Atomic Motions in Confined Polymer Films
Material properties at surfaces and interfaces often deviate from their bulk values. An example is viscoelastic properties that are important in many applications such as chemically amplified photoresists, lubricants, adhesives, etc. A viscosity increase of a liquid polymeric film, like those used for hard disc lubricants, could have deleterious consequences. Likewise, a viscosity decrease in a thin photoresist film could enhance photoacid diffusion and create image blur. It is clearly important to characterize the molecular level mobility in thin polymer films. Of particular interest are changes in the glass transition temperature Tg, since the glass transition corresponds to a sharp change in viscosity.
It is well known that atomic thermal motions lead to a decrease in the intensity of the elastically scattered neutrons. In a harmonic solid the intensity is related to the amplitude of the motion through the Debye-Waller factor, Iinc (Q) ~ exp(-1/3 Q2 <u2>), where Q is the scattering vector and <u2> is the mean-square atomic displacement. In a logarithmic plot of Iinc as a function of Q2, <u2> is proportional to the slope of a linear fit to the data. This approximation has proven very useful in studying the dynamics of polymers, both synthetic (Refer to Reference 1) and biological (Refer to Reference 2), despite the fact that their atomic motions are typically anharmonic.
The elastic incoherent neutron scattering intensities can be measured as a function of temperature on the N C N R High Flux Backscattering Spectrometer (H F B S), using the “Fixed Window” mode. Polycarbonate (P C) films, seen in Figure 1, are spin cast onto Si wafers, sectioned into small rectangles, and stacked into an Al sample cell. The scattering from the Si substrates and Al cell are negligible because their Bragg peaks are beyond the Q range of the instrument (0.25 Å-1 < Q < 1.75 Å-1); the scattering is dominated by the large incoherent cross-section of the amorphous PC.
Ideally there should be 200 mg of PC per cell to maximize the signal and minimize multiple scattering effects (90 % transmission). For the Si-supported films we are limited by 0.5 mg to 10 mg of PC per cell and a thus obtain a very weak signal (99 % transmission). However, the H F B S is sufficiently sensitive to observe this signal, as seen in Figure 2 where the temperature dependence of the elastic scattering is shown. The relative magnitude of the intensity change diminishes with decreasing film thickness, a clear indication of reduced atomic mobility in the thin PC films. Given the H F B S energy resolution of 0.85 µ eV, these intensity changes reflect changes in atomic motions at frequencies of 200 Megahertz or faster.
Graphics Caption FIGURE 1. A stack of 1014 Å P C films (notice the blue tint) on thin slabs of Si ready to go into the H F B S spectrometer.
Graphics Caption FIGURE 2. The normalized elastic intensities in the H F B S detector at Q = 0.36 Å-1. The decrease in intensity at high temperatures reflects the presence of atomic motion in the thin polymer films.
Graphics Caption FIGURE 3. <u2> as a function of temperature for the P C thin films. The inset reveals a stiffening of the harmonic (low temperature) force constant κ with decreasing film thickness h.
Graphics Caption FIGURE 4. <u2> as a function of thickness and temperature for similar P M M A (a) and P V C (b) films.
This reduced mobility is qualitatively reflected in Figure 3 as strong suppression of <u2>. We emphasize qualitatively because <u2> arises from a harmonic approximation. Nevertheless, there is a clear reduction in the amplitudes of the thermal motions, both above and below the calorimetric Tg, as the film thickness approaches the root-mean-square (rms) end-to-end length of PC (Ree = 160 Å).
At temperatures below 200 K, <u2> is small and the harmonic approximation is reasonable. In this regime the inverse of the slope of <u2> vs. T is proportional to an elastic force constant κ. The inset of Figure 3 displays a dramatic stiffening of κ with confinement. This implies that the atoms or molecules in P C become strongly caged by their environment in the thinner films.
Similar <u2> measurements on thin films of polymethyl methacrylate (P M M A; Ree = 915 Å) and polyvinyl chloride (P V C; Ree = 437 Å) are shown in Figures 4(a) and 4(b). There is a suppression of <u2> with decreasing film thickness above the bulk Tg that is qualitatively consistent with the P C films. However, below the Tg for bulk P M M A the suppression is less pronounced, and not observed in the P V C films. Careful inspection of Figures 3 and 4 suggests a relation between the magnitude of <u2> at Tg in the bulk and the degree of <u2> suppression in the thin glassy films.
Bulk P C has the largest <u2> at Tg, indicating extensive mobility. In fact, P C is well known for its segmental mobility in the glassy state, which is in part related to its superior impact resistance. Molecular mobility helps dissipate strain energy. (P C is used in bullet-proof glass.) This extensive mobility is also susceptible to thin film confinement, as also seen through the stiffening of the elastic constants from 1.9 N/m in bulk P C to 5.8 N/m in the 75 Å film. At the other extreme, bulk P V C has the smallest <u2> at Tg, indicating a lack of motions to be affected by confinement. This is also consistent with the strong intermolecular caging at low temperatures; both bulk P V C and the 220 Å film have large elastic force constants of 4.6 N/m. This lack of mobility reflects the observation that P V C is a brittle polymer that needs high levels of plasticization for practical use. Only when P V C is taken above the bulk Tg, where long-range motions dominate, do the effects of confinement become evident.
In short, we have seen two different responses to thin film confinement. While there is a general suppression of <u2> and the fast dynamics, polymers with greater mobility are more susceptible to confinement, especially below the calorimetric Tg. In contrast, polymers with strongly localized motions appear bulk-like, even in highly confined glassy films. More details of these studies can be found in our upcoming publications (Refer to References 3 through 5).
 B. Frick, D. Richter, Science 267, 1939 (1995).
 G. Zaccai, Science 288, 1604 (2000).
 C. Soles, J. Douglas, W. Wu, R. Dimeo, submitted to Phys Rev E.
 C. Soles, J. Douglas, W. Wu, H. Peng, D. Gidley, in preparation.
 C. Soles, E. Lin, J. Lenhart, R. Jones, W. Wu, D. Goldfarb, M. Angelopoulos, J. Vac. Sci. Tech. B, in press.
C. L. Soles, J. F. Douglas, and Wen-li Wu
National Institute of Standards and Technology
Gaithersburg, MD 20899-8541
R. M. Dimeo
N I S T Center for Neutron Research
National Institute of Standards and Technology
Gaithersburg, MD 20899-8562
Back to CHRNS Scientific Highlights
Last modified 30-May-2002 by website owner: NCNR (attn: Jeff Krzywon)