[CP2K-user] [CP2K:16905] How to find the global minimum...?

[Sam Broderick]

Hi

In researching global optimization, I stumbled on a reply be Ole Schütt, 
who is the author of the SWARM part of CP2K. He suggested that the methods 
are somewhat older and to look at ASE <https://wiki.fysik.dtu.dk/ase/>. I 
think this is what I will look into.

Cheers
Sam

On Friday, January 22, 2021 at 4:59:34 PM UTC+1 Monu Joy wrote:

> Thanks, Fangyong for the suggestion.
>
> On Fri, Jan 15, 2021 at 10:01 PM Fangyong Yan <fyya... at gmail.com> wrote:
>
>> Hi, Monu,
>>
>> The difference energy is -6235.818 - (-6235.824) = 0.006 Hartree = 
>> 15.753 kj/mol. 
>>
>> So if you use 800 K, the Boltzmann factor is: exp(-dE/R/T) = 
>> exp(-15.753*1000/8.314/800) = 0.09, which means you would have a 
>> probability of 0.09 to "jump" from the lower energy to higher energy, by 
>> the help of heat. 
>>
>> So I think NVT at 800K, may be helpful for exploring the potential energy 
>> surface, even within 10ps. 
>>
>> Regards,
>>
>> Fangyong
>>
>>
>> On Fri, Jan 15, 2021 at 8:08 PM Monu Joy <monuj... at gmail.com> wrote:
>>
>>> Hi Lucas & Fangyong
>>>
>>> Thanks for the help. I have performed two cell-opt calculations of 
>>> guest at host where the following total energies respectively represent the 
>>> guest at the center of the MOF pore and near to the pore surface -6235.818 
>>> and -6235.824. Do you guys think that this energy difference is 
>>> reasonable...?
>>>
>>> Also, do you by chance have any reference that shows local minimum 
>>> geometry modulate the property of a system more effectively...? 
>>>
>>> Thank you so much for your help!
>>>
>>> -Monu
>>>
>>>
>>> On Wednesday, 13 January, 2021, 10:01:30 am GMT-5, Lucas Lodeiro <
>>> eluni... at gmail.com> wrote: 
>>>
>>>
>>> Hi Monu and Fangyong,
>>>
>>> I disagree a little bit with Fangyong because an AIMD calculation is not 
>>> cheaper, and if the barriers between local minimas are high, you cannot see 
>>> a structural change with the MD, at least in the scale time of ps. Firstly, 
>>> I would try to optimize a couple of structures and measure the energy 
>>> differences, this will give you a little insight about the barriers (the 
>>> minimum energy for the barriers), If they are high, the AIMD probably 
>>> cannot move from one local minima to another. If they are little, you can 
>>> see the structural changes, but in this case, probably the global minima is 
>>> not sufficient to modulate the properties of the system, and some other 
>>> local minima have to be taken into account.
>>> Other way is to generate with a script random positions of the molecule 
>>> into the MOF and minimize them with a semiempirical formulation, as XTB (
>>> https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/QS.html#METHOD), 
>>> and rerun with the level of theory that you want just the low energy ones.
>>>
>>> Regards - Lucas Lodeiro
>>>
>>>
>>> El mié, 13 ene 2021 a las 0:07, Fangyong Yan (<fyya... at gmail.com>) 
>>> escribió:
>>>
>>> Hi, Monu,
>>>
>>> To explore the free energy surface better, you may need to start at a 
>>> high temperature, and then annealing down to the room temperature. And for 
>>> each temperature, you can plot the radial distribution function, for your 
>>> guest molecule and your MOF atoms, for example, at 800 K, 700 K, ...., 
>>> 300K. The reason starting at high temperature is that you can use heat to 
>>> get your guest molecule out of the potentially trapped free energy minima, 
>>> so you can explore the free energy minima more completely. 
>>>
>>> Regards,
>>>
>>> Fangyong
>>>
>>> On Tue, Jan 12, 2021 at 10:01 PM Fangyong Yan <fyya... at gmail.com> wrote:
>>>
>>> Hi, Monu,
>>>
>>> You can run a molecular dynamics simulation at room temperature, for 
>>> several picosecond, and then you can plot the radial distribution function, 
>>> between your guest molecule and the MOF, by doing this you would have a 
>>> general picture of the free energy surface of your guest molecule in the 
>>> MOF. Then based on the free energy minima structures, you can start to 
>>> optimize your guest-MOF complex structure. (the free energy surface may be 
>>> different from the potential energy surface, but at least by doing MD 
>>> simulation and calculating the free energy surface based on radial 
>>> distribution function, you can have a basic idea about the free energy 
>>> surface, which should not be quite different from potential energy surface).
>>>
>>> Regards,
>>>
>>> Fangyong
>>>
>>> On Tue, Jan 12, 2021 at 7:20 PM Monu Joy <monuj... at gmail.com> wrote:
>>>
>>> Hi there
>>>
>>> I have a general question: I would like to find out the global minimum 
>>> of a guest molecule within the pore of a MOF. I know I can place the guest 
>>> molecule in the pore and get the host-guest structure optimized, but if I 
>>> would place the guest molecule in another position of the same pore, there 
>>> also I can get the structure optimized. In this scenario how could I find 
>>> the exact most probable location of the guest molecule within the MOF pore 
>>> or its global minimum...?
>>>
>>> Any help would be highly appreciated...
>>>
>>> -Monu
>>>
>>> -- 
>>> 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+uns... at googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/1852672747.3320.1610497192914%40mail.yahoo.com 
>>> <https://groups.google.com/d/msgid/cp2k/1852672747.3320.1610497192914%40mail.yahoo.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>> -- 
>>> 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+uns... at googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/CAJjmc%2BKviWAFmj5UmYqMWL1Z5wFpQZjdKkhGS%2B7CTp7-6jaLgg%40mail.gmail.com 
>>> <https://groups.google.com/d/msgid/cp2k/CAJjmc%2BKviWAFmj5UmYqMWL1Z5wFpQZjdKkhGS%2B7CTp7-6jaLgg%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>> -- 
>>> 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+uns... at googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/CAOFT4PKCKjc0vOGZUkbjsDEEqJM_w_UxYK-PC%2Bt-vnzZzLhrgg%40mail.gmail.com 
>>> <https://groups.google.com/d/msgid/cp2k/CAOFT4PKCKjc0vOGZUkbjsDEEqJM_w_UxYK-PC%2Bt-vnzZzLhrgg%40mail.gmail.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>>> -- 
>>> 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+uns... at googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/cp2k/1361749041.8962.1610759305502%40mail.yahoo.com 
>>> <https://groups.google.com/d/msgid/cp2k/1361749041.8962.1610759305502%40mail.yahoo.com?utm_medium=email&utm_source=footer>
>>> .
>>>
>> -- 
>> 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+uns... at googlegroups.com.
>>
> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/cp2k/CAJjmc%2BKyP-zV-p%2B%2BZtUkTgqoWQnovqnUvaVpG7NQby7a8KURgQ%40mail.gmail.com 
>> <https://groups.google.com/d/msgid/cp2k/CAJjmc%2BKyP-zV-p%2B%2BZtUkTgqoWQnovqnUvaVpG7NQby7a8KURgQ%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .
>>
>
>
> -- 
> *Thanks with Regards*
>

-- 
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/b299be0c-ecdb-47f8-871b-cdbb52def373n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20220429/1440f74c/attachment-0001.htm>