Experimental band structure of potassium as measured by angle-resolved photoemission

Abstract

The bulk band structure of potassium along the [110) direction was measured using angleresolved photoemission from an epitaxial potassium film several thousand angstroms thick grown on a Ni(100) substrate. We find the occupied bandwidth to be 1.60+0.05 eV, which is narrower than the free-electron bandwidth of 2.12 eV and agrees with recent calculations of the quasiparticle self-energy. A narrow peak near the Fermi level which did not disperse with photon energy was observed for photon energies which, according to the nearly-free-electron model, should yield no direct transitions. A comparison of the binding energy and intensity of the anomalous peak as functions of photon energy is made to the calculations of Shung and Mahan [Phys. Rev. B 38, 3856 (1988)]. The discrepancies found are discussed in terms of an enhanced surface photoeffect in the photon energy range 20~%co~30 eV. For low photon energies, a bulk peak was also observed due to a surface umklapp process with an intensity comparable to the standard bulk (110) peak. The possible contributions to this strong surface umklapp process from a shear instability at the first few (110) atomic planes is discussed.

Shear displacement of the K(110) surface

“…Our study of the K(110) surface structure was motivated by the results of a previous angle-resolved photoemission study of the K(110) valence band [2]. An anomalously intense surface umklapp peak was observed, indicating the possible existence of a significant distortion of the surface structure from bulk termination.…”

“…S. Itchkawitz, (,) A. P. Baddorf, (2) H. L. Davis, (2) and E. W. Plummer (,) (,) Physics Department, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6396 i2) Solid State Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6032 (Received 9 July 1991) A low-energy electron diffraction analysis of the clean K(l 10) surface has been performed at 25 K, providing a quantitative structural determination of a clean alkali-metal surface. The K(110) surface exhibits a novel surface phase consisting of a lateral shear displacement between the first two surface planes while preserving their two-dimensional periodicity.…”

Vertex corrections in localized and extended systems

“…4͒ calculations only gave a narrowing of about 10% for a homogeneous electron gas of the same mean density, indicating a large impact of further many-body effects. This led to additional experimental and theoretical investigations [16][17][18][19][20][21] but the issue remains controversial. [22][23][24][25] For individual atoms, GW quasiparticle properties have been investigated previously by Shirley 31 These studies have shown that G 0 W 0 RPA in general gives quasiparticle properties which are much improved over DFT and Hartree-Fock methods and that, when calculated selfconsistently, GW also provides reasonably good total energies for atoms ͑with differences versus highly accurate reference methods being on the order of tens of millihartree/ electron͒.…”

Electronic Band Structure of Gd: A Consistent Description

“…Even fundamental questions such as the accuracy of the band structure methods for the ground-state description or the influence of many-body effects on the excitation spectra have not yet found an answer. In fact, band structure calculations based on the local spin density approximations (LSDA) predict a large bandwidth for the (5d6s) states in rare earths similar to that of the preceding elements (alkali and alkaline earth metals) in the Periodic Table. These latter cases [2][3][4][5] exhibit strongly dispersive photoemission features that are reasonably well interpreted within the paradigms based on effective single-particle approximations, although a quantitative description requires detailed consideration of many-body effects. In sharp contrast, angle-resolved photoemission from lanthanides does not display the expected dispersive behavior of the (5d6s)-bulk states.…”