COMMUNICATION

Redox Flow Batteries www.advmat.de

Phenothiazine-Based Organic Catholyte for High-Capacity

and Long-Life Aqueous Redox Flow Batteries
Changkun Zhang, Zhihui Niu, Sangshan Peng, Yu Ding, Leyuan Zhang, Xuelin Guo,
Yu Zhao,* and Guihua Yu*

redox-active polymers,[4] nitroxide rad-

Redox-active organic materials have been considered as one of the most ical compounds,[5] metal-complexes,[6]
promising “green” candidates for aqueous redox flow batteries (RFBs) due to and heteroaromatic compounds,[7] have
the natural abundance, structural diversity, and high tailorability. However, been explored as suitable materials for
many reported organic molecules are employed in the anode, and molecules with rechargeable solid-electrode batteries and
redox flow batteries (RFBs).
highly reversible capacity for the cathode are limited. Here, a class of hetero­
In aqueous RFBs, a series of the
aromatic phenothiazine derivatives is reported as promising positive materials redox-active organic materials have been
for aqueous RFBs. Among these derivatives, methylene blue (MB) possesses investigated: quiones,[7b,8] viologens,[5,9]
high reversibility with extremely fast redox kinetics (electron-transfer rate alloxazines,[10] phenazines,[11] TEMPOs
constant of 0.32 cm s−1), excellent stability in both neutral and reduced states, (2,2,6,6-tetramethyl-1-piperidinyloxy),[4,5]
and ferrocenes[12] (Figure  1a). Benefited
and high solubility in an acetic-acid–water solvent, leading to a high reversible
from the π-conjugated structure, alloxa-
capacity of ≈71 Ah L−1. Symmetric RFBs based on MB electrolyte demonstrate zine- and phenazine-based nitrogen-
remarkable stability with no capacity decay over 1200 cycles. Even concen- containing heteroaromatic molecules
trated MB catholyte (1.5 m) is still able to deliver stable capacity over hun- exhibit more stable electrochemical revers-
dreds of cycles in a full cell system. The impressive cell performance validates ibility than quione- and viologen-based
the practicability of MB for large-scale electrical energy storage. molecules when served as anolytes.[10a,11]
To date, only TEMPO, ferrocene and their
derivatives display relatively high redox
potential, and have been investigated as
Nowadays, the extensive exploitation of fossil fuels and the catholyte in aqueous RFBs.[4,5,12] However, the low ionic con-
increasing demand for large-scale electrical energy storage ductivity of the ion-selecting membrane in neutral electrolytes
call for sustainable and innovative technologies for generating limits the operating current density and rate performance.
and storing energy from renewable resources.[1] Compared Moreover, the demonstrated RFBs are still largely operated at
with inorganic materials that have been applied in commercial low concentrations though high solubility of these derivatives
batteries, redox-active organic materials are advantageous in have been reported (>2 m). The catholyte beyond 2 m of equiva-
structural diversity, resource sustainability, tunable properties, lent electron concentration has yet to be reported.
and potential material cost.[1b,2] Consequently, various kinds To explore positive redox-active molecules, one can intro-
of organic compounds including carbonyl compounds,[1b,3] duce S atoms or other electron withdrawing groups to increase
the electron affinity of the molecules, and thus to increase the
redox potentials.[3b,13] For instance, when substituting one N of
Dr. C. Zhang, S. Peng, Dr. Y. Ding, L. Zhang, X. Guo, Prof. G. Yu phenazine-based heteroaromatic compounds by S atom to form
Materials Science and Engineering Program thiazine-based compounds, a redox potential nearly 4.0 V was
and Department of Mechanical Engineering achieved in solid-electrode batteries.[7a,14] Adding a second S
The University of Texas at Austin
Austin, TX 78712, USA atom leading to thianthrene-based compounds raises the redox
E-mail: ghyu@austin.utexas.edu potential even higher.[13b,c] As one of the typical thiazine het-
Z. Niu, Prof. Y. Zhao eroaromatic compounds, phenothiazine (PTZ) and its deriva-
Institute of Functional Nano & Soft Materials tives (Figure 1b) are highly bioactive and have widespread use
Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices in the textile and pharmaceutical industries.[15] Most of the
Collaborative Innovation Centre of Suzhou Nano Science
reported PTZ derivatives possess high redox potential. How-
and Technology
Joint International Research Laboratory of Carbon-Based Functional ever, the application of PTZ-based redox-active molecules is
Materials and Devices mainly studied in nonaqueous systems in which PTZ deriva-
Soochow University tives have limited solubility especially for their charged radical
199 Renai Road, Suzhou Industrial Park, Suzhou, Jiangsu 215123, China states.[16] Besides, nonaqueous RFBs still fall behind the corre-
E-mail: yuzhao@suda.edu.cn
sponding aqueous counterparts in terms of ionic conductivity,
The ORCID identification number(s) for the author(s) of this article
ion-selecting membrane choices, and packaging process.[17] It
can be found under https://doi.org/10.1002/adma.201901052.
is, therefore, highly desirable to develop high capacity aqueous
DOI: 10.1002/adma.201901052 catholyte based on PTZ and its derivatives.

Adv. Mater. 2019, 1901052 1901052  (1 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advmat.de

Figure 1.  a) Comparison of various organic molecules for aqueous RFBs. The left axis represents the equivalent electron concentration (the max-
operated concentration of organic molecules × the number of electrons transferred). DHPS: 7,8-dihydroxyphenazine-2-sulfonic acid.[11] DHBQ:
2,5-dihydroxy-1,4-benzoquinone.[19] DHAQ: 2,6-dihydroxyanthraquinone.[20] 2,6-DBEAQ: 4,4′-((9,10-anthraquinone-2,6-diyl)dioxy)dibutyrate.[21]
ACA: 7/8-carboxylic acid.[10b] FMN-Na: flavin mononucleotide.[10a] MV: methyl viologen.[5] (SPr)2V: 1,1′-bis[3-sulfonatopropyl]-4,4′-bipyridinium.[22]
[(NPr)2TTz]: 4,4′- (thiazolo[5,4-d]thiazole-2,5-diyl)bis(1-(3-(trimethylammonio) propyl)pyridin-1-ium) tetrachloride.[23] [(Me)(NPr)V]Cl3: 1-methyl-10-
[3-(trimethylammonio)propyl]-4,40-bipyridinium trichloride.[9] AQS: anthraquinone-2-sulfonic acid.[24] AQDS: 9,10-anthraquinone-2,7-disulphonic
acid.[7b] BTMAP-Fc: bis((3-trimethylammonio)propyl)ferrocene dichloride.[25] FcNCl: (ferrocenylmethyl)trimethylammonium chloride.[12] HO-TEMPO.[5]
TEMPOSP: TEMPO-4-sulfate potassium salt.[26] TEMPTMA: N,N,N-2,2,6,6-heptamethylpiperidinyl oxy-4-ammonium chloride.[27] b) Molecular struc-
tures of PTZ-based derivatives. c) CV curves of 10 × 10−3 m PTZ-based molecules. MB: methylene blue. CPZ: chlorpromazine. PMZ: promethazine. MB
and Azure A tested at 3.0 m H2SO4 solution. CPZ and PMZ tested at pH = 1 solution.

According to substitution types, PTZ-based derivatives
could be roughly categorized into thiazine dye derivatives
(Type I) and N-substituted phenothiazine derivatives (N-PTZs,
Type II) (Figure 1b). The two types exhibit different redox reac-
tion mechanisms. Thiazine dye derivatives have a cation in the
tricyclic structure connecting one chloride or acetate anion,
and their redox reactions involve two-electron transfers at the
nitrogen position and aromatic rings. Meanwhile for N-PTZs,
they are able to donate one electron to form a radical-cation at the
nitrogen position, but further donating another electron would
result in significant structural instability of the molecule.[18] As
shown in Figure 1c and Table S1 in the Supporting Informa-
tion, the typical thiazine dye derivatives, methylene blue (MB),
and Azure A have redox potentials of 0.57 and 0.59 V versus
NHE (normal hydrogen electrode), respectively, which are
higher than the reported quinones and other heteroaromatic
molecules. The N-PTZs display even higher redox potential
(0.86 and 0.92 V versus NHE for chlorpromazine (CPZ) and
promethazine (PMZ) molecules at pH = 1, respectively).
Due to a highly hydrophobic tricyclic scaffold, the pristine
PTZ is insoluble in water. However, it is possible to enhance its
solubility by functionalization of hydrophilic side chains and/or
optimization of electrolytes. The improved solubility could be
achieved via adding an ammonium group at nitrogen or both
sides of ring (Table S1, Supporting Information). Decreasing
the symmetry of molecules could benefit to enhance solubility
in water. For instance, when dialkyl groups of MB are substi-
tuted by hydrogens (Azure A), the solubility is improved several
times greater than that of MB. However, it still cannot meet the
requirement of RFBs (≈0.6 m in H2SO4 solution), even though
MB was tried for RFBs with unclear battery chemistry and
incompetent cell performance, showing low solubility and poor
stability of MB in the reported conditions.[7d,28] Moreover, the
demonstrated performance lagged far behind of the reported
organic aqueous RFBs.[7b,10,11,28] In this work, we found that
MB exhibits a considerably high solubility by rational electro-
lyte optimization (≈1.8 m in H2O + acetic acid (AA) + H2SO4
mixture and >2.0 m in AA) and no deposition occurred after
>100 days storage in 1.5 m MB, and elucidated more clearly
the redox characteristics. Density functional theory (DFT) cal-
culation by implicit solvent mode reveals that MB has similar
solvation energies in H2O and AA solvents (−41.6 kcal mol−1

Adv. Mater. 2019, 1901052 1901052  (2 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Figure 2.  a) CV curves of 10 × 10−3 m MB molecules at different H2SO4 concentrations. b) Ipa/Ipc ratios, diffusion coefficient (D) and electron-transfer
rate constants (k0) at different H2SO4 concentrations (Ipa and Ipc are the anodic and cathodic currents at 10 mV s−1).
in H2O and −38.6 kcal mol−1 in AA). However, according to
the explicit solvent model that represents more realistic inter-
action between solvent and solute molecules (Figures S2–S3,
Supporting Information), the calculated adsorption energy of
MB in AA solvent (−214 kcal mol−1) is almost 4 times than in
H2O (−56 kcal mol−1). The stronger interaction between AA
and MB would reduce the lattice energy of MB and increase
its solubility in AA. Besides, the strong hydrogen bonding in
H2O–H2O molecules also inhibits the binding of H2O with MB
molecule, resulting in low solubility in H2O. Further UV–vis
and Fourier-transform infrared spectroscopy (FTIR) spectra
confirmed that there was no structural change for MB in AA
solvent (Figures S4 andS5, Supporting Information).
Electrochemical analysis and DFT calculations were per-
formed to investigate the electrochemical properties of MB
molecules. The redox potential of MB was pH dependent
(Figure  2a). At neutral condition, MB exhibited poor elec-
trochemical activity and reversibility. With the increasing of
acidity by adding H2SO4 (H2SO4 concentration ranging from
0.1 to 3.0 m), the redox potential gradually increased from
0.45 to 0.57 V versus NHE. In addition, the Ipa/Ipc (Ipa and Ipc
stand for the peak anodic and cathodic currents, respectively)
was promoted from 0.76 in 0.1 m H2SO4 to 1.02 in 3 m H2SO4
(Figure 2b), revealing significantly improved reversibility. The
diffusion coefficient of MB calculated by the Levich equation
is 2.05 × 10−6 cm2 s−1 in 0.1 m H2SO4. The diffusion coefficient
had a minor negligible change within the measured range of
H2SO4 concentration. Moreover, the electron-transfer rate con-
stant (k0) of MB (Figures S6–S9, Supporting Information) was
enhanced with the increase of H2SO4 concentration, reaching
0.32 cm s−1 in 3.0 m H2SO4 which is several orders of magni-
tude greater than VO2+/VO2+ redox couple[29] and the reported
redox-active organic materials.[5,7b,10]
As a heteroaromatic compound, MB molecule exhibits
two resonance forms, in which the positive charge is located
at the S atom of the middle ring or nitrogen of the side chain
(Figure  3a). The redox process of MB is dependent on the
acidity of the electrolyte due to the protonation/deprotonation of
MB involved in the redox reaction. In a weak acidic environment
(pH > 5.6), the redox reactions occur between the configuration
Adv. Mater. 2019, 1901052
www.advmat.de
of MB and Leuco-MB with two electrons and one proton to bal-
ance the charge. While in relatively strong acidic environment
(pH < 5.6), the redox reactions occur between the configuration
of MB and R-MB with two electrons and two protons to balance
the charge with the extra positive charge locating at the N of the
side chain. At pH = 5.6, there is a balance between the Leuco-
MB and R-MB.
To better understand the redox mechanism, the nucleus
independent chemical shift (NICS) and anisotropy of the
induced current density (ACID) calculations were conducted
on the electronic structures of different redox products
(Figure 3b). ACID is a magnetic indicator of aromaticity for
visualizing ring-currents and electron delocalization.[30] Along
the periphery of the rings, pristine MB showed clear diamag-
netic clockwise ring current, which was composed of one weak
aromatic thiazine ring (NICS(1)zz  =  −6.7 ppm, Figure 3c) and
the two aromatic benzenoid rings (NICS(1)zz  =  −16.8 ppm,
Figure 3c). Upon reduction, the benzenoid rings of both Leuco-
MB and R-MB exhibited strong aromatic while the central thia-
zine ring became antiaromatic, which was in consistent with
the ACID plot. Note that the two benzenoid rings of Leuco-
MB and R-MB were more aromatic than MB, suggesting good
stability of the reduced products. In addition, the Mayer bond
orders (Figure S10, Supporting Information) indicated that
both Leuco-MB and R-MB exhibited stable rigid molecular
structure though was slightly decreased in bond order for the
middle thiazine ring upon reduction.[31]
Furthermore, the energy levels of the lowest unoccupied
molecular orbital (LUMO) and the highest occupied molecular
orbital (HOMO) of MB, R-MB, and Leuco-MB were also cal-
culated using DFT. The HOMO plots display that the MB,
R-MB, and Leuco-MB are structurally stable because of effec-
tive electron delocalization and conjugation (Figure 3d). The
HOMO–HOMO energy level difference between MB and R-MB
(0.43 eV) is smaller than that of MB and Leuco-MB (0.73 eV),
suggesting a smaller energy barrier for the redox reactions.[32]
Moreover, R-MB shows a lower energy gap (Eg) than Leuco-MB,
indicating that the intrinsic electronic conductivity of R-MB
should be higher than for Leuco-MB. In addition, the chemical
and electrochemical stability of MB electrolyte was evaluated in
1901052  (3 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advmat.de

Figure 3.  a) pH-dependent redox process of MB. b–c) ACID plots and NICS values of MB, R-MB, and Leuco-MB. The molecular planes are placed
perpendicular to the magnetic field vector. Small green arrows are computed current density vectors. The isovalue for the surfaces is 0.05 a.u. The
paratropic/diatropic ring currents indicate antiaromaticity and aromaticity, respectively. The circles labeled by red and blue arrows indicate the dia-
tropic and paratropic ring current flow. d) Optimized structure and calculated energy levels of MB, R-MB, and Leuco-MB using DFT method under
M052x/6-31g(d) package.

acidic solution (0.1 m MB in H2O + AA (volume ratio of 3:1)
containing 3.5 m H2SO4). The electrolyte exhibited good sta-
bility in 2000 CV cycles without obvious decay (Figure S11a,
Supporting Information). Even stored in ambient condition for
120 days, this electrolyte still showed similar electrochemical
activity (Figure S11b, Supporting Information), demonstrating
its excellent chemical stability in acidic condition.
The symmetric cell is a direct probing method to test elec-
trochemical stability of MB electrolyte at full charge/discharge
condition.[33] A symmetric cell containing 0.2 m MB displayed
a capacity of 5.32 Ah L−1 at 80 mA cm−2, corresponding to a
utilization of 99.3% (Figure 4a). No capacity decay was observed
after 1200 cycles at this current density (Figure 4a) as well as
at other current densities (Figure S12, Supporting Informa-
tion). The high utilization ratio of MB together with capacity
retention indicated that MB was highly stable during full oxi-
dation and reduction cycling. Even at high concentration, MB
still maintained highly stable capacity with coulombic efficiency
(CE) and energy efficiency (EE) of nearly 100% (Figure 4b,c).
The symmetric cell based on 1.0 m MB produced a capacity
of 23.9 Ah L−1 (based on the full electrolytes) with excellent
capacity retention of 99.6% after 520 cycles (over 24 days) at
100 mA cm−2. Both of MB and R-MB molecules had stable
molecular structure and there was still no peak shift for the
UV–vis spectra after resting for above 3 months without any
inert protection (Figure S13, Supporting Information). The
good stability of R-MB molecules was also confirmed by the
unchanged charge/discharge capacity after standing for 48 h in
1.0 m symmetric cell (Figure S14, Supporting Information).
To evaluate the performance of MB in a full cell, vanadium
(II) (denoted as V(II)) anolyte (10 mL, 0.8 m V(II)) was paired
with MB catholyte (5 mL, 0.1 m). Excess V(II) anolyte was used
to determine the electrochemical stability of catholyte during
cycling. It is of noting that Lee et al. also assembled a flow cell
in which MB was taken as anolyte and V(V4+/V5+) as catho-
lyte, however the cell performance lagged far behind other
reported aqueous systems.[28] In this work, the V-MB flow cell
demonstrated capacities of 26.2, 25.6, 25.1, and 23.6 mAh at

Adv. Mater. 2019, 1901052 1901052  (4 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advmat.de

Figure 4.  a) Cycling capacities and CE of the symmetric RFB at 80 mA cm−2 with 0.2 m MB electrolyte (inset: charge/discharge profiles of the symmetric
RFB at different cycles). b) Charge/discharge profiles of the symmetric RFB at different cycles with 1.0 m MB electrolyte at 100 mA cm−2. c) Cycling
capacities and efficiencies of the symmetric RFB with 1.0 m MB electrolyte at 100 mA cm−2. d) Charge–discharge profiles of the V-MB RFB at different
current densities. e) Cycling capacities and efficiencies of the V-MB RFB with 0.1 m MB over 900 cycles at 80 mA cm−2. The data shown are selected
from every tenth of cycle.

60, 80, 100, and 140 mA cm−2, respectively (Figure 4c), and
had no sign of capacity decay over 900 cycles with CE nearly
100% (Figure 4d and Figure S15, Supporting Information).
The slight fluctuation of capacity should result from the varia-
tion of environment temperature. Compared with the capacity
retention and CE, the EE gradually decreased with the cycling.
The pheno­menon was observed in other organic RFBs.[10b,11]
The major reasons are the decreased ionic conductivity of V(II)
anolyte (Figure S16, Supporting Information) and the crossover
and oxidation of V(II) ions.[34]
The cell performance in concentrated electrolyte is important
when evaluating the potential of a new redox-active material for
practical RFBs. Till now, very limited organic compounds have
been evaluated with concentration higher than 1 m.[7b,10b,11]

Adv. Mater. 2019, 1901052 1901052  (5 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advmat.de

Figure 5.  a) Charge–discharge profiles of V-MB RFBs at different MB concentrations. b) Cycling efficiencies of V-MB RFB with 1 m MB catholyte at
different current densities (inset: corresponding charge–discharge profiles). c) Cycling capacities and CEs of the V-MB RFBs using 1.2 and 1.5 m MB
catholyte at 80 mA cm−2 (inset: charge–discharge profiles over time of V-MB RFB at 1.5 m MB).

Benefited from high solubility in H2O + AA solvent, the MB
catholyte was able to deliver capacities of 49, 59, and 70 Ah L−1
in 1.0, 1.2, and 1.5 m MB with the corresponding utilization of
91%, 92%, and 87% at 80 mA cm−2, respectively (Figure  5a),
which are superior to the reported aqueous organic electro-
lytes (Table S2, Supporting Information).[4,7b,10,11] At 1.0 m MB
catholyte, the EEs of the cell were about 76%, 73%, 70%, and
66% at 80, 100, 140, and 200 mA cm−2, respectively, with the
corresponding CEs of nearly 100% (Figure 5b and Figure S18,
Supporting Information). Even at 1.5 m, the flow cell delivered
performance high catholyte capacity of 71 Ah L−1 at 40 mA cm−2
with EE of ≈80% (Figure S17a,b, Supporting Information).
Cycling tests of the cell with 1.2 and 1.5 m MB catholyte dem-
onstrated that the V-MB RFB showed good capacity reten-
tion after 160 and 50 cycles (with capacity fade of 0.025% per
cycle and 0.52% per day for 1.2 m, 0.074% per cycle and 0.76%
per day for 1.5 m), respectively (Figure 5c,d and Figure S17d,
Supporting Information). The slight fluctuation of capacity and
EE was resulted from refreshing of the V(II) anolyte. Compared
with the long cycling at the symmetric cell (Figure 4b,c), the
major reasons for shortened cycling life at high concentra-
tion should be attributed to the unstable V(II) anolyte without
inert atmosphere protection and the crossover of vanadium
ions during long-term operation. Excess anolyte could ensure
the full discharge/charge of the MB catholyte in initial cycling.
However, the oxidation of V(II) and the crossover of vanadium
ions resulted in the lack of anolyte as the cycling proceeded,
thereby leading to the capacity fading of V-MB RFB. Experi-
ments confirmed that the capacity could be recovered with
refreshed anolyte (Figure S18, Supporting Information).
Thiazine dye derivatives can be optimized through function-
alization with electron-withdrawing groups to further increase
the redox potential and molecular polarity, thereby achieving
high cell voltage and enhanced solubility, which are the two
key factors related to the energy density of the proposed cell.
Selected thiazine dye derivatives were subjected to DFT calcula-
tions on the energy level of the molecular orbitals (Table S3,
Supporting Information). The LUMOs are linearly related to the
reduction potentials and molecules with more negative LUMO
possesses higher redox potential. The energy levels of LUMO
could be lowered when one N(CH3)2 group in MB molecule
was substituted by H, NH2, COOH, and N(CH3)(CH2)2NH3+
groups. Moreover, the corresponding solvation energy in water
showed more negative values than MB (Figure S19, Supporting

Adv. Mater. 2019, 1901052 1901052  (6 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Information), suggesting better solubility in water. Further-
more, N-PTZ derivatives such as PMZ and CPZ possess high
redox potential and good solubility that could further enhance
the energy density of RFBs. However, their redox reactions
involve the generation of free radicals, which are highly active
and may result in decreased stability. Appropriate molecular
modification and electrode optimization, thus, would be neces-
sary to tune the chemical properties in the future work.[15a,18]
In summary, PTZ-based heteroaromatic molecules represent
a promising candidate as redox-active material in aqueous RFBs.
As a representative PTZ-based derivative, MB exhibits excep-
tional reversibility as well as structural stability. Electrochemical
analysis reveals that both MB and its reduced state exhibit excel-
lent stability and high solubility in acidic solution. When coupled
with V(II) anolyte, the cell displayed stable capacity retention
in 900 cycles at 0.1 m MB and a high capacity of 71 Ah L−1
based on a 1.5 m MB catholyte. The demonstrated capacity and
cycling performance are superior to previous organic materials
(Table S2 and Figure S20, Supporting Information). In virtue of
the DFT theoretical modeling, better understanding of the PTZ-
based heteroaromatic class is achieved to help tailor high-per-
formance molecules with high redox potential as well as good
solubility. The introduction of PTZ-based redox-active molecules
paves the way toward the development of new organic cathode
materials for high-performance aqueous RFBs.
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
G.Y. acknowledges financial support from Exxon Mobil Corp through its
membership with University of Texas at Austin, Camille Dreyfus Teacher-
Scholar Award, and Sloan Research Fellowship. Y.Z. acknowledges the
financial support from the National Natural Science Foundation of
China (Grant No. 51772199), the Collaborative Innovation Centre of
Suzhou Nano Science & Technology, the Base for Introducing Talents
of Discipline to Universities, and the Joint International Research
Laboratory of Carbon-Based Functional Materials and Devices.
Conflict of Interest
The authors declare no conflict of interest.
Keywords
catholytes, energy storage, organic redox flow batteries, phenothiazine
Received: February 14, 2019
Revised: April 3, 2019
Published online:
[1] a) C. P. Grey, J. M. Tarascon, Nat. Mater. 2017, 16, 45; b) Y. Liang,
Y. Jing, S. Gheytani, K.-Y. Lee, P. Liu, A. Facchetti, Y. Yao, Nat. Mater.
2017, 16, 841; c) Y. Shi, L. Peng, Y. Ding, Y. Zhao, G. Yu, Chem. Soc.
Adv. Mater. 2019, 1901052 1901052  (7 of 8)
www.advmat.de
Rev. 2015, 44, 6684; d) M. R. Lukatskaya, B. Dunn, Y. Gogotsi, Nat.
Commun. 2016, 7, 12647.
[2] a) J. Winsberg, T. Hagemann, T. Janoschka, M. D. Hager,
U. S. Schubert, Angew. Chem., Int. Ed. 2017, 56, 686; b) Y. Zhao,
Y. Ding, Y. Li, L. Peng, H. R. Byon, J. B. Goodenough, G. Yu, Chem.
Soc. Rev. 2015, 44, 7968.
[3] a) Y. Ding, Y. Li, G. Yu, Chem 2016, 1, 790; b) Q. Zhao, Z. Zhu,
J. Chen, Adv. Mater. 2017, 29, 1607007; c) C. Zhang, Z. Niu, Y. Ding,
L. Zhang, Y. Zhou, X. Guo, X. Zhang, Y. Zhao, G. Yu, Chem 2018,
4, 2814; d) Y. Ding, G. Yu, Angew. Chem., Int. Ed. 2016, 55, 4772;
e) Y. Ding, G. Yu, Angew. Chem., Int. Ed. 2017, 56, 8614.
[4] T. Janoschka, N. Martin, U. Martin, C. Friebe, S. Morgenstern,
H. Hiller, M. D. Hager, U. S. Schubert, Nature 2015, 527, 78.
[5] T. Liu, X. Wei, Z. Nie, V. Sprenkle, W. Wang, Adv. Energy Mater. 2016,
6, 1501449.
[6] a) Y. Zhao, Y. Ding, J. Song, G. Li, G. Dong, J. B. Goodenough, G. Yu,
Angew. Chem., Int. Ed. 2014, 53, 11036; b) C. Yang, G. Nikiforidis,
J. Y. Park, J. Choi, Y. Luo, L. Zhang, S.-C. Wang, Y.-T. Chan, J. Lim,
Z. Hou, M.-H. Baik, Y. Lee, H. R. Byon, Adv. Energy Mater. 2018, 8,
1702897; c) X. Wei, L. Cosimbescu, W. Xu, J. Z. Hu, M. Vijayakumar,
J. Feng, M. Y. Hu, X. Deng, J. Xiao, J. Liu, V. Sprenkle, W. Wang, Adv.
Energy Mater. 2015, 5, 1400678; d) C. Xie, W. Xu, H. Zhang, X. Hu,
X. Li, Chem. Commun. 2018, 54, 8419.
[7] a) M. Kolek, F. Otteny, P. Schmidt, C. Muck-Lichtenfeld, C. Einholz,
J. Becking, E. Schleicher, M. Winter, P. Bieker, B. Esser, Energy
Environ. Sci. 2017, 10, 2334; b) B. Huskinson, M. P. Marshak,
C. Suh, S. Er, M. R. Gerhardt, C. J. Galvin, X. Chen, A. Aspuru-
Guzik, R. G. Gordon, M. J. Aziz, Nature 2014, 505, 195; c) C. Zhang,
L. Zhang, Y. Ding, S. Peng, X. Guo, Y. Zhao, G. He, G. Yu, Energy
Storage Mater. 2018, 15, 324; d) A. M. Kosswattaarachchi,
T. R. Cook, ChemElectroChem 2018, 5, 3437.
[8] a) Y. Ding, C. Zhang, L. Zhang, H. Wei, Y. Li, G. Yu, ACS Energy Lett.
2018, 3, 2641; b) K. Wedege, E. Dražević, D. Konya, A. Bentien, Sci.
Rep. 2016, 6, 39101; c) C. Zhang, L. Zhang, Y. Ding, X. Guo, G. Yu,
ACS Energy Lett. 2018, 3, 2875; d) L. Zhang, C. Zhang, Y. Ding,
K. Ramirez-Meyers, G. Yu, Joule 2017, 1, 623.
[9] C. DeBruler, B. Hu, J. Moss, X. Liu, J. Luo, Y. Sun, T. L. Liu, Chem
2017, 3, 961.
[10] a) A. Orita, M. G. Verde, M. Sakai, Y. S. Meng, Nat. Commun.
2016, 7, 13230; b) K. Lin, R. Gómez-Bombarelli, E. S. Beh, L. Tong,
Q. Chen, A. Valle, A. Aspuru-Guzik, M. J. Aziz, R. G. Gordon, Nat.
Energy 2016, 1, 16102.
[11] A. Hollas, X. Wei, V. Murugesan, Z. Nie, B. Li, D. Reed, J. Liu,
V. Sprenkle, W. Wang, Nat. Energy 2018, 3, 508.
[12] B. Hu, C. Debruler, Z. Rhodes, T. Liu, J. Am. Chem. Soc. 2017, 139,
1207.
[13] a) Y. Ding, C. Zhang, L. Zhang, Y. Zhou, G. Yu, Chem. Soc. Rev.
2018, 47, 69; b) M. E. Speer, M. Kolek, J. J. Jassoy, J. Heine,
M. Winter, P. M. Bieker, B. Esser, Chem. Commun. 2015, 51, 15261;
c) A. Wild, M. Strumpf, B. Häupler, M. D. Hager, U. S. Schubert,
Adv. Energy Mater. 2017, 7, 1601415; d) Y. Zhao, Y. Ding, J. Song,
L. Peng, J. Goodenough, G. Yu, Energy Environ. Sci. 2014, 7, 1990.
[14] a) G. Dai, X. Wang, Y. Qian, Z. Niu, X. Zhu, J. Ye, Y. Zhao, X. Zhang,
Energy Storage Mater. 2019, 16, 236; b) C. Peng, G.-H. Ning, J. Su,
G. Zhong, W. Tang, B. Tian, C. Su, D. Yu, L. Zu, J. Yang, M.-F. Ng,
Y.-S. Hu, Y. Yang, M. Armand, K. P. Loh, Nat. Energy 2017, 2, 17074.
[15] a) S. Mahajan, R. K. Mahajan, Adv. Colloid Interface Sci. 2013,
199–200, 1; b) B. Seger, K. Vinodgopal, P. V. Kamat, Langmuir 2007,
23, 5471.
[16] a) J. D. Milshtein, A. P. Kaur, M. D. Casselman, J. A. Kowalski,
S. Modekrutti, P. L. Zhang, N. Harsha Attanayake, C. F. Elliott,
S. R. Parkin, C. Risko, F. R. Brushett, S. A. Odom, Energy Environ.
Sci. 2016, 9, 3531; b) H. Chen, Y. Zhou, Y.-C. Lu, ACS Energy
Lett. 2018, 3, 1991; c) A. P. Kaur, N. E. Holubowitch, S. Ergun,
C. F. Elliott, S. A. Odom, Energy Technol. 2015, 3, 476; d) J. Huang,
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com
Z. Yang, M. Vijayakumar, W. Duan, A. Hollas, B. Pan, W. Wang,
X. Wei, L. Zhang, Adv. Sustainable Syst. 2018, 2, 1700131;
e) J. A. Kowalski, M. D. Casselman, A. P. Kaur, J. D. Milshtein,
C. F. Elliott, S. Modekrutti, N. H. Attanayake, N. Zhang,
S. R. Parkin, C. Risko, F. R. Brushett, S. A. Odom, J. Mater. Chem.
A 2017, 5, 24371.
[17] a) M. Park, J. Ryu, W. Wang, J. Cho, Nat. Rev. Mater. 2017, 2, 16080;
b) W. Liu, W. Lu, H. Zhang, X. Li, Chem. - Eur. J. 2019, 25, 1649;
c) S. Peng, L. Zhang, C. Zhang, Y. Ding, X. Guo, G. He, G. Yu, Adv.
Energy Mater. 2018, 8, 1802533; d) L. Zhang, Y. Ding, C. Zhang,
Y. Zhou, X. Zhou, Z. Liu, G. Yu, Chem 2018, 4, 1035.
[18] M. H. Parvin, Electrochem. Commun. 2011, 13, 366.
[19] Z. Yang, L. Tong, D. P. Tabor, E. S. Beh, M.-A. Goulet, D. De Porcel
linis, A. Aspuru-Guzik, R. G. Gordon, M. J. Aziz, Adv. Energy Mater.
2017, 1702056.
[20] K. Lin, Q. Chen, M. R. Gerhardt, L. Tong, S. B. Kim, L. Eisenach,
A. W. Valle, D. Hardee, R. G. Gordon, M. J. Aziz, M. P. Marshak,
Science 2015, 349, 1529.
[21] D. G. Kwabi, K. Lin, Y. Ji, E. F. Kerr, M.-A. Goulet, D. De Porcell
inis, D. P. Tabor, D. A. Pollack, A. Aspuru-Guzik, R. G. Gordon,
M. J. Aziz, Joule 2018, 2, 1894.
[22] C. DeBruler, B. Hu, J. Moss, J. Luo, T. L. Liu, ACS Energy Lett. 2018,
3, 663.
[23] J. Luo, B. Hu, C. Debruler, T. L. Liu, Angew. Chem., Int. Ed. 2018, 57, 231.
Adv. Mater. 2019, 1901052 1901052  (8 of 8)
www.advmat.de
[24] M. R. Gerhardt, L. Tong, R. Gómez-Bombarelli, Q. Chen,
M. P. Marshak, C. J. Galvin, A. Aspuru-Guzik, R. G. Gordon,
M. J. Aziz, Adv. Energy Mater. 2017, 7, 1601488.
[25] E. S. Beh, D. De Porcellinis, R. L. Gracia, K. T. Xia, R. G. Gordon,
M. J. Aziz, ACS Energy Lett. 2017, 2, 639.
[26] J. Winsberg, C. Stolze, A. Schwenke, S. Muench, M. D. Hager,
U. S. Schubert, ACS Energy Lett. 2017, 2, 411.
[27] T. Janoschka, N. Martin, M. D. Hager, U. S. Schubert, Angew.
Chem., Int. Ed. 2016, 55, 14427.
[28] W. Lee, Y. Kwon, Korean Chem. Eng. Res. 2018, 56, 890.
[29] H. Fink, J. Friedl, U. Stimming, J. Phys. Chem. C 2016, 120,
15893.
[30] a) D. Geuenich, K. Hess, F. Köhler, R. Herges, Chem. Rev. 2005,
105, 3758; b) Q. Wang, P. Hu, T. Tanaka, T. Y. Gopalakrishna,
T. S. Herng, H. Phan, W. Zeng, J. Ding, A. Osuka, C. Chi, J. Siegel,
J. Wu, Chem. Sci. 2018, 9, 5100.
[31] A. J. Bridgeman, G. Cavigliasso, L. R. Ireland, J. Rothery, J. Chem.
Soc., Dalton Trans. 2001, 0, 2095.
[32] T. Ma, Z. Pan, L. Miao, C. Chen, M. Han, Z. Shang, J. Chen, Angew.
Chem., Int. Ed. 2018, 57, 3158.
[33] J. Luo, B. Hu, C. Debruler, Y. Bi, Y. Zhao, B. Yuan, M. Hu, W. Wu,
T. L. Liu, Joule 2019, 3, 149.
[34] S. Roe, C. Menictas, M. Skyllas-Kazacos, J. Electrochem. Soc. 2016,
163, A5023.
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim