The ABINIT Users Mailing List ( CLOSED )

["Allan, Douglas C Dr" <AllanDC@Corning.com>]

Also, think of any given eigenstate as having arbitrarily many angular 
momentum components (for reasons that Jim described).  The angular momentum 
decomposition is simply a representation of the arbitrary eigenstate into 
components that each transform according to the various spherical harmonics, 
each of which has a definite eigenvalue under rotations.  Jim's is the more 
physical answer, but also keep in mind that any arbitrary wave function can 
be written as a sum of terms that are eigenstates of rotation (i.e. spherical 
harmonics).

To test abinit and your understanding, you can try calculating a single atom 
in a large box and see whether the angular momentum decomposition looks right 
to you.  If the box is too small then periodic image atoms can perturb the 
results a little.

Of course, there is always the small chance that a bug may exist, so such 
tests are worthwhile for your confidence and for the continued improvement of 
the code ...

Doug Allan

-----Original Message-----
From: James Raynolds [mailto:JRaynolds@uamail.albany.edu] 
Sent: Monday, June 30, 2003 8:41 AM
To: 'forum@abinit.org'
Subject: RE: [abinit-forum] Angular momentum decomposition of KS states.


Fabrizio,

        Placing the atom in the solid lowers the symmetry and so you can get 
all these other components due to interactions with the other atoms.

Jim Raynolds

-----Original Message-----
From: Fabrizio Cleri [mailto:cleri@casaccia.enea.it]
Sent: Monday, June 30, 2003 7:03 AM
To: forum@abinit.org
Subject: [abinit-forum] Angular momentum decomposition of KS states.


Hi,
we are doing some angular decomposition of wavefunctions in molecules 
based on the ABINIT capabilities discussed below...
however, I have some basic doubt: what does it mean the d, f, g 
component when I start from atoms which have only s and/or p electrons? 
In some cases we found C atoms with the f component being the 90% of 
the wf for a particular k and band...

thanks for your patience,
F. Cleri



Inizio del messaggio inoltrato:

> Da: verstraete@pcpm.ucl.ac.be
> Data: Gio 30 Gen 2003  18:00:13 Europe/Rome
> A: forum@abinit.org
> Cc: balkagl@sancharnet.in
> Oggetto: [abinit-forum] angular momentum decomposition of KS states.
> Rispondere-A: verstraete@pcpm.ucl.ac.be
>
> In answer to the mail below, a brief explanation of what you can now
> do to
> analyse the s, p, d, character of a given state:
>
> If cut3d is called, and a _WFK file is entered instead of a _DEN file,
> you
> get a menu for the analysis of the different electronic states.
>
> If appropriate (ie. more than 1 choice) you can choose a kpoint, a
> band, a
> spin polarisation, and (4.0.3?) a spinorial component.
>
> cut3d extracts the appropriate wavefunction, and asks you if you want
> to
> do its angular analysis. Typing 1 (yes) here will do the analysis:
> for each atom, the wavefunction surrounding the atom will be projected
> into s, p, d, f, and g components. The output is the relative 
> percentage
> of spdfg, for each atom, and then globally for the state. The 
> projection
> is fast, but depends on your cutoff energy, so for 100s of bands it can
> take a few minutes or even a 1/4 an hour.
>
> Afterwards, type 12 to exit the next menu (other operations on the 
> wavefunction), and 1 if you want a new state to analyse. Otherwise 2 
> to exit. Simple, hmmm?
>
> This process is still fairly painful if you don't script it (simple
> enough
> since cut3d reads from stdout) to loop over kpts and bands. I will in
> future combine this to give you angular momentum decomposition of the 
> DOS,
> but as yet I don't see any relativistic effects on my time, so we'll 
> have
> to wait.
>
>
> Matthieu
>
> --
> ===================================================================
> Matthieu Verstraete               mailto:verstraete@pcpm.ucl.ac.be
> PCPM, Boltzmann, pl. Croix du Sud, 1            tel: 010/ 47 86 81
> B-1348 Louvain-la-Neuve  Belgium                fax: 010/ 47 34 52
>
> ---------- Forwarded message ----------
> Date: Thu, 30 Jan 2003 16:05:43 +0100
> From: Xavier Gonze <gonze@pcpm.ucl.ac.be>
> To: Bal K. Agrawal <balkagl@sancharnet.in>
> Cc: mverstra@pcpm.ucl.ac.be
> Subject: Re: Abinit
>
> Dear Bal K. Agrawal,
>
> Bal K. Agrawal wrote:
>> Dear Prof. Gonze
>>    In the Abinit code we want to know the contributions of atomic 
>> orbitals to the electronic state of the system. Is it possible ? If 
>> yes, please explain us.
>> with best regards.
>> Prof. B. K. Agrawal
>> Physics Department
>> Allahabad University
>> INDIA
>
> I relay your demand to the appropriate person (Matthieu Verstraete, 
> who developed this for v4.0 , so this is very recent). Still, it would 
> be better to send such mail to forum@abinit.org , where all the 
> interested persons (both knowning the solution or interested to hear 
> about) will be contacted.
>
> Good continuation,
> Xavier Gonze
>
>
>
>
>


______________________________________
Fabrizio Cleri
ENEA, UTS Materiali e Nuove Tecnologie
Centro Ricerche Casaccia, s.p. 050
00100 Roma A.D., Italy
tel.: +39-06-30484825
fax: +39-06-30484729
e-mail: cleri@casaccia.enea.it
e-page: http://babylon.casaccia.enea.it/TMM.html