I have a C, H, N molecule and it’s pH-respinsive. I’d like to compare the homo-lumo gap of it and its after protonated (with three charges). Firstly, I constract molecules in Avogadro and then put them in orca to optimize their geometry. And finally get their homo-lumo gap. However, the gap results were not match the optical experiment result.
In my optical experiment, the charged molecule was emit at 560nm and the uncharged molecule was around 575nm. However, the homo-lumo gap of charged molecule was lower than (1.52eV) uncharged (1.93eV). I think the result was not correct, because the emission wavelength of charged molecule was blueshifted, so the gap should be increased. Could you help me to figure out this problem?
I’m not sure this is the best forum for this, I might suggest the more general Matter Modeling Stack Exchange, along with giving as many specifics about your calculation and task as you can.
You don’t mention much about the Orca calculation, and you don’t mention, for example TD-DFT to get excitation energies, or potentially the need for solvent calculations. When you compare a charged species to a neutral species, the difference in the electrostatics can definitely change the orbital eigenvalues … it’s generally harder to remove an electron from a charged species and easier to add an electron, so the HOMO and LUMO values will change. Computing TD-DFT excitation energies with a solvent model can help to correct for that.
But you’re also comparing to an emission energy, which is not necessarily going to line up with excitation energies either. To properly compute emission energies, you need to optimize the excited state geometry (e.g. ORCA Input Library - TDDFT)
Last point is that the differences you want on the experimental side is really small compared to the general accuracy of TDDFT (e.g., ~0.2 eV). The issue you’re seeing is large, but I just want to warn that you may not get an exact result or the accuracy you want.