# Problems with Lennard Jones potential parameters reading [closed]

So I was given this link, and was happy that I have now the epsilons and sigmas that I can use in Lennard-Jones potential calculations, but found out that the information in other sources is very different.

It says here:

"The ϵ and σ values for Xenon (Xe) are found to be 1.77 kJ/mol and 4.10 angstroms, respectively."

Which is different from the table in first link which says epsilon = 0.0238880 eV (2.304881456 kJ/mol) and sigma = 2.4945200 angstroms.

I understand that there could be some insignificant differences, but these are rather big ones. Maybe I am not correctly reading the table (ignoring cutoff) or something else?

• I'm not certain if this makes a difference, but your first link points to parameters for a shifted-LJ, where the energy values are shifted down by the energy value at some cutoff distance. I don't see how this would affect $\sigma$, but it may change the fit for $\epsilon$. The OpenKim model at least points sources for it's parameters in the README file. The ChemLibre text may have just made up values for the sake of example problems.
– Tyberius
Aug 24 at 13:23
• Its not just ChemLibre, but you might be right. Can I use the cutoff value to get more precise results? Aug 24 at 17:24
• +1 but I agree with Tyberius. The OpenKim data comes with references to publications whereas the ChemLibre data is just for some example problem in a textbook. Why not just use the OpenKim values and move on to the more exciting parts of your project? Since you are doing this as part of making a computer game, why do the sigma and epsilon values matter so much? Aug 24 at 18:56
• @NikeDattani you probably right ;) .. I will move on Aug 25 at 7:36
• Using cut or cut+shift can lead to very significant differences in results. If you are doing molecular dynamics, it is standard to use the shift, so that your force is smooth at the cut point. Where you cut at also makes a difference. make sure to cut at least 2.5*sigma, but for chemical potential calculations, cut at 5*sigma or more Aug 27 at 13:07