红外光谱结合偏最小二乘法快速测定二元乙丙橡胶中的结合乙烯量

FAST DETERMINATION OF BOUND ETHYLENE CONTENT OF ETHYLENE-PROPYLENE RUBBER BY FT-IR COMBINED WITH PARTIAL LEAST SQUARE REGRESSION

为快速测定二元乙丙橡胶中的结合乙烯量，以^13C NMR测量数据为基础数据，将红外光谱基团信息与基础数据相结合，采用偏最小二乘（PLS）化学计量学方法，建立测量二元乙丙橡胶中结合乙烯量的定量校正模型。对影响分析模型质量的参数进行优化并对红外光谱谱峰与结合乙烯量的相关性进行分析。结果表明，在波数为600～1300cm^-1范围内，采用二阶微分光谱预处理建立的模型质量较好。红外光谱中与结合乙烯量相关性较强的吸收主要是CH2-基团、端基乙烯和-C-C间的振动吸收。模型验证结果表明，在模型适用范围内，二元乙丙橡胶中结合乙烯量的模型预测值与^13C NMR实测值的误差在±1．5％以内，模型的精度能够满足工业生产对该定量分析的要求。

Quantitative calibration models for fast determination of bound ethylene content in ethylene-pro- pylene rubber were established by partial least square regression based on data of ^13C NMR measurement combined with information from mid-infrared spectroscopy. Parameters affected the models were optimized and the correlation between spectra and bound ethylene content was analyzed. Results show that in the wave number range of 600--1 300 cm^-1 ,the absorption of --CH2-- group,the vibrational absorption between end group of ethylene and-C-C-have great coherence with the bound ethylene content, models based on second order derivative spectra are in better quality. The results of validation show that under the model application scope, the deviation of bound ethylene contents between predicted value and actual measurement is within ±1.5 %, which could meet the requirement of quantitative analysis during industrial production.