G Model
CATTOD-10296; No. of Pages7
C. Megías-Sayago et al. / Catalysis Today xxx (2016) xxx–xxx
Table 1
2
Some recent studies report the base-free oxidation of glucose
studied glucose oxidation over gold supported on structured carbon
and ZrO2 at 110 ◦C and 0.3 MPa P(O2) and reported high initial glu-
cose conversion over Au/C (92.4%) in comparison to zirconia based
catalyst (12.7%), with conversion decrease of approx. 20% after 4
runs. Wang et al. [21] investigated base-free oxidation of glucose
over different CeO2 and ZrO2 supported gold catalysts at 65 ◦C and
0.23 MPa P(O2) and reported activity dependence on gold load-
ing, being the catalysts with lower gold content the most active.
Approximately 30% to 60% decrease of glucose conversion over
Au/CeO2 catalysts after 5 runs were observed. To palliate this effect
the authors proposed a catalyst treatment, calcination and/or base
washing before the reuse cycles, in order to improve the activ-
ity and to avoid fast deactivation. Those treatments suggest that
any change of the support state, such as particle size or acid-base
properties could influence in a great manner the activity towards
gluconic acid formation.
In this context, we report herein the base-free aerobic glu-
cose oxidation study under mild conditions over different gold
supported catalyst as a function of the support nature. Various
gold supported systems are screened, using ꢀ-Al2O3, CeO2 as
simple oxides and CeO2(20%wt)/Al2O3, CeO2(25%wt)/ZrO2, and
CeO2(50%wt)/ZrO2 as mixed oxides combinations. The effect of
support nature, reaction time, temperature and stirring rate on the
catalytic activity and product distribution are evaluated and the
catalyst reusability discussed.
Gold contents (%) for the fresh catalysts.
Au(%)
Au/Al
1.64
Au/Ce
3.94
Au/Ce/Al
Au/Ce25/Zr
2.31
Au/Ce50/Zr
2.39
after comparing the glucose concentration before and after the
reaction, Eq. (1). Owing to the wide variety of products that could
be obtained, selectivity was calculated on the base of the carbon
moles, as described in Eq. (2). Finally, yields were calculated by Eq.
(3). Carbon balance deviation from 100% was always less than 5%.
Glucose − Glucose
[
]
[
]
F
I
Conversion % =
× 100
(1)
(2)
(3)
( )
Glucose
[
]
I
Carbon mol of specific product
Carbon mol of total products
Selectivity % =
× 100
( )
Conversion %
( )
Yield % =
× Selectivity %
( )
( )
100
For the reuse of catalyst, higher amount of glucose (0.2 M, 15 mL)
and gold catalyst were used in order to maintain the same glucose-
to-catalyst ratio during the recycle runs. Between runs the catalyst
was recovered by filtration, and reused under the same reaction
conditions without any further pretreatment.
3. Results and discussion
Gold was deposited (2 wt.% nominal value) by direct anionic
viously by Ivanova et al. [22]. Gold precursor solution (around
10−4 M) was heated to 70 ◦C, and then contacted with the support
and 20 min later with NH3. The final solid was filtered, dried at
100 ◦C overnight and calcined at 300 ◦C during 4 h.
2. Experimental
2.1. Materials and methods
d-(+)-Glucose (anhydrous, 99%) was purchased from Alfa Aesar
and used as received. HAuCl4 (Johnson Matthey) was used as
gold precursor. All supports are commercially available solids
and were used in this study without any previous treatment: ꢀ-
Al2O3 (Sasol, hereinafter Al), CeO2(20%wt)/Al2O3 (Sasol, Ce/Al),
CeO2 (Ce), CeO2(25%wt)/ZrO2 (Ce25/Zr) and CeO2(50%wt)/ZrO2
(Ce50/Zr) (last three solids available from Daiichi Kigenso Kagaku
Kogio Co., Ltd.).
XRD measurements were carried out at room temperature on
Panalitycal X’Pert Pro diffractometer, equipped with Cu anode. All
diffractograms were recorded in the 10–90◦ 2 range, with 0,05◦
step size and 240 s acquisition time.
Gold loadings were determined by X Ray Fluorescence (XRF)
using Panalitycal AXIOS spectrometer with Rh tube of radiation.
Transmission electron microscopy (TEM) observations were
carried out on PHILIPS CM-200.
The products of glucose oxidation were identified and quantified
by HPLC using a Hi-Plex H column (300 × 7.7 mm) and refractive
index detector (Varian 360-LC) and MilliQ water as mobile phase.
Table 1 presents the actual gold loading of the fresh catalysts.
For Au/Al2O3 catalyst, a metal loss of around 18% from the nom-
inal value was detected, due to incomplete gold deposition. On the
other hand, all ceria-containing solids, present experimental values
close or even higher to the expected ones. The later suggests that,
although a complete metal deposition occurs some support loss is
also possible. In the case of the bare ceria support, almost the double
of gold loading is detected suggesting that the use of strongly basic
media during gold deposition could provoke support dissolution. In
addition, the support loss increases with the ceria content increase
within support composition. In the preparation process the amount
of gold precursor is always slightly higher than required to account
for the metal losses; nevertheless, the greater values observed may
be due to incident blends, or also to the fact that the commercial
support could contain other components, which are removed after
calcination.
The diffraction patterns of the prepared catalysts compared to
their corresponding supports are presented in Fig. 1.
The diffraction peaks corresponding to the gold metal phase are,
in general, not observed for the ceria containing catalysts, suggest-
ing an average size of the gold crystallites under the detection limit
of the technique (4 nm). Nevertheless, and despite the lower crys-
tallinity of bare alumina support, weak diffraction, which could be
attributed to gold at 77◦ 2, is observed for Au/Al sample. The later
suggests an average gold particles size slightly superior to 4 nm,
which cannot be properly quantified by using Scherrer equation
because of its low intensity.
It is worth to mention the CeO2 - ZrO2 solid solution formation
(Fig. 1B), independently to the Ce/Zr molar ratio, confirmed by the
diffraction shifts toward higher 2 with the increase of Zr content.
Transmission electron microscopy (TEM) was used to evalu-
ate the average gold particle size for Au/Al, Au/Ce/Al and Au/Ce
2.2. Catalytic test conditions
The catalytic tests were performed in a glass batch reactor
(50 mL) saturated with oxygen at atmospheric pressure (approx-
imate P(O2) of 0.1 MPa) with 5 mL 0.2 M glucose solution and
Glucose/Au molar ratio of 100. In a typical experiment, a 20 mL/min
pure oxygen flux was introduced in the reactor in order to sup-
ply an oxygen rich atmosphere. Then, the reactor was closed and
the mixture stirred at 600 rpm at various times and temperatures
(0 ◦C–120 ◦C temperature range) without base addition. After reac-
tion, 500 Lof sample was taken from the final mixture, diluted in
500 L of MilliQ water and immediately analyzed by HPLC. The con-
version, selectivity, yield and C balance calculations were based on
Please cite this article in press as: C. Megías-Sayago, et al., Gold catalysts screening in base-free aerobic oxidation of glucose to gluconic
Megías-Sayago
