LAMMPS Implementation of Streitz-Mintmire (CTIP) Variable Charge Potential

Gabriel_Plummer · June 22, 2022, 6:04pm

Hello LAMMPS Users,

I am looking to run some metal oxidation studies and am interested in using the Streitz-Mintmire (CTIP) potential form to handle the variable charges. I can see in the LAMMPS documentation that the original Streitz-Mintmire formalism has been implemented with the coul/streitz pair style. However, I would like to know if the valence charge bound corrections proposed by Zhou et al (Phys. Rev. B 69, 035402 (2004) - Modified charge transfer--embedded atom method potential for metal/metal oxide systems) have also been implemented? I do not see this anywhere in the documentation, but have seen it used in several recent publications making use of LAMMPS (for example: https://doi.org/10.1021/acs.chemmater.8b03969), and it is also referenced in the most recent LAMMPS publication (Redirecting). Any clarity on this issue would be appreciated.

Thanks,
Gabriel Plummer

akohlmey · June 22, 2022, 6:21pm

I have sent an email to the developer of the coul/streitz pair style to provide his comments on that.
When I look at the potential files (AlO.streitz and AlO.eam.alloy) it seems they are consistent with what you are looking for. However, it also looks like the corrections proposed by Zhou only affect the EAM potential file not the pair style itself.

Gabriel_Plummer · June 22, 2022, 6:33pm

Thank you, Axel. My main concern from preliminary testing is that when I apply compressive isotropic strain to an Al2O3 cell, the charges on Al and O readily exceed +3 and -2 respectively (this occurs at only around 2% strain). Including the charge bound corrections of Zhou et al should not allow this to happen. Those corrections also require the definition of q(min) and q(max) for each atomic species and a coefficient, omega. For the coul/streitz pair style I do not see any place for the definition of these parameters, which made me think it had not been implemented.

akohlmey · June 22, 2022, 7:14pm

That is definitely a question for the developer/contributor of the pair style. This sounds like it would be a feature that needs to be available in the charge-equilibration fix. Coincidentally, those fixes and the QEQ package were written by the same person. If you need his contact info please let me know.

Gabriel_Plummer · June 22, 2022, 7:22pm

Yes, that would be perfect if you would like to email me ([email protected]). Thanks!

kiran.sasikumar · November 12, 2024, 9:39am

Hello Gabriel and Alex,

Was there any movement regarding the implementation of Zhou et al. modified charge-transfer embedded atom method in LAMMPS? I ask this as the author of the paper you had referenced in your original question and since I am now working on making the LAMMPS extensions developed for that paper to work with newer versions of LAMMPS.

NOTE: To avoid confusion, I prefer to call Zhou’s modified charge-transfer EAM as CTIP and the implementation in LAMMPS as the Streitz-Mintmire (SM) potential. For single-metal oxides and under most equilibrium conditions, CTIP and SM should yield identical results. Zhou and Wadley’s CTIP has some other differences under-the-hood compared to the Strietz-Mintmire potential beyond the constraints for qmin and qmax. However, the parameter file only has two additional columns for the charge bounds.

As part of the work 6-7 years ago, we had implemented CTIP as an extension to LAMMPS. Unfortunately back then we had not distributed the CTIP code along with the paper and I had moved on to other priorities and jobs since.

It is now available at Lammps-Extensions / CTIP · GitLab. The only problem is that it was implemented for the 2015/2016 versions of LAMMPS. Due to some other requests I had to revisit this old piece of code and I am currently trying to make it work for the newer versions of LAMMPS.

Regards,
Kiran

stamoor · November 12, 2024, 3:49pm

github.com/lammps/lammps

Charge Transfer Interatomic Potential (CTIP) pair style and qeq fix

lammps:develop ← gplummer317:ctip

opened 09:06PM - 31 May 24 UTC

gplummer317

+21550 -32

**Summary** Implementation of CTIP, which is a variable charge potential and …an improved version of the existing coul/streitz potential (see Zhou et al., Phys Rev B, 69, 035402 (2004)). CTIP implements a charge bounding scheme which provides greater numerical stability to the potential and allows for more physical charge transfer in systems with >2 components. There are also several efficiency improvements to the original CTIP model which have been incorporated here: -Charge bounds enforced with a piecewise quadratic function which is compatible with the existing qeq matrix inversion algorithm. -Long range electrostatic interactions are calculated with a damped, shifted potential rather than Ewald summation (see pair_style coul/dsf). -Coulomb integrals are replaced with an empirical charge shielding function as in the reaxff potentials. **Related Issue(s)** **Author(s)** Gabriel Plummer - NASA Ames Research Center [email protected] [email protected] **Licensing** By submitting this pull request, I agree, that my contribution will be included in LAMMPS and redistributed under either the GNU General Public License version 2 (GPL v2) or the GNU Lesser General Public License version 2.1 (LGPL v2.1). **Backward Compatibility** **Implementation Notes** 3 new parameters have been added to the qfile which is read in as part of the fix qeq commands to enforce charge bounding - qmin, qmax, omega. This should not affect the other fix qeq styles which will not utilize these parameters. 2 new keywords have been added for fix qeq/ctip - cdamp and maxrepeat which control the damping of the coulomb potential and the self consistency of qeq calculations with the piecewise quadratic function, respectively. An Ni-O CTIP parametrization has been developed to perform tests on the new features - it utilizes the hybrid/overlay pair style in combination with an eam/fs potential. This parametrization can be made available by request if any testing is needed. **Post Submission Checklist** - [x] The feature or features in this pull request is complete - [ ] Licensing information is complete - [x] Corresponding author information is complete - [x] The source code follows the LAMMPS formatting guidelines - [x] Suitable new documentation files and/or updates to the existing docs are included - [x] The added/updated documentation is integrated and tested with the documentation build system - [x] The feature has been verified to work with the conventional build system - [x] The feature has been verified to work with the CMake based build system - [x] Suitable tests have been added to the unittest tree. - [x] A package specific README file has been included or updated - [x] One or more example input decks are included **Further Information, Files, and Links**

Gabriel_Plummer · November 12, 2024, 4:04pm

Hi Kiran,

Yes, as linked by Stan, our implementation is now available on the LAMMPS develop branch. Please read the description there before using though, as it is not identical to the original Zhou et al CTIP. There should be an example script to use it there as well. We currently have a paper in review, which will provide more details and example use cases. In the meantime, please feel free to message me with any questions.

Best,
Gabriel

kiran.sasikumar · November 12, 2024, 4:31pm

Thank you, Stan and Gabriel, especially to push this through and to have CTIP available as part of LAMMPS. I will test the code currently in the LAMMPS developer branch just to see how it compares against my old/naive implementation of the same, which likely does not have the efficiency improvements that you have incorporated. The reason I was digging through my old code was because I wanted to integrate CTIP with Echemdid (GitHub - nonofrio/LAMMPS-hacks-public: EChemDID and various USER packages for LAMMPS.) where the charge equilibration is dependent on an applied external potential. If my tests work out, then I think it would be beneficial to add that capability to your implementation in LAMMPS. I will keep you posted.

Regards,
Kiran