Catalysis Today 299 (2018) 303–316

Contents lists available at ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

Intermetallides as the catalysts for carbon dioxide reforming of methane MARK

a,⁎,1 a a b
Larisa A. Arkatova , Nikolai G. Kasatsky , Yury M. Maximov , Oleg V. Pakhnutov ,
Alexander N. Shmakovc
a
Department of Structural Macrokinetics, Tomsk Scientific Center, Siberian Branch, Russian Academy of Sciences, Academichesky Avenue 10/3, Tomsk 634021, Russia
b
Department of Chemistry, Tomsk State University, Lenin Avenue. 36, Tomsk 634050, Russia,
c
Boreskov Institute of Catalysis, Akademika Lavrentieva Avenue 5, Novosibirsk 630090, Russia

A R T I C L E I N F O A B S T R A C T

Keywords: Intermetallic catalysts for dry reforming of methane (DRM) based on Ni3Al with low content of Pt and Ru have
Dry reforming of methane been developed. Self-propagating high-temperature synthesis (SHS) was used as a method of catalyst synthesis.
Intermetallic Ion implantation was used to change the physical, chemical and catalytic properties of Ni3Al as a matrix for this
Self-propagating high-temperature synthesis new type of catalysts for DRM. Ions of Pt and Ru were accelerated in an electrical field and impacted into a solid
Ion implantation
Ni3Al. The catalytic performances were evaluated between 600 °C and 900 °C at atmospheric pressure. The feed
Platinum
Ruthenium
was CH4/CO2/He = 20/20/60 vol.% mixture. Particle size and chemical evolution of catalysts were studied by
XRD (in situ and ex situ), SEM, EDS, HRTEM + EDS and XPS. The active components were shown to be primarily
dispersed in the nearsurface layer of Ni3Al support as nanoparticles of size 5–10 nm, which were distributed
homogeneously or heterogeneously, depending on the catalyst composition. Spinel structure of some catalysts is
resistant to carbonization and provides high catalyst stability in DRM. Modification of catalyst by ion im-
plantation had several positive impacts: 1) high catalytic activity and stability in DRM, 2) reducing of carbon
deposits, 3) Pt and Ru prevent Ni phase sintering by avoiding particle coalescence.

1. Introduction carbon dioxide and 14% for methane [5]; hence, the output of these
gases must be reduced. It can be done using CO2 and CH4 as chemical
Natural gas conversion is the primary industrial method for the feedstock [4] for various chemical processes. Catalytic reforming of
production of synthesis gas or briefly, syngas, the precursor for che- methane with carbon dioxide is a highly promising reaction that is
micals and fuels. There are three approaches to such conversion: really important from an environmental and industrial perspective [4].
From an environmental point of view, this reaction involves two
Steam reforming (SRM) CH4 + H2O = CO + 3H2, ΔH0 =
greenhouse gases and therefore could provide their utilization. From an
206.3 kJ mol−1 (1)
industrial perspective, CO2-CH4 reforming provides an alternative route
Partial oxidation (POM) CH4 + 0.5O2 = CO + 2H2, ΔH0 = for syngas production with a H2/CO ratio near unity that can be used
−35.6 kJ mol−1 (2) for transformation into products with high added value, for example,
long-chain hydrocarbons and oxygenated chemicals, particularly di-
Dry reforming (DRM) CH4 + CO2 = 2CO + 2H2, ΔH0 = methyl ether (DME) that is expected to be a clean fuel for the 21st
247.3 kJ mol−1 (3) century [6,7].
and combinations of all these reactions [1]. Selecting an approach is From a thermodynamic standpoint, CO2-CH4 reforming is favoured
not a simple task because it is dependent on process objectives, avail- by high temperatures and low pressures. The main feature of this re-
ability of raw material and energy resources [2]. The use of fossil re- action is its endothermic nature which requires high temperatures; the
sources leads to strong/increased liberation of carbon dioxide and process can occur at temperatures higher than 640 °C at stoichiometric
further accumulation in the atmosphere. The main contributors to the ratio of CH4/CO2 = 1, temperatures higher than 800 °C are required to
global warming effect are CO2 and CH4 [3,4]. The relative contribution achieve acceptable conversions [8]. Such conditions often lead to cat-
of these molecules to global warming was estimated to be 47% for alyst deactivation because of accumulation of coke over catalyst surface

⁎
Corresponding author. Present address: Häradsvägen 17C, 784 34 Borlänge, Sweden.
E-mail addresses: larisa.arkatova@gmail.com, larisa-arkatova@yandex.ru (L.A. Arkatova).
1
The corresponding author Larisa A. Arkatova will handle correspondence at all stages of refereeing and publication, also post-publication. She moved to Sweden for permanent
residence after the completion of this work at the Department of Structural Macrokinetics of Tomsk Scientific Center (Siberian Branch) of the Russian Academy of Sciences.

http://dx.doi.org/10.1016/j.cattod.2017.09.021
Received 15 November 2016; Received in revised form 28 August 2017; Accepted 12 September 2017
Available online 18 September 2017
0920-5861/ © 2017 Elsevier B.V. All rights reserved.
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

and/or sintering of the active metal particles. The main side reactions of choosing a catalytic support with desired properties is very important
are methane thermolysis (or methane cracking) (4) and dis- at present time. The situation is well summarized in the recent review
proportionation of CO (or reverseBoudouard reaction) (5) [1,5]: [16]. As a result, it is worthwhile to develop stable and effective Ni-
based catalysts for DRM.
CH4 = C + 2H2, ΔH0 = 74.8 kJ mol−1 (4)
Intermetallic compounds exhibited a great potential since they
−1
2CO = C + CO2, ΔH = −172.2 kJ mol
0
(5) possess peculiarities of the crystal and electronic structure, they de-
monstrate very important optimal properties such as thermal stability,
In addition, two other reactions related to carbon deposition in close mechanical strength and high thermal conductivity. Takayuki Komatsu
relationship with the dry reforming process are reduction of CO2 (6) revealed catalytic properties of Ni- and Co-based intermetallic com-
and the reduction of carbon monoxide (7) [1]: pound for DRM at 1073 K [25]. Single-phase catalysts were synthesized
CO2 + 2H2 = C + 2H2O, ΔH0 = –90 kJ mol−1 (6) by arc-melting mixture of Ni or Co and a second element. Ni- and Co-
intermetallides containing Ta, Hf and Sc showed higher initial activity
−1
CO + H2 = C + H2O, ΔH = −131.3 kJ mol
0
(7) than Ni and Co powders.
Atomized Ni3Al was applied for steam reforming of methane [26]. It
Ginsburg et al. [9] calculated the correlation between the CO2/CH4
was revealed that
ratio and the process of carbon formation on the surface of Ni-based
fine Ni particles, produced on the outer surface of Ni3Al, survived on
catalyst. According to their model, the reactant ratio should be equal to
the catalyst surface after tests at all the temperatures. The catalyst was
2 in order to provide less carbon accumulation, i.e. the oxidizing system
stable and unchanged up to 700 °C, while sintering and oxidation oc-
must have an excess of CO2 as an oxidant.
curred at high temperatures, leading to the rapid deactivation.
In general, the catalysts of DRM can be divided into two main
The aim of this work is the development of novel materials on the
groups: supported noble metals (Rh, Pd, Pt, Ru) and non-noble transi-
base of Ni3Al intermetallic compound implanted with Ru and Pt as
tion metals (Fe, Co, Ni) [5]. Nickel is the most promising choice due to
active and stable catalysts for CO2-CH4 reforming. Due to its excellent
its much lower cost and comparatively high activity and selectivity. The
high temperature strength and good corrosion/oxidation resistance,
main disadvantage of Ni-based catalysts is a tendency to be deactivated
mechanical strength and high thermal conductivity [27], Ni3Al can
due to sintering, coking, phase transformations and loss of an active
become an attractive and promising candidate for untraditional cata-
component [1,4,5,10–19].
lysts for DRM [1,4]. Moreover, in this paper we focus on self-propa-
In any case, due to rather low price of nickel a number of efforts
gation high-temperature synthesis as a novel synthetic approach,
have been made to develop Ni-based catalysts to improve the perfor-
combined with the process of ion implantation [28] as a method of
mance [4,16] with lower content of carbon deposits on the catalyst
surface modification.
surface and higher stability against metal sintering. Many scientists
investigated supported Ni catalysts and they indicated that the process
2. Experimental
of modification of the supports with alkali- and alkaline-earth oxides
(such as La-doped Ni/Al2O3 [17], Ni/yttria-doped ceria catalysts [20])
2.1. SHS
is extremely important and modification plays one of the main roles in
the catalytic activity and in the coking resistance of the catalysts. Sci-
SHS is a method for producing inorganic materials whereby re-
entific group led by Sadykov [18] developed the coke resistant fluorite-
actants are ignited to spontaneously convert to products by exotheirmic
like mixed oxides promoted by Pt for CO2-CH4 reforming.
reactions. This method is implemented at high temperature therefore it
In terms of catalysts reported to be efficient for DRM, one of the
is ideally suited for synthesis of refractory materials with unusual
most commonly described involves Ni either directly or supported. The
properties (metallic alloys, powders, ceramic materials with high
variety of metallic oxides (Al2O3, MgO, SiO2, ZrO2, TiO2, CeO2, and
purity, corrosion-resistant compounds at high temperature or superhard
others) have been widely investigated as supports for CO2-CH4 re-
intermetallides). It should be noticed, that an intermetallic compound
forming [16].
can be defined as an ordered allow phase formed between two or more
Most noble metals (Pt, Ru, Rh and Ir) have been studied as catalysts
metallic elements, where an alloy phase is ordered in two or more
for DRM [4,16,19]. There are a lot of debates and controversies among
sublattices are required to describe its atomic structure. The ordered
authors about which of these metals shows the best performance
structure exhibits superior elevated-temperature properties because of
[11,19–24]. Hansen and Rostrup-Nielsen investigated a group of cata-
the long-range ordered superlattice, which reduces dislocation mobility
lysts based on Pt, Rh, Ru, Ir, Pd and Ni. They revealed that Rh and Ru
and diffusion processes at elevated temperatures.
showed high activities and selectivities without carbon formation [19].
The modern SHS process is based on the special phenomenon named
Wang and Ruckenstein [20] concluded that deactivation is strongly
“solid flame” (autowave process where initial, intermediate and final
dependent on the nature of the support for supported Rh catalysts.
products exist in a solid phase) which was discovered by A. Merzhanov,
Keeping in mind the fact that Rh seems to be the metal that better
I. Borovinskaya and V. Shkiro in 1967 and patented in 1972 [29–32].
fulfills the compromise between activity and stability, the required
SHS can be used for many applications, for example, highly active
aspect of an optimal catalyst would be to induce high conversions of
catalysts and stable supports have been developed for oxidation of CO,
CH4 without leading to deactivation and a comparatively low price. Pt
H2, NO, hydrocarbons, aldehydes, alcohols and methane to synthesis
is an attractive metal in order to be used as a catalyst for DRM due to its
gas, pyrolysis of diesel, petrol and naphtha, steam reforming of kero-
availability and relatively low price in comparison with Rh.
sene, combustion of carbon, hydrogenation, hydrodesulphurization,
Nevertheless, Nagaoka et al. detected that Pt particles are covered by
dehydrodimerization of CH4, ammonia synthesis, etc. [33].
coke monolayer on the surface of Pt/Al2O3 catalysts [21]. Among zir-
There is a more general concept, which is a combustion synthesis
conia supported metals of group VIII, Pt is one of the most active and
(CS). There are two modes by which CS can occur: 1) self-propagation
stable metals as Van Keulen and his colleagues have shown in [23].
high temperature synthesis (SHS), where a reactant compact is ignited
Some other catalysts, particularly Pt/ZrO2, Pt/α-Al2O3-ZrO2, showed
at one end using a heat W-coil, forcing the local reaction which then
significantly lower deactivation level than Ni/ZrO2 [24]. In spite of the
goes as a wave; 2) thermal explosion (TE) (or simultaneous combus-
fact that γ-Al2O3 is one of the most used supports in research, this oxide
tion), where the whole compact is heated up to the ignition temperature
is thermally unstable at high temperatures (> 600 °C) and undergoes
at which all components convert to the final product(s) [34]. In our
the phase transformation process into the more thermally stable α-
work intermetallic Ni3Al was prepared by SHS in the first mode
Al2O3 or/and Ni/Al2O3 tends to form NiAl2O4 spinel. Thus, the question
[4,29,30]. The powders of Ni and Al were preliminarily dried in

304
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

Table 1
Designation of the initial catalysts, their phase composition and lattice parameters of the main Ni3Al phase. BET surface area and Ni3Al dispersion before/after DRM. (Reaction conditions:
CH4:CO2:He = 20:20:60 vol.%, p = 1 atm, V(catalyst) = 1 cm3, V(CH4 + CO2) = 100 cm3 min−1, (Ni,Al) solid solution of aluminium in nickel.).

Catalyst Preparation method Pt or Ru dosage, Initial phase composition (in Lattice parameter BET surface area Particle diameter for
(ion cm−2) accordance with XRD) (nm) (m2 g−1) Ni3Al (nm)

Ni3Al SHS 0 Ni3Al + [(Ni,Al) + Ni] traces 0.3569 1.1/0.8 28/92

Pt/Ni3Al-1 SHS + II 1*1016 Ni3Al + [(Ni,Al) + Ni] traces 0.3569 1.1/0.9 27/67
Pt/Ni3Al-2 SHS + II 5*1016 Ni3Al + [(Ni,Al) + Ni] traces 0.3571 1.2/0.9 25/48
Pt/Ni3Al-3 SHS + II 1*1017 Ni3Al + (Ni,Al) + Ni 0.3573 1.3/1.1 21/36
Ru/Ni3Al-1 SHS + II 1*1016 Ni3Al +[(Ni,Al) + Ni] traces 0.3570 1.1/0.8 29/90
Ru/Ni3Al-2 SHS + II 5*1016 Ni3Al + (Ni,Al) + Ni traces 0.3572 1.1/0.9 27/71
Ru/Ni3Al-3 SHS + II 1*1017 Ni3Al + (Ni,Al) + Ni 0.3574 1.4/1.3 24/4

vacuum at 100 °С for 3 h. SHS was conducted in the same way as re- ion beam (between 50 and 200 keV) can be directed onto catalyst
ported before [1,4,28,35]. The basic chemical reaction was the fol- surface of Ni3Al. The feature of ion implantation is that it is not a
lowing: surface coating process but a method of penetration below the substrate
surface. One of the advantages of ion implantation is ability to implant
3Ni + Al = Ni3Al, ΔH0f (Ni3Al) = −157 kJ mol−1 (8)
ions that do not usually diffuse or are insoluble in the substrate mate-
In accordance with Eq. (8), the reaction releases a significant rial. The Mevva – V ion source is thoroughly described in [38].
amount of heat. It means that preheating the compact prior to ignition We applied the total ion current of 0.5 А at an accelerating voltage
is not necessary in order to gain the self-sustained combustion. of up to 50 kV. The area of the ion beam cross-section at the output of
The most important characteristic of synthesis is the cost-effective- the extracting system therewith amounted to 100 cm2. Ni3Al grains
ness of the SHS, which is associated with utilization of reaction heat were placed to the holder in front of ion beam with ion current density
instead of electric power, relatively high combustion temperature and of 2.3 mA cm−2, average ion energy of 60 keV, frequency 2 Hz in va-
burning velocity, and simplicity of facilities. Some of the typical para- cuum of 3*10−6 Torr. Three Pt- and three Ru-implanted Ni3Al catalysts
meters of SHS were the following: induction time for ignition (0.4 s), with different dosages (1*1016, 5*1016 and 1*1017 ion cm−2) were
burning velocity (1 cm s−1), combustion temperature (1200–1300 °C). prepared. For convenience and simplicity, the catalysts are referred to
In order to avoid any side effects, we did not specifically treat metal as Ru/Ni3Al-1, Ru/Ni3Al-2, Ru/Ni3Al-3 and in the same style as Pt/
powders, for example, in the atmosphere of hydrogen or other reductive Ni3Al-1, Pt/Ni3Al-2, Pt/Ni3Al-3, respectively. The phase composition
gases. First of all, it was important to avoid metal hydrides formation at and lattice parameters of prepared catalysts are presented in Table 1.
high temperatures during SHS [36] because majority of byproducts can
negatively influence on combustion kinetics. Secondly, metal powders 2.2. Catalytic tests
were produced by electroplating technique and, as a consequence, they
had an original high purity. Two types of Ni powders with different CO2-CH4 reforming was carried out in the catalytic set-up with a
particle size (PNE-1, Norilsk Nickel, 45–70 μm, 99.9% purity and PNK- temperature-programmable quartz tube reactor with inner diameter of
L5, Norilsk Nickel, < 10 μm, 99.7% purity) were used as one of the 6 mm and a length of 130 mm at temperature range of 600–900°С,
reactants. In the case of using coarse nickel powders (45–70 μm) the molar ratio CO2:CH4 = 1:1 and at ambient pressure. The test tem-
product expanded from the initial compact and had a porous structure. perature was measured with a thermocouple placed axially in the re-
Aluminum particles (ASD-4, Shelekhov, 5 μm, 99.7% purity) were used. actor inside a special thermocouple well with a diameter of 3 mm. The
In this work we used a double-acting pressing in order to gain optimal temperature gradient over the bed under DRM was less than 3 °C.
density. The porosity calculated in the current work was about 37%, Duration times of the stability tests were around 7200 min (120 h). The
which is optimal for a basic Ni3Al component. Powder mixtures were reactor was placed in a vertical tube furnace. Undiluted catalyst
prepared with the composition corresponding to stoichiometric ratio of (0.5–1 cm3 bulk volume) was used for testing. The procedure of pre-
Ni:Al = 3:1, and were dry mixed in a ball mill for 4 h. Controlling XRD liminary reduction was excluded because of several reasons: 1) Ni3Al
pattern of blended mixture demonstrated no signals other than those of can undergo the structural transformation that leads to decrease CH4
Ni and Al, with an indication that no mechanical alloying of initial and CO2 conversions; 2) during H2 pretreatment partially oxidized
powders occurred while the milling operating. Then mixed powders nickel is able to be reduced to metallic Ni and undergo anticipatory
were cold-pressed into cylindrical compact with a diameter of 2 cm and sintering. Before each experiment, heat treatment was carried out up to
height of 10 cm. SHS of pressed samples was performed in a special 600 °C in the atmosphere of helium (50 cm3/min). After 1 h at this
chamber under constant pressure in argon atmosphere. Heat impulse temperature, the feed was switched to reaction gases
was supplied to a butt part of the pressed intermediates. Ni3Al burned (CH4:CO2:He = 20:20:60 vol%, 100 cm3 min−1). Then reactor was
compacts obtained by the above-mentioned SHS method were me- heated to required temperature at 10 °C min−1 rate. When the tem-
chanically crushed and fractions with particle size of 400–600 μm and perature approached the value, the sampling of post-reactor gas phase
600–1000 μm were selected for catalytic investigations. was started. The reactor effluent was analyzed by on-line gas chroma-
The process of modification was performed in Mevva-V.RU (Metal tographic (GC) analysis (Chromos 1000 equipped with TCD and two
Vapour Vacuum Arc) set-up (designed in Lawrence Berkeley National columns with Carbosieve SII and Porapak Q.
Laboratory (USA) and developed by the group [37]. We used a method
of ion implantation, which is a surface treatment process where ions of
2.3. Characterization of catalysts
Pt and Ru are accelerated and made to penetrate the surface of Ni3Al.
The atoms of Pt and Ru are converted into an ion form by electron
The specific surface area of samples was determined by the BET
collisions in a plasma, focused into a stream using magnets and ac-
method by means of nitrogen adsorption using a ChemiSorb 2750
celerated by a voltage gradient towards the substrate. Ion implantation
analyzer (Micromeritics, USA) linked with mass-spectrometer QMS-300
is much more reproducible and controllable than other conventional
(Stanford Research System, USA). X-ray phase analysis was performed
surface treatments. After this process, the component surface requires
using a Shimadzu XRD-6000 diffractometer (Cu Ka radiation,
no further treatment prior to use. Using this approach, a high energy
λ = 0.154187 nm). All measurements were carried out before and after

305
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

all catalytic investigations. others are formed by peritectic reactions. Two of them, NiAl and Ni3Al
The evolution of the catalyst structure was investigated by means of have received the extensive majority of scientific attention because they
in situ XRD using the VEPP-3 station of Siberian Synchrotron and are considered to be perfect candidate materials for high temperature
Terahertz Radiation Center of Budker Institute of Nuclear Physics of SB structural and coating application [40]. It is understandable con-
RAS. To compare the phase structures of the samples, the count of in- sidering that these two intermetallic compounds have the highest
tensity of diffraction peaks was normalized (λ = 0.1731 nm). melting points in this system, rather low densities, acceptable strength
The catalyst morphology was examined by SEM using a scanning and high temperature corrosion resistance. In spite of the fact that NiAl
electron high-vacuum microscope VEGAII LMU coupled with an X-ray intermetallide has a higher melting point (1638 °C) than Ni3Al (T
energy dispersive spectroscopy (EDS) system (Oxford INCA Energy 350) melt. = 1385 °C), it suffers from low temperature brittleness and poor
and SE (secondary electron) and BSE (backscattering electron) detec- strength at rather high temperatures.
tors. Carbon content was determined by thermogravimetry-differential The nickel-rich part of the diagram is of fundamental interest for
thermal analysis (DTA-TG). The first part of measurements was con- research in the development of ordered alloys as new structural high-
ducted under oxidative (air velocity was 20 cm3 min−1) atmosphere temperature material since the yield strength of the intermetallic Ni3Al
with STA 409 Luxx (NETZSCH, Germany) using 30–50 mg of samples (γ’-phase) increases anomalously with temperature. More and more
and with 10 °C min−1 increase rate from room temperature to 1000 °C. scientists are interested in these compounds with the view of replacing
The second part of investigations was conducted in the atmosphere of Ni-based superalloys. The Ni3Al alloys are superior to the commercial
inert gas. alloys, especially in high-temperatures processes, in an oxidizing and
The particle size and state of the active components were estimated carburizing environments. Ni3Al can be obtained by the following
by HRTEM using JEM-2010 (JEOL, Japan) microscope equipped with peritectic reaction under equilibrium conditions [41]:
Gatan slow-scan camera for high-resolution observation and energy-
L (liquid) + β [(NiAl)] = γ’ [AlNi3] at 1395 °C (9)
dispersive X-ray microanalyzer (EDX) for elemental analysis of spe-
cimen surface. The catalysts were placed in an ultrasonic bath in where L is molten Ni + Al and β [(NiAl)] is β–phase of solid NiAl inter-
ethanol to disperse Pt/Ni3Al, Ru/Ni3Al and Ni3Al. Then samples were metallic compound.
transferred on a commercially available Cu grids (200 mesh Cu holey Ni3Al superalloy possesses the following attractive properties [40]:
carbon, SPI Supplies/Structure Probe, Inc). The proper sample is then 1) a high corrosion resistance in oxygen and carbon enriched atmo-
positioned in the microscope chamber and then, analysed by observing sphere up to 1100 °C, due to a formation of a continuous surface Al2O3
every tile of the grid, focusing on each nano-agglomerate. Images are layer, 2) a high fatigue strength resulting from the elimination of stress
taken at various magnifications, prior to image acquisition, and ele- concentrations on the second phase particles, 3) a high tensile and
mental analysis could assure the nature of metal or intermetallide. compression strength at 650–1100 °C, 4) a high corrosion resistance in
The elemental composition and electronic state of the elements were organic acids and bases, 5) an excellent high temperature wear re-
investigated by means of X-ray photoelectron spectroscopy (XPS). The sistance, 6) a relatively low density, 7) catalytic activity in different
patterns were acquired with a Physical Electronic 5700 spectrometer processes, for example, methanol decomposition [26], reforming of
equipped with a hemispherical electron analyser and MgKα X-ray ex- methane and others [1,4,28,33,35,36,42]. That is why Ni3Al was
citing source (1253.6 eV, 15 kV, 300 W). It has been used as an internal chosen as a support or even a catalyst for DRM.
patron for calibration C1 s (284.4 eV) considering a deviation ± 0.2 eV.
The binding energy (BE) scale was pre-calibrated using the positions of
the photoelectron of Au4f7/2 (BE = 84.0 eV) and Cu2p3/2 3.2. Structure of the materials on the base of Ni3Al
(BE = 932.67 eV) core level peaks. The samples were supported onto
double-sided conducting copper scotch tape. The binding energy of Fig. 1 demonstrates two spectral patterns with a distribution of
peaks was calibrated by the position of the C1 s peak (BE = 284.8 eV) elements (Al and Ni) along the line and the microstructure of the ma-
corresponding to the surface hydrocarbon-like deposits (CeC and CeH terial obtained in the plane wave propagation mode. The SHS compact
bonds). For the survey spectra the pass energy of the analyzer was was preliminary cut and polished for such an investigation.
50 eV, while for the narrow spectral regions it was 20 eV. For the There are no traces of NiAl3 at this stage and the porous micro-
quantitative analysis the integral intensities of the photoelectron structure shows Ni-rich globules distributed in the matrix. The central
spectra were corrected by the corresponding atomic sensitivity factors. part (or core) of the globules (2–20 μm, the bright spots) appears as
Spectral analysis and data processing were performed with XPS Peak virtually unreacted Ni. The core is surrounded by a band of Ni3Al (grey
4.1 program. The ionic etching technique was employed to remove colour). The dark-grey region (the original location of aluminum), is
possible impurities that could reside on the sample surface and to now a multiphase structure, consisting of small quantity of NiAl phase
analyze changes in atomic concentrations of elements along the depth. and Ni-Al solid solutions of different compositions. The black spots in
Etching was performed using an IQE 11/35 (SPECS) ion gun with the the picture are pores. The above microstructure was observed at room
energy of argon ions 2.5 kV and current density 10–12 mA/cm2. The temperature and is generally different from the microstructure existing
rate of surface etching (estimated with the use of calibrated thin InAs/ at high temperature. It should be noted that such a structure of the
SiО2and Al2O3 films) was ∼ 1 nm/min. Bombardment of the sample by material can undergo some further phase transformations at high
argon ions during 10 min allowed removal of a surface layer with a temperature of CO2-CH4 reforming as it will be shown later.
thickness of at least 10 nm. Additional information was obtained by transmission electron mi-
croscopy. Fig. 2 shows the structure of the material obtained by SHS in
3. Results and discussions the wave propagating mode.
TEM pictures clearly demonstrated the presence of rather small Ni3Al
3.1. Analysis of Al-Ni phase diagram and attractiveness of Ni3Al particles (50–200 nm), surrounded with alumina. As HRTEM and the
Fourier diffraction pattern confirmed, the crystal lattice corresponded to β-
Analyzing the Al-Ni phase diagram [39], we can conclude that there Al2O3. It means that small amount of alumina was also formed during SHS.
are five intermetallic compounds that can be formed by these two This oxide, in addition to the main function of forming the intermetallic
metals. They are: NiAl3, Ni2Al3, Ni5Al3, NiAl and Ni3Al. Only one Ni3Al, plays a positive role in the formation of protective oxide layer on
chemical compound (NiAl3) has a constant composition, and all the the surface of this material. Moreover, from a catalytic point of view, the
other intermetallic compounds have substantial homogeneity regions. presence of alumina can increase the conversions of CO2 and CH4 and
The intermetallic NiAl substance melts congruently, and all the prevent premature sintering of the catalyst.

306
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

Fig. 1. Backscattered electron image and distribution of Ni and Al in the initial Ni3Al catalyst prepared by self-propagation high-temperature synthesis (this half-finished Ni-Al sample
was preliminarily polished).

3.3. Phase composition of the materials obtained by SHS solution on the base of NiAl (β-phase) has a body-centered cubic lattice
type with the lattice parameter of 0.2887 nm. Nevertheless, this phase
Fig. 3(a,b) shows XRD patterns of fresh catalyst obtained by self- has very specific features, for example, its lattice parameter can be
propagating high-temperature synthesis. It is obvious that all the most increased or decreased, depending on the Al content [31]. The max-
intense diffraction peaks belong to the Ni3Al phase. The presence of imum lattice parameter is observed in the case of stoichiometric com-
superstructure reflections (100) and (110) (Fig. 4a) implies that Ni-Al position 1:1 (i.e. NiAl). In all other cases the lattice parameter decreases
system is well-ordered. It should be noted that γ’-phase (solid solution irrespective of the aluminum content. This class of materials is an in-
on the base of Ni3Al) can be obtained by peritectic reaction (9) at teresting subject due to the following peculiarity: the metal atoms are
1395 °C (under equilibrium conditions) but the value of Tcombustion in distributed randomly above the critical temperature (ordering tem-
SHS was slightly lower. It means that the reaction was not complete and perature), but they form two sublattices and occupy only very certain
the degree of conversion of the initial Ni and Al powders is not 100%. positions below this temperature [48]. During SHS, in the combustion
Moreover, intermetallic Ni3Al has a cubic structure (Cu3Au type) and its wave, not only simple chemical compound is formed but ordered, at
lattice parameter (0.3589 nm in pure Ni3Al) increases almost linearly least partially, NiAl phase of B2 type with a wide range of homogeneity
with increasing Al content. In our case the lattice parameter varied can be formed as well. According to the phase diagram (as mentioned in
between 0.3569 and 0.3573 nm. It means that the amount of aluminum paragraph 3.1), NiAl (β-phase) exists between 43 at.% and 70 at.% (62
in the intermetallic compound is slightly understated due to the for- and 83.9 wt.%) Ni at temperature of 1400°С and between 45 at.% and
mation of the intermetallic compound in the non-equilibrium SHS 60 at.% (64 and 76.5 wt.%) Ni at room temperature [49]. A well-known
conditions [43,44] i. e. the reaction between the components was not feature of the intermetallic compound NiAl is a change of the type of
completed. It means that the reactants can react later during the CO2- the solid solution within the homogeneity range. Alloys containing less
CH4 reforming process. than 50 at.% Al, are substitutional solid solution. The increase of alu-
Nevertheless, the long-range order exists, as shown by X-ray dif- minum content more than 50 at.% causes the structure reorganization
fraction, that is, the phase of the intermetallic compound is definitely and formation of subtraction solid solutions on the same basis while the
formed. Moreover, the lattice parameter of Ni3Al phase increases concentration of vacancies increases in the nickel sublattice. The sub-
during the process of DRM, which indicates expansion of the lattice. We traction solid solutions are formed due to the inability of Al to replace
attribute this increase to the dissolving of carbon after the process of Ni in the crystal lattice (in the Ni-Al alloy). It is associated with the
methane dissociation (and possibly, after the dissociation of CO2 and difference in atomic radii of pure nickel (0.124 nm) and aluminum
subsequent disproportionation of CO). (0.143 nm), or with the difference in the electron concentration. The
The diffraction peaks at 2θ of 44° and 52°, corresponding to metallic existence of these two types of solid solutions at a deviation from the
Ni (111) and (200), were observed in the catalysts prepared in the wave stoichiometric composition has a great influence on the properties of β-
propagating mode without any modification. In the case of catalyst phase. The presence of structural vacancies in the subtraction solid
preparation by SHS in a thermal explosion mode under near adiabatic solutions causes different changes in physical and mechanical proper-
conditions, the final product contained only Ni3Al phase. Since the most ties (hardness, tensile strength, density, brittleness, etc.). Therefore, we
active catalyst had multiphase structure, we decided to use SHS in the must take into consideration all the properties of this additional phase,
wave propagating mode for further investigations. as well as all conditions leading to the NiAl phase formation.
In addition, two different solid solutions on the base of Ni-Al were The most important point is that NiAl phase (stoichiometric or
detected. According to the PCPDFWIN database, they had compositions nonstoichiometric) can play a positive role in the structure of a catalyst,
as Al1.1Ni0.9 and Al0.96Ni1.04. The composition is very close to NiAl, namely: 1) it has a high enthalpy of formation, therefore, it improves
which has a structure of B2 type (CsCl) and it belongs to a class of the kinetics of combustion during SHS, 2) it is an inert phase (as we
ordered alloys. The phase diagrams or conditions for combustion investigated earlier [42]), so it can serve as a diluent or catalyst sup-
synthesis for the Ni-Al systems can be found in [45–47]. The solid port, 3) it is characterized by a high melting point (1638 °C), so it can

307
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

Fig. 2. TEM, HRTEM, EDS spectra and Fourier diffraction pattern of the catalyst based on Ni3Al (initial sample, after self-propagation high-temperature synthesis).

not be completely destroyed during CO2-CH4 reforming at 800–850 °C, 3.4. Catalytic activity and stability of Pt- and Ru-implanted Ni3Al in DRM
4) it has the highest heat resistance among all the other aluminum-
nickel alloys, so it can partially protect catalyst from oxidation. We have already investigated and reported the results related to
As for the Ru- and Pt-implanted catalysts, they had very similar catalytic activity of three intermetallic systems on the base of Ni3Al,
diffraction pictures as Ni3Al itself. Ru- or Pt- diffraction peaks were not NiAl and Ni2Al3 in DRM [42]. Intermetallic Ni3Al (containing a small
detected by XRD for both of the modified catalysts (even with the amount of metallic nickel phase) is the only one showing rather good
highest Ru and Pt dose used during the ion implantation). In general, it catalytic activity but it requires 3–4 h (sometimes even more, de-
was a predictable result because noble metals were implanted in a pending on the conditions) to reach the steady state due to additional
highly dispersed state, they could not form a single crystal phase, and phase transformations after SHS. Nevertheless, the main problem of
therefore, they could not produce any diffraction patterns. The only such a catalyst was deactivation due to coking [1]. Thus, taking into
feature that distinguished the original samples from implanted, was a consideration Pt and Ru abilities to activate the strong CeH bond (EC-
slight difference in the lattice parameters. The bulk lattice constant for H > 420 kJ mol−1) [50], and also extremely low carbon solubility in
Ni3Al phase in initial unmodified sample, a = 0.3569 nm was calcu- Ru (0.0004 wt.% at 820 °C) and in Pt (0.8 at.% at 876 °C) [43] we
lated from this XRD data (taken at room temperature). In the case of decided to incorporate such noble metals into the Ni3Al matrix, and to
modified catalysts, these values were the following: 0.3569 nm and use the specific method of catalyst modification by ion implantation to
0.3573 nm (for Pt/Ni3Al-1 and Pt/Ni3Al-3, respectively); 0.3570 nm provide very homogeneous Ru and Pt distribution.
and 0.3573 nm (for Ru/Ni3Al-1 and Ru/Ni3Al-3, respectively). Thus, The influence of additives on the catalytic performance is compared
the difference in lattice parameters is not significant. This could be in Fig. 4, which shows the dependence of CH4 and CO2 conversions on
attributed to the presence of defects of crystal lattice induced by ion reaction temperature for each catalyst.
implantation. The observed order of methane conversions over doped Ni3Al catalysts
is Pt/Ni3Al-3 > Ru/Ni3Al-3 > Pt/Ni3Al-2 > Pt/Ni3Al–1 ≥ Ru/Ni3Al-
2 > Ru/Ni3Al-1 > Ni3Al. The carbon dioxide conversion also showed
the same trend. The results have shown that in the case of Ni3Al

308
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

temperatures, for example:

a) 2CH 4 + CO2 = C2 H6 + CO + H2 O, ΔH 0 = 106 kJ mol−1 (11)

b) 2CH 4 + 2CO2 = C2 H6 + 2CO + 2H2 O, ΔH 0 = 284 kJ mol−1 (12)

In addition, the reactions that could be observed during DRM,

would most probably be related to reactions with H2, such as the re-
verse water-gas shift reaction (RWGS, at temperatures lower than
800 °C), which is the major cause of deviation from unity for H2/CO
ratio:

c) CO2 + H2 = CO + H2 O, ΔH 0 = 41 kJ mol−1 (13)

аnd methanation reactions (although they are not favorable at high

temperatures) [16],

d) CO + 3H2 = CH 4 + H2 O, ΔH 0 = −206.2 kJ mol−1 (14)

e) CO + 4H2 = CH 4 + 2H2 O,ΔH 0 = −165 kJ mol−1 (15)

In the case of Ru/Ni3Al and Pt/Ni3Al, the main products in DRM are
H2 and CO, as well as a negligible amount of water. The process se-
lectivity, represented as a H2/CO molar ratio, for all catalysts is less
than unity, as usual between 0.85 and 0.96 at relatively low tempera-
tures. This ratio decreases (up to 0.71–0.76) with temperature increase
up to 900 °C. This is in agreement with the results reported in [51,52].
In addition, this ratio decreases with the decrease of Ru and Pt content,
i.e. these noble metals induce H2 formation.
The effect of noble metals on catalytic performance can be ex-
plained by comprehensive studies performed by other investigators.
Many different conversion mechanisms have been proposed [13,14],
but one of the common steps is the activation of CH4 on the surface of
corresponding metals. The process runs with formation of adsorbed
carbon and followed by reaction with the CO2 (with subsequent CO
Fig. 3. XRD patterns of the catalyst on the base of Ni3Al phase after self-propagating high- formation). This process includes a methane dissociative adsorption
temperature synthesis: (a) – the whole diffraction pattern, (b) – analysis of the most in-
that leads to the formation of CHx species and the subsequent surface
tense peak.
reactions with H2 formation. The dissociation of CH4 is considered to be
the rate-limiting step in the SRM and DRM. Thus, in order to investigate
intermetallic samples the presence of Ru, and especially Pt, strongly en- the fundamental adsorption of CH4 on Ni, Pt, Ru single crystals were
hances the catalytic activity (Fig. 4a,b), with an effect which increases performed by different scientific groups [53–57]. Alfred B. Anderson
increasing Ru and Pt content. The most active catalysts are Pt/Ni3Al-3 explored mechanisms, activation energies, and orbital interactions as-
(1*1017 ion cm−2) (CH4 conversion is 95–96% at 850 °C), and Ru/Ni3Al-3 sociated with the oxidative addition of a methane CeH bond to several
(1*1017 ion cm−2) (CH4 conversion is 92–93% at 850 °C), while Ni3Al idealized clean transition metal surfaces [53]. He used an ASED mole-
catalyst showed only 72%. cular orbital theory and cluster method. The activation of CeH bond in
CO2 and CH4 conversions are not equal to each other at the same alkanes is the first step in oxidation catalysis. It is clear from simple
temperature as well as H2/CO is not equal unity because the side re- consideration of metal-hydrogen bong strengths (≤250 kJ mol−1) and
actions occur at different temperature ranges. Interaction between CeH bond strengths (434 kJ mol−1 for breaking the first CeH bond in
carbon dioxide and methane is of course a priority, however, from a the molecule of CH4) that oxidative addition and not hydrogen atom
thermodynamic point of view, some other reactions are favored at high abstraction will be the route followed on metal surfaces. In contrast,

Fig. 4. CH4 and CO2 catalytic conversions (a,b) over Ni3Al and Ru- and Pt-implanted catalysts at 600–900 °C. (Reaction conditions: CH4:CO2:He = 20:20:60 vol.%, p = 1 at., V
(CH4 + CO2) = 100 cm3 min−1).

309
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

O−x at oxide surfaces do abstract hydrogen atoms from CH4, forming reactive carbon occurs. In our case, the synergistic catalytic effects of
gas-phase methyl radicals. In case of Ni, activation energies Ea are Ru-Ni and Pt-Ni systems are possible due to the preparation method,
higher on single crystal surface than on films and supported metal which is extremely non-equilibrium. During the implantation process,
particles. The calculated activation energies for methane activation the ions with high energy are able to form new unsteady phases (such as
were: 64 kJ mol−1 for the close-packed Ni (111) surface, 29 kJ mol−1 Ru-Ni or Pt-Ni alloys or even ternary alloys including aluminum) after
for the more open Ni (100). This is as expected, being substntially less their penetration to the crystal lattice of Ni3Al. Moreover, each in-
than for the close-packed surface. The 43 kJ mol−1 activation energy dividual ion of Ru+4 (or Ru+6 , Ru+8) and Pt2+ (or Pt4+) produces
calculated for Pt (111) is close to experimentally determined value of different point defects in the crystal lattice of Ni3Al with formation of
about 46 kJ mol−1 for the (110) surface [54]. Pt (111) is predicted to be vacancies and interstitials. Due to a very strong collision with a target
more active in methane activation than Ni (111). These data are more Ni3Al crystal the transfer of a significant amount of energy to the target
or less in compliance with those obtained by D.W. Goodman’s scientific atom occurs. Then the target atoms can cause successive collision
group [55,56]. They investigated the kinetics of methane decomposi- events. Interstitials result when such atoms (or the original ion itself)
tion on low index planes of Ni single crystal surfaces. The experiments come to rest in the solid, but find no vacant space in the lattice to
were performed in a high-pressure reaction cell with surface char- reside. These point defects can migrate and cluster with each other,
acterization via AES in a contiguous UHV chamber subsequent to re- resulting in dislocation loops and other defects. This a topic of separate
action with methane. The reaction probability of methane was the deep investigation.
highest on Ni (110) and the lowest in Ni (111). The apparent activation Since Pt/Ni3Al-3 and Ru/Ni3Al-3 demonstrated the highest catalytic
energies for methane decomposition were found to be 50,2 kJ mol−1 activity, all the other measurements have been made with these cata-
for Ni (111), 55,6 kJ mol−1 for Ni (110), and 25,1 kJ mol−1 for Ni lysts. The stability results are illustrated in Fig. 5. It is clearly seen that
(100). The experiments in their lab indicated that CH4 dissociation on the catalytic activity is stable from the initial time of the reaction up to
Ni (100) proceeded predominantly via a precursor mediated me- 120 h only for the Pt-implanted catalyst, both of the conversions of
chanism with only a small contribution from a direct dissociation me- reagents are rather stable, values decreasing only slightly during all
chanism, though the other authors observed that direct dissociation period of testing.
mechanism was dominant [57]. At the same time, the Ru-implanted catalyst lost its activity, for
As for another noble metal (ruthenium), the methane dissociation example CH4 conversion decreased from 96 to 84%. The matrix itself,
on the Ru single crystal was carried out in the reactor chamber and then Ni3Al, requires 4–6 h, or even longer time (depending on the SHS
the surface species were detected by HREELS spectra subsequent to CH4 preparation mode) to reach the steady state. Ni3Al works steadily for
decomposition on Ru (1120) as a function of reaction temperature [55]. the first 32–34 h, but then both conversions start to decline rather
The methylidyne (CH), vinylidene (CCH2) and ethylidene (CCH3) spe- quickly. This observation suggests a matrix deactivation. This question
cies were detected on Ru (1120) and only CH and CCH2 on Ru (0001). is a complicated one and will be discussed later.
For both substrates, only the graphitic phase of carbon was observed Effect of CO2/CH4 molar ratio on catalytic activity was studied in
above 427 °C. It means that Ru (1120) was more reactive in dissociation our earlier work [59]. It was demonstrated that the composition of the
of CH4. This difference was attributed to the structural differences be- reaction mixture affects the conversion level of both components,
tween the two substrates. Thus, in spite of some peculiarities of cal- especially methane, for example, CH4 conversion increased from 64%
culations and experiments in the abovementioned studies, these data to 78% at CO2/CH4 = 3 and 800 °C. Nevertheless, the conversion of
are consistent with each other. Analyzing these results, we can conclude CO2 reduced from 73 to 36% at the same conditions. A favourable effect
that the dissociation of methane proceeds much better on precious of CO2 on DRM is caused by its ability to remove the carbon deposits,
metals (Ru, Pt) than on Ni. Moreover, the measured sticking coefficients i.e. to regenerate the catalyst according to the Boudouard reaction (5).
for Ru (0001) varied from 10−6 to 10−7 whereas this value varied from Nevertheless, it is not desirable to raise the CO2/CH4 molar ratio for
about 10−9 to 10−7 for Ni (111) and from 10−8 to 10−7 for Ni (100) more than 1.5, because the selectivity of DRM will be changed, H2/CO
[55]. It means that the number of adsorbed CH4 molecules on the will be greatly reduced (to 0.5–0.7) and, moreover, nickel can be oxi-
surface of Ru is much higher than on Ni regardless of the index facets. dized to nickel oxide, which is not active in CO2-CH4 reforming.
Besides, as shown by Crisafulli [58], ruthenium additions to nickel Thus, the ion implantation of Ni3Al matrix by Ru, and especially Pt,
catalysts greatly increase their activity due to the formation of nickel- leads to a remarkable increase of catalytic activity in DRM. Pt/Ni3Al
ruthenium bimetallic clusters. At the same time, highly dispersed nickel showed the best stability. The Ni3Al matrix experienced moderate de-
is formed, and, consequently, the process of formation of the more activation due to the presence of free nickel, which is very efficient

Fig. 5. Stability test for Ni3Al, Ru/Ni3Al-3 and Pt/Ni3Al-3 catalysts during CO2-CH4 reforming. (Reaction conditions: T = 900 °C, V(CO2:CH4:He = 20:20:60)) = 100 cm3 min−1.

310
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

stream, the intermetallide reflections shifted to the low-angle side. It

means that the lattice constant increases for 0.0015-0.002 nm while the
lattice constants of (Ni,Al) solid solution and metallic Ni remain the
same. Furthermore, the intensity of Ni3Al peaks becomes smaller
whereas the intensity of Ni peaks increases. This is mainly a reaction
impact.
The lattice parameter of initial unmodified Ni3Al sample,
a = 0.3569 nm (Table 1), was calculated from these XRD data (taken at
room temperature). In implanted catalysts these values were the fol-
lowing: 0.3573 nm (for Pt/Ni3Al-3) and 0.3574 nm (for Ru/Ni3Al-3), so
they are comparable ones and have the same magnitude. Nevertheless,
these slightly increased parameters show the presence of defects of
crystal lattice induced by ion implantation. After catalytic tests con-
ducted at 800° for 24 h, the Ni3Al lattice parameter becomes equal to
0.35796 nm for unmodified catalyst, 0.35807 nm for Pt/Ni3Al-3 and
0.35809 nm for Ru/Ni3Al-3, i.e. the lattice parameters have increased
after DRM. In the Pt- and Ru-implanted catalysts, as expected, the
phases of Ru and Pt are not detected in both cases, before and after
catalytic process. The main reasons are obvious, the content of noble
metals is very small and the conditions of ion implantation allowed to
spread Pt and Ru in a very ultrafine state (primarily in the atomic state).
By means of in situ XRD we confirmed that the phase of non-
stoichiometric Ni3AlC0.5 carbide is formed in the unmodified Ni3Al
Fig. 6. In situ X-ray diffraction measurements for Ni3Al catalyst during DRM during DRM. This phase was not detected in the implanted catalysts.
(CO2:CH4 = 1:1, V(CO2 + CH4 = 50 cm3 min−1, time-on-stream 1.5 h). The intermetallic Ni3Al phase, in comparison with Ni, has a similar
crystal lattice but with a slightly larger parameter constant. It means
element for CO2-CH4 reforming but it has a tendency to form carbon that carbon atoms are able to penetrate into the lattice of intermetallide
under operating conditions. So, the primary effect of promoters is to easier than into nickel one. In this case, the lattice parameter of Ni3Al
prevent catalyst deactivation by suppressing carbon deposition on the phase will be increased as we observed. There are no special hindrances
catalyst surface. Finally, combining a well dispersed Ru and Pt that for carbon diffusion in the Ni3Al sample. In the case of the implanted
promotes dissociation of CH4 on a support that allows a better inter- catalysts, well-dispersed Ru and Pt (with extremely low carbon solu-
action with CO2 may become a winning approach for the DRM reaction. bility and inability to form carbides), being in intimate contact with Ni,
physically and energetically impedes the diffusion of carbon atoms into
the Ni3Al crystal lattice. The mechanism for carbon formation on Ni
3.5. Bulk properties of the catalysts, ex situ and in situ XRD particles have been thoroughly debated in the literature [60,61], and
involves the formation of surface carbides, graphite, and dissolution
The initial phase composition and its different features for Ni3Al into the bulk.
catalyst were described in the paragraph 3.4. It was observed that the Thus, the XRD investigations revealed that a partial destruction of
broadening of the signals occurs and it can be caused by a high con- intermetallic compound Ni3Al occurs during DRM with the formation of
centration of microstrain and stacking faults in the sample, as well as by free nickel particles. The main effect of Pt and Ru consists of preventing
the presence of additional phases with similar parameters. It seems to the process of carbide formation and migration of carbon atoms inside
be quite logical, considering the essentially non-equilibrium nature of of the Ni3Al crystal lattice.
the self-propagating high-temperature synthesis.
As for the used catalyst, it shows primarily the peaks of Ni3Al, Ni-Al 3.6. Surface characterization
solid solutions and Ni, so the phase composition is not significantly
changed but several features appeared: 1) the Ni3Al phase still dom- 3.6.1. BET surface area
inates but its relative intensity decreased, 2) all the peaks became Since the catalysts on the base of intermetallic Ni3Al were prepared
sharper, indicating the annealing of defects and irregularities, 3) in- by SHS (which is characterized by high synthesis temperature
tensity of the Ni peak increased (implying either the partial destruction (Tadiabatic(Ni3Al) = 1293 °C)), their surface areas are expected to be
of the main phase of the intermetallic compound, or agglomeration of small. The BET surface areas of the intermetallides before and after
the dispersed nickel particles into larger ensembles). In addition, a DRM reaction are listed in Table 1. Initial surface areas ranged from
diffuse peak (within the range of small diffraction angles between 17 1,1 m2 g−1 for Ni3Al to 1,4 m2 g−1 Ru/Ni3Al-3). After testing of all the
and 25°) has appeared after catalytic testing during 24 h at 900 °C. It catalysts in DRM at high temperatures (between 600 °C and 900 °C) the
can be attributed to the phase of amorphous carbon. Apart from the surface areas decreased. In principle, it is expected that the Ni particles
basic phases, trace amount of corundum (less than a percent) are de- gradually aggregate and sinter on the surface of the catalyst. XRD in-
tected. vestigations confirmed the partial destruction of Ni3Al phase with Ni
In situ X-ray diffraction patterns of Ni3Al are shown in Fig. 6. The phase release. Nevertheless, the high catalytic activity per mass of
diffraction peaks of Ni3Al (200), Ni4-xAlx (200) and Ni (200) are pre- catalyst at such low specific areas is related to both the SHS materials’
sented. Since the increase in temperature provokes the expansion of the unique composition as well as their very high atomic defect con-
lattice, and respectively, it induces the peak shifting to the low-angle centration resulting from the conditions of self-propagating high-tem-
region, then, in order to prevent the temperature factor, we maintained perature synthesis.
the temperature at a constant level (850 °C, because the Anton Paar
oven chamber has temperature limitation) and performed DRM during 3.6.2. XPS measurements
the scheduled time (1.5 h). The sample was heated in He atmosphere The XP spectra of Ni2p of initial and used Ni3Al catalysts were
before the DRM reaction.No obvious changes are registrated in the XRD analysed. For initial sample Ni2p can be deconvoluted in three com-
patterns during DRM for the first 25–30 min, but after 45–50 min on ponents (at 852.04, 855.35 and 856.05 eV). These peaks are ascribed to

311
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

aluminides. Nevertheless, the most important results were obtained

during the SEM investigation of the samples used in DRM. The porous
structure of the initial samples remains only in Pt- and Ru-implanted
catalysts. The relatively smooth surface of the initial sample of Ni3Al is
altered considerably. Due to erosion and influence of intermediate ad-
sorbed substances, as well as the formation of nuclei of metallic nickel
phase and partial carbonization the surface has become friable. In
Fig. 8(b) the globular carbon clusters covering the surface of the matrix
are demonstrated. Carbon deposition may amount to 6–12% or some-
times even up to 16–22% (depending on the conditions) of the catalyst
and they may deactivate the catalyst by pore blocking. Carbon deposits
are not uniformly distributed also on the surface of Ni3Al, some areas
are completely covered by coke, while all the other areas are com-
pletely free from carbon deposits. It is based on supplementary EDS.
In the case of Pt-implanted catalysts, carbon deposition is almost not
Fig. 7. XP spectrum of Ru 3d5/2 for Ru/Ni3Al catalyst after DRM (850 °C, 24 h, the surface
observed (Fig. 8(d)). The only feature of such micrographs is the fact
was cleaned by argon ion beam bombardment).
that their contrast is slightly weaker, possibly, because of the adsorbed
molecules. The surface of the used Ru-implanted catalyst has inter-
1) Ni0, 2) Ni(II) (probably as NiO or NiAl2O4 due to the asymmetry of mediate (between Ni3Al and Pt/Ni3Al) structure (Fig. 8(f)), i.e. the
the signal), 3) Ni(II) as well. The Al2p spectrum corresponds to three strong structural changes (sintering, pore blockage or others) were not
chemical states for Al: 1) at 71.37 eV (metallic state of Al (Ni3Al)), 2) at observed, but the carbon deposits were detected in some places of the
73.96 eV corresponds to Al-oxide (AlOx or Al2O3) as amorphous com- catalytic surface. This is supposed to be the main reason of “light”
ponent, 3) at 76.15 eV is ascribed to possible formation of NiAl2O4 [4]. catalyst deactivation of this catalyst.
The deconvolution of spectra for Ni3Al used in DRM (at 900 °C, Thus, the catalysts composition does affect significantly the coke
24 h) led to the similar results for Ni2p (852.54, 855.80 and deposition. Ru, and especially Pt can enhance the rate of gasification of
857.05 eV). These signals can be associated with the same states of Ni. adsorbed carbon or carbon deposit precursors and depress the carbon-
In general, the XP spectra of Ni2p are relatively similar for modified and forming reactions.
unmodified catalysts but we can observe the shifting of BE (the main
signals: 852.56, 854.18 and 856.10 eV) and relative content of different 3.6.4. TEM investigation
states. The nickel state evolution from the initial state and then during In order to define the size, structure, morphology and state of ag-
DRM in catalysts has the opposite trends, i.e. Ni0 percentage decreases gregation of the metal particles, transmission electron microscopy in-
(8.62 → 7.21) in Ni3Al, increases (9.72 → 10.99) in Pt/Ni3Al catalyst, vestigations were performed. HRTEM revealed that the structure of
and increases in (9.23 → 10.13) in Ru/Ni3Al. initial catalyst based on Ni3Al includes different areas: 1) large areas
The most remarkable feature of XP spectra of Pt4d and Ru3d is that with the face-centered cubic lattice with the ordering strictly corre-
each of the metals does not change their oxidation state, both of them sponding to Ni3Al, 2) adjacent areas with higher interplanar spacing
are in the metallic state before and after DRM. For Pt/Ni3Al catalyst (0.210 and 0.211 nm) and the absence of distinctly determined inter-
only one species with a BE of 314.4 eV (corresponds to Pt0) is generally faces (because of the presence of structural defects and the regions with
formed on the catalyst surface [4]. In case of Ru/Ni3Al catalyst (Fig. 7), the coexistence of several phases), 3) small areas with a face-centered
two strong peaks appear at a binding energy of 280,6 eV and 284.6 eV. cubic lattice corresponding to Ni phase, 4) small areas with different
The first peak is associated with 3d5/2 of metallic ruthenium, and crystal lattice corresponding to α-, β- and δ-Al2O3.
the second one is ascribed to carbon. These results are well correlated After the catalytic tests in DRM, in addition to the above-mentioned
with the data of other researchers [62]. phases, carbon of different modifications was detected, for example,
Thus, ruthenium and platinum greatly improve the performance of amorphous carbon and well-crystallized graphite which is located
Ni3Al catalysts in DRM. The main reasons are: 1) better reducibility of around larger Ni particles (Fig. 9) of 30–50 nm. An increase in the re-
Ni+2 particles, 2) higher dispersion of the intermetallide (Table 1) and action temperature results in the decrease of d002 from 0.343 to
Ni particles, 3) noble metals in a highly dispersed state hinder diffusion 0.339 nm (from XRD data) which is close to that of perfect graphite
of carbon and thereby preventing deactivation by coking), 4) according (d002 = 0.335 nm).
to XPS, NiAl2O4 phase is formed during DRM. On one hand, this phase HRTEM examination also confirmed the existence of graphene
is not active in the reforming process and reduces the catalyst surface layers (with the thickness of 20–50 nm) and carbon nanofibers pro-
area, but on the other hand, it can play a positive role in small quan- duced from the interaction of Ni3Al catalyst with methane and carbon
tities because it has high thermal stability as well as may hinders the dioxide at 900 °C for 24 h. Rather small Ni particles provoke formation
coalescence of Ni particles and can also activate CO2 molecule for of filamentous carbon [1]. The average diameter of carbon filament and
dissociation [59]. the catalyst particle is identical (30–60 nm). As we discussed in para-
graph 3.6, the process of coke formation can be realized by the carbide
3.6.3. SEM observation cycle mechanism [60].
SEM investigations were performed for all the catalysts both before In Pt/Ni3Al and Ru/Ni3Al catalysts well-dispersed Pt and Ru parti-
and after catalytic tests in DRM for 24 h at 850 °C. Fig. 8 shows the cles with the size ranging from a few angstroms to 20–30 nm, are very
morphological features of Ni3Al, Pt/Ni3Al-3 and Ru/Ni3Al-3. The initial stable and maintain their shape and size after DRM even at high tem-
structures of all the catalysts are rough. The surfaces are typical splits of peratures above 800 °C. Fig. 10 demonstrates the location of Ru clusters
alloy. Some other areas are relatively smooth, with minor irregularities. distributed in such areas, where the phase of aluminum oxide domi-
The surfaces have many large pores with an average diameter of nated. In this observation metal crystal are highly dispersed.
5–30 μm and extensive cracks with an average width of 0.1–0.5 μm. The metal particle size is approximately of 5–35 nm only for visible
Modification by ion implantation does not significantly affect the SEM particles but majority of Ru particles are dispersed almost in atomic
morphology of the samples (Fig. 8(c,e)). The main structure and the state and are not detectable here. These Ru particles were formed
contrast of the matrix remained after implantation. It could be caused during the process of ion implantation and they are isolated. There is no
by a high strength and thermal stability of alloys based on nickel evidence of Ru clusters coalescence after 24 h on stream. Nevertheless,

312
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

Fig. 8. SEM micrographs of Ni3Al (a,b), Pt/Ni3Al

(c,d) and Ru/Ni3Al (e,f) catalysts: initial samples
(a,c,e), after CO2-CH4 reforming (b,d,f). (Time on
stream 24 h, T = 850 °C,V (CO2 + CH4)
= 100 cm3 min−1, CO2:CH4 = 1:1).

after a thorough inspection of various portions of the catalyst, existence quantitatively evaluated by DTA-TG. Depending on the time of opera-
of carbon deposits was proved in areas containing Ni particles. tion in DRM, as well as on the process temperature, the weight loss
The surface of Pt/Ni3Al catalyst was not covered with detectable amounted to 6–12%. For unmodified catalyst based on Ni3Al used in
carbon depositions. Instead, the weak coalescence of Ni particles was DRM for 24 h at 900 °C this amount was 6,03% at space velocity
found, which explains the increase of Ni peaks in XRD. 100 cm3 min−1 and 10,8% at 20 cm3 min−1. Carbon deposits over
Ni3Al catalyst were oxidized at 500–800 °C (one maximum at 590 °C
and another one at 680 °C). It can be associated with two different
3.6.5. DTA-TG measurements forms of deposits. The TG measurements of used Ru/Ni3Al revealed that
The amount of carbon depositions on the used catalysts was

313
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

Fig. 9. TEM and EDX spectrum of Ni3Al catalyst used

in DRM (time on stream 24 h, temperature range of
600–900 °C, V(CO2 + CH4) = 100 cm3 min−1,
CO2:CH4 = 1:1).

the content of carbon is in the range of 4–8%, depending on the oper- depth of more than 50–70 nm [63]. Ni particles in the Ni3Al structure
ating conditions. Carbon deposition on the used Pt/Ni3Al catalyst was are able to form graphitic carbon and then to be encapsulated by a
not detected by DTA-TG [4]. carbon layer (Fig. 9). In this case, the investigation of kinetics is the
Carbon deposits may vary from high molecular weight hydro- most appropriate step for future research in the field of development of
carbons to graphite, depending on the conditions under which the coke new type of catalysts on the base of intermetallides because the fun-
was formed [1,4,60,61,63]. In the case of Ni3Al containing the phase of damental principle for coke-insensitive reactions on metals is that de-
free nickel, two different forms of carbon deposits were detected, fila- activation rate depends greatly on the difference in rates of formation
mentous carbon and graphite [1]. Since not all form of carbon result in and gasification of carbon precursor [63].
loss of catalytic activity, it would be really important to reveal all the This work has shown that there are new materials having excellent
features of coke formation on the surface of intermetallic Ni3Al. At high properties which can be used for development new catalysts or supports
temperatures (800–900 °C for DRM) graphitic films encapsulate the for endothermic CO2-CH4 reforming. Moreover, we propose new effi-
metal surfaces and deactivation may be caused by means of precipita- cient technologies for modification of catalysts in order to overcome the
tion of atomic (carbidic) carbon dissolved in the Ni surface layers to a major current drawbacks of catalysts, i.e. carbonization.

Fig. 10. TEM micrograph and EDX spectra of Ru-implanted Ni3Al catalyst used in DRM (Time on stream 24 h, temperature range of 600–900 °C, V(CO2 + CH4) = 100 cm3/min,
CO2:CH4 = 1:1).

314
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

4. Conclusions References

Intermetallic Ni3Al is a modern engineering compound and belongs [1] Larisa Arkatova, Catal. Today 152 (2010) 170–176.
to the unique group of materials with a specific structure, strict order in [2] Nimir O. Elbashir, Claude Mirodatos, Anders Holmen, Dragomir B. Bukur, Catal.
Today 228 (2014) 1–4.
the crystal lattice, which is known as superalloy. The Ni3Al alloys are [3] C. Song, A.M. Gaffney, K. Fujimoto (Eds.), Proc. ACS Symp. Ser. American Chemical
superior to commercial alloys, especially for high-temperature pro- Society, Washington DC, 2002.
cesses, in an oxidizing and carburizing environment. The most attrac- [4] Larisa Arkatova, Oleg Pakhnutov, Alexandr N. Shmakov, Yury S. Naiborodenko,
Nikolai G. Kasatsky, Catal. Today 171 (2011) 156–167.
tive properties of Ni3Al intermetallics include: 1) high thermal con- [5] Jae Sung Choi, Kwang-Ik Moen, Young Gul Kim, Jae Sung Lee, Ched- Hyun Kim,
ductivity, 2) a high tensile and compression strength, 3) a high David L. Trimm, Catal. Lett. 52 (1998) 43–47.
corrosion resistance in oxygen and carbon enriched atmospheres up to [6] Kaoru Takeishi, Yoshimi Akaike, Appl. Catal. A Gen. 510 (2016) 20–26.
[7] T.M. Fleisch, A. Basu, R.A. Sills, J. Nat. Gas Sci. Eng. 9 (2012) 94–107.
1100 °C, 4) an excellent high temperature wear resistance, 5) a rela- [8] B. Fidalgo, J.A. Menendez, Syngas: Production, Application and Environmental
tively low density and 6) catalytic activity in decomposition of various Impact, (2013), pp. pp. 121–149 ISBN 978-1Q100-870-5.
chemical substances. These properties make them similar to ceramics. [9] J.M. Ginsburg, J. Pina, T.E.I. Solh, H.I. Lasa, Ind. Eng. Chem. Res. 44 (14) (2005)
4846–4854.
Nevertheless, unlike ceramic material, intermetallides possess excellent
[10] Muhammad Awais Naeem, Ahmed Sadeq Al-Fatesh, Wasim Ullahkhan, Ahmed
thermal and electric conductivity, which allows to maintain the same Elhag Abasaeed, A.H. Fakeeha, Int. J. Chem. Eng. Appl. 4 (5) (2013) 315–320.
temperature along the entire catalyst bed. That is why Ni3Al was chosen [11] W.K. Jozwiak, M. Nowosielska, J. Rynkowski, Appl. Catal. A 280 (2005) 233–244.
to fulfill the role of a catalyst or catalytic support for DRM. Moreover, in [12] W.D. Zhang, B.S. Liu, Y.L. Tian, Catal. Commun. 8 (2007) 661–667.
[13] M. Bradford, M. Vannice, Catal. Rev. Sci. Eng. 41 (1999) 1–42.
order to obtain the matrix structure that ensures partial delivery of [14] M.C. Bradford, M.A. Vannice, J. Catal. 173 (1998) 157–171.
nickel as an active component, containing a large number of defects and [15] Keisuke Abe, Genki Saito, Takahiro Nomura, Tomohiro Akiyama, Energy Fuel 30
irregularities, SHS method was chosen as a method of catalyst pre- (10 2016) 8457–8462.
[16] Jean-Michel Lavoie, Front. Chem. 2 (2014) 1–17.
paration since it has some good advantages, such as: simplicity, low [17] B.S. Liu, C.T. Au, Appl. Catal. A 244 (2003) 181–195.
power consumption, simple and small-sized equipment, high-perfor- [18] N.N. Sazonova, V.A. Sadykov, A.S. Bobin, S.A. Pokrovskaya, E. Gubanova, React.
mance, high purity of products and environmental safety. Kinet. Catal. Lett. 98 (1) (2009) 35–41.
[19] J.R. Rostrup-Nielsen, J.H. Bak-Hansen, J. Catal. 144 (1993) 38–49.
It was found that it is better to use coarser nickel powders [20] H.Y. Wang, E. Ruckenstein, Appl. Catal. A 204 (2000) 143–152.
(45–70 μm) instead of fine Ni powder (˂10 μm) in order to improve the [21] K. Nagaoka, K. Seshan, K. Aika, J. Lercher, J. Catal. 197 (2001) 34–42.
characteristics of catalytic material after preparation by SHS and to [22] J.C.B. Wang, S.Z. Hsiao, T.J. Huang, Appl. Catal. A 246 (2) (2003) 197.
[23] A.N.J. van Keulen, M.E.S. Hegarty, J.R.H. Ross, P.F. van den Oosterkamp, Stud.
avoid (or to reduce) the effect of material shrinkage. Surf. Sci. Catal. 107 (1997) 537–546.
The Ni3Al matrix experiences weak phase transformations during [24] F. Pompeo, N.N. Nichio, M.V.V.M. Souza, D.V. Cesar, O.A. Ferretti, M. Schmal,
DRM, supplying pure Ni nanoparticles, which serve as the active sites Appl. Catal. A 316 (2007) 175–183.
[25] Takayuki Komatsu, Tomoyuki Uezono, J. Jpn. Petrol. Inst. 48 (2) (2005) 76–83.
for methane and carbon dioxide activation. Catalytic tests showed ra-
[26] Y. Ma, Y. Xu, M. Demura, T. Hirano, Appl. Catal. B 80 (2007) 15–23.
ther good activity of Ni3Al matrix in DRM at temperatures of [27] K. Morsi, Mater. Sci. Eng. A 299 (2001) 1–15.
800–900 °C. Nevertheless, this catalyst undergoes partial deactivation [28] Larisa Arkatova, Proc. 17 Nordic Symp. on Catalysis, Lund, Sweden, 2016, pp.
because of carbon deposition in the form of graphitic layers and fila- 227–228.
[29] A. Merzhanov, I. Borovinskaya, V. Shkiro, Synthesis of Refractory Inorganic
mentary carbon. In some cases, the amount of carbon reached 12%, Compounds USSR Inventor’s Certificate, 255 221 (1967), Byull. Izobr., 10 (1971),
depending on the temperature and reaction time. Fr. Patent 2088668 (1972), US Patent 3726643 (1973), UK Patent 1 321 084, Jpn
Further modification has been performed by means of ion im- Patent 1 098 839 (1982).
[30] E.A. Levashov, A.S. Rogachev, V.V. Kurbatkina, Yu.M. Maximov, V.I. Yuhvid,
plantation. Ru and Pt were implanted in the Ni3Al matrix in a highly Advanced Materials and Technologies of SHS, MISiS, Moscow, 2011.
dispersed state by controlling the homogeneity of their distribution. [31] V.I. Itin, Yu.S. Naiborodenko, Self-Propagating High-Temperature Synthesis of
Combination of SHS and ion implantation gave several advantages of Intermetallic Compounds, TSU, Tomsk, 1989.
[32] A.G. Merzhanov, A.S. Mukasyan, Solid-Flame Combustion, Torus Press, Moscow,
our catalysts: (1) the presence of active components (Ru, Pt) in an ul- 2007.
tradispersed state, (2) homogenous distribution of active components, [33] Galina Xanthopoulou, Proc. XI Int Symp. of SHS, September 5–9, Greece, 2011, pp.
(3) incorporation of Ru and Pt into the Ni3Al crystal lattice, (4) a large 23–24.
[34] K. Morsi, Mater. Sci. Eng. A 299 (2001) 1–15.
number of structural defects and dislocations, (4) high thermal con-
[35] L.A. Arkatova, L.N. Kurina, Russ. J. Phys. Chem. A 84 (1) (2010) 19–24.
ductivity, thus avoiding local overheating of the catalysts and their [36] L.A. Arkatova, Russ. J. Phys. Chem. A 84 (4) (2010) 566–572.
further destruction. [37] A.G. Nikolaev, E.M. Oks, K.P. Savkin, G.Yu. Yushkov, I. Brown, R. MacGill, Patent
RU 48105 U1, 2005.
The catalytic tests showed that in the case of Ni3Al intermetallic
[38] I.G. Brown, Rev. Sci. Instrum. 65 (1994) 3061–3081.
samples the presence of Ru, and especially Pt, strongly enhances the [39] T.B. Massalski, Al-Ni, Binary Allow Phase Diagrams, 2nd ed., ASM International
catalytic activity, with an effect which increases with the rise of the Ru Metals Park, USA, OH, 1990, p. 181 v. 1.
and Pt content. The observed order of CH4 and CO2 conversions over [40] Pawel Jozwik, Wojciech Polkowski, Zbigniew Bojar, Materials 8 (2015) 2537–2568.
[41] K. Hilpert, D. Kobertz, V. Venugopal, M. Miller, H. Gerads, F.J. Bremer, H. Nickel, Z.
doped Ni3Al catalysts is Pt/Ni3Al-3 > Ru/Ni3Al-3 > Pt/Ni3Al- Naturforsch. 42a (1987) 1327–1333.
2 > Pt/Ni3Al-1 ≥ Ru/Ni3Al-2 > Ru/Ni3Al-1 > Ni3Al. The most ac- [42] L.A. Arkatova, T.S. Kharlamova, L.V. Galaktionova, L.N. Kurina, V.N. Belousova,
tive catalyst, Pt/Ni3Al-3 (Pt doze 1*1017 i cm−2), did not undergo de- Yu. S. Naiborodenko, N.G. Kasatsky, N.N. Golobokov, Russ. J. Phys. Chem. A 80
(2006) 1231–1236.
activation because of coking. This catalyst was very active and stable [43] N.P. Lyakishev, Phase Diagrams of Double Metal Systems, Mashinostroenie,
during 120 h in DRM. Moreover, this catalyst works without any ap- Moscow, 1996.
parent activation stage. [44] A. Biswas, S.K. Roy, Acta Mater. 52 (2004) 257–270.
[45] Y. Iwasawa, Tailired Metal Catalysts, D. Reidel Publishing Company, The University
of Tokyo, Japan), 1986.
Acknowledgements [46] O.S. Rabinovich, P.S. Grinchuk, M.A. Andreev, B.B. Khina, Physica B 392 (2007)
272–280.
[47] Boris B. Khina, J. Appl. Phys. 101 (2007) 063510, , http://dx.doi.org/10.1063/1.
This work was supported by the Department of Structural
2710443.
Macrokinetics of Tomsk Scientific Center (Siberian Branch of the [48] Alexander S. Rogachev, Alexander S. Mukasyan, Combustion for Material Synthesis,
Russian Academy of Sciences),. We gratefully acknowledge Prof. E.M. CRC Press, 2014.
[49] S.V. Kositsin, I.I. Kositsina, Usp. Fiz. Met. 9 (2008) 195–258.
Oks, Prof. G.Yu. Yushkov, and Dr. K.P. Savkin (Institute of High Current
[50] K. Arishtirova, B. Pavelec, R. Nikilov, J.L.G. Fierro, S. Damyanova, React. Kinet.
Electronics, Siberian Branch of the Russian Academy of Sciences) for Catal. Lett. 91 (2) (2007) 241.
the help in ion implantation and Dr. I.P. Prosvirin (Boreskov Institute of [51] Ahmed M. Gadalla, Barbara Bower, Chem. Eng. Sci. 43 (11 1988) 3049–3063.
Catalysis, Siberian Branch of the Russian Academy of Sciences) for XPS [52] Z. Shariatinia, Y. Khani, F. Bahadoran, Int. J. Environ. Sci. Technol. 13 (2016)
423–434.
measurements.

315
L.A. Arkatova et al. Catalysis Today 299 (2018) 303–316

[53] Alfred B. Anderson, John J. Maloney, J. Phys. Chem. 92 (3) (1988) 809–812. [59] L.A. Arkatova, Chem. Sustain. Dev. 19 (2011) 1–15.
[54] P.D. Szuromi, J.R. Engstron, W.H. Weinberg, J. Phys. Chem. 89 (1985) 2497–2502. [60] Inorganic Nanoparticles: Synthesis, Applications and Perspectives, in:
[55] T.V. Choudhary, D.W. Goodman, J. Mol. Cat. A 163 (2000) 9–18. Claudia Altavilla, Enrico Ciliberto (Eds.), CRC Press, Taylor & Fransis Group, Boca
[56] T.P. Beebe, D.W. Goodman, B.D. Kay, J.T. Yates, J. Chem. Phys. 87 (1987) Raton, London, New York, 2011.
2305–2315. [61] H. Papp, P. Schuler, Q. Zhuang, Top. Catal. 3 (1996) 299–311.
[57] P.M. Holmblad, J. Wambach, I. Chorkendorff, J. Chem. Phys. 102 (1995) [62] Du Juan, Yang Xiao-xi, Ding Jing, Wei Xiao-lan, Yang Jian-ping, Wang Wei-long,
8255–8263. Yang Min-lin, J. Cent. South Univ. 20 (2013) 1307–1313.
[58] C. Crisafulli, S. Scire, R. Maggiore, S. Minico, Galvagno, Cat. Lett. 59 (1) (1999) [63] Morris D. Argyle, Calvin H. Bartholomew, Catalysts 5 (2015) 145–269.
21–26.

316