yield. Davis et al. improved the selectivity for Table 1. Oxidative conversion of HMF with Au supported on various zeolites and
[
a]
metal oxides.
FDCA up to 65% by increasing the NaOH/HMF
[11]
Catalysts
d
Au
Structure
(pore size
Conversion
[%]
Yield [%]
HMFCA FDCA others
ratio from 2 to 20 over Au/TiO2. The reason may
be that the base can facilitate dehydrogenation of
hydroxyl, and reduce the amount of carboxylic acid
adsorbed on the Au surface, thus more active sites
are unoccupied. These recent researches reveal that
the oxidation of HMF to FDCA still exhibit prob- Au/H-MOR[
lems such as instability, ion leaching, and the need Au/Na-ZSM-5-38
of a high ratio of base, which need to be overcome.
Thus, it is still urgent to develop highly active, more
selective, and stable heterogeneous catalysts for
highly efficient realizing the efficient oxidation of
HMF to FDCA.
[
nm]
[
nm])
Au/Mg(OH)
Au/TiO
Au/CeO
2
ꢁ5–7 disorder
>99
>99
>99
96
87
92
>99
30
25
10
6
76
85
73
15
0
1
>99
0
14
9
2
21
15
9
0
3
1
2
10
10
disorder
disorder
2
25
64
85
90
0
b]
ꢁ3–5 channel (0.8)
[
[
b,c]
c]
20
15
1
channel (0.5)
channel (0.5)
cage (1.2)
cage (1.2)
–
Au/Na-ZSM-5-25
Au/HY
HY
no catalyst
1
–
27
24
0
[
a] Reaction conditions: catalyst (0.30 g, Au 1.5 wt%), HMF (0.317 g), H
2
O (4.6 g),
2
NaOH (0.4 g), 0.3 MPa O , 608C, 6 h, molar ratio of Au/HMF=1:110. [b] H-MOR
According to the current research status, con- and ZSM-5= channel-type zeolites. [c] The silica alumina ratio of Na-ZSM-5 was 38
and 25, respectively.
struction of new catalyst comprising extremely
[25,26]
small nanoclusters (<2 nm)
may be a promis-
ing method to break through the bottleneck in the
HMF oxidation reaction and achieve high yields of FDCA.
The small size endows nanoclusters an extremely high activi-
ty, but accompanies with the problem of easy aggregation
and instability which become more serious under severe
conditions. Recently, various protectors were widely em-
lytic performance in HMF oxidation to FDCA at 608C
under 0.3 MPa O in water, Typical experiments were con-
2
ducted over a certain amount of catalyst at 608C under
0.3 MPa O with the addition of NaOH in water. HMF did
2
not convert into FDCA without catalyst or only with the
HY support alone. When gold was supported on typical
metal oxide/hydroxide materials such as TiO , CeO , and
[27,28]
ployed to prevent aggregation, including dendrimers,
[29,30]
[31–33]
functional polymers,
or classic s-type ligands,
such
2
2
as carboxylic acids, thiols, and amines. Unfortunately, these
protectors, typically containing heteroatoms such as S and
N, which remarkably influence or even hamper the catalytic
performance. As widely industrialized catalyst, zeolites are
crystalline aluminosilicates with a diverse morphology and a
rigid microporous structure. We notice that the cage-type Y
zeolites own a typical supercage with 1.2 nm in diameter
and an open aperture of 0.74 nm, which can serve as a natu-
ral nanocage to encapsulate the extremely small nanoclus-
ters and prevent the inside nanoclusters from aggregation
and deactivation. However, this topic has not yet been ex-
tensively addressed.
Mg(OH) , the size of Au NPs is non-uniform, as shown by
2
TEM (see the Supporting Information, Figure S8). The di-
ameter of the Au particles in these catalysts was larger than
5 nm, some were even greater than 10 nm. All of them were
attached to the oxide surface, consisting with the fact that
there is no pore structure in these oxides. These catalysts
gave moderate reaction activity, achieving 76, 85, and 73%
FDCA yields, respectively. Until now, it is still difficult to
synthesize tiny gold nanoclusters, which are uniform and
highly dispersed on oxides surface in less than 1.5 nm, be-
cause these oxides lack a confined structure and thus cannot
restrict the growth of nanoclusters. In our case, even though
the oxide supports enable gold to disperse on the metal
oxide/hydroxide surface (TiO , CeO and Mg(OH) ) through
In the present study, we prepared Au/HY catalysts, in
which Au nanoclusters with 1 nm average size were encap-
sulated within a HY supercage, and applied this catalytic
system on HMF oxidation. The Au/HY catalyst exhibited
highly efficiency in the oxidation of HMF to FDCA. Special
attention has been paid to uncover the function of the su-
percage and the internal hydroxyl groups in the formation
of Au/HY, as well as in the promotion of the activity and
the stability of the catalyst. In our previous study on HMF
oxidation, several vanadium-based catalytic systems were
disclosed, achieving 2,5-diformylfuran (DFF) or maleic an-
2
2
2
charge effect, the obtained Au NPs easily aggregate, result-
ing in large particles.
However, it is remarkable that Au supported on HY zeo-
lite exhibited the top activity in HMF oxidation reaction,
[36]
achieving >99% conversion and >99% FDCA yield.
Contrasted with Au/HY, other zeolite-supported Au cata-
lysts such as Au/H-MOR and Au/Na-ZSM-5, exhibit lower
catalytic performance. For example, when Au was dispersed
in H-MOR zeolite featured with a one-dimensional struc-
ture or Na-ZSM-5 zeolite characterized by a three-dimen-
sional channel structure, HMF was oxidized into HMFCA
rather than the required target FDCA (<3%). The different
reaction activity is related to Au particles size effect; the
[34,35]
hydride as main products rather than FDCA.
Here, the
high efficiency in the oxidation of HMF to FDCA is realized
by Au nanoclusters encapsulated in a HY supercage.
[37–39]
smaller size gives much higher reaction activity.
Results and Discussion
We carefully checked the Au clusters in the zeolite by
TEM. The diameter of gold nanoclusters is calculated by ac-
counting more than 200 clusters. For H-MOR and Na-ZSM-
5 carrier, gold particles have a large size of more than 5 nm
Comparison of different supports: For comparison, Table 1
lists the common texture of various catalysts and their cata-
14216
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 14215 – 14223