No description available.

Understanding the composition of atmospheric organic particulate matter (OPM) is essential for predicting its effects on climate, air quality, and health. However, the polar oxygenated fraction (PO-OPM), which includes a significant mass contribution from carboxylic acids, is difficult to speciate and quantitatively determine by current analytical methods such as gas chromatography-mass spectrometry (GC-MS). The method of chemical derivatization and two-dimensional GC with time of flight MS (GC×GC/TOF-MS) was examined in this study for its efficacy in: 1) quantifying a high percentage of the total organic carbon (TOC) mass of a sample containing PO-OPM; 2) quantitatively determining PO-OPM components including carboxylic acids at atmospherically relevant concentrations; and 3) tentatively identifying PO-OPM components. Two derivatization reagent systems were used in this study: BF₃/butanol for the butylation of carboxylic acids, aldehydes, and acidic ketones, and BSTFA for the trimethylsilylation (TMS) of carboxylic acids and alcohols. Three α-pinene ozonolysis OPM filter samples and a set of background filter samples were collected by collaborators in a University of California, Riverside environmental chamber. Derivatization/GC×GC TOF-MS was used to tentatively identify some previously unidentified α-pinene ozonolysis products, and also to show the characteristics of all oxidation products determined. Derivatization efficiencies as measured were 40-70% for most butyl derivatives, and 50-58% for most trimethylsilyl derivatives. A thermal optical method was used to measure the TOC on each filter, and a value of the quantifiable TOC mass using a gas chromatograph was calculated for each sample using GC×GC separation and the mass-sensitive response of a flame ionization detector (FID). The TOC quantified using TMS and GC×GC-FID (TMS/TOCGC×GC FID) accounted for 15-23% of the TOC measured by the thermal-optical method. Using TMS and GC×GC/TOF-MS, 8.85% of the thermal optical TOC was measured and 48.2% of the TMS/TOCGC×GC-FID was semi-quantified using a surrogate standard. The carboxylic acids tentatively identified using TMS and GC×GC/TOF-MS accounted for 8.28% of the TOC measured by thermal optical means. GC×GC TOF-MS chromatograms of derivatized analytes showed reduced peak tailing due in part to the lesser interactions of the derivatized analytes with the stationary phase of the chromatography column as compared to the chromatograms of underivatized samples. The improved peak shape made possible the greater separation, quantification, and identification of high polarity analytes. Limits of detection using derivatization and GC×GC/TOF-MS were μL injected for a series of C2-C6 di-acids, cis-pinonic acid, and dodecanoic acid using both butylation and TMS. Derivatization with GC×GC/TOF-MS was therefore effective for determining polar oxygenated compounds at low concentrations, for determining specific oxidation products not previously identified in OPM, and also for characterizing the probable functional groups and structures of α-pinene ozonolysis products.

Cluster characteristics of the 3D Lennard-Jones, LJ, fluid are determined by Metropolis Monte Carlo computer simulations. The percolation probability and cluster distribution is calculated for several state points in the gas-liquid equilibrium region of the LJ fluid. The cluster number distribution is used to analyze the distribution of clusters above and below the percolation threshold. Using scaling theory, the critical exponent, Ï , is determined from the cluster distributions. Deviation from the scaling law is evaluated using a modified scaling law that includes a surface term. It is found that the surface term is unnecessary in the gas-like area of the phase diagram. The density profiles of large non-percolating clusters are calculated in order to study the surface structure of the clusters. The coordination number within a cluster is calculated directly in the simulation and, with the cluster energy, is used to discern the amount of "liquid-like" structure of the cluster. The radius of gyration, R g, as a function of cluster size determines the fractal dimension, D f of the non-percolating and clusters above and below the percolation threshold density. Finite size effects are briefly studied and presented for a few of points. / Master of Science