This thread is locked.Only browsing is available.
Top Page > Browsing
How to construct peudopotential of fictitious atom
Date: 2006/11/29 10:31
Name: Xinyuan Zhang   <>

The research I am doing requires to use a fictitious hydrogen atom with 1.5 nuclear charge and 1.5 electrons. I checked the adpack1.8 manual, but I am not sure I can do that. Is there a way for me to construct the pseudopotential so I can run the simulation in OpenMX?

Page: [1]

Re: How to construct peudopotential of fictitious atom ( No.1 )
Date: 2006/11/29 14:55
Name: T.Ozaki


Thank you for your suggestion.

The current version of OpenMX and ADPACK is not capable of treating
the fictious atoms with a fractional nuclear charge.

A couple of modifications are required to do that, while it is not
difficult in principle.

In the future release, the functionality will be considered.


Re: How to construct peudopotential of fictitious ( No.2 )
Date: 2006/12/01 04:54
Name: Xinyuan Zhang  <>

Hi Dr.Ozaki:
Thank you for your reply. I am also thinking whether it is possible to let the user to specify the occupation numbers in the next release. I think in that case the user can simulate the exciton problem. For example, the SiH4 has the occupations like 2 2 2 2 0 0.., but if we can specify the occupation numbers, we can use 2 2 2 1 1 0 0 simulate an exciton.
Thank you very much.

Re: How to construct peudopotential of fictitious atom ( No.3 )
Date: 2006/12/01 23:49
Name: Vasilii Artyukhov

I think that this feature (more precisely, first excited state) is available in SIESTA, but it's crude and not very useful within usual DFT. For excited states, you should probably try using Time-Dependent DFT. It is implemented in Octopus, ABINIT, CPMD and some other codes, but I doubt it can be easily implemented in usual order-N DFT schemes.

Page: [1]