Inter-comparison of carbon content in PM10 and PM2.5 measured with two thermo-optical protocols on samples collected in a Mediterranean site.

Scientific interest is focusing on different approaches for characterising organic carbon (OC), elemental carbon (EC) and equivalent black carbon (eBC), although EUSAAR2 protocol has been established and frequently used in EU for regulatory purposes. Discrepancies are observed due to thermal protocols used for OC/EC determinations and the effect of the chemical-physical properties of aerosol using optical measurements for eBC. In this work, a long-term inter-comparison of carbon measurements with two widely used protocols (EUSAAR2 and NIOSH870) was performed on PM2.5 and PM10 samples. The influence of the protocol on the evaluation of secondary organic aerosol (SOC) and on the correlation between EC and eBC was investigated. An extensive check of repeatability gave typical uncertainties of ~ 5% for TC and OC, and ~ 10% for EC for both thermal protocols. Results show that OC is statistically comparable between the two protocols but EC is significantly higher with EUSAAR2, especially during the warm season. The ratio OC/EC is lower with EUSAAR2, also showing a seasonality (lower values in the warm season) not observed with NIOSH870. Despite the differences in OC/EC ratios, the contribution of SOC to OC (~ 50%), evaluated using the EC-tracer method, did not differ significantly between the two protocols and for both size fractions. Further, SOC/OC ratios were comparable in cold and warm periods. eBC/EC ratios larger than one for both protocols were obtained, 1.62 (EUSAAR2) and 1.92 (NIOSH870), and also correlated with the ratio OC/EC for both protocols, especially in the cold season.

Traffic tracers in a suburban location in northern Spain: relationship between carbonaceous fraction and metals.

PM10 and black smoke were monitored at a suburban sampling station located in the northern Spanish city of Gijón. Thirty-two metals and total carbon (TC) (i.e., organic carbon (OC) and elemental carbon (EC)) were analyzed over a year. The study of air-mass origin based on 5-day back trajectories was carried out to assess its influence on the recovery data. Different strategies were implemented to infer the influence of traffic in the area. On average, TC accounted for 29 % of the PM10 fraction, with OC forming 77 % of this TC. The influence of traffic was clearly reduced during intense Atlantic advection episodes, when OC and EC decreased up to 0.39 and 0.22 µg C/m(3), respectively. In contrast, the highest values were reported during regional episodes, exceeding 10 µg C/m(3) of OC and 2 µg C/m(3) of EC. The correlation between EC and OC was found to notably improve when considering the days with high traffic flow (from R (2) = 0.46 to R (2) = 0.74). This pattern was also reproduced by black smoke and EC (from R (2) = 0.49 to R (2) = 0.59). Cu and Sn were found to be reliable traffic tracers given their high dependence on EC (R (2) = 0.82 and R (2) = 0.79, respectively). Nevertheless, Sn, Ba, and Sb showed a better correlation with Cu than EC, suggesting a common origin. In the case of Sn, R (2) improved from 0.79 to 0.91. The Cu/Sb ratio had a mean value of 6.6 which agrees with diagnostic criterions for brake wear particles. The relationships and ratios between EC, Cu, Sb, Sn, Ba, and Bi pointed out to non-exhaust emissions, playing a significant role in the chemical composition of PM10. Brake wear was presented as the most likely origin for Cu, Sb, and Sn.