A Portable Gas Chromatograph with Tunable Separation and Microsensor Array Detection: Design, Characterization, and Environmental Health Applications.

Abstract

The development of a prototype portable gas chromatograph (GC) with several novel design and operating features is described. Building on a previous design, this prototype incorporates modifications that enhance the capability for determining the components of complex mixtures of volatile organic compounds (VOCs). The instrument employs a miniature multi-adsorbent preconcentrator/injector (PCI), two series-coupled columns with fast, independent temperature-programming capabilities and junction-point pressure/flow control, and a thermostatted detector consisting of an array of microfabricated chemiresistor sensors coated with Au-thiolate monolayer-protected nanoparticles whose responses patterns can be used together with retention times to identify and quantify eluting VOCs. Scrubbed ambient air is employed as the carrier gas. The instrument was characterized, with a focus on the tradeoffs in performance associated with thermal and fluidic operating variables. The influences of flow rate and operating temperature on the responses from a microsensor array used as a GC detector are described for the first time. The determination of a 31-component mixture is achieved in a total analytical cycle time of 16 min, with projected limits of detection (LODs) in the parts-per-trillion range for many vapors, assuming a 1-L sample volume. Application of the instrument to the determination of vapor-phase markers of environmental tobacco smoke (ETS) and to breath biomarkers of lung cancer is illustrated. For the former application, an adsorbent pre-trap was developed to remove semi-volatile organic compounds from the sample stream. The two markers were successfully separated from the 34 most prominent co-contaminants found in smoking-permitted environments and detected at relevant concentrations. By combining the capabilities for retention-time tuning and chemometric vapor recognition the overall analytical cycle time was reduced by 16%. For the latter application, attempts were made to establish conditions necessary to analyze breath samples spiked with seven biomarkers and 30 endogenous and exogenous interferences. Approaches to removing water vapor co-adsorbed onto the PCI during sampling were explored, and the sample volumes and separation conditions required for practical application were determined. These investigations demonstrate the potential for this novel technology to solve problems in environmental health that demand on-site analysis of complex VOC mixtures.