Masahiro Watari

On-Line Monitoring of the Density of Linear Low-Density Polyethylene in a Real Plant by Near-Infrared Spectroscopy and Chemometrics

This paper reports on-line monitoring of the density of linear low-density polyethylene (LLDPE) by near-infrared (NIR) spectroscopy and chemometrics. The on-line monitoring was carried out not only in a laboratory but also in a real plant. We composed an on-line monitoring system for molten polymers consisting of a Fourier transform near-infrared (FT-NIR) spectrometer, input/output (I/O) module, a personal computer, and a sampling cell that we developed. We first compared NIR spectra of LLDPE in the solid and melt states and then developed calibration models that predict the density using partial least squares regression (PLS). The sample sets for developing prediction models were collected for three months at the plant, and the density of LLDPE was continuously monitored on-line for another three months using the model. The standard error of prediction (SEP) for the on-line monitoring of the density of LLDPE at the plant was ±2.1 mg/cm3 (range: 0.91–0.95 g/cm3).

Fourier Transform Raman Spectra of Linear Low-Density Polyethylene and Prediction of Their Density by Multivariate Data Analysis

Fourier transform Raman spectra have been measured for pellets of sixteen kinds of linear low-density polyethylene (LLDPE) with short branches and one kind of PE without any branch. Before we tried chemometrics analysis, the Raman spectra of LLDPE were investigated by comparing them with the spectrum of PE in order to explore the effects of the branches on the Raman spectra. Partial least-squares (PLS) regression was applied to the Raman spectra in the 1600–600 cm−1 region after multiplicative scatter correction (MSC) to propose a calibration model that predicts the density of LLDPE. The correlation coefficient was calculated to be 0.968, and the root mean square error of cross validation (RMSECV) was found to be 0.0018 g/cm3. The loadings plot of regression coefficients for the calibration model shows that, not only a sharp upward peak at 1417 cm−1 corresponding to the CH2 bending mode reflecting the crystallinity, but also a broad downward peak near 1308 cm−1 corresponding to the amorphous board band of LLDPE plays a key role in the prediction of their density. The chemometrics study has deepened the analysis of the Raman spectra of LLDPE. For example, the detailed analysis of the principal component weight loadings plots has elucidated the existence of bands due to the CH3 groups of branches and those arising from amorphous parts of LLDPE that are almost missing or hidden by other intense bands. In other words, the chemometrics analysis has enhanced spectral resolution.

Monitoring the melt-extrusion transesterification of ethylene-vinylacetate copolymer by self-modeling curve resolution analysis of on-line near-infrared spectra.

The transesterification of molten ethylene-vinylacetate (EVA) copolymers by octanol with sodium methoxide as catalyst in an extruder has been monitored by on-line near infrared (NIR) spectroscopy. A total of 60 NIR spectra were acquired for 37 min with the last spectrum recorded 31 min after the addition of octanol and catalyst was stopped. The experimental spectra show strong baseline fluctuations which are corrected for by multiplicative scatter correction (MSC). The chemometric methods of the orthogonal projection approach (OPA) and multivariate curve resolution (MCR) were used to resolve the spectra and to derive concentration profiles of the species. The detailed analysis reveals the absence of completely pure variables which leads to small errors in the calculation of pure spectra. The initial estimation of a concentration that is necessary as an input parameter for MCR also presents a non-trivial task. We obtained results that were not ideal but applicable for practical concentration control. They enable a fast monitoring of the process in real-time and resolve the spectra of the EVA copolymer and the ethylene-vinyl alcohol (EVAL) copolymer to be very close to the reference spectra. The chemometric methods used and the decomposed spectra are discussed in detail.

Near infrared spectroscopy and chemometrics analysis of linear low-density polyethylene

Near infrared (NIR) diffuse reflectance spectra have been measured using a rotating drawer for pellets of 16 kinds of linear low-density polyethylene (LLDPE) with short branches and PE without any branches to propose a calibration model which predicts their density and to increase the understanding of NIR spectra of LLDPE. The density of the LLDPE samples investigated was in the range 0.911–0.925 g cm−3. Partial least squares (PLS) regression has been applied to the original NIR spectra data set, their second derivatives and the spectra after multiplicative scatter correction (MSC) treatment to make up the models. The correlation coefficient was calculated to be 0.961, 0.965 and 0.970 for the original NIR spectra, their second derivatives and those with the MSC treatment, respectively, and the standard error of prediction (SEP) was found to be 0.001 g cm−3 for all the cases. The regression coefficients plot for the calibration models shows that bands at 1192, 1376 and 1698 nm due to the overtone and combination modes of the CH3 groups play important roles in the prediction of density.

Prediction of Ethylene Content in Melt-State Random and Block Polypropylene by Near-Infrared Spectroscopy and Chemometrics: Comparison of a New Calibration Transfer Method with a Slope/Bias Correction Method

This paper reports the prediction of the ethylene content (C2 content) in random polypropylene (RPP) and block polypropylene (BPP) in the melt state by near-infrared (NIR) spectroscopy and chemometrics. NIR spectra of RPP and BPP in the melt states were measured by a Fourier transform near-infrared (FT-NIR) on-line monitoring system. The NIR spectra of RPP and BPP were compared. Partial least-squares (PLS) regression calibration models predicting the ethylene (C2) content that were developed by using each RPP or BPP spectra set separately yielded good results (SECV (standard error of cross validation): RPP, 0.16%; BPP, 0.31%; correlation coefficient: RPP, 0.998; BPP, 0.996). We also built a common PLS calibration model by using both the RPP and the BPP spectra set. The results showed that the common calibration model has larger SECV values than the models based on the RPP or the BPP spectra sets individually and is not practical for the prediction of the C2 content. We further investigated whether a calibration model developed by using the BPP spectra set can predict the C2 contents in the RPP sample set. If this is possible, it can save a significant amount of work and cost. The results showed that the use of the BPP model for the RPP sample set is difficult, and vice versa, because there are some differences in the molar absorption coefficients between the RPP and BPP spectra. To solve this problem, a transfer method from one sample spectra (BPP) set to the other spectra (RPP) set was studied. A difference spectrum between an RPP spectrum and a BPP spectrum was used to transfer from the BPP calibration set to the RPP calibration set. The prediction result (SEP (standard error of prediction), 0.23%, correlation coefficient, 0.994) of RPP samples by the transferred calibration set and model showed that it is possible to transfer from the BPP calibration set to the RPP calibration set. We also studied the transfer from the RPP calibration set (the range of C2 content: 0–4.3%) to the BPP calibration set. The prediction result of C2 content (the range of C2 contents: 0–7.7%) in BPP by use of the calibration model based on the transferred BPP spectra from the RPP spectra showed that the transfer method is only effective for the interpolation of the C2 content range by the nonlinear change in the peak intensities with the C2 content.

Practical Calibration Correction Method for the Maintenance of an On-Line Near-Infrared Monitoring System for Molten Polymers

The present study has investigated a practical calibration correction method for an on-line monitoring system for molten polymers using a near-infrared (NIR) spectrometer. A partial least squares (PLS) calibration model for the ethylene (C2) content in melt polypropylene (PP) was developed for the investigation of changes in the performance of the on-line system before and after maintenance necessitated by the relocation. The predicted values for the C2 content from the spectra measured after maintenance by using the calibration model developed from the spectra collected before maintenance showed that there were some differences between the spectra obtained by the NIR spectrometer system before and after maintenance. The loadings from factor analysis suggested that the main cause for the differences in the system performance before and after maintenance was wavenumber shifts in the NIR spectra of PP in the melt state. Six popular standardization or calibration transfer methods (direct standardization (DS), piecewise direct standardization (PDS), additive correction (AD), multiplicative correction (MP), slope and bias (SB), and difference spectrum with interpolation (DSI)) were evaluated for the calibration correction of the on-line NIR monitoring system. However, the results of the evaluation showed that these standardization methods need more than two samples to obtain the high accuracy for the nonlinearity contained in the spectra set. From the standpoint of practical calibration in a real plant, the acceptable number of samples for the calibration is one or two. Moreover, recalibration using transferred spectra is not preferable because of the traceability for a calibration model. As a practical solution for a calibration correction in a real plant, a method considering wavenumber shift and path-length correction has been proposed in this study. The predicted results for the C2 content in the melt-state PP from the spectra measured after maintenance by using the proposed method have shown that the proposed method is useful for calibration correction in a real plant in spite of using only one sample.

Prediction of Ethylene Content in Melt-State Random and Block Polypropylene by Near-Infrared Spectroscopy and Chemometrics: Influence of a Change in Sample Temperature and its Compensation Method

This paper reports on the influence of a change in sample temperature, and a method for its compensation, for the prediction of ethylene (C2) content in melt-state random polypropylene (RPP) and block polypropylene (BPP) by near-infrared (NIR) spectroscopy and chemometrics. Near-infrared (NIR) spectra of RPP in the melt and solid states were measured by a Fourier transform near-infrared (FT-NIR) on-line monitoring system and an FT-NIR laboratory system. There are some significant differences between the solid- and melt-state RPP spectra. Moreover, we investigated the predicted values of the C2 content from the RPP or BPP spectra measured at 190 °C and 250 °C using the calibration model for the C2 content developed using the RPP or BPP spectra measured at 230 °C. The errors in the predicted values of the C2 content depend on the pretreatment methods for each calibration model. It was found that multiplicative signal correction (MSC) is very effective in compensating for the influence of the change of temperature for the RPP or BPP samples on the predicted C2 content. From the suggestion of principal component analysis (PCA) and difference spectrum analysis, we propose a new compensation method for the temperature change that uses the difference spectra between two spectra sets measured at different temperatures. We achieved good results using the difference spectra between the RPP/BPP spectra sets measured at 190 °C and 250 °C after correction and the calibration model developed with the spectra measured at 230 °C. The comparison between the method using MSC and the proposed method showed that the predicted error in the latter is slightly better than those in the former.

On-line monitoring of melt-extrusion transesterification of ethylene vinylacetate copolymers by near infrared spectroscopy and chemometrics

The melt-extrusion transesterification of ethylene/vinylacetate (EVA) copolymer to ethylene/vinylalcohol (EVAL) copolymers has been monitored by on-line near infrared (NIR) spectroscopy. A total of 60 NIR spectra were measured within 37 minutes after the initial addition of octanol (reagent) and catalyst (sodium methoxide) at the exit of the extruder by use of a fibre-optic probe. The most significant intensity change is observed for a band at 7089 cm−1 due to the first overtone of an OH stretching mode of the EVAL copolymers. We can monitor the progress of the reaction by plotting the peak intensity at 7089 cm−1 only. A principal component analysis (PCA) was carried out for the series of NIR spectra in the 7300–6900 cm−1 region. A score plot of PCA factor 1 is almost identical with the plot of the peak intensity at 7089 cm−1. Calibration models for predicting the vinyl acetate content in EVA copolymers have been developed by use of partial least squares (PLS) regression. The correlation coefficient and standard error of prediction are 0.96 and 0.85%, respectively, indicating that the described technique can be used to monitor the transesterification reaction.

Two-Dimensional Heterospectral Correlation Analysis of Water and Liquid Oleic Acid Using an Online Near-Infrared/Mid-Infrared Dual-Region Spectrometer:

Two-dimensional (2D) near-infrared (NIR) and mid-infrared (mid-IR) heterospectral correlation analyses were used to characterize temperature-dependent spectral variations of water and liquid oleic acid (OA), utilizing a dataset obtained with an online NIR/mid-IR dual-region spectrometer. The spectrometer facilitated sequential acquisition of both NIR (10000–4000 cm−1) and mid-IR (5000–1200 cm−1) spectra, which compose the spectral dataset required for 2D NIR/mid-IR heterospectral correlation analysis. Both NIR and mid-IR spectra were obtained under the same conditions by using the same sample compartment, more quickly and easily than is possible when using existing spectrometers. Successful 2D NIR/mid-IR correlation analysis was performed with the data collected with this instrument to characterize the temperature dependence of the molecular structures of water and pure liquid OA. Temperature-induced NIR/mid-IR spectral changes for water and OA were analyzed in detail, and band assignments in the NIR and mid-IR regions were elucidated by 2D NIR/mid-IR heterospectral correlation analysis. The results of this study indicate that liquid water consists of two major species, strongly hydrogen-bonded species and weakly hydrogen-bonded species, as well as one minor species. Additionally, OA was found to form an intermolecularly hydrogen-bonded species in which a single hydrogen bond of the dimer was broken; a mid-IR band at 1724 cm−1 was assigned to this species. Moreover, 2D NIR/mid-IR heterospectral correlation analysis revealed that NIR bands at 4690 and 4644 cm−1 also arose from intermolecularly hydrogen-bonded species. These results demonstrate that 2D NIR/mid-IR heterospectral correlation analysis is useful not only for NIR band assignments, but also for molecular structure studies. The spectrometer we developed makes this analysis even more accessible.

Development of a near-infrared/mid-infrared dual-region spectrometer for online process analysis.

A near-infrared (NIR) and mid-infrared (mid-IR) dual-region spectrometer having two immersion probes, a transmission probe for NIR, and an attenuated total reflection (ATR) probe for mid-IR has been developed for highly reliable process monitoring and deep process understanding. This spectrometer facilitates sequential acquisition of both NIR (10 000–4000 cm−1) and mid-IR (5000–1200 cm−1) spectra by switching the light path leading to the probes without the need for probe replacement. The use of a single light source and a single beam splitter enables achievement of a permanent alignment of the optical system and sequential data acquisition. The transmission NIR and ATR mid-IR probes designed and developed in the present study facilitate the acquisition of NIR/mid-IR spectra with optimized absorption intensities in both regions by simply placing the probes into a sample solution. The performance of the developed spectrometer was demonstrated in monitoring the ethanol fermentation process. NIR/mid-IR spectra of the fermentation solution with multiplicative scatter correction (MSC) represent the relative changes in the concentrations of glucose and ethanol in both regions. Principal component analysis (PCA) was performed on the MSC-treated spectra in the regions 6300–5650 cm−1, 4850–4300 cm−1, and 3500–2880 cm−1 to detect the end-point of the fermentation as an example of process monitoring. For all the regions, the score plot of the first principal component (PC) indicates that the fermentation progresses with the fermentation time and stops after 210 minutes and thus the end-point of the fermentation exists at around 210 minutes. The loading plot indicates that all of the first PCs are the relative changes in the concentrations of glucose and ethanol. This result reveals that the same chemical changes are observed in both transmission NIR and ATR mid-IR spectra. Multiple and simultaneous analysis was also performed, and intensity change in light scattering relating the growth of yeasts was monitored by the NIR spectra.