Fat Band Analyzer

[Dipankar Saha]

Hi,

What do you mean by saying_
'Min. value =  0 / Max. value =  1' for any projected band?

Best_
Dipankar

[Petr Khomyakov]

In simple words, changing these parameters is supposed to change the thickness of the fat bands. However, it seems to be working with Style -> 'Scaled points' only, e.g., keep max value at 1.0, and tune min from 0 to 0.9 to see how the fat band changes. To conclude, I think you have found a bug in the Fat Band Analyzer.

[Dipankar Saha]

Thanks a lot Petr !

I tried to vary Min. values (from 0 to 0.9), keeping Max. at 1, but did not see any change. Considering the thickness of the bands, it seems that the default projections can hardly be manipulated.

Howbeit, I just want to know:

1) Whether the overall band alignment is correct or not?

2) Are the individual atom wise projected bands properly aligned?

Best_
Dipankar Saha

[Petr Khomyakov]

1) Whether the overall band alignment is correct or not?

What do you mean by "band alignment" in the case of a bulk crystal? The band structure is correctly given, as well as the projections visualized with fat bands. It is just not possible to change the visual representation of the fat bands.   

2) Are the individual atom wise projected bands properly aligned?

I did not get the question. As I said the projections are correctly computed.

[Dipankar Saha]

The band structure is correctly given, as well as the projections visualized with fat bands.

Okay... That's fine!

As I said the projections are correctly computed.

Thank you for the information.

Best_
Dipankar

[Petr Khomyakov]

Just one thing, you mentioned that you had seen no changes while changing the min value. Did you do it with 'Scaled points' style or 'Lines'? I was able to tune the band thickness after switching to 'Scaled points', at least for the test calculation I did. It would be helpful to debug these thing, if you let us know what version of QuantumATK you are using and on which operating system running it. If possible, please post your python script.   

[NW]

Hello,
I have a question related to this topic. When we want to calculate the Fatband structure for an interface e.g. Graphene on SiC, how we can calculate the bandstructure just for Graphene (because we have carbon atoms both in SiC and Graphene)? as far as I know it is possible to define tags for graphene and then calculate the Fatband for tags. However, my question is how we can calculate the Fatband in this kind of structure for shells or orbitals? For example, how can I find changes in Pz of Graphene after building SiC/Graphene?

[Daniele Stradi]

Hi,

the FatBandstructure analysis object also allows for projections on tags.

Best regards,
Daniele

[NW]

Good news .
So how can I do that?

[Dipankar Saha]

Did you do it with 'Scaled points' style or 'Lines'? I was able to tune the band thickness after switching to 'Scaled points', at least for the test calculation I did.

Yeah... it is better to opt for 'scaled points' and "ProjectOnOrbitalsByElement" (instead of "ProjectOnElements").

It would be helpful to debug these thing, if you let us know what version of QuantumATK you are using and on which operating system running it.

Version 2018.06-SP1-1
Linux

If possible, please post your python script.

I will send you the script.


Best_
Dipankar Saha

[Dipankar Saha]

Good news .
So how can I do that?

If you made an interface, you must be having tags for the left and the right interfacing atoms.
E.g. :

bulk_configuration.addTags('Left Interface',  [.... no. of atoms....])
bulk_configuration.addTags('Right Interface', [.... no. of atoms....]) .

Then you need to modify  the "Fat Bandstructure" analysis details as_
projections=ProjectOnTags,

Thus the projection will be for_  'Left Interface' / 'Right Interface'  .
 

[NW]

I mean that I want to obtain Fatband for shells or orbitals of tags.

[Petr Khomyakov]

You have to create a tag, e.g., named 'Graphene', by selecting the carbon atoms comprising a supported  graphene monolayer, and then set

projections=[Projection(atoms=['Graphene'], l_quantum_numbers=[1], m_quantum_numbers=[0])]

in the Fat Band Structure analysis object in the python script. That would project DOS on graphene + pz orbital.

[NW]

Thank you Petr.

[Petr Khomyakov]

You may find more about setting different kinds of projections at https://docs.quantumwise.com/manual/Types/Projection/Projection.html. Note that there exists a typo in the usage examples, 'l_quantum_number' should be 'l_quantum_numbers'.

In fact, you may use projection generator to generate a projection needed, see https://docs.quantumwise.com/manual/Types/ProjectionGenerator/ProjectionGenerator.html.