Zinc deposition and dissolution in methanesulfonic acid onto a carbon composite electrode as the negative electrode reactions in a hybrid redox flow battery

Article abstract of DOI:10.1016/j.electacta.2011.04.111

Electrodeposition and dissolution of zinc in methanesulfonic acid were studied as the negative electrode reactions in a hybrid redox flow battery. Cyclic voltammetry at a rotating disk electrode was used to characterize the electrochemistry and the effect of process conditions on the deposition and dissolution rate of zinc in aqueous methanesulfonic acid. At a sufficiently high current density, the deposition process became a mass transport controlled reaction. The diffusion coefficient of Zn²⁺ ions was 7.5 × 10^-6 cm² s^-1. The performance of the zinc negative electrode in a parallel plate flow cell was also studied as a function of Zn²⁺ ion concentration, methanesulfonic acid concentration, current density, electrolyte flow rate, operating temperature and the addition of electrolytic additives, including potassium sodium tartarate, tetrabutylammonium hydroxide, and indium oxide. The current-, voltage- and energy efficiencies of the zinc-half cell reaction and the morphologies of the zinc deposits are also discussed. The energy efficiency improved from 62% in the absence of additives to 73% upon the addition of 2 × 10^-3 mol dm^-3 of indium oxide as a hydrogen suppressant. In aqueous methanesulfonic acid with or without additives, there was no significant dendrite formation after zinc electrodeposition for 4 h at 50 mA cm ^-2.

Full text of DOI:10.1016/j.electacta.2011.04.111

Electrochimica Acta 56 (2011) 6536–6546
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Zinc deposition and dissolution in methanesulfonic acid onto a carbon composite
electrode as the negative electrode reactions in a hybrid redox flow battery
∗
P.K. Leung, C. Ponce-de-León , C.T.J. Low, F.C. Walsh
Electrochemical Engineering Laboratory, Energy Technology Research Group, University of Southampton, Highfield, Southampton, SO17 1BJ, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
Electrodeposition and dissolution of zinc in methanesulfonic acid were studied as the negative elec-
trode reactions in a hybrid redox flow battery. Cyclic voltammetry at a rotating disk electrode was
used to characterize the electrochemistry and the effect of process conditions on the deposition and
dissolution rate of zinc in aqueous methanesulfonic acid. At a sufficiently high current density, the depo-
Received 24 January 2011
Received in revised form 26 April 2011
Accepted 27 April 2011
Available online 6 May 2011
2+
sition process became a mass transport controlled reaction. The diffusion coefficient of Zn ions was
−
6
2
−1
7
.5 × 10 cm s . The performance of the zinc negative electrode in a parallel plate flow cell was also
Keywords:
Dendrite
Electrodeposition
Flow battery
Methanesulfonic acid
Zinc
2+
studied as a function of Zn ion concentration, methanesulfonic acid concentration, current density,
electrolyte flow rate, operating temperature and the addition of electrolytic additives, including potas-
sium sodium tartarate, tetrabutylammonium hydroxide, and indium oxide. The current-, voltage- and
energy efficiencies of the zinc-half cell reaction and the morphologies of the zinc deposits are also dis-
cussed. The energy efficiency improved from 62% in the absence of additives to 73% upon the addition
−
3
−3
of 2 × 10 mol dm of indium oxide as a hydrogen suppressant. In aqueous methanesulfonic acid with
or without additives, there was no significant dendrite formation after zinc electrodeposition for 4 h at
−
2
5
0 mA cm
.
©
2011 Elsevier Ltd. All rights reserved.
1
. Introduction
Power generation sources need to develop energy storage capa-
electronegativity of zinc, its large negative potential in aqueous
medium and its widespread availability, the zinc redox couple has
been considered as the negative electroactive species in redox flow
batteries.
The large negative potential of the zinc redox couple allows a
high energy density when coupled with other electropositive redox
couples, such as chlorine/chloride [4,5], bromine/bromide [6,7], fer-
rocyanide/ferricyanide (hexacyanoferrate II/III) [8], ceric/cerous [9]
and oxygen/water [10]. The combination of zinc with these couples
facilitates a high energy density. During the charge–discharge cycle
of a redox flow battery, the electrodeposition of zinc:
bilities to be more efficient during off-peak hours [1]. Excess power
generation and transmission are underused every year due to
insufficient energy storage leading to problems, such as raised
volatility, reduced reliability and energy wastage [1,2]. The increas-
ing interest in renewables has drawn attention to the fact more
reliable, affordable, flexible and safer energy storage systems are
required.
Recently, redox flow batteries have emerged as possible systems
for energy storage. Some advantages over conventional energy
storage systems are that they can be installed anywhere, as stand-
alone systems and no special terrain or climate is necessary.
Compared to the conventional lead acid battery, redox flow bat-
teries are less costly to maintain. The modular nature of the redox
flow battery enhances its transportability and makes its construc-
tion and maintenance costs, lowest among other energy storage
systems [3]. Since 1970, a number of redox flow batteries have
been developed for applications in bulk energy storage, load-
levelling, strategic power supplies and electric vehicles. Due to the
charge
Zn(II) + 2e−
ꢀ
Zn
E
◦
= −0.76 V vs. SHE
(1)
discharge
is the reaction at the negative electrode.
In the case of a zinc–cerium hybrid redox flow battery, the pos-
itive electrode reaction is [11]:
charge
Ce(III) − e−
ꢀ
Ce(IV)
E
◦
≈ +1.28 − 1.72 V vs. SHE
(2)
discharge
The range of the cell potential difference for this pair of redox
couples depends on the type of electrolyte used. According to Plu-
rion Inc. (UK), the open-circuit cell voltage of a zinc–cerium flow
battery can be up to 2.4 V [12]. Recent attention has focused on
the application of aqueous methanesulfonic acid as an electrolyte
∗ Corresponding author. Tel.: +44 23 80598931; fax: +44 23 80597051.
E-mail address: capla@soton.ac.uk (C. Ponce-de-León).
0
013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2011.04.111

P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
6537
for redox flow batteries, as this electrolyte allows a high solubility
for a wide range of metal ions, including the highly electropositive
lead [13] and cerium ions [14]. Methanesulfonic acid has compara-
ble conductivity and is generally less corrosive than other common
electrolytes, such as sulfuric and hydrochloric acids [13].
trolled potentiostat using the General Purpose, Electrochemical
Software (GPES) Version 4.5.
2.3. Electrolysis in a flow cell
Although zinc electrodeposition has been practised for a long
time using acidic chloride [15–21] and sulfate [22–25] baths, there
have been few studies of zinc electrodeposition from methane-
sulfonic acid electrolytes [26]. Recent patents have suggested
that methanesulfonic acid can significantly decrease the dendritic
growth [10,27], thus increasing the life cycle in a zinc-based
flow battery. In contrast to the electroplating industry, where the
majority of the substrates are metals, carbon-based materials are
the commonest electrode materials used in redox flow batter-
ies [28–30], as in many cases they are chemically inert and do
not undergo dissolution and formation of oxides. In this work,
the electrochemistry of zinc in methanesulfonic acid was studied
via cyclic voltammetry at a glassy carbon rotating disk electrode.
In order to optimize the conditions for a zinc-based redox flow
battery, a range of electrodeposition and dissolution experiments
have been carried out in a parallel plate flow-cell using different
electrolyte compositions and operating parameters. A detailed dis-
cussion of zinc morphology and system efficiency is included in this
study.
Electrodeposition and dissolution of Zn in methanesulfonic
acid were carried out in a divided, parallel plate flow cell. A
cation-conducting Nafion® membrane (DuPont, NF117/H+) sepa-
rated the two compartments. Carbon polyvinyl-ester composite
BMC 18649, Engtegris GmbH, Germany) was used for both
(
the negative and the positive electrodes. Each electrode had
2
an area of 4.5 cm × 2 cm (9 cm ) with a gap of 2.0 cm between
the electrode surface and the membrane. The volume of each
3
electrolyte contained in separated tanks was 100 cm . Both pos-
itive and negative electrolytes were circulated at a mean linear
−1
flow velocity between 1.8 and 7.8 cm s
through the cell using
a peristaltic pump (Masterflex® Model 7553079, Cole-Parmer)
with high-pressure tubing (Masterflex® Norprene®, Cole-Parmer).
−3
The negative electrode compartment contained 0.5–2 mol dm
−3
zinc methanesulfonate in 0.5–3 mol dm
methanesulfonic acid
−3
while the positive electrode compartment contained 1.0 mol dm
methanesulfonic acid. The temperature of both electrolytes was
controlled at (293–343) ± 2 K by thermostat (Grant Model LV FG)
and circulated water through jackets fitted to each electrolyte
reservoir.
2
. Experimental details
At the negative electrode, zinc was electrodeposited at cur-
−
2
rent density in the range of 20–80 mA cm for 4 h. The electrode
was then left at open-circuit for 1 min followed by zinc dissolu-
tion using the same current density. The electrolysis during the
oxidation stopped when the zinc half-cell voltage shifted towards
more positive values than 0 V vs. Ag|AgCl to avoid the oxidation
of the substrate material. The reference electrodes were satu-
rated silver–silver chloride electrode, Ag|AgCl (ABB, Series 1400,
2.1. Chemicals
All chemicals used were analytical reagent grade from Alfa
Aesar (UK) and Sigma Aldrich (Germany); the solutions were
prepared with ultra-pure water (18 M ꢀ cm resistivity) from an
Elga water purification system. The solutions were purged with
a fast stream of dispersed nitrogen gas (The BOC Group, UK) for
−
3
0.1 mol dm
KNO3) placed at the entrance of the flow cell in
1
0 min to avoid interference from the oxygen reduction reac-
each channel. The electrochemical measurements were made by
BaSyTec (5 A/12 V, Germany) battery test system. The anode- and
cathode-potentials together with the cell potential difference were
continuously monitored throughout the experiment.
tion. Zinc(II) methanesulfonate solutions were produced by stirring
zinc carbonate basic (Alfa Aesar, UK, 99 wt.%) using a PTFE-coated
steel magnetic stirred bar (Fisherband, UK, 1 cm diameter, 4 cm
length) in 70% methanesulfonic acid (Sigma Aldrich, Germany).
The resulting zinc methanesulfonate solutions were colourless,
clean solutions with no precipitate. All electrolytic additives were
laboratory grade, supplied by Fluka (Austria) or Sigma Aldrich
2.4. Electrodeposition and microscopy
Zinc electrodeposition in the presence and absence of addi-
(
Germany).
−3
tives was carried out in an electrolyte containing 1.5 mol dm
−
3
zinc methanesulfonate and 1 mol dm methanesulfonic acid elec-
trolyte (60 cm ) in a 100 cm glass beaker. The applied current
3
3
2.2. Cyclic voltammetry
−
2
density was 50 mA cm for 4 h. Carbon polyvinyl-ester compos-
ite (Engtegris Inc., BMC 18649, Germany) electrodes of 8 cm × 1 cm
dimensions and with an exposed electrode area of 4 cm × 1 cm were
used as both cathode and anode. The electrodes faced each other
when fitted into a nylon holder with an interelectrode gap of 1 cm.
Before each experiment, the electrodes were polished using sili-
con carbide paper (grade P120), degreased with detergent then left
in an ultrasonic bath for 10 min in deionized water. The electrode
holder was immersed in a double walled water jacket container
that allows water circulation from the bath thermostat to maintain
Cyclic voltammetry was carried out in a divided, three-electrode
glass cell. The electrolyte volume contained in the working elec-
trode compartment was approximately 100 cm . The glass cell
was equipped with a water jacket connected to a Grant LV FG
water thermostat to control the electrolyte temperature in the
range of 20–60 C. The working electrode was a static glassy car-
bon disk electrode (area: 0.13 cm ) while the counter electrode
was a large area platinum mesh (projected area: 1 cm ). The work-
ing and counter electrode compartments were separated by a
3
◦
2
2
◦
®
+
the temperature of the solution constant, typically at 50 C.
Nafion membrane (Dupont, NF115/H ). The electrode potential
was measured against a saturated silver/silver chloride electrode
Throughout the experiment, the solution was stirred at 400 rpm
using an IKa yellowline® magnetic stirrer. After the electrodeposi-
tion and before taking the microscopic images, the samples were
rinsed with ultrapure water (Elga water purification system) for
−
ABB, Series 1400, 1.0 mol dm KNO ). For the cyclic voltammetry,
3
3
(
the potential was linearly swept from −0.8 to −1.4 V vs. Ag|AgCl at
−
1
a potential sweep rate in the range from 8 to 128 mV s . Cyclic
voltammograms were recorded over a wide range of electrolyte
1
0 s, air-dried and stored in a sample bag placed in a container
−3
with silica gel granules. The images of the zinc electrodeposits were
taken with a JEOL scanning electron microscope (model JSM 6500 F)
equipped with an energy-dispersive X-ray spectroscopic analyzer,
which was used to estimate the elemental composition.
composition including 10–80 mmol dm
zinc methanesulfonate
in 0.5 mol dm sodium methanesulfonate solution adjusted to pH
–4 with methanesulfonic acid. All electrochemical measurements
were made with an EcoChemie Autolab (PGSTAT20) computer con-
−3
0

6
538
P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
3
. Results and discussion
In this study, the overpotential for zinc electrodeposition was ca.
20 mV, as the nucleation potential on glassy carbon electrode is
−
3.1. Cyclic voltammetry
more negative than that on metallic electrode. For instance, Gomes
and da Silva Pereira [32] reported that the nucleation potential
of zinc electrodeposition was observed at ca. −0.8 V vs. SCE at a
stainless steel electrode from the sulfate bath in the absence of an
electrolytic additive. Such an overpotential was reported to have
important implications for the morphology of the deposit and the
current efficiency. Volmer and Weber [33] suggested that a larger
overpotential could lead to faster nucleation rate and prevent fur-
ther grain growth [33–37], which has been reviewed by Budevski
et al. [38]. Finer grain size and sometimes more uniform deposits
have been obtained at larger deposition overportentials, which was
found to be largely dependent on anions molecules [31,39,40], elec-
trolytic additives [36,41,42], temperature [31,43] and the influence
of pH on the formation of zinc complexes [44,45]. As suggested by
Morón et al. [46], the crystal size of zinc deposit was also influenced
by the exchange current density. Due to the reduced exchange cur-
rent density induced by the adsorption of electrolytic additives on
the electrode surface, compact and reflective coatings have been
obtained from an acid chloride bath.
Fig. 1a shows the cyclic voltammogram of Zn(II) ions using a
−1
glassy carbon electrode at 20 mV s
potential sweep rate in an
−3
electrolyte containing 0.01 mol dm zinc methanesulfonate and
−
3
0
.5 mol dm
sodium methanesulfonate adjusted at pH 4 with
methanesulfonic acid at 295 K. The potential was firstly swept from
−
0.8 V to −1.4 V vs. Ag|AgCl.
The voltammogram is similar to those reported in other acid
media. The cathodic scan shows the potential at −1.15 V vs. Ag|AgCl,
which is comparable to ca. −1.2 V vs. SCE reported in sulfate,
chloride and acetate electrolytes [31]. The sharp increase in cur-
rent reduction was due to the formation of fresh zinc nuclei,
and the further increase of the available deposition area. The
−
2
peak current density reached ca. 2.57 mA cm
at −1.22 V vs.
Ag|AgCl due to the reaction of Zn(II) ions already on the elec-
trode surface. After the peak, the current density decreased as
no more Zn(II) ions were present at the surface and new Zn(II)
species had diffused from the bulk of the electrolyte. Since the
standard electrode potential of zinc deposition is more nega-
tive than hydrogen evolution, hydrogen gas could evolve by the
cathodic reduction of hydrogen ions on the newly formed zinc
nuclei. Although the cyclic voltammogram does not show the
evolution of hydrogen gas, a fraction of the current is proba-
bly due to proton reduction. The scan was reversed at −1.4 V vs.
Ag|AgCl in order to avoid hydrogen evolution at more negative
potentials.
In terms of electron transfer, the zinc deposition reaction in
a methanesulfonate bath is considered to be a quasi-reversible
reaction, as the separation of peak potentials was 160 mV. Lit-
erature reports show that the separation of peak potentials for
the Zn(II)/Zn(0) reaction in chloride and sulfate electrolytes was
ca. 100 mV and 150 mV, respectively, in an electrolyte containing
−
3
−3
0.01 mol dm Zn(II) ions and 0.5 mol dm Na(I) ions. The results
in this paper suggest that zinc electrodeposition from methanesul-
fonaic bath shows a similar electrochemical behaviour to that from
a sulfate bath. Moreover, the ratio of anodic to cathodic charge
was found to be consistent with the previous findings in sulfate
bath. The charge ratio between the anodic and cathodic processes
in methaneulfonate bath was 82% while the reported value in a sul-
fate bath was 81% [47]. This suggests that some of the charge during
zinc electrodeposition was used for hydrogen evolution.
While sweeping the potential towards the positive direction,
zinc ions were still being reduced until the crossover potential at
1.16 V vs. Ag|AgCl on the foot of the stripping peak. At −1.14 V vs.
−
Ag|AgCl, solid zinc was no longer thermodynamically stable on the
electrode surface and the metallic zinc deposited on the electrode
oxidizes very quickly as the oxidation process was not diffusion
controlled and a sharp high current peak was observed. The for-
ward and reverse scans form a nucleation loop between −1.16 V
and −1.18 V vs. Ag|AgCl. The slightly more negative potential during
Zn(II) reduction observed during the forward scan shows that the
deposition of zinc on the less-conductive glassy carbon substrate
required slightly more energy than that on the freshly deposited
zinc surface.
Plotting the logarithm of the current density vs. the poten-
−
1
tial range of zinc nucleation, a Tafel slope of −52 ± 2 mV decade
was obtained which is almost the double the magnitude of the
−
1
theoretical calculation of −30 mV decade
for a two electron
transfer reaction. The value agrees with other works reported
in the literature: Ibrahim [43] reported that the Tafel slope was
7
(b)
(
a)
6
4
2
0
2
4
6
5
4
3
2
1
0
-
-
-1.4
-1.3
-1.2
-1.1
-1.0
-0.9
-0.8
0
2
4
6
8
10
1 0.5
12
-
Potential vs. Ag|AgCl / V
Square root scan rate / ( V s )
−
3
−3
Fig. 1. Electrochemistry of zinc electrodeposition at a glassy carbon electrode in methanesulfonic acid. Electrolyte: 0.01 mol dm Zn(CH3SO3)2 in 0.5 mol dm NaCH3SO3
−
1
◦
adjusted at pH 4 with CH3SO3H. (a) Cyclic voltammogram, from −0.80 to −1.40 V vs. Ag|AgCl at 20 mV s and 20 C. (b) Randles-Sevcik plot, using potential sweep rates of
−1
8
, 16, 32, 64 and 128 mV s
.

P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
6539
−
1
−3
−3
5
6 mV s at 0.09 mol dm zinc in 0.2 mol dm sodium sulfate. In
As shown in Fig. 2b, a linear relationship between the limiting cur-
rent density with the square root of the rotation rate is observed
and follows the Levich equation.
the investigation of Guerra in zinc sulfate, the Tafel slope was ca.
0 mV decade in neutral pH solution and ca. −30 mV decade
−1
−1
6
in highly acidic electrolyte. The smaller Tafel slope (high cathodic
current density) was due to the surface condition, as the hydrogen
adsorbed on the newly formed zinc nuclei reacted. This resulted in
the combined cathodic reactions of hydrogen evolution and zinc
deposition, leading to a sharp increase in current density [48].
Fig. 1b shows a plot of the cathodic peak current densities vs.
the square root of the potential sweep rate. For a mass transport
controlled reaction, a linear increase of the peak current density
with the square root of the potential sweep rate can be predicted
by the Randles-Sevcik equation:
jlim = 0.620zFD2/3ω1/2v1/6
c
(4)
where jlim is the limiting current density (A cm−2), F is Faraday’s
−
1
constant, ω is the rotation rate of the disk (rad s ) and v is the kine-
2
−1
matic viscosity of the electrolyte (cm s ). The linear behaviour
indicates that the reduction of Zn(II) ions is under complete mass
transport control. By using the gradient of Fig. 2b, a diffusion
coefficient for zinc ions of 7.8 × 10 cm s was estimated. This
value is slightly larger than the one obtained from the Randles-
Sevik equation as possibly both hydrogen adsorption and oxygen
reduction reactions are more significant under the mass transport
controlled regime and contribute to a higher degree to the observed
limiting current than in the cyclic voltammetry experiment. In
this study, the average diffusion coefficient of zinc ion obtained
−
6
2
−1
jp = 2.69 × 10 z1.5D0.5v0.5
where jp is the peak current density, c is the zinc ion concentration
in the electrolyte, z is the number of electrons involved the elec-
trode process, v is the potential sweep rate and D is the diffusion
coefficient of the zinc ions. From the Randles-Sevcik plot, the dif-
5
c
(3)
−
3
from the Levich and Randles-Sevik equations in 0.01 mol dm zinc
−
3
methanesulfonate and 0.5 mol dm sodium methanesulfonate at
−
6
2
−1
−6
2
−1
.
fusion coefficient for zinc ions was found to be 7.2 × 10 cm s
,
pH 4 is 7.5 × 10 cm s
−
6
2
−1
while Meas et al. [49] reported a value of 7.27 × 10 cm s in
hydrochloric acid. The value found here is comparable to previous
estimations for the diffusion coefficient of zinc ions reported in the
literature in other acid electrolytes: Yu et al. [31] calculated 2.1, 5.5
In the electroplating industry, high speed metal deposition is
usually required and increasing the mass transport is a common
strategy to provide a high current density and a fast electrodepo-
sition rate. In the literature, it has been reported that powdery and
burnt deposits can be avoided by increasing the limiting current
density via enhanced mass transport techniques [35,50]. Turbu-
lent flow and a high rate of cathode movement are often used for
industrial high speed zinc electrodeposition [51].
−6
2
−1
and 8.4 × 10 cm s in acetic, sulphuric and hydrochloric acids.
The current used for zinc electrodeposition in methanesulfonic acid
would be expected to be similar to that used in other common acids
but the less-corrosive nature of methanesulfonic acid and its abil-
ity to dissolve high levels of metal ions makes it more attractive for
energy storage applications.
3.2.2. Zinc ion concentration
Fig. 2c shows the linear sweep voltammogram of zinc elec-
trodeposition over a range of zinc concentrations of 10, 20, 40, 60
3
3
.2. Linear sweep voltammetry
−
3
−3
−3
and 80 × 10 mol dm at 600 rpm rotation speed in 0.5 mol dm
sodium methanesulfonate at pH 4 and 295 K. The electrode poten-
.2.1. Mass transport
As well as the charge-transfer process taking place on the
−
1
tial was swept from −1.0 to −2.0 V vs. Ag|AgCl at 20 mV s . As
the concentration of zinc ions increased, the nucleation potential
shifted towards positive values and changed linearly with the log-
arithm of Zn(II) ion concentration by 58 ± 2 mV per 10-fold change
in concentration. The magnitude of this potential shift is similar to
that obtained in sulfate electrolytes reported in the literature [43].
Fig. 2d shows the limiting current densities for zinc deposition as
a function of the concentration of the zinc ion. The limiting current
electrode surface, mass transport of the Zn(II) ions towards the
electrode contributes to the overall rate of the reaction. As the elec-
trode potential becomes gradually more negative during the linear
sweep voltammetry, the reduction of Zn(II) becomes dominated
by the rate at which the ions reach the electrode surface. Fig. 2a
shows a family of linear sweep voltammograms of Zn(II) reduction
on a glassy carbon rotating disk electrode at various speeds: 400,
−
2
9
00, 1600, 2500 and 3600 rpm. The electrolyte composition was
density of zinc electrodeposition increased by ca. 10 mA cm per
every 10 mM zinc ion concentration. The linear plot confirms the
fact that the system was under mass transport control according to
the Levich equation. The potential window of the limiting current
plateau decreases as the concentration of zinc increases suggesting
that hydrogen evolution was more favourable as more zinc was
deposited on the electrode surface.
the same as that used in the experiments of Fig. 1a and b. In order
to avoid excessive hydrogen evolution, the electrode potential was
swept from −1.0 V to −2.0 V vs. Ag|AgCl at 20 mV s . The reduction
of zinc was observed at the same potential as in the cyclic voltam-
mogtam, −1.2 V vs. Ag|AgCl, and well defined plateau of constant
current density where observed at potentials more negative than
−
1
−
1.4 V vs. Ag|AgCl. The constant supply of the zinc ions to the elec-
trode surface due to the rotation of the electrode at different rates
helps to maintain the Zn(II) ion concentration at the surface. An
increased rotation speed was found to shorten the plateau region,
leading to earlier hydrogen evolution.
3.2.3. Concentration of protons
Fig. 3a shows the linear sweep voltammogram of zinc elec-
trodeposition over a range of pH from 0 to 4 at the rotation speed
of 600 rpm. The electrolyte composition is the same as in Fig. 1.
The electrode potential was swept from −1.0 to −2.0 V vs. Ag|AgCl
The limiting current density for each rotating speed was mea-
sured at the mid-point of the plateau at −1.7 V vs. Ag|AgCl. The
supply rate of zinc ions to the cathode surface is necessary to equal-
ize with the rate of charge transfer. This is especially important
during the charge and discharge of a redox flow battery; if the
applied current density during the charge or discharge cycles is too
high compared to the limiting current density, hydrogen evolution
might take place instead of zinc electrodeposition or dissolution.
The evolution of hydrogen can also increase the electrolyte resis-
tance and reduce the effective plating area, which can result in
burnt deposit [50]. Under such conditions, the transport rate of zinc
ion to the surface of the negative electrode needs to be increased.
−
1
at 20 mV s . At higher pH values, the nucleation potentials were
shifted towards more positive values, which is due to the higher
conductivity of the electrolyte and to the increased concentrations
of protons. Large limiting current plateaus were observed at pHs 4
and 3 than at lower pHs 0, 1 and 2.
At increased electrolyte acidity from pH 2 to 0, the stability of
the zinc methanesulfonate complexes decreases and the reduction
of protons occur on the newly formed zinc nuclei. This results in an
undefined plateau at pH 0 and 1 and a slightly larger current density
plateau at pH 2. At the lowest pH zinc electrodeposition competes
with hydrogen evolution and the sharp increase in current density

6
540
P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
0
10
20
(b) 25
(a)
4
00 rpm
20
-
-
-
900 rpm
600 rpm
1
5
0
5
0
1
2
500 rpm
1
3
600 rpm
30
-
2.0
-1.8
-1.6
-1.4
-1.2
-1.0
0
5
10
15
20
25
-
1 0.5
Potential vs. Ag|AgCl, E / V
Square root of rotation rate / (rad s )
0
60
50
40
30
20
10
0
(d)
(c)
1
0 mM
-
-
-
-
-
-
10
20
30
40
50
20 mM
40 mM
60 mM
8
0 mM
60
-
2.0
-1.8
-1.6
-1.4
-1.2
-1.0
0
20
40
60
80
100
-
3
-3
Potential vs. Ag|AgCl, E / V
Zinc methanesulfonate concentration / ×10 mol dm
Fig. 2. Effect of rotation speed and zinc ion concentration on zinc electrodeposition in methanesulfonic acid at glassy carbon electrode (293 K). (a) The effect of rotation
speed on the cyclic voltammogram, (b) Levich plot, (c) the effect of zinc ion concentrations on the cyclic voltammogram and (d) the plot of zinc ion concentrations against
−
3
−3
.
the limiting current densities. Electrolyte: as in Fig. 1. Rotation speeds were 400, 900, 1600, 2500 and 3600 rpm. Zinc concentrations: 10, 20, 40, 60 and 80 × 10 mol dm
−
1
◦
The electrode was swept from −1.0 to −2.0 V vs. Ag|AgCl scanning at 20 mV s at 600 rpm and 20 C.
0
0
5
(a)
(b)
pH 4
pH 3
-
50
-
-
-
-
-
-
100
150
200
250
300
o
-
-
10
15
20 C
o
30 C
pH 2
o
4
5
6
0 C
0oC
0oC
pH 1
pH 0
-1.4
-20
-
25
-
2.0
-1.8
-1.6
-1.2
-1.0
-
2.0
-1.8
-1.6
-1.4
-1.2
-1.0
Potential vs. Ag|AgCl
Potential vs. Ag|AgCl / V
Fig. 3. Effect of electrolyte acidity and operating temperatures on zinc electrodeposition in methanesulfonic acid at glassy carbon electrode at 600 rpm. (a) Effect of electrolyte
acidity on the cyclic voltammogram, (b) effect of electrolyte temperature on the cyclic voltammogram. Electrolyte: as in Figs. 1 and 2. Solution acidity was pH 4, 3, 2, 1 and
◦
−1
◦
0
. Electrolyte temperature: 20, 30, 40, 50 and 60 C. The electrode was swept from −1.0 to −2.0 V vs. Ag|AgCl at 20 mV s and 20 C.

P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
6541
8
6
4
2
0
2
4
0
(
a)
(b)
(
iv)
-
2
(
ii)
(i)
iii)
-4
-
-
6
8
(
(iii)
(i)
(
ii)
i)
(iv)
-10
12
-14
-2.2
(
-
-
-
(
iv)
(
ii)
(
iii)
-1.4
-1.2
-1.0
-0.8
-2.0
-1.8
-1.6
-1.4
-1.2
-1.0
Potential vs. Ag|AgCl / V
Potential vs. Ag|AgCl / V
Fig. 4. Effect of electrolytic additive on cyclic voltammogram of zinc electrodeposition in methanesulfonic acid (a) at static electrolyte and (b) at 600 rpm. Electrolyte: as in
Figs. 1–3. Electrolytic additives were at 2 × 10− mol dm , (i) no additive, (ii) potassium sodium tartarate, (iii) tetrabutylammonium hydroxide and (iv) indium oxide.
3
−3
was due to the excessive hydrogen evolution. The competition of
the deposition of zinc and hydrogen evolution has been confirmed
by the reaction model reported by Cachet and Wiart [52], which
stated that the discharge of zinc ion is firstly inhibited by hydro-
gen adsorption followed by the hydrogen evolution on the surface
of the zinc deposit. At such highly acidic media, the zinc ion con-
centration on the electrode surface is relatively low and hydrogen
adsorption on the electrode is hence easy. In this case, the current
is mainly used for hydrogen evolution and the current efficiency of
zinc deposition decreases.
polarization of zinc ions reduction effectively. Since diffusion coeffi-
cient is a function of temperature, the limiting current densities will
◦
increase; every 10 C increase in the temperature of the electrolyte
−
2
lead to an increase in current density of ca. 5 mA cm . During the
zinc dissolution process (not shown) the charge of the zinc strip-
ping peak was higher at elevated temperature confirming that zinc
electrodeposition has been facilitated.
3.3. Effect of electrolytic additives on the cyclic voltammetry
The suppression of the hydrogen evolution reaction is neces-
sary to promote zinc electrodeposition. In the system used by
Plurion Inc., indium has been suggested as an additive to increase
the hydrogen overpotential giving a current efficiency almost 95%
compared to 70–80% without the additive [9,53]. Other finding sug-
gests that additives, such as Na-benzylnicotinate [54], polyethylene
glycol (PEG) [22,54], cetyltrimethyl ammonium bromide (CTAB)
Electrolytic additives are often employed in the plating indus-
try to obtain durable, uniform and more compact coatings as well
as to decrease the rate of undesired reactions. Fig. 4a shows the
cyclic voltammogram for zinc electrodeposition with the addi-
tion of three selected electrolytic additives: tetrabutylammonium
hydroxide, potassium sodium tartarate and indium oxide at a glassy
carbon static disk electrode. The concentration of the additives
[
32], tetrabutylammonium bromide (TEBACl) [55], perfluorosurfac-
tants [56], tartaric acid [57], succinic acid [57], thallium [9], indium
12], cellulose [58] are capable of suppressing the hydrogen evolu-
in all cases was 2 × 10−3 mol dm−3 in an electrolyte containing
0.01 mol dm−3 zinc methanesulfonate and 0.5 mol dm−3 sodium
[
methanesulfonate adjusted to pH 4 with methanesulfonic acid at
295 K. The potential was firstly swept from −0.8 V to −1.5 V vs.
Ag|AgCl then reversed. The reduction and oxidation processes can
be seen as the charge and discharge cycles of a half-cell zinc bat-
tery, respectively. Larger anodic peaks were observed with the
addition of the additives except with tetrabutylammonium hydrox-
ide. The ratio between the anodic and cathodic (discharge/charge)
processes (Table 1) was 67%, whereas when indium oxide and
potassium sodium tartarate were used, the ratio was approximately
88%. The addition of these two additives represented an improve-
ment compared to the experiment in the absence of additives
(82.3%). Despite the presence of indium oxide, single cathodic and
anodic peaks were observed.
Linear sweep voltammograms in the presence of the three elec-
trolytic additives were also carried out on a rotating vitreous carbon
electrode disc at 600 rpm. Well-defined limiting current densities
plateaus were observed in all the experiments but each additive
presented different limiting current density value, this could be
due to the additive reacting or adsorbing on the newly formed
zinc deposit on the carbon electrode surface, thus decreasing the
above surface area. Nucleation potentials of zinc in the presence of
indium oxide and potassium sodium tartrate shifted to more pos-
itive values as shown in Table 1. This can be beneficial for battery
tion reaction. In the plating industry, buffer reagents, for example
H BO , are often incorporated to stabilize the pH and avoid exces-
3
3
sive hydrogen evolution [59].
3.2.4. Operating temperature
Fig. 3b shows the linear sweep voltammograms of zinc elec-
trodeposition over a range of electrolyte temperatures from 20 to
◦
6
0 C on a glassy carbon rotating disk electrode at 600 rpm. The elec-
trolyte composition range is the same as that in the experiments
described in Figs. 1 and 2. Both, the nucleation potential of zinc
ions and hydrogen evolution potential shifted towards more posi-
tive values as the temperatures increased. The change in electrode
potential obeys the Nernst equation, as a lower driving force for zinc
electrodeposition and hydrogen evolution reactions is required as
a higher temperature.
The electrode potential for the deposition of Zn(II) shifted 60 mV
◦
◦
from −1.19 V vs. Ag|AgCl at 20 C to −1.13 V vs. Ag|AgCl at 60 C. Sim-
ilar findings were observed by Zhang et al. [60] who reported that
the cathodic potential decreased by ca. 65 mV when the tempera-
◦
◦
ture increased from 30 C to 45 C during zinc deposition in a sulfate
electrolyte with 1-butyl-3-methylimidazolium hydrogen sulfate as
an electrolytic additive. Elevated temperature can hence reduce the

6
542
P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
efficiencies, as lower charge voltages can be expected due to the
reduced overpotential and zinc electrodeposition taking place prior
to hydrogen evolution. The charge efficiency slightly improved
except when tetrabutylammonium hydroxide was used. Although
larger overpotentials obtained with tetrabutylammonium hydrox-
ide result in lower voltage efficiency, such overpotentials were
reported to be useful for refining the grain size and hence brighten-
ing the electrodeposit by partly covering the electrode surface and
blocking the growth of the zinc crystals [61].
3.4. Zinc half-cell efficiencies in a parallel plate flow cell
Aiming to achieve the optimum electrolyte compositions and
operating conditions for a zinc-based redox flow battery, the
electrodeposition and stripping of zinc were carried out over a
wide range of concentrations, electrolyte flow rates, tempera-
tures, applied current densities and electrolytic additives in a
divided parallel flow cell. Unless indicated otherwise, the elec-
−
3
trolyte composition was at 1.5 mol dm
zinc methanesulfonate
−
3
in 1 mol dm methanesulfonic acid in the negative compartment
while the other compartment contained 1 mol dm methanesul-
fonic acid. Electrodeposition of zinc on a carbon polyvinyl-ester
composite substrate was carried out at 50 mA cm and 50 C for
h. Table 2 summarizes the estimated half-cell efficiencies in terms
of voltage, charge and energy at different operating conditions.
−
3
−
2
◦
4
3.4.1. Effect of operating parameters
Fig. 5a–c shows the effect of current density, temperature and
electrolyte flow rate, respectively, on the charge–discharge curves
of the zinc half cell reaction. In the absence of additives, the curves
in Fig. 5a show that the largest energy efficiency was 70%, obtained
−
2
at a low current density of 20 mA cm (Table 2). This high effi-
ciency was attributed to the reduced ohmnic drop and improved
voltage efficiency as the self-discharge of zinc became more signif-
icant while discharging at low current density. At a higher applied
current density, the charge efficiencies were found to be higher but
overall efficiency decreased. Similar findings were also reported by
Saba and Elsherief [62] over a range of sulfuric acid and Zn(II) ion
concentrations.
Fig. 5b shows the electrode potential vs. time at different tem-
peratures during reduction and oxidation of zinc. Higher energy
efficiency was obtained at lower operating temperatures, despite
the fact that the solution conductivity and the resulting voltage
efficiency increased with temperature. The charge efficiency signif-
◦
◦
icantly decreases from 94% at 20 C to 68% at 70 C. This is because
higher temperature can lead to stronger depolarizing effect on
hydrogen evolution than zinc deposition [63]; therefore less zinc
was deposited at such a high temperature. Fig. 5c shows that as
the mean linear electrolyte flow velocity increased from 1.8 to
−
1
5
.8 cm s , an increase in charge efficiency was observed. This could
be due to the fact that at a high mean linear flow velocity the dif-
fusion layer brcomes thinner and facilitates the transport of zinc
ions to the working electrode. Despite this, larger voltage drop was
also observed at a higher mean linear flow velocity. Therefore, the
increase of electrolyte velocity did not significantly improve the
overall energy efficiency of the zinc half-cell reaction.
3.4.2. Effect of zinc ions and acid concentration
Electrolyte compositions, especially the concentration of acid,
were found to be crucial for the flow cell performance. Energy
−
3
efficiency was as high as 62% at 1 mol dm methanesulfonic acid
−
3
compared to 41% at 3 mol dm methanesulfonic acid. Alfantazi and
Dreisinger reported that electrical conductivity increase linearly
with sulfuric acid concentration; therefore higher voltage efficiency
was obtained [64]. Due to the more dominant hydrogen evolution
reaction and faster corrosion rate at higher acid concentration, a

P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
6543
Table 2
Half-cell efficiencies of electrodeposition and stripping of Zn on a carbon polyvinyl-ester composite substrate from a methanesulfonic acid bath in a parallel plate flow cell.
−
2
◦
−1
−3
Unless indicated otherwise, the current density was 50 mA cm , deposition time 4 h, temperature 50 C, mean linear flow velocity 3.9 cm s . The solution was 1.5 mol dm
Zn(CH3SO3)2 and 1 mol dm CH3SO3H.
−3
Operating parameter
Zinc half cell
Zinc half-cell charge
Zinc half-cell discharge
Zinc half-cell efficiencies (%)
overpotential −Á (V)
potential vs. Ag|AgCl (V)
potential vs. Ag|AgCl (V)
Voltage
Charge
Energy
Current density (mA cm−2)
2
5
8
0
0
0
0.26
0.41
0.41
−1.17
−1.28
−1.30
−0.91
−0.85
−0.87
78
68
63
90
91
92
70
62
62
◦
Temperature ( C)
22
40
50
70
0.45
0.43
0.41
0.27
−1.32
−1.31
−1.28
−1.19
−0.87
−0.88
−0.85
−0.92
66
67
68
71
94
94
91
68
62
62
62
52
CH3SO3H concentration (mol dm−3)
0
1
2
3
.5
0.42
0.41
0.36
0.34
−1.26
−1.28
−1.24
−1.24
−0.84
−0.85
−0.88
−0.89
67
68
71
72
94
91
87
57
62
62
62
41
2+
−3
Zn concentration (mol dm )
0
1
1
2
.5
0.33
0.34
0.41
0.38
−1.27
−1.24
−1.28
−1.21
−0.94
−0.91
−0.85
−0.83
74
74
68
69
78
83
91
92
58
62
62
64
.5
Mean linear flow velocity (cm s−1)
1
3
5
7
.8
.9
.8
.8
0.33
0.41
0.37
0.41
−1.24
−1.28
−1.25
−1.27
−0.92
−0.85
−0.88
−0.86
74
68
70
67
86
91
92
92
63
62
64
62
Additive (×0.002 mol dm−3)
No additive
0.41
0.35
0.44
0.27
−1.28
−1.26
−1.37
−1.21
−0.85
−0.91
−0.93
−0.94
68
72
68
78
91
86
39
94
62
62
26
73
Potassium sodium tartarate
Tetrabutylammonium hydroxide
Indium oxide
significant decrease in charge efficiency was observed. The increase
in zinc ion concentration was found to facilitate the electrodeposi-
tion process indicated by the improvement of the charge efficiency
but this does not improve the energy efficiency [64].
obtained could be attributed to the large hydrogen overpoten-
tial of indium as an additive also observed in the Plurion Inc.’s
zinc–cerium system [12], hence larger current values can be used
for zinc electrodeposition. As presented in the voltammetry study
in Section 3.3, lower deposition overpotential was observed. Hence,
an improved voltage efficiency was expected in this electrolysis
study, which is consistent with the previous finding of Yang and
Lin [66] in their alkaline zinc-air battery.
The other two additives did not have a significant improvement
on the half-cell efficiencies probably due to their strong adsorp-
tion and inhibiting effect (levelling property). Ganne et al. reported
that some electrolytic additives could slow down the charge trans-
fer reactions, poisoning the active kink sites and increasing the
deposition overpotential [37]. For instance, the charge and energy
efficiencies in the presence of tetrabutylammonium hydroxide
were only 39% and 26%, respectively (Table 2). As a result, only
indium oxide was suitable to be used as an electrolytic additive in
a zinc-based hybrid redox flow battery.
The lower charge cell voltages observed as the concentration
of Zn(II) ions increases can be explained by the Nernst equation.
Since the dissolution of zinc back to Zn(II) species becomes more
difficult at higher Zn(II) ion concentration in the vicinity of the elec-
trode surface, lower discharge voltages were obtained. Despite this,
higher Zn(II) ion concentration can be advantageous for redox flow
battery as it allows the battery to charge at high current densi-
ties without mass transport limitations. Such a high concentration
ensures that enough Zn(II) ions exist at all state of charges. In addi-
tion, two moles of methanesulfonic acid are formed per mol of zinc
methanesulfonate during the electrodeposition reaction as shown
in Eq. (1) [65]. In order to minimize the hydrogen evolution and
facilitate zinc electrodeposition, higher Zn(II) ion concentration is
preferred in the negative electrolyte.
3
.4.3. Effect of electrolytic additives in the parallel flat plate flow
cell
Three electrolytic additives from different categories were
3.5. Coating morphologies
Fig. 6a–d shows the microstructures of the zinc electrode-
−3
−3
−3
−3
tested at a concentration of 2 × 10 mol dm . Both indium oxide
posits using different additives at 2 × 10 mol dm : (a) no
additive, (b) potassium sodium tartarate, (c) tetrabutylammo-
nium hydroxide and (d) indium oxide. In the absence of additive,
hexagonal-like crystalline structure was aligned plate-by-plate
parallel to the substrate (Fig. 6a). This morphology is typical of
zinc deposition and has already been observed in the literature
[64,68]. On the uppermost of such hexagonal plates, a relatively
rough, irregular plate-like structure was observed because of the
continued growth and the stacking of the hexagonal platelet on
the top of one another during the prolonged deposition for 4 h.
[
66] and potassium sodium tartarate [67] are inorganic additives,
while tetrabutylammonium hydroxide is a surface active agent. In
order to determine whether they are suitable additives for battery
performance, charge–discharge experiments were carried out in
the presence of these additives. The electrode potential during the
charge/discharge cycle is shown in Fig. 5d. With the addition of
indium oxide, the energy efficiency improved significantly from
6
2% without additive to 73% (Table 2). In this case, higher charge
and voltage efficiencies were obtained. The high charge efficiency

6
544
P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
a) -1.6
b) -1.6
-1.4
(
(
Zn(II) to Zn(0)
Zn(0) to Zn(II)
Zn(II) to Zn(0)
Zn(0) to Zn(II)
-1.4
-1.2
-1.0
-0.8
-0.6
-0.4
-0.2
0 mA cm-2
o
8
22 C
50 mA cm-
2
o
50 C
-1.2
-2
o
2
0 mA cm
70 C
70 oC
5
50 mA cm-2
-1.0
2
0 mA cm-2
o
0
C
-0.8
0 mA cm-2
o
8
22
C
-0.6
-0.4
-0.2
0
.0
0.0
0
2
4
6
8
0
2
4
6
8
Time / hr
Time / hr
c) -
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
(d) -1.6
-1.4
(
Zn(II) to Zn(0)
Zn(0) to Zn(II)
Zn(II) to Zn(0)
No additive
Zn(0) to Zn(II)
Tetrabutylammonium hydroxide
-
-
-
-
-
-
-
-
1
Potassium sodium tartarate
1
.8 cm s
-1
-1.2
7.8 cm s
Indium oxide
-
1
5
.8 cm s
Indium
oxide
-1
-1.0
-1
5.8 cm s
1.8 cm s
-1
-0.8
7.8 cm s
No additive
-0.6
Tetrabutylammonium
hydroxide
-0.4
Potassium
sodium
tartarate
-0.2
0
.0
0.0
0
2
4
6
8
0
2
4
6
8
Time / hr
Time / hr
Fig. 5. The effect of current density, temperature, mean linear electrolyte flow rate and electrolytic additives of zinc half cell on a carbon polyl-ester composite substrate
−
2
◦
from a methanesulfonic acid bath in a parallel plate flow cell. Unless indicated, the current density was 50 mA cm , deposition time 4 h, temperature 50 C, mean linear flow
velocity 3.9 cm s and the solution was 1.5 mol dm Zn(CH3SO3)2 and 1 mol dm CH3SO3H.
−1
−3
−3
The mechanisms of such stacking behaviour from nucleation to
growth have been discussed by a proposed model of Lee et al. [69]
with schematic diagrams and microscope images in the medium of
zincate.
In general, the microstructure of the zinc deposit without addi-
tive is quite smooth indicated by a shiny appearance without any
nodule or dendrite. This finding is consistent with previous investi-
gations that methanesulfonic acid can reduce the dendritic growth
gested in the voltammetric study of this work. In the presence of
tetrabutylammonium hydroxide (Fig. 6c), the microstructure was
very different and was formed by nano size needle-shaped struc-
tures. Similar morphology has been reported by Gomes and da Silva
Pereira [68] in the presence of cetyl trimethyl ammonium bromide
in the sulfate medium. The needle-shaped microstructure was due
to the competition between nucleation and crystal growth by the
strong blocking effect of the absorbed surfactant as discussed in
Section 3.5. The strong blocking effect and the adsorption of tetra-
butylammonium hydroxide has been examined by the low charge
and voltage efficiencies in the deposition and stripping experiment
in a parallel flow cell (Section 3.4.3).
Among the additives tested, indium oxide had the highest
half-cell efficiencies and a significant effect on the cell voltage.
Due to the low deposition overpotential, zinc deposited at a
faster rate and boulder-like agglomerates were observed after
the addition of indium oxide. The boulder structure as shown in
Fig. 6d has been considered to be a precursor of dendrites but the
latter will not necessarily develop [74]. In this work, relatively
rough and nodular morphology was observed instead of dendrites.
On the coating surface, a small amount of indium (0.55 wt.%)
has been detected by the energy-dispersive spectroscopy. In all
[
10,27], which has been a major problem in the zinc-halogen redox
flow batteries, leading to short circuit and hence reduced life cycle
70–73]. In a zinc–bromine system, zinc dendrite was observed at
[
−2
current density as low as 15 mA cm [3]. Therefore, methanesul-
fonic acid is attractive for zinc-based redox flow battery, as it not
only minimizes the dendritic growth but also maintain high charge
efficiency (91%) as discussed in Section 3.4.
With the addition of potassium sodium tartrate, similar
microstructure of the rough, irregular plate-like structure was
observed as in the absence of additive (Fig. 6b). Comparing with the
previous micrograph, the hexagonal platelets were more difficult
to distinguish and were mostly covered by the irregular plate struc-
tures. This was due to the faster growth rate of the zinc platelets
at the lower overpotential of potassium sodium tartarate as sug-

P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
6545
Fig. 6. The effect of electrolytic additives on microstructure of zinc electrodeposits obtained on a carbon polyvinyl-ester composite for a methanesulfonic acid bath in a
−3
−3
beaker. Electrolytic additives 2 × 10 mol dm , (i) no additive, (ii) potassium sodium tartarate, (iii) tetrabutylammonium hydroxide and (iv) indium oxide. Electrolyte
compositions and operating conditions were as in Fig. 5 but stirring at 400 rpm.
electrodeposits, no dendrite has been observed after depositing
for 4 h at 50 mA cm in the absence or presence of the electrolytic
additives.
Acknowledgements
−2
Financial support has been provided by the Research Institute
for Industry (RIfI) at the University of Southampton. The authors
are grateful to Dr. L. Berlouis, University of Strathclyde for helpful
discussions. The authors particularly appreciate the training in elec-
tron microscopy provided by Miss Chao Ma and Dr Shuncai Wang.
This work represents part of P.K. Leung’s PhD research programme
on the development of zinc-based flow batteries for energy storage
and conversion technologies.
4
. Conclusions
Zinc electrodeposition from 0.01 mol dm−3 Zn(II) ions in
methanesulfonic acid is a mass transport controlled process and
−
6
2
−1
with a zinc ion diffusion coefficient of 7.5 × 10 cm s obtained
from the Randles-Sevick and Levich equations.
At a higher acid concentration, a significant decrease in
charge efficiency was observed due to more dominant hydro-
gen evolution and faster corrosion rate. Although higher zinc
ion concentration led to more difficult dissolution process, it is
still advantageous for redox flow battery application as not only
facilitates the deposition process, but also allows the battery to
charge at a high current density without mass transport limita-
tion and ensures enough Zn(II) ions at all state of charge. The
optimum electrolyte compositions for flow battery application
were suggested at high zinc(II) methanesulfonate concentration
References
[1] N. Tokuda, T. Kanno, T. Hara, T. Shigematsu, Y. Tsutsui, A. Ikeuchi, T. Itou, T.
Kumamoto, SEI Technol. Rev. 50 (1998) 88.
[
[
2] J. Makansi, J. Abboud, Energy Storage Council White Paper, 2002.
3] C. Ponce de León, A. Frías-Ferrer, J. González-García, D.A. Szánto, F.C. Walsh, J.
Power Sources 160 (2006) 716.
[
[
4] J. Jorné, J.T. Kim, D. Kralik, J. Appl. Electrochem. 9 (1979) 573.
5] T.R. Crompton, Battery Reference Book, 3rd ed., Elsevier Science & Technology
Books, Newnes, Oxford, England/Boston, 2000, Chapter 14.
[6] P.C. Butler, D.W. Miller, A.E. Verardo, 17th Intersociety Energy Conversion Engi-
neering conference, Los Angeles, 1982.
−3
(
(
1.5–2 mol dm ) and at low methanesulfonic acid concentration
i.e. 0.5 mol dm ).
[
[
7] H.S. Lim, A.M. Lackner, R.C. Knechtli, J. Electrochem. Soc. 124 (1977) 1154.
8] G.B. Adams, Electrically rechargeable battery, US Patent 4180623 (25/12/1979).
−3
In the presence and absence of the electrolytic additives tested,
no dendrite was observed in all zinc electrodeposits at 50 mA cm
[9] R.L. Clarke, B.J. Dougherty, S. Harrison, P.J. Millington, S. Mohanta, Cerium Bat-
teries, US Patent 2004/0202925 A1 (2004).
10] R. Clarke, Zinc air battery with acid electrolyte, US Patent 7582385 B2
−
2
[
◦
and 60 C over 4 h. The low deposition overpotential observed with
(
01/09/2009).
−3
−3
the addition of 2 × 10 mol dm potassium sodium tartarate and
indium oxide led to irregular, layer-like and granular microstruc-
tures, respectively. Among the three additives tested, only indium
oxide was suitable for zinc-based redox flow battery as energy effi-
ciency improved significantly from 62% in the absence of additives
to 73% due to the improvement of both charge and voltage efficien-
cies in the presence of additives.
[11] J. Bard, L.R. Faulkner, Electrochemical Methods – Fundamentals and Applica-
tions, 2nd ed., Wiley, 2001.
12] Plurion Systems Inc., http://plurionsystems.com/contact.html, 2011 (accessed
on 26.04.11).
[13] M.D. Gernon, M. Wu, T. Buszta, P. Janney, Green Chem. 1 (1999) 127.
[14] R.P. Kreh, R.M. Spotnitz, J.T. Lundquist, J. Org. Chem. 54 (1989) 1526.
[
[
15] J.Y. Lee, J.W. Kim, M.K. Lee, H.J. Shin, H.T. Kim, S.M. Parka, J. Electrochem. Soc.
51 (2004) C25.
16] K. Saber, C.C. Koch, P.S. Fedkiw, J. Mater. Sci. Eng. A341 (2003) 174.
1
[

6
546
P.K. Leung et al. / Electrochimica Acta 56 (2011) 6536–6546
[
17] M. Mouanga, L. Ricq, G. Douglade, J. Douglade, P. Ber c¸ ot, J. Surf. Coat. Technol.
[46] L.E. Morón, Y. Meas, R. Ortega-Borges, J.J. Perez-Bueno, H. Ruiz, G. Trejo, Int. J.
Electrochem. Sci. 4 (2009) 1735.
2
01 (2006) 762.
[
[
18] J.X. Yu, Y.Y. Chen, H.X. Yang, Q.A. Huang, J. Electrochem. Soc. 146 (1999) 1789.
19] G. Trejo, R. Ortega, Y. Meas, P. Ozil, E. Chaînet, B. Nguyen, J. Electrochem. Soc.
[47] J. Torrent-Burgués, E. Guaus, J. Portugaliae Electrochim. Acta 21 (2003) 179.
[48] E. Guerra, Evaluation of Zinc Electrodeposition Kinetics from Acidic Zinc Sul-
fate Solutions using a UPD Modified Platinum Substrate, University of British
Columbia, 2003. https://circle.ubc.ca/handle/2429/16770 (accessed 26.04.11).
[49] G. Trejo, H. Ruiz, R. Ortega Borges, Y. Meas, J. Appl. Electrochem. 31 (2001) 685.
[50] K. Sato, K. Yamato, K. Iozumi, Trans. Iron Steel Inst. Jpn. 23 (1983) 946.
[51] L.J. Durney, W.H. Safranek, Electroplating engineering handbook, in: High
Speed Electroplating, Common Technique for Sustaining High Current Density
and a Fast Deposition Rate, Chapman & Hall, 1998, Chapter 39.
[52] C Cachet, R. Wiart, Electrochim. Acta 44 (1999) 4743.
1
45 (1998) 4090.
[
[
[
20] D.S. Baik, D.J. Fray, J. Appl. Electrochem. 31 (2001) 1141.
21] M. Sanchez Cruz, F. Alonso, J.M. Palacios, J. Appl. Electrochem. 23 (1993) 364.
22] M.C. Li, S.Z. Luo, Y.H. Qian, W.Q. Zhang, L.L. Jiang, J.I. Shen, J. Electrochem. Soc.
1
54 (2007) D567.
[
[
23] M. Shanmugasigamani, J. Pushpavanam, Appl. Electrochem. 36 (2006) 315.
24] M.C. Li, L.L. Jiang, W.Q. Zhang, Y.H. Qian, S.Z. Luo, J.N. Shen, J. Solid State Elec-
trochem. 11 (2007) 549.
[
[
[
25] C.K. Sarangi, B.C. Tripathy, I.N. Bhattacharya, T. Subbaiah, S.C. Das, B.K. Mishra,
Miner. Eng. 22 (2009) 1266.
26] G. Nikiforidis, L. Berlouis, D. Hall, D. Hodgson, J. Power Sources. Available online
from 21 January 2011. doi:10.1016/j.jpowsour.2011.01.036.
27] G. Brodt, J. Haas, W. Hesse, H.U. Jäqer, Method for electrolytic galvanizing
using electrolytes containing alkane sulphonic acid, US Patent 2003/0141195
Al (31/07/2003).
[53] M. Yano, S. Fujitani, K. Nishio, Y. Akai, M. Kurimura, J. Power Sources 74 (1998)
129.
[54] L. Mirkova, G. Maurin, I. Krastev, C. Tsvetkova, J. Appl. Electrochem. 31 (2001)
647.
[55] B.C. Tripathy, S.C. Das, G.T. Hefter, P. Singh, J. Appl. Electrochem. 28 (1998) 915.
[56] J.L. Zhu, Y.H. Zhou, C.Q. Gao, J. Power Sources 72 (1998) 231.
[57] C.W. Lee, K. Sathiyanarayanan, S.W. Eom, H.S. Kim, M.S. Yun, J. Power Sources
159 (2006) 1474.
[58] C.W. Lee, K. Sathiyanarayanan, S.W. Eom, H.S. Kim, M.S. Yun, J. Power Sources
160 (2006) 161.
[59] Jr. Diaddario, L. Leonard, Method for improving the macro throwing power for
chloride zinc electroplating baths, US Patent 6143160 (2000).
[60] Q.B. Zhang, Y.X. Hua, T.G. Dong, D.G. Zhou, J. Appl. Electrochem. 39 (2009) 1207.
[61] E.M. Hofer, H.E. Hintermann, J. Electrochem. Soc. (1963) 103.
[62] A.E. Saba, A.E. Elsherief, Hydrometallurgy 54 (2000) 91.
[63] F. Galvani, I.A. Carlos, J. Met. Finish 95 (1997) 70.
[64] A.M. Alfantazi, D.B. Dreisinger, J. Hydrometallurgy 69 (2003) 99.
[65] A. Hazza, D. Pletcher, R. Wills, Phys. Chem. Chem. Phys. 6 (2004) 1773.
[66] C.C. Yang, S.J. Lin, J. Power Sources 112 (2002) 497.
[
[
[
[
[
[
[
28] M. Skyllas-Kazacos, F. Grossmith, J. Electrochem. Soc. 134 (1987) 2950.
29] H.T. Zhou, H.M. Zhang, P. Zhao, B.L. Yi, Electrochim. Acta 51 (2006) 6304.
30] C.M. Hagg, M. Skyllas-Kazacos, J. Appl. Electrochem. 32 (2002) 1063.
31] J.X. Yu, H.X. Yang, X.P. Ai, Y.Y. Chen, Russian J. Electrochem. 38 (2002) 321.
32] A. Gomes, M.I. da Silva Pereira, Electrochim. Acta 52 (2006) 863.
33] A.M. Volmer, Z. Weber, J. Phys. Chem. 119 (1926) 277.
34] J.C. Ballesteros, P. Díaz-Arista, Y. Meas, R. Ortega, G. Trejo, Electrochim. Acta 52
(
2007) 3686.
[
[
35] F.C. Walsh, M.E. Herron, J. Phys. D: Appl. Phys. 24 (1991) 217.
36] R.C. Kerby, H.E. Jackson, T.J. O’Keefe, Y.M. Wang, Metall. Mater. Trans. B 8 (1977)
6
61.
[37] F. Ganne, C. Cachet, G. Maurin, R. Wiart, E. Chauveau, J. Petitjean, J. Appl. Elec-
trochem. 30 (2000) 665.
[67] C.C. Hu, C.Y. Chang, J. Mater. Chem. Phys. 86 (2004) 195.
[68] A. Gomes, M.I. da Silva Pereira, Electrochim. Acta 51 (2006) 1342.
[69] S.K. Lee, J.H. Lee, Y.H. Kim, J. Electronic Mater. 36 (2007) 1442.
[70] Y. Ando, T. Ochiai, Electrolyte for zinc–bromine storage batteries, US Patent
4510218 (04/09/1985).
[
[
38] E. Budevski, G. Staikov, W.J. Lorenz, Electrochim. Acta 45 (2000) 2559.
39] B.Z. Jugovi c´ , T.L. Tri sˇ ovi c´ , J.S. Stevanovi c´ , M.D. Maksimovi c´ , B.N. Grgur, Elec-
trochim. Acta 51 (2006) 6268.
[
[
[
40] J. McBreen, E. Gannon, J. Electrochem. Soc. 130 (1983) 1667.
41] H. Fischer, Electrochim. Acta 2 (1960) 50.
42] H. Fischer, Elektrolytische Absecheichung and Electrokrystallisation der Met-
allen Springer Berlin, 1954.
[71] Y. Ando, Zinc dendrite inhibitor, US Patent 4479856 (1984).
[72] F.G. Will, Dendrite-inhibiting electrolytic solution and rechargeable aqueous
zinc-halogen cell containing the solution, US Patent 4074028 (02/14/1978).
[73] R.A. Putt, M.J. Montgomery, Metal-halogen battery having reduced dendrite
growth, US Patent 4218521 (08/19/1980).
[43] M.A.M. Ibrahim, J. Chem. Technol. Biotechnol. 75 (2000) 745.
[44] A. Zirino, M.L. Healy, Limnol. Oceanogr. 15 (1970) 956.
[45] A.E. Dong, J. Appl. Spectrosc. 27 (1977) 124.
[74] I.N. Justinijanovi c´ , J.N. Jovi c´ evi c´ , A.R. Despi c´ , J Appl. Electrochem. 3 (1973) 193.