Full opinion text
FINDINGS OF FACT AND CONCLUSIONS OF LAW KEELEY, District Judge. This is an action for the infringement of a chemical invention protected by U.S. Patent No. 5,053,407 (issued Oct. 1, 1991) (“the ’407 patent”). The patent, entitled “Optically Active Pyridobenzoxazine Derivatives and Anti-Microbial Use,” claims a compound named levofloxacin. The plaintiffs in this suit are Daiichi Pharmaceutical Co., Ltd. (“Daiichi”), levofloxaein’s inventor and patent-holder, and Johnson & Johnson, the parent company of Ortho-McNeil Pharmaceutical, Inc. and Johnson & Johnson Pharmaceutical Research & Development, LLC (collectively “Ortho”) , which holds a license to distribute levoflox-acin in the United States. The defendants are Mylan Laboratories, Inc. and Mylan Pharmaceuticals, Inc. (collectively “My-lan”). Mylan committed an act of infringement by filing two Abbreviated New Drug Applications (“ANDAs”) (Nos. 76-276 & 77-097) with the Food and Drug Administration (“FDA”), seeking permission to manufacture and distribute a generic version of levofloxacin before the ’407 patent expires. 35 U.S.C. § 271(e)(2). The ANDAs included a so-called “Paragraph IV” certification, which asserted that the ’407 patent is invalid. See 21 U.S.C. § 355(j)(2)(A)(vü)(3V). As statutorily required, Mylan notified Daiichi of its ANDA filing. See id. §§ 355(j)(2)(B)(i)-(ii). In response, Daiichi and Ortho (collectively “Daiichi/Ortho”) filed a timely suit for infringement on February 22, 2002. Id. § 355(j)(5)(B)(iii). Mylan essentially concedes infringement of the ’407 patent, but contends that the patent is invalid on a number of grounds. Between November 4, 2003 and December 22, 2003, the Court held a bench trial on four of these invalidity defenses — -prior invention, indefiniteness, inequitable conduct and obviousness. On December 5, 2003, the Court entered judgment as a matter of law against Mylan’s prior invention and indefiniteness defenses pursuant to Rule 52(c) of Federal Rules of Civil Procedure. The Court then heard evidence on inherent anticipation, the final invalidity issue, from May 24-26, 2004. In lieu of closing arguments, the parties submitted lengthy post-trial briefs. All briefing concluded on July 12, 2004. Pursuant to Rule 52(a) of the Federal Rules of Civil Procedure, the Court now states its findings of fact and conclusions of law. As discussed below, the Court concludes that Mylan has failed to meet its burden to prove the invalidity of the ’407 patent. I. BACKGROUND OF THE INVENTION Only claims 2 and 5 of the ’407 patent are in dispute in this case. Generally speaking, claim 2 is a compound claim covering levofloxacin, and claim 5 is a method claim covering the administration of “an antimicrobially effective amount” of levofloxacin to a patient. The invention at issue in both claims is levofloxacin, which is an enantiomeric compound. An enantiomer is one of a pair of isomers that are non-superimposeable mirror images of each other. This mirror image structure is often likened to the relative structures of a person’s right and left hands, and chemists normally refer to each enantiomer as either the dextro (Latin dexter, or right-handed) or levo (Latin laevus, or left-handed) enantiomer. The right/left nomenclature also stems from the fact that enantiomers are inherently “optically active.” That is, an enan-tiomer will rotate a plane of polarized light clockwise (dextrorotatory) or counterclockwise (levorotatory). Moreover, a given pair of enantiomers will always rotate polarized light in equal and opposite directions. For example, if the dextroro-tatory enantiomer rotates polarized light 90 to the right (clockwise), then the levoro-tatory enantiomer will rotate the polarized light 90 to the left (counterclockwise). Because enantiomers have identical chemical formulae, chemists distinguish between the chemical names of enantiom-eric pairs by preceding each with a symbol that reflects the direction the enantiomer rotates polarized light: “(+)” for dextroro-tatory enantiomers, and for levorota-tory enantiomers. Chemists also distinguish between enan-tiomers by designating an enantiomer as either “R” or “S” based upon the arrangement of certain atoms at the enantiomer’s “chiral center.” Where one enantiomer is an “R,” the other will be an “S.” When chemists first find or synthesize a given enantiomeric pair, the enantiomers always occur in a perfect 1:1 ratio. This solution of equal amounts of dextrorotato-ry and levorotatory enantiomers is known as a “racemic compound.” A racemic compound is optically inactive because, for every dextrorotatory enantiomer rotating polarized light to the right, there exists a levorotatory enantiomer rotating light to the left, resulting in a net rotation of zero. Chemists also have a specific nomenclature for racemic compounds — the chemical name is preceded by either “(±)” or “RS” (or both). In this case, ofloxacin is a racemic compound comprised of one dextrorotatory en-antiomer with an “R” configuration and one levorotatory enantiomer with an “S” configuration. Racemic ofloxacin is disclosed in U.S. Patent No. 4,382,892 (“the ’892 patent”). Levofloxacin, the subject of the ’407 patent, is the levorotatory isomer of ofloxacin with the “S” configuration. II. AMENDED CLAIM CONSTRUCTION A. Original Claim Construction The Court has previously construed the claims in the case at bar. See Ortho-McNeil Pharms., Inc. v. Mylan Labs., Inc., 267 F.Supp.2d 533 (N.D.W.Va.2003). Specifically, the Court held: An examination of the plain meaning of the claim language as understood by persons skilled in the art at the time of invention, the specification and the prosecution history indicate that “An S(-)-pyridobenzoxazine compound” and “S (-)-9-Fluoro-3-methyl-10-(4-methyl-l-piperazinyl)-7-oxo-2,3-dihydro-7H-py-rido[l,2,3-de] [l,4]benzoxazine-6-car-boxylic acid” refer to the levorotatory enantiomer of racemic ofloxacin, levo-floxacin. These terms do not refer to racemic ofloxacin. Furthermore, as demonstrated by the specification’s resolution methodology, as well as the specification and prosecution history’s repeated emphasis on levofloxacin’s unique pharmacological properties, the disputed language refers more specifically to a pharmaceutical preparation comprised principally of levofloxacin. Id. at 544. (emphasis added). The Court now recognizes that its prior construction suffered from several infirmities that need to be corrected. First, although that construction relies heavily on the specification and prosecution history, it does not explain what considerations justified their use as a source for claim limitations. Further, the Court failed to differentiate between the multiple claims at issue. Finally, the Court’s perspective on the claims has been sharpened by My-lan’s injection of new defenses. Accordingly, the Court finds it necessary to amend its claim construction. B. Standard of Law When construing patent claims, the Court must look first to the intrinsic evidence in the record: “The claims, the specification, and the prosecution history.” Markman v. Westview Instruments, Inc., 52 F.3d 967, 979 (Fed.Cir.1995) (en banc). The court does not look to each of these three sources equally; rather, they are a “hierarchy of analytical tools.” Digital Biometrics, Inc. v. Identix, Inc., 149 F.3d 1335, 1344 (Fed.Cir.1998). With these varying sources of information, the Court must be careful to always focus on interpreting the claim language as written. See Eastman Kodak Co. v. Goodyear Tire & Rubber Co., 114 F.3d 1547, 1552 (Fed.Cir.1997) (“... a construing court does not accord the specification, prosecution history, and other relevant evidence the same weight as the claims themselves, but consults these sources to give the necessary context to the claim language.”) Thus, the most important part of the court’s analysis is the claim language itself. See Digital Biometrics, 149 F.3d at 1344. (“The actual words of the claim are the controlling focus.”); see also Johnson Worldwide Assocs., Inc. v. Zebco Corp., 175 F.3d 985, 989 (Fed.Cir.1999) (“We begin, as with all claim interpretation analyses, with the language of the claims”). A court should interpret technical terms in claim language as having the meaning understood by persons skilled in the art, or having experience in the field at the time of invention. Hoechst Celanese Corp. v. BP Chems. Ltd., 78 F.3d 1575, 1578 (Fed.Cir.1996). Dictionaries have become not only a permissible source, but also a significant source, in ascertaining claim meaning. Tex. Digital Sys. v. Telegenix, Inc., 308 F.3d 1193, 1202-03 (Fed.Cir.2002), cert. denied, 538 U.S. 1058, 123 S.Ct. 2230, 155 L.Ed.2d 1108 (2003) (“Dictionaries are always available to the court to aid in the task of determining meanings that would have been attributed by those of skill in the relevant art to any disputed terms used by the inventor in the claims.”). As the Texas Digital court explained: “Dictionaries, encyclopedias and treatises, publicly available at the time the patent is issued, are objective resources that serve as reliable sources of information on the established meanings that would have been attributed to the terms of the claims by those of skill in the art.” Id. at 1202-03 (citations omitted). As with other sources of meaning, however, the dictionary definition must be disregarded if “the specification uses the words in a manner clearly inconsistent with the ordinary meaning reflected ... in a dictionary definition.” Id. at 1204 (emphasis added) (citations omitted). Next is the specification, which the Court must consult when construing the claim language. See Markman, 52 F.3d at 979 (“Claims must be read in view of the specification, of which they are a part.”); see also SciMed Life Sys., Inc. v. Advanced Cardiovascular Sys., Inc., 242 F.3d 1337, 1340-41 (Fed.Cir.2001) (stating that it is “proper” for a district court to follow Markman’s invocation); Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed.Cir.1996) (“[T]he specification is always highly relevant to the claim construction analysis.”) Importantly, the Court must guard against importing limitations into the claim language from the other intrinsic and extrinsic evidence where the claim language itself does not warrant it. See Johnson Worldwide, 175 F.3d at 989-90 (“[C]laim terms cannot be narrowed by reference to the written description or prosecution history unless the language of the claims invites reference to those sources.”); Electro Med. Sys., S.A. v. Cooper Life Sci., Inc., 34 F.3d 1048, 1054 (Fed.Cir.1994) (“[C]laims are not to be interpreted by adding limitations appearing only in the specification.”); SRI Int’l v. Matsushita Elec. Corp., 775 F.2d 1107, 1121 n. 14 (Fed.Cir.1985) (“Specifications teach. Claims claim.”). The third and final type of intrinsic evidence is the prosecution history. Markman, 52 F.3d at 980. The Court should consult the prosecution history to determine whether the patent applicant “consistently and clearly use[s] a term in a manner either more or less expansive than its general usage in the relevant community.” Resqnet.com, Inc. v. Lansa, Inc., 346 F.3d 1374, 1378 (Fed.Cir.2003) (citation omitted). For the prosecution history to limit the scope of the claim, however, the actions or statements that allegedly disavow a possible construction of the claim language must be “clear and unmistakable.” Id. (citation omitted). In addition to the intrinsic evidence, the Court may consult extrinsic evidence, such as expert testimony, if necessary. See Markman, 52 F.3d at 979 (“Expert testimony, including evidence of how those skilled in the art would interpret the claims, may also be used.”) (internal quotation marks omitted). To determine what one skilled in the art could understand, the court should consider the testimony of scientific expert witnesses. See AFG Indus., Inc. v. Cardinal IG Co., 239 F.3d 1239, 1249 (Fed.Cir.2001) (noting the value of having “scientific witnesses to aid the court in coming to a correct conclusion”); see also Key Pharms. v. Hercon Labs., Corp., 161 F.3d 709, 716 (Fed.Cir.1998) (“[T]rial courts generally can hear expert testimony for background and education on the technology implicated by the presented claim construction issues, and trial courts have broad discretion in this regard.”) Courts must be careful, however, that the expert testimony is being used only to ensure that the plain meaning is not contradicted by the understanding of one of ordinary skill in the art, not to rewrite the claim. See Voice Technologies Group v. VMC Sys., Inc., 164 F.3d 605, 614 (Fed.Cir.1999) (“When the intrinsic evidence is unambiguous, it is improper for the court'to rely on extrinsic evidence.”). When a claim term is amenable to two. or more interpretations based on the applicable record evidence,, it should be construed to preserve the patent’s validity. Harris Corp. v. IXYS Corp., 114 F.3d 1149, 1153 (Fed.Cir.1997); ACS Hosp. Sys., Inc. v. Montefiore Hosp., 732 F.2d 1572, 1577 (Fed.Cir.1984). C. Claim Construction, Revisited 1. Levofloxacin Distinguished from the Racemate Significantly, Mylan has failed to cast doubt on the Court’s key conclusion in its previous claim construction: the chemical structure referenced in claims 2 and 5, on the face of the claims, as confirmed by the specification and’ prosecution history, and as informed by extrinsic evidence, do not refer to racemic ofloxacin. The claims of the ’407 patent that must be construéd read: 1. An S(-)-pyridobenzoxazine compound represented by the formula (VI) wherein XI represents a halogen atom, R1 represents an alkyl group having 1 to 4 carbon atoms, and R3 represents an alkyl group having 1 to 3 carbon atoms. 2. S(-)-9-Fluoro-3-methy l-10-(4-methyl-l-piperazinyl)-7-oxo-2,3-dihydro-7H-pyrido[l,2,3-de] [l,4]benzoxazine-6-earboxylic acid according to claim 1. 5. A process for treating a patient in need of an antimicrobial therapy in claim 4 which comprises administering to said patient an antimicrobially effective amount of S(-)-9-Fluoro-3-methyl-10-(4 -methyl-l-piperazinyl)-7-oxo-2,3-dihydro-7H-pyrido[l,2,3-de] [l,4]ben-zoxazine-6-carboxylic acid. The parties continue to dispute the proper construction of the terms “[a]n S(-)-pyridobenzoxazine compound” and “S(-)-9-Fluoro-3-methyl-10-(4-methyl-l-piper-azinyl)-7-ox o-2,3-dihydro-7H-pyri-do[l,2,3-de][l,4]benzoxazine-6-carboxylic acid.” An examination of the plain meaning of the claim language as understood by persons skilled in the art at the time of invention, the specification and the prosecution history indicates that “[a]n S(-)-pyridoben-zoxazine compound” and “S(-)-9-Fluoro-3-methyl-10-(4-methyl-l-piperazinyl)-7-ox o-2,3-dihydro-7H-pyrido[l,2,3-de][l,4]benzoxazine-6-carboxylic acid” refer to the levorotatory enantiomer of ra-cemic ofloxacin, levofloxaein. These terms do not refer to racemic ofloxacin. Before addressing the dispute, the Court notes that the parties agree that, to one skilled in the art, claims 2 and 5 of the ’407 patent plainly refer to the “S(-)” optical isomer (enantiomer) of ofloxacin, levofloxa-cin. Despite this agreement, Mylan argues that the chemical name in claims 1 and 4 needs a plain-English “purity” qualification to avoid a breadth of coverage that would include the prior art racemic ofloxacin. There are a number of problems with this position, however. First, Mylan’s argument directly conflicts with its position that the plain language of the claim refers to levofloxaein. As discussed above, chemists skilled in the art regard levorotatory enantiomers as distinct from racemic compounds or the dextrorotatory enantiomer. Additionally, each type of compound has its own unique nomenclature. “S(-)” clearly designates the levorotatory enantiomer in this case. Had the inventor meant to designate the racemic compound, he would have used the designation “(±)” or “RS.” Even Mylan’s own expert testified at his deposition that it “would be an error” to use only the (-) symbol to designate a racemic compound, (Jordis Dep. at 22), and a chemist would not use a lone “S” to designate a racemic compound. (Id. at 176.) Mylan also asserts that In re Williams, 36 C.C.P.A. 756, 171 F.2d 319 (Oust. & Pat.App.1948), requires a plain-English “purity” limitation in claims for enantiom-ers. In Williams, the court was faced with an inventor seeking to patent an en-antiomer. The claim language “call[ed] for the laevo rotary form ‘substantially free from the dextro rotary form.’ ” Id. at 151, 171 F.2d 319. The Court did not construe the patent claims; indeed, aside from the preceding quotation, the claim language does not appear in the opinion at all. Thus, it is impossible to determine if the Williams court required the “substantially free” language as Mylan urges. The “substantially free” language certainly distinguishes the levorotatory enantiomer from the racemic compound. However, there is no indication that such a plain-English purity limitation is the only way to , distinguish the prior art. The case of In re May, 574 F.2d 1082 (Oust. & Pat.App.1978), suggests otherwise. May, too, involved the patentability of an enantiomer. The claims at issue in that case stated: 1. A method of affecting analgesic and morphine antagonistic activity without producing physical dependence in animals which comprises administering to an animal an effective dosage of an acid addition salt of the levo isomer of a compound of the structure where R is a lower alkyl group and R 1 is hydrogen or a lower alkyl group. 2. The method of claim 1 wherein said compound is -(-)-5,9-diethyl-2’hydroxy-2-methyl-6,7-benzomorphan. 3. The method of claim 1 wherein said compound is (-)-5-methyl-2’-hydroxy-2-methyl-6,7-benzomorphan. 4. The method of claim 1 wherein said compound is (-)-5-ethyl-2’-hydroxy-2-methyl-6,7-benzomorphan. 5. The method of claim 1 wherein said compound is -(-)-5-propyl-9-methyl-2’-hydroxy-2-methyl-6,7-benzomorphan. 6. The method of claim 1 wherein said salt is the hydrochloride. 7. The method of claim 6 wherein said compound is -(-)-5,9-diethyl-2’hydroxy-2-methyl-6,7-benzomorphan. 8. The method of claim 6 wherein said compound is (-)-5-methyl-2’-hydroxy-2-methyl-6,7-benzomorphan. 9. The method of claim 6 wherein said compound -is (-)-5-ethyl-2’-hydroxy-2-methyl-5,7-benzomorphan. 10. The method of claim 6 wherein said compound is -(-)-5-propyl-9-methyl-2-hydroxy-2-methyl-6,7-benzomorphan. 11. A pharmaceutical composition for internal.administration having an analgesic, non-addictive, morphine-antagonistic effect which comprises a pharmaceutical carrier and an effective amount of ■ an acid addition salt of — ( —)—5,9-diethyl — 2’—hydroxy—2—methyl—6,7—benzo-morphan. 12. The composition of claim 11 wherein said salt is the hydrochloride. 13. The composition of claim 11 wherein said salt is the acetate. Id. at 1084-85. -Claims 1 and 6 were rejected because they were specifically described in the pri- or art. Id. at 1089-90. The remaining claims were upheld. Notably, these claims contain no plain-English purity Imitations whatsoever. Instead, they distinguish the enantiomer from the racemic compound with the symbol “(-).” Mylan argues that the distinguishing language is not but rather the phrase “without producing physical dependence in animals,” which describes a chemical attribute unique to the levorotatory enan-tiomer. There are two problems with this analysis, however. First, the language emphasized by Mylan modifies the. method for which claim 1 sought patent protection, not the compound through which that method was effected. The compound itself is simply described as “an acid addition salt of the levo isomer of a compound [with the following structure .... ].” Thus, the description of the compound contains no plain-English purity limitation. Most importantly, the May court stated that, under common nomenclature, a chemical compound designated as is “limited to the levo enantiomer.” 574 F.2d at 1085. Thus, while it is certainly necessary to distinguish a new invention over the prior art, there is no indication that an inventor must use a plain-English purity limitation, as Mylan contends. Instead, an inventor may use anything that a person skilled in the relevant art would understand to limit the claim. In this case, the term “S(-)” clearly and plainly limits the claim language to the levorotatory enantiomer. Those skilled in the art clearly understand the term “S(-)” to affirmatively denote only the levorotatory enantiomer of a ra-cemic compound, and not the racemic compound itself. Furthermore, those skilled in the art clearly understand the terms “RS” or “(±)” to affirmatively denote a racemic compound. The inclusion of “S(-)” in the claim language, coupled with the obvious exclusion of “RS” or “(±),” militates against Mylan’s assertion that an additional plain-English purity limitation is necessary to distinguish the patented invention over the prior art racemic ofloxa-cin. This interpretation is not contradicted by the specification or the prosecution history. Indeed, it is directly confirmed by the prosecution history. As the Court observed in its March 31, 2003, Order: The prosecution history is replete with instances where the inventor distinguishes levofloxaein from the prior-art racemic ofloxacin. The first three claims of the ’407 patent were rejected by the patent examiner twice on the grounds that they were obvious in light of the prior art disclosure of racemic ofloxacin. Daiichi presented evidence of the differences between levofloxaein and ofloxacin until the examiner approved the patent as written. The Court must now revisit its conclusion that the claims cover “a pharmaceutical preparation comprised principally of levofloxaein” by reviewing the language of claims 2 and 5 in light of the intrinsic record. 2. The Levofloxaein “Compound” Claims 2 and 5, by incorporating the term from claims 1 and 4, respectively, refer to a “compound” with the structure S(-)-9-Fluoro-3-methyl-10-(4 -methyl-1piperazinyl)-7-oxo2,3-dihydro-7H-pyrido[l,2,3-de][l,4]benzoxazine-6-carboxylie acid. In its March 31, 2003 Order, the Court equated the term “compound” with “a pharmaceutical preparation” or drug. In reaching this conclusion the Court relied on the specification and the prosecution history, which referred to the properties of the compounds, not to confirm this limitation, but as its source. In light of the Federal Circuit’s recent decision in SmithKline Beecham v. Apotex, 365 F.3d 1306 (Fed.Cir.2004), the Court is compelled to revisit its conclusion that claim 2, although used as a drug, has been claimed as anything more limited than a compound defined solely by chemical structure. Mylan maintains that the word “compound” and the accompanying chemical formula comprise a broad compound claim to even a single molecule with that chemical structure. To the contrary, Daiichi/Or-tho argues that claim 2 must be construed to cover more than a single molecule of ofloxacin because the specification and prosecution history contain the additional limitation that the claimed compound be pharmaceutically effective. It emphasizes that a single molecule would be neither optically active nor pharmaceutically effective. At trial, Daiichi/Ortho and Mylan presented conflicting expert testimony as’ to how one of ordinary skill in the art would understand the term “compound.” My-lan’s expert on Patent and Trademark Office (“PTO”) procedures, George M. Gould, opined that a compound claim is the broadest possible claim, (Gould Tr. at 227.) Specifically, a “compound claim is directed to the molecule itself.” (Gould Tr. at 228.) This opinion was echoed by Dr. Mitscher, who stated that claim 2 “identifies a particular molecule.” (Mitscher Tr. at 728.) . To the contrary, Daiichi/Ortho’s expert, Dr. Klibanov, maintained that the term “compound” as used in claim 1 of the ’407 patent (and as otherwise incorporated therein) indicates that the patent is not referring to a “molecule or couple of molecules sitting somewhere.” (Klibanov Tr. at 1881). Under his definition of compound, the patent claim unambiguously refers to a drug: “The use of the word compound here implies that we are talking about a sizeable quantity of a material, a quantity that has certain properties, physical properties, chemical properties, and pharmacological properties.” (Id.) In support of its single molecule definition of “compound,” Mylan cites multiple cases in which a compound has been construed to refer a single molecule. See, e.g., Dow Chem. Co. v. Am. Cyanamid Co., 816 F.2d 617, 618 (Fed.Cir.1987) (“A nitrile is an organic compound containing a carbon-to-nitrogen triple bond and, depending on the rest of the molecule, can be classified as either aromatic or aliphatic”); Studiengesellschaft Kohle GmbH v. Eastman Kodak Co., 616 F.2d 1315, 1320 (5th Cir.1980) (“The simplest hydrocarbon molecule is a compound of one carbon atom and four hydrogen atoms and^ is commonly known as methane, represented by the chemical symbol CH 4.”). Most prominently, Mylan relies on the Federal Circuit’s recent decision in SmithKline Beecham v. Apotex, 365 F.3d 1306 (Fed.Cir.2004). As in the case at bar, Apotex involved a pharmaceutical patent. Engaging in a analysis similar to the one this Court followed in its prior order, the district court concluded that what appeared to be a claim defined only by chemical structure should be construed to cover only “commercially significant” quantities of that compound. Id. at 1310. On appeal, the Federal Circuit reversed this claim construction, holding that it improperly looked beyond the claim language because the “language [was] not ambiguous, but rather deseribe[d] a very specific compound.” Id. at 1313. Mylan further cites Schering Corp. v. Geneva Pharmaceuticals, 339 F.3d 1373, 1381 (Fed.Cir.2003), as evidence of the Federal Circuit’s approach to compound claims. Although Schering involved an agreed claim construction, the unambiguous language employed by the Court suggests that the Federal Circuit will read compound claims broadly: “[C]ompound claims ... broadly encompass compounds defined by structure only. Such bare compound claims include within their scope the recited compounds as chemical species in any surroundings, including within the human body as metabolites of a drug.” Id. Daiichi/Ortho does not cite any case law supporting a contrary construction of the term “compound.” Rather, the plaintiffs argue that, unlike the patentee in Apotex, they relied on unexpected results as the basis for establishing levofloxacin’s patent-ability over prior art ofloxacin, and thus clearly disavowed a claim to levofloxacin that did not demonstrate those unexpected results. Under a standard dictionary definition, “compound” means “a chemically distinct substance formed by union of two or more ingredients (as elements) in definite proportion by weight and with definite structural arrangement <water is a [compound] of oxygen and hydrogen>.” Webster’s Third New Int’l Dictionary 466 (2002). A “molecule” is “a unit of matter that is the smallest particle of an element or chemical combination of atoms (as a compound) capable of retaining chemical identity with the substance in mass.” Id. at 1455. Therefore, Mylan’s proposed definition of “compound” appears consistent with Apo-tex, Sobering and the dictionary definition. The Court’s construction of the disputed claims, however, does not end with that term. a. Optical Activity Limitation The Court has already concluded that “S(-)” in the claims indicates that the claimed substance is the levorotatory en-antiomer. The S describes the configuration in space. The parties dispute whether the (-) in the claim language limits the claim. The Court must, therefore, determine how one of ordinary skill in the art would read the (-). Dr. Klibanov testified that the (-) indicates that the compound is “optically active,” i.e., it will rotate a plane of polarized light counterclockwise. (Kli-banov Tr. at 1881.) Dr. Mitscher neither disputed nor conceded that actual optical activity was required to merit the S(-) designation, but admitted that each molecule of S(-) will contribute to the optical activity of a solution of S(-) molecules. (Mitscher Tr. at 6117.) Because the parties dispute the significance of the S(-), the Court must consult a dictionary. See International Union of Pure and Applied Chemistry (“IUPAC”), Basic Terminology of Stereochemistry (IUPAC Recommendations 1996), at http://www.ehem.qmul.ac.uk/iupac/ster-eo/NQ.html. IUPAC, one of the principal authorities on stereochemistry, defines “optical activity” as follows: Optical Activity A sample of material able to rotate the plane of polarisation of a beam of transmitted plane-polarised light is said to possess optical activity (or to be optically active). This optical rotation is the classical distinguishing characteristic (sufficient but not necessary) of systems containing unequal amounts of corresponding enantiomers. An enantiomer causing rotation in a clockwise direction (when viewed in the direction facing the oncoming light beam) under specified conditions is called dextrorotatory and its chemical name or formula is designated by the prefix (+)-; one causing rotation in the opposite sense is laevoro-tatory and designated by the prefix (-)-. Id. This definition comports with Dr. Kliba-nov’s testimony. According to IUPAC, “[a]n enantiomer ... causing rotation” in a counterclockwise direction is “laevorotato-ry and designated by the prefix Thus, the (-) indicates that the claims refer to an “optically active” compound. This reading of S(-) is confirmed by the specification. The specification repeatedly refers to the invention as “optically active.” ’407 patent, col. 1:6-11 (“[The] invention relates to optically active pyridobenzoxazine derivatives” and “optically active compounds of Ofloxacin and its analogs”); id. at col 1:25-26 (“The present inventors obtained optically active compounds of the racemic Ofloxacin .... ”); id. at col. 2:29-40 (“to provide optically active Ofloxacin and its analogs”; “to provide a novel intermediate ... useful for synthesizing optically active Oflox-acin”; “to provide a novel process for preparing optically active Ofloxacin and its analogs by the use of [that] intermediate”); id. at 2:65-67 (noting three methods for preparing “optically active Ofloxa-cin”). Thus, the specification accords with the meaning that a person of ordinary skill in the art would have attributed to the (-) designation. It is undisputed that more than a single molecule is required to manifest optical activity. (Klibanov. Summ. J. at 42). Not only was Dr. Klibanov’s testimony on this point unimpeached and uncontradict-ed, but Mylan’s expert, Dr. Mitscher, also admitted that a collection of levofloxacin molecules is required to exhibit optical activity, because a “single molecule would be too small” to measure its optical rotation. (Mitscher Tr. at 991.) Thus, on their face, the claims require at least enough levoflox-acin molecules to rotate polarized light. b. Purity Limitation The claims do not expressly refer to a specific minimum optical purity or percentage of the dextrorotatory enantiomer. In its prior claim construction, however, the Court described the claims as covering something “comprised principally of levo-floxacin.” That terminology gave rise to significant debate during trial. The language, however, was intended to capture a concept that was only reinforced at trial: a person of ordinary skill in the art at the time of the filing of the ’407 patent would have understood the claim to cover levo-floxacin with a purity that was highly, but less than 100 percent, optically pure. The Court did not reference a minimum purity because no minimum purity was claimed or identified in the intrinsic record. To determine what one skilled in the art could understand, the court should consider the testimony of scientific expert witnesses. See AFG Indus., Inc. v. Cardinal IG Co., 239 F.3d 1239, 1249 (Fed.Cir.2001). The Court is dealing here with highly technical terms. Therefore, it must rely on the education it received from the experts as to whether one of ordinary skill in the art would have read the reference as referring to a 100 percent pure substance. Because of the nature of enantiomeric separation at the time, one of ordinary skill in the art would not have read a claim to an enantiomer of ofloxacin as requiring 100 percent purity. Dr. Klibanov testified that when chemists refer to 100 percent purity they understand that “there is always that molecule hiding somewhere in the corner. There is no such thing as an absolutely 100 percent .0000 pure substance.” (Klibanov Tr. at 1899.) Mylan’s expert, Dr. Mitscher, implicitly conceded this when he calculated the optical purity of Example 6. He “arbitrarily” treated Example 7 as 100 percent pure, recognizing that it “may, in fact, not be 100 percent [pure],” but it was “the purest sample available” in the ’407 patent. (Mitscher Tr. at 741-42.) Further, Mitscher stated that the purity number he would require was “very high, 100 percent if possible, but you know as pure as can reasonably be obtained for a product at this time.” (Mitscher Tr. at 932.) Thus, although one of ordinary skill in the art would have understood the claim to the compound levofloxacin to be substantially pure levofloxacin, the realities of science would have led such a skilled artisan to conclude that the purity was not 100 percent. This is confirmed by the specification, particularly in the examples. As Drs. Klibanov and Mitscher testified, not all of the examples, and perhaps none of the examples, yielded 100 percent pure levofloxacin. Therefore, the Court declines to adopt such a limitation and construes claim 2 of the ’407 patent to cover substantially pure levofloxacin. 3. Construction Issues Unique to Claim 5 Claim 5 provides more description than the chemical structure described in claim 2. It is a claim to a method of administering levofloxacin. At issue is whether the language in the claim limits the claim in any other respect. The language “administering to said patient an antimicrobially effective amount” of levofloxacin could be read to provide a quantity limitation. Specifically, it could limit the claim to levofloxacin in quantities that are antimicrobially effective. The Federal Circuit was presented with a similar claim construction issue in Key Pharmaceuticals v. Hereon Laboratories Corp., 161 F.3d 709 (Fed.Cir.1998), which involved the construction of the term “pharmaceutically effective amount” in a patent for a patch that delivered nitroglycerin through the skin. It affirmed the district court’s use of extrinsic evidence, in the form of FDA ranges for the required quantity, to place a numerical value on “pharmaceutically effective.” Although the parties in that case disputed how much nitroglycerine would constitute a “pharma-ceutically effective” amount, there was no suggestion that the term did not limit the claim at issue. Applying that implied principle to the case at bar, the Court finds that the term “antimicrobially effective” limits the claim by requiring enough levo-floxacin molecules to exhibit antimicrobial activity. Further, claim 5 discusses “administering” the levofloxacin. Daiichi/Ortho maintains that this term limits the claim to something delivered from outside the body and thereby excludes in vivo production of levofloxacin. In response, Mylan argues that if ofloxacin were to become levofloxa-cin in vivo, an antimicrobially effective amount of levofloxacin would have been administered when ofloxacin was administered. It therefore maintains that the language does not exclude in vivo production. Mylan’s construction is supported by the Merriam Webster Medical Dictionary’s, definition of administer, “to give remedially (as medicine).” Whether levofloxacin formed as the claimed compound inside the body or outside the body, as long as it is given remedially as medicine, then levo-floxacin has been administered. Thus, claim 5 does not contain a preingestion limitation. 4. Claims Construed Claim 2 refers to a compound comprised of an optically active and substantially pure quantity of levofloxacin. Claim 5 is a method of administering an antimicrobially effective amount of the compound identified in claim 2. III. JUDGMENTS AS A MATTER OF LAW At the close of Mylan’s case-in-chief, Daiichi/Ortho moved for judgment on partial findings pursuant to Federal Rule of Civil Procedure 52(c). Rule 52(c) provides: If during a trial without a jury a party has been fully heard on an issue and the court finds against the party on that issue, the court may enter judgment as a matter of law against that party with respect to a claim or defense that cannot under the controlling law be maintained or defeated without a favorable finding on that issue .... The rule further requires that the judgment shall be supported by-findings of fact and conclusions of law. Id. The Court granted Daiichi/Ortho’s Rule 52(c) motion as to Mylan’s prior invention and indefiniteness defenses, making oral findings of fact and conclusions of law. The Court now issues written findings so that the record on appeal is clear. A. Prior Invention 1. Standard of Law Invalidity based on prior invention is a question of law that must be proven by clear and convincing evidence. Gambro Lundia AB v. Baxter Healthcare Corp., 110 F.3d 1573, 1576 (Fed.Cir.1997). Under the patent laws, a person shall be entitled to a patent unless: [B]efore such person’s invention thereof, the invention was made in this country by another 'inventor who had not abandoned, suppressed, or concealed it. In determining priority of invention ..‘. there shall be considered not only the respective dates of conception and reduction to practice of the invention, but the reasonable diligence of one who was the first to conceive and last to reduce to practice, from a time prior to conception by the other. 35 U.S.C. § 102(g)(2). Ordinarily, under. United States patent law, it is the date of conception, not the filing date or the date of .the actual production of the invention, that establishes the invention date. Scott v, Koyama, 281 F.3d 1243, 1245-46 (Fed.Cir.2002); Fina Oil and Chem. Co., 123 F.3d 1466, 1473 (Fed.Cir.1997). Conception alone, however, is not sufficient to prevail on the defense of prior invention; subsequent diligent reduction to practice must also be proven. Price v. Symsek, 988 F.2d 1187, 1190 (Fed.Cir.1993). This framework applies only to inventions conceived in the United States. See 35 U.S.C. § 102(g)(2). When an invention was not conceived in- the United States, but is the subject of an earlier-filed foreign patent application, the applicant may claim a priority date of the foreign- application if the foreign application meets certain requirements. Tronzo v. Biomet, Inc., 156 F.3d 1154, 1158 (Fed.Cir.1998). Nonetheless, applicants with a foreign patent may only receive credit for activities performed outside the United States that were included in a patent application. Scott, 281 F.3d at 1246 n. 2. Reliance on a foreign application’s date is a “constructive reduction to practice.” 35 U.S.C. § 104. Because a foreign patent holder does not receive recognition for conception, however, it will lose a priority contest if another party can prove conception in the United States before the filing date, of the foreign patent.. See id. at 1246. This is true even if- the inventor in the United States does not succeed in reducing the invention to practice until after the foreign filing date, as long as he or. she can prove diligence and ultimate success in a reduction to practice. See id. 2. Findings of Fact and Conclusions of Law a. The Japanese Patents The .’407 patent application was filed with the PTO on- June 5, 1986. One year before that filing in the United States, Daiichi filed the first of three Japanese patent applications, which, taken together, comprise the ’407 patent application. Each subsequent application disclosed a new process for making levofloxacin. These processes are referred to in the ’407 patent as Processes A, B, and C. Daiichi filed its first patent application claiming levofloxacin, Japanese Patent Application No. 60-134712 (the “ ’712 application”), in Japan on June 20, 1985. (PX 981.) That application disclosed Process A, which describes the synthesis of levo-floxacin by resolving an intermediate column that Daiichi had produced using a High Performance Liquid Chromatography (“HPLC”) column. Process A was limited in that it did not allow for the production of large quantities of levofloxa-cin. The levofloxacin produced by Process A was also less pure than that produced by the later-developed processes. Process A corresponds to Example 6, which is identical in the ’712 and ’407 patent applications. (DX 1 at col. 14, lines 6-45; DX 8 at 315-16.) The second Japanese patent application for levofloxacin, No. 30-2226499 (the “ ’499 application”), filed on October 11, 1985, disclosed Process B, which involved the production of levofloxacin using the lipase enzyme from Pseudomonas aeruginosa. This process allowed for the production of greater quantities of levofloxacin, which, in turn, provided the opportunity for more significant analysis and testing of the drug. The third Japanese application, filed on January 28, 1986, disclosed Process C. Process C was adapted from a process devised by Dr. John F. Gerster and presented at a September 1985 conference sponsored by the Interscience Conference on Antimicrobial Agents and Chemotherapy (“ICAAC”). Gerster’s work did not involve levofloxacin, but another fluoroqui-nolone called flumequine. Process C involves the use of L-tosyl proline as a reagent and allows for production of greater quantities of levofloxacin than either Process A or Process B. All three processes were included in the ’407 patent application. The ’407 patent included a total of seventeen (17) examples describing the production of levofloxacin and its intermediates, and covering three distinct processes. (DX 1.) On December 9, 1986, Bayer AG, another pharmaceutical company, submitted a patent application for levofloxacin. (DX 11.) Although that application went through a separate patent prosecution process, the PTO ultimately found that both patent applications were allowable. (Id.) Therefore, in June 1989, the PTO declared an interference proceeding to determine which party had priority, with the Daiichi applicants being declared the senior party and the Bayer AG applicants the junior party. (DX 8 at 276-80.) In connection with the interference proceeding, the Daiichi parties submitted a preliminary motion “to be Accorded Benefit of its Japanese Priority Applications.” (DX 17.148.) The Bayer AG parties opposed this motion on the ground that the Japanese applications did not disclose the entire subject matter of the ’407 patent. (DX 17.216.) In February 1991, the PTO’s Board of Patent Appeals and Interferences granted Daiichi’s motion over Bayer’s objection, and Daiichi was given the benefit of the priority date of its first Japanese patent application, June 20, 1985. (DX 17.414-19.) b. Mylan’s Claims Mylan’s case on prior invention had two essential components, both of which it needed to prove in order to prevail. First, Mylan challenged Daiichi’s June 20, 1985 priority date, claiming that the priority date should be October 11, 1985,- the date of its second patent application containing Process B. Second, it identified an alleged prior invention of levofloxacin in September 1985 by its expert, Dr. Mitscher, along with Dr. Daniel Chu, a scientist then employed by Abbott Laboratories. . If proven, that September 1985 invention would have pre-dated Daiichi’s second Japanese patent application by one month. For its prior invention defense, Mylan bore the burden of establishing in its casein-chief, by clear and convincing evidence, that (1) the ’712 patent application did not cover levofloxacin as claimed in the ’407 patent, and (2) Mitscher and Chu conceived of levofloxacin as claimed in the ’407 patent in mid-September 1985 and, thereafter, diligently reduced that invention to practice. i. Entitlement to the June 20, 1985 Priority Date Mylan maintains that Daiichi is not entitled to priority based on the filing of the ’712 application because1 it did not sufficiently describe levofloxacin. Specifically, Mylan contends that Example 6, which describes Process A and is also present in the specification of the ’407 patent, does not produce levofloxacin as claimed in the ’407 patent. (DX 1 at cols. 8-4; Mitch-er Tr. at 627-28.) Not only is the content of Example 6 identical in the ’712 and ’407 patents, but the MIC, or potency, data for levofloxacin and ofloxacin remained constant from the ’712 application to the ’407 application. (Mitscher Tr. at 911— 12.) For an inventor to claim priority based on the filing date of a foreign patent application, the earlier application, also known as the “parent,” must reasonably convey to one of skill in the art that the inventor possessed the later-claimed subject matter at the time the parent application was filed. Tronzo, 156 F.3d at 1158. Because it fulfills the written description requirement typically embodied by the specification, the disclosure must describe the claimed invention with all its limitations. Id. (observing that the written description requirement is a statutory mandate under 35 U.S.C. § 112). Whether a parent application sufficiently describes the invention is a highly fact-specific inquiry. See Vas-Cath Inc. v. Mahurkar, 935 F.2d 1555, 1562 (Fed.Cir.1991); Ralston Purina Co. v. Far-Mar-Co, Inc., 772 F.2d 1570,, 1575 (Fed.Cir.1985). Although “a disclosure in a parent application that merely renders the later-claimed invention obvious is not sufficient to meet the written description requirement,” Tronzo, 156 F.3d at 1158, a claim may be broader than the specific embodiment disclosed in the parent application’s specification. Ralston Purina, 772 F.2d at 1572. (citation omitted). Mylan does not assert that Example 6 failed to reasonably convey to a skilled artisan that the inventor possessed the compound yielded in that example. Instead, Mylan argues that Example 6 did not disclose a compound pure enough to be considered levofloxacin as claimed in the ’407 patent and embodied in examples yielding levofloxacin of greater optical purity. Mylan’s real argument appears to be that the foreign patent must be limited to whatever embodiment of the invention the inventor was capable of reducing to practice at the time the foreign patent was filed. That, hovyever, is not the law. The error in Mylan’s position is illustrated by examining the facts of Tronzo, the case on which Mylan principally relies. In Tronzo, the Federal Circuit denied a foreign patentee the priority date from its parent application. 156 F.3d at 1159. The case involved technology related to artificial hip sockets that included cup implants adapted for insertion into the hip bone. Id. at 1156. The parent application had described only conical cups, while the later United States patent application claimed hemispherical cups. Id. at 1159. The foreign patent holder maintained, however, that the foreign patent application had described a sufficient number of cups to ge-nerieally claim them. Id. The Federal Circuit found this assertion unsupportable because the foreign application had emphasized the conical shape as “extremely important,” and only discussed other cups to distinguish the conical ones as superior to the prior art. Id. Thus, the foreign application contained no reference to the later-claimed hemispherical cups and specifically limited the invention to conical cups. The facts in the case at bar differ significantly from those in Tronzo. There, the court denied the patent applicant the benefit of the foreign priority date because the new application claimed a different invention from what had been claimed in the parent application. Here, the difference between Example 6 and the later-developed, more highly pure embodiments of levofloxacin in the ’407 patent, is a difference of degree, not kind. It is clear from examining the claim and specification in the ’712 patent application that the inventors isolated the levorotatory enantiomer, and that they did so to gain the benefit of the isomer’s relatively superior properties and inform skilled artisans how to practice the claimed invention. It was not a claim to Process A itself, but to its product. Although Mylan suggests otherwise, the fact that additional processes for producing the claimed invention were included in the ’407 patent does not negate Daiichi’s claim to the invention as of the filing of the ’712 application. Mylan further argues that Example 6 is not within the scope of the ’407 patent because it does not produce levofloxacin with optical purity of “approximately ninety-five percent or better” based on the enantiomeric excess method of calculating optical purity. According to Mylan, “approximately ninety-five percent” is the purity floor established in the summary judgment testimony of Daiichi’s expert, Dr. Klibanov. (Klibanov Summ. J. at 44.) There are two different methods for measuring an enantiomer’s optical purity, the percentage method that Dr. Klibanov employed in obtaining the 95 percent figure, and the enantiomeric excess method employed by Mylan’s expert, Dr. Mitscher. Both values represent the ratio of the majority enantiomer to the minority enan-tiomer. In fact, to convert from the value under the percentage method to the value under the enantiomeric excess method, one need only perform an additional subtraction to exclude the presence of the dextro-rotatory enantiomer. (Mitscher Tr. at 902-03.) Although Mitscher maintains that the enantiomeric excess method is the only valid method for calculating for optical purity, he does not deny the mathematical relationship between that method and the percentage method. For purposes of deciding the prior invention issue the Court need not determine which method utilized by the experts is superior. What matters is that Drs. Klibanov and Mitscher employed different methods that yielded different results. Klibanov arrived at his optical purity estimation by calculating the optical purity of the least pure example in the ’407 patent (Example 6) using the percentage method. Because claims are typically read to cover the examples, he concluded that the optical purity of the least pure example was the lowest covered by the ’407 patent. Using the enantiomeric excess formula for determining optical purity, Mitscher concluded that the minimum optical purity of levofloxacin in the ’407 patent is 87.3% Therefore, he argued that Example 6 fails to meet the minimum purity level proffered by Klibanov, i.e., “approximately ninety-five percent or better.” Mitscher’s conclusion, however, is plainly erroneous because it compares Klibanov’s baseline purity estimation from Example 6, calculated by the percentage method, with a purity estimation of the same example calculated by the enantiomeric excess method. To make a proper comparison, both purity estimations must be calculated by the same formula. When converted accordingly (using either method), the purity estimations of both experts are equal. Thus, Mitcher’s opinion is illogical because it implies that, depending on the formula used, Example 6 can be less optically pure than the least optically pure example in the patent, which is also Example 6. Mitscher also does not fully embrace Klibanov’s assertion that the claims cover the examples. Although Klibanov and Mitscher agree that one of ordinary skill would normally look to the examples as a source of the optical purity limitation, Mitscher testified that the examples do not provide an optical purity limitation in the case at bar because of the language in the patent that the examples are not intended to limit the claims. (Klibanov Summ. J. at 43-44; Mitscher Tr. at 729-30, 914.) On that theory, however, Mitscher’s attempt to isolate Example 6 to determine optical purity is unavailing because that example is situated no differently than the other examples. Therefore, Mylan fails to prove by clear and convincing evidence that Daii-chi’s parent application failed to sufficiently describe levofloxacin. As such, the June 20,1985 priority date stands. ii. Alleged Prior Invention By Mitscher and Chu Even if Daiichi’s priority date rested solely on its ’499 application, filed oh October 11, 1985, Mylan still bore the burden at trial of proving conception by another inventor prior to that date. Conception requires that the inventor have “a specific, settled idea, a particular solution to the problem at hand, not just a general goal or research plan he hopes to pursue.” Burroughs Wellcome Co. v. Barr Labs., Inc., 40 F.3d 1223, 1228 (Fed.Cir.1994). It is only complete when the inventor has a reasonable expectation that he or she will produce the claimed invention. Hitzeman v. Rutter, 243 F.3d 1345, 1358 (Fed.Cir.2001). Further, its contours must be “so clearly defined in the inventor’s mind that only ordinary skill would be necessary to reduce the invention to practice, without extensive research or experimentation.” Burroughs Wellcome, 40 F.3d at 1228. • Conception is a question of law based on underlying factual findings. See Gambro Lundia AB v. Baxter Healthcare Corp., 110 F.3d 1573, 1576 (Fed.Cir.1997). It must be proven by clear and convincing evidence. See id. To satisfy this burden, inventors cannot prove conception through their own testimony alone. Trovan, Ltd. v. Sokymat SA, Irori, 299 F.3d 1292, 1302 (Fed.Cir.2002) (citing Price v. Symsek, 988 F.2d 1187, 1194 (Fed.Cir.1993)). Corroboration is required regardless of the inventor’s level of interest in the litigation. Finnigan Corp. v. Int’l Trade Comm’n, 180 F.3d 1354, 1369 (Fed.Cir.1999). “Whether the inventor’s testimony has been sufficiently corroborated is evaluated under a ‘rule of reason’ analysis.” Trovan, 299 F.3d at 1302. Under this analysis, courts evaluate “all pertinent evidence ... so that a sound determination of the credibility of the [alleged] inventor’s story may be reached.” Id. Factors to be considered in this analysis include: (1) delay between event and trial, (2) interest of witness, (3) contradiction or impeachment, (4) corroboration, (5) witnesses’ familiarity with details of alleged prior structure, (6) improbability of prior use considering state of the art, (7) impact of the invention on the industry, and (8) relationship between witness and alleged prior user. Juicy Whip, Inc. v. Orange Bang, Inc., 292 F.3d 728, 741 (Fed.Cir.2002). Although circumstantial evidence may be used to corroborate a claimed conception, physical records created contemporaneously with the alleged prior invention are preferred. See Trovan, 299 F.3d at 1303. Mylan asserts that its expert, Dr. Mitscher, together with Dr. Chu, a scientist then employed by Abbott Laboratories (“Abbott”), conceived of levofloxacin in September 1985. (Mitscher. Tr. at 755-56.) The evidence in support of this assertion was derived from the file in a PTO interference between Mitscher and scientists from Bayer, interference No. 102,548 (the “Mitscher-Bayer Interference”), and includes most prominently an affidavit executed by Mitscher on January 31, 1992 expressly for that proceeding. (DX 16.) Mitscher testified that he and Chu conceived of levofloxacin during a meeting between the two on or about September 18,1985. (Mitscher Tr. at 762-63.) Pages from Mitscher’s calendar reflecting that he was at Abbott during those days are the sole evidence of this meeting. (Mitscher Tr. at 764-65) No third party was present at the meeting, which Mitscher conceded was preliminary in nature. (Mitscher Tr. at 764, 934.) Attached to Mitscher’s 1992 affidavit was a document entitled “Confidential Memorandum of Invention,” authored by Chu and received at Abbott’s internal patent office on September 27, 1985. (DX 16.84; Mitscher Tr. at 765-66.) The Confidential Memorandum reports that Mitscher and Chu invented a novel method of producing the racemate and enantiom-ers of ofloxacin. (DX 16.) Although Mitscher testified that he and Chu had sketched a rough schematic during their September meeting that he believed Chu would have attached to the Confidential Memorandum, there is no documentary description of this alleged schematic attached to the existing copy of the Confidential Memorandum or in any other record. (Mitscher Tr. at 936-39.) Nor did Mitscher or any other witness testify from personal knowledge that there was ever any schematic attached to the Confidential Memorandum. Mitscher played no role in drafting the Confidential Memorandum, although he stated that he later viewed it in conjunction with the Mitscher-Bayer interference, and he has never seen it with any attachment. (Mitscher Tr. at 936-37). The oldest surviving written record containing any portion of the schematic that Mitscher claims he and Chu conceived in September 1985 is contained in the first entry of the laboratory notebook of Dr. Padam Sharma, a scientist working under Mitscher, made on December 25, 1985. (Mitscher Tr. at 937-38.) That entry, however, does not contain the entire schematic that Mitscher claims he and Chu devised. (Mitscher Tr. 938.) The schematic is not described fully until entries in Sharma’s notebook that are dated sometime after December 1985. (Id.) Indeed, there is no corroborated evidence at all that Mitscher, or anyone working with him or under his direction, ever made levofloxacin. The compound reflected in Sharma’s notebook in February 1986, PO-7, was a difluoro precursor, not levofloxacin, and none of the existing spectra analyzing it remains. (Mitscher Tr. at 955, 968-69.) Although Mitscher claims to have successfully made levofloxacin in March 1986, he admits that he never measured its optical rotation, nor does he have any existing data supporting this claim. (Mitscher Tr. at 952-56.) On April 25, 1986, Mitscher filed patent application number 4,777,253 (the “ ’253 patent”), which purported to claim a process for making levofloxacin, but that application did not provide any physical data or optical rotation information. (DX 28.) Mitscher testified that the schematic in his affidavit from the Mitscher-Bayer interference is a reproduction of the invention as it existed in September 1985. He further stated that one of ordinary skill could have reduced the invention to practice based on the schematic “using routine laboratory operations.” (Mitscher Tr. at 1225-26.) Thus, were the Court to credit Mitscher’s testimony in its totality, and find that testimony alone sufficient to satisfy Mylan’s burden, Mylan may have established conception. The Court, however, cannot give full credit to Mitscher’s testimony. As a rival inventor and an expert witness for Mylan, he clearly has some interest in the outcome of this litigation. More significantly, the Court has good reason to question the clarity of his memory in light of the fact that his testimony focused on a single meeting that took place eighteen (18) years before he testified. Thus, the rule that corroboration is required is especially forceful in the case at bar. At trial, Mylan offered the following as evidence of corroboration: Mitscher’s calendar pages, Chu’s Confidential Memorandum, Sharma’s laboratory notebook, and affidavits of Mitscher, Chu, and Sharma from the Mitscher-Bayer interference. Mylan’s corroboration is lacking, however, because none of those documents is sufficient, either viewed alone or together, to corroborate invention in September 1985. The documents created in September 1985 do not provide sufficient detail for the Court to determine tha