Dear Sumit, The easiest way to calculate Green’s function on arbitrary set of sites in your system is to specify additional virtual lead with zero self energy with following builder class: http://kwant-project.org/doc/1/reference/generated/kwant.builder.SelfEnergyL... Here is an example function, that can do it: def mount_vlead(sys, vlead_interface, norb): """Mounts virtual lead to interfaces provided. :sys: kwant.builder.Builder An unfinalized system to mount leads :vlead_interface: sequence of kwant.builder.Site Interface of lead :norb: integer Number of orbitals in system hamiltonian. """ dim = len(vlead_interface)*norb zero_array = np.zeros((dim, dim), dtype=float) def selfenergy_func(energy, args=()): return zero_array vlead = kwant.builder.SelfEnergyLead(selfenergy_func, vlead_interface) sys.leads.append(vlead) Then you can calculate Green’s gunction on interface of this lead. Basis in that case is specified on order of sites in vlead_interface. If I remember correctly, calculation time in that case scales as n*N*log(N), where N is number of sites in system, n is number of sites to calculate Green’s function in. With best regards, Viacheslav Ostroukh Instituut-Lorentz – Niels Bohrweg 2 – Room 259 – 2333 CA Leiden ostroukh@ilorentz.org slava@ostroukh.me +31 6 444 968 12 +38 099 721 76 06 From: Sumit Ghosh Sent: Tuesday, January 12, 2016 11:44 To: kwant-discuss@kwant-project.org Subject: [Kwant] Green's function for each block I am trying to calculate the Green's function for each layer/slice of a ribbon/wire. I can find the self energy and surface Green's function from which (in principle, I have not tried yet) it is possible find the Green's function for successive layer using RGF algorithm as discussed in this thread (http://thread.gmane.org/gmane.comp.science.kwant.user/647).
From the GreensFunction documentation, it looks like there is a way to specify the slice/block of the system and directly find the Green's function, but I don't find any example of that. Is there any way to do that ? (http://kwant-project.org/doc/1/reference/generated/kwant.solvers.common.Gree...)
For example consider this graphene ribbon. I added a random impurity to break the translational symmetry so that I can distinguish each slice. I can find the Green's function for the sites which are connected to leads only. How do I calculate the Green's function for each slice? Best, Sumit ------------------------------------------------------ from __future__ import division from math import sqrt from matplotlib import pyplot import kwant import numpy as np sin_30, cos_30 = (1 / 2, sqrt(3) / 2) graphene = kwant.lattice.general([(1, 0), (sin_30, cos_30)], [(0, 0), (0, 1 / sqrt(3))]) a, b = graphene.sublattices L=3;W=3; def box(pos): #scattering region x, y = pos return -L<x<L and -W<y<W def lshape(pos): #lead x,y = pos return -W<y<W def onsite(site): return kwant.digest.uniform(repr(site)) hoppings = (((0, 0), a, b), ((0, 1), a, b), ((-1, 1), a, b)) def scater(): sys = kwant.Builder() sys[graphene.shape(box, (0, 0))] = onsite sys[[kwant.builder.HoppingKind(*hopping) for hopping in hoppings]] = -1 lead = kwant.Builder(kwant.TranslationalSymmetry(graphene.vec((-1, 0)))) lead[graphene.shape(lshape, (0, 0))] = 0 lead[[kwant.builder.HoppingKind(*hopping) for hopping in hoppings]] = -1 sys.attach_lead(lead) sys.attach_lead(lead.reversed()) return sys sys=scater() #def family_colors(site): # return 0 if site.family == a else 1 #kwant.plot(sys, site_color=family_colors, site_lw=0.1, colorbar=False) sys = sys.finalized() # Self energy flead0 = sys.leads[0] flead1 = sys.leads[1] s_en = flead0.selfenergy(0.5) #Green's function g = kwant.greens_function(sys,0.5) gg = g.submatrix(1,0) -- Sumit Ghosh Spintronics Theory Group PSE Division KAUST This message and its contents, including attachments are intended solely for the original recipient. If you are not the intended recipient or have received this message in error, please notify me immediately and delete this message from your computer system. Any unauthorized use or distribution is prohibited. Please consider the environment before printing this email.
