Influence of the Support on the Catalytic Characteristics of the Deposited Palladium in the Liquid-Phase Hydrogenation of Diphenylacetylene

Article abstract of DOI:10.1134/S0023158417060027

Palladium catalysts on various types of supports were studied in the liquid-phase hydrogenation of diphenylacetylene. Samples of Pd/SiO₂–Al₂O₃, Pd/MgAl₂O₄, Pd/Al₂O₃, and Pd/TiO₂ were characterized by the chemisorption of the CO and IR spectroscopy of adsorbed CO. The use of n-hexane as the solvent increases the reaction rate, which can be explained by the better solubility of hydrogen in the liquid phase. It is established that the acid–base properties of the support do not affect the activity and selectivity of the catalysts in the reaction under study. However, they alter the electronic state of palladium. According to the catalytic tests, Pd/TiO₂ has the highest activity (turnover frequency) and selectivity to alkene. The comparison of the obtained catalytic data and the results of IR spectroscopy made it possible to conclude that this is due to the electron density redistribution between the palladium and TiO_x particles, which is caused by the strong metal–support interaction.

Full text of DOI:10.1134/S0023158417060027

ISSN 0023-1584, Kinetics and Catalysis, 2017, Vol. 58, No. 6, pp. 763–770. © Pleiades Publishing, Ltd., 2017.
Original Russian Text © G.O. Bragina, N.S. Smirnova, D.S. Krivoruchenko, P.V. Markov, G.N. Baeva, A.Yu. Stakheev, 2017, published in Kinetika i Kataliz, 2017, Vol. 58, No. 6,
pp. 743–750.
Influence of the Support on the Catalytic Characteristics
of the Deposited Palladium in the Liquid-Phase Hydrogenation
of Diphenylacetylene
G. O. Bragina, N. S. Smirnova, D. S. Krivoruchenko, P. V. Markov,
G. N. Baeva, and A. Yu. Stakheev*
Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Moscow, 119991 Russia
*
e-mail: st@ioc.ac.ru
Received February 2, 2017
Abstract—Palladium catalysts on various types of supports were studied in the liquid-phase hydrogenation of
diphenylacetylene. Samples of Pd/SiO –Al O , Pd/MgAl O , Pd/Al O , and Pd/TiO were characterized by
2
2
3
2
4
2
3
2
the chemisorption of the CO and IR spectroscopy of adsorbed CO. The use of n-hexane as the solvent
increases the reaction rate, which can be explained by the better solubility of hydrogen in the liquid phase. It
is established that the acid–base properties of the support do not affect the activity and selectivity of the cat-
alysts in the reaction under study. However, they alter the electronic state of palladium. According to the cat-
alytic tests, Pd/TiO has the highest activity (turnover frequency) and selectivity to alkene. The comparison
2
of the obtained catalytic data and the results of IR spectroscopy made it possible to conclude that this is due
to the electron density redistribution between the palladium and TiO particles, which is caused by the strong
x
metal–support interaction.
Keywords: liquid–phase hydrogenation of diphenylacetylene, IR-spectroscopy of adsorbed CO, Pd/TiO ,
2
strong metal–support interaction
DOI: 10.1134/S0023158417060027
INTRODUCTION
lytic properties of metallic nanoparticles [3]. This fac-
tor is key in the development of new-generation cata-
lysts and it requires a comprehensive study. There are
relatively few published sources on the role of the sup-
port in the selective hydrogenation of alkynes and
these sources mainly consider the process of ethylene
purification from admixtures of acetylenic com-
pounds [4–9]. Moreover, as noted by the authors [6],
the results of studies on the influence of the support
are hard to compare because of the differences in cat-
alyst preparation methods and, consequently, in the
dispersion of the resulting nanoparticles.
The selective hydrogenation of triple and double
C–C bonds is one of the main reactions for the syn-
thesis of a wide range of basic and fine organic synthe-
sis products. The stereoselective hydrogenation of
alkynes to the corresponding cis-alkenes and the
hydrogenation of triple carbon–carbon bonds in the
presence of other functional groups are the key pro-
cesses for obtaining raw materials for the needs of the
pharmaceutical industry, food industry, perfumery,
cosmetics, etc. These reactions are traditionally cata-
lyzed by CaCO -supported palladium with lead addi-
3
It is known from the literature that the metal–sup-
port interaction has a strong influence on the size,
shape, and electronic properties of metallic nanopar-
ticles [10, 11]. The authors [3] consider the porous
structure and texture of the support as one of the fac-
tors that determine catalytic characteristics. As noted
in [12], an increase in the specific surface of α-Al O
tives, which is known as the Lindlar catalyst [1]. How-
ever, its use is associated with serious disadvantages
such as the low selectivity for the target product, rapid
deactivation, toxicity of lead salts, and insufficient
mechanical strength. Thus, it remains important to
find and develop highly efficient heterogeneous cata-
lytic systems that would make it possible to transform
complex functionalized molecules with ultrahigh
selectivity under mild conditions [2].
2
3
led to an increase in the dispersion of palladium and
in the overall activity of the catalyst in acetylene
hydrogenation. In the case of liquid-phase hydroge-
The main function of the support in the supported nation of substituted alkynes which is preferably car-
metal catalysts is to provide the required dispersion of ried out in a solvent, the porous structure of the sup-
the active component and to prevent its sintering. In port determines the presence or absence of diffusion
addition, the support can significantly affect the cata- difficulties [13].
763

7
64
BRAGINA et al.
2
An important role is played by the acidic properties MgAl O spinel (Sasol, S
= 40 m /g), and TiO
2
4
BET
2
of the support because they significantly affect the
electronic state of the nanoparticles of the active com-
ponent [3]. In hydrogenation of 2-butyne-1,4-diol the
effect of a tenfold increase in the activity of a palla-
dium catalyst was demonstrated upon substituting the
2
(Saint-Gobain, S
0
= 40 m /g, sulfur content <
.3 wt %). The Pd/TiO catalyst was subjected to a
2
series of redox treatments in order to remove the resid-
ual sulfur. The sulfur removal process was monitored
by the IR spectroscopy of the adsorbed CO.
BET
basic MgO with the acidic SiO [14]. This effect was
2
caused by the fact that the acid nature of the support
Samples impregnated with an aqueous solution of
contributed to the formation of electron-deficient Pd [Pd(NH ) ]Cl were dried at room temperature for
3 4 2
nanoparticles, the adsorption of the electron-donor 24 hours and then calcined in an air flow (300 mL/min)
δ+
С≡С-bond on the Pd clusters increased, and, conse- at 500°C for 2 hours with temperature ramp
quently, the reaction rate increased as well. It should 0.5°C/min. The precalcined samples were reduced in
be noted that the selectivity for the target product was a 5% H /Ar flow at 500°C for 1 hour. The reduction
2
maximal when the supports had basic properties. The temperature of the Pd/TiO catalyst was decreased to
2
authors suggest that this effect is associated with an 300°C in order to reduce the strong metal–support
increase in the difference in the adsorption strength of interaction.
the reaction products. Thus, the electronic state of
metal nanoparticles can directly affect its catalytic
characteristics.
IR Spectroscopy of Adsorbed CO
Effects of the strong metal–support interaction
IR diffuse reflection spectra of adsorbed CO were
recorded on a Tensor 27 IR spectrometer (Bruker,
Germany) equipped with in situ a Harrick Diffuse
Reflectance Kit (Harrick, United Kingdom). A sample
of a crushed catalyst (20 mg) was placed in a tempera-
(
SMSI) can also be observed when using the supports
that can be partially reduced (ZnO, TiO , CeO , etc.):
2
2
these effects lead to the partial encapsulation of the
active component by the reduced support with the for-
mation of interphase active centers and bimetallic
compounds [15]. In this case, the electronic state of
the deposited metal, its crystal structure and, conse-
quently, its catalytic characteristics get significantly
modified [3, 16]. For example, the PdZn intermetallic
formation upon the reduction of Pd/ZnO [4] results in
an increase of the selectivity of the target product (2-
butene-1,4-diol) up to 99.9%. Moreover, as shown by
the authors [4], the structural rearrangement of the
active component (Pd, PdZn) that takes place during
the hydrogenation of acetylene strongly depends on
the support nature.
ture controlled cell with CaF windows. The sample was
2
reduced with a 5% H /Ar mixture (30 mL/min) at
2
3
50°C for 1 h before the measurements. In the case of
the Pd/TiO catalyst, the reduction temperature was
2
2
00°C. The cell was then cooled to 50°C in an argon flow
and the CO was then adsorbed (0.5 vol % CO/He,
0 min), after which the spectra were recorded
2
–1
(
500 scans, resolution 4 cm ). The stoichiometric
coefficients of the CO sorption on palladium (n,
Pd/CO), which are required for the calculation of the
dispersion of the metal, were approximately estimated
from the ratio of the absorption band areas of the linear
and bridge forms in the resulting spectra [17].
Our analysis of the published data showed that the
information on the influence of the support on the
catalytic properties of palladium in liquid-phase
hydrogenation of the substituted alkynes is very
scarce. For this reason, the aim of this work was to
study the effect of different types of supports on the
catalytic characteristics of palladium in liquid-phase
Determining the Dispersion of the Deposited Palladium
by Chemisorption of CO
The dispersion of the deposited palladium by the
hydrogenation of the substituted alkynes. Dipheny- chemisorption of CO was determined on an ASAP
lacetylene was used as the model compound.
2020 Chemi device (Micromeritics, United States).
The samples were vaccumed, then reduced in a hydro-
gen stream at 400°C for 30 min (for Pd/TiO , the tem-
2
EXPERIMENTAL
perature was 250°С), cooled to 35°С in an inert gas
flow, and then vacuumed again. The value of carbon
monoxide (II) adsorbed on the metal was determined
Preparation of Catalysts
The properties of monometallic samples deposited from the CO isotherms in the pressure range of 100–
on different types of supports are studied in this work. 500 mmHg. The number of surface palladium atoms,
Catalysts containing 0.5 wt % Pd were obtained by the dispersion, and the average particle size were cal-
incipient wetness impregnation. The following mate- culated, taking into account the Pd/СО sorption coef-
rials were used as the support: Al O (Sasol, S =
BET
ficients, according to the standard procedure [18]:
2
3
2
5
6 m /g; according to X-ray fluorescence analysis, it is
a mixture of γ- and δ-modifications), 30% SiO -Al O
3
aluminosilicate (Sasol, Siral-30, SBET = 450 m /g),
VnN А
2
2
N =
,
(1)
s
2
22414
KINETICS AND CATALYSIS
Vol. 58
No. 6
2017

INFLUENCE OF THE SUPPORT ON THE CATALYTIC CHARACTERISTICS
765
where N is the number of surface Pd atoms; V is the samples at the first reaction stage was calculated by
s
3
CO volume that is required to form a monolayer, cm
formula (4), where r is the alkyne hydrogenation rate
1
(
STP); n is the Pd/CO stoichiometric adsorption coef- (the number of DPA molecules converted in one sec-
ficient; and N is the Avogadro number.
ond) and N is the number of surface Pd atoms in the
s
А
To calculate the dispersion ((D), %), we used the weighted sample:
formula
r
N
S
1
TOFAC
=
.
(4)
VnM
mw × 224.14
D =
,
(2)
To determine the selectivities of the target olefin
where m is the sample weight, g; w is the mass fraction
of the metal, %; and M is the atomic weight of the
metal (Pd), g/mol.
The values of the dispersion and the average parti-
cle size (d ) were related by
(
S) and cis-stilbene (S ) formation, the reaction
CIS
mixture was analyzed by gas chromatography on a Crys-
tal 5000 instrument (Chromatec, Russia) equipped with
an HP5-MS column (5% phenyl dimethylsiloxane,
30 m, the internal diameter of 0.25 mm, and the car-
rier gas is helium) and a flame ionization detector. The
(3) total selectivity to alkene was calculated by formula
av
(
VPd aPd)
dav = 6 ×
,
D
(
5), where n and n are the mole fractions of the pro-
= –
3
where V is the volume, nm , occupied by an atom in duced olefin and alkane, respectively:
Pd
2
the metal and a is the area, nm , occupied by a sur-
Pd
n=
face Pd atom.
S =
.
(5)
n= + n−
The selectivity of cis-stilbene formation was calcu-
lated by formula (6), where n and n are
mole fractions of the cis and trans isomers of stilbene
diphenylethylene (DPE)):
Hydrogenation of Diphenylacetylene
cis-DPE
trans-DPE
The liquid phase hydrogenation of diphenylacety-
lene (DPA) (99%, Aldrich) was carried out in an auto-
clave reactor mounted on a magnetic stirrer. The reac-
tor was connected to a gas dosing system and an elec-
(
ncis-DPE
SCIS
=
.
(6)
tronic pressure sensor for the estimation of the H
2
ncis-DPE + ntrans-DPE
absorption. N-hexane (98%, Merck) and methanol
When comparing the selectivities for different cata-
lysts, we used the S and S values which were
(
99.8%, Merck) were chosen as the solvents. The reac-
tion temperature was 25°C and the initial hydrogen
80
cis80
pressure was 5 bar. The weight of the catalyst sample determined by the interpolation of the neighboring
and the intensity of mixing were selected individually points to the DPA conversion of 80%.
for each of the experiments to ensure that the process
proceeds in the kinetic region [19].
RESULTS AND DISCUSSION
IR Spectroscopy of Adsorbed CO
−
1
–
1
The reaction rate ((r), mol
g
min ) was calcu-
cat
H2
lated from the rate of hydrogen adsorption using the
dependence of the amount of absorbed hydrogen on the
reaction time. Since the hydrogenation of compounds
with a triple bond occurs in two stages, the reaction
rates were determined for each of them as follows: for
Infrared spectroscopy of adsorbed CO is a conve-
nient method for estimating the electron state of
deposited palladium nanoparticles, which provides
valuable information on the nature of the electronic
interaction between the metal clusters and
acidic/basic groups of the oxide support [20]. There
are two broad asymmetric absorption bands in the
diffuse reflection spectra of adsorbed CO (Fig. 1).
the first stage (r ), they were determined in the adsorp-
1
tion range of 0.1–0.6 eq. H ; and for the second stage
2
(
r ), in the adsorption range of 1.3–1.8 eq. H .
2
2
Chromatographic analysis showed that, at the first
–1
The band at ~2080 cm corresponds to the stretch-
stage, the amount of the product of the complete
hydrogenation is negligible; thus, the reaction of com-
plete hydrogenation can be ignored. Similarly, at the
second stage, the adsorption of hydrogen in the hydro-
genation of the residual DPA is also negligible.
ing modes of the adsorbed CO molecules that are lin-
early bound to the palladium atoms, whereas the
–
1
appearance of absorption bands at ~1850–1950 cm
indicates the adsorption of CO on two and three Pd
atoms [21].
The effectiveness of the kinetic control over the
course of the process was estimated from the ratio of
the hydrogenation rates of the initial alkyne at the first
and the second stages, r /r .
The presence of Bronsted and Lewis acid-base
centers on the support surface is known to have a sig-
nificant effect on the electronic properties of the
deposited metal [22]. The position of the absorption
peak of the linearly bound CO makes it possible to
1
2
–
1
The turnover frequency ((TOF ), s ) was used as
AC
a fundamental characteristic of the active center of the qualitatively evaluate the electron state of the palla-
catalyst (a surface palladium atom). TOF of the dium nanoparticles [20]. The shift of absorption band
AC
KINETICS AND CATALYSIS Vol. 58 No. 6 2017

7
66
BRAGINA et al.
2
.025
090
increase in the electron density on the particles of the
deposited palladium.
1953
0
It is known that the strong metal–support interac-
1
2
085
tion (SMSI) in the Pd/TiO catalyst leads to the
1
958
2
encapsulation of palladium by the TiO particles [23]
x
1959
and causes a redistribution of the electron density [24]
2073
2
3
to the active component. For the Pd/TiO sample, the
2
–1
absorption band of linearly bound CO is at 2091 cm ,
which agrees with FTIR-CO data [25] for the system
in which the SMSI effect is observed.
2091
The ratios of the peak areas of the bridge and linear
sorption are similar for the Pd/SiO –Al O ,
1
947
2
2
3
4
Pd/MgAl O , and Pd/Al O catalysts; these values are
2
4
2
3
in the range of 2.1–2.4 (Table 1). A significant differ-
2
150 2100 2050 2000 1950 1900 1850 1800
ence in the spectrum of the Pd/TiO sample lies in the
2
ν, cm–1
marked broadening and substantial increase in the
intensity of the adsorbtion band that corresponds to
the linear form of adsorbed CO. At the same time, the
intensity of the band that is typical of the bridged CO
adsorption decreases. This is presumably due to the
Fig. 1. IR diffuse reflection spectra of adsorbed CO for fol-
lowing catalysts: (1) Pd/SiO –Al O , (2) Pd/Al O , (3)
2
2
3
2 3
Pd/MgAl O , and (4) Pd/TiO .
2
4
2
decoration of the metal surface with TiO fragments
x
of the partially reduced support [26], which, in turn,
leads to a decrease in the number of centers that sup-
port a multipoint CO adsorption.
toward high wavenumbers indicates the electron-defi-
cient nature of the deposited palladium clusters. This
nature is due to the interaction of the metal particles
with the acid centers on the support surface. In con-
trast, the shifts of absorption bands toward the region
of lower ν values are usually due to an increase in the
donation of the electron density from a Pd atom to the
π-antibonding orbital of a CO molecule. These shifts
indicate the interaction of the metal particles with the
electron-donor (base) centers of the support.
The observed differences in the Pd/TiO spectrum
2
indirectly indicate a change in the stoichiometry of the
CO sorption (n) of the palladium particles. Since the
exact determination of the particle size of the catalysts
by instrumental methods (HRTEM or XRD) is asso-
ciated with large errors because of the low active com-
ponent content (0.5% by weight), a rough estimate of
the stoichiometric sorption coefficient was obtained
from the IR spectra for all the catalysts (Table 1). The
obtained values of n are correlated with the published
In the case of the Pd/SiO –Al O catalyst, the
2
2
3
maximum of the linearly adsorbed CO band was at
–1
2
090 cm , which is due to the strong acidity of the alumi- data, according to which n for oxide supports (SiO
2
nosilicate. For the Pd/Al O , the maximum of the linearly
2
3
and Al O ) is close to two [27]. For Pd/TiO , the CO
2
3
2
–1
adsorbed CO band is somewhat lower (~2085 cm ). This sorption stoichiometry strongly depends on the pre-
is consistent with the fact that the acid sites on the sur- liminary treatment of the support: this value is close to
face of aluminum oxide are weaker than those on unity at high catalyst reduction temperatures [25].
SiO –Al O . The significant absorption band shift
2
2
3
–1
toward the region of small wavenumbers (2073 cm )
Dispersion of Catalysts
in the spectra of the Pd/MgAl O catalyst is due to the
2
4
fact that the centers with electron-donor properties are
The dispersion of palladium (D) and its average
prevalent on the support surface. As mentioned above, particle size (dav) were calculated from the volumes of
the basic nature of the support contributes to an chemisorbed CO, while taking into account the stoi-
Table 1. Data of IR spectroscopy of adsorbed CO. Dispersion and average size of Pd particles
Ratio of peak areas
for bridge and linear
sorption, Br/Lin
n (Pd/CO, according
to IR spectroscopy)
Sample
Dispersion D, %
dav, nm
Pd/SiO –Al O
2.3
2.4
2.1
0.3
1.7
1.7
1.7
1.2
54.6
36.0
33.9
3.0
2.1
3.1
2
2
3
Pd/Al O3
2
Pd/MgAl O4
3.3
2
Pd/TiO2
36.8
KINETICS AND CATALYSIS
Vol. 58
No. 6
2017

INFLUENCE OF THE SUPPORT ON THE CATALYTIC CHARACTERISTICS
767
chiometric Pd/CO sorption coefficients (Table 1).
(а)
According to the obtained data, the Pd/SiO –Al O
3
–1
2
2
A , mol. H (mol DPA)
H
2
2
catalyst has the highest dispersion, which is probably
2
.0
due to the high specific surface area of its support
2
1.8
.6
.4
.2
.0
.8
.6
0.4
0.2
(
450 m /g). The dispersions of the Pd/Al O and
2
3
1
1
1
1
0
0
Pd/MgAl O catalysts are similar (36.0 and 33.9%,
respectively). The Pd/TiO sample is characterized by
2
4
2
Pd/Siral
the lowest dispersion (3.0%) and the largest size of Pd
particles (~37 nm). In this case, the low value of D can
be associated with the decoration of the metal with a
partially reduced support, which makes part of the
active component inaccessible to the adsorbed CO
molecules [28].
2
Pd/Al O3
2
Pd/MgAl O4
Pd/TiO2
0
20 40 60 80 100 120 140 160 180 200
Catalytic Characteristics of Palladium on Various
Supports in Hydrogenation of Diphenylacetylene
τ, min
(b)
Figures 2a and 2b show the kinetic curves of the
hydrogen adsorption for the Pd/SiO –Al O ,
A , mol. H2 (mol DPA)–1
2
2
3
H
2
Pd/MgAl O , Pd/Al O , and Pd/TiO catalysts. To
2.0
2
4
2
3
2
study the effect of the solvent’s polarity, the hydroge-
nation of diphenylacetylene was carried out in a polar
1
.8
.5
.3
.0
0.8
1
(
methanol) and in nonpolar (n-hexane) liquid phases.
Table 2 shows the calculated rates of the hydrogena-
1
tion of the triple and double bonds and the r /r ratios.
1
2
1
The TOF values were calculated for the catalysts in
AC
Pd/Siral
each series based on the obtained catalytic and
chemisorption data. The selectivities of the samples
were compared using the values of S at the 80% diphe-
nylacetylene conversion.
2
Pd/Al O3
0
.5
.3
2
Pd/MgAl O4
0
Pd/TiO2
Activity of the catalysts. In all cases, there was a
bend on the curves, which is caused by a decrease in
the reaction rate at the stage of olefin hydrogenation.
0
10 20 30 40 50 60 70 80 90 100
τ, min
It follows from Fig. 2 that Pd/SiO –Al O is the most
2
2
3
Fig. 2. Dependence of equivalent amount of absorbed
hydrogen on time of DPA hydrogenation in different sol-
vents: (a) methanol and (b) n-hexane. Reaction condi-
active catalyst in the hydrogenation of diphenylacety-
lene. According to the experimental data, the com-
plete hydrogenation of DPA on the Pd/SiO –Al O
tions: m = 2.5 mg, P = 5 bar, Т = 25°С, and [С
] =
cat
DPA
2
2
3
0
.160 mol/L.
sample occurs in 40–100 min (when using n-hexane
and methanol, respectively, as the solvent). The rates of
DPA hydrogenation on Pd/Al O and Pd/MgAl O are
2
3
2
4
Pd/TiO sample had the lowest activity. Since the
above-mentioned reaction is slowest on Pd/TiO , the
2
somewhat lower (Fig. 2): their values are within the
2
−
cat
1
–3
‒3
–1
range of 4.6 × 10 –17.8 × 10 molH2
g
min . The determination of the rate of the second stage is associ-
Table 2. Kinetic characteristics of hydrogenation of diphenylacetylene in different catalytic systems
Methanol n-Hexane
3
3
3
3
r × 10 ,
r × 10 ,
r × 10 ,
r × 10 ,
2
1
2
1
Sample
−
1
−1
cat
−1
cat
−1
cat
m, mg
r1/r2 TOFAC m, mg
r1/r2 TOFAC
molH2
g
molH2
g
molH2
g
molH2
g
cat
min–1
min–1
min–1
min–1
Pd/SiO –Al O
2.5
2.5
2.5
7.5
9.68
4.67
5.23
1.03
2.74
0.92
0.44
0.003
3.5
5.1
6.3
4.6
2.5
2.5
2.5
7.5
22.7
17.9
10.6
1.51
7.73
2.9
4.4
13.3
15.1
11.1
22.3
2
2
3
Pd/Al O3
4.09
2.04
0.016
2
Pd/MgAl O4
11.9
5.5
5.2
2
Pd/TiO2
333
12.1
92.1
KINETICS AND CATALYSIS Vol. 58 No. 6 2017

7
68
BRAGINA et al.
(
а)
The turnover frequency (TOF ) at the first reac-
AC
tion stage was calculated for all the samples. The
S , %, Scis80, %
80
TOF values obtained for the Pd/Al O , Pd/SiO –
S , %
Scis80, %
AC
2
3
2
8
0
100
Al O , and Pd/MgAl O catalysts in both series
2
3
2
4
95.6
–1
(
1
Table 2) are similar (4.6–6.3 s for methanol and
9
4.3
94.5
94.5
9
3.9
–1
9
5
0
5
0
1.1–15.1 s for n-hexane) within the experimental
error. This suggests that the nature of the active center
undergoes no particular change and that the type of
the support does not have a pronounced effect on the
8
9.2
9
8
8
86.8
catalytic activity. In both cases, however, the Pd/TiO
8
4.0
2
catalyst was characterized by the highest TOF values
AC
–1
(
12.1 and 22.3 s for the series of experiments with
methanol and n-hexane, respectively), which differ sig-
nificantly from the turnover frequency values for
Pd/Al O , Pd/SiO –Al O , and Pd/MgAl O . Our
7
5
2
3
2
2
3
2
4
results are in good agreement with the data of [28],
which studied the influence of the strong metal–sup-
port interaction on the catalytic characteristics in the
hydrogenation of phenylacetylene (PA)—a terminal
alkyne that is structurally similar to diphenylacety-
7
0
Pd/SiO –Al O
Pd/MgAl O
2 4
2
2
3
Pd/Al O
Pd/TiO2
2
3
lene. The TOF values calculated in the present study
AC
(b)
are comparable to those for the samples in the hydro-
genation of phenylacetylene on different supports. It
should be noted that [28] revealed no clear pattern of
the influence of the support acidity on the catalytic
activity. However, the authors [28] noted a sharp
S , %, Scis80, %
80
S , %
Scis80, %
100
80
95.4
95.1
9
4.0
9
5
0
5
0
93.3
increase (~tenfold) in the TOF for Pd/TiO in the PA
92.6
2
91.9
9
0.4
hydrogenation reaction, which is due to the SMSI
effect. Based on the above, it can be assumed that the
increase in TOF for the Pd/TiO system in the DPA
9
8
8
88.0
AC
2
hydrogenation is also associated with the strong
metal–support interaction.
The high ratio of the r /r rates provides an efficient
1
2
control of the hydrogenation process by stopping the
reaction after the stage alkyne to alkene transforma-
tion with the minimal losses in the form of the rehy-
drogenation product. The r /r value for the Pd/SiO –
7
5
1
2
2
70
Al O catalyst is the lowest (2.9–3.5), whereas the cor-
Pd/SiO –Al O
Pd/MgAl O
4
2
3
2
2
3
2
responding values calculated for the samples of
Pd/MgAl O and Pd/Al O are two to four times
Pd/Al O
Pd/TiO2
2
3
2
4
2
3
higher. The r /r ratios are maximal for the Pd/TiO
2
Fig. 3. Total selectivity for alkene (S , %) and selectivity for
1
2
80
cis-stilbene (S
, %) at 80% conversion of DPA for palla-
system (92.1 for the reaction in n-hexane and 333 for
the reaction in methanol). They exceed the corre-
sponding values for the other catalysts by more than an
order of magnitude (Table 2).
cis80
dium deposited on different supports. Solvents: a—metha-
nol and b—n-hexane.
ated with large errors. For this reason, the weight of
the catalyst sample was increased from 2.5 to 7.5 mg in
Selectivity for stilbene. In addition to evaluating the
activity and the reaction rates ratio, we also deter-
mined the selectivity for the target alkene (the total
order to calculate r and r (Table 2). It should be noted
1
2
that the complete hydrogenation requires a longer period content of cis- and trans-stilbene) for both series of
catalysts. According to Figs. 3a and 3b, the selectivities
at an 80% DPA conversion are comparable for the
series of catalysts with different solvents. The most
selective catalyst was Pd/TiO (S = 95.4–95.6%).
of time when methanol is used as the solvent. An increase
in the reaction rates that was observed for the experi-
ments with n-hexane is due to the better solubility of
hydrogen in this solvent (according to [29], at 25°C and
2
80
The lowest selectivity values (84.0 and 88.0% for
methanol and n-hexane, respectively) were observed
for the Pd/SiO –Al O , whereas the corresponding
3
3
1
0 atm, the solubility of H in n-C H is 1.85 cm /cm
2 6 14
3
3
and its solubility in CH OH is 0.9 cm /cm ).
3
2
2
3
KINETICS AND CATALYSIS
Vol. 58
No. 6
2017

INFLUENCE OF THE SUPPORT ON THE CATALYTIC CHARACTERISTICS
769
values for the Pd/Al O and Pd/MgAl O samples are
2. Ananikov, V.P., Khemchyan, L.L., Ivanova, Yu.V.,
Bukhtiyarov, V.I., Sorokin, A.M., Prosvirin, I.P.,
Vatsadze, S.Z., Medved’ko, A.V., Nuriev, V.N.,
Dil’man, A.D., Levin, V.V., Koptyug, I.V., Kovtu-
nov, K.V., Zhivonitko, V.V., Likholobov, V.A., Roma-
nenko, A.V., Simonov, P.A., Nenaidenko, V.G.,
Shmatova, O.I., Muzalevskii, V.M., Nechaev, M.S.,
Asachenko, A.F., Morozov, O.S., Dzhevakov, P.B.,
Osipov, S.N., Vorob’eva, D.V., Topchii, M.A.,
Zotova, M.A., Ponomarenko, S.A., Borshchev, O.V.,
Luponosov, Yu.N., Rempel’, A.A., Valeeva, A.A.,
Stakheev, A.Yu., Turova, O.V., Mashkovsky, I.S., Sys-
olyatin, S.V., Malykhin, V.V., Bukhtiyarova, G.A., Ter-
ent’ev, A.O., and Krylov, I.B., Russ. Chem. Rev., 2014,
vol. 83, p. 885.
2
3
2
4
similar (86.8–91.9%).
The results of evaluating the cis-stilbene content
showed that its value is rather high in the reaction mix-
ture (~93–95% at 80% DPA conversion) relative to
the total amount of alkene. Moreover, the S
values
cis80
are approximately the same for all the catalysts and the
differences are within the experimental error.
Since the reaction of diphenylacetylene hydroge-
nation is structurally sensitive, the selectivity of the
process is associated with the dispersion of the cata-
lyst. Previously, it was shown [30], that an increase in
the average size of the palladium particles from 1.5 to
2
2.0 nm makes it possible to increase the selectivity of
3. Crespo-Quesada, M., Cardenas-Lizana, F., Dessi-
moz, A.-L., and Kiwi-Minsker, L., ACS Catal, 2012,
vol. 2, p. 1773.
stilbene formation from 71 to 87%. It can be assumed
that the low values of S for the Pd/SiO –Al O cata-
80
2
2
3
4
5
6
. Crespo-Quesada, M., Yoon, S., Jin, M., Xia, Y.,
lysts (in comparison with those for Pd/MgAl O and
2
4
Weidenkaff, A., and Kiwi-Minsker, L., ChemCatChem,
Pd/Al O ) are associated with the presence of small
2
3
2014, vol. 6, p. 767.
palladium clusters. However, the high values of selec-
. Benavidez, A.D., Burton, P.D., Nogales, J.L., Jen-
kins, A.R., Ivanov, S.A., Miller, J.T., Karim, A.M.,
and Datye, A.K., Appl. Catal. A, 2014, vol. 482, p. 108.
. Ruta, M., Semagina, N., and Kiwi-Minsker, L.,
J. Phys. Chem., vol. 112, p. 13635.
tivity observed for Pd/TiO are caused not only by the
2
dispersion but also by the effects of the metal–support
interactions. The authors [28] attribute an increase of
the selectivity to styrene in the hydrogenation of phe-
nylacetylene on Pd/TiO to the charge transfer from
2
7
. Chesnokov, V.V., Podyacheva, O.Yu., and Richards, R.M.,
the TiO particles to the palladium clusters, which
x
Mater. Res. Bull., 2017, vol. 88, p. 78.
facilitates the desorption of styrene and prevents its
further hydrogenation. A similar explanation that con-
siders the increase of the electron density on palla-
dium was proposed to explain the increase in the yield
of ethylene in the gas-phase hydrogenation of acety-
8
. Komhom, S., Mekasuwandumrong, O., Panpranot, J.,
and Praserthdam, P., Ind. Eng. Chem. Res., 2009,
vol. 48, p. 6273.
9
. Komeili, S., Ravanchi, M.T., and Taeb, A., Appl. Catal.
A. Gen, 2015, vol. 502, p. 287.
lene on Pd/TiO [31].
2
1
0. Bukhtiyarov, V.I. and Slin’ko, M.G., Russ. Chem. Rev.,
001, vol. 70, p. 167.
1. Efremenko, I., J. Mol. Catal. A, 2001, vol. 173, p. 19.
2
1
CONCLUSIONS
1
2. Komhom, S., Praserthdam, P., Mekasuwandumrong, O.,
Although the acid–base properties of the support
influence the electronic state of palladium particles,
their effect on the activity and selectivity is relatively
small [32]. However, we established that the SMSI can
have a significant effect on the catalytic properties.
The obtained results make it possible to conclude that
and Panpranot, J., React. Kinet. Catal. Lett., 2008, vol. 94,
p. 233.
1
3. Marin-Astorga, N., Pecchi, G., Fierro, J.L.G., and
Reyes, P., Catal. Lett., 2003, vol. 91, p. 115.
1
4. Berguerand, C., Yuranov, I., Cárdenas-Lizana, F.,
Yuranova, T., and Kiwi-Minsker, L., J. Phys. Chem.,
2014, vol. 118, p. 12250.
Pd/TiO is a promising system for obtaining highly
2
selective catalysts for the liquid-phase hydrogenation
of alkyne compounds.
1
5. Unterberger, W., Jenewein, B., Klotzer, B., Penner, S.,
Reichl, W., Rupprechter, G., Wang, D., Schlogl, R.,
and Hayek, K., Catal. Lett., 2006, vol. 87, p. 215.
1
6. McCue, A.J. and Anderson, J.A., Front. Chem. Sci.
ACKNOWLEDGMENTS
Eng, 2015, vol. 9, p. 142.
1
7. Gubitosa, G., Berton, A., Camia, M., and Pernicone, N.,
Preparation of catalysts III. Proceedings of Third Interna-
tional Symposium, Amsterdam: Elsevier Science Publish-
ers BoV, 1983.
We thank Dr. G.I. Kapustin for determining the
dispersion of catalysts by the chemisorption of CO.
This work was supported by the Russian Science
Foundation, project no. 16-13-10530.
1
8. Bergeret, G. and Gallezot, P., Handbook of Heteroge-
neous Catalysis, Eds Ertl G., Knozinger H., Schuth F.,
Weitkamp J. Weinheim: Wiley-VCH Verlag GmbH&Co.
KGaA, 2008, vol. 8, p. 3298.
REFERENCES
1
. Chen, B., Dingerdissen, U., Krauter, J.G.E., Rotger- 19. Markov, P.V., Bragina, G.O., Baeva, G.N.,
ink, H.G.J.L., Mobus, K., Ostgard, D.J., Panster, P.,
Riermeier, T.H., Seebald, S., Tacke, T., and Trauth-
wein, H., Appl. Catal. A: Gen., 2005, vol. 280, p. 17.
Tkachenko, O.P., Mashkovskii, I.S., Yakushev, I.A.,
Kozitsyna, N.Yu., Vargaftik, M.N., and Stakheev, A.Yu.,
Kinet. Catal., 2015, vol. 56, p. 591.
KINETICS AND CATALYSIS Vol. 58 No. 6 2017

7
2
70
BRAGINA et al.
0. Huang, J., Jiang, Y., Vegten, N., Hunger, M., and 28. Weerachawanasak, P., Mekasuwandumrong, O.,
Baiker, A., J. Catal., 2011, vol. 281, p. 352.
Arai, M., Fujita, S.-I., Praserthdam, P., and Pan-
pranot, J., J. Catal., 2009, vol. 262, p. 199.
21. Cabilla, G.C., Bonivardi, A.L., and Baltanas, M.A.,
Catal. Lett., 1998, vol. 55, p. 147.
2
9. Kogan, V.B., Fridman, V.M., and Kafarov, V.V., Sprav-
ochnik po rastvorimosti. Binarnye sistemy. Tom 1. Kniga
1, Kafarov, V.V., Eds., Moscow: Izdatel’stvo AN SSSR,
1961.
2
2. Solymosi, F., Cat. Rev.-Sci. Eng., 1968, vol. 1, p. 233.
23. Suzuki, T. and Souda, R., Surf. Sci., 2000, vol. 448,
p. 33.
2
4. Weerachawanasak, P., Praserthdam, P., and Pan- 30. Markov, P.V., Turova, O.V., Mashkovsky, I.S., Khu-
pranot, J., J. Nanosci. Nanotechnol., 2014, vol. 14,
dorozhkov, A.K., Bukhtiyarov, V.I., and Stakheev,
p. 3170.
A.Yu., Mendeleev Commun., 2015, vol. 25, p. 367.
25. Zhu, B. and Jang, B.W.-L., J. Mol. Catal. A, 2014,
3
1. Panpranot, J., Kontapakdee, K., and Praserthdam, P.,
vol. 395, p. 137.
J. Phys. Chem. B, 2006, vol. 110, p. 8019.
2
6. Sa, J., Arteaga, G.D., Daley, R.A., Bernardi, J., and
Anderson, J.A., J. Phys. Chem. B, 2006, vol. 110,
p. 17090.
3
2. Dominguez-Dominguez, S., Berenguer-Murcia, A.,
Linares-Solano, A., and Cazorla-Amoros, D.,
J. Catal., 2008, vol. 257, p. 87.
27. Canton, P., Menegazzo, F., Polizzi, S., Pinna, F., Per-
nicone, N., Riello, P., and Fagherazzi, G., Catal. Lett.,
2
003, vol. 88, p. 141.
Translated by Yu. Modestova
KINETICS AND CATALYSIS
Vol. 58
No. 6
2017