Abstract:

To evaluate the ability of low field nuclear magnetic resonance (LF-NMR) in detecting oil adulteration, LF-NMR
and principal component analysis (PCA) were applied jointly to discriminate pure soybean oil from pure sesame oils (refined,
cold pressed and hot pressed), and partial least squares (PLS) regression was applied to analyze three sesame oils adulterated
with soybean oil and mixtures of two different sesame oils. The results showed that transverse relaxation time (0 to 900 ms)
was an important signal region for discrimination of oil samples. LF-NMR could completely distinguish four different pure
oil samples. The limit of detection (LOD) for sesame oil adulteration with soybean oil was 5% to 10%, and the LOD for
refined sesame oil adulteration with cold or hot pressed sesame oil was 10% to 20%. Therefore, LF-NMR can be used as a
rapid screening method to detect adulteration of oil.

Key words: low field nuclear magnetic resonance, oil adulteration, rapid detection, partial least squares, detection limit