Micro-frying of potato cubes
Micro-frying study was conducted with potato cubes following the previously reported method (Hwang, et al., 2013), which was slightly modified in the current study. Amino acid or amino acid salt (5.5 mM) and SBO (20.0 g) were weighed in a beaker and stirred with a magnetic bar to form a suspension. The suspension (6.0 g) was transferred to the same vial used for the heating study. For TBHQ, which is soluble in SBO, 0.02% (w/w) was dissolved in 20.0 g SBO, and the solution (6.0 g) was transferred to the vial. The ratio of surface area to volume of oil in this vial was 0.82 cm2/cm3. A potato was peeled, cut into 0.6 cm x 0.6 cm cubes, and soaked in DI water until frying. Vials were randomly placed in an oil bath, of which the temperature was maintained at 180 ºC on a hotplate using a temperature probe placed in the oil bath. After 1 h, the potato cube attached to the tip of a glass pipette was fried for 4 min every 20 min five times. After a 2 h and 40 min break while oil was heated at 180 ºC, frying was continued another five times in the same way. The oil was, then, left at 180 ºC for 1 h and 40 min. In this procedure, the total heating time was 8 h, and the total number of frying was 10 times. Oil samples were transferred to 1-dram vials and stored at -20 ºC until analyses. The oxidation of oil was determined in a week after frying.
Statistical analysis
Three experimental replicates were carried out for heating and frying studies. Since several sets of heating experiments were conducted in different days, control SBO samples were included in each set of heating experiments, which were then used for normalization of PTAG and NMR data. Two experimental replicates were carried out for heating studies for the synergistic effect and for the activity in five other vegetable oils. PTAG, fatty acid composition, and tocopherols were determined at least in duplicate. Data were analyzed with one-way analysis of variance (ANOVA) and Tukey-Kramer Honestly Significant Difference test atP < 0.05 using JMP 15 program (SAS Institute, Cary, NC, USA). Multivariate correlation analysis was conducted with the same program.