I am trying to do an opt+freq calculation with DFT using b3lyp/6-311g(2d,p) in order to calculate the HOMO/LUMO energies on the molecule shown below; a napthalene diimide (NDI) and two CH2-CH2-NH3 groups attached to the imide positions (first image). My issue is whenever I run the calculation, the R-NH3 deprotonates (second image), and the resultant MO energies are certainly incorrect. I want the R-NH3 to remain connected during the calculation. If I go to the redundant coordinate editor and create a build coordinate for the bonds between the nitrogen and 3 hydrogens with a set bond distance (third image), it still deprotonates. If instead I try to freeze the bond distances (fourth image), the bonds disappear (fifth image) and the MO energies are again well outside the range of what would be expected. Is there a better way to go about this?
Dear Daniel Powell ,
I think you must be giving wrong charge of the total system.
If its R-NH3, you have to allot +1 charge for this, as this is a quaternary type N (Nitrogen has one extra bond).
So, from the image you have posted you should allot total charge = 2 for your system. If you are calculating with total charge = 0 , it will optimize to give you C=C and NH3.
Regards.
Dear Daniel Powell ,
I think you must be giving wrong charge of the total system.
If its R-NH3, you have to allot +1 charge for this, as this is a quaternary type N (Nitrogen has one extra bond).
So, from the image you have posted you should allot total charge = 2 for your system. If you are calculating with total charge = 0 , it will optimize to give you C=C and NH3.
Regards.
Does R-NH3 really exists in the gas phase? You simulations is related to the gas phase at 0K. I assume that R-NH3 exists in a solution in this case water molecules are able to stabilize NH3 fragment. If it is correct you need to add H2O in your system.
As already said by Soumya Ranjan Dash first check the charge, and by Yuri Vladimirovich Barsukov check whether your molecule is stable in the gas phase. If you think these are not the origin of your problem: instead of redundant coordinates, you can can try to optimisations using Z-matrix to scan the N-H distance of the bond that breaks, to better understand what is going on. You could do that on a smaller model system... and check in the literature whether such H exchange has already been calculated for these chemical functions: if this exchange is real this is certainly already documented
- Badges
- Science topic