Resolving competitive adsorption through d-orbital tuning in plasmonic layered double hydroxides for multiplexed SERS

Multicomponent volatile organic compound (VOC) analysis remains challenging because competitive adsorption on sensing interfaces produces overlapping signals and obscures molecular fingerprints. Here we report a cation-directed plasmonic layered double hydroxide (LDH) platform in which secondary metal cations (Ni, Cu, or Co) modulate the d-orbital configuration of interfacial cationic sites, thereby differentiating molecule–surface interaction preferences and enabling multiplexed VOC detection. Density functional theory and time-resolved surface-enhanced Raman scattering (SERS) spectroscopy reveal distinct affinities, with Ni favoring π-conjugated aromatics, Cu interacting with amine- and nitro-substituted aromatics, and Co targeting carbonyl aromatics through Lewis acid–base interactions reinforced by hydrogen bonding. This cation-directed tuning enables functional group-specific signal amplification, yielding up to 10-fold enhancement in equilibrium SERS intensities compared with undoped LDHs. By integrating the three LDH variants into a multi-affinity array and coupling it with a two-stage transfer learning framework, models trained using pure-analyte spectra are adapted to previously unseen binary-to-quaternary mixtures without retraining. We achieve 100% classification accuracy with ≤6% deviation from nominal Raoult’s law-derived concentrations across eight representative aromatic VOC mixtures. This work demonstrates cation-directed electronic structure modulation in LDHs as a design principle for modulating plasmonic interfaces and achieving selective molecular recognition among representative aromatic VOCs. Volatile organic compound mixtures are difficult to analyze because competing adsorption obscures molecular fingerprints. Here, the authors tune metal cations in layered double hydroxides for selective SERS recognition and multiplexed sensing.