James C. Fang, Appellant, v. Elinor M. Hankins and William D. Emmons, Appellees, 399 F.2d 262 (C.C.P.A. 1968)

Fang appeals from the decision of the Board of Patent Interferences, which awarded to the senior party, Hankins and Emmons (Hankins), priority of invention of the subject matter of thirteen counts corresponding to claims 1-5, 7, 9, 10, 12, 14, 15, 17 and 19 of the Hankins patent.¹  Because Fang filed his application after the Hankins patent issued, he has the burden of proving priority beyond a reasonable doubt. Conner v. Joris, 241 F.2d 944, 44 CCPA 772, (1957).

The subject matter of the counts relates to certain monomeric acrylic acid and methacrylic acid esters of hydroxyalkyl oxazolidine compounds, polymers and copolymers of those monomers, and processes of preparing the monomers. Fang has presented testimony and documentary exhibits relating to activities beginning in March 1954 which, he contends, establishes reduction to practice of

(1) the specific monomers²  of counts 4, 5 and 7 — 3-methacryloxyethyl oxazolidine (hereafter MEO), 3-acryloxyethyl oxazolidine (AEO) and 3-methacryloxyethyl - 2 - spirocyclohexyl oxazolidine (MESO), respectively — all falling within the scope of count 1,

(2) the polymers and copolymers of those monomers set forth in counts 2, 3, 9, 10, 12, 14 and 15, and

(3) the processes of counts 17 and 19.

The board held with respect to some of Fang's evidence that it did not prove beyond a reasonable doubt that the materials and processes of the counts were actually prepared and carried out. In those instances where the board either found or assumed the compositions of the counts had been prepared, it held that Fang had not proved beyond a reasonable doubt that those compounds were useful at any time prior to Hankins' filing date. It is those holdings which frame the issues³  before us.

The story begins on March 30, 1954, when Nehf, Fang's assistant, reacted diethanolamine and cyclohexanone to form HESO, recording the results of that experiment on a page of Fang's notebook in evidence as Exhibit 6. Said the board:

* * * Exhibit 6 indicates that the HESO thus formed was refluxed with methylmethacrylate (and also with small amounts of hydroquinone and sodium) and that the material became "quite viscous" before the reaction was discontinued. With respect to this phase of the experiment Nehf testified:

Some MESO may have been prepared here since the reaction of the HESO with the methylmethacrylate did polymerize, and that would be what we would expect under the conditions that we ran this experiment.

It is noted that Exhibit 6 does not indicate that any product was recovered. Further, the record does not show that any analysis was made of any product that may have been formed. In our opinion the testimony of Nehf that "Some MESO may have been prepared" is insufficient to establish beyond a reasonable doubt, that MESO was actually formed. * * *

We find no error in that conclusion. Fang does not appear to press the matter seriously here.

On April 6-7, 1954, Nehf again reacted diethanolamine and cyclohexanone to form HESO, a known material in the prior art. About thirty six cubic centimeters of water was collected from the reaction mixture, corresponding to the theoretical amount of water expected if the reaction proceeded as planned. The product, identified by the code PRC-U93431 L-2 in Exhibit 7, was sent by Fang to Schwarzkopf Microanalytical Laboratory, an independent commercial laboratory, for a determination of the carbon, hydrogen and nitrogen content of the sample. The results, attested to by Drs. Otto and Francine Schwarzkopf, were consistent with the theoretical percentages of those elements in HESO. We may take the preparation of HESO as proved. Hankins does not argue otherwise.

The record shows nothing was done with that sample of HESO until some nine months later, when Gould, another of Fang's assistants, reacted HESO bearing the Code PRC-U93431 L-2 with methylmethacrylate and small amounts of hydroquinone, sodium and methanol. The record shows that the type of reaction sought to be carried out is known as "ester interchange" or "transesterification," in which the methyl alcohol component of the methylmethacrylate ester is "exchanged" for HESO to form MESO. Gould and Fang testified that MESO was prepared. The product was designated PRCE 19149 in Exhibit 12, a notebook record of the experimental procedure dated January 25, 1955, and samples thereof were submitted to Dr. Merrill for a carbon and hydrogen content determination and to Walter for a bromine number determination, the latter being a measure of the amount of unsaturated carbon linkages present. Both Merrill and Walter testified with respect to the procedural aspects of those analyses which are set forth in Exhibits 17 and 18. It appears the product was also analyzed for its saponification number (an indication of its ester content) and subjected to infrared analysis for a determination of its hydroxyl content, the results of which were recorded by Gould in Exhibit 16 along with the results of the carbon, hydrogen and bromine number analyses.

The board stated:

A controversy has arisen between the parties as to the hydroxyl content of PRCE 19149 as shown in Exhibit 16 and the significance thereof. It is stated in Exhibit 16 that "The hydroxyl content of this sample is in the same range as other samples submitted in this series (p-107, 108) according to I. R. analysis R. M. Z." The pages 107 and 108 appear as Exhibits 14 and 15. Gould testified that R. M. Z. were the initials of the analyst who reported the I. R. analysis. Page 108 of the notebook (Exhibit 15) includes a similar statement as to hydroxyl content and refers to page 107 (Exhibit 14). Exhibit 14 states as an objective "To determine purity of 5-methylol-5 ethyl-2-methyl-2-ethyl-1,3-dioxane" and includes the statement "No starting material (1,1,1-trimethylol propane) can be detected in the sample by I. R. analysis."

It is contended by Fang the I. R. analysis in Exhibit 14 clearly shows that the sample did not contain hydroxyl groups and the necessary conclusion is that the sample analyzed did not contain either the starting hydroxy compound nor the hydroxy-containing compound named at the top of the exhibit. It is to be noted that Exhibit 14 does not include any statement that the material is free from all hydroxy containing compounds. It is stated only that no 1,1,1-trimethylol propane, the starting material, can be detected by the I. R. analysis. In our opinion this statement cannot be construed to mean that no hydroxyl group is present when the objective was to determine the purity of a compound stated to have a methylol (a hydroxyl containing) group. We find no basis for concluding that none of the compound named in Exhibit 14 was present. Since Exhibit 16 indicates that the hydroxyl content of the sample of PRCE-19149 is "in the same range as" the sample of Exhibit 14 (stated to be a hydroxyl containing 1,3-dioxane compound) we agree with Hankins et al. that the logical conclusion would be that the sample of Exhibit 16 contains hydroxyl groups to substantially the same extent as the hydroxy compound of Exhibit 14. Gould testified * * * only that the hydroxyl content, as reported, is in the same range as other samples submitted in this series. It is conceded that MESO itself has no hydroxyl group.

Although both Gould and Chalmers, who signed Exhibits 14, 15 and 16 as a witness, testified (almost 10 years after the experiments and tests were performed) that it was their conclusion that the product tested (Exhibit 16) was MESO, it is our view that the showing in Exhibit 16 as to the hydroxyl content of the product raises a doubt as to whether the product PRCE-19149 was actually MESO. * * *

Here Fang urges the board erroneously assumed the PRCE-19149 product of Exhibits 12 and 16 contained hydroxyl groups without the benefit of any evidence that the products discussed in Exhibits 14-16 actually contained hydroxyl groups. Hankins points to a sentence⁴  in Exhibit 14 indicating that the product reported to have been analyzed therein had an actual hydroxyl percentage of 7.9% uncorrected for acid number as compared to a calculated value of 9.05%. We think the board was warranted in concluding that PRCE-19149 had a hydroxyl content "in the same range", viz. about 7.9%.

Alternatively, even if PRCE 19149 did have a hydroxyl content, says Fang, that fact is not necessarily inconsistent with that product being identified as MESO "because hydroxy compounds are obviously a common and natural impurity of MESO." He points out that the starting material, HESO, was an alcohol as was the solvent and theoretical by-product, methanol. It is his position that the infra-red analysis is not in conflict either with the testimony of Gould, Fang and Chalmers,⁵  all of whom concluded that MESO had been prepared, or with the results of the carbon, hydrogen and bromine analyses, all of which are said to correlate or be consistent with the identification of PRCE-19149 as MESO.

While we have given full weight to the evidence Fang has submitted to prove that PRCE-19149 was properly deemed to be MESO, his burden here, however, is to prove that fact beyond a reasonable doubt. Hankins has called our attention to other matters which we think have relevance in determining whether Fang has satisfied his burden of proof:

(1) Hankins questions whether the alleged HESO (PRC-U93431 L-2) made by Nehf nine months before Gould used it in the experiment of Exhibit 12 was in fact still HESO. Gould testified that he obtained that material somewhere "either in the laboratory, Dr. Fang's laboratory, or in a refrigerator." Hankins observes that there is no testimony as to the stability of HESO.⁶ 

(2) Dr. Merrill testified that the relative accuracy or maximum discrepancy between the carbon content actually found upon analysis and the theoretical carbon content should be about 0.2%. The discrepancy between the found and theoretical values for PRCE-19149 is three to four times that amount.

(3) An analysis for nitrogen content was regarded as important in order to distinguish between starting material, such as HESO, and final product, such as MESO. Chalmers testified:

Q25. * * * if you had products which you thought to be one or another of these compounds [MEO, AEO, or MESO] and analyzed them for carbon, hydrogen and nitrogen, which determination would be most important, if any, that is, the carbon, the hydrogen or the nitrogen, and why? A. In the analysis of MESO, MEO or AEO prepared from HESO or HEO as starting materials, the nitrogen analysis would be the most important since it would be the most severely affected by the presence of starting oxazolidine as an impurity. The change in the carbon and hydrogen content is of the order of 10 per cent whereas —

Q26. You mean if the reaction goes as planned? A. If the reaction goes as planned; — whereas the change in the nitrogen content is of the order of 30 per cent of the value, the final nitrogen value. If the reaction did not go as planned, this would be reflected in an adverse nitrogen value.

A similar conclusion reached by Fang with respect to identifying MEO would appear to apply equally as well to MESO:

Q378. Is the analysis for one of these elements more pertinent than that for the others in analyzing for this particular product? A. Yes. The particular important element in this analysis is the nitrogen analysis.

Q379. And why is that? A. Because it will then distinguish between HEO and MEO that we intend to obtain.

Q380. Could you explain further how the nitrogen analysis more clearly distinguishes those than the carbon and hydrogen analyses? A. In Exhibit No. 4 is shown clearly that MEO is a larger molecule than HEO. Therefore the percentage of nitrogen will differ markedly from MEO as against HEO.

There was no analysis for nitrogen content of PRCE-19149.

(4) As mentioned earlier, the saponification number was low, 131 in one determination and 151 in another as compared to 211 calculated. Exhibit 16 states that there was "(p)oor precision due to insufficient sample." Whose conclusion that was is not set forth. Dr. Merrill, who Fang says in his brief performed the saponification number determination, did not testify as to the reason for its inaccuracy.

Considering all the above factors, we agree with the board that Fang has not proved beyond a reasonable doubt that PRCE-19149 prepared by Gould was in fact MESO. For that reason, the process carried out by Gould cannot be regarded as an actual reduction to practice of the process of producing MESO recited in counts 17 and 19,⁷  nor can the copolymers which Gould made utilizing the alleged MESO be regarded as an actual reduction to practice of counts 2, 3 and 12. In view of our conclusion, we need not inquire further into whether the alleged MESO or the polymers made from it were useful.

Three and one-half years passed before further experimentation directed to a reduction to practice of the subject matter of the counts was conducted. Fang relies on certain experiments carried out by Mansi relating in general to the preparation of HEO, MEO and AEO, as well as polymers and copolymers of MEO and AEO, and also on Armour's experiments relating to the preparation of MESO. The experiments of Mansi which are of concern here were performed November 14-24, 1958, May 11-22, 1959, and June 15, 1959, while those of Armour were performed December 9, 1959.

For purposes of its decision, the board, with one exception, either assumed or found that Mansi and Armour did in fact prepare the monomers and polymeric materials of the counts, as recorded primarily in Exhibits 29, 31A-B, 41A-D, 43A-C, 45 and 47. It found, however, "no evidence in the record which could be considered as establishing that these polymers were useful for any particular purpose." The board also observed this court's decision in Blicke v. Treves, 241 F.2d 718, 44 CCPA 753 (1957), where it was stated:

* * * A composition of matter cannot be a patentable invention unless it has utility. * * * Accordingly, the invention of such a composition is not complete unless its utility is either obvious or is established by proper tests, regardless of whether the claims contain any specific reference to utility.

As this court has repeatedly stated, whether a composition of matter must be tested in order to establish a reduction to practice, and if so, what tests are necessary, is a question which must be decided on the basis of the facts of the particular case involved.

* * *

Said the board:

* * * We do not believe that the fact that a compound can be converted to a solid material (a polymer) without specifying what properties it may possess amounts to a showing of the utility of either the compound itself or the polymer. * * *

Fang does not deny that the record shows nothing further was done with the new and unobvious polymers and co-polymers said to have been prepared in the above exhibits. We cannot agree that the compositions of the counts were reduced to practice merely by making them in the present circumstances. Where, as here, no obvious usefulness satisfying the requirements of 35 U.S.C. § 101 would be apparent to one of ordinary skill from the record, we agree with the board that, without tests of an appropriate nature to determine the useful properties and to establish the usefulness of those polymers or copolymers, Fang has not proved that the new and unobvious subject matter of the counts was in fact useful beyond a reasonable doubt at any time prior to the filing date of the Hankins application fully describing the usefulness of that subject matter. Cf. Reiners v. Mehltretter, 236 F.2d 418, 43 CCPA 1019 (1956); Muskat v. Schmelkes, 140 F.2d 984, 31 CCPA 837 (1944).

Fang further relies on experiments conducted by Mansi in December 1958 and January 1959, the results of which were recorded primarily in Exhibits 36A-C, 38, and 40. The board analyzed with sufficient accuracy the testimony and documentary evidence relating thereto as follows:

It is also alleged that further preparations [of] HEO and MEO were carried out by Mansi on December 11 to 15, 1958, Exhibits 36A, 36B and 36C being the notebook record of these experiments. It appears that the product of these experiments designated by the code D49675 was subjected to analysis for carbon, hydrogen and nitrogen by the Schwarzkopf analytical laboratory and Mansi testified that from this analysis, he concluded that MEO had been prepared * * *. Mansi testified * * * that the product coded D49675 (MEO) was subjected to polymerization and copolymerization studies on December 17 and 18, 1958 and that solid materials were produced. Exhibit 38 shows that the MEO, and mixtures of MEO with methylmethacrylate and with styrene were heated with toluene and a catalyst, resulting in the production [of] viscous products in the case of MEO and mixtures of MEO with methylmethacrylate. Mansi testified that this work indicated that MEO copolymerizes with methylmethacrylate and that a homopolymer of MEO was formed.

Exhibit 40 is the notebook record of an experiment by Mansi performed on January 7, 1959 in which products of the above polymerization studies were coated onto "Bonderite 1000" panels and baked and the coated panels subjected to various tests. It appears that the panels were sent to "Physical Testing" for the Tukon hardness and "bump" test while Fang ran the curing tests * * *. It is urged that this work showed that polymers and copolymers of MEO were found to be excellent in several respects, and that Fang's work is corroborated by Mansi. Hankins et al. urge that since the person running the Tukon hardness tests did not testify the results of these tests are hearsay.

Since it is our conclusion that the tests results shown in Exhibit 40 do not establish that the alleged polymers possess utility, it is unnecessary for us to consider whether these results have been properly corroborated. Both Fang and Mansi testified * * * that the coating on panel number 1 was too thin and that a Tukon hardness test could not be run. It also appears that the coating on panel number 4 was also thin and did not pass the curing tests * * *. In the case of panel number 2 Mansi testified ". . . with the methyl ethyl Ketone it was not cured, or it wiped off completely." With respect to panel number 3 it appears that the coating when subjected to methyl ethyl ketone exhibited slight crazing or "a sort of chicken feet-like appearance."

We note that there is no indication that Fang or Mansi, at the time the tests were run, considered these tests as showing that the coatings were satisfactory for their intended purpose. We do not believe that the record demonstrates that, at the time of his alleged reduction to practice, Fang had a conviction of success, Ludwig v. Sohn et al., 51 CCPA 796, 801 O.G. 895, 1964 C.D. 127, 139 USPQ 500. In view of Fang's burden of proving his case beyond a reasonable doubt, we conclude that the utility of the alleged polymers shown in Exhibit 38 has not been established.

Our review of the record with due regard for appellant's arguments satisfies us that the board did not commit the reversible error of which Fang complains. We would simply add that it is well established that, when there is a reasonable doubt as to whether there has been an actual reduction to practice, the inventor's subsequent conduct may demonstrate that the acts relied on as an actual reduction to practice amount only to an abandoned experiment. Conner v. Joris, supra.

Here Fang's application in interference, filed after Hankins' patent issued, contained not only the claims of that patent but also every working example of that patent nearly verbatim. Curiously, Fang did not include in his application examples relating to the subject matter of Exhibits 36, 38 and 40. Fang also testified that, before he would make a suggestion that subject matter be patented, "I have to remain convinced that an idea is a worthwhile suggestion before I make such a proposal." At best, the earliest suggestion to patent the present subject matter was made after Hankins filing date. That is at least some indication, we think, that Fang did not regard the present subject matter "worthwhile" until that time, and that he did not have a conviction of success based on the evidence of record, particularly Exhibits 36, 38 and 40. There is, in our view, reasonable doubt that Fang reduced the subject matter of the counts to practice before Hankins' filing date.

One matter remains. Hankins has added certain material to the record. We regard approximately a third of that material to be properly part of the record and have considered it in reaching our decision, while the remaining material was unnecessary. Accordingly, one-third of the printing costs is assessed against appellant, and two-thirds against appellees.

The decision is affirmed.

Affirmed.

KIRKPATRICK, Judge, took no part in the decision of this case.