Ab initio molecular dynamics for liquid metals

Аб initio молекулярная динамика для жидких металлов
Georg Kresse, J. Häfner
1993-01-01

Hellmann-Feynman forcesab initio molecular dynamicsconjugate-gradient energy minimizationlocal-density approximationsubspace alignment
We present ab initio quantum-mechanical molecular-dynamics calculations based on the calculation of the electronic ground state and of the Hellmann-Feynman forces in the local-density approximation at each molecular-dynamics step. This is possible using conjugate-gradient techniques for energy minimization, and predicting the wave functions for new ionic positions using subspace alignment. This approach avoids the instabilities inherent in quantum-mechanical molecular-dynamics calculations for metals based on the use of a fictitious Newtonian dynamics for the electronic degrees of freedom. This method gives perfect control of the adiabaticity and allows us to perform simulations over several picoseconds.
1
Ab initio quantum-mechanical molecular-dynamics for liquid metals is performed by calculating the electronic ground state and Hellmann-Feynman forces at each MD step within the local-density approximation.
2
Conjugate-gradient energy minimization combined with subspace alignment prediction of wave functions enables efficient ground-state calculations for new ionic positions.
3
The method provides precise control of adiabaticity and permits stable simulations extending over several picoseconds.
4
The presented approach avoids instabilities inherent in methods that use fictitious Newtonian dynamics for electronic degrees of freedom in metals.

Liquid metals simulated with ab initio quantum-mechanical molecular dynamics

Accurate adiabatic ab initio molecular-dynamics simulation methodology based on electronic ground-state calculation and Hellmann–Feynman forces (using conjugate-gradient energy minimization and subspace alignment) enabling stable picosecond-scale simulations

Publication Details
Publication Date
1993-01-01
Journal
Publisher
ISSN
Access Type
Author Information
Authors
Georg Kresse
J. Häfner
Explore further
Open the scid.ai AI chat with a ready-made request: it will find papers on a similar topic and help build a literature review.
Find similar papers in the chat
Make a presentation
100%