SEAMM Packmol Plug-in

A SEAMM plug-in for building periodic boxes of fluid using Packmol

This plug-in takes the molecule in the current system and creates a periodic box containing many copies of the molecule in order to simulate a fluid.

• Free software: BSD license

• Documentation: https://molssi-seamm.github.io/packmol_step/index.html

• Code: https://github.com/molssi-seamm/packmol_step

History

2026.10.15 – Bugfix: molecules overlapped across the faces of a periodic cell

• Packmol’s periodic boundary conditions, used for periodic cells, do not keep molecules apart across the faces of the cell: in a cell of 512 waters, 25 to 53 pairs were closer than 1.8 Å, some at 0.8 Å, so the MD that followed could fail at once with “Lost atoms”. The molecules are now packed 1 Å inside each face, so they are at least 2 Å from those across it.

2026.10.12 – Forcefields from the Model Chemistry step

• The molecules are typed with a forcefield chosen in a Model Chemistry step (FF@oplsaa, …) as well as with the Forcefield step’s.

• Requires forcefield-step 2026.10.12.1.

2026.10.10 – Concentrations, documentation, and structures as sources keep their bonds

• The amount of each fluid molecule can be given as a proportion, a molality (mol/kg of solvent) or a molarity (mol/L), and the proportions by number of molecules or by mass, e.g. 1.0 M LiPF6 in EC:DMC 3:7 by weight. The molecules given as proportions are the solvent. Molarity is not offered when the volume comes from the Ideal Gas Law.

• The table of what was packed gives the requested and actual amounts of each molecule, its molarity and, for molecules other than the solvent, its molality.

• New user guide and getting-started pages, with three example flowcharts.

• Bugfix: with two or more structures from the database as sources (e.g. PF6/initial and LiEC4/initial), every source after the first lost its bonds, so the forcefield could not be assigned (PF6- was seen as a P plus six F atoms). The step also deleted the bonds of every structure in the database. Source structures are now left untouched.

• Bugfix: rectangular boxes sized from a volume, density or the Ideal Gas Law now have the side ratios given in the dialog; before, they were ignored.

• Bugfix: the “-” button of a molecule removed the wrong row after an earlier row had been removed.

• Bugfix: without a packmol.ini the step failed unless SEAMM was installed with conda. It now finds Packmol on the PATH and records it in packmol.ini.

• A source given only by its configuration name now uses the current system, and a solute’s proportion no longer changes the numbers of atoms or the mass.

• Flowcharts that get the amount of fluid from the density or the Ideal Gas Law now validate; the unused “gap” parameter says so.

• Removed the unused Docker image and its release job; the tests now run with the package’s own dependencies, on Python 3.12.

2026.3.1 – Internal: switching from deprecated library pkg_resources to importlib

2026.1.7 – Enhancement: using periodic packing now in Packmol

• Added support for the PBC boundary conditions now that they are implemented in Packmol. This seems to lead to considerably better packing, particulalrly with periodic solutes like MOF’s and zeolites.

2025.12.19 – Bugfix: initialization of random numbers

• Added the command to the Packmol input to set the random seed using the system clock so that each run is different.

2025.9.6 – Bugfixes: centering of solute and using variables for compounds

• Fixed a problem centering a solute in a cubic cell.

• Fixed problems using variables for the compounds, counts, etc. in the input.

• Fixed a problem if the number of molecules of a compound is zero. Now it is just ignored.

2025.6.20 – Added support for MACE PyTorch potentials

2025.3.9 – Update for new installer

• Changes in the installer required changes for codes that installed background codes, like PACKMOL. This update allows PACKMOL to be automatically installed.

2025.1.29 – Bugfixes: executable and SMILES

• The configuration for the executable was not set up correctly.

• There was a bug due to change in the SMILES interface for configurations.

2024.7.25 – Bugfix: initial setup of packmol.ini file

• The initial setup of the packmol.ini file did not add the information for Conda. This release fixes that problem.

2024.6.29 – Bugfix: bonding incorrect in some cases.

• The use of PDB with Packmol could in some cases lead to multiple bonds in the wrong place. This release fixes that problem.

2024.6.21.2 – Another internal release for Docker.

2024.6.21.1 – Internal release for Docker

• There was an internal issue creating the Docker image.

2024.6.21 – Switching to RDKit for SMILES

• Using RDKit for SMILES since we found some issues with OpenBabel, and also the atom typing uses RDKit, so this is more compatible.

2024.3.19 – Updated installer for new scheme

• packmol-step-installer now uses the new scheme, which supports both Conda and Docker installation.

• Added seamm-packmol Docker image

2024.1.16 – Adding support for containers

• Added the ability to work in Docker containers.

2023.9.6 – Using fractional coordinates for periodic systems.

• By convention, SEAMM is using fractional coordinates for periodic systems. The PACKMOL step was creating periodic structures with Cartesian coordinates, which can be a bit confusing, though it did not affect any reauls. This change fixes the issue.

2023.2.15 – Restructured documentation and moved to new theme.

2022.5.31 – Substantial enhancements

Added ability to solvate molecules and increased the options to include

• spherical regions

• cubic and rectangular regions

• cubic and rectangular unit cells for periodic systems

• input molecules from any combination of system/configurations and SMILES

• reworked GUI to be more intuitive.

2021.11.27 – Fixed bug with atom types.

2021.10.25 – Bugfix for problems with ideal gas approach.

2021.8.29 – Multiple different molecules available via SMILES

This release adds

• the ability to directly generate multiple different molecules from SMILES, giving the stoichiometry for packing into the fluid box.

• specifying the desired cell with the temperature and pressure, along with one of the volume, length of the box side, number of molecules or number of atoms. This uses the ideal gas law, so will only be reasonable at higher temperatures and lower densities.

Additionally, internally the release adds integration tests to check whether the module works properly in the SEAMM environment.

2021.6.3 – internal change for improvements in parsing the commandline

2021.5.25 – Added installer for Packmol executable.