This is quite typical as by default Morokuma decomposition analysis
uses guess=Hcore. While it cannot use Huckel guess at present,
it still can use guess=moread. Thus, converge monomers first in a
separate runs (you need to use exactly the same coordinates as in
the dimer). Then, provide $VEC1 and $VEC2 groups for Morocuma run.
You can find more details in the manual.
Hope this helps.
On Sat Feb 18 '12 0:51am, manoj wrote
> I was trying to apply Morokuma Decomposition analysis onto one of my H-bonded transition metal complex which is a dimer. Unfortunately the SCF is not converging. The input file is attached.
> $CONTRL SCFTYP=RHF RUNTYP=MOROKUMA MAXIT=400 ICHARG=0
> IREST=-1 MULT=1 UNITS=angstrom EXETYP=RUN
> INTTYP=HONDO ICUT=11 $END
> $system TIMLIM=5970 MEMORY=50000000 $end
> $smp csmtx=.t. call64=1 $end
> $p2p p2p=.t. dlb=.t. $end
> $BASIS GBASIS=N31 NGAUSS=6 NDFUNC=1 NPFUNC=1 $END
> $GUESS GUESS=HUCKEL $END
> $SCF NCONV=6 maxdii=30 $END
> $MOROKM MOROKM=.TRUE. BSSE=.TRUE. IATM(1)=45 $END
>monomer corrin single point run