Skip to content

Input File

Complete Input

! def2-SVP
!moread
%moinp "dae-rhf.gbw"
%pal nprocs 16 end

%casscf
nel 4
norb 4
nroots 3
mult 1
end

*xyzfile 0 1 dae.xyz

Line-by-Line

Line Meaning
! def2-SVP Basis set for the calculation.
!moread Tells ORCA to read in a starting set of molecular orbitals rather than generating an initial guess itself.
%moinp "dae-rhf.gbw" The specific orbital file to read, the converged RHF orbitals from the RHF tutorial.
%pal nprocs 16 end Runs the calculation in parallel across 16 cores.
%casscf ... end The CASSCF block, detailed below.
*xyzfile 0 1 dae.xyz Charge 0, multiplicity 1 (closed-shell singlet), geometry read from dae.xyz.

The %casscf Block

%casscf
nel 4
norb 4
nroots 3
mult 1
end
Keyword Value Meaning
nel 4 Number of active electrons.
norb 4 Number of active orbitals.
nroots 3 Number of electronic states to compute (state-averaged).
mult 1 Spin multiplicity of the states (singlet).

This defines a CAS(4,4) calculation over 3 singlet states, see The Active Space Concept for what choosing an active space actually means, and State-Averaged CASSCF for what computing multiple nroots at once implies for the orbital optimization.

Where the Geometry Came From

dae.xyz was not built by hand, it was converted directly from an already-optimized Gaussian geometry using Open Babel:

obabel -ig09 dae-opt.log -o xyz >> dae.xyz

This is the same conversion pattern used in the Electronic Structure Theory workflow, just with an xyz output format instead of a GAMESS input, Open Babel supports both from the same Gaussian log file.

Where the RHF Orbitals Came From

dae-rhf.gbw is not generated by this input, it's the converged orbital file produced by the separate RHF calculation covered in the RHF tutorial, read in here via %moinp. Reading in RHF orbitals rather than letting ORCA build its own initial guess gives the CASSCF orbital optimization a physically reasonable starting point, and is the same reason GUESS=MOREAD is used going from RHF into CIS in the GAMESS CIS tutorial.

Ask a Question

Have a question about quantum chemistry, or anything discussed on this site? Send it our way and we will reply by email.

Prefer to fill it out directly? Open the form.