A method for the quantitative analysis of organic compounds on submicron particulate matter (PM1) collected on quartz filters was developed. The compounds analyzed encompassed C22-C35 alkanes, polycyclic aromatic hydrocarbons (PAHs), quinones, levoglucosan, cis-pinonic acid and short chain dicarboxylic acids such as malonic, succinic, glutaric, adipic, suberic, azelaic, malic and phthalic acids. The method included extraction with a pressure liquid extraction system, sample filtration though glass fibre filter, fractionation by high performance liquid chromatography and subsequent analysis by gas chromatography coupled to mass spectrometry. The study of the extraction efficiency of different solvent mixtures showed that DCM:MeOH 1:1 was the one providing the highest recoveries for all compounds. Extraction temperatures of 100°C provided better results than 60°C or 80°C. This method provided comparable extraction efficiency and qualitative and quantitative data to those involving Soxhlet extraction. Method recoveries for alkanes, most PAH, quinones and polar compounds calculated from spiked real samples were 52-72%, 78-101%, 50-62% and 76-104%, respectively, reproducibilities were 2-28%, 7-29%, 10-27% and 5-28%, respectively, limits of quantification were 0.01-0.1ng/m(3), 0.01-0.27ng/m(3), 0.04ng/m(3) and 0.32-2.8ng/m(3), respectively, which affords the quantification of a broad number of primary and secondary organic constituents of submicron aerosols.

No abstract

No abstract

20The combined use of chemical analysis of organic molecules in atmospheric aerosols (PM1) 21 collected in situ in Barcelona and optical measurements with a light detection and ranging 22(LIDAR) instrument allowed the characterization of the smoke plume from a wildfire that 28(PAH), which was coincident with an increase of the aerosol density in the boundary layer (BL). 29According to air-mass trajectory modelling, the origin of the biomass burning particles was 30 ©2016. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ DOI:10.1016/j.apr.2015.12.003 2 related to the severe wildfire occurring 120 km northeast of the city. On the following days, no 1 significant contribution of wildfire smoke was found in the urban air (i.e. in the BL), although the 2 lidar detected particles aloft. The FLEXPART model was able to simulate the onset of the 3 wildfire smoke plume event in the urban center in the early morning of 23 July, but not the fast 4 transition of the plume at the surface, indicating that small-scale meteorological phenomena, 5 such as an intensive local sea-breeze was not fully characterized by this model, resulting to 6 simulated behavior of particles at the surface which was more representative of the conditions 7 aloft.

No abstract