Semiconductor device manufacturing: process – Coating with electrically or thermally conductive material – To form ohmic contact to semiconductive material
Reexamination Certificate
1999-08-17
2002-06-18
Potter, Roy (Department: 2822)
Semiconductor device manufacturing: process
Coating with electrically or thermally conductive material
To form ohmic contact to semiconductive material
C438S613000, C438S108000
Reexamination Certificate
active
06406989
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to a semiconductor device having a flip chip structure, a method of fabricating the same, and a wiring board used for the same.
2. Description of the Related Art
In order to fabricate electronic devices in a smaller size, electronic components have been made smaller and smaller with high performance and high integration being maintained.
A semiconductor device, one of electronic devices, is generally, entirely packaged with resin. However, in these days, a bare chip, which is a chip having no resin package, has been incorporated directly into a wiring board to thereby fabricate a semiconductor chip in a smaller size.
One of examples of a semiconductor device employing a bare chip is illustrated in FIG.
4
. The illustrated semiconductor device is comprised of a semiconductor pellet
1
, a wiring board
5
to be coupled to the semiconductor pellet
1
, and a resin layer
8
sandwiched between the semiconductor pellet
1
and the wiring board
5
for enhancing engagement therebetween.
The semiconductor pellet
1
is comprised of a semiconductor substrate
2
including a plurality of semiconductor elements (not illustrated) which are interconnected to one another to thereby constitute an electronic circuit, an insulating film (not illustrated) formed on a principal surface of the semiconductor substrate
2
, underlying electrodes
3
formed in regions where the insulating film is partially removed, and making electrical connection with the semiconductor elements, and bump electrodes
4
formed on the underlying electrodes
3
. The underlying electrodes
3
are generally made of aluminum. The bump electrodes
4
are formed by deposition of gold layers by plating, evaporation and so on, or by ball-bonding method wherein a gold wire is molten at a tip end thereof to thereby form a gold ball, which is in turn collapsed on an electrode with a tip end of capillary, and then the gold wire is cut so that the collapsed gold ball leaves on the electrode.
The wiring board
5
is comprised of an insulating substrate
6
made of insulating material such as resin and ceramics, and pad electrodes
7
formed on the insulating substrate
6
in facing relation with the bump electrodes
4
formed on the semiconductor substrate
2
. The pad electrodes
7
are electrically connected to an electrically conductive pattern (not illustrated) which is covered with a resist film (not illustrated) so that necessary portions are exposed outside, and which is electrically connected to other electronic components and/or external connection terminals to thereby constitute an electronic circuit. The electrically conductive pattern is formed generally by etching copper foil deposited on the insulating substrate
6
. The pad electrodes
7
are plated with nickel for preventing corrosion of the copper foil, and further plated with gold for enhancing electrical connection with the electrically conductive pattern.
The semiconductor pellet
1
and the wiring board
5
are connected to each other through the resin layer
8
. First, the semiconductor pellet
1
and the wiring board
5
are overlapped each other so that the bump electrodes
4
and the pad electrodes
7
are engaged to each other. Then, the semiconductor pellet
1
and the wiring board
5
are pressurized to thereby cause the bump electrodes
4
into plastic deformation. Then, resin is introduced into a space between the semiconductor pellet
1
and the wiring board
5
. The semiconductor pellet
1
and the wiring board
5
are kept under pressure due to contraction of the resin layer
8
which occurs when the resin cures. The resin layer
8
may be formed on the wiring board
5
before the semiconductor pellet
1
and the wiring board
5
are overlapped each other, or may be formed by introducing resin between the semiconductor pellet
1
and the wiring board
5
after they have been overlapped each other. With the semiconductor pellet
1
and the wiring board
5
kept under pressure, the semiconductor pellet
1
or the wiring board
5
is heated at about 200° C. for a few minutes in order to facilitate the resin layer
8
to cure.
In the illustrated semiconductor device, since the semiconductor pellet
1
is naked, namely, not packaged entirely with resin, a height measured from a surface of the insulating substrate
6
to a top surface of the semiconductor pellet
1
can be kept within 100 &mgr;m, which satisfies the requirement of fabricating a semiconductor device in a smaller size.
Japanese Unexamined Patent Publication No. 5-166881 has suggested a semiconductor device similar to the above-mentioned one. In the suggested semiconductor substrate, bump electrodes are formed on a semiconductor pellet by golden ball bonding method, and pad electrodes are connected to the bump electrodes through solder having a low melting point. Namely, the semiconductor pellet is mechanically and electrically connected to a wiring board without a resin layer such as the resin layer
8
illustrated in FIG.
1
.
Japanese Unexamined Patent Publication No. 7-153796 has suggested a semiconductor device including bump electrodes formed on a semiconductor substrate by golden ball bonding method, and pad electrodes formed on a wiring board so that the pad electrodes are formed with a contact hole having a reducing diameter towards an opening thereof After the bump electrode is partially inserted into the pad electrode, the bump electrode is deformed to thereby keep the bump electrode fixed in the hole.
In the above-mentioned semiconductor devices illustrated in FIG.
1
and suggested in the above-mentioned Japanese Unexamined Patent Publications, the semiconductor pellet and the wiring board thermally expands due to heat generated in the semiconductor device while the semiconductor device is in operation and/or externally transferred heat. Hence, if there is a big difference in thermal expansion coefficient between the semiconductor pellet and the wiring board, there would be generated a stress in the bump and pad electrodes having been firmly coupled to each other.
In the semiconductor device suggested in the firstly mentioned Japanese Unexamined Patent Publication, the bump and pad electrodes are firmly connected to each other by means of solder having a low melting point. Hence, if a stress generated due to a difference in thermal expansion coefficient between the semiconductor pellet and the wiring board concentrates on a contact portion of the semiconductor pellet and the wiring board, the solder might be cracked with the result of degradation of electrical contact between them.
In the semiconductor device suggested in the secondly mentioned Japanese Unexamined Patent Publication, the bump electrodes are partially inserted into a reverse-tapered contact hole of the pad electrodes, and then the bump electrodes are deformed to thereby fill the contact hole with the bump electrodes, which ensures electrically and mechanically firm connection between the semiconductor pellet and the wiring board. However, even though the bump and pad electrodes are firmly connected to each other, if a stress generated due to a difference in thermal expansion coefficient between the semiconductor pellet and the wiring board concentrates on a contact portion of the semiconductor pellet, the bump electrodes might be peeled off the semiconductor substrate. Thus, it was impossible to completely prevent degradation of electrical connection between the semiconductor pellet and the wiring board.
Thermal expansion and thermal contraction as mentioned above occur each time when a semiconductor device starts and finishes its operation. Hence, the above-mentioned problems caused by a difference in thermal expansion coefficient between a semiconductor pellet and a wiring board are serious problems in electronic devices which frequently repeats start-up and stop in operation.
To the contrary, the resin layer
8
connects an entire surface of the semiconductor pellet
1
to the wiring board
5
in the semiconductor device illustrat
Hutchins, Wheeler & Dittmar
NEC Corporation
Potter Roy
LandOfFree
Method of fabricating semiconductor device with bump electrodes 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 fabricating semiconductor device with bump electrodes, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of fabricating semiconductor device with bump electrodes will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2894559