Nonconstant electronic density of states tunneling inversion for A15 superconductors:Nb3Sn

Abstract

We reexamine the tunneling data on A15 superconductors by performing a generalized McMillan-Rowell tunneling inversion that incorporates a nonconstant electronic density of states obtained from band-structure calculations. For Nb 3 Sn, we find that the fit to the experimental data can be slightly improved by taking into account the sharp structure in the density of states, but it is likely that such an analysis alone is not enough to explain completely the superconducting tunneling characteristics of this material. Nevertheless, the extracted Eliashberg function displays a number of features expected to be present for the highest-quality Nb 3 Sn samples.

Specific heat ofNb3Sn: The case for a second energy gap

“…However, more recent high quality data on superconductor/insulator/superconductor junctions exclude gap anisotropy and give 2⌬͑0͒ / k B T c = 4.1. 30,48 Furthermore the measured specific heat is not compatible with a standard anisotropy for which ⌬͑k ជ ͒ is described by an ellipsoid.…”

“…According to the decomposition of the total DOS at the Fermi level into symmetryprojected components given by Mattheiss and Weber for the cubic phase, 51 83.4% of the bare DOS originates from the Nb d bands, in particular d͑ ͒ and d͑ ͒, the remaining part being essentially due to the Nb p bands. Taking into account the large renormalization 1 + Ϸ 3 for d states, 30 and assuming 1 + Ϸ 1.2 for the other ones, the fraction of the renormalized DOS which originates from the Nb d bands becomes Ϸ92.6%. This is quantitatively consistent with the experimental ratio 1 − x = 0.925 (Table II), and gives support to a scenario in which the superconducting coupling originates from the Nb d bands, while a minor gap or several small gaps are induced in the s and p bands by interband scattering or Cooper pair tunneling (for a review see Ref.…”

“…[46][47][48] Surface values may differ from bulk values; however it would be interesting to include two gaps in the inversion algorithm of the tunneling spectra to see how it would affect the results. 30 The effect of the field can only be discussed at a qualitative level. In analogy with MgB 2 , we define two values of the coherence length which are relevant for the carriers of the bands labeled S and L, respectively: …”

“…Above T Ϸ 0.3⌰ D , i.e., in the upper half of the temperature range of the present measurements, the electron-phonon renormalization contribution is expected to vanish, therefore we subtract an electronic specific heat reduced by a factor of 1 + Ϸ 3. 29,30 In practice, the uncertainty on the electronic contribution remains relatively unimportant over a broad temperature range. Anharmonic corrections are neglected.…”

Interplay between lattice dynamics and superconductivity in Nb3Sn thin films

“…It should be noted that α 2 (E) is not determined in a direct way but rather obtained by calculating the ratio between a PDOS measured by neutron scattering and an Eliashberg function α 2 F measured by electron tunneling on different samples 32,33 . There is in general still uncertainty on the precise shape of α 2 (E).…”