Iron pincer complexes as catalysts and intermediates in alkyl-aryl kumada coupling reactions

Article abstract of DOI:10.1021/om501122p

Iron-catalyzed alkyl-aryl Kumada coupling has developed into an efficient synthetic method, yet its mechanism remains vague. Here, we apply a bis(oxazolinylphenyl)amido pincer ligand (Bopa) to stabilize the catalytically active Fe center, resulting in isolation and characterization of well-defined iron complexes whose catalytic roles have been probed and confirmed. Reactivity studies of the iron complexes identify an Fe(II) "ate" complex, [Fe(Bopa-Ph)(Ph)₂]^-, as the active species for the oxidative addition of alkyl halide. Experiments using radical-probe substrates and DFT computations reveal a bimetallic and radical mechanism for the oxidative addition. The kinetics of the coupling of an alkyl iodide with PhMgCl suggests that formation of the "ate" complex, rather than oxidative addition, is the turnover-determining step. This work provides insights into iron-catalyzed cross-coupling reactions of alkyl halides.

Full text of DOI:10.1021/om501122p

Article
pubs.acs.org/Organometallics
Iron Pincer Complexes as Catalysts and Intermediates in Alkyl−Aryl
Kumada Coupling Reactions
Gerald Bauer,† Matthew D. Wodrich,‡ Rosario Scopelliti,† and Xile Hu*,†
‡
†Laboratory of Inorganic Synthesis and Catalysis and Laboratory for Computational Molecular Design, Institute of Chemical
́
Sciences and Engineering, Ecole Polytechnique Fed
Switzerland
́ ́
erale de Lausanne (EPFL), EPFL-ISIC-LSCI, BCH 3305, Lausanne CH 1015,
S
* Supporting Information
ABSTRACT: Iron-catalyzed alkyl−aryl Kumada coupling has
developed into an efficient synthetic method, yet its mechanism
remains vague. Here, we apply a bis(oxazolinylphenyl)amido
pincer ligand (Bopa) to stabilize the catalytically active Fe center,
resulting in isolation and characterization of well-defined iron
complexes whose catalytic roles have been probed and
confirmed. Reactivity studies of the iron complexes identify an
Fe(II) “ate” complex, [Fe(Bopa-Ph)(Ph)2]−, as the active species
for the oxidative addition of alkyl halide. Experiments using
radical-probe substrates and DFT computations reveal a
bimetallic and radical mechanism for the oxidative addition. The kinetics of the coupling of an alkyl iodide with PhMgCl
suggests that formation of the “ate” complex, rather than oxidative addition, is the turnover-determining step. This work provides
insights into iron-catalyzed cross-coupling reactions of alkyl halides.
between 1-bromooctane and (mesityl)MgBr under conditions
similar to the FeCl3-TMEDA protocol originally developed by
Nakamura and co-workers.7 They proposed a catalytic cycle
where a TMEDA-chelated Fe(II) was the active species.17
However, Bedford and co-workers showed that the reaction’s
active species was the homoleptic “ate” complex [Fe(mesityl)3]−.
TMEDA seemed to trap intermediates in the off-cycle of the
catalysis and suppressed side reactions.18 Bedford and co-
workers also isolated defined Fe(I) phosphine complexes that
were competent precatalysts in related Fe-catalyzed alkyl−aryl
INTRODUCTION
■
Iron-catalyzed cross-coupling of alkyl halides with aryl Grignard
reagents is among the most attractive methods for alkyl-aryl
coupling for a number of reasons. (1) Iron is inexpensive,
abundant, and nontoxic. (2) The iron-catalyzed alkyl−aryl
Kumada coupling generally has a high yield, a short reaction time
(within several hours), and a high functional group tolerance. (3)
The use of alkyl electrophiles as alkylating reagents allows the
coupling of functionalized alkyl groups.1−5 Whereas significant
progress has been made in developing new catalysts, reaction
conditions, and scope,6−14 the mechanism of this and related
iron-catalyzed coupling reactions of alkyl halides is only now
Negishi coupling,19,20 while Car
́
denas and co-workers obtained
EPR-based evidence for the involvement of Fe(I) carbene
species in alkyl−alkyl Kumada coupling.21
being unveiled. Furstner and co-workers showed that Fe “ate”
̈
A challenge in the mechanistic study of iron-catalyzed alkyl−
aryl Kumada coupling is the coordinative lability of ligands on Fe,
especially when the latter undergoes oxidation and spin state
changes. It is reported that TMEDA dissociates from the Fe(II)
center during alkyl−aryl Kumada coupling,18 and even a
bidentate phosphine ligand can leave an Fe ion during alkyl−
aryl Negishi coupling.20 Therefore, extrapolation of active iron
species from complexes isolated from the catalysis mixture or
defined precatalysts might be error-prone without considering
ligand dissociation. A strong chelating ligand can alleviate this
complication. Herein we employ a bis(oxazolinylphenyl)amido
pincer ligand to support the catalytically active Fe center. The
rigid tridentate chelate allows for the isolation and character-
ization of several intermediate species, whose catalytic roles are
complexes, in which the formal oxidation states of Fe range from
−2 to +2, might be catalytically active;8,15 treatment of FeCl3
with a large excess of aryl Grignard reagents could lead to Fe(0)
“ate” species.8 Further, they showed that a formal Fe(−2)
complex, [(Li(TMEDA))2Fe(C2H4)4], was a highly active and
selective catalyst for alkyl−aryl Kumada coupling.6 The low-
valent iron “ate” species would be difficult to generate in most in
situ catalyst systems where reduction of the initial Fe(III) salt to
Fe(0) or below by a Grignard reagent without β-hydrogen is
difficult. In these cases, Fe(I) and Fe(II) “ate” complexes are
more likely present. Neidig and co-workers reported that the
reaction of FeCl3 with MeMgBr in THF gave the homolytic
tetramethyliron(III) complex [MgCl(THF)5][FeMe4], which
then decomposed to give a Fe(I) species.16
Nagashima and co-workers observed the formation of
(TMEDA)Fe(mesityl)2 and (TMEDA)Fe(mesityl)Br
(TMEDA = tetramethylethylenediamine) from the coupling
Received: November 6, 2014
Published: December 17, 2014
© 2014 American Chemical Society
289
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
probed. Subsequent experiments using radical-probe substrates,
kinetic measurements, and DFT computations establish a
catalytic cycle in which an Fe(II) aryl “ate” complex activates
alkyl halide via a bimetallic oxidative addition pathway.
RESULTS
■
Synthesis, Characterization, and Reactivity of Cata-
lysts and Intermediates. The Fe(III) complex [Fe(Bopa-
Ph)Cl2] (1; Figure 1) was previously applied to iron-catalyzed
Figure 2. Molecular structure of complex 2. Hydrogen atoms are
omitted for clarity. The thermal ellipsoids are displayed at the 50%
probability level. Selected bond lengths (Å) and angles (deg): Fe(1)−
N(1), 2.114(2); Fe(1)−N(2), 2.1256(16); Fe(1)−O(2), 2.2906(15);
Fe(1)−Cl(1), 2.4152(9); N(1)−Fe(1)−N(2), 85.41(5); N(2)−
Fe(1)−N(2), 170.81(9); N(1)−Fe(1)−O(2), 91.03(5); N(2)−
Fe(1)−O(2), 86.32(7); N(1)−Fe(1)−Cl(1), 180.00(6), N(2)−
Fe(1)−Cl(1), 94.59(5).
Figure 1. Alkyl−aryl coupling reactions catalyzed by complex 1. Isolated
yields are reported. Legend: (a) reaction at −40 °C; (b) reaction at
room temperature.
configuration. The complex has a solution magnetic moment of
5.11 μB,24 consistent with an Fe(II) complex in a high-spin state.
Concomitant with the reduction of 1 to 2, 0.5 equiv of
biphenyl was formed, indicating that the phenyl anion in
PhMgCl was the electron donor in this process. Complex 2 is
unreactive toward (3-iodobutyl)benzene (3), suggesting that 2
requires further transformation during the catalytic process.
Indeed, further addition of PhMgCl to a solution of 2 shows an
immediate reaction that does not produce additional biphenyl,
suggesting that the Fe center is not further reduced. The
reactions of 2 with ArMgCl (Ar = Ph, o-Tol) gave [Fe(Bopa-
Ph)Ar] (4, Ar = Ph; 5, Ar = o-Tol). The crystal of 5 was obtained
in a THF−dioxane mixture, and its structure revealed the four-
coordinate nature of the complex in the solid state (Figure 3),
where the Fe ion adopts a tetrahedral geometry. While a crystal
structure of 4 could not be obtained, its composition was
confirmed by elemental analysis. Complexes 4 and 5 have
solution magnetic moments of 5.00 and 4.76 μB, respectively,24
consistent with each complex having a high-spin Fe(II) center. In
the presence of 1 equiv of PhMgCl, 4 decomposed with a half-life
of about 10 min at room temperature. The solution changed
from deep red to a turbid brown, and the NMR spectrum of the
solution suggested the formation of a Mg−Bopa−Ph adduct. No
biphenyl was observed. This evidence indicates that, in the
absence of alkyl halide, 4 may be decomposed by an excess
amount of PhMgCl via iron demetalation.24 However, this
decomposition is too slow to be catalytically relevant, except at
the end of the catalysis, where the iron complex has been
demetalated to form a presumable Mg−Bopa species.
hydrosilylation reactions.22,23 We believed that the pincer Bopa
ligand might also be suitable for use in iron-catalyzed cross-
coupling reactions. The bis(aryl)amido pincer backbone
provides both structural rigidity and chemical stability; the two
oxazolinyl donors can accommodate both Fe(III) and Fe(II) and
perhaps, for a short time, Fe(I) due to possible π back-bonding.
Complex 1 was first tested for the coupling of alkyl halides with
PhMgCl (Figure 1). The coupling proceeded smoothly at both
−40 °C and room temperature without needing an amine
additive. Both primary and secondary alkyl halides could be
coupled in high isolated yields, while either alkyl bromide or
iodide may serve as a suitable electrophile. Thus, complex 1 is a
competent precatalyst for iron-catalyzed alkyl−aryl Kumada
coupling. Like many other catalysts, the coupling of alkyl
chlorides and tertiary alkyl halides was inefficient.
Complex 1 served as the entry point of our mechanistic study.
The transformation of 1 by an aryl Grignard reagent was then
examined. Addition of 1 equiv of PhMgCl to a THF solution of 1
at room temperature produced [Fe(Bopa-Ph)Cl(THF)2] (2)
(Scheme 1). The complex could, alternatively, be independently
synthesized by lithiating the pincer ligand followed by a
metathesis reaction with FeCl2(THF)1.5. The crystal structure
of 2 (Figure 2) reveals an octahedrally coordinated Fe(II) center,
with the pincer Bopa ligand adapting the expected meridional
Scheme 1. Synthesis of Complexes 2, 4, and 5
To analyze the reactivity and the catalytic relevance of complex
4, this complex was reacted with 20 equiv of (3-iodobutyl)-
benzene (3) at −40 °C in a THF solution to give the C−C
coupled product 1,3-diphenylbutane (7) (Table 1). However,
the reaction was slow in comparison to the catalysis, as 50%
conversion (t1/2) required 40 min (entry 1, Table 1), whereas the
catalysis was generally complete within several minutes. The
difference in reaction rates suggests that 4 is not the active species
to activate alkyl halides in the catalytic reaction. In the presence
290
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
tane (7) and 3-(2-methylphenyl)-1-phenylbutane (8) were
produced. When the ratio of 4 to o-TolMgCl was increased,
the ratio of 7 to 8 increased in a similar manner (Table 2).
Second, the reactions of complex 5 with 3 in the presence of
PhMgCl were conducted,24 yielding coupling products contain-
ing both Ph and o-Tol groups. These results are more consistent
with the active species being the “ate” complex [Fe(Bopa-
Ph)Ar2]− (6), where the two aryl groups behave similarly in
catalysis.
Mechanism of Cross-Coupling. After the transmetalation
step (2 to 4) and the activation step (4 to 6) were examined, the
oxidative addition of alkyl halide was probed. The results of
radical clock experiments indicate that the oxidative addition
produces alkyl radicals (Scheme 2). For example, the coupling of
Scheme 2. Evidence for a Radical Pathway in the Activation of
Alkyl Halide
Figure 3. Molecular structure of complex 5. Hydrogen atoms and a
cocrystallized 1,4-dioxane molecule are omitted for clarity. The thermal
ellipsoids are displayed at the 50% probability level. Selected bond
lengths (Å) and angles (deg): Fe(1)−N(2), 2.032(4); Fe(1)−N(1),
2.033(4); Fe(1)−C(31), 2.053(5); Fe(1)−N(3), 2.083(4); N(2)−
Fe(1)−N(1), 90.25(15); N(2)−Fe(1)−C(31), 127.37(18); N(1)−
Fe(1)−C(31), 110.88(16); N(2)−Fe(1)−N(3), 87.65(15); N(1)−
Fe(1)−N(3), 114.31(15); C(31)−Fe(1)−N(3), 121.52(18).
Table 1. Influence of Additive on the Reaction of Complex 4
with 3
enantiomerically enriched 3-bromo-1-phenylbutane (9) led to
the racemic product (10) while coupling of (bromomethyl)-
cyclopropane (11) gave almost exclusively the ring-opened
product (12-L).
entry
additive
t1/2
We propose that the first step of oxidative addition is a single-
electron transfer between an iron(II) bis(aryl)-ate species,
represented here by 6, and an alkyl halide to give an Fe(III)
halide complex (13) and an alkyl radical (Scheme 3). The alkyl
radical then may engage in one of the three possible pathways to
give the coupling product: cage rebound, escape rebound, or
bimetallic oxidative addition.25 In the cage-rebound pathway
(pathway A, Scheme 3), the alkyl radical stays in the solvent cage
and recombines with 13 to give [FeIV(Ph)(alkyl)(X)] (14, X =
halide), which, upon reductive elimination, gives the coupling
product and regenerates complex 2 or its analogue (2-X). In the
escape-rebound pathway (pathway B, Scheme 3), the alkyl
radical first leaves the solvent cage and then re-enters it to
combine with 13 to give complex 14. Reductive elimination from
14 then yields the coupling product. In the bimetallic oxidative
addition pathway (pathway C, Scheme 3), the alkyl radical leaves
the solvent cage and combines with another molecule of Fe(II)
phenyl complex 4 to form a [FeIII(Ph)(alkyl)] complex (15).
Reductive elimination from 15 gives the coupling product and an
Fe(I) complex (16), which reacts with the Fe(III) complex 13 to
give Fe(II) complexes 2 and 4. In our considerations we neglect
the possibility that C−C coupling occurs via attack of free radical
on an Fe aryl species; Norbby and co-workers reported a
competitive Hammett study that ruled out radicals in the
coupling step.26
1
2
3
none
40 min
<15 s
<15 s
PhMgCl
PhLi
of PhMgCl and PhLi, the reaction of 4 with 3 was greatly
accelerated. The t1/2 value was less than 15 s (entries 2 and 3,
Table 1), which was comparable to the rate of the catalytic
coupling reaction. This result suggests that 4 reacted with
PhMgCl to form either an associated species, [Fe(Bopa-
Ph)Ph](PhMgCl), or an “ate” complex, [Fe(Bopa-Ph)Ph2]−,
as the catalytically active species.
To probe whether an associated species or an “ate” complex is
involved, two crossover experiments were performed. First, the
reaction of 4 with 3 in the presence of o-TolMgCl was examined
in THF at room temperature (Table 2). Both 1,3-diphenylbu-
Table 2. Influence of o-TolMgCl on the Reaction of Complex
4 with 3
amt of o-TolMgCl 4/o-TolMgCl
yield (%) 7/8
(conversn of 3)
7/8
entry
(equiv)
ratio
ratio
The reaction outcomes of the coupling of radical-probe
substrates can be used to distinguish the cage-rebound pathway
from the escape-rebound and bimetallic oxidative addition
pathways.25,27−30 In the current case, the coupling of 1-bromo-5-
hexene (17) with PhMgCl in the presence of 1 was employed for
1
2
3
4
1.0
0.8
0.6
0.4
1.0
1.3
1.7
2.5
37/45 (92)
32/33 (73)
33/30 (71)
33/19 (60)
0.8
1.0
1.1
1.7
291
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
Scheme 3. Three Possible Pathways for the Oxidative Addition of Alkyl Halide and the Consequent Reductive Elimination
this purpose (Scheme 4). Activation of 17 first gives the alkyl
radical (18) that can combine with an Fe-Ph species to give the
Scheme 4. Coupling of a Radical-Clock Substrate To
Differentiate the Pathways for Oxidative Addition and C−C
Bond Formation
linear coupling product 19. Radical 18 can also undergo an
intramolecular ring-closing rearrangement (k2 ≈ 105 s−1) to give
Figure 4. Ratio of 19/20 in the coupling of 17 with PhMgCl as a
the cyclized radical 18′.31 Combination of 18′ with an Fe-Ph
function of catalyst loading. The results are averaged over two
independent runs.
intermediate then gives the cyclized coupling product 20. The
ratio of 19 and 20 is a function of r1 and r2. For all scenarios r2 =
Scheme 5. Reaction of 4 with an in Situ Generated Alkyl
Radical
k2[18]. However, r1 depends upon the reaction pathway. In the
cage-rebound pathway, the recombination is considered an
intramolecular reaction and is zeroth order on the iron catalyst.
Hence, r1 = k1[18], and r1/r2 is independent of the concentration
of the iron catalyst. It follows that the ratio of 19/20 is
independent of catalyst loading. On the other hand, in the
bimetallic oxidative addition and escape-rebound pathways, the
combination of an alkyl radical with an iron ion is an
intermolecular reaction and is first order in the catalyst.
Therefore, r1 = k1[cat.][18], and the 19/20 ratio is first order
in catalyst loading.
Figure 4 shows that the 19/20 ratio is linearly dependent on
the loading of catalyst (1) in the coupling of 17 with PhMgCl.
This result eliminates the cage-rebound pathway; however, it is
consistent with either an escape-rebound or a bimetallic oxidative
addition mechanism.
4-phenylbutaneperoxoate (21).32 In the presence of 1 equiv of 4,
1,3-diphenylpropane was formed in 27% yield (relative to 4).
The formation of the alkyl−aryl coupled product in this process,
in turn, indicates that the bimetallic oxidative addition pathway
proposed in Scheme 3 is probable, although it is not a proof. As
the Fe(III) halide complex 13 cannot be isolated, an analogous
reactivity test cannot be conducted to probe the feasibility of the
escape-rebound mechanism.
To elucidate additional mechanistic details, we employed
To probe the feasibility of the bimetallic oxidative addition, the
reaction of 4 with an in situ formed alkyl radical was examined
(Scheme 5). If this mechanism is operating, the combination of 4
with an alkyl radical will lead to an alkyl−aryl coupling product,
as predicted by path C in Scheme 3. In the experiment a
phenylpropyl radical was generated by the photolysis of tert-butyl
density functional theory computations at the PBE033,34
-
dDsC35−38/TZ2P//M0639,40/def2-SVP level to determine the
reaction free energy profiles in implicit THF solvent (using the
COSMO-RS41 solvation model) of the bimetallic oxidative
addition and escape-rebound pathways (see the Supporting
Information for computational details).24 Owing to the
292
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
Figure 5. Reaction free energy profile and relevant structures for the bimetallic oxidative addition pathway computed at the unrestricted PBE0-dDsC/
TZ2P//M06/def2-SVP level (including THF implicit solvation using the COSMO-RS model).
extremely flat nature of the full catalyst potential energy surface
(PES), our computations employed a model of the Fe catalyst in
which some methyl groups replaced some phenyl groups. In the
model, the aryl nucleophile is represented using a tolyl moiety
while the isopropyl radical represents the alkyl group.
Figure 5 illustrates the bimetallic oxidative addition pathway.
Beginning from the reactant complex (4+iPr•), the isopropyl
radical first associates with the Fe(II) center by overcoming a
small transition state barrier, TS4,15 (+6.3 kcal/mol), thereby
forming a Fe(III) species characterized by a high-spin sextet,
Figure 6. Relative electronic energies (at the M06/def2-SVP level) of
15Sextet, in which both the aryl and alkyl groups are bound to the
the sextet (blue), quartet (red), and minimum energy crossing point
iron center. Initial computations on the sextet PES revealed the
(MECP, black) of 15. Values are given in kcal/mol.
reductive elimination step leading to final product formation to
be energetically inaccessible (e.g., >60 kcal/mol at the M06/
def2-SVP level). As a result, we examined the possibility that the
final product-forming steps occur on the PES of the intermediate
spin quartet state rather than sextet PES, as the energies of
(3.060 Å for 15Quartet vs 2.707 Å for 15′Quartet). From 15′Quartet
reductive elimination forms the final product (16). The highest
point on the bimetallic oxidative addition PES corresponds to the
reductive elimination TS located 10.8 kcal/mol above the
reactant complex.
The escape-rebound pathway represents a second possible
mechanism for forming the coupled product. Moving from the
high-spin quintet Fe(III) reactant complex (13+iPr•) to the TS
corresponding to association of the alkyl radical requires 10.6
kcal/mol of energy (Figure 7). The Fe(IV) intermediate 14 in
which the alkyl, aryl, and halogen groups are bound is roughly
isoenergetic with TS13,14 (TS13,14 → 14 −0.65 kcal/mol at the
15Sextet and 15Quartet lie within 4 kcal/mol (+6.6 kcal/mol at the
M06/def2-SVP level). Indeed, a minimum energy crossing point
(MECP) between the high (sextet) and intermediate (quartet)
spin states was located ∼11 kcal/mol above 15Sextet and ∼4 kcal/
mol above 15Quartet (Figure 6). Proceeding along the reaction
coordinate, 15Quartet adopts a more stable conformer, 15′Quartet
,
in which the carbon atoms of the aryl and alkyl groups that
ultimately form the new C−C bond are in closer proximity
293
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
Figure 7. Reaction free energy profile and relevant structures for the escape-rebound pathway computed at the unrestricted PBE0-dDsC/TZ2P//M06/
def2-SVP level (including THF implicit solvation using the COSMO-RS model).
M06/def2-SVP level). Product formation via reductive elimi-
nation requires an addition 3.4 kcal/mol of energy to overcome
the TS barrier (14 → TS14,2X). Given that the highest point on
this pathway lies higher in energy than proceeding through the
bimetallic oxidative addition route (14.3 vs 10.8 kcal/mol), the
DFT computations indicate that bimetallic oxidative addition is
the favored mechanistic pathway. Although the energy difference
is small given the uncertainty of DFT calculations, because the
concentration of Fe(II) species 4 is much higher than that of the
Fe(III) species 13 during catalysis, the bimetallic oxidative
addition should be the dominant reaction pathway.
Kinetics. The kinetics of the Fe-catalyzed alkyl−aryl coupling
was determined using complex 2 as the catalyst. Complex 1 was
not suitable for the kinetic studies, since the reduction from 1 to 2
is slow at −84 °C and leads to an induction period. The use of
complex 2 is justified because 1 and 2 have the same efficiency for
the reactions shown in Figure 1. The coupling reaction rates
between (3-iodobutyl)benzene (3) and PhMgCl (5 mol % of 2 as
catalyst) were measured using the initial rate approximation.24
Figure 8A,B shows the dependence of the reaction rate on the
concentrations of PhMgCl and catalyst, respectively. The data
approximately fit as being first order in Grignard reagent and
second order in catalyst. Only a small and random change in the
reaction rate was observed when the substrate (3) concentration
was varied (Figure 8C). This result suggested that the reaction is
zeroth order in alkyl iodide. To further confirm this, the reaction
profile of a given catalytic run was evaluated by the integrated rate
law (in the range of up to 77% conversion). The conversion of
the substrate could be fit with a first-order decay.24 Assuming a
constant concentration of catalyst, the result agrees with being
first order in Grignard and zeroth order in substrate. The
coupling of 3 and PhMgCl was also monitored by UV−vis
spectroscopy. Although the spectral change might be consistent
with the resting state of the catalyst being the Fe(II) phenyl
complex 4, definitive assignment is difficult due to overlap of
absorption bands.24
DISCUSSION
■
On the basis of the results described in the previous section, a
catalytic cycle can be proposed (Scheme 6). The Fe(III)
precatalyst [Fe(Bopa-Ph)Cl2] (1) is first reduced by PhMgCl to
form the Fe(II) catalyst [Fe(Bopa-Ph)Cl(THF)2] (2). Trans-
metalation of complex 2 with PhMgCl gives [Fe(Bopa-Ph)Ph]
(4), which is proposed as the resting state of the catalyst. Species
4 is activated by another molecule of PhMgCl to give the “ate”
complex [Fe(Bopa-Ph)(Ph)2]− (6). Furstner and co-workers
̈
showed that FeCl3 reacted with an excess of MeLi in ether to give
the analogous Fe(II) ate complex [(Me4Fe)(MeLi)](Li*Et2O)2,
which is capable of methylation of vinyl and acyl electrophiles.15
Neidig and co-workers showed that FeCl3 reacted with MeMgBr
in THF to give the Fe(III) ate complex [MgCl(THF)5][FeMe4],
which reductively eliminated ethane, supporting the catalytic role
of ate complexes.16 While the exact structure and composition
remain unclear, the two Ph groups in complex 6 appear
equivalent in reactivity. Species 6 reacts much faster than 4 with
alkyl halide; therefore, it is the relevant active species for
oxidative addition. Reaction of 6 with alkyl halide gives an alkyl
radical and an Fe(III) complex, [Fe(Bopa-Ph)(Ph)(X)] (13, X =
halide). The alkyl radical escapes the solvent cage and
recombines with another molecule of a Fe(II) aryl complex, i.e.
4 (it could be 6 as well), to give [Fe(Bopa-Ph)(Ph)(Alkyl)] (15).
Reductive elimination from 15 gives the alkyl−aryl coupled
product and the Fe(I) species [Fe(Bopa-Ph)] (16). Species 16
should be unstable and quickly react with the unstable Fe(III)
complex 13, to give the two Fe(II) complexes 2 and 4, which can
re-enter the catalytic cycle.
Several deviations from this catalytic cycle could be proposed.
For example, the alkyl radical might combine with species 6
rather than 4 to an Fe(III) species analogous to 15. This variation
does not significantly alter the overall catalytic cycle. On the
other hand, species 16 might react with an alkyl halide to
generate an alkyl radical and the Fe(II) species 2, which would
result in a different catalytic cycle where the transformation of 4
294
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
Furthermore, [Fe2(mesityl)4] reacts with (mesityl)MgBr to form
the tris(mesityl)ferrate (20), which could activate the alkyl
halide.18
The mechanism described here might be compared with those
recently proposed for Fe-catalyzed, TMEDA-assisted alkyl−aryl
Kumada coupling.18 When a bulky aryl Grignard reagent such as
mesitylMgBr is used, homoleptic [Fe(mesityl)3]− is the catalyst
and TMEDA does not serve as a ligand. If less bulky aryl
Grignard reagents are used, then the true catalyst is unknown,
although there is EPR evidence for an Fe(I) species in the
reaction mixture. In the present system, ligated Fe(II) complexes
are genuine catalysts for the coupling. This is due to the strong
chelating ability of the tridentate pincer ligand Bopa which, in
turn, facilitates the mechanistic study. As in the Fe-TMEDA
system for the coupling of less bulky Grignard reagent, the Fe(I)
species 16 is involved; however, this Fe(I) species is an unstable
intermediate rather than a resting species in the former. Stable
Fe(I) species have been shown as catalysts in Fe-catalyzed alkyl−
aryl Negishi coupling19,20 and alkyl−alkyl Kumada coupling.21
These Fe(I) species are supported by soft, neutral donors such as
phosphine and N-heterocyclic carbene ligands, which match well
with the low-valent Fe(I) center. In contrast, the nitrogen-based
anionic Bopa ligand renders the Fe(I) center highly reactive such
that species 16 cannot be observed or isolated.
The mechanism of this Fe catalysis might also be compared
with that of alkyl−alkyl Kumada coupling catalyzed by the nickel
pincer complex Nickamine.25 There are a number of similarities.
In both systems, transmetalation precedes oxidative addition,
and the oxidative addition follows a bimetallic radical pathway.
Furthermore, stable organometallic intermediates (Fe(II) Ar or
Ni(II) alkyl) each need to be activated by one molecule of
Grignard reagent to activate alkyl halide. It is tempting to
consider these as common features in Kumada coupling
reactions catalyzed by ligated first-row transition metals,
although more systems should be investigated. The Fe and Ni
systems exhibit significant differences as well. There is a
remarkable and surprising change of coordination geometries
of the Fe intermediates in the solid state. The Fe(III) halide
precatalyst is five-coordinate, the Fe(II) halide catalyst is six-
coordinate with two solvent ligands, and the Fe(II) aryl catalyst is
tetrahedral four-coordinate. It, however, cannot be excluded that
in the solution the Fe(II) species will bind additional solvent
molecules (e.g., THF) to form an octahedral geometry. In
contrast, the Ni catalysts remain Ni(II) with a square-planar
geometry. An additional difference is the reactivity of the active
species [Fe(Bopa-Ph)(Ph)2]MgCl and [Ni(N2N)(alkyl)]-
((alkyl)MgCl). In the former, the two phenyl groups are coupled
in similar probability with alkyl halide and it can be treated as a
genuine “ate” complex. In the latter, the original alkyl group is
coupled preferentially with alkyl halide; thus, the second
molecule of (alkyl)MgCl is only weakly associated.
Figure 8. Influence of the concentrations of Grignard reagent (A),
catalyst (B), and alkyl iodide (C) in the initial rates of the coupling of 3
with PhMgCl. The slope of log (rate) vs log (reagent) rounded to an
integral is the approximate rate order in (A) and (B). The data are
averaged over a minimum of three independent runs.
to 6 serves only as the initiation step. However, this possibility is
incompatible with 4 being the resting state and the catalysis being
first order in Grignard reagent. Furthermore, it is inconsistent
with the reaction profile of the catalysis that shows no change of
kinetic behavior over the reaction course.
The kinetics of the catalysis, i.e. second order in catalyst, first
order in Grignard, and zeroth order in alkyl iodide, suggests that
transmetalation, but not oxidative addition, is the turnover-
determining step. The second order in catalyst suggests that the
transformation of 4 to 6 goes through a bimetallic intermediate; a
possibility is shown in Scheme 7. Dimeric Fe(II) μ-aryl
complexes such as [Fe2(mesityl)4] (17; Scheme 8),42,43
[Fe2(2,4,6-iPr3C6H2)4] (18),43 and [{Ar*Fe(μ-Ph)}2] (19, Ar*
= C6H3-2,6-(C6H2-2,4,6-iPr3)2)44 have previously been reported.
CONCLUSION
■
By using the strongly chelating pincer ligand Bopa, we have been
able to prepare and isolate ligated Fe complexes which are
genuine catalysts and intermediates in Fe-catalyzed alkyl−aryl
Kumada coupling reactions. The easily accessible Fe(III)
dichloride precatalyst is first reduced by the Grignard reagent
to form the Fe(II) halide catalyst, which is transmetalated to form
the Fe(II) aryl catalyst. The structural elucidation of the Fe
catalysts reveals remarkable geometric changes of the Fe species.
Although the Fe(II) aryl catalyst activates alkyl halide to form the
alkyl−aryl coupling product, this reaction is too slow to be
295
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
a
Scheme 6. Proposed Catalytic Cycle for the Coupling of Alkyl Halide with PhMgCl
a
The X ligand in species 2 and 13 is identical with the X group in alkyl-X. The pincer Bopa-Ph is simplified for clarity.
mechanism described here is only confirmed for the Fe-Bopa
pincer system. However, this comprehensive mechanistic study
using well-defined Fe pincer catalysts should provide significant
new insights into the general understanding of Fe-catalyzed
coupling reactions of alkyl halides.
Scheme 7. Hypothetical Mechanism for the Transformation
of 4 to 6
EXPERIMENTAL SECTION
■
General Method. All manipulations were carried out under an inert
N2 atmosphere using standard Schlenk or glovebox techniques. The
solvents were purified and dried using a two-column solid-state
purification system. They were transferred to the glovebox in a Strauss
flask without exposure to air. The solvents were stored over molecular
sieves (3 Å). Deuterated solvents were purchased from Cambridge
Isotope Laboratories, Inc., and were degassed and stored over dried and
activated molecular sieves (3 Å). All other chemicals were purchased
from commercial sources and were degassed by standard freeze−
pump−thaw procedures prior to use. The Grignard reagents were
titrated prior to every use following the literature procedure.45 NMR
spectra were recorded on a 400 1H NMR chemical shifts were
referenced to the residual solvent peak as determined relative to TMS (δ
0 ppm). GC measurements were conducted on a GC instrument
equipped with a FID detector. GC-MS measurements were conducted
on a GC instrument equipped with MS and an FID detector.
Photochemical experiments were performed with a Rayonet Photo-
chemical Reactor, using Rayonet Photochemical Reactor Lamp of
wavelength 2537 Å for homogenic irradiation of the samples in the
photochemical reactor. The internal temperature was maintained within
a 40−60 °C range with the aid of an integrated mechanical ventilation
system. UV/vis absorption spectra were recorded with a UV−vis fiber
optic probe connected to a UV−vis spectrometer.
Synthesis. The following chemicals were synthesized according to
the literature methods: 2,2′-iminodibenzoic acid,46 (R)-(−)-phenyl-
glycinol,47 2,2′-iminodibenzoyl chloride, Bopa-Ph, (Bopa-Ph)Li,48
[Fe(Bopa-Ph)Cl2] (1),23 (R)-(3-bromobutyl)benzene,49 and tert-butyl
4-phenylbutaneperoxoate.32
Synthesis of [Fe(Bopa-Ph)Cl(THF)2] (2). (Bopa-Ph)Li (2.0 g, 1.1
equiv) was dissolved in 10 mL of THF, and FeCl2(THF)1.5 (930 mg, 1.0
equiv) was added. The solution was stirred overnight. The solvent was
evaporated, and the complex was redissolved in toluene and filtered over
Celite. The filtrate was concentrated and precipitated with pentane. The
solid was filtered off and recrystallized from THF/pentane, giving red
octahedral crystals. The crystals for the X-ray analysis were obtained by
diffusing pentane into a THF solution of 2. Yield: 2.6 g (93%), red
Scheme 8. Structure and Reactivity of Some Fe(II) μ-Aryl
Complexes
catalytically relevant. Instead, the Fe(II) aryl catalyst is further
transmetalated to form an Fe(II) bis(aryl) “ate” complex, which
is the true active species for oxidative addition of alkyl halide.
This oxidative addition proceeds via a bimetallic and radical
pathway in which two Fe(II) aryl species provide one electron
each. The kinetics of the coupling of an alkyl iodide with PhMgCl
shows a process that is second order in catalyst, first order in
Grignard reagent, and zeroth order in alkyl iodide. The turnover-
determining step is transmetalation of Fe(II) aryl catalyst to form
the active “ate” complex, which seems to proceed via a bimetallic
intermediate. The mechanisms of Fe-catalyzed coupling
reactions are likely ligand dependent, and the particular
296
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
octahedral crystals. Anal. Calcd for C38H40ClFeN3O2: C, 65.76; H, 5.81;
N, 6.05. Found: C, 65.34; H, 5.78; N, 6.04.
of THF (c = 12.6 mM). Inside the glovebox screw vials with a stirring bar
were filled with 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, and 1.0 mL of sol A, and
then 1.0 mL of sol B was added. The vials were filled up with THF to a
total volume of 2.0 mL. The vials were closed with a rubber septum. A
0.5 mL portion of sol C was put in 1.0 mL insulin syringes (the tip of the
needle was put in a rubber stopper to minimize the exposure to air). The
vials were taken out of the glovebox and attached to the Schlenk line by
piercing a needle through the septum. The following procedure was
done consecutively with every reaction vial. The vial was put in the ethyl
acetate slurry, and the contents were stirred for about 5 min (to be sure
that the temperature is constant). Then the rubber septum was removed
from the vial (while maintaining the nitrogen flow). sol C was added at
once. A 100 μL aliquot was taken at regular intervals (depending on the
concentration of Grignard reagent and hence its reaction rate) and
immediately pipetted into a GC vial containing 50 μL of acetonitrile.
The GC vials were then filled with diethyl ether and analyzed by GC (a
FI detector was used for quantification). The yields of 1,3-
diphenylbutane were determined with respect to naphthalene as an
internal standard. In order to determine the reaction rate, the data points
(up to 10% yield) were fitted to a linear plot. The reaction rates were
then logarithmized, averaged, and plotted versus the logarithm of the
PhMgCl concentration.
Computational Details. Geometries of all species were optimized
in the gas phase at the unrestricted M0639,40/def2-SVP level using the
“Ultrafine” grid in Gaussian09.50 The relative energetics of the various
spin states of the Fe complexes 4, 13, 14, and 15 were confirmed from
computations using both the M06 and OPBE functionals.33,51 The latter
functional features OPTZ exchange, which assists in the accurate
reproduction of energies of inorganic complexes with different spin
states.52−54 The M06/def2-SVP geometries of relevant compounds
were then recomputed as single-point energies using a density-
dependent dispersion correction35−38 appended to the PBE033,34
functional (PBE0-dDsC) with the triple-ζ Slater-type orbital TZ2P
basis set in ADF.55,56 Solvation corrections (in THF) employed the
continuum solvent model for realistic solvents41 (COSMO-RS), as
implemented in ADF. The minimum energy crossing point of 15 was
located using the “MECP Location Program” of Harvey.57 The
Supporting Information contains the computed Cartesian coordinates
of all the molecules reported in this study.
Synthesis of [Fe(Bopa-Ph)Ar] (4 and 5). Method A for 4 and 5. 2
(1.0 equiv) was dissolved in 5 mL of dioxane, and aryl Grignard (1.0
equiv) was added dropwise at room temperature. The reaction mixture
was stirred for 30 min. The stirring was stopped to precipitate the Mg
salts. The supernatant was slowly filtered over a pad of Celite. The
solvent was removed, and the residue was redissolved in a minimum
amount of toluene and filtered slowly over a pad of Celite. The resulting
solution was precipitated by adding pentane. The precipitate that
formed was filtered off and washed with pentane. Crystals of 5 were
obtained by overlayering a solution of 5 in THF/dioxane (1/1) with
pentane. Yield: 4, 75%; 5, 86%. Anal. Calcd for C37H31FeN3O2 (5) + 1.5
dioxane: C, 69.97; H, 5.87; N, 5.69. Found: C, 69.26; H, 5.75; N 5.66.
Method B for 4. 1 (798 mg, 1.0 equiv) was dissolved in 50 mL of THF
and 10 mL of dioxane was added (the addition of dioxane is crucial to
precipitate the formed MgCl2; otherwise, part of the complex starts to
decompose). A 1.40 mL portion of PhMgCl (1.83 M solution in THF,
1.9 equiv) was added dropwise over 1 h. The solids were filtered off over
a pad of Celite, and the solution was concentrated to about 20 mL and
then precipitated by adding pentane. The solid was filtered off and
redissolved in a minimum amount of benzene, filtered over Celite, and
precipitated by adding pentane. The solid was filtered off and thoroughly
washed with pentane. Yield of 4: 43%. Anal. Calcd for C36H29FeN3O2
(4) + 0.5 dioxane: C, 71.81; H, 5.23; N, 6.61. Found: C, 71.54; H, 5.55;
N, 6.42.
General Procedure for Coupling. Alkyl halide (0.5 mmol) was
dissolved in 3.0 mL of THF, and 1.0 mL of [Fe(Bopa-Ph)Cl2] stock
solution (25 mM) was added. The solution was brought to the
corresponding temperature (−40 °C or room temperature), and 0.33
mL of PhMgCl (1.86 M in THF) was added over a time period of 5 min.
The reaction mixture was stirred for another 10 min and was quenched
with water. The solution was acidified with HCl (1 M) and extracted
with 3 × 20 mL of CH2Cl2. The combined organic phases were dried
over Na2SO4, and the solvent was evaporated to dryness. The crude
product was purified by column chromatography (0−5% ethyl acetate in
hexane).
General Procedure for Radical-Probe Experiments. The
bromoalkane (0.25 mmol) and 1 (12.5 μmol) were weighed into a
vial and dissolved in THF (2.0 mL). PhMgCl in THF (0.30 mmol) was
added dropwise over a time period of 5 min at room temperature. After
addition the solution was stirred for another 10 min. The solution was
quenched with water and further acidified with HCl (1 M) and extracted
with 3 × 20 mL of CH2Cl2. The combined organic phases were dried
over Na2SO4, and the solvent was evaporated to dryness.
ASSOCIATED CONTENT
* Supporting Information
■
S
Text, figures, tables, and CIF and xyz files giving additional
experimental and computational details and characterization
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
Reaction of [Fe(Bopa-Ph)Ph] (4) with tert-Butyl 4-Phenyl-
butaneperoxoate under UV Irradiation. A solution of [Fe(Bopa-
Ph)Ph] (6.44 mM, 1.0 mL) and dodecane (internal standard) was put in
a J. Young NMR tube, and tert-butyl 4-phenylbutaneperoxoate (5.6 mg,
23.7 μmol) was added. The sample was put in a Rayonet Photochemical
reactor for 1.5 h. The reaction mixture was quenched with methanol,
and the coupling products were checked by GC/MS using dodecane as
an internal standard (an FI detector was used for quantification).
General Procedure for Kinetic Studies. The initial rate
approximation method was used, and the measurements were done
using catalyst 2. The determination of the order of Grignard reagent is
given as an example. In one experiment 8 reaction solutions were
prepared with variable Grignard concentrations. The reactions were
performed in a consecutive order to maintain the same reaction and
sampling conditions. For each reaction 10 GC samples were prepared.
To achieve a constant reaction temperature, a slurry of melting ethyl
acetate (mp −84 °C) was prepared before the experiment. The example
given below depicts one single experiment. In order to determine the
order of the reaction, the mean value of at least three independent
experiments was taken. Before the experiment three stock solutions
were prepared: (sol A) 1.0 mL of PhMgCl in THF (1.85 M) was diluted
to 5.0 mL with THF (c = 0.37 M); (sol B) (3-iodobutyl)benzene (289.6
mg, 1.11 mmol) and naphthalene (60.3 mg, 0.47 mmol) as an internal
standard were diluted to a 9.0 mL with THF (c = 0.124 M); (sol C)
[Fe(Bopa-Ph)Cl(THF)2] (43.8 mg, 63.1 μmol) was dissolved in 5.0 mL
AUTHOR INFORMATION
Corresponding Author
*E-mail for X.H.: xile.hu@epfl.ch.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by a starting grant from the European
Research Council under the European Community’s Seventh
Framework Programme (FP7 2007-2013)/ERC Grant agree-
ment no. 257096. M.D.W. acknowledges the BFE Curchod and
Prof. C. Corminboeuf (EPFL) for insightful discussions. The
Laboratory for Computational Molecular Design at EPFL is
acknowledged for providing computational resources.
REFERENCES
■
(1) Sherry, B. D.; Furstner, A. Acc. Chem. Res. 2008, 41, 1500−1511.
(2) Rudolph, A.; Lautens, M. Angew. Chem., Int. Ed. 2009, 48, 2656−
2670.
297
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298

Organometallics
Article
(3) Furstner, A.; Martin, R. Chem. Lett. 2005, 34, 624−629.
(4) Leitner, A. In Iron Catalysis in Organic Chemistry; Wiley-VCH:
Weinheim, Germany, 2008; pp 147−176.
(38) Steinmann, S. N.; Corminboeuf, C. Chimia 2011, 65, 240−244.
(39) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241.
(40) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.
(41) Klamt, A. WIREs Comput. Mol. Sci. 2011, 1, 699−709.
(42) Muller, H.; Seidel, W.; Gorls, H. J. Organomet. Chem. 1993, 445,
133−136.
(5) Bolm, C.; Legros, J.; Le Paih, J.; Zani, L. Chem. Rev. 2004, 104,
6217−6254.
(6) Martin, R.; Furstner, A. Angew. Chem., Int. Ed. 2004, 43, 3955−
3957.
(7) Nakamura, M.; Matsuo, K.; Ito, S.; Nakamura, E. J. Am. Chem. Soc.
(43) Klose, A.; Solari, E.; Floriani, C.; Chiesivilla, A.; Rizzoli, C.; Re, N.
J. Am. Chem. Soc. 1994, 116, 9123−9135.
(44) Ni, C. B.; Power, P. P. Organometallics 2009, 28, 6541−6545.
(45) Love, B. E.; Jones, E. G. J. Org. Chem. 1999, 64, 3755−3756.
(46) Paul, A.; Ladame, S. Org. Lett. 2009, 11, 4894−4897.
(47) McKennon, M. J.; Meyers, A. I.; Drauz, K.; Schwarm, M. J. Org.
Chem. 1993, 58, 3568−3571.
(48) Csok, Z.; Vechorkin, O.; Harkins, S. B.; Scopelliti, R.; Hu, X. L. J.
Am. Chem. Soc. 2008, 130, 8156−8157.
(49) Crosignani, S.; Nadal, B.; Li, Z. N.; Linclau, B. Chem. Commun.
2003, 260−261.
(50) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.;
Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.;
Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.;
Millam, M. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich,
S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J. Gaussian 09; Gaussian, Inc., Wallingford, CT, 2009.
(51) Handy, N. C.; Cohen, A. J. Mol. Phys. 2001, 99, 403−412.
(52) Conradie, J.; Ghosh, A. J. Phys. Chem. B 2007, 111, 12621−12624.
(53) Rotzinger, F. P. J. Chem. Theory Comput. 2009, 5, 1061−1067.
(54) Curchod, B. F. E.; Rotzinger, F. P. Inorg. Chem. 2011, 50, 8728−
8740.
2004, 126, 3686−3687.
(8) Furstner, A.; Martin, R.; Krause, H.; Seidel, G.; Goddard, R.;
Lehmann, C. W. J. Am. Chem. Soc. 2008, 130, 8773−8787.
(9) Ghorai, S. K.; Jin, M.; Hatakeyama, T.; Nakamura, M. Org. Lett.
2012, 14, 1066−1069.
(10) Nagano, T.; Hayashi, T. Org. Lett. 2004, 6, 1297−1299.
(11) Cahiez, G.; Habiak, V.; Duplais, C.; Moyeux, A. Angew. Chem., Int.
Ed. 2007, 46, 4364−4366.
(12) Czaplik, W. M.; Mayer, M.; Jacobi von Wangelin, A. Angew. Chem.,
Int. Ed. 2009, 48, 607−610.
(13) Bedford, R. B.; Betham, M.; Bruce, D. W.; Danopoulos, A. A.;
Frost, R. M.; Hird, M. J. Org. Chem. 2006, 71, 1104−1110.
(14) Bedford, R. B.; Betham, M.; Bruce, D. W.; Davis, S. A.; Frost, R.
M.; Hird, M. Chem. Commun. 2006, 1398−1400.
(15) Furstner, A.; Krause, H.; Lehmann, C. W. Angew. Chem., Int. Ed.
2006, 45, 440−444.
(16) Al-Afyouni, M. H.; Fillman, K. L.; Brennessel, W. W.; Neidig, M.
L. J. Am. Chem. Soc. 2014, 136, 15457−15460.
(17) Noda, D.; Sunada, Y.; Hatakeyama, T.; Nakamura, M.;
Nagashima, H. J. Am. Chem. Soc. 2009, 131, 6078−6079.
(18) Bedford, R. B.; Brenner, P. B.; Carter, E.; Cogswell, P. M.;
Haddow, M. F.; Harvey, J. N.; Murphy, D. M.; Nunn, J.; Woodall, C. H.
Angew. Chem., Int. Ed. 2014, 53, 1804−1808.
(19) Adams, C. J.; Bedford, R. B.; Carter, E.; Gower, N. J.; Haddow, M.
F.; Harvey, J. N.; Huwe, M.; Cartes, M. A.; Mansell, S. M.; Mendoza, C.;
Murphy, D. M.; Neeve, E. C.; Nunn, J. J. Am. Chem. Soc. 2012, 134,
10333−10336.
(20) Bedford, R. B.; Carter, E.; Cogswell, P. M.; Gower, N. J.; Haddow,
M. F.; Harvey, J. N.; Murphy, D. M.; Neeve, E. C.; Nunn, J. Angew.
Chem., Int. Ed. 2013, 52, 1285−1288.
(55) Guerra, C. F.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor.
Chem. Acc. 1998, 99, 391−403.
(21) Guisan-Ceinos, M.; Tato, F.; Bunuel, E.; Calle, P.; Cardenas, D. J.
Chem. Sci. 2013, 4, 1098−1104.
(56) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Guerra, C. F.; Van
Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001, 22,
931−967.
(22) Inagaki, T.; Ito, A.; Ito, J.; Nishiyama, H. Angew. Chem., Int. Ed.
2010, 49, 9384−9387.
(57) Harvey, J. N.; Aschi, M.; Schwarz, H.; Koch, W. Theor. Chem. Acc.
1998, 99, 95−99.
(23) Inagaki, T.; Phong, L. T.; Furuta, A.; Ito, J.; Nishiyama, H. Chem.
Eur. J. 2010, 16, 3090−3096.
(24) See the Supporting Information.
(25) Breitenfeld, J.; Ruiz, J.; Wodrich, M. D.; Hu, X. L. J. Am. Chem. Soc.
2013, 135, 12004−12012.
(26) Hedstrom, A.; Bollmann, U.; Bravidor, J.; Norrby, P.-O. Chem.
Eur. J. 2011, 17, 11991−11993.
(27) Kinney, R. J.; Jones, W. D.; Bergman, R. G. J. Am. Chem. Soc. 1978,
100, 635−637.
(28) Kinney, R. J.; Jones, W. D.; Bergman, R. G. J. Am. Chem. Soc. 1978,
100, 7902−7915.
(29) Biswas, S.; Weix, D. J. J. Am. Chem. Soc. 2013, 135, 16192−16197.
(30) Breitenfeld, J.; Wodrich, M. D.; Hu, X. L. Organometallics 2014,
33, 5708−5715.
(31) Fossey, J. S.; Lefort, D.; Sorba, J. Free radicals in organic chemistry;
Wiley: New York, 1995.
(32) Rueda-Becerril, M.; Sazepin, C. C.; Leung, J. C. T.; Okbinoglu, T.;
Kennepohl, P.; Paquin, J. F.; Sammis, G. M. J. Am. Chem. Soc. 2012, 134,
4026−4029.
(33) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77,
3865−3868.
(34) Adamo, C.; Barone, V. J. Chem. Phys. 1999, 110, 6158−6170.
(35) Steinmann, S. N.; Corminboeuf, C. J. Chem. Theory Comput.
2010, 6, 1990−2001.
(36) Steinmann, S. N.; Corminboeuf, C. J. Chem. Theory Comput.
2011, 7, 3567−3577.
(37) Steinmann, S. N.; Corminboeuf, C. J. Chem. Phys. 2011, 134.
298
dx.doi.org/10.1021/om501122p | Organometallics 2015, 34, 289−298