Firefly and PC GAMESS-related discussion club

 
Learn how to ask questions correctly  
 
 

 

Thomas Pijper
thomaspijper@hotmail.com

The reason for your convergence problems is a problem with your input, specifically the following lines:

 $CONTRL
   SCFTYP=RHF RUNTYP=OPTIMIZE DFTTYP=B3LYP1 MULT=1 ICHARG=0 COORD=ZMT UNITS=ANGS ECP=read
 $END

Any line in Firefly's input may not be longer than 80 or so characters. When you were performing the RHF calculation, the second line above was less than 80 characters long. However, when the DFTTYP option was added it became too long. As a result, Firefly did not see the "ECP=read" part and performed the calculation without the ECP (because the default is ECP=none).

As such, specifying these lines for example as:

 $CONTRL
   SCFTYP=RHF RUNTYP=OPTIMIZE DFTTYP=B3LYP1 MULT=1 ICHARG=0 COORD=ZMT 
   UNITS=ANGS ECP=read
 $END

fixes the problem.

Good luck!


Kind regards,
Thom


On Sun Dec 4 '16 3:59pm, Marko Mitic wrote
------------------------------------------
>Dear All,

>I'm new to Firefly. I'm trying to do optimization and frequency analysis of KBr molecule at level of B3LYP1 with Stuttgart ECP basis and their corresponding valence basis set. Diatomic KBr is just a start, we have in plan to study potassium bromide clusters.

>First, I have done RHF calculation (opt,freq) and everything is working great. I used MOs from RHF calculation for the second calculation, DFT one. The problem is that DFT optimization giving me an error, i.e. SCF not converged. Here is part of my input (I'm attaching the rest):

>!KBr molecule DFT opt and freq
> $CONTRL
>   SCFTYP=RHF RUNTYP=OPTIMIZE DFTTYP=B3LYP1 MULT=1 ICHARG=0 COORD=ZMT UNITS=ANGS ECP=read
> $END
> $CONTRL EXETYP=RUN INTTYP=HONDO ICUT=11 ITOL=30 D5=.T. NZVAR=1 MAXIT=100 $END
> $SYSTEM TIMLIM=300 MEMORY=30000000 $END
> $BASIS EXTFIL=.T. GBASIS=BAZA $END
> $GUESS GUESS=MOREAD NORB=27 $END
> $SCF DIRSCF=.T. $END
> $STATPT HSSEND=.T. NPRT=-2 NSTEP=200 $END
> $FORCE NVIB=2 VIBSIZ=0.005 $END
> $DATA
>potassium bromide, B3LYP1/9-ve PP(K),cc-pvtz-PP(Br)
>CNV 4

>K
>BR  1  rKBr

>rKBr=2.83
> $END

>The error I get at the end is:
>           ***** FAILURE TO LOCATE STATIONARY POINT, SCF HAS NOT CONVERGED
>     UPDATED HESSIAN, GEOMETRY, AND VECTORS WILL BE PUNCHED FOR RESTART

>I tried to change NORB, NSTEP, MAXIT, but didn't get any results. Also, I have tried with different functional (B3P86), but nothing.

>I have outputs from Gaussian 09 for this molecule, my friend do that for me. RHF from Firefly are the same as results from G09, DFT opt freq in G09 works fine.

>Please, if anyone can help me to solve this problem. Thanks in advance.

>I'm attaching files from RHF and DFT calculation.

>Best regards,
>Marko

[ This message was edited on Mon Dec 5 '16 at 11:24pm by the author ]