A. Taketoshi et al. / Applied Catalysis A: General 474 (2014) 257–262
261
activity, Au/ZrO , Au/ND, Au/La O , Au/Al O , Au/CeO , Au/SnO ,
that of a typical active gold catalyst. Accordingly, there is a large
room to improve the catalytic activity of Au/Al O . In Au/KB, the
2
2
3
2
3
2
2
Au/MnO , and Au/Co O (entries 1–9). The second one showed
2
3
4
2
3
decreases in both initial rate of reaction and conversion after 1 h
and is composed of Au/Fe O , Au/SiO , Au/TiO , Au/ZnO, Au/V O5,
diameters of primary particles of KB are 20–50 nm. The mean diam-
eter of gold NPs on KB was 6.5 nm ± 2.9 nm (Fig. 4h).
2
3
2
2
2
Au/NiO, Au/CuO, and Au/KB (entries 11–18). There seems to be no
clear correlations between the kind of metal oxide supports and the
enhancing effect on the catalytic activity for glucose oxidation. It is
likely that synergy between supported gold catalysts and glucose
oxidase may depend on the biological compatibility rather than on
physicochemical properties of the support materials.
3
.3. Decomposition of H O by gold catalysts
2 2
Table 2 shows the decomposition of H O2 on supported gold
2
catalysts and on support materials without gold deposition. All
materials show a certain degree of catalytic activity for H O2
2
It is interesting to note that two carbon materials show opposite
effect on the catalytic activity. Gold NPs deposited on KetjenBlack
having mesopores reduced the initial oxidation rate with glucose
oxidase by half, while gold NPs on NanoDiamond which were non-
porous enhanced the initial rate by 17%. Among supports alone
decomposition which avoids the deactivation of enzyme catalysts.
In the cases of Au/ZrO , Au/ND, Au/Al O , Pt/ZrO and ND, the com-
2
2
3
2
bination with glucose oxidase showed positive effect as expected.
In contrast, in the cases of Au/KB, ZrO , Al O and KB, they showed
2
2
3
negative effect. In particular, Au/KB highly depressed glucose oxi-
dation by glucose oxidase even though the deposition of noble
without gold deposition, ND, ZrO , Al O , and KB, only NanoDi-
2
2
3
amond presented higher conversion of glucose after 1 h reaction
metals increased the rate of H O2 decomposition. The fact means
2
(
entries 19–22).
that the co-presence of KB itself caused the depression of the
enzyme activity (Fig. 1). Therefore the compatibility of support
materials with glucose oxidase is one of the key factors for synergy
Fig. 1 shows that glucose oxidase alone could not attain 100%
conversion because of the deactivation by H O formed during
2
2
reaction. The combination with KB without gold deposition yielded
lower conversions than with Au/KB, suggesting that gold NPs could
facilitate the decomposition of H O . In fact, Table 2 shows that
effect as well as the ability for H O2 decomposition.
2
2
2
3
.4. Effect of H O on glucose oxidation with O2
2 2
deposition of gold NPs or Pt NPs on ZrO , Al O , ND, and KB
2
2
3
increased the rate of H O2 decomposition (entries 1–9). In con-
trast, the combination with ND without gold deposition yielded a
2
Table 3 lists glucose conversion with O2 alone and with O2
and H O . On glucose oxidase, the conversion with O alone was
2
2
2
little higher conversion than glucose oxidase alone. Because H O2
2
higher than with O2 and H O , indicating that the co-presence of
2
2
decomposition on ND was moderately fast similar to KB, it can be
assumed that the co-presence of ND enhances gluconic acid pro-
duction by glucose oxidase. Deposition of gold NPs on ND enabled
glucose oxidase to transform glucose with 100% conversion. This
is probably because Au/ND decomposes H O more rapidly and
H O depressed the oxidation reaction. On Au/ZrO , the conversion
2
2
2
was appreciably increased from 4 to 12%. This may partly explain
why the combination of enzyme with Au/ZrO2 catalyst exhibits
enhanced catalytic activity. The gold catalyst utilizes H O2 pro-
2
2
2
duced by glucose oxidation with the enzyme to gluconic acid and
avoids deactivation of glucose oxidase.
at the same time it decomposes H O2 to avoid the deactivation of
2
Figs. 2 and 3 show that Au/ZrO2 and Au/Al O3 which exhibited
2
enzyme by H O .
2
2
the highest catalytic activity for the aerobic oxidation of glucose in
strong alkali solution were not active at all under neutral conditions
4. Conclusions
(Table 1, entries 23 and 24). However, the combination with glucose
oxidase appreciably increased the catalytic activity both in terms
of initial rate and conversion attained after 1 h reaction though the
combination of their supports alone with glucose oxidase showed
negative effect.
Aerobic oxidation of glucose has been conducted in neutral
aqueous solution at room temperature to produce acid-rich glu-
conate by using a hybrid system of a biological catalyst like an
enzyme and artificial catalysts like supported gold catalysts. Gold
catalysts were chosen because they exhibit higher catalytic activity
at around room temperature than other metal catalysts.
3.2. Characterization of the catalysts
Although gold NPs could not be observed by a high-resolution
1
) Gold NPs supported on
and carbons showed positive synergy by the combination
with glucose oxidase, in particular, Au/ZrO , Au/Al O , and
a variety of base metal oxides
TEM, the diameters of primary particles of ND could be estimated
to be 5–10 nm for Au/ND. Fig. 4a shows a HAADF-STEM image as
well as a distribution of the diameter of gold NPs in Au/ND. By
this technique even tiny gold NPs could be observed because the
contrast was intensified by the second power of the atomic weight
2
2
3
Au/NanoDiamond.
) In the case of Au/ZrO and Au/Al O , the combination with their
2
2
2
3
2
2
supports alone showed negative effect whereas the gold cat-
alysts showed positive synergy caused by the following: they
of the element, namely, [197(Au)/12(C)] = (16.4) . Obviously, gold
NPs were highly and almost homogeneously dispersed. The mean
diameter and standard deviation of gold NPs were calculated to
be 2.4 nm ± 1.4 nm from about 30 images taken by a HAADF-STEM
decompose H O2 produced by glucose oxidase and protect the
2
enzyme from oxidative damages by H O2 and furthermore uti-
2
lize H O2 for glucose oxidation to gluconic acid.
(
Fig. 4b).
2
3
) In the case of Au/ND which is active for H O2 decomposition
The diameters of primary particles of ZrO2 were estimated to
2
but not active for the glucose oxidation with O2 and with O2
be 5–10 nm by HR-TEM. A HAADF-STEM image for Au/ZrO2 which
exhibited the most efficient combination effect is shown in Fig. 4c.
The mean diameter of gold NPs on ZrO2 was calculated to be
and H O , it can only prevent glucose oxidase from the oxida-
2
2
tive deactivation by H O . However, the presence of ND does
2
2
not cause the negative effect to glucose oxidase, in other words,
glucose oxidase is compatible with ND.
4
.2 nm ± 1.5 nm (Fig. 4d). In the case of Pt/ZrO , platinum NPs could
2
not be observed even by a HAADF-STEM. It is probable that plat-
inum NPs were too small and below 2 nm in diameter. The Au/Al O
2
3
catalyst was composed of a mixture of 20–50 nm particles and nee-
dles of Al O . Gold NPs could be found but they were dispersed
Acknowledgments
2
3
with a small population density (Fig. 4e). The gold particles are
1.1 nm ± 6.6 nm in diameter (Fig. 4f). This feature is different from
One of the authors (M. H.) is deeply grateful to Professor
Bernard Delmon for his continuous guidance, instruction, and
1