Modeling Of Ethylene Propylene Copolymerization Processes Towards Sequence Length Distribution

Since the development of the petrochemical industry,polyolefin has become one of the most important commercial materials.How to improve the market competitiveness of polyolefin products,to control and optimize the production conditions of olefin copolymerization process are important issues to be solved in the polyolefin industry.However,weak awareness of microstructures of polyolefins,results in unsatisfactory in optimization of process conditions,process simulation,online estimates and advanced controls thereof.The sequence length distribution(SLD),molecular weight distribution(MWD)and chemical composition distribution(CCD)constitute the main microstructures of olefin copolymers,which determine their mechanical properties and processing properties.Therefore,accurate modeling of their microstructure is beneficial for precise adjustment and optimization of properties.Meantime,accurate characterization of the microstructures of polyolefin including MWD,CCD and SLD,is the basis of kinetic study,determination of kinetic constants,microstructure simulation in olefin copolymerization.The main framework and contributions in this thesis are listed as follows:1.High temperatue gel permeation chromatography(GPC),1H-NMR and 13C-NMR are used to characterize the MWD,average copolymer composition and SLD of polyolefins.In order to measure the MWD of polyolefins with broad MWD,Mark-Houwink calibration,universal calibration,triple-detectors and light scattering dual-detectors calibrations are investigated and compared.It is found that the triple-detectors calibration is the best method to analyze GPC curves of polyethylene possessing a broad MWD together with branching.In addition,analysis methods of 13C-NMR spectra are established repectively for dyads/triads expression and monomer sequence length expression.2.Based on the general mechanism of olefin coordination copolymerization,and introducing concepts of dead sequence and live sequence,a common mechanism of chain growth and sequence growth of olefin copolymerization over a coordination catalyst is proposed,as well as a model that can simultaneously predict the(MWD)and SLD for olefin copolymerization.The model achieves high efficiency operation on gPROMS through solving population balance equations of different molecular and sequential structures.The sequence structure distribution simulated by this method presents more abundant sequence information than the conventional trias sequence distribution method.3.Based on the established microstructure model and characterization of the output variables,a kinetic parameter estimation method is established.Firstly,the parameter estimability is ranked by principal component analysis-eigenvalue method or orthogonalization method,and the ranking results are compared.Then,the estimable parameter subset is estimated.The estimation method includes step estimation method and Wu’s MSE-based approcah.In addition,Flory distribution is used to deconvolve the MWD cureves to deal with the kinetic parameter estimation of multi-site catalysts.4.The methods of microstructure modeling and parameter estimation of olefin copolymerization process were applied to a metallocene catalyst and a vanadium-based catalyt to simulate the ethylene-propylene copolymerization process.According to the reaction characteristics of different catalytic systems,the mechanism equations were adjusted.For the metallocene catalyst,a model for ethylene sequence and propylene sequence structure distribution is established.Then the orthogonalization method is combined with Wu’s MSE approach to estimate parameters in the model.Therefore,those estimated parameters can be applied in the microstructure model to predict the sequence structure distribution and the corresponding dynamic changes.For the vanadium-based catalyst,because of the presence of 2,1-insertion of α-olefin,a new monomer is introduced in the modeling process to convert the copolymerization reaction into a terpolymerization reaction.At the same time,due to the presence of the 2,1-insertion,the sequence structure is depicted as uninterrupted methylene sequence rather than monomer sequences.