Hello,
A yearly allocation for the LCRC cluster has been requested with the
following updated information:
Submitter/PI: Abelardo Ramirez-Hernandez
Project Name: nano-soft-assembly
Division: IME
Project title: Modeling Self Assembly in Nanostructured Soft Materials
Associated funding: PSE Strategic LDRD (FY2013-15)
Other Systems:
Science: The molecular design of functional self-assembling soft materials requires a profound understanding of the fundamental principles of the collective phenomena leading to self-assembly. The beauty and complexity of these materials resides in their capability to display hierarchical self-organization, from angstroms to microns. Thus, the description of these materials requires to span several length and time scales, to this end our work utilizes many different levels of description —atomistic detail, coarse-grained simulation, and continuum modeling— in order to probe assembly at those scales. Our efforts in this project will be focused on two important classes of soft matter systems: polymers and liquid crystals. We are developing a fundamental understanding of structure and dynamics during directed self-assembly in nanostructured, macromolecular and liquid crystalline materials through the concerted use of theory and simulations, tightly coupled to experimental
validation.
Project description: Self-assembly is an emergent property that arises as a consequence of collective phenomena occurring at a molecular level. In soft materials, it is common that such self-organization occur in a hierarchical way. The structural organization at different scales, from atomistic to mesoscopic levels, is controlled by the different interactions between the component molecular units. Such effective interactions can range from weak and isotropic, to strong and orientational-dependent potentials. The interplay of entropy (configurational and/or translational) and those interactions gives place to a rich array of phase behaviors. The understanding of the relationship between the molecular features and the associated self-assembly behavior provides critical knowledge for the design of new materials with tunable optical, rheological and structural properties. Our work utilizes many different levels of description going from atomistic detail to coarse-grained and co
ntinuum models, with the aim of capturing the relevant features at each of these scales. With these tools, we can probe assembly at length scales from nanometers to several microns.
Our efforts in this proposal will be focused on the self-assembly and dynamical behavior of two important classes of soft matter systems, polymers and liquid crystals. The proposal is organized into five thrusts: liquid crystals, biopolymers, polymer brushes, block polymers and software development.
Thrust 1: Liquid Crystals.
Liquid Crystals (LCs) form a distinctive class of materials with a measurable degree of orientational order, and anisotropic elastic and fluid properties. The interaction between LC molecules and a surface, can control the orientation of LC molecules at or near the surface to create local disoriented regions (topological defects) where the inherent orientations of mesophases are lost. Understanding the atomic-scale structure of these topological defects is essential for numerous purposes, from a fundamental point of view, such a LC defects closely resemble the defects encountered in cosmology and condensed matter physics, thus, their understanding can provide physical insights about their stability and dynamics. The LC topological defects can be experimentally observed using cross-polarized microscopy, but still very little is known about the atomistic structure due to their small size (less than 10 nm). A particularly interesting example, is the case of a LC droplet with per
pendicular orientation at the surface, with a single defect located in the center of the droplet. According to the continuum description of LC materials, this defect could be either a single ring or a point defect. Recent experiments have provided indirect evidence about the nanostructure of the core of the defect by templating the assembly of specific amphiphilic molecules into the core. The experimental observations support the formation of a nanometer-sized closed loop with approximate diameter of approx. 36 nm. While these observations indicate that the ring defect has lower free energy than point defect, we know that the presence of amphiphilic molecules might change the core structure of the LC defects from a point to a ring. Thus, this is still an open problem that detailed computational simulations can help to resolve.
During the FY 2018, we aim to employ atomistic molecular dynamics simulations to provide, for the first time, a molecular view of the core structure of the defects at the center of a LC droplet, with and without the presence of amphiphile molecules. Since the diameter of the defect is close to 30 nm, the LC droplet needs to be nearly 50 nm in diameter. We have already prepared the initial configurations of such a droplet, the system contains 179,389 molecules of 4-cyano-4'-pentylbiphenyl (5CB) with total number of 3,408,391 atoms. Such a huge atomistic system will require at least 500 nanoseconds of simulation to attain equilibrium, such calculations will make use of extensive computer resources. In addition to utilizing in-house codes, we use the Scalable Molecular Dynamics package (NAMD) to perform the simulations, this MD package has displayed near-optimal scaling in Blues, and we expect it will perform better in Bebop. The estimated amount of machine time required to comp
lete this ambitious project is expected to be a little more than 4,000,000 CPU hours.
Thrust 2: Biopolymers.
The aim of this thrust is the investigation of a particular type of polypeptides that plays causative roles in human diseases. Although these biopolymers do not share native structural or sequence homology, they exhibit a shared tendency to self-assemble into fibrillar structures called amyloid. These amyloidogenic polypeptides have been linked to type II diabetes, as well as a host of neurodegenerative diseases, including Alzheimer's disease and Huntington's disease. A common
hallmark of these illnesses is the accumulation of amyloid deposits in certain tissues in the body. Much effort has been spent to understand the mechanism of amyloid toxicity; this effort has shifted to early-stage amyloid aggregates due to substantial evidence suggesting that early-stage aggregates are toxic to cells and cause cellular damage at lipid membranes.
The study of the underlying kinetic mechanisms, and thermodynamics, of the aggregation process is critical in order to understand the role of these amyloidogenic polypeptides in disease progression and to develop effective therapeutics. Our group has strong expertise on utilizing molecular dynamics simulations combined with advanced sampling techniques to study protein folding and aggregation events. Recently, we have focused on human islet amyloid polypeptide (hIAPP), which is implicated in type II diabetes. We are interested in the study of the aggregation process of hIAPP molecules, from dimers to oligomers, through the use of atomistic molecular dynamics simulations (300k atoms total), utilizing advanced sampling methods. Given the extensive computer resources needed to compute the associated free energy landscapes, so far, only the aggregation of two hIAPP molecules to form a dimer has been performed. During the FY 2018, we aim to compute the minimum free energy pathways
, and free energy landscapes, associated to the supramolecular aggregation of hIAPP molecules into trimers and other higher-order oligomers. To address this, we will made use of the string method sampling technique in association with MD using GROMACS, which is powerful approach to compute such kinetic pathways. In order to do these calculations, as well as accompanying free energy calculations to extract the energetic barriers associated with the aggregation process, we anticipate the use of approximately 1,000,000 CPU hours.
Thrust 3: Polymer Brushes.
Polymer brushes are interesting because their responsiveness to environmental changes, such as temperature, stress, etc. Thus, by decorating surfaces with these brushes, macroscopic responses to specific environmental clues can be designed, for example, for controlling wettability in space and time. One of our goals is to understand the physics of polyelectrolyte brushes in the presence of multivalent salt solutions, both at- and out-of-equilibrium. FY2018 research plans will utilize the allocation on LCRC resources to run coarse-grained molecular dynamics simulations further exploring the physics of grafted polyelectrolyte brushes and the effects of added salt on the brush morphologies. With the success of our initial modeling of static properties, FY2018 modeling will focus on the effects of including explicit solvent molecules, as well as the examination of the rheological properties of polyelectrolyte brushes undergoing flow. We aim to look at confined nanochannels lin
ed with polyelectrolyte brushes, and how the structural aspects of these brushes influence electro-osmotic flow through these nanochannels. In continued collaboration with the Tirrell Lab at ANL/UC, we will use critical experimental details to inform our coarse-grained simulations. Simulations will employ a dissipative particle dynamics methodology using the code LAMMPS. For this work we request 1,500,000 CPU hours.
Thrust 4: Block Polymers.
Block polymers are promising materials for the large-scale fabrication of devices with nanoscale features such as high-density memory devices, computer chips and photovoltaic materials. Charged block polymers are a promising alternative for the next generation of batteries and photovoltaic devices, as they combine phase segregation and charge transport within domains sizes on the order of nanometers. One important study within this thrust, involves addressing the microscopic mechanisms of ion transport in charged block polymer systems, as well as a detailed description of the polymer-polymer interface, where important events, such as charge separation, occur. To this end, large-scale atomistic simulation of diblock copolymer systems will be performed. We will obtain morphologies from our theoretically-informed coarse-grained model (TICG) and backmap the atomistic detail. The simulations will then be properly relaxed and equilibrated before data collection takes place. These s
imulations will provide atomistic resolution into the interface of polymer morphologies, something classical coarse-grained and field-theory simulations are not able to do. Calculated physical properties will be compared to the amorphous systems. Further, these systems can be easily modified to study ion transport. One of the blocks can be functionalized with a methyl group, granting a positive charge to the polymer backbone. Then, a counterion can be placed and its dynamics, studied. We have found that Gromacs shows very good scaling for the atomistic model used to represent these systems, and it will be used to perform such calculations.
Another project within this thrust, involves the study of the kinetics of polymer healing and the effect of molecular weight, entanglements and polymer rigidity. For this study, we will use a microscopic molecular model of polymeric molecules that captures the effects of topological constraints. We will perform MD simulations to develop a fundamental understanding of the process of healing, including the effect of the molecular parameters on the time scale associated to this process. We will also use our coarse-grained approach of entangled block copolymers that includes soft intermolecular interactions and slip-springs, to improve our coarse-grained simulation approach for use in multi-scale studies of large scale, self-assembled multicomponent polymer systems. To address the large time and length scales associated to this study. We will made use of the GPU-nodes available on Blues to accelerate the calculations, making this study possible. For this thrust we request 1,500,
000 hours.
Thrust 5: Software Development.
The work of this thrust is to develop open source software part of SSAGES (Software Suite for Advanced General Ensemble Simulations) and COPSS (Continuum--Particle Simulation Software) which augment standard molecular dynamics and Monte-Carlo codes, permitting particle--field coupling and calculation of reaction coordinates, reactive pathways, and free energies. These suites are in constant development, and are part of the software distributed by the Midwest Integrated Center for Computational Materials (MICCoM), headquartered at Argonne National Laboratory. Most of the work in this thrust will be focused on optimizing the efficiency of our new developed codes, and Bebop is the ideal resource to perform this. For this thrust we request 2,000,000 hours.
Over the past year of LCRC supported research we have optimized the efficiency of our codes. For these projects, we will made use of NAMD, GROMACS, LAMMPS and
in-house codes using these software packages. We achieve excellent scaling, and make near-optimal use of LCRC resources.
Our recent work utilizing LCRC resources has contributed to the following publications. Several more are currently being prepared to be submitted
** Directed Self-Assembly of Nematic Liquid Crystals on Chemically Patterned Surfaces: Morphological States and Transitions
X. Li, J. C. Armas-Perez, J. A. Martinez-Gonzalez, X. Liu, H. Xie, C. Bishop, J. P. Hernandez-Ortiz, R. Zhang, J. J. de Pablo and P. F. Nealey
Soft Matter 12, 8595 (2016).
** Understanding Atomic-Scale Behavior of Liquid Crystals at Aqueous Interfaces
H. Ramezani-Dakhel, M. Sadati, M. Rahimi, A. Ram'irez-Hern'andez, B. Roux and J. J. de Pablo
J. Chem. Theory Comput. 13, 237 (2017).
** Segregation of Liquid Crystal Mixtures in Topological Defects
M. Rahimi, R. Zhang, H. Ramezani-Dakhel, A. Ram'irez-Hern'andez and J. J. de Pablo
Nat. Commun. 8, 15064 (2017).
** Directed Self-Assembly of Liquid Crystalline Blue-Phases into Ideal Single-Crystals
J A. Martinez-Gonzalez, X. Li, M. Sadati, Y. Zhou, R. Zhang, P. F. Nealey and J J. de Pablo
Nat. Commun. 8, 15854 (2017).
** Molecular Structure of Canonical Liquid Crystal Interfaces
M. Sadati, H. Ramezani-Dakhel, W. Bu, E. Sevgen, et al.
Journal of the American Chemical Society 139, 3841 (2017).
** Directed Self-Assembly of Colloidal Particles onto Nematic Liquid Crystalline Defects Engineered by Chemically Patterned Surfaces}
X. Li, J. C. Armas-Perez, J. P. Hernandez-Ortiz, C. G. Arges, et al.
ACS Nano 11, 6492 (2017).
** A Multi-chain Polymer Slip-Spring Model with Fluctuating Number of Entanglements:: Density Fluctuations, Confinement and Phase Separation
A. Ram'irez-Hern'andez, B. L. Peters, L. Schneider, M. Andreev, J. D. Schieber, M. M{"u}ller and J. J. de Pablo
J. Chem. Phys. 146, 014903 (2017).
** A Detailed Examination of Entangled Lamellae-Forming Block Copolymers
B. L. Peters, A. Ram'irez-Hern'andez, M. Andreev, M. Kr{"o}ger, M. M{"u}ller, J. D. Schieber and J. J. de Pablo
Macromolecules, In Press (2017).
** Multivalent Ions Induce Lateral Structural Inhomogeneities in Polyelectrolyte Brushes
J. Yu, N. E. Jackson, X. Xu, B. K. Brettman, M. Ruths, J. J. de Pablo and M. Tirrell
Sci. Adv., Under Review (2017).
** Multivalent Counterions Impede Lubrication of Polyelectrolyte Brushes
J. Yu, N. E. Jackson, X. Xu, M. Y. Kaufman, M. Ruths, J. J. de Pablo and M. Tirrell
Nat. Commun., Under Review (2017).
Industry partnership:
Project URL:
Current FY Hours Used: undetermined amount
New FY Requested allocation: 10000000
Q1: 2500000
Q2: 2500000
Q3: 2500000
Q4: 2500000
Justification: Below is the efficiency related to LAMMPS and NAMD on Bebop:
---- Table of NAMD efficiency ---- (for the atomistic model of LC)
nodes nanoseconds per day
1 1.7
2 4.8
4 5.8
6 7.7
10 10.0
20 15.87
40 24.39
60 26.31
---- Table of LAMMPS efficiency ---- (100,000 atoms)
nodes nanoseconds per day
1 4.85
2 8.99
4 16.23
8 25.78
16 36.54
32 52.00
Storage requirements: we require 4TB. The reason for this is because much of our simulations will involve the calculation of free energies and kinectic pathways, where several copies of large systems (involving millions of atoms) will be saved for post-processing analysis.
Thank You,
The LCRC Accounts System
