You are on page 1of 26

PROBING LOCAL STRUCTURE AND DYNAMICS IN BATTERY MATERIALS NEW

APPROACHES

Clare P. Grey
1,2
1
Chemistry Department, Lensfield Rd, Cambridge CB2 1EW, UK.
2
Chemistry Department, Stony Brook University, Stony Brook, NY 11794-3400, USA.

cpg27@cam.ac.uk

Keywords: NMR, insitu methods, PDF, diffraction, lithium ion batteries

This talk will focus on the use of NMR methods to
investigate local structure in battery materials, bringing
in other diverse local and long range probes of structure
to complement the characterization.
A brief overview of the use of NMR methods to
investigate local structure and the electronic properties of
paramagnetic materials will be provided, including
recent theoretical studies by groups world-wide to
calculate the NMR responses of these materials and help
in spectral assigment. These methods have now been
applied (for example) to a diverse set of materials, which
include the layered LiMO
2
materials (e.g.,
Li(Co
1/3
Ni
1/3
Mn
1/3
)O
2
, Li(Li
1/3
Mn
2/3
)O
2
, Li(Mn
1/2
Ni
1/2
)O
2

and solid solutions involving these compositions) and a
series of lithium iron (II) and (III) phosphates/silicates.
The development of insitu NMR methods will then be
reviewed. Applications of these approaches to study the
anode material Si, Li dendrite formation, and conversion
materials such as FeF
2
/FeF
3
will then be described.

293
TOWARDS CHEAPER CATHODES FOR LARGE-SCALE LITHIUM-ION BATTERIES

Anti Liivat, Torbjrn Gustafsson & Josh Thomas
ngstrm Advanced Battery Centre,
Department of Materials Chemistry, Uppsala University,
Box 538, SE-751 21, Uppsala, Sweden.

josh.thomas@mkem.uu.se


The development of larger, cheaper and necessarily
greener Li-ion batteries for EV and stationary energy-
storage applications is still presenting an apparently
insuperable challenge. In this context, the absence of
an ideal cathode material with a high energy- and
power-density is arguably the most serious bottleneck.
There are several underlying reasons for this; not least,
the cost of raw materials and the economy of the
synthesis process but, above all, limitations in
available chemistries to provide the desired
improvements in capacity. A breakthrough came with
the development of polyanion-based cathode materials;
typically, LiFePO
4
and Li
2
FeSiO
4
[1,2]. These materials
can be expected to be both cheaper and safer than
conventional transition metal (TM) oxides.
Unfortunately, they tend to exhibit poor electronic
conductivity a problem which must be addressed by
reducing particle-size or promoting conductivity
through some form of surface treatment. Doping
strategies are also highly desirable for these essentially
insulating materials, especially in the context of
promoting >1-electron redox reactions. Typically,
the materials combination on which we are pinning our
greatest hopes is iron-based polyanion compounds,
where the Fe-site is substituted by Mn to exploit the 2-
electron reaction involving a Mn
2+
Mn
4+
transition.
Recent advances in our efforts to address these
challenges will be reported.


This work has been supported by the Global Climate
and Energy Project of Stanford (GCEP), the Swedish
Energy Agency (STEM) and the Swedish Science
Research Council (VR).

[1] A.K. Padhi et al., J. Electrochem. Soc., 144, (1997) 1609.
[2] A. Nytn et al., Electrochem. Commun.,7, (2005)
156.



294
IN SITU RAMAN SPECTROSCOPY OF THIN FILM BATTERY: STRUCTURE CHANGE
AND PHASE BOUNDARY MOVEMENT OF Li
x
CoO
2


N. Kuwata,
1
Y. Matsuda, J. Kawamura,
1
O. Kamishima,
2


1
IMRAM, Tohoku University, Katahira 2-1-1, Sendai, 980-8577, Japan
2
Faculty of Engineering, Setunan University, Neyagawa, Osaka, 572-8508, Japan

kuwata@tagen.tohoku.ac.jp

Keywords: Solid-state battery, in situ Raman microscopy, Pulsed laser deposition, LiCoO
2


Introduction
All solid-state thin-film batteries (TFBs) using an
amorphous solid electrolyte have been expected as power
sources of various microelectronic devices. We report
recently that a thin-film battery of Li/Li
3
PO
4
/LiCoO
2
can
cycle more than 1000 cycles without capacity fading [1, 2].
The solid-state batteries are usually expected to have
a long cycle life. However, some thin-film batteries show
rapid capacity fading, because of a side reaction or an
irreversible reaction [3]. It is very important to know the
structural change during charge-discharge (in situ) for
understanding the mechanism of capacity fading.
In situ Raman spectroscopy is powerful tool to
detect the structure change of the lithium battery cathode
materials such as LiCoO
2
[4, 5]. In this study, we observed
the structural change of Li
x
CoO
2
cathode in the thin-film
batteries. The change of Raman spectra is related to the
electrochemical property.
Experimental
A thin-film battery of Li/Li
3
PO
4
/LiCoO
2
was
fabricated by pulsed laser deposition (PLD) and vacuum
thermal evaporation (VTE) [1]. A LiCoO
2
cathode of 200
nm thick was deposited by PLD using 4e Nd:YAG laser
on a Pt/SiO
2
substrate at 600C. An amorphous Li
3
PO
4

electrolyte of 1.5 m thick was deposited by PLD using
ArF excimer laser. Finally, a lithium metal anode of 3~5
m was grown by VTD. The thin-film battery was
measured by cyclic voltammetry in a vacuum-tight cell
with optical glass window. At the same time, micro Raman
scattering was measured using an Ar-ion laser (514.5 nm).
Laser power was less than 1mW to avoid annealing.
Results and discussion
Fig. 1 shows a variation of in situ Raman spectra of
TFB where LiCoO
2
, Co
3
O
4
and Li
3
PO
4
are observed. The
LiCoO
2
shows strong Raman bands of A
1g
and E
g

vibrations. The peak intensity and Raman shifts of the
Li
x
CoO
2
cathode varies reversibly depend on the battery
potential. The A
1g
and E
g
peak of Li
x
CoO
2
rapidly
decreases in intensity at around 3.9 V (two-phase region)
and shows lower wavenumber shift. After that, the peaks
are continuously shifted to lower wavenumber at4.0 to 4.4
V.
In addition, the intensity of A
1g
band shows unusual
hysteresis in the two-phase region. This behavior could be
attributed to the phase boundary movement of H1/H2
phases of Li
x
CoO
2
electrode. The heterogeneity of the two-
phase reaction (H1/H2) has been observed by Raman
microscopy.

Fig. 1. In-situ Raman spectra of TFB.

Fig. 2. Raman shift of A
1g
of Li
x
CoO
2
vs. capacity.
References
[1] N. Kuwata, N. Iwagami, Y. Tanji, Y. Matsuda, J.
Kawamura, J. Electrochem. Soc., 157 (4) A521-A527,
(2010).
[2] N. Kuwata, N. Iwagami, Y. Matsuda, Y. Tanji, J.
Kawamura, ECS Transactions, 16 (26) (2009) 53.
[3] N. Kuwata, J. Kawamura, K. Toribami, T. Hattori, N.
Sata, Electrochem. Commun. 6, 417 (2004).
[4] T. Itoh, H. Sato, T. Nishina, T. Matsue, I. Uchida, J.
Power Sources, 68 (1997) 333.
[5] T. Itoh, N. Anzue, M. Mohamedi, Y. Hisamitsu, M.
Umeda, I. Uchida, Electrochem. Commun., 2 (2000) 743.
295
STUDY OF CATHODE MATERIAL FOR LTHIUM ION BATTERY
BY NMR MICRO IMAGING

Yoshiki Iwai,

Daiki Ohno, Junichi Kawamura


Institute of Multidisciplinary Research for Advanced Materials,
Tohoku University, Sendai 980-8577

iy@mail.tagen.tohoku.ac.jp

Keywords: NMR imaging, lithium ion battery, cathode material

Introduction
Imaging techniques are very powerful tool for
scientific research. They provide us new different kind of
aspects from conventional one dimensional spectroscopy.
Recently we have performed NMR micro imaging to
lithium ion battery [1]. We have found out that proton (
1
H)
imaging of electrolyte (propylene carbonate: PC) which
was located near cathode material (LiMn
2
O
4
) showed
characteristic intensity enhancement and distortion. From
several considerations as experimental parameters or
charge-discharge cycling, we concluded the intensity
enhancement and distortion are related to paramagnets in
cathode material; however quantitative approach is not
established yet.
In this study, we consider a theoretical treatment
of intensity enhancement and distortion in NMR imaging
induced by magnetism of cathode material.
Experimental
A LiMn
2
O
4
electrode with polyvinylidene
difluoride (PVDF) and carbon coated on an aluminum foil
was arranged at the center in a glass tube. A polypropylene
(PP) was used as a retainer. The glass tube was filled with
LiClO
4
+PC (1mol/L) electrolyte. Bruker Avance 400 (9.4
T) was used for MRI measurement. The resonance
frequency of
1
H is 400 MHz. A pulse sequence for NMR
imaging is based on spin-echo method (msme), which is
installed in control software (paravision) produced by
bruker biospin. Imaging dimension was 25mm25mm,
spatial resolution and slice thickness was 195m (128128
pixels) and 0.8mm. Echo time (TE) and repetition time
(TR) was 3.495 ms and 300 ms.
Result and Discussion
Fig.1(a) shows NMR imaging of x-z cross section
(left hand side). A white arrow indicates a position of
LiMn
2
O
4
, which is arranged in x-y plane. One can see a
distortion around LiMn
2
O
4
. The intensity profile which is
measured along z-axis (dotted line in Fig.1 (a)) are shown
in Fig.(b). It shows intensity enhancement around center
position; which is also due to magnetism of LiMn
2
O
4
.
In theoretical treatment [2], a density ) (z
measured from NMR imaging is related to NMR
signal
) (t S
,
}
= ) exp( ) ( ) ( t z G i z t S
z

. (1)
where

,
z
G
is gyromagnetic ratio and magnetic gradient.
Basically, the term
z G
z

is a linear magnetic field


produced by a gradient probe. However, if a system
includes paramagnetic material it creates local
magnetization induced by high magnetic field. Taking into
account the effect, linear gradient term would change,
) (z B z G z G
z z
A +
(2)
where ) (x B A is a magnetic field produced by paramagnetic
material. Substituting the eq. (2) into eq. (1), one can
calculate an intensity profile of NMR imaging distorted by
paramagnetic material. The solid curve in Fig.1(b) is the
result of calculation, where we assume ) (z B A as Gaussian
shape. This calculation is good agreement with
experimental data. In addition, the magnetic susceptibility
evaluated from fitting parameter is comparable to literature
value of LiMn
2
O
4
at R.T. :
2
10 9 . 0

~ _ (emu/mol) [3]. It
supports the validity of this theoretical treatment. We
believe this technique shows promise as sensitive magnetic
microscope to show distribution of cathode material and/or
inhomogeneous charge-discharge process of lithium
battery.

References
[1] D. Ohno, Y. Iwai, J. Kawamura, Atom Indonesia, 36
(2010), 135-138, to be published
[2] Magnetic Resonance Imaging Physical Principles and
Sequence Design, E. Mark Haacke et al, WILEY-LISS,
p.615
[3] C. M.Julien, A. Ait- alah, A. Mauger, F. Gendron,
Ioncs, 12(2006), 21-32


Fig.1 NMR image of x-z cross section (a) and
intensity profile along z-axis (b)
(a) (b)
z
x
LiMn
2
O
4

A
296
STUDY OF DEFECTS IN LiFePO
4
BY USING PDF ANALYSIS AND MOSSBAUER
SPECTROSCOPY


S. Ferrari,
1
M. Bini,
1
D. Capsoni,
1
V. Massarotti
1
, P. Mustarelli,
1
G. Spina,
2
C. Leonelli
3
1
Dept. of Physical Chemistry, University of Pavia, Via Taramelli, 16 27100 Pavia, Italy
2
Dept. of Physics and Astronomy, Via Sansone, 1 50019 Sesto Fiorentino, Italy
3
Dept. Mat. & Environm. Engineering, Strada Vignolese, 905 41125 Modena, Italy


corresponding: stefania.ferrari@unipv.it

Keywords: lithium battery, LiFePO
4
, pair distribution function (PDF) analysis

LiFePO
4
is a very well known cathode material
for lithium ion batteries. Numerous studies have been
devoted in the last years to the optimization of this
material through the development of new synthesis
routes that favored the formation of particles with
specific morphologies (nanoparticles, hollow sphere,
oriented platelet like shaped) promising for the
electrochemical applications. Olivine prepared by
hydrothermal or precipitation methods has been
recognized as intrinsically defective due to the low
synthesis temperature and time. The most cited defect is
the lithium iron anti-site, which has been determined
both by experimental work [1] and by theoretical energy
calculation [2].
We prepared crystalline LiFePO
4
by means of an
innovative microwave assisted hydrothermal synthesis at
low temperature for few minutes, with the specific aim of
obtaining a defective material. The investigation of the
structural disorder has been carried out, for the first time,
to the best of our knowledge, through the Rietveld
method and the PDF analysis on synchrotron radiation
data (Fig. 1). The defect models have been supported by
Mossbauer spectroscopy, that allowed to characterize the
iron environment and oxidation state. Preliminary results
suggest the presence of Fe
3+
ions and of local disorder
that can be explained on the basis of a lowering of the
symmetry with respect to the orthorhombic Pnmb space
group of the olivine structure. The knowledge of the
defective structure of LiFePO
4
is of primary importance
for an exhaustive comprehension of the electrochemical
properties of this cathode material.





Fig. 1 Graphical comparison between calculated (red
line) and observed (black line) PDF.




References
[1] M. S. Whittingham, Chem. Rev. 104 (2004) 4271
[2] M. S. Islam et al., Chem. Mater. 17 (2005) 5085

297
ELECTROCHEMICAL IMPEDANCE STUDIES OF THIN FILM
COMPOSITE LiFePO
4
/C

G.Bajars, G.Kucinskis, J.Smits, J.Kleperis, A.Lusis

Institute of Solid State Physics, University of Latvia

gunars.bajars@gmail.com

Keywords: electrochemical impedance spectroscopy, LiFePO
4
/C composite, thin film electrode

Introduction
LiFePO
4
is emerging as a promising cathode material for
lithium-ion batteries. In spite of such attractive features
as low cost, environmentally benign and thermal
stability, LiFePO
4
requires further modifications to
overcome limitations involving low electronic
conductivity and slow lithium-ion diffusion. Much effort
has been develop to improve the electrochemical
properties of LiFePO
4
by reducing the particle size, by
coating LiFePO
4
particles with carbon or performing
synthesis of LiFePO
4
/C composite. However, up to now,
it has been difficult to describe the kinetic behavior of
LiFePO
4
materials used for lithium-ion batteries because
many factors including particle size, ionic and electronic
conductivity, as well as phase transition kinetics affect
their electrochemical performance.
Use of thin film electrodes enable to investigate
electrochemical properties of active material in detail.
This is especially useful for poorly conductive materials
because the thickness can be reduced to an extent which
does not significantly affect the electrochemical
behavior. In this study, LiFePO
4
/C composite thin films
prepared by AC magnetron sputtering were investigated
by electrochemical impedance spectroscopy (EIS), which
is a very powerful method to determine the rate of
individual electrode kinetic steps.
Experimental
LiFePO
4
/C thin films were deposited by magnetron
sputtering on polished stainless steel. The target (150
mm in diameter) used for sputtering was obtained by hot
pressing the LiFePO
4
/C powder on a copper pad heated
to 130 C. As-prepared LiFePO
4
/C thin films were
annealed at 550 C for 1 h in the argon atmosphere to
increase the crystallization degree. EIS measurements
were carried out at various states of charge by applying
an AC signal of 10 mV amplitude over the frequency
range from 65 kHz to 1 mHz using a Solartron 1287
electrochemical interface with a Solartron 1260
frequency response analyzer. These tests were performed
using three-electrode cell with metallic Li as counter and
reference electrodes in 1 M LiClO
4
-PC electrolyte.

Results and discussion
Three element equivalent circuit was constructed to
analyze the properties of LiFePO
4
/C electrode -
electrolyte interface (Fig.1). The dependence of
equivalent circuit parameters, such as charge transfer
resistance R
ct
, Warburg impedance Z
W
and constant
phase element CPE on the state of charge were obtained
during the charge-discharge processes. It was found that
during the charging (as well as discharging) processes a
simplified equivalent circuit without Warburg impedance
is obtained (Fig.2) meaning that under the polarization
conditions the Li
+
ion diffusion is not limiting stage.


Fig.1 Nyquist plots and equivalent circuit model for
LiFePO
4
/C 50 nm thin film at OCP (3.4 V)


Fig.2 Nyquist plots and equivalent circuit model for
LiFePO
4
/C 50 nm thin film with applied charging current
(100 A cm
-2
)

The minimum value of charge transfer resistance is
acquired in discharged state which corresponds to
maximum concentration of Li
+
ions in lithium iron
phosphate. During charging Li
+
ions are extracted from
LiFePO
4
/C electrode which is accompanied with increase
of R
ct
. Obviously, the highest charge transfer resistance
should be when FePO
4
phase is dominating in
LiFePO
4
/C electrode.

Conclusions
Well-crystallized LiFePO
4
/C composite thin films were
prepared by AC magnetron sputtering in combination
with successive annealing. The kinetic behavior of
LiFePO
4
/C thin films were measured by the EIS. The
analysis shows that the charge transfer resistance
increases in relation with an appearance of the FePO
4

phase. Obtained results gave evidence that during the
charge-discharge process the Li
+
ion diffusion is not
limiting stage in thin film LiFePO
4
.
298
SYNTHESIS, STRUCTURE AND ELECTROCHEMICAL PROPERTIES FOR
RUTHENIUM-SUBSTITUTED LITHIUM MANGANESE OXIDE

Daisuke Mori,
1
Hikari Sakaebe,
2
Masahiro Shikano,
2
Hiroshi Kojitani,
1
Kuniaki Tatsumi,
2

Yoshiyuki Inaguma
1
1
Department of Chemistry, Faculty of Science, Gakushuin University,
1-5-1 Mejiro, Toshima-ku, Tokyo 171-8588, Japan
2
Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and
Technology (AIST), 1-8-31 Midorigaoka, Ikeda, Osaka, 563-8577 Japan

daisuke.mori@gakushuin.ac.jp

Keywords: Lithium manganese oxide, Cathode material, Lithium battery

Introduction
Lithium batteries have attracted much attention
as the key issue for hybrid electric vehicle (HEV) and
electric vehicle (EV) because of an interest in reducing
emission of carbon dioxide and use of alternative energy
sources. In order to meet these demands, the
development of cathode materials having higher capacity
and energy density is necessary. Layered manganese
oxide, Li
2
MnO
3
, is a promising candidate as an
alternative cathode material having higher voltage, larger
theoretical capacity and lower cost. Li
2
MnO
3
has similar
layered structure to the practical cathode materials
represented as LiMO
2
(M = transition metal) with
layered- rock salt type structure. It consists of lithium
layer and lithium-manganese mixed layer situated
alternately between closed packing oxygen layers.
Li
2
MnO
3
synthesized at low temperature shows
moderately high specific capacity of 260 mAh g
-1
but
poor cycling performance [1]. These were attributed to
low electronic conductivity and structural instability of
during lithium de-intercalation. Whereas, Li
2
RuO
3
with
the similar layered structure to Li
2
MnO
3
shows relatively
high specific capacity and good reversibility because of
the tightly composed host lattice [2]. In addition, it
exhibits lower electronic resistivity of 10 cm at room
temperature. In the present study, we synthesized the
ruthenium-substituted Li
2
MnO
3
to enhance the cycling
performance by stabilizing the host structure during the
lithium (de-)intercalation and the electrochemical
activity due to a decrease in electrical resistivity. The
relationship between synthesis condition, phase relation,
structure and electrochemical properties was
investigated.

Experimental
The ruthenium-substituted lithium manganese
oxide was synthesized by solid-state reaction. Li
2
CO
3
,
MnO
2
and RuO
2
used as starting materials were mixed,
pelletized and then sintered at 600-1200 C for 6-48 hours
in an oxygen gas flow. The powdered samples obtained
were identified by X-ray diffraction method. The sample
composition analysis was performed by SEM-EDS. The
electrochemical properties were investigated using a flat-
type HS cell. The positive electrode comprised the
sample, acetylene black, and PVdF binder coated onto
aluminum current collector. 1 M LiPF
6
in EC/DMC with
a volume ratio of 1:2 and lithium metal were used as the
electrolyte and the negative electrode, respectively.

Results and discussion
Li
2
Mn
1-x
Ru
x
O
3
was synthesized at 1200 C. It
formed two solid solutions in the whole composition
range with a structural transition from Li
2
MnO
3
type to
Li
2
RuO
3
type structure between x = 0.6 and 0.8, although
Li
2
Mn
1-x
Ru
x
O
3
synthesized at 800 C clearly separated
into two phases except for the narrow composition range
of 0 x 0.05. The electrical resistivity of Li
2
Mn
1-
x
Ru
x
O
3
decreased with increasing ruthenium content and
exhibited the lowest resistivity of 5.7 10
2
O cm at room
temperature with the composition of x = 0.8. The
ruthenium substitution would induce to increase the
overlap integral composed of transition metal t
2g
orbitals
and oxygen 2p orbitals in edge-shared MO
6
octahedra.
Fig. 1 shows charge and discharge properties of initial
cycle for Li
2
Mn
1-x
Ru
x
O
3
. The charge and discharge
capacities increased with increasing ruthenium content.
Li
2
Mn
0.4
Ru
0.6
O
3
exhibited a discharge capacity of 192
mAh g
-1
in the first cycle and 169 mAh g
-1
in the tenth
cycle, which indicates a relatively good cycle retention
[3]. The ruthenium substitution is quite effective to
improve electrical conductivity and charge-discharge
performance.



Fig. 1. Charge and discharge profiles of Li
2
Mn
1-x
Ru
x
O
3

in the voltage range of 2.0-4.8 V at 1/10C rate.

References
[1] D.Y.W. Yu, K. Yanagida, Y. Kato, H. Nakamura, J.
Electrochem. Soc. 156 (2009) A417A424.
[2] H. Kobayashi, R. Kanno, Y. Kawamoto, M. Tabuchi, O.
Nakamura, M. Takano, Solid State Ionics, 82 (1995) 25.
[3] D. Mori, H. Sakaebe, M. Shikano, H. Kojitani,
K. Tatsumi, Y. Inaguma, J. Power Sources, in press.
299
Sn@C Nanoparticles in Hollow Carbon Fibers: A Novel Powerful Anode for Li-based Batteries
Yan Yu,
1
Joachim Maier
2


1
School of Material Science & Engineering, Beihang University, Xueyuan Rd. 37#, Beijing,100191, China.
2
Max Planck Institute for Solid State Research, Heisenbergstreet 1, D-70569, Germany


corresponding_yanyu@buaa.edu.cn

Keywords: Li-based BatteriesAnode, Tin

Introduction
As an anode material for Li-based batteries, metallic tin
(Sn) has attracted tremendous interest owing to its huge
theoretical capacity of about 990 mAh/g associated with
the formation of Li
4.4
Sn, which is significantly higher than
that of graphitic carbon (372 mAh/g corresponding to
LiC
6
)
[1]
. However, substantial capacity fading after cycling
is a major obstacle especially for Sn-based materials due to
the enormous volume change during Li alloying and
dealloying. These volume changes caused severe
mechanical strain that induces cracks in the electrode
structure and loss of electrical contact at the anode.
[3]
Many
efforts have been made to increase its performance, such as
modifying electrode configuration, decreasing the particle
size,
[4]
using thin-films,
[5]
and selecting an optimized
binder.
[6]
Recently, the most relevant results have been
obtained by preparing nanostructured tin dispersed in a
carbon matrix
[7]
and carbon-encapsulated hollow tin
nanoparticles.
[8]
It has been demonstrated that in order to
improve the electrochemical performance, this tin anode
materials should be designed with higher content of
metallic tin, smaller particle size, hollow structure, and
only thinner carbon coating layer. Here, we fabricated Sn-
C composite hollow nanofibers as anodes for Li-ion
batteries by coaxial electrospinning technique.
[9]


Experiment
Two viscous liquids were simultaneously fed through the
inner (core, containing tributyltin (TBT) and mineral oil
solution) and outer (sheath, containing polyacrylonitrile
(PAN) solution) capillaries, respectively. The average
diameter of as-spun nanofibers is about 200~300 nm. Then,
the nanofibers were soaked in n-octane overnight to extract
the mineral oil. Finally, they were carbonized at 1000
o
C in
Ar/H
2
atmosphere to obtain the Sn nanoparticles
encapsulated in carbon hollow nanofibers. The obtained
Sn-C composite was investigated by X-ray diffraction
(XRD), scanning electron microscopy (SEM), and
transmission electron microscopy (TEM). Electrochemical
behavior of the test materials was examined via swagelok-
type cells with lithium metal counter electrode, celgard
2400 membrane separator, and electrolyte of 1 M LiPF
6

dissolved in a mixture of ethyl carbonate (EC) and
dimethyl carbonate (DMC) (1:1 in volume ratio). They
were cycle in the voltage range between 3.0 V and 0.01 V
with a battery test system (Arbin MSTAT).
Conclusion
The Sn-C hollow nanofibers offer a buffer zone to
compensate the large volume change during the Li-Sn
alloying and de-alloying process, which result in the
improvement of electrochemical cycle life. Whats more,
the hollow carbon nanofibers also improve the electrical
contact during the cycling. This Sn-C composite is a
promising as a potential anode material for Li-based
Batteries. Moreover, coaxial electrospinning has proved
itself to be a powerful routine for the preparation of
nanomaterials with hollow core/shell architectures.
Acknowledgements
Dr. Yan Yu is grateful for the scholarship from the
Alexander von Humboldt foundation.


References
[1] Y. Idota, T. Kubota, A. Matsufuji, Y. Maekawa, T.
Miyasaka, Science 276, 1395 (1997)
2] O. Mao, R. L. Turner, I. A. Courtney, B. D.
Fredericksen, M. I. Buckett, L. J. Krause, J. R. Dahn,
Electrochem. Solid-State Lett. 2, 3 (1999).
[3] N. Li, C. R. Martin, J. Electrochem. Soc. 148, 164
(2001).
[4] J. O. Besenhard, J. Yang, M. Winter, J. Power Sources
68, 87 (1997).
[5] A. M. Wilson, J. R. Dahn, J. Electrochem. Soc. 142,
326 (1995).
[6] M. Wachtler, M. R. Wagner,M. Schmied, M. Winter, J.
O. Besenhard, J. Electroanal. Chem. 510, 12 (2001).
[7] G. L.Cui, Y. S. Hu, L. J. Zhi, D. Q.Wu, I. Frieberwirth,
J. Maien, K.Mllen, Small 3, 2066 (2007).
[8] W. M. Zhang, J. S. Hu, Y. G. Guo, S. F. Zheng, L.S.
Zhong, W.G. Song, L. J. Wan, Adv. Mater. 20, 1160
(2008).
[9] I. G. Loscertales, A. Barrero, I. Guerrero, R. Cortijo, M.
Marquez, A. M. Ganan-Calvo, Science 295, 1695 (202).
300
GRAPHITE PARTICLES EMBEDDED IN SiCO CERAMIC MATRIX: NEW STABLE
HIGH-POWER ANODE MATERIALS FOR LITHIUM-ION BATTERIES

M. Graczyk-Zajac, L.Toma, C. Fasel, R. Riedel

Institut fr Materialwissenschaft, Technische Universitt Darmstadt, Petersenstr. 23, 64287 Darmstadt, Germany
graczyk@materials.tu-darmstadt.de

Keywords: Li-ion battery, anode, SiOC-graphite composite

Introduction
In the recent years continuous work has been devoted to
the search for new electrode materials in lithium ion
batteries. Currently, mostly graphitic materials are used
as negative electrodes due to low price and high
reversibility, despite relatively low capacity (372 mAh/g)
and inadequacy for high power applications. SiOC
(silicon oxycarbide) and SiCN (silicon carbonitride)
ceramic anode materials have been intensely studied in
the middle of nineties by Dahns group [1]. Nowadays,
there is a genuine interest in electrode materials based on
SiOC and SiCN polymer-derived ceramics (PDCs)
because their chemical and physical properties can be
designed by varying the chemical composition of the pre-
ceramic polymer [2,3,4]. Recently, it has been
demonstrated that the SiCN-graphite composites display
outstanding properties for high current charge/discharge
applications [5].
Results and discussion
In the present work we investigated the SiOC-graphite
composite material. A commercially available siloxane
resin (Polyramic RD-684, from Starfire Systems) was
used as ceramic precursor. The polymer was mixed in
3:1 and 1:1 starting ratio with graphite (SLP 50 graphite
from Timcal Ltd.) and pyrolysed at 1100 and 1300 C.
We studied two series of composites, one pyrolysed
directly after mixing (series R) and second for which the
thermal treatment was preceded by UV-initiated cross-
linking step (series CR). This second experimental
procedure leads to higher carbon content in the SiOC
ceramic matrix as compared to the first series [6]. The
final powders were tested electrochemically in
conventional two electrodes cells by means of cyclic
voltammetry and constant current charging. Elemental
analysis indicated that the Si-O-C composite material is
considerably free carbon rich. The higher amount of free
carbon has been found for CR samples. XRD
measurements confirmed that R und CR ceramic
composites have amorphous nature. Raman spectroscopy
experiment allowed us to determine the form of
disordered carbon present in the samples. We have found
that both types of composite, R und CR, demonstrate
excellent electrochemical behavior in terms of capacity
and cycling stability. In Fig.1 the comparison of cyclic
voltammograms obtained for composite materials,
graphite and pure SiOC ceramic is presented. The best
electrochemical activity resulting in the highest capacity
has been registered for CR 1:1 1100C sample. Due to
the presence of disordered carbons a significant part of
capacity is recovered between 1 and 2V. Fig. 2 presents
the galvanostatic cycling of CR composites and pure
Polyramic ceramic with increasing charge/discharge
current. The highest capacity has been recovered for
sample CR 1:1 1100C, which demonstrates also the best
high rate capability.

Fig. 1 Cyclic voltammogram (2
nd
cycle) of composite
materials (CR), graphite and pure ceramic, 50V/s

Fig. 2 Galvanostatic charge/discharge of composite CR
1:1, 3:1 and pure Poly R pyrolysed at 1100C
Almost 300mAh/g is recovered within 2 hours
charge/discharge. In similar conditions commercial
graphite electrodes recovers only about 150mAh/g.

Conclusion
Our study shows that composite materials based on
graphite and SiOC polymer-derived ceramics are
excellent candidates for anodes in Li-ion batteries. The
presence of high amount of disordered carbons increases
the lithium insertion capacity of composite. Studies
aiming to explain the high rate capability of these
composites are currently in progress.

Acknowledgments
The financial support of SFB595 and SPP1473 DFG
programs is acknowledged.

References
[1] W. Xing et al, J. Electrochem. Soc., 144 (1997) 2410.
[2] D. Ahn et al, J. Power Sources, 195(2010) 3900
[3] M. Graczyk-Zajac et al, J. Eur. Ceram. Soc. 30
(2010) 3235.
[4] H. Fukui et al, ACS Appl. Mater. Interfaces, 2 (2010)
998.
[5] M. Graczyk-Zajac et al, submitted to J. Power
Sources.
[6] S. Martinz et al, J. Eur. Ceram. Soc., (2010), doi:
10.1016/j.jeurceramsoc.2010.11.019.
301
SELF-ASSEMBLY OF LITHIUM BATTERIES WITH 3D ARCHITECTURE

Phil Johns,
1
Matthew J. Lacey,
1
Matthew R. Roberts,
1

Gaber Enany
2
and John R. Owen
1

1
School of Chemistry, University of Southampton, Highfield, Southampton, SO17 1BJ, United Kingdom
2
University of

J.R.Owen@soton.ac.uk

Keywords: 3D-Microbattery, Electrodeposition
Self Assembly of a Battery Cell.

Conventional battery manufacture is a complex operation
involving the processing of materials to achieve a small
particle size, coating onto thin metal foil current collectors,
laminating across a separator and then rolling or folding
into a compact package. This is fine for conventional
batteries, but far from optimum when the requirement is for
unusual shapes and sizes or placement into an inaccessible
location. Self-assembly is described here as a process
which deposits a complete battery cell on a surface of any
shape to fill the available volume. The 3D microbattery
below is an example of an application where self-assembly
is the best technique.

3-D battery concepts have been described [1] as vertical
extensions of thin film (1D) cells to increase the energy
stored per area of footprint, E
f
. The extra dimensions refer
to the direction of the current, which in a 3D cell does not
need to be in the direction perpendicular to the film.
Decoupling the current path (L) from the film thickness
thus allows a short electrolyte path between cathode and
anode, thus reducing the diffusion time (~L
2
/D) that
determines the power density (P
f
~ E
f
/ ). This is important
because it enables the use of poorly conducting electrolytes
in high power applications.

The self-assembly technique described here requires:
- A conducting substrate of arbitrary shape
and processes for:
- Conformal deposition of one electrode.
- Conformal deposition of a solid electrolyte
- Volume-filling deposition of the other electrode.

Electrodeposition of a lithium-ion electrode
The deposition of manganese dioxide by anodic oxidation
of Mn
2+
salts is well known, forming Electrolytic
Manganese Dioxide (EMD). In this work we have
demonstrated conformal deposition of this material onto
Reticulated Vitreous Carbon (RVC) as shown in Fig. 1.

The samples were calcined following Thackeray [3] to
avoid contamination of the lithium electrolyte and to give
structural stability under cycling in LiPF
6
/EC/DME
electrolyte. Some cycling data are given in Fig. 2.



Electrodeposition of Polymer Electrolytes

Polymer electrolytes have been deposited onto MnO
2
coated RVC by electropolymerisation of vinyl monomers
such as acrylonitrile. (See further details in another paper at
this meeting [3].)

RVC A RVC A
RVC A
RVC | EMD
B
RVC | EMD
B
B RVC | EMD


Fig. 1 RVC coated with EMD.












Cycling of MnO
2
-coated RVC in lithium electrolyte.

Acknowledgement
The authors would like to thank the EU FP7 project
Superlion for continued support.

References
[1] J. Long, B. Dunn, D. Rolison, H. White, Chem. Rev.
104 (2004), 4463
[2] El-Enany, G.; Lacey, M. J.; Johns, P. A.; Owen, J. R.
Electrochemistry Communications. 11 (2009), 2320-
2323
[3]Thackeray, M. Progress in Solid State Chemistry. 1997,
25, 1-71.
[4]Matthew J. Lacey,
1
Maciej Sosna,
1
Matthew R.
Roberts,
1

John R. Owen
1
, this meeting.

Superlion

0.00 0.04 0.08 0.12 0.16
2.0
2.5
3.0
3.5
4.0
E

/

V

v
s
.

L
i
Discharge capacity / mAh cm
-2
0.1 C
0.2 C
0.4 C
302
TOWARDS A BETTER UNDERSTANDING OF THE SURFACE REACTIVITY OF
LAYERED LITHIUM OXIDES

N. Andreu, H. Martinez, I. Baraille, R. Dedryvre, D. Gonbeau
*
IPREM/ECP (UMR 5254), University of Pau, Hlioparc, 2 av. P. Angot, 64053 Pau cedex 9, France

*
danielle.gonbeau@univ-pau.fr

Keywords : layered oxides, Li-ion batteries, DFT calculations, surface reactivity

Introduction
LiCoO
2
is the most widely used positive electrode material
of todays lithium-ion batteries. Its theoretical capacity is
272 mAh.g-1, but only ~ 140 mAh.g-1 capacity is reached
in practical cells. This value corresponds to the
deintercalation of half of the Li
+
ions from LiCoO
2
to
Li
0.5
CoO
2
upon charging the battery up to a 4.2 V cutoff
voltage. To obtain higher capacities with LiCoO2 the cells
have to be charged at higher voltages than 4.2V, to remove
more Li
+
ions from the structure. However, increasing the
charge cutoff voltage can lead to structural degradation of
LiCoO
2
and dissolution of cobalt in the electrolyte [1].
Both phenomena result in an increased capacity fading
upon cycling. These last years, it was shown that
modifying the surface of LiCoO
2
particles by application of
a metal oxide or phosphate coating (especially by
aluminium based coating as Al
2
O
3
or AlPO
4
) can
significantly improve the capacity retention upon cycling at
high voltages without loss of the initial reversible capacity.
The appearance of a LiCo
1-x
Al
x
O
2
solid solution between
the material and the Al-based coating has been proposed
and then evidenced by X-ray photoelectron spectroscopy
(XPS) [2,3]. Compounds of this solid solution have been
the subject of structural and electrochemical studies.
Although they show lower reversible capacities due to the
presence of electrochemically inactive Al
3+
ions they show
higher lithium intercalation potentials than LiCoO
2
and are
effective to limit cobalt dissolution at 4.5V (vs Li
+
/Li)[4].
Since dissolution phenomena are linked to acid-base
interactions at the electrolyte/electrode interface, we have
explored the surface acid-base properties of these materials
as compared to LiCoO
2
. In this way, we carried out
adsorption experiments of gaseous probes (NH
3
or SO
2
)
followed by XPS analyses which allow the identification of
active sites (except the weakest ones) and a quantitative
determination of their concentration. The results obtained
evidenced that the initial great basicity of LiCoO
2
decreases
significantly in LiCo
1-x
Al
x
O
2
compounds as a function of x
, which makes these materials less sensitive to acidic attack
in the electrolyte than LiCoO
2
[5].
Results
In this work, our aim was to understand, through a
theoretical approach, the large difference in surface
behavior due to Al/Co substitution and experimentally
evidenced. In this way, we have investigated the adsorption
of gaseous molecules (SO
2
, CO
2
) on layered lithium oxides
such as LiCoO
2
and LiAlO
2
, considering both (001) and
(110) surfaces corresponding to the plane of stacking and
to the densest plane. The calculations have been performed
using the ab-initio total-energy and molecular program
VASP developped at the Universitat Wien [6], a plane
wave basis set (kinetic energy cut-off at 400 eV), PAW
potentials [7] and the generalized gradient approximation
with the Perdew-Wang functional [8]. In each case, the
positions of all the atoms were optimized with a fixed
volume cell. Different adsorption modes are compared and
reveal some complex processes with the coexistence of
several interactions with the electron transfers occurring.
Adsorption on LiAlO
2
surfaces clearly differ from those on
LiCoO
2
for which, for example, SO
2
binds to (110) surface
generating sulfite or sulfate species in relation with the
coverage increase. This study contributes to a better
understanding of the influence of the nature of the cations
and of the cristallographic planes on the surface reactivity.

Fig. 1: SO
2
adsorption on (110) surface of LiCoO
2

References
[1]G.G. Amatucci, J-M. Tarascon, L.C.Klein, Solid State
Ionics, 83 (1996) 167
[2] S. Verdier,L. El Ouatani, R. Dedryvre, F. Bonhomme,
P. Biensan, D. Gonbeau, J. Electrochem. Soc. 154 (2007)
A1088
[3]A.T.Appapillai, A.N.Mansour, J.Cho, Y.Shao-Horn,
Chem. Mater. 19 (2007) 5748
[4] S. T. Myung, N. Kumagai, S. Komaba, H. T. Chung,
Solid State Ionics 139 (2001) 47
[5] L.Daheron, R. Dedryvre, H. Martinez, D. Flahaut, M.
Mntrier, C. Delmas, D. Gonbeau, Chem. Mater. 21
(2009) 5607
[6] G. Kresse, J. Furthmller, Phys. Rev. B 54, 11 169
(1996) / G. Kresse, J. Hafner, Phys. Rev. B 47 (1993) 558
[7] G. Kresse, J. Joubert, Phys. Rev. B 59 (1999) 1758
[8] J. P. Perdew,Y. Wang, Phys. Rev. B 45 (1992) 13244
303
INFLUENCE OF Li-ION DIFFUSION DISTANCE ON THE NEGATIVE ELECTRODE
PROPRIES OF Si THIN PLATELETS FOR Li SECONDARY BATTERIES

Morihiro Saito,
1
Tomoyuki Yamada,
1
Chihiro Yodoya,
2
Akika Kamei,
2
Masato Hirota,
2
Toshio Takenaka,
2

Akimasa Tasaka,
1
and Minoru Inaba
1
1
Department of Molecular Chemistry and Biochemistry, Doshisha University, Kyotanabe, Kyoto, 610-0321, Japan
2
Frontier Center, Oike & Co., Ltd., Kamitoba, Minami-ku, Kyoto, 601-8121, Japan

Email: minaba@mail.doshisha.ac.jp

Keywords: Si thin platelets, high capacity, cycleability, effects of thickness

Introduction
Li-Si alloy systems have attracted much attention as the
negative electrodes in lithium-ion cells because of its high
theoretical capacity (ca. 4200 mAh g
-1
) [1]. However, the
capacity retention of Si is poor, and it shows a high
irreversible capacity in the first charge/discharge cycle.
The poor retention is ascribed to a large volume expansion
and contraction during the charge/discharge cycling, which
causes particle fracture and electrochemical pulverization
(Fig. 1 (a)). Therefore, relaxation of the stress caused by
the large volume change is important to obtain a good
cycleability. Though this concept has been realized by
using Si thin films, such the nano-order thin films limit the
capacity of the cells (Fig. 1 (b)). In the previous study, we
proposed thin platelet-type Si particles (Fig. 1 (c)) and
demonstrated that the shape of thin platelet was quite
effective to improve the cycleability [2-4].
In this study, we prepared five Si thin platelets (Si leaf
Powder

(Si-LP), Oike & Co., Ltd.) with different


thickness (50, 100, 200, 300 and 400 nm), and investigated
the effects of thickness on the negative electrode properties
of Si thin platelets in more detail.

Experimental
The test electrode was fabricated by coating slurry on a
copper foil as a current collector. The slurry was prepared
by mixing 83.3 wt% active material sample, 5.6 wt%
Ketjen Black (KB) as a conductive agent and 11.1 wt%
carboxymethyl cellulose sodium (NaCMC) salt as a binder.
The electrochemical evaluation was performed by constant
current-constant voltage (CC-CV) charge/discharge tests
with a coin-type cell. The counter electrode was Li foil.
The electrolyte solution was 1 M LiPF
6
dissolved in
EC/DEC (1:1, by volume). The surface morphologies of
the electrodes before and after CC-CV tests were examined
by a scanning electron microscope (SEM).

Results and Discussion
Fig. 2 shows the cycleability of Si-LP samples with
different thickness. In the thickness range of 50-200 nm, all
the Si-LP exhibited excellent cycleability (around 1500
mAh g
-1
at 50th cycle). On the other hand, the discharge
capacity of thicker Si-LP samples (300 and 400 nm)
drastically dropped in several cycles and it became hard to
charge and discharge after 10 cycles [5]. In the case of Si
thin films, it is reported that a good cycleability is kept up
to around 300 nm, which is in good agreement with those
of Si thin platelets. This result was supported the SEM
images of Si-LP samples after 50 cycles. Consequently, it
was found that reduction of thickness down to < 200 nm is
necessary to attain a high capacity retention of the Si-LP.












Fig. 1. Schematic images for morphology changes of Li-M
alloy materials during charging. (a) particle, (b) thin film,
and (c) platelet.













Fig. 2. Cycleability of Si-LP electrodes (Si-LP + KB +
NaCMC = 83.3 : 5.6 : 11.1) at C/6 in 1 M LiPF
6
/EC+DEC
(1:1). Potential rage: 0.02 1.5 vs. Li/Li
+
.

This work was supported by Kyoto Environmental
Nanotechnology Cluster from MEXT, Japan.

References
[1] R.A. Huggins, J. Power Sources, 81-82, (1999) 13.
[2] M. Saito et al., ECS Trans., 25 (2010) 101.
[3] K. Nakai et al., Electrochemistry, 78 (2010) 438.
[4] M. Saito et al., J. Power Sources, in press.
[5] H. Guo et al., Mater. Sci. Eng. B, 131 (2006) 173.
M
Li
x
M
(a) Large particles
Li
+
(b) Thin films
cracks
(c) Thin platelets
Li
+ Li
x
M
M
Li
+ Li
x
M
M
0 5 10 15 20 25 30 35 40 45 50
0
500
1000
1500
2000
2500
3000
C
a
p
a
c
i
t
y

/

m
A
h

g
-
1
Cycle number
50 nm
100 nm
200 nm
300 nm
400 nm
304
NEW HIGH CAPACITY MIXED CONDUCTING ANODES FOR
ALL SOLID STATE LITHIUM ION BATTERIES

Rinlee Butch Cervera
*
, Tsuyoshi Ohnishi, Kazunori Takada
*
, Takayoshi Sasaki

International Center for Materials Nanoarchitectonics, National Institute for Materials Science, Namiki, 1-1, Tsukuba,
Ibaraki 305-0044, Japan

cervera.rinleebutch@nims.go.jp; takada.kazunori@nims.go.jp

Keywords: anode, solid state lithium ion battery, mixed conductor, lithium iron silicon sulfide

Introduction
Rechargeable solid state lithium ion battery is one of the
promising and attractive clean and safe energy
technologies for applications such as in portable electronic
devices and hybrid electric vehicles. Finding lithium ion
battery anodes with high specific capacity is considered an
attractive route to lower battery weight, volume, and
ideally cost. Much effort has been devoted in developing
anode materials like core-shell structures
1
, nanowires
2
, and
complex composites
3
utilizing liquid electrolytes. However,
such approaches may have the opposite effect for solid
electrolytes. The size reduction in these approaches
increases the surface area, and thus the amount of the
electrolyte that should be there increases which decrease
the fraction of the active material which is quite notable in
solid electrolyte systems.
On the other hand, mass transfer is prohibited except for
lithium ions in solid electrolytes. This feature will suppress
morphological change upon cycling causing performance
degradation in the active material. In this study, Li-Si-Fe
sulfides were investigated as an example showing high
capacity and large morphological change. Their electrode
performance as an anode for all solid state lithium ion
batteries is reported.
Experimental
Li
2-2x
Fe
x
SiS
3
samples with x=0 to x=1.0 has been prepared
by ball milling and subsequent heat treatment.
Stoichiometric amounts of Li
2
S, SiS
2
, and FeS were
weighed in a glove box. The weighed materials were ball-
milled for 24 hours at 500 rpm using a Fritsch Pulverisette
high-energy planetary ball mill. The ball to sample weight
ratio is approximately 1.0. The as ball-milled samples were
pelletized and vacuum sealed in quartz ampoules and
heated at 600
o
C for 10 hours. The ground powders were
characterized by XRD, SEM, and their electrochemical
performance was evaluated. The electrode properties of
pure anode and mixed solid electrolyte were investigated in
two-electrode cell configuration. The solid electrolyte used
is a glass ceramic with the composition of 70Li
2
S-30P
2
S
5
.

Results
Electrode properties of Li
2
SiS
3
were reported for its thin
film samples fabricated by PLD to show high capacity and
efficiency
4
. In this study, we tried to investigate if such
electrode properties are also observe for bulk powder
samples, which are more suitable for practical and
commercial applications. Figure 1 shows the electrode
performance of Li
2-2x
Fe
x
SiS
3
with x ranging from 0 to 1.0.
As can be observed, as the Fe content increases, the
specific capacity and efficiency greatly increases. The
specific capacities of LiFe
0.5
SiS
3
and FeSiS
3
reach about
800 and 1100 mAh/g with about 87% and 82% initial
efficiencies, respectively. The results of structural,
morphological, and electrochemical analyses (cycling, rate
capability, impedance analysis) of these samples will be
reported in detail.

Fig. 1 Electrode performance of Li
2-2x
Fe
x
SiS
3


Summary
New high capacity anode materials for all solid state Li-
ion batteries have been developed which shows high
specific capacity reaching about 1100 mAh/g with high
coulombic efficiencies.

References
[1] Jiang, C. et. al. Nanotoday, 1 (2006) 28.
[2] Chan, C. et. al. Nature Nanotechnology, 3 (2008) 31.
[3] Magasinski, A. et.al. Nature Materials, 9 (2010) 353.
[4] Hang, B.T. et.al. J. Power Sources, 195 (2010) 3323.
3.0
2.5
2.0
1.5
1.0
0.5
0.0
P
o
t
e
n
t
i
a
l
/
(
V

v
s
.

L
i
/
L
i
+
)
1400 1200 1000 800 600 400 200 0
Specific capacity/mAhg
-1
0 2 4 6 8 10
x in Li
x
FeSiS
3
Li
2
SiS
3
Li
1.7
Fe
0.15
SiS
3
Li
1.5
Fe
0.25
SiS
3
Li
1.0
Fe
0.5
SiS
3
FeSiS
3
x=0
x=0.15
x=0.25
x=1.0
x=0.5
current density = 50uA/cm
2
305
HYDROGEN PERMEATION THROUGH DENSE CERAMIC OXIDES; APPLICATIONS,
PRINCIPLES, MATERIALS, AND CHALLENGES

Truls Norby
1
1
Department of Chemistry, University of Oslo, FERMiO, Gaustadalleen 21, NO-0349 Oslo, Norway.

truls.norby@kjemi.uio.no

Keywords: defects, protonic, electronic; conductivity, mixed, proton, electron; hydrogen permeation;membrane

Introduction and applications
Hydrogen permeate through a range of materials in a
number of ways, such as molecular diffusion through
meso- and microporous materials, polymers, and glass,
atomic diffusion through metals, and mixed proton-
electron diffusion through ionic oxides. The latter
materials require and tolerate relatively high
temperatures, are in principle quite robust chemically
speaking, and thus suitable for hydrogen purification or
extracting hydrogen from steam reformed gas or coal
(syngas or coal gas). The hydrogen gradient driving force
and transport is assured by permeate side pumping, inert
sweep gas, or oxidation by air to form water.[1] It has
long been proposed to also use such hydrogen permeable
membranes to make higher hydrocarbons from methane
by dehydrogenation. This may form part of novel gas-to-
liquid routes, all the time new gas reservoirs are being
discovered but limited by the cost of transporting gas.
Principles
There are two relevant processes of dissolution of mobile
protons in oxides, namely hydrogen dissolving as protons
(in hydroxide groups) and electrons,
/
O
x
O 2
e 2 OH 2 O 2 ) g ( H + = +
-

or water dissolving as protons and oxide ions. The oxide
ions normally fill oxygen vacancies
- - -
= + +
O
x
O O 2
OH 2 O v ) g ( O H
or more rarely - structurally empty sites as interstitials:
//
i O
x
O
x
i 2
O OH 2 O 2 v ) g ( O H + = + +
-

The electrons often end up as minority defects, but
contribute enough to conductivity to make the material a
mixed proton-electron conductor. The flux of hydrogen
(H) is in the simplest case equal to the flux of protons
(H
+
) and has hydrogen chemical potential (hydrogen
partial pressure) gradient as the only driving force.
However, while oxygen permeable ceramics tend to have
no protons and remain purely oxygen permeable,
hydrogen permeable ceramics tend to have a certain
concentration of oxygen defects (notably vacancies) and
thus mixed proton - oxide ion - electron conductivity.
They thus exhibit permeation of hydrogen, oxygen, and
water, depending on the gradients of the three partial
pressures. Such a mixture of fluxes can be undesirable or
beneficial in a real process. In the general case, the flux
equations become complex, and their integration over the
membrane non-analytical, so that numerical solutions
become helpful, even more so if surface limitations come
into play. And interpretation of flux measurements
which relies on analyses of gas composition and flow of
the outlets of the feed and permeate sides can become
very difficult. Measurements of total and partial
conductivities as well as surface kinetics are in this case
indispensable as supporting information.
Materials
At present, many studies are performed on the classic
acceptor-doped perovskites like SrCeO
3
. Their modest
hydrogen permeabilities are often limited by the
electronic conduction, and composites containing an
electron conducting phase such as Ni seem a way to go.
Lately, much interest has turned to lanthanum tungstate,
with simplified formula La
6
WO
12
. It was early on
recognised that the actual single phase region covered
compositions with less lanthanum, e.g. La
5.6
WO
11.4
. This
material undoped exhibits slightly higher hydrogen
permeation than doped SrCeO
3
and it seems remarkably
stable with respect to both evaporation and CO
2

tolerance. Doping it has only negative effects, illustrating
that neither the actual structure nor the defect structure is
known at present. The materials exhibit considerable
oxide ion conductivity at higher temperatures, depending
on pH
2
O. La and W can be fully or partially replaced by
other rare earths and Mo, respectively, and their ratio
varied, and in this way the properties in terms of proton,
oxide ion, and electronic conduction as well as other
properties can expectedly be tailored.
Some dense ceramic materials including YSZ and
many mixed proton - oxide ion electron conductors
appear to exhibit permeation of neutral hydrogen species.
The species and transport mechanism in operation have
however not been identified unambiguously, and the
phenomenon thus more mysterious than useful for now.
Membrane stability issues arising from metal cation
diffusion will be discussed briefly.
References
[1] T. Norby, R. Haugsrud, Dense ceramic membranes
for hydrogen separation, in Membrane Technology,
Volume 2; Membranes for Energy Conversion, K. V.
Peinemann, S. P. Nunes (eds.), Wiley-VCH, Weinheim,
ISBN 978-3-527-31481-2 (2008) pp 169-216.
306
MICROSCOPIC OXYGEN EXCHANGE PROPERTIES OF CGO-LSCF COMPOSITES

J. W. Druce and J.A. Kilner
Department of Materials, Imperial College London, South Kensington, London, SW7 2BP

j.kilner@imperial.ac.uk

Keywords: CGO, LSCF, Composites, Oxygen Exchange,
Introduction
Composites of Ce
0.9
Gd
0.1
O
1.95
La
0.6
Sr
0.4
Co
0.2
Fe
0.8
O
3-

(CGO-LSCF) have received much interest as potential
cathode materials for intermediate temperature solid
oxide fuel cells (IT-SOFCs) [1,2], but are also promising
materials for electrochemical oxygen separation devices.

The role of the two constituent phases in oxygen
transport in the composites, particularly oxygen surface
exchange, has not yet been fully evaluated.
Understanding this behaviour could suggest ways in
which the microstructure may be optimised to improve
performance. This contribution will report mesurements
of the oxygen exchange properties of grains of each
constituent phase in a composite of CGO-LSCF.
Experimental
Dual phase samples containing equal parts by weight of
CGO and LSCF were fabricated from commercially
available powders. The powders were mixed by ball
milling, and the resulting blend pressed into pellets and
sintered at 1250 C for 4 hours.
The IEDP technique [3,4], was used to study the role of
the two component phases in the oxygen exchange
process for the composites. Samples were annealed in a
volume of oxygen gas which had been enriched in the
18
O isotope. The spatial distribution of the tracer was
subsequently studied by means of high lateral resolution
focussed ion beam (FIB)-SIMS.
Results and Discussion
Previously reported characterization of the fabricated
samples [5] showed systematic changes in the lattice
parameters of the constituent phases, suggesting the
occurrence of some cation inter-diffusion upon sintering.
Despite this, sputter depth profiling measurements within
individual grains of each phase suggested that the oxygen
tracer diffusion coefficients (D*) of the constituent
phases had not changed significantly from those
measured for the starting materials.
Apparent surface exchange coefficients (k) for the two
constituent phases were estimated from the SIMS data.
These values are termed apparent, because they do not
necessarily represent an intrinsic material property, but
include the influence of neighbouring grains of the other
component phase.
Figure 1 shows the temperature dependence of these
apparent surface exchange coefficients for the two
phases, compared with those of the bulk materials. The
apparent surface exchange coefficients of the CGO phase
are around an order of magnitude greater for the grains in
the composites than for the bulk material. Conversely,
the values for the LSCF grains appear to be an order of
magnitude lower for the grains in the composites
compared to those reported for the bulk material.
Interestingly, the apparent surface exchange coefficients
for the CGO grains in the composites are actually higher
than those for LSCF grains in the composites.
Fig. 1 Apparent surface exchange coefficients for
constituent phases in CGO-LSCF composite

These observed changes are thought to be due either to
the cation inter-diffusion previously observed for these
materials, or due a cleaning effect by which the LSCF
dissolves impurities from the CGO to reveal a bare CGO
surface with a higher surface exchange coefficient than
previously thought.

Conclusions
The oxygen transport properties of a CGO-LSCF
composite have been measured on the microscopic level.
To these authors knowledge, this is the first time that
these quantities have been measured on this scale for
composite materials.

Apparent surface exchange coefficients of grains of the
CGO and LSCF phases were seen to increase and
decrease respectively by an order of magnitude
compared to values reported for the single phase starting
materials.

References
[1] A. Esquirol, J.A. Kilner and N. Brandon, SSI, 175
(2004) 327.
[2] V. Dusastre and J.A. Kilner, SSI, 126 (1999) 163.
[3] J. Kilner and R.A. De Souza, Proc. 17
th
Ris Int.
Symp. Mat. Sci., (1996) 41.
[4] R.A. De Souza and M. Martin, MRS Bull, 34 (2009)
907.
[5] J. Druce and J.A. Kilner, ECS Trans, 16(51), (2009)
327.
307
EXPERIMENTAL DETERMINATION OF THERMOCHEMICAL DATA AND LONG
TERM STABILITIES FOR VARIOUS MIEC MEMBRANE MATERIALS FOR THE USE
IN OXYFUEL PROCESSES

T. Markus,

A. Mbius, D. Schlehuber, L. Singheiser

Forschungszentrum Jlich GmbH, IEK-2, D-52425 Jlich, Germany

T.Markus@fz-juelich.de

Keywords: MIEC, Membrane, CCS, Oxyfuel, Stability

Introduction
A new generation of fossil fueled power plants is
necessary to reduce the emission of CO
2
to protect our
environment and to avoid global warming. New power
plant concepts already exist involving oxygen supply
using gas separation membranes. Such membranes are
made of ceramic high temperature materials like
perovskites. Due to the fact that those perovskites should
be used for gas separation in future power plants it is
necessary to get familiar with the materials and their
specific behaviour.

In this study the oxygen non-stoichiometry, the thermo
chemical expansion coefficient (TEC) and the thermo
chemical stability of mixed ionic and electronic
conducting materials (MIEC) within the system (La,
Ba)
1-x
(Ba, Sr)
x
Co
1-y
Fe
y
O
3-
(x = 0.5, 0.7; y = 0.2, 0.8)
were investigated as well as the long term behaviour of
the permeation in dependence of membrane thickness
and oxygen partial pressure difference between feed and
permeate side of the membrane.
Experimental
The materials were prepared by solid state reaction. The
powders were pressed into pills and sintered dense. The
TEC and oxygen non-stoichiometry depending on the
doping element on the A-site, the cation stoichiometry
and the oxygen content in the ambient atmosphere were
studied in the temperature range between 200 C and
1000 C. The thermo chemical stability was investigated
by annealing experiments in the temperature range
between 600 C and 800 C whereas different gas
mixtures were applied.
The microstructure was investigated on the basis of
cross-fractions using Scanning Electron Microscopy. For
phase investigations an energy dispersive X-ray detector
was used.
The long-term permeation measurements were conducted
at 800 C over ~3000 h. The membrane (d = 15 mm; h =
1.0 mm) was placed in a quartz glass sample holder. The
feed-side of the membrane was flushed with synthetic air
(100 ml/min) and the sweep-side with argon which has a
low oxygen partial pressure of 410
-6
atm (100 ml/min).
The oxygen transport was monitored using a mass
spectrometer. The oxygen flux is calculated on the basis
of 2
nd
Ficks law.
Results and Discussion
Our results show, that the influence of the doping
element on the A-site, the cation stoichiometry and the
oxygen content in the ambient atmosphere on the thermo
chemical behaviour is significant. The thermo chemical
stability was investigated in CO
2
- and water vapour
containing atmospheres at different temperatures. Within
this study all materials showed more or less strong
degradation in dry CO
2
-containing atmospheres. In
humid CO
2
-free atmospheres only the material
Ba
0.3
Sr
0.7
Co
0.2
Fe
0.8
O
3-
(BSCF3728) was stable for at
least 200 hours. All materials showed degradation in
humid CO
2
-containing atmospheres (see figure 1).



Fig. 1 SEM images of a cross-section and cross-fractions
of BSCF3728 samples annealed in 20 vol% H
2
O(g) in
70 vol% CO
2
in synthetic air

The long-term permeation test on a 1.0 mm thick
LSCF58 membrane shows a good stability of the oxygen
permeation depending on time in CO
2
-free air. Variations
are smaller than 4 % of the mean value and thus within
the measurement error.
During cooling down the sample cracked because of the
differences in thermal expansion coefficients of LSCF58,
sealing (gold) and housing (quartz glass). The
decomposition of LSCF58 is a result of the non-
negligible cation mobility at elevated temperatures,
which is described in literature as kinetic demixing [1],
[2].
References
[1] D. Schlehuber, et. Al., J. Membr. Sci 351 (2010) 16
[2] T. Grande et. Al., Solid State Ionics 177 (2006) 177
308
OXYGEN PERMEATION THROUGH Ta-DOPED SrCo
0.8
Fe
0.2
O
3-


Wei Chen,
1
Chusheng Chen,
2
Louis Winnubst
1, 2*


1
Inorganic Membranes, IMPACT Institute for Energy and Resources, University of Twente,
P.O. Box 217, 7500 AE Enschede, The Netherlands.
2
Laboratory of Advanced Functional Materials and Devices, Department of Materials Science
and Engineering, University of Science and Technology of China, Hefei, Anhui 230026, Peoples Republic of China

a.j.a.winnubst@utwente.nl

Keywords: CO
2
sensitivity, oxygen permeation

Introduction
The most significant anthropogenic greenhouse gas is CO
2
,
which arises mainly from use of fossil fuels. Burning fossil
fuels with CO
2
-diluted O
2
produces a concentrated CO
2
gas
stream and thus enables efficient CO
2
capture, and the
O
2
/CO
2
mixture, necessary for combustion, could be
supplied by using CO
2
to transport the oxygen as separated
by the Mixed Ionic and Electronic Conductor (MIEC)
membrane from the air. However, most of the MIEC
materials, which usually have the oxygen deficient
perovskite structure, are not stable in CO
2
containing
atmosphere. In this work, Ta
5+
was doped into the
SrCo
0.8
Fe
0.2
O
3-
(SCF) lattice, aiming to enhance its
chemical stability in CO
2
.
Experimental
Both SCF and SCFTa (Sr(Co
0.8
Fe
0.2
)
0.9
Ta
0.1
O
3-
) were
synthesized using an EDTA complexation/pyrolysis
process, and then calcined at 1000
o
C for 5 h. Membranes
with 1 mm thickness were sintered at 1200
o
C (SCF) and
1237
o
C (SCFTa) in air for 10 h. Details of the oxygen
permeation setup can be found elsewhere[1].
Results and Discussion
To examine the performance of SCF and SCFTa
membranes under working conditions, i.e. under a CO
2

atmosphere, long term oxygen permeation measurements
were performed at 900C. The results, as given in figure 1
showed for SCF, a permeability of around 110
-6

mol/cm
2
/s when helium was used as sweeping gas which is
in accordance with literature[2]. However, when CO
2
was
induced, the flux suddenly dropped to 7.210
-7
mol/cm
2
/s,
followed by a decrease to almost zero after 57 h. For
SCFTa, the initial oxygen permeation flux in helium was
about 910
-7
mol/cm
2
/s, less than that of SCF. When CO
2

was introduced into the system instead of helium, the flux
dropped to 8.510
-7
mol/cm
2
/s, and the decrease was about
6%, which was much smaller than that of SCF (30%), what
is more, the flux kept almost constant in the following 50
h (7.810
-7
mol/cm
2
/s after 50 h).
The degradation in oxygen permeation of SCF is caused by
a reaction between SCF and CO
2
[3]. The formation of
strontium carbonate at the surface of the membrane can
completely block the permeation of oxygen. Ta substitution
in SCF greatly improves the chemical stability of SCF in
CO
2
, and the oxygen permeation fluxes keep stable.


Fig. 1 Time dependence of oxygen permeation flux
through 1 mm membranes at 900
o
C
: SCF in He; : SCF in CO
2
;
: SCFTa in He; : SCFTa in CO
2

Conclusion
Ta doping effectively enhances the stability of SCF in CO
2
,
making it more suitable in an oxyfuel combustion process.


References

[1] S. Xie, W. Liu, K. Wu, P.H. Yang, G.Y. Meng,
C.S. Chen, Solid State Ionics 118 (1999) (1-2) 23.
[2] L. Qiu, T.H. Lee, L.M. Liu, Y.L. Yang, A.J.
Jacobson, Solid State Ionics 76 (1995) (3-4) 321.
[3] Q. Zeng, Y.B. Zu, C.G. Fan, C.S. Chen, Journal
of Membrane Science 335 (2009) (1-2) 140.

309
LSFTa PEROVSKITE BASED MEMBRANES FOR OXYGEN SEPARATION

Jonas Gurauskis
-
, Espen Wefring, rjan Fossmark Lohne, Hilde Lea Lein and Kjell Wiik

Department of Materials Science and Engineering, Norwegian University of Science and Engineering,
Sem Slandsvei 12, NO-7491, Trondheim, Norway.

-
jonas.gurauskis@material.ntnu.no

Keywords: ceramic membranes, surface structuring, oxygen permeability.

Introduction
Dense ceramic membranes made from mixed
oxygen-ionic and electronic conducting (MIEC)
perovskite oxides can selectively separate
oxygen at temperatures above 700 C
[1]
. Due to
their promising potential as a reliable source for
oxygen and syngas production, without
significant environmental impact, they are
attracting a lot of interest from the scientific
community. This led to a number of promising
MIEC materials
[2]
, but there are still a lot of
issues to be considered mainly related to long
term chemical and thermal stability of
membranes but also insufficient oxygen surface
exchange kinetics
[3]
.

In this work the La
0.2
Sr
0.8
Fe
0.8
Ta
0.2
O
3-o

(LSFT2882) perovskite membrane system will
be evaluated, adressing its stability issues and
the efect of structural surface modification on
oxygen permeation rates.

Preparation of membranes
MIEC membranes were prepared using
LSFT2882 composition powder obtained by the
spray pyrolysis method. Porous substrates were
made by uniaxially pressing ~3 mm thick discs
of LSFT2882 powder with carbon black as a
pore former. Dense and thin functional layer
deposition was performed by dip coating
technique using colloidaly stable LSFT2882
powder suspension. Sintering of membranes
was done at 1230 C with 2 hours of dwell time.
Cross section of obtained membrane is shown in
figure 1. In case of membrane with structural
surface modification, presintering step of dense
functional layer at 1100 C was carried out,
followed by deposition of porous LSFT2882
composition layer on top and /or chemical
etching.

Evaluation of membranes
Evaluation of prepared MIEC membranes was
carried out at temperatures between 700 and
1000 C. Oxygen flux measurements were
performed by gas chromotography technique as
a function of oxygen partial pressure.
Effect of operating conditions on stability of
thin and dense LSFT2882 membrane layer was
evaluated by Energy Dispersive X-ray analysis
on the cross section of post-mortem samples.



Fig. 1. Fracture cross section of LSFT2882
composition membrane. Thin and dense
functional layer (~15 m) on top of a porous
substrate.

References
[1] H.J.M. Bouwmeester, Catal. Today 82
(2003) 141.
[2] J. Sunarso et.al., J. Membrane Sci. 320
(2008)13.
[3] E. Bartonickova et.al., J. Eur. Ceram. Soc.
30 (2010) 605.


310
HYDROGEN PRODUCTION FROM WATER SPLITTING USING MIEC
MEMBRANES: DETERMINATION OF THE RATE LIMITING STEP

R. Franca,
1
A. Thursfield,
1
S. Abdullah,
1
A.Murugan
1
,
C. Buysse,
2
F. Snijkers,
2
A. Buekenhoudt
2
and Ian S. Metcalfe
1

1
School of Chemical Engineering and Advanced Materials, Newcastle University, Merz Court, NE1 7RU,
Newcastle-upon-Tyne, UK
2
Separation and Conversion Technology, VITO NV, Boeretang 200, 2400 MOL, Belgium

rafael.franca@ncl.ac.uk

Keywords: Mixed conductors, oxygen transport membrane,
18
O isotopic exchange, water-splitting

Introduction
Candidate mixed ionic electronic conducting (MIEC)
perovskite-type metal oxides have been studied for
their ability to uptake and transport oxygen. In
membrane form they offer the potential to supply
pure oxygen from an oxygen-containing atmosphere.
The pure permeated oxygen may be collected or
utilised in e.g., methane oxidation processes such as
synthesis gas production [1]. Recently, perovskite-
type materials have been tested for their ability to
split water under an oxygen chemical potential
gradient established across a membrane using a
reducing gas atmosphere such as methane or carbon
monoxide. The resultant oxygen is transported
across the membrane and the hydrogen produced on
the water-supply side is collected as a product. The
perovskite-type materials so far investigated for
water splitting include BaCo
x
Fe
y
Zr
1-x-y
O
3-o
[2],
La
0.3
Sr
0.7
FeO
3-o
[3] and La
0.7
Sr
0.3
Cu
0.2
Fe
0.8
O
3-o
[4].
We employ pulsed
18
O isotopic exchange techniques
on powder samples and on membranes during high
temperature operation to probe the surface oxygen
exchange and bulk oxygen transport. The employed
pulsed technique has recently been employed on
mixed conducting perovskites [5]. Here we
investigate the performance of perovskite-type
MIECs of general formula; M
1-x
Sr
x
Co
1-y
Fe
y
O
3-o

(M=La or Ba) for their potential for membrane-based
hydrogen production from water splitting.

Experimental
Oxygen permeation studies using Perovskite-type
MIEC microtubes were first conducted to evaluate
the capability of the membranes to permeate oxygen.
The next step was to study the production of
hydrogen by chemically-driven water splitting. In
this study, the water stream was fed to one side of
the membrane and methane as a reducing gas to the
other side. The reactions occurring on both sides of
the membranes were simultaneously monitored by
two independent gas chromatographs.

Results and Discussion
Preliminary results of oxygen permeation at 900C
showed that Ba
0.5
Sr
0.5
Co
0.8
Fe
0.2
O
3-o
(BSCF5582)
microtubes possess higher oxygen permeation rates
than La
0.6
Sr
0.4
Co
0.2
Fe
0.8
O
3-o
(LSCF6428). However
LSCF6428 demonstrated better potential for
hydrogen production from water splitting than
BSCF5582. To understand the rate determining
step associated with hydrogen production, studies
on surface oxidation and reduction processes were
conducted. This focused on systematically varying
the partial pressures and flow rates of the methane
and water. To analyse the influence of the reducing
gas on hydrogen production with the BSCF5582
membrane, the concentration of methane was
doubled from 2.5% to 5.0%. The mole fraction of
hydrogen produced increased from 0.05% to 0.08%
indicating some surface limitations. Investigation of
the oxidation step was also carried out by increasing
the water partial pressure. Further work will focus
on surface oxidation processes of perovskites from
the M
1-x
Sr
x
Co
1-y
Fe
y
O
3-o
(M=La or Ba) family by
H
2
18
O pulsed isotope exchange (PIE). Temperature
programmed reduction and oxidation cycles will
also be carried out to understand the kinetic
behaviour and reaction steps involved. A detailed
kinetic analysis of the system will be presented.

Conclusions
Hydrogen production from water splitting depends
on the catalytic capability of the perovskite-type
membrane surface to split water and activate
methane. Fast bulk transport of O
2-
through the
membrane is not necessarily the main
consideration; the limiting step can be a surface
process.

References
[1] A. Thursfield and I.S. Metcalfe, J. Mater.
Chem., 14 (2004), 2475.
[2] H. Jian, H. Wang, S. Werth, T. Schiestel and J.
Caro, Angewandte Chemie-International Edition,
47 (2008), 9341.
[3] A. Evdou, L. Nalbandian and V.T. Zaspalis, J.
Membr. Sci., 325 (2008), 704.
[4] C.Y. Park, T.H. Lee, S.E. Dorris and U.
Balachandran, ECS transactions, 13 (2008), 393.
[5] H.J.M. Bouwmeester, C. Song, J.Zhu, J. Yi M.
Van Sint Annaland and B.A. Boukamp, Phys.
Chem. Chem. Phys, 11 (2009) 9640.
311
SURFACE CHEMISTRY IN ELECTROCHEMICAL NOx SENSORS

Monika Backhaus-Ricoult

Corning Incorporated, Corning, NY 14381, USA

corresponding_backhausm@corning.com

Keywords: NOx sensor, electrochemical cell, LSF

Introduction
Increasingly stringent regulations for gaseous emissions
from automotive exhaust and industrial processes require
the use of selective sensors for species such as
hydrocarbons, CO
2
, CO, NO
x
, SO
x
and particulates.
These species have to be detected in-situ in the various
stationary and mobile applications, at high temperature,
with high sensitivity and fast response. Different
concepts have been explored for sensing NOx, such as
bulk conductivity changes of semi-conducting oxides,
NOx adsorption-impacted diffusivity of oxygen through
a porous oxide layer or the modified response of
electrochemical cells in presence of NOx.
Electrochemical sensors typically consist of an
electrolyte, a sensing electrode and a counter electrode,
and either electrical current or cell resistance can be used
for monitoring NOx. Mixed perovskites are suited NOx
sensing materials. NOx competes with oxygen for the
adsorption sites, adsorbed NOx affects the oxide surface
chemistry and impacts the oxygen incorporation rates.
Adsorbed NO is a reductant and can induce changes in
the perovskite surface chemistry and modify its oxygen
exchange rate. While the electrochemical response of the
sensors is widely studied, related changes in the sensing
oxides surface remain rather unexplored. In this work,
we combined in-situ surface spectroscopy and impedance
spectroscopy to bridge that gap in understanding.
Experimental
Model sensor cells were made by screen printing
lanthanum strontium iron oxide sensing electrode,
counter electrode and Pt-based electric contacts on a thin
zirconia electrolyte film. Characteristic changes in
cathode impedance with bias were measured with a
Solartron impedance analyzer at high temperature (300-
700C) under controlled gas flow (mixtures of O
2
, N
2
,
NO or NO
2
). Surface compositions of sensing layer and
electrolyte and surface chemical states were studied by
Scanning Photoelectron Microscopy (SPEM) at the
Synchtrone Trieste. The sensor cells were successfully
operated in the SPEM chamber under bias at 350-400C
in O
2
/NO gas mixtures at 2 10
-6
mbar total pressure.
Spatially-resolved spectra and image-spectra were
acquired at an incident beam energy of 650eV.



Fig.1 Electrochemical sensor cell used for surface spectroscopy
Results
Performance testing of the sensor cells revealed specific
resistance change patterns with bias depending on the
surrounding gas (O
2
, NO, NO
2
). In oxygen-rich
environment, the oxygen incorporation rate at the sensing
layer was found to decrease progressively with
increasing cathodic bias. At low oxygen levels, a more
complex behavior was found with an initial increase with
bias, followed by a decrease of the oxygen incorporation
resistance with bias. The presence of NO decreased the
bias-free resistance and pushed the changing point to
higher cathodic bias.
The bias-induced changes in the oxygen exchange rate
were assumed to be related to changes in the LSF surface
chemistry. SPEM in-situ investigations revealed that LSF
surfaces did not undergo any changes in their iron
oxidation state or chemisorbed surface oxygen over a
wide range of oxygen pressure. In absence of NO, the
LSF catalyst preserved a stable oxidation state of iron
over a wide range of cell operation conditions. In
presence of NO, NO was selectively adsorbed on the
perovskite surface and not on the electrolyte. In a
temperature range of 350-450C, the NO-adsorption was
reversible and responded fast to changes in gas
composition. In the SPEM chamber (relatively high
NO/O
2
ratios of 0.1-1 and low total pressure of 2 10
-6

mbar), the cell impedance decreased with NO addition.
Chemisorbed NO introduced different changes in LSF
surface chemistry, ranging from reduction of surface iron
to nitrate-type interactions with Sr, La and Fe.
In addition, differences in the overall surface chemistry
were observed in presence of NO. LSF surfaces typically
exhibited strong Sr-enrichment as result of high
temperature processing. Under cathodic bias
<0.3V/600C, the electrical field drove strontium
migration on the device surface and improved the oxygen
exchange at the perovskite surface as result of lower Sr-
segregation. In presence of NOx, due to strong
interaction of surface strontium with NO, Sr-surface
migration in the electrical field was delayed.
Simultaneous electrochemical and spectroscopy in-situ
investigation provided new insight in the senor operation.

Fig.2 Impedance of LSF sensor cell as function of bias in 15000ppm
O
2
, 500ppm NO

[1] Backhaus-Ricoult et al., Solid State Ionics 179, 891 [2008]
312
FOCUSED ION-BEAM MILLING OF MULTILAYERS FOR
DIRECT MEASUREMENTS FROM LAYERS AND INTERFACES

Y. Kuru,
1,2,*
H. Jalili,
1
Z. Cai,
1
B. Yildiz,
1
and H.L. Tuller,
2
1
Department of Nuclear Science & Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
2
Department of Materials Science & Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA

ykuru@mit.edu

Keywords: Multilayer, Interface, Focused Ion-Beam Milling

Introduction
The possibility to controllably produce multilayers of
complex oxides, enabled by recent advances in thin film
deposition techniques such as pulsed laser deposition and
molecular beam epitaxy, has stimulated research related to
the observation of novel properties (e.g. enhanced ionic
transport) within the layers of a multilayer and/or in the
vicinity of the interfaces [1,2].
Although the local properties of the multilayers is the focus
of numerous studies, a general and practical method
enabling the measurement of the properties of inner layers
and buried interfaces locally remains absent. While some
methods have recently been proposed to characterize the
local electronic properties of multilayers, they involve
either complicated instrumentation or are applicable to
only certain types of specimens [3,4]. In this study, we
show that the inner layers and interfaces of multilayers can
be exposed to ambient conditions by focused ion-beam
(FIB) milling using a small incidence angle, the angle
between the specimen surface and the Ga+ ion beam. As a
model, the La
0.65
Sr
0.35
MnO
3
(LSM) / SrTi
0.2
Fe
0.8
O
3
(STF)
multilayer system is employed since these materials are of
importance to a broad range of devices such as fuel cells,
sensors, and magnetic junctions.
Results and Discussion
FIB milling, using a low incidence angle (o), is
schematically illustrated in Fig. 1a. As part of the surface is
exposed by the Ga+ ion beam at an angle o, the thin layers
of the superlattice become enlarged by a factor of 1/sino on
the newly formed surface (cf. Fig. 1b). This makes it much
easier to investigate the properties of these layers and the
interfaces between them, locally, using scanning probe
techniques.
The different LSM and STF layers are clearly distinguished
in atomic force microscopy images at the FIB cut. The
scanning tunneling microscopy images show that the LSM
layers are brighter than the STF layers, suggesting that the
LSM layers are more conducting. The measured band gap
values (i.e. 1 and 2 eV for LSM and STF, respectively) by
scanning tunneling spectroscopy (STS) are consistent with
this conclusion (cf. Fig 1c). Moreover, the band gaps of
LSM (i.e. ~ 1 eV) and STF (i.e. ~ 2 3.2 eV) reported in
literature also are consistent with the measured band gap
values [5-7].


Fig. 1 (a) Schematic description of the FIB milling process
with a shallow incidence angle, allowing the visualization
of the inner layers and the interfaces. (b) The top view of
the multilayer specimen after FIB milling. (c) STS data
measured for LSM and STF layers.

Conclusions
We demonstrate that it is possible to reach the inner layers
and interfaces of multilayers by FIB milling by using a
small incidence angle o, and thereby enabling the
measurement of local electronic properties in the LSM/STF
model system using scanning probe methods.
Acknowledgements
This research is being funded by the MIT Energy Initiative
Seed Fund Program and by Basic Energy Sciences,
Department of Energy under award DE SC0002633.

References
[1] J. Chakhalian et al, Science, 318 (2007) 1114.
[2] A. Gozar et al, Nature, 455 (2008) 782.
[3] N.P. Guisinger et al, ACS Nano, 3 (2009) 4132.
[4] M. Basletic et al, Nature Mat., 7 (2008) 621.
[5] V.L. Kozhevnikov et al, J. Solid State Chem., 158
(2001) 320.
[6] K. Zhao et al, Appl. Phys. Lett., 88 (2006) 141914.
[7] W. Jung et al, J. Electrochem. Soc., 155 (2008) B1194.
313
LOCAL STRUCTURE AND IONIC CONDUCTION AT HETERO-INTERFACE
OF SOLID ELECTROLYTES

Hirotoshi Yamada,
1
Yusuke Oga,
1
Daisuke Fukimoto,
2
Koki Urita,
2
Isamu Moriguchi
2
1
Graduate School of Science and Technology, Nagasaki University
2
Department of Applied Chemistry, Faculty of Engineering, Nagasaki University

h-yama@nagasaki-u.ac.jp

Keywords: Solid electrolyte, Hetero-interface, Space charge layer, Vacancy

Introduction
To improve performance of all-solid-state lithium
secondary batteries, it is important to understand
phenomena at interfaces between electrode and electrolyte
materials and fabricate ideal interfaces. At the solid-solid
interfaces, formation of electrical double layer by
difference of chemical potential and the change of ionic
conduction behavior by disorder of crystal structure are
expected. In the present study we prepared nano-
composites of active materials and a solid electrolyte that
have large interface area, and report the local structure and
ionic conduction at nano heterogeneous interface.
Experimental
Li
2
SiO
3
was employed as a Li
+
conductor, whose nano
particle (ave. 54 nm in diameter) was obtained by ball-
milling of commercial powder (particle size: 1~10 m).
Electrochemically active TiO
2
, Li
x
TiO
2
or FePO
4
nano
particles was mixed with the Li
2
SiO
3
nano particles using a
ball-mill. As TiO
2
nano particle, anatase TiO
2
(JRC-TIO-1,
Ishihara Sangyo) with an average particle size of 11 nm
was used. Li
x
TiO
2
(x ~ 0.01) nano particle was obtained by
chemical lithiation to the TiO
2
nano particle using n-BuLi.
FePO
4
nano particle was prepared by delithiation of
LiFePO
4
(average particle size 800 nm) using NO
2
BF
4
.
Each nano particle was annealed at 500C under an Ar
flow before mixing in order to remove adsorbates and
defects formed by ball-milling. Average particle sizes of
the nano particles were estimated by SEM and BET surface
area. Crystal structure of the nano particles and the
composites was characterized by XRD, and ionic
conductivity was measured by two-terminal AC method.
The samples were handled under an Ar atomosphere.
Results and Discussions
SEM-EDX analyses confirmed that Li
2
SiO
3
-TiO
2

composite was uniformly mixed. XRD indicated that
crystal structures of both Li
2
SiO
3
and TiO
2
were not
changed by the mixing process. Ionic conductivity was
enhanced by mixing TiO
2
and reached maximum when
40% of TiO
2
was mixed. Activation energies of all samples
were same (~1.2 eV), meaning that ionic conduction
mechanism was same, i.e., vacancy migration. The results
suggest the ionic conductivity in the Li
2
SiO
3
-TiO
2
system
is enhanced by the space charge layer effect,[1] which is
supposed to be given rise to by difference of chemical
potential of lithium ions between the active materials and
the solid electrolytes. To investigate this model, Li
x
TiO
2

that exhibits higher chemical potential of lithium, (Li
+
),
than TiO
2
, and FePO
4
with lower (Li
+
) were used as the
active materials. As shown in Fig. 1, ionic conductivity
decreased by Li
x
TiO
2
mixing. This is because vacancies
didnt increase at the interface and insulating Li
x
TiO
2

decreased total ionic conductivity of the composite. This
model was also supported by the fact that the crystal
structure and the activation energy of the conductivity
didnt change. On the other hand, FePO
4
increased the
ionic conductivity and decreased the activation energy.
XRD patterns showed expansion of the unit cell volume of
Li
2
SiO
3
by 1.3% (Fig. 2). This suggests that FePO
4
trapped
many lithium ions strongly, and thus generated many
negative vacancies expanded the lattice by the electrostatic
repulsion, which decreased the migration enthalpy of the
vacancies.

References
[1] J. Maier, Prog. Solid State Chem. 23 (1995) 171.
Fig. 2 XRD profiles of pristine Li
2
SiO
3
, pristine FePO
4
and
Li
2
SiO
3
+FePO
4
nano-composite.
Fig. 1 Temperature dependence of the Li
+
conductivity of
composites.
C
o
n
d
u
c
t
i
v
i
t
y

(
S

c
m
-
1
)
1000 / T (K
-1
)
Li
2
SiO
3
Li
2
SiO
3
+TiO
2
Li
2
SiO
3
+Li
x
TiO
2
Li
2
SiO
3
+FePO
4
1.6 1.7 1.8 1.9 2.0
10
-8
10
-7
10
-6
10
-5
I
n
t
e
n
s
i
t
y

(
A
r
b
.

u
n
i
t
)
2u (deg. (CuKo))
FePO
4
Li
2
SiO
3
Li
2
SiO
3
+FePO
4
25 30 35
314
OPTIMIZATION OF THE INTERFACES IN SODIUM SULFUR BATTERY

Zhaoyin Wen, Jinduo Han, Shufeng Song, Ying Huang

CAS Key Laboratory of Materials for Energy Conversion, Shanghai Institute of Ceramics,
Chinese Academy of Sciences, 1295 DingXi Road, Shanghai 200050, PR China

zywen@mail.sic.ac.cn

Keywords: Sodium sulfur batteries; Interface; Anticorrosion; Sealant

Interfacial factors played significant effects to the
performances of NAS battery which has been developed
remarkably during the last several years. SICCAS has
made particular researches on the optimization of the
interfaces systematically, which were summarized in this
paper.
(1) The optimization of the interface between cathode
and electrolyte. Due to the poor electronic conductivity of
sulfur, a porous electronic conductor, was usually provided
in NAS battery to permit free access of the cathodic
reactant material to the neighborhood of the ceramic
electrolyte. However, the pore structure of the felt and the
deposition of sulfur on the interface hindered the transport
of the liquid reactant material. Thus during continued
operation of the cell, the electrochemical properties
degraded gradually. To increase the flow of the liquid
reactant and to restrain the deposition of sulfur on the
interface between the cathode and the electrolyte, alumina
fiber (as shown in Fig.1(a)) was applied as a interfacial
layer. The results indicated that by the arrangement of the
alumina fiber, low resistance paths were provided
extending from the electrolyte across the cathodic region,
and a good capacity was obtained (as shown in Fig.1(b)).

0 100 200 300 400 500 600
2000
2100
2200
2300
2400
2500
2600
2700
2800
V
o
l
t
a
g
e
/
m
V
Capacity/Ah
without alumina fiber
alumina fiber was applied
80A

Fig. 1 Morphology of the alumina fiber (a) and the charge
capacity of the battery with or without alumina fiber (b).
(2) The optimization of the interface between
cathode and current collector. In NaS battery, the cathode
active materials sodium polysulfide are in the molten
state and extremely corrosive at the relatively high
operation temperature (approx. 350
o
C). To improve the
anticorrosion performance of the container material
against the molten sodium tetrasulfide,
La
0.8
Sr
0.2
Co
0.2
Cr
0.8
O
3
coating was plasma-sprayed onto
the container material 316 L stainless steel. The
potentiodynamic polarization results indicated that the
corrosion current density of the coating was about 0.25
mAcm
-2
, just one twentieth of that of the uncoated
SS316L (5.30 mAcm
-2
). Moreover, the coating displayed
a higher E
corr
, a relatively lower passive current density
and a more stable passive region. The results from dip
immersion tests showed that after dip-immersion in
sodium tetrasulfide at 350
o
C for 150 days, the coating
was covered by a compact corrosion layer which was
mainly composed of La and Cr sulfides with a thickness
of about 20 m, implying that the passive films formed
on the coating provided effective protection against the
corrosion (Fig.2)

Fig. 2 Cross-section image of the coating after immersion
in molten sodium tetrasulfide at 350
o
C for 150 days.
(3) The optimization of the interface between - and
-Al
2
O
3
. For safety and performance reasons, a high-
integrity seal between - Al
2
O
3
and - -Al
2
O
3
is essential.
A series of glasses composed of Si, Al, B, Na, K, Li and Bi
was used as the sealant. As seen in Fig. 1, the joints were
pore and microcracks free. The glass joined well with both
the - and -Al
2
O
3
, The -Al
2
O
3
/glass/-Al
2
O
3
interfaces
did not present microcracks after 200-times thermal shock,
demonstrating satisfactory thermal expansion match
between glasses and -,- Al
2
O
3
. Moreover, the joints
exhibited a diffusion layer of about 5 m in glass/-Al
2
O
3
interfaces, indicating that the -Al
2
O
3
was chemically
active and the elements diffused between -Al
2
O
3
and glass.
The formation of thin diffusion layer was helpful to the
joining between the glass sealant and -Al
2
O
3
, but without
significant influence in the property of -Al
2
O
3
electrolytes.

Fig. 3 The interface of -Al
2
O
3
/glass/-Al
2
O
3
(a) and the
interface of glass/-Al
2
O
3
(b).
References
[1] S.F. Song, J. Solid State Electrochem., 14 (2010)1735.
[2] Z.Y. Wen, J. Power Sources, 184 (2008) 641.
Financial supports from NSFC, CSTM and STCSM are
acknowledged.
a
b
a
-Al
2
O
3
glass
-Al
2
O
3
b
-Al
2
O
3

glass
315
TRANSPORT PROPERTIES AND STRUCTURE OF INTERFACES IN N-TYPE BARIUM
TITANATE CERAMICS

W. Preis,
1
J. Hou,
1,2
W. Sitte,
1
Z. Zhang,
2
G. Dehm,
2
T. Frmling,
3
H. Hutter,
3
J. Fleig
3
1
Chair of Physical Chemistry, Montanuniversitt Leoben, A-8700 Leoben, Austria.
2
Erich Schmid Institute of Materials Science, Austrian Academy of Sciences and
Materials Physics, Montanuniversitt Leoben, A-8700 Leoben, Austria.
3
Institute of Chemical Technologies and Analytics, Vienna University of Technology, A-1060 Vienna, Austria.

wolfgang.preis@unileoben.ac.at

Keywords: Bulk and grain boundary resistivities, Schottky barrier model, grain boundary diffusion, HRTEM

Introduction
Barium titanate is an interesting model system for the
investigation of interfacial transport properties of oxide
ceramics. Additionally, BaTiO
3
-based electroceramic
components are of high technological importance owing to
wide-spread application as, e.g., multilayer ceramic
capacitors (MLCCs) and positive temperature coefficient
resistors (PTCRs).
It is the aim of this contribution to present an overview of
the electrical properties of bulk and grain boundaries, the
structure of grain boundaries, and fast diffusion processes
along interfaces in n-conducting BaTiO
3
ceramics. The
electrical properties have been investigated by impedance
spectroscopy, the grain boundary structures have been
characterized by high resolution transmission electron
microscopy (HRTEM), and diffusion has been studied by
means of conductivity relaxation (chemical diffusion) as
well as
18
O-tracer diffusion experiments.

Electrical properties
Donor doped BaTiO
3
typically shows a steep increase of
the grain boundary resistivity by several orders of
magnitude above the ferroelectric-paraelectric phase
transition temperature (Curie-point). This PTC-effect is
usually attributed to Schottky-barriers (space charge layers)
at the grain boundaries, where the electrons of n-
conducting bulk regions adjacent to the interface are
strongly depleted. A modified Schottky barrier model has
been developed recently [1] in order to describe the
temperature characteristics of the grain boundary resistivity
of n-type BaTiO
3
ceramics. This space charge model will
be discussed in detail and compared with experimental data
for bulk and grain boundary resistivities obtained from
impedance spectroscopy.

Structure of grain boundaries
The structures of various grain boundaries in n-type barium
titanate ceramics have been investigated in detail by means
of C
S
-corrected HRTEM, including low-angle grain
boundaries, coherent twin boundaries, and incoherent large
angle grain boundaries (see Fig. 1) [2]. Moreover, based on
atomic resolved HRTEM-images the structure of the Ti-
rich secondary phase located at the triple junctions will be
presented.


Fig. 1: HRTEM-image of a large angle grain boundary

Fast diffusion along interfaces
A careful analysis of conductivity relaxation curves,
obtained at 1100 s T/C s 1250, revealed extremely fast
diffusion along the grain boundaries and slow rate-
determining transport of cation vacancies into the grains of
n-type BaTiO
3
[3]. Fast
18
O tracer diffusion along the grain
boundaries has been confirmed recently by ToF-SIMS [4].
While no tracer diffusion is observed in the Ti-rich
secondary phase, fast
18
O transport is found at the interface
between BaTiO
3
grains and the secondary phase. In
addition, conductivity relaxation curves will be presented
which consist of two separate relaxation times (see Fig. 2)
due to fast diffusion along the grain boundaries and
subsequent slow transport (re-oxidation) processes in
adjacent bulk regions.
0 20 40 60 80 100
0.015
0.016
0.017
0.018
0.019
0.020
0.021
0.022
T = 900C
oxidation step
p(O
2
): 0.10 0.50 bar


o

/

S

c
m
-
1
t / hours

Fig. 2: Conductivity relaxation curve for oxidation step

References
[1] W. Preis, W. Sitte, Solid State Ionics, 177 (2006) 2549.
[2] J. Hou et al., J. Eur. Ceram. Soc., (2010),
doi:10.1016/j.jeurceramsoc.2010.11.016.
[3] W. Preis, W. Sitte, Solid State Ionics, 177 (2006) 3093.
[4] T. Frmling et al., (2011), in preparation.

(001)
(1-10)
[111]
(10-1)
(0-11)
316
ELUCIDATING EQUILIBRIUM & NON-EQUILIBRIUM SURFACE ELECTROCHEMISTRY
IN DOPED CERIUM OXIDE USING IN-SITU XPS

William C. Chueh
1,*
, Anthony H. McDaniel
1
, Farid El Gabaly
1
, Michael E. Grass
3
, Yong Hao
2
, Zhi Liu
3
,
Kevin F. McCarty
1
, Sossina M. Haile
2
, Hendrik Bluhm
3

1
Materials Physics Department, Sandia National Laboratories, Livermore, CA 94551 USA
2
Materials Science, California Institute of Technology, Pasadena, CA 91125 USA
3
Lawrence Berkley National Laboratory, Berkeley, CA 94720 USA

wnchueh@sandia.gov

Keywords: Samarium-doped ceria, ambient-pressure x-ray photoelectron spectroscopy, impedance spectroscopy

Activities of metal oxide electrocatalysts are strongly
coupled to surface ionic and electronic disorders, which are
dependent on the equilibrium- and over-potential.
However, for materials relevant to elevated-temperature
solid-state ionics, understanding of surface states is limited
due to the lack of tools to probe materials under the
conditions of interest, i.e., under ambient-pressure and high
temperatures. For mobile defects such as oxygen ions,
conventional ex-situ experiments are not likely to freeze-in
the high temperature property of surfaces and, sometimes,
even the bulk.

In this work, in-situ ambient-pressure X-ray photoelectron
spectroscopy (Advanced Light Source, BL 11.0.2) [1] is
used in conjunction with electrochemical impedance
spectroscopy to study the surface behavior of samarium
doped ceria, an important electrode material for solid oxide
fuel cell and catalyst for a wide range of reactions.
Specifically, microfabricated electrochemical cells
consisting of epitaxial ceria thin film on yttria-stabilized
zirconia electrolyte were examined under equilibrium
(open-circuit) and non-equilibrium (biased) conditions
under controlled H
2
-H
2
O environments between ~ 400 and
650 C.

The near-surface polaron concentration (Ce
3+
/(Ce
3+
+Ce
4+
))
is quantified using the Ce 4f (valance band) and Ce 3d
photoelectron peaks. Simultaneously, a model-based
impedance approach based on chemical capacitance is used
to quantify the polaron and oxygen vacancy concentration
in the bulk regions of the film [1]. Upon varying the bulk
polaron concentration by a factor of 30, it was found that
the near-surface concentration remains nearly fixed at
100 %, i.e. the surface is completely reduced. The high
stability of near-surface polarons upon varying the gas-
phase oxygen activity under isothermal conditions
indicates that the entropy of defect formation is low,
plausibly due to surface defect ordering.

This research was supported in part by the US Department
of Energy under Contracts DE-AC04-94AL85000 (Sandia)
and DE-AC04-94AL85000 (LBL) and by the Sandia
Laboratory Directed Research and Development program.


References

[1] D. F. Ogletree et al. Rev. Sci. Instrum. 73, 3872
(2002).
[2] W. C. Chueh & S.M. Haile, Phys. Chem. Chem. Phys.
11, 8144 (2009).

317