This is an appeal from an affirmance by the Board of Appeals of the rejection of claims 3, 4, and 7-9, which are all the claims remaining in appellants’ application serial No. 11,964, filed March 1, 1960, for “New Epoxy Compounds.”
The invention relates to epoxy esters of carbonic acid which may be termed di(3,4-epoxyhexahydrobenzyl) carbonates [equivalent name: di-, or bis-(3,4-epoxycyelohexylmethyl) carbonates].1 Those epoxy carbonate esters are light-colored monomers that are liquid or fusible at room temperature and can be cured with the usual hardeners for such polyepoxy *753(polyepoxide) compounds. Mixtures containing epoxy compounds of the invention and a hardener can be used as textile assistants, laminating resins, paints, lacquers, dipping resins, casting resins, fillers, putties, and adhesives, and are especially valuable as insulating compounds for electrical purposes. With reference to the groups giving rise to the particular nomenclature used hereinafter, the claimed epoxy carbonate esters may be depicted thus:
Z=a hydrocarbon radical of a polyalcohol or a polyphenol.
In order to appreciate the differences between the compounds claimed and those of the prior art, a short discussion of a method of preparing a compound representative of those claimed will be helpful. The claimed epoxy carbonate esters may be prepared by starting with a ring unsaturated cyclic alcohol, e. g.:
to which a carbonate group is added (by
O
reaction with phosgene, Cl-C-Cl, or a carbonic acid ester) to form the ester linkage by a known method. The ring double bonds are then epoxidized,
(with peracetic acid) to produce the claimed epoxy monomers.
As noted above in the diagram of the claimed compounds, there may be a single carbonate group linking two cyclohexylmethyl, [hexahydrobenzyl], rings, or, optionally, there may be an additional ester
0
link of the nature -Z-O-C-O- , the Z-link, inserted between the rings. If in the reaction which adds the carbonate group to the ring unsaturated cyclic alcohol, 2 moles of the alcohol are used, a compound containing a single carbonate link is formed, as, for example, in claim E:
3. The compound of the formula
*754If only one mole of alcohol is used, the resulting compound is the corresponding chloroformate:
Then the Z-link is added by reaction of with polyhydroxy compounds, e. g., bis-a proper amount of the chloroformate phenol A:
, producing, for example, the compound of claim 7:
7. The compound of the formula
"Generic cláim 9 covers both types and reads:
9. Epoxy compounds of the general formula
in which Ri and R5 taken together and R'i and R'5 taken together each form a member selected from the class consisting of two hydrogen atoms, two lower alkyl radicals of 1 to 4 carbon atoms and one methylene radical, R2, R'2, R3, R^, R4, R'a, Re, R/6, Rt, R7?, Rs, R's, R9 and R/9 each represent a member selected from the class consisting of hydrogen atoms and lower alkyl radicals of 1 to 4 carbon atoms, Z is a hydrocarbon radical with 2+(n-l).(p-l) free valences and is of the group *755consisting of the hydrocarbon radical of a polyalcohol and the hydrocarbon radical of a polyphenol, n represents a whole number of at least 1 and at the most 2, and p represents a whole number of at least 1.
Claims 4 and 8 differ from claims 3 and 7 respectively in that one hydrogen on the number 6 ring carbon is replaced by a methyl group.
The following references are relied on for the rejection of claims 3, 4 and 7-9 under 35 U.S.C. § 103:
Phillips et al. (I) 2,750,395 June 12, 1956
Mueller et al. 2,795,572 June 11, 1957
Newey 2,848,426 Aug. 19, 1958
Phillips et al. (II) 2,917,491 Dec. 15, 1959
A second ground of rejection involves generic claim 9 which was rejected “as failing to properly define the invention” within the meaning of 35 U.S.C. § 112. We do not reach that issue.
The Mueller et al. patent (Mueller) is directed to epoxy esters of carbonic acid and their polymers, which are thus of the same class of compounds as those disclosed by appellants. As do appellants, Mueller discloses and claims epoxy carbonate esters having either one or two carbonate ester linkages between alcohol moieties. Mueller’s epoxy esters are prepared in the same manner as appellants’. In the latter type, the two carbonate groups may be separated by a radical derived from di-nuclear dihydric phenols, e. g., bisphenol A, to give the following compound:
By comparison with appellants’ claim 7 it will be noted that only the terminal epoxy-containing moieties, derived from the starting reactant alcohols, are different.
Regarding the alcohol moiety of the Mueller epoxides, Mueller teaches and claims epoxyalkanols, epoxyalkoxyalkanols, expoxycyeloalkanols, and epoxyalkoxycycloalkanols. Within each of those classes of alcohols Mueller names preferred members. Pertinent here is the disclosure that:
Particularly preferred epoxy-substituted alcohols are the * * * epoxyeycloalkanols * * *“ and particularly those containing not more than 12 carbon atoms, such as * * * 3,4-epoxycyclohexanol * * * and the like.
As epoxy carbonate esters produced therefrom,2 Mueller specifically discloses:
The preferred epoxy esters of the invention are those derived from * * * (2) epoxy-substituted aliphatic and cycloaliphatic monohydric alcohols containing from 3 to 12 carbon atoms, such as for example, * * * di-(3,4-epoxycylohexyl) carbonate, * * *.
*756The structure of that compound is:
It can be seen by comparing that structure with the structure of the compound of appellants’ claim 3, that the carbonate group indirectly links the two rings through methylene bridges in appellants’ compound, while in Mueller’s compound the carbonate group is directly bonded to the rings.
The remaining references were cited to show the specific alcohol reactants used by appellants which would, when used in Mueller, result in epoxy carbonate esters having a methylene bridge, and, as called for in claims 4 and 8 and covered in generic claim 9, a methyl or other aliphatic group on the number 6 carbon of the cyclohexyl ring. All three secondary references are concerned with epoxy monomers (also termed epoxides) useful, as Phillips et al. (I) states, “in certain commercial applications such as, for example, surface coatings, laminates and plastic molding compositions.” Newey states that his epoxides may be cured with what are apparently well-known curing agents, and that when the epoxides
* * * are to be used in the preparation of castings or pottings the curing agent and the epoxy material are generally combined together and then poured into the mold or casting which may if desired contain additional material, such as electrical wires, apparatus, etc., and the mixture then heated to effect the cure.
Newey also discloses his epoxy compounds to be useful in adhesive and impregnating compositions, and that when amines are reacted with the epoxides “high molecular weight products instead of cross-linked cured products,” useful in making coating compositions, will be produced.
Exemplary alcohol reactants listed in Newey “include, among others, (3,4-epox-ycylohexyl) methanol * * * [and] 3,4-epoxycyclohexanol * * It will be recognized that those two alcohols are the epoxidized forms of appellants’ and Mueller’s unsaturated alcohols; Newey also discloses that the corresponding unsaturated alcohol is used as the starting material, to be subsequently epoxidized with the same reactant as used by appellants. The compound of Newey’s Example III shows the use of the 3,4-epoxycyclohexyl methanol:
*757The major difference between Newey and Mueller is that, like appellants, Mueller is directed toward epoxy carbonate esters while Newey’s epoxides are epoxy ethers, and thus do not contain the —0-0 II C-0- linkage. Also, while both Phillips references are directed to epoxides they are esters of dicarboxylic acids rather than of carbonic acid; they contain an 0 0 II II -O-C-R-C-O- linkage, rather than a 0 0 II II -O-C-O-Z-O-C-O linkage between the epoxycyclohexylmethyl rings.
I The Prima Facie Case of Obviousness
We agree with the board that upon these references the examiner has made a prima facie showing that the claimed epoxy carbonate esters are obvious. The board stated:
We agree with the Examiner that the Phillips et al. and Newey patents make the substitution of a 3,4-epox-ycyclohexyl-methyl alcoholic component for the 3,4-epoxycyclohexyl component of the Mueller et al. esters an obvious expedient. Mueller et al. describe the epoxy-substituted aliphatic and cycloaliphatic monohydric alcohols as being preferred epoxy-substituted alcohols in preparing their esters of carbonic acids
Newey’s disclosure of the equivalent or alternative use of epoxycyclohexyl and epoxycyclohexyl-methyl radicals as the terminal reactive radicals in his disclosed polyepoxides would suggest the alternative use of these radicals, with or without the substituents that Phillips et al. show to be conventional, as the terminal radicals in the Mueller et al. compounds.
It is quite clear from the references that appellants’ starting reactant alcohol is well-known in the art as a starting reactant to make other epoxy monomers useful for the same purposes as those disclosed by appellants. While the epoxides of Newey and the Phillips references are either ethers or carboxylic acid esters, rather than being carbonic acid esters as in Mueller and appellants’ application, it is clear that those differences are of no moment since all three types of compounds will undergo identical reaction with the identical reagent for the purpose of epoxidizing the cyclohexyl ring double bond. Further, from the record it appears that the three types of epoxy compounds have the same general properties as reflected by the same general end use. It is common ground that it is the epoxy group in all which is similarly affected by the same hardeners or curing agents.
We fully appreciate the fact that the alcohol reactant of Mueller is a secondary alcohol, and that of Newey and appellants is a primary alcohol. We fail to find this so significant as is urged since Mueller teaches that alcohols which may be used to produce epoxy carbonate esters may be either primary or secondary. Certainly primary and secondary alcohols are not to be taken as equivalents for all purposes, but from the above teaching of Mueller it would appear that the art suggests the use of the primary alcohol of Newey in place of the secondary alcohol of Mueller, for the purpose of making epoxy carbonate esters. This is particularly the case since Newey teaches the equivalence of the alcohol of Mueller and that of appellants as starting reactants.
Appellants also argue that there is no working example in Mueller of the di-(3,4-epoxycyclohexyl) carbonate, and that disclosure is merely the name of a compound “tucked away” among a number of other compounds. We cannot agree that in order to be a valid teaching for reference purposes only so much as is shown in the examples of a patent is available as prior art. Examples in a reference are merely that, exemplary of the broader disclosure, all of which is available for what it clearly teaches.
*758II The Affidavit
Appellants rely on an affidavit of Nikles, one of the joint inventors, as rebutting the examiner’s prima facie showing of obviousness. That affidavit purports to compare the Mueller epoxide, di3,4-epoxycyclohexyl carbonate, with appellants’ epoxide, di-3,4-epoxycyclohexylmethyl carbonate.' After preparation of the respective epoxides, samples of each are mixed with phthalic anhydride and tetrahydrophthalic anhydride hardeners. Two samples of each mixture are identically cured for 16 hours at 120°C, and for 5 additional hours at 160°C. Then “thermic aging” tests are performed. These tests consist of placing the sample moldings into an oven heated to 220°C. Both loss of weight and appearance of the moldings are determined after 5 hours and after 10 hours in the oven, with the following results:
Epoxy resin A (Mueller)=Di(3,4-epoxycyclohexyl) carbonate
Epoxy resin B (Appellants)^Di(3-4epoxycyclohexylmei/ii/J) carbonate
Regarding the appearance of the moldings, the affidavit states:
Moldings No. 1, 2, 5, 6 (from epoxy resin A) after 5 hours at 220°C already show extensive stress cracks and are dark colored, while
moldings No. 3, 4, 7 and 8 (from epoxy resin B) have remained intact.
From the facts shown in the affidavit, appellants conclude that any possible presumption of obviousness is rebutted, stating:
* * * The affidavit compares the thermal stability of the two compounds as represented by the loss of weight of moldings or castings obtained therefrom, which is a standard test for such products. The results obtained show that the loss of weight of the Mueller et al compounds is ten times greater than the loss of weight in applicants’ product, and further shows that in the Mueller et al product, the moldings show extensive stress cracks and are dark-colored, while the moldings from applicants’ compounds remain intact. The affidavit is clearly indicative of the unique properties of applicants’ products over the prior art compound which the Examiner regards as closest to applicants’ compounds. * * *
*759The board gave four reasons for according little weight to the affidavit:
* * * It is apparent, though, that: (1) this affidavit selects the phthalie anhydride and tetrahydrophthalic anhydride derivatives as a basis for comparison and that these compounds are only a small fraction of the possible derivatives described in the specification and patent; (2) the affidavit compares these derivatives as to a property which appellants’ specification does not even disclose as a general characteristic * * *; (3) these derivatives are prepared from only one of the compounds claimed, although the specification provides no disclosure that the derivatives of the compounds claimed are alike as to the characteristic tested; and (4) the breadth of claim 9 appears to make it very probable that compounds covered by claim 9 differ more from each other than they do from the compounds of the prior art.
The second objection does not seem well taken since Example 4 of the specification relates to an epoxy compound, in which the Z link is propyl, which is cured with phthalie anhydride and subsequently tested for thermal stability, the application reporting “Thermal stability according to Martens: 170°C.” 3 We also agree with appellants that the board upon reconsideration withdrew reasons 3 and 4 with respect to all but claim 9. In our view, however, we do not reach, with respect to claim 9, either the objections on points 3 or 4, nor the above noted rejection based on undue breadth.
The solicitor relies on In re Druey, 319 F.2d 237, 50 CCPA 1538, 1542; In re Surrey, 319 F.2d 233, 50 CCPA 1336, 1339, and In re Finley, 174 F.2d 130, 36 CCPA 998, as governing the result of this case with regard to the affidavit. In each of those cases the appellants attempted to rely on affidavit showings as establishing unexpected results or properties for the compounds claimed in order to overcome the prima facie case of obviousness as shown by references. This court held the affidavits to be insufficient to rebut the showing of obviousness since the compounds tested were end-products whereas the compounds claimed were intermediates. The solicitor points out that here epoxy monomers are the subject of the claims while a cross-linked, cured product is the subject of the affidavit tests, and thus, following the Druey, Surrey and Finley cases, the affidavit cannot be considered to rebut the prima facie case of obviousness based on the references.
We do not think those cases are to be read to establish a hard and fast rule that evidence of unexpected properties of the end-product is wholly irrelevant, and therefore inadmissible, on the issue of non-obviousness of an intermediate. While both the Druey and Surrey cases were primarily concerned with the relevance of the affidavits therein,4 we think that is the case because a sulfa drug or analgesic, the subject matter sought to be patented, is of such a nature as to be unpredictable in its action. Thus, considering that the action of drugs is generally highly specific, it can fairly be said that tests on the end-product drugs may not be relevant, and thus inadmissible, to show non-obviousness of intermediates. Clearly, the proffered evidence must run a greater risk of failing the relevancy test *760of admissibility in proportion to its remoteness from the claimed subject matter. Whether too remote is a question of law to be determined on the basis of the subject matter as a whole.
In contrast, the court in Finley appears to have been primarily concerned with the weight or materiality rather than the relevance of the tests on the end-product salt of the claimed ester, since the court noted the affiant averred the test on the salt was “a good measure of the stability of the [claimed] esters to heat.” Thus we think this court in stating, 174 F.2d at 135, 36 CCPA at 1005:
* * * we do not feel, viewing the record in the light of the authorities cited [which did not consider that unexpected results could rebut a prima facie showing of obviousness], that the grant of the patent sought would be proper * * *,
was indicating that the tests were either not material or not of such weight as to rebut the prima facie case of obviousness.
Here, we are satisfied that one of ordinary skill in this art would ascribe to the monomers the contributing cause for the difference in the cured products. Also, the end-product here does not itself react with anything else to show the unexpected property; the test of “thermic aging” or thermal stability is a test of the product itself rather than one of reactivity in and with a highly specific environment. Further, the cited art, in disclosing “potting” as a use of the epoxy monomers, suggests the property tested to be relevant in showing a relationship between the monomer and the cured end-product, since potting is well-known as encapsulation of electrical and electronic products in which thermal stability and freedom from cracking is important.
We think the unexpected result of less loss in weight and freedom from cracking during “thermic aging” 5 to be of such weight as to effectively rebut the prima facie case of obviousness based on the references, since nowhere therein are there any indications that such change in end-product properties would be expected to flow from the added methylene bridge in the monomer. Thus we reverse the decision of the board as to claim 3.
However, we do not view the affidavit as carrying claims 4 and 7-9. Claims 4 and 8 call for compounds having a methyl group on the number 6 carbon of the epoxycyclohexylmethyl ring. Claims 7 and 8 are specific to compounds in which a bisphenol A-derived radical separates the two carbonate groups linking the rings. While such variance in structure is suggested by the art to be prima facie obvious, that showing manifestly is not rebutted by a relevant affidavit which relates to the cured end-product of a compound which has no such methyl group or Z-link, and only one carbonate grouping. Thus we affirm as to claims 4, 7 and 8.
Further, we think it clear that the showing with respect to the cured end-product of claim 3 alone is of insufficient weight to overcome the prima facie showing of obviousness of the numerous epoxy monomers encompassed within generic claim 9. Thus we affirm as to claim 9.
In accordance with the above view, it is unnecessary to consider the board’s remaining objections to the affidavit, or to consider the separate rejection of claim 9 as unduly broad. The decision of the board as to claim 3 is reversed, and that as to claims 4 and 7-9 is affirmed.
Modified.