Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous Organic Redox Flow Batteries

Full text of DOI:10.1016/j.chempr.2019.07.006

Article
Bio-inspired Molecular Redesign of a Multi-
redox Catholyte for High-Energy Non-
aqueous Organic Redox Flow Batteries
Giyun Kwon, Kyunam Lee,
Myeong Hwan Lee, ..., Soo
Young Park, Ji Eon Kwon, Kisuk
Kang
mracle2@snu.ac.kr (J.E.K.)
matlgen1@snu.ac.kr (K.K.)
HIGHLIGHTS
Highly soluble and multi-redox
BMEPZ for non-aqueous redox
flow batteries
Investigation of kinetic properties
and redox mechanism of BMEPZ
The highest energy density
among non-aqueous all-organic
redox flow batteries
Bio-inspired molecular redesign
of high-performance catholyte
material
A highly soluble and multi-redox phenazine-based molecule, BMEPZ, is
redesigned through bio-inspiratioin as high-performance catholyte material for
NORFBs. A full-flow RFB based on BMEPZ/FL redox couple exhibits cell voltage of
1.2 and 2.0 V and stable cycling. The highest energy density among NORFBs is
attained based on the multi-redox capability and high solubility.
Kwon et al., Chem 5, 1–15
October 10, 2019 ª 2019 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2019.07.006

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
Article
Bio-inspired Molecular Redesign of
a Multi-redox Catholyte for High-Energy
Non-aqueous Organic Redox Flow Batteries
Giyun Kwon,1,3 Kyunam Lee,2,3 Myeong Hwan Lee,1 Byungju Lee,1 Sechan Lee,1 Sung-Kyun Jung,1
1,4,
Kyojin Ku,1 Jihyeon Kim,1 Soo Young Park,2 Ji Eon Kwon,2, and Kisuk Kang
*
*
SUMMARY
The Bigger Picture
Ever-growing attention to
Redox-active organic materials (ROMs) have recently attracted significant atten-
tion for redox flow batteries (RFBs) to achieve green and cost-efficient energy
storage. In particular, multi-redox ROMs have shown great promise, and further
tailoring of these ROMs would yield RFB technologies with the highest possible
energy density. Here, we present a phenazine-based catholyte material, 5,10-
bis(2-methoxyethyl)-5,10-dihydrophenazine (BMEPZ), that undergoes two sin-
gle-electron redox reactions at high redox potentials (ꢀ0.29 and 0.50 V versus
Fc/Fc+) with enhanced solubility (0.5 M in acetonitrile), remarkable chemical sta-
bility, and fast kinetics. Moreover, an all-organic flow battery exhibits cell volt-
ages of 1.2 and 2.0 V when coupled with 9-fluorenone (FL) as an anolyte. It shows
capacity retention of 99.94% per cycle over 200 cycles and 99.3% per cycle with
0.1 M and 0.4 M BMEPZ catholyte, respectively. Notably, the BMEPZ/FL couple
results in the highest energy density (ꢁ17 Wh Lꢀ1) among the non-aqueous all-
organic RFBs reported to date.
environmental preservation has
accelerated the global shift
toward renewable energy.
However, complete replacement
of fossil fuel is yet impossible
because of intrinsic intermittency
of renewable energy. Thus, large-
scale energy storage systems with
safety and cost-effectiveness are
necessary to solve this limitation.
Non-aqueous all-organic redox
flow batteries (NORFBs), which
store energy in redox-active
organic materials (ROMs)
dissolved in non-aqueous
INTRODUCTION
solution, have received massive
attention as promising candidates
for this application. Current
development of NORFBs is
hindered by limited choices and
performances of ROMs. We
present the BMEPZ/FL system
delivering the highest energy
density among NORFBs. We
discuss the redesign of the
With the ever-increasing global demand for the development of greener and sus-
tainable energy sources to mitigate the environmental concerns associated with fos-
sil fuels, renewable energy sources such as solar and wind power are becoming
affordable and broadly deployed. To achieve round-the-clock energy delivery, how-
ever, these power sources must be paired with scalable energy storage systems
(ESSs) owing to the significant mismatch between the energy supply and demand.1,2
Redox flow batteries (RFBs), which utilize redox-active materials dissolved in sepa-
rate electrolytes, are considered some of the most promising ESSs for modern
grid markets, and decoupling the energy and power is regarded facile for RFB sys-
tems.3–8 With the aim of pursuing the development of green energy technology,
research on RFBs has also shifted from conventional metal-based redox-active ma-
terials such as vanadium and zinc to redox-active organic materials (ROMs), which
are naturally abundant and potentially cost effective, safe, and chemically
tunable.9–11 In particular, recent studies on non-aqueous all-organic RFBs (NORFBs)
have demonstrated the great promise for achieving high energy densities in these
systems without the concerns associated with water electrolysis, which typically
limits the working voltage to a narrow range and thus leads to a rather low energy
density in aqueous RFBs.12–17
catholyte material, BMEPZ, based
on inspiration from biosystems.
Our results of multi-electron redox
material at high potentials with
enhanced solubility provide a
breakthrough in the realization of
high-energy-density RFBs.
Despite the great potential of organic RFBs, their practical energy density remains
very low.10 The energy density of ROM-based RFBs is dependent on the following
Chem 5, 1–15, October 10, 2019 ª 2019 Elsevier Inc.
1

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
three factors: (1) the selection of the anolyte and catholyte and their corresponding
redox potential difference, (2) the amount of ROMs dissolved, and (3) the number
of electrons participating in the redox reaction per ROM. Extensive research in
recent years has led to the identification of various promising ROMs, such as quinone
and viologen derivatives, that can be used as anolytes (i.e., n-type or reduction
type).18–27 For these anolyte ROMs, remarkable performance enhancements in
terms of the solubility and multi-electron redox activity have been
achieved.23,26,28,29 Nevertheless, only a limited number of ROMs have been
reported as promising catholyte materials (i.e., p-type or oxidation type).8,30
Research on these p-type (oxidation type) ROMs is still in its infancy, and high-energy
organic RFBs can only be realized when anolytes are coupled with appropriate
catholytes.12,25,31,32 Moreover, p-type ROMs capable of multi-electron redox reac-
tions are even rarer because their oxidized states, such as dications, are often highly
unstable in a solution.13
Considering a realistic limit of the ROM concentration in non-aqueous media,15,33
exploiting multi-electron redox from a single p-type ROM is indispensable to
achieve high energy density at a given concentration.34 Several attempts have
very recently been made to utilize multi-electron redox p-type ROMs as
catholytes.35–37 For instance, Kowalski et al. demonstrated that the chemical stabil-
ity of the dication state of phenothiazine-based molecules can be enhanced by intro-
ducing methoxy groups at the para positions to the nitrogen atom of the phenothi-
azine core. Accordingly, the second redox reaction of the phenothiazine catholyte
can be utilized reversibly in a non-aqueous bulk-electrolysis cell.36 However, critical
challenges in multi-electron redox catholyte ROMs must be addressed, such as their
low solubility (< 0.1 M) and lack of long-term stability in the flow cell.36,37 To realize
high-energy-density RFBs, there is thus an urgent demand for the development of
new p-type ROMs capable of stable and reversible multi-electron redox that exhibit
high solubility.
In this report, we present a high-energy-density NORFB by exploiting a multi-elec-
tron redox p-type ROM as a catholyte; this ROM was rationally designed by
mimicking the energy transduction in living organisms. This newly synthesized phen-
azine-based molecule, 5,10-bis(2-methoxyethyl)-5,10-dihydrophenazine (BMEPZ),
undergoes double-redox reactions at high redox potentials with enhanced solubility
in non-aqueous media. Its redox mechanism is carefully investigated using spectro-
scopic tools combined with a computational method. Its multi-electron redox reac-
tions are revealed to be remarkably reversible in acetonitrile (MeCN) solution and
highly stable in all redox states. In addition, this molecule outperforms previously re-
ported ROMs in terms of its mass- and charge-transfer kinetics. A full-flow RFB, in
which BMEPZ is dissolved as a catholyte at concentrations of up to 0.4 M, is prepared
1Department of Materials Science and
Engineering, Research Institute of Advanced
Materials (RIAM), Seoul National University, 1
Gwanak Road, Seoul 151-742, Korea
and it exhibits stable cycle performance, delivering the highest energy densities
(ꢁ17 Wh Lꢀ1) among NORFBs reported thus far. We believe that our material design
can pave the way for the practical use of NORFB by reducing its performance gap
2Center for Supramolecular Optoelectronic
with the state-of-the-art vanadium-based RFBs and aqueous all-organic RFBs
Materials (CSOM), Department of Materials
(AORFBs).38–40
Science and Engineering, Research Institute of
Advanced Materials (RIAM), Seoul National
University, 1 Gwanak-ro, Gwanak-gu, Seoul
08826, Korea
RESULTS AND DISCUSSION
3These authors contributed equally
Bio-Inspired Redesign of Catholyte Material
4Lead Contact
In biosystems, common redox-active molecules are known to undergo multi-elec-
*Correspondence: mracle2@snu.ac.kr (J.E.K.),
tron redox reactions in the electron transport chains. Among such biological redox
matlgen1@snu.ac.kr (K.K.)
systems, methanophenazine is known to act as a redox mediator in a respiratory
https://doi.org/10.1016/j.chempr.2019.07.006
2
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
Figure 1. A New Multi-electron Redox Catholyte Material with Enhanced Solubility, BMEPZ
(A) The rational molecular design strategy of BMEPZ inspired by biological redox systems.
(B) Schematic illustration of all-organic full-flow battery containing BMEPZ/FL electrolytes and CV
curves of the redox couple at a scan rate of 100 mV sꢀ1
.
chain of archaea Methanosarcina mazei Go1.41 The methanophenazine can trans-
¨
port two electrons accompanying translocation of two protons through reversible
reduction reactions (see Figure 1A).42 The redox core structure of the methanophe-
nazine consists of a pyrazine core bearing two imine-like nitrogen (–N=) atoms. By
accepting two electrons with two protons, the pyrazine core is reduced to form a di-
hydrophenazine (see the core structures and details of the redox reactions in Fig-
ure S1A), offering a redox potential of approximately ꢀ165 mV versus NHE. Because
of its suitable potential as an anolyte, several derivatives possessing the pyrazine
core were recently utilized as anolytes for AORFBs coupled with a Fe(CN)6 catho-
lyte.43 Pyocyanin, a redox-active molecule involved in the electron-transfer reactions
of Pseudomonas aeruginosa bacteria,44 bears a similar core structure as methano-
phenazine but has a substituent in the core, delivering much higher redox potential
in its deprotonation of NADH (Figures 1A and S1B).45,46 Similar to the pyrazine core,
the methylpyrazinium moiety as a redox core also undergoes a reduction by two
electrons; however, only one proton is involved to form a methyl-hydropyrazine
Chem 5, 1–15, October 10, 2019
3

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
because the quaternized N can be reduced to a tertiary amine-like N without a pro-
ton. In contrast to the proton compensation for the pyrazine/dihydropyrazine
couple, this methylated-pyrazine-core redox couple (i.e., methylpyrazinium/
methyl-hydropyrazine) can bear a positive charge in its oxidized form.
Comparison of these two bio redox systems led us to speculate that the absence of
the charge-relief process resulting from methylation likely leads to a higher redox
potential (E1/2 = ꢀ40 mV versus NHE) of pyocyanin,46 and we anticipated that an
additional methylation on the other imine-like N of the pyrazinium core would further
increase the redox potential. Indeed, we observed that 5,10-dimethyldihydrophena-
zine (DMPZ) could undergo redox reactions at higher potentials (E1/2 = ꢀ0.26 and
0.50 V versus Fc/Fc+, ꢁ0.40 and 1.16 V versus NHE, see Figure S1C) between a di-
methylpyrazine dication/dimethylpyrazine redox core couple.37,47 Moreover,
because of the increased redox potential, in contrast to methanophenazine and pyo-
cyanin, DMPZ naturally exists in its reduced form (neutral form) and reversibly un-
dergoes two stepwise single-electron oxidations to form a stable dication (i.e.,
p-type redox) and is thus suitable for use as a catholyte. Nevertheless, the intrinsi-
cally low solubility of DMPZ (ꢁ60 mM in MeCN) has been a critical bottleneck for
its practical application in RFBs.37 Despite its high redox potential and multi-elec-
tron redox capability, the actual energy densities of RFBs employing DMPZ as a cath-
olyte are far from practical application.
To this end, we attempted to re-tailor the DMPZ molecules to improve their solubility
while maintaining the reversible and stable redox capability and succeeded in ob-
taining a new derivative, BMEPZ, that bears two flexible 2-methoxyethyl chains
instead of the methyl groups (see step 4 in Figure 1A). It was expected that the flex-
ibility and bulkiness of the alkyl ether chains would not only improve the solubility
but also enhance the stability of the dication form by spatially hindering the
redox-active N atoms from undergoing side reactions.48–50 Synthesis of BMEPZ
was performed starting with the reduction of phenazine using Na2S2O4, followed
by N-substitution of both amino groups with 2-chloroethylmethyl ether. The molec-
ular structure of the synthesized BMEPZ was characterized using nuclear magnetic
resonance (NMR) spectroscopy, high-resolution mass spectroscopy (HRMS), and
elemental analysis (see the Experimental Procedures and Figure S2 for the synthetic
details and reaction schemes). The newly synthesized BMEPZ exhibited enhanced
solubility in MeCN as high as 0.5 M, which is one order of magnitude greater than
that of the original DMPZ. To confirm that the substitution that improved the solubi-
lity did not reduce its original redox capability, we conducted cyclic voltammetry
(CV) measurements of BMEPZ in MeCN solution. As shown in Figure S3, BMEPZ dis-
played two reversible redox peaks at voltages of ꢀ0.29 and 0.50 V versus Fc/Fc+,
which are identical to those of DMPZ. The electrochemical properties of BMEPZ
will be discussed in further detail in the following sections.
To construct the all-organic RFB, the commercially available 9-fluorenone (FL) was
employed as a typical anolyte ROM; FL undergoes an n-type redox reaction at
ꢀ1.50 V versus Fc/Fc+.12 Because FL is highly soluble and undergoes one-electron
redox reaction with good reversibility in MeCN, we could avoid unnecessary
complexity in evaluating the redox activities of BMEPZ in the flow cell.12 Figure 1C
shows the redox activities of BMEPZ and FL in a mixed electrolyte along with the mo-
lecular structures for the corresponding redox states. In the CV measurement, the
mixed electrolyte showed three reversible redox peaks with redox potentials and
current levels identical to those of the corresponding redox peaks of each material
in the separate electrolytes (Figure S3). Thus, the redox behaviors of the two ROMs
4
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
Figure 2. Characterization of BMEPZ
(A) Representative charge-discharge curve for bulk-electrolysis test of BMEPZ.
(B) Coulombic efficiencies and capacities for bulk-electrolysis test of BMEPZ with respect to cycle number.
(C) Linearly fitted Levich plots of limiting current (iL) as a function of the square root of the rotation rate (uꢀ1/2).
(D) Linearly fitted plots of logarithm of kinetics-controlled current (log ik) as a function of overpotential (h).
(E) Comparison of kinetic parameters (kinetic rate constant (k0) and diffusion coefficient (D) of reported redox-active materials in RFB system: this work
(blue stars) and reported RFBs (pink circles).
were unaffected by each other in any redox states in the mixed electrolyte sys-
tem.12,51 According to the CV results, the all-organic RFB exploiting the BMEPZ/
FL redox couple is expected to offer the cell voltages of 1.21 and 2.00 V.
Characterization of BMEPZ Catholyte
We conducted bulk electrolysis and rotating disk electrode (RDE) experiments to
investigate the intrinsic electrochemical properties of the newly synthesized BMEPZ
molecule. Bulk electrolysis was appropriate to investigate the inherent redox ability
and stability of BMEPZ under convection without considering the anolyte in the
actual RFB system. As shown in Figure 2A, BMEPZ exhibits two distinct plateaus
on charge and discharge at potentials consistent with the observations from the
CV measurements. In addition, the measured capacity (1.573 mAh) was remarkably
close to the theoretical value (1.608 mAh), indicating that all the BMEPZ molecules
effectively contributed to the two-step reactions of each single-electron redox.
Moreover, highly stable cycling was observed without any noticeable capacity
fading, as shown in Figure 2B, implying that the multi-redox reactions of BMEPZ
are reversible. The RDE tests revealed the kinetic properties of the BMEPZ and
BMEPZ+ species with respect to the mass-diffusion and charge-transfer rates. In
the linear sweep voltammetry (LSV) curves, the mass-transfer-controlled limiting cur-
rents (iL) increased with increasing rotation rate (u) in the range of 300–1500 rpm
(Figures S4A and S4B). Figure 2C shows that the linear relationship between iL
and the square root of u was well-fitted by the Levich equation;52 the diffusion co-
efficients (D) of BMEPZ and BMEPZ+ were calculated to be 1.02 and 1.12 3 10ꢀ5
cm2 ꢀ1, respectively (a detailed description of the calculations is provided in the
s
Chem 5, 1–15, October 10, 2019
5

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
electrochemistry section of Experimental Procedures). The kinetic-controlled cur-
rents (ik) at different overpotentials (h) were also determined by extrapolation in
the Koutecky´–Levich plots (Figures S4C and S4D).52 The logarithms of the exchange
currents (log i0) were determined to be ꢀ3.65 for BMEPZ and ꢀ3.28 for BMEPZ+ in
the linearly fitted plots of log ik versus h (Figure 2D), yielding kinetic rate constants
(k0) of 1.18 3 10ꢀ2 and 2.77 3 10ꢀ2 cm sꢀ1, respectively. Notably, even though
BMEPZ is bulkier than DMPZ, the kinetics of BMEPZ were on par with or even faster
than those of DMPZ (D of 1.58 3 10ꢀ5 and 8.82 3 10ꢀ6 cm2 sꢀ1, respectively, and k0
of 2.97 3 10ꢀ2 and 5.53 3 10ꢀ3 cm sꢀ1, respectively, for DMPZ and DMPZ+).37 In Fig-
ure 2E, the kinetic parameters for BMEPZ, BMEPZ+, and FL (blue stars) are compared
with those of previously reported materials for RFBs (pink circle dots) on a logarith-
mic scale. It is noteworthy that the BMEPZ/FL redox couple exceeds the others in
terms of both mass- and charge-transfer kinetics, implying that the redox couple
would be favorable for a low-polarization and high-efficiency RFB. In addition, a
scan-rate dependent CV study was conducted to verify the reversibility of BMEPZ/
FL redox couple. As shown in Figure S5, the current ratio between cathodic and
anodic peaks of both BMEPZ and FL was almost unity in the CV curves, implying their
superior reversibility. The redox couple showed the peak separations ranging from
80 to 100 mV with respect to the scan-rate. Furthermore, the Randles-Sevcik
equation was used to determine their diffusion coefficients.53 The diffusion
coefficients of BMEPZ, BMEPZ+, and FL were calculated to be 8.86 3 10ꢀ6
,
1.16 3 10ꢀ5, and 1.05 3 10ꢀ5 cm2 sꢀ1, respectively, which are consistent with those
from the RDE study.
Raman spectroscopy and natural population analysis (NPA) were conducted to verify
the redox mechanism of the BMEPZ molecule. In Figure 3A, the reversible changes
of the local bonding of BMEPZ (pink areas) upon charging and discharging are
shown. Notably, the peak changes mainly corresponded to the vibrational modes
of C–N–C and C=C bonds in the reduced diazabutadiene motif (N–C=C–N), which
is consistent with the previous observation for DMPZ (Table S1).37,47 In addition, the
NPA enabled visualization of the changes of the charge distribution of BMEPZ during
the redox reaction. From the results for BMEPZ, BMEPZ+, and BMEPZ2+ (Figure S6),
the change of the charge was plotted for each atom as colors ranging from white to
pink, as shown in Figure 3B. The most drastic color changes were observed at the
nitrogen atoms, followed by the conjugated carbons in the benzene rings, implying
that the redox reaction of BMEPZ mainly occurs in the diazabutadiene motif. This
finding supports the idea that the flexible side chains substituted to elevate the sol-
ubility did not affect the general redox mechanisms.
To explore the origin of the chemical stability of the BMEPZ molecule at all redox
states in the given MeCN-based electrolyte, density functional theory (DFT) calcula-
tions were conducted for BMEPZ, BMEPZ+, BMEPZ2+, and MeCN. It was observed
that the energy levels of the lowest unoccupied molecular orbital (LUMO) of BMEPZ
and the singly occupied molecular orbital (SOMO) of the radical BMEPZ+ were
located above the highest occupied molecular orbital (HOMO) level of MeCN.
More importantly, even the LUMO level of BMEPZ2+ was higher than the HOMO
level of MeCN, indicating that the electron transfer from MeCN to BMEPZ in any
oxidation state during the battery operation is prohibited. Similarly, the electron
transfer from BMEPZ in any oxidation state to MeCN is energetically unfavorable
because the HOMO (or SOMO) levels of BMEPZ and BMEPZ2+ (or BMEPZ+) are lower
than the LUMO level of MeCN. The above theoretical calculations indicate that para-
sitic side reactions based on electron transfer between BMEPZ and MeCN are less
likely to occur during battery operation (Figure S7).
6
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
Figure 3. Investigation of Redox Mechanism of BMEPZ
(A) Ex situ Raman analysis for BMEPZ in catholyte, demonstrating the reversible appearance and
disappearance of several peaks.
(B) NPA of BMEPZ, BMEPZ+, and BMEPZ2+. The depth of the pink color denotes the extent of the
charge change.
(C) Molecular geometries of BMEPZ, BMEPZ+, and BMEPZ2+
.
In addition, the structural changes in the optimized geometries of BMEPZ during the
redox reactions were observed. As shown in Figure 3C, the neutral BMEPZ has bent
conformation with an angle of 144.2ꢂ between the two phenyl ring planes. The
radical cation and dication, in contrast, show almost planar geometries with the an-
gles of 171.2ꢂ and 175.4ꢂ, respectively. It can be speculated that the planar molec-
ular geometries of the radical cation and dication effectively delocalized the addi-
tional charges, leading to the remarkable redox stability of BMEPZ.54,55 It is
noteworthy that many p-type ROMs except BMEPZ tend to be very unstable when
they are oxidized by two electrons, resulting in fast decay of the cycle
performance.13,31
Electrochemical Performance of BMEPZ/FL Flow Cell
Cycling tests of flow cells employing BMEPZ as the catholyte and FL as the anolyte
were performed using a custom flow cell (see the Experimental Procedures section
for further details) to investigate the electrochemical performance in a near-practical
system. Figure 4A presents a representative charge-discharge curve for the dilute
condition of 0.05 M BMEPZ and 0.1 M FL in a supporting electrolyte of 0.5 M bis(tri-
fluoromethane)sulfonimide lithium salt (LiTFSI) in MeCN at a current density of 20
mA cmꢀ2. The flow cell exhibited two well-defined plateaus at cell voltages of 1.2
and 2.0 V, which is consistent with the expectations from the CV curves in Figure 1B.
The plateaus of the same length at each voltage indicate that the two single-electron
Chem 5, 1–15, October 10, 2019
7

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
Figure 4. Electrochemical Performance of Flow Cell of 50 mM BMEPZ and 0.1 M FL
(A) Representative cell voltage versus time profile at a current density of 20 mA cmꢀ2. The inset
shows the color change of the catholyte as a function of the SOC.
(B) Cycling efficiencies and capacities with respect to cycle number.
redox reactions of BMEPZ contribute equally to the cell capacity in the flow cell sys-
tem. Additionally, we observed the color changes of the catholyte during redox re-
actions, which are presented as a function of the state of charge (SOC) with a step
size of SOC 25 in Figure 4A. The color changes roughly occurred in two distinct
stages, from yellow to green and from green to red, because of the double-redox
reactions of BMEPZ, which could be quantitatively probed using UV-vis spectros-
copy (Figure S8). The cycling efficiencies and capacity retention of the flow cell
are plotted in Figure 4B. The cycling of the flow cell has a Coulombic efficiency
of ꢁ96% with an energy efficiency of ꢁ70%, which were stably maintained over
200 cycles. Moreover, an initial discharge capacity of 2.08 Ah Lꢀ1 was achieved,
which is close to the theoretical capacity of 2.67 Ah Lꢀ1 and yields a material utiliza-
tion of ꢁ80%. Note that the initial capacity could be retained over 200 cycles (cycling
time of ꢁ129 h) with a capacity retention of 99.94% per cycle and 70% of theoretical
capacity could be achieved after 200 cycles (Figure S9). We attribute this highly
robust cycle performance of the cell to the high stability of BMEPZ+ radical cation
in the electrolyte as well as the stable dication form of BMEPZ. Typically, organic rad-
icals are often highly reactive and thus unstable, which is the key obstacle for
improving the cycle performance in an organic flow battery. In contrast, it is note-
worthy that the UV-vis spectra of BMEPZ+ in the electrolyte were virtually unchanged
over 24 h (Figure S10), indicating that BMEPZ+ has superior radical stability.12,56
We next performed a demonstration of the flow cell under the near-saturation con-
dition of BMEPZ to investigate the electrochemical behavior under practical cell con-
ditions. In Figure 5A, additional charge-discharge curves for flow cells using 0.1 M
and 0.4 M BMEPZ are presented. A proportional elevation of the capacity was
observed with respect to the concentration, and stable cycling of 99.94% per cycle
was achieved for the 0.1 M BMEPZ flow cell in Figure 5B. In addition, the 0.4 M
BMEPZ flow cell was successfully demonstrated with a material utilization of ꢁ75%
during discharge and a respectable capacity retention of 99.3% per cycle in
Figure 5C. Despite the superior stability of BMEPZ at all redox states, the capacity
decay in concentrated cell was non-negligible. Since longer cycling time was
needed for the concentrated cell at the same current density, more ROM molecules
would cross over through the microporous separator, represented by the low
Coulombic efficiency of 83%. Moreover, during the extended cycling time, the FL
radical anions at the charged state might have more chance to decay because of
their rather short lifetime.12 It can thus be expected that the development of
membrane which can suppress the cross-over issue and using more stable anolyte
material would improve cycle life in the concentrated cell. For comparison of the
8
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
Figure 5. Electrochemical Performance of BMEPZ/FL Flow Cell at High Concentrations
(A) Typical charge-discharge curves at different concentrations.
(B) Cycling efficiencies and capacities for 0.1 M BMEPZ and 0.2 M FL with respect to cycle number.
(C) Cycling efficiencies and capacities for 0.4 M BMEPZ and 0.8 M FL with respect to cycle number.
(D) Energy density plot of typical redox-active materials in organic RFBs. The molar energy density is the volumetric energy density divided by the
cycling concentration.
performance of the BMEPZ/FL flow cell with that of previously reported all-organic
RFBs involving both aqueous and non-aqueous media, the theoretical energy den-
sities of the RFBs at the cycling concentration are plotted in Figure 5D. The energy
densities of the RFBs are presented with respect to the intrinsic activity of the redox
couple (energy per mole) and its concentration in the solvent (energy per volume).
With regard to the dynamic properties of solvents and cost issue, the molar energy
density should be considered because ROMs with high molar energy densities are
advantageous to deliver high energy densities using relatively low concentrations
of ROMs. The figure shows that AORFBs (yellow areas) generally exhibit low intrinsic
energy densities limited by the low cell voltage (<1.4 V); therefore, a high-concen-
tration electrolyte is required to achieve a high volumetric energy density. For
instance, Janoschka et al. demonstrated the high volumetric energy density
(38 Wh Lꢀ1), which is comparable to the vanadium RFBs, but it needed 2.0 M solu-
tions of TEMPO derivative and methyl viologen.40 In contrast, a relatively high molar
energy density can be attained for NORFBs (purple areas) because of the wide elec-
trochemical window. Furthermore, NORFBs using multi-redox ROMs present a
promising pathway to realize high volumetric energy density with exceptionally
high molar energy densities, as illustrated in blue areas. However, in practice, their
Chem 5, 1–15, October 10, 2019
9

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
volumetric energy density rarely exceeds that of aqueous RFBs because of the
relatively low cycling concentration limited by the dynamic properties of non-
aqueous media and the lack of suitable membrane. As the multi-redox capability
is combined with enhanced solubility in the BMEPZ/FL system, the volumetric en-
ergy density (ꢁ17 Wh Lꢀ1) exceeded that of other NORFBs even at a lower concen-
tration (ꢁ0.4 M), showing a great potential of multi-redox NORFB systems. It can be
expected that further engineering to achieve solubility of phenazine derivatives over
1 M and the development of high-performance membrane for the non-aqueous sys-
tem would enable the phenazine-based NORFB to outperform vanadium-based and
aqueous organic systems.
Conclusions
We reported a multi-redox BMEPZ inspired by biosystems as a promising catholyte ma-
terial with the highest energy density demonstrated for organic RFBs. This ROM un-
dergoes two reversible redox reactions at high redox potentials of –0.29 and 0.50 V
versus Fc/Fc+ and exhibits outstanding electrochemical kinetics for the mass- and
charge-transfer processes, retaining high solubility in non-aqueous media. An all-
organic flow cell exploiting the BMEPZ/FL redox couple was successfully prepared
with two well-defined voltage plateaus at 1.2 and 2 V and highly robust cycling owing
to the remarkable chemical stability, delivering one of the highest energy density of re-
ported NORFBs. Although further engineering of ROMs and the development of high-
performance membrane are still needed for practical application of this system, these
findings on a new multi-electron redox material combined with the high solubility pro-
vide a breakthrough in the realization of high-energy-density RFBs.
EXPERIMENTAL PROCEDURES
Preparation of Materials
MeCN (anhydrous) and FL were purchased from Sigma–Aldrich and used without
further purification. The microporous separator (Celgard 4560) and LiTFSI were pur-
chased from Wellcos (Korea) and TCI Chemicals (Japan) and were treated under vac-
uum at 70ꢂC for 24 h to remove moisture. Tetrahydrofuran (THF) was distilled using
sodium and benzophenone before use. All the other reagents and solvents were
obtained from Sigma-Aldrich and Alfa Aesar Co. and used as received without
further purification. All the glassware, syringes, magnetic stirring bars, and needles
were thoroughly dried in a convection oven. The reactions were monitored using
thin layer chromatography (TLC) using commercial TLC plates (silica gel 60 F254,
Merck Co.).
Synthesis
Synthesis of 5,10-dihydrophenazine (1)
The 5,10-dihydrophenazine (1) was prepared following reported procedures.57
Phenazine (3 g, 16.6 mmol) was dissolved in ethanol (75 mL), and the solution was
heated to boiling. An aqueous solution (300 mL) containing Na2S2O4 (30 g, 0.17
mol) was added to the boiling solution. The solution turned purple immediately after
mixing; then, a greenish white precipitate formed. The precipitated solid was
collected by filtration, washed with water, and dried in vacuo to afford 2.7 g of a
greenish white solid. Because of its instability in air, the solid was stored under nitro-
gen without further purification and characterized.
Synthesis of 5,10-bis(2-methoxyethyl)-5,10-dihydrophenazine (2, BMEPZ)
In a flame-dried two-necked 100-mL round bottom flask, 1 (1 g, 5.5 mmol) was dis-
solved in freshly distilled THF (12 mL). To this solution, 1.6 M n-butyl lithium (n-BuLi)
in hexane (9.5 mL, 15.4 mmol) was added dropwise over 30 min at room
10
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
temperature. After an orange precipitate was formed, 2-chloroethyl methyl ether
(3.5 mL, 38.5 mmol) was added. Subsequently, the orange precipitate disappeared,
forming a reddish-brown solution. The solution was stirred overnight, and the reac-
tion mixture was then poured into water and extracted three times with ethyl acetate.
The organic layer was dried over anhydrous MgSO4 and purified using column chro-
matography on neutral alumina with ethyl acetate/n-hexane (1:19 v/v) as an eluent to
give a light-yellow powder. Yield: 78.2% (1.28 g), 1H-NMR (300 MHz, C6D6) d(ppm):
6.60–6.57 (m, 4H), 6.31–6.28 (m, 4H), 3.43 (t, J = 6.3 Hz, 4H), 3.25 (t, J = 6.3 Hz, 4H),
2.99 (s, 6H). 13C-NMR (125 MHz, C6D6) d(ppm): 137.75, 111.84, 69.15, 59.03, 46.35.
Anal. calc. for C18H22N2O2: C, 72.46; H, 7.43; N, 9.39; found: C, 72.42; H, 7.43; N,
9.38; HRMS (EI+): calc. for
(Figure S11).
C
18H22N2O2 (M+), 298.1681; found, 298.1684
Characterization
1H-NMR and 13C-NMR spectra were recorded on a Bruker Avance-300 and Avance-
500 NMR spectrometer, respectively. HRMS were obtained using a JEOL JMS-700
instrument. Elemental analysis was conducted using a Thermo Fisher Scientific Flash
2000 elemental analyzer.
In the solubility measurement, 30.2 mg of BMEPZ powder was fully dissolved in 2 mL
MeCN to make a 0.05 M solution first. Then, to the solution, the same amount
(30.2 mg) of BMEPZ powder was repeatedly added until it got saturated without
further dissolution.
For the electrochemical measurements, the solutions were prepared and evaluated
in an Ar-filled glove box under an inert atmosphere (<0.5 ppm O2, H2O). CV curves
of DMPZ, BMEPZ, FL, and the mixture of BMEPZ and FL (10 mM each) were obtained
using 0.1 M LiTFSI in MeCN as a supporting electrolyte. A three-electrode system
(a Pt counter electrode, a Ag/AgNO3 reference electrode, and Au working elec-
trode) was employed, and a scan rate of 100 mV sꢀ1 was used. Ferrocene (5 mM)
was used as an internal reference.
For the bulk-electrolysis test, a commercially available bulk-electrolysis cell
(MF-1056) was used. The cycling experiment used an electrolyte consisting of
1 mM BEMPZ and 0.5 M LiTFSI in MeCN (30 mL) and was stirred at 1,400 rpm. A
three-electrode system (a Pt counter electrode, Ag/AgNO3 reference electrode,
and reticulated vitreous carbon working electrode) was employed at a current of
1.608 mA (1C).
For the RDE test, a three-electrode system (a Pt counter electrode, Ag/AgNO3 refer-
ence electrode, and glassy carbon working electrode with 5-mm diameter) was em-
ployed for the LSV tests. The BMEPZ+ solution was prepared by oxidizing the BMEPZ
solution in the bulk-electrolysis experiment under a cut-off voltage of ꢀ0.11 V versus
Fc/Fc+. The working electrode was rotated from 300 to 1,500 rpm (in increments of
300 rpm) using a modulated speed rotator (AFMSRX; PINE). The LSV tests were con-
ducted with 1.0 mM ROMs in 0.5 M LiTFSI in MeCN at a scan rate of 5 mV sꢀ1. The
kinematic viscosity (y) of the 0.5 M LiTFSI in MeCN was measured to be 0.59 mm2 sꢀ1
following the standard test method ASTM D445 in Korea Polymer Testing &
Research Institute (Koptri, Korea). Using the slopes of the linearly fitted Levich plots
(Figure 2C) and the Levich equation (Equation 1), the diffusion coefficients (D) of the
ROMs were calculated (the slopes of BMEPZ and BMEPZ+ were determined to be
1.30 3 10ꢀ5 and 1.38 3 10ꢀ5 A radꢀ1/2 s1/2, respectively). Using the Koutecky´–Levich
equation (Equation 2), Koutecky´–Levich plots (Figures S4C and S4D) were obtained
Chem 5, 1–15, October 10, 2019
11

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
at different overpotentials, and the kinetic rate constants of the ROMs were calcu-
lated using Equation 3.
iL = 0:62nFAD2=3u1=2
y
ꢀ1=6C0
(Equation 1)
(Equation 2)
(Equation 3)
1
1
1
=
+
i
ik 0:62nFAD2=3u1=2
y
ꢀ1=6C0
i0 = nFAk0C0
Here, n is the number of electrons transferred (n = 1), F is the Faraday constant (F =
96; 485 C molꢀ1), A is the electrode area (A = 0:2 cm2), and C0 is the concentration of
the ROM ðC0 = 1:0 mMÞ.
For the scan-rate dependence CV study, CV curves of BMEPZ and FL (10 mM each)
were obtained using 0.5 M LiTFSI in MeCN as a supporting electrolyte at the
following scan-rates: 25, 50, 100, 200, and 300 mV sꢀ1. Randles-sevcik equation
(Equation 4) was used to calculate the diffusion coefficients of ROM.
1
2
ꢀ
ꢁ
nFsD
RT
ip = 0:4463nFAC
(Equation 4)
Here, A is the electrode area ðA = 7:07 mm2Þ, C is the concentration of the ROM
ðC = 10 mMÞ, and s is the scan-rate, respectively.
Raman spectra (LabRAM HR Evolution, Horiba) were recorded using capillary tubes
(inner diameter of 1.1–1.2 mm) with continuous-wave linearly polarized lasers (wave-
length: 532 nm). The laser beam was focused using a 503 objective lens, resulting in
a spot diameter of approximately 1 mm. The acquisition time and number of accumu-
lations were 10 s and 5, respectively.
For the UV-vis spectroscopy analysis, absorption spectra of the electrolytes in the
positive electrode compartment (diluted in MeCN, 5% v/v) were obtained using a
UV-vis spectrometer (Agilent Technologies, Cary 5000) with an optical glass cuvette
(Quartz; Hellma). Electrolytes containing 10 mM BMEPZ and 20 mM FL in the sup-
porting electrolyte of 0.5 M LiTFSI in MeCN were used.
Computational Details
Spin-unrestricted density functional theory (DFT)-type calculations were performed
using the Gaussian 09 package,58 including geometry optimization, energy evalua-
tion, eigenvalue calculation, NPA calculation, and vibration property prediction for
each molecule. The Becke–Lee–Yang–Parr (B3LYP) hybrid exchange-correlation
functional59–61 and triple-zeta valence polarization (TZVP) basis set62,63 were used
for the entire calculation, which have been demonstrated to reproduce experimental
results well.37,47,64,65 To model the solvation environment of acetonitrile (dielectric
constant = 38.8), the polarizable continuum model (PCM) scheme, implicit solvation
model, was introduced for the entire calculation. The LUMO and SOMO denote the
adiabatic LUMO and SOMO.
Flow Cell Test
A custom flow cell with backing plates (polyethylene-coated fiber glass), flow fields
(polytetrafluoroethylene (PTFE)), and gaskets (PTFE) was fabricated using materials
purchased from ILDO F&C (Korea). The flow cells were assembled with carbon felt
(XF30A; TOYOBO, Korea) as electrodes at both the anode and cathode side with
four pieces of microporous separators (Celgard 4560) sandwiched in between.
12
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
The active size of the flow cell was 2.0-cm wide 3 2.0-cm long (4 cm2), and a nor-
prene tubing (Masterflex) was used. The mixed electrolytes (13 mL in each half-cell
side) containing both BMEPZ and FL with the supporting electrolyte of 0.5 M LiTFSI
in MeCN were flowed through the felt electrodes at a flow rate of 80 mL minꢀ1 using
a pump (ShenChen). In the case of 0.4 M BMEPZ cell test, 0.8 M FL and 1.0 M LiTFSI
were used as the anolyte and the supporting electrolyte, respectively. Flow cell tests
were performed at a current density of 20 mA cmꢀ2 in constant-current mode using a
battery test system (WBCS 3000; WonA Tech, Korea), and free 20 cycles (5 cycles for
0.4 M BMEPZ and 0.8 M FL cell test) were conducted in advance for initial equilibra-
tion of the flow setup. The volumetric energy density was calculated based on the
entire electrolyte volume (26 mL).
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.chempr.
2019.07.006.
ACKNOWLEDGMENTS
We especially thank Dr. Minah Lee (Korea Institute of Science and Technology)
and Dr. Jihyun Hong (Korea Institute of Science and Technology) for providing inspi-
ration of using biomaterials as energy storage materials. This work was supported by
the National Research Foundation of Korea (NRF) grant funded by the Korean
Government (MSIP) (2015R1A2A1A10055991), Lotte Chemical, Creative Materials
Discovery Program through the National Research Foundation of Korea
(NRF) funded by the Ministry of Science, ICT and Future Planning (NRF-
2017M3D1A1039553), and Presidential Post-Doc. Fellowship Program through
the National Research Foundation (NRF) of Korea funded by the Ministry of Educa-
tion [2016R1A6A3A04008134(RIAM0417-20180002)].
AUTHOR CONTRIBUTIONS
G.K. designed the experiments with help from M.H.L., S.L., S.-K.J., K.K., and J.K. K.L.
performed the organic synthesis with help from S.Y.P. The computations were per-
formed by B.L. K.K. and J.E.K supervised the overall research. All authors discussed
the experiments and final manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: March 29, 2019
Revised: June 14, 2019
Accepted: July 12, 2019
Published: August 19, 2019
REFERENCES
1. Yang, Z., Zhang, J., Kintner-Meyer, M.C.W., Lu,
X., Choi, D., Lemmon, J.P., and Liu, J. (2011).
Electrochemical energy storage for green grid.
Chem. Rev. 111, 3577–3613.
4. Wang, Y., He, P., and Zhou, H. (2012). Li-
redox flow batteries based on hybrid
electrolytes: at the cross road between Li-ion
and redox flow batteries. Adv. Energy Mater. 2,
770–779.
6. Nguyen, T., and Savinell, R.F. (2010). Flow
batteries. Electrochem. Soc. Interface 19,
54–56.
7. Noack, J., Roznyatovskaya, N., Herr, T., and
Fischer, P. (2015). The chemistry of redox-flow
batteries. Angew. Chem. Int. Ed. Engl. 54,
9776–9809.
2. Dunn, B., Kamath, H., and Tarascon, J.M.
(2011). Electrical energy storage for the grid:
a battery of choices. Science 334, 928–935.
5. Darling, R.M., Gallagher, K.G., Kowalski, J.A.,
Ha, S., and Brushett, F.R. (2014). Pathways to
low-cost electrochemical energy storage: a
comparison of aqueous and nonaqueous
flow batteries. Energy Environ. Sci. 7, 3459–
3477.
8. Sevov, C.S., Samaroo, S.K., and Sanford, M.S.
(2017). Cyclopropenium salts as cyclable, high-
potential catholytes in nonaqueous media.
Adv. Energy Mater. 7.
3. Soloveichik, G.L. (2015). Flow batteries: current
status and trends. Chem. Rev. 115, 11533–
11558.
Chem 5, 1–15, October 10, 2019
13

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
9. Winsberg, J., Hagemann, T., Janoschka, T.,
Hager, M.D., and Schubert, U.S. (2017). Redox-
flow batteries: from metals to organic redox-
active materials. Angew. Chem. Int. Ed. Engl.
56, 686–711.
22. Lee, S., Kwon, G., Ku, K., Yoon, K., Jung, S.K.,
Lim, H.D., and Kang, K. (2018). Recent progress
in organic electrodes for Li and Na
C., et al. (2017). A stable two-electron-donating
phenothiazine for application in nonaqueous
redox flow batteries. J. Mater. Chem. A 5,
24371–24379.
rechargeable batteries. Adv. Mater. 30.
37. Kwon, G., Lee, S., Hwang, J., Shim, H.-S., Lee,
B., Lee, M.H., Ko, Y., Jung, S.-K., Ku, K., Hong,
J., et al. (2018). Multi-redox molecule for high-
energy redox flow batteries. Joule 2, 1771–
1782.
23. Hu, B., and Liu, T.L. (2018). Two electron
utilization of methyl viologen anolyte in
nonaqueous organic redox flow battery.
J. Energy Chem. 27, 1326–1332.
10. Wei, X., Pan, W., Duan, W., Hollas, A., Yang, Z.,
Li, B., Nie, Z., Liu, J., Reed, D., Wang, W., et al.
(2017). Materials and systems for organic redox
flow batteries: status and challenges. ACS
Energy Lett. 2, 2187–2204.
24. Lin, K., Chen, Q., Gerhardt, M.R., Tong, L., Kim,
S.B., Eisenach, L., Valle, A.W., Hardee, D.,
38. Li, L., Kim, S., Wang, W., Vijayakumar, M., Nie,
Z., Chen, B., Zhang, J., Xia, G., Hu, J., Graff, G.,
et al. (2011). A stable vanadium redox-flow
battery with high energy density for large-scale
energy storage. Adv. Energy Mater. 1, 394–400.
Gordon, R.G., Aziz, M.J., et al. (2015). Alkaline
quinone flow battery. Science 349, 1529–1532.
11. Park, M., Ryu, J., Wang, W., and Cho, J. (2017).
Material design and engineering of next-
generation flow-battery technologies. Nat.
Rev. Mater. 2, 16080.
25. Liu, T., Wei, X., Nie, Z., Sprenkle, V., and Wang,
W. (2016). A total organic aqueous redox flow
battery employing a low cost and sustainable
methyl viologen anolyte and 4-HO-TEMPO
catholyte. Adv. Energy Mater. 6.
39. Kear, G., Shah, A.A., and Walsh, F.C. (2012).
Development of the all-vanadium redox flow
battery for energy storage: a review of
technological, financial and policy aspects. Int.
J. Energy Res. 36, 1105–1120.
12. Wei, X., Xu, W., Huang, J., Zhang, L., Walter, E.,
Lawrence, C., Vijayakumar, M., Henderson,
W.A., Liu, T., Cosimbescu, L., et al. (2015).
Radical compatibility with nonaqueous
electrolytes and its impact on an all-organic
redox flow battery. Angew. Chem. Int. Ed. Engl.
54, 8684–8687.
26. Luo, J., Hu, B., Debruler, C., and Liu, T.L. (2018).
A p-conjugation extended viologen as a two-
electron storage anolyte for total organic
aqueous redox flow batteries. Angew. Chem.
Int. Ed. Engl. 57, 231–235.
40. Janoschka, T., Martin, N., Hager, M.D., and
Schubert, U.S. (2016). An aqueous redox-flow
battery with high capacity and power: the
TEMPTMA/MV system. Angew. Chem. Int. Ed.
Engl. 55, 14427–14430.
13. Milshtein, J.D., Kaur, A.P., Casselman, M.D.,
Kowalski, J.A., Modekrutti, S., Zhang, P.L.,
Harsha Attanayake, N.H., Elliott, C.F., Parkin,
S.R., Risko, C., et al. (2016). High current
density, long duration cycling of soluble
organic active species for non-aqueous redox
flow batteries. Energy Environ. Sci. 9, 3531–
3543.
27. Yang, Z., Tong, L., Tabor, D.P., Beh, E.S.,
Goulet, M.A., De Porcellinis, D., Aspuru-Guzik,
A., Gordon, R.G., and Aziz, M.J. (2018). Alkaline
benzoquinone aqueous flow battery for large-
scale storage of electrical energy. Adv. Energy
Mater. 8.
41. Beifuss, U., Tietze, M., Baumer, S., and
¨
Deppenmeier, U. (2000). Methanophenazine:
structure, total synthesis, and function of a new
cofactor from methanogenic archaea. This
work was supported by the Deutsche
Forschungsgemeinschaft
(Sonderforschungsbereich 416; grants De 488/
6-1 and De 488/4-2) and the Fonds der
Chemischen Industrie. We are grateful to Drs.
J. Paust and H. Jaedicke (BASF AG,
28. Huskinson, B., Marshak, M.P., Suh, C., Er, S.,
Gerhardt, M.R., Galvin, C.J., Chen, X., Aspuru-
Guzik, A., Gordon, R.G., and Aziz, M.J. (2014). A
metal-free organic–inorganic aqueous flow
battery. Nature 505, 195–198.
14. Ding, Y., Zhao, Y., Li, Y., Goodenough, J.B., and
Yu, G. (2017). A high-performance all-
metallocene-based, non-aqueous redox flow
battery. Energy Environ. Sci. 10, 491–497.
Ludwigshafen) and Dr. R. K. Muller (Hoffmann-
¨
15. Duan, W., Huang, J., Kowalski, J.A., Shkrob,
I.A., Vijayakumar, M., Walter, E., Pan, B., Yang,
Z., Milshtein, J.D., and Li, B. (2017). Wine-dark
sea in an organic flow battery: storing negative
charge in 2,1,3-benzothiadiazole radicals leads
to improved cyclability. ACS Energy Lett. 2,
1156–1161.
29. Kwabi, D.G., Lin, K., Ji, Y., Kerr, E.F., Goulet,
M.-A., De Porcellinis, D., Tabor, D.P., Pollack,
D.A., Aspuru-Guzik, A., Gordon, R.G., et al.
(2018). Alkaline quinone flow battery with long
lifetime at pH 12. Joule 2, 1894–1906.
La Roche Ltd., Basel) for generously providing
chemicals. Angew. Chem. Int. Ed. Engl. 39,
2470–2472.
42. Pierson, L.S., and Pierson, E.A. (2010).
Metabolism and function of phenazines in
bacteria: impacts on the behavior of bacteria in
the environment and biotechnological
processes. Appl. Microbiol. Biotechnol. 86,
1659–1670.
30. Zhang, C., Zhang, L., Ding, Y., Peng, S., Guo, X.,
Zhao, Y., He, G., and Yu, G. (2018). Progress
and prospects of next-generation redox flow
batteries. Energy Storage Mater. 15, 324–350.
16. Ding, Y., Li, Y., and Yu, G. (2016). Exploring bio-
inspired quinone-based organic redox flow
batteries: a combined experimental and
computational study. Chem 1, 790–801.
31. Wei, X., Duan, W., Huang, J., Zhang, L., Li, B.,
Reed, D., Xu, W., Sprenkle, V., and Wang, W.
(2016). A high-current, stable nonaqueous
organic redox flow battery. ACS Energy Lett. 1,
705–711.
43. Hollas, A., Wei, X., Murugesan, V., Nie, Z., Li, B.,
Reed, D., Liu, J., Sprenkle, V., and Wang, W.
(2018). A biomimetic high-capacity phenazine-
based anolyte for aqueous organic redox flow
batteries. Nat. Energy 3, 508–514.
17. Chen, H., Zhou, Y., and Lu, Y.-C. (2018). Lithium-
organic nanocomposite suspension for high-
energy-density redox flow batteries. ACS
Energy Lett. 3, 1991–1997.
32. Wei, X., Xu, W., Vijayakumar, M., Cosimbescu,
L., Liu, T., Sprenkle, V., and Wang, W. (2014).
TEMPO-based catholyte for high-energy
density nonaqueous redox flow batteries. Adv.
Mater. 26, 7649–7653.
44. Rada, B., and Leto, T.L. (2013). Pyocyanin
effects on respiratory epithelium: relevance in
Pseudomonas aeruginosa airway infections.
Trends Microbiol. 21, 73–81.
18. Zhao, Q., Zhu, Z., and Chen, J. (2017).
Molecular engineering with organic carbonyl
electrode materials for advanced stationary
and redox flow rechargeable batteries. Adv.
Mater. 29.
45. Tietze, M., Beuchle, A., Lamla, I., Orth, N.,
Dehler, M., Greiner, G., and Beifuss, U. (2003).
Redox potentials of methanophenazine and
CoB-S-S-CoM, factors involved in electron
transport in methanogenic archaea.
33. Zhang, J., Corman, R.E., Schuh, J.K., Ewoldt,
R.H., Shkrob, I.A., and Zhang, L. (2018). Solution
properties and practical limits of concentrated
electrolytes for nonaqueous redox flow
19. Kang, N., Park, J.H., Choi, J., Jin, J., Chun, J.,
Jung, I.G., Jeong, J., Park, J.G., Lee, S.M., Kim,
H.J., et al. (2012). Nanoparticulate iron oxide
tubes from microporous organic nanotubes as
stable anode materials for lithium ion batteries.
Angew. Chem. Int. Ed. Engl. 51, 6626–6630.
batteries. J. Phys. Chem. C 122, 8159–8172.
ChemBioChem 4, 333–335.
34. Li, Z., and Lu, Y.-C. (2018). Redox flow batteries:
want more electrons? Go organic! Chem 4,
2020–2021.
46. Wang, Y., and Newman, D.K. (2008). Redox
reactions of phenazine antibiotics with ferric
(hydr) oxides and molecular oxygen. Environ.
Sci. Technol. 42, 2380–2386.
20. Benniston, A.C., Hagon, J., He, X., Yang, S., and
Harrington, R.W. (2012). Spring open two-plus-
two electron storage in a disulfide-strapped
methyl viologen derivative. Org. Lett. 14,
506–509.
35. Kim, H.-s., Lee, K.-J., Han, Y.-K., Ryu, J.H., and
Oh, S.M. (2017). A comparative study on the
solubility and stability of p-phenylenediamine-
based organic redox couples for non-aqueous
flow batteries. J. Power Sources 348, 264–269.
47. Lee, M., Hong, J., Lee, B., Ku, K., Lee, S., Park,
C.B., and Kang, K. (2017). Multi-electron redox
phenazine for ready-to-charge organic
batteries. Green Chem. 19, 2980–2985.
21. Song, Z., and Zhou, H. (2013). Towards
sustainable and versatile energy storage
devices: an overview of organic electrode
materials. Energy Environ. Sci. 6, 2280–2301.
36. Kowalski, J.A., Casselman, M.D., Kaur, A.P.,
Milshtein, J.D., Elliott, C.F., Modekrutti, S.,
Attanayake, N.H., Zhang, N., Parkin, S.R., Risko,
48. Zhang, J., Yang, Z., Shkrob, I.A., Assary, R.S.,
Tung, S.o., Silcox, B., Duan, W., Zhang, J., Su,
C.C., Hu, B., et al. (2017). Annulated
14
Chem 5, 1–15, October 10, 2019

Please cite this article in press as: Kwon et al., Bio-inspired Molecular Redesign of a Multi-redox Catholyte for High-Energy Non-aqueous
Organic Redox Flow Batteries, Chem (2019), https://doi.org/10.1016/j.chempr.2019.07.006
dialkoxybenzenes as catholyte materials for
non-aqueous redox flow batteries: achieving
high chemical stability through bicyclic
substitution. Adv. Energy Mater. 7.
donor units in dye-sensitized solar cells. Chem.
Commun. (Camb.) 53, 2367–2370.
density. Phys. Rev. B Condens. Matter 37,
785–789.
60. Becke, A.D. (1993). Density-functional
thermochemistry. III. The role of exact
exchange. J. Chem. Phys. 98, 5648–5652.
55. Narayana, K.A., Casselman, M.D., Elliott, C.F.,
Ergun, S., Parkin, S.R., Risko, C., and Odom,
S.A. (2015). N-substituted phenothiazine
derivatives: how the stability of the neutral
and radical cation forms affects overcharge
performance in lithium-ion batteries.
49. Kowalski, J.A., Su, L., Milshtein, J.D., and
Brushett, F.R. (2016). Recent advances in
molecular engineering of redox active organic
molecules for nonaqueous flow batteries. Curr.
Opin. Chem. Eng. 13, 45–52.
61. Stephens, P.J., Devlin, F.J., Chabalowski, C.F.,
and Frisch, M.J. (1994). Ab initio calculation of
vibrational absorption and circular dichroism
spectra using density functional force fields.
J. Phys. Chem. 98, 11623–11627.
ChemPhysChem 16, 1179–1189.
50. Wang, W., Xu, W., Cosimbescu, L., Choi, D., Li,
L., and Yang, Z. (2012). Anthraquinone with
tailored structure for a nonaqueous metal–
organic redox flow battery. Chem. Commun.
(Camb.) 48, 6669–6671.
56. Armstrong, C.G., and Toghill, K.E. (2018).
Stability of molecular radicals in organic non-
aqueous redox flow batteries: a mini review.
Electrochem. Commun. 91, 19–24.
62. Schafer, A., Huber, C., and Ahlrichs, R. (1994).
¨
Fully optimized contracted Gaussian basis sets
of triple zeta valence quality for atoms Li to Kr.
J. Chem. Phys. 100, 5829–5835.
57. Sugimoto, A., Kotani, T., Tsujimoto, J., and
Yoneda, S. (1989). Preparation and
properties of electron donor acceptor
complexes of the compounds having capto-
dative substituents. J. Heterocycl. Chem. 26,
435–438.
51. Wei, X., Li, L., Luo, Q., Nie, Z., Wang, W., Li, B.,
Xia, G.-G., Miller, E., Chambers, J., and Yang, Z.
(2012). Microporous separators for Fe/V
redox flow batteries. J. Power Sources 218,
39–45.
63. Klaumunzer, B., Kroner, D., and Saalfrank, P.
¨
¨
(2010). (TD-)DFT calculation of vibrational and
vibronic spectra of riboflavin in solution.
J. Phys. Chem. B 114, 10826–10834.
64. Lim, H.-D., Lee, B., Zheng, Y., Hong, J., Kim, J.,
Gwon, H., Ko, Y., Lee, M., Cho, K., and Kang, K.
(2016). Rational design of redox mediators for
advanced Li–O2 batteries. Nat. Energy 1,
16066.
52. Bard, A.J., and Faulkner, L.R. (2002).
Fundamentals and Applications (Wiley).
58. Frisch, M.J., Trucks, G., Schlegel, H.B.,
Scuseria, G., Robb, M., Cheeseman, J.,
Scalmani, G., Barone, V., Mennucci, B., and
Petersson, G. (2009). Gaussian 09, revision A. 1,
27 (Gaussian Inc.), p. 34.
53. Compton, R.G., and Banks, C.E. (2011).
Understanding Voltammetry (World Scientific).
65. Hong, J., Lee, M., Lee, B., Seo, D.-H., Park, C.B.,
and Kang, K. (2014). Biologically inspired
pteridine redox centres for rechargeable
batteries. Nat. Commun. 5, 5335.
54. Chiykowski, V.A., Lam, B., Du, C., and
Berlinguette, C.P. (2017). Comparative analysis
of triarylamine and phenothiazine sensitizer
59. Lee, C., Yang, W., and Parr, R.G. (1988).
Development of the Colle-Salvetti correlation-
energy formula into a functional of the electron
Chem 5, 1–15, October 10, 2019
15