Abstract

The electronic structures of ten solid and liquid metals are investigated experimentally and theoretically, and the results of recent nuclear resonance measurements are correlated with already published data concerning magnetic susceptibility, electrical conductivity and atomic arrangement. It is concluded that the electronic structure of each of the metals investigated, with the notable exceptions of Bi and Ga, does not change appreciably at the melting point. Moreover, in Ga and Bi there is a correlation between a change in the electronic structure and structural rearrangements of neighboring atoms at short range; the other metals considered show neither sort of change. A principal conclusion is that a liquid metal possesses a band structure which is very like that of its solid, provided the short-range structure is not altered during melting. Included are hitherto unpublished values for the nuclear magnetic resonance line shifts in liquid Al, Sn, In, and Bi, and the nuclear spin relaxation times in solid and liquid Ga. The nmr shifts in solid and liquid Al are equal within experimental error. The same is true for Sn. This is in agreement with published results on the alkali metals and Hg, where the corresponding changes are known to be small, suggesting that the electronic structures of the respective solids and liquids are the same. The large quadrupole coupling in Ga, In, and Bi prevents observation of the nmr in the solid state. However, the value of the nmr shift in liquid In is consistent with values of the hyperfine coupling constant and the electronic specific heat at low temperatures, a situation which is true for most of the other metals. Ga and Bi are the exceptions for which the nmr shift in liquids is much larger than the hyperfine and low-temperature data would lead one to expect. This is taken as evidence for a reduced density of states at the Fermi surface in solid Ga and Bi. Confirmation is provided by the fact that the nuclear spin relaxation time of Ga increases sharply when the metal freezes. The measurement of relaxation in solid Ga is significant because the nuclear quadrupole resonance is directly observed and the relaxation process is of the same type, so that a change in its magnitude may also be related to a density of states.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call