Y. Kuang et al. / Applied Catalysis A: General 423–424 (2012) 52–58
53
Table 1
The temperature program for all analyses is as follows:
Characterizations of pristine, HCl washed, and HNO3 oxidized Co/NSC.
an oven temperature of 70 ◦C for 4 min, with a ramp rate of
10 ◦C/min to 200 ◦C for 30 min. Quantitative analysis was based
on area normalization.
Elemental analysis
C (%)
H (%)
N (%)
O (%)
Co/NSC
wCo/NSC
oCo/NSC
87.31
87.61
64.28
0.43
0.41
0.81
2.09
2.12
1.80
6.39
6.44
33.11
2.2. Preparation of nanoshell carbon-supported cobalt catalysts
Preparation of Co/NSC: 1 g of cobalt phthalocyanine was blended
with 4.156 g of phenol resin in 300 mL of acetone in a glass flask. The
materials were well dispersed by sonication and then acetone was
removed by evaporation. After drying the residue under reduced
pressure, the blend was heated to 800 ◦C under N2 flow over 1 h
and kept at 800 ◦C for 4 h. The obtained material was pulverized by
ball milling to obtain small particles and was heat treated again.
Preparation of HCl-washed Co/NSC (wCo/NSC): 0.5 g of Co/NSC
was dispersed in 200 mL of 37% HCl solution and was vigorously
stirred overnight. After filtration, the carbon material was washed
to neutral with deionized water. Such procedures were repeated
twice. The resulted carbon was then dried under vacuum for 20 h.
Preparation of HNO3-oxidized Co/NSC (oCo/NSC): 0.2 g of
Co/NSC was heated under reflux in 100 mL of 40% HNO3 solution
for 12 h. Then the carbon material was filtered and rinsed to neutral
with deionized water. The resulted oCo/NSC was dried under vac-
uum for 20 h. HNO3-oxidized activated carbon (oAC) was prepared
likewise.
XPS
C (%)
N (%)
O (%)
Co (%)
Co/NSC
wCo/NSC
oCo/NSC
90.45
85.92
75.42
1.04
0.97
1.93
8.30
13.11
22.65
0.20
0
0
EPMA
C (%)
N (%)
O (%)
Co (%)
Co/NSC
wCo/NSC
oCo/NSC
92.35
93.21
78.10
2.21
2.32
2.30
1.43
1.49
19.32
4.01
2.99
0.19
2-octanol was generally used in our condition optimization inves-
in 1,4-dixoane, t-butanol and heptane; however, the solvents con-
taining nitrogen or chlorine atom largely retarded the process.
Nevertheless, all those results are inferior to that obtained under
solvent free conditions in the conversion of 2-octanol. The tem-
perature effect is shown in Fig. 2, generally the conversion of
2-octanol increases with temperature with predetermined amount
of benzaldehyde. However, the selectivities to ketone product are
affected by elevated temperature because of side reactions. There-
fore, 110 ◦C was finally chosen as a trade-off between reaction
rate and selectivity. The BzH loading effect is shown in Table 2.
Upon fixed Co/NSC loading, BzH ranging from 1.25 equivalents to
2 equivalents were used to compare the respective alcohol conver-
lyst, only 40% conversion of 2-octanol was obtained under identical
reaction conditions. The optimized reaction condition was finally
obtained for the quantitative oxidation of 2-octanol as shown in
Table 3, entry 2. At 110 ◦C, in the presence of 5 mg of Co/NSC and
2 eq. of BzH added portion wise, 10 mmol of 2-octanol can be fully
oxidized with excellent selectivity to ketone after 16 h reaction.
Indeed, over 80% of BzH was converted to acid within 6 h; prolonged
reaction time was adopted to ensure the full conversion of BzH.
Without any catalyst, the BzH autoxidation was much slower, and
1).
2.3. General oxidation method
General oxidation method (2-octanol as an example): 5 mg of
Co/NSC, 1.62 mL (10 mmol) of 2-octanol were charged into a test
tube, followed with sonication for 20 s. The tube was then sealed
and evacuated with a vacuum pump, followed by the attachment of
an oxygen balloon. The reaction mixture was heated to 110 ◦C, using
an aluminum block. 2.08 mL (20 mmol) of benzaldehyde was added
portion wise over 4 h (25% per hour). After another 12 h reaction,
the resulted mixture was analyzed with GC–MS.
3. Results and discussion
Nanoshell carbon possesses shell-like carbon structures in the
range of 20–50 nm and can be easily prepared through the pyroly-
sis of a mixture of transition-metal complex and polymer precursor
[11–13]. In the process of pyrolysis, transition-metal is reduced to
nano-particles with sizes less than 40 nm. After pulverization, the
BET surface area of resulted carbon materials is generally between
200 and 400 m2 g−1. Co/NSC used in this study was prepared from a
mixture of Co(II) phthalocyanine (Co(II)Pc) and phenol resin, with
a cobalt content of ca. 4 wt% determined by EA and EPMA (Table 1).
XPS result shows that cobalt only accounts for ca. 0.2 wt% on the sur-
face (up to ca. 10 nm depth), indicating most of the cobalt particles
are in the bulk. TEM images, shown in Fig. 1, reveal the formation
of nanoshell carbon structures and the distribution of Co particles,
whose sizes are less than 20 nm.
3.3. The effect of the supporting material
To get an insight into the effect of nanoshell carbon on catalyst
performance, we treated Co/NSC with HCl wash and HNO3 oxida-
tion to obtain wCo/NSC and oCo/NSC, respectively (see Table 1 for
elemental characterization and Fig. 1 for TEM images). After the
HCl wash, the total cobalt content was decreased by ca. 1 wt% and
XPS detected no signal of cobalt, suggesting that the external or
surface cobalt has been all removed by HCl. Interestingly, after the
catalyst was oxidized with 40% HNO3 under reflux, the bulk cobalt
can reach much deeper than HCl into the carbon material, proba-
bly due to the difference in lipophilicity. Thus the structure feature
of Co/NSC and the effect of HCl wash and HNO3 oxidation can be
simply illustrated as shown in Fig. 3. On the other hand, wCo/NSC
exhibited a slightly reduced catalytic performance (Table 3, entry
3), while the performance of oCo/NSC largely declined (entry 4).
3.2. Optimization of alcohol oxidation conditions
In our preliminary screening of aldehydes, we found that
peracids from aliphatic aldehydes tend to decompose immediately
after formation rather than to react with alcohols in the absence
of ruthenium. In contrast, benzaldehyde (BzH) forms a stabler and
more oxidative peracid that is strong enough for alcohol oxidation
without the help of ruthenium. Therefore, benzaldehyde was cho-
sen as the peracid source in our following studies. The substrate