近红外反射光谱法测定PDA阳离子度的研究

以37份实验室自制二甲基二烯丙基氯化铵与丙烯酰胺的共聚物(PDA)为样品,用沉淀滴定法对PDA阳离子度进行测定,并进行近红外光谱扫描,分别采用特征峰和全波谱段作为输入层神经元,采用RBF神经网络的数学转换方法建立近红外反射光谱(NIRS)定标模型,采用小波分析对光谱进行降噪预处理,建立PDA阳离子度的近红外预测模型,并将预测值与沉淀滴定法测定值进行比较,其外部验证决定系数大于0.90,预测分布趋势良好,对预测值与实测值进行t检验,结果显示预测值与实测值差异不显著。结果表明,采用近红外光谱数据建立的定标模型预测PDA阳离子度具有较高可行性。

Determination of cationic degree in PDA with near infrared reflectance spectroscopy

Cationic degree has been investigated as an important factor in polyacrylamide materials. Diallyl dimethyl ammonium chloride and acrylamide (PDA) was grafted by free radical polymerisation of acrylamide monomer (AM) onto the cationic monomer dimethyl diallyl ammonium chloride (DMDAAC). In the present study, near infrared reflectance spectroscopy (NIRS) was used as a rapid and accurate method to determine the cationic degree in the PDA. In this experiment, the near infrared spectra of 37 PDA samples that were self-prepared in the laboratory from 900.00 to 1 700.00 were collected. The characteristic peaks and the entire spectrum segment as the input layer neurons in radical basis function (RBF) were investigated for establishing the mathematical conversion NIRS calibration mode. For reduction of the NIR spectrum noise, the wavelet analysis was used as pretreatment process. The measured value was determined by using precipitation titration and a comparison between the simulated value and measured value was made. It was found that the external validation determination coefficient was more than 0.9, and the simulation value is in good agreement with the measured value. The statistics analysis showed that there was no significant difference between simulated value and measured value. Therefore, the calibration model (RBF neural network) established in this paper exhibited a remarkable feasibility for predicting the cationic degree of PDA based on the near infrared spectroscopy.