Study On The Application Of Near Infrared Spectroscopy In The Process Of Epoxy Curing And Glutathione Reaction

Near-infrared(NIR)spectroscopy,as a process analysis technology(PAT),has been successfully and widely used in scientific research and production process control fields,which have shown its advantages such as rapid,nondestructive and simultaneous determination of many physical and chemical properties of substances.Therefore,it is of great practical significance to study and expand its application range.In addition,NIR analysis is based on Beer-Lambert law,which means the spectrum has a linear relationship with the optical path and the concentration of measured component,respectively.However,in fact,the collected NIR spectrum is the superposition of the spectral components of the measured component and the coexisting components.In addition to the spectral influence of coexisting components,morphological changes,which can lead to the error of the optical path,also have a serious impact on the spectra of the mixture of reactants and products during the complex reaction process,such as epoxy curing,it is difficult to separate the interested information from the interference signals using the common spectral preprocessing methods.In addition,the partial least squares(PLS)method is mostly used to establish multivariate model in NIR quantitative analysis.Building models refers to the tedious work of samples collection and determination of reference data,and is usually time consuming as well as high cost,hindering the large-scale application of NIR technology.Therefore,in this dissertation some researches of relevant methods and applications have been carried out as follows:1.Study on the analytical method of activity of amines and the mechanism of epoxy curing reaction process by NIR spectroscopy.The temperature of epoxy curing reaction and the reactivity of amino of the amine curing agent have an effect on the three-dimensional network structure of the product,which in turn effects on the performance of the epoxy resin.Therefore,it is of great significance for deeply understanding the curing mechanism and for controlling the product quality to study the reactivity and kinetics of the primary and secondary amines in epoxy curing process.In this chapter,an effective method of determining the reactivity and kinetics of primary and secondary amines in epoxy curing process is proposed by the evolutionary NIR spectra of the materials of the curing process combined with angle vector and multivariate curve resolution,MCR-ALS.The resins of bisphenol A epoxy resin(E51)and M-xylylenediamine(MXDA)were prepared at different curing temperatures(50,60,70,80,90?),respectively.The evolutionary spectra of the mixture of reactants and the intermediate products during the curing process were collected online.The characteristic absorption bands of the primary and secondary amines overlap heavily in the wavenumber range of 6700-6330 cm-1.The evolutionary spectra are transformed into angle vectors to eliminate the interference of the state changes on the spectra during the reaction.Then,the angle vectors were processed by MCR-ALS,and the curves of concentration for primary and secondary amine with reaction time were obtained.According to the curves,the curves of conversion rate and reaction rate for primary and secondary amines were calculated.The reactivity of amino can be expressed by the reaction rate.It is shown that the reactivity of primary amine is different with that of secondary amine during epoxy curing process.There is a competitive reaction between them,and their competitiveness can be expressed by k2(SA/k2(PA);the reactivity of primary and secondary amines varies with curing temperature,and k2(SA)/k2(PA)increases significantly when curing temperature exceeds 60?,which has a significant influence on the network structure of the formed resins;the kinetics of the curing reaction follows the self-catalytic Kamal model in the early stage and does the revised model of incorporating the diffusion factor after vitrification.2.A real time analytical method of determining the glutathione(GSH)content in the production process by NIR spectroscopy.GSH is an active polypeptide substance with physiological functions and has been used widely in the clinical diagnosis of cancer and other diseases,food additives,skin care products and so on.The demand for the products of GSH is rising rapidly in recent years and the quality of the products has been concerned.In general,GSH is prepared by enzyme-catalyzed reaction,and the product concentration of GSH is a key parameter of the quality control of the production process.However,the traditional analytical methods are time-consuming for the GSH concentration detection,which cannot meet the real-time demand of process quality control.Therefore,the real-time determination of GSH content has become a problem to be solved urgently in the actual production.In this chapter,a new method for the real-time determination of GSH content in the production process is proposed.A multivariate calibration model of determination of GSH concentration was established by NIR spectroscopy combined with various spectral processing and PLS and the SEC,SEP and RPD of the model are 0.222,0.224 and 6.29,respectively,which meet the requirement of actual production quality control.The problem of GSH analysis in the production process mentioned above has successfully solved.3.Preliminary exploration of a non-modeling analytical method of eliminating the interference of the dominated water for determination of GSH content.When NIR technology is used to detect the reaction system in which water is the main coexisting component,it is a challenge for NIR spectral analysis to eliminate the influence of water on the determination of the measured component due to the strong absorption of water.In addition,modeling involves massive work of collecting samples and determining the reference data.The disadvantages of long cycle and high cost have been hindering the large-scale application of NIR technology.In this chapter,a quantitative method of non-modeling with elimination of the interference of the dominated water is proposed.The mixed aqueous solution of amino acids and GSH was used as the subject to study.First,discrete wavelet transform(DWT)was used to eliminate the spectral baseline drift caused by the environment and water.Then the principal component analysis(PCA)combined with the projection algorithm was used to extract the pure spectral component of GSH from the spectrum of the solution.Based on Lambert-Beer law,the pure spectral component of GSH is used to achieve the determination of reaction product content.A non-modeling quantitative analysis method has been realized with a certain accuracy.