Hi, I am trying to do potential energy calculation of a triatomic molecule (e.g. carbon dioxide), using Gaussian.
I firstly calculated initial guess of molecular orbitals using HF method on 6-31G basis set.
Here is my code..(HF_Initial_Guess_code.txt).
I generated .chk file and a log file.
Next I used this .chk file to use CASSCF.
Here is my code..(co2_cassfcom.txt)
I took internuclear distance R starting from 1.8 Angstrom till 2.6 Angstrom, while the minimum of the potential is at 1.16 Angstrom.
The reason that I choose R well beyond the minimum is because at 1.16 Angs the calculation was not converging (Plot attached).
So my question is that can I use Scan and Opt keywords together? (I checked, but it did not work). Although I have seen examples where Optimization and Scan both are used, but Scan keyword is not used(R parameter is scanned with 'S' command). Since my task is to achieve Energy vs R in a tabular form (which can only be achieved using Scan), how can I get this done?
And is there any alternate way to get potential energy at each R value which is properly optimized?
Thanks in advance to those who look at such a long question and help me out.
Dear Kamal Kumar ,
To perform a scan I usually use the option : OPT=ADDREDUNDANT. After the coordinates and an empty line (for a bond), I put the following line:
B atom_1 atom_2 S number_of_steps steps
I don't know if this latter could help you but I hope it will give you ideas for your research !
According to https://www.cup.uni-muenchen.de/ch/compchem/geom/scan.html, you may use OPT=MODREDUNDANT or OPT=ADDREDUNDANT, depending on your system. It is also possible that you need to increase the number of cycles in order to achieve the optimization point, so you need to set a maximum number of cycles, like this:
#p opt=(modredundant,maxcycle=500)
I hope I've given you a satisfactory answer
Two commands can be used including OPT=MODREDUNDANT or OPT=ADDREDUNDANT . In addition, transition states can be localized in many cases through scanning one structural degree of freedom. This will, of course, only be successful if the reaction pathway can be described by essentially ONE structural parameter. This relaxed potential energy surface scan can be performed automatically in either a Z-Matrix or redundant internal coordinate system.
Thanks Corentin, Francisca and Majid for all your suggestions.
ModRedundant and MaxCycle really helped in my calculations (upto first excited state though). After first excited state there was again convergence failure. And, since I have to plot potential energy curves, I need a list of R and corresponding energy. Can I fetch these values from log file?
Dear Kamal, I'm not sure what are you referring to R, but I understand that you want to extract the reaction coordinate and energy from the log file. And yes, you can: you see the scan on Gaussview. Then, click on plots and save the data. After that, you need to extract it in Excel to do further calculations.
Dear Francisca,
You are right, R is the reaction co-ordinate. And since I do not have access to GaussView, is there any other way to extract those value from log file?
All I can think about the moment is that you need to write a script to extract the coordinate reaction and the energy.
Dear Kamal Kumar ,
I agree with Francisca J. Benítez , a script (awk for example and easy to do) would help you extract the information. Otherwise, you could do 'by hand' with a free software like Molden.
- Badges
- Science topic
- Similar topics
- Filing