Method of forming a metal fuse on semiconductor devices

Details Method of forming a metal fuse on semiconductor devices Method of forming a metal fuse on semiconductor devices

2002-12-18

2004-12-28

Nelms, David (Department: 2818)

Semiconductor device manufacturing: process

Coating with electrically or thermally conductive material

To form ohmic contact to semiconductive material

C438S132000, C438S281000, C438S618000

Reexamination Certificate

active

06835642

ABSTRACT:

FIELD OF THE INVENTION
This invention relates to a method of forming a metal fuse, and more particularly to a method of forming a metal fuse on or in a semiconductor device.
BACKGROUND OF THE INVENTION
It is known to use fuses for circuit repair of semiconductor devices. For example, as the memory device or the device with an embedded memory, the defective memory cells can be replaced by blowing the related fuses with the redundancy row or column of the cells. So the yield of the memory devices can be improved. Also, logic devices can be repaired or reconfigured by blowing such fuses. For example, it is common to initially fabricate a generic logic chip having a large number of interconnected logic gates. Thereafter, in a final processing step, the chip is customized to perform the desired circuitry by blowing fuses.
Conventional metal fuses are formed on the penultimate, antepenultimate or deeper layer. The thickness of the oxide remaining over the fuse links is difficult to control using etching technology, particularly in processes wherein the devices are manufactured with thinner and thinner layers. The thick oxide remaining over the fuse causes at least two problems. The first problem is that a higher laser energy is needed to penetrate the remaining oxide in order to cut the fuse links. The higher laser energy may result in micro-cracking of the inter-metal dielectric layer, so that the reliability of the device is decreased. The second problem is that the remaining fuse link causes the failure of the laser repair when an insufficient amount of laser energy is utilized. Further, moisture and other contaminants can diffuse through the deep opening in such devices in the area where the fuse is located.
FIGS. 1A-B
illustrate a prior art method of forming a conventional copper metal fuse and blowing the same.
FIG. 1A
is a cross sectional view of a conventional copper metal fuse semiconductor device. The copper metal fuse semiconductor device is provided by forming a conductive layer
22
, such as a polysilicon layer, above the semiconductor substrate
10
and an isolation oxide layer
20
. Then a first inter-level dielectric (ILD) layer
30
is formed and covers the entire substrate. Then an electrically conductive plug
32
is formed inside the first ILD layer
30
. Thereafter, a copper metal conductive layer (first metallization layer)
34
is formed inside the first ILD layer
30
and makes electrical contact with the conductive plug
32
.
Next, a first inter-metal dielectric (IMD) layer
40
is formed covering the first metallization layer
34
and the first ILD layer
30
. Then a conductive plug
42
is formed inside the first IMD layer
40
. Thereafter, a second metallization layer
44
is formed inside the first IMD layer
40
and makes electrical contact with the conductive plug
42
.
Next, a second IMD layer
50
is formed covering the second metallization layer
44
and the first IMD layer
40
. Then a conductive plug
52
is formed inside the second IMD layer
50
. Thereafter, a third metallization layer
54
is formed inside the second IMD layer
50
.
Next, a third IMD layer
60
, conductive plug,
62
and fourth metallization layer
64
are made in a similar manner as described above. Likewise, a fourth IMD layer
70
, conductive plug
72
and fifth metallization layer
76
are made in a similar manner as described above.
Next, a first passivation layer
92
such as silicon dioxide is formed over the fourth IMD layer
70
and fourth metallization layer
74
. A second passivation layer
94
, such as silicon nitride, may also be formed over the first passivation layer
92
. Thereafter, conventional photolithography and etching techniques are used to pattern the passivation layers
92
and
94
and to open a fuse window
96
. The IMD layer
60
, typically silicon dioxide, is etched back over the fuse
56
to leave a dielectric layer
66
over the fuse
56
as shown in FIG.
1
A.
Next the electrical probe test is performed to decide if the device cells or circuits need to be repaired. Thereafter, a laser beam
97
is emitted through the opening of the fuse window
96
and penetrates the remaining portion of the IMD layer (silicon dioxide)
66
to perform the laser repair. Thereafter, as shown in
FIG. 1B
, the fuse
56
is cut open by the laser beam. An opening
98
exposes the IMD layer
50
after the laser repair.
The fuse
56
is formed with the same mask as the conductive layer (third metallization layer)
54
so that thickness of the fuse
56
is the same as the conductive layer
54
. A thinner fuse cannot be produced using this prior art method.
In the conventional method of fabricating such a fuse, as shown in the prior art
FIGS. 1A-B
, the fuse
56
is positioned deep below the surface of the device. Therefore, the laser energy must be substantially high to implement the laser repair. Still further, when the fuse
56
is positioned too deep in the structure, it is difficult for the laser beam to reach a focal point without part of the laser beam being dispersed. Hence, a substantial amount of the laser power is wasted. Typically, in response, a higher laser power is applied in an effort to provide a higher repair rate. However, turning up the laser power can easily damage part of the device area, for example by causing micro-cracking, and thus reduces the reliability of the process. Because of the very narrow window provided when the fuse is located in a position very deep within the device, it becomes difficult to vaporize the remaining oxide
66
. When the thickness of the remaining oxide
66
is too thick, a greater amount of laser energy is required to blow the fuse and it is easy to cause the micro-cracking. However, if a lower laser energy is utilized to prevent micro-cracking, the fuse may not be sufficiently or completely cut. As a consequence, the laser energy window is very narrow in these prior art processes and devices. The present invention provides an improved method of forming a fuse on a semiconductor device, and in one embodiment forming a fuse on a semiconductor device produced using copper metallization techniques.
SUMMARY OF THE INVENTION
One embodiment of the invention includes a method of forming a metal fuse on the top metal conductive layer of a semiconductor device. Generally, the top metal conductive layer is thicker than the other metal conductive layers (metallization layers) in a semiconductor device. The present invention provides a method to reduce the thickness of the top metal fuse. In one embodiment, a specific additional mask is applied to form the metal fuse to reduce the thickness of the fuse. The method also includes forming a fuse window opening that is very shallow in the semiconductor device. The shallower opening allows for better control and removal of the remaining oxide left over the fuse during a fuse burning laser process. The thinner fuse and the thinner remaining oxide reduce the amount of laser energy required to vaporize the oxide and to cut the fuse. The location of the fuse also greatly enlarges the laser energy window that can be utilized to make laser repairs. The larger energy window results in a higher laser repair success ratio even if some deviation in the fabrication process occurs. Furthermore, device micro-cracking abnormality caused by using larger amounts of laser energy can be avoided. The prior art tendency to leave a metal fuse link as a result of using insufficient laser energy is also avoided.
Another embodiment of the invention includes a semiconductor device comprising:
a silicon based substrate, and a metallization layer overlying the silicon based substrate and a fuse portion, the metallization layer and the fuse portion being received in a dielectric layer, and the metallization layer having a thickness of at least 9000 angstroms, and the fuse portion having a thickness less than 4500 angstroms.
Another embodiment of the invention includes a method of making a semiconductor device having a thin fuse portion comprising: forming a first mask over a semiconductor device having a firs

Affiliated with

Also associated with

Rating

Method of forming a metal fuse on semiconductor devices 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 forming a metal fuse on semiconductor devices, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of forming a metal fuse on semiconductor devices will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3325912