Application of the integrating sphere method to separate the contributions of brown and black carbon in atmospheric aerosols.

Until about a decade ago, black carbon (BC) was thought to be the only light absorbing substance in the atmospheric aerosol except for soil or desert dust In more recent years, light absorbing polymeric carbonaceous material was found in atmospheric aerosols. Absorption increases appreciably toward short wavelengths, so this fraction was called brown carbon. Because brown carbon is thermally rather refractory, it influences the split between organic carbon (OC) and elemental carbon (EC) in thermal methods and, through its light absorption characteristics, leads to overestimations of BC concentrations. The goal of the present study was to extend the integrating sphere method to correct the BC signal for the contribution of brown carbon and to obtain an estimate of brown carbon concentrations. Humic acid sodium salt was used as proxy for brown carbon. The extended method is first tested on mixtures of test substances and then applied to atmospheric samples collected during biomass smoke episodes (Easter bonfires) in Austria. The resulting concentrations of black and brown carbon are compared to EC obtained with a widely used thermal method, the Cachier method (Cachier et al. Tellus 1989, 41B, 379-390) and a thermal-optical method (Schmid et al. Atmos. Environ. 2001, 35, 2111-2121), as well as to concentrations of humic like substances (HULIS) and to biomass smoke POM (particulate organic matter). Both the thermal methods were found to overestimate BC on days with large contributions of woodsmoke, which agrees with the findings of the method intercomparison study by Reisinger et at. (Environ. Sci. Technol. 2008, 42, 884-889). During the days of the bonfires, the Cachier method gave EC concentrations that were higher by a factor of 3.8 than the BC concentrations, while the concentrations obtained with the thermal-optical method were higher by a factor of 2.6.