Answers are grounded in this paper and cite their sources.
A minimum metallic conductivity ([sgrave]min) was proposed by the present author in 1972. In 1976 he showed that the existence or otherwise of this quantity depended on the index s, defined if the radius of an Anderson localized state at energy E tends to infinity as (E c–E)−s , where E c is the mobility edge. The quantity [sgrave]min exists only if s ⅔ (see also Stein and Krey 1979). Some theoretical treatments give s = ≥⅔, others give 0.6. Recent experiments suggest that there is a very narrow range of energies above E c in which, at T = 0, [sgrave] falls continuously to zero. This paper, therefore, starting from the Kubo–Greenwood formula, examines the consequences of the assumption that s = 0.6. It is found that, if this is so, there is indeed a narrow range in which [sgrave] at T = 0 climbs from zero to the calculated value of [sgrave]min; above [sgrave]min, [sgrave] is proportional to {N(E F)}2 as long as L ∼ a, where a is the distance between electron sites and L the mean free path. The temperature dependence of the conductivity is also discussed. Our conclusions for three dimensions differ from those of other recent approaches, such as those of Abrahams et al. and of Götze. At high temperatures the concept of a minimum metallic conductivity remains valid, and it is still legitimate to write [sgrave] = [sgrave]min exp {-(E c–E F)/kT} for conduction at a mobility edge.
No takes yet. Share an insight, caveat, or question.
N. F. Mott (1981) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: