|Posted on Thursday, July 02, 2009 - 3:12 pm: |
I am doing a series of transition state (Gaussian QST3) studies. This involves modifying various reactant and/or product structures to create transition structure guesses. In some situations I want to change the bonding patterns so the Editor feature to right-click on a bond and change the bond-order is very useful...unless I want to get rid of a bond...that option doesn't exist. If I just delete bond and add back in the atoms the Z-matrix atom numbering has changed; I have to manually renumber atoms. Is there any possibility of adding a zero-bond order option to the right-click dialog. I realize that this is really just a display and the underlying quantum mechanics doesn't care about the display..but this feature would enable me to quickly create modified structures much more efficiently.
Post Number: 108
|Posted on Thursday, July 02, 2009 - 3:22 pm: |
By design, and verified by my testing just now, adding / deleting BONDS does not change the z-matrix. The purpose for this is exactly as you say: mainly for looking at transition states. Adding / deleting ATOMS does re-generate the z-matrix.
However, this behavior would generate non-optimial z-matrices. For example, if you draw all the atoms and THEN all the bonds, the z-matrix that would result would be "random". Thus the procedure I described in the first paragraph is true ONLY if you get the geometry via the 'New Job Using This Geometry' button, i.e. if it came from another job. This makes sense, as the usual workflow in doing a TS job is optimization of the reactant or TS, followed by IRC calculation or something.
|Posted on Sunday, April 04, 2010 - 4:42 pm: |
To correct a z-matrix that has inherited or created horrible cross-molecule bonds:
1. Draw the bonds you want.
2. Go into Adjust mode, and for each bond you want to remove, click it and do Edit, Cut.
3. Now do a single point (Energy) calculation at a low level, checking "Cartesian coordinates" box under the Advanced tab.
4. Open your molecule again, and the correct bonds will be shown, and the z-matrix will be consistent with it.