Defining shape of leads to avoid generation of extra lattice sites
Hello, I am modeling a graphene lattice that is rectangular in shape. graphene = kwant.lattice.general([(1, 0), (sin_30, os_30)], [(0, 0), (0, 1 / sqrt(3))]) When you attach the leads (with symmetry parallel to the rectangle edges), extra lattice sites get added. I want to avoid this and my thinking is that you define a seperate lattice for the leads, such that it is a unit cell translatable like a square lattice, which in this case would be something like: lattice_lead = kwant.lattice.general([(1, 0), (0, 2*cos_30)], [(0, 0), (0, l(1 / sqrt(3))),(0.5,((1/sqrt(3))+(tan_30*0.5))),(0.5,((2/sqrt(3))+(tan_30*0.5)))]) My question is firstly, is this the way to go about it? And secondly, how does this interplay with the sublattices and thier hoppings?
Managed to find a working solution in the email archive: https://www.mail-archive.com/kwant-discuss@kwant-project.org/msg00172.html sym.add_site_family(graphene.sublattices[0], other_vectors=[(-1, 2)]) sym.add_site_family(graphene.sublattices[1], other_vectors=[(-1, 2)]) where graphene is the honeycomb lattice and sym is for example sym = kwant.TranslationalSymmetry((-1, 0)).
participants (1)
-
Henry Axt