HAL Id: hal-02057698
https://hal.archives-ouvertes.fr/hal-02057698
Submitted on 5 Mar 2019
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
β-NaMnO2: A High-Performance Cathode for
Sodium-Ion Batteries
Juliette Billaud, Raphaële J. Clément, A. Robert Armstrong, Jesús
Canales-Vázquez, Patrick Rozier, Clare P. Grey, Peter G. Bruce
To cite this version:
Juliette Billaud, Raphaële J. Clément, A. Robert Armstrong, Jesús Canales-Vázquez, Patrick Rozier,
et al.. β-NaMnO2: A High-Performance Cathode for Sodium-Ion Batteries. Journal of the American
Chemical Society, American Chemical Society, 2014, 136 (49), pp.17243-17248. �10.1021/ja509704t�.
�hal-02057698�
OATAO is an open access repository that collects the work of Toulouse
researchers and makes it freely available over the web where possible
Any correspondence concerning this service should be sent
to the repository administrator: tech-oatao@listes-diff.inp-toulouse.fr
This is an author’s version published in: http://oatao.univ-toulouse.fr/23055
To cite this version:
Billaud, Juliette and Clément, Raphaële J. and Armstrong, A. Robert and
Canales-Vázquez, Jesús and Rozier, Patrick and Grey, Clare P. and
Bruce, Peter G. β-NaMnO2: A High-Performance Cathode for Sodium-Ion
Batteries. (2014) Journal of the American Chemical Society, 136 (49).
17243-17248. ISSN 0002-7863
Official URL: https://doi.org/10.1021/ja509704t
β‑NaMnO
2
: A High-Performance Cathode for Sodium-Ion Batteries
Juliette Billaud,
†
Raphae
le J. Cle
ment,
‡
A. Robert Armstrong,
†
Jesu
s Canales-Va
zquez,
§
Patrick Rozier,
⊥
Clare P. Grey,
‡
and Peter G. Bruce*
,∥
†
School of Chemistry, University of St. Andrews, St. Andrews, Fife KY16 9ST, United Kingdom
‡
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, United Kingdom
§
Renewable Energy Research Institute, University of CastillaLa Mancha, 02071 Albacete, Spain
⊥
Institut Carnot CIRIMAT, CNRS UMR 5085, Universite
Paul Sabatier Toulouse III, Toulouse 31062, France
∥
Departments of Materials and Chemistry, University of Oxford, Oxford OX1 3PH, United Kingdom
*
S
Supporting Information
ABSTRACT: There is much interest in Na-ion batteries for
grid storage because of the lower projected cost compared
with Li-ion. Identifying Earth-abundant, low-cost, and safe
materials that can function as intercalation cathodes in Na-ion
batteries is an important challenge facing the field. Here we
investigate such a material, β-NaMnO
2
, with a different
structure from that of NaMnO
2
polymorphs and other
compounds studied extensively in the past. It exhibits a high
capacity (of ca. 190 mA h g
−1
at a rate of C/20), along with a
good rate capability (142 mA h g
−1
at a rate of 2C) and a good
capacity retention (100 mA h g
−1
after 100 Na extraction/
insertion cycles at a rate of 2C). Powder XRD, HRTEM, and
23
Na NMR studies revealed that this compound exhibits a
complex structure consisting of intergrown regions of α-NaMnO
2
and β-NaMnO
2
domains. The collapse of the long-range
structure at low Na content is expected to compromise the reversibility of the Na extraction and insertion processes occurring
upon charge and discharge of the cathode material, respectively. Yet stable, reproducible, and reversible Na intercalation is
observed.
1. INTRODUCTION
The renaissance of interest in sodium-based rechargeable
batteries has been driven by the greater and more uniform
Earth abundance of sodium, compared with lithium and, hence,
potentially lower cost.
1−6
The larger mass of Na, compared with
that of Li, leads to a lower specific capacity for sodium cells, with
respect to equivalent lithium cells, but this is no disadvantage for
static applications such as the storage of electricity on the grid. It
is the possibility of discovering sodium intercalation (insertion)
compounds that might outperform lit hium intercalation
compounds, leading to a new generation of sodium-based
rechargeable batteries, that is perhaps the most significant
motivation for the investigation of sodium intercalation
materials.
Potential sodium intercalation cathodes, such as 3D frame-
work compounds, especially those based on the NASICON
structure, have received considerable attention because of the
high Na
+
conductivity of the solid electrolyte, Na
3
Zr
2
Si
2
PO
12
,
with a similar structure.
1−3,7−10
The layered Li transition metal
oxide LiCoO
2
, and related materials have been the dominant
cathodes for lithium-ion cells.
11−16
Layered Na transition metal
compounds, NaMO
2
, exhibit extensive intercalation chemistry,
more so than their Li counterparts. For example, both NaFeO
2
and NaCrO
2
are electrochemically active in contrast to their
lithium analogues,
17,18
and NaMnO
2
compounds can sustain
sodium deintercalation without conversio n to t he spinel
structure, unlike layered LiMnO
2
.
19−21
A number of recent studies on sodium intercalation
compounds have focused on Earth abundant and, hence, low-
cost transition metals, especially Mn and Fe. Of the layered
Na
x
MnO
2
compounds, α-NaMnO
2
, which exhibits a monoclinic
distortion of the O3 crystal structure of LiCoO
2
(ABC oxygen
stacking), and P2−Na
0.67
MnO
2
(ABBA oxygen stacking) have
been widely studied as sodium positive electrode materi-
als.
19,22−24
The reversibility of the deintercalation process in
the P2 polymorph is enhanced by Mg doping.
25
Similarly,
substitution of Mn by lithium in P2 Na
x
[Mn
1−y
Li
y
]O
2
compounds leads to improved reversibility of the charge (that
is, the electrochemical Na insertion) process.
26,27
Solid solutions
of Mn and Fe, Na
x
[Mn
1−y
Fe
y
]O
2
, adopting either O3 or P2
structures, have also been investigated.
28,29
β-NaMnO
2
possesses a different layered structure from that
conventionally adopted by NaMO
2
type structures. Instead of
doi.org/10.1021/ja509704t
planar layers of MnO
6
octahedra that simply alter their stacking
sequence to generate the different polymorphs (O3, P2, P3,
etc.),
30
β-NaMnO
2
is composed of zigzag layers of edge sharing
MnO
6
octahedra between which Na
+
ions reside in octahedral
sites, Figure 1.
31
The structure possesses an orthorhombic symmetry, space
group Pmnm, with cell parameters a = 2.86, b = 4.79, c = 6.33 Å.
The c axis is perpendicular to the layers in this setting.
31
Sodium
deintercalation was first investigated by Mendiboure et al.,
demonstrating reversible removal of 0.15 Na at a potential of
around 2.7 V vs Na
+
/Na.
22
A phase formed upon electrochemical
Na extraction from NaMnO
2
was identified with the same space
group as the pristine (as-synthesized) phase, but with a
significant reduction of the Jahn−Teller distortion, associated
with Mn
3+
to Mn
4+
oxidation. Here we show that β-NaMnO
2
can
exhibit a first discharge capacity (that is, the capacity to
electrochemically reinsert Na in the material) as high as 190
mA h g
−1
, corresponding to the reinsertion of 0.82 Na per
formula unit. A discharge capacity of ca. 130 mA h g
−1
is retained
after 100 cycles.
2. EXPERIMENTAL METHODS
2.1. Synthetic Procedures. The β-NaMnO
2
samples were
prepared by solid-state synthesis. The solid-state route involved mixing
together Na
2
CO
3
and Mn
2
O
3
; 15% weight excess of sodium was used in
order to compensate for Na
2
O evaporation on firing. Two firing steps
were necessary, first at 950 °C for 24 h, after a temperature ramp of 1
°C/min, and second at 950 °C for 24 h, after ramping the temperature at
a faster heating rate of 5 °C/min. Both firing steps were performed on
pellets under oxygen flow, and followed by a quench to room
temperature. The samples were then transferred to an Ar- filled glovebox.
2.2. Powder X-ray Diffraction. Powder X-ray diffraction (XRD)
was performed on a Stoe STADI/P diffractometer operating in
transmission mode with Fe Kα
1
radiation (λ = 1.936 Å) and using a
capillary to avoid contact with the air. The Diffax program was used to
model the diffracted intensities.
32
2.3. In Situ X-ray Diffraction. In situ X-ray diffraction data were
collected on a Bruker D8 diffractometer operating in Bragg−Brentano
geometry with Cu Kα
1
radiation (λ = 1.5416 Å). The setup consisted of
an in situ cell with an X-ray transparent Beryllium window (thickness of
200 μm). To prevent Be oxidation at high potentials (above 3.8 V vs
Na
+
/Na), a protective Al foil (thickness 10 μm) was placed between the
Be window and the powder under study. The cell was connected to a
Biologic cycler and the evolution of the potential was recorded as a
function of the time.
2.4. Transmission Electron Microscopy. TEM was performed on
a Jeol JEM 2100 electron microscope operating at 200 kV and equipped
with a double-tilt (±25°) sample holder, an EDS detector (Oxford
Link), and an Orius SC200 CCD Camera. TEM specimens were
prepared by dispersing the oxides in dry hexane under an inert
atmosphere, and depositing a few drops of the suspension on to a holey
carbon-coated copper grid (EMS). TEM images were analyzed using the
Digital Micrograph software from Gatan.
2.5. Electrode Preparation. Composite electrodes were cast on
aluminum foil in an Ar-filled glovebox to prevent air oxidation. The
slurry was prepared by mixing β-NaMnO
2
, super S carbon, and Kynar
Flex 2801 as binder, in weight ratios of 75:18:7, in THF. Electrodes were
incorporated into coin cells (CR2325 type) with a sodium metal counter
electrode, and with an electrolyte solution composed of 1 M NaPF
6
in
ethylene carbonate/propylene carbonate/dimethyl carbonate, in weight
ratios of 45:45:10, respectively. Typical electrode active material
loadings were ca. 4−5 mg/cm
2
. The electrode used for in situ powder
X-ray measurements was prepared using the same composition as
described above (up to 20 mg per cell), but in powder form. The samples
for ex situ measurements (TEM, PXRD, and Na NMR) were prepared
by extracting the cathode material from the coin cells and washing it with
dry dimethyl carbonate (DMC). The solvent was then evaporated. All
steps were performed in an Ar-filled glovebox. The resulting powder was
stored in an Ar-filled glovebox for further characterization.
2.6. Electrochemical Measurements. Electrochemical measure-
ments were carried out at room temperature using a Maccor Series 4200
battery cycler.
2.7. Solid-State NMR. NMR experiments were performed under 60
kHz MAS, using a 1.3 mm double-resonance HX probe.
23
Na 1D spin
echo spectra were recorded at room temperature on a Bruker Avance III
200 wide-bore spectrometer, at a Larmor frequency of −77.9 MHz, and
23
Na NMR chemical shifts were referenced against NaCl.
23
Na spin echo
spectra were acquired using a 90° RF (radio frequency) pulse of 1 μsat
25.04 W, a 180° pulse of 2 μs at 25.04 W, and a recycle delay of 30 ms.
23
Na RF pulses were assumed to be selective for the
23
Na central
transition.
2.8. Chemical Analyses. Chemical analyses were performed by
Inductively Coupled Plasma (ICP) emission spectroscopy.
3. RESULTS AND DISCUSSION
3.1. Electrochemistry. β-NaMnO
2
samples were synthe-
sized, characterized and incorporated into electrochemical cells
as described in the experimental section. Load curves
(corresponding to electrochemical Na extraction and reinser-
tion) for β-NaMnO
2
are shown in Figure 2.
Figure 1. (a) Schematic representation of β-NaMnO
2
in the Pmnm
space group and (b) intergrowth model between α and β-NaMnO
2
.
MnO
6
octahedra are pink, NaO
6
octahedra are yellow, and O atoms are
red. Adapted from Abakumov et al.
36
copyright 2014 American
Chemical Society.
Figure 2. Load curves for β-NaMnO
2
at a rate of C/20 (10 mA g
−1
). The
1st, 2nd, 5th, and 10th Na extraction/reinsertion cycles are represented
in black, red, blue and green, respectively.
The 200 mA h g
−1
obtained on the first charge commences
with an extended plateau between NaMnO
2
and a phase with a
composition close to Na
0.57
MnO
2
, followed by a rising potential
interrupted by small steps at Na
0.49
MnO
2
and Na
0.39
MnO
2
.
These steps are still present upon discharge, but are less
pronounced, and the same extended voltage plateau is observed
at the end of discharge. There is an irreversible loss of capacity
corresponding to 25 mA h g
−1
(equivalent to 0.1 Na per formula
unit). The shapes of the load curves are almost invariant on
cycling, and only exhibit a small but continuous reduction in
capacity associated mainly with the voltage plateau. The 2.7 V
plateau is associated with the phase transition between the Jahn−
Teller distorted and undistorted structures and exhibits a small
polarization (below 150 mV).
3.2. Structural Characterization. 3.2.1. Structure of the
As-Prepared Material. To understand the structural changes
that accompany sodium deintercalation and reinsertion, the
structure of β-Na
x
MnO
2
was monitored by a combination of
powder X-ray diffraction, solid-state NMR and high-resolution
transmission electron microscopy. The PXRD pattern of the as-
prepared material is shown in Figure 3.
It is apparent from Figure 3 that the structure, although clearly
based on that of β-NaMnO
2
, cannot be described by the ideal
structure for this compound, see for example the (011) peak in
the region highlighted by the ellipse. Diffax has been used to
simulate the effect of introduci ng random stacking faults
corresponding to the insertion of a monoclinic α-NaMnO
2
cell
between two blocks of orthorhombic symmetry, as shown in
Figure 1.
32
Stacking faults are not uncommon in battery
materials.
33,34
Those observed here are similar to the micro-
twinning seen in the Ramsdellite form of MnO
2
, although in β-
NaMnO
2
there is also a Jahn−Teller distortion.
34
The stacking
faults in β-NaMnO
2
are most closely related to those in the
isostructural and similarly Jahn−Teller distorted LiMnO
2
material, albeit in a higher proportion (1−7% occurrence in
LiMnO
2
).
35
For β-NaMnO
2
, the experimental powder di ffrac-
tion pattern is well represented by a structural model composed
of 25% stacking faults. Simulations using other proportions of
stacking faults are shown in Supporting Information Figure S1.
While the agreement between the experimental powder
diffraction pattern and the simulation pattern with 25% stacking
faults is compelling, and certainly demonstrates that this material
cannot be described by a single structure, Diffax cannot reveal the
detailed nature of the structural complexity. A recent in-depth
HRTEM study of β-NaMnO
2
confirms that this material is
composed of structural motifs built from the α and β crystal
structures.
36
The two structures are energetically very similar and
they can form a low energy phase boundary (a twin plane), where
the MnO
6
layers in the α and β phases are oriented at
approximately 60° to each other, as indicated in Figure 1 b).
36
Intergrowth of blocks of the α and β-NaMnO
2
crystal structures
of different sizes leads to various intermediate structures. Our
own TEM data, coupled with NMR data, reinforce this recent
interpretation, as shown in Supporting Information Figure S2
and in Figure 4. Our as-prepared material is composed of regions
exhibiting an ideal β-like stacking sequence and regions in which
Figure 3. X-ray diffraction patterns of (a) the ideal β-NaMnO
2
structure
in the Pmnm space group, (b) simulated with 25% stacking faults, (c) as-
prepared β-NaMnO
2
and (d) after 5 cycles. Highlighted with the blue
circle is the major difference between the ideal and experimental data
with the (011) peak greatly broadened.
Figure 4. Ex situ
23
Na spin echo NMR spectra obtained at different
stages of the first electrochemical cycle, under an external field of 200
MHz and at a spinning frequency of 60 kHz. Spinning sidebands are
marked with an asterisk (*). The three regions containing the
resonances of Na atoms in a pure α environment, in a pure β
environment, and in the vicinity of a stacking fault are highlighted in
green, red, and blue, respectively. These regions are not valid for the
sample with lowest Na content (Na
0.236
MnO
2
), for which significant
disorder in the crystal structure leads to broadening of the NMR peaks,
and Mn
3+
to Mn
4+
oxidation induces a larger Fermi contact shift with all
Na resonances being shifted toward the LHS of the spectrum. The peak
near 0 ppm is due to Na
+
in a diamagnetic environment, most probably
from residual electrolyte or its decomposition products formed during
cycling.