Alex Granovsky
gran@classic.chem.msu.su
you do not need numerical gradients for this job.
Please try the following input:
$CONTRL SCFTYP=MCSCF RUNTYP=optimize inttyp=hondo icut=12 d5=.true. nzvar=1 $END $SYSTEM TIMLIM=9999999 MWORDS=100 $END $moorth nostf=1 nozero=1 tole=0 tolz=0 $end $BASIS gbasis=ano-vdzp extfil=.t. $END $GUESS GUESS=MOREAD NORB=140 $END $trans altpar=.t. mptran=2 dirtrf=.t. mode=112 $end $DRT NMCC=22 NDOC=7 NVAL=3 GROUP=C1 FORS=.t. $END $GUGDIA NSTATE=4 ITERMX=100 cvgtol=1d-8 memmax=999999 $end $GUGDM2 cutoff=1d-15 WSTATE(1)=1,1,1,1 $end $GUGEM CUTOFF=1.0D-20 $END $MCSCF CISTEP=GUGA soscf=.t. ISTATE=2 acurcy=3d-8 ENGTOL=1.0d-12 maxit=150 $END $mcaver target=pure jstate=3 conic=2 multiw=.t. $end $statpt method=conic nstep=200 $end $conic shift0=1d-3 $end
Kind regards,
Alex Granovsky
On Wed Sep 25 '13 4:35pm, Antonio Carlos Borin wrote
----------------------------------------------------
>Dear users,
>I am trying to compute a conical intersection (firefly8) between the S1 and S2 states (GS, S1, S2, S3 states employed). That is, S1 is the 2nd state and S2 is the 3rd state.
>I´m using the input:
> $CONTRL SCFTYP=MCSCF MPLEVL=0 MAXIT=100
> RUNTYP=optimize inttyp=hondo icut=12 d5=.true. nzvar=1 numder=1 $END
> $d5 d5=.true. f7=.true. g9=.true. $end
> $zmat dlc=1 auto=1 $end
> $SYSTEM TIMLIM=9999999 MWORDS=100 nojac=1 kdiag=0 $END
> $trans altpar=.t. mptran=2 dirtrf=.t. mode=112 $end
> $moorth nostf=1 nozero=1 tole=0 tolz=0 $end
> $BASIS gbasis=ano-vdzp extfil=.t. $END
> $GUESS GUESS=MOREAD NORB=140 norder=0 $END
> $DRT NMCC=22 NDOC=7 NVAL=3 GROUP=C1 FORS=.t. $END
> $GUGDIA NSTATE=4 ITERMX=100 cvgtol=1d-8 memmax=999999 $end
> $GUGDM2 cutoff=1d-15 WSTATE(1)=1,1,1,1 $end
> $GUGEM CUTOFF=1.0D-20 $END
> $MCSCF CISTEP=GUGA soscf=.t. ISTATE=2
> acurcy=1d-7 ENGTOL=1.0d-12 maxit=150 CASHFT=0.25
> $END
> $XMCQDPT NSTATE=4 ISTSYM=1 INORB=0 edshft=0.02
> wstate(1)=1,1,1,1 kstate(1)=1,1,1,1
> thrgen=1d-20 threri=1d-20 genzro=1d-20 thrcon=1d-8
> $END
> $mcaver target=mixed penlty=2 xgrad=.f. $end
> $numgrd
> order=6 delta=0.01 istate=2 jstate=3 ngrads=2 spline=4 confit=1 praxes=1
> $end
> $STATPT NSTEP=400 method=qa opttol=3d-5 fmaxt=1000 $END
>It is not working...it works only with "istate=1 jstate=3" keywords in the $numgrd group.
>Any hint?
>Best
>Antonio