Hello, 

Convergence problems are usually due to the initial geometry and the model chemistry; in your case it is QM (B3LYP/6-311++g(d,p)). This is especially the case in transition state optimizations. My suggestion is, first you must check your starting geometry for its correctness. It may be way different than the minimum (too distorted compared to the optimized geometry, incorrectly charged,or bonded etc.). Correctness of the initial geometry is essential. If you have an insight, you can manipulate your geometry manually before optimization. Second, you can try a pre-optimization - I usually do this at a nonhybrid functional, simply MMFF). Following to this you can tune your optimization at a certain level such as B3LYP/6-311++g(d,p). In fact your systems looks a bit large for QM. For such challenging optimizations another strategy can be starting small and working up to larger molecules and methods. In addition you can try different SCF keywords (http://gaussian.com/scf/). You can also check the below link:

https://www.researchgate.net/post/How_can_I_resolve_the_convergence_failure_error_in_Gaussian_09

Regards,

Serli

You can set the maximum number of optimization steps using the option MaxCycles=N. Alternately, you can restart the calculation from the last geometry (geom=check, guess=read). However, the main problem in your case appears to be that the energy landscape is flat and so perhaps Oleg's papers have some useful tips.

Thank you, B. R. Ramachandran , Serli Önlü for your useful comments. I am checking for the errors in my initial geometry as suggest by you. Oleg B. Gadzhiev sir I

Oleg B. Gadzhiev sir I have a flat potential energy surface, can you please provide any suggession how can i deal with it?

Probably this is due to convergence problems. There are many ways to deal with problems, such as starting from last geometry.

Also, you can try another two different ways:

1- Type in input the keyword:-

                                              SCF(maxcyc=number)

with this number bigger than 129. The other option is to try a quadratic convergence of the wavefunction using the keyword SCF(XQC).

2- Use a quadratic convergence of the wavefunction by typing the keyword:-

                                            SCF(XQC).

Furthermore, the basis sets: although they look fine, you need to try a different one.

All the best.

Regards,

 Khidhir

thank you Khidhir Alhameedi for your useful comment.

thank you, Oleg B. Gadzhiev. I will try this

Dear Keshav

If you are trying to optimize the reaction then the added keyword does not take care of that. It should be opt=ts (if one fragment) not just opt (which is for minimum). But, I do believe that you already know this, therefore, I think you meant the reactant or product complex of any reaction for which you need to get the minimum energy on PES (in that case, only opt is correct).

Now, for minimum case, you can do few things, e.g., maximize the no. of optimization cycles by opt=maxcycles=N. where N can be 300-400. So you can increase the cycle. Second, you should put the calculation of force constant to once. I also notice that you are not utilizing the full computer capacity, therefore, you can increase the processor number (for i3, total no. of cores/processors is 4) and RAM as well.

In addition, if the present procedure does not lead you to the desired output then take the final geometry of non-optimized structure and make that structure to initial guess and optimize again.

Let me know if it does not help you. 

I know that starting with HF or B3LYP/3-21g* gets me closer to the (local) optimum geometry before I start with the larger basis sets.  The faster running is HF/3-21g*, so that is what I use.  The other technique that I have used successfully has been mentioned above - take the geometry with the lowest energy from the log file (usually the last one) and use that as the input geometry to the next run.  Good luck, I know it is frustrating but you will get it if you already haven't.

Thank you shannon sir. I will do as you suggest 

Thank you,  Shannon Lieb I am trying your suggestions.

dear A. M Verma, I am trying to optimize the reaction complex not the transition structure. I have dual core processors only. also, PES for the reaction complex is nearly flat so increasing no of cycle don't work. I am trying to optimizing by providing force constant. I will tell if it works

Hi Keshav,

     If you are in a flat region of the PES, possibly relaxing the convergence criteria will help.  You can do that by using "opt=loose".  That worked for me when I was having the same problem.  Keep your fingers crossed.

Thank you, Shannon Lieb sir. I am trying this

Hallo, Keshaw!

    If your potential surface is very flat, you should also  increase the SCF accuracy e.g. to 10**(-9) by keywords SCF=Conver=9  or similarly.

   Best wishes

               mb

thank you Martin Breza