Detection of lard adulteration in selected edible oils using gas chromatography mass spectrometry and NIR spectroscopy

• Main Author: Hussain, Mutia Nurulhusna
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: Oils and fats, Edible; Lard oi; Near infrared spectroscopy
• Online Access: http://psasir.upm.edu.my/id/eprint/104744/1/FPV%202022%208%20IR.pdf; http://psasir.upm.edu.my/id/eprint/104744/
• Institution: Universiti Putra Malaysia

Adulteration of food products has become a common problem in many countries. Adulteration may take the form of substitution of one species for another, where the food products from one species have been mixed intentionally with either a similar substitute material or a cheaper species. However, the use of pork and lard is a serious matter in Islam because foods containing ingredients from pig sources are haram (unlawful or prohibited) for Muslims to consume. Therefore, a reliable technique for the detection of lard adulteration in food products is necessary to protect Muslim consumers from intentional or unintentional fraud. The objective of this study was to apply near-infrared (NIR) spectroscopy combined with certain chemometric techniques for the detection and quantification of lard adulteration in selected edible oils. An analysis of fatty acids of authentic oils (i.e., palm olein (PO), soybean oil (SB), and canola oil (CO)) due to lard adulteration can complement the NIR spectral measurements. The presence of lard adulteration in PO, SB, and CO was analysed using NIR spectroscopy combined with soft independent modelling class analogy (SIMCA) and partial least-squares (PLS) in the region of 1,350–2,450 nm. This method can discriminate between pure and adulterated samples. The results revealed that the models predicted the adulterants with error limits of ± 0.83, ± 1.67, and ± 0.99 weight/weight for PO, SB, and CO, respectively. The PCA-developed models were able to classify lard-adulterated edible oil mixtures with almost 100% certainty. The adulterants were quantified using their respective PLS models within the same error limits as mentioned above. Furthermore, the study was extended for the classification of food product systems, specifically in biscuit formulation to verify the ability of the proposed method to classify between biscuits containing lard and without lard in its formulation. The spectra from the NIR analysis showed three significant peaks in the region of interest: 1,400–1,500 nm, 1,600–1,700 nm, and 2,100–2,200 nm. In this case, the result is similar to a previous study using edible oils and lard as samples. Using the same established model, the samples of adulterated biscuits and biscuits without lard were 100% correctly classified. A chemometric PLS method was used to derive a NIR calibration model. However, the model used previously could not predict the validation data, resulting in a high value of root mean square error of prediction (RMSEP). Therefore, several limitations of the established model and recommendations have been discussed. In conclusion, this study shows that the utilisation of NIR spectroscopy combined with suitable chemometric techniques can provide a reliable analytical tool for detecting pork and lard adulteration. The findings from this study can serve as a basis of reference for the research in halal food authentication.