Nanofibrous α-, β-, γ- and α·γ-manganese dioxides prepared by the hydrothermal-electrochemical technique

Article abstract of DOI:10.1149/1.1583717

The hydrothermal-electrochemical method has been used for the synthesis of manganese dioxides from acidic MnSO₄ or A₂SO₄/MnSO₄ (A = Li, Na, K, NH₄) solutions. This method produced manganese dioxides of either the β-, γ- or α- structure types, mixtures of these structure types, α/γ or γ/β, or interconnected α- and γ-phases (α·γ). The structure of the material obtained can be controlled by adjusting the parameters of the synthesis: temperature, acidity of the solution, composition of the solution, or applied current density. The γ-MnO₂ materials synthesized are characterized by exceptionally low amounts of microwinning defects. These materials can be synthesized over a wide range of P_r values, including in the range usually reserved for chemically prepared (chemical manganese dioxide, CMD) materials. Materials exhibiting lower P_r values are favored when the pH or the temperature of the synthesis is decreased. The lowest P_r values were found in materials synthesized at 92°C and pH 0. A study of the morphology of the deposits has shown that it can be controlled by changing the experimental conditions.

Full text of DOI:10.1149/1.1583717

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D135
0
013-4651/2003/150͑8͒/D135/14/$7.00 © The Electrochemical Society, Inc.
Nanofibrous ␣-, ␤-, ␥- and ␣"␥-Manganese Dioxides Prepared
by the Hydrothermal-Electrochemical Technique
I. Synthesis and Characterization
z
Laurie I. Hill, Alain Verbaere, and Dominique Guyomard
Laboratoire de Chimie des Solides, Institut des Mat e´ riaux Jean Rouxel, CNRS-Universit e´ de Nantes,
44322 Nantes Cedex 3, France
The hydrothermal-electrochemical method has been used for the synthesis of manganese dioxides from acidic MnSO4 or
A SO /MnSO (A ϭ Li, Na, K, NH ) solutions. This method produced manganese dioxides of either the ␤-, ␥- or ␣- structure
2
4
4
4
types, mixtures of these structure types, ␣/␥ or ␥/␤, or interconnected ␣- and ␥-phases ͑␣•␥͒. The structure of the material
obtained can be controlled by adjusting the parameters of the synthesis: temperature, acidity of the solution, composition of the
solution, or applied current density. The ␥-MnO2 materials synthesized are characterized by exceptionally low amounts of
microtwinning defects. These materials can be synthesized over a wide range of P values, including in the range usually reserved
r
for chemically prepared ͑chemical manganese dioxide, CMD͒ materials. Materials exhibiting lower P values are favored when the
r
pH or the temperature of the synthesis is decreased. The lowest P values were found in materials synthesized at 92°C and pH 0.
r
A study of the morphology of the deposits has shown that it can be controlled by changing the experimental conditions.
©
2003 The Electrochemical Society. ͓DOI: 10.1149/1.1583717͔ All rights reserved.
Manuscript submitted July 8, 2002; revised manuscript received January 13, 2003. Available electronically June 16, 2003.
Manganese dioxides are of technological interest due to their
potential use as cathode materials for primary and secondary lithium
and alkaline batteries. Many manganese dioxide polymorphs are
known, including layered and tunnel structures. Two manganese di-
structure, the amount of which has been referred to as Mt ͑in
1
9
percent͒, instead of Tw; where the quantity 2Mt roughly corre-
sponds to Tw. More detailed information on the distribution statis-
tics of the microtwinning defects can be found in Ref. 18 and 19.
oxides of particular interest to the battery industry are ␣-MnO2
Using this method, the ␥-MnO materials prepared for commercial
2
͑
secondary batteries͒ and ␥-MnO ͑primary and secondary batter-
2
use are characterized by a P Ϸ 50 and Mt у 50. However, it has
r
ies͒. Both structures are built up of chains of edge sharing MnO6
octahedra ͑single or double͒ which share corners to form channels
been suggested that the reversible lithium insertion capacity of
␥-MnO materials should increase when the amounts of pyrolusite
2
2
0,21
through the structure. In ␣-MnO , with the Hollandite structure, the
and microtwinning defects decrease.
Thus, some research has
2
chains of octahedra connect to form 2 ϫ 2 as well as 1 ϫ 1 chan-
focused on efforts to synthesize pure ramsdellite or materials
approaching the ramsdellite limit.
2
0,22
nels. In ␥-MnO , it is widely accepted that the octahedra are con-
2
nected such that channels of dimensions 2 ϫ 1 ͑characteristic of
ramsdellite-MnO ) and 1 ϫ 1 ͑characteristic of pyrolusite-MnO )
are formed, as proposed by de Wolff.
With this goal in mind, and with the broader aim of stabilizing
new or modified MnO structures, we have chosen to explore the
2
2
2
1
hydrothermal-electrochemical synthesis technique for the prepara-
tion of manganese dioxides. Since hydrothermal conditions tend to
favor the formation of well-crystallized products, this technique may
␣
-MnO is usually prepared by chemical methods utilizing two
2
or more steps, with the large 2 ϫ 2 channels of the structure stabi-
ϩ
ϩ 2-8
allow the synthesis of better-crystallized MnO compounds than by
lized by large cations such as K or NH4 . However, for battery
applications, these large cations are thought to inhibit the mobility of
2
traditional electrochemical methods, and thus obtain materials con-
taining fewer defects. The hydrothermal-electrochemical method has
recently been used for the synthesis of a variety of compounds such
ϩ
ϩ
Li through the tunnels, and thus can be ion-exchanged for H ͑as
ϩ 9
ϩ
10,11
H O ) or Li to improve performance.
Alternatively, ␣-MnO2
3
23
24
ϩ
as mixed titanium oxides and LiMO2 (M ϭ Ni,Co) compounds,
containing only H O in the channels can also be synthesized di-
rectly by chemical methods.
25
3
as well as new structures of transition metal phosphates and
vanadates. It is noted that synthesis of EMD materials has been
11-15
Lithium containing ␣-MnO
26
2
ϩ 16
(
Li MnO ; x р 0.25) have been prepared by heating H O - or
x
ϩ
2
3
carried out at temperatures up to 108°C ͑corresponding to a pressure
6,17
NH -containing
␣-MnO with LiOH at temperatures у275°C.
27,28
4
2
of 2 bars͒.
We have recently reported some preliminary results
These proton- and lithium-containing ␣-MnO2 materials have
on ␥-MnO materials with low amounts of defects, approaching the
2
shown an increased cycled capacity compared to ␣-MnO materials
2
ramsdellite limit, as well as ␣-MnO2 materials using this
2
9-31
containing large cations.
method.
Here we present further details of the range of experi-
Synthesis of ␥-MnO for commercial use ͑electrolytic manga-
ments performed, including the effect of synthesis parameters ͑tem-
perature, pH, or current density͒ on the type of structure obtained; ␣,
2
nese dioxide, EMD͒ is performed by oxidative deposition from
2
ϩ
ϩ
ϩ
Mn /H SO solutions. These samples characteristically contain a
␤, ␥ or mixtures. The consequences of the presence of Li , Na ,
2
4
ϩ
ϩ
large amount of defects and their corresponding powder X-ray dif-
fraction ͑PXRD͒ patterns consist of only a few broad peaks ͑see the
review in Ref. 18͒. As a result, the characterization and classification
of these samples is quite difficult. Pannetier18 proposed a method of
K , or NH species as additives to the MnSO4 solution will be
4
ϩ
examined, with an emphasis on Li . Also presented will be the
physicochemical characterization of these materials ͑water content,
surface area, average Mn oxidation state͒ and a study of the mor-
phology.
characterizing ␥-MnO by the amount of pyrolusite defects in the
2
Ramsdellite structure, P ͑in percent͒, and microtwinning defects,
r
Experimental
Tw ͑in percent͒. Though this method is very useful in classifying
␥
-MnO samples, in some cases, it leads to a negative value for Tw.
2
Synthesis conditions.—All syntheses were performed in a
Teflon-lined stainless steel ͑SS͒ autoclave under autogeneous pres-
sure. The autoclave was equipped with outlets for connection of the
To address this problem, our group has developed a slightly different
definition for the distribution of the microtwinning defects in the
2
working ͑Au plate, 3.5 ϫ 4 cm ͒, counter ͑2 vitreous carbon plates,
2
3
.5 ϫ 4 cm each͒, and reference electrodes to an external power
z
E-mail: Laurie.Hill@cnrs-imn.fr
source ͑MacPile, Biologic, Claix, France͒. The reference electrode

D136
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
ϩ
ϩ
For some ␣-MnO compounds, an exchange of H for K was
2
performed. This was carried out by stirring the MnO in an aqueous
2
KOH ͑0.1 M͒/KNO3 ͑0.1 M͒ solution ͑atomic ratio K:Mn Ϸ 6;
pH Ϸ 12). After 2 days, the solution was filtered and the solid
washed several times with distilled water.
Characterization of the samples.—PXRD ͑Cu K␣ radiation͒ pat-
terns were obtained using a Siemens D5000 diffractometer in the
␪-2␪ configuration for the 2␪ range 10-75° ͑step size 0.03°, step
duration 23 s͒. Lattice parameters for ␣-MnO were determined by
2
a least-squares refinement of the powder diffraction data using the
3
2
program U-fit.
All samples were studied by thermogravimetric analysis/
differential scanning calorimetry ͑TGA/DSC͒ ͑SETARAM TG-DSC
11 system͒ which was done in air with a heating rate of 2°C/min.
For samples which were synthesized in the presence of Li, atomic
absorption spectroscopy by inductively coupled plasma ͑AAS/ICP
analysis͒ was performed to determine the Li content. The average
Mn oxidation state was determined by the redox back-titration tech-
nique ͑with an error of 0.01͒ by using first an acidic iron͑II͒ sulfate
3
3
solution and then a permanganate solution. The surface area of the
samples was measured by the Brunauer, Emmett, and Teller method
ϩ
͑
BET͒ using Kr as the adsorbing gas. Determination of K content
in ion exchanged ␣-MnO was carried out by quantitative energy
2
dispersive spectroscopy ͑EDS͒ analysis using a JEOL 5800 scanning
electron microscope ͑SEM͒ operating at 20 keV. Images of the de-
posits ͑both prior to and after grinding͒ were obtained using a JEOL
JSM 6400F SEM operating at 7 kV. Prior to imaging, samples were
coated with a fine layer ͑ϳ50 Å͒ of Au/Pd film to improve conduc-
tivity.
Figure 1. Schematic drawing of the experimental setup for electrochemical-
hydrothermal syntheses describes in this work.
Structural characterization of ␥-MnO materials.—For ␥-MnO
2
2
͑
Ag/AgCl͒ was an anodized Ag wire ͑the wire was first anodized in
34
containing samples, a method developed in our group which is a
a KCl solution, rinsed and placed in a fritted glass tube filled with
the reaction solution͒. This experimental setup can be seen in Fig. 1.
MnSO ͑0.3 M͒ acidic aqueous solutions were prepared using con-
centrated H SO4 ͑95-98%͒ to achieve the desired pH. For some
experiments, Li SO ͑1.5 M͒ was added to the MnSO solution, to
probe the effect of Li on the synthesis. The initial MnSO solution
had a pH of 3.7. The solutions pH after the addition of H SO ͑listed
as percentage of total volume in parentheses͒ were approximately
1
8
modification of that proposed by Pannetier, was used to determine
the relative concentrations ͑in percent͒ of pyrolusite (P ) and Mt
r
4
defects. This method, which is useful for samples with low amounts
of microtwinning ͑р50͒ only, is briefly described here, as it has not
been previously summarized in the English literature. A series of
2
2
4
4
ϩ
4
PXRD simulations were realized over the (P ,Mt) plane using the
r
34
2
4
program DIFFAX. From these simulated patterns, a plot of the
position of the ͑110͒ diffraction line vs. the amount of pyrolusite
‘‘Ϫ0.5’’ ͑calculated value based on 20% H SO ), 0 ͑ϳ10%͒, 1
2
4
intergrowth, P , at different levels of microtwinning was made as
r
͑
ϳ0.8%͒ to 2 ͑0.1ϳ%͒. For experiments involving the addition of
shown in Fig. 2a. A plot was also made of the difference in the
positions of peaks ͑221͒ and ͑240͒ ͑indices consistent with the
Li SO , the initial MnSO /Li SO solution had a pH of 7.7, allow-
ing syntheses to be performed in the pH range Ϫ0.5 to 7.7. Simi-
larly, some syntheses were also performed by adding A SO ͑1.0 M;
A ϭ Na, K, NH ) to probe the effects of these ions on the synthe-
sis. The total volume of the solution was 100 mL ͑35-40% filling of
the autoclave͒. Syntheses were performed at temperatures of 80, 92,
1
and 24 bars, respectively͒ with a constant applied current of 2, 10,
5
2
4
4
2
4
orthorhombic cell used in Ref. 18 for Ramsdellite-MnO ) vs. the
2
2
4
amount of microtwinning defects for different levels of pyrolusite
4
intergrowth as shown in Fig. 2b. The method of determining P and
r
Mt for a sample of ␥-MnO uses the above-mentioned two relation-
2
ships between line positions and P and Mt values. From the PXRD
r
13, 130, and 180°C ͑corresponding to pressures of 1.2, 2, 3.5, 10,
pattern, the 2␪ position of the ͑110͒ reflection is determined and the
difference in 2␪ between the positions of the ͑221͒ and ͑240͒ reflec-
tions is calculated. The position of the ͑110͒ reflection is used to
2
0, 250, or 1000 mA (J Ϸ 0.072, 0.36, 1.8, 8.9, or 36 mA/cm ,
respectively͒ to give 500-600 mg MnO product per synthesis. After
2
obtain a family of (P ,Mt) couples. Likewise, the difference in 2␪
r
the synthesis, the product was thoroughly rinsed with distilled water,
dried for 1-2 h at 50-75°C, then removed from the gold electrode.
Cyclic voltammetry ͑CV͒ experiments were carried out under the
same conditions as the syntheses, using a gold wire ͑0.8 mm diam͒
as the working electrode. Three full scans were recorded in poten-
tiodynamic mode from 0.4 to 1.8 V ͑initial forward scan started at
the open circuit voltage of the solution͒ at a rate of 1 mV/s ͑steps of
between the ͑221͒ and ͑240͒ reflections gives another family of
(
P ,Mt) values. When these two families of points are plotted on a
r
graph, the intersection of the two curves corresponds to the (P ,Mt)
r
value for the sample. The underlying process is iterative in nature;
determine P from the ͑110͒ line ͑assuming no microtwinning͒, then
r
given that value of P , determine Mt from ͑221͒ and ͑240͒ lines.
r
1
0 mV every 10 s͒ using a MacPile ͑Biologic, Claix, France͒ sys-
tem.
For comparison with the ␤-MnO2 phases synthesized electro-
chemically, ␤-MnO2 was also synthesized by heating Mn͑NO3)2
Using this value of Mt, the value of Pr can be redetermined and so
on. As an example to illustrate this method, consider the compound
G92-0-0-10 whose PXRD pattern is shown in Fig. 2c. The ͑110͒
peak is located at 2␪ ϭ 22.2°, ͑221͒ at 2␪ ϭ 56.1°, and ͑240͒ at
2
␪ ϭ 57.2°. Thus, from Fig. 2a, one obtains the following possible
•
4H O in a platinum crucible under flowing O at a rate of 80°C/h
2
2
to 425°C and maintained at this temperature for 24 h. Powder X-ray
diffraction ͑PXRD͒ of the product showed ␤-MnO2 as the only
phase. The black powder was then ground and the same heat-
treatment repeated to increase the crystallinity of the product.
(Pr ,Mt) couples: ͑21,0͒, ͑28,20͒, and ͑40.5,40͒. The difference be-
tween the ͑221͒ and ͑240͒ peaks is 2␪ ϭ 1.1°, and looking at Fig.
2b, one finds the following (Pr ,Mt) couples: ͑0,17.5͒, ͑40,23͒,
͑60,29͒, and ͑80,38͒. Plotting P vs. Mt for both sets of data, the
r

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D137
Figure 2. Graphs used in the method for
the determination of Pr and Mt de-
scribed in the text. ͑a͒ Graph showing
the variation in the 2␪ position of the
(
110)R peak in ␥-MnO as a function of
2
the amount of pyrolusite intergrowth,
P , for different amounts of microtwin-
r
ning defects, Mt. ͑Values determined
from simulated powder X-ray patterns
3
4
obtained with the program DIFFAX. ͒
b͒ Graph showing the variation in the
␪ difference between the (021)R and
240)R peaks in ␥-MnO2 as a function
͑
2
(
of the amount of microtwinning defects,
Mt, present for different amounts of py-
rolusite intergrowth, P . ͑Values deter-
r
mined from simulated powder X-ray pat-
terns obtained with the program
34
DIFFAX. ͒ ͑c͒ X-ray powder pattern for
sample G92-0-0-10 ͓␥-MnO2 ͑28,21͔͒.
͑
d͒ Graph showing the two curves ob-
tained from the (P ,Mt) couples found
r
from the graphs in ͑a͒ and ͑b͒ for the
X-ray powder pattern in ͑c͒. The inter-
section of the two curves gives the cor-
responding (P ,Mt) values for the
r
sample.
ϩ
graph shown in Fig. 2d is obtained and the value (P ϭ 28, Mt
rameters such as pH, temperature, presence of Li ͑or not͒, and
thickness of the deposit on the electrochemical measurements during
the synthesis.
r
ϭ 21) is deduced from the intersection. Comparison of the experi-
mental diffraction patterns with those simulated with similar struc-
tural parameters ͑determined using the above method͒ shows a
rather good agreement. It must be noted that the above method
cannot be used for high microtwinning levels ͑у50͒, i.e., samples
exhibiting only one broad peak in the vicinity of 2␪ ϭ 56°Ϫ57°,
and the method proposed by Pannetier18 must be used in such cases.
Figure 3a and b show the differences in the current density vs.
voltage curves for different solution pH and temperatures, respec-
tively. Figure 3a compares the current density vs. voltage curves for
solutions at 92°C in the absence of Li and two different pH values,
pH 0 and pH 1. The voltammograms show an initial increase in
current density which occurs at ϳ0.8 V for pH 0 and ϳ0.7 V for pH
, corresponding to the deposition of MnO according to Reaction 1.
ϩ
Results and Discussion
1
2
The half-wave potentials are close to the thermodynamic values cal-
culated from the Nernst law ͑see Table I͒, which indicates that the
overpotentials corresponding to Reaction 1 are low. At higher volt-
ages, ϳ1.6 V for pH 0 and Ͼ1.7 V for pH 1, a second increase in
current density occurs, which is much more drastic than the first
phenomenon ͑only the very beginning of this second phenomenon is
seen for pH 1͒, and corresponds to the oxidation of water according
to Reaction 2
Synthesis.—The formation of MnO at the electrode occurs ac-
cording to Reaction 1
2
Mn2ϩ͑aq͒ ϩ 2H O → MnO ͑s͒ ϩ 4Hϩ ϩ 2e
Ϫ
͓1͔
2
2
Syntheses were performed by varying the following four param-
ϩ
eters: bath temperature ͑T͒, pH, presence of A ͑or not; A ϭ Li,
Na, K, NH ), and applied current density. The alkaline or ammo-
4
ϩ
Ϫ
nium cations were added to the solution for a possible template
2H O → O ͑g͒ ϩ 4H ϩ 2e
͓2͔
2
2
effect ͑to stabilize the formation of the larger tunnels of ramsdellite͒
ϩ
or, in the case of Li , to possibly synthesize ␥-MnO samples con-
2
The voltages corresponding to water oxidation are much greater
than the calculated thermodynamic values ͑Table I͒, indicating a
large overpotential for this reaction. The dependence on pH does not
seem to follow the Nernst law ͑Ϫ0.145 pH at 92°C͒ at first glance.
However, since water oxidation occurs at the surface of deposited
ϩ
35-38
taining Li ͑e.g., ␥-Li0.33MnO ).
2
Some of the samples were annealed at 250°C for 16 h in air in
order to lower the water content with minimal structural change
before lithium battery studies.
MnO , it is possible that the different MnO2 samples deposited
2
Nomenclature of the synthesized samples.—Samples are denoted
could have a different catalytic effect on the oxygen evolution reac-
tion and their differences in electronic conductivity might affect the
rate of the current increase. The reverse scans show a negative cur-
rent density below ϳ0.8 or 0.7 V, depending on pH ͑0 or 1, respec-
tively͒, corresponding to the reverse of Reaction 1, i.e., the reduc-
tion and dissolution of MnO2 . Unexpectedly, at lower pH, the
maximum current corresponding to the deposition of MnO2 is
higher during the forward sweep, but on the reverse sweep the cur-
rent density is lower.
by a G, indicating galvanostatic synthesis mode, followed by the
quadruplet of data ͓T-pH-A͑cation added, 1 ϭ Li , 0 ϭ no)-I].
ϩ
For example, G113-0-1-10 indicates a material synthesized under
galvanostatic conditions at 113°C from a solution with a pH of 0 in
ϩ
the presence of Li and an applied current of 10 mA ͑corresponding
2
to a current density of 0.36 mA/cm ͒. In this notation, an ‘n’ ͑mean-
ing ‘‘negative’’͒ refers to solutions with a calculated pH of Ϫ0.5 ͑20
Vol % H SO ).
2
4
Voltammetric studies of the electrochemical synthesis.—In this
section, we will evaluate the influence of various experimental pa-
Figure 3b shows the voltammograms for solutions of pH 1, with-
out Li , at two different temperatures, 92 and 113°C. One can see
ϩ

D138
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
Table I. Redox potentials of the Mn2¿ÕMnO2 and H OÕO
2
2
couples under some conditions used in this study, calculated using
the Nernst equation and expressed vs. the AgÕAgCl reference elec-
0
trode †E„AgÕAgCl… ϭ 0.30 V vs. NHE‡. The normal potentials E
t
at the temperature
t
„in °C… were calculated from the
51
0
0
25
0
25
relationship: Et „in mV… ϭ E „in mV…À„t Ϫ 25…, where E
is the normal potential at 25°C taken from Ref. 52.
Temperature
E
Redox system
Mn2 /MnO2
͑°C͒
pH
͑V vs. Ag/AgCl͒
92
0
1
1
0
1
1
0.880
0.735
0.707
0.863
0.791
0.766
ϩ
92
113
92
H O/O2
92
2
113
current density observed on the reverse sweep is similar to that
ϩ
observed for the solution without Li .
Figure 5 shows three consecutive oxidation sweeps for a solution
ϩ
2ϩ
at 113°C, pH 1 in the absence of Li . The onset of Mn oxidation
occurs at lower voltage on the second and third sweeps, showing the
difference between deposition on a naked Au surface and a MnO2
surface. Also noticeable is the increase in current density on each
subsequent sweep, which can be explained by the increase in the
surface area as the deposit grows ͑not all MnO was reduced on the
2
reverse sweeps͒.
Figure 6 shows the evolution of the voltage as a function of time
2
during galvanostatic deposition at 0.36 mA/cm under various con-
ditions. These results parallel the findings from CV experiments. A
lower deposition voltage is observed with an increase in the tem-
perature or pH. Deposition of MnO2 occurs at approximately the
ϩ
same voltage in experiments with and without Li present. It is
noted that experiments performed under the same conditions show a
variation of up to 0.05 V in the stable deposition voltage. During the
synthesis, the voltage initially increases during the first minute be-
fore decreasing and settling to a nearly constant value in the first few
hours. The voltage stability during the synthesis is important for the
formation of uniform deposits with the same composition and struc-
ture throughout. XRD studies of both sides of the deposit ͑facing the
solution and facing the electrode͒ have revealed that while slight
Figure 3. Cyclic voltammograms performed using a gold wire working
differences were observed ͓variations in P ͑Ϯ5%͒, for example͔,
ϩ
r
electrode ͑a͒ for solutions at 92°C in the absence of Li and a pH of 0 or 1.
ϩ
͑
b͒ Cyclic voltammograms for solutions at pHϭ1 in the absence of Li at a
temperature of 92°C or 113°C. In both voltammograms, the arrows indicate
where deposition of MnO2 takes place during an electrochemical-
hydrothermal synthesis under the same conditions.
that the general shape of the curves is the same ͑more so than for
differing pH͒. When the temperature is increased, the half wave
potential of the deposition of MnO solid follows a simple variation
2
of the potential with temperature ͑see Table I͒. The current increase
corresponding to the oxidation of water also occurs at a slightly
lower voltage at higher temperatures, due to a decrease of the over-
potential with increasing temperature. The pseudo-plateau in the
oxidation current corresponding to the deposition of MnO is also
2
higher, as expected, due to lower kinetic limitation for the formation
of the MnO deposit.
2
Figure 4 shows the voltammograms of solutions with and with-
ϩ
out Li at pH 1 and 92°C. Here, the start of MnO deposition occurs
2
at roughly the same voltage ͑ϳ0.7 V͒ in both cases. The onset of the
oxidation of water occurs at a lower voltage ͑ϳ1.5 V͒ in the pres-
ϩ
ence of Li , which may be related to the same effects as proposed
above ͑catalytic effect and/or electrical conductivity͒. Surprisingly,
when Li is present in the solution, a lower current density is asso-
Figure 4. Cyclic voltammograms for solutions at 92°C and pH 1 with or
ϩ
ϩ
without Li present. The arrow indicates where deposition of MnO occurs
2
ciated with the deposition of MnO on the forward sweep, but the
during an electrochemical-hydrothermal synthesis under the same conditions.
2

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D139
Figure 5. Three consecutive CV oxidation sweeps for a solution at 113°C,
ϩ
pH 1 in the absence of Li .
the basic structure was maintained ͑i.e., ␣, ␤, ␥ or mixtures͒. For all
2
syntheses except those performed at J ϭ 36 mA/cm ͑G92-0-0-
1000͒, the measured voltage ͑generally less than 1 V vs. Ag/AgCl͒
was always lower than the voltage corresponding to the oxidation of
water, indicating all the current was consumed in the formation of
MnO . This is consistent with the measured reaction yield ͑faradic͒
2
near 100% for solutions with pH у 0 and an applied current density
2
between 0.36 and 8.9 mA/cm . However for more acidic solutions
͑
G92-n-0-10, G92-n-1-10, G113-n-0-10, and G113-n-1-10͒, the
measured reaction yield was significantly lower, only about 50-75%.
These solutions exhibited a deep rose color after the synthesis, in-
Ϫ
Figure 7. PXRD patterns representing some examples of the different struc-
ture types obtained by the electrochemical-hydrothermal method. The nota-
tion ␥/␤ or ␣/␥ indicates a macroscopic mixture of both phases, while ␣•␥
indicates a crystallographically interconnected mixture. Sample names rep-
dicating the formation of MnO4 during the synthesis, in agreement
39
with Pourbaix potential-pH diagrams, and thus explaining the
lower yield. In the case of G92-0-0-1000, the measured voltage was
2
ϩ
ϩ
у1.5 V vs. Ag/AgCl, indicating simultaneous oxidation of Mn
and water. The measured yield was thus significantly less than
00%. In the case of G92-0-0-2, the reaction yield was also signifi-
resent Synthesis mode ͑G͒ Temperature-pH-presence ͑1͒ or not ͑0͒ of Li
2
during synthesis-applied current ͑for 28 cm ͒. An ‘n’ indicates a calculated
1
pH of Ϫ0.5.
cantly below 100% ͑ϳ50%͒. Here, the solution did not appear to
Ϫ
contain an appreciable amount of MnO4 and the measured voltage
was below that corresponding to the oxidation of water. Given that
the length of the synthesis was ϳ10 days ͑compared to 2 days or
less͒, the possibility of other side reactions occurring was explored.
A search of the literature indicated that ␥-MnO can undergo the
2
2
ϩ40
following reaction in hot H SO in the presence of Mn
2
4
-MnO ͑s͒ ϩ Mn ͑aq͒ ϩ H → MnOOH͑s) ϩ Mn3ϩ͑aq͒
2
ϩ
ϩ
␥
2
͓
3͔
Thus it follows that if the above reaction took place during the
synthesis, a part of the oxidizing current could have been used to
3
ϩ
reoxidize the MnOOH to MnO ͑assuming Mn diffused away into
2
the solution͒, leading to a faradic yield less than 100%.
PXRD and analysis.—Materials obtained from MnSO4 and
MnSO /Li SO solutions.—PXRD patterns of representative materi-
4
2
4
als of each structure type synthesized here are shown in Fig. 7.
Results of all syntheses are summarized in Tables II and III. Results
of syntheses performed with an applied current density of 0.36
2
mA/cm are summarized in Fig. 8a and b in the form of ͑T, pH͒
phase diagrams for MnSO and MnSO /Li SO solutions, respec-
4
4
2
4
Figure 6. Evolution of the voltage as a function of time during galvanostatic
synthesis at 0.36 mA/cm under various conditions.
tively. The basic trends illustrated in Fig. 8a and b can be described
as follows: more acidic conditions and lower temperatures ͑and thus
2

D140
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
¿
Table II. Characterization of the samples obtained in the absence of Li . Sample names represent Synthesis mode „G… Temperature-pH-
¿
2
absence of Li „0… during synthesis-applied current „for 28 cm …. An ‘n’ indicates a calculated pH of À0.5.
Structural
parameters
Lattice
parameters
for ␣-MnO2
͑I centered
tetragonal
cell͒
͓
P (%),Mt(%)͔
r
for ␥-MnO2
͑Ϯ/-4%͒
before/after
annealing
BET
surface
area
Manganese
oxidation
state
Composition
͑derived from
TGA
Structure
from X-ray
diffraction
2
Sample
͑m /g͒
͑Ϯ0.01͒
measurements͒
G80-0-0-10
␥-MnO ϩ
2
a ϭ 9.833(7) Å
c ϭ 2.854(1) Å
23.7
ϩ3.82
MnO1.91•0.18H O
␣
-MnO2
2
b
(
␣•␥-MnO2)
G92-n-0-10
␥-MnO ϩ
a ϭ 9.821(6) Å
c ϭ 2.849(2) Å
14.2
ϩ3.89
MnO1.945•0.19H O
2
2
␣
-MnO2
G92-0-0-2
G92-0-0-10
G92-0-0-50
␥-MnO2
␥-MnO2
␣-MnO2
͑47,21͒/n.d.a
͑28,21͒/͑26,11͒
8.1
n.d.
28.5
ϩ3.96
ϩ3.99
ϩ3.89
n.d.
MnO1.995•0.10H O
2
a ϭ 9.825(4) Å
c ϭ 2.852(1) Å
MnO1.945•0.28H O
2
G92-0-0-250
G92-0-0-
␥-MnO2
␥-MnO2
͑ϳ40,у50͒/n.d.
͑ϳ40,у50͒/n.d.
56.2
139
ϩ3.87
ϩ3.84
MnO1.935•0.15H O
2
MnO1.92•0.24H O
2
1000
G92-1-0-10
␥-MnO ϩ
trace ␤-
͑49,25͒/͑49,17͒
13.0
ϩ3.92
MnO1.96•0.15H O
2
2
MnO2
G92-2-0-10
G113-n-0-10
G113-0-0-10
G113-1-0-10
␥-MnO2
␥-MnO2
␥-MnO2
͑47,26͒/͑51,19͒
͑42,21͒/͑41,12͒
͑48,23͒/͑45,14͒
͑62,15͒/n.d.
5.6
7.4
7.9
n.d.
n.d.
n.d.
ϩ3.96
ϩ3.96
ϩ3.94
MnO1.98•0.08H O
2
␥-MnO ϩ
n.d.
MnO1.97•0.06H O
2
2
␤
-MnO2
G113-2-0-10
␤-MnO ϩ
3.5
n.d.
n.d.
2
trace ␥-
MnO2
G130-0-0-10
G130-1-0-10
G180-1-0-10
␥-MnO2
␤-MnO2
␤-MnO2
͑67,15͒/n.d.
͑100,0͒/n.d.
͑100,0͒/-
3.9
n.d.
n.d.
n.d.
ϩ4.0
n.d.
MnO •0.11H O
2
2
n.d.
n.d.
a
n.d. indicates not determined.
As determined by electron diffraction.
b
lower pressures͒ favor the formation of materials with more open
structures, from ␤-MnO to ␥-MnO ͑ramsdellite͒ to ␣-MnO . Be-
low, these trends are described in more detail.
tures were always present. Looking at the diffraction patterns allows
a comparison of the relative amounts of ␤-MnO present in each of
2
2
2
2
ϩ
the samples. At pH 1, the presence of Li in the solution leads to a
In both Fig. 8a and b, the formation of pure ␤-MnO is favored
slightly higher proportion of pyrolusite in the sample. Increasing the
2
ϩ
at higher temperatures ͑у130°C, P у 10 bars͒ and pH ͑у1͒. It is
noted that ␤-MnO2 has not been previously synthesized electro-
pH to 2 in the absence of Li leads to a product which is mostly
pyrolusite, with a very small amount of ␥-MnO visible by PXRD.
2
ϩ
chemically. The formation of ␤-MnO at elevated temperatures is
2
However, in the presence of Li at a pH 7.7 ͑i.e., no added H SO ),
2
4
perhaps to be expected; the higher temperature and resulting higher
pressure stabilize the more compact pyrolusite structure. The
the proportion of ␥-MnO present in the sample is increased. In this
2
ϩ
case, it would appear that Li does act as a template or to stabilize
␤
-MnO synthesized here is pure and well crystallized, as can be
2
the more open ramsdellite structure, to an extent.
Increasing the acidity, or further lowering the temperature, leads
to a wide region where the formation of compounds with the
seen in Fig. 7. Electron microscopy studies are in progress to check
for the absence of defects. The PXRD patterns of the ␤-MnO com-
2
pounds synthesized here indicate lattice parameters that are not sig-
␥
-MnO structure is favored. For these MnO samples, the relative
2
2
nificantly different from those of the ␤-MnO synthesized by a solid
2
concentrations of pyrolusite (P ) and Mt defects were determined
r
state method or the values given in the JCPDS files ͑81-2261,72-
using the method described above. The corresponding structural pa-
1984,24-0735͒. In general, the compounds here do not show signifi-
rameters P and Mt are listed in Tables II and III and also noted on
r
cant differences compared to ␤-MnO2 synthesized by solid-state
methods.
On lowering the temperature ͑to 113°C, P Ϸ 3.5 bars͒ at the
more elevated pH values ͑у1͒, mixtures of ␥- and ␤-MnO2 are
Fig. 8a and b. The ␥-MnO PXRD patterns shown in Fig. 7 show
2
the effects of pyrolusite intergrowth. As P increases, the ͑110͒ re-
r
flection ͑peak at 2␪ ϳ 22°) shifts to higher angles ͓accompanied
by a shift to lower angle of the ͑130͒ reflection ͑peak at 2␪
ϳ 35°)], as described in Ref. 18. Also evident is a slight increase
in the distance ͑in degrees͒ between the ͑221͒ and ͑240͒ reflections
obtained. The well-defined peaks of ␤-MnO visible in the PXRD
2
patterns suggest that the mixtures are macroscopic in nature, though
the deposits appeared to be homogeneous both by eye and by in-
spection under the microscope ͑optical or scanning electron͒. XRD
patterns of both sides of the deposit revealed that while the relative
intensities of the ␤-MnO peaks with respect to the ␥-MnO peaks
(
2␪ ϳ 56 and ϳ57°, respectively͒.
To this point, the two T/pH phase diagrams in Fig. 8a and b are
ϩ
virtually the same, indicating little influence by the presence of Li
2
2
were slightly lower near the electrode than at the surface, both struc-
in the solution. However, for T р 92°C (P р 2 bars͒ or pH р 0,

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D141
¿
Table III. Characterization of the samples obtained in the presence of Li „3 M…. Sample names represent Synthesis mode „G… Temperature-
¿
2
pH-presence „1… of Li during synthesis-applied current „for 28 cm …. An ‘n’ indicates a calculated pH of À0.5.
Structural
parameters
͓
P (%),
Lattice
parameters
for ␣-MnO2
͑I centered
tetragonal
cell͒
r
Mt(%)] for
-MnO2
Ϯ/-4%͒
␥
BET
surface
Area
Manganese
oxidation
state
͑
Composition
͑derived from TGA
measurements͒
Structure
from XRD
before/after
annealing
2
Sample
͑m /g͒
͑Ϯ0.01͒
G80-0-1-10
␣-MnO2
␣-MnO2
␥-MnO2
a ϭ 9.842(4) Å
c ϭ 2.856(1) Å
a ϭ 9.809(1) Å
c ϭ 2.854(1) Å
23.4
5.9
ϩ3.77
ϩ3.87
Li0.03MnO1.90•0.18H O
2
G92-n-1-10
Li0.05MnO1.96•0.15H O
2
G92-0-1-10
G92-1-1-10
͑36,21͒/
n.d.a
9.0
ϩ3.97
ϩ3.96
MnO1.985•0.13H O
2
͑32,10͒
␥-MnO ϩ
͑46,23͒/
͑49,16͒
MnO1.98•0.09H O
2
2
trace ␤-
MnO2
G92-7.7-1-10
G113-n-1-10
G113-0-1-10
G113-1-1-10
␥-MnO2
͑41,20͒/
7.9
4.8
9.6
6.3
ϩ3.91
3.97
MnO1.955•0.08H O
2
͑
45,14͒
͑52,22͒/
60,17͒
͑63,23͒/
63,15͒
␥-MnO2
␥-MnO2
n.d.
͑
ϩ3.97
ϩ3.98
n.d.
͑
␥-MnO ϩ
͑65,7͒/n.d.
MnO1.99•0.04H O
2
2
␤
-MnO2
G113-7.7-1-10
␥-MnO ϩ
͑53,17͒/n.d.
9.7
ϩ3.89
n.d.
2
␤
-MnO2
G130-1-1-10
G180-1-1-10
␤-MnO2
␤-MnO2
͑100,0͒/n.d.
͑100,0͒/n.d.
n.d.
n.d.
ϩ4.0
ϩ4.0
n.d.
n.d.
a
n.d. indicates not determined.
ϩ
different results were obtained as a function of the presence of Li .
prepared at 92°C and pH Ͻ 0 presents a PXRD pattern indicative of
a macroscopic mixture of the two compounds ͑well defined, narrow
lines from both compounds͒. Electron diffraction using a transmis-
sion electron microscope ͑TEM͒ is consistent with this interpreta-
In a solution containing only MnSO , lowering the temperature to
4
8
0°C or increasing the acidity (pH Ͻ 0) led to samples which dis-
played the diffraction lines of both ␥- and ␣-MnO . The sample
2
Figure 8. ͑a͒ Temperature/pH phase diagram obtained from acidic MnSO solutions. ͑b͒ Temperature/pH phase diagram obtained from acidic MnSO /Li SO
4
4
4
2
ϩ
solutions. ͑c͒ Diagram showing the variation of the structure obtained as a function of current density for solutions at 92°C, pH 0 in the absence of Li . For
-MnO containing compounds, the corresponding (P ,Mt) structural parameters are noted.
␥
2
r

D142
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
Figure 11. Effect of heat treatment ͑250°C in air͒ on the structural param-
eters P and Mt of ␥-MnO containing compounds. The positions of classi-
cal HTMDs and unusual HTMDs are from Ref. 18 and 45, respectively.
Figure 9. Placement in the (P ,Mt) plane of ␥-MnO containing samples
r
2
r
2
obtained in this study. For reference, regions containing manganese dioxides
prepared by other methods have been added. The positions of classical
EMDs, and classical chemical manganese dioxides ͑CMDs͒ are from Ref.
18. Other EMDs were obtained at very low current densities ͑see Ref. 18 and
4
4͒ The position of Unusual EMDs is taken from reference 45. ᭹
lationship with the c axis in common.30 To reflect this structural
interconnectivity, and to differentiate this phase from the previously
ϭ ␥-MnO , ᭿ ϭ ␥-MnO ϩ ␤-MnO , ࡗ ϭ ␥-MnO ϩ ␣-MnO
2
2
2
2
2
described macroscopic mixture (␣/␥-MnO ), this compound is re-
2
ϩ
ferred to as ␣•␥-MnO . On the contrary, when Li is present in the
2
tion, the crystals displaying diffraction spots corresponding to either
- or ␣-MnO . This compound is thus referred to as ␣/␥-MnO .
However, at 80°C and pH 0, a material was obtained which exhib-
solution, under the same reaction conditions, pure ␣-MnO phases
2
␥
2
2
are obtained. At 92°C, pH Ͻ 0, ␣-MnO appears to be very well
2
crystallized by PXRD diffraction. However, at 80°C and pH 0, the
diffraction peaks are quite wide, in part due to a decrease in crys-
ited an PXRD pattern containing broad lines of both ␣- and
4
2
␥
-MnO , similar to that found in a compound previously reported
tallite size.
2
to contain some crystals exhibiting a ‘‘crystallographic interconnec-
The lattice parameters of the ␣-MnO compounds synthesized
2
4
1
tivity’’ ͑i.e., a coherent interface͒ between the ␣- and ␥- structures.
here are listed in Tables II and III. They are close to those found for
Further electron diffraction studies ͑TEM͒ of the sample prepared
here showed that most of the crystallites exhibited this type of crys-
tallographic interconnectivity.30 Examination of such crystallites
ϩ
ϩ
Li or H containing ␣-MnO2 compounds synthesized by other
11,13,14,16
methods.
The a lattice parameters for G92-n-1-10
(
␣-MnO , well crystallized͒ are slightly lower than for the other
2
͑
over regions measuring several tens of nanometers across͒ led to
␣
-MnO compounds synthesized in this study and the fit is better,
2
the simultaneous observation of diffraction spots corresponding to
both the ␣ and ␥͑Ramsdellite͒ structures, having an orientation re-
likely in relation to the increased crystallinity and sharp, well de-
fined diffraction peaks. After annealing at 250 or 300°C for 16 h, the
PXRD patterns ͑not shown͒ show pure ␣-MnO with no obvious
2
changes in degree of crystallinity. The PXRD patterns indicate
slightly lower a parameters ͑decrease of about 0.05 Å͒ and very
slightly higher c parameters ͑about 0.01 Å͒, consistent with the lit-
erature. The lattice parameters given for ␣-components of the ␣/␥-
and ␣•␥-MnO2 compounds ͑G92-n-0-10 and G80-0-0-10, respec-
Table IV. Characterization of the samples obtained in the pres-
¿
ence of cations other than Li . Sample names represent Synthesis
mode „G… Temperature-pH-A „A SO „1 M… added to solution…-
2
4
2
applied current „for 28 cm ….
Structural
parameters
͓
P (%),Mt(%)͔
r
for ␥-MnO2
Structure
͑Ϯ/-4%͒ before
Sample
from XRD
annealing
G92-0-Na-10
G92-1-Na-10
G92-7-Na-10
G113-0-Na-10
G113-1-Na-10
G113-1-K-10
␥-MnO2
␥-MnO2
␥-MnO2
␥-MnO2
␥-MnO2
␣-MnO2
␣-MnO2
͑30,20͒
͑35,22͒
͑32,18͒
͑51,24͒
͑42,16͒
Figure 10. Graph showing the variation of the amount of pyrolusite inter-
growth, P , as a function of solution pH at different temperatures and dif-
r
ferent solution compositions (MnSO or MnSO /Li SO ).
G92-1-NH -10
4
4
2
4
4

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D143
Figure 12. SEM photographs of
␣
-MnO2 samples: G92-n-1-10 ͑a͒ sur-
face of the deposit showing large pillar-
like crystals and ͑b͒ a cross-sectional
view of the deposit; G80-0-1-10 ͑c͒ sur-
face of the deposit showing star-like for-
mations of small fibers composed of
smaller needles, and ͑d͒ a cross-sectional
view of the deposit. The arrows in ͑b͒
and ͑d͒ indicate the direction of growth
during deposition; i.e., the surfaces of
the deposits are located to the right and
lower left of the two figures, respec-
tively.
tively͒ were calculated using only those peaks which did not overlap
bulk solution. After the initial decrease, the pH at the electrode
surface is expected to remain fairly constant at a steady state equi-
librium value, due to diffusion of species into the bulk solution.
with those of ␥-MnO , which lead to less precise fits.
2
In addition to the study of the effect of temperature, pH, and
ϩ
ϩ
presence of Li in the solution on the structure of MnO obtained, a
Thus as the current density is increased, the rate of production of H
2
ϩ
study was carried out to determine the effect of changing the current
density. G92-0-0-10 was used as a starting point as this material
ions could increase. Assuming the same rate of diffusion of H ions
away from the electrode, the local pH would decrease following an
increase in current density. Extrapolation of this reasoning would
displayed the lowest P and Mt values of the samples prepared thus
r
far. An experiment was performed at a current density of 0.072
indicate that ␥ structures would be obtained at 92°C and pH
2
mA/cm ͑G92-0-0-2͒ in the hopes of preparing a ␥-MnO com-
2
2
Ͻ
Ϫ 0.4 at a constant current density of 0.36 mA/cm , if such an
pound containing even fewer microtwinning defects ͑following the
trend of other EMDs shown in Fig. 9.18 See the following discus-
sion͒ and perhaps fewer pyrolusite defects as well. However, the
experiment were possible.
Analysis of ␥-MnO compounds.—Figure 10 summarizes the varia-
2
␥
-MnO ͑47,21͒ obtained at this current density contained approxi-
tion of the parameter P for each material as a function of the solu-
2
r
mately the same concentration of microtwinning defects (Mt
ϭ 21) and an increase in the amount of pyrolusite content (Pr
ϭ 47).
tion pH. As was mentioned earlier, it was expected that the addition
of Li to the solution might help to stabilize the Ramsdellite struc-
ϩ
ture, leading to lower values of P . At low temperature ͑92°C͒,
r
ϩ
Following this result, an experiment was carried out at the higher
there appears to be little or no influence of Li on the P values of
r
2
current density of 1.8 mA/cm . Under these conditions ͑G92-0-0-
the obtained materials. The results are slightly different at a tem-
ϩ
5
0͒, ␣-MnO was obtained with no indication of the presence of
2
perature of 113°C and pH р 1, where the presence of Li appears
␥
-MnO . The PXRD pattern appeared similar to that of G80-0-1-10
to favor higher amounts of pyrolusite content. However, at a
2
(
␣-MnO ), and indicated nearly the same lattice parameters. Fur-
pH Ͼ 1 and 113°C, the effect seems to be reversed, the presence of
Li leading to smaller amounts of pyrolusite. In the absence of Li
2
2
ther increasing the current density to 8.9 or 35.7 mA/cm ͑G92-0-0-
2
compounds. These compounds exhibited PXRD patterns similar to
traditional EMD samples, showing ϳ5 broad peaks in the 2␪ range
ϩ
ϩ
50 or G92-0-0-1000, respectively͒ leads once again to ␥-MnO2
at pH 2, the major phase obtained is ␤-MnO with only a small
2
amount of ␥-MnO present. In this case, the calculation of the struc-
2
tural parameters P and Mt is quite difficult due to the overlap of
r
1
0-75°. Analysis of these patterns indicated a pyrolusite content (Pr)
some of the peaks with the very intense peaks of ␤-MnO . How-
2
near 40 and high degree of microtwinning (Mt
у 50). The results of the study of the effects of current density are
summarized in Fig. 8c.
a
ϩ
ever, the presence of Li in the solution at pH 7.7 ͑i.e., no H SO4
2
added to the solution͒ leads to a mixture of ␥- and ␤-MnO , the
2
␥
-MnO2 exhibiting a lower level of pyrolusite intergrowth than
The trend observed for current densities between 0.072 and 8.9
2
found in the mixtures obtained at lower pH ͑ϭ1͒. Overall, the gen-
eral trend shows a decrease in the amount of pyrolusite content (Pr)
mA/cm , i.e., an initial decrease of P within the ␥ structure, fol-
r
lowed by a change to the ␣ structure, mirrors that observed for a
decrease in the pH of the bulk solution ͑see the data points obtained
at 92°C and different pH values in Fig. 8a͒. The structural variations
with current density can be attributed to a change in the local pH at
the electrode as the current is increased.43 Recall that H ions are
produced during the electrolysis according to Reaction 1, which will
lead to a decrease in the pH at the electrode surface compared to the
as the temperature or the pH ͑for pH Ͻ 2) is decreased, indepen-
ϩ
dent of the presence of Li . The ␥-MnO materials containing the
2
lowest rates of pyrolusite intergrowth were found at a temperature of
ϩ
92°C and pH 0.
Figure 9 shows the position in the (Pr ,Mt) plane of the ␥-MnO2
2
samples prepared at J ϭ 0.36 mA/cm in this work, compared to

D144
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
Figure 13. SEM photographs of
samples containing a mixture of ␣- and
␥
-phases: G92-n-0-10 (␣/␥-MnO ) ͑a͒
2
surface of the deposit showing rod-like
fibers, and ͑b͒ a cross-sectional view of
the deposit; G80-0-0-10 (␣•␥-MnO ) ͑c͒
2
surface of the deposit showing
pyramidal-type structures composed of
smaller needles, and ͑d͒ a cross-sectional
view of the deposit. The arrows in ͑b͒
and ͑d͒ indicate the direction of growth
during deposition; i.e., the surfaces of
the deposits are located to the upper
right and to the left of the two figures,
respectively.
previously known ␥-MnO compounds prepared electrochemically
changes in the amount of pyrolusite intergrowth, P , as well as a
2
r
͑
͑
EMDs͒ or by chemical methods ͓͑chemical manganese dioxides
decrease of about 10% in the degree of microtwinning, Mt. Addi-
tionally, for a number of samples, annealing leads to a decrease of
the microtwinning rate with an almost constant or even slightly de-
creasing P value, a property never reported for ␥-MnO materials
CMDs͔͒, taking into account the previous work1 ͑which uses Tw
8
as the microtwinning parameter͒, and assuming the relationship
Tw Ϸ 2Mt. As shown in the figure, classical EMD samples are
r
2
characterized by P values around 50% and high levels of micro-
twinning (Mt у 40). The amount of microtwinning in EMDs can
to this point. Even extending the time of annealing to 100 h at 250°C
filled triangles in the figure͒ does not lead to any significant in-
r
18
͑
be decreased with little change in the amount of pyrolusite inter-
crease in Pr . It is not clear at this point what the origin of this
unique behavior is, but work is in progress to explore the possibili-
ties of different kinds or distributions of defects in these materials
which could possibly be responsible.
growth by using a very low current density during the deposition
8,44
͑
‘‘other EMDs’’ in the figure͒.1
Previous work in our group
yielded a series of compounds ͑labeled as ‘unusual’ EMDs in the
figure͒ which were characterized by a slightly lower and more nar-
row range of P values as well as lower rates of microtwinning
Materials obtained from MnSO4 /A2SO4 solutions (A ϭ Na, K,
NH4).—The results of the syntheses carried out in the presence of
r
45
(
30 р Mt р 50). Classical CMDs are usually characterized by
ϩ
ϩ
ϩ
ϩ
lower rates of pyrolusite intergrowth and microtwinning defects than
Na , K , and NH ions are listed in Table IV. In the case of Na ,
4
found in EMD samples.18 The figure shows that the ␥-MnO con-
the results are similar to those obtained from MnSO4 and
2
taining samples synthesized here by the hydrothermal-
electrochemical method are characterized by low microtwinning
rates ͑ϳ20͒, significantly lower than what is normally characteristic
MnSO /Li SO solutions, but the P values of the ␥-MnO com-
4
2
4
r
2
pounds are lower. Notably, at pH 1 and a temperature of 113°C, the
ϩ
presence of Na in the solution suppressed the formation of
3
0
of EMDs. Also, ␥-MnO samples can be prepared over a wide
2
␤-MnO2 .
range of pyrolusite intergrowth rates ͑ϳ30-65͒, including low rates
usually observed only in CMDs.
ϩ
ϩ
4
In the case of K or NH , ␣-MnO was obtained at pH 1 and a
2
temperature of 113 or 92°C, respectively, regions where ␥ ͑92°C͒ or
Figure 11 shows the effects of annealing the ␥-MnO containing
2
␥/␤ ͑113°C͒ compounds were formed in MnSO or MnSO /Li SO
4
4
ϩ
2
4
samples synthesized here at 250°C for 16 ͑open circles͒ or 100 h
ϩ
solutions. The formation of ␣-MnO in the presence of K or NH4
2
͑
open or filled triangles͒. Annealing of ␥-MnO samples is routinely
2
ions is not surprising, as it is known these ions stabilize the
performed in order to decrease the amount of water contained in the
structure prior to electrochemical lithium insertion studies. These
samples are referred to as heat-treated manganese dioxides
2-8
2
ϫ 2 channels of this structure.
Morphology.—A detailed study of the morphology of the MnO2
͑
HTMDs͒. In addition to a decrease in the amount of water con-
compounds prepared by the hydrothermal-electrochemical method
4
2
tained in the structure, there is an accompanying change in structural
parameters, usually a large decrease in microtwinning, Mt, and an
has been published elsewhere. A brief summary of this study is
presented here.
increase in the amount of pyrolusite intergrowth, P , toward the
Of the ␣-MnO materials synthesized, two main surface mor-
r
2
pyrolusite limit. This results in materials occupying the region la-
beled as classical HTMD’s in Fig. 11.18 The so-called unusual HT-
MDs are the result of annealing unusual EMDs, and exhibit a lower
phologies were observed, large ͑ϳ1 ␮m diam͒ pillar-like crystals
͑G92-n-1-10͒, or star-like formations of small ͑15-25 nm in diam͒
needle-like fibers ͑G80-0-1-10, G92-0-0-50͒. These morphologies
are shown in Fig. 12a and c, respectively. The formation of the large
crystals seen in sample G92-n-1-10 is likely not only a result of the
temperature and acidity of the solution, but also due to a slower
range of P values, and a larger range of Mt values than classical
r
HTMDs.45 The samples prepared in this work represented in Fig. 11,
show that the annealing carried out here results mostly in small

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D145
Figure 14. SEM photographs of
␥
-MnO samples showing the effects of
2
changing the various synthesis param-
ϩ
eters: temperature, pH, presence of Li ,
and current density ͑J͒. ͑a͒ G92-0-0-10
͓
͓
͓
͓
͓
␥-MnO2 ͑28,21͔͒, ͑b͒ G113-0-0-10
␥-MnO2 ͑48,23͔͒, ͑c͒ G92-1-0-10
␥-MnO2 ͑49,25͔͒, ͑d͒ G92-0-1-10
␥-MnO ͑43,20͔͒, and ͑e͒ G92-0-0-250
2
␥-MnO (ϳ40, у 50)].
2
Ϫ
both samples the presence of individual crystallites or fibers is not
obvious.
deposition rate ͑because MnO4 is formed simultaneously͒. Looking
at the edges of broken pieces of the deposits, Fig. 12b and d, it can
be seen that the difference in the size of the particles is not only a
surface phenomenon, but also applies to the bulk of the film. From
this, it can be extrapolated that the morphology and particle size
differences will also be applicable after manual grinding.
Regarding the pure ␥-MnO2 phases, several surface morpholo-
gies were observed. In all cases, the diameter of the fibers was
significantly less than 1 ␮m. Some of these morphologies are shown
in Fig. 14, along with the trends observed when changing the vari-
ous synthesis parameters. The general trends illustrated in the figure
can be summarized as follows. A comparison of Fig. 14a ͕G92-0-
0-10 ͓␥-MnO2 ͑28,21͔͖͒ with 14b ͕G113-0-0-10 ͓␥-MnO2 ͑48,23͔͖͒
The samples containing both ␣- and ␥-phases ͓G92-n-0-10
(
␣/␥-MnO ) and G80-0-0-10 (␣•␥-MnO )] display different sur-
2
2
face morphologies than either of the pure phases ͑pure ␥-MnO
2
discussed below͒ and are shown in Fig. 13a and c, respectively.
and 14c ͕G92-1-0-10 ͓␥-MnO ͑49,25͔͖͒ shows that an increase in
2
G92-n-0-10 (␣/␥-MnO ) is characterized by rodlike ͑100 nm diam͒
the temperature or the pH of the solution led to a change in mor-
2
fibers, while G80-0-0-10 (␣•␥-MnO ) is characterized by
phology and an increase in fiber size. Comparison of Fig. 14a with
2
ϩ
pyramidal-type structures ͑ϳ100 nm across͒ composed of smaller
needles. It is important to note that both samples exhibited homoge-
neous surface morphologies. No visual evidence of two different
phases was found. A look at the edges of the deposits, Fig. 13b and
d, reveals that there is not much difference in the two samples. In
14d ͕G92-0-1-10 ͓␥-MnO ͑43,20͔͖͒ illustrates that addition of Li
2
to the solution led to little or no change in either the morphology or
the size of the fibers. Figure 14e indicates that an increase in the
applied current density led to a similar morphology with a slight
decrease in the size of the fibers. Finally, it is noted that no corre-

D146
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
Figure 15. SEM photographs of
-MnO2 samples representative of the
␥
two types of bulk morphology found
among the ␥-MnO2 compounds synthe-
sized in this study. Fibrous morphology
of G92-0-1-10 ͓␥-MnO2 ͑36,21͔͒ as
seen from ͑a͒ a cross-sectional view of
the deposit, and ͑b͒ the sample after
manual grinding with
pestle. Nonfibrous morphology of G92-
.7-1-10 ͓␥-MnO ͑41,21͔͒ as seen from
a mortar and
7
2
͑
c͒ a cross-sectional view of the deposit,
and ͑d͒ the sample after manual grinding
with a mortar and pestle. The arrows in
͑
a͒ and ͑c͒ indicate the direction of
growth during deposition.
lation is observed between the structural parameters and the ob-
served morphology of the samples, indicating that the synthesis pa-
rameters are the controlling factors. For example, samples G113-0-
curves for sample G92-1-0-10 ͓␥-MnO ͑49,25͔͒. The form of the
2
TGA/DSC curves is typical of the ␥-MnO materials obtained in
2
this study. Note that there is a continuous weight loss over the tem-
perature range studied, which corresponds to water loss followed by
the decomposition of the material to form Mn O at ϳ550°C. Con-
0
-10 ͓␥-MnO ͑48,23͔͒, G92-1-0-10 ͓␥-MnO ͑49,25͔͒ and G92-0-
2
2
1-10 ͓␥-MnO2 ͑43,20͔͒ ͑Fig. 14b, c, and d, respectively͒ are
2
3
characterized by approximately the same structural parameters, but
show clearly different morphologies. Similarly, samples G92-0-0-10
cerning the weight loss corresponding to water release, an initial
constant rate weight loss was observed from 20 to 200°C. Above
200°C, water was lost at a slightly slower rate until a noticeable loss
occurred at approximately 360°C. The inferred composition from
the TGA curve is MnO •0.15H O. Results for other ␥-MnO com-
͓
2
␥-MnO2 ͑28,21͔͒, G92-0-1-10 ͓␥-MnO2 ͑43,20͔͒ and G92-0-0-
50 ͓␥-MnO ͑ϳ40,у50͔͒ ͑Fig. 14a, d, and e, respectively͒ all ex-
2
hibited very similar morphologies but are characterized by quite
2
2
2
different P and Mt values. In summary, by changing the experi-
pounds studied here are listed in Tables II and III. The water con-
tents of the ␥-MnO2 synthesized here by the electrochemical-
hydrothermal method are close to or slightly less than for other
r
mental parameters, it is possible to control the morphology of the
deposit obtained, as well as the structure/structural parameters of the
material.
As in the case of the ␣-MnO and ␣/␥-MnO /␣•␥-MnO com-
2
2
2
pounds, the edges of the ␥-MnO deposits was also studied to get an
2
idea of the ‘‘bulk’’ morphologies of the samples in comparison with
the surface morphology. In looking at the ensemble of the ␥-MnO2
compounds, it was found that two main ‘‘bulk’’ morphologies were
exhibited; ‘‘fibrous’’ and ‘‘nonfibrous.’’ Two representative samples
are shown in Fig. 15. Figure 15a illustrates sample G92-0-1-10
͓
␥-MnO ͑36,21͔͒ with a fibrous morphology, while Fig. 15c shows
2
sample G92-7.7-1-10 ͓␥-MnO ͑41,21͔͒ with a nonfibrous morphol-
2
ogy. Figure 15b and d show these samples after grinding. Here it is
evident that the two different morphologies are preserved; Fig. 15b
contained needle-like particles, while the particles shown in Fig. 15d
appear to have no particular shape. The differences in morphology
were reflected in the PXRD patterns, the patterns of fibrous samples
often displaying preferential orientation ͑noted decrease in intensity
of (hkl) (l Þ 0) lines͒.
An inspection of ␤-MnO ͑G140-1-1-10͒ ͑not shown͒ revealed a
2
deposit consisting of very densely packed particles, such that no
specific form could be identified. Further investigation of other
␤
-MnO2 samples was not pursued, as the study was focused on
materials for lithium batteries and the ␤-MnO structure is not ex-
pected to be electrochemically active.
2
Figure 16. TGA/DSC curves ͑obtained in air at 2 K/min͒ for sample G92-
1
-0-10 ͓␥-MnO ͑49,25͔͒, typical of the ␥-MnO materials obtained in this
2
2
Physicochemical analysis.—Figure 16 shows the TGA/DSC
study.

Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
D147
sponding Mn vacancies present. An attempt was made to exchange
ϩ
ϩ
ϩ
the H for K to determine the amount of exchangeable H present
in sample G92-0-0-50. EDS/energy dispersive X-ray analysis of the
exchanged material showed an average of 0.08 K/Mn. However, it is
also possible that the exchange is accompanied by a redox reduction
5
0
reaction ͑involving Mn reduction͒, so this number may be taken as
an upper limit of the number of H /Mn if the exchange rate is
ϩ
assumed to be 100%. As a consequence of this evaluation, we be-
lieve that the ␣-MnO samples prepared in this work contained only
2
ϩ
a very small amount of protons or Li ions within the tunnels of the
structure, after dehydration. The low content of small sized ions in
␣
-MnO , together with the high average oxidation state of Mn
2
should be favorable for interesting insertion behavior in terms of
ϩ
Li diffusion kinetics and insertion capacity, respectively.
BET measurements were performed for most samples. Results
are tabulated in Tables II and III. For samples synthesized at an
2
applied current density of 0.36 mA/cm , the measured specific sur-
2
face areas are in the range of 5-10 m /g. However, as the applied
current density is increased, we see a corresponding increase in the
specific surface area. This increase in the specific surface area is
paralleled by an increase in the value of n in MnO •nH O, as mea-
Figure 17. TGA/DSC curves ͑obtained in air at 2 K/min͒ for sample G92-
0
-0-50 (␣-MnO ), typical of the ␣-MnO materials obtained in this study.
2
2
2
2
sured by TGA.
␥
-MnO2 compounds synthesized by other means.3,45 Considering
Conclusion
the shape of the TGA curve below 500°C, it appears that about half
of the weight loss can be attributed to ‘‘loosely bound’’ water. In-
deed, if one assumes that the water released below ca. 200°C can be
Using the hydrothermal-electrochemical synthesis method,
MnO compounds of the ␣-, ␤-, and ␥-structure types, as well as
2
two phase mixtures of the ␣/␥ and ␥/␤ structure types and intercon-
nected ␣- and ␥- structure types ͑␣•␥͒ have been prepared. For
MnSO or Li SO /MnSO solutions at pH 1 and T у 130°C, pure,
attributed to loosely bound water ͑following the model by
Ruetschi͒,46,47
then the so-called structural water is calculated to be
4
2
4
4
0
.06 H O per MnO .
2 2
well crystallized ␤-MnO was formed, the first time this structure
2
Figure 17 shows the TGA/DSC curves for sample G92-0-0-50
has been obtained electrochemically. At 113°C and pH у 1, mix-
(
␣-MnO ) which is representative of those obtained for the
2
tures of ␤- and ␥-MnO were obtained. Increasing the acidity ͑to
2
␣
-MnO compounds synthesized here. Roughly similar to the case
2
pH р 0) or decreasing the temperature to 92°C ͑for pH у 0), led
of ␥-MnO , the figure shows a mass loss corresponding to water
2
to the formation of pure ␥-MnO materials covering a wide range of
loss below ϳ450°C, followed by a mass loss at ϳ500°C corre-
2
pyrolusite intergrowth levels (30 р P р 70) and low levels of mi-
sponding to the formation of Mn O . However, in the case of
r
2
3
crotwinning defects (Mt Ϸ 20). This range includes defect concen-
trations which are significantly lower than typically found for
␣
-MnO , the weight loss is greater, consistent with the larger tun-
2
nels present in this structure. Also notable, is that the major mass
loss due to the release of water occurs below 200-250°C. The total
water content is calculated to be 0.28 H O per MnO . The same
␥
-MnO prepared by traditional electrochemical syntheses. In the
2
ϩ
absence of Li , when the pH is decreased to pH Ͻ 0 at 92°C or the
temperature is decreased to 80°C at pH 0, mixtures of ␣- and
2
2
experiment was performed for other samples to estimate the total
water content. Results are reported in Tables II and III. The water
␥
-MnO2 or interconnected ␣- and ␥-phases ͑noted ␣•␥͒, respec-
ϩ
tively, can be obtained. In the presence of Li , under these same
content of the ␣-MnO samples prepared here is similar to that of
2
11,13,16
samples prepared by other methods.
After annealing at 250 or
conditions, ␣-MnO2 is formed. This is the first time ␣-MnO2 with-
ϩ
3
00°C for 16 h, the water content of ␣-MnO samples was signifi-
out ‘‘stabilizing’’ cations such as K has been prepared electro-
2
cantly lower ͑Ͻ0.05 H O per MnO ). After heating at 100°C under
vacuum in preparation for lithium battery studies, a TGA study ͑not
chemically. Alternatively ␣-MnO2 can be synthesized at 92°C and
pH 0 if the current density is increased from 0.36 to 1.8 mA/cm .
Further increasing the current density under these conditions leads
to ␥-MnO2 materials characterized by high concentrations of mi-
2
2
2
shown͒ revealed that very little or no water ͑Ͻ0.005 H O/Mn) re-
2
mained in the structure.
Redox titration of the samples showed a manganese oxidation
crotwinning defects ͑у50͒. Heat-treatment of the ␥-MnO materials
2
state of ϩ4.00 for the ␤-MnO materials. The ␥-containing samples
showed average manganese oxidation states ranging from ϩ3.77 to
ϩ3.99, with the higher values obtained at higher temperature, lower
leads to slight decreases in the concentrations of microtwinning de-
fects, and in many cases a constant or slightly lower concentration
of pyrolusite defects, a behavior not reported to date. The effect of
2
ϩ
pH, or lower applied current density. These results are in agreement
Na in the solution favors ␥-MnO materials with slightly lower P
2
ϩ
r
with the literature.4
5,48,49
The Mn oxidation state ranged from ϩ3.77
ϩ
ϩ
values than those obtained with Li . Both K and NH4 stabilized
to ϩ3.89 for the ␣-MnO materials, following the same trends as in
the formation of ␣-MnO . It has been shown that by changing the
experimental parameters, in addition to controlling the structure, the
morphology of the materials obtained can also be controlled.
2
2
the ␥-MnO2 materials. AAS/ICP analysis of the ␤-MnO and
2
ϩ
␥
-MnO containing samples prepared in the presence of Li shows
2
very little to no Li in the final product ͑atomic ratio of Ͻ0.005
ϩ
References
Li:Mn͒. Samples of pure ␣-MnO prepared in the presence of Li
2
show an atomic ratio of 0.03-0.05 Li:Mn. Results are listed in Tables
II and III. EDS/EDX analysis of the samples prepared in the pres-
1. P. M. De Wolff, Acta Crystallogr., 12, 341 ͑1959͒.
. R. M. McKenzie, Miner. Mag., 38, 493 ͑1971͒.
2
3
4
. K. M. Parida, S. B. Kanungo, and B. R. Sant, Electrochim. Acta, 26, 435 ͑1981͒.
. T. Ohzuku and T. Hirai, in Manganese Dioxide Electrode Theory and Practice for
Electrochemical Applications, B. Schumm, Jr., R. L. Middaugh, M. P. Grotheer,
and J. C. Hunter, Editors, PV 85-4, p. 262, The Electrochemical Society Proceed-
ings Series, Pennington, NJ ͑1985͒.
ϩ
ence of Na indicated small amounts of Na in the product ͑р0.03
ϩ
Na/Mn͒. ␣-MnO prepared in the presence of K was found ͑AAS/
2
ICP analysis͒ to contain approximately 0.1 K per Mn.
ϩ
It is possible that those samples prepared in the absence of Li
5
. T. Ohzuku, M. Kitagawa, K. Sawai, and T. Hirai, J. Electrochem. Soc., 138, 360
͑1991͒.
ϩ
ϩ
contain H ͑or H O ) in the channels of the structure, with corre-
3

D148
Journal of The Electrochemical Society, 150 ͑8͒ D135-D148 ͑2003͒
6
7
. P. Botkovitz, P. Deniard, M. Tournoux, and R. Brec, J. Power Sources, 43-44, 657
1993͒.
. R. N. DeGuzman, Y.-F. Shen, E. J. Neth, S. L. Suib, C.-L. O’Young, S. Levine, and
J. M. Newsam, Chem. Mater., 6, 815 ͑1994͒.
30. L. I. Hill, R. Portal, A. Verbaere, and D. Guyomard, Electrochem. Solid-State Lett.,
͑
4, A180 ͑2001͒.
31. L. I. Hill, A. Verbaere, and D. Guyomard, J. New Mater. Electrochem. Sys. , 5, 129
͑2002͒.
32. M. Evain, U-Fit Computer Program, Institut des Mat e´ riaux de Nantes, Nantes,
8
9
. S. Bach, J. P. Pereira-Ramos, and N. Baffier, Solid State Ionics, 80, 151 ͑1995͒.
. M. Tsuji and M. Abe, Solvent Extr. Ion Exch., 2, 253 ͑1984͒.
France ͑1992͒.
1
0. M. A. Humbert, P. Biensan, M. Broussely, A. Lecerf, A. Doll e´ , and H. Ladhily, J.
Power Sources, 43-44, 681 ͑1993͒.
33. K. J. Vetter and N. Jaeger, Electrochim. Acta, 11, 401 ͑1966͒.
34. S. Sarciaux, Ph.D. Thesis, Universit e´ de Nantes, Nantes, France ͑1998͒.
1
1. J. Dai, S. F. Y. Li, K. S. Siow, and Z. Gao, Electrochim. Acta, 45, 2211 ͑2000͒.
2. M. H. Rossouw, D. C. Liles, and M. M. Thackeray, Mater. Res. Bull., 27, 221
1992͒.
35. L. Li and G. Pistoia, Solid State Ionics, 47, 241 ͑1991͒.
36. M. Yoshio, H. Noguchi, T. Miyashita, H. Nakamura, and A. Kozawa, J. Power
Sources, 54, 483 ͑1995͒.
37. G. Pistoia and A. Antonini, J. Electrochem. Soc., 144, 1553 ͑1997͒.
38. E. Levi, E. Zinigrag, H. Teller, M. D. Levi, D. Aurbach, E. Mengeritsky, E. Elster,
P. Dan, E. Granot, and H. Yamin, J. Electrochem. Soc., 144, 4133 ͑1997͒.
39. E. Deltombe, N. de Zoubov, and M. Pourbaix, in Rapport Technique du CEBEL-
COR, Gauthier, Villars and Cie, Paris ͑1963͒.
1
1
͑
3. Q. Feng, H. Kanoh, K. Ooi, M. Tani, and Y. Nakacho, Electrochem. Solid-State
Lett., 141, L135 ͑1994͒.
1
1
4. Q. Feng, H. Kanoh, Y. Miyai, and K. Ooi, Chem. Mater., 7, 148 ͑1995͒.
5. M. H. Rossouw, D. C. Liles, and M. M. Thackeray, Prog. Batteries Battery Mater.,
1
5, 8 ͑1996͒.
1
1
6. C. S. Johnson, D. W. Dees, M. F. Mansuetto, M. M. Thackeray, D. R. Vissers, D.
Argyriou, C.-K. Loong, and L. Christensen, J. Power Sources, 68, 570 ͑1997͒.
7. P. Botkovitz, R. Brec, P. Deniard, M. Tournoux, and G. Burr, Mol. Cryst. Liq.
Cryst. Sci. Technol., Sect. A, 244, 233 ͑1994͒.
40. W.-H. Kao, J. Electrochem. Soc., 136, 13 ͑1989͒.
41. C. S. Johnson, M. F. Mansuetto, M. M. Thackeray, Y. Shao-Horn, and S. A. Hack-
ney, J. Electrochem. Soc., 144, 2279 ͑1997͒.
42. L. I. Hill, H. Arriv e´ , and D. Guyomard, Ionics, 8, 161 ͑2002͒.
1
1
8. Y. Chabre and J. Pannetier, Prog. Solid State Chem., 23, 1 ͑1995͒.
9. S. Sarciaux, A. Le Gal La Salle, A. Verbaere, Y. Piffard, and D. Guyomard, Mater.
Res. Soc. Symp. Proc., 548, 251 ͑1999͒.
43. S. Peulon and D. Lincot, J. Electrochem. Soc., 145, 864 ͑1998͒.
44. C. Poinsignon, J. M. Amarilla, Y. Chabre, and J. Pannetier, Abstract CO5-7,
French-Canadian Conference Journ e´ es d’Electrochimie, Ecole Polytechnique de
2
2
2
0. M. M. Thackeray, M. H. Rossouw, R. J. Gummow, D. C. Liles, K. Pearce, A. de
Kock, W. I. F. David, and S. Hull, Electrochim. Acta, 38, 1259 ͑1993͒.
1. S. Sarciaux, A. Le Gal La Salle, A. Verbaere, Y. Piffard, and D. Guyomard, J.
Power Sources, 81-82, 656 ͑1999͒.
2. L. A. H. MacLean, J. M. Amarilla, C. Poinsignon, F. Le Cras, and P. Strobel, J.
Mater. Chem., 5, 1183 ͑1995͒.
Montreal ͑1997͒.
45. A. Le Gal La Salle, S. Sarciaux, A. Verbaere, Y. Piffard, and D. Guyomard, J.
Electrochem. Soc., 147, 945 ͑2000͒.
46. P. Ruetschi, J. Electrochem. Soc., 131, 2737 ͑1984͒.
47. P. Ruetschi, in Manganese Dioxide Electrode Theory and Practice for Electro-
chemical Applications, B. Schumm, Jr., M. P. Grotheer, R. L. Middaugh, and J. C.
Hunter, Editors, PV 85-4, p. 112, The Electrochemical Society Proceedings Series,
Pennington, NJ ͑1985͒.
48. T. Ohzuku, M. Kitagawa, and T. Hirai, J. Electrochem. Soc., 137, 40 ͑1990͒.
49. R. Williams, R. Fredlein, G. Lawrance, D. Swinkels, and C. Ward, Prog. Batteries
Battery Mater., 13, 102 ͑1994͒.
2
2
2
3. M. Yoshimura, W. Suchanek, and K.-S. Han, J. Mater. Chem., 9, 77 ͑1999͒.
4. K.-S. Han, S.-W. Song, and M. Yoshimura, Chem. Mater., 10, 2183 ͑1998͒.
5. L. Liu, X. Wang, R. Bontchev, K. Ross, and A. J. Jacobson, J. Mater. Chem., 9,
1
585 ͑1999͒.
2
2
2
6. X. Wang, L. Liu, and A. J. Jacobson, Z. Anorg. Allg. Chem., 624, 1977 ͑1998͒.
7. M. Mauthoor, Ph.D. Thesis, University of the Witwatersrand, Johannesburg ͑1995͒.
8. M. Mauthoor, A. W. Bryson, and F. K. Crundwell, Prog. Batteries Battery Mater.,
50. Q. Feng, H. Kanoh, Y. Miyai, and K. Ooi, Chem. Mater., 7, 1226 ͑1995͒.
51. B. Tr e´ millon, in Electrochimie Analytique et R e´ actions en Solutions Aqueuses, p.
1
6, 105 ͑1997͒.
13, Masson, Paris ͑1993͒.
52. M. Pourbaix, in Atlas d’Equilibres Electrochimiques, Gauthiers-Villars, Paris
͑1963͒.
´
´
2
9. L. I. Hill, R. Portal, A. Le Gal La Salle, A. Verbaere, and D. Guyomard, Electro-
chem. Solid-State Lett., 4, D1 ͑2001͒.