Tanabe Seiyaku Co., Ltd. and Marion Merrell Dow, Inc. (collectively "Tanabe") appeal from a final determination of the United States International Trade Commission ("Commission"), in In re Certain Diltiazem Hydrochloride & Diltiazem Preparations, No. 337-TA-349. Tanabe petitioned the Commission for a determination that the importation and sale of diltiazem by certain named respondents was a violation of United States law because the product was produced by a process that infringed a patent owned by Tanabe. The Commission denied the petition. Tanabe appealed to this court. We affirm the Commission's decision.

Section 1337 prohibits "[t]he importation into the United States ... of articles that ... are made, produced, processed, or mined under, or by means of, a process covered by the claims of a valid and enforceable United States patent." 19 U.S.C. § 1337(a) (1994)1. Tanabe in its complaint to the Commission alleged that respondents by the importation, sale for importation, and sale within the United States of diltiazem hydrochloride and diltiazem preparations, produced by a process infringing claim 1 of United States Patent No. 4,438,035 (the " '035 patent"), were in violation of § 1337. The complaint sought an investigation by the Commission and, based on the investigation, issuance of a permanent exclusion order and permanent cease and desist orders. The complaint named ten respondents; this appeal concerns the Commission's determination as it applies to four of the respondents, Orion Corporation Fermion, Interchem Corporation, Copley Pharmaceuticals, Inc., and Rhone-Poulenc Rorer, Inc. (collectively "Fermion").

The Commission referred the matter to an administrative law judge ("ALJ") for an evidentiary hearing and initial determination. The proceedings were suspended during reexamination proceedings of the patent, requested by certain of the respondents. The suspension was lifted after the United States Patent and Trademark Office ("PTO") confirmed the patentability of all claims of the '035 patent, including claim 1, the only claim at issue.

The ALJ, following extensive pre-trial proceedings and a two-week hearing involving all parties, in an exhaustive 334-page opinion concluded that claim 1 was not infringed by any of the respondents, was invalid because it was obvious in view of the prior art, and was unenforceable due to inequitable conduct in its procurement.2 On certification to the Commission, the Commission affirmed the ALJ's disposition of the matter, albeit on narrower grounds than those recited by the ALJ.

The Commission limited its affirmance to the ALJ's interpretation of claim 1 and his consequent conclusion that the '035 patent was not infringed by any of the respondents. The Commission did not find it necessary to reach the issues of validity and enforceability, and took no position on those issues. Further, the Commission, by not reviewing them, caused the ALJ's findings in his initial determination regarding importation and the existence of a domestic industry to become a part of the Commission's final determination.

The '035 patent is a method or process patent. It discloses and claims a chemical process for preparing diltiazem hydrochloride, a pharmaceutical product used to treat various cardiovascular diseases, sold in dosage form by Tanabe under the name CARDIZEM TM. There was an earlier, now expired, patent on the product itself, diltiazem hydrochloride, United States Patent No. 3,562,257 (the " '257 patent"), also assigned to Tanabe. This product patent and its prosecution play a significant part in the Commission's interpretation of claim 1 of the '035 patent, as shall be explained below.

That claim, claim 1 of the '035 patent, reads:

1. A method of preparing a benzothiazepine derivative of the formula:

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

In interpreting claim 1, and in assessing infringement under the doctrine of equivalents, the Commission3 considered the claim language, the patent specification, the prosecution history, and certain extrinsic evidence. The Commission noted that Tanabe filed a Statement of Art with the PTO dated January 6, 1983. The statement called the examiner's attention to the '257 patent (the earlier product patent) and one other reference. In distinguishing the claimed invention of the '035 patent from the '257 patent, Tanabe stated:

In contrast, Applicants' invention is the condensation of the acylated form of reference compound II (our II) without prior conversion to the alkali metal salt thereof but rather in the presence of potassium hydroxide in acetone or potassium carbonate in acetone, lower alkyl acetate, water-acetone, or water-lower alkyl acetate.

The Commission noted that in this statement, as well as in claim 1, Tanabe chose to describe its invention in the form of exact base-solvent combinations, rather than in terms of categories of bases and solvents, which Tanabe had done in its '257 patent, and in other patents.

The Commission also noted that Tanabe distinguished the invention claimed in the '035 patent over the prior art '257 patent on the basis that the precise base-solvent combinations claimed by Tanabe provide high yields under safe and economical conditions. Tanabe represented to the PTO: "[I]t is clear that the reference process yields are in the range of 65 to 70%.... Applicants' invention, on the other hand, gives yields which are no less than 87%. It is therefore clear that applicants' invention is patentable over the prior art."

The Commission noted that the '035 patent does not mention the use of any organic solvent other than acetone and lower alkyl acetates. The Commission found it "particularly noteworthy that although the '035 patent discloses a subclass of 'lower alkyl acetates,' it did not disclose a class or subclass of lower alkyl ketones. Rather, the disclosure of ketones was limited to a single ketone, i.e., acetone."

The Commission contrasted the language of claim 1 and the disclosure of only two solvents with Tanabe's disclosure of a range of solvents in the '257 patent. In that patent, Tanabe disclosed that the N-alkylation reaction is carried out in "a solvent (e.g. dioxane, toluene, xylene, dimethylsulfoxide)." The Commission concluded that one of ordinary skill in the art reading the '035 patent and file history "would conclude that the only solvents taught by the '035 patent as suitable for the claimed N-alkylation reaction were acetone and lower alkyl acetates, which in some cases may be mixed with water."

The Commission also considered statements made by Tanabe to foreign patent offices in prosecuting '035 counterpart applications in Finland and Israel, and before the European Patent Office ("EPO"). The '035 counterpart applications were initially rejected by all three of those patent offices, all citing United States Patent No. 3,075,967 to Krapcho (" '967 patent''). In response to those rejections, Tanabe argued that the invention was patentable over the '967 patent because Tanabe's five specific base-solvent combinations gave unexpectedly better results than other combinations of bases and solvents. Tanabe submitted a Comparative Test Report to show the examiners that the five specific base-solvent combinations were better than other base-solvent combinations. Tanabe argued to each of the three patent offices that its invention was not obvious, stating:

Judging from the facts (i) that [the '967 patent] teaches neither the use of potassium carbonate as the base nor the use of specific base-solvent combinations to be employed in the method of the present invention; (ii) that, when the condensation reaction was carried out by the use of sodium hydroxide or sodium carbonate as the base, the yield of the product was less than 10%; and (iii) that, even if potassium hydroxide or potassium carbonate was used as the base, the yield of the product was less than 30% in the case where dioxane, toluene or methanol was used, it is believed that the above mentioned advantages of the present invention have never been taught or suggested by [the '967 patent]. Thus [use of] the specific base-solvent combinations of the present invention is not obvious.

(emphasis added).

The Commission concluded that, based on claim language and the statements made to foreign patent offices as part of the prosecution history, Tanabe showed that it "intended to exclude all bases and solvents other than as particularly claimed, including those that might generally be thought of as equivalent, because the inventors believed that only through the unique base/solvent combinations stated could their requirements to produce diltiazem in high yield be realized."

In evaluating whether Fermion's process infringes under the doctrine of equivalents, the Commission noted that the burden was on Tanabe to show that the substitution of butanone for acetone was an insubstantial change. The Commission observed that acetone is chemically related to butanone and that a person skilled in the art might normally be expected to try to substitute butanone for acetone. However, it determined that a person skilled in the art would notice the specificity and exclusivity of claim 1 regarding the use of acetone in the N-alkylation reaction, and would conclude that other ketones, such as butanone, were not included because they did not work. It also noted that simply substituting butanone for acetone in the '035 patent examples would not generally work, unless conditions were carefully controlled, and that the '035 patent does not teach how to optimize critical process variables. The Commission concluded that:

In view of Tanabe's decision to use restrictive language in the claims of the '035 patent, specifying only acetone and no other keytone [sic, ketone] solvent, and given the admissions made to the PTO about the specificity of the claimed invention, and the further admissions Tanabe made to the EPO and other foreign patent offices, it has not been shown that Fermion's use of MEK is equivalent to the acetone covered by claim 1 of the '035 patent. Therefore, it has not been demonstrated that the accused Fermion process infringes claim 1 of the '035 patent.

On appeal, Tanabe asserts that the Commission erred in its construction of claim 1 and in its consequent determination that Fermion's process did not infringe under the doctrine of equivalents. Tanabe further asserts that the Commission erroneously invoked an estoppel against Tanabe regarding the proper range of equivalents based upon the purported subjective intent of the inventors, limitations not a part of claim 1, and irrelevant statements made to foreign patent offices.

The determination of whether an accused product or process infringes a claim in a patent is universally understood to involve two steps. First, we construe the claim asserted to be infringed to determine its meaning and scope. Second, we compare the properly construed claim to the accused product or process. Markman v. Westview Instruments, Inc., 52 F.3d 967, 976, 34 USPQ2d 1321, 1326 (Fed.Cir.1995) (in banc), aff'd, --- U.S. ----, 116 S. Ct. 1384, 134 L. Ed. 2d 577 (1996). In addressing these two steps, the burden is on the patent owner to establish infringement by a preponderance of the evidence. SmithKline Diagnostics Inc. v. Helena Lab. Corp., 859 F.2d 878, 889, 8 USPQ2d 1468, 1477 (Fed.Cir.1988). The patent owner must show that every limitation of the patent claim asserted is found in the accused process or product, either literally or under the doctrine of equivalents. Id.

Whether a product or process infringes the properly construed claims of a patent, literally or under the doctrine of equivalents, is a question of fact. Hilton Davis Chem. Co. v. Warner-Jenkinson Co., 62 F.3d 1512, 1520, 35 USPQ2d 1641, 1647 (Fed.Cir.1995) (in banc), rev'd on other grounds, --- U.S. ----, 117 S. Ct. 1040, 137 L. Ed. 2d 146 (1997). We review factual determinations of the Commission under the "substantial evidence" standard. 19 U.S.C. § 1337(c) (1994); 5 U.S.C. § 706 (1994). The proper construction of a claim, on the other hand, is solely a matter of law, over which on appeal we exercise complete and independent review. Markman, 52 F.3d at 979, 34 USPQ2d at 1329.

To determine the meaning and scope of a claim, we examine the claim language, the specification, and the prosecution history. Extrinsic evidence, that is, evidence outside the record before the PTO, such as expert testimony about how those skilled in the art would interpret certain language in the claim, may also be considered when appropriate as an inherent part of the process of claim construction and as an aid in arriving at the proper construction of the claim. Id.

To determine whether solvents other than those recited in claim 1 could be used in the claimed N-alkylation reaction without fear of infringement, a competitor would look to the claim language, the specification, and the patent's prosecution history. The Commission correctly noted that, in the specification and claims, Tanabe chose to define its invention in terms of specific base-solvent combinations, rather than in terms of categories of bases and solvents. A person skilled in the art would know that Tanabe could have used the term "lower alkyl ketone" to describe a class of ketone solvents including butanone and acetone. Tanabe used a similar term to describe a class of acetate solvents, claiming the use of a "lower alkyl acetate" solvent, either by itself or with water, rather than claiming only a specific acetate solvent.

For ketone solvents, Tanabe took a different approach to describing and claiming its invention, describing and claiming acetone as the only ketone solvent, to be used either by itself or with water. This specific claim language would suggest to a person skilled in the art that ketones other than acetone may not be useful for the N-alkylation reaction. In other words, the claim language itself, and the specification, suggest that substituting butanone for acetone is not an insubstantial change. Although we do not speculate on Tanabe's reasons for disclosing and claiming the use of only one ketone solvent, we note that:

the doctrine of equivalents is not available for the attainment in court of a scope of protection which encompasses subject matter deliberately removed from examination by the PTO during prosecution through narrow claiming.... [I]t is impermissible to erase under the doctrine of equivalents "meaningful limitations of the claim on which the public is entitled to rely in avoiding infringement."

Genentech, Inc. v. Wellcome Found., Ltd., 29 F.3d 1555, 1568 n. 41, 31 USPQ2d 1161, 1171 n. 41 (Fed.Cir.1994) (citation omitted).

In considering the prosecution history of the '035 patent, the Commission correctly noted that in a Statement of the Art filed by Tanabe with the PTO, Tanabe defined its invention in terms of the exact base-solvent combinations recited in claim 1. Tanabe also stated that its process resulted in "yields which are no less than 87%." Consequently, a review of the prosecution history by a competitor would reinforce the suggestion in the claim language and specification that using other ketone solvents, such as butanone, is not an insubstantial change from using acetone. Moreover, the prosecution history suggests that other ketone solvents may result in lower yields than the claimed solvents.

The sharply restricted nature of the claims has much to do with the scope we accord to the doctrine of equivalents. Tanabe contends that the Commission erred in limiting the scope of equivalents based upon the inventors' purported intent and in relying on the inventors' pre-application experiments. In support, Tanabe cites the following statement in a footnote in the initial decision:

Tanabe's decision not to disclose the use of other ketones might have been based on experimental failures, such as Tanabe's failed TZP N-alkylation experiment with potassium hydroxide as the base and toluene as the solvent. Testimony at the hearing by the inventors might have elucidated this subject further.

Tanabe also refers the court to other statements in the initial decision, including: "Tanabe's [1981] experiments with MEK either resulted in no product or impure product," and a statement in a footnote that "[t]his failure on Tanabe's part may explain why MEK was not claimed in the '035 patent." Tanabe then cites Atlas Powder Co. v. E.I. du Pont De Nemours & Co., 750 F.2d 1569, 224 U.S.P.Q. (BNA) 409 (Fed.Cir.1984), and argues that the Commission erred in relying on the inventors' pre-application experiments to determine that butanone is not interchangeable with acetone. Tanabe asserts that the Commission improperly relied on Tanabe's confidential laboratory notebooks in concluding that experiments not part of the prosecution history create an "experimental estoppel" that limits the application of the doctrine of equivalents in this case.

We do not read the Commission's opinion to establish any estoppel based on pre-application experiments performed by Tanabe. On the contrary, the Commission properly considered Tanabe's pre-application experiments in assessing whether there is a substantial difference between the claimed and accused processes. Tanabe's unsuccessful experiments with butanone indicate that the inventors did not consider butanone to be interchangeable with acetone for use in the claimed N-alkylation process.

The Commission also considered the statements made by Tanabe in prosecuting foreign counterparts to the '035 patent. As earlier explained, in response to rejections for obviousness Tanabe argued that its process was patentable over the prior art because its five specific base-solvent combinations gave unexpectedly better results than other combinations of bases and solvents. Tanabe submitted a Comparative Test Report to demonstrate the effectiveness of its base-solvent combinations. Tanabe argued before the EPO, and the Finland and Israel patent offices, that use of "the specific base-solvent combinations of the present invention is not obvious."

Tanabe argues that the Commission erred in considering foreign prosecution history, improperly finding a "foreign prosecution estoppel" to limit the application of the doctrine of equivalents. We do not read the Commission's decision to establish any estoppel related to the prosecution of foreign counterparts to the '035 patents. In evaluating infringement under the doctrine of equivalents, "representation[s] to foreign patent offices should be considered ... when [they] comprise relevant evidence." Caterpillar Tractor Co. v. Berco, S.p.A., 714 F.2d 1110, 1116, 219 U.S.P.Q. (BNA) 185, 188 (Fed.Cir.1983).

In the present case, the representations made to foreign patent offices are relevant to determine whether a person skilled in the art would consider butanone or other ketones to be interchangeable with acetone in Tanabe's claimed N-alkylation reaction. Because Tanabe represented that its "specific base-solvent combinations" distinguish its process from the prior art, Tanabe's statements to foreign patent offices suggest to a person skilled in the art that other solvents, including butanone, may not be interchangeable with the claimed solvents.

In determining that butanone is not interchangeable with acetone in the claimed N-alkylation reaction, the Commission reviewed experiments performed by Fermion and by Tanabe's expert, Professor Baldwin. The Commission found that Professor Baldwin's tests demonstrate that substituting butanone for the claimed solvents resulted in lower yields or productivity than that achieved in the examples in the patent. As for Fermion's tests, the Commission found that Fermion duplicated examples disclosed in the patent, substituting butanone for the disclosed solvents, and found that butanone generally produced substantially worse results.

There was an exception. The Commission found that when Fermion substituted butanone for acetone in Example 2, the reaction produced a slightly better yield with butanone. Tanabe argues that these results are "dispositive," and indicate that Fermion's process infringes under the doctrine of equivalents. Example 2 teaches conditions for performing the N-alkylation reaction using potassium carbonate as the base and acetone and water as the solvent system. Although when Fermion substituted butanone for acetone in this example, it achieved similar results to those reported in the patent for Example 2, the Commission also found that Fermion was unable to duplicate Example 2 using butanone in larger-scale pilot plant tests on two occasions.

To determine how to adjust reaction conditions to achieve consistent results in its N-alkylation process, Fermion conducted numerous experiments between 1983 and 1986. By performing these experiments, Fermion learned to adjust critical process conditions to achieve consistent high yields for the N-alkylation reaction. The extensive experimentation performed by Fermion, which involved a variety of solvents, bases, and reaction conditions, suggests that Fermion's activities are better described as "designing around" the '035 patent rather than "copying." Moreover, these experiments support the conclusion that butanone is not readily interchangeable with acetone in the claimed process.

Although the preceding review of the claim and its proper construction is more than sufficient to affirm the Commission's conclusion, the record shows that the parties also presented evidence on whether Fermion's process satisfies the function-way-result test for infringement under the doctrine of equivalents. See Hilton Davis, 62 F.3d at 1517-18, 35 USPQ2d at 1645. The Commission found that the claimed process operates by way of a "surface solvent phase," and that Tanabe offered no evidence that a surface solvent phase was present in the Fermion process.

The Commission's decision that Fermion's process does not infringe the '035 patent under the doctrine of equivalents reflects a proper construction of the patent claim, and to the extent it is dependent on factual determinations, is supported by substantial evidence. We have considered Tanabe's other arguments on appeal and find them to be unpersuasive.

AFFIRMED.