Bulletin of the American Physical Society
2006 APS March Meeting
Monday–Friday, March 13–17, 2006; Baltimore, MD
Session B32: Focus Session: Computational Nanoscience I |
Hide Abstracts |
Sponsoring Units: DCOMP DMP Chair: Andrew Canning, Lawrence Berkeley National Laboratory Room: Baltimore Convention Center 329 |
Monday, March 13, 2006 11:15AM - 11:51AM |
B32.00001: Losses due to phonon-phonon interactions in nanotube oscillators: from classical potentials through one-dimensional elasticity and many-body perturbation theory Invited Speaker: Phonon-phonon losses are an intrinsic loss mechanism and it is therefore important to calculate their magnitude as an upper bound on the quality factor of any nano-oscillator. We will present an approach to handling the problem of phonon-phonon interactions in nanotube oscillators which uses an empirical interatomic potential to compute the input parameters for a fully quantum-mechanical green function-based method. This approach allows us to compute losses at temperatures comparable to or below the Debye temperature ($>$ 500K), allowing comparison with experimental results. [Preview Abstract] |
Monday, March 13, 2006 11:51AM - 12:03PM |
B32.00002: Scalable Linear Electronic Structure Method for Free Standing Clusters Kab Seok Kang, James Davenport, David Keyes, James Glimm We have developed a Scalable Linear Augmented Slater Type Orbital (LASTO) method for electronic-structure calculations on free standing clusters. As with other linear methods we solve the Schrodinger equation using a mixed basis set consisting of numerical functions inside atom centered spheres matched onto tail functions outside. The tail functions are Slater type orbitals which are localized, exponentially decaying functions. To solve the Poisson equation between spheres, we use a finite difference method replacing the rapidly varying charge density inside the spheres with a smoothed density with the same multipole moments. We use multigrid techniques on the mesh which are well-known scalable solvers. This yields the Coulomb potential on the spheres which in turn defines the potential inside via a Dirichlet problem. To solve the linear eigen-problem, we use SCALAPACK, a well-developed package to solve large eigen-problems with dense matrices. We have tested the method on finite clusters of palladium and palladium hydride. Supported by the US Department of Energy under contract DEA02-98CH10886. [Preview Abstract] |
Monday, March 13, 2006 12:03PM - 12:15PM |
B32.00003: ONETEP: linear-scaling density-functional theory with plane waves Peter Haynes, Chris-Kriton Skylaris, Arash Mostofi, Mike Payne Plane waves are a popular choice of basis set for first-principles quantum-mechanical simulations based on density-functional theory because the implementation is straightforward and the completeness can be controlled systematically with a single parameter. The resulting simulations require a computational effort which scales as the cube of the system-size, which makes the cost of large-scale calculations prohibitive. Extended basis functions would appear to be an inappropriate choice for expanding the localized orbitals of linear-scaling methods or for embedding the calculation within a larger model. In spite of these apparent difficulties, the ONETEP linear-scaling method can achieve the same accuracy as traditional plane-wave calculations and overcomes the apparent difficulties mentioned above. An outline of the ONETEP method will be presented, focusing on its distinctive features and primarily the ability to optimize the localized orbitals in each particular environment. These optimized orbitals are known as non-orthogonal generalized Wannier functions within ONETEP, and justification for this term will be presented, in addition to results that demonstrate the scaling and accuracy of the method. [Preview Abstract] |
Monday, March 13, 2006 12:15PM - 12:27PM |
B32.00004: Acceleration of the Convergence in {\it ab initio} Atomic Relaxations Zhengji Zhao, Lin-Wang Wang, Juan Meza Atomic relaxations is often required to accurately describe the properties of nanosystems. In {\it ab initio} calculations, a common practice is to use a standard search algorithm, such as BFGS (Broyden-Fletcher-Goldfarb-Shanno) or CG (conjugate gradient) method, which starts the atomic relaxations without any knowledge of the Hessian matrix of the system. For example, the initial Hessian in BFGS method is often set to identity, and there is no preconditioning to CG method. One way to accelerate the convergence of the atomic relaxations is to estimate an approximate Hessian matrix of the system and then use it as the initial Hessian in BFGS method or a preconditioner in CG method. Previous attempts to obtain the approximated Hessian were focused on the use of classical force field models which rely on the existence of good parameters. Here, we present an alternative method to estimate the Hessian matrix of a nanosystem. First, we decompose the system into motifs which consist of a few atoms, then calculate the Hessian matrix elements on different motif types from {\it ab initio} calculations for small prototype systems. Then we generate the Hessian Matrix of the whole system by putting together these motif Hessians. We have applied our motif-based Hessian matrix in {\it ab initio} atomic relaxations in several bulk (with/without impurity) and quantum dot systems, and have found a speed up factor of 2 to 4 depending on the system size. [Preview Abstract] |
Monday, March 13, 2006 12:27PM - 12:39PM |
B32.00005: Electric field polarization calculations using charge patching method Lin-Wang Wang, Xavier Cartoixa Soler Charge patching method has been used to calculate the electronic structures of thousand atom nanosystems with ab initio accuracy. Within the charge patching method, the ab initio charge density of a nanostructure is patched together using charge motifs which are calculated from small prototype systems. However, the current charge patching method can only be used for systems without long range electric field. In this talk, we will present results which include the polarizations of charge motifs. These polarization motifs accurately describe the charge responses of a nanosystem under external electric fields, and the results agree well with direct ab initio calculations, including part of the local field effects. This motif polarization method enables us to calculate nanosystems such as charged impurities, quantum dots with permanent dipole moments, or charged quantum dots. It also makes it possible to calculate the charge density of a nanosystem selfconsistently when combined with the folded spectrum method in solving a few band edge electronic states. [Preview Abstract] |
Monday, March 13, 2006 12:39PM - 12:51PM |
B32.00006: A divide and conquer method for Bader decomposition Jun-ichi Hoshino, Kazuo Tsumuraya The ionicity of atoms in crystals or molecules is a measure of the bonding states among the atoms. There have been several methods to evaluate the electron charges that belong to each atom. Bader analysis is a direct method and divides up into atomic regions where the dividing surfaces are at a minimum in the charge density. [1] The algorithm, however, has computational difficulties to find the critical points for complex circumstance of atoms. A Henkelman’s algorithm [2] is free of the points, assign each point on a regular grid to one of the regions by following a steepest ascent (SA) method contrary to the Bader’s steepest descent (SD) algorithm. We have found the SA method separates the regions more precisely than the SD method, although they are reverse relation. The SA method however requires much memory capacity for all-electron densities which are oscillating in the core regions. We implement a divide and conquer method to calculate the core region separately from the other region, assess the proposed method, and compare it with the Henkelman’s decomposition. \begin{enumerate} \item R.Bader, Atoms in Molecules:A Quantum Theory, Oxford,1990. \item G.Henkelman,et al., Comp.Mater.Sci.(in press) \end{enumerate} [Preview Abstract] |
Monday, March 13, 2006 12:51PM - 1:03PM |
B32.00007: Direct enumeration of alloy configurations for semiconductor electronic structure properties Sirichok Jungthawan, Sukit Limpijumnong, Peter A. Graf, Kwiseon Kim, Wesley B. Jones, Gus L. W. Hart We present an approach to directly enumerating the electronic structure of all possible zincblende-based alloy configurations whose unit cell contains up to a specified number of atoms. This method allows us to map the space of bandgaps and effective masses versus alloy composition and atomic configuration. We demonstrate for GaInP alloys that a large range of bandgaps and masses are available for a given composition. By decomposing the space of possible atomic configurations into categories based on superlattice structure, we can identify trends in bandgap extrema. For example, bandgap maxima typically occur in [0 h k] superlattices where h is not equal to k, and minima typically occur in [1 1 1] superlattices. We focus on dilute alloys where the minority composition is below 10 percent. The empirical pseudo potential method (EPM) and folded spectrum method are used to solve the single particle Schr\"{o}dinger equation. The results from the EPM are compared with first- principle calculations. [Preview Abstract] |
Monday, March 13, 2006 1:03PM - 1:15PM |
B32.00008: Surface passivation optimization using DIRECT Kwiseon Kim, Peter A. Graf, Wesley B. Jones, Lin-Wang Wang The calculation of the electronic structure of a nanostructure must take into account surface effects. In experiments, the dangling bonds at the surface of a semiconductor nanostructure are passivated by other semiconductors or by organic ligands. In either case, photoluminescence measurements reveal that the emission comes from bulk-like, dot-interior states. These observations suggest that an approach to passivating a simulated nanostructure would be to attach “pseudo-atoms” to each dangling bond. Here we present an automated methodology for generating surface passivating pseudo potentials for bulk empirical pseudo potentials. Our method is based on the global optimization method DIRECT. We apply it to two materials, CdSe and InP. Incorporated into a larger computational nanoscience infrastructure, our work represents a much needed improvement in the usability of the empirical pseudo potential method. [Preview Abstract] |
Monday, March 13, 2006 1:15PM - 1:27PM |
B32.00009: Non-linear dynamics of the electron wave packet propagating through the resonant tunneling structure under the presence of the electron-photon and inter-electron interactions Masakazu Muraguchi, Kyozaburo Takeda We solved the TD Schrodinger equation numerically in the framework of the TD Hartree-Fock (HF) approach both in the real space and time, and studied the TD phenomena of an electron wave packet propagating through the time-modulated resonant tunneling structure (TMRTS). We suppose that a single electron Gaussian wave packet is injected into the TMRTS by varying its group velocity. For the inter-electron interaction in the TMRTS, we combine Poisson's equation with TD-HF equation. We found that the TD features of the wave function confined in the TMRTS (e.g., lifetime) strongly depend on the choice of the resonant states as well as the frequency and strength of the applied electric field. Furthermore, TD non-linear processes based on the multi-photon interaction are recognized. For extracting characteristics of these TD phenomena, we expanded the resulting wave function in terms of the RTS resonant sates at each time-step. This projection approach enables us not only to estimate the lifetime precisely but also provides a guiding principle to control the wave packet artificially. [Preview Abstract] |
Monday, March 13, 2006 1:27PM - 1:39PM |
B32.00010: Multi-Scale Modeling of Carbon Nanotube/Carbon Fiber/Epoxy Lamina S.J.V. Frankland, J.C. Riddick, T.S. Gates A carbon fiber/epoxy lamina in which the carbon fibers are coated with single-walled carbon nanotubes is modeled with a multi-scale method. The multi-scale model is designed to predict the effect of the carbon nanotubes on the constitutive properties of the lamina. Within the model both the nanotube volume fraction and nanotube distribution are varied. The multi-scale analysis links results from molecular dynamics and equivalent-continuum techniques with micromechanics and strength of materials models. The multi-scale method will be used in a parametric study to examine the relative effect of nanotube concentration, orientation, and distribution on the constitutive properties of the lamina. [Preview Abstract] |
Monday, March 13, 2006 1:39PM - 1:51PM |
B32.00011: Tight binding models derived from k-dot-p theory C. E. Pryor, M.-E. Pistol Calculations of the electronic properties of semiconductor nanostructures rely on one of three different methods: tight-binding, pseudopotentials, or k-dot-p theory. The first two are well suited to modeling small scale structures, however their parameters must be fitted to bulk properties, which can be a complicated procedure, especially for tight-binding. In contrast, k-dot-p theory is best at describing large nanostructures in which the placement of individual atoms is not important, and the parameters of k-dot-p theory are directly related to experimentally determined quantities. To bridge the gap between atomistic and large scale models, we will present a method for constructing tight-binding models directly from k-dot-p theory by considering a real-space representation of k-dot-p theory with finite differences on a grid which matches the desired crystal lattice. Conversely, given a tight-binding model it is also possible to construct an equivalent k-dot-p theory in the long wavelength limit [Preview Abstract] |
Monday, March 13, 2006 1:51PM - 2:03PM |
B32.00012: Mobility of Lithium and Hydrogen Ions in Nanotubes in Terms of Fokker-Plank Equation Andrew Kinchen, Yuriy Malozovsky We present the theory of mobility of Li and H ions in metallic nanotubes. We derived the mobility of ions in terms of the kinetic Fokker-Plank equation with the consideration of both the motion of an ion in the cylindrical periodic potential of a nanotube and interaction of an ion with lattice vibrations of the tubule. We argue that there is an optimum diameter of the tubule below which the mobility of ions is significantly reduced. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700