Electron density of states anomalies in disordered gold, silver and indium films

Critical Test for Altshuler-Aronov Theory: Evolution of the Density of States Singularity in Double PerovskiteSr2FeMoO6with Controlled Disorder

“…Altshuler and Aronov [2] studied the effect of electron-electron interaction in a disordered metallic medium, and predicted that the density of states (DOS) near the Fermi level (E F ) shows a singularity of |E − E F | 1/2 and the DOS at E F increases with increasing temperature in proportion to √ T . The theory has been applied to the low temperature conductivity of disordered metals such as disordered Au and Ag films [3], amorphous alloy Ge 1−x Au x [4], and transition metal chalcogenide Ni(S,Se) 2 [5].In a previous work, Sarma et al [6] have reported photoemission (PES) measurements on B-site disordered perovskites LaNi 1−x M x O 3 (M =Mn and Fe), which show MIT as a function of x, and shown that the disorder affects the DOS near E F in such a way that had been theoretically predicted by Altshuler and Aronov [2]. In a similar B-site substituted transition-metal oxide SrRu 1−x Ti x O 3 , which demonstrates MIT at x ∼ 0.3 (SrRuO 3 is metallic), the depletion of the DOS near E F has shown an unusual |E − E F | 1.2 dependence in both metallic and insulating phases [7].…”

“…Altshuler and Aronov [2] studied the effect of electron-electron interaction in a disordered metallic medium, and predicted that the density of states (DOS) near the Fermi level (E F ) shows a singularity of |E − E F | 1/2 and the DOS at E F increases with increasing temperature in proportion to √ T . The theory has been applied to the low temperature conductivity of disordered metals such as disordered Au and Ag films [3], amorphous alloy Ge 1−x Au x [4], and transition metal chalcogenide Ni(S,Se) 2 [5].…”

States-conserving density of states for Altshuler-Aronov effect: Heuristic derivation

“…The question is how to modify equation (8) to be valid also beyond the low energy limit. A direct self-consistent solution of the problem would be difficult.…”

“…Figure 3 shows the G(V ) curves measured [10] at 5 K and 25 K, normalized by G(0) at 5 K. The only difference between these curves is a pronounced AA singularity at 5 K and almost no AA singularity at 25 K, otherwise both curves coincide. Thus, the G(V ) curve at 25 K can play the role of G 0 (V ) and we can determine ρ(ǫ)/ρ 0 (0) [8,9] at 4.2 for indium (left) and T l0.7T e0.3 (right). The spectrum on the left originates from figure 1 of work [8], the spectrum on the right from figure 2 of work [9].…”

“…In Refs. [1,17] the integral in equation (8) was calculated assuming V (q) ≃ V (0). For V (q) = V (0) the integral diverges in the upper limit, therefore, the upper limit was restricted to q max = | ǫ | / D. Since V (q) ≃ V (0) only for q k s , the obtained result [equation (11)] holds only for | ǫ | Within this approach the term ρ(E F ) remains undetermined, it is usually determined experimentally [14,15].…”