Full opinion text
OPINION ROBERT J. WARD, District Judge. This is a patent infringement action brought by plaintiff Leesona Corporation (“Leesona”), a Massachusetts corporation and the owner of the two patents in issue, against defendant Varta Batteries, Inc. (“Varta”), a New York corporation and the American seller of the accused device. The patents are directed to electrochemical cells using specific electrode structures. The Varta product on which plaintiff bases this infringement action is a 1.4-volt zinc-air battery used to power hearing aids. It is known as the “Premium Plus TYP 4600 1.4V” battery, and because of its size and shape (the diameter of a shirt button and about twice as thick) the Varta battery, and others like it, is often called a “button cell.” The matter was tried to the Court sitting without a jury. The Leesona patents in issue currently are under exclusive license to Gould Inc., which in turn has a nonexclusive sublicense with Berec Group, Ltd., covering the same patents. The two patents on which this litigation centers were issued by the United States Patent Office to Glenn V. Elmore and Howard A. Tanner and to Anthony M. Moos. The Elmore and Tanner invention, No. 3,419,900 (“the ’900 patent”), was patented on December 31, 1968. The ’900 patent issued on an application filed December 22, 1966 (Serial No. 609,985), which was a continuation-in-part of an initial application filed March 4, 1960 (Serial No. 12,758), and a continuation application filed December 23, 1963 (Serial No. 332,812). The Moos invention, No. 3,276,909 (“the ’909 patent”), was patented on October 4, 1966, pursuant to an application filed April 6, 1961 (Serial No. 101,057). In its answer, Varta, denying infringement, claims as affirmative defenses that, on the various grounds addressed further in this opinion, the pertinent claims of the ’900 and ’909 patents are invalid and unenforceable. Varta also asserts four counterclaims against Leesona. The first counterclaim seeks a declaratory judgment pursuant to 28 U.S.C. § 2201 that the patents are invalid. The third counterclaim alleges that Leesona engaged in certain acts of unfair competition in violation of section 43(a) of the Lanham Act, as amended, 15 U.S.C. § 1125(a). In its second and fourth counterclaims, Varta seeks relief for alleged violations of federal and state antitrust laws. The federal antitrust counterclaim alleges violations of sections 1 and 2 of the Sherman Act, as amended, 15 U.S.C. §§ 1-2, and sections 1, 2, 3, 4 and 16 of the Clayton Act, as amended, 15 U.S.C. §§ 12, 13, 14, 15 & 26. The state antitrust counterclaim asserts that plaintiff’s actions constitute a violation of New York’s antimonopoly provisions, N.Y.Gen.Bus.Law art. 22. In a fifth and final counterclaim, Varta alleges acts of unfair competition under New York state law. Prior to trial the Court severed and stayed defendant’s antitrust counterclaims and stayed plaintiff’s claim for damages, to be litigated after the issues of patent validity and infringement were determined. Because of their close relation to Varta’s affirmative defenses, the patent-invalidity and federal and state unfair-competition counterclaims remained. At trial, at the close of defendant’s case, Leesona moved to dismiss Varta’s unfair-competition counterclaims. Conceding that it had not offered sufficient evidence in support of these counterclaims, defendant did not oppose the motion and Varta’s unfair-competition counterclaims were dismissed. Thus, remaining for decision at this time are the questions of patent infringement and patent validity for the ’900 and ’909 patents. For the reasons that follow, the Court finds that the relevant claims of both patents are valid, enforceable, and infringed by the Varta battery. A. BACKGROUND I. Preliminary Injunction Shortly after filing the complaint, plaintiff moved for a preliminary injunction to enjoin defendant from selling its allegedly infringing product. At oral argument on this application, held on January 9, 1980, the Court determined that the record before it was not sufficient to enable a decision on the preliminary injunction motion and that an evidentiary hearing was essential. See Dopp v. Franklin National Bank, 461 F.2d 873, 879 (2d Cir. 1972) (evidentiary hearing essential to resolve disputed facts on application for preliminary injunction). However, because the parties were not prepared to proceed directly to such a hearing, with the parties’ consent the Court directed that pursuant to Rule 65(a)(2), Fed.R.Civ.P., the trial of the infringement action on the merits — which has now been held — be advanced and consolidated with the evidentiary hearing on the motion for a preliminary injunction. Accordingly, the decision addresses the question of final injunctive relief, and the motion for a preliminary injunction is denied as moot. 2. Court Expert At the same time it ordered that the preliminary injunction application be consolidated with the full trial on the merits, the Court further directed that a court expert be appointed pursuant to Rule 706, Fed.R.Evid., and requested the parties to submit nominations. The name of one individual, Elton J. Cairns, Ph.D., appeared on both parties’ lists of nominees. Both sides agreed that Dr. Cairns, a professor of chemical engineering at the University of California, Berkeley, is an expert in the field of electrochemistry. There was no dispute that he is fully qualified to give expert testimony and render an opinion on the technical questions at issue in this litigation. After learning that Dr. Cairns was acceptable to both parties as a court expert, the Court communicated with him by telephone. Before he would agree to serve, Dr. Cairns requested an opportunity to review some background materials on the ease and the technical questions presented. The Court then mailed him a package of documents that included a summary of each side’s position (prepared by Leesona and Varta independently and submitted to the Court), a copy of each of the patents in issue and copies of other relevant patents, and certain other analytical information and literature references. After reviewing these materials Dr. Cairns agreed to serve, and pursuant to Rule 706(a) the Court entered a consent order designating him as a court-appointed expert. Shortly thereafter, at the Court’s request the parties prepared the technical questions they wished to have Dr. Cairns consider. The Court forwarded these questions to the expert. Dr. Cairns addressed the parties’ questions in a preliminary report dated July 29, 1980, which he submitted to the Court and copies of which the Court in turn furnished to counsel. The court expert was then examined on oral deposition in two three-day sessions, conducted by counsel without direct supervision by the Court. Dr. Cairns did not prepare a final written report subsequent to the July 29th preliminary report, but rather gave his final views, as discussed in detail below, in his oral testimony at trial after he had heard the testimony of the experts called by Leesona and Varta. The Court found Dr. Cairns’ testimony to be of enormous assistance. Not only was the court expert’s evaluation and testimony thorough and careful, but he also explained his views with a patience and clarity that any court trying a scientifically difficult patent case would find invaluable. The parties’ expert witnesses, as a reading of this opinion should reveal, also were quite helpful. Indeed, without their testimony the value of Dr. Cairns’ services, as well as the Court’s understanding of the technical issues, would have been severely diminished. But, in an action such as this, a court expert serves not only as a witness on whose opinion the Court can rely for assistance, but also as both a second set of ears for the court and a teacher who, unaffected by his having been called as a witness by one side or the other, can explain the technical significance of the evidence presented. Naturally a court will rely heavily on and give great weight to the testimony of its own expert, especially as able an expert as Dr. Cairns. Nevertheless, although a scientifically difficult patent case such as this is ideally suited to the appointment of a court expert pursuant to Rule 706, the appointment of such an expert does not remove from the trial court’s shoulders the burden of understanding and becoming fully familiar with the technical questions in the case. The court expert serves to enhance the trial court’s understanding, but it is the court, and not its expert, that decides the case. Thus, a court appointing a Rule 706 expert must be careful to remind counsel, particularly at the time of the court-appointed expert’s deposition, that the case is not to be tried to the court expert. Fortunately for this Court, counsel here well understood the court expert’s proper role, and their cooperation and assistance throughout this litigation enabled the Court to get the most benefit out of its decision to appoint Dr. Cairns. 3. The Parties’ Witnesses at Trial At trial, on its case-in-chief, plaintiff called two witnesses, Galen R. Frysinger, Ph.D., and Jose D. Giner, Ph.D. Both were qualified as expert witnesses in the field of electrochemistry. Defendant called four live witnesses, two independent expert witnesses in the field of electrochemistry, and two employees of Varta’s parent corporation, Varta A.G. Varta’s two independent experts were Karl Kordesch, Ph.D., and Paul Stonehart, Ph.D. The two Varta A.G. employees, Wolfgang Kloss and August Winsel, Ph.D., testified on the Varta manufacturing process for defendant’s metal-air battery and on the research and development that went into the production of the Varta battery. Dr. Winsel also was qualified as an expert in electrochemistry. In addition, Varta introduced the oral deposition testimony of four witnesses: Paul P. Johnson, John W. Cretzmeyer, Carl W. Zies Jr., and August J. Hipp. Johnson, now retired, was a high-level executive of Leesona during the period in which the inventions in issue were patented, having become the president of the plaintiff corporation in 1967. Cretzmeyer is a research chemist at Gould, plaintiff’s licensee, and Zies is the general manager of Gould’s Portable Battery Division. Hipp, also employed by Gould, is a patent lawyer. The Court called Dr. Cairns as its expert witness. B. THE PATENTS AND CLAIMS IN SUIT 1. In General The patents that are the subject matter of this case are directed to electrochemical devices used in the production of energy, commonly known as batteries and also referred to as electrochemical cells and fuel cells. The energy is produced by these devices through the chemical process known as the oxidation-reduction reaction. Energy generation results from the current flow between the oxidized reactant, which loses electrons, and the reduced reactant, which in turn gains electrons. The oxidation and reduction reactions occur at sites known as electrodes. In an electrochemical cell, current is produced in an external circuit by the electron flow from the electrode at which oxidation occurs (where electrons are lost) to the electrode at which the reactant is reduced (and where electrons are gained). A liquid or paste-like substance, the electrolyte, conducts current within the cell by carrying to the electrodes the positively and negatively charged ions created by the oxidation and reduction reactions. The fuel cell or battery must, of course, be able to contain the electrolyte. Thus, the electrochemical cell must be constructed in such a way that the electrolyte does not escape or leak out of the electrochemical cell package. The reactants fuel the oxidation and reduction reactions that create the electrochemical cell’s energy, and are consumed in the reactions. In metal-air batteries, for example, such as the Varta battery in this case, the oxidized and reduced reactants are a metal (e. g., zinc) and a gas from the air (e. g., oxygen). The metal-air battery contains the metal fuel in the battery package, but the gas is drawn from the air. Although both the metal and the gas are “consumed” in the electrochemical reactions — that is, they undergo a change in chemical form — the gas drawn from the air at the gas electrode is essentially in infinite supply and thus this electrode is said to be nonconsumable. On the other hand, the metal reactant at the metal electrode is contained in the battery itself and truly is consumed (that is, used up) in the chemical reactions. Electrochemical cells of the type at issue here also contain substances used to enhance or speed up the chemical reactions. These substances are known as catalytic or, more particularly, electrocatalytic materials. On .the issue of infringement in this case, it is the comparison of the Varta electrocatalyst with that disclosed in the Elmore and Tanner patent that presents the Court with the closest and most difficult question. Thus, the kind of electrochemical cell that is the subject matter of this lawsuit contains as its essential components two electrodes at which two fuel reactants are supplied, an electrolyte (and a means to contain it), and an electrocatalyst. In addition, at one or both of the electrodes externally supplied gas is the fuel reactant, and the electrode is nonconsumable. 2. Elmore and Tanner No. 3,419,900 In this action, plaintiff contends that the Varta metal-air battery infringes claim 4 of the ’900 patent. In claim 4 Elmore and Tanner describe an electrochemical cell for the direct generation of electrical energy consisting of two electrodes, an electrolyte in the space between them, and an electrocatalyst. The claim specifies that at least one of the electrodes be a nonconsumable gas electrode that consists of a “catalytic layer.” This electrocatalytic layer is described as a “substantially uniform” mixture of “metal containing electrocatalyst particles and a fluorocarbon polymer.” Claim 4 further specifies that the catalytic layer be exposed to the cell’s electrolyte and be sufficiently porous to permit the diffusion of gases into it. The electrocatalyst particles in the catalytic layer of the described electrode serve to catalyze the reactions in the cell. The catalytic layer is exposed on one side to the electrolyte and on the other to the reactant gas. The layer is exposed to the electrolyte so that the electrocatalyst particles can be “wetted” or “flooded” by the electrolyte. It is exposed to the gas so the gas reactant can pass into the porous catalytic layer and to the reactive sites, at the electrolyte-catalytic layer interface. Assuming validity over prior art, there is no dispute that the inventive feature of the Elmore and Tanner electrode structure is the use of a fluorocarbon polymer mixed with the electrocatalyst particles. The use of the fluorocarbon polymer, a hydrophobic substance, in the mixture that constitutes the Elmore and Tanner catalytic layer, or active mass, creates continuous pathways in the active mass through which the gas reactant can diffuse to pass to the reactive sites. This enables the electrochemical reactions to occur throughout the entire volume of the porous active mass, rather than at the surface of the catalytic layer as would occur if the hydrophobic polymer were not part of the catalytic mixture. As Dr. Cairns described it, the use of the hydrophobic fluorocarbon polymer allows the three phases of the reaction — the solid electrocatalyst particles, the liquid electrolyte, and the reactant gas — to coexist and interact at reaction sites throughout the catalytic layer. An electrochemical cell in which the oxidation and reduction reactions occur through the volume of the catalytic layer, such as the ’900 patent describes, is more efficient than a cell in which the reactions can take place only at the surface. The preferable fluorocarbon polymer disclosed in the ’900 patent is polytetrafluoroethylene (“PTFE”), familiarly known under the Du Pont tradename “Teflon.” PTFE is the fluorocarbon polymer used in the active masses of the gas electrodes both in the Varta battery and in the metal-air battery manufactured by Gould, plaintiff’s licensee. The mixing of the PTFE and the catalytic material to produce the electrode described in the ’900 patent results in a bonding of catalyst particles by the PTFE, and the electrode frequently is referred to as PTFE-bonded or Teflon-bonded. Thus, the PTFE in the Elmore and Tanner electrode serves both as a wetting-control agent and as a binding agent. The catalytic layer described in claim 4 consists of “a substantially uniform admixture of metal containing electrocatalyst particles and a fluorocarbon polymer.” The claim does not specify a minimum or maximum volume or weight proportion for the amount of metal in the “admixture.” But there is no question that the language of claim 4 does call for some metal in the catalytic layer. The important and difficult question, as discussed in detail below, part D(2) infra, is how much, if any, metal claim 4 should be read to require and whether any metal must be contained in the electrode for claim 4 of the ’900 patent to be infringed. The ’900 patent also discloses the use of “porous carbon blocks” in the gas electrode. As Elmore and Tanner describe this use of carbon, by example, the PTFE-metal mixture is impregnated into the pores of the carbon material. In general, it is now well recognized — and was at the time of the Elmore and Tanner patent applications— that carbon can be used in this way as an “extender” or “carrier” of the eleetrocatalytic material. When the metal catalyst is mixed with or impregnated into the carbon extender, the expensive metal catalyst (such as silver or platinum) is diluted and thus more efficiently used, with the result being greater current generation per unit of metal electrocatalyst. Used in this way, carbon does not function as an electrocatalyst. Plaintiff’s licensee Gould uses carbon in the oxygen electrode of its metal-air battery in this manner as a carrier on which the mixture of PTFE and its metal catalyst, manganese, is deposited. Although the ’900 patent unquestionably discloses the use of carbon as an extender or carrier, the question remains whether the use of eleetrochemically enhanced carbon as a catalyst (rather than in an inert form as an extender), either alone or in combination with a small amount of metal catalyst, is covered by the Elmore and Tanner patent. This is a question that the Court addresses in detail in its discussion of infringement below. 3. Moos No. 3,276,909 Plaintiff maintains that the Varta battery infringes claim 9 of the ’909 patent. Claim 9 is an independent claim, and there are no dependent claims directed to it. In general, the Moos patent discloses the use of porous “plastic sheets or films” as an inexpensive means of controlling the gas-electrolyte-electrocatalyst interface at which the current-generating reactions in an electrochemical cell occur. The porous plastic films, which can be constructed of any polymeric material, either hydrophobic or hydrophilic, are used to support the catalyst and any catalyst extender material (such as carbon) in the electrode and are in direct contact with the catalytic material. Fluorinated hydrocarbons are disclosed as among the acceptable polymers for such a film. PTFE, a fluorinated hydrocarbon polymer, is disclosed as an acceptable polymer membrane material. Claim 9 discloses a “fuel cell” with, as in the basic electrochemical cell described above, two electrodes, an electrolyte and an electrocatalyst, and specifically with at least one of the electrodes supported by “a porous hydrophobic polymer matrix.” Claim 9 prescribes as the electrocatalyst “a porous conductive catalytically activating metal layer,” which is to be in direct contact with the polymer matrix on one side. On its other side the catalytic layer is to be in direct contact with the electrolyte, and the side of the polymer matrix not in contact with the catalyst is, of course, to be oriented toward the gas. Thus, in a metal-air battery employing the Moos electrode structure of claim 9, the air or oxygen electrode would consist of a porous metal electrocatalyst backed with (i. e., bonded to) a matrix, or film or membrane, of porous hydrophobic polymer material. The polymer side of this membrane-backed electrode is exposed to and faces the atmosphere, and the electrocatalyst side is exposed to the electrolyte. The electrolyte in turn is of course in contact with the metal electrode of the battery. The particular advantage of Moos’ hydrophobic polymer membrane is in the control of the electrolyte. The membrane disclosed in claim 9 is sufficiently porous to permit gas to diffuse into the electrode but, because of its hydrophobic property, is not permeable to a liquid electrolyte. Thus, the electrolyte can freely wet the electrocatalyst yet not enter the gas “space” on the outside of the electrode. The use of a hydrophobic polymer electrode backing prevents the liquid electrolyte from escaping from the electrochemical cell by leaking through the electrode. As the court expert explained, a PTFE or other polymer backing prevents “the so-called weeping of electrolyte through the structure.” 4. Combined Use of the ’900 and ’909 Inventions Together, claim 4 of the Elmore and Tanner patent and claim 9 of the Moos invention can be used in an electrochemical cell, such as the Gould metal-air battery, containing a PTFE-bonded, PTFE-backed gas electrode. In Dr. Cairns’ view, the ’900 and ’909 patents represent two steps of advancement in electrochemical cell development. The first step, represented by the Elmore and Tanner invention, is the development of the PTFE-electrocatalyst mixture. The second step, described in the Moos patent, is, in the court expert’s opinion, the addition of the PTFE film backing to the PTFE-electrocatalyst mixture. Thus, an electrochemical cell combining the inventive features of the two patents may use PTFE in two separate applications — as the fluorocarbon called for in claim 4 of the ’900 patent and . as the hydrocarbon polymer in claim 9 of the ’909 patent. Whether the Varta battery, which employs PTFE both in its electrocatalyst mixture and as a backing for the oxygen electrode, infringes the ’900 patent or the ’909 patent is a question to which the Court turns below. To place the infringement discussion in its proper context, however, the Court first outlines briefly the 1976 Court of Claims decision that involved the patents in suit. C. PRIOR ADJUDICATION In Leesona Corp. v. United States, 530 F.2d 896, 208 Ct.Cl. 871 (1976) (per curiam), the Court of Claims affirmed and adopted the trial judge’s recommended decision (Cooper, J.), in which the trial court held, inter alia, that claims 4 and 7 of the ’900 patent and claim 9 of the ’909 patent were valid and infringed by the United States Government’s metal-air battery designated as the BB-626 ()/U (“BB-626”). In addition to drafting a formal opinion on the seven patents involved in Leesona Corp. v. United States, the trial judge also prepared forty-two separately numbered paragraphs of detailed findings of fact, the first twenty-three and the last three of which deal either specifically or generally with the ’900 and ’909 patents. 1. Validity On the question of validity of the ’900 and ’909 patents, the prior decision in the Court of Claims is not res judicata as to Varta, since defendant was not a party to that action. Carter-Wallace, Inc. v. Davis-Edwards Pharmacal Corp., 443 F.2d 867, 873 (2d Cir. 1971), cert. denied, 412 U.S. 929, 93 S.Ct. 2753, 37 L.Ed.2d 156 (1973). Similarly, since it was the United States, and not Varta, that lost on the claim of invalidity, no collateral estoppel effect can be accorded to the earlier decision and used against Varta and to the benefit of Leesona. Champion Spark Plug Co. v. Gyromat Corp., 603 F.2d 361, 367 & n.12 (2d Cir. 1979) (dictum), cert. denied, 445 U.S. 916, 100 S.Ct. 1276, 63 L.Ed.2d 600 (1980). Moreover, the decision by the Court of Claims is not controlling on this Court as a matter of stare decisis. Id.; Jamesbury Corp. v. Litton Industrial Products, Inc., 586 F.2d 917, 920 (2d Cir. 1978), cert. denied, 440 U.S. 961, 99 S.Ct. 1503, 59 L.Ed.2d 774 (1979). Nevertheless, the prior decision upholding the patents substantially strengthens the statutory presumption, 35 U.S.C. § 282, that arises out of the Patent Office’s determination of validity, Illinois Tool Works, Inc. v. Foster Grant Co., 547 F.2d 1300, 1303 (7th Cir. 1976), cert. denied, 431 U.S. 929, 97 S.Ct. 2631, 53 L.Ed.2d 243 (1977), at least as to those questions litigated in the prior action. For this reason, the Court directed Varta to concentrate its efforts on the validity question, as defendant has done, to matters defendant believed were not presented to the Court of Claims or not expressly considered in the earlier decision. Of course, since patent validity is a question of law, Philip v. Mayer, Rothkopf Industries, Inc., 635 F.2d 1056, 1061 (2d Cir. 1980), Varta may always argue, as it does here, see part E(1)(a) infra, that the application of the law of the Second Circuit to the factual determinations made by the Court of Claims compels a different result from that reached earlier. 2. Infringement On the question of infringement, with the possible exception of providing guidance to this Court on the construction of the claims of the two patents in issue, the Court sees no special importance to the prior adjudication. Accordingly, Leesona’s burden of proving infringement by a preponderance of the evidence is not lessened by the Court of Claims decision. It is safe to say that the factual issues before the Court of Claims on the question of infringement were considerably easier than those presented here. A quick comparison of the United States’ BB-626 and the Varta button cell (described in more detail below, part D(l)) makes this clear. First, as to claim 4 of the ’900 patent, the BB-626 battery is an electrochemical cell with one nonconsumable gas electrode consisting of a catalytic layer that is a mixture of a fluorocarbon polymer (Teflon) and a metal (platinum) electrocatalyst. The catalytic layer is porous and is exposed to the electrolyte. Although the Varta battery contains all of the above elements unlike the BB-626 it does not rely exclusively, if at all, on a metal electroeatalyst. Since claim 4 calls for “a metal containing electrocatalyst particles,” it can be argued, as defendant argues, that the gas electrode in Varta’s battery falls outside of the language of the claim; the BB-626 unquestionably comes within it. Second, with respect to claim 9 of the Moos patent, the BB-626 uses an electrode backing of a porous hydrocarbon polymer matrix (Teflon) and, as noted above, a metal (platinum) electrocatalyst. The Var-ta cell has the same Teflon backing, but again does not necessarily have a “metal” catalyst. Moreover, the BB-626 is closer to a classic “fuel cell” than the Varta battery, and it is easier to conclude that, as the ’909 patent specifies, the United States’ infringing cell has a “means for providing fuel cell reactants to [the] electrodes.” Inasmuch as the BB-626 and the Varta button cell draw oxygen continuously from the atmosphere to provide fuel to the gas electrode, both cells fall within this patent provision as to this electrode. But, unlike the Varta cell, the BB-626 has a mechanically replaceable metal (zinc) electrode; the Varta battery must be disposed of entirely when its metal reactant (also zinc) is fully consumed. Since the term “fuel cell,” as classically and most narrowly defined, is an electrochemical cell with fuel reactants continuously fed to both electrodes — or, to embellish on Moos’ language, a cell with the “means for [continuously] providing fuel cell reactants to [both] electrodes” — the BB-626, with a mechanically replaceable metal reactant, is more like a fuel cell as narrowly defined (which defendant argues the Moos patent specifies) than the Varta button cell. Thus, even if the Court were to accept the Court of Claims determination of infringement as the law of this case, which it does not, the scope of the Court’s inquiry on the question whether the Varta battery infringes either or both the ’900 and the ’909 patents would not be narrowed. A finding of noninfringement here would not be inconsistent with the Court of Claims decision in Leesona Corp. v. United States, supra. With this in mind and against the backdrop of the foregoing brief comparison of the Varta cell with the BB-626, the Court turns to the question of infringement. D. INFRINGEMENT Before addressing the question whether the Varta metal-air battery infringes either or both of the Leesona patents, it is essential in this technically difficult case that the Court present in some detail its factual determinations concerning the accused device. The following paragraphs describe the essential features of the oxygen electrode of the Varta button-cell hearing-aid battery and address a number of disputed issues with respect to the composition and operation of defendant’s electrode. After describing the electrode of the accused device, the Court turns first to the question of infringement of the ’900 patent and then to the question whether the Varta cell infringes the ’909 patent. 1. The Varta Premium Plus TYP 4600 1.4V Battery There is little, if any, dispute between the parties as to the components of the Varta battery or how the product is manufactured. Varta’s Kloss testified ably and at length on the production scheme for the air electrode of his employer’s battery as well as on the steps followed in the assembly of the entire button cell. The Court also was presented with analytic data on the contents of the Varta oxygen electrode. The real factual questions with respect to the Varta cell center not on what it is (or what is in it) but on how the battery works. a. Elements and Manufacturing Process. First, as to the product itself, the Varta button-cell hearing-aid battery is a metal-air electrochemical cell with a zinc electrode and an oxygen electrode. The oxygen (or gas) electrode contains an electrocatalyst layer bonded to a PTFE layer (or membrane). The Varta electrocatalyst consists of a mixture of PTFE and activated carbon. Both the PTFE, the trademarked product Hostaflon, and the activated carbon, known by the product name “Anthralur KC,” are purchased by Varta from other manufacturers. The PTFE-activated carbon catalyst layer is attached on one side to a silver-plated nickel screen that is used as a current collector and backed on the other side by the PTFE membrane. Varta prepares its electrocatalyst or active material by mixing the activated carbon and the PTFE, in a proportion by weight of approximately seventy percent activated carbon and thirty percent PTFE, in a large high-speed blending machine spinning at about 2500 to 3000 revolutions per minute. The resulting mixture of catalytically active material, to which no metal is added, is then rolled (or “nipped” between two rollers) to form the catalyst layer. The catalyst layer and the silver-plated nickel screen are pressed together by a second rolling operation, after which a third rolling operation bonds the catalyst-collector screen intermediate product to the electrode-backing layer of PTFE. From this ultimate three-layer product, which is produced in continuous strips approximately 100 meters long and ten centimeters wide, the small button-shaped oxygen electrodes are punched. b. Activated Carbon Catalyst. Although Varta does not intentionally add any metal to the catalytically active material in its electrode, the activated carbon that makes up seventy percent of the catalytic layer by weight does contain metal impurities. Moreover, some small amounts of metal in the form of silver may well dissolve from the silver-plated nickel screen into the electrolyte, and migrate to and be deposited on the activated carbon-PTFE mixture that constitutes the Varta catalyst layer. The two chemical analyses of the active material in Varta’s catalytic layer, introduced by plaintiff and described by Leesona’s expert witness, Dr. Giner, reveal the presence of metals such as cadmium, iron, magnesium, manganese, silver, vanadium, and zinc. The presence of zinc— which, at as much as 2500 parts per million parts (“ppm”) (or 0.25 percent), along with iron (at 3000 ppm or 0.30 percent) was observed in the greatest amount — could well be accounted for because, as noted, it is the reactant used in the metal electrode. Similarly, the silver found (at 750 to 1000 ppm, or 0.075 to 0.1 percent) may be attributable to the migration of silver from the collector screen. The other metals found, however, presumably result from the impurities in the activated carbon, Anthralur KC, that defendant purchases to use as its electrocatalytic material. Each of the above seven metals appears in the Elmore and Tanner patent as among the metals listed as catalytic materials that may be used in the electrode structures in the ’900 invention. Nevertheless, even though catalytically active metals listed in the patent are present in small amounts in the Varta electrocatalyst layer, the Court finds that the primary catalyst in defendant’s electrode is activated carbon. Although the presence of these small amounts of metals may have a significant effect on performance, Varta does not rely on the presence of the metals to produce a marketable metal-air battery. Dr. Cairns and Var-ta’s expert, Dr. Kordesch, agreed that it would be unwise for Varta to rely for electrode performance on the presence of metal impurities in the activated carbon or on the migration of silver from the collector screen to the catalytic material. By preparing test electrodes, Varta conducts performance tests on samples from all of the batches of Anthralur KC it purchases. But the Court is not persuaded that the fact that defendant conducts such tests alone demonstrates a reliance for electrochemical performance on the metal impurities in the activated carbon. Indeed, plainly Varta tests the Anthralur KC, among other things, to ascertain the quality of the carbon activation that enables carbon to function as an electrocatalyst. Plaintiff does not appear to dispute that activated carbon serves as the electrocatalyst in defendant’s electrode. Where Leesona and Varta disagree is over the enhancing effects on the carbon electrocatalyst of the small amounts of metal impurities in the Anthralur KC and of any silver that may dissolve from the collector screen into the electrolyte. Dr. Kordesch testified at length on the use of activated carbon as an electrocatalyst. None of the other expert witnesses, including Dr. Cairns, disagreed with Dr. Kordesch’s testimony that activated carbon may effectively be used as an electrocatalyst, at least where the carbon contains naturally occurring metal impurities. Despite Dr. Kordesch’s opinion, as the Court understands it, that, once activated, pure carbon, without metal impurities, can work effectively as an electrocatalyst, the experts agreed that to find and use such pure carbon would be, as Dr. Cairns put it, an “extremely unusual thing and require[ ] extraordinary efforts.” Presumably this would also be economically infeasible for mass production. Carbon activation, in essence, is the process by which the porosity of the carbon material is increased. This results in a manifold increase in the surface area of the carbon material. Dr. Kordesch testified that, for example, in the activation of graphitic carbon by superheated steam, the surface area per unit volume of the carbon is increased about eight fold. The increase in surface area and porosity enables the activated carbon electrocatalyst to function more efficiently by making the volume of the catalyst layer more accessible to the electrolyte and the reactant gas, and by providing increased surface area per unit volume over which the electrochemical reactions can occur. Dr. Kordesch testified that there are a number of processes by which carbon can be “activated” from an inactive form, such as graphite. One such process, the process Dr. Kordesch believes is used to produce Anthralur KC, the activated carbon used by Varta, involves exposing the unactivated carbon material to steam or carbon dioxide and then heating it up to a very high temperature, on the order of 900° C. The carbon material is “eaten away” by the superheated steam or carbon dioxide, thereby increasing the porosity and surface area of the material. Carbon can also be activated with metal salts. Although Dr. Kordesch did not explain the metal-salt activation process, presumably in this process the carbon material is exposed to an aqueous metal-salt solution that, like any corrosive liquid, eats away at the carbon to increase porosity. In a metal-salt process, Dr. Kordesch testified, metal impurities are added to the activated carbon material, while in a steam or carbon dioxide activation process very little in the way of metal impurities is added to the activated material. None of the experts explained to the Court how metal and carbon, or metal-and-carbon, electrocatalysts actually work to catalyze the chemical reactions in an electrochemical cell. The Court has considered this question on the basis of all the evidence received and is of the view that a metal electrocatalyst and a pure activated carbon electrocatalyst (assuming such pure activated carbon can be purchased) function somewhat differently, at least at the microscopic level. As the Court understands it, metal electrocatalysts enhance the chemical reactions directly, by causing the reactions to proceed faster, and thus serve as catalysts in a classic sense. Activated carbon, on the other hand, absent the influence of metal does not affect the chemical reactions directly, but rather affords more surface area over which the reactions can occur and thus enhances electrochemical performance, not by directly speeding up the reactions, but by allowing more reactant activity to occur at a slower rate. At the microscopic level, the Court finds this to be a significant difference, but at a macroscopic level, as discussed in part D(2)(b) infra, the Court sees this difference as far less important. c. PTFE Carbon Coating. Turning to the final aspect of the Varta electrode that requires attention, the Court finds that defendant’s mixing operation in its manufacturing process, in which PTFE and activated carbon are blended to form the Varta cell’s active mass, does not result, as defendant claims, in the coating of carbon with PTFE. Varta maintains that its high-speed mixing operation, in Dr. Kordesch’s words, “chops up” the long PTFE molecules and causes the PTFE to coat the active carbon rather than to become mixed with the catalyst material as the ’900 patent provides. The PTFE begins to vaporize in the Varta process, according to Dr. Kordesch, as a result of the heat generated by the fierce mechanical activity in defendant’s mixing operation. Thus, when the vaporized PTFE condenses, it acts more as a coating to the carbon catalyst particles than as a part of an admixture. Dr. Stonehart testified along the same lines as Dr. Kordesch. In fact, through Dr. Stonehart defendant took the rather unusual approach of introducing expert testimony by asking the expert witness to conduct a scientific experiment in open court. In this experiment Dr. Stonehart compared a sample of the PTFE-activated carbon mixture prepared by Varta at its manufacturing facilities in Germany with a sample mixture, prepared in court, of seventy percent Hostaflon, the PTFE used by Varta, and thirty percent Anthralur KC, defendant’s activated carbon material. Dr. Stone-hart dropped relatively equal amounts of the two sample mixtures into separate glass beakers containing the electrolyte used in the Varta cell. The result observed by the Court was that all of the Varta-prepared mixture floated at or near the surface of the electrolyte, while the mixture prepared in Court separated, part settling and part floating near the surface. In Dr. Stonehart’s opinion, this experiment demonstrated that the Varta electrode is more than a simple mixture. Such a simple mixture, in which no PTFE coating occurs, is demonstrated, in Dr. Stone-hart’s view, by the separation observed for the mixture prepared in court — the hydrophobic PTFE rose to the surface and the activated carbon settled. In contrast, Dr. Stonehart testified that all of the Varta PTFE-activated carbon mixture floated to the surface because the carbon had become coated, or “wetproofed,” by the PTFE. His view is that the coating of carbon by PTFE in the Varta electrode brings defendant’s electrode composition well within pri- or art, which had disclosed an activated carbon catalyst coated with a wetproofing material. Commenting on Dr. Stonehart’s in-court experiment, Dr. Cairns offered the view that the buoyancy observed for the Vartaprepared catalyst material was attributable not to PTFE coating but to an agglomeration of small PTFE particles and larger activated carbon particles. Although on the basis of this experiment he was prepared to recognize that PTFE-carbon mixtures prepared in a high-speed blender might exhibit different properties from such mixtures prepared otherwise, the court expert did not believe that in the Varta process PTFE would vaporize or begin to vaporize and then coat the activated carbon material. In Dr. Cairns’ view, Dr. Stonehart’s experiment merely demonstrated that separation would be likely to occur in a PTFEcarbon mixture that is not very thoroughly mixed. During the overnight break between the trial day on which Dr. Stonehart conducted his in-court experiment and the day on which Dr. Cairns testified, plaintiff conducted an electron microscope analysis of the Varta PTFE-activated carbon catalyst layer. Eight electron microscope “scans,” as they are called, performed at various electron voltage settings and degrees of magnification, were presented to Dr. Cairns in the form of black-and-white photographs. Although the court expert cautioned that he was not an expert in the particular scientific specialty of electronmicroscopy, the Court permitted him to give his interpretation of the scans. In Dr. Cairns’ view, the scans reflect a mechanical mixing of PTFE and activated carbon rather than the PTFE coating of carbon that would result from a chemical bonding of PTFE to the carbon particles or a PTFE vaporization and subsequent condensation on the carbon. Dr. Cairns testified that he thought the Varta mixture of PTFE and activated carbon could be separated, if there were an adequate means of mechanically manipulating the small PTFE and carbon particles. The Court finds that, while the Varta PTFE-activated carbon mixing process may result in some bonding of PTFE and carbon particles, it does not result in a coating of carbon with PTFE. Defendant’s process does not result in the near complete carbon coating described in prior art such as the Berl patent. 2. Infringement of Elmore and Tanner No. 3,419,900 The general test of patent infringement follows a two-step analysis. First, the Court must determine if the accused device reads literally on, or infringes directly, the language of the claims in the patent. Second, if the Court does not find that the patent has been infringed directly, it must then look to see if application of the doctrine of equivalents nevertheless leads to a finding of infringement. See Ziegler v. Phillips Petroleum Co., 483 F.2d 858, 868 (5th Cir.), cert. denied, 414 U.S. 1079, 94 S.Ct. 597, 38 L.Ed.2d 485 (1973). Assuming for its purposes here that the ’900 patent is valid, the Court addresses both steps in the infringement test in the following paragraphs to determine if the Varta metal-air battery infringes the Elmore and Tanner patent. After applying the test to the ’900 patent, the Court undertakes the same two-step analysis for the Moos ’909 invention. a. Direct Infringement. In considering whether a product directly or literally infringes a patent, “resort must be had in the first instance to the words of the claim.” Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 607, 70 S.Ct. 854, 855, 94 L.Ed. 1097 (1950). If the allegedly infringing device falls squarely within the literal language of the claims, a case of direct infringement is established. Id.; Continental Oil Co. v. Cole, 634 F.2d 188, 191 (5th Cir. 1981), petition for cert. filed, 49 U.S.L.W. 3932 (U.S. May 12, 1981) (No. 80-2037). The Court finds that the Varta button cell does not infringe the ’900 patent directly. The Varta battery, as described above, contains all of the elements specified in claim 4 of the ’900 patent, except one. No matter how persistently plaintiff argues to the contrary, the Court simply cannot read the words “metal containing electrocatalyst particles” in claim 4 to cover the Varta cell’s oxygen electrode. On the other hand, although defendant argues otherwise, the Court finds that the Varta gas electrode contains the “admixture” of fluorocarbon polymer (PTFE) and catalyst that claim 4 calls for. Even with a substantial bonding of PTFE and activated carbon particles in the catalytic layer of defendant’s electrode, the composition must still be regarded as a mixture of PTFE and carbon, and not as a coating of carbon particles with PTFE. Claim 4 specifies, inter alia, a catalytic layer in the gas electrode “containing a substantially uniform admixture of metal containing electrocatalyst particles and a fluorocarbon polymer.” Plaintiff argues that so long as any amount of metal is present in the electrocatalyst material of an otherwise identical device, claim 4 is directly infringed. The Court cannot agree. The Elmore and Tanner patent contemplates a metal electrocatalyst. Carbon, in activated or inactivated form, is not a metal. Moreover, the patent nowhere mentions carbon or activated carbon as a potential electrocatalyst. Yet it specifically lists some forty other “catalytic materials which may be used.” Though enhanced perhaps by the presence of small amounts of metal impurities and silver migrating from the collector screen, the Varta electrocatalyst cannot be said to be a metal catalyst. Plaintiff maintains that the words “metal containing” in claim 4 should be read as if there were a hyphen between them, i. e., as if claim 4 called for an “admixture of metal-containing electroeatalyst particles.” Read this way, of course, the electrocatalyst particles would only have to contain metal and not themselves be metal. The Varta cell, then, according to plaintiff, would literally infringe claim 4 because its activated carbon electrocatalyst is concededly “metal-containing.” The Court recognizes that this argument is not without support. Indeed, the abstract of the disclosure in the ’900 patent itself uses “metal-containing” (with hyphen). Nevertheless, construing the language of claim 4 in light of the disclosures of the patent as a whole, the relevant prior art, and the history of the patent application in the Patent Office, see, e. g., Decca Ltd. v. United States, 420 F.2d 1010, 1013 (Ct.Cl.) (per curiam), cert. denied, 400 U.S. 865, 91 S.Ct. 102, 27 L.Ed.2d 104 (1970), the Court finds that claim 4 cannot be read literally to encompass an activated carbon electrocatalyst, even if this catalyst contains some catalytically active metal materials. The “hyphen” discrepancy may very well be the result of an inartful drafting of claim 4. But, even if the Court were to construe the claim as if the words were hyphenated, it would still read the claim to require a metal electrocatalyst. There is no persuasive evidence that Elmore and Tanner considered activated carbon to be a potential catalytic material that could be used in the electrochemical cell they were disclosing. b. Infringement Under the Doctrine of Equivalents. The doctrine of equivalents permits a patent holder to extend the monopoly beyond the precise device patented to equivalent devices. The doctrine is founded on the theory that “if two devices do the same work in substantially the same way, and accomplish substantially the same result, they are the same, even though they differ in name, form or shape.” Graver Tank & Mfg. Co. v. Linde Air Products Co., supra, 339 U.S. at 608, 70 S.Ct. at 856 (quoting Machine Co. v. Murphy, 97 U.S. 120, 125, 24 L.Ed. 935 (1877)). Thus, as the doctrine has come to be stated in its well-recognized form, a device will be found to be infringing if it performs “substantially the same function in substantially the same way to obtain the same result.” Id. (quoting Sanitary Refrigerator Co. v. Winters, 280 U.S. 30, 42, 50 S.Ct. 9, 13, 74 L.Ed. 147 (1929)). See also Continental Oil Co. v. Cole, supra, 634 F.2d at 191, and N. V. Maatschappij Voor Industríele Waarden v. A. O. Smith Corp., 590 F.2d 415, 421 (2d Cir. 1978). The question of infringement by equivalence is one of fact, and a question on which expert testimony is of great assistance. As the Supreme Court instructed in Graver Tank & Mfg. Co. v. Linde Air Products Co., supra, 339 U.S. at 609, 70 S.Ct. at 857: Consideration must be given to the purpose for which an ingredient is used in a patent, the qualities it has when combined with the other ingredients, and the function which it is intended to perform. An important factor is whether persons reasonably skilled in the art would have known of the interchangeability of an ingredient not contained in the patent with one that was. In his preliminary report, Dr. Cairns gave an unconditional “yes” in answer to the question, put to him by plaintiff, whether the Varta electrochemical cell utilizing a non-consumable cathode having a catalytic layer which consists essentially of a mixture of active carbon and fluorocarbon polymer and a silver plated screen perform[s] substantially the same function in substantially the same way and obtain substantially the same result as does the electrochemical cell defined in claim 4 of the ’900 patent. The court expert reaffirmed this opinion at trial. In Dr. Cairns’ view, the Varta catalyst of activated carbon with its metal impurities is the equivalent of the metal catalyst disclosed in claim 4. There is no “important scientific difference, or difference in the art,” Dr. Cairns testified, “between electrode structures . . . and their operation in the case of electrocatalysts which are comprised of metal only and electrocatalysts which are comprised of metal and/or carbon, or, in fact, [of] carbon itself.” Nothing the Court heard at trial and none of the exhibits in evidence persuades it that the court expert’s opinion is unsound. Dr. Cairns testified that, in his view, the essential inventive feature of the ’900 patent is the fluorocarbon-electrocatalyst mixture. The Court recognizes that, at least at a microscopic level, there may well be a distinction between the way a metal catalyst catalyzes the reactions in an electrochemical cell such as the Elmore and Tanner patent discloses and the way an activated carbon electrocatalyst (at least a pure carbon catalyst) operates in this catalysis. But on a larger scale both activated carbon and metal catalysts, once mixed with PTFE, serve to enhance the energy-producing chemical reactions at the gas electrode in the cell. In both cases the PTFE-catalyst mixture enables the chemical reactions to occur throughout the volume of the catalytic layer. Moreover, to the extent that the metal impurities present in the Varta electrocatalyst have a positive effect on electrochemical performance, the difference even at a microscopic level diminishes further. It is enough that the court expert, a person reasonably skilled in the art, sees activated carbon as interchangeable as the catalyst ingredient with the metal disclosed in the ’900 patent. The Court therefore finds that the Varta cell performs substantially the same function (executing the desired electrochemical reactions to generate current) in substantially the same way (by using as its catalytic layer a mixture of a fluorinated hydrocarbon polymer and a catalyst material) to obtain the same result (enabling the electrochemical reactions to occur throughout the volume of the catalytic layer). Defendant argues that Leesona’s failure to investigate activated carbon as an electrocatalyst, coupled with the Elmore and Tanner patent’s failure to disclose carbon as a catalyst, should preclude application of the doctrine of equivalents. The Court sees no reason why such an “inverse doctrine of equivalents,” to use defendant’s term, should be applied in this case. Cf. Harris v. Allen, 15 F. 106 (N.D.Ill.1883) (narrowly construing patent claim but not mentioning doctrine of equivalents or its so-called “inverse”). The equivalents doctrine was developed to “prevent an infringer from stealing the benefit of the invention,” Royal Typewriter Co. v. Remington Rand, Inc., 168 F.2d 691, 692 (2d Cir.), cert. denied, 335 U.S. 825, 69 S.Ct. 50, 93 L.Ed. 379 (1948), merely by changing an ingredient that one reasonably skilled in the art would have realized or readily discovered was interchangeable. It is not necessary to a finding of infringement under the doctrine of equivalents that the Court determine that the patentees, or the patent holder Leesona in this case, had developed the interchangeable ingredient at the time the invention was patented. The Court need only find, as it does here, that persons reasonably skilled in the art — not necessarily the patentee— would have known that such an interchangeable ingredient could be developed. Leesona’s failure to investigate activated carbon as an electrocatalyst may properly be considered in construing whether the patent itself actually claims to cover such a catalytic material. But a patentee or patent holder’s determination not to investigate equivalent ingredients for its own patent ought never to preclude resort to the doctrine of equivalents to prove infringement on the basis of someone else’s use of those equivalents. After all, it will often be the case that a patent holder has little, if not the least, motivation to develop equivalent ingredients for its patent. Indeed, it is the person seeking to avoid the patent who has the greatest incentive to find an interchangeable ingredient that will take its device outside the infringement umbrella of the equivalents doctrine — that is, to find an ingredient that persons reasonably skilled in the art would not have known was interchangeable. Finally with respect to the doctrine of equivalents as it applies to the ’900 patent, the Court notes that defendant does not contend that plaintiff is limited in its resort to this doctrine by the doctrine of file wrapper estoppel. Varta does not contend that the file history of the Elmore and Tanner patent demonstrates that the patentees narrowed their claims by deleting references to carbon as an electrocatalyst. As has already been developed, Elmore and Tanner never disclosed activated carbon as an electrocatalyst material. 3. Infringement of Moos No. 8,276,909 a. Direct Infringement. The Court concludes that the accused Varta device does not directly infringe claim 9 of the ’909 patent. Claim 9 specifies a “means for providing ... reactants to [the] electrodes” and a “catalytically activating metal layer.” Inasmuch as the Varta cell contains neither the means for providing reactants to both electrodes nor a metal catalytic layer, the Court finds that the accused device is not described by the literal language of the claim. In the Court’s view, the Moos patent describes what “purists,” as Dr. Giner described them, would consider to be a classic fuel cell — that is, an electrochemical cell in which the reactants are continuously fed to both electrodes. The drawing in figure 4 of the Moos patent illustrates what the Court understands such a classic fuel cell to be. The Court, however, does not read claim 9 only to cover fuel cells as classically defined, and it is not Moos’ use of the term “fuel cell,” in and of itself, that causes the Court to read claim 9 as it does. Indeed, the Court declines to adopt a narrow, classic definition of “fuel cell.” Rather, it is the provision in claim 9 that requires a means for supplying reactants to both electrodes that compels the Court to construe the language of the claim only to cover an electrochemical cell where both reactants are continuously supplied (or at least, as with the BB-626, where the metal reactant can be mechanically replenished). For the reasons stated earlier in this opinion, the Court finds that the Varta button cell does not have a metal electrocatalyst. The ’909 patent, even more clearly than claim 4 of the ’900 patent, specifies a metal electrocatalyst. Although in Dr. Giner’s opinion the silver that migrates from the Varta collector screen and the metal impurities in defendant’s activated carbon constitute the “metal layer” called for in claim 9, in Dr. Cairns’ view the Varta cell does not contain a metal layer. Dr. Kordesch also testified that “[t]here is no metal layer in the Varta electrode.” Moreover, the ’909 patent nowhere discloses anything other than a metal electrocatalyst. b. Infringement Under the Doctrine of Equivalents. Turning to the question whether the accused device infringes claim 9 of the ’909 patent under the doctrine of equivalents, the Court notes first that in his preliminary report Dr. Cairns also gave an unconditional “yes” to the following question, similar to the one he was asked concerning claim 4 of the ’900 patent, and again put to him by plaintiff: Does the Varta cell utilizing a non-consumable cathode having a catalytic layer which consists essentially of a mixture of active carbon and fluorocarbon polymer and a silver plated screen with this catalyst layer in contact with a fluorocarbon polymer layer, with the fluorocarbon polymer layer being in contact with the reactant gas, perform substantially the same function in substantially the same way and obtain substantially the same result as does the invention of claim 9 of the ’909 patent? At trial Dr. Cairns confirmed that this was still his view. The Court finds that the essence of the Moos invention is the use of a hydrophobic polymer membrane, as a backing to a gas electrode, to prevent liquid electrolyte from leaking from the electrochemical cell, while at the same time permitting gas to diffuse into the electrode. Although it became clear to the Court that Dr. Cairns did not consider the Moos patent to be particularly strong because of the absence of scientific data in the patent demonstrating the performance capabilities of the electrode structures described by the patentee, it is equally clear to the Court that Moos’ contribution of a cell structure with a hydrophobic membrane electrode backing is recognized as significant development in the art. Dr. Cairns said as much in his preliminary report and at trial, and no other witness really disputed this conclusion. The Varta cell concededly employs the hydrophobic membrane backing disclosed by Moos. Defendant’s argument, however, is that nothing else about its electrochemical cell is described by the ’909 patent. In support of this argument, defendant makes three contentions. First, Varta asserts, the ’909 patent is directed to fuel cells. Defendant argues that its battery is not a fuel cell. Second, Moos makes no mention of a PTFE-catalyst mixture in one of the gas electrodes, yet Varta employs such a mixture in its oxygen electrode. Third, claim 9 of the Moos patent speaks of a metal catalyst layer. Varta maintains not only that its catalyst is not a metal layer, but also that it uses no metal catalyst at all. Concerning the first of these cont