Patents

DuPont v. Synvina: Overlapping Ranges and the Burden That Shifts

The Federal Circuit reversed a PTAB win for the patentee, holding that a claimed range overlapping the prior art creates a presumption of obviousness that applies in inter partes review just as in court.

September 15, 2025
An industrial chemical reactor vessel with temperature and pressure gauges
The patent claimed reaction conditions — temperature, pressure, catalyst, and solvent — for making a bio-based acid. Shutterstock
Educational content, not legal advice. This article explains general legal concepts. It does not create an attorney–client relationship. For your specific situation, consult a licensed attorney.

E.I. du Pont de Nemours & Co. v. Synvina C.V., No. 2017-1977 (Fed. Cir. Sept. 17, 2018), is the modern reference point for the obviousness of claimed numerical ranges. In an opinion by Judge Lourie, joined by Judges O’Malley and Chen, the Federal Circuit reversed a Patent Trial and Appeal Board final written decision that had upheld Synvina’s U.S. Patent No. 8,865,921 over DuPont’s inter partes review (IPR) challenge. The Board had treated the claimed reaction conditions as nonobvious; the Federal Circuit held that those conditions were the result of routine optimization of known variables, and — significantly — that the well-established burden-shifting framework for overlapping ranges applies in IPR proceedings, not just in district court litigation or examination.

At a glance

  • Case: E.I. du Pont de Nemours & Co. v. Synvina C.V., No. 2017-1977 (Fed. Cir. Sept. 17, 2018).
  • Panel: Judges Lourie, O’Malley, and Chen; opinion by Judge Lourie.
  • Patent: U.S. Patent No. 8,865,921, directed to a method of oxidizing 5-hydroxymethylfurfural (HMF) to make 2,5-furandicarboxylic acid (FDCA) under specified ranges of temperature, pressure, catalyst, and solvent.
  • Posture: Appeal from a PTAB final written decision in inter partes review finding the challenged claims not unpatentable.
  • Holding: Reversed. The claimed ranges overlapped the prior art, creating a presumption of obviousness; the burden then fell to the patentee to rebut it, and Synvina failed to do so. The burden-shifting framework applies in IPRs.
  • Doctrinal core: When a claimed range overlaps a prior-art range, a prima facie case of obviousness typically exists, and the patentee must show criticality, unexpected results, teaching away, or that the variable was not result-effective.

The chemistry and the claims

FDCA is a bio-based building block — a renewable substitute for petroleum-derived terephthalic acid used in plastics like PET. The ‘921 patent claimed a method of oxidizing HMF to FDCA carried out within particular windows: a temperature range, an oxygen partial-pressure range, a cobalt-and-manganese catalyst with a bromide promoter, and a solvent. The individual ingredients and the basic oxidation chemistry were known; the patentee’s claimed contribution lay in operating the reaction within the recited conditions.

DuPont, which had its own FDCA process, petitioned for IPR, arguing the claimed conditions overlapped or sat adjacent to ranges disclosed in prior-art references describing the same oxidation chemistry. The Board sided with Synvina, reasoning in part that DuPont had not shown a person of skill would have arrived at the specific claimed conditions and faulting the petitioner for not proving the variables “predictably affected” the outcome. DuPont appealed.

Overlapping ranges: the presumption restated

The Federal Circuit began with a doctrine that long predates IPRs. Drawing on cases such as In re Peterson and In re Aller, the court reaffirmed that “a prima facie case of obviousness typically exists when the ranges of a claimed composition overlap the ranges disclosed in the prior art.” The logic is that where the art already discloses operating conditions that overlap the claim, a skilled artisan working in that space would naturally explore the overlapping window. Picking a value within a range the prior art already taught is, absent something more, ordinary optimization rather than invention.

The court connected this to the result-effective variable doctrine from In re Antonie. A variable is “result-effective” when the prior art recognizes that it affects the outcome of the process. Once a variable is known to influence the result, optimizing it within known bounds is presumptively routine. The Board had demanded too much from the petitioner — effectively requiring proof that each variable “predictably affected” the result. The correct question, the Federal Circuit explained, is whether the prior art recognized the variable as one that affects the outcome at all; temperature, pressure, and catalyst loading in an oxidation reaction plainly were such variables.

The framework applies in IPR

Synvina’s most consequential argument was procedural. It contended that the overlapping-ranges burden-shifting framework — under which a patentee must come forward with evidence of nonobviousness once an overlap is shown — could not operate in an IPR, because the petitioner bears the burden of proving unpatentability by a preponderance of the evidence and that burden never shifts.

The Federal Circuit rejected the dichotomy. Drawing on the production/persuasion distinction articulated in Dynamic Drinkware, LLC v. National Graphics, Inc., the court held that the burden of persuasion indeed remains on the petitioner throughout an IPR — but the burden of production can still shift. Once the petitioner shows that the claimed range overlaps the prior art, it has produced enough to support a conclusion of obviousness, and “the burden of production falls upon the patentee to come forward with evidence” that the overlap does not establish obviousness. The overlapping-ranges presumption is a tool for analyzing the evidence, not an improper reallocation of the ultimate burden. There was, the court said, no reason the same analytical framework that governs examination and litigation should not govern IPR.

Synvina’s failed rebuttal

Having set the framework, the court walked through the avenues open to a patentee facing an overlapping range and found each unavailing on this record. A patentee can rebut by showing (1) the claimed range produces unexpected results — a difference in kind, not merely degree, from the prior art; (2) the prior art taught away from the claimed range; (3) the parameter was not recognized as result-effective; or (4) the disclosure of the broad range was so broad as not to invite routine optimization. Synvina offered evidence of commercial and other indicia, but the court found no showing of criticality or genuinely unexpected results tied to the claimed conditions, no teaching away, and no basis to treat temperature, pressure, and catalyst as anything other than classic result-effective variables. The optimization the patent reflected was the kind a skilled chemist performs as a matter of routine.

Open questions

  • How “overlapping” must a range be? The opinion addresses overlapping and closely abutting ranges. Where the prior art is adjacent rather than overlapping, or where the claim selects a narrow sliver of a vast disclosed range, the strength of the presumption is more contested.
  • What proof of criticality suffices? The court demanded evidence that the claimed range is “critical” or yields unexpected results different in kind. The quantum and form of that proof — comparative data, the right baseline, the right nexus — continues to generate disputes.
  • How far does the production-shift reach? Synvina confirms the framework operates in IPR, but its interaction with the petitioner’s enduring persuasion burden invites careful management by the Board, especially where the patentee’s rebuttal evidence is mixed.

Implications

  • For IPR petitioners: Identifying an overlap between the claimed range and the prior art is a powerful, economical path to obviousness. Frame the disputed parameters as recognized result-effective variables and let the production burden shift.
  • For patent owners: A claim defined by numerical ranges is fragile if the art discloses overlapping conditions. Defend it with criticality and unexpected-results data developed during prosecution, or with a genuine teaching-away showing — generic commercial-success evidence rarely rebuts an overlap.
  • For prosecutors: When claiming ranges, build a record of criticality. Comparative examples showing a difference in kind at the claimed boundaries are the most durable protection against a routine-optimization attack.
  • For the PTAB: Do not require petitioners to prove that variables “predictably affected” results. The question is whether the art recognized the variable as result-effective; if so, optimization within overlapping bounds is presumptively obvious.

Frequently asked questions

Does an overlapping range automatically make a claim obvious? No. It creates a prima facie case — a rebuttable presumption. The patentee can defeat it by showing criticality or unexpected results, teaching away, that the variable was not result-effective, or that the prior-art range was too broad to invite optimization.

What is a “result-effective variable”? A process parameter that the prior art recognizes as affecting the outcome — temperature or pressure in a reaction, for example. Once a variable is known to affect the result, optimizing it within known bounds is treated as routine and therefore presumptively obvious.

Why does the burden-shifting matter in an IPR? Because the petitioner always keeps the ultimate burden of persuasion by a preponderance of the evidence. Synvina clarifies that this does not prevent the burden of production from shifting to the patentee once an overlapping range is shown — the same analytical tool used in examination and litigation.

Authorities and sources