QUALITY ALTERATIONS OF FOUR FRYING FATS DURING LONG‐TERM HEATING (CONVENTIONAL ANALYSIS AND NIRS CALIBRATION)

Abstract

ABSTRACT Rapeseed oil, sunflower oil, pig (mangalica breed) and goose fat were heated (36 h, 4 h sampling intervals) at seven temperatures (140, 150, 160, 165, 170, 175 and 180C; n = 256) to follow quality alterations. Acid value (AV) increased permanently, while the increase of peroxide value (PV) was followed by a drop. Carbonyl number tended to increase, while p‐anisidine value (pAV) showed two‐peaked alterations. Total polar material (TPM) and dimeric and polymeric triacylglycerol content of oils increased up to over 60 and 25%, while by animal fats, these reached 45 and 19%, respectively. Near‐infrared spectroscopy (0.1 mm layer, transflectance mode, 800–2,400 nm wavelength, second derivative, modified partial least squares regression) was used for quantitative calibrations. Estimation accuracy of sampling event, heat‐sum (heating hours × temperature) exceeded R2 = 0.94, with over 0.9 cross‐validation values, analyzing fats separately. Calibrations for AV, PV, pAV and TPM reached practical application levels.PRACTICAL APPLICATIONSThe present approach aimed to follow‐up a heating regime frequently occurring in the culinary routine, in terms of heating temperature and duration. Four frying fats were used and conventional fat quality indices (acid value, peroxide value, carbonyl number, p‐anisidine value, total polar material, dimeric and polymeric triacylglycerols) were determined. Near‐infrared spectroscopy (NIRS) analysis was used to develop quantitative calibrations for the quality indices, and reached practical application levels for acid and peroxide values, p‐anisidine value and total polar material content. The spectra‐based estimation of the extent of use, i.e., the duration and the temperature, and as expressed together, the heat‐sum was predicted excellently. NIRS analysis details were set to be simple, i.e., no sample preparation, room temperature, easy sample application and sample‐cup cleaning. Thus, a rapid measurement methodology is presented for the accurate prediction and evaluation of frying fat deterioration, without toxic substances.