Low-coordinate iron(I) and manganese(I) dimers: Kinetic stabilization of an exceptionally short Fe-Fe multiple bond

Article abstract of DOI:10.1002/anie.201203711

Not just any old iron! The reduction of a bulky guanidinato iron(II) bromide complex yields a three-coordinate iron(I) dimer that possesses the shortest Fe-Fe interaction (2.127 A) reported to date. Magnetic, Moessbauer, and computational studies show the unprecedented compound to contain two high-spin iron(I) centers with significant multiple-bond character. A related dimer containing a rare example of an unsupported, carbonyl-free Mn-Mn bond is also described. Copyright

Full text of DOI:10.1002/anie.201203711

Angewandte
Chemie
DOI: 10.1002/anie.201203711
Metal–Metal Bonds
Low-Coordinate Iron(I) and Manganese(I) Dimers: Kinetic
À
Stabilization of an Exceptionally Short Fe Fe Multiple Bond**
Lea Fohlmeister, Shengsi Liu, Christian Schulten, Boujemaa Moubaraki, Andreas Stasch,
John D. Cashion, Keith S. Murray, Laura Gagliardi, and Cameron Jones*
The fundamental and applied chemistry of metal–metal
bonded complexes has rapidly expanded since Cottonꢀs
landmark report of metal–metal quadruple bonding in the
dianion, [Re2Cl8]2À, nearly 50 years ago.[1,2] Many of the
recent advances in the field have centered on the stabilization
of reactive low oxidation state/low coordination number M–
M bonded complexes using sterically imposing ligand systems.
Representative examples include the singly bonded zinc(I)
and magnesium(I) dimers, [Cp*ZnZnCp*] (Cp* = C5Me5)[3]
and [(DipNacnac)MgMg(DipNacnac)] (DipNacnac = [(DipNC-
Me)2CH], Dip = 2,6-iPr2C6H3),[4] and the quintuply bonded
chromium(I) dimer, [Ar’CrCrAr’] (Ar’ = 2,6-(Dip)2C6H3).[5]
The unprecedented bonding in the latter initiated a number
of studies aimed at preparing complexes with ever shorter
or Ni;[10] R = tBu, N(C6H11)2, or NiPr2), which are isostruc-
tural with 1, and which, for M = Co, possess the shortest
À
known Co Co bonds (ca. 2.14 ꢁ; for M = Ni, ca. 2.29 ꢁ).
However, unlike 1, these more electron-rich systems are
À
paramagnetic, and have lower M M bond orders, owing to
partial filling of metal-based anti-bonding MOs. We were
keen to extend this study to the preparation of the corre-
sponding iron(I) and manganese(I) dimers, as such com-
pounds are unknown and have the potential to exhibit rare
examples of multiple bonding between the metals, as well as
M–M distances intermediate between those of 1 and 2. We
were especially interested in iron(I) dimers, as the current
À
shortest known Fe Fe bond (2.198 ꢁ) exists in a related
mixed valence amidinate bridged complex, [FeI/II2{m-N,N’-
(PhN)2CPh}3] (3),[11] which has a formal bond order of 1.5. It is
of note that the Fe–Fe interaction in this high-spin complex
(S = 7/2) is ferromagnetic in nature, as opposed to the more
common antiferromagnetic interactions seen in related
dimers, such as [Ar’FeIFeIAr’] (4; Fe–Fe = 2.5151(9) ꢁ, S =
0).[12–14] Furthermore, guanidinato/amidinato iron(I) com-
plexes make attractive synthetic targets, as they will almost
certainly be highly reactive species that have considerable
potential for application in areas such as small molecule
activations and enzyme mimicry. (compare with the well-
developed chemistry of closely related, low-valent b-diketi-
minato iron complexes, such as [{(DipNacnac)Fe}2(m-N2)]).[15]
Our preliminary efforts to prepare low coordinate, multiply
bonded iron(I) and manganese(I) dimers are reported herein.
Cr Cr bonds.[6] These culminated in the isolation of quintuply
À
bonded amidinate and guanidinate bridged complexes, for
example, [Cr2{m-N,N’-(DipN)2CR}2] (1; R = H, Me, or NMe2),
which exhibit the shortest known metal–metal bonds (ca.
1.74 ꢁ).[7] As these bonds are essentially derived from the
filling of five metal-based bonding molecular orbitals (MOs),
the compounds are diamagnetic.
We have a strong background in developing bulky
amidinate and guanidinate ligands for the stabilization of
metal(I) dimers.[8] Most relevant to this study are the cobalt(I)
and nickel(I) dimers, [M2{m-N,N’-(DipN)2CR}2] (M = Co (2)[9]
[*] L. Fohlmeister, Dr. C. Schulten, Dr. B. Moubaraki, Dr. A. Stasch,
Prof. K. S. Murray, Prof. C. Jones
The
reduction
of
[{(Piso)FeII(m-Br)}2]
(Piso =
School of Chemistry, Monash University
PO Box 23, Melbourne, VIC, 3800 (Australia)
E-mail: cameron.jones@monash.edu
[(DipN)2CtBu])[16] with the mild and soluble magnesium(I)
reducing agent, [{(MesNacnac)Mg}2] (MesNacnac = [(MesNC-
Me)2CH], Mes = mesityl),[17] was initially carried out in
cyclohexane under an argon atmosphere. Although this led
to the desired iron(I) dimer, [FeI2(m-N,N’-Piso)2],[18] the
complex was susceptible to disproportionation in solution
and consistently co-crystallized with significant amounts (ca.
20%) of the square-planar homoleptic complex, [FeII(k2-
N,N’-Piso)2].[19] This was found to be disordered over the same
molecular site as the iron(I) dimer in the crystal lattice. As
a result, reliable spectroscopic and crystallographic data could
not be obtained for the compound. To overcome this problem,
the bulkier guanidinato iron(II) precursor, [{(Pipiso)FeII(m-
Br)}2] (Pipiso = [(DipN)2C(cis-2,6-Me2NC5H8)][20]), was pre-
pared (see the Supporting Information) and reacted with one
equivalent of [{(MesNacnac)Mg}2] to give the iron(I) dimer,
[FeI2(m-N,N’-Pipiso)2] (5), in good yield (57%), as a dark red-
brown crystalline solid (Scheme 1). It is of note that all
attempts to reduce [{(Pipiso)FeII(m-Br)}2] with more conven-
tional reducing agents, such as elemental Mg or K, led to
S. Liu, Prof. L. Gagliardi
Department of Chemistry and
Minnesota Supercomputing Institute, University of Minnesota
207 Pleasant St, SE Minneapolis, MN-55455-0431 (USA)
Dr. J. D. Cashion
School of Physics, Monash University
PO Box 27, Melbourne, VIC, 3800 (Australia)
[**] We thank the Australian Research Council (C.J. and K.S.M.) and the
U.S. Air Force Asian Office of Aerospace Research and Development
(grant FA2386-11-1-4110 to C.J.). L.G. and S.L. thank the U.S.
Department of Energy, Office of Basic Energy Sciences, Heavy
Element Chemistry program for support (USDOE/DE-SC002183).
The EPSRC is also thanked for access to the UK National Mass
Spectrometry Facility. Nicholas Chilton (Monash) is thanked for
carrying out D calculations.
Supporting information for this article, including details of the
synthesis and characterization of all new compounds, and full
details and references for the crystallographic and computational
studies, is available on the WWW under http://dx.doi.org/10.1002/
anie.201203711.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
The 1H NMR spectra of 5 under argon atmosphere in
solutions of [D6]-benzene and [D12]-cyclohexane are essen-
tially identical and revealed the compound to be para-
magnetic. When these solutions are exposed to a nitrogen
atmosphere, their 1H NMR spectra do not change. This
suggests that the compound does not react with arene
solvents or dinitrogen in solution at ambient temperature
(compare with the formation of [{(Piso)Fe}2(m-N2)] or
[(Piso)Fe(h6-toluene)] by the reduction of [{(Piso)Fe(m-
Br)}2] in the presence of dinitrogen and/or toluene).[16]
Similarly, the UV/Vis spectra of solutions of 5 in toluene
À
and hexane are very similar, further indicating that the Fe Fe
bond of the complex is resistant to reaction with arenes.
The effective magnetic moment of 5 at room temperature
([D6]-benzene, Evans method) was determined to be 8.1 mB,
which is somewhat higher than the spin-only predictions for
two non-interacting high-spin iron(I) centers (meff = 5.48 mB),
and a high-spin, ferromagnetically coupled system with an S =
3 spin state (meff = 6.93 mB). Moreover, solid state magnetic
susceptibility measurements of 5 (Figure 2) showed that its
magnetic moment is largely independent of temperature
Scheme 1. Syntheses of compounds 5 and 6 (by-products omitted).
intractable product mixtures. While compound 5 is extremely
air sensitive, it is stable in solution at room temperature for
several days before signs of disproportionation are visible
(namely, Fe metal deposition).
The molecular structure of 5 (Figure 1) shows it to be
isostructural with 1 and 2, in that the central [Fe2]2+ unit is
bridged by two delocalized guanidinate ligands. Both Fe
centers have T-shaped coordination geometries and the inner
Fe2N4C2 entity is essentially planar. The Fe–Fe distance of
2.1270(7) ꢁ in 5 is shorter than all of the more than 7000 Fe–
Fe interactions listed in the Cambridge Crystallographic
Database,[21] and it is markedly shorter than the sum of two
covalent radii for both low-spin iron (2.64 ꢁ) and high-spin
iron (3.04 ꢁ).[22] No crystallographic or spectroscopic evi-
dence was found for the presence of either bridging or
terminal hydride ligands at the iron centers. These facts
À
strongly suggest the presence of Fe Fe multiple bonding in 5.
Figure 2. Plot of the effective magnetic moment versus temperature
for dimer 5 in an applied field of 1 T. The solid line is a guide.
between 300 and 150 K, remaining at 7.95 mB. It then
decreases slowly to 6.3 mB at 8 K before rapidly decreasing
to 3.3 mB at 2 K. In an attempt to gain information on the
ground state spin, magnetization isotherms were determined
at 2–20 K in fields of 0–5 T (Supporting Information, Fig-
ure S6). Saturation was not achieved at 2 K and high fields,
with M = 2.5 NmB at 2 K and 5 T, which is similar to the results
recently obtained for the mixed-valence compound [FeI/II2{m-
N,N’-(PhN)2CH}3] (saturated value for S = 7/2; mS = 1= low-
2
est).[11b] The low M value signifies spin-orbit coupling, zero-
field splitting, and orbital degeneracy contributions, whereas
the lack of saturation reflects closely spaced energy levels
relative to the ground level (see below). This likely leads to
the high room temperature effective magnetic moment for
5,[23] and although S could not be unambiguously determined
through D value calculations on the compound, it is clear that
the ground state has a significantly high spin value.
Figure 1. Molecular structure of 5 (25% thermal ellipsoids; hydrogen
atoms omitted). Selected bond lengths (ꢀ) and angles (8): Fe(1)–Fe(2)
2.1270(7), N(1)–Fe(1) 1.938(3), N(4)–Fe(1) 1.953(3), N(2)–Fe(2)
1.940(3), N(5)–Fe(2) 1.952(3), N(1)–C(1) 1.349(4), N(2)–C(1)
1.357(4), N(4)–C(33) 1.349(4), N(5)–C(33) 1.348(4), N(1)-C(1)-N(2)
114.7(3), N(4)-C(33)-N(5) 114.7(3), N(1)-Fe(1)-N(4) 175.46(12), N(2)-
Fe(2)-N(5) 175.81(12).
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
To shed further light on the oxidation state of the iron
centers and the spin ground state of 5, the Mçssbauer
spectrum of the compound was measured in zero applied field
at 77 K (Figure S9). The dominant feature of the spectrum is
The manganese(I) analogue of 5 was seen as an attractive
synthetic target because of its potential reactivity, and
because it could exhibit a metal–metal bond order (and
length) between those of 1 and 5. With the aim of accessing
such a compound, attempts to prepare the dimeric mangane-
se(II) precursor complex, [{(Pipiso)MnII(m-Br)}2], were made,
though these were unsuccessful.[25] As a result, attention
turned to the preparation (see the Supporting Information)
and reduction of the related trinuclear complex, [{(Piso)MnII-
(m-Br)}3(THF)2]. Treatment of this complex with 1.5 equiv-
alents of the magnesium(I) reagent [{(MesNacnac)Mg}2]
resulted in the formation of the MnI dimer 6 in moderate
yield, as dark red-green dichroic crystals (Scheme 1). As
a crystalline solid, compound 6 has negligible solubility in
common organic solvents (such as, toluene, diethyl ether, and
THF), which made the acquisition of meaningful solution
spectroscopic data impossible.
a
quadrupole doublet with an isomer shift (IS) of
0.44(1) mmsÀ1 and a large quadrupole splitting (QS) of
2.39(1) mmsÀ1. Although Mçssbauer studies of high-spin
iron(I) compounds are comparatively rare, it is clear from
previous reports that the IS and QS values of such compounds
can be very dependent upon the nature and symmetry of the
surrounding ligands. Despite this, the Mçssbauer data for 5
are similar to those for several b-diketiminato iron(I) com-
pounds, for example, [(tBuNacnac)FeI(h2-HCCPh)] IS =
0.44(1) mmsÀ1, QS = 2.05(2) mmsÀ1 tBuNacnac = [(DipNCt-
(
Bu)2CH]).[24] In contrast, the Mçssbauer spectra of the related
mixed-valence complex [FeI/II2{m-N,N’-(PhN)2CH}3] (both Fe
centers are equivalent on the Mçssbauer timescale)[11b] yield
significantly different IS (0.65 mmsÀ1) and QS (0.32 mmsÀ1
)
Compound 6 is dimeric in the solid state (Figure 3) and
contains an unsupported Mn–Mn interaction, the length of
which (2.7170(9) ꢁ) is indicative of a single bond (compare
values than those found for 5. Although not unambiguous, the
Mçssbauer data for 5 are consistent with it possessing two
high-spin iron(I) centers (see below).
DFT and multiconfigurational CASSCF/PT2 calculations
(see the Supporting Information for full details) were carried
out on
a
simplified model of 5, [Fe2{m-N,N’-[(2,6-
Me2C6H3)N]2CNMe2}2] (5a), to gain further insight into its
À
electronic structure and the nature of the very short Fe Fe
bond in the compound. Geometry optimizations (DFT) were
initially carried out on 5a in its quintet (S = 2) and septet (S =
3) spin states. The geometry of the latter spin state converged
to be very close to that of 5 (with an Fe–Fe distance of
2.118 ꢁ), while the former state optimized with a markedly
À
shorter Fe Fe bond (2.013 ꢁ). Single point calculations
revealed that the septet spin state is favored with respect to
the quintet state by 6.5 kcalmolÀ1 (PT2) or 7.2 kcalmolÀ1
(DFT). Furthermore, the dominant electronic configuration
À
for the septet state (62.3%) yields a formal Fe Fe bond order
of two for 5a. This can be understood when it is considered
that, of the 14 metal-based valence electrons in the active
space of the dimer (Figure S10), eight occupy four bonding
MOs (s-and p-type), four singly occupy anti-bonding MOs
(s* and p*-type), and two singly occupy essentially non-
bonding orbitals (of d and d* character). This gives rise to
a bond order of two, and yields six unpaired electrons for the
S = 3 spin state dimer. However, when partial occupancies of
the CASSCF total ground state natural orbitals are taken into
account, an effective bond order (EBO)[6b] of 1.19 is obtained
(compared with an EBO of 1.29 for the S = 2 state). These
values are greater than those for the mixed-valence system
[FeI/II2{m-N,N’-(PhN)2CH}3], which has been reported as
having a formal Fe–Fe bond order of 1.5 and an EBO of
1.15.[11b] As a result, it is reasonable to conclude that the
exceptionally short Fe–Fe interaction in 5 is due to it
possessing significant multiple bonding character between
two high spin iron(I) centers. Moreover, a good agreement
was found between the experimental Mçssbauer IS and QS
values for 5 (see above), and those calculated for 5a (S = 3,
IS = 0.41 mmsÀ1, QS = 2.04 mmsÀ1). These data provide fur-
ther evidence for the proposed electronic structure of 5.
Figure 3. Molecular structure of 6 (25% thermal ellipsoids; hydrogen
atoms omitted). Selected bond lengths (ꢀ) and angles (8): Mn(1)–
Mn(2) 2.7170(9), Mn(1)–N(2) 2.076(12), C(1)–N(1) 1.308(4), C(1)–
N(2) 1.356(4), Mn(1)–C(18) 2.509(3), Mn(1)–C(19) 2.710(3), Mn(1)–
C(23) 2.766(3), N(1)-C(1)-N(2) 121.7(2), Mn(1)’-Mn(1)-N(2)
143.42(7). Symmetry operation: ’Àx+1, Ày+2, Àz+1.
with 2.90 ꢁ for [Mn2(CO)10]).[26] Interestingly, the ligand has
changed its coordination mode from N,N’-chelating in the
precursor to what is best described as N-,h3-arene-chelating in
6. Moreover, the backbones of the Piso ligands in the dimer
are largely electronically localized, and are therefore acting as
imino-amides. Although carbonyl-free manganese(I) dimers
are extremely rare, compound 6 is closely related to the b-
diketiminato coordinated complex, [{(DipNacnac)Mn}2] (7),
though in that compound the ligands N,N’-chelate the metal
centers.[27] The Mn Mn bond length in 7 (2.721(1) ꢁ) is
À
essentially identical to that in 6.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
The magnetic susceptibility of compound 6 in the solid
state was measured at 2–300 K (Figure 4) and the room
temperature effective magnetic moment was found to be
4.39 mB per dimer. This suggests a high spin state for the
magnetic coupling between two S1 = S2 = 5/2 spin centers,
very similar to that described for 7.
In conclusion, a bulky guanidinato-stabilized iron(I)
À
dimer, with the shortest Fe Fe bond reported to date, has
been prepared and shown by magnetic, Mçssbauer, and
computational studies to contain two high-spin iron(I) centers
that display significant multiple-bond character. A related
antiferromagnetically coupled manganese(I) dimer, contain-
À
ing a rare example of an unsupported, carbonyl free Mn Mn
single bond, has also been synthesized and fully characterized.
Investigations into the utility of these highly reactive com-
pounds as reagents for small molecule activations are under-
way in our laboratory.
Received: May 14, 2012
Revised: June 14, 2012
Published online: && &&, &&&&
Keywords: density functional calculations · iron · magnesium ·
.
manganese · metal-metal interactions
Figure 4. The temperature dependence (2–300 K) of the magnetic
susceptibility (main plot) and effective magnetic moment (insert plot)
of 6 (per dimer). The solid line (main plot) is that calculated for an
S=5/2:5/2 dimer model using the parameters given in the text.
[1] F. A. Cotton, L. A. Murillo, R. A. Walton, Multiple Bonds
Between Metal Atoms, 3rd ed., Springer, Berlin, 2005.
[2] F. A. Cotton, N. F. Curtis, B. F. J. Johnson, J. T. Mague, J. S.
Wood, C. B. Harris, W. R. Robinson, S. J. Lippard, Science 1964,
145, 1305 – 1307.
[3] I. Resa, E. Carmona, E. Gutierrez-Puebla, A. Monge, Science
2004, 305, 1136 – 1138.
[4] S. P. Green, C. Jones, A. Stasch, Science 2007, 318, 1754 – 1757.
[5] T. Nguyen, A. D. Sutton, M. Brynda, J. C. Fettinger, G. J. Long,
P. P. Power, Science 2005, 310, 844 – 847.
[6] a) F. R. Wagner, A. Noor, R. Kempe, Nat. Chem. 2009, 1, 529 –
536; b) M. Brynda, L. Gagliardi, B. O. Roos, Chem. Phys. Lett.
2009, 471, 1 – 10.
[7] a) C.-W. Hsu, J.-S. K. Yu, C.-H. Yen, G.-H. Lee, Y.-C. Tsai,
Angew. Chem. 2008, 120, 10081 – 10084; Angew. Chem. Int. Ed.
2008, 47, 9933 – 9936; b) A. Noor, G. Glatz, R. Mꢂller, M. Kaupp,
S. Demeshko, R. Kempe, Z. Anorg. Allg. Chem. 2009, 635, 1149 –
1152.
[8] C. Jones, Coord. Chem. Rev. 2010, 254, 1273 – 1289.
[9] a) C. Jones, C. Schulten, R. P. Rose, A. Stasch, S. Aldridge, W. D.
Woodul, K. S. Murray, B. Moubaraki, M. Brynda, G. La Mac-
chia, L. Gagliardi, Angew. Chem. 2009, 121, 7542 – 7546; Angew.
Chem. Int. Ed. 2009, 48, 7406 – 7410; b) for related rhodium(I)
chemistry, see: C. Jones, D. P. Mills, A. Stasch, Dalton Trans.
2008, 4799 – 4804.
manganese(I) ions (3d54s1 or 3d6), but the significantly lower
experimental value compared to the spin-only value for two
non-interacting manganese(I) ions (3d54s1: 9.80 mB or 3d6:
6.93 mB) is indicative of antiferromagnetic coupling between
the metal centers, with a Curie temperature (Tc) significantly
higher than 300 K. The decrease of c from 300 K to 15 K
points to an S = 0 coupled ground state at 0 K for the dimer,
even though c increases again below 15 K, presumably due to
the unavoidable presence of a small amount of paramagnetic
impurity. A good fit to an S = 5/2 (per Mn) Heisenberg model
(À2 JS1.S2) gave parameters: g = 2.00, J = À40.2 cmÀ1 (frac-
tion monomer 0.002). Use of an S = 2 (per Mn) model gave
similar parameters (see the Supporting Information), as the
S = 0, 1, and 2 coupled states are the only ones thermally
populated up to 300 K. It is of note that the related MnI dimer
7 showed broadly similar magnetic behavior to 6.[27]
Geometry optimizations (DFT) carried out on a simplified
model of 6, [{{tBuC[(2,6-Me2C6H3)N]2}Mn}2] (6a), showed
that only the geometry of the S = 5 spin state converged (Mn–
Mn = 2.741 ꢁ) close to the solid state geometry of 6. A
multiconfigurational analysis of 6a (S = 5) revealed a single
dominating electronic configuration (95.5%) and an EBO of
0.42 for the Mn–Mn interaction. From the results of previous
studies,[6b,11b] it is not surprising that this measure of bond
order provides a value less than unity, and it is clear that there
is a significant Mn–Mn interaction in 6a. The natural bond
orbitals for the compound show that its unpaired electrons are
essentially localized in the metal-centered 3d orbitals, while
the Mn 4s electrons are involved in forming a single bond
between the manganese centers (Figure S11). On the basis of
these calculations, the best interpretation of the results from
the magnetic measurements on 6 is that there is antiferro-
[10] C. Jones, C. Schulten, L. Fohlmeister, A. Stasch, K. S. Murray, B.
Moubaraki, S. Kohl, M. Z. Ertem, L. Gaglaiardi, C. J. Cramer,
Chem. Eur. J. 2011, 17, 1294 – 1303.
[11] a) F. A. Cotton, L. M. Daniels, L. R. Falvello, J. H. Matonic,
C. A. Murillo, Inorg. Chim. Acta 1997, 256, 269 – 275; b) for
a recent comprehensive spectroscopic and computational anal-
ysis of the closely related complex, [FeI/II2{m-N,N’-(PhN)2CH}3],
see: C. M. Zall, D. Zherebetskyy, A. L. Dzubak, E. Bill, L.
Gagliardi, C. C. Lu, Inorg. Chem. 2012, 51, 728 – 736.
[12] T. Nguyen, W. A. Merrill, C. Ni, H. Lei, J. C. Fettinger, B. D.
Ellis, G. J. Long, M. Brynda, P. P. Power, Angew. Chem. 2008,
120, 9255 – 9257; Angew. Chem. Int. Ed. 2008, 47, 9115 – 9117.
[13] For examples of paramagnetic compounds with Fe–Fe bonds,
see: a) H. Lei, J.-G. Guo, J. C. Fettinger, S. Nagase, P. P. Power, J.
Am. Chem. Soc. 2010, 132, 17399 – 17401; b) C. R. Hess, T.
Weyhermꢂller, E. Bill, K. Wieghardt, Angew. Chem. 2009, 121,
3758 – 3761; Angew. Chem. Int. Ed. 2009, 48, 3703 – 3706; c) A.
Klose, E. Solari, C. Floriani, A. Chiesi-Villa, C. Rizzoli, N. Re, J.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Am. Chem. Soc. 1994, 116, 9123 – 9135; d) F. A. Cotton, L. M.
Daniels, J. H. Matonic, C. A. Murillo, Inorg. Chim. Acta 1997,
256, 277 – 282.
[22] B. Cordero, V. Gomez, A. E. Platero-Prats, M. Reves, J.
Echeverria, E. Cremades, F. Barragan, S. Alvarez, Dalton
Trans. 2008, 2832 – 2838.
[14] For a discussion on ferromagnetic Fe – Fe interactions, see G. H.
Timmer, J. H. Berry, C. R. Chim. 2012, 15, 192 – 201.
[15] a) P. L. Holland, Acc. Chem. Res. 2008, 41, 905 – 914; b) M. M.
Rodriguez, E. Bill, W. W. Brennessel, P. L. Holland, Science
2011, 334, 780 – 783.
[16] R. P. Rose, C. Jones, C. Schulten, S. Aldridge, A. Stasch, Chem.
Eur. J. 2008, 14, 8477 – 8480.
[17] a) A. Stasch, C. Jones, Dalton Trans. 2011, 40, 5659 – 5672; b) S. J.
Bonyhady, C. Jones, S. Nembenna, A. Stasch, A. J. Edwards, G. J.
McIntyre, Chem. Eur. J. 2010, 16, 938 – 955.
[23] To rule out significant contributions from paramagnetic impur-
ities to the room temperature effective magnetic moment of 5,
the field dependency (at 0.01, 0.1, and 1 T) of meff for the
compound was examined (Supporting Information, Figure S7).
A small field dependency was observed below 70 K, thus
indicating the presence of a small amount of impurity. This is
unavoidable for such an air-sensitive compound. However, in the
temperature range of 150–300 K, the experiments pointed to
a negligible contribution from the impurity to the value of meff for
5.
À
[18] From one reduction a low yield of the C H activated product,
[24] S. A. Stoian, Y. Yu, J. M. Smith, P. L. Holland, E. L. Bominaar,
E. Mꢂnck, Inorg. Chem. 2005, 44, 4915 – 4922.
[(PisoÀH)FeII(m-H)Mg(THF)(MesNacnac)]
(PisoÀH = DipNC-
(tBu)N{2-iPr-6-[C(H)(Me)(CH2)]-C6H3}), was obtained. See
the Supporting Information for details of the crystal structure
and formation of this compound.
[25] The related guanidinato complexes, [{(Giso)MnII(m-X)2}2]
(Giso = [(DipN)2CN(C6H11)2], X = Cl or Br) were prepared
(see the Supporting Information), but their reductions gave
complex product mixtures.
[19] FeII(k2-N,N’-Piso)2] was subsequently intentionally synthesized
and fully characterized (see the Supporting Information).
Closely related square planar bis(amidinato)-iron(II) complexes
have been reported. C. A. Nijhuis, E. Jellema, T. J. J. Sciarone,
A. Meetsma, P. H. M. Budzelaar, B. Hessen, Eur. J. Inorg. Chem.
2005, 2089 – 2099.
[26] R. Bianchi, G. Gervasio, D. Marabello, Inorg. Chem. 2000, 39,
2360 – 2366.
[27] J. Chai, H. Zhu, C. Stꢂckl, H. W. Roesky, J. Magull, A. Bencini,
A. Caneschi, D. Gatteschi, J. Am. Chem. Soc. 2005, 127, 9201 –
9206. See also, the dianionic MnI dimer, [Mn2{(NDip)2SiMe2}2]2À
(Mn – Mn: 2.78198 ꢁ): D.-Y. Lu, J.-S. Yu, T.-S. Kuo, G.-H. Lee,
Y. Wang, Y.-C. Tsai, Angew. Chem. 2011, 123, 7753 – 7757;
Angew. Chem. Int. Ed. 2011, 50, 7611 – 7615.
[20] G. Jin, C, Jones, P. C. Junk, K.-A. Lippert, R. P. Rose, A. Stasch,
New J. Chem. 2009, 33, 64 – 75.
[21] As determined by a survey of the Cambridge Crystallographic
Database, May, 2012.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Metal–Metal Bonds
L. Fohlmeister, S. Liu, C. Schulten,
B. Moubaraki, A. Stasch, J. D. Cashion,
K. S. Murray, L. Gagliardi,
C. Jones*
&&&& — &&&&
Low-Coordinate Iron(I) and
Manganese(I) Dimers: Kinetic
Stabilization of an Exceptionally Short
Not just any old iron! The reduction of
a bulky guanidinato iron(II) bromide
complex yields a three-coordinate iron(I)
dimer that possesses the shortest Fe–Fe
interaction (2.127 ꢀ) reported to date.
Magnetic, Mçssbauer, and computa-
tional studies show the unprecedented
compound to contain two high-spin
iron(I) centers with significant multiple-
bond character. A related dimer contain-
ing a rare example of an unsupported,
À
Fe Fe Multiple Bond
À
carbonyl-free Mn Mn bond is also de-
scribed.
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!