[Wannier] Symmetry-adapted Wannier functions

[Valerio Vitale]

Dear Cao,
Just to clarify, the Symmetry-adapted routine does not work with non-collinear calculations for sure, i.e. in the case of spinors. 
However, for spin-polarised calculations there is nothing in the code that prevents to use the symmetry-adapted method on the two spin channels separately.
The SA method has never been tested by us in this case. In a non-magnetic system we believe that in principle the code should produce two equivalent sets of WFs, one for spin up and one for spin down. But we are not sure whether the result is reliable in more general systems. 
We are planning to do some tests in the near future, after the next Wannier90 release.

Best,

Valerio

Postdoctoral research associate
Imperial College London
Exhibition road
London SW7 2AZ
and
TCM Group, Cavendish Laboratory
19 JJ Thomson Avenue,
Cambridge





> On 22 Feb 2019, at 08:29, Valerio Vitale <vv284 at cam.ac.uk> wrote:
> 
> Dear Cao,
> this is correct. At the moment symmetry-adapted WFs cannot be constructed when you have a spin-dependent Hamiltonian, e. g. in soc. As far as I know, no DFT code is able to generate spin-dependent symmetry adapted WFs, sorry.
> I believe that the reason is no one has developed a working procedure in this case yet. We do  have plan to work on this in the future but at the moment only spin-independent WFs can be used with the SA method.
> 
> An alternative method that could work with spin-dependent WFs is the selectively localised Wannier functions method with constraints on their centres, see Wang et al. Phys. Rev. B 90, 165125.
> They claim that by fixing the centres of the WFs, you automatically obtain functions that respect the point group symmetry of that site, without enforcing any symmetry constrain. This feature has been implemented recently in Wannier90 and can be found  in the latest develop version on the official GitHub repository, but we haven't fully tested on spin-dependent WFs. 
> 
> Regards, 
> 
> Valerio 
> 
> On 22 Feb 2019 4:28 am, 曹志鹏 <cz7989 at 163.com> wrote:
> Dear developers and users,
>     I would like to construct symmetry-adapted wannier functions by using wannier90_2.1.0. I only found that Quantum espresso 6.3 has the interface to wannier which can generate 'seedname.dmn'. But when I tried this on Bi2Se3 with soc, the module pw2wannier90 reported 'non-collinear not implemented...'. So the symmetry-adapted wannier functions cannot be constructed by QE when soc is opened, is it right?
>     If I correct, is there any other DFT code can meet my requirement?
>     Thank you very much!
> 
> Best Regards,
> Cao Zhipeng
> 
> 
>  
> 
> _______________________________________________
> Wannier mailing list
> Wannier at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/wannier

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/wannier/attachments/20190222/dcd65ca6/attachment-0001.html>