Metal fusion bonding – Process – Plural joints
Reexamination Certificate
1999-07-30
2001-06-19
Ryan, Patrick (Department: 1725)
Metal fusion bonding
Process
Plural joints
C228S123100, C228S180220
Reexamination Certificate
active
06247637
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a method of integrated circuit assembly before encapsulation. More specifically, the present invention relates to soldering steps of such a method.
2. Discussion of the Related Art
In a conventional integrated circuit manufacturing method, each chip, resulting from the dicing of a semiconductive wafer on which are formed several chips, must be individually assembled on a heat sink. Each chip may be assembled on a connection grid, itself assembled on a heat sink, the sink and the connection grid being then separated by an isolator, for example, made of ceramic.
Each of the different elements (sink, ceramic, connection grid, chip) must be attached to at least another one of these elements. For this purpose, solderings have to be made between a first and a second element.
FIGS. 1A and 1B
illustrate a conventional soldering method. An upper element
11
(for example, a chip) is desired to be attached to a second lower element
12
(for example, a grid) by means of a solder paste
13
. Initially, as illustrated in
FIG. 1A
, first element
11
is laid on solder paste
13
. The distance separating the first and second elements
11
,
12
thus corresponds to thickness T
1
of deposited paste
13
, for example, about 40 &mgr;m. Such a paste is typically formed of balls
14
of fusible metals such as lead, tin, and silver in various combinations.
In the automatic deposition of the different assembly elements (chip, ceramic, grid), a mechanical pressure is exerted vertically from top to bottom, this pressure being generally high due to the absence of any clearance in the grasping system: the absence of any clearance is compulsory for a good positioning accuracy upon deposition.
This pressure “crushes” the joint before the fusion, the two elements to be soldered being then separated from each other by a minimum distance equal to the maximum diameter of the paste particles, that is, about 40 &mgr;m. Another disadvantage of this pressure upon deposition is to expel the paste towards the outside of the surface to be soldered.
Then, paste
13
is brought to the fusion point while exerting, on first element
11
, a light mechanical pressure homogeneously distributed on its surface.
FIG. 1B
illustrates the assembly obtained by such a method. Conventionally, first and second elements
11
,
12
are now only separated by a soldering joint (hard soldering)
16
of a thickness T
2
, for example, on the order of 15 &mgr;m, substantially one third of initial thickness T
1
of paste
13
.
A disadvantage of this method is linked to the very small final thickness T
2
, especially as compared to the contact surfaces involved, which are relatively large. Indeed, in the operation at high temperatures and during temperature variations, the heat coefficient difference between the two elements of different nature attached by soldering, combined with the small thickness of the soldering, can cause disconnection of the two elements. Such disconnection results in malfunctions of the circuit in which it occurs.
Thickness T
2
of the obtained soldering depends on the thickness T
1
of initially deposited solder paste. It has already been provided to increase thickness T
1
, by increasing the amount of deposited solder paste. But the high pressure during the deposition of the elements expels the paste outside the surface to be soldered, which results in new disadvantages, especially a problem of alignment of the elements, which tend to float.
It should be noted that these problems also arise when a soldering joint is created between a ceramic pad and a connection grid and between this pad and a sink. Indeed, if elements like the connection grid and the sink appear, upon assembly, in the form of strips of linked elements, the chips and ceramic pads are however individualized before this assembly and are thus not held laterally.
SUMMARY OF THE INVENTION
The present invention aims at overcoming the disadvantages of conventional methods.
Especially, the present invention aims at providing a novel method of soldering two integrated circuit elements which results in a strong solder bond, even under high operating temperatures and for large temperature variations.
To achieve these and other objects, the present invention provides for controlling the thickness of the soldering joint.
More specifically, the present invention provides a method of integrated circuit assembly before encapsulation including at least one step of soldering, under mechanical pressure, a first element on a second element, including temporarily maintaining a predetermined spacing, at least partially without solder paste, between the surfaces to be assembled of the first and second elements.
According to an embodiment of the present invention, the method includes using a solder paste comprising balls of a diameter greater than the individual diameter of other components of the deposited solder paste, to avoid filling the predetermined spacing defined by the ball diameter.
According to an embodiment of the present invention, the balls have a diameter much greater than the final thickness desired for the soldering.
According to an embodiment of the present invention, the ratio between the balls and the individual diameter of the other components is greater than 5, preferably on the order of 10.
According to an embodiment of the present invention, the balls have a diameter ranging between 0.1 and 0.6 mm, preferably on the order of 0.4 mm.
According to an embodiment of the present invention, the balls of greater diameter are formed of the same alloy as the other components of the solder paste.
According to an embodiment of the present invention, the method comprises the steps of depositing a layer of solder paste on the second element, depositing the balls by means of needles at the tips of which they are maintained before dispensing by solder paste.
According to an embodiment of the present invention, the needle tips are dipped in the solder paste layer, then in a bed of balls before being dipped again in the paste layer where the balls remain by capillarity.
According to an embodiment of the present invention, the first element is a ceramic pad and the second element is a metal heat sink.
According to an embodiment of the present invention, the first element is a heat sink and the second element is a connection grid.
The foregoing objects, features and advantages of the present invention will be discussed in detail in the following non-limiting description of specific embodiments in connection with the accompanying drawings.
REFERENCES:
patent: 3591839 (1971-07-01), Evans
patent: 5251806 (1993-10-01), Agarwala et al.
patent: 5361973 (1994-11-01), Ishii et al.
patent: 5385290 (1995-01-01), Degani
patent: 5425493 (1995-06-01), Interrante et al.
patent: 5667132 (1997-09-01), Chirovsky et al.
patent: 5868887 (1999-02-01), Sylvester et al.
patent: 32 08 604 (1983-09-01), None
French Search Report from Application No. 9809946, filed Jul. 30, 1998.
Patent Abstracts of Japan, vol. 006, No. 151 (E-124), Aug. 11, 1982, & JP 57 072357 A May 6, 1982.
Galanthay Theodore E.
Johnson Jonathan
Morris James H.
Ryan Patrick
STMicroelectronics S.A.
LandOfFree
Method of integrated circuit assembly does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of integrated circuit assembly, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of integrated circuit assembly will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2505620