Doctoral Dissertations

Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Food Science and Technology

Major Professor

Sharon L. Melton

Committee Members

Marjorie P. Penfield, F. Ann Draughon, Arnold M. Saxton


Four commercial frying oils of different degrees of unsaturation, designated as AA, BB, CC and CC-P, were obtained from 3 frying outlets daily for 2 weeks in order to determine levels of degradation in fresh and used frying oils as measured by chemical and physical characteristics. Chemical measurements included fatty acid composition (C18:2/C16:0 ratio), iodine values (IV) and levels of total polar components (TPC), free fatty acids (FFA), peroxide values (PV), polymers (POL), conjugated dienoic acids (CDA), tocopherols (TOC) and flavor volatiles. Physical measurements included dielectric constant (FOS), viscosity (VIS) and oil color (ΔE). Correlation coefficients among chemical and physical characteristics were calculated. The best prediction equations using measured characteristics were determined for estimation of TPC as an indicator of oil degradation.

IV in fresh oils were 74.1 for BB, 99.6 for CC, 102.7 for AA, and 108.4 for CC-P. Across 7-days frying, TPC, FFA, POL, FOS, CDA, VIS and ΔE generally increased and TOC, IV and C18:2/C16:0 ratio decreased. Nine volatile aldehydes were identified but not from all oils; the most abundant was t,t-2,4-decadienal. Level of each volatile was dependent upon compound, oil and sampling day. TPC, a reliable measure of frying oil deterioration, increased during 7-days frying from 4.8 to 16.9% in BB, 4.9 to 18.7% in CC and 2.7 to 27.6% in CC-P, and increased during 5-days frying from 5.8 to 16.9% in AA. The FFA in CC-P, which was in a steam pressure chicken fryer, increased from 0.05 to 6.15% during 7-days frying. FFA at 7-days frying in BB was 2.53% and in CC was 1.70%. TPC, FFA, FOS, and ΔE were highly correlated with one another (r ≥ 0.80). IV and the C18:2/C16:0 ratio also were highly correlated. POL was correlated with TPC (r = 0.57) and with VIS (r = 0.65), but VIS had a lower correlation with TPC (r = 0.30). Removing samples with FFA >3.0% from correlation calculations increased r to 0.60 between VIS and TPC showing that oil samples with high FFA affected that relationship. The best single variable model for TPC prediction included FOS (R2 = 0.88).

Files over 3MB may be slow to open. For best results, right-click and select "save as..."