Synthesis of MCM-41. This was synthesised using the
method of Rey et al.16 5 g Degussa Aerosil 200 silica was
slowly added to a solution of 21.3 g hexadecyl trimethyl
ammonium bromide that had been thoroughly mixed and
stirred with 28.75 ml water and 7.90 ml tetramethyl ammon-
ium hydroxide (25 wt.% solution in water). Once all the silica
was added, the resulting gel was stirred vigorously for at least
1 h, before being transferred to a polypropylene bottle, sealed
and placed in an oven at 90 ¡C for 24 h. The solidiÐed mixture
was then mixed thoroughly with distilled water, Ðltered o†
and washed with copious amounts of water to remove as
much surfactant as possible. The solid was then calcined by
Ðrst heating under nitrogen to 550 ¡C at a rate of 5 ¡C min~1
and holding the temperature at 550 ¡C for 1 h. The gas was
then changed to dry oxygen and the temperature held con-
stant for a further 6 h before cooling.
Ðtting procedure was carried out using the data without
Fourier Ðltering and the curved wave theory and multiple
scattering approach were adopted for the detailed extended
X-ray absorption Ðne structure (EXAFS) data analysis.
Hydrophobicity measurements.9 The hydrophobicity
index,19 HI, is the ratio of the masses of competitively
adsorbed octane and water. The higher the HI the more
hydrophobic the catalyst. In order to determine the HI for
these catalysts, the material (150 mg) was heated, prior to the
measurement to 250 ¡C in a constant stream of Ar. For the
adsorption experiment a stream of Ar, kept at a constant tem-
perature of 30 ¡C, was saturated with equal masses of octane
and water. The unadsorbed gases were measured by gas chro-
matography (GC) over a period of several h in order to deter-
mine the adsorbed quantities of water and octane.
Synthesis of V CPMCM-41 and V CPMCM-41. 2 g of cal-
2
4
Argon physisorption. These measurements were collected on
an Omnisorb 360 (Coulter) apparatus. The samples were out-
gassed, prior to the measurement, at 523 K and 10~4 mbar for
12 h. Adsorption isotherms were recorded using a static volu-
metric technique with a starting pressure of 5 Torr, which was
increased by a factor of 1.05 for every dosage step. Ar was
adsorbed at a temperature of 87.4 K. The pore-size distribu-
tions were calculated for the microporous samples using the
HorvathÈKawazoe model while, for the mesoporous materials
the BarrettÈJoynerÈHalenda theory was applied.20
cined MCM-41 was dehydrated under vacuum for 2 h at
200 ¡C and allowed to cool to 40 ¡C. The solid was then mixed
thoroughly with 75 ml anhydrous chloroform under an argon
atmosphere to which 0.17 g (or 0.34 g for the 4% catalyst)
VCp Cl was subsequently added. After 1 h, 1 ml tri-
2
2
ethylamine was added, and the mixture stirred under argon at
40 ¡C for 16 h. The pale-brown coloured solid was Ðltered o†
and washed with 250 ml chloroform. The calcination pro-
cedure involved heating the sample under nitrogen to 550 ¡C
at a rate of 5 ¡C min~1 and, after 30 min, changing the gas to
oxygen and holding the temperature at 550 ¡C for a further 3
h. The active catalyst so formed was white when fresh from
the oven, but gradually changed colour to yellow/orange on
exposure to the atmosphere. We note that a similar procedure
for synthesising a comparable catalyst has been reported.17
Catalysis
Cyclohexane oxidation reactions. The catalysts were heated
to a temperature of 400 ¡C for 5 h in a constant stream of Ar
to remove all traces of water. A 3 M solution of TBHP in
nonane (2 ml) was mixed with cyclohexane (6.48 ml) giving a
cyclohexane to TBHP ratio of 6 : 1. The mixture was degassed
before use by bubbling Ar through it overnight. 20 mg of the
catalyst and the reaction mixture were added to a glass micro-
reactor equipped with a magnetic stirrer, and was heated to
70 ¡C. Conversion was followed by GC.
Synthesis of V CMeMCM-41. 0.5 g V CMCM-41 (2%) was
2
2
dehydrated under vacuum for 2 h at 200 ¡C. Once cooled to
40 ¡C, a solution of 0.075 ml chlorotrimethylsilane (Me SiCl)
3
in 50 ml anhydrous chloroform was added, under an argon
atmosphere, and stirred at 40 ¡C for 16 h. The remaining green
solid was separated by Ðltration and washed with 250 ml chlo-
roform.
Cyclohexene oxidation reactions. These were performed
using 2 ml cyclohexene, 1 ml TBHP (80% solution in di-tert-
butyl peroxide), 6.5 ml acetonitrile, 0.5 ml mesitylene (as an
internal GC standard) and 75 mg catalyst. The reactions were
carried out at 30 ¡C under an argon atmosphere, using
undried reagents and catalyst. Since it has been reported that
vanadium may leach into solution, even under water-free con-
ditions, causing homogeneous rather than heterogeneous
catalysis,21,22 great care was taken to investigate the nature of
our catalytic systems. To prove the heterogeneous character of
the reactions, the solid catalyst was either centrifuged and
separated (cyclohexane oxidation reactions) or Ðltered o†
(cyclohexene oxidation reactions) after 60 min. No further
activity was observed in the remaining liquid showing that the
results obtained are not due to species leached into solution.
In addition, the catalyst was heated with one of the com-
ponents (i.e. either neat cyclohexane or neat tert-butyl
hydroperoxide) for 3 h, after which the solid catalyst was
removed by centrifugation. The missing component was then
added to the solution and the conversion of the homogeneous
mixture followed by GC. No catalysis was observed, and
hence the catalytic reactions are truly heterogeneous.
Characterisation
X-Ray di†raction. All the MCM-41 based catalysts were
characterised by XRD (Cu-Ka) employing a Siemens D-500
di†ractometer. They all showed the peaks characteristic of the
hexagonal form of MCM-41, with no evidence for extra-
framework formation of vanadium oxide or vanadate phases.
IR spectroscopy. In situ IR studies were performed using a
Perkin Elmer 1725X spectrometer employing a previously
reported in situ cell.18
XAS. In situ V K-edge XAS were recorded on station 8.1 at
the CLRC Daresbury Laboratory Synchrotron Radiation
Source, operating with a beam energy of 2 GeV and an
average current of ca. 200 mA. A double-crystal Si(111) mono-
chromator was employed, o†set to 50% of maximum beam
intensity for harmonic rejection. In a typical experiment 80
mg of the catalyst was pelletised and loaded into the in situ
cell described in an earlier publication.18 The catalysts were
calcined in pure oxygen at 550 ¡C prior to the room-
temperature data collection. Spectra were obtained in Ñuores-
cence mode using
a Canberra 13 element solid state
Ñuorescence detector. Four spectra were collected and aver-
aged for each sample. EXCALIB (to add all the data sets and
convert the raw data to energy vs. absorption coefficient),
EXBROOK (for pre- and post-edge background subtraction)
and EXCURV92 (for detailed curve Ðtting analysis) suite of
programs were used to analyse the XAS. X-Ray absorption
near-edge structure (XANES) data presented in this work
were obtained using the EXBROOK program. The curve-
Results and discussion
The mesoporous silica, MCM-41, was modiÐed with vana-
dium via a vanadocene dichloride precursor, two distinct
loadings being used with V : Si ratios of 2 : 100 and 4 : 100. (By
analogy with our earlier work12 on Ti-loaded MCM-41, using
titanocene dichloride, these materials are designated
3178
J. Chem. Soc., Faraday T rans., 1998, 94, 3177È3182