Cosden Oil & Chemical Co. v. American Hoechst Corp. ( 1982 )

543 F. Supp. 522 (1982)

COSDEN OIL & CHEMICAL COMPANY, Plaintiff,
v.
AMERICAN HOECHST CORPORATION, Defendant.

Civ. A. No. 76-37.

United States District Court, D. Delaware.

March 15, 1982. Supplemental Opinion June 23, 1982. As Modified June 23, 1982.

Douglas E. Whitney, Morris, Nichols, Arsht & Tunnell, Wilmington, Del., William D. Harris, Jr., Garland P. Andrews, Jerry R. Selinger, Richards, Harris & Medlock, Dallas, Tex., M. N. Cheairs, Monty L. Ross, Dallas, Tex., Joe A. Moss, American Petrofina, Inc., Dallas, Tex., for plaintiff.

E. Norman Veasey, Robert H. Richards, III, Jesse A. Finkelstein, Richards, Layton & Finger, Wilmington, Del., Francis J. Hone, Edward V. Filardi, Brumbaugh, Graves, Donohue & Raymond, New York City, John C. Wyman, Roche, Carens & DiGiacomo, Boston, Mass., for defendant.

OPINION

STAPLETON, District Judge.

These are consolidated actions involving the validity, infringement, ity of United States Patents No. 3,868,434 ("'434 patent") and No. 3,996,311 ("'311 patent"), which issued to Richard C. Westphal ("Westphal") and Paul Heinig ("Heinig"). Cosden Oil and Chemical Company ("Cosden") originally filed this suit asking for a declaratory judgment that the patents in suit are invalid, unenforceable and not infringed. It has since added allegation of fraud on the Patent Office. American Hoechst Corporation ("AHC"), as successor to Foster Grant Co.,^[1] the original assignee from Heinig and Westphal, seeks damages and injunctive relief against Cosden for infringement. Jurisdiction and venue are conferred upon this Court by 28 U.S.C. §§ 1338(a), 2201, 2202 and by 28 U.S.C. §§ 1391(a), 1391(b) and 1400(b). This opinion constitutes the Court's findings of fact and conclusions of law.

I. BACKGROUND FACTS

Styrene is a "monovinyl aromatic" liquid compound; it contains both an aromatic, or cyclic, substituent and an aliphatic substituent called a "vinyl" group having two carbon atoms connected by a double bond. The presence of this aliphatic double bond makes styrene a chemically reactive monomer which can be polymerized by heat or the addition of a catalyst to form homopolystyrene which has a clear or "crystal" appearance. This so-called "crystal polystyrene" can be molded to form clear plastic products having many desirable properties, such as good moldability, an attractive surface gloss, and good dimensional stability. Unfortunately, crystal polystyrene products suffer from relatively poor resistance to fracture under slight pressure. This characteristic detracts from the use of crystal polystyrene in certain applications and makes it wholly unsuitable for many others.

It has long been known that the impact resistance of crystal polystyrene may be improved by adding a small amount of a rubbery material to the styrene either before or after its polymerization. These "rubber toughened" materials are referred to as impact styrene polymer compositions, high impact polystyrene compositions (HIPS), or simply "impact polystyrene," and are currently used to make hundreds of items ranging from TV cabinets to refrigerator door linings.

Concerted efforts were made in the 1940's to commercialize impact styrene polymer compositions for wider end use applications. The early compositions were made either by dissolving a rubbery matter in styrene monomer and polymerizing the resulting mass, or by simple mechanical blending of the polystyrene and the rubbery material. These manufacturing techniques were well known by 1958 and are very similar to those used today.

In the early 1950's, Amos and his co-workers at the Dow Chemical Company ("Dow") added a step to the process by "agitating" or stirring the solution of rubbery material in styrene monomer. Agitation was shown to cure the early defect of large insoluble gels in the end products.^[2] This was the last significant change in the process; the focus of the industry after 1954 was on improving the compositions themselves by the substitution of new rubbers.

The prevailing rubber used by the mid-1950's was a styrene-butadiene type rubber, "SBR", which consisted of about 75% butadiene and 25% styrene.^[3] In 1956, "stereo-specific polymerization catalysts," invented by Karl Ziegler, were proved useful in the preparation of "stereoregulated" or "stereospecific" polybutadiene rubbers. These new catalysts made it possible to alter polybutadiene by directing and changing its microstructure. Among the many possibilities which resulted was the "high cis, low vinyl polybutadiene" rubber which was later used to make compositions encompassed by the claims of the patents in suit. "Vinyl" refers to the presence of 1,2-addition butadiene polymer units (called vinyl groups) in which only the first and second carbon atoms of the monomer form the polymer chain backbone, and the third and fourth carbons are in side chains:

"Cis" refers to the presence of Cis 1,4-polybutadiene polymer units, in which the first and fourth carbon atoms are disposed on the same side of a line joining the second and third carbon atoms:

The presence of these particular molecular configurations in quantities of at least 25% cis 1,4-butadiene, and not more than approximately 10% 1,2-addition butadiene, constitutes a "high cis, low vinyl polybutadiene" for purposes of this Opinion.

High cis, low vinyl polybutadiene first became known to the public in 1956 by reference made in a Belgian patent issued to Goodrich Gulf. In 1957, several other foreign patents issued, and an article was published by Short, Kraus and Thornton of the Phillips Rubber Co. ("the Short article"), extolling the virtues of the new material. By 1957, stereospecific rubbers had been sufficiently studied and evaluated that they were being recommended as a substitute for natural rubber in certain applications.

About this time, an effort was made in the industry to standardize the impact strength specification for various grades of impact polystyrene. Impact strength was the first property selected, evidently because it represents the most significant improvement over crystal polystyrene. Indeed, this is the property which enabled styrene polymer compositions to be used for a much larger number of products. The Izod impact test, which had been used since the 1920's to measure fracture resistance, was selected by the industry in 1956-57^[4] to establish representative values for crystal polystyrene and three grades of impact polystyrene (medium, high and super high impact polystyrene). For compositions made using the same type of rubber, the principal difference among the grades of impact polystyrene is determined by the amount of a rubber used in making the composition. The Izod impact strengths provided by high impact styrene polymer compositions made using about 6% SBR rubber were generally about 1 ft-lb/in. notch or less. This is contrasted with crystal polystyrene which has an Izod impact strength of about 0.3 ft-lb/in. notch.

In late March, 1958, Mr. Heinig of AHC found a box on his desk. Based on a label and the feel of the material, he concluded that it contained a rubber composition from the Phillips Petroleum Company. While the label contained the words "cis 1,4 polybutadiene", he did not know its microstructure. The box had been left by a Phillips' salesperson who had earlier suggested that his company's new stereospecific rubber should be tried in making impact polystyrene.^[5] The material was ultimately determined to be a high cis, low vinyl polybutadiene produced by Phillips.

Heinig went immediately to Westphal and instructed him to make impact polystyrene using this material. In April, 1958, Westphal polymerized three specimens of impact polystyrene: one using the Phillips' polybutadiene at a 6% level, one using a mixture of 3% Phillips' polybutadiene and 3% SBR 1006, and a "control" specimen using 6% SBR 1006. The three interpolymers were prepared using AHC's then current process of producing impact polystyrene which involved the sequential addition of two specific catalysts. The first catalyst added was Lupersol # 7 (75% tertiary butyl peracetate). Somewhat later in the polymerization cycle, a second catalyst addition was made, this time with a mixture of Lupersol # 7 and a dissimilar catalyst, di-tertiary butyl peroxide.

The two products made using the Phillips' rubber were described by Westphal and Heinig in their April 16, 1958 Memorandum of Invention to exhibit two "disadvantages": "poor color" and a "slightly slower rate of polymerization," and one advantage: good impact strength. These are the only advantages or disadvantages disclosed in writing by Westphal and Heinig in their initial Memorandum of Invention in the laboratory records; or in their subsequent memorandum to Leon Horne, then AHC's in-house patent counsel. There was no reference (comparative or otherwise) in any document prepared in April or May 1958 to the properties of creep yield or gloss. At trial, Mr. Heinig recalled that the gloss of the two products made using the Phillips' rubber was about the same as the gloss of the SBR-based control.

Some additional work was done by AHC during the summer of 1958, when Westphal polymerized three additional impact polystyrene specimens using Phillips' stereospecific rubber (with its unknown microstructure) at the 2%, 4% and 10% levels. During this time there was substantial correspondence between Mr. Heinig, Dr. Finestone, Heinig's supervisor, and Mr. Zallen, AHC's outside patent counsel, who was drafting the first patent application. To define the scope of the "invention" for Mr. Zallen, Heinig and Westphal generalized from a single stereospecific polybutadiene data point. They assumed that the Phillips' rubber must be one of the stereospecific polybutadienes discussed in the Short article and used that article to predict the cis content range over which stereospecific polybutadiene would enhance the impact strength of polystyrene. Specifically, they obtained the cis range from three places in the Short article, which showed that a lower limit of cis content of 25% was important: (1) a table; (2) a sentence which states that "without exception, the important physical properties change little between 25 and 80% cis content", and (3) a sentence which states that "[p]roperties of the polymer in the range of 25-80% cis are in some respects reminiscent of emulsion polybutadienes, but are superior to the latter in resilience and low temperature characteristics." The Short article also disclosed a maximum cis content of 95.2%. Heinig and Westphal admit that they very specifically limited the upper limit of cis content to be used in the claimed "invention" of the 1958 application to 95%.

On July 22, 1958, Heinig rewrote the description he earlier provided to Mr. Horne of the experiments conducted in April and May, 1958, to avoid disclosing the sequential addition/dual catalyst system actually employed. His substitute disclosures identified a single catalyst, benzoyl peroxide, with no mention of either the sequential addition system or the two actually-employed catalysts.

AHC filed an application for patent in the United States, with Westphal and Heinig named as co-inventors, on September 5, 1958. The application discussed the use of "cis 1,4 polybutadiene" in high impact polystyrene, wherein the end product was made either by (a) polymerizing polybutadiene in styrene or (b) by physically blending the polystyrene and polybutadiene, rather than utilizing chemical "grafting".

The only two rubbery materials referred to in the 1958 application were the SBR control and a stereospecific polybutadiene disclosed as being 80% cis. Each of the three claims recited a cis content of 25-95% and the specification taught that the use of 1,4 polybutadiene having a cis content of 25-95% was essential to the disclosed invention. The application contained no reference to the term "high cis, low vinyl," but made reference only to "high cis polybutadiene", "1,4 polybutadiene", or a combination thereof. While it disclosed polybutadiene with a vinyl content of 1% to 10%, the application contained no statement regarding the significance of the vinyl content, and during the pendency of this 1958 application, AHC made no reference to any importance of vinyl content. Consistent with Mr. Heinig's inability to find a disclosure of how to make cis 1,4 polybutadiene, there was no disclosure in the application of how to make stereospecific polybutadiene spanning the specified cis content range, except that a reference was made to the use of a "heterogeneous" catalyst system. The application disclosed no "advantageous" properties other than improved impact resistance.

During the prosecution of the 1958 application, several events occurred abroad. First, in September, 1960, AHC filed patent applications in France, Canada, Sweden, Australia, West Germany, Great Britain, Japan and the Netherlands. Each of these applications was filed on subject matter common to the subject matter of the 1958 application, by Westphal and Heinig or their representatives or assigns. Second, the French patent application, filed on September 9, 1960, issued to AHC as French Patent 1,273,982 ("the '982 French patent") on September 11, 1961. AHC was aware of its issuance no later than September 16, 1961.

AHC filed its first continuation-in-part application (CIP) on November 27, 1961. This 1961 application disclosed and claimed for the first time that the cis content of polybutadiene within the scope of the invention could be above 95%. The 1961 CIP also claimed for the first time a vinyl content limitation of "not more than about 10%".

The oath accompanying the 1961 CIP application failed to identify any of the foreign applications filed during 1960 or 1961 and erroneously stated that no such foreign applications had been filed. The application also failed to mention that the '982 French patent had issued.

AHC continued to be unsuccessful at the PTO, and rather than appealing the PTO rejection to an ultimate decision, opted to file its second CIP. The second CIP ("the 1965 CIP") was filed on May 21, 1965. This time the accompanying oath or declaration purported to identify AHC's foreign counterpart applications. However, AHC misstated the filing date of the '982 French patent.

On May 23, 1967, AHC filed its third CIP ("the 1967 CIP"). This application, for the first time, disclosed the allegedly unexpected properties of improved gloss and creep yield.

The 1967 CIP, which eventually issued as the '434 patent, was appealed to the PTO Board of Appeals three times. In 1969, the Board affirmed the Examiner's rejection of all claims, holding that the use of the claimed polybutadiene for improved impact in impact polystyrene was prima facie obvious and that AHC's showing of a single comparative data point for each of creep yield and gloss was inadequate to rebut this presumption. On the next appeal in 1973, the case was remanded for further consideration of art not cited by the Examiner and for reevaluation in view of a new CCPA decision.

On the third appeal, the Board of Appeals concurred with the Examiner that the claimed subject matter was prima facie obvious. In the course of its decision, the Board of Appeals held that the 1958 application did not support claims covering more than 95% cis:

We find that the 1958 application is very clear, in presenting an upper limitation of 95% cis 1,4 polybutadiene as well as a disclosure of 1-10% 1,2-addition butadiene content for the mixture. We are not convinced by the affidavits presented in support of appellants' position. (COS 696, p. 970).

However, the Board reversed the Examiner, holding that AHC's Rule 132 affidavits showing improvements in creep yield and gloss rebutted the showing of prima facie obviousness and rendered the subject matter patentable. As a result, the '434 patent issued.

On October 29, 1971, AHC filed its fourth CIP, which was prosecuted simultaneously with the 1967 CIP. The claims of the 1967 CIP differed from those of the 1971 CIP in that the 1967 CIP contained certain additional process limitations as to the manner of forming the polymer compositions which were not recited in the 1971 CIP. AHC told the PTO that

[a]ll of the claims in the parent application Serial No. 640,478 [the 1967 application] require that the composition be made by mass polymerization. The claims on appeal do not define any mass polymerization limitation. The claims in the parent application state that the rubber component must be dissolved in the styrene monomer. No such limitation appears in the claims presently on appeal. (COS 696).

These two process distinctions represent the only substantive differences between the claims of the '434 patent and the '311 patent. This application, too, was rejected several times, was appealed, and then abandoned in favor of a still further CIP.

The fifth and last CIP was filed on February 24, 1975, and eventually resulted in the '311 patent, which was issued on December 7, 1976.

II. INFRINGEMENT

Since the issuance of the '434 patent in February, 1975, Cosden has made and sold moldable impact styrene polymer compositions using Diene 35, Diene 55 and Taktene polybutadienes. Its use of Diene 35 and Diene 55 polybutadiene to make such compositions continues to the present. These compositions are made by dissolving about 5-10% of Diene or Taktene polybutadiene in styrene and polymerizing the styrene monomer under free-radical polymerization conditions until polymerization of the styrene monomer is substantially complete. Cosden's polymerization is carried out using mass polymerization techniques under agitation conditions to provide dispersed and substantially uniform compositions having enhanced impact value. Having so found, only two infringement issues remain: (1) is the vinyl content of the Diene polybutadiene "not more than about 10%" and its cis content "at least 25%"; and (2) is the cis content of the Taktene polybutadiene about 95% and its vinyl content "not more than about 10%".

AHC's experts, Dr. Wiles and Professor Levy, analyzed six samples of Diene polybutadiene secured from Cosden and represented by it to be the material used in making its styrene polymer compositions. The vinyl contents of those samples, as measured by Dr. Wiles, were 8.6, 8.8, 8.5, 8.6, 8.6 and 8.5% and, as measured by Professor Levy, were 8.6, 8.3, 8.7, 8.5, 8.5 and 8.9%. Those measurements are in close conformity with the values published by Firestone throughout the years for the vinyl content of Diene polybutadiene of 7.5% and 9.1 or 9.4%^[6] They are also relatively close to the measurements of Cosden's expert, Dr. Koenig, when he analyzed six samples of commercial Diene polybutadiene: 10.7, 10.8, 10.7, 10.8 and 10.8%.

Dr. Wiles values for the cis content of the six samples which he analyzed were 33.9, 34.5, 33.9, 33.9, 33.9 and 33.7%. Firestone's published value for the cis content is 35.9%. Dr. Koenig's figures were 37.7, 38.1, 37.6, 37.8, 37.5, 37.7%.

With respect to Taktene, Dr. Wiles found a cis content in two samples of 96.6 and 96.8% and a vinyl content of 1.4% and 1.3%. Professor Levy analyzed for vinyl content only and obtained values of 2.8% and 1.5%. Cosden submitted no test data for Taktene.

Based upon the Wiles and Levy data, I conclude that the Diene polybutadiene used by Cosden has a vinyl content of "not more than about 10%" and a cis content of "at least 25%". I also conclude that the Taktene used by Cosden has a cis content of "about 95%" and a vinyl content of "not more than about 10%".^[7]

I accept the data referenced above because it is the product of careful and skilled analysis and because each analysis seems internally consistent as well as confirmatory of the others. I discount the attacks made by Cosden on this data for reasons I will summarize below.

Initially, Cosden argues that the analyses done by Levy and Wiles should be disregarded because their respective methods of measurements were allegedly not practiced in the Spring of 1958, when the invention was conceived. Levy conducted nuclear magnetic resonance ("NMR") analyses, which concededly were not available in 1958, to determine microstructure as called for in the patent claims. Wiles used infrared ("IR") spectroscopy which was well-known in 1958 but which, according to Cosden, was practiced at that time only in ways which differ materially from those used here. I do not adopt Cosden's position on this point because I disagree with both its legal premise and its analysis of the record.

I agree that the state of the art of measurement in a particular field may be highly relevant in construing a patent and determining the scope of its claims. The statute charges the applicant with communicating his invention to those skilled in the art and his teachings are to be understood in light of the art as it then existed. In determining the boundaries of "not more than about 10%", for example, it is helpful to know the margin of error in the measurement techniques of the day. If those in the art are aware that the margin of measurement error is plus or minus 5%, it is unlikely that they would read "not more than about 10%" to exclude anything below 10.5%, and quite likely that they would read it to include a somewhat higher limit.^[8] But using the existing state of the art to determine the scope of the claims of a patent is something quite different from limiting proof of infringement to methods in existence on the date of invention. If the scope is determined in the context of the existing art, I perceive no advantage and considerable mischief^[9] in freezing measurement technology and disregarding new learning which can establish, almost beyond peradventure, the precise characteristics of the accused substance. I do not believe the law so requires.^[10]

Moreover, I do not believe Dr. Wiles' technique differed materially from the varying techniques reflected in the literature in the late 50's and early 60's. Indeed, he followed a technique described in a 1949 publication. As one would expect, those practicing IR spectroscopy at that time had various preferences about the best technique to follow. At least one of these preferences even led some astray for a short time; but it is not true in any practical sense that Dr. Wiles' IR analysis was not the kind of analysis being conducted in 1958. Indeed, Dr. Hampton, a pioneer in the field, indicated that Dr. Wiles' method of analysis was the only one being used in 1958. He testified that, while varying techniques would produce somewhat varying results, most skilled artisans were using the same method and were achieving substantially the same results, with an error rate in the neighborhood of plus or minus 5%.

Cosden does not seriously challenge Professor Levy's results other than on the ground that his method was too modern. Dr. Wiles, however, is said to have made three "critical errors": (1) his use of iodine monochloride titration, (2) his calculated recovery of only 90 to 92% polybutadiene of the Diene samples which he analyzed and (3) his application of an iteration technique to calculate extinction coefficients. I find no fault with Dr. Wiles' titration method; his data show that the reproducibility obtained was good and the precipitation which occurred was "during the titration", not "during the reaction." With respect to the finding of 90% unsaturation in the vinyl standard, both Hampton and Wiles testified that this would only lead one to infer that the sample was not pure polybutadiene^[11] and would cast no doubt on Wiles' results. Finally, the iteration technique used by Wiles was that taught by Hampton, and followed by Silas, Morero and Gevers.

III. VALIDITY

A. The Threshold Issues.

At the outset, three related issues will be considered. The first issue is the earliest filing date to which AHC is entitled under 35 U.S.C. § 120. The second issue is whether AHC complied with 35 U.S.C. § 112 in connection with the various applications leading up to the patents-in-suit. The third issue is whether AHC made an invention or discovery prior to 1967, when it for the first time disclosed the allegedly improved properties of gloss and creep yield which were the predicate for the issuance of the patents.

1. 35 U.S.C. § 120.

35 U.S.C. § 120 allows a later-filed application, under specified circumstances, to claim the benefit of one or more earlier-filed applications. It is well established that to come within the purview of Section 120, (1) a later application must, inter alia, disclose the same invention as has previously been disclosed in each prior application, and (2) each application must comply with 35 U.S.C. § 112. See, e.g., 35 U.S.C. § 120; Acme Highway Products Corp. v. D.S. Brown Co., 431 F.2d 1074, 1078 (6th Cir. 1970), cert. denied, 401 U.S. 956, 91 S. Ct. 977, 28 L. Ed. 2d 239 (1971); Bendix Corp. v. Balax, Inc., 421 F.2d 809, 816-17 (7th Cir. 1970), cert. denied, 399 U.S. 911, 90 S. Ct. 2203, 26 L. Ed. 2d 562 (1979), reh. denied, 414 U.S. 819, 94 S. Ct. 43, 38 L. Ed. 2d 51 (1973); Chromalloy American Corp. v. Alloy Surfaces Co., 339 F. Supp. 859, 874 (D.Del.1972).

None of the claims of the patents-in-suit which claim a cis content range of 25 to 100%^[12] is entitled to a 1958 filing date because the invention in the patents differs from the invention described in the 1958 application and is the result of changes made in the 1961 CIP. There was a crucial difference between the descriptions of the invention set forth, respectively, in the 1958 application and the 1961 CIP concerning the maximum cis content of the polybutadiene to be used in the "invention":

1958 Application                                        1961 CIP
"The invention comprises essentially a composition    "The invention comprises essentially a product
comprising 80-99% of the vinyl aromatic               comprising 80-99% of a vinyl aromatic material
polymer and 1-20% total rubber material,              and 20-1% total rubbery materials, said rubbery
wherein the composition contains at least 1% of       materials comprising (a) at least 1% by weight of
a 1,4 polybutadiene having a cis-polymer              the total product of a 1,4 polybutadiene having a
percentage of 25-95% of the total polymer             cis content of at least 25% and a vinyl or 1,2
forms present." (COS 696 [58], pp. 1-2). (Emphasis    addition content of not more than 10%. . . ."
added).                                               (COS 696 [61], p. 19). (Emphasis added).

This distinction between the effective maximum cis content of polybutadiene within the invention disclosed in the 1958 application and that disclosed in the 1961 CIP is also evident from a second comparison between the two applications:

1958 Application                                        1961 CIP
"Only when at least 1% of total rubbery content       "Only when at least 1% of the total rubbery
of the composition consists of a 1,4 polybutadiene    content of the composition consists of a 1,4
having 25-95% cis polymer content are the             polybutadiene having a cis content of at least
improvements of this invention obtained."             25% and a vinyl content of not more than 10%
(COS 696 [58], p. 3). (Emphasis added.)               are the improvements of this invention obtained."
(COS 696 [61], p. 21). (Emphasis added).

AHC, in 1958, twice described the necessity to the invention of using 1,4 polybutadiene with a cis polymer content of 25-95%, and each of the claims of the 1958 application precisely tracked this microstructure range.

In view of these contrasting descriptions of the invention, it is apparent that in 1958, 1,4 polybutadiene having less than 25% cis content or having more than 95% cis content was excluded from the impact polystyrene invention. The specification is unambiguous. As earlier noted, the PTO Board of Appeals so held in ruling that the 1958 application did not support the claims which included the 95-100% range. Not only are objective descriptions of the 1958 invention not the same as the 1961 invention, but it also was AHC's subjective intention in 1958 to so-limit the invention. Westphal and Heinig admitted that AHC purposefully and deliberately limited the invention in 1958 to 1,4 polybutadiene having a maximum cis content of 95% and AHC's contemporaneous documents confirm that, in 1961, AHC management believed the 1958 application was so-limited.^[13]

Thus, the 1958 application clearly was not intended to encompass the use of a 1,4 polybutadiene having greater than 95% cis content and, contrary to AHC's contention, in the context of that application, one of ordinary skill in the art would not read the disclosure of "typical cis polybutadienes having compositions substantially in the range 25-95% cis" as teaching the use of polybutadiene having a cis content in excess of 95%. It follows that all claims covering a cis content range of 25-100% are not entitled to an effective filing date prior to November 27, 1961, the first time such an invention was disclosed within the meaning of Section 120.

2. 35 U.S.C. § 112.

The first paragraph of 35 U.S.C. § 112 concludes with the requirement that the patent specification "... shall set forth the best mode contemplated by the inventor of carrying out his invention." The "purpose of this ... requirement is to restrain inventors from applying for patents while at the same time concealing from the public preferred embodiments of the inventions which they have in fact conceived." In re Gay, 309 F.2d 769, 772, 50 Cust. & Pat.App. 725 (1962).

At the time of the filing of the 1958 application, AHC was utilizing a sequential addition/dual catalyst system in which Lupersol No. 7 and di-tertiary butyl peroxide were added in two stages. This was considered to be "proprietary information", in Mr. Heinig's words, however, and Example 1 of the 1958 application, and all subsequent applications, referred to a single addition of benzoyl peroxide as a catalyst. This was done because AHC believed the process for making high impact polystyrene which involved the two step technique had commercial value.

It does not follow, however, that the 1958 application violated Section 112. The requirement is of disclosure of the best mode of carrying out the invention and the character of the claimed invention must be kept in mind in applying this section. The alleged invention of Westphal and Heinig was a composition, not a process.^[14] Consequently, all claims of both patents in suit are directed to styrene polymer compositions, not to the method or process of making such compositions. The best styrene polymer compositions known to the inventors at the time they filed their 1958 application were fully disclosed in that application, and contrary to Cosden's assertions, nothing in this record establishes that the process described in Example 1 produces a composition which is in any way inferior to compositions made with the sequential addition/dual catalyst system being used by AHC in 1958 in its commercial production.

The first paragraph of Section 112 also required a patent applicant to disclose his invention "in such full, clear, concise, and exact terms as to enable any person skilled in the art ... to make and use the same...." Cosden asserts that the 1958 application of Westphal and Heinig failed to comply with this enablement requirement and that AHC is, therefore, not entitled to rely upon that application for its priority date with respect to the prior art. Cosden's primary argument is that the 1958 application (as well as the subsequent applications of Westphal and Heinig) should have specified an analytical technique for determining the microstructure of the polybutadiene used in making the disclosed compositions.

As earlier indicated, I am not persuaded that there were numerous measurement technologies in 1958 which would produce materially different results. There was one technology — infrared spectroscopy — and there was no need to tell one skilled in the art what technology to use in determining microstructure. Contrary to Dr. Koenig's assertion, Mr. Hampton testified that, if a sample of polybutadiene had been cut up into small pieces and sent to different people skilled in the art of infrared analysis in 1958, the competent spectroscopists should have obtained similar microstructure values. The fact that individual practitioners of infrared spectroscopy may have had individual preferences in technique for practicing the art did not mean that the 1958 application was inadequate for failing to name some particular practitioner.

Cosden also faults AHC for failing to specify certain process conditions in the patent which can affect the quality of the resulting product. The "process conditions" necessary to practice the invention, however, were the conventional processes already known to those skilled in the art by virtue of their disclosure in prior art patents and publications. Lever Bros. Co. v. Proctor & Gamble Mfg. Co., 139 F.2d 633, 638-39 (4th Cir. 1944). While conditions such as stirrer speed were shown to have an impact on the properties of the high impact polystyrene, there is no evidence that any special process conditions were required to obtain the advantages of the invention.

Finally, Cosden contends that the patent fails to comply with Section 112 because it was issued on the basis of allegedly superior and unexpected properties for impact polystyrene over a critical polybutadiene microstructure range, which is not critical in fact. In support of this argument, Cosden relies heavily on the holding of Kwik Set, Inc. v. Welch Grape Juice Co., 86 F.2d 945 (2d Cir. 1936) that:

A patentee may not arbitrarily select a point in a progressive change and maintain a patent monopoly for all operations in that progressive change falling on one particular side of that arbitrarily selected point. It is only where the selected point corresponds with the physical phenomenon and the patentee has discovered the point at which that physical phenomenon occurs that the maintenance of a patent monopoly is admissible.... Id. at 947.

I think it apparent from the record that "no more than about 10%" vinyl is not a point at which a distinctive physical phenomenon occurs in the sense that the results on one side of that point are materially different from the results immediately on the other. Indeed, AHC does not argue that "about 10%" is a "go/no go" point. Such a finding is not necessary to patentability, however. Assuming that properties throughout the claimed range are markedly better than the properties of the prior art so that the inventor has indeed described how to get a composition substantially superior to prior art compositions, a point addressed hereafter, it should make no difference under Section 112 if the results somewhat above 10% are also substantially superior to the prior art. In that event, the applicant would have satisfied the requirements of Section 112 even though he may have claimed a monopoly of more limited scope than the one to which he was entitled.^[15]

3. 35 U.S.C. § 101.

Section 101 of the Patent Act provides that one who "invents or discovers" a new composition shall be entitled to a patent. Cosden argues that AHC is not entitled to the 1958 filing date because Heinig and Westphal had not discovered the alleged invention as of that time. It points out that the patent was issued solely on the basis of the allegedly unexpected properties of improved creep and gloss and that the first mention of these properties came in the 1967 CIP. While the argument is not without appeal, I believe the issue is no longer open in the Third Circuit. Assuming for the moment that these improvements are characteristic of the claimed compositions, neither the failure to expressly mention them in the 1958 application, nor, indeed, the failure of the inventors to appreciate the improvements at that time would operate to bar AHC from relying on the 1958 filing date. Eli Lilly and Co. v. Premo Pharmaceutical Laboratories, Inc., 630 F.2d 120 (3d Cir.), cert. denied, 449 U.S. 1014, 101 S. Ct. 573, 66 L. Ed. 2d 473 (1980); In re Davies, 475 F.2d 667 (Cust. & Pat.App. 1973).

B. Anticipation

1. Shell's Work.

Section 102(g) stipulates that one is not entitled to a patent if "before the applicant's invention thereof, the invention was made in this country by another who has not abandoned, suppressed or concealed it." Citing this Section, Cosden asserts that AHC is not entitled to a patent on any claims reciting a cis content in excess of 95% because workers at Shell had earlier conceived of that invention and reduced it to practice.

(a) AHC's Experience With Very High Cis Polybutadiene.

I have previously concluded that the 1958 patent application did not claim or disclose impact polystyrene made with very high cis polybutadiene (i.e., polybutadiene having a cis content in excess of 95%). As a result, AHC cannot claim a filing date or constructive reduction to practice with respect to the claims including this range until November 27, 1961. AHC nevertheless asserts, however, that Westphal and Heinig conceived of using very high cis polybutadiene in April of 1958 and reduced that conception to practice at least by November of 1960. In evaluating this claim, one must keep in mind the principles recently reviewed by Judge Schwartz in Grefco, Inc. v. Kewanee Industries, Inc., 499 F. Supp. 844, 848 (Del.1980):

The date of invention is deemed to be the date of the filing of the patent application. U.S. Expansion Bolt Co. v. Jordan Industries, Inc., 488 F.2d 566, 568 n. 3 (3d Cir. 1973). The burden rests with the inventor to demonstrate an invention or reduction to practice prior to the time of filing. See generally Rex Chainbelt Inc. v. Borg-Warner Corp., 477 F.2d 481, 487 (7th Cir. 1973); 1 Chisum, Patents § 3.08[3]; United Show [Shoe] Machine Corp. v. Brooklyn Wood Heel Corp., 77 F.2d 263 (2d Cir. 1935). Moreover, the uncorroborated and undocumented testimony of the patentee is insufficient to prove invention date.^[16]

Westphal and Heinig each testified that they conceived of making polystyrene with very high cis polybutadiene in April of 1958, but that they failed to mention this conception in the 1958 application because their patent counsel, Mr. Zallen, had advised that there should be no reference to a polybutadiene which they had no reason to believe had actually been produced. There is no corroboration of this testimony, however, and, based on all of the evidence, I conclude that it is, at least in part, post hoc rationalization.

Westphal and Heinig did not know the microstructure of the Phillips polybutadiene when they used it to make polystyrene and their memorandum of invention makes no mention of microstructure. Moreover, there is no mention of microstructure in Westphal's laboratory notebook or in the research department summaries between April and July 1958. At some point prior to the September 1958 filing, Westphal and Heinig were exposed to the microstructure information in the Short article, however, and the range claimed in the 1958 application was a guess based on the microstructure there reported. That article makes no reference to a polybutadiene having a cis content in excess of 95%. Nor is there any such reference in any of the drafts of the 1958 patent application or in any of the communications between Westphal and Heinig and their patent counsel.

We do know that, in June of 1958, AHC, and I believe it fair to infer Westphal and Heinig, were aware of the possibility that one could produce a 98% cis polybutadiene with stereospecific catalysts. Dr. Finestone at AHC wrote a letter to Mr. Zallen on June 25, 1958 so stating. Moreover, I can accept that Westphal and Heinig believed in the summer of 1958 that one could make an impact polystyrene with very high cis polybutadiene. After all, they had no reason to believe otherwise. On the other hand, the record as a whole indicates that they did not believe at the time of filing the application that one would get markedly improved impact strength, much less improved gloss and creep yield, in polystyrene made with very high cis polybutadiene. Their sole basis for conceiving that any polybutadiene, other than the Phillips' sample, would produce improved results was the Short article which they used to formulate a guess on the range throughout which improved results would be achieved. Accordingly, the most rational inference from all the evidence is that they did not formulate the conception in 1958 that very high cis polybutadiene would give improved results.

This situation appears to have continued until May of 1961 when AHC received a sample of 99% cis polybutadiene and used it to make impact polystyrene with improved results. While AHC repeatedly sought samples of high cis polybutadiene from rubber companies between September 1958 and May 1961, no reference is made to very high cis polybutadiene and the evidence concerning this period provides no basis for concluding that Westphal's and Heinig's appreciation for the potential of that material increased in any way.

(b) Shell's Experience With Very High Cis Polybutadiene.

At some time prior to November, 1957, Lunk, Schroeder and Doyle at Shell conceived of using very high cis polybutadiene as a toughening agent for making impact polystyrene. The testimony of the inventors to this effect was corroborated by three employees of Shell at the time, as well as by documentary evidence. Dr. Lunk's laboratory notebook, which described in detail experiments run in November 1957 and January-February 1958 using a 98.6% cis, .6% trans, and .8% vinyl polybutadiene, confirms that these Shell workers had the concept of very high cis polystyrene in mind as of November 1957 and that they in fact made polystyrene with that polybutadiene on several different occasions prior to April of 1958.

Only a single sample of very high cis polybutadiene rubber made at Shell was available to Lunk, Schroeder and Doyle during the 1957-58 winter. The November 1957 polymerization was not stirred. It showed an impact strength of 1.40 and 2.2 foot per pound per inch at the 3% and 4% rubber levels respectively. In December, Shell began making stirred polymerizations in accordance with the teachings of the Amos patent. The next three polymerizations with very high cis polybutadiene, conducted in January and February 1958, accordingly, were stirred.^[17] The January 1958 runs were carried out at the 3% and 4% levels. The 3% run, a relatively low rubber level, showed impact strength of .96 ft./lbs. per inch; the 4% run showed a rather modest impact, .58 and .66 ft./lbs. per inch. In February 1958, the same rubber was again run at the 4% level. It showed impact strengths of 1.21 and 1.46 ft./lbs. per inch. With the exception of the one 4% run in January 1958, which apparently was over-agitated, those impact results were substantially higher than Shell was contemporaneously achieving with SBR at comparable rubber levels. The November and February results at the 4% level were also better than those achieved by AHC at that level several months later with 95% cis polybutadiene.

Shell's work with very high cis polybutadiene was done with the expectation that it might produce a polystyrene with improved impact strength and Shell's experiments prior to April 1958 show that such a result was in fact achieved. The record affirmatively establishes, however, that the workers at Shell did not appreciate during that period of time that the improved results were attributable to the very high cis polybutadiene. The Shell monthly research reports do not report at all on the November 1957 to February 1958 work. Moreover, the January 1959 research report affirmatively concluded as follows:

Although properties similar to those obtained with S-1006 [i.e. SBR] were obtained with some experimental materials such as emulsion polybutadiene, and cis polybutadiene, none of these materials gave products superior to those from S-1006. Availability and ease of handling are other and perhaps not unimportant, factors in favor of S-1006.^[18]

A June 1959 report does show, however, that the Shell workers appreciated the connection between very high cis polybutadiene and higher impact strength by that time. In that publication, Lunk and Schroeder reported as follows:

Several types of cis-polybutadiene, differing in molecular weight and microstructure, have been investigated in the preparation of high impact polystyrene via bulk interpolymerization. Compared with SBR rubbers (i.e., S-1006) they showed the following advantages:

(1) Higher impact strength and elongation. At a rubber concentration of 5%, impact strengths of 1.5-2 ft. lb/in. were obtained, as compared to 0.9-1.2 ft. lb/in. with S-1006. The improvements in elongation were less spectacular but still substantial.

(2) Better resistance to high temperatures during the polymeriation reaction. With S-1006, considerable losses of impact strength and elongation occurred when the final temperature of the polymerization reaction was higher than 180° C. Interpolymers containing cis-polybutadiene suffered little or no losses at temperatures as high as 195° C.

(c) Analysis.

Whether or not there can be a conception of an invention without an appreciation by the inventor of unexpected, inherent properties which alone support its patentability has been a matter of some debate.^[19] I need not resolve that issue here, however, because under either view, Cosden prevails on its contention that Shell's work anticipates those claims of the patent-in-suit which include very high cis polybutadiene. If appreciation is not required, the evidence clearly and convincingly demonstrates that Shell both conceived this invention and reduced it to practice in the winter of 1957-58 before Heinig and Westphal received the Phillips sample. If appreciation is required, the record clearly and convincingly establishes that Shell had both conceived it and reduced it to practice on or before June 1959, while there is nothing in the record to support conception or actual reduction by AHC prior to May of 1961.

It follows that, under Section 102(g), all claims including the cis range over 95% are invalid unless it can be said that Shell "abandoned, suppressed or concealed" its invention. Clearly, it cannot. On January 25, 1960, Shell filed a United States patent application in the names of Lunk, Schroeder and Doyle covering the use of polybutadiene having a cis content above 90%, and preferably about 95%, as the toughening agent for making impact polystyrene, the subject matter of that application was carried forward in successive continuation-in-part applications for the next fourteen years until the last application in the chain was ultimately abandoned in 1974. AHC argues that this abandonment of the last patent application filed by Lunk, Schroeder and Doyle is evidence of abandonment of their 1957-1958 work. I disagree. The fourteen years work in attempting to secure patent protection provides more support for the conclusion that there was no intent to abandon, than the ultimate cessation of those unsuccessful efforts provides for a contrary conclusion.^[20]See Jacquard Knitting Mach. Co., Inc. v. Ordinance Gauge Co., 213 F.2d 503, 508 (3d Cir. 1954). Moreover, even if Shell had abandoned its invention in 1974, it would not help AHC since the relevant legal issue is whether there had been an abandonment as of 1961. Allen v. W.H. Brady Co., 508 F.2d 64 (7th Cir. 1974); Continental Copper and Steel Industries, Inc., 196 U.S.P.Q. 30, 37 (M.D. Pa.1976).

With respect to concealment and suppression, AHC contends that notebooks, records and reports relating to the work of Lunk, Schroeder and Doyle have never been made available to the public by Shell and that the abandoned patent application continued to be inaccessible to the public because of the PTO's policy of confidentiality set forth in 37 C.F.R. § 1.14.

Although many of Shell's internal corporate documents and laboratory records have not been made available to the public, this does not support an inference of "suppression or concealment." And, as far as the public availability of Shell's abandoned patent application is concerned, reference to that application in the Lunk '199 patent is sufficient to render it available to the public as of its January, 1964 issue date. In re Lund, 376 F.2d 982, 991, 54 Cust. & Pat. App. 1361 (1967); 37 C.F.R. 1.14(b).

Like abandonment, suppression or concealment must be intentional. English v. Heredero, 200 U.S.P.Q. 597, 600 (Bd.Pat. Interf.1978). Shell spent fourteen years trying to get a U.S. patent. It disclosed the Lunk, Schroeder and Doyle work in another U.S. patent to Lunk, which did issue. The work was also described in ten foreign published patent applications. All three inventors and their patent counsel testified that there was never, at any time, an intent to suppress or conceal the Lunk, Schroeder, Doyle work. The evidence is thus inconsistent with an intent to suppress or conceal.

2. The Amos '692 Patent.

The Amos '692 patent, in Example 6, taught a polystyrene composition obtained by polymerizing styrene monomer in the presence of a solution polymerized polybutadiene. Dr. Bucknell, Cosden's expert, explained that the cis content of this polybutadiene was in the range of 25% and the vinyl content was in the range of 18.5%. This is consistent with the data which AHC provided to the patent office. Cosden maintains that the 10% vinyl content limitations of the patents-in-suit is not "material" and that Example 6 of the Amos patent meets every other limitation of the patents-in-suit.

In addition to maintaining that the 10% vinyl limitation is material, AHC points out that Claims 1, 3-14, and 27 of the '131 require a graft interpolymer of styrene and 1,4 polybutadiene, while Claims 1, 3-14 of that patent specify that the composition must contain a discontinuous phase which is substantially insoluble in toluene at room temperature. These latter two limitations do not distinguish Amos, however. Amos clearly taught graft copolymers as well as linear polymers. With respect to solubility, the Amos '692 patent did not discuss whether the discontinuous phase of the impact polystyrene product was soluble or not. It dealt with the product as a whole. It appears, however, that the discontinuous phase of Amos is substantially insoluble.

I agree with AHC and the PTO, however, that the 10% vinyl limitation satisfactorily distinguishes the Amos '692 patent.

3. The Schramm '227 Patent; The Lunk '199 Patent; and the '982 French Patent.

Cosden cites the Schramm '277 patent; the Lunk '199 patent; and the '982 French patent as anticipations of the claims of the patents-in-suit that include the 95% to 100% range. Having reached the conclusion I reached on the Section 102(g), I need not decide these issues.

C. Obviousness.

For an invention to be held unpatentable under Section 103 of the Patent Act, it must be found that the differences between the invention and the prior art are such that the subject matter of the invention as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art. Clopay Corp. v. Blessings Corp., 422 F. Supp. 1312 (D.Del.1976).

The pertinent art in the context of this case is the commercial production of high impact polystyrene and the development of improved forms thereof. The record indicates that, as of September 1958 and at all times here pertinent, the persons skilled in this art were employed by major chemical companies, like Monsanto and Dow, which were the principal competitors in the impact polystyrene market. These firms employed groups of people who had specific responsibility for developing impact polystyrene with improved properties. They were thus "problem solvers". See, e.g., Systematic Tool & Machine Co. v. Walter Kidde & Co., Inc., 555 F.2d 342 (3d Cir.), cert. denied, 434 U.S. 857, 98 S. Ct. 178, 54 L. Ed. 2d 128 (1977); Komline-Sanderson Eng. Corp. v. Ingersoll Rand, 485 F. Supp. 973 (D.Del.), aff'd 639 F.2d 773 (3d Cir. 1980). Typically, these groups would include someone with post-graduate work in chemistry, as well as others with less formal education but many years of practical experience in this or related arts. Whatever the mix of education and on-the-job training, however, it is fair to characterize the members of these groups as well informed and highly skilled in the art.

As AHC readily concedes, it was well known in the art in September of 1958 that the impact properties of polystyrene could be enhanced by polymerizing the styrene monomer in the presence of rubbery polymers. The process customarily used at that time for accomplishing this objective was that described in the Amos '692 patent.^[21] Amos taught that the rubber should be dissolved in styrene monomer and stirred until it became a suitable prepolymer; the prepolymer was then placed in the second vessel where the polymerization would be completed. Amos also taught that polybutadiene was one of the rubbery modifiers which could be used in this way to "toughen" polystyrene. In particular, his example 6 described an experiment using a polybutadiene prepared by free radical polymerization. Amos does not specify the microstructure of the polybutadiene, but one skilled in the art would realize from his description of its preparation that it had a microstructure of about 25% cis and about 18% vinyl. This microstructure is, of course, quite close to that of polybutadienes included within the microstructure range referred to in the claims-in-suit and it seems clear that one skilled in the art would understand from Amos that impact polystyrene could be made with polybutadienes within the claimed range. Amos did not suggest, however, that higher cis (i.e., over 25%), lower vinyl (i.e., not more than about 10%) polybutadienes would produce a more desirable product than other polybutadienes or than prior art SBR rubber.

AHC further concedes that one skilled in the art in September of 1958 would know that emulsion polybutadiene, typically having a cis content of about 20% and a vinyl content of about 18%, could be used to toughen polystyrene. The teaching that emulsion polybutadiene could be utilized in making impact polystyrene^[22] is said by AHC to be the closest prior art.

High cis, low vinyl polybutadienes coming within the microstructure range cited in the patents in suit had been produced using stereospecific catalysts well prior to April of 1958. The Smith, et al patent,^[23] a Phillips patent filed in April of 1956, for example, disclosed a cis 4-type polybutadiene of generally high cis content. The Foster patent,^[24] a Firestone patent filed in August of 1956, disclosed the Diene type polybutadienes, of generally low to medium cis content. By October of 1957, Phillips had had an opportunity to do some in depth study of the characteristics of the new stereospecific polybutadienes. The Short article was published in that month.^[25] This article considered the effect of changes in the cis content on the physical properties of polybutadienes prepared with stereospecific catalysts systems and compared the properties of these compositions with those of emulsion polybutadiene and SBR. Short concluded that the high cis polybutadienes had "excellent resilience", and that the properties of polymers in the range of 25-80% cis "are in some respects reminiscent of emulsion polybutadiene but are superior to the latter in resilience and low temperature characteristics."^[26] Short also explained that a comparison of the stereospecific product with natural rubber and SBR revealed a "marked superiority of 1,4 polybutadiene at low temperatures."^[27]

Neither Short nor any other prior art reference suggested that if one polymerized styrene monomer in the presence of high cis, low vinyl polybutadiene, one would obtain a polystyrene with better impact strength, creep and gloss than could be obtained with emulsion polybutadiene or SBR. I am persuaded, however, that the Short article would motivate one of ordinary skill in the art to use Amos's teachings to make polystyrene with high cis, low vinyl polybutadiene in the expectation that the impact strength and low temperature properties of the resulting product would be superior to compositions made with emulsion polybutadiene or with SBR.

Contrary to AHC's contention, I believe those in the art of making Amos technology polystyrene did look to the properties of the rubber to predict characteristics of the polystyrene product. Accordingly, I believe that one of ordinary skill in the art would see in the resilience and low temperature characteristics of high cis, low vinyl polybutadiene an indication that one could make polystyrene with significantly improved impact strength and that such improvement would be particularly marked in low temperature applications. Richard A. Bishop, then Director of Research at AHC, for example, confirmed in an article published in 1955,^[28] that block and graft polymers differ from polymers "made using the same ratios of monomers in the usual process," because in block and graft polymers "properties of one material can be added onto the backbone of another material without changing the essential nature of the backbone."^[29] Mr. Bishop specifically suggested that a correlation existed between rubber resiliency and its ability to toughen polystyrene: "GR-S synthetic rubber is known to possess excellent resiliency and this material has been added to polystyrene to improve impact strength." Similarly, Safford^[30] taught that block and graft polymer "tend to retain the characteristic properties of the homopolymer of individual monomers." Indeed, Heinig, as well as Bucknall and Ruffing, confirmed that the art looked to the properties of the rubber as one indicator of the end product properties of impact styrene which might reasonably be expected.

While I do not suggest, based upon this evidence, that there was absolute predictability, I do find that those of ordinary skill in the art had a reasonable expectation that a graft copolymer of styrene and polybutadiene, made from polybutadiene having good low temperature properties and excellent resilience, would have the same superior properties. I believe this is borne out by the fact that, after dissemination of the information contained in the Short article, those practicing the art who could get their hands on high cis, low vinyl polybutadiene polymerized styrene monomer in the presence of that material within a very short time after they obtained it. As previously noted, Shell prepared impact polystyrene using high cis, low vinyl polybutadiene in late 1957, after receiving a small sample in October. Westphal and Heinig tried high cis, low vinyl polybutadiene immediately upon receiving a sample in April of 1958. Ruffing, at Dow, was unable to secure a sample of stereospecific butadiene until late 1958 and used that sample to make impact polystyrene on January 5, 1959. Indeed, within three years of Shell's initial work, not only these firms but Cosden, Firestone, Distrene, Monsanto, J.T. Baker Company, and the Richardson Company had all made impact polystyrene reinforced with stereo-specific polybutadiene. While it is true, as AHC contends, that the existence of high cis, low vinyl polybutadiene was publicly announced as early as 1956, I do not find this fact probative of non-obviousness since it was not readily available to artisans working in the polystyrene art until considerably later.

In reaching the conclusion that Amos and Short would motivate an artisan of ordinary skill to practice the invention of the patents in suit, I have not overlooked AHC's argument that there are counter-teachings in the prior art. While it had once been conventional wisdom that one needed a styrene and butadiene copolymer in order to make the modifer compatible with the styrene monomer, by 1958, it was known that such compatibility was not required when making a chemical graft of a rubber toughened plastic, rather than a mechanical blend.^[31]

Likewise, I do not read the Stoops^[32] and Safford^[33] patents to teach that high vinyl polybutadiene or low cis polybutadiene are preferable to other types of polybutadienes. Stoops does not refer to the microstructure of polybutadiene. He found that polybutadiene polymerized at 60°C. or below gave better results in his impact polystyrene than did polystyrene polymerized at 90°C., and speculated that the reason why this occurs is the "less highly branched structure" of the 60°C. polybutadiene, not because of vinyl content. While Stauffer utilized a polybutadiene having a vinyl content of over 30% in making a speciality impact polystyrene with decreased perma plastic flow and "excellent electrical properties", I do not read his teachings to suggest that high cis, low vinyl polybutadiene would not produce a polybutadiene with good impact and low temperature characteristics. Even if the hypothetical artisan had read Safford as AHC does, however, I believe that Short's work would quickly convince him of the error of Safford's suggestion.

Nor have I overlooked the significant commercial success that high cis, low vinyl polybutadiene has experienced. I understand Cosden to concede that virtually all high impact polystyrene is now made with either Diene or Taktene, both of which are within the claimed microstructure range. The probative value of commercial success evidence, however, depends in large part on whether, given the surrounding circumstances, it will support an inference that the market potential was sufficiently strong to motivate others who must have tried and failed to solve the problem. Here, we know there were no failures of others. As stereospecific polybutadiene became available commercially in 1959, 1960, 1961, it was accepted and used by the industry. Thus the evidence, taken as a whole, supports the inference of obviousness, rather than non-obviousness.^[34]

My conclusion about Amos and Short does not end the matter, however. The Board of Appeals of the PTO, on two occasions, reached substantially the same conclusion and nevertheless caused the patent to be issued. It did so because it was persuaded that an artisan would not have expected the use of high cis, low vinyl polybutadiene to produce a composition which was not only superior in impact strength, but also in gloss and creep yield as well.^[35]

Cosden acknowledges that a novel composition can be prima facie obvious and, nevertheless, be a patentable invention if it is characterized by surprisingly improved properties. Moreover, Cosden cannot seriously contend that based upon the prior art in 1958, one would expect the compositions of the patents in suit to have markedly improved creep yield and gloss. Its position is rather that these unexpected properties were non-existent in the 1958 invention, and do not occur throughout the claimed range. It suggests that any improvements in creep yield and gloss result from using commercial Diene and improved process conditions.

I am persuaded by the evidence that within the range of approximately 0 to 30% vinyl there is no positive correlation between vinyl content and improved creep and gloss and that not all compositions within the claimed range exhibit creep and gloss properties which are markedly better than those exhibited by polystyrene made with emulsion polybutadiene. This conclusion is based upon the testimony of Dr. Bucknall and on other evidence which tends to corroborate that testimony.

Dr. Bucknall testified that, in his opinion, vinyl content in the range of approximately 0 to 30% did not materially effect the creep and gloss characteristics of impact polystyrene. His opinion was based on his knowledge of the field and his review of testing data from three series of tests: the Labofina tests, the TNO 1 tests, and the TNO 2 tests. Finding Dr. Bucknall to be a highly qualified and objective expert witness and his testimony to be supported by the test data to which he referred, I credit this opinion.

The Labofina and TNO 1 data enabled Dr. Bucknall to conclude that over a range from 5% to 20% vinyl, vinyl content does not affect creep and gloss qualify and that compositions without the claimed range frequently have creep and gloss characteristics equal or superior to compositions within the claimed range. The Labofina test, for example, included three samples within the claimed range and seven without. Those without had vinyl content ranging between 12.1% and 19.6%; one had cis content of 21.5% cis and 19.5% vinyl, a microstructure approximately that of prior art emulsion polybutadiene and not far removed from Amos's Example 6. With respect to creep and gloss, the samples within the claimed range were superior to some of the samples without that range and inferior to others. In both categories, the 21.5%, 19.5% vinyl sample was superior to each of the samples within the claimed range.

In TNO 2, the testing compared the creep yield time and gloss of samples produced at varying stirrer speeds with Taktene (98% cis, 1% trans, 1% vinyl), emulsion polybutadiene (about 20% cis, 18% vinyl), and two rubbers of different viscosity having the same microstructure within the claimed range (i.e. 32% cis, 53% trans, 15% vinyl). With respect to creep yield time, Taktene performed no better than the other samples at some stirrer speeds and was inferior to all samples at other stirrer speeds. With respect to gloss, the Taktene samples were inferior to the emulsion sample and one of the medium cis samples at each speed. Taktene and the lower viscosity medium cis samples were about the same.

While AHC challenges the validity of this testing and Dr. Bucknall's testimony based thereon, I do not find its arguments persuasive. Dr. Bucknall did testify that he would need perhaps three years of testing to produce a definitive analysis of all of the factors which might affect impact, creep and gloss, but he did not indicate that one would need more testing than had been done in order to exclude vinyl content as one of those factors.

Nor am I persuaded that Cosden's testing provides an inadequate basis for teaching conclusions about creep and gloss because certain variables, other than vinyl content, were not maintained at a constant level. All process and ingredient variables relevant to creep and gloss were kept constant with the sole exceptions that stirrer speed and solution viscosity differed in some of the testing. The data, as indicated above, provided a basis for comparing the creep and gloss characteristics of polystyrenes produced at the same stirrer speed within and without the claimed range. Thus, the variation of stirrer speed does not impair one's ability to draw meaningful conclusions. With respect to solution viscosity, while it is true that the solution viscosity of the prepolymer varied as different rubbers were used, this is not significant for present purposes. The patent claims a range of compositions, asserting that impact polystyrene containing polybutadiene rubbers having microstructures outside that range. Since one of ordinary skill in the art would know that the solution viscosity of the prepolymer would change as one used different polybutadiene rubbers within and without the claimed range at the same rubber concentration, such an artisan would obviously not read the patent to claim superiority in creep and gloss of the claimed compositions solely over impact polystyrene produced with prior art rubbers having the same solution viscosity. AHC obviously did not read the patents to so indicate either. The comparative testing submitted by it to the PTO, for example, compared polystyrene produced with prepolymers having different solution viscosities.

The patent does not claim that any special process conditions are necessary to obtain the advantages of the invention. The only comparison which the terms of the patent make relevant is that between compositions inside the claimed range and those outside the range, so long as both are prepared in a manner known to those of ordinary skill in the art at the time. Cosden's testing was of compositions prepared in this manner.^[36]

Dr. Bucknall's opinion finds some additional support from the creep yield testing that AHC conducted between 1967 and 1970. That data will be discussed hereafter; it will suffice here to note that it suggests that Taktene and Diene reinforced polystyrene is not invariably superior to emulsion polybutadiene reinforced polystyrene in creep yield. Further corroboration of the Bucknall testimony is found in the fact that the polystyrene originally produced by Heinig and Westphal using Phillips polybutadiene (about 95% cis, less than 10% vinyl) did not have improved gloss, as well as in the fact that no claim of improved creep and gloss was made for approximately nine years after the invention.

Finally, the countervailing evidence tending to support the proposition that polystyrene compositions throughout the claimed range have surprisingly improved creep and gloss properties is not plentiful. While the PTO concluded that this proposition was true, as well be discussed hereafter, it did so on the basis of very sparse data which was not representative of the testing which produced it. Moreover, the most that the data now of record will support is that medium cis, low vinyl polybutadienes (i.e. 35-40% cis, 7-10% vinyl) like Diene, appear to produce better creep and gloss properties than do emulsion polybutadienes. Even if this proposition be accepted, however, it will not support the claims of the patents in suit.

Finally, AHC stresses its commercial success evidence and points out that Cosden has found it advantageous to operate within the claimed range rather than without. This record provides no basis, however, for inferring that compositions throughout the claimed range have experienced significant commercial success or that those which have experienced success have done so because of surprisingly superior creep and gloss, as opposed to improved impact or other properties.

Since the file wrapper discloses that the patents in suit were issued on the basis of the claimed compositions having markedly improved creep and gloss, the fact that the creep and gloss data submitted to the PTO was not representative of the testing which produced it, deprives the patents of the presumption of validity. Even if it were necessary, however, for Cosden to show by clear and convincing evidence that not all the claimed compositions are superior in creep and gloss properties, the Bucknall analysis and the testing which support it would satisfy that burden.

In Eli Lilly & Co. v. Premo Pharmaceutical Labs., 630 F.2d 120 (3d Cir.), cert. denied, 449 U.S. 1014, 101 S. Ct. 573, 66 L. Ed. 2d 473 (1981), the Third Circuit endorsed the rule of In re Papesch, 315 F.2d 381, 50 Cust. & Pat.App. 1084 (1963) that chemical compounds which are structurally obvious from the prior art are nonetheless patentable if the compounds have useful properties which would not have been expected from the prior art. This conclusion follows from the "inseparability" of a compound and its properties. The corollary to the Papesch rule is that a structurally obvious compound, or composition in this case, is patentable only if its properties are non-obvious. I have already concluded that it was obvious that the compositions claimed in the patent in suit would be a useful HIPS, and that they would have enhanced impact strength. While the PTO found additional improved properties to justify issuance of the patent only a subset of the compositions claimed in the patent manifest those properties.

Overbroad claims like the ones in suit pose two related problems under the statutory criteria. First, it cannot be said that the compositions which do not have improved creep yield or gloss are non-obvious, under Section 103, and a patent applicant cannot bootstrap a monopoly over a large range of obvious compositions merely by demonstrating that a few compositions within the range are not obvious. See 2 Chisum, Patents, § 7.03[7] at 7-41-42; In re Cavallito, 282 F.2d 357, 361, 48 Cust. & Pat.App. 711 (1960). Second, a patent which discloses a large number of compositions, only a few of which actually exhibit the properties of the invention, fails to satisfy the enablement requirement of Section 112. The patent in suit does not teach a practitioner with ordinary skill in the art how to distinguish those compositions which do give improved creep yield and gloss from those which do not. It leaves the patent reader to his or her own patient trial and error; it does not therefore "enable any person skilled in the art ... to make and use the same." 35 U.S.C. § 112. Such a disclosure fails to satisfy the most fundamental objective of the patent law, which is to disseminate useful information while preserving incentives for the inventor. The overbreadth of the claims in suit renders them invalid under Sections 103 and 112 of the Patent Act.

IV. ENFORCEABILITY.

A. The 1958 Application.

The only substantive allegation of fraud with respect to the 1958 application is the failure to cite the examiner to the Short article. Cosden has shown that Heinig and Westphal relied upon that article for an understanding of the properties of the new polybutadienes and for their definition of the microstructure range. Moreover, it is apparent from the prior discussion that the Short article was highly pertinent prior art. As AHC points out, however, before the patents in suit were issued, the PTO had considered numerous prior art references which discussed the improved properties over SBR and emulsion polybutadiene attributed to those polybutadienes in the Short article, such as resilience and low temperature properties. See, e.g., Naylor; Youngman; Foster; and Smith. Accordingly, it cannot be said that the Short article represents more pertinent art than that before the Patent Office at the time the patents issued. On the other hand, it is significant that when AHC chose not to disclose the Short article in 1958, it did not disclose any of these subsequently cited references and that the examiner did not discover any of them during the period between the filing of the patent in September of 1958 and the abandonment of the 1958 filing in November of 1961. While the failure to disclose the Short article does not, under these circumstances, render the patent in suit unenforceable, I consider it relevant evidence of the manner in which this application was prosecuted in the early years.

B. The 1961 Filing.

In September 1961, two months before the filing of the 1961 continuation in part, Westphal prepared the so-called Interim Report. This report represents three years of work on the Westphal-Heinig invention. It was important that a clear picture of the testing to-date be given in the Interim Report because it was being submitted to Foster Grant's President and other executives in order that they might make a determination on how to proceed in the developing of impact polystyrene.

The 1961 CIP added new examples 6-21. Examples 6-9 purport to compare SBR, emulsion polybutadiene (approximately 20% cis), Diene (approximately 35% cis), and cis-4 (approximately 95% cis) as impact modifiers. Examples 6, 8 and 9 were taken from the Interim Report. Table III of the Interim Report compared Diene, cis-4 and SBR impact strength data from one laboratory run which simultaneously polymerized three specimens. Table IV of the Interim Report compared Diene, cis-4 and SBR impact strength data from another laboratory run done at a different time. The cis-4 data from Table III was higher than the cis-4 data from Table IV. Example 9, designed to illustrate the performance of cis-4, took the impact value, 1.88, from Table III. The SBR "control" impact data from Table III was lower than the SBR impact data from Table IV and Example 6, designed to represent the performance of SBR, took the impact value 0.9, from Table III. On the other hand, the Diene impact data from Table IV was much higher than the Diene impact data from Table III (1.60 versus 1.18). The Diene example, example 8 of the 1961, CIP used the 1.60 value from Table IV. This data can be summarized as follows:

Patents-in-Suit             Interim Report                  Interim Report
Examples                   Table III                        Table IV
6      8      9         20-132(3)    6       5         20-152(2)    5       4
SBR       6%     0      0           6%         0       0           6%         0       0
Diene     0      6%     0           0          6%      0           0          6%      0
Cis-4     0      0      6%          0          0       6%          0          0       6%
Impact    0.9    1.6    1.88        0.9        1.18    1.88        1.04       1.60    1.84

The impression given, of course, is that both Diene and cis-4 give markedly improved impact values as contrasted with SBR.

Examples 10 through 12 disclosed interpolymers in which the rubbery toughening agent was a mixture of SBR and cis-4. A comparison between the data reported in Examples 10 through 12 and data reported on Table XI of the Interim Report demonstrates that this segment of the Interim Report was the source of the information reported in the CIP. The value reported for the control was altered, however. The data may be summarized as follows:

Patents-in-Suit                   Interim Report
Examples                     Experimental Run Nos.
10      11       12        20-103(3)      20-103(5)      20-103(6)
% SBR rubber         5.2     3.8      2.5         5.0            3.7            2.4
% cis-4 (95% cis)      0     1.0      2.1           0            1.0            2.0
Impact Strength      0.80    0.93     1.1         1.15           0.78           0.81[37]

Once again, material data within the scope of what was then claimed was misrepresented. The actual experiment suggested just the opposite of what the examples were intended to show — i.e. that additions of small amounts of cis-4 polybutadiene increased impact strength.

Examples 18-21 of the 1961 CIP, corresponding to Examples 18-21 of the '434 and the '311 patents, pertained to mechanical blends of polystyrene and rubbery material. Table X of the Interim Report also disclosed blends. That Table disclosed, inter alia, a polyblend using 7.5% SBR and a polyblend using 7.5% cis-4, as do Examples 21 and 18, respectively. However, once again the impact data reported in the Interim Report is at variance with the data AHC presented to the PTO:

Patent-in-Suit                   Interim Report
Example 21        Example 18               Table X
7.5%                  0.9                                       0.8
7.5% cis-4                              1.43                    0.72

Of these three misrepresentations, that in Examples 6, 8 and 9 is the most material in the context of the patents as finally issued. AHC does not suggest that the choice of Diene data to be included in the 1961 CIP was inadvertent. Rather, it argues that the experiment using Diene which is reflected on Table III was suspect because it had a slow polymerization rate and a high "viscosity at dump" and that the Diene result taken from Table IV and reported in Example 8 was "representative of the data produced" by its experimentation. Even if the factual premises of this argument be accepted, it would not explain why this Diene result was not compared with the SBR control in that experiment rather than the lower result of the SBR control from Table III. But, the problem is more substantial in any event.

The data in the Interim Report must have been either a complete record of the relevant data available to AHC or a nearly complete record which was representative of the data then available. I say this not only because Westphal, the author, testified that it was intended to be a complete picture of the experimentation to-date, but also because, if it were not either complete or at least representative, it would not have been presented to top management as the basis for making an important commercial decision. The Interim Report reported nine experiments making polystyrene with Diene at varying rubber levels between 4% and 9%. The six of the nine polystyrene samples produced in these experiments which were prepared using the laboratory filter press showed no improvements in impact strength over the SBR control. With respect to the three remaining samples, which were prepared using laboratory "cans", AHC claims to have considered the one from Table III "non-representative" for reasons related above and Westphal noted contemporaneously with respect to other samples that the impact values for both control and Diene were "unusually high" (i.e. 1.86 for the control and 2.52 for the Diene). The third Diene result, that from Table IV, was the only Diene result reported to the PTO in the 1961 CIP.

AHC contends that inventors are entitled to use some judgment regarding the reliability of data generated through experimentation in determining what data should be submitted to the PTO, just as they are entitled to utilize some judgment in submitting prior art references. Thus, according to AHC, inventors may omit "aberrational data" so long as the data submitted is representative. This argument may be accepted for present purposes, because it does not avail AHC. In addition to the fact that the sole data with respect to Diene was compared with a different control in order to enhance the claimed improvement, that data was not representative of AHC's research and was therefore misleading.

The impression conveyed by the 1961 CIP was that use of high cis, low vinyl polybutadiene in the lower part of the cis range claimed, as well as that in the higher part of the claimed range, would invariably provide polystyrene having impact strength significantly in excess of comparably produced polystyrene made with prior art SBR rubber. The single impact figure provided was not representative of the data which AHC then had which was relevant to this asserted proposition, however. While I believe the AHC personnel when they say they had more faith in "can" produced data than that which was filter press produced, this is not an answer. They nevertheless continued to generate and rely on filter press data for a number of purposes and obviously considered it to be of substantial value in making their own decisions. Moreover, this was not a case where AHC had extensive "can" data which contradicted the filter press data. It had only a single "can" result which it considered reliable. In these circumstances the single value provided, even if contrasted with the correct control, would not have been representative of AHC's research.

I am confident that the "adjustments" made in the data from the Interim Report were deliberate. They were also relevant in the sense that the PTO at that point in time considered the comparative data useful in addressing the issues presented by the claims of surprisingly improved impact strength then being made on behalf of high cis, low vinyl polybutadienes. It does not follow, however, that the claims in suit are invalid or unenforceable.

To render a patent invalid, a misrepresentation must be not only deliberate and intentional; it must also be material. Delong Corp. v. Raymond International, Inc., 622 F.2d 1135, 1145 (3d Cir. 1980); In re Frost Patent, 540 F.2d 601, 603-604 (3d Cir. 1976); Hercules, Inc. v. Exxon Corp., 497 F. Supp. 661, 689 (D.Del.1980). In general it may be said that a material misrepresentation is one which impairs the ability of "the Patent Office fairly to assess [the patent] application against the prevailing statutory criteria." Standard Oil Co. v. Montedison, S.p.A., 494 F. Supp. 370, 437 (D.Del.1980), aff'd, 664 F.2d 356 (3d Cir. 1981), quoting Monsanto Co. v. Rohm & Haas Co., 456 F.2d 592 (3d Cir. 1972); In re Frost Patent, supra, 540 F.2d note 9 at 604. The courts have applied a number of different specific tests of materiality, however, In the Frost case, Judge Wright analyzed what he referred to as the "objective but for" test, the "subjective but for" test and the "but it may have" test. 398 F. Supp. 1353, 1368-69 (D.Del.1975). The distinction among them is unimportant here. The significant feature of each standard for present purposes is that each asks whether the misrepresentation affected or may have affected the issuance of the patent, rather than whether it was simply relevant to an evaluation of a claim by the statutory criteria. 398 F. Supp. 1353, 1368-69 (D.Del. 1975), see also SCM Corp. v. RCA, 318 F. Supp. 433, 449 (S.D.N.Y.1970); CPC Int'l., Inc. v. Standard Brands, Inc., 385 F. Supp. 1057, 1069 (D.Del.1974).

In most cases, there will be no difference between these two questions. When an applicant undertakes to assume the risk of making misrepresentations to the PTO, it will normally be on a point which influences the final decision on patentability. Moreover, in jurisdictions, like Delaware, which require only a showing that the misrepresentation "may have" affected the issuance of the patent, all relevant information will be material information if the PTO has no occasion to explain the reasons for the issuance of the patent. Under the "might have" test uncertainty about whether the PTO relied upon the misinformation is resolved against the patentee. See Hercules, Inc. v. Exxon Corp., supra; SCM v. RCA, supra. But there is a difference in this case. The file wrapper reflects the basis upon which the patents were issued and makes clear that the PTO, notwithstanding AHC's claims of superior impact strength, did not rely on improved impact strength as a basis for patentability. Thus, in this case, unlike Frost, RCA or CPC, there is no burden of uncertainty to be borne. By serendipity, AHC's misleading data regarding superior impact strength did not bear any fruit.

This conclusion still does not close the inquiry, however. There is authority both in this Court and elsewhere^[38] which suggests that wrongful misrepresentations to the PTO may make claims unenforceable even when they are not material. Judge Wright, for example, made the following observations some years ago in Corning Glass Works v. Anchor Hocking Glass Corp., 253 F. Supp. 461, 470 (D.Del.1966), rev'd in part on other grounds, 374 F.2d 473 (3d Cir. 1967), cert. denied, 389 U.S. 826, 88 S. Ct. 65, 19 L. Ed. 2d 80 (1967):

Even though misrepresentations made to the Patent Office are not legally material to the issuance of a patent, nevertheless, this Court, being a court of equity, can and should refuse to enforce the patent if the Court finds the patentee made intentional misrepresentations to the patent examiner, i.e., if the patentee came into court with unclean hands.

The doctrine of unenforceability to which Judge Wright here refers is rooted in the equitable concept of "unclean hands" and focuses on deterring misrepresentations which necessarily destroy the fiduciary relationship between the patentee and the examiner and undermine the validity of the findings upon which courts must rely. Corning Glass Works, supra, 253 F.Supp. at 470.^[39] 4 Chisum, Patents, § 19.03[6] at 1984.

In the years since Corning Glass, however, the Court of Appeals for the Third Circuit has applied the unenforceability doctrine narrowly. In re Frost Patent Litigation, supra, and Delong Corp. v. Raymond Int'l., Inc., 622 F.2d 1135, 1146, n. 10 (3d Cir. 1980) appear to teach that, in applying the "unclean hands" doctrine in cases of this kind, it is not sufficient that the patent holder's fraud relate to the same subject matter as that involved in his claim of infringement; rather there must be causal connection between the fraud and the particular patent claim upon which the patent holder is relying. While these cases are factually distinguishable, they suggest to me that the shadowy distinction between invalidity and unenforceability has effectively been erased in the Third Circuit and that both doctrines require a showing that the wrongful conduct may have had an impact on the decision of the PTO.^[40] This requirement, which limits application of these doctrines to cases in which there are ill gotten gains to be stripped from the patent holder, does not significantly diminish their deterrent effect since, during prosecution, applicants must assume that any relevant information will turn out to be material.

Since it is clear that AHC's manipulation of Interim Report data on impact strength had no effect on the issuance of the patent, it may not serve Cosden as an effective defense.

C. The '982 French Patent.

While the '982 French patent, which issued prior to the first CIP application in 1961, was omitted from the 1961 oath and its filing date was misrepresented in the 1965 oath, these breaches of AHC's duty to the PTO are not material. Since the claims of the '982 French patent are co-extensive with those of the 1958 U. S. application, the French patent is not anticipatory of any invention which is not entitled to a 1958 filing date.

D. 1962-66: Emulsion Polybutadiene Test Results And Other Comparative Data Submitted To The PTO.

I am not persuaded that AHC misrepresented to the PTO its experience with emulsion polybutadiene during the early 1960s.

E. 1968: The Finestone Affidavit And Alleged Suppression Of The Finestone Japanese Affidavit.

Cosden contends that an affidavit filed in the PTO omitted relevant data which had been contained in an affidavit of the affiant previously filed in the Japanese patent office and thereby misled the PTO with respect to the improvement in impact strength provided by the compositions of the invention. There is no evidence to support this charge of fraud. While the impact values for the claimed compositions obtained in the experiments reported in the Japanese affidavit are somewhat lower than those submitted in the course of the prosecution of the patents in suit, the reported experiments nevertheless show a substantial improvement over the control SBR results obtained in those experiments. Moreover, given the specific purpose of the Finestone United States affidavit and the office action to which it was addressed, it was entirely appropriate to report the data reported and to omit the remaining data from the Japanese affidavit.

F. 1967-1974: Creep Yield And Gloss Data.

The 1967 CIP contains the first references to improved creep and gloss. It states:

In the following examples, comparative data are shown for varying amounts of vinyl aeromatic monomers and rubbery polymers. It will be noted that where an equivalent amount of high cis content, low vinyl content polybutadiene is substituted for conventional rubbery butadiene-styrene copolymer or emulsion polybutadiene, the impact strengths of the resultant products are greatly increased and that where intermediate amounts are employed, an intermediate increase in impact strength results.

Other advantages besides enhanced impact values at various temperatures have been surprisingly observed for the impact polymer compositions provided hereby. For example, objects have been molded from the compositions having improved surfaces, e.g., both as to smoothness and gloss. It has been observed the parts can be provided using these impact compositions which show significant enhancement of heat distortion temperatures, creep rupture times and flex modulus values, without sacrifice in impact values.

The only data initially submitted in support of these claims with respect to creep and gloss properties was an affidavit by Dr. Blanchette, an AHC employee, which compared the creep yield and gloss of a polystyrene product of a single run using Diene (35.9% cis, 7.8% vinyl) with those of a polystyrene product produced with prior art emulsion polybutadiene (20% cis, 18% vinyl). The affidavit showed a creep yield time for the former of 1400 minutes and 60 minutes for the latter.

In February of 1968, another Blanchette affidavit was filed which compared the gloss values of three compositions having a cis content of about 98% and a vinyl content of about 1-2%, with the gloss value of two emulsion polybutadiene products. The values shown for the former were between 30 and 36; the values for the latter were about 17.

As previously noted, the Board of Appeals, in 1970, affirmed the examiner's rejection of all claims on the ground that the claimed invention was prima facie obvious and that, while surprisingly improved properties of impact strength, creep and gloss were claimed, the record did not provide sufficient support for these claims. In the course of its opinion the Board commented as follows:

Having reached the conclusion that the combination of references is a proper one, rendering the appealed subject matter prima facie obvious, we turn to the matter of unexpected properties, vigorously advanced by appellants, relying on most of the affidavits listed above. The characteristic which is emphasized by appellants, and to which the major portion of the showing is directed, is impact strength.

... We will make a brief summary of the evidence to show that the compositions here claimed are not possessed of impact strengths unattainable with other polybutadienes. We do not consider it necessary to discuss the effects of other modifiers, such as SBR (GR-S) or polyisoprene, because the salient point is the unobviousness of the specified group of polybutadienes over other polybutadienes.

With respect to creep and gloss properties and the Blanchette affidavit, the Board found as follows:

In the creep yield point test and the gloss value test a single composition within the claimed range was compared with a single composition outside the claimed range.... There is no assurance from these extremely limited tests that the compositions encompassed by the appealed claims will generally exhibit the same order of superiority over polystyrenes modified by other polybutadienes.

In response to the Board's opinion, AHC filed an affidavit of Professor D'Alelio which adopted the two figures compared in the Blanchette affidavit and added creep yield point data for two additional runs, run A using Solprene and run B using Taktene. The affidavit provided the following data and reached the following conclusion:

The creep yield point values can then be summarized:

TABLE B
Creep Yield
Polymer                                 Reference for Data           Point (min.)
Prior Art
Interpolymer Styrene-Polybutadiene      Blanchette Affidavit              60
(20% cis;                               No. 2 "VI"
18% 1,2-addition)
---------------------------------------------------------------------------------
Westphal et al.
Interpolymer Styrene-Polybutadiene      Blanchette Affidavit
(35.9% cis;                             No. 2 ("VI")                    1400
7.8% 1,2-addition)

Interpolymer Styrene-Polybutadiene      Present Affidavit
(40-42% cis;                            (Run A)                         8040[41]
8-9%, 1,2-addition)
Interpolymer Styrene-Polybutadiene      Present Affidavit
(95% cis;                               (Run B)                          315
2-3% 1,2-addition)

Knowing the creep yield point of the prior art styrenepolybutadiene interpolymer (20% cis, 18% 1,2-addition) to be 60 minutes, I find it totally unexpected that the creep yield point of the styrene-polybutadiene interpolymers of Westphal et al. would be 5.25 (315/60), 23.3 (1400/60), and 134 (8,040/60) times that of the prior art interpolymer. To my knowledge, there is no way in which these superior results could have been predicted.

In my opinion, the interpolymers of Runs A and B, together with the interpolymer described in the Joseph A. Blanchette Affidavit No. 2 (Affidavit "VI") are definitely representative of interpolymers coming within the Westphal et al. claims and are sufficient to establish to my satisfaction the superiority of the narrow range of the Westphal et al. interpolymers over the interpolymers of the prior art.

The D'Alelio affidavit took the Blanchette gloss value figures and compared them with the gloss values for two other polystyrene compositions made with medium high cis polybutadienes. The figures reported for the two new compositions were 35 and 36.

Based on the creep yield data reported in the D'Alelio affidavit and the comparative gloss data there set forth, the Board concluded that the claimed composition had surprisingly improved creep and gloss properties and, under the authority of In re Davies, caused the patent to be issued.

During discovery in this action, Cosden unearthed evidence that, beginning in mid-1967, and continuing up to the filing of the D'Alelio affidavit in 1970, AHC had conducted more extensive testing than that reflected in that affidavit. Specifically, in each of the series of tests done by AHC, it tested SBR as well as emulsion polybutadiene against compositions within the claimed range. In addition, it had conducted this comparison testing at varying load levels, varying rubber content levels, and with and without plasticizing additives. Based upon this evidence, Cosden makes two claims. First, it is said that AHC fraudulently misled the PTO by concealing the fact that polystyrene produced with SBR often had superior creep characteristics to polystyrene produced with Taktene. Second, it is claimed that AHC fraudulently withheld testing results suggesting that the amount of rubber, amount of load, and the presence or absence of plasticizer make a difference in creep properties and that under some conditions, emulsion polybutadiene produces better creep qualities than polybutadienes within the claimed range.

AHC acknowledges that it realized it had data showing that, under at least some circumstances, SBR polystyrene had superior creep qualities to Taktene polystyrene. Its response to Cosden's first claim is that the only point of the challenged affidavits was to compare the creep and gloss properties of compositions within the claimed range with those of the compositions which the Board of Appeals had declared to be the closest prior art — emulsion polybutadiene, and indeed, that this is the only comparison material to the question of patentability. Cosden agrees that one in AHC's position was required only to test against the closest prior art but insists that, if testing is done against other relevant prior art, unhelpful test results resulting from that testing may not be suppressed. It stresses that the patented composition is an impact polystyrene which is claimed by AHC to be superior to prior art polystyrenes including SBR produced polystyrene. It points to the fact that throughout the file wrapper AHC claims superiority over SBR produced polystyrene, that the D'Alelio affidavit expressly claims superiority with respect to creep properties "over the interpolymers of the prior art", and that AHC specifically included SBR in its creep yield testing conducted for patent purposes.

While AHC's earlier insistence before the PTO of superiority over SBR impact polystyrene gives surface appeal to Cosden's claim regarding the omission of the SBR comparative data, I conclude that this argument is logically flawed. The materiality of the omitted data depends upon whether that data could reasonably be expected to be of help to the PTO in applying the statutory criteria to the alleged invention. Here the statutory criteria which Cosden claims makes the omitted data material is that of Section 103, non-obviousness. The creep yield performance of SBR produced polystyrene is simply not helpful on that issue, however. As the Board of Appeals put it "the salient point is the unobviousness of the specified group of polybutadienes over other polybutadienes."^[42]

Moreover, even if there were some basis in this record for suggesting that the creep yield properties of SBR impact polystyrene would be helpful in deciding whether one skilled in the art would have expected high cis, low vinyl impact polystyrene to have creep yield performance of the quality which is alleged to have been found, the PTO had clearly expressed its view that the SBR data was not relevant to a determination of obviousness and, ultimately, issued the patent without any analysis of the data comparing SBR impact polystyrene with the compositions of the invention. In this context, I cannot conclude that the failure to supply the SBR impact polystyrene data was an intentional effort to mislead the PTO or that it constituted gross negligence. In re Coordinated Pretrial Proceedings in Antibiotic Antitrust Actions, 676 F.2d 51 (3d Cir. 1982).

Cosden's second argument is of more weight. A comparison of the creep yield characteristics of emulsion reinforced polystyrene and compositions of the invention throughout the claimed range was of crucial importance to the PTO. The D'Alelio affidavit reported Taktene reinforced polystyrene to have a creep yield time 5.25 times that of emulsion polybutadiene reinforced polystyrene; Diene reinforced polystyrene was reported to be 23.5 times better. Clearly, this was material data which did influence the decision of the PTO to issue the patents.

When one examines all of the creep yield data possessed by AHC in 1970, the impression received is quite different from that conveyed by the D'Alelio affidavit.^[43] When one adjusts for duplication, AHC's testing reports indicate that it had a total of eight test series which were designed, inter alia, to provide a comparison between the creep yield characteristics of Taktene reinforced polystyrene and those of emulsion reinforced polystyrene.^[44] Four of these series, none of which was reported to the PTO, were run at a 5% rubber level. While reasonable persons attempting to read creep yield points may disagree about precisely where the yield occurs, allowing for such good faith dispute, it is fair to say that two of these four tests provide support for the proposition that Taktene is superior to emulsion polybutadiene. The other two showed substantially equal performance. Of the four tests run at a 6½% rubber level, one, utilizing plasticizer, also showed the creep yield characteristics to be about the same.

There were also eight test series which permitted a comparison of Diene reinforced polystyrene with emulsion reinforced polystyrene. Two of the three runs at a 6½% rubber level showed the Diene product to be superior. One test at that level showed substantially equal performance. Of the five series run at the 5% rubber level, only two support the proposition that the Diene product was substantially better. The remaining three, all run at different stress levels, showed the two compositions to be substantially equal in performance.

AHC acknowledges that these testing data were available to it at the time of the submission of the Blanchette and D'Alelio affidavits. The testing was supervised by its in-house patent counsel, L. G. Sinn, and it is apparent that a deliberate decision was made to report that which was reported and to withhold that which was withheld. AHC insists that it acted in good faith, however, and in support of this proposition, relies on two segments of the record. First, it notes that Dr. Bucknall himself questioned the reliability of the withheld data; second, it refers to a contemporary document, the Hanna Report of August 1, 1968, which is said to demonstrate that AHC's interpretation of its testing data was consistent with the information provided to the PTO.

Both sides apparently agree that there were problems with AHC's in-house testing of creep yield points. This fact provides no justification for AHC's conduct, however. First, it is apparent from the Hanna Report that AHC did regard these test results as having some value. More importantly, however, the creep yield data furnished to the PTO in the Blanchette affidavit came from the same in-house testing which produced the withheld data. AHC cannot treat the favorable results of its in-house testing as valid and the unfavorable and inconclusive results of that testing as invalid. As with AHC's use of impact data, I stress again that this is not a situation in which an applicant has a substantial amount of test data, submits to the PTO data which are fairly representative of those tests, and fails to report several aberrational data points. AHC had a limited amount of data which did not consistently support the proposition it was urging before the PTO. In these circumstances, it should have disclosed all of its in-house testing data for the PTO's consideration or none.

Nor do I find anything to aid AHC in the Hanna Report. While Hanna did express the opinion that Diene reinforced polystyrene was "superior in overall creep resistance," to polystyrene reinforced with Taktene, emulsion polybutadiene, or SBR, it expresses no opinion with respect to the relative merits of Taktene and emulsion polybutadiene and fails to provide any principled basis for distinguishing between the reported and the withheld data. In particular, it offers no rationale for reporting some of the 6% rubber level tests and withholding all 5% rubber level test results.

This case is virtually indistinguishable from In re Frost Patent, supra, in which the patent applicant withheld certain experimental data from the examiner while submitting other results which supported his claim. The Third Circuit observed:

General's attorneys knew of the Braidich data. It could be contended that the Braidich data was inconclusive, 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 would have given General the opportunity to present additional, more conclusive data.

540 F.2d at 609-610. Instead, General deprived the patent office of the opportunity to make an informed assessment of the evidence before it. AHC's selective presentation of creep yield data had the same effect.

By carefully selecting the tests results it put forward, AHC effectively prevented the patent office from applying its expertise to an evaluation of inconclusive scientific evidence. The effect is to deprive this Court and the public of the screening process which the law requires before granting a patent monopoly. Because it was done with that deliberate intention in mind, AHC's inequitable conduct renders the patents-in-suit unenforceable. See Monsanto v. Rohm & Haas, supra, 456 F.2d at 600.

I find no misconduct with respect to the presentation of gloss values. I note, however, that the PTO, as well apparently as AHC, had very little data upon which to determine whether compositions encompassed by the claims-in-suit generally exhibit a marked superiority in gloss over polystyrenes modified by other polybutadienes. As earlier noted, I consider this fact to be relevant to a determination of whether Cosden carried its burden on the Section 103 issues.

G. 1973: Suppression Of The Diene Data In The Westphal Rule 131 Affidavit.

Given the purpose of Westphal's Rule 131 affidavit, I find no impropriety in the failure to disclose to the patent office the material deleted from the exhibits to that affidavit.

V. CONCLUSION.

For the foregoing reasons, I conclude that (1) Cosden has produced and sold high impact polystyrene which would infringe certain of the claims-in-suit if those claims were valid, (2) the claims-in-suit which include the cis content range of 95% to 100% are invalid under Section 102(g), (3) all of the claims-in-suit are invalid under Sections 103 and 112, (4) all of the claims-in-suit are unenforceable as a result of AHC's conduct before the PTO, and (5) this case is "exceptional" within the meaning of 35 U.S.C. § 285.

SUPPLEMENTAL OPINION

In an opinion of March 15, 1982, this Court recorded its findings of fact and conclusions of law following a trial on the merits of this controversy. AHC has now moved to amend those findings and conclusions pursuant to Rule 52(b). The focus of the motion is on the findings concerning AHC's submission of test results to the PTO in support of its claim that Taktene and Diene are remarkably superior to emulsion polybutadiene with respect to creep yield.

AHC made the strong and unqualified claim before the PTO that Taktene reinforced polystyrene had been shown to have a creep yield time 5.25 times that of emulsion butadiene reinforced polystyrene and that Diene reinforced polystyrene had been shown to be 23.5 times better. In my original opinion, based on AHC's contemporary documents and Dr. Bucknall's testimony, I found that this claim was not representative of AHC's testing results which were, in fact, inconsistent and inconclusive. Moreover, I concluded as follows:

By carefully selecting the test results it put forward, AHC effectively prevented the patent office from applying its expertise to an evaluation of inconclusive scientific evidence. The effect is to deprive this Court and the public of the screening process which the law requires before granting a patent monopoly. Because it was done with that deliberate intention in mind, AHC's inequitable conduct renders the patents-in-suit unenforceable. See Monsanto v. Rohm & Haas, supra, 456 F.2d at 600.

Based on the foregoing findings of fact, I concluded, as a matter of law, that this was an "exceptional" case within the meaning of 35 U.S.C. § 285.

In support of its Rule 52(b) motion, AHC urges that the foregoing findings of fact are clearly erroneous because the evidence shows that the "meaningful" test data, rather than being inconsistent or inconclusive, strongly supports AHC's claim before the PTO. One can take this view, however, only if one treats as "unmeaningful" all tests in which the yield points occurred within 60 minutes or less. AHC urges this view based upon the testimony of expert witnesses O'Toole and Bucknall that creep yield testing can take "hundreds of hours" and upon the assertion of AHC's counsel "that tests which do not result in creep yield times of more than 60 minutes cannot be employed to establish valid creep yield comparisons between different impact polystyrene compositions." See proposed amended finding of fact AF-4. This new theory of AHC's is not supported by the record.

In its presentation to the PTO, AHC presented a single set of data in support of its unqualified claim of Taktene's superiority over emulsion polybutadiene. It showed a creep yield of 60 minutes for the emulsion polybutadiene and 315 minutes for the Taktene. The evidence relied upon by AHC will simply not support the view that 60 minutes versus 315 minutes is meaningful to the point of being conclusive while 60 minutes versus 60 minutes is meaningless. Indeed, it is apparent from both O'Toole's testimony and that of Bucknall that they did not regard the 60 minute data as meaningless.

No one connected with the selection of the data for presentation to the patent office testified that he or she considered the short term data meaningless and disregarded it for that reason. Similarly, the contemporary documentation does not reflect that AHC went through any such evaluative process. In particular, the Hanna Report, relied upon so heavily by AHC, does not treat the short term testing data as meaningless.

It is further interesting to note that Dr. Bucknall, in analyzing the TNO and Labofina test data, attributed significance to tests producing yield points at 60 minutes and below and he was not even cross-examined by AHC upon this aspect of his testimony.

Finally, I note that AHC's new theory is not reflected in its post-trial briefing. In that briefing, it finds creep yield times of 60 minutes for emulsion polybutadiene and 200 minutes for Taktene to be strong evidence of its claim before the PTO. Indeed, its reply brief draws inferences from Figure XII of Cosden's Exhibit 750 which is a graph reflecting a test with creep yield times below 60 minutes. (AHC Reply Br. p. 100).

Finally, AHC argues that, even if the data which it submitted to the PTO is not representative of its testing as a whole, its conduct does not rise to the level of the kind of fraud which properly occasions an award of counsel fees. I need not address here the issue of whether there are various degrees of fraud on the PTO some of which do, and some of which do not, justify counsel fees. This is not a case of neglect, gross or otherwise; I simply do not believe that anyone at AHC had a subjective evaluation of its testing results which would be consistent with the strong and unqualified claims made before the PTO. Moreover, as stressed in the original opinion, the selection of this test data was only one part of a course of conduct throughout the prosecution of the patent which was calculated to mislead and which makes an award of counsel fees in this case highly appropriate.

NOTES

[1] Most of the relevant facts predate the acquisition of Foster Grant by American Hoechst. "AHC" as used herein refers to Foster Grant prior to that acquisition and American Hoechst thereafter.

[2] The existence of the gels impaired the end products' aesthetic value and often weakened them as well. The "Amos Patent", U.S. Patent No. 2,694,692, issued on November 16, 1954.

[3] This rubber was developed during the Government Reserve Rubber Program as a substitute for natural rubber and thus was originally called "GR-S".

[4] The Izod Impact Test, which was standardized as an ASTM specification, subjects a standard notched test bar to a measured impact in an Izod Impact Test machine and the values obtained are recorded in terms of the resistance of the bar to a free-swinging arm as foot pounds per inch of notch (ft. lb/in. notch).

[5] At trial Heinig acknowledged that he so testified at his deposition. While he later recanted his deposition testimony, I accept his original version of the events of March 1958.

[6] Firestone's representative, Mr. John Musch, testified that these microstructure values published by Firestone for Diene polybutadiene are accurate.

[7] I am unpersuaded, however, that this material infringes claim 17 of the '434 patent which calls for "a 1,2-addition content of 3.3 percent."

[8] For this reason, I conclude that the Diene polybutadiene used by Cosden has a cis content of "about 35%" as that phrase is used in claim 10 of the '434 patent and claims 4 and 18 of the '311 patent. Similarly, even if I were persuaded by Dr. Koenig's work that Diene had a vinyl content of 10.8%, I would nevertheless conclude that its vinyl content is "not more than about 10%". AHC did argue during prosecution of the '434 patent that 17%, 18% and even 13% vinyl content in prior art was in excess of the claims of the '434. But this does not estop it from contending that "about 10%" includes as high as 10.8%.

[9] In order to make sure one was not infringing in a field where technological change is swift, for example, one would have to maintain "old art" as well as "new art" laboratories.

[10] See National Research Devel. v. Great Lakes Carbon, 410 F. Supp. 1108, 1116-1117 (D.Del. 1975).

[11] Contrary to Cosden's contention, Dr. Koenig's NMR analysis did not confirm the purity of the polybutadiene.

[12] I.e., Claims 1, 5, 8, 11, 13, 14 and 15 of the '434 patent and Claims 1, 2, 7, 8, 13, 14, 15, 16, 21, 22 and 27 of the '311 patent.

[13] Dr. Finestone reviewed the 1958 application prior to the filing of the 1961 CIP and complained that the cis content range of 25-95% is "1) too broad and 2) does not cover the desired higher limits...."

[14] While there is one claim in the 1958 patent application which Cosden describes as a "process claim", there are also composition claims which could support the claims of the patents in suit, even if Cosden's characterization be accepted.

[15] A patent with such a claim nevertheless permits one skilled in the art to obtain superior results by practicing the invention. By contrast, overbroad claims require the practitioner to conduct additional experiments before realizing the purported benefits of the invention. See pp. 543-544, infra (discussing the latter issue in relation to obviousness).

[16] This is as true in infringement actions as in an interference. Grefco, Inc. v. Kewanee Industries, Inc., supra; 1 Chisum, Patents § 3.08[3].

[17] I do not accept AHC's contention that the 1957-58 winter work at Shell is irrelevant because there was no agitation. While this was true of the November runs, it was not true of the January and February runs. Moreover, I am persuaded that by this point in time Amos's teachings were well known and the use of agitation was a matter of choice for the artisan of ordinary skill in the art.

[18] There is no persuasive record support for Cosden's argument that Shell appreciated the improved properties but failed to so note in its contemporaneous documents because very high cis impact polystyrene was not then commercially feasible.

[19] See discussion 3 Chisum, Patents § 10.04[6]. The Court of Customs and Patent Appeals has answered in the affirmative.

[20] The filing, prosecution and ultimate issuance of foreign applications, based upon the January, 1960 U.S. application, in Germany, Canada, Australia, Italy, The Netherlands, Japan, Great Britain, India, France, and Belgium, also belie the existence of any intent to abandon the work. The Belgium patent was published as early as July, 1961.

[21] U.S. Patent No. 2,694,692, issued November 16, 1954.

[22] See, e.g., Stoops, et al., U.S. Patent No. 2,754,282, filed January 15, 1953.

[23] Smith, et al., U.S. Patent No. 3,178,402, filed April 16, 1956.

[24] Foster, U.S. Patent No. 3,317,918, filed August 21, 1956.

[25] J. M. Short, V. Thornton and G. Kraus, Effect of Cis-Trans Ratio on the Physical Properties of 1,4 Polybutadienes. (AHC 465)

[26] AHC 465, pp. 1139-1140.

[27] AHC 465, p. 1130.

[28] Trends in Styrene Plastics, 32 Modern Plastics 154 (1955).

[29] Id. at p. 156.

[30] U.S. Patent No. 3,151,184, filed on March 27, 1956.

[31] E.g., M. Stafford Thompson, Gum Plastics (1958) at 61, 67 (COS 616).

When the resin monomer is polymerized in the presence of rubber the resin chain may become attached through a chemical bond to the rubber. The term "graft polymer" has been used to describe such a product because the resin molecule is "grafted" to the rubber molecule forming a gum plastic macromolecule. This type of polymerization makes possible combinations of chemically unrelated resins and rubbers by substituting a chemical bond for the physical attraction necessary to obtain impact strength in a simple blend plastic.

* * * * * *

Graft polymerization is a technique which allows efficient use of rubbers, such as polybutadiene, as an impact modifier in the absence of chemical relationship with the resin phase. A rubber such as polybutadiene has a high degree of chemical unsaturation providing points on the molecule where a resin chain can be chemically attached. Polymerization of the resin chains in the presence of this already polymerized rubber, enables the chains to attach themselves to the rubber. This yields a macromolecule with its separate resin and rubber phases which possesses the necessary adhesion of rubber to resin for developing shock resistance....

[32] U.S. Patent No. 2,754,282 (COS 556).

[33] U.S. Patent No. 3,151,184 (COS 577).

[34] 2 Chisum, Patents § 5.05.

[35] AHC maintains that the final decision of the Board of Appeals did not conclude that one of ordinary skill would expect significantly improved impact strength based on the prior art. The Board's opinion is capable of that interpretation. For present purposes, however, it is sufficient to note that there is no PTO finding that improved impact strength was non-obvious and that the patent issued based upon determinations that the alleged creep and gloss improvements were non-obvious.

[36] While Dr. Wiles testified that errors were committed in determining the microstructure of some of the samples in the Cosden testing, nothing he said indicates to me that the errors were of a magnitude that would undermine the points demonstrated by that testing.

[37] AHC admitted to the PTO that its data for Examples 10-12 was erroneous by the filing of a proposed Certificate of Correction (COS 1301), and this was confirmed by Mr. DeGrandi's testimony. (DeGrandi 2845, 2554). Clearly it was misleading or potentially misleading to the PTO. The date of this effort was some sixteen years after the initial representation was made, and while Cosden was well in the middle of its discovery.

[38] Duplan v. Deering Milliken Corp., 444 F. Supp. 648, 732 (D.S.C.1977), aff'd, 594 F.2d 979 (4th Cir. 1978), cert. denied, 444 U.S. 1015, 100 S. Ct. 666, 62 L. Ed. 2d 645 (1979).

[39] This doctrine originated in Precision Instrument Mfg. Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806, 65 S. Ct. 993, 89 L. Ed. 1381 (1945). Precision Instrument held two patents unenforceable although the inequitable conduct involved only one of the patent applications. See 4 Chisum, Patents § 19.03[6] at 19-85. See also 35 U.S.C. § 253.

[40] See In re Frost Patent, supra, 398 F. Supp. 1367-68; see also Digital Equipment Corp. v. Diamond, 653 F.2d 701 (1st Cir. 1981).

[41] Cosden maintains that this figure misrepresents the results of AHC's testing. The determination of the creep yield point of this particular sample, however, is a matter about which reasonable minds could differ, and this figure will not provide a basis for any claim of intentional misrepresentation or gross neglect.

[42] Cosden is, of course, correct when it says that Section 103 requires an alleged invention to be unobvious from all of the prior art. But, in this context, the relevant Section 103 issue is whether one of ordinary skill in the art, looking to all of the prior art, would find it surprising that the claimed compositions exhibit the creep and gloss properties which they are claimed to exhibit. If, as AHC's affidavits indicated, those properties were startlingly better than polystyrenes made with other polybutadienes, the closest compositions structurally to the claimed ones, it is difficult to see how a comparison between the claimed compositions and SBR polystyrene would be helpful on the relevant issue. If anything, given the excellent creep and gloss of crystal polystyrene, if SBR polystyrene had better properties than emulsion polybutadiene polystyrene, it would make the alleged invention even more surprising than one would find it on the basis of the emulsion polybutadiene data alone. If the claimed compositions are novel and unobvious from the prior art, it is not necessary for the applicant to show that their properties are superior to all compositions known to the prior art.

[43] I do not understand Cosden to charge Professor D'Alelio with fraud or gross neglect. He caused the additional compositions to be prepared, had them tested by the United States Testing Company, and accurately reported the results. There is no reason to believe he was aware of any data other than his own and that reported in the Blanchette affidavits. Moreover, if one accepts the Blanchette data at face value as D'Alelio apparently did, there is no reason to question the good faith of his opinion that the reported compositions were representative of the compositions within the scope of the patent. The problem arises only when one puts the D'Alelio affidavit in the context of the other data known to AHC.

[44] COS Nos. 655, 392A, 750, 189 and 190 are the AHC documents reflecting the testing information which it had. Dr. Bucknall's interpretation of this data is displayed in chart form on AHC 521.