[CPMD-list] what is the difference between Goedecker Pseudopotential and Troullier-Marins Pseudopotential?

[Axel Kohlmeyer]

On Wed, 28 Jun 2006, Chen Ying wrote:

CY> Dear all,
CY> 
CY> I know the both are norm-conserving, and Goedecker Pseudopotential (HGH) is
CY> relativistic, separable, dual-space Gaussian-type. Perhaps it is hard to say which
CY> one is better, but I don't know which one is more appropriate for a particular

there is no generic 'better', it all depends on the application and the
'quality' of the individual pseudopotential (which can vary a _lot_, so
the upbiquitous warnings about always testing your pseudopotentials are 
not a joke!).

CY> element or system. Is HGH more effective for metal, and TM for nonmetal? Then when

no, on average the goedecker type pseudopotentials are 'harder' than 
troullier-martins pseudopotentials. well, one can create them much
'harder' but a 'softer' pseudopotential needs a smaller basis set, 
i.e. a lower plane wave cutoff to represent the 'chemistry' of a
given atom correctly, so one always tries to find a good compromise
between softness and accuracy and then the goedecker type psp tend
to be harder. the fact that the goedecker pseudopotentials are more
frequently used for (transition) metals has its footing in the problems
with creating troullier-martins pseudopotentials without ghost states
(when using the kleinman-bylander scheme to handle the non-locality).

CY> I do simulations for a complicated system, can I use HGH for metal elements while

yes. you can mix and match pseudopotentials (as long as the dft 
functional is consistent) any which way you like. note, that the
'hardest' pseudopotential determines the required plane wave cutoff,
so there are certain combinations that make less sense.

CY> TM for nonmetal ones? Perhaps I should read more papers about these
CY> pseudopotentials instead of just asking for help. Thank you very much in advance.

thank you for answering the vital question yourself.

regards,
     axel.
CY> 
CY> 
CY> 
CY> 
CY> ------------------------------
CY> Best regards
CY>   achen
CY> _______________________________________________
CY> CPMD-list mailing list
CY> CPMD-list at cpmd.org
CY> http://cpmd.org/mailman/listinfo/cpmd-list
CY> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
   Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.