Hello.
I am using LAMMPS 64-bit 3Aug2022-MPI version to make fix widom. I want to get N2 solubility from excess chemical potential by TPI method. The system is polyimide polymer.In my simulation, the excess chemical potential is zero. What do we do about that
The input file is as follows.
Thank you for your consideration.
Best regard！
Yours sincerely，
Gaoyang
in.widom (4.1 KB)
N2.mol (359 Bytes)
after_deform.data (5.5 MB)
Fix widom has three properties that can be output. Are they all zero?
Your simulation run has only 100 steps. Why so few?
Thank you for your reply.
In my simulation, the average difference between the excess chemical potential and the potential energy is zero. My understanding of fix widom is that the output result is related to the number of insertions of each step, and the insertion time of each step is set to be 5000 times, and the simulation time will be extended later.
But I don’t think fix widom’s output should be zero despite the small number of steps.There are 17,730 atoms in the system. fix widom needs to run for so long that it can only run 30 steps a day. Is that normal?
Have a nice day
Well, what if it cannot find a suitable “hole” to insert the molecule?
… and since you disable the regular MD my wiping out the forces, there is no chance since atoms don’t move. For a condensed system and a difficult to insert molecule (versus a single atom), you would stand a better change by letting atoms evolve and making far fewer attempts and only do them not in every timestep, since the system cannot change much in just one step.
It is a consequence of your system and settings. Due to the use of kspace, you need to do a full force evaluation for each insertion attempt. With 5000 attempts per timestep, you have actually done the equivalent of 150000 MD steps without fix widom during the 30 steps. Fix widom is based on fix gcmc which in turn was conceived to more efficiently study the diffusion of noble gases in zeolites, the choice of locations for insertions is not very sophisticated (unlike in some MC codes) and thus the chances for successful insertion depend a lot on the density of the system…
Thank you very much for your reply, it is very important to me.
Previously, I tried fix gcmc with the same structure, and calculated the adsorption amount of N2 under 5 pressure gradients (maximum 10atm), showing that the number of N2 molecules in the system was zero. However, the problem was that I imported the structure into MaterialsStudio and calculated the adsorption isotherm with the sorption module. I guess there’s a place for the structure to insert atoms.
Based on the results of MS software, I have been modifying my lammps input file and made some attempts, but there is still no result. I’m going to cancel the fixed structure and try again.
Best regards,
Luo
Additionally，for fix gcmc, I have a question to ask you, at runtime, there are no atoms in the FIXGCMC group, is this an error? I provided a screenshot of the input file and the error.
in.adsorption (4.7 KB)
Hello Igy,
I would like to know if you can solve anything about the mistakes you were getting. I have the same mistakes, I get zero in the excess chemical potential. And at the end I can obtain the same 0 atoms in groupfix.
I really hope you can reply me.
I’m sorry I can’t help you. I did not solve this problem, you may need to talk to others.
Did you check another metodology to obtain excess chemical potencial? and which one?
One possible reason is the molecule template. If its format has any error, the insertion will fail.
Sorry, but this makes no sense. Molecule files are read and parsed with the molecule command. If there is an error in the file, it will be reported then.
Any reference to the molecule reference will use the internal data and thus cannot cause an error.