The Dow Chemical Company, Appellee, v. American Cyanamid Company, Appellant, 816 F.2d 617 (Fed. Cir. 1987)

This appeal is from the final judgment of the United States District Court for the Eastern District of Louisiana, 615 F. Supp. 471, 229 USPQ 171 (1985), holding that certain claims of plaintiff-appellee Dow Chemical Company's (Dow) U.S. Patents No. Re 31,430 ('430 reissue patent), No. 3,994,973 ('973 patent), and No. 3,642,894 ('894 patent), are not invalid and were infringed by defendant-appellant American Cyanamid Company (Cyanamid), and that Dow's U.S. Patent No. Re 31,356 (the '356 reissue patent) is not invalid but was not infringed by Cyanamid. The parties do not in this appeal challenge the findings on the infringement issue. We affirm the judgment on the issue of the validity of the claims in suit.

The four patents in suit relate to a chemical process of using a copper catalyst to convert an olefinic nitrile having 3 to 6 carbon atoms to its corresponding amide. Dow asserted that Cyanamid infringed its patents by using a process for catalytically converting acrylonitrile to acrylamide with a copper catalyst prepared by reducing a combination of copper compounds.

Acrylonitrile is a commonly produced chemical, and its conversion to acrylamide is an extremely valuable commercial process. That process, in which water is combined with acrylonitrile to produce acrylamide, is known as a "hydration" process, more specifically the "hydrolysis" of a nitrile, acrylonitrile, to its corresponding amide, acrylamide.

A nitrile is an organic compound containing a carbon-to-nitrogen triple bond and, depending on the rest of the molecule, can be classified as either aromatic or aliphatic. Aliphatic compounds can be either saturated or unsaturated. A saturated aliphatic compound is one in which all the bonds between the carbon atoms are single bonds, while an unsaturated compound is one in which at least two carbon atoms are joined by double (olefinic) or triple (acetylenic) bonds. Acrylonitrile is an unsaturated aliphatic nitrile, more specifically an olefinic nitrile expressed by the chemical formula CH2 = CH-C = N.

In the hydrolysis or hydration of acrylonitrile to acrylamide, a molecule of water reacts with the nitrile group of the acrylonitrile molecule to form acrylamide, CH2 = CH - COHN2 . The conventional industrial method of producing acrylamide involves mixing acrylonitrile with sulfuric acid and water to form acrylamide sulfate, which is then decomposed to acrylamide by reaction with ammonia, resulting in the formation of the by-product ammonium sulfate as well as acrylamide. The process now used by both Dow and Cyanamid, however, involves contacting the acrylonitrile with water in the presence of a copper catalyst, thereby forming acrylamide directly without the formation of any by-products.

For the purposes of this appeal, Cyanamid apparently concedes that its method for producing acrylamide would infringe the Dow patents. Cyanamid contends, however, that the Dow patents are invalid for claiming "unpatentably obvious" subject matter, arguing that one of ordinary skill in the art, who would be a Ph.D. chemist with industrial experience, having knowledge of the prior art, would have predicted and expected that certain copper catalysts, which had been employed generally in the nitrile conversion process, would also convert acrylonitrile to its corresponding amide.

The chemical process at issue is generally similar to the process that was at issue in Standard Oil Company v. American Cyanamid Company, 774 F.2d 448, 227 USPQ 293 (Fed. Cir. 1985) (Standard Oil) . In that case, this court affirmed a decision of Judge George Arceneaux of the same United States District Court for the Eastern District of Louisiana, holding claim 2 of Standard Oil's U.S. Patent No. Re 25,525 (the Greene patent) invalid for obviousness. The Greene patent also claimed a process for converting acrylonitrile to acrylamide using a different copper catalyst. In the Standard Oil case, the district court relied on two prior art references in support of its conclusion. Those references were U.S. Patent No. 1,891,055, issued in 1932 to Walter Reppe (the Reppe patent) and a series of articles written by Dr. Kenichi Watanabe (the Watanabe articles). Those same two references, along with the Greene patent and U.S. Patent No. 3,669,639, issued to Louis Haefele in 1968 (the Haefele patent), constitute the principal prior art which Cyanamid here asserts renders the Dow patents in suit invalid.

We note additionally that although the Standard Oil case and this one involve generally similar conversion processes--acrylonitrile to acrylamide using a copper catalyst--, the claims, patents, and records in these cases are dissimilar and thus we are in no way bound by that prior decision.

The Reppe patent, No. 1,891,055, uses copper catalysts to carry out the hydrogenation (the direct addition of hydrogen to the carbon-carbon bond of an unsaturated nitrile, to form a saturated nitrile, as distinguished from hydration, which is the addition of water to the carbon-nitrogen bond) of unsaturated nitriles to form an unsaturated amide. It discloses, inter alia, that after hydrogenation and separation of the copper catalysts from the reaction mixture, the saturated nitriles can be converted to their corresponding amides by the addition of water. Example 2 of the Reppe patent discloses the formation of propionitrile, a saturated aliphatic nitrile and its corresponding amide, propionamide, from crude acrylonitrile, using a catalyst prepared by the reduction of copper carbonate.2. The Watanabe Articles

Dr. Watanabe published a series of articles¹  confirming the efficacy of hydrating aliphatic and aromatic nitriles to their corresponding amides using metallic nickel catalysts. Dr. Watanabe also used metallic copper catalysts but with poor results; he stated that they were not as useful because of undesirable side reactions. A 1964 Watanabe article reports the details of experiments converting benzonitrile (an aromatic nitrile) to benzamide using, inter alia, copper catalysts, including precipitated copper, copper oxide and Urushibara copper, which is a metallic copper prepared by the reduction of copper chloride. The results of the experiments reported indicated that when the Urushibara copper catalyst was used to convert an aromatic nitrile to its corresponding amide the results were unsatisfactory.

The U.S. Greene patent, No. 3,381,034 (reissued as Re 28,525), discloses a homogeneous catalytic system consisting of soluble copper ions alone or in combination with metallic copper ions that will convert aromatic and aliphatic nitriles to their corresponding amides. The patent states that metallic copper alone is not an effective catalyst for converting acrylonitrile to acrylamide.

The Haefele U.S. patent, No. 3,366,639, describes the use of large amounts of the catalyst manganese dioxide to accomplish the hydration of a variety of nitriles, including aromatic, aliphatic, and olefinic nitriles, to their corresponding amides. It does not disclose the use of copper or any catalyst other than manganese dioxide but does state that its process is applicable to production of amides from nitriles in general.

C. The Patents at Issue and Their Prosecution

All of the Dow patents at issue here are continuations-in-part of an application filed in January of 1969 that eventually matured into U.S. Patent No. 3,597,481 (the '481 patent). The four patents that are asserted by Dow to be infringed by Cyanamid here are continuations-in-part of a June 1969 application that eventually became U.S. Patent No. 3,631,104 (the '104 patent), which was a continuation-in-part of the '481 patent application. The '104 patent was reissued as the '430 reissue patent on October 25, 1983.

The original claims of the '104 patent were directed to a process for converting aliphatic nitriles in general to their corresponding amides in the presence of a copper catalyst prepared by reducing various copper oxides. On February 26, 1981, ten years after the '104 patent issued, Dow filed a "no defect" reissue application, in which it disclosed several new references to the Patent and Trademark Office (PTO), including the German equivalent of the Reppe patent.

In the first office action on the reissue, the examiner rejected all claims directed to the hydration of aliphatic nitriles in general as obvious in view of the German Reppe patent. In response, Dow conceded that the Reppe patent disclosure might render their broader claims obvious but maintained that it disagreed with the examiner's decision. Dow proposed limiting the claims to hydration of olefinic nitriles of 3 to 6 carbon atoms, and the claims were then allowed.

Claim 26 of the '430 reissue patent, the one considered by the district court, reads:

26. The process of claim 16 wherein the nitrile is acrylonitrile.

Claim 16, not in issue, reads:

16. In the process for converting a nitrile to the corresponding amide by contacting an olefinic nitrile of 3 to 6 carbon atoms in the presence of water with a heterogeneous catalyst, the improvement comprising using a cupreous catalyst consisting essentially of reduced copper oxide, reduced copper chromium oxide, reduced copper-molybdenum oxide, unreduced copper-molybdenum oxide or mixtures thereof.

The first of a number of applications filed by Dow as continuations-in-part of the '104 patent application issued as the '894 patent. Claim 6, the one claim considered by the district court, is directed to the same basic process for converting an acrylonitrile to its corresponding amide. The difference between this claim and claim 26 of the '430 reissue patent is an additional step of "at least partially protecting the reduced catalyst from contact with oxygen after reduction."

The '973 patent is similar to the '430 reissue patent. Claim 13, the only claim considered by the district court, defines a process for converting an acrylonitrile to its corresponding amide "in the presence of a catalytic amount of copper prepared by reducing copper oxide." The district court held claim 13 infringed by Cyanamid.

A fourth Dow patent, also a continuation-in-part of the application for the '104 patent, issued September 11, 1973, as No. 3,758,578 (the '578 patent) and was reissued August 23, 1983, as the '356 reissue patent. Claim 10, the only claim in suit, defines the process of converting acrylonitrile to acrylamide using a copper salt catalyst. The district court held that claim 10 was not infringed because it found that the catalysts used by Cyanamid were prepared by reducing copper oxides, not copper salts.

The record indicates that there was a long-felt need to find a better way to manufacture acrylamide than by the sulfuric acid process. Dow urges that the shift to a catalytic process by both Dow and Cyanamid and Cyanamid's admitted adoption of a process based inter alia on information gleaned from the Dow patents are both objective factors tending to support the patentability of Dow's claimed process. Cyanamid argues that there was no evidence of nexus between the commercial success of Dow's acrylamide business and any aspect of the claimed invention, and further urges that the commercial success enjoyed by both Dow and Cyanamid was more probably attributable to a market demand for acrylamide.

Cyanamid conceded that the Dow patents provide one common disclosure to the public of a chemical process for the conversion of nitriles to amides, but noted at trial that the only distinction of Dow's invention from prior known processes is Dow's claim limitation to one subgroup of nitriles, namely, olefinic nitriles of 3 to 6 carbon atoms. Thus, Cyanamid asserted that the subject matter of the patents as a whole would have been obvious to one of ordinary skill in the art at the time that Dow's inventions were made. With respect to the obviousness of the claimed invention, Dow asserted that Cyanamid's copying of Dow's process, the great need to commercialize the process, and the commercial success enjoyed by Dow were compelling evidence of nonobviousness.

Trial to the district court sitting without a jury lasted seven days, during which time the district court heard the testimony of one of the Dow patentees, Dr. Ben Tefertiller, as well as two chemical experts on behalf of Cyanamid and an expert in the field of patent practice on behalf of Dow. The district court issued its decision, including findings of fact and conclusions of law, on August 14, 1985.

The only district court conclusion at issue is the legal question of nonobviousness. In addressing it the district court first noted the Graham inquiries, emphasizing objective evidence of nonobviousness such as long-felt need, unexpected results, commercial success and failure of others.

The court then turned to an analysis of the prior art and a comparison of each reference with the claimed invention, which the court characterized as "the hydration of acrylonitrile to acrylamide using a reduced copper compound as a catalyst." In discussing the Reppe patent, the district court found that it taught that reduced copper catalysts are useful for the hydrogenation of unsaturated nitriles to the corresponding saturated nitriles and that after hydrogenation the saturated nitrile can be converted to an amide by the addition of water. The district court concluded its analysis by stating that the Reppe patent "does not disclose the hydration of acrylonitrile to acrylamide using a reduced copper catalyst."

The district court found that the Watanabe articles indicated that the hydration of aliphatic nitriles was somewhat different from the hydration of aromatic nitriles, and that metallic copper catalysts "gave poor results as catalysts for hydration." In addition, the court noted that "Dr. Watanabe taught that the catalysts could be exposed to air during their preparation." The court concluded with the statement that " [t]he reference does not disclose the hydration of acrylonitrile to acrylamide with a reduced copper compound as a catalyst."

The district court acknowledged that the Greene patent "taught that soluble copper ions would convert aromatic and aliphatic nitriles, including acrylonitrile, to their corresponding amides" but also found that it stated "that metallic copper alone was not an effective catalyst for converting acrylonitrile to acrylamide." Similarly, the court noted that the Haefele patent "disclosed the use of large amounts of manganese dioxide as a catalyst with an organic solvent to convert nitriles, including acrylonitrile, to the corresponding amide," but found that " [t]he reference does not disclose the use of a reduced copper compound to catalyze the hydration of acrylonitrile to acrylamide."

The court then concluded "that there are differences between the subject matter claimed by Dow in the patent-in-suit [sic] and the prior art when taken as a whole, and that such differences would not be [sic, have been] obvious to those skilled in the art." The court found those differences to be, inter alia, that "the hydration of nitriles by the use of a copper catalyst obtained through the reduction of a copper compound was not predictable." The court also stated that the process claimed in the Dow patents was "limited to the conversion of certain unsaturated nitriles, particularly acrylonitrile, to its corresponding unsaturated amide." The court found that process to be different "from the prior art in the nature of the nitrile hydrated" because the prior art showed the hydration of only "saturated and aromatic nitriles."

Finally, the district court acknowledged several secondary considerations to support its conclusion that the Dow patents had not been proven invalid. The court found that Dow was able to build a full-size plant, utilizing its patented invention to replace the sulfuric acid process. The court was also impressed by the fact that the Dow process "produces high yields of pure acrylamide ... with little or no by-product." That fact, coupled with Cyanamid's apparent inability to develop its own commercial process based on the prior art, indicated to the court that Dow's patented invention would not have been obvious.

Two weeks later the court issued its judgment, holding the '430 reissue patent, the '894 patent, the '973 patent, and the '356 reissue patent valid and enforceable. The court then held that Cyanamid infringed claim 26 of the '430 reissue patent, claim 6 of the '894 patent, and claim 13 of the '973 patent, and permanently enjoined Cyanamid from practicing the processes of those claims.²  The injunction was stayed pending appeal to this court.

The district court correctly recognized that Reppe discloses a process in which unsaturated nitriles are converted to saturated ones by using a copper catalyst and in which the saturated nitrile is then converted to its corresponding amide but without the use of the copper catalyst as in the claimed inventions. The disclosure is of a two-step process, and the district court acknowledged that distinction in making its Graham inquiry.³ 

The district court recognized that the Watanabe articles disclose the use of copper catalysts to successfully convert aromatic and aliphatic nitriles to their corresponding amides, and Watanabe's statement that " [i]t has been found that the usual catalysts for hydrogenation are generally effective in the hydration of nitriles." The court found that Watanabe did not suggest the use of copper catalysts to convert acrylonitrile to acrylamide because his results showed that the yields of amide obtained in such a manner were very poor compared to the use of nickel catalysts, that copper catalysts gave unexpected side reactions, and that some copper catalysts caused the formation of amines rather than amides, all of which would have suggested the undesirability of a process as defined by the claims in issue. There is nothing in the record to support a conclusion that these findings are clearly erroneous, and thus we are not convinced that a mistake has been made.

The Greene reference shows that a catalyst composed of the combination of soluble copper ions and metallic copper will convert a wide range of nitriles to amides, including olefinic nitriles of 3 to 6 carbon atoms and specifically including acrylonitrile. That art also explicitly states, as found by the district court, that copper metal alone "is not an effective catalyst for [converting a nitrile to its corresponding amide]," and to that extent teaches away from the claimed invention's use of copper metal as a catalyst.

The Haefele patent relates only to the use as a catalyst of manganese dioxide to convert numerous different nitriles. The present case is concerned with copper catalysts.

With regard to objective evidence of nonobviousness, the district court found that "Dow's efforts were successful, not only scientifically, but also commercially in that Dow was able to build a full-size plant utilizing its patented invention that replaced the sulfuric acid process." The record reflects that this success was possible because the claimed invention unblocked a cost barrier to the market; it displaced the prior inefficient process burdened with the costs of extra reagents and an ammonium sulfate by-product. While Cyanamid's argument that Dow has not shown that its increase in sales resulted from the claimed invention may be well taken, there is other strong objective evidence of nonobviousness in the record as found by the district court.

The finding that the claimed invention resolved a long-felt need to do away with the wasteful sulfuric acid process and its unwanted by-products tends to show the nonobviousness of the claimed invention as does failure of others and copying; Cyanamid tried but failed to develop the claimed invention and copied it instead.

Based on all of the above, we conclude that the subject matter of the four claims of the four patents in suit would not have been obvious to one of ordinary skill in the art at the time those inventions were made. The Reppe prior art discloses copper metal catalysts and hydration, but when the distinction between using that copper catalyst in hydrogenation and not in hydration is made, that disclosure is not meaningful to the obviousness of the claimed invention; in Reppe, the copper catalyzed the saturation of the carbon-carbon bonds of the nitrile, not its hydration, to an amide, the subject matter of the claims in suit. The fact that the 1964 Watanabe article states that the usual hydrogenation catalysts are "generally effective" in the hydration of nitriles to amides might suggest the claimed invention when viewed in combination with Reppe if it were not for the fact that Watanabe also states that copper metal catalysts give poor yields and unexpected side reactions, the first of which the court found the claimed invention excelled in. Furthermore, taking the Greene reference as a whole in combination with Reppe and Watanabe, we can only conclude that its explicit statement that copper catalysts were ineffective for use in hydrating nitriles would lead one of ordinary skill in the art away from the claimed invention. Finally, the Haefele prior art adds little to the equation of obviousness. Lastly, the objective evidence of nonobviousness, particularly failure of others and long-felt need, lends much to our conclusion that the claimed invention fully satisfies the requirements of 35 U.S.C. § 103.

Accordingly, the district court's holding that the claims in suit of the four patents in suit, Re 31,430, No. 3,994,973, No. 3,642,894, and Re 31,356, are not invalid is affirmed.

AFFIRMED.

DAVIS, Circuit Judge, dissenting.

In my view, a contrary result is compelled by this court's earlier decision in Standard Oil Co. v. American Cyanamid Co., 774 F.2d 448, 227 USPQ 293 (Fed. Cir. 1985) (Sohio) .¹  The chemical process involved here is substantially the same process at issue in Sohio, although the challenged patents are different. Because of the general identity of the chemical process, there are holdings in Sohio that bear directly on the correctness of the District Court's decision in this case. The most important of these is our Sohio holding that the District Court there correctly "found that the combined teachings [of the Watanabe article and Reppe patent]²  would have indicated to one of ordinary skill in the art that copper³  was an effective agent for producing a catalysis in the process of converting a nitrile to an amide." 774 F.2d at 455, 227 USPQ at 298.⁴  This was a legal determination of obviousness, but in the current case Judge Beer made the precisely contrary determination that "the hydration of nitriles by the use of a copper catalyst obtained through the reduction of a copper compound was not predictable, and was not obvious." Because Sohio is a binding legal precedent on this exact point,⁵  the District Court's contrary legal holding should not stand. But that erroneous holding of the District Court was a mainstay of its opinion sustaining the validity of the patents at issue in this case.

Spelling out the matter somewhat further, my thought is that the District Court erred, as a matter of law, in holding that the Dow patents would not have been obvious under 35 U.S.C. § 103. The court first discussed each of the pertinent prior art references separately and, on that basis, was able to conclude that none of those references specifically disclosed the hydration of acrylonitrile to acrylamide using a reduced copper compound. It goes without saying, however, that it is not necessary, under Sec. 103, that each and every element of the claimed invention be disclosed in a prior art reference. Such an analysis is properly addressed only for the issue of anticipation under 35 U.S.C. § 102. W.L. Gore & Assocs. v. Garlock, Inc., 721 F.2d 1540, 1554, 220 USPQ 303, 313 (Fed. Cir. 1983), cert. denied, 469 U.S. 851, 105 S. Ct. 172, 83 L. Ed. 2d 107 (1984). Rather, the proper analysis under Sec. 103 is whether the invention, taken as a whole, would have been obvious to one skilled in the art. Environmental Designs Ltd. v. Union Oil Co., 713 F.2d 693, 698, 218 USPQ 865, 870 (Fed. Cir. 1983) cert. denied, 464 U.S. 1043, 104 S. Ct. 709, 79 L. Ed. 2d 173 (1984).

While the District Court did make two conclusions on obviousness purportedly based on the prior art considered as a whole, to my mind both conclusions must be rejected. First, the court decided that the prior art taught that the hydration of nitriles was not obvious since it was not predictable. As discussed supra, this conclusion runs directly contrary to this court's holding in Sohio and therefore should not be accepted since it involved the identical prior art. Second, the court concluded that the "process claimed in the Dow patents, therefore, differ from the prior art in the nature of the nitrile hydrated and the nature of the resultant amide" (e.g., acrylonitrile hydrated to acrylamide). This conclusion is unsupported by the District Court's own opinion which acknowledged that the Greene Patent (U.S. Letters Patent 3,381,034) taught that a copper catalyst could be used to convert "aromatic and aliphatic nitriles, including acrylonitriles, to their corresponding amides."

I would reverse the decision of the District Court.