TAKEAWAY: Be deliberate and strategic about what you remove from a provisional application when drafting a non-provisional application. Omitted technical details, examples, or definitions that support a narrower interpretation may change how claim terms are construed.
The Federal Circuit’s decision in FMC Corp. v. Sharda USA, LLC (No. 24-2335, Aug. 1, 2025) addressed whether a district court properly relied on material that appeared in a provisional patent application but was deleted from the later non-provisional applications that issued as the asserted patents.
FMC Corp. (FMC) owned U.S. Patents Nos. 9,107,416 and 9,596,857, both derived from a provisional application (U.S. Prov. Patent No. 60/752,979) and directed to insecticidal compositions combining bifenthrin and a cyano-pyrethroid. FMC accused Sharda USA, LLC’s (Sharda) insecticide “WINNER” of infringement and obtained a preliminary injunction based on the district court’s finding that Sharda’s product likely infringed and that the patents were valid. Central to the dispute was the term “composition” in the claims, as FMC urged that it encompassed only stable formulations, while Sharda argued for the plain and ordinary meaning, which would include both stable and unstable mixtures.
In construing “composition,” the district court relied heavily on statements in the provisional application and in another non-provisional patent related to the provisional application (U.S. Patent No. 8,153,145) that emphasized physical stability. The provisional application, for example, noted that “[a] problem in the art… is successfully achieving physical stability of a water-diluted mixture,” and the ’145 patent described methods and testing for “dilution stability.” Although none of those statements appeared in the issued ’416 and ’857 patents, the district court imported them into its claim construction. The district court held that “composition” meant “stable compositions” based on the “stability” language, rather than any combination of the recited ingredients.
The Federal Circuit reversed that reasoning. The Federal Circuit emphasized that while the provisional application contained numerous references to “stability,” FMC had deliberately deleted every such reference when drafting the non-provisional applications that became the asserted patents. The court viewed that deletion as highly significant, citing DDR Holdings v. Priceline.com and MPHJ Tech. v. Ricoh for the principle that a patentee’s omission of material between the provisional and non-provisional filings shows an evolution in the applicant’s intended meaning. Because FMC “culled all such references from the common specification,” the court held that a skilled artisan would not read “composition” to be limited to stable mixtures. Thus, importing the deleted matter from the provisional application was legal error. The Federal Circuit vacated the preliminary injunction and remanded.
The FMC Corp. v. Sharda USA, LLC decision offers several practical lessons for patent prosecutors when transitioning from a provisional to a non-provisional application:
- Be intentional with deletions: Anything removed when converting a provisional to a non-provisional can be viewed as an intentional narrowing or change in scope.
- Don’t rely on deleted content: Courts may not import limitations or definitions from a provisional if they were omitted from the issued patent.
- Draft provisionals carefully: Even though provisional applications can be filed quickly and without much structure, their language can shape later interpretations