Ex Parte Archie

Board of Patent Appeals and Interferences

Sep 3, 2008

09902374 (B.P.A.I. Sep. 3, 2008)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte CHARLES N. ARCHIE
____________

Appeal 2007-3815
Application 09/902,374
Technology Center 2800
____________

Decided: September 3, 2008
____________

Before KENNETH W. HAIRSTON, JOSEPH F. RUGGIERO,
and MAHSHID D. SAADAT, Administrative Patent Judges.
HAIRSTON, Administrative Patent Judge.

DECISION ON APPEAL
Appellant appeals under 35 U.S.C. § 134 from a final rejection of
claims 1 to 26. We have jurisdiction under 35 U.S.C. § 6(b).
We will reverse the rejection.

Appeal 2007-3815
Application 09/902,374

2
STATEMENT OF THE CASE
Appellant has invented a method of producing an optimum critical
dimension value that comprises structural measurements of a critical
dimension structure that are only relevant to a critical dimension of the
critical dimension structure (Figures 7 and 8; Specification 7, 8, and 11 to
13).
Claim 1 is representative of the claims on appeal, and it reads as
follows:
1. A method of producing an optimum critical dimension value, said
method comprising:
acquiring a waveform of data for a critical dimension structure;
determining a stepper focus parameter for said critical dimension
structure;
calculating an approximate critical dimension measurement for said
critical dimension structure;
calibrating said data of said waveform by determining at least three
best fit data parameters for improving a linearity of said waveform;
combining said stepper focus parameter with said approximate critical
dimension measurement and said best fit data parameters, wherein said
combining removes structural bias parameters from said approximate critical
dimension measurement; and
generating said optimum critical dimension value from said
combining, wherein said optimum critical dimension value comprises
structural measurements of said critical dimension structure that are only
relevant to a critical dimension of said critical dimension structure.
Appeal 2007-3815
Application 09/902,374

3
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Su US 6,388,253 B1 May 14, 2002
(filed Jun. 29, 1999)
Tanaka US 6,616,759 B2 Sep. 9, 2003
(filed Sep. 6, 2001)
The Examiner rejected claims 1 to 26 under 35 U.S.C. § 103(a) based
upon the teachings of Su and Tanaka.

ISSUE
Appellant contends that neither Su nor Tanaka, whether considered
singly or in combination, discloses the claimed steps of generating an
optimum critical dimension value based on three best fit data parameters in
combination with a stepper focus parameter (Br. 15 and 20). Thus, the issue
before us is whether the applied prior art teaches or would have suggested to
the skilled artisan generating an optimum critical dimension value based on
three best fit data parameters in combination with a stepper focus parameter
as set forth in the claims on appeal?

FINDINGS OF FACT
1. Appellant has not challenged the Examiner’s findings (Ans. 3 to 5)
that Su discloses the claimed steps of “acquiring a waveform of data for a
critical dimension structure,” “determining a stepper focus parameter for
said critical dimension structure,” and “calculating an approximate critical
dimension measurement for said critical dimension structure.”
Appeal 2007-3815
Application 09/902,374

4
2. The Examiner has acknowledged (Ans. 5 and 6) that “Su (253) as
applied above fails to teach a method of calibrating the waveform data by
determining at least three best-fit data parameters, and combining the best-fit
data parameters with a stepper focus parameter and a critical dimension
measurement to improve the linearity of the critical dimension waveform, as
recited in claims 1, 8, 10, 12, 19, and 20.”
3. Tanaka describes a method and system for monitoring a
semiconductor processing apparatus 1 by sensing the processing state of the
apparatus via sensing unit 3 (Figure 1A; Abstract; col. 3, ll. 31 to 43). The
sensed data is stored in storage unit 4 (col. 3, ll. 58 and 59). A storage unit 6
preserves measured values of processing results for each of a variety of
semiconductor devices (col. 4, ll. 1 and 12). A model equation generation
unit 7 fetches sensed data from storage unit 4 and measured values for a
sample of the same type of semiconductor device undergoing processing by
the semiconductor processing apparatus 1 from storage unit 6, and
“generates a model equation for predicting measured values of processing
results using the sensed data as explanatory variables” (col. 4, ll. 13 to 21).
The generated model equation is stored in model equation storage unit 8
along with other model equations generated for other semiconductor devices
(col. 5, ll. 38 to 43). Thereafter, when a semiconductor device of a certain
type is loaded into the semiconductor processing apparatus 1, and sensed by
sensor 3, a model equation corresponding to the loaded semiconductor
device is fetched from model equation storage unit 8 and loaded into model
equation based prediction unit 9 (col. 5, ll. 43 to 46). The model equation
prediction unit 9 calculates predicted values for processing results for the
Appeal 2007-3815
Application 09/902,374

5
semiconductor device, and the calculated results are sent to process recipe
control unit 10 (col. 5, ll. 47 to 56). The calculated processing results are fed
back to semiconductor processing apparatus 1 (Fig. 1A).
4. The model equation generation unit 7 in Tanaka uses normal
multiple regression to generate a model equation from sensed data for
prediction of processing results (Figs. 6 and 7; col. 5, ll. 8 to 34).
5. The model equation regression analysis may lead to a linear result
(Fig. 6; col. 5, ll. 34 to 37).

PRINCIPLES OF LAW
The Examiner bears the initial burden of presenting a prima facie case
of obviousness. In re Oetiker, 977 F.2d 1443, 1445 (Fed. Cir. 1992). If that
burden is met, then the burden shifts to the Appellant to overcome the prima
facie case with argument and/or evidence. See Id.
The Examiner’s articulated reasoning in the rejection must possess a
rational underpinning to support the legal conclusion of obviousness. In re
Kahn, 441 F.3d 977, 988 (Fed. Cir. 2006).

ANALYSIS
Although we agree with the Examiner that Tanaka improves the
linearity of the monitored semiconductor manufacturing data (Ans. 6;
Findings of Fact 4 and 5), we do not, however, agree with the Examiner that
“Tanaka (759) discloses a method of monitoring semiconductor
manufacturing that includes the generation of a model equation from three
parameters of sensed data, and using the best fit of the three parameters via
Appeal 2007-3815
Application 09/902,374

6
multiple regression to improve linearity of the data” (Ans. 6), or that “[i]t is
implied . . . that removing faulty process shapes in accordance with Tanaka
(759) is equivalent to improving linearity, as recited in claims 1, 8, 10, 12,
19, and 20” (Ans. 6). As indicated supra (Finding of Fact 3), Tanaka is
completely silent as to calibrating data of a waveform of data for a critical
dimension structure by determining at least three best fit data parameters for
improving linearity of that waveform, combining a stepper focus parameter
with an approximate critical dimension measurement for the critical
dimension structure and the best fit data parameter to remove structural bias
parameters from the approximate critical dimension measurement, and
generating an optimum critical dimension value from the combined
parameters and critical dimension measurement wherein the optimum
critical dimension value comprises structural measurements of the critical
dimension structure that are only relevant to a critical dimension of the
critical dimension structure as set forth in the claims on appeal.
In view of the foregoing, we agree with the Appellant that the applied
references neither teach nor would have suggested to the skilled artisan the
generation of an optimum critical dimension value based on three best fit
data parameters in combination with a stepper focus parameter as set forth in
the claims on appeal without the benefit of impermissible hindsight (Br. 15,
20, and 21).

CONCLUSION OF LAW
The Examiner has not established the obviousness of independent
claims 1, 8, 10, 12, 19, and 20, and the claims that depend therefrom.
Appeal 2007-3815
Application 09/902,374

7
ORDER
The obviousness rejection of claims 1 to 26 is reversed.

REVERSED

KIS

FREDERICK W. GIBB, III
Gibb & Rahman, LLC
2568-A RIVA ROAD
SUITE 304
ANNAPOLIS MD 21401