Dear Henry,

Attaching a superconducting lead to a system and attaching a finite superconducting part will usually lead to different results, especially for resonant systems. The only situation the two can lead to slightly similar results is when the Fermi energy lies in the gap of the superconductor. In that case, the lead will be equivalent to a slab of length equal to the penetration length. In the two cases, the conductance from left to right is zero.

I hope this helps,

Adel

On Thu, Mar 18, 2021 at 5:01 PM Henry Axt <henry.axt@gmail.com> wrote:

This is more of a physical question.

I am considering a system set up of a superconductor connected to a non-superconducting metal. As in: [Lead][Conductor][Superconductor]
In Kwant I set the superconductor up with an electron lattice and a hole lattice.

I can simulate this is two ways; one is the superconductor as a lead, and the other is the superconductor as part of the scattering system.

So, [Lead][Conductor Region][Superconducting Lead] or [Lead][Conductor Region][Superconductor Region]

If I look at the conduction between the electron and the hole lead on the left leads, I simulate slightly different conduction values for the case of the SC in the scattering region and the SC as a lead on the right side.

Physically, as I understand it, there should be basically no difference at least at low enough energy values. Is the slight difference due to this "low enough energy" threshold, or am I missing something physical or numerical from kwant that would cause this difference to exist regardless?