Preparation of molecularly imprinted polymer nanobeads for selective sensing of carboxylic acid vapors

Highlights: • Molecularly imprinted nanobeads prepared for sensing of typical carboxylic acid vapors in human body odor. • Nanosize and imprinting effect makes the MIP-beads highly sensitive to vapors at the ppm level. • MIP-nanobead-coated QCM sensor array developed for the vapor recognition. • A 96% classification rate achieved by applying Leave-One-Out Cross-Validation on the model of linear discrimination analysis. • MIP-nanobeads show promise as artificial receptors for vapor and odor sensing. - Abstract: The detection and discrimination of volatile carboxylic acid components, which are the main contributors to human body odor, have a wide range of potential applications. Here, a quartz crystal microbalance (QCM) sensor array based on molecularly imprinted polymer (MIP) nanobeads is developed for highly sensitive and selective sensing of typical carboxylic acid vapors, namely: propionic acid (PA), hexanoic acid (HA) and octanoic acid (OA). The MIP nanobeads were prepared by precipitation polymerization with methacrylic acid (MAA) as a functional monomer, trimethylolproane trimethacrylate (TRIM) as a crosslinker, and carboxylic acids (PA, HA and OA) as the template molecules. The precipitation polymerization resulted in nano-sized (150–200 nm) polymer beads with a regular shape. The polymerization conditions were optimized to give a functional monomer, crosslinker, and template ratio of 1:1:2. We investigated the imprinting effect using both QCM and GC/MS measurements comparing vapor absorption characteristics between the imprinted and non-imprinted (NIP) nanobeads. A four-channel QCM sensory array based on the NIP and the three types of MIP nanobeads was fabricated for sensing the three types of carboxylic acid vapor at concentrations on the ppm level. The output of the sensor array was analyzed by both a non-supervised method (principle component analysis: PCA) and supervised method (linear discrimination analysis: LDA). LDA showed a better discrimination ability than PCA. A 96%-classification rate was achieved by applying leave-one-out cross-validation to the LDA model. The high sensitivity and selectivity of the sensor array was attributed to the imprinting effect of the nano-sized polymer beads. The developed MIP nanobeads, together with other types of MIPs, show promise as materials for artificial receptors in vapor and odorant sensing.