On Fri, Oct 20, 2017 at 4:45 PM, Dripto Debroy <driptodeb...@gmail.com> wrote: > That makes a lot of sense, would I be correct in assuming that I would need > to define new lattices for each of the rows of square matrices before adding > the hoppings by hand?
You only need to define two lattices: one honeycomb, and one square for the lead. Best, Anton > > Much appreciated. > Dripto > > On Oct 20, 2017 5:52 AM, "Anton Akhmerov" <anton.akhmerov...@gmail.com> > wrote: >> >> Hi Dripto, >> >> The correct approach to implement that is to do the following: >> 1. add a sufficient number of sites from the square lattice to the >> scattering region, so that the lead may be attached to those (this >> should amount to 1 unit cell of the lead). >> 2. manually connect those sites to the honeycomb lattice. Since there >> are too many possibilities to do so, there's no way to make this >> choice automatically. >> 3. attach the lead as usual using Builder.attach_lead. >> >> Best, >> Anton >> >> On Thu, Oct 19, 2017 at 8:31 PM, Dripto Debroy <driptodeb...@gmail.com> >> wrote: >> > Hello all, >> > >> > I am hoping to use KWANT to model a graphene flake with two metal leads. >> > The >> > issue I am running in to is figuring out how to connect leads which have >> > a >> > square lattice and graphene which has a honeycomb lattice. Is this >> > something >> > that KWANT can handle? I was unable to find any documentation on mixing >> > lattice types and modeling heterojunctions in general. >> > >> > Thanks for your help, >> > Dripto