A recent comparative exercise found that different black carbon (BC) quantification methods produced widely varying results for a set of BC reference materials that included three laboratory-produced BC-rich materials, five environmental BC matrices and four samples of non-BC organic matter. This variation was attributed to a range of physical and chemical properties of the samples that could, in various ways, result in the over- or under-detection
of BC in some or all of the techniques. Here the most pertinent chemical and physical characteristics of the samples are presented, including elemental analysis, lightness measurements, BET surface area measurements and 13C NMR analysis. Amongst the BC-rich materials, soot and char could be distinguished from one another mainly on the basis of H/C and O/C ratios, NMR observability and BET surface area. The results indicate that the aromatic structures in the soot are more highly condensed, and this explains why some BC quantification techniques detect these two materials differently. The non-BC potentially
interfering materials were shown to share properties with the BC-rich materials (high C content, low lightness values and high aromaticity) that are used for certain BC quantification methods. This may lead to overestimation of BC unless these interfering materials are removed during pre-treatment. The environmental matrices were found to have relatively high amounts of metal oxides that have the potential to catalyse or inhibit thermal and chemical reactions during BC analysis.