Mechanism of a reversible CO2 capture monitored by the layered perovskite Li2SrTa2O7

Article abstract of DOI:10.1039/c002223n

We demonstrate for the first time, a new CO₂ capture ability monitored by a Ruddelsden-Popper compound. Under a humid CO₂ atmosphere, Li₂SrTa₂O₇ is transformed into LiHSrTa₂O₇ re

Full text of DOI:10.1039/c002223n

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Mechanism of a reversible CO capture monitored by the layered perovskite
2
Li SrTa O
2
2
7
a
a
b
a
Cyrille Galven, Jean-Louis Fourquet, Emmanuelle Suard, Marie-Pierre Crosnier-Lopez* and
a
Fran c¸ oise Le Berre
Received 1st February 2010, Accepted 19th February 2010
First published as an Advance Article on the web 24th March 2010
DOI: 10.1039/c002223n
We demonstrate for the first time, a new CO
compound. Under a humid CO atmosphere, Li
lithium hydroxide which combined with the atmospheric CO
carbonate is confirmed by IR, thermal analysis coupled with mass spectroscopy and diffraction
experiments (X-ray and neutron). The structural study of LiHSrTa performed with X-ray and
neutron diffraction data showed that the structure differs from that of LiHSrTa obtained by ionic
by the Li /H repartition in the interlayer spacing. In the case of the
2
capture ability monitored by a Ruddelsden–Popper
SrTa is transformed into LiHSrTa releasing
leads to Li CO . The presence of
2
2
2
O
7
2
O
7
2
2
3
2
O
7
2
O
7
+
+
exchange from Li
LiHSrTa studied in this paper, the Li and H ions are ordered, while in the other form, each cation
is unequally distributed on 2 sites. By heating, Li SrTa is recovered showing that the CO capture is
capture has been tested during six cycles.
2
SrTa
2
O
7
+
+
2
O
7
2
2
O
7
2
reversible and the cyclability of the CO
2
Introduction
As the CO concentration increases continuously in the atmo-
2
sphere, its capture therefore becomes the subject of a lot of studies
1
in order to reduce its amount in air. Numerous technologies for
this purpose have been described in the literature (absorption with
a solvent and selective membrane separation for example), but
only a few inorganic solid compounds have been reported as useful
2
3,4
5
6
7
candidates: Li
an example, the lithium orthosilicate Li
CO
2
ZrO
3
, Li
4
SiO
4
,
Bi
2
O
3
, Li
SiO
4
TiO
4
and CaSiO
3
. As
4
4
reacts reversibly with
3
,4
2
gas according to the reaction proposed by the authors:
Li
4
SiO
4
+ CO
2
(g) ↔ Li
2
SiO
3
+ Li
2
CO
3
To be a good candidate, the material needs to have high
absorbing efficiency and good cyclic performances, in order
to repeat the absorption/desorption process without efficiency
loss.
In the research of novel CO
2
absorbents, we recently highlighted
Fig. 1 XRD patterns of Li
2
2 7 2 2 7
SrTa O just synthesized (a), of Li SrTa O
this ability in some lithium members of the well known layered
perovskite family of the Ruddlesden–Popper (RP) phases of
old sample (b) and of Li SrTa O
2
2 7
fully transformed (c).
8
general formula A’
powder diffraction patterns of old Li
2
[An-1
B
n
O
3n+1]. We first observed that X-ray
9
SrTa
2
O
7
samples stored in
used to understand the transformation source: IR spectroscopy,
reaction with acidic solutions, thermal analysis (thermogravimet-
ric TG analysis coupled with mass spectroscopy MS), Scanning
Electron Microscope (SEM) analysis, X-rays and neutron powder
2
air and at room temperature over a few months exhibit in addition
to the Li SrTa diffraction lines (Fig. 1a), a new set of diffraction
2
2
O
7
peaks characteristic of a layered perovskite material (Fig. 1b). A
careful examination of the patterns revealed that the old samples
diffraction. All of them led us to conclude that Li
2
SrTa
2
7
O allows
were in fact a mixture of the two RP phases Li
2
SrTa
2
O
7
(Fig. 1a)
the CO capture when stocked in air. To our knowledge, and
despite the great number of studies published on these phases
since their discovery by Ruddlesden and Popper in 1957, this is
2
10
and LiHSrTa
2
O
7
(Fig. 1c) in proportion strongly dependent of
the storage duration time. After optimization of the synthesis to
obtain a totally transformed sample, several techniques were then
the first time that CO
lithium RP phase.
2
capture has been highlighted in a study of
The observation at room temperature of the slow spontaneous
CO capture process in the presence of the RP layered perovskite
Li SrTa and the elucidation of its mechanism are the aim of
the study described in this paper.
a
◦
Laboratoire des Oxydes et Fluorures, FR n 2575 IRIM2F, UMR CNRS
2
6
010, Facult e´ des Sciences et Techniques, Universit e´ du Maine, Avenue
O. Messiaen, F-72017, Le Mans, France
2
2
O
7
b
Institut La u¨ e-Langevin, 6 rue G. Horowitz, 38042, Grenoble Cedex 9, France
This journal is © The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 4191–4197 | 4191

View Article Online
Experimental
to dilute acid solutions, a strong effervescent reaction occurs
immediately due to CO formation and its departure from the
2
∑
IR spectroscopy
carbonated material. A few drops of dilute HCl acid are then
slowly deposited on a small portion of the powder placed on
a watch glass. The typical effervescent reaction is immediately
observed, confirming unambiguously the presence of carbonate
anions in the transformed material.
Infrared spectroscopy was performed on a Bomem Michelson
MB120 FTIR spectrometer with a diamond-anvil cell as a micro-
-
1
sampling device. The spectral resolution was 4 cm in the 650–
1
4
000 cm- range.
IR spectrum of the new material was also recorded (Fig. 2a).
-
1
Three observed bands (1440, 1088 and 865 cm ) are characteristic
∑
Thermal analysis
2
-
of the presence of the CO
1
3
carbonate group: the broad band at
-
1
440 cm is attributed to the antisymmetric u vibration mode,
3
TG-MS experiments were performed in the temperature range
◦
-1
while the two other thin bands at 1088 and 865 cm are respectively
attributed to the symmetric vibration mode u and the bending
All these bands are observed on the spectra of Li CO
Fig. 2b). The two remaining bands—one thin at 889 cm and
one broad under 800 cm —are attributed to the layered material
2
5–700 C under an inert atmosphere using a PFEIFFER-
OMNISTAR coupled with a SETARAM 92-1600 ATG.
1
13
mode u
2
.
2
3
-
1
(
∑
Scanning Electron Microscopy (SEM) analysis
-
1
10
Microstructure observations were performed using a Hitachi TM
LiHSrTa
NH Cl (Fig. 2c). The IR study shows then that the transformed
sample is a mixture of LiHSrTa and Li CO
2
O
7
synthesized in solid state through the reaction with
1
000 SEM equipped with a retrodiffused electron detector.
4
2
O
7
2
3
.
∑
X-ray and neutron powder diffraction
X-ray powder diffraction (XRD) data of the sample were collected
in air with a PANalytical X’pert Pro diffractometer using Cu-
Ka radiation equipped with the X’Celerator detector and a high
temperature attachment Anton Paar HTK1200. The data were
◦
collected with a 0.017 step scan increment and a step time of
◦
◦
2
10 s over an angular range of 5 ≤ 2q ≤ 130 .
The room temperature (RT) neutron powder diffraction (NPD)
pattern was recorded at ILL on the D2B high-resolution neu-
tron diffractometer. The experimental conditions used were:
◦
˚
l = 1.59431(1) A, angular range ( 2q) = 0.05–150.00, step scan
◦
increment ( 2q) = 0.05, counting time = 5 h.
For all the diffraction data, the refinements were carried out by
11
the Rietveld method with a pseudo-Voigt function using Fullprof
software. The background level was first chosen manually before
being refined.
12
Results and discussion
∑
Synthesis
As a long time of exposure to air is needed to obtain significant
transformation of the Li SrTa parent phase, we decided to
optimize the conditions to obtain quickly a fully transformed
sample. The action of a pure anhydrous CO flow does not result
2
2
O
7
2
in significant reaction but in the presence of water, the kinetics of
the transformation is increased. Typical conditions for a fast and
complete transformation are as follows: 1 g of a Li
2
SrTa
2
7
O powder
sample, placed in a platinum crucible, is loaded in a small Teflon
bomb (Parr) containing 1.7 g CO
2
dry ice and 5 drops of water;
◦
the bomb is heated at 140 C for 15 h, which corresponds to an
internal bomb pressure close to 30 bars, and then naturally cooled.
In these conditions, the powder X-ray diffraction analysis reveals
Fig. 2 IR spectra of our fully transformed sample (a), Li
2 3
CO (b),
LiHSrTa synthesized by solid state reaction from NH Cl (c).
2
O
7
4
that the Li
2
SrTa
2
7
O parent phase is fully transformed (Fig. 1c).
X-ray diffraction measurements performed on the transformed
material (Fig. 3) revealed that the main part of the XRD pattern
∑
Characterizations
Different techniques have been used to confirm the presence
of carbonate anions in our sample. As everybody knows, when
can be readily indexed on the basis of an orthorhombic unit cell
√
˚
in the Fmmm space group (no. 69), with a = 2a
p
ª 5.5 A,
√
2
-
˚
˚
the CO
3
carbonate group is submitted at room temperature
b = 2a
p
ª 5.5 A and c ª 18.8 A, where a is the lattice constant
p
4
192 | Dalton Trans., 2010, 39, 4191–4197
This journal is © The Royal Society of Chemistry 2010

View Article Online
Discussion
All these characterizations reveal unambiguously that the action
of CO and H O on Li SrTa leads to a mixture of LiHSrTa
and Li , this can be written by the following reaction:
2
2
2
2
O
7
2
O
7
2
CO
3
1
2
1
2
1
2
Li
2
SrTa
2
O
7
+
H
2
O + CO
2
→ LiHSrTa
2
O
7
+
Li
2
CO
3
(1)
This global reaction involves in fact two other reactions. The
first one corresponds to a partial protonation of the lithium RP
phase implying the release of a part of lithium and the formation
of LiOH according to:
Li
2
SrTa
2
O
7
+ H
2
O → LiHSrTa
2
O
7
+ LiOH
(2)
This reaction is followed by a second one, corresponding to the
carbonation of the lithium hydroxide:
Fig. 3 Observed, calculated and difference XRD patterns (pattern
matching mode) for the fully transformed sample. Vertical bars are related
to the calculated Bragg reflection positions corresponding to LiHSrTa
2
O
7
1
2
1
2
1
2
LiOH + CO
2
→
Li
2
CO
3
+
H
2
O
(3)
(
top – S.G. Fmmm) and Li
2
CO
3
(bottom, *, S.G. C2/c).
Since the residue obtained after TGA measurement corresponds
to Li SrTa , this means that reaction (1) is reversible and must
be rewritten as:
2
2
O
7
˚
˚
of the ideal cubic perovsvkite: a = 5.55841(3) A, b = 5.52912(4) A
˚
and c = 18.7810(1) A. These parameters are very close to those
1
2
1
2
1
2
10
Li SrTa O + H O + CO ↔ LiHSrTa O + Li CO
(4)
˚
˚
2
2
7
2
2
2
7
2
3
of LiHSrTa
2
O
7
phase (a = 5.5522(1) A, b = 5.5248(1) A and
+
+
˚
c = 18.7745(4) A) which can be prepared by Li /H partial
exchange from the mother phase Li SrTa either in dilute
HNO solution or in solid state reaction with NH Cl. However,
The reversibility has been confirmed by heating a fully trans-
◦
2
2
O
7
formed sample at 700 C during 6 h since this step allows
3
4
the recovery of the Li
2
SrTa
2
7
O mother phase. Moreover, the
at the side of that strongly dominating set of lines, a careful
examination of the pattern reveals the presence of very small
CO , according to the
Joint Committee of Powder Diffraction Standard (JCPDS) data
experimental mass variation observed (4.90%) on the TGA curve
is in agreement with the theoretical one corresponding to the
diffraction lines related to the presence of Li
2
3
departure of 0.5 mole of H
2
O and 0.5 mole of CO (5.11%).
2
(
Fig. 3).
∑
Structural study of LiHSrTa
2
O
7
To complete the characterization, we decided to realize a
thermal study. Fig. 4 shows the typical thermal analysis curves
TGA/MS) of the transformed material. On heating, the TGA
The structural study has been performed from neutron diffraction
data since this technique is more sensitive for the location of light
atoms as oxygen and lithium in samples containing heavy elements.
With such data, we confirmed unambiguously the presence of
(
curve shows a Dm/m relative mass loss close to 4.90% between RT
◦
and 700 C. The coupled MS experiment, carried out in order to
identify the components released from the sample under heating,
Li
2
CO and excluded the Fmmm space group deduced from
3
clearly shows that the weight loss is due to the simultaneous depar-
XRD data since it does not allow us to index all the lines of
the NPD pattern. In the Cmcm space group (no. 63), they are
perfectly indexed and a complete structural analysis of the two
◦
ture of CO
2
and H
2
O at 150 C. The solid residue obtained after
heating has been identified by X-ray diffraction and corresponds
to the parent phase Li SrTa
2
2
O
7
.
phases was then undertaken. We started the refinement from
9
the structural model described for Li
2
SrTa
2
O
7
for the perovskite
2
-
blocks [SrTa
2
O
7
]
and we tried to complete the structure from
subsequent Fourier-difference calculations. As encountered for
10
LiHSrTa
2
O
7
,
we did not find any reasonable solution in the
Cmcm space group while an interesting solution is found in the
2
-
corresponding acentric one Ama2 (no. 40). When the [SrTa
2
7
O ]
perovskite blocks refined atomic positions are judged stabilized,
the Fourier-difference calculations reveal four possible cationic
positions in the interlayer spacing. Two of these positions (Li1
+
and Li2) can be easily attributed to Li cations with an anionic
˚
tetrahedral environment and distances Li–O close to 2 A, while
the two others (named hereafter H1 and H2) can be attributed to
+
˚
H cations with distances H–O close to 0.9 A. Interestingly, these
+
+
four positions are the same as those found for Li and H cations
10
in LiHSrTa
close parameters of the two forms, we can assume now that the
new transformed material is in fact a mixture of a LiHSrTa
and Li CO . The occupation factors of the four cationic sites
2
O
7
. Taking into account this remark and the very
O
7
Fig. 4 TG-MS curves of a fully transformed sample showing the
2
◦
simultaneously departure of CO
2
and H
2
O as of 150 C. The full line is
2
3
related to the weight loss and the dashed lines are related to the MS curves.
Li1/Li2 and H1/H2 are then refined with a global constraint
This journal is © The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 4191–4197 | 4193

View Article Online
Table 1 Structure refinement results of LiHSrTa
2
O
7
from neutron data
Table 2 Atomic coordinates, Biso and bond valence sums (R) for
LiHSrTa
2 7
O from neutron diffraction data (S.G.: Ama2) (standard de-
viations are given in parentheses)
Space group
Ama2 (no. 40)
72
Pseudo-Voigt, 0.24(3)
Number of refined parameters
Peak shape, h
Cell parameters/ A˚
B/ A˚ 2
0.33(7) 5.2 (5)
0.25 0.753(1) 0.8872(2) 0.21(7) 5.1 (5)
0.25 0.252(1) 0.0006(4) 0.72(4) 2.3 (2)
0.25 0.723(1) 0.0097(4) 1.04(9) 2.0 (2)
Site s.o.f.
x
y
z
R
(Rexp)
a = 5.5534(1)
b = 5.5226(1)
Ta1 4b
Ta2 4b
Sr 4b
1
1
1
1
1
1
1
1
1
1
1
1
0.25 0.752(2) 0.1130
c = 18.7835(2)
576.1(1) A˚ /Z = 4
3
Cell volume/Z
Halfwidth parameters
u = 0.067(1)
v = -0.159(4)
w = 0.151(2)
x = 0.0004(4)
O1 4b
O2 4b
O3 4b
O4 4a
O5 4a
O6 4a
O7 4a
0.25 0.766(2) 0.2212(4) 0.7(1)
0.25 0.788(1) 0.7928(4) 1.3(1)
1.8 (2)
2.0 (2)
2.1 (2)
2.1 (2)
2.1 (2)
2.2 (2)
0.9 (1)
0.9 (1)
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.6191(6) 1.2(2)
0.4011(5) 0.5(2)
0.1045(5) 1.1(2)
0.9131(5) 0.6(1)
Asymmetry parameters
P
P
1
= 0.113(9)
= 0.038(4)
2
R
R
R
R
Bragg
p
3.99%
9.97%
11.00%
3.50%
9.88
Li
H
4a
4b
0.755(1)
0.9(2)
0.25 0.410(2) 0.7447(7) 3.2(2)
wp
exp
2
(the z coordinate for Ta1 atom was held invariant in order to define the
origin)
c
+
+
to the expected value: 4 Li cations and 4 H cations per unit-cell.
The calculations progressively fill two sites Li2 (0, 0, 0.76) and
H2 (0.25, 0.41, 0.74) while the other sites Li1 and H1 are emptied,
are reported in Table 2 while selected interatomic distances are
listed in Table 3.
leading to very good reliability factors: R
p
= 9.97%, Rwp = 11.00%,
We can clearly see on Fig. 6 that the structure of this LiHSrTa
2
O
7
RBragg = 3.99% and a good agreement between the observed and
form presents a small octahedral tilting inside the perovskite
blocks which could not be revealed from XRD data: inside the
calculated patterns (Fig. 5). Table 1 gathers the final cell and profile
parameters with the reliability factors for neutron diffraction data.
Atomic coordinates, thermal parameters and bond valence sums
perovskite blocks and as observed in the mother form Li
2
SrTa
2
O ,
7
◦
◦
the octahedra are slightly tilted (4.0 for Ta1O
6
and 3.3 for
Ta2O
n = 2 RP layered perovskite: the perovskite layers are formed
from two infinite TaO octahedra planes, the 12-coordinated site
being fully occupied by the large strontium cations, while the small
6
). The figure also shows that this phase is related to a
6
+
+
H and Li cations are located in the interlayer spacing. Compared
+
+
to the Li
2
SrTa
2
O
7
mother phase, the Li /H exchange leads to an
14
˚
increase of the c parameter as usually encountered: 18.7835(2) A
˚
for LiHSrTa
blocks are constituted of two types of infinite layers built from
corner-shared Ta1O and Ta2O octahedra. The Ta1O octahedra
are almost regular with Ta1–O distances ranging from 1.946(8) to
2
O
7
and 18.1822(3) A for Li
2
SrTa
2
O
7
. The perovskite
6
6
6
˚
˚
2
.034(8) A, leading to an average Ta1–O distance 1.972 A close to
15
˚
the sum of the ionic radii taken in Shannon’s table (1.99 A). In
˚
the Ta2O octahedra, a long distance Ta2–O1 (2.307(8) A) at the
6
˚
opposite of a short one Ta2–O3 (1.784(8) A) is observed with four
Fig. 5 Observed, calculated and difference NPD patterns for the fully
transformed sample. Vertical bars are related to the calculated Bragg
equivalent Ta2–O distances in the equatorial plane (2 ¥ 1.987(5)
and 2 ¥ 2.006(5) A). Despite this distortion, the average Ta2–O
˚
distance (2.013 A) is in good agreement with Shannon’s table and
˚
reflection positions corresponding to LiHSrTa
bottom). The inset shows a zoom view in which the presence of the Li
phase (*) is clearly visible.
2
O
7
(top) and Li
2
CO
CO
3
15
(
2
3
16
the calculated bond valence sums for the two Ta sites are quite
Table 3 Selected interatomic distances ( A˚ ) for LiHSrTa
O
from neutron diffraction data. (mean distances are given in italics)
Li tetrahedron
2
7
Ta octahedra
Sr polyhedron
H polyhedron
Ta1–O1 : 1 ¥ 1.946(8)
Ta1–O6 : 2 ¥ 1.959(6)
Ta1–O4 : 2 ¥ 1.967(6)
Ta1–O2 : 1 ¥ 2.034(8)
Sr–O7 : 2 ¥ 2.563(8)
Sr–O1 : 1 ¥ 2.607(8)
Sr–O5 : 2 ¥ 2.701(9)
Sr–O6 : 2 ¥ 2.771(9)
Sr–O1 : 2 ¥ 2.785(1)
Sr–O1 : 1 ¥ 2.926(8)
Sr–O4 : 2 ¥ 2.96(1)
<Sr–O> = 2.758 A˚
Li–O3 : 2 ¥ 2.03(1)
Li–O2 : 2 ¥ 2.069(9)
<Li–O> = 2.05 A˚
H–O2 : 1 ¥ 0.91(1)
H–O3 : 1 ¥ 2.28(1)
<
Ta1–O> = 1.972 A˚
Ta2–O3 : 1 ¥ 1.784(8)
Ta2–O5 : 2 ¥ 1.987(5)
Ta2–O7 : 2 ¥ 2.006(5)
Ta2–O1 : 1 ¥ 2.307(8)
<
Ta2–O> = 2.013 A˚
4
194 | Dalton Trans., 2010, 39, 4191–4197
This journal is © The Royal Society of Chemistry 2010

View Article Online
Li2–H1). Then, in order to test the sensitivity of the neutron
+
+
diffraction to a modification of the Li /H distribution, we
have simulated the neutron diffraction patterns for the following
two extreme situations: Li1/H1 fully occupied or Li2/H2 fully
occupied. Fig. 7 shows firstly that the differences are very weak
◦
and are essentially visible on two peaks ((115) at 34.2 2q and (211)
◦
at 37.8 2q) shown by arrows, and secondly that the comparison
of the experimental pattern with the two simulated ones is in favor
of the Li2/H2 distribution, in good agreement with the structural
refinement results.
Fig. 6 Projection of the LiHSrTa
layers and the location of Li and H ions in the interlayer spacing.
2
O
7
structure showing the perovskite
+
+
good and close to the expected value (5): 5.2 and 5.1 respectively
for Ta1 and Ta2 sites.
2
+
In the LiHSrTa
2
O phase as in the mother form, Sr cations are
7
located in the perovskite cages (coordination number CN = 12)
Fig. 7 NPD patterns of the fully transformed sample: experimental (a),
simulated with Li2 and H2 sites fully occupied (b) and simulated with Li1
and H1 sites fully occupied (c) (see text).
˚
˚
with Sr–O distances ranging from 2.563(8) A (Sr–O7) to 2.96(1) A
1
5
(
Sr–O4), values which can be compared to the ionic radii sum
2
+
˚
2
.79 A. It leads to the calculated bond valence sum 2.3 for Sr
(
expected value: 2).
The two small Li and H cations are both inserted in the
We can now ask us what differs between the form LiHSrTa
2
O
7
+
+
obtained by CO
2
absorption and the form LiHSrTa
2
O obtained in
7
+
interlayer spacing. Li ions are located on a 4a site (0, 0, z)
in a tetrahedral environment with four Li–O distances nearly
equivalent (2.03(1) to 2.069(9) A) and a little larger than the ionic
radii sum (1.94 A for Li in a 4 coordination number). Each LiO
tetrahedron links together two successive perovskite layers (Fig. 6):
two apical oxygens of each layer are used to form its anionic
tetrahedral environment. All the LiO tetrahedra are linked by
sharing one O–O edge and form infinite tetrahedra chains running
along the a axis. H ions fully fill a 4b site (1/4, y, z), leading to
a short H–O2 distance (0.91(1) A) with an apical oxygen of the
10
solid state reaction with NH
4
Cl. The main difference concerns
+
+
the Li /H organization in the interlayer spacing: from NH
4
Cl
˚
+
+
reaction, a disordered Li /H situation is observed with a partial
occupancy of two Li sites and two H sites, while from CO
absorption, each cation Li and H fully occupied only one
site leading to a perfectly ordered phase. As a consequence, in
15
+
˚
4
+
+
2
+
+
4
the ordered LiHSrTa
2
7
O form presented in this paper, all the
bond valence sum calculations give the expected ionic charge for
both cationic and anionic sites (ranging from 1.8 to 2.2 for the
oxygen sites) while in the disordered form, the bond valence sum
calculated for the two apical anionic sites are wrong despite the
+
˚
perovskite layer. This H–O distance is in good agreement with
those published on protonated phases. At the opposite of this
H–O2 short distance, a long H–O3 distance (2.28(1) A) is observed
with an apical oxygen of the adjacent perovskite layer. As for the
17
10
different cationic environments intended. Taking into account
˚
the good results obtained for the ordered form, we can reasonably
think that in the case of a statistical occupation of sites, the atomic
positions obtained from powder diffraction data are averaged and
do not reflect locally the true structure.
+
LiO
4
tetrahedra, the H sites are aligned along the a axis. For
+
+
the two cations Li and H , the bond valence sums are in good
agreement with the expected values (Table 2)
Only two different strictly ordered distributions of Li and H
ions can be imagined from the two available sites for each cation,
sites depicted from Fourier-difference calculations. Indeed, due to
the too short Li–H distances, Li1 and H2 sites (or Li2 and H1
16,18
In addition to this structural study, it is possible from the NPD
+
+
12
data to do a quantitative phase analysis. Indeed, for a mixture
of crystalline phases, the weight fraction w of each phase j can be
j
obtained from:
˚
sites) cannot be simultaneously occupied (1.5 A for Li1–H2 and
w
j
= {S
j
Z
j
M
j
V
j
/t
j
}/Sum(i){S
i
Z
i
M
i
V
i
/t
i
}
(5)
This journal is © The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 4191–4197 | 4195

View Article Online
where S
formula units per unit cell of the phase j, M
formula unit of the phase j, V
j and t is the Brindley particle absorption contrast factor for the
j
is the scale factor of the phase j, Z
j
is the number of
j
is the mass of the
j
is the unit cell volume of the phase
j
phase j. In a first approximation and when the phases have similar
absorption (in most of the cases), this last factor is nearly equal
to 1. In our case, the final refinement cycle gives scale factors
for LiHSrTa
.205(4), while the number of formula units are both equal to 4.
The numerical application of eqn (5) gives w(Li CO
) = 5.69% and
w(LiHSrTa
) = 94.31%, leading to an experimental molar ratio
n(LiHSrTa )/n(Li CO
) = 0.46(2), very close to the theoretical
value (0.5) implied by reaction (1). The fully transformed phase is
then a mixture of LiHSrTa and Li CO in a ratio of 2/1.
2
O
7
and Li
2
CO
3
respectively equal to 0.179(1) and
0
2
3
2
O
7
2
O
7
2
3
2
O
7
2
3
Finally, the transformed material has been observed by SEM
microscopy. The crystallites have the shape of thin platelets the
size of which does not exceed 10 mm (Fig. 8). The homogeneous
contrast observed with a detector of retrodiffused electrons implies
thus that the sample is homogeneous and does not clearly present
Fig. 9 Part of the XRD patterns of the mixture resulting from the first
absorption ((a) curve) and after the sixth absorption ((b) curve).
Conclusion
two phases with different chemical compositions. We must then
+
In this paper, we have demonstrated that the lithium RP phase
conclude that each crystallite of Li
on its entire surface, leading to the presence of Li
whole surface of each LiHSrTa grain. This most probable
2
SrTa
2
O
7
releases Li cations
Li
2
SrTa
2
O
7
is able to monitor spontaneously CO capture at room
2
2
CO on the
3
temperature when it is exposed to air under ambient conditions.
The capture kinetics can be increased with temperature and a
2
O
7
situation is supported by the fact that it is impossible to obtain
◦
humid CO
niques, the mechanism has been explained as follows: Li
slowly transforms into a half protonated form, with release of Li
cations in the form of lithium hydroxide. This hydroxide captures
easily CO leading to lithium carbonate (Li CO ). After these two
2
pressure. Thanks to several characterization tech-
pure Li
LiHSrTa
Aldrich) in spite of a homogeneous mixture carried out in an
agate mortar.
2
SrTa
2
O
7
by solid state reaction at 700 C in air from
O
7
(prepared as described in reference 10) and Li
2
CO
3
2
SrTa
2
O
7
+
2
(
2
2
3
successive reactions, a mixture is obtained according to the global
reaction:
1
2
1
2
1
2
Li
2
SrTa
2
O
7
+
H
2
O + CO
2
↔ LiHSrTa
2
O
7
+
Li
2
CO
3
This reaction is reversible since it is possible to recover the
mother lithium RP phase by heating the mixture without great
deterioration of its structure. If the CO
2
capture via Li
2
CO
3
formation is already known, it is interesting to note that the
operating temperatures of inorganic solid used actually are higher,
◦
2-6
◦
generally above 400 C (compared to 140 C for Li
2
SrTa
2
O
7
).
Despite the existence of the completely protonated form
1
4,19
H
Li
2
SrTa
SrTa
interesting from the CO
Lastly, to complete this study, we tested Li
the same experimental conditions and we observed a similar
CO capture ability with the formation of a mixture of the half
2
O
7
,
it is impossible to exchange the two lithium of
in our synthesis conditions, which would be very
capture point of view.
2
2
O
7
2
2
SrNb
2
7
O with
2
protonated RP form and Li
2
CO .
3
Fig. 8 SEM micrograph of the transformed material showing an uniform
contrast.
Acknowledgements
The authors thank Dr Olivier Mentr e´ (UCCS - UMR CNRS
181-Lille) for the coupled TGA/MS experiment, Dr Karim Adil
LdOF-UMR CNRS 6010-Le Mans) for the IR analysis and Mrs.
Finally and to test the cyclability of reaction (4), six cycles
absorption/desorption have been undertaken. Fig. 9 shows the
XRD pattern of the mixture resulting from the sixth absorption
8
(
Cuvier from Elexience society for SEM micrograph.
(
(b) curve) comparatively with that of the first absorption ((a)
curve). On this figure, one can see firstly a broadening of the
diffraction line after several cycles and secondly some additional
unknown diffraction peaks, which however does not deteriorate
References
1
K. S. Lackner, H.-J. Ziock and P. Grimes, Proceedings of the 24th
International Conference on Coal Utilization & Fuel Systems, 1999,
Clearwater, Florida; M. Kanniche, R. Gros-Bonnivard, P. Jaud and J.
the CO
2
absorption ability since a weight variation close to 5.11%
is observed for the sixth desorption.
4
196 | Dalton Trans., 2010, 39, 4191–4197
This journal is © The Royal Society of Chemistry 2010

View Article Online
Valle-Marcos, Energy Proc., 2009, 1, 4103; D. J. Barker, S. A. Turner,
P. A. Napier-Moore, M. Clark and J. E. Davison, Energy Proc., 2009,
11 H. M. Rietveld, J. Appl. Crystallogr., 1969, 2, 65.
12 J. Rodriguez-Carvajal, Program FULLPROF, version Feb 2007.
13 A. B. Ali, V. Maisonneuve, S. Houlbert, G. Silly, J. Y. Buzar e´ and M.
Leblanc, Solid State Sci., 2004, 6, 1237; N. Imanaka, T. Masui, Y.
Mayama and K. Koyabu, J. Solid State Chem., 2005, 178, 3601; R.
Khoshnavazi, A. Salimi and A. G. Moaser, Polyhedron, 2008, 27, 1303.
14 N. S. P. Bhuvanesh, M. P. Crosnier-Lopez, H. Duroy and J. L. Fourquet,
J. Mater. Chem., 2000, 10, 1685.
1
, 87; G. Pellegrini, R. Strube and G. Manfrida, Energy, 2010, 35, 851;
Aage Stangeland, Int. J. Greenhouse Gas Control, 2007, 1(4), 418.
K. Nakagawa and T. Ohashi, J. Electrochem. Soc., 1998, 145, 1344; K.
Nakagawa and T. Ohashi, Electrochemistry, 1999, 67, 618.
2
3
4
M. Kato, K. Nakagawa, K. Essaki, Y. Maezawa, S. Takeda, R. Kogo
and Y. Hagiwara, Int. J. Appl. Ceram. Technol., 2005, 2(6), 467.
K. Essaki, M. Kato and K. Nakagawa, J. Ceram. Soc. Jpn., 2006,
15 R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor.
Gen. Crystallogr., 1976, 32, 751.
1
14(1333), 739.
5
6
T. Esaka and K. Motoike, J. Alloys Compd., 2006, 408–412, 480.
N. Togashi, T. Okumura and K. Oh-ishi, J. Ceram. Soc. Jpn., 2007,
16 N. E. Brese and M. O’Keeffe, Acta Crystallogr., Sect. B: Struct. Sci.,
1991, 47, 192.
1
15(1341), 324.
17 S. Nishimoto, M. Matsuda, S. Harjo, A. Hoshikawa, T. Kamiyama,
T. Ishigaki and M. Miyake, J. Eur. Ceram. Soc., 2006, 26, 725; S.
Nishimoto, M. Matsuda, S. Harjo, A. Hoshikawa, T. Kamiyama, T.
Ishigaki and M. Miyake, J. Solid State Chem., 2006, 179, 1892.
18 I.D.Brown Private communication, LLB, version: March 2005, bound
to FullProf.
7
8
M. Wang and C.-G. Lee, Energy Convers. Manage., 2009, 50, 636.
S. N. Ruddlesden and P. Popper, Acta Crystallogr., 1957, 10, 538; S. N.
Ruddlesden and P. Popper, Acta Crystallogr., 1958, 11, 54.
T. Pagnier, N. Rosman, C. Galven, E. Suard, J. L. Fourquet, F. Le Berre
and M. P. Crosnier-Lopez, J. Solid State Chem., 2009, 182, 317.
9
1
0 C. Galven, J. L. Fourquet, E. Suard, M. P. Crosnier-Lopez and F. Le
19 M. P. Crosnier-Lopez and J. L. Fourquet, Solid State Sci., 2005, 7,
Berre, Dalton Trans., 2010, 39, 3212–3218.
530.
This journal is © The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 4191–4197 | 4197