Hi Chethan,

There can be hundreds of different causes. No help is possible unless you provide more information, such as complete error message and your input settings.

Hi Chethan

you must use out file of hf/3-21g for input file by b3lyp/6-31g or another basis set. If your problem dissolve, you must change first guess about input structures.

Hi dear chethan

you can use the out file of hf/3-21g for input file b3lyp/6-31g or another basis sets. If your problem dissolved you must change first guess about ts structure.

If you have a converged hf/3-21g TS with frequencies, you can use opt=readfc geom=check to re-use both the geometry and force constants from the lower level calculation to begin the higher level one. I have also seen, however, that hf/3-21g can sometimes give TS's that don't exist at higher levels at all - that is, the extremely low level of theory does a poor job sometimes with even the qualitative nature of the potential energy surface. But the others are correct - give us more details and we can give better help -

Dear all,

A better way, to face this issue is to re-optimize by using check file from HF/3-21G calculation with the keyword: # opt=(ReadFC,ts,noeigen) freq geom=allcheck guess=read at your desired level of theory (B3LYP/6-31G). Because their is different between HF and DFT method, we should calculate the step by step with HF information. This is my option. Thanks