In situelectrosynthesis of anthraquinone electrolytes in aqueous flow batteries

Article abstract of DOI:10.1039/d0gc02236e

We demonstrate the electrochemical oxidation of an anthracene derivative to a redox-active anthraquinone at room temperature in a flow cell without the use of hazardous oxidants or noble metal catalysts. The anthraquinone, generatedin situ, was used as the active species in a flow battery electrolyte without further modification or purification. This potentially scalable, safe, green, and economical electrosynthetic method is also applied to another anthracene-based derivative and may be extended to other redox-active aromatics.

Full text of DOI:10.1039/d0gc02236e

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Y. Jing, M. Wu, A.
A. Wong, E. Fell, S. Jin, D. A. Pollack, E. Kerr, R. G. Gordon and M. J. Aziz, Green Chem., 2020, DOI:
10.1039/D0GC02236E.
Volume 20
This is an Accepted Manuscript, which has been through the
Number
May 2018
Pages 1919-2160
9
7
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Page 1 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
1
In situ Electrosynthesis of Anthraquinone Electrolytes in Aqueous Flow Batteries
2
3
Yan Jing,1§ Min Wu,2§ Andrew A. Wong,2 Eric M. Fell,2 Shijian Jin,2 Daniel A. Pollack,3 Emily
F. Kerr,1 Roy G. Gordon,*,1,2 Michael J. Aziz*,2
4
5
1 Department of Chemistry and Chemical Biology, Harvard University, Cambridge,
Massachusetts 02138, United States
6
7
2 John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge,
Massachusetts 02138, United States
8
9
3 Department of Physics, Harvard University, Cambridge, Massachusetts 02138, United States
§ These authors contributed equally to this work.
10
11
* Correspondence: gordon@chemistry.harvard.edu; maziz@harvard.edu
Abstract
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
We demonstrate the electrochemical oxidation of an anthracene derivative to a redox-active
anthraquinone at room temperature in a flow cell without the use of hazardous oxidants or noble
metal catalysts. The anthraquinone, generated in situ, was used as the active species in a flow
battery electrolyte without further modification or purification. This potentially scalable, safe,
green, and economical electrosynthetic method is also applied to another anthracene-based
derivative and may be extended to other redox-active aromatics.
Introduction
Aqueous redox flow batteries (ARFBs) represent a class of devices for storing electrical energy
that are especially well suited for large-scale stationary deployment.1, 2 Vanadium redox flow
batteries, the most developed ARFB technology, have been limited by the high and fluctuating
price of vanadium.3
Anthraquinone-based aqueous redox flow batteries are considered as one class of the most
promising alternatives to vanadium redox flow batteries because they can be composed of earth-
abundant elements such as C, H, O, and N while providing comparable electrochemical
performance.4-9 However, reducing the production cost of anthraquinone-based electrolytes and
1

Green Chemistry
Page 2 of 18
View Article Online
DOI: 10.1039/D0GC02236E
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
improving their chemical stability are two major challenges preventing them from being cost-
competitive.9-14 Many factors can influence the synthesis cost of an organic molecule, including
the number, duration, complexity, and yields of the reaction steps, the reaction conditions (time,
temperature, and pressure), solvent and precursor costs, the cost of waste disposal, and economies
of scale. Likewise, a host of factors contributes to the stability, and by extension the long-term
viability, of redox-active organics including the chemical structure, solvent conditions, applied
potentials, and state of charge. Only through careful consideration of all of these factors can
commercial-scale organic ARFBs be viable storage solutions. Therefore, not only is the
development of a stable anthraquinone important, but the design of a potentially economical,
scalable, and green synthetic route toward targeted molecules is equally significant.11, 15
Electrochemically-mediated synthesis (electrosynthesis) enables the replacement of hazardous
oxidizing and reducing agents by electric current, or “clean” electrons, through an electrode and
has attracted considerable attention for both laboratory and industrial applications in multiple
fields of research.16-21 Compared to traditional thermochemical synthesis, electrosynthesis can be
significantly more environmentally benign due to reduced waste production and alternative
chemicals consumed.22, 23 However, the necessity of using specific solvents combined with
supporting electrolytes, along with their subsequent separations, are some of the primary hurdles
limiting the feasibility of electrosynthesis compared to thermochemical processes in many cases.16
As an example, anthraquinone is typically produced from anthracene, an inexpensive and
abundant component of coal tar and petroleum.24 Typically, hazardous oxidants such as
cerium(IV), chromium(VI), and vanadium(V) compounds dissolved in strong acids, sometimes at
elevated temperatures, are used to facilitate this thermochemical conversion.25 To minimize the
use of hazardous materials, often these consumed oxidants are electrochemically regenerated and
2

Page 3 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
reused for chemical oxidations,25-30 that is, a mediated or indirect electrochemical oxidation.
However, in both thermochemical conversion and mediated (indirect) electrochemical conversion,
isolating anthraquinone from these hazardous solutions can be time- and capital-intensive. Electro-
oxidations of anthracene and its derivatives at ~1 mM concentration have been performed
previously; however, the low concentrations of anthracene substrates and poor selectivity of the
reactions have prevented the method from being synthetically useful.31-36
Using a scalable flow cell setup,37 we demonstrate the capability to electrochemically oxidize
water-soluble anthracenes directly to anthraquinones in electrolytes without the use of strong
oxidants or catalysts, producing the desired negolyte (negative electrolyte) and ferrocyanide
posolyte (positive electrolyte) in situ. Compared to conventional thermochemical and
electrochemical methods, the new method is safe and potentially inexpensive because it eliminates
both the use of hazardous oxidants and the necessity of post-synthesis isolation of the products
from the supporting electrolytes. Taking advantage of a flow cell and bulk electrolysis setup, the
demonstrated electrosynthetic method is amenable to both continuous and batch processing.
Furthermore, we confirmed that the electrosynthetic method can also be extended to other
anthracene derivatives.
3,3'-(9,10-anthraquinone-diyl)bis(3-methylbutanoic acid) (DPivOHAQ) was recently reported
as an extremely stable and potentially inexpensive negolyte active species for organic ARFBs.38
However, the use of CrO3 in the synthesis can be highly toxic and explosive if produced in large
scale. Figure 1a shows the synthetic route for DPivOHAQ in three steps: 1) Through Birch
reduction, anthracene (AC) is converted to 9,10-dihydroanthracene (DHAC) at room temperature
(Figure S1). 2) After a FriedelCrafts reaction and subsequent oxidation by air in one pot, two
water-soluble groups are introduced and DHAC is re-oxidized to an AC derivative (Figure S2),
3

Green Chemistry
Page 4 of 18
View Article Online
DOI: 10.1039/D0GC02236E
73
74
75
76
77
78
79
forming 3,3'-(anthracene-diyl)bis(3-methylbutanoic acid) (DPivOHAC). The DPivOHAC
powder was then dissolved in water by adding KOH to deprotonate the carboxylic acid groups. 3)
Lastly, DPivOHAQ negolyte active species is produced by electrochemical oxidation in an
aqueous electrolyte without the need for further purification. Figure 1b illustrates how
DPivOHAQ and ferrocyanide active species can be produced in situ in the flow cell’s
electrosynthesis mode. These materials can directly serve as the active species in the negolyte and
the posolyte, respectively, of a flow battery in the same cell as illustrated in Figure 1c.
80
81
82
83
84
85
86
87
Figure 1. Preparation of DPivOHAQ and the corresponding flow battery. (a) The
DPivOHAQ synthetic route and conditions starting from anthracene. (b) The setup for
electrosynthesis of DPivOHAQ and ferrocyanide. (c) The flow battery setup with DPivOHAQ
negolyte (generated in situ) and ferrocyanide posolyte (generated in situ). DPivOHAC: 3,3'-
(anthracene-diyl)bis(3-methylbutanoic acid); DPivOHAC(COO) is deprotonated DPivOHAC.
DPivOHAQ: 3,3'-(9,10-anthraquinone-diyl)bis(3-methylbutanoic acid); DPivOHAQ(COO) is
deprotonated DPivOHAQ.
88
89
90
Figure 2a lists three different oxidation methods for DPivOHAQ synthesis. Conventionally,
anthracene derivatives can be chemically oxidized to their anthraquinone forms by oxidants such
as chromium oxide (CrO3) in strong acidic media at elevated temperature.38 To minimize the use
4

Page 5 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
91
92
93
94
95
96
97
of hazardous oxidants, the strategy of mediated electrochemical oxidation can be performed by
29
regenerating oxidants such as cerium(IV) compounds.26,
However, in both of these
thermochemical and indirect electrochemical oxidation processes, tedious and expensive isolation
of anthraquinone-based products from oxidants and acids is required. Taking advantage of the high
solubility of DPivOHAC in base, we demonstrate a synthetic route via direct electrochemical
oxidation in alkaline electrolyte with a flow cell. This method allows the complete elimination of
hazardous oxidants and costly separation processes.
98
99
100
101
102
103
Figure 2. Comparison of DPivOHAQ synthetic methods. (a) Thermochemical, mediated
(indirect) electrochemical, and direct electrochemical oxidation reactions to synthesize
DPivOHAQ. (b) Advantages of direct electrochemical oxidation in situ.
Experimental
104
105
106
107
108
109
110
Cell hardware
Glassy carbon was used as the working electrode for all three-electrode cyclic voltammetry (CV)
tests with a 5 mm diameter glassy carbon working electrode, an Ag/AgCl reference electrode
(BASi, pre-soaked in 3 M NaCl solution), and a graphite counter electrode. Both undivided cell
and divided cell were built for electrosynthesis. Flow battery experiments were constructed with
cell hardware from Fuel Cell Tech (Albuquerque, NM) assembled into a zero-gap flow cell
configuration. Pyrosealed POCO graphite flow plates with serpentine flow patterns were used for
5

Green Chemistry
Page 6 of 18
View Article Online
DOI: 10.1039/D0GC02236E
111
112
113
114
115
116
117
118
119
120
both electrodes. Each electrode comprised a 5 cm2 geometric surface area covered by AvCarb
HCBA woven carbon fiber without pretreatment, or Pt-coated Toray carbon paper without
pretreatment. The membrane is pre-soaked (1 M KOH for 24 hours) Nafion 212.
Undivided electrolytic cell setup (electrochemical oxidation vs. the HER)
Working electrode: carbon felt, where DPivOHAC(COO) was oxidized to DPivOHAQ(COO);
counter electrode: carbon rod, where water was reduced to hydrogen gas. While the electrolyte
was stirred, a constant potential (1.1 V vs. Ag/AgCl) was applied to the divided electrolytic cell
until 120% of the required coulombs were extracted from the working electrode.
Divided electrolytic cell setup (electrochemical oxidation vs. the ORR)
Anode: Commercial AvCarb HCBA (woven carbon cloth), where DPivOHAC(COO) was
121
122
123
124
125
126
oxidized to DPivOHAQ(COO); cathode: platinum coated Toray carbon paper, where humidified
air/oxygen was reduced to hydroxide. A constant voltage (1.8 V) was applied to the divided
electrolytic cell until the current decreased to 2 mA/cm2. The number of extracted electrons was
~1.2 times higher than the theoretical value.
Divided electrolytic cell setup (electrochemical oxidation vs. the reduction of ferricyanide)
Anode: AvCarb HCBA (woven carbon cloth), where DPivOHAC(COO) was oxidized to
127
128
129
130
131
132
DPivOHAQ(COO); cathode: AvCarb HCBA (woven carbon cloth), where potassium
ferricyanide was reduced to potassium ferrocyanide. A constant current density (20 mA/cm2) was
applied to the divided cell for at most 1.5 hours with a 1.2 V voltage cutoff; when either time or
voltage reached the limit, the potential was held (1.2 V vs. ferro-/ferricyanide) until the current
decreased to 2 mA/cm2. The number of extracted electrons was ~1.2 times higher than the
theoretical value.
6

Page 7 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
133
134
135
136
137
138
139
140
141
142
143
144
An aliquot (~250 µL) was transferred from the as-prepared anolyte to an Eppendorf® tube
(capacity: 1.5 mL) and acidified by a drop of concentrated HCl to obtain DPivOHAQ precipitate.
The final DPivOHAQ precipitate was re-dissolved in DMSO-d6 for 1H NMR measurement. The
yield was determined by peak integrations of spectrum. Faradaic efficiency (%) = yield (%) / 1.2.
More detailed information can be found in the Supplementary information.
Results and Discussion
In an electrolytic cell, an anodic oxidation half reaction must be accompanied by a cathodic
reduction half reaction. As shown in Table 1, we devise three different reduction half reactions to
be coupled with direct DPivOHAC electrochemical oxidation, i.e., the hydrogen evolution
reaction (HER), the oxygen reduction reaction (ORR), and the Fe(CN)63 to Fe(CN)64 reduction
reaction. The corresponding oxidation or reduction potentials for these reactions are listed in Table
1.
145
146
147
148
149
150
151
152
153
154
155
For the electrochemical oxidation of DPivOHAC to DPivOHAQ, two cell types are used, as
diagramed and described in Figures S3 and S4. A divided cell uses an ion exchange membrane to
separate the two half reactions, resembling the architecture of traditional fuel cells and ARFBs.
An undivided cell employs two electrodes suspended in electrolyte without the use of a membrane,
reflecting a bulk electrolysis cell.
Comparing these three overall reactions, the first one paired with the HER requires the highest
voltage; the second one paired with the ORR is known to have slow reaction kinetics and a high
3
4
overpotential;39 the third one paired with Fe(CN)6 to Fe(CN)6 reduction exhibits the lowest
overall reaction cell voltage, suggesting the least amount of energy will be required for
electrosynthesis. Another merit of the third reaction is the in situ generation of the desired negolyte
4
active species (DPivOHAQ) and posolyte active species Fe(CN)6 simultaneously. The
7

Green Chemistry
Page 8 of 18
View Article Online
DOI: 10.1039/D0GC02236E
156
157
158
159
160
161
162
163
164
disadvantage is that at least six equivalents of ferricyanide and hydroxide are used. Given the
similar reduction potentials of the ORR and of ferricyanide to ferrocyanide, an important direction
for future research is the concurrent reduction of oxygen and ferricyanide in order to achieve high
yields as well as lower ferricyanide usage. By using the same full cell configuration without
changing electrolyte reservoirs, carbon-based electrodes, or ion-exchange membranes, we can
immediately switch from electrosynthesis mode to flow battery mode for electrochemical energy
storage. In this configuration, neither hazardous oxidants nor purification steps are needed, nor is
waste generated. Furthermore, the reaction may proceed at room temperature with high atom
efficiency. The new synthesis is therefore potentially safe, green, economical, and scalable.
165
Table 1. Anodic, cathodic, and overall reactions for direct electrochemical oxidation.
166
167
168
169
170
171
172
173
The cyclic voltammogram (CV) of DPivOHAC at pH 14 (Figure 3a) indicates a peak oxidation
current at 1.14 V vs. SHE. This value is more positive than the standard redox potential of 0.40 V
vs. SHE for the oxygen evolution reaction (OER), and we expect that the OER will be a major side
reaction of electrosynthesis.
We then assembled a flow cell with DPivOHAC as the anolyte and K3Fe(CN)6 as the catholyte.
Galvanostatic electrolysis with a potentiostatic hold after reaching a potential limit of 1.2 V was
performed for ~4.5 hours to complete the electrosynthesis. The OER side reaction, evidenced by
8

Page 9 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
174
175
176
177
the observation of bubbles generated in the anolyte, precludes a faradaic efficiency of 100%. Thus,
the number of electrons extracted from the anolyte was ~1.2 times higher than the theoretical
number for complete conversion. A plateau appears at ~0.8 V against K3Fe(CN)6 (0.44 V vs. SHE)
in the voltage profile (Figure 3b).
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
Figure 3. Electrosynthesis and characterization of DPivOHAQ. (a) The cyclic voltammogram
(CV) of 0.1 M DPivOHAC in 1.0 M KCl + 1.0 M KOH aqueous solution. Scan rate: 0.1 V/s. (b)
The electrochemical oxidation was conducted by using a constant current (20 mA/cm2) with a
subsequent potential hold (1.2 V) until the current density decreased to 2 mA/cm2. (c) CV of
3
10 mM electrosynthesized DPivOHAQ (against Fe(CN)6 ) without purification and 10 mM
chemically synthesized DPivOHAQ with purification in 1 M KOH aqueous solutions,
1
respectively. Scan rate: 0.1 V/s. (d) H NMR spectra of (bottom to top): chemically synthesized
DPivOHAC (black); chemically synthesized DPivOHAQ (red); electrosynthesized DPivOHAQ
in an undivided cell (purple), 17.3% of DPivOHAC remained unreacted according to the
3
integration, yield: 82.7%; electro-synthesized DPivOHAQ in a divided cell against Fe(CN)6
(blue), 7.0% of DPivOHAC remained unreacted according to the integration, yield: 93.0%;
electrosynthesized DPivOHAQ in a divided cell against O2 (green), 0 % of DPivOHAC remained
unreacted according to the integration, yield: 100%. The deuterated solvent is DMSO-d6, and the
solvent peaks (DMSO and H2O) were removed to better display the peaks of interest. The
9

Green Chemistry
Page 10 of 18
View Article Online
DOI: 10.1039/D0GC02236E
193
194
195
electrosynthetic details are described under the headings Electrosynthesis I, II, and III in the
Supporting Information.
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
We compared the CV of DPivOHAQ produced by electrosynthesis against the reduction of
3
Fe(CN)6 to that of the chemically synthesized product at the same concentration to verify that
the reaction products are the same regardless of the synthetic procedure employed (Figure 3c). The
two CV curves show identical redox peaks and similar peak currents, indicating a high-yield
electrosynthesis process. 1H nuclear magnetic resonance (NMR) spectroscopy was used to further
examine the structure of electrosynthesized DPivOHAQ when using either a divided or undivided
cell (Figure S3) and to compare the spectra with those of the starting material, DPivOHAC, and
1
the chemically synthesized DPivOHAQ. The top three spectra in Figure 3d are the H NMR
spectra from electrosynthesized DPivOHAQ, in which the dominating peaks have the same
chemical shifts as those in the spectrum of chemically synthesized DPivOHAQ, further suggesting
the desired product was achieved.
Slightly different yields of DPivOHAQ were obtained when paired with the HER in an
undivided cell or with Fe(CN)63 reduction or the ORR in a divided cell (Figure S4). The 82.7%
yield when paired with the HER in an undivided cell could be explained by a molecular shuttling
effect; i.e., the electrosynthesized DPivOHAQ can first migrate to the cathode where it is reduced,
then diffuse back to the anode for re-oxidation. As a result, double counting of electrons can occur.
3
When paired with the Fe(CN)6 reduction half reaction, a yield of 93.0% was obtained. The
incomplete yield is likely due to the consumption and therefore decreased concentration of both
DPivOHAC and OH as the electrosynthesis continues, making further oxidation increasingly
difficult.
10

Page 11 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
216
217
The use of the ORR half reaction achieved almost 100.0% yield. This exceptional yield may be
attributed to the as-formed OH ions on the cathode (ORR) side crossing over to the anolyte and
218
219
220
compensating for any loss of OH ions on the anode side. Overall yields in excess of 80.0% for all
three conditions exceed many conventional reactions and are acceptable for direct flow battery use
without purification or separation.
Cycle number
b
0
300 600 900 1200 1500 1800 2100
a
78.5
78
1.2
1.1
1.0
0.9
0.8
0.7
0.6
100
95
90
85
80
75
Discharge
Coulombic efficiency
77.5
77
Theoretical capacity 96.5 C
76.5
76
Fitted fade rate: 0.014%/day
0
5
10
15
20
25
30
35
0
10 20 30 40 50 60 70 80 90 100
Capacity (C)
Time (days)
c
d
1.6
1.4
1.2
1.0
0.8
0.6
0.4
1.4
1.2
1.0
0.8
0.6
0.4
200
0
20%
Theoretical capacity 576 C
SOC
40%
60%
80%
~100%
-200
-400
-600
-400
-200
0
200
400
0
100
200
300
400
500
Current density (mA/cm2)
Capacity (C)
221
222
223
224
225
226
227
228
229
230
Figure 4. Full cell performance evaluation from Electrosynthesis III and IV. (a) A
representative chargedischarge profile with 0.1 M DPivOHAQ. Negolyte: 5 mL of 0.1 M
DPivOHAQ pH = ~13.5. Posolyte: 100 mL of 0.1 M potassium ferro-/ferricyanide solution [~0.06
M K4Fe(CN)6 and ~0.04 M K3Fe(CN)6] pH = ~13.6. (b) Discharge capacity (C) and coulombic
efficiency (%) vs. cycle number and time (days). Negolyte: 4.5 mL of 0.1 M DPivOHAQ. Posolyte:
100 mL of 0.1 M ferro-/ferricyanide solution [~0.06 M K4Fe(CN)6 and ~0.04 M K3Fe(CN)6].
Current density: 30 mA/cm2 with potential hold (cutoffs: 0.6 V, 1.2 V) until current decreased to
2 mA/cm2. (c) A representative chargedischarge profile with 0.5 M DPivOHAQ. Negolyte: 6
mL of 0.5 M DPivOHAQ. Posolyte: 100 mL of 0.5 M potassium ferro-/ferricyanide solution [~0.3
11

Green Chemistry
Page 12 of 18
View Article Online
DOI: 10.1039/D0GC02236E
231
232
233
234
235
236
M K4Fe(CN)6 and ~0.2 M K3Fe(CN)6]. Current density: 100 mA/cm2 with potential hold (cutoff:
0.4 V, 1.4 V) until current decreased to 2 mA/cm2. (d) Polarization curves of the 0.5 M
DPivOHAQ at the SOC of 20%, 40%, 60%, 80%, and ~100% respectively. Descriptions of
Electrosynthesis III and IV can be found in the Supporting Information.
To demonstrate the feasibility of switching from the electrosynthesis mode (when paired with
Fe(CN)63 reduction) to flow battery mode, we began chargedischarge cycling immediately upon
completion of the electrosynthesis, without performing any purification. Because other research
has reported that quinones and related compounds can decompose in the presence of light,40-42 we
wrapped the electrolyte reservoirs with aluminum foil to avoid light-induced decomposition during
cell cycling (Figures S13S15). Figure 4a shows the chargedischarge profile of a single cycle
with an open circuit voltage of ~1.0 V and a capacity of 84.0 coulombs. Given the 93.0% yield
found from the 1H NMR, the capacity utilization is 93.6%. Long-term cycling was then performed
to determine a temporal capacity fade rate of the full cell. Figure 4b demonstrates the discharge
capacity and coulombic efficiency over 33.2 days and 2271 cycles with a fitted fade rate of
0.014%/day and an average coulombic efficiency of 99.53%. This is consistent with the fade rate
of chemically synthesized DPivOHAQ.38 The extremely low capacity fade rate is attributed to the
chemical stability of the molecular structure. The CC covalent bond between the anthraquinone
core and the functionalizing chains is more robust in strong base and at elevated temperature than
the CO bond demonstrated in previous work.4, 5, 38 Furthermore, the two branched methyl groups
on the carbon connected to the anthraquinone (AQ) core may increase the stability of the
solubilizing chain even when exposed to harsh conditions.15
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
To examine the feasibility of this method for potential industry use, we further conducted
electrosynthesis with a higher concentration (0.5 M) of DPivOHAC at a higher current density
(100 mA/cm2) (See Figure S5). Figure 4c shows that 0.5 M electrosynthesized negolyte can deliver
72.9% of the theoretical capacity. We attribute the discrepancy between the delivered capacity and
12

Page 13 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
257
258
259
260
261
262
263
264
265
266
the theoretical capacity primarily to incomplete conversion (Figure S6). The capacity utilization is
81.9% if we consider that there is 11.0% unreacted DPivOHAC(COO) in the negolyte.
Additionally, the mass transport of active species at 0.5 M concentration may be another issue
limiting the full capacity utilization. The corresponding polarization curve at different states of
charge (SOC) is shown in Figure 4d. The peak power density exceeds 0.2 W/cm2 when at ~100 %
of SOC.
Given the total transfer of six electrons during the electrosynthesis of DPivOHAC to
DPivOHAQ, the high yields achieved in this work might be surprising. We hypothesize a three-
step successive two-electron transfer mechanism34, 35: first, when a potential is applied, anthracene
(AC) may react with three OH ions and donate two electrons to produce two water molecules and
267
268
269
270
271
272
273
274
275
276
277
the anthrone anion (A); second, A may further react with another three OH ions and donate
another two electrons to generate two water molecules and the deprotonated anthrahydroquinone
dianion (AQ2); third, AQ2 may further release two electrons to afford the anthraquinone species
(AQ). Complete electrochemical conversion in the third step has been well-documented at
negative potentials vs. ferro-/ferricyanide1, 2, 43 and should therefore be rapid at positive potentials
vs. ferro-/ferricyanide. The reverse reaction of the second step has recently been identified as a
side reaction in ARFBs, and the forward reaction is chemically feasible when exposed to O2 or
air.14, 38 Given the high voltage applied to the cell, it is thus plausible that the forward reactions
(AC to A to AQ2/AQ) can electrochemically proceed completely and swiftly.
Our group has also previously proposed a side reaction pathway for anthraquinones,14, 44 where
the anthrone anion (A) can be oxidatively dimerized to dianthrone (DA) chemically and/or
1
278
279
electrochemically. According to H NMR spectra (Figure 3d) and liquid chromatographymass
spectrometry (LCMS) results (Figure S7), neither DA nor Kolbe electrolysis-related byproducts45
13

Green Chemistry
Page 14 of 18
View Article Online
DOI: 10.1039/D0GC02236E
280
281
282
283
were detected (Scheme S1), suggesting that AC/AQ-related side reactions can be negligible when
a sufficient OH concentration is present to prevent dianthrone formation and a sufficiently low
voltage cutoff is chosen to prevent Kolbe electrolysis dimer formation. The major competing side
reaction is the OER, which, along with the reactions of AC to A to AQ2, will consume OH and
284
285
286
287
288
289
290
291
292
293
294
295
296
may lead to the formation of DA as a result of insufficient OH ions in the DPivOHAC solution
(see Electrosynthesis V in the SI). Interestingly, the dianthrones (Scheme S2), detected by
LCMS (Figure S11), are surprisingly redox-active when a broad voltage window is applied
(Figures S8 and S9 and Scheme S2). On the one hand, the OER can reduce faradaic efficiency; on
the other hand, the generated oxygen can serve as a mediator and chemically oxidize intermediates
(i.e., A, AQ2) to the final AQ form, i.e., mediated (indirect) electrochemical oxidation. Because
the entire process involves not only electrochemical oxidations, but also chemical oxidations, it is
more appropriate to call it an electrochemical–chemical oxidation process.46
In the proposed mechanism, the anthrone derivative is an intermediate in the electrochemical
oxidation. Anthrone formation has been identified as the major side reaction causing capacity fade
in previous work;14, 38 therefore, it is plausible that lost capacity of anthraquinone flow battery
systems may be recovered and anthraquinone lifetime extended by electrochemically oxidizing
anthrone to redox-active anthraquinone derivatives.
14

Page 15 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
297
298
299
300
301
302
303
Scheme 1. Proposed electrochemical oxidation mechanism. Three-step successive two-electron
transfer process from AC to A, A to AQ2, and AQ2 to AQ. The generated oxygen from the
OER side reaction may incur chemical oxidation processes including A to AQ2, AQ2 to AQ,
and oxidative dimerization (A to DA).
To demonstrate that the electrochemical oxidation can be applied to other anthracene
derivatives, we performed electrochemical oxidation of 4,4'-(9,10-dihydroanthracene-
diyl)dibutanoic acid (DBDHAC), where the molecular core is 9,10-dihydroanthracene.38 The 1H
NMR results indicate that DBDHAC can, like DPivOHAC, be electrochemically oxidized to the
final anthraquinone (Figure S12), DBAQ (4,4'-(9,10-anthraquinone-diyl)dibutanoic acid), which
has also been shown to be extremely stable.38
304
305
306
307
308
309
310
311
312
313
314
315
The shared precursor of DPivOHAQ and DBAQ, anthracene, is abundant in crude petroleum
and coal tar, and can be synthesized from benzene and benzyl alcohol (Scheme S3).47 The
precursor of DPivOHAQ, 3,3’-dimethyl acrylic acid, can be industrially produced from malonic
acid, a food acid; the precursor of DBAQ, succinic anhydride, can be industrially hydrogenated
from maleic anhydride and used as an important intermediate on an industrial scale. Thus, both
DPivOHAQ and DBAQ can be readily synthesized from commodity chemicals. Although the
synthetic cost of DPivOHAQ or DBAQ should be somewhat higher than that of 2,6-
15

Green Chemistry
Page 16 of 18
View Article Online
DOI: 10.1039/D0GC02236E
316
317
318
319
320
321
dihydroxyanthraquinone (DHAQ) due to more steps and more chemicals involved, the capital cost
of AORFBs that utilize finite-lifetime electrolytes can be viewed as including the total active cost,
which is the sum of the initial cost of redox-active materials and the present value of the future
costs of periodic electrolyte replacement.13 This can lead to an initial cost—lifetime trade-off in
the choice of electrolytes. Over an extended operational lifetime, the total active cost of
DPivOHAQ or DBAQ may be less than that of DHAQ due to their much longer lifetimes.14
322
323
Conclusion
324
325
326
327
328
329
330
331
This work demonstrates a potentially scalable, safe, green, and economical in situ
electrosynthetic method for anthraquinone electrolytes in a flow cell without the use of hazardous
oxidants or precious metal catalysts. The as-generated electrolytes, which are extremely stable,
can be immediately used in a redox flow battery without separation or purification. Other low-cost
compounds may also be amenable to this approach, providing a pathway to lower the cost of
electrochemical grid storage systems, thereby accelerating the development of a renewable energy
economy. The technique extends the opportunities for direct aqueous electrosynthesis to replace
thermochemical synthesis of value-added organics.
332
333
334
Supplementary Information
Supplementary Information can be found with this article online at
Acknowledgments
335
336
337
338
This research was supported by U.S. DOE award DE-AC05-76RL01830 through PNNL
subcontract 428977, Innovation Fund Denmark via the Grand Solutions project "ORBATS" file
no. 7046-00018B, and NSF grant CBET-1914543. D.A.P. acknowledges funding support from the
NSF Graduate Research Fellowship Program, no. DGE1144152 and DGE1745303.
339
Declaration of Interests
340
341
Harvard University has filed a patent application on the materials and the electrosynthetic methods
described in this paper.
342
16

Page 17 of 18
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02236E
343
REFERENCES
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
1.
2.
3.
4.
5.
6.
7.
8.
B. Huskinson, M. P. Marshak, C. Suh, S. Er, M. R. Gerhardt, C. J. Galvin, X. Chen, A. Aspuru-
Guzik, R. G. Gordon and M. J. Aziz, Nature, 2014, 505, 195198.
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 and M. P. Marshak, Science, 2015, 349, 15291532.
M. Moore, R. Counce, J. Watson and T. Zawodzinski, Journal of Advanced Chemical
Engineering, 2015, 5, doi:10.4172/2090-4568.1000140.
D. G. Kwabi, K. Lin, Y. Ji, E. F. Kerr, M.-A. Goulet, D. De Porcellinis, D. P. Tabor, D. A.
Pollack, A. Aspuru-Guzik, R. G. Gordon and M. J. Aziz, Joule, 2018, 2, 18941906.
Y. Ji, M.-A. Goulet, D. A. Pollack, D. G. Kwabi, S. Jin, D. De Porcellinis, E. F. Kerr, R. G.
Gordon and M. J. Aziz, Adv. Energy Mater., 2019, 9, 1900039.
A. Hollas, X. Wei, V. Murugesan, Z. Nie, B. Li, D. Reed, J. Liu, V. Sprenkle and W. Wang,
Nature Energy, 2018, 3, 508514.
C. Wang, X. Li, B. Yu, Y. Wang, Z. Yang, H. Wang, H. Lin, J. Ma, G. Li and Z. Jin, ACS Energy
Letters, 2020, 5, 411417.
J. D. Hofmann, F. L. Pfanschilling, N. Krawczyk, P. Geigle, L. Hong, S. Schmalisch, H. A.
Wegner, D. Mollenhauer, J. Janek and D. Schröder, Chemistry of Materials, 2018, 30, 762774.
D. G. Kwabi, Y. Ji and M. J. Aziz, Chem. Rev., 2020, 120, doi.org/10.1021/acs.chemrev.9b00599.
Z. Yang, L. Tong, D. P. Tabor, E. S. Beh, M.-A. Goulet, D. De Porcellinis, A. Aspuru-Guzik, R.
G. Gordon and M. J. Aziz, Adv. Energy Mater., 2018, 8, 1702056.
V. Dieterich, J. D. Milshtein, J. L. Barton, T. J. Carney, R. M. Darling, F. R. Brushett,
Translational Materials Research, 2018, 5, 034001.
S. Jin, E. M. Fell, L. Vina-Lopez, Y. Jing, P. W. Michalak, R. G. Gordon and M. J. Aziz, Adv.
Energy Mater., 2020, 10, doi.org/10.1002/aenm.202000100.
F. R. Brushett, M. J. Aziz and K. E. Rodby, ACS Energy Letters, 2020, 5, 879884.
M.-A. Goulet, L. Tong, D. A. Pollack, D. P. Tabor, S. A. Odom, A. Aspuru-Guzik, E. E. Kwan,
R. G. Gordon and M. J. Aziz, J. Am. Chem. Soc., 2019, 141, 80148019.
P. Anastas, N. Eghbali, Chem. Soc. Rev., 2010, 39, 301312.
D. S. P. Cardoso, B. Šljukić, D. M. F. Santos, C. A. C. Sequeira, Organic Process Research &
Development, 2017, 21, 12131226.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
M. Yan, Y. Kawamata, P. S. Baran, Chem. Rev., 2017, 117, 1323013319.
B. K. Peters, K. X. Rodriguez, S. H. Reisberg, S. B. Beil, D. P. Hickey, Y. Kawamata, M.
Collins, J. Starr, L. Chen, S. Udyavara, K. Klunder, T. J. Gorey, S. L. Anderson, M. Neurock, S.
D. Minteer and P. S. Baran, Science, 2019, 363, 838845.
E. J. Horn, B. R. Rosen, Y. Chen, J. Tang, K. Chen, M. D. Eastgate, P. S. Baran, Nature, 2016,
533, 7781.
19.
20.
21.
22.
23.
24.
A. Badalyan, S. S. Stahl, Nature, 2016, 535, 406410.
G. G. Botte, The Electrochemical Society Interface, 2014, 23, 4955.
P. M. Bersier, L. Carlsson and J. Bersier, Topics in Current Chemistry, 1994, 170, 116136.
C. Xia, Y. Xia, P. Zhu, L. Fan and H. Wang, Science, 2019, 366, 226231.
M. Granda, C. Blanco, P. Alvarez, J. W. Patrick and R. Menendez, Chem. Rev., 2014, 114, 1608
1636.
R. S. Tipson, in National Bureau of Standards Monograph 87, 1965, 149.
R. P. Kreh, R. M. Spotnitz and J. T. Lundquist, J. Org. Chem., 1989, 54, 15311535.
E. Oppermann, US Pat., US823,435A, 1906.
25.
26.
27.
28.
E. Steckhan, in Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH &
Co. KGaA, Weinheim, 2011, 12, DOI: 10.1002/14356007.o09_o04.
17

Green Chemistry
Page 18 of 18
View Article Online
DOI: 10.1039/D0GC02236E
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
29.
R. M. Spotnitz, R. P. Kreh, J. T. Lundquist and P. J. Press, Journal of Applied Electrochemistry,
1990, 20, 209215.
30.
31.
32.
33.
34.
R. P. Kreh, R. M. Spotnitz and J. T. Lundquist, Tetrahedron Letters, 1987, 28, 10671068.
E. J. Majeski, J. D. Stuart and W. E. Ohnesorge, J. Am. Chem. Soc.,1968, 90, 633636.
L. R. Faulkner, A. J. Bard, J. Am. Chem. Soc., 1968, 90, 62846290.
C. Amatore and A. R. Brown, J. Am. Chem. Soc., 1996, 118, 1482486.
O. Tovide, N. Jahed, C. E. Sunday, K. Pokpas, R. F. Ajayi, H. R. Makelane, K. M. Molapo, S. V.
John, P. G. Baker and E. I. Iwuoha, Sensors and Actuators B: Chemical, 2014, 205, 184192.
C. A. Paddon, C. E. Banks, I. G. Davies and R. G. Compton, Ultrason. Sonochem., 2006, 13,
126132.
V. D. Parker, Acta Chemica Scandinavica, 1970, 24, 27572767.
T. Noel, Y. Cao and G. Laudadio, Acc. Chem. Res., 2019, 52, 28582869.
M. Wu, Y. Jing, A. A. Wong, E. M. Fell, S. Jin, Z. Tang, R. G. Gordon and M. J. Aziz, Chem,
2020, 6, 14321442.
R. W. Zurilla, R. K. Sen and E. Yeager, J. Electrochem. Soc., 1978, 125, 11031109.
G. Maier, L. H. Franz, H.-G. Hartan, K. Lanz and H. P. Reisenauer, Chemische Berichte, 1985,
118, 31963204.
S. A. Carlson and D. M. Hercules, Analytical Chemistry, 1973, 45, 17941799.
B. E. Hulme, E. J. Land and G. O. Phillips, J. Chem. Soc. Faraday Trans. 1, 1972, 68, 1992
2002.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
M. Quan, D. Sanchez, M. F. Wasylkiw and D. K. Smith, J. Am. Chem. Soc., 2007, 129, 12847
12856.
S. Jin, Y. Jing, D. G. Kwabi, Y. Ji, L. Tong, D. De Porcellinis, M. A. Goulet, D. A. Pollack, R. G.
Gordon and M. J. Aziz, ACS Energy Letters, 2019, 4, 13421348.
H.-J. Schäfer, Topics in Current Chemistry, 1990, 152, 91151.
C. Costentin and J.-M. Savéant, Proc. Natl. Acad. Sci. U S A, 2019, 116, 1114711152.
H. E. Ungnade and E. W. Crandall, J. Am. Chem. Soc., 1949, 71, 30093010.
45.
46.
47.
417
18