Simulation of Crystallization Analysis Fractionation (Crystaf) of Linear Olefin Block Copolymers

Summary: Linear olefin block copolymers (OBCs) have microstructures that are unique among polyolefins and exhibit properties that are different from those of other polyolefin elastomers. Characterizing their chain microstructures is a challenging task, as conventional characterization techniques cannot probe directly block length distribution or composition. In this work, we used a Monte Carlo model to predict the microstructure details of OBCs and a modified version of the Crystaf model previously developed in our groups to describe theoretical Crystaf profiles for model OBCs. This model can be used as a tool to interpret Crystaf results of these interesting new polyolefins and to relate them to OBC microstructures. Effects of polymerization parameters on OBC microstructure and Crystaf profiles were also discussed.     

Determination of Operating Conditions of Ethylene/1-Octene Copolymerization Using Artificial Neural Network (ANN)

An artificial neural network (ANN) is applied to determine appropriate parameters in copolymerization of ethylene and 1-octene via metallocene catalytic system for producing a copolymer with desired chain microstructures. The polymerization parameters of interests are polymerization temperature, ethylene pressure, and the amount of hydrogen used. The ANN used is a feed-forward network with a back propagation learning method and has a 5-6-6-3 architecture. When comparing with both training and testing experimental data sets, it was found that ANN can provide a good guesstimation of polymerization parameters.     

A Mathematical Model for the Kinetics of Crystallization in Crystaf

Summary : A series of ethylene homopolymers and ethylene/1-hexene copolymers with different molecular weight distributions (MWD) and chemical composition distributions (CCD) was analyzed by crystallization analysis fractionation (Crystaf) at several cooling rates to investigate the effect of MWD, CCD, and cooling rate on their Crystaf profiles. Using these results, we developed a mathematical model for Crystaf that considers crystallization kinetic effects ignored in all previous Crystaf models and can fit our experimental profiles very well.     

Estimation of Polymerization Conditions Needed to Make Ethylene/1‐olefin Copolymers with Specific Microstructures Using Artificial Neural Networks

Two artificial neural network models (forward and inverse) are developed to describe ethylene/1‐olefin copolymerization with a catalyst having two site types using training and testing datasets obtained from a polymerization kinetic model. The forward model is applied to predict the molecular weight and chemical composition distributions of the polymer from a set of polymerization conditions, such as ethylene concentration, 1‐olefin concentration, cocatalyst concentration, hydrogen concentration, and polymerization temperature. The results of the forward model agree well with those from the kinetic model. The inverse model is applied to determine the polymerization conditions to produce polymers with desired microstructures. Although the inverse model generates multiple solutions for the general case, unique solutions are obtained when one of the three key process parameters (ethylene concentration, 1‐olefin concentration, and polymerization temperature) is kept constant. The proposed model can be used as an efficient tool to design materials from a set of polymerization conditions.

     

盐湖水化学类型的人工神经网络判别方法

研究了作为典型径向基函数网络之一的概率神经网络在盐湖水化学类型分类预测中的应用，验证了该方法的可靠性，得到了满意的分类预测结果。实验结果和网络结构分析表明，概率神经网络方法比熟知的反向传播算法(BP)网络要好。概率神经网络的研究应用为化学模式识别提供了一个新工具。     

An experimental and numerical study on crystallization analysis fractionation (Crystaf)

Crystallization analysis fractionation (Crystaf) is a new technique used to estimate the chemical composition distribution (CCD) of semi-crystalline copolymers. In this study, the effect of chain microstructure and operation parameters on Crystaf profiles was investigated using a series of ethylene/1-hexene copolymers and their blends. The Crystaf profiles were also modeled via stochastic simulation based on the distribution of average ethylene sequence lengths.     

Effect of Chain Microstructure and Cooling Rate on Crystaf Calibration Curves: An Experimental Study

Summary: Crystallization analysis fractionation (Crystaf) is a polymer characterization technique based on differences in chain crystallizabilities in a dilute solution during non-isothermal crystallization. Crystaf profiles, a weight distribution function of chains crystallized at each temperature, can be used to infer the chemical composition distribution (CCD) of copolymers when a Crystaf calibration curve, a relationship between peak crystallization temperature and average comonomer content, is known. In this investigation, the effect of the number average molecular weight, comonomer type, and cooling rate on Crystaf calibration curves were experimentally investigated. It was found that the cooling rate and comonomer type may strongly affect Crystaf calibration curves, while the influence of molecular weight is relatively subtle.     

On the Robustness of Forward and Inverse Artificial Neural Networks for the Simulation of Ethylene/1‐Butene Copolymerization

Forward and inverse artificial neural network (ANN) models are used to describe ethylene/1‐butene copolymerization with a model catalyst having two site types. The forward ANN predicts number and weight average molecular weights, average comonomer content, and polymer yield as a function of a set of polymerization conditions, while the inverse model estimates polymerization conditions needed to produce copolymers with desired microstructures. The forward model is found to be robust and resilient to random noise introduced into the datasets. The inverse model, however, leads to multiple solutions (several polymerization conditions can produce polymers with similar microstructures) and is sensitive to random noise in the data. Although the polymerization conditions estimated from inverse ANN are different from the model data, the estimated polymerization conditions are found to provide similar microstructures even with the random noise.     

Comparison of Fractionation Methods for the Characterisation of Short Chain Branching as a Function of Molecular Weight in Ethylene 1-Olefin Copolymers

Summary: Rapid and automated analysis of polyolefins is becoming essential for product development in industry. Quantifying short chain branching in ethylene 1-olefin copolymers is common practice. Several different methods are available to perform this type of analysis. Preparative fractionation followed by subsequent analysis of the fraction by SEC and NMR, SEC-FTIR and SEC-IR were studied towards their applicability in polyolefin research and product development environment. The method of choice is defined by prerequisites such as accuracy, labour and time demands but also in versatility and practicability. The most accurate method is limited in terms of sample throughput and the most practical method is limited towards resolution of very low branching. SEC-FTIR is capable to measure even heterogeneous low branched samples like bimodal high density polyethylene in rapid and satisfactory matter.     

Mathematical modeling of crystallization analysis fractionation of ethylene/1‐hexene copolymers

Crystallization analysis fractionation (Crystaf) is a polymer characterization technique for estimating the chemical composition distributions of semicrystalline copolymers. Although Crystaf has been widely used during the recent years, it is still a relatively new polymer characterization technique. More quantitative understanding of its fractionation mechanism is essential for further developments. In this work, three ethylene/1‐hexene copolymers with different 1‐hexene fractions, but similar number‐average molecular weights, were analyzed by Crystaf at several cooling rates. A mathematical model was proposed to describe the effect of comonomer fraction and cooling rate on Crystaf fractionation from a fundamental point of view. The model describes the experimental Crystaf profiles of ethylene/1‐hexene copolymers with different 1‐hexene fractions measured at distinct cooling rates very well. © 2007 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 45: 1010–1017, 2007     

神经元网络用于PCDD定量构效关系的研究

研究了不同PCDD（全名Polychlorinateddioxin）同系物分子结构的表达及特征参数的选择,应用神经元网络方法对其分子结构与色谱保留值进行了关联。对49种PCDD同系物在DWS往上不同温度下保留时间进行了预测,结果95％以上的数据点相对误差小于10％,而80％以上的数据点相对误差小于5％。     

人工神经网络方法用于多组分分光光度分析

本文介绍了人工神经网络（ANN）原理,详细讨论了网络参数的选择及其对网络预报结果的影响。并将人工神经网络应用于干扰严重的五组分体系的分光光度同时分析,其预测结果优于正交分解计算方法。     

Simultaneous Spectrophotometric Determination of Iron and Cobalt in Micellar Medium by Using a Principal Component Artificial Neural Network and Multivariate Calibration

A simple and reliable method for simultaneous spectrophotometric determination of iron(II) and cobalt(II) has been established. The method is based on complex formation with 1‐(2‐pyridylazo)‐2‐naphtol (PAN) in a micellar medium. Despite a spectral overlap, Fe²⁺ and Co²⁺ have been simultaneously determined with chemometric approaches involving principal component artificial neural network (PC‐ANN), principal component regression (PCR) and partial least squares (PLS). Various synthetic mixtures of iron and cobalt were assessed and the results obtained by the applications of these chemometric approaches were evaluated and compared. It was found that the PC‐ANN method afforded relatively better precision than that of PCR or PLS. The proposed method permits detection limits of 0.05 and 0.07 ng mL^?1 for Co and Fe, respectively. The influences of pH, ligand amount, solvent percentage and time on the absorbance were also investigated. The proposed method was also applied satisfactorily for the determination of Fe(II) and Co(II) in real and synthetic samples.     

Ethylene/1‐hexene copolymers synthesized with a single‐site catalyst: Crystallization analysis fractionation,modeling, and reactivity ratio estimation

A series of poly(ethylene‐co‐1‐hexene) samples made with rac‐ethylene bis(indenyl)zirconium dichloride/methylaluminoxane were analyzed by crystallization analysis fractionation (CRYSTAF). The nine samples had comonomer contents of 0–4.2 mol % 1‐hexene with a narrow range of molecular weights (34,000–39,000 g/mol). Because all the copolymer samples had narrow, unimodal chemical composition distributions, they were ideal as calibration standards for CRYSTAF. A linear calibration curve was constructed relating the peak crystallization temperature from CRYSTAF operated at a cooling rate of 0.1 °C/min and the comonomer content as determined by ¹³C NMR. Reactivity ratios for ethylene and 1‐hexene were estimated by the fitting of reactant liquid‐phase compositional data to the Mayo–Lewis equation. It was found that a value of the 1‐hexene reactivity ratio could not be unequivocally determined from the set of samples analyzed because the range of comonomer incorporation was too narrow. Stockmayer's bivariate distribution was used to model the fractionation process in CRYSTAF, and although a good fit to experimental CRYSTAF profiles was attained, the model did not fully describe the underlying crystallization phenomena. © 2002 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 40: 2595–2611, 2002     

近红外光谱-人工神经网络法用于粉末药品阿斯匹林的非破坏定量分析

研究了应用人工神经网络进行粉末药品的非破坏定量分析。使用阿斯匹林粉末药品的近红外漫反射一阶导数光谱数据建立人工神经网络模型,预测未知样品。讨论了影响网络的各参数,使用了新的网络评价标准－逼近度。     

人工神经网络—近红外光谱法用于甲氧苄胺嘧啶粉末药品的非？…

近红外(NIR)光谱分析技术已应用于制药、化妆品、烟草、食品、化学药品、聚合物、纺织品、油漆涂料、煤炭和石油工业等各个领域的质量监控.近年来,NIR光谱分析技术也应用于药品分析中,因该方法具有非破坏性,样品不需要复杂的预处理和分离即可直接测定.它可对药物进行定性和定量测定以及多晶、光学异构体和湿度的测定.近红外光谱法用于无损非破坏测定胶囊类以及片剂的研究已有报道[1,2].NIR光谱在使用中也有一定的局限性,主要是结构复杂,谱图重叠多,在进行定性和定量分析中需采用一定的数据处理才能获得准确可靠的分析结果.在定量分析中,…     

Nondestructive Quantitative Analysis of Analgini Powder Pharmaceutical by Near-Infrared Spectroscopy and Artifical Neural Network Technique

ABSTRACT

This paper demonstrates the usefulness of near-infrared (NIR) spectra and artificial neural network (ANN) in nondestructive quantitative analysis of pharmaceuticals. Real data sets from near-infrared reflectance spectra of analgini powder pharmaceutical were used to build up an artificial neural network to predict unknown samples. The parameters affecting the network were discussed. A new network evaluation criterion, the degree of approximation, was employed. The overfitting was discussed. Owing to the good nonlinear multivariate calibration nature of ANN, the predicted result was reliable and precise. The relative error of unknown samples was less than 2.5%