Mono- and dinuclear oxovanadium(V)calixarene complexes and their activity as oxidation catalysts

Article abstract of DOI:10.1021/ic061106j

The background of the investigation is constituted by reactive moieties and intermediates playing relevant roles on the surfaces of vanadiumoxide-based catalysts during the oxygenation/dehydrogenation of organic substrates. With the aim of modeling such species, a series of mono- and dinuclear charged and uncharged vanadium oxo complexes containing p-tert-butylated calix[4]arene and calix[8]arene ligands (denoted H₄B and H₈B″, respectively, in the protonated forms) has been synthesized and characterized: PPh₄[O=VB] (^PPh41), O=VB^OAc (2), PPh ₄[O₂V₂HB″] (3), and [μ-O(O=V(OMe)) ₂B^Me2] (4), where superscripts OAc and Me2 indicate that one or two protons of H₄B are substituted by these residues, respectively. These compounds were analyzed both in solution and by means of single-crystal X-ray crystallography; it turned out that the crystal structures are retained on dissolution (2 changed only from the paco to the cone structure). In the case of 4, it could be shown that the bulk product consists of a mixture of two isomers (4^t and 4^c) differing in the relative positions of the vanadium-bound methoxy groups. Subsequently, all compounds were tested as catalysts for the oxidation of alcohols with O ₂. It turned out that the two dinuclear complexes efficiently catalyze the oxidation of 1-phenyl-1-propargyl alcohol and fluorenol; in addition, they even show some activity with respect to the oxidation of dihydroanthracene. This may hint to a higher activity of dinuclear sites on the surfaces of heterogeneous catalysts as well.

Full text of DOI:10.1021/ic061106j

Inorg. Chem. 2006, 45, 8308−8317
Mono- and Dinuclear Oxovanadium(V)calixarene Complexes and Their
Activity as Oxidation Catalysts
Elke Hoppe, Christian Limberg,* and Burkhard Ziemer
Humboldt-UniVersita¨t zu Berlin, Institut fu¨r Chemie, Brook-Taylor-Strasse 2,
12489 Berlin, Germany
Received June 20, 2006
The background of the investigation is constituted by reactive moieties and intermediates playing relevant roles on
the surfaces of vanadiumoxide-based catalysts during the oxygenation/dehydrogenation of organic substrates. With
the aim of modeling such species, a series of mono- and dinuclear charged and uncharged vanadium oxo complexes
containing p-tert-butylated calix[4]arene and calix[8]arene ligands (denoted H4B and H8B′′, respectively, in the
protonated forms) has been synthesized and characterized: PPh4[O
(3), and [ -O(O
V(OMe))2BMe2] (4), where superscripts OAc and Me2 indicate that one or two protons of H4B are
d d ]
VB] (PPh41), O VBOAc (2), PPh4[O2V2HB′′
µ
d
substituted by these residues, respectively. These compounds were analyzed both in solution and by means of
single-crystal X-ray crystallography; it turned out that the crystal structures are retained on dissolution (2 changed
only from the paco to the cone structure). In the case of 4, it could be shown that the bulk product consists of a
mixture of two isomers (4t and 4c) differing in the relative positions of the vanadium-bound methoxy groups.
Subsequently, all compounds were tested as catalysts for the oxidation of alcohols with O2. It turned out that the
two dinuclear complexes efficiently catalyze the oxidation of 1-phenyl-1-propargyl alcohol and fluorenol; in addition,
they even show some activity with respect to the oxidation of dihydroanthracene. This may hint to a higher activity
of dinuclear sites on the surfaces of heterogeneous catalysts as well.
Introduction
with the possibilities and limitations of such an approach.3
This contribution deals with reactive moieties playing
relevant roles on the surfaces of vanadiumoxide-based
catalysts during the oxidation of organic substrates. Various
different surface species have been discussed as being active
in one way or the other,4 and a representative example is
the oxidative dehydrogenation (ODH) of methanol and
Despite significant research efforts directed toward un-
derstanding the general features and reaction steps of
oxidation catalysis mechanisms on surfaces, many systems
can still be regarded as being poorly defined,1 which may
in part be due to the fact that a lot of difficulties are
encountered during characterization of sites and reaction
intermediates on surfaces.2 Heterogeneous catalysts for
oxidation reactions thus sometimes have to be employed as
“black boxes”: Reactants are injected and products are
formed in good yields with good selectivities, but a complete
understanding of the origin of the catalytic behavior is not
yet available.
(3) Limberg, C. Angew. Chem. 2003, 115, 6112; Angew. Chem., Int. Ed.
2003, 42, 5932. Hunger, M.; Limberg, C.; Kircher, P. Angew. Chem.
1999, 111, 1171; Angew. Chem., Int. Ed. 1999, 38, 1105. Hunger,
M.; Limberg, C.; Kircher, P. Organometallics 2000, 19, 1044;
Borgmann, C.; Limberg, C.; Cunskis, S.; Kircher, P. Eur. J. Inorg.
Chem. 2001, 349; Hunger, M.; Limberg, C.; Kaifer, E.; Rutsch, P. J.
Organomet. Chem. 2002, 641, 9. Limberg, C.; Hunger, M.; Habicht,
W.; Kaifer, E. Inorg. Chem. 2002, 41, 3359. Roggan, S.; Limberg,
C.; Ziemer, B. Angew. Chem. 2005, 117, 5393; Angew. Chem., Int.
Ed. 2005, 44, 5259. Roggan, S.; Limberg, C.; Ziemer, B.; Brandt, M.
J. Organomet. Chem. 2005, 690, 5282. Siewert, I.; Limberg, C.;
Ziemer, B. Z. Anorg. Allg. Chem. 2006, 632, 1078. Wippert-Rodrigues,
C.; Limberg, C.; Pritzkow, H. Chem. Commun. 2004, 2734. Wippert-
Rodrigues, C.; Limberg, C.; Pritzkow, H. Eur. J. Inorg. Chem. 2004,
3644. Wippert Rodrigues, C.; Antelmann, B.; Limberg, C.; Pritzkow,
H. Organometallics 2001, 20, 1825. Borgmann, C.; Limberg, C.;
Kaifer, E.; Pritzkow, H.; Zsolnai, L. J. Organomet. Chem. 1999, 580,
214. Borgmann, C.; Limberg, C.; Zsolnai, L. Chem. Commun. 1998,
2729.
For several years, we have been dealing with the modeling
of surface intermediates occurring during heterogeneous
oxidation processes with molecular compounds as well as
* To whom correspondence should be addressed. E-mail:
Christian.limberg@chemie.hu-berlin.de.
(1) Reitz, J. B.; Solomon, E. I. J. Am. Chem. Soc. 1998, 120, 11467.
(2) Hunger, M.; Weitkamp, J. Angew. Chem. 2001, 113, 3040; Angew.
Chem., Int. Ed. 2001, 40, 2954.
8308 Inorganic Chemistry, Vol. 45, No. 20, 2006
10.1021/ic061106j CCC: $33.50
© 2006 American Chemical Society
Published on Web 08/29/2006

OxoWanadium(V)calixarenes as Oxidation Catalysts
Chart 1
Scheme 1
alkanes for which vanadia on SiO2, Al2O3, ZrO2, and other
oxidic supports proved to be catalytically efficient.5 Some
of the various working hypotheses are the following: the
active and selective sites consist of mononuclear tetrahedrally
coordinated V5+ (a in Chart 1) whereas dinuclear assemblies
(like b in Chart 1) and polynuclear moieties are responsible
for the loss in selectivity (oxygenation). On the other hand,
species with a square pyramidal oxo coordination geometry
around the vanadium center (c in Chart 1) have been
proposed to initiate the oxidation of butane to MSA on
vanadyl phosphate catalysts,6 and such units have been shown
to form if vanadia is deposited on certain metal surfaces,
too.7 We were interested in using these structural motifs as
an inspiration for attempts to mimic them in molecular
compounds. Two different ligand systems have established
themselves in the past for the modeling of oxidic surfaces
characteristic for transition-metal oxide catalysts or the
support belonging to a certain catalyst: the deprotonated
forms of silsesquioxanes8 (H3AR) and p-tert-butylated calix-
arenes9 (H4B, H6B′, H8B′′) (see Scheme 1).
After deprotonation, H3AR represents a tripodal ligand,
whereas H4B yields a tetrapodal ligand. Employing H3AR
and H4B thus enables empathizing the embedding of a
transition-metal ion in the midst of three or four oxo atoms
of an oxidic surface (for instance, a support), respectively.
On the other hand, ligands such as [B′]6- or a calix[8]arene
with eight phenolate units ([B′′]8-) provide more extended
platforms of oxo atoms, where polynuclear moieties can
arrange themselves, too.
Chart 2
so that it could not be isolated as such.10 Vanadium oxo
complexes of calix[4]arenes (B4-) are not known. There is
one example of a complex with a ligand derived from H4B
by monomethylation of a phenolic function, OdVVBMe (I,
see Chart 2)11 and a few vanadium oxo compounds contain-
ing calix[3]-, calix[6]-, and calix[8]arene ligands are known,
too, which, however, either are not homoleptic (i.e., the
vanadium centers are wearing ligands with donor atoms other
than O) or were not characterized structurally.12
Here, we describe our systematic studies concerning the
preparation of calixarene oxovanadium complexes with
varying nuclearity and the investigation of their reactivity
and potential as structural or functional models for active
sites on surfaces.
With the aim of modeling a corresponding vanadium oxo
single site, Feher and co-workers have already examined the
vanadium chemistry of ligands [AR]3-. They were able to
identify a corresponding [VdO]3+ complex in solution,
which, however, is involved in a dimerization equilibrium
Experimental Section
(4) Owens, L.; Kung, H. H. J. Catal. 1993, 144, 202. Evans, O. R.; Bell,
A. T.; Tilley, T. D. J. Catal. 2004, 226, 292.
General Procedures. All manipulations were carried out in a
(5) (a) Comite, A.; Sorrentino, A.; Capannelli, G.; Di Serio, M.; Tesser,
R.; Santacesaria, E. J. Mol. Catal. A 2003, 198, 151. (b) Zhou, R.;
Cao, Y.; Yan, S.; Deng, J.; Liao, Y.; Hong, B. Catal. Lett. 2001, 75,
107. (c) Blasco, T.; Galli, A.; Lopez Nieto, J. M.; Trifido, F. J. Catal.
1997, 169, 203. (d) Arakawa, H.; Aresta, M.; Armor, J. N.; Barteau,
M. A.; Beckman, E. J.; Bell, A. T.; Bercaw, J. E.; Creutz, C.; Dinjus,
E.; Dixon, D. A.; Domen, K.; Dubois, D. L.; Eckert, J.; Fujita, E.;
Gibson, D. H.; Goddard, W. A.; Goodman, D. W.; Keller, J.; Kubas,
G. J.; Kung, H. H.; Lyons, J. E.; Manzer, L. E.; Marks, T. J.;
Morokuma, K.; Nicholas, K. N.; Periana, R.; Que, L.; Rostrup-Nielson,
J.; Sachtler, W. M. H.; Schmidt, L. D.; Sen, A.; Somorjai, G. A.;
Stair, P. C.; Stults, B. R.; Tumas, W. Chem. ReV. 2001, 101, 953.
(6) Chen, B.; Munson, E. J. J. Am. Chem. Soc. 2001, 124, 1638.
(7) Schoiswohl, J.; Surnev, S.; Sock, M.; Ramsey, M. G.; Kresse, G.;
Netzer, F. P. Angew. Chem., Int. Ed. 2004, 43, 5546.
(8) Lorenz, V.; Edelmann, F. T. Z. Anorg. Allg. Chem. 2004, 630, 1147;
Cope´ret, C.; Chabanas, M.; Petroff Saint-Arroman, R.; Basset, J.-M.
Angew. Chem. 2003, 115, 164. Feher, F. J.; Budzichowski, T. A.
Polyhedron 1995, 14, 3239.
(9) Radius, U. Z. Anorg. Allg. Chem. 2004, 630, 957; Floriani, C. Chem.s
Eur. J. 1995, 5, 19.
glovebox, or else by means of Schlenk-type techniques involving
1
the use of a dry argon atmosphere. The H, 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 CH2Cl2,
1
CDCl3, or C6D6 as solvent at 20 °C. The H NMR spectra were
calibrated against the residual proton and natural abundance 13C
resonances of the deuterated solvent (CH2Cl2 δH 5.32 ppm, CDCl3
δH 7.26 ppm, and C6D6 δH 7.15 ppm) and the 51V NMR spectra
(10) Feher, F. J.; Walzer, J. F. Inorg. Chem. 1991, 30, 1689.
(11) (a) Hoppe, E.; Limberg, C.; Ziemer, B.; Mu¨gge, C. J. Mol. Catal. A
2006, 251, 34; b) Castellano, B.; Solari, E.; Floriani, C.; Scopelleti,
R. Inorg. Chem. 1999, 38, 3406.
(12) (a) Gibson, V. C.; Redshaw, C.; Elsegood, M. R. J. J. Chem. Soc.,
Dalton Trans. 2001, 767. (b) Hampton, P. D.; Daitch, C. E.; Alam,
T. M.; Pruss, E. A. Inorg. Chem. 1997, 36, 2879. (c) Castellano, B.;
Solari, E.; Floriani, C.; Scopelliti, R. Inorg. Chem. 1999, 38, 3406.
Also see ref 20a.
Inorganic Chemistry, Vol. 45, No. 20, 2006 8309

Hoppe et al.
against VOCl3 as a standard. Microanalyses were performed on a
Leco CHNS-932 elemental analyzer. Infrared (IR) spectra were
recorded using samples prepared as KBr pellets with a Digilab
Excalibur FTS 4000 FTIR-spectrometer.
arene, Li4B, was carried out according to the literature.19 p-tert-
Butyl-calix[4]arene (3.66 g, 5.64 mmol), 0.16 g (22.5 mmol) of Li
granulate and 2.89 g (22.5 mmol) of naphthalene were stirred in
60 mL of THF at room temperature overnight to give a pale orange
solution. After all volatiles were removed, the remaining residue
was washed with 30 mL of n-hexane and dried in vacuo, yielding
3.70 g (97%) of a white solid. This product was dissolved in 80
mL of toluene; the resulting solution was added to 2.67 g (5.47
mmol) of Ph4P[VO2Cl2] via cannula. After being refluxed for 20
h, the brown suspension was allowed to cool to room temperature.
After all volatiles were removed, the remaining solid was washed
with 80 mL of CH3OH and dried again. It was then redissolved in
30 mL of CH2Cl2 and precipitated by the addition of n-hexane,
yielding 4.98 g (85%) of a green, glittering solid.
Materials. Solvents were purified, dried, and degassed prior to
use. Pure nBu4NVO3,13 Ph4P[VO2Cl2],14 p-tert-butyl-calix[4]arene,15
dimethyl-p-tert-butyl-calix[4]arene,16 and p-tert-butyl-calix[8]-
arene17 were prepared according to the literature procedure.
VO(OCH3)3 was synthesized18 as described in the literature with
some modifications: To a solution of 4.8 mL (0.05 mol) of VOCl3
in 500 mL of toluene, heated to 80 °C, was added 6.1 mL (0.15
mol) of methanol while stirring. After 10 min, 21.1 mL (0.15 mol)
of triethylamine was added. The suspension was stirred at 80 °C
for another 15 min and then annealed to room temperature and
stirred for an additional hour. All volatiles were evaporated under
a vacuum and a temperature not higher than 40 °C. Sublimation of
the green solid at 60 °C in static vacuo yielded 2.8 g (35%) of a
yellow crystalline solid.
1H NMR (CD2Cl2, 297 K): δ 7.89 (m, 4H, Ph4P+), 7.76 (m,
2
8H, Ph4P+), 7.66 (m 8H, Ph4P+), 6.95 (s, 8H, HAr), 4.38 (d, J )
11.2 Hz, 4H, CH2), 2.91 (d, 2J ) 11.2 Hz, 4H, CH2), 1.18 (s, 36H,
C(CH3)3). 13C NMR (CD2Cl2, 197 K): δ 158.9 (q-CAr), 143.0 (q-
CAr), 136.1 (d, J ) 3 Hz, Ph4P+), 134.9 (d, J ) 10 Hz, Ph4P+),
131.0 (d, J ) 13 Hz, Ph4P+), 130.5 (q-CAr), 123.8 (CAr-H), 118.0
(d, J ) 89 Hz, Ph4P+), 34.7 (CH2), 34.1 (C(CH3)3), 31.8 (C(CH3)3).
31P NMR (CD2Cl2, 297 K): δ 23.76. 51V (CD2Cl2, 297 K): δ -297.
Anal. Found: C, 72.52; H, 6.79; Cl, 6.55. Calcd for C68H72O5PV
CH2Cl2: C, 72.94; H, 6.56; Cl, 6.24. Mp: 368 °C. IR (KBr, cm-1):
3046 (w), 2952 (s), 2918 (m), 2865 (m), 1587 (w), 1457 (s), 1441
(m), 1392 (w), 1361 (m), 1310 (m), 1286 (m), 1260 (s), 1207 (s),
1110 (s), 993 (s), 869 (w), 830 (m), 799 (s), 755 (w), 724 (s), 688
(m), 546 (m), 526 (s), 501 (w), 426 (s).
nBu4N [p-tert-Butyl-calix[4]arene Tetrahydroxylato Oxo-
vanadate(V)], NBu41. Triethylamine (0.17 mL, 1.23 mmol) was added
to a suspension of 200 mg (0.31 mmol) of p-tert-butyl-calix[4]-
arene, H4B, in 5 mL of CH2Cl2, and the resulting solution was
treated dropwise with a solution of 210 mg (0.62 mmol) of nBu4-
NVO3 in 5 mL of CH2Cl2. After being stirred for 72 h at room
temperature, the solution was evaporated to dryness and the resulting
green-brown solid was extracted with diethyl ether (2 × 10 mL).
Removal of all volatiles yielded a brown solid that was dissolved
in 2-3 mL of acetone. This solution was cooled to -50 °C, which
led to the precipitation of yellowish needles (that were discarded).
The overlaying brown solution was separated via filtration, and
evaporation of the solvent provided 172 mg (59% based on
employed H4B) of a crystalline brown solid (crude product). Further
purification can be achieved via recrystallization (cooling of a
saturated CH2Cl2 solution). This process, which leads to an
analytically pure sample, is, however, time-consuming and ac-
companied by significant losses in yield.
Oxo[O-acetyl-p-tert-butyl-calix[4]arene-trihydroxylato]vana-
dium(V), 2. A solution of CH3COCl (0.5 mL, 0.96 M) in CH2Cl2
(0.48 mmol) was added to a solution of 250 mg (0.24 mmol) of
PPh41 in 10 mL of CH2Cl2. Within 10 min, the color of the solution
changed from yellow-green to red-brown. After being stirred for
another 15 min, all volatiles were evaporated; the residual solid
was extracted with n-hexane (5 mL). This led to a ruby-colored
solution that was filtered from a pale mauve solid (predominantly
Ph4PCl). Removing the solvent from the filtrate yielded 160 mg
(89%) of a red solid.
2
1H NMR (CDCl3, 297 K): δ 6.96 (s, 8H, HAr), 4.47 (d, J )
2
11.2 Hz, 4H, CH2), 3.39 (m, 8H, N-(CH2)4), 2.99 (d, J ) 11.2
4
1H NMR (CDCl3, 297 K): δ 7.25 (d, J ) 2.4 Hz, 2H, HAr),
Hz, 4H, CH2), 1.65 (m, 8H, N-(CH2-CH2)4), 1.42 (m, 8H,
7.16 (d, 4J ) 2.4 Hz, 2H, HAr), 6.94 (s, 2H, HAr), 6.90 (s, 2H, HAr),
4.59 (d, 2J ) 13.5 Hz, 2H, CH2), 4.37 (d, 2J ) 13.1 Hz, 2H, CH2),
3.37 (d, 2J ) 13.1 Hz, 2H, CH2), 3.19 (d, 2J ) 13.5 Hz, 2H, CH2),
2.63 (s, 3H, CH3C(O)), 1.42 (s, 18H, C(CH3)3), 0.90 (s, 9H,
C(CH3)3), 0.82 (s, 9H, C(CH3)3). 13C NMR (CDCl3, 297 K): δ
167.9 (CAr), 161.3 (CAr), 154.0 (CAr), 147.8 (CAr), 147.1 (CAr), 144.9
(CAr), 142.0 (CAr), 130.7 (CAr), 129.8 (CAr), 125.9 (CAr), 125.1 (CAr),
124.7 (CAr-H), 124.4 (CAr-H), 124.3 (CAr-H), 124.2 (CAr-H), 33.0
(C(O)CH3), 32.6 (C(CH3)3), 32.5 (C(CH3)3), 31.7 (CH2), 31.6 (CH2),
30.4 (C(CH3)3), 29.5 (C(CH3)3), 29.3 (C(CH3)3). 51V NMR (C6D6,
297 K): δ -460 (s, br). Anal. Found: C, 74.01; H, 9.49. Calcd
for C46H55O6V 1.5 C6H14: C, 74.71; H, 8.66. Mp: 128 °C (dec).
IR (KBr, cm-1): 3524 (m), 3048 (w), 2960 (s), 2905 (m), 2868
(w), 1751 (s), 1481 (s), 1461 (m), 1393 (w), 1364 (s), 1280 (m,
br), 1206 (s, br), 1119 (m), 1012 (m), 996 (m), 943 (w), 922 (w),
871 (m), 818 (w), 799 (w), 694(vw), 596 (vw).
2
N-(CH2-CH2-CH2)4), 1.18 (s, 36H, C(CH3)3), 0.97 (t, J ) 6
Hz, 12H, N-((CH2)3-CH3)4). 13C NMR (CDCl3, 297 K): δ 156.2
(q-CAr), 140.1 (q-CAr), 127.5 (q-CAr), 121.4 (CAr-H), 56.3 (N-
(CH2)4) 32.4 (CH2), 31.4 (C(CH3)3), 29.2 (C(CH3)3), 21.7 (N-
(CH2-CH2)4), 17.27 (N-(CH2-CH2-CH2)4), 11.3 (N-((CH2)3-
CH3)4); 51V NMR (CDCl3, 297 K): δ -268. Anal. Found: C,
75.68; H, 9.94; N, 1.74. Calcd for C60H88NO5V: C, 75.52; H, 9.29;
N, 1.47. IR (KBr, cm-1): 3047 (w), 2961 (s), 2872 (m), 1458 (s,
br), 1420 (w), 1391 (w), 1391 (w), 1380 (m), 1360 (m), 1308 (m),
1283 (m), 1252 (m), 1204 (s), 1109 (w), 1030 (w), 978 (s), 920
(w), 883 (w), 870 (w), 829 (m), 798 (s), 758 (w), 727 (w), 696
(w), 677 (w), 544 (m), 498 (w), 432 (m), 422 (m).
Ph4P [p-tert-Butyl-calix[4]arene Tetrahydroxylato Oxovana-
date(V)], PPh41. Synthesis of the deprotonated p-tert-butyl-calix[4]-
(13) Day, V. W.; Klemperer, W. G.; Yagasaki, A. Chem. Lett. 1990, 19,
1267.
(14) (a) Ehrlich, P.; Engel, W. Z. Anorg. Allg. Chem. 1963, 322, 217. (b)
Fenske, D.; Shihada, A.-F.; Schwab, Dehnicke, K. Z. Anorg. Allg.
Chem. 1980, 471, 140.
(15) Gutsche, C. D.; Iqbal, M. Org. Synth. 1993, CV 8, 75.
(16) Gutsche, C. D.; Dhawan, B.; Levine, J. A.; No, K. H.; Bauer, L. J.
Tetrahedron 1983, 39, 409.
Ph4P[p-tert-butyl-calix[8]arene heptahydroxylato dioxodi-
vanadate(V)], 3. p-tert-Butyl-calix[8]arene (2.00 g, 1.5 mmol) and
0.43 mL (3.1 mmol) of triethylamine were added to a solution of
1.52 g (3.1 mmol) of Ph4P[VO2Cl2] in 80 mL of acetonitrile. After
being stirred at room temperature for 5 h, the black solution was
(17) Munch, J. H.; Gutsche, C. D. Org. Synth. 1993, CV 8, 80.
(18) Funk, H.; Weiss, W.; Zeising, N. Z. Anorg. Allg. Chem. 1958, 295,
327.
(19) Giullemot, G.; Solari, E.; Rizzoli, C.; Floriani, C. Chem.sEur. J. 2002,
8, 2072.
8310 Inorganic Chemistry, Vol. 45, No. 20, 2006

OxoWanadium(V)calixarenes as Oxidation Catalysts
Table 1. Crystal Data and Experimental Parameters for the Crystal Structure Analyses of Ph4P1, 2, 3, and 4
Ph4P1
2
3
4
formula
fw (g mol-1
T (K)
C68H72O5PV CH2Cl2
1136.09
150(2)
C46H55O6V
754.84
180(2)
C126H146N7O10PV2
2051.35
150(2)
C48H64O9V2
886.99
180(2)
)
wavelength (Å)
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
0.71073
tetragonal
P4/n
12.763(3)
12.763(3)
18.733(9)
90
90
90
3050(2)
2
1.237
0.71073
monoclinic
P21/a
11.752(2)
29.076(4)
12.547(2)
90
104.94(2)
90
4142.2(10)
4
0.71073
monoclinic
P21/n
17.0125(3)
28.8155(7)
23.6724(4)
90
99.1190(10)
90
11458.1(4)
4
1.189
0.238
4368
0.71073
orthorhombic
Pbcn
20.447(2)
11.381(2)
20.263(2)
90
90
90
4715.3(8)
4
1.089
Z
D (g cm-3
)
1.210
0.286
1608
µ (MoKR) (mm-1
F(000)
)
0.326
1200
0.252
1656
GOF
1.047
0.879
1.123
0.914
R
ind [I > 2σ(I)]
R1 ) 0.0402,
wR2 ) 0.1020
R1 ) 0.0550
wR2 ) 0.1135
-0.429/0.624
R1 ) 0.0454
wR2 ) 0.0925
R1 ) 0.0863
wR2 ) 0.1033
-0.392/0.632
R1 ) 0.0689
wR2 ) 0.1594
R1 ) 0.0862
wR2 ) 0.1677
-0.648/1.905
R1 ) 0.0596
wR2 ) 0.1568
R1 ) 0.1051
wR2 ) 0.1741
-0.350/0.667
Rind (all data)
∆Fmin/∆Fmax (e Å-3
)
Scheme 2
filtered and all volatiles were evaporated. The resulting black solid
was crystallized by cooling an acetonitrile solution, yielding 2.25
g (83%) of pure 3.
The 1H and 13C NMR spectroscopic data are in agreement with
those described in the literature20 before. Anal. Found: C, 74.07;
H, 7.04; N, 3.37. Calcd for C112H125O10PV2‚5CH3CN: C, 74.41;
H, 7.16; N, 3.55. Mp: >300 °C.
Dimethoxo-dioxo-µ-oxo[1,3-dimethyl-p-tert-butyl-calix[4]arene-
dihydroxylato]divanadium(V), 4. An orange solution of 0.50 g
(3.1 mmol) of OV(OCH3)3 and 1.06 g (1.6 mmol) of 1,3-dimethyl-
p-tert-butyl-calix[4]arene in 40 mL of THF was heated under reflux
for one night. The solution’s color changed to darkish red. All
volatiles were evaporated to dryness; the resulting solid was
extracted with acetonitrile and all volatiles were again evaporated.
The residual red solid was redissolved in 10 mL of CH2Cl2, and
after the addition of 30 mL of n-hexane, the solution was
concentrated in a vacuum until small amounts of a red precipitate
appeared. Cooling of the solution overnight to 5 °C yielded 1.1 g
(80%) of a crude product 4. A pure sample could be obtained by
recrystallization from a CH2Cl2/n-hexane solution and washing with
small amounts of n-hexane.
4
1H NMR (C6D6, 297 K, 4t): δ 7.30 (d, 2H, J ) 2.0 Hz, HAr),
4
4
7.22 (d, 2H, J ) 2.3 Hz, HAr), 6.87 (d, 2H, J ) 2.0 Hz, HAr),
4
6.81 (d, 2H, J ) 2.3 Hz, HAr).
4
1H NMR (C6D6, 297 K, 4c): δ 7.28 (d, 2H, J ) 2.3 Hz, HAr),
7.24 (d, 2H, J ) 2.3 Hz, HAr), 6.84 (s(br), 4H, HAr). 13C NMR
4
(CDCl3, 297 K): δ 168.5, 168.3, 152.7, 146.3, 145.8, 133.2, 133.0,
132.6, 132.4, 128.1 (CAr-H), 126.4(CAr-H), 126.2(CAr-H), 126.1-
(CAr-H), 125.9(CAr-H), 125.5 (CAr-H), 123.7 (CAr-H), 75.2 (VOCH3),
63.4 (OCH3), 63.0 (OCH3), 53.0 (C(CH3)3), 34.5 (CH2), 33.9 (CH2),
31.9(C(CH3)3), 31.2 (C(CH3)3). 51V NMR (C6D6, 297 K): δ -542.
Anal. Found: C, 65.73; H, 7.32. Calcd for C48H64O9V2: C, 65.00;
H, 7.27. IR (KBr, cm-1): 3047 (w), 2961 (s), 2920 (m), 2866 (w),
2820 (w), 1481 (s, br), 1362 (m), 1313 (w), 1300 (w), 1250 (m),
1207 (s), 1123 (m), 1059 (s), 1017 (m), 995 (m), 941 (w), 918
(w), 860 (m), 813 (vs), 798 (vs), 775 (w), 696 (w), 648 (w), 635
(m), 625 (w), 588 (w).
1H NMR (CDCl3, 297 K, 4c): δ 5.11 (s, 6H, VOCH3), 4.36 (d,
2H, 2J ) 12.5 Hz, CH2), 4.09 (s, 3H, CArOCH3), 4.05 (d, 2H, 2J )
12.5 Hz, CH2), 3.84 (s, 3H, CArOCH3), 3.36 (d, 2H, 2J ) 12.5 Hz,
CH2), 3.14 (d, 2H, 2J ) 12.5 Hz, CH2), 1.35 (s, 18H, CCH3), 0.84
1
(s, 9H, CCH3), 0.81 (s, 9H, CCH3). H NMR (CDCl3, 297 K, 4t):
δ 5.16 (s, 6H, VOCH3), 5.10 (d, 2H, 2J ) 12.5 Hz, CH2), 4.32 (d.
Protocol for the Oxidation Studies. Catalyst (0.01 mmol) was
dissolved in 1.5 mL of acetonitrile. After the addition of 500 mg
of molecular sieve powder (3 Å), 1.00 mmol alcohol in 0.5 mL of
acetonitrile was added. Subsequently, the mixture was stirred
vigorously for 3 h (or for 1 h when 3 h led to a complete conversion)
at 80 °C under a dioxygen atmosphere (oxygen was allowed to
pass through the reaction vessel for 15 s before the reaction, and
this procedure was repeated every 60 min). The mixture was then
cooled to room temperature, and the molecular sieve powder was
separated via filtration. In the case of high boiling alcohols (and
resulting carbonyl compounds), the acetonitrile solvent was then
simply evaporated in a vacuum. The conversion was determined
by 1H NMR spectroscopy. In the case of low boiling alcohols, the
2H, 2J ) 12.5 Hz, CH2), 3.94 (s, 6H, CArOCH3), 3.36 (d, 2H, 2J )
2
12.5 Hz, CH2), 3.15 (d, 2H, J ) 12.5 Hz, CH2), 1.35 (s, 18H,
CCH3), 0.82 (s, 18H, CCH3).
Because of serious overlap, the signals for the aryl protons cannot
be assigned unambiguously to the cis and trans isomers in CDCl3
as the solvent (which, however, better resolves the other regions).
They appear in the form of multiplets of equal integral at δ 7.23
(m, HAr), 7.10 (m, HAr), 6.64 (m, HAr), 6.56 (m, HAr). In C6D6
solution, the corresponding signals can be assigned:
(20) (a) Hofmeister, G. E., Hahn, F. E.; Pedersen, S. F. J. Am. Chem. Soc.
1989, 111, 2318. (b) Hofmeister, G. E.; Alvarado, E.; Leary, J. A.;
Yoon, D. I.; Pedersen, S. F. J. Am. Chem. Soc. 1990, 112, 8843.
Inorganic Chemistry, Vol. 45, No. 20, 2006 8311

Hoppe et al.
Scheme 3
1
reaction mixture was filtered and directly employed in H NMR
spectroscopic investigations.
contact of the metavanadate starting material with the acidic
protons in H4B triggers aggregation and condensation
reactions leading to II instead of a calixarene complex.
Hence, in the next attempt, NEt3 was added as a base, which
indeed enabled formation of the envisioned complex NBu4-
[OdVB], NBu41, (Scheme 3) in good yields. However, it is
difficult to completely separate NBu41 from small amounts
of unreacted H4B during the workup procedure. Analytically
pure NBu41 can be obtained via crystallization by cooling a
CH2Cl2 solution but only if significant losses in yield are
accepted. Alternatively, the complex anion 1 can be synthe-
sized starting from the lithium salt Li4B, which can be
obtained via reaction of H4B with lithium naphthalenide:
Corresponding treatment with tetraphenylphosphonium dichlo-
rodioxovanadate PPh4[VO2Cl2] leads to the elimination of
LiCl (and apparently also Li2O), yielding PPh4[OdVB],
PPh41, (Scheme 3) for which an X-ray crystal structure
analysis was performed (see Figure 1). As expected, the
vanadyl group in 1 is surrounded symmetrically by an O4
donor set, and 1 therefore represents the first model for such
a situation on an oxidic support.
All starting materials and products are known and commercially
available. The turnover frequencies (abbreviated TOF) were
calculated as the number of moles of substrate that a mole of catalyst
can convert per unit time.
Crystal Structure Determinations Suitable crystals of NBu41
could be obtained by cooling a saturated solution in CH2Cl2 to -30
°C. Single crystals of PPh41 and 5 were obtained by diffusion of
n-hexane at room temperature into a CH2Cl2 solution of PPh41 and
5, respectively. Suitable crystals of 2 were obtained by evaporation
of the solvent from a saturated hexane solution. Single crystals of
3 were obtained by cooling a saturated acetonitrile solution at 5
°C. Crystals of 4 were obtained by evaporation of the solvent from
a saturated acetonitrile solution at room temperature. The crystals
were mounted on a glass fiber and then transferred into the cold
nitrogen gas stream of the diffractometer (Stoe Stadi for PPh41, Stoe
IPDS for NBu41, 2, and 4, Stoe IPDS2T for 3 and 5) using MoKR
radiation, λ ) 0.71073 Å, and the structures were solved by direct
methods (SHELXS-97),21 refined versus F2 (SHELXL-97)22 with
anisotropic temperature factors for all non-hydrogen atoms (Table
1). All hydrogen atoms were added geometrically and refined by
using a riding model (except for H113 in the crystal structure of 3,
which was found).
The crystallographic data (apart from structure factors) of NBu41,
PPh42, 3, and 4 were deposited at the Cambridge Crystallographic
Data Center as supplementary publications CCDC 610225, 610226,
610227, 610228, and 610229, respectively. 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).
Results and Discussion
Syntheses and Structures of Mono- and Dinuclear
Calixarene Vanadium Oxo Complexes. For the above-
mentioned purpose first of all a set of various different
calixarene vanadium oxo complexes had to be prepared. To
obtain a complex containing an OdV3+ unit embedded in a
tetragonal oxo coordination environment, we reacted tetra-
butylammonium metavanadate with H4B. This, however, led
in good yields to the isolation of a pentanuclear vanadate II
(as evidenced by single-crystal X-ray analysis) that first
caught attention23 in 1989, as it represented a novel type of
structure in polyoxometallate chemistry being based on
corner-sharing tetrahedrons (see Scheme 2). Obviously, the
Figure 1. Molecular structure of the anion of PPh41. All hydrogen atoms,
the Ph4P+ cation, and a cocrystallized solvent molecule CH2Cl2 were omitted
for clarity. The molecule contains a crystallographic C4 axis through V
and O2. Selected bond distances (Å) and angles (deg): V-O2, 1.5917(8);
V-O1, 1.8858(4); V-O4, 0.3690(5); CAr-O1-V, 133.144(10); O2dV-
O1, 101.285(13).
(21) Sheldrick, G. M. SHELXS-97 Program for Crystal Structure Solution;
University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(22) Sheldrick, G. M. SHELXL-97 Program for Crystal Structure Refine-
ment; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(23) Day, V. W.; Klemperer, W. G.; Yaghi, O. M. J. Am. Chem. Soc. 1989,
111, 4518.
8312 Inorganic Chemistry, Vol. 45, No. 20, 2006

OxoWanadium(V)calixarenes as Oxidation Catalysts
Scheme 4
The cavity hosts a CH2Cl2 molecule, and the coordination
sphere around the vanadium atom can be described as being
almost ideal square pyramidal: as a crystallographic C4 axis
runs through V and O2, all V-OAr bonds have the same
length (1.8858(4) Å). However, the V center is not located
in the plane formed by the corresponding four O atoms but
0.3690(5) Å above it. The terminal V-O2 bond distance
(1.5917(8) Å) is very similar to that in I (1.598(5) Å),
whereas the other V-O bonds are somewhat longer than
comparable bonds in I (1.793(5), 1.811(6), and 1.825(5) Å).
Figure 2. Molecular structure of complex 2. All hydrogen atoms were
omitted for clarity. Selected bond distances (Å) and angles (deg): O1-V,
3.040(2); O2-V, 1.753(2); O3-V, 1.798(2); O4-V, 1.785(2); O5-V,
1.596(2); O1-V-O5, 79.89(8); O2-V-O3, 102.61(8); O2-V-O4, 115.95-
(9); O2-V-O5, 113.60(10); O3-V-O4, 110.34(9); O3-V-O5, 104.36-
(10); O4-V-O5, 109.14(10); CAr-O1-V, 99.04(13); CAr-O2-V, 206.0-
(2); CAr-O3-V, 106.5(2); CAr-O4-V, 145.5(2); Ar(O1)-Ar(O3), 15.42(11);
Ar(O2)-Ar(O4), 59.45(11).
PPh41 and NBu41 are comparatively stable compounds that
can even be handled in air for short periods of time. PPh41
does not react with alcohols such as ethanol and methanol,
and the solid is stable against water. According to the NMR
data, 1 shows in solution the same structure as in the
crystalline state: Singlet signals for the tBu protons and the
aryl protons as well as the fact that only two doublets are
observed for the methylene protons point to a C4 symmetry,
as expected from the results of X-ray crystallography.
The 51V NMR signal appears at -268 ppm for NBu41 in
CDCl3 and at -297 ppm for PPh41 in CD2Cl2, i.e., shifted to
lower field in comparison to the known neutral compound I
(-352 ppm).11a
For comparative reactivity studies, it seemed desirable to
also investigate a neutral calix[4]arene vanadium(V) oxo
complex. As mentioned, so far one representative has been
described (I) that, however, is difficult to isolate in the pure
state free from VIV impurities.11 Hence, we contemplated the
synthesis of a more easily accessible derivative. Indeed,
starting from PPh41, such a neutral complex can be obtained
in a clean and facile synthesis: the addition of CH3C(O)Cl
leads to the precipitation of PPh4Cl and formation of Od
VBOAc, 2 (see Scheme 4). The result of an X-ray crystal
analysis is shown in Figure 2.
line 2 adopts the partial cone (so-called “paco”) conformation
(CAr-V-O2 ) 206.0(2)°). Often, the corresponding “cone”
structures (with Cs symmetry) are very close in energy, so
that sometimes even both conformations are observed in the
solid depending on the conditions of crystallization,11a and,
furthermore, a situation favored in the solid (where packing
effects occur) might not be the one favored in solution.
Interconversion is possible in solution so that the most stable
structure is adopted if the activation barrier is not too high.
This is indeed the case for 2.
1
In the H NMR spectrum, three signals in the ratio 2:1:1
can be observed for the protons of the tBu groups, and for
the methylene protons, two pairs of doublets are found,
indicating a Cs symmetry with a mirror plane running through
the CAr-OAc moiety, the VdO group, and the phenolic unit
opposite to the acetylated V-O-CAr group. This suggests
that the cone structure is preferred in solution (or that all
phenoxy ligands in 2 switch very rapidly so that an averaged
spectrum is observed on the NMR time scale). In a Cs
symmetric structure, the two protons of the two different
aryl rings halved by the mirror plane should be magnetically
equivalent and show one singlet per aryl ring. The remaining
two aryl rings should be equivalent, but not the two protons
within an individual unit. Hence, for these two rings, two
doublets should be observed. This is exactly what is found
experimentally in the region where the aryl protons absorb,
which again points to a cone conformation.
The chemical shift in the 51V NMR spectrum amounts to
-460 ppm, i.e., the signal is shifted by more than 100 ppm
to higher field in comparison to I. Comparing the series 2,
I, and 1, it thus has to be noted that starting from a
tetrahedrally coordinated vanadyl group (2), via weak
coordination of an additional ligand (I), up to strong
coordination (1) leads to an increased deshielding of the
vanadium center. It is difficult to elucidate the exact reasons
for this trend, because structure and electronics change
simultaneously.
Whereas the coordination geometry around V in I has to
be described most appropriately as being distorted trigonal
bipyramidal because of the fact that the methoxy group binds
at a longer distance to V too (compare Chart 2), the vanadium
center in 2 is coordinated tetrahedrally by the O atoms O2,
O3, O4, and O5; there is no interaction with the O atom of
the acetyl unit (d ) 3.040(2) Å), and consequently, 2 can
be regarded as modeling species a in Chart 1. However, both
complexes I and 2 show almost equal VdO bond lengths;
the bonds to the aryloxy O atoms are somewhat shorter in 2
(1.753(2), 1.798(2), and 1.785(2) Å) than in I (1.793(5),
1.811(6), and 1.825(5) Å). As found in one of the molecular
structures reported for I,11a the calixarene moiety in crystal-
Inorganic Chemistry, Vol. 45, No. 20, 2006 8313

Hoppe et al.
for a similar titanium complex; this proposed structure is
identical with the one determined here by X-ray crystal-
lography.
The formation of 3 must have proceeded via the addition
of OH functions to VdO groups under elimination of HCl
and water (a similar type of reactivity has been reported only
recently for molybdenyl chemistry)24 so that two at first
independent OdV3+ units are coordinated by four aryloxide
ligands each, as in 1. However, two facts prohibit a
description of 3 as a dinuclear version of 1: First, one of
the phenol functions in H8B′′ is not deprotonated (i.e., a
coordinated alcohol function results), so that 3 contains only
a monoanion, and second, the ligand folds in a way that
allows for the coordination of five phenoxy functions at the
vanadium centers; that is, two of them within 3 adopt a
bridging binding modus. At the remaining coordination sites,
the terminal oxo ligands (O9 and O10) are bonded, being
positioned in a cavity defined by three aryl rings; the two
vanadium atoms are thus located in pseudooctahedral
environments. As mentioned already, at one of the calixarene
oxygen atoms (O5), a proton (H113) remained bonded, which
could be located in the crystal structure analysis (O-H )
0.93(8) Å); it undergoes a hydrogen bond to the opposite
calixarene oxygen atom (H‚‚‚O1 ) 1.61(8) Å, O5-H113‚
‚‚O1 ) 177(7)°). The vanadium atoms V1 and V2 are
positioned somewhat (0.2312(4) and 0.2202(4) Å) above the
ideal planes through O1, O2, O3, O8 and O4, O5, O6, O7,
respectively. The terminal VdO bond distances (V1-O9 )
1.589(2) Å, V2-O10 ) 1.599(2) Å) lie in the usual range,
being comparable to those displayed by I, 1, and 2. The
distance between the two vanadium atoms amounts to
3.3677(7) Å.
Figure 3. Molecular structure of the anion of 3. All hydrogen atoms, the
tert-butyl groups, and cocrystallized solvent molecules (acetonitrile) were
omitted for clarity. Selected bond distances (Å) and angles (deg): O1-
V1, 2.036(2); O2-V1, 1.849(2); O3-V1, 1.852(2); O4-V1, 2.212(2); O8-
V1, 1.992(2); O9-V1, 1.589(2); O5-V2, 2.141(2); O6-V2, 1.836(2); O7-
V2, 1.841(2); O4-V2, 1.968(2); O8-V2, 2.174(2); O10-V2, 1.599(2);
O5-H113, 0.93(8); O1-H113, 1.61(8); O5-H113-O1, 177(7); V1-O4-
V2, 107.20(8); V1-O8-V2, 107.77(8); V1-O4(O1, O2, O3, O8), 0.2313-
(4); V2-O4(O4, O5, O6, O7), 0.2202(4).
Scheme 5
Having accessed a dinuclear, singly charged vanadium oxo
complex, we also pursued the syntheses of a corresponding
neutral complex, ideally containing V-O-V units. Refluxing
H2BMe2 in THF solution together with VO(OCH3)3 overnight
leads to a dark red solution from which red crystals could
be obtained after work up. An X-ray diffraction analysis of
a selected single crystal revealed the molecular structure
of the first dinuclear calix[4]arene complex, [µ-O(Od
V(OMe))2BMe2], 4t (see Scheme 6 and Figure 4).
Like I, compound 2 shows only a slight sensitivity toward
oxygen, but it is very sensitive to water. In the absence of
both, it is stable in solution for long periods.
We next turned our attention to the synthesis of compounds
containing two oxovanadium units to also allow for a
comparison of the reactivity of dinuclear vs mononuclear
complexes. Accordingly, higher calixarenes were employed
as starting materials, as these offer room for more than one
vanadium center. First of all, H8B′′ was chosen to be reacted
with PPh4[VO2Cl2] in the presence of Et3N as a base. The
resulting product was fully characterized, and a single-crystal
X-ray analysis revealed the structure (Figure 3 and Scheme
5) and constitution of PPh4[O2V2HB′′], 3.
The vanadium centers each form only one bond to BMe2
.
They are coordinated by one aryloxy group with an
(ar)C-O-V angle that is characteristic of the paco confor-
mation; nonetheless, the paco conformation is not adopted
here, because none of the vanadium centers has a central
position. They have moved toward the lower rim being
bridged by an oxo ligand. The residual coordination sphere
is completed for each V center by one methoxy and one
terminal oxo ligand, so that the geometries around the V
atoms can be assigned as being tetrahedral. There is no
bonding interaction between the V atoms and the methoxy
groups of the ligand BMe2, as indicated by their distances to
vanadium, which are longer than the van der Waals distances.
A crystallographic C2 axis runs through the bridging O5 atom
It turned out that a salt containing the anion of 3 in
combination with a different cation had been reported before
by Pedersen and co-workers20 as the product of the conver-
sion of VO(OiPr)3 with H8B′′ in the presence of (R)-(+)-1-
(1-naphthyl)ethylamine. A single-crystal X-ray diffraction
study had not been performed, but the solution structure of
3 had been investigated in detail. A suggestion for a structure
was finally made on the basis of a structure determination
(24) Liu, L.; Zakharov, L. N.; Golen, J. A.; Rheingold, A. L.; Watson, W.
H.; Hanna, T. A. Inorg. Chem. 2006, 45, 4247.
8314 Inorganic Chemistry, Vol. 45, No. 20, 2006

OxoWanadium(V)calixarenes as Oxidation Catalysts
Scheme 6
and the center of the “calyx”. The distance between the two
vanadium centers amounts to 3.3060(2) Å. NMR spectro-
scopic studies with the bulk product showed that the reaction
in Scheme 6 leads not only to 4t with the two V-bound
methoxy groups in trans orientation as in Figure 4 but also
to an isomer 4c where they show a cis positioning (product
individual signals appear separated from each other in C6D6.
Three signals are apparent with an intensity allowing an
assignment to 4c: two doublets and a broad singlet with
doubled intensity. This can be explained as follows: the two
vanadium-ligated phenolic moieties are equivalent, so that
both show the same sets of signals. However, the protons
within each of those phenoxy groups are diastereotopic,
showing different chemical shifts; they couple with each
other to give two doublets. The aryl protons within the two
remaining aryl groups are equal, giving rise to singlets; the
two aryl rings differ from each other only slightly so that
the expected two singlets overlap to give a broad signal with
doubled integral. In 4t, there is no mirror plane running
through any of the aryl rings so that within each ring, the
two protons present are magnetically inequivalent. In turn,
pairwise, two of the aryl groups are equivalent; in total, 4
doublets with equal intensity should be expected, as observed.
From one experiment to the other, the relative yield of 4t to
4c varies somewhat, as evidenced by NMR spectroscopy.
Samples showing different ratios do not adjust within a
couple of hours, which already indicates that the isomers
are not in a rapidly adjusting equilibrium to each other.
Heating to 60 °C did not alter their ratio. The 51V NMR
signals for 4t and 4c are coincidentally isochronic (which is
not surprising, as the surroundings of the vanadium centers
vary only slightly in the second coordination sphere), so only
one signal is observed. The chemical shift of -542 ppm
compares best with that of 2, whose vanadium center also
shows a tetrahedral coordination sphere. In conclusion of
these investigations, it can thus be stated that 4t also retains
the structure it shows in the solid state; on this basis, it can
be safely assumed that the same is true for 4c.
1
ratio 3:2). This clearly becomes evident from H NMR
spectra recorded in C6D6 and CDCl3. As mentioned already
4t displays a C2 axis, whereas 4c exhibits a mirror plane
through the two methoxy units and the bridging O atom.
This leads for both isomers to two different types of
methylene groups and thus, as observed and resolved in
CDCl3 solution, to two pairs of doublets that can be easily
assigned to the individual isomers according to their intensity.
For 4t, two different types of tBu signals are expected,
whereas 4c should show three signals with an intensity ratio
of 2:1:1. These signal sets can be found in the experimental
spectrum, too, displaying the expected ratios. Furthermore,
in the region of the aryl-bound methoxy groups, three singlets
can be observed, two of them showing equal intensity.
Indeed, in 4c, these methoxy groups should be magnetically
different so that two signals are expected, whereas they are
equivalent in 4t, leading to only one singlet. The vanadium
bound methoxy groups are equal in both 4t and 4c, leading
to two singlets in a 3:2 ratio. The region of the aryl protons
cannot be resolved in CDCl3 because of overlap, but the
Reactivity of Compounds 1-4 with Respect to the
Activation of O2 for the Oxidation of Alcohols. As outlined
in the Introduction, the most prominent conversions that can
be carried out using vanadia on oxidic supports as hetero-
geneous catalysts are the ODH of methanol and alkanes.4,5,25
Naturally, molecular models such as those described here
will prove themselves unreactive in contact with alkanes, as
the rate determining step of the ODH is the first C-H bond
cleavage being characterized by a large activation barrier
(heterogeneous catalysis at elevated temperatures is required).
However, methanol and higher alcohols are less inert as
substrates, so their reactions can also be investigated with
molecular compounds in the liquid phase. Any information
derived from such studies might contribute to a more
Figure 4. Molecular structure of complex 4t. All hydrogen atoms were
omitted for clarity. Selected bond distances (Å) and angles (deg): O1-V,
3.056(2); O2-V, 1.788(4); O3-V, 1.756(4); O4-V, 1.593(4); O5-V,
1.766(2); O2-V-O3, 110.3(2); O2-V-O4, 108.4(2); O2-V-O5, 113.44-
(7); O3-V-O4, 104.3(2); O3-V-O5, 111.79(13); O4-V-O5, 108.1(1);
V-O5-V′, 138.796(3); O3-CH3, 1.455(8); V-O3-CH3, 130.1(4); VO2-
V′O2′, 9.86(9); Ar(O1)-Ar(O1′), 9.5(2); Ar(O2)-Ar(O2′), 88.0(2).
(25) Do¨bler, J.; Pritzsche, M.; Sauer, J. J. Am. Chem. Soc. 2005, 127, 10861.
Inorganic Chemistry, Vol. 45, No. 20, 2006 8315

Hoppe et al.
Table 2. Substrate Oxidation with O2 and 1 mol % 1-4 as Catalysts
TOF (h-1
)
yield (%) after 3 ha
PPh41
2
3
4
VO(acac)2
PPh41
2
3
4
VO(acac)2
cyclohexanol
1-hexanol
2-pentanol
benzyl alcohol
cinnamyl alcohol
crotyl alcohol
1-phenyl-1-propargylic alcohol
1-phenyl-1-propanol
9-fluorenol
0
0
0
0
0
0
0.7
0.3
0
0
0
0
0
1.7
0
2.3
0.7
3.0
0.7
0
0
0
0
0
0
0
0
20.3
0
15.3
1.3
15.3
1.3
0
0
0
0
12.3
50.0
9.0
60.0
2.7
24.0
0
0
0
0
0
0
0
2
1
0
0
0
0
0
0
0
5
0
7
2
15
2
0
0
0
10
21
0
77
2
0
0
0
0
61
0
46
4
46
4
0
0
0
13
100b
27
100c
8
3.3
7.0
0
25.7
0.7
86.0
10.0
0
100d
30
0
72
0
0
9,10-H-dihydroanthracene
fluorene
0
0
0
0
0
a The numbers in the right column specify the yield of oxidation product referring to the alcohol. b 50% yield after 1 h. c 60% yield after 1 h. d 86% yield
after 1 h.
comprehensive understanding of the methanol oxidation at
vanadia catalysts and they might as well provide ideas
concerning the (probably) more complex ODH of alkanes.
To test the compounds described above for any reactivity
in relation to the ODH and to thus test them as models, we
therefore decided to investigate the oxidation of a palette of
various different alcohols with O2 as the reagent. Of course,
there are numerous methods to oxidize alcohols with organic
or inorganic reagents,26 which, however, are often stoichio-
metric and/or lead to toxic waste products. The performance
of catalytic oxidations with O2 is therefore still an attractive
topic, and a lot of research has been devoted to it in the
past.26 Certainly, the background of the study presented here,
as mentioned, was an inquiry into whether compounds 1-4
could represent not only structural but also functional models
for the heterogeneously catalyzed ODH and not the improve-
ment of known catalysts for the aerobic oxidation of alcohols.
Nevertheless, the latter aspect was borne in mind as a
potential, beneficial quality of the investigation.
As we are dealing with VV compounds that are discussed
as functional models, they should contain functional moieties
already, i.e., the first step of a possible catalytic cycle should
consist of a reaction with the chosen alcohol. After a potential
oxidation reaction, O2 could serve to regenerate the active
species (for an initial O2 activation, some d-electrons would
be required: d0 complexes mostly do not activate O2). As a
reference, a system was chosen that has been shown to
catalyze the aerobic oxidation of 1-phenyl-1-propargyl
alcohol:27 OV(acac)2, which probably belongs to the category
of complexes that activates O2 first.
Figure 5. Substrate oxidation with O2 catalyzed by complexes 1-4. The
numbers refer to TOFs (h-1).
version could be observed for any of them, neither with
complexes 1-4 nor with the reference system. In case of
1-hexanol, this was not too surprising, as it represents a
primary alcohol, the oxidation of which is difficult with
vanadium compounds. Nonetheless, for 3, we were able to
observe some activity in the oxidation of other primary
alcohols, namely, in the cases of benzyl alcohol and cinnamyl
alcohol (TOFs of 3.3 and 7.0 h-1, respectively), and
OV(acac)2 performed even better (12.3 and 50.0 h-1), being
capable of oxidizing crotyl alcohol in addition with a TOF
of 9.0 h-1. When 1-phenyl-1-propargylic alcohol and fluo-
renol were employed, the following picture emerged: mono-
nuclear complexes 1 and 2 led to only low conversions.
However, the dinuclear complexes and the reference system
led to much higher TOFs between 15 and 86 h-1. With a
TOF of 86 h-1, the activity of 3 for the oxidation of fluorenol
was even almost 4 times higher than the one of OV(acac)2.
Catalytic amounts of the synthesized complexes were
dissolved in acetonitrile, and molecular sieve was added to
remove the water being produced during the oxidation. After
the addition of a chosen alcohol, the resulting mixture was
heated for 1-3 h at 80 °C under an oxygen atmosphere.
After work up, the turnover frequencies (TOFs) were
determined by NMR spectroscopy. The results are sum-
marized in Table 2. Three aliphatic alcohols were em-
ployed: 1-hexanol, 2-pentanol, and cyclohexanol. No con-
These findings could hint to the fact that two vanadium
centers are advantageous for an efficient catalysis and that
the activity of OV(acac)2 has the formation of a dinuclear
active species in solution as its basis.
(26) Arends, I. W. C. E.; Sheldon, R. A. In Modern Oxidation Methods;
Ba¨ckvall, J.-E., Ed.; Wiley-VCH: Weinheim, Germany, 2004; p 83.
(27) Maeda, Y.; Kakiuchi, N.; Matsumura, S.; Nishimura, T.; Kawamura,
T., Uemura, S. J. Org. Chem. 2002, 67, 6718.
8316 Inorganic Chemistry, Vol. 45, No. 20, 2006

OxoWanadium(V)calixarenes as Oxidation Catalysts
Having found that some of the complexes do in fact show
activity as catalysts for the alcohol oxidation, a nonfunc-
tionalized hydrocarbon was then employed to further exploit
their potential. A substrate with comparatively weak C-H
bonds was chosen, namely, 9,10-dihydroanthracene. Re-
markably, the two dinuclear complexes 3 and 4 again proved
to be active (even though the TOFs were much lower in
comparison to those for the alcohols), whereas the reference
system proved to be inactive. The oxidation products were
identified as 10% (3%) anthracene, 13% (6%) 10H-an-
thracene-9-one, and 7% (6%) 9,10-dihydroynthracene-9,10-
diol. These oxidations probably proceed via anthranol and
dihydroantracenediol by analogy with a corresponding
oxidation using a Ru catalyst.28 However, when fluorene,
whose C-H bonds are somewhat (ca. 23 kJ/mol) stronger,
was chosen as the substrate, none of the compounds
employed showed any activity.
The mechanism of oxidation for the catalysts employed
is still unclear (possibly, there is more than one). As
mentioned above, the first step in the cases of the active VV
catalysts will consist of a reaction with the alcohol, and in
fact, a significant color change can be observed on addition
of the alcohol for the active catalysts. A further color change
can then be noted on heating, together with O2. However,
so far, we have not been able to identify any of the potential
intermediates. More detailed and extensive studies might
provide some insight in future. Nevertheless, the results of
these preliminary investigations put forward some interesting
working hypotheses concerning the prerequisites for suc-
cessful catalysis (e.g., with respect to nuclearity) that will
be tested in subsequent research.
it turned out that it does not even transfer the terminal oxo
ligand to phosphines. A possible reason lies in the negative
charge that has to be accommodated for an OdV3+ unit if it
is surrounded by four additional negatively charged ligands,
and this conclusion could also hold for corresponding surface
species.
Conclusions
The synthesis and characterization of two mononuclear
and two dinuclear oxo vanadium calixarene complexes
(charged as well as neutral) have been achieved. Some of
these complexes proved to be efficient catalysts for the O2
oxidation of activated alcohols. It could be noted that the
two dinuclear complexes were significantly more active than
the two mononuclear compounds. Subsequent studies even
showed that the singly charged mononuclear complex 1
behaves completely inert toward other substrates under
various conditions, whereas the two dinuclear complexes
were capable of oxidizing a nonfunctionalized hydrocarbon
like 9,10-dihydroanthracene, although the TOFs are much
lower in comparison to the ones for the alcohol oxidations.
These findings may have implications on the discussion
which kind of species are active in oxidative dehydrogena-
tions with vanadia catalysts, as they suggest an increase in
activity through the cooperative behavior of two metal
centers. However, a larger number of complexes has to be
synthesized and more detailed mechanistic investigations
have to be performed to support this hypotheses. Especially,
the identification of an oxidation intermediate would be very
helpful. Current work deals with these issues.
Acknowledgment. We are grateful to the Deutsche
Forschungsgemeinschaft, the SFB 546, the Fonds der Che-
mischen Industrie, the BMBF and the Dr. Otto Ro¨hm
Geda¨chtnisstiftung for financial support. We also thank C.
Jankowski for the preparation of starting materials and C.
M. Klemm for performing catalytical tests.
Somewhat irritating was the apparent inertness of complex
1 that, at first sight, had not been expected. Therefore, some
further reactivity studies were performed. First of all,
stoichiometric reactions with tetramethylethylene, dihydroan-
thracene, and alcohols were performed. 1 was then tested as
catalysts for the oxidation of these hydrocarbons with O2 in
the presence of reducing agents such as Zn or Mg, and finally
as catalyst for the photooxidation of benzene and cyclohex-
ene with O2. In all these experiments, 1 behaved inert, and
Supporting Information Available: X-ray crystallographic data
in cif format and plots of the crystal structures showing the 30%
probability level ellipsoids. This material is available free of charge
via the Internet at http://pubs.acs.org.
(28) Kamata, K.; Kasai, J.; Yamaguchi, K.; Mizuno, N. Org. Lett. 2004,
6, 3577.
IC061106J
Inorganic Chemistry, Vol. 45, No. 20, 2006 8317