A symmetric aqueous redox flow battery based on viologen derivative

Article abstract of DOI:10.1016/j.cclet.2019.11.033

Due to the diversity and feasibility of structural modification for organic molecules, organic-based redox flow batteries (ORFBs) have been widely investigated, especially in aqueous solution under neutral circumstance. In this work, a symmetric aqueous redox flow battery (SARFB) was rationally designed by employing a bipolar redox active molecule (N,N'-dimethyl-4,4-bipyridinium diiodide, MVI₂) as both cathode and anode materials and combining with an anion exchange membrane. For one MVI₂ flow battery, MV²⁺/ MV^·+ and I^?/I₃^? serve as the redox couples of anode and cathode, respectively. The MVI₂ battery with a working voltage of 1.02 V exhibited a high voltage efficiency of 90.30percent and energy efficiency of 89.44percent after 450 cycles, and crossover problem was prohibited. The comparable conductivity of MVI₂ water solution enabled to construct a battery even without using supporting electrolyte. Besides, the bipolar character of MVI₂ battery with/without supporting electrolyte was investigated in the voltage range between ?1.2 V and 1.2 V, showing excellent stable cycling stability during the polarity-reversal test.

Full text of DOI:10.1016/j.cclet.2019.11.033

Journal Pre-proof
A symmetric aqueous redox flow battery based on viologen derivative
Shuang Liu, Meng Zhou, Ting Ma, Jian Liu, Qiu Zhang, Zhanliang
Tao, Jing Liang
PII:
S1001-8417(19)30680-1
https://doi.org/10.1016/j.cclet.2019.11.033
CCLET 5340
DOI:
Reference:
To appear in:
Chinese Chemical Letters
Received Date:
Revised Date:
Accepted Date:
24 September 2019
13 November 2019
19 November 2019
Please cite this article as: Liu S, Zhou M, Ma T, Liu J, Zhang Q, Tao Z, Liang J, A symmetric
aqueous redox flow battery based on viologen derivative, Chinese Chemical Letters (2019),
doi: https://doi.org/10.1016/j.cclet.2019.11.033
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
©
2019 Published by Elsevier.

Please donot adjust the margins
Communication
A symmetric aqueous redox flow battery based on viologen derivative
+
+

Shuang Liu , Meng Zhou , Ting Ma, Jian Liu, Qiu Zhang, Zhanliang Tao, Jing Liang
Key Laboratory of Advanced Energy Materials Chemistry (Ministry of Education), College of Chemistry, Nankai University, Tianjin 300071, China

Corresponding author.
E-mail address: liangjing@nankai.edu.cn
These authors contributed equally to this work.

Graphical abstract
2
A symmetric aqueous redox flow battery (SARFB) was assembled by using MVI as a bipolar molecule and AHA as an anion exchange
membrane. The SARFB exhibited well polar-reversal performances.
Article history:Received 24 September 2019 Received in revised form 13 November 2019 Accepted 14 November 2019
Available online
ABSTRACT
Due to the diversity and feasibility of structural modification for organic molecules, organic-based redox flow batteries (ORFBs) have been widely
investigated, especially in aqueous solution under neutral circumstance. In this work, a symmetric aqueous redox flow battery (SARFB) was rationally
designed by employing a bipolar redox active molecule (N,N'-dimethyl-4,4-bipyridinium diiodide, MVI
2
) as both cathode and anode materials and
serve as the redox couples of anode and cathode,
battery with a working voltage of 1.02 V exhibited a high voltage efficiency of 90.30% and energy efficiency of 89.44% after
50 cycles, and crossover problem was prohibited. The comparable conductivity of MVI water solution enabled to construct a battery even without
using supporting electrolyte. Besides, the bipolar character of MVI battery with/without supporting electrolyte was investigated in the voltage range
between −1.2 V and 1.2 V, showing excellent stable cycling stability during the polarity-reversal test.
2
+
·＋
−
−
combining with an anion exchange membrane. For one MVI
respectively. The MVI
2 3
flow battery, MV / MV and I /I
2
4
2
2
2
Keywords: Aqueous redox flow battery Bipolar redox active molecule Symmetric redox flow battery MVI Prohibited crossover problem
High coulombic efficiency
Redox flow batteries (RFBs) are an attractive electrochemical energy storage technology for large-scale energy storage of renewable
sources including wind and solar power generation [1-3]. All-vanadium redox flow battery (VRFB) has already been commercialized
and thoroughly investigated since 1980s [4-7]. However, the disadvantages including high cost of vanadium-based active materials, low
energy density and acidic operating condition limit the application of VRFBs [8, 9]. However, an obvious advantage of VRFB over other
—
——

Corresponding author.
E-mail address: liangjing@nankai.edu.cn
These authors contributed equally to this work.


Please donot adjust the margins
traditional RFBs system is that the crossover problem between the anolyte and the catholyte could be prevented effectively by employing
the same element of multi-valent vanadium couples (V(II)/V(III)) and (V(IV)/V(V)) in both half-cells.
Inspired by the bipolar feature of VRFBs, symmetric redox flow batteries (SRFBs) emerged [10]. By using only one material
possessing bipolar characters, battery system can be simplified drastically, the effort of searching two different materials for both anode
side and cathode side can be saved, and the crossover problem existed in non-SRFBs can be effectively prevented [11]. For example, the
SRFB with the application of commercially available 2-phenyl-4,4,5,5-tetramethylimidazoleline-1-oxyl oxide (PTIO) can operate at a
high cell potential of 2.5 V [12]. By using 5,10,15,20-tetraphenylporphyrin (H
V and an excellent cycling performance can be achieved by this SRFBs [13]. Although crossover is prohibited and high cell voltage (>
V) is achieved in these works, the inflammable and volatile character of organic solvent raises safety issues. In addition, the insufficient
2
TPP) as both anolyte and catholyte, a high voltage of 2.83
2
solubility of organic molecules, the high-cost and the side reaction restrict the further improvement of non-aqueous RFBs in
electrochemical performance. Thus, the symmetric aqueous redox flow batteries (SARFBs) would be one of the most promising
directions for future RFBs development [14].
2
Scheme 1. Working sketch of the aqueous MVI redox flow battery and the theoretic redox mechanism for both anode and cathode.
A few SARFBs have been reported so far. Winsberg et al. developed a SARFB by use of 2,2,6,6-tetramethylpiperidine-1-oxyl
(TEMPO)/Phenazine combi-molecule and achieved a cell voltage of 1.2 V. [14] However, the high-cost of TEMPO limits its large-scale
application. Zhu et al. reported a SARFB based on 1-(4-ferrocenyl-n-butyl)-1'-[3−(trimethylammonio)propyl]-4,4'-bipyridinium cation
3
+
(
Fc-bipy ), reaching a stable cycle performance of 4000 cycles. Nevertheless, due to the low cell voltage, the theoretical capacity was
restricted to 6.7 Ah/L [15]. Viologen compounds are a class of redox active molecules with two redox potentials, which have ever been
investigated in some RFBs [16-20]. For instance, Liu et al. reported a methyl viologen (MV)/4-hydroxy-2,2,6,6-tetramethylpiperidin-1-
oxyl (4-HO-TEMPO) ARFB with a theoretical energy density of 8.4 Wh/L [21]. However, the side reaction of the 4−HO−TEMPO
affects the long term stability. DeBruler et al. developed 1,1'-bis(3−sulfonatopropyl)-4,4'-bipyridinium ((SPr) V)/KI ARFBs [22].
2
Nevertheless, the oxidation potential and solubility of catholyte restrict the cell voltage to 1.0 V and the energy density to 6.7 Wh/L.

Recently, Luo et al. constructed a (SPr)
2
V/Br ARFB which possesses a high voltage of 1.51 V. [23] Nevertheless, the crossover issue
cannot be prevented. To the best of our knowledge, the application of N,N'-dimethyl-4,4-bipyridinium diiodide (MVI ) in a SARFB
2
system has never been reported before.
Here, MVI was synthesized through a one-step methylation reaction, and was introduced to the SARFBs, in which the bipyridinium
2
group acts as the anode redox active material while the iodide ion performs as cathode redox active material (Scheme 1). Cyclic
voltammetry (CV) revealed the bipolar character of the obtained organic molecule, with two reversible reactions at potentials of −0.66
V and 0.36 V versus Ag/AgCl, respectively. The non-pumped battery exhibited excellent stability and an average coulombic efficiency
of 99.30% for 450 cycles. When tested in the voltage range between −1.2 V and 1.2 V, the battery showed a stable polar-reversal
performance. MVI
2
demonstrated its capability as a bipolar redox-active material in subsequent multiple polarity reversals.
2
+
Earlier research showed that the divalent viologen cation (MV ) can undergo a two-step reduction to the neutral viologen (MV0).
·
＋
[
24] Generally, the first reduction to the monovalent radical cation (MV ) is fully reversible, while the second step may lead to the side
·＋
2+
reactions or precipitation due to the decreased solubility of the reduction product. The MV formed when the MV was reduced during
the charge process, and the reaction went inversely upon the discharge process. The redox reaction happened on the anode side exhibited

Please donot adjust the margins
·
＋
satisfied reversibility. Considering that MV is sensitive to oxygen, the oxygen sensitive character can be used to prove the formation
·
＋
of the charged product on the anode side. UV-vis spectroscopy was applied to detect the characteristic peak of the MV (Fig. S1 in
Supporting information). The measurement was carried out and the testing result was collected every two minutes to get series of
consequences. With time passing by, the strength of explored peak at both 395 nm and 605 nm declines gradually, while the peak appears
at 257 nm increases accordingly. At the very end, the former two peaks vanished and the later peak strengthened to a maximum value,
and the spectral peak return to the same with its original state. The visualized change of solution color from violet to transparent yellow
·
＋
2
+
also evidenced the oxidation process of the MV into the MV (insets of Fig. S1).
1
1
0
0
.5
.0
.5
.0
MV2+/MV·+
I-/I -
3
-0.66 V
0.36 V
1
.02 V
-
0.5
1.0
-
-1.0
-0.5
0.0
0.5
Potential (V vs. Ag/AgCl)
2
Fig. 1. Cyclic voltammetry of N,N'-dimethyl-4,4-bipyridinium diiodide at a scan rate of 100 mV/s. Condition: 2 mmol/L MVI and 1 mol/L NaCl in 20 mL
deionized water, glassy carbon working electrode, platinum plate counter electrode, Ag/AgCl reference electrode.
To avoid occurrence of the irreversible second-step reduction reaction, CV was carried out in the potential range from −0.85 V to 0.60
V. Note that the reduction of divalent viologen cation into monovalent radical cation takes place at −0.70 V and the oxidation occurs at
−
−
−
−
−
0.63 V (Fig. 1). For the cathode side, oxidation of I into I
total gives the symmetric system a theoretical cell voltage of 1.02 V, that is much higher than that of Fc-bipy -based SARFBs (0.7 V)
15] and comparable with that of the MV/ferrocene ARFB (1.05 V) [25] and (SPr) V/KI ARFB (1.0 V) [22]. With this voltage, a
theoretical energy density of 8.0 Wh/L can be achieved, which is comparable with that of the MV/4−HO−TEMPO (8.4 Wh/L) [21] and
SPr) V/KI ARFBs (6.7 Wh/L)[22]. CV at varied scan rates from 10 mV/s to 100 mV/s showed that the peak potentials of both oxidation
and reduction of anode material remain almost unchanged (Fig. S2 in Supporting information), indicating an outstanding reversibility of
MVI as anolyte. For the cathode side, only negligible peak shift can be observed with the increase of the scan rate. Furthermore, for
3
takes place at 0.48 V and reduction of I
3
into I occurs at 0.24 V. This in
3
+
[
2
(
2
2
both the anode and the cathode, the peak current follows a strict linear trend with the rise of square root of the scan rate for both the
oxidation and reduction reaction (Fig. S3 in Supporting information). These results imply an excellent electrochemically reversible redox-
reaction and a diffusion-controlled process [26].
Further kinetic investigation on anode side were performed via RDE measurements with rotation rates in the range of 0 rpm to 2500
rpm (Fig. 2a). Limiting current gradually increased with the increase of square root of rotation speed. Levich analysis showed good linear
relation between limiting current under different rotation rates and the boost of square root of rotation speed at −0.85 V versus Ag/AgCl
−
6
2
(
Fig. 2b), indicating a diffusion-controlled behavior of viologen cation. The diffusion coefficient was calculated to be 3.10×10 cm /s,
which is at typical magnitude for organic compounds reported in most of the previous research. [21, 27, 28] Further analysis with the
application of Koutecký-Levich equation [15], provided the mass-transfer independent current i (Fig. 2c). The linear fit of log|i | versus
over-potential yielded the exchange current i (Fig. 2d). According to Nicholson’s method [29], the calculated electron transfer rate
constant (k ) is calculated to be 0.69 cm/s, which is higher than some other organic active materials in ARFBs [14, 15].
k
k
0
0
a) 0
.0
b)
-4
-4
-4
-1.0x10
-2.0x10
-3.0x10
-4.0x10
-0.1
-0.2
-0.3
0 rpm
900 rpm
1225 rpm
1600 rpm
2500 rpm
1
2
4
6
00 rpm
25 rpm
00 rpm
25 rpm
y = -2.06×10-5x-7.23×10-6
-4
-
0.8 -0.6 -0.4 -0.2 0.0
3
6
9 .512 15 18
0 0.5 -0.5
Potential (V vs. Ag/AgCl)
(Rotation rate) (rad
s
)

Please donot adjust the margins
0
d) -2.6
c)
-
2.8
4
-
-
-
1x10
-3.0
4
2x10
3x10
-3.2
-
-
0.615 V
0.625 V
-0.655 V
-0.665 V
-0.675 V
-3.4
4
-0.635 V
0.645 V
y= -13.29x-3.66
-
-3.6
0
.1
0.2
0.3
-0.5
-0.08 -0.06 -0.04 -0.02 0.00
Over potential (V)
Rotation rate) (s0.5 rad
-
0.5
)
(
Fig. 2. Results of rotating disk electrode experiments of MVI
Ag/AgCl) versus the square root of the rotation velocity (Levich plot), (c) Koutecký-Levich plot and (d) Tafel plot with varied overpotential. Measured at a
concentration of c = 2 mmol/L MVI , in 0.1 mol/L NaCl(aq).
2
. (a) RDE voltammograms under different rotating speed. (b) The limiting current (at −0.85 V versus
0
2
Ion conductivity test showed that the water solution of MV-based organic salt exhibits excellent conductivity (0.5 mol/L MVI
2
, 51.5
mS/cm, 25 °C), which is acceptable compared with that of sodium chloride solution (1 mol/L NaCl, 87.5 mS/cm, 25 °C), implying a
2
+
high mobility of anion counter ions of MV . In this way, it is possible to construct a battery system without the supporting electrolyte
NaCl. To confirm the feasibility of NaCl-free SARFB system, non-pumped cell was made using 0.1 mol/L MVI (aq) without introduction
2
of NaCl. Electrochemical tests firstly evidenced that the battery worked properly for 200 consecutive cycles with average coulombic
efficiency of 98.46% (Figs. S4a and b in Supporting information). The decay of the capacity after first cycle (Fig. S4b) is due to the
unbalanced state of solution, which resulting in the second step reaction of viologen. With the charge and discharge going on, the solution
tends to balance, and the second step reaction of viologen vanished, resulting in the capacity decay (Fig. S4c in Supporting information).
The increased polarization of the charge/discharge curves caused by the void of supporting electrolyte decreases the voltage efficiency
of the battery, which in return reduces the energy efficiency. In this sense, supporting electrolyte needs to be introduced for further
investigation.
To optimize the best supporting electrolyte, charge/discharge curves under different sodium salt conditions are performed (Fig. S5
and Table S1 in Supporting information). It can be concluded that 1 mol/L NaCl can provide better electrochemical performance, other
electrolytes preferred in published research exhibited side reactions and/or increased polarization of the charge/discharge behaviour.
Therefore, 1 mol/L NaCl was filtered as the supporting electrolyte. The solubility of MVI
checked by UV-vis absorbance spectra (Fig. S6 in Supporting information). It can be calculated that the saturation concentration of MVI
in 1 mol/L NaCl aqueous solution is 0.40 mol/L and lower than that in pure water (0.58 mol/L, Fig. S7 in Supporting information) due
to salting out. Besides, to determine the suitable concentration of MVI in cells, non-pumped cells were constructed (Fig. S8 in Supporting
information) using MVI with different concentrations in 1 mol/L NaCl aqueous solution as electrolyte, respectively. After tested at a
constant current density of 2 mA/cm and a voltage range from 0.2 V to 1.2 V, cells with 0.05 mol/L MVI
performance (Fig. 3a). However, MVI with higher concentrations provided poor cycle performance or low capacity (Fig. S9 in
Supporting information). Besides, precipitation occurred with 0.1 mol/L MVI in 1 mol/L NaCl(aq) (Fig. S10 in Supporting information).
Because extra introduction of NaCl may decrease the solubility of the more hydrophobic and reduced monovalent viologen radical cation.
Therefore, 0.05 mol/L is the most suitable concentration of MVI 1 mol/L NaCl aqueous solution for battery operation.
2
in 1 mol/L NaCl aqueous solution was also
2
2
2
2
2
reached the most stable cycle
2
2
2
a)100
3
2
1
0
b)2
120
80
40
0
1
75
50
25
0
CE
VE
EE
Charge
Discharge
0
-
1
2
-
0
100 200 300 400
Cycle
448 452 456 460 464 468
Time (h)
c)120
d)100
16
1
8
4
2
8
0
0
0
75
CE
VE
EE
Charge
Discharge
5
0
5
0
CE
VE
EE
4
2
2
3
4
5
2
0
30
60
Cycle
90 120
-
Current Density (mA cm )
Fig. 3. Electrochemical performance of MVI
2
redox flow battery test. (a) Long-term cycling performance (0.2 V to 1.2 V) of a non-pumped battery at a constant
= 0.05 mol/L, in 1 mol/L NaCl(aq). (b) Polarity-reversal test of the symmetric redox flow battery.
2
current density of 2 mA/cm , concentration of active material c
0
2
2
(
c) Coulombic, voltage and energy efficiency of a flow cell depending on the applied current density from 2 mA/cm to 5 mA/cm . (d) Cycling performance (0.2
2
0
V to 1.2 V) of a flow battery at a current density of 4 mA/cm . Concentration of active material c = 0.1 mol/L, in 1 mol/L NaCl(aq).
According to the long-term cycling test of the non-pumped battery, the proper range of the working potential can be investigated. As
shown in Fig. 3a, the battery with 0.05 M MVI active material as the electrolyte exhibited an excellent stability and an average coulombic
2
efficiency of 99.30% for 450 cycles. Although there was a capacity increasing in previous cycles due to the membrane and electrode
activation process, an overall excellent performance can still be obtained. The small polarization of charge/discharge curves gives the
battery an average voltage efficiency of 90.07%, thus a high average energy efficiency of 89.44% is obtained. The above 99% coulombic
efficiency can be regarded as an indicator of the prevention of crossover problem [30]. It is worth noting that there was no colour change

Please donot adjust the margins
in catholyte after charge/discharge process and the original transparent yellow was maintained after 450 cycles (Fig. S11 in Supporting
information), suggesting that the crossover problem was prohibited as well. Attempt on polarity-reversal test of non-pumped symmetric
RFBs was investigated according to the following process, the assembled cell was charged/discharged in the voltage range of 0.2 V to1.2
V for ten cycles, and then the charge/discharge voltage was reversed in the range of −1.2 V to −0.2 V. The battery can work stably for
more than 500 hours (Fig. S12 in Supporting information). Further investigation of fully polarity-reversal test was implemented in a
voltage window of −1.2 V to 1.2 V for 100 cycles, achieving a high average coulombic efficiency of 99.93%. In addition, water
electrolysis was avoided completely (Fig. 3b and Fig. S13 in Supporting information).
Flow battery was further investigated combining 0.1 mol/L active material (MVI ) with 1 mol/L supporting electrolyte (NaCl). The
2
2
2
battery was tested at varied current density from 2 mA/cm to 5 mA/cm to give further investigation of the electrochemical performance.
2
As shown in Fig. 3c, a high energy efficiency of 74.94% can be achieved at a low current density of 2 mA/cm , even at a current density
2
of 5 mA/cm , the energy efficiency still exceeds 51.46%. Long-term cycling test of the flow battery was measured at a constant current
2
of 4 mA/cm and a voltage range from 0.2 V to 1.2 V (Fig. 3d). The capacity was able to reach up to 3.33 mAh, higher than that in non-
pumped battery with the same MVI
2
concentration (Fig. S9a). Moreover, the flow battery showed a capacity retention of 98.96% after
100 cycles charge/discharge test, higher than some other viologen-based RFBs [31], though it still remains to be improved. The
fluctuation of electrochemical result observed in Fig. 4d during the flow battery test may attribute to the electrolyte when circulated
through a peristaltic pump. Some inevitable jitters and temperature variation will cause the data to fluctuate as well. Although the overall
electrochemical performance of this MVI
and the low cost AHA anion exchange membrane reduces the total capital cost compared with nafion series membranes (e.g., nafion-
15, 117, 212), and the combination achieves better capacity than nafion-based batteries (Fig. S14 in Supporting information) and shows
2
-based-SARFB battery remains to be improved, the use of cheap and benign NaCl electrolyte
1
comparative energy efficiency compared with some similar works [16, 21, 25]. This design provides an effective strategy to solve the
crossover problems of ARFBs, and points out the direction for the development of green and low-cost SARFBs simultaneously. It can
be expected that if the concentration of MVI
electrochemical behaviours of SARFBs.
2
can be enhanced through molecule modification, there will be much outstanding
In conclusion, the abovementioned bipolar redox active material MVI
2
was synthesized through a simple one-step methylation
reaction, with a high yield of more than 95%. For the first time the small organic molecule with both cathodic and anodic redox active
centers were applied for symmetric aqueous redox flow batteries and its bipolar character was investigated. Based on the fact that the
organic salt possesses excellent conductivity which is even comparable to that of sodium chloride, battery test using only MVI solution
2
without any supporting electrolyte evidenced our hypothesis of constructing a system without supporting electrolyte. The non-pumped
battery tested from 0.2 V to 1.2 V exhibited high voltage efficiency of 90.07%, high energy efficiency of 89.44%, and high coulombic
efficiency of 99.30%. The crossover problem is prevented to some extent. Electrochemistry tests exhibit a well polar-reversal capability
of the bipolar characterized MVI , with a rather stable performance of an average coulombic efficiency of 99.93% during
2
charge/discharge in the voltage range from −1.2 V to 1.2 V. Flow battery achieved a good performance with higher concentration of
active material compared with non-pumped battery. This design is also an inspiration for other SARFBs.
Declaration of interests
☒
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
Acknowledgment
This work was supported by the National Key R&D Program of China (Nos. 2016YFA0202500 and 2016YFB0901502), the National
Natural Science Foundation of China (NSFC, Nos. 21673243, 51771094 and 21805141), the Ministry of Education (MOE) of China (No.
B12015) and Tianjin High-Tech (No. 18JCZDJC31500).

Please donot adjust the margins
References
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
1] G.L. Soloveichik, Chem. Rev. 115 (2015) 11533-11558.
2] M. Park, J. Ryu, W. Wang, J. Cho, Nat. Rev. Mater. 2 (2016) 16080-16098.
3] Z. Yuan, H. Zhang, X. Li, Chem. Commun. 54 (2018) 7570-7588.
4] M. Skyllas-Kazacos, R. Robins, US Patent, No. 4, 786, 567, 1986.
5] M. Skyllas-Kazacos, M.H. Chakrabarti, S.A. Hajimolana, F.S. Mjalli, M. Saleem, J. Electrochem. Soc. 158 (2011) R55-R79.
6] E. Sum, M. Skyllas-Kazacos, J. Power Sources 15 (1985) 179-190.
7] M. Skyllas-Kazacos, M. Rychcik, R.G. Robins, A.G. Fane, M.A. Green, J. Electrochem. Soc. 133 (1986) 1057-1058.
8] Á. Cunha, J. Martins, N. Rodrigues, F.P. Brito, Int. J. Energy Res. 39 (2015) 889-918.
9] M. Skyllas-Kazacos, G. Kazacos, G. Poon, H. Verseema, Int. J. Energy Res. 34 (2010) 182-189.
10] P.G. Rasmussen, US Patent, No. 8, 080, 327 B1, 2011.
11] R.A. Potash, J.R. McKone, S. Conte, H.D. Abruña, J. Electrochem. Soc. 163 (2016) A338-A344.
12] W.T. Duan, R.S. Vemuri, J.D. Milshtein, et al., J. Mater. Chem. A 4 (2016) 5448-5456.
13] T. Ma, Z. Pan, L. Miao, et al., Angew. Chem. Int. Ed. 57 (2018) 3158-3162.
14] J. Winsberg, C. Stolze, S. Muench, et al., ACS Energy Lett. 1 (2016) 976-980.
15] Y. Zhu, F. Yang, Z. Niu, et al., J. Power Sources 417 (2019) 83-89.
16] T. Janoschka, N. Martin, M.D. Hager, U.S. Schubert, Angew. Chem. Int. Ed. 55 (2016) 14427-14430.
17] K. Murugavel, Polym. Chem. 5 (2014) 5873-5884.
18] Y. Gu, W. Hong, W. Choi, et al., J. Electrochem. Soc. 161 (2014) H716-H721.
19] C. DeBruler, B. Hu, J. Moss, et al., Chem 3 (2017) 961-978.
20] B. Hu, Y. Tang, J. Luo, et al., Chem. Commun. 54 (2018) 6871-6874.
21] T.B. Liu, X.L. Wei, Z.M. Nie, V. Sprenkle, W. Wang, Adv. Energy Mater. 6 (2016) 1501449-1501457.
22] C. DeBruler, B. Hu, J. Moss, J. Luo, T.L. Liu, ACS Energy Lett. 3 (2018) 663-668.
23] J. Luo, W.D. Wu, C. Debruler, et al., J. Mater. Chem. A 7 (2019) 9130-9136.
24] C.L. Bird, A.T. Kuhn, Chem. Soc. Rev. 10 (1981) 49-82.
25] B. Hu, C. DeBruler, Z. Rhodes, T.L. Liu, J. Am. Chem. Soc. 139 (2017) 1207-1214.
26] J. Winsberg, C. Stolze, A. Schwenke, S. Muench, M.D. Hager, U.S. Schubert, ACS Energy Lett. 2 (2017) 411-416.
27] A. Orita, M.G. Verde, M. Sakai, Y.S. Meng, Nat. Commun. 7 (2016) 13230-13238.
28] K. Lin, Q. Chen, M.R. Gerhardt, et al., Science 349 (2015) 1529-1532.
29] R.S. Nicholson, Anal. Chem. 37 (1965) 1351-1355.
30] J. Winsberg, T. Hagemann, T. Janoschka, M.D. Hager, U.S. Schubert, Angew. Chem. Int. Ed. 56 (2017) 686-711.
31] B. Hu, T. L. Liu, J. Energy Chem. 27 (2018) 1326-1332.