No abstract

Classical methodology based on the application of "lters for sampling, followed by extraction and analysis, introduces severe artifacts for semi-volatile compounds like ammonium nitrate. These "lter methods do not meet the requirements for the assessment of the impact of aerosols on acidi"cation, air quality and especially on the radiative balance, in terms of required speed, detection limits and selectivity. These artifacts are avoided by using a steam jet aerosol collector sampler, based on scavenging of aerosols by droplet formation, in combination with on-line analytical techniques such as ion-chromatography for nitrate and membrane separation followed by conductivity detection for ammonium. The SJAC sampler combines very low blanks with high e$ciency of collection of particles. The ammonium detector and the IC system, based on 1-point internal standard calibration in combination with correction for curved calibration graphs, enables detection of ammonium and nitrate at background conditions, the detection limit is about 0.02 g m\ of ammonium and nitrate. Accuracy is, depending on ambient concentration, in the order of 5}10% relative, at a range of 0.05}50 g m\. The time resolution is 15}120 min, depending on required detection limit, and is short enough for continuously monitoring the chemical composition of aerosols. Quality assurance and quality control experiments and intercomparison experiments with classical "lter methods, thermo-denuder systems, denuder di!erence methods and other continuous monitoring techniques have shown that the results are reliable. The instrument has successfully been employed in "eld campaigns in Europe and the US.

[1] A comprehensive range of aerosol parameters was measured at an urban and a rural site in the Milan, Italy metropolitan region during summer smog events in summer 1998. Measurements were performed as part of the Pianura Padana Produzione di Ozono (PIPAPO) field campaign to determine the sensitivity of O 3 production to NO X and volatile organic carbon concentrations at several ground stations. Primary aerosol parameters (i.e., direct emissions) such as aerosol black carbon showed a distinct diurnal variation with maxima at about 0000 and 0800 central European summer time (CEST), in contrast to secondary aerosol parameters such as sulfate and nitrate. Aerosol number size distributions were measured under ambient conditions as well as after conditioning with volatility and hygroscopicity systems. A mode at d = 20-30 nm in the number concentration was found at 0800 CEST and exhibited high volatility at 110°C ($80% volume lost upon heating) but no hygroscopic behavior. Based on these measurements, small particles (d < 40 nm) are thought to consist mainly of hydrophobic particulate organic matter, rather than soot or H 2 SO 4 aerosols. Two distinct hygroscopic modes with average growth factors d/d 0 $ 1.02 and 1.21-1.28 were found for particles with dry (relative humidity of <30%) diameters d 0 = 50-200 nm. Submicrometer aerosols exhibited lower volatility at the rural than at the urban site, which is attributed to additional particulate mass produced by secondary particle formation.

No abstract

No abstract