Hanwha Q CELLS & Advanced Materials Corp.

Patent Trials and Appeals Board

Dec 3, 2020

IPR2019-01072 (P.T.A.B. Dec. 3, 2020)

Trials@uspto.gov Paper 38
571-272-7822 Date: December 3, 2020

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

LONGI GREEN ENERGY TECHNOLOGY CO., LTD., LONGI SOLAR
TECHNOLOGY CO., LTD., LONGI (H.K.) TRADING LTD., LONGI
(KUCHING) SDN. BHD., LONGI SOLAR TECHNOLOGY (TAIZHOU)
CO., LTD., LONGI SOLAR TECHNOLOGY (ZHEJIANG) CO., LTD.,
LONGI SOLAR TECHNOLOGY (HEFEI) CO., LTD., AND LONGI
SOLAR TECHNOLOGY (US), INC.,
Petitioner,

v.

HANWHA Q CELLS & ADVANCED MATERIALS CORPORATION,
Patent Owner.
____________

IPR2019-01072
Patent 9,893,215 B2
____________

Before CHRISTOPHER L. CRUMBLEY, JEFFREY W. ABRAHAM, and
JULIA HEANEY, Administrative Patent Judges.

CRUMBLEY, Administrative Patent Judge.

JUDGMENT
Final Written Decision
Determining All Challenged Claims Unpatentable
37 C.F.R. § 318(a)
IPR2019-01072
Patent 9,893,215 B2
2
I. INTRODUCTION
In this inter partes review, instituted pursuant to 35 U.S.C. § 314,
LONGi Green Energy Technology Co., LONGi Solar Technology Co., Ltd.,
LONGi (H.K.) Trading Ltd., LONGi (Kuching) Sdn. Bhd., LONGi Solar
Technology (Taizhou) Co., Ltd., LONGi Solar Technology (Zhejiang) Co.,
Ltd., LONGi Solar Technology (Hefei), Co., Ltd., and LONGi Solar
Technology (US), Inc. (collectively “Petitioner”) challenge the patentability
of claims 12–14 of U.S. Patent No. 9,893,215 (Ex. 1001, “the ’215 patent”),
owned by Hanwha Q Cells & Advanced Materials Corporation (“Patent
Owner”).
We have jurisdiction under 35 U.S.C. § 6. This Final Written
Decision, issued pursuant to 35 U.S.C. § 318(a) and 37 C.F.R. § 42.73,
addresses issues and arguments raised during trial. For the reasons
discussed below, we determine that Petitioner has shown by a preponderance
of the evidence that claims 12–14 of the ’215 patent are unpatentable.
A. Procedural History
On May 13, 2019, Petitioner requested an inter partes review of
claims 12–14 of the ’215 patent. Paper 1 (“Pet.”). Patent Owner filed a
Preliminary Response. Paper 7. On December 4, 2019, we instituted an
inter partes review of the challenged claims. Paper 12 (“Decision on
Institution” or “Dec. on Inst.”). Following institution, Patent Owner filed a
Patent Owner Response (Paper 18, “PO Resp.”), Petitioner filed a Reply
(Paper 26, “Pet. Reply”), and Patent Owner filed a Sur-Reply (Paper 29,
“PO Sur-Reply”).
Petitioner relies upon the declaration testimony of Dr. Sanjay
Banerjee (Ex. 1003) to support the Petition; Patent Owner took cross-
IPR2019-01072
Patent 9,893,215 B2
3
examination of Dr. Banerjee via deposition and submitted the transcript (Ex.
2007). Patent Owner relies upon the declaration testimony of Dr. Michael S.
Lebby (Ex. 2008); Petitioner took cross-examination of Dr. Lebby via
deposition and submitted the transcript (Ex. 1053).
Petitioner filed a Motion to Exclude, seeking to exclude certain
paragraphs of Dr. Lebby’s Declaration. Paper 32, “Mot.” Patent Owner
filed an Opposition (Paper 34, “Mot. Opp.”) and Petitioner filed a Reply
(Paper 35, “Mot. Reply”).
Oral hearing was requested by both parties. Papers 27, 28. A
consolidated oral hearing for this proceeding and Case IPR2019-01145, also
involving the ’215 patent, was held on September 9, 2020. A transcript of
the hearing in this proceeding has been entered into the record. Paper 37
(“Tr.”).
B. Related Proceedings
Petitioner and Patent Owner identify the following pending litigation
involving the ’215 patent: Certain Photovoltaic Cells and Products
Containing Same, Inv. No. 337-TA-1151 (U.S.I.T.C.) and Hanwha Q
CELLS & Advanced Materials Corp. v. LONGi Green Energy Technology
Co., Ltd. et al., No. 1:19-cv-00451-MN (D. Del.). Pet. 5–6; Paper 6, 1.
Patent Owner also identifies the following pending litigation involving the
’215 patent: Hanwha Q CELLS & Advanced Materials Corp. v. JinkSolar
Holding Co. Ltd. et al., No. 1:19-cv-00450-MN (D. Del.) and Hanwha Q
CELLS & Advanced Materials Corp. v. REC Solar Holdings AS et al., No.
1:19-cv-00452-MN (D. Del.). Paper 6, 1. As noted above, the ’215 patent is
the subject of another petition for inter partes review that was concurrently
filed: IPR2019-01145. Id.
IPR2019-01072
Patent 9,893,215 B2
4
C. The ’215 Patent
The ’215 patent, entitled “Method For Manufacturing a Solar Cell
With a Surface-Passivating Dielectric Double Layer, And Corresponding
Solar Cell,” issued February 13, 2018 and claims priority to a foreign
application filed November 14, 2007. Ex. 1001, [30], [45], [54]. According
to the ’215 patent, “[a] key requirement for achieving high degrees of
efficiencies in solar cells is very effective suppression of surface
recombination losses.” Id. at 1:14–16. The ’215 patent explains that, in
order to achieve this purpose, the surface of solar cells should be passivated
as effectively as possible, so that charge carrier pairs which are generated
inside the solar cell by incident light and which diffuse to the surfaces of the
solar cell substrate do not recombine at the solar cell surface, where they
would no longer be available to improve the efficiency of the solar cell. Id.
at 1:16–22. Good passivating results can be achieved using aluminum1
oxide layers which are deposited by means of atomic layer deposition
(ALD). Id. at 2:19–21. However, in atomic layer deposition, only a single
molecular layer of the material to be deposited is generally deposited on the
substrate surface within each deposition cycle. Id. at 2:23–25. Because a
deposition cycle typically lasts about 0.5 to 4 seconds, low deposition rates
are generally obtained using atomic layer deposition, where the thickness of
the deposition of the aluminum oxide layers may not be suitable for use as
an antireflection layer or as a back reflector. Id. at 2:25–32.
The ’215 patent describes a method for manufacturing silicon solar
cells with a dielectric passivating layer for reducing surface recombination

1 The ’215 patent refers to the element as “aluminium.” We substitute the
typical American spelling, “aluminum.”
IPR2019-01072
Patent 9,893,215 B2
5
losses in light of the challenges using an ALD process described above. Id.
at 2:57–60. The dielectric passivating layer is composed of two partial
layers: a first dielectric layer (i.e., very thin aluminum oxide containing
layer), which is formed by ALD, and a second dielectric layer (i.e., a thicker
layer made of silicon oxide, silicon nitride, or silicon carbide, for example),
which can be deposited on the first dielectric layer by means of plasma
enhanced chemical vapor deposition (“PECVD”), for example. Id. at 2:61–
67.
According to one embodiment, the first dielectric layer is deposited at
a thickness of less than 50 nm, preferably less than 30 nm, and more
preferably less than 10 nm. Id. at 5:40–43. According to the ’215 patent,
even at a very low thickness, the first dielectric layer offers very good
surface-passivating properties on account of its high quality which can be
achieved as a result of the ALD. Id. at 5:43–45. Further, according to the
embodiment, the second dielectric layer is deposited at a thickness of more
than 50 nm, preferably more than 100 nm, and more preferably more than
200 nm. Id. at 5:51–54. As described in the ’215 patent, the thickness of the
second dielectric layer can be adapted to its task as an antireflection layer.
Id. at 5:54–56.
Figure 1 of the ’215 patent is reproduced below:

IPR2019-01072
Patent 9,893,215 B2
6
Figure 1 is “a solar cell according to one embodiment of the present
invention.” Id. at 6:44–45. The ’215 patent describes that the solar cell
includes silicon wafer 1. Id. at 6:51–57. The solar cell further includes
aluminum oxide (Al2O3) layer 3 that is deposited onto the silicon wafer via
an ALD coating process. Id. at 6:63–7:11. The solar cell additionally
includes silicon oxide thin layer 5 that is subsequently coated onto the Al2O3
layer. Id. at 7:30–35. The silicon oxide layer, which is deposited by means
of PECVD, has a very high hydrogen content and therefore serves as a
source of hydrogen. Id. at 7:65–8:3. The hydrogen diffuses through the
ultrathin Al2O3 layer and passivates unsaturated silicon bonds (“dangling
bonds”) at the Si/Al2O3 interface, leading to very good surface passivation.
Id. at 8:3–7.
D. The Challenged Claims
Petitioner challenges claims 12–14 of the ’215 patent. Pet. 1. Claim
12, reproduced below, is illustrative of the subject matter of the challenged
claims:
12. A solar cell comprising:
a silicon substrate;
a first dielectric layer comprising alumin[]um oxide on a surface
of the silicon substrate; and
a second dielectric layer directly on a surface of the first
dielectric layer, materials of the first dielectric layer and the
second dielectric layer differing and hydrogen being embedded
into the second dielectric layer;
wherein the first dielectric layer has a thickness of less than 50
nm and is interposed between the surface of the silicon substrate
and the second dielectric layer.
Ex. 1001, 9:35–10:9.
IPR2019-01072
Patent 9,893,215 B2
7
Claims 13 and 14 depend from claim 12. Claim 13 requires that the
second dielectric layer comprise silicon nitride, silicon oxide, or silicon
carbide, while claim 14 further limits the thickness of the first dielectric
layer to less than 30 nm.
E. Instituted Grounds
We instituted an inter partes review of all claims challenged in the
Petition on the following grounds of unpatentability, each alleging
obviousness of the claims under 35 U.S.C. § 103(a)2:
Claims Challenged Reference(s)
12–14 Bhattacharyya3
12–14 Bhattacharya, Duerinckx,4 Hoex 20065
12–14 Hagino6
12–14 Hagino, Duerinckx, Hoex 2006

2 The Leahy-Smith America Invents Act (“AIA”), Pub. L. No. 112-29, 125
Stat. 284 (September 16, 2011), included revisions to 35 U.S.C. § 103 that
became effective on March 16, 2013. Because the ’215 patent issued from
an application filed before March 16, 2013, we apply the pre-AIA versions
of the statutory bases for unpatentability.
3 U.S. Patent Application Publication No. 2006/0102972 to Bhattacharyya
(Ex. 1005).
4 F. Duerinckx and J. Szlufcik, “Defect passivation of industrial
multicrystalline solar cells based on PECVD silicon nitride,” Solar Energy
Materials & Solar Cells 72 (2002) 231–246 (Ex. 1008).
5 B. Hoex, et al. “Ultralow surface recombination of c-SI substrates
passivated by plasma-assisted atomic layer deposited Al2O3,” Applied
Physics Letters 89, 042112 (2006) (Ex. 1007). We adopt Petitioner’s
convention of referring to the reference as “Hoex 2006,” to differentiate
from other articles by Hoex also made of record.
6 U.S. Patent Application Publication No. 2004/0112426 A1 to Hagino (Ex.
1006).
IPR2019-01072
Patent 9,893,215 B2
8
Pet. 6. Petitioner alleges that Bhattacharyya and Hagino are prior art to the
’215 patent under either 35 U.S.C. § 102(a) or (b), while Duerinckx and
Hoex 2006 are prior art under § 102(b). Pet. 19, 23, 27, 29. Patent Owner
has not contended that any of the asserted references are not prior art to the
’215 patent.
II. MOTION TO EXCLUDE
Petitioner moved to exclude paragraphs 1–24, 27–41, 43–77, 80–82,
85–105, 108, 111–12, 114–15, 119–30, 132–33, 135–43, 145–46, 148–50,
153–72, 175–76, 178–85, and 189–90 of Dr. Lebby’s Declaration (Ex.2008)
under Federal Rules of Evidence 401–03. Mot. 1. According to Petitioner,
these portions of Dr. Lebby’s testimony are either irrelevant to the issues
involved in the trial, or are unduly prejudicial, because they are “either not
cited at all in Patent Owner’s Response” or are “cited only as part of a
sweeping range of paragraphs.” Id. In its Reply, Petitioner also alleges that
any reliance on these paragraphs amounts to improper incorporation by
reference of arguments from one paper into another, in violation of our
Rules. Mot. Reply 1 (citing 37 C.F.R. § 42.6(a)(3)).
Patent Owner responds that the Board has never adopted a practice of
excluding all portions of a declaration that are “vaguely cited” in a paper, or
not cited at all, and that applying such a rule would be unworkable. Mot.
Opp. 2–3. In addition, the paragraphs to which Petitioner objects include
routine statements such as Dr. Lebby stating his name and retention by
Patent Owner (¶ 1) or declaring that his testimony is truthful (¶ 190), or
several blank “reserved” paragraphs. Id. at 1. Patent Owner argues that
these paragraphs, or others that relate to Dr. Lebby’s work and experience,
compensation, or the documents he reviewed, are not irrelevant or
IPR2019-01072
Patent 9,893,215 B2
9
prejudicial simply because they were not cited directly in a paper. Id. at 1–2.
According to Patent Owner, Petitioner has failed to establish that the
objected-to paragraphs are inadmissible under the Rules of Evidence, or that
the proper remedy would be exclusion of the paragraphs from the record. Id.
at 5–6 (citing Amerigen Pharm. Ltd. et al v. Jannsen Oncology, Inc.,
IPR2016-00286, Paper 86 at 45–46 (PTAB Jan. 17, 2018) (“We are unaware
of any legal basis for striking uncited paragraphs of an expert declaration.”).
Upon review, Petitioner’s Motion to Exclude fails to establish that
Petitioner is entitled to the remedy it seeks. To the extent the Motion seeks
exclusion of evidence under FRE 403 as overly prejudicial, Petitioner does
not adequately explain what prejudice it would suffer if we were to simply
disregard or give little weight to the uncited paragraphs of Dr. Lebby’s
testimony, rather than excluding them from the record. Indeed, as the Board
is not a jury, it would appear that the risk of unfair prejudice against which
Rule 403 guards is diminished, if not eliminated entirely. See, e.g., Schultz
v. Butcher, 24 F.3d 626, 632 (4th Cir. 1994) (“in the context of a bench trial,
evidence should not be excluded under 403 on the ground that it is unfairly
prejudicial”); see also Gulf States Utils. Co. v. Ecodyne Corp., 635 F.2d 517,
519 (5th Cir. 1981). The Board is capable of hearing relevant evidence and
weighing its probative value, including determining the weight to be given
Dr. Lebby’s testimony or disregarding it entirely.
Nor is it clear that the objected-to paragraphs are irrelevant to the
issues involved in the trial merely because they were uncited or cited as part
of a range of paragraphs. Some of the paragraphs provide important context
or foundation for the remainder of Dr. Lebby’s testimony, such as his
experience or the evidence he reviewed in forming his opinions. That
testimony does not become irrelevant simply because it is not cited in a
IPR2019-01072
Patent 9,893,215 B2
10
paper. We also note that some of the paragraphs that Petitioner objects to as
being cited in “sweeping range[s] of paragraphs” span three or five
paragraphs (Mot. 2), which do not appear to be categorically problematic.
Indeed, Petitioner concedes that some paragraphs from these ranges were
cited with more specificity in Patent Owner’s Sur-Reply. Id. at 5.
Evidence is relevant if it “has any tendency to make a fact more or
less probable than it would be without the evidence” and that fact is material
to determining the case. FRE 401. We have no reason to conclude the
paragraphs of Dr. Lebby’s testimony objected to by Petitioners fail to meet
this broad standard. While the Board has, in past cases, excluded exhibits
that a party submitted as “background” but never cited to any part of (see SK
Innovation Co., Ltd. v. Celgard, LLC, IPR2014-00679, Paper 58 at 49
(PTAB Sept. 25, 2015)), Dr. Lebby’s Declaration does not fall into this
category. We decline to parse an individual exhibit at the paragraph level to
exclude insufficiently cited-to portions from the record, without a clear
reason for undertaking such a task.
For these reasons, we deny Petitioner’s Motion to Exclude.
III. ANALYSIS OF UNPATENTABILITY
A. Claim Construction
For petitions such as this one, filed after November 13, 2018, claims
“shall be construed using the same claim construction standard that would be
used to construe the claim in a civil action under 35 U.S.C. 282(b),” which is
articulated in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005) (en
banc). 37 C.F.R. § 42.100(b) (2019). Under the Phillips standard, the
“words of a claim ‘are generally given their ordinary and customary
meaning,’” which is “the meaning that the term would have to a person of
IPR2019-01072
Patent 9,893,215 B2
11
ordinary skill in the art in question at the time of the invention, i.e., as of the
effective filing date of the patent application.” Phillips, 415 F.3d at 1312–13
(quoting Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed.
Cir. 1996). Only those terms in controversy need to be construed and only
to the extent necessary to resolve the controversy. See Nidec Motor Corp. v.
Zhongshan Broad Ocean Motor Co., 868 F.3d 1013, 1017 (Fed. Cir. 2017)
(quoting Vivid Techs., Inc. v. Am. Sci. & Eng’g, Inc., 200 F.3d 795, 803
(Fed. Cir. 1999)).
In our Institution Decision, we determined that no term of the
challenged claims required an express construction. Dec. on Inst. 8. During
trial, neither party proposed that we adopt any specific construction of a
claim term. Upon review of the record and the parties’ positions, we agree
that no term appearing in the challenged claims requires express
construction.
B. Level of Ordinary Skill in the Art
Factors pertinent to a determination of the level of ordinary skill in the
art include “(1) the educational level of the inventor; (2) type of problems
encountered in the art; (3) prior art solutions to those problems; (4) rapidity
with which innovations are made; (5) sophistication of the technology; and
(6) educational level of workers active in the field.” Envtl. Designs, Ltd. v.
Union Oil Co., 713 F.2d 693, 696–697 (Fed. Cir. 1983) (citing Orthopedic
Equip. Co. v. All Orthopedic Appliances, Inc., 707 F.2d 1376, 1381–82 (Fed.
Cir. 1983)). Not all such factors may be present in every case, and one or
more of these or other factors may predominate in a particular case. Id.
Moreover, “[t]hese factors are not exhaustive but are merely a guide to
determining the level of ordinary skill in the art.” Daiichi Sankyo Co. v.
IPR2019-01072
Patent 9,893,215 B2
12
Apotex, Inc., 501 F.3d 1254, 1256 (Fed. Cir. 2007). In determining a level
of ordinary skill, we also may look to the prior art, which may reflect an
appropriate skill level. Okajima v. Bourdeau, 261 F.3d 1350, 1355 (Fed.
Cir. 2001). Additionally, the Supreme Court informs us that “[a] person of
ordinary skill is also a person of ordinary creativity, not an automaton.”
KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 421 (2007).
Petitioner relies on the Declaration of Sanjay Banerjee, Ph. D. (Ex.
1003). Dr. Banerjee testifies that a person having ordinary skill in the art
would have “an advanced degree involving the discipline of electrical
engineering, applied physics, chemistry, or materials science, and at least
two years of experience designing, developing, or researching in the field,”
or alternatively, “a bachelor’s degree in electrical engineering, applied
physics, chemistry, or materials science, and at least three years designing,
developing, or researching in the field.” Ex. 1003 ¶ 46. We adopted this
definition for purposes of deciding whether to institute trial. Dec. on Inst. 9.
For purposes of this inter partes review, Patent Owner “does not contest this
level of skill.” PO Resp. 2. In view of this agreement, we adopt Dr.
Banerjee’s definition of the level of ordinary skill in the art.
C. Principles of Law
To prevail in challenging Patent Owner’s claims, Petitioner must
demonstrate by a preponderance of the evidence that the claims are
unpatentable. 35 U.S.C. § 316(e) (2012); 37 C.F.R. § 42.1(d) (2016). A
claim is unpatentable under 35 U.S.C. § 103 if “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
IPR2019-01072
Patent 9,893,215 B2
13
matter pertains.” KSR, 550 U.S. at 406. The question of obviousness is
resolved on the basis of underlying factual determinations, including: (1)
the scope and content of the prior art; (2) any differences between the
claimed subject matter and the prior art; (3) the level of ordinary skill in the
art; and (4) when available, evidence such as commercial success, long felt
but unsolved needs, and failure of others. Graham v. John Deere Co., 383
U.S. 1, 17–18 (1966); see also KSR, 550 U.S. at 407 (“While the sequence
of these questions might be reordered in any particular case, the [Graham]
factors continue to define the inquiry that controls.”).
The Supreme Court made clear that we apply “an expansive and
flexible approach” to the question of obviousness. KSR, 550 U.S. at 415.
Whether a patent claiming the combination of prior art elements would have
been obvious is determined by whether the improvement is more than the
predictable use of prior art elements according to their established functions.
Id. at 417. Reaching this conclusion, however, requires more than merely
showing that the prior art includes separate references covering each
separate limitation in a challenged claim. Unigene Labs., Inc. v. Apotex,
Inc., 655 F.3d 1352, 1360 (Fed. Cir. 2011). Rather, obviousness
additionally requires that a person of ordinary skill at the time of the
invention “would have selected and combined those prior art elements in the
normal course of research and development to yield the claimed invention.”
Id.
D. Obviousness Based on Hagino
Petitioner contends that claims 12–14 would have been obvious over
the disclosure of Hagino. Pet. 52–66.
IPR2019-01072
Patent 9,893,215 B2
14
1. Hagino (Ex. 1006)
Hagino, entitled “Solar Cell and Method of Manufacturing the Same,”
relates to “a solar cell for photoelectric conversion.” Ex. 1005, at [54]; ¶ 2.
Hagino’s Figure 1 is reproduced below.

Figure 1 illustrates a schematic cross-sectional view of solar cell 1. Id. ¶ 18.
Solar cell 1 includes dopant diffusion layer 6 formed on a side of a light-
receiving surface of a silicon wafer, and light-receiving surface passivation
film 3A formed on dopant diffusion layer 6. Id. ¶ 26.
Hagino’s Figures 3A–3F are reproduced below:

Figures 3A–3F of Hagino are schematic cross-section views showing a
process of manufacturing a solar cell. Id. ¶ 20. As shown in Figure 3C, a
IPR2019-01072
Patent 9,893,215 B2
15
silicon oxide film is formed as light-receiving surface passivation film 3A on
texture etching surface 2A that is on a light-receiving surface of silicon
wafer 1. Id. ¶ 28. Light-receiving surface passivation film 3A can be a
silicon film, a silicon nitride film, a titanium oxide film, or an aluminum
oxide film. Id. ¶¶ 29–30. As shown in Fig. 3E, a low-concentration dopant
diffusion layer is formed in first region 6A covered with light-receiving
surface passivation film 3A, and a high concentration dopant diffusion layer
is formed in second region 6B under the opening portion of light-receiving
surface passivation film 3A. Id. ¶ 28. As also shown in Fig. 3E, titanium
dioxide formed on the top of the wafer forms an antireflection film, where
the antireflection film includes, other than titanium dioxide, aluminum
oxide, tin oxide, silicon nitride, tantalum oxide, or the like. Id. ¶¶ 54–55.
2. Claim 12
Petitioner provides analysis setting forth how the disclosure of Hagino
allegedly meets each element of claim 12, and supports its analysis with the
testimony of Dr. Banerjee. Pet. 52–65; Ex. 1003. For example, Petitioner
asserts that Hagino discloses a solar cell, satisfying the preamble of claim
12.7 Pet. 52–53. Petitioner also sets forth how Hagino discloses a silicon
substrate, indicating Hagino’s “silicon wafer 1.” Id. at 53 (citing Ex. 1006
¶ 43, Fig.1). Patent Owner does not provide a substantive response on these
or several other limitations of the independent claim, or contend that the art
does not disclose them. Patent Owner’s response to this ground of

7 Petitioner also argues that the preamble of claim 12 is not limiting on the
scope of the claim. Pet. 30. Because each ground of unpatentability set
forth in the Petition discloses a solar cell, whether or not the preamble of
claim 12 is limiting has no effect on our evaluation of the grounds. We will
assume, without deciding, that the preamble is limiting.
IPR2019-01072
Patent 9,893,215 B2
16
unpatentability focuses on the question of whether the second dielectric
layer is embedded with hydrogen. PO Resp. 12–13.8 We have reviewed the
Petitioner’s evidence and arguments on the uncontested limitations of claim
12, and find that a preponderance of the evidence supports the conclusion
that Hagino teaches these limitations. We, therefore, turn to the disputed
limitation of the claim.
Hydrogen being embedded into the second dielectric layer
Petitioner does not contend that Hagino expressly discloses that its
second layer contains embedded hydrogen. Instead, Petitioner directs us to
Hagino’s disclosure that its silicon nitride layer is formed by “CVD using
silane, ammonia, hydrogen and the like as a base material.” Pet. 60 (citing
Ex. 1006 ¶ 29). Relying on the testimony of Dr. Banerjee, Petitioner argues
that this process would necessarily result in hydrogen being embedded in the
silicon nitride film. Id. (citing Ex. 1003 ¶¶ 206–210). Dr. Banerjee also
cites Hseih, a reference that explains that performing CVD using a reaction
“gas mixture of undiluted silane (SiH4) and pure ammonia (NH3)” resulted

8 Prior to institution, Patent Owner’s Preliminary Response also challenged
whether the Petition established that Hagino discloses a first dielectric layer
having a thickness of less than 50 nm. Paper 7, 36–45. Patent Owner did
not preserve any arguments on this issue by making them following
institution, in its Response. See Paper 13 at 7 (cautioning Patent Owner that
“any arguments for patentability not raised in the response may be deemed
waived.”); see also Tr. 47:11–18 (J. CRUMBLEY: “I didn't see anything in
the post institution briefing as to thickness in the Hagino reference, the
thickness of the layer.” MR. GLASS: “We did not argue that point in
response.”). Even if these arguments had been preserved, we have reviewed
Petitioner’s evidence supporting its contentions (Pet. 61–64) and find that
Hagino discloses a first dielectric layer with a thickness of less than 50 nm.
IPR2019-01072
Patent 9,893,215 B2
17
in a silicon nitride layer having embedded hydrogen. Ex. 1003 ¶ 207 (citing
Ex. 1021).
In our Institution Decision, we evaluated Petitioner’s evidence based
on the partial record available at the time, and in light of the arguments made
in Patent Owner’s Preliminary response, and concluded that
On this record, we are not convinced that a person of ordinary
skill in the art would have understood the antireflection layer of
Hagino to contain embedded hydrogen. Patent Owner is correct
that the First Embodiment of Hagino, relied on by Petitioner,
does not teach depositing the antireflection layer via CVD, and
CVD processing is key to Petitioner’s argument regarding
embedded hydrogen. Indeed, Hagino’s First Embodiment does
not mention CVD at all, for any of its layers. The only mention
of CVD in an embodiment of Hagino is for the passivation layer
of the Second Embodiment, on which Petitioner does not rely.
Dec. on Inst. 30.
In reaching these initial determinations, we relied on Patent Owner’s
arguments that, at least in the “First Embodiment” of Hagino on which
Petitioner primarily relies, the second-layer antireflection film is not formed
using CVD. Id. at 29. Instead, we concluded that Hagino uses a spin
coating and heating process to deposit and form the film in this embodiment.
Id. We noted that the portion of Hagino Petitioner cites for the disclosure of
CVD is in the general description of the preferred embodiments, and relates
to using CVD to deposit a silicon nitride passivation film. Id. (citing Ex.
1006 ¶ 29).
Despite our doubts, however, we instituted trial on this ground of
unpatentability, having found that other grounds in the Petition established a
reasonable likelihood that at least one claim was unpatentable. Id. at 31
(citing Guidance on the Impact of SAS on AIA Trial Proceedings (Apr. 26,
2018), https://www.uspto.gov/patents-application-process/patent-trial-and-
IPR2019-01072
Patent 9,893,215 B2
18
appeal-board/trials/guidance-impact-sas-aia-trial; see also SAS Inst., Inc. v.
Iancu, 138 S. Ct. 1348, 1359–60 (2018).)
During trial, Petitioner argued that our doubts regarding Hagino were
unfounded, and that we were mistaken in concluding that the reference
discloses using a spin-coating method for silicon nitride in the First
Embodiment. Pet. Reply 21. Rather, Petitioner argues that the spin-coating
process of the First Embodiment of Hagino is used only for forming
antireflection films of titanium dioxide or other oxides. Id. (citing Ex. 1006
¶¶ 54–55). While silicon nitride is listed as a potential material for an
antireflection film, Petitioner contends that the only method Hagino
discloses for creating a silicon nitride layer is CVD. Id. at 21–22. And,
because Hagino discloses performing this CVD in the presence of silane and
ammonia, this process would necessarily result in the silicon nitride second
layer of Hagino containing hydrogen. Id. at 22 (citing Ex. 1003 ¶¶ 206–
213).
Patent Owner responds that our Institution Decision was correct, that
there is no basis to interpret Hagino’s disclosure of a spin coating method in
its First Embodiment to exclude deposition of a silicon nitride layer. PO
Sur-Reply 21. Repeating its argument from the Preliminary Response,
Patent Owner points out that the only embodiment of Hagino that discloses
using CVD is the second, and only in the creation of a “back passivation
film.” Id. at 21–22.
Having the benefit of a fully developed record before us, and in light
of the parties’ expanded arguments regarding the disclosure of Hagino, we
find that the preponderance of the evidence supports Petitioner’s
interpretation of the reference. Petitioner is correct that the only deposition
IPR2019-01072
Patent 9,893,215 B2
19
process disclosed in Hagino for the deposition of a silicon nitride film is
CVD, in the following passage:
[I]n the solar cell in accordance with the present invention, it is
also preferable to form an amorphous silicon film or a silicon
nitride film as a passivation film. Here, an amorphous silicon
film can be formed by CVD (Chemical Vapor Deposition) using
silane, hydrogen, and the like as a base material. A silicon nitride
film can be formed by CVD using silane, ammonia, hydrogen
and the like as a base material.
Ex. 1006 ¶ 29. By contrast, the next paragraph of Hagino discloses a
contrasting method for creating a surface passivation film using titanium
oxide or aluminum oxide. Id. ¶ 30. “These films are formed by a vacuum
evaporation process.” Id. Thus, the reference generally distinguishes
between two deposition methods, depending on the material being deposited.
Turning to the First Embodiment of Hagino on which Petitioner relies,
the method recites applying a “Solution A” to the light-receiving surface of
the silicon wafer using a spin-coating method. Id. ¶ 50. The primary
purpose of this step is to diffuse an n-type dopant into the p-type
monocrystalline silicon when the wafer is heated to 800-950ºC. Id. ¶¶ 50,
51. But there can be a secondary purpose to this step: Hagino discloses that
Solution A may contain tetraisopropoxytitanium, which becomes a layer of
titanium dioxide when the wafer is heated. Id. ¶ 54. As a result, “at the
same time when the low concentration dopant diffusion layer is formed in
first region 6A and the high concentration dopant diffusion layer is formed
in second region 6B, titanium dioxide formed on the top of the wafer forms
an antireflection film 5.” Id. (emphasis added).
Hagino also provides alternative materials to titanium dioxide that can
be used as the antireflection film of its First Embodiment, including
“aluminum oxide, tin oxide, silicon nitride, tantalum oxide, or the like.” Id.
IPR2019-01072
Patent 9,893,215 B2
20
¶ 55. Notably, for the oxides listed, the metals for those oxides may be
mixed into solution A in place of, or together with, tetraisopropoxytitanium.
Id. (listing “aluminum, tin, tantalum, or the like”). No precursor for silicon
nitride is provided in this disclosure, and we find that Hagino did not
contemplate creating a silicon nitride layer using the “Solution A” spin-
coating method. Instead, we must turn to the only other disclosure for the
deposition of a silicon nitride layer contained in Hagino, which, as described
above, is a CVD process. Id. ¶ 29. Indeed, while Hagino’s First
Embodiment discloses the formation of the dopant diffusion layer and the
antireflection layer in a single step (involving Solution A), it alternatively
provides that “the antireflection film may be formed after the n-type dopant
diffusion.” Id. ¶ 56 (emphasis added). The “Solution A” spin-coating
method must not be the exclusive process for forming the antireflection film,
and the absence of silicon nitride from the list of materials applied though
Solution A is telling.
Dr. Banerjee’s testimony supports the conclusion that a person of
ordinary skill in the art would have interpreted Hagino’s disclosure in this
manner. Dr. Banerjee states that paragraph 29 of Hagino is “certainly . . .
the only described method for forming ‘silicon nitride’ layers” in the
reference. Ex. 1003 ¶ 211. We noted in our Institution Decision that this
appeared to not be the case, because silicon nitride could also be applied
through spin coating Solution A. Dec. on Inst. 30 n.10. But on the full
record now before us, we credit Dr. Banerjee’s testimony, because upon
further evaluation the Solution A spin-coating method does not pertain to the
deposition of a silicon nitride layer.
We have also reviewed Dr. Lebby’s testimony on this issue in
evaluating how a person of ordinary skill would have interpreted Hagino.
IPR2019-01072
Patent 9,893,215 B2
21
While Dr. Lebby states that the First Embodiment of Hagino does not
disclose using a CVD deposition method (Ex. 2008 ¶ 87), he does not
address how a person of ordinary skill in the art would have understood
Hagino to deposit a silicon nitride layer in the First Embodiment without
using CVD. In particular, Dr. Lebby did not indicate, in Hagino or
elsewhere, any disclosure that spin coating was a recognized method for
depositing a silicon nitride layer. The only method for depositing silicon
nitride recognized by Dr. Lebby was CVD as disclosed by the ’215 patent,
and he testified that CVD deposition of a silicon nitride layer in the presence
of silane leads to the “cracking” of silane and “typically you would . . . see
hydrogen as part of a deposited layer.” Ex. 1053, 55:6–57:20.
Based on the foregoing, we find that Hagino discloses deposition of a
surface layer of silicon nitride using CVD that corresponds to the second
layer of claim 12. We also credit Dr. Banerjee’s uncontradicted testimony
that performing CVD deposition of silicon nitride in the presence of silane,
as Hagino discloses (Ex. 1006 ¶ 29), would necessarily result in hydrogen
being embedded in the layer. Ex. 1003 ¶¶ 206–213.
3. Claims 13 and 14
Dependent claim 13 specifies that the second dielectric layer
comprises a material selected from the group consisting of silicon nitride,
silicon oxide and silicon carbide. Petitioner notes Hagino’s disclosure that
its antireflection film can be titanium dioxide, aluminum oxide, tin oxide,
silicon nitride, or tantalum oxide. Pet. 65 (quoting Ex. 1006 ¶ 55). Patent
Owner does not separately address this dependent claim. As discussed
above, we find that Hagino discloses a second layer comprised of silicon
nitride, thus meeting the additional limitation of claim 13.
IPR2019-01072
Patent 9,893,215 B2
22
Dependent claim 14 requires the first dielectric layer of claim 12 to
have a thickness of less than 30 nm. Petitioner relies on Hagino’s disclosure
that its passivation layer of silicon oxide may be from 3 to 30 nm in
thickness and Dr. Banerjee’s testimony that a person of ordinary skill in the
art would have recognized that a similar thickness may be used when
aluminum oxide is used as the passivation layer. Id. at 66 (citing Ex. 1003
¶ 242). Patent Owner does not separately address this dependent claim, and
we again note that any challenge Patent Owner may have made to this
limitation in the Preliminary Response was not preserved during the
instituted trial. We find that Hagino discloses a layer having the thickness
specified in claim 14.
For these reasons, we find that the preponderance of the evidence
supports a conclusion that the subject matter of claims 12–14 would have
been obvious to a person of ordinary skill at the time of the invention, in
light of the disclosure of Hagino.
E. Obviousness Based on Hagino, Duerinckx, and/or Hoex 2006
Petitioner also argues that claims 12–14 would have been obvious
over the combined teachings of Hagino, Duerinckx, and/or Hoex 2006. Pet.
66–71.
1. Duerinckx (Ex. 1008)
Duerinckx, entitled “Defect Passivation of Industrial Multicrystalline
Solar Cells Based on PECVD Silicon Nitride,” is an article related to
“efficiency improvements of solar cells based on . . . multicrystalline silicon
materials.” Ex. 1008, 231. Duerinckx describes that there are several
techniques for hydrogen passivation of silicon bulk and surfaces, including
“deposition of H-containing silicon nitride layers (SiNx:H).” Id. at 232.
IPR2019-01072
Patent 9,893,215 B2
23
Duerinckx further describes that hydrogen-containing silicon nitride layers
were becoming widely introduced in industrial crystalline silicon solar cell
processes due to “combining in one processing step an antireflection coating
deposition along with surface and bulk passivation.” Id. The deposition of
SiNx layers, including direct plasma enhanced chemical vapor deposition
(PECVD), effectively passivates the bulk of multicrystalline cells resulting
in an important efficiency improvement, according to Duerinckx. Id. at
232–35.
2. Hoex 2006 (Ex. 1007)
Hoex 2006, entitled “Ultralow Surface Recombination of c-Si
Substrates Passivated By Plasma-Assisted Atomic Layer Deposited Al2O3,”
relates to “[s]urface passivation of crystalline silicon (c-Si)” that improves
“performance of high efficiency industrial solar cells.” Ex. 1007, 042112-1.
Hoex 2006 describes using a high band-gap dielectric Al2O3 as a surface
passivation layer. Id. More specifically, Hoex 2006 describes preparing
ultrathin films of Al2O3 via plasma-assisted atomic layer deposition (PA-
ALD) with a thickness of 7–30 nm that provide a “good level of surface
passivation on p-type c-Si.” Id.
3. Claim 12
Petitioner’s analysis of this ground of unpatentability is similar to the
ground based solely on Hagino. But, to the extent that Hagino does not
teach hydrogen embedded in its antireflective second layer, Petitioner argues
that Duerinckx teaches this limitation. Pet. 66–69. And, if we were to
conclude that Hagino does not teach that the thickness of its passivation first
layer is less than 50 nm, Petitioner contends that Hoex 2006 teaches this
limitation. Id. at 69–71. As to the latter combination with Hoex 2006,
IPR2019-01072
Patent 9,893,215 B2
24
because Patent Owner has not contested the thickness of the first layer of
Hagino, we need not consider whether Hoex 2006 remedies any lack of
disclosure in Hagino. And Patent Owner does not dispute the disclosure of
Duerinckx, instead focusing primarily on the reason to combine the
references in the manner Petitioner proposes. We have reviewed Petitioner’s
uncontested contentions as to the disclosures of Hagino and Duerinckx, and
find that the preponderance of the evidence supports them. We, therefore,
focus our discussion on Patent Owner’s opposition, and its contention that a
person of ordinary skill in the art would not have combined the references.
PO Resp. 39–42.
Combination with Duerinckx
To the extent that the silicon nitride layer of Hagino does not
inherently contain hydrogen as a result of its deposition process, Petitioner
contends that a person of ordinary skill in the art would have applied
Duerinckx’s disclosure that embedding a silicon nitride layer with hydrogen
improves passivation to the second antireflective layer of Hagino. Pet. 67–
68. Petitioner observes that Duerinckx explicitly describes the deposition of
a silicon nitride insulating layer in a solar cell, and discloses that it is
advantageous to embed hydrogen in such a layer. Pet. 29–30, 45–46 (citing
Ex. 1008, Abstract, 232). According to Duerinckx, “[d]efects, impurities
and segregated impurities on extended defects can be passivated by
hydrogen,” and one of the known techniques for hydrogen passivation was
“deposition of H-containing silicon nitride layers.” Ex. 1008, 232. Dr.
Banerjee testifies that depositing such a layer on top of an aluminum oxide
layer, as disclosed in Hagino, would cause the hydrogen to diffuse through
the aluminum oxide layer to the surface of the silicon substrate, helping to
passivate the silicon. Ex. 1003 ¶¶ 216, 219 (citing Ex. 1023, 3823).
IPR2019-01072
Patent 9,893,215 B2
25
Petitioner contends that a person of ordinary skill in the art would have had
reason to combine Hagino with Duerinckx because both disclose layers of
silicon nitride in solar cells, and Duerinckx discloses the advantages of a
hydrogen-embedded silicon nitride layer in achieving increased passivation.
Pet. 67–68.
Patent Owner contends that a person of ordinary skill in the art would
not have combined Duerinckx with Hagino in the manner Petitioner asserts
because the two references are “fundamentally incompatible.” PO Resp.
39–42. First, Patent Owner argues that Hagino uses a spin-coating method
to apply its silicon nitride layer, while Duerinckx teaches a PECVD process
for depositing a single layer of silicon nitride. Id. at 40–41. In addition,
Hagino is a two-layer structure that uses both a passivation layer and an
antireflection layer, while Duerinckx discloses a single layer that performs
both functions. Id. Patent Owner contends that Duerinckx’s one-layer
system was “state-of-the-art at the time and would be simpler and
presumably less expensive to manufacture,” and there is no reason for a
person of ordinary skill in the art to make the combination Petitioner
proposes instead of simply replacing Hagino’s two-layer structure with the
one layer of Duerinckx. Id. at 41–42. Finally, Patent Owner argues that
even if Duerinckx provides a reason to improve a layer of Hagino by
embedding hydrogen, it would be the first passivation layer of Hagino that
would be improved by incorporating hydrogen, not the second antireflection
layer. Id. at 42. According to Patent Owner, because Hagino already
contains a passivation layer, there would be no reason to also modify the
antireflection layer to enhance passivation. Id.
Petitioner responds that Patent Owner’s argument is premised on the
mistaken interpretation that Hagino’s silicon nitride layer is formed by spin
IPR2019-01072
Patent 9,893,215 B2
26
coating. Pet. Reply 24. We agree, for the reasons discussed above in the
prior ground. Both references disclose depositing a silicon nitride layer
using CVD.
Nor are we persuaded that a person of ordinary skill in the art would
have only considered the wholesale substitution of Duerinckx’s single
silicon nitride layer for the dual-layer structure of Hagino. This is not the
combination proposed by Petitioner, which argues that the skilled artisan
would have recognized that the technique used in Duerinckx (increased
passivation using embedded hydrogen) would similarly improve the second
layer of Hagino. While an alternative modification might be the one
proposed by Patent Owner, it would more fundamentally reconfigure the
structure of Hagino’s solar cell. Patent Owner provides insufficient
evidence or argument to persuade us that the person of ordinary skill in the
art would have only considered a more drastic modification of Hagino, to he
exclusion of Petitioner’s proposed combination.
Finally, as to Patent Owner’s argument that a person of ordinary skill
in the art would have seen no passivation advantage in embedding hydrogen
in Hagino’s second layer because the device already contained a separate
passivation layer, Petitioner directs us to evidence that the hydrogen in a
second layer can diffuse and pass through a first layer to further passivate a
substrate. Pet. Reply 25 (citing Ex. 1003 ¶ 216; Ex. 1023, 3823, Ex. 1053,
23:1–21). Patent Owner does not directly respond to this evidence. We find
that, even if Hagino already contained a passivation layer, there would have
been reason for a skilled artisan to further improve the overall passivation of
the solar cell by implementing Duerinckx’s hydrogen embedding method in
a second silicon nitride layer.
IPR2019-01072
Patent 9,893,215 B2
27
For these reasons, we find that even if Hagino’s silicon nitride layer
does not inherently contain embedded hydrogen, a person of ordinary skill in
the art at the time of the invention would have combined the disclosures of
Hagino and Duerinckx, by modifying Hagino’s second silicon nitride layer
to embed hydrogen with the goal of further improving the passivation
characteristics of the solar cell. The resulting combination would teach or
suggest all elements of claim 12. We, therefore, conclude that the subject
matter of claim 12 would have been obvious over the combined disclosures
of Hagino, Duerinckx, and Hoex 2006.
4. Claims 13 and 14
Petitioner does not rely on the disclosures of Duerinckx or Hoex 2006
for the additional limitations of dependent claims 13 and 14. Rather,
Petitioner’s analysis as to these claims is the same as for the previous ground
based solely on Hagino. We find that the preponderance of the evidence
supports a conclusion of obviousness as to these claims for the same reasons
as discussed in the prior ground.
F. Obviousness Based on Bhattacharyya
Petitioner contends that claims 12–14 would have been obvious over
the disclosure of Bhattacharyya. Pet. 30–45.
1. Bhattacharyya (Ex. 1005)
Bhattacharyya, entitled “Optoelectronic Devices, Solar Cells,
Methods of Making Optoelectronic Devices, and Methods of Making Solar
Cells,” published on May 18, 2006. Ex. 1005, [43], [54]. Bhattacharyya
describes a solar cell comprising a first electrode, a first layer of
semiconductor material over the first electrode, a second layer of
semiconductor material over the first layer of semiconductor material, a
IPR2019-01072
Patent 9,893,215 B2
28
layer of semiconductor-enriched insulator over the second layer of
semiconductor material, a second electrode extending over one or more
segments of the layer of semiconductor-enriched insulator, and one or more
windows over one or more segments of the layer of semiconductor-enriched
insulator. Id. ¶ 7.
Bhattacharyya’s Figures 5 and 6 are reproduced below.

Figures 5 and 6 are diagrammatic, cross-sectional, fragmentary views of a
solar cell. Id. ¶¶ 15, 16. Bhattacharyya describes that Figure 5 illustrates
solar cell 200 supported by base 202. Id. ¶ 67. First electrode 204 is over
the base, first layer 206 of semiconductor material is over first electrode 204,
and second layer 208 of semiconductor material is over first layer 206. Id.
IPR2019-01072
Patent 9,893,215 B2
29
¶ 69. Layers 206 and 208 comprise conductive-doped silicon, conductively-
doped germanium, or conductively-doped silicon germanium. Id. Further,
Bhattacharyya describes that Figure 6 illustrates solar cell 300 that also
includes base 202, first electrode 204, first layer 206, and second layer 208.
Id. ¶ 82. Solar cell 300 further includes thin dielectric material 302 in direct
physical contact with second layer 208 of semiconductor materials and
further comprises semiconductor-enriched insulator 304 over the dielectric
material. Id. ¶ 83. Dielectric material 302 can be aluminum oxide or silicon
dioxide, and the semiconductor-enriched insulator can comprise one or both
of silicon-enriched silicon nitride and silicon-enriched silicon oxide. Id.
¶¶ 83, 85, 87.
2. Claim 12
Petitioner provides analysis setting forth how the disclosure of
Bhattacharyya allegedly meets each element of claim 12, and supports its
analysis with the testimony of Dr. Banerjee. Pet. 30–45; Ex. 1003. For
example, Petitioner asserts that Bhattacharyya discloses a solar cell,
satisfying the preamble of claim 12.9 Pet. 30. Petitioner also sets forth how
Bhattacharyya discloses a silicon substrate, indicating layers 206 and 208 as
shown in Figure 6 of the reference, which are said to comprise “single
crystal or polycrystalline silicon.” Id. at 31 (citing Ex. 1005 ¶¶ 82, 86, Fig.
6.). On these and several other limitations of the independent claim, Patent
Owner does not provide a substantive response or contend that the art does
not disclose these limitations. We have reviewed the uncontested assertions
of Petitioner on these points and find that a preponderance of the evidence
supports the conclusion that the uncontested limitations of claim 12 are

9 Again, we assume, without deciding, that the preamble is limiting.
IPR2019-01072
Patent 9,893,215 B2
30
disclosed by Bhattacharyya. We will, therefore, turn to the contested
limitations of the independent claim.
Wherein the first dielectric layer has a thickness of less than 50 nm
While Patent Owner does not dispute that Bhattacharyya discloses the
first dielectric layer, the parties disagree as to whether that first layer has a
thickness of less than 50 nm, as required by claim 12. Bhattacharyya does
not directly disclose a value for the thickness of its layer, but Petitioner
relies on the disclosure that layer 302 in Figure 6 is said to be “very thin,”
not just “thin.” Pet. 39. Petitioner notes that elsewhere in Bhattacharyya,
layers of “approximately 10 nanometers” and “from about 1 nanometer to
about 100 nanometers” are referred to as simply “thin.” Id. at 39–40 (citing
Ex. 1005 ¶¶ 47, 71, 80). If Bhattacharyya considers a “thin” layer to be 1–
100 nm, Petitioner argues a “very thin” layer must be even thinner than that,
or at the very least at the low end of the range. Id. at 40.
Petitioner also observes that layer 302 of Bhattacharyya’s Figure 6
embodiment must necessarily rely on quantum tunneling for electrons to
pass from the silicon layer below to the electrodes above. Pet. 41. Petitioner
also notes that in Figure 5 of Bhattacharyya, the first layer is explicitly said
to “function as a tunnel medium.” Ex. 1005 ¶ 71. Relying on the testimony
of Dr. Banerjee, Petitioner contends that a person of ordinary skill would
have recognized that the tunneling effect is increased the thinner the
intervening layer is made, and would have understood that layer 302 should
be as thin as possible. Pet. 42–43. Dr. Banerjee testifies that the number of
electrons that can tunnel through a layer is inversely proportional to the
thickness of that layer, and that a person of ordinary skill in the art would
have understood that the layer should be no more than 10 nm thick. Ex.
1003 ¶ 160 (citing Ex. 1019, 275–76 as “showing that electron tunneling
IPR2019-01072
Patent 9,893,215 B2
31
through Al2O3 film can occur at thicknesses up to 10 nm”). Dr. Banerjee
also testifies that it would have been obvious to keep the thickness of the
layer at much less than 50 nm, because the efficiency of the solar cell would
be inversely related to the thickness of the layer. Id. ¶ 161. He states that
this would provide another reason for the first layer of Bhattacharyya to be
maintained at a thickness of less than 10 nm. Id.
Patent Owner disagrees with Petitioner’s interpretation of
Bhattacharyya and the implications of tunneling. Patent Owner contends
that Dr. Banerjee has admitted that tunneling can occur through an
aluminum oxide layer thicker than 50 nm. PO Resp. 14 (citing Ex. 2005
¶ 248). In addition, according to Patent Owner, there is no reason to believe
that a person of ordinary skill in the art would have interpreted a “thin” layer
in a completely different context, such as the embodiments of Figures 2 and
3 of Bhattacharyya, as relevant to the understanding of what a “very thin”
dielectric layer means in the context of the Figure 6 embodiment. PO Resp.
15. Notably, the “thin” layer of Figures 2 and 3 is an electrically conductive
pure metal such as aluminum or gold that are not a dielectric layer, while
layer 302 is a dielectric layer. Id.; Ex. 1005 ¶ 47. According to Patent
Owner, “a [person of skill in the art] would understand that each layer must
be considered independently; what is ‘thin’ for one type of layer may
actually be ‘thick’ for another layer made from a different material. PO
Resp. 16 (citing Ex. 2008 ¶ 110). In addition, Patent Owner also observes
that Bhattacharyya does not always refer to layer 302 as “very thin,”
describing it as simply “thin” in a subsequent paragraph. Id. at 16 (citing
Ex. 1005 ¶ 87). As a result, a person of ordinary skill in the art “would
never have imparted precise, nanometer-scale measurements to words such
as ‘thin’ and ‘very thin.’” Id. at 15.
IPR2019-01072
Patent 9,893,215 B2
32
In response, Petitioner argues that it is irrelevant that some tunneling
can occur through layers over 50 nm, because in order to achieve the goals
of Bhattacharyya and provide an operable solar cell, the thickness of the
layer must be less than 10 nm. Pet. Reply 2. Petitioner observes that Dr.
Lebby’s testimony supports this conclusion, in that he stated that an
insulator 3.5 nm thick would be “greater than the amount that was believed
would allow tunneling to yield an operable solar cell.” Id. (citing Ex. 2008
¶ 106. “Thus, regardless of whether it is mathematically possible that one
single electron could tunnel through a thicker barrier,” Petitioner contends
that “both parties and their experts agree that the barrier in Bhattacharyya
would need to be less than 10 nm to yield an operable solar cell.” Pet. Reply
3.
This is a misinterpretation of Dr. Lebby’s testimony, Patent Owner
contends. In the passage Petitioner cites, Dr. Lebby was discussing a layer
as described in Hoex 2006, which includes an interfacial silicon oxide layer
between the substrate and the aluminum oxide layer. PO Sur-Reply 4 (citing
Ex. 2008 ¶ 106). According to Patent Owner, it is the presence of this
interfacial layer that makes thicknesses over 3.5 nm unconducive to
tunneling. Id. By contrast, Bhattacharyya excludes an interfacial layer.
Patent Owner also points to Dr. Banerjee’s testimony that tunneling
effectiveness is not dependent on layer thickness alone, but rather is
dependent on “many parameters.” Id. at 3–4 (citing Ex. 2007, 243:18–
244:12).
Upon reviewing the record as a whole, we believe Patent Owner has
the better argument on this issue, and find that Bhattacharyya does not
disclose that its first layer is necessarily thinner than 50 nm, much less the
10 nm thickness posited by Dr. Banerjee. Turning first to Petitioner’s
IPR2019-01072
Patent 9,893,215 B2
33
argument that Bhattacharyya’s characterization of a layer of 1–100 nm as
“thin” implies that a “very thin” layer must be less than 50 nm, we are not
persuaded that Bhattacharyya intended to use these terms with the degree of
exactitude Petitioner wishes to impart upon it. As Patent Owner correctly
notes, Bhattacharyya describes the same layer as both “thin” and “very thin,”
which indicates that the term is not being used with much precision. And
Petitioner provides no real explanation why the thickness of a “thin”
conductor in Bhattacharyya would have been relevant to the thickness of a
“very thin” dielectric layer.
Petitioner’s other rationale based on tunneling is more persuasive, to
some extent. Dr. Banerjee reasonably testifies why a person of ordinary
skill in the art would have recognized that for Bhattacharyya’s solar cell to
work, quantum tunneling must occur (Ex. 1003 ¶ 159), and Dr. Lebby
appears to agree with this assessment and states that the dielectric layer 302
of Bhattacharyya “must be thin enough to allow quantum tunneling.” (Ex.
2008 ¶ 105). And the experts appear to be in agreement that as a layer
becomes thinner, the ability for electrons to tunnel through that layer
increases. Thus, Dr. Banerjee’s conclusion that a person of ordinary skill in
the art would have had reason to make the first layer of Bhattacharyya
thinner is reasonable. Ex. 1003 ¶ 161.
That said, we cannot conclude based on the evidence before us that a
person of ordinary skill in the art would have concluded that a layer thinner
than 50 nm is required in Bhattacharyya, much less the 10 nm layer that Dr.
Banerjee testifies to. The only support for 10 nm that Dr. Banerjee cites for
his conclusion that “a Skilled Artisan would have recognized that the
geometry of FIG. 6 confirms that layer 306 would be less than 10 nm” is an
IPR2019-01072
Patent 9,893,215 B2
34
article by Yadavalli. Id. ¶ 160.10 But while Yadavalli describes a “tunnel
oxide barrier” of less than 10 nm (Ex. 1009, 275–76), Dr. Banerjee does not
explain how a “tunnel oxide barrier” applies to the solar cell of
Bhattacharyya, or why Yadavalli leads to a conclusion that a solar cell
cannot operate with a layer thicker than 10 nm.11
Nor have we found sufficient evidence in the record that a layer of 50
nm would result in an inoperable solar cell. Indeed, Bhattacharyya itself
indicates otherwise: its layer 210 is described as a “tunnel medium” and
may have a thickness of up to 100 nm (Ex. 1005 ¶ 71), strongly indicating
that in some cases a layer may be thicker than 50 nm while still permitting
sufficient tunneling to occur. Although layer 210 is silicon dioxide or
silicon nitride, not the aluminum oxide layer 302 of Bhattacharyya’s Figure
6 embodiment, Dr. Banerjee testified that he would expect the materials to
have similar tunneling properties. Ex. 2007 364:20–366:8.12

10 Dr. Banerjee also cites Hoex 2006 as supporting his conclusion that a
skilled artisan would have selected a 10 nm layer. (Ex. 1003 ¶¶ 164–166).
We discuss the disclosure of Hoex 2006 in the ground below, where it is
made part of the combination of references.
11 In the Reply, Petitioner cites to additional prior art references that
allegedly show layers less than 10 nm in solar cells. Pet. Reply 3. But the
fact that some solar cells include layers of less than 10 nm does not prove
Dr. Banerjee’s proposition that an operable solar cell must contain a layer of
less than 10 nm. Patent Owner also observes that the cited prior art layers
use different materials than the Al2O3 of Bhattacharyya, thus diminishing the
relevance of their disclosed layer thicknesses. PO Sur-Reply 5.
12 Q: “So in your opinion the 1 to 100 nanometer thickness of the
semiconductor-enriched insulator in Figure 5 of Bhattacharyya is helpful to
inform a person of ordinary skill in the art what the thickness of aluminum
oxide in Figure 6 of Bhattacharyya would need to be. Right?” A: “Especially
the silicon nitride portion. Because its band cap is so close . . . [to] aluminum
oxide.” Ex. 2007 365:21–366:8.
IPR2019-01072
Patent 9,893,215 B2
35
We also find that while tunneling efficiency is inversely proportional
to layer thickness, thickness is not the only variable that affects efficiency.
Dr. Banerjee’s testimony does not take into account other factors such as the
material of the layer, instead simply stating an absolute threshold of a
particular thickness. For this reason, we do not believe Dr. Lebby’s
testimony regarding a 3.5 nm layer in Hoex 2006 is the admission that
Petitioner asserts, because Petitioner’s analysis does not take into account
the fact that Hoex 2006 includes an interfacial layer of a different material
than Bhattacharyya.
Absent sufficient evidence that an Al2O3 layer of the type used in
Bhattacharyya would have been inoperable at thicknesses greater than 50
nm, we cannot conclude on the present record that a person of ordinary skill
in the art would have found it obvious to make the first layer of
Bhattacharyya at a thickness within the range of claim 12, much less the “no
more than 10 nm” range that Dr. Banerjee asserts. For these reasons, we do
not find that a preponderance of the evidence supports the conclusion that
claim 12, or its dependent claims 13 and 14, would have been obvious over
Bhattacharyya alone.
G. Alleged Obviousness Based on Bhattacharyya, Duerinckx, and
Hoex 2006
Petitioner contends that claims 12–14 would have been obvious over
the combined teachings of Bhattacharyya, Duerinckx, and Hoex 2006. Pet.
45–52.
1. Claim 12
Petitioner’s analysis of this ground of unpatentability is similar to the
ground based solely on Bhattacharyya. But, to the extent that Bhattacharyya
does not teach hydrogen embedded in the second layer of its Figure 6 solar
IPR2019-01072
Patent 9,893,215 B2
36
cell, Petitioner argues that Duerinckx teaches this limitation. Pet. 45–48.
And, as we have concluded that Bhattacharyya does not teach that the
thickness of the first layer of Figure 6 should be less than 50 nm, Petitioner
contends that Hoex 2006 teaches this limitation. Id. at 48–52. Because the
question of the combination with Hoex 2006 resolves this ground of
unpatentability, we focus on it here.
Combination with Hoex 2006
According to Petitioner, to the extent that Bhattacharyya does not
disclose the thickness of its first layer, the thickness would have been
obvious in light of the disclosure of Hoex 2006. Pet. 48–51. As set forth
above, Hoex 2006 describes excellent surface passivation using an
aluminum oxide layer as thin as 7 nm. Ex. 1007, 042112-1. Petitioner
contends that, given that Bhattacharyya also uses aluminum oxide in a solar
cell to “reduce surface recombination and to enhance collection and thereby
also enhance solar cell efficiency” (Ex. 1005 ¶ 87), a person of ordinary skill
in the art would have applied Hoex 2006’s disclosure regarding the thickness
of its Al2O3 film to Bhattacharyya. Pet. 49.
Patent Owner argues that “Petitioners arbitrarily force Hoex 2006’s
thin layer into the system of Bhattacharyya,” where in fact a person of
ordinary skill in the art would not have combined the references due to the
differences between the structures disclosed. PO Resp. 23–24. In particular,
Patent Owner notes that Hoex 2006 discloses that its structure has an
interfacial silicon oxide layer between its silicon substrate and the aluminum
oxide layer, whereas Bhattacharyya places aluminum oxide directly onto the
substrate. Id. at 24–25. Dr. Lebby testifies that the absence of an interfacial
layer in Bhattacharyya is significant because it contributes to the ability of
electrons to tunnel through the aluminum oxide layer, which allows the solar
IPR2019-01072
Patent 9,893,215 B2
37
cell to function. Ex. 2008 ¶¶ 113–115. Hoex 2006’s deposition process, by
contrast, necessarily results in the presence of an interfacial layer. Id. ¶ 106.
We rejected this argument in our Institution Decision, and find it no
more persuasive on the full record. As we said before, we do not understand
Petitioner’s proposed ground of unpatentability to rely on Hoex 2006’s
deposition method; rather, Petitioner cites Hoex 2006 as providing guidance
as to the appropriate thickness of the aluminum oxide layer of
Bhattacharyya. Patent Owner’s argument based on Hoex 2006’s deposition
method is beside the point.
Patent Owner responds, however, that Hoex 2006’s thickness cannot
simply be separated from the presence of its interfacial oxide layer. PO
Resp. 26–28. Dr. Lebby testifies that for a layer to be as thin as disclosed in
Hoex 2006, the presence of an interfacial layer is required in order to
contribute to the aluminum oxide’s high negative charge density and
promote field effect passivation. Ex. 2008 ¶ 106. In other words, if the
layer of Hoex 2006 were formed without an interfacial layer, as the
combination proposed by Petitioner contemplates, then Hoex 2006’s layer
would necessarily have to be made thicker to maintain its passivation
qualities. Id. Patent Owner contends that Petitioner does not attempt to
address these modifications, or whether they would take the thickness of
Bhattacharyya’s layer outside the range required by claim 12. PO Resp. 28.
Petitioner replies by noting that Hoex 2006 explicitly states that its
interfacial oxide layer “does not contribute to passivation.” Pet. Reply 16.
But the portion of Hoex 2006 that Petitioner cites is referring to the
passivation qualities of the oxide layer in itself, not the effect that the oxide
layer has on the passivation abilities of the aluminum oxide layer above it.
Ex. 1007, 2 (“the oxide layer formed during the O2 plasma exposure in itself
IPR2019-01072
Patent 9,893,215 B2
38
does not yield any surface passivation, also not after annealing.”). Nor do
we find persuasive Petitioner’s argument that because Hoex 2006’s
combined layers are 8.5 nm thick—less than the 10 nm cutoff proposed by
Dr. Banerjee—both layers can be incorporated into Bhattacharyya without
preventing tunneling. Pet. Reply 15. Again, this argument focuses only on
the thickness of the layer, to the exclusion of the materials of the layer, an
oversight in Dr. Banerjee’s testimony we addressed above.13
Finally, to counter Patent Owner’s argument that the layer of Hoex
2006’s Al2O3 layer would have to be made thicker if the interfacial layer
were removed, Petitioner cites to our Institution Decision, which pointed out
that Bhattacharyya uses two layers for passivation while Hoex 2006
discloses only one, so it was reasonable to conclude that Bhattacharyya
could achieve passivation with thinner layers. Pet. Reply 17 (citing Dec. on
Inst. 25). At the time, however, we had not been presented with Dr. Lebby’s
testimony, or Patent Owner’s argument and evidence regarding the effect of
Hoex 2006’s interfacial layer on tunneling and passivation characteristics.
Based on the fully developed record, we find that a person of ordinary skill
in the art would not have assumed that the Al2O3 layer of Hoex 2006 could
simply be transplanted into Bhattacharyya without making any adjustment to
its thickness. As a result, the fact that Hoex 2006 found its layer of 7 nm to
have “excellent surface passivation” qualities does not lead us to conclude

13 Nor do we find it persuasive that Petitioner cites a reference that shows a
barrier comprised of two different materials may actually increase tunneling
effectiveness. Pet. Reply 15 (citing Ex. 1044, 1353). Petitioner does not
explain why the fact a dual silicon oxide/silicon nitride layer stack may
improve tunneling has any bearing on the question of whether the silicon
oxide/aluminum oxide layers of Hoex 2006 permit tunneling at a thickness
of 8.5 nm.
IPR2019-01072
Patent 9,893,215 B2
39
that the Al2O3 layer of Bhattacharyya could simply be modified to a similar
thickness with a reasonable expectation of successfully achieving the same
results.
For these reasons, we cannot conclude that the preponderance of the
evidence establishes that a person of ordinary skill in the art would have
combined Bhattacharyya and Hoex 2006 to arrive at the claimed invention,
namely a solar cell having a first dielectric layer of less than 50 nm. As
Hoex 2006 does not remedy the deficiency found above in the ground based
on Bhattacharyya alone, and Petitioner does not contend that Duerinckx
remedies this deficiency, we need not address the combination of
Bhattacharyya with Duerinckx.
IV. CONCLUSION
Based on the evidence and arguments, Petitioner has demonstrated by
a preponderance of the evidence that:
(1) claims 12–14 are unpatentable under 35 U.S.C. § 103(a), as
having been obvious over the disclosure of Hagino;
(2) claims 12–14 are unpatentable under 35 U.S.C. § 103(a), as
having been obvious over the combined disclosures of Hagino,
Duerinckx, and Hoex 2006;
(3) claims 12–14 have not been proven unpatentable under 35 U.S.C.
§ 103(a) as having been obvious over the disclosure of
Bhattacharyya; and
(4) claims 12–14 have not been proven unpatentable under 35 U.S.C.
§ 103(a) as having been obvious over the combined disclosures of
Bhattacharyya, Duerinckx, and Hoex 2006.
IPR2019-01072
Patent 9,893,215 B2
40
V. ORDER
In consideration of the foregoing, it is hereby:
ORDERED that Petitioner’s Motion to Exclude is denied;
FURTHER ORDERED that claims 12–14 of U.S. Patent No.
9,893,215 B2 are unpatentable; and
FURTHER ORDERED that, because this Decision is final, a party to
the proceeding seeking judicial review of the Decision must comply with the
notice and service requirements of 37 C.F.R. § 90.2.

In summary:
Claims 35
U.S.C.
Reference(s) Claims
Shown
Unpatentable
Claims
Not shown
Unpatentable
12–14 § 103 Bhattacharyya 12–14
12–14 § 103 Bhattacharyya,
Duerinckx, and Hoex
2006
12–14
12–14 § 103 Hagino 12–14
12–14 § 103 Hagino, Duerinckx,
and Hoex 2006
12–14
Overall
Outcome
12–14

IPR2019-01072
Patent 9,893,215 B2
41
FOR PETITIONER:
Ching-Lee Fukuda
Jason Greenhut
Wonjoo Suh
SIDLEY AUSTIN LLP
clfukuda@sidley.com
jgreenhut@sidley.com
wsuh@sidley.com

FOR PATENT OWNER:
James Glass
Sean Gloth
QUINN EMANUEL URQUHART & SULLIVAN, LLP
jimglass@quinnemanuel.com
seangloth@quinnemanuel.com