Amine-stabilized small gold nanoparticles supported on AlSBA-15 as effective catalysts for aerobic glucose oxidation

Article abstract of DOI:10.1016/j.apcata.2014.01.024

The preparation of gold nanoparticles (AuNPs) supported on the alumina- and amine-modified silica SBA-15 and systematic studies of their structural properties and behaviour in glucose oxidation are presented. The AuNPs were deposited using either direct reduction or colloidal deposition method. Catalysts were characterized by nitrogen adsorption, XRD, XPS, STEM, HRTEM, ²⁷Al MAS NMR, ¹H NMR, DLS, HPLC, ICP MS and thermogravimetry to determine structure/activity relationship in the aerobic glucose oxidation. The highest catalytic activity ca. 800 mmol _GA/g_Au min was achieved for AuNPs prepared by a direct reduction method using SBA-15 first modified with alumina (1 wt.%) and then with aminopropyl groups serving as strong adsorption sites. It was notably larger than shown by commercial Au/Al₂O₃ and Au/TiO₂ catalysts (380 and 340 mmol_GA/g_Au min, respectively). Owing to excellent and stable dispersion of AuNPs and their small sizes <6 nm, the catalysts showed very good stability, demonstrated in 7 reaction cycles. The impact of specific functionalities on the size, dispersion and stability of AuNPs was discussed and high selectivity towards gluconic acid formation was demonstrated.

Full text of DOI:10.1016/j.apcata.2014.01.024

Applied Catalysis A: General 475 (2014) 203–210
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Amine-stabilized small gold nanoparticles supported on AlSBA-15 as
effective catalysts for aerobic glucose oxidation
Klaudia Odrozeka, Katarzyna Mareszb, Agnieszka Koreniukb, Krystian Prusikc,
Julita Mrowiec-Białon´ a,b,∗
a Department of Chemical Engineering and Process Design, Faculty of Chemistry, Silesian University of Technology, Strzody 7, 44-100 Gliwice, Poland
b Institute of Chemical Engineering Polish Academy of Sciences, Bałtycka 5, 44-100 Gliwice, Poland
c Institute of Materials Science, Silesian University, 75 Pułku Piechoty 1A, 41-500 Chorzów, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
The preparation of gold nanoparticles (AuNPs) supported on the alumina- and amine-modified silica SBA-
15 and systematic studies of their structural properties and behaviour in glucose oxidation are presented.
The AuNPs were deposited using either direct reduction or colloidal deposition method. Catalysts were
characterized by nitrogen adsorption, XRD, XPS, STEM, HRTEM, 27Al MAS NMR, 1H NMR, DLS, HPLC, ICP MS
and thermogravimetry to determine structure/activity relationship in the aerobic glucose oxidation. The
highest catalytic activity ca. 800 mmolGA/gAu min was achieved for AuNPs prepared by a direct reduction
method using SBA-15 first modified with alumina (1 wt.%) and then with aminopropyl groups serving as
strong adsorption sites. It was notably larger than shown by commercial Au/Al2O3 and Au/TiO2 catalysts
(380 and 340 mmolGA/gAu min, respectively). Owing to excellent and stable dispersion of AuNPs and their
small sizes <6 nm, the catalysts showed very good stability, demonstrated in 7 reaction cycles. The impact
of specific functionalities on the size, dispersion and stability of AuNPs was discussed and high selectivity
towards gluconic acid formation was demonstrated.
Received 4 October 2013
Received in revised form
20 December 2013
Accepted 10 January 2014
Available online 21 January 2014
Keywords:
Gold nanoparticles
Aerobic glucose oxidation
Mesostructured materials
Amine and alumina functionalization
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
appeared to be their direct synthesis and entrapment in a con-
fined space of mesoporous silica (MPS), cf. more recent excellent
Over two decades ago a breakthrough report of Haruta et al.
[1] on the surprisingly high catalytic activity of gold nanoparti-
cles (AuNPs) triggered a worldwide wave of intensive studies of
supported AuNPs catalysts [2–6]. Within a few years these studies
demonstrated the huge potentials of AuNPs to catalyse a number
of very important reactions under mild conditions and often with
quite exceptional selectivities [7,8], provided that their sizes were
invariably smaller than about 5–6 nm [9,10]. However, under liq-
uid reaction conditions the stabilization of minute gold particles
appeared to be a major challenge. The bonding of AuNPs to the oxide
surfaces was typically too weak to preclude their migration over
the surface, ending up with attachment to a growing cluster. These
weak bonding forces are clearly documented by very low values
of adsorption energies of AuNPs on oxide materials very precisely
determined in recent studies [10]. The most effective approach to
stabilize well dispersed AuNPs on the highly loaded oxide surfaces
review of Gutiérrez et al. [11]. Among various MPSs those of SBA-
15 family, owing to fairly large mesopores and thick/robust walls,
appeared to be the most suitable for this application [12,13], in
particular those additionally doped with alumina and grafted with
amine groups, which acted as strong adsorption sites for the anionic
gold precursors [14–17].
While those catalysts were originally aimed at the catalytic oxi-
dation of CO and cyclohexane, the rapidly growing interest in the
chemical synthesis of gluconic acid from glucose prompted us to
study their performance in this reaction using oxygen as a green
oxidant. The substantial interest in gluconic acid and its derivatives
stems from their very attractive properties: low toxicity, very low
corrosiveness and facile biodegradation, which prop up their exten-
sive application in various process industries [18,19]. This boosts
the demand which cannot easily be met by supply, owing to seri-
ous technical problems of the most common fabrication process
involving Aspergillus niger strains [18,20,21].
Various gold catalysts were already tested in a liquid phase
glucose oxidation: AuNPs supported on carbon [22–24], gold col-
loids in a “naked” form [25,26] and also deposited on metal oxides
[27–32]. Selectivity featured by those catalysts was very good,
ca. 99%. However, gold colloids and carbon-supported catalysts
∗
Corresponding author at: Department of Chemical Engineering and Process
Design, Faculty of Chemistry, Silesian University of Technology, Strzody 7, 44-100
Gliwice, Poland. Tel.: +48 32 2310811; fax: +48 32 2310318.
E-mail address: jmrowiec@polsl.pl (J. Mrowiec-Białon´ ).
http://dx.doi.org/10.1016/j.apcata.2014.01.024
0926-860X/© 2014 Elsevier B.V. All rights reserved.

204
K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
failed in long term stability tests, in contrast to those attached to
metal oxides which performed significantly better [30]. But AuNPs
supported on amine [12,15,16], alumina [9] or titania [30,33,34]
modified MPSs were tested mainly in CO oxidation and we are
familiar with only one study in which AuNPs supported on siliceous
SBA-15 were applied to promote the glucose oxidation [35]. For
this reason we decided to carry out more systematic studies of ami-
nated AlSBA-15 as supports for gold NPs to obtain glucose oxidation
catalysts with AuNPs deposited either by a colloidal deposition
(CD), as proposed in [36,37], or using a direct reduction (DR) of
pre-adsorbed HAuCl4 with NaBH4, as proposed by Ishida et al. [38].
An extensive characterization (TEM, XRD, XPS, nitrogen adsorp-
tion, 27Al NMR MAS, 1H NMR, ICP MS and thermogravimetry) of the
synthesized hybrids was carried out to determine their physico-
chemical properties and the size of AuNPs. The catalytic properties
were examined in a batch slurry oxidation of glucose using oxy-
gen as oxidant and compared with the commercial Au/TiO2 and
Au/Al2O3 (AUROlite) catalysts.
The nominal gold loading was 1 wt.% in both methods. The target
samples were designated as: Au(Y)NH2XAlS, where X represents
the nominal content of aluminium (1 or 10 wt.%), Y denotes the
gold deposition method (CD or DR, respectively) and S stands for
siliceous SBA-15 used as support. For comparison, similar samples,
but aluminium- or amine-free, were also prepared and analysed in
the same way.
2.3. Catalyst characterization
Textural characteristics of the materials were determined from
nitrogen adsorption isotherms measured by a Micromeritics ASAP
2020 instrument. The specific surface area (SBET) was calculated in
0.05–0.25 p/po range using Brunauer–Emmet–Teller (BET) method.
The pore size distribution was obtained from the adsorption branch
of isotherm using Barrett–Joyner–Halenda (BJH) method with
Kruk–Jaroniec–Sayari correction [41,42].
27Al MAS NMR spectra were recorded at 104.3 MHz with a
Bruker HP-WB high-speed MAS probe equipped with the 4 mm
zirconia rotor to record the spectra at the spinning speed of 8 kHz.
The amount of amine groups grafted onto silica surface
was determined by thermogravimetric method (Mettler Toledo
STAR851 thermobalance). Samples of ca. 15 mg were heated with
rate of 10 ◦C/min, in a standard 150 ␮l Pt crucibles under air flow
of 50 cm3/min TG, DTG and SDTA curves were recorded.
Scanning transmission electron microscopy (STEM) (Hitachi
HD-2300A) was used to analyse the size of gold nanoparticles as
well as surface morphology of the materials. The actual gold and
aluminium contents were determined by ICP-MS method (Thermo
Electron iCAP 6500 DUO spectrometer).
2. Experimental
2.1. Materials
The chemicals used in the synthesis and modification of
SBA-15 materials were as follows: Pluronic P123 and 3-
aminopropyltriethoxysilane (APTS) were from Sigma–Aldrich.
Chloroauric acid tetrahydrate (HAuCl4·4H2O), sodium borohydride
(NaBH4) and poly(vinyl) alcohol were from POCh (Poland). Alu-
minium butoxide ethylacetoacetate and tetraethoxysilane (99%,
TEOS) were from ABCR.
The size of Au crystallites was estimated from Scherrer equa-
tion. The X-ray diffraction patterns of the samples were collected
by a PAN analytical X’Pert Pro PW 3040/60 diffractometer. Instru-
mental broadening of diffraction lines was determined using LaB6
standard. The profile of main diffraction line from Au crystallites
was separated from the whole pattern assuming its symmetrical
shape.
X-ray photoelectron spectra (XPS) were recorded by a VG
Scientific ESCALAB-210 photoelectron spectrometer using a non-
monochromatized Al K␣ (Eh␯ = 1486.6 eV) radiation (14.5 kV;
20 mA). The binding energy (Eb) scale was calibrated against the
position of C 1s (Eb = 284.6 eV). The analysis chamber was oper-
ated under ultrahigh vacuum with a pressure of 5 × 10−7 Pa. Spectra
were recorded with 0.4 eV resolution and deconvolution was made
using AVANTAGE programme (Thermo Electric ver. 4.84), the back-
ground was fitted by a Shirley’s model.
2.2. Catalyst preparation
The gold catalysts were prepared in several successive steps.
First the siliceous SBA-15 materials were synthesized using an
original procedure [39] and then they were functionalized with
alumina. For that the siliceous SBA-15 was exposed for 5 h to satu-
rated water vapour at RT and then dried at 200 ◦C for 2 h to obtain
the loading of strong silanols of about 3 OH/nm2 [40]. After that,
they were dispersed in a solution of aluminium butoxide ethylace-
toacetate in hexane and stirred for 24 h at 40 ◦C. Following solvent’s
evaporation the solids were calcined at 500 ◦C during 6 h to obtain
AlSBA-15 composites with two alumina loads of 1 or 10 wt.%.
Aminopropyl functionalization of thus obtained AlSBA-15 mate-
rial was carried out using post-synthesis method with APTS as a
silylation reagent. First AlSBA-15, ethanol and APTS (1.5 mmol/g)
were stirred for 2 h at 50 ◦C under reflux, whereupon the solids
were recovered by filtration, washed with ethanol and then dried
overnight at 100 ◦C.
The size of colloidal AuNPs was measured by dynamic light
scattering technique using a Zetasizer S90 instrument (Malvern
Instrument, UK).
2.2.1. Colloidal deposition method (CD)
2.4. Oxidation of glucose
The colloidal solution of gold nanoparticles was prepared by
reducing HAuCl4 (aqueous solution of 10 g/l) with NaBH4 (freshly
prepared 0.1 M solution) under vigorous stirring in the presence of
2% PVA used as stabilizing agent. The colloidal solution was stirred
for 30 min at room temperature, afterwards the modified support
was added and mixing was continued for 2 h. After filtration and
washing with deionised water the solids were dried overnight at
90 ◦C.
The reactions of glucose oxidation were carried out at 60 ◦C in
a thermostated glass reactor, equipped with a mechanical stirrer
operating at 1000 rpm, using 120 cm3 of 1 M glucose solution. Oxy-
gen was bubbled through the suspension with flow rate of 1 l/min
at atmospheric pressure. Concentration of the catalyst was 3.2 g/l.
The pH of the reaction mixture was held constant in the range of
7.5–8.5 by neutralizing gluconic acid with 2 M aqueous solution of
potassium hydroxide. The catalysts selectivity, stability and activity
were determined. For selectivity investigations the concentrations
of glucose, potassium gluconate and potential by-products were
recorded by HPLC and confirmed by 1H NMR methods. In stability
studies, the catalyst was separated by filtration after each oxidation
reaction, washed with water, dried and then reused. The perfor-
mance of catalysts obtained was compared with the commercial
2.2.2. Direct reduction (DR)
An aqueous solution of HAuCl4 (10 g/l) was stirred with the mod-
ified SBA-15 material at room temperature during 30 min, then
0.1 M solution of NaBH4 was added dropwise to reduce the gold
precursor. After 2 h mixing the solids were separated, washed and
finally dried as described above.

K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
205
Table 1
Structural properties of materials.
No.
Sample
SBET
[m2/g]
VP
[cm3/g]
DBJH
[nm]
NH2
[mmol/g]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
SBA-15
10AlS
750
441
430
285
265
258
582
348
342
368
521
411
362
236
49
0.90
0.55
0.53
0.37
0.36
0.35
0.73
0.49
0.51
0.50
0.68
0.68
0.58
0.63
0.33
8.2
7.2
7.2
6.5
6.5
6.7
6.7
7.2
7.6
7.5
8.1
8.6
8.3
9.2
26
–
–
–
1.50
Au(DR)10AlS
NH210AlS
Au(DR)NH210AlS
Au(CD)NH210AlS
1AlS
1.40
b
–
–
1.50
NH21AlS
Au(DR)NH21AlS
Au(CD)NH21AlS
NH2S
Au(DR)NH2S
Au(CD)NH2S
1.32
b
–
1.50
1.29
b
–
a
Au/Al2O3
Au/TiO2
–
–
a
a
Gold content 1 wt.% 0.1.
Not determined because of the presence of PVA adsorbed on the AuNPs.
b
Fig. 2. X-ray diffraction patterns of functionalized and pristine silica.
AUROlite Au/Al2O3 and Au/TiO2 materials (1 wt.% 0.1% of Au),
purchased from Strem Chemicals, Inc., considered as benchmarks.
significant changes mainly in the values of surface area and pore
volume, whereas the attachment of gold NPs had minor effect.
The XRD analysis also confirmed the preservation of the ordered
pore structure after modifications, as can be concluded from three
clearly resolved peaks in the 2ꢀ range of 0.5–2◦, characteristic for
2d-hexagonally packed SBA-15 materials, shown in Fig. 2. The high-
est diffraction band corresponds to the (1 0 0) plane, while the less
intensive signals are characteristic for (1 1 0) and (2 0 0) planes.
These results confirmed that pristine silica SBA-15 had very well
ordered hexagonal structure and that its modification slightly dis-
torted that ordering.
3. Results and discussion
3.1. Structural properties
The SBA-15 silicas were already shown to preserve the original
pore structure even after incorporation of guest molecules. To
check this, all the materials, after each modification step, were
systematically analysed by a small angle X-ray diffraction and
low temperature nitrogen adsorption. Their structural properties:
specific surface area (SBET), pore volume (VP) and pore size (DBJH
)
are listed in Table 1 and Fig. 1 shows the evolution of nitrogen
adsorption/desorption isotherms and pore size distributions in
DR-series samples after each step. As was expected, all samples
display a type IV isotherm with H1 hysteresis, characteristic for
highly ordered mesoporous materials. The functionalized samples
also featured the same pore structure, but their texture parameters
decreased (Table 1) and structure ordering was slightly disrupted
(cf. Fig. 1 inset). A similar trend was also observed in other samples.
Clearly, the functionalization with alumina and amines resulted in
Aluminium was introduced into silica matrix by post-synthesis
method and its nominal content was set up to 1 and 10 wt.%. How-
ever, the ICP-MS analysis showed that the samples contained less
Al, namely 0.9 and 7 wt.%, respectively.
The state of the aluminium in the SBA-15 lattice was inves-
tigated by 27Al MAS NMR spectroscopy. The analyses were
performed for the most active catalysts, i.e. Au(DR)NH21AlS and
Au(DR)NH210AlS. The latter was also examined after four runs in
the glucose oxidation process. As can be seen from the spectra
displayed in Fig. 3 the aluminium was present in two coordina-
tion states: tetrahedral and octahedral, clearly evidenced by the
resonance peaks at ca. 50 and ca. 0 ppm, respectively. However,
in the catalyst with low aluminium content (spectrum 1) it was
solely in a tetrahedral coordination, unlike in the sample with
larger Al content (spectrum 2), where some (ca. 8%) octahedral alu-
minium atoms were detected. After four runs of glucose oxidation
the coordination of aluminium did not change (spectrum 3). The
data obtained from nitrogen adsorption and 27Al NMR analyses are
in good agreement with the results reported earlier [43]. Therefore,
in line with those observations we suppose that nearly all alu-
minium was located in micropores and their nearest vicinity when
its content was low. But when it was large the additional (extra)
aluminium islands were present on the surface of mesopores.
The amount of amino groups and their stability were evaluated
from the thermogravimetric analysis. Fig. 4 shows the DTG curves,
which represents a thermal decomposition in air of aminopropyl
groups attached to pure and alumina modified SBA-15 supports.
It can be seen that the decomposition of organic functionalities
attached to siliceous SBA-15 occurred in three steps. Firstly, weakly
bounded groups were removed (the highest rate of decomposition
was at 320 ◦C), next strongly bounded groups were decomposed
with the highest rate at 380 ◦C, and finally, the residual methylene
Fig. 1. Nitrogen adsorption/desorption isotherms and pore size distributions for
pristine and functionalized SBA-15.

206
K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
Fig. 5. DTG curves of Au(DR)NH2S, Au(DR)NH21AlS and Au(DR)NH210AlS catalysts.
Fig. 3. 27Al NMR spectra from: Au(DR)NH21AlS (1) and Au(DR)NH210AlS before (2)
and after 4 catalytic cycles (3).
The average diameter was less than 6 nm (Fig. 6A) and the parti-
cles were very well dispersed within the pores. In the alumina-free
sample gold particles were less finely dispersed and distribution of
their sizes was notably broader (Fig. 6B). The lack of amines resulted
in much larger particles with size in the range of 15–25 nm (Fig. 6C).
Unlike in DR method, the average size of gold colloids directly
deposited (CD) onto supports was ca. 10 nm, as found from dynamic
light scattering measurements (Fig. 7), and particles of much the
same sizes were also observed on the modified silica surface. Not
surprisingly perhaps, the smaller (6–8 nm) ones were embedded
inside mesopores and the larger ones (15–20 nm) attached on to
the external surface of silica particles (Fig. 6D). The gold content
determined by ICP-MS method was ca. 0.9% in all samples.
Selected catalysts were investigated by XPS method before and
after 4 catalytic cycles. Fig. 8 shows the Au 4f core-level spectra
recorded for Au(CD)NH21AlS, both fresh and used. All spectra are
fitted well by a single doublet line consisting of Au 4f7/2 and Au
4f5/2 peaks with an energy difference, the spin-orbit splitting, of
ca. 3.65 eV [45]. The Eb value of Au 4f7/2 is equal to 83.7 0.1 eV
(Table 2) for all samples and it is a bit lower than that observed for
metallic gold (84 eV). This difference can be attributed to charge
transfer from the oxide surface to AuNPs, and similar results were
also reported previously for gold-on-oxide systems [46,47]. Ionic
gold species (AuIII or AuI) were hardly detected on the catalysts
surface, but their presence cannot be ruled out because the applied
irradiation could change an electronic structure of the examined
catalyst [48]. The values of line-width FWHM (Table 2) increase
in the order: Au(CD)NH21AlS < Au(DR)NH2S < Au(DR)NH21AlS, in
accord with a decrease in particle sizes [49] estimated from TEM
observation. A decrease of FWHM recorded after four catalytic
cycles indicated a small increase in particle size.
groups bound to the silica surface were decomposed at ca. 500 ◦C.
The first step of decomposition is very well seen for alumina-free
sample (NH2S), and also in that with low Al content (NH21AlS),
but not when it was large. This indicates that aminopropyl groups
are more strongly bound to the alumina modified surface (e.g. in
NH210AlS with alumina located on mesopore surface) than to the
purely siliceous materials. In all pristine samples the amount of
organic groups was similar and of about 1.5 mmol/g (Table 1). Dur-
ing deposition of AuNPs the weakly bound groups were washed
out (Fig. 5) in a direct proportion to the initial value. Moreover, the
AuNPs bound to amino groups are more thermally stable, which is
evidenced by an additional minimum on DTG curve at ca. 470 ◦C.
This can be explained by a higher Brönsted acidity of alumina mod-
ified silica materials [44] which, in perfect accord with the rules of
sol-gel chemistry, facilitated the hydrolysis and condensation of
amine group’s precursor with silanols.
The size of AuNPs appeared to depend on both the deposition
method and the modification of silica surface. The smallest parti-
cles were obtained by DR method applied to silica modified with
both alumina and amines (Au(DR)NH210AlS and Au(DR)NH21AlS).
Moreover, it also appeared that Au(CD)NH21AlS (obtained by
a colloidal deposition) had ca. nine-fold larger atomic content
of surface gold than the corresponding sample, Au(DR)NH21AlS,
obtained from a direct reduction of gold precursor pre-adsorbed on
a modified silica surface (bulk samples had the same gold content
of 0.9 wt.%). This indicates that in DR samples most of gold NPs are
embedded in pores of the support, not “seen” by XPS, in perfect
agreement with TEM observations. After four catalytic cycles
the surface composition of Au(CD)NH21AlS changed significantly
(Table 2). The content of Au atoms on the surface decreased by
half and consequently lower values of the atomic ratios: Au/Al,
Au/Si and Au/N were determined. The surface compositions of
both DR catalysts, Au(DR)NH2S and Au(DR)NH21AlS, remained
Fig. 4. DTG curves of NH2S, NH21AlS and NH210AlS materials.

K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
207
Fig. 6. STEM images of Au(DR)NH21AlS (A), Au(DR)NH2S (B), Au(DR)10AlS (C) and Au(CD)NH21AlS (D).
quite similar, except for Au/N atomic ratio equal to 0.245, what
can be ascribed to a lower amine content in alumina-free sample
determined by thermogravimetry (cf. Table 1). After four reaction
runs of Au(DR)NH21AlS catalyst a noticeable increase in the atomic
ratios of Au/Si and Au/N were determined, which indicates that
etching of both silica and aminopropyl groups indeed took place
during reactions.
3.2. Catalytic performance
The catalytic activity appeared to depend strongly on surface
properties of the catalysts and hence the applied preparation
methods. AuNPs obtained by CD and DR on a pure silica SBA-15
appeared to be totally inactive in glucose oxidation with oxygen.
Colloidal NPs directly deposited onto Al modified SBA-15 were also
inactive, because they tended to agglomerate during deposition.
However, the application of DR approach to the same support
resulted in smaller gold particles deposits, and the material
(Au(DR)10AlS) showed some, but low catalytic activity (Fig. 9). The
maximum conversion obtained was only ca. 40% and the catalyst
was inactive in the second cycle, owing to rapid agglomeration
of AuNPs during glucose oxidation, accompanied and followed by
their washing out to the liquid. Used catalyst was white, without a
Fig. 7. Particles size distribution of colloidal gold obtained by DLS.
Table 2
The Au 4f7/2 binding energies, line-width FWHM and Au to Al, O, Si and N atomic ratio obtained from XPS analysis for CD and DR series of the catalysts.
Atomic ratio
Au/Al
Catalyst
Eb Au4f7/2 [eV]
FWHM [eV]
Au [at%]
Au/O
Au/Si
Au/Na
Au(CD)NH21AlSf
Au(CD)NH21AlS4c
Au(DR)NH2S
83.6
83.7
83.7
83.8
83.6
2.26
1.97
2.40
2.66
2.11
0.91
0.45
0.12
0.10
0.15
0.32
0.11
–
0.077
0.072
0.008
0.017
0.002
0.0018
0.0026
0.032
0.017
0.004
0.036
0.052
2.1
1.0
0.245
0.166
0.5
Au(DR)NH21AlSf
Au(DR)NH21AlS4c
a
N as –NH2; f: fresh catalyst, 4c: after 4 catalytic cycles.

208
K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
Fig. 9. Glucose conversion vs. time for DR and CD series of gold catalysts.
of aqueous glucose solution than of washing water, used dur-
ing the catalyst preparation, accelerated the removal of polymeric
ligands [50].
Among the CD-series the highest activity was recorded for
Au(CD)NH21AlS, but still it was less than observed in its DR coun-
terpart.
A precise comparison of activity of our catalysts with those
reported for AuNPs supported on carbon, Al2O3 and TiO2 [29,30,51]
was not easy because of significant differences in reaction con-
ditions (temperature, pH, glucose concentration) and also in
gold content. Therefore, we decided to compare activities of
Au(DR)NH21AlS and Au(CD)NH21AlS with those of commercial cat-
alysts AUROlite Au/TiO2 and Au/Al2O3 in a single catalytic cycle. It
appeared that activities of Au/TiO2 and Au/Al2O3 catalysts were
respectively, ca. 55% and 46% lower than that of Au(DR)NH21AlS
(Fig. 10) and similar in value to Au(CD)NH21AlS.
All the catalysts showed high selectivity (ca. 96%) towards glu-
conic acid. The 1H NMR analysis confirmed that the only by-product
was fructose created during glucose isomerisation and not due
to catalytic reaction. Its concentration was around 3% and the
remaining was the unreacted glucose.
Fig. 8. XPS Au 4f spectra of Au(CD)NH21AlS catalyst; (A) fresh, (B) after 4 catalytic
cycles.
characteristic pink colour observed in freshly prepared materials.
All amine functionalized materials showed significantly better
activity in glucose oxidation (Fig. 9). On the whole, activities of
the catalysts synthesized using DR method decreased in the order
Au(DR)NH21AlS > Au(DR)NH210AlS > Au(DR)NH2S > Au(DR)10AlS »
Au(DR)S. Inasmuch gold particles in two most active samples had
similar sizes and sizes of both catalysts particles were the same
(in the range of 5–30 ␮m), lower activity of the second sample
can be explained by internal diffusion limitation of mass transport
caused by smaller pores (Table 1). The third (alumina-free) catalyst
possessed larger gold particles of less uniform sizes and its activity
was ca. 38% lower.
Interestingly, in all CD catalysts characteristic induction time
was observed during the first run (Fig. 9). On the basis of XPS
data (Fig. 8, Table 2) we may rule out the effect of not fully
reduced gold. The most likely reason is the polyvinyl alcohol,
used as stabilizing agent in the colloidal AuNPs synthesis, which
hindered glucose molecules access to active gold atoms at the
beginning of the process. In its later stages higher temperature
Fig. 10. Activity of the synthesized and commercial gold catalysts in the first cat-
alytic cycle.

K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
209
Fig. 12. STEM images of Au(DR)NH21AlS (A) and Au(CD)NH21AlS (B) after 4th cat-
alytic cycles.
violet, clearly indicating that this process indeed occurred. The
more in-depth XRD analyses showed a threefold increase in gold
crystallite sizes. Most likely, the further use of this catalyst would
result in a steady increase of gold particle sizes and a rapid decrease
in their activity. After few operational cycles we observed changes
in structure parameters determined by nitrogen adsorption. Both
specific surface area, pore volume and pore sizes were larger than
in pristine catalysts and they were in between those obtained for
Fig. 11. Activity of the catalysts prepared using DR (A) and CD (B) method in suc-
cessive catalytic cycles.
3.3. Long-term stability
Investigations of the catalysts’ stability in the repeated batch
experiments (4–7 runs) showed that catalysts not only retained
their initial activity (Fig. 11) but it increased in the second run
and the largest increase was recorded for Au(DR)NH21AlS. The
activities were determined at the beginning of the reaction when
linear dependences of conversion against time were observed.
Small fluctuations in the activity in successive runs were prob-
ably caused by the changes in oxygen mass transfer [27,29,30].
The same relationship was observed already before in long-term
stability investigations of AuNPs supported on alumina [27].
The selected catalysts, after four operational cycles, were care-
fully analysed by XRD, XPS, TEM and nitrogen adsorption to check
the stability of gold particles and supports’ structure. It appeared
that in DR catalysts, Au(DR)NH21AlS and Au(DR)NH210AlS, the
initial sizes of AuNPs were well preserved (cf. Fig. 12A) and this
observation was also corroborated by the XRD analyses of Au
crystallites sizes (Scherrer method). In the pristine and used
catalysts the crystallite sizes remained the same (ca. 5.1 nm). More
significant changes in gold particle sizes were observed in the
CD catalysts (Fig. 12B). The agglomeration of gold nanoparticles
was observed not only in microscopic images, but it could also
be inferred from the change in the catalyst colour from pink to
Fig. 13. Glucose oxidation over Au(DR)NH21AlS – leaching test with fast filtration
of the catalyst.

210
K. Odrozek et al. / Applied Catalysis A: General 475 (2014) 203–210
alumina functionalized SBA-15 (10AlS) and the fresh catalyst (see
Table 1). The increase was due to etching of aminopropyl groups
and also of some silica, clearly evidenced by XPS analysis (cf. supra
3.1). To explain the role of amines during repeated operations,
the Au(DR)NH210AlS sample was calcined at 500 ◦C to remove
amines, and we hardly observed any change in both AuNPs size
and catalytic activity tests which followed. This proves that amines
play an important role during the synthesis of catalysts−by a DR
method. They act as strong adsorption sites for AuCl4 anions
and form a sort of weak covalent bonds with Au particles just
under development [52], to prevent their agglomeration during a
reduction process. Ultimately, this approach results in excellent
distribution of small AuNPs in pores, and not less importantly, it
effectively reduces the natural trend towards agglomeration, even
after removal of amine groups.
[3] G. Bond, D. Thompson, Gold Bull. 42 (2009) 247–259.
[4] M. Haruta, Chem. Rec. 3 (2003) 75–87.
[5] M. Haruta, Gold Bull. 37 (2004) 27–36.
[6] G. Schmid, B. Corain, J. Inorg. Chem. 17 (2003) 3081–3098.
[7] C. Corti, R. Holliday, D. Thompson, Appl. Catal. A: Gen. 291 (2005) 253–261.
[8] M. Haruta, M. Daté, Appl. Catal. A: Gen. 222 (2001) 427–437.
[9] C-W. Chiang, A. Wang, B-Z. Wan, C-Y. Mou, J. Phys. Chem. B 109 (2005)
18042–18047.
[10] C.T. Campbell, Acc. Chem. Res. 46 (2013) 1712–1719.
[11] L.-F. Gutiérrez, S. Hamoudi, K. Belkacemi, Catalysts 1 (2011) 97–154.
[12] M.T. Bore, H.N. Pham, E.E. Switzer, T.L. Ward, A. Fukuoka, A.K. Datye, J. Phys.
Chem. B 109 (2005) 2873–2880.
[13] B. Lee, Z. Ma, Z.T. Zhang, C. Park, S. Dai, Micropor. Mesopor. Mater. 122 (2009)
160–167.
[14] C.-M. Yang, P.-H. Liu, Y.-F. Ho, C.-Y. Chiu, K.-J. Chao, Chem. Mater. 15 (2003)
275–280.
[15] C-M. Yang, M. Kalwei, F. Schüth, K-J. Chao, Appl. Catal. A: Gen. 254 (2003)
289–296.
[16] Y.-S. Chi, H.-P. Lin, C.-Y. Mou, Appl. Catal. A: Gen. 284 (2005) 199–206.
[17] A.V. Biradar, T. Asefa, App. Catal. A: Gen. 435–436 (2012) 19–26.
[18] S. Ramachandran, P. Fontanille, A. Pandey, C. Larroche, Food Tech. Biotechnol.
44 (2006) 185–195.
[19] V. Singh, R. Kumar, Appl. Microbiol. Biotechnol. 75 (2007) 713–722.
[20] C. Dowdells, R.L. Jones, M. Mattey, M. Bencˇina, M. Legisˇa, D.M. Mousdale, Lett.
Appl. Microbiol. 51 (2010) 252–257.
[21] K. Buchholz, J. Seibel, Carbohydr. Res. 343 (2008) 1966–1979.
[22] S. Biella, L. Prati, M. Rossi, J. Catal. 206 (2002) 242–247.
[23] Y. Önal, S. Schimpf, P. Claus, J. Catal. 223 (2004) 122–133.
[24] H. Okatsu, N. Kinoshita, T. Akita, T. Ishida, M. Haruta, Appl. Catal. A: Gen. 369
(2009) 8–14.
Finally, to control the leaching of gold from our best catalyst
(Au(DR)NH21AlS) we performed a fast catalyst filtration test, after
reaching 50% of glucose conversion. This experiment showed (cf.
Fig. 13) that gold NPs were not washed out from the catalyst during
reaction and hence it really worked as a truly heterogeneous one.
4. Conclusions
[25] M. Comotti, C. Pina, R. Matarrese, M. Rossi, Angew. Chem. Int. Ed. 43 (2004)
5812–5815.
[26] A. Mirescu, U. Prüße, Catal. Commun. 7 (2006) 11–17.
[27] C. Baatz, U. Prüße, J. Catal. 249 (2007) 34–40.
[28] C. Baatz, N. Thielecke, U. Prüße, Appl. Catal. B: Environ. 70 (2007) 653–660.
[29] C. Baatz, N. Decker, U. Prüße, J. Catal. 258 (2008) 165–169.
[30] A. Mirescu, H. Berndt, A. Martin, U. Prüße, Appl. Catal. A: Gen. 317 (2007)
204–209.
[31] B. Gawel, K. Lambrechts, G. Øye, Mater. Sci. Eng. B: Adv. Funct. Solid State Mater.
177 (2012) 575–580.
[32] B. Gawel, K. Gawel, T.C. Hobæk, M. Yasuda, G. Øye, Mater. Chem. Phys. 137
(2012) 414–420.
[33] M. Ruszel, B. Grzybowska, M. Liniecki, M. Wójtowski, Catal. Commun. 8 (2007)
1284–1286.
[34] W. Yan, B. Chen, S.M. Mahurin, E.W. Hagaman, S. Dai, S.H. Overbury, J. Phys.
Chem. B 108 (2004) 2793–2796.
[35] C. Wang, Y. Li, H. Liu, L. Ye, C. Sun, Y. Fang, Y. Song, Adv. Mater. Res. 550–553
(2012) 220–224.
The highly active and stable catalysts for aerobic glucose oxi-
dation were obtained using gold nanoparticles embedded in silica
SBA-15 materials functionalized with aminopropyl groups and alu-
mina. The deposition of gold by a direct reduction of its precursor
on the modified support appeared to be far more effective to obtain
efficient catalyst than the deposition of colloidal gold on the same
support. In the former method amino gro−ups play a double role:
serve as strong adsorption sites for AuCl4 anions, which results
in their excellent dispersion inside pores, and stabilize small gold
nanoparticles during the reduction process. The embedding of gold
particles in the mesopores of SBA-15 appeared to be critical to pre-
vent their agglomeration during repeated oxidation processes, and
it also proved to be very effective means even after amine removal.
A positive effect of alumina incorporation can be explained by the
stronger binding and improved stability of aminopropyl groups
attached to the more acidic mixed oxide surface of AlSBA-15, with
a positive effect on the catalysts’ performance. XPS analysis con-
firmed the metallic state of catalytically active AuNPs and provided
a solid proof of gradual washing out of aminopropyl groups and
some silica during glucose oxidation reaction. Activity of the cata-
lysts devised in this study was significantly larger than those of the
commercial ones: AUROlite Au/TiO2 and Au/Al2O3, considered as
benchmarks.
[36] S. Biella, F. Porta, L. Prati, M. Rossi, Catal. Lett. 90 (2003) 23–29.
[37] K. Odrozek, K. Maresz, J. Mrowiec-Białon´ , Przem. Chem. 91 (2012)
1863–1867.
[38] T. Ishida, H. Watanabe, T. Bebeko, T. Akita, M. Haruta, Appl. Catal. A: Gen. 377
(2010) 42–46.
[39] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Science 279 (1998) 548–552.
[40] J. Mrowiec-Białon´ , Thermochim. Acta 443 (2006) 49–52.
[41] M. Jaroniec, L.A. Solovyov, Langmuir 22 (2006) 6757–6760.
[42] J. Choma, M. Jaroniec, Appl. Surf. Sci. 253 (2007) 5587–5590.
ˇ
[43] A. Zukal, H. Siklova, J. Cejka, Langmuir 24 (2008) 9837–9842.
[44] A. Corma, Chem. Rev. 97 (1997) 2373–2420.
[45] J.F. Moulder, W.F. Stickle, P.E. Sobol, K.D. Bomben, in: J. Chastain (Ed.), Hand-
book of X-ray Photoelectron Spectroscopy, Perkin-Elmer Corporation Physical
Electronics Division, USA, 1992.
Acknowledgements
[46] J. Radnik, C. Mohr, P. Claus, Phys. Chem. Chem. Phys. 5 (2002) 172–177.
[47] S.F. Arrii, Morfin, A.J. Renoupres, J.L. Rousset, J. Am. Chem. Soc. 126 (2004)
1199–1205.
[48] G.C. Bond, D.T. Thompson, Gold Bull. 33 (2000) 41–50.
[49] D. Dalacu, J.E. Klemberg-Sapieha, M. Martinu, Surf. Sci. 472 (2001) 33–40.
[50] J.A. Lopez-Sanchez, N. Dimitratos, C. Hammond, G.L. Brett, L. Kesavan, S. White,
P. Miedziak, R. Tiruvalam, R.L. Jenkins, A.F. Carley, D. Knight, C.J. Kiely, G.J.
Hutchings, Nat. Chem. 3 (2011) 551–556.
The authors wish to thank Professor A. Jarze˛bski for helpful dis-
cussions. Dr K. Rodak is thanked for her help in making the STEM
measurements. This work was financed by the National Science
Centre (Poland), project No DEC-2011/01/B/STB/03855.
[51] U. Prüße, M. Herrmann, C. Baatz, N. Decker, Appl. Catal. A: Gen. 406 (2011)
89–93.
References
[52] J.-P. Sylvestre, S. Poulin, A.V. Kabashin, E. Sacher, M. Meunier, J.H.T. Luong, J.
Phys. Chem. B 108 (2004) 16864–16869.
[1] M. Haruta, N. Yamada, T. Kobayahsi, S. Iijima, J. Catal. 115 (1989) 301–309.
[2] M. Haruta, Catal. Today 36 (1997) 153–166.