436
F. P. Silva et al.
oxidation of cyclohexene (Table 1, entry 7), affording very
low substrate conversion and good selectivity to the enone
product, 2-cyclohexen-1-one. Considering that the solid
support Fe3O4@SiO2 has no catalytic activity (Table 1,
entry 5), the higher activity of the supported CoO catalyst,
compared to the non-supported catalyst, can be reasonably
attributed to the good dispersion of the metal oxide over the
support surface, which increases the catalyst surface area.
The non-supported CoO nanoparticles are less stable
toward agglomeration, which cause reduction of the sur-
face area of the catalyst. Two more experiments were
performed with CoO catalyst supported on silica (SiO2-
CoO, Table 1, entry 8), prepared similarly to the magnetic
CoO catalyst (the solid shown in Fig. 4a), and a Co3O4
catalyst obtained after calcining the SiO2-CoO solid at
650 °C for 2 h (SiO2-Co3O4, Table 1, entry 9). The cata-
lytic results show that CoO is the most active cobalt cat-
alyst for the allylic oxidation of cyclohexene in the
conditions studied. Finally, the catalytic activity of the iron
oxide core Fe3O4 was also tested, and some conversion was
achieved (Table 1, entry 10). Therefore, the core shell
structure and the dense layer of silica, in the silica coated
iron oxide support, isolate the iron oxide from the reaction
medium giving non-detectable conversion of cyclohexene
(Table 1, entry 5), while uncoated iron oxide reaches ca.
30% conversion under similar catalytic conditions
(Table 1, entry 10).
Table 2 Oxidation of cyclohexene in different temperatures
Temperature Conversion Selectivity (%)
( °C)
(%)
O
O
O
OH
O
25
0
–
–
–
–
50
16.3
96.5
98.9
–
52.1
76.0
74.4
39.9
3.6
8.0
11.9
12.4
75
8.5
9.3
100
3.9
CoO particles that can be completely collected magneti-
cally. The deactivation experienced by our catalyst is likely
related to common inhibition processes found in oxidation
reactions, such as the adsorption of the oxygenated product
species at the catalyst surfaces and metal particle aggre-
gation. The results in Table 1 also showed that the deac-
tivation of the catalyst brought forth a consequential
selectivity to the allylic oxidation of cyclohexene to
2-cyclohexen-1-one, reaching 100% selectivity in the
fourth run.
Next, we examined the influence of temperature on the
oxidation experiments by varying the temperature from 25
to 100 °C. As shown in Table 2, the substrate conversion
and selectivity is highly dependent on the temperature. The
catalyst is inactive at 25 °C, and higher conversion and
selectivity were achieved above 75 °C.
In another set of experiments (Table 1, entry 2–4), the
same portion of Fe3O4@SiO2-CoO catalyst was reused in
successive oxidations of cyclohexene under similar con-
ditions. After each reaction, the reactor was cooled down to
room temperature and depressurized. The catalyst was
recovered magnetically, by placing a magnet in the reactor
wall, and the liquid was removed with a syringe and ana-
lyzed by GC–MS. The catalyst was dried by a vacuum
pump for 10 min and new portions of substrate were added
and submitted to the same reaction conditions. The recy-
cling procedure could be repeated for up to four successive
times without addition of fresh catalyst. Although high
conversions were achieved for all four runs, it is important
to mention that the catalyst gradually loses activity under
recycling, which can be confirmed by the decrease in
substrate conversion. In order to investigate the deactiva-
tion process we first analyzed the metal content in the
organic phase recovered after the magnetic separation of
the catalyst after each reaction (Table 1, entry 1–4) to
detect any cobalt metal leaching. The results were not
higher than 0.04 ppm Co (detected by ICP-AES) in each
sample, which corresponds to 0.02% of the original mass
of catalyst. Therefore, the loss of activity observed cannot
be related to catalyst mass loss. This can be attributed to
the efficient method used to coat the magnetite particles
and also to the strong interaction of the support and the
4 Conclusions
A new magnetically recoverable CoO-catalyst has been
prepared and characterized by TEM, HRTEM and XPS.
The catalyst showed high activity towards cyclohexene
oxidation and selectivity to the allylic oxidation product,
which increases when the catalyst is partially deactivated.
The catalytic results compared to other cobalt and iron
species show that CoO is the most active for the allylic
oxidation of cyclohexene in the conditions studied. The
magnetic property of the catalyst allowed a facile and fast
product separation with negligible Co leaching into the
products. The reaction method used has a strong environ-
mental appeal as molecular oxygen is the only chemical
used as oxidant, and the catalyst is easily recovered.
Acknowledgments The authors are grateful to the Brazilian agen-
cies FAPESP and CNPq, and to the INCT-Catalise for financial
support. The electron microscopy work has been performed with the
JEM-3010 ARP microscope at LNLS—Brazilian Synchrotron Light
Laboratory/MCT (Campinas, Brazil).
123