Catalytic activity of thermally treated Li3Fe 2(PO4)3 in the conversion of butan-1-ol

Article abstract of DOI:10.1016/j.mencom.2012.05.013

The effect of the thermal treatment of the LISICON catalyst in hydrogen and atmospheric oxygen on the activity and selectivity in butan-1-ol conversion was studied.

Full text of DOI:10.1016/j.mencom.2012.05.013

Mendeleev
Communications
Mendeleev Commun., 2012, 22, 150–151
Catalytic activity of thermally treated
Li Fe (PO ) in the conversion of butan-1-ol
3
2
4 3
Anna I. Pylinina* and Irina I. Mikhalenko
Peoples’Friendship University of Russia, 117198 Moscow, Russian Federation.
Fax: +7 495 952 0745; e-mail: pylinina@list.ru; imikhalenko@mail.ru
DOI: 10.1016/j.mencom.2012.05.013
The effect of the thermal treatment of the LISICON catalyst in hydrogen and atmospheric oxygen on the activity and selectivity in
butan-1-ol conversion was studied.
Solid-state ion conducting materials, also known as superionic
or fast ion conductors, have been intensively developed. These
materials can be used in solid-state electrochemical devices such
Table 1 XPS data for Li Fe (PO ) .
3 2 4 3
Atomic ratio
Binding energy/eV
P2p Fe2p O1s
Li Fe (PO )
4 3
3
2
1
,2
3
4–6
as sensors, membranes, and catalysts.
Solid electrolytes with the general formula Li M (PO )
4 3
P/Fe O/Fe
3
2
According to stoichiometry
Real surface
1.5
3.3
6
11.1
(
M = Sc, Cr, Fe or In), which are referred to as LISICONs,
133.9 712.8 531.7
7
,8
were discovered in 1983. The conductivity of these materials
caused by the transfer of lithium ions reaches ~10–2
Ω
–1
cm at
–1
3
00°C. Lithium phosphates can be prepared by solid state and
Alcohol dehydration with the formation of butene and dehydro-
genation with the formation of an aldehyde simultaneously occurred
over the test temperature range. The aldehyde yield was many
times higher than that of butene regardless of temperature in the
range of 250–370°C. This behaviour is difficult to explain. The
chemisorbed species OH and H O deactivate the active surface.
With increasing temperature, the surface coverage of dehydro-
genating inhomogeneous sites with butanol decreases, but the
heat of alcohol adsorption QA increases so that the apparent
hydrothermal syntheses, crystallization from melts and sol-gel
method.
Changes in the activity and selectivity of butanol decomposi-
tion on lithium–iron phosphates (LISICON and NASICON) and
the surface acidity after plasma-chemical treatment in glow dis-
charges of oxygen, hydrogen and argon were reported.
plasma treatment increased the catalytic activity and selectivity,
and the highest activity in the dehydration of butanol was observed
S
2
S
9
,10
The
with H as a plasma-forming gas.
activation energy E becomes formally equal to zero according
2
a
0
The aim of this work was to study the effects of thermal treat-
ment in hydrogen and atmospheric oxygen on the activity and
selectivity of the LISICON catalyst in butan-1-ol conversion.
The Li Fe (PO ) catalyst was prepared by a solid state syn-
to the E = E – aQ Brönsted–Polanyi–Semenov relationship
a
a
A
with heat effect fraction a.
Note that LISICON catalyst pretreatments had no significant
influence on dehydrogenating activity.
3
2
4 3
1
1
thesis. The product was white powder with a BET specific
The dehydration reaction on Li Fe (PO ) began at 250°C,
3
2
4 3
2
–1
surface area of 12 m g , as determined by nitrogen desorption
on a Gazochrom-1 gas analyzer. The specific surface areas of
the catalyst thermally treated in oxygen and hydrogen at 400°C
and a flow rate 1.5 dm h for 40 min were 14 and 13 m g ,
respectively.
and the yield of butane was 12% at 370°C [Figure 1(a), curve 1]
at a maximum total alcohol conversion of about 70%.
The activity of a working catalyst in dehydration differed from
that on a fresh surface: a maximum yield of butene was 4%
and the activation energy increased by a factor of 1.8 (Table 2).
Thus, the initial state of dehydration centres is unstable.
The catalytic behaviour of the Li–Fe phosphate (LISICON)
can be compared with that of Na–Zr phosphates (NASICON
3
–1
2
–1
The surface layer composition of Li Fe (PO ) was charac-
3
2
4 3
terized by X-ray photoelectron spectroscopy (XPS) (Table 1).
An XSAM-800 spectrometer with AlKa1,2 radiation (1486.6 eV)
was used. The C1s line (E = 285 eV) was used as an internal
reference.
Table 2 Product yields (N/108 mol h–1 m ), selectivity at (1) 300 and
–2
(
2) 360°C, activation energies E and logarithm of prefactors lnN in butanol
The conversion of butan-1-ol was performed in a U-shaped con-
tinuous-flow microreactor at atmospheric pressure and 200–400°C
with the use of 30 mg of the catalyst. Butan-1-ol diluted with He
was fed to the reactor at a partial pressure of 760 Pa; the total
a
0
dehydration reaction over Li Fe (PO ) catalyst: sample 1 – without treatment,
3
2
4 3
samples 2,3 – treatment in O and H , respectively.
2
2
NButene NButanal
SButene (%)
but
E
/
a
lnN 0b ut abut
flow rate was 1.1 dm–3 h . The reaction mixture was analyzed
on an LKhM chromatograph equipped with a 3 m stainless-steel
column containing Porapak Q and a flame-ionization detector at
–1
Sample
–1
kJ mol
1
2
1
2
1
2
Initial activity
1
50°C using helium as a carrier gas.
The XPS data (Table 1) show that the surface layer of LISICON
1
2
3
3.5 9.5 65 65
6.2 12 65 70
5
8
3
13
15
10
70
42
74
–2.8
–8
–2.2
5.5×10–3
1.8
2
8.2 60 60
does not correspond to the stoichiometric composition. The P/Fe
and O/Fe atomic ratios are higher by a factor of 2, which is
indicative of an excess phosphate concentration on the surface of
Li Fe (PO ) . The value of E (Fe2p) = 713±0.2 eV corresponds
Repeated experiments
1
2
3
0
6
3
3
73 73
11 65 65 8.5 13
60 60 4.8 12
0
4
121
51
69
+6.2
–6
–3.3
5.0×10–6
3
2
4 3
+
b
–5
8
7.5×10
3
to the Fe form.
©
2012 Mendeleev Communications. All rights reserved.
– 150 –

Mendeleev Commun., 2012, 22, 150–151
2
1
0
5
(a)
8
2
1'
6
1
1
0
5
4
2'
1
3
2
0
5
2
0
5
20
570
620
670
90
610
630
650
670
T/K
T/K
2
1
0
5
(b)
Figure 2 Effect of temperature on the selectivity ratio of butene formation
after treatments in O (1,1') and H (2,2') to ones before treatment (1',2' –
repeated experiments).
2
1
2
2
3
and drastically decreased after O treatment (by a factor of 180 –
2
1
0
5
–
3
1
/5.5×10 ). Note that for ‘working’ surface we obtained the
diminishing of N for both treatments, so the activating effect
0
is related to a decrease in E .
a
The selectivity ratio S2 /S 1b ut for initial activity remained
but
or 3
unchanged with temperature and close to 1 (Figure 2). However,
in repeated experiments, the ratio increased by a factor of 5–7
with decreasing temperature.
0
5
20
570
620
670
T/K
Thus, we found that thermal treatment exerted a smaller effect
on complex phosphate catalysts than plasma chemical activation,9
but this procedure prevents the deactivation of LISICON catalysts.
Figure 1 Temperature dependences of butene yield on LISICON catalyst
1) before and after thermal treatments in (2) O2 and (3) H : (a) initial
surface, (b) repeated experiments.
,10
(
2
Thermal treatment in O increased the yield of butene and caused
the formation of active catalytic centres with a lower binding
energy between butanol and lithium–iron phosphate.
2
catalysts), which were previously tested in the same reaction.5
The dehydration activity of NaZr (PO ) is superior to that of
2
4 3
Li Fe (PO ) but no dehydrogenation activity was observed under
3
2
4 3
these experimental conditions. This feature probably results from
differences between the LISICON and NASICON acid–base
References
1
N. Yamazoe and N. Miura, Sens. Actuators B, 1994, 20, 95.
5
,9
properties. The dehydrogenation activity can be attributed to
the reduction ability of the Fe (rather than Zr) ions located in
the channels of the Li(Na)SICON structure. The extent of this
reduction obviously depends on the nature of ion and on the
structural features of these phosphates.
2 D.-D. Lee and D.-S. Lee, IEEE Sens. J., 2001, 1, 214.
3 G. F. Tereshchenko, N. V. Orekhova, M. M. Ermilova, A. A. Malygin and
A. I. Orlova, Catal. Today, 2006, 118, 85.
4
S. N. Ienealem, S. G. Gul’yanova, T. K. Chekhlova, M. M. Ermilova,
A. I. Orlova, V. I. Pet’kov and A. G. Timakin, Zh. Fiz. Khim., 2000, 74,
2
273 (Russ. J. Phys. Chem. A, 2000, 74, 2082).
After thermal treatment in atmospheric oxygen, we obtained
the highest alcohol conversion and the highest butene yield. The
total butanol conversion was close to 90% at 370°C. At T < 350°C,
the yield of butene on sample 2 was twice as high as that on
sample 1 (Figure 1). The activation energy of olefin formation
was Ea = 42 kJ mol , which was lower by a factor of 1.7 than
that in case of sample 1. Note that the catalyst activity after O2
pretreatment was reproducible in repeated experiments.
5
A. I. Pylinina, E. P. Dobrova, I. I. Mikhalenko and T. V. Yagodovskaya,
Zh. Fiz. Khim., 2005, 79, 650 (Russ. J. Phys. Chem. A, 2005, 79, 552).
6 A. I. Pylinina, I. I. Mikhalenko, M. M. Ermilova, N. V. Orekhova and
V. I. Pet’kov, Zh. Fiz. Khim., 2010, 84, 465 (Russ. J. Phys. Chem. A,
2
010, 84, 400).
7
E. A. Genkina, L. N. Dem’yanets andA. K. Ivanov-Shits, Pis’ma Zh. Eksp.
Teor. Fiz., 1983, 38, 257 (JETP Lett., 1983, 38, 305).
but
–1
8 F. D’Yvoire, M. Paintard-Screpel, E. Bretey and M.De la Rochere, Solid
State Ionics, 1983, 9–10, 851.
After thermal treatment in hydrogen, the surface of this catalyst
had the lowest activity: the total butanol conversion was 75% at
9 A. I. Pylinina, I. I. Mikhalenko, A. K. Ivanov-Shits, T. V. Yagodovskaya
and V. V. Lunin, Zh. Fiz. Khim., 2006, 80, 1011 (Russ. J. Phys. Chem. A,
2
006, 80, 882).
3
70°C. The activation energy of butene formation over samples 1
1
1
0 A. I. Pylinina, I. I. Mikhalenko, T. V.Yagodovskaya andV. D.Yagodovskii,
Zh. Fiz. Khim., 2010, 84, 2372 (Russ. J. Phys. Chem. A, 2010, 84, 2172).
1 A. K. Ivanov-Shits and I. V. Murin, Ionika tverdogo tela (Solid State
Ionics), St. Petersburg University, St. Petersburg, 2000, vol.1 (in Russian).
–
1
and 3 was the same for ‘fresh’ surfaces (70 and 74 kJ mol ,
respectively), so the state of the dehydrating catalytic sites remains
unchanged after H pretreatment. Note that this state was more
2
stable because of the same LISICON activity after catalytic experi-
ments, as well as in the case of O treatment.
2
We analyzed the effect of treatment on the total number of
dehydrating surface centres N using the ratio a = N2or3 / N1
(Table 2). On a ‘fresh’LISICON surface, the number of dehydra-
tion centres increased by a factor of ~2 only after H treatment
Received: 23rd November 2011; Com. 11/3838
2
–
151 –