I started a job on gaussian to find the TS of the Cu oxidative addition. It takes a lot of time and I'm not even sure I will get to it. I would like to know if somebody have a general effective protocol to achive such task, like can I find the transition state threw multiple jobs by changing the redundant coordinate or by changing some keyword in the input? If so how do I choose those parameter? I put my input here so you can see what I did. I started from a known TS wich was from an iodophenyl instead of an iodoalkyne, so I just changed the phenyl group for an alkyne group. I started by fixing all the coordinates except those for the iodine and alkyne group.
Thank you very much.
Thanks for the answer, but what do you mean along the redundant coordinates? How do you choose them? I attached my input to my main question.
Chemically speaking, a standard protocol for 'getting' the TS for oxidative addition is by doing the following:
1. First optimize the final product, say it's LM(X)(Y); where X, Y correspond to the oxidatively added fragments (assuming of course that immediately after oxidative addition, both fragments are bound to the metal centre)
2. Scan the angle between X-M-Y and by doing an opt=redundant scan: Assuming you're using Gaussian/Gaussview you can do this by opening the optimized structure from Step 1, right clicking on Gaussview, selecting edit, selecting redundant coordinates. A new window (redundant coordinate editor) will open up. Push the 'add' button. The radio button next to it will be highlighted. Back in the Gaussview window, highlight X,M,Y. You will see that the atoms selected will show up in the 'redundant coordinate editor' window. From the 'unidentified' drop down menu in the editor window, select 'angle'. Below it, there should be another dropdown menu, select 'scan coordinate' here. The current value between X-M-Y will be visible, greyed out. Here's where a little intuition comes into play. Say the angle between X-M-Y is now 90deg, and as they reductively eliminate (the opposite of OA), the angle will decrease. So scan in steps of -10deg*, for lets say for up to 60 degrees; i.e. 5 steps. Save the file. *Note: minus 10 degrees because you're reducing the angle (RE, not OA).
3. At the top of the input file it should say "opt=modredundant" and at the bottom of the input file, you should see something like "A 3 1 2 S 5 -10.000000" A means angle, 3 1 2 are the atom indices, which in your case are X, M, Y. S means steps, so five steps, -10 is well, -10deg.
4. Run
5. Postrun, open the output file in Gaussview and right click and check scan results. Ideally the function of energy vs. reaction coordinate will show up with a maxima, which will correspond to a TS-like structure.
6. Save this structure, and edit the input file and run this geometry for a TS.
I just had a look at your input file!
It's always recommended to run a TS calc on a lower level first; so I'd use (at step 6 from above):
opt=(calfc,ts,noeigentest,maxcycles=200) M06/def2TZVPP geom=connectivity
Once you have a TS in gas phase, you can reopt in solvent. Don't do that in the beginning.
Also, calcall is a very expensive fail-safe. Use only when calcfc does not work.
Finally, noeigentest is almost always necessary, since your input file will have a few negative Eigen vectors in the initial freq test. With the noeigentest keyword, the Berny opt will follow (I think) the negative Eigen vector with the highest absolute value and 'iron-out' all the rest.
In an ideal scenario where your TS-like structure is very very close, you don't need noeigentest, but that is rarely the case.
Cheers!
- Badges
- Science topic