Heterogeneous Ice Nucleation from Laboratory-Generated and Field-Collected Aerosol Particles

Thumbnail Image
Issue Date
Wang, Bingbing
The Graduate School, Stony Brook University: Stony Brook, NY.
Aerosol particles can serve as atmospheric ice nuclei (IN) and thus initiate ice formation resulting in the formation of cirrus and mixed&ndashphase clouds and thereby indirectly affect the global radiative budget, hydrological cycle, and climate. Atmospheric ice crystal formation by heterogeneous nucleation is poorly understood and poses one of the largest uncertainties in predicting future climate. Here I present a laboratory study on heterogeneous ice nucleation efficiency and water uptake by various types of laboratory&ndashgenerated and field&ndashcollected organic&ndashcontaining particles for temperatures and relative humidity typical of the troposphere and lower stratosphere. Laboratory&ndashgenerated Suwannee river standard fulvic acid (SRFA) and Leonardite standard humic acid (Leonardite) particles served as surrogates of HUmic LIke Substances typically found in atmospheric particles. These organic particles nucleate ice efficiently via deposition mode (ice crystals form directly from the supersaturated water vapor) and immersion freezing (ice crystals form from the ice nucleus suspended in supercooled aqueous droplets) at relevant atmospheric conditions. Oxidation of Leonardite and SRFA particles by O<sub>3</sub> led to a decrease in deposition nucleation efficiency and to water uptake at lower temperatures for the former and to an increase in the lowest temperature at which deposition nucleation was observed for the latter. Thus, particle hygroscopicity may not be the only factor determining particle's ice nucleation efficiency. Laboratory&ndashgenerated amorphous secondary organic aerosol (SOA) particles from anthropogenic precursor gases such as naphthalene serving as surrogates of ubiquitous SOA in the atmosphere demonstrated the potential to act as deposition IN at temperatures below 230 K and at relative humidity (RH) with respect to ice (RH<sub>ice</sub>) below the homogeneous freezing limit. Water uptake was observed above 230 K followed by immersion freezing at temperatures between 230 and 242 K. The bulk atomic oxygen&ndashto&ndashcarbon (O/C) ratio of these SOA particles did not show a significant effect on deposition ice nucleation but on water uptake. Above 230 K particles with higher O/C ratio take up water at lower (RH than particles with low O/C ratio. These SOA particles may form a glassy, i.e. solid state, and nucleate ice via deposition mode below 230 K whereas they adopt a semi&ndashsolid state with lower viscosity at higher temperatures and take up water. Thus, phase state and viscosity affect the interaction of SOA with water vapor. Different types of anthropogenic and marine impacted particles collected within and around the urban environments of Los Angeles and Mexico City were investigated for their potential to nucleate ice. To relate the particle's ice nucleation efficiency with chemical composition, micro&ndashspectroscopic single particle analyses were applied by using computer controlled scanning electron microscopy with energy dispersive analysis of X&ndashrays (CCSEM/EDX) and scanning transmission X&ndashray microscopy with near edge X&ndashray absorption fine structure spectroscopy (STXM/NEXAFS). The chemical composition was found to play a crucial role in determining the water uptake and immersion freezing but less for deposition nucleation. Overall, these field&ndashcollected particles can serve as efficient IN at atmospheric conditions typical for cirrus and mixed phase cloud formation and exhibit distinctly different ice nucleation efficiencies compared to the laboratory generated organic proxies. The experimentally derived ice nucleation data were analyzed using classical nucleation theory (CNT) and the singular hypothesis approach (SH) providing heterogeneous ice nucleation rate coefficients (J<sub>het</sub> ) and cumulative IN spectrum (K), respectively, allowing either approach to be implemented in cloud resolving models. The experimentally derived J<sub>het</sub> and corresponding contact angles were determined as a function of temperature and RH<sub>ice</sub>and used to parameterize heterogeneous ice nucleation. Atmospheric implications of these findings and the proposed parameterizations are discussed.
185 pg.