Parameter choice for GAPW

[Ali Akbari]

Hi Marcella,

Many thanks for the quick reply. But I still have some questions. I do 
apologise in advance for asking too many questions, but you have a 
deep knowledge about the code & the implemented formalism and nobody can 
answer better than you.  

To give a better context for my questions I describe my system. It is a 
periodic system (TiO2 surface with an organic dye) and I want
to calculate XAS and XPS. Since (I think) all electron calculations are 
much heavier, I consider only the target atom with all electron 
and the rest with pseudo potentials. One general question is that how 
important is to change default values? Are the most important 
parameters the ones you wrote in the earlier post 6 years ago?
   
My details question are:

*ALPHA0_HARD* and EPS_3C_REDUCE
Compare to the previous discussion 6 years ago, It seems to me that 
 ALPHA0_SOFT is removed and ALPHA0_HARD default value is now 0 instead of 
10. What value do I need to choose? What should I do for EPS_3C_REDUCE?

*EPS_PPL *
EPS_PPL  default value is 0.01. Isn't it too rough?  

*EPS_DEFAULT*
It's written in the CP2K manual,  by setting EPS_DEFAULT, CP2K will all 
EPS_xxx variable to get an energy correct to up to EPS_DEFAULT. So how 
crucial is to set the variables EPS_******? Is there any EPS_**** that is 
not control by  EPS_DEFAULT?  


*LMAXN0 *and *LMAXN1*:
You wrote: The maximum possible value  is 2 times the maximum angular 
momentum present in the basis set.
Which basis set you refer to and how can I determine it (different atoms 
can have different basis sets)?
LMAXN1 default value is -1. What does it mean?


*QUADRATURE*:
There are three algorithms available. What is the criterion to choose?   


*HARD_EXP_RADIUS* and *EPSFIT*:
I have a C-H bond (length 1 A) in my system. The default values of 
HARD_EXP_RADIUS  are 0.6 A and  0.80 A for H and C respectively.
I treat C with all electron (aug-cc-pV5Z) and H with TZV2P-MOLOPT-GTH basis 
set. From your answer, I understand that I should choose smaller
values for  HARD_EXP_RADIUS.  What is your suggestion in this case? How 
should I choose an optimal value for situation like this?
Do I need to change  EPSFIT too?


*LEBEDEV_GRID* and *RADIAL_GRID*:
Both default values are 50 but you wrote before that 100 points on the 
radial grids, and 50 on the angular. Do they change if have a
mixture of all electron and pseudo potentials?



Thank you very much and cheers,
Ali



On Monday, 23 June 2014 10:21:15 UTC+3, Marcella Iannuzzi wrote:
>
> Hi Ali,
>
> Yes you can use pseudo potentials also with GAPW, and you can mix species 
> with pseudo potentials and species all electrons.
> Obviously, the PP species must have a basis sets optimised for PP 
> calculations.
> These most probably contain functions with relatively small exponents, 
> which means that most of them are going to contribute only to the soft 
> charge density.
> The partitioning in soft and hard terms is carried out in exactly the same 
> manner as for a full all electron calculation. 
> Since it is still a GAPW calculation, all GAPW parameters are still 
> needed. 
> Cheers
> Marcella
>  
>
> On Saturday, June 21, 2014 1:23:47 PM UTC+2, Ali Akbari wrote:
>>
>> Hi,
>>
>> I appreciate if someone tells me is it possible to use pseudo potential 
>> with GAPW
>> treatment? I was reading the following discussion to find the necessary 
>> parameters: 
>> https://groups.google.com/forum/#!msg/cp2k/RUFQScjSDn0/At_pJzz49_oJ
>>
>> What changes if I have a mixture of pseudo potential for some atoms and 
>> all
>> electron for others? e.g. Do I still need to choose the  HARD_EXP_RADIUS
>> such that there is no overlap between a pseudo atom and all electron atom?
>>
>> Many thanks and cheers,
>> Ali
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140623/38e2686f/attachment.htm>