TraPPE-UA and HH-Alkane for GROMACS

James W. Barnett jbarnet4@tulane.edu

This is a GROMACS implementation of the TraPPE-UA force field with optional HH-Alkane modifications for the C-OW cross interactions (see below).

To install, place the trapepua.ff folder where GROMACS can find it, like in a directory pointed to by the environmental variable GMXLIB:

git clone https://github.com/wesbarnett/trappeua.git
cd trappeua
cp -r trappeua.ff $GMXLIB/

Then include trappeua.ff/forcefield.itp in your topology file like you do for other forcefields. An example topology file is found here.

The TIP4P2005 water model is included, and is the only water model that should be used with the HH-Alkane modifications. Ethanol is also included as solvent (HH-Alkane modifications don't apply here).

There are several git branches in this repo, some with personal modifications to the force field. The only branch that has the official TraPPE-UA parameters along with the HH-Alkane modifications is the master.

Lastly, if you use this implementation, please read and cite the following references.

• M.G. Martin, and J.I. Siepmann, J. Phys. Chem. B, 102, 2569 (1998).
• M. G. Martin and J. I. Siepmann, J. Phys. Chem. B 103, 4508 (1999).
• B. Chen, J.J. Potoff, and J.I. Siepmann, J. Phys. Chem. B 105, 3093 (2001).
• H.S. Ashbaugh, L. Liu, and L.N. Surampudi. J.Chem. Phys. 135, 054510 (2011).
• J. L. F. Abascal and C. Vega, J. Chem. Phys 123 234505 (2005).

A few implementation notes

• TrAPPE-UA recommends that a cutoff of 1.4 nm should be used for LJ interactions.
• TraPPE-UA recommends using Ewald summation for long-range electrostatic calculations. You should probably use Particle-Mesh Ewald.
• The number of exclusions for intramolecular non-bonded interactions should be set to 3 for TraPPE-UA (intramolecular interactions separated by four or more bonds use the same potential as intermolecular interactions). This is pretty typical.
• Intramolecular 1-4 LJ and Coulomb interactions are excluded in TraPPE-UA, so you will need to remove the [ pairs ] list for a molecule. gen-pairs is by default "no" under [ defaults ] and no [ pairtypes ] section exists, so you'll get an error about any [ pairs ] section.
• Currently two .rtp files are included in the force field directory. Only a few molecules are present at the time. Note that you can easily create linear alkanes / alcohols by specifying the end groups with residue CH3 and the interior groups as residue CH2, as long as each atom has the name C.

NOTE: Please read the following carefully.

• You should set constraints to all-bonds, since TraPPeUA bonds are fixed. If you do not, bonds will be flexible and an arbitrarily large spring constant is set in the force field.
• You can choose use the HH-Alkane modifications by adding define = -DWITH_HHALK_MODS to your mdp files, thus using the original TraPPE-UA force field. The HHALK modifications only affect water-oxygen carbon cross interactions.

Disclaimer

I've done my best in transcribing and converting the parameters for usage with GROMACS. That said, it is up to the user to ensure that the parameters are correct and that simulations are set up according to the methods presented in the papers above. There are several other TraPPE papers out there with more parameters for different types of molecules - I haven't implemented those yet. This current implementation only includes linear and branched alkanes and alcohols. Be sure to check out the TraPPE website.