Characterization of binary/tertiary blends of SBR, NBR and PVC by IR spectroscopy.
Acrylonitrile-butadiene rubber (NBR) and poly(vinyl) chloride (PVC) are commonly blended to produce a thermoplastic material that provides improved thermal aging and resistance to ozone, solvents and tearing. These improvements of physical properties, along with cost reductions, have resulted in the use of NBR/PVC blends in a wide range of applications including hoses, gaskets, shoe soles, sealants and coatings in the electrical industry. In applications where excellent solvent resistance is not crucial, it is often desirable to replace a portion of the NBR with emulsion styrene butadiene rubber (E-SBR). E-SBR is cheaper and has better flexibility; unfortunately, NBR has limited compatibility with non-polar polymers such as SBR, polybutadiene (BR) and natural rubber (NR). However, the low acrylonitrile NBR grades can be blended with SBR over the full range of concentrations without significant deterioration of mechanical vulcanizate properties. In fact, a number of these blends are used in several critical ,applications. NBR/SBR blends are used to compensate the volume decrease in oil seal applications (ref. 1). In order to avoid poor morphology, a third elastomer is often added to act as a compatibilizer and to achieve a miscible blend. In the case of SBR/PVC blends, NBR is often used as the compatibilizer because the nonpolar polybutadiene portion of the NBR is miscible with the SBR phase, and the polar nature of NBR is compatible with the PVC phase. It has been shown that up to 30% of the NBR in NBR/PVC blends can be replaced by SBR without a significant decrease in mechanical properties (refs. 2 and 3).
The physical and mechanical properties of the blends of SBR, NBR and PVC are sensitive to small variations in the amounts of the individual polymers used. Therefore, there is a need for developing a variety of analytical tools to monitor blend compositions. Pyrolysis techniques have been employed extensively to characterize elastomers in the presence of nontransparent fillers. FTIR spectroscopy is also often used to characterize the composition of elastoblends. The ASTM D3677 (ref. 4) method for rubber identification by FTIR requires a lengthy sample preparation that involves an extraction step followed by pyrolysis. Frisone and coworkers (ref. 5) reported a multi-laboratory study of the qualitative and quantitative analysis of polymer blends using several techniques, including FTIR, which employed a modified ASTM method. Parker and Waddell (ref. 6) utilized photoacoustic FTIR to determine the acrylonitrile content in NBR and the styrene content of SBR. The photoacoustic FTIR technique allows a background spectrum of carbon black or fillers to be run prior to analyzing the compounded samples.
In this study, the feasibility of using mid-IR (MIR) and near-IR (NIR) spectroscopy to estimate polymer compositions of binary and tertiary blends of SBR, NBR and PVC was investigated. In the MIR method, the amount of incident radiation absorbed by a thin film is measured. The film is prepared using a hot press, and the resulting film stretched into a sample holder. As can be easily shown from Beer's law, utilizing the absorbance ratio of two characteristic peaks of an elastomer makes the pathlength less critical, and quantitative determinations become easy to carry on. The attenuated total internal reflectance FTIR (ATR-FTIR) and NIP, techniques require no sample preparation and monitor reflected light. The elastomer composition of SBR/NBR blends was obtained from a plot of the absorbance (MIR) ratio of a characteristic elastomer peak relative to the C=C stretching vibration of polybutadiene vs. the amount of each component in the blend. In contrast, ATR-FTIR allows for direct analysis of a gum rubber (rubber with no fillers or compounding ingredients) without extensive sample preparation steps. The use of an ATR reflective element eliminates scattering by the sample and maintains a constant pathlength. The pathlength, or depth of penetration, is dependent on the incident angle between the source and the reflective element.
SBR (23.5% styrene) and NBR (23.5% acrylonitrile) were dry-mixed at 60[degrees]C for eight minutes in a Brabender with 0.25 phr antioxidant added. Samples were prepared at 100/0, 80/20, 60/40, 40/60, 20/80 and 0/100 SBR/NBR levels. Tertiary blends of SBR, PVC and NBR were dry-mixed at 160[degrees]C for five minutes in a Brabender at levels of 10/30/60, 20/ 30/50, 30/30/40 and 50/30/20.
A Bruker Vector 33N FTIR/NIR equipped with a single bounce, horizontal ATR element (ZnSe crystal) and a NIR fiber optic probe was used to collect multiple (16) IR spectra, which were then signal averaged. The analysis time for the MIR and NIR determinations was [less than or equal to] 5 minutes. The sample preparation and specific experimental details are described for each technique:
* Mid-IR spectroscopy--thin films for MIR analysis were obtained by pressing approximately 0.2 g of sample for three minutes in a press held at 126[degrees]C and 35,000 psi. The film was then mounted onto a disposable adhesive IR card. Blends characterized by ATR were used without sample preparation and were placed directly on the ZnSe crystal. Infrared spectra of the samples were collected from 400-4,000 [cm.sup.-1] with a resolution of 2 [cm.sup.-1]. The resulting spectra were normalized, by setting the absorbance at 2,100 [cm.sup.-1] (inactive region) to zero, and the absorbance was measured at specific wavenumbers.
* Near-IR Spectroscopy--NIR spectra were collected from 4,000-12,000 [cm.sup.-1] with 8 [cm.sup.-1] resolution using a fiber optic probe. The resulting spectra were pre-processed using a vector normalization method. Three measurements of each standard were used to develop a NIR calibration, and a partial-least-squares (PLS) algorithm was used to cross-validate the model.
Results and discussion
The MIR region of FTIR spectroscopy can be very useful in qualitatively identifying individual elastomers in polyblends. Each elastomer usually has one or more unique absorption band that, if observed, may signify the presence of that particular polymer in the blend. Once the composition of the polyblend has been qualitatively determined, calibration curves can be established to quantitatively determine the polymer blend ratio of an unknown composition. Since NIR provides information attributable to overtone and combination bands of the fundamental MIR stretching vibrational bands, it is more difficult to qualitatively identify blends: however, if the blend components are well known, a calibration model can be quickly and easily developed to quantitatively determine the blend compositions. In this study, binary mad tertiary blends of SBR, NBR and PVC were characterized using MIR and NIR spectroscopy. Methodologies for determining the SBR content of SBR/NBR and SBR/PVC/NBR blends, as well as the accuracy of the technique are presented.
The frequencies measured in the blends of SBR, NBR and PVC are 1,602, 1,639, 2,237 and 1,254 [cm.sup.-1]. The absorbance at 1,602 [cm.sup.-1] is due to the carbon-carbon stretching vibration of the aromatic double bond (figure 1), and is assigned to the styrene incorporated into the elastomer. The absorbance at 1,639 [cm.sup.-1] is due to the stretching vibration of the olefinic portion of the rubber (figures 1 and 2), which is attributable to the butadiene in the elastomer.
[FIGURE 1 OMITTED]
The absorbance at 2,237 [cm.sup.-1] is a weak stretching vibration due to C=N (figure 2), which is attributable to acrylonitrile.
[FIGURE 2 OMITTED]
The absorbance at 1254 [cm.sup.-1] is due to a strong C[H.sub.2] wagging observed for C[H.sub.2]Cl (figure 3), which is attributable to PVC.
[FIGURE 3 OMITTED]
Determination of SBR content in SBR/NBR blends
The FTIR spectra of SBR/NBR blends clearly showed peaks that were characteristic of each polymer, providing for an easy qualitative assessment. Figure 4 shows the FTIR spectrum of a 60/40 SBR/NBR blend, where the characteristic peaks are labeled as: styrene (1,602 [cm.sup.-1]), butadiene (1,639 [cm.sup.-1]) and acrylonitrile (2,237 [cm.sup.-1]).
[FIGURE 4 OMITTED]
The SBR content in the SBR/NBR blends was determined from the ratio of peak heights of the absorbance peaks at 1,602 and 1,639 [cm.sup.-1], which correspond to styrene and butadiene respectively. A plot of the absorbance ratio (1,602 [cm.sup.-1]/1639 [cm.sup.-1]) vs. the % SBR (figure 5) yielded a straight line with [R.sup.2] = 0.9986. The NBR content was determined from the ratio of peak heights for the absorbance peaks at 2,237 [cm.sup.-1] and 1,639 [cm.sup.-1], which correspond to acrylonitrile and polybutadiene. The absorbance ratio of acrylnnitrile/butadiene (2,237 [cm.sup.-1]/ 1,639 [cm.sup.-1]) vs. % NBR was also plotted (shown vs. % SBR in figure 5) and yielded a straight line with [R.sup.2] = 0.9997. The SBR and NBR content were determined independently from the two calibration curves and normalized to predict the SBR/NBR blend ratio.
[FIGURE 5 OMITTED]
The SBR and NBR content in SBR/NBR blends was determined from the ATR absorbance peaks at 1,602 [cm.sup.-1] for styrene and 2,237 [cm.sup.-1] for acrylonitrile. A plot of absorbance at 1,602 [cm.sup.-1] and 2,237 [cm.sup.-1] vs. the SBR content of the SBR/NBR blends (figure 6) yielded a straight line with [R.sup.2] = 0.97 and [R.sup.2] = 0.997 for styrene and acrylonitrile, respectively. The SBR and NBR content were determined independently from the two ATR calibration curves and normalized to predict the blend ratio.
[FIGURE 6 OMITTED]
The raw NIR spectra of SBR and NBR over the data collection region are shown in figure 7. Since NIR bands are overtones and combinations of fundamental bands, interpretation of the data requires a factor analysis algorithm. A partial-least-squares (PLS) model with cross-validation was employed in the region of 5,443 to 6,103 [cm.sup.-1], following vector normalization preprocessing of the data. Figure 8 shows the correlation between NIR predictions and the calculated percent polymer in the SBR/NBR blends. The root-mean-squared error of cross-validation (RMSECV) and [R.sup.2] were 0.658 and 0.9996 for SBR and NBR, respectively.
[FIGURE 7 OMITTED]
Determination of SBR content in SBR/PVC blends
In this study, MIR was used to characterize SBR/PVC/NBR blends. PVC shows a characteristic peak in the FTIR spectrum at 1,254 [cm.sup.-1], which can be used to qualitatively determine the presence of PVC in the polyblends. Figure 9 shows the FTIR spectra of the 30/30/40 SBR/PVC/NBR blend. A plot of the absorbance peak ratio vs. percent SBR (figure 10) yielded a straight line with [R.sup.2] = 0.978 for styrene and [R.sup.2] = 0.996 for acrylonitrile. A plot of the peak absorbance measured by ATR vs. percent SBR in the blend (figure 11) yielded a straight line with [R.sup.2] = 0.99 and 0.987 for styrene and acrylonitrile, respectively. The presence of PVC in the SBR/NBR blends introduced no interference in the characterization. With appropriate standards, the blend ratio of unknown polyblends could easily be calculated from the absorbance ratio of PVC to polybutadiene or using the peak absorbance and the ATR-FTIR technique presented.
[FIGURES 8-11 OMITTED]
The NIR spectrum of PVC, SBR and NBR is shown in figure 12. The triblends and the individual polymers (SBR, PVC and NBR) were analyzed in triplicate and incorporated into the PLS model. Figure 13 shows the correlation between NIR predictions and the calculated percent polymer in the SBR/ PVC/NBR blends. The [R.sup.2] value for all calibrations was 0.9975 or better ([R.sup.2] and RMSECV values are reported in figure 13). The presence of PVC did not interfere or overlap with the NIR bands of SBR and NBR, and it was easily incorporated into the PLS model.
[FIGURES 12-13 OMITTED]
Determination of technique accuracy
Several test samples of known SBR/NBR blend ratios were analyzed by FTIR, ATR and NIR in order to verify and predict the error associated with the various methods of studying SBR/NBR blends. The SBR and NBR content were determined independently to obtain the blend composition. The predicted SBR/NBR blend ratios deviated from the true polymer blend by [less than or equal to] 6% for the MIR methods. The NIR calibration model predicting blend ratios also showed a maximum deviation of 6% from the calculated blend ratio. The accuracy of the predicted blend ratio obtained by FTIR, ATR and NIR was also compared to results obtained by TGA and pyrolysis-GC/M[S.sup.7] (table 1). The accuracy of the FTIR techniques evaluated and the analysis time were found to be acceptable. The FTIR-predicted blend ratios were comparable to those deter mined by TGA and pyrolysis-GC/MS.
Elastomers are commonly blended to improve the processing, vulcanizate or economic properties of the polymer. The physical, mechanical and dynamic properties of these polyblends arc sensitive to small variations in the amounts of the individual polymers. For this reason, there is a need for developing a variety of analytical tools to monitor blend compositions. In this study, methods for qualitatively and quantitatively characterizing SBWNBR and SBR/PVC/NBR blends have been developed using FTIR spectroscopy employing both the mid-and near-IR regions.
Results of this study show that FTIR spectroscopy is an excellent tool for quickly characterizing the elastomers that make up an elastoblend. By utilizing an absorbance ratio for peaks characteristic of the polymers, the SBR and NBR content were determined. For applications where preparing a film is not practical, ATR-FTIR was used to eliminate the need for an absorbance ratio. The peak absorbance of the characteristic components was monitored as a function of concentration. With an average deviation of less than 4% and analysis time of about five minutes, the MIR techniques presented in this study will be a valuable technique for the polymer industry. A comparative study performed in this work also shows that results obtained by the FTIR methods are also in good agreement with the well-accepted TGA and pyrolysis-GC/MS techniques (ref. 7).
NIR spectroscopy also proved a valuable tool in quickly and accurately determining the polymer ratio of elastoblends. A PLS algorithm was established for appropriate standards, and the model was tested by cross-validation. The average deviation of this technique was found to be within 2%, providing a very fast and accurate method for calculating blend ratios for elastoblends containing SBR, NBR and PVC.
Blends are widely used in the polymer industry, and there is a wide array of methods available for their characterization. FTIR and NIR were proven to be useful tools in studying binary and tertiary blends of SBR, NBR and PVC. Each FTIR technique has unique advantages; however, all methods of FTIR spectroscopy studied (MIR transmission, ATR-FTIR, and NIR) provide a fast and convenient means of accurately characterizing the polymer blends.
This article is based on a paper given at the October; 2001 meeting of the Rubber Division.
Table 1 - predicted SBR/NBR blend ratios for test samples Sample True SBR/NBR FTIR ATR-FTIR 1 89/11 95/5 85/15 2 79/21 77/23 76/24 3 70/30 73/27 66/34 4 67/33 67/33 63/35 5 50/50 45/55 44/56 6 30/70 27/73 26/74 Avg. % deviation: [+ or -] 3 [+ or -] 4 Sample NIR TGA (7) Py-GC/MS (7) 1 86/14 73/27 88/12 2 78/22 67/33 80/20 3 73/27 63/37 72/28 4 67/33 55/45 70/30 5 44/56 41/59 52/48 6 28/72 39/62 29171 Avg. % deviation: [+ or -] 2 [+ or -] 10 [+ or -] 2
(1.) H.H. Bertram, "Developments in Rubber Technology," 2nd ed., A. Whelan and K.S. Lee, Eds., Elsevier Applied Science Publishers, Barking, 1981, p. 51.
(2.) K.E. George, R. Joseph, D.J. Francis and K.T. Thomas, Polym. Eng. Sci., 27, 1,137 (1987).
(3.) J. Zhao, G.N. Ghebremeskel and J. Peasley, Rubber World, 219, 37 (1998).
(4.) Annual Book ASTM Stand. 09.01 (1990).
(5.) G.J. Frisone, D.L. Schwarz and R.A. Ludwigsen, Rubber& Plastics News, 21 (1997).
(6.) J.R. Parker and W.H. Waddell, J. Elastom. Plastics. 28, 140 (1996).
(7.) S.R. Shield, G.N. Ghebremeskel and C. Hendrix, ASC Rubber Division 159th Spring meeting, Providence, RI (2001).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||mid-IR and near-IR spectroscopy to estimate polymer compositions of binary and tertiary blends of SBR, NBR and PVC investigated|
|Author:||Shield, Stephanie R.|
|Date:||Jan 1, 2003|
|Previous Article:||Modular conveyor systems setting a new standard for manufacturing efficiency. (Process Machinery).|
|Next Article:||Utilization of the rubber process analyzer in Six Sigma programs.|