ProjectionList and angularMomenta..

[Cyrille]

Dear all

I wanted to write a script to calculate the density of states that is not only decomposed in term of anglular momenta (l=0,1,2) but also in terms of magnetic number m or more precisely in real harmonics px,py,pz (same for d..).
Is there an easy way to extract to do that?

thanks in advance

Cyrille

[Anders Blom]

No, this is currently not possible in ATK.

[Cyrille]

OK thanks
Maybe in a future version of atk:-)

Cyrille

[Anders Blom]

It's most definitely on the to-do list.

[Cyrille]

By the way there is another interesting (but more involved) feature related to my previous post: the possibility of calculating the density of states projected on a given molecular orbital. This can be very useful for instance to analyse  electronic features of a molecule deposited on a surface.  One can calculate first the MO of the isolated molecule and then calculate the PDOS (on these molecular orbitals) when this molecule is in contact with a surface..

It has been implemented in QE and proved to be very useful..

thanks a lot for all your work.

Cyrille

[Anders Blom]

I assume you mean project on the eigenstates of the isolated molecule?

[Cyrille]

Yes exactly.
Then you can also see how the molecular levels are broadened by interaction with substrate.

Cyrille

[Anders Blom]

Actually that can partly be done already now, since ATK has so-called MPSH analysis. This amounts to re-diagonalizing the total device Hamiltonian in the restricted space of the molecular orbitals. See https://quantumwise.com/documents/tutorials/latest/MolecularDevice/index.html/chap.analysis.html#sect1.analysis.mpsh
Also, the DeviceDensityOfStates can be projected onto the molecule atoms, and again that should give you more or less what you need, no?

We do have plans to include more projection options for the DOS too, however.

[WUT]

Dear Dr. Bloom,

I would also be interested in such a possibility, to project onto the eigenstates of an isolated molecule calculated with MPSH analysis. As you mentioned, it is possible to project the DOS onto the molecule atoms, but this way it is the projection onto atomic orbitals, right? Since all the relevant information are accessible (DOS of the full system as well as the MOs of the isolated molecule) it should be possible to project onto MOs. Are there any news on this or maybe a hint for an external script?

Best,
   Markus

[Anders Blom]

In the 2017 version we will introduce projections onto subshells (like px, py, etc). I agree projecting onto MOs can be useful too and I suppose the new framework would make that comparatively easy, but I'm not sure it can be done in a script. I will add the request to our to-do list and follow up.