1 Introduction

Heme enzymes are versatile catalysts in nature involved in the biosynthesis and biodegradation of compounds in biosystems.1,2,3,4,5,6,7,8,9 As such, they function as peroxidases and catalases as well as mono-oxygenases. The latter group are labelled the cytochromes P450 and contain a heme-iron bound to the protein via a cysteinate group.10,11 These P450 mono-oxygenases are found in the liver, where they participate in the degradation and metabolism of a variety of compounds, including drugs and xenobiotics.12,13 In addition to this, they are also involved in the biosynthesis of hormones such as estrogen.14,15 The P450 mono-oxygenases undergo a catalytic cycle that uses one molecule of O2, two electrons and two protons to convert an iron(III)-heme into an iron(IV)-oxo heme cation radical species called Compound I (Cpd I).16,17 Cpd I is known to react with substrates through oxygen atom transfer often in the form of aliphatic and aromatic hydroxylation, epoxidation, desaturation and sulfoxidation reactions.18,19

Since P450 Cpd I is highly reactive and difficult to trap and characterize, many experimental efforts have been directed toward biomimetic model complexes that have the structural features of the oxidant in its first-coordination sphere, but lack the protein.20,21,22,23,24 In particular, changes to the heme scaffold and replacing it with the structurally similar porphyrin, corrole, corrolazine or phthalocyanine ligands often changes the properties of the active oxidant and its reactivity patterns.25,26,27 These studies have given insight into the chemical and functional properties of the heme group in heme enzymes and how the intricate design of the enzyme active site enables its high reactivity and selectivity.28

In this work, we focus on the structure and reactivity of iron-phthalocyanine complexes as shown in Scheme 1. Thus, experimental studies of Sorokin et al.,29 characterized a biomimetic Cpd I species with tetra-tert-butyl-phthalocyanine equatorial ligand using UV-visible, electron paramagnetic resonance (EPR) and mass spectrometry techniques. These studies identified the iron(IV)-oxo species as a doublet spin ground state with unpaired spin density on the phthalocyanine ligand coupled to an iron radical. In addition, the same group prepared and characterized an N-bridged diiron oxo tetra(tert-butyl)phthalocyanine cation radical complex from the heterolytic O−O bond cleavage of the anionic diiron phthalocyanine hydroperoxo complex.30 The intermediate was characterised by UV-visible, EPR, Mössbauer and X-ray absorption spectroscopies. This N-bridged diiron-oxo species was shown to be able to oxidize methane at ambient temperatures and react with benzene through oxygen-atom transfer.

Scheme 1
scheme 1

Iron-phthalocyanine complex [FeIV(O)(Pc+•)(Cl)] and its reactivity with para-X-substituted thioanisole substrates.

Full size image

In analogous studies, a high valent iron(IV)-oxo tetramesitylporphyrin (TMP) π-cation radical complex, i.e. a Cpd I mimic, was prepared and characterised by several spectroscopic techniques including UV-visible, NMR, EPR, Mössbauer, and resonance Raman spectroscopies.31 Furthermore, the effect of the porphyrin structure and the electronic structure on the reactivity of iron(IV)-oxo porphyrin π-cation radical species was studied. Electron withdrawing substituents such as tetra-2,6-dichlorophenyl and tetra-pentafluorophenyl at the meso position of the porphyrin ring were shown to make the oxidant more effective for oxygen atom transfer reactions.32 Subsequent studies of Fujii et al explored the iron(IV)-oxo porphyrin complex with substituents at the pyrrole β-position, i.e. 2,7,12,17-tetramethyl-3,8,13,18–tetramesitylporphyrin (TMTMP).33 Drastic changes in the electronic states and magnetic properties of the complex were observed upon changes in the substitution patterns at the β-positions of the pyrrole rings as compared to the previously reported substitutions at the meso-position. The oxo iron porphyrin π-cation radical complexes with electron-withdrawing groups were shown to be more reactive as compared to the electron-donating substituents. The effect of the axial ligand on the structure and reactivity of the oxo iron(IV) porphyrin π-cation radical complexes was examined.34 These studies were followed by a series of density functional theory (DFT) calculations on the effect of the axial ligands on the substrate sulfoxidation by an iron(IV)-oxo porphyrin cation radical species and showed that the barriers correlated with the ionization potential of the substrate and the electron affinity of the oxidant.35

Apart from studies on heme-containing iron complexes, there is a plethora of studies reported on biomimetic non-heme iron systems. In particular, these studies focused on the effects of metal-ligand distances,36 electronic and steric comparisons,37 axial ligation,38 isolobal analogues39 on the reactivity of iron(IV)-oxo metal complexes have been brought to the forefront over the years. As P450 enzymes typically catalyse substrate sulfoxidation reactions, often as part of a drug metabolism reactions,40,41,42 it is often the reaction of choice for model reactions. The effect of sulfoxidation of para-substituted thioanisole by a non-heme iron(IV)-oxo complex with N4Py (N,N-bis(2-pyridylmethyl)-N-bis(2-pyridyl) methylamine) ligand was investigated.43 Detailed computational studies on the sulfoxidation of sulphides by this iron(IV)-oxo complex was conducted by us and rationalized with valence bond schemes.44,45,46 Interestingly, a mechanistic switch from sulfoxidation of thioanisole by non-heme iron(IV)-oxo to simple oxygen-atom transfer to the metal ion-coupled electron transfer (MCET) was observed in the presence of Sc3+.47 Also, upon the addition of Sc3+, the rate of the sulfoxidation reaction increased by 100 fold. Similarly, the addition of triflic acid (HOTf) and Sc(OTf)3 during the sulfoxidation reaction rendered proton-coupled electron transfer (PCET) and MCET as the driving forces.48 In addition, studies with various tetradentate and pentadentate iron(IV)-oxo complexes highlighted major differences in reactivity and spectroscopic properties.49 To understand the equatorial ligand effect of iron(IV)-oxo species better, we did a detailed computational study on the biomimetic P450 Cpd I model with phthalocyanine equatorial ligand. The work shows that this system has an unusual electronic configuration that should weaken its oxidative properties; although, it should be able to react with sulphides via oxygen atom transfer.

2 Experimental

2.1 Model set-up

We created an iron(IV)-oxo models with phthalocyanine (Pc) ligand based on computational structures published previously.50,51,52 These models were validated against experiment and shown to be good mimics of the actual system. Model A is the bare system with all phthalocyanine substituents abbreviated by hydrogen atoms and with chloride as the axial ligand, i.e. [FeIV(O)(Pc+•)(Cl)]. This system is overall charge-neutral and was calculated in the doublet and quartet spin states. Chloride was chosen as the axial ligand ion as it is isoelectronic to sulphide in its bound form and therefore is a good mimic of the enzymatic system.53

Despite the fact that previous work showed little effect of substituents on the periphery of porphyrinoid systems,54,55 we tested a second and more elaborate model based on model A but with four t-Bu substituents on the periphery of the phthalocyanine ring (model B), i.e. [FeIV(O)(PcBu+•)(Cl)].

The reactivity of model A and B with the substrate was tested with various para-X-thioanisole (X = Me, Cl, CN, H, NO2, OMe) for substrate sulfoxidation.

2.2 Computational methods and design

The structure and reactivity of complexes A and B were studied with density functional theory methods as implemented in the Gaussian-09 software package.56 In general, the unrestricted hybrid density functional method B3LYP57,58 was used in combination with a LACVP basis set (with core potential) on iron and 6-31G* on the rest of the atoms (basis set BS1) for geometry optimizations of local minima and transition states, analytical frequency calculations and constraint geometry scans.59,60,61 To correct the energies, single-point calculations with the LACV3P+ basis set on iron (with core potential) and 6-311+G* on the rest of the atoms (basis set BS2) was performed. These single-point calculations also included the continuum polarized conductor model with a dielectric constant mimicking acetone. Free energies are reported at 298 K and 1 atm and used unscaled vibrational frequencies.

Initially models A and B were geometry optimized in the gas phase at UB3LYP/BS1, but some test calculations were done with geometry optimizations that include a solvent model: structures As and Bs.

The methods used in this work have been extensively used in our groups and shown to reproduce experimental free energies of activation to within 3 kcal mol−1.62,63,64,65 Moreover, the methods for analogous problems were shown to correctly reproduce experimental selectivities and product distributions.66,67,68,69

3 Results and Discussion

We started with a series of geometry optimizations on the iron(IV)-oxo phthalocyanine cation radical complexes [FeIV(O)(Pc+•)(Cl)] and [FeIV(O)(PcBu+•)(Cl)] with and without a solvent model included in the calculations. Optimized geometries of 4,2A, 4,2B, 4,2As and 4,2Bs are shown in Figure 1. As can be seen, enlarging the model with four tert-Bu substituents on the periphery of the phthalocyanine ligand has very little effect on the Fe−O, Fe−Cl and Fe−N bond lengths and all of those change by less than 0.005 Å. In all cases, the Fe−O distance is short: 1.64 Å in the gas phase and 1.63 Å with a solvent model included. These Fe−O distances match previous calculations on Cpd I of P450 and biomimetic iron(IV)-oxo porphyrin complexes.70,71,72,73,74,75,76,77,78,79,80 Moreover, experimentally observed Fe−O distances on biomimetic Cpd I model complexes (both heme and non-heme) were found to be ranging between 1.62 and 1.67 Å;81,82,83,84,85,86 and hence, are of similar magnitude to the results obtained here. In the case of cytochrome c peroxidase and cytochrome P450, the bond lengths are even further extended up to 1.7−1.9 Å and 1.65 Å, respectively, although the former may correspond to a reduced or protonated species instead. Similarly, the distances observed for the non-heme model systems containing ligand frameworks like TMC (TMC = 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane) and N4Py are 1.646 and 1.64 Å, respectively. It appears; therefore, that the iron(IV)-oxo phthalocyanine cation radical is structurally similar to analogous porphyrin complexes, which would not explain the reactivity differences between the two systems.

Figure 1
figure 1

Optimized geometries of iron(IV)-oxo phthalocyanine cation radical complexes in the gas phase (models A/B) and with a solvent model included (models As/Bs). Bond lengths are in angstroms.

Full size image

To understand the electronic properties of the iron(IV)-oxo reactant complexes, we analysed the molecular orbitals in detail. Figure 2 displays the high-lying valence orbitals of typically seen in A, B, As and Bs. These orbitals take the z-axis along the Fe−O bond and the x- and y-axis in the phthalocyanine plane and in between two Fe−N bonds. There are five orbitals with metal 3d contributions that are labelled according to the 3d orbital of iron that is involved. The lowest in energy and doubly occupied in all structures is the δxx-yy orbital in the plane of the phthalocyanine group and is nonbonding. Higher in energy and singly occupied in the reactant complexes are the π*-type interactions of the metal 3dxz and 3dyz orbitals with a 2p orbital on the oxo group in the same plane. In the reactant states the σ* orbitals are virtual and represent the interactions along the Cl−Fe−O axis (σ*zz) and in the plane of the phthalocyanine between the iron and nitrogen atoms (σ*xy). There are also a number of π*-type phthalocyanine-based orbitals and two of those are shown on the right-hand side of Figure 2, namely the a1u and a2u orbitals. In P450 Cpd I, the a2u orbital is singly occupied and is filled during the substrate activation reaction mechanism.16,17,18,19,87,88 Hence, the electron affinity of the heme is an important factor in P450 reactivity.

Figure 2
figure 2

Molecular valence orbitals of iron(IV)-oxo phthalocyanine cation radical complexes in the gas phase (models A/B) and with a solvent model includes (models As/Bs). Examples are shown of the orbitals of structure A.

Full size image

The orbitals shown in Figure 2 are occupied with seven electrons and give an electronic ground state for A, B, As and Bs of δxx-yy2π*xz1π*yz1 a2u2 a1u1, whereby the a1u electron is down-spin in the doublet spin state and up-spin in the quartet spin state. This electronic ground state is different from that found in P450 Cpd I and biomimetic iron(IV)-oxo porphyrin cation radical complexes that always give an a2u radical and no a1u radical.70,71,72,73,74,75,76,77,78,79,80 As a matter of fact, a DFT study on a small iron(IV)-oxo complex in the gas phase gave an energy gap between the states with a2u and a1u singly occupied of well over 20 kcal mol−1.70,73 We attempted to swap molecular orbitals for all structures but in all cases, the SCF converged back to a state with a singly occupied a1u orbital. As such, the electronic configuration with the a1u orbital singly occupied is the ground state and well lower in energy than the one with a2u singly occupied with a phthalocyanine ligand. Interestingly, for a manganese(V)-oxo corrolazine cation radical system, Goldberg et al also found the a1u-type orbital higher in energy and singly occupied with the a2u-type orbital doubly occupied.89 Therefore, replacing the meso-carbon atom in porphyrin with a meso-nitrogen atom as in phthalocyanines changes the π-type orbitals on the equatorial ligand and brings the a2u-type orbital below the a1u-type orbital in energy. This also must affect the redox potential of the equatorial ligand and consequently its reactivity with substrates.

Subsequently, we investigated the reactivity of complexes A and B towards oxygen atom transfer to sulphides and used para-X-substituted thioanisole as a model. All reactions are concerted with a single oxygen atom transfer to form sulfoxide products, which is similar to previous substrate sulfoxidation reactions reported by P450 and model systems.35,90 The quartet spin state structures are the lowest in energy and well lower than the corresponding doublet spin state structures. The ordering is similar to those of the product complexes that also have the quartet spin structures well below the doublet spin. As such we will focus on the quartet spin structures only. Optimized geometries of the quartet spin oxygen atom transfer transition states (TSSO,As,X) for model As are given in Figure 3. In all structures the Fe−O bond has elongated with respect to the reactant complex to values ranging from 1.799–1.813 Å and hence all structures fall in a narrow window. A bit more fluctuation is seen in the Fe−Cl interaction that is only 2.539 Å in 4TSSO,As,NO2, while it is somewhat longer in 4TSSO,As,Me at 2.670 Å. As the sulfoxidation process involves an electron transfer into the σ*zz molecular orbital that is located along the O−Fe−Cl bond, this means both Fe−O and Fe−Cl distances elongate.

Figure 3
figure 3

UB3LYP/BS1 optimized sulfoxidation transition states as obtained for model As. Bond lengths are in angstroms and the imaginary frequencies in cm−1.

Full size image

The sulfoxidation transition states are characterized with relatively large imaginary frequencies for the S−O stretch vibration with values that range from i554 cm−1 for 4TSSO,As,NO2 to i733 cm−1 for 4TSSO,As,H. These structures have an S−O interaction that follows the trend of the imaginary frequencies with a small distance of 1.928 Å for 4TSSO,As,NO2 and the longest distance for 4TSSO,As,OMe with a value of 2.039 Å. We also optimized the transition states for the large model Bs and those structures are given in Figure 4. Most distances are very similar between models As and Bs and so are the imaginary frequencies in the transition states. Clearly, the addition of tert-Bu groups to the periphery of the phthalocyanine ligand has little effect on the structure and reactivity with substrates.

Figure 4
figure 4

UB3LYP/BS1 optimized sulfoxidation transition states as obtained for model Bs. Bond lengths are in angstroms and the imaginary frequencies in cm−1.

Full size image

The energetics of the substrate sulfoxidation barriers of models As and Bs are given in Table 1 as well as the corresponding σ Hammett parameters for the para-substituent.91 In general, we find the same trends whether enthalpies (\(\Delta\)E+ZPE) or free energies (\(\Delta\)G) are used. Moreover, changing the model with four additional tert-Bu substituents on the periphery of the phthalocyanine ligand has a small effect and generally lifts the barriers systematically by a few kcal mol−1. The addition of entropy and thermal corrections to the enthalpies gives a change in the numbers of about 12 kcal mol−1 for all structures and systems are shown in Table 1. Although it should be mentioned that gas-phase DFT calculations often overestimate entropy contributions and sometimes by as much as 50%.46

Table 1 Relative energies and free energies of para-X-substituted thioanisole sulfoxidation by models As and Bs in the quartet spin state.
Full size table

Next, we analysed the substrate sulfoxidation transition state energies and plotted those for the reaction of As with para-X-substituted thioanisole (X = OCH3, CH3, H, Cl, CN, NO2) against the σ-Hammett parameter and show the trends in Figure 5. 78 As can be seen, whether relative energies of free energies are used, a reasonably linear correlation with the σ-Hammett parameter is obtained. Relative energies (at \(\Delta\)E+ZPE level of theory) range from 8.91 kcal mol−1 for 4TSSO,As,OMe to 13.42 kcal mol−1 for 4TSSO,As,NO2 and as such cover a range of 4.5 kcal mol−1. We also calculated the transition states for the large model with four additional tert-Bu groups on the periphery of the phthalocyanine ligand and its trend is shown on the right-hand-side of Figure 5. The trend is not as nice as for the bare system, probably due to interactions of the approaching substrate with the equatorial ligand system, but a small increasing trend is seen.

Figure 5
figure 5

Relative energies and free energies of sulfoxidation transition states for para-X-substituted thioanisole with phthalocyanine model As (left-hand-side) and Bs (right-hand-side) as a function of the σ-Hammett parameter.

Full size image

Overall, the DFT calculations on the reaction of As and Bs with para-X-substituted thioanisole give barriers that are close in energy and a small increase of the free energy of activation with a decrease in Hammett parameter. Whether the trends in Figure 5 mean that there is no effect of the para-substituent (flat curve) or there is a small increase cannot be determined at this stage as the margin may be within the error of the calculations. Therefore, experimental studies will be needed to be performed to confirm this trend. Nevertheless, it appears the effect of the para-substituent in the biomimetic Compound I models studied here may be considerably smaller than that seen in, e.g. nonheme iron(IV)-oxo oxidants. Thus, previous studies on substrate sulfoxidation by a nonheme iron(IV)-oxo species gave an increasing trend with increasing Hammett parameter in analogy with experimental studies.92,93 In nonheme iron chemistry, the sulfoxidation reaction is accompanied with double electron transfer into the iron system, whereas in our system here, only one electron moves to the iron while the second one moves to the equatorial ligand. Also, studies on para-X-substituted styrene epoxidation by an iron(IV)-oxo porphyrin cation radical gave linear trends of the free energy of activation against the substrate ionization potential.94 The latter studies used a minimal model in the gas phase without substituents on the periphery. It may very well be that interactions of the second coordination sphere affect the trends as studied here.

To understand the electron transfer processes in the substrate sulfoxidation processes in the doublet and quartet spin states, we show the orbital occupation levels for the reactant (4,2Re) and product (4,2Pr) complexes in Scheme 2. As discussed above in Figure 2 the reactants have electronic configuration δxx-yy2π*xz1π*yz1 a2u2 a1u1 with the a1u electron either with up-spin in the quartet spin state or down-spin in the doublet spin state. During the substrate sulfoxidation process, two electrons are transferred from substrate to oxidant. The first electron regardless of the overall spin-state will move into the a1u orbital and make it closed-shell. In the doublet spin state the second electron is transferred into the π*xz orbital and fill it with two electrons, whereas in the quartet spin state the electron moves into the virtual σ*zz orbital. Although in porphyrin systems, the iron(III) complexes, such as the resting state in the catalytic cycle of P450 is in a doublet spin ground state,5,6,7,8,9,17,19 the sulfoxidation reaction often takes place on a dominant quartet spin state surface.35,90,95 Our studies with the phthalocyanine model; therefore, follow spin-state ordering and reactivity patterns analogous to related systems studied previously.

Scheme 2
scheme 2

Orbital occupation changes from reactants to products in the quartet (top) and doublet (bottom) spin states.

Full size image

To understand whether the transition states have an electronic configuration closer to the reactants or products state, we analysed the molecular orbitals of the transition states and the group spin densities and charges. Figure 6 displays the spin densities of the 4TSSO,As,X sulfoxidation transition states. As can be seen the structures have a spin of 1.6 (4TSSO,As,NO2) or 1.8 (all other transition states) on the FeO group, which implies that all structures have almost two unpaired electrons along the Fe–O bond. In addition, all structures keep a spin of 0.9 on the phthalocyanine ligand. Based on the comparison of the spin density data of Figure 6 and the electron transfer pathways in Scheme 2, it, therefore, follows that the transition states are early on the potential energy landscape and have an electronic configuration close to the reactant state. Moreover, the spin densities show that the two-electron transfer process in the sulfoxidation reaction happens sequentially, whereby the first electron transfer happens in the transition state, while the second one en route from the transition state to the product complex. We were not able to find a local minima for a one-electron transfer reaction and the geometry scans gave a direct collapse to the product complexes. As such, the second electron transfer in the sulfoxidation reaction is fast and occurs rapidly as soon as the S−O bond is formed.

Figure 6
figure 6

Group spin densities (ρ) of UB3LYP/BS1 optimized sulfoxidation transition states as obtained for model As.

Full size image

Often the reactivity of metal-oxo complexes is correlated with thermochemical properties of the oxidant, such as the one-electron electron affinity (EA) of the iron(IV)-oxo species and the O−H bond dissociation energy (BDE) of the iron(III)-hydroxo species.35,80,88,96,97,98,99,100,101,102,103,104 Thus, the EA is defined as the adiabatic energy difference between [FeIV(O)(Pc+•)(Cl)] and its one-electron reduced structure, namely [FeIV(O)(Pc)(Cl)]−. Similarly, the energy of [FeIV(O)(Pc+•)(Cl)] to abstract a hydrogen atom to form [FeIV(OH)(Pc)(Cl)] is calculated from the energy difference of these structures. The thermodynamic cycle shown in Figure 7 also includes the gas-phase acidity of the iron-hydroxo complex, which is calculated from the EA and BDE values and the ionization energy (IEH) of an isolated hydrogen atom. The latter was taken from the experimental data from the NIST database.99

Figure 7
figure 7

Thermochemical cycle for the iron(IV)-oxo phthalocynine model investigated here. Energies are in kcal mol−1 and represent \(\Delta\)E+ZPE data. The value of IEH is taken from Ref99.

Full size image

Previously, it was shown that for substrate hydroxylation reactions by metal-oxo complexes with a rate-determining hydrogen atom abstraction step, the corresponding rate constant correlates with the C−H bond energy (BDECH) of the bond that is broken.96,97 In particular, experimental work for several models showed that the natural logarithm of the rate constant of a hydrogen atom abstraction reaction is linearly proportional to BDECH. Computational studies confirmed these trends and showed that for a range of oxidants with the same substrate, the rate constant also correlates with the O−H bond strength of the iron-hydroxo complex that is formed in the process.97 Furthermore, in substrate epoxidation and sulfoxidation reactions the formation of the rate-determining O−C and O−S bonds could be mimicked with an O−H bond and hence their rate constants also showed a correlation with BDEOH.35,98 Therefore, we evaluated the BDEOH value of [FeIV(O)(Pc+•)(Cl)] and compared its value with those obtained with analogous iron(IV)-oxo-porphyrin and nonheme iron(IV)-oxo complexes from the literature.

The calculated EA for [FeIV(O)(Pc+•)(Cl)] is 104.7 kcal mol−1, while its BDEOH is 70.0 kcal mol−1. From the cycle shown in Figure 7 using the experimental ionization energy of a hydrogen atom of IEH = 313.9 kcal mol−1, we calculated a gas-phase acidity for the reduced iron(IV)-oxo species of \(\Delta\)Gacid = 348.6 kcal mol−1. The BDEOH that was calculated for a P450 Cpd I model was 87.4 kcal mol-1, which is considerably larger than the value obtained for the phthalocyanine system. This is probably the result of the higher energy a2u orbital in P450 Cpd I that makes electron abstraction easier than in the phthalocyanine system and hence will create a stronger O−H bond formation step. As a consequence, the phthalocyanine system will be a weaker oxidant than an analogous porphyrin system and likely react with substrates with substantially higher reaction barriers. Indeed, for ethylphenylsulfide sulfoxidation by a P450 Cpd I model [FeIV(O)(Por)(Cl)] Kumar et al., found a barrier of only 7.9 kcal mol−1,35 while in this work all thioanisole sulfoxidation barriers by the phthalocyanine system are at least several kcal mol−1 higher in energy. Also a study on the sulfoxidation of para-X-substituted thioanisole by several nonheme iron(IV)-oxo complexes gave low barriers for all systems; however, those studies identified a small upward slope in the Hammett plot in agreement with experimental data. It appears that for the system studied here the Hammett plot has a much smaller slope. The porphyrin complex reported in reference35 had a much larger electron affinity of around 120 kcal mol−1 and hence will abstract electrons much easier than the phthalocyanine system discussed here. The amount of charge transfer in the transition states will affect the slope of the Hammett plots.

4 Conclusions

A computational study is presented on the properties and reactivity of an iron(IV)-oxo phthalocyanine system with chloride axial ligand. The work shows that in contrast to an analogous P450 Cpd I system the electronic ground state has three unpaired electrons in π*xz, π*yz and a1u molecular orbitals. The latter gives the [FeIV(O)(Pc+•)(Cl)] system a weaker electron affinity and also a lower BDEOH value. Subsequent studies on para-X-thioanisole sulfoxidation were done and showed these reactions to be concerted with a single oxygen-atom transfer leading to sulfoxides. Overall, the calculations predict that the iron(IV)-oxo phthalocyanine system should react with sulphides through oxygen-atom-transfer reactions. However, the calculated barriers are somewhat higher in energy than analogous nonheme iron(IV)-oxo and biomimetic Cpd I systems.