and other oxidative processes (i.e., no ethers or sulfoxides)
and have a sufficiently high boiling point. The use of aprotic
polar solvents such as acetonitrile, DMF, and sulfolane re-
sulted in poor conversions (Table S1 in the Supporting Infor-
mation, entries 3–5), which can be explained by heteroatom
coordination by the solvent to the Pd centers, resulting in
[8d]
poisoning of the catalyst.
Interestingly, water could be
used as a solvent with moderate results when running the re-
action under an oxygen atmosphere. However, when per-
forming the reaction under an air atmosphere, a significantly
lower conversion was obtained (Table S1 in the Supporting
Information, entries 1 and 2). Aromatic solvents, such as tol-
uene, p-xylene, or trifluorotoluene (TFT) were found to be
the most suitable solvents for the aerobic oxidation of 1a
[14]
(
Table S1 in the Supporting Information, entries 6–8).
TFT was the best solvent when the reaction was run at
008C under an O2 atmosphere, giving acetophenone in
1
quantitative yield after 1 h (Table S1 in the Supporting In-
formation, entry 6). However, when the atmosphere was
changed to air only 92% conversion of the starting material
was obtained after 1 h. Although the aerobic oxidation ap-
peared to be slightly less efficient in p-xylene at 1008C, it
was possible to raise the temperature by 108C compared
with TFT (boiling point 1028C) and obtain the correspond-
ing carbonyl product in excellent yield in 1 h under an air
atmosphere (Table S1 in the Supporting Information,
entry 13). After several screening experiments with different
combination of solvents, catalyst loadings, and temperatures,
we found the most efficient and practical reaction conditions
to be 1a (0.8 mmol), Pd nanocatalyst (1.5 mol% of Pd), air
Figure 1. a) TEM image of well-dispersed Pd nanocatalyst, with Pd nano-
crystals of 1-2 nm. b) XPS of the Pd -AmP-MCF catalyst.
0
The large width of the Pd 3d core level suggests that the
II
Pd atoms bind to the aminopropyl groups in several ways.
0
The XPS of the reduced Pd nanocatalyst (Pd -AmP-
[15]
MCF) is given in Figure 1b. The Pd 3d core level was decon-
voluted into two spin-orbit split components found at 335.5
and 338.2 eV binding energy.
(1 atm) in p-xylene (2 mL) at 1108C.
Notably, the reaction in toluene is significantly slower
than that in p-xylene, mesitylene, and tert-butylbenzene,
which is unexpected because there is only a slight difference
in electronic properties (see the Supporting Information,
Figure S6). The origin of this solvent effect is not clear and
some possible explanations are considered in the Supporting
Information (p. S10, below Figure S6).
To study the scope of the Pd nanocatalyst catalytic
system, a variety of primary and secondary alcohols were
tested under the reaction conditions described above for the
aerobic oxidation (Table 1). The catalytic system demon-
strated a greater reactivity for benzylic and allylic alcohols
than for aliphatic ones.
From the results with the benzylic substrates (Table 1, en-
tries 1–5, 7 and 8, 10–13), we found that the overall presence
of substituents on the phenyl ring affects the course of the
reaction. The steric bulk most likely reduces the ability of
the substrate to coordinate to the nano-palladium, having a
negative influence on the rate of the reaction. When com-
paring electron-donating and electron-withdrawing groups
we found that electron-donating groups such as a methoxy
group in the para-position, had a beneficial effect, compared
with an electron-withdrawing group such as fluoro, which
leads to a slower reaction. This is in line with the redox po-
tentials and the fact that it is easier to oxidize an electron-
rich alcohol than an electron deficient one.
0
We assign the component at 335.5 eV to Pd , a value
slightly higher than literature references for clean Pd metal,
[13]
in which the Pd 3d core level is found at about 334.6 eV.
This shift in the binding energy scale by about 0.9 eV allows
II
us to identify the component at 338.2 eV as Pd . Thus, this
II
value for the Pd component corresponds well with the
II
II
binding energy found for Pd in the Pd -Amp-MCF sample
in Figure S4 (in the Supporting Information), given a shifted
energy scale. This shift is most likely to be due to the charg-
ing effect caused be the silica spheres with their highly insu-
lating electrical properties. Both samples were found to be
charging in the XPS measurements and the silica core levels
Si 2p and O 1s that flank the Pd 3d core level, were used to
compensate for the shifted binding energy scale. However,
II
Figure 1b shows that the major part of the Pd has been re-
0
II
duced to Pd and that the Pd component accounts for
<
10% of the total intensity.
Here, we investigate the application of the Pd nanocata-
lyst in the aerobic oxidation of alcohols under molecular
oxygen and air. 1-Phenylethanol (1a) was used as a model
substrate to study the effects of catalyst loading, reaction
temperature, and solvent. A suitable solvent for the aerobic
oxidation should be inert under the reaction conditions (i.e.,
no alcohols or ketones), be resistant to peroxide formation
&
2
&
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!