Controlled generation of black carbon particles from a diffusion flame and applications in evaluating black carbon measurement methods

Measurements of black carbon (BC) concentration have long been uncertain. Neither a BC standard or a technique that provides a reliable measurement of BC exist, which precludes the evaluation and optimization of BC measurement methods. In this manuscript, we describe the generation of BC particles (mass absorption efficiency similar to 8.5m(2) g(-1) and single scattering albedo similar to 0.2 at 530 nm) using an inverted diffusion flame. This flame is remarkably stable and can be used to generate a wide range of nearly constant concentrations of BC particles for many (e.g., 10 +) hours. The particles contain essentially no organic carbon (OC), so the quantification of BC mass using any of the thermal or thermal-optical analysis JOA) methods is straightforward instead of highly uncertain. In this case, the TOA measurement of BC can be used to evaluate the accuracy of other BC measurement methods. In this paper, we demonstrate the application of the diffusion flame in evaluating BC measurements made with filter-based light transmission methods, and in particular the aethalometer. We show that the amount of light attenuated by a BC-laden filter increases in less than constant proportion to the BC mass loading, and thus the effective BC attenuation coefficient decreases with increased BC mass loading. When sampling BC at constant concentration, the aethalometer erroneously reports decreasing concentrations of BC as its filter becomes increasingly loaded because it treats the attenuation coefficient as a constant. A simple method to correct erroneous aethalometer data when sampling aerosols with low single scattering albedo is presented. Another application of the diffusion flame considered in this paper is the development of BC standards for TOA. We envision the preparation of filter samples with known amounts of BC mixed with other aerosol constituents, which may help in understanding the uncertainty in and optimizing TOA measurements, and may be of use in future TOA method comparison studies. Toward this end, we demonstrate that the diffusion flame can be used to replicate filter samples with known amounts of BC. Additionally, we show that the combustion temperature of BC during TOA depends on sample composition, which suggests that the tempera ture-defined carbon fractions of some thermal and TOA methods may be of limited value. Published by Elsevier Ltd.