Opinion
Civ. A. No. 85-0459-R.
May 23, 1986.
James C. Roberts, Gary J. Spahn, Mays, Valentine, Davenport Moore, Richmond, Va., Dennis McInerney, George Wailand, P. Kevin Castel, Cahill, Gordon Reindel, New York City, Robert Hardy Falk, Hubbard, Thurman, Turner Tucker, Dallas, Tex., C. Frederick Leydig, Phillip H. Mayer, Charles S. Oslakovic, Leydig, Voit Mayer, Chicago, Ill., for plaintiffs.
Robert H. Patterson, Jr., Anne Marie Whittemore, McGuire, Woods Battle, Richmond, Va., Joseph M. Fitzpatrick, Fitzpatrick, Cella, Harper Scinto, New York City, for defendant.
FINDINGS OF FACT
JURISDICTION AND VENUE
1. Plaintiffs Akzo N.V. ("Akzo") and Aramide Maatschappij Vof. ("Aramide") instituted the present action by a complaint filed May 6, 1985, in the United States District Court for the Eastern District of Virginia, Richmond Division. Plaintiffs' complaint alleges that defendant, E.I. DuPont de Nemours Company ("DuPont"), has deliberately infringed U.S. Patent 4,308,374 ("the '374 patent"). Plaintiffs seek an injunction prohibiting infringement of the '374 patent by DuPont, an accounting of past damages, a determination that DuPont's alleged infringement was willful, and an award of attorneys' fees.
2. DuPont manufactures Kevlar fiber at its Spruance plant in Richmond, Virginia, using a solvent system which plaintiffs allege infringes the '374 patent.
3. DuPont has counterclaimed for a declaratory judgment that the '374 patent is unenforceable because of Akzo's fraud in its procurement, because the patent is invalid in view of the prior art, and because DuPont has not infringed the patent. DuPont also seeks an award of attorneys' fees by reason of Akzo's alleged fraud in the procurement of the '374 patent.
THE PARTIES
4. Plaintiff Akzo is a corporation organized and existing under the laws of The Netherlands and has its principal office and place of business in Arnhem, The Netherlands. (Stipulated to by parties).
5. Plaintiff Aramide is a partnership organized and existing under the laws of The Netherlands and has a place of business in Emmen, The Netherlands. (Stipulated to by parties).
6. Defendant DuPont is a corporation organized and existing under the laws of the State of Delaware, having its principal offices in Wilmington, Delaware, and its Spruance plant in Richmond, Virginia. (Stipulated to by parties).
7. DuPont makes Kevlar fiber at its Spruance plant using a solvent in the polymerization of PPD-T polymer which Akzo charges infringes the '374 patent. (Stipulated to by parties).
THE PATENT IN SUIT
8. U.S. Letters Patent No. 4,308,374 ("the '374 patent"), entitled "Process for the Preparation of Poly-p-phenyleneterephthalamide" was issued on December 29, 1981 to Akzo on an application by Leo Vollbracht and Teunis J. Veerman, filed November 15, 1979 under Serial Number 94,589. The Application Serial Number 94,589 is a continuation of Serial Number 915,585, filed June 15, 1978, which in turn is a continuation of Serial Number 658,750, filed February 17, 1976. An application for a patent for the same invention was filed by Akzo in the Netherlands on February 21, 1975 as Application Number 7,502,060. (Stipulated to by parties) (PTX-52, 374).
9. By an agreement dated March 4, 1983, Aramide acquired certain interests under the '374 patent, Akzo retaining responsibility to bring patent infringement actions against third parties. (Stipulated to by parties) (PTX-246, 247).
10. The invention for which the '374 patent was issued is a process by which poly-p-phenyleneterephthalamide [PPD-T] polymer having an inherent viscosity of at least 2.5 is made using the solvent N-methyl pyrrolidone ("NMP") containing at least 5% calcium chloride (caC12) salt. (PTX-374, DX-15, col. 1, lines 5-9). The patent states:
The present invention is based on the unexpected discovery that in mixtures of N-methyl pyrrolidone and at least 5 percent by weight of calcium chloride, there can be prepared poly-p-phenyleneterephthalamide having an inherent viscosity of 2.5 or higher. Such a property is unique considering that other combinations of an organic solvent and a salt, for instance as in the above-cited literature, only yield poly-p-phenyleneterephthalamide having a low inherent viscosity.
(PTX-375, DX-15, col. 2, lines 18-26).
11. All claims of the '374 patent require at least 5 percent by weight calcium chloride based on the weight of NMP. (Stipulated to by parties).
PRE-1975 DEVELOPMENT OF PPD-T SOLVENT PROCESS BY DUPONT
12. Generally, in order to produce commercial grade high-strength, low density fibers from aromatic polyamides, it is necessary to spin the fiber from a polymer of high molecular weight. High molecular weight of a polymer is indicated by relatively high inherent viscosity of the polymer. High inherent viscosity is desirable with respect to production of all polymers relevant to this action, including the polymer produced by the invention claimed in the patent in suit, i.e., PPD-T.
13. DuPont conducted research on aromatic polyamides at least as early as the mid-1950's.
14. In 1957, DuPont researchers discovered that among the amide-type solvents most effective in producing high molecular weight poly-p-phenylene isophthalamide ("MPD-I") were dimethylacetamide ("DMAc"), N-methylpyrrolidone (NMP), and hexamethylphosphoramide ("HMPA"). DuPont researchers also produced PPD-T polymer in a solution polymerization process using DMAc and HMPA. (PTX-357, 358).
15. In the mid-1960's, DuPont began research regarding polymerization of 1,4-B polymer (1,4 poly-p-benzamide or "Fiber B") spun from a solvent solution of NMP and calcium chloride or of NMP and lithium chloride ("LiCl"). (PTX-162).
16. In September and October 1966, Dr. Szucs, a DuPont scientist, reported that he had attempted to produce 1,4-B polymer in NMP but obtained a polymer of low inherent viscosity. The addition of calcium chloride or lithium chloride to the system "did not help." (PTX-163, 164).
17. In a DuPont memorandum dated October 19, 1966 it was reported: "Szucs has now definitely concluded that NMP is unsuitable for direct solution polymerization." (PTX-017).
18. By 1969, however, DuPont was producing 1,4-B polymer where polymerization was carried out in a solvent of NMP and lithium chloride. (DX-225).
19. By July 1969, DuPont targeted PPD-T rather than 1,4-B polymer as its principal candidate for commercial development of a low density, high strength fiber. At that time, PPD-T was produced in the laboratory from a 2:1 reaction mixture of HMPA and NMP. The polymer was of high inherent viscosity. (PTX-326).
20. From 1969 on, DuPont was aware that HMPA had toxic effects on humans. DuPont therefore began research to find substitute solvents for making PPD-T. (PTX-324).
21. At the time DuPont was initially investigating solvents for use in the manufacture of PPD-T in 1969-70, DuPont was concurrently investigating the polymerization of chloro-PPD-T, a polymer chemically closely related to PPD-T but of higher solubility. As a result of research in the field of chloro-PPD-T polymerization, it was known that chloro-PPD-T could be produced using either DMAc or NMP as the polymerization medium. (PTX-322, 323, 328).
22. In 1969, Dr. Fitzgerald of DuPont conducted experiments to polymerize chloro-PPD-T using DMAc/LiCl, DMAc/CaCl2, NMP/LiCl, or NMP/CaCl2 as the polymerization media. Dr. Fitzgerald concluded that to make chloro-PPD-T, DMAc/CaCl2 and NMP/CaCl2 were both "much worse than DMAc/LiCl or NMP/LiCl." Nevertheless, Dr. Fitzgerald achieved inherent viscosity in the range of 2.8-3.7 using DMAc/CaCl2 and NMP/CaCl2. (PTX-328).
23. Between 1969 and 1971, DuPont continued to search for a PPD-T solvent system not containing HMPA. During this time, it was determined that NMP by itself was not a suitable polymerization medium for PPD-T. (PTX-145, 165).
24. DuPont's first commercial unit, MDF-1 (Manufacturing Development Facility-1) used a solvent system of an HMPA/NMP mixture. (Heckert, Tr. 378).
25. When DuPont scaled up MDF-1 to use a continuous rather than a batch process, the solvent system used was 100% HMPA because DuPont desired to use one solvent in the continuous process. For polymerizing PPD-T, HMPA is a distinctly better solvent than NMP and therefore HMPA was chosen as the single solvent. (Heckert, Tr. at 378; Memeger Dep. p. 46).
26. Between the time in 1971 that 100% HMPA was selected as the solvent system to make PPD-T and the time in 1975 that HMPA was determined to be carcinogenic to animals, DuPont conducted virtually no research to develop an alternate solvent system for polymerizing PPD-T. (Schulz, Tr. at 438-39).
27. From 1971-74, considerable research was conducted at DuPont with respect to developing a vapor-phase polymerization system to produce PPD-T. Vapor-phase polymerization does not involve the use of an alternate solvent system, but in fact uses no solvent at all. (Schulz, Tr. at 438-41; PTX-371, 372).
28. DuPont's research into the use of vapor-phase polymerization was to determine the general applicability of vapor-phase polymerization to many polymer systems, including PPD-T. With respect to producing PPD-T, vapor-phase polymerization was determined not to be a viable alternative to the HMPA solvent system. (Schulz, Tr. at 438-41; PTX-371, 372).
29. At the time that DuPont decided to scale up its Kevlar process and to virtually discontinue research into alternate solvent systems, it was well known by DuPont researchers that there existed possible alternative solvent systems for the polymerization of PPD-T. The most promising of the systems that contained no HMPA were those composed of certain organic solvents, particularly NMP, DMAc, and TMU, which were combined with certain salts, particularly lithium chloride and calcium chloride. AKZO'S DEVELOPMENT OF THE NMP/CaCl2 SOLVENT SYSTEM
30. In the first half of 1971, Akzo researchers Vollbracht and Veerman investigated the solubility of PPD-T and PBA (1,4-B) in various solvents. Several different solvents were experimented with at that time, including NMP, DMAc, HMPA, TMU, and HMPA/NMP in a 2:1 ratio. In these experiments, the effect of the addition of the salts LiCl, CaCl2, and MgCl2 to these solvents also was investigated. Vollbracht and Veerman concluded that the addition of LiCl and CaCl2 caused some increase in the dissolving power of some of the solvents, LiCl generally causing a greater increase than CaCl2. (PTX-248).
31. On September 27, 1971, Dr. Vollbracht of Akzo recorded a discussion he had had with a former DuPont scientist, Dr. Szucs, who told Dr. Vollbracht of DuPont's early work on Fiber B (1,4-B polymer). Dr. Szucs revealed that polymerization of 1,4-B polymer was "carried out in N-methyl pyrrolidone (NMP); LiCl is used to keep the polymer in solution." (DX-225, p. 2, ¶ 11). Dr. Vollbracht concluded generally that Akzo's research, "begun not long ago, does appear to be on the correct path." (DX-225; cf. Vollbracht, Tr. at 201-02).
32. On November 9, 1971, Akzo distributed a report detailing the costs of polymerizing p-Nomex (i.e., PPD-T) in NMP/LiCl. The report noted the high cost of lithium chloride and stated that calcium chloride would be looked to as a possible replacement. (DX 312; cf. Vollbracht, Tr. at 203-07).
33. Veerman and others experimented through 1972 to investigate the solubility of PBA and PPD-T in various solvents and solvent mixtures, both with and without salt additives. In September, 1972, they reported that lithium chloride was the only salt found to be effective as a solvent aid. (PTX-255).
34. Akzo learned from DuPont patents that mixtures of HMPA/NMP were effective polymerization media for PPD-T. (DX-45 at p. 6). Akzo's own experiments revealed that a mixture of HMPA and NMP gave "the best results by far" for polymerizing PPD-T. (DX-234 at 2).
35. As early as February 1973, Akzo researchers decided that any solvent system using HMPA would be expensive and complicated, and would involve problems of procuring sufficient quantities of HMPA for commercial production of PPD-T. Akzo researchers therefore recommended that an alternative polymerization system be developed. (PTX-256; PTX-249; Vollbracht Tr. at p. 178).
36. As a result of the perceived problems in using HMPA, Akzo researchers at a January 9, 1974 meeting agreed to begin a program to find an acceptable polymerization solvent in which the use of HMPA was eliminated or at least reduced. At that meeting, discussion focused on the researchers' belief that without HMPA, only a low-molecular-weight product is formed. (PTX-261; Vollbracht Tr. at pp. 178-79).
37. At a working group meeting on PPD-T polymerization held February 6, 1974, Vollbracht and Veerman decided to investigate the solubility in various solvents of model substances and oligomers in an attempt to arrive at an alternate solvent. It was thought that since the solubility of the low molecular weight oligomers and of the model substances would be greater than that of PPD-T of high molecular weight, differences in the dissolving or swelling power of the solvents would then become more apparent. (PTX-362; DX-255 at 98-99; Vollbracht Tr. at 208-09).
38. Vollbracht and Veerman's first solubility test was conducted on February 13, 1974. The next day they tested NMP/CaCl2. (DX-220).
39. The solubility studies carried out by Vollbracht and Veerman were similar to those reported in the prior art of Federov, et al. The solvents tested by Vollbracht and Veerman included three of the four solvents discussed in the Federov article. The Akzo studies also tested the effect of adding CaCl2, a salt Federov reported as causing an increase in the inherent viscosity of polymers. Vollbracht and Veerman, however, tested the solvents using model compounds different from those used by Federov. (Compare DX-22 with PTX-265; see Vollbracht Tr. at 210-11).
40. At the time that these solubility studies were conducted in February 1974, Vollbracht and Veerman were unaware of Federov. (Vollbracht Tr. at 210-11).
41. At a working group meeting held on March 1, 1974, Veerman reported that the addition of CaCl2 improved the dissolving effect of some of the solvents, particularly NMP and DMAc. (PTX-265).
42. After further experimentation, Veerman and Vollbracht decided to concentrate on a solvent system of NMP/CaCl2. (PTX-250, 263, 266, 280).
43. By April 5, 1974, Vollbracht and Veerman had produced high molecular weight PPD-T using a solvent system of NMP and CaCl2. On April 5, 1974, Vollbracht and Veerman submitted an invention disclosure on the preparation of PPD-T in NMP/CaCl2. (PTX-264; Vollbracht Tr. at 208-10; stipulated to by parties).
44. Dr. Vollbracht later reported with respect to the solubility studies that "[t]he effect of LiCl has not been investigated. For practical purposes the use of the much cheaper CaCl2 is more attractive." (DX-45 at 11).
45. Vollbracht and Veerman continued to experiment with NMP/CaCl2 as a PPD-T polymerization solvent through November 1974. They found that higher monomer (PPD) concentrations require higher amounts of calcium chloride in order to maintain high inherent viscosity of the polymer. (PTX-268-71, 274, 277-79, 281-85, and 287-89).
46. By the fall of 1974, Akzo concluded that NMP/CaCl2 was a commercially feasible alternative to HMPA as a PPD-T polymerization solvent. (PTX-288).
47. On October 10, 1974, Akzo patent attorney Visseren proposed that an application be filed in the names of Vollbracht and Veerman on the use of NMP/CaCl2 as a PPD-T polymerization solvent. (PTX-286; stipulated to by parties).
48. On February 21, 1975, Akzo filed a patent application in the Netherlands entitled "Process for the preparation of poly-p-phenyleneterephthalamide" in which the process described used NMP/CaCl2 as the polymerization solvent for PPD-T. That application was accorded number 7502060. (PTX-252).
49. After development of the NMP/CaCl2 solvent system, Akzo continued to use the mixture of HMPA and NMP, disclosed in DuPont Blades patent, as Akzo's standard solvent for making PPD-T. In an August 1975 business plan to build its first pilot plant, Akzo indicated that it would use the HMPA/NMP solvent, although it indicated that "an alternative solvent system (NMP/CaCl2) [was] under development." (DX-172 at 15).
50. In a January 19, 1976 memorandum, Akzo patent attorney Visserin recommended that foreign counterpart patent applications based on Netherlands patent application 7502060 be filed in the United States, Germany, Great Britain, France, Italy, Japan, and the U.S.S.R. Visseren provided the following reasons for the proposal:
The present process is a good alternative to the known NMP-HMPT [NMP-HMPA] (hexamethylphosphortriamide) process, which has become very unattractive in view of a recent publication on the carcinogenic action of HMPT [HMPA]. Though the Dutch Patent Office has found a publication disclosing similar solvent-salt systems, from recent experiments at ARLA it must be concluded that the high molecular weight PPD-T required for Arenka fibre preparation only can be obtained in the system NMP/CaCl2. Accordingly, we believe that there is still a good chance of obtaining patent protection for the present process.
(PTX-251; stipulated to by parties).
DUPONT'S DEVELOPMENT OF THE NMP/CaCl2 SOLVENT SYSTEM
51. DuPont's Haskell Laboratory in September 1975 determined that HMPA produced nasal tumors in rats exposed to high concentrations of HMPA vapors. DuPont's Central Research and Development Department recommended at that time that HMPA be classified as an experimental carcinogen. (PTX-6, 8; DX-124).
52. DuPont immediately notified concerned agencies of the U.S. government, the scientific community and industry, including Akzo, of Haskell Laboratory's findings regarding the classification of HMPA as an experimental carcinogen. (DX-451; Weijland Tr. at 224-25; stipulated to by parties).
53. Upon discovering that HMPA was an experimental carcinogen, DuPont began research into several prospective alternatives to solvent systems using HMPA.
54. One of the areas DuPont decided to examine was the use of a nonsolvent system, particularly vapor phase polymerization. On September 22, 1975, DuPont authorized research into this area. (PTX-372).
55. DuPont researchers also investigated the use of solvent systems composed of non-amide-type solvents. (PTX-40, 84).
56. On September 26, 1975, eleven days after DuPont concluded that HMPA was an experimental carcinogen, Dr. Beer, a DuPont researcher, began his search for an alternate solvent system by targeting amide salt solvents of TMU, DMAc, and NMP as possible replacements for the HMPA solvent system. He identified LiCl, CaCl2, and MgCl2 as the salts he would use. (DX-91 at 9).
57. By the late 1960's, NMP and DMAc were well-known as strong amide solvents, and the addition of inorganic salts, such as lithium chloride and calcium chloride was known to greatly increase the solvating power of those organic solvents. (DX-20; Heckert Tr. at 389).
58. Dr. Beer also investigated numerous other solvent systems between September 1975 and January 1976. On December 1, 1975, Dr. Beer polymerized PPD-T of 4.6 inherent viscosity using a DMAc/LiCl system. (DX-91).
59. Dr. Katz, a DuPont researcher, began his search for an alternate solvent system in late September, 1975. He experimented primarily with NMP/LiCl and on December 10, 1975, polymerized PPD-T with an inherent viscosity of 5.64 using that solvent system. (PTX-75).
60. Dr. Miller, another DuPont researcher, began his search for an alternate solvent system in late October, 1975, concentrating on the use of NMP and tetrabutyl phosphonium chloride.
61. In late January, 1976, Dr. Zimmerman, Research Manager for Kevlar Research at the Carothers Laboratory, wrote a memorandum, copies of which were distributed to Dr. Miller and Dr. Katz, suggesting that calcium chloride be examined as an alternative to the more expensive lithium chloride salt used with NMP to polymerize PPD-T. R.A. Scheiderbauer, Technical Manager of DuPont's Industrial Fibers Division, also urged the development of calcium chloride against the wishes of the research team, which had been attaining high inherent viscosities with lithium chloride. (PTX-317, 319).
62. On January 30, 1976, Dr. Miller polymerized PPD-T using NMP/CaCl2, obtaining an inherent viscosity of 5.44. (PTX-058).
63. On February 2, 1976, Dr. Katz polymerized PPD-T using NMP/CaCl2, obtaining a maximum inherent viscosity of 1.14. (PTX-075). On February 10, 1976, he made PPD-T of 5.14 inherent viscosity using NMP/CaCl2.
64. By February 19, 1976, DuPont researchers had succeeded in making high molecular weight PPD-T polymer using an NMP/CaCl2 solvent system. DuPont decided to use an NMP/CaCl2 solvent system rather than an NMP/LiCl system in its continuous process plant because of economic and toxicological concerns involved with the NMP/LiCl system. (PTX-44, DX-111).
65. From the outset of DuPont's research for an alternate solvent system in the fall of 1975, it was known by DuPont researchers that both lithium chloride and calcium chloride were inorganic salts that, when added to certain amide solvents such as NMP and DMAc, could greatly assist in the polymerization of high molecular weight aromatic polyamides such as PPD-T. This knowledge was gained at least in part by research conducted by DuPont prior to its decision to use the HMPA solvent system.
66. DuPont suspended virtually all research into alternative solvent systems for polymerizing PPD-T once it decided to use the HMPA system in 1971. When the Alternate Solvent Task Force was formed in 1975 to find an alternative solvent to HMPA, Task Force researchers looked to prior amidetype solvent research conducted by DuPont before the HMPA solvent system was scaled up in 1971-72, and to prior art known in 1973 and earlier. It is this prior research which provided an advanced starting point for DuPont researchers and which enabled them to develop the alternative NMP/CaCl2 solvent system within 5 months.
67. Akzo developed its NMP/CaCl2 solvent system in less than three months. DuPont developed its NMP/CaCl2 solvent system in less than five months. Akzo and DuPont developed their respective systems independently of each other.
68. During its work to develop an alternate solvent system, DuPont repeatedly made high molecular weight PPD-T polymer using either NMP or DMAc with lithium chloride salt. (DX-94; Schulz Tr. at 421). DuPont in February 1976 also made PPD-T in DMAc containing amounts of calcium chloride both above and below the 5% level. (DX-92 at 124 and 142). Akzo's '374 patent therefore is incorrect when it states that "if use is made of combinations of a solvent and a salt which differ from those according to the invention, only polymers having a low inherent viscosity are obtained." (See DX-15, col. 6, lines 43-46).
69. During its work to develop an alternate solvent system, DuPont made high molecular weight PPD-T in NMP and calcium chloride above and below Akzo's claim limitation of "at least 5 percent." (DX-94 at 4). Akzo's '374 patent therefore is incorrect when it states that "use of percentages of calcium chloride that are outside the range of the present invention . . . only result in obtaining low values of the inherent viscosity." (DX-15, col. 6, lines 15-21).
70. In its work to develop an alternate solvent system, DuPont found that establishing the mole ratio of salt to PPD monomer was the key to obtaining high molecular weight PPD-T (DX-334; DX-96). This was a discovery previously made by Federov, et al. and reported in that prior art reference. DuPont's researchers, however, had not read Federov at that time. (DX-342, 345, 347).
71. A mole is a unit quantity in chemistry. It is the weight of a chemical substance in grams which corresponds to the sum of the atomic weights of all the atoms in the molecules. Each chemical substance has its own mole weight. (Stipulated to by parties).
PROSECUTION HISTORY OF THE '374 PATENT AND THE DUPONT APPLICATION
72. On February 17, 1976, Akzo filed its U.S. application under Serial No. 658,780 in the names of Vollbracht and Veerman.
73. The application was assigned to Examiner Harold Anderson of the U.S. Patent Office for examination. At a later date, Examiner Anderson also was assigned to examine the DuPont application. (Stipulated to by parties).
74. On June 1, 1976, DuPont filed U.S. application No. 691, 832, in the names of Miller and Katz, describing a process to polymerize PPD-T in NMP/CaCl2, where the solution contained 0.8 to 2 moles of calcium chloride per mole of PPD. (DX-125).
75. On or about September 1, 1976, DuPont filed a prior art letter with the Patent Office in the Miller and Katz application. The article by Federov, et al., "Synthesis of Polyterephthalamides of High Molecular Weight by Polycondensation in Amide-Salt Systems" (U.S.S.R., 1969) (hereinafter "Federov") was listed in the prior art letter. Nothing in that document suggested that Federov would adversely affect the patent application. (Hoes, Tr. at 310).
76. In the fall of 1976, DuPont saw Akzo's published German counterpart application. That application revealed to DuPont that Akzo had an earlier priority date on its U.S. patent application. (Hoes, Tr. at 255).
77. On November 22, 1976, Examiner Anderson rejected all of Akzo's claims. The prior art rejection was primarily for obviousness under 35 U.S.C. § 103 and the references cited were an article by Chodkowski, et al., published in Chemical Abstracts, Vol. 77, 20083X (1977) (hereinafter "Chodkowski") in view of one of DuPont's patents to Kwolek U.S. 3,063,966 (hereinafter "Kwolek"). (PTX-225, pp. 20-22).
78. On March 23, 1977, Akzo filed an amendment to its application and an affidavit by Herman Weijland. In his affidavit, Mr. Weijland stated that he had conducted experiments to obtain fibers using the process described in the Chodkowski reference and the process set forth in the Akzo application. Mr. Weijland reported that the process described in the Chodkowski reference yielded PPD-T of much lower inherent viscosity and fiber of much lower tenacity than those obtained using the process described in the Vollbracht application.
79. On May 12, 1977, the Examiner issued an interference between the Akzo and DuPont applications. The Examiner proposed a claim directed to the NMP/CaCl2 polymerization system for PPD-T and stated that he considered the claim to be allowable according to the disclosure at that time. The communication further stated, "If the suggested claim is neither allowable nor supported by the instant disclosure, applicants should so state, setting forth the reasons therefor." (DX-130).
80. On June 7, 1977, Akzo filed an amendment entering the claim proposed by the Examiner into the Vollbracht, et al. application. (Stipulated to by parties).
81. Following receipt of the Patent Office interference, DuPont evaluated the prior art then known to it. Between the time it filed its patent application and the time that it received the interference communication, DuPont had found additional prior art not previously known to it, including the Asakura, et al. patent (U.S. Patent 3,966,686) (hereinafter "Asakura patent"). (DX-131, Hoes Tr. at 296-98).
82. After evaluating the prior art, DuPont concluded that if an interference were to go forward on the claim suggested by the Examiner, Akzo would prevail, but that the suggested claim was not patentable by anyone. DuPont's legal department stated that the prior art was close, but that had it known of all of the prior art before the application was filed, the application probably would not have been filed at all. (DX-379).
83. Based on this evaluation, DuPont on June 8, 1977, filed a letter in the Patent Office in response to the Examiner's request setting forth the reasons that it considered the suggested claim not allowable over the prior art and abandoning its patent application. DuPont listed six references to prior art:
(1) Federov, et al., "Synthesis of Polyterephthalamide of High Molecular Weight by Polycondensation in Amide-Salt Systems," Polymer Science, pp. 2479-2491 (1970);
(2) "Fiber Forming Aromatic Polyamides," Man-Made Fibers, Volume 2, p. 301 (1968) (hereinafter "FFAP");
(3) The Mitsubishi Rayon patent, Japanese Patent Appln. Publn. 74/122594;
(4) Smith, et al., U.S. Patent 3,354,125;
(5) Asakura, et al., U.S. Patent 3,966,686; and
(6) Chodkowski, et al., article, Polimery (Chodkowski).
84. On August 11, 1977, the Examiner rejected all claims, including the claim proposed for interference. The Examiner cited additional art, including the Asakura patent, Federov, and the Chimura patent application, Japanese Patent Appln. Publn. 74/122594 (hereinafter "Chimura patent application"). The Examiner stated that "the substance of the references is such that the claimed process as a whole is clearly obvious." (DX-17, tab 6).
85. Akzo responded on December 2, 1977 with an amendment in which it argued that none of the references disclosed the process of making PPD-T in NMP containing calcium chloride. (PTX-225, pp. 75-79).
86. The Examiner again rejected all of Akzo's claims on the ground of obviousness. The Examiner stated, "It is not necessary that each reference alone teach every claimed limitation of the instant claims. The rejection is based on obviousness, not anticipation." (PTX-225, p. 81).
87. Akzo filed a continuation application, Serial No. 915,585, on June 15, 1978. (PTX-226).
88. On November 13, 1978, the Examiner again rejected Akzo's claims, stating, "All of the claimed limitations are obvious from these references [Kwolek, Beste, Chodkowski, Asakura, Federov, or FFAP] . . . The claimed inert ingredients are few and where the choices are few, as here, mere selection of the optimum combination is clearly obvious." (PTX-226, p. 27).
89. On March 12, 1979, Akzo filed an amendment and an affidavit by Herman Weijland. In its amendment, Akzo's patent attorney reviewed each of the prior art references and asserted that none of the references would suggest that high molecular weight PPD-T could be polymerized in NMP/CaCl2. In particular, Akzo's attorneys believed that the processes found in the prior art references either produced polymer of low molecular weight or did not refer to PPD-T at all. The attorneys concluded, in part: "Applicant's claimed process, which uses the system MP-CaCl2 [NMP/CaCl2], is the first really good alternative to the dangerous HMPA." (PTX-226, pp. 44-46).
90. In his declaration filed March 12, 1979, Mr. Weijland stated that he had conducted the experiment described in Example 1 of the Chimura patent application and attained an inherent viscosity of .31. (PTX-226, pp. 31-32). He knew, however, that the Chimura patent contained a typographical error, and therefore he conducted a second experiment in which the amounts of PPD and TCl were reduced by a factor of 10 by weight, but the amount of lithium chloride salt remained the same. An inherent viscosity of .72 was attained with this ratio. (PTX-226, p. 32). Mr. Weijland did not reduce the salt content to achieve the 1:1 mole ratio of salt to monomer specified in Federov because he believed that such ratio would result in low inherent viscosity. He also believed that while the precise nature of the error in Chimura was not known, it was possible that a higher level of salt would produce polymer of high viscosity, as is the case in the NMP/CaCl2 system.
91. On May 11, 1979, the Examiner again rejected all claims for obviousness based upon references previously relied on. The Examiner also stated that the processes described in Chimura and Asakura would produce polymers of high molecular weight. (PTX-226, pp. 63-65).
92. In an amendment filed on September 11, 1979, Akzo argued that the process described in its application was not obvious because of the criticality of forming the PPD-T in an NMP solution containing at least 5% by weight CaCl2. In that amendment, Akzo conceded that: "Federov, et al. discloses that a methyl pyrollidone (NMP) and calcium chloride solvent is suitable for the preparation of P-p-PTA [PPD]." Akzo argued, however, that Federov was distinguishable because it did not disclose the 5% claim limitation or that NMP/CaCl2 should be used as a substitute for HMPA because of the latter's harmful biological effects. (PTX-226, p. 77). An affidavit submitted by Mr. Weijland asserted that the Chimura process produced an inherent viscosity of only .31.
93. The Examiner again rejected the claims on September 27, 1979. (PTX-226, p. 79).
94. On December 29, 1979, Akzo filed a further continuation-in-part, Serial No. 94,589.
95. Before the Examiner could rule on the application, Akzo filed two preliminary amendments, a declaration of Mr. Weijland, and a translation of another Federov article. (PTX-227, pp. 63-82).
96. The second preliminary amendment asserted shortcomings of the Federov and Asakura processes and again argued that the only solvent system with "outstanding properties" was one using "a mixture of NMP and at least 5% CaCl2." (PTX-227, p. 67).
97. In Mr. Weijland's declaration dated November 15, 1979, he asserted that he reconstructed Examples 5 and 6 of the Asakura patent, but substituted calcium chloride for lithium chloride as set forth in the Examiner's rejection. The substitution of calcium chloride was done on a weight basis, rather than on a molar basis. This was done because it was Mr. Weijland's understanding that Federov directed that replacement of lithium chloride by calcium chloride should be conducted on a 1:1 weight basis. Such substitution reduced the molar ratio of salt to PPD by a factor of two and one-half. (PTX-227, pp. 68-70).
98. On June 9, 1980, applicants cited additional prior art to the Examiner which had been cited during the prosecution of the corresponding Dutch application. (PTX-227, pp. 83-85).
99. On June 25, 1980, the Examiner again rejected all of the pending claims on the art previously relied upon and additionally on the new Federov article. (PTX-227, pp. 86-88).
100. Akzo filed a Request for Reconsideration on October 23, 1980, arguing:
Applicants have found that it is possible to substitute N-methylpyrrolidone for hexamethylphosphoramide in the preparation of PPD-T having a viscosity of at least 2.5 only if the solvent contains at least 5% calcium chloride. This viscosity is required for spinning and hence this is an important feature on all the claims. (PTX-227, p. 96) (emphasis in original).
Akzo also argued specifically with respect to Asakura:
No one reading Asakura in the absence of applicants' present disclosure would learn that he should substitute N-methyl pyrrolidone containing calcium chloride in an amount of at least 5% for the heretofore commercially popular hexame-thylphosphoramide solvent.
* * * * * *
The references disclose various solvents as equivalents and not one suggests any reason for not using hexamethylphosphoramide or for selecting N-methyl pyrrolidone containing at least 5% calcium chloride. (Tab 21A, 21B, DX-19).
As regards Federov, Akzo again acknowledged that (Tab 21C, DX-19):
Federov et al disclose that a methyl pyrrolidone and calcium chloride solvent is suitable for the preparation of PPTA [PPD].
but Akzo argued that Federov did not mention the critical at least 5% calcium chloride (Tab 21C, DX-19);
However, this reference does not disclose that the N-methyl pyrrolidone should contain at least 5% calcium chloride or that it should be used as a substitute for hexamethylphosphoramide as a solvent to avoid undesirable biological effects. (Emphasis in original).
101. On January 8, 1981, the Examiner again rejected all of the claims for obviousness, but this time only on the basis of Asakura. The Examiner questioned how Akzo's claimed "`5 percent by . . . calcium chloride'" compared with Asakura's 20g. calcium chloride per 500 ml. NMP. (PTX-227, p. 112).
102. Akzo appealed on May 14, 1981, and filed its Brief on Appeal on May 28, 1981. (PTX-227, pp. 115-146).
103. Akzo argued in its Brief on Appeal that its process was capable of achieving high inherent viscosities, as opposed to the inherent viscosities of Asakura of 2 or less, because of "two new and unobvious steps:
a. use of the special combination NMP/CaCl2,
b. use of a CaCl2 content of 5% or more."
(PTX-227, pp. 134-5).
104. In its brief Akzo argued that Asakura was deficient because (Tab 29, DX-19):
No one reading Asakura without first having had the benefit of appellants' disclosure would learn from it that he should substitute N-methyl pyrrolidone containing calcium chloride in an amount of at least 5% for the heretofore commercially popular — but dangerous — hexamethylphosphoramide.
As regards the Examiner's earlier question (see Tab 22, DX-19):
"How does the claimed "5 percent by ___ calcium chloride compare with patentees 20 g. calcium chloride per 500 ml. N-methyl pyrrolidone?,"
Akzo responded by arguing (Tab 30, DX-19):
But even if one were to use NMP instead of dimethyl acetamide in Example 5 of Asakura, one still would not arrive at the process claimed here, because 20 g CaCl2 per 500 ml NMP would mean that the weight percentage of CaCl2, calculated on the NMP (specific weight 1.03 grams per ml), would amount to 3.88%. It should be noted that according to Table I, Run No. IIIc (page 12 of appellants' specification), a CaCl2 content of 4% by weight results in an inherent viscosity for the polymer of only 1.40.
105. After making certain minor amendments, the Examiner allowed the claims. The Notice of Allowance was issued on July 31, 1981. (PTX-227, p. 148; DX-19, Tab 32). No reasons were stated for finally allowing the claims.
106. Akzo's Vollbracht and Veerman patent issued as U.S. 4,308,374 on December 29, 1981.
107. In its patent prosecution, Akzo repeatedly stressed the significance of there being at least 5% calcium chloride in the solvent solution. This assertion was instrumental in the decision of the Patent Examiner to allow Akzo's claims.
108. At least as early as November 1976, Akzo was aware as a result of one of its technician's experiments that solvent solutions containing less than 5% calcium chloride were capable of producing PPD-T with an inherent viscosity of greater than 2.5. Those experiments found that a solution containing as low as 3.38% calcium chloride could produce PPD-T with an inherent viscosity as high as 4.80. (DX-47).
109. These test results were not revealed to the Patent Examiner at any point during the patent prosecution.
110. The test results were known by Mr. Visseren, Akzo's inside patent counsel, but not by Mr. Mosher, Akzo's U.S. patent attorney.
111. Mr. Visseren, in a handwritten note he made on the test report, acknowledged that the test results for the experiments using less than 5% calcium chloride "fall outside" the Akzo NMP/CaCl2 patent application.
112. Although PPD-T polymer of relatively high inherent viscosity can be produced with NMP/CaCl2 where there is less than 5% calcium chloride, there must be greater than 5% calcium chloride in the NMP/CaCl2 solvent solution in order to achieve high inherent viscosity at high polymer concentrations.
113. The most commercially viable NMP/CaCl2 solvent systems for producing PPD-T contain calcium chloride in proportions considerably higher than 5%. For example, DuPont's continuous polymerization system generally uses a solvent of NMP containing 9.1% calcium chloride; Akzo's batch system uses a solvent of NMP containing 12-13% calcium chloride.
114. Although there is no sharp dividing line in solvent properties just below or above 5% calcium chloride, Akzo selected the 5% claim recitation as a commercial limit because producing PPD-T with a solvent solution containing less than 5% calcium chloride does not produce enough polymer per volume of solution to be economically preferable. (Bailey, Tr. at 84-85).
THE DUPONT COMMERCIAL POLYMERIZATION PROCESS
115. After selecting NMP/CaCl2 for scale up in February 1976, DuPont spent the next two and a half years working to develop a continuous process for making high molecular weight PPD-T using NMP and CaCl2 (DX-121; Tr. Schulz at 424, 433-434).
116. DuPont chose to develop a continuous process because it would continuously make PPD-T polymer of good uniform molecular weight distribution at good yield and cost. A batch process, while convenient to work with in the laboratory, does not as readily lend itself to making high inherent viscosity PPD-T polymer with consistently good molecular weight distribution. (Tr. Schulz at 431-432). (Tr. Schulz at 428-429).
117. Akzo's commercial plant uses a batch process, not a continuous process. Akzo investigated the possibility of using a continuous process, but decided that research and development cost would be too high. (DX-173).
118. DuPont had problems making fiber with properties like those achieved with HMPA. Because DuPont had been manufacturing and selling Kevlar for several years, existing customers had developed applications for products with specifications tailored to the precise characteristics of Kevlar. As a result, DuPont had to make a fiber that was identical in all respects to that previously produced using HMPA and to satisfy its customers that it could be used successfully in a large number of different existing customer processes. (Stipulated to by parties; Tr. Schulz at 433-434).
119. DuPont found that the level of polymer "fines" must be controlled and made many other process improvements (DX-121 at 3). By March of 1977, fiber identical to that made using HMPA polymer was consistently made. It took substantially longer, however, to enable the DuPont customers to test and confirm the acceptability of the new yarn for their specific applications. (Stipulated to by parties; DX-121 at 3).
120. Construction of DuPont's commercial plant utilizing the NMP/CaCl2 continuous polymerization process began in 1980 and the plant commenced operation in October 1982. (Hoes Tr. at 273).
121. The continuous polymerization process which DuPont began using in 1982 at its Spruance plant in Richmond, Virginia generally uses approximately 9.1% CaCl2 in the premix with 6.4% PPD, thereby providing a CaCl2/PPD mole ratio of 1.4. Two commercial runs have used 4.4% CaCl2 in the premix along with 3.7% PPD, thereby providing a 1.2 CaCl2/PPD mole ratio. These runs employing 4.4% CaCl2 in the premix constituted a small proportion of the total production at the plant since it began operating in 1982.
122. The DuPont Spruance plant produces PPD-T having an inherent viscosity of 6.3. (Stipulated to by parties).
123. The manufacture of PPD-T at the Spruance plant comprises mixing PPD and TC in a "vigorous" manner. (Stipulated to by parties).
124. The polymerization solvent employed by DuPont in manufacturing PPD-T is "substantially anhydrous." (Stipulated to by parties).
125. The DuPont continuous process requires that at several different stages the temperature must be carefully regulated to prevent calcium chloride from precipitating out of solution.
126. DuPont is unaware whether at any stage the presence of its continuous polymerization process, there is any undissolved (solid) calcium chloride. (Schulz, Tr. at 444-446).
127. Dr. Edward Schulz's trial testimony with respect to the operation of DuPont's continuous polymerization process at the Spruance plant is incorporated herein by reference.
THE PRIOR ART AND OBVIOUSNESS
128. The level of ordinary skill in the art is a chemist, holding a doctoral degree, with practical experience in the polymerization of aramids. (Stipulated to by parties).
129. No prior art reference relied upon by DuPont specifies that a mixture of NMP and calcium chloride had been used as a polymerization medium to make PPD-T.
The Blades U.S. Patent No. 3,767,756
130. The Blades patent discloses how to make high molecular weight PPD-T polymer capable of being spun into fiber. That process includes vigorously mixing and reacting p-phenylenediamine (PPD) and terephthaloyl chloride (TCl) in a mixture of HMPA and NMP. When the reaction is complete, the resulting PPD-T polymer is separated from the solvent.
131. The Blades patent specifies that in order to obtain high molecular weight polymer, the reaction should take place under substantially anhydrous conditions.
132. The process described in Claim 2 of the '374 patent contains all of the elements of the process disclosed in the Blades patent with the exception of the choice of polymerization solvent. Claim 2 is typical of the claims made in the '374 patent.
133. Following the DuPont Blades patent, which disclosed the commercial technique for spinning Kevlar, there was a greater desire in the art to obtain PPD-T of higher inherent viscosity and the art more closely paid attention to the purity of the starting materials, the moisture content, and experimental technique.
134. The prior art reveals that at the time of the Blades patent the best solvent for polymerizing PPD-T was considered to be HMPA, but that NMP was considered to be the second best solvent.
135. Man-Made Fibers, published in 1968, disclosed the use of amide-salt solvent to produce high molecular weight aromatic polyamides:
The better solvents are amides such as dimethylacetamide (DMAc), N-methylpyrrolidone (NMP), and hexamethylphosphoramide (HMPA).
* * * * * *
For the solution polymerization of many aromatic polyamides there is no known organic solvent sufficiently powerful to keep the polymer in solution as its molecular weight builds up. The solvating power of many organic solvents is greatly increased, however, by the addition or inorganic salts such as lithium chloride and calcium chloride. Use of these mixed solvents has permitted preparation in solution of many high-molecular-weight polymers which have not been obtainable using organic solvents alone.
136. The non-HMPA organic solvents mentioned in Man-Made Fibers (i.e., NMP and DMAc), and the salts that Man-Made Fibers taught to combine with the solvents (i.e., lithium chloride and calcium chloride), were among the principal solvents and salts investigated by Akzo and by DuPont when those companies began their separate searches for an alternative to HMPA.
Antal, et al. British Patent No. 1,262,002 (Antal patent.)
137. The Antal British Patent, published on February 2, 1972, was not considered by the U.S. Patent Examiner during the prosecution of the '374 patent. The Antal patent discloses the preparation of the aramid polymer poly-p-benzamide, or "1,4-B," in amide solvents, including NMP and salts.
138. Poly-p-benzamide (1,4-B) is an aramid polymer similar to PPD-T polymer. 1,4-B polymer was the precursor to PPD-T.
139. Examples XIII through XVIII of the Antal patent demonstrate preparation of 1,4-B fiber in a spinning dope using either NMP or DMAc solvent and either lithium chloride or calcium chloride salt.
140. Although the Antal patent describes, inter alia, the use of NMP/CaCl2 and NMP/LiCl as spinning dopes, not as polymerization solvents, the patent teaches that those two mixtures are effective solvents for 1,4-B.
141. Example XX of the Antal patent discloses that amide-lithium chloride systems generally more effectively dissolve the polymer than amide-calcium chloride systems, but that calcium chloride is more effective than lithium chloride for increasing solvent power when the amide medium is NMP.
Federov et al. Publication ("Synthesis of Polyterephthalamide of High Molecular Weight by Polycondensation in Amide-Salt Systems," (U.S.S.R. 1989))
142. The Federov publication emphasizes the correlation between solvent power and the resulting polymer molecular weight. It also demonstrates the use of solubility studies on model compounds to predict the suitability of amide-salt mixtures as polymerization solvents.
143. Federov studied the effect of the better amide solvents on molecular weight in the synthesis of PPD-T and found that NMP is second only to HMPA in its effectiveness as a polymerization solvent.
144. Federov also studied the effect on PPD-T molecular weight of amide-salt solutions and found that lithium chloride and calcium chloride are clearly the better salts for increasing inherent viscosity, lithium chloride being preferable to calcium chloride.
145. Federov teaches that the mole ratio of salt to monomer is a key to making PPD-T of high inherent viscosity. In Federov, the suggested mole ratio is 1:1.
146. The level of PPD-T inherent viscosity was limited in the Federov study because the PPD monomer used did not receive preliminary purification, the mixture had a higher than desirable water content, or the process used was not as effective as other processes.
147. Federov polymerized PPD-T of highest inherent viscosity using a DMAc/LiCl solvent solution in which lithium chloride and PPD monomer were present in a 1:1 mole ratio.
Chimura Japanese Patent Application No. Sho 48/1973-36063
148. The Chimura publication, laid open on November 22, 1974 (Laid Open Publication No. Sho 49/1974-122594), describes a process for making PPD-T polymer of very high molecular weight.
149. Chimura indicates that very high molecular weight is desirable in polymerizing aromatic polyamides.
150. Chimura polymerized PPD-T using a process that followed the Federov publication teachings. Chimura polymerized PPD and TCl in a DMAc and lithium chloride system, using PPD and lithium chloride in approximately 1 to 1 proportions, as taught by Federov, and obtained PPD-T polymer having an inherent viscosity of greater than 5.
151. Chimura attained higher inherent viscosity than Federov using the Federov system because purer starting ingredients were used, the water content was lower, or both.
Asakura U.S. Patent No. 3,996,686
152. The Asakura patent issued on June 29, 1976 on a patent application filed December 24, 1974.
153. Asakura described the making of high molecular weight polymerization aramid polymers, including PPD-T, in amide-salt solvents:
Those polymers can be prepared by the low temperature solution polycondensation of p-phenylene diamine, p-amino benzoyl hydrazide and/or terephtyaloyl dihydrazide, and terephthaloyl chloride, or p-amino benzoyl chloride (or their halogenated or nitrated derivatives) in polar amide-type solvents. Examples of solvents are N, N-dimethyl-acetamide (DMAc), N-methyl-2-pyrrolidone (NMP), hexamethyphosphoramide, tetramethyl urea, etc. In order to increase the solubility and stability of prepared polymer solutions, an ionic inorganic salt, such as lithium chloride, magnesium chloride, calcium chloride, lithium nitrate, etc., can be added to the solvent. Organic tertiary amines, such as pyridine and dimethyl aniline, can also increase solubility and stability.
(Col. 3, lines 24-38).
154. In all six examples of the Asakura patent, either NMP or DMAc was used as a solvent and either calcium chloride or lithium chloride was used as a salt. In Example 5, Asakura made chloro-PPD-T using DMAc and calcium chloride and in Example 6 made chloro-PPD-T using NMP and lithium chloride.
155. Asakura does not indicate whether NMP/LiCl, NMP/CaCl2, DMAc/LiCl, or DMAc/CaCl2 is the preferred solvent for making PPD-T, but the examples in Asakura strongly suggest that at least one of those four systems is to be used.
General Findings With Respect to Obviousness
156. At the time that Akzo developed the solvent system set forth in the '374 patent, the prior art taught that aside from HMPA, the amide solvents NMP and DMAc were the preferred solvents for the polymerization of PPD-T. The same prior art disclosed calcium chloride and lithium chloride as salts which would increase the solvent power of those solvents.
157. The prior art as a whole in 1975 would indicate to anyone skilled in the art that to polymerize aromatic polyamides of high molecular weight without using HMPA, the best course would be to use a solvent system of DMAc or NMP in combination with calcium chloride or lithium chloride. This finding is borne out by the record evidence, which shows that Akzo and DuPont independently settled on the same solvent system within months of the respective times that they decided to stop using HMPA systems.
158. While there is no single prior art reference that puts NMP and calcium chloride together and carries out a PPD-T polymerization process, the prior art as a whole in 1975 taught that the combination of NMP solvent and calcium chloride salt to make high inherent viscosity PPD-T polymer was obvious. (Tr. Uhlmann at 525).
159. The "at least 5 percent" by weight of calcium chloride limitation set forth in the Akzo '374 patent claims is scientifically meaningless. There is no scientific basis that the relationship of "at least 5 percent" by weight calcium chloride based on the weight of NMP is necessary to obtain high molecular weight PPD-T. The "at least 5 percent" by weight of calcium chloride limitation set forth in the Akzo '374 patent claims does not serve to distinguish those claims from prior art.
160. At trial, DuPont's expert, Dr. Uhlmann, addressed the prior art as a whole and testified that based on his study of the prior art, the subject matter of the Akzo '374 patent claims taken as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art. The Court finds Dr. Uhlmann's testimony convincing.
Objective Factors with Respect to Obviousness
161. There was no long felt need for the NMP/CaCl2 solvent system. Akzo and DuPont independently developed the NMP/CaCl2 system within a few months of each one's perception of a need for an alternative to an HMPA system.
162. There is no history in the record of failure by others.
163. There is no documented commercial success of the '374 process. Akzo did not open its commercial plant for producing PPD-T polymer until March 1986. DuPont's Kevlar fiber was commercially successful] before Akzo's development of the patented system. No nexus exists between the claimed system of the '374 patent and DuPont's commercial Kevlar sales.
164. When DuPont filed the Miller, et al. patent application in 1976, it believed that the NMP/CaCl2 system was patentable over Federov, and the other prior art, although at that time DuPont was unaware of Asakura. DuPont believed that the system's patentability was a close question.
CONCLUSIONS OF LAW
JURISDICTION AND VENUE
1. Jurisdiction of the Court arises under the patent laws of the United States, including Title 35, U.S. Code § 271 and § 281 et seq., and Title 28, U.S. Code § 1338(a).
2. Venue properly lies in this district and division under Title 28, U.S. Code § 1400(b) and Rule 4 of the Local Rules of Practice of the Eastern District of Virginia.
3. The '374 patent is entitled under 35 U.S.C. § 119 to the benefit of the filing date of February 21, 1975, in The Netherlands of Application Number 7,502,060. (Stipulated to by parties).
VALIDITY OF THE '374 PATENT
4. The patent in suit is presumed to be valid under 35 U.S.C. § 282. This presumption can be overcome only by clear and convincing evidence. RCA Corp. v. Applied Digital Data Systems, Inc., 730 F.2d 1440, 1443-44 (Fed. Cir.), cert. dismissed, ___ U.S. ___, 105 S.Ct. 32, 82 L.Ed.2d 923 (1984).
5. The party asserting invalidity "not only has the procedural burden of proceeding first and establishing a prima-facie case, but the burden of persuasion on the merits remains with that party until final decision." Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1534 (Fed. Cir. 1983).
6. The decision of a Patent Examiner is never binding on a court. It is, however, "evidence the court must consider in determining whether the party asserting invalidity has met its statutory burden by clear and convincing evidence." Fromson v. Advance Offset Plate, Inc., 755 F.2d 1549, 1555 (Fed. Cir. 1985).
7. A party attacking the validity of a patent when the prior art relied on was previously considered by the Patent and Trademark Office:
has the added burden of overcoming the deference that is due to a qualified government agency presumed to have properly done its job, which includes one or more examiners who are assumed to have some expertise in interpreting the references and to be familiar from their work with the level of skill in the art and whose duty it is to issue only valid patents.American Hoist Derrick Co. v. Sowa Sons, Inc., 725 F.2d 1350, 1359 (Fed. Cir.), cert. denied, ___ U.S. ___, 105 S.Ct. 95, 83 L.Ed.2d 41 (1984).
8. Patentability is dependent on three conditions: utility and novelty as articulated and defined in 35 U.S.C. § 101 and 102, and nonobviousness as set forth in 35 U.S.C. § 103. Graham v. John Deere Co., 383 U.S. 1, 12, 86 S.Ct. 684, 691, 15 L.Ed.2d 545 (1966).
9. A patent is invalid because of obviousness if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. 35 U.S.C. § 103. See Standard Oil Co. v. American Cyanimid Co., 774 F.2d 448, 453 (Fed. Cir. 1985). It is irrelevant whether or not the inventor knew of the prior art. EWP Corp. v. Reliance Universal Inc., 755 F.2d 898, 907 (Fed. Cir.), cert. denied, ___ U.S. ___, 106 S.Ct. 131, 88 L.Ed.2d 108 (1985).
10. In Graham, the Supreme Court enunciated the test a factfinder must apply for determining obviousness:
Under § 103, the scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobviousness of the subject matter is determined. Such secondary considerations as commercial success, long felt but unsolved need, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy.383 U.S. at 17-18, 86 S.Ct. at 693-694, quoted in Vandenberg v. Dairy Equipment Company, 740 F.2d 1560, 1564 (Fed. Cir. 1984).
The Prior Art
11. Akzo's repeated assertions that none of DuPont's prior art references discloses the use of an NMP/CaCl2 system to polymerize PPD-T are for the most part irrelevant. DuPont's claim that the `374 patent is unpatentable is based on Section 103, not Section 102. As the Federal Circuit recently stated:
A reference must be considered for everything it teaches by way of technology and is not limited to the particular invention it is describing and attempting to protect. On the issue of obviousness, the combined teachings of the prior art as a whole must be considered.EWP Corp., 755 F.2d at 907 (emphasis in original).
12. The evidence presented at trial shows clearly and convincingly that the invention as a whole set forth in the claims of Akzo's `374 patent would have been obvious to one of ordinary skill in the art at the time of the alleged invention.
13. At the time of the Akzo invention, it was known from the prior art, especially the Blades patent, that the process for making PPD-T involved vigorously mixing and reacting p-phenylenediamine (PPD) with terephthaloyl chloride (TCl) in a solvent.
14. It was known from the prior art at the time of the Akzo invention that the reaction should take place under substantially anhydrous conditions.
15. At the time of the Akzo invention, the prior art taught that a key to producing commercial-grade aramid fiber from PPD-T was high inherent viscosity of the PPD-T.
16. The Blades patent and the Chimura patent application taught the importance of low water content and the use of pure materials when polymerizing aromatic polyamides.
17. The evidence shows that Akzo represented to the Patent Examiner, after numerous rejections of Akzo's claims, that the two features distinguishing Akzo's invention from the prior art were the use of NMP/CaCl2 as the solvent system, and Akzo's discovery that over 5% calcium chloride by weight relative to NMP was essential to producing PPD-T of high inherent viscosity.
18. Material representations made to the PTO (with or without amendment) to define and explain the claimed invention for the purpose of distinguishing it from the prior art are pertinent to the proper interpretation of the claim language. Builders Concrete, Inc. v. Bremerton Concrete Products, Co., 757 F.2d 255, 258 (Fed. Cir. 1985).
19. The Court concludes that these two representations, and particularly the "at least 5 percent" calcium chloride limitation, formed the basis of the Examiner's decision to change his mind and to allow Akzo's claims.
20. A person skilled in the art at the time of the Akzo invention would have been aware from disclosures in the prior art, particularly in Man-Made Fibers, the Federov publication, and the Antal patent, that the best non-HMPA amide solvents for polymerizing PPD-T were NMP and DMAc.
21. The prior art at the time of the Akzo invention, especially Man-Made Fibers, the Federov publication, and the Antal patent, taught that the best salts to use to increase the solvent power of amide solvents were calcium chloride and lithium chloride.
22. At the time of the Akzo invention, it would have been obvious to a person skilled in the art seeking an alternative solvent to HMPA that the best course would be to experiment with combinations of DMAc or NMP and lithium chloride or calcium chloride.
23. It also would have been obvious to a person skilled in the art that while lithium chloride generally was a slightly better salt than calcium chloride for increasing the solvent power of an amide solvent, the Antal patent taught that calcium chloride was more effective when the solvent was NMP. It also was obvious to one skilled in the art, and was known by Akzo researchers, that calcium chloride was much less expensive than lithium chloride, and therefore was an attractive alternative salt.
24. The DuPont Blades patent taught the entire process claimed in the `374 patent except for the choice of solvent. The choice of a solvent system once HMPA was ruled out essentially boiled down to four alternatives obvious to one skilled in the art: NMP/LiCl, NMP/CaCl, DMAc/LiCl, DMAc/CaCl2.
25. Federov taught that solubility studies are the most effective experimental technique for determining the potential effectiveness of an amide-salt polymerization solvent. Conducting such a solubility study on NMP/CaCl2, NMP/LiCl, DMAc/LiCl, and DMAc/CaCl2 would have made it obvious to one skilled in the art that calcium chloride was the best alternative salt, given its effectiveness as a salt when combined with NMP to form a polymerization solvent for PPD-T, and given its relative inexpensiveness. Akzo conducted such solubility studies to arrive at the NMP/CaCl2 system.
26. The Court concludes that the choice of the NMP/CaCl2 system as an alternative to an HMPA-based system was obvious to one skilled in the art at the time of the Akzo invention. Indicative of this fact, although not dispositive of the issue, is the evidence that DuPont and Akzo arrived at the same alternative system independently of one another, each within a few months of beginning its research.
27. The Court has found that the "at least 5 percent" calcium chloride limitation that Akzo so vigorously asserted during its prosecution of the patent was and is scientifically meaningless. The Court concludes that the "at least 5 percent" by weight calcium chloride limitation set forth in Akzo's patent claims could not serve to distinguish Akzo's invention from the prior art. The Court also concludes by clear and convincing evidence that had that limitation not been represented to the Examiner as critical, neither he nor any other reasonable Patent Examiner would have approved Akzo's claims.
Secondary Factors
28. Several Federal Circuit opinions direct that secondary factors such as commercial success, product copying, failure of others, and long felt need must always be taken into account, "not just when the decisionmaker remains in doubt after reviewing the art." Stratoflex, 713 F.2d at 1539, quoted in Cable Electric Products, Inc. v. Genmark, Inc., 770 F.2d 1015, 1026 (Fed. Cir. 1985); see also Pentec, Inc. v. Graphic Controls Corp., 776 F.2d 309, 315 (Fed. Cir. 1985).
29. Nevertheless, a "nexus between the merits of the claimed invention and the evidence of secondary considerations is required in order for the evidence to be given substantial weight in an obviousness decision." Simmons Fastener Corp. v. Illinois Tool Works, Inc., 739 F.2d 1573, 1575 (Fed. Cir. 1984), cert. denied, ___ U.S. ___, 105 S.Ct. 2138, 85 L.Ed.2d 496 (1985) (quoting Stratoflex, 713 F.2d at 1539).
30. With respect to commercial success arising from the `374 patent, the Court can find no such nexus. In fact, the Court cannot find any commercial success at all. Akzo opened its commercial aramid fiber plant in March 1986. There is no evidence before the Court regarding any commercial success of Akzo's aramid fiber produced using the solvent claimed in the `374 patent, or indicating that sales of that fiber "represent a substantial share of any definable market or whether the profitability per unit is anything out of the ordinary in the industry involved." See Cable Electric Products, 770 F.2d at 1026-27.
31. The Court concludes that there was no long felt need for the NMP/CaCl2 solvent system and that there is no evidence of failures by others to develop an alternate solvent system to replace HMPA. Once the need for the alternate system was perceived independently by Akzo and Dupont, the two companies developed the same system within a few months of beginning their respective research programs. Although Akzo correctly points out that both it and DuPont tried many solvent systems that could not successfully make high molecular weight PPD-T, such minor setbacks are to be expected when conducting research on the scale engaged in by these companies.
32. Furthermore, the fact that DuPont had failed to produce an alternate system to polymerize PPD-T in the 1960's is irrelevant. At that time, DuPont's focus was on Fiber-B as the fiber and on HMPA as the solvent. HMPA/NMP and HMPA were found to be such superior solvents for PPD-T that research into alternative solvent systems was unnecessary until HMPA was discovered to be carcinogenic.
33. The fact that DuPont filed a patent application on the same invention claimed in the `374 patent does not preclude DuPont from arguing the obviousness of Akzo's claims, nor does it prevent the Court from concluding in light of the entire record that the invention was obvious. Medtronic, Inc. v. Cardiac Pacemakers, Inc., 721 F.2d 1563, 1582 (Fed. Cir. 1983). The Court, however, does consider that fact as evidence tending to support the presumption of validity. Nevertheless, the Court notes that DuPont considered the patentability of the NMP/CaCl2 solvent system to be a close question at best and that DuPont's application was made with the knowledge that the prior art might well preclude a Patent Examiner from approving the claims. The Court concludes that this evidence pales in light of the clear obviousness of the subject matter of Akzo's invention.
34. The Court concludes that DuPont has sustained its burden and has shown by clear and convincing evidence that the subject matter of the Akzo `374 patent claims is obvious within the meaning of 35 U.S.C. § 103. The Court therefore holds that the Akzo `374 patent is invalid.
WILLFUL INFRINGEMENT
35. As the Court has concluded that the Akzo `374 patent is invalid, it cannot find that DuPont infringed the `374 patent, willfully or otherwise. Akzo's claims therefore are hereby ordered dismissed.
FRAUD OR INEQUITABLE CONDUCT
36. The various decisions of the Federal Circuit indicate that a two-prong test is used to determine whether an applicant has procured a patent by such conduct as will render the patent unenforceable and give rise to a finding of what has been called "fraud on the Patent Office," and which is now known as "inequitable conduct." See, J.P. Stevens Co., Inc. v. Lex Tex Ltd., Inc., 747 F.2d 1553, 1559 (Fed. Cir. 1984), cert. denied, ___ U.S. ___, 106 S.Ct. 73, 88 L.Ed.2d 60 (1985). Such a finding generally is based on nondisclosure of a pertinent reference during the prosecution process.
37. To support a finding of "inequitable conduct," there must first be proof, by clear and convincing evidence, that the non-disclosed information has a threshold degree of materiality. Atlas Powder Co. v. E.I. DuPont de Nemours Co., 750 F.2d 1569, 1577-1578 (Fed. Cir. 1984). Second, there must be proof of a threshold intent on the part of the applicant. Id. at 1578. The degree of materiality and the level of intent must then be established and balanced to determine whether "inequitable conduct" has occurred. J.P. Stevens v. Lex Tex, 747 F.2d at 1560.
38. The Federal Circuit advocates a sliding scale, such that the greater the materiality of the non-disclosed reference, the less intent is needed to show that the applicant has engaged in "inequitable conduct". American Hoist Derrick Co., 725 F.2d at 1363.
39. With regard to the materiality of the information, the Federal Circuit has said that it may be proven by any of four criteria, but that the Patent Office standard (i.e., 37 C.F.R. § 1.56(a) — whether there is substantial likelihood that a reasonable Examiner would have considered the omitted or false information important in deciding whether to allow the application to issue as a patent) is the appropriate starting point. American Hoist Derrick Co., 725 F.2d at 1363.
40. If a sufficient level of materiality is found, the Court must then determine the intent of the applicant. In making this determination, hindsight analysis should be avoided. Instead, one should look to the facts and circumstances facing the actor at the time of the patent prosecution. See Kansas Jack, Inc. v. Kuhn, 719 F.2d 1144, 1152 (Fed. Cir. 1983). In addition, "simple negligence, oversight or an erroneous judgment made in good faith" is insufficient to result in a holding of "inequitable conduct". J.P. Stevens v. Lex Tex, 747 F.2d at 1560.
41. The Court has reviewed the evidence pertaining to the prosecution of the `374 patent and to the tests conducted by Akzo. The Court concludes that DuPont has failed to demonstrate by clear and convincing evidence that Akzo engaged in "inequitable conduct" in the prosecution of the `374 patent.
The Court is especially unable to conclude by clear and convincing evidence that the requisite intent is present in this case. It is unclear whether Akzo intended to represent that the 5% floor was a scientific or a commercial limitation. It is also impossible for the Court to determine whether the experiments conducted by Akzo during the patent prosecution and the conclusions reported therefrom are indicative of bad faith, sloppy experimentation, or simply a belief by Akzo researchers that further experimentation would be futile given their prior experience and the prior art as known by them.
DUPONT'S COUNTERCLAIM
42. An order shall issue dismissing Akzo's complaint with prejudice; declaring that United States Patent No. 4,308,374 is invalid; and declaring that United States Patent No. 4,308,374 has not been infringed by DuPont. All other relief prayed for by DuPont in its counterclaim shall be denied.