Large-scale measurement of airborne particulate sulfur☆

Abstract We describe an aerosol sampling and analysis system which represents an integral approach to large-scale monitoring of airborne paniculate matter. During our two-year participation in the St. Louis, Missouri, Regional Air Pollution Study (RAPS), 34,000 size-fractionated samples were collected by automated dichotomous samplers characterized by a particle size cutpoint of 2.4 μm. The total mass of the paniculate matter was measured by beta-particle attenuation and the elemental composition, including sulfur, was determined by photon-excited X-ray fluorescence. The long-term performance of the system will be reported Potential systematic effects related to the sampling and analysis of sulfur particles are treated here in detail. Both the accuracy and precision of sulfur measurement are estimated to be 2%. While the X-ray attenuation correction required is typically only a few per cent, a larger correction is required for a small fraction of the samples due to the migration of the sulfur into the filter. This correction is derived from the ratio of sulfur determinations made on the front and back surfaces of the membrane filter. Laboratory and field experiments have shown insignificant gaseous SO 2 conversion on the type of filters employed in the study. Preliminary data on the composition and the temporal and spatial distribution of the St Louis aerosol are presented. About 90% of the sulfur was found in the fine-particle fraction. Sulfur variations were significantly slower than those of the trace elements. Sulfates usually constitute about 35% of the total fine-particle mass, but may rise to 41% during an ‘episode’. The long-term (4 month average) sulfur data indicate that the background air masses arriving at St Louis from the west and north were about 30% lower in paniculate sulfur than those from the east and south. Also, an urban station may experience local increases in sulfur level up to a factor of two greater than the general background. Short-term (6 h average) data indicate that the effects of stationary SO 2 sources extend for long distances, (at least 40 km) and are highly directional in character.