Clues for a standardised thermal-optical protocol for the assessment of organic and elemental carbon within ambient air particulate matter

Along with some research networking programmes, the European Directive 2008/50/CE requires chemical speciation of fine aerosol (PM2.5), including elemental (EC) and organic carbon (OC), at a few rural sites in European countries. Meanwhile, the thermal-optical technique is considered by the European and US networking agencies and normalisation bodies as a reference method to quantify EC–OC collected on filters. Although commonly used for many years, this technique still suffers from a lack of information on the comparability of the different analytical protocols (temperature protocols, type of optical correction) currently applied in the laboratories. To better evaluate the EC–OC data set quality and related uncertainties, the French National Reference Laboratory for Ambient Air Quality Monitoring (LCSQA) organised an EC–OC comparison exercise for French laboratories using different thermal-optical methods (five laboratories only). While there is good agreement on total carbon (TC) measurements among all participants, some differences can be observed on the EC / TC ratio, even among laboratories using the same thermal protocol. These results led to further tests on the influence of the optical correction: results obtained from different European laboratories confirmed that there were higher differences between OCTOT and OCTOR measured with NIOSH 5040 in comparison to EUSAAR-2. Also, striking differences between EC TOT / EC TOR ratios can be observed when comparing results obtained for rural and urban samples, with ECTOT being 50% lower than ECTOR at rural sites whereas it is only 20% lower at urban sites. The PM chemical composition could explain these differences but the way it influences the EC–OC measurement is not clear and needs further investigation. Meanwhile, some additional tests seem to indicate an influence of oven soiling on the EC–OC measurement data quality. This highlights the necessity to follow the laser signal decrease with time and its impact on measurements. Nevertheless, this should be confirmed by further experiments, involving more samples and various instruments, to enable statistical processing. All these results provide insights to determine the quality of EC–OC analytical methods and may contribute to the work toward establishing method standardisation