Alumina-Catalyzed Epoxidation with Hydrogen Peroxide
FULL PAPERS
higher activity is due to the amount and/or acidity of
surface hydroxy groups which we are presently inves-
tigating.
Experimental Section
Synthesis of the Catalysts
The commercial alumina Rc was purchased from Fluka (type
507C, neutral, 100 125 mesh). The surface area, obtained by
BET (Micrometrics ASAP 2010 with nitrogen as probe
molecule), was 195 m2/g, and the average pore diameter was
5.1 nm. Sol-gel alumina R1 was prepared by mixing of
aluminum isopropoxide (9.39 g, 46 mmol) with nitric acid
Figure 3. Conversion of (S)-limonene catalyzed by different
aluminas.
À
1
(8 mol L , 2.2 mL, 17.6 mmol) and water (3.2 mL, 0.18 mol) at
758C for 30 min. No homogeneous solution was obtained. The
sol was dried for 3 days at open air. The gel was treated under a
flow of synthetic air at 1008C, 2008C, and 4008C for 24 h each.
The alumina obtained was non-porous with a surface area of
180 m2/g. Sol-gel alumina R2 was prepared as reported by
Buelna and Lin,[5] by dissolving aluminum isopropoxide
(9.12 g, 45 mmol) in water (22.3 mL, 1.23 mol) at 858C for
cyclohexene, since the commercial alumina already
allows conversions above 90%. The higher activity of R4
in the epoxidation of limonene is probably due to the
same reasons already explained for the epoxidation of
cyclohexene.
À
1
1 h. Nitric acid (8 mol L , 0.39 mL, 3.1 mmol) was added and
the sol refluxed for 12 h. After distilling off the water, it was
dried at 358C for 48 h. The gel was calcined as described before.
The alumina obtained was porous (pore diameter of 3.4 nm)
with a surface area of 293 m2/g. Sol-gel alumina R3 was
obtained using a recipe of Wang et al.[6] by dissolving aluminum
isopropoxide (10.2 g, 50 mmol) in 2-propanol (15 mL,
0.2 mol). A solution of oxalic acid (0.7 g, 5.6 mmol) in water
(7.5 mL, 0.41 mol) was added and the mixture refluxed for 3 h.
The sol was dried at 708C and then calcined as described
before. The alumina obtained was non-porous with a surface
area of 290 m2/g. Sol-gel alumina R4 was prepared using the
same procedure of alumina R3, but substituting aluminum
isopropoxide by aluminum sec-butoxide (10.0 g, 40.6 mmol).
The sol obtained was dried and calcined as described for R3.
The alumina obtained was non-porous with a surface area of
287 m2/g.
In our previous publications[2,3] we already mentioned
that some molecular oxygen is formed during the
catalytic reactions. Small amounts of water retard the
decomposition of hydrogen peroxide,[3] however, the
epoxidation requires an excess of hydrogen peroxide. In
order to understand when and how much molecular
oxygen is formed, we monitored its formation during the
epoxidation of limonene in the presence of commercial
alumina Rc. The profile of the molecular oxygen
formation is the same as the profile of the conversion
of limonene. Molecular oxygen is produced constantly
during the reaction, reaching a maximum of 4.2 mmol
after 300 min. This means that 42% of the active oxygen
is consumed in the production of molecular oxygen,
while the rest is used for the epoxidation of the
limonene. It should be mentioned that in this reaction
approximately 14% of diepoxide 4 is formed, explaining
why more than 50% of the active oxygen are incorpo-
rated in the epoxidation products.
Catalytic Reactions
Solutions of anhydrous hydrogen peroxide in ethyl acetate
À
were prepared by the Dean Stark method as described
before[2,3] and contained approximately 7 mmol of hydrogen
peroxide per gram of ethyl acetate. The other reagents were
used without pre-treatment.
Conclusion
A mixture of the alkene (10 mmol, Fluka), hydrogen
peroxide (20 mmol), di-n-butyl ether (5 mmol, internal stand-
ard) and ethyl acetate (10 mL) was heated under reflux with
magnetic stirring for 1 h. A sample was taken for GC analysis
(blank) and the reaction started by addition of the alumina
(0.50 g). Aliquots of the reaction mixture were taken after the
indicated reaction times for a total of 9 h. The samples were
analyzed using a Hewlett-Packard HP 5890 Series II gas
chromatograph equipped with an HP 5 capillary column (25 m
 0.2 mm  0.33 mm film thickness) and a flame ionization
detector (FID). Products were quantified using calibration
Alumina can be conveniently recycled for three times in
the epoxidation of alkenes with hydrogen peroxide/
ethyl acetate, however, it strongly looses its activity after
the fourth reaction. The reason for this deactivation is
not yet understood. The sol-gel alumina synthesized in
this work by using oxalic acid and aluminum sec-
butoxide in the synthesis gel is more active than the
commercial alumina. The surface area and the non-
porosity of this alumina were not different from the
other aluminas tested. We, therefore, believe that the curves obtained with standard solutions. Selectivities are
Adv. Synth. Catal. 2002, 344, 911 914
913