From rezarasuli at physics.sharif.edu Tue Dec 1 19:53:55 2009 From: rezarasuli at physics.sharif.edu (Reza Rasuli) Date: Tue, 1 Dec 2009 22:23:55 +0330 Subject: [Wannier] Convergence problem of wannierisation In-Reply-To: References:

Message-ID: Dear all Can one answer this please? 1- What errors would a Bloch orbital approach in a periodic cell give, for calculation of conductance, compared to the MLWF? 2- Does the K-mesh in PWSCF input depend on the MLWF basis? Is the convergence alluded to better, compared to a plane-wave basis? I will be grateful if you respond me. -- Reza Rasuli -------------- next part -------------- An HTML attachment was scrubbed... URL: From nicolas.poilvert at gmail.com Tue Dec 1 21:27:27 2009 From: nicolas.poilvert at gmail.com (nicolas poilvert) Date: Tue, 1 Dec 2009 15:27:27 -0500 Subject: [Wannier] Convergence problem of wannierisation In-Reply-To: References: Message-ID: <66bc743b0912011227m368f5638t5400d057046299ad@mail.gmail.com> Dear Reza, For the first question, I don't know if you are talking about methodology or technical details. If it is about methodology, then I would say that the calculation of the Quantum Conductance can be done in 2 approches. The first one would be called "waveguide approach" in which the Bloch states are used to compute the "transmission function" of the system. The second approach is a "tight-binding approach" in which the Hamiltonian matrix in a localized basis is used to compute the same "transmission function". Both approaches are actually mathematically equivalent and so if you would use two different computer codes each focusing on one of the two above mentioned "approach" to compute the conductance, then the answer should be the same! For the second question, I am a little unsure of what you are asking. If by the K-mesh you mean the uniform K-mesh used at the Non-Self-Consistent step before the Wannierisation with Wannier90 then YES, this mesh in PWscf determines the quality of the computed MLWF. If you are referring to the Self-Consistent step, the mesh at this step can be whatever you want, knowing that the important thing to keep in mind is to obtain a well-converged ground state charge density. Cheers, Nicolas On Tue, Dec 1, 2009 at 1:53 PM, Reza Rasuli wrote: > Dear all > Can one answer this please? > 1- What errors would a Bloch orbital approach in a periodic cell give, for > calculation of conductance, compared to the MLWF? > 2- Does the K-mesh in PWSCF input depend on the MLWF basis? Is the > convergence alluded to better, compared to a plane-wave basis? > I will be grateful if you respond me. > -- > Reza Rasuli > > > _______________________________________________ > Wannier mailing list > Wannier at quantum-espresso.org > http://www.democritos.it/mailman/listinfo/wannier > > -- POILVERT Nicolas PhD candidate, Dpt of Materials Science and Engineering Massachusetts Institute of Technology 77, Massachusetts avenue Cambridge, MA 02139 USA work: (617) 452-4212 nicolas.poilvert at gmail.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From jhouska at kfy.zcu.cz Fri Dec 18 09:48:24 2009 From: jhouska at kfy.zcu.cz (Jiri Houska) Date: Fri, 18 Dec 2009 09:48:24 +0100 Subject: [Wannier] Cubefile shift Message-ID: <20091218094824.9g50d7hz44g0804c@webmail.zcu.cz> Dear all, when printing cubefiles using wannier90 and wannier_plot=.true. wannier_plot_supercell = 2 or 3 (did not matter) wannier_plot_format = cube , after visualization (VMD) I realized that atomic positions and the orbital positions are obviously shifted (of non-integer times cell size). The shift appears to be different for different MLWFs. Please is is avoidable and/or is there a known reason for that? Thanks! Jiri Houska University of West Bohemia From a.mostofi at imperial.ac.uk Mon Dec 21 15:40:28 2009 From: a.mostofi at imperial.ac.uk (Arash Mostofi) Date: Mon, 21 Dec 2009 14:40:28 +0000 Subject: [Wannier] Cubefile shift In-Reply-To: <20091218094824.9g50d7hz44g0804c@webmail.zcu.cz> References: <20091218094824.9g50d7hz44g0804c@webmail.zcu.cz> Message-ID: <4B2F88DC.1050302@imperial.ac.uk> Dear Jiri, The routine that writes the cube files translates each MLWF centre to the home unit cell, defined by the volume of space that has fractional crystallographic coordinates in the range [0,1]. But this should always shift the position of the centre by an integer number of cell lengths in each direction as compared to the MLWF centres that are, for example, displayed in the Wannier90 output file at the end of the Wannierisation process. The atomic positions of the system are also written to the cube file. As the code is written at the moment, these are not shifted to the home unit cell. In general MLWF centres do not coincide with atomic centres, so it is not unusual that they differ by non-integer multiples of the lattice vectors. Hope that helps, Arash PS The subroutine that writes cube files is called internal_cube_format and can be found in plot.F90 (in the src directory of the W90 distribution). :------------------------------------------------------------: : Dr. Arash A. Mostofi :: a.mostofi at imperial.ac.uk : : Lecturer and RCUK Fellow :: : : Depts. of Materials & Physics :: : : Imperial College London :: T +44 (0)207 594 8154 : : London SW7 2AZ, United Kingdom :: F +44 (0)207 594 6757 : :------- http://www.cmth.ph.ic.ac.uk/people/a.mostofi -------: Jiri Houska wrote: > Dear all, > when printing cubefiles using wannier90 and > > wannier_plot=.true. > wannier_plot_supercell = 2 or 3 (did not matter) > wannier_plot_format = cube > > , after visualization (VMD) I realized that atomic positions and the > orbital positions are obviously shifted (of non-integer times cell > size). The shift appears to be different for different MLWFs. Please > is is avoidable and/or is there a known reason for that? > > Thanks! > Jiri Houska > University of West Bohemia > > _______________________________________________ > Wannier mailing list > Wannier at quantum-espresso.org > http://www.democritos.it/mailman/listinfo/wannier