Hi everyone,
i want to simulate interaction between SnO2 and TiO2. By investigating some literatures, i got pair style and parameters for SnO2 and TiO2, respectively. I tried to apply the “pair_style hybrid” command to correlate SnO2 and TiO2.
Following is my idea. If it is nonsense, please hints other practicable plains.
atom_style: Sn—1 , O(s)—2 , Ti-----3, O(Ti)—4
Syntax:
pair_style hybrid buck/coul 10.0 buck/coul 10.0 eam.alloy ?(Unkown)
pair_coeff 3 3 buck/coul * * (Known）
pair_coeff 3 4 buck/coul * * （Known）
pair_coeff 4 4 buck/coul * * （Known）
Whether is this method right? If i know the interatomic potential between Sn and Ti, then their interactions can be simulated？
Any comments are appreciated.
Hai-Long Yao
State Key Laboratory for Mechanical Behavior of Materials
School of Materials Science and Engineering, Xi’an Jiaotong University
Xianning West Road 28, Xi’an, Shaanxi, 710049 China
Tel: +86 29 82665299
Fax: +86 29 83237910
E-mail：yhl206@…127…
Hi everyone,
i want to simulate interaction between SnO2 and TiO2. By
investigating some literatures, i got pair style and parameters for SnO2
and TiO2, respectively. I tried to apply the "pair_style hybrid" command to
correlate SnO2 and TiO2.
Following is my idea. If it is nonsense, please hints other practicable
plains.
atom_style: Sn---1 , O(s)---2 , Ti-----3, O(Ti)---4
Syntax:
pair_style hybrid buck/coul 10.0 buck/coul 10.0 eam.alloy ?(Unkown)
pair_coeff 1 1 buck/coul * * （Known）
pair_coeff 1 2 buck/coul * * （Known）
pair_coeff 2 2 buck/coul * * （Known）
pair_coeff 1 3 eam.snti.alloy * * (Unknown)
pair_coeff 1 4 buck/coul * * （Known）
pair_coeff 2 3 buck/coul * * （Known）
pair_coeff 2 4 buck/coul * * （Known）
pair_coeff 3 3 buck/coul * * (Known）
pair_coeff 3 4 buck/coul * * （Known）
pair_coeff 4 4 buck/coul * * （Known）
Whether is this method right? If i know the interatomic potential between Sn
and Ti, then their interactions can be simulated？
three things:
- you cannot just mix and match potentials. you have to understand
what assumptions they were parameterized on.
- you a hybrid approach (which is in essence a mechanical coupling of
two systems) usually only works well, if you have multiple separate
systems and you have a reasonable argument for using a simple
mechanical interaction via a simple, pairwise additive potential.
- you cannot use the eam/alloy interaction for the unknown part. you
are not modelling an alloy and - more importantly - any manybody type
potential, requires you to assign atom *types* to the potential (i.e.
cross-terms *and* self-interaction terms), so using your assignment
cannot work.
Hi Dr. Axel,
Thank you for reply. Unfortunately, I did not understand your mentioned second thing.
> you a hybrid approach (which is in essence a mechanical coupling of
>two systems) usually only works well, if you have multiple separate
>systems and you have a reasonable argument for using a simple
>mechanical interaction via a simple, pairwise additive potential.
Did you mean that should do multiple seperate systems, like TiO2, SnO2.
Meanwhile, apply a pairwise additive potential for TiO2/SnO2 (eg. Sn-Ti-O potential?)
to practice mechanical interaction bewteen TiO2 and SnO2?
Hi Dr. Axel,
Thank you for reply. Unfortunately, I did not understand your mentioned second thing.
> you a hybrid approach (which is in essence a mechanical coupling of
>two systems) usually only works well, if you have multiple separate
>systems and you have a reasonable argument for using a simple
>mechanical interaction via a simple, pairwise additive potential.
Did you mean that should do multiple seperate systems, like TiO2, SnO2.
Meanwhile, apply a pairwise additive potential for TiO2/SnO2 (eg. Sn-Ti-O potential?)
to practice mechanical interaction bewteen TiO2 and SnO2?
what i mean is that you need a consistent model. if you have a TiO2
crystal and an SnO2 crystal and want to study their interaction, you
are more likely to get a meaningful simulation with pair style hybrid
than if you want to model one crystal containing Sn, Ti and O.
Since you have oxides, you may still face problematic issues due to
polar surfaces. apparently, both sets of potentials use partial
charges, but they may have a different "balance" the two components of
the potential, so that they would not be compatible. similarly, each
of the potentials may be suitable for bulk materials, but not for
surfaces. and so on. there is many little things to consider with
multi-component systems.
It means that using the "pair_style hybrid" means that there are only interactions between atoms in TiO2 and SnO2, respectively.
And no interaction for atoms between TiO2 crysatl and SnO2 crystal.
could you point out what kind of "a consistent model" i should make?
No. You define interactions for all pairs of atoms. However, for some systems this approach to combine potential is meaningful and for others less so. That is what I mean.
Axel.