Laboratory investigations of ultrafine particle growth due to the oxidation of atmospheric volatile organic compounds

Abstract

Atmospheric aerosols, or particles suspended in the Earth’s atmosphere, affect both human health and climate around the world. Ultrafine particles (<100 nm in diameter) can have direct negative impacts on human health, mainly through inhalation. These particles have limited influence on climate, however, until the diameter exceeds about 50-100 nm, where they become able to serve as cloud condensation nuclei (CCN) and participate in cloud formation. Clouds significantly reduce the intensity of incoming solar radiation by reflecting this energy back into space, resulting in a significant cooling effect. Growth of smaller ultrafine particles into the CCN active size range primarily occurs through the condensation and/or partitioning of gas-phase organics onto the particles, but the chemical processes and kinetics associated with particle growth are complex and not well understood. ☐ This dissertation focuses on simulating atmospheric gas- and particle-phase chemistry in a flow tube reactor to gain a better understanding of what conditions and parameters affect how quickly a particle grows. First, a kinetic model was developed and applied to atmospherically relevant conditions typically used in the flow tube reactor experiments to characterize the dynamic particle growth within the reactor. These simulations examine the uptake of organic molecules produced by oxidation of a precursor compound and how uptake depends on the ability of particle size, composition, and phase to promote particle-phase reactions. Particle growth is represented by a parameter called the growth factor (GF), which is defined as the fraction of precursor oxidation products able to grow the particle. For a variety of simulated conditions, it is shown that the GF is dynamic and can change over time in the flow tube, but particle growth can be accurately described by a single GF value obtained from the measured outlet minus inlet particle diameter change. ☐ Next, a series of flow tube experiments were performed to investigate how variations in particle size, composition, and phase can affect the GFs. Here, seed particles consisting of ammonium sulfate or ammonium bisulfate with varying water contents are mixed with a gas-phase organic precursor (isoprene or α-pinene) and ozone in the flow tube, and GFs are determined from the outlet minus inlet diameter change. For particles grown by isoprene ozonolysis, ammonium sulfate exhibited higher GFs than ammonium bisulfate, and for both compositions, the GFs increased with increasing water content, either when the amount of adsorbed water on effloresced particles or the relative humidity increased, or by deliquescing the particles. GFs were relatively unchanged with increasing particle diameter, suggesting that the growth kinetics were surface-limited rather than volume-limited. These same trends were observed for particles grown by α-pinene ozonolysis, though the GFs were uniformly larger for α-pinene than isoprene, which can be linked to the greater size of α-pinene, which produces larger and less volatile oxidation products. In all cases, the GFs measured in this work were equal to or larger than the yield of highly oxidized molecules reported for ozonolysis of the respective organic precursor, suggesting that semi-volatile products of ozonolysis are also taken into the particles, most likely by the formation of oligomers. ☐ Finally, offline molecular analysis of aerosol exiting the flow tube reactor was performed under conditions examined for growth experiments. Due to the significant amount of gas-phase oxidation products in the air flow exiting the reactor with respect to particulate matter, the use of charcoal denuders was required to reduce the overwhelming signal from gas-phase species. While the gas-phase signal was still very large with the denuders, it did become possible to identify a few products unique to the particle phase, and these generally contained nitrogen and/or sulfur atoms in addition to carbon, hydrogen, and oxygen. More importantly, the molecular species observed from both the gas and particle phases were oligomeric, likely formed through the reaction of peroxy radicals and/or hydroperoxides directly in the gas phase, in the particle phase, or on the surface of the collection filter fibers. The results suggest that the large GFs measured in this work do indeed arise from the oligomerization of semi-volatile ozonolysis products, and the formation of these oligomers is enhanced by increasing water content at or near the particle surface. On the other hand, very acidic particles can inhibit the formation of oligomers and/or cause oligomers to decompose, thereby decreasing the GF. ☐ The results of this dissertation suggest that models of CCN formation from new particle formation should incorporate particle composition into the calculations of ultrafine particle growth in the atmosphere. With simple changes in composition and/or water content, significant changes in particle growth are observed, highlighting that a singular growth parameter should not be utilized for particle growth in CCN models. By improving the complexity of climate models, the accuracy of their predictions inherently improves, allowing for more informed decisions on how humanity addresses problems associated with the ever-changing environment.