Dear Octaviors, You can solve this by including few cells from the lead in the definition of the system and then attach the lead.
I hope this helps Adel On Wed, Aug 12, 2020 at 6:47 PM <octavi...@gmail.com> wrote: > I am using two leads (left and right) where their widths are larger the > scattering region. For instance, the sites of the scattering region are > between the vertical interval y1 <= y <= y2, so its width is W = y2-y1 +1. > The > sites of the leads are of course between 0 <= y <= W_lead. So, W < W_lead. > The problem is that attach_lead doesn't work, unless one includes sites in > the scattering region for 0<= y < y1, y2<y<=W_lead. Thus I did not define > hoppings for those zones, however attach_lead() imposes spurious horizontal > hoppings at the frontiers, otherwise it gives an error message. it seems > that attach_lead needs to connect the whole frontier of any lead. Is there > any way for avoiding these spurious horizontal hoppings ?. > -- Abbout Adel
