RASCI from CASSCF orbitals

Questions
#1

Dear Qchem Community,

I am trying to do a RASCI calculation starting from previously converged CASSCF orbitals because I wanted slightly better orbitals than the HF one. However in doing so the energy of the orbitals is completely wrong and I don’t know why. Here you have the input of casscf:

$molecule
0 1 
coord
$end
$rem
GUI              2
UNRESTRICTED FALSE
MAX_SCF_CYCLES   100
scf_algorithm    DIIS
EXCHANGE         hf
CAS_METHOD       2
CAS_M_S         0
CAS_N_ELEC      8
CAS_N_ORB       8
CAS_N_ROOTS     8
MEM_TOTAL       8000
MEM_STATIC      2000
CAS_SAVE_NAT_ORBS       TRUE
MAX_CASSCF_CYCLES       400
CAS_USE_RI              TRUE
AUX_BASIS               RIJK-def2-svp
BASIS            def2-svp
$end

And here the RASCI input:

$molecule
0 1 
coord
$end
$rem
GUI              2
SCF_GUESS        READ
SKIP_SCFMAN      true
scf_algorithm    DIIS
EXCHANGE         hf
CORRELATION      rasci
UNRESTRICTED     False
BASIS            def2-svp
!RAS-CI
RAS_ROOTS        10
RAS_ELEC         2
RAS_ACT          2
RAS_OCC          170   !SOMO is 171
N_FROZEN_CORE    155
N_FROZEN_VIRTUAL 485   !462
RAS_DO_HOLE      false
RAS_DO_PART      false
SET_ITER         200
RAS_SPIN_MULT    0
RAS_PRINT        2
!MEMORY
MEM_TOTAL        8000
MEM_STATIC       2000
MAX_CIS_CYCLES   200
$end

I don’t know if it is relevant, but I noticed that if I read the fchk files and I noticed that in the mo energies section the file print the occupation number insted:

Alpha Orbital Energies                     R   N=         869
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00  2.00000000E+00
  2.00000000E+00  2.00000000E+00  1.95860603E+00  1.95208081E+00  1.94303420E+00
  1.00010871E+00  1.00000015E+00  5.34494109E-02  5.16369993E-02  4.10836883E-02
  3.60318087E-14  2.73087508E-14  2.56748322E-14  2.26716960E-14  2.13774709E-14
  2.01754683E-14  1.90715559E-14  1.90083567E-14  1.78080141E-14  1.76899983E-14
  1.65075577E-14  1.52138273E-14  1.51445208E-14  1.49731500E-14  1.43664098E-14
  1.43362863E-14  1.42160373E-14  1.40762473E-14  1.38276447E-14  1.38018957E-14
  1.29914629E-14  1.29539434E-14  1.27447568E-14  1.23413523E-14  1.22142709E-14
  1.11967049E-14  1.09875103E-14  1.08851698E-14  1.06295839E-14  9.69886080E-15
  9.68039258E-15  9.42683461E-15  9.11515037E-15  8.93719177E-15  8.72030290E-15
  8.52217256E-15  8.28488145E-15  8.27397326E-15  7.99600221E-15  7.84066997E-15
  7.80690806E-15  7.75779988E-15  7.73774024E-15  7.29750129E-15  7.19279556E-15
  7.15939664E-15  7.12596775E-15  6.85950079E-15  6.79254110E-15  6.69310866E-15
....

I am pretty sure the input are loaded and saved in the right way because if I do the same using DFT orbitals I do not encounter similar error

Thank you for your help

#2

The ASCI (= CASSCF) code uses the orbital energy block for occupation numbers instead, since orbital energies aren’t well-defined anyway for a multideterminant wave function. In older versions of Q-Chem these are still labeled (erroneously) as orbital energies although that’s been updated in recent versions.