Grottus mechanism

[huan... at mail.huji.ac.il]

Dear DJ MG,

The CPK representation only shows the "fixed bond". It is good for the 
snapshot published in journals, since it is only a picture. However, when 
displaying your simulation with proton transfer event, the OH bond will 
break in H3O^+ part and will form a new OH bond in the H2O part. It is a 
dynamic process. As a result, the "dynamicbond" is more suitable than CPK 
representation.

Let's go back to your input file.

As you said, the system you want to simulate is a dimer cluster. So, in the 
&POISSON section, you need to specify the parameter of PERIODIC as NONE, 
not XYZ. 
In addition, you need to use the Martyna-Tuckerman POSSION_SOLVER, as shown 
in  https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/DFT/POISSON/MT.html
This Poisson solver was proposed for the isolated molecules or clusters in 
the simulation box.

Keeping the default values for the parameters of ALPHA and REL_CUTOFF would 
be fine.

&POISSON
  PERIODIC NONE
  POISSON_SOLVER  MT
  &MT
     ALPHA 7.0
     REL_CUTOFF 2.0
  &END MT
&END POISSON


In the QS section,  it is better to set EXTRAPOLATION_ORDER as 3. 

In the &SUBSYS  &CELL section, you also need to specify the PERIODIC as 
NONE, which was not shown in your input file.
You can try to increase the dimension of your simulation box, perhaps 
increase to 15.0 angstrom. 

 &SUBSYS
   &CELL
       PERIODIC  NONE
       ABC  15.0  15.0  15.0
   &END CELL
...


I hope those suggestions would be helpful.

All the best,
Huan
.



On Thursday, July 14, 2016 at 1:41:44 PM UTC+3, Dj MG wrote:
>
> OK thank you Huan. I owe you one. I'm beginner in cp2k and english is not 
> my Mother tongue. So that only i can do is try exercises in cp2k website 
> and try and try more. OK i will try your suggestion. And your first 
> suggestion was succeed to change drawing method. My question is why in CPK 
> is failed to visualize the H+ transport?. The fact a lot of journal using 
> CPK model. 
>
> My system is small dimer cluster, any suggestion for this case?
>
> DJ MG
> Cheers Always
>
> On Wednesday, July 13, 2016 at 4:52:11 PM UTC+7, hua... at mail.huji.ac.il 
> wrote:
>>
>> Dear Dj MG,
>>
>> I checked your files. Actually, the proton has transferred from H3O^+ to 
>> H2O in your short trajectory.
>>
>> You may use the program called VMD to visualize the position file, which 
>> is your "H3O+_H2O-pos-1.xyz". 
>>
>> By adding the representation with the "Drawing Method" of "dynamicbonds", 
>> and setting the "Distance Cutoff" as 1.2, you will be able to observe the 
>> proton transfer event.
>>
>> The questions in your input file are:
>>
>> 1) Why you set the "MULTIPLICITY" as 3 in your input file? 
>> The proton does not have electron, and all electrons in H3O^+ H2O 
>> system are in-pair. In my opinion, the multiplicity for this system should 
>> be 1.
>>
>> 2) What kind of system would you like to simulate? Periodic system or 
>> just a small dimer cluster?
>>
>> 3) The temperature printed in your output file fluctuated quite large. 
>> The target temperature you set is 313 K, however, as shown in 
>> your "H3O+_H2O-1.ener" file, the maximum value even reached to 1339 K, 
>> whereas the minimum value dropped down to 99 K. That seems quite 
>> unreasonable. You need to choose a better thermostat, like Nose-Hoover 
>> chain thermostat or CSVR thermostat.
>>
>> All the best,
>> Huan
>> .
>>
>>
>> On Tuesday, July 12, 2016 at 6:19:40 PM UTC+3, Dj MG wrote:
>>>
>>> Hello cp2k lovers. 
>>> I have project which aims to simulate proton transfer through grottus 
>>> mechanism using aimd.
>>> I got a problem, why H atoms from H3O+ can't move to H2O?
>>> I attached my files below.
>>>
>>>
>>> Cheers Always
>>> ...
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20160714/7557a4b1/attachment.htm>