This job illustrates how to calculate the energies of electronic states of formic acid resulting from one-electron excitation.
Data-set:
bonds First, optimize the geometry of formic acid O 0.0 0 0.0 0 0.0 0 0 0 0 C 1.20 1 0.0 0 0.0 0 0 0 0 O 1.32 1 116.8 1 0.0 0 2 1 0 H 0.98 1 123.9 1 0.0 0 3 2 1 H 1.11 1 127.3 1 180.0 0 2 1 3 cis c.i.=4 meci oldgeo Now calculate the excited states of formic acidThere are two calculations in this data set. In the first calculation, the geometry of formic acid is optimized. No keywords are needed for this operation, but because a blank line is not easy to see, the keyword BONDS has been added. BONDS is not necessary, but is used here only to show where the keyword line is.
The second calculation is the Configuration Interaction (C.I.) calculation. This uses the optimized geometry of formic acid from the previous calculation (using OLDGEO). Four C.I. keywords are used here:
CIS: Use only the ground state and those states that can be created by excitation of one electron.
C.I.=4: Use an active space of four molecular orbitals. Since this keyword is unqualified, these four M.O.s are the HOMO, HOMO-1, LUMO, and LUMO+1. That is, the two highest occupied and two lowest unoccupied M.O.s The C.I. thus involves four electrons.
MECI: Print out the microstates and the States.
Other keywords, such as SINGLET are unnecessary here, because no geometric operations are being done.
In the output, there are a total of nine states: one for the ground state and eight resulting from one-electron excitations. In most C.I. calculations, the ground state would be stabilized by C.I., but in this case Koopmans' theorem applies. This says that no state resulting from one-electron excitation from a ground state configuration will interact with the ground state. The ground state is included in the CIS calculation, to indicate its position in the energy spectrum.
The PMx-type methods in MOPAC are not accurate for predicting excitation energies, where possible use INDO instead. INDO is similar to ZINDO.