Journal of Power Sources 372 (2017) 221–226

Contents lists available at ScienceDirect

Journal of Power Sources

journal homepage: www.elsevier.com/locate/jpowsour

Zinc electrodeposition from flowing alkaline zincate solutions: Role of MARK

hydrogen evolution reaction
Jan Dundáleka, Ivo Šnajdra, Ondřej Libánskýa, Jiří Vránaa, Jaromír Pocedičb, Petr Mazúrb,
Juraj Koseka,b,∗
a
Department of Chemical Engineering, University of Chemistry and Technology, Prague, Technická 5, 166 28 Prague 6, Czech Republic
b
New Technologies - Research Centre, University of West Bohemia, Univerzitní 8, 306 14 Pilsen, Czech Republic

H I G H L I G H T S

• Hydrogen evolution reaction has a complex role during zinc electrodeposition.

• Current efficiency of zinc deposition is lowered by hydrogen evolution.
• Rising hydrogen bubbles increase zinc limiting current density.
• Current density ratio is not suitable as zinc morphology indicator.

A R T I C L E I N F O A B S T R A C T

Keywords: The hydrogen evolution reaction is known as a parasitic reaction during the zinc electrodeposition from alkaline
Zinc electrodeposition zincate solutions and is thus responsible for current efficiency losses during the electrolysis. Besides that, the
Hydrogen evolution rising hydrogen bubbles may cause an extra convection within a diffusion layer, which leads to an enhanced
Alkaline solutions mass transport of zincate ions to an electrode surface. In this work, the mentioned phenomena were studied
Flow battery
experimentally in a flow through electrolyzer and the obtained data were subsequently evaluated by mathe-
Zinc morphology
Mathematical modeling
matical models. The results prove the indisputable influence of the rising hydrogen bubbles on the additional
mixing of the diffusion layer, which partially compensates the drop of the current efficiency of the zinc de-
position at higher current flows. Moreover, the results show that the current density ratio (i.e., the ratio of an
overall current density to a zinc limiting current density) is not suitable for the description of the zinc deposition,
because the hydrogen evolution current density is always involved in the overall current density.

1. Introduction researches [9–16] paid attention to determine factors, which influence

A zinc metal has been historically a very popular negative electrode
material and hence was used especially in primary batteries due to its
good electrochemical behavior, high electrochemical equivalence, non-
toxicity, low-cost and availability [1]. Therefore, researchers attempted
to develop also secondary or flow batteries based on the zinc electro-
chemistry especially in alkaline solutions, where a zinc electrode can be
combined with a nickel [2–4] or an air [5–7] electrode, but the wide-
spread use of zinc-based secondary batteries is restrained mainly by the
poor cycle life of the zinc electrode [8].
The poor cycle life is caused by shape changes and dendritic growth
of the electrode [8]. These phenomena are linked with an electro-
chemical dissolution and deposition of the zinc metal during cycling, so
it is important to control the morphology of the zinc deposits. Many
the final zinc morphology, and among them an electrolyte convection,
an applied current density, a zincate concentration and a presence of
additives in the electrolyte were proved to play the key role. Although
the electrolyte convection was found as an important factor affecting
the zinc electrodeposition, the effect of convection caused by rising
hydrogen bubbles has been studied minimally.
This work aims to study the influence of the hydrogen evolution
reaction on the zinc electrodeposition from the flowing alkaline zincate
solutions. Since the hydrogen evolution reaction is thermodynamically
preferred, as the value of the standard reduction potential for the hy-
drogen reduction is higher than for the zinc reduction, it can be ex-
pected that the coulombic efficiency of the zinc deposition will be
lowered by the presence of the hydrogen gas evolution [17]. On the
other hand, the rising hydrogen bubbles may influence the mass-

∗
Corresponding author. Department of Chemical Engineering, University of Chemistry and Technology, Prague, Technická 5, 166 28 Prague 6, Czech Republic.
E-mail address: juraj.kosek@vscht.cz (J. Kosek).

http://dx.doi.org/10.1016/j.jpowsour.2017.10.077
Received 23 August 2017; Received in revised form 16 October 2017; Accepted 24 October 2017
0378-7753/ © 2017 Elsevier B.V. All rights reserved.
J. Dundálek et al.
transfer rate of zincate ions to the zinc electrode, since the effect of
mass-transfer enhancement by rising bubbles within a mass-transfer
boundary layer was reported in literature [18]. Therefore, the final
influence of the hydrogen evolution reaction on the zinc electro-
deposition is complex. Another issue related to the presence of the
hydrogen evolution reaction is the characterization and the prediction
of the deposited zinc metal morphology. For these purposes, a current
density ratio (i.e., the ratio of an overall current density to a zinc lim-
iting current density) is often used. Its value equal to 1 should in-
tuitively mean the mass transfer limitation of the zinc electrodeposi-
tion, which leads to the dendritic growth of zinc metal. However, some
researchers [12,13] haven't observed such correlation of the current
density ratio with the zinc morphology.
The presented study consists of experimental electrodepositions of
zinc metal in the electrolyzer with a convective flow of the electrolyte
and of subsequent data evaluation using mathematical models de-
scribing the process with and without the dependence of the zincate
mass-transfer rate on the hydrogen gas evolution. This approach en-
ables us to reveal the complicated role of the hydrogen evolution on the
zinc electrodeposition, especially the influence of the rising hydrogen
bubbles on the limiting current density of the zinc deposition, which is
linked with the mass-transfer rate of the zincate ions. It is also de-
monstrated that the correlation of the morphology of the zinc deposits
with the current density ratio, which is commonly used for the char-
acterization of deposits, is not generally applicable when used as a
deposition criterion.
2. Experimental part
2.1. Experimental setup
The experiments were conducted in the in-house-designed and
constructed flow through cell (see Fig. 1). The cathode for the zinc
deposition was made of carbon composite plate (PPG 86, SIGRACET)
with a square shaped electrode active area of 2.25 cm2. The anode for
the oxygen evolution was made of the rolled nickel foam of the same
size as the cathode. These two electrodes were parallel to each other
with a gap of 1.0 cm. The electrolyte inlet was connected to the bottom
of the cell to ensure upward flow of the electrolyte. The electrolyte flow
rate of 6 cm3 s−1 was maintained by a peristaltic pump with a flow
output stabilized by an air vessel. The zinc deposition was conducted
using a three-electrode arrangement, in which the electric current was
supplied by a power supply and both current and voltage responses
were measured by universal multimeters. The Hg/HgO reference elec-
trode was connected at the bottom of the cell via a capillary tube ended
by a fritted glass.
As the electrolyte, solutions of 8 mol dm−3 KOH (p.a. purity) con-
taining various concentrations of ZnO (p.a. purity) were used in a great
222
Journal of Power Sources 372 (2017) 221–226
excess. The precise ZnO concentrations were determined by a chelato-
metric titration. Before each experiment, the cathode was polished with
a sandpaper of P2500 grit size, degreased with ethanol (95% vol. + 5%
vol. methanol) and weighted on the analytical balance. Then the zinc
was deposited at selected constant current densities by an electric
charge of 45 A s (an equivalent of 0.152 mg zinc). Immediately after the
deposition, the cathode with the deposited zinc metal was rinsed in
ethanol, left for 5 min at temperature of 50 °C to evaporate possible
residues and weighted again. Based on the weighting, the coulombic
efficiency of the zinc deposition was calculated using the Faraday's laws
of the electrolysis as follows:
z Zn F
mZn,deposited
ηC =
MZn
Q (1)
The oxidation of zinc deposits during the mentioned treatment,
which could influence the value of the coulombic efficiency based on
the weighting, hasn't happened, since the coulombic efficiencies were
during the first few depositions determined also by the chelatometric
titration and the provided values were very similar to the values of the
weighting method. Consequently, the optical analysis of the zinc de-
posits was carried out using a scanning electron microscope.
2.2. Mathematical approach
The obtained data from the zinc deposition in the flow through cell
were regressed with two mathematical models – the first one reflecting
the additional mixing of the diffusion layer by the rising hydrogen
bubbles and the second one neglecting this influence. The only che-
mical reactions considered for the mathematical description are zinc
deposition and the hydrogen evolution as follows:
2 H2O + 2 e− → H2 + 2 OH− (2)
− −
[Zn(OH)4] 2−
+2e → Zn + 4 OH (3)
An overall current density at the electrode is given by the sum of
corresponding current densities of the proposed chemical reactions
according to the following equation:
j = jH2 + jZn (4)
The hydrogen evolution reaction is described by the Tafel kinetics
(see equation (5)) and it is assumed that kinetics constants are identical
for the hydrogen evolution that takes place on the deposited zinc of
different morphology as well as on the carbon composite plate.
zH F
jH2 = −j0,H2 exp ⎡−αC,H2 2 (E − EH02 ) ⎤
⎣ RT ⎦ (5)
The zinc deposition reaction is expressed by the Butler-Volmer ki-
netics combined with a mass transport limitation of the zinc reduction
Fig. 1. Scheme of the flow through deposition cell.
J. Dundálek et al. Journal of Power Sources 372 (2017) 221–226

Table 1
List of model parameters.

Symbol Parameter Value Unit Source

R universal gas constant 8.314 J mol−1 K−1

F Faraday's constant 96485 C mol−1
T absolute temperature 298.15 K
p atmospheric pressure 101325 Pa
d gap between electrodes 1.0⋅10−2 m *
w width of electrode channel 1.5⋅10−2 m *
l length of electrodes 1.5⋅10−2 m *
v flow speed of electrolyte 4.0⋅10−2 m s−1 *
c KOH concentration of potassium hydroxide 8 mol dm−3 *
cZn concentration of zincate anion [0.243; 0.473; 0.767] mol dm−3 *
ρ density of electrolyte 1326 kg m−3 +
η dynamic viscosity of electrolyte 2.536⋅10−3 Pa s +
DZn diffusion coefficient of zincate anion in electrolyte 6.0⋅10−10 m2 s−1 [21]
EZn0 standard reduction potential of zinc reaction −1.285 V vs. SHE [22]
z Zn number of electrons involved in zinc reaction 2 – based on the equation (2)
EH02 standard reduction potential of hydrogen reaction −0.828 V vs. SHE [22]
z H2 number of electrons involved in hydrogen reaction 2 – based on the equation (3)

* Based on experimental apparatus and operating conditions.

+ Taken for the pure 8 mol dm−3 KOH solution [23].

Table 2
List of parameters obtained by regression.

Symbol Parameter Value with assumption of additional mixing Value without assumption of additional mixing Unit

αC,Zn cathodic charge transfer coefficient for zinc reaction 0.249 0.346 –
j0,Zn exchange current density of zinc reaction 10.1 6.07 mA cm−2
αC,H2 cathodic charge transfer coefficient for hydrogen reaction 0.103 0.126 –
j0,H2 exchange current density of hydrogen reaction 0.0189 0.00836 mA cm−2
dbubbles diameter of rising hydrogen bubbles 1.37⋅10−5 – m

as follows: diffusion layer, the thickness of the diffusion layer is obtained from the
following equation [19]:
⎧− ⎛1 − jZn ⎞ exp ⎡−αC,Zn zZn F (E − EZn
0
) ⎤⎫
⎪ jlim,C,Zn ⎣ RT ⎦⎪ 1 1 1
jZn = j0,Zn ⎝ ⎠ = +
⎨ z Zn F 0 ⎬ δ δconv 2 δ bubbles 2 (13)
⎪+ exp ⎡
⎣
(1 − αC,Zn ) RT (E − EZn ) ⎤
⎦ ⎪ (6)
⎩ ⎭
where the thickness of the diffusion layer given by a hydrodynamic
The limiting current density of the zinc reduction is calculated using convection δconv is determined using equations (8-12) and the thickness
the film theory by the following equation: of a diffusion layer given by the rising hydrogen bubbles δ bubbles is
cZn computed from a criterial description [19] as follows:
jlim,C,Zn = −z Zn FDZn
δ (7)
dbubbles
δ bubbles =
When the additional mixing of the diffusion layer caused by the Shbubbles (14)
rising hydrogen bubbles is neglected, then the thickness of the diffusion
layer δ is equal to the thickness of a diffusion layer given by a hydro- Shbubbles = 0.93Re bubbles 0.5Sc0.487 (15)
dynamic convection δconv . The hydrodynamic convection is obtained
from the analytical derivation of the Navier-Stokes equation for a liquid dbubbles v bubbles ρ
Re bubbles =
flow between plan parallel plates assuming no influence of the de- η (16)
posited zinc metal on the flow [19] as follows:
jH2 RT
d eq v bubbles =
δconv = z H2 Fp (17)
Sh conv (8)
For the evaluation of experimental data, an integral value of the
2wd coulombic efficiency obtained from the deposition experiments was
d eq =
w+d (9) approximated by a differential value of a current efficiency, since the
1
amount of the deposited zinc in the experiments was in the order of
d eq 3 milligrams. The current efficiency is defined as the ratio of the zinc
Sh conv = 1.85 ⎛ Reconv Sc⎞
⎜ ⎟

⎝ l ⎠ (10) reaction current density to the total electrode current density by the
following equation:
d eq vρ
Reconv = jZn
η (11) ηC ≈ ηj =
j (18)
η
Sc = Used model parameters are listed in Table 1. Parameters obtained
DZn ρ (12)
by regression of experimental data are listed in Table 2 and they were
In the case of the model considering the additional mixing of the estimated by a nonlinear optimization using the Interior Point

223
J. Dundálek et al. Journal of Power Sources 372 (2017) 221–226

Fig. 2. Deposited zinc structures at v = 4 cm s−1, cZn = 0.243 mol dm−3 and: a) j = 20 mA cm−2 (the mossy), b) j = 40 mA cm−2 (the crystalline), c) j = 65 mA cm−2 (the
crystalline dendrites), d) j = 180 mA cm−2 (the dendrites).

Fig. 3. Experimentally obtained and mathematically mod-

eled polarization curves of the zinc electrode with the in-
dicated morphology of zinc deposits at v = 4 cm s−1 and
zincate concentration of: a) cZn = 0.243 mol dm−3, b)
cZn = 0.473 mol dm−3 and c) cZn = 0.767 mol dm−3.

Optimizer [20] of the following objective function: density ratio (CDR ) and the current density ratio of the zinc deposition
(CDRZn ). The definition of CDR is taken from the literature [12,13]:
f= ∑ ∑ [(Eexp − E (jexp ))2 + (ηC − ηj (jexp ))2]
cZn jexp (19) j
CDR =
The obtained zinc deposits were linked to the value of the current jlim,C,Zn (20)

224
J. Dundálek et al. Journal of Power Sources 372 (2017) 221–226

Fig. 4. Experimentally obtained and mathematically mod-

eled current efficiencies of the zinc deposition with the
indicated morphology of zinc deposits at v = 4 cm s−1 and
zincate concentration of: a) cZn = 0.243 mol dm−3, b)
cZn = 0.473 mol dm−3 and c) cZn = 0.767 mol dm−3.

between the mossy and the crystalline morphology, which is char-

acterized by simultaneous deposition of both these structures, was de-
tected rarely in our experiments. On the contrary, the transient state
between the crystalline and the dendritic morphology was observed
frequently. In this transition state, some of the zinc crystals grow up
rapidly above the electrode surface like the zinc dendrites. Since this
structure occurred frequently in our experiments, it is considered to be
a new morphology pattern, which will be further referred as the crys-
talline dendritic.
Fig. 3 and Fig. 4 illustrate the influence of the hydrogen evolution
on the zinc electrode polarization curves and on the current efficiency
of the zinc deposition. The hydrogen evolution reaction starts to run
significantly near the mass transfer limitation of the zinc deposition,
since the shape of the zinc electrode polarization curve lacks the pla-
teau of the zinc limiting current density. The occurrence of the hy-
drogen evolution side reaction is responsible for the drops of the current
efficiency of the zinc deposition, which rapidly decreases at higher
current densities, where the morphologies of zinc deposits are crystal-
line dendritic or dendritic.
More detailed information about the zinc deposition is given by the
prediction of the process using the mathematical models with the re-
gression parameters (see Figs. 3 and 4). Although the models aren't
perfect in the prediction of the zinc electrode polarization curves at the
high current densities due to the use of the simple Tafel equation for the
description of the hydrogen evolution with a complex reaction me-
Fig. 5. Comparison of the CDR and the CDRZn calculated by the mathematical model chanism, the models confirm the experimentally observed influence of
considering the additional mixing of the diffusion layer by rising bubbles at v
the hydrogen evolution reaction on the zinc deposition phenomena like
= 4 cm s−1.
the missing plateau of the zinc limiting current and the drop of the
current efficiency at the higher current densities, because the Tafel
while CDRZn is defined within this work as follows: equation is a reasonable simplification of the hydrogen evolution at low
jZn rates of gas generation. Moreover, the models show another impact of
CDRZn = hydrogen evolution – an additional mixing of the diffusion layer by
jlim,C,Zn (21) rising bubbles. The bubbles contribute to the convection in the diffusion
layer as they grow at the electrode surface and then they detach from
the electrode. This enhancement of the mass transport of the zincate ion
3. Results and discussion to the electrode surface improves the current efficiency of the zinc
deposition at higher applied current densities. Therefore, the mathe-
The zinc deposits (see Fig. 2) occurred in morphologies standardly matical prediction, which does not consider the additional mixing of
reported in literature [9,10] – the mossy, the crystalline (also referred the diffusion layer, always underestimates the experimentally observed
as the compact or the boulder) and the dendritic. The transient state

225
J. Dundálek et al.
Fig. 6. Prediction of the zinc deposition morphology map for the electrolyzer at v
= 4 cm s−1 using the mathematical model that includes the additional mixing of the
diffusion layer by rising bubbles and CDRZn approach.
Table 3
List of the used boundary values of CDRZn for three morphology changes.
Morphology change Value of CDRZn
mossy – crystalline 0.600
crystalline – crystalline dendrites 0.950
crystalline dendrites – dendrites 0.998
current efficiency.
Besides that, the presence of the hydrogen evolution reaction
complicates the description of the zinc deposition process based on the
current density ratio. As the overall current density at the electrode
includes always the contribution of the current density of the hydrogen
evolution, the CDR may take values higher than 1 (see Fig. 5), which is
understood as the mass-transfer limitation of the zinc deposition, but
the dendritic morphology of the zinc deposits is not observed. Another
difficulty of the CDR usage is linked with the prediction of the limiting
current density of the zinc reduction, which is supposed to be constant,
but it is in reality variable due to the additional mixing of the diffusion
layer caused by the rising bubbles as it has been shown and discussed
above.
As a better morphology indicator, it is more convenient to use the
current density ratio of the zinc deposition CDRZn . Its value may range
due to the definition only within the interval from 0 to 1 (see Fig. 5), so
the CDRZn is more suitable for the characterization and the prediction of
the morphology of the deposited zinc metal. On the other hand, the
correct use of CDRZn requires the implementation of mathematical
models of the zinc deposition process that include the hydrogen evo-
lution and the influence of the rising bubbles on the diffusion layer.
Once are these models developed, they can be used for the prediction of
zinc deposition morphology maps, which provide the information about
the morphology of zinc deposits at selected flow conditions in depen-
dence on current density and zincate concentration. The morphology
map can be used for the adjustment of operation conditions of the flow
batteries based on the alkaline zinc chemistry during the charging or
the electrolysis of the flowing alkaline zincate solution to obtain the
desired morphology of zinc deposits. An illustrative example of such
approach is shown in Fig. 6, where the zinc deposition morphology map
is generated for the presented flow through electrolyzer using the
morphology boundary parameters listed in Table 3, which were de-
termined from experiments. Based on the generated morphology map, a
226
Journal of Power Sources 372 (2017) 221–226
long-term electrolysis with the varying applied current may be carried
out to obtain the zinc deposits with the required morphology.
4. Conclusions
The complex role of the hydrogen evolution reaction on the zinc
electrodeposition was studied experimentally with the support of
mathematical modeling. The hydrogen evolution reaction started to
occur near the limiting current density of the zinc deposition and it was
responsible for the drop of the current efficiency during the zinc de-
position. On the other hand, the rising hydrogen bubbles caused the
additional mixing of the diffusion layer, thus the mass transport of the
zincate ion to the electrode surface was enhanced by this extra con-
vection, so this phenomenon partially compensated the drop of the
current efficiency. It was shown that due to the complex influence of
the hydrogen evolution reaction during the zinc electrodeposition, the
commonly used current density ratio (CDR ) is not suitable as the
morphology indicator of the zinc deposits, since the hydrogen evolution
current density is involved in the overall current density. Our work
demonstrates that it is more convenient for the morphology prediction
to use the current density ratio of the zinc deposition (CDRZn ) with the
consideration of the additional mixing of the diffusion layer caused by
the rising hydrogen bubbles. Although such approach requires some
modeling effort, it represents a huge benefit in terms of the under-
standing of the underlying processes.
Acknowledgements
Financial support from specific university research fund (MŠMT no.
20/2017) is gratefully acknowledged.
The result was also developed within the CENTEM project, reg. no.
CZ.1.05/2.1.00/03.0088, cofunded by the ERDF as part of the Ministry
of Education, Youth and Sports OP RDI programme and, in the follow-
up sustainability stage, supported through CENTEM PLUS (LO1402) by
financial means from the Ministry of Education, Youth and Sports under
the National Sustainability Programme I.
References
[1] D. Linden, T.B. Reddy, Handbook of Batteries, McGraw-Hill, 2002.
[2] J. McBreen, J. Power Sources 51 (1994) 37–44.
[3] A.K. Shukla, S. Venugopalan, B. Hariprakash, J. Power Sources 100 (2001)
125–148.
[4] J. Cheng, L. Zhang, Y.-S. Yang, Y.-H. Wen, G.-P. Cao, X.-D. Wang, Electrochem.
Commun. 9 (2007) 2639–2642.
[5] J.-S. Lee, S. Tai Kim, R. Cao, N.-S. Choi, M. Liu, K.T. Lee, J. Cho, Adv. Energy Mater.
1 (2011) 34–50.
[6] Y. Li, H. Dai, Chem. Soc. Rev. 43 (2014) 5257–5275.
[7] P. Pei, K. Wang, Z. Ma, Appl. Energy 128 (2014) 315–324.
[8] X.G. Zhang, Secondary batteries - zinc systems | zinc electrodes: overview,
Encyclopedia of Electrochemical Power Sources, Elsevier, Amsterdam, 2009, pp.
454–468.
[9] R.Y. Wang, D.W. Kirk, G.X. Zhang, J. Electrochem. Soc. 153 (2006) C357–C364.
[10] R.D. Naybour, Electrochim. Acta 13 (1968) 763–769.
[11] J.W. Diggle, R.J. Fredericks, A.C. Reimschuessel, J. Mater. Sci. 8 (1973) 79–87.
[12] Y. Ito, X. Wei, D. Desai, D. Steingart, S. Banerjee, J. Power Sources 211 (2012)
119–128.
[13] D. Desai, X. Wei, D.A. Steingart, S. Banerjee, J. Power Sources 256 (2014) 145–152.
[14] A. Gavrilović-Wohlmuther, A. Laskos, C. Zelger, B. Gollas, A.H. Whitehead, J.
Energy Power Eng. 9 (2015) 1019–1028.
[15] C. Zelger, J. Laumen, A. Laskos, B. Gollas, Electrochim. Acta 213 (2016) 208–216.
[16] N.D. Nikolić, P.M. Živković, J.D. Lović, G. Branković, J. Electroanal. Chem. 785
(2017) 65–74.
[17] M.M. Saleh, J.W. Weidner, B.G. Ateya, J. Electrochem. Soc. 142 (1995) 4113–4121.
[18] G.M. Whitney, C.W. Tobias, AIChE J. 34 (1988) 1981–1995.
[19] I. Roušar, K. Micka, A. Kimla, Electrochemical Engineering vol. 2, Academia, 1989.
[20] A. Wächter, T.L. Biegler, Math. Program. 106 (2006) 25–57.
[21] Z. Mao, R.E. White, J. Electrochem. Soc. 139 (1992) 1105–1113.
[22] E. Deiss, F. Holzer, O. Haas, Electrochim. Acta 47 (2002) 3995–4010.
[23] P.M. Sipos, G. Hefter, P.M. May, J. Chem. Eng. Data 45 (2000) 613–617.