[CP2K-user] How to calculate the MD simulation containing the radical.

Du qq2855... at gmail.com
Mon Apr 27 09:40:27 UTC 2020


Dear Vladimir,

I'm so sorry that I made mistakes, what I meant are sulfate radical and 
hydroxyl radical. 

And I'll try your suggestions: UKS, MULTIPLICITY, BS. 

Thank you so much!

Yours,

Du

在 2020年4月20日星期一 UTC+8下午11:20:17,Vladimir Rybkin写道:
>
> Dear Du,
>
> it's a matter of trial and error and finding the right combination of 
> total charge and spin. To me it's not clear what OH- is. OH* is a radical, 
> hydronium, on the other hand, has a closed shell. SO4- is also a strange 
> beast. If you set total spin to 2 and charge to -2, you'll likely get OH* 
> and SO4(2-). There is an option to play with the guess:
>
>
> https://manual.cp2k.org/cp2k-6_1-branch/CP2K_INPUT/FORCE_EVAL/SUBSYS/KIND/BS.html
>
> Yours,
>
> Vladimir
>
> среда, 15 апреля 2020 г., 3:39:58 UTC+2 пользователь Du написал:
>>
>> Dear Vladimir,
>>
>> Thank you! 
>>
>> Here is the thing, I want to simulate the SO4- radical and the OH- 
>> radical, in a bulk of 40 water molecules. I don't know whether cp2k 
>> recognize them as radicals or just ions
>> ? As you mentioned, I've already set the UKS T in the input profile. 
>>
>> Yours,
>>
>> Du
>> 在 2020年4月14日星期二 UTC+8下午6:54:02,Vladimir Rybkin写道:
>>>
>>> Dear Du,
>>>
>>> if both radicals are doublet and the product is singlet you should set 
>>> total spin to 0 while keeping UKS true. This should work automatically.
>>>
>>> Yours,
>>>
>>> Vladimir
>>>
>>> понедельник, 13 апреля 2020 г., 14:25:56 UTC+2 пользователь Du написал:
>>>>
>>>> Hello All,
>>>>
>>>> I have some doubts about the MD simulation, when both the reactant are 
>>>> radical, is there any special parameter should be set in the input profile? 
>>>> or Can cp2k recognize the radical species automatically?
>>>>
>>>> Thanks!
>>>>
>>>> Du
>>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200427/90b2cefb/attachment.htm>


More information about the CP2K-user mailing list