your questions is not clear. can you elaborate you question such as what type of structure you are trying to optimize what pseudopotential you are using. And most important thing what is the error that you are getting. it will be more helpful if you upload input and output files so that the error can be traced out

Thank you for your reply. I have trying for Pd(II) coordination complexes with organic Ligands with "b3lyp/genecp" program in Gaussian 09. I have attached the partial output file and attached it in the question. Please look at it and give your suggestion. 

Hi Sudh,

U can try with LanL2DZ basis set instead of B3LYP. I guess u will get good results. I tried few of my palladium and even some platinum complexes with this basis set and got good results.

Good Luck

yes i had seen that partial output file but that was not reflecting any error message that is why i asked you for input and output files

thanks

@deb Thank you for your suggestion... 

@Kurshid Ayub Sir, I am too unable to see the error from the out put file. That is the reason I am not proceeding further to rectify it. I am attaching the partial input file. Please help me out.

try scf=(fermi,novaracc) 

or 

Scf=(xqc,fermi,novaracc)

in the route line

and the let me know about the outcome. the latter one is more expensive computationally

if this fails also then send me the complete output file

thanks

Try switching on the magnetism. Pd is notorious for being difficult to converge since it sits on the verge on magnetic and non-magnetic behavior in the periodic table.

I don't know the size of your system, but in case you have many ligands, what I've been using and works fine is:

First  do a geometry optimization with a smaller basis set for all of your atoms or use ONIOM(HF:DFT) for example; B3LYP with your chosen basis set for ther metal and some of the atoms directly linked to it and HF with a smaller basis set for the ligands.

Then you could use the previous converged structure as initial geometry for the calculation at the desired level of theory, I think the number of scf cycles is excessive (5000) . I would use less cycles (1000) and maybe it could be useful if you decrease the convergence criterion (10^-5) to get a starting converged geometry.

thank you @Vaibhav Kaware for your suggestion. I am sure that Pd(II) is diamagnetic

@Ana E. Torres thank you for your suggestions  

@Khurshid Ayub Sir, I have submitted the files with your suggestions.  It may take 4-6 days to complete the program. Hope, it will end p with normal termination. Thank you once again for your inputs. 

Mr. Ganta,

Can you upload a whole inpute content of your complex or any other PdII-complex which computation causes for same problems

Metal complexes are tedious and challenging problem for electronic structure calculation. My question to you is what is spin multiplicity of the complex? Next question is whether the system has any kind of symmetry? Typically in case of open shell d-block elements is is unpaired electron and occupation number is the problematic part.

@Prakah Sir, The spin multiplicity of the palladium(II) complex is 1 (no unpaired electrons in the metal) and the complex generally possesses D4 symmetry. I have done for the other systems for Pd(II) with different ligand which ended up with normal termination. But this complex is not terminating smoothly.

@Bojidarka B. Ivanova madam, I will try to send input file of a similar complex.

Mr. Ganta,

The multiplicity is determined by the number of electrons but for the whole system, not only taking into account the metal ion. 

@Bojidarka B. Ivanova madam, I am sure that the multiplicity is 1.

Thank you everyone for suggestions. finally I got optimized structure with the keyword opt=rfo.