Abstract: Attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) was used to determine the selenium methionine content in rice flour in order to develop a fast and environmentally friendly method for the determination of selenium in food. IR spectra of samples with added selenomethionine were collected by an attenuated total reflection mid-infrared spectrometer. The partial least square method was used to develop a mathematical model and the model was validated by TQ Analyst software. The results showed that the infrared spectra in the range of 1 600–650 cm-1 were selected for analysis. After derivative preprocessing of the original infrared spectra, good prediction results were obtained by using either the leave-one-out method or the leave-more-out method. Comparatively, the leaveone- out model with first derivative pretreatment was simpler in calculation and gave more accurate prediction. The coefficient of determination for calibration (RC 2 ) of the left-one-out model was greater than 0.997, the number of factors was 6, the root mean square error of calibration (RMSEC) was 1.76–1.9, the coefficient of determination for prediction (RP 2) was 0.986 8, the root mean square error of prediction (RMSEP) was 3.27, and the relative prediction deviation (RPD) was 8.612 9. The predicted values with a relative deviation of less than 20% and 10% accounted for 92.91% and 82.27% of the total, respectively.

Key words: attenuated total reflection infrared spectroscopy, TQ Analyst software, selenium, quantitative analysis, model