Dear Sunghwan and Dario,

I confirm Dario's guess. From a technical point of view, current implementation 
of turboTDDFT does not allow to use smearing. There are no conceptual problems 
to extend turboTDDFT such that to be able to treat systems with a smearing, but 
it has not been implemented yet.


Best regards,

Iurii



--

Dr. Iurii Timrov
Postdoctoral Researcher
Swiss Federal Institute of Technology Lausanne (EPFL)
Laboratory of Theory and Simulation of Materials (THEOS)
CH-1015 Lausanne, Switzerland

________________________________
From: [email protected] <[email protected]> on behalf of 
dario rocca <[email protected]>
Sent: Sunday, July 24, 2016 10:52 PM
To: PWSCF Forum
Subject: Re: [Pw_forum] Turbo TDDFT for a quantum dot

Dear Sunghwan,
That's my guess. If you set occupations='fixed' (the default) for the ground 
state calculation, the turboTDDFT will not complain that your system is 
metallic.
If your system has a 0.5 eV gap there will not be any problem. If the gap goes 
below 0.1 eV it migh be hard or impossible to converge the ground state 
calculation with  occupations='fixed'. In that case the turboTDDFT cannot be 
used.
According to me as long as the ground state energy converges with 
occupations='fixed' the turboTDDFT will run smoothly and give reasonable 
results for your nanoparticles.
Once again this is my guess, you should try.
Best,
Dario


On Sat, Jul 23, 2016 at 6:45 PM, SungHwan Choi 
<[email protected]<mailto:[email protected]>> wrote:
Dear Dario,
Thank you for a kind reply. You mean that if it is converged without smearing 
option, then there is no problem on doing turboTDDF calculation. Am I correct?
Sincerely,
Sunghwan Choi


========================================
Sunghwan Choi
Ph.D candidate
Computational Quantum Molecular Science Lab
Department of Chemistry, KAIST, Republic of Korea

2016-07-23 16:43 GMT+09:00 dario rocca 
<[email protected]<mailto:[email protected]>>:
Dear Sunghwan,
I have not used the code on systems with the characteristics you describe. This 
is my guess: As far as the ground state of your quantum dot converges without a 
smearing the turboTDDFT code can be used and will be stable. While the 
independent electron polarizability of a 0 gap system diverges for \omega-->0, 
the Hartree part in the TDDFT kernel will increase the value of the gap and/or 
will "push" the oscillator strength to higher energies.
Best,
Dario


On Fri, Jul 22, 2016 at 6:37 PM, SungHwan Choi 
<[email protected]<mailto:[email protected]>> wrote:
Dear all,
I am checking whether turboTDDFT program is suitable for my project. My project 
is to calculate absorption spectrum of metallic quantum dot. I have two 
questions.

1) From other post, I read the fact that turboTDDFT cannot calculate a metallic 
system. Here, how do we define metallic system? just system that have no band 
gap? then for small metallic quantum dot can be a non-metallic system since it 
has small band gap (0.1~0.5eV). case that fractional occupation does not 
happen? then, we can control the occupation parameter not to make fractional 
occupation.

2) A metallic quantum dot has a size-dependency for band gap. Thus, if size 
become larger, then band gap become narrow. When turboTDDFT failed to predict 
an absorption spectrum? and do you know a numerical trend of artifact, which 
come from low band gap?

By the way, does this limit come from algorithm or implementation of 
turboTDDFT? As far as I understood there is no limit of algorithm itself. I 
think there is some practical issues on there. Would you tell me where some 
troubles come up?

Sincerely
Sunghwan Choi


========================================
Sunghwan Choi
Ph.D candidate
Computational Quantum Molecular Science Lab
Department of Chemistry, KAIST, Republic of Korea


_______________________________________________
Pw_forum mailing list
[email protected]<mailto:[email protected]>
http://pwscf.org/mailman/listinfo/pw_forum


_______________________________________________
Pw_forum mailing list
[email protected]<mailto:[email protected]>
http://pwscf.org/mailman/listinfo/pw_forum


_______________________________________________
Pw_forum mailing list
[email protected]<mailto:[email protected]>
http://pwscf.org/mailman/listinfo/pw_forum

_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum

Reply via email to