Kwant-discuss February 2021

kwant-discuss@python.org

15 participants
17 discussions

LDOS on semi-infinite topological insulator
by Gerson J. Ferreira 13 Feb '21

13 Feb '21

Is it possible to obtain the edge or surface states of a TI from its LDOS on a semi-infinite slab? I'm attaching a single lead to a scattering region consistent of a single layer of the same H as the lead, then I ask for the LDOS at the scattering region, which would be the surface of the semi-infinite system. The problem is that kwant uses the propagating modes of the lead to calculate the LDOS, and consequently I only get the surface states on energy ranges where the bulk bands exist, but I'm interested in the LDOS at the surface and within the bulk gap.

1 0

Dear Kwant developers
by Marc Vila 11 Feb '21

11 Feb '21

Dear Kwant developers and community, Short version: in the source code of Bands module, it is said that one can get the eigenvectors using the argument "return_eigenvectors = True". However, this gives me an error. Is there currently a way to obtain the eigenvectors using Bands? Longer verion + context: I want to obtain the eigenvalues and eigenvectors of a quasi-1d system with a given width. For 2D, I've used wraparound and hamiltonian_submatrix to then diagonalize the Hamiltonian and obtain both the energies and eigenvectors. However, for a quasi-1d system, where one translational symmetry of the system exists, this does not work. Either doing wraparound with "keep = 0" (for example) from a system with 2 translational symmetries or simply finalizing a system with only 1 translational symmetry produces a system which is width independent. I also tried to use the lead's hamiltonian with this approach (https://mail.python.org/archives/list/kwant-discuss@python.org/thread/7FV3F…) but I was not successful (and I felt there should be an easier way). So then I thought about the Bands module (https://kwant-project.org/doc/1/reference/generated/kwant.physics.Bands#kwa…) which gives the energies of the bands (that I knew already and been using it for a long time) but in principle does not give the eigenvectors. I've checked the source code and in there it says that the eigenvectors can be added to the output of Bands with the argument "return_eigenvectors = True". However, this does not work and python says that this is not a valid keyword argument. In the same post as before (https://mail.python.org/archives/list/kwant-discuss@python.org/thread/7FV3F…) it is said that such "return_eigenvectors = True" was forgotten to be included in Kwant's version 1.3 and would be added in version 1.4. However, I'm using verion 1.4.2 and this is not the case. Will this be possible in the future and how can I obtain the eigenvectors of a quasi-1d system? Thank you! Marc ------------------------ Marc Vila Tusell Catalan Institute of Nanoscience and Nanotechnology (ICN2) - Theoretical and Computational Nanoscience Group Additional information: https://scholar.google.es/citations?user=h2V4iNIAAAAJ&hl=es http://icn2.cat/en/theoretical-and-computational-nanoscience-group https://www.researchgate.net/profile/Marc_Vila_Tusell https://www.becarioslacaixa.net/marc-vila-tusell-BI00042?nav=true https://orcid.org/0000-0001-9118-421X

2 1

How to kwant.discretize this hamiltonian?
by Victor Regis 06 Feb '21

06 Feb '21

Hi all, I need to use the following hamiltonian to test if the rest of my code is correct, however I haven't been able to implement it in Kwant. Reading the examples, all the discretize ones are for polynomial dependences on k but mine isn't. H = sin(k_x)*sigma_x +sin(k_y)*sigma_y +B*(2+M-cos(k_x)-cos(k_y))*sigma_z When I implement it on a square grid I obtain the following output: # Discrete coordinates: x y # Onsite element: _cache_0 = ( array([[ 1.+0.j, 0.+0.j], [ 0.+0.j, -1.+0.j]])) _cache_1 = ( array([[-1.+0.j, 0.+0.j], [ 0.+0.j, 1.+0.j]])) _cache_2 = ( array([[-1.+0.j, 0.+0.j], [ 0.+0.j, 1.+0.j]])) _cache_3 = ( array([[ 2.+0.j, 0.+0.j], [ 0.+0.j, -2.+0.j]])) _cache_4 = ( array([[0.+0.j, 1.+0.j], [1.+0.j, 0.+0.j]])) _cache_5 = ( array([[0.+0.j, 0.-1.j], [0.+1.j, 0.+0.j]])) def onsite(site, B, M, cos, k_x, k_y, sin): _const_0 = (cos(k_x)) _const_1 = (cos(k_y)) _const_2 = (sin(k_x)) _const_3 = (sin(k_y)) return (B*M) * (_cache_0) + (B*_const_0) * (_cache_1) + (B*_const_1) * (_cache_2) + (B) * (_cache_3) + (_const_2) * (_cache_4) + (_const_3) * (_cache_5) My issues are: 1. The constants are just my sines and cosines of k_x/k_y, so what happened here? 2. I think because of that it can't find the hopping terms; 3. It doesn't plot any lattice, however if I set hamiltonian = "k_x+k_y" i does plot the square lattice. I know the hamiltonian can be linearized in k, but if possible I want to implement it as it is. Thanks in advance.

2 1

Installation Kwant
by alessandro.lodi＠materials.ox.ac.uk 03 Feb '21

03 Feb '21

Hi guys, I'm running kwant on WSL2 Ubuntu 20.04 and I followed along with your installation procedure. I've tried to follow along with the examples you provide on the website and the workshop you had a couple of months ago and when it's time to calculate smatrix.transmission(1, 0) for a very simple 1-D scattering problem it doesn't. When I run the python test I get a few errors back, mainly from the test_mumps.py source. I understand it's a bit vague said in this way but I have tried to copy and paste the screenshot of my error and the text file but it seems the post window doesn't allow such thing. I'm a beginner, so any help would be appreciated. Thanks a lot,

3 2

KWANT for polarization dependent photocurrent in topological insulators?
by david.j.thames＠gmail.com 02 Feb '21

02 Feb '21

Hello I am particularly interested in studying polarization dependent photocurrent in topological insulators. I came across several papers using nonequilibrium Green's function (NEGE) method to do this topic. I just attached them for your information (Nature Communications: 8, 1037 and Physical Review Research 2, 023055, 2020). I am a novice for KWANT. I wonder if it can be applied to calculate polarization dependent photocurrent in topological insulators (for example to reproduce the figure 1b and 2b in Nature Communications: 8, 1037 ). If so, are there any examples to demonstrate the computational protocol? Thank you, David

1 0

Eigenvectors of H(k) for different k values
by Nayra Alvarez 02 Feb '21

02 Feb '21

Hello :) I'm trying to calculate the Berry curvature. Is there a way to get access to the eigenvectors of H (k) for different k values just diagonalizing the following hamiltonian H_k? -----------Code------------------------------------------------------------------------------- wrapped = kwant.wraparound.wraparound(sys).finalized() def ham(sys,k_x,k_y=None, **params): k= momentum_to_lattice(sys, [k_x] if k_y is None else [k_x,k_y]) p=dict(zip(sys._momentum_names,k),**params) return sys.hamiltonian_submatrix(params=p,sparse=False) H_k= ham(wrapped,kx,ky) ------------------------------------------------------------------------------------------------------ Thanks for your time and help! Nayra

2 4

Figsize when ax is enabled in kwant.plot
by abbout.adel＠gmail.com 01 Feb '21

01 Feb '21

Dear All, I am trying to plot the density on top of the plot of the current. I use for that "ax" to combine the two plots. I kind of noticed that the two plots do not have the same size (although the system is the same!) I know that when "ax" is not None, "figsize" is disabled. Is there a way to fix the size of the two plots to be the same such that they can be combined? Regards, Adel

2 1

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

Kwant-discuss February 2021