PM(2.5) exposure assessment using FTIR spectroscopy

Fourier transform infrared (FTIR) spectra of outdoor, indoor, and personal fine particulate matter (PM2.5) samples were collected during the Relationship of Indoor, Outdoor, and Personal Air (RIOPA) study. Results were used to chemically-characterize PM2.5, with particular attention to the poorly characterized organic fraction. Quantitative information relevant to PM2.5 exposure assessment was obtained from the FTIR spectra by applying the calibration (regression) algorithm called partial least squares (PLS). PLS was used to develop predictive relationships between raw FTIR spectral signal and other properties of PM2.5 samples. In addition, the use of PLS to estimate the outdoor contribution to indoor PM2.5 concentrations was explored. Using PLS, it was calculated that at most 85% of indoor PM2.5, 49% of personal PM2.5, and 75% of indoor OC were of outdoor origin.; PLS models developed from functional group loadings and spectra of laboratory-generated calibration standards successfully quantified sulfate, aliphatic carbon-hydrogen functionalities (CH), and carbonyl (as glutaric acid) functional groups in concurrently collected outdoor, indoor, and personal PM2.5 samples. Nitrate and ammonium functional groups were under-predicted, presumably due to losses in sampling and/or analysis. A larger portion of particulate organic matter was quantified at the functional group level than is typically identified by GC-MS analysis. An advantage of the FTIR-PLS method over both GC-MS and PM2.5 functional group quantitation methods traditionally used in aerosol science is that functional group quantities can be calculated from hundreds of samples in a matter of a few minutes.; PM2.5 species mass balances from RIOPA and other recent studies provide growing evidence that indoor-generated PM2.5 is largely predominantly. Concentrations of aliphatic CH in many indoor and personal PM2.5 samples were substantially greater than in outdoor samples, decreasing the polarity of organic PM2.5 in those samples and suggesting the presence of indoor and personal activity sources of compounds that contain this functional group, such as alkanes and phthalate esters. Examination of quantified organic functional groups suggests that indoor and personal organic PM2.5 are considerably more non-polar than outdoor organic PM 2.5, which is a new substantial finding of this dissertation and is important to understanding the behavior and fate of PM2.5.