Date of Award

Fall 9-2016

Embargo Period

11-7-2016

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Chemical Engineering

Advisor(s)

Neil Donahue

Abstract

Volatile organic compounds released by plants or through processes such as combustion reacts with oxidative species in the air, such as ozone or hydroxyl radicals. Smog chamber studies are conducted to determine the chemistry that these organic precursors undergo, and the products that are formed. These products span orders of magnitude in volatility, making the tracking of each individual species a difficult process. The volatility basis set (VBS), which separates species based on volatility, has been shown to be an effective framework to track the formation of aerosols from these products. In the first part of this work, a 2-dimensional VBS model is used to investigate the introduction of NOx to -pinene aging. A new dimension is added to the VBS to track the formation and aging of organonitrates. The results show that higher volatility precursors produce less aerosol mass, while lower volatility precursors produce more, which is consistent with prior experiments of NOx effects. In addition, the model shows that the detection of small concentrations of nitrate ions can still indicate presence of substantial organonitrate mass. The formation of aerosols from -pinene ozonolysis experiments at CLOUD are modeled in the next part of this work. CLOUD experiments show the production of low volatility organic compounds, or (E)LVOCs, from -pinene ozonolysis contribute to the growth of nucleated particles. The inclusion of a Kelvin effect is necessary to reproduce particle growth rates at small diameters (< 4 nm). Flux balance calculations from the dynamic VBS model show that the raw distribution of products seen by the nitrate-CIMS cannot fully explain the particle growth, and indicate that product masses must be higher. When the model accounts for the charging efficiency of LVOCs in the nitrate-CIMS, it is capable of reproducing the growth of particles from these experiments. Lastly, the yields of E(LVOC)s required to reproduce the data in the previous chapter appear to contradict yields from prior -pinene experiments. We explore potential explanations for this disagreement. By treating the chamber model as a dynamical process, the model demonstrates that high yields will appear lower due to the delay between (E)LVOC formation and condensation. While the results still show an overprediction by the current model, it indicates that a dynamical treatment is indeed necessary to capture the condensation of vapors to particles.

Share

COinS