C. Feldmann et al.
+
1.3
ꢀ1
ꢀ1
of Sn
is expected. This analysis is also consistent with the
3.9 mms ; SnO: 2.7 mms ) and molecular FeꢀSn clusters
ꢁ
0
+III
ꢀ1
formal oxidation states Fe and Sn
for a recently discovered barbell-shaped [FeI(CO)
cluster. In accord with these formal oxidation states, the
Fe Mulliken charge
ACHTUNGTRENNUNG{ SnI} I ] and [FeI(CO) ACHNUTGTREUNNGN( SnI ) ] , whereas the Sn Mulliken
6 4 3 3
charge is much more positive for [FeI(CO)
Table 3).
that were adopted
(e.g., [{(SnCl )Fe(CO) } ]: 2.01 mms ; [{(SnBr )Fe(CO) } ]:
2
4
2
2
4 2
ꢀ
ꢀ1 [18]
A
H
U
G
R
N
N
(SnI ) ]
2.40 mms ).
The quadrupole splitting parameter of
3
3 2
[16]
ꢀ1
DE =1.5(1) mms reflects the non-cubic electron density
distribution around the tin nuclei. In sum, the experimental
data are again in agreement with approximating a formal
oxidation state of Sn . In addition to Sn, we aimed at
Fe Mçssbauer spectroscopy. Owing to the high concentra-
Q
[17]
is very similar for [{Fe(CO) } -
3
4
2
ꢀ
ꢀ
2
ꢀ
3 2
+1.3
119
A
H
U
G
R
N
U
G
3
5
7
(
tion of iodine, however, the iron atoms are strongly shield-
ed, so that no dedicated signal was obtained. Only a signifi-
cant enrichment of the sample with Fe would solve this
problem.
[
17]
Table 3. PBE0/dhf-SVP Mulliken charges
q(Fe) and q(Sn) of
5
7
2
+
[
{Fe(CO)
3
}
4
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
{SnI}
6
]
in comparison with selected reference compounds.
Compound
q(Fe) [e]
q(Sn) [e]
After elucidating the bonding situation and valence state,
the question regarding role and benefit of the ionic liquid
toward the synthesis of 1 and 2 still remains. In view of their
excellent redox stability, ionic liquids have already been
suggested as suitable reaction media for carbonyls. On the
other hand, only few studies report on such reactions to
2
ꢀ
[
[
{Fe(CO)
{Fe(CO)
3
3
}
}
4
4
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
{SnI}
6
6
I
4
]
ꢀ0.87
ꢀ0.92
ꢀ0.90
ꢀ0.89
+0.36
+0.35
+0.52
2
+
A{ SnI}
C
H
T
U
N
G
T
R
E
N
N
U
N
G
]
ꢀ
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
[FeI(CO)
3
A( SnI
C
H
T
U
N
G
T
R
E
N
N
U
N
G
3
)
2
]
[19]
Fe(CO)
5
To further verify our quantum chemical analysis, we con-
date. This includes the synthesis of a tetranuclear [{DPIM}-
119
+
ducted
preparation of 20 mg amounts result in phase-pure
BMIm] [{Fe(CO) } Sn I ] (1; FT-IR, TG, EDX), the much
Sn Mçssbauer spectroscopy studies. Although
A
H
U
G
R
N
U
G
2
6 2
1
ꢀ
ꢀ
ꢀ
1
[R ImC HCo (CO) ]X (X: PF , BF , B CAHTUNGTRNEGNU( C H ) ; R : vinyl,
6 5 4
allyl) or [EMIm]K[Co(CO) ] , in which [Co(CO) ] units
4 2 4
are interlinked by K . Interestingly, Co (CO) was used as
2 8
[
2
3
4
6
10
larger quantities (>500 mg) needed for Mçssbauer spectro-
scopy contain the adamantane-like cluster only as a minority
phase (Figure 5). This finding can be rationalized based on
a starting material for all the above compounds, whereas
Fe(CO) —to our surprise—has not yet been considered as a
5
starting material for the preparation of new carbonyl clus-
ters in ionic liquids. In contrast, its use for the synthesis of
0
[21]
Fe nanoparticles has been described recently. In addition
to the redox stability of ionic liquids, their weakly coordinat-
ing properties and inertness can also be considered as bene-
[20,22]
ficial aspects.
cally stabilized by alkyl or aryl groups, “purely inorganic”
clusters, such as [{Fe(CO) } ACHTGNUTRENN{NUG SnI} I ] , have so far been
3 4 6 4
reckoned as less stable.
In summary, the first bimetallic adamantane-like cluster
was obtained with [{Fe(CO)3} ACHTNUGTRNEUNG{ SnI} I ] in the compounds
4 6 4
Though, FeꢀSn carbonyl clusters are typi-
2
ꢀ
[2–7]
2ꢀ
[
BMIm] [{Fe(CO) } Sn I ] (1) and [BMIm] [S][{Fe ACHTUNGTRENNUNG( CO) } -
2 3 4 6 10 6 3 4
AHCTUNGTRNNEUGS n I ] (2). The valence states of iron and tin were verified
6 10 2
1
19
based on bond-length considerations, FT-IR and Sn Mçss-
bauer spectroscopy as well as with DFT calculations. As a
1
19
result, iron shows an electron density similar to Fe(CO)5
Figure 5. Sn Mçssbauer spectrum of the adamantane-like [{Fe-
2ꢀ
ꢀ
A( CO)
C
H
T
U
N
G
T
R
E
N
N
U
N
G
3
}
4
Sn
6
I
10
]
cluster and the barbell-shaped [FeI(CO)
3
A
H
U
G
R
N
U
G
3
)
2
]
cluster
and can be assumed to exhibit a formal oxidation state of
ꢁ
0
with transmission integral fit (red) at 78 K (see the Supporting Informa-
tion).
Fe . Tin is positively charged and can be estimated to be
+1.3
2ꢀ
Sn . The “purely inorganic” [{Fe(CO)3} ACHTUNGTRNEU{GN SnI} I ] cluster
4 6 4
in 1 and 2 is not stabilized by alkyl/aryl ligands. Its forma-
tion can be ascribed to the synthesis in ionic liquids with
their excellent redox stability and their weakly coordinating
properties.
the reactivity of the starting materials and the resulting car-
bonyl cluster compounds (see the Supporting Information).
The spectra could be well reproduced by a main signal at an
ꢀ
1
isomer shift of d=1.93(1) mms , exhibiting significant
ꢀ
1
quadrupole splitting of DE =1.32(1) mms . This signal
Experimental Section
Q
ꢁ
0
+
stems from the smaller, barbell-shaped [Fe I(CO)3ACHTUNGTRENNUNG
III
ꢀ
[16]
A
H
U
G
R
N
N
2
[{Fe(CO)
3
}
4
Sn
6
I
10
]
(1): SnI
4
(100 mg, 1 equiv) and Fe(CO)
[NTf ]. This
5
3
2
ꢀ
1
ACHTUNGTRENNUNG
2
tion, a weak, quadrupole-splitted signal at d=3.0(1) mms
solution was sealed in a glass ampoule and heated at 1308C for 4 days.
2
ꢀ
represents the adamantane-like [{Fe(CO)3}4
see the Supporting Information). Here, the isomer shift is
in good agreement with divalent tin compounds (e.g., SnI2:
A
H
U
G
R
N
N
{SnI} I ] cluster
ꢀ1
6
4
After being cooled to room temperature with a rate of 1 Kh , dark-red
(
crystals of 1 were obtained. The crystals were separated from the ionic
liquid by filtration through a glass filter and washed with the pure ionic
&
4
&
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!