Author

Date Approved

2026

Degree Type

Open Access Thesis

Degree Name

Master of Science (MS)

Department or School

Chemistry

Committee Member

Cory Emal, Ph.D.

Committee Member

Heather Holmes, Ph.D.

Committee Member

Gregg Wilmes, Ph.D.

Abstract

In recent years, varietal honey has been a massive target of adulteration through mislabeling and the addition of other sugars. Unethical companies do this to cut production costs while still charging the consumer full price. Previous studies have used nuclear magnetic resonance (NMR) to unravel possible adulteration in honey, but currently, there are no standard rapid methods to authenticate varietal honey. In our research, we collected NMR signatures (“fingerprints”) and combined them with linear discriminant analysis (LDA) to predict the varietal, country, and region of varietal honey. As a part of the project, we investigated whether an adjustment to the spectral data based on a signal-to-noise cutoff would provide better predictive ability. Our results are inconclusive as to whether a signal-to-noise adjustment provides a meaningful advantage; however, they demonstrate that combining NMR data with LDA enables prediction of honey varietal, country, and region at rates significantly better than random chance.

Share

COinS