ABINIT is an open-source software distributed under GNU-GPL licence which is designed for the calculation of various properties of a system composed of atomic nuclei and electrons within Density Functional Theory (DFT). DFT is an ab initio theory that does not require any parameters from experiments. The development of ABINIT is a collaborative international project [1] involving several developers in France (1), Belgium (2,3), Canada (4,5), Spain (6), USA (7), etc. The Laboratoire Matière en Conditions Extrêmes (LMCE) is one of the principal group contributing to this project participating since more than 20 years to the development of the code.
ABINIT describes the electronic wavefunctions on a plane waves basis, without making any assumption on the nature of the materials to be described. It consists of several modules, interacting with each others, each implementing a formalism derived from DFT. First and foremost, the code allows to compute the ground state of electrons, which gives access to the total energy, the interatomic forces and the stress tensor of the system under study. Those quantities are then used in static studies (ie. atomic nuclei at fixed positions), but also dynamical ones, to compute different thermodynamic properties of matter.
The code can also compute the responses of a system to different perturbations, giving access to various properties such as vibrational spectra, elastic constants or the dielectric tensor. A specific module also implements the many-body perturbation theory (MBPT) which can be used to obtain excited states of matter. The Dynamical Mean Field Theory (DMFT), which is an extension of DFT particularly efficient to describe strong correlation effects between electrons, has also been implemented at the LMCE.
Several members of the LMCE are involved in many developments in the ABINIT code regarding in particular the optimization of performances on supercomputers [2], the “Projector Augmented-Wave” (PAW) formalism [3], the description of electronic correlations (DFT+U method and DMFT) [4], the computation of response functions using Density Functional Perturbation Theory (DFPT) [5], the description of excited states in time-dependent DFT (TDDFT), spectroscopy calculations or the description of finite temperature effect in materials.
At the LMCE, ABINIT is mostly used to predict the behavior of matter at the microscopic scale in conditions that are not experimentally accessible, typically at very high temperature and/or pressure, and to compare to experimental results in order to mutually validate the experimental setup and theoretical models. In particular, one can mention the computation of equations of states of matter from the solid to the plasma state, the determination of the atomic structure of materials subjected to high stresses, the study of diffusion of a chemical specie in solids, the determination of phase diagrams (including solid-liquid transitions), …
Figure 1:
Scalability of ABINIT as a function of the number of CPUs. Black curve: Chebyshev filtering algorithm using MPI parallelisation. Red curve: LOBPCG algorithm (block conjugated gradient) using hybrid hybride MPI+openMP parallelisation. Test case: 1960 atoms of Gallium oxyde (8700 electronic states).
Figure 2:
Adiabatic transfer of an oxygen -type hole polaron in the BaSnO3 perovskyte (charge transfer insulator). The charge localisation of the polaron requires the use of DFT+U. The minimum energy path has been obtained using the Nudge Elastic Band (NEB)
Figure 3:
X-ray absorption near edge structure spectra (XANES) of gold. Properly accounting for spin-orbit coupling in the ABINIT calculations (full lines) is necessary to retrieve the experimental results (dash line). The colors present correspond to different types of electronic transitions.
Publications
- A. H. Romero, D. C. Allan, B. Amadon, G. Antonius, T. Applencourt, L. Baguet, J. Bieder, F. Bottin, J. Bouchet, E. Bousquet, F. Bruneval, G. Brunin, D. Caliste, M. Côté, J. Denier, C. Dreyer, P. Ghosez, M. Giantomassi, Y. Gillet, O. Gingras, D. R. Hamann, G. Hautier, F. Jollet, G. Jomard, A. Martin, H. P. C. Miranda, F. Naccarato, G. Petretto, N. A. Pike, V. Planes, S. Prokhorenko, T. Rangel, F. Ricci, G.-M. Rignanese, M. Royo, M. Stengel, M. Torrent, M. J. van Setten, B. Van Troeye, M. J. Verstraete, J. Wiktor, J. W. Zwanziger, X. Gonze, “ABINIT: Overview and focus on selected capabilities”, J. Chem. Phys., 152, 124102 (2020) DOI
- A. Levitt, M.Torrent, “Parallel eigensolvers in plane-wave Density Functional Theory”, Comp. Phys. Comm., 187, 98 (2015) DOI
- M. Torrent, F. Jollet, F. Bottin, G. Zérah, X. Gonze, “Implementation of the projector augmented-wave method in the ABINIT code: Application to the study of iron under pressure”, Comp. Mat. Sci., 42, 337 (2008) DOI
- B. Amadon, “First-principles DFT+DMFT calculations of structural properties of actinides: Role of Hund’s exchange, spin-orbit coupling, and crystal stucture”, Phys. Rev. B, 94, 115148 (2016) DOI
- A. Martin, M. Torrent, R. Caracas, “Projector Augmented-Wave Formulation of Response to Strain and Electric-Field Perturbation within Density Functional Perturbation Theory”, Phys. Rev. B, 99, 094112 (2019) DOI
Researchers involved
B. Amadon, L. Baguet, R. Béjaud, A. Blanchet, F. Bottin, F. Brieuc, F. Gendron, G. Geneste, F. Jollet, V. Recoules, M. Torrent