Halide Abstraction Competes with Oxidative Addition in the Reactions of Aryl Halides with [Ni(PMenPh(3−n))4]

Abstract Density functional theory (DFT) calculations have been used to study the oxidative addition of aryl halides to complexes of the type [Ni(PMenPh(3−n))4], revealing the crucial role of an open‐shell singlet transition state for halide abstraction. The formation of NiI versus NiII has been rationalised through the study of three different pathways: (i) halide abstraction by [Ni(PMenPh(3−n))3], via an open‐shell singlet transition state; (ii) SN2‐type oxidative addition to [Ni(PMenPh(3−n))3], followed by phosphine dissociation; and (iii) oxidative addition to [Ni(PMenPh(3−n))2]. For the overall reaction between [Ni(PMe3)4], PhCl, and PhI, a microkinetic model was used to show that our results are consistent with the experimentally observed ratios of NiI and NiII when the PEt3 complex is used. Importantly, [Ni(PMenPh(3−n))2] complexes often have little, if any, role in oxidative addition reactions because they are relatively high in energy. The behaviour of [Ni(PR3)4] complexes in catalysis is therefore likely to differ considerably from those based on diphosphine ligands in which two coordinate Ni0 complexes are the key species undergoing oxidative addition.


BENCHMARK OF FUNCTIONALS
* Single point energies on B3LYP-D3 geometries using the LANL2DZdp basis set/ECP for I, LANL2LTZ(f) for Ni and 6-311+G(d,p) for the rest of atoms, with free energy corrections from B3LYP-D3 calculations.
The evaluation of different functionals in the key step with PhI and PMe 3 as ligand shows minor differences among all of them. The open-shell single electronic structure for the halogen abstraction transition state was found to be the most stable electronic structure in all the cases (using the wavefunction stability evaluation of g09). We selected B3LYP-D3 as the functional because the dissociation energy of the phosphine was the most consistent with experimental data and the evaluation of the product ratio provides a full agreement with the experimental observations.

ALTERNATIVE MECHANISMS (A) Outer-sphere electron transfer from [Ni(PMe 3 ) 4 ] to ArX
Electron transfer from [Ni(PMe 3 ) 4 ] to ArX could in theory occur via an outer sphere mechanism, to form [Ni(PMe 3 ) 4 ] + and the radical anion of ArX; the latter would then spontaneously dissociate Xto form an aryl radical. This was explored using Marcus-Hush theory, where the barrier to electron transfer can be estimated using equations S1-S3. This assumes that the major energetic barrier is the rearrangement of the products immediately following the electron transfer event. This has been successfully used to describe the reactions of aryl halides with 'super electron donors'. 22 ΔG ‡ ≈ (λ i /4) · (1 + ΔG/λ i ) 2 (S1) λ i (species) = (E N (R C ) -E N (R N )) + (E C (R N ) -E C (R C )) (S3) • λ i is the internal reorganisation energy; λ i (NiP4 + ) is the internal reorganisation energy of [Ni(PMe 3 ) 4 ] + ; and λ i (ArX) is the internal reorganisation energy of ArX radical anion • ΔG is the free energy change of the reaction • E N (R N ) and E C (R C ) are the energies of the neutral and charged species in their respective geometries; E C (R N ) and E N (R C ) are the single point energies of the neutral and charged geometries as charged and neutral species, respectively.
While the initial result with [Ni(PMe 3 )] 4 and PhI in THF was sensible (ΔG ‡ = 15.6 kcal mol -1 ), barriers for PhBr and in toluene and hexane were far too high to account for the Ni I products observed, albeit in reduced quantities, in these reactions (see Table S2). The major contributor to these energies was the very large value of ΔG for this process in non-polar solvents.

(B) Inner-sphere electron transfer from [Ni(PMe 3 ) 3 ] to ArX
Inner-sphere electron transfer from [Ni(PMe 3 ) 3 ] to ArX could occur via complex [Ni(PMe 3 ) 3 (XAr)], with a concomitant change in multiplicity from singlet to triplet. This cannot be described by a simple transition state, and instead requires the geometry at which the singlet and triplet have the same energy to be located, termed the minimum energy crossing point (MECP).
The software tool published by Harvey was used, and the MECP could be located for a number of examples. However, the energies of MECPs for aryl bromides were too high to compete with oxidative addition to [Ni(PMe 3 ) 2 ] (see Table S3). Energies are quoted as a range, because these structures typically have imaginary frequencies when either or both of the singlet and triplet frequency calculations are performed; there is a small discrepancy in their free energies, because the MECP calculation seeks the geometry at which the potential energy E is the same for both singlet and triplet.  The dissociation energy of the second phosphine is always higher that the oxidative addition transition state from [Ni(PMe 3 ) 3 ], so this mechanism is unfavourable in the three cases. We could not find the barrier for PhI oxidative addition, due to the high reactivity of this species with Ni(P) 2 , which forms the Ni(II) product 9 directly.

MICROKINETIC MODEL (A) Reaction between [Ni(PMe 3 ) 4 ] to PhI in toluene
We used the COPASI program package to run the reactions used in the microkinetic model, with the following equations: The activation energies and the associated kinetic constants are in Table 4   Table S4. Calculated activation free energies and kinetic constants of kinetic model of the reaction between [Ni(PMe 3 ) 4 ] and PhI in toluene.
The conditions used to analyze the time course of the reaction were 0.01s in 20 steps, obtaining the results in Table S5. The initial concentrations were taken from the experiments (0.005 M for species S9 A and B and solvent concentration calculated from the density at 25ºC (12.2 M for THF). The numbers represent the concentration of labelled species in mol per litre (Table S5).

(B) Reaction between [Ni(PMe 3 ) 4 ] to PhBr in THF
We ran the same microkinetic model for the reaction between Ni complex and PhBr in THF. The reactions were: The activation energies and the associated kinetic constants are in Table S6:  ) 3.6 1.45e+10 4 1 3.6 1.45e+10 5 9.5 6.700e+5 The reactions 3 and 4 are also barrierless. We applied the same methodology that above, using the experimental value of 0.456 mPa·s for the viscosity of THF. 25 The associated activation free energy was calculated from Eyring equation and is 3.6 kcal/mol.