Dear Researchers,
I am trying to find a TS. It is a computational DFT calculation using Gaussian09. I am using the following input section:
%chk=tssb36.chk
%mem=1000MB
%nproc=4
# b3lyp/6-311+g(d,p) opt(calcfc,ts,noeigen) freq=noraman scrf(cpcm,solvent=ch3cn) scf=maxcyc=400
It optimizes the structure but the TS is not appropriate as frequency result. It gives a single imaginary frequency but with a very low magnitude as follows:
1 2 3
A A A
Frequencies -- -24.6042 15.4233 19.5119
Red. masses -- 5.1457 10.4436 11.1897
Frc consts -- 0.0018 0.0015 0.0025
IR Inten -- 11.6011 3.2653 0.6836
It is a phosphoryl (>HP=O) reaction center. How can I get the proper TS in this situation? Advance thanks for supportive information and helps.
It's been a while since I did such calculations but as far as I recall it is a three-step procedure:
1) You do a single point calculation for your presumed TS.
2) You perform the frequency calculation for that. That should ideally yield one large and some small imaginary.
3) You perform the actual optimization in which you specify with regard to which imaginary frequency it should be done.
It's been a while since I did such calculations but as far as I recall it is a three-step procedure:
1) You do a single point calculation for your presumed TS.
2) You perform the frequency calculation for that. That should ideally yield one large and some small imaginary.
3) You perform the actual optimization in which you specify with regard to which imaginary frequency it should be done.
Dear Jürgen Weippert , Thank you very much for your very kind answer. According to your process, I did all. Unfortunately, while I perform the original optimization, the result is deviated and not reach to the expected transition state. Is there any other suggestion to overcome the problem? I already tried several times. Is it possible to solve by fixing an entering bond using modredundant?
When you say it did not reach the expected transition state, did it reach a reasonable transition state with a single imaginary frequency or something that does not make sense at all?
There are some ideas:
1) You might try to use varying starting geometries. Especially in homolytic bond openings I remember that the calculation only converged to the right geometry if I was extremely close to the desired structure beforehand.
2) Gaussian provides the option to optimize using a quadratic synchronous transit (http://gaussian.com/qst2/) which means that you provide optimized starting and product geometries (QST2) and eventually also a TS geometry (QST3). This method creates quadratic paths through the phase space and tries to find the TS near the maximum of the path. In my calculations QST2 never worked, but QST3 often helped to find a TS that I couldn't reach in the conventional way.
Anyway, transition states can be beasts, so don't let them disencourage you.
Dear Jürgen Weippert , Thank you again for your kind prompt reply. According to your previous suggestions, in the second step, I found a pretty good only imaginary frequency. Then during optimization with it, it optimized but not properly. Just as before!
Thank you for more suggestions. I have no experience with QST2 or 3, But I am trying. I shall communicate again with you soon.
Thank you again for your help and suggestions.
By generating a starting structure for a TS optimization based upon the reactants and products that the TS connects, that this technique is known as a QST2 optimization.
If you have already optimized the reactants and the products, you can set up a QST2 calculation in the following way in Gaussview:
Open the optimized reactant, creating a new molecule group.
Open the optimized product, adding the molecule to the molecule group containing the reactant by selecting add all files to active molecule group from the target popup menu at the bottom of the open files so the molecule group will now contain two structures.
Use the Gaussian calculation setup to specify the TS optimization and set the optimize to a field to TS (QST2).
Also, the QST2 feature additionally requires that the atom numbering be consistent between the two structures. for example, the nitrogen atom has the same atom number in the reactants and in the product.
The job is now ready for submission and if you examine the input file for QST2 optimization, you find it has the following format:
(Note that,the input file contains two titles for QST2/ three titles for QST3 and two/three complete molecule specifications).
%chk=file.chk %nprocshared=n %mem=nGB
# M062X/6-31++G(d,p) Opt=QST2 Freq
Title card for reagents
Q M (First molecule specification)
Cartesian coordinates for reagents
—blank line—
and also for products
QST3 calculations are similar to QST2 calculations, but they require three molecular structures as input: the reactants, the products and a starting guess for the TS (the order of the first two does not matter, but the starting TS guess must be the third molecule specification). After selecting the TS (QST3) optimization type in the optimize to a field on the job type panel of the Gaussian calculation setup, you must also specify which structure is the transition state guess via the geometry to use for TS Guess field.
You can also use these links:
http://gaussian.com/opt/
- Badges
- Science topic