Dear Senthil,

Thank you for your answer. I'm trying like you told me and I hope it will work ;-)

I setup the job in Gaussian 03 like this:

#p opt b3lyp/6-31g geom=connectivity

With best wishes

Dear Senthil,

I tried two times but the system crash at link l502

How can I solve this problem please? Thanks in advance.

Many regards

Link 502 refers to solving SCF equations. Maybe your job hasn’t enough cycles to go. Gaussian’s default is 20. Use the keyword maxcycles=N using a bigger number, 100 or 200. Best luck

CRASHES AT LINK 502. They can be avoided by using this statement:

int=nofofcou

It generally occurs when Gaussian 03 is used ni more than processor. I always insert this statement in my G03 jobs. After this I had never a crash in that link. I hope this will be useful to you.

Dear Milagros and dear Juan-Sebastian,

I thank you for your help. I'm trying both strategies ;-)

I will let you know about the result.

Dear Milagros and dear Juan-Sebastian,

I performed the job but now the system crash at link l701.

Maybe I did something wrong...

I put in attached document both files containing the setup and the results after crash.

The second file

The last one.

Thanks ;-)

Hi, initially try to optimize the geometry of the large molecules using semi empirical (SE) methods such as AM1. And then submit the SE optimized geometry to higher method i.e HF/6-31g.. and B3lyp/6-31g. Depending on your molecule choose the suitable method and basis set.

In the keyword section mention opt=maxcycle=100 and scf=maxcycle=500.

Hope this will work.

Best wishes

Joshi LK

Are you doing full gopt? If yes the sugestion of Dr. Kumar is right. If you still have problems we may repeat here (with your permission of course) your calculation to see what is happening. Good luck.

I saw your log file. Try increasing the %mem to 100MW or 1GB if your computer allows you to. Your molecule is big enough to use all that memory.

I would like to make a geometry optimization in order to get the more favourable geometry of my molecule with the value of the corresponding dipole moment. I tried this job:

#p fopt b3lyp/6-31g int=nofofcou pop=full

but the system crashed after few hours (please see log file)

Now, I'm trying this job:

#p fopt=maxcycle=100 scf=maxcycle=500 b3lyp/6-31g int=nofofcou pop=full with 1GB of memory

I don't know if I made something wrong... If someone would like to try she/he can ;-)

I would like also to thank you for all interesting remarks and comments!

Thank you!

sfc=qc when optimization crashes. It consumes a lot of time more (I use G03).

The remaining instructions for full geom. opt. are

# opt b3lyp/6-31g(d,p) geom=connectivity int=nofofcou

I see from your log that you are using G98 for Windows and that the geometry optimization was carried out without any problem. Are you sure tat the size of hard disk (MaxDisk=20GB) is enough? I am thinking about the space for the integrals. The end of your log file does not show any abnormal finishing....

Dear all,

After one night of calculation, the Gaussian's job completed without any problem.

The job route is:

#p fopt=maxcycle=100 scf=maxcycle=500 b3lyp/6-31g int=nofofcou pop=full with %mem=1GB %nproc=1

Thanks all of you for your help ;-)

Best wishes,

Eric