Use of a Continuous Stirred Tank Reactor for the Study of Aqueous Aerosol Chemistry
<p>Atmospheric aerosol chemistry is important in areas ranging from urban air pollution to cloud formation. It has long been supposed that droplet-phase reactions account for a significant fraction of the atmospheric conversion of SO₂ to sulfate. Among such reactions is the manganese-catalyzed...
| Summary: | <p>Atmospheric aerosol chemistry is important in areas ranging from urban air pollution to cloud formation. It has long been supposed that droplet-phase reactions account for a significant fraction of the atmospheric conversion of SO₂ to sulfate. Among such reactions is the manganese-catalyzed aqueous-phase oxidation of SO₂. Whereas the role of aqueous phase SO₂ oxidation in the dilute solutions characteristic of fog and cloud droplets (diameter > 10 µm) has been reasonably well established, the role of comparable reaction in submicron aerosols is uncertain. In this thesis a reactor system is developed to carry out gas-aerosol reactions under humid, ambient-like conditions. The apparatus consists of a continuous stirred tank reactor (CSTR) in which the growth of the aqueous aerosol is measured. Absence of mass transfer limitation, coagulation, and nucleation ensure that particle growth is direct evidence of reaction. Special care is taken to minimize size biasing of the aqueous aerosol in the electrostatic classifier used to measure the reactor feed and effluent distributions. Aerosol behavior in the reactor is modeled assuming an ideal CSTR and, given the solution thermodynamics and equilibrium chemistry, the effluent distribution can be predicted using one of the proposed reaction rate mechanisms.</p>
<p>Experiments were performed using a pure MnSO₄ or a MnSO₄-Na₂SO₄ mixture feed aerosol. The relative humidity ranged from 86 to 94% and 0.1 ppm < p<sub>SO₂</sub>, < 50 ppm. The slow, approximately constant reaction rate of Bronikowski and Pasiuk-Bronikowska (1981) (R ~ 2 x 10⁻⁴ Ms⁻¹) was found to best predict the observed growth over the entire range of operating conditions. The various rate expressions proposed for this system in the literature resulted in varying estimates of growth. When reactor conditions were similar to those at which the rate expression was determined, the agreement between the predicted and observed distributions improved. This indicates that use of a rate expression beyond its specified range may result in erroneous predictions.</p>
|
|---|