247 F.Supp. 51 (1965)
Carl Max BRENNER, and Hans Rudolf Rickenbacher, Plaintiffs,
v.
David L. LADD, Commissioner of Patents, Defendant.
Civ. A. No. 1542-63.
United States District Court District of Columbia.
October 7, 1965.
A. Ponack, Washington, D. C., for plaintiffs.
Clarence W. Moore, Sol., Washington, D. C., for defendant.
JACKSON, District Judge.
This is a civil action pursuant to 35 U.S.C. § 145 wherein plaintiffs seek an adjudication that the defendant, Commissioner of Patents, should be authorized to issue to plaintiffs a patent containing claims 12 to 18 and 20 to 22, inclusive, of application Serial No. 784,054, filed December 31, 1958, entitled "Process for the Preparation of L-Alpha-Amino-Epsilon-Caprolactam".
The present invention concerns a process for the preparation of L-alpha-amino-epsilon-caprolactam (hereinafter referred to as aminocaprolactam, or acl), by resolution of the racemic compound DL-aminocaprolactam with L-2-pyrrolidone-5-carboxylic acid (hereinafter referred to as L-pca) as the resolving agent. The purpose for resolving optically inactive DL-acl into the optically active L-acl isomer is that the latter substance may be hydrolyzed to form the amino acid L-lysine, which in turn is useful as an additive to animal or human foodstuffs, in particular to supplement vegetable proteins having a poor content of L-lysine. The present invention is also concerned with achieving an economical commercial process for the preparation of L-acl by recycling L-pca and D-acl, by-products of the process which would otherwise be wasted.
Claims 13, 16, and 20 are representative, and read as follows:
Claim 13.
In a process for the production of L - alpha-amino-epsilon-caprolactam, from DL-alpha-amino-epsilon-caprolactam, the step of reacting DL-alpha-amino-epsilon-caprolactam with substantially an equimolar amount of L-2-pyrrolidone-5-carboxylic acid in solution in an alkanol having at most two carbon atoms, and separating the resultant difficultly soluble L-alpha-amino-epsilon-caprolactam L-2-pyrrolidone-5-carboxylate from the solution.
Claim 16.
The pure, isomer free, L-alpha-amino-epsilon-caprolactam L-2-pyrrolidone-5-carboxylate, melting at about 198-201°C.
Claim 20.
An independent entity consisting solely of crystalline L-alpha-amino-epsilon-caprolactam L-2-pyrrolidone-5-carboxylate.
Claim 14 is narrower than claim 13 in that the alkanol is specified to be methanol.
Claim 15 is directed to the step of racemizing the by-product D-acl to form additional DL-acl starting material which is then recycled in the process. Claim 12 defines the entire process, including the resolution step of claims 13 and 14 and the racemization step of claim 15. Product claims 17 and 21, directed to L-amino-caprolactam itself, and claims 18 and 22, directed to the hydrochloride of L-acl, are similar in form to above claims 16 and 20 for the L-acl.L-pca salt.
The references relied upon by the Patent Office in rejecting plaintiffs' claims are as follows:
Dearborn et al. 2,528,267 October 31, 1950
Francis et al. 2,876,218 March 3, 1959
Karrer "Organic Chemistry", 4th English Ed., pp. 102-4 (Elsevier),
(1950).
Karrer "Organic Chemistry", pp. 92-93 (1938) (Nordmann).
Gilman "Organic Chemistry" Vol. I, Chapter 3, Part IV, pp. 176-200
(1938) (Wiley).
The Dearborn et al. patent discloses a process for the resolution of racemic amines with L-pca. The general types of amines disclosed by Dearborn et al. are alkyl amines, aralkyl amines, and alkaloids. Two of the three examples of this patent are concerned with the resolution of racemic alkyl amines, and the remaining example relates to the resolution of an aralkyl amine with L-pca.
The Francis et al. patent discloses the preparation of alpha-amino-epsilon-caprolactam and its conversion to lysine by hydrolysis, proceeding through the intermediate aminocaprolactam hydrochloride.
Gilman and Karrer are general background references which describe racemic modifications and methods for their resolution, including the method of chemical conversion to diastereoisomers. Karrer discloses that all ordinary syntheses give rise to optically inactive racemates and that the method most frequently used for the separation of these racemic modifications is chemical resolution.
The Patent Office has taken the position that plaintiffs' method claims 12 to 15 are unpatentable over Dearborn et al. and Francis et al. taken in view of Gilman and Karrer on the ground that it would be obvious to attempt to resolve the DL-acl of Francis et al. with the L-pca of Dearborn et al. in view of the disclosures of chemical resolution contained in Gilman and Karrer.
Product claims 16 and 20 were rejected on the ground that the L-acl.L-pca salt is the inherent product of an obvious process, the process of claims 12 to 15 being considered obvious by the Patent Office.
Claims 17 and 21 for L-acl, and 18 and 22 for L-acl.HCl were rejected on the ground that these optically active isomers are obvious and unpatentable over the corresponding racemic compounds disclosed in Francis et al.
Testimony given at the trial of this case established that the over-all process claimed by plaintiffs is considered an "ingenious" one, by skilled organic chemists. Further testimony established the well known unpredictability of chemical reactions in general, and especially the unpredictability in this particular field of chemical resolution of racemic compounds with optically active resolving agents.
While it was also established that Dr. Louis Fieser was unaware of the existence of the Dearborn et al. patent at the time he characterized plaintiffs' process as "ingenious" on pp. 1033-1034 of the book "Advanced Organic Chemistry", co-authored by Dr. Fieser in 1958, and while one of the plaintiffs, Dr. Max Carl Brenner, testified that he likewise was unaware of the Dearborn et al. patent at the time of the invention, it must be recognized at the outset that the Dearborn et al. patent, which is the principal reference relied upon by the Patent Office, is considered by this Court to be a fully effective reference for all that it teaches or suggests to those of ordinary skill in the resolution art to which it relates.
It is immaterial whether or not an applicant for a patent actually knew of a reference at the time of his invention. It is likewise immaterial that men of ordinary skill in the art to which a reference relates were generally unaware of the existence or contents of a reference at the time of an invention. See Siegel v. Watson, 105 U.S.App.D.C. 344, 267 F.2d 621 (1959). Inventors and other skilled workers in a particular field are conclusively presumed to have constructive knowledge of all references which are available to the public in printed form at the time of making an invention. This is true regardless of whether the reference is a domestic patent as here, or a foreign patent or printed publication in English or a foreign language, or whether the reference is an article in a widely circulated publication or an obscure technical journal.
Proceeding now from the determination that the Dearborn et al. patent is available as a reference for all that it discloses or suggests, the issue now presented is exactly what would be suggested by this reference to an organic chemist of ordinary skill in the art of resolving racemic modifications by conversion into diastereoisomers.
Initially, the types of racemic amines whose possible resolution with L-pca would be suggested to skilled chemists would be the three types disclosed in the middle of Column 2 of the patent, namely, alkyl amines, aralkyl amines, and alkaloids. Evidence of record shows that plaintiffs' DL-acl starting material, a cyclic amine with a cyclically bound amide group, is not one of these three types of amines.
On the one hand, it is true that the suggestion of resolution would not be limited to these three types of amines. On the other hand, it is certain that the construction which the Patent Office contends should be given to this principal reference is too broad. On the basis of the statement in the certificate of correction for the Dearborn et al. patent that "all of the heretofore mentioned racemic amines which are resolved by practicing the present novel process are more basic in character than is the resolving agent, pyrrolidone carboxylic acid", the Patent Office contends that the resolution of all racemic amines which are in fact more basic than pca would be suggested to an organic chemist of ordinary skill reading this Dearborn et al. patent.
In view of the evidence of record in this case establishing the unpredictability of chemical resolutions, this Court holds the opinion that no skilled chemist would misinterpret such a statement, to the effect that all racemic amines which are in fact resolved by the process of Dearborn et al. are more basic than pca, to mean that all racemic amines which are more basic than pca would probably be resolved by this process.
More important, however, than the types of amines whose resolution with pca would be suggested to the skilled chemist upon reading this principal reference is the clear suggestion that, in order to obtain the desired pure L-isomer, the racemic amine should be resolved with D-pca rather than the L-pca which plaintiffs used.
All three examples of Dearborn et al. show that, when L-pca is used in the process of resolution, the D-isomer is precipitated. Likewise, when Dearborn et al. seek to obtain the L-amine isomer from the mother liquor in pure form, D-pca is used for this further resolution. Lines 51 to 53 of Column 2 of the Dearborn et al. patent contain the statement that the desired amine isomer in purified form is obtained by resolving with an isomer of pyrrolidone carboxylic acid having an optical rotation opposite to that of the amine.
With the foregoing in mind, it must be considered whether a skilled chemist faced by the whole problem which plaintiffs were attempting to solve would consider it obvious to resolve the racemic aminocaprolactam of Francis et al. with the L-pca of Dearborn et al.
The Patent Office has taken the position that the problem to be solved was merely the resolution of racemic aminocaprolactam into its constituent optically active enantiomers. However, as fully established by testimony at the trial of this case, the real problem which plaintiffs faced was of a much more complicated nature. It was not sufficient for plaintiffs merely to resolve their aminocaprolactam starting material, in which case recovery of either the D-isomer or the L-isomer would have been satisfactory. In order to achieve the desired ultimate objective of an economical, commercially feasible process for the preparation of metabolically active L-lysine, it was necessary for plaintiffs, first, to obtain the desired L-acl isomer in pure form, uncontaminated by either the D-acl isomer or the DL-acl racemate. Second, for economic purposes it was necessary for plaintiffs to minimize the formation of by-products which could not be reused in the process; and third, it was necessary for plaintiffs to devise a process for converting the pure L-acl isomer to the desired end product, L-lysine, without racemization which would form the undesired DL-lysine.
In order to achieve the desired goal of an economical process, it was necessary for plaintiffs to locate an operable resolving agent for the racemic aminocaprolactam of the Francis et al. patent which would lead to the formation of the desired pure L-acl intermediate in simple fashion. With this background in mind, this Court is of the opinion that a skilled organic chemist reading Dearborn et al. in conjunction with Francis et al. would conclude that L-pca might be an operable resolving agent for DL-acl; but that, if operable at all, the process would probably lead to the precipitation of the D-acl-L-pca diastereoisomer. The skilled chemist would be likely to expect that the desired L-isomer would remain in solution in the mother liquor in the form of L-acl-L-pca contaminated by that portion of the precipitated D-acl.L-pca which is soluble in the solvent. As was clearly established by testimony at the trial, there is no such thing as absolute insolubility, and therefore, it would be expected that the L-isomer in the solution would be contaminated by some dissolved D-acl.L-pca salt.
While the Dearborn et al. patent might further suggest recovery of the purified L-acl isomer by "further resolution which an isomer of pca having a rotation opposite to that of the amine", namely, D-pca, this Court is of the opinion that an organic chemist of ordinary skill seeking an economical, simple process to obtain pure L-acl from DL-acl would reject the aforesaid hypothetical process as uneconomical and would continue to search for another resolving agent other than pca.
In view of the determination that the skilled chemist faced with plaintiffs' entire problem would not consider it obvious to react the DL-acl of Francis et al. with the L-pca of Dearborn et al., plaintiffs' process, wherein DL-acl is successfully resolved in a simple manner with L-pca leading to the unexpected precipitation of L-acl.L-pca diastereoisomer in sufficiently pure form for conversion to L-lysine through the intermediates L-acl and L-acl.HCl, is considered unobvious subject matter. This Court therefore concludes that process claims 12, 13, and 14 which recite this unobvious subject matter are patentable over the prior art relied upon by the Patent Office.
Process claim 15, on the other hand, merely recites the step of racemizing the by-product D-acl in the mother liquor by heating this D-acl together with a base to a temperature of 150°C165°C in order to obtain additional DL-acl starting material. While such racemization may provide for further economy in the process by producing additional DL-acl from what would otherwise be a waste product, this step of racemizing by the application of heat in the presence of a base is considered to be obvious and unpatentable, especially in view of the knowledge of skilled chemists that racemization occurs readily with compounds, such as D-acl, which contain a carbonyl group adjacent to an asymmetric carbon atom bearing a hydrogen atom. Therefore, claim 15 should not be included in the patent to be granted.
Proceeding now to a consideration of product claims 16 to 18 and 20 to 22, it is noted that the racemic compounds aminocaprolactam and aminocaprolactam hydrochloride are disclosed in the Francis et al, patent upon which the Patent Office relies in its determination that the L-acl isomer of claims 17 and 21 and the L-acl.HCl isomer of claims 18 and 22 are unpatentable over the two corresponding racemic compounds of Francis et al. This Court agrees with the Patent Office position that, in the absence of unexpected or unobvious beneficial properties, an optically active isomer is unpatentable over either the isomer of opposite rotation or, as in this case, the racemic compound itself. See In re Adamson et al., 275 F.2d 952, 47 CCPA 841 (1960); Sterling Drug Inc. v. Watson, 135 F.Supp. 173 (D.C.D.C.1955). No beneficial properties for L-acl or L-acl.HCl have been shown other than their utility as intermediates in the production of L-lysine. This utility is considered obvious in view of the teachings of Francis et al. that DL-acl may be hydrolyzed to DL-lysine through the intermediate DL-acl.HCl. Claims 17, 18, 21, and 22 should not be granted to plaintiffs.
While this determination that the L-acl of claims 17 and 21 is obvious and unpatentable according to 35 U.S.C. § 103 makes it unnecessary for the Court to consider the issue raised under 35 U.S.C. § 102(e),[1] it should be noted that plaintiffs' L-acl is not considered by this Court to be anticipated by the solution of DL-acl disclosed in Francis et al., even through a racemic compound such as DL-acl may dissociate in solution as disclosed by Karrer. A solution of L-acl contaminated by D-acl is not equivalent to a solution of pure L-acl, the latter being required for conversion into the desired end product, L-lysine.
With regard to the subject matter of claims 16 and 20, namely, the L-acl.L-pca salt, the Patent Office contends that this is the obvious or inherent product of an obvious process and therefore unpatentable under 35 U.S.C. § 103. While, as previously stated, this Court does not consider plaintiffs' process to be obvious in view of the combination of Dearborn et al. and Francis et al., it is noted that, even if it were considered obvious to a chemist of ordinary skill to attempt to resolve the DL-acl of Francis et al. with the L-pca of Dearborn et al., the obvious product of this hypothetical process according to the clear suggestions of Dearborn et al. would be the D-acl.L.-pca salt, rather than the L-acl.L-pca diastereoisomer which is actually precipitated in plaintiffs' process. Since the process itself is considered unobvious, and further since the salt precipitated would be unobvious even if the process were considered obvious, the L-acl.L-pca salt of claims 16 and 20 is deemed patentable over the references cited by the Patent Office. The L-acl.L-pca salt of claims 16 and 20 is useful as an intermediate in the production of the desired end product L-lysine, and thus the utility requirement of the patent law is satisfied. In re Nelson et al., 280 F.2d 172, 47 CCPA 1033 (1961).
Lastly, the Patent Office has rejected the two sets of product claims 16-18 and 20-22 as being duplicates of each other, relying upon 35 U.S.C. § 112, second paragraph. While this rejection is considered well taken with respect to claims 17 and 21 to L-acl and 18 and 22 to L-acl.HCl, the subject matter of these four claims is already considered obvious and unpatentable over the DL-acl and DL-acl.HCl of Francis et al. On the other hand, claims 16 and 20, which are considered to define unobvious and patentable subject matter, are not really duplicates of each other. Claim 16 is the broader claim and covers the L-acl.L-pca salt in either liquid or solid form. Claim 20 is narrower and defines the "crystalline" salt. Both claims 16 and 20 should, therefore, be allowed.
In summary, the subject matter of plaintiffs' process claims 12, 13, and 14, and product claims 16 and 20 is considered unobvious and patentable over the references relied upon by the Patent Office. Plaintiffs are therefore entitled to receive a patent for their invention, as specified in those five claims, and the Commissioner of Patents is authorized to issue such patent to plaintiffs on compliance with the requirements of law. Plaintiffs' Complaint is dismissed as to claims 15, 17, 18, 21 and 22.
The above Opinion includes Findings of Fact and Conclusions of Law pursuant to Rule 52(a) of the Federal Rules of Civil Procedure.
NOTES
[1] Subsection (e) of 35 U.S.C. § 102 is applicable since the Francis et al. patent issued subsequent to the filing date of plaintiffs' application, on an application filed prior to plaintiffs' filing date.