dissenting.
I do not join the panel majority in its reversal of the Commission’s finding of literal infringement. No error of fact or law has been shown in the Commission’s determination; indeed the Commission’s findings are scarcely mentioned. Instead, my colleagues have made findings contrary to the detailed and unchallenged text of the patent specifications, and have construed the claims so that they exclude a major part of the invention described in the patents. I must, respectfully, dissent.
The basic principle of determining the scope of patent claims is that the claims provide the legal definition of the invention that is patented, based on the descriptive text and controlled by the prosecution history, with cognizance of the prior art. A patentee can reduce or disclaim claim scope to cover less than what is described in the specification. Here, however, there was no disclaimer of the scope set forth in the patent specifications and claims; there is no prior art to limit the claims in the way selected by the panel majority; and there is no reason to insert an absolute numerical limit of “about 4%” protic material into the claims that do not contain a numerical limit, when the specifications of both patents demonstrate significantly higher percentages. There was no evidence contradicting the evidence of the experts concerning the range of protic material set forth in the specifications’ text and illustrated in the specific examples.
The Commission correctly construed the “controlled amount of protic material” to match the content of the specifications and claims. The Commission’s findings concerning the amount of protic material shown in the specifications are supported by substantial evidence, the statutory standard for review of the agency’s findings, 19 *1142U.S.C. § 1337(c), and its claim construction is in accordance with law. The panel majority has seriously erred in discarding the Commission’s findings and conclusions, for they are not only supportable on the required standard of review, but they also are correct.
The products at issue are N-(l,3-dime-thylbutyl)-N’-phenyl-p-phenylenediamine (6-PPD) and 4-aminodiphenylamine (4-ADPA), from which 6-PPD is derived. The complaint charged that the 4-ADPA and 6-PPD are produced using a process that is covered by claims 30 and 61 of Patent No. 5,117,063 and claims 7 and 11 of Patent No. 5,608,111. The '111 patent issued from a continuation of an application that led to Patent No. 5,453,541 (not in suit here), which was a continuation-in-part to the application that led to the '063 patent. The patents describe and claim conducting the process by the direct reaction of aniline and nitrobenzene in the presence of a solvent, a base and a “controlled amount” of protic material as set forth in step (b) of the overall process. Claim 61 of the '063 patent is representative:
61. A method of producing alkylated p-phenylenediamines comprising the steps of:
a) bringing aniline and nitrobenzene into reactive contact in a suitable solvent system;
b) reacting the aniline and nitroben-zene in a confined zone at a suitable temperature, and in the presence of a suitable base and controlled amount of protic material to produce one or more 4-ADPA intermediates;
c) reducing the 4-ADPA intermediates under conditions which produce 4-ADPA; and
d) reductively alkylating the 4-ADPA of Step c).
The patents explain that the presence of a controlled amount of protic material produces the environmental advantage of avoiding halide waste products, and reduces manufacturing and material costs because it permits the direct reaction of aniline and nitrobenzene, instead of requiring the preparation of intermediate compounds as in the prior art.
The claims in suit do not include numerical limits for the “controlled amount” of protic material, and the specifications define the amount as varying with the solvent, base, base cation, and the like. The '063 patent states (with emphases added):
A “controlled amount” of protic material is an amount up to that which inhibits the reaction of aniline with nitrobenzene, e.g., up to about 4% H20 based on the volume of the reaction mixture when aniline is utilized as the solvent. The upper limit for the amount of protic material present in the reaction varies with the solvent. For example, when DMSO is utilized as the solvent and tetramethylammonium hydroxide is utilized as the base, the upper limit on the amount of protic material present in the reaction is about 8% H20 based on the volume of the reaction mixture. When aniline is utilized as a solvent with the same base, the upper limit is 4% H20 based on the volume of the reaction mixture. In addition, the amount of protic material tolerated will vary with type of base, amount of base, and base cation, used in the various solvent systems. However, it is within the skill of one in the art, utilizing the teachings of the present invention, to determine the specific upper limit of the amount of protic material for a specific solvent, type and amount of base, base cation and the like.
Id., col. 4, lines 48-68.. The “teachings of the present invention” are extensive and *1143detailed. The Commission defined “controlled amount” as “an amount up to that which inhibits the reaction of aniline and nitrobenzene.” This definition is in accord with the patent specifications, but it is rejected by the panel majority, who limit “controlled amount” to the part of the patentee’s statement about using up to 4% water where there is excess aniline and no other solvent, and exclude all of the other variables that affect the amount of water, as demonstrated in the specifications’ text and specific examples. The panel majority promotes the number that is described for one condition, to a limit under all conditions, contrary to the specifications. It is not the judicial role to change the invention. Although the panel majority acknowledges that limits above 4% “could be found in the specification” and were so found by the Commission, the panel majority discards this undisputed fact.
The patents describe the conditions in which 4-ADPA intermediates are produced by the direct reaction of aniline and nitrobenzene. The '063 patent specification includes twelve specific examples, varying the amount and nature of the base, the solvent, the cation, the protic material, and the reaction conditions. Additional examples are in the continuing '111 patent, illustrating further variation in the protic material. In the '063 patent, Example 1 was conducted at room temperature, using tetramethylammonium hydroxide (TMAH) dihydrate1 as the base. Example 2 shows the effect of varying the temperature. Example 3 shows varying the amount of water from 2.2 to 4.7%, and also uses methanol as the protic material. In Example 3, water in amounts of 0, 10, 50 and 100 pH is added to the reaction. The percentage of water in the reaction includes both this added water and the water introduced by the TMAH dihydrate. See '063 patent, col. 9, tbl. 2 note. The other Examples show that the amount of water present in the dihydrate is part of the protic material that is present and is calculated as such. See id. col. 11, lines 14-18 & tbl.6.
Example 4 shows the use of various solvents; Example 5 is directed to various bases; Example 6 is a comparison with a prior art method; Example 7 varies the ratios of the reactants; Example 8 shows varying the total amount of water in steps of 2.3%, 3.5%, 6%, 9.75%, and 14.7%; Example 9 shows varying the amount of the TMAH dihydrate base while maintaining the amount of added protic material constant at 4.7%. Example 10 shows anaerobic conditions at 50°C and contains 10% water 2 from the dihydrate; the Commission *1144describes this example as a preferred embodiment. Example 11 demonstrates the production of the tetramethylammonium ion salt of 4-NDPA and p-NDPA; Example 12 illustrates the conversion of 4-ADPA to produce 6-PPD.
In the '111 patent, additional examples, particularly Examples 13 and 15, again show the presence of significantly more than 4% water. The panel majority declines to consider the additional examples, which show 10.8% water for Example 13 and 13.8% water for Example 15. The Commission had no need to rely on the additional examples because it found the '063 examples sufficient. Nor is the issue one of “new matter,” in using the continuation-in-part to construe claims of an ancestor patent, for the charge is infringement of both or either patent.
The specifications explain the effect of each variable, and state that “the amount of protic material tolerated will vary with type of base, amount of base, and base cation, used n the various solvent systems.” '063 patent, col. 4 lines 61-63. The specifications further explain that “[T]he reactions can be successfully performed by conventional modifications known to those skilled in the art, e.g., by appropriate adjustments in temperature, pressure and the like, by changing to alternative conventional reagents such as other solvents or other bases, by routine modification of reaction conditions, and the like, or other reactions disclosed herein or otherwise conventional, will be applicable to the method of this invention.” Id. col. 6, lines 35-43.
The panel majority, in discussing some of the examples, ignores the dihydrate that is present throughout the examples, and finds that only Example 10 of the '063 patent can be viewed as having more than 4% water. The majority’s complaint that Example 10 “is not even directed toward illustrating the control of the amount of protic material to be used in the reaction,” maj. op. at 13, is inapt, for Example 10 shows 92.8% yield of the desired 4-ADPA intermediates in conditions that include 9-10% of protic material. '063 patent, col. 12, lines 11-13. Thus my colleagues find “clear definitional language [of a 4% limit] set forth in the specification” although that number is in a sentence with the signal “for example,” and ignore the text and examples showing a higher range. These erroneous appellate findings of scientific fact directly contradict the findings of the Commission, made upon extensive and detailed evidence and argument on the content of the specifications, including the variables discussed and exemplified in the '063 and '111 patents. Instead, the panel majority selects the parts of the specifications that show water in the 4% range, and ignores the description and examples that show other amounts of water. The majority’s de novo ruling is contrary to the testimony of experts for both sides, contrary to Commission expertise, and contrary to the rules of claim construction. The Commis*1145sion’s finding warrants appropriate deference. See, e.g., Federal Power Comm’n v. Fla. Power & Light Co., 404 U.S. 453, 466, 92 S.Ct. 637, 30 L.Ed.2d 600 (1972) (“The court may not ... ignore the conclusions of the experts and the Commission and put itself in the absurd position of substituting its judgment for theirs on controverted matters of hydraulic engineering.” (quoting United States ex rel. Chapman v. Fed. Power Comm’n, 191 F.2d 796, 808 (4th Cir.1951), aff'd, 345 U.S. 153, 73 S.Ct. 609, 97 L.Ed. 918 (1953))).
Claims are construed as they would be understood by persons in the field of the invention, for those are the persons by and for whom patents are written. Such persons are charged with the specification, the prosecution history, and general knowledge in the field of the invention. Phillips v. AWH Corp., 415 F.3d 1303, 1316-17 (Fed.Cir.2005) (en banc). It is beyond debate that the entire specification must be consulted in construing the claims. Pfizer, Inc. v. Teva Pharms. USA, Inc., 429 F.3d 1364, 1373 (Fed.Cir.2005) (“It is necessary to consider the specification as a whole, and to read all portions of the wriU ten description, if possible, in a manner that renders the patent internally consistent.”); Slimfold Mfg. Co. v. Kinkead Industries, Inc., 810 F.2d 1113, 1116 (Fed.Cir.1987) (“Claims are not interpreted in a vacuum, but are part of and are read in light of the specification.”).
When the entire specification including the specific examples is consulted, rather than selected snippets, the correct claim scope is apparent from the specifications. Abbott Labs. v. Andrx Pharms., Inc., 473 F.3d 1196, 1210-11 (Fed.Cir.2007) (declining to limit claim term to description following the word “is” in the specification where to do so would exclude disclosed examples); Verizon Servs. Corp. v. Vonage Holdings Corp., 503 F.3d 1295, 1305 (Fed. Cir.2007) (“We normally do not interpret claim terms in a way that excludes disclosed examples in the specification.”); Invitrogen Corp. v. Biocrest Mfg., L.P., 327 F.3d 1364, 1369 (Fed.Cir.2003) (district court’s claim construction erroneously excluded an embodiment described in an example in the specification); Budde v. Harley-Davidson, Inc., 250 F.3d 1369, 1379-80 (Fed.Cir.2001) (the specification should be considered as a whole and read “if possible, in a manner that renders the patent internally consistent”).
The labored justification by the majority of its ruling that “about 4%” is the maximum amount of protic material permitted by the specifications does not suggest how the Commission erred, or explain a claim construction that excludes much of the content of the specifications and is contrary to the expert testimony. For example, expert witness David Crich testified, without dispute, that a person of ordinary skill would recognize that the description of 4% water is directed to reactions with aniline as the solvent at conditions of ambient temperature and pressure. The panel majority inexplicably attributes “no weight” to this expert testimony by arguing that since “controlled amount” has no universal meaning in chemistry, the court can ignore how persons skilled in this field of chemistry would understand the term as applied in this field. As resolved in Phillips, 415 F.3d at 1313, “[t]he inquiry into how a person of ordinary skill in the art understands a claim term provides an objective baseline from which to begin claim interpretation. That starting point is based on the well-settled understanding that inventors are typically persons skilled in the field of the invention and that patents are addressed to and intended to be read by others of skill in the pertinent art.” (Citations omitted.)
*1146The cases on which the majority relies do not support, and indeed contravene the majority’s approach to claim construction. In Durel Corp. v. Osram Sylvania, Inc., 256 F.3d 1298, 1303-04 (Fed.Cir.2001) the court held that the entirety of the specification, including the specific examples, must be considered in determining the scope of the term “oxide coating,” and not a single broader statement. In Multiform Desiccants, Inc. v. Medzam, Ltd., 133 F.3d 1473, 1478 (Fed.Cir.1998) the court stated that the patentee’s lexicography could not be enlarged by resort to extrinsic dictionary definitions, when the intended scope was clearly set forth in the entirety of the specification “without ambiguity or incompleteness,” a sound rule that emphasizes the controlling role of the specification. These cases support the Commission’s holding, not that of the panel majority. Equally inapplicable is the majority’s citation of Rheox, Inc. v. Entact, Inc., 276 F.3d 1319, 1327 (Fed.Cir.2002), for unlike Rheox, here the prosecution history does not require exclusion of preferred embodiments.
The majority also relies on Telemac Cellular Corp. v. Topp Telecom, Inc., 247 F.3d 1316 (Fed.Cir.2001) and North American Container, Inc. v. Plastipak Packaging, Inc., 415 F.3d 1335 (Fed.Cir.2005) as supporting its claim construction. These are cases where the construction of claims is narrower than the breadth of the specification, on specific facts where such narrowing was required, such as by “means-plus-function” construction, or by an express disclaimer during prosecution as in North American Container, 415 F.3d at 1345-46, 1348. In contrast, in the present case the words “controlled amount” do not have an unambiguous meaning that dictates the exclusion of disclosed embodiments, nor is there any suggestion of disclaimer. These decisions do not support the majority’s arbitrary appellate limitation of claim scope, contrary to the content of the specifications. See, e.g., Conoco, Inc. v. Energy & Envtl. Int’l, L.C., 460 F.3d 1349, 1358 (Fed.Cir.2006) (it is generally improper for courts to import a numerical limitation into a patent claim that contains no numerical limit); Modine Mfg. Co. v. U.S. Int’l Trade Comm’n, 75 F.3d 1545, 1551 (Fed.Cir.1996) (“Ordinarily a claim element that is claimed in general descriptive words, when a numerical range appears in the specification and in other claims, is not limited to the numbers in the specification or the other claims.”) The patents’ teachings that the amount of protic material can vary with the reaction conditions, teachings heavily reinforced by actual experimental examples, were readily understood by the Commission to show, as the patentee stated, that the invention can be practiced over a range of conditions. The extensive data in the patent specifications were not challenged. No error has been shown in the Commission’s conclusion that “controlled amount” is not limited to 4%.
Our appellate obligation is to impart consistency, predictability, and guidance to patent claiming, whereby the patent-user community can rely on a technologically correct and legally consistent interpretation of patent claims. The health of innovative technology requires confidence in objective rules of claim construction, and in uniform judicial application of the rules. The panel majority adds inconsistency and unpredictability by arbitrarily limiting the scope of the claimed invention in a way that conflicts with the teachings of the specifications and the knowledge in the field of the inventions. I respectfully dissent.
. TMAH dihydrate contains two molecules of water per molecule of tetramethylammonium hydroxide. The specific Examples in the patents variously add water in liquid form, or as the dihydrate, or both together.
. My colleagues call it a “complex calculation” to calculate the percentage of water in the total volume of reactants, maj. op. n. 7, and decline to consider this evidence; the AU, the Commission, and everyone else in the record and briefs treat this calculation as routine. In response to a question from the bench during oral argument, counsel submitted the following explanation:
Specifically, the question was asked of how one skilled in the art would be able to
calculate the percentage of water in the reaction disclosed in Example 10 of the patents-in-suit.
Example 10 discloses "the reaction of aniline, nitrobenzene and tetramethylam-monium hydroxide dihydrate ('TMAH dih-ydrate') under anaerobic conditions at 50°C.” JA 000269. Col. 11, line 64-Col. 12, line 1. The term “dihydrate” means that there are two water molecules for each molecule of TMAH. JA 000561, lines 13-22. Example 10 discloses that 0.42 moles of TMAH dihydrate are present in the reaction. Because there are two moles of water (dihydrate) for one mole of TMAH, there are .84 moles of water in the reaction. JA 000567, lines 4-20. Because the amount of *1144water and of aniline and nitrobenzene and TMAH are disclosed in Example 10, the percentage of water in the reaction can be calculated using the method described at JA 000568, line 6 through JA 000569, line 21. The method of calculating the amount of water in a reaction where the amount of TMAH dihydrate, aniline and nitrobenzene are known is also found in the prosecution history of the '111 patent, as is discussed in the Final Initial and Recommended Determination at JA 000137-138. Sinorgchem's experts agreed that Example 10 has 9 to 10 percent water, and that it could be calculated. JA 000151.
Letter from counsel Mark G. Davis dated September 14, 2007. The experts were in agreement that Example 10 shows the presence of 9-10% water. It is undisputed that a person of ordinary skill in the field of the invention can readily calculate the amount of water.