[Influence of optical path length on NIR analysis results for trace metal determination in Chinese rice wine]

The prediction performance of near infrared (NIR) spectra with different optical path-length for trace metal (potassium, calcium, magnesium, zincum, and iron) determination was investigated. NIR transmission spectra of Chinese rice wine were collected in rectangular quartz cuvette with different optical path lengths (1, 2, 5 and 10 mm) using Fourier transform near infrared (FT-NIR) spectrometer in the wavelength range of 800-2 500 nm with air as the reference. The reference data for potassium, calcium, magnesium, zinc, and iron were determined by atomic absorption spectroscopy (AAS). Calibration models were developed by partial least squares (PLS) regression. The PLS models of NIR spectra group with 5 mm path length gave the best calibration result. The determination coefficients (r2) for potassium, calcium, magnesium, zincum, and iron were 0.93, 0.85, 0.93, 0.72, and 0.66, respectively, and the root mean square error of cross validation (RMSECV) for the five elements were 26.5, 35.6, 4.63, 0.26, and 0.64 mg x L-(-1), respectively, whereas the models established by NIR spectra group of 10 mm path-length was the worst. And the r2 values for potassium, calcium, magnesium, zincum, and iron were 0.61, 0.65, 0.63, 0.09, and 0.25, respectively. The results indicated that the optical path length has an influence on the NIR analysis results for trace metal determination in Chinese rice wine, and that the appropriate path length for the NIR analysis should be determined by comparison analysis.