Dielectric and fluorescence spectroscopy techniques for determination of palm olein degradation during batch deep fat frying

• Main Author: Ibrahim, Nur Ul Atikah
• Format: Thesis
• Language: English
• Published: 2017
• Online Access: http://psasir.upm.edu.my/id/eprint/70959/1/FK%202017%207%20-%20IR.pdf; http://psasir.upm.edu.my/id/eprint/70959/

Degradation of frying oil by series of complex chemical reaction during frying process will lead to the formation of undesirable compounds which affecting the quality of frying oil. Conventional chemical methods for frying oil quality assessment are well known to be time consuming, require skilled operator and involve significant amount of harmful solvents. In this study, dielectric and fluorescence spectroscopy techniques were used to monitor the degradation of refined, bleached, and deodorized palm olein (RBDPO) in batch deepfat frying. Intermittent frying experiment was conducted for five consecutive days. In total, 30 batches of French fries were fried at the temperature of 185±5⁰C in each day. The dielectric constant, fluorescence intensity, total polar compounds (TPC) and free fatty acids (FFA) were measured for analysis. The results showed that there were generally significant differences in the mean of dielectric constant and intensity of fluorescence emission and excitation spectra of RBDPO over five days of frying (P<0.0001). The mean dielectric constant of RBDPO increased from 3.09 to 3.17 as the frying cycle increased. Meanwhile, the mean measured fluorescence intensity of RBDPO for emission and excitation spectra were generally decreased for all prominent peaks observed. For example, mean measured fluorescence intensity decreased from 201.83 to 20.59 arbitrary unit (a.u) and from 240.04 to 24.73 a.u at emission peak of 476 nm and excitation peak of 396 nm, respectively. PLS model developed using dielectric spectroscopy data showed good potential to predict the degradation of RBDPO with the application of GA selector with R2 of 0.91 and 0.95 for TPC and FFA, respectively and RMSECV of 1.06 %. Whereas, PLS model constructed using fluorescence emission and excitation spectra data showed substantially good potential to predict the degradation of RBDPO even without the application of GA. For example, PLS model developed by correlating fluorescence emission spectra data to TPC and FFA without GA selector had R2 of 0.98 and 0.99 for TPC and FFA, respectively with RMSECV of 0.52 % and 0.01 %, respectively. The result from this study indicated that both spectroscopy data were significantly sufficient to be used for monitoring the degradation of RBDPO during batch deep fat frying. This study could help effort on investigating reliable techniques for in-situ determination of physicochemical properties of palm olein.