Development and Characterization of a Novel, Low-Cost Method for Measurement of Volatile Organic Compounds

Abstract

Measurements of atmospheric pollutants are crucial for improving our understanding of atmospheric chemistry, managing air quality, and estimating exposure to compounds that have profound impacts on human health. Low cost sensors (LCS), due to order-of-magnitude reductions in power usage, maintenance needs, and purchase cost compared to research-grade reference instruments, have the potential to greatly expand the spatiotemporal resolution of these measurements. While there are several commercially-available LCS that can measure environmental volatile organic compounds (VOCs), an important class of hazardous pollutants, these sensors can only make non-specific “broadband” measurements and have, to date, been underutilized in research.

This thesis describes the development, characterization, optimization, and use of a novel low-cost instrument for measuring environmental VOCs. This instrument utilizes an array of low-cost VOC sensors representing three fundamentally different sensor types. It also takes advantage of user-controlled parameters that achieve greater degrees of differentiation between responses of sensors with the same measurement type. In the first part of this work, we describe the instrument itself, as well as a laboratory study that characterizes sensor responses to environmentally relevant VOCs. Though environmental applications pose unique challenges that can’t be completely addressed in the laboratory, our results demonstrate that this instrument can give quantitative, chemically specific information about VOCs.

The second part of this work is based on measurements made as part of a collaborative indoor air quality campaign, where our low-cost VOC instrument and co-located reference monitors made measurements of realistic indoor VOC sources. Results from an LCS-derived matrix factorization analysis were compared to an independent factor analysis of reference VOC measurements, demonstrating that our uncalibrated low-cost data can provide quantitative and qualitative information about VOC sources and composition. Based on this comparison analysis, we describe a procedure for sensor selection that allows us to evaluate the relative importance of specific sensors or sensor types in providing information about VOC composition and sources, helping future similar LCS array applications to avoid measurement redundancies and minimize material cost.

Overall, the results from this thesis show that this LCS instrument can provide useful, quantitative information about VOC sources and composition at a fraction of the size and cost of a research-grade instrument--opening the possibility of widespread and spatially distributed measurements of VOCs in air quality and chemistry contexts, especially for indoor air.