Au/Pt/TiO2 catalysts prepared by redox method for the chemoselective 1,2-propanediol oxidation to lactic acid and an NMR spectroscopy approach for analyzing the product mixture

Article abstract of DOI:10.1016/j.apcata.2014.11.039

1,2-propanediol obtained from the waste glycerol is the most prominent substrate to produce lactic acid (LA) through the sustainable green catalytic oxidative route. We showed that bimetallic Au/Pt/TiO₂ catalysts that could be easily prepared by the redox reaction with preadsorbed hydrogen possessed a high activity in the formation of lactic acid from 1,2-propanediol through selective primary hydroxyl group oxidation by an oxygen-air mixture or an air. The most efficient catalyst with the highest TON value was the sample with the lowest Au content. Even the addition of 0.025 wt.% of gold to the 1%Pt/TiO₂ catalyst by the redox method resulted in the enhanced 1,2-propanediol conversion up to 96% with the 91% selectivity to LA. The high oxidation activity of the Au/Pt/TiO₂ catalysts was ascribed to the strong interaction between the two metals in Au/Pt bimetallic particles and between the bimetallic particles and the support, which was proved by X-ray photoelectron spectroscopy (XPS) analysis, Fourier transform infrared spectroscopy of adsorbed CO (DRIFTS) technique, and temperature programmed reduction method (TPR-H₂). The reaction products were identified and analyzed with the use of 1D and 2D nuclear magnetic resonance spectroscopy (NMR) methods, which also allowed suggesting a plausible up-to-date scheme of 1,2-propanediol catalytic oxidation.

Full text of DOI:10.1016/j.apcata.2014.11.039

Applied Catalysis A: General 491 (2015) 170–183
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Au/Pt/TiO2 catalysts prepared by redox method for the
chemoselective 1,2-propanediol oxidation to lactic acid and an NMR
spectroscopy approach for analyzing the product mixture
Elena Redinaa,∗, Alexander Greisha, Roman Novikova,b, Anastasiya Strelkovac,
Olga Kirichenkoa, Olga Tkachenkoa, Gennady Kapustina, Ilya Sinevd, Leonid Kustova,e
a N.D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 47 Leninsky Prospect, 119991 Moscow, Russian Federation
b V.A. Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, 32 Vavilov Street, 119991 Moscow, Russian Federation
c Lomonosov Moscow State University of Fine Chemical Technologies, 86 Vernadsky Prospect, 119571 Moscow, Russian Federation
d Lehrstuhl Technische Chemie, Ruhr-Universitat Bochum, 44780 Bochum, Germany
e Chemistry Department of Moscow State University, Leninskie Gory 1, Bldg.3, 119992 Moscow, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 August 2014
Received in revised form 2 November 2014
Accepted 26 November 2014
Available online 11 December 2014
1,2-propanediol obtained from the waste glycerol is the most prominent substrate to produce lactic acid
(LA) through the sustainable “green” catalytic oxidative route. We showed that bimetallic Au/Pt/TiO2 cat-
alysts that could be easily prepared by the redox reaction with preadsorbed hydrogen possessed a high
activity in the formation of lactic acid from 1,2-propanediol through selective primary hydroxyl group
oxidation by an oxygen–air mixture or an air. The most efficient catalyst with the highest TON value
was the sample with the lowest Au content. Even the addition of 0.025 wt.% of gold to the 1%Pt/TiO2
catalyst by the redox method resulted in the enhanced 1,2-propanediol conversion up to 96% with the
91% selectivity to LA. The high oxidation activity of the Au/Pt/TiO2 catalysts was ascribed to the strong
interaction between the two metals in Au/Pt bimetallic particles and between the bimetallic particles
and the support, which was proved by X-ray photoelectron spectroscopy (XPS) analysis, Fourier trans-
form infrared spectroscopy of adsorbed CO (DRIFTS) technique, and temperature programmed reduction
method (TPR-H2). The reaction products were identified and analyzed with the use of 1D and 2D nuclear
magnetic resonance spectroscopy (NMR) methods, which also allowed suggesting a plausible up-to-date
scheme of 1,2-propanediol catalytic oxidation.
Keywords:
Gold–platinum catalyst
Redox method
Bimetallic catalyst preparation
Lactic acid
1,2-propanediol oxidation
NMR
DRIFTS
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
The first works on glycerol to LA transformation were focused
on the hydrothermal base-induced reaction [5–7]. Kishida et al. [5]
In the recent years renewable bio-based feedstock sources have
become increasingly important in the chemical industry. Glycerol,
as a large-scale by-product in biodiesel production, is a versatile
feedstock for the synthesis of a wide range of different chemicals
[1–4]. Among them lactic acid (LA) and lactates are the promising
target substances. They are common chemicals in the food indus-
try and cosmetic manufactures, and their demand increases due
to expanding markets for biodegradable synthetics, ecologically
friendly chemicals. Furthermore, poly-lactic acid is a well-known
biodegradable polymer usable as packing materials (Nature Works
polymers).
first succeeded in the lactic acid synthesis via the hydrothermal
conversion of glycerol with NaOH under an inert atmosphere. The
yield of lactic acid was 90 mol.% at 300 ◦C. Under those conditions,
the glycerol conversion and LA yield depended on the alkaline metal
nature and its concentration ranging within 0.25–2.5 M [6]. The
high LA yield of 85–90% was obtained even at low temperatures
of 250–290 ◦C [7]. The supported precious metals (Au, Ir, Pd, Pt, Rh,
Ru) catalyzed the glycerol conversion in alkaline solutions under an
inert or reducing atmosphere of 2–4 MPa and decreased the tem-
perature of the glycerol conversion to 180–200 ◦C, but the LA yield
was low [8–11].
Replacement of the inert gas for oxygen resulted in glycerol
oxidation over precious metal catalysts at temperatures as low
as 35–80 ◦C; the major products being glyceric and tartronic acids
[12–16]. Only in a few papers LA was claimed to be a major prod-
uct of glycerol oxidation. The highest LA yield was 86% at complete
∗
Corresponding author. Tel.: +7 915 2694149.
E-mail address: redinalena@yandex.ru (E. Redina).
http://dx.doi.org/10.1016/j.apcata.2014.11.039
0926-860X/© 2014 Elsevier B.V. All rights reserved.

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
171
glycerol conversion over the sol-prepared 1%Au–Pt/TiO2 catalysts
at 90 ◦C and 1 bar O2, but the reaction rate was too low [17]. LA
was obtained using a 1%Au–Pt/CeO2 system with the 80% selectiv-
ity and almost complete conversion at 100 ◦C under 5 bar of oxygen
pressure in strong alkali media [18]. The synergistic effect of Au–Pt
interaction on the glycerol oxidation was revealed previously [19]
and resulted only in an increase of the glycerol conversion and
selectivity to tartronic acid at the same reaction conditions. How-
ever, the reasons behind so different effects are still unclear.
Catalytic glycerol non-base oxidation proceeds mostly to glyc-
eric acid, the highest conversion and selectivity being reached over
AuPt/H-mordenite catalysts [20], AuPt/hydrotalcite catalysts [21],
Au/Pd/MgO or Au/Pt/MgO [22], and trimetallic Au/Pt/Pd nanopar-
ticles supported on TiO2 [23].
Lactic acid was obtained either at milder conditions or at higher
yields using the glycerol derivatives. For the hydrothermal conver-
sion of dihydroxyacetone in acid solutions in the presence of Cr3+
and Al3+ salts, a complete conversion was reached at 140 ◦C with a
selectivity to LA 80–90% and 90–95%, respectively [24]. Glyceralde-
hyde can be converted at the same conditions, yet the LA yield is
lower. In basic solutions, the lactic acid yield of 59% was achieved
with Ca(OH)2 at 55 ◦C and an OH−:substrate molar feed ratio of 2
[25].
In fact, glycerol can be easily converted into 1,2-propanediol
(1,2-PD) [26,27], which is a good source to produce lactic acid
using a facile “green” route via the selective oxidation of 1,2-PD
with oxygen or air. Recently, it has been shown that supported
gold–platinum catalysts were effective in polyols transformation
into carboxylic acids with lactic acid as a case [17,18,28,29]. It
should be mentioned that only an Au-containing catalyst exhib-
ited a sufficient catalytic activity and selectivity depending on the
Au:Pt molar ratio, whereas the activity of Pt–Pd and Au–Pd cata-
lysts was found to be much lower and/or a loss of regioselectivity
toward primary OH-group oxidation was observed [29,30].
One can see from this brief outlook of the published data that
the Au–Pt combination is the most beneficial in selective pri-
mary OH-group oxidation of polyols. Anyway, most studies on
partial oxidation of alcohols were carried out using the methods
of Au–Pt catalyst preparation requiring the preliminary synthesis
of a bimetallic sol with the following deposition on a support. It
has been shown that simple room-temperature borohydride co-
reduction of appropriate aqueous metal salt solutions yielded alloy
nanoparticles in the bulk immiscible Au–Rh, Au–Pt, Pt–Rh, and
Pd–Rh systems [31]. However, the fine sols synthesized should
be stabilized using protective agents that are to be removed com-
monly in order to avoid limitation of the reagent mass transfer to
the catalyst surface [32–35].
NPs in the parent catalyst. That is why it is so important to choose
an appropriate technique to prepare the parent catalyst with the
optimal metal NPs features. While the particle size is smaller, the
number of active sites is larger where the adsorption of the oxidized
metal or reductant proceeds.
Here we report a selective oxidation of 1,2-propanediol (1,2-
PD) to lactic acid (LA) with high yields over the Au/Pt/TiO2 samples
obtained by the redox method.
We also aimed at developing an approach suitable for the cor-
rect identification and quantification of the obtained products. The
high-performance liquid chromatography is commonly applied to
analyze the obtained products in the polyols and sugars oxidation
reactions. Description of the method and conditions used are avail-
able elsewhere [29,45]. The drawback of the application of this
method is a requirement to use strong acids as eluents because
of the alkali media of a reaction product mixture. Another possible
way of analysis of the mixture of carboxylic acids in the form of
salts implies the use of gas–liquid chromatography with the pre-
vious steps of neutralization, esterification, and extraction of the
obtained esters, or silylation of the acids [46]. Normally, the method
is time-consuming because of the extraction stage of the products,
which are soluble both in water and in polar organic solvents. It
is a good solution to use NMR spectroscopy, but there is another
problem arising here: the overlapping of a huge peak of water with
other peaks in a diluted water solution makes the identification
of the all products hardly possible. Some approaches have been
elaborated to fix the problem. For example, 1H NMR spectroscopy
was used to identify the products obtained in 1,2-propanediol elec-
trooxidation on an Au electrode in a basic medium [47]. In order
to record the proton spectra and suppress the signal of water the
reaction was conducted in D2O [47]. In our paper, we propose the
method of quantitative analysis and comprehensive precise iden-
tification of the products, even minor ones, by 1D (1H NMR) and 2D
nuclear magnetic resonance spectroscopy techniques: COSY [48],
DOSY [49], HSQC [50,51], HMBC [52], which were applied without
product separation/isolation from the aqueous mixture.
2. Experimental
All reagents were analytically grade: HAuCl4·3H2O (Aurat),
H2PtCl6·6H2O (Aurat), 1,2-propanediol (Acros Organics). Support
TiO2 (P-25, Degussa) with a BET surface area of 45 m2 g−1 was used.
2.1. Catalysts preparation
2.1.1. Pt/TiO2 preparation
In our work, we propose a facile redox method to obtain Au–Pt
catalysts with strong interaction between the two metals. The
redox method of bimetallic particles preparation includes two
stages. The first step is the synthesis of a supported monometallic
parent catalyst. The second stage is a modification of this parent cat-
alyst by the deposition of a second metal on the surface of the metal
nanoparticles (NPs) in the parent catalyst. This can be achieved by
the direct redox reaction between the modifier metal in the oxi-
dized form and the reduced parent metal due to different standard
redox potentials of the metal couple [36–40]. Another approach
of this method is an interaction of the modifier with the reduc-
tant preadsorbed on metal NPs in the parent catalysts [41–44]. It is
worth mentioning that the redox potential of the modifier should
be higher than that of the parent metal or preadsorbed reductant.
This makes the redox method to be the best choice to prepare
gold-containing bimetallic catalysts as gold possessed the highest
standard redox potential among the over d-elements.
The Pt (1 wt.%) monometallic catalyst was prepared by the
method of platinum hydroxide deposition under pH control. The
initial 0.1 M H2PtCl6 solution was diluted in a proper amount of
deionized water to the concentration 1.03 mmol L−1. The pH value
of the obtained solution was adjusted to 6.8–7.0 by slow titration
with 0.1 M Na2CO3. Then TiO2 was added to the solution. The con-
centration of the support in the solution was 2 g L−1. The suspension
was then kept under stirring at RT for 30 min, and the pH value
decreased spontaneously to 2.5. No separation of the suspension
was observed at these conditions if stirring was switched off. Then
the suspension was heated up to 65 ◦C and kept under stirring for
30 min followed by a step-by-step increase of the temperature up
to 80 ◦C with a slow increase of pH again to the value 6.9–7.0 by
2−
Na2CO3 addition to facilitate hydrolysis of [Pt(OH)xCl6−x
]
ions.
The suspension was kept at 80 ◦C for 1 h. After that it was separated
from the parent solution and solid. Hot filtration was performed to
obtain a solid catalyst, which was washed several times with hot
water to remove adsorbed chloride ions from the catalyst surface.
A qualitative test with KI (0.6 wt.%) and HCl proved no platinum
The characteristics of the bimetallic catalysts prepared by the
redox method strongly depend on the size and morphology of metal

172
E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
left in the parent solution and, thus, almost 100% of its deposition
onto the TiO2 surface. The catalyst was dried in air at RT for 1 h
and then at 60 ◦C overnight. Finally, the catalyst was reduced in a
hydrogen flow at 200 ◦C to obtain metallic platinum nanoparticles
(NPs) supported on TiO2. The reduction temperature was chosen in
accordance with the TPR-H2 data (See Supporting Information S2).
by CO adsorption and spectra recording. Evacuation of adsorbed CO
was also performed at different temperatures: 20, 100, 200 ◦C.
2.2.3. Temperature programmed reduction with hydrogen
(TPR-H2)
TPR measurements were performed in the lab-constructed flow
system. The 1Pt/TiO2 and 1Au/1Pt/TiO2 catalysts with a weight of
140–170 mg were pretreated in an argon flow at 80 ◦C for 90 min.
Then the catalyst sample was cooled in an Ar flow to −50 ◦C prior
to the TPR experiment. Heating from −50 to 850 ◦C was carried out
at the rate of 10 ◦C/min in a 4.6% H2–Ar gas mixture supplied with a
space velocity of 30 ml min−1. Then the sample was kept at 850 ◦C
until the hydrogen consumption ceased.
2.1.2. Au/Pt/TiO2 preparation
A series of Au/Pt/TiO2 catalysts (Au:Pt = 0.025:1; 0.05:1; 1:1
nominal atomic ratio) were prepared by a redox reaction with
preadsorbed hydrogen.
First, Pt–O/TiO2 (powder) was reduced in a U-shaped quartz
reactor with hydrogen at 200 ◦C and at a hydrogen flow rate
30 ml min−1. After that, the system was cooled down and hydro-
gen adsorption was carried out for 1 h at RT. Then the hydrogen
flow was switched off and a proper amount of an HAuCl4 solu-
tion (C = 1–7 mmol L−1) was added to the reduced platinum catalyst
under the hydrogen atmosphere avoiding the contact of the sam-
ple with air. Decolorization of the acid solution was immediately
observed. The catalyst was periodically stirred and kept in the
closed reactor for 30 min. Finally, the parent solution was separated
from the solid by centrifugation. To determine the completeness
of the gold deposition, an aliquot of a parent solution (2 ml) was
collected and iodometric titration of gold was carried out with
Na2S2O3 [53]. The possible leaching of platinum was also checked
with the solution of KI (0.6 wt.%) in acidic media. In all cases the
gold deposition was close to 99–100% and no leaching of platinum
was detected. Then the solid was centrifuged and washed several
times with distilled water. To remove chlorine ions, the washing
steps were repeated until a qualitative reaction with a solution of
AgNO3 was not observed. The catalysts were dried under vacuum
at 40 ◦C and then at 60 ◦C overnight.
2.2.4. Scanning transmission electron microscopy (STEM)
Target-oriented approach was utilized for the optimization
of the analytical measurements [55]. The samples morphology
was studied using a Hitachi SU8000 field-emission scanning elec-
tron microscope (FE-SEM). Before measurements, the samples
were mounted on 3 mm copper grids and fixed in a grid holder.
Images were acquired in the bright-field STEM mode at the 30 kV
accelerating voltage. The average particle size was calculated as
dav = ꢀnidi/n, where ni represents the number of particles with the
diameter di, n is the total number of calculated particles. The num-
ber of particles used for calculation of the particle size distribution
for each sample was 300–400, and at least 10 STEM images for
each catalyst were used to obtain the statistics of the particle size
distribution.
2.3. Catalyst testing
The oxidation reaction was performed in the liquid phase in
the batch mode using a lab-constructed mini-autoclave reactor.
1,2-Propanediol (0.5 M solution in distilled water, 1 ml) was mixed
with NaOH (1 M solution in distilled water, 0.5 ml) to obtain the
NaOH:sub ratio 1:1. The powdered catalysts (50 or 5 mg) were
added to the solution (molar ratio substrate:metal = 100–2000).
The reactor was charged with an oxygen–air mixture (78% O2/N2)
or an air to a 5 bar pressure and heated to the required temperature
(90 or 110 ◦C) under vigorous stirring. The contact time varied from
30 min to 20 h. The temperature of the reactor was maintained for
the desired reaction time, then it was cooled down in ice water,
and the gas phase products were collected in a gas bag. The liquid
products were filtered and the filtrate was collected as well.
The recyclability test was conducted with catalysts dried in air
without washing after the previous runs.
The monometallic 1%Au/TiO2 catalyst was prepared as well by
the DPU technique described elsewhere [54].
The catalysts were marked as x Au, x Pt, or x Au/y Pt, where x
and y were wt.%.
The metal loading was calculated as:
m
(metal) m
(support)−1 100%.
2.2. Catalyst characterization
2.2.1. X-ray photoelectron spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS) measurements were
performed on XSAM-800 with MgK␣ (1253 eV) X-ray source at
10 kV and 20 mA. The binding energies were calibrated with the
C 1s peak (285 eV) as the external standard reference while Ti 2p
(458.5 eV) was used as an additional internal standard reference. In
order to preserve the gold cations from possible reduction, the XPS
data were registered at −50 ◦C. The pressure in the experimental
chamber was kept at 3.7 × 10−10 Tоrr.
2.4. Product analysis
The liquid product analysis was performed by nuclear mag-
netic resonance (NMR) spectroscopy. 1H and 13C NMR spectra were
recorded with a Bruker AVANCE II 300 spectrometer (300.1 MHz)
in an H2O/D2O (10–20%) solution; the external standard was DSS.
Assignments of 1H and 13C signals and determination of the struc-
tures were made with the aid of 2D DOSY, COSY, editing-HSQC,
and HMBC spectra with or without suppression of the water sig-
nal when necessary. The presaturation and/or the gradient “W5”
[56,57] and “3919” WATERGATE [58] sequences were used for the
water signal suppression.
Yields and molar ratios of the products were calculated by inte-
gration of the signal in 1H NMR spectra on the basis of 3H. Sodium
benzoate was used as the external standard to calculate the yields.
The experimental error was within 3%. The time of recording of pro-
ton spectra was 5–10 min. The acids obtained were in the form of
sodium salts, but further in the text they have been named as just
acids for simplicity.
2.2.2. Diffuse reflectance Fourier-transform IR spectroscopy of
adsorbed CO (DRIFTS)
DRIFT spectra were recorded using a NICOLET “Protege” 460
spectrometer in the interval of 6000–400 cm−1 at a resolution of
4 cm−1 (500 scans). The adsorption of CO was performed at room
temperature (20 ◦C) and CO equilibrium pressure of 15 Torr. Before
the experiments, the sample was evacuated at a temperature not
exceeding the one used to treat the catalyst during its prepara-
tion. According to the DRIFTS experiment, the fractioned sample
(0.25–0.5 mm) was treated in a vacuum at 200, 250, 300, 400 and
500 ◦C for 30 min. After treating the sample at each temperature,
it was reduced in situ with hydrogen (60 Torr) for 30 min at the
same temperature. Then the sample was evacuated at the reduc-
tion temperature and cooled down to room temperature followed

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
173
reduction to Au0 with adsorbed hydrogen shifts the equilibrium
(Eq. (2b)) to the right, thereby increasing dissolution of Pt0. More-
over, the oxidation of Pt atoms with a low coordination number
to Ptn+ (Eq. (4)) and a further reduction of Ptn+ on Pt atoms with a
high coordination number by hydrogen occur in the HCl solution
as well (Eq. (3)). The processes described by (Eqs. (1) and (3) could
proceed simultaneously resulting in the formation of Au–Pt species
on the Pt surface, as in the present work the redox Au deposition
was performed in the hydrogen atmosphere. Hence, adsorption of
a new portion of hydrogen occurred on the platinum surface, and
a platinum chloride complex from the solution was reduced by
new portions of adsorbed hydrogen (Eq. (3)). The larger particles
and the contact zones between different particles should serve as
the centers for Pt deposition resulting in the growth of both the
particles and the contact areas between them, formally, sintering
Au and Pt nanoparticles. This phenomenon could result not only
in the growth of Pt NPs, but in their coverage with Pt deposit.
When the Au deposition was performed under Ar instead of H2 we
observed Pt leaching into the solution.
An epitaxial growth of gold (NPs) on Pt nanoparticles could
not be excluded in the lack of hydrogen as well. At room tem-
perature, the Pt NPs were proved to serve not only as an initial
catalytic reducing agent of the Au ions in the solution, but as the
seed nucleation platforms for the epitaxial growth with no crys-
tal modification resulting in AuPt heterodimers of the 2–4 times
larger size in the 0.4 mM solution of HAuCl4 in an organic solvent
(1,2-dichlorobenzene) [59].
Gas products were analyzed by GC with a Polysorb 6000 column
and a TC detector. It should be noted that some amount of CO2 was
in the form of sodium carbonate.
3. Results and discussion
3.1. Characterization of the catalysts
The prepared catalysts were studied by the different physico-
chemical methods and their properties were discussed in terms of
the preparation procedure used. STEM images of the monometallic
Au and Pt catalysts and of the 1Au/1Pt sample prepared by a redox
method are presented in Fig. 1.
A uniform distribution of isolated Au-NPs with a mean diameter
4 nm was obtained for the Au/TiO2 sample prepared using the DPU
procedure. In the STEM images of the Pt/TiO2 catalysts (Fig. 1b),
the isolated highly uniform Pt NPs with an extremely narrow size
distribution were observed. The mean particle size was 2 nm. Thus,
we have prepared highly dispersed Pt NPs with a narrow size distri-
bution on the surface of titanium oxide using platinum hydroxide
deposition under pH control.
Redox deposition of Au in a small amount of 0.05 and 0.025 wt.%
did not change the particle size distribution. Quite a different pic-
ture was observed in the sample with a high Au loading (Fig. 1c).
A bimodal NPs size distribution was characteristic of high-loaded
1Au/1Pt catalysts. There were small NPs along with the big ones
with mean diameters 3 and 10 nm, respectively. The lager par-
ticles have a complex morphology: they look like closely packed
aggregates of particles. The wide well-defined contact boundaries
between the areas are clearly seen for the big NPs in Fig. 1d due
to the different contrasts of Au and Pt-rich species forming the
large particle. The used catalyst did not change its morphology and
particle size distribution (Fig. 1e).
Table 1 represents the conditions used for the catalyst prepa-
ration by the redox method. Real redox potentials were calculated
according to the Nernst’s Eqs. (5) and (6):
ꢀ
ꢁ
RT
nF
[ox]
[red]
EM
EH
= E0
+
· ln
(5)
(6)
+
/M
ꢂ
ꢃ
[H+]2
pH
Redox-deposition of Au on the Pt NPs resulted in small bimetallic
Au/Pt NPs uniformly dispersed on the TiO2 surface when prepar-
ing the Au/Pt catalysts with a low Au content. The process of gold
deposition can be described by the set of equations (Eqs. (1)–(1b)).
RT
=
· ln
nF
2
The larger is the difference between the redox potentials of Au
and Pt electrode couples, the higher is the possibility of the direct
redox reaction to proceed, and two types of Au deposition can be
taken into account for the catalysts with a high Au:Pt ratio.
2[AuCl4]− + 3H2(ads) = 2Au0 + 6H+ads + 8Cl−
[AuCl4]− + 3e = Au0 + 4Cl− E0 = +0.99 V
(1)
(1a)
(1b)
Gold was found to deposit preferentially at high Miller-index
planes of Pt NPs during the direct redox reaction [41]. In our case,
this could lead to a larger size of Au NPs deposited. For a low
gold coverage, the quantity of adsorbed hydrogen was sufficient
to reduce all the introduced gold ions, gold was reduced by pread-
sorbed hydrogen and its deposition occurred on the entire Pt NPs
surface. For this reason, the size of particles in the bimetallic cat-
alysts with a low Au:Pt atomic ratio was very small. Indeed, for
the higher Au:Pt atomic ratio, we have obtained a bi-modal dis-
tribution of Au/Pt particles because of two parallel reactions of Au
deposition on the Pt NPs surface. As the contribution of the par-
ticles with the mean size about 10 nm is suppressed compared to
the contribution of the particles with the average diameter 3 nm, it
could be concluded that the rate of Au reduction with preadsorbed
hydrogen was higher than that of the direct redox reaction.
H2 − 2e = 2H+
E0 = +0.00 V
ads
At the same time, the growth of NPs was observed in the 1Au/1Pt
catalyst while the Au redox-deposition. A similar increase in the
size of the supported Pt nanoparticles from 1.2 to 6.5 nm after
treatment in the comparable conditions (pH 1, 1 h) was observed
previously when the amount of introduced Au exceeded 1/3 of
the amount of preadsorbed hydrogen atoms [41]. The fact was
attributed to the catalytic effect of [AuCl4]− ions on the mass
transport of platinum via the solution. However, there are several
reasons behind these phenomena in our case. When the Au:Pt ratio
is high, two parallel pathways of redox gold deposition are proba-
ble: the redox reaction with preadsorbed hydrogen (Eqs. (1)–(1b))
and the direct redox reaction (Eqs. (2)–(2b)) followed by Pt reduc-
tion (Eq. (3)).
Strong interaction between Au and Pt monometallic phases was
confirmed by the XPS method. The binding energy of Pt 4f differed
from the expected one for the metallic Pt species (71.1 eV) in the
1Au/1Pt catalyst. The shift of doublets of Pt 4f to the higher BE,
72.3 eV, and Au 4f to the lower BE, 84.3 eV, in comparison with the
expected values [31] (71.1 eV and 84 eV, respectively) indicated the
strong interaction between Au and Pt in 1Au/1Pt catalysts giving
rise to slightly positively charged metallic Pt NPs (Ptı+) and neg-
atively charged metallic Au NPs (Auı−) (Fig. 2a and b; Table 2). A
similar shift of Pt BE to the higher values with increasing Au load-
ing was reported previously for Au(core)–Pt(shell)/C catalysts [60].
[AuCl4]− + Pt = Au0 + [PtCl4]2−
(2)
(2a)
(2b)
(3)
[AuCl4]− + 3e = Au0 + 4Cl− E0 = +0.99 V
Pt + 4Cl− + 2e = [PtCl4]2− E0 = +0.73 V
[PtCl4]2− + 2H2(ads) − 2e = Pt0 + 4H+ + 4Cl−
Pt0 + 4Cl− + 4H+ + 2e = [PtCl4]2− + 2H2
(4)
The appearance of a significant amount of chloride ions, which
are generated in the HAuCl4 solution (Eq. (1)) in the course of

174
E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
Fig. 1. STEM images of (left): (A) 1%Au/TiO2; (B) 1%Pt/TiO2; (C) 1%Au/1%Pt/TiO2; (D) bimetallic Au–Pt particles in 1%Au/1%Pt/TiO2 (magnified); (E) 1%Au/1%Pt/TiO2-used after
the reaction provided at 110 ◦C, p(78%O2/N2) 5 bar for 20 h; and particle size distribution for each catalysts (right).

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
175
Table 1
The conditions used for Au deposition by the redox method and evaluated real redox potentials.
Catalyst
T (K)
C[AuCl4]− (mol/l)
pH
EH
E1
E2[PtCl4]2−/Pt (V)
Au deposition
Direct RR
H+/H2 (V)
[AuCl4]−/Au (V)
Reduction by
H2
0.025Au/1Pt
1Au/1Pt
295
0.63
8.16
3
1
−0.15
−0.12
1.11
1.06
0.99
0.84
−
+
(E1≈E2)
(E1 > EH
+
(E1 > EH
)
)
+
(E1 > E2)
Table 2
XPS characterization of Au–Pt catalysts.
Sample
BE (eV)Ti 3s
BE (eV)Au 4f7/2
BE (eV)Pt 4f7/2
Pt/Ti
Surface
Au/Ti
Au/Pt
Bulk
Au/Pt
1Au/1Pt
0.05Au/1Pt
0.025Au/1Pt
62.0
62.4
62.2
83.4
–
–
72.3
71.2
71.4
0.0243
0.0222
0.0200
0.0142
0.0025
0.0018
0.583
0.114
0.091
1
0.05
0.025
Strong Au–Pt interaction in the alloy Au–Pt/CeO2 catalyst resulting
in a major contribution of Pt2+ for the platinum species in the XPS
spectrum of the Pt 4f region was pointed out in another work as
well [46].
for this fact. Multilayer deposition of Au and Pt could lead to partial
coverage of Au NPs with platinum as it was described above. More-
over, La Parola et al. [61] reported the decrease in the surface Au/Pt
ratio determined by XPS in AuPt/SiO2 catalysts while increasing
the thickness of the platinum shell on the surface of the gold
core during the aging of the solution of Au and Pt precursors with
mercaptopropyltriethoxysilane. Another reason was the presence
of big Au/Pt agglomerates, which were observed by STEM.
DRIFT spectra of monometallic Pt/TiO2 and bimetallic
1Au/1Pt/TiO2 samples detected in the presence of 15 Torr of
CO after in situ reduction of the samples at 200, 250, 300, 400 and
500 ◦C are depicted in Figs. 3 and 4. Three bands are observed at
2185, 2068 and 1837 cm−1 assigned to linear adsorption of CO
on the support Ti4+–CO, linear adsorption of CO on the metallic
platinum and CO adsorbed in the bridging form on metallic
platinum, respectively, for the monometallic Pt catalyst reduced
in situ at 200 ◦C (Fig. 3) [62–64]. A higher calcination temperature
led to the blue shift of the band at 2068 cm−1 associated with
the Pt0–CO, but after the reduction at 500 ◦C the intensity of this
band drastically decreased as the intensity of other bands. The
DRIFT spectra of the bimetallic Au/Pt catalyst reduced in situ at
200 ◦C exhibited four bands of adsorbed CO at 2185, 2107, 2069,
1852 cm−1 (Fig. 4). The new band at 2107 cm−1 was attributed
to the CO adsorbed on metallic Au (Au0–CO) in accordance with
the published data [65]. Noteworthy that it can also be related
In contrast, the Pt 4f BE in the 0.05Au/1Pt sample coincided
with that of metallic platinum, and BE in the 0.025Au/1Pt cata-
lysts was just slightly shifted (practically within an experimental
error) to the higher values than expected ones (Table 2). Anyway,
we assume that when depositing a very small amount of gold onto
the Pt-NPs surface by the redox method, the most active sites of
platinum nanoparticles covered by preadsorbed hydrogen should
be involved in the oxidation–reduction reaction and strong Au–Pt
contacts could be formed. The determination of Au 4f binding ener-
gies in the low-loaded Au–Pt samples was difficult due to the low
intensity of Au 4f signals.
The Au:Pt atomic ratio on the sample surface of the low-loaded
Au/Pt samples considerably exceeded the ratio in the bulk (Table 2)
due to the higher dispersion of Au than Pt and coverage of the
Pt surface with Au species. In spite of the fact that the bulk gold
concentration in the 0.025Au/1Pt sample was twice lower than
that in the 0.05Au/1Pt catalyst, the surface Au/Pt ratio was almost
the same. This indicates the higher gold surface enrichment in
the low-loaded 0.025Au/1Pt catalyst. For the 1Au/1Pt sample, we
observed a different effect. The surface Au:Pt ratio was almost
twice lower than that in the bulk. There were two possible reasons
Fig. 2. High-resolution XPS spectra of Au/Pt/TiO2 catalysts: (A) Pt 4f region and (B) Au 4f + Pt 4f + Ti3s region.

176
E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
3,0
2,5
2,0
1,5
1,0
0,5
0,0
2,5
Pt0 - CO
200oC
2061
250oC
300oC
400oC
500oC
2,0
1,5
1,0
0,5
0,0
2068
ads. CO, 20oC, 15 Torr
vac. 20oC, 1 h
2051
vac. 100oC, 30 min
vac. 200oC, 30 min
Ti4+ - CO
2185
2031
1837
1832
Pt0 Pt0
2300
2200
2100
2000
1900
1800
1700
Wavenumber, cm-1
2300
2200
2100
2000
1900
1800
1700
Wavenumber, cm-1
Fig. 5. DRIFT spectra of CO (15 Torr) adsorbed on 1%Pt/TiO2 catalyst (reduced in situ
at 200 ◦C) and after evacuation of CO at different temperatures.
Fig. 3. DRIFT spectra of adsorbed CO (15 Torr, 20 ◦C) after in situ reduction of the
1%Pt/TiO2 catalyst.
The strength of CO adsorption in the bridged form on Pt NPs in
the bimetallic Au/Pt catalyst was weaker than in the monometal-
lic Pt/TiO2 catalyst, as the frequency of the band of CO adsorbed in
the bridging form for the 1Au/1Pt catalyst was higher (1852 cm−1
against 1837 cm−1) and thus adsorbed CO was much more eas-
ily removed during evacuation (even at 20 ◦C) compared to the
monometallic Pt sample. This effect can be accounted for by cover-
ing of Pt NPs by Au NPs in the bimetallic catalyst as a result of direct
redox deposition of gold.
If we compare the frequency of the linearly adsorbed CO for Pt
and 1Au/1Pt catalysts, i.e. the Pt0–CO band, after the evacuation
at 100 ◦C for 30 min (Table 3), we will see that below 300 ◦C the
frequency of the Pt–CO band is higher for the 1Au/1Pt catalysts and
close to the band at 2060 cm−1 previously attributed to CO adsorbed
on the bimetallic Au–Pt nanoparticles [61]. Anyway, this could be
also due to CO adsorption on bigger Pt NPs in the bimetallic 1Au/1Pt
catalyst, the presence of such NPs was proved by the STEM study.
At the higher temperature within 300–400 ◦C, the frequency of the
Pt0–CO band in the 1Au/1Pt sample became lower compared to
that for the monometallic Pt catalyst, and we can assume that gold
addition to the Pt catalyst led to its higher resistance to the particle
growth at high-temperature treatment.
to Au−–CO band [66], the presence of Au with a partial negative
charge in the 1Au/1Pt catalyst being observed by XPS in this work.
The other bands were the same as for the Pt/TiO2 sample. Higher
temperatures of in situ reduction of the sample also led to the
blue shift of the Pt0–CO band as for the monometallic Pt catalyst.
In addition, the band assigned to the Au0–CO band disappeared
after the reduction at 300 ◦C. The blue shift of the Pt0–CO band
and disappearance of the Au0–CO band after the high-temperature
reduction could be caused by the Pt and Au particle growth. It is
known that the strength of CO adsorption on metal nanoparticles
is size-dependent and small NPs adsorb CO stronger than the big
ones, especially it is noticeable in the case of Au NPs [67–69].
For the monometallic Pt catalyst, we also found that an increase
of the evacuation temperature after the CO adsorption led to a red
shift of the Pt0–CO bands (Fig. 5). The band of linearly adsorbed
CO on Pt0 in the spectra of the bimetallic Au/Pt catalysts was also
shifted to the low-frequency region of the spectra, and the Au0–CO
band completely disappeared even after evacuation at room tem-
perature (20 ◦C) like the other bands (Fig. 6). The shift of the
Pt0–CO bands to the lower frequencies was attributed to dimin-
ishing of dipole interaction between the adsorbed CO molecules.
2.0
2.5
200o
250o
300o
400o
500o
Pt0 - CO
2069
2063
2.0
1.5
1.0
0.5
0.0
1.5
ads. CO, 20oC, 15 Torr
2058
vac. 20oC, 1 h
Au0 - CO
Ti4+ - CO
vac. 100oC, 30 min
vac. 200oC, 30 min
1.0
2107
2185
2036
0.5
Pt0 Pt0
1852
0.0
2300
2200
2100
2000
1900
1800
1700
2300
2200
2100
2000
1900
1800
1700
Wavenumber, cm-1
Wavenumber, cm-1
Fig. 4. DRIFT spectra of adsorbed CO (15 Torr, 20 ◦C) after in situ reduction of the
1%Au/1%Pt/TiO2 catalyst.
Fig. 6. DRIFT spectra of CO (15 Torr) adsorbed on 1%Au/1%Pt/TiO2 catalyst (reduced
in situ at 200 ◦C) and after evacuation of CO at different temperatures.

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
177
0.90
0.85
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0.45
A
B
0.60
0.55
0.50
0.45
0.40
0.35
0.30
200
300
400
500
200
300
400
500
Reduction temperature, oC
Reduction temperature, oC
Fig. 7. Ti4+–CO/Pt0–CO intensity ratio in DRIFT spectra of linear adsorbed CO (15 Torr, 20 ◦C) for (A) 1%Pt/TiO2 and (B) 1%Au/1%Pt/TiO2 catalysts.
Moreover, we suggested that gold redox deposition on TiO2 sup-
ported Pt NPs was also responsible for the noticeable SMSI (strong
metal support interaction) effect. Comparing the ratio between
the intensity of the Ti4+–CO and Pt0–CO bands, we observed an
increase of the Ti4+–CO/Pt0–CO ratio for the Pt/TiO2 sample with an
increase of the reduction temperature from 200 to 500 ◦C (Fig. 7A).
This ratio for the bimetallic Au/Pt catalyst was much lower, and
after the reduction of the sample at 400 ◦C it sharply diminished
(Fig. 7B). This could be attributed to the higher rate of the TiO2
reduction with the formation of Ti3+ cations due to the effect of
strong Au–Pt and TiO2 interaction in the bimetallic catalyst. Since
Ti3+ cations do not form carbonyls, they cannot be seen in DRIFT
spectra. Thus, the lower intensity ratio of the Ti4+–CO and Pt0–CO
bands was observed. At the same time, the geometrical factor of
SMSI, i.e. covering of metal NPs with the reducible support [70–72],
can be responsible for the increase in the Ti4+–CO/Pt–CO ratio with
increasing reduction temperature for both samples. Therefore, the
results of the DRIFTS study also proved the selective Au deposition
on the Pt NPs by the redox method used and the effect of strong
Au–Pt and TiO2 interaction.
c
b
a
-50
150
350
550
750
Temperature, oC
Fig. 8. TPR profiles for the TiO2 (a), 1Pt/TiO2 (b), and 1Au/1Pt/TiO2 (c) samples.
clearly demonstrated that the redox deposition of Au on the TiO2
supported Pt NPs enhanced the TiO2 reduction, and the higher
hydrogen consumption was observed (Table 4).
Moreover, the later was also proved by TPR-H2 method (Fig. 8).
In the presence of Pt NPs and Au/Pt NPs the bulk TiO2 reduc-
tion started from 200 ◦C, and the unusual low-temperature peak
could be attributed to Ptı+ or Ti4+ reduction on the perimeter con-
tact of the Pt NPs or Au/Pt bimetallic particles and support. It was
3.2. 1,2-propanediol oxidation
All the prepared catalysts were active in 1,2-PD oxidation. Lac-
tic acid (LA), pyruvic acid (PvA), acetic acid (AcA), hydroxyacetone
(HA), acetone, CO2 were detected as reaction products. A typical
1H NMR spectrum of the liquid product mixture and the signals
assignment are depicted in Fig. 9, for more details of the struc-
ture identification by 1D and 2D NMR spectroscopy, see Supporting
Information S4–S13. The product yields data are shown in Table 5.
Almost complete conversion of 1,2-PD was observed after 2 h
of the reaction suspension stirring at 90 ◦C under a 5 bar pres-
sure of the oxygen–air mixture (78 vol.% O2) (Table 5). The main
Table 3
The frequency of stretching vibrations of linear Pt carbonyl in DRIFT spectra of
adsorbed CO after in situ reduction of 1%Pt/TiO2 and 1%Au/1%Pt/TiO2 at different
temperatures.
Temperature of in situ
reduction of the
Frequency of stretching
vibrations of Pt–CO
sample with H2 (◦C)
linear carbonyla (cm−1
)
1Pt/TiO2
2051
1Au/1Pt/TiO2
2058
200
250
300
400
500
2051
2056
Table 4
H2 consumption during 1Pt/TiO2 and 1Au/1Pt/TiO2 reduction.
2069
2063
Catalyst
Temperature
region (◦C)
H2 consumption (␮mol)
H2/g cat
2072
2068
1Pt
−50–200
3.55
83.5
No band
No band
200–850
1Au/1Pt
−50–200
14.7
138.5
a
CO was adsorbed (15 Torr) at room temperature and then the sample was evac-
uated for 30 min at 100 ◦C.
200–850

178
E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
Fig. 9. Typical 1H NMR spectra of the reaction product mixture (for more details see Supporting Information S5–S13).
product was lactic acid, although Au/TiO2 was active in the acetic
acid formation with almost the same selectivity as for LA. The
monometallic platinum catalyst appeared to be more active and
selective to lactic acid than the Au/TiO2 sample, which is in an
agreement with the findings of another research group [29]. In con-
trast, the gold–platinum combination mostly gave rise to pyruvic
acid as a by-product (Table 5).
The bimetallic sample with the lowest Au loading showed the
highest selectivity in 1,2-PD oxidation to LA. The addition of a
gold amount as small as 0.025 wt.% to the Pt/TiO2 sample by the
redox method significantly improved its selectivity to 91% at the
96% conversion. But it was observed that further increase in the
gold content in the bimetallic catalysts led to the lower values of
the selectivity to LA mostly because of higher amounts of pyruvic
acid formed (Fig. 10). The obtained results were somewhat differ-
ent from the previously published data [30]. It was reported that
the increase in the Au:Pt atomic ratio from 0.05 to 0.95 resulted in
as low LA selectivity as 56% due to the higher yield of acetic acid,
that was 14% even at 30 ◦C, but pyruvic acid was not detected at all.
This difference could be due to the state of Au and Pt because of the
use of a different preparation procedure in our work.
The reaction was carried out using air as an oxidant instead of
the oxygen–air mixture. Results are presented in Fig. 11. Oxidation
of 1,2-PD with an air over the Au/TiO2 catalyst led to the increased
selectivity to LA and 1,2-PD conversion, and we observed the
decrease in the amount of acetic acid and CO2. It is known that
Table 5
The effect of the catalyst composition on the 1,2-PD conversion and product distribution in 1,2-PD oxidation to lactic acid.
Entry
Catalyst
X1,2-Pd (%)
Yield (mol.%)
LA
PvA
AcA
HA
Cond. prod
Acetone
CO2
1
2
3
4
5
6
7
8
1Pt
1Au
89
47
100
100
96
6
63
17
58
70
87
6
100
Trace
5
1
13
9
2
0
0
0
6
13
6
6
3
0
0
0
1
0
1
2
1
0
0
0
7
1
4
3
1
0
0
0
1
0
6
14
18
7
2
0
1Au/1Pt
0.05Au/1Pt
0.025Au/1Pt
0.025Au/1Pta
0.025Au/1Ptb
TiO2
Trace
3
0
0
0
0
–
–
0
0
Reaction conditions: 0.5 M 1,2-PD; 1 ml; n(1,2-PD):n(NaOH) = 1; m cat. 50 mg; 90 ◦C; 2 h; p(78% O2/N2) = 5 bar.
a
The reaction was performed under the inert atmosphere at p(He) = 5 bar.
b
LA was used as the substrate, reaction time 1 h, LA, lactic acid; PvA, pyruvic acid; AcA, acetic acid; HA, hydroxyacetone; 1,2-PD, 1,2-propanediol; Cond. prod, condensation
products, in the case of Au–Pt catalysts they are 2-hydroxy-2-methyl-4-oxopentanedioic and 2,2-dihydroxypropanoic acid.

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
179
Fig. 10. Au/Pt/TiO2 catalysts efficiency for acids. LA, lactic acid; AcA, acetic acid;
PvA, pyruvic acid, HA, hydroxyacetone, 1,2-PD, 1,2-propanediol. Reaction condi-
tions: 0.5 M 1,2-PD; 1 ml; n(1,2-PD):n(NaOH) = 1; 50 mg cat.; 90 ◦C, 2 h, p 5 bar of
78% O2/N2.
CO2 forms surface carbonates over the gold catalysts supported
on reducible oxides resulting in catalysts deactivation [73,74]. The
use of air, which acted as a milder and more selective oxidant than
pure oxygen, therefore could prevent CO2 formation and Au/TiO2
catalyst deactivation by surface carbonates. It led to the higher
conversion of 1,2-PD and higher selectivity to LA. In the case of
the Au–Pt catalysts, the reaction with air instead of oxygen–air
mixture also resulted in the higher selectivity to LA because
of suppressed 1,2-PD oxidation to pyruvic acid, but the lower
conversion of 1,2-PD was obtained for the same reaction time.
Anyway, the activity and selectivity of the bi-metallic sample were
higher than those of the monometallic Au catalyst. The observed
higher selectivity to LA, when using air instead of oxygen, is not in
line with the previously reported results for the activated carbon
supported 1%Au–Pt nanoparticles: the selectivity to LA on those
catalysts dropped from 86% (at 78% conversion) to 81% (at 38%
conversion) if the reaction was run in air [30].
Fig. 12. Liquid product distribution vs. time obtained at different temperatures: (A)
90 ◦C and (B) 110 ◦C over 1%Au/1%Pt/TiO2 catalyst. ᭹, lactic acid;
, pyruvic acid; ꢀ,
acetic acid; , hydroxyacetone; ꢁ, 1,2-propanediol. Reaction conditions: 0.5 M 1,2-
PD, 1 ml; m cat. 50 mg; n(1,2-PD):n(NaOH) = 1; 90 ◦C or 110 ◦C; p(78% O2/N2) = 5 bar,
reaction time 0.5–20 h; the molar concentration was calculated on the basis of the
liquid phase.
the full 1,2-PD conversion and an increased molar concentration
of lactic, pyruvic, acetic acids, and hydroxyacetone (which is a tau-
tomeric form of lactaldehyde under the basic or acidic conditions).
At the same time, the pH of the liquid dropped from 10–12 to
7–8 because of obtained acids neutralization by NaOH. The fur-
ther increase of the reaction time led to the increase of the amount
of acetic and pyruvic acids in the liquid phase, while the LA con-
centration remained almost the same and HA concentration fell to
the trace amount. At the higher temperature, 110 ◦C, deeper and
fast oxidation of 1,2-PD was observed (Fig. 12B). Almost full 1,2-PD
conversion was achieved in 1 h of the reaction. The concentration of
LA dropped in the whole range of the reaction time, while the con-
centration of PvA and AcA was higher than that obtained at 90 ◦C. In
addition, higher concentration of HA was observed in the reaction
mixture after 2 and 6 h experiments. The increase of the reaction
time also led to the simultaneous increase of PvA and AcA concen-
tration and HA disappearance. According to the results obtained,
we can conclude that HA was the intermediate product, which led
preferably to PvA under the pH close to neutral and to LA in the
basic medium. Severe conditions (high temperature and prolong
reaction time) were responsible for unselective primary and sec-
ondary OH-group oxidation of 1,2-PD with higher amount of HA
formation and it also led to deeper 1,2-PD oxidation with C C bond
scission.
The influence of the temperature and contact time on the prod-
uct distribution in the liquid phase was studied over the 1Au/1Pt
catalyst (Fig. 12A). The experiments showed that the increase of
the reaction time from 30 min to 2 h at 90 ◦C led to the achieving
Comparison of the initial oxidation activity of the bimetallic
catalysts under the low conversion conditions (Table 6) revealed
that the highest TON value and 100% selectivity were obtained for
the 0.025Au/1Pt catalyst. It is worth mentioning that oxidation of
1,3-propanediol (1,3-PD) was rather hard to perform and a 10%
Fig. 11. Product selectivity and conversion in respect to oxygen or air used as an
oxidant. LA, lactic acid; AcA, acetic acid; PvA, pyruvic acid, HA, hydroxyacetone, 1,2-
PD, 1,2-propanediol. Reaction conditions: 0.5 M 1,2-PD, 1 ml; m cat. 50 mg; n(1,2-
PD):n(NaOH) = 1; 90 ◦C; 2 h; p 5 bar of 78% O2/N2or air.

180
E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
Table 6
TON values and product yields in 1,2-Pd oxidation to lactic acid over the Pt and bi-metallic Au–Pt catalysts.
Entry
Catalyst
Yield (mol.%)
LA
AcA
FA
Acetone
CO2
X1,2-Pd (%)
SLA (%)
TONe
1
2
3
4
5
6
7
8
1Pt
12
35
21
13
26
0
Trace
3
1
Trace
0
4
3
0
0
0
0
0
6
1
0
0
0
6
1
0
0
0
7
0
12
44
23
13
26
10
47
26
100
81
92
100
100
–
82
113
69
87
181
–
1Au/1Pt
Trace
0
0
0
1Au/1Pta
0.05Au/1Pt
0.025Au/1Pt
0.025Au/1Ptb
1Au/1Pt-usedc
0.025Au/1Pt-usedd
0
36
26
Trace
0
77
100
107
181
Trace
Reaction conditions: 0.5 M 1,2-PD; 1 ml; n(1,2-PD):n(NaOH) = 1; m cat. 5 mg; 90 ◦C; 2 h; p(78% O2/N2) = 5 bar.
a
0.5 M 1,2-PD; 1 ml; n(1,2-PD):n(NaOH) = 1; m cat. 5 mg; 60 ◦C; 2 h; p(78% O2/N2) = 3 bar.
b
0.5 M 1,3-PD; 1 ml; n(1,3-PD):n(NaOH) = 1; m cat. 5 mg; 90 ◦C; 2 h; p(78% O2/N2) = 5 bar.
c
The catalyst was reused after the reaction performed at 110 ◦C under 5 bar of 78% O2/N2 for 20 h; 0.5 M 1,2-PD; 1 ml; n(1,2-PD):n(NaOH) = 1; without washing; m cat.
5 mg.
d
The catalyst was used for three times.
e
TON was calculated as mol (LA) mol (Au + Pt)−1 h−1 LA, lactic acid; AcA, acetic acid; FA, formic acid; 1,2-PD, 1,2-propanediol.
conversion was achieved on the 0.025Au/1Pt catalyst with only
acetic and formic acids formation (Table 6, entry 6). The 1Au/1Pt
sample was rather active too in LA formation from 1,2-PD, although
the selectivity to LA was lower because of further 1,2-PD oxidation
to acetic acid and CO2. The activity of the 1Au/1Pt catalyst was also
tested under milder conditions: 60 ◦C and 3 bar of the oxygen–air
mixture (Table 6, entry 3). In spite of the fact that the selectivity to
LA was high (92%), the TON value was far lower than that obtained
for the 0.025Au/1Pt catalyst at 90 ◦C and 5 bar of the oxygen–air
mixture. The initial activity of the 0.05Au/1Pt catalyst was below
those of other bimetallic catalysts with an increased and decreased
Au:Pt atomic ratio, but exceeded the activity of Pt/TiO2. Thus,
there is an optimal Au:Pt atomic ratio. There were several reasons
under this finding. According to the XPS data, platinum and gold
in the high-loaded 1Au/1Pt sample were not fully metallic in their
character indicating noticeable electron donation from platinum
to gold because of strong Au–Pt interaction. The DRIFTS and TPR
studies also proved enhanced interaction between the Au/Pt NPs
and TiO2 support. All that resulted in the high oxidation activity
of the 1Au/1Pt sample. As it was already mentioned, gold in the
0.025Au/1Pt sample was supposed to be deposited on the most
active sites of Pt NPs. Hence, this should lead to strong Au–Pt inter-
action but without formation of big nanoparticles, and gold was
highly dispersed on the surface of Pt NPs in contrast to the 1Au/1Pt
sample. In should be also noticed that the peripheral interaction
of small bimetallic Au/Pt NPs and TiO2 proved by the DRIFTS-CO
study can be the region of oxygen adsorption that is responsible for
the 1,2-PD oxidation, as it was proposed in [75] for CO oxidation.
The higher is the dispersion of Au/Pt NPs, the higher is the number
of peripheral sites for oxygen adsorption. Therefore, the deposition
of as low as 0.025 wt.% of gold on the Pt NPs active sites increased
their oxidation activity, but made them more selective in primary
hydroxyl group oxidation due to the specific Au interaction with
the most active sites of Pt NPs and, thus, the C C bond scission was
avoided in comparison to the 1Au/1Pt catalyst with a high gold
content. The twofold increased gold content in 0.05Au/1Pt catalyst
gave rise to the lower Au dispersion on the surface of Au/Pt NPs than
in the 0.025Au/1Pt sample, which was obtained by XPS study. We
presumed that while the redox deposition of gold in amount higher
than 0.025:1 atomic ratio, not only the most active sites of Pt NPs
with preadsorbed hydrogen served as centers for Au deposition,
but the whole accessible surface of Pt NPs covered by hydrogen
was involved. Therefore, the 0.05Au/1Pt catalyst lose the specific
activity that was characteristic of the 0.025Au/1Pt sample. Big
bimetallic NPs, the surface of which was enriched with Pt, and the
higher gold content in the 1Au/1Pt catalyst could lead to the higher
amounts of reactive O-species. Such active O-species easily react
with the substrate giving the products of unselective 1,2-PD oxida-
tion, i.e. oxidation of the secondary hydroxyl group along with the
primary OH-group, and C C bond cleavage. Hence, the electronic
effects, the surface properties of Au/Pt NPs, and the Au:Pt atomic
ratio in the Au/Pt/TiO2 catalysts governed the oxidation activity of
the catalysts and their selectivity in 1,2-PD oxidation to LA.
The Au–Pt catalyst showed an excellent stability. No particle
growth was observed after the reaction run at 110 ◦C and 5 bar of
the oxygen–air mixture for 20 h over 1Au/1Pt (Fig. 1e). The catalysts
0.025Au/1Pt and 1Au/1Pt almost did not lose their activity (Table 6,
entries 7and 8) and no leaching was observed by analytical methods
used to determine the presence of Au and Pt in the solution [53].
3.2.1. Proposed scheme of 1,2-propanediol oxidation
According to the results obtained and taking into account the
published data, we can propose a general scheme of the 1,2-PD
oxidation over Au/Pt/TiO2 catalysts prepared by the redox method
(Fig. 13).
Lactic acid could be formed as a result of both direct 1,2-
PD oxidation and via the pyruvaldehyde formation followed by
the intramolecular Cannizzaro reaction. The presence of the small
amounts of hydroxyacetone (which is a tautomeric form of lac-
taldehyde or 2-hydroxypropanal) and pyruvic acid in the product
mixture proved that the Cannizzaro reaction took place in addition
to direct oxidation, especially for platinum-containing samples. We
can also claim that, as far as the oxidation of LA used as a sub-
strate did not occur over the 0.025Au/1Pt catalyst (Table 5, entry
7), pyruvic and acetic acids were formed through the intermediate
products oxidation, probably, through hydroxyacetone oxidation,
while pyruvic acid could be oxidized to acetic acid as well [76].
Other minor products were observed in 1H NMR spectra of the
product mixture under the conditions used at full 1,2-PD conver-
sion. When running the reaction over 1Au/1Pt catalysts for 2–20 h
and at 90–110 ◦C, 2,2-dihydroxypropanoic acid (5%, mol/mol of
liquid products) and 2-hydroxy-2-methyl-4-oxopentanedioic (9%,
mol/mol of liquid products) acids (Figs. 9 and 13) were formed.
Smaller amounts of these acids were also found in the reaction
mixture for 1,2-PD oxidation over the monometallic Pt catalyst at
severe conditions only: 110 ◦C and 24 h. 2,2-Dihydroxypropanoic
acid was also detected in the product mixture for 1,2-PD oxidation
over the 0.05Au/1Pt catalyst under the same conditions, but the
yield was lower, 3% (mol/mol of liquid products). The structures
of the acids were determined by a set of 1H–1H COSY, 1H–13
C
HSQC, and 1H–13C HMBC NMR spectroscopy techniques (See
Supporting Information S5–S13). To the best of our knowledge,
the formation of these acids in the reaction of 1,2-PD oxidation has
not been reported yet. Probably, these products could originate

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
181
OH
OH
Propane-1,2-diol (1,2-PD)
H2O
H2O
H2O2
H2O2
O2
O2
OH
O
OH-
O
O2
O
OH
Cat.
2-hydroxypropanal
Hydroxyacetone (HA)
Acetone
O2
OH
O2
O2
HO
O2
Cat.
Cat.
Cat.
Cat.
O2
HOOC
HO
OH
OH-
O
2,2-dihydroxypropanoic acid
O
COOH
Cannizzaro reaction
COOH
Lactic acid (LA)
HO
Cat.
COOH
2-oxopropanal
Acetic acid (AcA)
PvA, t
O2
HO
O
Cat.
HOOC
O
2-hydroxy-2-methyl-4-oxopentanedioic acid
COOH
Pyruvic acid (PvA)
Aldol cond.
OH-
OH-
COOH
O
OH
HO
OH
H O
OH-
2
HO
H+
COOH
HOOC
HOOC
2-hydroxy-2-methyl-4-oxopentanedioic acid
2,2-dihydroxypropanoic acid
Fig. 13. Proposed scheme of the 1,2-PD oxidation to lactic acid. Marked structures were proved by 1H, 13C NMR spectroscopy with the aid of 2D techniques HSQC, HMBC,
COSY (see Supporting Information S5–S15).
from the radicals recombination. It seems possible that hydrogen
peroxide may be generated in situ through water and/or OH−
ions oxidation over Au–Pt/TiO2 catalysts, and H2O2 decomposition
gave rise to hydroperoxide or hydroxyl radicals. Hydrogen perox-
ide formation was observed over the Au–Pd supported catalyst.
H2O2 formed O-centered radicals (detected by ESR with DMPO
{5,5-dimethyl-1-pyrroline N-oxide} spin trapping), which
enhanced the methane to methanol oxidation reaction [77,78].
The promoting effect of hydroxyl radicals obtained from OH−
was discussed for aqueous-phase CO and glycerol oxidation over
Au/C catalysts [14,79]. Thus, it should be noted that the reason
of the high oxidation activity of the Au/Pt/TiO2 catalyst, espe-
cially of the 1Au/1Pt/TiO2 sample, also could be due to hydrogen
peroxide formation. Recently, Hutchings et al. [80] reported the
enhanced activity of Pt-doped bimetallic Au–Pd/CeO2 catalyst
in H2O2 formation. Moreover, H2O2 could be activated on the
TiO2 support as surface-stabilized peroxo- or/and hydroperoxo
At the same time, it should be noted that the concentra-
tion of the minor acids in the liquid phase correlated with the
concentration of pyruvic and lactic acid. As it was shown, the
oxidation of 1,2-PD under the harsh conditions led to the higher
amount of pyruvic acid in the liquid, and the higher amount
of the minor acids was observed as well. Therefore, we could
also suggest that 2,2-dihydroxypropanoic acid was formed by the
hydration of the enol form of pyruvic acid in the alkali media.
The 2-hydroxy-2-methyl-4-oxopentanoic acid could be formed
via the aldol condensation reaction of the PvA under the acid
or base catalysis. However, even after providing the reaction
of 1,2-PD oxidation for 2 h the pH of the liquid was close to
neutral, which inhibited the enolization of PvA, the acid catal-
ysis having been also restricted due to acids neutralization by
NaOH.
4. Conclusion
−
species (Ti–O2 or/and Ti–OOH), as it was proved by Bonino
et al. [81], or just decomposed to O-centered radicals leading to
non-selective secondary hydroxyl group oxidation in 1,2-PD, and
The Au/Pt/TiO2 catalysts prepared by a simple redox method
are promising systems for obtaining of lactic acid through the 1,2-
propanediol oxidation by molecular oxygen or just by air, with the
selectivity to lactic acid increasing when air was used as an oxidant.
The most efficient catalyst with the highest selectivity to LA was
the sample with the lowest Au content. The 0.025%Au/1%Pt/TiO2
catalyst produced lactic acid with a 91% selectivity at a 96% con-
version of 1,2-propanediol. The high activity of the low-loaded
C
C bond scission under the more severe conditions. The fact
that large amounts of CO2 were formed during glycerol oxidation
even under mild conditions (see Supporting Information Table S1)
also could indicate the possible assistance of the radical route of
the reaction as far as glycerol is a more reactive substrate than
1,2-PD [76].

182
E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
0.025%Au/1%Pt/TiO2 catalyst could be due to the presence of the
highly dispersed bimetallic Au–Pt particles of the size about 1–2 nm
with strong Au–Pt synergetic effect and strong interaction of Au/Pt
NPs with TiO2 support. An increase of the amount of Au deposited
led to the more pronounced effects of the Au–Pt interaction in the
bimetallic NPs and also of the interaction of the bimetallic Au/Pt
NPs with support, which was proved by XPS, DRIFTS-CO, TPR-
H2 methods. The high-loaded 1%Au/1%Pt/TiO2 catalyst provided
deeper 1,2-PD oxidation at 90 ◦C and 5 bar of the oxygen–air mix-
ture than low-loaded 0.025%Au/1%Pt/TiO2 sample, while at milder
reaction conditions its TON value was lower than TON obtained for
the 0.025%Au/1%Pt/TiO2 catalyst.
Redox-prepared Au/Pt/TiO2 catalysts showed an excellent sta-
bility proved by repeated catalytic tests, STEM study, and no
leaching of metallic NPs was observed.
Application of NMR spectroscopy methods allowed accurate
products analysis and their precise identification in the reaction
mixture. We obtained the acids not detected in the previously
published works. The general scheme of 1,2-PD oxidation, which
included the possible radical processes, was suggested.
[21] D. Tongsakul, S. Nishimura, K. Ebitani, ACS Catal. 3 (2013) 2199–2207.
[22] G.L. Brett, Q. He, C. Hammond, P.J. Miedziak, N. Dimitratos, M.A. Sankar, A.
Herzing, M. Conte, J.A. Lopez-Sanchez, C. Kiely, G.J. Hutchings, Angew. Chem.
Int. Ed. 50 (2011) 10136–10139.
[23] 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
7 (2014) 1326–1334.
[24] C.B. Rasrendra, B.A. Fachri, I.G.B.N. Makertiharthra, S. Adisasmito, H.J. Heeres,
ChemSusChem 4 (2011) 768–777.
[25] S. Lux, M. Siebenhofer, Catal. Sci. Technol. 3 (2013) 1380–1385.
[26] Z. Huang, F. Cuia, J. Xue, J. Zuo, J. Chen, C. Xia, Catal. Today 183 (2012) 42–51.
[27] R.B. Mane, A.M. Hengne, A.A. Ghalwadkar, S. Vijayanand, P.H. Mohite, H.S. Pot-
dar, C.V. Rode, Catal. Lett. 135 (2010) 141–147.
[28] M. Sankar, N. Dimitratos, P.J. Miedziak, P.P. Wells, C.J. Kiely, G.J. Hutchings,
Chem. Soc. Rev. 41 (2012) 8099–8139.
[29] Yu. Ryabenkova, P.J. Miedziak, N.F. Dummer, S.H. Taylor, N. Dimitratos,
D.J. Willock, D. Bethell, D.W. Knight, G.J. Hutchings, Top. Catal. 55 (2012)
1283–1288.
[30] Yu. Ryabenkova, Q. He, P.J. Miedziak, N.F. Dummer, S.H. Taylor, A.F. Carley, D.J.
Morgan, N. Dimitratos, D.J. Willock, D. Bethell, D.W. Knight, D. Chadwick, C.J.
Kiely, G.J. Hutchings, Catal. Today 203 (2013) 139–145.
[31] E.R. Essinger-Hileman, D. DeCicco, J.F. Bondi, R.E. Schaak, J. Mater. Chem. 21
(2011) 11599–11604.
[32] M. Comotti, C.D. Pina, M. Rossi, J. Mol. Catal. A: Chem. 251 (2006) 89–92.
[33] C.L. Bianchi, P. Canton, N. Dimitratos, F. Porta, L. Prati, Catal. Today 102–103
(2005) 203–212.
[34] A. Villa, D. Wang, G.M. Veith, F. Vindignid, L. Prati, Catal. Sci. Technol. 3 (2013)
3036–3041.
[35] J.A. Lopez-Sanchez, N. Dimitratos, C. Hammond, G.L. Brett, L. Kesavan, S. White,
P. Miedziak, R. Tiruvalam, R.L. Jenkins, A.F. Carley, D. Knight, C.J. Kiely, G.J.
Hutchings, Nat. Chem. 3 (2011) 551–556.
[36] J. Barbier, J.C. Menezo, C. Montassier, J. Naja, G. Del Angel, J.M. Dominguez, Catal.
Lett. 14 (1992) 37–43.
[37] Z. Bodnár, T. Mallát, S. Szabó, J. Petró, Stud. Surf. Sci. Catal. 63 (1991) 459–468.
[38] I. Bakos, S. Szabó, J. Electroanal. Chem. 547 (2003) 103–107.
[39] E.A. Redina, A.A. Greish, I.V. Mishin, G.I. Kapustin, O.P. Tkachenko, O.A.
Kirichenko, L.M. Kustov, Catal. Today (2014), http://dx.doi.org/10.1016/j.
cattod.2013.11.065.
[40] M. Bonarowska, B. Burda, W. Juszczyk, J. Pielaszek, Z. Kowalczyk, Z. Karpinski,
Appl. Catal. B: Environ. 35 (2001) 13–20.
[41] J. Barbier, P. Marécot, G. DelAngel, P. Bosch, J.P. Boitiaux, B. Didillon, J.M.
Dominguez, I. Schifter, G. Espinoza, Appl. Catal. A: Gen. 116 (1994) 179–186.
[42] C. Micheaud, P. Marecot, M. Guerin, J. Barbier, Appl. Catal. A: Gen. 171 (1998)
229–239.
Acknowledgments
Electron microscopy characterization was performed in the
Department of Structural Studies of Zelinsky Institute of Organic
Chemistry, Moscow. This work was partially supported by the
Russian Basic Research Foundation grant OFI-M-13-03-12234. The
financial support of the Ministry of Education and Science of Rus-
sian Federation (project no. RFMEFI61314X0012) is acknowledged.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.apcata.
2014.11.039.
[43] O.A. Kirichenko, E.A. Redina, N.A. Davshan, I.V. Mishin, G.I. Kapustin, T.R.
Brueva, L.M. Kustov, W. Li, C.H. Kim, Appl. Catal. B: Environ. 134–135 (2013)
123–129.
[44] G. Budroni, S. Kondrat, S. Taylor, D. Morgan, A. Carley, G. Hutchings, Catal. Sci.
Technol. 3 (2013) 2746–2754.
References
[45] S. Demirel, K. Lehnert, M. Lucas, P. Claus, Appl. Catal. B: Environ. 70 (2007)
637–643.
[46] R. Kumar, P. Purushothamana, J. van Haveren, D.S. van Es, I. Melián-Cabrera,
J.D. Meeldijk, H.J. Heeres, Appl. Catal. B: Environ. 147 (2014) 92–100.
[47] C. Alonso, J. Gonzalez-Velasco, J. Electroanal. Chem. Interfacial Electrochem.
248 (1988) 193–208.
[48] W.P. Aue, E. Bartholdi, R.R. Ernst, J. Chem. Phys. 64 (1976) 2229–2246.
[49] D. Wu, A. Chen, C.S. Johnson, J. Magn. Reson. A115 (1995) 260–264.
[50] L.E. Kay, P. Keifer, T. Saarinen, J. Am. Chem. Soc. 114 (1992) 10663–10665.
[51] J. Schleucher, M. Schwendinger, M. Sattler, P. Schmidt, O. Schedletzky, S.J.
Glaser, O.W. Sorensen, C. Griesinger, J. Biomol. NMR 4 (1994) 301–306.
[52] A. Bax, M.F. Summers, J. Am. Chem. Soc. 108 (1986) 2093–2094.
[53] S.I. Ginzburg, K.A. Kladyshevskaya, Handbook on Chemical Analysis of Platinum
Metals and Gold, Nauka, Moscow, 1965 (in Russia).
[54] R. Zanella, S. Giorgio, C.-H. Shin, C.R. Henry, C. Louis, J. Catal. 222 (2004)
357–367.
[55] V.V. Kachala, L.L. Khemchyan, A.S. Kashin, N.V. Orlov, A.A. Grachev, S.S.
Zalesskiy, V.P. Ananikov, Russ. Chem. Rev. 82 (2013) 648–685.
[56] M. Liu, X. Mao, C. He, H. Huang, J.K. Nicholson, J.C. Lindon, J. Magn. Reson. 132
(1998) 125–129.
[1] A. Behr, J. Eilting, K. Irawadi, J. Leschinski, F. Lindner, Green Chem. 10 (2008)
13–30.
[2] C.-H. Zhou, J.N. Beltramini, Y.-X. Fana, G.Q. Lu, Chem. Soc. Rev. 37 (2008)
527–549.
[3] B. Katryniok, H. Kimura, E. Skrzyn´ ska, J.-S. Girardon, P. Fongarland, M. Capron,
R. Ducoulombier, N. Mimura, S. Paul, F. Dumeignil, Green Chem. 13 (2011)
1960–1979.
[4] P.S. Shuttleworth, M. De Bruyn, H.L. Parker, A.J. Hunt, V.L. Budarin, A.S. Matharu,
J.H. Clark, Green Chem. 16 (2014) 573–584.
[5] H. Kishida, F. Jin, Z. Zhou, T. Moriya, H. Enomoto, Chem. Lett. 34 (2005)
1560–1561.
[6] Z. Shen, F.M. Jin, Y.L. Zhang, B. Wu, A. Kishita, K. Tohji, H. Kishida, Ind. Eng.
Chem. Res. 48 (2009) 8920–8925.
[7] C.A. Ramírez-Lypez, J.R. Ochoa-Gymez, M. Fernandez-Santos, O. Gymez-
Jiminez-Aberasturi, A. Alonso-Vicario, J.S. Torrecilla-Soria, Ind. Eng. Chem. Res.
49 (2010) 6270–6278.
[8] E.P. Maris, W.C. Ketchie, M. Murayama, R.J. Davis, J. Catal. 251 (2007) 281–294.
[9] M. Checa, F. Auneau, J. Hidalgo-Carrillo, A. Marinas, J.M. Marinas, C. Pinel, F.J.
Urbano, Catal. Today 196 (2012) 91–100.
[57] G. Zheng, W.S. Price, Prog. Nucl. Mag. Reson. Spectr. 56 (2010) 267–288.
[58] M. Piotto, V. Saudek, V. Sklenar, J. Biomol. NMR 2 (1992) 661–666.
[59] S.I. Lim, M. Varon, I. Ojea-Jiménez, J. Arbiol, V. Puntes, J. Mater. Chem. 21 (2011)
11518–11523.
[10] F. Auneau, S. Noël, G. Aubert, M. Besson, L. Djakovitch, C. Pinel, Catal. Commun.
16 (2011) 144–149.
[11] F. Auneau, L.S. Arani, M. Besson, L. Djakovitch, C. Michel, F. Delbecq, P. Sautet,
C. Pinel, Top. Catal. 55 (2012) 474–479.
[60] S. Yan, S. Zhang, Int. J. Hydrogen Energy 37 (2012) 9636–9644.
[61] V. La Parola, M. Kantcheva, M. Milanova, A.M. Venezia, J. Catal. 298 (2013)
170–178.
[62] K.I. Hadjiivanov, J. Chem. Soc. Faraday Trans. 94 (13) (1998) 1901–1904.
[63] K.I. Hadjiivanov, G.N. Vayssilov, Adv. Catal. 47 (2002) 307–511.
[64] A.A. Davydov (Ed.), Molecular Spectroscopy of Oxide Catalyst Surfaces, Wiley
Interscience Publisher, 2003, p. 466.
[65] M.A. Bollinger, M.A. Vannice, Appl. Catal. B: Environ. 8 (1996) 417–443.
[66] M.S. Chen, Y. Cai, Z. Yan, D.W. Goodman, J. Am. Chem. Soc. 128 (2006)
6341–6346.
[67] C. Lemire, R. Meyer, Sh.K. Shaikhutdinov, H.-J. Freund, Surf. Sci. 552 (2004)
27–34.
[12] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.J. Hutchings, Phys.
Chem. Chem. Phys. 5 (2003) 1329–1336.
[13] S. Demirel, M. Lucas, J. Wärnå, T. Salmi, D. Murzin, P. Claus, Top. Catal. 44 (2007)
299–305.
[14] W.C. Ketchie, M. Murayama, R.J. Davis, Top. Catal. 44 (2007) 307–317.
[15] D. Wang, A. Villa, F. Porta, L. Prati, D. Su, J. Phys. Chem. C 112 (2008) 8617–8622.
[16] L. Prati, P. Spontoni, A. Gaiassi, Top. Catal. 52 (2009) 288–296.
[17] Y. Shen, S. Zhang, H. Li, Y. Ren, H. Liu, Chem. Eur. J. 16 (2010) 7368–7371.
[18] R. Kumar, P. Purushothaman, J. van Haveren, D.S. van Es, I. Melián-Cabrera, J.D.
Meeldijk, H.J. Heeres, Appl. Catal. B: Environ. 147 (2014) 92–100.
[19] L. Prati, A. Villa, C. Campione, P. Spontoni, Top. Catal. 44 (2007) 319–323.
[20] A. Villa, G.M. Veith, L. Prati, Angew. Chem. 122 (2010) 4601–4604;
A. Villa, G.M. Veith, L. Prati, Angew. Chem. Int. Ed. 49 (2010) 4499–4502.
[68] C. Ruggiero, P. Hollins, Surf. Sci. 377–379 (1997) 583–586.

E. Redina et al. / Applied Catalysis A: General 491 (2015) 170–183
183
[69] L. Fan, N. Ichikuni, Sh. Shimazu, T. Uematsu, Appl. Catal. A: Gen. 246 (2003) 87.
[70] S.J. Tauster, S.C. Fung, R.L. Garten, J. Am. Chem. Soc. 100 (1978) 170–175.
[71] S.J. Tauster, S.C. Fung, J. Catal. 55 (1978) 29.
[72] S.J. Tauster, S.C. Fung, R.T.K. Baker, J.A. Horsley, Science 211 (1981) 1121.
[73] Y. Denkwitz, Z. Zhao, U. Hörmann, U. Kaiser, V. Plzak, R.J. Behm, J. Catal. 251
(2007) 363–373.
[74] W. Deng, M. Flytzani-Stephanopoulos, Angew. Chem. 118 (2006) 2343–2347;
W. Deng, M. Flytzani-Stephanopoulos, Angew. Chem. Int. Ed. 45 (2006)
2285–2289.
[77] J.K. Edwards, B.E. Solsona, P. Landon, A.F. Carley, A. Herzing, C.J. Kiely, G.J.
Hutchings, J. Catal. 236 (2005) 69–79.
[78] M.H. Ab Rahim, M.M. Forde, R.L. Jenkins, C. Hammond, N. Dimitratos, J.A. Lopez-
Sanches, A.F. Carley, S.H. Taylor, D.J. Willock, D.M. Murphy, G.J. Hutchings,
Angew. Chem. Int. Ed. 52 (2013) 1280–1284.
[79] W.C. Ketchie, M. Murayama, R.J. Davis, J. Catal. 250 (2007) 264–273.
[80] J.K. Edwards, J. Pritchard, L. Lu, M. Piccinini, G. Shaw, A.F. Carley, D.J.
Morgan, C.J. Kiely, G.J. Hutchings, Angew. Chem. Int. Ed. 53 (2014)
2381–2384.
[75] A.P. Markusse, B.F.M. Kusterand, J.C. Schouten, Catal. Today 66 (2001) 191–197.
[76] N. Dimitratos, J.A. Lopez-Sanchez, M. Sankar, J.M. Anthonykutty, G. Brett,
A.F. Carley, S.H. Taylor, D.W. Knight, G.J. Hutchings, Green Chem. 11 (2009)
1209–1216.
[81] F. Bonino, A. Damin, G. Ricchiardi, M. Ricci, G. Spano, R. D’Aloisio, A.
Zecchina, C. Lamberti, C. Prestipino, S. Bordiga, J. Phys. Chem. B 108 (2004)
3573–3583.