IPC for LAMMPS

Implementation of Inverse Patchy Colloids for LAMMPS.

We provide all that is necessary to simulate IPCs with LAMMPS: sample .in files and startingstates, a program to print and tabulate the IPC potential, and generators of startingstates. Anybody can start simulating IPCs in LAMMPS in few minutes. For any help, feel free to contact us here on GitHub.

Installation Instructions

Requires: a version of g++ recent enough to compile c++14, and CMake 3.5 or newer.

There are two ways to install this package.

The simplest is to use git: clone the repo using git clone https://github.com/Zirbo/IPC_LAMMPS.git. That's it. The scripts will take care of everything else.

If you don't want to use git, it's a bit more complicated. You have to download the repository as zip, unpack it, install the Eigen library via package manager or from source, and the point to it in the compilation instructions in the last line of IPC_LAMMPS/1-make_potentials/sources/CMakeLists.txt. You might come to issues if the version of g++ and Eigen that you are using are not compatible with each other.

We have the following directories:

1-make_potentials

Contains programs that compute the potentials in LAMMPS format, and some helpers.

from_contact_values_to_potentials.sh

(requires a c++14-compatible version of g++ and uses Eigen library, which will be downloaded automatically)

Requires as input (modify the file to supply the inputs!) the geometric parameters and the desired contact values of the potential in the three EE, EP1 and P1P1 orientations (plus EP2, P1P2, P2P2 for asymmetric models) or BB, BP and PP for Janus, and prints out the potential in an output directory, named target_<modelname>_<symmetry>_contact.

The directory contains a lammpspot dir that contains your potential. In addition there will be three directories:

• lammpspot_angular_plots
• lammpspot_radial_plots
• potential_on_path whose content can be plotted using gnuplot or xmgrace, so you can quickly visualize your potential. See the reference paper [3] for an explanation of the orientations.

Finally there is also a file, inputfile.dat, that recaps your input values.

from_epsilon_to_potentials.sh

Requires as input (modify the file to supply the inputs!) the geometric parameters and the epsilons from the mapping (see the reference papers!) and prints out the potential in the output directory named target_<modelname>_<symmetry>_epsilons.

It is a wrapper to the same program used by from_contact_values_to_potentials.sh. Outputs and requirements are exactly the same.

compute_ipc_geometry.py

(requires python3)

In case you want to follow the IPC geometry, this python script helps you determine eccentricity and patch radius from patch amplitude and interaction range, or viceversa. Check the inline help with -h.

2-startingstate_creators

Contains Python3 scripts that can be used to generate FCC startingstates for Janus and two-patch IPCs. There's also one cubic lattice for two-patch only. All of them have an inline help, so run them with the -h flag to get explanations.

3-lammps_inputfiles

Each subdirectory of it contains a sample .in file and a sample startingstate. To run run a simulation you have to:

• copy th .in file to the directory where you want to run
• copy in that directory the potential and startingstate that you want to use
• adapt the .in file with paths to startingstate, potentials and eccentricities.
• run LAMMPS:

$ mpirun -np <num_cpus> </lammps/exe/path> -in run_*.in -var seed $RANDOM

where the seed is used to initialize the velocities ($RANDOM is a bash command). Your LAMMPS distribution needs to be compiled with the MOLECULE package.

There is also a sample tk script that you can load in VMD (Visual Molecular Dynamics) to visualize your trajectory:

$ vmd -e vmdscript.tk </path/to/trajectory.lammpstrj>

4-postprocess

!!! currently WIP !!!

• does not support Janus
• does not support Asymmetric IPCs
• works with any symmetric IPC, although the code still refers to IPCs

Run the build.sh script to obtain the binary; a c++11-compatible version of g++ and CMake 3.5 are required.

Run the obtained binary without arguments to see what arguments are required. You need a valid trajectory, the directory with the potentials used to run the simulation, and an inputfile with eccentricity and patch radius.

Will be extended to all IPCs as soon as possible :)

9-advanced

Contains other startingstate generators and visualization tools that we have been using in our research. You are welcome to peep, but they are not all well documented, so use them at your own risk ;)

Have fun!

We hope that you will have as much fun playing with IPCs as we did :D S.F. & E.B.

References

[1] "Inverse patchy colloids: from microscopic description to mesoscopic coarse-graining" E. Bianchi, G. Kahl, and C. N. Likos - Soft Matter 7, 8313 (2011)

[2] "Molecular dynamics simulations of inverse patchy colloids" S. Ferrari, G. Kahl, E. Bianchi - Eur. Phys. J. E 41, 43 (2018)

[3] "" S. Ferrari, E. Locatelli, E. Bianchi - in prep.