Ex Parte Fratti et al

Board of Patent Appeals and Interferences

Sep 28, 2009

10628941 (B.P.A.I. Sep. 28, 2009)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte ROGER A. FRATTI and WARREN K. WASKIEWICZ
__________

Appeal 2009-009128
Application 10/628,941
Technology Center 2800
____________

Decided: September 28, 2009
____________

Before CATHERINE Q. TIMM, BEVERLY A. FRANKLIN, and
MICHAEL P. COLAIANNI, Administrative Patent Judges.

COLAIANNI, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134 the final rejection of claims
1-16. We have jurisdiction over the appeal pursuant to 35 U.S.C. § 6(b).
We AFFIRM.
Appellants disclose power transistor devices and techniques to control
curvature in such devices (Spec. 1). Specifically, Appellants disclose
Appeal 2009-009128
Application 10/628,941

forming a stress compensation layer 410 atop a DMOS device 100 to affect
the curvature of the device (Spec. 9; Fig. 4).
Claims 1, 7, and 10 are illustrative:
1. A method for controlling curvature of a power transistor device
comprising a device film formed on a substrate, the method comprising the
steps of:

thinning the substrate, the device having an overall residual stress
attributable at least in part to the thinning step; and

applying a stress compensation layer to a surface of the device film,
the stress compensation layer having a tensile stress sufficient to
counterbalance at least a portion of the overall residual stress of the device.

7. The method of claim 1, wherein the steps of thinning and
applying are performed repeatedly until a desired curvature is attained.

10. The method of claim 1, wherein the stress compensation layer
applied to the surface of the device maintains the curvature of the device.

The Examiner relies on the following prior art references as evidence
of unpatentability:

Oda US 6,091,121 Jul. 18, 2000
Shibib US 6,559,011 B1 May 6, 2003

Sergey Savastiouk, et al., Atmospheric Downstream Plasma, European
Semiconductor 1-4 (June 1998).

Appellants appeal the following rejection:
Claims 1-16 are rejected under 35 U.S.C. § 103(a) as being
unpatentable over Oda in view of Savastiouk and Shibib.
Appellants argue independent method claim 1 and device claims 13
and 16 as a group. We select claim 1 as representative of the group, with
2
Appeal 2009-009128
Application 10/628,941

claims 2-6, 8, 13-16 standing or falling therewith. Dependent method claims
7, 11, and 12 are argued as a group of which we select claim 7 as
representative. We further address Appellants’ arguments regarding
separately argued dependent method claims 9 and 10.

Claim 1
ISSUE
Have Appellants shown that the Examiner reversibly erred in
determining that: (1) the combined teachings of Oda, Savastiouk and Shibib
teach or suggest “applying a stress compensation layer to a surface of the
device film, the stress compensation layer having a tensile stress sufficient to
counterbalance at least a portion of the overall residual stress of the device”
as recited in claim 1 and (2) there is a reason to combine Savastiouk’s
thinning process with Oda? We decide these issues in the negative.

PRINCIPLES OF LAW
The applicant bears the procedural burden of showing error in the
Examiner’s rejections. See, e.g., In re Kahn, 441 F.3d 977, 985-86 (Fed.
Cir. 2006) (“On appeal to the Board, an applicant can overcome a rejection
[under § 103] by showing insufficient evidence of prima facie obviousness”)
(citation and internal quote omitted).
Where the claimed and prior art products are identical or substantially
identical, the PTO can require an applicant to prove that the prior art
products do not necessarily or inherently possess the characteristics of the
claimed product. In re Best, 562 F.2d 1252, 1255 (CCPA 1977).
3
Appeal 2009-009128
Application 10/628,941

When assessing whether a claimed invention would have been
obvious, a court must ask “whether the improvement is more than the
predictable use of prior art elements according to their established
functions.” KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 417 (2007).
Applicant’s motivation does not control the obviousness analysis;
rather, “any need or problem known in the field of endeavor at the time of
invention and addressed by the patent can provide a reason for combining
the elements in the manner claimed.” KSR, 550 U.S. at 420.
FACTUAL FINDINGS (FF)
1. Appellants do not dispute the combination of Shibib and Oda
(App. Br. and Reply Br. generally).
2. Oda discloses a semiconductor device having layers of silicon
nitride and silicon oxide atop a gate electrode to prevent moisture
infiltration to the gate electrode and to control the amount of stress
applied to the semiconductor structure (col. 6, ll. 34-67). Oda
discloses that the stress in the layers may be tailored such that the
tensile stress is relaxed by the compressive stress of another layer
to prevent applying stress directly to the gate oxide film 3 and gate
electrode 4 (col. 7, ll. 40-60).
3. Oda further discloses that the amount of stress applied to the
semiconductor structure may be optimized by controlling the RF
frequency and/or thickness of the layers (col. 6, ll. 55-67). Oda
exemplifies controlling the stress in the layers so that the residual
tensile stress is half of its original value (col. 8, ll. 50-57).
4. Oda discloses that the multi-layer film may include nitride (i.e.,
silicon nitride) and oxide (i.e., silicon oxide) layers (col. 5, ll. 60-
4
Appeal 2009-009128
Application 10/628,941

67, col. 6, ll. 1-19). The nitride layer has a tensile stress of 1 x 1010
dynes/cm3 and the compressive stress in Oda’s oxide layer is 1.5 x
109 dynes/cm3, which indicates that a portion of the tensile stress
of the nitride layer is necessarily available to counteract a
“portion” of the residual stress of the thinned device (col. 6, ll. 34-
52).
5. Savastiouk discloses that backgrinding (i.e., thinning) techniques
were known in the semiconductor art and that such techniques
produced residual stress that caused wafer bow and warp
(Savastiouk 1).
6. The Specification defines “stress compensation layer” as including
“one or more thin films or other material layers applied to a device
film in order to counterbalance at least a portion of an overall
residual stress of a device” (Spec. 9).
7. Appellants disclose that the stress compensation layer may be
made of silicon nitrides, silicon oxides, silicon oxynitrides,
oxynitrides, nitrides, and combinations thereof (Spec. 9).
ANALYSIS
Appellants argue that Oda, Savastiouk and Shibib fail to teach or
suggest “applying a stress compensation layer to a surface of the device film,
the stress compensation layer having a tensile stress sufficient to
counterbalance at least a portion of the overall residual stress of the device”
as recited by claim 1 (App. Br. 3 and 5). Appellants contend that Oda’s
layers 12, 15, and 17 are not disclosed as providing a tensile stress sufficient
to counterbalance any amount of overall residual stress from the substrate
thinning (App. Br. 5). Appellants also contend that Oda discloses that the
5
Appeal 2009-009128
Application 10/628,941

layers in the multi-layer film merely compensate for stresses inherent in
layers adjacent to the multi-layer film (App. Br. 5).
The Examiner finds that Oda discloses forming a multi-layer film
made of the same materials disclosed by Appellants, such that it would be an
inherent property of the material to provide a tensile stress sufficient to
counterbalance at least a portion of the residual stress in a thinned substrate
(Ans. 6). This finding by the Examiner properly shifted the burden to
Appellants to establish that the prior art does not possess the particular
characteristic (i.e., counterbalancing a portion of the residual stress of the
device). Best, 562 F.2d at 1255.
Appellants respond by stating that even though the film may
inherently have a tensile stress, it does not necessarily have a tensile stress
sufficient to counterbalance at least a portion of an overall residual stress of
the device (Reply Br. 4). However, Appellants have not shown error in the
Examiner’s determination that Oda’s film would inherently have a residual
stress sufficient to balance at least “a portion” of the residual stress caused
by a thinning operation taught by Savastiouk. Indeed, the example relied
upon by the Examiner (col. 6, ll. 44-45; Ans. 6) shows that the tensile stress
in Oda’s nitride layer is 1 x 1010 dynes/cm3 and the compressive stress in
Oda’s oxide layer is 1.5 x 109 dynes/cm3, which indicates that a portion of
the tensile stress of the nitride layer is necessarily available to counteract a
“portion” of the residual stress of the thinned device as claimed.
Appellants further argue that there is no motivation to combine
Savastiouk’s thinning process with Oda because Savastiouk does not
disclose or suggest that “thinner wafers may be used to control curvature of
a power transistor device, as taught by the present invention” (App. Br. 4).
6
Appeal 2009-009128
Application 10/628,941

Appellants’ argument improperly requires the Examiner’s reason for
combining the prior art references to be the same as Appellants’ reason. To
the contrary, any need or problem known in the field of endeavor can
provide a reason for combining the elements in the manner claimed. KSR,
550 U.S. at 420
Appellants’ motivation argument fails to show reversible error in the
Examiner’s stated reason for combining Savastiouk’s thinning process with
Oda’s semiconductor manufacturing method: to adapt the device to current
technology and to increase the number of chips per wafer (Ans. 4).

Claim 7
ISSUE
Have Appellants shown reversible error in the Examiner’s
determination that it would have been obvious to repeat the steps of thinning
and applying stress compensation layers until a desired curvature is attained
as recited in claim 7? We decide this issue in the negative.

PRINCIPLES OF LAW
In addition to the principles of law noted above, we rely on the
following principles:
The test for obviousness is what the combined teachings of the
references would have suggested to one of ordinary skill in the art. In re
Keller, 642 F.2d 413, 425 (CCPA 1981).
The obviousness “analysis need not seek out precise teachings
directed to the specific subject matter of the challenged claim, for a court
7
Appeal 2009-009128
Application 10/628,941

can take account of the inferences and creative steps that a person of
ordinary skill in the art would employ.” KSR, 550 U.S. at 417-18.

ADDITIONAL FINDINGS OF FACT
8. The Examiner stated that the subject matter of claim 7 would have
been obvious over the combined teachings of Oda, Shibib and
Savastiouk “in order to have better control over the characteristics
of the device and by using known monitoring techniques” (Ans. 5).

ANALYSIS
Appellants argue that none of the prior art teaches repeatedly
performing the thinning and applying steps until a desired curvature is
obtained (Reply Br. 5). Appellants contend that the prior art fails to disclose
that thinning of the substrate would produce a curvature of the device (App.
Br. 5).
Appellants’ argument that the prior art does not teach that thinning
produces curvature is unpersuasive because, as the Examiner points out,
Savastiouk teaches that thinning the substrate produces residual stress that
causes wafer bow and warp (Ans. 9; FF 5). Accordingly, combining
Savastiouk’s thinning technique with Oda’s semiconductor process would
have produced a thinner semiconductor substrate having residual stress. As
noted above, Oda exemplifies an embodiment where the tensile stress of the
nitride layer is greater than the compressive stress of the oxide layer thereby
counterbalancing a portion of the residual stress inherent in the thinned
wafer of the Oda and Savastiouk combination.
8
Appeal 2009-009128
Application 10/628,941

Appellants’ argument that the art does not teach the “repeated” feature
of claim 7, fails to show error in the Examiner’s determination that one of
ordinary skill would have known to monitor and repeat the thinning and
stress compensation layer applying steps to have better control over the
characteristics of the device (Ans. 5). In fact, Appellants contend that the
Examiner’s determination is “incorrect” based on Appellants’ argument that
the cited references do not disclose that thinning the substrate would produce
a curvature (App. Br. 5). This argument was shown above to be factually
erroneous.
That none of the cited art, individually, expressly teaches repeated
thinning and applying steps is of no moment because the Examiner may take
account of inferences and creative steps that a person of ordinary skill would
employ. KSR, 550 U.S. at 417-18. The Examiner’s analysis takes account
of the inferences and creative steps (i.e., monitoring and repeated
application) that would have been employed by one of ordinary skill, which
Appellants have not shown to be in error.
Indeed, Oda discloses that it was known to optimize the stress in the
film by controlling the RF frequency during deposition and/or the layer
thickness of multiple layers (i.e., repeated depositions). Therefore, it would
have been obvious to thin the substrate as taught by Savastiouk and apply
layers of material as taught by Oda to optimize the desired stress and thus
curvature of the device. The thinning process is taught by Savastiouk as
introducing residual stress into the device that may cause the device to bow
or warp. Oda teaches controlling the stress in the multi-layer film to
counteract stresses in various layers of the device. Accordingly, the
9
Appeal 2009-009128
Application 10/628,941

combined teachings of the references would have suggested the invention of
claim 7.
In fact, Appellants’ arguments that “Oda, Savastiouk, or Shibib”
(emphasis added) fail to disclose the claim 7 subject matter indicates that
Appellants improperly attack the references individually, instead of
addressing what the combined teachings would have suggested to one of
ordinary skill the art. Keller, 642 F.2d at 425.
Accordingly, Appellants have not shown that the Examiner reversibly
erred in determining that it would have been obvious to repeat the steps of
thinning and applying stress compensation layers until a desired curvature is
attained as recited in claim 7.

Claims 9 and 10
ISSUE
Have Appellants shown that the Examiner reversibly erred in
determining that Oda’s nitride layer (i.e., stress compensation layer) would
have inherently changed (claim 9) or maintained (claim 10) the curvature of
a power transistor device as claimed? We decide this issue in the negative.

PRINCIPLES OF LAW
We rely on the principles of law noted earlier in this Decision.

FINDINGS OF FACT
9. The Examiner determines that Oda’s nitride film inherently has a
tensile stress such that applying the nitride film to a thinned
semiconductor device as taught by Savastiouk would have
10
Appeal 2009-009128
Application 10/628,941

inherently affected (changed or maintained) the curvature of the
device (Ans. 5 & 9).
10. Claims 9 and 10 depend from claim 1, which recites that “the
stress compensation layer [has] a tensile stress sufficient to
counterbalance at least a portion of the overall residual stress of the
device” (claim 1).
ANALYSIS
Appellants argue that there is no suggestion in Oda that a stress
compensation layer applied to the surface of the device changes or maintains
the curvature of the device (App. Br. 6). Appellants contend that Oda
merely discloses a protecting insulator film having a compressive stress for
relaxing a tensile stress of the protecting nitride film (App. Br. 6).
While Oda does disclose that the oxide film relaxes the nitride film
(col. 7, ll. 50-53), Oda further exemplifies that the tensile stress is not
necessarily eliminated (FF. 4). Accordingly, a portion of the tensile stress
would still be present and inherently change or maintain the curvature of the
thinned substrate by affecting a portion of the residual stress as required by
claim 9.
Accordingly, Appellants have not shown reversible error with the
Examiner’s determination that Oda’s layers would inherently affect (change
or maintain) the curvature of the device.
Moreover, Oda discloses that the stress in the various layers can be
optimized by controlling the RF frequency during deposition and/or
thickness of the layer (FF 3). In fact, Oda exemplifies that the stress in the
various layers may be optimized to a desired value, which may include
halving the tensile stress of the nitride layer (FF 3). Therefore, it would
11
Appeal 2009-009128
Application 10/628,941

have been obvious to optimize the stress in the multilayer film to any desired
level.
The desired level of stress would have included optimizing the stress
level to counterbalance the stress caused by the thinned substrate and the
tensile stress in the nitride layer. When the stress in the nitride layer is
optimized to completely counterbalance the tensile stress in the nitride layer,
the multilayer would maintain the curvature of the thinned substrate (i.e., it
would have no affect on the curvature).
Similarly, when the tensile stress of the nitride layer is optimized to
only partially counterbalance the tensile stress in the nitride layer, as in
Oda’s example (FF 3), then a portion of the tensile stress would be available
to counterbalance the residual stress in a thinned substrate and thereby
change the curvature of the device to some extent.
For the above reasons, we affirm the Examiner’s § 103 rejection of
claims 9 and 10 over Oda in view of Savastiouk and Shibib.

DECISION
We affirm the Examiner’s § 103 rejection of claims 1-16 over Oda in
view of Savastiouk and Shibib.

ORDER
The Examiner’s decision is affirmed.

No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1).

12
Appeal 2009-009128
Application 10/628,941

AFFIRMED

ssl

RYAN, MASON & LEWIS, LLP
1300 POST ROAD
SUITE 205
FAIRFIELD, CT 06824
13