Dataset Paper Dataset for Modelling Reaction Mechanisms Using Density Functional Theory : Mechanism of ortho-Hydroxylation by High-Valent Iron-Oxo Species

Modelling reaction mechanisms using density functional theory is one of the popular routes to underpin the course of a chemical reaction. Although numerous publications have come out in this area, the pitfall of modelling such reactions and explicitly publishing the entire data set (structures, energies, coordinates, spin densities, etc.) which lead to the conclusions are scarce.Herewe have attempted to set a trendwherein all the computed data to underpin the reactionmechanismof ortho-hydroxylation of aromatic compounds by high-valent iron-oxo complexes (FeIII–OOH, FeIV=O, and FeV=O) are collected. Since the structure, energetics and other details of the calculations can be employed in future to probe/understand the reactivity pattern of such species, establishing the data set is justified. Here by analysing the presented results we also discuss in brief the presented results.


Introduction
Heme and nonheme metal catalytic reactions of aliphatic/ aromatic hydrocarbons are an important tool in pharmaceutical industry [1,2].Iron catalyzed hydroxylations of organic compounds have been reported by several catalysts and studied by several experimentalists and theoreticians [3][4][5][6][7][8].The Fe III -OOH, Fe IV =O, and Fe V =O species are reported to be involved in the hydroxylation of aromatic and aliphatic compounds [3,4].At several instances, although the presence of Fe III -OOH species has been directly detected, high-valent iron-oxo species are invoked to explain the reactivity of many mononuclear heme/nonheme iron enzymes.Particularly in the last decade the Fe IV =O species are very popular as they have been hypothesized as potential oxidants in several aliphatic/aromatic hydroxylation reactions [9][10][11][12][13].Besides the Fe IV =O, the Fe V =O intermediates also captured attention very recently in the hydroxylation reactions [3,4] (Figure 2).
Detection of Fe V =O species under ambient conditions is a challenging task and thus this species has been detected only at a few occasions [14][15][16].Of particular interest here is the recently reported ortho-hydroxylation of aromatic compounds by a putative iron(V)-oxo species [3,4].In this dataset paper, we aim to list all the data which are collected over the years to underpin this reaction mechanism.In our original study [17], we use [Fe II (TPA) (CH 3 CN) 2 ] 2+ (tpa = tris(2-pyridylmethyl)amine) complex (shown in Figure 1) which performs ortho-hydroxylation very selectively [3].Experimentally the mechanism of this regiospecific reaction is probed by various sets of tools such as spectroscopy and 18 O labelling experiments [3].Since hydrogen peroxide is employed as the oxidant in this chemistry, the oxygen atom of the H 2 O 2 is found to be incorporated in the product.The mechanistic study suggests the Fe III -OOH species as an obvious starting point, but during the course of reaction, the Fe IV =O and Fe V =O species could in principle be generated, which then catalyze the reaction.These two species are generated by either homolytically or heterolytically cleaving the O-O bond of the Fe III -OOH species (Figures 3, 4, and 5).
In our earlier mechanistic study we have modelled the reaction using DFT and assumed (i) Fe III -OOH, (iv) Fe IV =O, and (iii) Fe V =O as the possible oxidants and also probed the intrinsic mechanistic details on how each of these oxidants bifurcates as the reaction proceeds.The calculations were performed with B3LYP-D, B3LYP, wB97XD, B97D, M06-2X, OLYP, TPPSh, and MP2 functionals and the results within the employed functionals deviate to certain extent.As B3LYP-D is one of the state-of-the-art functionals, all the earlier results are discussed at this level and here we are presenting the employed dataset which leads to the mechanistic conclusions that this particular reaction is triggered specifically by an Fe V =O species and that species prefers to undergo electrophilic attack to facilitate the hydroxylation reaction.

Methodology
All calculations were carried out by using the Gaussian 09 program [18].The geometry optimizations have been performed initially using B3LYP [19,20].Basis set LanL2DZ for Fe [21][22][23][24] and a 6-31 G [25] basis set for the other atoms like C, H, N, and O are employed.Accuracy of B3LYP has been well documented by the work of Neese, Shaik, Siegbahn, Solomon, and Ryde earlier [26][27][28][29][30][31][32].Many reported density functionals have issues of obtaining correct spin ground state.Some specific functionals such as OLYP [33] and TPSSh [34] are reported to predict correct spin state for iron complexes [35,36].However these functionals lack features like dispersion.Due to the absence of dispersion, the transition states are too loosely bound leading to the increase of the barrier heights.Dispersion corrected functionals such as M06-2X [37], B97D [38], and wB97XD address this issue to a certain extent [39], while addition of dispersion to the popular B3LYP (due to the work of Grimme) is found to be more robust and reliable [38,40].In case of iron-oxo complexes, B3LYP and B3LYP-D [38] along with TPSSh were reported to predict correct spin state in comparison to the CCSD(T) method [41].
Here we applied eight levels of theory (B3LYP, B3LYP-D, wB97XD, B97D, M06-2X, OLYP, TPPSh, and MP2 [42]) for the mechanistic study of ortho-hydroxylation reaction of the benzoic acid by using [Fe III (TPA)(OOH)] 2+ catalyst (see Figure 1).We optimized all structures with B3LYP, B3LYP-D, B97D, wB97XD, and M06-2X and we also performed single point calculation with OLYP, TPSSh, and MP2 method on the B3LYP optimized geometries.With the exception of MP2, all methods support heterolytic cleavage.The performance of MP2 is the most unsatisfactory among the methods tested.This result is consistent with previous work indicating that high level quantum chemical methods (e.g., the spectroscopically oriented CI (SORCI) method of Neese [43]) fail to predict the correct ground state of Fe IV =O species.Among the tested methods, our results illustrate that B3LYP, B3LYP-D, and wB97XD are consistently predicting the correct spin state compared to experimental data [3,15].Since the performances of all three functionals are nearly the same, here we intend to discuss B3LYP results, and comparison to other functionals has been made where appropriate.

Dataset Description
This Dataset Paper consists of 30 items which are described as follows.
Dataset Item 1 (Spectra).This computed energy profile for pathway I a is optimized with B3LYP-D function.The computed values (Δ) are given in kJ/mol.Dataset Item 23 (Spectra).This computed energy profile for a relax scan for the generation from trans-Fe III -OOH complex is optimized with B3LYP function.The computed values (Δ) are given in kJ/mol.Dataset Item 24 (Spectra).This computed energy profile for a relax scan for the generation from trans-Fe III -OOH complex is optimized with B3LYP-D function.The computed values (Δ) are given in kJ/mol.

Concluding Remarks
The presented exhaustive data offer one possible way to collect a dataset from theoretical modelling of reaction mechanisms where all the computed parameters for a particular reaction studied earlier are listed.The collection of dataset has a great advantage; apart from the fact that it can help verify the energies and structures, the mechanism of reactions can also be qualitatively predicted if the electronic structures of all the relevant species (reactant, intermediates, transition states, and products) are available.
III -OOH is optimized with B3LYP function.The computed values (Δ) are given in kJ/mol.Dataset Item 3 (Spectra).This computed energy profile for heterolytic cleavage of O-O bond of cis-Fe III -OOH is optimized with B3LYP-D function.The computed values (Δ) are given in kJ/mol.Dataset Item 4 (Spectra).This computed energy profile for heterolytic cleavage of O-O bond of cis-Fe III -OOH is optimized with B97D function.The computed values (Δ) are given in kJ/mol.Dataset Item 5 (Spectra).This computed energy profile for ortho-hydroxylation by C-H bond activation (pathway II a ) by Fe V =O oxidant is optimized with B3LYP function.The computed values (Δ) are given in kJ/mol.Dataset Item 6 (Spectra).This computed energy profile for ortho-hydroxylation by C-H bond activation (pathway II a ) by Fe V =O oxidant is optimized with B3LYP-D function.The computed values (Δ) are given in kJ/mol.Dataset Item 7 (Spectra).This computed energy profile for ortho-hydroxylation by C-H bond activation (pathway II a ) by Fe V =O oxidant is optimized with wB97XD function.The computed values (Δ) are given in kJ/mol.Dataset Item 8 (Spectra).This computed energy profile for ortho-hydroxylation by C-H bond activation (pathway II a ) by Fe V =O oxidant is optimized with B97D function.The computed values (Δ) are given in kJ/mol.

Figure 3 :
Figure 3: Schematic mechanism for the ortho-hydroxylation reaction by proximal oxygen attack by Fe III -OOH species (pathway I a ).

Figure 4 :
Figure 4: Schematic mechanism for the ortho-hydroxylation reaction by Fe V =O oxidant where reaction may be proceeded via either C-H activation or electrophilic way.

Figure 5 :
Figure 5: Schematic mechanism for the ortho-hydroxylation reaction by Fe IV =O oxidant where reaction may be proceeded via either C-H activation or electrophilic way.
Schematic mechanism for the formation of intermediates Fe IV =O and Fe V =O species.