Rockwell International Corporation (“Rockwell”) appeals the decision on summary judgment of the United States Court of Federal Claims in
Rockwell International Corp. v. United States,
This appeal was submitted for our decision following oral argument on March 3, 1998. We have jurisdiction over a final decision of the Court of Federal Claims pursuant to 28 U.S.C. § 1295(a)(3) (1994). Because the Court of Federal Claims erred in holding on summary judgment that the inventions of the ’098 patent would have been obvious, we vacate that portion of the decision and remand the case for further proceedings. On the cross-appeal, we affirm the decision of the Court of Federal Claims that genuine issues of material fact prevented summary judgment that the claims of the ’098 patent were anticipated.
BACKGROUND
Rockwell filed suit against the United States on August 30, 1993, alleging infringement of various claims of its ’098 patent and identifying sixteen government contractors alleged to have participated in the infringement. 2 SDL, one of those contractors, received a notice from the United States under RCFC 3 14(a)(1) & (c) but initially did not appear because SDL believed that its liability as an indemnitor was not high enough. In May 1995, however, Rockwell sued SDL for infringement of the same patent in the United States District Court for the Northern District of California. In June 1995, SDL intervened in the Court of Federal Claims action as a third-party defendant, and in September 1995, the district court action was suspended pending disposition of the validity *1361 issues in the action pending in the Court of Federal Claims.
Prior to SDL’s intervention, Rockwell and the United States had obtained a scheduling order that trial would proceed in phases based on categories of accused infringing equipment as follows: (1) night vision equipment; (2) photo voltaic cells; and (3) laser diodes in semiconductor devices. Claim construction and validity were to be decided in phase 1 and applied to all three phases. This appeal and cross-appeal relate to whether the ’098 patent is invalid.
The ’098 patent claims a process called MOCVD 4 for growing Group III/V semiconductor materials. Chemical vapor deposition (“CVD”) is a process for depositing a thin film of material onto a substrate by reacting the constituent elements in gaseous phase. CVD processes are used to produce thin, single crystal films called epitaxial films. Prior to the invention of the ’098 patent, other CVD processes had been used to grow similar Group III/V semiconductor films.
The ’098 patent identifies specific groups of reactants to be used in CVD processes. The reactant supplying the Group III element is to be an organometallic alkyl, and the Group V reactant is to be a hydride or a halide-free alkyl compound. Dr. Harold M. Manasevit first filed his patent application on the MOCVD process on February 13, 1968. After numerous continuations and continuations-in-part, the application issued as the ’098 patent on January 11,1983.
In phase 1 of the litigation, Rockwell asserted infringement of at least independent claims 1, 2, and 3. Defendants asserted that claims 1, 2, and 3 are invalid under sections 102, 103, and 112 of Title 35. SDL further asserted invalidity for failure to claim statutory subject matter under section 101. Rockwell filed a motion for summary judgment that claims 1 and 3 were not invalid, which was denied. Defendants filed motions for summary judgment of anticipation, which were denied, and of obviousness, which were granted.
The trial court held that because neither Rockwell’s own motion for summary judgment that claims 1 and 3 are not invalid nor its consolidated opposition to Defendants’ motions urging the invalidity of claims 1, 2, 3, 11, 35, 40, 44, 50, 55-58, 66, and 72 separately defended independent claim 2 or any of the other claims of the ’098 patent, claim 2 and all of the other claims stand or fall with the decision on claims 1 and 3.
See id.
at 492;
see also Rockwell Int’l Corp. v. United States,
The trial court found that during prosecution, the claims of the ’098 patent were limited to a CVD process using a cold-wall reactor.
Rockwell,
DISCUSSION
Summary judgment is appropriate where “there is no genuine issue as to any material fact and ... the moving party is entitled to a judgment as a matter of law.” RCFC 56(c). In determining the propriety of summary judgment, credibility determinations may not be made, and the evidence must be viewed favorably to the nonmovant, with doubts resolved and reasonable inferences drawn in the nonmovant’s favor.
See SRI Int’l v. Matsushita Elec. Corp.,
775 F.2d
*1362
1107, 1116,
“[I]n rendering a decision on a motion for summary judgment, a court must View the evidence presented through the prism of the substantive evidentiary burden’ that would inhere at trial.”
Monarch Knitting Mach. Corp. v. Sulzer Morat Gmbh,
Obviousness is a question of law based on underlying factual determinations.
See Richardson-Vicks, Inc. v. Upjohn Co.,
I. Claim Construction
The first step in any invalidity or infringement analysis is claim construction.
See Beachcombers v. WildeWood Creative Prods., Inc.,
The trial court first determined that the preambles of independent claims 1 through 3 operated as claim limitations.
6
See Rockwell,
If the preamble language is not used, there would be no claim limitations regarding the growing of epitaxial films or the type of substrate used in the process. The patent when read in its entirety reads toward the process of the growth of a particular type of film, Group III-V semiconductor, and a certain process for doing that very growth.
Id.
The patent’s written description repeatedly discloses that it is directed to a process for producing an epitaxial film of a Group III7V semiconductor.
See, e.g.,
’098 patent, col. 1,11.18; col. 2,11. 56-57; col. 3,11. 51-52; col. 4, 11. 46-47; col. 5, 11. 29-31; col. 8, 11. 13-14; col. 10, 11. 42-43. Similar language is contained in the preambles. “When the claim drafter chooses to use both the preamble and the body to define the subject matter of the claimed invention, the invention so defined, and not some other, is the one the patent protects.”
Bell Communications Research, Inc. v. Vitalink Communications
*1363
Corp.,
The trial court further interpreted the independent claims as reflecting a narrowing of the claim scope during prosecution, concluding that the claimed “open reactor” must be an “open
cold-ivall
reactor.”
Rockwell,
Although the parties agree that the claims are limited to cold-wall reactors, SDL contends that the trial court improperly adopted the definition of cold-wall reactor asserted by Rockwell instead of the definition asserted by SDL. In response to the third Office action rejection based on Reuhrwein, Dr. Mansevit referenced “the article by Stringfellow et al., supra, at pages 50 and 54 which diseuss[es] and distinguishes] cold wall versus hot wall apparatus and related organometallic processes.” 7 In that article, the authors defined a cold-wall system as one “where only the graphite pedestal is heated (by rf induction heating) and the SiO walls remain cool even though no special water or forced air cooling is employed.” Id. at 490. The trial court held that Dr. Manasevit incorporated the definition of cold-wall reactor from the Stringfellow et al, article. Id. The trial court thus defined a cold-wall reactor as one in which heat is applied at the substrate and the walls are kept significantly cooler than the substrate.
SDL agrees that the Stringfellow article was an appropriate place to find the definition of cold-wall reactor. However, SDL asserts that the definition adopted by the trial court misreads the article and that the definition supplied by its expert, the same Dr. Stringfellow, should be adopted. According to SDL, Dr. Stringfellow’s summary, in his declaration, of the statements made in his article lead to the following definition of a cold-wall reactor: “a reactor in which the walls of the reaction chamber are kept cool enough to avoid significant pyrolysis of the reactants on or near the walls of the reactor.” Declaration of Gerald B. Stringfellow, Ph.D., in support of the Cross-Motion of SDL, Inc. for Summary Judgment and Response to Motion for Summary Judgment by Rockwell, pp. 17-20 (May 2, 1996). Thus, SDL asserts that not only must a cold-wall reactor’s walls be significantly cooler than the substrate, but also be cool enough that no significant pyrolysis occurs on or near the walls.
We agree, however, with the trial court’s analysis of the definition of cold-wall reactor. As the trial court noted, although SDL is correct that “a cold-wall reactor may have significantly less pyrolysis on the walls of the chamber than the hot-wall reactor, [that] is the reason to use a cold-wall reactor and not the meaning of a cold-wall reactor.”
II. Anticipation
Anticipation under 35 U.S.C. § 102 requires the disclosure in a single piece of prior art of each and every limitation of a claimed invention.
See Electro Med. Sys. S.A. v. Cooper Life Sciences,
The four prior art patents analyzed by the trial court for both its anticipation and its obviousness determinations were: U.S. Pat. No. 8,266,270 (the “Miederer Patent”), U.K. Pat. No. 778,383 (the “Scott Patent”), U.S. Pat. No. 3,364,084 and U.K. Pat. No. 1,011,-979 (the “Ruehrwein Patents”). (The only patent in this group that was cited as a reference during the prosecution of the ’098 patent was the U.S. Ruehrwein patent.)
The trial court determined that genuine issues of material fact remained concerning what was disclosed in the prior art cited by Defendants.
Rockwell,
The question of whether the Miederer patent contains the single crystal substrate element is a dispute of fact that is material.
... The question of whether the terms of Miederer patent provides a[sic] sufficient basic chemistry information to enable one skilled in the art to achieve epitaxial single crystalline semiconductors, presents a material issue of fact that needs to be clarified with further ventilation through expert testimony.
Id. at 504-05. In addition, the trial court held that “[t]here is a genuine dispute of material fact as to whether the [Scott] patent terms provide sufficient basic chemistry information to enable one skilled in the art to grow epitaxial single crystalline semiconductors.” Id. at 505. Finally, the trial court determined that “[t]he information in the summary judgment papers is not sufficient to resolve the differences in expert opinion” as to whether the Ruehrwein patents taught single crystal growth using organometallics. Id. Hence, of the four prior art patents cited by Defendants, the trial court did not find that any one taught single crystal growth of a Group III/V semiconductor using orga-nometallics. Defendants have not shown that the trial court erred in identifying these genuine issues of material fact. Nor on appeal have Defendants indisputably shown that any of the four patents teaches such single crystal growth of Group IU/V semiconductors. The trial court’s determination, therefore, that there is a genuine dispute of material fact as to whether any of the four prior art references relied upon by Defendants taught a successful process for the growth of an epitaxial film using organome-tallics is affirmed.
III. Obviousness
Under 35 U.S.C. § 103(a),
[a] patent may not be obtained ... 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 matter pertains.
In determining obviousness, the invention must be considered as a whole without the benefit of hindsight, and the claims must be considered in their entirety.
See W.L. Gore & Assocs. v. Garlock, Inc.,
A. Genuine Issues About the Content of the Prior Art Precluded a Determination of Obviousness on Summary Judgment
After determining that genuine issues of material fact existed with respect to the prior art cited to prove anticipation, the trial court nevertheless accepted the assertions of Defendants as to the content of the same prior art patents for its obviousness analysis. From these assertions, it found that each limitation of the process claimed in the ’098 patent “is taught by the combination of the Miederer, Scott and Ruehrwein patents, or elsewhere in the CVD prior art.”
Rockwell,
Before the trial court and on appeal, Rockwell disputes this determination, contending that Defendants’ obviousness challenge “does not include ‘a single piece of prior art reporting the growth of gallium arsenide (or any other III/V semiconductor) single crystals using organometallie reagents,’ ”
id.,
the same limitation whose presence in the prior art the trial court found materially disputed. That prior art patents may have described failed attempts
9
or attempts that used different elements
10
is not enough. The prior art must be enabling.
11
See Motorola, Inc. v. Interdigital Tech. Corp.,
B. Inferences May Not Be Drawn Against the Nonmovant on Summary Judgment
I. The Trial Court Cannot Infer that One of Ordinary Skill in the Art Would Have Known How to Achieve Epitaxial Growth
Rockwell argues that the lack of any teaching in the prior art that single crystal
*1366
films of III/V semiconductor materials may
successfully be grown
is dispositive as to obviousness. The trial court held, however, that “[o]bviousnss standards under section 103 do not require such a showing. Invalidity for obviousness requires a showing that the prior art viewed as a whole would have suggested the claimed process to one skilled in the art.”
Rockwell,
The trial court had earlier determined with respect to anticipation that none of the prior art patents taught single crystal growth of Group III/V semiconductors using organome-tallics. Therefore, with respect to obviousness, the trial court could not simply find that these four patents, when combined with each other and unidentified “other CVD prior art,” taught the very limitation that admittedly none of them taught separately. Such a determination required the assumption or inference against Rockwell, the nonmovant, that somewhere in some prior art single crystal growth of Group III/V semiconductors was taught, and one of ordinary skill in the art would have known. Although there may be cases when the addition of the missing limitation would be clear to anyone looking at the problem, this is not one of those cases.
Single crystal growth of a Group III/V semiconductor using an organometallic reagent is a specific limitation of the claims of the ’098 patent. The trial court’s determination, therefore, that “[t]he Obviousness Claims Charts show that Miederer, Scott and Ruehrwein patents, when combined contain all of the elements included in claims 1 and 3 of the ’098 patent,”
id.
at 497, is not legally sufficient. It cannot support the conclusion that all of the claims in the ’098 patent are invalid for obviousness on the summary judgment record because doing so requires drawing of an adverse inference. Rather “[t]he consistent criterion for determination of obviousness is whether the prior art would have suggested to one of ordinary skill in the art that this process should be carried out and would have a reasonable likelihood of success.”
Dow,
We agree with the trial court that assessing the prior art patents involved material factual issues genuinely in dispute, precluding summary judgment of anticipation. We must take the next step, however, and hold that these material issues also precluded a conclusion of obviousness on summary judgment. Because of the presence of genuine issues of material fact regarding whether on this record the prior art patents taught a successful process for epitaxial growth of Group III/V semiconductors using organome-tallics, Defendants have not proven by clear and convincing evidence that the invention defined in the claims of the ’098 patent would have been obvious to one of ordinary skill in the art and that therefore they are entitled to judgment. We hold, on this record, that the Court of Federal Claims legally erred in determining on summary judgment that the claimed invention would have been obvious.
2. Inferences Adverse to the Nonmovant Regarding Secondary Considerations of Nonobviousness Are Also Impermissible on Summary Judgment
Evidence of secondary considerations is but a part of the “totality of the evidence” that is used to reach the ultimate conclusion of obviousness.
See Richardson-Vicks,
CONCLUSION
The decision that Defendants had not proven anticipation so as to support a grant of summary judgment is affirmed. Further, the evidence in the summary judgment record raises genuine issues of material fact concerning the obviousness determination. Accordingly, the summary judgment of the Court of Federal Claims holding the ’098 patent invalid for obviousness is vacated and the case is remanded for further proceedings on obviousness, anticipation and any other issues necessary to conclude the lawsuit.
AFFIRMED-IN-PART, VACATED-IN-PAJRT, and REMANDED-IN-PART.
COSTS
Each party shall bear its own costs.
Notes
. SDL, Inc. is also known as Spectra Diode Labs, Inc.
. Each of the contractors apparently had an interest in this action because their respective contracts with the United States may have obligated them to indemnify the United States should it lose a litigation as a defendant regarding items covered by their procurement contracts.
See Rockwell Int'l Corp. v. United States,
.Rules of the Court of Federal Claims.
. MOCVD: Metal-organic chemical vapor deposition. A technique for growing thin layers of compound semiconductors in which metal organic compounds, having the formula MRx, where M is a Group III metal and R is an organic radical, are decomposed near the surface of a heated substrate wafer, in the presence of a hydride of Group V element. McGraw-Hill Dictionary of Scientific and Technical Terms 1250 (5th ed. 1994).
. In its appeal, Rockwell challenges the trial court’s disposal of all of the claims of the '098 patent. Because we vacate and remand the trial court's summary judgment of obviousness on the asserted claims, we need not address whether all of the claims of the '098 patent stand or fall with the asserted ones.
. Independent claim 1 is representative for purposes of this appeal and is reproduced below:
1. An organo-metallic process for producing an epitaxial film of Group III-V semiconductor disposed on a single crystal substrate, said process employing an open reactor and comprising the steps of:
heating said substrate in said open reactor,
introducing into said open reactor a fist [sic] material containing a hydride or halide-free alkyl compound of at least one Group V constituent of said semiconductor, and, as a second material, at least one halide-free alkyl compound containing at least one of the Group III constituents of said semiconductor, and
exhausting said open reactor to pressure not greater than one atmosphere.
. The inventor was citing G.B. Stringfellow and H.T. Hall, 43 J. Ciystal Growth 47 (1978).
. Defendants assert that the works of Drs. Stearns and Thomas are prior art, but the trial court found that it was unclear on the summary judgment record if Drs. Stearns and Thomas reduced the invention to practice prior to February 7, 1967, the date of reduction to practice of the invention claimed in the '098 patent. The trial court does not appear to have used these works in its anticipation analysis, but to the extent that the trial court used these teachings on summary judgment either of anticipation or of obviousness, it erred because of its finding that a genuine issue of material fact existed regarding reduction to practice.
. For example, the trial court found that the Miederer patent taught the use alkyl gallium along with AsC13 to deposit a GaAs layer on a semiconductor rod substrate, but the trial court also found a material issue of fact as to whether one could grow epitaxial layers by following the procedure disclosed in Miederer. In fact, there was evidence by the inventor himself that the example in his patent did not accomplish epitaxial growth: “I did not accomplish my ultimate goal of growing a single crystal film of gallium arsenide; rather, the film I grew was polycrystalline." Although the inventor further stated that his “inability to grow epitaxial layers was not the result of any deficiency in the process ... but rather the result of the inadequacy of the facilities available ...,” there were differences between the failed Miederer process and the successful one disclosed in the '098 patent that also may have contributed to that failure.
. For example, the trial court found that there was a genuine issue of material fact as to whether the Ruehrwein patents specifically disclosed Group III alkyls in combination with alkyls or hydrides of Group V. Furthermore, the trial court found that “SDL's expert in his deposition stated that organometallie [pyrolysis] in a hot wall chamber would not create single crystalline epitaxy.”
Rockwell,
. The district court found a disputed issue of fact with respect to the enablement of epitaxial layer growth using the Scott patent. Rockwell's expert asserted that Scott would not enable a knowledgeable person to produce epitaxial film no matter what the operator's skill, but SDL's expert asserted otherwise.
