Answers are grounded in this paper and cite their sources.
A order-N method for calculating the electronic structure of general (non-tight-binding) potentials is presented. The method uses a combination of the ``purification''-based approaches used by Li, Nunes, and Vanderbilt, and Daw, and a representaiton of the density matrix based on ``traveling basis orbitals.'' This method gives a total energy form that has the form of a cubic multicomponent Landau theory. The method is applied to several one-dimensional examples, including the free-electron gas, the ``Morse'' bound-state potential, a discontinuous potential that mimics an interface, and an oscillatory potential that mimics a semiconductor. The method is found to contain several physical effects that are hard to obtain in real-space total-energy functionals: Friedel oscillations, quantization of charge in bound states, and band-gap formation. Quantitatively accurate agreement with exact results is found in most cases. Possible advantages with regard to treating electron-electron interactions and arbitrary boundary conditions are discussed.
No takes yet. Share an insight, caveat, or question.
A. E. Carlsson (1995) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: