Journal of Electroanalytical Chemistry 799 (2017) 134–141

Contents lists available at ScienceDirect

Journal of Electroanalytical Chemistry

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

Electrochemical impedance studies of Na/MnO2 primary cells MARK

⁎
Brij Kishore, Munichandraiah Nookala
Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bengaluru 560012, India

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

Keywords: Primary Na/MnO2 cells are expected to gain importance for replacement of primary Li/MnO2 cells in future akin
Na/MnO2 primary cells to the Na-ion cells as alternative for Li-ion cells. Amorphous MnO2 is prepared by a drop-wise addition of KMnO4
Depth-of-discharge solution to MnSO4 solution, and coin cells of Na/MnO2 are fabricated in a non-aqueous electrolyte of Na salt. The
Surface film electrochemical impedance spectroscopy (EIS) data provides a high resistance of Na metal due to the surface
Delayed response
passive film. On subjecting the cell for discharge, the surface film causes a delay response of the cell voltage and
Electrochemical impedance spectroscopy
the closed circuit voltage reaches the normal discharge level following dielectric breakdown of the film. The EIS
data measured at different stages of cell discharge are subjected to non-linear least square fitting with the aid of
an appropriate equivalent circuit. The impedance parameters are examined to throw light on state-of-charge of
Na/MnO2 primary cells.

1. Introduction Zn/HgO primary cells by differential AC resistance method. Measure-

ments of current-voltage characteristics, and AC impedance at different
There has been increasing interest on Na-based non-aqueous re- currents and frequencies were reported. Slopes of current-voltage
chargeable batteries in research and development laboratories for a curves above a certain cell current were nearly constant. These slopes
viable substitution for Li-based batteries [1,2]. This is because of in- had linear relation to the charge expended and were used as a basis for
creasing demand for Li-ion batteries in spite of the fact that the global state-of-charge indication. Impedance parameters and SOC estimation
resources of Li are limited and they are distributed unevenly [3,4]. studies of Zn/MnO2 and Mg/MnO2 primary cells were reported by
Primary batteries based on Li are also widely used for several appli- Gopikanth and Sathyanarayana [17]. The problem of non-destructive
cations [5]. Similar to the substitution of Li in Li-ion batteries resulting determination of SOC of these cells was examined from the view point
in Na-ion batteries, Na-based primary batteries are expected to be vi- of impedance parameters and open-circuit voltage at equilibrium. The
able substitutes for Li-based primary batteries. Recently, we have re- equivalent series/parallel resistance and capacitance, modulus of im-
ported studies on Na/MnO2 primary cells and compared the results with pedance and phase shift of commercial Zn/MnO2 and prototype Mg/
Li/MnO2 primary cells, which were fabricated and evaluated under MnO2 cells were measured at different SOC values. For an un-dis-
similar experimental conditions [6]. The performance of Na/MnO2 cells charged Mg/MnO2 cell, the Nyquist plot was a straight line with slope
is comparable to that of Li/MnO2 cells, except that the discharge pla- greater than unity, which was ascribed due to protective surface film
teau voltage of the former is less by about 0.4 V. present on the Mg anode. However, the Nyquist plot exhibited a
Electrochemical impedance spectroscopy (EIS) is a useful technique semicircle, after the cell was subjected to partial discharge, due to
to evaluate various properties of electrochemical systems [7–10]. The breakdown of surface film on Mg. The area under the semicircle de-
technique involves application of a small alternating current (AC) signal creased with a decrease in SOC indicating a decrease in polarization
over a wide frequency range on the electrochemical cell and measuring resistance with a progressive discharge of the cells. In the case of Zn/
current response. The analysis of EIS data allows separation of resistive MnO2 primary cells, there was a progressive change from charge-
and capacitive components of the electrochemical cell and, therefore, transfer control of impedance at SOC ≈ 1 to diffusion control at
several electrochemical parameters can be evaluated. Several battery SOC ≈ 0. It was concluded that determination of SOC was not possible
systems were studied by EIS as a non-destructive technique to evaluate for Mg/MnO2 and Zn/MnO2 cells by measuring equivalent series/par-
state-of-charge (SOC) and kinetic parameters of electrochemical reac- allel resistance and capacitance, modulus of impedance or phase shift
tions [11–15]. There are only a few reports available in the literature on due to large fluctuations caused by film break down and reformation
impedance studies of primary cells. Winter et al. [16], studied SOC of processes. In another study of aged Mg/MnO2 primary cells, impedance

⁎
Corresponding author.
E-mail address: [email protected] (M. Nookala).

http://dx.doi.org/10.1016/j.jelechem.2017.05.038
Received 27 December 2016; Received in revised form 16 May 2017; Accepted 22 May 2017
Available online 23 May 2017
1572-6657/ © 2017 Elsevier B.V. All rights reserved.
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

parameters were examined at several SOC values and also ageing time carbonate (Aldrich) and propylene carbonate (Aldrich) (1:1 by volume)
dependence of partially discharged cells [18]. It was shown that the was used as the electrolyte. Coin cells were assembled in argon filled
frequency (fmax) corresponding to the maximum value of the imaginary MBraun glove box model Unilab. Galvanostatic discharge experiments
component of the high frequency semicircle exhibited a nearly linear were carried out using Biologic SA multichannel potentiostat/galva-
increase from about 2 Hz (virgin cell) to about 40 Hz (fully discharged nostat model VMP3. Electrochemical impedance spectroscopy mea-
cell). It was thus concluded that fmax was a suitable parameter for the surements were carried out using a CH Instrument model CHI608C. For
estimation of SOC or residual capacity of aged Mg/MnO2 cells. For a the purpose of recording EIS at intermittent stages of cell discharge, the
partially discharged Mg/MnO2 cell, resistance parameters increased cell was galvanostatically discharged at 100 μA for the required dura-
linearly with ageing time. tion, disconnected from the potentiostat/galvanostat and allowed at
In the present work, Na/MnO2 non-aqueous primary cells are as- rest for 30 min before recording EIS. The cell was again discharged for
sembled and studied by EIS at several stages of discharge. The EIS data the required duration and EIS was measured following the above steps.
are analyzed using an appropriate equivalent circuit. The variation of All experiments were carried out at 22 ± 1 °C.
impedance parameters with depth-of-discharge (DOD) is studied to
identify the cell's state-of-charge. 3. Results and discussions

2. Experimental There are various synthetic routes known for the preparation of
MnO2, which include oxidation of Mn2+ ions by a suitable oxidizing
The performance of Na/MnO2 primary cell strongly depends on the agent, reduction of Mn7+ by a reducing agent and electrochemical
crystallographic nature of MnO2, which in turn depends on the syn- deposition [19,20]. In the present study, the synthesis of MnO2 was
thetic route [19,20]. The performance of cells assembled using poorly carried out by adding drop-wise a neutral aqueous solution of KMnO4 to
crystalline MnO2 was superior to that of cells made with crystalline an aqueous solution of MnSO4. The redox reaction between Mn7+ and
MnO2 [6]. In the present study, amorphous MnO2 was synthesized by a Mn2+ results in the formation of MnO2.
different route. It was synthesized by redox reaction between KMnO4
2KMnO4 + 3MnSO4 + 2H2 O → 5MnO2 + K2 SO4 + 2H2 SO4 (1)
(SRL) and MnSO4·H2O (HIMEDIA). In brief, 0.727 g of KMnO4 was
dissolved in 50 ml of double distilled (DD) water and added drop wise The reaction was carried out at room temperature. The product was
to 50 ml aqueous solution containing 1.166 g of MnSO4·H2O under washed and dried at 70 °C.
constant stirring. After the complete addition, the contents were stirred There are several reports on synthesis of MnO2 from the redox re-
for additional 6 h. The product was collected by filtration, washed co- action between Mn2+ and Mn7+ ions [21,22]. DeGuzman et al., pre-
piously with DD water, then with ethanol and dried at 70 °C for about pared MnO2 of different crystal phases by varying pH and temperature
12 h. The quantity of sample prepared in each batch was about 1.0 g. from the redox reaction between KMnO4 and MnSO4 [21]. Chen et al.,
The MnO2 samples were characterized by powder X-ray diffraction prepared α-MnO2 from KMnO4 and H2SO4 at 70–95 °C by the redox
(XRD) using a Philips X'Pert Pro diffractometer at 45 kV and 30 mA reaction between Mn7+ and Mn2+, the latter being formed by reduc-
with Cu Kα (λ = 1.5418 Å) as the radiation source and the data were tion of KMnO4 in strong acidic conditions [22]. Crystalline MnO2 na-
collected at 2θ range from 10 to 90° at a rate of 2° per min. The mor- noparticles were prepared by mixing aqueous solutions of KMnO4 and
phology was examined using a FEI Co scanning electron microscope MnSO4 at ambient conditions and aging for different durations at 25
(SEM) model Sirion. Nitrogen adsorption-desorption isotherms were and 85 °C. It is understood from the above reports that the redox re-
recorded at −196 °C using Micromeritics surface area analyzer model action between KMnO4 and MnSO4 produces different crystalline
ASAP2020. The specific surface area was calculated by the Brunauer- phases by varying the reaction conditions. Unlike the formation of
Emmett-Teller (BET) method from the adsorption branch. The pore size crystalline phases of MnO2 in these reports, an amorphous MnO2 is
distribution was calculated by Barrett-Joyner-Halenda (BJH) method obtained in the present study from the reaction between aqueous so-
from desorption branch of the isotherm. The oxidation states of Mn on lutions of KMnO4 and MnSO4, as detailed below. The formation of
the surface of the sample was examined by high resolution X-ray pho- amorphous MnO2 is attributed to a slow drop-wise addition of KMnO4
toelectron spectroscopy (XPS: Kratos Axis Ultra DLD system) with dual solution to MnSO4 solution.
Al source of 1486.6 eV, 300 μm × 700 μm aperture size and the C 1s The MnO2 sample was subjected to chemical analysis by ICP and
signal as the reference in ultrahigh vacuum. The stoichiometry was TGA. There was a loss of 11.29 wt% up to 250 °C when the sample was
checked via Inductively Coupled Plasma-Optical Emission Spectrometry subjected to TGA. The results of chemical analysis after applying cor-
(ICP-OES) with a Perkin Elmer Optima 2100DV Analytical Instruments. rection of 11.29 wt% water was Mn 60.32, K 1.47, and O 33.90 wt%.
Thermogravimetric analysis (TGA) was carried out under argon flow at Thus the composition of the sample becomes MnK0.033O2.17. Taking
a heating rate of 10 °C min−1 from room temperature to 900 °C using account of water present in the as prepared sample, the manganese
TG 209 F1 Iris, NETZSCH instrument. dioxide formula thus becomes MnK0.033O2.17.0.6H2O. The hydrated
For fabrication of electrodes, the active material (70 wt%), acet- oxide is referred to as MnO2 hereafter.
ylene black (20 wt%, Alfa Aesar) and polytetrafluoroethylene suspen- It is known that MnO2 forms in different polymorphs depending on
sion (10 wt%, Aldrich) were mixed in a mortar and ground well. The the way MnO6 octahedra are inter-linked [23]. Powder XRD pattern
dough, thus, formed was rolled into a sheet on a glass plate. Stainless (not given) of MnO2 sample prepared in the present study indicated that
steel (SS) mesh of 0.3 mm thickness and mesh size 80 was used as the it was poorly crystalline. Minor broad peaks were observed at 2θ values
current collector. The pre-treatment of SS mesh was carried out by first of 36.6°, 41.1°, 54.7° and 65.9° corresponding to the α-phase. For α-
etching it in 5 N HCl (SDFCL), followed by etching in 5 N HNO3 MnO2, the octahedra are linked at the corners to form a (2 × 2) and a
(SDFCL). It was then washed thoroughly with DD water, rinsed with (1 × 1) channels. This structure allows the Na+ ion diffusion. The XPS
acetone (SDFCL) and air dried. The rolled sheet of electrode material spectrum for Mn 2p3/2 signal of the sample is shown in Fig. 1. On de-
was transferred onto a mesh of 16 mm diameter such that the loading convolution of the spectrum [24], oxidation states of +2, +3 and +4
level of the active material was 10–12 mg cm−2. The electrodes were are identified for Mn, but the major oxidation state is +4. The mor-
then dried at 100 °C under reduced pressure for 12 h. Coin cells CR2032 phology of MnO2 sample as seen in scanning electron micrograph
(Hohsen Corporation, Japan) were assembled where Na (SDFCL) metal (Fig. 2a) appears globules, which consist of short fibers as reflected in
foil (about 30 mg cm−2) was used as the anode and MnO2 as the transmission electron micrograph (Fig. 2b). Nitrogen adsorption/deso-
cathode. Whatman filter paper grade GF/C was used as the separator rption isotherms and BJH curves of S70 sample are shown in Fig. 3.
and 1 M NaPF6 (Aldrich) dissolved in a mixed solvent of ethylene There is a wide hysteresis loop in the isotherm (Fig. 3), which is an

135
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

220

Quantity of N2 adsorbed / cm3 g-1

3 -1 -1
200

dV/dD / cm g nm
0.4

0.3
180
4+ 0.2
Mn
Intensity / a.u.

160 0.1

140 0.0
0 1 2 3 4 5 6 7 8 9 10
120 Pore diameter / nm

100
Mn3+
80

Mn2+
60
40
638 640 642 644 646 648
20
Binding energy / eV 0.0 0.2 0.4 0.6 0.8 1.0
o
Fig. 1. X-ray photoelectron spectra of MnO2 sample in Mn 3P3/2 region.
Relative pressure (p/p )
Fig. 3. Nitrogen adsorption/desorption isotherm of MnO2 sample and in the inset is BJH
indication of porous nature of the sample. The BET specific surface area pore distribution curve.
for the sample is 184 m2 g−1. The BJH curve (Fig. 3 inset), shows a very
narrow pore size distribution at pore diameter of 3.5 nm. 3.5 3.5
A Na/MnO2 coin cell was subjected to galvanostatic discharge at
0.50 mA (0.25 mA cm−2) and the voltage profile is presented in Fig. 4. 3.0
The open circuit voltage of the cell before discharge is 3.32 V. On in- 3.0

Cell voltage / V
Cell voltage / V
itiation of galvanostatic discharge, there is a sudden decrease in cell 2.5

voltage followed by a rise to about 3.05 V. Thereafter, the cell voltage

decreases gradually down to the cut-off voltage of 1.0 V. The time re- 2.5 2.0

quired to complete the discharge is 7 h (i.e. C/7 rate) and the discharge 1.5
capacity obtained is 3.5 mAh. This value is close to the theoretical value
2.0
(308 mAh g−1), which is calculated on the basis of MnO2. The dis- 1.0
charge voltage profile is similar to that of Li/MnO2 cells, which are 0 1 2 3 4 5 6 7
assembled under similar conditions to that of Na/MnO2 cells [6]. The 1.5 Time / h
discharge capacity of both Li/MnO2 and Na/MnO2 cells are also the
same. However, the voltage of Na/MnO2 cell is less than that of Li/
MnO2 cell by about 0.4 V, which arises due to the difference in po- 1.0
tentials of Li/Li+ and Na/Na+ electrodes.
Nyquist and Bode plots of a fresh Na/MnO2 cell are shown in Fig. 5. 0 1 2 3 4 5 6 7
The Nyquist plot (Fig. 5a) consists of a high frequency intercept on the Time / h
real axis, a semicircular region in the frequency range from 100 kHz to
1 Hz and a linear region, which is parallel to the real axis. The Bode Fig. 4. Discharge curve of Na/MnO2 cell during continuous discharge and the curve with
intermittent discharges in the inset.
plots (Fig. 5b) suggest a resistive region at low frequency (phase angle
(θ): −2° and impedance (| Z |): 3.8 kΩ) as well as at high frequency
regions (θ: −10° and | Z|: 100 Ω), and a capacitive behavior with θ: the flow of current. Among these components, the anode and cathode
−60° at mid frequency region. The cell consists of Na metal foil as the are expected to contribute significantly to the total cell impedance. In
anode and MnO2 powder pressed on SS mesh as the cathode in the non- order to understand which of the two electrodes decide the cell im-
aqueous electrolyte. The cell impedance is expected due to the con- pedance, Na/Na and MnO2/MnO2 symmetrical cells were assembled
tribution of all cell components, which constitute the electric circuit for and their impedance spectra were recorded at the open circuit voltage,

Fig. 2. (a) Scanning electron micrograph and (b) trans-

mission electron micropgraph of MnO2 sample.

136
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

4000 9000
a a
3500
7500
3000
6000
2500
-Z'' / Ω

-Z" / Ω
2000 4500
31 Hz
1500 81 Hz 14 Hz
3000
253 Hz
1000 4.5 Hz
0.01 Hz
1 kHz 1.4 Hz 0.08 Hz
500 1500

0
0 500 1000 1500 2000 2500 3000 3500 4000 0
Z' / Ω 0 1500 3000 4500 6000 7500 9000
Z' / Ω
1000
150
b b
60 125
-Phase / Degree

800
45 100

-Z'' / Ω
30 75

600
15
-Z'' / Ω 50

0 25
400
0
3.6 0 25 50 75 100 125 150
Z' / Ω
log (|Z| / Ω)

3.2
200
2.8
2.4
0
2.0
0 200 400 600 800 1000
1.6 Z' / Ω
-2 -1 0 1 2 3 4 5
log (freq / Hz) Fig. 6. Nyquist plots of (a) Na/Na and (b) MnO2/MnO2 symmetrical cells.

Fig. 5. (a) Nyquist and (b) Bode EIS plots of as assembled Na/MnO2 cell. The measured
spectrum is shown by data points and the fit spectrum by smooth curve. Frequencies 3.20
corresponding to some data points are indicated in (a). The NLLS fit was carried out V1
between 100 KHz and 1 mHz. 3.15

which was close to 0.0 V in each case (Fig. 6). It is seen that the re-
Cell voltage / V

3.10
sistance measured from low frequency region for Na/Na symmetrical
cell is about 9 kΩ, whereas it is about 50 Ω for MnO2/MnO2 symme- 3.05 delay time
trical cell. Thus, the impedance of Na/Na cell is greater than that of
MnO2/MnO2 cell by about 180 times. MnO2 electrodes are fabricated V3
3.00
using high surface area MnO2 particles pressed on SS mesh current
collecting substrate. As a result, MnO2 electrodes have low impedance.
2.95
As Na and MnO2 electrodes in a Na/MnO2 cell are present in series
combination, the impedance characteristics of the cell (Fig. 5) are at-
2.90
tributed essentially to the contribution of Na anode of high impedance.
This is due the fact that Na metal is covered with a passivating layer in V2
2.85
contact with the electrolyte. The passive film acts as a barrier for the 0 5 10 15 20 25 30 35 40 45 50
flow of charge across the Na anode. The presence of passive film is
confirmed by analysis of discharge voltage profile, which is discussed Time / s
below in detail. Fig. 7. Voltage transient (delay action) of Na/MnO2 coin cell on initiation of galvano-
The voltage profile of Na/MnO2 coin cell at the beginning of gal- static discharge at 100 μA.
vanostatic discharge at 100 μA is shown in Fig. 7. The cell voltage drops
from the open circuit voltage of 3.17 V to 2.87 V instantaneously at the Na/MnO2 cell is 13 s (Fig. 7). This delay time is caused by the surface
commencement of galvanostatic discharge, followed by an increase protective film present on the metal, which initially acts as a barrier for
approaching about 3.0 V and then the voltage follows a normal course the passage of current through the cell. Thus, there is a decrease in the
of cell discharge. Thus, there is a time lag between the initiation of cell voltage due to the film resistance. In the non-aqueous Li primary
discharge and commencement of actual discharge. This time lag is cells consisting of soluble SOCl2 or SO2Cl2, which serves as both the
known as the ‘delay action’ or ‘delay time’ in the literature of Li metal- solvent and the cathode active material, the Li metal undergoes oxi-
as well as Mg metal-based battery systems [25–31]. The delay time of dation resulting in the formation of film of LiCl on its surface. The LiCl

137
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

film formed at the initial stages of contact of Li with the solution acts as where f# is the AC frequency corresponding to the maximum of the
a barrier to protect the underlying Li metal and also as the inter-elec- imaginary component (−Z”) on the low frequency semicircle.
trode separator. When the cell is subjected to discharge, the LiCl film However, the Nyquist plot of Na/MnO2 cell shown in Fig. 5a ap-
undergoes a break-down or rupture, thus, allowing the Li surface to pears as a single semicircle. The semicircle is depressed and distorted,
make contact with the electrolyte and undergo oxidation. A delay ac- which is likely due to porous nature of the surface film. The EIS spec-
tion is observed at this stage [25,26]. In a similar way, aqueous based trum of fresh Na/MnO2 cells (Fig. 5) was subjected to non-linear least
Mg/MnO2 cells also exhibit delay action when the oxide film present on squares (NLLS) fitting procedure using the equivalent circuit (Fig. 8). As
Mg breaks down on initiation of cell discharge [29–31]. The delay time the low frequency data do not exhibit Warburg, the Warburg element is
behavior observed for Na/MnO2 cell in the present study (Fig. 7) is excluded for the purpose of data fit. The capacitances are replaced by
similar to Li/SOCl2 or Li/SO2Cl2 cells [26,27]. Thus, it is inferred that constant phase elements (CPE). A CPE is defined as
the Na foil used as the anode in Na/MnO2 cell is covered with a pas-
sivating layer, which acts as a barrier for the electrochemical oxidation ZCPE = 1 ((jω) nY) (5)
of Na in the non-aqueous electrolyte. Thus, the high impedance ob-
where ω = 2πf, f being the frequency and n is a constant. For n = 0, the
served in Fig. 5 for Na/MnO2 cell and also in Fig. 6a for Na/Na sym-
CPE represents a resistance with R = Y−1; for n = 1, a capacitance
metrical cell is associated with the resistance of passivating film. On
with C = Y; for n = 0.5, a Warburg and for n = −1, inductance with
initiation of galvanostatic discharge, the surface film undergoes di-
L = Y−1 [32]. The necessity of using CPE in the place of capacitance
electric breakdown or rupture at weak spots resulting in exposure of
arises because of its microscopic distribution over finite size electrodes,
film-free regions of Na surface. It is expected that now a fraction of Na
which possess a large number of surface defects such as kinks, jags, and
metal surface is covered with film and the rest (film-free region) of the
edges, local charge inhomogeneities, two-or-three-phase regions, ad-
surface can allow the electrochemical oxidation of Na.
sorbed species, and variations in composition and stoichiometry [32].
The reaction expected at the film-free region of the Na anode during
The impedance parameters obtained from the NLLS fit procedure are
discharge of Na/MnO2 cell is given in Eq. (2).
listed in Table 1. The theoretical spectrum generated from the fit pro-
Na → Na+ + e− (2) cedure is shown as smooth curves in both Nyquist and Bode plots and
The electrical equivalent circuit appropriate for the Na electrode the experimental data points are presented as open circles. It is seen
with film covered and film free regions is shown in Fig. 8a. The film that the theoretical spectrum passes through the data points and the χ2
covered region is characterized by a parallel combination of the film of NLLS fit is 1 × 10−4. These observations suggest the impedance
resistance (Rf) and its capacitance (Cf). The surface of Na exposed to the parameters obtained are reliable. The total resistance (Rs + Rf + Rct) is
electrolyte is characterized by charge-transfer resistance (Rct) of reac- 4210 Ω. This value is close to the high frequency intercept of Bode plot
tion (2), diffusion controlled Warburg resistance (W) and double-layer (Fig. 5b). The phase angle is about −60° suggesting capacitance nature
capacitance (Cdl). The Nyquist plot of impedance is expected to consist of the system in the frequency range around 500 Hz.
of two semicircles and a linear spike as shown in Fig. 8b. The high A Na/MnO2 cell was discharged at C/7 rate to depth of discharge
frequency impedance of the Nyquist plot provides the value of ohmic (DOD) of 0.04, allowed at rest for about 15 min and impedance spec-
resistance (RΩ) of the cell. The value of Rf can be calculated from the trum was recorded (Fig. 9). DOD is the ratio of charge removed from
diameter of high frequency semicircle and film capacitance Cf from the the cell by discharge to the total capacity of the cell. It is seen from both
following relationship. Nyquist and Bode plots that the cell resistance has decreased to 500 Ω
from 4210 Ω measured before the discharge (Fig. 5). At the com-
Cf = 1⁄ (2πf ∗Rf ) (3) mencement of discharge, the surface film is ruptured exhibiting voltage
where f* is the AC frequency corresponding to the maximum in the dip and delay action (Fig. 7) resulting in a decreased coverage of the
imaginary component (−Z”) on the high frequency semicircle. Simi- surface film. A large fraction of the Na surface is free from the surface
larly, Rct can be calculated from the diameter of the high frequency film and therefore the film resistance is expected to decrease. The EIS
semicircle and Cdl from the following equation. spectrum was subjected to NLLS fit procedure, and the fit parameters
are listed in Table 1. As expected, there is a decrease of Rf to 112 Ω from
Cdl = 1⁄ (2πf #R ct) (4) 1620 Ω measured before the cell is subjected to a short discharge.

Fig. 8. (a) Equivalent circuit and (b) Nyquist representation of

EIS.

138
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

Table 1
Impedance parameters of a Na/MnO2 cell at different depth of discharge values. Yf and nf correspond to constant phase element of surface film on Na; Ydl and ndl correspond to constant
phase element of double-layer, χ2 is goodness of NLLS fitting and other symbols are defined in the text.

DOD RΩ (Ω) Rf (Ω) Yf (F snf − 1) nf Rct (Ω) Ydl (F sndl − 1) ndl f⁎ (Hz) χ2

0.000 60.0 1620.0 8.96 × 10− 5 0.34 2530.00 2.75 × 10− 6 0.890 37.56 0.00014
0.036 52.6 112.6 6.85 × 10− 5 0.434 343.80 5.85 × 10− 6 0.859 214.80 0.00011
0.071 59.5 87.3 3.90 × 10− 5 0.494 192.70 7.41 × 10− 6 0.849 371.10 0.00011
0.214 56.8 88.8 2.36 × 10− 5 0.534 127.10 1.04 × 10− 5 0.819 664.10 0.00002
0.357 60.8 83.4 9.63 × 10− 6 0.519 78.79 1.31 × 10− 5 0.899 966.80 0.00002
0.500 63.8 86.3 6.58 × 10− 6 0.695 65.69 1.44 × 10− 5 0.811 966.90 0.00005
0.643 69.7 75.9 2.70 × 10− 6 0.748 61.56 1.16 × 10− 5 0.783 1172.00 0.00009
0.786 67.3 87.5 3.23 × 10− 6 0.776 48.53 1.07 × 10− 5 0.769 1465.00 0.00013
0.929 71.7 88.8 2.58 × 10− 6 0.789 46.24 1.28 × 10− 5 0.742 1465.00 0.00016
1.000 71.1 88.3 2.32 × 10− 6 0.781 52.45 1.41 × 10− 5 0.761 1172.00 0.00015

600
a 300
a
500 250

400 200
-Z'' / Ω

-Z'' / Ω
300 150
1 kHz 312 Hz
200
3.7 kHz 100 0.01 Hz
69 Hz 0.01 Hz
21 kHz 8 kHz 1.4 kHz 0.03 Hz
81 kHz
100 81 kHz 175 Hz 0.3 Hz
50

0
0 100 200 300 400 500 600 0
Z' / Ω 0 50 100 150 200 250 300
Z' / Ω
35
b
-Phase / Degree

30
25 15 b
-Phase / Degree

20 12
15
9
10
5 6
0 3
0

2.8
2.5
2.6 2.4
log (Z / Ω)

log (Z / Ω)

2.4 2.3
2.2 2.2
2.0 2.1
1.8 2.0
-2 -1 0 1 2 3 4 5 1.9
log (freq / Hz) -2.1 -1.1 0.1 1.1 2.1 3.1 4.1 5.1
log (freq / Hz)
Fig. 9. (a) Nyquist and (b) Bode plots of Na/MnO2 cell at DOD = 0.04. The measured
spectrum is shown by data points and the fit spectrum by smooth curve. Frequencies Fig. 10. (a) Nyquist and (b) Bode plots of Na/MnO2 cell at DOD = 0.8. The measured
corresponding to some data points are indicated in (a). spectrum is shown by data points and the fit spectrum by smooth curve. Frequencies
corresponding to some data points are indicated in (a).

Assuming Rf is a measure of coverage (Θ) of surface film, Θ can be

defined as density (io) of reaction (2) is calculated using Eq. (7)

Θ = Rf(DOD) Rf(initial) (6) i o = −RT (FAR ct) (7)

The surface coverage on Na electrode at DOD = 0.04 becomes 0.07. where R is gas constant, T is absolute temperature, F is Faraday con-
This indicates that a large fraction of area of Na electrode has become stant and A is electrode area. As Θ is very low as discussed above, A is
free from its surface film after initiation of galvanostatic discharge. approximated to the geometrical area of Na electrode. From the Rct
Similar to the decrease in Rf, there is a decrease in Rct from 2530 Ω value of 344 Ω, the value of io calculated at 25 °C for Na electrode, thus,
before discharge to 344 Ω at DOD = 0.04. A substantial decrease in Rct becomes 38 μA cm−2 at DOD = 0.04 (it is assumed that the area of Na
indicates that the reaction (2) in the non-aqueous electrolyte is more electrode does not undergo variation by cell discharge).
facile after the dielectric rupture of surface film. The exchange current The Na/MnO2 cell was subjected to discharge at C/7 rate and EIS

139
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

100
a 1600 200 b

Film resistance (Rf) / Ω

Solution resistance (Rs) / Ω
80 1400 160

Film resistance (Rf) / Ω

1200 120

60 1000 80

800
40
40
600
0
0.0 0.2 0.4 0.6 0.8 1.0
400 Depth of discharge
20
200

0 0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
a Depth of discharge Depth of discharge

1600
c d
Charge transfer resistance (Rct) / Ω

Charge transfer resistance (Rct) / Ω

2500 500
1400

400 1200
2000
300 1000
1500

f* / Hz
200 800

1000 100 600

0 400
500 0.0 0.2 0.4 0.6 0.8 1.0
Depth of discharge
200
0
0
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Depth of discharge Depth of discharge
Fig. 11. Dependence of (a) RΩ, (b) Rf, (c) Rct, and (d) f⁎ of Na/MnO2 cell on depth of discharge. The Y-axes between DOD = 0.04 and 1.0 are expanded and shown in insets of (b) and (c).

data were measured at several stages of discharge. The voltage versus τ = Rf Cf = 1 2πf ∗ (8)
time of discharge curve (Fig. 4 inset) shows that the cell voltage in- ⁎
an increase in f indicates a decrease in τ of surface film present on Na.
creases by about 0.20 V at open-circuit conditions from the closed
discharge circuit condition after terminating the discharge at the re- 4. Conclusions
quired DOD. The EIS measurements were made at the equilibrated
open-circuit cell voltage. The two overlapped semicircles as observed Primary Na/MnO2 cells are fabricated employing amorphous MnO2
above at DOD ≈ 0.04, were seen well separated at higher DODs. Ny- and electrochemical impedance measurements are carried out at dif-
quist and Bode plots recorded at DOD ≈ 0.80, for instance, are pre- ferent stages of cell discharge. The cell impedance is dominated by Na
sented in Fig. 10. The presence of two semicircles and the low fre- metal anode due to the surface film present on its surface initially for an
quency linear Warburg region are clearly seen in the Nyquist plot undischarged cell. On initiation of cell discharge, the surface passive
(Fig. 10a). The Bode plots also reflect a pair of time constants. The film undergoes dielectric breakdown resulting in a significant decrease
impedance parameters obtained from fitting of the data at all DODs are in cell impedance. The impedance parameters are examined in view of
presented in Table 1. The values of f⁎ corresponding to the high fre- identifying a suitable parameter that exhibits a strong dependence on
quency semicircle are also listed in Table 1. The variations of RΩ, Rf, Rct depth of discharge of the cell. The frequency maximum f⁎ increases
and f⁎ with DOD are presented in Fig. 11. It is seen that there is an from about 100 Hz for the undischarged cell to about 1400 Hz for a
increase in ohmic resistance (RΩ) from 60 Ω to about 70 Ω when the completely discharged cell. Thus, f⁎ is a suitable impedance parameter
Na/MnO2 cell is discharged completely. This marginal linear increase which strongly depends on the depth of discharge of Na/MnO2 primary
(Fig. 11a) is likely due to variation in electrolyte composition and also cell.
die to progressive percolation of the electrolyte into the pores of
amorphous MnO2 cathode powder. Although there is a significant de- References
crease in Rf of undischarged cell from 1600 Ω to about 100 Ω of a
slightly discharged (DOD = 0.04) cell due to surface film break down, [1] N. Yabuuchi, K. Kubota, M. Dahbi, S. Komaba, Research development on sodium-
the variation of this parameter during the subsequent cell discharge is ion batteries, Chem. Rev. (2014), http://dx.doi.org/10.1021/cr500192f.
not significant (Fig. 11b). There is a slight decrease in Rf when the cell [2] V. Palomares, M. Casas-Cabanas, E. Castillo-Martínez, M.H. Han, T. Rojo, Update on
Na-based battery materials. A growing research path, Energy Environ. Sci. 6 (2013)
undergoes discharge from DOD = 0.04 to DOD = 1.00. Thus, there is a
2312–2337, http://dx.doi.org/10.1039/c3ee41031e.
marginal decrease in coverage of surface film on Na metal. Similar to [3] M. Sawicki, L.L. Shaw, Advances and challenges of sodium ion batteries as post
the decrease of Rf, there is a decrease in Rct also (Fig. 11c). Although the lithium ion batteries, RSC Adv. 5 (2015) 53129–53154, http://dx.doi.org/10.1039/
variations in Rf and Rct are not significant to indicate DOD of Na/MnO2 C5RA08321D.
[4] J. Tang, A.D. Dysart, V.G. Pol, Advancement in sodium-ion rechargeable batteries,
cells, the frequency maximum f⁎ exhibits substantial variation with Curr. Opin. Chem. Eng. 9 (2015) 34–41, http://dx.doi.org/10.1016/j.coche.2015.
DOD (Fig. 11d). The value of f⁎ is about 20 Hz for an undischarged cell, 08.007.
and it gradually and nearly linearly increases to about 1400 Hz at [5] D. Linden, T.B. Reddy (Eds.), Linden's Handbook of Batteries, fourth ed., Mcgraw-
hill, 2010.
DOD = 0.9. Thus, f⁎ is expected as useful for predicting DOD of Na/ [6] B. Kishore, G. Venkatesh, N. Munichandraiah, A Na/MnO2 primary cell employing
MnO2 cell. As f⁎ is related to time constant (τ) of the film capacitance as poorly crystalline MnO2, J. Electrochem. Soc. 162 (2015) A839–A844, http://dx.

140
B. Kishore, M. Nookala Journal of Electroanalytical Chemistry 799 (2017) 134–141

doi.org/10.1149/2.0301506jes. Cryst. Growth 310 (2008) 716–722, http://dx.doi.org/10.1016/j.jcrysgro.2007.11.

[7] Q. Xu, J.J. Davis, The diagnostic utility of electrochemical impedance, 113.
Electroanalysis 26 (2014) 1249–1258, http://dx.doi.org/10.1002/elan.201400035. [20] S.C. Pang, S.F. Chin, C.Y. Ling, Controlled synthesis of manganese dioxide nanos-
[8] A. Nechache, M. Cassir, A. Ringuedé, Solid oxide electrolysis cell analysis by means tructures via a facile hydrothermal route, J. Nanomater. (2012) ID607870, http://
of electrochemical impedance spectroscopy: a review, J. Power Sources 258 (2014) dx.doi.org/10.1155/2012/607870.
164–181, http://dx.doi.org/10.1016/j.jpowsour.2014.01.110. [21] R.N. DeGuzman, Y.-F. Shen, E.J. Neth, S.L. Suib, C.-L. O'Young, S. Levine,
[9] M. Mahbuburrahman, N. Chandradebnath, J.J. Lee, Electrochemical impedance J.M. Newsam, Synthesis and characterization of octahedral molecular sieves (OMS-
spectroscopic analysis of sensitization-based solar cells, Isr. J. Chem. 55 (2015) 2) having the Hollandite structure, Chem. Mater. 6 (1994) 815–821, http://dx.doi.
990–1001, http://dx.doi.org/10.1002/ijch.201500007. org/10.1021/cm00042a019.
[10] D. Kashyap, P.K. Dwivedi, J.K. Pandey, Y.H. Kim, G.M. Kim, A. Sharma, S. Goel, [22] Y. Chen, C. Liu, F. Li, H.M. Cheng, Preparation of single-crystal α-MnO2 nanorods
Application of electrochemical impedance spectroscopy in bio-fuel cell character- and nanoneedles from aqueous solution, J. Alloys Compd. 397 (2005) 282–285,
ization: a review, Int. J. Hydrog. Energy 39 (2014) 20159–20170, http://dx.doi. http://dx.doi.org/10.1016/j.jallcom.2004.12.049.
org/10.1016/j.ijhydene.2014.10.003. [23] S. Devaraj, N. Munichandraiah, Effect of crystallographic structure of MnO2 on its
[11] M. Li, Li-ion dynamics and state of charge estimation, Renew. Energy 100 (2016) electrochemical capacitance properties, J. Phys. Chem. C 112 (2008) 4406–4417,
44–52, http://dx.doi.org/10.1016/j.renene.2016.06.009. http://dx.doi.org/10.1021/jp7108785.
[12] S. Kumar, P.K. Nayak, K.S. Hariharan, N. Munichandraiah, Temperature and po- [24] Z. Huang, W. Zhou, C. Ouyang, J. Wu, F. Zhang, J. Huang, Y. Gao, J. Chu, High
tential dependence electrochemical impedance studies of LiMn2O4, J. Appl. performance of Mn-Co-Ni-O spinel nanofilms sputtered from acetate precursors, Sci.
Electrochem. 44 (2014) 61–71, http://dx.doi.org/10.1007/s10800-013-0601-y. Rep. 5 (2015) 10899, http://dx.doi.org/10.1038/srep10899.
[13] E. Panabière, N. Emery, S. Bach, J.-P. Pereira-Ramos, P. Willmann, A kinetic study [25] A.N. Dey, Lithium anode film and organic and inorganic electrolyte batteries, Thin
of electrochemical lithium insertion in Li7MnN4 by impedance spectroscopy, J. Solid Films 43 (1977) 131–171, http://dx.doi.org/10.1016/0040-6090(77)
Alloys Compd. 663 (2016) 624–630, http://dx.doi.org/10.1016/j.jallcom.2015.12. 90383-2.
121. [26] A. Leef, A. Gilmour, Voltage delay in lithium non-aqueous battery systems, J. Appl.
[14] F. Paloukis, C. Elmasides, F. Farmakis, P. Selinis, S.G. Neophytides, N. Georgoulas, Electrochem. 9 (1979) 663–669, http://dx.doi.org/10.1007/BF00614959.
Electrochemical impedance spectroscopy study in micro-grain structured amor- [27] K. Il Chung, J.S. Lee, Y.O. Ko, Electrical analysis of Li/SOCl2 cell connected with
phous silicon anodes for lithium-ion batteries, J. Power Sources 331 (2016) electrochemical capacitor, J. Power Sources 140 (2005) 376–380, http://dx.doi.
285–292, http://dx.doi.org/10.1016/j.jpowsour.2016.09.062. org/10.1016/j.jpowsour.2004.08.043.
[15] T. Osaka, D. Mukoyama, H. Nara, Review—development of diagnostic process for [28] I. Nainville, A. Lemarchand, J.P. Badiali, Passivation of a lithium anode: a simu-
commercially available batteries, especially lithium ion battery, by electrochemical lation model, Electrochim. Acta 41 (1996) 2855–2863, http://dx.doi.org/10.1016/
impedance spectroscopy, J. Electrochem. Soc. 162 (2015) A2529–A2537, http://dx. 0013-4686(96)00113-2.
doi.org/10.1149/2.0141514jes. [29] B.V. Ratnakumar, S. Sathyanarayana, The delayed action of magnesium anodes in
[16] J.J. Winter, J.T. Breslin, R.L. Ross, H.A. Leupold, F. Rothwarf, A state-of-charge test primary batteries, part I. Experimental studies, J. Power Sources 10 (1983)
for zinc-mercuric oxide primary cells, J. Electrochem. Soc. 1421 (1975) 1434–1438. 219–241, http://dx.doi.org/10.1016/0378-7753(83)80078-0.
[17] M.L. Gopikanth, S. Sathyanarayana, Impedance parameters and the state-of-charge. [30] S. Sathyanarayana, B.V. Ratnakumar, The delayed action of magnesium anodes in
III. Zinc-manganese dioxide and magnesium-manganese dioxide dry batteries, J. primary batteries. Part II. Theoretical studies, J. Power Sources 10 (1983) 243–261,
Appl. Electrochem. 9 (1979) 581–585, http://dx.doi.org/10.1007/BF00610945. http://dx.doi.org/10.1016/0378-7753(83)80079-2.
[18] N. Munichandraiah, Electrochemical impedance studies of a decade-aged magne- [31] B.V. Ratnakumar, Passive films on magnesium anodes in primary batteries, J. Appl.
sium/manganese dioxide primary cell, J. Appl. Electrochem. 29 (1999) 463–471. Electrochem. 18 (1988) 268–279, http://dx.doi.org/10.1007/BF01009274.
[19] X. Zhang, W. Yang, J. Yang, D.G. Evans, Synthesis and characterization of α-MnO2 [32] J.R. Macdonald, Impedance spectroscopy, Ann. Biomed. Eng. 20 (1992) 289–305,
nanowires: self-assembly and phase transformation to β-MnO2 microcrystals, J. http://dx.doi.org/10.1007/BF02368532.

141