Isolated system Geometry Optimizations

[Babgen Manookian]

Hey Everyone,

I am a new cp2k user and have been working on a test system in order to get a feel for how the program works. My test system is a cyclohexane molecule and I am interested in determining the energy difference between boat and chair conformations. I am using a wavelet poisson solver with periodicity turned off and the molecule centered in the cell. I attached the template input file which I used to create the different inputs with varying cells size.

What I have noticed is that as I increase my cell size the energy does not converge on a single value, instead it seems to oscillate. The graphs below show the energies of each conformation and their differences as functions of cell size. I am very curious as to what is causing these oscillations. If I have a single cyclohexane molecule centered in my cell, as I increase my cell size, I would think that the extra vacuum space around the molecule will not affect the energy. 

If there is anyone out there who could provide any sort of input on this it would be much appreciated.

Thanks,

Bobby

[hut...@chem.uzh.ch]

Hi

the problem is connected to the fact that the plane wave basis
depends on the box size. The same is true for the XC integration grid.
Now with a very small cutoff for the PW this will show the dramatic effects
you have.
To solve this proble you have to use a much higher cutoff, especially the
relative cutoff. I would suggest values of 80 and 400, respectively.
The smoothing you use for the XC will also have an effect. I don't know
how bad this is in your case. I would start without the smoothing and
make later some tests to see if it helps.

In general, it is a good idea for such tests to increase accuracy in all
parts of the calculation.

regards

Juerg

--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----To: cp2k <cp...@googlegroups.com>
From: Babgen Manookian
Sent by: cp...@googlegroups.com
Date: 02/14/2017 07:21PM
Subject: [CP2K:8695] Isolated system Geometry Optimizations

--

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...@googlegroups.com.

To post to this group, send email to cp...@googlegroups.com.

Visit this group at https://groups.google.com/group/cp2k.

For more options, visit https://groups.google.com/d/optout.



[attachment "cychex_boat_temp.inp" removed by Jürg Hutter/at/UZH]

[Babgen Manookian]

Hey Juerg,

Thank you so much for the reply. I re-converged my cutoff values for the larger cell size and found that 400Ry and 80Ry produce energies which have much less deviation. What I do notice is that there is still the oscillation of the energy as the cell size changes. Is there a reason for this oscillation occurring?

My thoughts on this are that the planewaves are being cut at the cell wall at different values as the cell size changes which in turn causes the energy to oscillate. I would imagine if I could force the calculation to set the density equal to zero at the cell wall then there would be no oscillation. I am not sure if this reasoning is correct.

Bobby 

[hut...@chem.uzh.ch]

Hi

yes, the changes (jumps) in energy when changing the cell size is a
well known phenomenon in plane wave calculations. You can avoid it
by using a constant plane wave instead as constant cutoff calculation.
In CP2K you get this by defining a reference cell (REF_CELL).

regards

Juerg
--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----To: cp2k <cp...@googlegroups.com>
From: Babgen Manookian
Sent by: cp...@googlegroups.com

Date: 02/17/2017 10:18PM
Subject: [CP2K:8716] Re: Isolated system Geometry Optimizations

[Babgen Manookian]

Hey Juerg,

So I looked in the input manual and could not find the REF_CELL keyword. Also, it is still a little unclear what is actually being changed in terms of the calculation. Could you elaborate on that for me please?

Does this have anything to do with setting density to be zero at the cell wall?

Thank you in advance,

Bobby 

On Tuesday, February 14, 2017 at 1:21:43 PM UTC-5, Babgen Manookian wrote:

[hut...@chem.uzh.ch]

Hi

sorry, the correct keyword is

CP2K_INPUT / FORCE_EVAL / SUBSYS / CELL / CELL_REF

this allows you to do constant number of grid point calculations
instead of constant cutoff. It results in smoother energy vs. volume
curves.

regards

Juerg
--------------------------------------------------------------
Juerg Hutter                         Phone : ++41 44 635 4491
Institut für Chemie C                FAX   : ++41 44 635 6838
Universität Zürich                   E-mail: hut...@chem.uzh.ch
Winterthurerstrasse 190
CH-8057 Zürich, Switzerland
---------------------------------------------------------------

-----cp...@googlegroups.com wrote: -----To: cp2k <cp...@googlegroups.com>
From: Babgen Manookian
Sent by: cp...@googlegroups.com

Date: 02/25/2017 09:39PM
Subject: [CP2K:8745] Re: Isolated system Geometry Optimizations