The following is an abstract for the selected article. A PDF download of the full text of this article is available here. Members may download full texts at no charge. Non-members may be charged a small fee for certain articles.
Classification of Vegetable Oils by FT-IR
Volume 51, Number 8 (Aug. 1997) Page 1118-1124
Dahlberg, Donald B.; Lee, Shawn M.; Wenger, Seth J.; Vargo, Julie A.
The Fourier transform infrared (FT-IR) spectra of 27 brands of 10 types of cooking oils and margarines were measured without temperature control. Attempts to predict the vegetable source and physical properties of these oils failed until wavelength selection and multiplicative signal correction (MSC) were applied to the FT-IR spectra. After pretreatment of the data, principal component analysis (PCA) was totally successful at oil identification, and partial least-squares (PLS) models were able to predict both the refractive indices [standard error of estimation (SEE) 0.0002] and the viscosities (SEE 0.52 cP) of the oils. These models were based predominately on the FT-IR detection of the cis and trans double-bond content of the oils, as well as small amounts of defining impurities in sesame oils. Efforts to use selected wavelengths to discriminate oil sources were only partially successful. These results show the potential utility of FT-IR in the fast detection of substitution or adulteration of products like cooking oils.