2
L.-S. Bai et al. / Tetrahedron Letters xxx (2014) xxx–xxx
to prepare rGO. It was characterized by FT-IR, XRD, XPS, TG, and
TEM. FT-IR spectra of GO and rGO-1 are shown in Figure 1. For
GO, the following characteristic peaks were observed: 3347 cmꢀ1
for hydroxyl OAH, 1726 cmꢀ1 for carbonyl C@O, 1619 cmꢀ1 for
aromatic C@C, 1411 cmꢀ1 for carboxyl CAO, 1223 cmꢀ1 for epoxy
CAO, and 1053 cmꢀ1 for CAO. This indicates that GO contains
abundant oxygen-containing functional groups. After reduction
by hydrazine hydrate, the peaks associated to oxygen-functional
groups decreased significantly, and the peak at 1726 cmꢀ1 for
carbonyl C@O totally disappeared. The peak at 1619 cmꢀ1 moved
to 1555 cmꢀ1, indicating formation of aromatic C = C groups.
Figure 2 shows the X-ray diffraction (XRD) patterns of GO and
rGO-1. For GO, a sharp diffraction peak displays at 11.8°, corre-
sponding to an interlayer d-spacing of 0.749 nm as the oxygen-
functional groups located on the graphene sheet. After being
reduced by hydrazine hydrate, no sharp peaks were observed, indi-
cating that most of the oxygen-containing groups were removed
during the reduction process. XPS analysis showed that rGO-1
had a C/O ratio of 7.0, which was much greater than that of GO
(1.2). This also supported the removal of most of the oxygen-con-
taining groups.
Figure 2. XRD patterns of GO, (A) pristine rGO-1, (B) rGO-1 after being used one
time, (C) and rGO-1 after being used five times (D).
TG analysis of rGO-1 is shown in Supplementary data. There
was no apparent mass loss at around 200 °C, indicating that little
amount of oxygen-functional groups remained after reduction of
GO. The morphology of rGO-1 was observed by TEM (see Supple-
mentary data). TEM image showed large aggregated or spread
graphene sheets with numerous wrinkles.
transformation (entry 18, Table 1). The reaction was sluggish in
the absence of rGO-1 (entry 19, Table 1). Thus, the optimum reac-
tion conditions were found as 10 wt% of rGO-1, 30 °C, EtOH as sol-
vent, and under open-air.
The dehydrogenations of various hydrazo compounds (1b–k)
were examined and the results are summarized in Table 2. For
the substrates having electron-donating groups, the reactions took
place readily to give the corresponding azo products in good to
excellent yields (entries 1–2, 4–8). Changing rGO loading to
5 wt % and 2 wt % led to extension of the reaction time and
decrease of the yields (see Supplementary data, Table 1S). Sym-
metric substrates in the reaction were prepared by the reductive
coupling of the corresponding nitro compounds. Among them,
3,30-dimethyl hydrazobenzene was difficult to isolate, so its etha-
nol solution was directly used without purification and the yield
of 2c was based on m-nitrotoluene. In the cases of hydrazo com-
pounds having an electron-withdrawing group, the oxidative
dehydrogenation proceeded more slowly (entries 9 and 10). Exten-
sion of reaction time did not result in the increase of yields. Chang-
ing rGO loading to 5 wt% resulted in a poor yield of 1i even for a
long reaction time under reflux (see Supplementary data, Table 1S).
After dehydrogenation reaction of hydrazobenzene (1a), rGO-1
was recovered by filtering the solid out, washing thoroughly with
ethanol, and drying. Then it was reused in the reaction. At sixth
run, no obvious loss of catalytic activity was observed (Table 3).
There was no significant change in IR spectra (Figure 1) and TG
curves (see Supplementary data) of rGOs before first run, after first
run, and after fifth run.
As reaction temperature was raised from 10 °C to 30 °C, the
reaction time was reduced and the yield of azobenzene (2a)
increased (entries 1–3, Table 1). However, when the temperature
changed from 30 °C to 40 °C, the yield of 2a decreased (entry 4,
Table 1). Furthermore, the amount of catalyst was changed. The
yield of 2a reached to 97% satisfactorily when the loading of
rGO-1 was 10 wt % (entries 5–8, Table 1). Subsequently, different
solvents were screened (entries 9–13, Table 1). The results demon-
strated that ethanol, acetone, and acetonitrile could serve as the
solvent but ethanol was the best choice. Meanwhile, the reaction
was carried out under nitrogen atmosphere. The yield of 2a was
fairly poor, which indicated that oxygen in air could be the final
oxidant (entry 14, Table 1). In addition, other rGOs were prepared
in the presence of NaBH4, NaHSO3, and hot water, which were
marked as rGO-2, rGO-3, and rGO-4, respectively. Their catalytic
activities were much lower than that of rGO-1 (entries 15–17,
Table 1). While graphite power was not able to catalyze this
At this stage it is hard to propose a detailed mechanism of this
reaction. Recently, Loh et al. reported that unpaired electrons in the
holes and at the edges of ba-GO sheets could activate O2 to form
O2Åꢀ and led to aerobic oxidative coupling of benzylamines.9c The
ba-GO in their research was prepared by treatment of GO with base
and acid sequentially. Without regard to the carboxyls at the edges,
the structure of basal plane of ba-GO was very similar with that of
rGO.12 Enlightened by their mechanism study, we envisage that
rGO might activate O2 to form O2Åꢀ similarly, then hydrazo
compound 1 was oxidized by O2Åꢀ to produce azo compound 2
and H2O2 (Scheme 1). Meanwhile, an electron was released and
transferred back to rGO, then rGO structure with unpaired
electrons was recovered. However, pristine GO has less unpaired
electrons at the edges of its sheets,9c so it showed lower catalytic
activity than rGO (entries 3 vs 18, Table 1).
Figure 1. FT-IR of pristine rGO-1, (A) rGO-1 after being used one time, (B) rGO-1
after being used five times, (C) and GO (D).