Chronoamperometric versus galvanostatic preparation of manganese oxides for electrochemical capacitors

Article abstract of DOI:10.1149/1.3625581

Chronoamperometric and galvanostatic methods of manganese dioxide electrodeposition have been compared in this work for their ability to produce high specific capacitance manganese dioxide electrodes for supercapacitor applications. When directly compared in terms of the charge passed (and hence the mass of manganese dioxide deposited) and the timeframe of electrodeposition, chronoamperometry most often led to superior performing electrodes. In the best case, an electrode was prepared that attained 2986 F/g. The results of this work were interpreted in terms of the manganese dioxide deposition mechanism and its impact on deposit morphology. Furthermore, it was concluded that the loss of charge during electrodeposition, as either the result of Mn³ intermediate species diffusing away from the electrode, or the competing oxygen evolution process, significantly influenced the ability to estimate the amount of manganese dioxide prepared.

Full text of DOI:10.1149/1.3625581

A1160
Journal of The Electrochemical Society, 158 (10) A1160-A1165 (2011)
0
013-4651/2011/158(10)/A1160/6/$28.00 VC The Electrochemical Society
Chronoamperometric Versus Galvanostatic Preparation of
Manganese Oxides for Electrochemical Capacitors
a
b
c
a,*,z
Andrew D. Cross, Alban Morel, Tony F. Hollenkamp, and Scott W. Donne
a
Discipline of Chemistry, University of Newcastle, Callaghan, NSW 2308, Australia
Ecole Polytechnique de l’universit e´ de Nantes, 44306 Nantes Cedex 3, France
CSIRO Energy Technology, PO Box 312, Clayton South, VIC 3169, Australia
b
c
Chronoamperometric and galvanostatic methods of manganese dioxide electrodeposition have been compared in this work for
their ability to produce high specific capacitance manganese dioxide electrodes for supercapacitor applications. When directly
compared in terms of the charge passed (and hence the mass of manganese dioxide deposited) and the timeframe of electrodeposi-
tion, chronoamperometry most often led to superior performing electrodes. In the best case, an electrode was prepared that attained
2
986 F/g. The results of this work were interpreted in terms of the manganese dioxide deposition mechanism and its impact on de-
3
þ
posit morphology. Furthermore, it was concluded that the loss of charge during electrodeposition, as either the result of Mn in-
termediate species diffusing away from the electrode, or the competing oxygen evolution process, significantly influenced the
ability to estimate the amount of manganese dioxide prepared.
VC 2011 The Electrochemical Society. [DOI: 10.1149/1.3625581] All rights reserved.
Manuscript submitted April 18, 2011; revised manuscript received June 28, 2011. Published August 16, 2011.
Introduction
are capable of achieving relatively high cell voltages (2.0–4.0 V),
one of their major disadvantages, along with their flammability, tox-
icity and relatively high cost, is their high resistivity. This is due to
larger electrolyte ions, which causes a decreased ionic mobility,
which in turn results in a somewhat lower specific power. As a
result, electrochemical capacitor systems utilising aqueous electro-
lytes have received considerable attention, despite their lower oper-
ating voltages, which are typically in the order of 1 V.
The importance of energy and energy storage.— One of the
major concerns faced by the modern world is the delivery of a con-
stant and efficient supply of energy. The combustion of fossil fuels
such as petroleum, natural gas and coal has for many years been the
traditional source of energy, and indeed accounted for 85% of the
1
world’s primary sources of energy. Fossil fuels are not without their
drawbacks; indeed, in recent years, concern over the emission of
greenhouse gases produced from combustion of fossil fuels, and their
suspected role in climate change, as well as the limited supply of oil,
coal and natural gas reserves has led to an increased focus on devel-
oping more renewable sources of energy. These renewable sources
As the energy of a capacitor is related to the capacitance and
voltage of the device by
CV2
E ¼
[1]
2
(
solar, wind, geothermal, etc.) tend to be implemented on the smaller
scale, and deliver power intermittently. To counteract this, any
excess energy produced must be stored, so as to be utilised during
power sags and compensate for the decreased output of the primary
power generator.
where E is the energy stored (J), C is the capacitance (F) and V is
the cell voltage (V), many efforts have been made to compensate
for the limited width of the potential window in aqueous-based elec-
trochemical capacitors. Some previously known electroactive mate-
rials have recently been examined due to their high capacitance.
Energy storage devices.— With current levels of technology,
there are many methods of energy storage, including mechanical
5
–14
These materials have included transition metal oxides,
amorphous hydrated RuO exhibiting the highest reported capaci-
tance (ꢀ760 F/g in acidic electrolytes
the positive limit of the potential window, with some studies report-
with
2
(
(
hydroelectric, flywheel and compressed air storage) and thermal
thermal reservoirs such as water, and molten salts of fluorides,
5
,15–17
), and can contribute to
chlorides and nitrates) methods. These are relatively low-grade,
with efficiencies in the order of 50–90%.
1
8–20
2
,3
ing an extension of the potential window to 1.4 V,
thereby
achieving higher energy and power densities when compared to
other systems. However, despite these improvements, the cell volt-
age is still too low to achieve satisfactory energy for practical appli-
cations. Additionally, the prohibitive cost of ruthenium, especially
as its oxide, low pore volume, and toxicity of such compounds con-
tributes to the impracticability of capacitors made from such materi-
als. Oxides of other transition metals, such as manganese, have been
widely investigated as the positive electrode in asymmetric systems,
with a carbon material (such as activated carbon or graphite) as the
Storage as chemical energy is also an alternative; however, com-
bustion of hydrocarbons is also somewhat inefficient, with the typi-
cal efficiency of an internal combustion engine in the range of 35%.
Electrochemical storage and conversion in devices such as batteries,
fuel cells and capacitors, however, is highly efficient and reversible
in many cases. The relative merits of each can be summarised using
4
a typical Ragone plot. Batteries and fuel cells have a relatively
high specific energy, but a low specific power. Conversely, capaci-
tors have a low specific energy and a high specific power, making
them well suited for applications requiring high current in a short
time period. Also included in this category are supercapacitors, or
electrochemical capacitors, which have similar power densities as
conventional capacitors, but energy densities several orders of mag-
nitude larger. The focus of this work is on further improving the
energy density of electrochemical capacitors.
21
negative electrode.
Manganese dioxide (commonly referred as MnO , but more
accurately a non-stoichiometric arrangement of [Mn(O , OH ,
2
2
ꢁ
ꢁ
zþ
H
2
O)
6
]
octahedra) have a long history as cathode materials in
aqueous and non-aqueous batteries, and have proven to be a promis-
ing category of electrochemical capacitor electrode materials. Previ-
ous findings have shown that films of manganese dioxide deposited
using chronoamperometry exhibited specific capacitances >2000
Materials for electrochemical capacitors.— Existing commercial
electrochemical capacitors are composed of activated carbon elec-
trodes with an organic electrolyte. They exhibit high reversibility in
charge/discharge cycling due to the mechanism of charge storage in
the electric double layer. Despite the fact that organic electrolytes
22
F/g. In these cases it was deduced that careful control of the mass
transport characteristics by varying certain parameters during depo-
sition was important – specifically, it was demonstrated that lower
deposition voltages, and high acid/low Mn(II) concentrations in the
electrolyte produced the best performing electrodes. These condi-
tions led to slower mass transport, which apparently has a favour-
able effect on material morphology.
*
z
Electrochemical Society Active Member.
E-mail: scott.donne@newcastle.edu.au
Downloaded on 2014-11-14 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 (10) A1160-A1165 (2011)
A1161
This work.— This study aims to compare the performance of
electrodes deposited using chronoamperometry (constant potential),
to those using a galvanostatic (constant current) deposition protocol.
The aim is to pass equivalent amounts of charge during deposition,
but vary the current density. In addition to material performance,
insight into the manganese dioxide deposition mechanism will also
be achieved.
Experimental
Electrodeposition.— All deposition solutions contained a combi-
nation of 0.01 M MnSO (AnalaR, >98%) and 0.1 M H SO (Ajax
4 2 4
Finechem, 95–98%). The working electrodes used during deposition
and cycling were Radiometer Analytical XM150 platinum electro-
des (1 cm diameter, 0.785 cm area), with a saturated calomel elec-
trode (SCE) as a reference, and a graphite rod as a counter electrode.
Initially, so as to determine appropriate voltages for deposition, a
linear sweep voltammetry experiment was performed using a Per-
kin-Elmer VMP multichannel potentiostat/galvanostat, and the cell
voltage was swept from the open circuit potential (ꢀ0.6 V) to 0 V,
and then to 1.6 V (vs. SCE) at a scan rate of 5 mV/s.
After obtaining the linear sweep voltammogram, a potential in
the non-diffusion limited region was selected (1.2 V), and was used
as the step potential (from the open circuit potential) in a chro-
noamperometic deposition. Chronoamperometric deposition was
performed initially for 30 s, and the total charge passed during this
deposition was calculated via integration of the current vs. time
plot. To compare the effects of varying the electrodeposition
method, galvanostatic depositions were also performed. Here the
equivalent amount of charge passed during the 30 s CA deposition
was passed over varying times ranging from 1–60 s. This process
was repeated for charges equivalent to 5, 15 and 60 s of chronoam-
perometry at 1.2 V.
2
Figure 1. Mechanistic pathways for the electrodeposition of manganese
dioxide.
electrodeposited manganese dioxide, as per our previous work in this
area. The voltammogram in Fig. 2 shows two closely overlapping
22
anodic waves at ꢀ1.253 and ꢀ1.364 V which correspond to the two
2þ
2
stage oxidation of Mn to MnO . What is interesting, and very rele-
vant for our ability to estimate the amount of electrodeposited man-
ganese dioxide, is that at slightly higher and overlapping potentials
there is evidence for the oxygen evolution reaction occurring. How-
ever, since we have chosen a non-diffusion limited potential (1.2 V
vs SCE) to carry out our chronoamperometry experiments, the con-
tributions made by oxygen evolution to the total anodic charge
passed are expected to be minimal.
A typical chronoamperogram (i vs t) is shown in Fig. 3, in which
case the potential was stepped from the open circuit potential to
1
.2 V vs SCE for 30 s. It shows the expected spike in current after
the imposition of the step potential, as well as the subsequent decay
in current with time. What was unexpected was the slight increase
in current after longer deposition times. This phenomenon, after
consideration of the details of a chronoamperometry experiment,
Electrochemical cycling.— The electrodeposited MnO was then
2
electrochemically tested via cyclic voltammetry. After deposition,
whether by chronoamperometry or galvanostatic means, the depos-
ited films were washed thoroughly with Milli-Q water, patted dry
with paper towel and then transferred to the cycling electrolyte (0.5
23
and the factors that influence its behaviour, is believed to be due
to a changing electrode area as a result of manganese dioxide depo-
sition. Similar results have been obtained in our research group with
24
M Na SO ) together with a saturated calomel reference and graphite
4
2
the electrodeposition of poly(terthiophene). For the data shown in
Fig. 3 it is not possible to estimate the increase in area of the elec-
trode (as was done in the literature case) because the conditions
counter electrode. The cell was cycled between 0–0.8 V vs SCE for
0 cycles at a scan rate of 5 mV/s, using the same Perkin-Elmer
VMP as for the deposition.
5
Results and Discussion
Manganese dioxide electrodeposition mechanism.— Since the
nature of the manganese dioxide materials we will be producing
will be dependent on the conditions under which they were made, it
is appropriate to begin by presenting the currently accepted mecha-
nisms for manganese dioxide electrodeposition. There are several
deposition mechanisms that have been proposed for the oxidation of
2
þ
2
Mn to MnO . The two main pathways are dependent on acid con-
centration, and are hence referred to as the low acid and high acid
mechanisms. As shown as Path A in Fig. 1, the low acid mechanism
2
þ
3þ
3þ
involves oxidation of Mn to Mn , hydrolysis of the Mn to pre-
cipitate MnOOH on the electrode surface, and then further solid
state oxidation to give MnO
1
2
. The high acid process (Path C in Fig.
) also begins with an oxidation step; however, the next step is a dis-
3þ
4þ
2þ
proportionation (2Mn ! Mn þ Mn ), and a final hydrolysis of
2
. While possible, there has been no
4
þ
the resulting Mn to give MnO
experimental evidence presented in the literature to support the
direct two electron oxidation of Mn to Mn (Path B in Fig. 1).
2
þ
4þ
Electrodeposition data.— As outlined in the Experimental sec-
tion, the appropriate non-diffusion limited voltage had to be ascer-
tained from a linear sweep voltammogram in the chosen electrolyte.
The linear-sweep voltammogram obtained from 0.1
SO is shown in Fig. 2. This electrolyte
þ 0.01 M MnSO
was chosen because of its ability to produce high performing
M
Figure 2. Linear sweep voltammogram for the electrodeposition of manga-
nese dioxide from an electrolyte of 0.01 M MnSO þ 0.1 M H SO . Scan
H
2
4
4
4
2
4
rate 5 mV/s.
Downloaded on 2014-11-14 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1162
Journal of The Electrochemical Society, 158 (10) A1160-A1165 (2011)
Figure 3. Chronoamperometric data (i vs t) from an electrolyte of 0.01 M
MnSO SO . Step potential 1.2 V vs SCE. Inset: Charge passed
þ 0.1 M H
during each deposition.
Figure 4. Potential (vs. SCE) versus deposition time for the galvanostatic
depositions equivalent to 15 s chronoamperometry in 0.1 M H SO þ 0.01 M
4
2
4
2
4
MnSO . Current densities used are indicated on each data set.
4
under which electrodeposition occurred were not diffusion limited,
As a result of holding the current density at a constant value, as
opposed to a constant potential during chronoamperometry, the
potential during galvanostatic deposition changes over time. Exam-
ples of this for galvanostatic depositions equivalent to 15 s of chro-
noamperometry in 0.1 M H SO þ 0.01 M MnSO are shown in
2
þ
and hence the concentration of Mn at the electrode surface was
not zero, and variable. As such, the Cottrell equation for chronoam-
perometry experiments could not be applied.
2
3
After chronoamperometry was performed at 5, 15, 30 and 60 s,
the charge passed during each deposition was calculated via numeri-
cal integration of the current passed during deposition. This data is
included as an inset in Fig. 3, and also in Table I. This table also
contains the calculated theoretical mass of manganese dioxide de-
posited based on the charge passed. To demonstrate the conversion
from chronoamperometry to galvanostatic deposition, the charge
2
4
4
Fig. 4, where an interesting trend in the potential data can be seen.
For short timeframe deposition experiments (<15 s), and hence
higher current density experiments, the electrode potential increases
continually, finishing within the range 1.25–1.65 V depending on
the current density used. For longer duration experiments (>15 s),
and hence lower current densities, the potential exhibits a maxi-
mum, after which it drops and eventually plateaus. Once the plateau
was reached the electrode remains at this potential for the remainder
of the deposition.
2
passed during the 30 s deposition experiment was 12.425 mC/cm ,
so for galvanostatic deposition with the same charge, and for exam-
ple a deposition time of 30 s, the constant current will be equal to
0
2
.414 mA/cm . Table I also contains a listing of the galvanostatic
It is appropriate at this time to consider what the potential
changes represent in terms of the electrodeposition mechanisms
shown in Fig. 1. With the electrolyte conditions we are using, in
currents and times to match the charge passed during each of the
chronoamperometry experiments.
Table I. Data pertaining to the chronoamperometric and galvanostatic depositions of manganese dioxide.
Chronoamperometry
Time (s)
Charge (mC)
5
15
30
9.75
4.394
60
27.34
12.315
1.88
0.848
7.59
3.419
Galvanostatic deposition
Mass MnO
2
(lg)
2
Time (s)
Current (mA/cm )
1
2
5
2.348
1.186
0.476
0.237
0.158
0.119
0.094
0.079
0.051
9.473
4.784
1.925
0.963
0.642
0.481
12.176
6.150
2.474
1.239
0.781
0.495
0.619
0.412
0.414
12.646 (2.74 s)
6.698
3.355
10
15
20
25
30
45
60
2.238
1.343
1.119
0.746
0.320
0.212
0.160
0.580
0.560
11.962 6 0.184
0.039
0.833 6 0.009
0.274
4.343 6 0.071
Ave. MnO
2
Mass (lg)
3.391 6 0.018
Downloaded on 2014-11-14 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 (10) A1160-A1165 (2011)
A1163
particular the high acid concentration (0.1 M H
2
SO
4
), it is reasona-
deposition experiments were performed at the same temperature
ꢂ
3
þ
ble to presume that any Mn intermediate will have a relatively
long lifetime in solution before either hydrolysis or disproportiona-
tion occurs. What we propose is that at the start of the deposition
experiment the rapid increase in potential is due primarily to the
(22 6 1 C) and using the same electrolyte conditions, it is reasona-
ble to assume that the crystallographic phase of manganese dioxide
formed during deposition is the same. Therefore, any differences in
cycling performance are due to differences in the morphology of the
electrode film. It is also reasonable to expect that an electrode with a
relatively high surface area and porosity will result in a higher ca-
pacitance than a denser electrode with a reduced surface area. As a
result of this, any phenomena that is seen to occur in cycling data is
related to the morphology, which in turn is due to the differing con-
ditions during deposition, most notably the current and voltages
used or achieved during deposition.
Figure 5 shows typical cyclic voltammograms of an electrode-
posited manganese dioxide prepared in this work. As can be seen in
the figure, the voltammogram exhibits the expected “box” shape of
a supercapacitive electrode material. Furthermore, there is a high
degree of consistency with cycling as shown by the 25th cycle in
Fig. 5, as well as the anodic:cathodic charge ratio (inset), which is
essentially constant over the cycle range examined. To evaluate per-
formance we have determined the specific capacitance for each elec-
trode after 50 cycles. The mass of manganese dioxide used for the
normalization procedure was determined from the amount of charge
passed during the electrodeposition experiment, assuming that
3þ
formation of a soluble Mn intermediate. The maximum in the
3þ
electrode potential is obtained as a result of the Mn becoming
supersaturated in the vicinity of the electrode, after which either hy-
drolysis (to MnOOH, with subsequent solid state oxidation to
2
þ
MnO ) or disproportionation (to form Mn in solution and solid
2
MnO on the electrode) leads to the deposition of a lower potential
2
solid material on the electrode surface. The plateau represents the
quasi-equilibrium two phase potential between the solid manganese
dioxide produced and the bulk electrolyte concentration.
With this description of the mechanism occurring at the electrode
surface we are now in the position of being able to interpret the impli-
cations of a changing potential on the properties of the deposited
manganese dioxide, in particular the amount of material deposited. In
the first case, given the acid concentration of the electrolyte being
3þ
used it is possible that the soluble Mn intermediate formed may not
have sufficient time to either hydrolyse or disproportionate and de-
posit onto the electrode before the deposition experiment has been
completed. In fact, the increasing potential seen in Fig. 4 would indi-
cate that little deposition of a solid material onto the electrode has
occurred. The impact of this is that charge is effectively lost to the
electrolyte without leading to a manganese dioxide deposit on the
electrode. Furthermore, it means that we will overestimate the amount
of manganese dioxide on the electrode surface because of this loss of
2
MnO was the stoichiometric product. As was mentioned previ-
ously, there may be complications associated with this estimate of
mass, particularly for the short time frame, high current density
experiments, where some of the charge may be lost due the solubil-
3þ
ity and stability of Mn intermediate species, as well as from the
competing oxygen evolution reaction. Calculating the specific ca-
pacitance enables a comparison firstly between the use of different
deposition current densities, as well as whether the deposition was
carried out chronoamperometrically or galvanostatically.
3þ
charge (due to uncaptured Mn ) to the bulk of the electrolyte.
Another complicating factor associated with constant current
deposition and hence an uncontrolled potential is the fact that com-
peting reactions can also occur. From the linear sweep voltammo-
gram in Fig. 2 it was highlighted that manganese dioxide deposition
overlaps with oxygen evolution on the anode. Again, for the short
time frame, high current depositions oxygen evolution will be com-
peting with manganese dioxide electrodeposition. This overlapping
reaction can consume charge that would otherwise be associated
with manganese dioxide deposition, again leading to an overesti-
mate of the amount of material deposited.
Figure 6 shows the relationship between deposition time and spe-
cific capacitance for all of the experiments conducted. To begin, the
chronoamperometric data in this figure (large squares) shows a very
clear exponential drop in specific capacitance as the deposition
time, and hence deposition charge, was increased. This is consistent
with our previous work in the area, and is an indication of how the
density of the deposit increases as the deposition time increases.
In an attempt to confirm the mass of manganese dioxide
deposited onto the substrate duplicate electrodes were prepared and
washed thoroughly with Milli-Q ultra-pure water to remove any
remaining plating electrolyte. The electrode was then re-
dissolved in acidified (5% v/v H SO ) H O , with the resultant solu-
2
4
2 2
tion analysed for total manganese content using ICP-AES (Varian
Liberty Series II). Over the entire range of manganese dioxide elec-
trode materials prepared it was determined that using the i-t integral
(
2
charge) to determine the amount of MnO underestimated the elec-
trode mass by 5–10%. Nevertheless, for the purposes of this study,
the mass calculation using charge was still employed.
As a final comment in this section, it is anticipated that the cur-
rent used for deposition will dramatically affect the morphology of
the electrodeposited manganese dioxide, and hence the performance
of the material as a supercapacitor electrode. From our past experi-
2
2
ence with electrodeposited manganese dioxide we expect that the
extent of disorder within the deposit will be greatest for the higher
current density depositions, while the use of a low current density
will lead to a relatively dense, low porosity deposit. These statements
on morphology are supported by experimental evidence such as cat-
ion vacancy fraction, BET surface area, and crystal structure on bulk
deposits of manganese dioxide. For thin films of manganese dioxide,
such as we are producing here, characterizing their properties is
exceedingly difficult since there is very little material available for
characterization. In fact, the electrochemical cycling as a supercapa-
citor electrode in a suitable electrolyte is probably the most appropri-
ate method for evaluating the deposited material morphology.
Cycling performance.— Given that during the course of this
study the conditions used during deposition were similar; i.e., all
Figure 5. Typical cyclic voltammogram from an electrode prepared in this
work. Inset: Anodic:cathodic charge ratio for the corresponding electrode.
Downloaded on 2014-11-14 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A1164
Journal of The Electrochemical Society, 158 (10) A1160-A1165 (2011)
capacitance with deposition time was quite interesting. For longer
time frame deposits (>20 s) the specific capacitance decayed expo-
2
nentially from 1023 F/g (for a 20 s deposit using 0.481 mA/cm ) as
the deposition time increased. However, for shorter duration experi-
ments the specific capacitance was seen to increase from 733 F/g
2
for a 1 s deposit using 9.47 mA/cm ) up to a maximum at 20 s. We
(
believe that this phenomenon is a result of an inaccurate estimate of
the amount of manganese dioxide deposited onto the electrode sub-
strate. Under these circumstances we have a relatively small amount
of charge passed during deposition, and if some of that is lost as a
3
þ
result of soluble Mn formation, or from the competing oxygen
evolution reaction, then the overestimate of manganese dioxide de-
posited onto the substrate will be much more significant. Given that
for these shorter timeframe experiments the deposition potential is
3
þ
also high, indicating either excessive Mn formation or oxygen
evolution (see Fig. 4) as a likely explanation.
The same result is also apparent for the 5 s chronoamperometry
2
experiment (2.40 mC/cm of deposition charge) which generated
a specific capacitance of 2986 F/g. All of the galvanostatic deposi-
tions carried out with the same charge were inferior to the chro-
noamperometrically prepared sample; however, a maximum in
specific capacitance was also observed here. As before, this maxi-
mum is due to the combination of a small amount of deposition
charge and the high current density used for the short time frame
experiments leading to an overestimate of the amount of manga-
nese dioxide deposited.
Figure 6. Specific capacitance for each electrode. (n) Chronoamperometric
deposition, and galvanostatic data corresponding to (n) 60 s, (^) 30 s, (~)
1
5 s, and (ꢃ) 5 s of chronoamperometric charge.
Summary and Conclusions
In this study we have compared the ability of chronoamperome-
try and galvanostatic methods to electrodeposit thin films of manga-
nese dioxide for use as supercapacitor electrodes. The following
points are significant outcomes from the work:
The chronoamperometry experiment conducted over 60 s (34.82
2
mC/cm ) led to a very low specific capacitance, with only 45 F/g
being achieved. With such a relatively long time frame experiment,
with a relatively low current density used, it was to be expected that
this deposit was quite dense and hence of low capacitance. For all of
the corresponding galvanostatic depositions, the specific capacitance
was similar, and relatively constant irrespective of the deposition
current density used. For the short time frame deposition experi-
ments done here, we may have overestimated the amount of manga-
nese dioxide present, and so the specific capacitance here may
indeed be more enhanced than what was recorded. Nevertheless,
the average specific capacitance under these circumstances was
(
i) During chronoamperometric deposition an increase in the
current towards the end of the experiment is evidence of the elec-
trode area increasing as a result of the manganese dioxide
morphology.
(
ii) The mechanism of manganese dioxide electrodeposition
plays a significant role in determining the properties of the resultant
deposit. From the data collected, chronoamperometry is the pre-
ferred method for smaller amounts of charge (i.e. 1.88 and 7.59
mC), where as for larger amounts of charge (9.75 and 27.34 mC),
the performance of galvanostatically prepared electrodes is compa-
rable to those prepared from chronoamperometry.
4
2 6 5 F/g with no trends apparent in the data.
For the 30 s chronoamperometry experiment (12.42 mC/cm of
2
charge passed during deposition) a specific capacitance of 378 F/g
was achieved, with the equivalent galvanostatic deposit (0.412 mA/
cm for 30 s) resulting in a similar specific capacitance (402 F/g).
(
iii) For these conditions (short timeframe galvanostatic deposition
2
with high current density and low charge deposits) complications were
also identified relating to the mass of electrodeposited manganese
dioxide on the substrate. Under these circumstances, a significant pro-
These values are very similar to those obtained from the use of
lower deposition current densities (longer deposition times); i.e.,
3þ
2
17 F/g for the 45 s deposition (0.274 mA/cm ) and 388 F/g for
portion of the Mn intermediate was lost to the electrolyte leading to
4
the 60 s deposition (0.205 mA/cm ). It is to be expected, based on
2
an overestimate of the amount of deposited manganese dioxide. Oxy-
gen evolution as a competing anodic reaction can also contribute to an
inaccurate estimate of deposited manganese dioxide.
22
our previous work, that lower current densities result in a denser
electrode film, and hence a lower specific capacitance. When the
deposition time was much shorter, considerable improvements in
capacitive performance were observed. Indeed, a maximum specific
(
iv) The specific capacitance obtained for each electrode was
used as a tool to characterize the morphology of the deposited man-
ganese dioxide. Again, short timeframe experiments were preferred
because of their ability to produce high performance electrodes,
implying a deposit with high surface area and porosity. Longer time
frame experiments led to poor performing electrodes, indicating
much denser, less porous films.
2
capacitance was observed at for a 1 s deposition (12.176 mA/cm ),
with a value of 1032 F/g, which decreased steadily as the deposition
time increased until the aforementioned values for longer deposition
times were reached. What this data is beginning to suggest is that
for very short timeframe deposition experiments only the nuclei of
manganese dioxide crystallites are being produced on the substrate
surface. These nuclei have a very high surface area and hence when
cycled, lead to a very high specific capacitance. With continued dep-
osition, crystal growth predominates effectively closing off pores,
decreasing surface area, leading to a denser deposit and hence a
much lower specific capacitance.
Acknowledgments
AC acknowledges the financial support provided by the Univer-
sity of Newcastle and CSIRO Energy Technology in the form of a
postgraduate scholarship.
The deposit made using a 15 s chronoamperometry experiment
9.67 mC/cm ) had a specific capacitance of 1059 F/g. For the corre-
References
2
(
sponding galvanostatically prepared deposits the trend in specific
1.
http://www.eia.doe.gov/oiaf/ieo/world.html, accessed April 14, 2011.
2. A. Sari and K. Kaygusuz, Sol. Energy, 71, 365 (2001).
Downloaded on 2014-11-14 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 158 (10) A1160-A1165 (2011)
A1165
3
4
. R. F. Post, K. Fowler, and S. F. Post, Proc. IEEE, 81, 462 (1993).
. D. Ragone in Proceedings of Society of Automotive Engineers Conference, Detroit,
MI, May (1968).
15. J. P. Zheng, P. J. Cygan, and T. R. Jow, J. Electrochem. Soc., 142, 2699 (1995).
16. J. P. Zheng, Electrochem. Solid-State Lett., 2, 359 (1999).
17. M. C. Santos, A. J. Terezo, V. C. Fernandes, E. C. Pereira, and L. O. S. Bullhoes,
J. Solid State Electrochem., 9, 91 (2005).
18. Y.-G. Wang, Z.-D. Wang, and Y.-Y. Xia, Electrochim. Acta, 50, 5641 (2005).
19. B.-O. Park, C. D. Lokhande, H.-S. Park, K.-D. Jung, and O.-S. Joo, J. Mater. Sci.,
39, 4313 (2004).
20. X.-F. Wang, Z. You, and D.-B. Ruan, Chin. J. Chem., 24, 1126 (2006).
21. M. S. Hong, S. H. Lee, and S. W. Kim, Electrochem. Solid-State Lett., 5, A227 (2002).
22. A. D. Cross, A. Morel, A. Cormie, A. F. Hollenkamp, and S. W. Donne, J. Power
Sources, 196, 7847 (2011).
23. A. J. Bard and L. R. Faulkner, Fundamentals of Electrochemistry, 2nd ed., John
Wiley & Sons: New York, (2001).
5. J. P. Zheng and T. R. Jow, J. Electrochem. Soc., 142, L6 (1995).
6. H. H. Kim and K. B. Kim, Electrochem. Solid-State Lett., 4, A62 (2001).
7. J. K. Chang and W. T. Tsai, J. Electrochem. Soc., 150, A1333 (2003).
8. C. C. Hu and T. W. Tsou, J. Power Sources, 115, 179 (2003).
9. J. W. Long, A. L. Young, and D. R. Rolison, J Electrochem. Soc., 150, A1161
(
2003).
1
1
1
1
1
0. H. Kim and B. N. Popov, J. Electrochem. Soc., 150, D56 (2003).
1. R. N. Reddy and R. G. Reddy, J. Power Sources, 124, 330 (2003).
2. N. L. Wu, Mater. Chem. Phys., 75, 6 (2002).
3. V. Srinivasan and J. W. Weidner, J. Electrochem. Soc., 144, L220 (1997).
4. H. K. Kim, T. Y. Seong, J. H. Lim, W. I. Cho, and Y. S. Yoon, J. Power Sources,
24. D. Wakeham, S. W. Donne, W. J. Belcher, and P. C. Dastoor, Synth. Met., 158,
661 (2008).
1
02, 167 (2001).
Downloaded on 2014-11-14 to IP 128.82.252.58 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).