Abstract

Background: Rendang is one of the traditional foods of West Sumatra made from processed beef, which is vulnerable to adulteration with non-halal meats such as pork.

Objective: This research aimed to develop the method for discriminating beef fat and lard (pig and wild boar) of rendang using a Fourier Transform Infra-Red (FTIR) combined with chemometric analysis.

Material and Methods: The training data set was rendang made of the binary mixture of beef and pork (pig and wild boar) with 11 concentrations. The testing data set was rendang obtained from several food stalls in Padang, West Sumatra Indonesia. Fat was extracted using the soxhlet method and then analyzed using ATR-FTIR spectroscopy at 4000-400 cm-1 wave numbers. The discriminant analysis (DA) was used to discriminate fat, while the Principle Component Regression (PCR) and Partial Least Square (PLS) models were used for quantification analysis.

Results: The result showed DA could classify rendang without misclassification at wave numbers of 1400-1000 cm-1 (pig) and 2800-1000 cm-1 (wild boar). About seven of ten commercial rendang were identified as beef, but three of them failed using the Beef-Pork-DA model. Meanwhile, five commercial rendang were classified as mixture and wild boar using the Beef-Wild boar DA model. In quantitative analysis, the best PCR model for pig was at a wave number of 1500-1000 cm-1 with a calibration R2 value of 0.9946, predicted R2 0.9926, RMSEC 0.0327, and RMSEP 0.0464. The best PLS model for wild boars was at wave numbers 1400-1000 cm-1 with the R2 calibration value of 0.9974, R2 prediction 0.9875, RMSEC 0.0229, RMSEP 0.0572.

Conclusion: FTIR spectroscopy combined with chemometrics successfully classified and quantified beef fat pork, and wild boar fat in rendang. However, further validation of the five rendang testing datasets using the DNA-based method that failed to identify as beef using the FT-IR method is necessary.

Keywords: Beef, chemometrics, fat, FT-IR, pork, wild boar, soxhlet extraction.