Doubts about reference weight in CASPT2 calculations

FedCpp · 2024-04-20 15:22:46

I'm performing a CASSCF/CASPT2 calculation on an aromatic system with 182 electrons (24 of which are pi-electrons) in the gas phase,
and I need to characterize its absorption spectrum.
I'm performing various tests considering different active spaces, using the default options (i.e., IPEA 0.25).
Unfortunately, I notice that the reference weight in the CASPT2 state is low.

For example, for a CASPT2 calculation on top of SA(8)/CASSCF(6,6)/6-31G**, I find:

Reference weight: 0.44286
Reference weight: 0.43781
Reference weight: 0.43479
Reference weight: 0.43581
Reference weight: 0.43422
Reference weight: 0.43414
Reference weight: 0.43441
Reference weight: 0.42728

For the root 1 for example:

 Computing the right-hand side (RHS) elements
 --------------------------------------------
  Using RHSALL2+ADDRHS algorithm

 Variance of |WF0>:      12.0088395650

 The contributions to the second order correlation energy in atomic units.
-----------------------------------------------------------------------------------------------------------------------------
  IT.      VJTU        VJTI        ATVX        AIVX        VJAI        BVAT        BJAT        BJAI        TOTAL       RNORM
-----------------------------------------------------------------------------------------------------------------------------
   1    -0.004192   -0.010342   -0.008266   -0.094946   -0.178485   -0.021610   -0.346153   -2.856568   -3.520562    0.003291
   2    -0.004225   -0.010393   -0.008309   -0.095218   -0.178656   -0.021659   -0.346322   -2.856804   -3.521587    0.000335
   3    -0.004224   -0.010395   -0.008308   -0.095231   -0.178664   -0.021661   -0.346330   -2.856809   -3.521622    0.000033
   4    -0.004224   -0.010395   -0.008308   -0.095230   -0.178662   -0.021661   -0.346328   -2.856807   -3.521613    0.000003

  Correlation energy /Case, /Symm, and sums:
 VJTU     -0.00422385 -0.00422385
 VJTIP    -0.00872812 -0.00872812
 VJTIM    -0.00166670 -0.00166670
 ATVX     -0.00830767 -0.00830767
 AIVX     -0.09522970 -0.09522970
 VJAIP    -0.08322757 -0.08322757
 VJAIM    -0.09543444 -0.09543444
 BVATP    -0.01850275 -0.01850275
 BVATM    -0.00315799 -0.00315799
 BJATP    -0.15413198 -0.15413198
 BJATM    -0.19219570 -0.19219570
 BJAIP    -1.78469886 -1.78469886
 BJAIM    -1.07210776 -1.07210776
 Summed:  -3.52161308 -3.52161308
-----------------------------------------------------------------------------------------------------------------------------

  FINAL CASPT2 RESULT:

      Reference energy:       -1034.1609110920
      E2 (Non-variational):      -3.5216130847
      Shift correction:          -0.0353390117
      E2 (Variational):          -3.5569520964
      Total energy:           -1037.7178631884
      Residual norm:              0.0000003521
      Reference weight:           0.44286

      Contributions to the CASPT2 correlation energy
      Active & Virtual Only:         -0.0299684087
      One Inactive Excited:          -0.4457812241
      Two Inactive Excited:          -3.0458634519


++ Denominators, etc.
--------------------------------------------------------------------------------------------------------------
 Report on small energy denominators, large coefficients, and large energy contributions.
  The ACTIVE-MIX index denotes linear combinations which gives ON expansion functions
  and makes H0 diagonal within type.
  DENOMINATOR: The (H0_ii - E0) value from the above-mentioned diagonal approximation.
  RHS VALUE  : Right-Hand Side of CASPT2 Eqs.
  COEFFICIENT: Multiplies each of the above ON terms in the first-order wave function.
 Thresholds used:
         Denominators: 0.3000
         Coefficients: 0.0250
 Energy contributions: 0.0050

CASE  SYMM ACTIVE-MIX  NON-ACTIVE INDICES          DENOMINATOR     RHS VALUE       COEFFICIENT     CONTRIBUTION
AIVX     1  Mu1.0002  In1.088 Se1.095               0.71452322     -0.02916399      0.03633980     -0.00105981
AIVX     1  Mu1.0010  In1.088 Se1.095               0.81134512      0.03985521     -0.04509153     -0.00179713
AIVX     1  Mu1.0016  In1.088 Se1.095               0.85614522     -0.02392402      0.02640851     -0.00063180
AIVX     1  Mu1.0018  In1.088 Se1.095               0.88448935     -0.02918559      0.03140910     -0.00091669
AIVX     1  Mu1.0012  In1.086 Se1.098               0.94402266     -0.03589174      0.03576130     -0.00128354
VJAIM    1  Mu1.0002  Se1.096 In1.087 In1.086       0.94221988      0.02720961     -0.02680031     -0.00072923
BVATP    1  Mu1.0003  Se1.096 Se1.096               0.86100626     -0.03107640      0.03313801     -0.00102981
BVATP    1  Mu1.0003  Se1.098 Se1.098               0.95657805     -0.02777542      0.02728972     -0.00075798
BJATM    1  Mu1.0002  In1.087 Se1.098 Se1.096       0.95707837     -0.05067670      0.04973331     -0.00252032
BJAIP    1  Se1.095 In1.085 Se1.095 In1.085         0.98684735      0.02735369     -0.02624597     -0.00071792
BJAIP    1  Se1.097 In1.085 Se1.097 In1.085         1.09552058      0.03456875     -0.03133476     -0.00108320
BJAIP    1  Se1.096 In1.087 Se1.096 In1.087         0.96121674      0.03380285     -0.03293962     -0.00111345
BJAIP    1  Se1.095 In1.088 Se1.095 In1.088         0.87855162      0.04606607     -0.04853729     -0.00223592
BJAIP    1  Se1.097 In1.088 Se1.097 In1.088         0.98722485      0.02703315     -0.02595438     -0.00070163
BJAIM    1  Se1.095 In1.085 Se1.096 In1.088         0.98703610      0.05394989     -0.05203567     -0.00280732
--
--

But the energy and character of the first two excited states I'm interested in are satisfactory with experiments.

The same happens if I increase the active space (SA(8)/CASSCF(8,8)), the Reference weight is below 0.5 for all roots.

Are there any suggestions?

Thanks to anyone who will help me.

Ignacio · 2024-04-22 08:00:20

https://molcasforum.univie.ac.at/viewtopic.php?id=860

Molcas Forum

Announcement

#1 2024-04-20 15:22:46

Doubts about reference weight in CASPT2 calculations

#2 2024-04-22 08:00:20

Re: Doubts about reference weight in CASPT2 calculations

Board footer