Trustees of Boston University v. Everlight Electronics Co., Ltd. et al

Filing 1252

Chief Judge Patti B. Saris: MEMORANDUM and ORDER entered. After a review of the record and the case law, I deny the motion for summary judgment on the ground that the patent is invalid for indefiniteness. Defendants' Motion for Summary Judgment of Invalidity Docket No. 878 is DENIED. (Geraldino-Karasek, Clarilde)

UNITED STATES DISTRICT COURT
DISTRICT OF MASSACHUSETTS
)
)
)
) Consolidated Civil Action No.
v.
) 12-11935-PBS
)
EVERLIGHT ELECTRONICS CO., LTD.,)
et al.,
)
Defendants.
)
)
)
TRUSTEES OF BOSTON UNIVERSITY, )
Plaintiff,
)
) Civil Action No. 12-12326-PBS
v.
)
)
EPISTAR CORPORATION, et al.,
)
Defendants.
)
)
)
TRUSTEES OF BOSTON UNIVERSITY, )
Plaintiff,
)
) Civil Action No. 12-12330-PBS
v.
)
)
LITE-ON INC., et al.,
)
Defendants.
)
________________________________)
TRUSTEES OF BOSTON UNIVERSITY,
Plaintiff,

MEMORANDUM AND ORDER
June 9, 2015
Saris, Chief Judge.
I.

INTRODUCTION

Defendants move for summary judgment of invalidity of U.S.
Patent No. 5,686,738 (the ‘738 patent), which relates to a twostep method of preparing gallium-nitride (GaN) films, a common
1

component of blue-colored light-emitting diodes (LEDs). The
invalidity dispute involves the “non-single crystalline buffer
layer” limitation in all asserted patent claims. The Court has
construed the claim

limitation “non-single crystalline” to mean

“polycrystalline, amorphous or a mixture of polycrystalline and
amorphous.” Trustees of Boston University v. Everlight
Electronics Co., Ltd., 23 F. Supp. 3d 50, 62 (D. Mass. 2014).
Defendants argue that all asserted claims1 of the ‘738 patent are
invalid under 35 U.S.C. § 112, ¶¶ 1 and 2 on the grounds that (1)
the patent’s specification has no written description of a “nonsingle crystalline” limitation; (2) the disclosure does not

1

Claim 1 of the patent provides:
1. A semiconductor device comprising:
a substrate, said substrate consisting of a
material selected from the group consisting
of (100) Silicon, (111) silicon, (0001)
sapphire, (11-20) sapphire, (1-102) sapphire,
(111) gallium aresenide, (100) gallium
aresenide, magnesium oxide, zinc oxide and
silicon carbide;
a non-single crystalline buffer layer having
a thickness of about 30A to about 500A,
comprising a first material grown on said
substrate, the first material consisting
essentially of gallium nitride; and
a first growth layer grown on the buffer
layer, the first growth layer comprising
gallium nitride and a first dopant material.

(Emphasis added).

2

enable a person skilled in the art to make and use the invention;
and (3) the term “non-single crystalline” is indefinite. The
Court assumes familiarity with the underlying technology. See id.
at 53-57. After hearing (Docket No. 1065), Defendants’ motion is
DENIED.
II. STANDARD OF REVIEW
Issued patents are presumed valid under the Patent Act, 35
U.S.C. § 282. As a result, the party challenging the validity of
patent claims bears the burden of showing by clear and convincing
evidence that the patent is invalid. AK Steel Corp. v. Sollac &
Ugine, 344 F.3d 1234, 1238-39 (Fed. Cir. 2003).
In deciding a case on summary judgment, the Court views the
facts in the light most favorable to the non-moving party and
makes all reasonable inferences in that party’s favor. O’Connor
v. Steeves, 994 F.2d 905, 907 (1st Cir. 1993). Summary judgment
is appropriate when no genuine issue exists as to any material
fact, and the moving party is entitled to judgment as a matter of
law. Eli Lilly & Co. v. Barr Labs, Inc., 251 F.3d 955, 962 (Fed.
Cir. 2001). “When evaluating a motion for summary judgment, the
court views the record evidence through the prism of the
evidentiary standard of proof that would pertain at a trial on
the merits.” Id. “Thus, a moving party seeking to invalidate a
patent at summary judgment must submit such clear and convincing

3

evidence of invalidity so that no reasonable jury could find
otherwise.” Id.
III. DISCUSSION
A. Written Description of “Non-Single Crystalline” Limitation
Defendants argue that the ‘738 patent’s specification does
not contain an adequate written description of a GaN buffer layer
that is “amorphous, polycrystalline or a mixture of amorphous and
polycrystalline.” Defendants contend that the specification does
not expressly teach a non-single crystalline buffer layer for a
completed device, pointing out that the specification never
mentions the terms “non-single” or “polycrystalline.” While they
acknowledge that the specification describes an “amorphous”
buffer layer, Defendants contend that it mentions “amorphous” to
describe an intermediate state that is crystallized during the
second step: as such, there is no amorphous buffer in the
completed device. Defendants submitted extrinsic evidence to
support their core contention that the patent’s inventor, Dr.
Theodore Moustakas, actually invented a “single crystalline”
buffer layer, but added the modifier “non-single crystalline” on
the suggestion of the Patent and Trademark Office examiner.
Docket No. 881-3, Yoches Decl., Ex. 8C, Excerpts from File
History of the ‘738 Patent, at 11-12. Defendants argue this was
an opportunistic change unsupported by the specification.

4

A patent’s written description “must convey with reasonable
clarity . . . that, as of the filing date sought, [the patentee]
was in possession of the invention, and demonstrate that by
disclosure in the specification of the patent.” Carnegie Mellon
Univ. v. Hoffman-La Roche Inc., 541 F.3d 1115, 1122 (Fed. Cir.
2008) (internal quotation omitted). Assessing such “possession as
shown in the disclosure” requires “an objective inquiry into the
four corners of the specification from the perspective of a
person of ordinary skill in the art.” Ariad Pharm., Inc. v. Eli
Lilly & Co., 598 F.3d 1336, 1351 (Fed. Cir. 2010) (en banc). A
“mere wish or plan” for obtaining the claimed invention is not an
adequate written description. Regents of the Univ. of Cal. v. Eli
Lilly & Co., 119 F.3d 1559, 1566 (Fed. Cir. 1997). The
sufficiency of a patent’s written description is ordinarily a
question of fact, but “[a] patent also can be held invalid [as a
matter of law] for failure to meet the written-description
requirement based solely on the face of the patent
specification.” Centocor Ortho Biotech, Inc. v. Abbott Labs., 636
F.3d 1341, 1347 (Fed. Cir. 2011).
The specification of the ‘738 patent states that the buffer
layer is amorphous at “the low temperatures of the nucleation
step” and “can” be “crystallized” by heating. ‘738 patent at
2:40-42. In the description of the preferred embodiment, the
patent states, “The buffer is the only part of the film which is

5

highly defective.”2 ‘738 patent at 4:49-50. As defense counsel
conceded, the term “crystallized” includes a polycrystalline
buffer layer, so the specification does disclose a
polycrystalline buffer. Dkt. No. 1029, Ex. 22, Piner Depo. 44:56, 45:9-12 (pointing out that there can also be a polycrystalline
layer with amorphous regions). Plaintiffs emphasize the word
“can” to argue that the specification does not require that the
amorphous layer ever crystallize. The plain meaning of “can” is
“to be able to; to have the ability, power or skill to” or “to
have the possibility.” Random House Unabridged Dictionary 302 (2d
ed. 1993). Plaintiff’s expert, Dr. Piner, a professor of Physics
and Material Sciences, Engineering and Commercialization at Texas
State University, testified that “the ‘738 patent discloses that
the buffer layer may be purely amorphous.” Dkt. No. 945-2, Piner
Decl. ¶ 35. Dr. Piner added that, in his own research, he has
“directly observed a monocrystalline GaN growth layer ‘on’3 a
100% amorphous buffer layer.” Id. at ¶ 37. Consistent with this
plain meaning, Piner testified that “can crystallize” does not
mean “must be crystallized.” Id. at ¶ 35.
Despite this, Defendants argue that the patent specification

2

Because a single crystal can be “highly defective,” the
term “highly defective” in the specification does not resolve the
question as to whether the buffer layer is monocrystalline.
3

The Court defined the term “grown on” to mean “formed
directly or indirectly above.” Trustees of Boston University, 23
F. Supp. 3d at 59.
6

does not adequately describe a completely amorphous buffer layer
in the completed device. In their view, the amorphous film
crystallizes during the two-step process, and any further growth
takes place on the crystallized GaN buffer layer. Therefore,
Defendants insist that “can” crystallize should be interpreted to
mean “must” crystallize. They point out that there is no evidence
that one could make the claimed semiconductor device with a
purely amorphous buffer layer. Defendant’s expert, Dr. Eugene
Fitzgerald, a professor of Material Sciences at MIT, testified as
follows:
Q.

And during the high temperature growth step
described in the ‘738 patent, the increase in the
temperature causes the amorphous gallium nitride
buffer layer to crystallize into a polycrystalline
or mixed polycrystalline and amorphous layer of
gallium nitride. Is that correct?

A.

Yes. It can do either. It could - if you’re at
high enough temperature, it could recrystallize
into a single crystal. If you have the first layer
too thick, it could - it could crystallize into
amorphous and polycrystalline. There’s a variety
of outcomes. Yes.

Dkt. No. 945-3, Fitzgerald Depo. at 39:9-20. Defendants also cite
the testimony of inventor Dr. Moustakas, who stated that one
“would not be able to grow a single crystal on an amorphous
buffer layer.” Dkt. No. 945-4, Moustakas Depo. 200:18-204:4; but
see id. 88:21-89-8 (stating he had grown a growth layer on top of
a purely amorphous layer). The parties’ experts, both eminently
qualified, thus disagree whether a monocrystalline growth layer

7

may grow on an amorphous buffer layer. This debate is better
resolved in the context of enablement. Summary judgment on
whether the patent adequately describes an amorphous buffer zone
in the claimed device is denied.
B. Enablement
Defendants assert that the ‘738 patent fails to teach how a
person skilled in the art would grow a single-crystalline growth
layer directly on top of an amorphous buffer layer.
To meet the enablement requirement, a patent specification
must enable “one skilled in the art . . . [to] practice the
claimed invention without undue experimentation.” AK Steel Corp.,
344 F.3d at 1244 (Fed. Cir. 2003) (citation omitted). The Federal
Circuit has explained that the written description requirement
and the enablement requirement are distinct, but they “often rise
and fall together.” Ariad, 598 F.3d at 1352.
Defendants argue that the plaintiffs have not enabled the
full scope of their claimed invention. After the Markman hearing,
I construed the term “non-single crystalline buffer layer” to
mean “a layer of material that is not monocrystalline, namely,
polycrystalline, amorphous, or a mixture of polycrystalline and
amorphous.” Trustees of Boston University, 23 F. Supp. 3d at 6263. I drew this construction from Dr. Moustakas’ express
definition of “non-single crystalline” during the prosecution of
the parent patent, ‘819, noting that his limitation of this term

8

for the parent application “applies with equal force to the ‘738
patent.” Id. Based on this tripartite definition, Defendants
argue that the patent must enable a semiconductor device
containing a buffer layer that is polycrystalline, one containing
a buffer layer that is amorphous, and one containing a buffer
layer that is a mixture of polycrystalline and amorphous. Since,
in Defendants’ view, Plaintiffs have failed to show that the
buffer layer can ever be purely amorphous, they contend the
patent fails on enablement grounds.
Defendants base this argument in a number of cases
indicating that, for a patent to be valid, the “scope of the
claims must be less than or equal to the scope of the
enablement.” Sitrick v. Dreamworks, LLC, 516 F.3d 993, 995 (Fed.
Cir. 2008) (where court construed claims to encompass an audio
signal able to connect to both video games and movies, patent
held invalid because it enabled only how to connect signal to
video games); see also Automotive Techs. Int’l, Inc. v. BMW of
North America, Inc., 501 F.3d 1274, 1277 (Fed. Cir. 2007) (where
court construed phrase “corresponding structure” to include both
mechanical switches and electronic switches, holding patent
invalid because it enabled only mechanical switch); LiebelFlarsheim Co. v. Medrad, Inc., 481 F.3d 1371, 1378-79 (Fed. Cir.
2007) (where claim encompassed injectors with pressure jackets
and injectors without jackets, holding patent invalid because it

9

only enabled injectors with jackets); Tronzo v. Biomet, Inc., 156
F.3d 1154, 1158-59 (Fed. Cir. 1998) (holding that “the disclosure
must describe the claimed invention with all its limitations”).
Plaintiffs argue that a person of ordinary skill in the art could
construct, without undue experimentation, a semiconductor device
containing a fully amorphous buffer layer.4
Factual disputes remain regarding whether the full scope of
the ‘738 patent is sufficiently enabled. It is clear, as BU’s
expert Dr. Piner stated, that at least some version of the
claimed device can be created without any undue experimentation,
since the patent teaches process parameters that are “clear,
sufficiently concise, and sufficiently complete.” Piner Decl. ¶
38. Even Defendants’ expert, Dr. Fitzgerald, stated that he had
built a device using the process described by the ‘738 patent.
Dkt. No. 945-3, Fitzgerald Depo. at 40:21-41:2. His tutorial
explains how to make the device. Dkt. No. 295-1.
The harder question is whether the patent enables a device
with a purely amorphous buffer layer. Defendants present evidence
that Dr. Moustakas’s contemporaneous research into galliumnitride films dealt with single-crystalline buffer layers as

4

At the hearing, Plaintiffs did remark in passing that
“the case law is pretty clear that you only need to enable one
mode of making the claimed invention.” Docket No. 1072 at 58.
However, Plaintiffs did not press this argument, either during
oral argument or in the briefs, and cited no support in the case
law.
10

opposed to non-single,5 and argue that it is “impossible” to grow
a single-crystalline growth layer directly on an amorphous buffer
layer. Dkt. 879 at 4. But plaintiffs raise several points in
response. For one thing, as discussed above, Dr. Piner testified
that:
I have directly observed a monocrystalline GaN growth layer on
a 100 % amorphous buffer layer . . . Thus, my own published
experience confirms that the growth layer described by the
‘738 patent can be grown on a purely amorphous buffer layer.
Dkt. No. 945-2, Piner Decl. ¶ 37. For another, Plaintiffs
highlight that additional layers may exist between an amorphous
buffer layer and a monocrystalline growth layer under the Court’s
construction of the phrase “grown on,” which may mean “formed
indirectly above.” Trustees of Boston University, 23 F. Supp. 3d
at 59 (emphasis added). Finally, Plaintiffs challenge Defendants’
reliance on the views of Dr. Gopinath Menon, who, they argue, is
currently a plastic surgeon and thus not a person of ordinary
skill in the art. Dkt. No. 945 at 16. Based on this evidence,
factual disputes preclude summary judgment on enablement.

5

Defendants cite these articles: Epitaxial growth of zinc
blende and wurtzitic gallium nitride thin films on (001) silicon,
59(8) Appl. Phys. Lett. 944 (1991); Epitaxial growth and
characterization of zinc-blende gallium nitride on (001) silicon,
71 J. Appl. Phys. 10 (1992); A Comparative Study of GaN Films
Grown on Different Faces of Sapphire by ECR-Assisted MBE, 242
Mat. Res. Soc’y Symp. Proc. (Symp. G - Wide Band-Gap
Semiconductors) 427, 427-28 (1992); Growth by GaN by ECR-assisted
MBE, 185 Physica B: Condensed Matter 36, 45 (1993).
11

C. Claiming the Subject Matter and Definiteness
Defendants’ last argument is that the asserted claims in the
‘738 patent are invalid under 35 U.S.C. § 112, ¶ 2 because (1)
there is a contradiction between the claims and the
specification; (2) the claims recite a limitation not disclosed
in the specification; and (3) the term “non-single crystalline”
is indefinite. The first two arguments appear to involve the same
issues addressed in the discussion of enablement and written
description. The third argument raises new issues.
Title 35 U.S.C. § 112, ¶ 2 states that a specification
“shall conclude with one or more claims particularly pointing out
and distinctly claiming the subject matter which the inventor or
a joint inventor regards as the invention.” The Federal Circuit
has stated that ¶ 2 contains two requirements. First, the claims
must set forth what the applicant regards as his invention. Allen
Eng’g Corp. v. Bartell Indus., Inc., 299 F.3d 1336, 1348 (Fed.
Cir. 2002). Second, the claim must set forth the invention with
“sufficient particularity and distinctness, i.e., the claim must
be sufficiently definite.” Id. (quotation marks omitted); see
also Nautilus, Inc. v. Biosig Instruments, Inc., ___ U.S. ___,
134 S. Ct. 2120, 2124 (2014).
For a patent specification to be invalid for indefiniteness,
it must be "insolubly ambiguous," Halliburton Energy Servs. Inc.
v. M-I LLC, 514 F.3d 1244, 1249 (Fed. Cir. 2008), such that

12

"reasonable efforts at claim construction prove futile." Exxon
Research and Engineering Co. v. United States, 265 F.3d 1371,
1375 (Fed. Cir. 2001). Imprecise language is not enough to render
a claim indefinite, because only a "reasonable degree of
particularity" is required. Id. at 1381. "Claims amenable to more
than one construction should, when it is reasonably possible to
do so, be construed to preserve their validity." Karsten Manuf.
Corp. v. Cleveland Gold Co., 242 F.3d 1376, 1384 (Fed. Cir.
2001). A Court's claim construction "need not always purge every
shred of ambiguity," Acumed LLC v. Stryker Corp., 483 F.3d 800,
806 (Fed. Cir. 2007), and "some line-drawing problems may be left
to the trier of fact." CardioFocus, Inc. v. Cardiogenesis Crop.,
827 F. Supp. 2d 36, 43 (D. Mass. 2008). As the Federal Circuit
has observed:
Even if it is a formidable task to understand a claim, and the
result not unanimously accepted, as long as the boundaries of
a claim may be understood, it is sufficiently clear to avoid
invalidity for indefiniteness.
Invitrogen Corp. v. Biocrest Mfg., L.P., 424 F.3d 1374, 1383
(Fed. Cir. 2005).
Defendants argue that the term “non-single crystalline” is
indefinite because there is no accepted method in the field for
determining whether a substance is single-crystalline versus nonsingle crystalline. Dr. Fitzgerald has testified that the easiest
way to determine crystallinity is by a Transmission Electron
Microscopy (TEM) diffraction. Dkt. No. 945-3, Fitzgerald Depo.
13

59:16-60:3. In his declaration, he stated: “There is no clear and
accepted boundary between a single crystalline and a
polycrystalline material.” Dkt. No. 882 at 2. However, he also
acknowledged that certain materials are considered single
crystalline or polycrystalline but “when the crystallinity of a
material is in the middle range between the two extreme,
determining the crystallinity category of such a material is
highly subjective and depends on the characterization technique.”
Id.; see also Molnar Depo. at 154:18-20 (“Again, I think single
crystal versus polycrystal is a somewhat subjective term.”).
While the experts disagree on how to read the TEM diffraction
patterns in this case, they do not disagree that TEM is a
reliable method for determining crystallinity. Piner Decl. ¶ 47.
Similarly, both experts agree that certain kinds of highangle “grain boundaries” indicate polycrystallinity, although
they dispute whether a person of ordinary skill in the art would
determine that the low-angle grain boundaries in the Exemplar
buffer layers render those layers “non-single crystalline” or
simply indicate defects in a monocrystalline layer. See Trustees
of Boston University v. Everlight Electronics Co, Ltd., 2015 WL
2400760 at *3 (D. Mass. May 20, 2015) (denying Defendants’ motion
for summary judgment as to infringement of the ‘738 patent).
Finally, both experts agree that a person of ordinary skill
would define a single-crystalline GaN layer by its defect
14

density. See Dkt. No. 945-3, Fitzgerald Depo. at 54:10-55:2
(stating that in the gallium nitride field, “a high quality
single crystal . . . might have a defect density of 10 to the 8th
dislocations per centimeter squared”); Dkt. No. 945-2, Piner
Decl. at ¶ 49 (discussing the defect density in a non-single
crystalline GaN buffer law as > 1010 or even 1012). While the
precise lines the experts draw with respect to dislocations
differ, there are accepted lines for determining with reasonable
certainty the boundaries for defect density between single
crystalline and non-single crystalline layers. In sum, though, on
the margin, there may be some ambiguity as to whether a layer is
monocrystalline or polycrystalline, some imprecision does not
invalidate a claim.
After a review of the record and the case law, I deny the
motion for summary judgment on the ground that the patent is
invalid for indefiniteness.
ORDER
Defendants’ Motion for Summary Judgment of Invalidity
(Docket No. 878) is DENIED.

/s/ PATTI B. SARIS
Patti B. Saris
Chief United States District Judge

15