Hello,
A change in allocation has been requested:
Requester: bair (Ray Bair)
Project: nano-soft-assembly
Title: Modeling Self Assembly in Nanostructured Soft Materials
Description: The Blues cluster at Argonne's Laboratory Computing Resource Center is ideal for performing mid-size simulations of self assembly. We will use these simulations both to supplement work we have proposed as part of the ALCF INCITE program, and to produce independent results and publications. We request 499,999 hours to be used in fiscal year 2014. We anticipate use of Blues by 10-12 group members, running primarily parallel jobs.
Theory and simulation are indispensable in targeting macromolecular self-assembly. Computational experiments are often able to examine relevant parameters more quickly, cheaply, and completely than those in a traditional laboratory. Experiment and simulation thus exist in a symbiosis where predictions are made, tested, and verified. Our group's work examines self-assembly in four contexts: block copolymer self-assembly, liquid crystal-based materials, glassy materials, and polypeptide aggregation and folding. In each, we are developing new models and techniques capable of probing exciting new phenomena. Our work utilizes many different levels of description---atomistic detail, coarse-grained simulation, and continuum modeling---in order to probe assembly at length scales from nanometers to several microns. In
conjunction with experiment, these become a powerful tool for probing and explaining material behavior.
To date, our work utilizing LCRC resources (Fusion) has contributed to the following publications:
``Modeling the polydomain-monodomain transition of liquid crystal elastomers,'' J. K. Whitmer, Raj Shekhar, T. F. Roberts, N. L. Abbott and
J. J. de Pablo. Phys. Rev. E 87, 020502(R) (2013).
``Liquid-Crystal Mediated Nanoparticle Interactions and Gel Formation,'' J. K. Whitmer, A. A. Joshi, T. F. Roberts and J. J. de Pablo. Journal of Chemical Physics, J. Chem. Phys. 138, 194903 (2013).
``Particle--Defect interactions in Liquid-Crystal Nanodroplets,'' J. K. Whitmer, J. C. Armas-Perez, A. A. Joshi, T. F. Roberts and J. J. de Pablo. in review at Phys. Rev. Lett. (2013).
``Measuring Liquid Crystal Elastic Constants with Free Energy Perturbations,'' A. A. Joshi, J. K. Whitmer, O. Guzman and J. J. de Pablo. in review at Soft Matter (2013).
``Adsorption of Ions in Non-polarizable Water Models,'' J. K. Whitmer, A. A. Joshi and J. J. de Pablo. in review at Langmuir (2013).
``Model Vapor-Deposited Glasses: Growth Front and Composition Effects'', I. Lyubimov, M. D. Ediger and J. J. de Pablo in review at J. Chem. Phys. (2013).
C.-C. Liu, A. Ramirez-Hernandez, E. Han, G. S. W. Craig, Y. Tada, H. Yoshida, H. Kang, S. Ji, P. Gopalan, J. J. de Pablo and P. F. Nealey, Macromolecules 46 (4), 1415-1424 (2013).
S. Xiong, D. R. Dunphy, D. C. Wilkinson, Z. Jiang, J. Strzalka, J. Wang, Y. Su, J. J. de Pablo and C. J. Brinker, Nano Letters 13 (3), 1041-1046 (2013).
LCRC resources were also used in testing of many of the new methods for block copolymer assembly described below. Renewal of our project on Blues will enable continuing work on these topics.
In block polymer work, we are developing and testing new methodologies for the prediction of self-assembling morphologies using theoretically-informed coarse-grain (TICG) methods. Block copolymer lithography is considered as an alternative to photolithography for the manufacturing of next generation electronic circuits. Using molecular simulations, we are trying to predict the block copolymer self assembly in presence of various guiding fields like chemoepitaxy, graphoepitaxy, and flow. These systems show that within block copolymer systems, highly ordered states may be achieved with an
exceptionally low area density of defects. For semiconductor manufacturing, we aim to introduce ever-smaller features, which can bring an increased tendency to form defects. The presence of any such defects can ruin a lithographic process, thus it is preferred to eliminate them entirely. To do so, we must understand their formation and relaxation processes. In these non-equilibrium process, the history of preparation and full dynamics is important. To treat these situations, TICG methods are being extended to include hydrodynamic interactions and entangled polymer dynamics. These simulations enable selection of optimal process conditions for defect free self assembly amenable to industrial application.
Our methodologies at equilibrium are being extended in order to treat polydisperse polymer blocks, and polymer swelling by monomeric solvent
molecules. We are especially interested solvent annealing processes, where a block copolymer film is assembled in the presence of solvent vapor. Particular morphologies are achieved which are a function of the solvent type and vapor
pressure. The subsequent process of solvent evaporation further affects the morphology. Theoretical understanding of this process is very limited. Our studies involve Monte Carlo simulations of block copolymer using augmented
TICG methods. Our model is computationally inexpensive and has capability of being executed on parallel machines, thus giving us the ability to access large length and time scales pertinent to the self assembly process. Coupling these additional complexities to dynamic complexities will permit extremely accurate structural predictions which may be experimentally verified.
We are also interested in investigating the conditions under which a myriad of droplets with a well-defined morpology may be placed with long-range order onto a substrate. For these systems, a promising solution is to use chemically
patterned surfaces defined by regions of differing wettability. This increases
the number of parameters which must be explored in determining morphology diagrams, which therefore require extensive computational work to map. Work related to block copolymer research, on the formation of equilibrium coacervate phases, is currently in progress. These solutions of oppositely charged polyelectrolytes (often polypeptides) exhibit a poorly understood liquid--liquid phase separation which results in the formation of two immiscible phases, each of which is primarily water. Simulating the phase coexistence of polyelectrolyte rich and poor regions is the first step toward understanding the phase behavior and physical properties of the coacervate phase. The phase diagram is determined by various parameters, such as the length of the
polyelectrolytes, the charge density along the chain, the amount of salts present. The goal of the project is to use grand-canonical Monte
Carlo simulations to establish the coexistence curve under various polymerization, pH, and salt conditions and examine the effects of these controlling parameters. The results will be closely compared with predictions of the standard model for describing the formation of the
coacervate phase, the Voorn-Overbeek model, and be used to test the limitation of the model. The simulations envisioned is highly coarse-grained, includes multiple species, treats long-range
interactions explicitly, and needs to explore a large window of parameter space. The phase separation itself is interesting, and
exhibits many subtle nuances, sensitive to the molecular character of the polyelectrolytes---for instance, the stereochemistry of monomeric units in a polypeptide strongly influences coacervation behavior. We are examining the clustering of oppositely charged molecules under the
influence of salt in three regimes: at the continuum TICG level, with electrostatic interactions explicitly resolved alongside standard polymeric forces; at the coarse-grained ``blob'' level using charged bead-spring chains capable of describing most of the assembly physics; and at atomistic level, where we probe the influence of molecular conformations on the aggregation of individual polypeptides.
One of the main research focuses within our group is understanding the principles underlying formation of so called stable glasses, materials
with enhanced thermodynamic and mechanical properties. Experiment and simulation have demonstrated that the slow relaxations permitted in vapor deposited glasses imbue them with extraordinary stability, similar to glasses which have been aged for thousands of years. We are
performing a series of coarse-grained molecular dynamic simulations to study model glass formers using the LAMMPS package. In our simulations, we prepared binary Lennard Jones glasses by a process
that mimics physical vapor deposition on a substrate. As in experiment, our simulations are able to demonstrate mechanical probabilities of glasses aged orders of magnitude longer than the
deposition process. We now aim to understand what structural features are present in these glasses, and how these structures relate to equilibrium solids.
Liquid Crystals (LCs) are being explored as novel detectors for toxic bio-molecules, an example being bacterial endotoxin. Endotoxin is a
lipopolysaccharide present in the outer membrane of Gram-negative bacteria. The lipid-A portion of endotoxin is responsible for its toxic effects within human body and can produce septic shock even with very limited exposure to endotoxin molecules. Current methods for determining the presence of endotoxin involve coagulation of blood from horse-shoe crabs, an expensive,
environmentally damaging, and unreliable process. Recent experimental studies have shown that when LCs are confined to spherical geometry (droplets), their behavior may be drastically changed. Consequently, a wide range of new
morphologies are presented, opening the possibility of new promising technological applications. Experiments have shown that LC droplets are sensitive to endotoxin at concentrations as low as one picogram per mililiter, provoking a phase transition on the droplets which allows the use of LC droplets as a biosensors. The exact mechanism by which endotoxin molecules cause this morphology change is not clear. Hence, we are performing molecular
simulation to study this in detail. This involves both the study of how lipid-A affects ordering of the nematic liquid crystal at the water--oil
interface, and examination of the processes which lead to morphology transitions within LC droplets. This work will enable manufacture of
droplet-sensors which are balanced on a knife-edge, ready to transform upon exposure to single endotoxin molecules.
We are also examining extensions of TICG formalism to treat nematic fields using a continuum, Q-tensor-based Landau--de Gennes free energy expansion. The resulting functional is minimized by discretize the field of Q-tensors onto a mesh of points. In our methods, we explicitly consider the energy cost of
point fluctuations, resulting in a spin-like model which allows Monte-Carlo simulation of the nematic field. Allowing fluctuations in this way prevents a liquid-crystalline system from being stuck in local free energy minima, and
thus permits prediction of defect and morphology structures occurring in experiment---minimization to local (rather than global) minima plagues
standard continuum methods. This work is being used to study the effect of confinement on LCs. The goal of this research is to explain the effect and the generation of new morphologies by changing various parameters such as: the
size of the system, anchoring strength, anchoring type, mechanical constants. The other properties to be considered are chirality and the applied
electric or magnetic field. Results of my research shall have a direct application primarily for the generation of new nanodevices. This work further allows simple extensions to mobile particles embedded in a liquid crystalline
phase, whose directionally-dependent interactions are useful for self-assembling explicit structures. Coarse-grained simulations and
experiments have shown a distinct tendency for nanoparticles within the nematic phase to aggregate, and for nanoparticles in droplets to seek the defects within the region. Coupling these two effects will lead to robustly
assembling patchy particles which can be used in further materials synthesis. Further, these aggregates may be themselves used to prime a
morphological transition within a liquid crystal droplet sensor. Coupling nematic and smectic ordering to block-copolymer assembly, in order to create thick films with sharply defined features, is also of interest.
Finally, we intend to use Blues to devise and implement new free energy computation strategies. Metadynamics is a recently developed technique which has found broad application for driving phase transitions and obtaining free energy surfaces for molecular systems as a function of a small subset of coordinates (order parameters). This enables the identification of transition mechanisms, and the configurations leading to morphology or phase transitions. Its use relies on the convergence of a free-energetic estimate to the true free energy surface through systematic addition of biasing potentials
to create a ``flat histogram'', where all order parameter configurations are equally likely. Several issues exist with the process of convergence. Importantly, convergence in many systems is slow, and the free energy surface may only be known to a limited degree of accuracy. Further, the presence of rigid bounds on an order
parameter leads to systematic oversampling for which only approximate correction methods exist. Our group has developed robust strategies,
such as flux-tempering, for metadynamics convergence, and new convergence-criterion-based boundary correction methods. Alongside
these, we are developing new free energetic methods which are self-adapting, in order to restructure sampling on the fly and speed
convergence by preventing oversampling.
Current: undetermined amount
Justification:
Requested: 125000
A specific reason has been given:
we are running feverishly the prototype calculations that we’ll use for the BES center proposal, and these and other deadlines have pushed us into a corner, so we need to increase out allocation for Q4.
This needs to be approved and the final allocation amount decided upon.
Thank You,
The LCRC Accounts System
