 |
NARSTO Workshop on Innovative Methods for Emission Inventory Development and Evaluation University of Texas, Austin October 14-17, 2003 |
 |
An Innovative Method for Source Identificationand Apportionment of PM: Isotope Measurements of CarbonSpecies (OC/EC) via Thermal Desorption/Combustion & GC-IRMS
L. Huang, W. Zhang, J. Brook, D. Ernst, A.Chivulescu, G. Lu, and S. SharmaMeteorological Service of Canada, Toronto, ON M3H 5T4 Canada
Carbonaceous aerosols play an important role in environment related issues, e.g. air quality, human health and global climate. However, the relative contribution of different emission sources, their formation mechanism, their atmospheric transformation and transport are not well understood. Because of their carbonaceous nature, 13C/12C ratios (expressed as d13C, i.e. relative deviation to the international standard in ‰) of bulk carbon fractions (i.e., organic carbon and elemental carbon) in airborne PM could provide valuable information for their sources and formation processes. It is known that aerosols originate from different emission sources, e.g. fossil fuel combustion, biomass burning and vegetation emissions etc., through different formation processes, i.e. primary combustion emission (high temperature) and secondary photochemical production (low temperature). The d13C of EC is expected to remain constant once formed during combustion because of its inert nature and thus, should reflect the source signature. In contrast, the d13C of OC (organic carbon) is expected to undergo varying amounts of change due to photochemical processes, and this should depend upon the original compound and on the history of the air mass. Therefore, the distribution of d13C among bulk carbon fractions may provide valuable information for PM formation processes and correspondent sources.
In this study, we have developed a method using a step-wise heating thermal desorption/combustion OC/EC analyzer coupled with GC-IRMS (Gas Chromatography separation, followed by a Isotopic Ratio Mass Spectrometer analysis) to measure 13C/12C ratios of bulk carbon fractions from filter samples of atmospheric particle. These fractions are released at different temperature ranges and different redox conditions. In the current method we focus on three fractions, i.e., low OC (organic carbon), high temperature OC plus carbonate carbon (CC) and EC (elemental carbon). The overall precision of the method is ~ 0.3‰ and the results compare well with the conventional off-line method, differing by an average of 0.1‰, which is the accuracy of the method.
We have applied the method to study fine (PM2.5) particles collected at several contrasting locations. These were: three surface sites in the Lower Fraser Valley (LFV) of British Columbia (Golden Ears Park, which is expected to be significantly influenced by biogenic sources; Cassiar Tunnel, which is dominated by motor vehicle emissions and Slocan Park, which is a typical urban background area and could show the results of different sources mixing), an aircraft-based platform flying well-within the free troposphere over the LFV, and Alert, NWT, a baseline station in high Arctic (summer and winter). As expected, differences in d13C and its distribution patterns of bulk carbon fractions among the sampling locations were observed. The findings of the study will be presented. It is concluded that the combination of the isotopic compositions of different carbon components and the correspondent abundances of the carbon components is likely to be a useful tool to constrain source identification and apportionment for ambient particle samples.