GROMACS version: 2021.3
GROMACS modification: Yes/No
Here post your question
I want to calculate the interfacial tension between an alkane and water in absence and presence of a surfactants. With this I want to determine how much reduction in IFT occurs in presence of surfactant. For this purpose, I first equilibrated my system in NVT (after proper minimization with 100 kj/mol/nm force) and then made an NPT run with semiisotropic coupling wherein the compressibility of x/y was set to 0 (to maintain a constant area in these directions) and the compressibility along z direction was 4.5e-05. But the average surface tension is coming out to be negative (I used the surf*surftens command with the NPT.edr file to get the value of surface). Also I tried using the Pzz, Pxx and Pyy tensor to calculate the surface tension, both these methods give a negative value of surface tension. What should I do?
Read this ref. Also, ensure that your FF are properly parameterized
I would suggest also checking the refs that are cited therein, as well as the technical documentation of the codes that implement MD simulation.
There is always the possibility of small mistakes, especially in more recent papers/reviews. E.g. the paper linked says that GROMACS and NAMD don’t support pressure profiles along the Z axis. But there is at least a modified version of GROMACS that has been used in the past and NAMD supports those in the standard version.
Ultimately while no paper is technically perfect, they generally provide useful information that one can use to train themselves sufficiently well to understand all sources of error in the result, before they present it or publish it.
Giacomo
thanks for the help… I will read the paper. and the FF is properly parametrized. I am using OPLSAA forcefield for the job. The parameters for the surfactants were obtained from Jorgensen group’s ligpargen server and for the dodecane molecule, I am using parameters from a referred paper, which has been already published and gives results which are in good agreement with the experiments.
I have one additional question. How can I carry out an NPA_normalT ensemble simulation? How is this constant area condition maintained. Is this achieved via a semiisotropic pcoupletype with compressibility set to 0 along x/y direction?