TMEDA in iron-catalyzed Kumada coupling: Amine adduct versus homoleptic "ate" complex formation

Article abstract of DOI:10.1002/anie.201308395

The reactions of iron chlorides with mesityl Grignard reagents and tetramethylethylenediamine (TMEDA) under catalytically relevant conditions tend to yield the homoleptic "ate" complex [Fe(mes)₃] ^- (mes=mesityl) rather than adducts of the diamine, and it is this ate complex that accounts for the catalytic activity. Both [Fe(mes) ₃]^- and the related complex [Fe(Bn)₃] ^- (Bn=benzyl) react faster with representative electrophiles than the equivalent neutral [FeR₂(TMEDA)] complexes. Fe^I species are observed under catalytically relevant conditions with both benzyl and smaller aryl Grignard reagents. The X-ray structures of [Fe(Bn) ₃]^- and [Fe(Bn)₄]^- were determined; [Fe(Bn)₄]^- is the first homoleptic σ-hydrocarbyl Fe^III complex that has been structurally characterized. Casting iron in catalytic roles: Chelating diamine ligands such as TMEDA are routinely used in iron-catalyzed cross-coupling reactions of aryl Grignard reagents. However, it was found that under catalytically relevant conditions, there is little evidence for their coordination to iron centers; homoleptic anionic organoiron species are produced instead. Copyright

Full text of DOI:10.1002/anie.201308395

.
Angewandte
Communications
DOI: 10.1002/anie.201308395
Homoleptic Iron Complexes
TMEDA in Iron-Catalyzed Kumada Coupling: Amine Adduct versus
Homoleptic “ate” Complex Formation**
Robin B. Bedford,* Peter B. Brenner, Emma Carter, Paul M. Cogswell, Mairi F. Haddow,
Jeremy N. Harvey, Damien M. Murphy, Joshua Nunn, and Christopher H. Woodall
Abstract: The reactions of iron chlorides with mesityl
Grignard
reagents
and
tetramethylethylenediamine
(TMEDA) under catalytically relevant conditions tend to
yield the homoleptic “ate” complex [Fe(mes)3]À (mes = mesi-
tyl) rather than adducts of the diamine, and it is this ate
complex that accounts for the catalytic activity. Both
[Fe(mes)3]À and the related complex [Fe(Bn)3]À (Bn =
benzyl) react faster with representative electrophiles than the
equivalent neutral [FeR2(TMEDA)] complexes. FeI species are
observed under catalytically relevant conditions with both
benzyl and smaller aryl Grignard reagents. The X-ray
structures of [Fe(Bn)3]À and [Fe(Bn)4]À were determined;
[Fe(Bn)4]À is the first homoleptic s-hydrocarbyl FeIII complex
that has been structurally characterized.
Scheme 2. Previously proposed catalytic cycle.[4]
I
ron-catalyzed couplings of aryl Grignard reagents and alkyl
halides (Scheme 1)[1] are run under a wide variety of
The catalytic cycle shown in Scheme 2 has been proposed
as a model for the Fe/TMEDA-catalyzed coupling of aryl
Grignard reagents with alkyl halides, based on the reactions
of isolated intermediates 1 and 2.[4] Formation of 1 under
1
catalytic conditions was supported by H NMR spectroscopy
of a single reaction mixture similar to that used for catalysis,
though in the absence of the electrophile and without an
excess of the nucleophile.
Scheme 1. Iron-catalyzed coupling of aryl Grignard reagents with alkyl
halides.
Species 1 was formed as the principal paramagnetic iron-
containing species from a mixture of FeCl3, mesMgBr (mes =
mesityl), and TMEDA in a 1:3:8 ratio. One equivalent of the
Grignard reagent presumably serves to reduce the FeIII to FeII
(with concomitant formation of bimesityl), leaving two
equivalents to furnish complex 1. However, under catalytic
conditions, the amount of Grignard reagent relative to iron is
much higher. We found that using this mixture of FeCl3,
Grignard reagent, and TMEDA (1:3:8) in THF, complex 1 is
generated along with at least two other paramagnetic
species,[5] the major one of which is the previously reported
FeII ate complex [Fe(mes)3]À (4).[6,7] Upon addition of more
equivalents of the Grignard reagent, the proportion of 4
increases until, after the addition of four extra equivalents, 4
was the sole paramagnetic species that could be observed by
1H NMR spectroscopy.
conditions, although chelating diamines, such as TMEDA
(tetramethylethylenediamine), are used in many instances.[2,3]
Herein, we report our preliminary studies to elucidate the
possible role(s) of TMEDA under various catalytic condi-
tions. Specifically, we address three questions: 1) What is the
role of TMEDA in catalytic reactions with model compounds
that are derived from bulky aryl Grignard reagents? 2) How
representative are these models for reactions with less bulky
Grignard reagents? 3) Can oxidation states below FeII be
accessed in the presence of TMEDA and Grignard reagents?
[*] Prof. Dr. R. B. Bedford, Dr. P. B. Brenner, P. M. Cogswell,
Dr. M. F. Haddow, Prof. Dr. J. N. Harvey, J. Nunn, Dr. C. H. Woodall
School of Chemistry, University of Bristol
Cantock’s Close, Bristol, BS8 1TS (UK)
E-mail: r.bedford@bristol.ac.uk
Scheme 3 illustrates that complex 4 could also be pro-
duced by the reactions of mesMgBr with complex 1, or with
[{Fe(mes)}2(m-mes)2] (5), or with either FeCl2 or FeCl3 in the
Dr. E. Carter, Prof. Dr. D. M. Murphy
School of Chemistry, Cardiff University, Main Building
Park Place, Cardiff CF10 3TB (UK)
1
presence or absence of TMEDA. The H NMR spectrum of
[**] We thank the EPSRC and Pfizer for funding through the Bristol
Chemical Synthesis Doctoral Training Centre (P.M.C.) and for
student (J.N.) and postdoctoral (P.B.B., E.C.) support. TMEDA=
tetramethylethylenediamine.
a 1:1 mixture of 4 and TMEDA shows only signals of 4;
increasing the amount of TMEDA (in the absence of
mesMgBr) leads to the formation of small amounts of
complex 1 until at 10 equivalents the ratio of 4/1 is approx-
imately 1:1.3.[5]
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201308395.
1804
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1804 –1808

Angewandte
Chemie
Sterically demanding ligands often kinetically stabilize
organometallic complexes with respect to reductive elimina-
tion. Hence, the behavior of mesityl iron species may not
reflect that of other Grignard reagents. Indeed, aryl Grignard
reagents with 2,6-dimethyl groups perform poorly in cross-
coupling reactions, giving either low yields after many hours,[4]
or no reaction.[3a,15] Therefore, we also investigated the
reactivity of sterically less demanding Grignard reagents,
including BnMgCl. This species undergoes cross-coupling in
the presence of iron(III) chloride and TMEDA, albeit
accompanied by some homo-coupling (Scheme 5).
Scheme 3.
Complex 4 was also found to be the only paramagnetic
species that can be observed by 1H NMR analysis of the
reaction mixture of a representative catalytic cross-cou-
pling.[8] This does not preclude the presence of 1 at a concen-
tration below the NMR detection limit, and if 1 were able to
react much more rapidly with the electrophile than 4 does,
then it could still be responsible for catalytic turnover despite
its very low concentration. However, both the data in
Scheme 4 and reaction profiles (Supporting Information,
Scheme 5. Iron-catalyzed cross-couplings of benzyl Grignard reagents.
Yields were determined by H NMR spectroscopy. [a] 1,2-Diphenyl-
ethane and 1,2-di-(3-methoxyphenyl)ethane were produced in 35% and
14% yield, respectively. R=3-methoxybenzyl. [b] 1,2-diphenylethane
(8%) and traces of hexadiene were formed. R=allyl.
1
The reaction of FeCl2 with BnMgCl (4 equiv) gave the
new 12-electron ate complex [FeBn3]À (8), which is stable at
room temperature. Complex 8 could also be produced from
FeCl3 or [NEt4][FeCl4] with BnMgCl (ꢀ 4 equiv;
Scheme 6).[16] Whereas TMEDA could not efficiently displace
the mesityl group in 4 to yield 1, excess TMEDA reacted with
8 to yield [FeBn2(TMEDA)] (9)[17] and the reaction with
(À)-sparteine gave 10 (Scheme 6).[18] In both cases the
reaction was reversed by the addition of excess BnMgCl.
Complex 8 reacted with an excess of 3-methoxybenzyl
bromide to give 6 and the homocoupled products at À408C,
as seen for the catalytic reaction.[9] By contrast, the TMEDA
complex 9 did not react under the same conditions, which
indicates that this TMEDA adduct cannot be part of the
catalytic cycle.
Scheme 4.
Figures S5 and S6)[9] show that complex 4 reacts significantly
faster with bromooctane than 1 does. Furthermore, the
reaction with 1 displays a pronounced induction time of at
least five minutes, which demonstrates conclusively that it
cannot be part of the catalytic cycle.
The X-ray structure of 8 with a mixed magnesium halide/
triflate counterion is shown in Figure 1.[19] Structurally
characterized mononuclear homoleptic s-organometallic
iron(II) complexes are rare and, unlike 8, either rely on
Taken together, the data above indicate that TMEDA is
not coordinated to the complexes that are observed under
catalytic conditions, and that TMEDA is unlikely to play
a role in the primary catalytic cycle. Therefore, homoleptic
complex 4 should be considered in place of complex 1 in the
catalytic cycle shown in Scheme 2.[10,11] What then is the role
of TMEDA?[12] We found that the yield of cross-coupled
product in a representative catalytic reaction was not
influenced by the presence or absence of TMEDA.[13]
However, in the latter case, a smaller amount of starting
material was recovered, and greater amounts of octene and
octane were formed, which suggests that TMEDA must play
a role in suppressing competing, unselective pathways.[14] It
seems likely that TMEDA traps intermediates that are
formed off-cycle and would otherwise lead to less selective
manifolds being populated.
Scheme 6.
Angew. Chem. Int. Ed. 2014, 53, 1804 –1808
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1805

.
Angewandte
Communications
coordinate to FeII benzyl intermediates, it does not appear to
coordinate to FeIII in the presence of magnesium salts.
The formation of the FeIII species 13 from 8 presumably
occurs through disproportionation, which implies that iron
species with oxidation states below FeII must also have been
formed. Although we did not observe these reduced species
directly, the reaction of 8 with the chelating diphosphine 1,2-
bis(diphenylphosphino)benzene (dpbz) that was followed by
EPR spectroscopy[9] revealed the formation of the previously
reported FeI complex 14,[25] whereas crystals isolated from the
reaction mixture were confirmed to be 13a by X-ray
crystallography).
Figure 1. Structure of [{(thf)3Mg}2(m-Cl)2(m-OTf)][FeBn3], [{(thf)3Mg}2-
(m-Cl)2(m-OTf)][8]. Ellipsoids set at 50% probability; hydrogen atoms
are omitted for clarity.
Finally, we turned our attention to the Fe/TMEDA-
catalyzed coupling reactions of smaller aryl Grignard
reagents. Unlike with the mesityl Grignard reagent, which
gives yellow homogeneous solutions throughout the course of
the reaction, the appearance of the reaction mixtures for
coupling reactions of less sterically encumbered aryl Grignard
reagents is highly dependent on the reaction conditions.
Rapid addition of the Grignard reagent results in black
suspensions of catalytically active zero-valent iron nano-
particles;[15] dropwise addition typically gives red transient
intermediates in otherwise yellow solutions.[26] The facile
reduction to Fe0 nanoparticles[27,28] contrasts profoundly with
the stability of the FeII mesityl complexes described above,
reflecting the stability of the latter with respect to reductive
elimination.[29]
bulky ligands, such as [Fe(mes*)2] (mes* = C6H2-2,4,6-tBu3)
and [Fe{C(SiMe)3}2],[20,21] or they are ate complexes with
stabilizing Li···C interactions, such as [Li(Et2O)2][Li-
(dioxane)][FePh4] (11), [Li3(Me)(Et2O)2][FeMe4] (12), and
[Li(Et2O)]2[Fe(1-naphthyl)4].[22,23]
The geometry at the iron center of 8 is approximately
trigonal planar with C-Fe-C angles ranging from 116.1(3) to
À
124.3(2)8. The Fe C bond lengths range from 2.088(5) to
2.115(6) ꢀ, and are longer than those in [FeMes*2] and
[Fe{C(SiMe)3}2], but similar to those for the anionic complex
11.
When samples of complex 8 were left to crystallize, we
reproducibly obtained crystals of a second species: the
13-electron homoleptic iron(III) tetrabenzyl ate complex 13
with a [MgCl(thf)5]+ counterion (13a; Figure 2). To the best
of our knowledge, the only structurally characterized homo-
In an attempt to study the transient red intermediates that
were observed in catalytic reactions with dropwise addition of
the Grignard reagent,[30] we reacted 4-TolylMgBr with FeCl3
in THF in the presence or absence of TMEDA at À308C. This
gave red solutions with decomposition of the iron-containing
species above À208C. Definitive structural information on
the nature of the various species that are present in these
solutions has so far been elusive,[31] but EPR spectra of the
solutions (with and without TMEDA) revealed the presence
of an S = 1= species, which is consistent with the formation of
2
Figure 2. Structures of 13, with [MgCl(thf)5]+ (left; 13a) and
[Mg3Cl5(TMEDA)3]+ (right; 13b) as the counterions. Ellipsoids set at
50% probability; hydrogen atoms and toluene solvate (13b) are
omitted for clarity.
an FeI complex.[5] Furthermore, addition of dpbz led to the
formation of the previously reported iron(I) phosphine
complex [Fe(4-tolyl)(dpbz)2].[25]
In summary, we have shown that the reactions of iron
chlorides with bulky aryl Grignard reagents and TMEDA
under catalytically relevant conditions tend to yield homo-
leptic organoiron complexes rather than adducts of the
diamine. Whereas TMEDA does (reversibly) react with
a homoleptic tribenzyl FeII ate complex, it does not react
with a tetrabenzyl FeIII ate species, but prefers to coordinate
to the magnesium-based counterion instead. Both [Fe(mes)3]À
and [Fe(Bn)3]À react faster with electrophiles than the
equivalent neutral [FeR2(TMEDA)] complexes.
In light of the organoiron complexes that were observed
in this study, catalytic pathways for which the lowest oxidation
state is FeII appear plausible with bulky aryl groups (such as
mesityl) that resist reductive elimination. However, the facile
formation of zero-valent nanoparticles with smaller aryl
Grignard reagents and the observed formation of FeI species
with both benzyl and tolyl Grignard reagents all strongly
suggest the involvement of catalytically relevant species with
oxidation states below FeII when Grignard reagents with
leptic s-organometallic iron(III) complex that has been
reported previously is the approximately square planar
[Li(thf)4][Fe(C6Cl5)4].[24] By contrast, the iron center in
complex 13a is essentially tetrahedral with C-Fe-C angles in
À
the range of 106.05(7)–112.54(7)8. The Fe C bond lengths in
13a range from 2.081(2) to 2.101(2) ꢀ and are comparable to
À
À
the Fe C bond length of the terminal Fe CH3 moiety in the
tetrahedral FeII complex 12 (2.095(4) ꢀ) despite the differ-
ence in oxidation state, but are significantly shorter than the
À
other Fe C bonds in 12 (average 2.185(4) ꢀ), which feature
secondary C···Li interactions.[22] A second crystal structure for
the anion 13 was obtained after reaction of complex 8 and
TMEDA (13b; Figure 2). In this case, the counter-cation is
[Mg3Cl5(TMEDA)3]+, indicating a pronounced preference for
coordination of the TMEDA to magnesium rather than to the
FeIII center. Therefore, although TMEDA can (reversibly)
1806 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1804 –1808

Angewandte
Chemie
[16] The 1H NMR spectrum of 8 showed peaks at d 945.4, 31.4, À49.5,
substituents that are smaller than the mesityl moiety are
exploited.[32]
and À76.5 ppm. In several of the spectra, a small, unassigned
peak at d 65.5 ppm was observed. The measured meff (5.7 BM) of
the solution is consistent with a high-spin FeII complex.
[17] a) D. H. Hill, A. Sen, J. Am. Chem. Soc. 1988, 110, 1650; b) D. H.
Hill, M. A. Panez, A. Sen, J. Am. Chem. Soc. 1994, 116, 2889.
[18] S. C. Bart, E. J. Hawrelak, A. K. Schmisseur, E. Lobkovsky, P. J.
Chirik, Organometallics 2004, 23, 237.
Received: September 25, 2013
Revised: October 31, 2013
Published online: January 21, 2014
Keywords: catalysis · cross-coupling · Grignard reagents · iron
.
[19] The counterion was formed by metathesis with trimethylsilyltri-
flate, which enabled the crystal structure to be solved. For details
of a poor-quality structure of [{(thf)3Mg}2(m-Cl)3][FeBn3], see the
Supporting Information.
[20] H. Mꢁller, W. Seidel, H. Gçrls, Angew. Chem. 1995, 107, 386;
Angew. Chem. Int. Ed. Engl. 1995, 34, 325.
[21] A. M. LaPointe, Inorg. Chim. Acta 2003, 345, 359.
[22] A. Fꢁrstner, R. Martin, H. Krause, G. Seidel, R. Goddard, C. W.
Lehmann, J. Am. Chem. Soc. 2008, 130, 8773.
[23] T. A. Bazhenova, R. M. Lobovskaya, R. P. Shibaeva, A. K.
Shilova, M. Gruselle, G. Leny, E. Deschamps, J. Organomet.
Chem. 1983, 244, 375.
[24] P. J. Alonso, A. B. Arauzo, J. Forniꢃs, M. Angeles Garcꢄa-
Monforte, A. Martꢄn, J. I. Martꢄnez, B. Menjꢅn, C. Rillo, J. J.
Sꢆiz-Garitaonandia, Angew. Chem. 2006, 118, 6859; Angew.
Chem. Int. Ed. 2006, 45, 6707.
[25] C. J. Adams, R. B. Bedford, E. Carter, N. J. Gower, M. F.
Haddow, J. N. Harvey, M. Huwe, M. A. Cartes, S. M. Mansell,
C. Mendoza, D. M. Murphy, E. C. Neeve, J. Nunn, J. Am. Chem.
Soc. 2012, 134, 10333.
[1] For reviews, see: a) C. Bolm, J. Legros, J. Le Paih, L. Zani, Chem.
Rev. 2004, 104, 6217; b) A. Fꢁrstner, R. Martin, Chem. Lett. 2004,
33, 624; c) B. D. Sherry, A. Fꢁrstner, Acc. Chem. Res. 2008, 41,
ˇ
1500; d) W. M. Czaplik, M. Mayer, J. Cvengros, A. J. von Wan-
gelin, ChemSusChem 2009, 2, 396.
[2] For stoichiometric addition of TMEDA, see: M. Nakamura, K.
Matsuo, S. Ito, E. Nakamura, J. Am. Chem. Soc. 2004, 126, 3686.
[3] For the use of catalytic amounts of TMEDA, see: a) R. B.
Bedford, D. W. Bruce, R. M. Frost, M. Hird, Chem. Commun.
2005, 4161; b) G. Cahiez, V. Habiak, C. Duplais, A. Moyeux,
Angew. Chem. 2007, 119, 4442; Angew. Chem. Int. Ed. 2007, 46,
4364.
[4] D. Noda, Y. Sunada, T. Hatakeyama, M. Nakamura, H.
Nagashima, J. Am. Chem. Soc. 2009, 131, 6078.
[5] See the Supporting Information for spectra.
[6] a) V. W. Seidel, K.-J. Lattermann, Z. Anorg. Allg. Chem. 1982,
488, 69; b) M. Irwin, R. K. Jenkins, M. S. Denning, T. Krꢂmer, F.
Grandjean, G. J. Long, R. Herchel, J. E. McGrady, J. M.
Goicoechea, Inorg. Chem. 2010, 49, 6160.
[7] The spectrum is comparable to that of a genuine sample of 4 that
was prepared according to Ref. [6a].
[26] See the Supporting Information for a video.
[27] It has been suggested that an Fe0 complex can be produced
under catalytically relevant conditions (see Ref. [22]), but this
was based on an erroneous crystal structure determination. The
structure is better modelled as an FeII dihydride. For the
incorrect structure determination, see: T. A. Bazhenova, R. M.
Lobkovskaya, R. P. Shibaeva, A. E. Shilov, A. K. Shilova, M.
Gruselle, G. Leny, B. Tchoubar, J. Organomet. Chem. 1983, 244,
265. For the corrected structure, see: J. M. Jefferis, G. S. Giro-
lami, Organometallics 1998, 17, 3630.
[8] Cross-coupling of OctBr and mesMgBr catalyzed by FeCl3/
TMEDA as described in Ref. [4]. 1H NMR spectrum recorded at
room temperature after 70 min (ca. 10% conversion into 3).
[9] See the Supporting Information.
[10] 1H NMR spectroscopy of the reaction of
4 with OctBr
(0.8 equiv) indicated the presence of residual amounts of 4
along with peaks at d 131, 106, and 30 ppm, tentatively assigned
to the complex [FeBr(mes)2]À, as a very similar spectrum is
obtained on reacting 5 with [NBu4]Br. We therefore suggest that
complexes 1 and 2 should be replaced with 4 and [FeBr(mes)2]À,
respectively, in the highly simplified cycle in Scheme 2.
[28] An FeÀII/Fe0 pathway has been proposed because stoichiometric
and catalytic reactions of [Li(TMEDA)]2[Fe(h2-C2H4)4] are
much faster than those of model complexes in oxidation states
0, I, or II (see Ref. [22]). As these models contain a cyclo-
pentadienyl (Cp) or
a pentamethylcyclopentadienyl (Cp*)
[11] For a review summarizing early suggestions on the role of
ligand, a simpler explanation may be that these ligands, and
not the oxidation state of the pre-catalysts, are responsible for
the observed rate retardation. The FeÀII/Fe0 hypothesis suggests
that the tested model Fe0 complex, namely [Li(TMEDA)]-
[CpFe(C2H4)2], should be as effective as the FeÀII complex if the
Cp ligand did not hamper productivity; however, it displayed the
lowest catalytic activity.
homoleptic FeII ate complexes in C C bond formation, see: T.
À
Kauffmann, Angew. Chem. 1996, 108, 401; Angew. Chem. Int.
Ed. Engl. 1996, 35, 386.
[12] TMEDA will also coordinate to ArMgX, increasing its nucleo-
philicity and thus the rates of transmetalation; for a discussion,
see: O. Vechorkin, V. Proust, X. Hu, J. Am. Chem. Soc. 2009, 131,
9756.
[13] The cross-coupling of OctBr and mesMgBr catalyzed by FeCl3/
TMEDA as described in Ref. [4] gave 3 in 35% yield. Repeating
the reaction without TMEDA gave 3 in a yield of 36%. The
former reaction gave 64% of recovered bromooctane, < 1%
octane, and < 1% octene, the latter gave these compounds in
57%, 2%, and 5% yield, respectively.
[29] The nanoparticles may simply act as a resting-state reservoir for
active homogeneous species in higher oxidation states. Indeed,
they react rapidly when added dropwise to a large excess of the
electrophile (see the Supporting Information for a video), which
suggests that at least a part of, if not the entire catalytic cycle may
occur in solution.
[30] A previous study focused on the isolation of model complexes
using ArLi rather than ArMgX reagents under similar con-
ditions (Ref. [22]); however, it cannot be assumed that the same
species are formed. Indeed, lithium reagents give significantly
poorer performance in iron-catalyzed cross-coupling reactions.
For instance, the reaction of CyBr (Cy = cyclohexyl) with
PhMgBr catalyzed by FeCl2 (5 mol%) and TMEDA
(10 mol%) at RT gives CyPh in 76% yield, the equivalent
reaction with PhLi provides CyPh in only 38%; also see example
in Ref. [22].
[14] This is reflected in both the lower selectivities of 4 and 5 in the
reactions described in Scheme 4 and in the different appearances
of the reaction mixtures, with the less selective reactions yielding
black suspensions of reduced iron. Early in the reaction of OctBr
with 4, the selectivity for 3 is high, but rapidly decreases as the
reaction proceeds (Figure S6). This implies that the iron species
that are formed while 4 is consumed account for the competing
side-reactions.
[15] R. B. Bedford, M. Betham, D. W. Bruce, S. A. Davis, R. M.
Frost, M. Hird, Chem. Commun. 2006, 1398.
Angew. Chem. Int. Ed. 2014, 53, 1804 –1808
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1807

.
Angewandte
Communications
[31] FeCl2 and 3 or 5 equivalents of 4-tolylMgCl (THF, À308C) gave
a red solution with paramagnetically shifted peaks in the
1H NMR spectrum at d 292.3, 117.9, and 99.2 ppm. The stoichi-
ometry is consistent with the formation of the ate complex [Fe(4-
tol)3]À, analogous to 4. Repeating the reaction with 1–10 equiv-
alents of TMEDA gave increasing amounts of a second species
(d 221.7, 95.3, and 89.4 ppm), possibly owing to the formation of
[Fe(4-tol)4]2À, an analogue of [Fe(Ph)4]2À (Ref. [22]), or [FeCl(4-
tol)3]À. No peaks for TMEDA coordinated to a paramagnetic
iron center were observed, but we cannot rule out the possibility
that TMEDA coordinates to an NMR-silent Fe center.
[32] For recent examples of the putative involvement of FeI species in
cross-couplings with aryl Grignard reagents, see: a) A. Hed-
strçm, E. Lindstedt, P.-O. Norrby, J. Organometal. Chem. 2013,
748, 51; b) A. Hedstrçm, U. Bollmann, J. Bravidor, P.-O. Norrby,
Chem. Eur. J. 2011, 17, 11991; c) J. Kleimark, A. Hedstrçm, P.-F.
Larsson, C. Johansson, P.-O. Norrby, ChemCatChem 2009, 1, 152.
1808
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1804 –1808