Ex Parte Langdo et al

Patent Trial and Appeal Board

Jan 17, 2014

11073780 (P.T.A.B. Jan. 17, 2014)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________________

Ex parte THOMAS A. LANGDO, MATTHEW T. CURRIE,
RICHARD HAMMOND, ANTHONY J. LOCHTEFELD, and
EUGENE A. FITZGERALD
____________________

Appeal 2011-006333
Application 11/073,780
Technology Center 2800
____________________

Before CHUNG K. PAK, PETER F. KRATZ, and CATHERINE Q. TIMM,
Administrative Patent Judges.

TIMM, Administrative Patent Judge.

DECISION ON REQUEST FOR REHEARING

Appellants request rehearing of our Decision of October 31, 2013. In
conformance with MPEP §1214.01(II), Appellants submit the Request in
two papers, which we will refer to as Request (I) and Request (II). Request
(I) requests rehearing of our decision to sustain the rejection of claims 61,
Appeal 2011-006333
Application 11/073,780

2
64, and 66 under 35 U.S.C § 103(a) as obvious over Chu1 in view of
Kakizaki.2 Request (II) requests rehearing of the new ground of rejections
we entered with respect to claims 92, 93, and 110. We address both papers
in this Decision as a single Request for Rehearing.
A. Request I: The Rejection over Chu in view of Kakizaki
With respect to the rejection of claims 61, 64, and 66 under 35 U.S.C.
§ 103(a) over Chu in view of Kakizaki, Appellants contend we
misapprehended or overlooked the requirements of claim 61 in sustaining
the rejection (Req. Reh’g I at 2).
Claim 61 reads as follows:
61. A method for forming a structure, the method
comprising:
providing a first substrate having a first strained
semiconductor layer formed thereon;
performing a step to increase a bond strength between the
first strained semiconductor layer and an insulator layer
disposed on a second substrate and to which the first strained
semiconductor is subsequently bonded;
after performing the step to increase the bond strength,
bonding the first strained semiconductor layer to the insulator
layer disposed on the second substrate; and
removing the first substrate from the first strained
semiconductor layer, the first strained semiconductor layer
remaining bonded to the insulator layer,
wherein bonding comprises achieving a bond strength
greater than or equal to about 1000 milliJoules/meter squared
(mJ/m2) at a temperature less than approximately 600 °C.

1 Chu et al., US 6,649,492 B2, issued Nov. 18, 2003.
2 Kakizaki et al., US 6,326,279 B1, issued Dec. 4, 2001.
Appeal 2011-006333
Application 11/073,780

3
(Claims App’x at Br. A2 (emphasis added).)
According to Appellants, we did not address Appellants’ argument
that neither Chu nor Kakizaki teaches or suggests the requirement of the last
clause of claim 61 (Req. Reh’g I at 2, citing Br. 43-44, and further directing
our attention to Br. 41).
We disagree. We addressed this argument in Section G of the
Decision (Decision 13-14).
Appellants contend that bonding that achieves the recited bond
strength is not taught by the references (Req. Reh’g I at 2). We agree with
Appellants that neither Chu nor Kakizaki expressly teaches the level of
bonding required by the last clause of claim 61. However, such an express
teaching is not required. The obviousness analysis “need not seek out
precise teachings directed to the specific subject matter of the challenged
claim, for a court [or this Board] can take account of the inferences and
creative steps that a person of ordinary skill in the art would employ.” KSR
Int'l Co. v. Teleflex Inc., 550 U.S. 398, 418 (2007). “If a person of ordinary
skill can implement a predictable variation, § 103 likely bars its
patentability.” Id. at 417.
As pointed out by Appellants, both Chu and Kakizaki seek to perform
bonding (Req. Reh’g I at 2). Moreover, bond strength was a known
parameter sought to be increased: As we pointed out, “[t]he Examiner relies
upon Kakizaki for a teaching of increasing bond strength by exposing an
insulating layer to be bonded to a nitrogen plasma or oxygen plasma to
activate its surface (Ans. 9; see also Kakizaki, col. 7, ll. 33-37 and col. 15, ll.
24-27).” (Decision 13.) Therefore, there was a reason within the art to
apply the plasma treatment step of Kakizaki to increase bond strength of
Appeal 2011-006333
Application 11/073,780

4
Chu’s bond. Consistent with the teachings of Kakizaki, “Appellants' own
Specification provides evidence that nitrogen plasma and oxygen plasma
activation increases bond strength between silicon-based layers,” as we
stated in our Decision, Id. This is not disputed by Appellants (Req. Reh’g
I).
Given that there was a reason within the art to increase bond strength
by adopting a plasma treatment that is the same or similar to the plasma
treatment Appellants themselves disclose will result in the claimed bond
strength, it is reasonable to conclude that a bond strength within the claimed
range would result in the process suggested by Chu in combination with
Kakizaki. When a claimed property value necessarily flows from the
conventional practice of the prior art, an express teaching of the property
value is not required in order to support a conclusion of obviousness. See In
re Kubin, 561 F.3d 1351, 1357 (Fed. Cir. 2009) (Prior art need not explicitly
discuss a property that is necessarily present in the thing formed when
following the process of the prior art); Perricone v. Medicis Pharmaceutical
Corp., 432 F.3d 1368, 1378 (Fed. Cir. 2005) (“If Pereira discloses the very
same methods, then the particular benefits must naturally flow from those
methods even if not recognized as benefits at the time of Pereira's
disclosure.”). It is knowledge within the art of the use of plasma treatment
to increase bond strength that forms the foundation of the rejection. The
result, i.e., the bond strength obtained, follows from that knowledge.
Plasma treatment was known to result in increased bond strength and
it appears reasonable that bond strengths within the claimed range would
have been predictable and expected. “The combination of familiar elements
Appeal 2011-006333
Application 11/073,780

5
according to known methods is likely to be obvious when it does no more
than yield predictable results.” KSR, 550 U.S. at 416.
Appellants have not convinced us that we overlooked or
misapprehended the requirements of claim 61 in sustaining the Examiner’s
rejection of claims 61, 64, and 66 as obvious over Chu in view of Kakizaki.

B. Request II: New Grounds of Rejection
In our Decision, we reversed the Examiner’s decision to reject claim
92 as anticipated by Chu, but affirmed the Examiner’s decision to reject
claims 93 and 110 as obvious over Chu (Decision 3-7). Because claims 93
and 110 depend from claim 92, we held that broader claim 92 was rendered
obvious based on the grounds of rejection applied against claims 93 and 110
(Decision 6-7). Because the Examiner did not list claim 92 as rejected as
obvious over Chu, we denominated our affirmance with regard to claim 92 a
new ground of rejection (id.).
In our Decision, we also determined that the Examiner made a faulty
finding in rejecting claims 92 and 93 as anticipated by Godbey3 (Decision
7). We, however, found that Godbey supported a finding of anticipation on
different grounds (Decision 8). Therefore, we affirmed the Examiner’s
rejection of claims 92 and 93, but denominated our affirmance as involving a
new ground of rejection (id.). The Examiner further rejected claim 110 as
obvious over Godbey (Ans. 6). We determined that Godbey anticipated
claim 110 and, therefore, entered a new ground of anticipation by Godbey
with respect to claim 110 (Decision 8).

3 Godbey et al., US 5,013,681, issued May 7, 1991.
Appeal 2011-006333
Application 11/073,780

6
With respect to the new grounds of rejection of claims 92, 93, and 110
as obvious over Chu and claims 92, 93, and 110 as anticipated by Godbey,
Appellants contend that we did not establish that the strained layer of either
of these references has an initial misfit dislocation density capable of being
reduced (Req. Reh’g II at 2).
Because Appellants’ arguments are confined to limitations found in
claim 92, we select that claim as representative for deciding the issue on
rehearing. Claim 92, with the limitations of particular interest highlighted
reads as follows:
92. A method for forming a structure, the method
comprising:
providing a first substrate having a dielectric layer
disposed thereon;
forming a semiconductor layer on a second substrate, the
semiconductor layer having an initial misfit dislocation density;
bonding the semiconductor layer to the dielectric layer;
removing the second substrate, the semiconductor layer
remaining bonded to the dielectric layer; and
reducing the initial misfit dislocation density in the
semiconductor layer.
(Claims App’x at Br. A4-A5 (emphasis added).)
With regard to Chu, there is no dispute that Chu epitaxially deposits a
strained layer 100 (Req. Reh’g II at 2; see also Decision 4; Br. 20-21). Or
that “[s]train occurs because the relatively high Ge concentration layer 140,
upon which layer 100 is deposited or grown, has larger lattice spacing than
Appeal 2011-006333
Application 11/073,780

7
the Si based layer 100.” (Decision 4-5, citing Br. 20-21; Chu, col. 6, ll. 33-
40 and Vossen,4 cited by Appellants at Br. 20.)
According to Appellants, a strained layer may be defect-free (Req.
Reh’g II at 2-3). Appellants cite a passage of Vossen, a reference supplied
in the Evidence Appendix of the Appeal Brief, as well as a statement in Chu
in support of an argument that Chu does not expressly or inherently teach
that layer 100 has an initial misfit dislocation density capable of being
reduced in a subsequent step (Req. Reh’g II at 3).
The passage of Chu referred to by Appellants is found in a discussion
of the Background of the Invention relating to carrier mobility in the
semiconductor raw material used in device fabrication (Chu, col. 1, ll. 16-
48). According to Chu, there is great difficulty in keeping carrier mobility
high in devices with small submicron transistors (Chu, col. 1, ll. 28-30).
Chu states that
A promising avenue toward better carrier mobility is to modify
slightly the semiconductor that serves as raw material for device
fabrication. It has been known, and recently further studied, that
tensilely strained Si has intriguing carrier properties. A Si layer
embedded in a Si/SiGe heterostructure grown by UHV-CVD
has demonstrated enhanced transport properties, namely carrier
mobilities, over bulk Si. In particular, a 90-95% improvement
in the electron mobility has been achieved in a strain Si channel
n-MOS (Metal Oxide Semiconductor transistor, a name with
historic connotations for Si Field- Effect-Transistors (FET)) in
comparison to a bulk Si n-MOS mobility. Similarly, a 30-35%
improvements in the hole carrier mobility has been obtained for
a strained Si channel p-MOS, in comparison to bulk silicon p-
MOS. The great difficulty lies in the production of a layer of
tensilely strained Si, or SiGe, that are of high enough

4 John L. Vossen & Werner Kern, THIN FILM PROCESSES II. 370-442
(Academic Press Inc. 1991).
Appeal 2011-006333
Application 11/073,780

8
crystalline quality, namely free of dislocations and other
defects, to satisfy the exceedingly elevated demands of
microelectronics.
(Chu, col. 1, ll. 30-48 (emphasis added).)
Chu conveys that it is difficult to produce strained layers that are free
of dislocations and other defects.
Vossen, in discussing pseudomorphic growth states that:
The growth of defect-free, generally lattice-mismatched
materials can take place only at specific compositions at which
the lattice parameter matches the substrate. There is an
intermediate growth regime for lattice-mismatched materials in
which defect-free, but highly strained thin layers of material can
be grown on a substrate [201, 202]. This growth mode is often
referred to as pseudomorphic growth and is schematically
shown in Fig. 11.
(Vossen, p. 413, ll. 5-11 (emphasis added).)
The caption for Figure 11 states that
Pseudomorphic growth of strained layer structures results in a
tetragonal distortion of the growing layer. Thick layers of
lattice-mismatched materials relax to their bulk lattice
parameter with the generation of extended defects.
(Fig. 11 caption.)
Vossen then describes the heteroepitaxial growth process, explaining
that distortion builds up a strain in the material as the layer is grown with a
large amount of energy stored in the structure (Vossen, p. 413, ll. 11-21).
According to Vossen,
The amount of energy stored in the structure eventually exceeds
a critical value required for the creation of strain-relieving
defects such as dislocations [201,202]. It is therefore possible
to grow highly mismatched materials, without extended defects,
to a limited thickness as shown in Fig. 11. The exact maximum
Appeal 2011-006333
Application 11/073,780

9
thickness prior to the formation of extended defects, commonly
referred to as the critical thickness, is materials-specific. Defect
generation proceeds through a nucleation, propagation, and
multiplication process. These processes are generally
temperature-dependent, as well as being dependent on the
material combination and, perhaps, the interfacial structure.
There have been several theoretical predictions for the critical
thickness as a function of strain or lattice mismatch. The initial
pioneering work of Matthews and Blakeslee serves as a good
working estimate of this critical thickness [201,202]. This
estimate is based on an energy balance between the strain
energy in the pseudomorphic structure and the energy
associated with the formation of the requisite number of misfit
dislocations. Other theoretical estimates are based on the
inclusion of dislocation nucleation effects or the consideration
of the interfacial stresses. The exact value of the critical
thickness allowable in a particular layered structure still
requires experimental verification.
(Vossen, p. 413, l. 21 to p. 414, l. 18 (emphasis added).)
Even though Vossen uses the term “defect-free” in the initial
discussion, in the context of the further discussion, and given the disclosure
in Chu that it is difficult to produce strained layers without defects, it is
reasonable to conclude that the strained layer one of ordinary skill in the art
would form when following the process of Chu would have some defects
and hence an initial misfit dislocation density.
Appellants cite to no evidence indicating that Chu’s etching process
would not reduce defect density in the same or similar manner as Appellants
etching process reduces such defects. In other words, there is no convincing
evidence that the initial defect density of Chu’s strained layer would not be
capable of being reduced.
With respect to Godbey, Appellants contend that “[n]othing in
Godbey expressly teaches a misfit dislocation density in layer 24.” (Req.
Appeal 2011-006333
Application 11/073,780

10
Reh’g II at 3.) Again, Appellants state that a strained layer may be defect-
free (id.). However, the evidence cited above indicates that forming defect-
free strained layers is difficult and it is reasonable to conclude that some
misfit defects would result when forming the strained layer 24 of Godbey.
The subject Request has been granted to the extent that the Decision
has been reconsidered, but denied with respect to making any changes
therein.

DENIED

cdc