**Due to the range of formats, we are not able to guarantee that the interatomic potentials will work with particular software packages. Rather, we provide the interatomic potentials with appropriate citation information, and it is up to the user to determine whether a potential will work with given software and produce the expected results.**

Steve Plimpton has a very nice description of various potential and file formats on the LAMMPS website. These include the **Embedded-atom method (EAM) funcfl** and **setfl** formats, **Finnis-Sinclair** potentials, and **Modified embedded-atom method (MEAM)** potentials. More detailed information can be found there. In addition to LAMMPS, the EAM funcfl and setfl formats are compatible with DYNAMO, the molecular dynamics program developed by Daw, et al. (1), though recent versions of LAMMPS have made some modifications to the header information. The EAM/FS format was designed for use of Finnis-Sinclair (2) interatomic potentials within LAMMPS.

The IMD molecular dynamics program also supports a range of interatomic potential types and formats. The file formats are described in the user guide found at http://www.itap.physik.uni-stuttgart.de/~imd/userguide/potformat.html.

SPaSM TABLE files have the following format:

* LINE 1 : Number of particle types, Number of tables per particle (this number should be set to 6 for 1 particle type)

* -- Begin Table 0 --

* LINE 2 : I J 0 (Table 0 for pair(I,J))

* LINE 3 : Number of grid points N

* LINE 4 : R_start R_Cutoff

* R.start n.nnnn

* ...

* R_Cutoff n.nnnn

*

** -- Begin Table 1--

* LINE 1 : I J 1 (Table 1 for electron density function rho(I,J))

* LINE 2 : Number of grid points N

* LINE 3 : R_start R_Cutoff

* R.start n.nnnn

* ...

* R_Cutoff n.nnnn

*

* -- Begin Table 2 --

* 1 : I J 2 (Table 2 for embedding function F(I,J))

* 2 : Number of grid points N

* 3 : rho_start rho_Cutoff

* rho.start n.nnnn

* ...

* rho_Cutoff n.nnnn

------------------------------------------------------------------------------------------------

The number of tables per particle should be set to 6 per type: since the input data is converted into the following internal tables

* Table function

----------------------------------------------------------------------------------------------

* 0 = Phi(r) - Tabulated in terms of r

* 1 = rho(r) - Tabulated in terms of r

* 2 = F(p) - Tabulated in terms of p

* 3 = Phi'(r)/r - Tabulated in r

* 4 = rho'(r)/r - Tabulated in r

* 5 = F'(p) - Tabulated in p

SPaSM formatting information courtesy of Ramon Ravelo (U. Texas El Paso).

__Modified Embedded-Atom Method (MEAM):__

The **"Modified Embedded-Atom Method"** or **"MEAM"** format extends the embedded-atom method to include angularly-dependent forces. The original MEAM references (including functional forms) are:

M.I. Baskes, "Application of the embedded-atom method to covalent materials: A semiempirical potential for silicon," Phys. Rev. Lett. **59**, 2666-2669 (1987).

M.I. Baskes, "Modified embedded-atom potentials for cubic materials and impurities," Phys. Rev. B **46**, 2727 - 2742 (1992).

Additional information can be found on the LAMMPS website.

__Angular-Dependent Potential (ADP):__

The **"Angular-Dependent Potential"** or **"ADP"** format also extends the embedded-atom method to include angularly-dependent forces. More information (including equations) c an be found in the following reference: Y. Mishin, M.J. Mehl, and D.A. Papaconstantopoulos, "Phase stability in the Fe-Ni system: Investigation by first-principles calculations and atomistic simulations," Acta Mat. **53**, 4029 (2005). DOI :10.1016/j.actamat.2005.05.001.

Prof. Mishin requested the following note be included: "The equation appearing in the Appendix on page 4040 contains a typing error: the sign before 1/3 in the last line must be negative." He provided the corrected equation for the angular-dependent force contributions in ADP_Forces.jpg or ADP_Forces.pdf.

**"ADP tabulated functions"** are pairs of (x,f(x)) values as described in the headers of those files.

More information about the **"GULP"** format is available at http://nanochemistry.curtin.edu.au/gulp/.

__MoldyPSI:__

The **"MoldyPSI"** format combines a parameter file with the Fortran90 routines to calculate the EAM functions and their derivatives.

__Table:__

The **"table"** formats are described in the headers of the files. They are pairs of (x,f(x)) values such as (ρ,F(ρ)) as described in the references associated with each interatomic potential.

In addition to the citation information for the interatomic potential, please include "NIST Interatomic Potentials Repository: http://www.ctcms.nist.gov/potentials" in the references.

Send the potential to potentials@nist.gov with format and citation information. A prototype page will be created that will be posted publicly upon the approval of the submitter.

LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) is developed and maintained at Sandia National Laboratories.

The purpose of this project is to provide a repository of interatomic potentials for atomistic simulations (e.g. molecular dynamics) with comparison tools and reference experimental and ab-initio data in order to facilitate the evaluation of these potentials for particular applications. Our goal is not to judge that any particular potential is "the best" because the best interatomic potential may depend on the problem being considered. For example, some interatomic potentials which have been fit only to the properties of solid phases may model solid surface properties better than one fit with solid and liquid properties. However, the second potential will probably better represent properties which have a strong liquid contribution (e.g. crystal-melt interfacial properties). Additionally, we are not limiting the repository to a single class of material (e.g. metals), interatomic potential format (e.g. Embedded-Atom Method), or software package. As we obtain interatomic potentials for other materials or in other formats, we will include them with proper website modifications.

There can be several reasons for this. One is file formatting. Different developers and software packages use different data and file formats to present their interatomic potentials.

Even with the same file/data format, invariant transformations in the EAM format mean that different parameterizations of the interatomic potentials can look completely different but yield the same physical properties. This is especially true for alloys. The issue of invariant transformations is discussed in several places. Among them are:

- Y. Mishin, "Interatomic potentials for metals," in
*Handbook of Materials Modeling*, edited by S. Yip (Springer, Dordrect, The Netherlands, 2005), Chap. 2.2, pp. 459-478. - A.E. Carlsson, "Beyond pair potentials in elemental transition metals and semiconductors," in
*Solid State Physics*, Volume 43, Ehrenreich and Turnbull (eds.) (1990).

Please send an email to potentials@nist.gov which includes the publication information. This will be used to create a list of papers citing the website in order to help us better understand how the website is being used

Date Created: October 5, 2010 | Last updated: November 20, 2020