The analysis of food authenticity and the detection of adulteration are major issues in the food industry and are attractive for consumers. Concerns about meat authenticity are increasing recently, due to great fraud scandals [1]. In the present research, the feasibility of diffuse reflectance infrared Fourier transform Spectroscopy (DRIFTS) (MIR; 4000 – 650 cm-1; 2500-15384 nm) in combination with
chemometrics was used for meat speciation of beef, chicken and pork. Spectral data were pre-processed using different techniques and explored by a Principal Component Analysis (PCA) to find out differences among pure samples. Partial Least Squares Discriminant Analysis (PLS-DA) with Median Center and 2nd derivative (21) (Savitzky–Golay) was used for meats classification.
Meat samples of beef (19 foreshank and hind shank samples) and chicken (30 breast and drumstick samples) were collected from local butchers in Tehran and Yazd cities, IR Iran in 2019. Thirteen pork samples (from different parts of body) were collected from different countries in 2019. Samples were first cleaned, removing the remaining skin and fat that could interfere in the analysis, and they were minced by a Moulinex (1000 W) mincer to homogenize them. Due to the high percentage of moisture in meat and the high intensity absorbance of the bands that appeared with the vibration of the water bonds, samples were dried in order to be able to observe the specific bands from protein and lipids in more detail. The samples were dried with freeze-dryer for 24 h. The spectra were recorded as transmittance values at each data point from 650 to 4000 cm−1 at a resolution of 8 cm−1. The spectra of each sample were taken as an average of 60 successive scans. Measurements are calibrated against the background gold. Spectral datasets were divided into calibration (70%) and validation (30%) sets with duplex algorithm. For PLS-DA model, sensitivity and specificity values in the validation set were 89% and 100% for beef, 100% and 100% for chicken, and 100% and 94% for pork, respectively. The overall accuracy of the method was 95%. In summary, a combination of infrared spectroscopy technology with PLS-DA can be used as an effective method to identify beef, chicken and pork meat samples.