Cocatalytic effect of palladium and zinc in the condensation of alcohol carbon backbones into hydrocarbons

Article abstract of DOI:10.1134/S0023158411020030

The results of the direct conversion of ethanol and its mixture with glycerol into a C₄-C₁₀₊ alkane and olefin fraction in the presence of Pd-, Zn-, and Pd-Zn-containing catalysts, which were prepared by supporting homo- and heterometallic acetate complexes onto the surface of γ-Al₂O₃, are reported. It was found that, in the presence of mono- and bicomponent Pd-ZnO (Pd; ZnO)/γ-Al₂O ₃ systems, selectivity in the formation of alkanes, olefins, or their mixtures in the target fraction can be controlled as a result of the cocatalytic effect of active components that are responsible for the catalyst activity in condensation and hydrogenation reactions. The structures of the active components were studied and the genesis of the catalytic systems was characterized using XAFS, XPS, and XRD analysis. It was found that the addition of glycerol considerably increased the yield of the target hydrocarbon fraction.

Full text of DOI:10.1134/S0023158411020030

ISSN 0023ꢀ1584, Kinetics and Catalysis, 2011, Vol. 52, No. 2, pp. 258–272. © Pleiades Publishing, Ltd., 2011.
Original Russian Text © A.V. Chistyakov, M.V. Tsodikov, V.Yu. Murzin, F.A. Yandieva, Ya.V. Zubavichus, N.Yu. Kozitsyna, A.E. Gekhman, V.V. Kriventsov, I.I. Moiseev, 2011,
published in Kinetika i Kataliz, 2011, Vol. 52, No. 2, pp. 266–281.
Cocatalytic Effect of Palladium and Zinc in the Condensation
of Alcohol Carbon Backbones into Hydrocarbons
A. V. Chistyakova, M. V. Tsodikova, V. Yu. Murzina, F. A. Yandievaa, Ya. V. Zubavichusb,
N. Yu. Kozitsynac, A. E. Gekhmanc, V. V. Kriventsovd, and I. I. Moiseeva
aTopchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, Moscow, 119991 Russia
bKurchatov Institute Russian Research Center, Moscow, 123182 Russia
cKurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences, Moscow, 119991 Russia
dBoreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia
eꢀmail: chistyakov@ips.ac.ru
Received July 2, 2010
Abstract—The results of the direct conversion of ethanol and its mixture with glycerol into a C4 – C10+ alkane
and olefin fraction in the presence of Pdꢀ, Znꢀ, and Pd–Znꢀcontaining catalysts, which were prepared by
supporting homoꢀ and heterometallic acetate complexes onto the surface of
found that, in the presence of monoꢀ and bicomponent Pd–ZnO (Pd; ZnO)/
γ
ꢀAl2O3, are reported. It was
γꢀAl2O3 systems, selectivity in
the formation of alkanes, olefins, or their mixtures in the target fraction can be controlled as a result of the
cocatalytic effect of active components that are responsible for the catalyst activity in condensation and
hydrogenation reactions. The structures of the active components were studied and the genesis of the catalytic
systems was characterized using XAFS, XPS, and XRD analysis. It was found that the addition of glycerol
considerably increased the yield of the target hydrocarbon fraction.
DOI: 10.1134/S0023158411020030
INTRODUCTION
nature of mutual effects of active components with
each other and the support surface has been poorly
studied. The interaction mechanism mainly depends
on the distance between active components distribꢀ
uted over the support surface and on the charge state of
these components, which depends on the thermal preꢀ
treatment procedure.
Alcohol conversion with the predominant formaꢀ
tion of alkanes containing at least a doubled number of
carbon atoms in their skeletons, as compared with that
in the parent alcohol, is referred to as a reductive
dehydration reaction [1]. This reaction has been studꢀ
ied previously on catalytic systems having a proꢀ
nounced hydrogenating function, such as polymetallic
intermetallides, molten iron, and a commercial aluꢀ
mina–platinum catalyst (APꢀ64) [1–3]:
In this work, catalysts based on the acetate comꢀ
plexes of Pd, which were responsible for the occurꢀ
rence of reaction path (I), and a typical Zn/Al2O3 acid
catalyst for oxygenate condensation reactions, which
was responsible for reaction paths (II) and (III), were
chosen as test materials.
m
C2H5OH + 2[H]
CnH2
+
+ 2
mH2O.
(I)
n
Note that the conversion of alcohols in the presꢀ
ence of nanosized metal oxide catalysts occurs with
the formation of olefins in two alternative directions:
the condensation and crossꢀcondensation of the carꢀ
bon skeletons of various alcohols [4–6] in accordance
with the general reaction scheme
The use of heterometallic acetate complexes as
precursors, in which palladium and zinc atoms
approach each other within a single structure, is the
most interesting problem. These complexes result in
the formation of heterometallic active centers on the
2RНOH
R1HOH + R2HOH
[R=R] + 2H2O,
(II)
surface of ꢀAl2O3, as described previously [9, 10], to
γ
[R1 = R2] + 2H2O. (III)
affect activity and selectivity in the reaction of ethanol
conversion into aliphatic hydrocarbons (the reductive
dehydration of ethanol).
These results suggest multivariant reaction paths in
the conversion of alcohols depending on the nature
and structure of active centers.
The aim of this work was to study the reaction of
To control selectivity and activity, heterogeneous ethanol conversion into hydrocarbons on these cataꢀ
catalysts containing bimetallic active components lysts and to characterize the structure and catalytic
supported onto a substrate, which are prepared based activity of systems prepared from monoꢀ and bimetalꢀ
on heterometallic metal complex compounds, have lic acetate complexes. In addition, it was important to
been widely used in recent years [7–10]. However, the study the effect of the pretreatment procedure and the
258

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
amounts of active components supported onto the surꢀ
259
face of
γ
ꢀAl2O3 on genesis and catalytic properties.
C(2)
C(6)
C(1)
O(5)
O(2)
C(5)
EXPERIMENTAL
O(1)
O(6)
Commercial ꢀAl2O3 with the following characterꢀ
γ
istics was used as the support: spherical granules with a
sphere size of 2.0–3.0 mm; specific surface area, no
less than 190 m2/g (OST [Branch Standard] 153ꢀ39.2ꢀ
022ꢀ2002); pore volume, 0.65 cm3/g (OST 153ꢀ39.2ꢀ
023ꢀ2002); iron impurity content, no more than
0.14 wt %; sodium impurity content, no more than
0.1 wt %; sorption capacity for insulating gas, no more
than 10 g/kg; and sorption capacity for hydrogen
(deuterium) fluoride, no less than 7 g/kg.
Pd(1)
Zn(1)
O(9)
O(3)
C(4)
O(4)
C(3)
O(7)
O(8)
C(7)
C(8)
The acetate complexes Zn(OAc)2
Pd3(OAc)6, and Pd( ꢀOOCMe)4Zn(OH2) were used
⋅
2H2O,
Fig. 1. Structure of the binuclear acetate complex of pallaꢀ
dium with zinc Pd( ꢀOOCMe) Zn(OH )
μ
µ
.
2
4
to prepare monoꢀ and bimetallic catalytic systems
[11]. The palladium atom in a bimetallic complex is
strongly bound to the zinc atom by four acetate
The metalꢀcontaining catalytic systems thus preꢀ
bridges, and the metal–metal distance is 2.57 Å, which pared were thermally treated under various conditions
is close to the sum of the van der Waals radii of pallaꢀ before catalytic tests. The concentrations of metals in
dium and zinc atoms (2.53 Å). Figure 1 shows the the synthesized catalyst samples were determined by
structure of the binuclear acetate complex of pallaꢀ inductively coupled plasma atomic emission specꢀ
dium with zinc.
trometry (ICP AES). Table 1 summarizes the used
metal complex precursors and pretreatment proceꢀ
dures.
Ethanol of analytical grade was used without preꢀ
liminary purification. The catalytic experiments were
performed in a laboratory flowꢀcirculation system
with a fixed catalyst bed in Ar under previously found
optimum conditions for ethanol conversion: temperaꢀ
The complexes were supported onto ꢀAl2O3 from
solutions in methanol by impregnation in accordance
with the following procedure: aluminum oxide granꢀ
γ
ules calcined at 350°С were placed in a hermetically
sealed vessel and thoroughly purged with argon. In
flowing argon, a calculated amount of an acetate
complex solution was measured with consideration for
incipient wetness impregnation and added to the aluꢀ
minum oxide; the granules were stirred at regular
intervals. After 2 h, the granules were dried in air for
ture, 350°С; pressure, 50 atm; ethanol supply rate,
0.6 h–1; and gas circulation rate, 50 cm3/min [5, 6].
Ethanol was supplied to an evaporator using a highꢀ
15 h and then in a vacuum oven at 80°С for 5 h; thereꢀ precision plunger feeder (HPP 5001), and vapor from
after, they were calcined in a muffle furnace (Ar) at the evaporator arrived at a reactor. The reaction prodꢀ
500°С for 5 h.
ucts arrived at a cooled gas–liquid separator, after
Table 1. Catalytic systems and preparation procedures
Pretreatment procedure
Active components
(concentration, wt %)
Starting complex
for supporting onto γꢀAl2O3
No.
stage 1
stage 2
I
Pd (1.07)
Pd3(OAc)6
Calcination in flowing argon;
10 h at 450°C in flowꢀ
ing hydrogen; 10 l/h
Т
= 500
°
C, 5 h
II
Zn (1.03)
Zn(OAc)2
⋅
2H2O
″
″
″
″
″
III
IIIа
IV
PdꢀZn (0.72 : 0.43)
PdꢀZn (2.40 : 1.43)
Pd+Zn (0.70 : 0.41)
Pd(
μꢀOOCMe)4Zn(OH2)
″
″
″
*
″
Pd3(OAc)6; Zn(OAc)2
⋅ 2H2O;
stepꢀbyꢀstep supporting
V
Pd/Zn (0.72 : 0.43)
Pd(μꢀOOCMe)4Zn(OH2)
Reduction in flowing hydrogen;
5 K/min to 450°C; 8 h at 450°C
in flowing hydrogen
″
* Catalyst IIIa with a triple concentration of active components was prepared in accordance with a procedure identical to that for catalyst III.
KINETICS AND CATALYSIS Vol. 52 No. 2 2011

260
CHISTYAKOV et al.
which the condensed liquid fraction was collected in a into account using the C 1s spectrum in accordance
receiver. After the separator, a mixture of argon with with a traditional procedure. The C–CH state was
uncondensed reaction products was returned to the recognized in the spectrum, and an energy of 285.0 eV
reaction volume of the system using a circulation was ascribed to it. The calibration of the spectra of
pump.
The gaseous reaction products were analyzed by
onꢀline gas chromatography. The С1–С5 hydrocarbon
other samples was performed using the Al 2p peak in
the spectrum of the support. The spectra were meaꢀ
sured at room temperature.
gases were determined on a Kristallꢀ4000 chromatoꢀ
The charge states and local structures of metalꢀ
containing active components were studied by XAFS
spectroscopy at the Structural Materials Science staꢀ
tion of the Kurchatov Center of Synchrotron Radiaꢀ
graph (FID; He, 70 cm3/min; 120°С;
P
= 0.65 MPa;
HPꢀPLOT/Al2O3 column, 50 m
×
0.32 mm).
The analysis for CO, СО2, and Н2 was performed on
a Kristallꢀ4000 chromatograph (TCD; Ar (high purity);
tion and Nanotechnology [8]. The Kꢀedge absorption
SKT column, 150
×
0.4 cm; 130°С; 30 ml/min). Low
spectra of zinc and palladium were measured in the
transmission mode using two ionization chambers
filled with the corresponding mixtures of nitrogen and
concentrations of CO (<0.4 vol %) were determined
using a Riken Keiki gas analyzer with an IR cell
(model RIꢀ550A).
argon to provide 20 and 80% absorption for I0 and It
,
respectively. For the monochromatization of a synꢀ
chrotron radiation beam, monoblock cut Si(111) and
Si(220) monochromators were used (for the К edges of
Zn and Pd, respectively). The catalyst powder was
pressed as a pellet 1.5 mm in thickness before the meaꢀ
surements. Programs from the IFEFFIT package [12]
were used for the processing of spectra, and photoelecꢀ
tron scattering phases and amplitudes used for the
refinement of local structure parameters were calcuꢀ
lated using the FEFF program [13, 14]. The range of
The liquid organic reaction products in aqueous
and organic phases were identified by gas chromatogꢀ
raphy–mass spectrometry (GC–MS) on MSD 6973
(Agilent) and Automassꢀ150 (Delsi Nermag) instruꢀ
ments; EI = 70 eV; sample volume, 1
umn, 0.32 500 cm, Df = 0.52 mm; 50
10 K/min; 270°С inj = 250° ; constant flow rate of
1 ml/min; split ratio, 1 : 100–200; CPSilꢀ5, 0.15
250 cm, Df = 1.2 mm; 50° (8 min); 10 K/min;
270° inj = 250° inj = 2.2 bar; split ratio, 1 : 300.
μ
l; HPꢀ5MS colꢀ
×
°С
(5 min);
;
T
С
×
С
С;
T
С; P
electron wave numbers (k) in the model refinement
Quantitatively, the concentrations of organic subꢀ
stances were determined by GLC on a Varian 3600
procedure was 3.0–13.0 Å–1.
instrument; SEꢀ30 Chromtec column, 0.25
Df = 0.3 mm; 50° (5 min); 10 K/min; 280°
250° Pinj = 1 bar; split ratio, 1 : 200; FID; trifluoꢀ
×
250 cm,
The diffraction patterns of catalysts were also meaꢀ
sured with the use of synchrotron radiation at the
Structural Materials Science station of the Kurchatov
Center of Synchrotron Radiation and Nanotechnolꢀ
ogy. The diffraction patterns were measured in the
transmission geometry (Debye–Scherrer) using Fujiꢀ
Film photosensitive imaging plates as a detector. The
С
С;
Tinj =
С;
romethylbenzene was used as an internal standard for
an organic layer, and internal normalization was used
for an aqueous layer. The ethanol content of the aqueꢀ
ous phase was determined by GC–MS from a ratio
between the integrated signals of ethanol and water
using the absolute calibration technique.
The charge states of metalꢀcontaining active cataꢀ
lyst components were studied by XPS.
The photoelectron spectra were measured at room
temperature on an XSAMꢀ800 spectrometer from
Kratos (United Kingdom) using nonmonochromatic
wavelength was
15 min.
λ
= 0.69 Å, and the exposure time was
RESULTS AND DISCUSSION
Catalytic Activity and Selectivity of Pd,ZnꢀContaining
Systems in the Reductive Dehydration of Ethanol
Mg
K radiation whose power was no higher than 90 W.
α
On the test catalysts, ethanol is converted into С1–
С11 alkanes and olefin hydrocarbons, carbon oxides,
liquid oxygenates, and water. Table 2 summarizes the
balance of the yields of reaction products on a carbon
basis (mol %) in the presence of monoꢀ and bimetallic
catalysts.
The spectra were measured under conditions of a conꢀ
stant relative energy resolution with a step of 0.1 eV.
The measurements were performed at a pressure of
~5
10–8 Pa in the analytical chamber. The analyzed
×
spectra were approximated by a Gaussian profile or a
sum of these profiles, and the background due to secꢀ
ondary electrons and photoelectrons that underwent
It was found that the maximum yield of the target
energy losses was approximated by a straight line. The fraction of С3–С10 aliphatic hydrocarbons (~50 wt %)
energy scale of the spectrometer was calibrated in was reached upon catalysis using a Pd–Znꢀcontaining
accordance with a standard procedure using the folꢀ system prepared by supporting the heterometallic aceꢀ
lowing binding energies: Cu
368.3 eV; and Au 4 7/2, 84.0 eV. Quantitative analysis (Fig. 2). With the use of monometallic systems conꢀ
was performed based on element sensitivity factors taining Pd ( ) or Zn (II) with active component conꢀ
2p3/2, 932.7 eV; Ag
3d5/2
,
tate complex Pd(μꢀOOCMe)4Zn(OH2) on Al2O3 (III)
f
I
from the software provided by the manufacturer. The centrations equivalent to the concentration in the hetꢀ
samples were fixed with a doubleꢀsided Scotch tape. erometallic system, a decrease in the yield of the target
The surface charging for support spectra was taken fraction by 25 or 10 wt %, respectively, was observed.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
261
Table 2. Balance on carbon in the conversion of ethanol in the presence of catalysts I–III*
Pd (catalyst I)
C supplied, % C obtained, %
100.00
Zn (catalyst II)
C supplied, % C obtained, %
100.00
Pd–Zn (catalyst III)
C supplied, % C obtained, %
100.00 2.62
С1
11.79
20.78
12.06
19.31
0.86
С1
0.28
24.10
3.48
32.14
3.12
7.96
2.25
1.52
0.14
0
С1
С2
С2
С2
21.77
2.22
28.46
2.78
13.68
2.19
3.74
0
С3
С3
С3
С4
С4
С4
С5
С5
С5
С6
6.10
С6
С6
С7
1.26
С7
С7
С8
2.03
С8
С8
С9
0.20
С9
С9
С10
С11
Oxygenates
Aromatics
СO
0.48
С10
С11
Oxygenates
Aromatics
СO
С10
С11
Oxygenates
Aromatics
СO
0
0.12
0
0
6.79
23.75
1.27
0
17.54
0.97
2.99
1.04
100.00
1.60
14.23
2.38
CO2
Total
CO2
Total
0
CO2
Total
100.00
100.00
* The data are given in mol %.
Note that alkanes were predominant among the circulates in a gas mixture through a catalyst bed, into
carbon chain growth [15].
products of ethanol conversion on the Pdꢀcontaining
catalyst, whereas the target fraction contained 70 wt %
olefins and only 30 wt % alkanes on the Znꢀcontaining
catalyst. A more detailed analysis of the reaction prodꢀ
ucts (Table 3) indicated that, in the case of the Znꢀ
containing system, a considerable yield of ethylene
(126.61 mmol) and the complete absence of carbon
monoxide and carbon dioxide were observed in conꢀ
trast to the Pdꢀcontaining system, in which the yields
of carbon oxides were 193.13 and 32.35 mmol, respecꢀ
tively, and ethylene was absent (Table 3).
We can also hypothesize that the formation of tarꢀ
get products (С3+ alkanes) on the palladiumꢀcontainꢀ
ing catalyst results from the participation of surface
ethylene, which is formed in the reaction of ethylene
dehydration, in carbon chain growth followed by
hydrogenation at palladiumꢀcontaining sites, as
shown in the following reaction scheme:
Yield of the С3–С10 fraction, wt %
50
45
40
Olefins
Alkanes
These results suggest radically different reaction paths
in the conversion of ethanol on Pdꢀ and Znꢀcontaining
systems. Figure 3 shows the distribution of hydrocarbons
in the target fraction for these catalysts (I, II).
35
30
25
20
15
10
5
Thus, the composition of products prepared on the
palladiumꢀcontaining system was close to the compoꢀ
sition prepared in the presence of a commercial aluꢀ
mina–platinum catalyst [15]. This fact suggests the
identity of reaction paths in ethanol conversion on the
above systems. Previously, it was found that a possible
reaction path in ethanol conversion into linear alkanes
with even numbers of carbon atoms is the incorporaꢀ
tion of ethylene, which is formed from ethanol and
0
Pd (I)
Zn (II)
Pd–Zn (III)
Fig. 2. Yields of С –С alkane and olefin fractions
3
10
obtained from ethanol in the presence of monoꢀ and bimeꢀ
tallic catalysts.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

262
CHISTYAKOV et al.
Table 3. Yields (mmol) of the conversion products of 700 mmol of ethanol on Pdꢀ and Znꢀcontaining systems
Pd (
I
)
Zn (II
olefins
Aliphatic hydrocarbon
)
Pd–Zn (III)
Products
alkanes
olefins
others
alkanes
others
alkanes
olefins
others
s
С1
С2
160.01
141.01
99.40
–
0
–
–
–
4.00
43.45
26.18
–
–
–
–
36.34
92.68
51.09
–
–
–
–
126.61
59.21
58.02
78.91
С3–С10
6.68
Other products
Oxygenates
Aromatics
СO
–
–
–
–
–
–
–
–
–
–
11.38
2.01
–
–
–
–
–
–
28.97
1.32
0
–
–
–
–
–
–
–
–
–
–
28.39
1.58
–
–
–
–
193.13
32.35
18.07
41.39
14.34
0.50
CO2
0
H2
5.60
–H O
2
alkanes by reaction (I), and carbon monoxide result
from the reaction of ethanol dehydrogenation folꢀ
lowed by the decarbonylation of acetaldehyde:
H3C–CH2–OH
H2C=CH2
(IV)
+H2
+C H
2
+C H
2
4
4
…
CnH2n
CnH2n + 2.
As was found previously [5], it is most likely that
hydrogen, which is required for the formation of
O
–H2
H3C–CH2–OH
H3C–C
CH4 + CO. (V)
H
The predominance of linear alkanes, whose conꢀ
centration was higher than 90 wt %, in the reaction
products suggests the above reaction path. Note that
alkanes with even numbers of carbon atoms were preꢀ
dominant (Fig. 3a, Table 4). Alkanes with odd numꢀ
bers of carbon atoms, whose concentration was much
lower (see Fig. 3a), can result from the decarbonylaꢀ
tion reactions of acetaldehyde homologs. The water
gas shift reaction can occur simultaneously to cause
the formation of carbon dioxide:
Yield, wt %
14
(a)
12
Olefins
Alkanes
10
8
6
4
CO + H2O
CO2 + H2.
2
This was indirectly supported by the equivalent
yields of carbon oxides and alkanes with odd numbers
of carbon atoms: 204.29 and 225.48 mmol, respecꢀ
tively.
0
C3 C4 C5 C6 C7 C8 C9 C10 C11
(b)
20
18
16
14
12
10
8
6
4
2
0
Olefins
Alkanes
Note that the high hydrogenating activity of
reduced palladium negatively affects the oligomerizaꢀ
tion of olefins because of the occurrence of a competꢀ
ing reaction of ethylene hydrogenation, which results
in the termination of carbon chain growth. The similar
yields of С3+ alkanes (27 wt %) and ethane (20 wt %)
suggest similar rates of the competing reactions. The
process selectivity also considerably decreases because
of the formation of methane and carbon oxides, whose
total yield was close to 35 wt %.
C3
C4
C5
C6
C7
C8 C10
Fractions
With the use of the zincꢀcontaining system, the most
likely reaction mechanism seems involve the dehydroꢀ
genation of ethanol with the formation of acetaldehyde
Fig. 3. Fractional composition of С –С hydrocarbons
3
10
obtained from ethanol in the presence of catalysts (a)
(b) II
I and
.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
263
Table 4. Ratios between the yields of the nearest hydrocarbon
homologs formed upon ethanol conversion on palladiumꢀ
containing catalyst I
followed by the aldol condensation of acetaldehyde to
form monomethylꢀsubstituted carbon chains and the
reduction of the resulting oxygenates [16]:
–H2
Homologs
Yield ratio
C2H5OH
CH3COH
+(n – 3)AcH
С4/C3
C6/C5
C8/C7
C10/C9
1.52
6.53
9.75
2.38
(VI)
HOCnH2nCOH
CH3CH2CH(CH3)CmH2m
11 and – 3.
+H2
,
where
n
≤
m = n
The low hydrogenating activity of the zinc catalyst
does not provide for the occurrence of reductive dehydraꢀ
tion to selectively form alkanes (reaction scheme (VI)).
However, the composition of olefin products is consisꢀ
tent with this reaction path (Fig. 3b). It was found that
the concentration of hydrocarbons with branched
structures was 50–60% (3ꢀmethylꢀsubstituted derivaꢀ
tives were predominant) after the hydrogenation of a
С3+ olefin fraction in the presence of the wellꢀknown
Ni/Al2O3 catalyst for this reaction.
As found previously, a number of oxygenates such
as diethyl ether and acetaldehyde were also converted
into aliphatic hydrocarbons in the presence of a comꢀ
mercial alumina–platinum catalyst (APꢀ64) as the
contact time was increased [15]. In this context, the
oxygenates, which are produced in a sufficiently high
yield, can be considered as raw materials for recycling
in order to obtain an additional amount of hydrocarꢀ
bons.
Table 5. Yields (mmol) of the conversion products of
700 mmol of ethanol in the presence of catalytic system IIIa
Products
Alkanes
Olefins
Others
Hydrocarbons
С1
С2
97.50
77.43
47.54
–
–
–
–
0.17
1.48
С3–С9
Other products
Oxygenates
Aromatics
СO
–
–
–
–
–
–
–
–
–
–
26.17
0
289.30
53.11
123.10
CO2
Note that the yield of ethylene was high (126.61 mol).
As noted above, ethylene may participate in the forꢀ
mation of linear aliphatic hydrocarbons upon catalysis
by the palladium system. However, it is likely that the
accumulation of ethylene in reaction products was due
to a small contribution of conversion in accordance
H2
mum yield of aliphatic hydrocarbons is higher than
80% on a theoretical basis.
with reaction scheme (IV) in the presence of system II
.
To intensify the process of ethanol conversion, we
performed a reaction with the use of catalyst IIIа, in
which the concentration of Pd–Zn active compoꢀ
nents was higher than that in the above example by a
factor of 3.
Thus, the hypothesis of the occurrence of both of
the reaction paths in the presence of the Pd–Zn cataꢀ
lyst based on a heterometallic precursor seems reasonꢀ
able.
As can be seen in Table 3 and Fig. 2, upon the conꢀ
version of ethanol on palladium–zinc system III, the
liquid products contained approximately the same
amounts of alkanes and alkenes.
This fact suggests an additive catalytic effect of
palladium and zinc in the reactions of ethanol conꢀ
version. It is most likely that this effect was due to
the complementarity of ethanol conversions in
reactions (I) and (II). As can be seen in Table 3,
small amounts of ethylene and carbon oxides were
present in the reaction products; this fact indicates
that the reaction simultaneously occurred via reacꢀ
tion paths (IV)–(VI).
In Table 5, it can be seen that the yields of alkanes,
carbon monoxide, methane, and hydrogen noticeably
increased in the presence of catalyst IIIа (as compared
with catalyst III). The yield of the target fraction of
aliphatic hydrocarbons decreased by a factor of >3;
olefins were also almost completely absent from the
reaction products. This behavior of the catalytic sysꢀ
tem suggests that an increase in the amount of pallaꢀ
diumꢀcontaining centers resulted in the predomiꢀ
nance of reaction path (IV) and the suppression of
activity in the process of carbon skeleton condensation
via reaction path (VI).
In this context, we can hypothesize that, as the
amount of active components was increased, pallaꢀ
diumꢀcontaining centers became predominant in ethꢀ
anol conversion and the role of zincꢀcontaining cenꢀ
Thus, the experimental results demonstrate that
the process that occurs in the presence of palladium–
zinc catalytic system III, which was prepared from a
heterometallic precursor, is most selective for a fracꢀ
tion of aliphatic hydrocarbons. In this case, a maxiꢀ ters decreased.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

264
Yield of C3–C10 fraction, wt %
CHISTYAKOV et al.
of ~2. The yield of oxygenates remained at an approxꢀ
imately the same level as that in the case of a catalyst
based on a heterometallic complex. Thus, on the catꢀ
alyst in which active components are supported nonꢀ
uniformly and separated by considerable distances,
the effect of the cocatalytic reaction enhancement
considerably decreased.
Previously, a procedure for the preparation of a
Pd–Zn alloy by supporting the heterometallic comꢀ
plex Pd( ꢀOOCMe)4Zn(OH2) followed by reduction
50
45
40
35
30
25
20
15
10
5
Olefins
Alkanes
μ
in hydrogen has been developed [9, 10].
We used this procedure to prepare catalyst
V. The
tests of this catalyst in ethanol conversion demonꢀ
strated that the yield of the target fraction of С3–С10
hydrocarbons was almost the same that the yield
observed with the system containing only palladium
(Fig. 4, Table 6). Alkanes dominated in the products of
ethanol conversion, and ethylene was almost absent.
Nevertheless, a decrease in the yields of carbon oxides
and methane and a simultaneous increase in the yield
of olefins suggest the occurrence of reaction paths typꢀ
ical of the zincꢀcontaining system (Fig. 3, Table 2). It
is likely that the formation of intermetallide clusters
considerably intensified reaction paths (IV) and (V),
0
Pd/Zn (
V
)
Pd + Zn (IV
)
Pd–Zn (III)
Fig. 4. Yields of the products of ethanol conversion in the
presence of bimetallic catalysts prepared by various proceꢀ
dures.
We found that the catalyst pretreatment procedure
and the nature of the metal complex precursors used
exerted a considerable effect on process selectivity.
It is believed that, upon supporting active compoꢀ which are characteristic of the palladiumꢀcontaining
nents from a heterometallic complex, these compoꢀ system. However, a number of zinc atoms were distribꢀ
nents were arranged on the support surface at disꢀ uted over the surface of
ꢀAl2O3.
γ
tances close to the sum of the van der Waals atomic
radii of the metals to provide the bifunctional characꢀ
ter of these clusters [9]. This was demonstrated above
using ethanol conversion as an example. To test this
hypothesis, we prepared catalyst IV by the separate
To study the crossꢀcondensation reaction of the carꢀ
bon skeletons of different alcohols, we used glycerol as a
coreactant for ethanol. We found that glycerol added to
ethanol in an amount of no more than 30 wt % was an
active coreactant in the formation of С4–С10+ aliꢀ
phatic hydrocarbons to considerably increase the yield
of these hydrocarbons. The composition of reaction
products essentially depends on the composition of
palladium–zincꢀcontaining systems.
supporting of homometallic precursors Zn(OAc)2
⋅
2H2O and Pd3(OAc)6 onto the support.
As can be seen in Fig. 4 and Table 6, the yield (wt %)
of the target fraction of hydrocarbons on this catalyst
decreased by a factor of >2, and the yield of the
Table 7 summarizes the results of the combined
ethane–ethylene fraction also increased by a factor conversion of 70 wt % ethanol and 30 wt % glycerol on
Table 6. Yields (mmol) of the conversion products of 700 mmol of ethanol in the presence of Pdꢀ and Znꢀcontaining systems
Pd + Zn (IV
)
Pd/Zn (
olefins
Aliphatic hydrocarbons
V
)
Pd–Zn (III)
Products
alkanes
olefins
others
alkanes
others
alkanes
olefins
others
С1
С2
53.12
101.25
25.37
–
–
–
–
79.86
165.16
64.55
–
–
–
–
36.34
92.68
51.09
–
–
–
–
197.82
46.05
0.61
1.36
58.02
78.91
С3–С10
Other products
Oxygenates
Aromatics
СO
–
–
–
–
–
–
–
–
–
–
28.64
0.11
–
–
–
–
–
–
68.51
0
–
–
–
–
–
–
–
–
–
–
28.39
1.58
–
–
–
–
56.33
16.93
1.48
85.15
20.27
8.60
41.39
14.34
0.50
CO2
H2
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
265
Table 7. Yields (wt %) of the conversion products of 70 wt % ethanol and 30 wt % glycerol on Pdꢀ and Znꢀcontaining systems
Pd (
I
)
Zn (II
olefins
Aliphatic hydrocarbons
)
Pd–Zn (III)
Products
alkanes
olefins
others
alkanes
others
alkanes
olefins
others
С1
2.13
8.41
–
–
–
–
0.18
6.97
–
–
–
–
0.17
2.24
–
–
–
–
С2–С3
24.35
24.08
16.04
37.32
10.82
27.05
С4–С10
15.80
14.25
11.00
Other products
Oxygenates
Aromatics
СO
–
–
–
–
–
–
–
–
–
–
16.09
2.40
7.08
2.72
0.01
–
–
–
–
–
–
16.64
2.21
2.42
3.86
0.05
–
–
–
–
–
–
–
–
–
–
35.14
2.64
5.16
1.87
0.01
–
–
–
–
CO2
H2
the palladiumꢀcontaining system. Note that the addiꢀ of a mixture of 70% ethanol and 30% glycerol to the
tion of glycerol facilitated an increase in the total yield
of hydrocarbons by ~10 wt % because of an increase in
the yield of olefins. Among other distinctive features,
note an increase in the yield of oxygenates from 5 wt %
in ethanol conversion to 15 wt % in the combined
reaction. The yield of carbon oxide and carbon dioxide
decreased from >30 wt % and the yield of methane
decreased from 12 wt % in the case of ethanol to 2 wt %
in the conversion of the mixture of the alcohols.
It is most likely that this fact suggests the competiꢀ
tive chemisorption of ethanol and glycerol at catalytic
active sites. Zincꢀcontaining system II exhibited the
highest activity in the crossꢀcondensation reaction. In
the presence of this system, the yield of the alkane–
olefin fraction increased to 50 wt %. Data given in
yields of hydrocarbons prepared by the conversion of
only ethanol.
The above data indicate that, with the use of glycꢀ
erol as a coreactant, the yield of the ethane–ethylene
fraction decreased by a factor of 2, and the yield of С5–
С7 hydrocarbons increased by a factor of 4; in the case
of С10, the yield increased by a factor of 6.
Thus, the composition of the catalytic system preꢀ
pared based on aluminum oxide and acetate homoꢀ
and heterometallic complexes essentially affected the
direction of С2–С3 alcohol conversion reaction into a
fraction of С4–С10 aliphatic hydrocarbons. It is eviꢀ
dent that the found high sensitivity of the reaction was
Table 7 indicate that the increase in the yield of the tarꢀ due to different structures of catalytically active comꢀ
get hydrocarbon fraction was due to an increase in the
yield of olefins; in this case, the yield of alkanes
remained at the same level (12–15 wt %). The increase
in the yield of unsaturated hydrocarbons can be
explained by the fact that the glycerol molecule conꢀ
tains three hydroxyl groups and a triple amount of
hydrogen is required for the saturation of C–C bonds
upon its deoxygenation, as compared with ethanol.
Note that the yield of propylene increased considerꢀ
ably and the yield of propane decreased. It is likely that
the remaining 12% glycerol participated in the
increase of the yield of hydrocarbons, mainly, the С5+
fraction, by reaction (III).
In the presence of heterometallic palladium–zincꢀ
containing system III, the concentration of the
alkane–alkene fraction decreased mainly as a result of
an increase in the yield of oxygenates (Table 7).
Data given in Table 8 can be used to estimate the
contribution of glycerol to the formation of one or
another hydrocarbon. Table 8 summarizes the ratios of
the yields of hydrocarbons prepared by the conversion
plexes.
Table 8. Ratios between the yields of hydrocarbons obtained
from ethanol (e) and a mixture of ethanol and glycerol (m)
Pd (
alkanes
I
)
Zn (II
alkanes
)
Yield ratio
olefins
olefins
С1m/С1e
0.36
0.54
1.15
0.98
1.79
2.07
0.83
1.61
–
–
0.05
0.12
0.23
0.29
22.64
0.49
4.72
0.42
–
С2m/С2e
С3m/С3e
С4m/С4e
С5m/С5e
С6m/С6e
С7m/С7e
С8m/С8e
С9m/С9e
0.46
1.75
1.07
3.31
1.26
3.85
1.85
6.15
0.87
2.34
0.67
1.70
1.27
0
0.73
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

266
CHISTYAKOV et al.
Intensity, arb. units
9000
O 1
s
8000
7000
6000
5000
4000
3000
Pd 3d
O KVV
Al 2
s
Al 2p
C 1
s
Pd MNN
C KVV
Pd 4p
2000
1000
0
1000
800
600
400
200
0
Binding energy, eV
Fig. 5. Survey photoelectron spectrum of Pd–Zn catalyst III
.
Intensity, arb. units
Intensity, arb. units
8
1
2
3
200
150
100
50
1023.3
1
2
1021.8
3
4
0
0
−
50
355
−
4
350
345
340
335
330
325
1050 1045 1040 1035 1030 1025 1020 1015
Binding energy, eV
Binding energy, eV
Fig. 6. The Pd 3
) a sample after an experiment, (
and ( ) an initial sample.
d
photoelectron spectra of catalyst III
:
Fig. 7. The Zn 2
) a sample after an experiment, (
and ( ) an initial sample.
p
photoelectron spectra of catalyst III
:
(1
2
) a reduced sample,
(
1
2
) a reduced sample,
3
3
GENESIS OF CATALYTICALLY ACTIVE
COMPONENTS UPON SUPPORTING
THE PRECURORS AND PRETREATMENT
The Pd 3d spectrum (Fig. 6) includes data for catꢀ
alyst III in initial and reduced states and after operaꢀ
tion. Peaks with binding energies of 337.1 and 335.1 eV
correspond to the states Pd2+ and Pd0, respectively. The
experimental data indicate that Pd mainly occurred as
PdO in the initial catalyst sample, whereas it was parꢀ
tially reduced to palladium metal upon treatment with
hydrogen and after an experiment.
Study of the States of Active Components
by Xꢀray Photoelectron Spectroscopy
Figure 5 shows a survey spectrum of initial Pd–Zn
catalyst III. According to quantitative analysis, the
surface composition of the catalyst corresponds to the
Figure 7 shows the spectrum of the Zn 2
initial, reduced, and spent Pd–Zn catalysts. The
p region for
formula C25.1O40.8Al30.5Pd3Zn0.6
.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
267
Zn
ZnO
Pd
PdO
Pd initial
Pd–Zn initial
Pd–Zn initial
Pd–Zn reduced
Pd–Zn spent
Pd spent
Pd–Zn spent
Pd–Zn reduced
9640
9660
9680
9700
9720
, eV
24325 24350 24375 24400 24425 24450
E
E
, eV
Fig. 8. The
K
ꢀedge XANES spectra of Zn for the samples
Fig. 9. The
K
ꢀedge XANES spectra of Pd for the samples
of Pd–Zn catalyst III
.
of Pd–Zn catalyst III.
charge states Zn2+ and Zn0 in the spectrum correspond ZnAl2O4 with Zn in the tetrahedral positions of the
spinel structure.
to lines at 1023.3 and 1021.8 eV, respectively. In the
experimental spectra, zinc occurred in the state Zn2+
(1022.9 and 1022.6 eV); however, a shift of peaks corꢀ
responding to the charge state Zn2+ allowed us to conꢀ
clude that, in all of the catalysts, Zn occurred as the
The reductive treatment of the Pd–Zn catalyst
with hydrogen at 450°
cant changes in the
Zn: an inflection appeared in the lowꢀenergy slope of
the line, which is characteristic of Zn0. In this case, the
fraction of zeroꢀvalent zinc was no higher than 30%.
Analogous changes in XANES spectra were observed
for a bimetallic catalyst after a catalytic cycle. In this
case, the fraction of reduced Zn was somewhat lower
(~20%).
С
for 10 h resulted in insignifiꢀ
Kꢀedge absorption spectrum of
oxide ZnxAl2 – xO4 with a spinel structure.
δ
Study of the Structure of Active Components
by XAFS Spectroscopy
Figures 8 and 9 show the experimental
XANES spectra of Zn and Pd. For comparison, data
for the reference samples of Pd, PdO, Zn, and ZnO
Kꢀedge
An analysis of the Kꢀedge XANES spectra of Pd
(Fig. 9) suggests that the state of palladium in the iniꢀ
tial catalysts (both monoꢀ and bimetallic) was close to
Pd2+. After reductive treatment or catalysis, the
absorption spectrum underwent considerable changes
are also given. The
Kꢀedge position of Zn absorption
for freshly prepared catalysts (both pure zinc and
zinc–palladium) coincides with that of ZnO; however, to approach the XANES spectrum characteristic of
the nearꢀedge fine structure is dramatically different Pd metal.
(Fig. 8). Therefore, zinc in the catalysts occurs in the
oxidized state Zn2+ rather than as ZnO. Taking into
account published data [17] on a very characteristic
XANES spectrum for ZnAl2O4 with a developed fine
structure, we can hypothesize that the major portion of
Figures 10 and 11 show atomic radial distribution
(ARD) curves for Zn and Pd, which were obtained by
the Fourier transform of experimental EXAFS specꢀ
tra. Reference curves for Pd, PdO, Zn, and ZnO and
the results of the optimization of structural models are
also given. Tables 9 and 10 summarize the local enviꢀ
zinc ions diffused into ꢀalumina grains to form a
γ
nearꢀsurface nanocrystalline phase identical to ronment parameters of Pd and Zn atoms. The experiꢀ
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

268
CHISTYAKOV et al.
Pd–Pd, Zn
Zn–Zn, Pd
Zn...Zn
Pd...Pd
Zn–O
Pd–O
Zn
Pd
PdO
ZnO
Pd initial
Pd, Zn initial
Pd, Zn initial
Pd spent
Pd, Zn spent
Pd, Zn spent
Pd, Zn reduced
Pd, Zn reduced
0
1
2
3
4
5
6
R, Å
0
1
2
3
4
5
6
R, Å
Fig. 10. ARD curves for the local environment of Zn in
Pd–Zn catalyst III (solid lines) and the results of the
refinement of structural models (points).
Fig. 11. ARD curves for the local environment of Pd in Pd
) and Pd–Zn (III) catalysts (solid lines) and the results of
the refinement of structural models (points).
(
I
mental results support the above hypothesis that Zn in atoms. The Pd–Pd distance is 2.80 Å, which is someꢀ
the Pd–Zn catalyst occurs as a spinel structure what greater than that in bulk palladium. This increase
because a peak due to the second coordination sphere in the Pd–Pd bond length can be due to hydride forꢀ
was observed, and this peak did not coincide with the mation [18]. On the contrary, a considerable shortenꢀ
peak of zinc in ZnO. The ARD curves of reduced samꢀ ing of Pd–Pd bonds was observed in the case of
ples exhibited a clearly pronounced maximum, which reduced palladium–zinc catalysts. Moreover, Pd–Zn
corresponded to a metal–metal bond; this fact sugꢀ bonds (2.56–2.58 Å) should be involved into the strucꢀ
gests the partial formation of a zinc metal phase or the tural model in order to obtain data consistent with
formation of an alloy of Pd and Zn.
experimental results; this is direct evidence for the forꢀ
mation of an intermetallide. Note that the formation
of palladium–zinc alloy nanoparticles with similar
local structure parameters was observed previously as a
result of the reductive treatment of double palladium–
zinc carboxylate [19].
As judged from the ARD curve, the local environꢀ
ment of palladium in the initial catalysts is close to
PdO. This suggests that palladium atoms form a nanꢀ
odispersed oxide phase on the support surface.
After treating with hydrogen or performing a cataꢀ
lytic cycle, palladium was almost completely reduced
to a metal state, as supported by the disappearance of
a peak due to Pd–O in the ARD curve and the appearꢀ
ance of a peak due to Pd–Pd(Zn). Differences in the
position and shape of the latter peak for monometallic
palladium and bimetallic zinc–palladium catalysts
suggest the possible formation of an alloy. This
hypothesis was supported by a more detailed simulaꢀ
tion of the local environment of palladium atoms
(Table 9). In the case of a spent monometallic catalyst,
Study of the Structure of Active Components
by Xꢀray Diffraction Analysis
Figure 12 shows XRD data for heterometallic sysꢀ
tem III in an initial state and after reduction and catalꢀ
ysis. Note that diffraction data are difficult to interpret
because of the occurrence of broad peaks from the iniꢀ
tial
γꢀAl2O3 matrix.
As can be seen in Fig. 12, the spectra of the initial
the local environment of Pd consists of palladium Pd–Znꢀcontaining catalytic system exhibited reflecꢀ
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
269
Table 9. Local structure parameters obtained by simulating the Kꢀedge EXAFS spectra of Zn: coordination number (N), interꢀ
2
atomic distance (R), and Debye–Waller factor (σ )
2
Sample
Coordination sphere
N
R,
Å
σ
,
Å2
Zn
Zn–Zn
Zn–Zn
Zn–O
6.0
6.0
4.0
6.0
6.0
3.9
4.2
2.8
0.8
3.4
2.7
0.6
0.8
2.5
2.67
2.90
1.96
3.16
3.27
1.94
3.45
1.92
2.57
3.42
1.90
2.46
2.54
3.41
0.0065
0.0104
0.0046
0.0050
0.0050
0.0079
0.0110
0.0041
0.0060
0.0085
0.0045
0.0112
0.0043
0.0095
ZnO
Zn...Zn
Zn...Zn
Zn–O
Pd–Zn initial
Pd–Zn spent
Zn...Zn
Zn–O
Zn–Pd
Zn...Zn
Zn–O
Pd–Zn reduced
Zn–Zn
Zn–Pd
Zn...Zn
Table 10. Local structure parameters obtained by simulating the Kꢀedge EXAFS spectra of Pd: coordination number (N), inꢀ
2
teratomic distance (R), and Debye–Waller factor (σ )
2
Sample
Coordination sphere
N
R,
Å
σ
,
Å2
Pd
Pd–Pd
Pd–O
12
2.74
2.02
3.03
3.42
2.00
3.02
3.40
2.80
2.01
3.04
3.41
2.05
2.58
2.72
1.99
2.56
2.68
0.0067
0.0017
0.0050
0.0050
0.0026
0.0050
0.0050
0.0067
0.0017
0.0050
0.0050
0.0124
0.0056
0.0124
0.0003
0.0091
0.0087
PdO
4.0
4.0
8.0
3.9
2.5
1.7
8.9
4.2
2.6
2.5
1.3
2.0
8.2
0.4
3.4
4.9
Pd…Pd
Pd…Pd
Pd–O
Pd initial
Pd…Pd
Pd…Pd
Pd–Pd
Pd–O
Pd spent
Pd–Zn initial
Pd…Pd
Pd…Pd
Pd–O
Pd–Zn spent
Pd–Zn
Pd–Pd
Pd–O
Pd–Zn reduced
Pd–Zn
Pd–Pd
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

270
CHISTYAKOV et al.
PdO
(101)
PdH0.6Pd
(111) (111)
PdZn
(111)
Pd spent
Pd–Zn spent
Pd–Zn reduced
Pd–Zn initial
γꢀAl2O3
30
35
40
45
50
, deg
2θ
Fig. 12. Xꢀray diffraction data for
γ
ꢀAl О , Pd (
I
), and Pd–Zn (III) catalysts. The Pd catalyst was prepared from a palladium aceꢀ
2
3
tate complex (for comparison with published data, the angle scale was converted to the Cu
K
wavelength λ = 1.5406 Å).
α
1
tions characteristic of a PdO phase [20]. After the step metal phase [23]. The analogous formation of a
hydride phase was also observed in the monometallic
Pd catalyst, as demonstrated by EXAFS data.
of reduction, the diffraction pattern of the catalyst
sample exhibited no peaks due to the PdO phase and
peaks characteristic of the palladium hydride PdH0.6
[21] and a Pd–Zn alloy (1 : 1) [22] appeared. Conseꢀ
The decrease in the intensity of a peak correspondꢀ
ing to a Pd–Zn alloy after catalysis suggests that alloy
particles can be active sites in the test reaction. Previꢀ
quently, in the course of hydrogen treatment at 450°С,
Pd(II) was reduced to Pd(0). In this case, the phases of ously, the high selectivity of catalysts based on Pd–Zn
alloys in the steam conversion reaction of methanol
was demonstrated [24–26].
palladium hydride and an alloy with zinc were formed.
After a catalytic experiment with the Pd–Zn catalyst,
a decrease in the peak intensity corresponding to the
Pd–Zn alloy and an increase in the intensity of a peak
due to PdH0.6 were observed. In this case, we can state
Based on the experimental structural data, we can
conclude that the phase composition of active compoꢀ
nents essentially depends on pretreatment conditions
the occurrence of reflections corresponding to a Pd (Table 11).
Table 11. Phase composition of active components at the stages of catalyst pretreatment
Phase composition after pretreatment
stage 2
No.
Brief notation
stage 1
I
Pd
PdO
ZnxAl2O3 +
[PdH0.6 + Pd]
II
III
IV
V
Zn
ZnxAl2O3 +
x
δ
x
δ
Pd–Zn
Pd + Zn
Pd/Zn
PdO + ZnxAl2O3 +
PdO + ZnxAl2O3 +
[PdH0.6 + PdZn]
[PdH0.6 + Pd + PdZn] + ZnxAl2O3 +
[PdH0.6 + Pd] + ZnxAl2O3 +
x
x
δ
δ
x
δ
x
δ
[PdH0.6 + PdZn]
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

COCATALYTIC EFFECT OF PALLADIUM AND ZINC IN THE CONDENSATION
271
The evolution scheme of phase transformations of yield of aliphatic hydrocarbons, which is as high as
active components at all stages of preparation and 40–50 wt % on a starting ethanol basis, and a conꢀ
catalysis for a bimetallic catalyst, which exhibited the siderable amount of branchedꢀchain structures
highest activity, is given below:
(to 50 wt %).
Ar, 500°C
These conclusions were supported by selectivity
changes depending on the phase composition of the
catalyst formed upon various catalyst pretreatment
procedures. In the case of a long reductive treatment
with Н2, a higher concentration of the Pd–Zn alloy
and a lower concentration of oxide with the structure
of a complex spinel are formed on the surface of aluꢀ
mina. In this case, a more considerable contribution of
the reductive dehydration of ethanol is observed,
which leads to the formation of linear alkanes. Upon
the formation of a catalytic system by thermal treatꢀ
ment in Ar followed by reduction in H2, to 80% zinc
atoms on the surface of alumina are incorporated into
the structure of a complex oxide; the remaining zinc
atoms enter into the composition of PdxZny intermetalꢀ
lide clusters. Palladium metal clusters are also formed
on the support surface. In this case, the yield of the
hydrocarbon fraction increases because of reaction (II),
which leads to the formation of branchedꢀchain oleꢀ
fins. In this case, the yield of С1–С2 light alkanes and
carbon oxides considerably decreases.
Pd(II)Zn(II)(OAc)4(OH)2/Al2O3
H , 450°C
PdO
2
+
ZnxAl2O3 + x
[PdH0.6 + PdZn]
δ
catalysis
+
ZnxAl2O3 + x
[PdH0.6 + Pd + PdZn]
δ
+
ZnxAl2O3 + x
.
δ
From the above scheme, it follows that metal and
oxide clusters are formed on the support surface in the
course of the preparation of a Pd–Znꢀcontaining catꢀ
alyst; these clusters can possess hydrogenating and
condensing functions.
CONCLUSIONS
The supporting of the bimetallic complex
Pd(
acetate complexes Zn(OAc)2
onto the surface of ꢀAl2O3 followed by thermal treatꢀ
μ
ꢀOOCꢀMe)4Zn(OH2) and the monometallic
⋅
2H2O and Pd3(OAc)6
γ
ment and reductive activation results in the formation
of two types of active components: (1) a hydrogenating
component as the clusters of Pd metal, Pd hydride,
and PdxZny alloy and (2) oxide components as the
mixed zinc–aluminum spinel ZnxAl2 – xO4 . The
results of the study on the catalytic activity δof the
resulting systems in the reactions of ethanol converꢀ
sion demonstrated that the active components formed
from a bimetallic precursor exert a complementary
effect to increase the total yield of hydrocarbons by
10–15%. At the same time, the selectivity for alkane
and alkene formation directly depends on the compoꢀ
sition of the catalytic system. The Pdꢀcontaining catꢀ
alyst exhibited an enhanced activity in the reductive
dehydration of ethanol to gasolineꢀfraction alkanes,
primarily, of linear structure, in accordance with overꢀ
all reaction scheme (I). The Znꢀcontaining system
with the spinel structure exhibits selectivity for the
condensation of ethanol carbon skeleton into alkenes
of the above fraction, but containing to 50%
branchedꢀchain hydrocarbons according to reaction
(II). The complementary effect of active components,
which results in an increase in the formation of a С3–
С10 alkane–alkene fraction in the reaction occurs in
the presence of palladium metal clusters and complex
zinc–aluminum oxide distributed over the surface of
In general, it is believed that the bifunctional
mechanism of alcohol conversion consists in the conꢀ
densation reaction of a carbon skeleton, which occurs
at acid sites, followed by hydrogen transfer to olefins
on the resulting clusters of palladium hydride.
It is most likely that, upon the preparation of a catꢀ
alyst from a heterometallic precursor, an optimum
ratio between closely spaced metal and oxide sites is
formed on the support surface after thermal treatment
followed by reduction. As a result of more closely
spaced hydrogenating and condensing active sites, the
spillover of hydrogen to unsaturated alkenes chemiꢀ
sorbed on the surface is facilitated and, correspondꢀ
ingly, a smaller amount of hydrogen is consumed for
the deep hydrogenation of ethanol to methane.
Thus, we demonstrated for the first time that the
selectivity of ethanol conversion can be controlled by
changing catalyst composition. This makes it possible
to obtain aliphatic hydrocarbon fractions with alkane
concentrations higher than 90 wt %, aliphatic hydroꢀ
carbon fractions with alkene concentrations to
70 wt %, and an alkane–olefin fraction.
γ
ꢀAl2O3, probably, at a short distance from each other.
The addition of glycerol to ethanol mainly results
The reductive dehydration reaction of ethanol in an increase in the yield of alkenes, and it decreases
occurs with the release of large amounts of methane, the possibility of affecting the composition of reaction
ethane, and carbon monoxide, which are formed by products upon changes in the catalyst composition.
ethylene hydrogenation reaction and reactions (IV) However, glycerol is an active coreactant, which parꢀ
and (V), to cause a decrease in the total yield of the tarꢀ ticipates in hydrocarbon chain growth according to
get С4–С10 fraction. It is most likely that the formation reaction (III); it leads to a decrease in the yields of
of carbon dioxide was due to the water gas shift reacꢀ methane and ethane and an increase in the yield of the
tion. The conversion of ethanol by reaction (II) with target fraction of aliphatic hydrocarbons to 55 wt %,
the formation of alkenes is characterized by a higher that is, more than 80% on a theoretical basis.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011

272
CHISTYAKOV et al.
12. Chernyshov, A.A., Veligzhanin, A.A., and Zubaviꢀ
REFERENCES
chus, Y.V., Nucl. Instrum. Methods Phys. Res., Sect. A
,
1. Tsodikov, M.V., Kugel, V.Ya., Slivinskii, E.V.,
Zaikin, V.G., Mordovin, V.P., Colon, G., Hidalgo, M.C.,
and Navio, J.A., Langmuir, 1999, vol. 15, p. 6601.
2009, p. 95.
13. Ravel, B. and Newville, M., J. Synchrotron Radiat.
,
2005, vol. 12, p. 537.
2. Tsodikov, M.V., Kugel, V.V., Yandieva, F.A., Morꢀ
dovin, V.P., Gekhman, A.E., and Moiseev, I.I., Pure
Appl. Chem., 2004, vol. 76, no. 9, p. 1769.
14. Ankudinov, A.L., Ravel, B., Rehr, J.J., and Conradꢀ
son, S.D., Phys. Rev. B: Condens. Matter, 1998,
vol. 58, p. 7565.
3. Tsodikov, M.V., Kugel’, V.Ya., Yandieva, F.A.,
Kliger, G.A., Glebov, L.S., Mikaya, A.I., Zaikin, V.G.,
Slivinskii, E.V., Plate, N.A., Gekhman, A.E., and Moiꢀ
seev, I.I., Kinet. Katal., 2004, vol. 45, no. 6, p. 904
15. Yandieva, F.A., Tsodikov, M.V., Chistyakov, A.V.,
Kugel’, V.Ya., Zubavichus, Ya.V., Veligzhanov, A.A.,
Kitaev, L.E., Yushchenko, V.V., Gekhman, A.E., and
Moiseev, I.I., Kinet. Katal., 2010, vol. 51, no. 4, p. 572
[Kinet. Catal. (Engl. Transl.), vol. 45, no. 6, p. 854].
[Kinet. Catal. (Engl. Transl.), vol. 51, no. 4, p. 548].
16. Tsodikov, M.V., Chistyakov, A.V., Yandieva, F.A.,
Gekhman, A.E., and Moiseev, I.I., Second German–
Russian Seminar on Catalysis, Munich, 2010, p. 22.
17. Revel, R., Bazin, D., Elkaim, E., Kihn, Y., and Dexꢀ
pert, H., J. Phys. Chem. B, 2000, vol. 104, p. 9828.
18. Davis, R.J., Landry, S.M., Horsley, J.A., and Boudart, M.,
4. RF Patent 2391133, 2010.
5. Tsodikov, M.V., Yandieva, F.A., Kugel, V.Ya., Chistyaꢀ
kov, A.V., Gekhman, A.E., and Moiseev, I.I., Catal.
Lett., 2008, vol. 121, nos. 3ꢀ4, p. 199.
6. Tsodikov, M.V., Tezisy III Rossiiskoi konf. “Aktual’nye
problemy neftekhimii” (Proc. III Russian Conf. on Curꢀ
rent Topics in Petroleum Chemistry), Zvenigorod,
Moscow oblast, 2009, p. 156.
Phys. Rev. B: Condens. Matter, 1989, vol. 39, p. 10 580.
19. Kozitsyna, N.Yu., Nefedov, S.E., Dobrokhotova, Zh.,
Ikorskii, V.N., Stolyarov, I.P., Vargaftik, M.N., and
Moiseev, I.I., Ross. Nanotekhnol., 2008, no. 3, p. 100.
7. Ichikawa, J.M., Catal. Lett., 1979, vol. 59, p. 127.
8. Maksimov, Yu.V., Matveev, V.V., Suzdalev, I.P.,
Slinkin, A.A., Fedorovskaya, E.A., Kutyreva, N.A.,
20. Waser, J., Levy, H.A., and Peterson, S.W., Acta Crystalꢀ
logr., 1953, no. 6, p. 661.
21. Khodyrev, Yu.P., Baranova, R.V., Imamov, R.M., and
Khomenko, T.I., and Kadushin, A.A., Kinet. Katal.
1989, vol. 30, p. 1198.
,
Semiletov, S.A., Izv. Akad. Nauk SSSR, Neorg. Mater.
,
1978, no. 14, p. 1645.
9. Tkachenko, O.P., Stakheev, A.Yu., Kustov, L.M.,
Mashkovsky, M., Berg, GrunertW., Kozitsyna, N.Yu., 22. Nowotny, H. and Bittner, H., Monatsh. Chem., 1950,
Nefedov, S.E., Dobrokhotova, Zh.V., Shilov, V.I.,
Vargaftik, M.N., and Moiseev, I.I., Catal. Letters, 2006,
vol. 112, p. 155.
no. 81, p. 679.
23. Hull, A.W., Phys. Rev., 1921, vol. 17, p. 571.
24. Neyman, K.M., Lim, K.H., Chen, Z.ꢀX., Moskaꢀ
leva, L.V., Bayer, A., Reindl, A., Borgmann, D.,
Denecke, R., Steinruck, H.ꢀP., and Rosch, N., Phys.
Chem. Chem. Phys., 2007, no. 9, p. 3470.
10. Mashkovsky, I.S., Tkachenko, O.P., Stakheev, A.Yu.,
Kozitsyna, N.Yu., Nefedov, S.E., Vargaftik, M.N.,
Cano, F.M., and Molenbroek, A.M., Proc. 13th Nordic
Symp. on Catalysis, Göteborg, 2008, p. 194.
25. Iwasa, N., Mayanagi, T., Masuda, S., and Takezawa, N.,
React. Kinet. Catal. Lett., 2000, vol. 69, no. 2, p. 350.
11. Kozitsyna, N.Yu., Nefedov, S.E., Dolgushin, F.M.,
Cherkashina, N.V., and Vargaftik, M.N., Moiseev. I.I., 26. Fukuhara, C., Kamata, Y., and Igarashi, A., Appl.
Inorg. Chim. Acta, 2006, vol. 359, p. 2072.
Catal., A, 2007, vol. 330, p. 108.
KINETICS AND CATALYSIS Vol. 52
No. 2
2011