Infection Diagnostics Enabled by Selective Adsorption of Breath-Based Biomarkers in Zr-Based Metal-Organic Frameworks

Exhaled breath contains trace levels of volatile organic compounds (VOCs) that can reveal information about metabolic processes or pathogens in the body. These molecules can be used for medical diagnosis, but capturing and accurately measuring them is a significant challenge in chemical separations. A highly selective nanoporous sorbent can be used to capture target molecules from a breath sample and preconcentrate them for use in a detector. In this work, we present a combined predictive modeling-experimental validation study in which five Zr-based metal-organic frameworks (MOFs) were identified and tested. These MOFs display good selectivity for a variety of VOCs known to be indicators of viral infections such as influenza and COVID-19. We first used molecular simulation to identify promising MOF candidates that were subsequently synthesized and tested for recovery of a variety of VOCs (toluene, propanal, butanone, octane, acetaldehyde) at concentrations of 20 ppm in humid nitrogen. We show that MOF-818, PCN-777, and UiO-66 have particularly good selectivity for the target molecules in the presence of humidity. These three MOFs each recover around 40-60% of the targets (with the exception of acetaldehyde) at up to 95% relative humidity. MOF-818 recovers 63% of butanone and 60% of toluene at 80% relative humidity. Recovery for acetaldehyde is lower across all MOFs at high humidity, but notably, MOF-808 recovers 90% of acetaldehyde at 60% humidity.