Molcas Forum

jxzou

Member

Registered: 2020-09-11

Posts: 8

Is the MRCI method in OpenMolcas uncontracted or internally contracted

Hi, I apologize if my question is stupid, and appreciate for any helpful suggestions.

Which kind of/What type is the MRCI method/program in OpenMolcas? Is it uncontracted, full internally contracted (FIC, or PC), or strongly contracted?

I conducted a CASSCF(5,5)-MRCISD calculation for a molecule with identical basis set in various software packages, their CASSCF energies are identical, MRCISD energies are given below
-74.026517 a.u. (MRCISD in OpenMolcas, by using &MRCI)
-74.028027 a.u. (uncontracted MRCISD in OpenMolcas, by using &RASSCF CIonly)
-74.028027 a.u. (uncontracted MRCISD in ORCA)
-74.028028 a.u. (uncontracted MRCISD in Gaussian, by using RASCI)
-74.017809 a.u. (FIC-MRCISD in ORCA)
Davidson corrections are not included in all above energies. All core orbitals are not frozen.

So it seems that the MRCISD energy of OpenMolcas is higeher than that of uncontracted MRCISD, but lower than that of FIC-MRCISD.
The input section of MRCI in .input file are shown below. Please let me know if there is anything can be added or improved

...
&MOTRA
LUMORB
Frozen = 0

&GUGA
Spin = 2
Inactive = 5
Active = 5
NACTel = 5
CIAll = 1

&MRCI

Last edited by jxzou (2020-09-12 10:52:41)