Year of Award


Document Type


Degree Type

Master of Science (MS)

Degree Name


Department or School/College

Department of Chemistry

Committee Chair

Christopher P. Palmer

Commitee Members

Michael DeGrandpre, Tony Ward


PM2.5, wood stove, EPA, levoglucosan, resin acids, PAH


University of Montana

Subject Categories

Analytical Chemistry | Environmental Chemistry


Results are presented for the comparative analysis of the PM2.5 (Particulate Matter <2.5 µm) emissions from an EPA certified and a Traditional style wood stove using western larch. A total of 92 Quartz QMA 47mm filters were collected using a BGI PM2.5 SSC (Sharp Cut Cyclone) sampler from each stove type and analyzed on a gas chromatograph-mass spectrometer against blank and deuterated internal standards. The results were analyzed using a Welch’s t-test (α > 0.05) to statistically differentiate between stove designs for temperature, mass, levoglucosan, resin acids (abietic and dehydroabietic), and polycyclic aromatic hydrocarbons [PAH] (acenaphthene, anthracene, benz(a)anthracene, pyrene, and retene). There was no statistical difference in levoglucosan PM2.5 mass fraction between the stove types, yielding a mean levoglucosan fraction of 9.25% and 95% confidence interval of 8.43% to 10.25%. This suggests a uniform breakdown of cellulose to levoglucosan without considerable secondary byproducts regardless of wood stove design, and a 95% confidence interval for the conversion factor to calculate total woodstove PM2.5 from levoglucosan. Significant differences were observed for the resin acids, which both yielded smaller fractions in the EPA stove than the traditional, and for mean stove operating temperatures, the EPA stove was between 55.9°C and 102.3°C higher than the Traditional stove. Mass and PAH results require more data in order to be clearly interpreted with respect to this study. These results, along with previously published studies, add to the body of knowledge regarding EPA and traditional wood stove analysis and wood combustion conversion factors for use in source apportionment studies.



© Copyright 2015 Virginia M. Porden