Anisole hydrodeoxygenation over Ni-Cu bimetallic catalysts: The effect of Ni/Cu ratio on selectivity

Article abstract of DOI:10.1016/j.apcata.2013.10.046

The activity and selectivity of model NiCu-SiO₂ catalysts with the high metal loading (90%) and various Ni content was studied in anisole hydrotreatment at 280 C and a hydrogen pressure of 6 MPa in a batch reactor. To obtain the alloys with the homogeneous phase composition, the catalysts were prepared by simultaneous decomposition of metal salts with the subsequent stabilization with 10 wt% SiO₂. The composition of the Ni _yCu_1-y alloy surface was estimated from the combination of XPS and XRD data. On the basis of obtained kinetic data, a reaction scheme of anisole conversion was proposed. The scheme includes two parallel routes, one of which leads to C_arO bond cleavage with the formation of benzene, which is then converted to cyclohexane (HDO route), while the second route leads to hydrogenation of the aromatic ring of anisole with the formation of methoxycyclohexane and cyclohexanol (HYD route). The experimental dependence of anisole conversion was described by the first-order kinetics with respect to the reagents. The effect of the surface composition of the active phase Ni _yCu_1-y on the specific catalytic activity was examined. According to the proposed reaction scheme, the dependence of the first-order rate constants for both routes on the nickel content in the active component of the catalysts was determined. The selectivities of both reaction routes of anisole conversion (HDO and HYD) were found to be independent of the Ni content except for the nickel-rich and copper-rich sides. The specific catalytic activity in the HDO route rises with an increase in the nickel content in the whole range of Ni loading.

Full text of DOI:10.1016/j.apcata.2013.10.046

Applied Catalysis A: General 470 (2014) 261–270
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Anisole hydrodeoxygenation over Ni–Cu bimetallic catalysts: The
effect of Ni/Cu ratio on selectivity
Sofia A. Khromovaa,b,∗, Andrey A. Smirnova, Olga A. Bulavchenkoa,b, Andrey A. Saraeva,
Vasiliy V. Kaicheva,b, Sergey I. Reshetnikova, Vadim A. Yakovleva,b
a Boreskov Institute of Catalysis SB RAS, pr. Akad. Lavrenteva, 5, 630090 Novosibirsk, Russia
b Novosibirsk State University, Pirogova str.2, 630090 Novosibirsk, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 20 March 2013
Received in revised form 24 October 2013
Accepted 25 October 2013
Available online 12 November 2013
The activity and selectivity of model NiCu–SiO2 catalysts with the high metal loading (90%) and various
Ni content was studied in anisole hydrotreatment at 280 ◦C and a hydrogen pressure of 6 MPa in a batch
reactor. To obtain the alloys with the homogeneous phase composition, the catalysts were prepared
by simultaneous decomposition of metal salts with the subsequent stabilization with 10 wt% SiO2. The
composition of the NiyCu1−y alloy surface was estimated from the combination of XPS and XRD data. On
the basis of obtained kinetic data, a reaction scheme of anisole conversion was proposed. The scheme
includes two parallel routes, one of which leads to Car O bond cleavage with the formation of benzene,
which is then converted to cyclohexane (HDO route), while the second route leads to hydrogenation of the
aromatic ring of anisole with the formation of methoxycyclohexane and cyclohexanol (HYD route). The
experimental dependence of anisole conversion was described by the first-order kinetics with respect to
the reagents. The effect of the surface composition of the active phase NiyCu1−y on the specific catalytic
activity was examined. According to the proposed reaction scheme, the dependence of the first-order rate
constants for both routes on the nickel content in the active component of the catalysts was determined.
The selectivities of both reaction routes of anisole conversion (HDO and HYD) were found to be indepen-
dent of the Ni content except for the nickel-rich and copper-rich sides. The specific catalytic activity in
the HDO route rises with an increase in the nickel content in the whole range of Ni loading.
© 2013 Elsevier B.V. All rights reserved.
Keywords:
Ni–Cu alloys
Catalytic hydrotreatment
Anisole
Hydrogenation
Deoxygenation
1. Introduction
of the feed composition, which causes most of research groups to
use model compounds [6–8] or their mixtures [9–12] for investiga-
Fast pyrolysis, which allows one to produce liquid organic prod-
ucts with the high yield, is considered to be a promising approach
for the processing of lignocellulosic biomass. However, pyrolysis oil
(or bio-oil) can not be used as a fuel or fuel additive due to the high
oxygen content (20–40%) [1], which results in such undesirable
properties as low thermal and chemical stability, low heating value,
immiscibility with the fossil fuels and high acidity, and therefore,
requires an upgrade by oxygen removal and increasing of hydro-
gen content. Hydrodeoxygenation (HDO) process is one of the most
promising routes to upgrade pyrolysis bio-oils for producing liquid
transportation fuels [2]. It was shown previously that Ni–Cu cata-
lysts are active in hydrodeoxygenation of pyrolysis oil and its model
compounds [3–5]. The study of the kinetics and reaction mecha-
nisms of pyrolisys oil HDO is complicated due to the multiplicity
tions of hydrodeoxygenation kinetics. This approach allows one to
work at the reaction conditions similar to the conditions of real feed
hydrotreatment, as well as makes it possible to study the kinetics
and mechanisms of the reactions and thus to observe possible inter-
actions between the components of the model mixture. It is known
that the hydrotreatment reactions of oxygen-containing aromat-
ics proceeds through two independent routes, one of which leads
to the formation of aromatic hydrocarbons, and the other leads to
aliphatic oxygen-containing products [13–16].
The addition of Cu to Ni-based catalysts results in signifi-
cant changes in catalytic activity and selectivity compared to Ni
monometallic catalysts. For example, copper can prevent the for-
mation of carbon and the sintering of the active phase particles [17].
Lee et al. [18] studied the suppression of carbon deposition in the
methane reforming on NiCu/Al2O3 catalysts. They showed that the
addition of 1 wt% of Cu to a Ni/Al2O3 catalyst increases its activity
and stability. Besides, the addition of Cu to Ni catalysts enhances the
activity of some hydrogenation reactions, such as hydrogenation
of ethylene [19,20] and 1,3-butadiene [21]. The catalytic activity of
Ni–Cu alloys is a function of their composition. These changes in the
∗
Corresponding author at: Boreskov Institute of Catalysis SB RAS, pr. Akad.
Lavrenteva, 5, 630090 Novosibirsk, Russia. Tel.: +7 3833269453.
E-mail addresses: khromova@catalysis.ru, khromova@gmail.com
(S.A. Khromova).
0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.10.046

262
S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
catalytic properties are associated with changes in the electronic
properties of the active component in the alloy, with the influence
of the geometrical effect, or with a combination of these effects [22].
According to the electron-band theory, a group VIII metal, such as
nickel, has an incomplete d-band, and a metal such as copper (1B
group) has a filled d-band. Therefore, the substitution of Ni atoms
by Cu can result in filling of the d-zone and in significant changes
of the catalytic activity [23]. Thus, the purpose of the present study
was to investigate the effect of Ni content (the active component of
Ni–Cu bimetallic catalysts) on the selectivity of product formation
in anisole conversion; anisole was chosen as a model compound
for lignin depolymerization products in pyrolysis oil.
volume Vꢁ (from ultimate adsorption at a relative pressure of
P/P = 1), the micropore volume V , and the mean pore size.
␮
0
2.2.5. X-ray photoelectron spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS) was applied for the
chemical analysis of the spent catalyst surface. All the catalysts
were pretreated with 1 bar H2 at 350 ◦C for 30 min in the high-
pressure cell of the spectrometer. The XPS measurements were
performed on a SPECS’s photoelectron spectrometer equipped with
an XR-50 X-ray source with a double Al/Mg anode and a PHOIBOS-
150-MCD-9 hemispherical electron energy analyzer. Before the XPS
analysis, all the catalysts were additionally reduced in 1 bar H2 at
350 ◦C for 30 min in a special high-pressure cell connected directly
to the spectrometer [24]. The core-level spectra were obtained
using the non-monochromatic Mg K␣ radiation in the fixed pass
energy mode under ultrahigh vacuum conditions. Charging effects
were corrected for by setting the Cu2p3/2 peak to a position at
932.67 eV. Relative concentrations of elements were determined
from the integral intensities of the XPS peaks using the cross-
sections according to Scofield [25].
2. Experimental
2.1. Catalyst preparation
To obtain the alloys with the homogeneous phase composi-
tion, the catalysts were prepared by simultaneous decomposition
of metal salts with subsequent stabilization with 10 wt% SiO2.
The appropriate amounts of Ni(NO3)2·6Н2O and Cu(NO3)3·3Н2O
obtained from Reahim–Samara (Samara, Russian Federation) were
heated while stirring to form a viscous substance, then calcined
in air at 400 ◦C for 2 h and cooled down to the room temperature,
and grounded into powders. Then the samples were impregnated
with the appropriate amount of tetraethoxysilane Si(OEt)4 (Penta-
91, Ltd., Moscow, Russia) dissolved in ethanol according to their
water capacity, dried at 100 ◦C and calcined in air at 400 ◦C for
2 h. The amounts of the metal precursors were calculated to obtain
a required Ni–Cu composition. Before the reaction, the catalysts
were activated by reduction with hydrogen; the reduction temper-
ature (350 ◦C) was determined from the temperature programmed
reduction (TPR) procedure.
The metallic surface area of the reduced samples and the num-
ber of the active sites were determined by CO pulse chemisorption
measurements using a Chemosorb analyzer (“Modern laboratory
equipment”, Novosibirsk, Russia). CO uptakes were measured at
25 ◦C after prereduction at 350 ◦C.
2.3. Experimental setup
Catalytic hydrotreatment reactions of anisole were carried out
in a high-pressure setup (Autoclave Engineers, USA) in a sealed
300-ml stainless steel batch reactor (EZE Seal type). The reactor
was equipped with a magnetic stirrer, a thermocouple, a pressure
sensor, and a system for controlling stirring rate, temperature, and
pressure (Fig. 1). Before the reaction, the catalysts (1 g of a fine
powder) were activated directly in the reactor by the reduction
in a flow of 100% H2 (100 ml/min) for 1 h at 350 ◦C and 0.1 MPa.
After the activation, 50 ml of 6 wt% anisole (99%, ACROS Organ-
ics, Geel, Belgium) in hexadecane (99%, Sigma–Aldrich, Steinheim,
Germany) was placed into the reactor at 0.1 MPa and 25 ◦C with-
out access of air to prevent the catalyst oxidation. The reactions
of anisole were carried out at isothermal conditions (280 ◦C) and a
total pressure of 6 MPa. The hydrogen pressure was kept the same
in all the experiments. The reaction conditions were selected in
terms of the required hydrogen excess 1:10 (anisole, mol %: H2 mol
%) to provide the first-order kinetics with respect to the organic
reagents; the temperature of 280 ◦C is close to the typical tem-
perature of the hydrodeoxygenation. During the reaction, liquid
products were taken at certain intervals and analyzed. The stirring
rate was 2000 rpm, the catalysts were used in the form of fine pow-
der, the reaction was carried out in a tenfold excess of hydrogen
with respect to the anisole. Also the special tests were carried out
to ensure that the reaction proceeds in the kinetics region without
the diffusion limitations.
2.2. Catalyst characterization
2.2.1. Temperature programmed reduction
Catalyst samples (0.05 g) were placed in a U-tube quartz reactor
and treated in a reducing atmosphere (10 vol% of H2 balanced in Ar
at a flow rate of 20 ml/min) with a constant heating rate of about
8 ◦C/min up to 800 ◦C. The hydrogen concentration in the outlet
stream during reduction was measured with a thermal conductivity
detector (TCD).
2.2.2. Elemental analysis
The carbon content in the catalysts after the reaction was
determined using a Vario El III elemental analyzer, CHNS version
(Elementar, Germany).
2.2.3. X-ray diffraction
The phase composition of the spent catalysts was studied using
a D8 (Bruker, Germany) X-ray diffractometer using CuK␣ radiation
˚
ꢀ = 1.5418 A. The quantitative phase analysis was performed using
2.4. Product analysis
the Rietveld method. The error in the determination of Ni and Cu
˚
lattice parameter is 0.001 A.
Qualitative analysis of liquid products of the anisole HDO and
HYD was carried out using a Varian Saturn 2000 GC/MS spectrome-
ter equipped with an ion trap and an HP-5 quartz capillary column
(stationary phase: 5% phenyl–95% dimethylpolysiloxane, column
length 30 m, inner diameter 0.25 mm). Quantitative analysis of the
liquid products was performed using a Hromos GC 1000 chro-
matograph equipped with a Zebron ZB-35HT INFERNO capillary
column (stationary phase: 35% phenyl–65% dimethylpolysiloxane,
30 m × 0.32 mm × 0.25 ␮m). The injector and detector tempera-
tures were 280 and 260 ◦C, respectively. The following temperature
2.2.4. Texture characteristics
Texture characteristics of the reduced catalysts were measured
at the liquid nitrogen temperature using an ASAP-2400 automated
volumetric adsorption analyzer (Micromeritics Instrument. Corp.,
USA). Before the analysis, the samples were calcined at 150 ◦C and
pressure 0.13 Pa for 4 h. The analysis time was varied depending
on the particular sample. The resulting adsorption isotherms were
used to calculate the specific surface area ABET, the total pore

S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
263
Fig. 1. Set-up scheme. V1–V5 – valves; DPC – downstream pressure controller; MFC – mass flow controller; SP – safety pressure rupture disk; DCM – DC motor; R – reactor;
EF – electric furnace; M – manometer; BPR – back pressure regulator; CV1,CV2 – control valves.
program was used for analysis: 70 ◦C hold for 10 min, 15 ◦C/min for
4 min, 280 ◦C hold for 6 min. A 0.3 ␮l sample was injected into the
GC with a chromatographic syringe; argon was used as a carrier
gas. The reaction mixture components were identified by retention
times on the chromatogram, which were determined separately for
each component in the calibration process. The internal normal-
ization was used for the quantitative calculation of the component
concentrations in the mixture.
The conversion of anisole was determined as:
ꢀ
ꢁ
CA
CA0
X = 1 −
× 100%
where CA is the concentration of anisole and CA0 is the initial con-
centration of anisole.
C
j
ꢂ
The selectivity for product j was determined as: Sj =
×
C
i
i
100%where Cj is the concentration of product j, and Ci is the con-
centration of product i.
Fig. 2. H2-TPR profiles of the catalysts.
3. Results and discussion
3.1. Catalyst characterization results
behavior of the active phase precursors was studied to determine
the temperature of pretreatment in hydrogen before the reaction.
The reduction profiles of some samples are presented at Fig. 2.
TPR provides that reduction proceeds in one step, most of the
samples were fully reduced at temperatures below 400 ◦C. The
position of the maximum of hydrogen consumption peak for
bimetallic catalysts depends on the copper content, with higher
Cu contents leading to a lower reduction temperature. Robertson
et al. attributed more facile reduction of CuO in the bimetallic
Ni–Cu samples to molar free energy of reduction being lower for
this oxide (−100.65 kJ/mol at 25 ◦C) than for NiO (−12.31 kJ/mol at
25 ◦C) [28]. The TPR curve of monometallic Ni–SiO2 shows that its
reduction proceeds at higher temperatures with the maximum at
420 ◦C. The reduction of the bulk nickel oxide occurs at 400–450 ◦C
[29,30]. The addition of Cu to nickel catalysts promotes nickel oxide
reduction [31,32]. Our observation is in line with this proposal. The
positions of the corresponding peaks of hydrogen consumption for
the bimetallic samples are shifted to lower temperature (Fig. 2).
The phase composition of the active component of the catalysts
after the reaction was determined by XRD and the results are given
The catalysts represent the model systems with a high content
of the metallic active component and a small addition of the stabi-
lizing component – SiO2. The chemical composition of the catalysts
activated in hydrogen at 350 ◦C, metallic surface areas, surface areas
and pore volumes of the fresh reduced samples are given in Table 1.
The catalytic activity can be predicted from the adsorption proper-
ties of the catalyst: the higher is the number of adsorbed molecules,
the higher is the activity. In the case of bimetallic NiCu catalysts, the
heat of adsorption and the amount of adsorbed CO decrease with
an increase in the copper content in the catalyst [26]. This effect
can be explained by the changes in the ratio of bridged and linear
forms of adsorbed CO on the surface of the alloy [27].
The catalysts preparation method (the simultaneous decompo-
sition of metals salts with subsequent stabilization) was used to
obtain a homogeneous phase of the active component precursors.
After the reduction of the catalyst, this phase converts into a
NixCu1−x solid solution (or alloy), whose composition depends
on a nickel and copper content in the sample. The reduction
Table 1
Chemical composition, surface area and pore volume of the reduced fresh catalysts determined by the BET method.
Sample
Ni5Cu85–SiO2
Ni15Cu75–SiO2
Ni30Cu60–SiO2
Ni45Cu45–SiO2
Ni60Cu30–SiO2
Ni75Cu15–SiO2
Ni85Cu5–SiO2
Ni–SiO2
Ni (% wt.)
5
15
30
45
60
75
85
90
Cu (% wt.)
85
20
0.036
0.002
0.27
75
23
0.065
0.002
0.38
60
51
0.120
0.007
0.56
45
48
0.138
0.006
0.99
30
42
0.119
0.003
1.28
15
40
0.117
0.003
1.78
5
44
0.119
0.004
2.65
0
73
0.108
0.012
4.26
SBET (m2/g)
Vmesopores (cm2/g)
Vmicropores (cm2/g)
SCO (m2/g)

264
S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
Table 2
XRD data and phase composition of the active component in the spent samples.
Ni5Cu85–SiO2
Ni15Cu75–SiO2
Ni30Cu60–SiO2
Ni45Cu45–SiO2
Ni60Cu30–SiO2
Ni75Cu15–SiO2
Ni85Cu5–SiO2
Cua lattice parameter (Å)
Nia lattice parameter (Å)
Alloy composition
3.615
–
Cu0.95Ni0.05
3.615
3.556
Cu0.95Ni0.05
Ni0.64Cu0.36
3.615
3.554
Cu0.95Ni0.05
Ni0.67Cu0.33
3.614
3.552
Cu0.95Ni0.05
Ni0.68Cu0.32
3.616
3.543
Cu0.95Ni0.05
Ni0.78Cu0.22
3.616
3.534
Cu0.95Ni0.05
Ni0.85Cu0.15
–
3.536
Ni0.86Cu0.14
a
Alloy.
in Table 2. According to XRD data, the metallic reflexes are shifted
toward smaller angles as compared with the reference data [33]
which indicates the Ni–Cu solid solution (or alloy) formation. The
composition of the alloys was according to the literature data on
the Ni lattice parameter as a function of nickel content in Ni–Cu
solid solution [34].
the literature data [36–40], the Cu metal and Cu2O are character-
ized by the Cu2p3/2 binding energy in the range of 932.4–932.8 eV,
while the binding energy for CuO is between 933.6 and 934.6 eV.
Moreover, the Cu2p spectra of different Cu2+ compounds are char-
acterized by high-intensity shake-up satellites at approximately
9 eV higher binding energy than the main Cu2p3/2 and Cu2p1/2
peaks, whereas these satellites are absent in the spectra of metal-
lic copper and Cu2O. Therefore, the shape of the Cu2p spectra of
the catalysts under study (Fig. 4) indicates that copper mainly is
in the Cu0 or Cu1+ state. The Cu metal and Cu2O characterize close
values of the Cu2p3/2 binding energy; however, they are best distin-
guished from the X-ray-excited Cu LMM Auger spectra. Usually to
identify the chemical state of copper, the modify Auger parameter
(˛) is used, which is a sum of the Cu2p3/2 binding energy and the
Cu LMM kinetic energy. For all the catalyst, this parameter equals
1851.2–1851.3 eV (Table 3) that is typical for copper in the metallic
state. In the literature [36–40], this parameter for Cu0 and Cu1+ is
in the range of 1851.0–1851.4 and 1848.7–1849.3 eV, respectively.
Hence, the treatment in hydrogen leads to the full reduction of cop-
per to the metallic state. Some additional peaks in the spectra of
Ni15Cu75–SiO2 are due to the differential charge effect.
The estimation of NixCu1−x alloys composition was carried out
using Vegard’s rule according to which the cell parameters of the
substitutional solid solution are linear functions of its components
mole fractions [35]. All the samples with the Ni content in the range
of 15–85 wt% after the reduction contain two types of solid solu-
tions, one of which has a constant composition Cu0.95Ni0.05 and the
other NixCu1−x, with variable composition, is enriched with nickel.
The amounts of Cu0.95Ni0.05 and NixCu1−x phases depend on the
composition of the catalyst (Fig. 3). The greater the amount of Ni in
the catalyst, the greater its content in the second type solid solu-
tion. The dependence of the percent of the Ni0.05Cu0.95 and NixCu1−x
phases on the Ni content in the catalyst indicates that, with increas-
ing nickel content in the catalyst, the amount of the Cu-enriched
phase decreases and the amount of Ni-enriched phase becomes
greater.
The [Ni]/[Cu], [Ni]/[Si], and [Cu]/[Si] atomic ratios and the chem-
ical composition of the active component of the catalysts under
study determined by XPS data are presented in Table 3. The atomic
ratio [Ni]/[Si] varies in the narrow range of 0.095–0.17 for all the
catalysts. In contrast, the atomic ratio [Cu]/[Si] changes signifi-
cantly. For the catalysts with a low loading of nickel (Ni5Cu85,
Ni15Cu75, and Ni30Cu60) the atomic ratio [Cu]/[Si] varies in the
range of 0.67–0.80, whereas for the catalysts with a high load-
ing of nickel (Ni75Cu15 and Ni85Cu5) this values is approximately
0.03–0.06 only. It indicates that a mean average size of the nickel
particles increases with loading of nickel while copper localizes
mainly on the surface of these particles. Of course these data do
not exclude the formation of Ni–Cu alloys.
The analysis of the Ni2p spectra (Fig. 4) points out that Ni in
all the catalyst, at least in surface and subsurface layers, mainly
is in the metallic state as well. The spectra consist of two sharp
peaks which correspond to the Ni2p3/2–Ni2p1/2 spin-orbit doublet
and low-intensity satellites which can be associated with energy
loss peaks due to Plasmon excitation in metal [41–43]. The Ni2p3/2
binding energy is in the range of 852.4–853.0 eV that agrees with
the literature data [41–43] where for the Ni metal and NiO the
Ni2p3/2 binding energy refers in the range of 852.4–853.0 and
854.5–854.6 eV, respectively. The scattering in the Ni2p3/2 binding
energy (Table 3) can be associated with the formation of Ni–Cu solid
solutions or alloys. Recently it was found that in the Ni1−xCux alloys
the Ni2p3/2 binding energy decreased with the increasing copper
content [44]. The shift of the Ni2p3/2 binding energy achieved 0.5 eV
when the concentration of copper changed between 0 and 90%.
A similar shift is observed in our experiments that points to the
formation of Ni–Cu alloys in the catalysts under study. All the
Si2p spectra (not shown) consist of a symmetric single peak at
104.1–104.8 eV, which can be assigned to silicon in the Si4+ state.
The XPS method gives information on the total composition
of the surface, whereas, according to the X-ray diffraction data
given above, most of the samples within the active catalyst com-
ponent have two phases, one of which is enriched with copper
(Ni0.05Cu0.95). In this connection, we estimated the composition of
the surface for the NixCu1−x phase, which is active in hydrodeoxy-
genation process.
Fig. 4 presents the Cu2p, Ni2p, and Si2p core-level spectra of the
reduced catalysts under study. It is well known that the analysis
of the Cu2p spectrum allows unambiguously to distinguish Cu2+
cations from copper in the metallic state and Cu1+. According to
The composition of the surface of the NixCu1−x component was
calculated using the formula:
n − ϕ1 · 0.05
y =
ϕ2
where y is the Ni content on the surface of the active component of
the NiyCu1−y composition, n is the Ni content according to the XPS
data, ϕ1 is the Ni0.05Cu0.95 phase content, ϕ2 is the NixCu1−x phase
content.
Fig. 3. Ni0.05Cu0.95 and NixCu1−x phases percentage (calculated using the Rietveld
method) vs. Ni content in metallic active component of the catalysts.

S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
265
Table 3
XPS data. Si2p, O1s, Ni2p3/2, modified Cu LMM Ager parameter Si2p, O1s, Ni2p3/2, and modified Cu LMM Ager parameter, atomic ratios, and surface chemical composition of
the spent catalysts.
Catalysts
Si2p
Ni2p3/2 (%)
O1s (%)
˛
[Ni]/[Cu]
[Ni]/[Si]
[Cu]/[Si]
Chemical composition
Ni5Cu85–SiO2
Ni15Cu75–SiO2
Ni30Cu60–SiO2
Ni45Cu45–SiO2
Ni60Cu30–SiO2
Ni75Cu15–SiO2
Ni85Cu5–SiO2
104.13
104.21
104.61
104.62
104.36
105.56
104.78
852.43
851.20852.67
852.62
852.77
852.79
533.65
533.64
534.04
534.20
533.75
535.06
534.21
1851.2
1851.21851.3
1851.3
1851.3
1851.2
0.12
0.22
0.16
0.45
1.1
0.08
0.16
0.12
0.11
0.17
0.095
0.16
0.67
0.74
0.80
0.24
0.15
0.06
0.03
Ni0.11Cu0.89
Ni0.18Cu0.82
Ni0.14Cu0.86
Ni0.31Cu0.69
Ni0.53Cu0.47
Ni0.61Cu0.39
Ni0.84Cu0.16
852.99
853.03
1851.2
1851.2
1.6
5.3
Fig. 4. Cu2p and Ni2p core-level spectra of the catalysts. All spectra are normalized on the intensity of the Si2p core-level spectra.
The content of each phase for all the catalysts was estimated
from the data obtained by X-ray analysis (Fig. 3). Table 4 presents
data on the composition of the surface of the NixCu1−x phase. It
can be seen that, with increasing nickel content in the sample,
its contents on the surface and in the volume of metal particles
increase as well. The particle surface of the active component is
enriched with copper. This is due to the fact that surfaces of most
alloys are enriched by the specific metal (segregation effect) due
to the desire of the system to minimize the total free energy. The
surface is enriched with the component which has lower surface
free energy. Therefore, the surface is enriched with copper, which
has lower surface free energy (ꢂCus = 1.670 J/m2, ꢂNis = 1.850 J/m2
[45]). The Ni content in the samples on the surface and in the vol-
ume increases, when its content in the catalysts is increased.
3.2. The effect of Ni content on activity and selectivity of
NiCu–SiO2 catalysts in anisole hydrotreatment
The main products obtained during the hydrotreatment of
anisole are methoxycyclohexane and cyclohexane; the inter-
mediate components are cyclohexanol and benzene. The minor
intermediate, which is formed from anisole and converted sequen-
tially into cyclohexane, is benzene. The concentration variations
of the species in the reaction mixture as a function of reaction
time are shown in Fig. 5. The proposed reaction scheme of anisole
conversion, which includes two parallel routes, one of which gives
the aromatic product and the other gives the aliphatic product, is
given in Fig. 6. The first pathway leads to Car O bond cleavage with
the formation of benzene, which is then converted to cyclohexane
Table 4
The calculated surface composition of NiyCu1−y phases. CHNS analysis results. The carbon content in the spent catalysts.a
Ni5Cu85–SiO2
Ni15Cu75–SiO2
Ni30Cu60–SiO2
Ni45Cu45–SiO2
Ni60Cu30–SiO2
Ni75Cu15–SiO2
Ni85Cu5–SiO2
Ni–SiO2
Ni (% at.)
Cu (% at.)
C (% wt.)
11
89
0.31
45
55
0.32
20
80
0.49
36
64
0.44
56
44
0.45
62
38
0.52
84
16
0.69
100
0
0.83
a
The samples were washed with acetone and dried in air before the analysis. Catalytic reaction conditions: T = 280 ◦C, p = 6 MPa H2, 6% of anisole in hexadecane, and
mcatalyst = 1 g.

266
S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
Fig. 5. Concentration profiles of the reactant and products as a function of reaction time. Reaction conditions: 50 ml of 6 wt% anisole in hexadecane, 280 ◦C, 6 MPa, mcatalyst = 1 g.
(HDO route). The second pathway is hydrogenation of the aromatic
ring of anisole with the formation of methoxycyclohexane and
cyclohexanol (HYD route).
The experimental dependence of anisole conversion was
described by the first-order kinetics with respect to the reagents.
The kinetic modeling was based on the proposed reaction network;
the calculations were performed using the following simplified
model:
ꢂ
dC
i
= Ri, where Ri =
jrj and rj = kjCj, with the boundary con-
dt
ditions: t = 0, C1 = 1 (anisole), Ci = 0, i = 2,. . .,5. Here kj is first-order
rate constant, Cj is molar fraction of each component, and rj is
reaction rate.
Fig. 6. Scheme of anisole conversion pathways (ki – first-order rate constants,
k1 = kHDO, k2 + k3 = kHYD, k0 = kHDO + kHYD), the reaction was carried out at T = 280 ◦C,
p = 6 MPa H2, 6% of anisole in hexadecane, and mcatalyst = 1 g.

S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
267
Fig. 7. Selectivity of anisole hydrotreatment products formation vs. conversion for some of the catalyst, symbols – experimental results, line – modeling. The reaction was
carried out at T = 280 ◦C, p = 6 MPa H2, 6% of anisole in hexadecane, and mcatalyst = 1 g.
Fig. 8. (a) The surface area of the metallic active component determined by CO pulse chemisorption vs. Ni content in the active component (NiCu) of NiCu–SiO2 catalyst; (b)
the initial values of the rate constants of anisole conversion k0 at T = 280 ◦C and p = 6 MPa H2 vs. Ni content in the active component (NiCu), k0 = k1 + k2 + k3.
The system of equations was solved using the
Runge–Kutta–Merson numerical integration algorithm; the
modeling was performed by minimizing the residual sum of
squares between experimental and calculated results. The results
of modeling for some of the catalysts, representing the selectivity
as a function of conversion, are given in Fig. 7.
On the basis of the results of kinetic modeling, the presence
of the reaction pathway corresponding to the direct conversion
Fig. 9. (a) The initial values of the rate constants of anisole conversion vs. nickel surface area, determined by CO pulse chemisorption, (b) the dependence of NiCu–SiO2
catalysts specific activity on the content of nickel on the surface of NiyCu1−y. The reaction was carried out at T = 280 ◦C, p = 6 MPa H2, 6% of anisole in hexadecane, and
mcatalyst = 1 g.

268
S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
of anisole into cyclohexanole (k2, Fig. 6) was identified. The pro-
posed reaction scheme is in the agreement with the data presented
in the majority of studies on the kinetics and mechanisms of
the hydrotreatment of oxygen-containing aromatics, according to
which this process involves two independent routes, one of which
leads to the formation of aromatic hydrocarbons, and the second
leads to aliphatic oxygen-containing products. For example, in the
hydrogenation of phenol on a sulfided CoMo/Al2O3 catalyst, the
first reaction pathway is the cleavage of Car O bond and the sec-
ond route leads to the hydrogenation of the phenol aromatic ring
to form an intermediate – cyclohexanol, which is then rapidly con-
verted to cyclohexene or cyclohexane [13]. Gevert et al. [14] studied
the kinetics of 2,6-dimethylphenol and 3,5-dimethylphenol hydro-
genation and hydrodeoxygenation in the presence of a sulfided
CoMo/Al2O3 catalyst at a hydrogen pressure of 5 MPa and at tem-
peratures of 240–330 ◦C. Their results also indicate the presence
of two parallel reaction routes, one of which is the direct deoxy-
genation without hydrogenation of the aromatic ring, while the
second is the hydrogenation of the benzene ring without oxygen
removal. The authors attribute this effect to the involvement of two
types of active sites in the process [16]. Those centers, on which
the hydrogenation of aromatic ring occurs, possess the electron-
acceptor properties; the reactant adsorption on these sites is planar
– by ␲-system of the ring. The centers of the second type have the
electron-donor character, and the reactant is adsorbed vertically,
through a heteroatom, which results in the cleavage of the Car
bond.
O
The results of the elemental analysis of the catalysts after the
reaction, which indicate the formation of carbon deposits on the
metal surface during the reaction, are given in Table 4.The higher
the Ni content in the catalyst the higher the mass of the carbon
deposit. This may lead to the faster deactivation of the samples
with the high nickel loading as a result of the active sites blocking
by carbon.
The increase in the nickel content in the active component
(NiCu) of the NiCu–SiO2 catalyst results in the higher specific sur-
face area of the metal per gram of the catalyst (Fig. 8a). The same
trend is observed for the dependence of the initial rate constants of
anisole conversion k0 for the bimetallic samples (Fig. 8b).
The dependence of the initial reaction rate constants on the spe-
cific surface area of nickel is linear for bimetallic catalysts (Fig. 9a),
while the value for Ni–SiO2 is lower, which demonstrates the dif-
ference between the activities of bi- and of monometallic catalysts.
In addition, this relationship shows a correlation of the catalytic
properties of NiCu systems with the number of active sites (sur-
face nickel atoms). It is essential that the alloying may have an
impact not only on the major reactions, but also on side reactions.
For example, according to van Barneveld’s and Ponec’s observations
[46], the activity of Cu–Ni alloy per surface unit in the hydrogena-
tion of benzene is lower than that of pure Ni in the temperature
range 200–220 ◦C. However, at temperatures above 220 ◦C, the
alloy is more active than Ni, because the side reactions of CH4 and
other non-cyclic hydrocarbons and carbon deposition, which occur
on Ni, may result in the deactivation of Ni catalysts. The specific
catalytic activity, which is related to the unit surface area of nickel,
also increases with increasing the nickel content on the surface of
the active phase NiyCu1−y up to 40–60 at % Ni. The further increase
in the nickel content, however, leads to a drop in activity of the
bimetallic samples (Fig. 9b).
Fig. 10. NiCu alloys catalytic activity in hydrogenation reactions as a function of
Ni content: (a) specific activity of the alloy in cyclopropane hydrogenation at 90 ◦
C
(adapted from [J.M. Beelen, V. Ponec, W.M.H. Sachtler, J. Catal. 28 (1973) 376–380]);
(b) initial activity of the alloy in ethylene hydrogenation: open symbols are for sam-
ples pretreated by heating the alloy in hydrogen at 300 ◦C, solid symbols are for
samples homogenized by heating to 500 ◦C in hydrogen, triangles are for films in
which the copper was deposited on the nickel, the open circles are for nickel evapo-
rated on top of the copper, the open square represents a film in which the copper and
nickel were deposited simultaneously (adapted from [J.S. Campbell, P.H. Emmett,
J. Catal. 7 (1967) 252–262]); (c) ethylene hydrogenation on NiCu films at −73◦C
(adapted from [M.K. Gharpurey, P.H. Emmett, J. Phys. Chem. 65 (1961) 1182–1184]).
Such changes in the catalytic properties can be caused by
changes in the electronic properties of the alloy particles. According
to the electron-band theory, a Group VIII metal, such as nickel, has
free d-orbitals delocalized in the conductivity band, whereas metals
from Group 1B, for example copper, have free d-electrons. There-
fore, the significant changes in the catalytic activity of Ni-based
alloys (solid solutions with varying composition) due to the filling

S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
269
Fig. 11. Selectivities of CAr O bond hydrogenolysis and anisole aromatic ring hydrogenation over NiCu–SiO2 catalysts at T = 280 ◦C and p = 6 MPa H2 as a function of the
catalyst composition (a) and the composition of alloy surface (b).
of the d-zone because of interaction with a Group 1B metal can be
expected. When alloying nickel, palladium, or platinum (0.4–0.6
holes per atom) with a Group 1B metal, such as copper, the d-band
is completely filled when the concentration of the Group 1B metal
is in the range 40–60 at %. At or near this composition, the catalytic
activity of alloys is expected to fall [22]. Fig. 10 demonstrates the
literature data on the activity of Ni–Cu alloys as a function of their
composition in various reactions.
The data obtained in the present study on the catalytic activity
of Ni–Cu alloys of various composition in anisole hydrotreatment
at T = 280 ◦C and p = 6 MPa H2 is in a good agreement with the
literary data presented in Fig. 10, from which it can be concluded
that the main factor affecting the catalytic activity of the alloy in
this process is the electronic structure of the active site, rather
than a geometric factor.
The hydrogenolysis of the C C bond in the ethane molecule
has been studied by Sinfelt et al. over a series of Ni–Cu catalysts
prepared by coprecipitation [34]. They observed the significant
suppression of the catalytic activity in hydrogenolysis with the
addition of copper to nickel in the whole range of Ni content in
the alloy. Satterfield attributed this decrease in activity to the geo-
metric effect, i.e. when a group of metal atoms on the catalyst
surface acts as an active site for adsorption and reaction (ensemble
effect) [46]. Gandarias et al. [47] also studied a C C bond cleav-
age, but in glycerol. To check whether this reaction is sensitive
to the size of the Ni ensemble, they analyzed the amount of C < 3
products of glycerol conversion. They found that concentration of
C < 3 products exponentially decreased with reducing the Ni sur-
face proportion, which indicates that the C C bond cleavage in
glycerol requires a whole ensemble of contiguous Ni atoms. The
size of the Ni ensemble in a Ni–Cu alloy is lower than in pure
ring is considered a structure-insensitive reaction [46], it can be
expected that the activity of the nickel–copper alloys in this reac-
tion is constant over the entire range of nickel concentrations in
the alloy. For example, Takeuchi et al. reported that the catalytic
activity of granular Ni–Cu alloys in the hydrogenation of benzene
at 190 ◦C did not depend on the alloy composition [48]. Another
such example was described by Van der Plank and Sachtler [49] for
activity of Ni–Cu alloy films at 190 ◦C. On the contrary, the entirely
different results were obtained by Cadenhead et al. [50]. They stud-
ied the hydrogenation of benzene at temperatures 98–182 ◦C over
series of copper–nickel alloys using a microcatalytic method and
found that the specific activities decrease with increasing the cop-
per content and then fall sharply at a copper content more than
80%. The same results were reported by Hall and Emmett [51]. As
it was noted above, nickel–copper catalysts appeared to be active
in the hydrotreamnent of pyrolysis oil for production of motor
fuel components. The upgrading of crude pyrolysis oil involves
the hydrogenation processes (including hydrogenation of aromatic
compounds) and the removal of oxygen, i.e. hydrogenolysis of the
C
O bond. Unlike the hydrogenation of aromatic and aliphatic com-
pounds, there is a lack of published data on the hydrogenolysis of
the C O bond over the nickel–copper alloys. In this paper, the effect
of the catalyst composition on the activity of Ni–Cu alloys in the
hydrogenolysis of a Car O bond of a model compound – anisole –
is reported.
As it follows from Fig. 11, the selectivities of two reaction routes
of anisole conversion (hydrogenolysis (HDO) of the Car O bond and
hydrogenation (HYD) of anisole aromatic ring) seem to be indepen-
dent of the Ni content except for the nickel-rich and copper-rich
sides. For these two extreme ranges of composition, there is an
increase in HDO selectivity and a decrease in HYD selectivity, espe-
cially significant for the monometallic nickel sample, which is in
agreement with the trend observed by Cadenhead et al. [50].
nickel, which results in a decline of the catalyst activity in C
C
hydrogenolysis. Since the hydrogenation of the benzene aromatic
Fig. 12. NiCu–SiO2 catalysts specific activity in anisole HDO (hydrogenolysis of CAr O bond) vs. content of nickel in the active phase of the catalyst (a) and on the surface of
NiyCu1−y alloy (b). The reaction was carried out at T = 280 ◦C, p = 6 MPa H2, 6% of anisole in hexadecane, and mcatalyst = 1 g.

270
S.A. Khromova et al. / Applied Catalysis A: General 470 (2014) 261–270
Fig. 12 shows the correlations between the specific activities
References
of different NiCu–SiO2 samples in the Car O bond cleavage in
anisole, which proceeds via the formation of benzene as a minor
intermediate with its subsequent conversion into cyclohexane. In
both cases (activity as a function of nickel content in the active
metallic phase and as a function of the composition of NiyCu1−y
alloy surface calculated from combination of XPS and XRD data),
the specific catalytic activity in the HDO route rises with the
increase of nickel content in the whole range of Ni concentrations.
This is the main difference in behavior between the HDO and HYD
activities of NiCu–SiO2 catalysts. The total rate constant of anisole
conversion k0 are lower for the monometallic nickel sample that
it is expected from the metallic surface area and Ni content in this
sample (Fig. 8) compared to the bimetallic catalysts.
[1] F. de Miguel Mercader, M.J. Groeneveld, S.R.A. Kersten, R.H. Venderbosch, J.A.
Hogendoorn, Fuel 89 (2010) 2829–2837.
[2] Z.H. He, X. Wang, Catal. Sustain. Energy 1 (2012) 28–52.
[3] V.A. Yakovlev, S.A. Khromova, O.V. Sherstyuk, V.O. Dundich, D.Y. Ermakov, V.M.
Novopashina, M.Y. Lebedev, O. Bulavchenko, V.N. Parmon, Catal. Today 144
(2009) 362–366.
[4] A.R. Ardiyanti, S.A. Khromova, R.H. Venderbosch, V.A. Yakovlev, H.J. Heeres,
Appl. Catal. B: Environ. 117/118 (2012) 105–117.
[5] A.R. Ardiyanti, S.A. Khromova, R.H. Venderbosch, V.A. Yakovlev, I.V. Meliꢀan-
Cabrera, H.J. Heeres, Appl. Catal. A: General 449 (2012) 121–130.
[6] C.-L. Lee, D.F. Ollis, J. Catal. 87 (1984) 325–331.
[7] O.I. Senol, T.R. Viljava, A.O.I. Krause, Catal. Today 100 (2005) 331–335.
[8] A.Y. Bunch, X. Wang, U.S. Ozkan, J. Mol. Catal. A: Chem. 270 (2007) 264–272.
[9] C.N. Satterfield, S.H. Yang, J. Catal. 81 (1983) 335–346.
[10] C.-L. Lee, D.F. Ollis, J. Catal. 87 (1984) 332–338.
[11] E.O. Odebunmi, D.F. Ollis, J. Catal. 80 (1983) 65–75.
[12] E.O. Odebunmi, D.F. Ollis, J. Catal. 80 (1983) 76–89.
[13] F.E. Massoth, P. Politzer, M.C. Concha, J.S. Murray, J. Jakowski, J. Simons, J. Phys.
Chem. B 110 (2006) 14283–14291.
4. Conclusions
[14] S.B. Gevert, M. Eriksson, P. Eriksson, F.E. Massoth, Appl. Catal. A 117 (1994)
151–162.
[15] J.B. Bredenberg, M. Huuska, P. Toropainen, J. Catal. 120 (1989) 401–408.
[16] B.S. Gevert, J.-E. Otterstedt, F.E. Massoth, Appl. Catal. 31 (1987) 119–131.
[17] T.V. Reshetenko, L.B. Avdeeva, Z.R. Ismagilov, A.L. Chuvilin, V.A. Ushakov, Appl.
Catal. A: General 247 (2003) 51–63.
On the basis of the present study of the activity and selectiv-
ity of NiCu–SiO2 catalysts with various compositions in anisole
hydrotreatment at 280 ◦C and a hydrogen pressure of 6 MPa, the
following conclusions can be drawn:
[18] J.-H. Lee, E.-G. Lee, O.-S. Joo, K.-D. Jung, Appl. Catal. A: General 269 (2004) 1–6.
[19] M.K. Gharpurey, P.H. Emmett, J. Phys. Chem. 65 (1961) 1182–1184.
[20] R.J. Best, W.W. Russell, J. Phys. Chem. 76 (1954) 838–842.
[21] M. Kang, M.W. Song, T.W. Kim, K.L. Kim, Can. J. Chem. Eng. 80 (2002) 63–70.
[22] J.R. Anderson, Structure of Metallic Catalysts, Acad. Press, N.Y., 1975.
[23] Y. Wang, T. Gao, React. Kinet. Catal. Lett. 70 (2000) 91–96.
[24] M.V. Bykova, D.Y. Ermakov, V.V. Kaichev, O.A. Bulavchenko, A.A. Saraev, M.Y.
Lebedev, V.A. Yakovlev, Appl. Catal. B 113/114 (2012) 296–307.
[25] J.H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8 (1976) 129–137.
[26] L.J. Gijzeman, C.M.A.M. Mesters, F. Labohm, J.W. Geus, J. Mol. Catal. 25 (1984)
193–206.
[27] E. Asedegbega-Nieto, A. Guerrero-Ruíz, I. Rodríguez-Ramos, Thermochim. Acta
434 (2005) 113–118.
[28] S.D. Robertson, B.D. McNicol, J.H. De Baas, S.C. Kloet, J.W. Jenkins, J. Catal. 37
(1975) 424–431.
[29] B. Mile, D. Stirling, M.A. Zammit, A. Lovell, M. Webb, J. Catal. 114 (1988)
217–229.
[30] O. Clause, L. Bonneviot, M. Che, J. Catal. 138 (1992) 195–205.
[31] L. De Rogatis, T. Montini, B. Lorenzut, P. Fornasiero, Energy Environ. Sci. 1 (2008)
501–509.
1. The reaction scheme of anisole conversion was proposed, which
includes two parallel routes. The first pathway leads to the Car
O
bond cleavage with the formation of benzene, which is then con-
verted to cyclohexane (HDO route). The second pathway leads
to hydrogenation of the aromatic ring of anisole with the for-
mation of methoxycyclohexane and cyclohexanol (HYD route).
On the basis of the results of kinetic modeling, the presence of
the reaction pathway corresponding to the direct conversion of
anisole into cyclohexanole was identified.
2. The increase of the nickel content in the active component (NiCu)
of bimetallic NiCu–SiO2 catalysts results in the increase of the
initial rate constants of anisole conversion K0. The dependence
of the initial reaction rate constants on the specific surface area
of nickel is linear for bimetallic catalysts. The initial reaction rate
constant for Ni–SiO2 is lower than for bi-metallic catalysts, and
this demonstrates the difference in their activities.
[32] J.C. Rodríguez, A.J. Marchi, A. Borgna, E. Romeo, A. Monzón, in: G.W. Roberts,
J.J. Spivey, B.H. Davis (Eds.), Studies in Surface Science and Catalysis, Elsevier,
Lexington, KY, U.S.A., 2001, pp. 37–44.
3. According to XRD data, all the samples with Ni content in the
range of 15–85 wt% after the reduction contain two types of solid
solutions, one of which has a constant composition Cu0.95Ni0.05
and the second solution NiyCu1−y, with variable composition, is
enriched with nickel. The results of XPS analysis show that the
surface composition of the active component particles differs
from their bulk composition. The composition of the NiyCu1−y
alloy surface was estimated from the combination of XPS and
XRD data.
4. The specific catalytic activity, which is related to the unit surface
area of nickel, increases with an increase in the nickel content on
the surface of the active phase NiyCu1−y with the drop in activity
at 40–60 at % of Ni.
5. The selectivities of two reaction routes of anisole conversion
(hydrogenolysis (HDO) of the Car O bond and hydrogenation
(HYD) of anisole aromatic ring) were found to be independent of
the Ni content except for the nickel-rich and copper-rich sides.
The specific catalytic activity in the HDO route rises with the
increase in the nickel content over the whole range of Ni loading.
[33] PDF. # 040850.
[34] J.H. Sinfelt, J.L. Carter, D.J.C. Yates, J. Catal. 24 (1972) 283–296.
[35] A. West, Solid State Chemistry and its Application, John Wiley & Sons, New
York, 1985, pp. 742.
[36] B.R. Strohmeier, D.E. Levden, R.S. Field, D.M. Hercules, J. Catal. 94 (1985)
514–530.
[37] M.C. Biesinger, L.W.M. Lau, A.R. Gerson, R.St.C. Smart, Appl. Surf. Sci. 257 (2010)
887–898.
[38] S. Poulston, P.M. Parlett, P. Stone, M. Bowker, Surf. Interface Anal. 24 (1996)
811–820.
[39] J. Batista, A. Pintar, D. Mandrino, M. Jenko, V. Martin, Appl. Catal. A 206 (2001)
113–124.
[40] V.I. Bukhtiyarov, V.V. Kaichev, I.P. Prosvirin, Top. Catal. 32 (2005) 3–15;
M. Richter, M.J.G. Fait, R. Eckelt, M. Schneider, J. Radnik, D. Heidemann, R. Fricke,
J. Catal. 245 (2007) 11–24.
[41] C.P. Li, A. Proctor, D.M. Hercules, Appl. Spectrosc. 38 (1984) 880–886.
[42] A.F. Carley, S.D. Jackson, J.N. O’Shea, M.W. Roberts, Surf. Sci. 440 (1999)
L868–L874.
[43] V.V. Kaichev, A.Y. Gladky, I.P. Prosvirin, A.A. Saraev, M. Hävecker, A. Knop-
Gericke, R. Schlögl, V.I. Bukhtiyarov, Surf. Sci. 609 (2013) 113–118.
[44] P.F. Barbieri, A. de Siervo, M.F. Carszzolle, R. Landers, G.G. Kleiman, J. Electron
Spectrosc. Relat. Phenom. 135 (2004) 113–118.
[45] G.A. Somorjai, Y. Li, Introduction to Surface Chemistry and Catalysis, A Wiley-
Interscience Publication John Wiley & Sons. inc, New York, 1994, pp. 667.
[46] C.H. Satterfield, Heterogeneous Catalysis in Practice, M.: Mir, 1984, pp. 169.
[47] I. Gandarias, J. Requies, P.L. Arias, U. Armbruster, A. Martin, J. Catal. 290 (2012)
79–89.
[48] T. Takeuchi, M. Sakaguchi, I. Miyoshi, T. Takabatake, Bull. Chem. SOC. Jpn. 35
(1982) 1390.
[49] P. van der Plank, W.M.H. Sachtler, J. Catal. 7 (1967) 300–303.
[50] D.A. Cadenhead, N.G. Masse, J. Phys. Chem. 70 (1966) 3558.
[51] W.K. Hall, P.H. Emmett, J. Phys. Chem. 62 (1958) 816.
Acknowledgments
The Ministry of education and science of Russia (contracts No.
16.740.11.0683 and 16.526.12.6012) is acknowledged for financial
support.