Department of Chemistry
University of Oregon
Polymer liquids are ubiquitous materials. Many polymers, such as proteins, cellulose, silk, rubber, DNA, and cellular filaments are found in nature. Many others such as elastomers, fibers, plastics, polymer liquid crystals, conducting polymers, glasses and composite materials are produced synthetically. Understanding the dynamics of polymer fluids is a field of interest for biophysicists (e.g., solving the mechanisms of protein folding) as well as engineers (e.g., understanding the freezing of the dynamics at the glass transition). The theoretical tools to investigate these systems include condensed matter theory, statistical mechanics, and advanced computer simulations.
Because of the large number of relevant variables that characterizes these highly complex systems, in many cases only purely phenomenological theoretical approaches are available. Nevertheless, it is important to develop microscopic theoretical approaches that correlate the local chemical structure with the macroscopic properties (e.g., to predict the glass transition temperature of a newly synthesized polymer material). The achievement of such a microscopic understanding is a key step in the development of tailored synthesis of new materials with specific technological and/or biological properties.
In this presentation I show how a conventional microscopic approach for the dynamics of polymer fluids can be extended to treat systems of increasing complexity. An analysis of computer simulations shows the underlying physical picture missing from the conventional approach, and suggests a way to develop a new theory of polymer fluid dynamics using traditional statistical mechanics (1).
- M. G. Guenza, Cooperative Dynamics in Unentangled Polymer Fluids, Phys.Rev.Lett. 88, 025901-1 (2002).
Theory and Simulations of Multicomponent Polymeric Materials: Recent Advances and Open Challenges
The morphological self-assembly patterns exhibited by multicomponent
polymers and complex fluids have attracted considerable interest in the
recent years. The possibility of achieving well defined long-range order
in these systems, and on length scales continuously tunable by tuning the
chemistry of the molecule, has opened up a number of avenues for applications.
One of the main challenges confronting success in the realization of
self-assembly applications is to achieve an ability to tune the resulting
morphologies of self-assembly in a controllable manner in such systems.
Indeed, a number of experiments have unequivocally demonstrated that the
macroscopic properties and response of these self-assembled systems is
crucially dependent on the morphology of self-assembly. Consequently, an
understanding and an ability to tune the morphology of self-assembly is
expected to play an important role in complementing the synthesis aspect.
However, a purely experimentally based program in pursuit of the above
objective proves complex --- primarily, due to the broad parameter space
accessible to synthetic chemists. Consequently, there has been a significant
effort in the development of computational tools that can enable a rational
exploration of the parameter space.
Atomistically faithful computer simulations of self-assembly in dense
phases of soft materials prove to be difficult or impossible for many
systems of practical interest. This talk will focus on some of the recent
developments in computational tools for predicting the self-assembly and
thermodynamics of polymeric and complex fluid materials, which include
coarse-grained particle-based simulation approaches
and field-theory based simulation approaches. I will specifically
focus on the advantages and the disadvantages of the two approaches and
illustrate their utility explicitly through results pertaining to the
thermodynamics of different complex fluids. In addition, I will also
indicate a few outstanding challenges, viz. in the context of systems
involving short-range repulsions or in situations wherein information on the
dynamics of complex fluids is desired. These features render the exclusive
use of one or the other technique cumbersome – thereby necessitating
a hybrid approach. I will also expound upon some of the recent developments
in such contexts. (a) (b) (a) Field-theoretic and (b) Particle-based simulations of the lamellar phase of a diblock copolymer. References: 1. "Field-Theoretic Computer Simulation Methods for Polymers and Complex Fluids," G. H. Fredrickson, V. Ganesan and F. Drolet, Macromolecules, 35, 16 (2002). 2. "Multiscale problems in polymer science: Simulation approaches," K. Kremer and F. Muller-Plathe, MRS Bull., 26, 205 (2001).
Progress in Theoretical Understanding of High Temperature Superconductivity
In this talk, I shall review recent progress in our theoretical
understanding of high Tc superconductivity and other strongly correlated
materials. In particular, I shall discuss the global phase diagram of
high Tc cuprates in terms of effective field theories, in
relation to recent experiments.
Exploring the Complexity of Colossal Magnetoresistant Manganites and
High-Tc Cuprates using Computational Techniques
The study of Strongly Correlated Electrons is among the most active
and exciting areas of research in
Condensed Matter Physics. Several promising materials of the correlated
electron family have been synthesized
and investigated. These compounds include the manganese oxides -with
resistivities that change by enormous factors upon the application of small
fields- and the copper oxides -with high superconducting critical temperatures.
In the last couple of years, a qualitatively new phenomenon common
to these compounds and others has been unveiled. They present slowly
fluctuating nanoscale inhomogeneities driven by interactions at the microscopic
level.
These inhomogeneities translate into coexisting small nanometer scale
regions (clusters), with properties that change from cluster to cluster
such as their properties to conduct electricity. Phase competition is at
the heart of cluster formation, as illustrated in the figure. The understanding
of these effects is only at the early stages and a huge effort worldwide
is currently under way to investigate further this phenomenon. This may
originate in a new type of quantum-mechanical ordering driven by the competition
between spin exchange, kinetic energy, lattice effects, and long-range
Coulomb interactions. Computer simulations have been at the forefront of
theoretical investigations
in this context, providing new ways of learning about these complex
materials [1].
[1] A. Moreo et al., Science 283, 2034 (1999); E. Dagotto, T. Hotta,
and A. Moreo, Physics Reports 344, 1 (2001);
Competing, Coexisting, Hidden and Quasi-Orderings of
Strongly Correlated Electrons: Insights from Models
In recent years it has become increasingly clear that high-Tc
copper-oxide materials, layered organic compounds, and other
strongly-correlated electron systems exhibit a rich variety of different
orders. Exotic
superconductivity, antiferromagnetism, charge segregation such as stripes,
and non-Landau Fermi liquids are just some of the phases believed to
appear. Two approaches to deal with complex forms of order have been
advocated: The first is basically to admit defeat, give up on any attempt
to link microscopic models to the observed physics, and focus instead on
phenomenology. This view of condensed matter physics is sometimes
compared to that of biology, as there is no attempt to gain a
first-principles understanding. The second approach, the focus of this
talk, is to solve various microscopic models of strongly-interacting
electrons with the goal of discerning phases that arise over significant
regions of parameter space. A combination of systematic analytic and
numerical methods can yield reliable phase diagrams in many instances, and
a wealth of different orders typically appear. I will illustrate the
approach by focusing on the two-leg ladder systems. The rich phase
diagram contains competing, coexisting, and hidden forms of order, as well
quasi long-range ordering tendencies. Of paramount importance to gaining
insight into real materials is the need to search through high-dimensional
parameter spaces.
Undergraduate Education in Biological Physics at the University of Pennsylvania
The University of Pennsylvania has a strong tradition of
interdisciplinary research, but not so much of interdisciplinary
undergraduate course material. So it came as a surprise when we
offered a course on Biological Physics and found undergraduates
enrolling from eight different majors (physics, biology, biochemistry,
biophysics, materials science, and chemical, mechanical, and
bio-engineering). Evidently undergraduate interest in this subject is
strong. Even though many related courses exist in other departments,
still many students are interested in the physics point of view.
Figuring out what belongs in such a one-semester course, and how to
present it as a coherent story, was a big challenge. I'll describe
what we came up with, along with some other ideas about how to
construct the course.
The whole exercise has led me to write a textbook, with support from
NSF's DUE:
http://www.physics.upenn.edu/~biophys/frontmatter.pdf
Improved Electrostatics for
Large-Scale Biomoleculer Simulations
Reliable biomolecular simulations are extremely difficult, because they
involve complex systems such as macromolecules bathed in solvent,
long-range electrostatic interactions, correlation effects, and a high
sensitivity to both temperature and dynamical effects.
During the past decade, large-scale codes such as AMBER and CHARMM, which are
based on empirical, classical molecular potentials have been the mainstay
of biomolecular simulations. Although much has been learned, there is a clear
understanding that such codes have their limitations. As a consequence, new
multiscale methods that aim at improving on these codes are under development
by many groups. These methods attempt to handle different parts of the
biomolecular system at the appropriate computational level, ranging all the
way from reference-quality ab initio quantum calculations,
to first-principles “density-functional” molecular dynamics,
classical atomistic mechanics, and continuum methods.
In particular, an accurate and numerically efficient treatment of the
classical electro-static force fields is absolutely essential for meaningful
biomolecular simulations, and for a smooth interface to quantum codes. Why is
this? Classical codes such as AMBER and CHARMM assign “partial charges”
to virtually every atom in a simulation in order to model the interatomic
potentials, so that these represent the most important non-bonded interactions
for biomolecules. There are therefore two main problems associated with the
treatment of classical electrostatics: (i) how does one eliminate
artifacts associated with the point charges in any classical model, and
thereby improve the electrostatic poten-tials in a physically meaninful way?;
(ii) how does one efficiently simulate the very costly long-range
electrostatic interactions? In this talk, we will address these issues
and present some of our recent solutions to some of these challenges.
Frustration and
Fractionalization in Quantum Magnets
Magnets with competing interactions are a richly varied class
of systems which represent a substantial and continuing challenge
in themselves. They are also of special interest as it has
long been suspected that they can support phases with fractionalized
excitations, i.e. those with quantum numbers different from the
fundamental constituents, above one dimension. The canonical example
of this is the long sought "resonating valence bond" (RVB) state
conjectured by Anderson.
I will review a body of recent work that is substantial progress
on both fronts in interesting, interlinked ways. This involves
results on "order by disorder" and "disorder by disorder" in
frustrated quantum Ising models, RVB phases in quantum dimer models,
duality relations between the Ising and dimer models, the equivalence
of quantum dimer models to Ising gauge theories and the characterization
of the RVB state as a topologically ordered state.
Liquid-crystals, defined as states of matter intermediate
in their properties between fully disordered isotropic liquids
and fully ordered crystals, are ubiquitous in nature. Recent
transport measurements[1] on two-dimensional electron systems in
moderate magnetic fields suggest the existence of a spontaneously
orientationally-ordered, compressible liquid state. I will
discuss a number of electronic liquid-crystal phases[2-4] as
candidates for the experimentally observed anisotropic state.
First-Principles Calculations in Materials Physics:
Successes and Challenges
During the last decade, methodologies for performing "first-principles"
calculations have largely matured and been optimized on our modern computers.
The most widely used model involves implementing density functional theory
with an approximation at the local or gradient level, and utilizing the
pseudopotential method with a plane-wave basis set. Various materials
properties including phase stability, phonon frequencies, atomic diffusion,
electronic structure, etc. can be reliably predicted. Ab initio molecular
dynamics simulations have thus become feasible. Accurate materials parameters
can be extracted for further model calculations of collective properties
such as the superconducting transition temperature. These calculations have
also started making impacts in the chemical, pharmaceutical, and biotechnology
industries. In this talk, I will give an overview of these calculations and
their successes. I will also discuss our recent work on modeling the quantum
size effects in two-dimensional nanostructures [1,2].
The two challenges we are facing are: (1) how to make these calculations
more efficient without losing much of the accuracy, and (2) how to make them
more accurate without significantly reducing the efficiency. A lot of efforts
have been devoted to the first challenge in recent years. The second one
requires reexamination of the approximations that have been included. All
modern density functional calculations depend on models of the exchange and
correlation energy functional, yet the exact form for an inhomogeneous system
is completely unknown. We have used coupling-constant integration and
variational quantum Monte Carlo techniques to calculate the quantities of
central importance in density functional theory for an anisotropic silicon
crystal. By comparing the "exact" exchange-correlation energy density for
a prototype semiconductor with results obtained from various approximations,
we hope to identify directions for improved implementation of density
functional theory [3-5].
Mysteries of the Amorphous State: Jamming, Plasticity
and Localization in the Absence of Long-range Order
Non-crystalline solids exist as metals, ceramics, semiconductors and
polymers and find wide application in industry. Yet the atomic-level
mechanisms that control the mechanical properties of non-crystals are poorly
understood compared to the comparable mechanisms in crystals. The disordered
nature of the amorphous state makes the identification of these mechanisms
challenging, but this same disorder may permit a statistical treatment of the
material that would be impossible in the context of crystals.
The work I will describe uses molecular dynamics (MD) simulation to
extract the salient features of the deformation process. These simulations
are particularly timely because they interface with a number of recent
advances in metallurgy, microscopy and theories of the glass transition.
Recent advances in the production of bulk metallic glasses have sparked
interest in using these materials for structural applications. However, the
dominant failure mode of these materials is the spontaneous formation of
severe plastic localization in the absence of thermal softening. MD
simulations can potentially aid in identifying the precise mechanism of
this localization process. At the same time new fluctuation microscopy
methods are being developed which provide insight into the structure of
non-crystalline materials. These methods are able to quantitatively
characterize amorphous materials. The structural interpretation of these
microscopic signatures can also be greatly aided by computer simulation.
Simulations of the glass transition have recently revealed the onset of
significant dynamical heterogeneity in the supercooled liquid. Computer
simulations are helping to reveal the relationship between this aspect of
liquid structure and the mechanical response of the quenched glass under
shear stress [1].
Research along these lines has already revealed a number of interesting
aspects of the plastic dynamics of non-network forming glasses. The
regions of rearrangement are both local and directional. This provides
some natural constraints on their incorporation in a constitutive theory for
the material [2]. The facture behavior is very sensitive to small changes in
the inter-atomic potential in ways that can be understood via these atomic
scale mechanisms [3]. Frictional contact of these materials results
in significant mechanical mixing, stress assisted diffusion and the formation
of a mechanically worked layer near the surface that may control
tribological properties [4]. In addition the mechanism of deformation
in a localizing region may be significantly different than that in
the bulk, involving nano-cavitation and the diffusion of the resulting excess
volume to the surface. The details of this transport process are currently
under investigation.
The Role of Quantum Mechanics in "Virtual Aluminum Castings"
Increasing demands to further reduce emissions and simultaneously
improve fuel economy in automobiles has expanded the need for
lightweight materials (such as Al, Mg, and their alloys). In
order to optimize alloy design and processing conditions to
quickly achieve Al-alloy castings with suitable mechanical
properties, researchers at Ford Research Laboratory are developing
the Virtual Aluminum Castings methodology: a suite of predictive
computational tools that span length scales from atomistic to
macroscopic to describe alloy microstructure, precipitation,
solidification, and ultimately, mechanical properties.
The role of first-principles atomistic computations in the Virtual
Aluminum Castings methodology will be described, as will the connection
between these atomistic methods and other computational approaches
(phase-field microstructural models, computational thermodynamics
methods, cluster expansion methods, etc.). Because of their highly
accurate and predictive nature, there is a growing desire to use
these types of theoretical approaches to predict properties of new,
experimentally unexplored, or difficult-to-synthesize solids.
Application to problems of precipitation, thermal growth, and
microstructure evolution during heat treatment has proved very fruitful.
Combining these quantum-mechanical results with other modeling and
experimental efforts, one can suggest heat treatments which
optimize thermal stability and hardness of industrial alloys.
From Coherence and Strong Correlations to
Entanglement and Quantum Control: Exploring A New
Interface Between AMO and Condensed Matter Physics
Recent experimental and theoretical developments open up several
emerging sub-fields on the boundary of traditional disciplines of
condensed matter physics and quantum optics/atomic physics.
This talk will describe several ideas from this emerging interface
and emphasize their relationship to other areas such
as experimental quantum information science.
In particular, we will discuss how recently demonstrated
techniques for controlling cold atoms in optical lattices could be used
for "engineering" quantum phases and quantum correlated atomic states.
Specifically we will show that ultra-cold atoms could be used to probe
fundamental problems in many-body physics such as the origin of high
temperature superconductivity. Similar techniques can be used to
accurately study the effects of decoherence and dissipation on quantum
phase transitions and to "engineer" exotic many-body states.
We will also show that small ensembles of strongly interacting atoms could
display mesoscopic phenomena that resemble similar effects in
nanoscale solid-state devices. In particular, we will discuss how the
controlled optical excitation of atoms into the Rydberg states can result
in a so-called "dipole blockade". This effect can be
used for generation of non-trivial collective atomic states as well
as non-classical photonic states.
Ideas from quantum optics that can be applied for coherent manipulations
of mesoscopic condensed matter systems will be discussed as an
outlook.
What Do Quantum Gases Do For Materials Theory?
The field of cold atoms has become highly interdisciplinary. It
brings together
researchers from condensed matter physics, atomic physics, quantum optics,
nuclear physics, and quantum information.
In this talk, we shall discuss a number of fundamental issues in cold atoms
which are particularly relevant for condensed matter physics.
We shall begin with an overview on the recent experimental and theoretical
developments of the field, and the hot topics today. We shall explain the
importance of these topics and how they relate to the deep issues in quantum
many-body theory.
We shall in particular discuss the following topics which are under intense
experimental and theoretical research today:
Department of Chemistry
University of Texas at Austin
Department of Physics
Stanford University
NHMFL-Tallahassee
Florida State University
J. Burgy et al., Phys. Rev. Lett. 87, 277202 (2001); M. Moraghebi,
S. Yunoki and A. Moreo, Phys. Rev. Lett. 88, 187001 (2002) ; and
references therein.
Phase diagram of a model used by Burgy et al. (see ref.[1])
to analyze the competition between ordered states. The red and blue points
denote results in the presence of quenched disorder. The insets show phase
competition in real manganites and cuprates. Details will be discussed
in the presentation.
Department of Physics
Brown University
Department of Physics and Astronomy
University of Pennsylvania
Additional course materials are here:
http://www.physics.upenn.edu/~pcn/Course/280.html
Department of Physics
North Carolina State University
Department of Physics
Princeton University
Department of Physics
University of Colorado Boulder
School of Physics
Georgia Institute of Technology
Department of Materials Science and Engineering
University of Michigan
Ford Research Laboratory
Dearborn, MI
Department of Physics
Harvard University
Department of Physics
The Ohio State University
Last modified August 29, 2002
Daryl Hess, hess@physics.georgetown.edu
Amy Liu, liu@physics.georgetown.edu