SiO2-, Cu-, and Ni-supported Au nanoparticles for selective glycerol oxidation in the liquid phase

Article abstract of DOI:10.1016/j.jcat.2014.08.003

We tested for the first time the efficiency of SiO₂-, Cu-, and Ni-supported Au in deep glycerol oxidation in a diluted and viscous H₂O₂/H₂O liquid phase. Acetic acid (AA), the C₂ oxidate, was preferentially formed in such a system. High conversion (100%) and AA yields (90%) were observed for the sol-gel SiO₂-supported Au in diluted solutions. Although with the increase of glycerol concentration in the viscous liquid phase these values decreased to ca. 40% (conversion) and 20% (AA yield), the addition of acetonitrile improved the AA yield to ca. 40%, while the surfactants were found to be capable of a many-fold enhancement of the catalyst activity at the room temperature highly viscous liquid phase. High performances were also observed for the bimetallic Au/Cu and Au/Ni catalysts obtained by nano-Au transfer; however, these catalysts were destroyed during the reaction by the Cu or Ni leaching effect.

Full text of DOI:10.1016/j.jcat.2014.08.003

Journal of Catalysis 319 (2014) 110–118
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
SiO -, Cu-, and Ni-supported Au nanoparticles for selective glycerol
2
oxidation in the liquid phase
a
a,d
a
a
b
b
Maciej Kapkowski , Piotr Bartczak , Mateusz Korzec , Rafal Sitko , Jacek Szade , Katarzyna Balin ,
c
a,
⇑
Józef Lel a˛ tko , Jaroslaw Polanski
a
Institute of Chemistry, University of Silesia, Szkolna 9, 40-006 Katowice, Poland
Chełkowski Institute of Physics, ul. Uniwersytecka 4, 40-007 Katowice, Poland
Institute of Materials Science, University of Silesia, 75 Pułku Piechoty 1A, 41-500 Chorzów, Poland
NANO-CHEM-TECH, Ligocka 90A/14, 40-568 Katowice, Poland
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
We tested for the first time the efficiency of SiO -, Cu-, and Ni-supported Au in deep glycerol oxidation in
2
Received 14 June 2014
Revised 6 August 2014
Accepted 6 August 2014
a diluted and viscous H
2
O
2
/H
2
O liquid phase. Acetic acid (AA), the C
2
oxidate, was preferentially formed in
-supported
such a system. High conversion (100%) and AA yields (90%) were observed for the sol–gel SiO
2
Au in diluted solutions. Although with the increase of glycerol concentration in the viscous liquid phase
these values decreased to ca. 40% (conversion) and 20% (AA yield), the addition of acetonitrile improved
the AA yield to ca. 40%, while the surfactants were found to be capable of a many-fold enhancement of
the catalyst activity at the room temperature highly viscous liquid phase. High performances were also
observed for the bimetallic Au/Cu and Au/Ni catalysts obtained by nano-Au transfer; however, these cat-
alysts were destroyed during the reaction by the Cu or Ni leaching effect.
Keywords:
Au nanoparticles
Glycerol oxidation
Bimetallic Au/Cu
Au/Ni
Ó 2014 Elsevier Inc. All rights reserved.
Heterogeneous catalysis
Silica supported nanogold
Silica support
1
. Introduction
Renewable naturally sourced carbohydrates, amino acids, and
reactions have been developed; e.g., catalytic glycerol conversions
have been thoroughly reviewed [4]. Nanocatalysis is an interesting
option in this area.
triglycerides are available in vast quantities in our environment.
This biomass, a product of living organisms, could be used as valu-
able feedstock for chemical processing; however, we need novel
chemistry to transform large amounts selectively and efficiently in
their natural state without extensive functionalization and protection
Nanocatalysts are extremely sensitive toward structure differ-
entiation, and their activity and selectivity depend not only on
nanometal and support type but also on size, shape, and composi-
tion [5]. Thus, optimization of such materials is an open issue. In
fact, more efficient catalysts are still being sought to run the reac-
tions with higher yields and higher selectivity under mild condi-
tions. Other important problems to be addressed include
reducing the fraction of noble metals, facilitating the catalyst sep-
aration, improving reusability, and reducing contamination of the
final products. Gold nanoparticles (Au NPs) catalyze a variety of
reactions [6–9]. As they tend to agglomerate, they are usually sup-
ported on carriers to form more stable catalytic systems. Generally,
Au NPs are available on a variety of supports, from carbon-like
graphite to inorganic materials. Basically, the first should be
wettable by apolar reagents and solvents, while the latter should
be wettable by polar ones. Wettability, and consequently catalyst
availability for the reactants, is of crucial importance for the
reaction progress; e.g., the recently oxidation of cyclohexene and
[
1]. For this reason, biomass conversion has received increasing
attention in contemporary chemistry. Glycerol yielded as a by-
product in biodiesel production is one of the most widely-available
biosourced chemicals, making it an attractive target of investiga-
1 3
tions. The oxidation of glycerol could yield a variety of C to C oxy-
genates, which are potentially valuable chemicals in chemical and
pharmaceutical applications or intermediates in organic synthesis.
This has triggered a growing interest in new methods for this pro-
cess. Although there has been tremendous progress in recent years
in this area, a number of problems remain to be solved [2,3].
Available enzymatic or stoichiometric methods are often wasteful
and economically inefficient. Alternatively, a variety of catalytic
⇑
Corresponding author.
E-mail address: polanski@us.edu.pl (J. Polanski).
D
-glucose over nano-Au/SiO
2
in water has been compared. In this
-glucose complying with polar SiO
reaction, polar polyhydroxyl
D
2
http://dx.doi.org/10.1016/j.jcat.2014.08.003
021-9517/Ó 2014 Elsevier Inc. All rights reserved.
0

M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
111
support and a polar solvent (H
perature, whereas nonpolar cyclohexene violating the polarity rule
needed the addition of surfactants to react efficiently [10].
Glycerol oxidation catalyzed on Au NPs, pioneered by Hutchings
et al. [11,12], has been exhaustively studied recently [8,9,13–17].
Possible glycerol oxidates can be connected by a complex reaction
network [3,16,21]. Scheme 1 presents a variety of products yielded
from such catalytic systems, with nanometals supported on carbon
2
O) reacted smoothly at room tem-
widest wettability and availability for the reacting molecules. This
would be especially important for highly viscous undiluted glyc-
erol solutions. Thus, a suspension of the silica-supported catalyst
should provide a system in which the reaction proceeds smoothly,
while hydrogen peroxide could be a compatible convenient, safe,
and green reagent in such a system. It is also a model oxidant, since
it was reported that aerobic oxidation on Au NPs proceeds with
in situ formation of H
oxidation of glycerol on SiO
investigated, despite the polarity match between glycerol, SiO
and H /H O systems. Recently, we reported a new method for
transferring SiO -supported nanometals to a variety of other sup-
2
O
2
[16]. To the best of our knowledge, the
or inorganic carriers. The C
3
oxygenates, of which dihydroxyace-
2
-supported nano-Au has never been
tone is the most desired product, designate oxidation without C–
C cleavage. Possible products and catalytic systems are specified
2
,
2
O
2
2
in Tables 15 and 16 of Ref. [3]. In contrast, oxidation to C
investigated as a potential source of glycolic acid [18] and C
ucts. In turn, CO and HCOOH are products of the highest chain
decomposition level.
Glycerol oxidation by aqueous hydrogen peroxide in an auto-
clave reactor on Au/C, Au/graphite, or Au/TiO under the presence
of NaOH provided a mixture of glycolate, glycerate, and tartronate
18]. Moreover, oxidation of glycerol to glyceric acid using an Au/
graphite system in aqueous sodium hydroxide under basic free
19] or mild conditions [11] also appeared to be relatively selective.
2
has been
2
1
prod-
ports [22]. As Au/Cu nanoparticles have been recently developed
as an interesting catalyst for the oxidation of various alcohols
[23], here we prepared bimetallic nano-Au supported on Cu or Ni
grains, to test their performance in glycerol oxidation. Despite
the complex network of possible reactions (Scheme 1), a process
of deep glycerol oxidation proceeds on the investigated catalysts
preferentially to acetic acid (AA).
2
2
[
[
2. Experimental
The influence of various carbonaceous supporting materials on the
Au/C catalyst performance in glycerol oxidation was described by
Gil et al. [20]. More generally, the influence of various support
materials on catalytic glycerol oxidation was reviewed by Katry-
niok et al. [16].
In this study, we tested the efficiency of the deep oxidation of
glycerol in the presence of unmodified silica-supported nanogold
catalyst in aqueous 30% hydrogen peroxide, under mild conditions
in aqueous glycerol solutions of various dilutions. We assumed
that polar glycerol or a glycerol/water mixture would prefer a
relatively hydrophilic catalyst carrier, which should provide the
2.1. Preparation of Au NPs on silica or Cu and Ni carriers
The series of nanocatalysts, namely Au/silica or Au/Cu and Au/
Ni were prepared according to the procedure optimized (for details
see: Supplementary material).
2.2. Glycerol oxidation
Nano-Au catalyst (20 mg, 0.2–20.0
lmol Au) was suspended in
a mixture of 1.0 mL of 30% hydrogen peroxide (10 mmol H O ) and
2
2
Scheme 1. A complex reaction network of glycerol oxidation [16].

1
12
M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
0
.5 mL (0.5–13.6 mol/L) glycerol (Fisher BioReagentsÒ – Glycerol
For Molecular Biology) by sonication at room temperature for
0 min (RK 52 H, Bandolin Electronics, 35 kHz). Reagents were stir-
concentrations determined by XPS are lower than those deter-
mined by EDXRF. This suggests that in the 0.1% Au/SiO catalyst,
in addition to the Au NPs located directly on the catalyst surface,
some of their important parts are located within SiO pores that
are out of view of the XPS analysis. It should be noted that X-rays
have much larger penetration range compared to silica
particles. The information depth 99% for any element that
yields 99% of the element intensity is given by the formula
99% = 4.6/ (E ,E , where is the density of the sample and
) Â
(E ,E ) = (E )csc(/ ) + (E )csc(/ ) is the total mass-attenuation
coefficient of the sample. (E ) and (E ) represent the mass atten-
uation coefficients of the sample at the primary E and fluorescent
radiation E (analytical line of Au, i.e., L at 9.71 keV), and / and /
are the incidence and take-off angles, respectively. The information
depth d99% calculated for gold (Au L line) in silica particles is ca.
530 m. This is much larger than the diameter of silica particles,
2
1
red at 770 rpm in a sealed tube (septa system) placed in a thermo-
stated oil bath at 80 °C for 24 h. The resulted reaction mixture was
centrifuged and decantated. The supernatant was dissolved into
deuterated water and analyzed using H and C NMR. For quanti-
tative determination of the reaction products, we used an external
standard procedure with Coaxial Small Volume NMR Insert tubes
2
a
1
13
d
d
v
v
l
0
i
q
1
q
(
ARMAR Chemicals) and hydroquinone as a reference substance.
0
i
0
l
i
2
Additionally, the 2D COSY and HMQC methods were used to iden-
tify and quantify products. The spectra were recorded on the Bru-
ker Avance 400 or 500 spectrometers with TMS as internal
l
0
l
i
0
i
a
1
2
standard (400 MHz, H, 101 MHz C or 500 MHz, 1H, 126 MHz
1
13
1
3
C) at room temperature. The signal from water was suppressed
a
using 90 water-selective pulses (zggpwg). Optionally this oxida-
tion procedure was modified by the addition of 1.0 mL acetonitrile
l
i.e., 500–1200 nm. Therefore, in contrast to XPS results which
reveal the surface structure of the samples, EDXRF provides us with
the representative bulk composition of the catalysts.
(
19.10 mmol) or surfactants: Sulforkanol (sodium laureth sulfate –
SLES), Triton X-100, PEG 400, ca. (0.05 wt.%).
Also, SEM and TEM analyses indicate that a surface texture of
0
.1% Au/SiO
2
evidently differs from that of the higher Au content
catalysts. In Fig. 1, we compare the 0.1% Au/SiO and 1%
systems. This figure shows that Au NPs are deeply embed-
catalyst (Fig. 1a and b), whereas
Au sticks out of the silica surface for the 1.0% Au/SiO sample
Fig. 1c). Probably, the Au solution could have entirely penetrated
3
. Results and discussion
Au/SiO
Au/SiO
2
2
2
A variety of nano-Au/C supported catalytic systems have been
ded into silica for the 0.1% Au/SiO
2
developed recently for the selective oxidation of glycerol [26];
however, there have been no reports on the possible application
of SiO -supported Au NPs in glycerol processing. In turn, silica-sup-
2
ported Nb- and W-oxide, if applied to the gas phase dehydration of
glycerol under argon, appeared to yield acrolein [27]. Recent devel-
opments in this field were discussed in [28]. Additionally, various
mixed oxide catalysts were used in glycerol oxydehydration,
another variant of catalytic glycerol processing that was also tested
in the gas phase [29].
2
(
into the porous silica surface during the reduction step if it were
present in a low 0.1% fraction. The contents of Au determined for
1
.0% Au/SiO
2
by EDXRF and XPS analyses are shown in Table 1,
system, these values do not
entry 2. Similar to the 0.1% Au/SiO
2
agree and XPS showed ca. twice as high concentrations as EDXRF.
However, in this case, we observed a relative surface Au enrich-
ment. With a larger amount of Au salt used during the reduction
process, there is insufficient space within the pores and Au NPs
had to spread across the surface.
3.1. The catalysts preparation and structure
Over the past few years, a number of techniques have been
developed for the production of nanosized metallic particles and
their distribution on different carriers [30,31]. The methods in
use, based on ‘‘the bottom-up’’ and ‘‘the top-down’’ techniques,
still suffer from some disadvantages, including the broad-sized dis-
tribution of nanoparticles and their tendency to aggregate or form
clumps [30,31]. To minimize these problems, we recently devel-
oped a novel innovative method for the formation of bimetallic
Pd catalysts [22].
We used amorphous silica synthesized by the sol–gel [25] tech-
nique as a basic carrier. SEM observations indicated that silica
obtained by this method exhibited a regular spherical shape, a con-
trolled size distribution, and a uniform porous surface (Fig. S1, Sup-
plementary material). This regular shape was preserved in Au NPs
supported on the SiO
24].
The EDXRF spectrum of 0.1% Au/SiO
several peaks at 8.45, 9.71, 10.26, 11.44, and 13.38 keV, which cor-
respond to the Au lines Ll, L , Ln, Lb, and L , respectively (Fig. S2,
2
carrier obtained using the Stöber method
[
2
2
, in contrast to SiO , shows
Herein, we tested to determine whether this approach can also
be used for other bimetallic systems, in particular, Cu- and Ni-sup-
a
c
Supplementary material). The quantitative EDXRF analysis reveals
ported Au NPs. After some modifications, fumed silica (f-SiO ) was
2
the presence of minor (Ca) and trace elements (Cr, Mn, Fe, Ni, Cu,
also tested as potential target carrier supporting Au NPs. The 1%
and Pt), in both SiO
Supplementary material). The determined concentration of Au,
.092%, is very close to the designed and expected value of 0.1%
Table 1, entry 2). In contrast, the XPS analysis, if used to determine
the Au concentration on the catalyst surface, indicated entirely dif-
2
and Au/SiO
2
(for more details see Table S1,
Au/SiO2 system was chosen for the intermediate carrier. Accord-
ingly, bimetallic or f-SiO2 catalysts were synthesized by transfer
of Au NPs from 1% Au/SiO2 to the target carrier. For Cu or Ni, the
ingredients were suspended in deionized water, placed in an ultra-
0
(
sound bath, and stirred. Then, SiO was digested. We assumed that
2
ferent values. For the 0.1% Au supported on sol–gel, SiO
2
surface Au
the appropriate digesting solvent should fulfill the following
requirements: It should digest only the intermediate carrier (i.e.,
silica), and it must be inert both for the target carrier and metallic
nanoparticles. We used 40% aqueous NaOH as a digesting medium.
XPS results are shown in Fig. 2. Once again, a comparison of the
EDXRF and XPS analyses (Table 1, entries 4 and 5) reveals the sur-
face Au enrichment effect for these new catalysts. The low porosity
Table 1
Au content as determined by EDXRF and XPS analyses.
Catalyst
Au concentration, % (m/m)
EDXRF
XPS
1
2
3
4
5
1.0% Au/SiO
0.1% Au/SiO
1.0% Au/f-SiO
1.0% Au/Cu
1.0% Au/Ni
2
2
0.711 ± 0.042
0.092 ± 0.0018
1.12 ± 0.034
1.18 ± 0.051
1.11 ± 0.030
1.42 ± 0.05
0.04 ± 0.05
0.2 ± 0.05
7.2 ± 0.1
2
of Cu and especially Ni determines that, in comparison with SiO , a
larger fraction of Au NPs are directly available on the surfaces of
these catalysts. SEM and TEM analyses prove that Au–Cu or Au–
Ni contact is formed in these catalysts (Supplementary material
Fig. S3a, S3b, S3c, S3d, S4, S5). However, the residual debris of
a
2
50.2 ± 0.1
a
OrisilÒ 380.
2
the original Au/SiO conglomerates can still be detected on the

M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
113
3.2. Glycerol oxidation in liquid aqueous phase
We assumed that treatment of the concentrated glycerol solu-
tions under mild conditions would potentially be a major advan-
tage for processing waste glycerol. Intuitively, the glycerol
concentration that influences the viscosity of a system should
strongly control the glycerol contact with the catalyst surface. In
practice, the catalytic reactions of glycerol in water solutions have
been performed previously in a relatively diluted solutions, e.g., in
the liquid phase: 0.6 mol/L (mol of glycerol per liter of the reaction
mixture, if recalculated using data from the literature; glycerol to
2
H O
2
molar ratio amounted to 1:4) [18] or in the vapor phase:
Ar/glycerol/H O = 5:1:21 [27] or N /H O/Gly = 46/48/6 [29].
In Table 2, we specified the performance of catalytic Au/SiO
systems in the oxidation of glycerol in the diluted liquid phase;
.2 mol/L glycerol, whereas glycerol to H molar ratio amounted
to 1:37. For the new catalysts, i.e., SiO -supported Au, both the
conversion and selectivity of the process can reach as high as
00% (Table 2, entries 1 and 3), respectively, while in these condi-
tions, the Au/C system provided much lower conversion (Table 2,
entry 7). AA (Au/SiO ) or glycolic acid (GA) and aldehyde (Au/C)
2
2
2
2
0
2 2
O
2
1
2
were the main products observed for the reactions, respectively.
The performance of the Au/C system, typically used for catalytic
glycerol oxidation [3], compares well to published data for the
1
2 2
% Au/C (0.6 mol/L glycerol to H O molar ratio 1:4), which pro-
vides only slightly higher conversions of ca. 40% [18].
In Table 3, we reported the results of glycerol oxidation where
we designed the dilutions amounting to 1.0 mol/L (mol of glycerol
per liter of the reaction mixture), while the glycerol/H O molar
2 2
ratio took a value of 1:7. Thus, the dilutions were slightly lower
than those usually used in the experiments reported in the litera-
2
ture (0.6 mol/L). For the new catalysts, i.e., SiO -supported Au, both
the conversion and AA selectivity of the process were slightly
lower than those obtained for more diluted solutions (Table 2),
reaching ca. 90% (conversion) or more than 85% (AA selectivity),
as indicated in Table 3, entries 1 and 3. Also, for the Au/C system
(
Table 3, entry 4), the conversion is similar to that specified in
Table 2, entry 7, i.e., it is similar to that obtained for more diluted
glycerol solutions. AA (Au/SiO ) and glycolic aldehyde (GLAD) (Au/
2
C) were the main products observed for the reactions, respectively.
To test glycerol reactivity in undiluted solutions, we mixed
together glycerol and 30% aqueous H
system with 4.5 mol/L glycerol in a reaction mixture in which
the glycerol/H molar ratio amounted to 1:1.5. Glycerol was
reacted under low temperatures, i.e., below 80 °C. If tested, in undi-
luted solutions the performance of the 1.0% Au/C system appeared
relatively low, not exceeding a conversion of 8%. Glyceric acid
(GLA), also a typical product of glycerol oxidation in the presence
of C-supported Au NPs, appeared to be the main product of the
2 2
O . This provided a reaction
2 2
O
Fig. 1. TEM images of silica supported Au NPs in 0.1% Au/SiO
c).
2 2
(a, b), 1.0% Au/SiO
(
metal surface, especially Cu (Supplementary material Fig. S4 and
S5).
For the preparation of f-SiO
modified the procedure by preprocessing the intermediate 1%
Au/SiO catalyst by sonication with magnesium oxide. Then, the
silica digested in aqueous NaOH provided intermediate MgO-sup-
ported Au NPs. This material, if sonicated with f-SiO , gave a con-
glomerate from which MgO could be removed in acidic solutions
of glacial acetic acid as a digesting solvent. The resulting system
is a highly porous silica structure ornamented with Au NPs (for
TEM and SEM images see Supplementary material Fig. S6). In this
2
(Orisil)-supported Au NPs, we
2
reaction (Table 4, entry 4). The performance of the SiO -supported
Au catalysts in the undiluted liquid phase is detailed in Table 4. It is
seen that the conversions reached a maximum value of ca. 40% for
2
0.1% Au/SiO
80% with a high AA yield for Au/f-SiO
appeared now to be the most selective SiO
2
(Table 4, entry 1). The selectivity to AA could reach ca.
(Table 4, entry 3); this
-based catalyst. The
2
2
2
carbon C-supported system yielded other products under lower
selectivities (Table 4, entry 4). In a high-viscosity liquid phase glyc-
erol oxidation to AA, the performance of the 0.1% Au/SiO system is
2
context, the structure resembles that of the 0.1% Au/SiO
2
. Further
given by an outstanding TON of 26932. A low Au fraction and a
high glycerol concentration accompanied by a high conversion rate
contribute to this result. This compares with a TON value of 497 for
1% Au/C catalyst, indicating much lower activity. Practically, AA
could not be found in the reaction mixture in the 1% Au/C system.
evidence is provided by a comparison of the EDXRF vs. XPS analy-
ses (Table 1, entry 3), which indicates a high discrepancy of the
surface and bulk Au concentrations. However, unlike highly mono-
dispersive sol–gel silica, Orisil (which is a polydispersive system), if
used as the Au support, also resulted in Au/f-SiO
2
with a much less
2
It is worth noticing that the f-SiO (Orisil) support appeared to give
defined structure. Apart from the small Au clusters below 5 nm, we
can, incidentally, find here much larger Au conglomerates.
the highest AA yield in comparison with other SiO
(Table 4, entries 1, 2, 3).
2
systems

1
14
M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
Fig. 2. XPS multiplets of Ni 2p (a) and Cu 2p (b) before and after reaction. The bars represent the contributions from the various oxidation states of Cu and Ni, derived from
fitting the data obtained before reaction.
Table 2
Catalytic performance of SiO
a
2
and C supported Au NPs in diluted glycerol solutions at 80 °C.
Catalyst
TONb
TOFb (hÀ1
)
Conv. (mol%)
Selectivity to productsc (mol%)
Yield to acetic acid (%)
GLAD
GA
TA
AA
FA
OS
2.9
11.0
0.7
4.6
4.6
5.3
7.4
1
2
3
4
5
6
7
0.1% Au/SiO
1.0% Au/SiO
1.0% Au/f-SiO
2.0% Au/SiO
5.0% Au/SiO
10.0% Au/SiO
1.0% Au/C
2
2
2660
201
266
125
45
111
8
11
5
2
1
100.0
75.5
100
94.3
84.4
86.0
35.1
2.2
1.6
0
2.2
1.3
0
0
1.8
0
0.3
3.2
0,7
25.9
0
1.3
0
0.4
0
0
90.0
84.3
99.3
92.5
90.9
94.0
0
4.9
0
0
0
0
90.0
63.7
99.3
87.2
76.7
80.8
0
d
2
2
2
2
23
93
0
0
4
66.7
0
a
b
c
0
.2 mol/L of glycerol in the reaction mixture (glycerol/H
2 2
O molar ratio 1:37), 20 mg of catalyst (0.2–20.0 lmol Au), 80 °C, 24 h, 770 rmp.
Turnover number (TON) or turnover frequency (TOF) based on the total gold content in the material.
GLAD – glycolaldehyde, GA – glycolic acid, TA – taratronic acid, AA – acetic acid, FA – formic acid, OS – others.
Fumed silica OrisilÒ 380
d
The influence of glycerol dilution in the reaction mixture on the
concentration. As could be expected, this effect is also much less
distinct for the SiO -supported Au then for the C-based catalyst
(Table 2, entries 1 vs. 7 and Table 4, entries 1 vs. 4). This can be
explained by the substantial difference in wettability of the
reaction yield and selectivity is shown in Fig. 3. It is clear that the
performance of the catalysts significantly decreases with an
increase in glycerol concentration. This effect appeared, how-
2
ever, to be less pronounced for the 0.1% Au/SiO
2
system, where
2
catalyst supports, i.e., SiO and C, if reacting with the relatively
the largest conversions were observed independent of glycerol
viscous aqueous glycerol solutions.

M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
115
Table 3
a
2
Catalytic performance of SiO and C supported Au NPs in diluted glycerol solutions at 80 °C.
Catalyst
TONb
TOFb (hÀ1
)
Conv. (mol%)
Selectivity to productsc (mol%)
Yield to acetic acid (%)
GLA
GLAD
GA
0.7
4.1
0.3
TA
AA
FA
OS
5.0
11.6
8.6
1
2
3
4
0.1% Au/SiO
1.0% Au/SiO
1.0% Au/f-SiO
1.0% Au/C
2
12,411
570
1335
504
517
24
56
87.5
40.2
94.1
35.5
5.5
0
0
19.5
4.8
0
0
0
0.5
0
54.8
79.5
90.6
0
14.5
0
0
48.0
32.0
85.3
0
2
d
2
21
7.3
40.1
10.9
34.4
7.3
a
b
c
1
.0 mol/L of glycerol in the reaction mixture (glycerol/H
2 2
O molar ratio 1:7), 20 mg of catalyst (0.2–2 lmol Au), 80 °C, 24 h, 770 rmp.
Turnover number (TON) or turnover frequency (TOF) based on the total gold content in the material.
GLA – glyceric acid, GLAD – glycolaldehyde, GA – glycolic acid, TA – taratronic acid, AA – acetic acid, FA – formic acid, OS – others.
Fumed silica OrisilÒ 380
d
Table 4
a
2
Catalytic performance of SiO supported Au NPs in undiluted glycerol solutions at 80 °C.
Catalyst
TONb
TOFb (hÀ1
)
Conv. (mol%)
Selectivity to productsc (mol%)
Yield to acetic acid (%)
GLAD
HPA
GLA
GA
AA
FA
OS
1
2
3
4
5
6
7
8
0.1% Au/SiO
1.0% Au/SiO
1.25% Au/f-SiO
1.0% Au/C
C
2
26,932
1606
1875
497
–
–
–
–
1122
67
78
21
–
–
–
–
40.2
24.0
35.0
7.4
21.5
17.0
17.4
18.7
14.2
0
8.4
12.5
21.8
29.3
33.3
30.4
0
0
3.7
0
0
0
7.1
15.2
1.5
0
0
0
0
6.2
1.8
0
0
0
48.6
29.9
80.3
0
0
0
20.9
0
1.9
14.8
70.1
3.8
50.0
25.5
17.0
16.7
10.9
19.5
7.2
28.1
0
0
0
2
2
1.9
31.3
3.6
4.9
4.8
4.4
0
47.3
48.8
38.1
39.1
SiO
f-SiO
Nonea
2
2
0
0
0
0
a
b
c
4
.5 mol/L of glycerol in the reaction mixture (glycerol/H
Turnover number (TON) or turnover frequency (TOF) based on the total gold content in the material.
GLA – glyceric acid, GA – glycolic acid, GLAD – glycolaldehyde, HPA – hydroxypyruvic acid, AA – acetic acid, FA – formic acid, OS – others.
2 2
O molar ratio 1:1.5), 20 mg carrier or catalyst (0.2–2.5 lmol Au), 80 °C, 24 h, 770 rmp.
2 2 2
Fig. 3. Conversion of glycerol vs. glycerol concentration, 0.1% Au/SiO ( ), 1.0% Au/SiO ( ) and 5.0% Au/SiO ( ). Reaction conditions: 20 mg of catalyst (0.2–10.0 lmol Au),
8
0 °C, 24 h, 770 rpm.
Additionally, we performed blind experiments without Au NPs
to carefully test glycerol reactivity in the systems investigated
Table 4, entries 5–8). It is interesting that glycerol conversion
mentioning that the addition of acetonitrile allowed us to decrease
the reaction temperature to 60 °C vs. 80 °C in acetonitrile-free con-
ditions; however, a nominal glycerol concentration after the addi-
tion of acetonitrile was lower 2.7 mol/L vs. ca. 4.5 mol/L in the
acetonitrile-free conditions. Moreover, the addition of acetonitrile
(Table 5, entries 2 (+) and 4 (À)) also increases the selectivity of
(
can reach about 20% in such conditions, but the reaction products
appeared to be completely different than those formed in the pres-
ence of Au NPs (Table 4, entries 1–4). Non-catalytic and quasi-
homogeneous glycerol oxidation in aerobic conditions was investi-
gated recently by Skrzynska et al. [17], who also investigated oxi-
dation in non-catalytic glycerol solutions (but basic) at high
temperatures and similarly noted different reaction paths for cata-
lytic and non-catalytic processes. An interesting fact is that, in our
systems of undiluted solutions, a conversion can be higher without
Au (Table 4, entries 5–8) than for C-supported Au NPs (Table 4,
entry 4). However, completely different products were formed in
2
the AA formation for the f-SiO -supported Au system.
Paradoxically, independent of glycerol dilution, higher conver-
sion and AA selectivities were observed for 0.1% Au NPs (Table 2,
entry 1 or Table 4, entry 1) and not for the catalyst containing
higher Au contents 1–10% (Table 2, entries 2 and 4–6 or Table 4,
entries 1 and 3). This seems to prove a heterogeneous character
of the process and the importance of the specific highly porous
2
structure of the 0.1% Au/SiO system, where Au is placed in pores.
these processes, i.e., formic acid for Au free, vs. AA for SiO
2
-sup-
This is discussed in Section 3.1.
ported Au (Table 4, entries 1–3), or C
Au (Table 4, entry 4).
In non-diluted glycerol solutions, the selectivity to AA can be
increased under the addition of acetonitrile (Table 5). It is worth
3
oxygenates for C-supported
It has been recently reported that Au/Cu nanoparticles can form
an interesting catalyst for the oxidation of various alcohols [23].
Thus, we prepared the related nano-Au/Cu and nano-Au/Ni
catalysts. The catalytic performance of these novel Cu- and

1
16
M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
Table 5
Catalytic performance of SiO
2
supported Au NPs in undiluted glycerol solutions at 60 °C with (+) and without (À) acetonitrile.
Catalyst
3
CH CN
Conv. (mol%)
Selectivity to productsc (mol%)
Yield to acetic acid (%)
GLA
GA
GLAD
HPA
AA
FA
OS
0.5
5.4
5.7
a
1
2
3
4
1.0% Au/SiO
1.25% Au/f-SiO
1.0% Au/SiO
1.25% Au/f-SiO
2
+
40.2
23.6
14.9
33.7
0
0
0
7.9
0
0
0
0
21.6
0
0
0
14.8
99.5
88.2
91.4
0.7
0
4.8
0
40.0
20.9
13.6
0.3
a
2
2
+
1.6
2.9
0
b
2
À
b
À
41.3
13.7
a
b
c
2
.7 mol/L of glycerol in the reaction mixture (glycerol/H
Glycerol: 4.5 mol/L (glycerol/H molar ratio 1:1.5).
GLA – glyceric acid, GA – glycolic acid, GLAD – glycolaldehyde, HPA – hydroxypyruvic acid, AA – acetic acid, FA – formic acid, OS – others.
2 2
O /acetonitrile molar ratio 1:1.5:2.8), 20 mg catalyst (2.0–2.5 lmol Au), 60 °C, 24 h, 770 rmp.
2 2
O
Table 6
a
Catalytic performance of bimetallic Cu or Ni supported Au NPs.
Catalyst
Conv. (mol%)
TONb
TOFb (hÀ1
)
Selectivity to productsc (mol%)
Yield to acetic acid (%)
HPA
GLA
GLAD
GA
AA
FA
1.4
9.1
19.1
18.0
OS
4.3
3.1
31.2
36.1
1
2
3
4
1.0% Au/Cu
1.0% Au/Ni
Cu
41.2
39.6
40.5
35.7
2762
2657
–
115
111
–
5.0
0
0
0
0
0
16.2
0
3.6
0
2.2
0
85.7
80.2
3.4
35.3
31.8
1.4
7.6
27.9
13.5
Ni
–
–
28.8
3.6
1.3
a
b
c
4
2 2
.5 mol/L of glycerol in the reaction mixture (glycerol/H O molar ratio 1:1.5), 20 mg carrier or catalyst (0.2–2.5 lmol Au), 80 °C, 24 h, 770 rmp.
Turnover number based on the total gold content in the material – Turnover frequency based on the total gold content in the material.
GLA – glyceric acid, GA – glycolic acid, GLAD – glycolaldehyde, HPA – hydroxypyruvic acid, AA – acetic acid, FA – formic acid, OS – others.
Table 7
Catalytic performance of 0.1% Au/SiO
2
at room temperature,a if enhanced by surfactant addition.
Surfactant
Conv. (mol%)
TONb
TOFb (hÀ1
)
Selectivity to productsc (mol%)
Yield to acetic acid (%)
AA
OS
1
2
3
4
Sulforokanol
Triton X-100
PEG 400
8.8
7.9
9.4
4.7
5876.1
5312.6
6278.5
3192.9
244.8
221.4
261.6
133.0
68.8
88.4
80.6
30.2
31.2
11.6
19.4
69.8
6.0
7.0
7.6
1.4
None
a
b
c
4
2 2
.5 mol/L of glycerol in the reaction mixture (glycerol/H O molar ratio 1:1.5) surfactant (0.05 wt.%), 25 °C, 24 h, 770 rpm.
Turnover number (TON) or turnover frequency (TOF) based on the total gold content in the material.
AA – acetic acid, OS – others.
Ni-supported Au catalysts in glycerol oxidation in non-diluted
solutions is given in Table 6. Basically, the results are slightly better
In the Au/Ni bimetallic catalyst, a structure of the Ni 2p XPS
multiplet indicates Ni as the main component with a small
2 3
O
than those obtained for the similar Au/SiO
2
2
was observed. In particular, AA was the main product, while a for-
mation of glyceric acid (GLA) was not observed. Similar to the SiO
2
system (Table 4, entry
). Maximal TON for the bimetallic Au/Cu system amounted to
762, while high selectivity toward a formation of the C products
amount of NiO (Fig. 2a). The reaction did not change this. The sit-
uation is similar for the Cu support where two oxidation states can
be detected: a relatively weak contribution from CuO represented
within the Cu 2p3/2 component by the line with binding energy of
about 935 eV, and the characteristic broad satellite structure situ-
ated in the region 940–945 eV (Fig. 2b). Their counterparts are vis-
ible within the Cu 2p1/2 line. The dominant contribution comes
2
2
support, we performed a blind test performing oxidation on Cu or
Ni (Table 6, entries 3 or 4). This provided similar conversions to the
Au-catalyzed reaction (Table 6, entries 1 or 2) but with much lower
selectivities. Moreover, other products were yielded then those
obtained in the catalytic process.
Although Au supported on Cu or Ni gave repeatable oxidation
results, these catalysts appeared to be consumed during the reac-
tion. After about three hours, we observed that the reaction mix-
ture color changed to green for nickel or blue for copper, which
suggested oxidation of Cu or Ni supports accompanied by their dis-
solution. In order to check the amount of released metal, we deter-
mined the concentrations of Cu and Ni in the reaction mixture
shown in Fig. 2 and Table S2 (Supplementary material). Thus, a
process of metal digestion accompanied oxidation to carboxylic
2
from either Cu O or Cu. The positions of the lines for metallic Cu
and oxide I are roughly the same [34], and we cannot distinguish
between these two oxidation states. The reaction led to a slight
change in the ratio between two oxidation states as it partly
reduced CuO. Its contribution, as derived from fitting, decreased
from about 46% to 37%.
Thus far, the results of the exhaustive oxidation of glycerol indi-
cated that catalyst availability is of major importance for the
observed performance of the system. Below, we report the results
of the experiments in which we tested possible applications of the
surfactants to improve this feature (Table 7). We observed that a
conversion and the yield of AA can be almost doubled at room tem-
perature when Sulforoktanol, PEG 400, or Triton X-100 were used
as surfactants.
2 2
acids, which in the presence of H O , possibly through peroxyacids,
can oxidize the supporting metals and digest them into the solu-
tion. A similar effect was described earlier [33]. An analysis of
the Cu redox behavior indicates that this can be a complex process
that includes Cu oxidation and comproportionation [31,32].
In a series of additional experiments, we tested the catalytic
2
stability of the SiO -supported systems during glycerol oxidation
in the liquid phase. The catalyst separation was facilitated via

M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
117
Fig. 4. Catalyst recycling: the change of glycerol conversion (solid lines) and yield to acetic acid (dotted lines) for 0.1% Au/SiO
2
(
) and 1.0% Au/SiO
2
(
) with recycle number.
Reaction conditions: glycerol concentration 0.2 mol/L of glycerol concentration in reaction mixture (glycerol/H
0 °C, 24 h, 770 rpm.
2 2
O molar ratio 1:37), 20 mg of catalyst (0.2–2.0 lmol Au),
8
centrifuge assistance. A facile recycling method involved catalyst
filtration, washing, and drying. The results of these experiments
are detailed in Fig. 4. Catalyst reusability dependent on the individ-
found to be capable of a many-fold enhancement of the catalyst
activity. However, this was relatively low at the room temperature
highly viscous liquid phase.
ual Au/SiO
system. In this case, the initial 100% conversion was observed in
cycles and decreases to ca. 60% after 4 cycles. However, the AA
yield decreased already in the second cycle to ca. 70%, while the
Au content after 4 cycles decreased to 25%–10% (EDXRF) of the ini-
tial Au content in the catalyst.
2
system appeared to be the best for the 0.1% Au/SiO
2
2
In summary, SiO -supported Au NPs can form an interesting
catalytic system for deep selective glycerol oxidation to acetic acid
in undiluted viscous liquid solutions. This seems especially inter-
esting for the processing of glycerol wastes.
2
Acknowledgments
The research was co-financed by the National Research and
Development Center (NCBiR) under Grant ORGANOMET no:
PBS2/A5/40/2014 Maciej Kapkowski and Mateusz Korzec appreci-
ates the support of the Doktoris fellowships.
4
. Conclusions
We assumed that a high catalyst reactivity accompanied by the
support wettability and availability should be of major importance
in viscous glycerol solutions. Thus, we investigated as a potential
catalyst for glycerol processing SiO
be compatible with the polar reactant, oxidant, and solvent, i.e.,
glycerol, H , and H O, respectively. Such a catalyst has not been
tested previously in glycerol processing. In fact, tested Au/SiO sys-
Appendix A. Supplementary material
2
-supported Au NPs that should
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2014.08.003.
O
2 2
2
2
tems appeared more reactive than Au/C systems. Since Au/Cu
nanoparticles have been recently reported as an interesting cata-
lyst for the oxidation of various alcohols, we prepared bimetallic
catalysts here. We performed a broad series of experiments in
References
[
1] C.J. Li, B.M. Trost, PNAS 105 (2008) 13197.
[2] N. Dimitratos, J.A. Lopez-Sanchez, G.J. Hutchings, Top. Catal. 52 (2009) 258.
[
[
3] M. Besson, P. Gallezot, C. Pinel, Chem. Rev. 114 (2014) 1827.
4] C.H. Zhou, H. Zhao, D.S. Tong, L.M. Wu, W.H. Yu, Catal. Rev. – Sci. Eng. 55 (2013)
2
order to compare Au/SiO , Au/Cu, and Au/Ni systems, some of
which were prepared by the innovative nano-Au transfer method.
Bimetallic Au/Cu and Au/Ni catalysts appeared relatively efficient;
however, they were destroyed during the reaction by the Cu or Ni
leaching effect.
3
69.
[5] S. Philippe, P. Karine (Eds.), Nanomaterials in Catalysis Introduction, in: S.
Philippe, P. Karine (Eds.), Concepts in Nanocatalysis, Viley-VCH, Weinheim,
2
013, pp. 1–5.
6] Y. Zhang, X. Cui, F. Shi, Y. Deng, Chem. Rev. 112 (2012) 2467.
[7] C.D. Pina, E. Falletta, M. Rossi, Chem. Soc. Rev. 41 (2012) 350.
[
We observed that, within the complex network of possible reac-
[
8] M.D. Hughes, Y.J. Xu, P. Jenkins, P. Mcmorn, P. Landon, D.I. Enache, F. Carley,
G.A. Attard, G.J. Hutchings, F. King, E.H. Stitt, P. Johnston, K. Griffin, C.J. Kiely,
Nature 437 (2005) 1132.
[9] A.S.K. Hashmi, G.J. Hutchings, Angew. Chem. Int. Ed. 45 (2006) 7896.
10] P. Bujak, P. Bartczak, J. Polanski, J. Catal. 295 (2012) 15–21.
tions, a process of deep glycerol oxidation proceeded on Au/SiO
catalysts preferentially to acetic acid. The selectivity of glycerol
oxidation to C products was investigated thoroughly. It is not a
coincidence, because only C oxygenates can form the single carb-
onous products of glycerol processing, while any oxidation to C
must also yield the C oxygenate (C ? C + C ). This means that,
2
3
[
[
3
11] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, G.J. Hutchings, Chem. Commun.
2
7
(2002) 696.
1
3
2
1
[12] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.J. Hutchings, Phys.
Chem. Chem. Phys. 5 (2003) 1329.
13] R. Schoevaart, T. Kieboom, Top. Catal. 27 (2004) 1.
14] J. Ma, W. Yu, M. Wang, X. Jia, F. Lu, J. Xu, Chin. J. Catal. 34 (2013) 492–507.
[15] S. Gil, M. Marchena, C.M. Fernández, L. Sánchez-Silva, A. Romero, J. Luís
Valverde, Appl. Catal. A 450 (2013) 189.
although the process can be fully selective, we obtain at least
two products.
[
[
Thus, the reactivity to C
contrast to the C and C products. Herein, we described the sys-
tems providing high selectivity to acetic acid, the C glycerol oxy-
3
products has been well described, in
1
2
[
16] B. Katryniok, H. Kimura, E. Skrzy n´ ska, J.S. Girardon, P. Fongarland, M. Capron,
2
R. Ducoulombier, N. Mimura, S. Paul, F. Dumeignil, Green Chem. 13 (2011)
genate. High conversions (100%) and acetic acid yields (90–99%)
were observed for the best catalysts in the diluted aqueous glycerol
solutions. Although in a relatively viscous liquid phase these values
decreased to ca. 40% and 20%, the addition of acetonitrile could
improve the acetic acid yield to ca. 40%, while surfactants were
1
960.
[17] E. Skrzy n´ ska, J. Ftouni, J.S. Girardon, M. Capron, L. Jalowiecki-Duhamel, J.F.
Paul, F. Dumeignil, ChemSusChem 5 (2012) 2065–2078.
18] M. Sankar, N. Dimitratos, D.W. Knight, A.F. Carley, R. Tiruvalam, C.J. Kiely, D.
Thomas, G.J. Hutchings, ChemSusChem 2 (2009) 1145.
[
[19] A. Villa, G.M. Veith, L. Prati, Angew. Chem. Int. Ed. 49 (2010) 4499.

1
18
M. Kapkowski et al. / Journal of Catalysis 319 (2014) 110–118
[
[
[
[
20] S. Gil, L. Muñoz, L. Sánchez-Silva, A. Romero, J.L. Valverde, Chem. Eng. J. 172
2011) 418–429.
21] A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem. 10 (2008)
3.
22] M. Korzec, P. Bartczak, A. Niemczyk, J. Szade, M. Kapkowski, P. Zenderowska, K.
Balin, J. Lel a˛ tko, J. Polanski, J. Catal. 313 (2014) 1.
23] Y. Sugano, Y. Shiraishi, D. Tsukamoto, S. Ichikawa, S. Tanaka, T. Hirai, Angew.
Chem. Int. Ed. 52 (2013) 5295.
24] K.S. Rao, J. Colloid Interface Sci. 289 (2005) 125.
25] H. Okudera, A. Hozumi, Thin Solid Films 434 (2003) 62.
26] M. Pagliaro, M. Rossi, The Future of Glycerol, second ed., RSC Publishing
Cambridge, 2010.
[27] M. Massa, A. Andersson, E. Finocchio, G. Busca, J. Catal. 307 (2013) 170.
[28] B. Katryniok, S. Paul, F. Dumeignil, ACS Catal. 3 (2013) 1819.
[29] J. Deleplanque, J.L. Dubois, J.F. Devaux, W. Ueda, Catal. Today 157 (2010) 351.
[30] H. Skaff, T. Emrick, in: V. Rotello (Ed.), Nanoparticles: Building Blocks for
Nanotechnology, Springer Science and Business Media Inc., New York, 2004, p.
32.
[31] N.H. Nguyen, V. Percec, J. Polym. Sci., Part A: Polym. Chem. 49 (2011) 4241.
[32] K.L. Chavez, D.W. Hess, J. Electrochem. Soc. 148 (2001) 640.
[33] S. Sahu, N.C. Kauri, M. Kundu, Braz. J. Chem. Eng. 28 (2011) 251.
[34] M.C. Biesinger, L.W.M. Lau, A.R. Gerson, R.St.C. Smart, Appl. Surf. Sci. 257
(2010) 887.
(
1
[
[
[