| Summary: | The mineralogy of atmospheric particles affects the amount of light scattered and absorbed by these particles as they pass through the atmosphere. A classification system described in this thesis was developed to automatically infer mineralogy of particles based on their measured elemental chemical composition and the elemental chemical composition of known minerals. When comparing the particle's elemental composition to preset ranges for each mineral, if a particle falls within that range, it is marked as elementally similar to that mineral. When each ratio is converted to python code and applied to elemental composition output files, each particle can be labeled as any chemically similar mineral. The elemental chemical composition of hundreds of thousands to millions of particles atmospheric particles entrapped in glacial ice samples can be measured using single particle Inductively Coupled Plasma-Time of Flight Mass Spectrometry (spICP-TOFMS) in about 2 hours (consuming only about 4 mL of melted ice). A spICP-TOFMS can measure the total number of particles per mL, elemental composition of each particle, and the particle size (mass equivalent diameter). This system of inferring mineralogy was assessed by comparison to Transmission Electron Microscope-Energy Dispersive X-ray Spectroscopy (TEM-EDXS) analysis of a limited (50) number of particles from Taylor Glacier samples with manual inference of mineralogy. The system described in this thesis was also used to infer mineralogy of mineral standards whose elemental composition was measured using spICP-TOFMS. After comparisons to multiple methods of inferring mineralogy, the system within this thesis was used to analyze samples of Taylor glacier measured by spICP-TOFMS. A one-year embargo was granted for this item. Academic Major: Earth Sciences Academic Major: Chemistry
|
|---|