Early detection of lithium-ion battery (LIB) electrolyte leakage in the vapor phase is important for
battery safety, yet vapor-phase surface-enhanced Raman spectroscopy (SERS) remains challenging
because weak gas–surface interactions limit analyte residence within plasmonic hot spots. Here, we
report a hybrid graphene/plasmonic SERS platform for vapor-phase detection of LIB electrolyte
components based on periodic Ag nanoparticle (AgNP) multimer arrays integrated with a monolayer
graphene overlayer. The substrate is fabricated by capillary-assisted particle assembly (CAPA)
followed by a unified poly(vinyl alcohol) (PVA)-assisted hot-press transfer process, enabling both
the transfer of ordered AgNP arrays to glass and spatially selective graphene integration. This
approach preserves nanoscale ordering while creating a four-region architecture on a single chip,
allowing the individual and combined contributions of graphene and the plasmonic array to be
evaluated under identical vapor-exposure conditions. Optical characterization shows a broadband
plasmonic response dominated by interparticle coupling within AgNP multimers, with spectral overlap
across the 532 nm excitation and Raman-scattering window. Upon exposure to vapors from a commercial
LiPF6 electrolyte containing ethylene carbonate (EC) and ethyl methyl carbonate (EMC), no
analyte-attributable Raman features are observed from bare glass, graphene on glass, or the AgNP
array alone. In contrast, the graphene-coated AgNP region yields clear vapor-phase Raman signatures
assignable to both EC and EMC. These results show that detectable vapor-phase electrolyte signatures
emerge only from the combined graphene–plasmonic architecture, consistent with a hybrid interfacial
effect in which graphene may increase the local surface population of volatile molecules while the
AgNP multimers provide localized electromagnetic enhancement. This work establishes a scalable
hybrid-transfer strategy for ordered vapor-phase SERS substrates and highlights graphene-coated
plasmonic arrays as promising material platforms for molecularly specific LIB leak detection.