How to get parameters of interaction between atoms of H2 and C6H6?

[Mikhail_V]

HI, I want to reproduce results from the article (attached).

I need to get parameters of interaction between atoms of H2 and C6H6 for Morse equation.
I easily got results of interation between MOLECULES H2 and C6H6, shown in Table S2 'The first-principles and Morse force field (eq.1)...', but I do not know how to get parameters between ATOMS of H2 and C6H6, shown in Table S4
I used ORCA for my caclulations.
Any suggestions are welcome

 

[SpectraCat]

The data in table S4 are still for the intermolecular potentials as far as I can see. They are simply giving the 1-D Morse potential parameters for the intermolecular bonds formed in various geometries. See figure S6 for some examples.

 

[Mikhail_V]

The data in table S4 are still for the intermolecular potentials as far as I can see. They are simply giving the 1-D Morse potential parameters for the intermolecular bonds formed in various geometries. See figure S6 for some examples.

Thank you for your answer, but how can I repeat these calculations. I mean, how can I get 1-D Morse potential parameters. For example for C_R ---H_A (Table S4) do I need to use specific technique to exclude the influence of other atoms in cluster?

 

[SpectraCat]

Thank you for your answer, but how can I repeat these calculations. I mean, how can I get 1-D Morse potential parameters. For example for C_R ---H_A (Table S4) do I need to use specific technique to exclude the influence of other atoms in cluster?

No, what you need to do is calculate the intermolecular potential at various points along the specified intermolecular coordinate, with all other intermolecular parameters fixed. You then fit those points to some analytical form .. in this case a Morse potential. Note that calculating the intermolecular potential in this way is non-trivial. You typically should do a partial minimization, where the intermolecular geometry is kept fixed at each desired configuration, but the individual molecular geometries are allowed to relax ... I doubt it matters much in this case, since the potentials seem pretty shallow, but it is usually better to be on the safe side. Furthermore, extracting the intermolecular potential at each point is also non-trivial .. usually this is done by the counterpoise method of Boys and Bernardi. Most molecular structure packages have counterpoise corrections as built-in routines.

 

[Mikhail_V]

No, what you need to do is calculate the intermolecular potential at various points along the specified intermolecular coordinate, with all other intermolecular parameters fixed. You then fit those points to some analytical form .. in this case a Morse potential. Note that calculating the intermolecular potential in this way is non-trivial. You typically should do a partial minimization, where the intermolecular geometry is kept fixed at each desired configuration, but the individual molecular geometries are allowed to relax ... I doubt it matters much in this case, since the potentials seem pretty shallow, but it is usually better to be on the safe side.

I calculated the intermolecular potential at various points along the specified intermolecular coordinate (H_ --- H_A, please, see attachment mentioned above), the potential well is -0,117 kcal/mol (much higher). All other atoms were kept fixed via constraints. How can I exclude the impact of other atoms of system C6H6---H2 in simulated potential H_---H_A?

Furthermore, extracting the intermolecular potential at each point is also non-trivial .. usually this is done by the counterpoise method of Boys and Bernardi. Most molecular structure packages have counterpoise corrections as built-in routines.

Thank you. This correction is really important.

 

[SpectraCat]

Ok, I think I understand better what you were asking in your original question. I see now that the authors do appear to have done some sort of atomistic decomposition of the different parts of their interaction potential. That means that they treated the overall interaction as a sum over all all of the individual atomic interactions .. that is why they have their Morse potential parameters indexed by atom in tables S3 & S4 .. their use of the term "atom type" in several places supports this interpretation. It makes a kind of sense as well, because they are trying to generate force field parameters .. I am just surprised that they are assuming isotropic potentials and completely ignoring the angular dependence of the potential which is certainly significant. Personally I think this is *highly* approximate, and I would not expect it to give good agreement with experimental data ... however, that is only my opinion, and I am not an expert in the field of force field generation. I strongly urge you to read the authors earlier work in that field, since they may explain the reasons why they can ignore angular dependence.

Anyway, you can do this same treatment yourself easily enough, and see if you get the same results. You just need to set up the sums properly, which is a bit tedious and tricky, but quite doable. You will also have more fitting parameters to deal with of course .. one set for each atom type.

One other point .. it doesn't look like the authors used the counterpoise corrections for their potentials, based on the equation on page S2, so you may want to try fitting both your corrected and uncorrected surface to see which gives better agreement with the paper.

Hope this helps.