MCSCF calculation of a lithium triplet dimer (edited on April 7th)


Click here for full thread
Clicked A Few Times
Hi, all.

I'm trying to calculate a lithium triplet dimer using MCSCF module in NWChem 6.1.
It was supposed to reproduce the result of D. Danovich, et al, J. Am. Chem. Soc. 1999, 121, 3165-3174 (10.1021/ja982913n).
They included the orbitals 1sigmag 1sigmau 2sigmag 2sigmau 1piu 2piu 3sigmag 3sigmau 1pig 2pig in the active space for CASSCF.
I assumed that it is 1s2s2p orbitals of a lithium dimer, so I assigned ACTIVE=10 and ACTELEC=6 in my NWChem input.
But the calculation stopped with errors like below:

         %%%%% MacroIteration   1 %%%%%


          >>>| Valence CI:            -15.817898641011
0:0:pnga_create_config:ga_set_data:wrong dimension specified:: 0
(rank:0 hostname:node17 pid:10738):ARMCI DASSERT fail. ../../ga-5-1/armci/src/common/armci.c:ARMCI_Error():208 cond:0
Last System Error Message from Task 0:: No such file or directory


MPI_ABORT was invoked on rank 0 in communicator MPI COMMUNICATOR 4 DUP FROM 0
with errorcode 0.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.




mpirun has exited due to process rank 0 with PID 10738 on
node node17 exiting without calling "finalize". This may
have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).



Here's my input file.

start 5.0_dimer
echo

geometry noautosym units angstrom
    Li                        0.00000000     0.00000000     0.00000000
Li 0.00000000 0.00000000 5.00000000
END

memory 2000 mb

charge 0

BASIS spherical
 Li   library  def2-qzvpp
end

SCF
 UHF
TRIPLET
VECTORS output scf.movecs
END

TASK SCF ENERGY

MCSCF
 VECTORS input scf.movecs
ACTIVE 10
ACTELEC 6
MULTIPLICITY 3
END

TASK MCSCF ENERGY


Interestingly, the other MCSCF calculation of a lithium triplet dimer was successful where I set ACTIVE=10 and ACTELEC=2.
Is there any restriction to set active electrons, for example, in valence shells?
Thanks in advance.

+) ps.
I've found that the same error occurs when I calculate MCSCF(2,2) of triplet hydrogen dimer(H2).
I guess that this kind of error is due to no inactive space in core orbitals, but it is only my guess anyway.
Does anyone have clear idea or answer?