Hydranautics appeals from the August 30, 1991 judgment and order of the United States District Court for the Southern District of California which determined that FilmTec holds legal title to Cadotte U.S. Patent 4,277,344, the '344 patent is valid, and Hydranautics willfully infringed that patent. FilmTec Corp. v. Hydranautics, Case No. 90-563 GT (M) (S.D.Cal. Aug. 30, 1991). Because we conclude that the Saline Water Conversion Act of 1971 provides that the United States is entitled to ownership of the '344 patent, FilmTec cannot maintain the instant suit. We therefore reverse the district court's judgment and vacate its holdings that the '344 patent is valid and infringed.

This appeal involves a patent resulting from research performed under a government contract. In 1976, the Office of Water Research and Technology in the U.S. Department of the Interior entered into a research contract (the Contract) with the North Star Division of Midwest Research Institute (MRI), a not-for-profit research organization. The purpose of the Contract was "research and development on in-situ-formed condensation polymer membranes for seawater desalination by reverse osmosis."1 When subjected to high pressure, reverse osmosis membranes retain salt and allow water to pass through. Their performance capabilities are measured in terms of their flux and their salt rejection. "Flux" refers to the rate at which purified water passes through the membrane and may be stated in terms of gallons of water per square foot of membrane per day (gfd). "Salt rejection" refers to the percentage of salt that a membrane removes from treated water.

The government entered into the Contract with MRI under authority of the Saline Water Conversion Act of 1971 (the Act), Pub.L. No. 92-60, 85 Stat. 159, which had the broad purpose of developing means for desalinizing salt water so as to make it suitable for municipal, industrial, and agricultural uses. Section 6(d) of the Act provided that all patents resulting from contracts made pursuant to it, such as the Contract, will "be available to the general public." 85 Stat. at 161. Furthermore, Congress provided that "title to ... inventions made or conceived in the course of or under any contract or grant pursuant to the ... Saline Water Conversion Act of 1971 ... shall be governed by ... section[ ] 9 ... of the Federal Non-nuclear Energy, Research, and Development Act of 1974^ERDA...." Pub.L. No. 94-316, 90 Stat. 694 (effective date June 22, 1976). Section 9 of the FNERDA provides in pertinent part:

(a) Whenever any invention is made or conceived in the course of or under any contract of the Administration ... and the Administrator [of the Energy Research and Development Administration] determines that--

(1) the person who made the invention was employed or assigned to perform research ... and the invention is related to the work he was employed or assigned to perform, or that it was within the scope of his employment duties ...

title to such invention shall vest in the United States, and if patents on such invention are issued they shall be issued to the United States, unless in particular circumstances the Administrator waives all or any part of the rights of the United States to such invention in conformity with the provisions of this section.

(b) Each contract entered into by the Administration with any person shall contain effective provisions under which such person shall furnish promptly to the Administration a written report containing full and complete technical information concerning any invention ... which may be made in the course of or under such contract.

Pub.L. No. 93-577, 88 Stat. 1878, 1887-88. Section 9(g) of the FNERDA further provides that the Administrator may grant exclusive or partially exclusive licenses in any invention to which title is vested in the United States. 88 Stat. at 1889.

In light of Congressional intent to ensure that the government receive title to all patents resulting from research contracts authorized by the Saline Water Conversion Act, the Contract explicitly provided that MRI

does hereby grant to the Government the full and entire domestic right, title and interest in [any invention ... [conceived or first actually reduced to practice] in the course of or under this contract or any subcontract].

submit ... written information concerning the conception or actual reduction to practice ... of every invention made by [MRI] pertaining to the work called for in this contract which was conceived or first actually reduced to practice ... during ... the term of this contract, which invention would be a Subject Invention if made under this contract, but which the Contractor believes was made outside the performance of work required under this contract.

[f]ailure to furnish such information [ ] shall, in any subsequent proceeding, place on the Contractor the burden of going forward with the evidence to establish that such invention is not a Subject Invention. If such evidence is not then presented the invention shall be deemed to be a Subject Invention.

The Contract became effective on July 15, 1976 and extended through January 15, 1978. MRI hired John E. Cadotte to carry out the reverse osmosis membrane research contemplated by the Contract. On November 17, 1977, Cadotte reported in his MRI laboratory notebook the formation of three composite polyamide membranes, all made by the interfacial reaction of trimesoyl chloride (TMC), an aromatic acyl chloride having three acyl chloride groups, and metaphenylene diamine (MPD), an aromatic diamine, at 130? C. Cadotte tested two of these TMC/MPD membranes for their ability to desalinize seawater at 1500 psi. Cadotte found that these membranes exhibited fluxes of 12.7 gfd and 9.5 gfd with salt rejections of 95.0 and 88.5%, respectively. The record does not indicate that MRI reported this work to the government as required by the Contract.

In the summer of 1977, while he was still working at MRI, Cadotte and other MRI employees began to discuss forming a new company called FilmTec Corporation to develop new reverse osmosis membrane technologies. On December 20, 1977, E.E. Erickson, the president of FilmTec, sought a research contract for FilmTec with the Office of Water Research and Technology. In the research proposal that Cadotte helped to prepare, Erickson stated that former MRI staff members would carry out the proposed research program as a "follow-on" to the recent research at MRI under the Contract. Erickson further stated that "FilmTec was founded by [MRI] staff as a spin-off of MRI to conduct research and development and to manufacture products based on the polymer thin film technology that was developed at MRI."

Cadotte left MRI on December 31, 1977 and joined FilmTec as one of its four co-founders. On February 23, 1978, working in FilmTec's research laboratory, he reacted TMC and MPD and formed two composite reverse osmosis membranes. Cadotte's February experiments at FilmTec essentially duplicated his November experiments at MRI: he reacted the same proportions of TMC and MPD in the same manner for the same amount of time. The sole difference between the November and February experiments was that Cadotte reduced the membrane drying temperature from 130? C to 100? C. As a result, the TMC/MPD membranes that Cadotte produced in February exhibited fluxes of 13.1 and 11.0 gfd at a 98.0% salt rejection rate, better than those obtained the previous November.

Approximately one year later, on February 29, 1979, Cadotte filed a patent application relating to his interfacially polymerized reverse osmosis membranes. Cadotte assigned his rights in that application and any resulting patent to FilmTec, which duly recorded this assignment in the United States Patent and Trademark Office. Cadotte's patent application ultimately issued as the '344 patent, entitled "Interfacially Synthesized Reverse Osmosis Membrane." FilmTec currently has record title to the '344 patent. The government has not taken action to determine its rights to the '344 patent under the Contract.

Hydranautics manufactures two reverse osmosis membranes known as CPA-2 and HPM. HPM is formed by the interfacial condensation of, inter alia, TMC and MPD, while CPA-2 is also formed from TMC and MPD. Other reactants were also involved, but their presence is not material to the issue before us. Accordingly, FilmTec sued Hydranautics for infringement of claims 1 and 6 of the '344 patent. Hydranautics asserted in defense that under the Contract, the United States, not FilmTec, has legal title to the '344 patent and that FilmTec therefore lacks standing to maintain its infringement suit. The district court held that FilmTec has legal title to the '344 patent because Cadotte invented the claimed membrane after he commenced work at FilmTec, not while he was employed at MRI, and the '344 patent was not invalid and was infringed.2DISCUSSION

Hydranautics argues that Cadotte conceived the claimed invention during the term of the Contract and that, under the Contract, the United States has title to the '344 patent. According to Hydranautics, FilmTec cannot maintain the present suit. This case thus turns on whether the United States or FilmTec is entitled to ownership of the '344 patent. We conclude that the district court erred in deciding that the invention of the '344 patent was made while Cadotte was employed at FilmTec and in failing to correctly apply the provisions of the Contract, the Saline Water Conversion Act, and the FNERDA.

At the outset we address FilmTec's argument that Hydranautics cannot assert that the United States has title to the '344 patent. FilmTec argues that one cannot raise a third party's equitable claim of title as a defense to patent infringement. As support for its broadly-stated proposition, FilmTec points to Dorr-Oliver, Inc. v. United States, 432 F.2d 447, 165 U.S.P.Q. (BNA) 517 (Ct.Cl.1970). In Dorr-Oliver, the record title holder of a patent sued the United States for patent infringement under 28 U.S.C. § 1498. The United States asserted as a defense that the plaintiff was not the "owner" of the patent during the period of infringement because plaintiff's employer, AMF, had a superior equitable claim of title pursuant to an employment contract requiring the plaintiff to assign all patents to AMF. The court permitted the United States to assert its title defense, but stated:

In patent litigation between private parties, equitable rights of ownership of strangers to the suit cannot be raised as defenses against the legal titleholder of a patent.

Id. at 451, 165 USPQ at 517 (citing Sigma Eng'r Service, Inc. v. Halm Instr. Co., 33 F.R.D. 129 (E.D.N.Y.1963)) (emphasis added). Despite this statement, the court in Dorr-Oliver permitted the defendant to raise its title defense. It did so because of the anti-assignment statute which "prohibit[ed] the assignment of any claim for patent infringement against the United States which arose prior to the time a plaintiff acquired title to the patent." Id. at 450 n. 2 (citing Brothers v. United States, 52 Ct. Cl. 462, 466 (1917), aff'd, 250 U.S. 88, 39 S. Ct. 426, 63 L. Ed. 859 (1919)).3 The court considered that, if a third party was the lawful title-holder, then the present plaintiff could not recover from the government. Dorr-Oliver in fact supports, rather than undermines, Hydranautics' position in this case since the Court of Claims in Dorr-Oliver applied a federal statute to preclude the record title-holder from asserting a patent infringement claim.

In this case, the district court did not address the controlling provisions of the Saline Water Conversion Act, the FNERDA, or the Contract. Section 9(a) of the FNERDA clearly provides that title to any invention made or conceived under a FNERDA contract "shall vest" in the United States. There is no doubt that Cadotte's MRI work was funded by a contract subject to Section 9(a) of the FNERDA. Thus we only need to decide whether the invention of the '344 patent was made or conceived while Cadotte was at MRI, a matter that we will address infra.

However, our decision concerning FilmTec's inability to maintain the present suit is also based on the fact that Hydranautics was a third party beneficiary of the Contract. Section 6(d) of the Saline Water Conversion Act, the legislation under which the Contract was let, provides:

All research within the United States contracted for ... under authority of this Act, shall be provided for in such manner that all ... patents ... resulting from such research developed by Government expenditure will ... be available to the general public [emphasis added].

85 Stat. at 161. Because, as this statute makes clear, Congress intended that inventions made under the Contract be available to the "general public," of which Hydranautics is a member, it would be contrary to the intent of Congress to permit FilmTec to preclude Hydranautics from practicing such an invention. Hydranautics is clearly a member of the general public to whom inventions made under the Contract must be available.4 We now consider whether the district court erred in determining that the invention of the '344 patent was not made at MRI.

The district court's opinion dealt mostly with questions of infringement. However, it held that Cadotte did not invent the claimed membranes under the Contract because his November experiments, as evidenced by his MRI laboratory notebook, did not "embody the '344 invention." Specifically, the district court concluded that since the November experiments did not result in a membrane with the flux and salt rejection rate limitations of the '344 claims, Cadotte did not invent the '344 membrane during the term of the Contract. The court cited Cadotte's testimony that he did not find the results of his November experiments particularly encouraging, and that he did not even remember those experiments when he reacted TMC and MPD again in February at FilmTec after leaving MRI.

Hydranautics argues that the district court erred as a matter of law by resolving the issue of title to the '344 patent according to the patent claims rather than according to the terms of the Contract. We agree. As we determined above, under the Contract the government has title to all inventions on in situ condensation polymer membranes conceived or first actually reduced to practice by Cadotte at MRI between July 15, 1976 and January 15, 1978. We conclude that the invention of the '344 patent was conceived while Cadotte was at MRI and that the district court erred in not so holding.

Conception is a legal determination that we review de novo. Hybritech Inc. v. Monoclonal Antibodies, Inc., 802 F.2d 1367, 231 U.S.P.Q. (BNA) 81, 87 (Fed.Cir.1986) (citing Barmag Barmer Maschinenfabrik AG v. Murata Machinery, Ltd, 731 F.2d 831, 837, 221 U.S.P.Q. (BNA) 561, 565 (Fed.Cir.1984)). Conception is the "formation in the mind of the inventor, of a definite and permanent idea of the complete and operative invention, as it is hereafter to be applied in practice." Hybritech, 802 F.2d at 1376, 231 USPQ at 87 (citing 1 Robinson On Patents 532 (1890)).

The claimed invention of the '344 patent is a membrane material which is distinguished by the polyfunctional acyl halide (TMC) and polyamine (MPD) components of the polyamide desalinizing layer.5 Cadotte prepared membranes from these reactants in November 1977 at MRI. He tested these membranes for their ability to desalinize seawater by reverse osmosis and found that they exhibited fluxes of 12.7 gfd and 9.5 gfd with salt rejections of 95.0 and 88.5%, respectively.6

FilmTec argues that Cadotte's November 1977 membranes did not constitute conception of the claimed membranes because their flux and salt rejection were below the limits stated in the claims and below that required for commercial use. However, those facts will not defeat a finding that Cadotte conceived the '344 invention within the meaning of the Contract. The claimed invention is a composition, claimed by its structure, the key limitations of the claims being the chemical reactants, and Cadotte conceived that invention while at MRI.7 Cadotte made the membranes with the stated reactants, and the results sufficed to show that Cadotte's invention went beyond the minimum requirements of operability; they actually worked to desalinize seawater. Since commercial success is not required for a reduction to practice, it certainly is not required for a conception. See DSL Dynamic Sciences, Ltd. v. Union Switch & Signal, Inc., 928 F.2d 1122, 1126, 18 USPQ2d 1152, 1155 (Fed.Cir.1991) (citing King Instrument Corp. v. Otari Corp., 767 F.2d 853, 861, 226 U.S.P.Q. (BNA) 402, 407 (Fed.Cir.1985), cert. denied, 475 U.S. 1016, 106 S. Ct. 1197, 89 L. Ed. 2d 312 (1986)).

The fact that the materials may not have met the performance recitations of the claims is also not conclusive. The membranes were surely within the structural scope of the claims. The key distinction between the invention of the '344 patent and the prior art is the presence of TMC, which has three acyl groups in the acid chloride component of the reaction as opposed to two such groups in the prior art. Cadotte relied on the distinction between di- and polyfunctional acyl halides in the prosecution of his patent application, arguing that a triacyl halide has superior crosslinking ability due to its extra crosslinking. At trial, Cadotte underscored the importance of TMC by testifying that his discovery of its ability to produce crosslinking, sidebranching, and a "super" molecular structure was an important contribution to his discovery of the reaction. Cadotte stated:

There are many, many, many amines available, likewise, many, many acid chlorides. And over that long time, I know I had done hundreds, maybe thousands, of different combinations of acid chloride reactions and polyamine reactions. So it turns out this reaction is actually very unique.... I have been using both [TMC and MPD] for some time ... and there was nothing in there that suggested to me that if I brought these two reagents together, that they'd form [the] extremely superior membrane that actually resulted.

It is clear that the invention conceived in November 1977 at MRI is the invention claimed in the '344 patent. The claims cover a large number of materials, no doubt having varied performance characteristics. Cadotte may well have refined the invention when he went to FilmTec, and he may have then found the best conditions for making his compositions. The record is clear, however, that his work in February 1978 was on the same invention that he conceived in November 1977. Were we to find that inclusion of narrow performance limitations in the claims could serve to expel the claimed invention from operation of the Contract under which the invention was made, we would be defeating the intentions of the parties to the Contract, who agreed that inventions made thereunder belong to the United States.8

The district court noted that Cadotte did not find the results of his November work "particularly encouraging," that in February he did not remember the November experiments, and that he was a credible witness. The court speculated that Cadotte might not have remembered what led him to the February experiments and that he did not keep the invention to himself so he could take it with him to FilmTec.

We will not invade the province of the district court to judge matters of credibility, difficult as it is for us to understand how February experiments on the same compositions were not a direct sequel to experiments three months earlier. What we do conclude, however, is that in November 1977, Cadotte conceived reverse osmosis membranes made from the interfacial reaction of TMC and MPD and showed that they performed their "intended function beyond a probability of failure." This is the subject matter of both the Contract and the invention of the '344 patent, notwithstanding the difference in performance limitations between language of the patent claims and the products of the November experiments. Cadotte therefore conceived the '344 invention under the Contract. The United States, by virtue of the Saline Water Conversion Act and the FNERDA, is entitled to ownership of the invention.

FilmTec argues that Cadotte was not bound by the Contract, and that Cadotte, as the inventor of the claimed invention, had title to the '344 patent. This is not the point since the provisions of the legislation governing the Contract control. Under the FNERDA, title to inventions made by an employee hired to perform that research automatically vested in the United States. The Contract required MRI to include a provision in its employment contracts with employees hired to do research under it that the government would retain title to all inventions conceived or reduced to practice during its term. Cadotte conceded that it did so and that he used government support and MRI's facilities in conducting his research. Thus, when the invention was conceived by Cadotte, title to that invention immediately vested in the United States by operation of law. He had no right to assign it to FilmTec; the statute had divested him of all of his interest. Although the formalities of the statute, including issuance to the United States, were not complied with, that omission does not change the effect of the law. As noted earlier, the Administrator could have waived title to the invention, or granted an exclusive or partially exclusive license to any party, including MRI, FilmTec, or Hydranautics. However, the Administrator has not done so. The result, therefore, is that title remained vested in the United States. In light of our disposition of this case, we need not review the district court's holding that Hydranautics infringed the '344 patent.9

Because we conclude that Cadotte conceived the claimed invention in November 1977, within the term of the Contract, and under the terms of the governing legislation title vested in the United States, FilmTec cannot maintain this suit for patent infringement.

Costs to Hydranautics.

REVERSED.