Welcome To Support Community

Materials Studio

Advanced Search
Ask Search:
K_KK_K 

COMPASS force field deal with diaspore crystal

Dear all, I met some problems about COMPASS force field in dealing with diaspore crystal shown as follows. Any help is much appreciated.Thank you advance.
User-added image
1. The Geometry Optimization task(Forcite) with COMPASS force field has been done on diaspore crystal for which first GO by CASTEP was successful. The output txt file shown that "Energy contributors with missing parameters" and "Energy contributors using automatic parameters" as Picture 1.
The error in COMPASS output file
Refering to the Step 2 of "Polymer interactions with a metal oxide surface", I manually applied forcefield types and then deleted all the bonds in the structure and the error was fixed. So how it can be interpreted?The first way is wrong?
Howerver, I found that the sum charge of the structure is not 0. Forcefield assigned of charge was adopted. How this happen?
User-added image


2、The (010) surface of diaspore with 19.017 angstorm thickness was created  and bulk atoms Cartesian position were constrainted. Same like the pre-operations, the Geometry Optimization was done with Ultra-fine quality.Unexpected, the output file shown "WARNING Convergence criteria are not satisfied",even in the Corse precision.
The structure of dia(010) which has been GO is very odd as follows which was magnified significantly:
User-added image

I wonder if the surface thickness  & non-bond cut off distance probelm. How can I choose an appropriate thickness?
I found "As the non-bond cutoff distances in the forcefield settings are 9.5 Å, the thickness of your surface must be more than 9.5 Å. For the 0 0 -1 plane it is approximately 13 Å." in help of MS for "Polymer interactions with a metal oxide surface" and  "The number of layers in the structure should be chosen so that the depth of the surface is greater than the non-bond cutoff used in the calculation" in "Determining the location of SO2 on the Ni(111) surface with Adsorption Locator". However, there is one problem about the latter in which Ni(111) surface thickness is 12.207 angstorm, but non-bond cut off distance with Fine is 15.5 angstorm. Isn't that a contradiction?


3、What is the meaning of following description appeared in “If geometry optimization fails”content for Forcite?
User-added image

That's all. Thank you a lot!
Best Answer chosen by K_K
reinierreinier
Hi K_K,

I think the main problem is that COMPASS does not have parameters for oxygen in this environment. It ends up assigning the default type “o”, leading to a charge mismatch amongst others.

You could try the CLAYFF forcefield, which is available here. Like COMPASS, it has typing rules that require bonds to be present. But you must delete the bonds before the calculation, as otherwise the exclusion rules prevent ionic interaction (as well as giving you a warning about missing parameters).

Regards,
Reinier
 

All Answers

reinierreinier
Hi K_K,

I think the main problem is that COMPASS does not have parameters for oxygen in this environment. It ends up assigning the default type “o”, leading to a charge mismatch amongst others.

You could try the CLAYFF forcefield, which is available here. Like COMPASS, it has typing rules that require bonds to be present. But you must delete the bonds before the calculation, as otherwise the exclusion rules prevent ionic interaction (as well as giving you a warning about missing parameters).

Regards,
Reinier
 
This was selected as the best answer
K_KK_K
The problem has been solved by Charging use QEq method.Firstly, I manually assign Forcefield type of each atom in COMPASS. Then,Charging use QEq method with initial charge of zero was adopted to assign charges. Finally, delete all the Al-O bonds and donnot check calculate automaticlly button and Use Current charge method.
K_KK_K
reinier
Thank you a lot. I will try CLAYFF forcefield.