Bulletin of the American Physical Society
APS March Meeting 2011
Volume 56, Number 1
Monday–Friday, March 21–25, 2011; Dallas, Texas
Session H18: Electronic Structure II |
Hide Abstracts |
Sponsoring Units: DCOMP Chair: David Prendergast, Lawrence Berkeley National Laboratory Room: D172 |
Tuesday, March 22, 2011 8:00AM - 8:12AM |
H18.00001: Spectral element solution of the Kohn-Sham atom Kristopher Andersen, Noam Bernstein, John Pask Electronic structure calculations of atoms are important in nuclear physics, and are necessary input for most methods to construct first-principles effective potentials (i.e., pseudopotentials and projector augmented wave potentials). The standard method to solve the atomic problem within Kohn-Sham density functional theory is the shooting method. In this work, the more robust spectral element method is applied to the 1D atomic radial equation. The spectral element method provides a strict, upper-bound on the absolute error in the Kohn-Sham eigenvalues and wavefunctions enabling the solution to be converged to a well controlled accuracy. The results of this method are compared to the extensive ``NIST Atomic Reference Data for Electronic Structure Calculations'' database for elements H to U, providing a more rigourous assessment of this dataset than previously available. [Preview Abstract] |
Tuesday, March 22, 2011 8:12AM - 8:24AM |
H18.00002: Ab initio calculations of optical absorption spectra: Solution of the Bethe-Salpeter equation within density matrix perturbation theory Dario Rocca, Deyu Lu, Huy-Viet Nguyen, Giulia Galli We present an approach to compute optical absorption spectra from first principles, which is suitable for the study of large systems and gives access to spectra within a wide energy range. In this approach, the quantum Liouville equation is solved iteratively within first order perturbation theory, with a Hamiltonian containing a static self-energy operator [1]. This is equivalent to solving the Bethe-Salpeter equation. Explicit calculations of single particle excited states and inversion of dielectric matrices are avoided using techniques based on Density Functional Perturbation Theory [1,2]. The calculation and inclusion of GW quasi-particle corrections within this framework are discussed. The efficiency and accuracy of our approach are demonstrated by computing optical spectra of solids, nanostructures and dipeptide molecules exhibiting charge transfer excitations. \\[4pt] [1] D.Rocca, D.Lu and G.Galli, J. Chem. Phys. 133, 164109 (2010). \\[0pt] [2] H. Wilson, F. Gygi and G. Galli, Phys. Rev. B , 78, 113303, (2008). [Preview Abstract] |
Tuesday, March 22, 2011 8:24AM - 8:36AM |
H18.00003: Ab-initio calculations of absorption spectra of nanowires by solving the Bethe-Salpeter Equation Yuan Ping, Dario Rocca, Deyu Lu, Giulia Galli A first principle approach to the solution of the Bethe Salpeter equation without empty electronic states has been recently developed [1], which makes possible the calculations of absorption spectra of relatively large systems (with several hundreds of electrons). We present applications of this approach to quasi-one dimensional systems, including chains of hydrogen molecules and Si nanowires. We discuss techniques to further improve the performance of absorption spectra calculations, and present a general scheme to accurately integrate the divergence in the screened exchange integrals. Finally, in the case of Si nanowires, we discuss the effect of surface reconstruction in shaping optical absorption spectra.\\[4pt] [1] D. Rocca, D. Lu and G. Galli, J. Chem. Phys. 133, 164109 (2010) [Preview Abstract] |
Tuesday, March 22, 2011 8:36AM - 8:48AM |
H18.00004: Efficient GW calculations using eigenvalue-eigenvector decomposition of the dielectric matrix Huy-Viet Nguyen, T. Anh Pham, Dario Rocca, Giulia Galli During the past 25 years, the GW method [1] has been successfully used to compute electronic quasi-particle excitation spectra of a variety of materials. It is however a computationally intensive technique, as it involves summations over occupied and empty electronic states, to evaluate both the Green function (G) and the dielectric matrix (DM) entering the expression of the screened Coulomb interaction (W). Recent developments have shown that eigenpotentials of DMs can be efficiently calculated without any explicit evaluation of empty states [2]. In this work, we will present a computationally efficient approach to the calculations of GW spectra by combining a representation of DMs in terms of its eigenpotentials [3] and a recently developed iterative algorithm [4]. As a demonstration of the efficiency of the method, we will present calculations of the vertical ionization potentials of several systems. [1] L. Hedin, Phys. Rev. 139, A796 (1965). [2] H.-V. Nguyen and S. de Gironcoli, Phys. Rev. B 79, 205114 (2009); H. F. Wilson, D. Lu, F. Gygi, and G. Galli, Phys. Rev. B 79, 245106 (2009). [3] D. Lu, F. Gygi, and G. Galli, Phys. Rev. Lett. 100, 147601 (2008). [4] P. Umari, G. Stenuit, and S. Baroni, Phys. Rev. B 81, 115104 (2010) [Preview Abstract] |
Tuesday, March 22, 2011 8:48AM - 9:00AM |
H18.00005: A TDLDA+U approach on strongly hybridized Frenkel excitons in Mott insulators and implications to TDDFT and GW+BSE Chi-Cheng Lee, H.C. Hsueh, Wei Ku The applicability of nowadays first-principles approach on local excitations of strongly correlated systems is unknown. We therefore derived the dynamical linear response of LDA+U functional within the framework of TDDFT.\footnote{Chi-Cheng Lee et al., Phys. Rev. B 82, 081106(R) (2010).} The strength and weakness of LDA+U functional in describing charge excitations of strongly interacting Mott insulators is examined via this TDLDA+U method. Formulated using real-space Wannier functions, a computationally inexpensive framework gives detailed insights into the formation of tightly bound Frenkel excitons with reasonable accuracy. Specifically, a strong hybridization of multiple excitons is found to significantly modify the exciton properties. Furthermore, our study exposes a significant generic limitation of adiabatic approximation in TDDFT with hybrid functionals and in existing Bethe-Salpeter-equation approaches, advocating the necessity of strongly energy-dependent kernels in future development. Finally, a superatom approach beyond TDLDA+U will also be discussed. [Preview Abstract] |
Tuesday, March 22, 2011 9:00AM - 9:12AM |
H18.00006: Level alignment at covalently bonded metal-organic interfaces within the GW approximation Jeffrey Neaton, Isaac Tamblyn, Su Ying Quek, Stanimir Bonev, Pierre Darancet Accurate calculations of orbital energies for molecules chemisorbed on metal surfaces are important for understanding energy conversion, molecular scale transport, and charge transfer events at metal electrodes. Here, using density functional theory (DFT) and many-body perturbation theory within the GW approximation (GWA), we report the orbital energies of a well-studied molecule, benzene diamine (and derivatives), covalently bonded to aluminum and gold surfaces. For chemisorbed derivatives on Al surfaces, we predict a shift in the highest occupied molecular orbital resonance energy greater than 1 eV relative to the DFT result. We discuss our GWA results in the context of a model self-energy approach based on prior work [1], which can be applied to larger systems at greatly reduced computational cost. \\[4pt] [1] J. B. Neaton, M.S. Hybertsen, and S.G. Louie, PRL, 97, 216405 (2006) [Preview Abstract] |
Tuesday, March 22, 2011 9:12AM - 9:24AM |
H18.00007: GW approach to degenerate systems Johannes Lischner, Jack Deslippe, Steven G. Louie Many-body perturbation theory based on the GW approximation to the electron self energy describes accurately in first-principles calculations the electronic (quasiparticle) excited states of solids, clusters and molecules. However, despite the multitude of important systems with degenerate ground states, ranging from open-shell atoms and molecules to magnetic defects in solids, the GW approach has been applied almost exclusively to closed-shell systems. In this talk, we discuss some of the problems with existing GW calculations for degenerate systems, such as spin contamination, the multiplet problem, and the proper definition of the Green function in open-shell systems. Different formulations to overcome these problems are explored. [Preview Abstract] |
Tuesday, March 22, 2011 9:24AM - 9:36AM |
H18.00008: Ab-initio theory of spin fluctuations in magnets Vladimir Antropov, Liqin Ke, Mark van Schilfgaarde, Mikhael Katsnelson We propose a framework for a true ab initio theory of magnetism, based on many-body perturbation theory (MPBT). It fits in naturally with methods based MPBT such as the GW approximation; but the approach can be implemented as an extension to any existing static method for electronic structure such as the local spin density approximation to density functional theory, to include spin fluctuations. Initially we calculated the spin fluctuation contributions using random phase approximation. The self consistency procedure similar to the one used in Moryia-Kawabata theory can be naturally implemented. The fluctuation dissipation theorem is used to calculate the reduction of the mean field magnetic moment in itinerant magnets. The applications of the technique includes traditional 3d ferromagnetic metals, their alloys and compounds and 5f systems. [Preview Abstract] |
Tuesday, March 22, 2011 9:36AM - 9:48AM |
H18.00009: A GW-based many-body perturbation theory approach for investigating materials with strong spin-orbit coupling Bradford Barker, Jack Deslippe, Oleg Yazyev, Steven Louie Spin-orbit coupling is an essential ingredient in understanding the electronic properties of materials of recent interest. We have developed a means of incorporating spin-orbit coupling to the quasiparticle excitations in solids within the GW approach. We apply our method to the properties of materials with heavy ion cores. This work was supported by National Science Foundation Grant No. DMR10-1006184, the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. Computational resources have been provided by NERSC. [Preview Abstract] |
Tuesday, March 22, 2011 9:48AM - 10:00AM |
H18.00010: Self-consistent band-structure calculations at GW quality and DFT expense Stephan Lany, P. Graf, M. d'Avezac, A. Zunger GW provides rather accurate quasi particle energies where approximate DFT methods tend to fail, yet are too computer intensive for complex~inorganic materials (large systems or dense Brillouin zone sampling) that are currently of interest, e.g. for energy conversion. We explore the possibility that the trends of the GW quasi-particle energy corrections due to the non-local and energy dependent self-energy $\Sigma $(r,r',$E)$ can be captured by atomic potentials that do not significantly increase the computational effort of a standard DFT calculation. We proceed in 4 steps: (i) Perform GW reference calculations for II-VI and III-V semiconductors (ii) Define atomic potentials that are added to the DFT Hamiltonian. Here we extend the concept of the non-local external potentials (NLEP) of Ref. [1], now allowing for two parameters per atom type and angular momentum. (iii) Fit the NLEP parameters to the GW test set. (iv) Finally, we test transferability by applying the potentials to the III$_{2}$-VI$_{3}$ and II$_{3}$-V$_{2}$ compounds that were not included in the fitting set. \\[4pt] [1] S. Lany, H. Raebiger, A. Zunger, Phys. Rev. B 77, 241201(R) (2008). [Preview Abstract] |
Tuesday, March 22, 2011 10:00AM - 10:12AM |
H18.00011: The sc-COHSEX+GW and the static off-diagonal GW approaches to quasiparticle wavefunctions and energies Jack Deslippe, Manish Jain, Georgy Samsonidze, Marvin Cohen, Steven Louie Within the conventional GW approach, density functional theory (DFT) is typically used as a mean-field starting point; the self-energy operator is evaluated to 1st order in the DFT Green's function G$_{0}$ and screened Coulomb interaction W$_{0}$. The quasiparticle energies are calculated from diagonal elements of $\Sigma$ in the DFT orbital basis. This approach works extraordinarily well for many materials but has limitations when the DFT states are far from the quasiparticle wavefunctions. In such cases, off-diagonal elements of $\Sigma$ in the mean-field basis are large and the full $\Sigma$ matrix is needed. The slow convergence of the off-diagonal elements make approaches requiring the explicit construction of this matrix prohibitively expensive. We present two alternative approaches based on the static (COHSEX) approximation that efficiently include the mean-field off-diagonal matrix element effects: a sc-COHSEX+GW approach where a renormalized basis is obtained from a self-consistent evaluation of quasiparticle wavefunctions in the static approximation and a less intensive treatment of just the off-diagonal elements within the COHSEX approximation. We show examples of the approaches for molecules and crystalline systems. Support by NSF DMR10-1006184, DOE DE-AC02-05CH11231. [Preview Abstract] |
Tuesday, March 22, 2011 10:12AM - 10:24AM |
H18.00012: Constructing unoccupied states for G$_0$W$_0$ quasiparticle calculations from plane-waves Georgy Samsonidze, Manish Jain, Jack Deslippe, Marvin L. Cohen, Steven G. Louie Standard methods of first-principles calculations of the quasiparticle energies within the G$_0$W$_0$ scheme require summing over large numbers of unoccupied states. The generation of these states within the ab initio pseudopotential plane-wave density functional theory (DFT) quickly becomes a bottleneck of the calculation with increasing system size, especially in low-dimensional systems. In this work, we propose a method for approximating the high-energy continuum and resonant states in low-dimensional systems. The continuum and resonant states above a chosen energy are replaced with symmetrized plane-waves and localized DFT states computed with short-range localized basis functions (such as in the SIESTA code), respectively. The Gram-Schmidt process is used to orthogonalize these constructed high-energy unoccupied states. The method opens a route towards precise G$_0$W$_0$ quasiparticle calculations in large low-dimensional systems using a small number of unoccupied DFT states. This work was supported by NSF Grant No. DMR10-1006184, the U.S. DOE under Contract No. DE-AC02-05CH11231. Computational resources have been provided by NSF through TeraGrid at NICS and DOE at LBNL's NERSC. [Preview Abstract] |
Tuesday, March 22, 2011 10:24AM - 10:36AM |
H18.00013: On-site screened Coulomb interactions for localized electrons in transition metal oxides and defect systems Bi-Ching Shih, Peihong Zhang Electronic and structural properties of strongly correlated material systems are largely determined by the strength of the on-site Coulomb interaction. Theoretical models devised to capture the physics of strongly correlated materials usually involve screened Coulomb interactions as adjustable parameters. We present first-principles results for the screened on-site Coulomb and exchange energy for transition metal oxides. The dielectric screening is calculated within the random phase approximation and the localized electrons are represented by maximally localized Wannier functions. We further extend our study to calculate on-site Coulomb interactions for localized defect states in semiconductors. We acknowledge the computational support provided by the Center for Computational Research at the University at Buffalo, SUNY. This work is supported by the National Science Foundation under Grant No. DMR-0946404 and by the Department of Energy under Grant No. DE-SC0002623. [Preview Abstract] |
Tuesday, March 22, 2011 10:36AM - 10:48AM |
H18.00014: Optimizing Generalized Norm-Conserving Pseudopotentials D.R. Hamann The ``generalized'' method permits the construction of norm-conserving pseudopotentials at energies that do not correspond to bound atomic states, giving added flexibility in the treatment of angular-momentum channels for which no bound states exist.\footnote{D. R. Hamann, Phys. Rev. B \textbf{40}, 2980 (1989).} An effective method for optimizing the convergence of pseudopotential calculations with plane-wave-basis cutoff energy requires atomic wave functions with decaying tails, and has not been applicable to such ``generalized'' states.\footnote{A. M. Rappe,\textit{ et al.}, Phys. Rev. B \textbf{41}, 1227 (1990).} By introducing a potential well outside the core radius for selected angular-momenta, an artificial decaying tail can be produced for positive-energy states. This permits the application of the optimization method, and we find convergence behavior comparable to that for ordinary bound states. In practice, we terminate the positive-energy all-electron wave function smoothly with an exponential or Gaussian tail, and never need to treat the implied well potential explicitly. The projectors to form fully-nonlocal operators\footnote{L. Kleinman and D. M. Bylander, Phys. Rev. Lett. \textbf{48}, 1425 (1982).} can be terminated at the core radii as usual, despite differences of the semi-local potentials outside the well radii. [Preview Abstract] |
Tuesday, March 22, 2011 10:48AM - 11:00AM |
H18.00015: Development of a Semi-empirical Hamiltonian for Phosphorus for Quantum Mechanics Based Simulations of Phosphorous-based Nanostructures Paul Tandy, Christopher Leahy, Ming Yu, C.S. Jayanthi, S.Y. Wu We have developed a parameterized semi-empirical Hamiltonian for phosphorous for simulation studies of phosphorous-based nanostructures including phosphorous-doped silicon nanowires.This Hamiltonian models the environment-dependent electron-ion and ion-ion interactions and electron-electron correlations, by capturing the salient features of \textit{ab initio} Hamiltonians/\textit{ab initio} methods, ($e.g$., electron screening and charge self-consistency).Such a semi-empirical Hamiltonian has been shown to be successful in predicting the properties of intermediate-sized silicon, boron, and carbon clusters and extended structures of boron and silicon [1-4]. We optimized the parameters of our Hamiltonian for phosphorous by fitting the properties of bulk (black phosphorous) and small clusters (P$_{2}$ to P$_{10})$ as obtained by our method to \textit{ab initio} calculations. It is expected that such a Hamiltonian will have the predictive power to enable the study of larger phosphorous based nanostructures that are not possible via \textit{ab initio} studies. \\[4pt] [1] C. Leahy, et al, Phys. Rev. B 74,155408 (2006). [2 ]P. Tandy, et al, Bulletin of the APS,2009 APS March Meeting Vol. 54, Num.1, Sess. D26, [3] Ming Yu, et al, J. Chem. Phys. 130,184708 (2009). [4] Ming Yu, S.Y. Wu, and C.S. Jayanthi, Physica E 42, 1 (2009). [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