Magnetic circular dichroism and computational study of mononuclear and dinuclear iron(iv) complexes

The electronic structures of mononuclear and dinuclear iron(iv) complexes are studied using magnetic circular dichroism and wavefunction-based ab initio methods, and then correlated with their similar reactivities toward H- and O-atom transfer.

The effective transition dipole moment products obtained from the global eff parameter optimization yield fractional polarization factors as indicated in the insets.
Energies, values and Boltzmann populations of S = 1 magnetic sublevels as a 〈 〉,〈 〉 function of the applied magnetic field. Figure S4. The energies (a), the expectation value of (b) and the corresponding 〈 〉 Boltzmann populations (c and d) of S = 1 magnetic sublevels as a function of the applied magnetic field for a system with g x,y,z = 2.0, D = 28 cm -1 and E/D = 0. Figure S5. The energies (a), the expectation values of (b) and the corresponding 〈 〉 Boltzmann populations (c and d) of S = 1 magnetic sublevels as a function of the applied magnetic field for a system with g x,y,z = 2.0, D = 28 cm -1 and E/D = 0.
Derivation of the excited states arising from the 1b 2 →2b 1 transition of complex 1 In our CASSCF/NEVPT2 calculations, we found five low-lying excited states arising form the 1b 2 →2b 1 transition, for which the corresponding electron configurations are shown in Figure S6. In configurations A and B, one of the 2e orbitals is doubly occupied, the linear combinations of these two configurations yield the excited states of symmetry A 2 and B 2 , respectively.
The linear combinations of configurations C, D, E and F generate three triplet excited states of 3 B 1 and 3 A 1 (×2) symmetry, respectively, in addition to the M s = 1 component of the spin-flip excited state 5 A 1 .
On the basis of the above wave-functions, we can compute the energy differences among them as follows, is the average energy of the two 3 A 1 state, is the on-site Coulomb integral and  Figure S6. Low-lying electron configurations involves in the 1b 2 →2b 1 transition .
For the transition 3 E(2e→2b 1 ) , For the transition 3 E(1b 2 →2e) , Determination of the MCD C-term sign of the A 2 (1b 2 →2b 1 ) transition In the effective C 4v symmetry of complex 1, the 1b 2 →2b 1 transition results in the excited states of A 1 (×2), A 2 , B 1 and B 2 symmetry, respectively. Only the transition to A 2 is dipole-allowed and polarized along the z-direction. However, because the actual symmetry of complex 1 is C s , all five excited states are mixed and all transitions are symmetry-allowed. One excited state is computed to appear at 16440 cm -1 , which may contribute to the z-polarized transition around 17000 cm -1 .
Band 1 which is assigned as 3 A 2 (1b 2 →2b 1 ) is in fact a two-electron transition and hence has vanishing intensity. Lowing the symmetry from C 4v to C s allows band 1 to borrow intensity from intense transitions, which are close in energy. Thus, band 1 has no well-defined polarization property (x, y, z = 38%, 38%, 24%), as deduced from the VTVH analysis. The spin-orbit coupling (SOC) of 3 A 2 (1b 2 →2b 1 ) with E(2e→2b 1 ) provide a feasible source for the positive C-term signal of band 1 (see below). The SOC with other excited states, such as 3 E(1b 2 →2e), may also contribute to this positive C-term signal because of the complex polarization property of band 1.
However, due to the low intensity, the absolute polarization of the transitions 3 E(1b 2 →2e) cannot be determined unambiguously, which hampers the further analysis.
Using the same approach to determine the MCD C-term sign of E(2e→2b 1 ), we can show that the SOC between A 2 (1b 2 →2b 1 ) and E(2e2b 1 ) leads to the positive C-term MCD sign of band 1, if the transition energies of E(2e2b 1 ) are higher than that of A 2 (1b 2 →2b 1 ) as found experimentally.
If , Rotation of the d xy orbital to d yz along the y-direction results in a negative overlap, thus . Calculated MCD spectrum of complex 2 Figure S12. Computed MCD spectrum of complex 2.