LAUSR

ABSTRACT Approaches for improving the performance of near-infrared diffuse reflectance spectroscopy (NIRDRS) have been extensively studied. The silver mirror was found to be a substrate to enhance the detection ability of this technique. For further optimization of the method, experimental and chemometric efforts were made in this study. Lysozyme samples of different concentrations were spotted on the silver mirror for measuring the spectra. Low-resolution spectra with repeated measurements were used for increasing the signal-to-noise ratio, and the chemometric methods of variable selection and nonlinear modeling were used to make the quantitative model more effective and accurate. The results show that the quality of the spectra was improved and a satisfactory quantitative model was obtained after optimization. The maximum deviation for the prediction samples is only 2.98 µg, which is much lower than the reported values in previous papers. Significant advances were obtained for NIRDRS for trace analysis.