LATTICE DYNAMICS OF METALS (ION DISTORTION, PHASE EXCITATION)

YAO RONG WANG, Purdue University

Abstract

Lattice dynamics of potassium is employed to study the exchange and correlation of a conduction electron gas. A longstanding but yet unresolved problem is whether or not the exchange and correlation correction G(,+)((')q) to the dielectric function of an electron gas has a sharp peak at q = 2k(,f). Using pseudopotential theory and a model dielectric function that has a peak in its G(,+)((')q), we find that the pho- non frequencies of potassium in a certain region of reciprocal space are altered in a unique way (by the peak) and that this effect can be resolved by neutron scattering experiments. We suggest that the question of whether or not there is a prominent peak in G(,+)((')q) can be settled by carrying out inelastic-neutron scattering experiments on potassium to measure phonon dispersion curves along a 0.28,0.14,(xi) trajectory in reciprocal space. Electron clouds of large atoms undergo significant distortion during lattice vibration. The shell model, which divides each ion into an inner core and a last-filled electron shell, is combined with the pseudopotential method to provide a theory of phonon spectra. The dynamic pseudopotential theory is applied to the lattice dynamics of twelve cubic metals. The proposed core and shell pseudopotentials involve just one parameter each; and these are determined for Na, K, Rb, Cu, Ag, Au, Ca, Sr, Ba, Yb, Al and Pb. Distortion of the ions is found very important in the heavy metals. A recently discovered anomaly in the phonon dispersion of Ba can be attributed to exchange interactions between the (last-filled) shell and the conduction-electron sea. The (so far unmeasured) spectra for fcc Sr and bcc Yb are predicted. The phase excitation of an incommensurate charge-density-wave (CDW), i.e., a phason, is a special kind phonon. Its effect on nuclear-magnetic-resonance (NMR) of metallic potassium is investigated. Due to its CDW structure, the NMR line of potassium should exhibit significant broadening because the Knight shift depends on the con- duction electron density surrounding each nucleus. Nevertheless, experiments on metallic potassium at 1.5K have not revealed such an effect. We show that the thermal excitation of phasons can cause sufficient motional narrowing to explain this observation. (Abstract shortened with permission of author.)

Degree

Ph.D.

Subject Area

Condensation

COinS