Molcas Forum

freshman1234

Member

Registered: 2017-02-23

Posts: 18

the error in structure optimiztion of excited states by molcas

This is my out file about the error:

 Convergence not reached yet!
 The predicted energy change is too large:   -30039.019249745212
 This can't be right!
 This job will be terminated.

And at the last of the out file,it says

--- Stop Module:  slapaf at Thu Feb 23 07:56:28 2017 /rc= _INTERNAL_ERROR_ ---
Non-zero return code - check program input/output
--- Module auto spent 6 days and 6 hours
Non-zero return code - check program input/output
--- Stop Module:  auto at Thu Feb 23 07:56:49 2017 /rc= _INTERNAL_ERROR_ ---
--- Module auto spent 1 week and 0 days
--- Stop Module:  auto at Thu Feb 23 07:56:49 2017 /rc= _INTERNAL_ERROR_ ---
--- Module auto spent 1 week and 0 days

My input file is as follow,

>>COPY /home /ex-lowThrs-4.RasOrb INPORB

>>> export MOLCAS_REDUCE_PRT=NO
>>>Do while
&SEWARD &END
Expert
Basis set
H    / inline
End of Basis
Low Cholesky
Douglas-Kroll
R02O
>>>If(Iter=1)<<<
 &SCF
>>>EndIf<<<
&RASSCF &END
TITLE
 use scforb
 Spin
   3
 Nactel
 26 0 2
 Ras1
 0
 Ras2
 14
 Ras3
 4
 Inactive
 80
 ChoLesky
LumOrb
 CiRoot
 10 10
 1 2 3 4 5 6 7 8 9 10
 1 1 1 1 6 0 0 0 0 0
 Thrs
 1.0e-05,1.0e-1,1.0e-1
Rlxroot
 5
 &SLAPAF
 Thrs
 1.0D-06 3.0D-04
>>>EndDo
&GRID_IT
 All; Ascii
End of Input

I appreciate the one that help me solve the problem.

Last edited by freshman1234 (2017-02-23 07:12:02)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: the error in structure optimiztion of excited states by molcas

What is that? No basis set?

That kind of CIRoot (non-equal weights) is usually asking for trouble too.

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: the error in structure optimiztion of excited states by molcas

I just noticed that you replied to this thread by clicking "Report" instead o "Quote" or using the reply box. This is what you wrote:

Because the basis set is too long,so I just leave it out.And if I set the CIRoot at equal weight, the excited state I want will disappear.

Have you checked that the root you want remains stable? As you try to optimize it, it will be stabilized, and it may become a lower root, and then everything's messed up.