Arylcalcium hydrides as precursors to alkoxides and aryloxides of calcium

Article abstract of DOI:10.1002/ejic.200390184

Calcium atoms react with +I-substituted benzene derivatives under cocondensation conditions to yield arylcalcium hydrides. With toluene, tert-butylbenzene, and trimethyl(phenyl)silane the reaction showed no selectivity for C-H activation, resulting in the formation of each of the three possible isomers, while with m-xylene the reaction resulted in selective activation of the bond meta to the CH3 groups. Treatment of (tert-butylphenyl)calcium hydride with di- and trisubstituted phenols such as 2,6-di-tert-butylphenol, 2,6-di-tert-butyl-4-methylphenol, and 2,4,6-tri-tert-butylphenol resulted in the formation of calcium aryloxides in yields > 95%. [Ca(2,4,6-tBu₃C₆H₂O)₂ (THF)₃] crystallises from a THF solution as a distorted trigonal bipyramid, with two THF ligands in the trans-axial positions and the third THF ligand and both aryloxide groups in equatorial positions. The Ca-O_Aryl bond length was found to average at 2.181(3). A. The Ca-O-C_Aryl angles are almost linear, with Cal-O1-C1 and Ca1-O2-C19 being 173.9(3)° and 178.8(3)°, respectively, while the O1-Cal-O2 angle was determined as 157.04(12)°. A reaction between (tert-butylphenyl)calcium hydride and triphenylmethanol resulted in the formation of the corresponding monomeric calcium bis(alkoxide) Ca(OCPh₃)₂(THF)₄ in 95% yield. The structure was determined as a distorted octahedron with the alkoxide ligands in a cis-equatorial arrangement. The Ca-O bond length was determined at 2.1609(17) A. The Ca-O-C_Aryl angle of 177.46(16)° is almost linear, while the O1-Cal-O1A angle was determined at 110.12(19)°. ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003).

Full text of DOI:10.1002/ejic.200390184

FULL PAPER
Arylcalcium Hydrides as Precursors to Alkoxides and Aryloxides of Calcium
John P. Dunne,[a] Matthias Tacke,*[a] Carola Selinka,[b] and Dietmar Stalke[b]
Keywords: Calcium / CϪH activation / Cocondensation / Hydrides
˚
Calcium atoms react with +I-substituted benzene derivatives
under cocondensation conditions to yield arylcalcium hy-
drides. With toluene, tert-butylbenzene, and trimethyl(phen-
yl)silane the reaction showed no selectivity for C−H activa-
tion, resulting in the formation of each of the three possible
length was found to average at 2.181(3) A. The Ca−O−CAryl
angles are almost linear, with Ca1−O1−C1 and Ca1−O2−C19
being 173.9(3)° and 178.8(3)°, respectively, while the
O1−Ca1−O2 angle was determined as 152.04(12)°. A reaction
between (tert-butylphenyl)calcium hydride and triphenylme-
isomers, while with m-xylene the reaction resulted in select- thanol resulted in the formation of the corresponding mono-
ive activation of the bond meta to the CH3 groups. Treatment
of (tert-butylphenyl)calcium hydride with di- and trisubsti-
meric calcium bis(alkoxide) Ca(OCPh3)2(THF)4 in 95% yield.
The structure was determined as a distorted octahedron with
tuted phenols such as 2,6-di-tert-butylphenol, 2,6-di-tert-bu- the alkoxide ligands in a cis-equatorial arrangement. The
˚
tyl-4-methylphenol, and 2,4,6-tri-tert-butylphenol resulted in Ca−O bond length was determined at 2.1609(17) A. The
the formation of calcium aryloxides in yields
Ͼ
95%.
Ca−O−CAryl angle of 177.46(16)° is almost linear, while the
O1−Ca1−O1A angle was determined at 110.12(19)°.
[Ca(2,4,6-tBu3C6H2O)2(THF)3] crystallises from a THF solu-
tion as a distorted trigonal bipyramid, with two THF ligands
in the trans-axial positions and the third THF ligand and both
aryloxide groups in equatorial positions. The Ca−OAryl bond
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
Another synthetic method is to use bulk calcium metal
for reaction, but a major difficulty encountered with this
technique is its relative inertness to reaction. There are a
number of different methods to activate the bulk metal, in-
cluding the formation of an activated powder by displace-
ment of calcium by potassium from a calcium dihalide
(Rieke calcium),[2] or by decomposition of alkaline earth
metal/anthracene complexes (Bogdanovic method).[3] Al-
ternatively, the metal can be dissolved in liquid ammonia[4]
or dispersed in THF.[5]
Metal vapour synthesis (MVS) takes a different approach
to the above methods to overcome the inertness of bulk
calcium metal. The above procedures are in reality a clean-
ing of the metal surface. Metal atom techniques are differ-
ent, in that they utilise the vaporisation enthalpy of calcium
as the reaction enthalpy. This technique has received only
sporadic attention to date. In the late 1980s, MVS was used
to prepare [Ca(cot)(THF)n][6] and the bis(cyclopentadienyl)-
calcium complex [Ca{η-C5H3-1,3-(SiMe3)2}2(THF)],[7] and
more recently [Ca{CH(SiMe3)2}2(THF)3],[8] while in the
mid 1980s Mochida et al.[9] reported the CϪH bond activa-
tion of benzene by calcium atoms to form PhCaH, which
was not isolated but characterised by its organic products
as outlined in Figure 2. In these systems the high reactivity
of calcium atoms is stored in the products’ highly reactive
bonds.
Until the mid 1980s, complexes of calcium were among
the longest known, yet least understood class of com-
pounds. Since the early 1990s, however, there has been re-
newed interest in this field.[1] The reasons for this slow de-
velopment have been due to numerous factors. Firstly, cal-
cium complexes have rather poor solubility in hydrocarbon
solvents and many are thermally unstable.[1] Secondly, there
are only a limited number of starting reagents and preparat-
ive techniques for the formation of calcium compounds.[1]
The most popular preparative technique is to use non-
elemental forms of calcium (as shown in Figure 1), utilising
a salt such as CaI2 or calcium bis(trimethylsilyl)amide as
the source of calcium. Another method is by deprotonation
with an activated hydrocarbon, as shown in Equation (C)
in Figure 1.
Figure 1. Reactions using non-elemental forms of calcium
[a]
Department of Chemistry, University College Dublin,
Belfield, Dublin 4, Ireland
Recently performed theoretical work showed that the
molecule ‘‘PhCaH’’ should be able to exist in donor solv-
[b]
Institut für Anorganische Chemie, Universität Würzburg,
Am Hubland, 97074 Würzburg, Germany
1416  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ1948/03/0407Ϫ1416 $ 20.00ϩ.50/0 Eur. J. Inorg. Chem. 2003, 1416Ϫ1425

Arylcalcium Hydrides as Precursors to Alkoxides and Aryloxides of Calcium
FULL PAPER
Figure 2. Cocondensation reactions with calcium atoms
Figure 3. Calculated structures for PhCaH in donor solvents
ents such as THF in different aggregation states.[10] As a Results
means of developing a convenient theoretical model for ar-
ylcalcium hydride in donor solvents, ammonia was used to
simulate an amine and the elimination of H2 was sup-
Experimental Study
Under cocondensation conditions, calcium atoms react
pressed. These calculations show that a concentration-de- with ϩI-substituted benzene derivatives to form extremely
pendent equilibrium exists in donor solvents (as outlined in air- and moisture-sensitive, dark red solids in yields ranging
Figure 3). At low concentrations of ‘‘PhCaH’’, a mono- from 78 to 97%. When calcium atoms were cocondensed
meric species is energetically favoured, while at higher con- with benzene in the presence of THF (as outlined in
centrations a tetrameric species with triply bridging hydride Scheme 1), a dark red, pyrophoric solid of composition
ligands exists. However, at average concentrations of PhCaH·2THF was obtained in a yield of 78%. An IR spec-
‘‘PhCaH’’ the energetically favoured structure is a dimer.
trum in Nujol mull showed a strong vibration at 1094 cmϪ1
Calcium alkoxides are used as CVD precursors and for and a weaker signal at 1031 cmϪ1. We assign these as the
sol-gel synthesis.[1,11] As well as this, alkaline earth metal CaϪH vibration at 1094 cmϪ1 and CϪH bond deforma-
alkoxides are noted for their rich variety of structures and tions of the phenyl ring at 1031cmϪ1, based on similar
reactions.[1] This is due to the conformational and bonding trends seen in our DFT-modelled vibrations. Because of the
flexibility of the alkoxide ligand (e.g., µ1-, µ2-, and µ3-bond- low solubility of the product in organic solvents and its ex-
ing modes are known). Crystallographically identified tremely high sensitivity to air and moisture, however, NMR
structural types range from discrete monomeric species spectroscopic data were not obtained for this or any other
such as [Ca(BHT)2(THF)3] (BHT
methylphenolate)[12] to large polynuclear aggregates such as °C with SiMe3Cl (as shown in Scheme 1) and the organic
[Ca9(OCH2CH2OMe)18(HOCH2CH2OMe)2].[13]
derivative trimethyl(phenyl)silane was identified by GC MS
ϭ 2,6-tert-butyl-4- arylcalcium hydride species. Instead, 1a was treated at Ϫ30
The aim of this work was firstly to use calcium atoms to and NMR techniques. Further characterisation to show the
activate the aromatic CϪH bonds of ϩI-substituted ben- presence of the hydride moiety was performed by treatment
zene derivatives, thus forming arylcalcium hydrides, and of 1a with the hydride-abstracting agent Ph3CϩBF4Ϫ. The
secondly to use these arylcalcium hydrides as starting re- product isolated from this reaction was triphenylmethane,
agents for the formation of calcium alkoxides and aryl- as shown in Scheme 2.
oxides. To date, relatively few examples of monomeric alk-
A change of the aromatic compound from benzene to
oxides and aryloxides have been structurally characterised toluene resulted in a higher yield of 97%. The elemental
[examples include [Ca(BHT)2(THF)3] and [Ca(cox)2- analysis of the solid compound showed that it was com-
(THF)3]·THF {cox
ϭ
C(Ph)2(CH2C6H4Cl-4)}.[12] The posed of a C7H7CaH·2THF moiety. As with the benzene
reasons for this arise from difficulties in preventing oligo- cocondensation product, a CaϪH vibration was found at
merisation or polymerisation of the calcium alkoxides and 1092 cmϪ1. In order to ascertain which CϪH bond was
their low solubilities in most solvents. For this purpose a activated, 1b was converted into an organic compound by
series of sterically bulky alcohols and phenols were chosen, treating it with chlorotrimethylsilane. In this case, GC MS
in order to prevent oligomerisation and increase solubility detected the presence of three different isomers in similar
in hydrocarbons.
abundances. In conjunction with NMR spectroscopic data,
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425
1417

J. P. Dunne, M. Tacke, C. Selinka, D. Stalke
FULL PAPER
Scheme 1. Cocondensation of calcium atoms with ϩI-substituted benzene derivatives in the presence of THF
Ϫ
Scheme 2. Reaction of arylcalcium hydrides with Ph3CϩBF4
these were identified as the ortho, meta, and para isomers dicating the presence of a hydride ligand, as outlined in
of
trimethyl(methylphenyl)silane,
as
outlined
in Scheme 2.
Scheme 1.[14Ϫ16] As before, 1b was treated with
A change of the ϩI benzene derivative to m-xylene re-
Ph3CϩBF4Ϫ, resulting in the isolation of triphenylmethane sulted in a lower yield of 85%. As in the previous cocon-
as outlined in Scheme 2. No attack at the thermodyn- densation products, elemental analysis shows the product
amically favoured benzyl position was detected.
to be of the composition C8H9CaH·2THF, and the CaϪH
When tert-butylbenzene was used instead of toluene (as vibration was detected at 1095 cmϪ1. To confirm the regio-
outlined in Scheme 1), a slightly lower yield of 93% was chemistry of this reaction, 1d was again treated with chloro-
obtained. Like the previous cocondensation products, ele- trimethylsilane. Unlike in the toluene and tert-butylben-
mental analysis showed it to be composed of
a
zene cocondensations, though, GC MS and NMR showed
C10H13CaH·2THF moiety, and a CaϪH vibration was that a single species had been formed, and this was identi-
found at 1094 cmϪ1. As in the toluene cocondensation, 1c fied as (3,5-dimethylphenyl)trimethylsilane, as outlined in
Ϫ
was treated with chlorotrimethylsilane to identify which Scheme 1.[20] Treatment of 1d with Ph3CϩBF4 as outlined
CϪH bonds were activated. As in the case of the toluene in Scheme 2 confirmed the presence of the hydride ligand.
reaction product, GC MS and NMR showed the presence
When trimethyl(phenyl)silane was cocondensed with cal-
of ortho, meta, and para isomers of (tert-butylphenyl)trime- cium atoms in the presence of THF, a similar lack of select-
thylsilane in similar abundances, as outlined in ivity was observed. The isolated dark red solid (86% yield)
Scheme 1.[17Ϫ19] As in the previous cocondensations, 1c was had the composition C9H13SiCaH·2THF. As in the previ-
treated with Ph3CϩBF4Ϫ to form triphenylmethane, thus in- ous cocondensation products, a CaϪH vibration was found
1418
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425

Arylcalcium Hydrides as Precursors to Alkoxides and Aryloxides of Calcium
FULL PAPER
at 1094 cmϪ1. For ease of characterisation, 1c was treated
with dimethyl disulfide rather than chlorotrimethylsilane in
order to elucidate the selectivity of this reaction. GC MS
measurements showed the presence of three different iso-
mers in similar abundances. In conjunction with NMR
spectroscopic data, these were identified as the ortho, meta,
and para isomers of trimethyl[4-(methylthio)phenyl]silane,
as outlined in Scheme 1.[17,19] As in all the previous cocon-
Ϫ
densations, 1c was treated with Ph3CϩBF4 (as outlined in
Scheme 2) to confirm the presence of a hydride ligand.
These arylcalcium hydrides can be redissolved in THF
and utilised as reagents for the formation of new calcium
compounds. When arylcalcium hydrides were treated at
room temperature with bulky phenols (3aϪ3c), the corres-
ponding bis(aryloxides) were formed in excellent yields (all
Ͼ 95%), as outlined in Scheme 3. IR and NMR techniques
identified all species as bis(aryloxides) of calcium. A char-
acteristic 13C NMR signal for the ipso-carbon atom
(CϪOϪ) can be seen at δ ഠ 164.4 (for 3a), 163.9 (for 3b)
and 163.4 ppm (for 3c). The calcium aryloxides 3b and 3c
have been reported previously and the NMR and IR data
agree with the literature.[12,21] However, no crystal structure
for 3c had been reported in the literature. Integration of the
THF signals in the 1H NMR shows that 3a has a coordina-
tion number of six, while 3b has the lower coordination
number of five.
Figure 4. Crystal structure of Ca[2,4,6-tBu3C6H2O]2(THF)3 (3c)
˚
Table 1. Selected bond lengths [A] and angles [°] for 3c
˚
Bond
Bond length [A]
Angle
[°]
Ca1ϪO1
Ca1ϪO2
Ca1ϪO3
Ca1ϪO4
Ca1ϪO5
O1ϪC1
2.180(3)
2.183(3)
2.399(3)
2.395(3)
2.393(3)
1.325(5)
1.336(5)
O3ϪCa1ϪO4
O1ϪCa1ϪO2
O1ϪCa1ϪO5
O2ϪCa1ϪO5
C1ϪO1ϪCa1
C19ϪO2ϪCa1
177.78(12)
152.08(12)
103.39(10)
104.45(11)
173.8(3)
178.7(3)
O2ϪC19
The calcium aryloxide 3c crystallises as a monomer with
two aryloxide ligands and a single THF donor molecule in
the equatorial plane and two THF donors in axial posi-
tions. The Ca atom and three equatorial oxygen atoms are
˚
located in a plane with a mean deviation of 0.03 A. The
axial (THF)OϪCaϪO(THF) angle of 177.78(12)° is almost
linear, but the equatorial ArOϪCaϪOAr angle of
152.08(12)° is markedly widened from the expected 120° for
an ideal trigonal bipyramid. As a consequence, the angles
for the other equatorial ligands are smaller [103.93(10)° for
O1ϪCa1ϪO5 and 104.45(11)° for O2ϪCa1ϪO5]. The C6
planes of the aryl ligands intersect at an angle of 8.2°. The
Ϫ
˚
two calciumϪoxygen ( OAr) bond lengths of 2.183(3) A
(av.) are almost identical. The CaϪO(THF) bonds are al-
˚
most invariant at 2.395(3) A. The carbonϪoxygen bonds in
the alkoxide ligands are identical within estimated standard
˚
˚
deviations [1.324(5) A for O1ϪC1 and 1.336(5) A for
O2ϪC19], and hence considerably shorter than standard
[22]
˚
CϪO single bonds (1.43 A).
In addition, the CaϪOϪC
Scheme 3. Synthesis of calcium aryloxides using 1c
angles of 173.8(3)° for Ca1ϪO1ϪC1 and 178.7(3)° for
Ca1ϪO2ϪC19 are almost linear. Similar bond shortening
A single crystal of 3c suitable for X-ray diffraction was and linearity was found in the analogous complex
grown over 5 d at ϩ5 °C in THF. The result of the structure {[Ca(BHT)2(THF)3]·THF[12] and in the thiolate complex
determination is depicted in Figure 4. The coordination [Ca(SMes)2(THF)4] (Smes ϭ 2,4,6-tBu3C6H2S)}.[23]
number for the calcium is five, with two aryloxide and three
The arylcalcium hydride 1c also reacts with alcohols to
THF ligands arranged in a distorted trigonal-bipyramidal form the corresponding bis(alkoxides) of calcium. When tri-
coordination polyhedron around the calcium centre. Two phenylmethanol was allowed to react with 1c, it produced
and a half further uncoordinated THF molecules could be the previously unrecorded calcium alkoxide 4 in 94% yield,
detected as lattice solvent. Bond lengths and angles are as shown in Scheme 4. A single crystal of 4 was grown for
listed in Table 1.
X-ray diffraction over 10 d at ϩ5 °C in THF. The coordina-
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425
1419

J. P. Dunne, M. Tacke, C. Selinka, D. Stalke
FULL PAPER
tion polyhedron around the central calcium atom is made 110.16(9)°, while the angle between the two equatorial THF
up of six oxygen atoms, two of the Ph3COϪ anions in adja- ligands at the calcium atom is considerably smaller at
cent equatorial positions and four of THF molecules in the 76.3(4)°. This deformation in the octahedral geometry can
remaining equatorial and axial positions (Figure 5). An ad- clearly be attributed to the greater steric demand of the al-
ditional THF molecule is found uncoordinated in the crys- koxide ligands than of the THF ligands. Both axial THF
tal lattice. Bond lengths and angles are described in full ligands are bent away from the alkoxide ligands. Again, the
in Table 2.
angle CaϪOϪCAryl, at 177.47(13)°, indicates near linearity.
The CaϪO bond length in the alkoxide ligand is 2.1609(16)
˚
A, and hence almost as long as in 3c. The OϪCAryl single
˚
bonds in 4 Ϫ at 1.375(2) A Ϫ are slightly longer than in 3c.
Theoretical Study
In order to gain further insight into the proposed com-
positions of the arylcalcium hydrides a series of theoretical
calculations were performed. As in previous reported theor-
etical work,[10] the AKR4/6-31g** basis set combination
was used. This has proved to give reasonable geometries
without greatly increasing the calculation time. As before,
ammonia was chosen as a model Lewis base to decrease
calculation time.
The aim of this theoretical study was to examine a pos-
sible mechanism for this reaction. The first step examined
was the direct insertion of calcium atoms into the CϪH
bond. This is unlikely, since the calculated reaction enthalpy
of ϩ10.8 kcal/mol for the formation of PhCaH (8) is too
large to occur at 77 K (outlined in Figure 6). This would
indicate that a highly endothermic reaction is occurring at
such a low temperature. The calculated CaϪH bond length
Scheme 4. Synthesis of 4 using 1c
˚
for 8 was 2.037 A and the CaϪC bond length was found to
˚
be 2.324 A. The optimised structure 8 (as shown in detail in
Table 3) showed an agostic interaction between the calcium
atom and the ortho CϪH bond of the phenyl ligand. This
˚
agostic bond length was calculated as 2.534 A and the angle
between the phenyl ligand, the calcium atom and the hy-
dride ligand (CϪCaϪH angle) was found to be 144.8°. This
bending of the CϪCaϪH angle in conjunction with an
agostic CϪH bond indicates coordinative unsaturation of
the calcium centre in 8. This is to be expected, since a coor-
dination number of two for calcium is extremely low, par-
ticularly in view of the lack of steric bulk in the phenyl and
hydride ligands. It would be reasonable to expect that, in
the presence of donor solvents such as THF, a higher coor-
dination number would be achieved through the coordina-
tion of donor solvent molecules.
Figure 5. Crystal structure of Ca[OCPh3]2(THF)4
˚
Table 2. Selected bond lengths [A] and angles [°] for 4
˚
Bond
Bond length [A]
Angle
[°]
Ca1ϪO1
Ca1ϪO2
Ca1ϪO3
O1ϪC1
C1ϪC2
C1ϪC8
C1ϪC14
2.1609(16)
2.440(2)
2.482(7)
1.375(2)
1.565(3)
1.550(3)
1.565(3)
O1ϪCa1ϪO1A
O1ϪCa1ϪC3A
O3ϪCa1ϪO3A
O2ϪCa1ϪO2A
C1ϪO1ϪCa1
110.16(9)
87.9(2)
76.3(4)
166.67(8)
177.47(13)
The distorted octahedral arrangement is evident from the
deviation from linearity of the angle O2ϪCa1ϪO2A, at
166.67(8)°. The calcium atom and four equatorial oxygen
Figure 6. Reaction of calcium atoms with benzene
In a previous theoretical study,[10] it was reported that
atoms are located in a plane with a mean standard devi- phenylcalcium hydride exists in donor solvents as a hydride-
˚
ation of 0.22 A. The angle O1ϪCa1ϪO1A between the two bridged dimer with a coordination number of six. Here it
triphenylmethoxy ligands and the calcium atom is is extended to show that the trans ‘‘PhCaH·3NH3’’ dimer
1420
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425

Arylcalcium Hydrides as Precursors to Alkoxides and Aryloxides of Calcium
FULL PAPER
CϪH bonds of toluene, tert-butylbenzene, trimethyl-
(phenyl)silane, and m-xylene, as outlined in Scheme 1. It is
clear from the derivatisation reactions with chlorotrimethyl-
Table 3. Comparison of bond lengths and IR frequencies in calcu-
lated arylcalcium hydride species
Ϫ
silane (as shown in Scheme 1) and Ph3CϩBF4 (shown in
PhCaH trans-PhCaH·3NH3 cis-PhCaH·3NH3
Scheme 2) that the backbone of 1a is composed of the an-
ionic ligands PhϪ and HϪ. However, the elemental analysis
indicates that the composition is (PhCaH·2THF)x: The ex-
istence of coordinated Lewis base molecules such as THF
is a well-known facet of calcium chemistry.[1]
˚
CaϪH [A]
2.037 2.305
2.324 2.552
2.318
2.528
2.575
3.532
849 (sym)
930 (antisym)
˚
CaϪC [A]
˚
CaϪN [A]
N/A
N/A
N/A
2.569
3.461
991 (sym)
1048 (antisym)
˚
CaϪCa [A]
νs(CaϪH) [cmϪ1
]
νas(CaϪH) [cmϪ1] N/A
Theoretical calculations show the tendency towards oli-
gomerisation through hydride bridging and the uptake of
THF to fill the coordination sphere of the calcium atom.
These calculations showed that the formation of PhCaH is
energetically unfavourable by ϩ10.8 kcal/mol and that an
aggregated state through hydride bridging is favoured over
a monomeric species.[6] These calculations also showed that
the favoured coordination number is six and that a solvent-
dependent equilibrium exists in donor solvents. The calcula-
tions show that the aggregation state can be elucidated from
the IR data on the CaϪH vibration. The calculated IR fre-
quency for a terminal hydride ligand (e.g., monomeric
PhCaH·4NH3) is 897 cmϪ1. When the hydride ligand is
bridging, as in a dimeric species, the calculated CaϪH vi-
bration is higher, at 1048 cmϪ1. The calculated data for this
dimeric species are in good agreement with the experiment-
ally measured value of 1094 cmϪ1. However, the possibility
of higher aggregates such as tetramers (in which the hydride
ligand would be triply bridging) or an oligomeric species
through mixed bridging of the phenyl and hydride ligands
cannot be gauged from these data.
species 9 has a reaction enthalpy Ϫ13.9 kcal/mol lower than
that of the cis dimer species 10, as described in Figure 7.
For the trans species 9, the calculated CaϪH bond length
˚
˚
is 2.305 A, the CaϪC bond length is 2.552 A and the
˚
CaϪN bond length is 2.569 A; these values are outlined in
Table 3. The CaϪH antisymmetric vibration, which is IR-
active, is calculated as 1078 cmϪ1. Slightly different values
(as shown in Table 3) are seen for 10. The calculated CaϪH
˚
bond (2.318 A) is slightly longer than in 9, while the CaϪC
˚
bond length was found to be 2.528 A, shorter than in 9.
The CaϪH antisymmetric vibration in 10 is 930 cmϪ1. This
value is lower than that of 9 (1078 cmϪ1). It should be pos-
sible to distinguish whether ‘‘solvated PhCaH’’ exists in
either cis or trans form in solution by IR methods.
Selectivity of C؊H Activation with Calcium Atoms
Lithium atoms have been reported to activate aromatic
CϪH bonds of ϩI-substituted benzene derivatives select-
ively in the para position.[24] By use of an approach similar
to Mochida’s and by treatment of the ‘‘ArCaH’’ with chlor-
otrimethylsilane at Ϫ30 °C, easy identification of the activ-
ated positions on the aromatic ring by NMR and GC MS is
possible. On expanding this reaction type to other aromatic
compounds [e.g., toluene, tert-butylbenzene, trimethyl-
Figure 7. Energy difference between cis and trans isomers of
PhCaH·3NH3
Discussion
Composition of Arylcalcium Hydrides
Mochida originally proposed that the product from the (phenyl)silane, and m-xylene], it was expected that selective
reaction between benzene and calcium atoms was CϪH bond activation in the para position would be ob-
‘‘PhCaH’’,[9] but he failed to isolate this compound. In- served, in a reaction series analogous to that of lithium
stead, it was allowed to react with D2O and SiMe3Cl. On atoms.[20] However, this was not observed; instead, a lack of
repeating this experiment in THF, we isolated a dark red, selectively was noted. All benzene derivatives except xylene
air- and moisture-sensitive solid in 78% yield, as shown in experienced the activation of the ortho, meta, and para
Scheme 1. This is an extremely high yield for an MVS ex- CϪH bonds. This lack of selectivity is different to the case
periment, and it appears moreover that the loss in yield of the lithium atom reaction, in which the para position was
stems from a relatively high-melting matrix in the vessel, selectively attacked.[24] Another difference was that xylene,
which hinders collection of the entire sample. In the ben- which had failed to react with lithium atoms, reacted with
zene cocondensation, for example the matrix is a benzene/ calcium atoms. [Calcium atoms react with mesitylene, which
THF mixture. The melting point of benzene is 5.5 °C and does not react with lithium atoms, to yield an extremely
the melting point of THF is Ϫ103 °C. However, the toluene pyrophoric, insoluble, red solid.]
cocondensation has a lower melting matrix of toluene/THF
One similarity to the lithium atom reaction is that the
(melting point of toluene Ϫ95 °C) and hence a higher yield. reaction between calcium atoms and toluene does not result
This reaction type was expanded to include other ϩI-sub- in the activation of the thermodynamically favoured benzyl
stituted benzene derivatives, as in the similar lithium atom position. In the case of lithium atoms, it was shown the a
reaction.[20] Calcium atoms were able to activate the key intermediate was the formation of a π complex between
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425
1421

J. P. Dunne, M. Tacke, C. Selinka, D. Stalke
FULL PAPER
the benzene ring and an Li2 cluster, which favours the ac- has been shown that calcium possesses low-lying d orbitals
tivation of the aromatic CϪH bonds.[24]
that influence the structure of molecules.[27Ϫ29] It can be
very difficult to separate steric and electronic effects.[30] It
has been demonstrated that the MϪOϪC angle has no
simple correlation with the degree of π-bonding.[25] A sim-
ilar comparison can be made between the calcium alkoxide
4 and the previously reported species 6 and 8.[12,31] There is
a pronounced difference between the structures. In 4, the
crystal was grown from THF and crystallised as a monomer
of coordination number six, while compounds 6 and 8 were
grown from toluene and crystallised as dimers with lower
coordination numbers of four and three, respectively. The
CaϪO bond length for the terminal alkoxide in 6 is 2.105
Reactions of Arylcalcium Hydrides
These arylcalcium hydrides can be redissolved in THF
and allowed to react with sterically bulky phenols and alco-
hols to form bis(aryloxides)/-(alkoxides) of calcium in im-
proved yields (all Ͼ 95%).[12]
The structure of 3c is similar to those of 5 and 7, which
are shown in Figure 8 and have been reported previ-
ously.[12,23] It is monomeric, in a distorted trigonal-bipyr-
amidal arrangement, with the CaϪO(phenolate) bond
˚
length of 2.182(3) A av. for 3c being shorter than that for 5
˚
A, shorter than the same CaϪO bond length, of 2.1609(17)
˚
at 2.197 A. Like those in 5 and 7, the aryloxide ligands in
˚
A, in 4 and more akin to the CaϪO bond lengths found in
3c are in the equatorial positions. Another similarity be-
tween 3c and 5 is the almost linear CaϪOϪC angle. It is
not clear what causes this linearity (and shortened CaϪO
bond) in calcium aryloxides.[12] One possible explanation is
the steric load on the ligands in both 3c and 5; another
might lie in π-bonding.[25,26] This is not as unreasonable as
it appears, since aryloxides are good π-donor ligands and it
˚
˚
the aryloxide species 3d [2.182(3) A] and 8 (2.121 A). Unlike
those in the above species 3c, 4, and 5, the CaϪOϪC angle
in 6 is bent at 167.4°, while in 4 the CaϪOϪC angle is
177.46(14)°. This is almost linear and similar in value to
those in the aryloxide species 3c and 5.
Conclusion
Calcium atoms react unselectively with toluene, tert-bu-
tylbenzene and trimethyl(phenyl)silane to form the ortho,
meta, and para isomers of each arylcalcium hydride. How-
ever, when m-xylene was cocondensed with calcium atoms
in the presence of THF, selective CϪH bond activation in
the position meta to the methyl groups was observed. These
arylcalcium hydrides are useful reagents for the formation
of calcium bis(alkoxides). The reactions between (tert-bu-
tylphenyl)calcium hydride and different tert-butyl-substi-
tuted phenols resulted in the formation of the correspond-
ing aryloxides in high yields. In Ca(2,4,6-tBu3C6H2-
O)2(THF)3 the five oxygen atoms furnish a distorted tri-
gonal-bipyramidal coordination polyhedron with the
sterically bulky aryloxide ligands in the equatorial posi-
tions. In a similar reaction, triphenylmethanol yielded the
alkoxide [Ca{OC(C6H5)3}2(THF)4] in a similar high yield.
This compound shows a distorted octahedron of six oxygen
atoms around the central metal atom and the two alkoxide
ligands in a cis equatorial geometry.
Experimental Section
General: All experimental procedures were performed by use of
standard Schlenk techniques under dry argon. Solvents such as
THF, benzene, toluene, m-xylene, and tert-butylbenzene were dried
and distilled prior to use from sodium benzophenone ketyl under
argon. Trimethyl(phenyl)silane was dried and distilled from calcium
dihydride under argon. Pentane was dried and distilled prior to use
from sodium/potassium amalgam under argon. The solvents were
then degassed by three freeze-pump-thaw cycles. Triphenylcarben-
ium tetrafluoroborate and 18-crown-6 were purchased from Ald-
rich and dried under high vacuum for 2 d. Celite was stored in an
oven at 120 °C overnight and was dried under vacuum for 3 h
before use. Approximately 6 cm of Celite was used for each filtra-
Figure 8. A comparison of some selected group-II bis(alkoxides)
1422
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425

Arylcalcium Hydrides as Precursors to Alkoxides and Aryloxides of Calcium
FULL PAPER
1
tion. H and 13C NMR spectra were obtained with either a Varian stirring was continued, while the solution was allowed to warm to
300 MHz or a Varian 500 MHz NMR machine. All chemical shifts
room temperature. Water (15 mL) was added to this solution. The
are reported in ppm and are referenced to TMS. All the GC MS organic layer was separated and dried with magnesium sulfate, and
samples were run as ethyl acetate solutions with a Finnigan Trace
GC MS, on a room temp. X-5MS 15-m column. Warwick Analyt-
ical Service, Warwick, UK, performed the elemental analysis.
the solvent was removed under reduced pressure to yield the corres-
ponding substituted silane as a dark red oil.
Derivatisation of 1b with Chlorotrimethylsilane: All three possible
trimethyl(methylphenyl)silane isomers were detected by GC MS
Typical Cocondensation Experiment: Calcium (1.5 g, 0.0375 mol)
was vaporised from an alumina crucible over 90 min (5.6 V and
56 A) and cocondensed with a mixture of the aromatic compound
(20 mL) and THF (80 mL). The pressure measured at the bottom
of the cocondensation vessel was kept below 10Ϫ5 mbar to avoid
gas-phase reactions between the metal atoms and reactants. The
apparatus used has been described previously.[24] At the end of the
experiment, the frozen matrix was allowed to warm up under ar-
gon. The molten mixture flowed through an internal drain into a
Schlenk tube. The solution was filtered through 6 cm of Celite and
the solvent was evaporated under reduced pressure to yield a dark
red, pyrophoric solid {(ArCaH·2THF)x} in yields ranging from 78
to 97%.
1
([Mϩ]: m/z ϭ 164.1). Trimethyl(2-methylphenyl)silane: H and 13C
NMR spectroscopic data agree with literature values.[14] Trime-
thyl(3-methylphenyl)silane: 1H and 13C NMR spectroscopic data
agree with literature values.[15] Trimethyl(4-methylphenyl)silane: 1H
and 13C NMR spectroscopic data agree with literature values.[16]
Treatment of 1c with Chlorotrimethylsilane. All three possible (tert-
butylphenyl)trimethylsilane isomers were detected by GC MS
([Mϩ]: m/z ϭ 206.4). (2-tert-Butylphenyl)trimethylsilane: 1H and
13C NMR spectroscopic data agree with literature values.[17] (3-tert-
1
Butylphenyl)trimethylsilane: H and 13C NMR spectroscopic data
agree with literature values.[18] (4-tert-Butylphenyl)trimethylsilane:
1H and 13C NMR spectroscopic data agree with literature values.[19]
Cocondensation of Benzene with Calcium Atoms in THF: This
Treatment of 1e with Chlorotrimethylsilane: Only one compound
was detected by GC MS ([Mϩ]: m/z ϭ 178.4), corresponding to
(3,5-dimethylphenyl)trimethylsilane as identified by 1H and 13C
NMR data.[20]
cocondensation yielded
a
dark red, pyrophoric solid
{(PhCaH·2THF)x}(1a) in 78% yield. C6H5CaH·2THF (262.4):
calcd. C 64.1, H 8.4, Ca 15.3; found C 63.5, H 7.9, Ca 15.2. IR
(Nujol mull; major peaks): ν˜ ϭ 3369 (s), 1285 (m), 1224 (s), 1094
(vs), 1031 (w), 981 (m), 905 (s), 837 (s), 751 (m), 726 (m), 696
Treatment of 1d with Dimethyl Disulfide: Dimethyl disulfide
(0.07 mL, 0.73 mmol) in THF (5 ml) was added dropwise at Ϫ30
°C to a stirred solution of 1d (1.0 g, 0.091 mmol, dissolved in
50 mL of THF). The workup was performed as for the derivatis-
ation above. All three isomers of trimethyl[(methylthio)phenyl]-
silane were detected by GC MS ([Mϩ]: m/z ϭ 196.4). Trimethyl[2-
(methylthio)phenyl]silane: 1H NMR (CDCl3): δ ϭ 0.33 (s, 9 H,
SiMe3), 2.52 (s, 3 H, SCH3), 7.31 (2 H, Ar-H), 7.5 (2 H, ArH)
ppm.13C NMR (CDCl3): δ ϭ Ϫ1.1 (SiMe3), 15.6 (SCH3), 124.4
(CmetaЈ), 125.9 (Cmeta), 129.0 (Cpara), 132.5 (CorthoЈ), 138.1 (Cipso
CSiMe3), 144.1 (Cortho CSMe). Trimethyl[3-(methylthio)phenyl]sil-
ane: 1H and 13C NMR spectroscopic data agree with literature
values.[17] Trimethyl[4-(methylthio)phenyl]silane: 1H and 13C NMR
spectroscopic data agree with literature values.[19]
(m) cmϪ1
.
Cocondensation of Toluene with Calcium Atoms in THF: This
cocondensation yielded brown, pyrophoric solid
a
{(C7H7CaH·2THF)x} (1b) in 97% yield. C7H7CaH·2THF (276.4):
calcd. C 65.2, H 8.8, Ca 14.5; found C 64.4, H 8.8, Ca 14.3. IR
(Nujol mull; major peaks): ν˜ ϭ 3368 (s), 1285 (m), 1223 (s), 1093
(vs), 1032 (w), 986 (m), 908 (s), 837.9 (s), 752 (m), 728 (m), 691
(m) cmϪ1
.
Cocondensation of tert-Butylbenzene with Calcium Atoms in THF:
This cocondensation yielded dark red, pyrophoric solid
a
{(C10H13CaH·2THF)x} (1c) in 93% yield. C10H13CaH·2THF
(318.5): calcd. C 67.9, H 9.5, Ca 12.6; found C 67.0, H 9.2, Ca
12.3. IR (Nujol mull; major peaks): ν˜ ϭ 3370 (s), 1264 (s), 1112 Typical Hydride Abstraction Reaction of the Arylcalcium Hydride
(s), 1094 (vs), 1033 (w), 982 (m), 835 (s), 761.6 (m), 723 (m), 697
with Ph3CBF4: The appropriate arylcalcium hydride (0.52 g,
1.9 mmol) was dissolved in toluene (30 mL) and 18-crown-6 (0.5 g,
1.9 mmol) and placed in a pressure-equalising dropping funnel.
This solution was added slowly at Ϫ30 °C and with vigorous stir-
ring to a suspension of triphenylcarbenium tetrafluoroborate
(0.65 g, 1.9 mmol) in toluene (20 mL). After completion of the ad-
dition, the dark red solution was stirred at Ϫ30 °C for 1 h. Any
unchanged solid was removed by filtration and the toluene solvent
was removed in vacuo to yield triphenylmethane.
(m) cmϪ1
.
Cocondensation of m-Xylene with Calcium Atoms in THF: This
cocondensation yielded dark red, pyrophoric solid
a
{(C8H9CaH·2THF)x} (1d) in 85% yield. C8H9CaH·2THF (290.5):
calcd. C 66.2, H 9.0, Ca 13.8; found C 65.9, H 8.7, Ca 13.6. IR
(Nujol mull; Major peaks): ν˜ ϭ 3372 (s), 1288 (m), 1220 (s), 1096
(vs), 10323 (w), 989 (m), 912 (s), 838.4 (s), 757 (m), 732 (m), 686
(m)cmϪ1
.
Typical Synthesis of Calcium Bis(alkoxides) by Use of 1c: Com-
pound 1c (1.27 g, 0.001 mol) was dissolved in THF (150 mL). The
substituted alcohol or phenol (0.008 mol), dissolved in THF
(50 mL), was added dropwise to this solution over 15 min at Ϫ30
°C. The solution was kept at Ϫ30 °C with vigorous stirring until
the evolution of H2 had ceased. The solution was then allowed to
warm up slowly to room temperature. The solvent was removed
under vacuum to yield a beige, moisture-sensitive solid. This was
washed twice with pentane and the remaining solid was dried under
vacuum to yield the corresponding bis(alkoxide) of calcium in high
yields ranging from 95 to 98%.
Cocondensation of Trimethyl(phenyl)silane with Calcium Atoms in
THF: This cocondensation yielded a dark red, pyrophoric solid
{(C9H13CaH·2THF)x} in 86% yield. C9H13SiCaH·2THF (334.6):
calcd. C 61.0, H 9.0, Ca 12.0, Si 8.4; found C 59.8 H 8.8, Ca 12.3,
Si 8.0. IR (Nujol mull; major peaks): ν˜ ϭ 3372 (s), 1591 (w), 1306
(w), 1260 (s), 1206 (m), 1114 (m), 1094 (vs), 1034 (m), 1019 (s), 837
(m), 801 (br), 751 (m), 725 (s), 698 (m), 618 (m) cmϪ1
.
Typical Derivatisation of the Arylcalcium Hydride with Chlorotri-
methylsilane: Chlorotrimethylsilane (0.06 mL, 0.73 mmol) in THF
(5 mL) was added dropwise at Ϫ30 °C to a stirred solution of aryl-
calcium hydride (1.0 g, dissolved in 50 mL of THF). The dark red Synthesis of [Ca(2,6-tBu2C6H2O)2(THF)3] (3a): This reaction
colour turned yellow over a period of 5 min. After this time, the
yielded 0.53 g (96% yield) of 3a as an off-white solid. 1H NMR
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425
1423

J. P. Dunne, M. Tacke, C. Selinka, D. Stalke
FULL PAPER
(500 MHz, C6H6) δ ϭ 7.3 (d, 4 H), 6.7 (t, 2 H), 3.4 (s, 12H THF),
1.9 (s, 12 H, THF), 1.4 (s, 36 H, CH3) ppm. 13C NMR (500 MHz,
C6H6) 164.4 (Cipso), 137.4 Cortho), 125.0 (Cmeta), 112.8 (Cpara), 69.1
(THF), 35.4 (CCH3)3, 31.4 (CH3), 25.1 (THF) ppm. IR (Nujol
mull; major vibrations): ν˜ ϭ 2887 (m), 1416 (m) 1382 (s), 1228 (m),
gram for empirical absorption correction.[35] In addition to the cal-
cium complex 3c, the unit cell also contained 2.5 noncoordinating
THF molecules. The THF molecule (O7ϪC57) is disordered and
was refined to split occupancies of 0.68/0.32. The disordered tert-
butyl moiety (C12ϪC14) was refined to a split occupancy of 0.65/
0.35. All non-hydrogen atoms were refined anisotropically. All hy-
drogen atoms were geometrically idealised and refined with a riding
model. In addition to the calcium complex 4, the unit cell also
contains one noncoordinating THF molecule. The coordinated
THF molecule (O3ϪC27) was disordered and could be refined to
split occupancies of 0.73/0.27. Bond length and similarity restraints
were used in the refinement of the disorder. All non-hydrogen
atoms were refined anisotropically. The hydrogen atoms were geo-
metrically idealised and refined with a riding model. CCDC-190905
(3c) and -190906 (4) contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge at
www.ccdc.cam.ac.uk/conts/retrieving.html [or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
1037 (s), 898 (m), 809 (m), 519 (m) cmϪ1
.
Synthesis of [Ca(2,6-tBu2-4-MeC6H2O)2(THF)3] (3b): This reaction
1
yielded 0.55 g (98% yield) of 3b as an off-white solid. H and 13C
NMR spectroscopic data (500 MHz C6H6) agree with literature
values.[12]
Synthesis of [Ca(2,4,6-tBu3C6H2O)2]·3THF (3c): This reaction
1
yielded 0.59 g (97%) of 3c as an off-white solid. H and 13C NMR
(500 MHz C6H6) agree with literature values.[32] X-ray quality crys-
tals were grown from a concentrated THF solution at ϩ5 °C over
a period of one week.
Synthesis of [Ca{OC(C6H5)3}2(THF)3] (4): This reaction yielded
0.77 g (95%) of 4 as a pale white solid. A colourless single crystal
of 4 for X-ray diffraction studies was grown from a concentrated
THF solution at ϩ5 °C over 10 d. 1H NMR (500 MHz, C6H6) δ ϭ
7.30 (d, 12 H, CH C6H5), 7.10Ϫ7.01 (m, 18 H, C6H5), 3.41 (m, 16
H, THF), 1.36 (m, 16 H, THF) ppm. IR (Nujol mull; major bands):
ν˜ ϭ 3056 (s), 2919 (m), 1490 (m), 1444 (m), 1230 (s), 1031 (vs),
1015 (s), 823 (m), 776 (m), 754 (m), 727 (m), 700 (s), 671 (m), 617
CB2 1EZ, UK; Fax: (internat.)
deposit@ccdc.cam.ac.uk].
ϩ 44-1223/336-033; E-mail:
Theoretical Methods: For determination of the structures, spectro-
scopic data, and reaction enthalpies of intermediates in the reaction
mechanisms, the application of theoretical methods has proven ad-
vantageous. For this purpose the program package
GAUSSIAN98[36] implemented on a DEC Alpha workstation
(500 MHz CPU/256 MB RAM) or an Origin 200 eight-processor
cluster (SGI, 180 MHz CPU/2 GB RAM) was used. Ab initio cal-
culations were performed at the DFT level of theory (B3LYP) with
the 6-31G** basis set for C, H, and N. The chosen general basis
set for calcium was AKR4, which was used by us in another study
on organocalcium complexes.[6] Harmonic vibrational frequencies,
calculated at the same level, characterised stationary points and
gave the zero-point energies. The difference in the sum of the elec-
tronic and the zero-point energies were interpreted as reaction en-
thalpies at 0 K.
(m), 512 (m) cmϪ1
.
X-ray Crystallographic Study of 3c and 4: The data for 3c and 4
were collected at 173(2) K from shock-cooled crystals with a
BRUKER SMART-APEX diffractometer (graphite-mono-
chromated Mo-Kα radiation, λ ϭ 71.073 pm), equipped with a low-
temperature device.[32] The crystal data are summarized in Table 4.
The structure was solved by direct methods (SHELXS-NT97)[33]
and refined by full-matrix, least-squares methods against F2
(SHELXL-NT97).[34] R values defined as R1 ϭ Σ||Fo| Ϫ |Fc||/Σ|Fo|,
wR2 ϭ [Σw(F2o Ϫ Fc2)2/Σw(F2o)2]0.5, w ϭ [σ2(F2o) ϩ (g1P)2 ϩ g2P]Ϫ1
P ϭ 1/3[max(F2o,0) ϩ 2Fc2]. SADABS 2.0 was employed as a pro-
,
Table 4. Crystallographic data for [Ca{2,4,6-tBu3C6H2O}2(THF)3] (3c) and [Ca(OPh3)2(THF)4] (4) at T ϭ 173 K
3c
4
Empirical formula
Formula mass
Crystal system
Space group
C48H82CaO5ϩ2.5C4H8O
959.48
monoclinic
P21/c
9.942(2)
35.933(8)
16.138(3)
C58H70CaO7
919.22
orthorhombic
Pccn
14.564(8)
16.580(9)
21.182(13)
˚
a [A]
˚
b [A]
˚
c [A]
β [°]
94.206(4)
5750(2)
90
5115(5)
3
˚
V [A ]
Z
4
4
Calcd. density [Mg/m3]
Absorption coeff. [mmϪ1
F(000)
1.108
0.158
2120
1.194
0.174
1976
]
Crystal size [mm]
θ range [°]
0.3 ϫ 0.3 ϫ 0.2
1.39Ϫ22.46
62291
11975 [R(int) ϭ 0.0612]
99.9
7481/636/761
1.192
0.6 ϫ 0.5 ϫ 0.3
1.86Ϫ25.03
48730
5776 [R(int) ϭ 0.0274]
100.0
4528/110/367
1.156
Reflections collected
Independent reflections
Completeness at θmax. [%]
Data/restraints/param.
Goodness-of-fit on F2
Final R indices [I Ͼ 2σ(I)]
R indices (all data)
R1 ϭ 0.0890, wR2 ϭ 0.1979
R1 ϭ 0.0949, wR2 ϭ 0.2015
0.601/Ϫ0.417
R1 ϭ 0.0550, wR2 ϭ 0.1313
R1 ϭ 0.0575, wR2 ϭ 0.1331
0.336/Ϫ0.266
Largest diff. peak/hole [e·AϪ3
]
1424
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425

Arylcalcium Hydrides as Precursors to Alkoxides and Aryloxides of Calcium
FULL PAPER
J. W. Zwickler, L. W. Jeneskens, J. Organomet. Chem. 1995,
496, 117Ϫ125.
P. A. Deck, C. L. Beswick, T. J. Marks, J. Am. Chem. Soc.
1998, 120, 1772Ϫ1784.
S. R. Drake, D. J. Ottway, M. B. Hursthouse, K. M. Abdul
Malik, Polyhedron 1992, 11, 1995Ϫ2007.
P. Rademacher, Strukturen organischer Moleküle, VCH,
Weinheim, 1987, p. 56.
Acknowledgments
[20]
[21]
[22]
[23]
This work was supported by the Centre for High-Performance
Computing Applications at University College Dublin, Ireland.
C. S. and D. S. thank the Deutsche Forschungsgemeinschaft and
the Fonds der Chemischen Industrie for financial support.
[1]
For some recent reviews see: [1a] T. P. Hanusa, Polyhedron 1990,
U. Englich, K. Ruhlandt-Senge, Z. Anorg. Allg. Chem. 2001,
627, 851Ϫ856.
[1b]
9, 1345Ϫ1362.
T. P. Hanusa, Chem. Rev. 1993, 93,
[1c]
[24] [24a]
[24b]
1023Ϫ1036.
D. Stalke, Angew. Chem. Int. Ed. Engl. 1994,
M. Tacke, Chem. Ber. 1995, 128, 91Ϫ92.
M. Tacke,
33, 2168Ϫ2171. [1d] M. Westerhausen, Coord. Chem. Rev. 1998,
Chem. Ber. 1996, 129, 1369Ϫ1371. [24c] M. Tacke, Eur. J. Inorg.
Chem. 1998, 537Ϫ541.
Tacke, Eur. J. Inorg. Chem., in press.
[1e]
[24d]
176, 157Ϫ210.
M. Westerhausen, Angew. Chem. Int. Ed.
J. P. Dunne, M. Bockmeyer, M.
2001, 40, 2975Ϫ2977. [1f] T. P. Hanusa, Coord. Chem. Rev. 2000,
[1g]
[25]
[26]
210, 329Ϫ367.
2559Ϫ2571.
T. P. Hanusa, Organometallics 2002, 21,
G. D. Smith, P. E. Fanwick, I. P. Rothwell, Inorg. Chem. 1990,
29, 3221Ϫ3226.
T. W. Coffindaffer, B. D. Steffy, I. P. Rothwell, K. Folting, K.
Huffman, J. C. Streib, J. Am. Chem. Soc. 1989, 111,
4742Ϫ4749.
M. Kaupp, P. v. R. Schleyer, J. Am. Chem. Soc. 1992, 114,
491Ϫ497.
M. Kaupp, P. v. R. Schleyer, H. Stoll, H. Preuss, J. Am. Chem.
Soc. 1991, 113, 6012Ϫ6020.
M. Kaupp, P. v. R. Schleyer, H. Stoll, H. Preuss, J. Chem. Phys.
1991, 94, 1360Ϫ1366.
[2]
T. C. Wu, R. Xiong, R. D. Rieke, J. Org. Chem. 1990, 55,
5045Ϫ5051.
B. Bogdanovic, Acc. Chem. Res. 1988, 21, 261Ϫ267.
Bogdanovic, Main Group Met. Chem. 1990, 13, 341.
S. R. Drake, D. J. Otway, S. P. Perlepes, Main Group Met.
Chem. 1991, 14, 243.
K. Mochida, T. Yamanishi, J. Organomet. Chem. 1987, 332,
247Ϫ252.
D. S. Hutchings, P. C. Junk, W. C. Patalinghug, C. L. Raston,
A. H. White, J. Chem. Soc., Chem. Commun. 1989, 973Ϫ974.
L. M. Engelhardt, P. C. Junk, C. L. Raston, A. H. White,
Chem. Commun. 1988, 1500Ϫ1501.
P. B. Hitchcock, M. F. Lappert, G. A. Lawless, B. Royo, Chem.
Commun. 1990, 17, 1141Ϫ1143.
K. Mochida, H. Takeuchi, Y. Hiraga, H. Ogawa, Organometal-
lics 1987, 6, 2293Ϫ2297.
J. P. Dunne, M. Tacke, S. Fox, J. Mol. Struct. THEOCHEM
2001, 543, 157Ϫ166.
[3] [3a]
[3b]
B.
[27]
[28]
[29]
[4]
[5]
[6]
[30]
[31]
R. Fleischer, D. Stalke, Inorg. Chem. 1997, 36, 2413Ϫ2419.
G. B. Deacon, C. M. Forsyth, P. C. Junk, J. Organomet. Chem.
2000, 607, 112Ϫ119.
[7]
[8]
[32] [32a]
[32b]
D. Stalke, Chem. Soc. Rev. 1998, 27, 171.
T. Kottke,
[32c]
R. J. Lagow, D. Stalke, J. Appl. Crystallogr. 1996, 29, 465.
T. Kottke, D. Stalke, J. Appl. Crystallogr. 1993, 26, 615.
G. M. Sheldrick, Acta Crystallogr., Sect. A 1990, 46, 467.
G. M. Sheldrick, SHELXL-NT97, Program for Crystal Struc-
ture Refinement, Universität Göttingen, 1997.
[9]
[33]
[34]
[10]
[11] [11a] D. C. Bradley, Chem. Rev. 1989, 89, 1317Ϫ1322. [11b] W. A .
Wojtczak, P. F. Fleig, M. J. Hampden-Smith, Adv. Organomet.
Chem. 1996, 40, 215Ϫ329. [11c] J. S. Matthews, W. S. Rees, Adv.
Inorg. Chem. 2000, 50, 173Ϫ192.
[35]
[36]
G. M. Sheldrick, Program for empirical absorption correction,
Universität Göttingen, 2000.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgom-
ery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Mil-
lam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Po-
melli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P.
Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A.
G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Re-
plogle, J. A. Pople, Gaussian98, Revision A.7, Gaussian, Inc.,
Pittsburgh, PA, 1998.
[12]
K. F. Tesh, T. P. Hanusa, J. C. Huffman, C. J. Huffman, Inorg.
Chem. 1992, 31, 5572Ϫ5579.
S. C. Geol, M. A. Matchett, M. Y. Chiang, W. E. Buhro, J.
[13]
Am. Chem. Soc. 1991, 113, 1844Ϫ1845.
J. A. Albanese, D. E. Gingrich, C. D. Schaeffer Jr, S. M. Coley,
[14]
J. C. Otter, J. Organomet. Chem. 1989, 365, 23Ϫ35.
H. J. Reich, J. E. Holladay, J. D. Mason, W. H. Sikorski, J. Am.
[15]
Chem. Soc. 1995, 117, 12137Ϫ12150.
L. Hevesi, M. Dehon, R. Crutzen, A. Lazarescu-Grigore, J.
[16]
Org. Chem. 1997, 62, 2006Ϫ2012.
[17]
F. Effenberger, D. Hablich, Liebigs Ann. Chem. 1979, 6,
842Ϫ857.
[18]
C. D. Schaeffer, J. J. Zuckerman, C. H. Yoder, J. Organomet.
Chem. 1974, 80, 29Ϫ35.
C. A. van Walree, X. Y. Lauteslager, A. M. A. van Waginingen,
Received August 12, 2002
[19]
[I02456]
Eur. J. Inorg. Chem. 2003, 1416Ϫ1425
1425