Bismuth as a modifier of Au-Pd catalyst: Enhancing selectivity in alcohol oxidation by suppressing parallel reaction

Article abstract of DOI:10.1016/j.jcat.2012.04.021

Bi has been widely employed as a modifier for Pd and Pt based catalyst mainly in order to improve selectivity. We found that when Bi was added to the bimetallic system AuPd, the effect on activity in alcohol oxidation mainly depends on the amount of Bi regardless its position, being negligible when Bi was 0.1 wt% and detectably negative when the amount was increased to 3 wt%. However, the selectivity of the reactions notably varied only when Bi was deposited on the surface of metal nanoparticles suppressing parallel reaction in both benzyl alcohol and glycerol oxidation. After a careful characterization of all the catalysts and additional catalytic tests, we concluded that the Bi influence on the activity of the catalysts could be ascribed to electronic effect whereas the one on selectivity mainly to a geometric modification. Moreover, the Bi-modified AuPd/AC catalyst showed possible application in the production of tartronic acid, a useful intermediate, from glycerol.

Full text of DOI:10.1016/j.jcat.2012.04.021

Journal of Catalysis 292 (2012) 73–80
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Bismuth as a modifier of Au–Pd catalyst: Enhancing selectivity in alcohol
oxidation by suppressing parallel reaction
Alberto Villa a, Di Wang b, Gabriel M. Veith c, Laura Prati a,
⇑
a Università di Milano, Dipartimento CIMA ‘‘L.Malatesta’’, via Venezian 21, I-20133 Milano, Italy
b Institut für Nanotechnologie, Karlsruher Institut für Technologie, Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany
c Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Bi has been widely employed as a modifier for Pd and Pt based catalyst mainly in order to improve selec-
tivity. We found that when Bi was added to the bimetallic system AuPd, the effect on activity in alcohol
oxidation mainly depends on the amount of Bi regardless its position, being negligible when Bi was
0.1 wt% and detectably negative when the amount was increased to 3 wt%. However, the selectivity of
the reactions notably varied only when Bi was deposited on the surface of metal nanoparticles suppress-
ing parallel reaction in both benzyl alcohol and glycerol oxidation. After a careful characterization of all
the catalysts and additional catalytic tests, we concluded that the Bi influence on the activity of the cat-
alysts could be ascribed to electronic effect whereas the one on selectivity mainly to a geometric modi-
fication. Moreover, the Bi-modified AuPd/AC catalyst showed possible application in the production of
tartronic acid, a useful intermediate, from glycerol.
Received 15 February 2012
Revised 24 April 2012
Accepted 28 April 2012
Available online 9 June 2012
Keywords:
Bi-modified catalyst
AuPd alloy
Selective alcohol oxidation
Tartronic acid
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
out on this subject, the nature of the promoter role is still under
debate. Proposed models include the formation of a complex
The liquid phase oxidation of alcohols to the corresponding car-
bonyl compounds over supported metal catalysts has been exten-
sively studied in the last decade [1,2] especially involving the use
of molecular oxygen as the oxidant because of its attractive
‘‘green’’ technology, with water as main co-product. Research has
historically focused on the use of Pt and Pd catalysts [3–5]; how-
ever, a drawback in using these catalysts is the deactivation due
to metal overoxidation and irreversible adsorption of (by)products
or intermediates [6]. On the contrary, gold has been demonstrated
to be an active and long-life catalyst for oxidizing alcohol in the
presence of O2 [7–9]. However, gold as the catalyst presents
the limitation of the often compulsory presence of a base [7–9].
The base enhances the rate of the reaction but also drastically af-
fects the selectivity, mainly producing carboxylate instead of the
desired aldehyde. This problem was partly over passed by using
bimetallic systems based on Au and Pd or Pt metals [10–18].
Pd or Pt based catalysts have also been modified by metal pro-
moters, such as Bi and Pb [19,20]. When used alone Bi and Pb are
inactive in the alcohol oxidation, but when associated with noble
metal, they increase the catalytic performance [21] also modifying
the selectivity [21,23]. Despite the large number of studies carried
among the noble metal, the promoter and the reactant [24–26];
a bifunctional catalyst [27,28]; an alloy formation [29–31], or
even a homogeneous catalysis by leached promoter [32]. The
most accepted model involves the blocking of a fraction of active
sites, thus providing different substrate coordination route that in
turn modified the activity and the selectivity of the original cata-
lyst [33].
Different preparation routes were developed aiming the bis-
muth atoms being on the metal and not on the support
[28,34,35]. Indeed, it is generally accepted that bismuth directly
deposited on the noble metal is mostly effective, whereas bis-
muth atoms on the support are only ‘‘spectator species’’ [35].
We already reported the preparation of completely alloyed Au–
Pd on carbon catalyst and ascribed its high catalytic performances
in the liquid phase oxidation of alcohol to the isolation of Pd sin-
gle sites [13,20]. In the present paper, we investigated the effect
of Bi addition to the fully characterized, well alloyed Au–Pd bime-
tallic catalyst in the oxidation of benzyl alcohol and glycerol in
order to study the activity and the regio- as well as chemo-selec-
tivity of these modified catalysts. We added different amount of
Bi (0.1–3%) using two different procedures, the first for assuring
the deposition of Bi on Au–Pd nanoparticles, the second for the
direct deposition on the support. This latter methodology allowed
us also to study the effect of Bi on the formation of the alloyed
phase.
⇑
Corresponding author.
E-mail address: Laura.Prati@unimi.it (L. Prati).
0021-9517/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2012.04.021

74
A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
connected to a 5000 ml reservoir charged with oxygen (300 kPa).
2. Experimental
2.1. Materials
The oxygen uptake was followed by a mass-flow controller con-
nected to a PC through an A/D board, plotting a flow time diagram.
Glycerol oxidation: glycerol 0.3 M and the catalyst (substrate/
total metal = 1000 mol/mol) were mixed in distilled water (total
volume 10 ml) and 4 eq of NaOH. The reactor was pressurized at
300 kPa of oxygen and set to 50 °C. Once this temperature was
reached, the gas supply was switched to oxygen and the monitor-
ing of the reaction started. The reaction was initiated by stirring.
Samples were removed periodically and analyzed by high-
NaAuCl4Á2H2O, Na2PdCl4, and BiO(NO3) were from Aldrich
(99.99% purity) and activated carbon from Camel (X40S;
SA = 900–1100 m2/g; PV = 1.5 ml/g; pH 9–10). Before use, the car-
bon was suspended in HCl 6 M and left under stirring for 12 h and
then washed several times with distilled water by decantation until
the pH of the solution reached values of 6–6.5. At the end, the carbon
was filtered off and dried for 5–6 h at 150 °C in air. The final water
content was evaluated to be <3%. NaBH4 of purity >96% from Fluka,
polyvinylalcohol (PVA) (Mw = 13,000–23,000 87–89% hydrolysed)
from Aldrich was used. Gaseous oxygen from SIAD was 99.99% pure.
performance liquid chromatography (HPLC) using
a column
(Alltech OA- 10,308, 300 mm_7.8 mm) with UV and refractive in-
dex (RI) detection to analyze the mixture of the samples. Aqueous
H3PO4 solution (0.1 wt%) was used as the eluent. Products were
identified by comparison with the original samples.
Benzyl alcohol oxidation: benzyl alcohol and the catalyst (alco-
hol/total metal = 5000 mol/mol) were mixed in cyclohexane (ben-
zyl alcohol/cyclohexane 50/50 vol/vol; total volume, 10 ml). The
reactor was pressurized at 200 kPa of oxygen and set to 80 °C.
The reaction was initiated by stirring. Periodic removal of samples
from the reactor was performed. Identification and analysis of the
products were done by comparison with the authentic samples by
GC using a HP 7820A gas chromatograph equipped with a capillary
2.2. Catalyst preparation
2.2.1. Au–Pd bimetallic catalysts
NaAuCl4Á2H2O (0.072 mmol) was dissolved in 140 ml of H2O, and
PVA (2% w/w) was added (0.706 ml). The yellow solution was stirred
for 3 min, and 0.1 M NaBH4 (2.15 ml) was added under vigorous
magnetic stirring. The ruby red Au(0) sol was immediately formed.
An UV–visible spectrum of the gold sol was recorded to check the
complete AuClÀ4 reduction and the formation of plasmon peak.
Within a few minutes of sol generation, the gold sol (acidified until
pH 2 by sulfuric acid) was immobilized by adding activated carbon
under vigorous stirring. The amount of support was calculated as
having a final metal loading of 0.73 wt% when Au/Pd was prepared.
After 2 h, the slurry was filtered and the catalyst washed thoroughly
with distilled water (neutral mother liquors). ICP analyses were per-
formed on the filtrate using a Jobin Yvon JV24 to verify the total me-
tal loading on carbon. The Au/C was then dispersed in 140 ml of
water; Na2PdCl4 (10 wt% in Pd solution) (0.0386 ml) and PVA (2%
w/w) (0.225 ml) were added. H2 was bubbled (50 ml/min) under
atmospheric pressure and room temperature for 2 h. After addi-
tional 18 h, the slurry was filtered and the catalyst washed thor-
oughly with distilled water. ICP analyses were performed on the
filtrate using a Jobin Yvon JV24 to verify the metal loading on car-
bon. 1 wt% has been the total metal loading.
column (HP-5 30 m  0.32 mm, 0.25
lm Film, made by Agilent
Technologies) and TCD detector. Quantification of the reaction
products was done by the external calibration method.
2.4. Characterization
2.4.1. Catalyst characterization
(a) The metal content was checked by ICP analysis of the filtrate
or alternatively directly on catalyst after burning off the car-
bon, on a Jobin Yvon JY24.
(b) Morphology and microstructures of the catalysts are charac-
terized by transmission electron microscopy (TEM). The
powder samples of the catalysts were ultrasonically dis-
persed in ethanol and mounted onto copper grids covered
with holey carbon film. A Philips CM200 FEG electron micro-
scope, operating at 200 kV and equipped with EDX DX4 ana-
lyzer system and a FEI Titan 80–300 electron microscope,
operating at 300 kV and equipped with EDX SUTW detector
were used for TEM observation.
(c) XPS – X-ray photoelectron spectroscopy (XPS) measurements
were performed with a PHI 3056 spectrometer equipped with
an Al anode source operated at 15 kV and an applied power of
350 W and a pass energy of 5.85 eV. Samples were mounted
on In foil since the C1s binding energy, from the carbon in
the samples, was used to calibrate the binding energy shifts
of the sample (C1s = 284.8 eV).
2.2.2. Au–Pd–Bi trimetallic catalysts
2.2.2.1. Procedure A. Catalysts were prepared following the proce-
dure reported for Au–Pd bimetallic catalyst, except that activated
carbon (Camel X40S; SA = 1100–1200 m2 gÀ1; pH 8–9) that was
loaded by different amount of Bi (0.1–3 wt%) priori. To acidic
H2O (pH 1 H2SO4), a 10À2 M stock solution of BiO(NO3) and PEG–
BDE (polyethyleneglycol reacted with bisphenol diglycidyl ether)
(2% w/w stock solution) were added, under nitrogen and vigorous
magnetic stirring. After 3 min, an alkaline NaBH4 solution (14 M)
was added obtaining a brown sol. An UV–visible spectrum of the
Bi sol was recorded, checking for complete reduction. After
1 min, 1 g of carbon was added as having the proper Bi loading
(0.1; and 3 wt%) and it was maintained under vigorous stirring
for 0.5 h. The slurry was then filtered, and the catalyst washed
thoroughly with distilled water (neutral mother liquors). The two
catalysts have been labeled as Au–Pd@0.1%Bi and Au–Pd@3%Bi.
3. Results and discussion
The two step procedure for preparing Au–Pd on carbon catalyst
was demonstrated to produce AuPd alloyed NPs of single composi-
2.2.2.2. Procedure B. In this case, the Bi nanoparticles, prepared as
in A, have been added after the immobilization of AuPd alloy.
The two catalysts have been labeled as 0.1%Bi@Au–Pd and
3%Bi@Au–Pd.
Table 1
Statistical median and standard deviation of particle size analysis for Au–Pd–Bi
catalysts.
Catalyst
Statistical median (n)
Standard deviations
1% Au–Pd
3.4
3.5
3.6
3.4
3.4
1.4
1.4
1.4
1.4
1.4
1% Au–Pd@ 0.1% Bi
1% Au–Pd@ 3% Bi
0.1%Bi@ 1% Au–Pd
3%Bi@ 1% Au–Pd
2.3. Catalytic tests
Reactions were carried out in a 30 ml glass reactor equipped
with a thermostat and an electronically controlled magnetic stirrer

A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
75
palladium [14]. It was assumed that the synergistic effect can be
considered as a sum of two different effects: the geometric and
the electronic one, the geometric deriving from the isolation of
Pd active site by Au [13,22,23].
We proceeded to modify this catalyst by the addition of con-
trolled amounts of Bi (0.1–3 wt%). On the basis of the literature
[36,37], it was expected that by depositing Bi on the preformed
Au–Pd/AC, a selective deposition of Bi on the metallic NPs can be ob-
tained. On the contrary, by previously impregnating the support, a
random dispersed Bi NPs could be expected. A perturbation in the
AuPd alloy formation could be also envisaged in this latter case. A
complete characterization of the catalysts by TEM and ICP analyses
has been then performed. The preparation and the detailed charac-
terization of the Au–Pd/AC homogenously alloyed single phase
catalyst have been already reported somewhere else [13]. The syn-
thesized single alloyed Au6Pd4 particles have a median particle size
of 3.4 nm (Table 1). By replacing pristine activated carbon with 0.1%
Bi on activate carbon (0.1 wt%Bi/AC), the resulted tri-metallic cata-
lyst shows a very similar structure as the Au6Pd4/AC. An overview
TEM micrograph is shown in Fig. 1. The metal particles are homoge-
neously distributed on the activated carbon, and the size histogram
can be fitted by LogNormal function with median value of 3.53 nm
(Table 1). The 0.1 wt% Bi could not be detected by EDX, but the ICP
analysis carried out by burning off the carbon revealed its presence.
Despite the Bi concentration on each particle was too low to be de-
tected by EDX, we could also assume that the dispersion is good
since no segregation of Bi or Bi-rich alloy particles have been ob-
served. By increasing the Bi loading on AC to 3 wt%, the characteris-
Fig. 1. Overview TEM image of 1% Pd–Au@0.1% Bi on AC.
tion with multiply twinned structure highly dispersed on activated
carbon [13]. In the liquid phase oxidation of alcohols, this catalyst
highlighted an extraordinary synergistic effect between gold and
Fig. 2. The overview TEM image of 4Pd@6Au/3% Bi on AC from (a) the area with small particles and (c) the area with big and irregular particles. (b and d) The EDX spectra
corresponding to (a and c).

76
A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
tics of the catalyst changed while maintaining a similar particle size,
(median value of the size 3.58 nm) (Table 1). In this case, inhomoge-
neous composition is observed in different regions. Fig. 2a and c are
the TEM micrographs from two randomly selected regions of the 3%
Bi catalyst. From the corresponding EDX spectra and the quantita-
tive data shown in Fig. 2b and d, it can be seen that the ratios among
Au, Pd, and Bi vary largely even without obvious change of the par-
ticle morphology. Therefore, we concluded that Bi/AC (especially
when 3 wt% loaded) produces an inhomogeneous catalyst, differing
from AC alone, which allows the uniform Au–Pd alloy formation. On
the contrary when Bi was added after the immobilization of AuPd
NPs, the alloyed structure of AuPd is maintained at any loading of
Bi (0.1% or 3%). Fig. 3a shows the overview STEM image of
0.1%Bi@Au–Pd/AC, where EDX spectrum (Fig. 3b) was acquired from
the area within the red box and Bi signal was only at noise level but
the ICP analysis carried out by burning off the carbon revealed its
presence. In the case of 3% Bismuth, different Bi forms coexisted,
the Bi alloyed with Au–Pd and the segregated Bi. A typical overview
Fig. 3. The overview STEM image of (a) 4Pd@6Au/0.1% Bi on AC and (b) 4Pd@6Au/3% Bi on AC; (c) the EDX spectra corresponding to (a and d) element mapping on 4Pd@6Au/
3% Bi on AC with the integrated EDX spectrum from the mapping area.

A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
77
Fig. 5. Benzyl alcohol oxidation with 1% Au–Pd and 1% Au–Pd + 0.1%Bi.
than the initial feed, implying Bi segregation. The segregated Bi
was detected in some areas, as the example shown in Fig. 4 with
the STEM image and the corresponding EDX spectrum.
The catalytic performances of Au–Pd and Au–Pd–Bi catalysts
have been evaluated in the oxidative dehydrogenation of benzyl
alcohol (Table 2) and glycerol (Table 3).
Benzyl alcohol oxidation has been performed in a batch reactor
at 200 kPa of oxygen and thermostatted at 80 °C, using cyclohexane
as solvent (benzyl alcohol/cyclohexane 50/50 vol/vol, AuPd/sub-
strate molar ratio 1/5000). When a small amount of Bi (0.1 wt%)
was added directly to the support, a negligible effect on both activity
and selectivity was detected compared to Au–Pd/AC (TOF 2278 hÀ1
and 2261 hÀ1, respectively) (Table 2). However, when the Bi amount
was increased from 0.1 wt% to 3 wt%, the catalytic performance of
the modified catalyst decreased (TOF 2278 hÀ1 for Au–Pd/0.1%Bi,
and 1245 hÀ1 for Au–Pd/3%Bi) even the selectivity remained similar
(selectivity to benzaldehyde at 90% conversion = 88–89%). It should
be noticed that the catalyst obtained using the support pre-treated
Fig. 4. Bi segregation in 3%Bi/AuPd/AC.
STEM image of trimetallic particles supported on carbon is shown in
Fig. 3c. Fig. 3d is the STEM–EDX spectrum imaging of some particles,
where each particle consists of all the three metals. The integrated
spectrum from the whole mapping area and the quantification re-
sult are also shown in Fig. 3d. The Bi concentration is much lower
Table 2
Comparison of bi and tri-metallic catalyst activities in benzyl alcohol oxidation.
b
Catalysta
TOF (hÀ1
)
Selectivity (%)c
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
1% Au–Pd
2261
2278
1245
2546
1643
8
8
9
2
1
89
88
89
95
97
1
1
0
0
0
2
3
2
3
2
1% Au–Pd@ 0.1% Bi
1% Au–Pd@ 3% Bi
0.1%Bi@ 1% Au–Pd
3%Bi@ 1% Au–Pd
a
b
c
Reaction condition: alcohol/metal 5000/1 (mol/mol), 80 °C, pO2 2 atm, 1250 rpm.
TOF calculated after 15 min of reaction.
Selectivity at 90% conversion.
Table 3
Comparison of bi and tri-metallic catalyst activities in glycerol oxidation.
b
Catalysta
TOF (hÀ1
)
Selectivity (%)c
Glycerate
Glycolate
Tartronate
Oxalate
1% Au–Pd
3503
3538
1856
3897
2856
76
74
73
75
74
19
18
20
12
15
2
4
5
10
11
1
0
1
1
1
1% Au–Pd@ 0.1% Bi
1% Au–Pd@ 3% Bi
0.1%Bi@ 1% Au–Pd
3%Bi@ 1% Au–Pd
a
b
c
Reaction condition: alcohol/metal 1000/1 (mol/mol), 50 °C, pO2 3 atm, 1250 rpm.
TOF calculated after 15 min of reaction.
Selectivity at 90% conversion.

78
A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
with 3 wt% Bi did not allow the formation of a homogeneous alloy.
Therefore, we considered that the results could also be affected by
other effects than the presence of Bi, such as the different composi-
tion, not the size as it was similar, of metallic nanoparticles. We thus
considered the catalytic behaviors of the systems obtained by the
subsequent modification of Bi on preformed Au–Pd/AC that assured
the presence of a single alloy composition. Surprisingly, we ob-
served the same trend of activity as the previous one (Table 2):
the slight increase of the catalytic activity observed when the
0.1 wt% of Bi was added (TOF 2278 hÀ1 for AuPd/AC and 2546 hÀ1
for 0.1%Bi@AuPd/AC) drastically drop down when the amount of
Bi was increased to 3 wt% (TOF 1643 hÀ1). We could thus conclude
that the decreasing of activity could be correlated with the increas-
ing of Bi from 0.1 wt% to 3 wt%. However, as the characterization of
3%Bi@AuPd/AC revealed a partial segregation of Bi, we also tested a
physical mixture of 3%Bi/AC and AuPd/AC in order to determine the
potential role of isolated Bi nanoparticles on AC. This experiment
showed a perfectly overlapped reaction profile for the reaction car-
ried out with the AuPd/AC or with a physical mixture of 3%Bi/AC and
AuPd/AC. This finding confirmed the spectator role of Bi on AC
(Fig. 5). Therefore, we were also able to state that the detrimental ef-
fect of Bi on the activity of AuPd/AC is due to Bi near located to the
active sites and possibly a consequence of a selective inhibition of
some AuPd active sites by an intimate contact with Bi. Probably
when the amount of Bi is low (0.1 wt%), this inhibiting effect on
the catalytic activity is reduced. This proximity effect can be re-
vealed in almost the same extent regardless the Bi position (un-
der/on) with respect to AuPd nanoparticles. Therefore, we mainly
ascribed the detrimental effect of the Bi presence to an electronic ef-
fect. To investigate the electronic properties of the catalysts, X-ray
photoelectron spectroscopy (XPS) measurements were performed
on three AuPd/C based samples (Au–Pd/C; 0.1%Bi@AuPd/C and
3%Bi@AuPd/C). However, due to peak overlap between the gold
4d5/2 and Pd 3d5/2 peaks at 335 eV, and the low concentrations
Scheme 1. Proposed mechanism for toluene formation.
Scheme 2. Reaction scheme for glycerol oxidation under basic conditions.

A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
79
Table 4
Selectivity of AuPd/AC and 0.1%Bi@AuPd/AC at different temperature.
Catalysta
Selectivityb
50 °C
60 °C
70 °C
GLYC
TART
GLY
GLYC
TART
GLY
GLYC
TART
GLY
Au–Pd
0.1%Bi Au–Pd
76
75
2
10
19
12
70
54
5
34
23
13
22
10
20
34
45
42
a
Reaction condition: alcohol/metal 1000/1 (mol/mol), pO2 3 atm, 1250 rpm.
Selectivity at 90% conversion. GLYC = Glycerate, TART = Tartronate, GLY = Glycolate.
b
of metals, relative to support material, identification of peak shifts
due to changes in electronic structure is impossible. Therefore,
XPS technique cannot definitively confirm the origin for this change
in activity to electronic effects. However, given the incorporation of
Bi, it is highly probably that the electronic structures of individual
clusters changes significantly affecting the bonding between the
reactant and the catalyst. However, looking at the selectivity of
the benzyl alcohol oxidation, we found an additional effect that
strictly depended on the position of Bi. In fact, interestingly, when
Bi is deposited on the metallic nanoparticles, a significant increase
in the selectivity to benzaldehyde was observed mainly at the ex-
pense of toluene formation that in both cases is reduced to 1–2%
from 8% to 9% in the other cases when Bi modified the AC (Table
2). Toluene is expected to be formed by reaction of the intermediate
metal-hydride with alcohol instead of O2 [38] (Scheme 1). Thus,
most probably the blocking of a fraction of active sites occurred
and, according to previous studies on Bi-modified catalysts [33], this
can also provide a different substrate coordination route slowing
down its reaction with the metal-hydride. On the contrary when
Bi was deposited on the AC, this effect on selectivity resulted negli-
gible thus supporting that a geometric effect could be the main
responsible of selectivity modification. However, it cannot be ex-
cluded that in this latter case also an electronic effect could be active
as well.
activity and highest selectivity to tartronate. By increasing the tem-
perature from 50 °C to 60 °C (Table 4), the selectivity to tartronate
increased to 34% from 10% without any appreciable increasing of
selectivity to glycolate. When the temperature was raised to 70 °C,
the selectivity to tartronate did not vary but a strong increase of gly-
colate selectivity (42%), deriving from oxidative C–C carbon bond
cleavage. Therefore, for optimizing the tartronate production, we
maintained the temperature in the range of 50–60 °C but we pro-
longed the reaction time. Fig. 6 showed the reaction profiles of glyc-
erol oxidation carried out at 60 °C in the presence of AuPd/AC
(Fig. 6a) and of 0.1Bi%@AuPd/AC (Fig. 6b). As clearly shown by
Fig. 6, AuPd/AC is not able to perform the consecutive oxidation of
glycerate to tartronate, whereas 0.1Bi%@AuPd/AC efficiently cata-
lyzed the reaction providing the complete conversion of glycerate
to tartronate after 14 h with a isolated yield of tartronate of 78%.
For better understanding this latter aspect, we tested the cata-
lyst in the glycerol oxidation, where different OH groups could
be preferentially oxidized. This reaction was performed in water
under basic conditions (0.3 M Glycerol solution, glycerol/
NaOH = 4 mol/mol, glycerol/metal = 1000 mol/mol) (Table 3). The
trend of the activity of all the modified catalysts with respect to
the AuPd/AC one is in line with the activities observed in the ben-
zyl alcohol oxidation. The addition of 0.1 wt% Bi slightly increased
the activity, in particular when Bi has been deposited on preformed
AuPd/AC, whereas a consistent decreasing of activity occurred
when 3 wt% Bi is present. However, the most interesting aspect lies
on the selectivity of the reaction. The selectivity to the main prod-
uct glycerate did not show significant variations. However, con-
cerning with the distribution of the other reaction products,
consistent differences have been evidenced. Indeed, considering
the two reaction pathways (a and b) of Scheme 2, it should be
noted that tartronate is formed through the consecutive oxidation
of glycerate, whereas glycolate represented a probe for the oxida-
tive degradation [8]. Therefore, from the data of Table 3, we could
conclude that when Bi is deposited on the metal nanoparticles, it
promotes the consecutive reaction of glycerol to tartronate (path
b, Scheme 2) instead of the degradation (path a, Scheme 2). This
finding confirmed what reported in the literature in the case of
Pd or Pt catalysts modified by Bi [36,38,39]. The maximum yield
(58%) of tartronate was obtained under basic conditions by Kimura
et al. doping a Ce–Pd–Pt/AC catalyst with Bi [39].
As tartronate represents a useful chemical for pharmaceutical
application [40–42] and as an anticorrosive agent [43], we studied
more in detail this reaction focusing our attention on the
0.1%Bi@AuPd/AC catalyst, being the catalyst showing the highest
Fig. 6. Reaction profile of (a) 0.1%Bi@AuPd/AC and (b) AuPd/AC at 60 °C.

80
A. Villa et al. / Journal of Catalysis 292 (2012) 73–80
[6] B.N. Zope, R.J. Davis, Green Chem. 13 (2011) 3484.
4. Conclusions
[7] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.J. Hutchings, Phys.
Chem. Chem. Phys. 5 (2003) 1329.
Bi has been used as modifier of AuPd/AC catalyst and tested in
the catalytic oxidation of benzyl alcohol in cyclohexane and glyc-
erol in water. The deposition of Bi on AC has been shown to have
no effect on particle size of the AuPd metallic nanoparticles but
did not allow a homogeneous composition of the alloyed phase
even no segregation of the active metal (Au or Pd) has been de-
tected. Conversely, the deposition of Bi on preformed AuPd/AC
did not affect the alloy composition even partial segregation of Bi
has been observed by increasing the loading of Bi.
From a catalytic point of view, in both the catalytic tests, it has
been observed that from an activity point of view, Bi has a detri-
mental effect more consistent for a higher loading (3 wt%). The ef-
fect on the activity of the catalysts was observed either the Bi was
deposited under or on the AuPd nanoparticles but experiments car-
ried out on a physical mixture of Bi/AC and AuPd/AC allowed to
conclude that the effect is determined by the proximity of Bi to
AuPd nanoparticles and could be mainly electronic. However, a
sensible effect on the selectivity was observed depending of the
relative position between Bi and AuPd NPs, being present only
when Bi is deposited on preformed AuPd alloyed particles. We as-
cribed the effect mainly to the blocking of a fraction of active sites
that possibly provided a different substrate coordination route
even we could not exclude an additional electronic effect. In the
case of benzyl alcohol oxidation, this produced a suppression of
the parallel reaction pathway that yield to toluene thus increasing
the benzaldehyde production; in the case of glycerol oxidation, this
promoted the consecutive reactions increasing the tartronate pro-
duction. As the production of tartronate is of importance from an
industrial point of view, the reaction was optimized obtaining a
yield of 78% at full conversion of glycerol.
[8] F. Porta, L. Prati, J. Catal. 224 (2004) 397.
[9] S. Demirel-Gulen, M. Lucas, P. Claus, Catal. Today 102–103 (2005) 166.
[10] N. Dimitratos, J.A. Lopez-Sanchez, D. Lennon, F. Porta, L. Prati, A. Villa, Catal.
Lett. 108 (3–4) (2006) 147.
[11] D.I. Enache, J.K. Edwards, P. Landon, B. Solsona-Espriu, A.F. Carley, A.A. Herzing,
M. Watanabe, C.J. Kiely, D.W. Knight, G.J. Hutchings, Science 11 (2006) 362.
[12] W. Hou, N.A. Dehm, R.W.J. Scott, J. Catal. 253 (2008) 22.
[13] D. Wang, A. Villa, F. Porta, D. Su, L. Prati, Chem. Commun. (2006) 1956.
[14] A. Villa, D. Wang, N. Dimitratos, D. Su, V. Trevisan, L. Prati, Catal. Today 150
(2010) 8.
[15] A. Villa, G.M. Veith, L. Prati, Angew. Chem. Int. Ed. 49 (2010) 4499.
[16] N. Dimitratos, A. Villa, D. Wang, F. Porta, D. Su, L. Prati, J. Catal. 244 (2006) 113.
[17] G.L. Brett, Q. He, C. Hammond, P.J. Miedziak, N. Dimitratos, M. Sankar, Andrew
A. Herzing, M. Conte, J.A. Lopez-Sanchez, C.J. Kiely, D.W. Knight, S.H. Taylor, G.J.
Hutchings, Angew. Chem. Int. Ed. 50 (2011) 10136.
[18] A. Tsuji, K.V. Rao, S. Nishimura, A. Takagaki, Kohki Ebitani, ChemSusChem 4
(2011) 542.
[19] C. Mondelli, D. Ferri, J. Grunwaldt, F. Krumeich, S. Mangold, R. Psaro, A. Baiker,
J. Catal. 252 (2007) 77.
[20] A. Deffernez, S. Hermans, M. Devillers, Appl. Catal. A 282 (2005) 303.
[21] T. Mallat, Z. Bodnar, A. Baiker, Stud. Surf. Sci. Catal. 78 (1993) 377.
[22] P.C.C. Smits, B.F.M. Kuster, K. van der Wiele, H.S. van der Baan, Carbohydr. Res.
153 (1986) 227.
[23] H. Kimura, A. Kimura, I. Kokubo, T. Wakisaka, Y. Mitsuda, Appl. Catal. A 95
(1993) 143.
[24] H. van Bekkum, in: F.W. Lichtenthaler (Ed.), Carbohydrates as Organic Raw
Materials, VCH, Weinheim, 1991, p. 289.
[25] P.C.C. Smits, B.F.M. Kuster, K. van der Wiele, H.S. van der Baan, Appl. Catal. 33
(1987) 83.
[26] M. Besson, P. Gallezot, in: R.A. Sheldon, H. van Bekkum (Eds.), Fine Chemicals
through Heterogeneous Catalysis, Wiley–VCH, Weinheim, 2001, p. 491.
[27] R. Parsons, T. VanderNoot, J. Electroanal. Chem. 257 (1988) 9.
[28] T. Mallat, Z. Bodnar, A. Baiker, O. Greis, H. Strübig, A. Reller, J. Catal. 142 (1993)
237.
[29] T. Tsujino, S. Ohigashi, S. Sugiyama, K. Kawashiro, H. Hayashi, J. Mol. Catal. 71
(1992) 25.
[30] M. Wenkin, C. Renard, P. Ruiz, B. Delmon, M. Devillers, Stud. Surf. Sci. Catal.
108 (1997) 391.
[31] S. Karski, I. Witonska, J. Mol. Catal. A 191 (2003) 87.
[32] M. Wenkin, P. Ruiz, B. Delmon, M. Devillers, J. Mol. Catal. A 180 (2002) 141.
[33] T. Mallat, A. Baiker, J. Patscheider, Appl. Catal. A 79 (1991) 59.
[34] T. Mallat, Z. Bodnar, P. Hug, A. Baiker, J. Catal. 153 (1995) 131.
[35] C. Keresszegi, J.-D. Grunwaldt, T. Mallat, A. Baiker, J. Catal. 222 (2004) 268.
[36] A. Villa, D. Wang, D. Su, G.M. Veith, L. Prati, Phys. Chem. Chem. Phys. 12 (2010)
2183.
[37] P. Fordham, M. Besson, P. Gallezot, Appl. Catal., A 133 (1995) L179.
[38] M. Sankar, E. Nowicka, P.J. Miedziak, G.L. Brett, R.L. Jenkins, N. Dimitratos, S.H.
Taylor, D.W. Knight, D. Bethell, G.J. Hutchings, Faraday Discuss. 145 (2010)
341.
Acknowledgments
Financial support by Fondazione Cariplo and Euminafab are
gratefully acknowledged. A portion of this research was sponsored
by the Laboratory Directed Research and Development Program of
Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for
the U.S. Department of Energy (GMV).
[39] H. Kimura, T. Imanaka, Y. Yokota, Kao Corp., JP199395253, 1994.
[40] P. Gallezot, Catal. Today 37 (1997) 405.
[41] G. Caselli, M. Monatovanini, C.A. Gandolfi, M. Allegretti, S. Fiorentino, L.
Pellegrini, G. Mellillo, R. Bertini, W. Sabbatini, R. Anacardio, G. Calvenna, G.
Sciortino, A. Teti, J. Bone Miner. Res. 12 (1997) 972.
[42] C.A. Gandolfi, L. Cotini, M. Mantovanini, G. Caselli, G. Clavenna, C. Omini,
DompeFarmaceutici S.P.A, 1994, WO9410127.
References
[1] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037.
[2] A. Corma, H. Garcia, Chem. Soc. Rev. 37 (2008) 2096.
[3] M. Besson, P. Gallezot, Catal. Today 57 (2000) 127.
[4] A.B. Crozon, M. Besson, P. Gallezot, New J. Chem. 22 (1998) 269.
[5] P. Vinke, D. deWit, A.T.J.W. de Goede, H. van Bekkum, New Developments in
Selective Oxidation by Heterogeneous Catalysis, Studies in Surface Science and
Catalysis, vol. 72, Elsevier, Amsterdam, 1992. p. 1.
[43] P.M. Bizot, B.R. Bailey, P.D. Hicks, Nalco Chemical Company, WO9816475.