[CP2K-user] [CP2K:11259]

[Patrick Gono]

Dear Akshay,

the input file seems reasonable to me, at least on first inspection. I've
looked at the trajectory as well. Proton transfer via the Grotthuss
mechanism can be observed, for example, around snapshot 3700 if you focus
on the water molecules with oxygen atoms No. 61, 70, and 79. Note that VMD
numbers atoms starting from 0 (unlike CP2K, which starts at 1), hence using
VMD's query tool you should look for oxygen atoms 60, 69, and 78. The water
molecules exchange hydrogen atoms in the snapshots leading up to here.
Using the CPK representation this becomes apparent through long bonds
between the oxygen of one water molecule and the hydrogen now belonging to
another water molecule. For a more comprehensive analysis, I suggest
writing a script that would, for example, track the distance of each
hydrogen atom from the closest oxygen atom in the initial structure. This
should make jumps in the distance apparent, and hence serve as a good
representation of proton transfer.

Since you are using GGA, I would expect the excess positive charge to be
spread out instead of being localized on a specific hydrogen atom. That's
why the proton transport happens all over the palce over the length of the
trajectory. This might or might not be important for the purpose of your
studies. In this paper:
https://pubs.acs.org/doi/abs/10.1021/acs.jpcb.6b03876 it is argued that the
hybrid PBE0(0.4) functional (40% exact exchange) and the van der Waals
rVV10 correction with a b parameter of 5.3 reproduce the structural and
electronic properties of liquid water quite well. Hybrid functionals are
quite efficient in CP2K thanks to the ADMM method, and the more accurate
rVV10 vdW correction has basically little to no overheard as far as my
experience goes. So this simulation should be feasible. However, I am not
really sure that such hybrid functionals would accurately model excess
charge in water. I trust you know more about this topic from your own
literature search.

Yours sincerely,
Patrick Gono

On Wed, 13 Feb 2019 at 05:31, Akshay Malik <akshayma... at gmail.com>
wrote:

> Dear Patrick,
> Since I have done AIMD of simple ions in water that is out of 64 water
> molecules replaced one water molecule and placed H+ ion in it so i am
> expecting to see grotthus mechanism. I am attaching the input files and
> trajectory files. Please give it a quick look and reply if I am doing
> something wrong.
>
> Yours sincerely
> Akshay Malik
>  WATER-pos-1.xyz
> <https://drive.google.com/file/d/1r2HPG-MvTK6UvZFjgXnM-IkEfZhSIeiy/view?usp=drive_web>
>  water.inp
> <https://drive.google.com/file/d/1SUH4LUEqy1bAZ0AiOrx0gATttp7K4szy/view?usp=drive_web>
>  wateralter.xyz
> <https://drive.google.com/file/d/1mKGUwOqDbGH49Un4ciK2Kc1nPTdMRMxe/view?usp=drive_web>
>
> On Mon, Feb 11, 2019 at 6:34 PM Patrick Gono <patric... at gmail.com>
> wrote:
>
>> Dear Akshay,
>>
>> A visualization software is just that -- a tool that allows you to
>> visually inspect the simulated trajectory. If no proton transfer happened,
>> there is nothing to visualize. Assuming that there is some proton transfer
>> happening, the CPK representation should be actually the best. Here, the
>> bonds are displayed between close-by atoms in the initial configuration,
>> and then remain during the simulation. Hence, if a proton did move from one
>> water molecule to another, you should see ridiculously stretched O-H bonds
>> in your MD. The Dynamic Bonds representation only uses a very simple cutoff
>> distance bellow which a bond is displayed. So, naturally, there will be
>> some bond breaking and formation of new bonds during the course of the
>> simulation, but it may be hard to spot this using this representation.
>> Plus, if you use a large cutoff, hydrogen bonds to neighboring water
>> molecules might be shown as real bonds occasionally, which is not what you
>> probably are looking for.
>>
>> If you have dozens and dozens of picoseconds worth of trajectory to comb
>> through, and you do not trust the CPK representation for some reason, you
>> can also write a short script to analyze the trajectory yourself. For
>> example, you could calculate the two shortest bonds for each oxygen atom in
>> the first snapshot, and then track the length of these bonds over time. At
>> some point, if a proton transfer occurs, the length of one of these bonds
>> will exceed a given threshold (say, around 1.3 A). You then simply output
>> the snapshot index and go look it up in VMD to see what is happening. But
>> this is basically what the CPK representation already visualizes.
>>
>> Are you sure that proton transfer can be observed in your simulation? Are
>> you simulating pure water, or do you also include an extra proton (or
>> H3O+)? What parameters did you use for the simulation? How long is your
>> trajectory in physical terms -- do you expect to observe proton transfer on
>> this timescale? Could you share your CP2K input, initial water structure
>> file, and possibly (if size allows) the simulated trajectory, please? You
>> can upload the trajectory to dropbox or google drive and send me the link
>> if it is too large.
>>
>> Yours sincerely,
>> Patrick Gono
>>
>> On Fri, 8 Feb 2019 at 13:46, Akshay Malik <akshayma... at gmail.com>
>> wrote:
>>
>>> Dear Patrick,
>>>
>>> I have used VMD to analyze the trajectory file but I am not able to see
>>> the quantum effects such as proton transfer in CPK Representation but in
>>> DynamicBonds representation there is some sort of bond making and breaking.
>>> Can I consider DynamicBonds representation as quantum effects?
>>>
>>> On Fri, Feb 8, 2019 at 4:37 PM Patrick Gono <patric... at gmail.com>
>>> wrote:
>>>
>>>> Dear Akshay,
>>>>
>>>> I suggest VMD, which can be found at
>>>> https://www.ks.uiuc.edu/Research/vmd/
>>>>
>>>> Yours sincerely
>>>> Patrick Gono
>>>>
>>>> On Fri, 8 Feb 2019 at 05:35, Akshay Malik <akshayma... at gmail.com>
>>>> wrote:
>>>>
>>>>> I have done AIMD of liquid water by using cp2k. I want to see the bond
>>>>> breaking and proton transfer by grotthuss mechanism. Please suggest some
>>>>> visualisation softwares.
>>>>>
>>>>> --
>>>>> 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 post to this group, send email to cp... at googlegroups.com.
>>>>> Visit this group at https://groups.google.com/group/cp2k.
>>>>> For more options, visit https://groups.google.com/d/optout.
>>>>>
>>>> --
>>>> 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 post to this group, send email to cp... at googlegroups.com.
>>>> Visit this group at https://groups.google.com/group/cp2k.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>> --
>>> 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 post to this group, send email to cp... at googlegroups.com.
>>> Visit this group at https://groups.google.com/group/cp2k.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>> --
>> 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 post to this group, send email to cp... at googlegroups.com.
>> Visit this group at https://groups.google.com/group/cp2k.
>> For more options, visit https://groups.google.com/d/optout.
>>
> --
> 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 post to this group, send email to cp... at googlegroups.com.
> Visit this group at https://groups.google.com/group/cp2k.
> For more options, visit https://groups.google.com/d/optout.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20190213/7ed49ae0/attachment.htm>