Glycerol Oxidation Using MgO- and Al2O3-supported Gold and Gold–Palladium Nanoparticles Prepared in the Absence of Polymer Stabilizers

Article abstract of DOI:10.1002/cctc.201800074

Au and AuPd nanoparticles supported on MgO and Al₂O₃ were employed for the selective aqueous phase oxidation of glycerol under basic conditions. Catalysts were prepared by sol-immobilization without the addition of a stabilizing agent such as polyvinyl alcohol (PVA), which is generally added to stabilize the noble metal sol prior to immobilization. The obtained materials prepared with and without stabilizing agent were active for glycerol oxidation and showed similar catalytic performances—implying that the stabilizing polymer is not required to obtain active materials. Depending on the support used, it was possible to tailor the selectivity towards the desired oxidation products by using catalysts prepared with or without stabilizing agent. PVA-free Au/γ-Al₂O₃ exhibited a remarkably high selectivity towards tartronic acid (40 % at 97 % conversion), which was not observed for Au/γ-Al₂O₃ prepared with PVA (27 % at isoconversion). Selective glycerol oxidation performed under base-free conditions over AuPd/MgO catalysts also corroborated the previous results that the presence of a stabilizing polymer is not required to prepare active catalysts by sol-immobilization. Thus, a facile way to circumvent the inherent drawbacks encountered by the use of polymer stabilizers during catalyst preparation is presented herein. Experimental results suggest that the presence of the polymer stabilizers can affect the reaction pathways and control selectivity.

Full text of DOI:10.1002/cctc.201800074

Accepted Article
Title: Glycerol oxidation using MgO and Al2O3 supported gold and
gold-palladium nanoparticles prepared in the absence of polymer
stabilisers
Authors: Georgios Dodekatos, Laura Abis, Simon Freakley, Harun
Tüysüz, and Graham John Hutchings
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201800074
Link to VoR: http://dx.doi.org/10.1002/cctc.201800074
A Journal of
www.chemcatchem.org

10.1002/cctc.201800074
ChemCatChem
Glycerol Oxidation using MgO and Al2O3 supported gold and
gold-palladium nanoparticles prepared in the absence of polymer
stabilisers
Georgios Dodekatosa, Laura Abisb, Simon J. Freakleyb, Harun Tüysüza and Graham J. Hutchingsb.
a
Max-Planck-Institut für Kohlenforschung, Kaiser-Wilhelm-Platz 1, D-45470 Mülheim an der Ruhr,
Germany.
b Cardiff Catalysis Institute, Cardiff University, Main Building, Park Place, Cardiff CF103AT.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Abstract
Au and AuPd nanoparticles supported on MgO and Al2O3 supports were employed for the
selective aqueous phase oxidation of glycerol under basic conditions. Catalysts were prepared
by sol-immobilisation without the addition of a stabilizing agent such as polyvinyl alcohol
(PVA), which is generally added to stabilize the noble metal sol prior to immobilisation on.
The obtained materials prepared with and without stabilizing agent were active for glycerol
oxidation and showed similar catalytic performances – implying that the stabilizing polymer
is not required to obtain active materials. Depending on the support used, it was possible to
tailor the selectivity towards desired oxidation products by using catalysts prepared with or
without stabilizing agent. PVA-free Au/-Al2O3 exhibited a remarkably high selectivity
towards tartronic acid (40% at 97% conversion), which was not observed for Au/-Al2O3
prepared with PVA (27% at isoconversion). Selective glycerol oxidation performed under
base-free conditions over AuPd/MgO catalysts also corroborated the previous results that the
presence of a stabilizing polymer is not required to prepare active catalysts by sol-
immobilisation. Thus, a facile way is presented herein to circumvent the inherent drawbacks
encountered by the use of polymer stabilizers during catalyst preparation and the
experimental results in fact suggest that the presence of the polymer stabilisers can affect the
reaction pathways and control selectivity.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Introduction
As a by-product of biodiesel production glycerol represents a highly functionalised molecule
which can itself be considered a starting material for the production of many industrially
desirable products.[1-3] With the shift towards biomass derived fuels the selective
functionalization of by-products from these processes will emerge as a key technology in
achieving greener chemical production in the future. Selective oxidation using heterogeneous
catalysis can provide a route from glycerol to valuable oxidised products such as glyceric
acid, tartronic acid and dihydroxyacetone.[4-5] In these processes, the catalyst activity and
selectivity is determined by factors such as the choice of metal, the control of the particle
size, the nature of the support material and the pH of the reaction solution.
Monometallic Au[6-9] and bimetallic AuPd[10-12] nanoparticles have been extensively
studied for the selective oxidation of glycerol under basic and base-free[13-15] conditions
where high activity can be achieved with high selectivity towards C3 oxygenates such as
glyceric and tartronic acid. It has also been shown that Au/TiO2 catalysts are capable of
producing vinyl chloride monomers out of alkynes.[16] Commonly, the catalysts are prepared
by a sol-immobilisation method[17] where solutions of colloidal Au and AuPd nanoparticles
are formed prior to immobilisation on a high surface area support material such as TiO2 or
MgO. Usually the solutions are stabilised with polymer additives such as
polyvinylpyrrolidone (PVP) or polyvinyl alcohol (PVA) to prevent coalescence of particles
during the preparation. This method of preparing catalysts allows control of the nanoparticle
synthesis in solution prior to immobilisation by varying the amount of polymer added and the
rate of reduction using strong reductants such as NaBH4.[5, 18] However, the presence of these
polymers on the metal surface can lead to several disadvantages, since they can partially or
completely block the access of the reactant to the metal surface, decreasing the catalyst
activity, and heat treatments or refluxing treatments are typically used to remove polymers
before use.[19]
Recently, the preparation of gold and gold palladium nanoparticles by colloidal methods in
the absence of polymer stabilisers has been reported for both quasi-homogeneous catalysts in
solution phase reduction of 4-nitrophenol[20] and supported catalysts for selective
oxidation[21]. We reported that despite a larger particle size distribution, catalysts prepared
without the addition of polymers were similar in activity to catalysts prepared by the
extensively reported method.[21] However, the selectivity was affected by the absence of the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
polymer stabilisers suggesting that the presence of polymer additives can control the reaction
pathways occurring on the metal particle surface.
Here, we report the synthesis of supported monometallic and bimetallic nanoparticles of
Au and AuPd by the sol-immobilisation method previously reported in the literature but
without the addition of any stabiliser polymer using MgO and Al2O3 as support materials and
reveal the extent of these effects on the catalytic performance for glycerol oxidation is highly
dependent on the choice of support.
Results
We initially prepared Au colloidal solutions both in the presence and absence of a PVA
stabiliser (PVA:metal = 0.65 by weight) before immobilisation on various MgO (NanoActive
MgO plus (MgOnano), Nanoscale Corporation, and MgO BDH (MgOBDH), Analar) and Al2O3
(-Al2O3, Aldrich, and -Al2O3, Alfa Aesar) support materials to give a nominal metal
loading of 1 wt%. This method was investigated for the general applicability of preparing
catalyst materials in the absence of polymers on supports other than TiO2, since we could
show that TiO2 supported Au and AuPd NPs prepared without capping agent exhibited
similar catalytic performances compared to the catalysts prepared with capping agent for
benzyl alcohol and glycerol oxidation.[21] In addition, bimetallic AuPd catalysts were also
prepared on MgOnano with and without the addition of PVA. As depicted in
a)
b)
Au/MgOnano wo PVA
Au/MgOBDH wo PVA
AuPd/MgOnano wo PVA
Au/-Al2O3 wo PVA

Au/-Al2O3 wo PVA
-Al2O3
 -Al2O3
Au
Mg(OH)2
MgO
Au
10
20
30
40
50
60
70
80
10
20
30
40
50
60
70
80
2  / °
2  / °
Figure 1a, Au and AuPd NP catalysts supported on MgO mainly consisted of hydrated
Mg(OH)2, which is due to the aqueous conditions of the sol-immobilisation procedure and is
consistent with former reports.[14] The pristine supports prior to noble metal loading showed
differing levels of hydration (Supporting Information, Figure S1a and b), which were
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
transformed to Mg(OH)2 after the catalyst synthesis. No differences in the diffraction patterns
of the catalyst materials can be detected between syntheses conducted with and without PVA
(Supporting Information, Figure S1). Characteristic principle reflections for the Au and AuPd
alloy metals cannot be observed due to the overlap with the (011) reflection of Mg(OH)2 and
the low amount of metal loaded. Nonetheless, a weak reflection around 45° for Au/MgOnano
and to a lesser extent for Au/MgOBDH prepared without PVA shows that Au is deposited onto
the samples. The reflection is not visible for samples prepared with PVA (Supporting
Information, Figure S1 a and b) which indicates that larger Au crystallites are deposited on
the surface of the surfactant-free catalysts, as it is also evidenced by TEM analysis
In the case of the Al2O3 supports, the - and -phases can readily be identified from the
diffraction patterns obtained for the catalysts prepared without PVA (Figure 1b). Both
supports were loaded with Au NPs and the reflections for Au can clearly be observed for
Au/-Al2O3 but remain undetected for the -phase. It is important to note that for the -Al2O3
the Au reflections were not observed for the sample prepared with PVA (Supporting
Information, Figure S1c), which indicates a lower Au loading compared to the sample
prepared without PVA and is discussed below in more detail.
a)
b)
Au/MgOnano wo PVA
Au/MgOBDH wo PVA
AuPd/MgOnano wo PVA
Au/-Al2O3 wo PVA

Au/-Al2O3 wo PVA
-Al2O3
 -Al2O3
Au
Mg(OH)2
MgO
Au
10
20
30
40
50
60
70
80
10
20
30
40
50
60
70
80
2  / °
2  / °
Figure 1. XRD patterns of a) Au/MgO and AuPd/MgO materials, and b) Au/Al2O3 materials. “wo PVA” indicates that no
capping agent was used during the catalyst synthesis.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Table 1 reports the elemental analysis via MP-AES of the catalyst samples after digestion
in aqua regia. The results show no significant differences in the gold loading for samples for
each MgO support material prepared with PVA and without PVA suggesting that the absence
of polymer does not hinder the immobilisation step significantly. The AuPd/MgOnano sample
prepared without PVA shows an 1:1 molar ratio determined by MP-AES for Au and Pd as
expected and is comparable to the sample prepared with PVA stabilizer.
Table 1. Elemental analysis via MP-AES for Au and AuPd NPs supported on various
MgO and Al2O3 supports prepared with (w) and without (wo) PVA during sol-
immobilisation.
Sample
Nominal amount Au
Elemental analysis
or Au/Pd (wt% /
Au / wt% or Au/Pd
wt%)
(wt% / wt%)
Au/MgOnano w PVA
Au/MgOnano wo PVA
Au/MgOBDH w PVA
Au/MgOBDH wo PVA
Au/-Al2O3 w PVA
Au/-Al2O3 wo PVA
Au/-Al2O3 w PVA
Au/-Al2O3 wo PVA
AuPd/MgOnano w PVA
AuPd/MgOnano wo PVA
1
1
1
1
1
1
1
1
0.93
0.99
0.81
0.80
0.1
0.93
1
0.86
0.65 / 0.35
0.65 / 0.35
0.63 / 0.33
0.73 / 0.38
Interestingly, the deposition of Au NPs was not quantitative on -Al2O3 when PVA was
used as stabilizing agent (
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Table 1). Only 0.1 wt% Au was determined by MP-AES instead of the expected 1 wt%.
The poor immobilisation of Au on -Al2O3 was also evident during the synthesis procedure.
Discolouration of the Au sol after adding the support was not observed to the same extent as
for the other materials, which indicates an incomplete immobilisation. Alteration of the pH
during the immobilisation procedure did not improve the deposition of Au on -Al2O3. On
the other hand, omitting PVA under otherwise identical preparation conditions resulted in an
almost complete deposition of Au. This implies that the absence of a capping agent can also
be beneficial in achieving target metal deposition for the catalyst synthesis via sol-
immobilisation on certain support materials. Apparently, the interaction of the naked Au NPs
with the -Al2O3 is stronger than the PVA-capped Au NPs. However, it is shown later that
even the low-loaded PVA-capped Au/-Al2O3 catalysts exhibited a remarkable high catalytic
performance for glycerol oxidation. In the case of -Al2O3, no distinct differences in
immobilisation were observed between stabilizer-free and PVA-capped Au NPs. Error!
Reference source not found. depicts representative XPS profiles for AuPd/MgOnano catalysts
between 320 and 370 eV and for Au/MgOnano catalysts between 80 and 100 eV. XPS analysis
of the nature of the catalyst surface is challenging due to the overlap of the Pd 3d /Au 4d
region with the Mg KLL Auger structure and also the overlap of the Mg 2s region with the
Au 4f photoemission peaks. Nevertheless, by comparison with similarly treated neat MgO
samples, we can be confident of the Mg 2s /Au 4f fitting, which reveals solely metallic gold
(binding energy between 83.2 and 83.9 eV; see Table 2) in all samples irrespective of the
presence of PVA in the preparation. The shift of the Au 4f7/2 binding energies to lower values
generally is an indication for ultra-small clusters (< 1 nm) or for flat, unidimensional Au
particles[22], which can be excluded in our case as evidenced by TEM analysis. Another
reason lies in the interaction between the Au NPs and the support which seems to be the case
for the Au/MgO samples. The electron transfer from the support to the Au NP should result
in the lowered 4f7/2 binding energies[22] and was previously observed for Au/MgO and
Au/MgAl-LDH catalysts.[23-24] This effect is more pronounced for MgOBDH than for the
MgOnano support (Table 2).
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
a)
b)
Mg Auger
Pd 3d
Mg Auger
Au 4f5/2
Au 4f7/2
Pd 3d
Au 4d
Mg 2s
Au 4d
AuPd/MgO w PVA
Au/MgO w PVA
AuPd/MgO wo PVA
MgO
Au/MgO wo PVA
MgO
370
360
350
340
330
320 110
105
100
95
90
85
80
75
Binding energy / eV
Binding energy / eV
Figure 2. Representative XPS spectra of a) Au 4d and Pd 3d for AuPd/MgOnano, and b) Au 4f for Au/MgOnano materials
prepared with (w) and without (wo) PVA during sol-immobilisation.
Due to the complications of fitting the Pd 3d region, it is not possible to accurately
determine the Pd surface concentrations. However, for those where it is possible to
confidently fit the Au 4d and Mg KLL Auger structure, it was determined that the Pd was in
the metallic form with a BE of around 335 eV (Error! Reference source not found.a and
Table 2).
Table 2. Physical parameters determined by TEM and XPS analysis for Au/MgO
and AuPd/MgO materials prepared with and without PVA during sol-
immobilisation.
Catalyst
Stabilizer
Mean Au
Binding energy / eV
particle size
Au 4f
Pd 3d
/ nm
2.6
4.3
3.0
6.5
7.9
6.7
Au/MgOBDH
Au/MgOnano
AuPd/MgOnano
PVA
none
PVA
none
PVA
none
83.2
83.3
83.9
83.8
83.8
83.6
-
-
-
-
335.1
334.8
Transmission electron microscopy (TEM) shown in Error! Reference source not found.
reveals that Au deposition on MgOnano, MgOBDH, and -Al2O3 without PVA results in bigger
nanoparticles compared to the samples prepared with PVA – with, in the case of the MgO
supports, a poorly shaped morphology. Samples prepared with PVA contained spherical like
particles which are uniform in shape. The mean particle diameter and the particle size
distributions are shown in Error! Reference source not found.c, f, i. Depending on the
support used, the mean Au particle size is slightly changed for the PVA-capped Au NPs.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Notably, -Al2O3 apparently stabilizes the PVA-free Au NPs better than the other supports.
Indeed, no deformed Au NPs are observed for these materials, which is not the case for the
other supports (see Supporting Information). The MgO and -Al2O3 supports (Supporting
Information, Figure S2) show besides spherical Au NPs, which were taken into account for
the histograms, also irregularly shaped NPs. This effect is even more pronounced for
AuPd/MgOnano catalysts (Supporting Information, Figure S3). The fact that the support plays
a pivotal role in stabilizing the noble metal NPs has been shown previously.[17] For instance,
PVA seems to be capable of stabilizing the size of Au NPs immobilized on active carbon.
Tetrakis(hydroxymethyl-)phosphonium chloride (THPC) on the other hand, a weaker
stabilizing agent, does not stabilize the particles on active carbon resulting in bigger Au NPs.
However, by using TiO2 as support, THPC is capable of maintaining the Au particle size after
immobilisation. In fact, AuPd/TiO2 samples prepared with and without PVA, as reported in
our previous publication[20], showed only minor changes in the particle size distribution.
Nonetheless, the same trend was observed that without stabilizing agent the AuPd NPs were
slightly larger. Analogous samples of AuPd NPs on MgOnano, however, show an inverted
behaviour with respect to particle size distribution (Supporting Information, Figure S3).
Samples prepared with PVA apparently have a bigger particle mean diameter compared to
samples prepared without PVA. Nonetheless, the PVA seems to have a stabilizing effect on
the shape of the AuPd NPs, whereas PVA-free Au NPs are highly irregular. This fact might
also distort the histogram for the particle size distribution since measuring the diameter for
non-spherical NPs is to some extent ambiguous.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Figure 3. TEM micrographs of a), b) Au/MgOnano, d), e) Au/MgOBDH, and g), h) Au/-Al2O3 prepared with (w) and without
(wo) PVA, respectively. c), f), i) Corresponding histograms of particle size distributions for Au NPs (at least 300 counts for
each sample).
Catalytic Glycerol Oxidation
The materials prepared with and without PVA were employed for selective aqueous phase
glycerol oxidation in order to investigate the effect of the polymer stabiliser on the catalytic
performance. The general reaction pathway is illustrated in Scheme 1, which is based on the
observations obtained in this study and of reaction pathways reported in the literature[4, 25]
.
The first oxidative dehydrogenation step forms glyceraldehyde or dihydroxyacetone, which
are in equilibrium. Glyceraldehyde is an unstable intermediate which is readily oxidized to
glyceric acid, which also can be further oxidized to form tartronic acid as consecutive
product. It has to be noted that, based on the catalytic results obtained herein and reported in
our previous publication,[21] tartronic acid is likely to be also a primary product formed at the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
catalyst surface. Subsequently, either from glyceric acid or eventually from tartronic acid, C-
C cleavage products are formed such as glycolic and oxalic acid. The formation of lactic acid
follows another route over a dehydration and Cannizaro rearrangement step.
Scheme 1. General and simplified reaction scheme for glycerol oxidation according to our results and ref.[25]
.
As shown in Figure 4a, no significant changes in glycerol conversion can be observed
between Au/MgO and Au/-Al2O3 – both catalysts reach a conversion of around 98% after
3 h reaction time. Furthermore, the effect of the capping agent used during the material
synthesis on the conversion seems to be negligible. This is noteworthy since TEM
characterization indicated that the preparation without capping agent resulted in larger, ill-
shaped Au NPs (Error! Reference source not found. and Supporting Information). It is
generally described in the literature that smaller Au NPs show an improved catalytic activity
24-26]
for glycerol oxidation.[8-9,
Hence, this observation might imply that, besides the
detrimental effect on the particle morphology, the absence of PVA can have a positive impact
on the catalytic performance.
Similar observations were made for Au/TiO2 catalysts prepared with and without PVA in our
previous report.[21] The activity of the catalysts reported there seems to be comparable to the
activity of the monometallic catalysts presented herein, which might imply that the support
does not play a pivotal role for the reaction. Nonetheless, it is shown below that by selecting
the right support remarkable differences can be observed in the selectivity toward the
products. Hence, the impact of the support cannot be neglected when this new synthesis
approach is used in order to prepare active catalysts for glycerol oxidation.
The selectivity towards glyceric and tartronic acid is not significantly changed for
Au/MgOBDH catalysts (Figure 4 b). On the other hand, for Au/-Al2O3 catalysts, the impact of
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
the capping agent on the selectivity is more obvious. The selectivity profiles depicted in
Figure 4 b) reveal that PVA-free catalysts show a higher selectivity towards tartronic acid at
the expense of glyceric acid. Interestingly, PVA-free Au/-Al2O3 exhibits after 5 h reaction
time a remarkably high selectivity of 42% towards tartronic acid compared to 33% for the
catalysts with PVA. It has to be emphasized that the selectivities obtained for tartronic acid in
this study resemble the selectivities reported by Sankar et al.[27] who achieved, as pointed out
by Davis et al.[28], the highest selectivity for tartronic acid over Au/TiO2 and Au/C catalysts
but at a much higher reaction temperature of 120 °C. These catalysts were prepared by the
sol-immobilisation method with PVA as stabilizing agent as presented here. Generally for
monometallic catalysts, similar high selectivities were observed for Pd/C catalysts.[29] For
bimetallic systems, the work of Villa et al. can be given as example where Bi-modified AuPd
catalysts exhibited high selectivities towards tartronic acid.[30] In this light, it is noteworthy to
mention that the absence of a stabilizing agent during the sol-immobilisation procedure can
alter the catalyst properties and tailor the selectivity towards desired products in glycerol
oxidation. However, it has to be pointed out that comparison of catalytic performances
between papers is difficult due to the different reaction parameters, the necessity of
comparing the selectivities at isoconversions, and the necessity of results free from mass
transport limitations.[28]
Further supports were employed to prepare active Au/MgO and Au/-Al2O3 catalysts for
glycerol oxidation. As shown in Table 3, Au/MgOnano catalysts prepared with and without
PVA show similar conversions after 3 h reaction time (reaction profiles are presented in the
Supporting Information, Figure S4 and S5). Interestingly, the presence of PVA increases the
selectivity towards tartronic acid to 22% compared to 14% for the PVA-free catalyst. This
trend is further enhanced after full consumption of glycerol, where it is apparent that more
tartronic acid is formed at the expense of glyceric acid (Supporting Information, Figure S4). –
which is in good accordance to the consecutive reaction pathway shown in Scheme 1. For
Au/-Al2O3, as shown before in Figure 4 and also summarized in Table 3, the opposite is the
case, where the absence of PVA during the synthesis results in a higher selectivity towards
tartronic acid. In the case of Au/-Al2O3, no differences in selectivities depending on the
presence or absence of PVA can be observed. Hence, it can be concluded that the effect of
the addition of the stabilizing agent on the catalytic performance seems to be highly
dependent on the support used.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Remarkably, 0.1 wt% Au/-Al2O3 catalysts prepared with PVA exhibited a similar
glycerol conversion as PVA-free 1 wt% Au/-Al2O3, although the Au loading for the former
was highly diminished (
Table 1, for reaction profiles see Supporting Information, Figure S5). With this
observation, not only could we show that a surfactant-free sol-immobilisation method allows
depositing higher amounts of Au on -Al2O3 (whereas with PVA this seems not possible; see
above), but we also demonstrated that low Au loadings of around 0.1 wt% are sufficient for
the PVA-capped Au/-Al2O3 catalyst to exhibit remarkably high glycerol conversions. It is
apparent that the speciation of the Au NPs plays the predominant role in the activity than the
actual loaded amount, which was demonstrated also for Au/TiO2 by Rogers et al.[31] By
carefully optimizing the sol-immobilisation preparation of the catalyst, Au clusters – which
were the minor species deposited on TiO2 – contributed remarkably to the catalytic
performance. A similar case might be true for 0.1 wt% Au/-Al2O3 presented here, where
primarily active Au species were deposited on the Al2O3 surface and surplus PVA-capped Au
NPs were washed off. This is corroborated by the fact that PVA-capped 0.1 wt% Au/-Al2O3
shows a higher mean Au particle size (7.8 nm) than PVA-free 1 wt% Au/-Al2O3 (5.2 nm;
Supporting Information, Figure S2). This implies that not only a decreased number of Au
NPs but also a strongly decreased number of smaller Au NPs, which can still be detected by
TEM analysis, is present for PVA-capped 0.1 wt% Au/-Al2O3. This is in contradiction to
the expectation that a high number of small Au NPs should lead to a better catalytic
performance for glycerol oxidation. Hence, it might also be in this case evident that the Au
speciation is the key parameter for a high activity – which, however, could not be
investigated in more detail in this study. Nonetheless, the results suggest that any higher
amount of Au loading contributes to the catalytic performance only to a negligible extent.
Generally, it is difficult to assign which Au particle size range of the catalysts presented is
the most active for glycerol oxidation. For instance, PVA-free and PVA-capped Au/MgO and
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
Au/-Al2O3 exhibited different mean Au particle sizes (Error! Reference source not
found. c, f, i) but essentially the same conversion profiles (Figure 4 and Supporting
Information Figure S4) whether or not PVA was present – with changes only for the
selectivities. Furthermore, no clear trend was observed for the change in selectivity with
respect to the stabilizer (see above). It has to be kept in mind that superimposing effects of
the stabilizer make it difficult to completely assign the catalytic results to an effect of the
particle size.
Bimetallic AuPd catalysts were also deposited on MgOnano and -Al2O3 to see if the
preparation method with or without PVA has an influence on the catalytic performance. As
shown in Figure 4c and d for AuPd/MgOnano and in Table 3 for AuPd/-Al2O3 (for reaction
profiles see Supporting Information Figure S6), the stabilizing agent plays no decisive role
for preparing a catalyst with high activity for glycerol oxidation. Especially in the case of
AuPd/MgOnano, it is obvious that both catalysts, with and without PVA during the synthesis,
behave essentially the same, which was also demonstrated for AuPd/TiO2 catalysts in an
earlier study conducted from some authors of the present work.[20] This is remarkable since
TEM analysis (Supporting Information, Figure S3) revealed that only with the use of PVA
well-formed AuPd NPs could be obtained. Also here, the particle size distribution seems to
play a minor role for the catalytic performance.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
TA | Au/MgOBDH w PVA
GA | Au/MgOBDH w PVA
TA | Au/-Al2O3 w PVA
GA | Au/-Al2O3 w PVA
TA | Au/MgOBDH wo PVA
GA | Au/MgOBDH wo PVA
TA | Au/-Al2O3 wo PVA
GA | Au/-Al2O3 wo PVA
a)
b)
70
60
50
40
30
20
10
0
100
80
60
40
20
0
Au/MgOBDH w PVA
Au/MgOBDH wo PVA
Au/-Al2O3 w PVA
Au/-Al2O3 wo PVA
1
2
3
4
5
0
1
2
3
4
5
Time / h
Time / h
c) 100
d)70
60
TA | AuPd/MgOnano w PVA
GA | AuPd/MgOnano w PVA
TA | AuPd/MgOnano wo PVA
GA | AuPd/MgOnano wo PVA
80
60
40
20
0
AuPd/MgOnano w PVA
AuPd/MgOnano wo PVA
50
40
30
20
10
0
0
1
2
3
4
5
1
2
3
4
5
Time / h
Time / h
Figure 4. Conversion and selectivity profiles of glycerol oxidation for a), b) 1 wt% Au/MgOBDH and 1 wt% Au/-Al2O3, and
c), d) AuPd (1:1 molar ratio) supported on MgOnano. Reaction conditions: 0.3 M glycerol, 2:1 NaOH/glycerol, 500:1
glycerol/metal (for Au), 385:1 glycerol/metal (for AuPd, based on Au), 10 mL, 60 °C, 3 bar O2, stirring speed 1200 rpm.
TA: tartronic acid; GA: glyceric acid. Hollow forms denote PVA-free catalysts; solid lines indicate TA selectivity and
dashed lines indicate GA selectivity.
Table 3. Summarized conversions, selectivities, and carbon mass balances for glycerol oxidation after 3 h reaction time over
catalysts employed in this study. For reaction conditions see the caption in Figure 4. Generally, catalysts contained approx. 1 wt%
Au loading if not otherwise specified.
Catalyst
Conversion
/ %
Selectivity / %
Oxalic acid
Carbon
mass
balance / %
Glyceric
acid
Tartronic
acid
Glycolic
acid
Formic acid
Lactic acid
Acetic acid
Au/MgOnano w PVA
Au/MgOnano wo PVA
98
92
52
60
22
14
3
5
3
2
2
4
9
8
5
94
97
10
Au/MgOBDH w PVA
Au/MgOBDH wo PVA
97
95
48
47
28
29
5
4
3
4
3
2
5
4
8
9
94
92
Au/-Al2O3 w PVAa)
Au/-Al2O3 wo PVA
76
84
67
65
7
8
8
5
1
1
6
4
9
3
5
98
99
11
98
97
55
43
27
40
4
3
3
6
3
2
3
1
6
5
95
98
Au/-Al2O3 w PVA
Au/-Al2O3 wo PVA
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
AuPd/MgOnano w PVA
100
35
32
38
35
2
2
9
8
1
2
10
15
5
6
89
90
AuPd/MgOnano wo PVA
1000
100
100
44
53
35
28
3
2
6
3
2
2
8
9
3
3
92
93
AuPd/-Al2O3 w PVA
AuPd/-Al2O3 wo PVA
a) Au loading for this sample was 0.1 wt% (see
Table 1).
It is highly desired to work under base-free conditions for glycerol oxidation. In an earlier
study, Hutchings and co-workers showed that MgO in conjunction with bimetallic AuPt or
AuPd NPs are suitable catalysts for performing glycerol oxidation at neutral pH.[14] Hence,
AuPd/MgO catalysts prepared with and without PVA were also employed for base-free
glycerol oxidation and the catalytic results are presented in Figure 5. It is clearly visible that
the PVA-free catalyst exhibited only a slightly inferior catalytic performance than the
AuPd/MgO prepared with PVA. Moreover, the selectivity profiles show the same behaviour
with only slight changes in selectivities towards glyceric acid. Also in this case it is proven
that the absence of PVA during the sol-immobilisation preparation of catalysts does not show
a detrimental impact on the catalytic performance for glycerol oxidation.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
a)
b)
80
70
60
50
40
30
20
10
0
60
40
20
TA | AuPd/MgOnano w PVA
GA | AuPd/MgOnano w PVA
DHA | AuPd/MgOnano w PVA
TA | AuPd/MgOnano wo PVA
GA | AuPd/MgOnano wo PVA
DHA | AuPd/MgOnano wo PVA
AuPd/MgOnano w PVA
AuPd/MgOnano wo PVA
0
0
1
2
3
4
5
1
2
3
4
5
Time / h
Time / h
Figure 5. a) Conversion and b) selectivity profiles for glycerol oxidation under base-free conditions over
1 wt% AuPd/MgOnano catalysts prepared with and without PVA during sol-immobilisation. Reaction conditions: 0.3 M
glycerol, 385:1 glycerol/metal (based on Au), 10 mL, 60 °C, 3 bar O2, stirring speed 1200 rpm. TA: tartronic acid; GA:
glyceric acid; DHA: dihydroxyacetone. Hollow forms denote PVA-free catalysts; solid lines indicate TA selectivity, dashed
lines indicate GA selectivity, and dotted lines indicate DHA selectivity.
Conclusions
The results presented in this study show that the absence of any stabilizing agent during
the sol-immobilisation method of catalyst preparation still produces active catalysts for
glycerol oxidation. In particular, it was shown that, depending on the support used, the
catalytic performance either remains the same independent of the presence or absence of
PVA as stabilizing agent or shows alterations in the product selectivity while maintaining
similar catalytic activity. This indicates that the supports can influence the metallic Au NPs
differently and, thus, result in different implications on the effect of the stabilizer on the
catalytic performances. Therefore, our method allows to prepare catalysts with unprecedented
catalytic properties. For instance, it was possible for Au/-Al2O3 catalysts prepared without
PVA to tailor the selectivity towards tartronic acid to remarkable 40% at nearly full
conversion after 3 h reaction time. On the other hand, PVA-capped or PVA-free Au/MgO,
AuPd/MgO, and AuPd/-Al2O3 catalysts exhibited only negligible deviations in the catalytic
performances. Up to now, mainly the effects of the type of stabilizing agent or of the used
amount on the catalytic performance were investigated. It is thus remarkable that the total
absence of a stabilizing agent during the sol-immobilisation synthesis still yields active
catalysts for glycerol oxidation, although the Au mean particle sizes were slightly increased
for the PVA-free catalysts. In addition, we demonstrated that PVA-free AuPd/MgO catalysts
are also active for glycerol oxidation under base-free conditions. It is thus shown that
detrimental effects by the protecting polymers, which in the past were generally inevitable for
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
the sol-immobilisation method, can be avoided by simply omitting their addition during the
catalyst synthesis. Finally, it was shown that the sol-immobilisation technique without PVA
allows a quantitative deposition of Au NPs on -Al2O3, whereas with PVA only a small
fraction of the Au sol was immobilized on the support; however, this small fraction revealed
to be highly active for glycerol oxidation. These results pave the way for new catalyst
syntheses, which might not have been possible by sol-immobilisation due to the poor noble
metal sol-support interaction.
Acknowledgements
This work is supported by MAXNET Energy research consortium of the Max Planck Society and the
Cluster of Excellence RESOLV (EXC 1069) funded by the Deutsche Forschungsgemeinschaft.
Experimental Section
Catalyst Preparation
HAuCl4·3H2O (Sigma-Aldrich, ≥49.0%), PdCl2 (Sigma-Aldrich, Reagent Plus® 99%), MgO
(MgOnano denotes NanoActive MgO Plus from NanoScale Corporation and MgOBDH denotes
MgO from BDH Analar), Al2O3 (-Al2O3, Aldrich, 99.8 %, < 10 micron; -Al2O3, Alfa
Aesar, 99.97%), PVA (Sigma-Aldrich, average molecular weight MW = 9000-10000 g mol−1,
80% hydrolysed), H2SO4 (J. T. Baker), and NaBH4 (Aldrich, ≥99.99%) were used without
further purification. Generally, milli-Q water (18.2 M) was used for the aqueous solutions.
Monometallic Au and bimetallic AuPd NPs deposited on MgO or Al2O3 (in total 1 wt% metal
loading) were prepared by sol-immobilisation using either PVA as stabilizing polymer ligand
or without any addition of stabilizer. In a typical synthesis of 1 g of monometallic sample, an
aqueous solution of HAuCl4·3H2O (0.8 mL, 12.5 mg mL-1 Au) was added to 400 mL of
deionised water under vigorous stirring, followed by the addition of PVA (1 wt% aqueous
solution, PVA/Au = 0.65 wt/wt). Subsequently, a freshly prepared aqueous solution of
NaBH4 (0.1 M, NaBH4/Au = 5 mol mol-1,) was added rapidly into the vortex to form a red
sol. After 30 min of sol generation under stirring at room temperature, the colloid was
immobilised by adding 0.99 g of the desired support (MgO or Al2O3) and, for the Al2O3
supported catalyst, 8 drops of concentrated H2SO4. After 1 h of continuous stirring, the slurry
was filtered, the catalyst washed thoroughly with deionized water (1 L) and dried at 110 °C
for 16 h.
Syntheses for the bimetallic AuPd samples proceeded in the same way, but with the further
addition of an aqueous PdCl2 solution (6 mg mL-1 Pd, total metal loading 1 wt%,
Au:Pd = 1 mol mol-1). In this case, the stabilizer-to-metal ratio was 1.2 wt wt-1. The
immobilisation steps were carried out in the same manner as previously described.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
For the stabilizer free variant, both monometallic and bimetallic analogues were prepared as
described above but the addition of stabilizer to the preparation was omitted. The PVA-free
sample was immobilised on each support after 30 min of sol generation.
Glycerol Oxidation
Glycerol oxidation under basic conditions was performed in a glass reactor (Colaver®)
positioned in a thermostatically controlled oil bath at 60 °C and at 3 bar O2 pressure, under
continuous stirring (1200 rpm). Glycerol (5 mL, 0.6 M, Sigma-Aldrich, anhydrous, >99.5%)
and NaOH (5 mL, 1.2 M, Sigma Aldrich BioXtra, >98%, pellets, anhydrous) were added to
the reactor to give a total reaction volume of 10 mL (0.3 M glycerol, 2:1 glycerol:NaOH).
The glycerol/metal mole ratio was 500:1 for the monometallic catalysts and 385:1 for the
bimetallic AuPd catalysts based on Au. The total reaction time employed was 4 h, with
sampling performed after 30, 60, 120 and 240 min of reaction. Samples were quenched and
diluted 1:10 in deionised water before analysis by HPLC. Product analysis was carried out
using an Agilent 1260 Infinity HPLC with a Metacarb 67H column with a 0.1 wt% solution
of phosphoric acid as mobile phase.
Catalyst Characterization
Transmission Electron Microscopy (TEM) was performed using a JEOL 2100 microscope
with a LaB6 filament operating at 200 kV or a Hitachi H-7100 with 100 kV acceleration
voltage. Samples were prepared by dispersing the powder catalyst in ethanol and allowing a
drop of the suspension to evaporate on a lacey carbon film supported over a 300 mesh copper
TEM grid.
XPS was performed on a Kratos Axis Ultra-DLD photoelectron spectrometer, using
monochromatic Al K radiation at 144 W (12 mA × 12 kV) power. High resolution and
survey scans were performed at pass energies of 40 and 160 eV respectively. Magnetically
confined charge compensation was used to minimize sample charging and the resulting
spectra were calibrated against the C 1s line at 284.7 eV. Deconvolution of the XPS spectra is
complicated due to the overlap of the Au 4f5/2 peak with the Mg 2s photoelectron line. To
account for this, we apply a fixed spin orbit splitting of 3.67 eV for the Au 4f doublet, with an
area ratio of 4/3 and assume the FWHM of both peaks are equal.
MP-AES (Agilent 4100) was performed on the samples at the end of the reaction. The slurry
was centrifuged in order to separate the solid from the liquid. The liquid was then recovered
by a syringe and filtered (0.45 μm PTFE Fisherbrand®) to make sure no solid residue was
left. Finally, an aliquot was appropriately diluted in order to lower the base concentration and
analysed.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
References
[1]
M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi, C. Della Pina, Angew. Chem. Int. Ed. 2007,
46, 4434-4440.
[2]
[3]
[4]
C. H. Zhou, J. N. Beltramini, Y. X. Fan, G. Q. Lu, Chem. Soc. Rev. 2008, 37, 527-549.
A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem. 2008, 10, 13-30.
B. Katryniok, H. Kimura, E. Skrzynska, J. S. Girardon, P. Fongarland, M. Capron, R.
Ducoulombier, N. Mimura, S. Paul, F. Dumeignil, Green Chem. 2011, 13, 1960-1979.
A. Villa, N. Dimitratos, C. E. Chan-Thaw, C. Hammond, L. Prati, G. J. Hutchings, Acc.
Chem. Res. 2015, 48, 1403-1412.
S. Carrettin, P. McMorn, P. Johnston, K. Griffin, G. J. Hutchings, Chem. Commun. 2002,
696-697.
S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C. J. Kiely, G. J. Hutchings, Phys. Chem.
Chem. Phys. 2003, 5, 1329-1336.
[5]
[6]
[7]
[8]
S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C. J. Kiely, G. A. Attard, G. J. Hutchings,
Top. Catal. 2004, 27, 131-136.
[9]
F. Porta, L. Prati, J. Catal. 2004, 224, 397-403.
[10]
[11]
[12]
D. Wang, A. Villa, F. Porta, D. S. Su, L. Prati, Chem. Commun. 2006, 1956-1958.
W. C. Ketchie, M. Murayama, R. J. Davis, J. Catal. 2007, 250, 264-273.
M. Sankar, N. Dimitratos, P. J. Miedziak, P. P. Wells, C. J. Kiely, G. J. Hutchings, Chem.
Soc. Rev. 2012, 41, 8099-8139.
[13]
[14]
A. Villa, G. M. Veith, L. Prati, Angew. Chem. Int. Ed. 2010, 49, 4499-4502.
G. L. Brett, Q. He, C. Hammond, P. J. Miedziak, N. Dimitratos, M. Sankar, A. A. Herzing,
M. Conte, J. A. Lopez-Sanchez, C. J. Kiely, D. W. Knight, S. H. Taylor, G. J. Hutchings,
Angew. Chem. Int. Ed. 2011, 50, 10136-10139.
[15]
[16]
S. A. Kondrat, P. J. Miedziak, M. Douthwaite, G. L. Brett, T. E. Davies, D. J. Morgan, J. K.
Edwards, D. W. Knight, C. J. Kiely, S. H. Taylor, G. J. Hutchings, ChemSusChem 2014, 7,
1326-1334.
J. Oliver-Meseguer, A. Domenech-Carbo, M. Boronat, A. Leyva-Perez, A. Corma, Angew.
Chem. Int. Ed. 2017, 56, 6435-6439.
[17]
[18]
[19]
S. Campisi, M. Schiavoni, C. E. Chan-Thaw, A. Villa, Catalysts 2016, 6, 21.
L. Prati, A. Villa, Acc. Chem. Res. 2014, 47, 855-863.
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. 2011, 3, 551-556.
[20]
[21]
C. Deraedt, L. Salmon, S. Gatard, R. Ciganda, R. Hernandez, J. Ruiz, D. Astruc, Chem.
Commun. 2014, 50, 14194-14196.
L. Abis, S. J. Freakley, G. Dodekatos, D. J. Morgan, M. Sankar, N. Dimitratos, Q. He, C. J.
Kiely, G. J. Hutchings, ChemCatChem 2017, 9, 2914-2918.
[22]
[23]
S. Arrii, F. Morfin, A. J. Renouprez, J. L. Rousset, J. Am. Chem. Soc. 2004, 126, 1199-1205.
J. Feng, C. Ma, P. J. Miedziak, J. K. Edwards, G. L. Brett, D. Li, Y. Du, D. J. Morgan, G. J.
Hutchings, Dalton Trans. 2013, 42, 14498-14508.
[24]
[25]
C. Xu, Z. Wang, X. Huangfu, H. Wang, RSC Adv. 2014, 4, 27337-27345.
C. D. Evans, S. A. Kondrat, P. J. Smith, T. D. Manning, P. J. Miedziak, G. L. Brett, R. D.
Armstrong, J. K. Bartley, S. H. Taylor, M. J. Rosseinsky, G. J. Hutchings, Faraday Discuss.
2016, 188, 427-450.
[26]
[27]
[28]
[29]
[30]
N. Dimitratos, J. Lopez-Sanchez, D. Lennon, F. Porta, L. Prati, A. Villa, Catal. Lett. 2006,
108, 147-153.
N. Dimitratos, A. Villa, C. L. Bianchi, L. Prati, M. Makkee, Appl. Catal., A 2006, 311, 185-
192.
W. C. Ketchie, Y. L. Fang, M. S. Wong, M. Murayama, R. J. Davis, J. Catal. 2007, 250, 94-
101.
M. Sankar, N. Dimitratos, D. W. Knight, A. F. Carley, R. Tiruvalam, C. J. Kiely, D. Thomas,
G. J. Hutchings, ChemSusChem 2009, 2, 1145-1151.
S. E. Davis, M. S. Ide, R. J. Davis, Green Chem. 2013, 15, 17-45.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800074
ChemCatChem
[31]
[32]
[33]
C. L. Bianchi, P. Canton, N. Dimitratos, F. Porta, L. Prati, Catal. Today 2005, 102, 203-212.
A. Villa, D. Wang, G. M. Veith, L. Prati, J. Catal. 2012, 292, 73-80.
S. M. Rogers, C. R. A. Catlow, C. E. Chan-Thaw, D. Gianolio, E. K. Gibson, A. L. Gould, N.
Jian, A. J. Logsdail, R. E. Palmer, L. Prati, N. Dimitratos, A. Villa, P. P. Wells, ACS Catal.
2015, 5, 4377-4384.
This article is protected by copyright. All rights reserved.