Semiconductor device manufacturing: process – With measuring or testing – Optical characteristic sensed
Reexamination Certificate
2002-12-23
2004-04-20
Thompson, Craig A. (Department: 2813)
Semiconductor device manufacturing: process
With measuring or testing
Optical characteristic sensed
C438S014000, C438S800000, C438S010000, C702S040000, C702S058000, C702S081000
Reexamination Certificate
active
06723574
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to methods and systems of quantifying uniformity of measured quantities on semiconductor wafers, and more particularly, to improved methods and systems for characterizing and analyzing nonuniformities on semiconductor wafers and providing feedback and control to preceding semiconductor manufacturing processes.
2. Description of the Related Art
Semiconductor wafers undergo numerous processes during the semiconductor manufacturing process. Layers may be added, patterned, etched, removed, polished and many other processes. After each process the wafer is typically examined to confirm the previous process was completed with an acceptable level of errors or nonuniformities. The various operating variables (e.g., event timing, gas pressure, concentrations, temperatures, etc.) of each process the wafer is processed through are recorded so that any changes in any variable may be quickly identified and potentially correlated to any errors or nonuniformities discovered when the wafer is examined.
FIG. 1A
shows a typical etched wafer
100
. A top layer of material was mostly removed from the wafer in the etch process except for a portion
106
of the top layer. For clarity purposes, the portion
106
is a portion of a layer or ultrathin film. A notch
104
is typically included in each wafer
100
so that the wafer can be oriented (aligned) in the same position during the various manufacturing processes. The portion
106
is a nonuniform portion of the surface of the wafer
100
and therefore can be termed a nonuniformity. As shown, the portion
106
is in the approximate form of a ring or annular shape where the top layer was removed from the center and around the edges of the wafer
100
.
FIG. 1B
shows another typical etched wafer
120
. A portion
108
of a top layer remains, when the top layer was mostly removed in the etch process. The portion
108
is typically termed an azimuthal-type nonuniformity on the surface of wafer
120
because the nonuniformity
108
is not the same at the same radius around the wafer
120
.
Prior art approaches to describing nonuniformities
106
,
108
include subjective, verbal descriptions such as “center-fast” for annular nonuniformity
106
or “left side slow” for azimuthal nonuniformity
108
. Center-fast generally describes wafer
100
because material from the center of the wafer
100
is removed faster than the material in the annular region
106
. However, center-fast does not provide a specific, objective and quantitative description of the nonuniformity
106
. Similarly, left side slow describes wafer
120
because the etch process removed material from the left side region
108
slower than the other regions of the wafer
120
but left side slow also fails to provide a specific, objective and quantitative description of the nonuniformity
108
.
The descriptions of the nonuniformities
106
,
108
are used to provide feedback to correct errors and inconsistencies in the etch and other preceding processes that were performed on the wafers
100
,
120
. The descriptions of the nonuniformities
106
,
108
can also be used to track the impact of the nonuniformities
106
,
108
on subsequent semiconductor manufacturing process and on metrics from completed semiconductor devices (e.g., device yields, performance parameters, etc.)
As nonuniformities become smaller and smaller, the nonuniformities become less symmetrical and also more difficult to accurately describe with the subjective, verbal descriptions.
FIG. 1C
shows a typical wafer
150
with multiple, asymmetrical nonuniformities
152
A-G. The nonuniformities
152
A-G can be smaller and are less symmetrical than nonuniformities
106
,
108
in part because the various variables in the etch and other previous processes are very stringently controlled. The subjective, verbal descriptions have therefore become insufficient to accurately describe the nonuniformities
152
A-G so that further improvements in the preceding processes can be successfully completed.
A more objective description of wafer uniformity is referred to as a 3-sigma uniformity metric. The 3-sigma uniformity metric quantifies a standard deviation of measurements of some quantity of the wafer. By way of example, the 3-sigma can be an expression of the deviations in thickness of the wafer detected by an array of measurement points across the wafer.
FIG. 1D
shows a typical 49-point array used in completing a scan of wafer
160
. The thickness of the wafer
160
is measured at each of the 49 points. The 49-points are arranged with a center point
162
, and three concentric rings
164
,
168
,
172
. The inner ring
164
has 8 evenly spaced points. The intermediate ring
168
has 16 evenly spaced points. The outer ring
172
has 24 evenly spaced points. The rings
164
,
168
,
172
are typically approximately equally spaced radially from the center point
162
. Each of the points in the rings
164
,
168
,
172
and the center point
162
is typically assigned to represent a given portion of the wafer
160
. For example, a typical wafer
160
has a 3 mm edge exclusion zone on the outer perimeter of the wafer
160
. The rings
164
,
168
,
172
and the center point
162
are spaced equidistant and therefore each of the 49 points represent about {fraction (1/49)}
th
of the area of the wafer
160
, less the 3 mm exclusion zone (i.e., the outer edge of the wafer where expected process abnormalities occur). Because nonuniformities do not suddenly appear under a single scan point, the nonuniformities are automatically smoothed due to the choice of measuring points.
The measured thicknesses can be correlated to other aspects of the wafer such as an etch rate at the particular measured point A standard deviation (SD) and mean of the etch rates at these 49 points are determined. The 3-sigma nonuniformity metric equal to [3*(SD)/mean]/100, expressed as percentage is typically reported. The 3-sigma metric effectively compresses or summarizes all of the individually measured point etch rates to one summary value. However, the 3-sigma metric does not provide any information about the relationship between the etch rates at the different measured points. This relationship can become important when higher uniformity is achieved. The relationship can help identify differences between different etch patterns with the same 3-signal nonuniformity metric.
Many prior art approaches apply a Fourier or a Bessel decomposition on the measured data to better describe a shape and magnitude of the nonuniformity
106
,
108
,
152
A-G. However, Fourier and Bessel decompositions are an effort to force-fit the shape of the nonuniformity to a predetermined Fourier and Bessel defined shape, rather than determine the actual shape of the nonuniformity
106
,
108
,
152
A-G. The Fourier and Bessel decompositions are therefore only estimating the magnitude of the nonuniformity in the forced-fit shape. While the Fourier and Bessel decompositions provide additional objective descriptions of the nonuniformities, the Fourier and Bessel decompositions still do not accurately describe either the shape or the magnitude of the nonuniformity
106
,
108
,
152
A-G.
In view of the foregoing, there is a need for an improved system and method of objectively and accurately quantifying a nonuniformity and correlating the nonuniformity to a any relevant change in the system (for e.g. process variable, hardware change).
SUMMARY OF THE INVENTION
Broadly speaking, the present invention fills these needs by providing a system and method for quantifying a nonuniformity and correlating the nonuniformity to process variables. It should be appreciated that the present invention can be implemented in numerous ways, including as a process, an apparatus, a system, computer readable media, or a device. Several inventive embodiments of the present invention are described below.
One embodiment includes a method for determining a multiple uniformity metrics of a semiconductor wafer man
Bailey III Andrew D.
Misra Pratik
Yadav Puneet
Lam Research Corporation
Martine & Penilla LLP
Thompson Craig A.
LandOfFree
Method for quantifying uniformity patterns and including... 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 for quantifying uniformity patterns and including..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for quantifying uniformity patterns and including... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3236701