Bulletin of the American Physical Society
APS March Meeting 2021
Volume 66, Number 1
Monday–Friday, March 15–19, 2021; Virtual; Time Zone: Central Daylight Time, USA
Session R19: Density Functional Theory and Beyond IIILive
|
Hide Abstracts |
Sponsoring Units: DCOMP DCP DCMP DPOLY Chair: Koblar Jackson, Central Michigan Univ |
Thursday, March 18, 2021 8:00AM - 8:12AM Live |
R19.00001: Analyticity with respect to external potential in DFT and implications for Kohn-Sham computation Paul Lammert
|
Thursday, March 18, 2021 8:12AM - 8:24AM Live |
R19.00002: Density Functional Theory Study of the Optical and Electronic Properties of With-Defect Semiconductors using a Tuned Screened Range-Separated Hybrid Kirk Lewis, Ashwin Ramasubramaniam, Sahar Sharifzadeh Tuned and screened range-separated hybrid (SRSH) methods have emerged as an alternative to state of the art many-body perturbation theory (MBPT) calculations of the optoelectronic properties of materials. Specifically, it has been shown that SRSH hybrid methods can approach the quantitative accuracy of MBPT at the cost of hybrid DFT for a variety of dissimilar molecules and both bulk and monolayer crystals. Here, we test the accuracy of the time-dependent (TD) SRSH approach for describing the optoelectronic properties of defective semiconductors by the study of point defects in bulk GaN. We first show that the predicted quasiparticle gap and low-energy excitation spectra of (TD)-SRSH and GW/BSE agree well in both pristine GaN and GaN with a single nitrogen vacancy, establishing the accuracy of the method. Aided by the reduced computational cost of (TD-)SRSH, we then report on a series of technologically relevant point defects and complexes in GaN. This study indicates that TD-SRSH is a promising and computationally feasible approach for quantitatively accurate, first-principles modeling of defective semiconductors. |
Thursday, March 18, 2021 8:24AM - 8:36AM Live |
R19.00003: Accurately predicting electron affinities with Koopmans spectral functionals Edward Linscott, Nicola Colonna, Riccardo De Gennaro, Nicola Marzari Density functional theory (DFT) is a popular method for electronic-structure calculations. But while Kohn-Sham eigenvalues can loosely mirror experimental quasiparticle energies, there is formally no connection between the two (except for the HOMO in exact DFT). Furthermore, the presence of self-interaction errors in semi-local DFT can make those eigenvalues an even poorer proxy for quasiparticle energies [1]. |
Thursday, March 18, 2021 8:36AM - 8:48AM Live |
R19.00004: Optical absorption spectra from model exchange-correlation (XC) kernels SANTOSH NEUPANE, Niraj K Nepal, Adrienn Ruzsinszky We use model exchange-correlation (XC) kernels in the framework of time dependent density functional theory (TDDFT) to obtain the optical absorption spectra of different bulk materials. We also calculate the optical absorption spectra by solving the Bethe-Salpeter equation (BSE) for the two-particle Green’s function. We test various kernels such as JGM and JGM-G [1] on different bulk materials, and compare them with the Random Phase Approximation (RPA). In addition, we compare all the results with the experimental spectra when available. We find that model xc kernels built upon exact physical constraints are reasonably accurate for the optical response properties of bulk solids. |
Thursday, March 18, 2021 8:48AM - 9:00AM Live |
R19.00005: Band-gap of bulk solids and two-dimensional bent nanoribbons from first-principles Bimal Neupane, Hong Tang, Niraj K Nepal, Adrienn Ruzsinszky
|
Thursday, March 18, 2021 9:00AM - 9:12AM Live |
R19.00006: Barrier Heights of BH76 Database with PZ-SIC and Locally-scaled Self Interaction Correction Methods Prakash Mishra, Yoh Yamamoto, Koblar Jackson, Tunna Baruah, Rajendra R Zope We investigate the performance of the Perdew-Zunger (PZ) self-interaction correction (SIC) method and recent locally scaled SIC (LSIC) method[1] for predicting barrier heights in chemical reactions using the BH76 database. Self-interaction error (SIE) is pronounced when molecules or solids are not in equilibrium, such as when chemical bonds are stretched or broken during chemical reactions. We determine the SIC using Fermi-Löwdin orbitals[2]. We find that removing SIE using the PZ-SIC energy functional in the FLO-SIC framework reduces the overall errors, but in PZ-SIC-LSDA the barriers are still too small compared to reference values. Applying LSIC-LSDA improves the barrier heights in almost every case resulting in better agreement with experiment and the accurate reference values from the higher-level calculations. |
Thursday, March 18, 2021 9:12AM - 9:24AM Live |
R19.00007: Quantifying and reducing different sources of errors in DFT calculations Stefan Vuckovic, Suhwan Song, Eunji Sim, Kieron Burke Density functional theory (DFT) calculations are ubiquitous in different branches of chemistry and physics. While in principle, DFT is an exact theory, in practice, it must rely on approximations. In this talk, I will describe a set of approaches for disentangling different sources of errors in approximate DFT calculations. I will discuss how errors in approximate molecular geometries [1] and approximate electronic densities [2–4] affect the overall accuracy of DFT calculations. Then I will explain how these insights can be used to improve the performance and to reduce the cost of DFT calculations with implications for both molecules and surfaces. |
Thursday, March 18, 2021 9:24AM - 9:36AM Live |
R19.00008: Density sensitive analysis for evaluating density functional theory approximations to exchange-correlation energies Ryan J. McCarty, Stefan Vuckovic, Suhwan Song, John Kozlowski, Eunji Sim, Kieron Burke We developed a density sensitivity difference measure using theory from density-corrected Density Functional Theory that provides a physically-motivated comparison of exchange-correlation functional approximations. Analyzing the comparative density sensitivities with machine-learning reveals striking trends among standard functionals. Comparative differences between approximate functionals are more easily quantified than absolute errors, and are indifferent to the intent or construction of the functional. Evaluating individual molecules with this approach indicates clear molecular groupings, highlighting the similarities or differences of various external potentials. Our analysis provides a new method for evaluating DFT functionals, and enables a new data driven approach for dataset generation. |
Thursday, March 18, 2021 9:36AM - 9:48AM Live |
R19.00009: Accelerating the Fermi-Löwdin Orbital Descriptor Optimizations for Self-Interaction Free Density Functional Theory Calculations. Md Nageeb Bin Zaman, Koblar Alan Jackson, Juan Ernesto Peralta The Fermi-Löwdin Self Interaction Correction (FLO-SIC) method was introduced to address the shortcomings of standard density functional approximation calculations by removing the spurious electron self-interaction [1-4]. Within this method, a set of Fermi orbital descriptors (FODs) need to be optimized to minimize the energy. Current FOD optimization algorithms require large numbers of steps and the number of steps grows with systems size, that is, as the number of FODs increases. Moreover, every step in this optimization process requires the time-consuming evaluation of single-particle potential energy functions. Here we propose a strategy to reduce the number of single particle potential evaluations, making the FOD optimization step in FLO-SIC calculations more efficient. We will present proof-of-concept results for small molecules and compare with existing optimization schemes. |
Thursday, March 18, 2021 9:48AM - 10:00AM Live |
R19.00010: A new method for initializing Fermi orbital descriptors for FLO-SIC calculations Duyen Nguyen, Koblar Alan Jackson, John Perdew, Mark Pederson, Juan Ernesto Peralta Fermi orbital descriptors (FODs) play a key role in Fermi-Löwdin orbital self-interaction correction (FLO-SIC) calculations used to remove one-electron self-interaction from approximate density functional calculations on an orbital-by-orbital basis. Optimal FODs are obtaining by minimizing the SIC energy, and, in this process, identifying an initial set of FODs becomes crucial for practical applications. Here we propose a novel generator for automatically initializing FODs without requiring much user input based on the minimization of a “pseudo energy” expression that involves a Coulomb electron density attraction, a FOD-FOD short-range repulsion, and an exchange-like density repulsion term. We implemented and tested this method for molecules involving a variety of bonding situations and found that it successfully reproduces FOD configurations that are in qualitative good agreement with Lewis theory. For spin-unpolarized molecules, this method underestimates and overestimates the separation of lone pair FODs in the second- and third-row elements, respectively, while the distances between double and triple bond FODs are slightly exaggerated. For spin-polarized systems, this method can also provide good FODs for radicals and transition states. |
Thursday, March 18, 2021 10:00AM - 10:12AM Live |
R19.00011: Fragment Electron Populations in Partition Density Functional Theory Kui Zhang, Adam Wasserman Partition Density Functional Theory (P-DFT) is a density-based embedding method that partitions a system into fragments by minimizing the sum of fragment energies subject to two constraints: (1) That the sum of fragment densities equals the density of the system; (2) That the sum of fragment electron populations equals the total number of electrons. To perform this constrained minimization, we study a two-stage procedure in which the sum of fragment energies is lowered when electrons flow from fragments of lower electronegativity to fragments of higher electronegativity. The global minimum is reached when all electronegativities are equal. The non-integral fragment electron populations are dealt with in two different ways: (1) by using fractionally occupied orbitals (FOO) and (2) ensemble (ENS) treatments. Although these two methods lead to the same total energy and density, they lead to different fragment properties and partial charges. We compare exact P-DFT calculations with results obtained from the Local-Density Approximation (LDA) for heteronuclear diatomic molecules. We find that the electron numbers transferred in ENS are generally smaller than that in FOO, and explain why. |
Thursday, March 18, 2021 10:12AM - 10:24AM Live |
R19.00012: Unharmonic adiabatic potential by short-range correlation effect enlarging C33 of crystalline graphite Koichi Kusakabe, Akira Nagakubo, Hirotsugu Ogi, Kensuke Murashima, Mutsuaki Murakami In our previous work using the LDA+U+RPA method [1], we reported a theoretical value of C33 of graphite as large as 48GPa when cRPA estimation of the on-site U for 2p-orbitals is used. Renormalization in Wannier functions determined at each point on the adiabatic potential surface is essential. Fixed localized orbital cannot cause the enhancement of C33 from a value by ACFDT-RPA, which is more than 20% smaller than the observed C33 of defect-free monocrystalline graphite. In this presentation, after discussing relevance of the multi-reference extension of DFT[2] for beyond-RPA approaches, we open comparison among several +U approaches with the double-counting term. |
Thursday, March 18, 2021 10:24AM - 10:36AM Live |
R19.00013: Efficient First-Principles Approach with a Pseudohybrid Density Functional for Extended Hubbard Interactions Sang-Hoon Lee, Young-Woo Son For massive database-driven materials research, there are increasing demands for both fast and accurate quantum mechanical computational tools. Contemporary density functional theory (DFT) methods can be fast sacrificing their accuracy or be precise consuming a significant amount of resources. Here, to overcome such a problem, we present a DFT method that exploits self-consistent determinations of the on-site and inter-site Hubbard interactions (U and V ) simultaneously and obtain band gaps of diverse materials in the accuracy of GW method at a standard DFT computational cost. To achieve self-consistent evaluation of U and V , we adapt a recently proposed Agapito-Curtarolo-Buongiorno Nardelli pseudohybrid functional for U to implement a new density functional of V . This method is found to be appropriate for considering various interactions such as local Coulomb repulsion, covalent hybridization and their coexistence. We also obtained good agreements between computed and measured band gaps of low dimensional systems, thus meriting the new approach for large-scale as well as high throughput calculations for various bulk and nanoscale materials with higher accuracy. |
Thursday, March 18, 2021 10:36AM - 10:48AM Live |
R19.00014: Reverse-Engineering the Exchange-Correlation hole for the SCAN and r2SCAN Functional Luis Lopez Macias, John Perdew, Jianwei Sun The success of density functional theory (DFT) of electronic structure is dependent on the development of accurate and efficient density functional approximations (DFAs). These DFAs are limited by the approximation of its exchange-correlation (XC) energy which can be defined by the XC hole. Understanding this XC hole has played a vital role developing DFAs and explaining their success and limitations. We present a construction of the XC hole model for SCAN [1] and the recent r2SCAN [2] by reverse engineering from known exact hole constraints. The hole models are tested for atoms and simple molecules. |
Thursday, March 18, 2021 10:48AM - 11:00AM Live |
R19.00015: Beyond-DFT database of spectral function for correlated materials Subhasish Mandal, Kristjan Haule, Karin M Rabe, David Vanderbilt Recent trends in condensed matter physics greatly rely on the database and data science-driven materials discovery. The existing materials databases, constructed in the spirit of the materials genome initiative, are built almost exclusively by DFT engines and are very often making incorrect predictions for many correlated materials. As for qualitative predictions of excited-state properties usually require beyond-DFT methods, various methods going beyond DFT, such as meta-GGAs, hybrid functionals, GW, & DMFT have been developed to describe the electronic structure of correlated materials, but it is unclear how accurate these methods can be expected to be when applied to a given material. It is thus of pressing interest to compare their accuracy as they apply to different categories of materials, and at the same time, to build up a database for beyond-DFT methods [1]. We discuss a systematic study of these methods on various training sets of moderately and strongly correlated materials starting from elemental metallic systems to Fe-pnictides and chalcogenides, and various perovskites and compare with experimental photoemission data where available. |
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