Semiconductor device manufacturing: process – Making field effect device having pair of active regions... – Having insulated gate
Reexamination Certificate
2002-04-19
2004-02-10
Tsai, Jey (Department: 2812)
Semiconductor device manufacturing: process
Making field effect device having pair of active regions...
Having insulated gate
C438S294000
Reexamination Certificate
active
06689666
ABSTRACT:
TECHNICAL FIELD
The present claimed invention generally relates to semiconductors. More specifically, the present claimed invention relates to a method of forming a semiconductor device.
BACKGROUND ART
Various techniques known in the art can be used to fabricate a semiconductor device such as an integrated circuit or processor. In general, these techniques typically involve applying a layer of material to an underlying substrate or over a preceding layer, and then selectively removing the material using an etch process. Using these techniques, the components of a semiconductor device, perhaps comprising different types of material, can be accurately formed and placed.
One type of component used by semiconductor devices is an isolation device. An isolation device, in general, includes a stacked gate isolated from an adjacent stacked gate by a shallow trench. The isolation device also typically includes a spacer formed on the sidewalls of the stacked gate.
Prior Art
FIGS. 1A through 1F
illustrate a prior art process for forming spacers in an isolation device. For simplicity of discussion, the process is described for a single spacer
50
(
FIG. 1F
) formed on the sidewall
11
of a stacked gate
10
adjacent to a shallow trench
12
. Shallow trench
12
is filled with a material such as high density plasma (HDP) oxide.
Referring first to
FIG. 1A
, a residual oxide layer
14
(e.g., silicon dioxide, SiO
2
) is formed over stacked gate
10
(including sidewall
11
) and shallow trench
12
. The residual oxide layer
14
is formed during annealing of the silicon. Usually, the residual oxide layer
14
includes a pre-implant oxide layer and a layer of oxide grown by a gate anti-reflective coating (GARC) anneal. Typically, the pre-implant oxide layer is approximately 50 to 75 Angstroms (Å) in thickness, and the oxide layer grown by the GARC anneal is approximately 95 Å in thickness. However, the total thickness of the residual oxide layer
14
is actually something less than the sum of the thicknesses of the pre-implant oxide layer and the oxide layer grown by the GARC anneal. A typical residual oxide layer
14
is approximately 100 Å in depth, although it may be less.
With reference to
FIG. 1B
, a liner layer
16
of a first material, typically TEOS (tetraethylorthosilicate), is deposited over stacked gate
10
(including sidewall
11
) and shallow trench
12
. Liner layer
16
typically has a thickness of approximately 150 Å.
Next referring to
FIG. 1C
, a layer
18
of a second material, typically nitride, is deposited over the liner layer
16
. Referring now to
FIG. 1D
, an etch of layers
16
and
18
is performed, removing the nitride and essentially all of the TEOS from the horizontal surfaces of the isolation device; however, a thin layer of TEOS typically remains on the surface
20
over shallow trench
12
. Also, a layer
16
of TEOS and a layer
18
of nitride also remain on the sidewall of stacked gate
10
. The residual oxide layer
14
, underlying the other layers, also remains.
With reference to
FIG. 1E
, a layer
22
of material, typically nitride, is deposited over the remaining portions of layers
16
and
18
. Referring now to
FIG. 1F
, an etch of layer
22
is performed to remove layer
22
from the horizontal surfaces of the isolation device and to form a spacer
50
having a prescribed (design) thickness T1. Spacer
50
is thus formed of layers
16
,
18
and
22
using a process that includes two etches. The residual oxide layer
14
, underlying the other layers, also is present along sidewall
11
.
A problem with the process illustrated by
FIGS. 1A through 1F
is that, during the second etch, relatively significant gouging of the HDP oxide in shallow trench
12
often occurs. The liner layer
16
is reduced to a thin layer, or effectively removed, during the first etch. Any remaining portion of layer
16
is not sufficiently thick to withstand the second etch and serve as a protective layer for the shallow trench
12
for the duration of the second etch. Consequently, shallow trench
12
is exposed during the second etch, allowing the HDP oxide to be gouged by the etch.
As a result of the gouging, isolation issues may be introduced, reducing the effectiveness of the isolation device. If these isolation issues are not detected or corrected, the performance of the semiconductor device may also be affected. Detection and correction of the gouging can reduce the yield (throughput) of the fabrication process and increase the unit cost of the semiconductor device.
Accordingly, what is needed is a method and/or system that can be used to form spacers in an isolation device, but without gouging the shallow trench filler material. The present invention provides a novel solution to this need.
DISCLOSURE OF THE INVENTION
Embodiments of the present invention provide a method and system thereof that can be used to form spacers in an isolation device, but without gouging the shallow trench filler material.
In one embodiment, an oxide layer is produced on a sidewall of a stacked gate and over a shallow trench adjacent to the stacked gate. The thickness of the oxide layer is sufficient to withstand a subsequent etch. A first layer of material is deposited over the oxide layer. In a first etch, the first layer is reduced to a first thickness along the sidewall. Because the oxide layer has a depth sufficient to withstand the first etch, the oxide layer serves as a protective layer for the shallow trench during the first etch. Accordingly, a protective liner layer does not need to be deposited in addition to the oxide layer.
In one embodiment, a second layer is deposited to a second thickness over the first layer subsequent to the first etch. A third layer is deposited over the second layer. In a second etch, the third layer is reduced to a third thickness along the sidewall. During the second etch, the second layer serves as a protective layer for the shallow trench. In the present embodiment, the first, second and third layers in combination form a spacer having a thickness corresponding to the first, second and third thicknesses.
In one embodiment, the first layer comprises nitride. In one embodiment, the second layer comprises TEOS (tetraethylorthosilicate). In another embodiment, the third layer comprises nitride. In yet another embodiment, the shallow trench is substantially filled with material comprising high density plasma (HDP) oxide. In another embodiment, the oxide layer is a residual oxide layer comprising a pre-implant oxide layer and an oxide layer grown by a gate anti-reflective coating (GARC) anneal. In one embodiment, the depth of the residual oxide layer is between approximately 150 and 180 Angstroms.
In its various embodiments, the present invention protects the shallow trench (e.g., the HDP oxide filler material) from gouging during etching, particularly during the second etch. This aspect of the present invention is especially beneficial as the length (in time) of the second etch is increased in order to remove nitride stringers at the core-periphery interface.
REFERENCES:
patent: 6521519 (2003-02-01), Shimizu et al.
Thio Hsiao-Han
Wang Zhigang
Yang Nian
Advanced Micro Devices , Inc.
Tsai Jey
LandOfFree
Replacing a first liner layer with a thicker oxide layer... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Replacing a first liner layer with a thicker oxide layer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Replacing a first liner layer with a thicker oxide layer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3312444