# ORCA: How to plot an adiabatic potential in dihydrogen H2 molecule?

I stared to study ORCA, and I try to obtain classical results for dihydrogen as for example. I need to get a starting point to understand what needs to be done. So, How to plot an adiabatic potential in dihydrogen H2 molecule with ORCA calculation?

! RHF OPT def2-QZVPP
%geom Scan
B 0 1 = 1.0, 3.0, 12
end
end

* xyz 0 1
H        0.000000       0.00000        0.00000
H        0.800000       0.00000        0.00000
*


• An optimisation only finds you the minimum of the curve, it doesn’t trace out the entire curve; for that you need a scan. Feb 6, 2021 at 11:38
• @orthocresol How and what output files should I use to build? Feb 6, 2021 at 11:50
• In addition to the answer, plus the ORCA manual and input library which I suggested on your last question, I can also recommend youtube.com/playlist?list=PLxm4o04Xz3Bb7q5EKZNHnA8peEuJQYk0L. Feb 6, 2021 at 12:11

By adiabatic, I presume you mean the Born-Oppenheimer approximation (which is usually used). If you need a plot of potential against H-H distance, then you need to do a relaxed potential energy surface (PES) scan.

! UHF OPT def2-QZVPP
%geom Scan
B 0 1 = 0.3, 1.3, 30
end
end

* xyz 0 1
H       -4.61685        1.79381        0.00000
H       -4.14986        1.26166        0.00000
*


The Scan directive in %geom section will perform a relaxed geometry scan (although in this case you have only two atoms so there is no difference between relaxed and unrelaxed). The B 0 1 = 0.3, 1.3, 30 line will scan the bond (B) between atom 0 and atom 1 (in Orca, atom counting starts from 0). The scanning will start from the bond distance 0.31Å and will end at 1.3Å and will go through 30 steps (i.e. the PES will have 30 points in total).

Another thing is that you need to use the unrestricted hartree-fock (UHF) because when the H atom splits the two electrons go into different orbitals, so RHF will give the wrong behaviour at higher bond distances (it will force double occupancy in one orbital, resulting in $$\ce{H+}$$ and $$\ce{H-}$$ fragments). UHF gives the right behaviour but may not reproduce the correct energy. Multi-configurational methods might be required.

The input file that I have shown will show you the energy behaviour of the ground state (in your graph, $$\mathrm{U_S}$$). Plotting the first excited state (antibonding) i.e. the $$\mathrm{U_A}$$ graph would be more difficult. You would probably need to do some sort of orbital rotation to converge to the excited triplet state, and then use that as the guess for the geometry scan (or maybe do TDDFT). I don't know much about that.

After running the calculation, the energy values at each bond length will be printed out at the end of the .out file, and also as a .dat file. You can import the .dat file into Excel or other programs to get the actual visual plot.

Edit: If I am not mistaken, the $$\mathrm{U_A}$$ graph indicates the first excited triplet state, so converging to that UHF solution is easier, all you have to do is to set the multiplicity to 3 (*xyz 0 3). Using UHF/def2-SVP, the scan looks like this—

As you can see, the two curves are crossing whereas they shouldn't really cross. I am not entirely sure why this happens, but I suspect that using multiconfigurational methods will solve this. Also note that the energy plotted is the absolute SCF energy, so the graph's don't go to zero at high bond distances.

• "absolute SCF energy, so the graph's don't go to zero at high bond distances." Of course, graph's go to -0.5 Eh. "I suspect that using multiconfigurational methods will solve this" How can I do this in ORCA? Feb 6, 2021 at 19:15
• @Sergio These two guides sites.google.com/site/orcainputlibrary/cas-calculations/… , sites.google.com/site/orcainputlibrary/cas-calculations have a lot of information on setting up CASSCF on Orca. Feb 6, 2021 at 19:35
• If you need higher level methods (for very accurate energies) you can try CASPT2 (for that you have to look into the orca manual) Feb 6, 2021 at 19:36
• Where in that input file does it say that the atoms will be hydrogen atoms? I referenced your answer here: mattermodeling.stackexchange.com/a/4830/5. Apr 26, 2021 at 19:56
• @user1271772 I only posted the modified part of the input as the question had the full input. I have edited it to include the coordinates as well, so now it's a full input file. Apr 26, 2021 at 20:24