ESTIMATION OF QUALITY ATTRIBUTES AND FRYING TIMES OF COTTONSEED AND SUNFLOWER OIL BLEND DURING DEEP-FAT FRYING OF POTATO CHIPS

The chemical, physical and some sensory changes induced upon deep-fat frying of potato chips in a blend of
cottonseed and sunflower oil were investigated. The chemical changes were correlated with the frying times aiming
at establishing prediction models for estimating frying times. The oil blend became dark in colour after each successive
frying. The correlation coefficient between the frying time and A400 was very high (r=0.9967). Similar trend was
also observed in the relationship between the frying times and A450. Oil viscosity increased significantly with increasing
frying times indicating evidence of thermal decomposition of frying oil blend. The correlation coefficient between
the frying times and the refractive index was high (r=0.9898). Slight and continuous losses in unsaturated fatty
acids of the oil blend during deep-fat frying of potato chips have taken place.
The peroxide value (meq O2/kg oil) increased initially from 3.86 to 11.81 after 10 fryings, then dropped to 5.23
after 20 fryings followed by another rapid increase to 19.76 (meq. O2/kg oil) after 30 fryings. The UV absorption at
232nm increased initially from 0.4 to 1.5 after only 5 fryings and unchanged from the 5th frying to the 20th frying then
dropped after 30 fryings. A considerable increase in the TBA value has occurred up to the 20th frying and then the
TBA value showed no change for the next 10 fryings. The correlation coefficient between the frying time and TBA
value was 0.93.
The rate of polar compounds formation for the first 5 fryings was 3.62 mg/100mg oil/hr. whereas from the 5th
frying to the 30th frying such rate was reduced to only 1.12 mg/100mg oil/hr. Sensory evaluation of the oil blend indicated
that odour and colour scores were inversely related to the frying time. The most important and accurate mathematical
models which express the relationship between the independent and dependant variables were established.
The frying time could be successfully and accurately predicted by wise and intelligent chose of dependant variables
incorporated in the best fit equations.