Application Of Near-infrared Spectroscopy In E-gelatin、Gasoline、and Crystallization Analysis Of Glucose

Near-infrared spectroscopy(NIR)can be used for qualitative and quantitative analysis in real time.It plays an important role in identifying material authenticity and in determining multi-properties of products for optimization and control of process industrial process.In this thesis,several NIR analysis methods were studied,including the rapid classification and identification methods of the E-gelatin and the dried donkey skin raw materials,the method of model transfer between the channels of the multi-channel online near-infrared spectrometer,and the method of testing the meso-stability area of crystallization process of glucose.The contributions of the thesis are as follows:1.Study on the rapid classification and identification methods of the E-gelatin and the dried donkey skinThe E-gelatin,being a long-standing old-style product,is well-known at home and abroad,it and its raw materials(donkey skin)are expensive.the adulteration of using other animal skins sometimes occurs,therefore,it is necessary to develop the rapid classification and identification methods of the E-gelatin and the dried donkey skin for market monitoring and for testing in acquisition of raw materials.In this chapter,the rapid classification and identification methods of the E-gelatin and the dried donkey skin was proposed by using near-infrared spectral combined with pattern recognition.The samples of different brands of E-gelatin,the dried donkey skin,other animal skin,the E-gelatin cakes were acquired from the manufactures and the market.Their NIR spectra were collected and their identification models established,respectively.The identification model for E-gelatin and other animal skin gelatin and that for donkey skin and other animal skin have a correct recognition rate of 100%.There are two different brands of E-gelatin cake models,with a correct recognition rate of 95%,and of 70%,respectively.2.Study on the method of model transfer between the channels of the multi-channel online near-infrared spectrometerMulti-channel on-line near-infrared spectrometer can detect on-line the multiple properties of products at multiple parts of a single set of production apparatus or the multi-properties of one kind of product at multiple sets of the apparatus for process optimization and control.Usually,the workload is large,the cycle long,and the cost high for modeling.However,there are optical and mechanical differences between the different channels of the multi-channel online near-infrared spectrum,the spectrum of different channels is very different,and the model built using the spectrum collected by one channel cannot be used for detection of other channels.In this chapter,a method of model transfer between different channels of a multi-channel on-line near-infrared spectrometer was proposed by using angular vector.The experiment was carried out on the pilot apparatus of catalytic reformatting in the China Petroleum and Petrochemical Research Institute.An online Fourier near-infrared spectrometer with 3-channels was installed at three sets of pilot apparatus of catalytic reformatting for collecting the NIR spectra of reformate gasoline.Forty-two gasoline samples were collected,and octane number and aromatics content of the samples were determined by using standard methods.The quantitative models of octane number and aromatic content were established using raw spectra for each channel.The samples of other channels were predicted by the models of channel 1.The near-infrared spectrum is converted into angular vector.The quantitative models of octane number and aromatic content were established using angular vector.The performance of models of using NIR spectra and of using angular vectors are compared for model transfer between different channels.The results show that the models of using angular vector can effectively adapt for the different channels.3.Study on the determination method of glucose metastable zoneCrystallization is an important means of separating and purifying substance.During the crystallization process,the metastable zone(MSZW)is between the solubility curve and the super solubility curve,which is related to multi-properties of solute and solvent,the presence or absence of stirring and the intensity of stirring,the presence or absence of seed crystals and the number and size of seed crystals,the cooling rate and so on.Therefore,the width of MSZW will change as the change of operating conditions.Wether the metastable region is too wide or too narrow,it’s not conducive to the growth of the crystal.In order to make the product having higher purity and ideal particle size,and the crystallization process can be controlled in the metastable region.Therefore,the width of the MSZW is a key parameter for designing and optimizing the crystallization process.In this chapter,the glucose crystallization process is selected as the study object.The methods of measuring saturation by infrared analysis and supersaturation by near infrared analysis,and the solubility equation of glucose in water is obtained:C0=1.114*10-4T2-4.166*10-2T+3.526(R2=0.9605)and super Solubility equation:C0=2.182*10-4T2-1.031*10-1T+12.62(R2=0.9596).Therefore,we can get the width of the MSZW of the glucose.