The rational design of a redox-active mixed ion/electron conductor as a multi-functional binder for lithium-ion batteries

Article abstract of DOI:10.1039/d0ta08913c

A redox-active mixed ion and electron conductor (redox-active MIEC) is presented as a binder for the lithium titanate anodes of lithium-ion batteries. The redox-active MIEC binder (symbolized by PT*-G_mC_n) was designed to be (1) electrically conductive along its conjugated thiophene backbone (PT = polythiophene), (2) redox-active from its succinimide moiety (* = redox-active) and (3) ionically conductive by adopting glyme (G) branches. It was superior to the practically used PVdF binder in terms of lithium ion diffusivity and electrical conductivity (1.4× and 15?000×, respectively). High capacity was guaranteed, particularly at high rates due to its MIEC nature of PT*-G_mC_n, while an additional capacity was achieved from its redox activity.

Full text of DOI:10.1039/d0ta08913c

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
The rational design of a redox-active mixed ion/
electron conductor as a multi-functional binder for
lithium-ion batteries†
Cite this: J. Mater. Chem. A, 2021, 9,
4751
Chihyun Hwang,‡a Jungho Lee,‡ab Jihong Jeong,a Eunryeol Lee,a Jonghak Kim,a
a
a
Seoyoung Kim,a Changduk Yang
and Hyun-Kon Song
*
*
A redox-active mixed ion and electron conductor (redox-active MIEC) is presented as a binder for the
lithium titanate anodes of lithium-ion batteries. The redox-active MIEC binder (symbolized by PT*-GmCn)
was designed to be (1) electrically conductive along its conjugated thiophene backbone (PT
¼
polythiophene), (2) redox-active from its succinimide moiety (* ¼ redox-active) and (3) ionically
conductive by adopting glyme (G) branches. It was superior to the practically used PVdF binder in terms
of lithium ion diffusivity and electrical conductivity (1.4ꢀ and 15 000ꢀ, respectively). High capacity was
guaranteed, particularly at high rates due to its MIEC nature of PT*-GmCn, while an additional capacity
was achieved from its redox activity.
Received 10th September 2020
Accepted 14th January 2021
DOI: 10.1039/d0ta08913c
rsc.li/materials-a
polyaniline and poly(3,4-ethylene dioxythiophene) were effec-
tive in letting electric conduction pathways reach active mate-
Introduction
rials and resultantly helpful in reducing ohmic polarization,
particularly during fast charging or discharging.10–15 Ionic
conductive binders such as polyethylene oxide (PEO) and poly-
acrylonitrile (PAN) facilitated the access of lithium ions to active
materials to suppress the concentration polarization develop-
ment.16–19 However, an additional capacity can be provided by
using a redox-active binder.20,21
As practical energy storage devices, lithium-ion batteries (LIBs)
are essential for portable electronic devices and electric vehi-
cles.1–3 The reason to conquer markets is their high energy
density at high operating voltages with self-discharge limited.4
The energy density of LIBs has steadily improved since its rst
appearance in the market, reaching four times that of the rst
LIB.5 Despite remarkable improvements, the growing markets
of large-scale energy storage systems as well as electric vehicles
are demanding higher energy densities.6 To overcome the
saturated energy densities of LIBs, even reactions alternative to
the current chemistry of LIBs have been actively tried and
tested.7,8
In this work, we designed and synthesized redox-active
mixed ion/electron conductors (redox-active MIEC) as the
binder for an anode of LIB. All the merits of aforementioned
ionically/electrically conductive and redox-active binders were
integrated into a single polymer. The redox-active MIEC binders
were constructed by attaching ionically conductive branches
(containing a unit of PEO) to an electrically conductive back-
bone (polythiophene) having a redox-active pendant. The
number of ionically conductive branches was controlled to
deconvolute the contribution of the ionically conductive nature
to LIB cell performances from that of the electrically conductive
nature. The redox-active MIEC binder presented in this work
was superior to the practically used PVdF binder in terms of
lithium ion diffusivity and electrical conductivity (1.4ꢀ and
15 000ꢀ, respectively). To demonstrate benets obtained from
the improved properties of the redox-active MIEC binder,
a Li4Ti5O12 spinel (LTO) was chosen as the active anode material
because the redox plateau potential of LTO is around 1.5 V
versus Li/Li+ similar to that of the redox-active MIEC binder. The
LTO cell based on the redox-active MIEC binder was superior to
cells based on PVdF and its ionically nonconductive control,
exhibiting higher capacity (due to its redox activity) and
Keeping the present chemistry of LIBs, we still have chances
to increase the energy density by touching the non-active
components of an active layer laminated on
a current
collector. In the active layer of LIB electrodes, a polymer binder
integrates all components together and to the current collector,
while carbon black provides electric pathways from the current
collector to active material particles. There is every possibility
that the polymer binder obstructs electric conduction and mass
transfer of lithium ions. To eliminate the drawbacks of practical
binders, electrically or ionically conductive polymers have been
employed.9 Electrically conductive polymer binders such as
aSchool of Energy and Chemical Engineering, UNIST, Ulsan 40419, Korea. E-mail:
yang@unist.ac.kr; philiphobi@hotmail.com
bDepartment of Chemistry, Purdue University, West Lafayette, IN 47907, USA
† Electronic supplementary information (ESI) available. See DOI:
10.1039/d0ta08913c
‡ Both authors contributed equally to this work.
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 4751–4757 | 4751

View Article Online
Journal of Materials Chemistry A
Paper
improved rate capability (due to its ionic and electric G1C2, PT*-G2C1, and PT*-G3 had sufficiently high Mn of 70.0,
conductivities).
32.0, 24.0, and 48.0 kDa, respectively, with narrow PDI values of
1.17–2.05. The thermogravimetric analysis (TGA) showed good
ꢁ
thermal stability of all polymers over 400 C (Fig. S2†).
Experimental section
Characterization
4,4-Dioctyl-4H-cyclopenta[2,1-b:3,4-b0]dithiophene (M1). To
a suspension solution of 4H-cyclopenta[2,1-b:3,4-b0]dithiophene
(CPDT) (1 g, 5.6 mmol) in 20 mL of DMSO, 1-bromooctane
(2.38 g, 12.3 mmol) and potassium iodide (15 mg) were added at
room temperature. The reaction mixture was cooled in an ice
bath, and potassium hydroxide (0.95 g, 16.9 mmol, 3.0 equiv.)
was added in one portion. The resulting solution was stirred at
room temperature overnight and then quenched with water and
extracted with diethyl ether (3 ꢀ 100 mL). The combined
organic layer was dried over MgSO4. Aer removing the solvent
using a rotary evaporator, the residue was puried by column
chromatography on silica gel (hexane) to get compound M1
(1.58 g, 70%). 1H NMR (400 MHz, CDCl3 d): 7.14 (d, 2H), 6.93 (d,
2H), 1.81 (t, 4H), 1.35–1.00 (m, 24H), 1.00–0.89 (m, 4H), 0.84 (t,
6H).
4,4-Bis(2-(2-ethoxyethoxy)ethyl)-4H-cyclopenta[2,1-b:3,4-b0]
dithiophene (M2). The procedure is the same as that used for
synthesizing compound M1, with 1-bromo-2-(2-ethoxyethoxy)
ethane (2.42 g, 12.3 mmol), yielding a yellow liquid (1.15 g,
50%). H NMR (400 MHz, CDCl3 d): 7.15 (d, 2H), 6.96 (d, 2H),
3.39–3.31 (m, 8H), 3.31–3.28 (m, 4H), 2.99 (t, 4H), 2.28 (t, 4H),
1.14 (t, 6H).
(4,4-Dioctyl-4H-cyclopenta[1,2-b:5,4-b0]dithiophene-2,6-diyl)
bis(trimethylstannane) (CPDT–C). M1 (1 g, 2.5 mmol) was dis-
solved in anhydrous THF (25 mL) and kept at ꢂ78 ꢁC under
nitrogen ow. Then, n-butyllithium (2.2 mL, 5.5 mmol, 2.5 M in
hexane) was slowly added. Subsequently, aer stirring at ꢂ78 ꢁC
for 1 h, a trimethyltin chloride solution (5.5 mL, 5.5 mmol,
1.0 M in THF) was added. The reaction mixture was then slowly
warmed to room temperature and stirred overnight. The
mixture was quenched by adding water and extracted with
diethyl ether 3 times. The organic layer was dried over anhy-
drous MgSO4 and concentrated under reduced pressure. The
nal product was obtained by column chromatography (AlOx, n-
hexane contained 10% of trimethylamine) to give CPDT–C
1H NMR spectra were recorded using a VNMRS 400 MHz spec-
trophotometer using CDCl3 as the solvent and tetramethylsi-
lane (TMS) as the internal standard. The number-average (Mn)
and weight average (Mw) molecular weights, and polydispersity
index (PDI) of the polymer products were determined by gel
permeation chromatography (GPC) with Waters 150C GPC
using a series of monodisperse polystyrene as standards in
1,2,4-trichlorobenzene (HPLC grade) at 100 ꢁC. Thermogravi-
metric analysis (TGA) was performed using a Simultaneous TGA
instrument (TA Instruments, USA) at
a heating rate of
5
ꢁC minꢂ1. Cells were galvanostatically lithiated and deli-
thiated between 1.5 and 3 V (WonAtech WBCS 3000). Electro-
chemical impedance spectroscopy analysis was performed
using an electrochemical workstation (Biologic, VSP-300) in the
ꢁ
frequency range of 10 mHz to 1 MHz at 25 C.
Electrodes and electrochemical cells
1
A synthesized binder was dissolved in tetrahydrofuran (THF) by
stirring at 60 ꢁC for 1 day to prepare 5 wt% binder solution. An
electrode slurry was prepared by mixing lithium titanate (LTO;
Ishihara, < 1 mm in Fig. S1†), carbon blacks (Timcal, Super P)
and binder in 80 : 10 : 10 proportion (wt%). The slurry was
ꢁ
coated onto a copper foil and dried at 110 C under vacuum.
The slurry-coated foil was punched to a disk of 14 mm diameter.
The loading density of LTO was xed at 3 mA h cmꢂ2. Coin-type
half cells were constructed by sandwiching a polyethylene
separator (Asahi, NH716) between the lithium metal and the
LTO-based electrodes. The electrolyte was 1 M LiPF6 in
a mixture of ethylene carbonate (EC) and dimethyl carbonate
(DMC) 1 : 1 in volume.
Synthesis
All starting materials were purchased from Aldrich and Acros (1.4 g, 77%). 1H NMR (400 MHz, CDCl3, d): 6.94 (m, 2H), 1.78 (m,
and used without further purication. All solvents were of ACS 4H), 1.21 (m, 20H), 0.85 (m, 10H), 0.85 (m, 10H), 0.38 (t, 18H).
grade unless otherwise noted. All the polymer precursors in this
(4,4-Bis(2-(2-ethoxyethoxy)ethyl)-4H-cyclopenta[2,1-b:3,4-b0]
study were prepared analogous to the reported procedures, and dithiophene-2,6-diyl)bis(trimethylstannane) (CPDT–G). The
the synthetic scheme is shown in Scheme S1.†22–27 With procedure is the same as that used for synthesizing compound
different combinations of monomers (CPDT–C, CPDT–G, TPD– CPDT–C, with M2 (1 g, 2.4 mmol), yielding a viscous liquid
1
C, and TPD–G), polymerization was performed via the Stille (0.89 g, 50%). H NMR (400 MHz, CDCl3 d): 6.96 (s, 2H), 3.41–
cross-coupling reaction in the presence of Pd2(dba)3 (2 mol%) 3.28 (m, 12H), 2.99 (t, 4H), 2.28 (t, 4H), 1.14 (t, 6H), 0.38 (t, 18H).
as a catalyst and P(o-tol)3 (8 mol%) as a ligand for 2 days. The
crude polymers were puried via a sequential Soxhlet extraction (TPD–C).
1,3-Dibromo-5-octyl-4H-thieno[3,4-c]pyrrole-4,6(5H)-dione
1,3-Dibromo-4H-thieno[3,4-c]pyrrole-4,6(5H)-dione
with methanol, acetone, n-hexane, and chloroform. The chlo- (TPD) (1 g, 3.2 mmol) and K2CO3 (0.44 g, 6.4 mmol) were dis-
roform fractions of polymers were concentrated and precipi- solved in 50 mL of DMF. Then, 1-bromooctane (0.93 g, 4.8
tated in methanol. Aer that, the precipitated polymers were mmol) was added in one portion to the reaction mixture. The
collected using a membrane lter (pore size, 0.45 mm) and dried reaction was stirred at room temperature overnight, removing
in a high vacuum oven. The average molecular weights (Mn) and the solvent using a rotary evaporator, and the residue was
polydispersity index (PDI) were determined by high- puried by column chromatography on silica gel (hexane : DCM
temperature gel-permeation chromatography (HT-GPC) at 1 : 1) to get compound TPD–C (1.0 g, 74%). 1H NMR (400 MHz,
ꢁ
100 C with 1,2,4-trichlorobenzene as the eluent. PT*-C3, PT*-
4752 | J. Mater. Chem. A, 2021, 9, 4751–4757
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
CDCl3 d): 3.58 (t, J ¼ 7.2 Hz, 2H), 1.69–1.55 (m, 2H), 1.43–1.15
(m, 10H), 0.87 (t, J ¼ 6.7 Hz, 3H).
1,3-Dibromo-5-(2-(2-ethoxyethoxy)ethyl)-4H-thieno[3,4-c]
pyrrole-4,6(5H)-dione (TPD–G). The procedure is the same as
that used for synthesizing compound TPD–C, except for using 1-
bromo-2-(2-ethoxyethoxy)ethane (0.94 g, 4.8 mmol). A white
solid was isolated by column chromatography (0.71 g, 52%). 1H
NMR (400 MHz, CDCl3 d): 3.92 (t, 2H), 3.72 (t, 2H), 3.65–3.61 (m,
2H), 3.57–3.45 (m, 4H), 1.18 (t, 3H).
PT*-C3. A mixture of CPDT–C (145.7 mg, 0.2 mmol) and
TPD–C (84.6 mg, 0.2 mmol) was dissolved in dry toluene (4 mL),
and the solution was degassed for 20 min. Subsequently,
Pd2(dba)3 (3.7 mg, 4 mmol) and P(o-tol)3 (4.9 mg, 8 mmol) were
added, and the mixture was stirred for 2 days at reux
temperature. The resulting crude polymer material was
precipitated in methanol and puried by repetitive Soxhlet
extractions with methanol, acetone, n-hexane and chloroform.
The chloroform fraction was again precipitated in methanol
and ltered, yielding a blue solid (110 mg, 79%). Mn ¼ 70.0 kDa,
Mw ¼ 123.0 kDa, PDI ¼ 1.75.
Fig. 1 PT*-GmCn as a redox-active and mixed ionically/electrically
conductive (MIEC) polymer. CPDT ¼ 4H-cyclopenta[2,1-b:3,4-b0]
dithiophene; TPD ¼ 4H-thieno[3,4-c]pyrrole-4,6(5H)-dione.
(Fig. 1). Detailed synthetic routes of monomers and polymeri-
zation are described in Experimental section and Schemes S1
and S2 in ESI.† The 4H-cyclopenta[2,1-b:3,4-b0]dithiophene
(CPDT)-based monomer unit has a fused-resonance structure
over thiophene–pentadiene–thiophene to allow electrons to be
delocalized. The 4H-thieno[3,4-c]pyrrole-4,6(5H)-dione (TPD)-
based monomer unit is a thiophene having a redox-active suc-
cinimide functionality (cyclic imide). The Pd-catalyzed Stille
polymerization proceeded via a thiophene–thiophene linkage to
have a conjugated structure developed between the units. Both
of the monomers were designed to have one or two linear
strands, R, which were attached to the sp3 carbon of the pen-
tadiene of the CPDT unit and the nitrogen of succinimide of the
TPD unit. Two different functional strands were used, which
had a similar length (or the same number of atoms) but
different atomic identities. An ethylene glycol diethyl ether (G ¼
–CH2CH2OCH2CH2OCH2CH3) strand was used for providing
ionic conductivity, which was a fraction of PEO or polyethylene
glycol (PEG) as a representative lithium ion conductor.28 As an
PT*-G1C2. A mixture of CPDT–C (145.7 mg, 0.2 mmol) and
TPD–G (85.4 mg, 0.2 mmol) was dissolved in dry toluene (4 mL),
and the solution was degassed for 20 min. Subsequently, Pd2(dba)3
(3.7 mg, 4 mmol) and P(o-tol)3 (4.9 mg, 8 mmol) were added, and the
mixture was stirred for 2 days at reux temperature. The resulting
crude polymer material was precipitated in methanol and puried
by repetitive Soxhlet extractions with methanol, acetone, n-hexane
and chloroform. The chloroform fraction was again precipitated in
methanol and ltered, yielding a blue solid (100 mg, 71%). Mn ¼
32.0 kDa, Mw ¼ 66.0 kDa, PDI ¼ 2.05.
PT*-G2C1. A mixture of CPDT–G (147.2 mg, 0.2 mmol) and
TPD–C (84.6 mg, 0.2 mmol) was dissolved in dry toluene (4 mL),
and the solution was degassed for 20 min. Subsequently,
Pd2(dba)3 (3.7 mg, 4 mmol) and P(o-tol)3 (4.9 mg, 8 mmol) were
added and the mixture was stirred for 2 days at reux temper-
ature. The resulting crude polymer material was precipitated in
methanol and puried by repetitive Soxhlet extractions with
methanol, acetone, n-hexane and chloroform. The chloroform
fraction was again precipitated in methanol and ltered,
yielding a blue solid (110 mg, 79%). Mn ¼ 24.0 kDa, Mw ¼ 39.0
kDa, PDI ¼ 1.61.
ionically non-conductive control, an n-octyl group (C ¼ –C8H17
)
strand was used for comparison. The synthesized
polythiophene-backboned binders were designated as PT*-
GnCm, where PT is the polythiophene-type backbone, * is the
redox-active succinimide unit, G is the ethylene glycol diethyl
ether group, C is the octyl group, m is the number of G strands
and n is the number of C strands.
PT*-G3. A mixture of CPDT–G (147.2 mg, 0.2 mmol) and
TPD–G (85.4 mg, 0.2 mmol) was dissolved in dry toluene (4 mL),
and the solution was degassed for 20 min. Subsequently,
Pd2(dba)3 (3.7 mg, 4 mmol) and P(o-tol)3 (4.9 mg, 8 mmol) were
added, and the mixture was stirred for 2 days at reux
temperature. The resulting crude polymer material was
precipitated in methanol and puried by repetitive Soxhlet
extractions with methanol, acetone, n-hexane and chloroform.
The chloroform fraction was again precipitated in methanol
and ltered, yielding a blue solid (100 mg, 71%). Mn ¼ 48.0 kDa,
Mw ¼ 56.0 kDa, PDI ¼ 1.17.
The synthesized polymers showed high electrolyte-philicity,
which is the priority requirement of electrode-constituent
materials.29 A drop of the electrolyte used in cell tests spread
better on the synthesized thiophene-based polymers than on
the practically used PVdF. The contact angles of PT*-G3 and
PT*-C3 (26ꢁ and 38ꢁ, respectively) were much smaller than that
of PVdF (81ꢁ) (Fig. 2a). Both of the synthesized thiophenes were
less swollen by the electrolyte than PVdF (Fig. 2b) even though
they were more electrolyte-philic. The practical binder soaked
the electrolyte as large as 19% to 28% of its mass for 2 h and
20 h respectively, while the all-alkane-tethered thiophene (PT*-
C3) did at only 13% to 19%. The alkyl chain of PT*-C3 is possibly
responsible for the low electrolyte absorption due to its
nonpolar nature against the polarity of electrolyte molecules:
Results and discussion
The electrically conductive backbone of the redox-active MIEC
binder stemmed from two thiophene-based monomer units
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 4751–4757 | 4753

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 2 Electrolyte-philicity. (a) Contact angles on binder films. (b)
Swellability of binder films after 2 h and 20 h wetting. The swellability is
defined as a ratio of absorbed electrolyte mass to polymer mass,
indicating electrolyte mass uptake. (c) Electrical conductivity of active
layers consisting of LTO : carbon black : binder at 8 : 1 : 1 on slide
glasses. (d) Randle–Sevcik plots of cathodic peak current (ipc) versus
Fig. 3 Lithiation and delithiation of LTO. (a) Potential profiles of the
first discharge and charge at 0.1C (0.3 mA cmꢂ2). Voltage (E) was
measured versus Li/Li+. Initial coulombic efficiencies were indicated.
(b) Differential capacity (dQ/dV) profiles along potential, obtained from
(a). (c) Capacity (Q) retention along repeated cycles of lithiation and
delithiation at 1C. (d) Rate dependency of capacity (Q). The same rates
were used for lithiation and delithiation.
root square of scan rate (v1/2). The same active layers used in (c) on
+
a copper foil were used. The values of lithium ion diffusivity (DLi
)
calculated from the slopes were indicated.
e.g., dielectric constant ¼ 2 for octane ꢃ 90 for ethylene
carbonate. The amount of electrolyte uptake of the synthesized
thiophene-backbone polymer increased signicantly by replac-
ing the alkyl branch with the glyme-type branch: from 13% to
19% for 2 h and 17% to 22% for 20 h aer the PEGylation.
The cathodic stability of the PT*-GmCn binders were
conrmed (Fig. S3†). During the 300-time repeated potential
sweeps in the LTO working potential range between 1 V and 3 V
versus Li/Li+, the electrochemical ngerprints of the binders in
the cyclic voltammograms were not changed. Therefore, the
PT*-GmCn binders were possibly used with lithium titanate
(LTO) having its faradaic plateau around 1.5 V versus Li/Li+.
However, they were considered to be inappropriate as the
cathode binder. The PT*-G3, as a representative of the PT*-
GmCn, was oxidized around 3.8 V versus Li/Li+ (Fig. S4†).
The electrical conductivity of the electrode active layer
abruptly increased from the order of 10ꢂ2 to 102 in S cmꢂ1 by
replacing PVdF with PT*-GmCn (Fig. 2c): 0.013 S cmꢂ1 for PVdF,
210 S cmꢂ1 for PT*-C3 and 182 S cmꢂ1 for PT*-G3. Samples were
prepared by coating a mixture of the active material, carbon
black and binder (80 : 10 : 10 by weight) on slide glasses. LTO
was used as the active material. As expected, electrical
conductivity was determined by the polythiophene-type (PT)
backbone, independent of branches. All PT*-GmCn showed
a similar value of electrical conductivity. The adhesion of the
PT*-GmCn binders to copper current collectors was stronger
than that of the popularly used PVdF binder (Fig. S5†).
+
better than PVdF in terms of lithium diffusion (slope ꢄ DLi ).
+
More interestingly, the value of DLi increased with the amount
of G, conrming that the G branch plays the main role of
lithium ion conduction. The G-dependency of DLi was expected
when we designed the binders because ethylene oxide units
have been well known to be responsible for lithium ion
conduction.
+
Lithiation and delithiation of LTO were compared between
binders to investigate the effects of electric and ionic conduc-
tion and redox activity of PT*-GmCn on the faradaic process. LTO
was chosen as the active material because it was redox-active at
the same potential range, where the succinimide moiety of PT*
was redox-active. Reversible lithiation of LTO was conrmed at
0.1C for all binders (Fig. 3a). The plateau potentials of lithiation
and delithiation did not show serious differences around 1.55 V,
which is the ideal reduction potential of the electrochemistry in
galvanostatic potential proles.29 At a closer look on the
differential capacity proles along potential (Fig. 3b), however,
the potential gap between lithiation and delithiation was
recognized. The benets obtained from the highly conductive
nature of PT-GmCn were conrmed by the activation polariza-
tion gap (DEdLi–Li). The MIEC binder (PT*-G3) was superior to
the electrically conductive but ionically non-conductive binder
(PT*-C3), which was much superior to the non-conductive
binder (PVdF): DEdLi–Li ¼ 8.84 mV for PT*-G3 < 10.8 mV for
PT*-C3 ꢃ 19.46 mV for PVdF. Moreover, the initial coulombic
efficiency increased from 88% for PVdF to 90% for PT*-GmCn
due to more efficient utilization of LTO in the conduction-
improved environments (Fig. 3a).
Cyclic voltammograms (CVs) of LTO were obtained to
calculate the lithium ion diffusivity (DLi ).29 Cathodic and anodic
+
peaks responsible for lithiation and delithiation of LTO at 1.4 V
and 1.7 V versus Li/Li+ respectively were clearly observed in CVs
(Fig. S6†). The linearity of Randles–Sevcik plots of peak current
(ip) versus root square of scan rate (v1/2) conrmed that
lithiation/delithiation of LTO was diffusion-limited (Fig. 2d).
The comparison of the slopes of the Randles–Sevcik plots
between different binders indicated that all PT polymers were
When PVdF was replaced with PT*-GmCn, the delithiation
capacity of the LTO electrode increased from 157 mA h gꢂ1 to
>170 mA h gꢂ1 due to the redox electrochemistry of PT* between
]O and –OLi (Fig. 3a). The redox ngerprint was found in dQ/
dV curves (Fig. 3b). The cathodic peaks at 1.6 V and 1.35 V
4754 | J. Mater. Chem. A, 2021, 9, 4751–4757
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
conrmed the redox of the succinimide moiety of PT*, which attached to the PT* backbone (Fig. 3d; Potential proles at 6C in
was not found in the PVdF-based LTO electrode. The lithiation Fig. 4a). The MIEC PT* with three glyme branches (PT*-G3) was
capacity originating from the reduction of succinimide of PT*- the best in terms of the rate capability due to its highest lithium
G3 was estimated to be ꢄ20 mA h gꢂ1 from the dQ/dV curves, ion diffusivity, followed by G2C1 and G1C2. PT*-C3 was the
which corresponded to the capacity difference between PT*-G3 poorest because all branches were occupied by ionically non-
and PVdF (19 mA h gꢂ1 ¼ 176 mA h gꢂ1 ꢂ 157 mA h gꢂ1). The conductive octane without ether. Therefore, the mass transfer
higher capacities of PT* binders were stably achieved during the of lithium ion should be considered more seriously especially at
repeated cycles of lithiation followed by delithiation both at 1C high C-rates. The PT*-C3 cell did not work at 10C, its capacity
(Fig. 3c). More than 98% of the initial capacities were obtained falling to less than 15% of the 1C capacity at the rate.
for PT*-G3 and PT*-C3 aer the 300 cycles. However, the PVdF-
To investigate the effects of the G and C branches of PT*-
based cell showed a relatively poor cyclability of 95% capacity GmCn binders on lithiation/delithiation of LTO, impedances
retention aer the 300 cycles. In terms of the coulombic effi- were measured in situ during the galvanostatic lithiation at 6C
ciencies, the PT*-based binders were superior to the PVdF (Fig. 4b and c).30 Sinusoidal current signals depending on
control (Fig. S7†).
frequencies were overlapped on a constant current at 6C.
The capacity difference between PT* and PVdF and between Qualitatively, the PT*-based LTO electrodes showed two charge
PT*'s became more serious as the current to lithiate and transfer processes indicated by two overlapped semi-circles in
delithiate LTO increased from 1C to 10C (Fig. 3d; Potential Nyquist plots.31 Considering that the PVdF-based LTO electrode
proles at 6C in Fig. 4a). A series of the PT* binders were showed a single charge transfer (indicated by a single semi-
kinetically superior to the PVdF counterpart especially due to circle) relevant to lithiation of LTO, the redox electrochemistry
their mixed-conductive nature. The PVdF cell did not work at C of succinimide moiety of PT* backbone is considered as the
rates higher than 6C, but PT*-G3 exhibited 54% of the 1C additional charge transfer process. The size of the overlapped
capacity even at 10C (Fig. 3d). The ohmic drop of the PT*-based semi-circles on abscissa was used as a measure of the overall
electrode at the initial period of lithiation was signicantly less charge transfer resistance (Rct). All binders showed a similar
than those of the PVdF-based electrodes due to their electrically trend of Rct change along lithiation even if their capacity spans
conductive nature (Fig. 4a). The plateau potential of the PT* were different: Rct decreased initially, reached a plateau at
electrodes in potential proles at 6C was found to be 1.55 V a minimum value and then increased at the end of lithiation.
corresponding to the thermodynamic potential of lithiation/ There was an insignicant difference in Rct between PT*
delithiation of LTO. However, 150 mV overpotential was binders up to 40 mA h gꢂ1. Before the critical capacity point
required for the PVdF electrode, showing the plateau potential (40 mA h gꢂ1), a serious increase in Rct of PVdF was observed,
at 1.4 V.
In comparison between PT* binders, as expected, the rate
while PT* still showed a minimized plateau.
Electrically conductive nature of binders was considered to
capability was strongly dependent on the number of G branches be responsible for the difference between PVdF and PT*‘s in the
initial lithiation at the high rate (6C) before 40 mA h gꢂ1 (Fig. 4).
The highly electrically conductive PT* environment
surrounding the active material allowed more effective utiliza-
tion of the redox-active mass and/or discouraged ohmic voltage
drop to keep the potential prole far from a cut-off voltage
(Fig. 4a). Aer the critical capacity point (40 mA h gꢂ1), the onset
points of the Rct increase were met in the increasing order of the
number of G branches from C3, G1C2, G2C1 to G3. The G branch
responsible for the ionic conduction (high lithium diffusivity
for PT*-G3 in Fig. 2d) played a major role to decrease Rct in the
latter part of lithiation (Fig. 4c). The concentration polarization
at the end of lithiation of LTO as well as the activation polari-
zation at the middle of lithiation is possibly improved by
adopting the ionically conductive binders.
To clarify the technological advancement of the MIEC PT*-G3
binder particularly at high rates, its capacity at 10C was
compared with the capacities at high currents presented by
previously reported binder works (Fig. 5 and Table S1†).31–40
A
limited number of publications were found for binders for LTO.
Four different ionically conductive binders used for LTO pre-
sented capacities at limited currents. An ionically conductive
beta-phase PVdF binder showed the best value, 0.29 mA h cmꢂ2
at 10 mA cmꢂ2, allowing only 2 min operation.31 Therefore, the
rate capability of 2.1 and 1.7 mA h cmꢂ2 at 15 mA cmꢂ2 and 30
mA cmꢂ2 respectively presented by the MIEC PT*-G3 in this
Fig. 4 Electrochemical impedance spectra in situ measured during
lithiation of LTO at 6C. (a) Potential profiles. (b) Impedance spectra at
0, 25, and 80 mA h gꢂ1 and fully lithiated state. (c) Charge transfer
resistance (Rct).
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 4751–4757 | 4755

View Article Online
Journal of Materials Chemistry A
Paper
due to its MIEC nature of PT*-GmCn, while additional capacity
was achieved from its redox activity. We believe that a rational
design of polymer binders possibly plays a crucial role to
improve the performances of LIBs without changing or
touching active materials.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
This work was supported by NRF (2018R1D1A1B07042773,
2017M1A2A2087813,
2018R1A2A1A05077194,
(P0002068), UNIST (1.200092.01) and Ulsan City (1.200042),
Korea.
2019M1A2A2065614,
2016R1A5A1009926), MOTIE
Fig. 5 Comparison of rate performances of LTO and silicon cells
having lithium metal as a counter electrode. This work was compared
with other works previously reported. Each point indicates the capacity
at the maximum current used in the corresponding publication. The
data were taken from binder-related works listed in Table S1.†
Capacities and currents were normalized by a geometric electrode
area since the area-normalized quantities were practically more
meaningful rather than the mass-normalized ones.
Notes and references
1 J. M. Tarascon and M. Armand, Nature, 2001, 414, 359–367.
2 M. Armand and J. M. Tarascon, Nature, 2008, 451, 652–657.
3 V. Etacheri, R. Marom, R. Elazari, G. Salitra and D. Aurbach,
Energy Environ. Sci., 2011, 4, 3243–3262.
4 B. Dunn, H. Kamath and J. M. Tarascon, Science, 2011, 334,
928–935.
work is recognized to be extraordinarily superior over the up-to-
date binders. The capacity values at the high currents equiva-
lent to 5C and 10C are considered to be high, assuming that
commercialized LIB cells have loadings around 4 mA h cmꢂ2
.
These high capacities were obtained even at such high rates
since the high electric and ionic conductivities of the MIEC PT*-
G3 allowed high loading of LTO at 17 mg cmꢂ2 responsible for
3 mA h cmꢂ2 (c.f., 10 mg LTO cmꢂ2 for the beta-PVdF work31).
From the viewpoint of the loading mass, moreover, the high rate
superiority of MIEC PT*-G3 with LTO was more emphasized
when it was compared with novel binders for high-capacity
silicon anodes. A more mass of LTO is required for achieving
the same capacity of silicon: 17 mg LTO cmꢂ2 versus 1.0 mg Si
cmꢂ2 required for 3 mA h cmꢂ2. In spite of the loading mass
burden of LTO, it was easily seen in the same comparison
charge (Fig. 5 and Table S1†) that MIEC PT*-G3 with LTO was
much superior to the best binder for silicon (a crosslinked
composite binder of PEO and PEI as an ionic conductor and
PEDOT:PSS as an electric conductor39): 1.7 mA h cmꢂ2 at 30 mA
cmꢂ2 for MIEC PT*-G3 with LTO versus 2.0 mA h cmꢂ2 at 11 mA
cmꢂ2 for MIEC c-PEO–PEDOT:PSS/PEI with Si. When consid-
ering that the loading density of silicon in the composite binder
work was just 1.3 mg cmꢂ2, it is commendable that MIEC PT*-
G3 supported the excellent high rate performances of the LTO
5 J. B. Goodenough and K. S. Park, J. Am. Chem. Soc., 2013, 135,
1167–1176.
6 C. Y. Wang, T. Xu, S. H. Ge, G. S. Zhang, X. G. Yang and Y. Ji,
J. Electrochem. Soc., 2016, 163, A1944–A1950.
7 J. Janek and W. G. Zeier, Nat. Energy, 2016, 1.
8 X. P. Gao and H. X. Yang, Energy Environ. Sci., 2010, 3, 174–
189.
9 M. Y. Wu, X. C. Xiao, N. Vukmirovic, S. D. Xun, P. K. Das,
X. Y. Song, P. Olalde-Velasco, D. D. Wang, A. Z. Weber,
L. W. Wang, V. S. Battaglia, W. L. Yang and G. Liu, J. Am.
Chem. Soc., 2013, 135, 12048–12056.
10 D. Shao, H. X. Zhong and L. Z. Zhang, ChemElectroChem,
2014, 1, 1679–1687.
11 H. Wu, G. H. Yu, L. J. Pan, N. A. Liu, M. T. McDowell,
Z. A. Bao and Y. Cui, Nat. Commun., 2013, 4, 1943.
12 H. Zhao, Z. H. Wang, P. Lu, M. Jiang, F. F. Shi, X. Y. Song,
Z. Y. Zheng, X. Zhou, Y. B. Fu, G. Abdelbast, X. C. Xiao,
Z. Liu, V. S. Battaglia, K. Zaghib and G. Liu, Nano Lett.,
2014, 14, 6704–6710.
13 G. Liu, S. D. Xun, N. Vukmirovic, X. Y. Song, P. Olalde-
Velasco, H. H. Zheng, V. S. Battaglia, L. W. Wang and
W. L. Yang, Adv. Mater., 2011, 23, 4679.
cell embracing such a heavy loading burden of 17 mg cmꢂ2
.
14 M. Ling, J. X. Qiu, S. Li, C. Yan, M. J. Kiefel, G. Liu and
S. Q. Zhang, Nano Lett., 2015, 15, 4440–4447.
Conclusion
In this work, we presented a redox-active MIEC as the binder for 15 H. Zhao, Y. Wei, R. M. Qiao, C. H. Zhu, Z. Y. Zheng, M. Ling,
LIBs. PT*-GmCn was designed to be (1) electrically conductive
along its conjugated thiophene backbone (PT ¼ polythiophene),
(2) redox-active from its succinimide moiety (* ¼ redox-active)
Z. Jia, Y. Bai, Y. B. Fu, J. L. Lei, X. Y. Song, V. S. Battaglia,
W. L. Yang, P. B. Messersmith and G. Liu, Nano Lett., 2015,
15, 7927–7932.
and (3) ionically conductive by adopting glyme (G) branches. 16 Z. G. Xue, D. He and X. L. Xie, J. Mater. Chem. A, 2015, 3,
High capacity of LTO was guaranteed especially at high rates
19218–19253.
4756 | J. Mater. Chem. A, 2021, 9, 4751–4757
This journal is © The Royal Society of Chemistry 2021

View Article Online
Paper
Journal of Materials Chemistry A
17 J. Liu, Q. Zhang, T. Zhang, J. T. Li, L. Huang and S. G. Sun, 30 Y. Ko, C. Hwang and H. K. Song, J. Power Sources, 2016, 315,
Adv. Funct. Mater., 2015, 25, 3599–3605. 145–151.
18 A. Nishimoto, M. Watanabe, Y. Ikeda and S. Kohjiya, 31 W. J. Song, S. H. Joo, D. H. Kim, C. Hwang, G. Y. Jung, S. Bae,
Electrochim. Acta, 1998, 43, 1177–1184.
Y. Son, J. Cho, H. K. Song, S. K. Kwak, S. Park and S. J. Kang,
Nano Energy, 2017, 32, 255–262.
32 S. Karuppiah, S. Franger and K. Nallathamby,
ChemElectroChem, 2018, 5, 343–349.
33 F. De Giorgio, A. La Monaca, A. Dinter, M. Frankenberger,
K. H. Pettinger and C. Arbizzani, Electrochim. Acta, 2018,
289, 112–119.
19 D. Y. Oh, Y. J. Nam, K. H. Park, S. H. Jung, K. T. Kim, A. R. Ha
and Y. S. Jung, Adv. Energy Mater., 2019, 9, 1802927.
20 J. Kim, H. S. Park, T. H. Kim, S. Y. Kim and H. K. Song, Phys.
Chem. Chem. Phys., 2014, 16, 5295–5300.
21 C. Benoit, D. Demeter, D. Belanger and C. Cougnon, Angew.
Chem., Int. Ed., 2016, 55, 5318–5321.
22 Z. B. Henson, P. Zalar, X. F. Chen, G. C. Welch, T. Q. Nguyen 34 B.-R. Lee and E.-S. Oh, J. Phys. Chem. C, 2013, 117, 4404–
and G. C. Bazan, J. Mater. Chem. A, 2013, 1, 11117–11120. 4409.
23 J. Brebels, E. Douvogianni, D. Devisscher, R. T. Eachambadi, 35 D. Liu, Y. Zhao, R. Tan, L.-L. Tian, Y. Liu, H. Chen and F. Pan,
J. Manca, L. Lutsen, D. Vanderzande, J. C. Hummelen and
W. Maes, J. Mater. Chem. C, 2018, 6, 500–511.
24 I. H. Jung, J. Yu, E. Jeong, J. Kim, S. Kwon, H. Kong, K. Lee,
H. Y. Woo and H. K. Shim, Chem.–Eur. J., 2010, 16, 3743–
3752.
Nano Energy, 2017, 36, 206–212.
36 T. M. Higgins, S.-H. Park, P. J. King, C. Zhang, N. McEvoy,
N. C. Berner, D. Daly, A. Shmeliov, U. Khan, G. Duesberg,
V. Nicolosi and J. N. Coleman, ACS Nano, 2016, 10, 3702–
3713.
25 C. C. Ho, S. Y. Chang, T. C. Huang, C. A. Chen, H. C. Liao, 37 J. Ryu, S. Kim, J. Kim, S. Park, S. Lee, S. Yoo, J. Kim,
Y. F. Chen and W. F. Su, Polym. Chem., 2013, 4, 5351–5360.
26 R. M. W. Wolfe and J. R. Reynolds, Org. Lett., 2017, 19, 996–
999.
27 B. Xu, X. P. Yi, T. Y. Huang, Z. L. Zheng, J. X. Zhang, A. Salehi,
V. Coropceanu, C. H. Y. Ho, S. R. Marder, M. F. Toney,
N.-S. Choi, J.-H. Ryu and S. Park, Adv. Funct. Mater., 2020,
30, 1908433.
38 C. Zhang, Q. Chen, X. Ai, X. Li, Q. Xie, Y. Cheng, H. Kong,
W. Xu, L. Wang, M. S. Wang, H. Yang and D. L. Peng, J.
Mater. Chem. A, 2020, 8, 16323–16331.
J. L. Bredas, F. So and J. R. Reynolds, Adv. Funct. Mater., 39 W. Zeng, L. Wang, X. Peng, T. Liu, Y. Jiang, F. Qin, L. Hu,
2018, 28, 1803418.
P. K. Chu, K. Huo and Y. Zou, Adv. Energy Mater., 2018, 8,
1702314.
40 C. Hwang, Y.-G. Cho, N.-R. Kang, Y. Ko, U. Lee, D. Ahn,
J.-Y. Kim, Y.-J. Kim and H.-K. Song, J. Power Sources, 2015,
298, 8–13.
28 D. Zhou, D. Shanmukaraj, A. Tkacheva, M. Armand and
G. X. Wang, Chem, 2019, 5, 2326–2352.
29 D. Kim, C. Hwang, J. Jeong, W. J. Song, S. Park and
H. K. Song, ACS Appl. Mater. Interfaces, 2019, 11, 43039–
43045.
This journal is © The Royal Society of Chemistry 2021
J. Mater. Chem. A, 2021, 9, 4751–4757 | 4757