Anal. Chem. 1998, 70, 1768-1772
Phase and Curing Behavior of Polybutadiene/Diallyl Phthalate Blends Monitored by FT-IR Imaging Using Focal-Plane Array Detection Sung Joon Oh and Jack L. Koenig*
Department of Macromolecular Science, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, Ohio 44106-7202
The phase morphology and interfacial regions of cured and uncured polybutadiene/diallyl phthalate (PBD/DAP) blends that exhibit upper critical solution temperature (UCST) behavior have been characterized using an FT-IR microscopic imaging technique with focal-plane array detection. The observations of domain size and structure were based on monitoring characteristic infrared C-H, CdO, and C-O modes of the blend components. The basic phase morphology of the uncured blend was monitored immediately after sample preparation and after 24 h. The average domain sizes of the uncured blend increased with aging by coalescence of smaller particles. The chemical compositions of the separated domains were determined by measuring ratios of integrated intensities and analyzed using a calibration curve. Finally, the blends were cured above the UCST and an apparently homogeneous network was obtained. There is considerable effort to modify the useful properties of elastomers by introducing a coagent which is copolymerized with the elastomer and contributes positively to the performance properties of the elastomer.1-4 The macroscopic properties are determined by microscopic and molecular factors such as the degree of phase separation, the domain sizes, interfaces, and the composition of the different phases.4,5 One of the challenges in coagent elastomer blend technology is relating the initial domain distributions in the uncured state to those of the cured state, as these blends often exhibit upper critical solution temperature (UCST) separation behavior. This phase analysis problem is made difficult by the lack of optical contrast between the coagent and the elastomer in the domains of the uncured state. The optical contrast can be enhanced by staining techniques, but this approach is limited. Our approach is to use FT-IR microscopy to map the different multiphase domains. FT-IR microscopy does (1) Cornell, J. A.; Winters, A. J.; Halterman, L., Jr. Rubber Chem. Technol. 1970, 43, 613-623. (2) Keller, R. C. Rubber Chem. Technol. 1988, 61, 238-254. (3) Dikland, H. G.; Van Der Does, L.; Bantjes, A. Rubber Chem. Technol. 1993, 66, 196-212. (4) Dikland, H. G.; Ruardy, T.; Van Der Does, L.; Bantjes, A. Rubber Chem. Technol. 1993, 66, 693-711. (5) Dikland, H. G.; Sheiko, S. S.; van der Does, L.; Moller, M.; Bantjes, A. Polymer 1993, 34, 1773-1775.
1768 Analytical Chemistry, Vol. 70, No. 9, May 1, 1998
not require any staining procedure for contrast development, as the characteristic absorption frequencies and their corresponding intensities of the blend components provide the necessary image contrast. As traditionally used, FT-IR spectroscopy provides sampleaverage measurements of the composition of multicomponent samples. No spatially resolved spectral information is obtained, so whether the sample is heterogeneous or homogeneous cannot be determined. FT-IR microscopy has been used to map the spatial distribution of multiphase systems,6-9 but, until recently, the point-scanning approaches have been slow and time-consuming and suffer from illumination and positioning problems. Recently, the coupling of an infrared focal-plane array detector (FPA) to a step-scan interferometer has provided an instrumental advantage that substantially improves FT-IR microscopic measurements.10-13 The multiple detector elements of the FPA make it possible to acquire 4096 spatially resolved spectra with acquisition times of a few minutes with constant illumination. The spatial resolution is determined by the dimensions of the detector elements on the FPA (7.8 µm), and the spectral resolution is determined by the step-scan interferometer. This FPA-enhanced FT-IR microscopic technique has been applied to a number of biological systems,14 plastic waste identification,15 and polymer systems.16,17 The FPA-enhanced FT-IR system represents the state of the art in infrared chemical imaging analysis, as it combines (6) Mizakoff, B. K.; Taga, K.; Keller, R. Appl. Spectrosc. 1993, 47, 1476-1483. (7) McFarland, C. A.; Koenig, J. L.; West, J. L. Appl. Spectrosc. 1993, 47, 321329. (8) Challa, S. R.; Wang, S. Q.; Koenig, J. L. Appl. Spectrosc. 1995, 49, 267272. (9) Sommer, A. J.; Kanton, J. E. Appl. Spectrosc. 1991, 45, 1633-1640. (10) Lewis, E. N.; Treado, P. J.; Reeder, R. C.; Story, G. M.; Dowrey, A. E.; Marcott, C. M.; Levin, I. W. Anal. Chem. 1995, 67, 3377-3381. (11) Lewis, E. N.; Kidder, L. H.; Arens, J. F.; Peck, M. C.; Levin, I. W. Appl. Spectrosc. 1997, 51, 563-567. (12) Lewis, E. N.; Levin, I. W. Appl. Spectrosc. 1995, 49, 672-678. (13) Treado, P. J.; Levine, I. W.; Lewis, E. N. Appl. Spectrosc. 1994, 48, 607615. (14) Lewis, E. N.; Gorbach, A. M.; Marcott; C.; Levin, I. W. Appl. Spectrosc. 1996, 50, 263-269. (15) van den Broek, W. H. A. M.; Wienke, D.; Melssen, W. J.; Feldhoff, R.; HuthFehre, T.; Kantimm, T.; Buydens, L. M. C. Appl. Spectrosc. 1997, 51, 856865. (16) Snively, C. M.; Koenig, J. L., submitted to Macromolecules. (17) Bhargava, R.; Koenig, J. L. Appl. Spectrosc., in press. S0003-2700(97)01290-0 CCC: $15.00
© 1998 American Chemical Society Published on Web 03/21/1998
Figure 1. Infrared spectroscopic images of phase separation of PBD/DAP blends (50 wt % DAP) taken immediately after sample preparation. The phase-separated domain sizes are smaller than those after 24 h (Figure 2). Submicrometer phase separation is not detected due to instrument resolution.
Figure 2. Infrared spectroscopic images of phase separation of PBD/DAP blends (50 wt % DAP) taken 24 h after sample preparation. Red denotes high intensity and blue denotes low intensity in IR absorbance. DAP domains in PBD matrix and interfacial regions are observed.
spectroscopy with visualization. In particular, it allows chemical imaging of a broad range of heterogeneous systems. The objective of this study is to apply the FPA-enhanced FTIR imaging technique to a spatial analysis of the domains in blends of polybutadiene/diallyl phthalate (PBD/DAP) in the uncured and cured states. The domain compositions, sizes, shapes, and interfaces are monitored immediately after sample preparation and after 24 h of storage at room temperature. These blends exhibit UCST behavior, and the composition of the domains reflects the phase separation behavior. The changes in domain sizes and composition with time (prior to curing) are driven by minimization of the interfacial area. Upon heating the phase-separated blend above the USCT for purposes of curing, a homogeneous cured sample results.
EXPERIMENTAL SECTION Polybutadiene (PBD, Aldrich Co.) with a molecular mass of 4500 g/mol was used. The vinyl content of this polymer was 45%, and cis and trans content was 55%. Diallyl phthalate (DAP, Lancaster Co.) was used as cure coagent, and dicumyl peroxide (DCP, Aldrich Co.) was used as cure initiator. The chemical structures of PBD and DAP are shown in Chart 1. For the characterization of the morphology by FT-IR imaging, 50 wt % of DAP, 48 wt % of PBD, and 2 wt % of DCP were dissolved in cyclohexane. The solution was cast onto a NaCl plate and then dried on a hot plate at 80 °C for several minutes. The dried blend on the NaCl plate was covered by another NaCl plate. The thickness of polymer layer was kept constant by using 5-µm-thick glass spacers. The sample was annealed at 80 °C for an additional Analytical Chemistry, Vol. 70, No. 9, May 1, 1998
1769
Figure 3. Particle size distribution by image analysis (a) immediately after sample preparation and (b) 24 h after sample preparation. The increase in particle size and the decrease in particle numbers are observed. Volume fraction of DAP increased from 15.6 (a) to 43.5 vol % DAP (b). The increase in volume fraction results from the coalescence of submicrometer particles.
Chart 1. Chemical Structure of (a) Polybutadiene (PBD), X/Y ) 45/55, MW ) 4500 g/mol, and (b) Diallyl Phthalate (DAP)
2 min and then cooled to room temperature slowly. The images were taken immediately after sample preparation and again after 24 h. A Bio-Rad FTS 6000 Stingray FT-IR imaging spectrometer was used for the acquisition of image data. This spectrometer consists of an UMA500 microscope coupled to an FTS 6000 step-scan interferometer. A focal-plane array mercury-cadmium-telluride (MCT) detector was attached to this microscope. The 4096 spectra were acquired from different positions in a 500-µm × 500µm area. From these spectra, 64 × 64 pixel images were constructed. The spatial resolution of the images was estimated as 7.8 µm from the dimension and the pixel numbers. A Bio-Rad FTS 60 FT-IR spectrometer was also used to obtain the spectra of the individual components prior to FT-IR imaging experiments. The images acquired from the spectrometer were analyzed by the SigmaScan image analysis program from Jandel Scientific Software. RESULTS AND DISCUSSION The FT-IR spectra of the respective neat components of PBD and DAP were taken from the FTS 60. The C-H (3100-2800 cm-1) and CdC stretching (1630 cm-1) modes from PBD and CdO (1730 cm-1) and C-O stretching (1280 cm-1) modes from DAP were identified. From the comparison of these two spectra, the differences in absorbance intensities in the C-H stretching, CdO stretching, and C-O stretching modes were used as probes for FT-IR imaging contrasts. The blend morphologies are shown in Figures 1 and 2. The images were taken 0 (Figure 1) and 24 h (Figure 2) after the sample preparation. The DAP concentration of the blend was 50 wt %. The PBD C-H (2950 cm-1, a), DAP CdO (1730 cm-1, b), and DAP C-O stretching (1280 cm-1, c) images are shown in Figures 1 and 2. The pseudocolor of each image denotes the 1770 Analytical Chemistry, Vol. 70, No. 9, May 1, 1998
Figure 4. Calibration curves for quantitative analysis. The peak area ratio of the ester C-O stretching region (1340-1180 cm-1) to C-H stretching region (3140-2676 cm-1) was used. Data were fitted linearly for the concentration determination.
intensity of the absorbance in the FT-IR spectra. This is in the order of red > orange > yellow > green > blue, where red denotes the highest absorbance and blue denotes the lowest absorbance. Absorbance limits are shown with the images. PBD has a higher absorbance intensity in the C-H stretching band than DAP; therefore the intensity in panel a is blue for the DAP phase and yellow or green for the PBD phase. The DAP CdO and C-O stretching images are also shown in panels b and c with higher absorbances of the red color. The morphology changes on aging and is readily observed by comparing Figure 1 to Figure 2. The distinctive differences between Figures 1 and 2 are the domain size distribution of blends. The domain size distribution in Figure 1 is dependent on the cooling rate, as the kinetics of +k phase separation dominates. The image in Figure 1 that was taken immediately after the sample preparation has smaller DAP domains than the one taken 24 h later (Figure 2).
Figure 5. Interfacial region detected by FT-IR imaging: (a) image of DAP domain; (b) FT-IR spectra from the image across the interface; and (c) interfacial region monitored by the change in C-H stretching intensity at 2950 cm-1. The spectral positions of the aperture across the interface for five spectra (b) are shown in panel a as a solid line.
The image analysis results are shown in Figure 3. The average diameters of DAP domains were calculated from the measured area (A) and the equation 2 (A/π)1/2. An increase in particle sizes and a decrease in the number of particles were observed on aging. The driving force of the growth in particle sizes is the minimization of interfacial area.18,19 When the system is sufficiently mobile, the particle sizes grow to reduce the free energy of the system. The volume fraction of DAP before aging is calculated as 15.6% from the image analysis, whereas the value after aging is 43.5%. It is thought that the increase in DAP domain is due to coalescence of submicrometer-sized particles in the PBD matrix. These submicrometer-sized PBD particles are not spatially resolved from the image due to the limit in instrument resolution. The coalescence of larger domains is readily observed in Figure 2, in which two large particles are starting to merge into each other. Another advantage of this FT-IR spectral imaging is the capability of quantitative analysis. Two calibration curves were constructed for the analysis and are shown in Figure 4. One calibration curve (Figure 4a) is for the PBD-rich phase, and the other (Figure 4b) is for the DAP-rich phase. The aim of constructing two calibration curves is to avoid significant errors from phase separation. The selected concentration regions for the calibration curves are apparently homogeneous within the instrument resolution. Another convenience is that data can be fitted linearly within the small concentration region. The integration value of the DAP C-O stretching mode (1340-1180 cm-1) was normalized to the integration value of C-H stretching (3140(18) Park, D. W.; Roe, R. J. Macromolecules 1991, 24, 5324-5329. (19) Jang, B. Z. Rubber Chem. Technol. 1984, 57, 291-306.
2667 cm-1). The calibration data were fitted by a least-squares method to determine the individual concentrations of each phase. The choice for spectrum extraction in the image is arbitrary. The peak area ratio of C-O stretching to C-H stretching modes was calculated first, and then the DAP concentration was determined from the calibration curves. The measured value is 24.9 wt % ( 3.5 in the PBD matrix and 91.3 wt % ( 4.3 in the DAP domain before aging. After aging of the sample at room temperature, the DAP concentration dropped to a value of 17.8 wt % ( 4.1 in PBD matrix and increased to 88.3 wt % ( 4.9 in the DAP domain. The high concentration of DAP in the PBD phase before aging is thought to be due to submicrometer particles of DAP in PBD. The decrease in DAP concentration in the DAP domain on aging arises from the inclusion of submicrometer PBD particles into the DAP domain during aging process. The interfacial region can also be identified from the images. The existence of an interfacial region is the result of interdiffusion of the two molecules at the interface. The thickness of the interfacial region can be estimated by monitoring the change in the C-H stretching intensity across the interface. The spatial resolution of the instrument is determined by the dimensions of elements in the focal-plane array detector. As a 64 × 64 pixel image corresponds to a 500-µm × 500-µm area, one pixel position has a spatial resolution of 7.8 µm. Figure 5 shows the DAP domain image (a) and the C-H stretching absorbance spectra across the interface (b). The spectral positions of the aperture across the interface for five spectra (b) are shown in panel a as a solid line. The C-H stretching intensity changes are shown in panel c. The thickness of the interfacial region is estimated as Analytical Chemistry, Vol. 70, No. 9, May 1, 1998
1771
15.6 µm from plot c. The reliability of the measurement is limited by the focal-plane array detector’s spatial resolution. The blend sample was cured at 150 °C for 15 min, and the cured image changes from heterogeneous to homogeneous. UCST behavior of this blend was confirmed by optical microscope examination, with the cloud point around 75 °C. As the temperature is increased, the blends change from two-component (heterogeneous) to single-component (homogeneous). At the same time, the peroxide starts to decompose, and the curing reaction begins. The homogeneous morphology at 150 °C is locked in by the curing reaction. The curing mechanism involves allylic hydrogen abstraction from PBD molecules.3 The allylic hydrogen is easily abstracted because the polymer radical can be stabilized by resonance with the double bond. The DAP polymerizes on the polymer radicals to form cyclized or uncyclized structures. The pendant, uncyclized allyl groups react with another PBD radical, forming the cross-
1772 Analytical Chemistry, Vol. 70, No. 9, May 1, 1998
linked structure. Other side reactions can also occur, but the above mechanism is thought to be predominant. CONCLUSIONS The morphology of PBD/DAP blends that have UCST behavior can be characterized by FT-IR imaging with focal-plane array detection. The average domain sizes of the uncured blends increase by coalescence of smaller particles. The chemical compositions of the separated domains were determined by measuring ratios of integrated intensities and analyzed using a calibration curve. The blends were cured above the upper critical solution temperature, and an apparently homogeneous network was obtained. Received for review November 26, 1997. February 7, 1998. AC971290D
Accepted
