Predicting limits of detection in real-time sweat-based human performance monitoring

Abstract

Sweat-based human performance monitoring devices offer the possibility of real-time emotional and cognitive awareness in both civilian and military applications. Broad applicability and point of use necessitate non-invasive, printable, flexible, wearable chemical sensors with low power consumption. Sweat fluidics must enable movement of sweat across the sensor compartment within 1 minute to assure only fresh sweat is at the chemical sensor. The sensor material should have reaction kinetics to capture a sufficient number of target molecules for quantification in real-time (< 1minute). Chemical selectivity is critical in complex biofluids such as sweat, which may be comprised of 800+ biomarkers. Given these constraints, there continues to be significant technological barriers for translation from laboratory-based proof-of-concept demonstrations and scalable manufacturing of devices. Using finite element simulations, we focus on determining which sweat flow geometry and chemical capture dynamics are best suited to meet temporal performance requirements. Two common sensing approaches are compared and contrasted: bio-recognition chemical adsorption events and electrochemical detection. Responsivity of both mechanisms is shown to be highly dependent on fluid dynamics, analyte capture efficiency, analyte concentration, and reaction kinetics. Key metrics of temporal response and capture efficiency will be discussed for a number of state of the art electronic sensor materials, with a focus on the validity of printable platforms.