Stabilization of aryl-calcium, -strontium, and -barium compounds by designed steric and π-bonding encapsulation

Article abstract of DOI:10.1002/anie.200501494

(Chemical Equation Presented) Pentafluorophenyl complexes of the heavier alkaline-earth metals (Ca, Sr, Ba) are stabilized by a novel sterically crowded triazenido ligand. The pendent arene rings of the aryl-substituted triazenide are π bonded to the alka

Full text of DOI:10.1002/anie.200501494

Angewandte
Chemie
Alkaline-Earth Metal Complexes
We recently succeeded in the preparation of aryl-substi-
tuted, sterically crowded triazenes. Ligands of this type may
be synthesized in excellent yields by the reaction of different
substituted 2-lithiobiphenyls with the m-terphenyl azide 1,
followed by hydrolysis (Scheme 2).
DOI: 10.1002/anie.200501494
Stabilization of Aryl–Calcium, –Strontium, and
–Barium Compounds by Designed Steric and
p-Bonding Encapsulation
Sven-Oliver Hauber, Falk Lissner, Glen B. Deacon, and
Mark Niemeyer*
In the last decade a major theme of organometallic chemistry
has been the design and development of alternative ligand
systems capable of stabilizing monomeric metal complexes
while provoking novel reactivity. Exploration of this field is
driven by the potential use of these complexes in catalysis and
organic synthesis. Examples of monoanionic chelating N-
donor ligands that have received much recent attention
(Scheme 1) include the b-diketiminate (I)[1] and the amidinate
Scheme 2. Synthesis of the triazene ligands 2a and 2b.
In a first attempt to test their properties, we have used the
obtained triazenes to stabilize pentafluorophenyl compounds
of the heavier alkaline-earth metals calcium, strontium, and
barium. The heteroleptic pentafluorophenyl triazenides are
accessible in tetrahydrofuran as solvent by a convenient one-
pot transmetalation/deprotonation[7] reaction from the tri-
azene 2a (HN3ArAr’), bis(pentafluorophenyl)mercury, and
the corresponding alkaline-earth metal (Scheme 3). After
crystallization from n-heptane, either the THF-free com-
Scheme 1. A comparison of some monoanionic chelating N ligands.
(II)[2] ligand systems. Much less attention has been given to
the closely related triazenides (III).[3,4] This may be attributed
to the lack of suitable ligands that are sterically crowded
enough to prevent undesirable ligand redistribution reactions
and allow better control of the electronic and steric properties
at the metal center.
Triazenides are weaker donors than the isoelectronic
amidinates and the related b-diketiminates, and should
induce greater electrophilicity at a bonded metal atom.[5]
This is reflected by the results of an NBO (natural bond
orbitals) analysis of the energy-minimized structures[6] of the
model anions 1,3-diphenyl-1,3-diketiminate (IM), 1,3-
diphenyl-1,3-diazaallyl (IIM), and 1,3-diphenyltriazenide
(IIIM) (see Supporting Information), which shows an NPA
(natural population analysis) charge for the chelating N atoms
of À0.54, À0.60, and À0.38, respectively.
Scheme 3. Synthesis of 3–5.
[*] Dipl.-Chem. S.-O. Hauber, Dr. F. Lissner, Priv.-Doz. Dr. M. Niemeyer
Institut für Anorganische Chemie
Universität Stuttgart
pounds [M(C6F5)(N3ArAr’)] (M = S r 4(), Ba (5)) or the
solvate [Ca(C6F5)(N3ArAr’)(thf)] (3) were isolated in good
yields. It is remarkable that attempts to replace the penta-
fluorophenyl substituents by a second triazenide ligand have
not been successful so far. Apparently, the steric bulk of the
latter prevents further substitution or ligand redistribution
and therefore formation of the homoleptic complexes.[8]
Solutions of 4 or 5 in aromatic or aliphatic solvents show
considerable thermal stability and can be stored at ambient
Pfaffenwaldring 55, 70569 Stuttgart (Germany)
Fax : (+49)711-685-4241
E-mail: mn@lanth.de
Prof. Dr. G. B. Deacon
School of Chemistry
Monash University
Clayton 3800 (Australia)
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2005, 44, 5871 –5875
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5871

Communications
Table 1: Important structural parameters [Š, 8] for the compounds
[(ArAr’N3)M(C6F5)] (3–5) and the DFT-calculated model complex [(2-
MesC6H4)(2-TripC6H4)N3SrC6F5] (4M).
temperature for months with only minor decomposition. This
may be contrasted with the behavior of some related
pentafluorophenyl compounds of the di- and trivalent rare
earth metals, which are thermally much more labile.[9]
Compared to 4 and 5, solutions of the calcium complex 3
are more sensitive and decompose within weeks at room
temperature or within two days at 708C. We have not been
able to isolate a well-defined decomposition product so far.
However, NMR spectroscopic evidence, in particular a
19F NMR signal at d = À72.3 ppm, points to the formation of
a molecular fluoride, presumably [CaF(N3ArAr’)(thf)], as
one of the main products.
M=Ca(thf) M=Sr
4M
M=Ba
À
N1 N2
1.338(7)
1.325(6)
4
1.314(7)
1.307(8)
3
1.305
1.307
3
1.309(5)
1.306(4)
3
À
N2 N3
Coord. no. (M)
À
M N1
2.371(5)
2.496(6)
53.3
0.850
2.499(11)
0.797
2.576(6)
2.548(5)
48.9
0.406
2.673(7)
0.712
2.591
2.596
48.7
0.038
2.638
0.246
2.740(3)
2.711(3)
46.3
0.706
2.808(5)
1.086
À
M N3
À
À
N1 M N3
[a]
M/N3
À
M C
[b]
M/C6F5
Compounds 3–5 were examined by X-ray crystallogra-
phy.[10] The molecular structures of the Ca and Ba derivatives
are shown in Figures 1 and 2, respectively, while important
structural parameters for all complexes are summarized in
Table 1. Despite the relatively large ionic radii of the M2+ ions
(Ca: 1.14 Š; Sr: 1.32 Š; Ba: 1.49 Š for coordination number
four[11]) the size of the h2-bonded triazenide ligands enforces
the formation of strictly monomeric complexes in which the
metal atoms possess apparent low coordination numbers of
three (4, 5) and four (3). It should be noted that the deviation
of the metal atoms from the N3 plane increases in the order Sr
(0.406 Š), Ba (0.706 Š), Ca (0.850 Š), while the bite angles
À
M O
2.308(5)
112.2
À
À
C M N2
114.8
127.0
113.1
hn/hn[c]
h5
h6/h5
h6/h5
h5/h5
M···C(Mes)[d]
M···C(Trip)[d]
M···F[e]
3.015–3.132 3.165–3.306 3.310–3.400 3.327–3.430
3.078–3.283 3.277–3.470 3.286–3.406
3.414
3.076
2.881
3.417
[a] Displacement of M from the N3 plane. [b] Displacement of M from the
C6F5 plane. [c] Hapticityof the M··· p-arene interaction to the Mes and
Trip rings. [d] M···C(arene) distances considered to be bonding.
[e] Closest M···F contact.
N1-M-N3 decrease in the order Ca (53.25(18)8), Sr
(48.89(17)8), Ba (46.33(10)8).
À
The average M N distances to the triazenide ligands
(Ca = 2.434, Sr= 2.562, Ba = 2.726 Š) are comparable to
those of the amidinates [{PhC(NSiMe3)2}2M(thf)2] (M = Ca
2.431 Š,[12a] Sr 2.584 Š[12b]), [{HC(NDip)2}2M(thf)n] (Dip =
2,6-diisopropylphenyl, M = Ca 2.383 Š (n = 1), Sr 2.585 Š,
Ba 2.728 Š (n = 2)),[12d] and [{PhC(NSiMe3)2}2Ba(dme)(thf)]
(2.779 Š),[12c] which contain five- to seven-coordinate metal
atoms, and are longer than those of the four-coordinate
homoleptic b-diketiminates [(Dip-nacnac)2M] (Dip-nacnac =
(Dip)NC(Me)C(H)C(Me)N(Dip)) (M = Ca 2.379 Š, Sr
2.510 Š, Ba 2.712 Š).[13] Taking into account the apparent
lower coordination numbers in the triazenides 3–5 the longer
À
M N bond lengths are most likely a result of the decreased
donor ability of the triazenide relative to the amidinate and b-
diketiminate ligands and/or a consequence of the additional
metal–p-arene interactions (see below). Furthermore, the
Figure 1. Molecular structure of 3 with thermal ellipsoids set to 30%
probability. Hydrogen atoms are omitted for clarity.
À
average Ca N distance of 2.434 Š in the THF solvate 3 is
considerably shorter than the corresponding values found in
the eight-coordinate triazenide [Ca{N3(tol)2}2(dme)2] (av
2.524 Š) (tol = 4-Me-phenyl),[8] which is the only derivative
of the heavier alkaline-earth metals currently available for
comparison.
Well-defined s-bonded organometallic compounds of the
heavier alkaline-earth metals calcium, strontium, and barium
are scarce and are currently restricted to some structurally
characterized trimethylsilylmethanides or -benzyls, di- and
triphenylmethanides, and acetylides.[14] The heteroleptic pen-
tafluorophenyls 3–5 therefore are the first structurally
authenticated Ar–Ca, Ar–Sr, and Ar–Ba compounds. The
À
observed M C distances are toward the short end of the range
À
À
of M C bonds reported for other organyls. The Ca C bond
(2.499 Š) is comparable to the value in the four-coordinate
alkyl [Ca{CH(SiMe3)2}2(dioxane)2] (2.48 Š)[15] but slightly
longer than the distance in the two-coordinate compound
Figure 2. Molecular structure of 5 with thermal ellipsoids set to 50%
probability. Hydrogen atoms are omitted for clarity.
5872
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5871 –5875

Angewandte
Chemie
[Ca{C(SiMe3)3}2] (2.46 Š).[16] The Sr C and Ba C distances of
2.673(7) and 2.808(5) Š, respectively, seem to be the shortest
such bonds reported so far. The deviations of the metal atoms
from the least-squares plane defined by the carbon atoms of
the bonded C6F5 ligands also merit comment. They increase in
the order Sr (0.712 Š), Ca (0.797 Š), Ba (1.086 Š). These
unusually large displacements are explained by the highly
ionic, nondirectional character of the metal–carbon bonds,[6]
which is further increased by the electronic influence of the
fluoro substituents. As a result, relatively weak intra- or
intermolecular forces such as steric repulsion or p-stacking
may lead to severe distortions (see Figure S5 in the Support-
ing Information). It is notable that similar, although even
larger, displacements in the range 0.75 to 1.50 Š have been
for the unsolvated ytterbium thiophenolate [Yb(S-2,6-
À
À
Trip2C6H3)2].[19] In addition, low-temperature 19F NMR
experiments for 4 and 5 show coalescence of the signals of
the ortho-C6F5 fluorine atoms at 218 K and 203 K, with
corresponding DG values of 39.1 and 34.5 kJmolÀ1, respec-
tively. Unfortunately, we have not yet been able to determine
which intra- or intermolecular processes are responsible for
this fluxional behavior. According to preliminary quantum-
chemical calculations (see Supporting Information), the
binding energy for the Sr···F interaction in the model complex
[H2N3SrC6F5] is estimated to be only 22.1 kJmolÀ1 at the MP2
level of theory. DFT calculations for the more sophisticated
model compound [{(1-MesC6H4)(1-TripC6H4)N3}Sr(C6F5)]
(4M), in which the noncoordinating Mes group in 4 has been
replaced by a hydrogen atom, reveal a shorter Sr···F contact
(2.881 Š for 4M versus 3.076 Š for 4) but weaker p-coordi-
nation (av M···C 3.363 Š for 4M versus 3.210 Š for 4). Since it
is known that DFT methods tend to underestimate weak
interactions such as p-arene bonding to electropositive
metals,[19] it is plausible that in the real systems 3–5 these
interactions are actually stronger and dominate over any
M···F coordination. As a result, decomposition pathways
involving ortho-fluoride elimination are effectively blocked.
In summary, we have designed and prepared very bulky
aryl-substituted triazenide ligands that have permitted the
synthesis of the first structurally characterized aryl com-
pounds of the heavier alkaline-earth metals Ca, Sr, and Ba.
Kinetic stabilization in the solid-state and in solution through
steric and electronic saturation of the metal centers is
achieved by additional p-arene interactions to the pendent
aryl substituents. These novel, sterically crowded triazenides
may find use as ancillary ligands in catalysis and we are
currently exploring this field.
observed for the aryl lanthanides [Ln(Dpp)2(thf)2] (Dpp =
2,6-Ph2C6H3), which also contain highly polar M C bonds.
[17]
À
Perhaps the most striking features in the solid-state
structures of 3–5 are the additional metal–p-arene interac-
tions[18] with the pendent arms of the biphenyl and terphenyl
groups in the triazenide ligands. These arene rings apparently
compete with THF to bind the MII cations, thus allowing the
isolation of solvent-free derivatives.[19] The calcium ion in the
THF solvate 3 interacts with only one mesityl (Mes) ring in an
h5 fashion,[20] with Ca···C distances in the relatively narrow
range 3.015(7)–3.132(7) Š. In the solvent-free compounds 4
and 5 the metal ions show h6- or h5-p-arene interactions to one
Mes group and also to the triisopropylphenyl (Trip) ring of
the triazenide ligand. The resulting, rather distorted pseudo-
tetrahedral coordination of the metal cations is reflected in
part by the angles N2-M-C51 (Ca: 112.28; Sr: 114.88; Ba:
113.18) and X1-M-X2 (Sr: 129.58; Ba: 145.08), where X1 and
X2 define the centroids of the coordinated arene rings and the
bidentate triazenido ligand occupies only one coordination
site. The importance and specificity of the (N3ArAr’) ligand,
and in particular its p-bonding capability, is shown by the
failure of a reaction analogous to Scheme 3 between Ca, Sr, or
Ba, Hg(C6F5)2, and the bulky N,N’-bis(2,6-diisopropylphen-
yl)formamidine.[12d] Only decomposition products were
obtained, plausibly derived from an unstable [ML(C6F5)]
species.
Experimental Section
All manipulations were carried out under strictly anaerobic and
anhydrous conditions under argon. The starting materials 2-iodo-
2’,4’,6’-triisopropylbiphenyl,[21]
2,4,6,2’’,4’’,6’’-hexamethyl-1,1’:3’,1’’-
terphenyl-2’-azide,[22] and bis(pentafluorophenyl)mercury[23] were
prepared by known procedures.
À
A relatively short Sr···F52 C52(ortho) contact of 3.076 Š
is observed in the solid-state crystal structure of the Sr
HN3ArAr’ (2a): n-Butyllithium (44.4 mmol, 2.5m hexane solu-
tion) was added at 08C to a solution of 2-iodo-2’,4’,6’-triisopropylbi-
phenyl (18.1 g, 44.4 mmol) in diethyl ether (250 mL) and stirring was
continued for 2 h. The clear solution of the aryl lithium compound
thus obtained was then treated with small portions of 2,4,6,2’’,4’’,6’’-
hexamethyl-1,1’:3’,1’’-terphenyl-2’-azide (15.8 g, 44 mmol). After
warming to ambient temperature and stirring for an additional 12 h,
the orange solution was quenched with water (400 mL). The aqueous
phase was separated and extracted with diethyl ether (3 ” 100 mL).
The organic phases were combined, repeatedly washed with water,
and dried with Na2SO4. Filtration followed by solvent removal in
vacuo afforded 2a as a yellow solid. Analytically pure, pale yellow
crystals were obtained by recrystallization from hot acetone. Yield:
25.2 g (39.6 mmol, 89%); m.p. 174–1768C; 1H NMR (400.1 MHz,
[D6]benzene): d = 0.87, 1.00, 1.13 (3 ” d, 3JH,H = 6.9 Hz, 3 ” 6H; o + p-
CH(CH3)2), 2.04 (s, 12H; o-CH3), 2.18 (s, 6H; p-CH3), 2.54 (sept,
3JH,H = 6.9 Hz, 2H; o-CH(CH3)2), 2.69 (sept, 3JH,H = 6.9 Hz, 1H; p-
CH(CH3)2), 6.79 (s, 4H; m-Mes), 6.85–7.14 (m, 7H; various aryl-H),
7.00 (s, 2H; m-Trip), 8.70 ppm (s, 1H; NH); 13C NMR (62.9 MHz,
[D6]benzene): d = 20.9 (o-CH3), 21.1 (p-CH3), 23.8, 24.0, 24.5 (o + p-
CH(CH3)2), 30.7 (o-CH(CH3)2), 34.5 (p-CH(CH3)2), 121.4, 122.2,
À
À
derivative 4, which is accompanied by different Sr C51 C52/
C56 angles (105.7(5)8 and 139.6(6)8). In contrast, the other
pentafluoroaryl derivatives show considerably longer M···F
contacts of more than 3.41 Š. To clarify if the metal···arene
and metal···fluorine interactions also persist in solution,
variable-temperature NMR experiments were performed
for the most soluble compounds 4 and 5. At 298 K, the
1H NMR spectra in [D14]methylcyclohexane show only one
set of signals for the Mes substituents, which, however, splits
at lower temperatures. Upon warming, coalescence is
detected at 238 K (4) and 258 K (5), with corresponding
energy barriers of 49.0 and 53.1 kJmolÀ1, respectively. This
observation can be interpreted in terms of a hindered rotation
À
around the N1 C1 bond and therefore the persistence of the
metal–p-arene interactions in solution. A similar behavior,
with a rotational barrier of 54.3 kJmolÀ1, was recently
observed and confirmed by quantum-chemical calculations
Angew. Chem. Int. Ed. 2005, 44, 5871 –5875
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5873

Communications
126.8, 128.3, 129.8, 130.8 (aromatic CH), 113.7, 125.6, 134.8, 135.8,
the Gaussian 03 adaptation of the NBO program (E. D.
Glendening, A. E. Reed, J. E. Carpenter, F. Weinhold, NBO
Version 3.1). For the model complex 4M the following NPA
charges were calculated: + 1.728 (Sr), À0.505 (N1), À0.509
(N3), À0.637 (C51).
136.1, 137.3, 139.3, 146.2, 147.5, 149.3 ppm (aromatic C). elemental
analysis calcd (%) for C45H53N3: C 84.99, H 8.40, N 6.61; found: C
84.88, H 8.41, N 6.64.
[Ba(C6F5)(N3ArAr’)]
(5):
Bis(pentafluorophenyl)mercury
(1.07 g, 2.00 mmol) was added at ambient temperature to a stirred
mixture of barium ingots (2.5 g, 18.2 mmol) and 2a (1.27 g,
2.00 mmol) in THF (60 mL). Stirring was continued for 14 h, where-
upon the solvent was removed under reduced pressure. The green
residue was treated with n-heptane, and solid materials were
separated by centrifugation. The volume of the resulting deep
yellow solution was reduced to incipient crystallization under reduced
pressure. Storage in a freezer at À208C for several days afforded 5 as
a bright yellow, crystalline material. Yield: 1.48 g (1.58 mmol, 79%);
[7] G. B. Deacon, C. M. Forsyth, S. Nickel, J. Organomet. Chem.
2002, 647, 50.
[8] Solvated bis- and tris(triazenide) complexes with much less
bulky ligands are known for Mg and Ca: S. Westhusin, P.
Gantzel, P. J. Walsh, Inorg. Chem. 1998, 37, 5956.
[9] a) G. B. Deacon, C. M. Forsyth, Organometallics 2003, 22, 1349;
b) L. Maron, E. L. Werkema, L. Perrin, O. Eisenstein, R. A.
Andersen, J. Am. Chem. Soc. 2005, 127, 279; c) M. L. Cole, G. B.
Deacon, P. C. Junk, K. Konstas, Chem. Commun. 2005, 1581.
[10] Shock-frozen crystals in Paratone N, programs SHELXTL 5.03
and SHELXL-97, refinement with all data on F2, non-hydrogen
atoms anisotropic, hydrogen atoms at calculated positions.
Crystal data for 3: pale yellow prism (0.20 ” 0.20 ” 0.10 mm3)
from n-heptane at 208C, C55H60CaF5N3O, Mr = 953.92, mono-
clinic, space group P21/n, a = 9.380(3), b = 23.159(5), c =
1
m.p. 144–1708C (decomp); H NMR (250.1 MHz, [D6]benzene): d =
0.63, 0.78, 1.18 (3 ” d, 3JH,H = 6.8 Hz, 3 ” 6H; o + p-CH(CH3)2), 2.04 (s,
3
12H; o-CH3), 2.19 (s, 6H; p-CH3), 2.57 (sept, JH,H = 6.8 Hz, 2H; o-
CH(CH3)2), 2.90 (sept, 3JH,H = 6.8 Hz, 1H; p-CH(CH3)2), 6.46–7.27
(m, 7H; various aryl-H), 6.78 (s, 4H; m-Mes), 7.04 ppm (s, 2H; m-
Trip); 13C NMR (62.9 MHz, [D6]benzene): d = 20.8 (o-CH3), 21.0 0(p-
CH3), 23.1, 24.2, 24.2 (o + p-CH(CH3)2), 30.5 (o-CH(CH3)2), 33.8 (p-
CH(CH3)2), 116.9, 122.0, 122.0, 123.1, 128.6, 129.0, 129.1, 129.4
(aromatic CH), 128.8, 131.5, 136.6, 137.4, 142.0, 142.4, 148.0, 150.3,
150.6, 152.1 ppm (aromatic C); signals for the C6F5 group could not be
detected or assigned completely due to signal overlap and their weak
appearance; 19F NMR (235.4 MHz, [D6]benzene): d = À159.0 (m, 2F;
m-C6F5), À154.2 (tt, 3JF,F = 20.3, 4JF,F = 3.3 Hz, 1F; p-C6F5),
À113.2 ppm (m, 2F; o-C6F5); 15N NMR (40.6 MHz, [D6]benzene):
d = 109.5 (s; -NNN-), À69.3, À70.4 ppm (s; -NNN-); elemental
analysis calcd (%) for C51H52BaF5N3: C 65.21, H 5.58, N 4.47;
found: C 64.76, H 5.76, N 4.52.
23.142(6) Š, b = 93.93(3)8, V= 5015(3) Š3, Z = 4, 1calcd
=
1.211 gcmÀ1, m(MoKa) = 0.184 mmÀ1, Siemens P3 diffractometer,
T= 173 K, 2Vmax = 488, 8420 (Rint = 0.106) collected and 7881
unique reflections, 612 parameters, 7 restraints, absorption
correction by Y-scans, R1 = 0.117 for 2929 reflections with I >
2s(I), wR2 = 0.163 (all data), GOF = 1.071. Crystal data for 4:
yellow block (0.15 ” 0.08 ” 0.06 mm3) from n-heptane at À158C,
C51H52F5N3Sr, Mr = 889.58, monoclinic, space group P21/n, a =
16.051(3), b = 14.970(3), c = 19.128(4) Š, b = 106.79(3)8, V=
4399.9(15) Š3,
Z = 4,
1calcd = 1.343 gcmÀ1
,
m(MoKa) =
1.285 mmÀ1, Nonius Kappa CCD diffractometer, T= 100 K,
2Vmax = 56.68, 50525 (Rint = 0.149) collected and 10346 unique
reflections, 571 parameters, 0 restraints, absorption correction
from symmetry-related measurements, R1 = 0.124 for 8127
reflections with I > 2s(I), wR2 = 0.298 (all data), GOF = 2.33.
Crystal data for 5: yellow block (0.15 ” 0.06 ” 0.04 mm3) from n-
heptane at À208C, C51H52BaF5N3, Mr = 939.30, monoclinic,
space group P21/n, a = 16.4690(4), b = 15.5775(3), c =
18.3051(3) Š, b = 108.2063(11)8, V= 4461.00(16) Š3, Z = 4,
Details of the preparation and characterization of compounds 2b,
3, and 4, and additional spectroscopic data for 2a and 5 are given in
the Supporting Information.
Received: April 30, 2005
Keywords: barium · calcium · fluorinated ligands · N ligands ·
.
strontium
1calcd = 1.399 gcmÀ1
,
m(MoKa) = 0.946 mmÀ1
,
Nonius Kappa
CCD diffractometer, T= 100 K, 2Vmax = 56.68, 74322 (Rint
=
[1] J. Barker, M. Kilner, Coord. Chem. Rev. 1994, 133, 219.
[2] L. Bourget-Merle, M. F. Lappert, J. R. Severn, Chem. Rev. 2002,
102, 3031.
0.175) collected and 10877 unique reflections, 599 parameter, 1
restraint, absorption correction from symmetry-related mea-
surements, R1 = 0.060 for 7895 reflections with I > 2s(I), wR2 =
0.114 (all data), GOF = 1.536. CCDC-268134 (3),CCDC-268135
(4), and CCDC-268136 (5) contain the supplementary crystallo-
graphic 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.
[3] D. S. Moore, S. D. Robinson, Adv. Inorg. Chem. Radiochem.
1986, 30, 1.
[4] K. Vrieze, G. van Koten, in Comprehensive Coordination
Chemistry, Vol. 2 (Ed.: G. Wilkinson), Pergamon, Oxford,
1987, pp. 189 – 244.
[5] P. Gantzel, P. J. Walsh, Inorg. Chem. 1998, 37, 3450.
[6] The Gaussian 03 package (Revision B.01, Gaussian Inc.,
Pittsburgh PA, 2003) was used for all energy and frequency
calculations. The geometries of the molecules were optimized
using either density functional theory (DFT) with the functional
B3LYP for the model ligands IM–IIIM (all with implied C2
symmetry) and the model complex 4M or second-order Møller–
Plesset perturbation theory (MP2) for two conformers of the
model compound [H2N3SrC6F5] (Cs symmetry). In each case the
GDIISalgorithm with the TIGHT convergence criterion was
used. A quasi-relativistic 10-valence-electron pseudopotential
was employed for the heavy atom Sr (M. Kaupp, P. von R.
Schleyer, H. Stoll, H. Preuss, J. Chem. Phys. 1991, 94, 1360). The
corresponding 6s6p5d valence basis was augmented by one set
of f-functions. The basis sets for C, H, N, and F were either 6-
311 + G** (IM–IIIM) or 6-31G* (4M, [H2N3SrC6F5]). All sta-
tionary points were characterized by analytical or numerical
frequency analyses. The natural bond orbital analysis employed
[11] R. D. Shannon, Acta Crystallogr. Sect. A 1976, 32, 751.
[12] a) M. Westerhausen, W. Schwarz, Z. Naturforsch. B 1992, 47,
453; b) M. Westerhausen, H. D. Hausen, W. Schwarz, Z. Anorg.
Allg. Chem. 1992, 618, 121; c) M. Westerhausen, W. Schwarz, Z.
Anorg. Allg. Chem. 1993, 619, 1455; d) M. L. Cole, P. C. Junk,
New J. Chem. 2005, 29, 135.
[13] S. Harder, Organometallics 2002, 21, 3782.
[14] a) T. P. Hanusa, Coord. Chem. Rev. 2000, 210, 329; b) M.
Westerhausen, Angew. Chem. 2001, 113, 3063; Angew. Chem.
Int. Ed. 2001, 40, 2975; c) J. S. Alexander, K. Ruhlandt-Senge,
Eur. J. Inorg. Chem. 2002, 2761.
[15] F. G. N. Cloke, P. B. Hitchcock, M. F. Lappert, G. A. Lawless, B.
Royo, J. Chem. Soc. Chem. Commun. 1991, 724.
[16] C. Eaborn, S. A. Hawkes, P. B. Hitchcock, J. D. Smith, Chem.
Commun. 1997, 1961.
[17] a) G. Heckmann, M. Niemeyer, J. Am. Chem. Soc. 2000, 122,
4227; b) M. Niemeyer, Acta Crystallogr. Sect. E 2001, 57, m578.
5874
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5871 –5875

Angewandte
Chemie
2+
[18] For examples of p-arene interactions to Ca2+, S r , and Ba2+, see:
a) M. Westerhausen, M. Krofta, N. Wiberg, J. Knizeh, H. Nꢀth,
A. Pfitzner, Z. Naturforsch. B 1998, 53, 1489; b) L. Bonomo, E.
Solari, R. Scopelliti, C. Floriani, Chem. Eur. J. 2001, 7, 1322;
c) G. Guillemot, E. Solari, C. Rizzoli, C. Floriani, Chem. Eur. J.
2002, 8, 2072; d) F. Feil, S. Harder, Eur. J. Inorg. Chem. 2003,
3401.
[19] M. Niemeyer, Eur. J. Inorg. Chem. 2001, 1969.
[20] The assignment of the hapticity of the metal–p-arene interac-
À
tions in compounds 3–5 is based on the evaluation of the M C
distances and the angle between the M–centroid vector and the
normal of the arene plane (see Supporting Information).
[21] N. Hartmann, M. Niemeyer, Synth. Commun. 2001, 31, 3839.
[22] J. Gavenonis, T. D. Tilley, Organometallics 2002, 21, 5549.
[23] G. B. Deacon, J. E. Cosgriff, E. T. Lawrenz, C. M. Forsyth, D. L.
Wilkinson, in Synthetic Methods of Organometallic and Inor-
ganic Chemistry, Vol. 6 (Eds.: W. A. Herrmann, F. T. Edelmann),
Thieme, Stuttgart, 1997, pp. 48 – 49.
Angew. Chem. Int. Ed. 2005, 44, 5871 –5875
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5875