Variable charge potential in Lammps

LAMMPS LAMMPS Mailing List Mirror
1

Dear all

I am a new user of lammps, so I apologize if I ask something trivial…

I have to model metal-oxide interfaces (Al-Al2O3, in particular).

To this purpose I need variable-charge potentials.
In literature I could find some potentials with different levels of refinements, such as:

1 - Streitz, Mintmire, PRB 50, 11996 (1994)

2 - X. W. Zhuo et al, PRB 69, 035402 (2004)

3 - A. Hasnaoui et al, Surface Science 579, 47 (2005) {Vashishta’s group}

4 - T. J. Campbell et al, PRB 71, 205413 (2005) {Vashishta’s group}

5 - Lazic & Thijsse, Comput. Mat. Sci. 53, 483 (2012)

For my purposes (and for stability reasons) I would need from potential 2 on.

I have found some posts on the Mailing list where the Al2O3 is discussed

(but couldn’t figure out if just an ionic was considered or in company with some metals…).

But when I looked in the files of the ReaxFF and COMB potentials I could not
find “Al” in the elements (I installed the version of last 7th September 2013).

Also I could find online some presentations by Taho Liang of 2013 (from the Phillpott group) discussing the availability of Al2O3 potential in COMB-3 for lammps.

Some other presentations, instead, where saying that its implementation is still under way.

My questions are, therefore:

Are these variable charge potentials implemented in lammps ?

What is the proper file where I can find them ?
What potential is implemented ?

If not, do we have an estimation of when it will be implemented or do you
have any advice on how implement it myself ? Can I find online some guidelines to understand the structure of the pair_potential files in order to implement that properly so to eventually make it available to the community also ?

Thank you very much in advance for your help.

Best regards
Daniele Scopece

2

Dear all

I am a new user of lammps, so I apologize if I ask something trivial...

I have to model metal-oxide interfaces (Al-Al2O3, in particular).
To this purpose I need variable-charge potentials.
In literature I could find some potentials with different levels of
refinements, such as:
1 - Streitz, Mintmire, PRB 50, 11996 (1994)
2 - X. W. Zhuo et al, PRB 69, 035402 (2004)
3 - A. Hasnaoui et al, Surface Science 579, 47 (2005) {Vashishta's group}
4 - T. J. Campbell et al, PRB 71, 205413 (2005) {Vashishta's group}
5 - Lazic & Thijsse, Comput. Mat. Sci. 53, 483 (2012)

For my purposes (and for stability reasons) I would need from potential 2
on.

I have found some posts on the Mailing list where the Al2O3 is discussed
(but couldn't figure out if just an ionic was considered or in company with
some metals...).

But when I looked in the files of the ReaxFF and COMB potentials I could not
find "Al" in the elements (I installed the version of last 7th September
2013).

Also I could find online some presentations by Taho Liang of 2013 (from the
Phillpott group) discussing the availability of Al2O3 potential in COMB-3
for lammps.
Some other presentations, instead, where saying that its implementation is
still under way.

My questions are, therefore:
Are these variable charge potentials implemented in lammps ?

What is implemented in the official LAMMPS distribution is documented
on the LAMMPS webpage and in the bundled manual. all implementations
come with references and relevant descriptions, so you can match them
to the publications yourself. there may be additional implementations
of potentials by individual people, you have to contact them for their
sources.

What is the proper file where I can find them ?

proper file for what? potential file with the parameters? LAMMPS ships
only a few validated ones. there are *far* to many parameterizations
to ship all of them. where possible, the documentation provides
references for how to get them. otherwise you have to refer to the
corresponding paper and/or their authors.

What potential is implemented ?

see above.

If not, do we have an estimation of when it will be implemented or do you
have any advice on how implement it myself ? Can I find online some

impossible to say. implementing complex potentials is not a simple
task and requires suitable programming skills, a good understanding of
the relevant physics, perseverance, care and - sometimes - luck. it is
easier to implement a new potential that is similar to an existing
one, since you can take the existing source code as a template.

guidelines to understand the structure of the pair_potential files in order
to implement that properly so to eventually make it available to the
community also ?

the only way to understand code is by reading it. usually it is better
to start with a simple example and build your skills and get
acquainted with how LAMMPS works. there are available discussions in
the manual and online documentations and there is also the developer
guide with some additional insight into the inner workings of LAMMPS.

in general, please always remember TANSTAAFL.

axel.

3

For additional ReaxFF potentials the reax doc
page suggests you contact Adri van Duin and
gives the web site.

Ray Shan (here at Sandia) is working on
variable-charge potentials. He can answer
some of your other Qs.

Steve

4

Daniele,

Papers 2-5 of the five papers you cited are all based on the first
one, Streitz and Mintmire (SM) potential. Brief glimpse of papers 2-5
I found that they all use the same SM coulomb and charge transfer
scheme. Streitz and Mintmire will be release soon, and then maybe
somebody else, including you, can include more variations.

The 3rd generation COMB (Liang, Shan, Sinnott, Phillpot and
co-workers) will also be released soon, which will include Al2O3.

Currently the only variable charge potential for Al-O is reaxff, and
you can request its force field parameter file from Prof. van Duin.

Ray