This case involves a controversy over a charge of fraud on the Patent Office. The complexity derives not from the law but from the facts. To determine whether there has been fraud on the Patent Office, particularly in a case such as this, where the crux of the alleged fraud consists of an omission of data, a court must attempt to explain the subject matter of the claimed invention. We affirm in part and reverse in part. See page 611 below.

The patent involved in this litigation is No. 3,072,582, the "Frost" patent. General Tire & Rubber Company ("General"), the assignee of this patent, brought this action,1 alleging that the Frost patent had been infringed by the defendants and seeking damages and injunctive relief. The defendants answered with a counterclaim under the antitrust laws and with, among several other defenses, a charge that General Tire committed fraud on the Patent Office. The district court determined that the fraud issue was separable from the other issues in the case.2 After a trial of the fraud issue alone, the district judge held the Frost patent invalid, dismissed General's complaints, and ordered that General be enjoined from enforcing the patent against the defendants and their customers.3 The district judge expressly determined that there was no just cause for delay and directed that final judgment be entered in accordance with F.R.Civ.P. 54(b).4 General timely filed a notice of appeal. We have jurisdiction over this appeal under 28 U.S.C. § 1291 and 28 U.S.C. § 1292(a)(1), see W. L. Gore & Associates, Inc. v. Carlisle Corp., 529 F.2d 614, 617-18 (3d Cir. 1976).

The district judge's exhaustive and learned opinion recognized that the burden of proof in a patent fraud case is that the fraud must be proved clearly, unequivocally and convincingly. In finding that this demanding standard had been met in this case, the district judge stated:

"(S)ince the evidence on which this Court is relying comes primarily from uncontested documents, the defendants have carried their burden of adducing clear and convincing proof."

The fact that the greater part of the evidence in this case consists of uncontested documents is insufficient, in and of itself, to meet the high burden of proof. Since we are in just as good a position as the district court to evaluate documentary evidence,5 we have undertaken an independent and comprehensive review of this evidence. On the basis of this review, we agree with the district court that the burden of establishing the existence of fraud on the Patent Office has been met in this case.

The patent claims a method of making foamed polyether urethanes by using certain inert stable blowing agents6 to produce at least part or all of the bubbles or porosity in the foam, and claims the product, foamed polyether urethane, so made. The chemistry involved in the creation of the polyether urethane foams was known prior to the Frost invention and was not the subject of this patent.

A polyether urethane is made by reacting a nonlinear slightly branched polyether glycol or polyol with a diisocyanate or by reacting a linear ether glycol with a mixture of di- and tri-isocyanates.7 Prior to the invention disclosed in the Frost patent, foaming was customarily produced by using an excess of diisocyanate over that needed to help form the polyether urethane; the excess of diisocyanate would react with water in the mixture to produce carbon dioxide gas (CO 2) and to produce substituted urea by-products.8 The CO 2 produced by this reaction was the blowing agent. It was generated as soon as the ingredients were mixed. According to the Frost patent,

"One of the disadvantages of this uncontrollable release of gas is the formation of a solid crust on the top surface of the polyurethane foam . . . because there is loss of CO 2 from the surface before the poly-urethane is strong enough to hold the CO 2 as bubbles to provide cellular foam. The crust must be cut off before further use with great waste."

Patent No. 3,072,582, column 2, lines 17-24, App. at 989. Moreover, the substituted urea by-products generated by the CO 2 forming reaction were reactive and "enter(ed) to some extent, the crosslinking reactions which produce(d) undesirable stiffness where flexible-type foam products (were) desired." Id. at lines 44-46. Another claimed disadvantage of the prior art was that the diisocyanate was very expensive; "the very substantial excess used to react with water to evolve carbon dioxide . . . represent(ed) an economic limitation." App. at 1001. Yet another problem with the diisocyanate and water method was that the polyether urethane would shrink upon setting.

The Frost blowing agents are easily liquified gases inert to the reactants and the polyether urethane. By carefully selecting the blowing agent, the behavior of the foaming process can be precisely controlled. Water and excess diisocyanate can be partly or completely replaced in the reaction, thus reducing the amount of substituted urea by-product. Using a Frost blowing agent results in less shrinkage, the elimination of the solid crust on the surface of the foam, the creation of a foam with more uniform characteristics, and, because the Freons often used as Frost blowing agents have excellent thermal insulation characteristics, the creation of an excellent insulating material. Polyether urethane foams can be made to be flexible or inflexible, with many gradations in between. A common use of inflexible foams is as insulation material, for instance in refrigerators. Flexible foams find common application as cushioning material, as in furniture.

The claimed fraud on the Patent Office involves the failure of General to report to the patent examiner the results of certain work on flexible polyether urethane foams performed by two of General's employees, Braidich and Van Wagenen, particularly in light of some statements contained in an affidavit presented to the examiner by another General employee, Nicholas. The Nicholas affidavit reported results of experiments carried out by Nicholas with inflexible polyether urethane foams. The significance of these omissions cannot be evaluated outside of the context of the patent prosecution. This is so because a finding of fraud on the Patent Office depends on subsidiary findings that the omission or misrepresentation was material,9 and that the patentee possessed the intent to defraud, or that "willfulness or recklessness indicating a disregard for his duty of frankness . . .." Xerox Corp. v. Dennison Manufacturing Co., 322 F. Supp. 963, 969 (S.D.N.Y.1971).

The story began on October 20, 1955, when the application of Charles B. Frost was received at the Patent Office. That application (Serial No. 541,823) (the "parent application") claimed a method of making polyester and polyether urethane ("polyurethane") foams utilizing as blowing agents easily liquified gases, preferably the halogen substituted lower molecular weight alkane gases,10 inert to the other ingredients used. All of the examples provided in the application involved polyesters, but it was claimed that "(a)ll or part of the polyesters used in the examples may be substituted by polyethers." App. at 1015.

The following remarks appeared in an amendment to the application, filed December 11, 1957:

"The present invention is directed to unusually soft flexible cellular polyurethane elastomers and a process for producing the same.

"When a halogen substituted alkane gas is used as at least part of the blowing agent . . . the resulting material has better cushioning properties than a standard polyurethane foamed material made without the halogen substituted alkane gas."

App. at 1049a. On June 24, 1958, the patent examiner responded by rejecting all the claims, citing German Patent No. 860,109 ("Hochtlen," DX 5). That patent disclosed that blowing could be achieved in polyester urethanes by the use of "vaporizable solvents, which themselves do not react with the diisocyanates . . .." App. at 1255a. In example 3 of the Hochtlen patent, methylene chloride was listed as one of the ingredients, presumably to serve as the "vaporizable solvent" referred to above. The examiner pointed out that methylene chloride was a halogen substituted alkane.

General attempted to distinguish Hochtlen on the grounds that the boiling point of methylene chloride was above the upper limit of the boiling point claimed in Frost. See App. at 1080a. On May 4, 1959, the examiner again rejected the application, citing Hochtlen, and he stated:

"(I)t is not seen wherein the difference in boiling points is critical or patentable. Mere assertions as to superior or unpredictable results are of no weight in the absence of proper corroborating evidence."

App. at 1084a. The examiner also cited Belgium Patent No. 519,547 ("Wingfoot," DX 8A), which disclosed the use of certain volatile solvents, including carbon tetrachloride (a halogen substituted alkane) and isopropyl ether (a dialkyl ether).11 Wingfoot described the role of these solvents as viscosity-decreasing agents, which function by lubricating the molecular structure of the foam. App. at 1260a. The examiner pointed out that even though Wingfoot did not spell out that the solvents claimed would function as blowing agents, "(i)t is obvious that the volatile solvents of Wingfoot would volatilize during the exothermic12 foaming reaction." App. at 1142a.

On April 1, 1959, General filed a continuation in part application ("CIP," Serial No. 803,381), pertaining to polyether urethane foams.13 General then attacked the examiner's reference to Hochtlen on two grounds: (1) Hochtlen disclosed the use of blowing agents for polyesters, while the Frost CIP dealt with blowing agents for polyethers; and (2) the methylene chloride used in the Hochtlen process provided "unsatisfactory" results in polyethers. App. at 1130a.14 It also attacked the Wingfoot reference on two grounds: (1) Wingfoot did not disclose that its "viscosity decreasers" were used as blowing agents; and (2) certain of the Wingfoot agents did not produce satisfactory results in polyether urethanes. App. at 1132a. Moreover, it appears that Wingfoot, as well as Hochtlen, pertained to polyesters. See App. at 1263a.

On September 22, 1960, the examiner responded by asserting that the prior art revealed an essential equivalence between polyesters and polyethers.15 App. at 1141a. He also stated:

"Mere assertions as that the methylene chloride . . . of Hochtlen is unsatisfactory are of no moment unless supported by proper corroborating evidence. The comparative tests in the specification cannot be given any weight since they are not under oath. . . .

". . . It is immaterial that some of the solvents of Wingfoot may not yield the desired results since it would be a mere matter of routine experimentation and determination for one skilled in the art to separate the desirable solvents from the undesirable solvents."

App. at 1142a. The examiner also explicitly rejected the Frost patent's product claims on the grounds that the foams produced by the Frost process were not patentably different from the foams produced by the traditional CO 2 blowing process.

"This is especially true of the flexible products which the claims read on. The blowing agent would be forced out due to the flexing of the foam. Furthermore, the claims read on an open cell16 structure."

App. at 1143a.

On March 20, 1961, General submitted its second amendment to its CIP. Summarizing an interview one of General's patent attorneys had with the examiner, the amendment stated:

"It is believed that the Examiner will recall that substantial agreement was reached that the above-entitled application contained patentable subject matter requiring only a restriction of the claims to the superior blowing agents . . . coupled with a showing (of superiority) as compared to several materials recited in the references cited."

App. at 1150a. General thus expressed the belief that the examiner would be satisfied by a limiting of the claimed blowing agents to exclude those revealed by Hochtlen and Wingfoot, and by a showing that those blowing agents continued to be claimed were superior to the Hochtlen and Wingfoot agents. General also continued to assert that polyesters and polyethers were not interchangeable, but it appears that this contention was made for record purposes only and not with any expectation that the examiner would change his position on this point. See App. at 1151a.

Appended to this amendment was an affidavit from Dr. Louis Nicholas, one of General's scientists. The affidavit contained the results of experiments he carried out, showing "the superiority of the blowing agents of applicant's invention as compared to any suggestion of the prior art . . .." App. at 1156a. All the experiments dealt with inflexible foams blown without the addition of any water to produce CO 2, and this fact was not hidden from the examiner. The Nicholas affidavit was clearly a response to the examiner's invitation to demonstrate the superiority of the Frost agents over all prior art agents in the blowing of polyether urethane foams.

The examiner continued to reject the process claims, but this time solely for technical reasons. App. at 1187a. General responded with yet another amendment on June 8, 1961, which contained the following statement:

"Having complied with all formal matters and shown the superiority of the blowing agents of applicant's invention . . . as compared to any suggestion of the prior art, it is believed that an allowance of the method claims is in order and such action is requested."App. at 1199a. On September 8, 1961, the examiner accepted the Frost method (process) claims, withdrawing the references to Hochtlen and Wingfoot. App. at 1220a.

He continued, however, to reject the product claims on the grounds that the foams produced were not patentably different than the foams produced by the prior art CO 2 method.17 General had attempted to convince the examiner that even the flexible polyether foams produced by the Frost method were patentably superior to flexible foams made by the prior art CO 2 process, even though the greater part of the blowing agent was not retained in a flexible foam. The comparison discussed by General was between Frost blown flexible foams and flexible foams blown solely by the water-CO 2 process.18 An affidavit of George Gmitter, a scientist employed by General as a group leader of a research team, pointed out that the water-CO 2 process generates urea linkages.

"That in contrast a reaction product as above but utilizing either (1) a blowing agent in accordance with the above application, e. g., 'Freon,' as the only blowing or expansion force or (2) partially 'Freon' and partially CO 2 as the expansion force to give the same density cellular or foam product would be characterized by a polymer structure having less urea linkages and less cross-linkings. This would be due to the fact that less CO 2 would be needed and therefore, less HOH (water) would be required, the latter, of course, being the material which leads to urea linkages."

App. at 1216a-17a (emphasis in original). The urea linkages are undesirable, particularly in a flexible foam, because they cause the foam to be stiffer and more harsh. The more such linkages, the less satisfactory will be the flexible foam, and conversely the fewer such linkages, the softer and more yieldable the foam. App. at 1217a.

General appealed the examiner's rejection of the product claims to the Board of Patent Appeals. The examiner continued to rely on the prior art revealing the water-CO 2 blowing process to reject the product claims, and did not cite either Hochtlen or Wingfoot.19 The Board reversed the examiner, stating:

"We find no reference to the use of foaming agents other than carbon dioxide produced by the reaction of water and excess isocyanate and cannot assume from this record that the use of physical foaming agents . . . is known as an equivalent thereof in the foaming of polyether-polyurethane foams."20

App. at 1223a.

II. The Alleged Omissions or Misrepresentations

The crux of the alleged fraud was the failure of General to report to the examiner the results of certain experiments carried out with Hochtlen, Wingfoot and Frost gases as auxiliary blowing agents in the making of flexible foams.

Dr. Donald Van Wagenen carried out some experiments in February and March 1958, using a machine called the Fay Foamer, to produce flexible polyether urethane foams. In a report to General's patent department on July 21, 1958, on Van Wagenen's experiments, the following statement appears:

"D. Van Wagenen stated there is no interest in the idea (the Fay Foamer) at the present time, for the reason that auxiliary blowing can be more successfully and easily accomplished by incorporating into the prepol, low boiling liquids such as methylene chloride or Freon 11.

"These liquids can be readily mixed into the prepol at atmospheric pressure, and good foams are produced by using standard equipment and methods."

App. at 1282a. The term "good foam" is not very helpful, for we cannot tell how the foams made with methylene chloride compared with those made with Freon 11, or those made without any auxiliary blowing agent. Three months later General's outside patent counsel sent a letter to Dr. Gmitter, with a copy to General's patent department, requesting that tests be carried out using Hochtlen and Wingfoot agents to blow flexible polyether urethane foams. The purpose of these experiments was to enable the claims to be limited and to distinguish the prior art agents. App. at 1298-99a. The experiments carried out in response to this letter, rather than the Van Wagenen experiments, should be closely examined.

Dr. Gmitter assigned Dr. Emery Braidich to perform the requested work on flexible foams. App. at 570a. Dr. Gmitter is the same person who assigned Dr. Nicholas to work on inflexible foams. Braidich used the water-CO 2 method as the primary blowing agent in all his experiments. General contends that this fact alone justifies their failure to give the resulting data to the examiner because the CO 2 blowing masked the comparative effects of the other blowing agents tested. Since, however, flexible foams could not, by and large, be successfully blown without the addition of water, General's contention fails. See App. at 572a, 574a-75a.

The results of Dr. Braidich's experiments were reported to General's patent counsel in the form of tables, with no descriptive language. App. at 1307a. There was nothing in this report indicating whether there was shrinkage or how good the cell structure was. This information apparently is very important, App. at 606a; no explanation of why such descriptive information was not included in the Braidich report has been given. In analyzing the Braidich data, we have relied on the testimony of Dr. Kurt C. Frisch, an expert who testified for the plaintiff. According to his testimony, a major purpose of using an auxiliary blowing agent in the making of a flexible foam is to produce a softer foam; i. e., one with fewer urea linkages. He testified that a softer foam would be indicated by lower numbers in two columns of the tables prepared by Braidich per cent. rebound and 25% compression deflection. App. at 607a. Dr. Frisch pointed out that in the third table presented in the Braidich report, Freon 11 (a Frost agent) scored significantly better than both methylene chloride (a Hochtlen agent) and CO 2 alone. App. at 607a-08a; see App. at 1308a. The Braidich report contained two other tables, each representing different formulations of the polyether urethane. The data in the first table confirmed the superiority of foams produced by the Frost method over those produced by the prior art methods, but the results in table 2 were substantially different. The following table summarizes Braidich's second table:

   Auxiliary     Type of      %        25%
 Blowing Agent    Agent    Rebound  Comp. Def.
---------------  --------  -------  ----------
None              ---        29     7 3/4
Pentane          Frost       34     7
Hexane           Frost       30     6 3/4
Freon 11         Frost       32     7 1/2
Carbon
  tetrachloride  Wingfoot    28     7
Acetone          Wingfoot    32     7

As can be seen, the Frost agent Freon 11 scored worse than carbon tetrachloride, a Wingfoot agent, in both categories. The best of the Frost agents in this particular experiment, hexane, was better than carbon tetrachloride in one category but worse in another.

General could have in fact believed that all the examiner was interested in was a showing that the prior art agents were not equivalents for all purposes of the Frost agents. Once they could demonstrate this, coupled with the exclusion of the prior art agents from their claims, they could forget Hochtlen and Wingfoot.21 This would have reflected wishful thinking but wishful thinking is endemic in advocacy. The Nicholas experiments showed that inflexible foams blown with Hochtlen and Wingfoot agents tended to shrink while those blown with Frost agents did not.22 This was probably due to some reaction between the Hochtlen and Wingfoot gases and the polyether urethane. See App. at 598, 671, 1210. General's patent attorneys could reasonably have believed that this phenomenon would likewise occur in flexible foams despite the Braidich data. See App. at 682-83. Braidich, it will be recalled, used some water, thus producing some CO 2 blowing, as well as Hochtlen, Wingfoot and Frost gases. While water is almost always required in order to produce an acceptable flexible foam, the CO 2 produced would mask the effect of the blowing agents concerning which the examiner desired a comparison.23

The advocate's mind is quick to fasten on a distinction (in the instant case, the relative shrinkage of the foams), often forgetting or avoiding inquiry into whether that distinction makes a significant difference. The Braidich data indicated that the reaction with polyether urethane attributable to the Hochtlen and Wingfoot agents was possibly less important with respect to flexible foams than to inflexible foams.24 One reason for this possible result is that flexible foams generally are principally open-celled that is, there are relatively few integral unruptured bubbles in the foam containing some of the blowing agent employed in the formation of the foam while inflexible foams generally are principally closed-celled. Moreover, the fact that much of the blowing of flexible foams is necessarily accomplished by the water-CO 2 process means that any solvent effect of an auxiliary blowing agent will be less pronounced than the solvent effect of that agent when used to provide all the blowing in an inflexible foam simply because it will be masked by the CO 2.25

General's attorneys knew of the Braidich data.26 It could be contended that the Braidich data was inconclusive,27 and that General legitimately did not inform the examiner of it for the sake of avoiding needless confusion. It would have been proper for General to have made this contention before the examiner. The examiner could have accepted that contention, or he could have given General the opportunity to present additional, more conclusive data. In the file wrapper of the parent application, there appears an affidavit of Charles Frost, setting forth the results of experiments comparing the characteristics of flexible polyester28 urethane foams blown with Frost agents and those blown with CO 2. App. at 1057a. It is obvious, then, that General's attorneys knew how to make a meaningful comparison of the flexible foams. Had General revealed the existence of the Braidich data at the same time that it submitted the Nicholas affidavit, we expect that the examiner would have required comparisons such as those discussed in the Frost affidavit. Without such further tests, the examiner could not have known whether the Frost agents were superior with respect to flexible foams as well as with respect to inflexible foams.29

In this context, we find several statements made in the Nicholas affidavit30 to be extremely troubling. The affidavit included a blanket statement that the use of Frost blowing agents produced "markedly superior results to those obtained by utilization of any of the solvent materials disclosed in any of the reference patents31 . . . ." App. at 1158a. General's attorneys knew32 that there existed data casting some doubt on the accuracy of this statement. Also contained in the affidavit was the following paragraph:

"In my opinion, the effects of the different blowing agents as set forth in Table I . . . would be the same in a 'one-shot' type flexible formulation wherein all the foaming ingredients are introduced into a mechanically driven shear type mixing head at one and the same time. (The latter type of foaming operation is conventionally utilized for foaming on a commercial scale but control thereof on a laboratory scale is more difficult.)"

App. at 1163a. Again, General's attorneys should not have allowed this statement to go to the examiner without some indication that data existed casting doubt on the correctness of Nicholas' expectations.

We are also troubled by General's failure to call to the attention of the Board of Patent Appeals during its appeal of the examiner's denial of its product claims33 the existence of the Hochtlen and Wingfoot patents. See DX-3 at 129-188. It will be recalled that the examiner rejected these claims by citing the prior art disclosing the water-CO 2 blowing method. He believed that any properties of inflexible foams blown with Frost agents different from inflexible foams blown by another method would be the obvious result of the differences in the gases retained in the foam. DX-3 at 172-177. Moreover, since most of the blowing agent escapes from flexible foams, he believed there would be no differences in the foams produced. General countered by discussing the inherent structural differences (i. e., the number of urea linkages) between flexible foams made using the water-CO 2 process exclusively and foams made with an auxiliary blowing agent. E. g., DX-3 at 185.34 General should have informed the Board that use of Frost agents was not the only means to cause auxiliary blowing; that Hochtlen and Wingfoot agents could also be used.35

The question in this case is whether the patent monopoly at issue came out of a background free from an intent to deceive or other inequitable conduct. As this court has previously stated:

"The concept of misrepresentation as applied to patent infringement cases admits to no fixed parameters and promulgates no specific dogma. At best it is an abbreviated expression of basic equitable maxims inherent in the law of patents; a recognition that a part of the quid pro quo for the acquisition of a patent monopoly is an insistence that the circumstances surrounding the application be 'free from fraud or other inequitable conduct.' "

Monsanto Co. v. Rohm & Haas Co., 456 F.2d 592, 597-98 (3d Cir. 1972), quoting Precision Instrument Manufacturing Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806, 816, 65 S. Ct. 993, 89 L. Ed. 2005 (1945). Our exhaustive review of the evidence has convinced us that, "under the 'totality of the circumstances,' " Monsanto, supra at 600, General failed to satisfy this requirement.36

Since the refusal of courts to enforce patents in cases such as this is founded on equitable notions, we possess the equitable discretion to choose whether to deny enforcement to the Frost patent in part or in whole. Under the totality of the circumstances, we have concluded that this record requires a denial of enforcement of Patent No. 3,072,582 only to the extent that its claims purport to apply to flexible polyether urethane foams.37

The judgment of the district court will be affirmed in part and reversed in part, and the case will be remanded for further proceedings consistent with this opinion.