Reactions of carbaalanes with HBF4 and HCl: A μ3-bridging fluorine atom in (AlEt)7(C=CHPh)2(CCH2Ph)3(μ3 -F) versus a terminal chlorine atom in (AlMe)7(AlCl)(CCH2Ph)5H

Article abstract of DOI:10.1021/om010584t

The carbaalane cluster (AlEt)₇(C=CHC₆H₅)₂(CCH₂ C₆H₅)₃H (2) reacted with HBF₄·OEt₂ to yield the fluorine derivative (AlEt)₇(C=CHC₆H₅)₂ (CCH₂C₆H₅)₃(μ₃-F) (4) by the replacement of its bridging hydrogen atom and the cleavage of a B-F bond. The fluorine atom bridges three aluminum atoms with the Al-F distances in a narrow range between 1.97 and 2.00 A?. A similar reaction was observed with the arachno-carbaalane (AlMe)₈(CCH₂C₆H₅)₅H (1). However, an impure product was isolated which could not be purified by repeated recrystallization. In contrast, use of HCl instead of HBF₄·OEt₂ as a proton donor did not result in substitution of the bridging hydrogen atom of 1, but a methyl group at the bottom of the cluster was replaced by release of methane. The compound (AlMe)₇(AlCl)(CCH₂C₆H₅)₅ (μ-H) (6), which possesses a terminally bonded chlorine atom, was formed.

Full text of DOI:10.1021/om010584t

5478
Organometallics 2001, 20, 5478-5484
Rea ction s of Ca r ba a la n es w ith HBF 4 a n d HCl: A
µ3-Br id gin g F lu or in e Atom in
(AlEt)7(CdCHP h )2(CCH2P h )3(µ3-F ) ver su s a Ter m in a l
Ch lor in e Atom in (AlMe)7(AlCl)(CCH2P h )5H
Werner Uhl,* Frank Breher, and Bernhard Neumu¨ller
Fachbereich Chemie der Philipps-Universita¨t, Hans-Meerwein-Strasse,
D-35032 Marburg, Germany
Arne Lu¨tzen and Wolfgang Saak
Fachbereich Chemie, Carl-von-Ossietzky Universita¨t, Postfach 2503,
D-26111 Oldenburg, Germany
J o¨rg Grunenberg
Institut fu¨r Organische Chemie, Technische Universita¨t Braunschweig, Hagenring 30,
D-38106 Braunschweig, Germany
Received J une 29, 2001
The carbaalane cluster (AlEt)7(CdCHC6H5)2(CCH2C6H5)3H (2) reacted with HBF4‚OEt2
to yield the fluorine derivative (AlEt)7(CdCHC6H5)2(CCH2C6H5)3(µ3-F) (4) by the replacement
of its bridging hydrogen atom and the cleavage of a B-F bond. The fluorine atom bridges
three aluminum atoms with the Al-F distances in a narrow range between 1.97 and 2.00
Å. A similar reaction was observed with the arachno-carbaalane (AlMe)8(CCH2C6H5)5H (1).
However, an impure product was isolated which could not be purified by repeated
recrystallization. In contrast, use of HCl instead of HBF4‚OEt2 as a proton donor did not
result in substitution of the bridging hydrogen atom of 1, but a methyl group at the bottom
of the cluster was replaced by release of methane. The compound (AlMe)7(AlCl)(CCH2C6H5)5-
(µ-H) (6), which possesses a terminally bonded chlorine atom, was formed.
In tr od u ction
(2), which has a singular structure with two CdC double
bonds and one vertex of the Al8 cube of 1 unoccupied.2
The structure of the second type of carbaalanes, (AlR)7-
The hydroalumination of dialkyl(ethynyl)aluminum
derivatives with the corresponding dialkylaluminum
hydrides afforded novel compounds which contain poly-
hedral clusters of carbon and aluminum atoms.1-3
Owing to obvious similarities to carbaboranes, we
suggested the name carbaalanes for this new and now
well-established class of compounds.4 Two different
types of structures were verified by crystal structure
determinations. One possessing the general formula
(AlR)8(CCH2R′)5(R′′) (1: R ) Me, Et; R′ ) CH3, Ph; R′′
) H, CtCPh, CtCMe; Scheme 1) may be derived from
a slightly distorted cube of aluminum atoms, five faces
of which are bridged by C-CH2R′ groups, while the
sixth face is occupied by a hydrogen atom or an ethynyl
ligand. These compounds have 16 cluster electron pairs,
which, in accordance with the Wade rules,5 verify an
arachno configuration. Incomplete hydroalumination
caused the formation of (AlEt)7(CdCHPh)2(CCH2Ph)3H
(CCH2R′)4(R′′)2 (3: R′′ ) H, CtCPh; Scheme 1) is closely
2- 6
related to that of the closo-borate anion [B11H11
] .
However, the number of electron pairs needed for a closo
configuration, 12, can only be calculated for the alumi-
num compounds if the bridging acetylenide or hydride
anions were taken as exo ligands not involved in the
cluster bonding. As was shown by quantum-chemical
calculations, all these clusters require a delocalized
bonding description with, however, only weak direct Al-
Al interactions.3 The charge separation in the car-
baalane clusters is as expected, with positively charged
aluminum atoms and negative charges localized at the
carbon and bridging hydrogen atoms. We have begun
investigations of the chemical reactivity of these com-
pounds and report here on reactions with protic re-
agents which should lead to the replacement of the
bridging hydrogen atoms.
(1) Uhl, W.; Breher, F. Angew. Chem. 1999, 111, 1578; Angew.
Chem., Int. Ed. 1999, 38, 1477.
Resu lts a n d Discu ssion
(2) Uhl, W.; Breher, F.; Lu¨tzen, A.; Saak, W. Angew. Chem. 2000,
112, 414; Angew. Chem., Int. Ed. 2000, 39, 406.
(3) Uhl, W.; Breher, F.; Grunenberg, J .; Lu¨tzen, A.; Saak, W.
Organometallics 2000, 19, 4536.
(4) Uhl, W.; Breher, F. Eur. J . Inorg. Chem. 2000, 1.
(5) Wade, K. Adv. Inorg. Chem. Radiochem. 1976, 18, 1.
Tr ea tm en t of (AlEt)7(CdCHP h )2(CCH2P h )3H (2)
w ith HBF 4‚OEt2. Gas evolution was observed when the
(6) Volkov, O.; Dirk, W.; Englert, U.; Paetzold, P. Z. Anorg. Allg.
Chem. 1999, 625, 1193.
10.1021/om010584t CCC: $20.00 © 2001 American Chemical Society
Publication on Web 11/13/2001

Reactions of Carbaalanes with HBF4 and HCl
Organometallics, Vol. 20, No. 25, 2001 5479
Sch em e 1
alkenyl carbaalane 2 was treated with a solution of
tetrafluoroboric acid in diethyl ether at -50 °C (eq 1).
groups attached to aluminum exhibit chemical shifts
between 0.7 and 3.1 ppm, which is in the normal range
of carbon atoms bonded to coordinatively saturated
aluminum atoms.7 In contrast, the cluster carbon atoms
not involved in CdC double bonds resonate at an
unusually low field (δ 32.5 and 39.6), which is quite
characteristic for carbaalanes1-3 and may be caused by
the particular bonding situation of the clusters. The
alkenylidene carbon atoms coordinated by three alumi-
num atoms have a shift of 156.2 ppm. 27Al NMR data
of organoaluminum derivatives are sensitive to the
coordination numbers of aluminum atoms, and shifts
of about 170 ppm were recorded for a tetrahedral
coordination sphere, while an octahedral environment
gave shifts of about 0 ppm.8 The 27Al NMR shifts of
carbaalanes derived from Al8 cubes were observed in a
broad range between 85 and 170 ppm.9 Their assign-
ment and clear separation are difficult, owing to the line
widths of the usually very broad 27Al NMR resonances
and the partial overlap of signals. Compound 4 showed
resonances with maxima at δ 74 (W1/2 ) 12 000 Hz) and
-28 (W1/2 ) 10 000 Hz). Clearly resolved shoulders were
detected at about 110 and 20 ppm. Much more data
must be collected in future work, in conjunction with
quantum-chemical calculations, to allow the secure
assignment of resonances to particular aluminum atoms
of the cluster. This may be very helpful for a more
sophisticated understanding of the bonding situation of
carbaalanes. We were not able to detect a 19F NMR
signal in the range between -100 and -160 ppm, which
is characteristic of AlF groups containing µ2-bridging
fluorine atoms (see below for quantum-chemical calcula-
Orange crystals of the product, (AlEt)7(CdCHPh)2(C-
CH2Ph)3(µ3-F) (4), were isolated after evaporation of the
solvent and recrystallization of the residue from cyclo-
pentane in 48% yield. Characterization of 4 showed that
the bridging hydrogen atom was replaced by a single
fluorine atom. This reaction requires the cleavage of at
least one B-F bond of the tetrafluoroborate anions,
which may be favored by the coordination of each
fluorine atom by three aluminum atoms of the cluster.
The 1H and 13C NMR spectra are very complicated,
owing to the low molecular symmetry with five chemi-
cally different ethyl groups and further owing to the
coupling between fluorine and carbon or hydrogen.
Despite the employment of various NMR techniques we
were not able to unambiguously assign all observed
resonances. The alkenylidene proton is shifted to low
field (δ 9.28), similar to the case for the starting
compound 2.2 The carbon atoms of the terminal ethyl
(7) (a) Uhl, W.; Cuypers, L.; Schu¨ler, K.; Spies, T.; Strohmann, C.;
Lehmen, K. Z. Anorg. Allg. Chem. 2000, 626, 1526. (b) Uhl, W.; Spies,
T.; Haase, D.; Winter, R.; Kaim, W. Organometallics 2000, 19, 1128.
(c) Uhl, W.; Spies, T.; Saak, W. Z. Anorg. Allg. Chem. 1999, 625, 2095.
(d) Uhl, W.; Spies, T.; Koch, R.; Saak, W. Organometallics 1999, 18,
4598. (e) Uhl, W.; Spies, T.; Koch, R. J . Chem. Soc., Dalton Trans. 1999,
2385. (f) Uhl, W.; Hannemann, F. Eur. J . Inorg. Chem. 1999, 201. (g)
Uhl, W.; Koch, M.; Wagner, J . Z. Anorg. Allg. Chem. 1995, 621, 249.
(8) (a) Benn, R.; Rufin˜ska, A.; Lehmkuhl, H.; J anssen, E.; Kru¨ger,
C. Angew. Chem. 1983, 95, 808; Angew. Chem., Int. Ed. Engl. 1983,
22, 779. (b) Benn, R.; Rufin˜ska, A. Angew. Chem. 1986, 98, 851; Angew.
Chem., Int. Ed. Engl. 1986, 25, 861. (c) Hinton, J . F.; Briggs, R. W. In
NMR and the Periodic Table; Harris, R. K., Mann, B. E., Eds.;
Academic Press: London, 1978.
(9) Uhl, W.; Breher, F.; Mbonimana, A.; Gauss, J .; Haase, D.; Lu¨tzen,
A.; Saak, W. Eur. J . Inorg. Chem., in press.

5480 Organometallics, Vol. 20, No. 25, 2001
Uhl et al.
knowledge, a similar µ3 bridging has not been observed
before in aluminum chemistry but was found in the
gallium compound Mes6Ga6F4O411 with a similar length-
ening of the element-fluorine distances compared to the
µ2 situation.
The carbaalane 2 underwent a rearrangement reac-
tion upon heating to yield almost quantitatively an
Al7C4 cluster (3).2 The same reaction was not observed
with the fluoro-containing cluster 4, which is stable even
at elevated temperatures. Ethereal solutions of 4 are
sensitive toward irradiation, and exposure to daylight
gave a new product which, to date, could not be isolated
and characterized. A reaction similar to that in eq 1 was
observed upon treatment of the arachno-carbaalane 1
(R ) Me, R′ ) Ph, R′′ ) H) with HBF4‚OEt2. However,
equimolar amounts of the acid gave a 3:1 mixture of a
novel product (5) and the starting compound 1, which
could not be separated by repeated recrystallization. To
completely consume 1, we employed an excess of the
acid. However, a solid precipitated which was insoluble
in organic solvents and which could not be character-
ized. NMR spectra of 5 recorded of the mixture verify
that the bridging hydrogen atom was replaced by a
fluorine atom and that the structure of the cluster
remained unchanged with a cube of eight aluminum
atoms and five faces bridged by CCH2Ph groups. A
resonance at -121.4 ppm was observed in the 19F NMR
spectrum in the expected range for AlF compounds.10
Tr ea tm en t of (AlMe)8(CCH2P h )5H (1) w ith Hy-
d r ogen Ch lor id e. To gain deeper insight into the
reaction of the arachno-carbaalane 1 with protic re-
agents and to isolate a pure compound, we treated 1
with a small excess of an ethereal solution of HCl. Gas
evolution was observed at low temperature (-70 °C, eq
2). After evaporation of the solvent at -20 °C and
F igu r e 1. Molecular structure of 4. The thermal ellipsoids
are drawn at the 40% probability level. Methyl, phenyl,
and benzyl groups are omitted for clarity. Important bond
lengths (Å): Al(1)-Al(2) ) 2.598(1), Al(1)-Al(3) ) 2.552(2),
Al(1)-C(1) ) 2.056(5), Al(1)-C(2) ) 2.152(4), Al(2)-Al(4)
) 2.691(2), Al(2)-Al(5) ) 2.662(2), Al(2)-C(1) ) 2.042(2),
Al(2)-C(2) ) 2.071(4), Al(2)-C(3) ) 2.128(4), Al(3)-F(1)
) 1.966(4), Al(3)-Al(5) ) 2.788(2), Al(3)-C(2) ) 2.019(4),
Al(4)-C(1) ) 2.148(6), Al(4)-C(3) ) 2.045(4), Al(5)-F(1)
) 2.000(1), Al(5)-C(2) ) 2.050(4), Al(5)-C(3) ) 2.009(4),
C(3)-C(300) ) 1.359(5).
tions).10 An organogallium compound, Mes6Ga6F4O4, is
known with a µ3-fluorine ligand and a pseudo-µ3-fluorine
ligand, for which only small upfield shifts were observed
compared to the µ2 situation.11 The nonobservance of a
resonance due to the fluorine atom in 4 may be caused
by the quadrupole moment of aluminum, the coupling
to chemically different aluminum atoms, and an unusu-
ally strong line broadening.
The molecular structure of 4 is depicted in Figure 1.
It contains an Al7C5 cluster, which may be derived from
an Al8 cube if one vertex remains unoccupied. It has
three Al4 rings which are bridged by CCH2C6H5 groups.
At the open face three Al3 triangles result, two of which
are bridged by the alkenylidene ligands and one by the
fluorine atom. The structure is almost identical with
that of the starting compound 2.2 The length of the Cd
C bonds (1.359(5) Å) corresponds to that of 2 and is only
slightly longer than the standard value.12 The Al-C
distances to the terminally bonded ethyl groups are
quite normal (1.95 Å on average). Longer separations
(2.009(4)-2.148(6) Å) are detected in the cluster, similar
to data reported before for other carbaalanes.1-3 Three
ranges of Al-Al distances are observed (2.575, 2.677,
and 2.788 Å on average). The longer ones belong to the
fluorine-bridged Al3 triangle (Al(3)-Al(5)); the shortest
ones are to the aluminum atom Al(1) (Al(1)-Al(2),
Al(1)-Al(3)), which is opposite to the open face and
occupies the most intact part of the cluster. The fluorine
atom bridges the atoms Al(3), Al(5), and Al(5′) with
Al-F distances of 1.966(4) Å (Al(3)) and 2.000(1) Å
(Al(5)). These distances are much longer than those
observed for µ2-Al-F-Al bridges, which usually have
values between 1.78 and 1.85 Å.10 To the best of our
recrystallization of the product 6 from toluene, a color-
less, solid product was isolated which usually had an
impurity of about 10% of the starting compound 1. A
pure sample of 1 was obtained in very small quantity
by recrystallization from a very dilute pentane solution.
(10) Neumu¨ller, B. Coord. Chem. Rev. 1997, 158, 69.
(11) Neumu¨ller, B.; Gahlmann, F. Angew. Chem. 1993, 105, 1770;
Angew. Chem., Int. Ed. Engl. 1993, 32, 1701.
(12) March, J . Advanced Organic Chemistry, 3rd ed.; Wiley: New
York, 1985.

Reactions of Carbaalanes with HBF4 and HCl
Organometallics, Vol. 20, No. 25, 2001 5481
small line width (W1/2 ) 1500 Hz). However, the signal
has a very broad base covering a range of almost 150
ppm, which may indicate the overlap of several reso-
nances.
The molecules of 6 showed a disorder in the solid state
with the chlorine atom distributed over three positions
attached to Al(1), Al(3), and Al(4). The molecule with
the highest occupation factor (0.6) is shown in Figure
2. The structure of the cluster is almost unchanged in
comparison with that of the starting compound 1,1 with
the important restriction that only seven Al atoms are
bonded to methyl groups, while one (Al(1)) is attached
to a terminal chlorine atom. Thus, a methyl group at
the closed part of the cluster is replaced, which may be
caused by the higher electron density at positions
completely enclosed in the electronically delocalized
cluster bonding. The Al-Al distances at the hydrogen-
bridged face of the cluster are very long (2.789 Å on
average) compared to those of the remaining edges of
the Al8 cube (2.594 Å). The open face has a diamond-
shaped distortion with two short (1.89 Å average) and
two long Al-H distances (2.11(5) and 2.32(5) Å). The
Al-C distances in the cluster (2.005(5)-2.129(6) Å) are
longer than those observed to the terminal methyl
groups. The shortest ones involved the aluminum atoms
of the hydrogen-bridged face and the Al-C bonds to the
carbon atom C(50) opposite to the bridging hydrogen
atom. The introduction of the chlorine substituent seems
to have no significant influence on the bond lengths in
the cluster. Both compounds 4 and 6 are excellent
starting materials for further investigations, and salt
elimination reactions may provide access to a great
variety of derivatives.
F igu r e 2. Molecular structure of 6. The thermal ellipsoids
are drawn at the 40% probability level. Benzyl groups are
omitted for clarity. Only one of the disordered positions of
the chlorine atom is shown. Important bond lengths (Å):
Al(1)-Al(2) ) 2.599(2), Al(1)-Al(4) ) 2.588(2), Al(1)-Al(5)
) 2.588(2), Al(1)-C(10) ) 2.061(5), Al(1)-C(40) ) 2.076(5),
Al(1)-C(50) ) 2.046(5), Al(1)-Cl(1) ) 2.096(5), Al(2)-Al(3)
) 2.603(2), Al(2)-Al(6) ) 2.596(2), Al(2)-C(10) ) 2.129(6),
Al(2)-C(20) ) 2.113(6), Al(2)-C(50) ) 2.029(5), Al(3)-Al(4)
) 2.599(2), Al(3)-Al(7) ) 2.597(2), Al(3)-C(20) ) 2.087(5),
Al(3)-C(30) ) 2.120(5), Al(3)-C(50) ) 2.095(6), Al(4)-Al(8)
) 2.580(2), Al(4)-C(30) ) 2.090(5), Al(4)-C(40) ) 2.114(5),
Al(4)-C(50) ) 2.050(5), Al(5)-Al(6) ) 2.800(2), Al(5)-Al(8)
) 2.795(2), Al(5)-C(10) ) 2.020(5), Al(5)-C(40) ) 2.015(5),
Al(6)-Al(7) ) 2.781(2), Al(6)-C(10) ) 2.036(5), Al(6)-C(20)
) 2.016(5), Al(7)-Al(8) ) 2.780(2), Al(7)-C(20) ) 2.005(5),
Al(7)-C(30) ) 2.021(5), Al(8)-C(30) ) 2.037(5), Al(8)-
C(40) ) 2.013(5).
Qu a n tu m -Ch em ica l Ca lcu la tion s. In contrast to
5, the compound (AlEt)7(CdCHC6H5)2(CCH2C6H5)3(µ3-
F) (4) did not give an observable resonance in the 19F
NMR spectrum in the expected range of -100 to -160
ppm. To verify that range for carbaalanes also, we
carried out quantum-mechanical calculations on (AlH)7-
(CdCH2)2(CH)3(µ3-F) (I) and (AlH)7(CdCHC6H5)2(CH)3-
(µ3-F) (III and IV) as model clusters (see Figure 3) of
the real cluster 4 and (AlH)8(CH)5(CH)3F (II) as a model
system of the impure product 5. All calculations em-
ployed B3LYP hybrid density functional13 and standard
double-ú basis sets 6-31(G)d and 6-31+G(d).14 This
theoretical model seems to be a reasonable compromise
between accuracy and computer time in describing the
fundamental properties of carbaalanes.3
Increasing the excess of HCl did not result in complete
consumption of 1. Instead, the concentration of poly-
halogenated byproducts increased. Five resonances in
1
the H NMR spectrum with an intensity ratio of 1:2:2:
1:1 for the methyl groups attached to aluminum clearly
show that the expected substitution of the bridging
hydrogen atom did not occur. Its resonance is still
present at δ 5.04, which is only slightly shifted com-
pared to the corresponding signal of 1 (δ 5.23).1 Thus, a
remarkable and unexpected reaction occurred in which
the proton did not attack the negatively charged hy-
drogen atom at the open face of the cluster; instead, a
methyl group terminally attached to an aluminum atom
at the opposite site of the cluster was replaced with
release of methane. Thus, the chlorine atom does not
occupy a bridging position between two or more alumi-
num atoms but prefers a terminal position. The unex-
pectedly different reaction behavior of HCl and HBF4‚
OEt2 may depend on the strength of these acids and
the different Al-X bond energies, but we need to
undertake more systematic studies to clearly answer
that question. The replacement of one methyl group at
the bottom of the cluster by a chlorine atom is in
complete agreement with the NMR spectroscopic ob-
servation of five chemically different types of methyl
groups. Single groups are attached to the atoms Al3,
Al5, and Al7 (see Figure 2), while pairs of equivalent
groups are at the opposite atoms (Al2, Al4) and (Al6,
Al8), respectively. Despite the complicated molecular
symmetry only a single resonance was observed in the
27Al NMR spectrum (δ 132), which seems to have a
In a first step, we did a full geometry optimization
on structures I and II (see Figure 3) followed by
calculations of the Hessian matrixes to characterize the
nature of the stationary points. The B3LYP/6-31+G(d)
optimized geometry of (AlH)7(CdCH2)2(CH)3(µ3-F) I (Cs
(13) Becke, A. D. J . Chem. Phys. 1993, 98, 5648.
(14) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian 98; Gaussian, Inc.: Pittsburgh, PA, 1998.

5482 Organometallics, Vol. 20, No. 25, 2001
Uhl et al.
resonance of 4 in the characteristic range of Al-F
compounds. The fact that we did not observe a signal
may thus indeed be the consequence of an unusually
strong line broadening.
In a second step, we augmented our model system and
included the phenyl rings on the CdC double bonds in
order to model their electronic effects on the 19F shield-
ing in (AlH)7(CdCHC6H5)2(CH)3(µ3-F) (III). We had to
reduce the diffuse functions to a 6-31G(d) basis because
of limited computing power. Starting with the X-ray
geometry of (AlEt)7(CdCHC6H5)2(CCH2C6H5)3(µ3-F) (4),
we generated our model cluster III (see Figure 3) in Cs
symmetry. A full geometry optimization was done,
ending up with the stationary point IV, which was
characterized as a true minimum (no imaginary fre-
quencies). According to our calculations, the 19F NMR
shift is more or less invariant (cluster III, -155 ppm;
cluster IV, -162 ppm) to the presence or orientation of
the CdCHC6H5 phenyl rings.
Exp er im en ta l Section
All procedures were carried out under purified argon.
n-Pentane and cyclopentane were dried over LiAlH4 and
toluene and diethyl ether over Na/benzophenone. The car-
baalanes (AlMe)8(CCH2C6H5)5H (1) and (AlEt)7(CdCHC6H5)2-
(CCH2C6H5)3H (2) were synthesized according to literature
procedures.1,2 Commercially available solutions of HBF4 (54%)
and HCl (1 M) in diethyl ether (Aldrich) were used without
1
further purification. H and 13C NMR spectra were recorded
on a 500 MHz NMR spectrometer at room temperature. 19F
NMR spectra were recorded on a 300 MHz NMR spectrometer.
27Al NMR spectra were taken on a 400 MHz spectrometer. The
1
resonances were assigned on the basis of H, 13C, 19F, H-H-
COSY, HMQC, and HMBC spectra. Shifts are reported on the
δ scale in ppm relative to residual nondeuterated solvent
signals (1H) or the signals of the deuterated solvent (13C) as
internal standards and relative to CFCl3 as external standard
in case of the 19F NMR experiments or relative to [Al(H2O)6]3+
as external standard for 27Al NMR spectra.
F igu r e 3. Optimized geometry of the model clusters
(AlH)7(CdCHR)2(CH)3(µ3-F) (I, III, and IV; Cs symmetry)
and (AlH)8(CH)5(F) (II; C4v symmetry) at the B3LYP/6-
31+G(d) level of theory.
Syn th esis of th e Flu or ocar baalan e (AlEt)7(CdCHC6H5)2-
(CCH2C6H5)3F (4). (AlEt)7(CdCHPh)2(CCH2Ph)3H (2; 1.501
g, 1.66 mmol) was dissolved in 50 mL of diethyl ether and
cooled to -50 °C. One equivalent (227 µL) of a 54% solution of
tetrafluoroboric acid in diethyl ether was added slowly. Gas
evolution was observed, and a light orange solid precipitated.
After the mixture was warmed to room temperature, the
solvent was distilled off under vacuum. The highly viscous
residue was thoroughly evacuated to completely remove all
volatile impurities. It was treated with cyclopentane to give a
suspension of a light orange solid, which was filtered. That
solid was insoluble in organic solvents and was not character-
ized further. The filtrate was concentrated and cooled to -30
°C. The orange product 4 precipitated, which is very hygro-
scopic and rapidly decomposes in solution in daylight. Yield:
0.722 g (48%). Dec pt (argon, sealed capillary): 160 °C. Anal.
Calcd for C54H68Al7F (925.0): Al, 20.4; C, 70.1; H, 7.4. Found:
Al, 20.1; C, 70.5; H, 7.3. Molar mass (cryoscopically in
symmetry) is a minimum on the potential energy
surface (no imaginary frequencies). The equilibrium
distances (1.968 and 2.131 Å) between the µ3-bridging
F and the Al atoms are in agreement with experimental
data. The B3LYP/6-31+G(d) optimized geometry of
(AlH)8(CH)5(µ4-F) (II) in C4v symmetry as a model
system of the product (AlMe)8(CCH2C6H5)5F (5) shows
three small (<100 cm-1) imaginary frequencies. Releas-
ing the symmetry constraint during the optimization
moves the F atom from a symmetric µ4-bridging position
(Al-F ) 2.188 Å) to a pseudo µ3-bridging position (Al-F
) 2.023, 2.133, 2.827 Å). This leads to a true minimum
in Cs symmetry, which is 8.34 kcal/mol lower in energy
than the C4v symmetric third-order saddle point. Nev-
ertheless, the calculated 19F NMR resonance using the
GIAO method15 is quite invariant (C4v symmetry, -153
ppm; Cs symmetry, -148 ppm). Taking into account the
simplicity of our model system compared to the real
compound (AlMe)8(CCH2C6H5)5F (5) (δ(19F) -121 ppm),
the agreement is quite good. A similar chemical shift of
-160 ppm was calculated for the 19F NMR resonance
in (AlH)7(CdCH2)2(CH)3(µ3-F) (I) as a model compound
of 4. This verifies that we have to expect the 19F
1
benzene): found m/e 907. H NMR (d8-toluene, 500 MHz; δ):
benzyl group at C1, 7.51 (2 H, m, o-H), 7.30 (2 H, m, m-H),
7.15 (1 H, m, p-H), 3.68 (2 H, s, CH2); benzyl groups at C2,
7.38 (4 H, m, o-H), 7.22 (4 H, m, m-H), 7.10 (2 H, m, p-H),
2
4
3.47 (2 H, dd, CH2, J HH ) -13.7 Hz, J HF ) 4.3 Hz), 3.42 (2
2
4
H, dd, CH2, J HH ) -13.7 Hz, J HF ) 9.0 Hz); CdCHPh
groups: 7.18 and 7.06 (m, C6H5, further assignment uncertain),
9.28 (2 H, s, CdCH); ethyl group at Al1*, 1.20 (3 H, t, J HH
3
)
3
8.1 Hz, CH3), 0.12 (2 H, q, J HH ) 8.1 Hz, CH2); ethyl groups
3
at Al2 and Al2′, 1.20 (6 H, dd, J HH ) 8.1 and 8.1 Hz, CH3),
(15) Wolinski, K.; Hilton, J . F.; Pulay, P. J . Am. Chem. Soc. 1990,
112, 8251.
2
3
0.62 and 0.48 (each 2 H, dq, J HH ) -14.5 Hz, J HH ) 8.1 Hz,

Reactions of Carbaalanes with HBF4 and HCl
Organometallics, Vol. 20, No. 25, 2001 5483
3
Ta ble 1. Cr ysta l Da ta , Da ta Collection
P a r a m eter s, a n d Str u ctu r e Refin em en t Deta ils for
Com p ou n d s 4 a n d 6
CH2); ethyl group at Al3*, 1.40 (3 H, t, J HH ) 8.1 Hz, CH3),
3
0.75 (2 H, q, J HH ) 8.1 Hz, CH2); ethyl group at Al4*, 1.32 (3
3
3
H, t, J HH ) 8.1 Hz, CH3), 0.53 (2 H, q, J HH ) 8.1 Hz, CH2);
ethyl groups at Al5 and Al5′, 0.64 (6 H, dd, 3J HH ) 8.1 and 8.1
4
6
2
Hz, CH3), 0.01 and -0.18 (each 2 H, dq or m, J HH ) -14.9
formula
cryst syst
space group
Z
C
54H68Al7F
C52H57Al8Cl
triclinic
3
3
Hz, J HH ) 8.1 Hz, the second resonance shows J HF ) 11.2
Hz, CH2); (the asterisks denote the fact that some uncertainties
exist concerning the individual assignment of the ethyl
groups). 13C NMR (d8-toluene, 125.8 MHz; δ): benzyl group
at C1, 148.8 (i-C), 128.5 (o-C), 129.2 (m-C), 126.3 (p-C), 35.7
(CH2); benzyl groups at C2 and C2′, 148.7 (i-C), 128.5 (o-C),
129.8 (m-C), 126.7 (p-C), 34.3 (CH2); phenyl groups at C300
and C300′, 142.3 (i-C), 130.6, 129.2, 128.6, 128.3 and 125.5
(phenyl, individual assignment not possible), 184.2 (CdCHPh);
ethyl group at Al1*, 10.0 (CH3), 2.1 (CH2); ethyl groups at Al2
and Al2′, 10.2 (CH3), 2.8 (CH2); ethyl group Al3*, 10.5 (CH3),
0.7 (CH2); ethyl group at Al4*, 9.2 (CH3), 2.1 (CH2); ethyl
orthorhombic
Pnma; No. 6216
4
P1h; No. 216
2
temp, K
D
193(2)
1.187
193(2)
1.195
calcd, g/cm3
a, Å
b, Å
c, Å
20.0859(7)
17.1674(7)
15.0162(6)
90
90
90
11.833(2)
13.386(4)
16.716(5)
81.90(2)
82.41(2)
85.48(2)
2593.5(12)
0.242
R, deg
â, deg
γ, deg
V, 10-30 m3
µ, mm-1
cryst dimens, mm
radiation
2θ range, deg
index ranges
5177.9(3)
0.179
2
groups at Al5 and Al5′, 7.6 (CH3), 3.1 (CH2, J CF ) 17 Hz);
0.55 × 0.45 × 0.40
0.55 × 0.40 × 0.25
2
2
32.5 (C1), 39.6 (C2, J CF ) 11 Hz), 156.2 (C3dC, J CF ) 23
Hz) (the asterisks denote uncertain assignments). 27Al NMR
(d8-toluene, 104 MHz; δ): δ 110 (sh), 74 (W1/2 ) 12 000 Hz),
20 (sh), -28 (W1/2 ) 10 000 Hz). IR (CsBr plates, paraffin,
cm-1): 1945 vw, 1873 vw, 1798 vw, 1683 vw, 1598 w (phenyl);
1578 w ν(CdC); 1505 m δ(CH), 1457 vs, 1377 vs (paraffin);
1302 m, 1224 w, 1196 w, 1169 m, 1155 m δ(CH), ν(CC); 1074
w, 1029 w ν(CC); 977 m, 957 br m, 930 sh, m, 907 m, 835 w
(phenyl); 747 s δ(Ph); 721 vs; 705 sh, s, 690 m Al-C and C-C
bending modes; 653 br s, 609 m, 558 m δ(Ph), ν(AlC); 502 w,
474 m, 447 m, 365 br w, 288 w ν(AlC), ν(AlF).
Mo KR; graphite monochromator
4.06-52.00
5.38-49.94
-24 e h e 24
-21 e k e 21
-18 e l e 18
-10 e h e 0
-11 e k e 11
-14 e l e 14
no. of unique rflns
5222 (Rint
0.0825)
306
)
6434 (Rint
0.0465)
566
)
no. of params
R1 (rflns I > 2σ(I))
wR2 (all data)
max/min residual
electron density,
1030 e/m3
0.0678
0.1962
0.893/-0.546
0.0741
0.1984
0.692/-0.601
Syn th esis of th e Im p u r e F lu or oca r ba a la n e (AlMe)8-
(CCH2C6H5)5F (5). The arachno-carbaalane (AlMe)8(CCH2C6-
H5)5H‚0.5(cyclopentane) (1; 0.638 g, 0.72 mmol) was dissolved
in 25 mL of diethyl ether and cooled to -50 °C. At this
temperature 1 partially precipitated. One equivalent (98 µL)
of a 54% solution of tetrafluorboric acid in diethyl ether was
added slowly. Gas evolution occurred, and the precipitate
dissolved upon warming to room temperature. The mixture
was stirred for 4 h. A small quantity of a colorless solid
precipitated, which was filtered. The solvent was removed in
vacuo, and the residue was recrystallized from toluene (+20
to -30 °C). An amorphous solid was obtained which contains
up to 30% of the starting cluster 1. Many attempts at
recrystallization from different solvents gave no enrichment
of 5. The NMR data for 5 were recorded by employing such a
mixture. 1H NMR (C6D6, 500 MHz; δ): four equivalent benzyl
groups, 7.13-6.99 (20 H, m, phenyl groups, no individual
assigment owing to overlap), 3.01 (8 H, d, 4J HF ) 6.4 Hz, CH2);
benzyl group opposite to the fluorine atom, 7.51 (2 H, m), 7.36
(2 H, m), 7.21 (1 H, m), 3.53 (2 H, s, CH2); methyl groups of Al
solved in toluene, and filtered. A colorless, crystalline solid,
6, was obtained upon cooling the solution to -30 °C. The
product contained about 10% of the starting compound 1. We
did not succeed in reducing the concentration of that impurity
by repeated recrystallization from different solvents. Yield:
0.335 g (56%). 1H NMR (C6D6, 500 MHz; δ): phenyl groups
(a), 7.73 (2 H, m, o-H), 7.35 (2 H, m, m-H), 7.19 (1 H, m, p-H);
phenyl groups (b), 7.31 (4 H, m, o-H), 7.13 (4 H, m, m-H), 7.01
(2 H, m, p-H); phenyl groups (c) ,7.09 (4 H, m, o-H), 7.05 (4 H,
m, m-H), 6.97 (2 H, m, p-H), further assignment uncertain;
3.70 (2 H, s, CH2 and C50); remaining CH2 groups, 3.46 and
2.94 (each 2 H, d, 2J HH ) 14.0 Hz), 3.06 and 2.95 (each 2 H, d,
2J HH ) -13.5 Hz); methyl groups at Al3, Al5, and Al7, -0.26,
-0.81, and -1.21 (each 3 H, s); remaining methyl groups,
-0.46 and -0.49 (each 6 H, s); 5.04 (1 H, s, AlH). 13C NMR
(C6D6, 125.8 MHz; δ): δ 148.1, 148.0, 147.70, 147.67, 137.8,
130.3, 130.1, 130.0, 129.5, 129.4, 129.4, 129.3, 128.5, 127.8,
126.8, 126.7, 126.6, 126.3, 125.6 (phenyl, further assignment
uncertain); 34.9 (CH2 at C50), remaining CH2 groups, 34.7 and
33.8; cluster carbon atoms, 34.5 and 29.4, one resonance is
missing; methyl groups, -11.3, -11.8, and -12.3 (Al3, Al5,
Al7), -12.4 and -12.9 (Al2, Al4, Al6, Al8). 27Al NMR (C6D6,
104 MHz; δ): δ 132 (W1/2 ) 1500 Hz). IR (CsBr plates, paraffin,
cm-1): 1599 m, 1582 vw (phenyl); 1491 s (δ(CH3)); 1453 vs,
1377 vs (Nujol); 1301 m, 1190 s, 1154 m, 1075 w, 1029 w (ν-
(CC)); 1003 vw, 961 s, 933 m, 906 m, 823 m (phenyl); 751 vs,
712 vs, 690 vs (δ(Ph), ν(AlC)); 586 m, 552 m, 530 m (ν(AlC));
504 m, 464 m (ν(AlCl)); 450 vw, 429 m (ν(AlC)).
3
atoms attached to fluorine, -0.67 (12 H, d, J HF ) 4.3 Hz);
remaining methyl groups, -0.27 (12 H, s). 13C NMR (C6D6,
125.8 MHz; δ): four equivalent benzyl groups, 148.4 (i-C),
130.2, 128.0, 126.6, 34.9 (CH2); benzyl group opposite to the
fluorine atom, 148.1 (i-C), 129.4, 129.2, 126.2, 35.1 (CH2);
methyl groups of Al atoms attached to fluorine, -11.0 (2J CF
)
12.7 Hz); methyl groups at the remaining Al atoms, -9.1;
cluster carbon atoms, 38.7 (2J CF ) 13.6 Hz, four equivalent C
atoms), 10.0. 19F NMR (C6D6, 282 MHz; δ): -121.4. 27Al NMR
(C6D6, 104 MHz; δ): 62 (W1/2 ) 6500 Hz), -48 (W1/2 ≈ 13 000
Hz).
Cr ysta l Str u ctu r e Deter m in a tion s. Single crystals of
compound 4 were obtained by very slow concentration of a
solution in cyclopentane at room temperature. Extraction of
solid 6 with n-pentane and cooling of the solution to 0 °C gave
single crystals in trace amounts. The crystallographic data of
compounds 4 and 6 were collected with a STOE imaging plate
diffractometer and the four-cycle CAD 4 diffractometer, re-
spectively. Relevant crystal data, data collection parameters,
and results of the structure refinement are summarized in
Table 1. The structures were solved by direct methods and
refined with the program SHELXL-9717 by a full-matrix least-
Syn t h esis of t h e Ch lor oca r b a a la n e (AlMe)7(AlCl)-
(CCH2C6H5)5H (6). 1‚0.2(cyclopentane) (0.594 g, 0.69 mmol)
was dissolved in 30 mL of diethyl ether and cooled to -70 °C.
That solution was treated with 1.2 equiv (0.82 mL) of a 1 M
solution of HCl in diethyl ether. Gas evolution was observed.
The mixture was slowly warmed to -20 °C, and all volatile
components were distilled at this temperature. Higher tem-
peratures gave secondary reactions, yielding a mixture of
products which could not be separated. After evaporation of
all volatiles the product is stable in solution at room temper-
ature. The colorless residue was thoroughly evacuated, dis-

5484 Organometallics, Vol. 20, No. 25, 2001
Uhl et al.
squares method based on F2. Compound 6 showed a disorder
of the chlorine atom with three distinct positions: Al(1)-Cl(1)
(occupation factor 61%), Al(3)-Cl(3) (16%), and Al(4)-Cl(4)
(21%). The corresponding carbon and chlorine positions as well
as the occupation factors were refined with restrictions of bond
lengths. One n-pentane molecule was enclosed per formula
unit of 6.
Ack n ow led gm en t. We are grateful to the Deutsche
Forschungsgemeinschaft and the Fonds der Chemischen
Industrie for generous financial support.
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
coordinates, isotropic and anisotropic displacement param-
eters, and all bond lengths and angles. This material is
available free of charge via the Internet at http://pubs.acs.org.
Further details of the crystal structure determinations are
available from the Cambridge Crystallographic Data Center
on quoting the depository numbers CCDC-158662 (4) and
CCDC-158663 (6).
(16) International Tables for Crystallography, Space Group Sym-
metry; Hahn, T., Ed.; Kluwer Academic: Dordrecht, Boston, London,
1989; Vol. A.
(17) (a) Sheldrick, G. M. SHELXTL PLUS, release 4.1; Siemens
Analytical X-Ray Instruments, Inc., Madison, WI, 1990. (b) Sheldrick,
G. M. SHELXL-93, Program for the Refinement of Structures; Uni-
versita¨t Go¨ttingen, Go¨ttingen, Germany, 1993.
OM010584T