A New Method for Accurate Signal Processing in Measurements of Elemental Mercury Vapor by Atomic Fluorescence Spectrophotometry

Thursday, 17 December 2015
Poster Hall (Moscone South)
Jesse L Ambrose II, University of Washington, Bothell, WA, United States and Daniel A Jaffe, University of Washington Bothell Campus, Bothell, WA, United States
The most widely used method for quantifying atmospheric Hg is gold amalgamation pre-concentration, followed by thermal desorption (TD) and detection via atomic fluorescence spectrophotometry (AFS). Most AFS-based atmospheric Hg measurements are carried out using commercial analyzers manufactured by Tekran® Instruments Corp. (instrument models 2537A and 2537B). A generally overlooked and poorly characterized source of analytical uncertainty in these measurements is the method by which the raw Hg AFS signal is processed. In nearly all applications of Tekran® analyzers for atmospheric Hg measurements, researchers rely upon embedded software which automatically integrates the Hg TD peaks. However, Swartzendruber et al. (2009; doi:10.1016/j.atmosenv.2009.02.063) demonstrated that the Hg TD peaks can be more accurately defined, and overall measurement precision increased, by post-processing the raw Hg AFS signal; improvements in measurement accuracy and precision were shown to be more significant at lower sample loadings. Despite these findings, a standardized method for signal post-processing has not been presented. To better characterize uncertainty associated with Tekran® based atmospheric Hg measurements, and to facilitate more widespread adoption of an accurate, standardized signal processing method, we developed a new, distributable Virtual Instrument (VI) which performs semi-automated post-processing of the raw Hg AFS signal from the Tekran® analyzers. Here we describe the key features of the VI and compare its performance to that of the Tekran® signal processing method.