High-valent oxoiron porphyrins refer to iron complexes of porphyrin ligands in which the oxidation state of the iron ion is higher than þ3 and a terminal oxo group is bound to the iron center. The first high-valent oxoiron(IV) porphyrin-cation radical complex was prepared by the oxidation of Fe(TMP)Cl with m-CPBA in CH2Cl2CH3OH at 78C.31 The green species, formulated as [(TMP)ỵFeIVẳO]ỵ, has been well-characterized by various spectroscopic tech- niques such as UV–vis,31 NMR,31,69 EPR,69 Mo¨ssbauer,69,70 resonance Raman,69,71 magnetic circular dichroisom,72 and EXAFS spectroscopies.73 The [(TMP)ỵFeIVẳO]ỵcomplex was also prepared using other oxidants such as ozone,74dimethyldioxirane,75and iodosylbenzene.76
8.12.3.1 Oxoiron(IV) Porphyrin Cation Radical Complexes 8.12.3.1.1 Porphyrin ligand effects
The success of generating [(TMP)ỵFeIVẳO]ỵas a model of compounds I of heme-containing enzymes prompted studies on the electronic effect of porphyrin ligands on the reactivities and spectroscopic properties of high-valent oxoiron(IV) porphyrin complexes.77–79As electron-deficient iron(III) porphyrin complexes are better catalysts in catalytic oxygenation reactions, oxoiron(IV) porphyrin -cation radicals bearing electron-deficient porphyrin ligands show high reactivities in hydrocarbon oxygenations (Figure 5(b)).51,80 For example, while [(TPFPP)ỵFeIVẳO]ỵcomplex hydroxylates alkanes rapidly and efficiently, [(TMP)ỵFeIVẳO]ỵis notcapable of hydroxylating alkanes under the identical conditions.51 In addition, oxoiron(IV) porphyrin-cation radicals of electron-rich porphyrin ligands (e.g., [(TMP)ỵFeIVẳO]ỵ) reactfastwith hydroperoxides (Scheme 2, pathway A). On the other hand, oxoiron(IV) porphyrin -cation radicals of electron-deficient porphyrin ligands (e.g., [(TPFPP)ỵFeIVẳO]ỵ) react faster with olefins than with hydroperoxides (Scheme 2, pathway B). These results demonstrate that the electronic nature of porphyrin ligands affects the reactivities of oxoiron(IV) porphyrin-cation radicals significantly.51
The electronic effect of porphyrin ligands on the spectroscopic properties of oxoiron(IV) porphyrin -cation radicals has been investigated by placing electron-donating and electron- withdrawing substituents atmeso- and pyrrole-positions (Figure 5(b)).69,80–83The EPR spectra of oxoiron(IV) porphyrin -cation radicals containing substituents at meso-positions exhibit typical Sẳ3/2 EPR signals, indicating the strong ferromagnetic coupling between the ferryl iron and porphyrin radical spins.82 As the electron-withdrawing power of the meso-substituent increases, the energy of the a2u orbital would be stabilized relative to the a1u orbital via the interaction of the aryl and porphyrin -orbitals because of the large spin density at the meso- position in thea2uorbital; there is instead a node at themeso-positions in thea1uorbital (Figure 5a).80,84 When the meso-substituent is highly electron-withdrawing such as pentafluorophenyl, the EPR spectrum exhibits a broad signal around gẳ2,83 suggesting a weak ferromagnetic interaction between ferryl iron and porphyrin -cation radical spins. The result further suggests that the a2u orbital energy becomes lower and the HOMO of the porphyrin radical is switched from the a2u radical state to the a1u radical state (Figure 5(b)). The 1H NMR spectra of oxoiron(IV) porphyrin -cation radicals are sensitive to the electron-withdrawing power of meso-substituents.80The pyrrole proton signals shift upfield with the increase of the electronega- tivity of themeso-substituent, and the spectral change is interpreted with the mixing of thea2uand a1uorbitals due to the lowering the energy of thea2uorbital relative to thea1uorbital.80
The electronic effects of pyrrole -substituted porphyrins are different from those of meso- substituted porphyrins. The EPR spectra of oxoiron(IV) porphyrin -cation radicals containing substituents at pyrrole-positions show signals at g?3.1 and g||2.0,82 which are similar to those of compounds I of Micrococcus lysodeikticus catalase and ascorbate peroxidase.85,86 The
1H-NMR spectra show large downfield shifts of-methyl protons and small shifts of the meso- protons.80 The EPR and 1H-NMR spectral features indicate that the HOMO of the porphyrin radical is a1u orbital and there is a weak ferromagnetic interaction between ferryl iron and porphyrin radical spins.82In addition, in contrast to themeso-substituted porphyrins, the electro- nic structure of oxoiron(IV) porphyrin-cation radicals containing pyrrole-substituents is not altered much by the electron-withdrawing pyrrole -substituents because of the small spin densities of the a2u and a1u orbitals at the pyrrole -positions and the slight stabilization of both orbitals by the electron-withdrawing substituents (Figure 5b).
8.12.3.1.2 Axial ligand effects
Axial ligands of heme-containing enzymes, such as a cysteine thiolate in cytochromes P450 and chloroperoxidases, a histidine imidazole in peroxidases, and a tyrosine phenolate in catalases, are
N N
N N
Ar
Ar Ar
Ar FeIV
H3C
H3C
CH3 H3C
Cl
Cl
Cl
Cl
Cl Cl
F
F F F
F
N N
N N
FeIV Ar
Ar
Ar Ar
Magnetism a2u
a1u Orbital energy
Strong ferromagnetic Weak ferromagnetic
Magnetism a1u a2u Orbital energy
Weak ferromagnetic Weak ferromagnetic +ã
+ã
O
O
Ar =
a1u a2u
Redox potential of
iron(III) porphyrin Low High
(a)
(b) Reactivity Low High
Reactivity Low High
Figure 5 (a) Electron spin distribution of porphyrin atomic orbitals ofa1u(left) anda2u(right) symmetries.
Black and white circles represent signs of the upper lobe of the-AOs. (b) Electron-withdrawing effects of the meso- and pyrrole -substitued porphyrins on the reactivity and electronic structure of oxoiron(IV)
porphyrin-cation radicals (after Fujii).52
believed to play vital roles in generating and regulating the reactivities of high-valent oxoiron porphyrin intermediates.87,88 Among the heme-containing enzymes, the unique monooxygenase activity of cytochromes P450 has been attributed to the presence of a cysteine thiolate ligand (see Figure 1). Thus, a variety of thiolate-coordinated heme models have been synthesized and studied in elucidating the role(s) of the thiolate ligand.89 Early models were designed to mimic the physical properties of cytochromes P450 such as the appearance of a Soret band at 450 nm when iron(II) porphyrins bind to CO.90,91For an example, Collman and Groh synthesized iron(II) complexes of ‘‘mercaptan-tail’’ porphyrins (Figure 6a) that showed the resemblance of the spectral properties of the mercaptide-Fe(II)-CO to those of cytochrome P450 enzymes.92Another remarkable structural model was a picket fence iron porphyrin complex which showed a thermal stability of O2 binding.93 In the 1990s, Hirobe and co-workers synthesized an alkanethiolate- coordinated iron porphyrin complex (Swan Resting, SR complex), in which bulky groups on the RS coordination face were introduced to protect the thiolate ligation from oxidation (Figure 6b).94,95 The SR complex was highly stable and retained its axial thiolate coordination during catalytic oxidations. Furthermore, the reactivities of the SR complex were different to the Fe(TPP)Cl complex but similar to cytochromes P450. For example, peroxyphenylacetic acid was heterolytically cleaved by the SR complex (Equation (8)), whereas homolytic OO bond cleavage of peroxyphenylacetic acid was the predominant pathway when the catalyst was Fe(TPP)Cl (Equation (9)). The SR complex in alkane hydroxylation also showed a stronger hydrogen-abstraction ability than the chloride- or imidazole-ligated iron porphyrin complexes.
In 2001, Naruta and co-workers synthesized iron porphyrin complexes bearing an alkanethiolate axial ligand and hydroxyl groups inside molecular cavities and demonstrated that the oxy form of the heme models was stabilized by the hydrogen bonding between the bound oxygen and the hydroxyl groups inside molecular cavities:96
PhCH2CO3H SR complex PhCH2CO2H + compound I
benzene, 25°C ð8ị
PhCH2CO3H
PhCH3 + PhCH2OH + PhCHO [PhCH2COOã] + compound II Fe(TPP)Cl
benzene, 25°C
ð9ị
Compared to the many examples of iron(III) porphyrins with a thiolate axial ligand, the preparation of high-valent oxoiron porphyrins bearing a thiolate ligand has been less successful, probably, due to the easy oxidation of the thiolate ligand by oxidants. So far, only one oxo- iron(IV) porphyrin -cation radical complex bearing a thiolate axial ligand has been prepared as a model of chloroperoxidase (Figure 6c).97
R N
N N
N Fe
NH S
O
N
N N
N Fe
NH S
O O
N
N N
N Fe
O S
HN C O
O NH
C O HN
C O
R (R = NHCOC(CH3)3)
(a) (b) (c)
Figure 6 Structures of iron porphyrin complexes coordinating an axial thiolate ligand as models of cytochrome P450 enzymes.
Axial ligand effects on the reactivities and spectroscopic properties of oxoiron(IV) porphyrin -cation radicals have been demonstrated with iron porphyrins bearing simple anionic ligands.
Gross and co-workers reported that there is a pronounced axial ligand effect on the epoxidation of olefins by (TMP)ỵFeIVẳO(X) complexes with different anionic ligands (XẳF, MeOH, Cl, MeCO2, CF3SO3, and ClO4).74,98 All of the (TMP)ỵFeIVẳO(X) complexes except (TMP)ỵFeIVẳO(ClO4) epoxidize styrene to styrene oxide, and the rate of styrene epoxidation by the (TMP)ỵFeIVẳO(X) complexes is in the order of F>>MeOH>Cl>MeCO2>>CF3SO3
. Gross, Kincaid, and co-workers also reported the effect of axial ligands on the spectroscopic properties of (TMP)ỵFeIVẳO(X) complexes.98,99 The EPR spectra of the (TMP)ỵFeIVẳO(X) complexes show typicalSẳ3/2 signals irrespective of the axial ligands, whereas the 1H-NMR -pyrrole chemical shifts and resonance Raman v(FeẳO) bands are sensitive to the nature of the axial ligands. According to the spectral data of1H-NMR and resonance Raman, the (TMP)ỵFeIVẳO(X) complexes can be divided into two groups: one group with stronger electron donors such as F, Cl, and MeCO2 and the other with weaker electron donors such as MeOH, CF3SO3
, and ClO4
.98,99
8.12.3.2 Oxoiron(V) Porphyrins
Oxoiron(V) porphyrins, which are an isoelectronic form of oxoiron(IV) porphyrin -cation radicals, have been reported in:
(i) the reaction of [Fe(TDFPP)(F)2]andm-CPBA in the presence of excess fluoride salt at 25C;100
(ii) the addition of a small amount of MeOH to a CH2Cl2solution of [(TDCPP)ỵFeIVẳO]ỵ at 90C;101and
(iii) the reaction of [Fe(TDCPP)]þwith F5PhIO in the presence of a small amount of MeOH at90C.101
The color of the oxoiron(V) porphyrins is red, and the UV-vis spectra of the complexes exhibit bands at 420 nm (Soret) and 540 nm. These features are very similar to those of (Porp)FeIVẳO species. The (TDFPP)FeVẳO(F) is stable even at room temperature, whereas (TDCPP)FeVẳO(OCH3) is only stable at low temperature. The latter complex has shown a reactivity toward olefin epoxidation at 90C, butitreacts 10 times slower than [(TDCPP)ỵFeIVẳO]ỵ.101 Although the spectroscopic and chemical properties, including
1H NMR, EPR, magnetic susceptibility, iodometric titration, and the reactivity with organic sub- strates, of the red complexes are distinct from those of the corresponding oxoiron(IV) porphyrin and oxoiron(IV) porphyrin-cation radical complexes,100,101the exact oxidation state of the iron ion of the red species is ambiguous. Very recently, Dey and Ghosh carried out density functional theory calculations on the oxoiron(V) porphyrin species and concluded that there is no true oxoiron(V) porphyrin species and an axial ligand setconsisting of an oxide and a fluoride favors an oxoiron(IV) porphyrin-cation radical as the ground state.102
8.12.3.3 Oxoiron(IV) Porphyrins
Oxoiron(IV) porphyrins, one oxidizing equivalent above the resting ferric state, are known as compound II in the catalytic cycle of peroxidases and catalases. The (Porp)FeIVẳO complexes can be generated by: (i) the homolytic OO bond cleavage of (Porp)FeIII-O-O-FeIII(Porp), which is formed by the addition of dioxygen to iron(II) porphyrins in the presence of a nitrogen base, (ii) the chemical oxidation of iron(III) porphyrins by m-CPBA and PhIO under certain circum- stances, (iii) the electrochemical oxidation of hydroxoiron(III) porphyrins, and (iv) the reactions of iron(III) porphyrins with hydroperoxides (ROOH) in aqueous or organic solvents.103 More detailed preparation methods and physical properties of various oxoiron(IV) porphyrin com- plexes are summarized in recentreviews.77,79,103
It has been considered that the oxoiron(IV) porphyrins are such poor oxidants that they can only oxygenate triphenylphosphine to triphenylphosphine oxide.104 However, an olefin epoxida- tion by an oxoiron(IV) porphyrin complex was demonstrated by Groves and co-workers, in which (TMP)FeIVẳO, prepared by ligand metathesis of (TMP)ỵFeIII(ClO4)2 over basic alumina,
reacted with olefins readily at room temperature.105 As observed in oxomanganese(IV and V) porphyrin chemistry,61 the reactivity pattern of the (TMP)FeIVẳO complex was very different from that of [(TMP)ỵFeIVẳO]ỵ. A nearly equimolar mixture ofcis- and trans--methylstyrene oxide (cis/trans ratio of 1.2) was produced in the epoxidation of cis--methylstyrene by (TMP)FeIVẳO (Equation (10)), whereas the [(TMP)ỵFeIVẳO]ỵ complex afforded cis-oxide productpredominantly (cis/transratio of11) (Equation (11)). In addition, a concave (upward) Hammett plot was observed in the reactions of (TMP)FeIVẳO with a series of substituted styrenes, whereas a linear Hammett plot with a large negative slope (ẳ 1.9) was obtained with the [(TMP)ỵFeIVẳO]ỵcomplex.106Other reactivity studies with (Porp)FeIVẳO complexes, such as the porphyrin ligand effect (Section 8.12.3.1.1),51 have notyetbeen carried outin hydrocarbon oxygenation reactions:
CH3
O CH3
O CH3 +
Ratio of cis-oxide : trans-oxide = 1.2: 1 (TMP)FeIV=O
ð10ị
[(TMP)+•FeIV=O]+ CH3
O CH3
O CH3 +
Ratio of cis-oxide : trans-oxide = 11: 1
ð11ị
8.12.3.4 Mechanisms of Hydrocarbon Oxygenations
An early proposed mechanism of alkane hydroxylation by cytochrome P450 was a concerted insertion process in which an oxoiron(IV) porphyrin -cation radical transfers its oxygen to unactivated CH bonds via an insertion reaction (Scheme 7, concerted insertion pathway).107 In the late 1970s, the currently most accepted ‘‘oxygen rebound’’ mechanism was enunciated by Groves and co-workers.108 In the ‘‘oxygen rebound’’ mechanism (Scheme 7, cage-radical path- way), the oxoiron(IV) porphyrin -cation radical abstracts a hydrogen atom from the substrate (RH) to form a carbon radical and an FeIV-OH complex in a cage so that they cannot diffuse away from each other. These two species then recombine in the rebound step to produce the hydroxylated product and the resting state of the heme iron.109,110 The ‘‘oxygen rebound’’
FeIV +.
O
C H FeIV
O C H
FeIV OH
• C
cage
FeIII C OH concerted
insertion
cage-radical (oxygen rebound)
+
rebound
Scheme 7
mechanism was supported by numerous experimental observations such as high kinetic isotope effects,108,111–115
partial positional or stereochemical scrambling,116–118 and allylic rearrange- ments119,120 in hydroxylation reactions by cytochromes P450 and iron porphyrin models. The formation of alkyl bromide in alkane hydroxylations by synthetic iron(III) porphyrins and oxidants in the presence of CCl3Br was another piece of supporting evidence for the generation of a carbon radical in the ‘‘oxygen rebound’’ mechanism.109,110
The advent of ‘‘radical clock’’ substrates121,122in alkane hydroxylations by cytochromes P450 provided strong evidence for the involvement of carbon radicals and allowed the estimation of the rate of the oxygen rebound step. If a carbon radical is produced by a hydrogen atom abstraction of a ‘‘radical clock’’ substrate by an oxoiron(IV) porphyrin-cation radical, the carbon radical either rebounds to the iron(IV)-OH complex to give an unrearranged alcohol product or rear- ranges to another ring-opened carbon radical which then rebounds to the iron(IV)-OH complex to yield the rearranged alcohol product (Scheme 8).123 The ratio of unrearranged/rearranged alcohol products allows the calculation of the rate of oxygen rebound (kOH) with the known rearrangement rate of the carbon radical (kr). When Ortiz de Montellano and Stearns used bicyclo[2.1.0]pentane (8) as a radical clock in cytochrome P450-catalyzed hydroxylation reaction, a 7:1 mixture of unrearranged alcohol (9), and rearranged alcohol (10) was obtained (Scheme 8).123 With the ratio and the rate of bicyclo[2.1.0]-2-yl radical rearrangement of 2109s1 (kr), the rate of oxygen rebound (kOH) was calculated to be 1.41010s1. Very recently, Ortiz de Montellano, Groves, and co-workers reinvestigated the oxidation of norcarane ((11), krẳ2108s1) by four different cytochrome P450 enzymes and observed the formation of radical- and cation-derived rearrangement products (Equation (12)).124By calculating the relative amounts of unrearranged product (12) to radical-derived rearranged product (13) the oxygen rebound rate (kOH) was calculated to be21010s1, a number very similar to that obtained in the oxidation of bicyclo[2.1.0]pentane.123 The authors further suggested that the cation-derived product(14) was formed via an electron-transfer process:
(Porp)+ãFeIV=O (Porp)FeIV-OH (Porp)FeIV-OH
kOH
HO
(Porp)FeIII (Porp)FeIII
OH kr
(8)
(9) (10)
Unrearranged Rearranged
•
•
Scheme 8
OH
OH
OH
+ +
(12) (13) (14)
(11)
Unrearranged (Radical-derived)(Cation-derived) P450
ð12ị
Despite the accumulated evidence for the ‘‘oxygen rebound’’ mechanism, recent results using more sophisticated and ultrafast ‘‘radical clocks’’ have challenged the ‘‘oxygen rebound’’ mechan- ism and revived the concerted insertion mechanism. Atkinson and Ingold used several radical clocks including phenyl-substituted cyclopropanes in P450-catalyzed hydroxylation reactions.125
Interestingly, the computed rates of the oxygen rebound step varied by nearly three orders of magnitude, depending on the ‘‘radical clock’’ substrates. In addition, there was no correlation between the amounts of rearranged products and the rate constants for radical rearrange- ments.126,129 The latter phenomenon was explained with the generation of cationic species by an iron(III)-hydroperoxo intermediate (see Section 8.12.4.2).129,130 A striking result of using hypersensitive radical probes such as (15) (Scheme 9), (trans,trans-2-tert-butoxy-3-phenylcyclo- propyl) methane, and methylcubane was the incredibly fast oxygen rebound rate (kOHẳ 1.41013s1).126–128 Thus, the lifetime of the radical species is in the range of 80–200 fs, and such an extremely short lifetime implies that the radical is not a true intermediate but a component of a reacting ensemble (or transition structure). Accordingly, a concerted nonsynchronous hydroxylation mechanism was invoked in which a ‘‘side-on’’ approach of an oxygen atom to the CH bond was suggested as opposed to the linear CHO array of a conventional abstraction mechanism.126–129
CH3
H
CH2OH
H OH
+ other products +
(15) (16) (17)
Unrearranged Rearranged
CH2 • H
kr = 3 x 1011 s–1 • P450
Product ratio of (16) to (17) = 48 ± 13
Scheme 9
Recent developments in theoretical calculations of compound I models have shed light on the structures and reactivities of oxoiron(IV) porphyrin -cation radicals.131–134 Notably, Shaik, Schwarz, and co-workers proposed a two-state reactivity (TSR) in CH bond activation by a model for compound I of cytochromes P450.134,135As shown schematically in Figure 7, alkane hydroxylation proceeds by a two-state rebound mechanism of two closely lying spin states, the high-spin quartet state (4A2u) and low-spin doublet state (2A2u), of an oxoiron(IV) porphyrin -cation radical. The CH bond activation by the two spin states (4A2u and 2A2u) involves hydrogen abstraction-like transition states (4TSHand2TSH) followed by the formation of carbon radical and iron(IV)-hydroxide intermediate clusters (4CI and 2CI). The low-spin intermediate cluster (2CI) collapses to the (Porp)Fe-alcohol product in a virtually barrierless fashion, making the low-spin process a nonsynchronous reaction with no true intermediate. On the other hand, the high-spin intermediate cluster (4CI) of iron(IV)-hydroxide and the alkyl radical undergoes rebound with a distinct transition state (4TSReb) and a significant barrier, resulting in the existence of a true radical intermediate. Thus, the TSR scenario provides a reasonable reconciliation of the apparent contradiction between the oxygen rebound and concerted insertion mechanisms.133–137
Mechanisms for olefin epoxidation by oxoiron(IV) porphyrin -cation radicals are not as well understood as those for alkane hydroxylations, and no major experimental advances have been made in the past decade. Possible intermediates that have been proposed in olefin epoxidations are depicted inScheme 10: (i) a metallaoxetane (a), (ii) a carbon radical (b), (iii) a carbocation (c), (iv) an alkene-derived-cation radical (d), and (v) a concerted oxygen insertion (e).2,138 Mostof the mechanisms were proposed on the assumption that oxoiron(IV) porphyrin -cation radicals were generated as epoxidizing intermediates in catalytic epoxidation reactions, but recent studies provided evidence that multiple oxidants can be involved in olefin epoxidations (Section 8.12.4.2).
So more experimental work, especially with in situ generated oxoiron(IV) porphyrin -cation radicals,139is necessary to clarify the mechanisms of olefin epoxidations.