Method to remove an oxide seam along gate stack edge, when...

Details Method to remove an oxide seam along gate stack edge, when... Method to remove an oxide seam along gate stack edge, when...

2002-08-07

2004-03-30

Chaudhari, Chandra (Department: 2813)

Semiconductor device manufacturing: process

Making field effect device having pair of active regions...

Having insulated gate

C438S595000

Reexamination Certificate

active

06713350

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates generally to the field of semiconductor integrated circuits and, more particularly, to the use of self-align contact etch processes in the fabrication of memory devices.
BACKGROUND OF THE INVENTION
The semiconductor industry requires processes for achieving higher device density within a given die area. This is particularly true in memory circuit fabrication, in which the number and density of memory cells on a given chip have dramatically increased. The increase in density is a result of downsizing of the individual semiconductor devices, due partly to advances in photolithography and directional (anisotropic) plasma etching. As the horizontal device feature sizes continue to decrease to submicron dimensions, it is necessary to use self-alignment techniques to relax the alignment requirements and improve critical dimension (CD) control. One such technique is called a self-align contact (SAC) etch, in which a pair of adjacent gate stacks are utilized to align an etched opening in an insulating layer.
A conventional SAC process includes first providing a substrate on which there are at least two memory cells (such as MOS devices or flash memory cells), then forming an insulating layer, such as silicon oxide, on the substrate. Each of the two or more memory cells includes a polysilicon gate and spacers on the sidewalls of the gate. The two cells have a common source/drain region located between the gates of the cells. The insulating layer is patterned to form a self-align contact opening which exposes the common source/drain region. A conductive layer is deposited in the self align contact opening to form a contact. SAC etch processes primarily involve dry or plasma etches, typically utilizing a C
x
F
y
(x>1)-type plasma chemistry, such as, for example C
4
F
8
, C
5
F
8
, or C
4
F
6
in combination with other diluent gases.
The increase in packing density also places increased demands on many aspects of the fabrication process. Alignment of features from one level to the next is critically important, particularly the alignment of contact holes with underlying structures. The miniaturization of the devices makes the formation of interconnect structures difficult. To maintain sufficient electrical communication, the interconnect structure must be formed in precise alignment with the underlying active region. At the same time, the area of the interconnect structure interfacing with the active area must be maximized. Thus, as device sizes decrease, there is less room for misalignment errors of the interconnect structure.
Moreover, misalignment during an etch process can create problems resulting from the unintended etching or removal of an oxide seam (or spacer) along a gate stack, which forms between the gate stack and nitride spacers when spacer process formation includes the deposition of an oxide liner. In a flash cell, for example, if the misaligned SAC etch removes the oxide seam, metal will be deposited during subsequent plug fill metallization steps into the space where the seam previously was located, thereby causing the control to become shorted to both the floating gate and the substrate.
FIGS. 1-3
illustrate this problem.
FIG. 1
depicts a pair of typical flash cell gate stacks
1
positioned adjacent to each other on a substrate
2
at an intermediate stage prior to SAC etching. Each stack includes a tunnel oxide layer
3
positioned on top of the substrate
2
. A polysilicon floating gate
4
is positioned on top of the tunnel oxide layer
3
and an oxide
itride/oxide (ONO) layer
5
is positioned on top of floating gate
4
. A polysilicon control gate
6
is typically formed on top of the ONO layer
5
with a tungsten silicide layer
7
formed on top of the control gate
6
. A cap
8
, which is typically either nitride or TEOS is positioned on the tungsten silicide layer
7
. Between the gate stacks is the area
9
in which the contact will be etched into the substrate.
FIGS. 2A-2D
illustrate processing steps that may be performed subsequent to the intermediate stage illustrated in FIG.
1
. First, an oxide layer
10
, preferably a TEOS layer with a thickness of between about 50 Å and about 300 Å, more preferably, 100 and 200 Angstroms, is deposited over the surface of the device (FIG.
2
A), to include sidewall S/D oxidation. Next, dry etching is used to remove horizontal TEOS surfaces, leaving vertical surfaces to form oxide spacers (also referred to herein as seams)
11
, along the gate stack edge (FIG.
2
B). To form a nitride spacer, a nitride layer
12
is deposited over the surface of the device, as shown in FIG.
2
C. To complete nitride spacer formation, a dry etch is used, resulting in gate stacks having oxide seams (spacers)
11
separating nitride spacers
13
from gate stack
1
(FIG.
2
D). The oxide seams
11
are exposed at the top
14
, which can present a problem during the subsequent SAC etch process.
Following a first dielectric layer deposition, a self-align contact etch is used to remove the dielectric from the common region between the gate stacks to prepare the contact region
15
for metal deposition.
FIG. 3
illustrates a problem that occurs when the self-align contact etch
17
is misaligned so as to hit a gate stack, rather than merely the contact region between adjacent gate stacks. Because the first level dielectric
16
and the exposed oxide seam
11
are composed of the same material, the etch plasma chemistry, which is selective to this material, etches out the oxide seam as well as the intended dielectric, leaving a void where the seam was. Once the seam is etched out, the resulting void would then be filled with metal. Because the metal deposited in the vertical space where the oxide seam had been is conductive, the control gate would be shorted to the floating gate and to the substrate (in a flash memory cell for example), or the gate would be shorted to the substrate (in a device other than a flash cell, such as, for example, a standard MOS transistor). This shorting results in a non-functional device.
There is thus a need for a method which mitigates or eliminates the above disadvantages. In particular, there is a need for a method which prevents problems, such as shorting between gate stack components, that occur when a misaligned self-align contact etch unintentionally removes an oxide seam or spacer.
SUMMARY OF THE INVENTION
The present invention provides methods of preventing a short from occurring in a gate stack during a self-align contact etch process. According to one aspect of the invention, a memory device has at least two adjacent gate stacks positioned on a substrate. Each of the stacks is separated from a vertical nitride spacer by a vertical oxide spacer exposed at its top end. A portion of the vertical oxide spacer is removed from its exposed top end, a nitride layer is formed over the memory device sufficient to replace the removed portion, and then the memory device is exposed to a self-align contact etch process.
The present invention also relates to a memory device that has a gate stack, a vertical spacer adjacent to the gate stack, in which the vertical spacer has a lower oxide portion and an upper nitride portion, and a continuous nitride layer overlaying the spacer and the gate stack.
The present invention also relates to a method of fabricating memory devices. A preferred embodiment of the invention includes the steps of forming a continuous oxide layer over a gate stack positioned on a substrate, the continuous oxide layer having horizontal and vertical surfaces, removing the horizontal surfaces of the continuous oxide layer, and forming a first continuous nitride layer over the gate stack, the vertical oxide surfaces and the substrate. In the preferred embodiment, the first continuous nitride layer has horizontal and vertical surfaces, with each vertical oxide surface thereby becoming an oxide spacer. The present invention may also include the steps of removing the horizontal surfaces of the first continuous nitride layer, thereby e

Affiliated with

Also associated with

Rating

Method to remove an oxide seam along gate stack edge, when... 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 to remove an oxide seam along gate stack edge, when..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method to remove an oxide seam along gate stack edge, when... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3268939