Abstract
Using quantum dot artificial atoms as a simple toy model, we reflect on the question of whether spin density functional theory (SDFT) can accurately describe correlation effects in low-dimensional fermion systems. Different expressions for the local density approximation of the exchange-correlation energy for the two-dimensional electron gas, such as the much-used functional of Tanatar and Ceperley, and the recent suggestion by Attaccalite et al., are compared with the results of a numerical diagonalization of the many-body Hamiltonian matrix in the limit of small electron numbers. For systems with degeneracies, as shown in the present work for the example of a spin triplet with S = 1, the direct comparison with configuration interaction (CI) methods demonstrates that the spin representation of SDFT may, under certain circumstances, produce artificial energy splittings between states that belong to the same spin multiplet. For a singlet ground state with S = Sz = 0, however, the correlation functions of the CI solutions confirm the spin-density wave states found earlier within the SDFT method.
Original language | English |
---|---|
Pages (from-to) | 817-825 |
Number of pages | 9 |
Journal | International Journal of Quantum Chemistry |
Volume | 105 |
Issue number | 6 |
DOIs | |
Publication status | Published - 20 Dec 2005 |
Externally published | Yes |
Keywords
- Configuration interaction calculations
- Density functional method
- Quantum dots
- Spin symmetry
- Spin-density waves