48
T. Yumura et al.
To elucidate mechanisms for the methane hydroxylation by sMMO as well as
pMMO, quantum chemistry calculations, especially density functional theory (DFT)
calculations have been performed by various research groups. In this chapter, we will
compare a variety of computational methodologies using different models of MMO
applied to elucidate mechanisms for the methane hydroxylation with an aim to clarify
key factors in determining the reactivity of MMO.
2 History of Computational Approach
To gain a better understanding of mechanisms for the methane hydroxylation by
sMMO, several groups employed quantum chemistry calculations [41–101], especially density functional theory (DFT) calculations. Advantage of the use of DFT
calculations is that one can track the direct methane hydroxylation by MMOH Q on
atomic-level. Employing DFT calculations to investigate the methane activation by
MMOH Q , one should be careful to select its model suitable to describe the iron(IV)
coordination environment, as well as the spin state suitable to describe magnetic
features obtained experimentally. Briefly, we summarize in Table 1 details of computational methods and diiron models (the coordination number of the dual iron cation,
and the spin state) in published papers. Table 1 shows that these computational studies
differ from one another in terms of selection of the diiron models and their spin states.
Pioneering papers were published at 1996 and 1997 by Yoshizawa et al. who
performed extended Hückel calculations to investigate the dioxygen activation and
conversion of methane to methanol by the diiron active site of MMOH Q [41–43]. As
shown in Fig. 3, Yoshizawa et al. found that a five-coordinate iron cation contributes
to activate a methane C–H bond, because it can allow to attractively interact with
methane by utilizing a nonbonding d orbital at its frontier orbital region. Then,
methane distortion from T d symmetry to C 3v or D 2d symmetries is important in the
attractive orbital interactions [43]. This orbital interaction results in the formation
of a direct Fe–C coupling, which facilitates the electron transfer from the highest
occupied molecular orbital (HOMO) of methane (C–H bonding character) to the
unoccupied nonbonding d orbital [46, 102]. This electron transfer induces to activate
a methane C–H bond. The findings obtained from extended Hückel calculations are
qualitative [41–43], however, they can provide fundamental insights on roles of the
orbital interactions in the activation of a methane C–H bond.
The publication by Yoshizawa et al. stimulated to quantitatively study the methane
to methanol conversion by MMOH Q by means of density functional theory (DFT)
calculations. Representative models used in the papers are displayed in Fig. 4. At
1997, DFT calculations with B3LYP functional (DFT/Hartree–Fock hybrid functional) [103–107] were carried out by Yoshizawa et al. [44, 46] and Siegbahn et al.
[57] who used diiron models, whose iron centers are coordinated by formate, water,
and OH groups. See models labeled as (a), (b), and (k) in Fig. 4. Unfortunately, these
models are too small to represent the coordination sphere of iron cations of sMMO.
These studies were followed by Morokuma group at 1999 [66] and Friesner group
T. Yumura et al.
To elucidate mechanisms for the methane hydroxylation by sMMO as well as
pMMO, quantum chemistry calculations, especially density functional theory (DFT)
calculations have been performed by various research groups. In this chapter, we will
compare a variety of computational methodologies using different models of MMO
applied to elucidate mechanisms for the methane hydroxylation with an aim to clarify
key factors in determining the reactivity of MMO.
2 History of Computational Approach
To gain a better understanding of mechanisms for the methane hydroxylation by
sMMO, several groups employed quantum chemistry calculations [41–101], especially density functional theory (DFT) calculations. Advantage of the use of DFT
calculations is that one can track the direct methane hydroxylation by MMOH Q on
atomic-level. Employing DFT calculations to investigate the methane activation by
MMOH Q , one should be careful to select its model suitable to describe the iron(IV)
coordination environment, as well as the spin state suitable to describe magnetic
features obtained experimentally. Briefly, we summarize in Table 1 details of computational methods and diiron models (the coordination number of the dual iron cation,
and the spin state) in published papers. Table 1 shows that these computational studies
differ from one another in terms of selection of the diiron models and their spin states.
Pioneering papers were published at 1996 and 1997 by Yoshizawa et al. who
performed extended Hückel calculations to investigate the dioxygen activation and
conversion of methane to methanol by the diiron active site of MMOH Q [41–43]. As
shown in Fig. 3, Yoshizawa et al. found that a five-coordinate iron cation contributes
to activate a methane C–H bond, because it can allow to attractively interact with
methane by utilizing a nonbonding d orbital at its frontier orbital region. Then,
methane distortion from T d symmetry to C 3v or D 2d symmetries is important in the
attractive orbital interactions [43]. This orbital interaction results in the formation
of a direct Fe–C coupling, which facilitates the electron transfer from the highest
occupied molecular orbital (HOMO) of methane (C–H bonding character) to the
unoccupied nonbonding d orbital [46, 102]. This electron transfer induces to activate
a methane C–H bond. The findings obtained from extended Hückel calculations are
qualitative [41–43], however, they can provide fundamental insights on roles of the
orbital interactions in the activation of a methane C–H bond.
The publication by Yoshizawa et al. stimulated to quantitatively study the methane
to methanol conversion by MMOH Q by means of density functional theory (DFT)
calculations. Representative models used in the papers are displayed in Fig. 4. At
1997, DFT calculations with B3LYP functional (DFT/Hartree–Fock hybrid functional) [103–107] were carried out by Yoshizawa et al. [44, 46] and Siegbahn et al.
[57] who used diiron models, whose iron centers are coordinated by formate, water,
and OH groups. See models labeled as (a), (b), and (k) in Fig. 4. Unfortunately, these
models are too small to represent the coordination sphere of iron cations of sMMO.
These studies were followed by Morokuma group at 1999 [66] and Friesner group