Coupling of benzyl halides with aryl Grignard reagents catalyzed by iron(III) amine-bis(phenolate) complexes

Article abstract of DOI:10.1016/j.jorganchem.2013.03.034

Reaction of benzylamino-N,N-bis(2-methylene-4,6-di-tert-amylphenol), H ₂L1, with anhydrous ferric chloride in the presence of a base yields FeCl(THF)L1 (1). In the solid state, complex 1 exists as a monomeric iron(III) species with a distorted trigonal bipyramidal geometry. Complex 1 is an air-stable, non-hygroscopic, single-component catalyst for C-C cross-coupling of aryl Grignard reagents with benzyl halides, including chlorides. Moderate to good yields of cross-coupled products can be obtained in diethyl ether at room temperature. Preliminary investigations include the screening of electron-donating and electron-withdrawing groups on both the benzylic substrate and the aryl Grignard reagent.

Full text of DOI:10.1016/j.jorganchem.2013.03.034

Journal of Organometallic Chemistry 737 (2013) 32e39
Contents lists available at SciVerse ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Coupling of benzyl halides with aryl Grignard reagents catalyzed by iron(III)
amine-bis(phenolate) complexes
,
,
Elliott F. Chard a, Louise N. Dawe a b, Christopher M. Kozak a
*
a Department of Chemistry, Memorial University of Newfoundland, St. John’s, NL, A1B 3X7 Canada
b X-ray Diffraction Laboratory, Centre for Chemical Analysis, Research and Training, Memorial University of Newfoundland, St. John’s, NL, A1B 3X7 Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
Reaction of benzylamino-N,N-bis(2-methylene-4,6-di-tert-amylphenol), H2L1, with anhydrous ferric
chloride in the presence of a base yields FeCl(THF)L1 (1). In the solid state, complex 1 exists as a
monomeric iron(III) species with a distorted trigonal bipyramidal geometry. Complex 1 is an air-stable,
non-hygroscopic, single-component catalyst for CeC cross-coupling of aryl Grignard reagents with
benzyl halides, including chlorides. Moderate to good yields of cross-coupled products can be obtained in
diethyl ether at room temperature. Preliminary investigations include the screening of electron-donating
and electron-withdrawing groups on both the benzylic substrate and the aryl Grignard reagent.
Ó 2013 Elsevier B.V. All rights reserved.
Received 13 November 2012
Received in revised form
13 March 2013
Accepted 22 March 2013
Keywords:
Cross-coupling
Iron
N,O-ligands
Grignards
1. Introduction
inconvenient on a large scale because it is highly hygroscopic and
yields vary according to its purity and commercial origin [24].
Transition metal-catalyzed cross-coupling, including nickel-
and palladium-catalyzed KumadaeCorriu couplings of Grignard
reagents with organohalides [1e3], is an important class of CeC
bond forming reactions [4,5]. Following early reports by Kochi [6],
and in the interest of sustainability, iron-based catalysts have been
experiencing a renaissance [2,7e11]. Iron catalysts are comple-
mentary to Ni and Pd in that they successfully couple alkyl halides
with Grignard reagents, which is not easily achieved using Ni or Pd
Although Fe(acac)3 is a more convenient, less hygroscopic starting
material [19], amine additives are often employed to achieve high
conversions and yields of cross-coupled products [24,25]. An easy
to use, non-hygroscopic single component catalyst precursor is,
therefore, highly desirable.
The catalytic formation of diarylmethane motifs is a very
important synthetic tool, with applications for the preparation of
pharmaceuticals and biologically active compounds [30]. Tradi-
tionally, diarylmethane compounds have been formed by
palladium-catalyzed coupling, including Suzuki-Miyaura [31e33]
and Stille [34,35] reactions.
To date, the formation of diarylmethanes by the iron-catalyzed
coupling of benzyl halides with aryl Grignard reagents has been
reported to be unsatisfactory, giving low yields and poor selec-
tivity resulting in the formation of homo-coupled by-products.
Compared to iron-catalyzed systems, copper or palladium-based
catalysts have been found to be more effective for the cross-
coupling of benzyl halides with aryl Grignard reagents
[31,32,36e41].
In 2009, Bedford and co-workers reported an efficient method
for the iron-catalyzed construction of diarylmethane compounds
[18]. Initial attempts focused on the cross-coupling of benzyl ha-
lides with aryl Grignard reagents in the presence of iron-based
catalysts. However, low yields and poor selectivity with respect to
cross-coupled products resulted. As a result, a route to these
due to competing
b-hydride elimination. However, unactivated
alkyl halides [2,3], particularly alkyl chlorides, continue to pose a
challenge and only a few examples of CeC bond formation using
alkyl chlorides have been reported [12e16]. Also, there have been
few reports of the synthesis of diarylmethane compounds via iron-
catalyzed coupling of aryl Grignards with benzyl halides [17], and
others have found these products required the use of aryl zinc
nucleophiles because aryl Grignard reagents proved unsatisfactory
[18]. A catalytic system that can address these shortcomings is
required.
The breadth of iron-catalyzed methods used for Kumada-type
cross-coupling is diverse [12,13,19e29]. Use of FeCl3, whether on
its own or to generate an active catalyst in situ, is often
* Corresponding author. Tel.: þ1 709 864 8082; fax: þ1 709 864 3702.
E-mail address: ckozak@mun.ca (C.M. Kozak).
0022-328X/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2013.03.034

E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
33
products was developed via the iron-phosphine-catalyzed Negishi
coupling of benzyl halides with diaryl zinc reagents. Although
diaryl zinc reagents can be obtained from the corresponding
Grignard reagents, from an atom-economic perspective it would be
desirable to develop an iron-based catalytic system in which aryl
Grignard reagents can be used directly.
Single crystals of 1 and 2 suitable for X-ray diffraction were
obtained by slow cooling of saturated toluene solutions to ꢀ35 ꢁC
inside a nitrogen filled glove box. The solid-state molecular
structures of 1 and 2 are shown in Fig. 1. Crystallographic and
refinement data are given in Table 1 and selected bond lengths
and angles in Table 2. The coordination around the iron atom in
both complexes is distorted trigonal bipyramidal with the trig-
We previously reported the synthesis of iron chloride complexes
of amine-bis(phenolate)-ether ligands [42] and their use in cata-
lytic cross-coupling of aryl Grignard reagents with primary and
secondary alkyl halides [17]. In order to generate more reactive
catalysts, yet still maintain the robust nature of the catalyst pre-
cursor, we turned our attention to tridentate amine-bis(phenolate)
ligands [43]. The previously reported iron(III)chloride amine-
bis(phenolate) used as the cross-coupling catalyst possessed
moderately sterically demanding groups on the phenolates (2-tert-
butyl,4-methyl phenolates) and an n-propyl amine. This complex
showed good to excellent activity for the cross-coupling of aryl
Grignard reagents with alkyl halides, including a small number of
benzyl halides. Encouraged by these preliminary results, we sought
to investigate a broader range of benzyl halide electrophiles as
coupling partners. For this study, we investigated the effect of using
bulkier groups on both the phenolate and amine sites in terms of
the structural and catalytic properties on the resulting complexes.
The complex used in our previous investigation [43] is noticeably
hygroscopic and obtaining pure, anhydrous material was some-
times problematic. By using the larger tert-amyl (tert-pentyl)
groups on the phenol fragment and a benzyl amine, we believed
that we could obtain a complex that was less hygroscopic. Herein
we report two new structurally authenticated amine-
bis(phenolate) iron(III) complexes and our initial investigations of
their use in CeC cross coupling of aryl Grignard reagents with
benzyl halides.
onality index value
s (as defined by Addison and Reedijk [44]) of
0.805 for 1 and of 0.783 for 2. The metals are bonded to two
phenolate oxygen atoms and a chloride ion (in 1) or bromide ion
(in 2), which define the trigonal planes of the bipyramids. The
sums of bond angles about the equatorial planes are 359.89ꢁ and
359.85ꢁ for
1 and 2, respectively, indicating nearly perfect
planarity. The central nitrogen atoms of the ligands and the oxy-
gen atoms of the THF ligands occupy apical sites giving O(3)e
Fe(1)eN(1) bond angles of 172.0(2)ꢁ and 171.77(12)ꢁ for 1 and 2,
respectively. The O(3)eFe(1)eN(1) angles are distorted from the
ideal linear geometry and are bent away from the phenolate
groups and directed toward the halide ions. The FeeOphenolate
distances in 1 and 2 are experimentally equivalent, with Fe(1)e
ꢀ
O(1) and Fe(1)eO(2) lengths of 1.854(6) and 1.848(6) A, respec-
tively, for 1 and 1.8491(18) and 1.842(4) A for 2. The Fe(1)eCl(1)
ꢀ
2. Results and discussion
2.1. Synthesis and characterization of 1 and 2
The tridentate amine-bis(phenol) ligand precursor H2L1,
abbreviated H2[O2N]RR R (where R ¼ R0 ¼ t-Am, R00 ¼ benzyl) was
readily synthesized by a modified Mannich condensation reaction,
in which the 2,4-disubstituted phenol, amine and formaldehyde
were heated to reflux in water for 12 h. Monometallic complexes 1
and 2 were prepared in high yield by two methods (Scheme 1). One
route involves reacting the proligand H2L1 with FeCl3 in THF then
neutralizing the resulting dark purple solution with NEt3. The
second route involves first reacting the proligand with NaH to form
Na2L1. The sodium compound is then reacted with FeBr3 to
generate the desired product and sodium bromide. MALDI-TOF
mass spectrometry was carried out on 1 and 2 using anthracene
as the matrix and the mass spectra showed the characteristic
fragment ions [M ꢀ THF]þ and [M ꢀ THFeX]þ (X ¼ Cl for 1, Br for 2).
0
00
Fig. 1. Molecular structures (ORTEP) and partial atom labeling of 1 (top) and 2 (bot-
Scheme 1. Two routes to the synthesis of 1 and 2.
tom). Ellipsoids shown at 50% probability. Hydrogen atoms omitted for clarity.

34
E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
Table 1
2.2. Cross-coupling catalysis studies
Crystallographic and structure refinement data for 1 and 2.
Previous studies with related Fe(III) complexes supported by
tetradentate amine-bis(phenolate)-ether ligands suggest that
diethyl ether is superior toTHF as a solvent for the cross-coupling of
Grignard reagents with alkyl halides [17]. In addition, it was pre-
viously found that reactions performed at room temperature gave
superior results to those conducted at lower temperatures. There-
fore, diethyl ether was the solvent of choice for the current study
and all reactions were performed at room temperature. The first
group of cross-coupling reactions investigated involved the reac-
tion between benzyl bromide (or benzyl chloride) and a series of
Grignard reagents (Table 3). An initial reaction of benzyl bromide
with phenylmagnesium bromide (PhMgBr) in the presence of 1
gave a 30% yield of cross-coupled product after 30 min at room
temperature. Similar results were obtained when bromide con-
taining complex 2 was used (Table 3, entry 1). Significant yields of
the bibenzyl and biaryl homocoupled by-products were also ob-
tained. Benzyl chloride could also be used as the electrophilic
partner, generating a similar yield of the desired cross-coupled
product by both 1 and 2, except higher yield of bibenzyl product
was observed when complex 2 was used (entry 2). The reaction of
benzyl bromide with o-tolylmagnesium bromide gave a moderate
yield (86%) of the cross-coupled product after 30 min at room
temperature (entry 3). In previous reports from our group, the re-
action between benzyl bromide and o-tolylmagnesium bromide
gave yields of 60% and 68% in the presence of octahedral (amine)
bis(phenolato)FeIII(acac) complexes and trigonal bipyramidal
iron(III) halide complexes (supported by tetradentate amine-
bis(phenolate) ligands), respectively [17,49]. Surprisingly, when
benzyl chloride was employed, a higher yield (94%) of the cross-
coupled product was found (entry 4). We previously obtained a
95% yield of cross-coupled products from benzyl chloride and o-
tolylmagnesium bromide in the presence of related tridentate
amine-bis(phenolate) iron(III) complexes [43]. A yield of 52% was
reported in the presence of octahedral (amine)bis(phenolato)
FeIII(acac) complexes [49]. Using p-tolylmagnesium bromide,
however, gave slightly lower yields than o-tolylmagnesium bro-
mide with the respective benzyl halide (entries 5 and 6) generating
higher yields of the bibenzyl and biaryl homocoupled by-products.
A similar finding was also observed in the presence of octahedral
(amine)bis(phenolato)FeIII(acac) complexes and trigonal bipyra-
midal iron(III) chloride complexes with tetradentate amine-
bis(phenolate) ligands [17,49]. Bedford and co-workers reported
the iron-catalyzed Negishi coupling of benzyl bromide with the
diaryl zinc reagent prepared from p-tolylmagnesium bromide gave
a 76% isolated yield of the desired cross-coupled product [18].
Because there was generally little difference observed between the
catalytic activity of 1 and 2, subsequent coupling experiments were
performed using 1 only. When benzyl bromide was reacted with 4-
methoxyphenylmagnesium bromide (4-anisylmagnesium bro-
mide) in the presence of 1, a 21% yield of the cross-coupled product
was found along with large quantities of the bibenzyl by-product
(entry 7). We had previously reported that the reaction between
benzyl bromide and 4-anisylmagnesium bromide resulted in no
yield of the cross-coupled product when iron(III) chloride com-
plexes with tetradentate amine-bis(phenolate) ligands were used
as the pre-catalyst [17]. The Negishi-type arylation between benzyl
bromide and the corresponding diaryl zinc reagent prepared from
4-anisylmagnesium bromide resulted in a 95% isolated yield of the
cross-coupled product [18]. Reacting benzyl bromide with 4-
fluorophenylmagnesium bromide (4-FPhMgBr) also resulted in a
poor yield (21%) of the cross-coupled product (entry 8). Large
quantities of the homocoupled biaryl and bibenzyl products were
formed instead. Surprisingly, benzyl bromide was found to couple
Compound
1
2
Chemical formula
Color
Habit
C
Red
Prism
45H67ClFeNO3
C45H67BrFeNO3
Dark red
Prism
Formula mass
Crystal system
761.33
Monoclinic
25.441(12)
10.907(4)
31.379(15)
90
94.00(3)
90
8686(7)
163(1)
I2/a (#15)
8
1.164
MoK
0.446
3288
16,461
7386
0.1000
0.1181
0.3701
1.094
805.78
Monoclinic
39.512(3)
10.9112(4)
25.084(2)
90
126.271(3)
90
8718.8(11)
163(1)
C2/c (#15)
8
1.228
MoKa
1.302
3432
54,433
9027
0.0364
0.0622
0.1715
1.093
ꢀ
a [A]
ꢀ
b [A]
ꢀ
c [A]
a
b
g
[ꢁ]
[ꢁ]
[ꢁ]
3
ꢀ
Unit cell V [A ]
Temperature [K]
Space group
Z
Dc/g cmꢀ3
Radiation type
Absorption,
F(000)
a
m
[mmꢀ1
]
Reflections measured
Independant refl’s
Rint
R1 (I > 2s
(I))a
wR(F2) (I > 2 (I))b
s
Goodness of fit on F2
CCDC Ref
894993
910516
a
R1 ¼ S((jFoj ꢀ jFcj)/SjFoj).
b
wR2 ¼ [S(w(F2o ꢀ F2c)2)/Sw(Fo2)2]1/2
.
ꢀ
distance of 2.237(3) A in 1 is shorter than the FeeCl distances
found in previously reported trigonal bipyramidal iron(III)
chloride-bridged dimers with related tridentate amine-
bis(phenolate) ligands [43]. Additionally, the FeeCl distance
observed in 1 is slightly shorter than the FeeCl distances reported
in similar iron(III) trigonal bipyramidal complexes possessing
tetradentate amine-bis(phenolate) ligands [42]. The Fe(1)eBr(1)
distance in 2 is shorter than FeeBr distances reported in other
five-coordinate iron(III)-bromide complexes having multidentate
ligands in trigonal bipyramidal geometries [45e47] but longer
than in complexes having square pyramidal geometries [42,48].
The central nitrogen donor in the ligand backbone exhibits an Fee
ꢀ
N(1) bond length of 2.190(6) A in 1 and 2.185(2) in 2. The Fe(1)e
O(3) bond lengths are 2.151(6) and 2.145(2) A for 1 and 2,
respectively.
ꢀ
Table 2
ꢁ
ꢀ
Selected bond lengths (A) and angles ( ) for 1 and 2.
1
2
Fe(1)eO(1)
Fe(1)eO(2)
Fe(1)eO(3)
Fe(1)eN(1)
Fe(1)eCl(1)
Fe(1)eBr(1)
O(1)eFe(1)eO(2)
N(1)eFe(1)eCl(1)
N(1)eFe(1)eBr(1)
N(1)eFe(1)eO(3)
Cl(1)eFe(1)eO(3)
Br(1)eFe(1)eO(3)
O(1)eFe(1)eCl(1)
O(1)eFe(1)eBr(1)
O(1)eFe(1)eO(3)
O(2)eFe(1)eCl(1)
O(2)eFe(1)eBr(1)
O(2)eFe(1)eO(3)
O(1)eFe(1)eN(1)
O(2)eFe(1)eN(1)
1.854(6)
1.848(6)
2.151(6)
2.190(6)
2.237(3)
1.8491(18)
1.842(4)
2.145(2)
2.185(2)
2.3808(8)
124.80(14)
123.7(3)
96.12(17)
96.79(8)
172.0(2)
91.72(18)
171.77(12)
91.36(10)
121.49(20)
120.17(11)
86.89(9)
86.7(2)
114.7(2)
114.88(9)
88.17(12)
88.03(8)
88.6(2)
88.0(2)
89.4(2)
89.41(11)

E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
35
Table 3
Cross-coupling of benzyl bromide or benzyl chloride with Grignard reagents by complexes 1 and 2.
Entry
ArMgBr
Alkyl halide
Product
Yielda (%)
30d (28)
Yieldb AreAr (%)
Yieldc bibenzyl (%)
1
Ph
50 (50)
30 (32)
<5 (25)
10 (10)
<3 (20)
30 (32)
6 (10)
30
2
3
4
5
6
7
Ph
32 (30)
86 (80)
94 (80)
49 (35)
91 (85)
21
65 (55)
18 (20)
20 (20)
45 (50)
40 (35)
30
o-tolyl
o-tolyl
p-tolyl
p-tolyl
4-anisyl
8
9
4-FPh
21
95
60
30
2,6-Me2Ph
<5
Trace
10
vinyl-MgBr
Trace
Trace
30
a
Spectroscopic yields determined by GCeMS using dodecane as an internal standard.
Percent yield of biaryl by-product is given with respect to Grignard reagent.
Percent yield of bibenzyl by-product is given with respect to benzyl halide.
Yields given for reactions catalyzed by 1, except those in brackets, which were catalyzed by 2.
b
c
d
with the sterically crowded 2,6-dimethylphenylmagnesium bro-
as the aryl Grignard reagent, a slightly higher yield of cross-coupled
product (38%) was obtained (entry 3). However, high quantities of
the bibenzyl and biaryl homocoupled by-products were generated.
The Negishi-type arylation between 4-methylbenzyl bromide and
the corresponding diaryl zinc reagent of p-tolylmagnesium resulted
in an 86% isolated yield of the cross-coupled product as reported by
Bedford and co-workers [18]. Surprisingly, in the presence of 1, 4-
methylbenzyl chloride was found to couple with p-tolylmagne-
sium in a high yield of 85% (entry 4). When the electron donating
methyl group of 4-methylbenzyl bromide was replaced by a weakly
electron withdrawing bromide group (4-methylbenzyl bromide),
the desired cross-coupled product was obtained in a modest yield of
67% (entry 5) giving higher yields of the biphenyl homocoupled by-
product. According to reports by Bedford and co-workers, the iron-
catalyzed Negishi coupling of 4-methylbenzyl bromide with the
corresponding diaryl zinc reagent of p-tolylmagnesium bromide
gave an 80% isolated yield of the desired cross-coupled product [18].
Interestingly, when the strongly electron withdrawing substrate 4-
(trifluoromethyl)benzyl bromide was employed, a higher yield (76%)
mide (2,6-Me2PhMgBr) in an excellent yield of 95% (entry 9). Pre-
viously, a 78% yield of cross-coupled product from benzyl bromide
and 2,6-dimethylphenylmagnesium bromide was obtained when
using a related tridentate amine-bis(phenolate) iron(III) complex
[43]. When benzyl bromide was reacted with vinylmagnesium
bromide (in the presence of 1), very poor selectivity resulting in the
formation of homocoupled by-products was found. Only trace
quantities of the desired product were obtained (entry 10) while
high quantities of the bibenzyl by-product formed instead.
A series of para-substituted benzyl halides were screened as
cross-coupling reaction partners. When 4-methylbenzyl bromide
was reacted with 4-anisylmagnesium bromide in the presence of 1
at room temperature, only 19% yield of the desired cross-coupled
product was obtained (Table 4, entry 1) and high quantities of the
homocoupled bibenzyl product were formed instead. Reacting 4-
methylbenzyl bromide with 4-fluorophenylmagnesium bromide
(4-FPhMgBr) also resulted in a poor yield (13%) of the cross-coupled
product (entry 2). When p-tolylmagnesium bromide was employed

36
E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
Table 4
Cross-coupling of para-substituted benzyl halides with aryl Grignard reagents catalyzed by 1.
Entry
ArMgBr
Alkyl halide
Product
Yielda (%)
19
Yieldb AreAr (%)
Yieldc bibenzyl (%)
25
1
4-anisyl
40
2
4-FPh
13
50
45
3
4
5
p-tolyl
p-tolyl
p-tolyl
38
85
67
40
10
40
20
10
20
6
p-tolyl
p-tolyl
Trace
50
10
0
7
76
10
a
Spectroscopic yields determined by GCeMS using dodecane as an internal standard.
Percent yield of biaryl by-product is given with respect to Grignard reagent.
Percent yield of bibenzyl by-product is given with respect to benzyl halide.
b
c
of the cross-coupled product was obtained (entry 7). Bedford and
co-workers found a 59% isolated yield of the cross-coupled product
for the Negishi coupling of 4-(trifluoromethyl)benzyl bromide with
the corresponding diaryl zinc reagent of p-tolylmagnesium bromide
[18]. As shown in Table 4, entry 6, the introduction of an ester group
at the para position of the benzyl halide only gave trace quantities of
the desired product resulting in the generation of high quantities of
the homocoupled biaryl product instead. The poor activity for cross-
coupling is likely attributed to nucleophilic attack of the Grignard
reagent at the ester functional group.
Cross-coupling reactions with meta- and ortho-substituted
benzyl halides were also screened. 3-Methoxybenzyl bromide was
found to give low to modest yields depending on the aryl Grignard
reagent used. In the presence of p-tolylmagnesium bromide, a
moderate yield (72%) of the cross-coupled product was obtained
(Table 5, entry 1). A higher yield of 92% was reported for the Negishi
coupling of 3-methoxybenzyl bromide with the diaryl zinc reagent
prepared from p-tolylmagnesium bromide [18]. Surprisingly, 3-
methoxybenzyl chloride could also be used as the electrophilic
partner, generating a higher yield (91%) of the desired cross-
coupled product (entry 2). Unfortunately, 3-methoxybenzyl bro-
mide gave a low yield of the cross-coupled product when reacted
with 4-anisylmagnesium bromide in the presence of 1 (entry 3). In
bromide was reacted with p-tolylmagnesium bromide in the
presence of 1, only trace quantities of the desired cross-coupled
product were generated with high yields of the biaryl homo-
coupled by-product (entry 4). When 2-(bromomethyl) benzonitrile
was reacted with p-tolylmagnesium bromide, no yield of the cross-
coupled product was obtained (entry 5), possibly due to attack of
the Grignard reagent at the nitrile group. In fact, the reaction
selectively generated the biaryl homocoupled by-product. The re-
action between 2-(trifluoromethyl)benzyl bromide and p-tol-
ylmagnesium bromide gave a low yield (24%) of the cross-coupled
product (entry 6). Low yields of the bibenzyl and biphenyl homo-
coupled by-products were also obtained. The Negishi coupling of 2-
(trifluoromethyl)benzyl bromide with the corresponding diaryl
zinc reagent of p-tolylmagnesium bromide gave a higher yield
(64%) of the cross-coupled product [18].
2.3. Mechanistic considerations
For many of the cross-coupling reactions attempted, bibenzyl
homocoupling by-products were observed. Previously, for the re-
action of dichloroethane with Grignard reagents, Hayashi and co-
workers proposed a mechanism suggesting that benzyl halides
could undergo radical reactions in the presence of reduced metals
[50]. We also proposed a similar mechanism for the reaction be-
tween dichloromethane and Grignard reagents [51]. Nakamura and
fact,
a high quantity of the unreacted starting material 3-
methoxybenzyl bromide was found. When 2-bromobenzyl

E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
37
Table 5
Cross-coupling of meta- and ortho-substituted benzyl halides with aryl Grignard reagents.
Entry
ArMgBr
Alkyl halide
Product
Yielda (%)
Yieldb Ar-Ar (%)
Yieldc bibenzyl (%)
1
p-tolyl
72
10
14
2
3
p-tolyl
91
24
15
50
0
4-anisyl
10
4
5
p-tolyl
p-tolyl
trace
0
70
70
trace
0
6
p-tolyl
24
25
25
a
Spectroscopic yields determined by GCeMS using dodecane as an internal standard.
Percent yield of biaryl by-product is given with respect to Grignard reagent.
Percent yield of bibenzyl by-product is given with respect to benzyl halide.
b
c
Fürstner have also reported mechanisms where the iron-catalyzed
cross-coupling of alkyl halides with aryl Grignard reagents pro-
ceeds via a radical process [12,20]. Benzyl halides can undergo
oxidative addition (OA) at a reduced iron centre or undergo a single
electron transfer (SET) reaction with the reduced centre generating
an arylmethyl radical, which subsequently undergoes radical
many of the cross-coupling reactions attempted and consequently
the low yields of the desired cross-coupled product. As shown in
Scheme 2, after the iron(III) pre-catalyst is reduced by the aryl
Grignard reagent, the catalytically active iron species, [Fe] can
either undergo oxidative addition (Path B) with the benzyl halide or
take part in a single electron transfer (SET) side reaction (Path A)
with the benzyl halide generating an arylmethyl radical, and in
turn, 0.5 equivalents of the bibenzyl homocoupled by-product. If
oxidative addition at the reduced iron centre occurs, the resulting
benzylironhalide complex is expected to undergo transmetallation
with the aryl Grignard to form an arylbenzyliron complex. Reduc-
tive elimination of the arylbenzyliron complex would then
generate the desired cross-coupled product along with regenera-
tion of the reduced iron species. The exact nature of the reduced
iron species under these conditions is not known. However, upon
addition of the Grignard reagent to the benzyl halide and catalyst
mixture, the solution darkens considerably, suggesting the forma-
tion of particulate iron may be occurring. Nanoparticulate iron has
been proposed as the catalytic species in iron-catalyzed cross-
coupling reactions [29]. Copper impurities in iron salts were shown
to highly influence the yields of CeN, CeO and CeS bond formation
reactions [52] and the copper-catalyzed coupling of aryl Grignard
reagents with alkyl halides, including benzyl bromides, was suc-
cessful but limited by low functional group tolerance [36].
Furthermore, since the traditional metals used for Kumada
coupling of Grignard reagents with aryl or vinyl halides are palla-
dium or nickel, we used inductively coupled plasma-mass spec-
trometry (ICP-MS) to determine the degree of contamination by Cu,
coupling (Scheme 2) [50,51].
A similar mechanism may be
responsible for the bibenzyl homocoupled byproduct observed in
FeIII
ArMgX
Reduction
Reductive
Path A
Single electron
transfer
ArCH2X
elimination
Ar−Ar
ArCH2Ar'
[Fe]
Ar
•
[Fe]X + ArCH2
[Fe]
Ar'
MgX2
X
0.5 ArCH2CH2Ar
[Fe]
Path B
Oxidative addition
Ar'MgX
Ar
Transmetallation
Scheme 2. Catalytic cycle for the generation of bibenzyl homocoupled by-products
(Path A) and diarylmethane compounds (Path B).

38
E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
Ni, Pd and other metals in compound 1 and in the FeCl3 used to
prepare it. ICP-MS showed compound 1 contained only 1.3 ppm of
Cu impurity, whereas Ni and Pd concentrations were extremely low
at 100 and 25 ppb (parts per billion), respectively. By comparison,
the FeCl3 used for the synthesis of 1 was found to contain 110 ppm
of Ni, 100 ppm of Cu, and only 3 ppb of Pd. We were therefore
pleased that in terms of metal-contaminants, our complex proved
to be significantly more pure than the anhydrous ferric chloride
from which it was made.
Elan DRCII ICP-MS. 100 mg of a purified and dried sample was
dissolved in distilled water and concentrated nitric acid then
heated on a hot plate overnight. 60 g of Nanopure water was added
to the samples, which were then centrifuged and diluted to 10%
concentration for analysis.
4.2. Syntheses
4.2.1. H2[O2N]AmAmBn (H2L1)
3. Conclusions
To a stirred mixture of 2,4-di-tert-amylphenol (28.83 g,
123.2 mmol) in 100 mL of deionized water was added 37% aqueous
formaldehyde (10 mL, 123.2 mmol) followed by slow addition of
benzyl amine (6.59 g, 61.5 mmol). The reaction mixture was heated
to reflux for 12 h. Upon cooling, the mixture separated into two
phases. The upper phase was decanted and the remaining white
mass of waxy material was triturated with cold methanol to give an
analytically pure, white powder (27.91 g, 76%). Anal. Calcd for
These investigations demonstrated that iron(III) complexes
supported by tridentate amine-bis(phenolate) ligands catalyze the
cross-coupling of aryl Grignard reagents with benzyl halides,
although with highly variable yields and significant homo-coupled
by-products. In the presence of 1 or 2, the coupling of o-tol-
ylmagnesium bromide with benzyl halides (including chlorides)
gave cross-coupled products in good yields. The system also
showed moderate to good reactivity for sterically demanding nu-
cleophiles, such as 2,6-dimethylphenylmagnesium bromide. Un-
fortunately, the reaction of the more sterically congested ortho-
substituted benzyl halides with p-tolylmagnesium bromide gave
very low yields of the desired cross-coupled product generating
mainly homocoupled by-products.
C
41H61NO2: C, 82.08; H, 10.25; N, 2.33. Found: C, 82.13; H, 10.22; N,
2.32%. 1H NMR (500 MHz, 295 K, CDCl3):
d
7.38 (m, ArH, 2H); 7.31
(m, ArH, 3H); 7.08 (d, 4J ¼ 2.2 Hz, ArH, 2H); 6.86 (d, 4J ¼ 2.2 Hz, ArH,
2H); 3.62 (s, ArCHHN, 4H); 3.53 (s, NCH2Ph, 2H); 1.88 (q,
3J ¼ 7.49 Hz, CH2, 4H); 1.56 (q, 3J ¼ 7.49 Hz, CH2, 4H); 1.34 (s,
C(CH3)2, 12H); 1.23 (s, C(CH3)2, 12H); 0.63 (t, 3J ¼ 7.45 Hz, CH2CH3,
6H); 0.62 (t, 3J ¼ 7.45 Hz, CH2CH3, 6H). 13C{1H} NMR (125 MHz,
298 K, CDCl3):
d 152.4 (ArCOH); 139.9 (Ar); 138.0 (Ar); 134.5 (Ar);
4. Experimental
130.0 (Ar); 129.3 (Ar); 128.3 (Ar); 126.3 (Ar); 126.2 (Ar); 121.6 (Ar);
58.7 (ArCH2N); 57.4 (ArCH2N); 38.9 (C(CH3)2); 37.7 (CH2CH3); 37.5
(CH2CH3); 33.4 (C(CH3)2); 29.0 (C(CH3)2); 28.15 (C(CH3)2); 10.0
(CH2CH3); 9.6 (CH2CH3). HRMS (TOF MS EIþ): (m/z): [M]þ calcd. For
4.1. General methods and materials
C
41H61NO2, 599.4702; found, 599.4711. MP range (ꢁC): 127.4e128.9.
ꢀ
THF was stored over 4 A molecular sieves and distilled from
sodium benzophenone ketyl under nitrogen. Anhydrous toluene
and diethyl ether were purified using an MBraun solvent purifica-
tion system. Anhydrous FeCl3 (97%, Aldrich) was used for the syn-
thesis of 1 and FeBr3 (99%, Strem) used for the synthesis of 2. All
other reagents were purchased from Strem, Aldrich or Alfa Aesar
and used without further purification. All CeC cross-coupling re-
actions were performed under an atmosphere of dry oxygen-free
nitrogen by means of standard Schlenk techniques or by using an
MBraun LabmasterDP glove box. Benzyl halides and Grignard re-
agents were used without further purification. Dodecane was used
as an internal standard for GCeMS analysis and diethyl ether was
used for sample preparation. Acetophenone was used as an internal
standard for 1H NMR analysis.
4.2.2. FeCl(THF)L1 (1)
To a THF solution (50 mL) of recrystallized H2L1 (2.00 g,
3.33 mmol) under nitrogen was added a solution of anhydrous FeCl3
(597 mg, 3.33 mmol) in THF resulting in an intense purple solution.
To this solution was added triethylamine (674 mg, 6.66 mmol) and
the resulting mixture was stirred for 2 h. After stirring, the dark
purple solution was filtered through Celite. Removal of solvent un-
der vacuum yielded a dark purple product (1.81 g, 71%). The purple
product was dissolved in minimal toluene and was placed in the
freezer for 48 h during which a thin layer of white precipitate
appeared. The supernatant was decanted and filtered through Cel-
ite. Crystals suitable for X-ray diffraction were obtained by slow
evaporation of this toluene solution (1.408 g, 56%). Anal. Calcd for
NMR spectra were recorded on a Bruker Avance III 300 MHz
instrument with a 5 mm-multinuclear broadband observe (BBFO)
probe. MALDI-TOF MS spectra were performed using an ABI QSTAR
XL Applied Biosystems/MDS hybrid quadrupole TOF MS/MS system
equipped with an oMALDI-2 ion source. Samples were prepared at
a concentration of 10.0 mg/mL in toluene. Anthracene was used as
the matrix, which was also mixed at a concentration of 10.0 mg/mL.
UVevis spectra were recorded with an Ocean Optics
USB4000 þ fiber optic spectrophotometer. HRMS spectra were
recorded using a Waters Micromass GCT Premier High Resolution
mass spectrometer equipped with an electron ionization source
and an orthogonal acceleration time-of-flight (oa-TOF) mass
analyzer. Melting point data were collected on a MPA100 OptiMelt
Automated Melting Point System. Elemental analyses were carried
out by Canadian Microanalytical Services Ltd. Delta, BC, Canada. Gas
chromatography mass spectrometry (GCeMS) analyses were per-
formed using an Agilent Technologies 7890 GC system coupled to
an Agilent Technologies 5975C mass selective detector (MSD). The
chromatograph is equipped with electronic pressure control, split/
splitless and on-column injectors, and an HP5-MS column. Induc-
tively Coupled Plasma-Mass Spectrometry was carried out on an
C
45H67ClFeNO3: C, 70.99; H, 8.87; N,1.84. Found: C, 71.25; H, 9.03; N,
2.10. (MALDI-TOF) m/z (%, ion): 599.445 (100, [M ꢀ FeeCleTHF]þ),
653.375 (40, [M ꢀ CleTHF}]þ), 688.328 (8, [M ꢀ THF}]þ). UVevis
(methanol) lmax, nm (
4.2.3. FeBr(THF)L1 (2)
3 ): 600 (2750), 330 (3950), 250 (6610).
A
THF solution (50 mL) of recrystallized H2L1 (2.00 g,
3.33 mmol) was added dropwise to a NaH suspension (320 mg,
13.3 mmol) in THF at ꢀ78 ꢁC. Upon returning to room temperature,
the sodium salt of the ligand was added dropwise via filter cannula
to a THF solution of anhydrous FeBr3 (0.985 g, 3.33 mmol) at ꢀ78 ꢁC
resulting in an intense purple solution. After stirring for 2 h, the
solvent was removed via vacuum to give a dark purple powder. The
purple product was extracted with minimal toluene and the
resulting solution was filtered through Celite. Crystals suitable for
X-ray diffraction were obtained by slow evaporation of this toluene
solution (2.255 g, 84%). Anal. Calcd for C45H67BrFeNO3: C, 67.08; H,
8.38; N, 1.74. Found: C, 66.87; H, 8.12; N, 2.05. (MALDI-TOF) m/z (%,
ion): 599.445 (40, [M ꢀ FeeBreTHF]þ), 653.363 (100, [M ꢀ Bre
THF}]þ), 734.288 (5, [M-THF]þ), 805.225 (1, [M]þ).

E.F. Chard et al. / Journal of Organometallic Chemistry 737 (2013) 32e39
39
[5] C. Bolm, J. Legros, J. Le Paih, L. Zani, Chem. Rev. 104 (2004) 6217.
[6] J.K. Kochi, Acc. Chem. Res. 7 (1974) 351.
4.3. General catalytic method
[7] W.M. Czaplik, M. Mayer, J. Cvengros, A. Jacobi von Wangelin, ChemSusChem 2
(2009) 396.
Catalyst 1 or 2 as appropriate (0.10 mmol) in CH2Cl2 (3 mL) was
added to a 30 mL Schlenk flask and the solvent removed under
vacuum. Et2O (5 mL) and the alkyl halide (2.0 mmol) were added to
the catalyst under dry nitrogen. A solution of Grignard reagent
(4.0 mmol) was added dropwise under vigorous stirring. The
resulting mixture was stirred at room temperature for 30 min and
the reaction was quenched with HCl(aq) (2.0 M, 5 mL). The organic
phase was extracted with Et2O (5 mL) and dried over MgSO4. The
organic phase was then passed through a plug of silica and the
diethyl ether was removed under vacuum. The resulting products
were analyzed by GCeMS (using dodecane as internal standard)
and NMR spectroscopy (using acetophenone as internal standard).
[8] A. Fürstner, Angew. Chem. Int. Ed. 48 (2009) 1364.
[9] A. Rudolph, M. Lautens, Angew. Chem. Int. Ed. 48 (2009) 2656.
[10] B.D. Sherry, A. Fürstner, Acc. Chem. Res. 41 (2008) 1500.
[11] A. Fürstner, R. Martin, Chem. Lett. 34 (2005) 624.
[12] M. Nakamura, K. Matsuo, S. Ito, B. Nakamura, J. Am. Chem. Soc. 126 (2004)
3686.
[13] R. Martin, A. Fürstner, Angew. Chem. Int. Ed. 43 (2004) 3955.
[14] J.R. Zhou, G.C. Fu, J. Am. Chem. Soc. 125 (2003) 12527.
[15] A.C. Frisch, N. Shaikh, A. Zapf, M. Beller, Angew. Chem. Int. Ed. 41 (2002) 4056.
[16] J.H. Kirchhoff, C.Y. Dai, G.C. Fu, Angew. Chem. Int. Ed. 41 (2002) 1945.
[17] R.R. Chowdhury, A.K. Crane, C. Fowler, P. Kwong, C.M. Kozak, Chem. Commun.
(2008) 94.
[18] R.B. Bedford, M. Huwe, M.C. Wilkinson, Chem. Commun. (2009) 600.
[19] T. Nagano, T. Hayashi, Org. Lett. 6 (2004) 1297.
[20] R.B. Bedford, D.W. Bruce, R.M. Frost, J.W. Goodby, M. Hird, Chem. Commun.
(2004) 2822.
4.4. X-ray crystallography
[21] A. Fürstner, A. Leitner, Angew. Chem. Int. Ed. 41 (2002) 609.
[22] A. Fürstner, A. Leitner, M. Mendez, H. Krause, J. Am. Chem. Soc. 124 (2002)
13856.
[23] R.B. Bedford, D.W. Bruce, R.M. Frost, M. Hird, Chem. Commun. (2005) 4161.
[24] G. Cahiez, V. Habiak, C. Duplais, A. Moyeux, Angew. Chem. Int. Ed. 46 (2007)
4364.
[25] G. Cahiez, C. Duplais, A. Moyeux, Org. Lett. 9 (2007) 3253.
[26] K. Bica, P. Gaertner, Org. Lett. 8 (2006) 733.
[27] R.B. Bedford, M. Betham, D.W. Bruce, A.A. Danopoulos, R.M. Frost, M. Hird,
J. Org. Chem. 71 (2006) 1104.
[28] A. Fürstner, H. Krause, C.W. Lehmann, Angew. Chem. Int. Ed. 45 (2006) 440.
[29] R.B. Bedford, M. Betham, D.W. Bruce, S.A. Davis, R.M. Frost, M. Hird, Chem.
Commun. (2006) 1398.
[30] Y.Q. Long, X.H. Jiang, R. Dayam, T. Sanchez, R. Shoemaker, S. Sei, N. Neamati,
J. Med. Chem. 47 (2004) 2561.
[31] G.A. Molander, M.D. Elia, J. Org. Chem. 71 (2006) 9198.
[32] M.J. Burns, I.J.S. Fairlamb, A.R. Kapdi, P. Sehnal, R.J.K. Taylor, Org. Lett. 9 (2007)
5397.
[33] J. Luis Serrano, L. Garcia, J. Perez, E. Perez, J. Garcia, G. Sanchez, P. Sehnal, S. De
Ornellas, T.J. Williams, I.J.S. Fairlamb, Organometallics 30 (2011) 5095.
[34] C.M. Crawforth, S. Burling, I.J.S. Fairlamb, R.J.K. Taylor, A.C. Whitwood, Chem.
Commun. (2003) 2194.
[35] C.M. Crawforth, I.J.S. Fairlamb, R.J.K. Taylor, Tetrahedron Lett. 45 (2004) 461.
[36] W. Dohle, D.M. Lindsay, P. Knochel, Org. Lett. 3 (2001) 2871.
[37] S.M. Nobre, A.L. Monteiro, Tetrahedron Lett. 45 (2004) 8225.
[38] M. McLaughlin, Org. Lett. 7 (2005) 4875.
[39] B. Liegault, J.-L. Renaud, C. Bruneau, Chem. Soc. Rev. 37 (2008) 290.
[40] R. Kuwano, M. Yokogi, Org. Lett. 7 (2005) 945.
[41] L. Chahen, H. Doucet, M. Santelli, Synlett (2003) 1668.
[42] K. Hasan, C. Fowler, P. Kwong, A.K. Crane, J.L. Collins, C.M. Kozak, Dalton Trans.
(2008) 2991.
Diffraction data for 1 and 2 were obtained on a Rigaku Saturn 70
CCD area detector with a SHINE optic and graphite monochromated
Mo-Ka radiation, and equipped with an X-stream 2000 low tem-
perature system. The data were processed using CrystalClear [53]
software and corrected for Lorentz and polarization effects and
absorption [54]. Neutral atom scattering factors for all non-
hydrogen atoms were taken from the International Tables for X-
ray Crystallography [55]. The structures were solved by direct
methods using SIR92 [56] and expanded using Fourier techniques
(DIRDIF99) [57]. All non-hydrogen atoms were refined anisotropi-
cally. Hydrogen atoms were refined using the riding model. One of
the t-amyl groups in each structure was disordered with two ori-
entations. These were refined as PARTs with SIMU restraints. The
THF groups in 1 and 2 were also disordered and modeled as another
two PARTS, with SIMU and DFIX (for CeC bonds) restraints.
Anomalous dispersion effects were included in Fcalc [58]; the values
for D D
f0 and f00 were those of Creagh and McAuley [59]. The values
for the mass attenuation coefficients are those of Creagh and
Hubbell [60]. All calculations were performed using the Crystal-
Structure [61] crystallographic software package except for
refinement, which was performed using SHELXL-97 [62].
[43] X. Qian, L.N. Dawe, C.M. Kozak, Dalton Trans. 40 (2011) 933.
[44] A.W. Addison, T.N. Rao, J. Reedijk, J. van Rijn, G.C. Verschoor, J. Chem. Soc.
Dalton Trans. (1984) 1349.
Acknowledgements
[45] R.K. Dean, C.I. Fowler, K. Hasan, K. Kerman, P. Kwong, S. Trudel, D.B. Leznoff,
H.B. Kraatz, L.N. Dawe, C.M. Kozak, Dalton Trans. 41 (2012) 4806.
[46] M.D. Fryzuk, D.B. Leznoff, E.S.F. Ma, S.J. Rettig, V.G. Young, Organometallics 17
(1998) 2313.
[47] G.A. Abakumov, V.K. Cherkasov, M.P. Bubnov, L.G. Abakumova, V.N. Ikorskii,
G.V. Romanenko, A.I. Poddel’sky, Russ. Chem. Bull. 55 (2006) 44.
[48] Y. Ohgo, S. Neya, T. Ikeue, M. Takahashi, M. Takeda, N. Funasaki, M. Nakamura,
Inorg. Chem. 41 (2002) 4627.
Financial support was provided by the Canada Foundation for
Innovation (CFI), Newfoundland and Labrador Research Development
Corporation (NL RDC), the Natural Sciences and Engineering Research
Council(NSERC)ofCanadaandMemorialUniversityofNewfoundland.
Appendix A. Supplementary material
[49] K. Hasan, L.N. Dawe, C.M. Kozak, Eur. J. Inorg. Chem. (2011) 4610.
[50] T. Nagano, T. Hayashi, Org. Lett. 7 (2005) 491.
[51] X. Qian, C.M. Kozak, Synlett 6 (2011) 852.
[52] S.L. Buchwald, C. Bolm, Angew. Chem. Int. Ed. 48 (2009) 5586.
[53] J.W. Pflugrath, Acta Crystallogr. Sect. D. 55 (1999) 1718.
[54] A.C. Larson, Crystallographic Computing, Munksgaard, Copenhagen, 1970.
[55] D.T. Cromer, J.T. Waber, International Tables for X-ray Crystallography, The
Kynoch Press, Birmingham, England, 1974.
CCDC 894993 and 910516 contain the supplementary crystal-
lographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
[56] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. Burla, G. Polidori,
M. Camalli, J. Appl. Cryst. 27 (1994) 435.
Appendix B. Supplementary data
[57] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, R. de Gelder, R. Israel,
J.M.M. Smits, DIRDIF99, University of Nijmegen, Netherlands, 1999.
[58] J.A. Ibers, W.C. Hamilton, Acta Crystallogr. 17 (1964) 781.
[59] D.C. Creagh, W.J. McAuley, International Tables for Crystallography, vol. C,
Kluwer Academic Publishers, Boston, 1992, pp. 219e222.
[60] D.C. Creagh, J.H. Hubbell, International Tables for Crystallography, vol. C,
Kluwer Academic Publishers, Boston, 1992, pp. 200e206.
[61] CrystalStructure 3.7.0, Crystal Structure Analysis Package, Rigaku and Rigaku/
MSC, The Woodlands, TX, 2000e2005.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jorganchem.2013.03.034.
References
[1] D.J. Cardenas, Angew. Chem. Int. Ed. 42 (2003) 384.
[2] A.C. Frisch, M. Beller, Angew. Chem. Int. Ed. 44 (2005) 674.
[3] M.R. Netherton, G.C. Fu, Adv. Synth. Catal. 346 (2004) 1525.
[4] A. de Meijere, F. Diederich, Metal-catalyzed Cross-coupling Reactions, second
ed., Wiley-VCH, Weinheim, 2004.
[62] SHELX-97-Programs for Crystal Structure Determination (SHELXS) and
Refinement (SHELXL), University of Gottingen, Gottingen, Germany,
1997.