Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process in conventional and membrane reactors

Article abstract of DOI:10.1016/j.cattod.2015.12.017

In this paper synthesis and catalytic properties of new catalysts based on double lithium-zirconium phosphate (LiZr₂(PO₄)₃) with monoclinic NASICON-type structure, doped by indium, niobium and molybdenum are discussed. The obtained samples with particle size of 50-300 nm were characterized by X-ray diffraction, scanning electron microscopy and X-ray microanalysis. The synthesized samples exhibit catalytic activity in the dehydration and dehydrogenation reactions of ethanol conversion. The main products were acetaldehyde, diethyl ether, hydrogen, C₂- and C₄-hydrocarbons. Indium- and molibdenum-doped samples were characterized by high activity in dehydrogenation processes, while niobium-doped was more active in dehydration processes. The highest selectivity in diethyl ether formation was achieved for LiZr₂(PO₄)₃ and Nb-doped samples (90 and 60% at 300°C). The highest hydrogen yield (up to 60%) was obtained with the use of In-doped catalyst. LiZr₂(PO₄)₃ and Mo-doped samples are also noticeable for high C₄-hydrocarbons formation, selectivity to which reaches 60% at 390°C. Use of a 100% hydrogen selective palladium-ruthenium alloy membrane increases hydrogen yield by 20%.

Full text of DOI:10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol
conversion process in conventional and membrane reactors
Andrey B. Ilin∗, Natalia V. Orekhova, Margarita M. Ermilova, Andrey B. Yaroslavtsev
A.V Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, Leninsky Pr. 29, Moscow, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
In this paper synthesis and catalytic properties of new catalysts based on double lithium-zirconium
phosphate (LiZr2(PO4)3) with monoclinic NASICON-type structure, doped by indium, niobium and molyb-
denum are discussed. The obtained samples with particle size of 50–300 nm were characterized by
X-ray diffraction, scanning electron microscopy and X-ray microanalysis. The synthesized samples
exhibit catalytic activity in the dehydration and dehydrogenation reactions of ethanol conversion. The
main products were acetaldehyde, diethyl ether, hydrogen, C2- and C4-hydrocarbons. Indium- and
molibdenum-doped samples were characterized by high activity in dehydrogenation processes, while
niobium-doped was more active in dehydration processes. The highest selectivity in diethyl ether for-
mation was achieved for LiZr2(PO4)3 and Nb-doped samples (90 and 60% at 300 ◦C). The highest hydrogen
yield (up to 60%) was obtained with the use of In-doped catalyst. LiZr2(PO4)3 and Mo-doped samples are
also noticeable for high C4-hydrocarbons formation, selectivity to which reaches 60% at 390 ◦C. Use of a
100% hydrogen selective palladium-ruthenium alloy membrane increases hydrogen yield by 20%.
© 2016 Elsevier B.V. All rights reserved.
Received 20 September 2015
Received in revised form
15 December 2015
Accepted 19 December 2015
Available online xxx
Keywords:
Nasicon catalyst
Membrane catalysis
Ethanol
Dehydration
Dehydrogenation
1. Introduction
radiation stability. The variability of structure allows both iso- and
heterovalent doping [6,7]. This defines a wide range of possible
Low-temperature fuel cells are the most promising alternative
energy sources. For their implementation the new ways of hydro-
gen production based on renewable raw materials are developed.
One of the most promising methods is bio-alcohols conversion. At
the same time, alcohols can be used for obtaining a number of
commercially important products, such as hydrocarbons, ethers,
aldehydes, ketones, hydrogen, and so on. In this regard, new cat-
alysts for the selective alcohols conversion into desired products
need to be found.
Complex phosphates are considered as one of the promising
types of catalysts for the alcohols conversion. Materials with NASI-
CON (NA Super Ionic CONductor) structure type have general
formula AxB2(ZO4)3, where A - is alkali or alkaline earth element,
B - polyvalent element (Zr, Ti, Sc etc.), Z - phosphorus or silicon.
Their structure is presented by edge-linked BO6 octahedra and ZO4
tetrahedra. A-cations are located in the cavities of this structure
[1,2]
applications of these compounds [8–10] as lithium power sources
[11,12], in fuel cells [13], in sensors [14–16], in radionuclide clean-
ing [17] and in catalysis [18–20].
The ability to heterovalent substitution is important for the cat-
alytic applications, since it allows to vary the number and strength
of acid (Brønstedand Lewis) and redox centers at the surface. Chem-
ical and thermal stability allows to use catalysts based on such
compounds in conditions in which the metal catalysts are degraded.
Therefore, attempts to use these compounds as catalysts in alcohol
conversion were made by the number of authors [21–28].
During the catalytic process certain products could be selec-
tively separated or fed to the reaction mixture through membrane
[29–31]. The main advantage of this method is the possibility
of thermodynamic equilibrium shifting due to removal of prod-
uct from reaction mixture. This is especially evident in reactions
with hydrogen participation when membranes based on palla-
dium alloys are used, because of their high hydrogen permeability
[32–34]. Hydrogen, obtained in such reactors, requires no further
purification due to high selectivity of such membranes. Nowadays
many researches are devoted to carrying out catalytic processes in
membrane reactors [35–39].
Interest in these compounds is determined primarily by their
high ionic conductivity [3–5], which is achieved due to presence
of conduction channels in their structure [2]. Other advantages
of the NASICON-type compounds are high chemical, thermal and
The aim of this work was to obtain the LiZr2(PO4)3 NASICON-
type phosphates with heterovalent substitution of the zirconium
and phosphorus ions and their testing in the process of ethanol
conversion. It was assumed that the substitution of some zirconium
∗
Corresponding author.
E-mail address: Novsel25@yandex.ru (A.B. Ilin).
http://dx.doi.org/10.1016/j.cattod.2015.12.017
0920-5861/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
2
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
tion equilibrium. For this purpose the obtained catalysts were
investigated in the membrane reactor with Pd-Ru membrane.
Advantages of this rector and membrane properties were discussed
also.
2. Experimental
In this work the synthesis of the following compounds
was carried out: LiZr2(PO4)3, Li1 0.1Zr1.9M0.1(PO4)3 (M = In
and Nb), LiZr1.8In0.1Nb0.1(PO4)3, Li0.9Zr2P2.9Mo0.1O12 and
Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12, referred to below as LZP, LZInP,
LZNbP, LZInNbP, LZPMo and LZInNbPMo, respectively. The synthe-
sis was performed according to the Pechini method [28,40–43].
For this purpose a corundum crucible was used. ZrOCl2 * 8H2O,
citric acid, Li2CO3 and NH4H2PO4 were successively dissolved in
a mixture of ethylene glycol (2 ml) and deionized water (10 ml).
The pH of the solution was quickly adjusted to 5.5 by adding con-
centrated ammonia solution to prevent precipitation of zirconium
phosphate. The resulting solution was kept in the oven at 95 ◦C
(24 h), 150 ◦C (24 h) and at 350 ◦C (4 h) for sequential removal of
water and other gaseous components. After thoroughly grinding
the resulting mixture was subjected to a final annealing at 750 ◦C
for 10 h to form a NASICON phases. For LZInP synthesis In2O3
was dissolved in a minimum amount of hot concentrated nitric
acid. This solution was added to ethylene glycol and deionized
water first. NbCl5 dissolved in the concentrated hydrochloric
acid was used as the niobium source for LZNbP. The mixture of
NH4H2PO4 with (NH4)6Mo7O24 was used instead of NH4H2PO4 for
LZPMo synthesis. The samples, simultaneously doped with several
elements, were obtained similarly.
Fig. 1. The scheme of the used membrane reactor.
X-ray diffraction (XRD) of samples was performed using X-ray
diffractometer Rigaku D/Max-2200 (CuK␣1 radiation), for spectra
processing and qualitative analysis the Rigaku Application Data
Processing software package was used. The particle size (coherent
scattering region) was estimated on the base of the X-ray diffraction
line broadening with the use of Scherrer equation:
Fig. 2. X-ray patterns of the synthesized compounds.
Li1.1Zr1.9In0.1(PO4)3, Li0.9Zr1.9Nb0.1(PO4)3, Li0.9Zr2P2.9Mo0.1O12
LiZr1.8In0.1Nb0.1(PO4)3, 6 - Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12
1
-
LiZr2(PO4)3,
2
-
-
3
-
4
-
, 5
.
k × ꢀ
d =
(1)
(B − b) × cos ꢁ
where k = 0.89–Scherrer constant, ꢀ = 1.5406 Å - the wavelength of
the radiation used, B - the half-width at half-maximum of peak
(2ꢁ), b - the instrumental broadening (2ꢁ), ꢁ - the angle of the peak
position. LaB6 powder (Standard Reference Material® 660a) was
used as the standard for determining the instrumental broadening.
The specific surface area was determined by the BET method
with the use of Micromeritics ASAP 2020. The analysis was carried
out in the relative pressures (p/po) area 0.01–0.99.
The micrographs of the samples were obtained by a scanning
electron microscope (SEM) Carl Zeiss NVision 40 with an attach-
ment for X-ray microanalysis. The accelerating voltage was 1 kV.
Catalytic properties were investigated in a conventional flow-
type quartz tube reactor in the helium or argon stream with
further identification of reaction products by chromatograph
Crystallux–4000 M with thermal conductivity detector, columns
HayeSep T 60/80 mesh (2 m, 150 ◦C, 30 cm3/min, He), SKT–6 (2 m,
150 ◦C, 30 cm3/min, He), Mole Seive 5 A (2 m, 25 ◦C, 30 cm3/min, Ar).
For these experiments, the 0.3 g of catalyst was mixed with milled
quartz (d = 250–500 ␮m) and placed in the reactor with a 6 mm
internal diameter so that the catalyst layer was 17 cm in length.
To create the desired concentration of ethanol vapor, the carrier
gas was passed at a volumetric flow rate of 20 cm3/min through a
bubbler with ethanol at 11 ◦C.
Fig. 3. A micrograph of a sample with composition Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12
.
ions on indium or niobium will change the acid-base properties of
catalysts and their activity in dehydration processes. At the same
time, replacement of the phosphorus on molybdenum will make
them active in redox dehydrogenation processes.
The doped lithium-zirconium phosphate with NASICON struc-
ture can have catalytic activity in dehydrogenation and dehydration
processes. It was assumed that a Pd-based membrane allows to
extract hydrogen from the reaction mixture and displace the reac-
The scheme of the used membrane reactor is shown on Fig. 1.
Palladium-ruthenium alloy foil (ruthenium proportion 6 mass%)
tightened between two copper seal rings was used as a membrane
with a thickness of 70 micron and active area of 7 cm2. 0.3 g of
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
3
Table 1
The specific surface area and particle size characteristics of the synthesized compounds, estimated by different techniques.
Composition
Specific surface area, m2/g
Particle size, nm
The average particle size estimated from the surface area, nm
LiZr2(PO4)3
16
11
12
20
25
1
1
1
1
1
49
81
73
40
51
55
2
4
4
2
2
3
87
174
157
94
74
229
Li1.1Zr1.9In0.1(PO4)3
Li0.9Zr1.9Nb0.1(PO4)3
Li0.9Zr2P2.9Mo0.1O12
LiZr1.8In0.1Nb0.1(PO4)3
Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12
8.0 0.8
catalyst mixed with 2.2 g of silica (fraction 250–500 ␮m) was placed
into the bottom part of the reactor. Two flows of the carrier gas
were passed at a volumetric flow rate of 20 cm3/min in each part.
The bottom flow was saturated with ethanol. In the experiment
with an impermeable foil a stainless steel Plate 1 mm in thickness
was used.
Acidity of catalysts was measured by thermoprogrammed des-
orption of NH3 with the use of sorption analyzer USGA-101.
0.1–0.2 g of the sample was calcined at 550 ◦C in dry helium and
then cooled down to 60 ◦C. NH3 adsorption was carried out at 60 ◦C
for 30 min using the NH3/N2 (1:1) mixture. Then the temperature
was increased to 100 ◦C in order to remove physically-adsorbed
NH3. The final measurements were carried out in 60–800 ◦C tem-
perature range in dry helium flow (30 ml/min) with heating speed
of 8◦/min.
3. Results and discussion
3.1. The structure of the catalysts
According to X-ray phase analysis, all samples synthesized are
single phase with monoclinic NASICON structure (Fig. 2). The calcu-
lated particle size, estimated with the use of X-ray diffraction line
broadening, ranged from 50 to 80 nm (see Table 1). According to
the SEM data their size consist 50–200 nm (Fig. 3).
The specific surface area of the catalysts synthesized lies in the
range of 8–25 m2/g (Table 1). The average particle size calculated
from these data exceeds the value obtained from X-ray diffraction
line broadening no less than 2 times (see Table 1). Along with the
SEM data this allows us to conclude that a significant fraction of
particles are aggregated. The indium and niobium doping leads to
the increase of particle size and the decrease in samples specific
surface area.
The conversion (X), selectivity (S), yield (Y) of the samples were
determined using the following formulas:
3.2. Catalytic transformation of ethanol in a quartz tube
flow-type reactor
ϕ0 − ϕ1
X =
S =
Y =
× 100%
× 100%
(2),
(3),
(4),
ϕ0
Catalytic tests carried out in the conventional flow-type tube
quartz reactor showed that all synthesized samples exhibit activity
in the dehydration and dehydrogenation reactions of ethanol.
The blank experiments, which were carried out only with quartz
particles (SiO2, 250–500 ␮m) allow us to conclude that the alcohol
conversion is catalyzed only by the synthesized phosphates (Fig. 4).
The main products of the processes are diethyl ether, acetaldehyde,
hydrogen, C2-hydrocarbons (ethylene and ethane, formed by the
hydrogenation of ethylene by hydrogen during the reaction) and
C4-hydrocarbons. CO, CO2 and CH4 formation was observed also,
but their quantities were rather small compared to the main
products.
ϕi
ϕ0 − ϕ1
ϕi
ϕ0 − ϕ1
where ϕ0 and ϕ1 are the initial and final volume fractions of alcohol,
respectively; ϕi is the part of alcohol consumed to form product i.
Permeation rate was calculated as volume of hydrogen in cm3,
passed through 1 cm2 of membrane per minute. The productivity of
the samples was calculated as the amount of product (in millimoles)
formed per one gram of catalyst per hour.
The sample with composition of LiZr2(PO4)3 is the most active
in the formation of C2- and C4-hydrocarbons and diethyl ether. It
shows the maximum ether yield, which reaches of 50% of all other
Fig. 4. The dependence of ethanol conversion versus temperature for catalysts with composition LiZr2(PO4)3 (1), Li1.1Zr1.9In0.1(PO4)3 (2), Li0.9Zr1.9Nb0.1(PO4)3 (3),
LiZr1.8In0.1Nb0.1(PO4)3 (4), Li0.9Zr2P2.9Mo0.1O12 (5), Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12 (6), SiO2 blank test (7).
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
4
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 5. Yield of C2-hydrocarbons (1), C4-hydrocarbons (2), acetaldehyde (3), diethyl ether (4) for the catalysts with composition LiZr2(PO4)3 (a), Li1.1Zr1.9In0.1(PO4)3, (b),
Li0.9Zr1.9Nb0.1(PO4)3 (c), Li0.9Zr2P2.9Mo0.1O12 (d), LiZr1.8In0.1Nb0.1(PO4)3 (e), Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12 (f).
products (Fig. 5a). Nb doping results in enhanced dehydration prop-
erties and causes the increase in the ethylene yield, which may be
associated with an increasing of the acidity of active sites due to
stable oxidation states (0 and +3). However, it is known that stabil-
ity of some monovalent indium compounds increases with rise of
a temperature.
For samples doped with several elements simultaneously ethy-
lene is the main product. It is the product of a monomolecular
dehydration reaction:
the higher PO4 groups polarization by Nb5+ ions. The doping by
3−
indium contrary leads to suppression of acid activity. At the same
time doping by indium and molybdenum increases the dehydro-
genation properties (yield of acetaldehyde). In the case of indium,
this result can be considered quite surprising since it has only two
C2H5OH → C2H4+H2O, ꢂH0298 = +28, 99 kJ/mol
(5)
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
5
Fig. 6. The temperature dependence of selectivity to C2-hydrocarbons (a), C4-hydrocarbons (b), diethyl ether (c) for catalysts with composition LiZr2(PO4)3 (1),
Li1.1Zr1.9In0.1(PO4)3 (2), Li0.9Zr1.9Nb0.1(PO4)3 (3), Li0.9Zr2P2.9Mo0.1O12 (4), LiZr1.8In0.1Nb0.1(PO4)3 (5).
At first glance, it is similar to the diethyl ether formation reac-
tion. But last reaction requires the simultaneous participation of
two ethanol molecules. It should be ensured by the close location
of the two centers with the dehydrogenation activity. In samples
containing both niobium and indium, activity of some dehydrat-
ing centers increases and decreases for others. So the probability
of close location of two active sites decreases. Therefore, despite
the high activity in reactions of ethylene formation, their activity
in ether formation is suppressed.
For the formation of C4-hydrocarbons the simultaneous pres-
ence of centers active in dehydration and dehydrogenation is
required. This reaction proceeds most active for the undoped and
doped by molybdenum samples. At the same time it is interest-
ing that in Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12, in which both types
of centers must be present, rate of this reaction is very small.
It can be explained by the positive charge of centers containing
molybdenum and niobium, which occupied phosphorus and zir-
conium positions respectively. Meanwhile indium ions located in
zirconium positions, in contrast, have a negative charge. There-
fore, the centers doped by molybdenum and indium have a
tendency to association, niobium and molybdenum—to repul-
sion. This makes it unlikely location of the two centers having
enhanced activity in the dehydration and dehydrogenation reac-
Li0.9Zr1.9Nb0.1(PO4)3 catalysts dimethyl ether is the main prod-
uct and acetaldehyde dominates in all other cases. However,
at lower temperatures the selectivity is not as important as at
higher, since the reaction rate and conversion is very low. For
Li0.9Zr1.9Nb0.1(PO4)3 high selectivity to diethyl ether (60%) per-
sists up to 300 ◦C. For all other catalysts several products are
formed simultaneously at higher temperatures. An exception is
LiZr1.8In0.1Nb0.1(PO4)3, for which at above 350 ◦C the ethylene is
the dominant product with selectivity is more than 90%. And valu-
able is the fact that for LiZr2(PO4)3 and Li0.9Zr2P2.9Mo0.1O12 at high
temperatures C4-hydrocarbons are formed with selectivity up to
60%. It should be noted also that the molybdenum doping reduces
the temperature of the reaction on 60 ◦C.
The LiZr2(PO4)3 and Nb-doped samples have the highest con-
centration of high acidity centers (Fig. 7a,c). At lower temperatures
these catalysts showed the highest selectivity in diethyl ether for-
mation (curves 1, 3 of Fig. 6 c). In the high temperature range all acid
sites take part in the reaction route, thus, a sample with higher spe-
cific surface area posses higher dehydration properties. This is most
typical for LiZr1.8In0.1Nb0.1(PO4)3 sample with 25 m2/g of specific
surface area and high concentration of acid centers. The selectivity
of ethylene formation reaches 80% at 400 ◦C for this sample (curve
5 of Fig. 6a).
tions in Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12
Highest selectivity (close to 100%) is achieved at low tem-
peratures for all the catalysts (Fig. 6). For LiZr2(PO4)3 and
.
The indium-doped sample shows the highest activity in hydro-
gen formation (Fig. 8a). Most likely, this is due to the fact that the
function of dehydration in the sample is suppressed, while dehy-
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
6
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
Fig. 7. Rates of NH3 desorption for the samples with composition LiZr2(PO4)3 (a), Li1.1Zr1.9In0.1(PO4)3 (b), Li0.9Zr1.9Nb0.1(PO4)3 (c), Li0.9Zr2P2.9Mo0.1O12 (d),
LiZr1.8In0.1Nb0.1(PO4)3 (e), Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12 (f).
drogenation function increases. In Li0.9Zr2P2.9Mo0.1O12 not only
dehydrogenation activity enhanced, but the activity of the dehy-
dration increased also. At the same time, at low temperatures,
when the dehydrogenation rate is low, it is possible that a part
of the hydrogen produced is consumed by the partial molybdenum
reduction. In this regard indium-doped catalyst was selected for
the experiments in a membrane reactor
3.3. Transformation of the ethanol in the membrane reactor
Compared to the experiments with the impermeable stain-
less steel foil, the reactor with palladium-ruthenium alloy as a
membrane, Li1.1Zr1.9In0.1(PO4)3 catalyst increases the hydrogen
production by 20% (Fig. 9a). Increase in the ethanol conversion is
also observed (Fig. 9b). These effects are attributed to the shifting of
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
7
Fig. 8. Hydrogen yield for the catalysts with composition LiZr2(PO4)3 (1), Li1.1Zr1.9In0.1(PO4)3 (2), Li0.9Zr1.9Nb0.1(PO4)3 (3), Li0.9Zr2P2.9Mo0.1O12 (4) - (a), and LiZr2(PO4)3 (1),
LiZr1.8In0.1Nb0.1(PO4)3 (5), Li0.9Zr1.8In0.1Nb0.1P2.9Mo0.1O12 (6) - (b).
Fig. 9. Hydrogen productivity in experiments with impermeable foil (1) and the membrane (2, 3, 4), where (2) - retentate, (3) - permeate, (4) - the sum of the permeate and
retentate - (a) and ethanol conversion in membrane reactor with impermeable foil (1) and the membrane (2) - (b).
Fig. 10. The dependence of hydrogen permeation on (P20,5-P10,5) - (a) and (P2-P1) - (b).
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017

G Model
CATTOD-9948; No. of Pages8
ARTICLE IN PRESS
8
A.B. Ilin et al. / Catalysis Today xxx (2016) xxx–xxx
the thermodynamic equilibrium due to removal of hydrogen from
the reaction mixture in the membrane reactor.
References
[1] L.O. Hagman, P. Kierkegaard, Acta Chem. Scand. 22 (1968) 1822.
[2] H. Kohhler, H. Schulz, Mat. Res. Bull. 20 (1985) 1461–1471.
[3] Hong H.Y.-P, Crystal structures and crystal chemistry in the system
Na1+xZr2SixP3-xO12//, Mater. Res. Bull. 11 (2) (1976) 173–182.
[4] S. Roy, P.P. Kumar, Solid State Ionics 253 (2013) 217–222.
[5] P. Zhang, H. Wang, Q. Si, M. Matsui, Y. Takeda, O. Yamamoto, N. Imanishi,
Solid State Ionics 272 (2015) 101–106.
[6] I.A. Stenina, A.B. Yaroslavtsev, Z. Neorg. Khim. 51 (2006) 95-114.
[7] V.I. Petkov, Rus. Chem. Rev. 7 (2012) 606–637.
[8] G.-A. Nazri, G. Pistoia, Lithium Batteries: Science and Technology, Kluwer
Academic Publishers, Boston, Dordrecht, New York, London, 2004, pp. 708.
[9] D. Morgan, G. Ceder, M.Y. Saidi, J. Barker, J. Swoyer, H. Huang, G. Adamson, J.
Pow. Sour. 119–121 (2003) 755–759.
[10] M.Y. Saidi, J. Baker, H. Huang, J.L. Swoyer, G. Adamson, J. Pow. Sour. 119–121
(2003) 266–272.
[11] A.D. Robertson, A.R. West, A.G. Ritchie, Solid State Ionics. 104 (1997) 1–11.
[12] S.R.S. Prabaharan, S. Ramesh, M.S. Michael, K.M. Begam, Mat. Chem. Phys. 87
2–3 (2004) 318–326.
[13] G. Pasciak, K. Prociow, W. Mielcarek, B. Gornicka, B. Mazurek, J. Europ. Ceram.
Soc. 21 (10–11) (2001) 1867–1870.
[14] M. Cretin, P. Fabry, J. Eur. Ceram. Soc. 19 (1999) 2931–2940.
[15] T. Kida, H. Kawate, K. Shimanoe, N. Miura, N. Yamazoe, Solid State Ionics.
136–137 (2000) 647–653.
The recovery rate of hydrogen does not change with the temper-
ature increasing and its average mean is about 35%. The hydrogen
productivity increases significantly with increase in hydrogen
conversion. The permeate contains hydrogen without any other
impurities.
Hydrogen permeates through membrane via
a solution-
diffusion mechanism [44]. Thus, the hydrogen transfer is affected
by its diffusion through the bulk metal and hydrogen pressure dif-
ference in retentate and permeate zones. Hydrogen permeation (J)
can be described by the equation:
P × (pn2 − pn1)
J =
(6)
l
where P is hydrogen permeability constant, l is a membrane thick-
ness, p1 and p2 are the partial hydrogen pressures in the permeate
and retentate zones respectively, n is a constant depending on the
diffusion mechanism.
If the atomic hydrogen diffusion in the membrane bulk is the
rate-controlling stage, n value should be equal to 0.5 because of
equilibrium between molecular hydrogen in gaseous phase and dis-
solved atomic hydrogen. (Sieverts law [45]). But Sieverts law is not
complied in our case (Fig. 10a). The dependence becomes linear for
n value equal to 1 (Fig. 10b). This means that the rate-controlling
stage involves hydrogen in molecular form only. It is reasonable
to suppose that it took place in retentate zone because of the
competitive adsorption of other reaction products and hydrogen
on membrane surface [46–48]. In the permeate zone only hydro-
gen and sweep gas were found, so the competitive adsorption in
this zone cannot occur. Also the experiments with different carrier
gases (Ar and He) were carried out and it was established that even
helium doesn’t pass through membrane which means that there
are no holes in membrane through which molecular hydrogen can
pass in permeate zone.
[16] G. Pas´ciak, W. Mielcarek, K. Prociów, J. Warycha, Ceram. Int. 40 (8B) (2014)
12783–12787.
[17] R.D. Ambashta, M.E.T. Sillanpää, J. Environ. Radioact. 105 (2012) 76–84.
[18] P. Agaskar, R. Grasselli, D. Buttrey, B. White, Stud. Surf. Sci. Catal. 110 (1997)
219.
[19] A.I. Orlova, V.L. Pet’kov, S.T. Gul’yanova, M.M. Ermilova, S.L. Ienealem, O.K.
Samuilova, T.K. Chekhlova, V.M. Gryaznov, Russ. J. Phys. Chem. A. 73 (11)
(1999) 1767–1769.
[20] I.I. Mikhalenko, E.I. Povarova, A.I. Pylinina, Nauchn. Vedom. Belarus. Gos. Univ.
Ser.: Mat. Fiz. 27 (130) (2012) 169–174, 11.
[21] M.A. Aramendia, V. Borau, J.M. Marinas, F.J. Romero, Chem. Lett. (1994)
1361–1364.
[22] A. Serghini, R. Brochu, M. Ziyad, J. Vedrine, J. Chem. Soc. Faraday Trans. 87 (15)
(1991) 2487–2493.
[23] Y. Brik, M. Kacimi, F. Bozon-Verduraz, M. Ziyad, Microporous Mesoporous
Mater. 43 (2001) 103–112.
[24] A.B. Il’in, S.A. Novikova, M.V. Sukhanov, M.M. Ermilova, N.V. Orekhova, A.B.
Yaroslavtsev, Inorg. Mater. 48 4 (2012) 397–401.
[25] E.I. Povarova, A.I. Pylinina, I.I. Mikhalenko, Russ. J. Phys. Chem. A 86 (6) (2012)
935–941.
[26] A.I. Pylinina, I.I. Mikhalenko, Mendeleev Commun. 22 (2012) 150–151.
[27] A.I. Pylinina, I.I. Mikhalenko, Russ. J. Phys. Chem. A. 87 (3) (2013) 372–375.
[28] A.B. Il’in, M.M. Ermilova, N.V. Orekhova, A.B. Yaroslavtsev, Inorg. Mater. 51 (7)
(2015) 778–784.
[29] I.F.J. Vankelecom, K.A.L. Vereruysse, P.E. Neys, et al., Top. Catal. 5 (1998)
125–132.
Within an hour the hydrogen stream in the membrane reactor
decreases by 40% due to the catalyst carbonization. Calcination in
an air stream allows to regenerate the catalyst and to improve its
activity up to its original level.
[30] V. Diakov, A. Varma, Chem. Eng. Sci. 57 (2002) 1099–1105.
[31] R. Tesser, M.D. Serio, E. Santacesaria, Catal. Today 77 (2003) 325–333.
[32] A. Iulianelli, P. Ribeirinha, A. Mendes, A. Basile, Renew. Sustainable Energy
Rev. 24 (2014) 355–368.
[33] A. Iulianelli, A. Basile, Catal. Sci. Technol. 1 (2011) 366–379.
[34] E. Yu, A.A. Mironova, M.M. Lytkina, M.N. Ermilova, L.M. Efimov, N.V. Zemtsov,
G.P. Orekhova, G.N. Karpacheva, D.N. Bondarenko, A.B. Muraviev,
Yaroslavtsev, Int. J. Hydrogen Energy 40 (2015) 3557–3565.
[35] B. Krishna, R. Nair, M.P. Harold, Chem. Eng. Sci. 61 (2006) 6616–6636.
[36] A. Basile, A. Parmaliana, S. Tosti, A. Iulianelli, F. Gallucci, C. Espro, J. Spooren,
Catal. Today 137 (2008) 17–22.
[37] A. Iulianelli, S. Liguori, T. Longo, P. Pinacci, A. Basile, Int. J. Hydrogen Energy 35
(2010) 3170–3177.
[38] A. Iulianelli, A. Basile, Int J. Hydrogen Energy 35 (2010) 3159–3164.
[39] P.K. Seelam, S. Liguori, A. Iulianelli, P. Pinacci, V. Calabrò, M. Huuhtanen, R.
Keiski, V. Piemonte, S. Tosti, M. De Falco, A. Basile, Catal. Today 219 (2012)
42–48.
4. Conclusion
The single-phase catalysts were obtained based on dou-
ble lithium-zirconium phosphate with NASICON-type structure
LiZr2(PO4)3 doped by In, Nb, and Mo, with a particle size less
than 200 nm. All samples exhibit activity in the ethanol dehydra-
tion and dehydrogenation processes. The main products are the
C2- and C4-hydrocarbons, diethyl ether, acetaldehyde and hydro-
gen. It has been shown that the doping by indium increases the
yield of acetaldehyde and hydrogen, doping by niobium - diethyl
ether and ethylene. Molybdenum incorporation increases the cat-
alyst activity in the ethanol dehydrogenation rate and reduces the
temperature of the reaction on 60 ◦C. It is worth mentioned that C4-
hydrocarbons are formed on LiZr2(PO4)3 and Li0.9Zr2P2.9Mo0.1O12
catalysts with selectivity up to 60% at high temperatures.
[40] M.P. Pechini, US Patent 3.3306.97. 1967.
[41] M. Kakihana, M. Yoshimura, Bull. Chem. Soc. Jpn. 72 (1999) 1427–1443.
[42] C.R. Mariappan, C. Galven, M.-P. Crosnier-Lopez, F. Le Berre, O. Bohnke, J. Solid
State Chem. 179 (2) (2006) 450–456.
[43] F. Ejehi, S.P.H. Marashi, M.R. Ghaani, D.F. Haghshenas, Ceram. Int. 38 (8)
(2012) 6857–6863.
[44] F.A. Lewis, The Palladium Hydrogen System, Academic Press, London, 1967.
[45] A. Sieverts, W. Kumbhaar, Ber. Deut. Chem. Gas. 43 (1910) 893.
[46] M. Vandrucci, F. Borgognioni, A. Moriani, A. Santucci, S. Tosti, Hydrogen
permeation through Pd–Ag membranes: surface effects and Sieverts’ law, Int.
J. Hydrogen Energy 38 (2013) 4144–4152.
Carrying out the process in a membrane reactor leads to the
increase in the hydrogen yield due to the thermodynamic equilib-
rium shift and increase in the ethanol conversion.
Acknowledgement
[47] W.-H. Chen, M.-H. Hsia, Y.-L. Lin, C.-C. Yang, Appl. Energy 113 (2014) 41–50.
[48] A. Basile, Top. Catal. 51 (2008) 107–122.
This work was supported by the Russian Foundation for Basic
Research, grant no. 13-08-00660.
Please cite this article in press as: A.B. Ilin, et al., Catalytic activity of LiZr2(PO4)3 nasicon-type phosphates in ethanol conversion process
in conventional and membrane reactors, Catal. Today (2016), http://dx.doi.org/10.1016/j.cattod.2015.12.017