# pair_style buck/long/coul/long command

Accelerator Variants: buck/long/coul/long/omp

## Syntax

pair_style buck/long/coul/long flag_buck flag_coul cutoff (cutoff2)

• flag_buck = long or cut

long = use Kspace long-range summation for the dispersion term 1/r^6
cut = use a cutoff
• flag_coul = long or off

long = use Kspace long-range summation for the Coulombic term 1/r
off = omit the Coulombic term
• cutoff = global cutoff for Buckingham (and Coulombic if only 1 cutoff) (distance units)

• cutoff2 = global cutoff for Coulombic (optional) (distance units)

## Examples

pair_style buck/long/coul/long cut off 2.5
pair_style buck/long/coul/long cut long 2.5 4.0
pair_style buck/long/coul/long long long 4.0
pair_coeff * * 1 1
pair_coeff 1 1 1 3 4


## Description

The buck/long/coul/long style computes a Buckingham potential (exp/6 instead of Lennard-Jones 12/6) and Coulombic potential, given by

$\begin{split}E = & A e^{-r / \rho} - \frac{C}{r^6} \qquad r < r_c \\ E = & \frac{C q_i q_j}{\epsilon r} \qquad r < r_c\end{split}$

$$r_c$$ is the cutoff. If one cutoff is specified in the pair_style command, it is used for both the Buckingham and Coulombic terms. If two cutoffs are specified, they are used as cutoffs for the Buckingham and Coulombic terms respectively.

The purpose of this pair style is to capture long-range interactions resulting from both attractive 1/r^6 Buckingham and Coulombic 1/r interactions. This is done by use of the flag_buck and flag_coul settings. The Ismail paper has more details on when it is appropriate to include long-range 1/r^6 interactions, using this potential.

If flag_buck is set to long, no cutoff is used on the Buckingham 1/r^6 dispersion term. The long-range portion can be calculated by using the kspace_style ewald/disp or pppm/disp commands. The specified Buckingham cutoff then determines which portion of the Buckingham interactions are computed directly by the pair potential versus which part is computed in reciprocal space via the Kspace style. If flag_buck is set to cut, the Buckingham interactions are simply cutoff, as with pair_style buck.

If flag_coul is set to long, no cutoff is used on the Coulombic interactions. The long-range portion can calculated by using any of several kspace_style command options such as pppm or ewald. Note that if flag_buck is also set to long, then the ewald/disp or pppm/disp Kspace style needs to be used to perform the long-range calculations for both the Buckingham and Coulombic interactions. If flag_coul is set to off, Coulombic interactions are not computed.

The following coefficients must be defined for each pair of atoms types via the pair_coeff command as in the examples above, or in the data file or restart files read by the read_data or read_restart commands:

• A (energy units)

• rho (distance units)

• C (energy-distance^6 units)

• cutoff (distance units)

• cutoff2 (distance units)

The second coefficient, rho, must be greater than zero.

The latter 2 coefficients are optional. If not specified, the global Buckingham and Coulombic cutoffs specified in the pair_style command are used. If only one cutoff is specified, it is used as the cutoff for both Buckingham and Coulombic interactions for this type pair. If both coefficients are specified, they are used as the Buckingham and Coulombic cutoffs for this type pair. Note that if you are using flag_buck set to long, you cannot specify a Buckingham cutoff for an atom type pair, since only one global Buckingham cutoff is allowed. Similarly, if you are using flag_coul set to long, you cannot specify a Coulombic cutoff for an atom type pair, since only one global Coulombic cutoff is allowed.

Styles with a gpu, intel, kk, omp, or opt suffix are functionally the same as the corresponding style without the suffix. They have been optimized to run faster, depending on your available hardware, as discussed on the Speed packages doc page. The accelerated styles take the same arguments and should produce the same results, except for round-off and precision issues.

These accelerated styles are part of the GPU, USER-INTEL, KOKKOS, USER-OMP and OPT packages, respectively. They are only enabled if LAMMPS was built with those packages. See the Build package doc page for more info.

You can specify the accelerated styles explicitly in your input script by including their suffix, or you can use the -suffix command-line switch when you invoke LAMMPS, or you can use the suffix command in your input script.

See the Speed packages doc page for more instructions on how to use the accelerated styles effectively.

## Mixing, shift, table, tail correction, restart, rRESPA info

This pair style does not support mixing. Thus, coefficients for all I,J pairs must be specified explicitly.

This pair style supports the pair_modify shift option for the energy of the exp() and 1/r^6 portion of the pair interaction, assuming flag_buck is cut.

This pair style does not support the pair_modify shift option for the energy of the Buckingham portion of the pair interaction.

This pair style supports the pair_modify table and table/disp options since they can tabulate the short-range portion of the long-range Coulombic and dispersion interactions.

This pair style write its information to binary restart files, so pair_style and pair_coeff commands do not need to be specified in an input script that reads a restart file.

This pair style supports the use of the inner, middle, and outer keywords of the run_style respa command, meaning the pairwise forces can be partitioned by distance at different levels of the rRESPA hierarchy. See the run_style command for details.

## Restrictions

This style is part of the KSPACE package. It is only enabled if LAMMPS was built with that package. See the Build package doc page for more info.

## Default

none

(Ismail) Ismail, Tsige, In ‘t Veld, Grest, Molecular Physics (accepted) (2007).