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 C19: Density Functional Theory and Beyond ILive
|
Hide Abstracts |
Sponsoring Units: DCOMP DCP DCMP DPOLY Chair: James Furness, Tulane Univ |
Monday, March 15, 2021 3:00PM - 3:12PM Live |
C19.00001: A spectral scheme for Kohn-Sham Density Functional Theory of helical structures Shivang Agarwal, Amartya Banerjee Based on the observation that one of the most successful methods for solving the Kohn-Sham equations for periodic systems - the plane-wave method - is a spectral method based on eigenfunction expansion, we formulate and implement a spectral method designed towards solving the Kohn-Sham equations for helical structures. Various important technological materials such as nanotubes (of arbitrary chirality), nanowires, nanoribbons and miscellaneous chiral structures from chemistry and biology constitute examples of helical structures, and such systems are often associated with fascinating material properties. |
Monday, March 15, 2021 3:12PM - 3:24PM Live |
C19.00002: Dissociation limit and the scaled-down self-interaction correction Santosh Adhikari, Biswajit Santra, Kushantha Withanage, Koblar Alan Jackson, Adrienn Ruzsinszky The Perdew-Zunger self-interaction correction (PZ-SIC), when applied to a (semi)-local density functional approximation, removes spurious self-interaction error (SIE) on an orbital-by-orbital basis, making the functional exact for all one-electron systems. However, PZ-SIC is overcorrecting for many-electron systems. There were several schemes introduced in the past to scale-down PZ-SIC in many-electron regions. These so-called exterior scaling schemes were able to predict better equilibrium properties like atomization energies, first ionization energies, etc. compared to regular PZ-SIC. The problem of fractional-charge dissociation of heteronuclear molecules and charged molecular ions fixed by unscaled PZ-SIC methods returns with these exterior scalings. A recently introduced scheme [1] based on the interior scaling of PZ-SIC shows promising results for equilibrium properties like first ionization energies, electron affinities, atomization energies, barrier heights of chemical reactions, etc. In this work we are investigating the behavior of this recently proposed scheme on the dissociation of heteronuclear molecules. |
Monday, March 15, 2021 3:24PM - 3:36PM Live |
C19.00003: Kinetic energy density using machine learning for orbital-free density functional calculations Mohammed Al Ghadeer, Abdulaziz Al-Aswad, Fahhad Alharbi Inspired by the remarkable ongoing progress of the data-driven science approach, a predictive model is prepared to develop accurate one-dimensional kinetic energy density functionals (KEDF) using Machine Learning (ML). Starting from possible analytical forms of kinetic energy density [1,2] and by utilizing a variety of solvable models, an accurate Linear Regression model is statistically trained to estimate the kinetic energy as functionals of the density. The mean relative accuracy for even a small number of randomly generated potentials is found to be better than the standard KEDF by several orders of magnitudes. As more different potentials of model problems are mixed, the coefficients of the linear model significantly approach the analytic values of Thomas-Fermi (TF) and von Weizsäcker (vW), suggesting the reliability of the statistical training approach. This work can provide an important step toward more accurate large-scale orbital free density functional theory (OFDFT) calculations. |
Monday, March 15, 2021 3:36PM - 3:48PM Live |
C19.00004: Machine learning accurate exchange and correlation functionals of the electronic density Sebastian Dick, Marivi Fernandez Here, we review recent efforts to use machine learning (ML) methods for the creation of density functionals. We showcase our own framework, NeuralXC, which is based on a projection of the electron density onto localized atomic orbitals and a functional parametrized by neural networks. The functionals thus created are designed to lift the accuracy of a baseline method towards that provided by more accurate reference calculations, all while maintaining their efficiency. We show that a meaningful representation of the physical information contained in the training data is learned, making the functionals transferable across systems. Challenges on the path to a truly universal ML-functional are outlined and possible future approaches are discussed. |
Monday, March 15, 2021 3:48PM - 4:00PM Live |
C19.00005: Data-Driven Exchange-Correlation Functional Design for Transferability and Interpretability Kyle Bystrom, Boris Kozinsky Due to its computational efficiency compared to other quantum mechanical methods, Density Functional Theory (DFT) is a popular tool in computational chemistry. Recently, machine learning (ML) has been explored as a tool to develop more accurate exchange-correlation (XC) functionals, but more work is required to design ML models which are transferable across chemical space and can be interpreted in the context of conventional functional design. To this end, we introduce two developments to design functionals that are transferable, obey exact theoretical constraints, and have separate exchange and correlation parts. First, we design a Gaussian Process-based exchange-only functional that obeys the uniform scaling rule and approximately matches the homogeneous electron gas limit. Second, we explore the use of the exchange energy density (both exact and ML) as a parameter for the correlation functional, effectively resulting in a more flexible local hybrid without gauge ambiguity issues. The accuracy of these approaches is competitive with semi-empirical functionals and recent ML models for atomization energies, ionization potentials, and barrier heights. |
Monday, March 15, 2021 4:00PM - 4:12PM Live |
C19.00006: New approximations for the exchange-correlation potential using connector theory Ayoub Aouina, Matteo Gatti, Lucia Reining In the Kohn-Sham formulation of density functional theory (DFT) [1], the ground-state density of interacting electrons can be obtained from a fictitious system of independent particles in an effective potential. Even though DFT is in principle exact the effective potential contains an unknown quantity called the exchange-correlation (xc) potential. In this talk we propose a new approximation to the xc potential using a general approach called "Connector Theory" (COT) [2]. This approach is a prescription of how to use data from models to calculate quantities in materials. COT is in principle exact but in practice approximations are needed to make it useful. After introducing the general scheme of this approach, we explain how to use the COT in order to build non-local functionals of the density for the xc potential. We show how well the connector approximation describes the non-local functional and we compare with previous approximations. Finally, after implementing the connector xc potential in the self-consistent Kohn-Sham loop, we discuss the quality of the resulting electronic densities and band structures. |
Monday, March 15, 2021 4:12PM - 4:24PM Live |
C19.00007: Enhancing the accuracy of interior-scaled Perdew-Zunger self-interaction correction Puskar Bhattarai, Biswajit Santra, Kamal Wagle, Yoh Yamamoto, Rajendra R Zope, Koblar Jackson, John Perdew The Perdew Zunger self-interaction correction (PZ-SIC)1 improves the performance of density functional approximations (DFAs) for the properties that involve significant self-interaction error (SIE), as in stretched bond situations, but overcorrects for equilibrium properties. This overcorrection is often reduced by LSIC2, local scaling of the PZ-SIC to LSDA. Here we propose a new scaling factor to use in an LSIC-like approach that satisfies an additional important constraint: the correct coefficient of Z in the asymptotic expansion of the exchange-correlation (xc) energy for atoms of atomic number Z, which is neglected by LSIC. LSIC+ applied to LSDA works better for many properties than LSDA-LSIC and the PBE GGA and gives comparable results to the SCAN meta-GGA. Unlike LSIC and LSIC+, SCAN-sdSIC3, seems to describe weak bonds correctly. However, sdSIC spoils the correct asymptotic behavior of the xc potential. Interior-scaled PZ-SIC with improved iso-orbital indicators might be developed to work more widely than the methods considered here. |
Monday, March 15, 2021 4:24PM - 4:36PM Live |
C19.00008: Accelerate stochastic calculation of random-phase approximation correlation energy difference with atom-based correlated sampling Yu-Chieh Chi, Chen Huang A kernel polynomial method (KPM) is developed to calculate the random phase approximation (RPA) correlation energy. In the method, RPA correlation energy is formulated in terms of the eigenvalues of a matrix that is the product between the Coulomb potential and Kohn-Sham (KS) linear response function. The integration over the eigenvalues is then calculated using KPM. Since it is often the energy difference between two systems that is of much interest in practice, another focus of this work is to develop a method to accelerate the convergence of such energy-difference calculations. The method is termed atom-based correlated sampling (ACS). The performance of ACS is examined by calculating the isomerization energy of acetone to 2-propanol and the energy of water-gas shift reaction. Using ACS, the convergences are accelerated by 3.6 and 4.5 times, respectively. Methods developed in this work are expected to be useful for calculating energy differences between systems that mainly differ in certain local regions, such as calculating adsorption energies of molecules on metal surfaces for surface catalysis. |
Monday, March 15, 2021 4:36PM - 4:48PM Live |
C19.00009: Can the Hartree-Fock kinetic energy exceed the true kinetic energy? Steven Crisostomo, Mel Levy, Kieron Burke Hartree-Fock (HF) theory has remained an important tool for quantum chemical calculations since its earliest appearances in the late 1920s. Despite a near century of development, the sign of the difference between the true and Hartree-Fock kinetic energies remains unclear for most systems. Intuition suggests that the true kinetic energy should always be larger than the HF value, but we know of no proof of this. For atoms or ions, the virial theorem shows that the true kinetic energy is always larger than its HF counterpart. Beginning from a generalized virial theorem derived from density scaling considerations, we derive a general expression for the kinetic energy difference, that applies to all systems. Calculations on Hooke’s atom illustrate this relation and show that the difference of the kinetic energies always remains positive, thereby not providing a counterexample. |
Monday, March 15, 2021 4:48PM - 5:00PM Live |
C19.00010: Self-consistency in the Fermi-Löwdin orbital self-interaction correction method using the Krieger-Li-Iafrate approximation Carlos Diaz, Tunna Baruah, Rajendra R Zope The Perdew-Zunger (PZ) self-interaction correction (SIC) method provides a way to remove one-electron self-interaction errors on an orbital-by-orbital basis. It requires use of local orbitals as use of Kohn-Sham orbitals leads to the size-extensivity problem. Pederson and coworkers have shown that use of Fermi-Löwdin orbitals (FLOs) simplifies implementation of PZ-SIC. In this talk we present a self-consistent implementation of FLO-PZ-SIC using the Krieger-Li-Iafrate approximation (KLI) to the optimized effective potential (OEP) and compare it to Jacobi-like self-consistent implementation of Pederson et al. [1]. Since a single Hamiltonian is diagonalized in FLO-SIC-KLI it, unlike the Jacobi method, also provides a correction to unoccupied orbitals. We compare the results obtained using the FLO-SIC-KLI method with the FLO-SIC-Jacobi scheme for a wide array of properties. Our results show that FLO-SIC-KLI provides comparable results to the FLO-SIC-Jacobi for a wide array of properties. Similar to the differences between Hartree-Fock and exact exchange OEP HOMO-LUMO gaps, we find that the HOMO-LUMO gaps in FLO-SIC-KLI are smaller than FLO-SIC-Jacobi gaps. |
Monday, March 15, 2021 5:00PM - 5:12PM Live |
C19.00011: Assessing Local Hybrid Density Functionals for the Prediction of Exchange Coupling Constants in Transition Metal Complexes Henry Fitzhugh, James Furness, Jianwei Sun The prediction of Spin-Hamiltonian exchange coupling parameters is key for successful modelling of multi-center transition metal complexes (MTMCs) such as single molecule magnets and biocatalyst analogues. Local hybrid functionals, which include a spatially varying proportion of exact exchange, have been proposed as a method for improving the performance of density functional theory (DFT) with these highly magnetic systems. Given DFT is one of the few methods efficient enough for analysis of large MTMCs, it is crucial to understand and improve upon performance and reliability for calculation of coupling parameters and magnetic contributions to energy levels. Here we examine the performance of local hybrid and meta-GGA functionals, including the new r2SCAN functional, for the prediction of coupling parameters in several di-nuclear transition metal complexes. Natural population analysis and qualitative density comparisons rationalize the variation in functional performance and demonstrate the need for further local hybrid functional development. |
Monday, March 15, 2021 5:12PM - 5:24PM Live |
C19.00012: Towards an orbital-free kinetic energy density functional for molecular systems Omololu Akin-Ojo New kinetic energy density functionals (KEDFs) for orbital-free density functional theory (OFDFT) involving many-atoms systems are proposed. These KEDFs utilize the densities and information from properties of the constituent atoms. The performance of these KEDFs are presented for densities calculated from Hartree Fock theory, Kohn-Sham density functional theory and for those obtained by minimizing energy functionals involving these new KEDFs. |
Monday, March 15, 2021 5:24PM - 5:36PM Live |
C19.00013: Building a database of two-dimensional material properties using the SCAN functional Can Ataca, Daniel Wines, Gracie Chaney, Jaron A Kropp, Fatih Ersan The strongly constrained and appropriately normed (SCAN) meta-GGA functional has reportedly performed exceptionally well for density functional theory (DFT) calculations involving different crystalline systems, accurately capturing lattice constants and van der Waals (vdW) interactions. The computational cost of SCAN is less than that of the hybrid functional methods as well. SCAN has been applied to several three-dimensional systems, but has not been widely used for two-dimensional (2D) materials such as transition metal (M) monochalcogenides (MX), M dichalcogenides (MX2), and M trichalcogenides (MX3). We provide a comprehensive set of data obtained by SCAN, hybrid functionals (HSE06), and PBE. Specifically, we compare lattice constants, bandgaps, and cohesive energies. We also study optical properties with the GW approximation and Bethe-Salpeter equation (BSE), using wavefunctions obtained from SCAN and PBE. Our goal is to benchmark these results and create a full database to determine how SCAN performs compared to other well established DFT functionals. This work is the terminal study for benchmarking different DFT functionals and will guide further theoretical studies involving 2D materials and electronic structure method development. |
Monday, March 15, 2021 5:36PM - 5:48PM Live |
C19.00014: Non-additive kinetic potential functional Vnad from analytic inversion: all-electron and pseudopotential calculations Mojdeh Banafsheh, Leeor Kronik, Tim Gould, Tomasz Adam Wesolowski, David A Strubbe The non-additive kinetic potential functional Vnad is a key issue in density-dependent embedding methods, such as Frozen Density Embedding Theory and Partition-DFT. Vnad is a bifunctional of pairs of specific electron densities ρA and ρB. We previously reported the exact analytical inversion procedure to generate reference Vnad for weakly overlapping ρA and ρB (M. Banafsheh, T.A. Wesolowski, Int. J. Quant. Chem. 118 (2018): e25410). We discuss the constraints on the choice of electron densities to ensure their admissibility. The potential is constructed for various diatomic systems of four electrons at different interatomic distances for which the atomic densities are weakly overlapped. The results are compared to common kinetic functional approximations to assess their quality in this regime. Vnad is also presented for some diatomic systems including more than 4 electrons. We investigate the behavior around the nuclei, where cusps appear in the density, and compare results for all-electron and pseudopotential calculations. We demonstrate the application in embedding theory by solving for a subsystem with Vnad as an additional potential. In addition, the well known step structure associated with molecular dissociation is studied from the analytically inverted potential. |
Monday, March 15, 2021 5:48PM - 6:00PM Live |
C19.00015: Investigation of Electronic and Optical Properties of Copper-Cysteamine with Halogens Noura Alkhaldi, Muhammad Huda, Wei Chen, Nil Kanatha Pandey, Manbo Zhang Copper cysteamine (Cu-Cy) is molecular solids with layers structure. It has a strong luminescence that can be used in imaging. Besides, Cu-Cy is a photosensitizer which can be activated by visible light, X-rays, microwaves, and ultrasound to generate reactive oxygen species (ROS) to treat the cancer and infection diseases. Understanding the electronic and optical properties of Cu-Cy is crucial. In this presentation, density functional theory (DFT) is used to study the electronic and optical properties of Cu-Cy-X, with X= F, Cl, Br, I. We have started with studying the stability of Cu-Cy structures considering different states. Different spin-multiplicities along with spin-orbit-coupling (SOC) is considered to understand the electron transition from the occupied to the unoccupied bands. Point defects are made as well in the pristine structures of Cu-Cy to study how those affect the stability of the structures. Our computed data is in good agreement with the experimental results. |
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