Analyte detection in complex samples using a biomimetic, non-spectroscopic sensing method

Abstract

Optical sensing techniques for organic and biological molecules traditionally require the collection of complete, well-defined, vibrational spectra of a sample. The vibrational absorption bands of the spectra are identified, and statistical methods are then utilized to determine the moieties and concentration of target analytes present in a given sample. In this work, we present a new, non-spectroscopic alternative technique to molecular vibrational sensing. This biomimetic method, based on human color vision, uses only three broad, overlapping infrared (IR) optical filters to discriminate between chemicals with similar vibrational absorption bands. Unique detection vectors are defined by the interaction of a given chemical’s absorption bands with the three filter channels. Identification of the analytes present in a sample are then determined based on these detection vectors. We present multiple studies that demonstrate the ability of this approach to clearly discriminate between molecules with similar infrared vibrational absorption bands. We show that this method has the ability to precisely identify specific analytes in the presence of potential interferents with similar infrared absorption bands in the same sample. An optical filter based sensor that operates in the mid-IR using low power components, requiring no spectral scanning has been developed using this technique, and results using this sensor are shown. This method has the potential to lead the development of small, rugged optical sensors for non-invasive diagnostics and sensing of biological fluids.