Date of Award


Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)




Brian G. Frederick

Second Committee Member

Elizabeth A. Stemmier

Third Committee Member

Bruce L. Jensen


The identification and quantitation of organic compounds in complex mixtures by GC-MS is adversely affected by co-elution, because the acquired spectra are not pure, as well as by column bleed and ion chamber contaminants. Manual deconvolution, to purify such complex mass spectra needs expertise and is very labor intensive, but it is necessary to correctly identify compounds. The Automated Mass Deconvolution and Identification System (AMDIS) associates peaks in extracted ion chromatograms and then extracts a series of “components,” whose pure mass spectra can be matched in a library search. However, some compounds, for instance isomers, have similar mass spectra, and hence can be hard to uniquely identify. We have created a library of mass spectral data and retention indices for authentic standards and evaluated the ability of AMDIS to correctly identify and quantitate compounds in a complex pyrolysis oil sample. We propose that by lowering the mass spectral match criterion and imposing a statistically-based retention index constraint that the rate of false positives and false negatives can be decreased with the overall goal of increasing the rates of true positive and true negative identification.

The recently developed Formate Assisted Pyrolysis (FAsP) method has been applied to purified lignin and is a potentially valuable and renewable source of phenols, which are of current interest for production of phenol-formaldehyde resins and wood based panels. We extracted phenolics from Lignin fast pyrolysis oil by an alkali extraction and applied GC/MS analysis to determine die amount of phenolics that can be obtained from bio-oil and the efficiency of the extraction process. Quantitative analysis using calibration standards for selected compounds was carried out to assess the accuracy of the assumptions used in AMDIS to reduce complex chemical analysis to a practical level.