concurring.
Because the alleged best mode in this case was an intermediate, not the claimed invention, the district court correctly concluded that the best mode requirement “does not compel disclosure of the unclaimed method.” Bayer AG, 129 F.Supp.2d at 721. On this basis, I would affirm. I write to underscore the district court’s correct application of the statutory test for best modes. The alleged best mode in this case does not fall within the scope of the claims. Therefore, this case simply does not require creation of a new test for best modes.
I.
Title 35 requires disclosure of “the best mode contemplated by the inventor of carrying out his invention.” 35 U.S.C. § 112 (1994). The most important words in this phrase are “his invention.” These words invoke the claims. Based on this direct statutory language, the bulk of this court’s precedent states that the disclosure necessary to satisfy the best mode requirement depends on the scope of the claimed invention. Teleflex, 299 F.3d at 1329-33; Eli Lilly, 251 F.3d at 963 (“[T]he extent of information that an inventor must disclose *1324depends on the scope of the claimed invention.”); see also ante, at 1315-16.
One of these precedents deserves special emphasis as the earliest Federal Circuit case to address the rule for identifying a best mode. DeGeorge was very clear that the Board erred by defining the claim too broadly and identifying an alleged best mode beyond the proper claim scope:
The board found no best mode in the DeGeorge applications. The board’s analysis, however, was influenced by its erroneous claim construction.... Because the properly construed count does not include a word processor, failure to meet the best mode requirement here should not arise from an absence of information on the word processor.
DeGeorge, 768 F.2d at 1325. Thus, this court’s earliest precedent explains that the claimed invention sets the bounds of the best mode inquiry. The DeGeorge court certainly did not inquire about a material effect on properties of the invention.
Moreover this emphasis on the claimed invention did not begin with the Federal Circuit. Indeed this court’s predecessor followed the same standard:
The claimed invention is a blend of uniformly random ethylene-methacrylic acid copolymer and polyethylene.... Appropriate inquiries into the best mode requirement should pertain to the contemplated mode of carrying out the invention, which in this case is the blend and not the starting materials.
In re Brebner, 455 F.2d at 1404 (footnote omitted) (emphases added). The Court of Customs and Patent Appeals limited the scope of “appropriate” best mode inquiries to the scope of the claims. Indeed the present case is very similar to Brebner. In Brebner, the claims did not recite the starting materials; here, the claims do not recite an intermediate. In both cases, “appropriate inquiries” into the best mode need not exceed the scope of the claims.
II.
With the “scope of the claimed invention” rule governing the identification of best modes, this court should have halted its analysis when the district court correctly applied that rule. Up to the point of acknowledging the claimed invention, this Bayer opinion reflects well the bulk of this court’s best mode jurisprudence. Then, inexplicably and without support in the statute or case law, this Bayer opinion widens its best mode net to capture the properties of the claimed invention and further sweeps in any material effect or impact on those properties.1
*1325A closer look at the policy and rationale for the best mode requirement discloses the disturbing implications of extending the best mode requirement beyond the reach of the claims. In most instances, the best mode requirement is self-enforcing. If an inventor does not disclose a critical trade secret within the best mode requirement, that nondisclosure puts the value of the entire patented invention at risk — a risk beyond the requirements of § 112. Competitors in the same technology can, and invariably will, discover the undisclosed trade secret and claim it in a separate patent application. When that application ripens into a patent, the competitor will have a blocking patent that could compromise much of the value of the original patent. Therefore, an informed patent applicant will never withhold a genuine best mode. Informed patent applicants will always either disclose the best mode in the original patent’s specification (often as a dependent claim) or, if the trade secret is not part of the claimed invention (as in this case), file a separate patent application on the separate innovation (also as in this case). Because informed patent applicants know to avoid best mode problems, this § 112 requirement is invariably little more than a trap for the uninformed applicant — usually a university or independent inventor without corporate legal resources. Because the best mode requirement is a trap for the unwary, the Federal Circuit has wisely followed the statutory “scope of the claimed' invention” rule to confine the reach of this snare.
When extended beyond the scope of the claimed invention, the best mode requirement becomes as insidious and destructive as a hidden land mine. One of the cases emphasized by this opinion, Dana, 860 F.2d 415, illustrates those disturbing implications. In Dana, the inventor claimed a seal apparatus, not any method at all, let alone a method of treating elastomeric material to ensure its longevity. Having invented a unique seal apparatus, the inventor could not have guessed that the best mode would reach out to encompass a process to increase the useful life of one component of the invention a process that was already well known in the prior art to boot. Dana, 860 F.2d at 419. Nonetheless, this dourt invalidated the patent because the undisclosed method affected the life of the elastomeric material and thus the satisfactory performance 'of the seal.2
At the outset, the Dana relationship test could sponsor a potentially boundless inquiry into any undisclosed method or property that could affect the satisfactory performance of the invention. Patent law in general is not concerned with the performance of an invention, let alone its satisfactory performance. Hildreth v. Mastoras, 257 U.S. 27, 34, 42 S.Ct. 20, 66 L.Ed. 112 (1921) (“The machine patented may be imperfect in its operation; but if it embodies the generic principle and works ... it is enough.”); Decca, Ltd. v. United States, *1326544 F.2d 1070, 1077, 191 USPQ 439, 444, 210 Ct.Cl. 546 (1976) (“The mere fact that the system has some drawbacks, or that under certain postulated conditions it may not work ... does not detract from the operability of the disclosed equipment to perform its described function.”); Nat’l Recovery Tech. v. Magnetic Separation Sys., 166 F.3d 1190, 1196, 49 USPQ2d 1671, 1676 (Fed.Cir.1999). Notwithstanding this bedrock principle of patent law, Dana would inexplicably make satisfactory performance the critical identifier for best modes.
Sadly this Bayer opinion incorporates Dana within its “material effects” test: “In Dana, ... the inventors failed to disclose subjective preferences that related to the use of the claimed inventions, and the undisclosed information materially affected the properties of the claimed inventions.” Ante, at 1319. Apparently this opinion, like Dana, would include all uses and properties of the invention within the best mode.
Indeed if the rule for identifying a best mode expands to capture potentially innumerable unclaimed “uses” and “properties,” as in Dana, the best mode requirement becomes a minefield for wary and unwary alike. At the outset, one might question the need for a new test. The Federal Circuit has identified best modes for twenty years without a material effect or properties test. Moreover, the Federal Circuit already has a claimed invention rule based on the language of § 112.
In expanding the best mode test to accommodate Dana, this new Bayer test unfortunately creates new conflicts with many cases in which this court found no best mode. Thus this new material effect test contravenes much of the calculus the Federal Circuit has employed in best mode cases. For instance, even this court’s most recent pronouncement on best mode law could not stand under the new test. In Teleflex, the claimed invention was a connector in an automobile shift cable that facilitated easier access for service. Instead of perceiving some new material effect test, Teleflex applied the prior binding Brebner-DeGeorge claimed scope rule to identify best modes:
The claims do not mention any particular material, hardness, or material matching for the clip. The claims do not mention any particular thickness for the clip. Thus, the information alleged to be part of the best mode and alleged to be missing from the disclosure is unclaimed subject matter.
Teleflex, 299 F.3d at 1332. Once again, in Teleflex, the claims specified a clip. One property of a clip is hardness. Another is thickness. Another is the metallic material for the clip. Under this new proposed Bayer test, these properties of the claimed clip (each with a material effect on its use and properties) would have triggered a best mode violation. Contrary to this proposed test, however, the Federal Circuit detected no alleged best mode within the scope of the claims and, for that reason alone, rejected the best mode allegations. Moreover this court in Teleflex examined nearly every case catalogued by this present opinion and found no new tests to identify best modes.
Another conflict with the proposed new test arises in Brebner, the earliest best mode case. 455 F.2d 1402. In Brebner, the inventor claimed a chemical compound used to process film. The inventor did not disclose the starting materials necessary to make the claimed compound. No doubt these undisclosed starting materials had a material effect on the properties of the invention because, without these starters, the invention could not be made. Nonetheless, this court’s predecessor did not apply a material effect test or examine the *1327compound’s properties, but simply determined that the claimed invention is “the blend and not the starting materials.” Id. at 1404. Under the material effect test, this necessary starting material would have led to a best mode violation. Again, by expanding the best mode test to accommodate opinions like Dana, this new test creates new conflicts with this court’s other precedents.
Another recent case, Eli Lilly, similarly shows the difficulty of expanding the best mode test. While discussing Eli Lilly, this court misses much of its significance. In Eli Lilly, the patents claimed a compound in Prozac and a method of using the compound. The alleged best mode violation was failure to disclose p-trifluoromethyl-phenol, the starting material necessary to make the claimed compound. The Federal Circuit found no best mode violation:
Neither patent ... claims p-trifluorome-thylphenol ... or a method for synthesizing it. Thus, while the best mode for developing fluoxetine hydrochloride involves use of p-tri[ ], the claimed inventions do not cover p-tri[ ].... As a result, the best mode requirement does not compel disclosure of [the inventor’s] unclaimed method for synthesizing p-tri[ ].
Eli Lilly, 251 F.3d at 964. With those words, the Eli Lilly case followed the “claimed invention” rule. Eli Lilly certainly did not assess the starting material’s “effect on the properties” of the invention. As a starting, material,, the “p-tri” had a direct effect on the claimed invention and certainly its properties and uses as well,but that did not compel a best mode violation.
This Bayer opinion characterizes Eli Lilly as within its new rule based on a few words in Eli Lilly that distinguish the starting material from the “intrinsic quality” of the invention. Of course a starting-material differs from the end product, but that is not the proposed test. This Bayer case’s bare characterization does not contend, nor could it, that a starting material does not materially affect the properties of the invention. Under the broad proposed test, Eli Lilly and many other prior cases might have reached a different result. In sum, this Bayer opinion’s test for “effects on properties” of the invention seems to have unspecified breadth.
This new test also conflicts with the Federal Circuit’s analysis in numerous cases. For instance, Northern Telecom applies the claimed invention rule and eschews a “material effect on properties” test to identify a best mode. In that ease, the patent claimed “a process for gaseous etching of aluminum and aluminum oxide that employs a gaseous trihalide.” N. Telecom, 215 F.3d at 1287. The district court concluded that the best mode of carrying out the invention was fine line etching. Because aluminum silicon alloy was “necessary” to achieve this etching, “the district court held that disclosure of aluminum silicon alloy was required to satisfy the best mode requirement.” Id. at 1286. Judge Clevenger corrected this misapplication of § 112:
As we have repeatedly held, the contours of the best mode requirement are defined by the scope of the claimed invention.... Northern Telecom asserts that the district court erred in requiring disclosure of a mode for carrying out something other than the claimed invention. ... We agree with Northern Tele-com .... What is claimed in claim 1 is a process for plasma etching of aluminum and aluminum oxide in the presence of a gaseous trihalide. Fine line etching is simply not part of the claimed invention.
Id. at 1286-87. Although the district court found that the undisclosed alloy had - a necessary effect on the use the invention, *1328this court found no best mode violation because it applied the scope of the invention rule. This court did not apply an effect on properties test.
In sum, a review of the totality of this Circuit’s best mode cases, as opposed to a few cases that found a violation, shows that this court does not use an “effect on properties” test to identify best modes, but instead uses a scope of the claimed invention test. Thus, to my eyes, this opinion fails in its effort to erect a new test that is, in any event, beyond the facts of this case.
III.
As I mentioned at the outset, this district court did its job well. It determined that the scope of the claimed invention did not include any intermediates. Because the defendants could not identify any alleged best mode within the scope of the claimed invention, the trial judge disposed of the best mode allegation as outside the statutory reference to the “invention.” In fact, as this court’s opinion notes, the inventor fully disclosed the intermediate, but not in this patent’s specification. Instead the inventor disclosed the intermediate in a separate patent application. The proposed best mode in this case was so far removed from the scope of the claimed invention that it was itself a separate invention.
Thus, the district court got it exactly right. It applied Federal Circuit law and deserves commendation. Instead, this court purports to use this easy case to erect a new best mode test. Fortunately, both this court’s failure to find a best mode in this case and the wealth of prior case law render this Bayer case mostly dicta. Otherwise, the next district judge encountering a best mode case would have to ask several imponderable questions: What is the Federal Circuit rule for the reach of the best mode rule? Even under this case, what is the test to identify a best mode— scope of the claimed invention, necessary relationship to performance of the claimed invention, or material effect on the properties of the claimed invention? What is a “property?” What is a “material effect?” How “material” is “material?” The district court correctly decided this easy best mode case. This court certainly did not need to plant any new traps in the best mode minefield.
. In dicta, this court has purported to apply the best mode beyond the scope of the claims: "[M]ost of the cases in which we have said that the best mode requirement was violated addressed situations where the inventor failed to disclose non-claimed elements that were nevertheless necessary to practice the best mode.” Chemcast, 913 F.2d at 927. Despite its language about "non-claimed elements,” the Chemcast claim recited a "locking portion ... more rigid than said base portion.” Id. at 925. Thus, the best mode violation in that case^ — a grommet of specified rigidity — was within the scope of the claims. Chemcast's comment about "non-claimed elements” was purely dicta. Contrary to the Chemcast dicta, an undisclosed feature only becomes a "best mode” candidate — by statutory definition — if it arises within the scope of the claimed invention.
In Chemcast, the inventor’s specification disclosed a grommet of "rigid castable resinous material ... for example, polyurethane or polyvinyl chloride ... [with] a durometer hardness reading of 70 Shore A or harder.” Chemcast, 913 F.2d at 929. No doubt the inventor thought that this disclosure was sufficient. After all, the inventor disclosed a preferred embodiment. Moreover, Chemcast Corporation’s actual product, R 4467, was indeed PVC with a hardness reading above Shore A 70 — consistent with that disclosure.
The Federal Circuit, however, perceived concealment — the second prong of the best *1325mode requirement. This court noted that R 4467 had a material hardness of 75, on the Shore D scale — "three hardness scales removed from the 70 Shore A hardness mentioned in the specification.” Id. at 930. Thicircuit further stressed that the inventors had "expended several months and many hundred man-hours” developing R 4467. Id. at 929. Chemcast is a concealment case, indeed an enthusiastic application of the concealment standard. In its zeal to reach the concealment issue, the Federal Circuit stretched its purported definition of best modes to encompass "unclaimed features,” but that certainly does not override prior binding case law.
. On' the basis of the inventor’s admission of the-merits of the fluoride treatment, the Federal Circuit quickly confined Dana to its peculiar facts. See Wahl Instruments, Inc. v. Acvious, Inc., 950 F.2d 1575, 1580, 21 USPQ2d 1123, 1126 (Fed.Cir.1991).