乌醋溶液中乌洛托品含量的近红外光谱温度校正模型的研究

Study on Temperature Calibration Model of Hexamethylenetetramine(HA) Content in HA-HAc Solution by Near Infrared Spectroscopy Method

乌洛托品-醋酸(HA-HAc)溶液是制备奥克托今(HMX)的重要原材料。为了快速测定HA-HAc溶液中HA的含量,利用近红外(NIR)光谱技术,加入温度因子,建立了HA定量分析的温度校正模型。讨论了温度、光谱预处理方法对模型的影响。结果表明,温度会影响该溶液的吸光度,含温度因子的模型的平均相对误差减小了81.74%,模型准确度提高;光谱预处理可以消除外界因素对红外谱图采集的干扰,有利于提高建模的准确度,最佳预处理方法为一阶导数+Norris-derivative法。温度校正模型的评价参数为:相关系数R2=0.9997,均方差RMSEC=0.0388、内部交互验证均方差RMSECV=0.0441,预测均方差RMSEP=0.0350;模型的重复性标准偏差为0.0153%,此方法重复性为0.1298%,显示该法可用于HA-HAc溶液中HA含量的快速测定。

Hexamethylenetetramine-acetic acid（ HA-HAc） solution is an important raw material for preparing Octogen（ HMX）. In order to rapidly determine the HA content in HA-HAc solution,the temperature calibration model of quantitatively analyzing HA was established by near-infrared（ NIR） spectroscopy technology and adding temperature factor. The effect of temperature and spectral pretreatment method on the model was discussed. Results show that temperature will affect the absorbance of the solution,the average relative error of the model with temperature factor is reduced by 81.74%,and the accuracy of the model is improved. The spectral pretreatment can eliminate the interference of external factors to infrared spectrum collection,which is beneficial to improve the accuracy of modeling. The best pretreatment method is the first derivative ＋ Norris-derivative method. The evaluation parameters of the temperature correction model are correlation coefficient R2=0.9997,the root-mean square error of correlation RMSEC = 0.0388,the root-mean square error of cross-validation RMSECV = 0. 0441,and the root-mean square error of prediction RMSEP = 0.0350. The standard deviation of model repeatability is 0.0153% and the repeatability of this method is 0.1298%,revealing that the method can be used for fast determination of HA content in HA-HAc solution.