LED based NIR spectroscopy for detection of lard adulteration in palm oil via chemometrics / Katrul Nadia Basri

• Main Author: Basri, Katrul Nadia
• Format: Thesis
• Language: English
• Published: 2017
• Subjects: Oil-bearing plants. Wax plants; Technological change
• Online Access: https://ir.uitm.edu.my/id/eprint/37198/1/37198.pdf; https://ir.uitm.edu.my/id/eprint/37198/
• Institution: Universiti Teknologi Mara

Adulteration is a major concern in religion and health. In this study, palm oil was used as it is one of the world most frequent used cooking oil. Near infrared spectroscopy technique is a reliable method to detect adulteration in cooking oil due to less sample preparation, rapid scan and applicable on on-site. Short wave near infrared spectroscopy (NIR) method was implemented in this experiment to detect the presence of lard adulteration in palm oil. MicroNIR was set up in two different scan modes to study the effect of path length to the performance of spectral measurement. Pure and adulterated palm oil sample were classified using soft independent modeling class analogy (SIMCA) algorithm with model accuracy more than 0.95 reported for both transflectance and transmission modes. Additionally, by employing partial least square (PLS) regression, the coefficient of determination (R2 ) of transflectance and transmission were 0.9987 and 0.9994 with root mean square error of calibration (RMSEC) of 0.5931 and 0.6703 respectively. In order to remove the uninformative variables, cumulative adaptive reweighted sampling (CARS) has been performed. The result of R2 and RMSEC after variable selection for transflectance and transmission were improve significantly. Based on the result of classification and quantification analysis, the transmission mode has yield better prediction model compared to the transflectance mode. Based on the critical wavelength obtained in CARS, 1200 nm and 1450 nm LED were chosen to replace the halogen tungsten as a light source for the NIR spectrometer. Using LED, the model accuracy obtained to detect pure palm oil is 0.99 which shows the classification result for LED is better compared using halogen tungsten light source. R2 acquired of PLS model for LED is 0.9970 for calibration and 0.9962 for prediction. RMSEC and RMSEP value are 0.8886 and 1.0173. For quantification analysis, the results of LED and halogen tungsten light source were on par.