Issues in the quantitation of functional groups by FTIR spectroscopic analysis of impactor-collected aerosol samples

Fourier Transform Infrared (FTIR) spectroscopy is used to complement total carbon and molecular level measurements in the study of organic particulate matter because it is a relatively simple analytical procedure that provides chemically useful details (i.e., functional group and bond information) missing in total carbon measurements. FTIR characterizes the entire aerosol, whereas molecular methods identify only a small fraction of the organic mass. The need to measure the size-resolved chemical composition of ambient aerosols often necessitates the collection of aerosol samples with a single-jet impactor, such as a Hering low pressure impactor (LPI). Single-jet impactor samples have a nonuniform (conical) deposit geometry that results in a variable and inconsistent path length across the sample deposit and may also result in the enhancement of infrared beam scattering. This makes quantitation of LPI samples by traditional procedures based on Beer's Law impossible. Therefore we applied a ratio method developed by Cliff and Lorimer (1972) for energy dispersive spectroscopy to FTIR analyses of LPI samples. In this study the prospects and limitations for quantitative measurement of functional groups by FTIR spectroscopic analysis of impactor-collected aerosol samples is examined using laboratory-generated standards and size-segregated samples collected with a LPI in the Smoky Mountains. FTIR spectroscopy provides semiquantitative estimates of functional group concentrations (precision 50%) in aerosol samples. However, Infrared (IR) beam scattering due to the physical geometry of impactor deposits is a significant source of uncertainty, and concentrations of some functional groups appear to be substantially underestimated due to the clipping of high wave number IR peaks by rising baselines. The rising baselines result from increases in IR beam scattering with wave number and are related to the sample deposit geometry. Future sampling or analytical methodologies must overcome the bias that results from irregular sample deposit geometry.