MULTI-MODAL CHEMICAL CHARACTERIZATION OF ATMOSPHERIC PARTICLES

Atmospheric aerosols are solid and liquid particles emitted from a range of natural and anthropogenic sources, and that impact Earth’s climate directly by interacting with solar radiation, as well as indirectly through modifications to the properties and lifecycles of clouds. Furthermore, atmospheric particles yield substantial implications on air quality, visibility, and human health. While the impact of aerosols on the planet has been broadly defined, accurate forecasting of atmospheric particle processes remains challenging due to their complex physicochemical properties. Highly variable aerosol characteristics include size, morphology, viscosity, elemental and molecular composition, hygroscopicity, mixing state, and light-absorption. Moreover, aerosols experience transformations as they evolve during transport downwind of the emission source. Aerosol evolution is dictated (between many other factors) by ambient conditions, such as relative humidity, temperature, and sunlight activity. This dissertation aims at providing a comprehensive characterization of atmospheric particles both at the bulk and single-particle level by implementing a unique combination of offline and online instrumentation.

The first chapter of this dissertation describes sources of atmospheric particles, as well as aerosol properties frequently examined to quantify their impact on our planet, such as chemical composition and light absorption. The second chapter delves into the wide range of techniques implemented in this study to characterize laboratory-generated and field-sampled aerosols. Notably, online measurements of chemical composition and optical properties were acquired with aircraft-deployed mass spectrometers and a particle-into-liquid sampler. These were frequently used to complement single-particle analysis employed with offline electron and X-ray microscopy techniques.

The third chapter describes a systematic approach to infer the viscosity of organic particles based on their morphology. Specifically, particles deform upon impacting a rigid surface during sampling, and the degree of deformation is highly influenced by the viscoelastic properties. Highly viscous and solid particles will retain their shape after sampling, whereas liquid-like particles will flatten drastically. Hence, we expanded on a semi-quantitative approach to infer the viscosity of particles based on their measured height-to-width aspect ratios. The fourth chapter discussed bulk measurements of chemical composition of smoke plumes emitted during wildfires in Western United States. An aerosol mass spectrometer was employed to quantify the mass concentration of key chemical species and their subsequent evolution during plume transport. Analyzed samples corresponded to daytime and nighttime particulate, which provided valuable insights on the impact of photochemical reactions on the composition and evolution of biomass burning particles. The fifth chapter serves as a follow-up study for that discussed in Chapter 4. Biomass burning particles were deposited onto substrates and taken for further chemical imaging. Scanning electron microscopy coupled with X-ray microanalysis provided single-particle information on the size, morphology, and elemental composition of aerosols sampled at different locations of the smoke plume. More detailed chemical information was acquired using synchrotron-based X-ray microscopy coupled with near-edge X-ray absorption fine structure spectroscopy. This technique distinguished between organic carbon, soot, and inorganic species, while also determining the contribution of functional groups, including alkenes, aliphatic, and carboxyl groups. Chemical imaging measurements were examined with respect to real-time optical data acquired onboard research aircraft. This facilitated correlating the chemical and light-absorbing properties of particles. The sixth chapter discusses a multi-modal, novel approach to distinguish between sources of soot-containing particles. Atomic force microscopy, integrated with Raman spectroscopy, was implemented for a screening of the morphological and spectral features of individual particles. Subsequently, automated μ-Raman was used to acquire the spectra of large ensembles of particles that are considered representative of the whole particle population. Emission sources of soot particles were then determined following two curve-fitting approaches previously established.

Overall, the studies discussed in this dissertation provide a comprehensive understanding of aerosol characteristics at the single-particle level, which is often overlooked by atmospheric model when predicting the impact of atmospheric particles on climate.