Electrochemical quartz crystal microbalance study of amorphous MnO 2 prepared by anodic deposition

Article abstract of DOI:10.1016/j.electacta.2011.11.106

In this work, the electrochemical behavior and the corresponding mass variation of amorphous manganese oxide (denoted as a-MnO_x) are examined simultaneously in a neutral electrolyte containing 10 mM Na ₂SO₄ without or with NaHCO₃ or Na ₂HPO₄ by cyclic voltammetry with a quartz crystal microbalance (QCM). a-MnO_x was anodically deposited onto the gold-sputtered quartz electrode by means of a potentiostatic method at 1.0 V for 60 s. The morphology and microstructure of a-MnO_x with different potential windows of CV cycling in the above electrolyte were examined by means of the scanning electron microscopy (SEM). From this EQCM study, a-MnO _x is unstable between 1.0 and 1.2 V in 10 mM Na₂SO ₄ because of significant dissolution of a-MnO_x. The repeated potential cycling promotes the electrocatalytic activity of a-MnO _x for oxygen evolution, which also enhances the dissolution of a-MnO_x. The dissolution phenomenon and oxygen evolution are successfully suppressed by the formation of insoluble manganese carbonate or the adsorption of phosphate by adding NaHCO₃ or Na₂HPO ₄ in the Na₂SO₄ electrolyte, enlarging the potential window for the charge/discharge of a-MnO_x.

Full text of DOI:10.1016/j.electacta.2011.11.106

Electrochimica Acta 61 (2012) 124–131
Contents lists available at SciVerse ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Electrochemical quartz crystal microbalance study of amorphous MnO2 prepared
by anodic deposition
Yi-Hsuan Chua, Chi-Chang Hua,∗,1, Kuo-Hsin Changa,b,1
a Department of Chemical Engineering, National Tsing Hua University, Hsin-Chu 30013, Taiwan
b Department of Chemical Engineering, National Chung Cheng University, Chia-Yi 621, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work, the electrochemical behavior and the corresponding mass variation of amorphous man-
ganese oxide (denoted as a-MnOx) are examined simultaneously in a neutral electrolyte containing
10 mM Na2SO4 without or with NaHCO3 or Na2HPO4 by cyclic voltammetry with a quartz crystal microbal-
ance (QCM). a-MnOx was anodically deposited onto the gold-sputtered quartz electrode by means of a
potentiostatic method at 1.0 V for 60 s. The morphology and microstructure of a-MnOx with different
potential windows of CV cycling in the above electrolyte were examined by means of the scanning elec-
tron microscopy (SEM). From this EQCM study, a-MnOx is unstable between 1.0 and 1.2 V in 10 mM
Na2SO4 because of significant dissolution of a-MnOx. The repeated potential cycling promotes the elec-
trocatalytic activity of a-MnOx for oxygen evolution, which also enhances the dissolution of a-MnOx. The
dissolution phenomenon and oxygen evolution are successfully suppressed by the formation of insolu-
ble manganese carbonate or the adsorption of phosphate by adding NaHCO3 or Na2HPO4 in the Na2SO4
electrolyte, enlarging the potential window for the charge/discharge of a-MnOx.
Received 3 September 2011
Received in revised form
21 November 2011
Accepted 25 November 2011
Available online 6 December 2011
Keywords:
a-MnOx
EQCM
Supercapacitor
Potential window
Cycle stability
© 2011 Elsevier Ltd. All rights reserved.
1. Introduction
[1,2,4]. In comparing with EDLCs, higher specific capacitance can
be achieved by using redox-active materials and supercapacitors of
Supercapacitors have been widely recognized as a complemen-
tary charge storage assistance to rechargeable batteries, fuel cells,
wind power, and solar cells because of their high charge/discharge
efficiency, high power properties, and very long cycle life [1,2].
Applications of this kind of devices become more important to
the cases which require transient but high/peak power pulses of
charge/discharge for time-dependent usage [3] because superca-
pacitors with a much lower equivalent series resistance (ESR),
in comparison with rechargeable batteries, can capture and store
pulse energy during the braking process, leading to a much higher
energy efficiency [3,4].
In general, supercapacitors can be divided into three cate-
gories according to the charge storage mechanism of devices. The
so-called electrical double layer capacitors (EDLCs) use the elec-
trostatic charge separation at the electrolyte–electrode interface,
showing the importance of increasing the electrolyte-accessible
surface area [5–7]. Pseudocapacitors with a mechanism similar
to rechargeable batteries employ fast superficial redox reactions
the asymmetric type are developed to further enhance the energy
density of devices [8]. Moreover, supercapacitors of the asymmetric
type can be further divided into two forms, i.e., double-layer/redox
and redox/redox types since asymmetric indicates that the active
materials in the anode are different from that in the cathode.
Accordingly, developing electrode materials promising to EDLCs
or pseudocapacitors are equally important to the application of
asymmetric type.
Crystalline ruthenium dioxide (RuO2) in the hydrous form has
been widely recognized as the best electrode materials for pseudo-
capacitors because of its high specific capacitance, good electronic
conductivity, and cycle stability [9,10]. However, due to cost and
environment considerations, other transition metal oxides, such as
MnOx [11], CoOx [12], VOx [13], Fe3O4 [14], Ni–Co oxides [15,16],
and WO3 [17] are recommended to replace RuO2. Among these
pseudocapacitor electrode materials, manganese oxides in both
amorphous and crystalline forms are the most promising candi-
date because of their excellent reversibility and acceptable specific
capacitance [11,18] although the cycle stability of MnOx is not sat-
isfactory [19]. Fortunately, the cycle stability of amorphous MnOx
(denoted as a-MnOx) was found to be significantly improved by
∗
introducing suitable anions (e.g., HPO4 or HCO3−) into the elec-
Corresponding author at: Department of Chemical Engineering, National Tsing
2−
Hua University, 101, Section 2, Kuang-Fu Road, Hsin-Chu 30013, Taiwan.
Tel.: +886 3 573 6027; fax: +886 3 573 6027.
trolyte [20] or by adding Mn3O4 into the active material [21].
In developing pseudocapacitive materials, gaining insights on
the mechanisms of charge storage and performance degradation
E-mail address: cchu@che.nthu.edu.tw (C.-C. Hu).
ISE Member.
1
0013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2011.11.106

Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
125
is of both academic and practical interests. For example, Na+ and
H3O+ cations from aqueous electrolytes were proposed to involve in
the charge storage process of MnOx, evidenced by X-ray photoelec-
tron spectroscopic (XPS) [18] and electrochemical quartz crystal
−
microbalance (EQCM) [22] analyses. Moreover, HPO42− and HCO3
were proposed to reduce the dissolution of Mn2+ by forming indis-
soluble Mn2+ salts [20]. However, the potential region responsible
for the formation of Mn2+ and the deposition rate of the indissol-
uble Mn2+ salts are important to improve the oxide degradation
issue. In this work, the voltammetric and quartz crystal microbal-
ance responses were employed to gain the degradation information
2−
of a-MnOx−in neutral aqueous media. The introduction of HPO4
and HCO3 on the formation of indissoluble Mn2+ salts and the
improvement of a-MnOx cycle stability was also investigated.
2. Experimental
Amorphous MnOx was deposited onto an Au/Ti-sputtered
quartz crystal working electrode in the EQCM cell under the
potentiostatic mode at 1.0 V for 1 min. The size of the whole
EQCM electrode is 0.205 cm2. The diameter of the quartz crystal
is 13.7 mm. The diameter of the gold electrode for a-MnOx deposi-
tion is 5.11 mm. The reference crystal has an oscillation frequency of
8.000 MHz and the working crystal oscillation frequency is between
7.995 and 7.950 MHz. The mass of a-MnOx was simultaneously
monitored by the quartz crystal microbalance. After deposition,
the electrode was washed with deionized water several times and
dried in an oven at 85 ◦C overnight. The deposition bath contained
5 mM Mn(CH3COO)2·4H2O (Showa EP, Japan). All solutions used
in this work were prepared with deionized water produced by a
reagent water system (Milli-Q SP, Japan) at 18 Mꢀ cm. All reagents,
not specified, without further purification are Merck, GR. Solution
temperature was maintained at room temperature (under an air
condition at 25 ◦C).
Cyclic voltammetric (CV) and EQCM responses were conducted
by an electrochemical analyzer (CHI4051a, CH Instruments Inc.) in
a solution containing 10 mM Na2SO4 with/without Na2HPO4 (J.T.
Baker AR, USA) or NaHCO3 (Hanawa EP, Japan). The reference and
counter electrodes were Ag/AgCl (Argenthal, 3 M KCl, 0.207 V vs.
SHE at 25 ◦C) and a platinum wire (99.95%, 0.5 mm in diameter and
5 cm in length), respectively. The surface morphologies of a-MnOx
before and after CV cycling were examined by a field-emission
scanning electron microscope (FE-SEM, Hitachi S4800-type I).
Fig. 1. (a) Current–time curve of a-MnOx deposited at 1.0 V for 1 min from a solu-
tion containing 5 mM Mn(AcO)2·4H2O. (b) The corresponding (1) first derivative of
mass–time curve and (2) mass–time curve for the deposition of a-MnOx.
increases at a high rate and then, increases linearly with deposition
time beyond 13 s. In order to correlate the deposition current and
the deposition rate, the first derivative of mass–time curve is added
as curve 1 in Fig. 1b. From a comparison of Fig. 1a and curve 1 in
Fig. 1b, the rate of a-MnOx deposition and the current–time curve
are very similar. Hence, the current efficiency of a-MnOx deposi-
tion is expected to be high. This speculation is also supported by
the result that the mass of a-MnOx has been found to be directly
proportional to the passed charge [24]. Based on the Faraday law,
the theoretical mean deposition rate of a-MnOx is 50 ng/s when the
total charge of this 60-s deposition process is 4 mC and m/n = 0.35.
(n + m)Mn2+ + 2(n + m)H2O → (MnO2)n(MnOOH)m
3. Results and discussion
+ (4n + 3m)H+ + (2n + m)e
(1)
3.1. EQCM and morphology studies of as-deposited a-MnOx
This result indicates that the mean oxidation state of a-MnOx is 3.74.
However, due to the hydrous nature of as-deposited a-MnOx, the
amount of Mn species in the deposit should be lower. Accordingly,
the mean oxidation state of a-MnOx should be lower than 3.74,
which was reported to be 3.43 [25].
Fig. 1 shows the typical current–time and mass–time curves
during the electrodeposition process of a-MnOx. In Fig. 1a, upon
imposing the potential at 1.0 V, the deposition current rises to a high
value due to the double-layer charging effect and the extremely
high concentration gradient of Mn2+ at the electrode–electrolyte
interface since Mn2+ at the electrode surface must be effectively
consumed and the concentration of Mn2+ decreases immediately
to 0 when the electrode potential is stepped to 1.0 V. The later
statement is supported by the fact that the applied overpoten-
tial for depositing a-MnOx in this work is about 0.6 V because the
onset deposition of a-MnOx from Mn(CH3COO)2·4H2O was found
to be 0.4 V and the deposition process was under diffusion con-
trol at potentials positive than ca. 0.75 V [23]. The gradual decrease
in current to 46 ␮A with prolonging the deposition time to 60 s is
attributed to the continuous decrease in the concentration gradient
of Mn2+ at the electrode–electrolyte interface with the deposi-
tion time [11]. On curve 2 of Fig. 1b, the mass of a-MnOx initially
Fig. 2a and b shows the SEM images of a bare EQCM electrode
and that deposited with a-MnOx, respectively. Clearly, the mor-
phology of sputtered Au/Ti is relatively dense and compact with
clear grain boundaries. The as-deposited a-MnOx is a highly porous
film with a uniform nanostructure of a three-dimensional (3D)
network. In addition, no obvious grain boundaries can be found
on this deposit and several pores are covered with sheet-like a-
MnOx films. The mean diameter of these MnOx nanowires is about
10 nm meanwhile the diameter of pores is about 30–50 nm, favor-
able for the electrolyte penetration during the charge/discharge
process for the supercapacitor application. The morphology of
this as-deposited a-MnOx film resembles that prepared by cyclic
voltammetry between 0.4 and 1.0 V for 30 cycles in our previous
work [25], which generally showed a higher specific capacitance.

126
Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
Fig. 3. (a) Cyclic voltammograms and (b) ꢁm–E curves of a-MnOx measured
between −0.2 and 1.2 V at 25 mV s−1 for 5 cycles in 10 mM Na2SO4. The solid lines
stand for the positive scans and the dash lines represent the negative scans.
3.2. EQCM and morphology studies of a-MnOx in Na2SO4 without
Na2HPO4 and NaHCO3
Fig. 3a and b respectively shows the first 5 cycles of CV and ꢁm–E
(mass variation against potential) curves measured in 10 mM
Na2SO4 at 25 mV s−1 between −0.2 and 1.2 V for a-MnOx. In Fig. 3a,
the first CV cycle shows the typical capacitive behavior of a-MnOx
in Na2SO4 [11]. This CV curve demonstrates a rectangular shape
in the whole potential range although two pairs of redox peaks
are respectively visible in the potential ranges from −0.2 to 0.2 V
and from 0.8 to 1.2 V. The corresponding ꢁm–E curve (see cycle
1 in Fig. 3b) can be qualitatively divided into three regions from
the mass–charge ratio information, which correspond to the three
potential ranges: −0.2 ∼ 0.2 V, 0.2 ∼ 0.8 V, and 0.8 ∼ 1.2 V. In the
most negative potential range, although there is a pair of broad
redox peaks on cycle 1 in Fig. 3a, the mass variation in this poten-
tial region is minor and reversible on both positive and negative
sweeps. This result suggests that the redox transition in this poten-
tial region is reversible, which mainly involves H+ (and minor
H3O+/Na+) because the mass–charge ratio is much lower than 19
and 23 for H3O+ and Na+, respectively. The mass–charge ratio (MCR)
is defined as follows [22]:
Fig. 2. Surface morphologies of (a) a bare EQCM electrode and (b) surface and
(c) cross section images of an a-MnOx-coated EQCM electrode. The a-MnOx was
deposited at 1.0 V for 1 min from a solution containing 5 mM Mn(AcO)2·4H2O.
From Fig. 2c, a-MnOx deposited onto the Au/Ti-sputtered quartz
electrode is an extremely thin film (ca. 33 nm) and the 132-nm
film is the Au/Ti layer. These statements are supported by the EDX
analyses (not shown here). Based on the mass and thickness of a-
MnOx, the density of this a-MnOx film is about 4.64 g cm−3 which is
lower than the theoretical densities of ␣-MnO2 (5.026 g cm−3 [26])
and Mn3O4 (4.856 g cm−3 [26]), consistent with the morphology
results.
MCR ≡ dM/(dQ/−96, 500)
(2)
where dM is the differential mass change of the electrode (g) and
dQ is the differential charge change (coulomb). MCR represents the
mass gain (or loss) when 1 mole of electrons have been passed.
Based on the potential-pH diagram of Mn [27], the redox cou-
ple is reasonably assigned to be Mn(III)/Mn(II) between −0.2 and
0.2 V. Thus, the redox transition of Mn(III)/Mn(II) involving cation

Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
127
Fig. 4. SEM images of a-MnOx with CV cycling for (a) 3, (b) 7, (c) 9, and (d) 10 segments between −0.2 and 1.2 V at 25 mV s−1 in 10 mM Na2SO4.
expelling/intercalating from/into a-MnOx can be simply expressed
as Eq. (3):
Mn(O−C+)2,surface, MnO2−C+surface, and MnO2,surface). Since the mass
loss in this potential region is negligible, Mn(O−C+)2,surface reduced
from MnO2−C+
on a-MnOx should not be responsible for
the dissolution of a-MnOx. Note that the ideal capacitive per-
formance of a-MnOx between 0.2 and 0.8 V is attributed to the
partial oxidation/reduction of a-MnOx, which also involves cation
expelling/intercalating from/into the oxide [18,22,25]:
surface
Mn(O−C+)2,surface ↔ MnO2−C+surface + C+ + e
(3)
where the major species of C+ is H+ with minor amount of
Na+ and H3O+ [18,22]. It should be mentioned here that the
major Mn species within a-MnOx should be Mn(IV) from sev-
eral previous XPS analyses [18,25] while the pseudocapacitance
of a-MnOx mainly comes from the superficial redox species (i.e.,
(MnO2−C+)n,surface ↔ MnO2(MnO2−C+)n−1,surface + C+ + e
(4)
Fig. 5. The fifth cycle of (a and b) cyclic voltammograms and (c and d) ꢁm–E curves of a-MnOx measured (1) between 0 and 1.0 V, (2) between −0.2 and 1.0 V, (3) between
−0.2 and 0.8 V, and (4) between −0.2 and 1.2 V at 25 mV s−1 in 10 mM Na2SO4. The solid lines stand for the positive scans and the dash lines represent the negative scans. (e)
The corresponding mass–charge plot of a-MnOx measured between 0 and 1.0 V at 25 mV s−1 in 10 mM Na2SO4.

128
Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
The idea for the partial oxidation/reduction of a-MnOx in this poten-
tial region comes from the Synchrotron XRD and X-ray absorption
near-edge structure (XANES) analysis [22]. Clearly, the mass loss
is significant and the rate of mass recovery in this potential region
of the negative sweep is higher than the rate of mass loss in the
same potential region of its preceding positive sweep. Thus, the
electrochemical reactions involve not only the redox transition of
superficial Mn species but also the dissolution and redeposition
of a-MnOx (see below). In the most positive potential range, irre-
versible dissolution is found on both positive and negative sweeps.
Moreover, the rate of a-MnOx dissolution on the negative sweep is
higher than that on its preceding positive sweep. This phenomenon
is also visible on cycles 2–5 in Fig. 3b, revealing that dissolution in
the negative sweep is enhanced by the dissolution occurring on
its corresponding positive sweep. This effect results in an increase
of the dissolved Mn(IV) concentration at the electrode–electrolyte
vicinity, favoring the redeposition of a-MnOx between 0.8 and 0.2 V
on the negative sweep. Accordingly, the rate of mass recovery
between 0.8 and 0.2 on the negative sweep is higher than the rate
of mass loss in the same potential region on its preceding positive
sweep. Therefore, the electrochemical reactions in this potential
region are very complicated, at least, involving two reactions (i.e.,
Eqs. (4) and (5)):
accelerated by the reduction reaction corresponding to the shoul-
der. The above evidences reveal that the electrochemical processes
between 1.0 and 0.2 V on the negative sweep include redox tran-
sition of superficial Mn species and redeposition of a-MnOx. Fifth,
the voltammetric currents between 0.2 and −0.2 V on the negative
sweeps become higher and higher with the cycle number in Fig. 3a.
However, overlapping of negative (dash lines) and positive (solid
lines) sweeps between −0.2 and −0.05 V from the negative sweep
of cycle 4 (or cycle 3) to the positive sweep of cycle 5 (or cycle 4)
and the horizontal ꢁm–E response from −0.2 to 0.3 V on the posi-
tive sweep of cycle 5 in Fig. 3b suggest that oxide redeposition does
not occur significantly in this potential range.
In order to confirm the dissolution process, Fig. 4a–d shows
the typical SEM images of four EQCM electrodes deposited with a-
MnOx, which were potentiodynamically swept between −0.2 and
1.2 V for 3, 7, 9, and 10 sweeps of CV, respectively. Here, the starting
potential of all CVs was set at −0.2 V and every CV cycle contains
positive (from −0.2 to 1.2 V) and negative (from 1.2 to −0.2 V)
sweeps. A comparison of Fig. 2b and 4a, the morphology of an as-
deposited a-MnOx is similar to that of the a-MnOx with 3 sweeps
of CV (the electrode potential is interrupted at 1.2 V). This result is
due to the minor mass loss in the first two cycles of CV between
−0.2 and 1.2 V (see curves 1 and 2 in Fig. 3b). On the other hand,
several pin holes are visible in Fig. 4b, revealing the obvious disso-
lution of a-MnOx after 7 sweeps of CV. This phenomenon is more
obvious in Fig. 4c, due to the accumulative mass loss in previous
two CV cycles (from sweep 5 to sweep 8) and the dissolution of
MnO2(MnO2−C+)n−1,surface + 2ıH2O
→ MnO2(MnO2−C+)n−1−ı,surface + ıMn4+ + 4ıOH− + ıC+ + ıe
(5)
Note here that the dissolved species is Mn(IV) species although the
precipitation of MnHPO4 or Mn(HCO3)2 (i.e., the insoluble salts of
Mn2+) were proposed to inhibit the dissolution of a-MnOx [20]. The
exact reasons responsible for this contradiction are not clear while
the formation of insoluble Mn(HCO3)4−2y(CO3)y and the adsorption
2−
of HPO4
were found to effectively inhibit the dissolution of a-
MnOx polarized at highly positive potentials in Section 3.3.
From an examination and detailed comparisons of cycles 1–5
in Fig. 3a and b, some additional features have to be mentioned.
First, an irreversible process due to the oxygen evolution reac-
tion (OER) is clearly found at/beyond the third cycle from Fig. 3a.
This result indicates that the repeated potential cycling promotes
the electrocatalytic activity of a-MnOx for the OER. This effect not
only causes the irreversible reaction of oxygen evolution but also
enhances the dissolution of a-MnOx due to oxygen evolution (see
ꢁm–E responses on the positive sweep of cycle 3 between 1.0 and
1.2 V). As a result, the mass loss in this potential region on both pos-
itive and negative sweeps is very obvious when the cycle number is
equal to/above 3 from Fig. 3b. Second, the mass loss (with the excep-
tion of cycle 1) becomes more and more obvious with the cycle
number and the net mass changes for cycles 1–5 are equal to +29.5,
−5.8, −214.7, −622.6, and −1160.5 ng, respectively. This result sup-
ports the statement that the dissolution of a-MnOx is enhanced by
the dissolution process occurring in previous cycles, probably due
to the more active a-MnOx freshly redeposited onto the electrode.
Third, on the negative sweeps of CV and ꢁm–E curves, cathodic
currents and mass recovery commence at the same potential (ca.
1.0 V). This result suggests that some MnOx can be immediately
redeposited when oxygen evolution is ceased and MnO2 (not dis-
solved in the electrolyte) is reduced. This phenomenon is not visible
when the upper potential limit is set at 1.0 V (see below). Fourth, a
cathodic peak centered at 0.68 V appears on cycle 3 and becomes
more and more obvious with increasing the cycle number in Fig. 3a.
This peak suggests the reduction of Mn(IV) and redeposition of a-
MnOx. Since there is a shoulder in the same potential range on the
negative sweeps of cycles 3–5 in Fig. 3b and since the slope of mass
recovery is higher after this shoulder, redeposition of a-MnOx is
Fig. 6. SEM images of a-MnOx with CV cycling (a) between −0.2 and 0.8 V, (b)
between −0.2 and 1.0 V for 5 cycles at 25 mV s−1 in 10 mM Na2SO4.

Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
129
Fig. 7. (a and b) Cyclic voltammograms and (c and d) ꢁm–E curves of a-MnOx measured between −0.2 and 1.2 V at 25 mV s−1 for 5 cycles in 10 mM Na2SO4 with the addition
of 0.3 mM (a and c) NaHCO3 and (b and d) Na2HPO4.
a-MnOx from 1.0 to 1.2 V in sweep 9. The morphology of resul-
tant a-MnOx with 10 sweeps of CV is clearly different from that
of an as-deposited a-MnOx from a comparison of Fig. 2b and 4c.
This phenomenon is attributed to the accumulative mass loss from
sweep 5 to sweep 10 although certain amount of a-MnOx has been
significantly redeposited back.
2, suggesting that the lower potential limit of CV shows insignif-
icant influences on the redox transitions of superficial MnOx. On
the other hand, the negative shift of the lower potential limit of CV
from 0 to −0.2 V significantly promotes the capacitive currents on
the positive sweep from a comparison of curves 1 and 2 in Fig. 5a.
These capacitive currents are attributed to the reverse reaction of
the cathodic peak centered at ca. −0.17 V. The above statement is
supported by the very similar i–E responses on the positive sweeps
of curves 2 and 3 in Fig. 5a because the lower potential limit of CV
for both CV curves is the same (−0.2 V). Interestingly, for curves
1 and 2 with the same upper potential limit of CV (1.0 V), the i–E
responses on their negative sweeps approximately follow the same
trace and their capacitive currents are generally higher than that
on the negative sweep of curve 3 from 0.8 to 0.1 V. Based on all
the above results and discussion, the anodic charges accumulated
from 0.8 to 1.0 V will be consumed in the following whole negative
sweep meanwhile the cathodic charges spent from 0.1 to −0.2 V can
be recovered in the following whole positive sweep. Since the mass
loss for curves 1–3 in Fig. 5c is minor and negligible, a-MnOx pre-
pared by anodic deposition is stable between −0.2 and 1.0 V during
this 5-cycle test.
In order to gain more information about the dissolution and
redeposition of a-MnOx, the morphology of a-MnOx after 5 cycles
of CV within −0.2 ∼ 0.8 V and −0.2 ∼ 1.0 V are shown as Fig. 6a and
b, respectively. From a comparison of Figs. 2b and 6a, the morphol-
ogy of a-MnOx is not changed by the 5-CV cycling between −0.2
and 0.8 V. This result reveals the excellent stability of a-MnOx in
this potential range even though the presence of a reduction peak
on the negative sweep of curve 3 in Fig. 5a. When the potentials are
cycled between −0.2 and 1.0 V, the uniform 3D network structure is
still visible in Fig. 6b while the structure of original nanosheets cov-
ered on the pores become loose. Thus, the main network structure
of a-MnOx is undamaged while the sheet-like materials are rela-
tively unstable. Note that when CV was set between −0.2 and 1.2 V
(see Fig. 4d), there are some larger holes in the oxide matrix while
Fig. 5 shows the fifth cycle of CVs and the corresponding ꢁm–E
curves measured in 10 mM Na2SO4 at 25 mV s−1 within different
potential windows for a-MnOx. On curve 1 in Fig. 5a, as the poten-
tial window is set between 0 and 1.0 V, the i–E responses generally
show the typical capacitive behavior of a-MnOx although a shoul-
der is visible at potentials positive to 0.8 V on the positive sweep. In
Fig. 5d, the corresponding ꢁm–E responses show several features.
First, mass slightly decreases with the positive shift in electrode
potentials from 0 to ca. 0.2 V. In this potential range, the corre-
sponding electrochemical reaction is presumably ascribed to Eq.
(3). Second, the mass loss becomes very obvious with the positive
shift in electrode potentials from 0.2 to 0.8 V. From the potential-
pH diagram of Mn [27], the electrochemical reaction is reasonably
attributed to the Mn(III) oxidation to Mn(IV) (see Eq. (4)). The more
obvious mass change between 0.2 and 0.8 V suggests that H3O+ and
Na+ may involve mainly in the oxidation from 0.2 to 0.8 V because
the mass–charge ratio (MCR) in this potential window is around
19 (see Fig. 5e), which is close to the MCR of H3O+ and Na+. Third,
when electrode potentials are between 0.8 and 1.0 V on the positive
sweep, an obvious increase in mass is found while this phenomenon
is not clear from 1.0 to 0.8 V on the negative sweep. This mass-
increasing phenomenon, consistent with the anodic currents in
this potential region, is not attributable to the dissolution of MnOx.
Accordingly, in the same potential range on the negative sweep,
mass increase is invisible when voltammetric currents are switched
to be cathodic. Fourth, at potentials negative to 0.8 V on the negative
sweep, a relatively smaller slope is found in comparison with the
ꢁm–E responses in the same potential range on the positive sweep.
Similar results are obtained in the same potential region on curve

130
Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
some tiny nanoparticles are also visible. The above results indicate
that some a-MnOx originally filled in the pores dissolves in the elec-
trolyte, which can be redeposited back to the network structure,
according to the discussion for the dissolution and redeposition
phenomena in Fig. 3.
3.3. EQCM and morphology studies of a-MnOx in Na2SO4 with
Na2HPO4 and NaHCO3
Komaba et al. [20] proposed that a-MnOx dissolution was
successfully suppressed by adding small amounts of NaHCO3 or
Na2HPO4 into neutral aqueous electrolytes. This phenomenon can
be further clarified through the EQCM study. In this work, CV and
ꢁm–E curves were measured in the 10 mM Na2SO4 electrolyte con-
taining 0.3 mM NaHCO3 or Na2HPO4 and the initial 5-cycle results
are shown in Fig. 7. Fig. 7a and b shows the typical capacitive behav-
ior of a-MnOx in neutral media [11] although the upper limit of
−
CV is 1.2 V. This result reveals the powerful ability of HCO3 and
HPO4
2−
in depressing the a-MnOx dissolution in the highly posi-
tive potential region. The voltammetric currents of the redox peaks
at potentials positive than 0.65 V on both sweeps are monotonously
decreased with the continuous potential cycling with the absence
of the irreversible OER at potentials positive to ca. 1.0 V. On the
other hand, voltammetric currents at potentials negative to 0.3 V on
both sweeps are slightly increased with the CV cycling. The former
result suggests the formation of insoluble Mn salts on the a-MnOx
surface inhibiting the redox transition of oxyhydroxyl Mn species
(e.g., the Mn(IV)/Mn(III) transition). The latter result suggests that
the redox ability of the Mn(III)/Mn(II) couple is slightly enhanced
by potential cycling in aqueous media, which is commonly found
for oxides [16,28–30]. The above statements are supported by the
ꢁm–E results shown in Fig. 7c and d. From a detailed comparison of
Fig. 7a–d, several features have to be mentioned. First, the voltam-
metric responses of a-MnOx measured in Na2SO4 with 0.3 mM
NaHCO3 and Na2HPO4 are approximately the same (see Fig. 7a and
b). However, the mass of a-MnOx at −0.2 V significantly increases
with the cycle number of CV for the electrolyte containing NaHCO3
with the exception of the first cycle, while this phenomenon is
very minor for the electrolyte containing Na2HPO4. The above dif-
ference in the ꢁm–E responses may indicate different inhibition
Fig. 8. SEM images of a-MnOx with CV cycling between −0.2 and 1.2 V for 5 cycles
at 25 mV s−1 in 10 mM Na2SO4 with the addition of 0.3 mM (a) NaHCO3 and (b)
Na2HPO4.
0.4–0.8 V and 0.8–1.2 V, respectively, on the positive sweep may
−
mechanisms by using HCO3 and HPO42−. For HCO3−, an insolu-
2−
involve desorption and adsorption of HPO4 as well as the redox
ble salt (e.g., Mn(HCO3)4−2y(CO3)y, y is between 0 and 2) should
be deposited onto the substrate, which continuously increases the
deposit mass. For HPO42−, the specific adsorption of anions at the
deposit–electrolyte interface is believed to depress the a-MnOx dis-
solution. Accordingly, the largest mass increase is only 15 ng and
transitions of superficial oxyhydroxyl Mn species. Thus, the almost
constant and low slope on the negative sweep of the ꢁm–E curves
suggests the pure redox transitions of superficial oxyhydroxyl Mn
species. Fourth, the maximum mass difference on every CV cycle
decreases with the cycle number in Fig. 7c but not in Fig. 7d, reveal-
ing the stabilization of a-MnOx through adding dilute NaHCO3 and
Na2HPO4 although the effect of the former compound is more
obvious.
a slight decrease in mass to −3.8 ng at the end of cycle 5 due
2−
to the weak adsorption be−tween HPO4
and the superficial a-
MnOx species. Using HCO3 causes an obvious increase in mass
(50 ng), which is higher than using HPO42−. The presence of pre-
cipitated carbonates and adsorbed phosphates has been confirmed
through the EDX analyses (not shown here). Second, in Fig. 7c,
mass decreases with the positive shift in potentials from −0.2 to
ca. 0.8 V on the positive sweep and an obvious increase in mass
is found with the further positive shift in potentials. A relatively
lower rate in mass increase is found from 1.1 V on the positive
sweep to ca. 0.2 V on the negative sweep, then, a decrease in mass is
found from 0.2 to −0.05 V, and again a relatively lower rate in mass
decrease is found from −0.05 V on the negative sweep to ca. 0.2 V
on the positive sweep. The obvious increase in mass on the positive
sweep at potentials positive than ca. 0.8 V suggests the deposition
of insoluble salts which inhibit the electrochemical activation of
a-MnOx for the OER found in the absence of NaHCO3 and Na2HPO4
(see the 3rd–5th cycles in Fig. 3a). Third, in Fig. 7d, the lower rate
in mass changes is found on both sweeps at potentials negative
to ca. 0.4 V. However, the obvious mass decrease and increase in
Fig. 8a and b shows the SEM images of a-MnOx after the CV
cycling between −0.2 and 1.2 V in 10 mM Na2SO4 with 0.3 mM
NaHCO3 and Na2HPO4, respectively. Apparently, there are no big
holes in the network structure of both deposits. The surface mor-
phologies of both deposits are very similar to the as-deposited film
(see Fig. 2b). A comparison of all SEM images supports the state-
ment that the formation of insoluble Mn(HCO3)4−2y(CO3)y or the
2−
adsorption of HPO4 effectively inhibit the dissolution of a-MnOx
polarized at highly positive potentials although Komaba et al. [20]
proposed that manganese phosphate and carbonate insoluble in
water are responsible for this inhibition.
4. Conclusions
Amorphous manganese oxide (a-MnOx) anodically deposited at
1.0 V for 60 s were employed for the EQCM study (i.e., CV and ꢁm–E

Y.-H. Chu et al. / Electrochimica Acta 61 (2012) 124–131
131
responses) in 10 mM Na2SO4 without or with 0.3 mM NaHCO3
or Na2HPO4. The as-deposited a-MnOx showed a highly porous
morphology with a 3D network nanostructure. From the surface
morphology and EQCM study of a-MnOx, this oxide was unsta-
ble between 1.0 and 1.2 V in 10 mM Na2SO4 because of significant
dissolution. The repeated potential cycling promoted the electro-
catalytic activity of a-MnOx for the OER which further enhanced the
dissolution of a-MnOx due to gas evolution. The oxide dissolution
and gas evolution phenomena were successfully suppressed by the
formation of insoluble Mn(HCO3)4−2y(CO3)y or the adsorption of
HPO42− by adding 0.3 mM NaHCO3 or Na2HPO4 in the neutral elec-
trolyte. This merit enlarged the charge/discharge potential window
of a-MnOx for the supercapacitor application.
[4] B.E. Conway, Electrochemical Supercapacitors, Scientific Fundamentals and
Technological Applications, Kluwer Academic/Plenum, New York, 1999.
[5] H.Y. Lee, J.B. Goodenough, J. Solid State Chem. 144 (1988) 220.
[6] C.C. Hu, J.H. Su, T.C. Wen, J. Phys. Chem. Solids 68 (2007) 2353.
[7] S. Nomoto, H. Nakata, A. Yoshioka, H. Yoneda, J. Power Sources 97–98 (2001)
807.
[8] A.D. Pasquier, I. Plitz, J. Gural, S. Menocal, G. Amatucci, J. Power Sources 113
(2003) 62.
[9] C.C. Hu, K.H. Chang, M.C. Lin, Y.T. Wu, Nano Lett. 6 (2006) 2690.
[10] K.H. Chang, C.C. Hu, C.Y. Chou, Chem. Mater. 19 (2007) 2112.
[11] C.C. Hu, T.W. Tsou, Electrochem. Commun. 4 (2002) 105.
[12] C.C. Hu, T.Y. Hsu, Electrochim. Acta 53 (2008) 2386.
[13] C.C. Hu, C.M. Huang, K.H. Chang, J. Power Sources 185 (2008) 1594.
[14] S.Y. Wang, N.L. Wu, J. Appl. Electrochem. 33 (2003) 345.
[15] C.C. Hu, C.Y. Cheng, Electrochem. Solid State Lett. 5 (2002) A43.
[16] T.Y. Wei, C.H. Chen, H.C. Chien, S.Y. Lu, C.C. Hu, Adv. Mater. 22 (2010) 347.
[17] K.H. Chang, C.C. Hu, C.M. Huang, Y.L. Liu, C.I. Chang, J. Power Sources 196 (2011)
2387.
[18] M. Toupin, T. Brousse, D. Be’langer, Chem. Mater. 16 (2004) 3184.
[19] J.K. Chang, C.H. Huang, M.T. Lee, W.T. Tsai, M.J. Deng, I.W. Sun, Electrochim.
Acta 54 (2009) 3278.
[20] S. Komaba, A. Ogata, T. Tsuchikawa, Electrochem. Commun. 10 (2008)
1435.
[21] C.C. Hu, C.Y. Hung, K.H. Chang, Y.L. Yang, J. Power Sources 196 (2011) 847.
[22] S.L. Kuo, N.L. Wu, J. Electrochem. Soc. 153 (2006) A1317.
[23] Y.S. Chen, C.C. Hu, Y.T. Wu, J. Solid State Electrochem. 8 (2004) 467.
[24] Y.T. Wu, C.C. Hu, J. Electrochem. Soc. 151 (2004) A2060.
[25] C.C. Hu, C.C. Wang, J. Electrochem. Soc. 150 (2003) A1079.
[26] D.R. Lide, Handbook of Chemistry and Physics, CRC Press, 1992.
[27] M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions, National
Association of Corrosion Engineers, Houston, TX, 1966.
Acknowledgments
The financial support of this work, by the National Science Coun-
cil of ROC, Taiwan under contract nos. NSC 100-3113-E-006-001-,
NSC 98-2221-E-007-078-MY3, NSC 97-2221-E-007-078-MY3, and
the boost program from Low Carbon Energy Research Center of
NTHU, is gratefully acknowledged.
References
[28] C.C. Hu, Y.H. Huang, K.H. Chang, J. Power Sources 108 (2002) 117.
[29] C.C. Hu, K.H. Chang, T.Y. Hsu, J. Electrochem. Soc. 155 (2008) F196.
[30] F. Ataherian, K.-T. Lee, N.-L. Wu, Electrochim. Acta 55 (2010) 7429.
[1] R. Kötz, M. Carlen, Electrochim. Acta 45 (2000) 2483.
[2] J.W. Long, B. Dunn, D.R. Rolison, H.S. White, Chem. Rev. 104 (2004) 4463.
[3] J.R. Miller, A.F. Burke, Electrochem. Soc. Interface (Spring) (2008) 57.