Polymer dynamics in multicomponent systems.

     The long stringy nature of polymer molecules gives them properties that cannot be found in any other material.  We work to tie together the viewpoints of polymer physicists and polymer chemists.  Physicists (stereotypically) are looking for universal properties, i.e., in what ways do polymer molecules all act the same?  Chemists tend to emphasize the peculiarities of polymer behavior and ascribe them to particular details of the molecular structure.  When possible, we try to integrate these perspectives, attempting to understand from the molecular structure exactly how the “universal” properties will be exhibited.

        Almost all applications of polymers utilize copolymers, polymer blends, or composites, as opposed to homopolymers.  Miscible polymer blends are a good starting point for understanding the properties of these composite systems; they are also technologically important and have flow properties which cannot currently be predicted. 

        In most polymer melts and blends, conformational(segmental) dynamics are the fundamental motions which drive longer length scale rearrangements.  These dynamics, which occur on the length scale of a few Angstroms, ultimately control the viscosity, elasticity, and glass transition of the system. In polymer blends, the segmental dynamics of each component are modified from the values which they have as pure components and in general are not equal to each other.  Composition fluctuations thus lead to spatial variations in the mobility of the blend.  Two major questions regarding polymer blends are: 

1) How can the segmental dynamics of each component in the blend be predicted from the properties of the pure homopolymers? 

2) Given knowledge of the segmental dynamics of each component, how can the long time transport and flow properties of the blend be predicted?

     NMR measurements have a special role to play in the study of polymer blends, since isotopic labelling allows each component to be selectively interrogated. However, the successful application of NMR relaxation time measurements to polymer blends requires the ability to turn relaxation parameters into a quantitative description of the distribution of segmental relaxation times.

 Simulation of Polyisoprene / Polystyrene Blends

by Roland Faller (UC-Davis)

 Using multi-field measurements, we have recently shown that this is possible in both single component systems and in miscible blends and block copolymers.  An essential feature of our approach is the comparison with large scale molecular dynamics computer simulations.  These simulations provide a check on the interpretation of our measurements, while the measurements also check the simulations.

 Recent powerpoint presentation on polymer mixture dynamics

	Supercooled liquids/diffusion in thin films.

    If crystallization is avoided upon cooling, a liquid will become more viscous and gradually transform into an amorphous solid which we call a glass.   Part of the interest in this process is that a new type of matter results by a purely kinetic process, i.e., the glass transition does not result from changes in structure.   Glasses of many types (polymeric, inorganic, saccharide...) play important roles in technologies.

Spatially heterogeneous dynamics
Simulation by Glotzer group (U. of Michigan)

Much of the phenomenology of supercooled liquids and glasses transcends a particular class of materials.  We study molecules such as o-terphenyl and glycerol; these are arguably the simplest systems which supercool and form glasses.  These are ideal systems for uncovering the essential features of glass formation and insights from the study of these materials can be transferred to technologically relevant systems.As an example, our studies on low molecular weight glass formers have suggested the explanation for the anomalously fast diffusion of solvents, antioxidants, and plasticizers through polymers, as well as some puzzling rheological properties.

Other areas where an understanding of mobility of supercooled liquids is essential include

pharmaceuticals (will an amorphous drug crystallize during storage?), enzyme and tissue preservation (can water be removed from a saccharide solution to yield a glassy matrix while preserving biological structures?), and food products (can crystallization of LifeSavers be suppressed?).

In the last few years, we have shown that, as the glass transition is approached, dynamics become increasingly spatially heterogeneous, i.e., the dynamics in one region of the sample may be orders of magnitude faster than the dynamics a few nanometers way.  One important consequence of this heterogeneity is that translational diffusion coefficients can be 1000 times larger than expected.  Our current work involves vapor deposition of thin films of deuterated and hydrogenous glass formers onto a cold substrate, followed by subsequent interdiffusion.  Using neutron reflectivity (NR) and thermally programmed desorption (TPD), these experiments have extended the range of measured diffusion coefficients

Self-diffusion of tris-naphthyl benzene

  by 6 orders of magnitude.  Neutron reflectivity measurements provide a direct test for the presence of nanometer-size dynamic heterogeneities.

 Powerpoint presentation tutorial on dynamics in supercooled liquids and polymers

 Powerpoint presentation on supercooled liquids (ACS meeting, Sept 2003)

 Recent powerpoint presentation on "Unanswered questions about supercooled liquids and glasses"

	Creation of Exceptionally Stable Glasses

    We have prepared what are very likely the most stable glasses ever made in a laboratory. In an afternoon, we can make glasses that would require at least 4000 years to prepare using any published methodology. Our glasses have useful material properties. For example, because they are more dense and energetically more stable than ordinary glasses, they resist crystallization and water uptake. These stable glasses may have immediate technological relevance, e.g., to stabilize amorphous pharmaceuticals. More fundamentally, as compared to ordinary glasses, our glasses are much deeper in the energy landscape that controls the thermodynamics and kinetics of an amorphous system. There has been intense recent activity aimed at understanding fundamental issues of glass formation, such as the Kauzmann entropy paradox, in terms of the energy landscape. Our stable glasses are ideal for exploring these fundamental issues.

   We have prepared stables glasses of two organic molecules by vapor deposition onto substrates held somewhat below the glass transition temperature. In this temperature range, the top few nanometers of a glass can be substantially more mobile than the interior of the glass. If deposition is sufficiently slow, molecules sticking to the surface have the opportunity to sample many configurations before they get buried and immobilized. This configuration sampling allows the top layer to attain or nearly attain equilibrium at the substrate temperature. Subsequent layers are similarly equilibrated. In this way, the usual kinetic constraints to the production of stable glasses can be avoided. Given the deep connections between glasses and other energy landscape problems (e.g., protein folding), we hope that this work have a multidisciplinary impact. This is a collaborative effort with Robert McMahon and Lian Yu (UW-Madison) and Sushil Satija (NIST Center for Neutron Research).

PowerPoint presentation on "Creation of Exceptionally Hiking down the energy landscape:
How vapor deposition and surface mobility produce exceptionally stable glasses"

PowerPoint presentation on "Exceptionally stable organic glasses with low enthalpy and
high kinetic stability prepared by vapor deposition" (APS meeting, March 2008)

PowerPoint presentation on "Comparison of surface mobility on polymeric and low molecular weight glass-formers" (APS meeting, March 2008)

	Crystal growth from glasses and supercooled liquids

    Amorphous materials are often useful as precursors to crystalline materials. For example, if we can learn to control crystallization in amorphous aluminum alloys, we should be able to produce a material stronger (and significantly lighter!) than steel. In other cases, spontaneous crystallization destroys the useful properties of amorphous materials. The pharmaceutical industry has developed many drugs that cannot be marketed because the crystalline form of these drugs is not sufficiently soluble. It has been shown that drugs delivered as glasses can be more than 20 times as active as the crystalline form, because of enhanced solubility. However, this is only possible if the glassy drug is stable in the package for two years without crystallizing.

Currently no one can accurately predict crystallization under the conditions relevant to these applications – the current theory cannot predict correctly either nucleation rates or crystal growth rates. For example, the accompanying figure shows that the kinetic part of crystal growth rates is not inversely proportional to the viscosity, as predicted by current theory. There are good reasons to believe that these failures occur because of the heterogeneous dynamics described above. The current theory treats the disordered state as dynamically uniform.

With Professor Lian Yu and his group (in the School of Pharmacy), we are studying crystal growth in thin films of organic glasses. We have found evidence for fast crystal growth at the surface.

 Another interesting feature of this surface crystal growth in the way in which the crystals grow more than 100 nm out of the surface. This striking behavior seems to require either that the crystal is pushed out of the glass or that molecules climb up the sides of the crystal to attach to the top.

PowerPoint presentation on Influence of heterogeneous dynamics on crystallization

PowerPoint presentation on Fundamental Studies of Crystallization of Organic Materials

	Dynamics near polymer interfaces.

    In many situations dynamics near a glassy polymer surface are important.  For example, polymer structures with dimensions less than 50 nm are projected to be routinely prepared by lithography within 6 years.  

Collapse behavior of resist structure

These small structures are weaker than would be expected based on the properties of larger polymer samples; this weakness will interfere with many possible applications.   Our current hypothesis is that this weakness results from dynamics near free surfaces being orders of magnitude faster than those away from the interface.  We use confocal microscopy to measure the dynamics of dye molecules near the surface of small polymer structures. The rotational and translational motion of the dyes can be quantitatively related to the polymer dynamics near the surface.

PowerPoint presentation on "Rapid molecular motion near the surfaces of polymeric and low molecular weight glass-formers"

PowerPoint presentation on "Segmental dynamics of thin polymer films probed by dye reorientation." (APS meeting, March 2008)