Exchange Contributions and Plasmon Spectra of One‐Dimensional Systems

Abstract

(b), and J. RWALDS (b) The dynamical exchange decoupling method for including exchange in the dielectric function of the electron gas is applied to one-dimensional systems with a tight-binding energy band. Depending on the bandwidth and the position of the Fermi energy in the band, dynamical exchange effects are shown to have appreciable effects on the plasmon dispersion. For propagation of the plasmon along the strand axis, and in the long-wavelength limit, exchange effects drastically lower the slope of the plasmon dispersion. For non-parallel incidence, the exchange effects are less pronounced, and extended over a smaller wave vector region.Die dynamische Austauschentkopplungsmethode fur die Einbeziehung des Austauschs in die dielektrische Funktion des Elektronengases wird auf eindimensionale Systeme mit einem ,,tightbinding"-Energieband angewendet. Es wird gezeigt, daI3 abhjingig von der Breite der Binder und der Lage der Fermienergie im Band die dynamischen Austauscheffekte einen betrichtlichen Einflu13 auf die Plasmonendispersion haben. Fur die Ausbreitung eines Plasmons in Richtung der ,,Strand"-Achse und im Grenzfall lsnger Wellen erniedrigen die Austauscheffekte drastisch den h s t i e g der Plasmonendispersion. Fur nichtparallelen Einfall sind die Austauscheffekte geringer ausgepragt und uber einen kleineren Bereich des Wellenvektors ausgedehnt.

Plasmon excitations in quasi-one-dimensionalK0.3MoO3

“…However, the latter has not been treated theoretically up to now with respect to the plasmon excitation modes for our case of a quasi-onedimensional electron system represented by coupled 1D strands. 10,40,41 We deliberately desist from the inclusion of both a proper treatment of the crystal local-field effects and of corrections for exchange 32,37,38,50,51 and correlation 37,51,52 to the RPA in order to bring into sharper focus the main physical processes controlling the observed plasmon dispersion in this compound.…”

“…The central difference in the conclusions of the various RPA-based theories concerns the energy of the plasmon excitations perpendicular to the chains, and originates solely from the level at which they treat the coupling between the 1D chains. If besides the interchain Coulomb interaction an additional coupling through a transverse electron hopping t Ќ is forbidden, an angular dependence given by cos() is expected in the long-wavelength limit, 32,[36][37][38][39] whereas for t Ќ 0 the plasmon energy remains finite even for ϭ90°. 30 This statement holds irrespective of any corrections for exchange and/or correlation.…”

“…A cos dependence for the plasmon energy was thereby arrived at, as predicted from all theoretical models neglecting the interchain coupling as far as mediated by a finite transverse electron-hopping probability. 32,[36][37][38][39] Although this procedure may be regarded as a reasonable attempt to extract some information from the optical data, it is by no means a direct probe of the angular dependence of the plasmon modes which, in any case, cannot be excited by transverse electromagnetic waves, i.e., in an optical experiment.…”