[CP2K-user] [CP2K:16703] Re: Point charges to simulate an electric field?

[Бранислав Миловановић]

Hi,
Ok, thank once again for clarifying.
Mulliken population analysis gives expected values for these fake atoms.

Best
BM

petak, 11. mart 2022. u 13:13:43 UTC+1 mattwa... at gmail.com je napisao/la:

> Hi, 
> those you list are Hirshfeld charges that are defined geometrically 
> without reference to a basis set (intergrated density in some radius or 
> similar). If you look at Mulliken charges those should be as you expect.
> At a glance the setup looks right to me.
> Matt
>
>
> On Sunday, 6 March 2022 at 00:28:30 UTC dr.bra... at gmail.com wrote:
>
>> Thanks.
>> OK, I think I get it now. 
>> So I have set CHARGE keyword to 25, since I want to have 25 fake H atoms 
>> and each to represent smeared charge of +0.28e, and the rest of my system 
>> in neutral.
>> My KIND section now:
>>
>>
>>
>>
>>
>>
>> *    &KIND X      ELEMENT H      BASIS_SET NONE      POTENTIAL 
>> GTH-PBE-q1      CORE_CORRECTION -0.88    &END KIND*
>>
>> However, when I perform population analysis I find finite population on 
>> these fake atoms:
>>
>>                            
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> *Hirshfeld Charges  #Atom  Element  Kind  Ref Charge     Population       
>>              Net charge      1       H      1       0.280          0.046   
>>                       0.234      2       H      1       0.280         
>>  0.047                         0.233      3       H      1       0.280     
>>      0.052                         0.228      4       H      1       0.280 
>>          0.055                         0.225      5       H      1       
>> 0.280          0.046                         0.234      6       H      1   
>>     0.280          0.051                         0.229.*
>> *.*
>> *.*
>>
>> How is this possible when they are described with no basis at all?
>>
>> Best,
>> BM
>>
>> On Wednesday, March 2, 2022 at 5:04:08 PM UTC+1 mattwa... at gmail.com 
>> wrote:
>>
>>> There will be the core charge of the original atom as well - so you need 
>>> neutralise the atom core (-1 in your case) then add your charge back.
>>> Yes - the charge keyword will alter the number of electrons in the 
>>> system so if you want these to be virtual protons not hydrogen atoms then 
>>> you do need to remove the electron as well with the charge keyword.
>>> If your overall system is charged then normal issues of charged periodic 
>>> systems apply.. 
>>> Matt
>>>
>>> On Monday, 28 February 2022 at 12:37:58 UTC dr.bra... at gmail.com wrote:
>>>
>>>> Additionally, should I in such cases change total charge od the system?
>>>>
>>>> On Wednesday, February 23, 2022 at 2:09:59 PM UTC+1 Бранислав 
>>>> Миловановић wrote:
>>>>
>>>>> Thanks, just want to check if this is right way to do this:
>>>>> I defined atoms labeled *X* in the coord file and then I added this 
>>>>> section into the SUBSYS:
>>>>>
>>>>>     &KIND X
>>>>>       ELEMENT H
>>>>>       GHOST .TRUE.
>>>>>       BASIS_SET NONE
>>>>>       POTENTIAL GTH-PBE-q1
>>>>>       CORE_CORRECTION 0.12   # 3/25
>>>>>     &END KIND
>>>>>
>>>>> Is this the right way to divide for example 3*e* charge over 25 fake 
>>>>> hydrogen atoms?
>>>>>
>>>>> Branislav
>>>>>
>>>>> On Monday, February 21, 2022 at 4:31:49 PM UTC+1 mattwa... at gmail.com 
>>>>> wrote:
>>>>>
>>>>>> Multiple options - most directly:
>>>>>>
>>>>>> https://manual.cp2k.org/cp2k-9_1-branch/CP2K_INPUT/FORCE_EVAL/SUBSYS/KIND.html#list_CORE_CORRECTION
>>>>>> lets you change the core charge on a pseudopotential. If you make a 
>>>>>> fake hydrogen and adjust charge appropriately you'd get close to what you 
>>>>>> were after (won't be a point charge but gaussian smeared charge).
>>>>>>
>>>>>> There are other routines that could load an external potential on a 
>>>>>> grid or make use of the QM/MM code in some form.
>>>>>> Matt
>>>>>>
>>>>>>
>>>>>> On Monday, 21 February 2022 at 15:24:34 UTC dr.bra... at gmail.com 
>>>>>> wrote:
>>>>>>
>>>>>>> Dear cp2k users,
>>>>>>>
>>>>>>> Is it possible to introduce ghost atoms with no basis set and with 
>>>>>>> finite non-integer point charge in AIMD calculations? Something like MM 
>>>>>>> point charges.
>>>>>>>
>>>>>>> I want to try to simulate behaviour of some small molecules in the 
>>>>>>> electric field which originates from a particular sequence of the point 
>>>>>>> charges.
>>>>>>>
>>>>>>> Thanks
>>>>>>> Branislav
>>>>>>>
>>>>>>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/afc642ee-74fb-4582-808a-00a887223fc7n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20220312/a4f6bd21/attachment.htm>