Ex Parte Dautet et al

Patent Trial and Appeal Board

May 13, 2016

11725661 (P.T.A.B. May. 13, 2016)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
11/725,661 03/20/2007 Henri Dautet 41067-6106 6544
57449 7590 05/13/2016
SHEEHAN PHINNEY BASS & GREEN, PA
c/o PETER NIEVES
1000 ELM STREET
MANCHESTER, NH 03105-3701
EXAMINER
SHINGLETON, MICHAEL B
ART UNIT PAPER NUMBER
2815
MAIL DATE DELIVERY MODE
05/13/2016 PAPER
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________________

Ex parte HENRI DAUTET and RICHARD SEYMOUR
____________________

Appeal 2014-001887
Application 11/725,661
Technology Center 2800
____________________

Before ADRIENE LEPIANE HANLON, CATHERINE Q. TIMM, and
JAMES C. HOUSEL, Administrative Patent Judges.

TIMM, Administrative Patent Judge.

DECISION ON APPEAL1

STATEMENT OF CASE
Appellants2 appeal the Examiner’s decision to reject claims 1, 6–8,
10–12, 14–21, 28, 29, and 31–37 under 35 U.S.C. § 103(a). Claims 1, 6–8,

1 In our opinion below, we refer to the Specification filed March 20, 2007
(Spec.), Final Office Action mailed Dec. 10, 2012 (Final), the Appeal Brief
filed June 4, 2013 (Appeal Br.), the Examiner’s Answer mailed Aug. 15,
2013 (Ans.), and the Reply Brief filed Oct. 15, 2013 (Reply Br.).
2 Appellants identify the real party in interest as Excelitas Canada Inc.
Appeal Br. 1.
Appeal 2014-001887
Application 11/725,661

2
10–12, 14–21, 28, 29, and 31–35 stand rejected as obvious over Kobayashi3
in view of McCann4 and Neilson.5 To reject claims 36 and 37, the Examiner
adds Nishizawa6 or Bui7 or Veliadis.8 We have jurisdiction under
35 U.S.C. §§ 6(b) and 134(a).
We AFFIRM.

The claims are directed to an avalanche photodiode with an active
substrate bonded to a handle substrate. The active substrate includes a
heavily doped layer at the bonded surface. Claim 1, with emphasis on
limitations at issue in the appeal, is reproduced below:
1. An avalanche photodiode comprising:
an electrooptically active substrate, wherein the active
substrate includes a portion of a first wafer, wherein the active
substrate has a lightly doped silicon region, and wherein the
active substrate includes a heavily doped layer at a first silicon
surface of the active substrate;
a handle substrate including heavily doped silicon,
wherein the handle substrate includes a portion of a second
wafer, and wherein a second silicon surface of the handle
substrate is directly bonded to the first silicon surface of the
active substrate; and
an avalanche photodiode active area formed in the active
substrate including a high field region for generating avalanche
current gain,

3 Kobayashi, US 5,596,186, patented Jan. 21, 1997.
4 McCann et al., US 2004/0087109 A1, pub. May 6, 2004.
5 Neilson et al., US 6,054,369, pat. Apr. 25, 2000.
6 Nishizawa, US 4,727,649, pat. Mar. 1, 1988.
7 Bui et al., US 7,655,999 B2, pat. Feb. 2, 2010.
8 Veliadis, US 7,863,647 B1, pat. Jan. 4, 2011.
Appeal 2014-001887
Application 11/725,661

3
wherein the heavily doped layer is adapted to improve
interface quality between the active substrate and the handle
substrate.
Claims Appendix of the Appeal Br. (emphasis added).

OPINION
As Appellants focus their arguments on the rejection of claim 1, we
do the same.
Kobayashi teaches an avalanche photodiode with an n+ silicon
substrate 10 upon which is grown an n- epitaxial growth layer 11 that serves
as the active substrate. Kobayashi, col. 6, l. 66–col. 7, l. 25; Fig. 1. The two
layers of silicon are present but not made by the direct bonding process
recited in claim 1. Kobayashi does not expressly teach a heavily doped layer
at the surface of the active substrate adapted to improve interface quality. In
Kobayashi, the n+ substrate 10 has a higher doping concentration than the n-
epitaxial growth layer 11. Kobayashi, col. 7, ll. 3–6 (layer 11 has an
impurity concentration of 1x1014 cm-3 and substrate 10 has an impurity
concentration of 1x1017 cm-3).
The Examiner advances several bases for rejection.
First, although Kobayashi’s photodiode is not made by directly
bonding an active substrate to a handle substrate, the Examiner finds that the
process differences do not patentability distinguish the claimed photodiode
from the one taught by Kobayashi. Final 4–5. To support this rejection, the
Examiner relies upon McCann as extrinsic evidence supporting the finding
that the structure is not patentably different. Id.
In another basis for rejection, the Examiner concludes it would have
been obvious to have replaced the substrate of Kobayashi with two wafers
Appeal 2014-001887
Application 11/725,661

4
bonded together given that the two alternative structures are art-recognized
equivalents as taught by McCann. Final 5–7. The Examiner relies upon
Neilson as teaching that it was known in the prior art to heavily dope prior to
bonding to improve the interface. Final 7.
We will consider the issues for the obviousness of substituting
Kobayashi’s epitaxial growth substrate 11 and substrate 10 with McCann’s
bonded substrates first.
McCann is directed to a method of direct bonding two silicon wafers.
McCann ¶ 1. McCann indicates that the method can be used to form a range
of devices including PIN diodes and bipolar devices where two silicon layers
of different resistivity are required. McCann ¶¶ 2–3.
McCann describes the direct bonding method as an alternative that
overcomes some of the problems with the epitaxial growth method.
McCann ¶ 3. McCann fusion bonds the two substrates together in a way that
minimizes interfacial oxide and stress at the bond interface. McCann ¶¶ 9–
12. Because oxide is minimized at the bond interface, and effectively
eliminated, dopant diffused into one of the wafers diffuses through the bond
interface to the other wafer without hindrance. McCann ¶ 82.
Appellants contend that the ordinary artisan would have had no reason
to modify Kobayashi’s avalanche photodiode in view of McCann’s “fairly
generic disclosure,” and the Examiner has not presented a compelling reason
why the modification would have been obvious. Appeal Br. 13–14. But
McCann’s “fairly generic disclosure” specifically identifies the direct
bonding method as a substitute for epitaxially grown layers on silicon
substrates to form bipolar devices where two silicon layers of different
resistivity are required. The photodiode of Kobayashi is just such a
Appeal 2014-001887
Application 11/725,661

5
structure. Moreover, McCann discusses the pros and cons of using two
directly bonded substrates instead of an epitaxially grown layer on a
substrate. Both the direct bonding and epitaxial growth methods were
known in the art, and the evidence indicates that those of ordinary skill in the
art would have reasonably expected the direct bonding method to solve the
problems with epitaxial growth discussed in McCann. “The combination of
familiar elements according to known methods is likely to be obvious when
it does no more than yield predictable results.” KSR Int'l Co. v. Teleflex Inc.,
550 U.S. 398, 416 (2007).
A preponderance of the evidence supports the Examiner’s finding of a
reason to substitute the epitaxial growth method of Kobayashi with the direct
bonding method of McCann.
Appellants further contend that none of the prior art references “would
have suggested adding a ‘heavily doped layer’ to ‘improve interface quality’
between Kobayashi’s epitaxial growth layer 11 to Kobayashi’s silicon
substrate 10, as recited by claim 1.” Appeal Br. 14; Reply Br. 6. But the
prior art need not suggest adding a ‘heavily doped layer’ to ‘improve
interface quality’ in order to meet the requirements of the claim. This is
because the claim is directed to an avalanche photodiode; a structure. Thus,
the claim language “the heavily doped layer is adapted to improve interface
quality between the active substrate and the handle substrate” must be read
to determine how the language limits the structure of the photodiode; the
layer need not be added for the reason set forth by the claim. See In re
Michlin, 256 F.2d 317, 320 (CCPA 1958) (“It is well settled that
patentability of apparatus claims must depend upon structural limitations and
not upon statements of function.”). At best, the “improve interface quality”
Appeal 2014-001887
Application 11/725,661

6
language may limit the range of dopant concentrations within the heavily
doped layer. We turn to the Specification to determine whether and how the
language limits the dopant concentration.
In our search to determine the concentrations of dopants that would
“improve interface quality” and thus be encompassed by the language of the
claim, we find little guidance in the Specification. All we find is the naked
statement that “[h]eavily doped silicon layer 20 improves interface quality
between substrates 12 and 14.” Spec. 8:19–22. The Specification does not
discuss what concentrations of dopant come under the umbrella of “heavily
doped,” what concentrations “improve interface quality,” or even what
property improvements constitute an improvement in interface quality. The
Specification also states that, alternatively, an oxide layer may be used to
improve interface quality. Spec. 8:23–9:2. Thus, what constitutes an
improvement to interface quality is further clouded. In the absence of any
indication that a particular concentration is required to improve interface
quality, we determine the active layer need only have a heavily doped layer
having some concentration of the dopant.
Kobayashi dopes substrate 10 to a concentration of 2x1017 cm-3, and
epitaxial growth layer 11 to a concentration of 1x1014 cm-3. Kobayashi, col.
7, ll. 4–6. The Examiner finds, and Appellants do not dispute, that substrate
10 of Kobayashi is heavily doped. Final 6; Appeal Br. 7, 9–12, 14. McCann
suggests replacing the epitaxially grown layer 11 on substrate 10 with two
substrates such that there would be two substrates in place of epitaxially
grown layer 11 bonded to a substrate 10. McCann teaches that dopant
diffused into one of the substrates diffuses unhindered through the bond
interface to the other substrate. McCann ¶ 82. Diffusion of dopant from the
Appeal 2014-001887
Application 11/725,661

7
heavily doped substrate 10 of Kobayashi into the bonded substrate replacing
epitaxial layer 11 (per the bonding method of McCann) would result in a
heavily doped layer in the active substrate at the concentrations as required
by the “improve interface quality” language of claim 1.
Moreover, the Examiner relies upon Neilson as evidence that it was
conventional to heavily dope prior to bonding to improve the interface, and
concludes it would have been obvious to one of ordinary skill to further
employ such doping “to control the density of the recombination centers to
be that desired as it is taught by Neilson.” Final 7. Appellants do not
address this finding and conclusion, but merely state that none of the
references suggest what is claimed. Appeal Br. 14. Appellants have not
identified a reversible error in the Examiner’s finding and conclusion in this
regard.
As to the Examiner’s alternate basis of rejection that relies upon
Kobayashi with McCann used as extrinsic evidence, Appellants have not
convinced us of a reversible error in this basis for rejection. Appellants as
well as Kobayashi and McCann all create a structure with two layers of
material, each material being silicon. Whether directly bonding or
epitaxially growing silicon onto silicon, the result is a two-layer silicon
structure. Where one defines the boundary between the heavily doped lower
substrate and the heavily doped layer in the active substrate does not seem to
result in a patentable distinction between the claimed and prior art devices.
McCann indicates that dopant diffuses across the interface.
Appellants rely upon arguments addressed above for all the claims
and both rejections. Appellants have not identified a reversible error in
either rejection.
Appeal 2014-001887
Application 11/725,661

8
CONCLUSION
We sustain the Examiner’s rejections.

DECISION
The Examiner’s decision is affirmed.

TIME PERIOD FOR RESPONSE
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).

AFFIRMED