Plant Biomarker Recognition by Molecular Imprinting Based Localized Surface Plasmon Resonance Sensor Array: Performance Improvement by Enhanced Hotspot of Au Nanostructure

Detection of plant volatile organic compounds (VOCs) enables monitoring of pests and diseases in agriculture. We previously revealed that a localized surface plasmon resonance (LSPR) sensor coated with a molecularly imprinted sol-gel (MISG) can be used for cis-jasmone vapor detection. Although the selectivity of the LSPR sensor was enhanced by the MISG coating, its sensitivity was decreased. Here, gold nanoparticles (AuNPs) were doped in the MISG to enhance the sensitivity of the LSPR sensor through hot spot generation. The size and amount of AuNPs added to the MISG were investigated and optimized. The sensor coated with the MISG containing 20 μL of 30 nm AuNPs exhibited higher sensitivity than that of the sensors coated with other films. Furthermore, an optical multichannel sensor platform containing different channels that were bare and coated with four types of MISGs was developed to detect plant VOCs in single and binary mixtures. Linear discriminant analysis, k-nearest neighbor (KNN), and naïve Bayes classifier approaches were used to establish plant VOC identification models. The results indicated that the KNN model had good potential to identify plant VOCs quickly and efficiently (96.03%). This study demonstrated that an LSPR sensor array coated with a AuNP-embedded MISG combined with a pattern recognition approach can be used for plant VOC detection and identification. This research is expected to provide useful technologies for agricultural applications.