Union Carbide Corporation, Plaintiff-appellee, v. the Dow Chemical Company, Defendant-appellant, 682 F.2d 1136 (5th Cir. 1982)

Annotate this Case
US Court of Appeals for the Fifth Circuit - 682 F.2d 1136 (5th Cir. 1982) Aug. 9, 1982

Fulbright & Jaworski, Jefferson D. Giller, Guy L. McClung, III, Houston, Tex., Sidney Joseph Walker, Clute, Tex., Lloyd S. Jowanovitz, Midland, Mich., for defendant-appellant.

Leydig, Voit, Osann, Mayer & Holt, Gerald Rose, C. Frederick Leydig, Chicago, Ill., Baker & Botts, Garrett R. Tucker, Jr., Houston, Tex., for plaintiff-appellee.

Appeal from the United States District Court for the Southern District of Texas.

Before CLARK, Chief Judge, GEE and REAVLEY, Circuit Judges.

CLARK, Chief Judge:

The Union Carbide Corporation (Carbide) sued the Dow Chemical Company (Dow) for infringement of Carbide's process patent No. 3,242,219 for making the chemical bisphenol-A. The trial court refused to adopt Dow's narrow construction of the '219 patent's claims. It also rejected Dow's arguments concerning the patent's obviousness and Carbide's misrepresentations to the Patent Office. Accordingly, it found that the '219 patent is valid, and that it has been infringed by Dow.

On appeal, Dow presses the identical arguments it urged below, couching them in terms of this court's duty to decide almost every issue as a matter of law. We find that the issues to be resolved depend most heavily on the factual record developed below. The record amply supports the factual findings and legal conclusions of the trial court. We affirm.

The '219 patent teaches a process for producing a chemical called bisphenol-A. This chemical is an essential ingredient in several plastics, including epoxy resins and Lexan polycarbonate "glass." The application for the patent was filed on December 31, 1957, by Carbide's employees, Alfred G. Farnham, Francis N. Apel, and Howard L. Bender. Each of the eleven claims in the application were rejected by the patent examiner, and then by the patent office's Board of Appeals. On appeal, however, the Court of Customs and Patent Appeals (CCPA) reversed the Patent Office's determination, and allowed three of the claims.1  Application of Farnham, 52 C.C.P.A. 1118, 342 F.2d 455 (1965).

The claims allowed by the CCPA teach that bisphenol-A is produced by reacting a member of a class of chemicals known as "ketones" with a member of a class of chemicals known as "phenols" in the presence of a catalyst. Water is a by-product of this reaction. The specified catalyst consists of acidized cation exchange resins, also called ion exchange resins. These resins resemble tiny beads. Each bead is an inert resin to which acid groups are then chemically attached. The acid sites thus formed on the resins, rather than the resins themselves, act as the catalyst.

The issue of infringement centers almost entirely on the proper construction of the claims as they pertain to "drying" the resin catalyst. Claim 1 states, the reactants must be contacted with a resin catalyst "which has been treated so as to be substantially free of water." Similarly, Claim 2 states that the reactants are contacted with a resin catalyst "which is substantially free of water."

The removal of moisture from the resin is central to the patent's operation and validity. See Farnham, 342 F.2d at 458. The disagreement here, however, concerns how and when this "drying" occurs as taught by the '219 patent process as compared to the manner in which the drying is accomplished in Dow's process.

Dow argues that if the trial court had given a properly narrow construction to that which is claimed by the '219 patent's "drying" feature, it would not have found infringement. The construction Dow would impress on the patent would limit the words "(a catalyst) which has been treated so as to be substantially free of water" to apply only to those processes where the catalyst has been predried to less than 3% moisture prior to its initial contact with the reactants. Dow contends that its process neither predries the catalyst nor reduces its moisture to 3% before contacting it with the reactants.

To arrive at this particular interpretation, Dow urges us to look not only at the literal wording of the claims, but also at the prosecution history of the patent. Graham v. John Deere Co., 383 U.S. 1, 33, 86 S. Ct. 684, 701-702, 15 L. Ed. 2d 545 (1966). It is undisputed that of all the claims sought in the application only those three with the limitation that the catalyst be substantially free of water were accepted by the CCPA, and that moisture reduction was critical to validity. The CCPA was told that resins having a 50% water content were unsuitable for producing bisphenol-A, and was given various examples of how the resins could be predried. These examples covered both types of processes-batch and continuous.2  Dow asserts, however, that when the '219 application was filed, Carbide had actually run only batch experiments and had not yet discovered that predrying was "unnecessary" in a continuous process. Dow claims the '219 patent does not teach that the initial flow of the reactants will dry the catalyst and let the reaction commence. It contends that Carbide made this discovery prior to the appeal to the CCPA, but that Carbide failed to inform the CCPA that predrying was "unnecessary" in a continuous process. To the contrary, Dow says various statements in Carbide's brief to the CCPA were to the effect that the resin had to be dried to 3% water content before it came in contact with the reactants. Dow refers to the following extract from the brief Carbide filed with the CCPA:

Consider the question of Sussman's coal exchanger (a resin catalyst). Its water content, even in the air dried condition of 15-30% stops the reaction dead-indeed it prevents it from even starting. Example 5 of ('219's) specification shows that at only 3% water, the reaction gave no conversion of acetone .... Claims 3 and 7 cover this feature, (the removal of water from the catalyst) but all the references are completely devoid of any such pretreatment requirements or teachings.

This passage was relied upon in the decision of the CCPA. Farnham, 342 F.2d at 458.

From this history, Dow would derive two limitations to be placed on the patented process-1) that it applies only where the resin is pretreated, and 2) that the pretreatment must reduce the resin's moisture to 3% prior to contact with the reactants.3 

The first of Dow's limitations is correct. In reversing the Board of Appeals, the CCPA explicitly stated "appellants appear to have discovered that the desired reaction required pretreatment of the ion exchange resin catalyst." 342 F.2d at 458. Where Dow falls short is in the factual basis for its theory of non-infringement, namely, that Dow does not pretreat the resin in its continuous process.

Dow's own detail of its continuous process describes the beginning catalyst as having an initial water content of 60%.4  In the first step, a mixture which is 90% acetone and 10% water is flowed in a reverse direction through the glass tubes containing the catalyst. The acetone absorbs some of the water and leaves the resin catalyst with as much as a 27% water content. However, this process also leaves a quantity of unwanted nitrogen in the reactor. Removal of the nitrogen is accomplished by a second reverse flow of the phenol and ketone reactants themselves through the catalyst tubes which forces the nitrogen out of the reactor. When this happens, the catalyst is dried even further. The next step is to start a continuous flow of the reactants forward through the catalyst resins, which produces bisphenol-A. The desired reaction begins whenever the catalyst is sufficiently dry. Because of this, the reaction sometimes commences during the second reverse flow when nitrogen is being expelled.

Dow's position is that neither the initial reverse flow of he acetone-water misture nor the second reverse flow of reactants to expel nitrogen constitutes a pretreatment. The trial court correctly rejected this contention.5  We affirm its finding of literal infringement of a pretreatment process.

Dow also makes the assertion that it merely practices what is described in rejected claim five. Rejected claim five teaches the step of continuously contacting phenol and acetone with the resin catalyst, but does not mention predrying the catalyst. Dow's own description of its process belies this assertion as it does the first. Both factually and legally, the various steps Dow admits it performs in its basic procedure prior to commencing a continuous reactant flow constitute pretreatment.

On a few occasions in 1977 Dow "started up" by flowing only the reactants themselves through the catalyst in order to dry it. Dow argues strenuously that this continuous process is precisely what rejected claim five teaches. Dow asserts that by finding this process infringed the '219, the district court extended the patent's scope to a process which is in the public domain by virtue of the CCPA's rejection of claim five. This is untenable. Claim five was rejected for failing to disclose that the reaction won't start in a continuous reaction until the catalyst is dry, as is disclosed by claim two. Thus, what is left in the public domain by the rejection of claim five is a process where the catalyst is not dried. Dow concedes its catalyst is dried. In so doing it has not practiced claim five and "skirted" the protection of the patent. Kolene Corp. v. Motor City Metal Treating, Inc., 440 F.2d 77, 82 (6th Cir.), cert. denied, 404 U.S. 886, 92 S. Ct. 203, 30 L. Ed. 2d 169 (1971). Rather, it has appropriated the patentee's idea as taught in claim two. Id. See also Garrett Corp. v. United States, 190 Ct. Cl. 858, 422 F.2d 874, 881-82, cert. denied, 400 U.S. 951, 91 S. Ct. 242, 27 L. Ed. 2d 257 (1970). Additionally, we do not read the claims or specification as suggesting by way of limitation that the drying cannot be accomplished with the reactants themselves. See Ziegler v. Phillips Petroleum Co., 483 F.2d 858, 869 (5th Cir.), cert. denied, 414 U.S. 1079, 94 S. Ct. 597, 38 L. Ed. 2d 485 (1973). A patent is not to be limited to the preferred embodiment shown in the specification. Id. This is particularly so here, where the heart of the invention is the need for drying rather than the means by which it is done.

The second claim limitation which Dow urges-that the resin be dried to 3% moisture before contact with the reactants-is similarly strained. Dow asserts that at the moment of first contact with the reactants its catalyst contains up to 27% water. Presumably, we are to find noninfringement because the catalyst has not been lowered to 3% water at the time of the initial reactant contact. But this presupposes first, that the claims should be read to incorporate a 3% requirement, and second, that drying must be accomplished entirely before the reactants are flowed in. We find no merit in either of these suppositions.

As mentioned above, Dow relies primarily on a statement in Carbide's brief to the CCPA, which says, "example 5 of the appellants' specifications shows that at only 3% water, the reaction gave no conversion of acetone." The trial court rejected this contention because the referenced example makes clear that the 3% figure refers to the amount of water in the total reaction mixture, not just in the catalyst. Dow's expert witness conceded on cross-examination that this construction was proper. The trial court concluded that the 3% figure was correct and fair when it was related to water in the reaction mixture. We find no error in these factual findings. Nor is there error in the trial court's corresponding legal construction of the claims.6 

Because the claims do not contain a 3% water limitation for the catalyst, Dow's other supposition necessarily falls, namely that there must be predrying of the catalyst to less than 3% water prior to the initial contact with the reactants. As the trial court found, there is absolutely nothing in the literal words of the claim which indicates that all of the drying must be accomplished before the initial contact with the reactants, and nothing in the specifications or file wrapper. Rather, the patent merely requires whatever amount of drying is needed to permit the reaction to commence.

Dow seeks to engraft narrow interpretations on a patent, the validity of which is based on what the district court properly termed a generic concept-that the bisphenol-A reaction sill not start until the catalyst is substantially dry. Dow's own description of its process fits squarely within the scope of that patented concept: "During the time that the phenol and acetone mixture (was) flowing upwardly through the reactor to remove the nitrogen each of those reactants (was) drying the catalyst as each is a drying agent. When the catalyst became sufficiently dry, the reaction to produce bisphenol-A started regardless of whether the reactant stream was flowing upwardly or downwardly at the time." Dow's explanation that it does not begin to produce bisphenol-A until the catalyst is "sufficiently dry," in our view is consistent with the basic teaching of the '219 patent.

Dow's other noninfringement argument is that it uses a different catalyst from that specified in the patent. Dow claims that it reacts cation exchange resins with dimethylthiazolidine (DMT), and that it owns a patent for this different type of catalyst. Dow argues that the DMT catalyst has a different reaction mechanism and yields a substantially greater rate of reactant conversion. Dow says that, in any event, Carbide's catalyst would not make a commercially acceptable product if used in Dow's process.

The district court found, and Dow does not seriously dispute, that DMT is a promoter of the basic cation exchange resin called for in the patent. We find no error in this factual finding. We agree that Dow's acquisition of the DMT patent does not alter the conclusion that it continues to practice Carbide's patented bisphenol-A process.7  "(T)he grant of a patent on an improvement of a patented article does not excuse infringement of the dominant patent." Ziegler v. Phillips Petroleum Co., 483 F.2d 858, 879 (5th Cir.), cert. denied, 414 U.S. 1079, 94 S. Ct. 597, 38 L. Ed. 2d 485 (1973). "(A)n unlicensed improver is an infringer." Id.

Dow contends the patent is invalid for obviousness, see 35 U.S.C. § 103, and that the precedential effect this court would ordinarily attach to a decision of the CCPA is weakened here by Carbide's failure to present all the pertinent prior art to that court. See, e.g., Ingersoll-Rand Co. v. Brunner & Lay, Inc., 474 F.2d 491, 496 (5th Cir.), cert. denied, 414 U.S. 865, 94 S. Ct. 125, 38 L. Ed. 2d 117 (1973).

Dow first attacks as incomplete and erroneous the trial court's appreciation of the prior art presented to the CCPA. Dow maintains the trial court failed to point out that certain references disclosed the use of anhydrous (dry) catalysts, other than cation exchange resins, in bisphenol-A reactions, including continuous processes. Conversely, Dow says the trial court neglected to acknowledge the use of predried cation exchange resins for reactions other than those that produce bisphenol-A.

Beyond that, Dow cites prior art references which it produced at trial but which Carbide failed to present to the CCPA. These likewise showed the use of dried cation exchange resins for other than bisphenol-A reactions; also, two Dow brouchers state pretreatment in those instances was "usually necessary." Dow argues that had the CCPA had before it this additional state of the art evidence, it would have found the patent obvious to a person having ordinary skill in the art.

Our review of Dow's references satisfies us that the trial court was correct in its determination that the art presented to the CCPA was at least as pertinent as that which Dow produced at this trial. Dow's exposition serves to confirm the basic outline of the scope of the prior art enunciated by the trial court: 1) references showing the bisphenol reaction catalyzed by acid catalysts other than ion exchange resins; 2) references showing the use of cation exchange resins for reactions other than the bisphenol reaction; and 3) miscellaneous references.8  This initial finding by the trial court describes the basic categories into which all of Dow's references fit. Dow's contention that the trial court did not appreciate, or at least point out, the specific ramifications of the individual references, is untenable. The district court made a thorough examination of all the references cited by Dow and, in large part, agreed with Dow as to what those references taught. But the court found, and we agree, that none of them filled one crucial gap-the absolute necessity of drying cation exchange resins in order to produce bisphenol-A. The question, then, with the scope of the prior art and its differences from the claims correctly established, is whether filling that gap would have been obvious to a person having ordinary skill in the art at the time the invention was made. Graham v. John Deere Co., 386 U.S. at 17, 86 S. Ct. at 693-694.

The trial court found nonobviousness based on the fact that this was the first cation exchange resin catalyzed reaction known to science which required that the resin be substantially free of water to produce bisphenol-A. The trial court was entitled to rely on the repeated failures of Dow's scientists to achieve this very invention as evidence of nonobviousness. Graham v. John Deere Co., 383 U.S. at 36, 86 S. Ct. at 703 (citing Marconi Wireless Co. v. United States, 320 U.S. 1, 60, 63 S. Ct. 1393, 1420, 87 L. Ed. 1731 (1943)). Dow seeks to shrug off the experiments of five teams of highly trained scientists as "spasmodic and desultory," and states that there is no evidence that its experimenters had before them the most pertinent art. We cannot say the trial court's rejection of these explanations was clearly erroneous. Dow again asserts that since predrying is unnecessary in continuous reactions, we should disregard Dow's unsuccessful batch experiments. As we have stated previously, the evidence belies the premise for this assertion. Dow failed to rebut the presumption of validity by "significantly more than a mere preponderance" of the evidence, Ludlow Corp. v. Texlite Rubber & Chem. Co., 636 F.2d 1057, 1059 (5th Cir. 1981). Indeed, it did not rebut it at all.

Finally, we note that as soon as Dow's scientists discovered the "trick" to making bisphenol-A with dried cation exchange resins, it sought a patent for that discovery, only to find that Carbide had filed a previous application in a foreign country.

Dow claims the '219 patent is invalid and unenforceable by virtue of the misrepresentations made by Carbide in prosecuting the patent. Dow relies on two statements it characterizes as "false." The first is the reference to 3% water content discussed above, made in the brief to the CCPA. The second is contained in the affidavit of a Carbide vice president vouching for the commercial success of the '219 patent, and particularly its use at Carbide's Marietta, Ohio, plant. Dow contends that the affidavit failed to reveal that Dow was using an esterified acid resin in its Marietta plant.9 

We agree, of course, wholeheartedly with Dow's elucidation of the "high degree of candor and good faith" required of patent applicants. Parker v. Motorola, 524 F.2d 518, 535 (5th Cir. 1975), cert. denied, 425 U.S. 975, 96 S. Ct. 2175, 48 L. Ed. 2d 799 (1976). We also agree, however, with the trial court, that neither statement relied upon by Dow constituted a misrepresentation. As noted above, the 3% water statement, read carefully and in context, is correct and fair.

The statement in the affidavit, "that the catalytic process described in the claims of the aforementioned patent application ... is presently in use at the (Marietta) plant," was likewise found to be correct and fair by the trial court. As with its argument concerning the DMT-promoted catalyst, Dow would have us overlook the fact that the process patent is generic. The esterification of the acid resin is an improvement only. There can be no doubt that the process taught by the '219 patent was in use at the Marietta plant.

The district court correctly determined that patent No. 3,242,219 is valid and was infringed. The judgment appeal from is



See the appendix at the end of this opinion. The claims sought and allowed are set out in full


A batch process, as the name implies, is one in which all the ingredients are at once poured into a container and stirred. In the continuous bisphenol-A process, the catalyst resins are packed into long glass tubes. The reactants flow through the glass tubes and the bisphenol-A and other reaction products are separated in the outflow stream


Dow also asserts that Carbide's failure to inform the CCPA that it is unnecessary to predry in a continuous reaction constituted a misrepresentation sufficient to render the patent invalid and unenforceable. We discuss this contention below


Dow has varied its procedures somewhat over the years. For simplicity, the text outlines the most recent procedure as described in Dow's brief. The various changes in the procedures do not alter our conclusions as to infringement


In its findings of fact the trial court states:

Dow's catalyst is always dried before it is used to produce bisphenol-A. Dow must always reduce the water content of the catalyst from about 50-65 percent to about 4 percent or less before the bisphenol-A reaction commences. Dow has always pretreated the catalyst whether it is with phenol, as the patent teaches, with a mixture of acetone and water, or with the reaction feed mixture. These drying agents are used to remove substantially all of the water so that the reaction can be catalyzed by the cation exchange resin.


We also agree with the district court that the statement in the dissent to the companion case before the CCPA, Application of Apel, 52 C.C.P.A. 1122, 342 F.2d 459, 462 (1965), concerning 3% water content in the catalyst finds no support in the record and should be discounted


Dow's patent states that the DMT catalyst is an improved acid cation exchange resin catalyst


As Carbide points out, one of the additional references, a Dow brochure, "teaches away" from the invention. It states "(i)f water is a product of the reaction, the resin should not be dried as it will soon equilibrate with the water produced." The district court noted this and other references which "taught away."


An esterified acid resin is a type of acidized cation exchange resin which has been reacted with alcohol. This reaction replaces a portion of the acid sites on the resins with esters