Approximate Prediction of Chemical Changes in Peanut Oil during Intermittent Deep Frying Process Using UV-Visible Spectroscopy

Aims: The study aimed at predicting some quality changes in peanut oil during intermittent frying of carbohydrate and protein-based foods, using visible spectroscopic and chemometric methods.

Study Design: Completely Randomized Design and Multivariate Linear Regression were used to achieve this study.

Place and Duration of Study: The study took place at the Department of Food Science and Technology, Federal University of Technology, Akure between February and September 2017.

Methodology: Equal weight of yam chips and marinated chicken [carbohydrate (CHO) and protein-based (PRO) foods, respectively] were fried at 170°C for 20, 40, 60, 80, 100, 120, and 140 min with oil samples taken and topped at every interval. Changes in quality parameters such as colour density, free fatty acid (FFA), acid value (AV), peroxide (PV and saponification values (SV), K-extinction coefficients (K232nm, K266nm,K270nm and K274nm), and ΔK, with time, were determined. UV-Visible spectra (350 – 800 nm) of the oil samples were taken, and the data were elaborated with Principal Component Analysis (PCA) and Partial Least Square (PLS) regression techniques.

Results: Reduction in oxidative stability measured as increased values of FFA, PV, K-extinction values and ΔK were observed in all the samples and were particularly more pronounced (p = 0.05) in PRO-fried oils than those of CHO. Similarly, colour density increased linearly as frying time advanced in PRO-fried oil. PCA models of quality and spectra data revealed clear distinctions between PRO and CHO-fried oil samples. PLS regression coefficients showed that FFA (0.95), PV (0.92), SV (0.94), ΔK (0.98) and colour (0.95) were satisfactorily predicted; despite the relatively small sample size (15).

Conclusion: Non-destructive spectroscopic quality screening of vegetable oils during frying could facilitate rapid detection of degradation and the extent to which it can be reused. However, a large sample size is required to validate its reliability.