Opinion
Appeal No. 81-517.
August 20, 1981.
William H. Vogt, III, Morris N. Reinisch, Marcus J. Millet, New York City, for appellants.
Joseph F. Nakamura, Sol., and Gerald H. Bjorge, Associate Sol., Washington, D.C., for Patent and Trademark Office.
Appeal from the Patent and Trademark Office Board of Appeals.
Before MARKEY, Chief Judge, and RICH, BALDWIN, MILLER and NIES, Judges.
This is an appeal from the Patent and Trademark Office ("PTO") Board of Appeals ("board") affirming the examiner's rejection of claims 1-31 under 35 U.S.C. § 103 as unpatentable over Nutting et al. ("Nutting") in view of either Rector or Aeillo in application serial No. 742,094, filed November 16, 1976, for "Decaffeination Process," which is a continuation-in-part of application serial No. 605,717, filed August 18, 1975, which is a continuation-in-part of application serial No. 527,870, filed November 27, 1974. We reverse.
United States Patent No. 3,361,571, issued January 2, 1968.
United States Patent No. 1,716,323, issued June 4, 1929.
Aeillo, "The Distribution Coefficients of Diuretics and Narcotics, and the Theory of Narcosis," Biochem. Z., 124, at 192 205 (1921).
The Invention
The invention relates to processes for preparing a decaffeinated vegetable material, such as coffee or tea, suitable for use in preparation of beverages. Prior art decaffeination techniques employ conventional, potentially toxic solvents, such as chloroform, carbon tetrachloride, and trichloroethylene, as the decaffeination medium. The problems with these solvents are that they can affect the flavor and remove desirable constituents of the beverage.
Appellants teach decaffeination with the use of edible fats. These fats may include soybean oil, corn oil, olive oil, coffee oil and others. Appellants accomplish decaffeination by contacting a caffeine-containing vegetable material, in either solid or aqueous extract form, with a liquid, water-immiscible fatty material. The fatty material removes caffeine from the vegetable material. The caffeine-laden fatty material is then separated from the decaffeinated vegetable material by centrifuging or decanting.
The following claims are exemplary:
1. A process for producing a decaffeinated vegetable material suitable for use in preparation of beverages, which comprises:
a) contacting a caffeine-containing vegetable material with a liquid, water-immiscible fatty material capable of removing caffeine therefrom;
b) maintaining said vegetable material and said fatty material in contact for a time sufficient to transfer caffeine from said vegetable material to said fatty material; and
c) separating the decaffeinated vegetable material from the caffeine-laden fatty material.
9. A process for producing a decaffeinated coffee or tea suitable for use in preparation of beverages, which comprises:
a) contacting caffeine-containing coffee or tea in extract or solid form with a liquid, water-immiscible fatty material capable of removing caffeine therefrom;
b) maintaining contact between said coffee or tea and said fatty material for a time sufficient to transfer a part of the caffeine from said coffee or tea to said fatty material;
c) separating the partially decaffeinated coffee or tea from the caffeine-laden fatty material;
d) contacting the partially decaffeinated coffee or tea with additional fatty material capable of removing caffeine therefrom;
e) maintaining contact between the partially decaffeinated coffee or tea and the additional fatty material for a time sufficient to transfer a further amount of caffeine from said partially decaffeinated coffee or tea to said additional fatty material; and
f) separating the further decaffeinated coffee or tea from the caffeine-laden additional fatty material.
21. A process for producing decaffeinated coffee suitable for use in preparation of beverages, which comprises:
a) extracting green coffee beans with water to transfer caffeine from the beans to the water;
b) separating the caffeine-laden aqueous phase from the beans;
c) contacting the caffeine-laden aqueous phase with a liquid, water-immiscible fatty material capable of removing caffeine therefrom, selected from the group consisting of safflower oil, soybean oil, corn oil, peanut oil, coffee oil, triolein, olive oil, and lard;
d) maintaining the caffeine-laden aqueous phase and the liquid fatty material in contact for a time sufficient to transfer caffeine from the aqueous phase into the fatty material;
e) separating the caffeine-liberated aqueous phase from the caffeine-laden fatty material;
f) recycling the caffeine-liberated aqueous phase to the same or different green beans to remove caffeine from the beans; and
g) regenerating the caffeine-laden fatty material by removing caffeine therefrom to permit its reuse for decaffeination of aqueous extracts of green coffee beans.
The Prior Art
Nutting discloses a conventional decaffeination process in which solvents, preferably trichloroethylene, are utilized to separate caffeine from the other coffee constituents. Prior to decaffeination, however, Nutting requires that a de-oiling step be performed to remove coffee oil from the materials undergoing the decaffeination process.
Rector discloses a method of coffee beverage preparation in which roasted coffee beans are ground with oil and pulverized to form a powder which is useful as a flavoring material. Rector states:
In practicing the present invention, the roasted coffee is first ground to any required fineness, even to the point of impalpability, in an edible oil or fat. By grinding the coffee in an oily medium a much finer product results than when the coffee is ground alone. The fatty medium should preferably be a neutral oil, i.e., an odorless and tasteless oil, thereby avoiding any undesirable odor or taste in the product. If desired, the grinding medium may consist of an oil charged with coffee substance, such as oil expressed from coffee.
The ground paste is next placed in a press and the surplus of oil removed. Any desirable percentage of oil may be left in the product, but it is preferable to allow from twelve to fifteen percent to remain, as this is the usual percentage of fat in the coffee bean.
The cake which is left in the press may then be reduced to a powder in an ordinary pulverizer. This powder, being very finely divided, is largely soluble and can be used for making coffee, as a flavoring material, or for many other uses. It can be combined with malted or powdered milk and sugar to form a prepared beverage. The insoluble portion of the coffee is so finely ground that it remains in suspension in the beverage in an unobjectionable form.
The oil which was expressed from the paste is now used for grinding a fresh batch of coffee, and is again expressed from the paste. This cyclic process is carried on as long as desired. Successive extractions leave the oil more heavily charged with the aromatic and stimulative elements of the coffee. Thus the process becomes more efficient as it continues, until a point is reached where the oil is practically pure coffee oil, and it then continues at maximum efficiency. If it is preferred not to use all the expressed oil to grind other batches of fresh coffee, it may be used as a flavoring constituent in itself, the oil having a fine, fresh, coffee flavor without any of the fibrous constituents of the coffee bean.
The paste produced by grinding the coffee in oil may also be used as a flavoring material or in producing confections. [Emphasis supplied.]
Aeillo discusses the lipoid theory of narcotics and disputes that theory on the basis that it rests upon an inaccurate experimental premise. Aeillo states:
The lipoid theory of the narcotics was formulated more than twenty years ago by Hans Horst Meyer and by Overton simultaneously and independently.
It is based, not on the special chemical properties of the compounds and their narcotic effects, but on the physical-chemical behavior of the narcotic compounds, i.e. their solubility in fatty oils, in particular distribution coefficients of these compounds between water and oil. According to it, the effect of the narcotic agents depends on the distribution coefficients between oil and water. The tables prepared by Hans Horst Meyer of threshold old values of the narcotic agents and the distribution coefficients between oil and water show the threshold values of narcotic substances inversely proportional to the distribution coefficient. . . .
. . . .
. . . [O]ne must argue against the experimental part of the theory, on which, after all, the entire theory rests, that it is built upon a conclusion by analogy rather than an experiment which corresponds to the actual conditions. In fact, Baum's experiments, which Loewe already criticized as being partially inadequate, refer, not to the behavior of the narcotics and to the partition coefficients between blood and brain lipoids, but to the solubility ratio of the narcotics between water and oil. Now water is an entirely different solvent than blood or serum, as is known from many facts. . . . On the other hand . . . it is not the fatty masses in the body that the narcotics act on, but the brain and the rest of the nervous system. Also, the water-insoluble fat differs greatly from the complicated mixture of the diverse lipoids. . . .
Our statements are intended only to show, to begin with, that to infer from the system water-oil to the system blood-lipoid mixture-protein is contrary to the facts, and that the actual situation is quite different.
. . . .
It was of great interest to us to investigate the behavior of another group of compounds; in particular we wanted to see if the group of diuretics is lipoid-soluble, what the distribution coefficient is between oil and water, and what differences result when one phase, water, is replaced by serum. With the experience we had thus gained, it was logical to extend this modification, water replaced by serum, to some soporifics. . . . [I]t was of great interest to see how such exciting and diuretic substances, which certainly are not narcotics but their physiological antagonists, behave in the test arrangement of Hans Horst Meyer and Baum, and how these results change when one solvent, water, is replaced by serum, studying both the solubility of the compounds and the distribution coefficient in relation to oil. None of these agents can be claimed to be a soporific. On the contrary, caffeine is an antagonist of the soporifics, and we know that in caffeine intoxication ample quantities of caffeine are found in the brain, so that there can be no doubt that the brain is highly receptive for caffeine. We conducted our experiments as follows:
Method
[To avoid deviations that occur at higher concentrations, Aeillo uses a 1% caffeine solution. Of this solution 50 cc was mixed with 50 cc of purest olive oil. The mixture was machine shaken for two hours until emulsified and foamy. The liquid was allowed to settle for six hours and then was centrifuged. The remaining traces of oil were removed by adding salt and filtering.]
. . . As this aqueous solution contained salt and we could therefore not obtain its weight by evaporation of the solution, we extracted it exhaustively in the shaking machine with chloroform, or better still with a continuous extractor for solvents heavier than water, and completely evaporated the filtered chloroform solution in a weighed crystallizing dish. . . . After shaking and centrifuging, we took 10 cc of the clear solution, admixed them with solid common salt, exhaustively extracted them with chloroform after filtering, concentrated this chloroform solution in a weighted dish and weighed the dish again. We then converted to the total volume, and from the difference between the content of the original solution and the content of the solution after shaking with oil we determined the quantity passed into the oil.
For caffeine there resulted the partition coefficient:
0.1000 — 0.0651 = 349 = 0.53. _______________ ___ 0.0651 651
. . . This having been established, we had to investigate now how the distribution coefficient varies when water is replaced by serum. For this purpose the methodology had to be changed as follows:
[A mixture of 50 cc serum and 50 cc olive oil was shaken for two hours, allowed to stand all night, centrifuged until clear and filtered. The caffeine was assayed by acidifying the serum with dilute acetic acid. If a clear filtrate was not obtained, then common salt was again added and rinsed with boiling water; if the volume of water was too great, then the mixture was concentrated and extracted exhaustively with chloroform. The partition coefficient between serum and oil was found to be .18; therefore, caffeine was found to be more soluble in serum than in water.]
It is clear from this experiment that the determination of the partition coefficient between oil and water is far from representing the real conditions in the organism, that a change in one solvent, replacement of water by the more realistic serum, completely alters the numerical values . . . . It is clear from these experiments that, whether the theory is correct or incorrect, the numerical experimental material cannot be used in support of this theory . . . .
. . . [I]t must be mentioned that according to the experiments of Verkade (9) that solubility of substances in olive oil is an isolated phenomenon and does not offer analogies to any other fat. Especially the differences between olive oil and castor oil with respect to the solubility of salicylic acid and benzoic acid are huge; they amount to a multiple of 4 to 6. . . .
The Board Opinion
The board stated:
. . . It is apparent that the main difference between the presently claimed decaffeinating process and that described in Nutting, for example, lies in the particular organic solvent utilized as the decaffeinating medium. However, we are convinced that it would have been obvious within the meaning of Section 103 to utilize a liquid water-immiscible fatty material as the decaffeinating medium for producing decaffeinated vegetable materials in view of the disclosures in the Rector and Aiello references.
Rector, for example, clearly teaches that a liquid fatty material will extract the aromatic and stimulative elements from roasted coffee. Such is particularly evident from the disclosure at page 1, lines 59-61 wherein patentee states that "successive extractions leave the oil more heavily charged with the aromatic and stimulative elements of the coffee." This teaching clearly indicates that the stimulative elements, which obviously includes caffeine, are soluble in the fatty oils utilized by Rector. In fact, the Rector reference is deemed to fairly suggest the process steps recited in claim 1 as patentee teaches contacting a caffeine containing vegetable material with a liquid water-immersible [ sic] fatty material, page 1, lines 25 to 29, for a time sufficient to transfer caffeine (stimulative elements) from the vegetable to the fatty material and subsequently separating the decaffeinated vegetable material from the fatty material containing caffeine.
It is also apparent from the disclosure in the Aiello reference that caffeine is soluble in olive oil, a fatty material within the scope of the present claims. . . .
Since it is readily apparent from the foregoing teachings in Rector and Aeillo that caffeine is soluble in fatty materials, we are convinced that it would have been obvious to one of ordinary skill in the art to utilize such materials to remove caffeine from the caffeine-containing materials of Nutting. Appellants have merely substituted one known caffeine solvent for another such solvent and have only obtained the expected results. . . .
OPINION
We are persuaded that the board has misinterpreted the disclosure of Rector, has erroneously considered a nonanalogous publication (Aeillo), and has failed to consider the teachings of the references as a whole.
Rector produces a beverage flavoring material by using an edible oil or fat as a "grinding medium" which achieves a "much finer product . . . than when the coffee is ground alone." The resulting oily paste is then pressed to remove all but 12-15% of the oil so that "the usual percentage of fat in the coffee bean" is retained. The oil extracted by pressing is then used to grind successive batches of roasted coffee. The board and the examiner have focused on the Rector statement, "Successive extractions leave the oil more heavily charged with the aromatic and stimulative elements of the coffee." The board states that this teaching "clearly indicates that the stimulative elements . . . are soluble in the fatty oils utilized by Rector." This is speculation. See In re Rice, 481 F.2d 1316, 1318, 178 USPQ 478, 479 (Cust. Pat.App. 1973). Moreover, Rector's statement must be read in full context. He also states: "Thus the process becomes more efficient as it continues, until a point is reached where the oil is practically pure coffee oil, and it then continues at maximum efficiency." (Emphasis supplied.) Thus, Rector is concerned with a physical extraction (grinding and pressing) process and not a chemical (solvent) extraction process. He does not suggest that oil is "capable of removing caffeine," as required in appellants' process.
Although the Rector process would operate immediately at highest efficiency if pure coffee oil were used, Rector states that the process can be started with other types of oils. As each batch of coffee beans is ground to form a cake, the residual oil from that batch contains a higher percentage of coffee oil to be used in each subsequent batch. After a number of batches have been processed, the surplus oil will be essentially pure coffee oil, and, as stated by Rector, the process then "continues at maximum efficiency."
We regard Aeillo as nonanalogous art, which cannot properly be considered pertinent prior art under 35 U.S.C. § 103. In In re Wood, 599 F.2d 1032, 1036, 202 USPQ 171, 174 (Cust. Pat.App. 1979), this court stated:
In resolving the question of obviousness under 35 U.S.C. § 103, we presume full knowledge by the inventor of all the prior art in the field of his endeavor. However, with regard to prior art outside the field of his endeavor, we only presume knowledge from those arts reasonably pertinent to the particular problem with which the inventor was involved. [Citation omitted.] The rationale behind this rule precluding rejections based on combination of teachings of references from nonanalogous arts is the realization that an inventor could not possibly be aware of every teaching in every art. Thus, we attempt to more closely approximate the reality of the circumstances surrounding the making of an invention by only presuming knowledge by the inventor of prior art in the field of his endeavor and in analogous arts.
The determination that a reference is from a nonanalogous art is therefore twofold. First, we decide if the reference is within the field of the inventor's endeavor. If it is not, we proceed to determine whether the reference is reasonably pertinent to the particular problem with which the inventor was involved.
Both the instant claims and Nutting involve decaffeination of vegetable materials; whereas, Aeillo compares the solubility of a diuretic solution, such as a caffeine solution combined with an oil/serum mixture, to the same solution combined with an oil/water mixture. He determines that caffeine is "more soluble in serum than in water." From this he concludes that the Meyer/Overton lipoid theory of narcotics, which was based upon experiments using an oil/water mixture, is inaccurate because an oil/water mixture does not approximate the substances found in the human body. Thus, Aeillo's disclosure is not "within the field of the inventor's endeavor." Further, Aeillo is not pertinent to appellants' problem because he is not concerned with either beverage preparation or decaffeination of vegetable materials. There is no common environment which could form a "close relationship" between either the claimed invention or Nutting on the one hand and Aeillo on the other to logically require consideration of Aeillo. In re Antle, 58 CCPA 1382, 1387, 444 F.2d 1168, 1171-72, 170 USPQ 285, 287-88 (1971). An earlier statement by this court in In re Van Wanderham, 54 CCPA 1487, 1494, 378 F.2d 981, 988, 154 USPQ 20, 25 (1967), is particularly appropriate:
Our determination here is not without difficulty. However, we think the difficulty arises from not considering the subject matter as a whole and instead focusing on the scientific principle involved . . . .
In this case, the board erred by focusing on the affinity of olive oil for caffeine without considering the subject matter of Aeillo as a whole and the impropriety of the Aeillo reference, as pointed out above.
The board has relied on "isolated teachings of the prior art without considering the over-all context within which those teachings are presented." In re Mercier, 515 F.2d 1161, 1166, 185 USPQ 774, 778 (Cust. Pat.App. 1975). Thus, the Nutting reference is concerned with decaffeination using a conventional, oil-free process: coffee oil is removed, and then the caffeine is extracted, using toxic solvents such as chloroform, carbon tetrachloride, and trichloroethylene. Because Nutting teaches removal of oil before decaffeination, he teaches away from the concept of using oil to decaffeinate. Although Rector involves beverage preparation, he is not concerned with the problems associated with decaffeination of beverages or vegetable materials. Consequently, he does not suggest either appellants' problem or solution. In re Rice, 481 F.2d at 1318, 178 USPQ at 479. Even assuming that a person skilled in the pertinent art would have considered Aeillo, this reference does not teach or suggest decaffeination of vegetable materials as required by the claims. Aeillo uses a 1% caffeine solution — not vegetable materials. Further, he notes that the caffeine solution is three times more soluble in an oil/serum mixture than in an oil/water mixture. This teaches away from the use of oil alone as a caffeine extraction medium. There is no suggestion that oil can be used to extract caffeine from vegetable materials or that it is desirable to use oil alone to extract caffeine even in solution.
In view of the foregoing, we hold that the teachings of the references, considered as a whole, would not have rendered appellants' invention obvious.
The decision of the board is reversed.
REVERSED.
My review of appellants' claims and the references relied upon by the PTO convinces me that the decision of the board holding that the claimed invention would have been obvious to one of ordinary skill in the art is correct.
Claim 1 broadly sets forth a process for decaffeinating vegetable material by "contacting" the material (in ground, crushed, or whole form) with liquid oil for a time sufficient to "transfer" caffeine to the oil, which is then "separated" from the vegetable material. No patentable distinction is seen between such a process and the Rector process which teaches grinding coffee with oil and pressing out the oil which, Rector states, is charged "with aromatic and stimulative elements of the coffee." The contrast drawn by the majority between "physical extraction" in Rector and "chemical extraction" in appellants' process is wholly without support in claim 1.
Moreover, although in certain dependent claims, such as claim 8, the volatiles (i.e., aromatics) are removed prior to caffeine extraction to avoid possible loss in the oil used as the extractant, claim 1 contains no such limitation. Thus, both caffeine and volatiles may be removed in the very same manner as in Rector.
Although Rector is not directed toward the production of decaffeinated coffee, this does not negate his teaching that caffeine is transferred to the oil. Rector specifically teaches:
Successive extractions leave the oil more heavily charged with the aromatic and stimulative elements of the coffee. . . . If it is preferred not to use all the expressed oil to grind other batches of fresh coffee, it may be used as a flavoring constituent in itself.
To use the oil itself as a coffee flavoring, clearly an appreciable quantity of "aromatic and stimulative elements" in the coffee must have been transferred to the oil.
Rector must be combined with Nutting to render obvious those embodiments of the claimed process in which an aqueous extract of the vegetable material is decaffeinated. This combination is tenable since both references deal with processes for the preparation of coffee products which result in caffeine-laden by-products. Nutting employs an organic solvent to decaffeinate an aqueous extract from which the volatile constituents have first been removed (to save for later restoration and thus prevent their loss during decaffeination) and which has also been de-oiled. The purpose of this de-oiling is not disclosed. Contrary to the majority's reasoning, the teaching of the removal of original oils before decaffeination does not teach away from substitution of oil for organic solvent in the decaffeination step. Accordingly, I find no contraindication to the use of Rector's edible oils to extract caffeine in the Nutting process.
The by-products are the above-mentioned flavoring oil produced by the Rector process and the solvent after use in the Nutting process, from which he states the caffeine may be recovered and sold.
Finally, I do not find the Aeillo reference, submitted for consideration by appellants, to be nonanalogous art. The majority fails to note that one of the problems indicated by appellants in the use of the prior art solvents, e.g., trichloroethylene or chloroform, is that such "solvents themselves are often detrimental" (to health), concern, therefore, being directed to their use in production of comestibles. This being the problem and coffee being a well-known diuretic, one of ordinary skill in the coffee art would be led simply by the title of the Aeillo article, which indicates it deals with distribution coefficients of diuretics, to consider Aeillo's research.
No distinction can be made between appellants' starting material and Aeillo's use of a 1% caffeine solution in his tests, as drawn by the majority. Appellants' aqueous extracts of the caffeine-containing vegetable material are of unspecified caffeine content. Accordingly, the article must be evaluated for what it teaches a person skilled in the art, namely, that caffeine is soluble in olive oil. While a lower value of solubility is found for oil as compared to serum, (i.e., 0.18) than for oil as compared to water (i.e., .53), both results show solubility of caffeine in oil. As specifically stated therein:
[F]rom the difference between the content of the original solution and the content of the solution after shaking with oil we determined the quantity [of caffeine] passed into the oil.