Volatile organic compounds in industrial, urban, and suburban areas: Sources and exposures.

This research was aimed at evaluating and refining sampling and analytical methods for airborne volatile organic compounds (VOCs), and at characterizing concentrations and potential exposures of VOCs found in indoor and ambient air in industrial, urban and suburban communities. A new analytical strategy of combining selective ion monitoring (SIM) and scan mode mass spectrometer analyses was developed and evaluated. This strategy improved sensitivity and selectivity without extra cost or calibration efforts. An intermittent active sampling method for collecting VOCs, which has not been previously evaluated, was compared to continuous active and passive sampling methods with the aim of obtaining long-term integrated measurements. Results obtained by the three methods agreed over a wide concentration range after accounting for the sampling rate. Intermittent sampling provides greater flexibility with respect to sampling period and flow rate, and enables the use of multi-bed adsorbents that increase the range of VOCs that can be monitored. VOC concentrations were measured inside and outside of 159 residences in suburban (Ann Arbor), urban (Ypsilanti) and urban/industrial (Dearborn) communities in southeastern Michigan from 2004 to 2005. A total of 53 and 46 VOCs were detected indoors and outdoors, respectively. Outdoors, benzene, toluene, p,m-xylene and carbon tetrachloride had the highest concentrations, and differences were seen between cities and seasons. Factor analyses identified four types of outdoor sources: vehicle exhaust/gasoline vapor, industrial solvents, biogenic emissions, and industrial sources. Indoors, benzene, toluene, p,m-xylene, n-heptane, alpha-pinene and d-limonene had the highest concentrations. Indoor to outdoor concentration ratios ranged from 1 to 10 for most compounds. Higher indoor concentrations were associated with the presence of attached garages, recent renovations, indoor smoking, residence age, infrequent window/door opening, high CO2 concentration, lower ventilation, and community. Factor analyses identified up to ten VOC source types in residences. This is one of the most comprehensive VOC surveys conducted in the US since 2000. A semi-parametric probability framework was developed to characterize VOC mixtures in residences. High concentration VOC mixtures resulted from the presence of several features (e.g. strong sources and ventilation conditions), and many characteristics of the observed mixtures could be represented using multivariate and correlated lognormal distributions.PhDAtmospheric sciencesEarth SciencesEnvironmental scienceHealth and Environmental SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/126728/2/3276194.pd