Tandem C-H activation/arylation catalyzed by low-valent iron complexes with bisiminopyridine ligands

Article abstract of DOI:10.1002/chem.201304459

Tandem C-H activation/arylation between unactivated arenes and aryl halides catalyzed by iron complexes that bear redox-active non-innocent bisiminopyridine ligands is reported. Similar reactions catalyzed by first-row transition metals have been shown to involve substrate-based aryl radicals, whereas our catalytic system likely involves ligand-centered radicals. Preliminary mechanistic investigations based on spectroscopic and reactivity studies, in conjunction with DFT calculations, led us to propose that the reaction could proceed through an inner-sphere C-H activation pathway, which is rarely observed in the case of iron complexes. This bielectronic noble-metal-like behavior could be sustained by the redox-active non-innocent bisiminopyridine ligands. A radical choice! A low-valent iron complex with non-innocent bisiminopyridine ligands performs C-H activation/arylation of unactivated aryl compounds (see figure). The reaction likely involves ligand-based radicals, whereas previously reported iron-based systems imply substrate-based radicals.

Full text of DOI:10.1002/chem.201304459

DOI: 10.1002/chem.201304459
Full Paper
&
CÀH Activation
Tandem CÀH Activation/Arylation Catalyzed by Low-Valent Iron
Complexes with Bisiminopyridine Ligands
Elise Salanouve, Ghania Bouzemame, Sꢀbastien Blanchard, Etienne Derat, Marine Desage-
El Murr,* and Louis Fensterbank*[a]
Dedicated to Professor Max Malacria on the occasion of his 65th birthday
Abstract: Tandem CÀH activation/arylation between unacti-
vated arenes and aryl halides catalyzed by iron complexes
that bear redox-active non-innocent bisiminopyridine li-
gands is reported. Similar reactions catalyzed by first-row
transition metals have been shown to involve substrate-
based aryl radicals, whereas our catalytic system likely in-
volves ligand-centered radicals. Preliminary mechanistic in-
vestigations based on spectroscopic and reactivity studies,
in conjunction with DFT calculations, led us to propose that
the reaction could proceed through an inner-sphere CÀH ac-
tivation pathway, which is rarely observed in the case of iron
complexes. This bielectronic noble-metal-like behavior could
be sustained by the redox-active non-innocent bisiminopyri-
dine ligands.
cent ligands,[3] which can act as electronic reservoirs, and thus
provide the metal with the additional electron(s) needed to
perform bielectronic redox elementary steps, such as oxidative
addition and reductive elimination.[4] Indeed, rationalization of
the mechanisms involved in iron catalysis often involves sub-
strate-centered radical pathways, which tend to impair both
scope and selectivity. In this prospect, the use of non-innocent
ligands to promote two-electron redox events could bypass
these limitations.[5] Among non-innocent redox ligands, bisimi-
nopyridines have been shown to participate in redox changes
that occur in the complex through stabilization of one or more
electrons on the ligand core, and thus engage in ligand-based
redox behavior.[6] These bisiminopyridine[7] ligands have been
popularized through their successful use with iron in polymeri-
zation,[7b–d] cycloisomerization,[7e] cycloaddition,[7f,g] hydrosilyla-
tion,[7h] and hydroboration reactions.[7i]
Introduction
Noble metals reign in the realm of catalysis and their ability to
sustain two-electron redox events that occur at the metal
center plays a large part in this success. However, growing
concerns with regards to the foreseeable stock depletion of
noble metals and inherent cost-related issues are driving
chemists to revisit the chemistry of their more modest cousins:
the first-row transition metals. Among these, the well-estab-
lished contribution of iron to organometallic chemistry[1] has
resulted in a significant number of synthetic applications and
helped deem this metal a potential alternative to noble metals
hegemony. More specifically, well-defined low-valent iron spe-
cies have lately elicited great excitement over their ability to
participate in useful catalytic endeavors and shed light on the
benefits obtained through more precise control of the coordi-
nation sphere.[2] These species show promise in terms of reac-
tivity, however, such low-valent complexes are still only mar-
ginally used and synthetic efforts towards their development
are of prime interest.
Iron-based reactions overall dubbed CÀH activation are
being reported at great pace[8] and stoichiometric inner-sphere
CÀH activations that involve iron cyclometalated species have
been reported.[9] However, apart from a few notable excep-
tions,[10] strictly speaking, metal-catalyzed inner-sphere catalytic
CÀH activation pathways have so far been the private preserve
of the noble metals. Among these transformations, one of the
trademark reactions of noble metals is the highly studied
carbon–carbon bond formation. Reliable iron-catalyzed cross-
coupling alternatives have now been widely developed with
a variety of functionalized partners[1d,11] but the specific case of
tandem CÀH activation/CÀC cross-coupling has been put
under scrutiny by recent studies. Tandem iron-catalyzed direct-
ed CÀH activation/arylation[12] has been achieved by Nakamura
by using diarylzinc or Grignard reagents as the coupling part-
ner,[10c,13] and reports of related iron-catalyzed direct arylation
of unactivated arenes with aryl halides have been simulta-
One of the strategies to enlarge the applicability of these
low-valent iron species lies in the use of redox-active non-inno-
[a] E. Salanouve, G. Bouzemame, Dr. S. Blanchard, Dr. E. Derat,
Dr. M. Desage-El Murr, Prof. L. Fensterbank
Sorbonne Universitꢀs, UPMC Univ Paris 06
UMR 8232, LabEx MiChem
Institut Parisien de Chimie Molꢀculaire (IPCM)
F-75005, Paris (France)
CNRS, UMR 8232, Institut Parisien de Chimie Molꢀculaire (IPCM)
F-75005, Paris (France)
E-mail: marine.desage-el_murr@upmc.fr
louis.fensterbank@upmc.fr
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304459.
Chem. Eur. J. 2014, 20, 1 – 9
1
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
neously disclosed by the groups of Charette[14] and Lei[15] in
2010. Although synthetically useful, these reactions do not
seem to proceed by genuine inner-sphere CÀH activation and
the mechanisms proposed involve substrate-centered radicals.
Although such catalytic developments of CÀH activation strat-
egies with iron-based systems are being reported, well-defined
iron complexes with bisiminopyridine ligands have not been
tested yet in CÀH activation strategies or cross-coupling reac-
tions. A recent report based on DFT calculations suggests that
benzylic CÀH activation with a diazocarbene coupling partner
should be possible.[16] Mindful of these challenges, we decided
to investigate this transformation and report the first tandem
CÀH activation/arylation of unactivated benzene with aryl bro-
mides catalyzed by iron complexes that bear bisiminopyridine
ligands.
mation of the catalyst did not provide significantly different re-
sults (Table 1, entry 7). Therefore, we selected complex
[1·FeBr2] as the catalyst for the rest of our study. X-ray diffrac-
tion studies performed on a single crystal of deep-blue
[1·FeBr2], grown by slow evaporation from a solution in CD2Cl2,
confirmed a geometry (see the Supporting Information) very
close to that of previously described [1·FeCl2].[18]
A subsequent base screen showed that KHMDS allowed for
complete conversion (Table 2, entry 1) and use of 1.5 equiva-
lents gave comparable results (Table 2, entry 2). However, use
Table 2. Screening of bases.
Results and Discussion
Entry
Base ([equiv])
Conversion [%]
GC yield [%]
Initial studies were focused on arylation with bromoanisole by
using benzene as the solvent (100 equiv) and potassium hexa-
methyl disilazide (KHMDS; 2 equiv) as the base at 808C for
18 h. Under these conditions, low amounts of the desired
product, 4-phenylanisole, and degradation products were ob-
tained when using FeBr2 alone, and the use of NaBEt3H to gen-
erate a reduced iron catalyst[17] did not yield any of the expect-
ed product (Table 1, entries 1 and 2).
1
2
3
4
5
6
7
8
9
10
11
12
13
KHMDS (2)
100
100
92
56
53
42 (35)[a]
KHMDS (1.5)
KHMDS (1.3)
KHMDS (0.5)
LiHMDS (1.5)
NaHMDS (1.5)
Cs2CO3 (1.5)
tBuOK (1.5)
DBU (1.5)
56
7[a]
11
30
0
10
0
0
100
100
0
49
52
43
52
40
100
DABCO (1.5)
Ph2NPh-4-OMe (1.5)
LDA (1.5)
14
7
19
Table 1. Catalyst screening.
LiTMP (1.5)
[a] Isolated yield.
of less than 1.5 equivalents resulted in incomplete conversion
and lower yields (Table 2, entries 3 and 4). The influence of the
counterion was found to be critical; LiHMDS and NaHMDS pro-
vided lower yields (Table 2, entries 5 and 6). Cesium carbonate
proved completely ineffective (Table 2, entry 7) and bulky terti-
ary nitrogen bases (1,8-diazabicyclo[5.4.0]undec-7-ene (DBU),
1,4-diazabicyclo[2.2.2]octane (DABCO), and Ph2N(4-MeO)Ph;
Table 2, entries 9–11) gave similarly disappointing results, with
lower conversions and little or no expected product. Interest-
ingly, though often used successfully in arylation reactions,
tBuOK provided the expected product in only 10% yield
(Table 2, entry 8). Stronger bases such as lithium diisopropyl-
amide (LDA) and lithium 2,2,6,6-tetramethylpiperidide (LiTMP;
Table 2, entries 12 and 13) were also far less efficient than
KHMDS, which suggested that the chemical structure of the
base, rather than high pKa values, was crucial for the reaction.
The influence of the metal source was also evaluated and it
was observed that, when combined with ligand 1, Fe(OAc)2
performed similarly to FeBr2, whereas FeCl2 afforded a slightly
lower yield (Table 3, entries 4–6). Furthermore, copper salts
(Table 3, entries 7 and 8) were found to be completely ineffi-
cient and resulted in complex mixtures of degradation prod-
ucts, which suggested that the catalytic activity was indeed
iron based.[19]
Entry
Catalyst
Reducing agent
([mol%])
Conversion
[%]
GC yield[a]
[%]
1
2
3
4
FeBr2
FeBr2
none
100
100
100
100
100
100
100
<5
0
NaBEt3H (32)
NaBEt3H (32)
NaBEt3H (32)
none
none
none
[1·FeBr2]
1+FeBr2
[1·FeBr2]
[1·FeCl2]
1+FeBr2
38
21
56
45
53
5
6[b]
7[b]
[a] Yields were determined by GC analysis with trimethylbenzene as an in-
ternal standard. [b] KHMDS (1.5 equiv).
In the presence of NaBEt3H and base, the bisiminopyridine
complex [1·FeBr2] gave the coupling product in 38% yield,
whereas the same amount of free ligand mixed in situ with
FeBr2 provided a lower yield of 21% (Table 1, entries 3 and 4).
The yield could be further improved to 56% by omission of
NaBEt3H, which also gave a cleaner reaction (Table 1, entry 5).
Catalyst [1·FeCl2] gave a slightly lower yield (Table 1, entry 6),
and premixing the ligand with the iron source for in situ for-
&
&
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
2
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Table 3. Influence of the metal source.
Table 5. Substrate scope.
Entry
ArBr
Conversion
[%]
ArÀPh
ArÀAr
Entry
Catalyst
Conversion [%][c]
GC yield [%]
[%]
[%]
1
[1·FeBr2]
[1·FeBr2]
[1·FeCl2]
1+FeBr2
100
100
100
96
90
93
53
54
45
53
42
52
0
2[a]
3[b,c]
4[c]
5[c]
6[c]
7[c]
8[c]
1
2
3
4
100
53
35
20
63
23
–
100
100
100
1+FeCl2
1+Fe(OAc)2
1+CuCl2
1+Cu(OAc)2
80
81
–
0
[a] Iron-source purity=99.995% (Alfa Aesar). [b] KHMDS (2 equiv). [c] Pre-
complexation mixing time=3 h.
10
5
100
35
3
Table 4. Screening of the ligand structure.[a]
6
7
8
100
100
100
26
7
–
–
16
trace
9
10
11
100
100
92
50
55
20
–
–
Entry
Ligand
Conversion [%]
GC yield [%]
1
2
3
4
5[c]
6
7[c]
8[d]
R=CH3, R’=iPr, R’’=H; 1
R=R’=CH3, R’’=H; 2
R=H, R’=R’’=CH3; 3
96
100
93
53
40
41[b]
17[e]
22[e]
17[e]
28[e]
18
98
R=CH3, R’=iPr, R’’=H; 4
>95
>95
>95
100
trace
R=R’’=H, R’=CH3; 5
terpyridine
12
75
–
–
[a] Reaction conditions: bromoanisole (1 equiv), benzene (100 equiv),
KHMDS (1.5 equiv), FeBr2 (15 mol%), ligand (15 mol%), 808C, 18 h.
[b] Yield determined by GC analysis with trimethylbenzene as an internal
standard. [c] Ligand (30 mol%). [d] FeCl2 used instead of FeBr2, KHMDS
(2 equiv). [e] Isolated yield.
13
14
15
100
100
100
41
7
–
–
–
trace
With optimal conditions in hand, we aimed to explore the
effect of the ligand structure. The best results were obtained
with the sterically hindered ligand 1 and reduction of the
steric bulk—either at the aryl ring by replacement of the diiso-
propylphenyl moiety with o,o’-dimethylphenyl (ligand 2) or at
the imine sp2 carbon (ligand 3)—resulted in slightly decreased
yields (Table 4, entries 2 and 3). Monoiminopyridine ligands,
used in iron-catalyzed dimerization of 1,3-dienes,[20] 1,4-addi-
tion of olefins to dienes,[21] hydroboration,[22] and hydrosilyla-
tion reactions,[23] were also tested and a sharp decrease in
yield was observed, regardless of the substitution pattern and
the amount used (metal/ligand ratio of 1:2 and 1:1; Table 4, en-
tries 4–7). Terpyridine gave similarly low conversions (Table 4,
entry 8).
16
17
100
100
–
–
–
–
18
100
–
–
poor, if any, conversion. Second, homocoupling products,
which can arise from the aryl halide, were only detected along-
side the expected product, suggesting that these two products
might share a common intermediate in the catalytic cycle.
However, the coupling product can be observed in the ab-
sence of the homocoupling adduct.
Investigation of the substrate scope revealed two interesting
features (Table 5). First, the reaction exhibits a strong bias to-
wards electron-rich scaffolds and all electron-deficient aromatic
or heteroaromatic compounds that were tested gave very
The influence of the halogen in the halide partner was eval-
uated and similar yields were obtained with iodides and bro-
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
3
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
copper,[26] and cobalt.[27] For the vast majority of reaction con-
ditions, the use of a base (such as tBuOK or LiHMDS) and a bi-
dentate nitrogen-based ligand (such as N,N’-dimethylethylene-
diamine (DMEDA) or bathophenanthroline) along with the
metal source was mandatory. Concomitant reports by Shi,[28]
Lei,[29] and Hayashi[30] even indicated that the metal source was
superfluous; the coupling reactions occurred in the sole pres-
ence of base and ligand, which tipped the scale towards
a seemingly organocatalytic approach. However, later com-
ments by Studer and Curran[31] suggested that HAS (homolytic
aromatic substitution) was most likely to be involved and that
these results might not indicate organocatalytic or metal-cata-
lyzed reactions, but rather HAS or metal-initiated base-promot-
ed HAS reactions.
Table 6. Halide scope and competition.[a]
Entry
X
Conversion [%]
GC yield [%][b]
1
2
3
4
Br
I
Cl
100
100
80
53
55
15[c]
46[c]
Br/Cl (1:1)
80(Br)/22(Cl)
[a] Reaction conditions: aryl halide (1 equiv), benzene (100 equiv).
[b] Yield determined by GC analysis with trimethylbenzene as an internal
standard. [c] Isolated yield.
Addition of radical scavengers, such as TEMPO or galvinoxyl
(1 equiv), to our reaction conditions was found to induce com-
plete inhibition of the reaction. Though indicative of the for-
mation of radicals in the catalytic pathway, these experiments
cannot distinguish between ligand- versus substrate-based
radicals. Further studies were conducted to determine whether
the radical(s) involved were substrate- or ligand-centered. Sub-
stitution of benzene with toluene afforded the expected cou-
pling products in 46% overall yield and a relative o/m/p distri-
bution of 2.2:1.95:1 (see the Supporting Information). Relative
to the ratios reported by Charette (3.1:1.9:1) and Lei
(2.47:1.9:1) for the same reaction, the observed pattern was
found to be closer to a statistical distribution and, therefore,
less in favor of an ortho effect, usually consistent with the pres-
ence of aryl radicals. Furthermore, this distribution does not
correlate with the regioselectivities observed in radical aromat-
ic substitutions on related systems.[32] Also noteworthy is the
fact that poor yield (10%; Table 2, entry 8) was obtained when
potassium tert-butoxide was used, which is usually very effi-
cient in HAS reactions. A KIE of 1.23 was measured (see the
Supporting Information), which indicated that breaking of the
CÀH bond might not be the rate-limiting step.[33]
mides (Table 6, entries 1 and 2), whereas aryl chlorides were
much less reactive towards coupling; only 15% yield was ob-
tained (Table 6, entry 3). This reactivity trend was confirmed
when an equimolar mixture of bromoanisole (1 equiv) and
chloroanisole (1 equiv) was used (Table 6, entry 4); almost all of
the bromoanisole reacted, whereas only 22% of the chloroani-
sole was converted, which confirmed that aryl bromides were
preferentially converted over aryl chlorides.
Complete conversion of the aryl bromide is always observed,
but the comparatively modest yields (ꢀ50%) obtained with
our system led us to investigate the reaction mixture for po-
tential byproducts (Scheme 1). It was found that the expected
Furthermore, reaction with a dihalide probe (4-chloro-bro-
mobenzene; Scheme 2a) for radical-anion mechanisms, de-
scribed by Bunnett and Creary,[34] provides a mixture of 4-
chlorobiphenyl (6; 15%), the result of monocoupling at the
bromo-substituted carbon atom, and para-terphenyl (7; 15%),
the result of biscoupling at both the bromo- and chloro-substi-
tuted carbon atoms. In radical-type mechanisms that involve
substrate-based radicals, the intramolecular electron-transfer
step that occurs on such dihalide probes would generate
mainly terphenyl 7, as observed by Kwong and Lei.[29] Conse-
quently, the significant amounts of 6 obtained do not fit well
with a substrate-based radical mechanism, nor does the pres-
ence of trace amounts of dichlorobiphenyl 8 that arise from
homocoupling at the bromo-substituted carbon atom, which is
totally unprecendented in such reactions.
Scheme 1. Product distribution for the CÀH arylation.
biaryl product is usually formed along with varying amounts of
the biaryl product from homocoupling of the aryl bromide. In
contrast with previously reported studies[14,15] of iron-catalyzed
CÀH arylation, in which biphenyl (a result of benzene homo-
coupling) was isolated as a byproduct, the only homocoupling
product we observe arises, instead, from the aryl bromide.
With a 100:1 ratio of benzene/bromoanisole in the reaction
mixture, this unusual observation could imply that different
mechanistic pathways operate in these reactions.
The procedures reported by the groups of Charette and Lei
call for catalytic amounts of iron salts (Fe(OAc)2 and FeCl3, re-
spectively) and no directing or activating group is required.
Preliminary mechanistic studies, which included kinetic isotope
effect (KIE) measurements and the introduction of radical scav-
engers, such as 2,2,6,6-tetramethylpiperidine N-oxide (TEMPO),
suggested that a radical pathway was at work in both systems.
Interestingly, similar coupling reactions have been reported
with other first-row transition metals,[24] such as nickel,[25]
To confirm this hypothesis, subsequent radical-trap experi-
ments were carried out on substrates 9 and 10 (Scheme 3).
Under conditions that involve substrate-based radicals, these
compounds are known to undergo intramolecular 6-endo-trig
cyclization to yield compound 12 as the major product.[24a,27a,35]
However, upon reaction with our catalytic system, the iodinat-
&
&
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
4
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Scheme 2. Dihalide probe for the CÀH arylation: results with a) our system
and b) under HAS conditions.
Scheme 3. Intramolecular radical-trap experiments.
ed substrate 9 did not provide any cyclized adduct 12 but
rather the reduced product 11, albeit in modest yield (28%).
The brominated substrate 10 provided reduced product 11
(10%), along with the cyclized adduct 12 (27%). Both sub-
strates 9 and 10 are known to favor cyclization in good yield
under conditions that involve substrate-based radicals, through
a mechanism via the same aryl-radical intermediate. However,
in our case, the low yield and discrepancy observed between
the iodo- and bromo-substrates under our conditions does not
support this pathway and these results suggest that substrate-
based radicals are probably not operative in this system.
Figure 1. 3D evolution profile of the in situ IR spectrum of [1·FeBr2] upon
portionwise addition of KHMDS (0.1 equiv increments) up to 2 equiv, then at
3 equiv a) n˜ =1300–1100 cmÀ1, b) n˜ =960–790 cmÀ1. c) In situ IR bands upon
catalysis: consumption of KHMDS (decrease at n˜ =1110 and 1080 cmÀ1), for-
mation of the coupling product (increase at n˜ =1180 cmÀ1). Substrate was
added at t=05:10:00.
In situ IR experiments performed on a solution of catalyst
[1·FeBr2] in benzene, to which increasing amounts of KHMDS
were introduced, showed complete consumption of up to two
equivalents of KHMDS before a stationary point was reached
(Figure 1a and b). Addition of more KHMDS (up to 15 equiv;
a similar ratio to the catalytic reaction conditions) only resulted
in the appearance and growth of the bands from KHMDS (in
particular, n˜ =1110 and 1080 cmÀ1), up to the maximum solu-
bility of KHMDS in benzene. Heating this solution to 808C did
not lead to changes in the IR spectrum (except for very slight
intensity variations), which attests to the stability of the mix-
ture in the absence of substrate. Addition of bromoanisole
(10 equiv) to the mixture under catalytic conditions induces
immediate consumption of KHMDS, attested to by a decrease
of the bands at n˜ =1110 and 1080 cmÀ1 (Figure 1c). Concomi-
tant growth of a band at n˜ =1180 cmÀ1 is observed, related to
the formation of the cross- and homo-coupling products,
which both display such a band.
The complete consumption of the catalyst under portion-
wise addition of KHMDS in benzene was accompanied by
a color change of the reaction mixture from dark blue to deep
green. The paramagnetic NMR spectrum of a solution of
[1·FeBr2] with KHMDS (2 equiv) in C6D6 was recorded and
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
5
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Figure 2. a) Spin density calculated for the reagent in the quintet spin state. b) Energy profiles for CÀH activation with an FeII catalyst. Values are in kcalmolÀ1
and the ground state of the reagent was taken as a reference. Values in plain text correspond to enthalpy, values in italics to Gibbs free energy. c) Possible
transition state.
showed a shift of the peaks in the paramagnetic region, which
suggested modifications of the complex structure. The spec-
trum was rather complicated and could only confirm the for-
mation of new unidentified paramagnetic species in solution.
However, full interpretation is difficult due to the instability of
the system under the NMR conditions. Attempts to record
a clean X-band EPR spectrum of this solution ([1·FeBr2] and
KHMDS (2 equiv) in benzene) only led to reproducible complex
spectra that precluded straightforward interpretation. However,
the fact that this solution displays EPR activity is noteworthy
because the starting complex [1·FeBr2] is X-band EPR silent.
These results collectively suggest that single-electron transfer
could have occurred between the starting complex and
KHMDS.
iron, one beta unpaired electron on the ligand, and one alpha
unpaired electron on the base (HMDS). Thus, triplet or quintet
spin states both emerge from a formal reduction of the ligand
by the base, which could be expected with regards to the
well-known redox properties of this ligand–iron complex. Sen-
sibly, higher in energy is a septet state; its electronic picture is
basically identical to that of the triplet or quintet state, except
that all the unpaired electrons are of alpha spin.
The transition state that corresponds to the CÀH activation
involves the three components (catalyst, benzene, base) in
a concerted fashion. This could formally be regarded as an
acid–base reaction, in which the base abstracts the proton and
the iron–phenyl bond is created simultaneously (Figure 2C).
However, because the HMDSÀ base also has strong radical
character, one could also consider the transformation as a hy-
drogen atom abstraction concomitant with a radical iron–
carbon coupling. With regards to the energetics, the barrier in
the quintet state is slightly below 20 kcalmolÀ1, which is com-
patible with the experimental conditions. The septet spin state
gives a rather high barrier of 38 kcalmolÀ1, which is related to
the fact that all the unpaired electrons are of alpha spin and,
thus, cannot rearrange to create bonds. Finally, the reaction is
clearly favorable for the products, with an exothermic DH=
À16 kcalmolÀ1 for the quintet ground state. It could be ex-
pected that the second equivalent of KHMDS could reduce the
FeII complex obtained after the CÀH activation step. Further-
more, the KIE of this step was calculated and a theoretical
value of 1.37 was found, which is a close match to the experi-
mental value of 1.23 and provides good theoretical support for
the proposed model for the CÀH activation step.
DFT calculations were conducted to rationalize the CÀH acti-
vation step. The CÀH activation on the iron(II) complex in pres-
ence of benzene and KHMDS (1 equiv) was scrutinized by
using a classical DFT functional (UB3LYP) in conjunction with
a double-dzeta polarized basis set (Figure 2). Several spin
states (triplet, quintet, and septet) were tested for reagents,
transition states, and products. It appears that the quintet spin
state is the ground state for the starting complex [LFeBr2],
which is not surprising for this family of d6 complexes. We as-
sumed that, upon interaction with KHMDS, loss of KBr would
result in a [LFeIIBr]+/HMDSÀ ion pair. When this ion pair is
formed, the calculations showed that four unpaired electrons
remain localized in the d orbitals of iron. At the same time,
electron transfer occurs from HMDSÀ to the ligand, which re-
sults in a single electron localized on the nitrogen of the
former HMDSÀ unit antiferromagnetically coupled to another
radical localized inside the aromatic ring of the ligand. Thus,
even in the presence of the reactants, the ground state is still
a quintet. The corresponding quintet spin density is shown in
Figure 2. Moreover, a triplet state is close in energy, around
3 kcalmolÀ1 above this quintet state in DH and more stable by
0.5 kcalmolÀ1 in DG, with two alpha unpaired electrons on
Conclusion
We have developed a tandem CÀH activation/arylation of un-
activated arenes with aryl bromides catalyzed by iron com-
plexes with bisiminopyridine ligands. Experimental data are
&
&
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
6
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Angew. Chem. 2011, 123, 2264–2266; Angew. Chem. Int. Ed. 2011, 50,
2216–2218.
clearly in favor of a mechanism distinct to that previously re-
ported for iron-based catalytic systems and, especially, not
compatible with homolytic aromatic substitution. Based on
preliminary theoretical studies, we propose that the system op-
erates through a metal-based inner-sphere CÀH activation that
involves concomitant electron transfer from the ligand and the
substrate, Further investigation of the mechanism is currently
underway. This strategy expands the scope of first-row base
metal catalysis and should undoubtedly promote their chemis-
try into exciting new venues.
[3] a) C. K. Jørgensen, Coord. Chem. Rev. 1966, 1, 164–178; b) P. J. Chirik, K.
Wieghardt, Science 2010, 327, 794–795; c) V. Lyaskovskyy, B. de Bruin,
ACS Catal. 2012, 2, 270–279; d) V. K. K. Praneeth, M. R. Ringenberg, T. R.
Ward, Angew. Chem. 2012, 124, 10374–10380; Angew. Chem. Int. Ed.
2012, 51, 10228–10234; e) O. R. Luca, R. H. Crabtree, Chem. Soc. Rev.
2013, 42, 1440–1459; f) W. I. Dzik, J. I. van der Vlugt, J. N. H. Reek, B. de
Bruin, Angew. Chem. 2011, 123, 3416–3418; Angew. Chem. Int. Ed. 2011,
50, 3356–3358.
[4] a) A. L. Smith, K. I. Hardcastle, J. D. Soper, J. Am. Chem. Soc. 2010, 132,
14358–14360; b) M. R. Haneline, A. F. Heyduk, J. Am. Chem. Soc. 2006,
128, 8410–8411; c) K. J. Blackmore, J. W. Ziller, A. F. Heyduk, Inorg. Chem.
2005, 44, 5559–5561.
[5] S. Blanchard, E. Derat, M. Desage-El Murr, L. Fensterbank, M. Malacria, V.
Mouriꢅs-Mansuy, Eur. J. Inorg. Chem. 2012, 376–389.
Experimental Section
[6] a) D. Zhu, I. Thapa, I. Korobkov, S. Gambarotta, P. H. M. Budzelaar, Inorg.
Chem. 2011, 50, 9879–9887; b) W. Kaim, Inorg. Chem. 2011, 50, 9752–
9765; c) B. de Bruin, E. Bill, E. Bothe, T. Weyhermꢄller, K. Wieghardt,
Inorg. Chem. 2000, 39, 2936–2947; d) P. H. M. Budzelaar, B. de Bruin,
A. W. Gal, K. Wieghardt, J. H. van Lenthe, Inorg. Chem. 2001, 40, 4649–
4655.
General procedure for aryl coupling
Complex [1·FeBr2] (78.4 mg, 0.1125 mmol, 0.15 equiv) and benzene
(2 mL) were added to a flame-dried Schlenk flask that had been
previously evacuated and backfilled with nitrogen (ꢂ3). KHMDS
(236 mg, 1.125 mmol, 1.5 equiv) was added under a N2 stream,
then the aryl bromide (0.75 mmol, 1 equiv) and benzene (4 mL)
were added. The resulting brownish solution was stirred for 18 h at
808C under nitrogen. After cooling to room emperature, the mix-
ture was filtered through a plug of silica gel and eluted with Et2O
or AcOEt. The solvent was removed in vacuo and the residue was
purified by silica gel column chromatography (pentane/CH2Cl2 or
cyclohexane/AcOEt) to afford the desired coupling product and, in
some cases, the homocoupling product. The yields were deter-
mined by GC analysis with 1,3,5-trimethylbenzene as an internal
standard.
[7] For a review, see: a) V. C. Gibson, C. Redshaw, G. A. Solan, Chem. Rev.
2007, 107, 1745–1776; b) B. L. Small, M. Brookhart, J. Am. Chem. Soc.
1998, 120, 7143–7144; c) G. J. P. Britovsek, M. Bruce, V. C. Gibson, B. S.
Kimberley, P. J. Maddox, S. Mastroianni, S. J. McTavish, C. Redshaw, G. A.
Solan, S. Strçmberg, A. J. P. White, D. J. Williams, J. Am. Chem. Soc. 1999,
121, 8728–8740; d) A. M. Tondreau, C. Milsmann, A. D. Patrick, H. M.
Hoyt, E. Lobkovsky, K. Wieghardt, P. J. Chirik, J. Am. Chem. Soc. 2010,
132, 15046–15059; e) K. T. Sylvester, P. J. Chirik, J. Am. Chem. Soc. 2009,
131, 8772–8774; f) B. R. D’Souza, T. K. Lane, J. Louie, Org. Lett. 2011, 13,
2936–2939; g) S. K. Russell, E. Lobkovsky, P. J. Chirik, J. Am. Chem. Soc.
2011, 133, 8858–8861; h) A. M. Tondreau, C. C. H. Atienza, K. J. Weller,
S. A. Nye, K. M. Lewis, J. G. P. Delis, P. J. Chirik, Science 2012, 335, 567–
570; i) J. V. Obligacion, P. J. Chirik, Org. Lett. 2013, 15, 2680–2683.
[8] C.-L. Sun, B.-J. Li, Z.-J. Shi, Chem. Rev. 2011, 111, 1293–1314.
[9] N. Yoshikai, Iron-Catalyzed CÀC Bond Formation with CÀH Bond Activa-
tion in PATAI’S Chemistry of Functional Groups, Wiley, 2012.
[10] a) W. D. Jones, G. P. Foster, J. M. Putinas, J. Am. Chem. Soc. 1987, 109,
5047–5048; b) T. Hatanaka, Y. Ohki, K. Tatsumi, Chem. Asian J. 2010, 5,
1657–1666; c) N. Yoshikai, S. Asako, T. Yamakawa, L. Ilies, E. Nakamura,
Chem. Asian J. 2011, 6, 3059–3065.
[11] a) A. Fꢄrstner, R. Martin, H. Krause, G. Seidel, R. Goddard, C. W. Lehmann,
J. Am. Chem. Soc. 2008, 130, 8773–8787; b) A. Fꢄrstner, Angew. Chem.
2009, 121, 1390–1393; Angew. Chem. Int. Ed. 2009, 48, 1364–1367.
[12] D. Alberico, M. E. Scott, M. Lautens, Chem. Rev. 2007, 107, 174–238.
[13] a) N. Yoshikai, A. Matsumoto, J. Norinder, E. Nakamura, Angew. Chem.
2009, 121, 2969–2972; Angew. Chem. Int. Ed. 2009, 48, 2925–2928;
b) E. Nakamura, N. Yoshikai, J. Org. Chem. 2010, 75, 6061–6067.
[14] a) F. Vallꢀe, J. J. Mousseau, A. B. Charette, J. Am. Chem. Soc. 2010, 132,
1514–1516; b) J. J. Mousseau, A. B. Charette, Acc. Chem. Res. 2013, 46,
412–424.
Acknowledgements
This research was funded by the Agence Nationale de la Re-
cherche, ANR JCJC 2011, (“UNILVERSAL” project, grant ANR-11-
JS07-004-01). The authors thank the Universitꢀ Pierre & Marie
Curie (UPMC, Paris 6) and the Centre National de la Recherche
Scientifique (CNRS). L.F. wishes to thank the Institut Universi-
taire de France (IUF). Clꢀment Nique is acknowledged for the
synthesis of ligand 3. X-ray crystallography was performed by
Lise-Marie Chamoreau and Geoffrey Gontard (UPMC). The au-
thors also thank the RENARD network and, more specifically,
Dr. J. Von Bardeleben and Dr. J.-L. Cantin from INSP (UMR 7588,
CNRS - UPMC, Paris 6) for access to the EPR spectrometer and
useful discussions.
[15] W. Liu, H. Cao, A. Lei, Angew. Chem. 2010, 122, 2048–2052; Angew.
Chem. Int. Ed. 2010, 49, 2004–2008.
[16] A. Varela-Alvarez, D. G. Musaev, Chem. Sci. 2013, 4, 3758–3764.
[17] S. C. Bart, K. Chłopek, E. Bill, M. W. Bouwkamp, E. Lobkovsky, F. Neese, K.
Wieghardt, P. J. Chirik, J. Am. Chem. Soc. 2006, 128, 13901–13912.
[18] B. L. Small, M. Brookhart, A. M. A. Bennett, J. Am. Chem. Soc. 1998, 120,
4049–4050.
Keywords: catalysis · CÀH activation · iron · IR spectroscopy ·
redox-active ligands
[19] S. L. Buchwald, C. Bolm, Angew. Chem. 2009, 121, 5694–5695; Angew.
Chem. Int. Ed. 2009, 48, 5586–5587.
[20] a) H. Tom Dieck, J. Dietrich, Chem. Ber. 1984, 117, 694–701; b) H. Tom
Dieck, J. Dietrich, Angew. Chem. 1985, 97, 795–796; Angew. Chem. Int.
Ed. Engl. 1985, 24, 781–783.
[21] B. Moreau, J. Y. Wu, T. Ritter, Org. Lett. 2009, 11, 337–339.
[22] J. Y. Wu, B. Moreau, T. Ritter, J. Am. Chem. Soc. 2009, 131, 12915–12917.
[23] J. Y. Wu, B. N. Stanzl, T. Ritter, J. Am. Chem. Soc. 2010, 132, 13214–
13216.
[24] a) H. Li, C. Sun, M. Yu, D. Yu, B. Li, Z. Shi, Chem. Eur. J. 2011, 17, 3593–
3597; b) B. Pieber, D. Cantillo, C. O. Kappe, Chem. Eur. J. 2012, 18, 5047–
5055.
[1] a) C. Bolm, J. Legros, J. Le Paih, L. Zani, Chem. Rev. 2004, 104, 6217–
6254; b) B. Plietker, Iron Catalysis in Organic Chemistry: Reactions and
Applications, Wiley-VCH, Weinheim, 2008; c) A. Correa, O. Garcꢃa Man-
cheÇo, C. Bolm, Chem. Soc. Rev. 2008, 37, 1108–1117; d) B. D. Sherry, A.
Fꢄrstner, Acc. Chem. Res. 2008, 41, 1500–1511; e) S. Enthaler, K. Junge,
M. Beller, Angew. Chem. 2008, 120, 3363–3367; Angew. Chem. Int. Ed.
2008, 47, 3317–3321.
[2] a) A. Fꢄrstner, K. Majima, R. Martꢃn, H. Krause, E. Kattnig, R. Goddard,
C. W. Lehmann, J. Am. Chem. Soc. 2008, 130, 1992–2004; b) A. Fꢄrstner,
R. Martin, K. Majima, J. Am. Chem. Soc. 2005, 127, 12236–12237;
ˇ
c) W. M. Czaplik, M. Mayer, J. Cvengros, A. J. von Wangelin, ChemSu-
sChem 2009, 2, 396–417. For a highlight, see: d) O. Garcꢃa MancheÇo,
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
7
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[25] a) G. Xie, T. Li, X. Qu, J. Mao, J. Mol. Catal. A 2011, 340, 48–52; b) O. Ko-
bayashi, D. Uraguchi, T. Yamakawa, Org. Lett. 2009, 11, 2679–2682.
[26] a) R. J. Phipps, M. J. Gaunt, Science 2009, 323, 1593–1597; b) H.-Q. Do,
R. M. K. Khan, O. Daugulis, J. Am. Chem. Soc. 2008, 130, 15185–15192.
[27] a) W. Liu, H. Cao, J. Xin, L. Jin, A. Lei, Chem. Eur. J. 2011, 17, 3588–3592;
b) C. T. To, T. L. Chan, B. Z. Li, Y. Y. Hui, T. Y. Kwok, S. Y. Lam, K. S. Chan,
Tetrahedron Lett. 2011, 52, 1023–1026.
[28] C.-L. Sun, H. Li, D.-G. Yu, M. Yu, X. Zhou, X.-Y. Lu, K. Huang, S.-F. Zheng,
B.-J. Li, Z.-J. Shi, Nat. Chem. 2010, 2, 1044–1049.
[29] W. Liu, H. Cao, H. Zhang, H. Zhang, K. H. Chung, C. He, H. Wang, F. Y.
Kwong, A. Lei, J. Am. Chem. Soc. 2010, 132, 16737–16740.
[31] A. Studer, D. P. Curran, Angew. Chem. 2011, 123, 5122–5127; Angew.
Chem. Int. Ed. 2011, 50, 5018–5022.
[32] D. I. Davies, D. H. Hey, B. Summers, J. Chem. Soc. C 1971, 2681–2684.
[33] E. M. Simmons, J. F. Hartwig, Angew. Chem. 2012, 124, 3120–3126;
Angew. Chem. Int. Ed. 2012, 51, 3066–3072.
[34] J. F. Bunnett, X. Creary, J. Org. Chem. 1974, 39, 3611–3612.
[35] For iodo substrate, see: a) C.-L. Sun, Y.-F. Gu, W.-P. Huang, Z.-J. Shi,
Chem. Commun. 2011, 47, 9813–9815; b) D. Sustac Roman, Y. Takahashi,
A. B. Charette, Org. Lett. 2011, 13, 3242–3245; for bromo substrate, see:
ref. [24a], [27a], and [28].
[30] E. Shirakawa, K. Itoh, T. Higashino, T. Hayashi, J. Am. Chem. Soc. 2010,
132, 15537–15539.
Received: November 14, 2013
Published online on && &&, 0000
&
&
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
8
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
&
CÀH Activation
E. Salanouve, G. Bouzemame,
S. Blanchard, E. Derat,
M. Desage-El Murr,* L. Fensterbank*
&& – &&
Tandem CÀH Activation/Arylation
Catalyzed by Low-Valent Iron
Complexes with Bisiminopyridine
Ligands
A radical choice! A low-valent iron
complex with non-innocent bisimino-
pyridine ligands performs CÀH activa-
tion/arylation of unactivated aryl com-
pounds (see figure). The reaction likely
involves ligand-based radicals, whereas
previously reported iron-based systems
imply substrate-based radicals.
Chem. Eur. J. 2014, 20, 1 – 9
www.chemeurj.org
9
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ