Langmuir 2008, 24, 2539-2544
2539
Segmental Dynamics of Poly(methyl acrylate)-d3 Adsorbed on Anopore: A Deuterium NMR Study Macduff O. Okuom,† Burak Metin,‡ and Frank D. Blum* Departments of Chemistry, Materials Science and Engineering, and Materials Research Center, Missouri UniVersity of Science and Technology, Rolla, Missouri 65409 ReceiVed October 7, 2007. In Final Form: NoVember 21, 2007 The segmental dynamics of poly(methyl acrylate-d3) (PMA-d3) adsorbed in the pores of anopore membranes has been investigated using deuterium NMR over the temperature range 25-80 °C. The onset of the NMR glass-transition temperature (Tg) for the adsorbed samples was approximately 15 °C higher than that for the bulk sample. The adsorbed polymer contained segments with restricted mobility (glassy), even at the highest temperatures studied, at which the bulk polymer showed only mobile segments. The spectra from samples with different adsorbed amounts of PMA-d3, between 1.1 and 4.2 mg/m2, were similar in their temperature-dependent mobilities. Neither was there much difference in the spectra of PMA-d3 on anopore samples with pore sizes of 0.2 and 0.02 µm. However, for a solvent-washed sample with an adsorbed amount of 0.7 mg/m2, additional restriction in PMA-d3 mobility was observed.
Introduction Many technological advances involve the use of polymeric composites. The interactions between polymers and surfaces often play crucial roles in determining the properties of these composites. As such, it is important to characterize the interfaces in polymer composites, especially with respect to the behavior of the polymeric component. Since the amount of polymer affected by the presence of and interaction with the surface is very small, spectroscopic techniques sometimes have the sensitivity to probe the behavior of adsorbed polymer. In recent years, a number of studies have focused attention on the properties of adsorbed polymers. Differences in adsorbed and bulk polymer systems have been readily observed in properties such as the glass-transition temperature (Tg). Interfacial polymers may have increased or decreased Tg’s depending on their interaction with the adjacent interfaces, local geometries, and other factors.1,2 Of particular interest to us has been the adsorption of carbonyl-containing polymers on metal-oxide particles. The hydrogen bonding of these polymer’s carbonyl groups to surface hydroxyls3 undoubtedly are responsible for changes in the mobility of polymer segments close to the silica surface. Adsorbed PMMA on silicon (with an oxide surface) or pure silica showed an increased Tg as determined using optical/scattering techniques like ellipsometry,4,5 or neutron reflectivity,6,7 though the details like the breadth of the transition are a bit difficult to discern from * To whom correspondence should be addressed. E-mail: fblum@ mst.edu. † Present address: Department of Chemistry, University of Pittsburgh, 219 Parkman Ave., Pittsburgh, PA 15260. ‡ Present address: SoloPower Inc., 1635 McCandless Dr., Milpitas, CA 95035. (1) Forrest, J. A.; Dalnoki-Veress, K. AdV. Colloid Interface Sci. 2001, 94, 167-196. (2) Alcoutlabi, M.; McKenna, G. B. J. Phys.: Condens. Matter 2005, 17, R461-R524. (3) Kulkeratiyut, S.; Kulkeratiyut, S.; Blum, F. D. J. Polym. Sci., Part B: Polym. Phys. 2006, 44, 2071-2078. (4) Keddie, J. L.; Jones, R. A. L.; Cory, R. A. Faraday Discuss. 1994, 98, 219-230. (5) Sharp, J. S.; Forrest, J. A. Phys. ReV. E 2003, 67, 031805/1-031805/9. (6) Wu, W.-L.; van Zanten, J. H.; Orts, W. J. Macromolecules 1995, 28, 771774. (7) Lin, E. K.; Wu, W. I.; Satija, S. K. Macromolecules 1997, 30, 7224-7231.
these techniques. Thermal techniques, like local thermal analysis8 and modulated differential calorimetry (MDSC),9-11 also showed increases in the PMMA Tg due to adsorption on silica. The derivative MDSC mode has been shown to be capable of separating a 1.1 nm tightly bound layer and more loosely bound PMMA.10 Other related carbonyl-containing polymers adsorbed on oxide surfaces, including mechanical studies,12,13 also exhibited increases in Tg (or decreased rates of dynamics) compared to bulk. In our group, we have studied adsorbed polymers on small particles at very small adsorbed amounts because, at these levels, the altered behavior of the polymer segments at the interfaces can be readily observed. In particular, deuterium NMR has provided significant detail on the dynamics of polymer segments in the glass-transition region. For poly(vinyl acetate) (PVAc) on silica,14 a motional gradient was shown to exist with more mobile segments located at the polymer-air interface and the less mobile segments at the polymer-silica interface. Additional verification of the nature of the interfacial interactions was found with deuterated poly(methyl acrylate) (PMA).15-17 When adsorbed PMA was overlayered with an unlabeled polymer, the segments at the polymer-air interface were shown to lose their extra mobility.17 Deuterium NMR studies of PMA adsorbed on silica have also provided additional details on adsorbed polymers including effects of adsorbed amount,15 molecular mass,18 and relationship to MDSC experiments.19 In these cases, the interaction of the (8) Fryer, D. S.; Nealey, P. F.; de Pablo, J. J. Macromolecules 2000, 33, 6439-6447. (9) Porter, C. E.; Blum, F. D. Macromolecules 2000, 33, 7016-7020. (10) Blum, F. D.; Young, E. N.; Smith, G.; Sitton, O. C. Langmuir 2006, 22, 4741-4744. (11) Kabomo, M.; Blum, F.; Kulkeratiyut, S.; Kulkeratiyut, S. J. Polym. Sci., Part B: Polym. Phys., in press. (12) Wang, X. P.; Xiao, X. D.; Tsui, O. K. C. Macromolecules 2001, 34, 4180-4185. (13) Primak, S. V.; Jin, T.; Dagger, A. C.; Finotello, D.; Mann, E. K. Phys. ReV. E 2002, 65, 031804/1-031804/4. (14) Blum, F. D.; Xu, G.; Liang, M.; Wade, C. G. Macromolecules 1996, 29, 8740-8745. (15) Lin, W.-Y.; Blum, F. D. Macromolecules 1997, 30, 5331-5338. (16) Lin, W.-Y.; Blum, F. D. Macromolecules 1998, 31, 4135-4142. (17) Lin, W. Y.; Blum, F. D. J. Am. Chem. Soc. 2001, 123, 2032-2037. (18) Metin, B.; Blum, F. D. J. Chem. Phys. 2006, 125, 054707/1-054707/9. (19) Blum, F. D.; Lin, W.-Y.; Porter, C. E. Colloid Polym. Sci. 2003, 281, 197-202.
10.1021/la703103j CCC: $40.75 © 2008 American Chemical Society Published on Web 02/06/2008
2540 Langmuir, Vol. 24, No. 6, 2008
polymer and surface was limited to silica. However, there is clear need for extension of this work to determine the dynamics of adsorbed PMA on other surfaces. Anopore is one such interesting substrate appropriate for such an extension. Anopore is an inorganic membrane material that is made by the anodic oxidation of aluminum20 and consists of relatively uniform cylindrical pores with oxide interfaces. Anopore has been used as a confining substrate for adsorption. It appears that adsorbed liquid crystals in anopore can show loose ordering,21 high degrees of ordering,22 and at least one adsorbed polymer, polydimethylsiloxane-d6, moderate ordering,13 consistent with relatively flat chain adsorption. It is not known whether other polymers would tend to orient inside of anopore. In this paper, we report the study of the dynamics of PMA-d3 on anopore using deuterium NMR. We examine the process of adsorption and the effect of adsorbed amounts and pore sizes on the polymer segmental dynamics. Experimental Section PMA-d3 with molecular masses of 33 and 37 kDa with polydispersities of 1.07 and 1.18, respectively, measured by gel permeation chromatography using an Optilab DSP interferometer refractometer and Dawn EOS light-scattering instrument (both from Wyatt Technology, Santa Barbara, CA) were used. For the purposes of this study, these molecular masses were assumed to be equivalent. A detailed description of the synthesis of PMA-d3 can be found elsewhere.23 Basically, methyl acrylate-d3 was prepared by the reaction of acryloyl chloride and methanol-d4. The resulting monomer was polymerized by atom-transfer radical polymerization (ATRP). The 33 and 37 kDa polymers formed were found, by proton NMR, to be 51% and 50% meso-dyads, respectively. The bulk glasstransition temperatures were determined to be 2.7 and 3.5 °C, respectively, using differential scanning calorimetry (DSC) (DSC 2920 Modulated DSC, TA Instruments, Newcastle, DE). Protonated PMA of molecular mass 33 kDa and polydispersity 1.15 was also prepared by the same method, for use in the isotherm measurements. Anopore membranes (Whatman, Florham Park, NJ) were purchased as 60 µm thick discs, with 0.2 or 0.02 µm diameter pore sizes, and were cut into small 4 × 10 mm pieces. Adsorption was carried out by immersing the pieces of anopore in a solution of PMA-d3 in chloroform for 5 min. Different adsorbed amounts were made from solutions as follows: a 10% solution was used to make the 4.2 mg/m2 samples, and a 6% solution was used for the 1.1 and 1.4 mg/m2 samples. After adsorption, the samples were removed from the solution and allowed to dry in air for at least 45 min. One sample of 33 kDa PMA-d3 on a 0.2 µm pore size anopore was dried, washed with chloroform, and then dried again. Partial adsorption isotherms were made using the protonated PMA of a molecular mass similar to the deuterated samples. Samples were retrieved from the anopore-PMA solution mixture, at 5 min intervals, patted with tissue paper and dried, or washed in solvent and dried. The adsorbed amounts were determined using a thermogravimetric analyzer (TGA, Hi-Res Thermogravimetric Analyzer 2950, TA Instruments, Newcastle, DE). Scanning electron microscope (SEM) images of anopore and PMA-d3 adsorbed anopore were taken on an Hitachi S-4700 FESEM microscope. The samples were coated with palladium, in an argon atmosphere. When possible, the sample was cut into two pieces so that either side could be mounted on the goniometer and examined. The deuterium NMR spectra were obtained using a Varian VXR400/S spectrometer. The quadrupole-echo pulse sequence (delay90y-τ-90x-τ-acquisition) was used with a 2H frequency at 61.39 MHz (20) O’Sullivan, J. P.; Wood, G. C. Proc. R. Soc. Ser. A 1970, 317, 511-543. (21) Crawford, G. P.; Yang, D. K.; Zumer, S.; Finotello, D.; Doane, J. W. Phys. ReV. Lett. 1991, 66, 723-726. (22) Crawford, G. P.; Steele, L. M.; Ondris-Crawford, R.; Iannacchione, G. S.; Yeager, C. J.; Doane, J. W.; Finotello, D. J. Chem. Phys. 1992, 96, 77887796. (23) Metin, B.; Blum, F. D. J. Chem. Phys. 2006, 124, 054908/1-054908/10.
Okuom et al.
Figure 1. Partial adsorption isotherms for PMA on anopore for adsorbed samples dried (filled squares) and washed with chloroform (open squares). using an 8 mm coil single-tuned deuterium probe (Doty Scientific, Columbia, SC). The 90° pulse width used was 2.7 µs with an echo time of 30 µs. The number of scans collected for the bulk sample was 128, and from 4096 to 16384 scans were collected for the adsorbed samples. The time-domain spectra were Fourier-transformed starting at the peak of the echo. No line broadening was used. For the bulk and adsorbed samples, spectra were taken at temperatures of 25-65 °C, at 5 °C intervals. The anopore-adsorbed samples were broken into small pieces and stacked together with the same relative orientation and put in the NMR sample tubes. They were tested for segmental orientation by rotating the samples 90° and then taking the spectra. Since the cylindrical holes in the samples were perpendicular to the flat surfaces, orientations parallel and perpendicular to the pores were probed. Simulations of the experimental spectra were made based on the Multiple Axis Quadrupolar Exchange program, MXQET.24 The program includes the effect of a finite pulse width on the 2H NMR line shapes. The simulations were based on a Pake powder pattern with a reduced quadrupolar-coupling constant (QCC) of 50 kHz to account for the fast rotation of the methyl groups. The motion was assumed to be that of small jumps, in this case simulated by nearest neighbor jumps between sites on the vertices of a truncated icosahedron (soccer-ball shape) following the procedure of Metin and Blum.23 Other jump models with fewer sites, e.g., tetrahedral (4 jump sites), octahedral (6 jump sites), dodecahedral (20 jump sites), and icosahedra (12 jump sites) had jumps that were too large to successfully match the experimental spectra.23 A series of 94 basis spectra were simulated with varying jumping rates from 100 to 1 × 1011 jumps/s. A mathematical routine was applied to fit the experimental line shapes to a superposition of simulated spectra with MATLAB.25
Results The adsorbed amounts for some of the anopore samples as a function of time are shown in Figure 1. Additional data points (not shown) at 720 min had the same value as those at earlier times. The adsorbed amounts remained relatively constant after the first 5 min of immersion. These samples averaged about 1.9 mg/m2. The adsorbed amount was reduced considerably for the washed samples to about 0.26 mg/m2. The 2H NMR variable temperature spectra for bulk 33 kDa PMA-d3 at 25-65 °C are shown in Figure 2. The spectra for the bulk polymer were similar to those previously described.18 At 25 °C, a Pake powder pattern was evident with “horns” at (18 kHz that are rather smooth at the ends. The bulk sample began to show the presence of a central resonance at 35 °C in addition to the dominant Pake powder pattern. A fairly broad feature, with powder pattern characteristics, was apparent at 40 °C. At 45 °C, the powder pattern collapsed to a broad single resonance. (24) Greenfield, M. S.; Ronemus, A. D.; Vold, R. L.; Vold, R. R.; Ellis, P. D.; Raidy, T. E. J. Magn. Reson. 1987, 72, 89. (25) MATLAB. MATLAB; The Mathworks, Inc.: Natick, MA.
Dynamics of PMA-d3 on Anopore
Figure 2. Experimental (s) and simulated (‚‚‚) 2H NMR spectra for bulk 33 kDa PMA-d3.
Figure 3. Experimental (s) and simulated (‚‚‚) 2H NMR spectra for 33 kDa PMA-d3 adsorbed (1.1 mg/m2) on 0.02 µm pore size anopore.
This central resonance got narrower with increases in temperature. Part of the spectrum at 50 °C was magnified 5×, verifying the absence of any significant amount of residual powder pattern in the wings of the spectrum. The resonance continued to narrow with increasing temperature. Variable temperature 2H NMR spectra for the 0.02 µm pore size anopore-adsorbed 33 kDa PMA-d3 (at 1.1 mg/m2) are shown in Figure 3. At 25 °C, the spectrum consisted of a Pake powder pattern. The outer horns were sharper than those for the bulk sample at the same temperature. At 40 °C, the powder pattern remained, though the horns were a little more rounded. At 45 °C, a central resonance began to fill in the center of the Pake
Langmuir, Vol. 24, No. 6, 2008 2541
Figure 4. Experimental (s) and simulated (‚‚‚) 2H NMR spectra for 33 kDa PMA-d3 adsorbed (1.4 mg/m2) on 0.2 µm pore size anopore.
Figure 5. Experimental (s) and simulated (‚‚‚) 2H NMR spectra for 37 kDa PMA-d3 adsorbed (4.2 mg/m2) on 0.2 µm pore size anopore.
powder pattern. As the temperature was increased, the central resonance became larger while the intensity of the outer horns was reduced. At 55 °C and above, almost all of the Pake powder pattern collapsed as the central resonance increased due to faster segmental motion. The magnifications in the spectra at 55 and 65 °C showed only a hint of the shoulders in the powder patterns. Both the parallel and perpendicular orientations of the pores (with respect to the magnetic field) produced Virtually identical spectra. Spectra obtained from the samples on the larger (0.2 µm) pore size anopore, 33 kDa (at 1.4 mg/m2) and 37 kDa PMA-d3 (at 4.2 mg/m2) are shown in Figures 4 and 5, respectively. In essence, the spectra for these two samples were effectively the same as
2542 Langmuir, Vol. 24, No. 6, 2008
Okuom et al.
Figure 6. Experimental (s) and simulated (‚‚‚) 2H NMR spectra for 33 kDa PMA-d3 adsorbed (0.7 mg/m2) on 0.2 µm pore size anopore (washed sample).
those shown in Figure 3 for the 0.02 µm anopore. Apparently, a change in the anopore pore size or increases in the adsorbed amount do not significantly affect the spectra. Some minor differences between the spectra occur; for example, there is a bit more of a hint of the powder pattern shoulders in the expanded spectra of the 4.2 mg/m2 sample in Figure 5. In some contrast to the similarity of the other spectra, the spectra of the washed sample, shown in Figure 6 (33 kDa PMAd3), had moderate differences. This sample was made by washing part of the sample whose spectrum is shown in Figure 4. The signal-to-noise ratios of the spectra in Figure 6 were also smaller because of the smaller amount of polymer present. We note that some of these spectra were taken at different temperatures than those in Figures 2-5. At the two lower temperatures, the spectra were similar to those of the other three, higher adsorbed amount samples. At 55 °C, broader shoulders are apparent as is a central component. At 65 °C, there were still noticeable, but small, shoulders in the spectrum. These shoulders are apparent even at temperatures as high as 79 °C, when that area of the spectrum is expanded. The simulated spectra were based on the MXQET FORTRAN program.24 In this case, the spectra were simulated as from nearest neighbor jumps on the vertices of a truncated icosahedron (soccerball shape).18,23 Figures 2-6 include the simulated spectra as dotted lines along with the experimental spectra. The fits to each of the spectra were quite good. To simplify the results of the fitting, we have grouped the sets of spectra together with respect to the motional rates as slow (1.0 × 102 to 1.00 × 104), intermediate (1.1 × 104 to 1.5 × 105), or fast (2.0 × 105 to 1.00 × 1011 jumps/s). The distributions of these for different samples at different temperatures are shown in Figure 7. In general, the amounts of intermediate, and then fast, components increased with increased temperatures. The bulk sample had the largest amount of intermediate and faster material, the moderate and larger adsorbed amount samples were pretty similar to each other, and the washed sample had the most restricted components. For each simulated spectrum, the correlation functions of each component used in the simulation were added together to generate a correlation function, C(t). Similar to our findings for PMA in
Figure 7. Graphical representation of motional rate distribution of the simulated spectra for the different samples grouped into components in the slow, intermediate, and fast regimes.
bulk and on silica,18,23 a Kohlrausch-Williams-Watts (KWW) function with a fixed width parameter (β ) 0.63) provided a best fit of the correlation function for each spectrum. C(t) ) exp(-t/τ)β
(1)
The fitted curves were integrated to yield the average correlation times, 〈τ〉. The average correlation times for the different samples are shown in Figure 8 as a function of temperature. As expected, the correlation times decrease with increasing temperature and those for the adsorbed samples are considerably slower than those for the bulk samples. An example of an electron micrograph of an anopore sample with adsorbed polymer is shown in Figure 9 for PMA (4.2 mg/ m2) on 0.2 µm pore size anopore. The anopore shows the holes (dark areas) associated with the nominal 0.2 µm pores. The light areas are the supporting oxidized aluminum. When the membrane is turned over, the other side of it shows thinner walled polyhedra. The presence of small amounts of polymer could not be detected with certainty, although in some cases there appears to be some duller gray areas around the holes that may be due to adsorbed polymer.
Discussion The partial adsorption isotherms of PMA on anopore (Figure 1) show that the maximum adsorbed amount is attained after 5
Dynamics of PMA-d3 on Anopore
Langmuir, Vol. 24, No. 6, 2008 2543 〈3 cos2 θ(t) - 1〉 ) (1/2)〈3 cos2 β(t) - 1〉(3 cos2 χ(t) - 1) (3)
Figure 8. Average correlation times for bulk and adsorbed PMA samples as a function of temperature.
Figure 9. Scanning electron microscopy picture of 0.2 µm anopore with 4.2 mg/m2 PMA adsorbed. The scale bar is 100 nm.
min of exposure. There was also not any significant time dependence of the adsorbed amounts over the time ranges studied. Washing the adsorbed sample with solvent considerably reduced the adsorbed amount; i.e., some of the polymer could be washed off the anopore. This behavior contrasts with that of PMA-d3 or PVAc-d3 adsorbed on silica.14,15 On silica, these polymers maintained a much larger adsorbed amount after washing. We believe that this is an indication of either (or both) a weaker interaction between anopore and PMA and/or a different structure of the adsorbed polymer. For the deuterium nucleus (spin quantum number, I ) 1), the quadrupolar splitting, ∆νq, is given by26 ∆νq ) (3/4)(e2qQ/h)[3 cos2 θ(t) - 1 - η sin2 θ(t) cos2 φ(t)]
(2)
where e2qQ/h is the quadrupole coupling constant, θ and φ are the spherical polar angles for the orientation of the principal axis system of the electric field tensor relative to the applied static magnetic field, B0, and η is the asymmetry parameter of the quadrupolar tensor. In general, the spherical polar angles vary with time, t. In our case, for a methyl group undergoing rapid rotation about its symmetry axis, the value of η may be taken as 0. For a C-D bond or a methyl group in particular, the 3 cos2 θ(t) - 1 term can be expanded as (26) Abragam, A. The Principles of Nuclear Magnetism; Clarendon Press: Oxford, 1961; p 599.
where 〈 〉 represents the time average, β(t) is the angle between the B0 and the rotation axis, and χ is the angle between the rotation axis and the C-D bond. Since χ is 70.5° for a methyl deuteron, the 3 cos2 χ - 1 term reduces the quadrupole splitting to 1/3 of its static value. Values for QCC of methyl groups are typically on the order of 150-170 kHz.27 Solid-state deuterium NMR line shapes are dominated mainly by the quadrupolar interactions27 and are sensitive to the orientation of labeled methyl groups on the polymer chain. Due to the coupling of C-D bond axes with cooperative motions of the polymer chains, deuterium NMR can yield a wealth of information about the dynamics through glass-transition temperature.28 The powder pattern for a solid sample collapses to a single line with increasing temperature due to changes in the mobility of methyl groups and the backbone motion in the polymer chain.15 The results of the variable temperature NMR on the bulk sample confirm this behavior (Figure 2). At 25 °C, the bulk sample exhibited a Pake powder pattern with a quadrupolar coupling constant of 50 kHz, typical of the rotating C-D of methyl groups.29 The smooth edges of the horns could be a result of some relatively more mobile segments in the presence of more rigid segments. The simulations show the presence of slowsand intermediatesrange motional components. A middle peak, characteristic of faster segmental mobility, begins to appear at 35 °C in the presence of a dominant Pake powder pattern. Motional rates in the intermediate range dominate in this region. The powder pattern collapses at 45 °C as a broad central resonance takes its place. These changes are due to increased rates of segmental motion with increasing temperature. As the temperature increases, the central resonance gets even narrower as the segments in the fast motional components begin to dominate the spectra. The anopore-adsorbed samples showed a gradual collapse of the powder pattern with an increase in temperature; however, the residual powder patterns and the broader signals are characteristic of the effects of hydrogen-bonded segments. For 33 kDa PMAd3 with 1.1 mg/m2 adsorbed on 0.02 µm pore size anopore, Figure 3, a Pake powder pattern was obtained at 25 °C. This pattern was similar to that of the bulk sample at the same temperature, except for sharper edges near the horns. These sharper edges are associated with larger fractions of segments in the slow motional rate region, as confirmed by the simulations. As the temperature increased, the edges of the horns became more rounded as the rates of motion of the segments increased. A central resonance emerged at 45 °C and its intensity increased as the temperature increased, while that of the powder pattern decreased. At 55 °C, the powder pattern collapsed and was replaced by a broad peak. Even though a 5× magnification did not show evidence of a residual powder pattern at 60 and 65 °C, the broadness of the resonance including a sharp center was an indication of a wide range of motional rates present. There was remarkable similarity between the variable temperature NMR spectra of PMA-d3 adsorbed on a 0.02 µm pore size, with an adsorbed amount of 1.1 mg/m2 (Figure 3), and that adsorbed on the 0.2 µm pore size anopore, with adsorbed amounts of 1.4 mg/m2 (Figure 4) and 4.2 mg/m2 (Figure 5). Comparison with the washed sample (Figure 6) spectra revealed significant (27) Ulrich, A. S.; Grage, S. L. 2H NMR. In Solid State NMR of Polymers; Ando, I., Asakura, T., Eds.; Elsevier: Amsterdam, 1998; p 190. (28) Pschorn, U.; Roessler, E.; Sillescu, H.; Kaufmann, S.; Schaefer, D.; Spiess, H. W. Macromolecules 1991, 24, 398-402. (29) Harris, R. K. Nuclear Magnetic Resonance Spectroscopy; Pitman: Marshfield, MA, 1983.
2544 Langmuir, Vol. 24, No. 6, 2008
Okuom et al.
differences, especially in the persistence of the Pake powder pattern at 65 °C and above. Differences between the bulk and anopore samples are attributed to the interaction of the aluminalike surfaces and the polymer. It is therefore apparent that the bulk sample had relatively high mobility, while at the other extreme, the washed sample showed the most significant motional restriction. At 45 °C, the Pake powder patterns from the different adsorbed samples looked very similar. However, at 55 °C, the washed sample showed a well-defined powder pattern (sharp shoulders) in the presence of a sharp central resonance. The residual Pake pattern is quite evident, even at the highest temperature (79 °C) studied for the washed sample. The difference in motional rates between the bulk and the adsorbed samples can be attributed to restrictions due adsorption. The similarities and differences in the spectra can also be identified through the simulations, which fit the spectra quite well. The results of the simulations have been simplified through conversion into three general categories. Slow jumps (