Journal of Power Sources 244 (2013) 196e206

Contents lists available at SciVerse ScienceDirect

Journal of Power Sources

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

Technological potential and issues of polyacrylonitrile based nanober

non-woven separator for Li-ion rechargeable batteries
Yeon-Joo Kim a, Hyun-Soo Kim a, Chil-Hoon Doh a, Seok Hong Kim b, Sang-Min Lee a, *
a
b

Battery Research Center, Korea Electrotechnology Research Institute, Changwon 641-120, Republic of Korea
R&D Center, Finetex EnE, 350, Bukyang-dong, Hwasung-si, Gyeonggi-do, Republic of Korea

h i g h l i g h t s
< PAN separator has higher ionic conductivity and lower thermal shrinkage.
< Battery employing PAN separator shows superior performance over PE separators one.
< PAN separator couldnt prevent metal complex from being deposited on anode surface.

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 11 October 2012
Received in revised form
25 January 2013
Accepted 29 January 2013
Available online 6 February 2013

We investigate on the technological potential and issues of non-woven separator as an alternative to

conventional separator in terms of various evaluation points. Thus, optimal evaluation systems for new
nonwoven polyacrylonitrile (PAN) separator are proposed to compare with commercialized polyethylene
(PE) separator, and corresponding stacking-type pouch batteries are used as an evaluation tool. Cell
performance, electrochemical stability, and thermal stability for both separators are compared. Based on
the comparison results, each separators effects are discussed in detail. In comparison to PE separator, the
non-woven separator exhibits distinct advantage in microporous structure, leading to superior reliability
of cell. Based on the understanding of non-woven separator, cell performances of the separator at different charge/discharge conditions are discussed in terms of ion transport of the separator and AC
impedance of the cell. Moreover, the problems of nonwoven separator to be solved are clearly stated
through the phenomenological analyses. The overall evaluation results indicate that the non-woven PAN
separator has many advantages over the PE separator, which, in turn, contribute to superior cell performances. Therefore, the PAN nonwoven membrane can be the most promising candidate for separators
of the next generation of lithium-ion battery if the problems to be stated will be solved.
2013 Elsevier B.V. All rights reserved.

Keywords:
Lithium-ion battery
Micro-porous membrane separator
Non-woven separator
Ion tranport

1. Introduction
Lithium-ion batteries are widely used for electronic devices
such as mobile phones, laptop computers, and digital cameras due
to their high energy density and long life cycle. With the search for
a solution as alternative propulsion system, lithium-ion batteries
are also expected to be an alternative power source for Plug-in
Hybrid Electric Vehicle (PHEV), particularly with increasing needs
for alternative energy and environment protection, due to the advantages in longer cycle life, higher voltages and energy density
when compared with other rechargeable batteries [1e6].

* Corresponding author. Tel.: 82 55 280 1663; fax: 82 55 280 1590.

E-mail addresses: sangma@keri.re.kr, sangminlee71@gmail.com (S.-M. Lee).
0378-7753/$ e see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jpowsour.2013.01.166

As lithium-ion batteries used in the electric car sector require

high-capacity and high-power performance, certain safety features
(i.e. ignition, explosion, etc.) become more critical. Lithium,
a highly-reactive anode material, may cause an explosion if it
comes into direct contact with the cathode [7e13]. Therefore, an
electrically insulating membrane to separate the anode from the
cathode is critical component for the safety of a battery. Most
membranes used for lithium-ion batteries are polyolen types.
However, thermal contraction coupled with physical damage occurs when polyolen is exposed to high temperatures over 100
C
due to its physical properties and manufacturing process, which
could result in a possible short-circuiting inside a battery. There
have been numerous studies to overcome these problems related to
current polyolen separators [14,15].
A micro-porous membrane is necessary for separating positive and
negative electrodes and for maintaining liquid electrolyte between

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

both electrodes. However, the existence of a separator increases the

cell resistance by a factor and reduces the volumetric energy density of
the battery. Polyolen micro-porous membranes have been used as
major separators for lithium ion batteries. But, the rate capabilities of
the separators are not enough for high-power applications, such as
electric vehicles, hybrid electric vehicles and robots. The micro-porous
membrane separators have some disadvantages that need to be
improved, such as low wettability, low porosity of about 40% [16], and
thermal shrinkage that causes short circuits between electrodes under unusual heat generation. Recently, various approaches to overcome these shortcomings of polyolen based separators have been
reported. Most reports focused on many alternatives of polyolen
based separators such as non-woven separators and inorganic composite [17e21,24]. In contrast, non-woven separators have higher
porosity (60e90%) and higher air permeability [22] than polyolen,
which could be used to increase the rate capability of lithium ion
batteries. In addition, excellent thermal tolerance at high temperature
up to 180
C can be expected, thanks to low thermal-shrinkage.
Some notable studies focusing on non-woven, ceramic composite, co-extrusion & electrolyte afnity technologies have been
very active. Also, there have been some ndings on the application
of a PAN separator to a rechargeable lithium battery, utilizing
aramid-based non-wovens and PAN-based nano-bers. Cho et al.
[23] has reported on the electrochemical performance of PAN
nanober based non-woven separator for lithium ion batteries.
They showed the superior performances of non-woven separator
based battery in view of its reliability such as rate capability and
cycle life, etc. Furthermore, the thermal stability of battery was
reported to be remarkably improved by applying PAN based nonwoven based separator [18]. There still, however, remain some
process issues to be solved due to the inferiority in its physical
properties such as low tensile strength, and puncture strength [24].
In spite of this physical drawback, the PAN based non-woven separator has been regarded as effective for thermal battery safety and
low cost, as the demand for large-capacity batteries for electric cars
has signicantly increased. Most importantly, the PAN based nonwoven separator was not overall seen if it will be feasible component in many different aspects of battery.
In this study, an optimal evaluation system for current PE separator and new PAN separator was come up with, and corresponding stacking-type pouch batteries were designed. Batteries
with PE separators were electrically, thermally, electrochemically
and chemically compared to those with PAN separators. Based on
the comparison results, each separators effects were extensively
investigated and analyzed.
2. Experiments
2.1. Fabrication of the PAN-based non-woven separator
The non-woven separator was fabricated by electro-spinning
process. The electro-spinning solution was prepared by dissolving
polyacrylonitrile (PAN, Mw 150,000, Aldrich) powder in N, Ndimethylformamide (DMF, Aldrich) at room temperature. The prepared solution was put into 10 ml syringes with metal needles. A
power supply (SHV200/Conver Tech) was used to provide the potential difference (1.9 V) for electro-spinning process. Needle-tocollector distance was set as 18 cm and electro-spun bers were
collected on an aluminum foil. After that, the electrospun bers were
hot-pressed to be non-woven separator of 25 mm thickness at 70
C.

197

capacity of 300 mAh. Before making the full cell, the design specications for the anode/cathode/electrolyte were taken into consideration. Enhanced natural graphite materials (mostly IT & highoutput uses) were used as active materials for an anode. The core
anode material is natural graphite with coreeshell type, while the
surface of natural graphite is coated with amorphous carbon layer.
In addition, SBR/CMC water-based binders were added to attain
a loading level (20 mgcm2) per unit area and electrode density of
1.51 gcc1. Typical high-output active materials such as Li(Ni1/3Co1/
3Mn1/3)O2 (LNCMO) were used for the cathode, while super-P and
PVDF binder were used as a conducting agent and a binder,
respectively. The cathode was designed to attain a corresponding
loading level (39 mgcm2) and electrode density (3.64 gcc1). The
electrolyte specications were set at 1 M LiPF6, EC/EMC 1:2(v/
v) VC 2 wt. %. Furthermore, PE separators (provided by Celgard
Co. Ltd.) with similar thickness (w26 mm) were compared to PAN
separators in order to evaluate each separator types performance.
In order to establish a full-cell evaluation, the design specications
for a corresponding battery were set at an n/p ratio of 1.1 and
a current density of 2.5 mAcm2. In addition, there was a 2 mm
variation in the widths of the anode, cathode and separator, in order to prevent any short-circuiting between the anode and cathode.
2.3. Analysis of PAN-separator-based rechargeable lithium batteries
FE-SEM (Hitachi Co. S-4800) was used to observe the structures
of PAN separator and PE separator, and each separators electrochemical reliability, electrochemical stability and thermal stability
were accordingly evaluated.
2.3.1. Electrochemical reliability
For evaluation of electrochemical reliability, the related charge/
discharge performances were measured in a voltage range between
3.0 V and 4.2 V (by CC (constant current)eCV (constant voltage)) on
charge/discharge curves. Rate capability was evaluated at current of
0.2 C charge and 0.1 C discharge. Also, the cyclic durability was
measured at room temperature (25
C) and high temperature
(45
C) while the C-rate was set at 1 C. For the analysis on electrochemical properties of each separator, low-temperature discharge properties and high-rate discharge capacity were evaluated
and discussed with physical properties of separator. The lowtemperature discharge properties and ion conductivity were tested with 0.2 C-charging & 0.5 C-discharging between 2.75 V and
4.2 V at 20, 10, 0, and 25
C. At each temperature, the ion conductivity of each fully-charged separator was derived from AC
impedance spectra. AC impedance was analyzed for the fullycharged cell using IM6 (ZAHNER, Germany). The frequency range
was measured in 1 mHze100 kHz, and amplitude was measured in
10 mV. The high-rate discharge capacity was also tested with 0.2 Ccharging & 1 C/2 C/4 C/10 C-discharging between 2.75 V and 4.2 V.
2.3.2. Electrochemical stability
For the evaluation of electrochemical stability during cell operation, the LSV (Linear Sweep Voltammetry) analysis was conducted
on each separator. The measurement was performed in a twoelectrode electrochemical system consisting of a SUS (stainless steel
blocking electrode) working electrode, and lithium counter electrodes at room temperature. LSV was scanned between 3 V and 6 V at
a rate of 20 mV s1. The full cell with each separator was fully charged
up to 4.2 V under CC/CV trickling charging mode and then the surface
of each separator was analyzed with SEM-EDS (Hitachi Co. S-4800).

2.2. Design of PAN-separator-based rechargeable lithium batteries

PAN separators provided by Finetex EnE, Inc. were stacked by
the stack-fold method to make full-pouch cells with the design

2.3.3. Thermal stability

For the evaluation of thermal stability, each separator was
positioned between slide glasses and subjected to certain tensile

198

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

Fig. 1. SEM images of A) PE separator and B) PAN separator applied in pouch type full
cell.

Table 1
Brief physical properties of separators.
Property

PE separator

PAN separator

Thickness (mm)
Tensile strength (MPa)
Porosity (%)
Ionic conductivity (mScm1)
Electrolyte uptake (%)

26
159
44
0.63
w116

25
32
54
0.76
w174

shear stress. Then, the temperature was raised to 150
C and 180
C
at 5
C per minute, and both separators were kept for 12 h at each
temperature in order to compare their contraction percentages. In
general, thermal abuse tolerance was studied using calorimetric
techniques on full cell. In this work, ARC (Accelerated Rate Calorimetry) was used to see how the thermal contraction characteristics of separators affect the thermal tolerance of pouch full cell
[25].
A maximum temperature of 450
C was set to prevent uncontrolled thermal runaway and explosive decomposition of the cells
in the ARC apparatus (TIAX Model). The ARC is operated in a heat,
wait, search mode to detect the onset point of self-heating. The
ARC increases the systems temperature in discrete steps (heat),
waits for the thermal transients to decay (wait) and then monitors the temperature of the cell for a xed time period (search). If
the cell temperature is not increasing above a threshold value,
typically 0.02
C min1, the temperature is increased by another
step and the process repeated. If the cell temperature increases at
rate equal or above the threshold value, the ARC switches to the
exothermic mode during which the ARC temperature closely
matches the cell temperature, thus maintaining the adiabatic state.
The normal mode of operation of ARC terminates an experiment by
cooling the sample once it reaches a set upper temperature limit or
maximum temperature. The ARC experiment closely simulates

Table 2
1st and 2nd cycle data for pouch type full cell with (A) PE and (B) PAN separators.
Cycle PE separator based cell
PAN separator based cell
No
C-capacity D-capacity Efciency C-capacity D-capacity Efciency
(mAh)
(mAh)
(%)
(mAh)
(mAh)
(%)
1st
2nd

372
332

318
325

85
97

375
338

320
336

85
99

Fig. 2. Comparison of charge and discharge curves of pouch type full cell with PE and
PAN separators at charge (0.2 C)edischarge (0.1 C) rate.

a thermal abuse environment that includes moderately high temperatures for relatively long periods of time. The details on ARC
apparatus are described in Ref. [26]. In this experiment, the battery
sample were heated between 50 and 400
C at the rate of
5
C min1 in the search for self-heating at the sensitivity threshold
of 0.05
C min1.
3. Results and discussion
3.1. Comparisons of physical properties between PE and PAN based
separators
The PE separator shows (Fig. 1(A)) a net structure with many
open pores and these pores are uniformly distributed. On the other
hand, as shown in Fig. 1(B), a number of nano-bers can be
observed by SEM analysis. The bers exhibit homogeneous diameters and any observable beads do not exist on the bers. In
contrast to a PE separator having large number of small-sized micropores, of which average diameter appears to be below 0.1 mm,

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

199

Fig. 5. Comparison of Gurley index after 100th cycles at room temperature and elevated temperature.

Fig. 3. Comparison of capacity retention of pouch type full cell with PE and PAN
separators at (A) room temperature (25
C) and (B) elevated temperature (45
C).

the pristine PAN non-woven shows excessively large-sized (0.3e

0.5 mm) pores that are arbitrarily distributed between the PAN bers. It is well known that various physical properties of separator
are closely related with its morphologies. Table 1 shows brief
physical properties of the PAN separator, compared with that of the
conventional PE separator. The measured thickness of the PAN
separator is in the range of 26 mm, of which values are similar to
that of the PE separator. In spite of similar thickness, the PAN based
non-woven separator has much lower tensile strength than conventional PE separator. In general, the non-woven separator doesnt
have enough tensile strength to apply the current process of cell
fabrication due to its morphological characteristics. Thus, the stack
and folding process are employed to assemble the evaluation cell
due to lack of tensile strength of PAN separator in this work. Indeed,
the tensile strength of non-woven separator can be further
improved by optimizing several key parameters such as ber
diameter and molecular weight, etc [27,28]. In particular, the PAN
separator showed roughly 10% higher porosities than that of the PE
separator. The ion conductivity of the PAN separator reaches
7.6  101 S cm1 at 25
C, which is higher than that of the PE
separator (6.3  101 S cm1) at room temperature. Besides, the
PAN separator is easily wetted by the liquid electrolyte and also
affords larger electrolyte uptake (PAN separator w174% vs. PE
separator w116%). These physical properties of separator were
measured by referring description in previous reports [29].
3.2. Comparisons of unit cell performance between PE and PAN
based separators
3.2.1. Formation charge/discharge curves
The performances of batteries with the PE separator and PAN
separator were investigated. At the rst cycle, the cells were
Table 3
Gurley index of the PE separator and PAN separators after 100th cycle at 25
C, and
45
C.

Fig. 4. Comparison of separator morphologies after 100th cycles at room temperature

(A:PE, C:PAN) and elevated temperature (B:PE, D:PAN).

PE separator

Gurley (s)

PAN separator

Gurley (s)

As-prepared
After 100th cycle
at 25
C
After 100th cycle
at 45
C

712.9
828.9

As-prepared
After 100th cycle
at 25
C
After 100th cycle
at 45
C

30.8
32.9

894.5

40.9

200

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

charged up to 4.2 V under constant currenteconstant voltage

mode, and then discharged to 2.75 V under constant current mode.
The initial chargeedischarge cycles were tested with a 0.1 C-rate.
The following formation cycle chargeedischarge tests were
charged up to 4.2 V (0.2 C) under constant currenteconstant voltage mode, and then discharged to 3 V (0.1 C) under constant current
mode.
A cells design capacity was set at 320 mAh. A pouch batterys
0.2 C standard capacity was measured to be 101e105% of the

corresponding designed capacity (325 mAhw336 mAh), and the

battery became fully activated (reversible efciency over 99%) in
the second cycle. (Table 2) In addition, the charge/discharge
behavior was observed to be almost similar between twoseparators. Any unstable voltage proles were not observed for
the cells with the both separator as shown in Fig. 2. Both cells
shows the similar discharge capacity at the 1st cycle but in case of
PAN based cell, the reversible efciency of more than 99% was
attained earlier than that of PEs.

Fig. 6. Comparison of high-rate discharge properties of pouch type full cell with PE and PAN separators. (A) Change of discharge curves with C-rates (0.1, 1, 2, 4 and 10C) after CCeCV
charging at 0.2 C to 4.2 V at room temperature for PE based pouch type full cell. (B) Change of discharge curves with C-rates (0.1, 1, 2, 4 and 10C) after CCeCV charging at 0.2 C to
4.2 V at room temperature for PAN based pouch type full cell.

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

3.2.2. Cycle properties

To observe capacity deterioration behavior of each battery with
repeated charge and discharge at 25
C and 45
C, the electrochemical durability at the 1 C/1 C charge/discharge rate was
examined. In terms of the battery life, at room temperature (25
C),
both separators showed very similar cycle life without any capacity
fading. On the other hand, the newly developed PAN-separatorbased battery showed comparatively better durability at high
temperature (45
C), as shown in the Fig. 3.
To observe the morphological change between the two separators after 100 cycles, each battery was fully discharged at the end of

201

the 100th cycle. Then, each pouch cell was disassembled, and SEM
analysis was performed to see how the surface morphology of each
separator had been changed (due to oxidation reaction, etc.). As
shown in Fig. 4, either separators morphology was hardly changed
at room temperature. At high temperature, the PE based cell,
however, might be observed to be a little gradual deterioration in
battery life due to the blocking of air holes in separator. Although
the PAN separator had a thin organic lm on its surface, it could be
observed that this type of lm would be not enough to block the
separators air holes. Therefore, it is conrmed that the cyclic degradation was meaningfully attributed to separator pore clogging

Fig. 7. Comparison of discharge curves with different temperatures (25
C, 0
C, 10
C, and 20
C) for pouch type full cell with (A) PE and (B) PAN separators.

202

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

Table 4
Temperature dependence of discharge capacity for pouch type full cell with (A) PE
and (B) PAN separators.
Temperature

25
C
0
C
10
C
20
C

PE separator based cell

Capacity
(mAh)

CT/C25
(%)

321
282
126
16

100
87.9
39.4
4.9

PAN separator based cell

C

Capacity
(mAh)

CT/C25
(%)

322
271
156
32

100
84.1
48.4
9.9

induced by electrochemical oxidation in interface between electrode and separator during cyclic electrochemical operation.
Fig. 5 shows comparison of each separators Gurley value with
as-prepared one after 100th cycle at room temperature (25
C) and
high temperature (45
C). Less Gurley value implies that it takes
less for a corresponding separator to penetrate through a certain
volume Compared to the PE separator, the PAN separator showed
less Gurley value, which could be also expected by physical parameters of each separator. That means the air-penetration path is
reduced by a well-formed air hole structure (Table 3).
3.2.3. Rate capabilities & low-temperature discharge-ability
The high-rate and low-temperature discharge properties for
pouch stack cells with one of the separators were evaluated. As
shown in Figs. 6 and 7, the cell applying PAN based separator with
higher air porosity showed relatively better high-rate properties
and low temperature discharge-abilities. Temperature dependence
of discharge capacity for pouch type full cell with PE and PAN
separators is summarized in Table 4.
In general, at extremely high C-rate (i.e. 10 C, etc.), concentration
polarization associated with the ion conductivity inside an electrolyte dominates over an activation polarization that is triggered
by the interface resistance between the active material and the
electrolyte. The difference in discharge capacity between batteries
applying two kinds of separators seems to be getting noticeable at
10 C rate as shown in Fig. 6. It indicates the concentration polarization of batteries studied in this paper depends on separators
properties considering the rate capability of higher C-rate. Because
the concentration polarization is closely related to ionic conductivity determined by a separators morphologies, the AC impedance
of battery was measured to determine the ionic conductivity in
battery associated with a separators air hole size. The ionic conductivity (s) of battery can be derived from eq (1).

s L=Rb $S

(1)

Where L and S are thickness and area of the separator, respectively,

and Rb is the bulk resistance obtained from the AC impedance test.
The bulk resistance of battery is determined by intercept point
on real impedance between circle and x-axis at high frequency
region in AC impedance spectra. The PAN separator showed higher
ionic conductivity values that were derived from the real impedance values at high-frequency (100 kHz) than that of conventional
separator based battery, as shown in the Fig. 8. It is, therefore,
conrmed that the battery applying PAN based non-woven separator shows better high-rate and low-temperature performances
due to higher ionic conductivity. In addition, based on the
temperature-dependent impedance changes, the change of charge
transfer resistance has a close relationship with temperature rather
than separator morphologies. In particular, the changes in such
values relating to the bulk resistance components measured at high
frequency were more noticeable in case of battery with conventional separator.

Fig. 8. Comparison of Nyquist plots with different temperatures (25
C, 0
C, 10
C,

and 20
C) for pouch type full cell with (A) PE and (B) PAN separators.

Based on the AC impedance test results, therefore, the PAN

separators superior performance at high-rate and low-temperature
conditions was mainly due to the size and distribution of its air
holes.
3.3. Electrochemical stability
3.3.1. Electrochemical window stability (linear sweep voltammetry
study)
In order to see how the PAN separator electrochemically reacts
at high-voltage conditions (i.e. electrochemical oxidation condition), the related tests were implemented from two perspectives.

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

First, in order to observe the oxidation reaction and side reaction at

certain high-voltage conditions, a cell with SS (Working electrode:
Stainless steel blocking electrode)/separator-electrolyte (with
LiPF6, w/o LiPF6)/Li (Reference electrode, counter electrode) was
made for the LSV (Linear Sweep Voltammetry) test. As shown in the
Fig. 9(A), there was almost no change in the onset voltage, which
indicates the sudden occurrence of an oxidation reaction. Then, an
LSV test that excluded LiPF6 to remove any oxidation reaction
current, which might be triggered by anion in dissociated salt, was
conducted. As shown in Fig. 9(B), the very little parasitic current
was generated for both separators even at high voltages up to 6 V,
implying that there was no additional oxidation reaction triggered
by the PAN separator.

Fig. 9. Linear sweep voltammograms of each cell with SS/separator/Li pole at a scan
rate 20 mV s1 applying EC:EMC 1:2 (v/v) VC 2 wt. % (A) with LiPF6, and (B)
without LiPF6.

203

3.3.2. Metal ion transfer

A full cell for each separator was made with graphite/LNCMO
and tested for trickling charging, which is a way of charging a battery at constant voltage (i.e. 4.3 V and 4.5 V) until the current
reaches zero. In addition, SEM/EDS analysis was conducted to check
for any metal dissolution on a fully-charged anodes surface or any
metal dissolution on a corresponding cathodes surface. This test is
used to prove that a PAN separators air hole size can be regarded as
a positive dynamical property like high-rate and low-temperature
properties, but is closely related to the movement of undesired
metal ions triggered by the shuttle mechanism.
As shown in the Figs. 10 and 11, after continuous charging at 4.2 V
for hours, the PE based cell had no metal complex deposition on the
interface between anode and separator, while the PAN based one
had a trace of metal complex deposition, especially for such metals
as Mn, Co and Ni. A similar phenomenon occurred when the battery
life tests were repeated at room temperature and high temperature.
Metal ion elution phenomenon triggered by the oxidation reaction
between a fully-charged anodes surface and the electrolyte surface
cannot be avoided, and this phenomenon becomes more apparent at
high temperature and high-voltage conditions. Therefore, the PAN
separator normally goes through a more inuential metal ion
transition phenomenon, and this kind of phenomenon becomes
more serious at certain conditions (i.e. high temperature, repeated
battery life evaluation, abnormal charging, etc.).

Fig. 10. Current proles during continuous charging at 4.2 V to zero current for pouch
type full cell with (A) PE and (B) PAN separators.

204

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

Fig. 11. Energy Dispersive Spectroscopy (EDS) analysis on anode surface after continuous charging at 4.2 V to zero current for pouch type full cell with (A) PE and (B) PAN separators.

3.4. Thermal stability

3.4.1. Thermal contraction
In order to evaluate certain heat-resistant properties of separators developed in this study, test conditions similar to the 150
C

hot-box test (a standard test to evaluate the physical conditions of

a jelly roll inside a real battery and its safety) were established. Each
separator was positioned between slide glasses and subjected to
certain tensile shear stress. Then, the temperature was raised up to
150
C and 180
C at 5
C per minute, and both separators were kept

Fig. 12. Comparison of thermal contraction behavior for each separator after 12 h at 150
C, and 180
C.

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

parameters on thermal safety of battery can be dened from temperature behavior with heating time. As shown in Fig. 13(A),
regarding runaway temperature, the thermal runaway phenomenon of PAN based battery occurs at higher onset temperature
(159
C) than that of PE based battery (141
C). In addition, the PAN
based battery has lower self-exothermic reaction rate (i.e. PAN
(0.78
C min1) < PE (2.46
C min1)) at 130
C, thereby showing
retardation effects on temperature and time until a thermal runaway of battery actually occurs (Fig. 13(B)). Therefore, based on
the PAN separators thermal contraction properties, it can be concluded that its thermal stability is much better than that of PE
separator.

10000
1000

Temperature rate (deg C/min)

205

100
10
1
0.1

4. Conclusion

PE separator

0.01

PAN separator

0.001
0

100

200

300

Temperature (C)

400

500

(A)
500
PE separator
PAN separator

Temperature (C)

400

300

Cell performance, electrochemical stability, and thermal stability properties for both PE and PAN separators were extensively
compared and analyzed. The overall evaluation result indicates
that the cell applying PAN based non-woven separator showed
superior performance over the PE separators one. The better
performances of PAN based cell was conrmed to be mainly
attributed to higher ionic conductivity and lower thermal shrinkage properties of PAN based non-woven separator compared with
conventional PE separator. However, in the continuous charging
test, the PE separator didnt allow any transfer of metal ion, which
was dissolved by oxidation reaction on the cathode surface, across
air holes in separator, while the cell applying PAN separator had
metal complex deposition on the anode surface, especially for such
metals as Mn, Co and Ni. Going forward with this study, it is
necessary to develop a composite separator that can prevent any
metal transfer followed by deposition phenomenon while maintaining the PAN separators positive performance, based on higher
ionic conductivity.

200
Acknowledgment
This work was supported by the Korea Evaluation Institute of
Industrial Technology funded by the Ministry of Knowledge
Economy (MKE-2012-10040033) and the National Research Foundation of Korea Grant funded by the Korean Government (MEST)
(NRF-2011-C1AAA001-0030538).

100

0
0

100

200

300

400

Time (min)

(B)
Fig. 13. Comparison of (A) exothermic rate and (b) thermal run-away behavior for
pouch type full cell during accelerated rate calorimetric test.

References
[1]
[2]
[3]
[4]
[5]

for 12 h at each temperature in order to compare their contraction

percentages. As shown in the Fig. 12, the PAN separator showed
a contraction percentage close to 0% at 150
C, while the PE separator showed a contraction percentage greater than 10% in both
width and length. Moreover, at 180
C, the PAN separator showed
a contraction percentage equal to only 4%, while the PE separator
showed a contraction percentage greater than 25%.
3.4.2. ARC behavior
In order to check the actual effects of thermal contraction
behavior on a pouch full cell, ARC (Accelerated Rate Calorimetry)
tests were performed. This test is designated to check the relative
heat safety of battery. The ARC behavior of battery is mainly characterized by onset temperature for thermal runaway and selfexothermic heating rate at corresponding temperature. These key

[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]

M. Armand, J.M. Tarascon, Nature 451 (2008) 652.

D. Linden, T.B. Reddy, Handbook of Batteries, third ed., McGraw-Hill, 2002.
F. Croce, G.B. Appetecchi, L. Persi, B. Scrosati, Nature 394 (1998) 456.
G. Venugopal, J. Moore, J. Howard, S. Pendalwar, Journal of Power Sources 77
(1999) 34.
B.L. Luan, G. Campbell, M. Gauthier, X.Y. Liu, I. Davidson, J. Nagata, M. Lepinay,
F. Bernier, S. Argue, ECS Transactions 25 (2010) 59.
R.J. Brodd, H.M. Friend, J.C. Nardi, Lithium Ion Battery Technology, ITE-JEC
Press, 1995.
U. von Sacken, E. Nodwell, A. Sundher, J.R. Dahn, Solid State Ionics 69
(1994) 284.
S. Tobishima, J.I. Yamaki, Journal of Power Sources 81 (1999) 882.
J.I. Yamaki, in: M. Wakihara, O. Yamamoto (Eds.), Lithium Ion Batteries, 83,
Kodansha and Wiley-VCH, Tokyo (Japan), 1998.
J. Cho, Y.J. Kim, T.J. Kim, B. Park, Chemistry of Materials 13 (2001) 18.
H.J. Kweon, S.J. Kim, D.G. Park, J. Power Sources 88 (2000) 255.
R.A. Leising, M.J. Palazzo, E.S. Takeuchi, K.J. Takeuchi, Journal of the Electrochemical Society 148 (2001) A838.
A.N. Dey, Journal of the Electrochemical Society 118 (1971) 1547.
A.M. Christie, S.J. Lilley, E. Staunton, Y.G. Andreev, P.G. Bruce, Nature 433
(2005) 50.
A. Manuel Stephan, European Polymer Journal 42 (2006) 21.
F.G.B. Ooms, E.M. Kelder, J. Schoonman, N. Gerrits, J. Smedinga, G. Callis,
Journal of Power Sources 97e98 (2001) 598.
H.S. Jeong, J.H. Kim, S.Y. Lee, Journal of Materials Chemistry 20 (2010) 9180.
T.H. Cho, M. Tanaka, H. Onishi, Y. Kondo, T. Nakamura, H. Yamazaki, S. Tanase,
T. Sakai, Journal of Power Sources 181 (2008) 155.

206

Y.-J. Kim et al. / Journal of Power Sources 244 (2013) 196e206

[19] Y. Lee Min, J.W. Kim, N.S. Choi, J. Lee An, W.H. Seol, J.K. Park, Journal of Power
Sources 139 (2005) 235.
[20] M. Tanaka, T.H. Cho, T. Nakamura, T. Tarao, M. Kawabe, T. Sakai, Electrochemistry 78 (2010) 982.
[21] Y. Liang, L. Ji, B. Guo, Z. Lin, Y. Yao, Y. Li, M. Alcoutlabi, Y. Qiu, X. Zhang, Journal
of Power Sources 196 (2011) 436.
[22] P. Kritzer, J.A. Cook, Journal of the Electrochemical Society 154 (2007) A481.
[23] T.H. Cho, T. Sakai, S. Tanase, K. Kimura, Y. Kondo, T. Tarao, M. Tanaka, Electrochemical and Solid-State Letters 10 (7) (2007) A159.

[24] T.H. Cho, M. Tanaka, H. Ohnishi, Y. Kondo, M. Yoshikazu, T. Nakamura, T. Sakai,

Journal of Power Sources 195 (13) (2010) 4272.
[25] E.P. Roth, D.H. Doughty, Journal of Power Sources 128 (2004) 308e318.
[26] I. Uchida, H. Ishikawa, M. Mohamedi, M. Umeda, Journal of Power Sources 821
(2003) 119e121.
[27] P.X. Ma, R. Zhang, Journal of Biomedical Materials Research 46 (1999) 60.
[28] D.H. Reneker, I. Chun, Nanotechnology 7 (1996) 216.
[29] K.J. Kim, J.H. Kim, M.S. Park, H.K. Kwon, H.S. Kim, Y.J. Kim, Journal of Power
Sources 128 (2012) 298e302.