Available for INDO-based CI calculations only. The active space for double excitations includes n M.O.s. If only n is specified, the n M.O.s which ‘bracket’ the occupied-virtual energy levels are selected. (see C.I.=n for details). In C.I.D.=(n,m), n is the number of M.O.s in the active space, and m is the number of doubly filled (that is, not empty or partially filled) levels to be used.
By default, if CISD is specified, an active space of 10 M.O.s (or the same active space as single excitations, if this is smaller) is used.
To generate double excitations, all single excitations specified by C.I. are generated. For each single excitation, all double excitations within the active space generated by C.I.D. are generated. For example, if C.I.=6, one of the single excitations is (HOMO-2, LUMO+2). If C.I.D.=2, this single excitation will be used to generate the double excitation (HOMO-2, HOMO, LUMO, LUMO+2), where both M.O.s involved in the second excitation are within the smaller active space specified by C.I.D., even though the original single excitation involves orbitals outside the C.I.D. active space.
For a CISD calculation, the number of excitations that are generated is approximately:
(CI(m) * CI(n-m)) * (CID(m) * CID(n-m))
Even if the MAXCI keyword is used to limit the size of the CI matrix, generating and computing the energies of doubly excited electron configurations can become computationally expensive if a large active space for double excitations is used.