Ex Parte Sheppard et al

Board of Patent Appeals and Interferences

Mar 27, 2008

10440727 (B.P.A.I. Mar. 27, 2008)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte SCOTT THOMAS SHEPPARD, SCOTT THOMAS ALLEN,
and JOHN WILLIAMS PALMOUR
____________

Appeal 2007-4373
Application 10/440,727
Technology Center 2800
____________

Decided: March 27, 2008
____________

Before JOSEPH F. RUGGIERO, ANITA PELLMAN GROSS, and
MAHSHID D. SAADAT, Administrative Patent Judges.

SAADAT, Administrative Patent Judge.

DECISION ON APPEAL
STATEMENT OF THE CASE
Appellants appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1, 2, 4-6, 9, 10, 13, 14, 17, 18, 21, and 22 which are all of the claims
Appeal 2007-4373
Application 10/440,727

pending in this application as claims 3, 7, 8, 11, 12, 15, 16, 19, and 20 have
been canceled. We have jurisdiction under 35 U.S.C. § 6(b).1
Appellants’ invention relates to a high electron mobility transistor
(HMET) where a two dimensional electron gas (2DEG) formed at the
junction of two semiconductor materials having different bandgap energies
provide the high electron mobility area. The transistor comprises a semi-
insulating silicon carbide substrate, an aluminum nitride buffer layer on the
substrate, an insulating gallium nitride layer on the buffer layer, an active
structure of aluminum gallium nitride on the gallium nitride layer, a
passivation layer on the aluminum gallium nitride active structure, and
respective source, drain, and gate contacts to the aluminum gallium nitride
active structure. (Spec. 3). According to Appellants, the passivation layer
appears to minimize or eliminate the problem of unterminated chemical
bonds at the surface by trapping a proportion of the electrons that would
otherwise flow in the 2DEG of a HEMT. (Spec. 7-8).
Claim 5, which is representative of the claims on appeal, reads as
follows:
5. A high electron mobility transistor for generating high
power signals within the RF spectrum, comprising:

a substrate that is sufficiently non-conductive to reduce
parasitic charge in the substrate;

a transition layer on said substrate;

a Group III nitride heterojunction on said transition layer, said
heterojunction comprising:

1 An oral hearing for this application was held on March 12, 2008.
2
Appeal 2007-4373
Application 10/440,727

an undoped gallium nitride layer on said transition layer, and

an active region on said gallium nitride layer, said active region
comprising:

a first undoped A1GaN layer on said gallium nitride layer;

a conductively doped A1GaN layer on said first undoped
A1GaN layer, and

a second undoped A1GaN layer on said doped A1GaN layer;

respective source, gate, and drain contacts electrically
connected to said heterojunction; and

a passivation layer on said active region, said passivation layer
at least partially covering said source, gate, and drain contacts;

wherein a two dimensional electron gas develops in said
gallium nitride layer directly beneath said first undoped A1GaN layer
of said active region, said two dimensional electron gas providing a
conductive path from said source contact to said drain contact.

The Examiner relies on the following prior art in rejecting the
appealed claims:
O’Neil US 5,739,557 Apr. 14, 1998
Razeghi US 5,831,277 Nov. 3, 1998

Kenneth K. Chu, High Speed High Power AlGaN/GaN Heterostructure
Field Effect Transistors with Improved Ohmic Contacts, IEEE/Cornell
Conference on Advanced Concepts, 399-406 (1997).

Claims 5, 9, 10, 13, 18, 21, and 22 stand rejected under 35 U.S.C.
§ 103(a) based upon the teachings of Chu and O’Neil.
Claims 1, 2, 4, 6, 14, and 17 stand rejected under 35 U.S.C. § 103(a)
based upon the teachings of Chu, O’Neil, and Razeghi.
3
Appeal 2007-4373
Application 10/440,727

Rather than reiterate the opposing arguments, reference is made to the
Briefs (Substitute Appeal Brief filed Aug. 26, 2006 and Reply Brief filed
May 15, 2006) and the Answer (mailed Nov. 14, 2006) for the respective
positions of Appellants and the Examiner.
We affirm.

ISSUES
1. Under 35 U.S.C. § 103(a), with respect to the appealed claims
5, 9, 10, 13, 18, 21, and 22, would the ordinarily skilled artisan have found it
obvious to combine Chu and O’Neil to render the claimed invention
unpatentable?
2. Under 35 U.S.C § 103(a), with respect to the appealed claims 1,
2, 4, 6, 14, and 17, would one of ordinary skill in the art at the time of the
invention have found it obvious to combine Chu, O’Neil, and Razeghi to
render the claimed invention unpatentable?

FINDINGS OF FACT
The following findings of fact (FF) are relevant to the issue involved
in the appeal and are supported by substantial evidence.
O’Neil
1. O’Neil relates to a heterostructure field effect transistor that
includes improved ohmic contact metallization (col. 1, ll. 6-8).
2. O’Neil’s disclosure addresses three problems with existing
heterostructure transistors: 1) ohmic contact metallization, 2) oxidization of
the AlGaAs surface when it is exposed to air, 3) controlled etching of the
highly doped cap layer (col. 1, ll. 23-44).
4
Appeal 2007-4373
Application 10/440,727

3. The transistor stack is capped with a GaAs passivation layer 20,
an AlAs etch stop layer 22, and an InGaAs cap layer 24 (col. 3, ll. 12-14).
4. The passivation layer 20 prevents oxidation of AlGaAs layer 18
during the etch process (col. 4, ll. 1-2).
5. A dielectric layer 42, formed of Si3N4, is deposited over the
stack to define the ohmic contact areas 50 (Fig. 4; col. 4, ll. 5-15).
6. A thick dielectric layer 60, formed of SiO4, Si3N4, or some
other oxynitride material, is deposited to further define via openings 62 for
deposition of interconnect metal (Fig. 6; col. 4, ll. 34-41).
Razeghi
7. Razeghi provides a GaN-based composition in a superlattice
structure (Abstract).
8. The substrates used by Razeghi are disclosed to be GaAs, Si,
Al2O3, MgO, SiC, ZnO, LiGaO2, LiAlO2, MgAl2O4 or GaN (col. 3, ll. 37-
39).
9. Razeghi further discloses that before forming the AlGaAs
active layers, a buffer layer of AlN, a contact layer, and n-doped
confinement layers of GaN and AlGaN are deposited on the substrate (col. 4,
ll. 44-47).

PRINCIPLES OF LAW
The test for obviousness is what the combined teachings of the
references would have suggested to one of ordinary skill in the art. See In re
Kahn, 441 F.3d 977, 987-88 (Fed. Cir. 2006), In re Young, 927 F.2d 588,
591 (Fed. Cir. 1991) and In re Keller, 642 F.2d 413, 425 (CCPA 1981).
5
Appeal 2007-4373
Application 10/440,727

“Section 103 forbids issuance of a patent when ‘the differences
between the subject matter sought to be patented and the prior art are such
that the subject matter as a whole would have been obvious at the time the
invention was made to a person having ordinary skill in the art to which said
subject matter pertains.’” KSR Int’l Co. v. Teleflex Inc., 127 S. Ct. 1727,
1734 (2007).

ANALYSIS
1. Rejection of claims 5, 9, 10, 13, 18, 21, and 22
The Examiner reads substantially all of the claimed features related to
the stacked layers on the AlGaN/GaN heterostrucutre field effect transistor
(HFET) of Chu and further relies on O’Neil for adding a passivation layer on
the active region of the transistor for partially covering the source, drain, and
gate contacts of the HFET (Ans. 4-5). Appellants contend that the devices
disclosed in Chu and O’Neil are incompatible since they use different cap
layers and only O’Neil includes a passivation layer and an etch stop layer
(App. Br. 11-12).
Initially, we note that Appellants’ arguments appear to be based on
using the passivation layer 20 of O’Neil in the device of Chu and do not
address the silicon nitride dielectric layers 42 and 60, which are what the
Examiner characterizes as the passivation layer. As argued by the Examiner
(Ans. 7-8), the dielectric layers 42 and 60 are formed on the active region
and at least partially cover the source, drain, and gate contacts (FF 5-6).
While O’Neil calls layer 20 a passivation layer (FF 3), the function and
material of layer 20 is so different from a silicon nitride layer that it cannot
be what the Examiner cited as a “silicon nitride top passivation layer” (Ans.
6
Appeal 2007-4373
Application 10/440,727

7). The only silicon nitride layers in O’Neil are dielectric layers 42 and 60
which are formed of Si3N4, or some oxynitride material (FF 5-6). It is also
important to note that calling a dielectric top layer a “passivation layer” is
commonly known to those skilled in the art. This characterization is
consistent with Appellants’ own Specification (¶ [0030]) wherein the
passivation layer 23 is indicated as a silicon nitride or silicon dioxide layer.
We therefore do not agree with Appellants that the stated differences
between the structure disclosed by O’Neil and Chu are significant enough to
discourage their combination. In fact, consistent with KSR, one of ordinary
skill in the art would have found it obvious to use the dielectric passivation
layer of O’Neil over the structure disclosed by Chu to benefit from the
protection against environmental contaminants and mechanical damages that
otherwise exposed surfaces of the device would have likely incurred. We
also agree with the Examiner (Ans. 8) that the protection provided by the
passivation layer against the above mentioned elements would be available
to HFET structures regardless of the type and thickness of the cap layer and
any other layers under the dielectric passivation layer.
Based on our analysis of Chu and O’Neil, we agree with the
Examiner’s assessment of the teaching value of O’Neil regarding using a
passivation layer as a solution with predictable results. Therefore, it is our
view that the Examiner has set forth a prima facie case of obviousness,
which Appellants have not successfully overcome by presenting either
convincing arguments or evidence to support the patentability of the claims.
Additionally, Appellants have presented no arguments in the Reply Brief to
indicate that the teachings in the applied references would have discouraged
a person of ordinary skill from applying the dielectric passivation layer
7
Appeal 2007-4373
Application 10/440,727

taught in O’Neil in combination with Chu to rebut the Examiner’s prima
facie case.
Therefore, in view of the analysis above, we find that one of ordinary
skill in the art would have combined Chu and O’Neil in order to benefit
from the protection provided by a dielectric passivation layer. Accordingly,
since we find no error in the Examiner’s position, we sustain the 35 U.S.C.
§ 103(a) rejection of independent claim 5, as well as claims 9, 10, 13, 18, 21,
and 22 argued together with claim 5 as one group, over Chu and O’Neil.
2. Rejection of claims 1, 2, 4, 6, 14, and 17
With respect to the rejection of independent claim 1, Appellants take
the position that Razeghi’s device is incompatible with the structures
disclosed in Chu and O’Neil and, therefore, could not be combined (App.
Br. 13-14). Appellants further contend that the Examiner’s selection of SiC
from the list of substrates disclosed in Razeghi is not supported by any
reason based on a teaching or suggestion (App. Br. 14).
The Examiner relies on using the SiC substrate as a common substrate
disclosed to be suitable for epitaxially growing an AlN buffer layer before
forming a GaN-based layered structure (Ans. 10). The Examiner further
points out that the choice of substrate does not depend on what additional
layers are further stacked over the substrate and the buffer layer, whereas
other factors such as cost and availability do matter (id.).
We agree with the Examiner’s rationale in combining Razeghi with
Chu and O’Neil and using a SiC substrate since the different substrates
disclosed by Razeghi are all suggested to be suitable for forming a GaN
active layer (FF 7-8). Again, we find the rationale for the combination
proposed by the Examiner to be consistent with KSR, which held obvious the
8
Appeal 2007-4373
Application 10/440,727

combination of familiar elements according to known methods when the
results are predictable. Here, a SiC substrate is used for forming the stacked
layers of a heterostructure device where an AlN buffer layer and a GaN
active layer, similar to those disclosed and claimed by Appellants, are
formed on the substrate. The predictability of the proposed combination is
further supported by the fact that Razeghi lists the SiC substrate as equal and
compatible with other semi-insulating substrates such as GaAs used in
O’Neil and Al2O3 (or sapphire) used in Chu.
For all of the previously discussed reasons and in view of our analysis
of Razeghi, we find no error in the Examiner’s position that using a SiC
substrate as the semi-insulating substrate in combination with Chu and
O’Neil would be recognized by the skilled artisan as an obvious alternative
substrate. Therefore, we sustain the 35 U.S.C. § 103(a) rejection of claim 1,
as well as claims 2, 4, 6, 14, and 17 which are not argued separately, over
Chu, O’Neil, and Razeghi.

DECISION
The decision of the Examiner rejecting claims 1, 2, 4-6, 9, 10, 13, 14,
17, 18, 21, and 22 is affirmed.

9
Appeal 2007-4373
Application 10/440,727

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)(iv).

AFFIRMED

tdl/gw

SUMMA, ALLAN & ADDITON, P.A.
11610 NORTH COMMUNITY HOUSE ROAD SUITE 200
CHARLOTTE, NC 28277
10