Molcas Forum

Fancy

Member

From: Beijng.China

Registered: 2016-07-09

Posts: 15

NO convergence after many iteraction in an opt encited state work

Hi, everyone!
Recently, in my work of optimized an excited state, the energy is stop going down after many iterations and became up and down, I tired to change the maxstep between 0.1~1.2, and added the "cartesian"， but it doesn't work, the input file of opt part as follows

***
>>> If ( Iter = 1 ) <<<
&SCF
Title
T5 minimum optimization
>>> EndIf <<<
&RASSCF &END
TITLE
 use scforb
 Spin
  3
 Nactel
 10 0 0
 Ras2
 10
 Inactive
 86
 Ciroot
 8 8 1
 LumOrb
 Thrs
 1.0e-05,1.0e-1,1.0e-1
 Rlxroot
 8
 &SLAPAF
 CARTesian
 MaxStep
 0.2
 Thrs
 1.0D-04 1.0D-03
 >>>EndDo
 End of Input

the out file of iteractiobn part is as follows

*****************************************************************************************************************
*                                  Energy Statistics for Geometry Optimization                                  *
*****************************************************************************************************************
                       Energy     Grad     Grad              Step                 Estimated   Geom     Hessian
Iter      Energy       Change     Norm     Max    Element    Max     Element     Final Energy Update Update Index
  1  -2456.64286263  0.00000000 0.164558 0.085694 lnm108  -0.138291* lnm052    -2456.65813704 RS-RFO  None    0
  2  -2456.64071580  0.00214683 0.057741-0.039212 lnm108   0.105954* lnm002    -2456.64255011 RS-RFO  BFGS    0
  3  -2456.64175571 -0.00103991 0.056955-0.039242 lnm108   0.083708* lnm008    -2456.64478258 RS-RFO  BFGS    0
  4  -2456.64183546 -0.00007975 0.025566 0.011637 lnm108  -0.115415* lnm003    -2456.64249118 RS-RFO  BFGS    0
  5  -2456.64269552 -0.00086006 0.325224-0.160733 lnm108   0.045132* lnm007    -2456.65195035 RS-RFO  BFGS    0
  6  -2456.64211963  0.00057588 0.064123 0.043176 lnm108  -0.168238* lnm007    -2456.65613605 RS-RFO  BFGS    0
  7  -2456.64233028 -0.00021065 0.025336 0.016589 lnm108   0.060718  lnm007    -2456.64253779 RS-RFO  BFGS    0
  8  -2456.64271300 -0.00038272 0.028035 0.018305 lnm108   0.022579  lnm007    -2456.64287164 RS-RFO  BFGS    0
  9  -2456.64284057 -0.00012757 0.028531 0.015050 lnm108  -0.045633  lnm001    -2456.64316714 RS-RFO  BFGS    0
 10  -2456.64269049  0.00015008 0.191511 0.088408 lnm108   0.033794  lnm012    -2456.64682760 RS-RFO  BFGS    1
 11  -2456.64209652  0.00059398 0.142232-0.097399 lnm108  -0.227502* lnm001    -2456.64514834 RS-RFO  BFGS    1

Could someone help me to solve this problem? I would apperciate it! Thank you again！

Last edited by Fancy (2018-10-31 10:15:40)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,201

Re: NO convergence after many iteraction in an opt encited state work

Same as in the other thread. With such a high root it's likely you have root flipping. Use the Track keyword and/or optimize an intersection between roots 8 and 7.