二阶导数预处理法在中红外光谱定量分析中的应用研究

Application of second derivative pretreatment in quantitative analysis using mid-infrared spectrum

介绍了运用最小二乘法建立傅立叶变换中红外光谱定量分析模型的原理和方法。以苯甲酸和邻苯二甲酸氢钾为实验材料获取红外吸收光谱，采用MAT—LAB工程语言编程，分别以吸收光谱和二阶导数光谱，进行中红外光谱定量分析建模波数信息选择，优选出2-4个波数点的光谱信息，建立了中红外光谱定量分析模型。用此模型预测混合物的量，预测值与实际值的相对误差小于5．0％。

In this paper, the principle and method of the quantitative analysis of fourier transform infrared spectroscopy, using Least-square as a method of information treatment, were introduced. The wavelength information was selected by matlab language design program in order to establish the quantitative analysis modelswith mid-infrared spectroscopy. Using benzoic acid and potassium hydrogen phthalate as experiment materials, quantitative analysis models to determine the mixture content were established using infrared spectrum or second derivative spectrum with two or four wavelength information as analytical information. Finally, an experiment with the mixture of benzoic acid and the potassium hydrogen phthalate was performed and the result shows that the relative error is smaller than 5.0% for the predicted result in comparison with the reat value.