Research Projects
Below is a selection of projects that are representative of the type of research opportunities available in the Computational Materials Science REU. Projects can be found in three main research areas: surface science, biology, and nanoscience. All projects will involve collaboration, using cyberspace-enabled tools, with multiple groups at the four universities participating in the REU.
Computational surface science including thin-film growth and catalysis
T. Rahman, S. Stolbov, and A. Kara - UCF PhysicsComputational surface science uses a variety of approaches including molecular dynamics (MD), density-functional theory (DFT), and kinetic Monte-Carlo (KMC) to elucidate reactions and molecular transport at surfaces. We will offer projects on both thin-film growth and heterogeneous catalysis. Understanding and controlling thin-film growth processes is one of the grand challenge projects. The significance of the research is that it offers the possibility of materials by design. In the case of epitaxial growth, deposited atoms diffuse on the surface before being trapped or attached to others. The details of the morphological evolution of the film depend on the local surface geometry and electronic structure. Students will learn by using empirical potentials to elucidate the microscopic energetics of surface diffusion. The kinetics can be simulated using a combination of MD, molecular-statics (MS), and KMC to examine the evolution and growth of the film. We will focus on low Miller index surfaces of Cu, Ag, and Ir, targeting cases of homo- and hetero-epitaxy to make contact with experimental data. Activation energies will be calculated and used in standard KMC and our recently-developed self-learning KMC. Another focus area is heterogeneous catalysis. The emerging hydrogen economy has generated enormous interest in novel catalysts for hydrogen production, hydrogen oxidation, and oxygen reduction. We will use DFT simulation to predict the catalytic properties of nanostructured catalysts. We will focus on CO oxidation at Au, Pt, and Pt-Ru alloy nanoparticles on oxide surfaces, including ceria, titania, and silica. Recent experimental results from Prof. Beatriz Roldan-Cuenya at UCF have demonstrated strong dependence on the reducibility of the oxide support, although the mechanisms are not well understood. REU students will interact with Prof. Roldan-Cuenya and her students and postdocs through meetings and lab tours. The computational REU students will see firsthand the interaction of theory and experiment. The surface science group works closely together with a mix of undergraduate and graduate students.
Computer simulation studies of biomolecular machines: Chemomechanical coupling and bio-recognition
W. Yang - FSU Department of ChemistrySplendid biomolecular machines are functionally robust and efficient. For instance, in DNA repair system, enzymes are required to specifically locate damaged bases, which are subtly different from correct ones; in immunological defense system, T-cell are activated to kill alien species by recognizing trivially varied molecular fragments. From biophysical viewpoint, underlying physical problems can be defined as "chemomechanical coupling", which ensures multiple time-scale events coupled with efficient energy transduction, and "bio-recognition", which guarantees subtle chemical information to be transformed accurately. The bottleneck of understanding these processes is the missing link between biochemical/biophysical observations and static structures. Biomolecular simulation is an ideal approach to forge such gap. In our studies, free energy simulation, path sampling and QM/MM calculation are the major approaches taken in the Yang group; they are featured by multi-time-scale/ multi-length-scale treatments and parallel solutions. Currently, the Yang group has several novel techniques developed in each individual area described above, which have been implemented in the program CHARMM. So, participating students will be educated on performing molecular simulation in CHARMM in multiple time/length frameworks with QM calculation, MD/MC simulation, and continuum treatment as technical basis. Especially, students will be exposed to problem-oriented algorithm development in dealing with complex biomolecular systems. Data analysis and graphic treatment will be accessory parts of this training.
Multiscale algorithms for materials simulations
A. El-Azab - FSU Department of Scientific Computing and Materials Science ProgramM. Gunzburger - FSU Department of Scientific Computing
Attempts are now underway to achieve concurrent coupling of temporal and spatial scales in materials modeling in order to understand the impact of the macroscopic conditions on the mesoscale structure of materials; including defects and defect interactions, interfaces, and phase transitions. In this regard, the fundamental material description is of atomistic type, which is then either coarse-grained or coupled with a continuum material representation to enable the prediction of both atomic and continuum scale variables simultaneously. A number of methods emerged lately which have the potential to fulfill this promise, including coarse-grained molecular dynamics methods, bridging scale method, heterogeneous multiscale method, and the quasicontinuum (QC) approach. An effort is underway at FSU to investigate a class of these approaches from the mathematical and computational points of view. This effort focuses on the development of general atomistic-to-continuum connection strategies, based on methods such as QC but with an enriched statistical mechanical foundation of the coarse-grained atomic system. Incoming REU students will work closely with Drs. El-Azab and Gunzburger and their doctoral students on various mathematical and computational aspects of this topic. Specifically, the REU students will learn about the rationale behind reduced-order model development in material simulations, the mathematical formalisms of such models, and the computational techniques used to implement these models, such as nonlinear optimization, finite element methods, grid generation, and advanced visualization techniques. Realizing the highly complex level of these topics, the REU students' involvement in research will include one or more of the much simpler tasks of computer generation of crystals, preparing molecular dynamics and molecular statics runs used to illustrate the connections between atomic and continuum models, computer visualization, grid generation, to name some. The REU students will be introduced to the topic of multiscale modeling of materials through lectures by Drs. El-Azab and Gunzburger, which will be based on a graduate course on Multiscale Materials Modeling at FSU.
Simulation of materials microstructure under reactor environment
A. El-Azab - FSU Department of Scientific Computing and Materials Science ProgramNuclear energy is an important part of our nation's energy portfolio. The successful and safe operation of nuclear reactors relies on the availability of materials that withstand the aggressive environment of nuclear reactor cores. Intense neutron irradiation induces atomic scale damage in the core materials and alters their properties. At Florida State University, the Materials Theory Group is developing computational modeling solutions to this problem, ranging from atomic scale investigation of the nature of the induced materials defects and their evolution and changes in the materials microstructure. Incoming REU students will work closely with Dr. El-Azab and doctoral students on various mathematical and computational aspects of radiation effects in materials. Specifically, the REU students will learn how the reactor environment affects materials at the microscopic level and how the materials defects rearrange at the mesoscale in ways that lead to materials degradation. The students will also participate in research using molecular dynamics models of defect interactions in materials and phase field theory of microstructure changes.
Modeling the rheology and crystallization of entangled polymer melts
S. Shanbhag - FSU Department of Scientific ComputingSynthetic polymers are employed as key components in products as diverse as artificial organ implants, lubricants in hard disk drives, bullet-proof vests, drug delivery vehicles, and light weight composites for spacecrafts. Polyolefins, which includes polyethylene and polypropylene, constitute about half of the synthetic polymer produced worldwide on a volume basis. The Brownian motion of a polymer molecule at temperatures above the melting point governs its rheology and hence processibility. Similarly, the motion of a polymer molecule as it is cooled below its melting temperature dictates how it organizes into crystalline and amorphous domains, which, in turn, governs solid-state properties like toughness and strength. Several unresolved, and technologically important questions, that are simultaneously amenable to undergraduate research can be posed. For example, (i) what is the diffusivity of cyclic (or linear/branched) polymers in an environment of linear (or branched/cyclic) polymers? (ii) what are the rates of homogeneous nucleation in branched polyolefins? (iii) how can we describe the rheology of melts of single chain crystals? Depending on the problem chosen, the student will be exposed to either one, or several, of a suite of simulation methods that includes molecular dynamics, Brownian dynamics, lattice Monte Carlo, slip-link models, and mean field theory. The student will learn to design the simulation system, write scripts to piece together existing code, and interpret and visualize results. To enrich the research experience, the student will interact extensively with other graduate students/postdocs, and attend joint polymer research group meetings held weekly.
Molecular modeling of self-assembly and transport in amphiphilic systems
D. Kopelevich - UF Chemical EngineeringSurfactants (or amphiphiles) are molecules that contain both hydrophobic and hydrophilic segments. In aqueous solution, surfactants spontaneously self-assemble into a variety of microstructures that play a key role in numerous applications, ranging from nanostructured templates to drug delivery vehicles. In addition to their industrial uses, self-assembled structures of amphiphilic molecules, such as cell membranes, are building blocks for various biological systems. In the proposed research projects the students will investigate dynamics of self-assembly, transitions between different self- assembled structures, and transport of solute molecules within complex amphiphilic phases. These studies will be performed by molecular dynamics simulations using a software package such as GROMACS. Therefore, the students will be educated on modeling within the nanometer-nanosecond length- and time-scale region. In addition, the students will be exposed to research on multi-scale modeling of amphiphilic systems. The students will also learn to use molecular visualization tools such as VMD to monitor the simulation results.
Oxidation of metal catalyst surfaces
A. Asthagiri - UF Chemical EngineeringMetals are used as catalysts in numerous industrial applications, such as the catalytic converter to reduce the toxicity of the emissions from automobiles. To improve and design new catalysts we must have a fundamental atomic-level understanding of their reactive properties. While there have been many studies of reactive properties of metals, much of this work is not relevant to real-world catalysts that are used in oxygen-rich conditions. For example, operating internal combustion engines under oxygen-rich conditions can significantly enhance fuel efficiency and lower the emissions of hydrocarbons and CO. Unfortunately, however, lean combustion also generates high levels of NOx compounds. Under oxygen-rich conditions the metal surface becomes oxidized and a range of complex oxide phases form. This project will examine the atomic-level steps that lead to oxide formation on the metal catalyst surface using various simulation tools. The benefits of such simulations are that we can explicitly track the dynamics and complex interactions in the system and correlate atomic-level behavior with the reactive properties of the catalyst. The work is done in close collaboration with the experimental lab of Dr. Weaver at UF. The REU student will use quantum mechanical (QM) codes in our group to extract energetics and rates of particular atomic-level processes.
Complex metal boro-hydride hydrogen storage materials
V. Bhethanabotla - USF Chemical EngineeringAdvanced complex boro-hydrides that are light-weight, low-cost and have high hydrogen storage capacity are very desirable for on-board vehicular applications. Boro-hydride complexes as hydrogen storage materials have recently attracted great interest as they exceed the 2010 DOE target for hydrogen storage. These materials are the focus of this research. Designing efficient complex boro-hydride materials requires both experimental studies and modeling across different length and time scales. Working closely with Ph.D. student Mr. Pabitra Choudhury, the REU student will utilize a direct method lattice dynamics approach using ab initio force constants to calculate the phonon dispersion curves for several complex boro-hydrides to establish stability of the crystal structures at finite temperatures. He/she will use density functional theory (DFT) to calculate electronic structures and the ab initio force constants for the direct method lattice dynamics calculations. These computational simulations will be utilized in understanding the crystal structure, nature of bonding in complex boro-hydrides, and mechanistic studies on catalytic doping to improve the kinetics and reversibility, and the hydrogen dynamics, with the objective of lowering the decomposition temperature. Since a combined theoretical and experimental approach can better lead us to designing a suitable complex material for hydrogen storage, the REU student will work closely with experimentalists in the group who synthesize, dope and characterize these materials.
Quantum mechanical modeling of single molecule nanodevices
I. Oleynik - USF PhysicsThis project aims at a quantum mechanical investigation of electron and spin transport through single organic molecular devices as well as their relationship to the atomic and electronic properties. A REU student will study atomic and electronic properties of the molecular switch that exhibits hysteresis in the current-voltage (I-V) curve. This switching has been recently observed in a series of experiments but fundamental mechanisms of this unusual behavior are currently unknown. First-principles density- functional calculations (DFT) will be performed to investigate the atomic and electronic properties of bipyridyl-dinitro oligophenylene-ethylene dithiol (BPDN) molecular devices. Both neutral and positively and negatively charged molecules will be considered. The REU students will also screen several other classes of molecules to determine the existence of two different molecular conformations: one for a neutral molecule and another for a negative molecular ion. Based on this knowledge, two modes of hysteresis behavior: the polaron and exciton mechanisms of conductance hysteresis will be thoroughly investigated. In addition, phenomenon of the Coulomb blockade will also be considered. The major focus will be on establishing the structure- property relationship: how the chemical structure of different functional groups affects the switching behavior of these molecular switches.
Electronic properties of carbon nanotubes
L.M. Woods - USF PhysicsCarbon nanotubes have drawn much interest due to their unique structure and electronic properties as well as their potential applications in developing different devices, such as energy storage elements, functionalized structures, sensors, biocompatible agents, bearing devices, etc. Carbon nanotubes are seamless rolled graphite sheets, which can be metallic or semiconducting depending on the way the rolling is done. Of particular importance is the ability of nanotubes to interact with different materials. In fact, this ability is in the basis of the majority of envisioned devices involving nanotubes. In this study, the main focus will be to investigate the adsorption properties of carbon nanotubes to different DNA building blocks, such as adenine, guanine, cytosine, and thymine. This is motivated by the increasing role of carbon nanotubes in the biological fields. In addition, the role of different imperfections, such as deformations and defects, in the adsorption process will also be studied. The studies will be performed using state of the art density functional theory (DFT) methods, using computer codes such as VASP and Abinit, which are available at the USF computing facilities. The students will also learn how to model realistic nanostructured systems within DFT methods and extract necessary data from the simulations. In addition, undergraduates will be exposed to large scale computing systems and gather experience how to use them for simulating materials properties.
Thermal conductivity in complex oxides
P. Schelling - UCF PhysicsS. Phillpot - UF Materials Science and Engineering
In gas-turbines, the turbine blades are coated with an oxide to protect the metal blade from melting. The standard thermal barrier coating (TBC) material is yttria-stabilized zirconia (YSZ) due to its good strength and toughness and low thermal conductivity. However, there is significant interest in identifying new materials with even lower thermal conductivity. New materials with lower thermal conductivity will enable operation at higher temperatures and hence improved efficiency. Complex oxides are a potential new direction for improved TBC. Experiment and also simulation by Schelling and Phillpot have elucidated the mechanisms resulting in low thermal conductivity in YSZ and also fluorite-based pyrochlore materials for a wide range of compositions. We propose to elucidate heat transport in complex oxides based on the monoclinic zirconolite and hexagonal zirkelite structures. The prototype composition for zirkelite and zirconolite is CaZrTi2O7 and hence has a composition related to the pyrochlore A2B2O7. Other possible compositions exhibit these structures, including substitutions of Eu, Y, Th, Ce, Nb, and Mn. Using atomistic simulation, we will create property-composition maps that connect ion size and dynamical properties to the computed thermal conductivity. Computational study will performed using simple Buckingham potentials for the interactions, fit over a range of compositions. Density-functional theory (DFT) calculations will be used to provide data on lattice parameters and elastic constants where experimental data might be lacking. We will also elucidate the possibility of phase transitions and explore the issue of phase stability. The overall management of the project will involve Schelling at UCF providing any required first-principles results on structures and properties. The REU student will work with both groups to develop classical models and to perform the thermal conductivity calculations.
Other participating faculty
S. Sinnott - UF Materials Science and EngineeringHai-Ping Cheng - UF Physics
A. Bhattacharya - UCF Physics
A. Masunov - UCF Chemistry