Microwave-activated carbon dioxide reforming of propane over Ni/TiO2 catalysts

Article abstract of DOI:10.1007/s11172-016-1662-y

The use of microwave activation in Ni/TiO₂-catalyzed carbon dioxide reforming of propane increases the catalytic activity and significantly reduces the coke formation in comparison with conventional thermal heating. During microwave activated reaction, C₂—C₃ olefins were formed, apart from CO and H₂, and the selectivity to olefins reached 6%. It was suggested that exposure to microwave radiation may induce local high-temperature heating of catalytically active phases and catalyst sites, which is not inherent in conventional heating. According to X-ray absorption spectroscopy (XAS = XANES + EXAFS), unlike conventional thermal heating in a hydrogen flow, on exposure to microwave radiation, the Ni²⁺ cations are partly reduced to Ni⁰.

Full text of DOI:10.1007/s11172-016-1662-y

Russian Chemical Bulletin, International Edition, Vol. 65, No. 12, pp. 2820—2824, December, 2016
2820
Microwaveꢀactivated carbon dioxide reforming
of propane over Ni/TiO2 catalysts
A. L. Tarasov,a O. P. Tkachenko,a O. A. Kirichenko,a and L. M. Kustova,b
аN. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences,
47 Leninsky prosp., 119991 Moscow, Russian Federation.
Eꢀmail: lmk@ioc.ac.ru
bDepartment of Chemistry, M. V. Lomonosov Moscow State University,
Build. 3, 1 Leninskie gory, 119992 Moscow, Russian Federation.
The use of microwave activation in Ni/TiO2ꢀcatalyzed carbon dioxide reforming of proꢀ
pane increases the catalytic activity and significantly reduces the coke formation in comparison
with conventional thermal heating. During microwave activated reaction, C2—C3 olefins were
formed, apart from CO and H2, and the selectivity to olefins reached 6%. It was suggested that
exposure to microwave radiation may induce local highꢀtemperature heating of catalytically
active phases and catalyst sites, which is not inherent in conventional heating. According to
Xꢀray absorption spectroscopy (XAS = XANES + EXAFS), unlike conventional thermal heating
in a hydrogen flow, on exposure to microwave radiation, the Ni2+ cations are partly reduced to Ni0.
Key words: carbon dioxide reforming, propane, nickel, catalysts, Xꢀray absorption spectroscopy.
Carbon dioxide reforming of methane can be promisꢀ
The carbon dioxide reforming of propane is also
a promising process. The catalysts tested for reforming of
ing for industrial production of hydrogen, synthesis gas
with a specified H2/CO ratio in the "gasꢀto liquid" (GTL)
processes, and for greenhouse gas disposal.1 Carbon diꢀ
oxide reforming of light C1—C3 paraffins to syngas is an
endothermic catalytic transformation of hydrocarbons. As
an example, consider the reforming of methane:
5
propane include Mo—Ni/Al2O3 and chromium oxideꢀ6
and gallium oxideꢀbased7 systems. The catalytic reducꢀ
tion of carbon dioxide accompanied by oxidative deꢀ
hydrogenation of propane over Cr2O3, ZnO, or Ga2O3
catalyst is described by the following equation:
CH4 + CO2 = 2 CO + 2 H2 (ΔH298K = 247 kJ mol–1).
CO2 + C3H8 = CO + C3H6 + H2O.
It was found that this reaction proceeds without forꢀ
mation of side gaseous products such as methane or ethane
only in the presence of gallium oxide.7
For this reaction to proceed to a high conversion,
a temperature above 700 °C and a catalyst such as Co/SiO2,
Co/MgO, or Co/CaO are required.2 Some authors used
the Ni—MexOy/γꢀAl2O3 compositions as catalysts.3 The
redox properties of the surface were controlled by modifyꢀ
ing Ni/γꢀAl2O3 by weakly basic oxides with a moderate
redox potential (MexOy = CeO2, La2O3, etc.). The addiꢀ
tion of CeO2 to the Ni/γꢀAl2O3 catalyst decreases the
contribution of the water shift reaction, which hence inꢀ
creases the process selectivity to CO. Furthermore, surꢀ
face modification with oxides increases the rate of interꢀ
action of carbon fragments with reactive oxygen species
on the surface and with the oxide lattice oxygen; this supꢀ
presses the carbon deposition on the catalyst surface.
Previously, carbon dioxide reforming of ethane in the
presence of nanostructured cesium oxide has been deꢀ
scribed.4 In this case, carbon monoxide formation is
accompanied by the parallel oxidative dehydrogenation of
ethane to give ethylene:
An important problem of carbon dioxide reforming is
coke formation. Fast catalyst deactivation induced by
coking necessitates the search for new bimetallic cataꢀ
lysts. It was shown8 that addition of minor amounts of
molybdenum considerably increases the stability of Ni/Al2O3
catalysts. For successful synthesis of efficient catalysts, it
is important to obtain highly dispersed metals (Ni, Cr)
with active surface, which would be stable in highꢀtemꢀ
perature processes and as inert to coking as possible. Also,
the efficiency of reforming increases in the presence of
carbon dioxide as a result of partial hydrogen release from
the process and decrease in the coking intensity.9
Our previous studies10 demonstrated that the use of
microwave energy in situ for reactions with noble metalꢀ
based catalysts can increase their catalytic activity 3—5ꢀ
fold and decrease the temperature of the onset of dehydroꢀ
genation of polycyclic hydrocarbons, for example, terꢀ
cyclohexane dehydrogenation, by 100—200 °C.11 The key
C2H6 + CO2 = C2H4 + CO + H2O.
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 12, pp. 2820—2824, December, 2016.
1066ꢀ5285/16/6512ꢀ2820 © 2016 Springer Science+Business Media, Inc.

Microwaveꢀactivated CO2 reforming over Ni/TiO2
Russ.Chem.Bull., Int.Ed., Vol. 65, No. 12, December, 2016 2821
the temperature was controlled by moving the reactor in the
microwave oven chamber, which had an electric field gradient
along the height. In the case of thermal heating, the temperature
was specified by a Termodat 17 controller.
The amount of carbon deposits was determined by weighing
the reactor before and after the experiment.
advantage of microwave energy transfer over thermal heatꢀ
ing is that the energy is supplied through radiation rather
than through heat transfer. This enables fast energy peneꢀ
tration into the bulk of materials (catalysts) that absorb
the microwave radiation. A previous study12 defined a
number of supports and catalysts that are able to absorb
microwave radiation to be heated to high temperatures
needed for carbon dioxide conversion of propane. These
materials include metals (Pt, Ni) supported on carbon,
metal (Ta, W) carbides, and titanium oxide.
The purpose of this study was to compare the carbon
dioxide reforming of propane carried out with thermal and
microwave heating of the Ni/TiO2 catalyst. The choice
of the catalyst was due to the fact that nickel is traditioꢀ
nally used in the carbon dioxide reforming reactions of
lower C1—C3 paraffins, while the TiO2 support can absorb
the microwave energy, being thus heated to high temꢀ
peratures.
The gas at the reactor outlet was analyzed on a Crystallux
chromatograph with a heat conductivity detector and two colꢀ
umns: 5 Å molecular sieves (2 m) for quantitative determination
of H2, CH4, and CO and HayeSepꢀQ (3 m) for analysis of CO,
CO2, C2H4, C2H6, C3H6, and C3H8. For quantitative analysis of
gases (vol.%), two loops of invariable volume were used. Figure 1
shows an example of chromatograms of reaction products conꢀ
taining all of these gases.
XꢀRay absorption spectroscopy (XAS) of the Ni(5%)/TiO2
catalyst. Nickel Kꢀedge Xꢀray absorption spectra (8333 eV) were
measured on a HASYLAB synchrotron (E4 station) (DESY,
Hamburg) using a twoꢀcrystal Si(111) monochromator tuned to
70% of the maximum intensity (to eliminate the higher harmonꢀ
ics from the Xꢀray beam). The spectra were recorded in the
transmission mode at liquid nitrogen temperature to reduce the
Debye—Waller disorder factor. The samples were pressed into
pellets 13 mm in diameter. For energy calibration, a nickel foil
was placed between the second and third ionization chambers
during recording the spectra. As standard nickel compounds, Ni
foil, NiO, and Ni2(OH)2CO3 were used; their spectra were reꢀ
corded under similar conditions.
Experimental
Preparation of the catalysts. TiO2 powder (Pꢀ25, Degussa,
phase composition: anatase (75%) and rutile (25%), specific surꢀ
face area 50 m2 g–1) served as a support. The nickel precursor
was deposited on the support surface and then reduced to Ni0.
For this purpose, an excess of the initial solution containing
0.37 mol L–1 of nickel nitrate and 3.3 mol L–1 of urea was added
in portions to the support. The volume of the solution was taken
in such a way that the final catalyst contained 5 wt.% Ni. The
initial solution was acidified to pH 3.4 with dilute HNO3 in
order to avoid Ni2+ adsorption on the support during preparaꢀ
tion of the suspension. The suspension thus formed was stirred
for 6 h at 96 °C. Thermal hydrolysis of urea resulted in precipitaꢀ
tion of the poorly soluble highly dispersed nickel precursor. Therꢀ
mogravimetric data for the precipitate obtained by similar heatꢀ
ing of the initial solution without the support suggests that the
precursor is nickel hydroxy carbonate (Ni2(OH)2CO3). The susꢀ
pension was cooled down and filtered. The precipitate was
washed many times with cold water with decantation, stirred for
1 h in hot water, separated on a filter, and additionally washed
on the filter with cold water. Then the resulting Ni(5%)/TiO2
catalyst sample was dried for 12 h at 110 °C .
The data were processed using the VIPER program.13 The
amplitudes and scattering phases of the neighboring atoms were
calculated by the ab initio FEFF8.10 program14 for NiO and Ni
foil. The experimental and theoretical spectra were compared in
both the reciprocal (k, wavenumber) and real (r, coordinate)
spaces. The shell radius (distance to the nearest atom), coordiꢀ
2
nation number, Debye—Waller factor (σ ), and the deviation
between the calculated and experimental energetic absorption
edge positions (ΔE) used as the fitting parameters to fit the theoꢀ
retical spectra to the experimental data.
Three catalyst samples were studied: one dried at 110 °C,
one reduced in an H2 flow at 350 °C (conventional heating), and
one reduced at 310 °C (microwave heating).
V/mV
290
Commercial catalysts, NiO(6—10%)/Al2O3 (GIAPꢀ8), and
Cr2O3(5—8%)/Al2O3 (GIAPꢀ14) manufactured by Maxamꢀ
Chirchiq (Uzbekistan) were used as references.
230
CO2
CH4
CO
Catalytic activity measurements. The catalyst (1 g) was
charged into a flow type reactor, which was a quartz tube with an
internal diameter of 7 mm. The catalyst reduction and carbon
dioxide reforming of propane were carried out using microwave
heating or conventional thermal heating. The reactor was placed
into an electrically heated furnace or, when operating with the
Ni(5%)/TiO2 catalyst, into a Vigor domestic microwave oven.
170
110
CO
C2H4
50
CH4
H2
C3H8
C3H6
C2H6
Then the catalyst was heated in a hydrogen flow (15 mL min–1
)
to 310—350 °C and kept for 2 h. Then the H2 flow was replaced
by C3H8 : CO2 = 1 : 1 (mol.) fed at a space velocity of 4000 h–1
and the catalysts were heated to the reaction temperature.
The temperature was measured by thermocouple placed in
the middle of the catalyst bed. In the case of microwave heating,
2
4
6
8
10 12 14 16 18 20 t/min
Fig. 1. Example of chromatograms of the products of propane
reforming catalyzed by Ni(5%)/TiO2 (0—4 min, analysis of 5 Å
sieves; 4—21 min, analysis on a HayeSepꢀQ column).

2822
Russ.Chem.Bull., Int.Ed., Vol. 65, No. 12, December, 2016
Tarasov et al.
D
Results and Discussion
1
2
3
4
5
6
2.0
The NiO/Al2O3 and Cr2O3/Al2O3 catalytic systems
used in this study for comparison with the Ni(5%)/TiO2,
catalyst cannot be heated in a microwave field; therefore,
they were tested only after thermal heating. Table 1 sumꢀ
marizes characteristics of the carbon dioxide reforming of
propane for the test samples and experimental conditions.
It can be seen from Table 1 that in the presence of the
thermally heated (700 °C) catalyst, C2—C3 olefins are
formed only over the Cr2O3/Al2O3 sample, which is typiꢀ
cal of aluminum chromium catalysts.9 The reactions catꢀ
alyzed by Ni/TiO2 and Ni/Al2O3 give mainly syngas with
the H2 : CO > 2 ratio in the gas products, which is favorꢀ
able for the subsequent use of this mixture in methanol
synthesis. However, the reaction with nickel catalysts is
accompanied by substantial coke deposition, which hamꢀ
pers the practical use of these catalysts.15
1.5
1.0
0.5
8330 8340 8350 8360 8370 8380 E/eV
Fig. 2. Normalized XANES spectra of the nickel reference comꢀ
pounds (1, Ni foil; 2, Ni2(OH2)CO3; 3, NiO) and Ni catalyst
(5%)/TiO2 (4, initial; 5, after microwaving, 310 °C, H2; 6, after
conventional thermal treatment, 350 °C, H2).
Unlike thermal heating, microwave activation of the
Ni(5%)/TiO2 catalyst results in small amounts of C2—C3
olefins in the reaction products. In the presence of microꢀ
waveꢀactivated Ni(5%)/TiO2, the conversion of propane
and CO2 is considerably higher at the same temperature,
while the carbon deposition on the catalyst decreases more
than twofold. This finding suggests that microwave actiꢀ
vation of the carbon dioxide reforming is a promising trend
for development of novel heterogeneous catalysts.
A comparative study of the electronic state and local
environment of nickel in the Ni(5%)/TiO2 sample after
conventional thermal heating and microwave activation
was performed by Xꢀray absorption spectroscopy (XAS =
XANES + EXAFS). Figure 2 shows the K XANES spectra
of the Ni(5%)/TiO2 catalyst and reference nickel comꢀ
pounds normalized to one Ni atom.
the spectrum of Ni2(OH)2CO3. Hence, the active sites in
this sample exist as Ni2+ cations in the same coordination
state as nickel ions in the presumed Ni2(OH)2CO3 precurꢀ
sor. The reduction of the dried sample in an H2 flow in
a microwave field at 310 °C and conventional reduction in
an H2 flow at 350 °C result in decreasing intensity of the
white line.
Figure 3 shows the Ni K EXAFS spectra (Fourier transꢀ
form of the oscillating part of Xꢀray absorption spectra)
for the Ni(5%)/TiO2 catalyst and reference nickel comꢀ
pounds.
The EXAFS spectra of the initial dried Ni(5%)/TiO2
catalyst differ from the spectra of Ni foil and NiO and
resemble the spectrum of Ni2(OH)2CO3. The spectrum of
the catalyst pretreated in air at 110 °C exhibits two peaks
at an uncorrected distance of ~1.5 and ~2.5 Å. The posiꢀ
tions of these peaks correspond to the peaks in the specꢀ
The spectrum of the initial dried sample (see Fig. 2)
differs from the spectra of Ni foil and NiO and resembles
Table 1. Characteristics of carbon dioxide conversion of propane at 650 °C (gas mixture space velocity 4000 h–1, C3H8 : CO2=
= 1 : 1, mol.)
Characteristics
Ni(5%)/TiO2
Ni(5%)/TiO2
NiO(6—10%)/Al2О3
Сr2О3(5—8%)/Al2О3
Type of heating
Т/°С
Microwave
650
Thermal
650
Thermal
650
Thermal
700
Gas composition (vol.%), including:
H2
СО
СH4
CO2
С2Н4
С2Н6
С3Н6
С3Н8
46.9
34.0
12.8
10.8
11.8
11.8
10.2
11.7
12.8
89.8
95.2
44.2
18.7
14.6
12.2
10.0
10.0
10.0
10.3
27.5
58.0
50.2
44.9
22.4
27.9
14.8
10.0
10.0
10.0
10.0
40.6
100.0
75.8
21.5
20.0
25.9
13.2
11.8
11.2
12.1
14.3
10.4
84.8
53.3
Carbon deposits (%)/h
Conversion of C3H8 (%)
Conversion of CO2 (%)

Microwaveꢀactivated CO2 reforming over Ni/TiO2
Russ.Chem.Bull., Int.Ed., Vol. 65, No. 12, December, 2016 2823
Table 2. EXAFS data for the Ni(5%)/TiO2 catalyst and references
2
Sample
Initial
Pair
Distance/Å
C.N.
σ •103/Å2
ΔE/eV
Ni—O
Ni—Ni
2.056 0.009
3.106 0.005
4.8 0.4
5.3 0.4
5 1
3 1
2 1
2 1
After treatment at 310 °C,
H2, microwave
Ni—O
Ni—Ni
Ni—Ni
Ni—O
Ni—Ni
2.073 0.005
2.514 0.010
3.089 0.006
2.080 0.006
3.075 0.008
4.2 0.2
0.9 0.1
6.8 0.5
4.2 0.2
8.5 0.8
7 1
4 1
12 1
7 1
2 1
2 1
2 1
2 0.7
2 1
at 350 °C, H2
10 1
trum of Ni2(OH)2CO3 that belong to Ni—O and Ni—Ni
atomic pairs. In the spectrum of the sample reduced at
310 °C in an H2 flow in a microwave field, a new peak
appears at ~2 Å, which coincides with the first peak in the
spectrum of metallic Ni foil (Ni—Ni atomic pair). In the
spectrum of the sample after conventional reduction at
350 °C in an H2 flow, this peak is absent.
Table 2 presents the average coordination numbers and
the real distances found in the calculation of the first and
second spheres.
According to calculation, the coordination number of
the Ni—O atomic pair is 4.8 0.4 in the initial (airꢀdried)
sample and 4.2 0.2 for the samples treated in H2 flow
both in a microwave field at 310 °C and in the conventionꢀ
ally reduced sample at 350 °C. The real distance in this
atomic pair (~2.06 Å) in the airꢀdried sample increases to
2.07 Å after its reduction in a microwave field and to
2.08 Å after conventional reduction at 350 °C. The real
distance in the Ni—Ni atomic pair characteristic of a simꢀ
ilar atomic pair in NiO is ~3.10 Å in the initial sample,
~3.09 Å in the sample reduced in the microwave field, and
~3.08 Å in the conventionally reduced sample. The calcuꢀ
lated data for the Ni—O atomic pair in the catalyst proved
to be close to the results obtained for the simulated
spectrum of the Ni2(OH)2CO3 reference, whereas data for
the Ni—Ni atomic pair similar to that observed in the
spectrum of nickel oxide were intermediate between the
data for the second coordination sphere for the two referꢀ
ences (Ni2(OH)2CO3 and NiO). The coordination numꢀ
ber for this Ni—Ni atomic pair increased from 5.3 0.4 to
6.8 0.5 upon the reduction of dried sample in the microꢀ
wave field and to 8.5 0.8 upon conventional reduction.
The calculation showed also that after reduction of the
catalyst sample in H2 in the microwave field, a peak inꢀ
herent in metallic nickel in Ni foil with a Ni—Ni distance
of ~2.52 Å appears in its spectrum. The coordination numꢀ
ber in this shell is close to 1.
Thus, microwaveꢀactivated carbon dioxide reforming
of propane catalyzed by Ni(5%)/TiO2 results in higher
propane and CO2 conversion and lower coking as comꢀ
pared with the thermally induced reaction and in appearꢀ
ance of C2—C3 olefins in the products. XꢀRay absorption
spectroscopy demonstrated that, unlike conventional thermal
heating in a hydrogen flow, on exposure to microwave
radiation, the Ni2+ cations are partly reduced to Ni0. A change
in the local environment of catalytically active sites affects
the selectivity of carbon dioxide reforming of propane.
This work was financially supported by the Russian
Science Foundation (Project No. 14ꢀ33ꢀ00001).
FT(χk2)
Ni—Ni
References
10
Ni—O
Ni—Ni
8
1. S. A. M. Said, M. Waseeuddin, D. S. A. Simakov, Renewꢀ
able and Sustainable Energy Reviews, 2016, 59, 149.
2. E. Ruckenstein, H. Y. Wang, Catal. Lett., 2001, 73, № 2, 99.
3. S. A. Solov´ev, E. V. Gubareni, Ya. P. Kurilets, Evraziiskii
Simpozium po Innovatsiyam v katalize i elektrokhimii [Eurꢀ
asian Symposium on Innovations in Catalysis and Electrochemꢀ
istry] 2010, p. 74 (in Russian).
4. R. X. Valenzuela, G. Bueno, A. Solbes, F. Sapiсa, E. Marꢀ
tínez, V. C. Corberán, Topics in Catalysis, 2001, 15, No. 2, 181.
5. A. Siahvashi, D. Chesterfield, Adesoji A. Adesina, Chem.
Engin. Sci., 2013, 93, 313.
6
5
4
6
4
2
2
1
3
2
4
6
8
d/Å
Fig. 3. EXAFS spectra of the nickel reference compounds (1, Ni
foil, 2, Ni2(OH)2CO3, 3, NiO) and Ni catalyst (5%)/TiO2
(4, initial; 5, after microwaving, 310 °C, H2; 6, after convenꢀ
tional thermal treatment, 350 °C, H2).
6. N. A. Gaidai, K. Esmaili, E. A. Smolenskii, T. N. Myshenꢀ
kova, Tekhnol. Nefti i Gaza [Oil and Gas Technologies], 2013,
86, No. 3, 29 (in Russian).

2824
Russ.Chem.Bull., Int.Ed., Vol. 65, No. 12, December, 2016
Tarasov et al.
7. P. Michorczyk, P. Kus´trowski, A. Kolak, M. Zimowska,
Catal. Commun., 2013, 35, 95.
8. A. Siahvashi, D. Chesterfield, A. A. Adesina, Chem. Engin.
Sci., 2013, 93, 313.
9. K. Esmaili, R. M. Kunusova, N. A. Gaidai, Tekhnol. Nefti i
Gaza [Oil and Gas Technologies], 2012, 5, 33 (in Russian).
10. L. M. Kustov, A. V. Zeigarnik, Proc. 7th Europ. Congr. on
Catalysis, Sofia, 2005, OFꢀ11.
13. K. V. Klementiev, VIPER for Windows (Visual Processing
in EXAFS Researches), freeware, www.desy.de/~klmn/
viper.html.
14. A. L. Ankundinov, B. Ravel, J. J. Rehr, S. D. Conradson,
Phys. Rev., 1998, 58, 7565.
15. C. H. Bartholomew, Appl. Catal. A: Gen., 2001, 212, 17.
11. J. S. Sung, K. Y. Choo, T. H. Kim, A. L. Tarasov, A. V.
Soldatov, L. M. Kustov, Int. J. Hydrogen Energy, 2008,
33, 4116.
12. Report on the State Contract of the Ministry of Science
and Education No. 11.519.11.5021 of October 12, 2011
(in Russian).
Received September 20, 2016;
in revised form October 18, 2016