Improving vacuum ultraviolet spectroscopy using instrumental modifications and chemometric applications.

Abstract

Gas chromatography (GC) is a method of separating volatile and semi-volatile analytes in the gas phase. GC is often paired with a detection technique such as mass spectrometry (MS). GC-MS is a widely-used technique that allows for both qualitative and quantitative measurements of mixtures of gas-phase analytes. GC and MS libraries of hundreds of thousands of chemicals allow rapid and accurate identification of most analytes using GC-MS. Vacuum Ultraviolet (VUV) spectroscopy is a technique that, like MS, is useful for both qualitative and quantitative measurements of gas-phase molecules. Although VUV spectroscopy has been practiced since the late nineteenth century, it is only recently that a GC-compatible VUV absorption spectrometer has become commercially available. GCVUV provides near-universal detection for gas-phase molecules, has rapid data acquisition speeds, and can provide analyte-specific responses. In other words, GC-VUV’s niche overlaps with that of GC-MS. Whenever a new chemical measurement technology emerges, it is often combined with existing technologies (often called “hyphenation”). Additionally, the data provided by the new chemical measurement technology are often able to be manipulated and analyzed using novel methods. This dissertation documents the combination of VUV spectroscopy with existing technologies as well as the analysis of VUV spectroscopy data with novel chemometric methods. Chapter two details the combination of VUV spectroscopy with cryogenic preconcentration. With cryogenic preconcentration, gases can be trapped over an extended period of time and then released quickly. Combination with cryogenic preconcentration improves the poor sensitivity and small dynamic range of VUV spectroscopy. Chapters three and four discuss the hyphenation of VUV spectroscopy with GC and MS. VUV spectroscopy has many strengths of chemical identification that MS lacks and vice versa. By combining VUV spectroscopy with GC and MS, chemical identification is enhanced. Chapter five surveys a method of VUV data analysis that combines library searching with existing chemometric deconvolution techniques. This method improves chemometric deconvolution of VUV data, especially when the convoluted analyte signals are unknown. Chapter six summarizes the work and explores future directions, including automation of library searching, enhancement of VUV linear dynamic range, and improvements to library searching and VUV-MS capabilities.