二维相关光谱技术在含能材料快速分析中的应用研究

发布时间：2018-05-03 07:30

本文选题：二维相关光谱 + 近红外光谱　；参考：《光谱学与光谱分析》2016年S1期

【摘要】：以单基发射药为研究对象,制备了二苯胺(DPA)含量范围为0.8%~6%的40个发射药样品,获得其在不同二苯胺含量下的动态近红外光谱,并以二苯胺为外扰,采用二维相关光谱技术对吸收峰进行准确归属,提取了二苯胺的特征波段(4 575~4 960,5 330~5 835,6 050~6 650和6 735~7 080cm-1)。以二维相关谱提取的特征波段和间隔偏最小二乘(iPLS)及反向间隔偏最小二乘(BiPLS)筛选的特征波段分别建立二苯胺偏最小二乘模型,模型的相关系数Rc和Rp均大于0.98,均方根误差RMSEC、RMSECV和RMSEP均较小,说明二维相关光谱技术提取的特征波段建立的二苯胺定量模型与iPLS和BiPLS模型的预测性能均较优,这提高了建模波段选择的化学解释性。该技术为快速分析含能材料提供了一种途径。
[Abstract]:A total of 40 propellants with diphenylamine (DPA) content ranging from 0.8% to 6% were prepared, and their dynamic near infrared spectra were obtained under different diphenylamine contents, and diphenylamine was used as an external disturbance. The absorption peak of diphenylamine was accurately assigned by two-dimensional correlation spectroscopy. The characteristic bands of diphenylamine were extracted in the characteristic band of 4575U ~ (496) ~ (5 330) ~ (5 835) ~ 6 050 ~ (6 650) and 6 735 ~ (7) 080 cm ~ (-1) ~ (-1) of diphenylamine. The biphenylamine partial least squares model was established by using the characteristic bands extracted from the two-dimensional correlation spectra and the characteristic bands screened by the interval partial least squares (PLS) and the reverse interval partial least squares (BiPLS), respectively. The correlation coefficients R _ c and R _ p of the model are both greater than 0.98, and the root mean square error of RMSECV and RMSEP are both small, which indicates that the quantitative model of diphenylamine and the model of iPLS and BiPLS obtained by two-dimensional correlation spectrum technique have better prediction performance. This improves the chemical interpretation of the modeling band selection. This technique provides a way for rapid analysis of energetic materials.
【作者单位】：南京理工大学国家特种超细粉体工程技术研究中心;
【分类号】：O657.3;TQ560.72

【相似文献】