Mikko P. Haataja (PhD, McGill University) |
Welcome to the homepages of the COMBO
(COmputational Materials &
group! I am the leader of a group of computational and theoretical materials scientists, whose research focuses on
physics of materials and biophysics. The underlying theme of our research can be stated
as "mesoscale modeling of hard and soft matter systems". Our research activities in
hard matter systems are currently supported by
the National Science Foundation (NSF) and Department of Energy (DOE), while support for the soft matter activities are
provided by the NSF.
Current work includes
studies of microstructure formation during solid-solid phase transformations and
solidification, dislocation dynamics, mechanical properties of metallic glasses, and
compositional domain formation on surfaces for catalysis applications. More recently, my group has extended some of the
coarse-grained methods developed in hard matter systems to address important biophysical phenomena in soft matter
systems, such as structure and dynamics of compositional lipid microdomains in membranes. To give
you a flavor of my research interests, a brief description
of some of the problems I've been working on recently can be found below. The list of recent papers also
includes links to PDF versions of the published/submitted papers.
PRESENT & FUTURE RESEARCH ACTIVITIES
One problem of long-standing interest to me has to do
with extending the time scales of atomic scale simulations. To this end, several years ago,
we introduced a continuum "phase-field crystal model" capable of simulating crystalline phases with atomic
scale spatial resolution across diffusive time scales [for details, see K. R. Elder
et al., PRL 88, 245701 (2002)]. Together with Peter Stefanovic and
Nick Provatas, we have extended this method to include elastic
relaxation through the propagation of "quasi-phonons" [Stefanovic et al., PRL 95, 225504 (2006)]. More recently, we have studied
the formation of strain-induced compositional domains on metallic surfaces [Muralidharan and Haataja, Phys. Rev. Lett. (2010)].
The figures below illustrate the behavior of CoAg film on Ru(0001) surface. Left panel: symmetric composition. Middle panel:
Ag-rich composition. Note the coexistence between a nanoscale alloy and a dislocated pure Ag phase. Right panel: Co-rich composition.
Note the coexistence between a pseudomorphic alloy phase and Co-rich phase.
Additionally, we have investigated the growth of multilayer thin film systems using the PFC approach. Left panel: so-called herringbone
pattern emerges from the deposition of 1 ML of Ag on Ru(0001) substrate. Middle panel: Nucleation of islands on top of misfit dislocations
(shown in red circles) within ML1. Here, ML2 atoms (shown in blue) have a misfit relative to both Ag and Ru.
[In other words, we have a A/B/C-type multilayer system.]
Right panel: A/B/A-type growth. Here, Ru atoms comprising ML2 form elongated islands as directed by the herringbone patterns in ML1.
Dislocations in crystalline pure metals and alloys continue to be a
topic of national importance and interest, as the development of novel structural materials
requires the development of physically-based, effective constitutive equations at the
mesoscale. To this end, we have employed a parallel, three-dimensional level-set code to simulate the dynamics of isolated dislocation lines
and loops in an obstacle-rich environment. This system serves as a
convenient prototype of those in which extended, one-dimensional objects interact with obstacles and the
out-of-plane motion of these objects is key to understanding their pinning-depinning behavior. In contrast to earlier
models of dislocation motion, we incorporate long-ranged interactions among dislocation segments and obstacles to study the effect of
climb on dislocation dynamics in the presence of misfitting penetrable obstacles/solutes, as embodied in an effective climb mobility.
Our main observations are as follows. First, increasing climb mobility leads to more effective
pinning by the obstacles, implying increased strengthening. Second,
decreasing the range of interactions significantly reduces the effect of climb.
The dependence of the critical stress on obstacle concentration and misfit strength is also explored and
compared with existing models. In particular, our results are shown to be in
reasonable agreement with the Friedel-Suzuki theory. The figures below illustrate dislocation morphologies at low, intermediate, and
high external stresses, as well as the effective glide mobility extracted from the simulations. Note the extensive deformation of the
dislocation line at intermediate values of the external stress.
[Chen et al., Phys. Rev. B 81, 054104 (2010).]
We are also interested in developing a mechanistic understanding of the migration of low angle grain boundaries (LAGB) in terms
of the collective dynamics of the constituent dislocations. The image below left displays an LAGB of mixed character during steady-state
migration driven by external stress, while the figure on the right displays the steady-state migration velocity of the GB as a function of
external stress and dislocation climb mobility. [Lim, Srolovitz, Cai, and Haataja, in preparation.]
We have also recently investigated mechanical properties of metallic glasses and metallic glass-(nano)crystalline material
composites through simulations of a novel phase-field model. The figures below
illustrate the deformation of such a composite, in which the ductile particles are spherical. Left panel: representative microstructure. Middle panel: deformed shape and local
shear strain prior to shear band nucleation. Right panel: deformed shape and local
shear strain after shear band nucleation. [Abdeljawad and Haataja, Phys. Rev. Lett. (2010)]
Additionally, we have investigated the role of particle morphology on mechanical properties of BMG composites. The figures below
illustrate the deformation of a composite, in which the ductile particles are ramified. Left panel: representative microstructure.
Middle panel: deformed shape and local
shear strain prior to shear band nucleation. Note the existence of incipient shear
bands emanating from the tips of the dendrites, and constrained between dendrites.
Right panel: deformed shape and local shear strain after shear band nucleation and propagation.
[Abdeljawad et al., Applied Physics Letters (2011)]
We have also recently worked on developing physically-based models which explain morphologies of wurtzite nanoribbons.
In particular, such nanoribbons have been experimentally observed to spontaneously bend into nanorings or nanoarcs. Our theory suggests that bending
is controlled by dopant density, spontaneous polarization, and nanoribbon thickness. The figures below display the crystal
structure of ZnO and predicted nanoribbon curvature as a function of thickness H and dopant density [C. Majidi, Z. Chen, D. J. Srolovitz, and
M. Haataja, JMPS (2010)].
We are also interested in understanding and predicting the spatio-temporal evolution of
morphologies, deformation, and damage in multiphase, multicomponent materials
for energy conversion and storage under extended use. The figures below display continuum simulations of
stress evolution and damage accumulation due to redox reaction in a Solid Oxide Fuel Cell anode (Ni-YSZ).
In the figures below, Ni phase is shown in orange, YSZ in green, and pore phase in white. Leftmost panel: experimentally obtained
Ni-YSZ microstructure employed in the simulations. Panel 2nd from left: spatial distribution of accumulated damage during stress
evolution. Panel 2nd from right: two-dimensional cross-section of microstructure. Rightmost panel: regions of large tensile
stress within the YSZ phase are localized within highly constrained YSZ domains.
[F. Abdeljawad, G. J. Nelson, W. K. S. Chiu, and M. Haataja, work in progress].
We have recently launched a research effort in soft condensed matter systems.
More specifically, we are currently investigating (1) self-organization and
non-equilibrium dynamics in lipid bilayers, (2) molecular structure and properties of
lipid bilayers, and (3) self-assembly of detergents in bulk and on surfaces. The
first figure from the left shows a typical microstructure in a three component
lipid bilayer undergoing phase separation in the presence of non-equilibrium flux
of lipid to and from the monolayer in view. This "lipid recycling" leads to a
steady-state where the domains attain a characteristic size dependent on the
recycling rate. We argue that this recycling-induced steady-state
can explain the presence of the elusive
"lipid rafts" thought to exist in the mammalian plasma membrane [Fan et al., PRL 2008]. The middle frame demonstrates the
effects of spatially-varying recycling rate on domain formation. In particular, rapid changes in the local
recycling rate may lead to emergence of extended, transient rafts. Upon homogenizing the recycling rate,
the membrane microstructure becomes homogeneous, as shown in the panel on the right [Fan et al., PRE 2010].
We have also developed a comprehensive classification scheme to experimentally distinguish between competing theoretical
models for raft formation [Fan et al, PRL 2010]. The method is based on quantifying the spatio-temporal fluctuations of the
raft domains via speckle autocorrelation function and static structure factor; the combination of the two
provides the means to identify the operative raft formation mechanism in living cells. Below, typical microstructures
corresponding to the five existing theoretical models are displayed (I - critical fluctuations [Veatch et al., 2008]; II - pinning by
immobile membrane proteins [Yethiraj & Weisshaar, 2007]; III - stochastic recycling above critical point [Edidin, 1999]; IV - stochastic recycling below
the critical point [Fan, Sammalkorpi, and Haataja]; and V - coupling to lipid reservoir [Foret, 2005]).
The corresponding data for the temporal decay of fluctuations and static structure factor are shown below. Collecting data over a
broad range of wavenumbers allows one to experimentally distinguish between the raft formation scenarios.
We have also investigated the role of membrane and exterior solvent hydrodynamics on membrane critical dynamics [Haataja, PRE (2009)] and
coarsening dynamics [Fan, Han, and Haataja, J. Chem. Phys. (2010)]. The two snapshots from the left below illustrate the coupling between
compositional dynamics and membrane hydrodynamic flow fields, while the snapshot on the right displays the flow fields induced within
the solvent due to compositional membrane evolution; this flow field subsequently back-reacts with the membrane flow field and influences
the phase separation process. In particular, we have demonstrated that the 2D nature of the membrane flow fields coupled
to a 3D exterior fluid gives rise to novel coarsening behavior.
Finally, the leftmost figure displays a snapshot from
a simulation where micelle formation takes place. In addition to lipids and detergents,
we are interested in the behavior of synthetic molecular switches in biology. To this end,
we have constructed physically-based models for the tether length dependence of signal integration proteins, shown
schematically in the figure on the right [Van Valen, Haataja, and Phillips, Biophys. J. 96, 1275 (2009)].
 T. Han and M. Haataja,
"Compositional interface dynamics in symmetric and asymmetric planar lipid bilayer membranes",
submitted to Soft Matter (2012).