The flat surface (10) of the recess (12).
The method of manufacturing a solder bump according to claim 6, wherein the contour shape in a) is substantially circular. 11. The flat surface (10) of the recess (12).
The solder bump manufacturing method according to claim 6, wherein the contour shape in a) is substantially polygonal. 12. The method of manufacturing a solder bump according to claim 8, wherein the recessed portion (12) has a deepest portion at a defined position. 13. The contour shape decreases as the depth of the recess (12) increases.
The method for manufacturing a solder bump according to. 14. The method for manufacturing a solder bump according to claim 12, wherein the contour shape gradually decreases as the depth of the recess (12) increases. 15. The method of manufacturing a solder bump according to claim 1, wherein the solder ball forming member (10) is composed of a single plate. 16. The solder ball forming member (10) comprises a plurality of stacked plates, at least one of the plurality of plates having a through hole, the through hole forming a part of the recess. The method of manufacturing a solder bump according to claim 1, wherein 17. The solder ball forming member (10) has the concave portion (12) having a contour shape on the flat surface (10a), a bottom portion, and an annular side wall portion connecting the flat surface and the bottom portion. And the solder ball forming member is composed of first and second plates removably superposed, the first plate having the flat surface and a through hole forming an annular side wall of the recess. The step of transferring the solder balls by forming the bottom of the recess by the surface of the second plate, the step of transferring the second plate to the first plate with the solder balls held by the first plate is performed. 2. The method according to claim 1, further comprising a step of moving away from the plate, and a step of bringing a member on which the solder bumps are to be formed relatively close to a surface of the first plate opposite to the flat surface. Of manufacturing solder bumps of. 18. The step of transferring the solder balls further comprises the step of extruding the solder balls held on the first plate toward a member on which the solder bumps are to be formed. The method for manufacturing a solder bump according to. 19. A solder ball forming member (10) having a flat surface (10a) and a plurality of recesses (12) arranged in a predetermined pattern on the flat surface is prepared. The concave portion (12) of (10) is filled with a solder paste (14) composed of solder particles (16) and a flux (15), and the solder ball forming member (10) is heated so that each concave portion (1)
Solder particles contained in the solder paste (14) of 2) are melted and rounded by the surface tension.
0) is formed, the solder ball forming member (10) is cooled, and the solder ball (20) is formed into a solder ball forming member (10).
A method of manufacturing a solder ball, characterized in that the solder ball is taken out from the concave portion (12). 20. The squeegee (18) of claim 19, wherein the step of filling the solder ball forming member with the solder paste in the concave portion moves the squeegee (18) along the flat surface of the solder ball forming member. Solder ball manufacturing method.