Discrimination of various poly(propylene) copolymers and prediction of their ethylene content by near‐infrared and Raman spectroscopy in combination with chemometric methods

Near‐infrared (NIR) diffuse reflectance (DR) spectra and Fourier‐transform (FT) Raman spectra were measured for 12 kinds of block and random poly(propylene) (PP) copolymers with different ethylene content in pellets and powder states to propose calibration models that predict the ethylene content in PP and to deepen the understanding of the NIR and Raman spectra of PP. Band assignments were proposed based calculation of the second derivatives of the original spectra, analysis of loadings and regression coefficient plots of principal component analysis (PCA) and principal component regression (PCR) (predicting the ethylene content) models, and comparison of the NIR and Raman spectra of PP with those of linear low‐density polyethylene (LLDPE) with short branches. PCR and partial least squares (PLS) regression were applied to the second derivatives of the NIR spectra and the NIR spectra after multiplicative scatter correction (MSC) to develop the calibration models. After MSC treatment, the original spectra yield slightly better results for the standard error of prediction (SEP) than the second derivatives. A plot of regression coefficients for the PCR model shows peaks due to the CH₂ groups pointing upwards and those arising from the CH₃ groups pointing downwards, clearly separating the bands due to CH₃ and CH₂ groups. For the Raman data, MSC and normalization were applied to the original spectra, and then PCR and PLS regression were carried out to build the models. The PLS regression for the normalized spectra yields the best results for the correlation coefficient and the SEP. Raman bands at 1438, 1296, and 1164 cm^?1 play key roles in the prediction of the ethylene content in PP. The NIR chemometric evaluation of the data gave better results than those derived from the Raman spectra and chemometric analysis. Possible reasons for this observation are discussed. © 2002 Wiley Periodicals, Inc. J Appl Polym Sci 87: 616–625, 2003     

Heat consumption forecasting using partial least squares, artificial neural network and support vector regression techniques in district heating systems

Effective management of district heating networks depends upon the correct forecasting of heat consumption during a certain period. In this work short-term forecasting for the amount of heat consumption is performed first to validate the three forecasting methods: partial least squares (PLS) method, artificial neural network (ANN), and support vector regression (SVR) method. Based on the results of short-term forecasting, one-week ahead forecasting was performed for the Suseo district heating network. Data of heat consumption and ambient temperature during January and February in 2007 and 2008 were employed as training elements. The heat consumption estimated was compared with actual one in the Suseo area to validate the forecasting models.     

Measurement of Whole Soybean Fatty Acids by Near Infrared Spectroscopy

Whole soybean fatty acid contents were measured by near infrared spectroscopy. Three calibration algorithms—partial least squares (PLS), artificial neural networks (ANN), and least squares support vector machines (LS-SVM)—were implemented. Three different validation strategies using independent sets and part of calibration samples as validation sets were created. There was a significant improvement of the prediction precision of all fatty acids measured on relative concentration of oil compared with previous literature using PLS (standard error of prediction of 0.85, 0.42, 1.64, 1.67, and 0.90% for palmitic, stearic, oleic, linoleic and linolenic acids respectively). ANN and LS-SVM methods performed significantly better than PLS for palmitic, oleic and linolenic acids. Calibration models developed on relative concentrations (% of oil) were compared to prediction models created on absolute fatty acid concentration (% of weight) and corrected to relative concentration by multiplying by the predicted oil content. While models were easier to develop in absolute concentration (higher coefficients of determination), the multiplication of errors with the total oil content model resulted in no net precision improvement.     

Discrimination of edible oil products and quantitative determination of their iodine value by Fourier transform near-infrared spectroscopy

This work demonstrates the application of partial least squares (PLS) analysis as a discriminant as well as a quantitative tool in the analysis of edible fats and oils by Fourier transform near-infrared (FT-NIR) spectroscopy. Edible fats and oils provided by a processor were used to calibrate a FT-NIR spectrometer to discriminate between four oil formulations and to determine iodine value (IV). Samples were premelted and analyzed in gass vials maintained at 75°C to ensure that the samples remained liquid. PLS calibrations for the prediction of IV were derived for each oil type by using a subset of the samples provided as the PLS training set. For each oil formulation (type), discrimination criteria were established based on the IV range, spectral residual, and PLS factor scores output from the PLS calibration model. It was found that all four oil types could be clearly differentiated from each other, and all the validation samples, including a set of blind validation samples provided by the processor, were correctly classified. The PLS-predicted IV for the validation samples were in good agreement with the gas chromatography IV values provided by the processor. Comparable predictive accuracy was obtained from a calibration derived by combining samples of all four oil types in the training set as well as a global IV calibration supplied by the instrument manufacturer. The results of this study demonstrate that by combining the rapid and convenient analytical capabilties of FT-NIR spectroscopy with the discriminant and predictive power of PLS, one can both identify oil type, as well as predict IV with a high degree of confidence. These combined capabilities provide processors with better control over their process.     

紫外光谱结合偏最小二乘法监测氯化1-羟乙基-3-甲基咪唑离子液体合成反应进程

采用紫外吸收光谱结合偏最小二乘回归建立一种同时快速测定氯化1-羟乙基-3-甲基咪唑离子液体合成反应体系中N-甲基咪唑(IM1)和1-羟乙基-3-甲基咪唑氯化物(IM2)的方法。配制25个样本作为校正集,采用偏最小二乘回归方法和留一交互验证法分别建立IM1和IM2的回归模型,并采用5个样本作为独立验证集评价模型的预测性能。校正模型用于预测氯化1-羟乙基-3-甲基咪唑离子液体合成体系中IM1和IM2含量。结果表明紫外光谱法结合偏最小二乘回归可实现氯化1-羟乙基-3-甲基咪唑离子液体合成过程的近实时监测。     

Copolymerization of ethylene and propylene catalyzed by novel magnesium chloride supported,vanadium‐based catalysts

Two novel magnesium chloride supported, vanadium‐based Ziegler–Natta catalysts with 9,9‐bis(methoxymethyl)fluorene and di‐i‐butyl phthalate as internal donors were prepared and used in the copolymerization of ethylene and propylene. The catalytic behaviors of these catalysts were investigated and compared with those of traditional magnesium chloride supported, vanadium‐based catalysts without internal donors. Differential scanning calorimetry, gel permeation chromatography, and ¹³C‐NMR spectroscopy analysis were performed to characterize the melting temperatures, molecular weights, and molecular weight distributions as well as structures and compositions of the products. The copolymerization kinetic results indicated that the novel catalyst with 9,9‐bis(methoxymethyl)fluorene as an internal donor had the highest catalytic activity and optimal kinetic behavior in ethylene–propylene copolymerization with an ethylene/propylene molar ratio of 44/56. Low‐crystallinity and high‐molecular‐weight copolymers were obtained with these novel magnesium chloride supported, vanadium‐based catalysts. The reactivity ratio data indicated that the catalytic systems had a tendency to produce random ethylene–propylene copolymers. © 2008 Wiley Periodicals, Inc. J Appl Polym Sci, 2009     

Multivariate determination of cloud point in palm oil using partial least squares and principal component regression based on FTIR spectroscopy

A rapid FTIR spectroscopic method was developed for quantitative determination of the cloud point (CP) in palm oil samples. Calibration samples were prepared by blending randomized amounts of palm olein and palm stearin to produce a wide range of CP values ranging between 8.3 and 47.9°C. Both partial least squares (PLS) and principal component regression (PCR) calibration models for predicting CP were developed by using the FTIR spectral regions from 3000 to 2800 and 1800 to 1600 cm⁻¹. The prediction capabilities of these calibration models were evaluated by comparing their standard errors of prediction (SEP) in an independent prediction set consisting of 14 palm oil samples. The optimal model based on PLS in the spectral range 1800-1600 cm⁻¹ produced lower SEP values (2.03°C) than those found with the PCR (2.31°C) method. FTIR in conjunction with PLS and PCR models was found to be a useful analytical tool for simple and rapid quantitative determination of CP in palm oil.     

The functions of crystallizable ethylene‐propylene copolymers in the formation of multiple phase morphology of high impact polypropylene

The functions of crystallizable ethylene‐propylene copolymers in the formation of multiple phase morphology of high impact polypropylene (hiPP) were studied by solvent extraction fractionation, transmission electron microscopy (TEM), selected area electron diffraction (SAED), nuclear magnetic resonance (¹³C‐NMR), and selected reblending of different fractions of hiPP. The results indicate that hiPP contains, in addition to polypropylene (PP) and amorphous ethylene‐propylene random copolymer (EPR) as well as a small amount of polyethylene (PE), a series of crystallizable ethylene‐propylene copolymers. The crystallizable ethylene‐propylene copolymers can be further divided into ethylene‐propylene segmented copolymer (PE‐s‐PP) with a short sequence length of PE and PP segments, and ethylene‐propylene block copolymer (PE‐b‐PP) with a long sequence length of PE and PP blocks. PE‐s‐PP and PE‐b‐PP participate differently in the formation of multilayered core‐shell structure of the dispersed phase in hiPP. PE‐s‐PP (like PE) constructs inner core, PE‐b‐PP forms outer shell, while intermediate layer is resulted from EPR. The main reason of the different functions of the crystallizable ethylene‐propylene copolymers is due to their different compatibility with the PP matrix. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

基于智能优化算法的软测量模型建模样本优选及应用

根据目标工况合理选择训练样本,是建立软测量模型的关键。传统的训练集样本选择方法难以充分利用因变量信息,而且难以综合考虑样本对模型的影响。为了解决上述问题,本文提出一种基于智能优化算法的训练集样本选择模型,定义了损失函数和样本压缩率,通过权重因子将二者融合为多目标适应度函数,可调整优化算法的寻优方向,使算法能够同时对建模样本组合结构与样本数量寻优,因此极大提高了所选建模样本的质量。为了验证方法的有效性,以汽油调和过程中采集的汽油近红外光谱-研究法辛烷值数据以及柴油近红外基准数据为例,与偏最小二乘、局部权重偏最小二乘等多种方法进行了比较研究,并分析了建模样本对软测量模型的影响。结果表明,本文方法在大规模降低训练集样本规模的同时能够保证软测量模型的精度和泛化性,非常适合工业应用。     

Photodegradation of ethylene/propylene/polar monomers,co‐, and terpolymers. II. Prepared by Ni catalyst systems

The copolymers of ethylene/propylene as well as their terpolymers with polar monomers were prepared by Ni‐catalyst systems and their photodegradation behavior was studied by Fourier transform infrared spectroscopy. The polar monomers used to synthesize co‐ and terpolymers of ethylene/propylene/polar monomer were 5‐hexen‐1‐ol, 10‐undecen‐1‐ol, acrylamide, methylmethacrylate, acrylonitrile, and methylvinyl ketone. The morphological changes of the irradiated samples were determined by scanning electron microscopy. The photodegradation kinetics has also been studied. The surface damage caused by polychromatic irradiation (λ ≥ 290 nm) at 55 °C in atmospheric air is presented in different micrographs. The rate of photo‐oxidative degradation is very fast in terpolymers containing polar monomers as compared with copolymers and homopolymers. The morphological study of the photodegraded samples showed a very good correlation with the photodegraded results. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 104: 1783–1791, 2007     

Model-based control of a molten carbonate fuel cell (MCFC) process

To improve availability and performance of fuel cells, the operating temperature of molten carbonate fuel cells (MCFC) stack should be strictly maintained within a specified operation range, and an efficient control technique should be employed to meet this objective. While most modern control strategies are based on process models, many existing models of MCFC are not ready to be applied in synthesis and operation of control systems. In this study, we developed an auto-regressive moving average (ARMA) model and machine learning methods of least squares support vector machine (LS-SVM), artificial neural network (ANN) and partial least squares (PLS) for the MCFC system based on input-output operating data. The ARMA model showed the best tracking performance. A model predictive control method for the operation of MCFC system was developed based on the proposed ARMA model. The control performance of the proposed MPC methods was compared with that of conventional controllers using numerical simulations performed on various process models including an MCFC process. Numerical results show that ARMA model based control provides improved control performance compared to other control methods.     

Influence of the Form II/Form I crystal polymorphism of random copolymers of butene‐1 with ethylene or propylene on the peel behavior of peel films

Peel films of blends of low density polyethylene (LDPE) and random isotactic copolymers of butene‐1 with either ethylene (iPB‐Eth) or propylene (iPB‐Prop) were investigated regarding the effect of the copolymer composition on both the Form II mesophase to Form I crystal transformation of the copolymers, and the time‐dependent peel behavior of their blends with LDPE in peel films. In general, there is observed a decrease of the peel force with increasing concentration of both ethylene and propylene co‐units in random iPB‐1 based copolymers and their blends with LDPE, after completion of the Form II to Form I transformation. Thus, to tailor the peel force, either the content of the peel component in the blends, or the concentration of ethylene or propylene co‐units in the peel component may be varied. The effect of ethylene co‐units in the random copolymers on the peel force is distinctly larger than that of propylene co‐units. Parallel to the Form II to Form I transition of butene‐1 based copolymers, the peel force decreases with a rate which depends on the copolymer composition. The Form II to Form I transition in iPB‐Prop copolymers proceeds distinctly faster than in iPB‐Eth copolymers of identical concentration of co‐units. POLYM. ENG. SCI., 55:1749–1757, 2015. © 2014 Society of Plastics Engineers     

Determination of Mechanical and Optical Properties of Eucalyptus Kraft Pulp by NIR Spectrometry and Multivariate Calibration

Abstract

Multivariate data analysis and NIR spectrometry were used to determine the mechanical and optical properties of eucalyptus kraft pulps with different chemical compositions and refined to different levels. Tear (TrID), tensile (TsID), burst (BuID), and bending (BeID) indexes and elastic modulus (EM), stretch (ST), and breaking length (BL) were the mechanical properties measured. Measurement of beating degree (SR) was also achieved. Light scattering (LS) and light absorption (LA) coefficients were the optical properties measured. Mechanical and optical properties were modeled using NIR spectra obtained on pulp hand sheets by diffuse reflectance and application of the partial least squares (PLS) method. Models with two to seven PLS components and very good predictive ability were established after testing the first‐derivative, Kubelka‐Munk, or a combination of both as pre‐processing techniques. Models were validated by using cross‐validation methodology and a comparison of measurements using conventional methods for new samples. The predictive models can reduce time in traditional measurements in the pulp and paper industry and are suitable for direct application using “at‐line” conditions. As an additional benefit, improvements in process monitoring and paper quality can be achieved.     

Multilayered core–shell structure of the dispersed phase in high‐impact polypropylene

The multiphase morphology of high impact polypropylene (hiPP), which is a reactor blend of polypropylene (PP) with ethylene–propylene copolymer, was investigated by transmission electron microscopy, selected area electron diffraction, atomic force microscopy, and field‐emission scanning electron microscopy techniques in conjunction with an analysis of the hiPP composition and chain structure based on solvent fractionation, ¹³C‐NMR, and differential scanning calorimetry measurements. A multilayered core–shell structure of the dispersed phase of hiPP in solution‐cast films and the bulk was observed. The inner core was mainly composed of polyethylene (including its long blocks) together with part of PP, the intermediate layer was ethylene–propylene random copolymer, and the outer shell consisted of ethylene–propylene multiblock copolymers. The formation process and controlling factors of the multilayered core–shell structure are discussed. This kind of multiphase morphology of hiPP caused the material to possess both a high rigidity and high toughness. © 2008 Wiley Periodicals, Inc. J Appl Polym Sci, 2008     

Study of crystallization and melting behavior of polypropylene‐block‐polyethylenes copolymers fractionated from polypropylene and polyethylene mixtures

A series of ethylene–propylene block copolymer fractions of differing compositions, while still retaining broad molecular weight distributions, were obtained by fractionation of polypropylene (PP) and polyethylene (PE) copolymers prepared by sequential polymerization of ethylene and propylene. The crystallization and melting behavior of the polypropylene‐block‐polyethylene fractions were studied. It was observed that the major component could suppress crystallization of the minor component, leading to a decrease in crystallinity and melting temperature. Non‐isothermal crystallization showed that crystallization of the ethylene block was less influenced by composition and cooling rate than the propylene block. At fast cooling rates, the ethylene block could crystallize prior to the propylene block. Isothermal crystallization kinetics experiments were also conducted. We found that the block copolymers with minor ethylene components had smaller Avrami exponents (n ≈ 1.0), hence indicating a reduced growth dimension of the PE crystals by the pre‐existing PP crystals. On the other hand, the ethylene block exhibited much larger Avrami exponents in those block copolymers with major ethylene contents. Copyright © 2004 Society of Chemical Industry     

Offset‐Free Inferential Feedback Control of Distillation Compositions Based on PCR and PLS Models

An offset‐free inferential feedback control strategy for distillation composition control using principal component regression (PCR) and partial least squares (PLS) models is presented in this paper. PCR and PLS model based software sensors are developed from process operational data so that the top and bottom product compositions can be estimated from multiple tray temperature measurements. The PCR and PLS software sensors are then used in the feedback control of the top and bottom product compositions. With this strategy the problem of substantial time delay in composition analyzer based control and of substantial bias in single tray temperature control can be overcome. A practically very important issue in software sensor based feedback control is that static control offsets often exist due to a static estimation bias, especially when the process operating condition changes. A technique for eliminating the static estimation bias and the resulting static control offsets through mean updating of process measurements is proposed in this paper. Applications to a simulated methanol‐water separation column demonstrate the effectiveness of this control strategy.     

Measurement of soybean fatty acids by near-infrared spectroscopy: Linear and nonlinear calibration methods

A key element of successful development of new soybean cultivars is availability of inexpensive and rapid methods for measurement of FA in seeds. Published research demonstrated applicability of NIR spectroscopy for FA profiling in oilseeds. The objectives of this study were to investigate the applicability of NIR spectroscopy for measurement of FA in whole soybeans and compare performance of calibration methods. Equations were developed using partial least squares (PLS), artificial neural networks (ANN), and support vector machines (SVM) regression methods. Validation results demonstrated that (i) equations for total saturates had the highest predictive ability (r ²=0.91–0.94) and were usable for quality assurance applications, (ii) palmitic acid models (r ²=0.80–0.84) were usable for certain research applications, and (iii) equations for stearic (r ²=0.49–0.68), oleic (r ²=0.76–0.81), linoleic (r ²=0.73–0.76), and linolenic (r ²=0.67–0.74) acids could be used for sample screening. The SVM models produced significantly more accurate predictions than those developed with PLS. ANN calibrations were not different from the other two methods. Reduction in the number of calibration samples reduced predictive ability of all equations. The rate of performance degradation of SVM models with sample reduction was the lowest.     

In‐line monitoring of hydrogen peroxide in two‐phase reactions using raman spectroscopy

Hydrogen peroxide is an environment‐friendly oxidizer, which is used in several chemical processes. However, safety necessitates the determination and control of the concentration of hydrogen peroxide during oxidation reactions. We propose a methodology to monitor hydrogen peroxide in disperse two‐phase reaction mixtures based on in‐line Raman spectroscopy. We compare indirect hard modeling (IHM), peak integration (PI), and partial least squares (PLS). Building predictive PLS and PI calibration models is challenging, whereas the IHM calibration is easy to develop. These methods show good accuracy for known samples (root mean square error of cross validation [RMSECV] of 0.3–0.7 wt %) compared to the classic titration method (RMSECV of 0.4 wt %). After calibration, inline monitoring during reaction is performed demonstrating that the concentration of hydrogen peroxide can be successfully monitored in a fast and reliable way by Raman spectroscopy. The IHM seems to give slightly better inline predictions. © 2017 American Institute of Chemical Engineers AIChE J, 63: 3994–4002, 2017     

Local dynamic partial least squares approaches for the modelling of batch processes

The application of multivariate statistical projection based techniques has been recognized as one approach to contributing to an increased understanding of process behaviour. The key methodologies have included multi‐way principal component analysis (PCA), multi‐way partial least squares (PLS) and batch observation level analysis. Batch processes typically exhibit nonlinear, time variant behaviour and these characteristics challenge the aforementioned techniques. To address these challenges, dynamic PLS has been proposed to capture the process dynamics. Likewise approaches to removing the process nonlinearities have included the removal of the mean trajectory and the application of nonlinear PLS. An alternative approach is described whereby the batch trajectories are sub‐divided into operating regions with a linear/linear dynamic model being fitted to each region. These individual models are spliced together to provide an overall nonlinear global model. Such a structure provides the potential for an alternative approach to batch process performance monitoring. In the paper a number of techniques are considered for developing the local model, including multi‐way PLS and dynamic multi‐way PLS. Utilising the most promising set of results from a simulation study of a batch process, the local model comprising individual linear dynamic PLS models was benchmarked against global nonlinear dynamic PLS using data from an industrial batch fermentation process. In conclusion the results for the local operating region techniques were comparable to the global model in terms of the residual sum of squares but for the global model structure was evident in the residuals. Consequently, the local modelling approach is statistically more robust.