1028
J IRAN CHEM SOC (2012) 9:1021–1031
did not occur and the starting nitro compound was recov-
ered unchanged. These findings confirm that the type of
alcohol plays an important role in this reaction. Propan-2-ol
was the optimal alcohol for the reduction of nitro
compounds.
that the corresponding amines were obtained under both
conditions. The difference in the reaction times shows that
the heating mechanism is different in the two methods.
The reduction of several substrates was also investigated
over a bulk BiFeO3 catalyst under optimized conditions.
This sample with an average particle size of 15 lm and a
BET surface area of 4.25 m2/g was prepared by the classic
solid-state reaction of Bi2O3 and Fe2O3 at 900 °C. The
results in Table 4 clearly indicate that the nanosized
BiFeO3 is much more active than the bulk BiFeO3 sample.
The BiFeO3 nanoparticles have a large surface-to-volume
ratio and consequently exhibit increased surface activity as
compared to the bulk sample. It is well known that a higher
specific surface area of nanomaterials provides much
higher density of active sites as well as greater contact area
between reactants and catalyst to achieve higher efficiency.
The recyclability and stability of the catalyst were also
investigated. For this purpose, the catalyst was separated
via magnetic separation at the end of the reduction of
nitrobenzene, dried and activated at 200 °C for 1 h. The
recovered catalyst was reused in the next run under the
same conditions. The results in Table 5 indicate that there
is no appreciable difference in yield after four runs. The
recovered BiFeO3 after the fourth run was characterized by
FT–IR, XRD and magnetic measurements (VSM). From
the comparison of Fig. 8 with Figs. 2, 3 and 6, it was
concluded that the structure and magnetic property of the
recycled BiFeO3 catalyst did not show observable changes
under the reaction conditions and were not affected by the
reactants. This could be attributed to the high stability of
the perovskite-type structure of BiFeO3 nanoparticles. On
the other hand, no detectable leaching of Fe was observed
in the first or the fourth run of the reaction by ICP-AES
analysis. Also, in an experiment when the catalyst was
separated from the reaction mixture, a short time (10 min)
after beginning microwave irradiation and the reaction
filtrate was further irradiated, no extra formation of amine
was observed via GC–MS analysis, even after 20 min.
These findings also confirm that the reaction catalyzed by
BiFeO3 nanoparticles is heterogeneous in nature.
To explore the potential of this catalytic system, the
reduction of a variety of aromatic nitro compounds was
studied under optimized reaction conditions. The results in
Table 3 show that aromatic nitro compounds containing
various electron donating/withdrawing groups were con-
verted to the corresponding amines in excellent yields
([90 %) within very short reaction times of 7–22 min. As
seen in Table 3, electron withdrawing/donating groups do
not have a significant influence on the reaction times and
yields. In all cases, amines were found to be the only
product of the reactions and the usual side products of nitro
reduction such as azoxy, azo and hydrazo compounds were
not observed in the final product. The present method was
highly chemoselective in the presence of sensitive func-
tional groups. As we can see, hydrogenlyzable or reducible
groups such as halogens, -OH, -NH2, -OCH3, -CN, -COMe,
-CHO, -COOH and -COOEt were not affected by the
reaction conditions. Halogenated nitro compounds were
reduced to the amino compounds without losing halides
(Table 3, entries 6–11). The catalyst also shows promise
for the regioselective reduction of dinitro compounds, such
that selective reduction of one nitro group in the presence
of the other was accomplished (Table 3, entries 16–18).
Nitro compounds containing hydroxyl, cyano and carbonyl
groups were also reduced to the corresponding amino
compounds and the reduction of these functional group and
other carbonyl groups was not observed (Table 3, entries
19–24). Furthermore, this reduction was also successfully
carried out on bulkier molecules such as 1-naphthylamine
with high yield (Table 3, entry 25). In some cases, the
activity was significantly influenced by the position of the
substituents on the aromatic ring. For example, the pres-
ence of a methyl/halide/amine group, ortho or meta to the
nitro group increased the reaction time to a larger extent
than at the para position probably due to steric hindrance
effects. This new reduction system was easily scaled up
and used for the synthesis of aniline at a several-gram
scale. The reduction of nitrobenzene into aniline was per-
formed on a 50-mmol scale. The yield of product was
slightly lower compared with the 5-mmol scale (Table 3,
entries 1 and 26).
The reaction was then studied in the presence of other
nanosized perovkite-type and simple metal oxides. The
results are presented in Table 6. As can be seen, Fe(III)-
containing ferromagnetic oxides such as LaFeO3, SmFeO3,
Fe2O3 and Fe3O4 were almost equally effective in the
reaction (Table 6, entries 1–5), while LaAlO3, LaCoO3,
LaCrO3 and LaMnO3 were not effective (Table 6, entries
6–10). This finding confirms that the ferromagnetic prop-
erties of the catalyst play a vital role in the reaction. From
these observations, we can suggest that the enhancement in
the reduction rate of the aromatic compounds in the pres-
ence of ferromagnetic oxides is related to the phenomenon
of ‘‘nonequilibrium local heating’’ reported in the literature
The reduction of nitro substrates was also conducted
using conventional heating under the same conditions. The
results were compared with the microwave method in
Table 3. As expected, conventional heating required a
longer reaction time for the maximum conversion. From
this table, it is evident that microwave heating dramatically
accelerated the reaction rate. It is, however, noteworthy
123