[molpro-user] caspt2 convergence problem reg

Fri Dec 9 13:59:49 GMT 2011

Dear Molpro developers and subscribers,

I want to  calculate first 3 a' states of the b7 cluster geometry with cs point groupusing multi state caspt2 calculation. 
First I tried multi state - multi reference caspt2 calculations using the following input. ***,coupling constants
  memory,750,m
  basis = cc-pvdz
  geomtyp = xyz
  geometry = {7,
   testing
   b        0.570766   -0.003545    0.000000
   b        0.226880    1.761462    0.000000
   b       -0.251612    0.843642    1.225464
   b       -0.251612   -0.841773    1.230988
   b        0.208804   -1.761654    0.000000
   b       -0.251612   -0.841773   -1.230988
   b       -0.251612    0.843642   -1.225464
  }
 { rhf;occ,12,6;wf,36,1,0;}
 { multi;occ,15,8;closed,9,4;wf,35,1,1;state,3;}
 { rs2,mix=3;state,3}
  exit
The output is 
  ITER. STATE  ROOT     SQ.NORM     CORR.ENERGY   TOTAL ENERGY   ENERGY CHANGE       DEN1      VAR(S)    VAR(P)      TIME
    1     1     1     1.11458986     0.00000000  -172.53686711     0.00000000    -0.47173386  0.11D+00  0.15D+00  4086.88
    1     2     2     1.14040849     0.00000000  -172.45666411     0.00000000    -0.48674798  0.14D+00  0.15D+00  4086.88
    1     3     3     1.13323093     0.00000000  -172.45126929     0.00000000    -0.47685790  0.13D+00  0.15D+00  4086.88
 NO CONVERGENCE IN CIRS2II. MAXITI= 50  DE= 0.134D-17  VAR= 0.637D+01
    2     1     1     1.28565163    -0.49207255  -173.02893966    -0.49207255    -0.00268331  0.19D-02  0.92D-03  6293.54
    2     2     2   347.69521990    -6.60643600  -179.06310011    -6.60643600    -0.01546806  0.88D+00  0.11D-02  6293.54
    2     3     3  2852.28923595   -90.80603719  -263.25730648   -90.80603719    -0.00861259  0.19D+01  0.11D-02  6293.54
 NO CONVERGENCE IN CIRS2II. MAXITI= 50  DE= 0.000D+00  VAR= 0.376D+04
    3     1     1     1.29270395    -0.49631131  -173.03317842    -0.00423876    -0.00257874  0.71D-04  0.27D-04  8484.71
    3     2     2150405.19031346 -2413.03808658 -2585.49475069 -2406.43165058    -0.01942981  0.55D+01  0.11D-02  8484.71
    3     3     3************************************************************    -0.01654675  0.12D+03  0.22D-02  8484.71
 NO CONVERGENCE IN CIRS2II. MAXITI= 50  DE= 0.000D+00  VAR= 0.214D+07
    4     1     1     1.29410982    -0.49644548  -173.03331259    -0.00013418    -0.00003599  0.36D-05  0.14D-05 10684.36
    4     2     2************************************************************    -0.04655063  0.49D+02  0.16D-01 10684.36
    4     3     3************************************************************    -0.36395025  0.27D+04  0.20D+00 10684.36
 NO CONVERGENCE IN CIRS2II. MAXITI= 50  DE= 0.318D-17  VAR= 0.122D+10
    5     1     1     1.29441444    -0.49645203  -173.03331914    -0.00000654    -0.00008324  0.25D-06  0.87D-07 12941.86
    5     2     2************************************************************    -0.24838833  0.23D+03  0.11D+00 12941.86
    5     3     3************************************************************    -4.49998709  0.25D+05  0.21D+01 12941.86
 NO CONVERGENCE IN CIRS2II. MAXITI= 50  DE= 0.000D+00  VAR= 0.693D+12
    6     1     1     1.29447027    -0.49645243  -173.03331954    -0.00000041    -0.00000054  0.19D-07  0.68D-08 15164.80
    6     2     2************************************************************    -0.67145839  0.11D+04  0.33D+00 15164.80
    6     3     3************************************************************    -9.08453778  0.11D+06  0.38D+01 15164.80
 NO CONVERGENCE IN CIRS2II. MAXITI= 50  DE= 0.419D-17  VAR= 0.395D+15
and goes on to iteration 30, finally saying convergence problem. I read 
through the manual, want to check any intruder problem occurring. 
Even
 though fourth state is not energetically nearer to the third state and 
energy of neighboring molecular orbitals are not degenerate , I 
increased 
the no of states to mix to 4 and calculated the above 
states with single state - single reference caspt2 (as ms - mr - caspt2 
taking so much time)  with 
following input. 
  ***,coupling constants
  memory,750,m
  basis = aug-cc-pvdz
  geomtyp = xyz
  geometry = {7,
  testing
   b        0.570766   -0.003545    0.000000
   b        0.226880    1.761462    0.000000
   b       -0.251612    0.843642    1.225464
   b       -0.251612   -0.841773    1.230988
   b        0.208804   -1.761654    0.000000
   b       -0.251612   -0.841773   -1.230988
   b       -0.251612    0.843642   -1.225464
  }
{ rhf;occ,12,6;wf,35,1,1;}
 { multi;occ,15,8;closed,9,4;wf,35,1,1;state,4;}
 { rs2,mix=4,init;state,1,1;}
 { rs2,mix=4;state,1,2;}
 { rs2,mix=4;state,1,3;}
  exit
First two state calculations went fine but the last state {rs2,mix=4;state,1,3} is not converging instead it is diverging. 
Can you please suggest me how can I converge these energies. 
Thanks you very much, 
                                                                                                                                                                                                                                    Yours sincerely 

                                                                S.Rajagopala Reddy
                                                                                      Prof.Mahapatra lab
                                                                                      School of Chemistry
                                                                                  University of Hyderabad

                                                                S.Rajagopala Reddy
                                                                                      Prof.Mahapatra lab
                                                                                      School of Chemistry
                                                                                  University of Hyderabad