Vanadium calixarene complexes as molecular models for supported vanadia

Article abstract of DOI:10.1016/j.molcata.2006.02.045

Reaction of dimethylated tert-butyl calix[4]arene, H₂A^Me2, with O{double bond, long}VCl₂(thf)₂ unexpectedly led to the isolation of Cl{single bond}V^IVA^Me, 2, whose crystal structure was det

Full text of DOI:10.1016/j.molcata.2006.02.045

Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
Vanadium calixarene complexes as molecular
models for supported vanadia
∗
¨
Elke Hoppe, Christian Limberg , Burkhard Ziemer, Clemens Mugge
Humboldt-Universita¨t zu Berlin, Institut fu¨r Chemie, Brook-Taylor-Str. 2, 12489 Berlin, Germany
Available online 29 March 2006
Abstract
Reaction of dimethylated tert-butyl calix[4]arene, H2AMe2, with O VCl2(thf)2 unexpectedly led to the isolation of Cl VIVAMe, 2, whose crystal
structure was determined. Hence, one of the MeO–C bonds present in H2AMe2 has been cleaved and one equivalent of HCl eliminated leaving
three phenolate and one phenolether function to coordinate at the V centre. 2 smoothly reacts with O2 to give O VVAMe, B, i.e. the chloride
ligand has been replaced by a terminal oxo ligand with concomitant oxidation of the vanadium atom to the oxidation state +5. B crystallises
isostructurally with 2; crystallisation of a mixture of 2 and B even leads to crystals that contain both compounds beside each other. However, when
B was co-crystallised together with H2AMe2, from acetonitrile a different structure with a partial cone conformation was obtained. The mechanism
by which B is formed from 2 is discussed. To obtain information concerning the fate of the chloride ligand in 2, a derivative, Ph3SiO–VIVAMe, 3,
has been synthesised and investigated with respect to its behaviour in the presence of O2: it reacts the same way as 2.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Vanadium; Oxo complexes; Models; Calix arenes; Oxygen
1. Introduction
various working hypotheses assumes that the active and selective
sites contain mononuclear tetrahedrally coordinated V5+ centres
[4] (Chart 1).
Calix[4]aren ligands, A, have established themselves in the
past for the modelling of oxidic surfaces characteristic for transi-
tion metal catalysts or the support belonging to a certain catalyst
[6], and their valency can be adjusted by successive methylation
of the phenolate units. Here, we report the employment of such
ligands for the mimicking of V O sites.
Despite significant research efforts directed toward under-
standing the general features and reaction steps of oxidation
catalysis mechanisms on surfaces, many systems can still be
regarded as poorly defined [1], which may in part be due to the
fact that a lot of difficulties are encountered during characterisa-
tion of reaction intermediates on surfaces [2]. For several years,
we have been dealing with the modelling of surface intermedi-
ates occurring during heterogeneous processes with molecular
compounds as well as with the possibilities and limits of such
an approach [3].
2. Experimental
This contribution deals with reactive moieties and intermedi-
ates playing relevant roles on the surfaces of vanadiumoxide-
based catalysts during the oxygenation/dehydrogenation of
organic substrates. A portfolio of different surface species has
been considered to be involved in one way or the other [4]. A
representative example is the oxidative dehydrogenation (ODH)
of alkanes for which heterogeneous vanadiumoxide catalysts
containing SiO2, Al2O3, TiO2, ZrO2, MgO as well as zeolite
supports proved to represent efficient catalysts [5]. One of the
2.1. General procedures
All manipulations were carried out in a glove-box, or else
by means of Schlenk-type techniques involving the use of a
dry argon atmosphere. The 1H, 13C and 51V NMR spectra
were recorded on a Bruker AV 400 NMR spectrometer (1H
400.13 MHz, 13C 100.63 MHz, 51V 105.25 MHz) with toluene-
d8 or C6D6 as solvent at 20 ◦C. The H NMR spectra were
1
calibrated against the residual proton and natural abundance 13
C
resonances of the deuterated solvent (toluene-d8 δH 2.10 ppm
and C6D6 δH 7.15 ppm), the 51V NMR spectra against VOCl3
as a standard. The ESR spectra were recorded on a ERS 300
∗
Corresponding author.
E-mail address: Christian.limberg@chemie.hu-berlin.de (C. Limberg).
1381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.02.045

E. Hoppe et al. / Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
35
Chart 1.
spectrometer (X-Band (9 GHz) ZWG) on solid and liquid
assays with CH2Cl2 as solvent at 20 ◦C. Microanalyses were
performed on a Leco CHNS-932 elemental analyser. Infrared
(IR) spectra were recorded using samples prepared as KBr
pellets with a Digilab Excalibur FTS 4000 FTIR-spectrometer.
via a syringe. While stirring at r.t. over night the colour of the
solution turned red-brown and small amounts of a brown solid
precipitated. Subsequently, all volatiles were removed, and the
residue was extracted with hexane (5 ml). Evaporation of the
solvent yields B as a brown powder (0.13 g, 89.0%).
The analytical as well as the IR and 1H NMR spectroscopic
data were in agreement with those described in the literature
[9b] before. 51V NMR (toluene-d8, δ [ppm]): −352 (br).
2.2. Materials
Solvents were purified, dried and degassed prior to use. Pure
VOCl2(thf)2 [7] and dimethyl-p-tert-butyl-calix[4]arene [8]
were prepared according to the literature procedure. NaOSiPh3
was prepared using NaH and HOSiPh3 in THF.
2.5. (O-methyl-p-tert-butyl-calix[4]arene)-
triphenylsilyloxy-vanadium(IV) (3)
To a stirred red-brown solution of 2 (0.25 g, 0.335 mmol) in
THF (5 ml) a solution of NaOSiPh3 (0.10 g, 0.335 mmol) in THF
(5 ml) was added. The colour changed rapidly to yellow brown.
After stirring for 1 h at r.t. the solvent was evaporated in vacuum.
Extracting with hexane and cooling of the saturated solution to
4 ◦C yields fine black needles of 3 (0.24 g, 74.1%).
ESR of the solid: g = 1.88328, a = 0.405 GHz, mp
213–215 ◦C. IR (KBr): ν˜ [cm−1] 3047 w, 2958 s, 2823
m, 2871 m, 1428 s, 1304 w, 1262 w, 1203 w, 1187 w, 1117 s,
1005 m, 905 vs b, 740 w, 712 s, 697 s, 622 w, 602 w, 510 s,
458 w. Anal. found: C 76.20, H 7.93, calc. for C63H70O5SiV:
C 76.72, H 7.15%.
2.3. (O-methyl-p-tert-butyl-calix[4]arene)-
chloro-vanadium(IV) (2)
2.3.1. Method A
A solution of VOCl2(thf)2 (0.52 g, 1.85 mmol) in toluene
(15 ml) was added to a solution of dimethyl-p-tert-butyl-
calix[4]arene (1.25 g, 1.85 mmol) in toluene (15 ml) and this
mixture was refluxed for 24 h. From the resulting red-brown
solution all volatiles were removed and the residue was extracted
with hexane (2 × 10 ml). Cooling a hexane solution to −30 ◦C
yields 1.28 g (92.7%) of 2.
2.3.2. Method B
2.6. Crystal structure determinations
A solution of VCl4(thf)2 (0.25 g, 0.68 mmol) in toluene (5 ml)
was added to a solution of dimethyl-p-tert-butyl-calix[4]arene
(0.46 g, 0.68 mmol) in toluene (5 ml) and this mixture was
refluxed for 16 h. From the resulting red-brown solution all
volatiles were removed. The residue was extracted with CH2Cl2
(10 ml) and after evaporation to dryness again with hexane
(10 ml). Removal of the solvent yields complex 2 (0.37 g,
72.6%).
ESR: g = 1.88328, a = 0.203 GHz (solution in CH2Cl2) and
g = 1.86430, a = 0.401 GHz (solid), mp 119 ◦C, anal. found: C
73.14, H 7.92, Cl 5.25. Calc. for C45H55O4VCl: C 72.42, H 7.43,
Cl 4.75%.
Single crystals of 2 were obtained by cooling a saturated
pentane solution to −30 ◦C or else by slow evaporation of the
solvent from a saturated solution of 2 in toluene in a glovebox.
Single crystals of 3 were obtained by slowly cooling a saturated
hexane solution to 4 ◦C. Suitable crystals of B in the paco con-
formation were obtained by evaporation of the solvent from a
saturated acetonitrile solution. The crystals were mounted on a
glass fiber and then transferred into the cold nitrogen gas stream
of the diffractometer (Stoe IPDS for 2 and 3, Stoe IPDS2T for B)
˚
using Mo K␣ radiation, λ = 0.71073 A, and the structures were
solved by direct methods (SHELXS-97) [10], refined versus F2
(SHELXL-97) [11] with anisotropic temperature factors for all
non-hydrogen atoms (Table 1). All hydrogen atoms were added
geometrically and refined by using a riding model.
2.4. (O-methyl-p-tert-butyl-calix[4]arene)-
oxo-vanadium(V) (B)
The crystallographic data (apart from structure factors) of 2,
3 and B were deposited at the Cambridge Crystallographic Data
Centre as supplementary publication no. CCDC 282429, 282427
To a stirred black brown solution of 2 (0.15 g, 0.20 mmol) in
acetonitrile (5 ml) a 1000-fold excess of dry oxygen was added

36
E. Hoppe et al. / Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
Table 1
Crystal data and experimental parameters for the crystal structure analyses of 2, 3 and B
2
3
B
Formula
Weight (g. mol−1
Temperature (K)
C45H55ClO4V·1.5C7H8
884.48
180(2)
C63H70O5SiV
986.254
180(2)
C45H55O5V·C46H60O4·3C2H3N
)
1526.93
100(2)
˚
Wavelength (A)
0.71073
Monoclinic
I2/a
0.71073
Orthorombic
Pn21a
0.71073
Triclinic
P1
Crystal system
¯
Space group
˚
a (A)
24.976(3)
13.2350(10)
29.927(3)
90
94.91(2)
90
9856.3(17)
8
1.192
21.556(4)
23.200(3)
25.196(3)
90
90
90
12600(3)
8
1.04
0.216
4136
11.9765(13)
18.105(2)
20.653(3)
94.147(9)
91.838(9)
100.667(9)
4384.7(8)
2
˚
b (A)
˚
c (A)
α (◦)
β (◦)
γ (◦)
3
˚
V (A )
Z
Density (g cm−3
µ(Mo K␣) (mm−1
F(0 0 0)
)
1.157
0.171
1644
)
0.299
3776
GoF
0.787
1.026
1.075
Rind [I > 2σ(I)]
R1 = 0.0492
wR2 = 0.0868
R1 = 0.1196
wR2 = 0.1029
−0.341/0.273
R1 = 0.0634
wR2 = 0.1723
R1 = 0.0828
wR2 = 0.1789
−0.449/1.094
R1 = 0.1244
wR2 = 0.1856
R1 = 0.2432
wR2 = 0.2272
−0.486/0.570
Rind (all data)
−3
˚
ꢀρmin/ꢀρmax (eA
)
H4A, H2AMe2, in toluene at 111 ◦C for 24 h in order to syn-
thesise a complex 1 as depicted in Scheme 1. A pure product
was obtained in good yield, whose spectroscopic characterisa-
tion, however, showed that the expected product 1 had not been
formed. Finally, an X-ray crystal structure analysis revealed the
constitution 2 for the product, a compound that had been previ-
ouslysuggestedby Florianiandco-workers fortheproductofthe
a reaction between the complex Cl–VIIIAMe2and iodine on the
basis of an elemental analysis and decomposition studies [9a].
The crystal structure of complex 2 is depicted in Fig. 1. Each
molecule hosts a toluene molecule in the cavity and additional
0.5 equivalents of co-crystallised toluene are present within the
unit cell. The coordination polyhedron around the vanadium
centre resembles a distorted trigonal bipyramid (τ = 0.6) [12],
with O2, O4 and Cl atoms defining the equatorial plane and the
O1 and O3 atoms in the axial positions. The distortions from a
and 282428. Copies of the data can be ordered free of charge
on application to CCDC, 12 Union Road, Cambridge CB2 1EZ
(fax: +44 1223 336 033; e-mail: deposit@ccdc.cam.ac.uk).
3. Results and discussion
With regard to calixarene based vanadium oxo complexes,
only one precedent complex has been reported containing a
ligand derived from A by monomethylation of one phenolic
function, AMe, and an [O V]3+ unit [9b] (see B in Chart). Nat-
urally, there is also interest in reduced complexes that could
model reduced species formed after oxidation reactions, and in
this context VIV and VIII are important. We therefore decided to
investigate the chemistry of VIV on an oxo surface modelled by
calixarene ligands in order to study possible reoxidation reac-
tions. Accordingly, we reacted O VCl2(thf)2 with dimethylated
Scheme 1.

E. Hoppe et al. / Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
37
Table 2
Dihedral angles (◦) between planar moieties (E (reference plane) refers to
the least-squares mean plane defined by the C7, C14, C21 and C28 bridging
methylenic carbons)
2
3
B
E A
E B
E C
E D
A C
B D
113.18(6)
135.99(8)
93.90(2)
133.50(7)
130.94(5)
91.27(6)
111.1(3)
114.5(2)
123.7(2)
115.0(1)
170.2(2)
130.2(2)
114.0(2)
132.62(12)
119.74(11)
146.38(11)
132.24(13)
99.03(14)
˚
Contact distances (A) between para-carbon atoms of opposite aromatic rings
C4–C17
C10–C24
7.650(1)
9.039(6)
8.192(9)
9.427(9)
7.602(10)
9.609(9)
Displacement of vanadium to the O3-atom in a trigonale bipyramide
0.2256(7) 0.2384(9) 0.2211(13)
∆
in a flame-sealed NMR tube. After the reaction no resonance
could be detected for MeOH or MeCl. There was only one new
signal with the expected integral: a singlet at 2.2 ppm, which
we assign to heptadeuterated o-xylene that could have formed
via Lewis acid catalysed methylation of d8-toluene solvent with
MeCl.
Fig. 1. Molecular structure of complex 2. Solvent molecules and all hydro-
gen atoms were omitted for clarity. Selected bond distances (A) and angles
(◦): V–Cl 2.2327(11), V–O1 2.324(2), V–O2 1.759(3), V–O3 1.825(2),
V–O4 1.767(3), Cl–V–O1 91.44(6), Cl–V–O2 109.74(8), Cl–V–O3 96.24(7),
Cl–V–O4 109.67(8), O1–V–O3 172.32(9), O2–V–O4 136.11(13), V–O1–C1
117.5(2), V–O2–C13 156.1(2), V–O3–C20 115.6(2), V–O4–C27 152.8(2).
˚
During the characterisation of 2 we noticed, that it is sensitive
to both water and O2. Hydrolysis leads to H3AMe as well as
unidentified vanadium oxo halides. The reaction with oxygen
astonishingly leads to B, which crystallises isotypically with 2
[9b]. Hence, if oxygen is not rigorously excluded during the
synthesis of 2 and the product finally purified by crystallisation,
the crystals contain cocrystallised 2 and B. This leads to peculiar
effects: subsequent to an experiment during which some O2 had
been present, the crystals obtained were selected and their cell
determined, in order to make sure that it is identical with the cell
found for 2. Having confirmed that, the crystals were dissolved
in d8-toluene. A 1H NMR spectrum then showed a sharp set of
signals belonging to B, and a 51V NMR spectrum displayed a
resonance at −352 ppm. However, at the same time the solution
can be investigated by ESR revealing a signal at g = 1.88328
characteristic for an uncoupled d1-system (2). Accordingly, the
crystals contain mainly 2 but also some B and the only evidence
for that in the crystal structure would be an apparent shortening
of the V Cl bond with rising concentration of B.
As shown by Radius [14] by X-ray crystallography and NMR
investigation d0 calix[4]arene complexes (e.g. those of Mo and
Ti) can undergo isomerisation reactions: the cone conforma-
tion can change into the partial cone conformation (“paco”).
DF calculations gave evidence that this phenomenon is mainly
driven by the co-ligands at the metal centers (strongly directing
co-ligands favour the “paco” conformation). We found evi-
dence for this kind of isomerism also in the case of B: when
2 was synthesised the way depicted in Scheme 1 (that left some
H2AMe2 unreacted) and treated straight away with O2 without
further purification, the crystals obtained from such a reac-
tion mixture in acetonitrile contained B in a paco-conformation
(Fig. 2) together with co-crystallised H2AMe2 and three ace-
tonitrile molecules, two of which are hosted in the calixarene
regular trigonal bipyramid are indicated by the O1–V–O3 angle
(172.32(1)◦) and by the displacement of the metal atom from
˚
the equatorial plane toward the O3 atom (0.2256(7) A). The
V O2, V O3 and V O4 bond distances (1.758(3), 1.824(2),
◦
˚
1.767(2) A) as well as the wide angles V–O2–C13 (156.1(2) )
and V–O4–C27 (152.8(2)◦) are in agreement with the existence
of V O ␲ bonding. However, the V–O3–C20 angle (115.6(2)◦)
is particularly acute [13], and this value is close to the V–O1–C1
angle (117.5(2)◦) defined by the coordinated methoxy group. As
a consequence of the presence of the methoxy group (V–O1:
˚
2.324(2) A) the macrocycle adopts an elliptical cross section
conformation with the opposite A and C aryl rings (C con-
tains the methoxy group) pushed outward and the opposite B
and D rings inward to the cavity, as indicated by the dihedral
angles they form with the “reference” plane and by the distances
between opposite para-carbon atoms (Table 2). Within the cav-
ity the methyl group of guest molecule toluene points towards
the vanadium centre.
How was 2 formed? Obviously, instead of the expected two
equivalents of HCl only one equivalent was eliminated. Further-
more, the oxo ligand and also a methyl group of the original lig-
and AMe2 is missing. Thus, formally MeOH has been eliminated
(probably as a result of Lewis acid catalysed ether cleavage) in
addition to HCl; MeOH can of course react with HCl under elim-
ination of water to give MeCl. In order to clarify the fate of the
methyl group GC/MS measurements were performed both with
the gas phase and the liquid phase after a corresponding reac-
tion in toluene. However, no additional organic product could be
detected this way. Correspondingly, a reaction was carried out

38
E. Hoppe et al. / Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
cavity of B; the residual one is found in the cavity of H2AMe2
.
Apparently, in case of this d0–V O complex both conforma-
tions are very close in energy and the environment from which
thecompoundiscrystallised(solvent, co-crystallisedmolecules)
determine which conformation is adopted in the crystal. It was
not possible to freeze this process out on the NMR time scale.
While the cone structure of B [9b] is – as mentioned – very
similar to the structure discussed above for 2 displaying quite
different V O bonds, in the paco structure all aryloxidic V–O
bonds are comparable (O2–V 1.793(5), O3–V 1.811(6), O4–V
˚
1.825(5) A), and in this conformation the methoxy group can
˚
approach the V centre more closely (O1–V 2.244(6) A). The
coordination sphere can still be described as trigonal bipyrami-
˚
dal (τ = 0.78) [12], and the V O bond length (1.598(5) A) is sim-
˚
ilar to the one reported for the cone conformation (1.585(8) A)
[9b].
As we were interested in the conversion of 2 to B via molec-
ular oxygen, the reaction was investigated in another NMR tube
experiment. A solution of 2 in d8-toluene was treated with a 50-
fold excess of dry oxygen, and the reaction was followed with
the aid of 1H NMR spectroscopy. Directly after the addition of
dioxygen the characteristic signal set for B could be observed,
and its intensity continuously rose at the expense of the rather
broad signals belonging to paramagnetic 2. The reaction took 2
daystoreachcompletion(itisofcoursemuchfasterinaflaskthat
can be stirred), and during this time a diamagnetic reaction inter-
mediate containing ligand AMe became apparent that could not
be detected anymore at the end of the reaction. How is B formed
Fig. 2. Molecular structure of complex B. Uncoordinated ligand, solvent
molecules and all hydrogen atoms were omitted for clarity. Selected bond dis-
◦
˚
tances (A) and angles ( ): V–O1 2.244(6), V–O2 1.793(5), V–O3 1.811(6),
V–O4 1.825(5), V–O5 1.598(5), O5–V–O1 86.3(2), O5–V–O2 121.3(2),
O5–V–O3 99.5(2), O5–V–O4 109.2(2), O1–V–O3 171.6(2), O2–V–O4
124.6(2), V–O1–C1 143.6(5), V–O2–C13 119.1(4), V–O3–C20 117.9(4),
V–O4–C27 130.2(4).
Scheme 2.

E. Hoppe et al. / Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
39
from 2 and dioxygen? Most likely the initiating step consists of
an end-on addition of O2 to the Lewis acidic VIV centre to form a
VIVCl(␩1–O O)unit. Subsequently, anelectroncanbeexpected
to flow from the vanadium centre to the oxygen ligand so that
a vanadium(V)superoxide moiety is produced (Scheme 2). The
next step should represent a reaction with a second molecule of
1 so that a peroxide unit is formed. In order to give 2 the O O
bond then has to cleave homolytically, but since the V-centres are
already in their highest oxidation states, a V O bond can only
form if an additional electron is provided. As, moreover, at some
stage the chloro atom has to leave the molecule, one possibility
that would solve both of these issues is the elimination of Cl radi-
calsthatsubsequentlyabstractHatomsfromthesolvent. Inorder
to test for this, we first of all repeated the reaction in the presence
of PMe3 (a chloro radical would abstract an H atom anywhere
to yield HCl which could be trapped by PMe3, and the resulting
phosphonium cation could easily be identified via 31P NMR).
However, the whole system behaved differently in the presence
of PMe3 and we were not able to draw any conclusions from that
approach (O PMe3 formed but the active species might not have
been B).
Of course, there is the possibility that not Cl atoms but chlo-
ride anions are eliminated in the last step of Scheme 2, but
then electrons would have to be provided externally. Assumed,
one calixarene molecule decomposes, this could bare reducing
equivalents for many V centres at the same time, so that the par-
tial decomposition of calixarene molecules could account for all
electrons the reaction in Scheme 2 needs. After the reaction of
2 with O2 in acetonitrile and extraction of B with hexane, the
residue contains a white solid. This solid tested positively for
chloride and it also contained organic compounds that derive
Fig. 3. Molecular structure of complex 3. The second independent molecule
and all hydrogen atoms were omitted for clarity. Selected bond distances
◦
˚
(A) and angles ( ): V–O1 2.325(4), V–O2 1.756(4), V–O3 1.844(4), V–O4
1.790(4), V–O5 1.807(3), Si–O5 1.639(4), Si–C46 1.897(5), Si–C52 1.875(5),
Si–C58 1.873(5), O5–V–O1 86.9(1), O5–V–O2 111.2(2), O5–V–O3 100.5(2),
O5–V–O4 111.2(2), O1–V–O3 172.3(2), O2–V–O4 132.1(2), V–O1–C1
114.5(3), V–O2–C13 145.8(3), V–O3–C20 115.9(3), V–O4–C27 168.8(4),
V–O5–Si 146.6(2), C46–Si–C52 106.8(2), C52–Si–C58 110.6(2), C46–Si–C58
112.3(2), O5–Si–C46 109.8(2), O5–Si–C52 110.0(2), O5–Si–C58 107.3(2).
1
from the calixarene moiety as evidenced by H NMR spec-
troscopy. However, we were not able to identify its detailed
constitution.
A further option has to be mentioned: if the dinuclear species
shown in Scheme 2 were capable of transferring chloro atoms
to unreacted molecules of 2, the mechanism shown in Scheme 2
could apply without the generation of free Cl radicals: instead,
Cl2VVAMe molecules would result, whose hydrolysis would
yield further equivalents of B and HCl. As stated above all
reactions were performed under rigorously dry conditions so
that there should not have been any scope for hydrolysis, but
as it is difficult to be absolutely sure, we decided to test this
hypothesis. We synthesised complex 3 via reaction of 2 with
NaOSiPh3 in thf, and crystals were grown from hexane. The
crystal structure is shown in Fig. 3. The arrangement of the lig-
and donor atoms around the V centre is very similar to the one
of 2.
The unit cell contains two independent molecules of 3 (nei-
ther of which hosts a solvent molecule) with similar bond dis-
tances and angles. For the discussion, the data of one molecule
were selected. The V–O5–Si angle (146.6(2)◦) is in a normal
range for a V–O–SiPh3 group [15]. The coordination sphere
around the silicon atom is almost perfectly tetrahedral with
bond angles O5–Si–C1ꢀ (109.8(2)◦), O5–Si–C2ꢀ (110.0(2)◦)
and O5–Si–C3ꢀ (107.3(2)◦) as well as C1ꢀ–Si–C2ꢀ (106.8(2)◦),
C1ꢀ–Si–C3ꢀ (112.3(2)◦) and C2ꢀ–Si–C3ꢀ (110.6(2)◦).
The reaction of 3 with O2 was investigated again in an
NMR tube experiment. As in the case of 2, the characteristic
signals for B appeared, whose integral rose until the reaction
had reached completion. As in contrast to two Cl ligands two
Ph3SiO ligands would hardly find sufficient space at a VAMe
complex metal fragment, the last mechanistic suggestion men-
tioned above (requiring the presence of water) can be excluded.
Moreover, – considering that the formation of Ph3SiO• radicals
is even more unlikely than the formation of Cl• (for compari-
son, the bond dissociation energy of H3SiO–H is significantly
higher than the one of HCl [16]) – the mechanism involving the
elimination of radicals does not seem feasible. We thus prefer
the suggestion that Cl as well as Ph3SiO leave the corresponding
peroxo complexes as anionic species supported by an external
electron source.
4. Conclusions
In conclusion, we have synthesised the calixarene vana-
dium(IV) chloro complex, 2, in a reaction that was meant to
produce a vanadium(IV)oxo complex. Apparently, the starting

40
E. Hoppe et al. / Journal of Molecular Catalysis A: Chemical 251 (2006) 34–40
material O VCl2(thf)2 prefers to react with the potential lig-
and H2AMe2 via cleavage of a aryl–O–Me unit and retains a
chloro rather than an oxo ligand. The chloro complex reacts
with dioxygen to give the vanadium(V)oxo complex B, which is
conformationally flexible and crystallises in the cone or partial
cone conformation. The first step leading to the formation of B
certainly represents the binding of O2 to the VIV centre. The
O–O bond cleavage probably involves a second molecule of 2.
The subsequent steps remain unclear; they result in chloride and
organic decomposition products. This type of reactivity is not
restricted to 2: replacement of chloride by triphenylsiloxide (3)
leads to the same results.
C. Borgmann, C. Limberg, E. Kaifev, H. Pritzkow, L. Zsolnai, J.
Organomet. Chem. 580 (1999) 214;
S. Roggan, C. Limberg, B. Ziemer, Angew. Chem. Int. Ed. 44 (2005)
5259–5262;
S. Roggan, C. Limberg, B. Ziemer, M. Brandt, J. Organomet. Chem.
690 (2005) 5282–5289.
[4] L. Owens, H.H. Kung, J. Catal. 144 (1993) 202;
O.R. Evans, A.T. Bell, T.D. Tilley, J. Catal. 226 (2004) 292.
[5] (a) A. Comite, A. Sorrentino, G. Capannelli, M. Di Serio, R. Tesser, E.
Santacesaria, J. Mol. Catal. A 198 (2003) 151;
(b) R. Zhou, Y. Cao, S. Yan, J. Deng, Y. Liao, B. Hong, Catal. Lett. 75
(2001) 107;
(c) T. Blasco, A. Galli, J.M. Lopez Nieto, F. Trifido, J. Catal. 169 (1997)
203;
(d) H. Arakawa, M. Aresta, J.N. Armor, M.A. Barteau, E.J. Beck-
man, A.T. Bell, J.E. Bercaw, C. Creutz, E. Dinjus, D.A. Dixon, K.
Domen, D.L. Dubois, J. Eckert, E. Fujita, D.H. Gibson, W.A. God-
dard, D.W. Goodman, J. Keller, G.J. Kubas, H.H. Kung, J.E. Lyons,
L.E. Manzer, T.J. Marks, K. Morokuma, K.N. Nicholas, R. Periana, L.
Que, J. Rostrup-Nielson, W.M.H. Sachtler, L.D. Schmidt, A. Sen, G.A.
Somorjai, P.C. Stair, B.R. Stults, W. Tumas, Chem. Rev. 101 (2001)
953.
Acknowledgements
We are grateful to the Deutsche Forschungsgemeinschaft, the
SFB 546, the Fonds der Chemischen Industrie, the BMBF and
the Dr. Otto Ro¨hm Geda¨chtnisstiftung for financial support. We
also would like to thank C. Jankowski for the preparation of
[6] U. Radius, Z. Anorg. Allg. Chem. 630 (2004) 957;
C. Floriani, Chem. Eur. J. 5 (1995) 19.
¨
starting materials as well as R. Stoßer and A. Zehl for ESR
[7] R.J. Kern, J. Inorg. Nucl. Chem. 24 (1962) 1105–1109.
[8] C.D. Gutsche, B. Dhawan, J.A. Levine, K.H. No, L.J. Bauer, Tetrahedron
39 (1983) 409–426.
[9] (a) B. Castellano, E. Solari, C. Floriani, Organometallics 17 (1998)
2328–2336;
measurements.
References
[1] J.B. Reitz, E.I. Solomon, J. Am. Chem. Soc. 120 (1998) 11467.
[2] M. Hunger, J. Weitkamp, Angew. Chem. 113 (2001) 3040;
M. Hunger, J. Weitkamp, Angew. Chem. Int. Ed. Engl. 40 (2001)
2954.
(b) B. Castellano, E. Solari, C. Floriani, R. Scopelleti, Inorg. Chem. 38
(1999) 3406–3413.
[10] SHELXS-97, G.M. Sheldrick, Program for Crystal Structure Solution,
University of Go¨ttingen, 1997.
[3] C. Limberg, Angew. Chem. 115 (2003) 6112;
C. Limberg, Angew. Chem. Int. Ed. 42 (2003) 5932;
M. Hunger, C. Limberg, P. Kircher, Angew. Chem. 111 (1999)
1171;
[11] SHELXL-97, G.M. Sheldrick, Program for Crystal Structure Refinement,
University of Go¨ttingen, 1997.
[12] A.W. Addison, T.N. Rao, J. Reedijk, J. van Rijn, G.C. Verschoor, J.
Chem. Soc. Dalton Trans. (1984) 1349–1956.
M. Hunger, C. Limberg, P. Kircher, Angew. Chem. Int. Ed. Engl. 38
(1999) 1105;
M. Hunger, C. Limberg, P. Kircher, Organometallics 19 (2000) 1044;
C. Borgmann, C. Limberg, S. Cunskis, P. Kircher, Eur. J. Inorg. Chem.
(2001) 349–352;
M. Hunger, C. Limberg, E. Kaifer, P. Rutsch, J. Organomet. Chem. 641
(2002) 9–14;
C. Limberg, M. Hunger, W. Habicht, E. Kaifer, Inorg. Chem. 41 (2002)
3359–3365;
[13] L.R. Chamberlain, L.D. Durfee, P.E. Fanwick, L.M. Kobriger, S.L.
Latesky, A.K. McMullen, B.D. Steffey, I.P. Rothwell, K. Folting, J.C.
Huffman, J. Am. Chem. Soc. 109 (1987) 6068–6076.
[14] U. Radius, Inorg. Chem. 40 (2001) 6637–6642.
[15] M. Rost, H. Go¨rls, W. Imhof, W. Seidel, K. Thiel, Z. Anorg. Allg.
Chem. 624 (1998) 1994–2000.
[16] H3SiOH (122.7 kcal mol−1):. M.D. Allendorf, C.F. Melius, P. Ho, M.R.
Zachariah, J. Phys. Chem. 99 (1995) 15285–15293;
HCl (103.1 kcal mol−1):. K.P. Huber, G. Herzberg, Molecular Spectra
and Molecular Structure Constants of Diatomic Molecules, Van Nos-
trand, New York, 1979.
S. Roggan, C. Limberg, B. Ziemer, Angew. Chem. 117 (2005)
5393–5397;