Oxo Species - Wiley Online Library

0 downloads 0 Views 2MB Size Report
[(LNHC)FeIV(O)(MeCN)]2+ (species 1) and its corresponding spin density plot. b). Optimized structure of the ground state of the [(porphyrin)FeIVO(SH)]- (species.
A Journal of

Accepted Article Title: Axial vs. Equatorial Ligand Rivalry in Controlling the Reactivity of Iron(IV)-Oxo Species: Single-State vs. Two-State Reactivity Authors: Gopalan Rajaraman, Ravi Kumar, and Azaj Ansari This manuscript has been accepted after peer review and appears as an Accepted Article online prior to editing, proofing, and formal publication of the final Version of Record (VoR). This work is currently citable by using the Digital Object Identifier (DOI) given below. The VoR will be published online in Early View as soon as possible and may be different to this Accepted Article as a result of editing. Readers should obtain the VoR from the journal website shown below when it is published to ensure accuracy of information. The authors are responsible for the content of this Accepted Article. To be cited as: Chem. Eur. J. 10.1002/chem.201800380 Link to VoR: http://dx.doi.org/10.1002/chem.201800380

Supported by

10.1002/chem.201800380

Chemistry - A European Journal

FULL PAPER Axial vs. Equatorial Ligand Rivalry in Controlling the Reactivity of Iron(IV)-Oxo Species: Single-State vs. Two-State Reactivity

Abstract: High-valent iron-oxo species are known for its very high reactivity and this aspect has been studied in detail over the years. Particularly the role of axial ligands in fine-tuning the reactivity of the iron(IV)-oxo species are studied in detail. The role of equatorial ligands in fine-tuning the reactivity of such species are rarely explored and is of prime importance in the development of nonheme chemistry. Here, we have undertaken a detailed DFT calculations on [(LNHC)FeIV (O)(CH3CN)]2+ (1) species NHC (L =3,9,14,20-tetraaza-1,6,12,17-tetraazoniapenta-cyclohexane cosane-1(23),4,6(26),10,12(25),15,17(24),21-octaene) in comparison to compound II of cytochrome P450 [(Porphyrin)FeIV(O)(SH)]- (2) to probe this aspect. The electronic structures of 1 and 2 are found to vary significantly and this has led to a large variation in the reactivity. Particularly, strong equatorial ligand present in 1 destabilize the quintet states significantly as compared to species 2. To fully understand the reactivity pattern of this species, we have modelled the hydroxylation of methane by species 1 and 2. Our calculations reveal that species 1 reacts via low-lying S=1,  pathway and generally available S=2, σ pathway is not energetically accessible. In addition to possessing a significant barrier for C-H bond activation, the -OH rebound step is also computed to have a large barrier height leading to a marked difference in reactivity between these two species. Of particular relevance here is the observation of pure triplet state reactivity for species 1. Besides we have also attempted to test the role of axial ligands in fine-tuning the reactivity of species 1 and our results demonstrate that in contrast to heme systems, the axial ligands in 1 do not significantly influence the reactivity. This highlights the importance of designing the equatorial ligands to fine-tune reactivity of high-valent iron(IV)-oxo species.

aliphatic and aromatic hydrocarbons.[1] In heme iron enzymes such as cytochrome P450, iron(IV)-oxo cation radical porphyrin compound (Cpd I) and iron(IV)-oxo porphyrin compound (Cpd II) species are suggested as the [2] reactive intermediates. Between the two, Cpd I has biological significance and it has been explored in detail over the years. Cpd II is also found to activate C-H bonds of alkanes.[2i, 3] To understand their reactivity and also to find a way to fine-tune the reactivity, several model complexes are synthesized in the laboratory to mimic the reactivity of the metalloenzymes.[1, 2c, 4] A number of efforts have been made to synthesis and characterize heme and non-heme iron model complexes .[1a, 2] Their reactivities toward C-H bond activation have been studied by means of experiments and theory. Among others, one possible way to fine-tune the reactivity is to alter the nature of the a xial ligands. This has been proposed and studied in detail for cytochrome P450 enzymes. In cytochrome P450 and related model complexes , [2a, 5] weaker axial ligands are found to decrease the reactivity of the species towards Hydrogen Atom Transfer (HAT) reactions. For example –SH axial ligand shows higher reactivity as compared to imidazole axial ligand. [2a, 5]

Introduction High-valent iron-oxo species are key intermediates in the catalytic cycles of heme and non-heme iron enzymes that insert an oxygen atom from dioxygen into inert C-H bonds of

[a]

Mr. Rav i Kumar, Prof . Dr. Gopalan Rajaraman Department of Chemistry Indian Institute of Technology Bombay, Powai Mumbai, India E-mail: [email protected] Dr. Azaj Ansari, Department of Chemistry , Central Univ ersity of Hary ana, India 123031.

**



GR would like to acknowledge the f inancial support f rom the Gov ernment of India through Department of Science and Technolo gy (EMR/2014/000247) and gen erous computational resources f rom NPSF. AA also would like to acknowledge the f inancial support from DST-SERB (ECR/2016/001111).RK thanks CSIR f or a f ellowship. Both the authors contributes equally . Supporting inf ormation f or this article is av ailable on www under http://dx.doi.org/10.1002/anie.xxxxx

Scheme 1. The hy droxy lation process by tetracarbene and cytochrome P450 cpd-II iron(IV)-oxo reagents with v ariable axial ligands: 1 NHC IV 2+ NHC IV + ([(L )Fe (O)(CH 3CN)] ), 1a ([(L )Fe (O)(SH)] ), 1b NHC IV 2+ NHC IV 2+ ([(L )Fe (O)(N(CH 3)3 )] ), 1c ([(L )Fe (O)(N(CH 3)3)] ), 1d NHC IV 2+ IV ([(L )Fe (O)] ), 2 ([(Porphy rin)Fe (O)(SH)] ) and 2a IV ([(Porphy rin)Fe (O)(CH 3CN)]).

Models reported to mimic non-heme enzymes are based on aminopyridine ligands which offer weak to moderate ligand fields. While the role of axial ligand in bio-mimic non-heme iron(IV)-oxo models have been tested,[6] how substantially stronger equatorial ligand alters the reactivity has not been established? This is an attractive idea and has been explored lately. Particularly, carbene based organometallic

This article is protected by copyright. All rights reserved.

Accepted Manuscript

Ravi Kumar†, Azaj Ansari† and Gopalan Rajaraman* [a]

10.1002/chem.201800380

Chemistry - A European Journal

catalysts are synthesized and characterized.[7] These catalysts are also found to perform C-H bond activation, epoxidation, and N-H insertion.[7-8] In this direction, efforts have been made to fine tune the equatorial ligands ,[9] particularly synthesis and reactivity of iron(IV)-oxo complex possessing macrocyclic tetracarbene ligand ([(L NHC)FeIV(O)(CH 3CN)]2+ (1) where L NHC=3,9,14,20tetraaza-1,6,12,17-tetraazoniapenta-cyclohexa cosane1(23),4,6(26),10,12 (25),15,17(24),21-octaene) gained attention.[10] In complex 1, the carbon atoms of the carbenes are found to coordinate on the equatorial positions leading to a stable iron(IV)-oxo species. Here we have undertaken a detailed theoretical study to answer the following intriguing questions : (i) why species 1 exhibits high stability compared to the other iron(IV)-oxo species reported? (ii) how comparable are the electronic structure and reactivity of species 1 and 2? (iii) can the axial ligand be used to fine-tune the reactivity in 1? (Scheme 1)

Computational Details In the present work, all the calculations were performed using established procedures.[5c, 11] Geometry optimization is carried out with Gaussian 09 software,[12] whereas all the spectroscopic parameters are calculated with the ORCA 3.0 [13] program package incorporating COSMO solvation effects. The geometries were optimized using the Grimme’s dispersion corrected unrestricted B3LYP functional (UB3LYP-D2).[14] We carried out optimizations and frequency calculations with LANL2DZ double ζ-quality basis set with the Los Alamos effective core potential for Fe and a 6 -31G basis set for C, H, O, N, P, S; [15] then perform single-point [11e, 16] energy calculations using a TZVP basis set for all the atoms. To ascertain the role of basis sets, geometry optimizations are also carried with LANL2TZ basis set for Fe and 6-31G* basis set for other atoms. The structural parameters and the computed energetics are found only marginally altered by this higher basis set (see Table S5 and S6 in ESI). Frequency calculations were performed on the optimized structures to verify that they are minima on the potential energy surface (PES) and also to obtain free energy corrections. The quoted DFT energies are UB3LYPD2/TZVP solvation energy including free-energy corrections at the temperature of 298.15 K unless otherwise mentioned. Since most the species studied in this work contain +2 charges, the gas phase optimizations at lower basis set are prone to self-interaction errors (SIE). To avoid this issue, we have also performed solvent phase optimizations for selected species, however, only minor alterations in the values are noted and these are compared in Table S3 in ESI. Further, the optimized geometries were verified by animating frequency using chemcraft software.[17] The solvation energies were computed at the UB3LYP-D2 level using polarizable continuum model (PCM) with acetonitrile as a solvent. Spin density visualizations are done using chemcraft [18] software. Calculation of all spectroscopic parameters incorporates a relativistic effect via a zeroth-order regular approximation method (ZORA) as implemented in ORCA

suite.[19] The Mössbauer isomer shift () was calculated based on the calibration constants reported by Römelt et al. and 0.16 barn was used for the calculation of quadruple moment of 57Fe nuclei.[20] We have performed the state average complete active space self-consistent field (SACASSCF) calculations to compute the zero field splitting (ZFS) parameters of species 1. Here, dynamic correlations are incorporated using N-electron valence perturbation theory (NEVPT2) method. We employed the def2-TZVP basis set for this calculation. The active space for CASSCF calculations comprises five Fe IV-based orbitals with four electrons in them (d 4 system; CAS(4,5) setup). We considered five quintets excited states, thirty-five triplets excited states and twenty-two singlet excited states in our calculations. The zero-field splitting is then extracted using effective Hamiltonian approach implemented in ORCA program. Here we have employed the following notation to specify particular species, for example, 1--3TS1 denote, TS1 corresponding to species 1 in the triplet surface for the  reaction channel.

Results Electronic Structure of species 1 and 2 To answer the first and the second questions, we have computed the electronic structures of species 1 and 2. Calculations yield S=1 as the ground state for 1, where the Fe-O bond length is estimated to be 1.659 Å (see Figure 1a and Table S1 of ESI). This and other structural parameters are in accordance with the X-ray structure reported.[10] The Fe-C distances are estimated to be 1.986 Å and 2.043 Å and these are shorter and stronger than the equatorial Fe -N distances computed for 2. The electronic configuration for the ground state (S=1) of the species 1 is computed to be (xy)2(* xz)1(*yz) 1(*z2) 0(* x2- y2) 0 (see Figure 2a) and a strong  o verlap between O-px/y orbitals and Fe-d xz/Fe-dyz orbitals are detected here. Besides , as the carbene ligands are non0.927 O C4

Fe

C1

C3

1.103

C2 N

(a) 0.960

O N4

Fe

N1

N3 1.132 N2

S

(b) Figure 1. B3LY P-D2 a) optimized structure of the ground state of the NHC IV 2+ [(L )Fe (O)( MeCN)] (species 1) and its corresponding spin density plot. b) IV Optimized structure of the ground state of the [(porphy rin)Fe O(SH)] (species 2) and its corresponding spin density plot.

This article is protected by copyright. All rights reserved.

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

6.77 6.26

*xz

σ*x2-y2

6.84

σ*z2

σ*z

6.57

σ*x2-y2

O

xy

xy

0.75

1.26

1.22

1.03

 xy

*yz

*xz

*yz

1.00

z

2

xy

0.0

0.80

0.0

x

Fe (a)

(b)

y Figure 2. Eigenv alue plot f or the ground state (S=1) of the (a) species 1 and (b) species 2 (in eV).

planar, the  bonding orbitals of the carbenes are found to have bonding interactions with the *xz and * yz orbitals leading to a reduction in xy and *xz /yz gap (0.25 e V, see Figure 2a). Unlike the conventional iron(IV)-oxo species, very strong equatorial ligation of the tetracarbene ligand 2 2 destabilize the * x - y orbital significantly in 1. To a certain extent, this also destabilizes the *z2 orbital. Lower *z2 energy also implies that the Fe-C equatorial ligand field is much stronger than the axial Fe-O and Fe-N bonds. As both these orbitals are destabilized, this creates a large energy penalty for the S=2 high spin state, placing this at 93.1 kJ/mol higher in energy compared to the S=1 state. Such a large energy gap (one of the largest energy gap known) reduces the reactivity and contributes to the stability of species 1.[10] Besides the optimized structure shows that each ethylene forms a weak C-H...O interaction with the ferryl o xygen atom and this brings additional stability for the species 1 (see Figure 1a). This stabilization can be compared to the tripodal iron(IV)-oxo reported possessing additional stability due to three N-H...O hydrogen bonding interactions.[21] To ascertain quantitatively the strength of this interaction, we have performed NBO second-order perturbation theory donor-acceptor analysis which place this interaction energy as large as 8 kJ/mol. Spin density at the iron center for the ground state of species 1 is computed to be 1.103 (see Figure 1a). A significant spin density is also computed at the ferryl o xygen atom (0.927) in species 1 suggesting oxyl radical character (see Table S2 in ESI).[2a, 5a] The spin density on the oxygen atom of 2 is estimated to be

0.960 which is larger than that observed for the 1 (see Figure 1b). The electronic configuration for the quintet state of the 1 1 1 2 1 2 2 0 species 1 is (xy) (*xz) (* yz) (*z ) (* x -y ) . The Fe-O distance is estimated to be 1.755 Å, while the Fe-N bond of the acetonitrile is 3.899 Å, leading to a distorted square pyramidal geometry. The Fe-O distance found for the S=2 state is much longer than that observed for the S=1 state and this is due to the presence of an unpaired electron in the antibonding * z2 orbital which reduces the Fe-O bond order. The elongation of Fe-N bond in the S=2 state is due to strong pseudo-Jahn-Teller distortion which favours elongation along the z-direction and compression along the Fe-C bond for species 1. Our attempt to search for the second pseudo-Jahn-Teller isomer with longer Fe-C bonds and shorter axial bonds reveal no defined second minima. (Other isomer found to lie at ~27.2 kJ/mol higher in energy, see Figure S11 in ESI for relaxed scan profile). Usually, nonheme complexes at S=2 state will possess unpaired electron 2 2 in *x -y orbital exhibiting compression along the z-direction, here strong equatorial ligation and weak axial ligation of the MeCN, pushes the * x2- y2 orbital very high in energy leading to a different electronic state (see Figure 3). Although, for species 1, there is a reduction in the radical character at the ferryl o xygen, significant spin density is still detected at the ferryl o xygen atom and this suggests a possible C-H bond activation for this species. To fully comprehend the implication of the electronic structure, we have computed the spectral features of species 1.

This article is protected by copyright. All rights reserved.

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

FULL PAPER Reactivity of species 1 and 2 towards hydrox ylation of methane:

1.236 O C4 C1

2.699 C2

N

Figure 3. B3LY P-D2 optimized structure of the quintet state (S=2) of the NHC IV 2+ [(L )Fe (O)( MeCN)] (species 1) and its corresponding spin density plot.

Calculations yield Mossbauer isomer shift  of -0.139 mms -1 and the quadrupole splitting parameter EQ of 2.97 mms-1 for the S=1 state. The  value computed is relatively smaller than the  value reported for other iron(IV)-oxo species (in the range of 0.15-0.20 mms-1).[10] The lower isomer shift for species 1 reflects the presence of strong equatorial σ-donor macrocyclic tetracarbene ligand with charge donation into the 4s-orbital of the iron centre as supported by the NBO calculations. This charge transfer also adds to the stability of the iron(IV)-oxo species. The zero-field splitting parameter of [13, 22] the S=1 state is estimated using NEVPT2 approach -1 -1 yield D=16.52 cm with E/D of 0.015 cm and these are in excellent agreement with the earlier reports [10, 23 ] (see in Table 1). The estimated axial D parameter is relatively smaller compared to other iron(IV)-oxo triplet state suggesting a reduction in the spin-orbit coupling for species 1. Very large ZFS parameter generally suggests strong mixing of the spin states - a desired condition to observe two-state reactivity. Since the magnitude of D parameter is small for species 1, this is likely to further affect its reactivity.[24]

To probe the reactivity of species 1 vs. 2, we have performed [27] the computational study on methane activation. This is particularly chosen as methane is known to be a very inert substrate (Scheme 2). Although the reactivity of species 1 has not been known when this manuscript was written, very recently Mayer’s group established the reactivity of species 1 with various substrates with relatively weaker C−H bonds such as 1,4 cyclohexadiene, 9,10-dihydroanthacene, 9Hxanthene and 9H-fluorene.[28] Besides reactivity of Fe(IV)=O carbene species possessing slightly different architecture has been documented and its C-H bond activation and oxygen atom transfer abilities have been demonstrated earlier.[29]

Table 1. Computed spectroscopic parameters f or the species 1. Fe =O

IV

∆EQ

δ(mms )

D(cm )

S=2

-1.78

-0.117

-

-

S=1

2.97

-0.139

16.52

0.015

S=0

-

-

-

-

[10]

exp

3.08

-1

-0.13

-1

16.40

E/D

-

To compare the electronic structure and the reactivity of this species to the heme iron(IV)-oxo species, we have chosen to study [(Porphyrin)Fe IV(O)(SH)]- (2) species (see Figure 1b). This is the active site structure of compound II (cpd-II) in cytochrome P450 [25] which has shown to exhibit moderate reactivity compared to the compound I of the cytochrome P450. Besides, this particular model complex has also been prepared and its catalytic abilities towards C-H bond activation and epoxidation of olefins are tested. [2d, 2f] Similar to cpd-I of cytochrome P450, this species also found to exhibit strong axial ligand effect on the reactivity. Our calculations reveal that the ground state of 2 to be a triplet with the S=2 found to lie at 38.7 kJ/mol higher in energy (see Table 2).[26] The electronic configuration for the ground state (S=1) of the species is computed to be (xy) 2(* xz)1(* yz)1 (* x2- y2) 0 (*z2) 0 (see Figure 2b). Unlike in species 1, the *xz / yz orbitals are pure Fe-O orbital and the xy- *xz /yz gap is estimated to be 0.42 eV and this is much larger than that computed for species 1.

Scheme 2. The schematic mechanism proposed f or C-H bond activ ation of methane by iron(IV)-oxo species.

The hydroxylation of methane is expected to proceed through a C-H bond activation step (via TS1) followed by the formation of a radical intermediate. The rebound of -OH group from the intermediate to the radical species via TS2 lead to the hydroxylation product. For the C-H bond activation step, two different pathways, σ and  based on the Fe-O-H angles are proposed.[30] In the σ pathway, the electron from the σ* orbital C-H bond is expected to occupy *z2 orbital of the metal. In the case of  pathway the electron from the C-H bond is expected to occupy * xz/* yz orbitals of the metal. This is shown in orbital evolution diagram depicted in Figure 4. In σ pathway (via σ-TS1), the Fe-O-H angle is expected to be close to linearity, while in the  (via -TS1) pathway, the angle close to 120 degrees are [30] expected.

This article is protected by copyright. All rights reserved.

Accepted Manuscript

C3 Fe

10.1002/chem.201800380

Chemistry - A European Journal

Computed potential energy surface for the h ydroxylation reaction of species 1 is shown in Figure 5. For species 1, we have computed five different C-H bond activation barrier heights for the S=1, 0 and 2 surfaces. The barrier height o n 3 1-- TS1 channel is 123.2 kJ/mol which is estimated to be the lowest and for the 1-σ-3 TS1, 1--5TS1 and 1-σ-5TS1 channels, the barriers are estimated to be 183.2 kJ/mol, 197.1 kJ/mol and 221.9 kJ/mol, respectively. For the S=0 singlet surface (1-1TS1), a much higher barrier of 238.1 kJ/mol was obtained and this ruled out the possibility of S=0 participating in this reaction. The newly forming O-H bond distance at the TS1 is found to be shorter at the triplet state (1.148 Å/1.108 Å for 1-σ-3TS1/1--3TS1 species) compared 5 5 to the quintet states (1.168 Å/1.204 Å for 1-σ- TS1/1-- TS1 species, see Table S1 in ESI). The computed Fe-O-H angle 3 for the lowest triplet transition state (1-- TS1) is 114.8 which correspond to  pathway (see Figure 6a). For the S=2 pathway, the Fe-O-H angle is estimated to be 110.0 suggesting a  pathway (1--5TS1).

Figure 4. Orbital occupancy diagrams for the H-abstraction processes and corresponding orb ital selection rules f or predicting transition-state structures in species 1. The Fe-O-H angle giv en in f igures are obtained f rom the calculated results.

3

Figure 6. HOMO of transition states f or (a) 1- - TS1 species exhibiting 5 interaction with the C-H bond (b) 1-σ- TS1 species exhibiting σ-interaction.

1-1TS1 238.1 221.9 1-σ-5TS1 5 1-π- TS1 197.1 1-σ-3TS1 183.2

1-π-5TS2

191.1 1-σ-3TS2

121.1

93.1

1-1R

1-π-3TS1

138.7

123.2 1-5R

1-5Int 77.2

84.0

1-3Int 1-5P

0.0

14.3

1-3R 1-3P

Figure 5. B3LY P-D2 computed potential energy surf ace (∆G in kJ/mol) f or methane hy droxy lation by [(L

NHC

IV

)Fe (O)( MeCN)]

This article is protected by copyright. All rights reserved.

2+

(species 1).

-41.7

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

FULL PAPER

Species

1

1a

1b

1c

1d

2

2a

5

93.1

97.6

78.6

86.2

50.6

38.7

49.2

3

0.0

0.0

0.0

0.0

0.0

0.0

0.0

1

R

121.1

129.0

124.3

120.9

111.6

123.0

133.4

σ- TS1

5

221.9

223.4

152.5

240.6

189.5

137.1

-

5

197.1

189.5

129.6

152.8

121.3

156.7

144.9

3

183.2

141.0

71.2

165.0

107.1

-

-

3

123.2

128.3

120.5

118.0

138.0

110.7

125.5

TS1

238.1

258.4

236.9

231.9

259.0

226.3

-

7

158.9

259.1

105.1

142.9

124.6

72.3

56.9

5

84.0

114.7

101.4

69.3

97.2

90.5

139.8

3

77.2

81.8

71.8

68.6

50.4

56.4

71.9

5

191.1

221.1

-

151.8

177.0

-

3

138.7

156.2

184.0

128.6

138.6

132.9

131.4

TS2

208.3

202.8

-

-

265.9

167.4

-

R R

 - TS1 σ- TS1  - TS1 1

Int Int Int

 - TS2 σ- TS2 1

5

14.3

40.2

-51.1

-11.0

-11.1

-91.4

-48.4

3

-41.7

-11.5

-110.1

-50.1

-77.2

-69.2

-72.6

1

-42.4

4.2

-42.2

-50.9

-8.6

-39.6

-57.3

P P P

Generally, S=2 state reacts via σ-pathway, as this maximizes the exchange energy leading to stabilization of the transition state and the observation of two-state reactivity (TSR). To further probe the reactivity of 1, we have also explored other reaction channels, where we have computed σ pathway of the S=1 (1-σ-3TS1) as well as for S=2 states (1-σ-5TS1) (see Figure 6b). These transition states computed are found to be even higher in energy at 183.2 kJ/mol and 221.9 kJ/mol (60.0 kJ/mol and 24.8 kJ/mol higher than the corresponding  pathway). This clearly suggests that expected lowering of the kinetic barrier due to the choice of various reaction channels are absent in species 1 leading to the observation of a pure triplet state reactivity. In all the computed TS1 structures, clearly significant spin densities are detected at the carbon atom of the methane molecule (see Figure 7 and Table S2 in ESI) and this suggests that the reaction proceeds via radical pathway as shown in Scheme 2. The hydrogen atom abstraction leads the formation of the Fe III-OH intermediate. For this intermediate (1-Int), a triplet state is found to be the ground state, with the quintet state lying at 6.8 kJ/mol higher in energy. Thermodynamically, the formation of this intermediate (Fe III-OH) is found to be endothermic by 77.2 kJ/mol. In the next step -OH rebound is expected to occur. In some cases, -OH rebound step is computed to be the rate determining step [26] and therefore the estimation of this step becomes important to fully understand the reactivity pattern. For species 1, rebound barriers are estimated to be 138.7 kJ/mol at the triplet surface and 191.1 kJ/mol at the quintet surface. We have also estimated the barrier height at the singlet surface and this is estimated to be very high (208.3 kJ/mol, see Table 2).

0.761

C H C4

0.505

O C3

Fe

C1

0.944

C2 N

(a) H

C

N4 N1

Fe

0.792 0.476

O N3

0.949

N2

S

(b)

Figure 7. B3LY P-D2 a) optimized structure of the lowest transition state of the C-H bond activ ation by the [(L )Fe (O)( MeCN)] (species 1) (1- - TS2) and its corresponding spin density plot. b) optimized structure of the lowest IV transition state of the C-H bond activ ation by the [(Porphy rin)Fe O(SH)] NHC

IV

2+

3

(species 2) (2- - TS2) and its corresponding spin density plot. 3

Earlier Neese and co-workers reported [30a] that the reactivity of iron(IV)=O group are expected to follow σ-pathway for the H-abstraction followed by a -pathway for the rebound step or vice versa. For the rebound at the 1-σ-3TS2 transition state, the Fe-O-C bond angle is computed to be 157.7° suggesting a σ-pathway (see Figure 8 and Table S1 in ESI).[26, 31] Thus the reaction follows  pathway for TS1 followed by σ-pathway for the hydroxylation step. Very large barrier computed for the rebound step suggests that both the C-H bond activation and -OH rebound steps are decisive kinetic barriers. However here as well, the quintet state barriers are estimated to be very high in energy ruling out the possibility of S=2 state participating in the reaction. Thus the

This article is protected by copyright. All rights reserved.

Accepted Manuscript

Table 2. B3LY P-D2 computed relativ e energies (kJ/mol) of species 1-1d and 2-2a.The potential en ergy surf ace f igures corresponding to the energies giv en here are giv en in Figure S1-S5 and S9 in ESI.

10.1002/chem.201800380

Chemistry - A European Journal

whole reaction is expected to occur in triplet surface following rather a single state reactivity. The final hydroxylation product is found to poss ess a triplet ground state and its formation is exothermic by 41.7 kJ/mol. Again, as both the reactant and the products C

C4 C1

Fe N

0.947

-0.054

O C3

1.346

C2

Figure 8. B3LY P-D2 a) optimized structure of the lowest transition state of the NHC IV 2+ 3 O-H rebound of the [(L )Fe (O)(CH 3CN)] (species 1) (1-σ- TS2) and its corresponding spin density plot.

possess same spin multiplicity, TSR scenario can be ruled out. To compare the reactivity of species 1 with species 2, we have also computed the corresponding C-H bond activation transition states (see Figure 7b). For species 2, the barrier heights are estimated to 110.7 kJ/mol, 137.1 kJ/mol, 156.7 kJ/mol and 226.3 kJ/mol for 2--3TS1, 2-σ-5 TS1, 2--5TS1 1 and 2- TS1 species, respectively (see Table 2 and Figure S1 in ESI). For species 2 as well, the difference in barrier heights are larger suggesting a possible triplet state reactivity in the initial step and this has been witnessed also in the earlier work.[26] The formation of the radical intermediate is found to be endothermic (56.4 kJ/mol at triplet surface) and the rebound steps are estimated to have the barrier heights 3 of 132.9 kJ/mol, 177.0 kJ/mol, and 167.4 kJ/mol for 2-σ- TS2, 2--5TS2 and 2-σ-1TS2 transition states , respectively. Although reaction likely to proceed in the triplet surface until the rebound surface, the final hydroxylation product found to have quintet ground state and its formation is exothermic by 91.7 kJ/mol. Since the final product has different multiplicity than the reactant, this suggests a possible spin crossover prior to the formation of product and invokes a likely TSR scenario for this reaction. Role of axial ligands in the Electronic Structure and Reactivity of species 1 and 2 To assess the role of a xial ligand in fine-tuning the reactivity, a series of complexes possessing thiol (-SH; species 1a), tertiary-amine (N(CH 3)3; species 1b), tertiary-phosphine (P(CH 3)3; species 1c) and without any ligand (species 1d) at the axial position of 1 are modelled. While such axial ligands have been routinely used on non-heme iron(IV)-oxo chemistry, e xperimentally variation of the axial ligands for 1 has not been tested yet. Additionally, we have also modelled species 2 with axial acetonitrile coordination (2a) to compare and contrast the results with heme models. Our computed results show that independent of the nature of the axial ligation, the ground state is a triplet in all cases. The gap between the triplet and the quintet states are estimated to be 93.1, 97.6, 78.6, 86.2, 50.6, 38.7 and 49.2 kJ/mol for species 1, 1a, 1b, 1c, 1d, 2 and 2a respectively (see Table 2). Clearly, the quintet states are very high-lying in energy for all

axial ligands tested for s pecies 1 with an exception of 1d, while for species 2 only moderate gaps are detected even with acetonitrile at the axial position. The S=1 state of 1a-1d possess the following electronic 2 1 1 2 0 2 2 0 configuration, (xy) (* xz) (* yz) (* z ) (* x - y ) same as that of species 1. Among the axial ligands tested, thiol group has greater electron donating (basicity) ability followed by tertphosphine, tert-amine, and acetonitrile. This is reflected in the Fe-O bond lengths with –SH group exhibiting the longest Fe-O bond length with both carbene and porphyrin ligand moiety. This is followed by P(CH 3)3, N(CH 3)3,CH3CN and noaxial ligand (see Table S1 of ESI).[5d] The amount of spin density on the oxygen centre particularly unravels the electrophilic/nucleophilic nature of the metal-oxygen bonds. All the six species studied here have an oxyl-radical character at the ferryl o xygen centre. This invariably suggests that all the Fe IV=O species studied behave as an electrophile. The largest oxylradical character was detected for species 2a followed by 2. Within various NHC based species, the largest spin densities are detected for 1 and the smallest detected for 1d. Quite interestingly, both in porphyrin as well as NHC moiety, acetonitrile group found to promote strong radical character. The S=2 state of species 1a-1c on the other hand possess the electronic configuration (xy)1(*xz) 1(* yz)1(* x2-y2)1(*z 2)0 where *z2 orbital is strongly destabilized compared to *x2-y2. (E(*x2-y2-*z2) ~ 2 .7 e V in all three cases). This is similar to the S=2 state of species 2 but contrary to species 1 and 1d. Due to this electronic configuration, the Fe-O bond lengths for species 1a-1c are barely altered compared to the triplet state (in the range of 1.658 Å to 1.683 Å). However, the Fe-C bonds in these species are found to be longer (Fe -C avg are 2.004 Å and 2.152 Å for 1 and 1a respectively, see Table S1 in ESI) compared to the S=2 state of species 1. This suggests stabilization of a pseudo-Jahn-Teller compression along Fe=O direction for the S=2 species . Stronger axial donation in 1a-1c compared to MeCN/no-axial ligand in 1 and 1d, lead to stabilization of pseudo-J-T., compressed geometries for species 1a-1c. To understand the role of axial ligands on the reactivity further, we have also computed the C-H bond activation of methane by species 1a-1d and 2a. For species 1a, the barrier heights are estimated to 128.3 kJ/mol, 189.5 kJ/mol and 258.4 kJ/mol for the triplet, quintet, and singlet spin surface respectively. Despite possessing strong basic ligand in the axial position, the barrier heights are barely altered in 1a compared to 1 and the quintet state still lies very high in energy. The Fe-O-H bond angle for 1a-3TS1 and 1a-5TS1 are 113.3° and 110.2° and this suggests -pathway for the C-H bond activation on both the surfaces .[30a] To rule out the possibility that σ-pathway of S=1 and S=2 could be lower for 1a, we have also computed the corresponding transition states, 1a-σ-3TS1 and 1a-σ-5TS1 and these are estimated to be 12.7 kJ/mol and 33.9 kJ/mol higher compared to 1a-3 TS1 and 1a--5TS1 states, respectively (see Table 2 and Figure S2 in ESI). The Fe-O-H angle for S=2  pathway is estimated to be 124.9 degrees and this is far away from linearity suggesting weaker overlap and hence a larger

This article is protected by copyright. All rights reserved.

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

kinetic barrier along this channel. To estimate the energy penalty required to form a linear Fe-O-H bond, we have performed the single point energy calculation on 1a-σ-5TS1 species by fi xing the Fe-O-H angle to 180°and this geometry is found to 10.7 kJ/mol higher in energy compared to the ground state of 1a-σ-5TS1 and our attempt to obtain another transition state with linear angle did not succeed. It is interesting to note here that, despite species 1a possessing different electronic configuration compared to species 1, the estimated kinetics are very similar. A very close look at the electronic structure of the 1a--5TS1 reveals that the unpaired electron present in *z2 orbital and not the lower lying * x2-y2 found in the reactant state. This suggests that during the course of the reaction, the computed pseudo J-T., compressed S=2 switches to pseudo J-T., elongated state resulting in a reactivity similar to that observed for species 1. To assert this further, the Fe-O bond lengths at quintet state of the reactant and 1/1a--5TS1 transition states can be compared. For species 1, the Fe-O bond length elongates moderately from 1.775 Å to 1.943 Å at the 1--5TS1 transition state while in species 1a this change is significant from 1.683 Å to 2.041 Å (s ee Table S1 in ESI). As the ferryl oxygen accept the H atom, the Fe=O  characters weakens leading to a longer Fe-O bond length at the transition state and hence a pseudo J.T., elongated structure. The MO analysis also confirms this picture where *x2-y2 is found to be empty at the transition state. The longer Fe-O distance is also accompanied by the concomitant shorter Fe-C distances for the 1a--5TS1 species (Fe-C avg are 2.152 Å 5 and 2.035 Å at the S=2 reactant at 1a-- TS1, respectively). Thus despite the stronger donation by the -SH group, the reactivity patterns are unaltered. In the next step, the H-atom abstraction lead to the formation of Fe III-OH intermediate possessing S=1 ground state and its formation is exotherm ic by 81.8 kJ/mol. For the rebound step, the triplet state (1a-3TS2) is found to be lower in energy and here the Fe-O-C bond angle is computed to be 147.6° suggesting -pathway as seen in species 1. The formation of the hydroxylated product is exothermic (11.5 kJ/mol) and possesses an S=1 ground state and does not necessities a spin-crossover from the reactant to the product formation. We have also explored the reactivity of species 1b, 1c and 1d, to further conclude the overall reactivity pattern (see Table 2 and Figure S3-S5 in ESI). The computed geometry and electronic structures for species 1b and 1c are very similar to that of 1a and thus have not been elaborated further (see Figure S6 and S7 in ESI). Likewise, the electronic structure of 1d is similar to species 1 (see Figure S8). The C-H bond activation barriers for 1b (1c) [1d] are estimated to be 120.5 (118.0) [138.0] kJ/mol, 71.2 (165.0) [107.2] kJ/mol, 129.6 (152.8) [121.3] kJ/mol and 152.5 3 3 5 (240.6) [189.5] kJ/mol for - TS1, σ- TS1, - TS1 and σ5 TS1 species, respectively (see Table 2 and Figure S3-S5 in ESI). For all four species, the lowest lying transition state is S=1, however there are some difference. Particularly, species 1b and 1d are found to proceed via σ-3TS1 .i.e. via

triplet σ channel while in all the other cases , the reactions are found to be channelized through the triplet  pathway. A closer look at the transition state structure reveals that in 1b, the axial N(CH 3)3 group cleaves and weakly hydrogen bonded to the NHC ligand (s ee Figure S8 in ESI). This is very similar to the transition state structure computed for 1d where no axial ligand is present. The reacti vity of species 1c is found to be similar to species 1a. For species 1d, however, the 1d--5TS1 transition state is only 14.2 kJ/mol higher than the triplet transition state and this suggests a possible TSR scenario. A similar conclusion has also been derived for species 1d for other chemical transformations such as epoxidation.[29] The difference in energy between the lowest lying S=1 and S=2 states are found to be very high for all species except 1d and this rules out the possibility of twostate reactivity for these species (1-1c). The formation of the radical intermediate (int1) is found to be endothermic for species 1a-1d and for this intermediate S=1 is found to be the ground state in all three cases . The –OH rebound barriers are estimated to be 184.0 kJ/mol, 128.6 3 3 kJ/mol and 138.6 kJ/mol for the 1b-σ- TS2, 1c-σ- TS2 and 3 [32] 1d-σ- TS2 transition states , respectively. The product also has a triplet ground state and its formation is exothermic in all three cases (110.1 kJ/mol, 50.1 kJ/mol and 77.2 kJ/mol for 1b, 1c and 1d, respectively). Besides, for species 2a, the C-H bond activation step has barrier heights of 125.5 kJ/mol and 144.9 kJ/mol for the 2a--3TS1 and 2a--5TS1 states which are higher than that observed for species 2 (see Table 2 and Figure S9 in ESI). The intermediate formation is found to be endothermic in nature and the rebound barrier computed on the triplet surface is estimated to be 131.4 kJ/mol (2a-σ-3TS2) and this is similar to that observed for species 2.

Discussion TSR reactivity is one of the celebrated concept in high-valent iron(IV)-oxo species and has been termed as one the reason for its very high reactivity observed. Over the years , many factors influencing this reactivity has been explored and the donor abilities of the axial ligand have been emphasized as the most import factor. This has been witnessed both in experiments and theory. However, in all the cases tested, the equatorial ligands are only moderate donors compared to the oxygen atom and the axial ligand put together. If this scenario changes, such as in the studied case, this may dramatically influence the reactivity pattern. Electronic structure of species 1 and 2 differs drastically both in the triplet ground state and in the important first excited quintet state. The most striking difference are the energies of 2 σ* orbitals where *z is found to be lower in energy for 1 while *x2-y2 found to be lower in energy for 2. Higher reactivity of iron(IV)-o xo group stems from lower barrier height at the transition state for the quintet states. Generally, the quintet state reacts via σ-pathway whereby it accepts an α electron into the *z2 orbital. This also maximizes the exchange energy and there by stabilizes the transition state. For species 2 and other non-heme iron(IV)-oxo species

This article is protected by copyright. All rights reserved.

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

having empty low-lying *z2 orbital [33], this scenario is true leading to higher reactivity. [34]

8

*z2

Energy (in eV)

6

*x2-y2

4

*yz/*xz 2

xy

0

1

1a

1b

1c

1d

2

2a

S=2

8

* z2 * x2-y2

Energy (in eV)

6

4

2

*yz/*xz

xy 0

xy

1

1a

1b

1c

1d

2

2a

S=1 Figure 9. Orbital energy diagram of the species 1-1c, 2 and 2a in their triplet and quintet states.

On the other hand, for species 1, as the *z2 orbital is already filled, only * x2- y2 can accept α electron in the σpathway. As this orbital is strongly destabilized in species 1, this more than compensate the additional exchange stabilization available at the  S=2 pathway leading to the lower unconventional  S=2 pathway. This precludes the possibility of S=2 participating in the reactivity of species 1. The absence of TSR is clearly visible beyond the first transition state, in stabilizing the S=1 state for the intermediate, possessing significantly lower barrier for the triplet state at the rebound step and leading to the formation of an Fe(II) product possessing an unusual S=1 state. Thus the entire reaction proceeds in triplet surface leading to a single state reactivity. For species 2, the barrier heights computed for the triplet and quintet are closer but not very close to suggest a TSR scenario. However, the Fe(II) product has an S=2 quintet ground state suggesting a spin

crossover after the rebound step leading to the possibility of observing TSR. Earlier theoretical study performed on this species for other chemical transformation also clearly affirms this picture.[28-29] Although, the axial ligands are found to influence the reactivity of several heme and non-heme iron(IV)-oxo, the reactivity pattern computed for species 1a-1c are very similar to that computed for 1. Clearly, the a xial ligand modulates the electronic configuration of S=2 state, with the very strong donor destabilizing (see Figure 9, top) the *z2 orbital. However, triplet state orbital ordering and reactivity's are unaltered across the series rendering very similar reactivity pattern across the series studied (see Figure 8, bottom). The axial ligand found to particularly influence the *xz/yz orbital, however here we have not witnessed any noticeable changes in the orbital splitting/ordering. This is perhaps due to the -acceptor abilities of carbene, which could compensate this effect offered by the axial ligands (see Figure 9, bottom and Figure S10 in ESI where such interactions are noted in NBO donor-acceptor interaction). Besides the nature of S=2 state found to switch from one pseudo J-T., isomer to another during the formation of the transition state (TS1), for this reason, the  S=2 pathway is found to be higher in energy compared to the unconventional  S=2 pathway for these species (with the exception of species 1b). As seen in species 1, the occupation of * x2- y2 is required at the  S=2 pathway and this is unlikely to be influenced by the nature of axial ligand. This dictates again a very similar reactivity across the series studied. For species 1d, where no axial ligands present, calculation depict different picture where  S=1 and  S=2 transition states are close lying in energy in the rate-determining C-H bond activation step. This unequivocally suggests TSR reactivity for this species as has been witnessed also for other tested reactions.[29] To compare and contrast the reactivity among all the species computed, we have taken the lowest energy pathway computed for species 1, 1a-d and 2 and 2a and plot them together in Figure 10. Clearly, the barrier heights computed at the  S=1 surface for the 1 and 1a-c are unaltered and S=2 σ-pathway is high-lying. For 2 and 2a larger difference in the barrier heights are noted but closer margin and multiplicity of the product suggest TSR scenario. Significant barrier heights are noted also for the rebound step for all species and this suggests that both the steps are important in deciding the kinetics of the reaction. For species 1, 1a, 1c and 1d the reaction found to proceed via  S=1 path in the C-H activation step followed by σ S=1 path at rebound step leading to S=1 product. For 1b, however slight difference are noted where reaction found to proceed via σ S=1 path in the C-H bond activation step. For species 2 on the other hand, the reaction proceeds via σ S=1 followed by the  S=1 pathway leading to an S=2 product.

This article is protected by copyright. All rights reserved.

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

FULL PAPER 184.0

180

160

--- 1 --- 1a --- 1b --- 1c --- 1d --- 2 --- 2a

156.1

140

128.3 123.2 120.5

120

100

131.4

125.5 118.0 110.7 107.1

101.4 84.0

81.8

80 60

TS1

40

138.6138.7 132.9 128.6

71.9 56.4

TS2

50.4

20 0

0.0

Int -11.5

-20

-40

R

-41.7 -50.1

-60

-69.2

-72.6

-80

-77.2

100

-110.1

120

P

140

Figure 10. B3LY P-D2 computed potential energy surf ace (∆G in kJ/mol) for the comparison of the triplet state reactiv ity of species 1, 1a, 1b, 1c,1d, 2 and 2a

1

Conclusions

2

3

A

From the above calculations, it is very clear that species 1 does not exhibit two-state reactivity. This is essentially due to very strong equatorial donations pushing the σ* x2- y2 orbital to higher energy. This along with already destabilized *z2 due to strong Fe IV=O bond, enforce the reaction to proceed only in the triplet surface. The axial tuning by altering the *z2 orbital energy does not bring forth larger transformations to make quintet state accessible. This is clearly witnessed when both the σ and  pathways are calculated for the quintet state/triplet states. Unlike the usual iron(IV)-oxo reactivity, where σ, S=2 is low-lying in energy, here S=2 also prefer to react via  pathway leading to sluggish reactivity (with the exception of 1b). Besides the C-H bond activation step, there is also a substantial barrier for rebound step and

4

5

this hinders the reactivity further. Secondly, the axial ligands found to have very little effects on the reactivity and the barrier height computed at the σ triplet surface for various axial ligands are very similar affirming the above statement. To this end, a comprehensive DFT calculations have been employed to probe the electronic structure and reactivity of iron(IV)-oxo species having strong equatorial carbene ligands. These species are found not to exhibit two-state reactivity as the quintet states are found to lie very high in energy and the axial ligands are not influencing significantly the electronic structure/reactivity of this species. Both these points are contrary to the established concepts both in heme and non-heme iron(IV)-oxo chemistry and stress the need to focus also on designing appropriate equatorial ligands in fine-tuning the reactivity.

Keyw ords: Equatorial ligation • iron( IV)-oxo • N- heterocyclic carbene • tw o-state reactivity • DFT

References [1] a) W. N. Oloo, L. Que Jr, Acc. Chem. Res. 2015, 48, 2612-2621; b) C. Krebs, D. Galonic Fujimori, C. T. Walsh, J. M. Bollinger Jr, Acc. Chem. Res. 2007, 40, 484-492; c) W. Nam, Acc. Chem. Res. 2007, 40, 522-531.

[2] a) S. Shaik, S. Cohen, Y. Wang, H. Chen, D. Kumar, W. Thiel, Chem. Rev. 2010, 110, 949-1017; b) P. R. Ortiz de Montellano, Chem. Rev. 2009, 110, 932-948; c) S. Shaik, D. Kumar, S. P. de Visser, A. Altun, W. Thiel, Chem. Rev. 2005, 105, 2279-2328; d) S. Kozuch, T. Leifels, D. Meyer, L. Sbaragli, S. Shaik, W.-D. Woggon, Synlett. 2005, 2005, 675-684; e) B. Meunier, S. P. De Visser, S. Shaik, Chem. Rev. 2004, 104, 3947-3980; f) S. P. de Visser, F. Ogliaro, P. K. Sharma, S. Shaik, J. Am. Chem. Soc.

This article is protected by copyright. All rights reserved.

Accepted Manuscript

200

10.1002/chem.201800380

Chemistry - A European Journal

2002, 124, 11809-11826; g) W. Nam, Y. M. Goh, Y. J. Lee, M. H. Lim, C. Kim, Inorg. Chem. 1999, 38, 3238-3240; h) M. Sono, M. P. Roach, E. D. Coulter, J. H. Daw son, Chem. Rev. 1996, 96, 2841-2888; i) J. T. Groves, Z. Gross, M. K. Stern, Inorg. Chem. 1994, 33, 5065-5072; j) T. E. Nemo, B. Evans, J. Am. Chem. Soc. 1981, 103, 2884-2886. [3] W. Nam, S.-E. Park, I. K. Lim, M. H. Lim, J. Hong, J. Kim, J. Am. Chem. Soc. 2003, 125, 14674-14675. [4] a) D. F. Leto, A. A. Massie, D. B. Rice, T. A. Jackson, J. Am. Chem. Soc. 2016; b) D. Schroder, S. Shaik, H. Schwarz, Acc. Chem. Res. 2000, 33, 139-145; c) P. Comba, M. Kerscher, W. Schiek, in Prog. Inorg. Chem., John Wiley & Sons, Inc., 2007, pp. 613-704; d) W. Nam, Y.-M. Lee, S. Fukuzumi, Acc. Chem. Res. 2014, 47, 1146-1154. [5] a) S. Shaik, S. Cohen, Y. Wang, H. Chen, D. Kumar, W. Thiel, Chem. Rev. 2009, 110, 949-1017; b) S. P. de Visser, L. Tahsini, W. Nam, Chem. Eur. J. 2009, 15, 5577-5587; c) S. P. de Visser, J. Biol. Inorg. Chem. 2006, 11, 168-178; d) T. Kamachi, T. Kouno, W. Nam, K. Yoshizawa, J. Inorg. Biochem. 2006, 100, 751-754. [6] a) H. Hirao, L. Que, W. Nam, S. Shaik, Chem. Eur. J. 2008, 14, 1740-1756; b) C. V. Sastri, J. Lee, K. Oh, Y. J. Lee, J. Lee, T. A. Jackson, K. Ray, H. Hirao, W. Shin, J. A. Halfen, J. Kim, L. Que, S. Shaik, W. Nam, Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 19181-19186; c) C. V. Sastri, M. J. Park, T. Ohta, T. A. Jackson, A. Stubna, M. S. Seo, J. Lee, J. Kim, T. Kitagawa, E. Münck, J. Am. Chem. Soc. 2005, 127, 1249 -12 d) T A Jackson, J ohde, M S Seo, C V Sastri, DeHont, A Stubna, T hta, T itagaw a, E M nck, W. Nam, J. Am. Chem. Soc. 2008, 130, 12394-12407. [7] Zhong, A C Lindhorst, F J Groche, F E hn, Chem. Rev. 2017, 117, 1970-2058. [8] D. A. Sharon, D. Mallick, B. Wang, S. Shaik, J. Am. Chem. Soc. 2016, 138, 9597-9610. [9] a) S. A. Cramer, D. M. Jenkins, J. Am. Chem. Soc. 2011, 133, 19342-19345; b) F. E. Hahn, V. Langenhahn, T. Lügger, T. Pape, D. Le Van, Angew. Chem., Int. Ed.. 2005, 44, 3759-3763. [10] S. Meyer, I. Klaw itter, S. Demeshko, E. Bill, F. Meyer, Angew. Chem., Int. Ed.. 2013, 52, 901-905. [11] a) S. P. de Visser, Angew. Chem. 2006, 118, 1822-1825; b) S. J. Kim, R. Latifi, H. Y. Kang, W. Nam, S. P. de Visser, Chem. Commun. 2009, 1562-1564; c) A. Ansari, A. Kaushik, G. Rajaraman, J. Am. Chem. Soc. 2013, 135, 4235-4249; d) A. Ansari, M. Ansari, A. Singha, G. Rajaraman, Chem. Eur. J. 2017, 23, 10110-10125; e) A. Ansari, G. Rajaraman, Phys. Chem. Chem. Phys. 2014, 16, 14601-14613; f) P. Jayapal, A. Ansari, G. Rajaraman, Inorg. Chem. 2015, 54, 11077-11082. [12] Gaussian 09, Revision 02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr. , J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, :. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc. Wallingford, CT, 2009. [13] F. Neese, WIRES 2012, 2, 73-78. [14] S. Grimme, J. Comput. Chem. 2006, 27, 1787-1799. [15] a) R. Ditchfield, W. J. Hehre, J. A. Pople, J. Chem. Phys. 1971, 54, 724-728; b) P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299-310. [16] a) A. Schäfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571-2577; b) A. Schäfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829-5835; c) B. Pandey, A. Ansari, N. Vyas, G. Rajaraman, J. Chem. Sci. 2015, 127, 343-352; d) M. Jaccob, A. Ansari, B. Pandey, G. Rajaraman, Dalton Trans. 2013, 42, 16518-16526.

[17] G. Zhurko, D. Zhurko, Lite version build 2005, 8, 2005. [18] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999-3094. [19] a) M. Radoul, M. Sundararajan, A. Potapov, C. Riplinger, F. Neese, D. Goldfarb, Phys. Chem. Chem. Phys. 2010, 12, 72767289; b) M. Roemelt, S. Ye, F. Neese, Inorg. Chem. 2009, 48, 784. [20] F. Neese, J. Am. Chem. Soc. 2006, 128, 10213-10222. [21]C. E. MacBeth, R. Gupta, K. R. Mitchell-Koch, V. G. Young, G. H. Lushington, W. H. Thompson, M. P. Hendrich, A. Borovik, J. Am. Chem. Soc. 2004, 126, 2556-2567. [22] S. F. Ye, F. Neese, J. Chem. Theory Comput. 2012, 8, 23442351. [23] S. Ye, C. Kupper, S. Meyer, E. Andris, R. Navrátil, O. Krahe, B. Mondal, M. Atanasov, E. Bill, J. Roithová, F. Meyer, F. Neese, J. Am. Chem. Soc. 2016, 138, 14312-14325. [24] a) J. N. Harvey, M. Aschi, H. Schwarz, W. Koch, Theor. Chem. Acc. 1998, 99, 95-99; b) H. Chen, W. Lai, S. Shaik, J. Phys. Chem. B 2011, 115, 1727-1742. [25]a) K. L. Stone, L. M. Hoffart, R. K. Behan, C. Krebs, M. T. Green, J. Am. Chem. Soc. 2006, 128, 6147-6153; b) E. Derat, S. Shaik, J. Am. Chem. Soc. 2006, 128, 8185-8198. [26] A. Rosa, G. Ricciardi, Inorg. Chem. 2012, 51, 9833-9845. [27] M. Ansari, N. Vyas, A. Ansari, G. Rajaraman, Dalton Trans. 2015, 44, 15232-15243. [28] C. Kupper, B. Mondal, J. Serrano-Plana, I. Klaw itter, F. Neese, M. Costas, S. Ye, F. Meyer, J. Am. Chem. Soc. 2017, 139, 8939-8949. [29] F. G. Cantú Reinhard, S. P. de Visser, Chem. Eur. J. 2017, 23, 2935-2944. [30] a) S. Ye, F. Neese, Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 1228-1233; b) B. Pandey, M. Jaccob, G. Rajaraman, Chem. Commun. 2017, 53, 3193-3196. [31] G. Ricciardi, E. J. Baerends, A. Rosa, ACS Catal. 2015, 6, 568579. [32] L. Ji, A. Franke, M. Brindell, M. Oszajca, A. Zahl, R. van Eldik, Chem. Eur. J. 2014, 20, 14437-14450. [33]S. N. Dhuri, M. S. Seo, Y. M. Lee, H. Hirao, Y. Wang, W. Nam, S. Shaik, Angew. Chem., Int. Ed.. 2008, 47, 3356-3359. [34] S. Hong, P. Kumar, K. B. Cho, Y. M. Lee, K. D. Karlin, W. Nam, Angew. Chem., Int. Ed.. 2016, 55, 12403-12407.

This article is protected by copyright. All rights reserved.

Accepted Manuscript

FULL PAPER

10.1002/chem.201800380

Chemistry - A European Journal

FULL PAPER Entry for the Table of Contents

Accepted Manuscript

TOC: Strong equatorial ligation suppress the two-state reactivity and suggest a way to finetune the reactivity of FeIV =O species using equatorial ligands.

This article is protected by copyright. All rights reserved.