Validation of an assay for the determination of levoglucosan and associated monosaccharide anhydrides for the quantification of wood smoke in atmospheric aerosol.

Biomass burning is becoming an increasing contributor to atmospheric particulate matter, and concern is increasing over the detrimental health effects of inhaling such particles. Levoglucosan and related monosaccharide anhydrides (MAs) can be used as tracers of the contribution of wood burning to total particulate matter. An improved gas chromatography-mass spectrometry method to quantify atmospheric levels of MAs has been developed and, for the first-time, fully validated. The method uses an optimised, low-volume methanol extraction, derivitisation by trimethylsilylation and analysis with high-throughput gas chromatography-mass spectrometry (GC-MS). Recovery of approximately 90 % for levoglucosan, and 70 % for the isomers galactosan and mannosan, was achieved using spiked blank filters estimates. The method was extensively validated to ensure that the precision of the method over five experimental replicates on five repeat experimental occasions was within 15 % for low, mid and high concentrations and accuracy between 85 and 115 %. The lower limit of quantification (LLOQ) was 0.21 and 1.05 ng m(-3) for levoglucosan and galactosan/mannosan, respectively, where the assay satisfied precisions of ≤20 % and accuracies 80-120 %. The limit of detection (LOD) for all analytes was 0.105 ng m(-3). The stability of the MAs, once deposited on aerosol filters, was high over the short term (4 weeks) at room temperature and over longer periods (3 months) when stored at -20 °C. The method was applied to determine atmospheric levels of MAs at an urban background site in Leicester (UK) for a month. Mean concentrations of levoglucosan over the month of May were 21.4 ± 18.3 ng m(-3), 7.5 ± 6.1 ng m(-3) mannosan and 1.8 ± 1.3 ng m(-3) galactosan.